- Kernel - mailweb.openeuler.org

[PATCH openEuler-5.10 000/119] Backport 5.10.83 LTS
by Zheng Zengkai 14 Jan '22

14 Jan '22

Backport 5.10.83 LTS patches from upstream. Adrian Hunter (1): mmc: sdhci: Fix ADMA for PAGE_SIZE >= 64KiB Albert Wang (1): usb: dwc3: gadget: Fix null pointer exception Alex Deucher (1): drm/amdgpu/gfx9: switch to golden tsc registers for renoir+ Alexander Aring (1): net: ieee802154: handle iftypes as u32 Alexander Mikhalitsyn (1): shm: extend forced shm destroy to support objects from several IPC nses Amit Cohen (1): mlxsw: spectrum: Protect driver from buggy firmware Arjun Roy (1): tcp: correctly handle increased zerocopy args struct size Christophe Leroy (1): powerpc/32: Fix hardlockup on vmap stack overflow Dan Carpenter (4): usb: chipidea: ci_hdrc_imx: fix potential error pointer dereference in probe staging: rtl8192e: Fix use after free in _rtl92e_pci_disconnect() drm/nouveau/acr: fix a couple NULL vs IS_ERR() checks drm/vc4: fix error code in vc4_create_object() Daniele Palmas (1): USB: serial: option: add Telit LE910S1 0x9200 composition Danielle Ratson (1): mlxsw: Verify the accessed index doesn't exceed the array length David Hildenbrand (2): proc/vmcore: fix clearing user buffer by properly using clear_user() s390/mm: validate VMA in PGSTE manipulation functions Davide Caratti (1): net/sched: sch_ets: don't peek at classes beyond 'nbands' Diana Wang (1): nfp: checking parameter process for rx-usecs/tx-usecs is invalid Dylan Hung (1): mdio: aspeed: Fix "Link is Down" issue Eric Dumazet (3): mptcp: fix delack timer ipv6: fix typos in __ip6_finish_output() tcp_cubic: fix spurious Hystart ACK train detections for not-cwnd-limited flows Florent Fourcot (2): netfilter: ctnetlink: fix filtering with CTA_TUPLE_REPLY netfilter: ctnetlink: do not erase error code with EINVAL Florian Fainelli (3): ARM: dts: BCM5301X: Fix I2C controller interrupt ARM: dts: BCM5301X: Add interrupt properties to GPIO node ARM: dts: bcm2711: Fix PCIe interrupts Guo DaXing (1): net/smc: Fix loop in smc_listen Hans Verkuil (1): media: cec: copy sequence field for the reply Heiner Kallweit (1): lan743x: fix deadlock in lan743x_phy_link_status_change() Helge Deller (1): Revert "parisc: Fix backtrace to always include init funtion names" Holger Assmann (1): net: stmmac: retain PTP clock time during SIOCSHWTSTAMP ioctls Huang Jianan (1): erofs: fix deadlock when shrink erofs slab Huang Pei (2): MIPS: loongson64: fix FTLB configuration MIPS: use 3-level pgtable for 64KB page size on MIPS_VA_BITS_48 Jakub Kicinski (2): tls: splice_read: fix record type check tls: fix replacing proto_ops Jason Gerecke (1): HID: wacom: Use "Confidence" flag to prevent reporting invalid contacts Jeff Layton (1): ceph: properly handle statfs on multifs setups Jesse Brandeburg (1): igb: fix netpoll exit with traffic Jiri Olsa (1): tracing/uprobe: Fix uprobe_perf_open probes iteration Joakim Zhang (2): net: stmmac: fix system hang caused by eee_ctrl_timer during suspend/resume net: stmmac: platform: fix build warning when with !CONFIG_PM_SLEEP Joerg Roedel (1): iommu/amd: Clarify AMD IOMMUv2 initialization messages Juergen Gross (9): xen: sync include/xen/interface/io/ring.h with Xen's newest version xen/blkfront: read response from backend only once xen/blkfront: don't take local copy of a request from the ring page xen/blkfront: don't trust the backend response data blindly xen/netfront: read response from backend only once xen/netfront: don't read data from request on the ring page xen/netfront: disentangle tx_skb_freelist xen/netfront: don't trust the backend response data blindly tty: hvc: replace BUG_ON() with negative return value Karsten Graul (1): net/smc: Fix NULL pointer dereferencing in smc_vlan_by_tcpsk() Kumar Thangavel (1): net/ncsi : Add payload to be 32-bit aligned to fix dropped packets Maciej Fijalkowski (1): ice: fix vsi->txq_map sizing Marek Behún (2): PCI: aardvark: Deduplicate code in advk_pcie_rd_conf() net: marvell: mvpp2: increase MTU limit when XDP enabled Mark Rutland (1): sched/scs: Reset task stack state in bringup_cpu() Marta Plantykow (1): ice: avoid bpf_prog refcount underflow Mathias Nyman (2): usb: hub: Fix usb enumeration issue due to address0 race usb: hub: Fix locking issues with address0_mutex Maurizio Lombardi (1): nvmet: use IOCB_NOWAIT only if the filesystem supports it Michael Kelley (1): firmware: smccc: Fix check for ARCH_SOC_ID not implemented Mike Christie (1): scsi: core: sysfs: Fix setting device state to SDEV_RUNNING Miklos Szeredi (1): fuse: release pipe buf after last use Minas Harutyunyan (1): usb: dwc2: gadget: Fix ISOC flow for elapsed frames Mingjie Zhang (1): USB: serial: option: add Fibocom FM101-GL variants Nathan Chancellor (1): usb: dwc2: hcd_queue: Fix use of floating point literal Nicholas Kazlauskas (1): drm/amd/display: Set plane update flags for all planes in reset Nicholas Piggin (1): KVM: PPC: Book3S HV: Prevent POWER7/8 TLB flush flushing SLB Nikolay Aleksandrov (3): net: nexthop: fix null pointer dereference when IPv6 is not enabled net: ipv6: add fib6_nh_release_dsts stub net: nexthop: release IPv6 per-cpu dsts when replacing a nexthop group Nitesh B Venkatesh (1): iavf: Prevent changing static ITR values if adaptive moderation is on Noralf Trønnes (1): staging/fbtft: Fix backlight Ondrej Jirman (1): usb: typec: fusb302: Fix masking of comparator and bc_lvl interrupts Pali Rohár (4): PCI: aardvark: Update comment about disabling link training PCI: aardvark: Implement re-issuing config requests on CRS response PCI: aardvark: Simplify initialization of rootcap on virtual bridge PCI: aardvark: Fix link training Peng Fan (1): firmware: arm_scmi: pm: Propagate return value to caller Pierre-Louis Bossart (1): ALSA: intel-dsp-config: add quirk for JSL devices based on ES8336 codec Russell King (Oracle) (2): net: phylink: Force link down and retrigger resolve on interface change net: phylink: Force retrigger in case of latched link-fail indicator Sakari Ailus (1): ACPI: Get acpi_device's parent from the parent field Shin'ichiro Kawasaki (1): scsi: scsi_debug: Zero clear zones at reset write pointer Sreekanth Reddy (1): scsi: mpt3sas: Fix kernel panic during drive powercycle test Srinivas Kandagatla (3): ASoC: qdsp6: q6routing: Conditionally reset FrontEnd Mixer ASoC: qdsp6: q6asm: fix q6asm_dai_prepare error handling ASoC: codecs: wcd934x: return error code correctly from hw_params Stefano Garzarella (1): vhost/vsock: fix incorrect used length reported to the guest Stefano Stabellini (2): xen: don't continue xenstore initialization in case of errors xen: detect uninitialized xenbus in xenbus_init Steve French (1): smb3: do not error on fsync when readonly Steven Rostedt (VMware) (2): tracing: Fix pid filtering when triggers are attached tracing: Check pid filtering when creating events Takashi Iwai (5): ALSA: ctxfi: Fix out-of-range access ALSA: hda/realtek: Fix LED on HP ProBook 435 G7 staging: greybus: Add missing rwsem around snd_ctl_remove() calls ASoC: topology: Add missing rwsem around snd_ctl_remove() calls ARM: socfpga: Fix crash with CONFIG_FORTIRY_SOURCE Thinh Nguyen (2): usb: dwc3: gadget: Ignore NoStream after End Transfer usb: dwc3: gadget: Check for L1/L2/U3 for Start Transfer Thomas Zeitlhofer (1): PM: hibernate: use correct mode for swsusp_close() Tim Harvey (1): mmc: sdhci-esdhc-imx: disable CMDQ support Todd Kjos (1): binder: fix test regression due to sender_euid change Tony Lu (2): net/smc: Ensure the active closing peer first closes clcsock net/smc: Don't call clcsock shutdown twice when smc shutdown Trond Myklebust (1): NFSv42: Don't fail clone() unless the OP_CLONE operation failed Varun Prakash (1): nvmet-tcp: fix incomplete data digest send Vladimir Oltean (2): net: mscc: ocelot: don't downgrade timestamping RX filters in SIOCSHWTSTAMP net: mscc: ocelot: correctly report the timestamping RX filters in ethtool Volodymyr Mytnyk (1): net: marvell: prestera: fix double free issue on err path Weichao Guo (1): f2fs: set SBI_NEED_FSCK flag when inconsistent node block found Werner Sembach (1): ALSA: hda/realtek: Add quirk for ASRock NUC Box 1100 Will Mortensen (1): netfilter: flowtable: fix IPv6 tunnel addr match Ziyang Xuan (1): net: vlan: fix underflow for the real_dev refcnt yangxingwu (1): netfilter: ipvs: Fix reuse connection if RS weight is 0 Documentation/networking/ipvs-sysctl.rst | 3 +- arch/arm/boot/dts/bcm2711.dtsi | 8 +- arch/arm/boot/dts/bcm5301x.dtsi | 4 +- arch/arm/mach-socfpga/core.h | 2 +- arch/arm/mach-socfpga/platsmp.c | 8 +- arch/mips/Kconfig | 2 +- arch/mips/kernel/cpu-probe.c | 4 +- arch/parisc/kernel/vmlinux.lds.S | 3 +- arch/powerpc/kernel/head_32.h | 6 +- arch/powerpc/kvm/book3s_hv_builtin.c | 5 +- arch/s390/mm/pgtable.c | 13 + drivers/acpi/property.c | 11 +- drivers/android/binder.c | 2 +- drivers/block/xen-blkfront.c | 126 +++++--- drivers/firmware/arm_scmi/scmi_pm_domain.c | 4 +- drivers/firmware/smccc/soc_id.c | 2 +- drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 46 ++- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 4 +- .../gpu/drm/nouveau/nvkm/subdev/acr/gm200.c | 6 +- .../gpu/drm/nouveau/nvkm/subdev/acr/gp102.c | 6 +- drivers/gpu/drm/vc4/vc4_bo.c | 2 +- drivers/hid/wacom_wac.c | 8 +- drivers/hid/wacom_wac.h | 1 + drivers/iommu/amd/iommu_v2.c | 6 +- drivers/media/cec/core/cec-adap.c | 1 + drivers/mmc/host/sdhci-esdhc-imx.c | 2 - drivers/mmc/host/sdhci.c | 21 +- drivers/mmc/host/sdhci.h | 4 +- .../net/ethernet/intel/iavf/iavf_ethtool.c | 30 +- drivers/net/ethernet/intel/ice/ice_lib.c | 9 +- drivers/net/ethernet/intel/ice/ice_main.c | 18 +- drivers/net/ethernet/intel/igb/igb_main.c | 2 +- .../net/ethernet/marvell/mvpp2/mvpp2_main.c | 14 +- .../marvell/prestera/prestera_switchdev.c | 6 +- drivers/net/ethernet/mellanox/mlxsw/minimal.c | 4 + .../net/ethernet/mellanox/mlxsw/spectrum.c | 5 + .../ethernet/mellanox/mlxsw/spectrum_ptp.c | 3 + .../ethernet/mellanox/mlxsw/spectrum_router.c | 3 + .../mellanox/mlxsw/spectrum_switchdev.c | 4 + drivers/net/ethernet/microchip/lan743x_main.c | 12 +- drivers/net/ethernet/mscc/ocelot.c | 11 +- drivers/net/ethernet/netronome/nfp/nfp_net.h | 3 - .../ethernet/netronome/nfp/nfp_net_ethtool.c | 2 +- drivers/net/ethernet/stmicro/stmmac/stmmac.h | 1 + .../net/ethernet/stmicro/stmmac/stmmac_main.c | 139 +++++---- .../ethernet/stmicro/stmmac/stmmac_platform.c | 44 +++ drivers/net/mdio/mdio-aspeed.c | 7 + drivers/net/phy/phylink.c | 26 +- drivers/net/xen-netfront.c | 272 ++++++++++------- drivers/nvme/target/io-cmd-file.c | 4 +- drivers/nvme/target/tcp.c | 7 +- drivers/pci/controller/pci-aardvark.c | 235 +++++++-------- drivers/scsi/mpt3sas/mpt3sas_scsih.c | 2 +- drivers/scsi/scsi_debug.c | 5 + drivers/scsi/scsi_sysfs.c | 2 +- drivers/staging/fbtft/fb_ssd1351.c | 4 - drivers/staging/fbtft/fbtft-core.c | 9 +- drivers/staging/greybus/audio_helper.c | 8 +- drivers/staging/rtl8192e/rtl8192e/rtl_core.c | 3 +- drivers/tty/hvc/hvc_xen.c | 17 +- drivers/usb/chipidea/ci_hdrc_imx.c | 18 +- drivers/usb/core/hub.c | 24 +- drivers/usb/dwc2/gadget.c | 17 +- drivers/usb/dwc2/hcd_queue.c | 2 +- drivers/usb/dwc3/gadget.c | 39 ++- drivers/usb/serial/option.c | 5 + drivers/usb/typec/tcpm/fusb302.c | 6 +- drivers/vhost/vsock.c | 2 +- drivers/xen/xenbus/xenbus_probe.c | 27 +- fs/ceph/super.c | 11 +- fs/cifs/file.c | 35 ++- fs/erofs/utils.c | 8 +- fs/f2fs/node.c | 1 + fs/fuse/dev.c | 10 +- fs/nfs/nfs42xdr.c | 3 +- fs/proc/vmcore.c | 16 +- include/linux/ipc_namespace.h | 15 + include/linux/sched/task.h | 2 +- include/net/ip6_fib.h | 1 + include/net/ipv6_stubs.h | 1 + include/net/nl802154.h | 7 +- include/xen/interface/io/ring.h | 278 ++++++++++-------- ipc/shm.c | 189 +++++++++--- kernel/cpu.c | 7 + kernel/power/hibernate.c | 6 +- kernel/sched/core.c | 4 - kernel/trace/trace.h | 24 +- kernel/trace/trace_events.c | 10 + kernel/trace/trace_uprobe.c | 1 + net/8021q/vlan.c | 3 - net/8021q/vlan_dev.c | 3 + net/ipv4/nexthop.c | 35 ++- net/ipv4/tcp.c | 4 +- net/ipv4/tcp_cubic.c | 5 +- net/ipv6/af_inet6.c | 1 + net/ipv6/ip6_output.c | 2 +- net/ipv6/route.c | 19 ++ net/mptcp/options.c | 3 +- net/ncsi/ncsi-cmd.c | 24 +- net/netfilter/ipvs/ip_vs_core.c | 8 +- net/netfilter/nf_conntrack_netlink.c | 6 +- net/netfilter/nf_flow_table_offload.c | 4 +- net/sched/sch_ets.c | 8 +- net/smc/af_smc.c | 12 +- net/smc/smc_close.c | 6 + net/smc/smc_core.c | 35 +-- net/tls/tls_main.c | 47 ++- net/tls/tls_sw.c | 23 +- sound/hda/intel-dsp-config.c | 9 + sound/pci/ctxfi/ctamixer.c | 14 +- sound/pci/ctxfi/ctdaio.c | 16 +- sound/pci/ctxfi/ctresource.c | 7 +- sound/pci/ctxfi/ctresource.h | 4 +- sound/pci/ctxfi/ctsrc.c | 7 +- sound/pci/hda/patch_realtek.c | 28 ++ sound/soc/codecs/wcd934x.c | 3 +- sound/soc/qcom/qdsp6/q6asm-dai.c | 19 +- sound/soc/qcom/qdsp6/q6routing.c | 6 +- sound/soc/soc-topology.c | 3 + 119 files changed, 1539 insertions(+), 815 deletions(-) -- 2.20.1

1 119

[PATCH openEuler-5.10 01/16] USB: gadget: detect too-big endpoint 0 requests
by Zheng Zengkai 14 Jan '22

14 Jan '22

From: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> stable inclusion form stable-v5.10.85 commit 7193ad3e50e596ac2192531c58ba83b9e6d2444b bugzilla: 185937 https://gitee.com/openeuler/kernel/issues/I4DDEL CVE: CVE-2021-39685 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- Sometimes USB hosts can ask for buffers that are too large from endpoint 0, which should not be allowed. If this happens for OUT requests, stall the endpoint, but for IN requests, trim the request size to the endpoint buffer size. Co-developed-by: Szymon Heidrich <szymon.heidrich(a)gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Chen Jun <chenjun102(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> --- drivers/usb/gadget/composite.c | 12 ++++++++++++ drivers/usb/gadget/legacy/dbgp.c | 13 +++++++++++++ drivers/usb/gadget/legacy/inode.c | 16 +++++++++++++++- 3 files changed, 40 insertions(+), 1 deletion(-) diff --git a/drivers/usb/gadget/composite.c b/drivers/usb/gadget/composite.c index a8704e6498ab..426132988512 100644 --- a/drivers/usb/gadget/composite.c +++ b/drivers/usb/gadget/composite.c @@ -1648,6 +1648,18 @@ composite_setup(struct usb_gadget *gadget, const struct usb_ctrlrequest *ctrl) struct usb_function *f = NULL; u8 endp; + if (w_length > USB_COMP_EP0_BUFSIZ) { + if (ctrl->bRequestType == USB_DIR_OUT) { + goto done; + } else { + /* Cast away the const, we are going to overwrite on purpose. */ + __le16 *temp = (__le16 *)&ctrl->wLength; + + *temp = cpu_to_le16(USB_COMP_EP0_BUFSIZ); + w_length = USB_COMP_EP0_BUFSIZ; + } + } + /* partial re-init of the response message; the function or the * gadget might need to intercept e.g. a control-OUT completion * when we delegate to it. diff --git a/drivers/usb/gadget/legacy/dbgp.c b/drivers/usb/gadget/legacy/dbgp.c index e1d566c9918a..e567afcb2794 100644 --- a/drivers/usb/gadget/legacy/dbgp.c +++ b/drivers/usb/gadget/legacy/dbgp.c @@ -345,6 +345,19 @@ static int dbgp_setup(struct usb_gadget *gadget, void *data = NULL; u16 len = 0; + if (length > DBGP_REQ_LEN) { + if (ctrl->bRequestType == USB_DIR_OUT) { + return err; + } else { + /* Cast away the const, we are going to overwrite on purpose. */ + __le16 *temp = (__le16 *)&ctrl->wLength; + + *temp = cpu_to_le16(DBGP_REQ_LEN); + length = DBGP_REQ_LEN; + } + } + + if (request == USB_REQ_GET_DESCRIPTOR) { switch (value>>8) { case USB_DT_DEVICE: diff --git a/drivers/usb/gadget/legacy/inode.c b/drivers/usb/gadget/legacy/inode.c index 71e7d10dd76b..04b9c4f5f129 100644 --- a/drivers/usb/gadget/legacy/inode.c +++ b/drivers/usb/gadget/legacy/inode.c @@ -110,6 +110,8 @@ enum ep0_state { /* enough for the whole queue: most events invalidate others */ #define N_EVENT 5 +#define RBUF_SIZE 256 + struct dev_data { spinlock_t lock; refcount_t count; @@ -144,7 +146,7 @@ struct dev_data { struct dentry *dentry; /* except this scratch i/o buffer for ep0 */ - u8 rbuf [256]; + u8 rbuf[RBUF_SIZE]; }; static inline void get_dev (struct dev_data *data) @@ -1333,6 +1335,18 @@ gadgetfs_setup (struct usb_gadget *gadget, const struct usb_ctrlrequest *ctrl) u16 w_value = le16_to_cpu(ctrl->wValue); u16 w_length = le16_to_cpu(ctrl->wLength); + if (w_length > RBUF_SIZE) { + if (ctrl->bRequestType == USB_DIR_OUT) { + return value; + } else { + /* Cast away the const, we are going to overwrite on purpose. */ + __le16 *temp = (__le16 *)&ctrl->wLength; + + *temp = cpu_to_le16(RBUF_SIZE); + w_length = RBUF_SIZE; + } + } + spin_lock (&dev->lock); dev->setup_abort = 0; if (dev->state == STATE_DEV_UNCONNECTED) { -- 2.20.1

1 15

[PATCH openEuler-5.10 000/146] Backport LTS 5.10.82
by Zheng Zengkai 14 Jan '22

14 Jan '22

Backport LTS 5.10.82 patches from upstream. Adrian Hunter (2): scsi: ufs: core: Fix task management completion scsi: ufs: core: Fix task management completion timeout race Alex Elder (1): net: ipa: disable HOLB drop when updating timer Alexander Antonov (2): perf/x86/intel/uncore: Fix filter_tid mask for CHA events on Skylake Server perf/x86/intel/uncore: Fix IIO event constraints for Skylake Server Alexander Mikhalitsyn (1): ipc: WARN if trying to remove ipc object which is absent Alistair Delva (1): block: Check ADMIN before NICE for IOPRIO_CLASS_RT Alvin Lee (1): drm/amd/display: Update swizzle mode enums Amit Kumar Mahapatra (1): arm64: zynqmp: Do not duplicate flash partition label property Anatolij Gustschin (1): powerpc/5200: dts: fix memory node unit name AngeloGioacchino Del Regno (1): arm64: dts: qcom: msm8998: Fix CPU/L2 idle state latency and residency Arjun Roy (3): net-zerocopy: Copy straggler unaligned data for TCP Rx. zerocopy. net-zerocopy: Refactor skb frag fast-forward op. tcp: Fix uninitialized access in skb frags array for Rx 0cp. Baoquan He (1): s390/kexec: fix memory leak of ipl report buffer Bart Van Assche (1): MIPS: sni: Fix the build Bjorn Andersson (1): pinctrl: qcom: sdm845: Enable dual edge errata Bongsu Jeon (1): net: nfc: nci: Change the NCI close sequence Chao Yu (1): f2fs: fix incorrect return value in f2fs_sanity_check_ckpt() Chengfeng Ye (1): ALSA: gus: fix null pointer dereference on pointer block Christian Lamparter (1): ARM: BCM53016: Specify switch ports for Meraki MR32 Christophe JAILLET (1): platform/x86: hp_accel: Fix an error handling path in 'lis3lv02d_probe()' Christophe Leroy (2): powerpc/8xx: Fix Oops with STRICT_KERNEL_RWX without DEBUG_RODATA_TEST powerpc/8xx: Fix pinned TLBs with CONFIG_STRICT_KERNEL_RWX Colin Ian King (1): MIPS: generic/yamon-dt: fix uninitialized variable error David Heidelberg (1): ARM: dts: qcom: fix memory and mdio nodes naming for RB3011 Dmitry Baryshkov (1): clk: qcom: gcc-msm8996: Drop (again) gcc_aggre1_pnoc_ahb_clk Eric Dumazet (1): net: reduce indentation level in sk_clone_lock() Eryk Rybak (3): i40e: Fix correct max_pkt_size on VF RX queue i40e: Fix changing previously set num_queue_pairs for PFs i40e: Fix ping is lost after configuring ADq on VF Ewan D. Milne (1): scsi: qla2xxx: Fix mailbox direction flags in qla2xxx_get_adapter_id() Fabio Aiuto (1): staging: rtl8723bs: remove possible deadlock when disconnect (v2) Gao Xiang (1): f2fs: fix up f2fs_lookup tracepoints Grzegorz Szczurek (2): iavf: Fix for setting queues to 0 i40e: Fix display error code in dmesg Guanghui Feng (1): tty: tty_buffer: Fix the softlockup issue in flush_to_ldisc Guo Zhi (1): scsi: advansys: Fix kernel pointer leak Hans Verkuil (1): drm/nouveau: hdmigv100.c: fix corrupted HDMI Vendor InfoFrame Hans de Goede (1): ASoC: nau8824: Add DMI quirk mechanism for active-high jack-detect Heiko Carstens (1): s390/kexec: fix return code handling Hyeong-Jun Kim (1): f2fs: compress: disallow disabling compress on non-empty compressed file Ian Rogers (1): perf bpf: Avoid memory leak from perf_env__insert_btf() Imre Deak (1): drm/i915/dp: Ensure sink rate values are always valid Jacob Keller (1): iavf: prevent accidental free of filter structure James Clark (1): perf tests: Remove bash construct from record+zstd_comp_decomp.sh James Smart (1): scsi: lpfc: Fix list_add() corruption in lpfc_drain_txq() Jan Kara (1): udf: Fix crash after seekdir Jedrzej Jagielski (1): i40e: Fix creation of first queue by omitting it if is not power of two Jesse Brandeburg (1): e100: fix device suspend/resume Joel Stanley (1): clk/ast2600: Fix soc revision for AHB Johan Hovold (1): drm/udl: fix control-message timeout Jonathan Davies (1): net: virtio_net_hdr_to_skb: count transport header in UFO Josef Bacik (2): fs: export an inode_update_time helper btrfs: update device path inode time instead of bd_inode Jérôme Pouiller (1): staging: wfx: ensure IRQ is ready before enabling it Karen Sornek (1): i40e: Fix warning message and call stack during rmmod i40e driver Keoseong Park (1): f2fs: fix to use WHINT_MODE Leon Romanovsky (2): RDMA/netlink: Add __maybe_unused to static inline in C file ice: Delete always true check of PF pointer Li Yang (2): ARM: dts: ls1021a: move thermal-zones node out of soc/ ARM: dts: ls1021a-tsn: use generic "jedec,spi-nor" compatible for flash Like Xu (1): perf/x86/vlbr: Add c->flags to vlbr event constraints Lin Ma (3): NFC: reorganize the functions in nci_request NFC: reorder the logic in nfc_{un,}register_device NFC: add NCI_UNREG flag to eliminate the race Linus Walleij (1): ARM: dts: ux500: Skomer regulator fixes Lu Wei (1): maple: fix wrong return value of maple_bus_init(). Luis Chamberlain (1): firmware_loader: fix pre-allocated buf built-in firmware use Maher Sanalla (1): net/mlx5: Lag, update tracker when state change event received Masami Hiramatsu (1): tracing/histogram: Do not copy the fixed-size char array field over the field size Mateusz Palczewski (1): iavf: Fix return of set the new channel count Matthew Hagan (1): ARM: dts: NSP: Fix mpcore, mmc node names Matthias Brugger (1): arm64: dts: rockchip: Disable CDN DP on Pinebook Pro Maxim Levitsky (1): KVM: nVMX: don't use vcpu->arch.efer when checking host state on nested state load Maxime Ripard (3): ARM: dts: sunxi: Fix OPPs node name arm64: dts: allwinner: h5: Fix GPU thermal zone node name arm64: dts: allwinner: a100: Fix thermal zone node name Meng Li (1): net: stmmac: socfpga: add runtime suspend/resume callback for stratix10 platform Michael Ellerman (2): powerpc/dcr: Use cmplwi instead of 3-argument cmpli KVM: PPC: Book3S HV: Use GLOBAL_TOC for kvmppc_h_set_dabr/xdabr() Michael Walle (2): arm64: dts: hisilicon: fix arm,sp805 compatible string arm64: dts: freescale: fix arm,sp805 compatible string Michal Maloszewski (1): i40e: Fix NULL ptr dereference on VSI filter sync Michal Simek (1): arm64: zynqmp: Fix serial compatible string Mike Christie (3): scsi: target: Fix ordered tag handling scsi: target: Fix alua_tg_pt_gps_count tracking scsi: core: sysfs: Fix hang when device state is set via sysfs Mitch Williams (1): iavf: validate pointers Nathan Chancellor (2): hexagon: export raw I/O routines for modules hexagon: clean up timer-regs.h Nguyen Dinh Phi (1): cfg80211: call cfg80211_stop_ap when switch from P2P_GO type Nicholas Nunley (2): iavf: check for null in iavf_fix_features iavf: free q_vectors before queues in iavf_disable_vf Nick Desaulniers (2): sh: check return code of request_irq arm64: vdso32: suppress error message for 'make mrproper' Nicolas Dichtel (1): tun: fix bonding active backup with arp monitoring Nikolay Borisov (1): btrfs: fix memory ordering between normal and ordered work functions Ondrej Mosnacek (1): selinux: fix NULL-pointer dereference when hashtab allocation fails Paul Cercueil (1): clk: ingenic: Fix bugs with divided dividers Pavel Skripkin (2): net: bnx2x: fix variable dereferenced before check net: dpaa2-eth: fix use-after-free in dpaa2_eth_remove Pierre-Louis Bossart (4): ASoC: SOF: Intel: hda-dai: fix potential locking issue ALSA: intel-dsp-config: add quirk for APL/GLK/TGL devices based on ES8336 codec ALSA: hda: hdac_ext_stream: fix potential locking issues ALSA: hda: hdac_stream: fix potential locking issue in snd_hdac_stream_assign() Piotr Marczak (1): iavf: Fix failure to exit out from last all-multicast mode Punit Agrawal (1): net: stmmac: dwmac-rk: Fix ethernet on rk3399 based devices Raed Salem (1): net/mlx5: E-Switch, return error if encap isn't supported Randy Dunlap (8): ALSA: ISA: not for M68K sh: fix kconfig unmet dependency warning for FRAME_POINTER sh: math-emu: drop unused functions sh: define __BIG_ENDIAN for math-emu mips: BCM63XX: ensure that CPU_SUPPORTS_32BIT_KERNEL is set mips: bcm63xx: add support for clk_get_parent() mips: lantiq: add support for clk_get_parent() x86/Kconfig: Fix an unused variable error in dell-smm-hwmon Roger Quadros (1): ARM: dts: omap: fix gpmc,mux-add-data type Roi Dayan (1): net/mlx5: E-Switch, Change mode lock from mutex to rw semaphore Rustam Kovhaev (1): mm: kmemleak: slob: respect SLAB_NOLEAKTRACE flag Sasha Levin (1): Revert "perf: Rework perf_event_exit_event()" Sean Christopherson (1): x86/hyperv: Fix NULL deref in set_hv_tscchange_cb() if Hyper-V setup fails Selvin Xavier (1): RDMA/bnxt_re: Check if the vlan is valid before reporting Shawn Guo (1): arm64: dts: qcom: ipq6018: Fix qcom,controlled-remotely property Sohaib Mohamed (1): perf bench futex: Fix memory leak of perf_cpu_map__new() Sriharsha Basavapatna (1): bnxt_en: reject indirect blk offload when hw-tc-offload is off Stefan Riedmueller (1): clk: imx: imx6ul: Move csi_sel mux to correct base register Steven Rostedt (VMware) (1): tracing: Add length protection to histogram string copies Surabhi Boob (1): iavf: Fix for the false positive ASQ/ARQ errors while issuing VF reset Sven Peter (1): usb: typec: tipd: Remove WARN_ON in tps6598x_block_read Sven Schnelle (1): parisc/sticon: fix reverse colors Tadeusz Struk (1): tipc: check for null after calling kmemdup Takashi Iwai (1): ASoC: DAPM: Cover regression by kctl change notification fix Teng Qi (1): iio: imu: st_lsm6dsx: Avoid potential array overflow in st_lsm6dsx_set_odr() Tetsuo Handa (1): sock: fix /proc/net/sockstat underflow in sk_clone_lock() Tony Lindgren (2): bus: ti-sysc: Add quirk handling for reinit on context lost bus: ti-sysc: Use context lost quirk for otg Uwe Kleine-König (1): usb: max-3421: Use driver data instead of maintaining a list of bound devices Valentine Fatiev (1): net/mlx5e: nullify cq->dbg pointer in mlx5_debug_cq_remove() Vincent Donnefort (1): sched/core: Mitigate race cpus_share_cache()/update_top_cache_domain() Wen Gu (1): net/smc: Make sure the link_id is unique Xin Long (2): tipc: only accept encrypted MSG_CRYPTO msgs net: sched: act_mirred: drop dst for the direction from egress to ingress Yang Yingliang (2): usb: musb: tusb6010: check return value after calling platform_get_resource() usb: host: ohci-tmio: check return value after calling platform_get_resource() hongao (1): drm/amdgpu: fix set scaling mode Full/Full aspect/Center not works on vga and dvi connectors arch/arm/boot/dts/bcm-nsp.dtsi | 4 +- arch/arm/boot/dts/bcm53016-meraki-mr32.dts | 22 ++ arch/arm/boot/dts/ls1021a-tsn.dts | 2 +- arch/arm/boot/dts/ls1021a.dtsi | 66 +++--- arch/arm/boot/dts/omap-gpmc-smsc9221.dtsi | 2 +- .../boot/dts/omap3-overo-tobiduo-common.dtsi | 2 +- arch/arm/boot/dts/qcom-ipq8064-rb3011.dts | 6 +- .../arm/boot/dts/ste-ux500-samsung-skomer.dts | 8 +- arch/arm/boot/dts/sun8i-a33.dtsi | 4 +- arch/arm/boot/dts/sun8i-a83t.dtsi | 4 +- arch/arm/boot/dts/sun8i-h3.dtsi | 4 +- .../arm64/boot/dts/allwinner/sun50i-a100.dtsi | 6 +- .../dts/allwinner/sun50i-a64-cpu-opp.dtsi | 2 +- .../boot/dts/allwinner/sun50i-h5-cpu-opp.dtsi | 2 +- arch/arm64/boot/dts/allwinner/sun50i-h5.dtsi | 2 +- .../boot/dts/allwinner/sun50i-h6-cpu-opp.dtsi | 2 +- .../arm64/boot/dts/freescale/fsl-ls1088a.dtsi | 16 +- .../arm64/boot/dts/freescale/fsl-ls208xa.dtsi | 16 +- arch/arm64/boot/dts/hisilicon/hi3660.dtsi | 4 +- arch/arm64/boot/dts/hisilicon/hi6220.dtsi | 2 +- arch/arm64/boot/dts/qcom/ipq6018.dtsi | 2 +- arch/arm64/boot/dts/qcom/msm8998.dtsi | 20 +- .../boot/dts/rockchip/rk3399-pinebook-pro.dts | 4 - .../dts/xilinx/zynqmp-zc1751-xm016-dc2.dts | 4 +- arch/arm64/boot/dts/xilinx/zynqmp.dtsi | 4 +- arch/arm64/kernel/vdso32/Makefile | 3 +- arch/hexagon/include/asm/timer-regs.h | 26 --- arch/hexagon/include/asm/timex.h | 3 +- arch/hexagon/kernel/time.c | 12 +- arch/hexagon/lib/io.c | 4 + arch/mips/Kconfig | 3 + arch/mips/bcm63xx/clk.c | 6 + arch/mips/generic/yamon-dt.c | 2 +- arch/mips/lantiq/clk.c | 6 + arch/mips/sni/time.c | 4 +- arch/powerpc/boot/dts/charon.dts | 2 +- arch/powerpc/boot/dts/digsy_mtc.dts | 2 +- arch/powerpc/boot/dts/lite5200.dts | 2 +- arch/powerpc/boot/dts/lite5200b.dts | 2 +- arch/powerpc/boot/dts/media5200.dts | 2 +- arch/powerpc/boot/dts/mpc5200b.dtsi | 2 +- arch/powerpc/boot/dts/o2d.dts | 2 +- arch/powerpc/boot/dts/o2d.dtsi | 2 +- arch/powerpc/boot/dts/o2dnt2.dts | 2 +- arch/powerpc/boot/dts/o3dnt.dts | 2 +- arch/powerpc/boot/dts/pcm032.dts | 2 +- arch/powerpc/boot/dts/tqm5200.dts | 2 +- arch/powerpc/kernel/head_8xx.S | 13 +- arch/powerpc/kvm/book3s_hv_rmhandlers.S | 4 +- arch/powerpc/sysdev/dcr-low.S | 2 +- arch/s390/include/asm/kexec.h | 6 + arch/s390/kernel/ipl.c | 3 +- arch/s390/kernel/machine_kexec_file.c | 18 +- arch/sh/Kconfig.debug | 1 + arch/sh/include/asm/sfp-machine.h | 8 + arch/sh/kernel/cpu/sh4a/smp-shx3.c | 5 +- arch/sh/math-emu/math.c | 103 ---------- arch/x86/Kconfig | 3 +- arch/x86/events/intel/core.c | 4 +- arch/x86/events/intel/uncore_snbep.c | 4 + arch/x86/hyperv/hv_init.c | 3 + arch/x86/kvm/vmx/nested.c | 22 +- block/ioprio.c | 9 +- drivers/base/firmware_loader/main.c | 13 +- drivers/bus/ti-sysc.c | 110 +++++++++- drivers/clk/clk-ast2600.c | 12 +- drivers/clk/imx/clk-imx6ul.c | 2 +- drivers/clk/ingenic/cgu.c | 6 +- drivers/clk/qcom/gcc-msm8996.c | 15 -- .../gpu/drm/amd/amdgpu/amdgpu_connectors.c | 1 + .../drm/amd/display/dc/dcn20/dcn20_resource.c | 4 +- .../amd/display/dc/dml/display_mode_enums.h | 4 +- drivers/gpu/drm/i915/display/intel_dp.c | 11 + .../drm/nouveau/nvkm/engine/disp/hdmigv100.c | 1 - drivers/gpu/drm/udl/udl_connector.c | 2 +- drivers/iio/imu/st_lsm6dsx/st_lsm6dsx_core.c | 6 +- drivers/infiniband/hw/bnxt_re/ib_verbs.c | 12 +- .../ethernet/broadcom/bnx2x/bnx2x_init_ops.h | 4 +- drivers/net/ethernet/broadcom/bnxt/bnxt_tc.c | 2 +- .../net/ethernet/freescale/dpaa2/dpaa2-eth.c | 4 +- drivers/net/ethernet/intel/e100.c | 18 +- drivers/net/ethernet/intel/i40e/i40e.h | 2 + drivers/net/ethernet/intel/i40e/i40e_main.c | 160 +++++++++------ .../ethernet/intel/i40e/i40e_virtchnl_pf.c | 121 ++++------- .../net/ethernet/intel/iavf/iavf_ethtool.c | 30 ++- drivers/net/ethernet/intel/iavf/iavf_main.c | 14 +- drivers/net/ethernet/intel/ice/ice_main.c | 3 - drivers/net/ethernet/mellanox/mlx5/core/cq.c | 5 +- .../net/ethernet/mellanox/mlx5/core/debugfs.c | 4 +- .../net/ethernet/mellanox/mlx5/core/eswitch.c | 11 +- .../net/ethernet/mellanox/mlx5/core/eswitch.h | 2 +- .../mellanox/mlx5/core/eswitch_offloads.c | 28 +-- drivers/net/ethernet/mellanox/mlx5/core/lag.c | 28 ++- .../net/ethernet/stmicro/stmmac/dwmac-rk.c | 5 + .../ethernet/stmicro/stmmac/dwmac-socfpga.c | 24 ++- drivers/net/ipa/ipa_endpoint.c | 2 + drivers/net/tun.c | 5 + drivers/pinctrl/qcom/pinctrl-sdm845.c | 1 + drivers/platform/x86/hp_accel.c | 2 + drivers/scsi/advansys.c | 4 +- drivers/scsi/lpfc/lpfc_sli.c | 1 + drivers/scsi/qla2xxx/qla_mbx.c | 6 +- drivers/scsi/scsi_sysfs.c | 30 ++- drivers/scsi/ufs/ufshcd.c | 57 +++--- drivers/scsi/ufs/ufshcd.h | 1 + drivers/sh/maple/maple.c | 5 +- drivers/staging/rtl8723bs/core/rtw_mlme_ext.c | 7 +- drivers/staging/rtl8723bs/core/rtw_recv.c | 10 +- drivers/staging/rtl8723bs/core/rtw_sta_mgt.c | 22 +- drivers/staging/rtl8723bs/core/rtw_xmit.c | 16 +- .../staging/rtl8723bs/hal/rtl8723bs_xmit.c | 2 - drivers/staging/wfx/bus_sdio.c | 17 +- drivers/target/target_core_alua.c | 1 - drivers/target/target_core_device.c | 2 + drivers/target/target_core_internal.h | 1 + drivers/target/target_core_transport.c | 76 +++++-- drivers/tty/tty_buffer.c | 3 + drivers/usb/host/max3421-hcd.c | 25 +-- drivers/usb/host/ohci-tmio.c | 2 +- drivers/usb/musb/tusb6010.c | 5 + drivers/usb/typec/tps6598x.c | 2 +- drivers/video/console/sticon.c | 12 +- fs/btrfs/async-thread.c | 14 ++ fs/btrfs/volumes.c | 21 +- fs/f2fs/f2fs.h | 3 +- fs/f2fs/super.c | 4 +- fs/inode.c | 7 +- fs/udf/dir.c | 32 ++- fs/udf/namei.c | 3 + fs/udf/super.c | 2 + include/linux/fs.h | 2 + include/linux/perf_event.h | 1 - include/linux/platform_data/ti-sysc.h | 1 + include/linux/trace_events.h | 2 +- include/linux/virtio_net.h | 7 +- include/net/nfc/nci_core.h | 1 + include/rdma/rdma_netlink.h | 2 +- include/sound/hdaudio_ext.h | 2 + include/target/target_core_base.h | 6 +- include/trace/events/f2fs.h | 12 +- include/uapi/linux/tcp.h | 2 + ipc/util.c | 6 +- kernel/events/core.c | 142 ++++++------- kernel/sched/core.c | 3 + kernel/trace/trace_events_hist.c | 14 +- mm/slab.h | 2 +- net/core/sock.c | 189 +++++++++--------- net/ipv4/tcp.c | 122 ++++++++--- net/nfc/core.c | 32 +-- net/nfc/nci/core.c | 34 +++- net/sched/act_mirred.c | 11 +- net/smc/smc_core.c | 3 +- net/tipc/crypto.c | 4 + net/tipc/link.c | 7 +- net/wireless/util.c | 1 + security/selinux/ss/hashtab.c | 17 +- sound/core/Makefile | 2 + sound/hda/ext/hdac_ext_stream.c | 46 +++-- sound/hda/hdac_stream.c | 4 +- sound/hda/intel-dsp-config.c | 22 +- sound/isa/Kconfig | 2 +- sound/isa/gus/gus_dma.c | 2 + sound/pci/Kconfig | 1 + sound/soc/codecs/nau8824.c | 40 ++++ sound/soc/soc-dapm.c | 29 ++- sound/soc/sof/intel/hda-dai.c | 7 +- tools/perf/bench/futex-lock-pi.c | 1 + tools/perf/bench/futex-requeue.c | 1 + tools/perf/bench/futex-wake-parallel.c | 1 + tools/perf/bench/futex-wake.c | 1 + .../tests/shell/record+zstd_comp_decomp.sh | 2 +- tools/perf/util/bpf-event.c | 6 +- tools/perf/util/env.c | 5 +- tools/perf/util/env.h | 2 +- 174 files changed, 1413 insertions(+), 964 deletions(-) delete mode 100644 arch/hexagon/include/asm/timer-regs.h -- 2.20.1

1 146

[PATCH openEuler-5.10 01/16] drm/i915/guc: Update to use firmware v49.0.1
by Zheng Zengkai 14 Jan '22

14 Jan '22

From: John Harrison <John.C.Harrison(a)Intel.com> mainline inclusion from mainline-5.11-rc1 commit c784e5249e773689e38d2bc1749f08b986621a26 bugzilla: 185850 https://gitee.com/openeuler/kernel/issues/I4DDEL CVE: CVE-2020-12363, CVE-2020-12364 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- The latest GuC firmware includes a number of interface changes that require driver updates to match. * Starting from Gen11, the ID to be provided to GuC needs to contain the engine class in bits [0..2] and the instance in bits [3..6]. NOTE: this patch breaks pointer dereferences in some existing GuC functions that use the guc_id to dereference arrays but these functions are not used for now as we have GuC submission disabled and we will update these functions in follow up patch which requires new IDs. * The new GuC requires the additional data structure (ADS) and associated 'private_data' pointer to be setup. This is basically a scratch area of memory that the GuC owns. The size is read from the CSS header. * There is now a physical to logical engine mapping table in the ADS which needs to be configured in order for the firmware to load. For now, the table is initialised with a 1 to 1 mapping. * GUC_CTL_CTXINFO has been removed from the initialization params. * reg_state_buffer is maintained internally by the GuC as part of the private data. * The ADS layout has changed significantly. This patch updates the shared structure and also adds better documentation of the layout. * While i915 does not use GuC doorbells, the firmware now requires that some initialisation is done. * The number of engine classes and instances supported in the ADS has been increased. Signed-off-by: John Harrison <John.C.Harrison(a)Intel.com> Signed-off-by: Matthew Brost <matthew.brost(a)intel.com> Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio(a)intel.com> Signed-off-by: Oscar Mateo <oscar.mateo(a)intel.com> Signed-off-by: Michel Thierry <michel.thierry(a)intel.com> Signed-off-by: Rodrigo Vivi <rodrigo.vivi(a)intel.com> Signed-off-by: Michal Wajdeczko <michal.wajdeczko(a)intel.com> Cc: Michal Winiarski <michal.winiarski(a)intel.com> Cc: Tomasz Lis <tomasz.lis(a)intel.com> Cc: Joonas Lahtinen <joonas.lahtinen(a)linux.intel.com> Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio(a)intel.com> Signed-off-by: Joonas Lahtinen <joonas.lahtinen(a)linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20201028145826.2949180-2-John… Signed-off-by: Chen Jun <chenjun102(a)huawei.com> Reviewed-by: Xiu Jianfeng <xiujianfeng(a)huawei.com> Reviewed-by: Jason Yan <yanaijie(a)huawei.com> Signed-off-by: Chen Jun <chenjun102(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> --- drivers/gpu/drm/i915/gt/intel_engine_cs.c | 3 +- drivers/gpu/drm/i915/gt/uc/intel_guc.c | 18 --- drivers/gpu/drm/i915/gt/uc/intel_guc_ads.c | 131 +++++++++++++++---- drivers/gpu/drm/i915/gt/uc/intel_guc_fwif.h | 80 +++++------ drivers/gpu/drm/i915/gt/uc/intel_guc_reg.h | 5 + drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c | 27 ++-- drivers/gpu/drm/i915/gt/uc/intel_uc_fw.h | 2 + drivers/gpu/drm/i915/gt/uc/intel_uc_fw_abi.h | 6 +- 8 files changed, 176 insertions(+), 96 deletions(-) diff --git a/drivers/gpu/drm/i915/gt/intel_engine_cs.c b/drivers/gpu/drm/i915/gt/intel_engine_cs.c index a19537706ed1..c940ac3aae2f 100644 --- a/drivers/gpu/drm/i915/gt/intel_engine_cs.c +++ b/drivers/gpu/drm/i915/gt/intel_engine_cs.c @@ -305,8 +305,9 @@ static int intel_engine_setup(struct intel_gt *gt, enum intel_engine_id id) engine->i915 = i915; engine->gt = gt; engine->uncore = gt->uncore; - engine->hw_id = engine->guc_id = info->hw_id; engine->mmio_base = __engine_mmio_base(i915, info->mmio_bases); + engine->hw_id = info->hw_id; + engine->guc_id = MAKE_GUC_ID(info->class, info->instance); engine->class = info->class; engine->instance = info->instance; diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc.c b/drivers/gpu/drm/i915/gt/uc/intel_guc.c index 942c7c187adb..6909da1e1a73 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_guc.c +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc.c @@ -213,23 +213,6 @@ static u32 guc_ctl_feature_flags(struct intel_guc *guc) return flags; } -static u32 guc_ctl_ctxinfo_flags(struct intel_guc *guc) -{ - u32 flags = 0; - - if (intel_guc_submission_is_used(guc)) { - u32 ctxnum, base; - - base = intel_guc_ggtt_offset(guc, guc->stage_desc_pool); - ctxnum = GUC_MAX_STAGE_DESCRIPTORS / 16; - - base >>= PAGE_SHIFT; - flags |= (base << GUC_CTL_BASE_ADDR_SHIFT) | - (ctxnum << GUC_CTL_CTXNUM_IN16_SHIFT); - } - return flags; -} - static u32 guc_ctl_log_params_flags(struct intel_guc *guc) { u32 offset = intel_guc_ggtt_offset(guc, guc->log.vma) >> PAGE_SHIFT; @@ -291,7 +274,6 @@ static void guc_init_params(struct intel_guc *guc) BUILD_BUG_ON(sizeof(guc->params) != GUC_CTL_MAX_DWORDS * sizeof(u32)); - params[GUC_CTL_CTXINFO] = guc_ctl_ctxinfo_flags(guc); params[GUC_CTL_LOG_PARAMS] = guc_ctl_log_params_flags(guc); params[GUC_CTL_FEATURE] = guc_ctl_feature_flags(guc); params[GUC_CTL_DEBUG] = guc_ctl_debug_flags(guc); diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_ads.c b/drivers/gpu/drm/i915/gt/uc/intel_guc_ads.c index d44061033f23..7950d28beb8c 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_guc_ads.c +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_ads.c @@ -10,11 +10,52 @@ /* * The Additional Data Struct (ADS) has pointers for different buffers used by - * the GuC. One single gem object contains the ADS struct itself (guc_ads), the - * scheduling policies (guc_policies), a structure describing a collection of - * register sets (guc_mmio_reg_state) and some extra pages for the GuC to save - * its internal state for sleep. + * the GuC. One single gem object contains the ADS struct itself (guc_ads) and + * all the extra buffers indirectly linked via the ADS struct's entries. + * + * Layout of the ADS blob allocated for the GuC: + * + * +---------------------------------------+ <== base + * | guc_ads | + * +---------------------------------------+ + * | guc_policies | + * +---------------------------------------+ + * | guc_gt_system_info | + * +---------------------------------------+ + * | guc_clients_info | + * +---------------------------------------+ + * | guc_ct_pool_entry[size] | + * +---------------------------------------+ + * | padding | + * +---------------------------------------+ <== 4K aligned + * | private data | + * +---------------------------------------+ + * | padding | + * +---------------------------------------+ <== 4K aligned */ +struct __guc_ads_blob { + struct guc_ads ads; + struct guc_policies policies; + struct guc_gt_system_info system_info; + struct guc_clients_info clients_info; + struct guc_ct_pool_entry ct_pool[GUC_CT_POOL_SIZE]; +} __packed; + +static u32 guc_ads_private_data_size(struct intel_guc *guc) +{ + return PAGE_ALIGN(guc->fw.private_data_size); +} + +static u32 guc_ads_private_data_offset(struct intel_guc *guc) +{ + return PAGE_ALIGN(sizeof(struct __guc_ads_blob)); +} + +static u32 guc_ads_blob_size(struct intel_guc *guc) +{ + return guc_ads_private_data_offset(guc) + + guc_ads_private_data_size(guc); +} static void guc_policy_init(struct guc_policy *policy) { @@ -48,26 +89,37 @@ static void guc_ct_pool_entries_init(struct guc_ct_pool_entry *pool, u32 num) memset(pool, 0, num * sizeof(*pool)); } +static void guc_mapping_table_init(struct intel_gt *gt, + struct guc_gt_system_info *system_info) +{ + unsigned int i, j; + struct intel_engine_cs *engine; + enum intel_engine_id id; + + /* Table must be set to invalid values for entries not used */ + for (i = 0; i < GUC_MAX_ENGINE_CLASSES; ++i) + for (j = 0; j < GUC_MAX_INSTANCES_PER_CLASS; ++j) + system_info->mapping_table[i][j] = + GUC_MAX_INSTANCES_PER_CLASS; + + for_each_engine(engine, gt, id) { + u8 guc_class = engine->class; + + system_info->mapping_table[guc_class][engine->instance] = + engine->instance; + } +} + /* * The first 80 dwords of the register state context, containing the * execlists and ppgtt registers. */ #define LR_HW_CONTEXT_SIZE (80 * sizeof(u32)) -/* The ads obj includes the struct itself and buffers passed to GuC */ -struct __guc_ads_blob { - struct guc_ads ads; - struct guc_policies policies; - struct guc_mmio_reg_state reg_state; - struct guc_gt_system_info system_info; - struct guc_clients_info clients_info; - struct guc_ct_pool_entry ct_pool[GUC_CT_POOL_SIZE]; - u8 reg_state_buffer[GUC_S3_SAVE_SPACE_PAGES * PAGE_SIZE]; -} __packed; - static void __guc_ads_init(struct intel_guc *guc) { struct intel_gt *gt = guc_to_gt(guc); + struct drm_i915_private *i915 = gt->i915; struct __guc_ads_blob *blob = guc->ads_blob; const u32 skipped_size = LRC_PPHWSP_SZ * PAGE_SIZE + LR_HW_CONTEXT_SIZE; u32 base; @@ -99,13 +151,25 @@ static void __guc_ads_init(struct intel_guc *guc) } /* System info */ - blob->system_info.slice_enabled = hweight8(gt->info.sseu.slice_mask); - blob->system_info.rcs_enabled = 1; - blob->system_info.bcs_enabled = 1; + blob->system_info.engine_enabled_masks[RENDER_CLASS] = 1; + blob->system_info.engine_enabled_masks[COPY_ENGINE_CLASS] = 1; + blob->system_info.engine_enabled_masks[VIDEO_DECODE_CLASS] = VDBOX_MASK(gt); + blob->system_info.engine_enabled_masks[VIDEO_ENHANCEMENT_CLASS] = VEBOX_MASK(gt); + + blob->system_info.generic_gt_sysinfo[GUC_GENERIC_GT_SYSINFO_SLICE_ENABLED] = + hweight8(gt->info.sseu.slice_mask); + blob->system_info.generic_gt_sysinfo[GUC_GENERIC_GT_SYSINFO_VDBOX_SFC_SUPPORT_MASK] = + gt->info.vdbox_sfc_access; + + if (INTEL_GEN(i915) >= 12 && !IS_DGFX(i915)) { + u32 distdbreg = intel_uncore_read(gt->uncore, + GEN12_DIST_DBS_POPULATED); + blob->system_info.generic_gt_sysinfo[GUC_GENERIC_GT_SYSINFO_DOORBELL_COUNT_PER_SQIDI] = + ((distdbreg >> GEN12_DOORBELLS_PER_SQIDI_SHIFT) & + GEN12_DOORBELLS_PER_SQIDI) + 1; + } - blob->system_info.vdbox_enable_mask = VDBOX_MASK(gt); - blob->system_info.vebox_enable_mask = VEBOX_MASK(gt); - blob->system_info.vdbox_sfc_support_mask = gt->info.vdbox_sfc_access; + guc_mapping_table_init(guc_to_gt(guc), &blob->system_info); base = intel_guc_ggtt_offset(guc, guc->ads_vma); @@ -118,11 +182,12 @@ static void __guc_ads_init(struct intel_guc *guc) /* ADS */ blob->ads.scheduler_policies = base + ptr_offset(blob, policies); - blob->ads.reg_state_buffer = base + ptr_offset(blob, reg_state_buffer); - blob->ads.reg_state_addr = base + ptr_offset(blob, reg_state); blob->ads.gt_system_info = base + ptr_offset(blob, system_info); blob->ads.clients_info = base + ptr_offset(blob, clients_info); + /* Private Data */ + blob->ads.private_data = base + guc_ads_private_data_offset(guc); + i915_gem_object_flush_map(guc->ads_vma->obj); } @@ -135,14 +200,15 @@ static void __guc_ads_init(struct intel_guc *guc) */ int intel_guc_ads_create(struct intel_guc *guc) { - const u32 size = PAGE_ALIGN(sizeof(struct __guc_ads_blob)); + u32 size; int ret; GEM_BUG_ON(guc->ads_vma); + size = guc_ads_blob_size(guc); + ret = intel_guc_allocate_and_map_vma(guc, size, &guc->ads_vma, (void **)&guc->ads_blob); - if (ret) return ret; @@ -156,6 +222,18 @@ void intel_guc_ads_destroy(struct intel_guc *guc) i915_vma_unpin_and_release(&guc->ads_vma, I915_VMA_RELEASE_MAP); } +static void guc_ads_private_data_reset(struct intel_guc *guc) +{ + u32 size; + + size = guc_ads_private_data_size(guc); + if (!size) + return; + + memset((void *)guc->ads_blob + guc_ads_private_data_offset(guc), 0, + size); +} + /** * intel_guc_ads_reset() - prepares GuC Additional Data Struct for reuse * @guc: intel_guc struct @@ -168,5 +246,8 @@ void intel_guc_ads_reset(struct intel_guc *guc) { if (!guc->ads_vma) return; + __guc_ads_init(guc); + + guc_ads_private_data_reset(guc); } diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_fwif.h b/drivers/gpu/drm/i915/gt/uc/intel_guc_fwif.h index a6b733c146c9..79c560d9c0b6 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_guc_fwif.h +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_fwif.h @@ -26,8 +26,8 @@ #define GUC_VIDEO_ENGINE2 4 #define GUC_MAX_ENGINES_NUM (GUC_VIDEO_ENGINE2 + 1) -#define GUC_MAX_ENGINE_CLASSES 5 -#define GUC_MAX_INSTANCES_PER_CLASS 16 +#define GUC_MAX_ENGINE_CLASSES 16 +#define GUC_MAX_INSTANCES_PER_CLASS 32 #define GUC_DOORBELL_INVALID 256 @@ -62,12 +62,7 @@ #define GUC_STAGE_DESC_ATTR_PCH BIT(6) #define GUC_STAGE_DESC_ATTR_TERMINATED BIT(7) -/* New GuC control data */ -#define GUC_CTL_CTXINFO 0 -#define GUC_CTL_CTXNUM_IN16_SHIFT 0 -#define GUC_CTL_BASE_ADDR_SHIFT 12 - -#define GUC_CTL_LOG_PARAMS 1 +#define GUC_CTL_LOG_PARAMS 0 #define GUC_LOG_VALID (1 << 0) #define GUC_LOG_NOTIFY_ON_HALF_FULL (1 << 1) #define GUC_LOG_ALLOC_IN_MEGABYTE (1 << 3) @@ -79,11 +74,11 @@ #define GUC_LOG_ISR_MASK (0x7 << GUC_LOG_ISR_SHIFT) #define GUC_LOG_BUF_ADDR_SHIFT 12 -#define GUC_CTL_WA 2 -#define GUC_CTL_FEATURE 3 +#define GUC_CTL_WA 1 +#define GUC_CTL_FEATURE 2 #define GUC_CTL_DISABLE_SCHEDULER (1 << 14) -#define GUC_CTL_DEBUG 4 +#define GUC_CTL_DEBUG 3 #define GUC_LOG_VERBOSITY_SHIFT 0 #define GUC_LOG_VERBOSITY_LOW (0 << GUC_LOG_VERBOSITY_SHIFT) #define GUC_LOG_VERBOSITY_MED (1 << GUC_LOG_VERBOSITY_SHIFT) @@ -97,12 +92,37 @@ #define GUC_LOG_DISABLED (1 << 6) #define GUC_PROFILE_ENABLED (1 << 7) -#define GUC_CTL_ADS 5 +#define GUC_CTL_ADS 4 #define GUC_ADS_ADDR_SHIFT 1 #define GUC_ADS_ADDR_MASK (0xFFFFF << GUC_ADS_ADDR_SHIFT) #define GUC_CTL_MAX_DWORDS (SOFT_SCRATCH_COUNT - 2) /* [1..14] */ +/* Generic GT SysInfo data types */ +#define GUC_GENERIC_GT_SYSINFO_SLICE_ENABLED 0 +#define GUC_GENERIC_GT_SYSINFO_VDBOX_SFC_SUPPORT_MASK 1 +#define GUC_GENERIC_GT_SYSINFO_DOORBELL_COUNT_PER_SQIDI 2 +#define GUC_GENERIC_GT_SYSINFO_MAX 16 + +/* + * The class goes in bits [0..2] of the GuC ID, the instance in bits [3..6]. + * Bit 7 can be used for operations that apply to all engine classes&instances. + */ +#define GUC_ENGINE_CLASS_SHIFT 0 +#define GUC_ENGINE_CLASS_MASK (0x7 << GUC_ENGINE_CLASS_SHIFT) +#define GUC_ENGINE_INSTANCE_SHIFT 3 +#define GUC_ENGINE_INSTANCE_MASK (0xf << GUC_ENGINE_INSTANCE_SHIFT) +#define GUC_ENGINE_ALL_INSTANCES BIT(7) + +#define MAKE_GUC_ID(class, instance) \ + (((class) << GUC_ENGINE_CLASS_SHIFT) | \ + ((instance) << GUC_ENGINE_INSTANCE_SHIFT)) + +#define GUC_ID_TO_ENGINE_CLASS(guc_id) \ + (((guc_id) & GUC_ENGINE_CLASS_MASK) >> GUC_ENGINE_CLASS_SHIFT) +#define GUC_ID_TO_ENGINE_INSTANCE(guc_id) \ + (((guc_id) & GUC_ENGINE_INSTANCE_MASK) >> GUC_ENGINE_INSTANCE_SHIFT) + /* Work item for submitting workloads into work queue of GuC. */ struct guc_wq_item { u32 header; @@ -336,11 +356,6 @@ struct guc_policies { } __packed; /* GuC MMIO reg state struct */ - - -#define GUC_REGSET_MAX_REGISTERS 64 -#define GUC_S3_SAVE_SPACE_PAGES 10 - struct guc_mmio_reg { u32 offset; u32 value; @@ -348,28 +363,18 @@ struct guc_mmio_reg { #define GUC_REGSET_MASKED (1 << 0) } __packed; -struct guc_mmio_regset { - struct guc_mmio_reg registers[GUC_REGSET_MAX_REGISTERS]; - u32 values_valid; - u32 number_of_registers; -} __packed; - /* GuC register sets */ -struct guc_mmio_reg_state { - struct guc_mmio_regset engine_reg[GUC_MAX_ENGINE_CLASSES][GUC_MAX_INSTANCES_PER_CLASS]; - u32 reserved[98]; +struct guc_mmio_reg_set { + u32 address; + u16 count; + u16 reserved; } __packed; /* HW info */ struct guc_gt_system_info { - u32 slice_enabled; - u32 rcs_enabled; - u32 reserved0; - u32 bcs_enabled; - u32 vdbox_enable_mask; - u32 vdbox_sfc_support_mask; - u32 vebox_enable_mask; - u32 reserved[9]; + u8 mapping_table[GUC_MAX_ENGINE_CLASSES][GUC_MAX_INSTANCES_PER_CLASS]; + u32 engine_enabled_masks[GUC_MAX_ENGINE_CLASSES]; + u32 generic_gt_sysinfo[GUC_GENERIC_GT_SYSINFO_MAX]; } __packed; /* Clients info */ @@ -390,15 +395,16 @@ struct guc_clients_info { /* GuC Additional Data Struct */ struct guc_ads { - u32 reg_state_addr; - u32 reg_state_buffer; + struct guc_mmio_reg_set reg_state_list[GUC_MAX_ENGINE_CLASSES][GUC_MAX_INSTANCES_PER_CLASS]; + u32 reserved0; u32 scheduler_policies; u32 gt_system_info; u32 clients_info; u32 control_data; u32 golden_context_lrca[GUC_MAX_ENGINE_CLASSES]; u32 eng_state_size[GUC_MAX_ENGINE_CLASSES]; - u32 reserved[16]; + u32 private_data; + u32 reserved[15]; } __packed; /* GuC logging structures */ diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_reg.h b/drivers/gpu/drm/i915/gt/uc/intel_guc_reg.h index 1949346e714e..b37fc2ffaef2 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_guc_reg.h +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_reg.h @@ -118,6 +118,11 @@ struct guc_doorbell_info { #define GEN8_DRB_VALID (1<<0) #define GEN8_DRBREGU(x) _MMIO(0x1000 + (x) * 8 + 4) +#define GEN12_DIST_DBS_POPULATED _MMIO(0xd08) +#define GEN12_DOORBELLS_PER_SQIDI_SHIFT 16 +#define GEN12_DOORBELLS_PER_SQIDI (0xff) +#define GEN12_SQIDIS_DOORBELL_EXIST (0xffff) + #define DE_GUCRMR _MMIO(0x44054) #define GUC_BCS_RCS_IER _MMIO(0xC550) diff --git a/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c b/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c index 80e8b6c3bc8c..ee4ac3922277 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c +++ b/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c @@ -44,23 +44,19 @@ void intel_uc_fw_change_status(struct intel_uc_fw *uc_fw, * List of required GuC and HuC binaries per-platform. * Must be ordered based on platform + revid, from newer to older. * - * TGL 35.2 is interface-compatible with 33.0 for previous Gens. The deltas - * between 33.0 and 35.2 are only related to new additions to support new Gen12 - * features. - * * Note that RKL uses the same firmware as TGL. */ #define INTEL_UC_FIRMWARE_DEFS(fw_def, guc_def, huc_def) \ - fw_def(ROCKETLAKE, 0, guc_def(tgl, 35, 2, 0), huc_def(tgl, 7, 5, 0)) \ - fw_def(TIGERLAKE, 0, guc_def(tgl, 35, 2, 0), huc_def(tgl, 7, 5, 0)) \ - fw_def(ELKHARTLAKE, 0, guc_def(ehl, 33, 0, 4), huc_def(ehl, 9, 0, 0)) \ - fw_def(ICELAKE, 0, guc_def(icl, 33, 0, 0), huc_def(icl, 9, 0, 0)) \ - fw_def(COMETLAKE, 5, guc_def(cml, 33, 0, 0), huc_def(cml, 4, 0, 0)) \ - fw_def(COFFEELAKE, 0, guc_def(kbl, 33, 0, 0), huc_def(kbl, 4, 0, 0)) \ - fw_def(GEMINILAKE, 0, guc_def(glk, 33, 0, 0), huc_def(glk, 4, 0, 0)) \ - fw_def(KABYLAKE, 0, guc_def(kbl, 33, 0, 0), huc_def(kbl, 4, 0, 0)) \ - fw_def(BROXTON, 0, guc_def(bxt, 33, 0, 0), huc_def(bxt, 2, 0, 0)) \ - fw_def(SKYLAKE, 0, guc_def(skl, 33, 0, 0), huc_def(skl, 2, 0, 0)) + fw_def(ROCKETLAKE, 0, guc_def(tgl, 49, 0, 1), huc_def(tgl, 7, 5, 0)) \ + fw_def(TIGERLAKE, 0, guc_def(tgl, 49, 0, 1), huc_def(tgl, 7, 5, 0)) \ + fw_def(ELKHARTLAKE, 0, guc_def(ehl, 49, 0, 1), huc_def(ehl, 9, 0, 0)) \ + fw_def(ICELAKE, 0, guc_def(icl, 49, 0, 1), huc_def(icl, 9, 0, 0)) \ + fw_def(COMETLAKE, 5, guc_def(cml, 49, 0, 1), huc_def(cml, 4, 0, 0)) \ + fw_def(COFFEELAKE, 0, guc_def(kbl, 49, 0, 1), huc_def(kbl, 4, 0, 0)) \ + fw_def(GEMINILAKE, 0, guc_def(glk, 49, 0, 1), huc_def(glk, 4, 0, 0)) \ + fw_def(KABYLAKE, 0, guc_def(kbl, 49, 0, 1), huc_def(kbl, 4, 0, 0)) \ + fw_def(BROXTON, 0, guc_def(bxt, 49, 0, 1), huc_def(bxt, 2, 0, 0)) \ + fw_def(SKYLAKE, 0, guc_def(skl, 49, 0, 1), huc_def(skl, 2, 0, 0)) #define __MAKE_UC_FW_PATH(prefix_, name_, major_, minor_, patch_) \ "i915/" \ @@ -371,6 +367,9 @@ int intel_uc_fw_fetch(struct intel_uc_fw *uc_fw) } } + if (uc_fw->type == INTEL_UC_FW_TYPE_GUC) + uc_fw->private_data_size = css->private_data_size; + obj = i915_gem_object_create_shmem_from_data(i915, fw->data, fw->size); if (IS_ERR(obj)) { err = PTR_ERR(obj); diff --git a/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.h b/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.h index 23d3a423ac0f..99bb1fe1af66 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.h +++ b/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.h @@ -88,6 +88,8 @@ struct intel_uc_fw { u32 rsa_size; u32 ucode_size; + + u32 private_data_size; }; #ifdef CONFIG_DRM_I915_DEBUG_GUC diff --git a/drivers/gpu/drm/i915/gt/uc/intel_uc_fw_abi.h b/drivers/gpu/drm/i915/gt/uc/intel_uc_fw_abi.h index 029214cdedd5..e41ffc7a7fbc 100644 --- a/drivers/gpu/drm/i915/gt/uc/intel_uc_fw_abi.h +++ b/drivers/gpu/drm/i915/gt/uc/intel_uc_fw_abi.h @@ -69,7 +69,11 @@ struct uc_css_header { #define CSS_SW_VERSION_UC_MAJOR (0xFF << 16) #define CSS_SW_VERSION_UC_MINOR (0xFF << 8) #define CSS_SW_VERSION_UC_PATCH (0xFF << 0) - u32 reserved[14]; + u32 reserved0[13]; + union { + u32 private_data_size; /* only applies to GuC */ + u32 reserved1; + }; u32 header_info; } __packed; static_assert(sizeof(struct uc_css_header) == 128); -- 2.20.1

1 15

[PATCH OLK-5.10] x86/tsc: Make cur->adjusted values in package#1 to be the same
by LeoLiuoc 14 Jan '22

14 Jan '22

When resume from S4 on Zhaoxin 2 packages platform that support X86_FEATURE_TSC_ADJUST, the following warning messages appear: [ 327.445302] [Firmware Bug]: TSC ADJUST differs: CPU15 45960750 --> 78394770. Restoring [ 329.209120] [Firmware Bug]: TSC ADJUST differs: CPU14 45960750 --> 78394770. Restoring [ 329.209128] [Firmware Bug]: TSC ADJUST differs: CPU13 45960750 --> 78394770. Restoring [ 329.209138] [Firmware Bug]: TSC ADJUST differs: CPU12 45960750 --> 78394770. Restoring [ 329.209151] [Firmware Bug]: TSC ADJUST differs: CPU11 45960750 --> 78394770. Restoring [ 329.209160] [Firmware Bug]: TSC ADJUST differs: CPU10 45960750 --> 78394770. Restoring [ 329.209169] [Firmware Bug]: TSC ADJUST differs: CPU9 45960750 --> 78394770. Restoring The reason is: Step 1: Bring up. TSC is sync after bring up with following settings: MSR 0x3b cur->adjusted Package#0 CPU 0-7 0 0 Package#1 first CPU value1 value1 Package#1 non-first CPU value1 value1 Step 2: Suspend to S4. Settings in Step 1 are not changed in this Step. Step 3: Bring up caused by S4 wake up event. TSC is sync when bring up with following settings: MSR 0x3b cur->adjusted Package#0 CPU 0-7 0 0 Package#1 first CPU value2 value2 Package#1 non-first CPU value2 value2 Step 4: Resume from S4. When resuming from S4, Current TSC synchronous mechanism cause following settings: MSR 0x3b cur->adjusted Package#0 CPU 0-7 0 0 Package#1 first CPU value2 value2 Package#1 non-first CPU value2 value1 In these Steps, value1 != 0 and value2 != value1. In Step4, as function tsc_store_and_check_tsc_adjust() do, when the value of MSR 0x3b on the non-first online CPU in package#1 is equal to the value of cur->adjusted on the first online CPU in the same package, the cur->adjusted value on this non-first online CPU will hold the old value1. This cause function tsc_verify_tsc_adjust() set the value of MSR 0x3b on the non-first online CPUs in the package#1 to the old value1 and print the beginning warning messages. Fix it by setting cur->adjusted value on the non-first online CPU in a package to the value of MSR 0x3b on the same CPU when they are not equal. Signed-off-by: LeoLiu-oc <LeoLiu-oc(a)zhaoxin.com> --- arch/x86/kernel/tsc_sync.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/x86/kernel/tsc_sync.c b/arch/x86/kernel/tsc_sync.c index 3d3c761eb..19a0bed9b 100644 --- a/arch/x86/kernel/tsc_sync.c +++ b/arch/x86/kernel/tsc_sync.c @@ -190,6 +190,8 @@ bool tsc_store_and_check_tsc_adjust(bool bootcpu) if (bootval != ref->adjusted) { cur->adjusted = ref->adjusted; wrmsrl(MSR_IA32_TSC_ADJUST, ref->adjusted); + } else if (cur->adjusted != bootval) { + cur->adjusted = bootval; } /* * We have the TSCs forced to be in sync on this package. Skip sync -- 2.20.1

1 0

[PATCH OLK-5.10] Add support for PxSCT.LPM set based on actual LPM circumstances
by LeoLiuoc 14 Jan '22

14 Jan '22

AHCI Spec shows that the PxSCTL.IPM field in each port must be programmed to disallow device initiated LPM requests when HBA can not support transitions to the LPM state. The current LPM driver has no restrictions on LPM transitions when enabling device initiated LPM. Signed-off-by: LeoLiu-oc <LeoLiu-oc(a)zhaoxin.com> --- drivers/ata/libata-sata.c | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-) diff --git a/drivers/ata/libata-sata.c b/drivers/ata/libata-sata.c index c16423e44..7e1296b55 100644 --- a/drivers/ata/libata-sata.c +++ b/drivers/ata/libata-sata.c @@ -394,9 +394,18 @@ int sata_link_scr_lpm(struct ata_link *link, enum ata_lpm_policy policy, case ATA_LPM_MED_POWER_WITH_DIPM: case ATA_LPM_MIN_POWER_WITH_PARTIAL: case ATA_LPM_MIN_POWER: - if (ata_link_nr_enabled(link) > 0) + if (ata_link_nr_enabled(link) > 0) { /* no restrictions on LPM transitions */ scontrol &= ~(0x7 << 8); + /* if Host does not support partial, then disallows it, + * the same for slumber + */ + if (!(link->ap->host->flags & ATA_HOST_PART)) + scontrol |= (0x1 << 8); + + if (!(link->ap->host->flags & ATA_HOST_SSC)) + scontrol |= (0x2 << 8); + } else { /* empty port, power off */ scontrol &= ~0xf; -- 2.20.1

1 0

[PATCH OLK-5.10] Add support for PxSCT.LPM setting based on actual LPM circumstances
by LeoLiuoc 14 Jan '22

14 Jan '22

In AHCI spec, PhyRdy Change Interrupt cannot coexist with Link Power Management (LPM). While LPM driver would disable PhyRdy Change interrupt without checking whether HBA supports LPM or not before enabling it. Therefore, if enabling LPM from LPM driver by default, we can not disable it from BIOS without influence Hotplug out capability. Signed-off-by: LeoLiu-oc <LeoLiu-oc(a)zhaoxin.com> --- drivers/ata/ahci.c | 9 +++++++++ drivers/ata/libata-eh.c | 4 ++++ include/linux/libata.h | 4 ++++ 3 files changed, 17 insertions(+) diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c index f8059eed3..2d1ac1b5b 100644 --- a/drivers/ata/ahci.c +++ b/drivers/ata/ahci.c @@ -1883,6 +1883,15 @@ static int ahci_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) else dev_info(&pdev->dev, "SSS flag set, parallel bus scan disabled\n"); + if (hpriv->cap & HOST_CAP_PART) + host->flags |= ATA_HOST_PART; + + if (hpriv->cap & HOST_CAP_SSC) + host->flags |= ATA_HOST_SSC; + + if (hpriv->cap2 & HOST_CAP2_SDS) + host->flags |= ATA_HOST_DEVSLP; + if (pi.flags & ATA_FLAG_EM) ahci_reset_em(host); diff --git a/drivers/ata/libata-eh.c b/drivers/ata/libata-eh.c index b6f92050e..ff6e660d6 100644 --- a/drivers/ata/libata-eh.c +++ b/drivers/ata/libata-eh.c @@ -3237,6 +3237,10 @@ static int ata_eh_set_lpm(struct ata_link *link, enum ata_lpm_policy policy, unsigned int err_mask; int rc; + /* if Host does not support lpm, then sets no LPM flags*/ + if (!(ap->host->flags & (ATA_HOST_PART | ATA_HOST_SSC | ATA_HOST_DEVSLP))) + link->flags |= ATA_LFLAG_NO_LPM; + /* if the link or host doesn't do LPM, noop */ if (!IS_ENABLED(CONFIG_SATA_HOST) || (link->flags & ATA_LFLAG_NO_LPM) || (ap && !ap->ops->set_lpm)) diff --git a/include/linux/libata.h b/include/linux/libata.h index 5f550eb27..81124ddc2 100644 --- a/include/linux/libata.h +++ b/include/linux/libata.h @@ -260,6 +260,10 @@ enum { ATA_HOST_PARALLEL_SCAN = (1 << 2), /* Ports on this host can be scanned in parallel */ ATA_HOST_IGNORE_ATA = (1 << 3), /* Ignore ATA devices on this host. */ + ATA_HOST_PART = (1 << 4), /* Host support partial.*/ + ATA_HOST_SSC = (1 << 5), /* Host support slumber.*/ + ATA_HOST_DEVSLP = (1 << 6), /* Host support devslp.*/ + /* bits 24:31 of host->flags are reserved for LLD specific flags */ /* various lengths of time */ -- 2.20.1

1 0

[PATCH OLK-5.10] Fix some bugs like plugin support and sata link stability when user enable ahci RTD3
by LeoLiuoc 14 Jan '22

14 Jan '22

The kernel driver now will disable interrupt when port is suspended, causing plugin not work. It is useful to make plugin work，when the plugin interrupt is enabled with power management status. Adding pm_request_resume() is to reume port for plugin or PME signal waking up controller which in D3. with the AHCI controller frequently enters into D3 and leaves from D3, the identify cmd may be timeout when controller resumes and establishes a connect with the device.it is effective to delay 10ms between controller resume and port resume，with link’s smooth transition. with non power management request and power management competing with each other in queue, it is often found that block IO hang 120s when system disk is suspending or resuming.it is now guaranteed that PM requests will enter the queue no matter other non-PM requests are waiting. Increase the pm_only counter before checking whether any non-PM blk_queue_enter() calls are in progress. Meanwhile, the new blk_pm_request_resume() call is necessary to occur during request assigned to a queue when device is suspended. Signed-off-by: LeoLiu-oc <LeoLiu-oc(a)zhaoxin.com> --- drivers/ata/ahci.c | 13 +++++++++++++ drivers/ata/libahci.c | 14 ++++++++++++++ 2 files changed, 27 insertions(+) diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c index 33192a8f6..617cfcb2b 100644 --- a/drivers/ata/ahci.c +++ b/drivers/ata/ahci.c @@ -868,6 +868,19 @@ static int ahci_pci_device_runtime_resume(struct device *dev) if (rc) return rc; ahci_pci_init_controller(host); + + /* controller delay for 10ms when being reusmed && + * port resume for Zx platform + */ + if (pdev->vendor == PCI_VENDOR_ID_ZHAOXIN) { + ata_msleep(NULL, 10); + for (rc = 0; rc < host->n_ports; rc++) { + struct ata_port *ap = host->ports[rc]; + + pm_request_resume(&ap->tdev); + } + } + return 0; } diff --git a/drivers/ata/libahci.c b/drivers/ata/libahci.c index fec2e9754..a37eac335 100644 --- a/drivers/ata/libahci.c +++ b/drivers/ata/libahci.c @@ -826,6 +826,12 @@ static void ahci_power_down(struct ata_port *ap) void __iomem *port_mmio = ahci_port_base(ap); u32 cmd, scontrol; + /* port suspended enable Plugin intr for Zx platform */ + struct pci_dev *pdev = to_pci_dev(ap->host->dev); + + if ((pdev->vendor == PCI_VENDOR_ID_ZHAOXIN) && (!ap->link.device->sdev)) + writel(PORT_IRQ_CONNECT, port_mmio + PORT_IRQ_MASK); + if (!(hpriv->cap & HOST_CAP_SSS)) return; @@ -1703,6 +1709,7 @@ static void ahci_error_intr(struct ata_port *ap, u32 irq_stat) struct ata_eh_info *active_ehi; bool fbs_need_dec = false; u32 serror; + struct pci_dev *pdev; /* determine active link with error */ if (pp->fbs_enabled) { @@ -1791,6 +1798,13 @@ static void ahci_error_intr(struct ata_port *ap, u32 irq_stat) ata_ehi_push_desc(host_ehi, "%s", irq_stat & PORT_IRQ_CONNECT ? "connection status changed" : "PHY RDY changed"); + /* when plugin intr happen,now resume suspended port for Zx platform */ + pdev = to_pci_dev(ap->host->dev); + if ((pdev->vendor == PCI_VENDOR_ID_ZHAOXIN) && + (ap->pflags & ATA_PFLAG_SUSPENDED)) { + pm_request_resume(&ap->tdev); + return; + } } /* okay, let's hand over to EH */ -- 2.20.1

1 0

[PATCH OLK-5.10] EHCI:Clear wakeup signal locked in S0 state when device plug in
by LeoLiuoc 14 Jan '22

14 Jan '22

If we plug in a LS/FS device on USB2 port of EHCI, it will latch a wakeup signal in EHCI internal. This is a bug of EHCI for Some project of ZhaoXin. If enable EHCI runtime suspend and no device attach. PM core will let EHCI go to D3 to save power. However, once EHCI go to D3, it will release wakeup signal that latched on device connect to port during S0. Which will generate a SCI interrupt and bring EHCI to D0. But without device connect, EHCI will go to D3 again. So, there is suspend-resume loop and generate SCI interrupt Continuously. In order to fix this issues, we need to clear the wakeup signal latched in EHCI when EHCI suspend function is called. Signed-off-by: LeoLiu-oc <LeoLiu-oc(a)zhaoxin.com> --- drivers/pci/pci-driver.c | 6 +++++- drivers/usb/host/ehci-hcd.c | 21 +++++++++++++++++++++ drivers/usb/host/ehci-pci.c | 4 ++++ drivers/usb/host/ehci.h | 1 + 4 files changed, 31 insertions(+), 1 deletion(-) diff --git a/drivers/pci/pci-driver.c b/drivers/pci/pci-driver.c index 8b587fc97..c33d7e0a6 100644 --- a/drivers/pci/pci-driver.c +++ b/drivers/pci/pci-driver.c @@ -518,7 +518,11 @@ static int pci_restore_standard_config(struct pci_dev *pci_dev) } pci_restore_state(pci_dev); - pci_pme_restore(pci_dev); + if (!((pci_dev->vendor == PCI_VENDOR_ID_ZHAOXIN) && + (pci_dev->device == 0x3104) && + ((pci_dev->revision & 0xf0) == 0x90)) || + !(pci_dev->class == PCI_CLASS_SERIAL_USB_EHCI)) + pci_pme_restore(pci_dev); return 0; } diff --git a/drivers/usb/host/ehci-hcd.c b/drivers/usb/host/ehci-hcd.c index 8aff19ff8..586e2735d 100644 --- a/drivers/usb/host/ehci-hcd.c +++ b/drivers/usb/host/ehci-hcd.c @@ -1142,6 +1142,27 @@ int ehci_suspend(struct usb_hcd *hcd, bool do_wakeup) return -EBUSY; } + /*clear wakeup signal locked in S0 state when device plug in*/ + if (ehci->zx_wakeup_clear == 1) { + u32 __iomem *reg = &ehci->regs->port_status[4]; + u32 t1 = ehci_readl(ehci, reg); + + t1 &= (u32)~0xf0000; + t1 |= PORT_TEST_FORCE; + ehci_writel(ehci, t1, reg); + t1 = ehci_readl(ehci, reg); + usleep_range(1000, 2000); + t1 &= (u32)~0xf0000; + ehci_writel(ehci, t1, reg); + usleep_range(1000, 2000); + t1 = ehci_readl(ehci, reg); + ehci_writel(ehci, t1 | PORT_CSC, reg); + udelay(500); + t1 = ehci_readl(ehci, &ehci->regs->status); + ehci_writel(ehci, t1 & STS_PCD, &ehci->regs->status); + ehci_readl(ehci, &ehci->regs->status); + } + return 0; } EXPORT_SYMBOL_GPL(ehci_suspend); diff --git a/drivers/usb/host/ehci-pci.c b/drivers/usb/host/ehci-pci.c index e87cf3a00..a5e27deda 100644 --- a/drivers/usb/host/ehci-pci.c +++ b/drivers/usb/host/ehci-pci.c @@ -222,6 +222,10 @@ static int ehci_pci_setup(struct usb_hcd *hcd) ehci->has_synopsys_hc_bug = 1; } break; + case PCI_VENDOR_ID_ZHAOXIN: + if (pdev->device == 0x3104 && (pdev->revision & 0xf0) == 0x90) + ehci->zx_wakeup_clear = 1; + break; } /* optional debug port, normally in the first BAR */ diff --git a/drivers/usb/host/ehci.h b/drivers/usb/host/ehci.h index 59fd523c5..8da080a54 100644 --- a/drivers/usb/host/ehci.h +++ b/drivers/usb/host/ehci.h @@ -219,6 +219,7 @@ struct ehci_hcd { /* one per controller */ unsigned need_oc_pp_cycle:1; /* MPC834X port power */ unsigned imx28_write_fix:1; /* For Freescale i.MX28 */ unsigned is_aspeed:1; + unsigned zx_wakeup_clear:1; /* required for usb32 quirk */ #define OHCI_CTRL_HCFS (3 << 6) -- 2.25.1

1 0

[PATCH OLK-5.10] XHCI:Fix some device identify fail when enable xHCI runtime suspend
by LeoLiuoc 14 Jan '22

14 Jan '22

If plug out device form xhci with runtime suspend enabled. On the one hand, driver will disconnect this device and send disabled slot command to xhci. On the other hand, without no device connect to xhci, PM core will call xhci suspend function to let xhci go to D3 to save power. However there is a temporal competition to get xhci lock between disable slot command interrupt and xhci suspend. If xhci suspend function get xhci lock first, then this function will clear xhci command ring. It will get error command trb when driver to handle disable slot command interrupt. This is a serious error for driver and driver will cleanup xhci. So，any device connect to this xhci port again will not be recognized. In order to fix this issues, we let disable slot command interrupt ISR to get xhci lock first. So, add a delay in xhci suspend function before to get xhci lock. Signed-off-by: LeoLiu-oc <LeoLiu-oc(a)zhaoxin.com> --- drivers/usb/host/xhci-pci.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/usb/host/xhci-pci.c b/drivers/usb/host/xhci-pci.c index 583a8e895..83aa8e179 100644 --- a/drivers/usb/host/xhci-pci.c +++ b/drivers/usb/host/xhci-pci.c @@ -272,6 +272,8 @@ static void xhci_pci_quirks(struct device *dev, struct xhci_hcd *xhci) xhci->quirks |= XHCI_LPM_SUPPORT; xhci->quirks |= XHCI_ZHAOXIN_HOST; } + if (pdev->vendor == PCI_VENDOR_ID_ZHAOXIN) + xhci->quirks |= XHCI_SUSPEND_DELAY; /* See https://bugzilla.kernel.org/show_bug.cgi?id=79511 */ if (pdev->vendor == PCI_VENDOR_ID_VIA && -- 2.20.1

1 0

[PATCH OLK-5.10] rtc: Fix set RTC time delay 500ms on some Zhaoxin SOCs
by LeoLiuoc 14 Jan '22

14 Jan '22

When the RTC divider is changed from reset to an operating time base, the first update cycle should be 500ms later. But on some Zhaoxin SOCs, this first update cycle is one second later. So set RTC time on these Zhaoxin SOCs will causing 500ms delay. Skip setup RTC divider on these SOCs in mc146818_set_time to fix it. Signed-off-by: LeoLiu-oc <LeoLiu-oc(a)zhaoxin.com> --- drivers/rtc/rtc-mc146818-lib.c | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) diff --git a/drivers/rtc/rtc-mc146818-lib.c b/drivers/rtc/rtc-mc146818-lib.c index 2ecd8752b..96d9d0219 100644 --- a/drivers/rtc/rtc-mc146818-lib.c +++ b/drivers/rtc/rtc-mc146818-lib.c @@ -171,8 +171,17 @@ int mc146818_set_time(struct rtc_time *time) save_control = CMOS_READ(RTC_CONTROL); CMOS_WRITE((save_control|RTC_SET), RTC_CONTROL); +#ifdef CONFIG_X86 + if (!((boot_cpu_data.x86_vendor == X86_VENDOR_CENTAUR || + boot_cpu_data.x86_vendor == X86_VENDOR_ZHAOXIN) && + (boot_cpu_data.x86 <= 7 && boot_cpu_data.x86_model <= 59))) { + save_freq_select = CMOS_READ(RTC_FREQ_SELECT); + CMOS_WRITE((save_freq_select|RTC_DIV_RESET2), RTC_FREQ_SELECT); + } +#else save_freq_select = CMOS_READ(RTC_FREQ_SELECT); CMOS_WRITE((save_freq_select|RTC_DIV_RESET2), RTC_FREQ_SELECT); +#endif #ifdef CONFIG_MACH_DECSTATION CMOS_WRITE(real_yrs, RTC_DEC_YEAR); @@ -190,7 +199,14 @@ int mc146818_set_time(struct rtc_time *time) #endif CMOS_WRITE(save_control, RTC_CONTROL); +#ifdef CONFIG_X86 + if (!((boot_cpu_data.x86_vendor == X86_VENDOR_CENTAUR || + boot_cpu_data.x86_vendor == X86_VENDOR_ZHAOXIN) && + (boot_cpu_data.x86 <= 7 && boot_cpu_data.x86_model <= 59))) + CMOS_WRITE(save_freq_select, RTC_FREQ_SELECT); +#else CMOS_WRITE(save_freq_select, RTC_FREQ_SELECT); +#endif spin_unlock_irqrestore(&rtc_lock, flags); -- 2.20.1

1 0

[PATCH OLK-5.10 v2 1/3] arm64: Add memmap parameter for pmem
by Zhuling 14 Jan '22

14 Jan '22

hulk inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4O31I?from=project-issue CVE: NA Add "!" in memmap parameter, use memmap(memmap=nn[KMG]!ss[KMG]) reserve memory for pmem which register persistent memory in arm64. Signed-off-by: Zhuling <zhuling8(a)huawei.com> --- Documentation/admin-guide/kernel-parameters.txt | 9 +++++++-- arch/arm64/mm/init.c | 6 ++++++ 2 files changed, 13 insertions(+), 2 deletions(-) diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt index 64be32b..ab42db4 100644 --- a/Documentation/admin-guide/kernel-parameters.txt +++ b/Documentation/admin-guide/kernel-parameters.txt @@ -2829,10 +2829,15 @@ will be eaten. memmap=nn[KMG]!ss[KMG] - [KNL,X86] Mark specific memory as protected. + [KNL,X86,ARM64] Mark specific memory as protected. Region of memory to be used, from ss to ss+nn. - The memory region may be marked as e820 type 12 (0xc) + [X86]The memory region may be marked as e820 type 12 (0xc) and is NVDIMM or ADR memory. + [ARM64]Reserve memory for persistent storage when the kernel + restart or update. the data in PMEM will not be lost and can + be loaded faster + Example: + memmap=100K!0x1a0000000 memmap=<size>%<offset>-<oldtype>+<newtype> [KNL,ACPI] Convert memory within the specified region diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c index 6ebfabd..b6512dc 100644 --- a/arch/arm64/mm/init.c +++ b/arch/arm64/mm/init.c @@ -56,6 +56,8 @@ s64 memstart_addr __ro_after_init = -1; EXPORT_SYMBOL(memstart_addr); +phys_addr_t start_at, mem_size; + #ifdef CONFIG_PIN_MEMORY struct resource pin_memory_resource = { .name = "Pin memory", @@ -442,6 +444,10 @@ static int __init parse_memmap_one(char *p) start_at = memparse(p + 1, &p); memblock_reserve(start_at, mem_size); memblock_mark_memmap(start_at, mem_size); + } else if (*p == '!') { + start_at = memparse(p+1, &p); + pmem_start = start_at; + pmem_size = mem_size; } else pr_info("Unrecognized memmap option, please check the parameter.\n"); -- 2.9.5

1 2

openEuler kernel技术分享第十七期 & 双周例会
by openEuler conference 14 Jan '22

14 Jan '22

您好！ Kernel SIG 邀请您参加 2021-12-31 14:00 召开的ZOOM会议(自动录制) 会议主题：openEuler kernel技术分享第十七期 & 双周例会会议内容： 14:10-15:30 openEuler kernel 技术分享第17期 - cgroup介绍会议链接：https://us06web.zoom.us/j/83118407147?pwd=elVzMVhGc2YzaDQ1TGorN2tPa1NhZz09 温馨提醒：建议接入会议后修改参会人的姓名，也可以使用您在gitee.com的ID 更多资讯尽在：https://openeuler.org/zh/ Hello! openEuler Kernel SIG invites you to attend the ZOOM conference(auto recording) will be held at 2021-12-31 14:00, The subject of the conference is openEuler kernel技术分享第十七期 & 双周例会, Summary: 14:10-15:30 openEuler kernel 技术分享第17期 - cgroup介绍 You can join the meeting at https://us06web.zoom.us/j/83118407147?pwd=elVzMVhGc2YzaDQ1TGorN2tPa1NhZz09. Note: You are advised to change the participant name after joining the conference or use your ID at gitee.com. More information: https://openeuler.org/en/

3 2

openEuler kernel SIG 双周例会
by openEuler conference 14 Jan '22

14 Jan '22

您好！ Kernel SIG 邀请您参加 2022-01-14 14:00 召开的ZOOM会议(自动录制) 会议主题：openEuler kernel SIG 双周例会会议内容：议题：ebpf增加机制以支持非主线helper从而保证一定的兼容性会议链接：https://us06web.zoom.us/j/86252266487?pwd=VjRNTVJzdkVwaXl6K2V1VUcxTWpEQT09 温馨提醒：建议接入会议后修改参会人的姓名，也可以使用您在gitee.com的ID 更多资讯尽在：https://openeuler.org/zh/ Hello! openEuler Kernel SIG invites you to attend the ZOOM conference(auto recording) will be held at 2022-01-14 14:00, The subject of the conference is openEuler kernel SIG 双周例会, Summary: 议题：ebpf增加机制以支持非主线helper从而保证一定的兼容性 You can join the meeting at https://us06web.zoom.us/j/86252266487?pwd=VjRNTVJzdkVwaXl6K2V1VUcxTWpEQT09. Note: You are advised to change the participant name after joining the conference or use your ID at gitee.com. More information: https://openeuler.org/en/

2 1

[PATCH OLK-5.10] scsi: core: Stop using DRIVER_ERROR
by Zhang Yi 14 Jan '22

14 Jan '22

From: Hannes Reinecke <hare(a)suse.de> mainline inclusion from mainline-5.14-rc1 commit ced202f7bd78eb6a79c441a8b217e0f3d38bccfc category: bugfix bugzilla: 185813 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… --------------------------- Return the actual error code in __scsi_execute() (which, according to the documentation, should have happened anyway). And audit all callers to cope with negative return values from __scsi_execute() and friends. [mkp: resolve conflict and return bool] Link: https://lore.kernel.org/r/20210427083046.31620-7-hare@suse.de Reviewed-by: Bart Van Assche <bvanassche(a)acm.org> Reviewed-by: Christoph Hellwig <hch(a)lst.de> Signed-off-by: Hannes Reinecke <hare(a)suse.de> Signed-off-by: Martin K. Petersen <martin.petersen(a)oracle.com> Signed-off-by: Zhang Yi <yi.zhang(a)huawei.com> --- drivers/ata/libata-scsi.c | 8 ++++++++ drivers/scsi/ch.c | 3 ++- drivers/scsi/cxlflash/superpipe.c | 2 +- drivers/scsi/scsi.c | 2 ++ drivers/scsi/scsi_ioctl.c | 4 +++- drivers/scsi/scsi_lib.c | 15 +++++++++------ drivers/scsi/scsi_scan.c | 4 ++-- drivers/scsi/scsi_transport_spi.c | 2 +- drivers/scsi/sd.c | 15 +++++++++------ drivers/scsi/sd_zbc.c | 2 +- drivers/scsi/sr_ioctl.c | 4 ++++ drivers/scsi/ufs/ufshcd.c | 2 +- drivers/scsi/virtio_scsi.c | 2 +- include/scsi/scsi.h | 7 +++++-- 14 files changed, 49 insertions(+), 23 deletions(-) diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c index 48b8934970f3..c5129b9e3afd 100644 --- a/drivers/ata/libata-scsi.c +++ b/drivers/ata/libata-scsi.c @@ -409,6 +409,10 @@ int ata_cmd_ioctl(struct scsi_device *scsidev, void __user *arg) cmd_result = scsi_execute(scsidev, scsi_cmd, data_dir, argbuf, argsize, sensebuf, &sshdr, (10*HZ), 5, 0, 0, NULL); + if (cmd_result < 0) { + rc = cmd_result; + goto error; + } if (driver_byte(cmd_result) == DRIVER_SENSE) {/* sense data available */ u8 *desc = sensebuf + 8; cmd_result &= ~(0xFF<<24); /* DRIVER_SENSE is not an error */ @@ -490,6 +494,10 @@ int ata_task_ioctl(struct scsi_device *scsidev, void __user *arg) cmd_result = scsi_execute(scsidev, scsi_cmd, DMA_NONE, NULL, 0, sensebuf, &sshdr, (10*HZ), 5, 0, 0, NULL); + if (cmd_result < 0) { + rc = cmd_result; + goto error; + } if (driver_byte(cmd_result) == DRIVER_SENSE) {/* sense data available */ u8 *desc = sensebuf + 8; cmd_result &= ~(0xFF<<24); /* DRIVER_SENSE is not an error */ diff --git a/drivers/scsi/ch.c b/drivers/scsi/ch.c index cb74ab1ae5a4..0e7d1214c3d8 100644 --- a/drivers/scsi/ch.c +++ b/drivers/scsi/ch.c @@ -198,7 +198,8 @@ ch_do_scsi(scsi_changer *ch, unsigned char *cmd, int cmd_len, result = scsi_execute_req(ch->device, cmd, direction, buffer, buflength, &sshdr, timeout * HZ, MAX_RETRIES, NULL); - + if (result < 0) + return result; if (driver_byte(result) == DRIVER_SENSE) { if (debug) scsi_print_sense_hdr(ch->device, ch->name, &sshdr); diff --git a/drivers/scsi/cxlflash/superpipe.c b/drivers/scsi/cxlflash/superpipe.c index 5dddf67dfa24..3ebb7ac86a64 100644 --- a/drivers/scsi/cxlflash/superpipe.c +++ b/drivers/scsi/cxlflash/superpipe.c @@ -369,7 +369,7 @@ static int read_cap16(struct scsi_device *sdev, struct llun_info *lli) goto out; } - if (driver_byte(result) == DRIVER_SENSE) { + if (result > 0 && driver_byte(result) == DRIVER_SENSE) { result &= ~(0xFF<<24); /* DRIVER_SENSE is not an error */ if (result & SAM_STAT_CHECK_CONDITION) { switch (sshdr.sense_key) { diff --git a/drivers/scsi/scsi.c b/drivers/scsi/scsi.c index 6ad834d61d4c..f561caabc771 100644 --- a/drivers/scsi/scsi.c +++ b/drivers/scsi/scsi.c @@ -495,6 +495,8 @@ int scsi_report_opcode(struct scsi_device *sdev, unsigned char *buffer, result = scsi_execute_req(sdev, cmd, DMA_FROM_DEVICE, buffer, len, &sshdr, 30 * HZ, 3, NULL); + if (result < 0) + return result; if (result && scsi_sense_valid(&sshdr) && sshdr.sense_key == ILLEGAL_REQUEST && (sshdr.asc == 0x20 || sshdr.asc == 0x24) && sshdr.ascq == 0x00) diff --git a/drivers/scsi/scsi_ioctl.c b/drivers/scsi/scsi_ioctl.c index 14872c9dc78c..d34e1b41dc71 100644 --- a/drivers/scsi/scsi_ioctl.c +++ b/drivers/scsi/scsi_ioctl.c @@ -101,6 +101,8 @@ static int ioctl_internal_command(struct scsi_device *sdev, char *cmd, SCSI_LOG_IOCTL(2, sdev_printk(KERN_INFO, sdev, "Ioctl returned 0x%x\n", result)); + if (result < 0) + goto out; if (driver_byte(result) == DRIVER_SENSE && scsi_sense_valid(&sshdr)) { switch (sshdr.sense_key) { @@ -133,7 +135,7 @@ static int ioctl_internal_command(struct scsi_device *sdev, char *cmd, break; } } - +out: SCSI_LOG_IOCTL(2, sdev_printk(KERN_INFO, sdev, "IOCTL Releasing command\n")); return result; diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c index 0a265a3e3dc6..9352a309fe55 100644 --- a/drivers/scsi/scsi_lib.c +++ b/drivers/scsi/scsi_lib.c @@ -245,20 +245,23 @@ int __scsi_execute(struct scsi_device *sdev, const unsigned char *cmd, { struct request *req; struct scsi_request *rq; - int ret = DRIVER_ERROR << 24; + int ret; req = blk_get_request(sdev->request_queue, data_direction == DMA_TO_DEVICE ? REQ_OP_SCSI_OUT : REQ_OP_SCSI_IN, rq_flags & RQF_PM ? BLK_MQ_REQ_PM : 0); if (IS_ERR(req)) - return ret; - rq = scsi_req(req); + return PTR_ERR(req); - if (bufflen && blk_rq_map_kern(sdev->request_queue, req, - buffer, bufflen, GFP_NOIO)) - goto out; + rq = scsi_req(req); + if (bufflen) { + ret = blk_rq_map_kern(sdev->request_queue, req, + buffer, bufflen, GFP_NOIO); + if (ret) + goto out; + } rq->cmd_len = COMMAND_SIZE(cmd[0]); memcpy(rq->cmd, cmd, rq->cmd_len); rq->retries = retries; diff --git a/drivers/scsi/scsi_scan.c b/drivers/scsi/scsi_scan.c index 8e474b145249..57002529174f 100644 --- a/drivers/scsi/scsi_scan.c +++ b/drivers/scsi/scsi_scan.c @@ -599,7 +599,7 @@ static int scsi_probe_lun(struct scsi_device *sdev, unsigned char *inq_result, "scsi scan: INQUIRY %s with code 0x%x\n", result ? "failed" : "successful", result)); - if (result) { + if (result > 0) { /* * not-ready to ready transition [asc/ascq=0x28/0x0] * or power-on, reset [asc/ascq=0x29/0x0], continue. @@ -614,7 +614,7 @@ static int scsi_probe_lun(struct scsi_device *sdev, unsigned char *inq_result, (sshdr.ascq == 0)) continue; } - } else { + } else if (result == 0) { /* * if nothing was transferred, we try * again. It's a workaround for some USB diff --git a/drivers/scsi/scsi_transport_spi.c b/drivers/scsi/scsi_transport_spi.c index c37dd15d16d2..a9bb7ae2fafd 100644 --- a/drivers/scsi/scsi_transport_spi.c +++ b/drivers/scsi/scsi_transport_spi.c @@ -127,7 +127,7 @@ static int spi_execute(struct scsi_device *sdev, const void *cmd, REQ_FAILFAST_TRANSPORT | REQ_FAILFAST_DRIVER, RQF_PM, NULL); - if (driver_byte(result) != DRIVER_SENSE || + if (result < 0 || driver_byte(result) != DRIVER_SENSE || sshdr->sense_key != UNIT_ATTENTION) break; } diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c index 97713f595a29..3fc184e9702f 100644 --- a/drivers/scsi/sd.c +++ b/drivers/scsi/sd.c @@ -1673,7 +1673,7 @@ static unsigned int sd_check_events(struct gendisk *disk, unsigned int clearing) &sshdr); /* failed to execute TUR, assume media not present */ - if (host_byte(retval)) { + if (retval < 0 || host_byte(retval)) { set_media_not_present(sdkp); goto out; } @@ -1734,6 +1734,9 @@ static int sd_sync_cache(struct scsi_disk *sdkp, struct scsi_sense_hdr *sshdr) if (res) { sd_print_result(sdkp, "Synchronize Cache(10) failed", res); + if (res < 0) + return res; + if (driver_byte(res) == DRIVER_SENSE) sd_print_sense_hdr(sdkp, sshdr); @@ -1842,7 +1845,7 @@ static int sd_pr_command(struct block_device *bdev, u8 sa, result = scsi_execute_req(sdev, cmd, DMA_TO_DEVICE, &data, sizeof(data), &sshdr, SD_TIMEOUT, sdkp->max_retries, NULL); - if (driver_byte(result) == DRIVER_SENSE && + if (result > 0 && driver_byte(result) == DRIVER_SENSE && scsi_sense_valid(&sshdr)) { sdev_printk(KERN_INFO, sdev, "PR command failed: %d\n", result); scsi_print_sense_hdr(sdev, NULL, &sshdr); @@ -2194,7 +2197,7 @@ sd_spinup_disk(struct scsi_disk *sdkp) ((driver_byte(the_result) == DRIVER_SENSE) && sense_valid && sshdr.sense_key == UNIT_ATTENTION))); - if (driver_byte(the_result) != DRIVER_SENSE) { + if (the_result < 0 || driver_byte(the_result) != DRIVER_SENSE) { /* no sense, TUR either succeeded or failed * with a status error */ if(!spintime && !scsi_status_is_good(the_result)) { @@ -2379,7 +2382,7 @@ static int read_capacity_16(struct scsi_disk *sdkp, struct scsi_device *sdp, if (media_not_present(sdkp, &sshdr)) return -ENODEV; - if (the_result) { + if (the_result > 0) { sense_valid = scsi_sense_valid(&sshdr); if (sense_valid && sshdr.sense_key == ILLEGAL_REQUEST && @@ -2464,7 +2467,7 @@ static int read_capacity_10(struct scsi_disk *sdkp, struct scsi_device *sdp, if (media_not_present(sdkp, &sshdr)) return -ENODEV; - if (the_result) { + if (the_result > 0) { sense_valid = scsi_sense_valid(&sshdr); if (sense_valid && sshdr.sense_key == UNIT_ATTENTION && @@ -3614,7 +3617,7 @@ static int sd_start_stop_device(struct scsi_disk *sdkp, int start) SD_TIMEOUT, sdkp->max_retries, 0, RQF_PM, NULL); if (res) { sd_print_result(sdkp, "Start/Stop Unit failed", res); - if (driver_byte(res) == DRIVER_SENSE) + if (res > 0 && driver_byte(res) == DRIVER_SENSE) sd_print_sense_hdr(sdkp, &sshdr); if (scsi_sense_valid(&sshdr) && /* 0x3a is medium not present */ diff --git a/drivers/scsi/sd_zbc.c b/drivers/scsi/sd_zbc.c index 01088f333dbc..6834e029906a 100644 --- a/drivers/scsi/sd_zbc.c +++ b/drivers/scsi/sd_zbc.c @@ -116,7 +116,7 @@ static int sd_zbc_do_report_zones(struct scsi_disk *sdkp, unsigned char *buf, sd_printk(KERN_ERR, sdkp, "REPORT ZONES start lba %llu failed\n", lba); sd_print_result(sdkp, "REPORT ZONES", result); - if (driver_byte(result) == DRIVER_SENSE && + if (result > 0 && driver_byte(result) == DRIVER_SENSE && scsi_sense_valid(&sshdr)) sd_print_sense_hdr(sdkp, &sshdr); return -EIO; diff --git a/drivers/scsi/sr_ioctl.c b/drivers/scsi/sr_ioctl.c index ffcf902da390..4c1de11e69fb 100644 --- a/drivers/scsi/sr_ioctl.c +++ b/drivers/scsi/sr_ioctl.c @@ -205,6 +205,10 @@ int sr_do_ioctl(Scsi_CD *cd, struct packet_command *cgc) cgc->timeout, IOCTL_RETRIES, 0, 0, NULL); /* Minimal error checking. Ignore cases we know about, and report the rest. */ + if (result < 0) { + err = result; + goto out; + } if (driver_byte(result) != 0) { switch (sshdr->sense_key) { case UNIT_ATTENTION: diff --git a/drivers/scsi/ufs/ufshcd.c b/drivers/scsi/ufs/ufshcd.c index 930f35863cbb..b8462f54cb72 100644 --- a/drivers/scsi/ufs/ufshcd.c +++ b/drivers/scsi/ufs/ufshcd.c @@ -8354,7 +8354,7 @@ static int ufshcd_set_dev_pwr_mode(struct ufs_hba *hba, sdev_printk(KERN_WARNING, sdp, "START_STOP failed for power mode: %d, result %x\n", pwr_mode, ret); - if (driver_byte(ret) == DRIVER_SENSE) + if (ret > 0 && driver_byte(ret) == DRIVER_SENSE) scsi_print_sense_hdr(sdp, NULL, &sshdr); } diff --git a/drivers/scsi/virtio_scsi.c b/drivers/scsi/virtio_scsi.c index 6dac58ae6120..fabbad5cda60 100644 --- a/drivers/scsi/virtio_scsi.c +++ b/drivers/scsi/virtio_scsi.c @@ -355,7 +355,7 @@ static void virtscsi_rescan_hotunplug(struct virtio_scsi *vscsi) if (result == 0 && inq_result[0] >> 5) { /* PQ indicates the LUN is not attached */ scsi_remove_device(sdev); - } else if (host_byte(result) == DID_BAD_TARGET) { + } else if (result > 0 && host_byte(result) == DID_BAD_TARGET) { /* * If all LUNs of a virtio-scsi device are unplugged * it will respond with BAD TARGET on any INQUIRY diff --git a/include/scsi/scsi.h b/include/scsi/scsi.h index 7ac50dbee475..6f227b610681 100644 --- a/include/scsi/scsi.h +++ b/include/scsi/scsi.h @@ -256,10 +256,13 @@ static inline int scsi_is_wlun(u64 lun) * This returns true for known good conditions that may be treated as * command completed normally */ -static inline int scsi_status_is_good(int status) +static inline bool scsi_status_is_good(int status) { + if (status < 0) + return false; + if (host_byte(status) == DID_NO_CONNECT) - return 0; + return false; /* * FIXME: bit0 is listed as reserved in SCSI-2, but is -- 2.31.1

1 0

[PATCH OLK-5.10] SCSI: spfc: Update lb mode acquired solution
by Yanling Song 14 Jan '22

14 Jan '22

Ramaxel inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4QLBP CVE: NA Changes: 1. Read lb mode from chip 2. Disable fake vf function 3. Remove unnecessary code Signed-off-by: Yanling Song <songyl(a)ramaxel.com> Reviewed-by: Yun Xu <xuyun(a)ramaxel.com> --- .../ethernet/ramaxel/spnic/hw/sphw_cfg_cmd.h | 3 +- .../ethernet/ramaxel/spnic/hw/sphw_hw_cfg.c | 3 + .../ethernet/ramaxel/spnic/hw/sphw_hw_cfg.h | 3 + drivers/scsi/spfc/hw/spfc_cqm_bat_cla.c | 43 +-- drivers/scsi/spfc/hw/spfc_cqm_bitmap_table.c | 10 +- drivers/scsi/spfc/hw/spfc_cqm_main.c | 299 +----------------- drivers/scsi/spfc/hw/spfc_cqm_main.h | 3 - drivers/scsi/spfc/hw/spfc_cqm_object.c | 21 -- 8 files changed, 29 insertions(+), 356 deletions(-) diff --git a/drivers/net/ethernet/ramaxel/spnic/hw/sphw_cfg_cmd.h b/drivers/net/ethernet/ramaxel/spnic/hw/sphw_cfg_cmd.h index 23644958e33a..63b89e71c552 100644 --- a/drivers/net/ethernet/ramaxel/spnic/hw/sphw_cfg_cmd.h +++ b/drivers/net/ethernet/ramaxel/spnic/hw/sphw_cfg_cmd.h @@ -40,7 +40,8 @@ struct cfg_cmd_dev_cap { u8 sf_svc_attr; u8 func_sf_en; - u16 rsvd_sf; + u8 lb_mode; + u8 smf_pg; u32 max_conn_num; u16 max_stick2cache_num; diff --git a/drivers/net/ethernet/ramaxel/spnic/hw/sphw_hw_cfg.c b/drivers/net/ethernet/ramaxel/spnic/hw/sphw_hw_cfg.c index 5d6a53307f0a..4b2674ec66f0 100644 --- a/drivers/net/ethernet/ramaxel/spnic/hw/sphw_hw_cfg.c +++ b/drivers/net/ethernet/ramaxel/spnic/hw/sphw_hw_cfg.c @@ -112,6 +112,9 @@ static void parse_pub_res_cap(struct sphw_hwdev *hwdev, else cap->sf_en = false; + cap->lb_mode = dev_cap->lb_mode; + cap->smf_pg = dev_cap->smf_pg; + cap->timer_en = (u8)timer_enable; /* timer enable */ cap->host_oq_id_mask_val = dev_cap->host_oq_id_mask_val; cap->max_connect_num = dev_cap->max_conn_num; diff --git a/drivers/net/ethernet/ramaxel/spnic/hw/sphw_hw_cfg.h b/drivers/net/ethernet/ramaxel/spnic/hw/sphw_hw_cfg.h index 1147beeb76ad..1b48e0991563 100644 --- a/drivers/net/ethernet/ramaxel/spnic/hw/sphw_hw_cfg.h +++ b/drivers/net/ethernet/ramaxel/spnic/hw/sphw_hw_cfg.h @@ -201,6 +201,9 @@ struct service_cap { u8 timer_en; /* 0:disable, 1:enable */ u8 bloomfilter_en; /* 0:disable, 1:enable*/ + u8 lb_mode; + u8 smf_pg; + /* For test */ u32 test_mode; u32 test_qpc_num; diff --git a/drivers/scsi/spfc/hw/spfc_cqm_bat_cla.c b/drivers/scsi/spfc/hw/spfc_cqm_bat_cla.c index 30fb56a9bfed..0c1d97d9e3e6 100644 --- a/drivers/scsi/spfc/hw/spfc_cqm_bat_cla.c +++ b/drivers/scsi/spfc/hw/spfc_cqm_bat_cla.c @@ -53,17 +53,6 @@ cqm_bat_fill_cla_common_gpa(struct cqm_handle *cqm_handle, gpa.acs_spu_en = 0; } - /* In fake mode, fake_vf_en in the GPA address of the BAT - * must be set to 1. - */ - if (cqm_handle->func_capability.fake_func_type == CQM_FAKE_FUNC_CHILD) { - gpa.fake_vf_en = 1; - func_attr = &cqm_handle->parent_cqm_handle->func_attribute; - gpa.pf_id = func_attr->func_global_idx; - } else { - gpa.fake_vf_en = 0; - } - memcpy(&cla_gpa_h, &gpa, sizeof(u32)); bat_entry_standerd->cla_gpa_h = cla_gpa_h; @@ -379,13 +368,8 @@ s32 cqm_bat_update(struct cqm_handle *cqm_handle) CQM_PTR_CHECK_RET(buf_in, CQM_FAIL, CQM_ALLOC_FAIL(buf_in)); buf_in->size = sizeof(struct cqm_cmdq_bat_update); - /* In non-fake mode, func_id is set to 0xffff, indicating the current func. - * In fake mode, the value of func_id is specified. This is a fake func_id. - */ - if (cqm_handle->func_capability.fake_func_type == CQM_FAKE_FUNC_CHILD) - func_id = cqm_handle->func_attribute.func_global_idx; - else - func_id = 0xffff; + /* In non-fake mode, func_id is set to 0xffff */ + func_id = 0xffff; /* The LB scenario is supported. * The normal mode is the traditional mode and is configured on SMF0. @@ -545,19 +529,6 @@ s32 cqm_cla_fill_buf(struct cqm_handle *cqm_handle, struct cqm_buf *cla_base_buf spu_en = 0; } - /* fake enable */ - if (cqm_handle->func_capability.fake_func_type == - CQM_FAKE_FUNC_CHILD) { - fake_en = 1ULL << 62; - func_attr = - &cqm_handle->parent_cqm_handle->func_attribute; - pf_id = func_attr->func_global_idx; - pf_id = (pf_id & 0x1f) << 57; - } else { - fake_en = 0; - pf_id = 0; - } - *base = (((((cla_sub_buf->buf_list[i].pa & CQM_CHIP_GPA_MASK) | spu_en) | fake_en) | @@ -1248,14 +1219,8 @@ s32 cqm_cla_update(struct cqm_handle *cqm_handle, struct cqm_buf_list *buf_node_ } } - /* In non-fake mode, set func_id to 0xffff. - * Indicates the current func fake mode, set func_id to the - * specified value, This is a fake func_id. - */ - if (cqm_handle->func_capability.fake_func_type == CQM_FAKE_FUNC_CHILD) - cmd.func_id = cqm_handle->func_attribute.func_global_idx; - else - cmd.func_id = 0xffff; + /* In non-fake mode, set func_id to 0xffff. */ + cmd.func_id = 0xffff; /* Mode 0 is hashed to 4 SMF engines (excluding PPF) by func ID. */ if (cqm_handle->func_capability.lb_mode == CQM_LB_MODE_NORMAL || diff --git a/drivers/scsi/spfc/hw/spfc_cqm_bitmap_table.c b/drivers/scsi/spfc/hw/spfc_cqm_bitmap_table.c index 4e482776a14f..21100e8db8f4 100644 --- a/drivers/scsi/spfc/hw/spfc_cqm_bitmap_table.c +++ b/drivers/scsi/spfc/hw/spfc_cqm_bitmap_table.c @@ -405,14 +405,8 @@ s32 cqm_cla_cache_invalid(struct cqm_handle *cqm_handle, dma_addr_t gpa, u32 cac cmd.gpa_h = CQM_ADDR_HI(gpa); cmd.gpa_l = CQM_ADDR_LW(gpa); - /* In non-fake mode, set func_id to 0xffff. - * Indicate the current func fake mode. - * The value of func_id is a fake func ID. - */ - if (cqm_handle->func_capability.fake_func_type == CQM_FAKE_FUNC_CHILD) - cmd.func_id = cqm_handle->func_attribute.func_global_idx; - else - cmd.func_id = 0xffff; + /* In non-fake mode, set func_id to 0xffff. */ + cmd.func_id = 0xffff; /* Mode 0 is hashed to 4 SMF engines (excluding PPF) by func ID. */ if (cqm_handle->func_capability.lb_mode == CQM_LB_MODE_NORMAL || diff --git a/drivers/scsi/spfc/hw/spfc_cqm_main.c b/drivers/scsi/spfc/hw/spfc_cqm_main.c index a4c8e60971b1..52cc2c7838e9 100644 --- a/drivers/scsi/spfc/hw/spfc_cqm_main.c +++ b/drivers/scsi/spfc/hw/spfc_cqm_main.c @@ -11,21 +11,8 @@ #include "sphw_crm.h" #include "sphw_hw.h" #include "sphw_hw_cfg.h" - #include "spfc_cqm_main.h" -static unsigned char cqm_lb_mode = CQM_LB_MODE_NORMAL; -module_param(cqm_lb_mode, byte, 0644); -MODULE_PARM_DESC(cqm_lb_mode, "for cqm lb mode (default=0xff)"); - -static unsigned char cqm_fake_mode = CQM_FAKE_MODE_DISABLE; -module_param(cqm_fake_mode, byte, 0644); -MODULE_PARM_DESC(cqm_fake_mode, "for cqm fake mode (default=0 disable)"); - -static unsigned char cqm_platform_mode = CQM_FPGA_MODE; -module_param(cqm_platform_mode, byte, 0644); -MODULE_PARM_DESC(cqm_platform_mode, "for cqm platform mode (default=0 FPGA)"); - s32 cqm3_init(void *ex_handle) { struct sphw_hwdev *handle = (struct sphw_hwdev *)ex_handle; @@ -63,19 +50,6 @@ s32 cqm3_init(void *ex_handle) goto err1; } - /* In FAKE mode, only the bitmap of the timer of the function is - * enabled, and resources are not initialized. Otherwise, the - * configuration of the fake function is overwritten. - */ - if (cqm_handle->func_capability.fake_func_type == CQM_FAKE_FUNC_CHILD_CONFLICT) { - if (sphw_func_tmr_bitmap_set(ex_handle, true) != CQM_SUCCESS) - cqm_err(handle->dev_hdl, "Timer start: enable timer bitmap failed\n"); - - handle->cqm_hdl = NULL; - kfree(cqm_handle); - return CQM_SUCCESS; - } - /* Initialize memory entries such as BAT, CLA, and bitmap. */ if (cqm_mem_init(ex_handle) != CQM_SUCCESS) { cqm_err(handle->dev_hdl, CQM_FUNCTION_FAIL(cqm_mem_init)); @@ -287,64 +261,6 @@ void cqm_service_capability_init(struct cqm_handle *cqm_handle, cqm_service_capability_init_fc(cqm_handle, (void *)service_capability); } -s32 cqm_get_fake_func_type(struct cqm_handle *cqm_handle) -{ - struct cqm_func_capability *func_cap = &cqm_handle->func_capability; - u32 parent_func, child_func_start, child_func_number, i; - u32 idx = cqm_handle->func_attribute.func_global_idx; - - /* Currently, only one set of fake configurations is implemented. - * fake_cfg_number = 1 - */ - for (i = 0; i < func_cap->fake_cfg_number; i++) { - parent_func = func_cap->fake_cfg[i].parent_func; - child_func_start = func_cap->fake_cfg[i].child_func_start; - child_func_number = func_cap->fake_cfg[i].child_func_number; - - if (idx == parent_func) { - return CQM_FAKE_FUNC_PARENT; - } else if ((idx >= child_func_start) && - (idx < (child_func_start + child_func_number))) { - return CQM_FAKE_FUNC_CHILD_CONFLICT; - } - } - - return CQM_FAKE_FUNC_NORMAL; -} - -s32 cqm_get_child_func_start(struct cqm_handle *cqm_handle) -{ - struct cqm_func_capability *func_cap = &cqm_handle->func_capability; - struct sphw_func_attr *func_attr = &cqm_handle->func_attribute; - u32 i; - - /* Currently, only one set of fake configurations is implemented. - * fake_cfg_number = 1 - */ - for (i = 0; i < func_cap->fake_cfg_number; i++) { - if (func_attr->func_global_idx == - func_cap->fake_cfg[i].parent_func) - return (s32)(func_cap->fake_cfg[i].child_func_start); - } - - return CQM_FAIL; -} - -s32 cqm_get_child_func_number(struct cqm_handle *cqm_handle) -{ - struct cqm_func_capability *func_cap = &cqm_handle->func_capability; - struct sphw_func_attr *func_attr = &cqm_handle->func_attribute; - u32 i; - - for (i = 0; i < func_cap->fake_cfg_number; i++) { - if (func_attr->func_global_idx == - func_cap->fake_cfg[i].parent_func) - return (s32)(func_cap->fake_cfg[i].child_func_number); - } - - return CQM_FAIL; -} - /* Set func_type in fake_cqm_handle to ppf, pf, or vf. */ void cqm_set_func_type(struct cqm_handle *cqm_handle) { @@ -358,41 +274,20 @@ void cqm_set_func_type(struct cqm_handle *cqm_handle) cqm_handle->func_attribute.func_type = CQM_VF; } -void cqm_lb_fake_mode_init(struct cqm_handle *cqm_handle) +void cqm_lb_fake_mode_init(struct cqm_handle *cqm_handle, struct service_cap *svc_cap) { struct cqm_func_capability *func_cap = &cqm_handle->func_capability; - struct cqm_fake_cfg *cfg = func_cap->fake_cfg; - func_cap->lb_mode = cqm_lb_mode; - func_cap->fake_mode = cqm_fake_mode; + func_cap->lb_mode = svc_cap->lb_mode; /* Initializing the LB Mode */ - if (func_cap->lb_mode == CQM_LB_MODE_NORMAL) { + if (func_cap->lb_mode == CQM_LB_MODE_NORMAL) func_cap->smf_pg = 0; - } else { - /* The LB mode is tailored on the FPGA. - * Only SMF0 and SMF2 are instantiated. - */ - if (cqm_platform_mode == CQM_FPGA_MODE) - func_cap->smf_pg = 0x5; - else - func_cap->smf_pg = 0xF; - } + else + func_cap->smf_pg = svc_cap->smf_pg; - /* Initializing the FAKE Mode */ - if (func_cap->fake_mode == CQM_FAKE_MODE_DISABLE) { - func_cap->fake_cfg_number = 0; - func_cap->fake_func_type = CQM_FAKE_FUNC_NORMAL; - } else { - func_cap->fake_cfg_number = 1; - - /* When configuring fake mode, ensure that the parent function - * cannot be contained in the child function; otherwise, the - * system will be initialized repeatedly. - */ - cfg[0].child_func_start = CQM_FAKE_CFUNC_START; - func_cap->fake_func_type = cqm_get_fake_func_type(cqm_handle); - } + func_cap->fake_cfg_number = 0; + func_cap->fake_func_type = CQM_FAKE_FUNC_NORMAL; } s32 cqm_capability_init(void *ex_handle) @@ -465,10 +360,9 @@ s32 cqm_capability_init(void *ex_handle) func_cap->gpa_check_enable = true; - cqm_lb_fake_mode_init(cqm_handle); + cqm_lb_fake_mode_init(cqm_handle, service_capability); cqm_info(handle->dev_hdl, "Cap init: lb_mode=%u\n", func_cap->lb_mode); cqm_info(handle->dev_hdl, "Cap init: smf_pg=%u\n", func_cap->smf_pg); - cqm_info(handle->dev_hdl, "Cap init: fake_mode=%u\n", func_cap->fake_mode); cqm_info(handle->dev_hdl, "Cap init: fake_func_type=%u\n", func_cap->fake_func_type); cqm_info(handle->dev_hdl, "Cap init: fake_cfg_number=%u\n", func_cap->fake_cfg_number); @@ -517,153 +411,6 @@ s32 cqm_capability_init(void *ex_handle) return err; } -void cqm_fake_uninit(struct cqm_handle *cqm_handle) -{ - u32 i; - - if (cqm_handle->func_capability.fake_func_type != - CQM_FAKE_FUNC_PARENT) - return; - - for (i = 0; i < CQM_FAKE_FUNC_MAX; i++) { - kfree(cqm_handle->fake_cqm_handle[i]); - cqm_handle->fake_cqm_handle[i] = NULL; - } -} - -s32 cqm_fake_init(struct cqm_handle *cqm_handle) -{ - struct sphw_hwdev *handle = cqm_handle->ex_handle; - struct cqm_func_capability *func_cap = NULL; - struct cqm_handle *fake_cqm_handle = NULL; - struct sphw_func_attr *func_attr = NULL; - s32 child_func_start, child_func_number; - u32 i; - - func_cap = &cqm_handle->func_capability; - if (func_cap->fake_func_type != CQM_FAKE_FUNC_PARENT) - return CQM_SUCCESS; - - child_func_start = cqm_get_child_func_start(cqm_handle); - if (child_func_start == CQM_FAIL) { - cqm_err(handle->dev_hdl, CQM_WRONG_VALUE(child_func_start)); - return CQM_FAIL; - } - - child_func_number = cqm_get_child_func_number(cqm_handle); - if (child_func_number == CQM_FAIL) { - cqm_err(handle->dev_hdl, CQM_WRONG_VALUE(child_func_number)); - return CQM_FAIL; - } - - for (i = 0; i < (u32)child_func_number; i++) { - fake_cqm_handle = kmalloc(sizeof(*fake_cqm_handle), GFP_KERNEL | __GFP_ZERO); - if (!fake_cqm_handle) { - cqm_err(handle->dev_hdl, - CQM_ALLOC_FAIL(fake_cqm_handle)); - goto err; - } - - /* Copy the attributes of the parent CQM handle to the child CQM - * handle and modify the values of function. - */ - memcpy(fake_cqm_handle, cqm_handle, sizeof(struct cqm_handle)); - func_attr = &fake_cqm_handle->func_attribute; - func_cap = &fake_cqm_handle->func_capability; - func_attr->func_global_idx = (u16)(child_func_start + i); - cqm_set_func_type(fake_cqm_handle); - func_cap->fake_func_type = CQM_FAKE_FUNC_CHILD; - cqm_info(handle->dev_hdl, "Fake func init: function[%u] type %d(0:PF,1:VF,2:PPF)\n", - func_attr->func_global_idx, func_attr->func_type); - - fake_cqm_handle->parent_cqm_handle = cqm_handle; - cqm_handle->fake_cqm_handle[i] = fake_cqm_handle; - } - - return CQM_SUCCESS; - -err: - cqm_fake_uninit(cqm_handle); - return CQM_FAIL; -} - -void cqm_fake_mem_uninit(struct cqm_handle *cqm_handle) -{ - struct sphw_hwdev *handle = cqm_handle->ex_handle; - struct cqm_handle *fake_cqm_handle = NULL; - s32 child_func_number; - u32 i; - - if (cqm_handle->func_capability.fake_func_type != CQM_FAKE_FUNC_PARENT) - return; - - child_func_number = cqm_get_child_func_number(cqm_handle); - if (child_func_number == CQM_FAIL) { - cqm_err(handle->dev_hdl, CQM_WRONG_VALUE(child_func_number)); - return; - } - - for (i = 0; i < (u32)child_func_number; i++) { - fake_cqm_handle = cqm_handle->fake_cqm_handle[i]; - cqm_object_table_uninit(fake_cqm_handle); - cqm_bitmap_uninit(fake_cqm_handle); - cqm_cla_uninit(fake_cqm_handle, CQM_BAT_ENTRY_MAX); - cqm_bat_uninit(fake_cqm_handle); - } -} - -s32 cqm_fake_mem_init(struct cqm_handle *cqm_handle) -{ - struct sphw_hwdev *handle = cqm_handle->ex_handle; - struct cqm_handle *fake_cqm_handle = NULL; - s32 child_func_number; - u32 i; - - if (cqm_handle->func_capability.fake_func_type != - CQM_FAKE_FUNC_PARENT) - return CQM_SUCCESS; - - child_func_number = cqm_get_child_func_number(cqm_handle); - if (child_func_number == CQM_FAIL) { - cqm_err(handle->dev_hdl, CQM_WRONG_VALUE(child_func_number)); - return CQM_FAIL; - } - - for (i = 0; i < (u32)child_func_number; i++) { - fake_cqm_handle = cqm_handle->fake_cqm_handle[i]; - - if (cqm_bat_init(fake_cqm_handle) != CQM_SUCCESS) { - cqm_err(handle->dev_hdl, - CQM_FUNCTION_FAIL(cqm_bat_init)); - goto err; - } - - if (cqm_cla_init(fake_cqm_handle) != CQM_SUCCESS) { - cqm_err(handle->dev_hdl, - CQM_FUNCTION_FAIL(cqm_cla_init)); - goto err; - } - - if (cqm_bitmap_init(fake_cqm_handle) != CQM_SUCCESS) { - cqm_err(handle->dev_hdl, - CQM_FUNCTION_FAIL(cqm_bitmap_init)); - goto err; - } - - if (cqm_object_table_init(fake_cqm_handle) != CQM_SUCCESS) { - cqm_err(handle->dev_hdl, - CQM_FUNCTION_FAIL(cqm_object_table_init)); - goto err; - } - } - - return CQM_SUCCESS; - -err: - cqm_fake_mem_uninit(cqm_handle); - return CQM_FAIL; -} - s32 cqm_mem_init(void *ex_handle) { struct sphw_hwdev *handle = (struct sphw_hwdev *)ex_handle; @@ -671,49 +418,35 @@ s32 cqm_mem_init(void *ex_handle) cqm_handle = (struct cqm_handle *)(handle->cqm_hdl); - if (cqm_fake_init(cqm_handle) != CQM_SUCCESS) { - cqm_err(handle->dev_hdl, CQM_FUNCTION_FAIL(cqm_fake_init)); - return CQM_FAIL; - } - - if (cqm_fake_mem_init(cqm_handle) != CQM_SUCCESS) { - cqm_err(handle->dev_hdl, CQM_FUNCTION_FAIL(cqm_fake_mem_init)); - goto err1; - } - if (cqm_bat_init(cqm_handle) != CQM_SUCCESS) { cqm_err(handle->dev_hdl, CQM_FUNCTION_FAIL(cqm_bat_init)); - goto err2; + return CQM_FAIL; } if (cqm_cla_init(cqm_handle) != CQM_SUCCESS) { cqm_err(handle->dev_hdl, CQM_FUNCTION_FAIL(cqm_cla_init)); - goto err3; + goto err1; } if (cqm_bitmap_init(cqm_handle) != CQM_SUCCESS) { cqm_err(handle->dev_hdl, CQM_FUNCTION_FAIL(cqm_bitmap_init)); - goto err4; + goto err2; } if (cqm_object_table_init(cqm_handle) != CQM_SUCCESS) { cqm_err(handle->dev_hdl, CQM_FUNCTION_FAIL(cqm_object_table_init)); - goto err5; + goto err3; } return CQM_SUCCESS; -err5: - cqm_bitmap_uninit(cqm_handle); -err4: - cqm_cla_uninit(cqm_handle, CQM_BAT_ENTRY_MAX); err3: - cqm_bat_uninit(cqm_handle); + cqm_bitmap_uninit(cqm_handle); err2: - cqm_fake_mem_uninit(cqm_handle); + cqm_cla_uninit(cqm_handle, CQM_BAT_ENTRY_MAX); err1: - cqm_fake_uninit(cqm_handle); + cqm_bat_uninit(cqm_handle); return CQM_FAIL; } @@ -728,8 +461,6 @@ void cqm_mem_uninit(void *ex_handle) cqm_bitmap_uninit(cqm_handle); cqm_cla_uninit(cqm_handle, CQM_BAT_ENTRY_MAX); cqm_bat_uninit(cqm_handle); - cqm_fake_mem_uninit(cqm_handle); - cqm_fake_uninit(cqm_handle); } s32 cqm_event_init(void *ex_handle) diff --git a/drivers/scsi/spfc/hw/spfc_cqm_main.h b/drivers/scsi/spfc/hw/spfc_cqm_main.h index 1b8cf8bdb3b7..cf10d7f5c339 100644 --- a/drivers/scsi/spfc/hw/spfc_cqm_main.h +++ b/drivers/scsi/spfc/hw/spfc_cqm_main.h @@ -328,9 +328,6 @@ void cqm_mem_uninit(void *ex_handle); s32 cqm_event_init(void *ex_handle); void cqm_event_uninit(void *ex_handle); u8 cqm_aeq_callback(void *ex_handle, u8 event, u8 *data); -s32 cqm_get_fake_func_type(struct cqm_handle *cqm_handle); -s32 cqm_get_child_func_start(struct cqm_handle *cqm_handle); -s32 cqm_get_child_func_number(struct cqm_handle *cqm_handle); s32 cqm3_init(void *ex_handle); void cqm3_uninit(void *ex_handle); diff --git a/drivers/scsi/spfc/hw/spfc_cqm_object.c b/drivers/scsi/spfc/hw/spfc_cqm_object.c index b895d37aebae..165794e9c7e5 100644 --- a/drivers/scsi/spfc/hw/spfc_cqm_object.c +++ b/drivers/scsi/spfc/hw/spfc_cqm_object.c @@ -155,8 +155,6 @@ struct cqm_qpc_mpt *cqm3_object_qpc_mpt_create(void *ex_handle, u32 service_type struct cqm_qpc_mpt_info *qpc_mpt_info = NULL; struct cqm_handle *cqm_handle = NULL; s32 ret = CQM_FAIL; - u32 relative_index; - u32 fake_func_id; CQM_PTR_CHECK_RET(ex_handle, NULL, CQM_PTR_NULL(ex_handle)); @@ -180,25 +178,6 @@ struct cqm_qpc_mpt *cqm3_object_qpc_mpt_create(void *ex_handle, u32 service_type return NULL; } - /* fake vf adaption, switch to corresponding VF. */ - if (cqm_handle->func_capability.fake_func_type == - CQM_FAKE_FUNC_PARENT) { - fake_func_id = index / cqm_handle->func_capability.qpc_number; - relative_index = index % cqm_handle->func_capability.qpc_number; - - cqm_info(handle->dev_hdl, "qpc create: fake_func_id=%u, relative_index=%u\n", - fake_func_id, relative_index); - - if ((s32)fake_func_id >= - cqm_get_child_func_number(cqm_handle)) { - cqm_err(handle->dev_hdl, CQM_WRONG_VALUE(fake_func_id)); - return NULL; - } - - index = relative_index; - cqm_handle = cqm_handle->fake_cqm_handle[fake_func_id]; - } - qpc_mpt_info = kmalloc(sizeof(*qpc_mpt_info), GFP_ATOMIC | __GFP_ZERO); CQM_PTR_CHECK_RET(qpc_mpt_info, NULL, CQM_ALLOC_FAIL(qpc_mpt_info)); -- 2.32.0

1 0

[PATCH v1 OLK-5.10 0/4] Add four modifications
by Yanling Song 14 Jan '22

14 Jan '22

1.Remove unused functions about ceq 2.Remove unused clp hardware channels 3.Remove the code of polling mode 4.Remove the code about little endian and big endian conversion Yanling Song (4): net/spnic:Remove unused functions about ceq net/spnic:Remove unused clp hardware channels net/spnic:Remove the code of polling mode net/spnic:Remove the code about little endian and big endian conversion .../net/ethernet/ramaxel/spnic/hw/sphw_cmdq.c | 58 +-- .../ethernet/ramaxel/spnic/hw/sphw_common.h | 12 - .../net/ethernet/ramaxel/spnic/hw/sphw_crm.h | 2 - .../net/ethernet/ramaxel/spnic/hw/sphw_csr.h | 13 - .../net/ethernet/ramaxel/spnic/hw/sphw_eqs.c | 104 +--- .../net/ethernet/ramaxel/spnic/hw/sphw_hw.h | 2 - .../ethernet/ramaxel/spnic/hw/sphw_hwdev.c | 31 -- .../ethernet/ramaxel/spnic/hw/sphw_hwdev.h | 2 - .../net/ethernet/ramaxel/spnic/hw/sphw_hwif.c | 27 +- .../net/ethernet/ramaxel/spnic/hw/sphw_mbox.c | 20 +- .../net/ethernet/ramaxel/spnic/hw/sphw_mbox.h | 2 - .../net/ethernet/ramaxel/spnic/hw/sphw_mgmt.c | 487 ------------------ .../net/ethernet/ramaxel/spnic/hw/sphw_mgmt.h | 50 -- .../net/ethernet/ramaxel/spnic/hw/sphw_mt.h | 1 - .../net/ethernet/ramaxel/spnic/spnic_nic_io.h | 6 +- .../net/ethernet/ramaxel/spnic/spnic_nic_qp.h | 7 +- drivers/net/ethernet/ramaxel/spnic/spnic_rx.c | 39 +- drivers/net/ethernet/ramaxel/spnic/spnic_tx.c | 20 +- 18 files changed, 41 insertions(+), 842 deletions(-) -- 2.23.0

1 4

[PATCH openEuler-1.0-LTS] ext4: Fix BUG_ON in ext4_bread when write quota data
by Yang Yingliang 13 Jan '22

13 Jan '22

From: Ye Bin <yebin10(a)huawei.com> mainline inclusion from mainline-v5.17 commit ce85548ab4295234b4f8e63a0eea0c157d2f6b25 category: bugfix bugzilla: 185930 CVE: NA ----------------------------------------------- We got issue as follows when run syzkaller: [ 167.936972] EXT4-fs error (device loop0): __ext4_remount:6314: comm rep: Abort forced by user [ 167.938306] EXT4-fs (loop0): Remounting filesystem read-only [ 167.981637] Assertion failure in ext4_getblk() at fs/ext4/inode.c:847: '(EXT4_SB(inode->i_sb)->s_mount_state & EXT4_FC_REPLAY) || handle != NULL || create == 0' [ 167.983601] ------------[ cut here ]------------ [ 167.984245] kernel BUG at fs/ext4/inode.c:847! [ 167.984882] invalid opcode: 0000 [#1] PREEMPT SMP KASAN PTI [ 167.985624] CPU: 7 PID: 2290 Comm: rep Tainted: G B 5.16.0-rc5-next-20211217+ #123 [ 167.986823] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS ?-20190727_073836-buildvm-ppc64le-16.ppc.fedoraproject.org-3.fc31 04/01/2014 [ 167.988590] RIP: 0010:ext4_getblk+0x17e/0x504 [ 167.989189] Code: c6 01 74 28 49 c7 c0 a0 a3 5c 9b b9 4f 03 00 00 48 c7 c2 80 9c 5c 9b 48 c7 c6 40 b6 5c 9b 48 c7 c7 20 a4 5c 9b e8 77 e3 fd ff <0f> 0b 8b 04 244 [ 167.991679] RSP: 0018:ffff8881736f7398 EFLAGS: 00010282 [ 167.992385] RAX: 0000000000000094 RBX: 1ffff1102e6dee75 RCX: 0000000000000000 [ 167.993337] RDX: 0000000000000001 RSI: ffffffff9b6e29e0 RDI: ffffed102e6dee66 [ 167.994292] RBP: ffff88816a076210 R08: 0000000000000094 R09: ffffed107363fa09 [ 167.995252] R10: ffff88839b1fd047 R11: ffffed107363fa08 R12: ffff88816a0761e8 [ 167.996205] R13: 0000000000000000 R14: 0000000000000021 R15: 0000000000000001 [ 167.997158] FS: 00007f6a1428c740(0000) GS:ffff88839b000000(0000) knlGS:0000000000000000 [ 167.998238] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 167.999025] CR2: 00007f6a140716c8 CR3: 0000000133216000 CR4: 00000000000006e0 [ 167.999987] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 168.000944] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 168.001899] Call Trace: [ 168.002235] <TASK> [ 168.007167] ext4_bread+0xd/0x53 [ 168.007612] ext4_quota_write+0x20c/0x5c0 [ 168.010457] write_blk+0x100/0x220 [ 168.010944] remove_free_dqentry+0x1c6/0x440 [ 168.011525] free_dqentry.isra.0+0x565/0x830 [ 168.012133] remove_tree+0x318/0x6d0 [ 168.014744] remove_tree+0x1eb/0x6d0 [ 168.017346] remove_tree+0x1eb/0x6d0 [ 168.019969] remove_tree+0x1eb/0x6d0 [ 168.022128] qtree_release_dquot+0x291/0x340 [ 168.023297] v2_release_dquot+0xce/0x120 [ 168.023847] dquot_release+0x197/0x3e0 [ 168.024358] ext4_release_dquot+0x22a/0x2d0 [ 168.024932] dqput.part.0+0x1c9/0x900 [ 168.025430] __dquot_drop+0x120/0x190 [ 168.025942] ext4_clear_inode+0x86/0x220 [ 168.026472] ext4_evict_inode+0x9e8/0xa22 [ 168.028200] evict+0x29e/0x4f0 [ 168.028625] dispose_list+0x102/0x1f0 [ 168.029148] evict_inodes+0x2c1/0x3e0 [ 168.030188] generic_shutdown_super+0xa4/0x3b0 [ 168.030817] kill_block_super+0x95/0xd0 [ 168.031360] deactivate_locked_super+0x85/0xd0 [ 168.031977] cleanup_mnt+0x2bc/0x480 [ 168.033062] task_work_run+0xd1/0x170 [ 168.033565] do_exit+0xa4f/0x2b50 [ 168.037155] do_group_exit+0xef/0x2d0 [ 168.037666] __x64_sys_exit_group+0x3a/0x50 [ 168.038237] do_syscall_64+0x3b/0x90 [ 168.038751] entry_SYSCALL_64_after_hwframe+0x44/0xae In order to reproduce this problem, the following conditions need to be met: 1. Ext4 filesystem with no journal; 2. Filesystem image with incorrect quota data; 3. Abort filesystem forced by user; 4. umount filesystem; As in ext4_quota_write: ... if (EXT4_SB(sb)->s_journal && !handle) { ext4_msg(sb, KERN_WARNING, "Quota write (off=%llu, len=%llu)" " cancelled because transaction is not started", (unsigned long long)off, (unsigned long long)len); return -EIO; } ... We only check handle if NULL when filesystem has journal. There is need check handle if NULL even when filesystem has no journal. Signed-off-by: Ye Bin <yebin10(a)huawei.com> Reviewed-by: Jan Kara <jack(a)suse.cz> Link: https://lore.kernel.org/r/20211223015506.297766-1-yebin10@huawei.com Signed-off-by: Theodore Ts'o <tytso(a)mit.edu> Cc: stable(a)kernel.org Signed-off-by: Ye Bin <yebin10(a)huawei.com> Reviewed-by: Zhang Yi <yi.zhang(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- fs/ext4/super.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 587817ac60990..b89f431ec78d3 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -6245,7 +6245,7 @@ static ssize_t ext4_quota_write(struct super_block *sb, int type, struct buffer_head *bh; handle_t *handle = journal_current_handle(); - if (EXT4_SB(sb)->s_journal && !handle) { + if (!handle) { ext4_msg(sb, KERN_WARNING, "Quota write (off=%llu, len=%llu)" " cancelled because transaction is not started", (unsigned long long)off, (unsigned long long)len); -- 2.25.1

1 0

[PATCH openEuler-1.0-LTS] PM: hibernate: use correct mode for swsusp_close()
by Yang Yingliang 13 Jan '22

13 Jan '22

From: Thomas Zeitlhofer <thomas.zeitlhofer+lkml(a)ze-it.at> stable inclusion from linux-v4.19.219 commit 68945e943519df1532e598fafab16ac54488933f --------------------------------------------------- [ Upstream commit cefcf24b4d351daf70ecd945324e200d3736821e ] Commit 39fbef4b0f77 ("PM: hibernate: Get block device exclusively in swsusp_check()") changed the opening mode of the block device to (FMODE_READ | FMODE_EXCL). In the corresponding calls to swsusp_close(), the mode is still just FMODE_READ which triggers the warning in blkdev_flush_mapping() on resume from hibernate. So, use the mode (FMODE_READ | FMODE_EXCL) also when closing the device. Fixes: 39fbef4b0f77 ("PM: hibernate: Get block device exclusively in swsusp_check()") Signed-off-by: Thomas Zeitlhofer <thomas.zeitlhofer+lkml(a)ze-it.at> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki(a)intel.com> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Ye Bin <yebin10(a)huawei.com> Reviewed-by: Zhang Yi <yi.zhang(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- kernel/power/hibernate.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/kernel/power/hibernate.c b/kernel/power/hibernate.c index 72ef9df2a29cd..fc6466351e62f 100644 --- a/kernel/power/hibernate.c +++ b/kernel/power/hibernate.c @@ -686,7 +686,7 @@ static int load_image_and_restore(void) goto Unlock; error = swsusp_read(&flags); - swsusp_close(FMODE_READ); + swsusp_close(FMODE_READ | FMODE_EXCL); if (!error) hibernation_restore(flags & SF_PLATFORM_MODE); @@ -896,7 +896,7 @@ static int software_resume(void) /* The snapshot device should not be opened while we're running */ if (!atomic_add_unless(&snapshot_device_available, -1, 0)) { error = -EBUSY; - swsusp_close(FMODE_READ); + swsusp_close(FMODE_READ | FMODE_EXCL); goto Unlock; } @@ -925,7 +925,7 @@ static int software_resume(void) pm_pr_dbg("Hibernation image not present or could not be loaded.\n"); return error; Close_Finish: - swsusp_close(FMODE_READ); + swsusp_close(FMODE_READ | FMODE_EXCL); goto Finish; } -- 2.25.1

1 0

[PATCH openEuler-1.0-LTS] Revert "watchdog: Fix check_preemption_disabled() error"
by Yang Yingliang 13 Jan '22

13 Jan '22

hulk inclusion category: bugfix bugzilla: 173968, https://gitee.com/openeuler/kernel/issues/I3J87Y CVE: NA --------------------------- This reverts commit b2e484e966ac2f211f8f8cf56083e9bc4d5d3c6a. When CONFIG_LOCKDEP and CONFIG_DEBUG_LOCKDEP are enabled, it detects the following error: [ 10.145007] BUG: sleeping function called from invalid context at mm/slab.h:418 [ 10.145394] in_atomic(): 1, irqs_disabled(): 0, pid: 1, name: swapper/0 [ 10.145765] Preemption disabled at: [ 10.145978] [<ffff000008f8e7b4>] hardlockup_detector_perf_init+0x20/0x100 [ 10.146770] CPU: 6 PID: 1 Comm: swapper/0 Not tainted 4.19.90+ #3 [ 10.148242] Hardware name: linux,dummy-virt (DT) [ 10.148572] Call trace: [ 10.148667] dump_backtrace+0x0/0x190 [ 10.148765] show_stack+0x24/0x30 [ 10.148875] dump_stack+0xa4/0xf8 [ 10.148964] ___might_sleep+0x150/0x180 [ 10.149065] __might_sleep+0x58/0x90 [ 10.149199] kmem_cache_alloc_trace+0x244/0x2b0 [ 10.149308] perf_event_alloc+0x74/0x680 [ 10.149402] perf_event_create_kernel_counter+0x2c/0x190 [ 10.149516] arch_probe_cpu_freq+0x84/0x1ac [ 10.149611] hw_nmi_get_sample_period+0xb8/0x180 [ 10.149713] hardlockup_detector_event_create+0x28/0xfc [ 10.149827] hardlockup_detector_perf_init+0x24/0x100 [ 10.149943] watchdog_nmi_probe+0x14/0x1c [ 10.150037] lockup_detector_init+0x58/0x98 [ 10.150173] kernel_init_freeable+0x10c/0x1c4 [ 10.150298] kernel_init+0x18/0x110 [ 10.150422] ret_from_fork+0x10/0x18 In 'b2e484e966ac ("watchdog: Fix check_preemption_disabled() error")', we tried to fix check_preemption_disabled() error by disabling preemption in hardlockup_detector_perf_init(), but missed that function perf_event_create_kernel_counter() may sleep. The preemption is always disabled, the problem that wanted be fixed is not existed, so just revert this commit. Fixes: b2e484e966ac ("watchdog: Fix check_preemption_disabled() error") Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> Reviewed-by: Xiongfeng Wang <wangxiongfeng2(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- kernel/watchdog_hld.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/kernel/watchdog_hld.c b/kernel/watchdog_hld.c index a5aff8ffa48ce..43832b1023693 100644 --- a/kernel/watchdog_hld.c +++ b/kernel/watchdog_hld.c @@ -508,17 +508,14 @@ void __init hardlockup_detector_perf_restart(void) */ int __init hardlockup_detector_perf_init(void) { - int ret; + int ret = hardlockup_detector_event_create(); - preempt_disable(); - ret = hardlockup_detector_event_create(); if (ret) { pr_info("Perf NMI watchdog permanently disabled\n"); } else { perf_event_release_kernel(this_cpu_read(watchdog_ev)); this_cpu_write(watchdog_ev, NULL); } - preempt_enable(); return ret; } #endif /* CONFIG_HARDLOCKUP_DETECTOR_PERF */ -- 2.25.1

1 0

[PATCH OLK-5.10] SCSI: spfc: Update lb mode acquired solution
by Yanling Song 13 Jan '22

13 Jan '22

Ramaxel inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4QLBP CVE: NA Changes: 1. Read lb mode from chip 2. Disable fake vf function 3. Remove unnecessary code Signed-off-by: Yanling Song <songyl(a)ramaxel.com> Reviewed-by: Yun Xu <xuyun(a)ramaxel.com> --- .../ethernet/ramaxel/spnic/hw/sphw_cfg_cmd.h | 3 +- .../ethernet/ramaxel/spnic/hw/sphw_hw_cfg.c | 3 + .../ethernet/ramaxel/spnic/hw/sphw_hw_cfg.h | 3 + drivers/scsi/spfc/hw/spfc_cqm_bat_cla.c | 43 +-- drivers/scsi/spfc/hw/spfc_cqm_bitmap_table.c | 10 +- drivers/scsi/spfc/hw/spfc_cqm_main.c | 299 +----------------- drivers/scsi/spfc/hw/spfc_cqm_main.h | 3 - drivers/scsi/spfc/hw/spfc_cqm_object.c | 21 -- 8 files changed, 29 insertions(+), 356 deletions(-) diff --git a/drivers/net/ethernet/ramaxel/spnic/hw/sphw_cfg_cmd.h b/drivers/net/ethernet/ramaxel/spnic/hw/sphw_cfg_cmd.h index 23644958e33a..63b89e71c552 100644 --- a/drivers/net/ethernet/ramaxel/spnic/hw/sphw_cfg_cmd.h +++ b/drivers/net/ethernet/ramaxel/spnic/hw/sphw_cfg_cmd.h @@ -40,7 +40,8 @@ struct cfg_cmd_dev_cap { u8 sf_svc_attr; u8 func_sf_en; - u16 rsvd_sf; + u8 lb_mode; + u8 smf_pg; u32 max_conn_num; u16 max_stick2cache_num; diff --git a/drivers/net/ethernet/ramaxel/spnic/hw/sphw_hw_cfg.c b/drivers/net/ethernet/ramaxel/spnic/hw/sphw_hw_cfg.c index 5d6a53307f0a..4b2674ec66f0 100644 --- a/drivers/net/ethernet/ramaxel/spnic/hw/sphw_hw_cfg.c +++ b/drivers/net/ethernet/ramaxel/spnic/hw/sphw_hw_cfg.c @@ -112,6 +112,9 @@ static void parse_pub_res_cap(struct sphw_hwdev *hwdev, else cap->sf_en = false; + cap->lb_mode = dev_cap->lb_mode; + cap->smf_pg = dev_cap->smf_pg; + cap->timer_en = (u8)timer_enable; /* timer enable */ cap->host_oq_id_mask_val = dev_cap->host_oq_id_mask_val; cap->max_connect_num = dev_cap->max_conn_num; diff --git a/drivers/net/ethernet/ramaxel/spnic/hw/sphw_hw_cfg.h b/drivers/net/ethernet/ramaxel/spnic/hw/sphw_hw_cfg.h index 1147beeb76ad..1b48e0991563 100644 --- a/drivers/net/ethernet/ramaxel/spnic/hw/sphw_hw_cfg.h +++ b/drivers/net/ethernet/ramaxel/spnic/hw/sphw_hw_cfg.h @@ -201,6 +201,9 @@ struct service_cap { u8 timer_en; /* 0:disable, 1:enable */ u8 bloomfilter_en; /* 0:disable, 1:enable*/ + u8 lb_mode; + u8 smf_pg; + /* For test */ u32 test_mode; u32 test_qpc_num; diff --git a/drivers/scsi/spfc/hw/spfc_cqm_bat_cla.c b/drivers/scsi/spfc/hw/spfc_cqm_bat_cla.c index 30fb56a9bfed..0c1d97d9e3e6 100644 --- a/drivers/scsi/spfc/hw/spfc_cqm_bat_cla.c +++ b/drivers/scsi/spfc/hw/spfc_cqm_bat_cla.c @@ -53,17 +53,6 @@ cqm_bat_fill_cla_common_gpa(struct cqm_handle *cqm_handle, gpa.acs_spu_en = 0; } - /* In fake mode, fake_vf_en in the GPA address of the BAT - * must be set to 1. - */ - if (cqm_handle->func_capability.fake_func_type == CQM_FAKE_FUNC_CHILD) { - gpa.fake_vf_en = 1; - func_attr = &cqm_handle->parent_cqm_handle->func_attribute; - gpa.pf_id = func_attr->func_global_idx; - } else { - gpa.fake_vf_en = 0; - } - memcpy(&cla_gpa_h, &gpa, sizeof(u32)); bat_entry_standerd->cla_gpa_h = cla_gpa_h; @@ -379,13 +368,8 @@ s32 cqm_bat_update(struct cqm_handle *cqm_handle) CQM_PTR_CHECK_RET(buf_in, CQM_FAIL, CQM_ALLOC_FAIL(buf_in)); buf_in->size = sizeof(struct cqm_cmdq_bat_update); - /* In non-fake mode, func_id is set to 0xffff, indicating the current func. - * In fake mode, the value of func_id is specified. This is a fake func_id. - */ - if (cqm_handle->func_capability.fake_func_type == CQM_FAKE_FUNC_CHILD) - func_id = cqm_handle->func_attribute.func_global_idx; - else - func_id = 0xffff; + /* In non-fake mode, func_id is set to 0xffff */ + func_id = 0xffff; /* The LB scenario is supported. * The normal mode is the traditional mode and is configured on SMF0. @@ -545,19 +529,6 @@ s32 cqm_cla_fill_buf(struct cqm_handle *cqm_handle, struct cqm_buf *cla_base_buf spu_en = 0; } - /* fake enable */ - if (cqm_handle->func_capability.fake_func_type == - CQM_FAKE_FUNC_CHILD) { - fake_en = 1ULL << 62; - func_attr = - &cqm_handle->parent_cqm_handle->func_attribute; - pf_id = func_attr->func_global_idx; - pf_id = (pf_id & 0x1f) << 57; - } else { - fake_en = 0; - pf_id = 0; - } - *base = (((((cla_sub_buf->buf_list[i].pa & CQM_CHIP_GPA_MASK) | spu_en) | fake_en) | @@ -1248,14 +1219,8 @@ s32 cqm_cla_update(struct cqm_handle *cqm_handle, struct cqm_buf_list *buf_node_ } } - /* In non-fake mode, set func_id to 0xffff. - * Indicates the current func fake mode, set func_id to the - * specified value, This is a fake func_id. - */ - if (cqm_handle->func_capability.fake_func_type == CQM_FAKE_FUNC_CHILD) - cmd.func_id = cqm_handle->func_attribute.func_global_idx; - else - cmd.func_id = 0xffff; + /* In non-fake mode, set func_id to 0xffff. */ + cmd.func_id = 0xffff; /* Mode 0 is hashed to 4 SMF engines (excluding PPF) by func ID. */ if (cqm_handle->func_capability.lb_mode == CQM_LB_MODE_NORMAL || diff --git a/drivers/scsi/spfc/hw/spfc_cqm_bitmap_table.c b/drivers/scsi/spfc/hw/spfc_cqm_bitmap_table.c index 4e482776a14f..21100e8db8f4 100644 --- a/drivers/scsi/spfc/hw/spfc_cqm_bitmap_table.c +++ b/drivers/scsi/spfc/hw/spfc_cqm_bitmap_table.c @@ -405,14 +405,8 @@ s32 cqm_cla_cache_invalid(struct cqm_handle *cqm_handle, dma_addr_t gpa, u32 cac cmd.gpa_h = CQM_ADDR_HI(gpa); cmd.gpa_l = CQM_ADDR_LW(gpa); - /* In non-fake mode, set func_id to 0xffff. - * Indicate the current func fake mode. - * The value of func_id is a fake func ID. - */ - if (cqm_handle->func_capability.fake_func_type == CQM_FAKE_FUNC_CHILD) - cmd.func_id = cqm_handle->func_attribute.func_global_idx; - else - cmd.func_id = 0xffff; + /* In non-fake mode, set func_id to 0xffff. */ + cmd.func_id = 0xffff; /* Mode 0 is hashed to 4 SMF engines (excluding PPF) by func ID. */ if (cqm_handle->func_capability.lb_mode == CQM_LB_MODE_NORMAL || diff --git a/drivers/scsi/spfc/hw/spfc_cqm_main.c b/drivers/scsi/spfc/hw/spfc_cqm_main.c index a4c8e60971b1..52cc2c7838e9 100644 --- a/drivers/scsi/spfc/hw/spfc_cqm_main.c +++ b/drivers/scsi/spfc/hw/spfc_cqm_main.c @@ -11,21 +11,8 @@ #include "sphw_crm.h" #include "sphw_hw.h" #include "sphw_hw_cfg.h" - #include "spfc_cqm_main.h" -static unsigned char cqm_lb_mode = CQM_LB_MODE_NORMAL; -module_param(cqm_lb_mode, byte, 0644); -MODULE_PARM_DESC(cqm_lb_mode, "for cqm lb mode (default=0xff)"); - -static unsigned char cqm_fake_mode = CQM_FAKE_MODE_DISABLE; -module_param(cqm_fake_mode, byte, 0644); -MODULE_PARM_DESC(cqm_fake_mode, "for cqm fake mode (default=0 disable)"); - -static unsigned char cqm_platform_mode = CQM_FPGA_MODE; -module_param(cqm_platform_mode, byte, 0644); -MODULE_PARM_DESC(cqm_platform_mode, "for cqm platform mode (default=0 FPGA)"); - s32 cqm3_init(void *ex_handle) { struct sphw_hwdev *handle = (struct sphw_hwdev *)ex_handle; @@ -63,19 +50,6 @@ s32 cqm3_init(void *ex_handle) goto err1; } - /* In FAKE mode, only the bitmap of the timer of the function is - * enabled, and resources are not initialized. Otherwise, the - * configuration of the fake function is overwritten. - */ - if (cqm_handle->func_capability.fake_func_type == CQM_FAKE_FUNC_CHILD_CONFLICT) { - if (sphw_func_tmr_bitmap_set(ex_handle, true) != CQM_SUCCESS) - cqm_err(handle->dev_hdl, "Timer start: enable timer bitmap failed\n"); - - handle->cqm_hdl = NULL; - kfree(cqm_handle); - return CQM_SUCCESS; - } - /* Initialize memory entries such as BAT, CLA, and bitmap. */ if (cqm_mem_init(ex_handle) != CQM_SUCCESS) { cqm_err(handle->dev_hdl, CQM_FUNCTION_FAIL(cqm_mem_init)); @@ -287,64 +261,6 @@ void cqm_service_capability_init(struct cqm_handle *cqm_handle, cqm_service_capability_init_fc(cqm_handle, (void *)service_capability); } -s32 cqm_get_fake_func_type(struct cqm_handle *cqm_handle) -{ - struct cqm_func_capability *func_cap = &cqm_handle->func_capability; - u32 parent_func, child_func_start, child_func_number, i; - u32 idx = cqm_handle->func_attribute.func_global_idx; - - /* Currently, only one set of fake configurations is implemented. - * fake_cfg_number = 1 - */ - for (i = 0; i < func_cap->fake_cfg_number; i++) { - parent_func = func_cap->fake_cfg[i].parent_func; - child_func_start = func_cap->fake_cfg[i].child_func_start; - child_func_number = func_cap->fake_cfg[i].child_func_number; - - if (idx == parent_func) { - return CQM_FAKE_FUNC_PARENT; - } else if ((idx >= child_func_start) && - (idx < (child_func_start + child_func_number))) { - return CQM_FAKE_FUNC_CHILD_CONFLICT; - } - } - - return CQM_FAKE_FUNC_NORMAL; -} - -s32 cqm_get_child_func_start(struct cqm_handle *cqm_handle) -{ - struct cqm_func_capability *func_cap = &cqm_handle->func_capability; - struct sphw_func_attr *func_attr = &cqm_handle->func_attribute; - u32 i; - - /* Currently, only one set of fake configurations is implemented. - * fake_cfg_number = 1 - */ - for (i = 0; i < func_cap->fake_cfg_number; i++) { - if (func_attr->func_global_idx == - func_cap->fake_cfg[i].parent_func) - return (s32)(func_cap->fake_cfg[i].child_func_start); - } - - return CQM_FAIL; -} - -s32 cqm_get_child_func_number(struct cqm_handle *cqm_handle) -{ - struct cqm_func_capability *func_cap = &cqm_handle->func_capability; - struct sphw_func_attr *func_attr = &cqm_handle->func_attribute; - u32 i; - - for (i = 0; i < func_cap->fake_cfg_number; i++) { - if (func_attr->func_global_idx == - func_cap->fake_cfg[i].parent_func) - return (s32)(func_cap->fake_cfg[i].child_func_number); - } - - return CQM_FAIL; -} - /* Set func_type in fake_cqm_handle to ppf, pf, or vf. */ void cqm_set_func_type(struct cqm_handle *cqm_handle) { @@ -358,41 +274,20 @@ void cqm_set_func_type(struct cqm_handle *cqm_handle) cqm_handle->func_attribute.func_type = CQM_VF; } -void cqm_lb_fake_mode_init(struct cqm_handle *cqm_handle) +void cqm_lb_fake_mode_init(struct cqm_handle *cqm_handle, struct service_cap *svc_cap) { struct cqm_func_capability *func_cap = &cqm_handle->func_capability; - struct cqm_fake_cfg *cfg = func_cap->fake_cfg; - func_cap->lb_mode = cqm_lb_mode; - func_cap->fake_mode = cqm_fake_mode; + func_cap->lb_mode = svc_cap->lb_mode; /* Initializing the LB Mode */ - if (func_cap->lb_mode == CQM_LB_MODE_NORMAL) { + if (func_cap->lb_mode == CQM_LB_MODE_NORMAL) func_cap->smf_pg = 0; - } else { - /* The LB mode is tailored on the FPGA. - * Only SMF0 and SMF2 are instantiated. - */ - if (cqm_platform_mode == CQM_FPGA_MODE) - func_cap->smf_pg = 0x5; - else - func_cap->smf_pg = 0xF; - } + else + func_cap->smf_pg = svc_cap->smf_pg; - /* Initializing the FAKE Mode */ - if (func_cap->fake_mode == CQM_FAKE_MODE_DISABLE) { - func_cap->fake_cfg_number = 0; - func_cap->fake_func_type = CQM_FAKE_FUNC_NORMAL; - } else { - func_cap->fake_cfg_number = 1; - - /* When configuring fake mode, ensure that the parent function - * cannot be contained in the child function; otherwise, the - * system will be initialized repeatedly. - */ - cfg[0].child_func_start = CQM_FAKE_CFUNC_START; - func_cap->fake_func_type = cqm_get_fake_func_type(cqm_handle); - } + func_cap->fake_cfg_number = 0; + func_cap->fake_func_type = CQM_FAKE_FUNC_NORMAL; } s32 cqm_capability_init(void *ex_handle) @@ -465,10 +360,9 @@ s32 cqm_capability_init(void *ex_handle) func_cap->gpa_check_enable = true; - cqm_lb_fake_mode_init(cqm_handle); + cqm_lb_fake_mode_init(cqm_handle, service_capability); cqm_info(handle->dev_hdl, "Cap init: lb_mode=%u\n", func_cap->lb_mode); cqm_info(handle->dev_hdl, "Cap init: smf_pg=%u\n", func_cap->smf_pg); - cqm_info(handle->dev_hdl, "Cap init: fake_mode=%u\n", func_cap->fake_mode); cqm_info(handle->dev_hdl, "Cap init: fake_func_type=%u\n", func_cap->fake_func_type); cqm_info(handle->dev_hdl, "Cap init: fake_cfg_number=%u\n", func_cap->fake_cfg_number); @@ -517,153 +411,6 @@ s32 cqm_capability_init(void *ex_handle) return err; } -void cqm_fake_uninit(struct cqm_handle *cqm_handle) -{ - u32 i; - - if (cqm_handle->func_capability.fake_func_type != - CQM_FAKE_FUNC_PARENT) - return; - - for (i = 0; i < CQM_FAKE_FUNC_MAX; i++) { - kfree(cqm_handle->fake_cqm_handle[i]); - cqm_handle->fake_cqm_handle[i] = NULL; - } -} - -s32 cqm_fake_init(struct cqm_handle *cqm_handle) -{ - struct sphw_hwdev *handle = cqm_handle->ex_handle; - struct cqm_func_capability *func_cap = NULL; - struct cqm_handle *fake_cqm_handle = NULL; - struct sphw_func_attr *func_attr = NULL; - s32 child_func_start, child_func_number; - u32 i; - - func_cap = &cqm_handle->func_capability; - if (func_cap->fake_func_type != CQM_FAKE_FUNC_PARENT) - return CQM_SUCCESS; - - child_func_start = cqm_get_child_func_start(cqm_handle); - if (child_func_start == CQM_FAIL) { - cqm_err(handle->dev_hdl, CQM_WRONG_VALUE(child_func_start)); - return CQM_FAIL; - } - - child_func_number = cqm_get_child_func_number(cqm_handle); - if (child_func_number == CQM_FAIL) { - cqm_err(handle->dev_hdl, CQM_WRONG_VALUE(child_func_number)); - return CQM_FAIL; - } - - for (i = 0; i < (u32)child_func_number; i++) { - fake_cqm_handle = kmalloc(sizeof(*fake_cqm_handle), GFP_KERNEL | __GFP_ZERO); - if (!fake_cqm_handle) { - cqm_err(handle->dev_hdl, - CQM_ALLOC_FAIL(fake_cqm_handle)); - goto err; - } - - /* Copy the attributes of the parent CQM handle to the child CQM - * handle and modify the values of function. - */ - memcpy(fake_cqm_handle, cqm_handle, sizeof(struct cqm_handle)); - func_attr = &fake_cqm_handle->func_attribute; - func_cap = &fake_cqm_handle->func_capability; - func_attr->func_global_idx = (u16)(child_func_start + i); - cqm_set_func_type(fake_cqm_handle); - func_cap->fake_func_type = CQM_FAKE_FUNC_CHILD; - cqm_info(handle->dev_hdl, "Fake func init: function[%u] type %d(0:PF,1:VF,2:PPF)\n", - func_attr->func_global_idx, func_attr->func_type); - - fake_cqm_handle->parent_cqm_handle = cqm_handle; - cqm_handle->fake_cqm_handle[i] = fake_cqm_handle; - } - - return CQM_SUCCESS; - -err: - cqm_fake_uninit(cqm_handle); - return CQM_FAIL; -} - -void cqm_fake_mem_uninit(struct cqm_handle *cqm_handle) -{ - struct sphw_hwdev *handle = cqm_handle->ex_handle; - struct cqm_handle *fake_cqm_handle = NULL; - s32 child_func_number; - u32 i; - - if (cqm_handle->func_capability.fake_func_type != CQM_FAKE_FUNC_PARENT) - return; - - child_func_number = cqm_get_child_func_number(cqm_handle); - if (child_func_number == CQM_FAIL) { - cqm_err(handle->dev_hdl, CQM_WRONG_VALUE(child_func_number)); - return; - } - - for (i = 0; i < (u32)child_func_number; i++) { - fake_cqm_handle = cqm_handle->fake_cqm_handle[i]; - cqm_object_table_uninit(fake_cqm_handle); - cqm_bitmap_uninit(fake_cqm_handle); - cqm_cla_uninit(fake_cqm_handle, CQM_BAT_ENTRY_MAX); - cqm_bat_uninit(fake_cqm_handle); - } -} - -s32 cqm_fake_mem_init(struct cqm_handle *cqm_handle) -{ - struct sphw_hwdev *handle = cqm_handle->ex_handle; - struct cqm_handle *fake_cqm_handle = NULL; - s32 child_func_number; - u32 i; - - if (cqm_handle->func_capability.fake_func_type != - CQM_FAKE_FUNC_PARENT) - return CQM_SUCCESS; - - child_func_number = cqm_get_child_func_number(cqm_handle); - if (child_func_number == CQM_FAIL) { - cqm_err(handle->dev_hdl, CQM_WRONG_VALUE(child_func_number)); - return CQM_FAIL; - } - - for (i = 0; i < (u32)child_func_number; i++) { - fake_cqm_handle = cqm_handle->fake_cqm_handle[i]; - - if (cqm_bat_init(fake_cqm_handle) != CQM_SUCCESS) { - cqm_err(handle->dev_hdl, - CQM_FUNCTION_FAIL(cqm_bat_init)); - goto err; - } - - if (cqm_cla_init(fake_cqm_handle) != CQM_SUCCESS) { - cqm_err(handle->dev_hdl, - CQM_FUNCTION_FAIL(cqm_cla_init)); - goto err; - } - - if (cqm_bitmap_init(fake_cqm_handle) != CQM_SUCCESS) { - cqm_err(handle->dev_hdl, - CQM_FUNCTION_FAIL(cqm_bitmap_init)); - goto err; - } - - if (cqm_object_table_init(fake_cqm_handle) != CQM_SUCCESS) { - cqm_err(handle->dev_hdl, - CQM_FUNCTION_FAIL(cqm_object_table_init)); - goto err; - } - } - - return CQM_SUCCESS; - -err: - cqm_fake_mem_uninit(cqm_handle); - return CQM_FAIL; -} - s32 cqm_mem_init(void *ex_handle) { struct sphw_hwdev *handle = (struct sphw_hwdev *)ex_handle; @@ -671,49 +418,35 @@ s32 cqm_mem_init(void *ex_handle) cqm_handle = (struct cqm_handle *)(handle->cqm_hdl); - if (cqm_fake_init(cqm_handle) != CQM_SUCCESS) { - cqm_err(handle->dev_hdl, CQM_FUNCTION_FAIL(cqm_fake_init)); - return CQM_FAIL; - } - - if (cqm_fake_mem_init(cqm_handle) != CQM_SUCCESS) { - cqm_err(handle->dev_hdl, CQM_FUNCTION_FAIL(cqm_fake_mem_init)); - goto err1; - } - if (cqm_bat_init(cqm_handle) != CQM_SUCCESS) { cqm_err(handle->dev_hdl, CQM_FUNCTION_FAIL(cqm_bat_init)); - goto err2; + return CQM_FAIL; } if (cqm_cla_init(cqm_handle) != CQM_SUCCESS) { cqm_err(handle->dev_hdl, CQM_FUNCTION_FAIL(cqm_cla_init)); - goto err3; + goto err1; } if (cqm_bitmap_init(cqm_handle) != CQM_SUCCESS) { cqm_err(handle->dev_hdl, CQM_FUNCTION_FAIL(cqm_bitmap_init)); - goto err4; + goto err2; } if (cqm_object_table_init(cqm_handle) != CQM_SUCCESS) { cqm_err(handle->dev_hdl, CQM_FUNCTION_FAIL(cqm_object_table_init)); - goto err5; + goto err3; } return CQM_SUCCESS; -err5: - cqm_bitmap_uninit(cqm_handle); -err4: - cqm_cla_uninit(cqm_handle, CQM_BAT_ENTRY_MAX); err3: - cqm_bat_uninit(cqm_handle); + cqm_bitmap_uninit(cqm_handle); err2: - cqm_fake_mem_uninit(cqm_handle); + cqm_cla_uninit(cqm_handle, CQM_BAT_ENTRY_MAX); err1: - cqm_fake_uninit(cqm_handle); + cqm_bat_uninit(cqm_handle); return CQM_FAIL; } @@ -728,8 +461,6 @@ void cqm_mem_uninit(void *ex_handle) cqm_bitmap_uninit(cqm_handle); cqm_cla_uninit(cqm_handle, CQM_BAT_ENTRY_MAX); cqm_bat_uninit(cqm_handle); - cqm_fake_mem_uninit(cqm_handle); - cqm_fake_uninit(cqm_handle); } s32 cqm_event_init(void *ex_handle) diff --git a/drivers/scsi/spfc/hw/spfc_cqm_main.h b/drivers/scsi/spfc/hw/spfc_cqm_main.h index 1b8cf8bdb3b7..cf10d7f5c339 100644 --- a/drivers/scsi/spfc/hw/spfc_cqm_main.h +++ b/drivers/scsi/spfc/hw/spfc_cqm_main.h @@ -328,9 +328,6 @@ void cqm_mem_uninit(void *ex_handle); s32 cqm_event_init(void *ex_handle); void cqm_event_uninit(void *ex_handle); u8 cqm_aeq_callback(void *ex_handle, u8 event, u8 *data); -s32 cqm_get_fake_func_type(struct cqm_handle *cqm_handle); -s32 cqm_get_child_func_start(struct cqm_handle *cqm_handle); -s32 cqm_get_child_func_number(struct cqm_handle *cqm_handle); s32 cqm3_init(void *ex_handle); void cqm3_uninit(void *ex_handle); diff --git a/drivers/scsi/spfc/hw/spfc_cqm_object.c b/drivers/scsi/spfc/hw/spfc_cqm_object.c index b895d37aebae..165794e9c7e5 100644 --- a/drivers/scsi/spfc/hw/spfc_cqm_object.c +++ b/drivers/scsi/spfc/hw/spfc_cqm_object.c @@ -155,8 +155,6 @@ struct cqm_qpc_mpt *cqm3_object_qpc_mpt_create(void *ex_handle, u32 service_type struct cqm_qpc_mpt_info *qpc_mpt_info = NULL; struct cqm_handle *cqm_handle = NULL; s32 ret = CQM_FAIL; - u32 relative_index; - u32 fake_func_id; CQM_PTR_CHECK_RET(ex_handle, NULL, CQM_PTR_NULL(ex_handle)); @@ -180,25 +178,6 @@ struct cqm_qpc_mpt *cqm3_object_qpc_mpt_create(void *ex_handle, u32 service_type return NULL; } - /* fake vf adaption, switch to corresponding VF. */ - if (cqm_handle->func_capability.fake_func_type == - CQM_FAKE_FUNC_PARENT) { - fake_func_id = index / cqm_handle->func_capability.qpc_number; - relative_index = index % cqm_handle->func_capability.qpc_number; - - cqm_info(handle->dev_hdl, "qpc create: fake_func_id=%u, relative_index=%u\n", - fake_func_id, relative_index); - - if ((s32)fake_func_id >= - cqm_get_child_func_number(cqm_handle)) { - cqm_err(handle->dev_hdl, CQM_WRONG_VALUE(fake_func_id)); - return NULL; - } - - index = relative_index; - cqm_handle = cqm_handle->fake_cqm_handle[fake_func_id]; - } - qpc_mpt_info = kmalloc(sizeof(*qpc_mpt_info), GFP_ATOMIC | __GFP_ZERO); CQM_PTR_CHECK_RET(qpc_mpt_info, NULL, CQM_ALLOC_FAIL(qpc_mpt_info)); -- 2.32.0

1 0

[PATCH openEuler-5.10 01/30] net: hns3: refactor hns3 makefile to support hns3_common module
by Zheng Zengkai 12 Jan '22

12 Jan '22

From: Jie Wang <wangjie125(a)huawei.com> mainline inclusion from mainline-br26_refactor2 commit 5f20be4e90e6 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4Q02P CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… ---------------------------------------------------------------------- Currently we plan to refactor PF and VF cmdq module. A new file folder hns3_common will be created to store new common APIs used by PF and VF cmdq module. Thus the PF and VF compilation process will both depends on the hns3_common. This may cause parallel building problems if we add a new makefile building unit. So this patch combined the PF and VF makefile scripts to the top level makefile to support the new hns3_common which will be created in the next patch. Signed-off-by: Jie Wang <wangjie125(a)huawei.com> Signed-off-by: Guangbin Huang <huangguangbin2(a)huawei.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Reviewed-by: Jian Shen <shenjian15(a)huawei.com> Reviewed-by: Yue Haibing <yuehaibing(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> --- drivers/net/ethernet/hisilicon/hns3/Makefile | 14 +++++++++++--- .../net/ethernet/hisilicon/hns3/hns3pf/Makefile | 12 ------------ .../net/ethernet/hisilicon/hns3/hns3vf/Makefile | 10 ---------- 3 files changed, 11 insertions(+), 25 deletions(-) delete mode 100644 drivers/net/ethernet/hisilicon/hns3/hns3pf/Makefile delete mode 100644 drivers/net/ethernet/hisilicon/hns3/hns3vf/Makefile diff --git a/drivers/net/ethernet/hisilicon/hns3/Makefile b/drivers/net/ethernet/hisilicon/hns3/Makefile index 7aa2fac76c5e..32e24e0945f5 100644 --- a/drivers/net/ethernet/hisilicon/hns3/Makefile +++ b/drivers/net/ethernet/hisilicon/hns3/Makefile @@ -4,9 +4,8 @@ # ccflags-y += -I$(srctree)/$(src) - -obj-$(CONFIG_HNS3) += hns3pf/ -obj-$(CONFIG_HNS3) += hns3vf/ +ccflags-y += -I$(srctree)/drivers/net/ethernet/hisilicon/hns3/hns3pf +ccflags-y += -I$(srctree)/drivers/net/ethernet/hisilicon/hns3/hns3vf obj-$(CONFIG_HNS3) += hnae3.o @@ -14,3 +13,12 @@ obj-$(CONFIG_HNS3_ENET) += hns3.o hns3-objs = hns3_enet.o hns3_ethtool.o hns3_debugfs.o hns3-$(CONFIG_HNS3_DCB) += hns3_dcbnl.o + +obj-$(CONFIG_HNS3_HCLGEVF) += hclgevf.o +hclgevf-objs = hns3vf/hclgevf_main.o hns3vf/hclgevf_cmd.o hns3vf/hclgevf_mbx.o hns3vf/hclgevf_devlink.o + +obj-$(CONFIG_HNS3_HCLGE) += hclge.o +hclge-objs = hns3pf/hclge_main.o hns3pf/hclge_cmd.o hns3pf/hclge_mdio.o hns3pf/hclge_tm.o \ + hns3pf/hclge_mbx.o hns3pf/hclge_err.o hns3pf/hclge_debugfs.o hns3pf/hclge_ptp.o hns3pf/hclge_devlink.o + +hclge-$(CONFIG_HNS3_DCB) += hns3pf/hclge_dcb.o diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/Makefile b/drivers/net/ethernet/hisilicon/hns3/hns3pf/Makefile deleted file mode 100644 index d1bf5c4c0abb..000000000000 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/Makefile +++ /dev/null @@ -1,12 +0,0 @@ -# SPDX-License-Identifier: GPL-2.0+ -# -# Makefile for the HISILICON network device drivers. -# - -ccflags-y := -I $(srctree)/drivers/net/ethernet/hisilicon/hns3 -ccflags-y += -I $(srctree)/$(src) - -obj-$(CONFIG_HNS3_HCLGE) += hclge.o -hclge-objs = hclge_main.o hclge_cmd.o hclge_mdio.o hclge_tm.o hclge_mbx.o hclge_err.o hclge_debugfs.o hclge_ptp.o hclge_devlink.o - -hclge-$(CONFIG_HNS3_DCB) += hclge_dcb.o diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3vf/Makefile b/drivers/net/ethernet/hisilicon/hns3/hns3vf/Makefile deleted file mode 100644 index 51ff7d86ee90..000000000000 --- a/drivers/net/ethernet/hisilicon/hns3/hns3vf/Makefile +++ /dev/null @@ -1,10 +0,0 @@ -# SPDX-License-Identifier: GPL-2.0+ -# -# Makefile for the HISILICON network device drivers. -# - -ccflags-y := -I $(srctree)/drivers/net/ethernet/hisilicon/hns3 -ccflags-y += -I $(srctree)/$(src) - -obj-$(CONFIG_HNS3_HCLGEVF) += hclgevf.o -hclgevf-objs = hclgevf_main.o hclgevf_cmd.o hclgevf_mbx.o hclgevf_devlink.o -- 2.20.1

1 29

[PATCH OLK-5.10 0/8] revert bcache prefetch temporarily
by Zheng Zengkai 12 Jan '22

12 Jan '22

This patch set introduce many conflicts while backporting mainline bcache patches, revert it temporarily. Zheng Zengkai (8): Revert "bcache: always record start time of a sample" Revert "bcache: do not collect data insert info created by write_moving" Revert "bcache: Rewrite patch to delay to invalidate cache data" Revert "bcache: Add a sample of userspace prefetch client" Revert "bcache: Delay to invalidate cache data in writearound write" Revert "bcache: inflight prefetch requests block overlapped normal requests" Revert "bcache: provide a switch to bypass all IO requests" Revert "bcache: add a framework to perform prefetch" Documentation/admin-guide/bcache.rst | 4 - drivers/md/bcache/Makefile | 2 +- drivers/md/bcache/acache.c | 586 --------------------------- drivers/md/bcache/acache.h | 79 ---- drivers/md/bcache/bcache.h | 3 - drivers/md/bcache/btree.c | 4 +- drivers/md/bcache/request.c | 161 +++----- drivers/md/bcache/request.h | 45 -- drivers/md/bcache/stats.c | 13 - drivers/md/bcache/stats.h | 3 - drivers/md/bcache/super.c | 6 - drivers/md/bcache/sysfs.c | 12 - include/trace/events/bcache.h | 22 - samples/acache_client/Makefile | 13 - samples/acache_client/connect.c | 144 ------- samples/acache_client/connect.h | 74 ---- samples/acache_client/main.c | 133 ------ 17 files changed, 59 insertions(+), 1245 deletions(-) delete mode 100644 drivers/md/bcache/acache.c delete mode 100644 drivers/md/bcache/acache.h delete mode 100644 samples/acache_client/Makefile delete mode 100644 samples/acache_client/connect.c delete mode 100644 samples/acache_client/connect.h delete mode 100644 samples/acache_client/main.c -- 2.20.1

2 9

[PATCH openEuler-5.10 01/17] tcp: Add some stub info for KABI consistency
by Zheng Zengkai 11 Jan '22

11 Jan '22

From: Baisong Zhong <zhongbaisong(a)huawei.com> hulk inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4PY1Q CVE: NA -------------------------------- We add stub info in some structures to maintain the consistency of KABI Signed-off-by: Baisong Zhong <zhongbaisong(a)huawei.com> Reviewed-by: Wei Yongjun <weiyongjun1(a)huawei.com> Reviewed-by: Yue Haibing <yuehaibing(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> --- include/linux/skbuff.h | 2 ++ include/net/ipv6.h | 2 +- include/net/sock.h | 3 +++ 3 files changed, 6 insertions(+), 1 deletion(-) diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 9e3a454d2377..d485f17ff33a 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -36,6 +36,7 @@ #include <linux/splice.h> #include <linux/in6.h> #include <linux/if_packet.h> +#include <linux/llist.h> #include <net/flow.h> #include <net/page_pool.h> #include <linux/kabi.h> @@ -732,6 +733,7 @@ struct sk_buff { }; struct rb_node rbnode; /* used in netem, ip4 defrag, and tcp stack */ struct list_head list; + struct llist_node ll_node; }; union { diff --git a/include/net/ipv6.h b/include/net/ipv6.h index bd1f396cc9c7..c0273ae50296 100644 --- a/include/net/ipv6.h +++ b/include/net/ipv6.h @@ -344,9 +344,9 @@ struct ipcm6_cookie { struct sockcm_cookie sockc; __s16 hlimit; __s16 tclass; + __u16 gso_size; __s8 dontfrag; struct ipv6_txoptions *opt; - __u16 gso_size; }; static inline void ipcm6_init(struct ipcm6_cookie *ipc6) diff --git a/include/net/sock.h b/include/net/sock.h index 712bb7b09f96..b3d451878640 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -63,6 +63,7 @@ #include <linux/atomic.h> #include <linux/refcount.h> +#include <linux/llist.h> #include <net/dst.h> #include <net/checksum.h> #include <net/tcp_states.h> @@ -405,6 +406,8 @@ struct sock { struct sk_buff *head; struct sk_buff *tail; } sk_backlog; + struct llist_head defer_list; + #define sk_rmem_alloc sk_backlog.rmem_alloc int sk_forward_alloc; -- 2.20.1

2 17

[PATCH OLK-5.10] arm64: Add memmap parameter and register pmem
by Zhuling 10 Jan '22

10 Jan '22

hulk inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4O31I?from=project-issue CVE: NA ------------------------------ Register pmem in arm64: Use memmap(memmap=nn[KMG]!ss[KMG]) reserve memory and e820(driver/nvdimm/e820.c) function to register persistent memory in arm64. when the kernel restart or update, the data in PMEM will not be lost and can be loaded faster. this is a general features. driver/nvdimm/e820.c: The function of this file is scan "iomem_resource" and take advantage of nvdimm resource discovery mechanism by registering a resource named "Persistent Memory (legacy)", this function doesn't depend on architecture. We will push the feature to linux kernel community and discuss to modify the file name. because people have a mistaken notion that the e820.c is depend on x86. If you want use this features, you need do as follows: 1.Reserve memory: add memmap to reserve memory in grub.cfg memmap=nn[KMG]!ss[KMG] exp:memmap=100K!0x1a0000000. 2.Insmod nd_e820.ko: modprobe nd_e820. 3.Check pmem device in /dev exp: /dev/pmem0 Signed-off-by: Zhuling <zhuling8(a)huawei.com> --- arch/arm64/Kconfig | 22 ++++++++++++++++ arch/arm64/kernel/Makefile | 1 + arch/arm64/kernel/pmem.c | 36 ++++++++++++++++++++++++++ arch/arm64/kernel/setup.c | 11 ++++++++ arch/arm64/mm/init.c | 63 ++++++++++++++++++++++++++++++++++++++++++++++ drivers/nvdimm/Kconfig | 6 +++++ drivers/nvdimm/Makefile | 1 + 7 files changed, 140 insertions(+) create mode 100644 arch/arm64/kernel/pmem.c diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig index 405e5ce..2231eac 100644 --- a/arch/arm64/Kconfig +++ b/arch/arm64/Kconfig @@ -1322,6 +1322,28 @@ config RODATA_FULL_DEFAULT_ENABLED This requires the linear region to be mapped down to pages, which may adversely affect performance in some cases. +config ARM64_PMEM_RESERVE + bool "Reserve memory for persistent storage" + default n + help + Use memmap=nn[KMG]!ss[KMG](memmap=100K!0x1a0000000) reserve + memory for persistent storage. + + Say y here to enable this feature. + +config ARM64_PMEM_LEGACY_DEVICE + bool "Create persistent storage" + depends on BLK_DEV + depends on LIBNVDIMM + select ARM64_PMEM_RESERVE + help + Use reserved memory for persistent storage when the kernel + restart or update. the data in PMEM will not be lost and + can be loaded faster. + + Say y if unsure. + + config ARM64_SW_TTBR0_PAN bool "Emulate Privileged Access Never using TTBR0_EL1 switching" help diff --git a/arch/arm64/kernel/Makefile b/arch/arm64/kernel/Makefile index 169d90f..f615325 100644 --- a/arch/arm64/kernel/Makefile +++ b/arch/arm64/kernel/Makefile @@ -68,6 +68,7 @@ obj-$(CONFIG_ARM64_PTR_AUTH) += pointer_auth.o obj-$(CONFIG_SHADOW_CALL_STACK) += scs.o obj-$(CONFIG_ARM64_MTE) += mte.o obj-$(CONFIG_MPAM) += mpam/ +obj-$(CONFIG_ARM64_PMEM_LEGACY_DEVICE) += pmem.o obj-y += vdso/ probes/ obj-$(CONFIG_COMPAT_VDSO) += vdso32/ diff --git a/arch/arm64/kernel/pmem.c b/arch/arm64/kernel/pmem.c new file mode 100644 index 0000000..d1efdc8 --- /dev/null +++ b/arch/arm64/kernel/pmem.c @@ -0,0 +1,36 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Copyright(c) 2021 Huawei Technologies Co., Ltd + * + * Derived from x86 and arm64 implement PMEM. + */ +#include <linux/platform_device.h> +#include <linux/init.h> +#include <linux/ioport.h> +#include <linux/module.h> + +static int found(struct resource *res, void *data) +{ + return 1; +} + +static int __init register_e820_pmem(void) +{ + struct platform_device *pdev; + int rc; + + rc = walk_iomem_res_desc(IORES_DESC_PERSISTENT_MEMORY_LEGACY, + IORESOURCE_MEM, 0, -1, NULL, found); + if (rc <= 0) + return 0; + + /* + * See drivers/nvdimm/e820.c for the implementation, this is + * simply here to trigger the module to load on demand. + */ + pdev = platform_device_alloc("e820_pmem", -1); + + return platform_device_add(pdev); +} +device_initcall(register_e820_pmem); + diff --git a/arch/arm64/kernel/setup.c b/arch/arm64/kernel/setup.c index 92d75e3..1a8a4d2 100644 --- a/arch/arm64/kernel/setup.c +++ b/arch/arm64/kernel/setup.c @@ -70,6 +70,11 @@ static int __init arm64_enable_cpu0_hotplug(char *str) __setup("arm64_cpu0_hotplug", arm64_enable_cpu0_hotplug); #endif +#ifdef CONFIG_ARM64_PMEM_RESERVE +extern struct resource pmem_res; +#endif + + phys_addr_t __fdt_pointer __initdata; /* @@ -305,6 +310,12 @@ static void __init request_standard_resources(void) } } +#ifdef CONFIG_ARM64_PMEM_RESERVE + if (pmem_res.end && pmem_res.start) + request_resource(&iomem_resource, &pmem_res); +#endif + + static int __init reserve_memblock_reserved_regions(void) { u64 i, j; diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c index 6ebfabd..3b6907f 100644 --- a/arch/arm64/mm/init.c +++ b/arch/arm64/mm/init.c @@ -56,6 +56,8 @@ s64 memstart_addr __ro_after_init = -1; EXPORT_SYMBOL(memstart_addr); +phys_addr_t start_at, mem_size; + #ifdef CONFIG_PIN_MEMORY struct resource pin_memory_resource = { .name = "Pin memory", @@ -111,6 +113,18 @@ static void __init reserve_pin_memory_res(void) */ phys_addr_t arm64_dma_phys_limit __ro_after_init; +static unsigned long long pmem_size, pmem_start; + +#ifdef CONFIG_ARM64_PMEM_RESERVE +struct resource pmem_res = { + .name = "Persistent Memory (legacy)", + .start = 0, + .end = 0, + .flags = IORESOURCE_MEM, + .desc = IORES_DESC_PERSISTENT_MEMORY_LEGACY +}; +#endif + #ifndef CONFIG_KEXEC_CORE static void __init reserve_crashkernel(void) { @@ -403,6 +417,26 @@ static int __init reserve_park_mem(void) return -EINVAL; } #endif +static bool __init is_mem_valid(unsigned long long mem_size, unsigned long long mem_start) +{ + if (!memblock_is_region_memory(mem_start, mem_size)) { + pr_warn("cannot reserve mem: region is not memory!\n"); + return false; + } + + if (memblock_is_region_reserved(mem_start, mem_size)) { + pr_warn("cannot reserve mem: region overlaps reserved memory!\n"); + return false; + } + + if (!IS_ALIGNED(mem_start, SZ_2M)) { + pr_warn("cannot reserve mem: base address is not 2MB aligned!\n"); + return false; + } + + return true; +} + static int need_remove_real_memblock __initdata; @@ -442,7 +476,17 @@ static int __init parse_memmap_one(char *p) start_at = memparse(p + 1, &p); memblock_reserve(start_at, mem_size); memblock_mark_memmap(start_at, mem_size); +<<<<<<< HEAD + } else if (*p == '!') { + start_at = memparse(p+1, &p); + + pmem_start = start_at; + pmem_size = mem_size; + + }else +======= } else +>>>>>>> 374db2be8805428b4f54b5ef793b0d3f5069c5f9 pr_info("Unrecognized memmap option, please check the parameter.\n"); return *p == '\0' ? 0 : -EINVAL; @@ -464,6 +508,20 @@ static int __init parse_memmap_opt(char *str) } early_param("memmap", parse_memmap_opt); +#ifdef CONFIG_ARM64_PMEM_RESERVE +static void __init reserve_pmem(void) +{ + if (!is_mem_valid(mem_size, start_at)) + return; + + memblock_remove(pmem_start, pmem_size); + pr_info("pmem reserved: 0x%016llx - 0x%016llx (%lld MB)\n", + pmem_start, pmem_start + pmem_size, pmem_size >> 20); + pmem_res.start = pmem_start; + pmem_res.end = pmem_start + pmem_size - 1; +} +#endif + void __init arm64_memblock_init(void) { const s64 linear_region_size = BIT(vabits_actual - 1); @@ -638,6 +696,11 @@ void __init bootmem_init(void) reserve_quick_kexec(); #endif +#ifdef CONFIG_ARM64_PMEM_RESERVE + reserve_pmem(); +#endif + + reserve_pin_memory_res(); memblock_dump_all(); diff --git a/drivers/nvdimm/Kconfig b/drivers/nvdimm/Kconfig index b7d1eb3..9567cca 100644 --- a/drivers/nvdimm/Kconfig +++ b/drivers/nvdimm/Kconfig @@ -132,3 +132,9 @@ config NVDIMM_TEST_BUILD infrastructure. endif + +config PMEM_LEGACY + tristate "Pmem_legacy" + select X86_PMEM_LEGACY if X86 + select ARM64_PMEM_LEGACY_DEVICE if ARM64 + diff --git a/drivers/nvdimm/Makefile b/drivers/nvdimm/Makefile index 0407753..6f8dc92 100644 --- a/drivers/nvdimm/Makefile +++ b/drivers/nvdimm/Makefile @@ -3,6 +3,7 @@ obj-$(CONFIG_LIBNVDIMM) += libnvdimm.o obj-$(CONFIG_BLK_DEV_PMEM) += nd_pmem.o obj-$(CONFIG_ND_BTT) += nd_btt.o obj-$(CONFIG_ND_BLK) += nd_blk.o +obj-$(CONFIG_PMEM_LEGACY) += nd_e820.o obj-$(CONFIG_OF_PMEM) += of_pmem.o obj-$(CONFIG_VIRTIO_PMEM) += virtio_pmem.o nd_virtio.o -- 2.9.5

3 3

[PATCH openEuler-5.10] BMA: Fix format string compile warning in arm32 builds
by Zheng Zengkai 10 Jan '22

10 Jan '22

driver inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4ETXO CVE: NA ----------------------------------------- Fix following build warnings in arm32 builds: drivers/net/ethernet/huawei/bma/edma_drv/bma_devintf.c: In function ‘bma_cdev_add_msg’: drivers/net/ethernet/huawei/bma/edma_drv/bma_pci.h:92:20: warning: format ‘%ld’ expects argument of type ‘long int’, but argument 5 has type ‘size_t {aka unsigned int}’ [-Wformat=] drivers/net/ethernet/huawei/bma/veth_drv/veth_hb.c: In function ‘veth_recv_pkt’: drivers/net/ethernet/huawei/bma/veth_drv/veth_hb.c:74:37: warning: format ‘%llx’ expects argument of type ‘long long unsigned int’, but argument 7 has type ‘dma_addr_t {aka unsigned int}’ [-Wformat=] Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> Reviewed-by: Kefeng Wang <wangkefeng.wang(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> --- .../bma/cdev_veth_drv/virtual_cdev_eth_net.c | 14 +++++++------- .../net/ethernet/huawei/bma/edma_drv/bma_devintf.c | 2 +- .../net/ethernet/huawei/bma/edma_drv/edma_host.c | 2 +- drivers/net/ethernet/huawei/bma/veth_drv/veth_hb.c | 4 ++-- 4 files changed, 11 insertions(+), 11 deletions(-) diff --git a/drivers/net/ethernet/huawei/bma/cdev_veth_drv/virtual_cdev_eth_net.c b/drivers/net/ethernet/huawei/bma/cdev_veth_drv/virtual_cdev_eth_net.c index 04ea55f4e0a2..d8c3ab655566 100644 --- a/drivers/net/ethernet/huawei/bma/cdev_veth_drv/virtual_cdev_eth_net.c +++ b/drivers/net/ethernet/huawei/bma/cdev_veth_drv/virtual_cdev_eth_net.c @@ -605,7 +605,7 @@ static int edma_veth_cut_tx_packet_send(struct edma_eth_dev_s *eth_dev, do_queue_rate_limit(eth_dev->ptx_queue); while (length > 0) { - LOG(DLOG_DEBUG, "length: %u/%lu", length, len); + LOG(DLOG_DEBUG, "length: %u/%zu", length, len); if (length > BSPPACKET_MTU_MAX) { /* fragment. */ @@ -1689,7 +1689,7 @@ static ssize_t cdev_copy_packet_to_user(struct edma_eth_dev_s *dev, start = dev->rx_packet[dev->rx_packet_head].packet + g_read_pos; LOG(DLOG_DEBUG, - "User needs %ld bytes, pos: %u, total len: %u, left: %ld.", + "User needs %zu bytes, pos: %u, total len: %u, left: %zd.", count, g_read_pos, dev->rx_packet[dev->rx_packet_head].len, left); if (left <= 0) { /* No more data in this message, retry. */ @@ -1721,7 +1721,7 @@ static ssize_t cdev_copy_packet_to_user(struct edma_eth_dev_s *dev, } LOG(DLOG_DEBUG, - "Copied bytes: %ld, pos: %d, buf len: %lu, free_packet: %d.", + "Copied bytes: %zd, pos: %d, buf len: %zu, free_packet: %d.", length, g_read_pos, count, free_packet); if (packet) { @@ -1807,11 +1807,11 @@ ssize_t cdev_read(struct file *filp, char __user *data, if (!data || count >= MAX_PACKET_LEN) return -EFAULT; - LOG(DLOG_DEBUG, "read begin, count: %ld, pos: %u.", count, g_read_pos); + LOG(DLOG_DEBUG, "read begin, count: %zu, pos: %u.", count, g_read_pos); length = cdev_copy_packet_to_user(dev, data, count); - LOG(DLOG_DEBUG, "read done, length: %ld, pos: %u.", length, g_read_pos); + LOG(DLOG_DEBUG, "read done, length: %zd, pos: %u.", length, g_read_pos); return length; } @@ -1837,7 +1837,7 @@ ssize_t cdev_write(struct file *filp, const char __user *data, g_peer_not_ready = 0; } - LOG(DLOG_DEBUG, "data length is %lu, pos: %u (%u/%u)", + LOG(DLOG_DEBUG, "data length is %zu, pos: %u (%u/%u)", count, g_read_pos, pdev->ptx_queue->pshmqhd_v->count, pdev->ptx_queue->pshmqhd_v->total); @@ -1859,4 +1859,4 @@ MODULE_DESCRIPTION("HUAWEI CDEV DRIVER"); MODULE_LICENSE("GPL"); module_init(edma_cdev_init); -module_exit(edma_cdev_exit); \ No newline at end of file +module_exit(edma_cdev_exit); diff --git a/drivers/net/ethernet/huawei/bma/edma_drv/bma_devintf.c b/drivers/net/ethernet/huawei/bma/edma_drv/bma_devintf.c index 7817f58f8635..acf6bbfc50ff 100644 --- a/drivers/net/ethernet/huawei/bma/edma_drv/bma_devintf.c +++ b/drivers/net/ethernet/huawei/bma/edma_drv/bma_devintf.c @@ -497,7 +497,7 @@ int bma_cdev_add_msg(void *handle, const char __user *msg, size_t msg_len) hdr->sub_type = priv->user.sub_type; hdr->user_id = priv->user.user_id; hdr->datalen = msg_len; - BMA_LOG(DLOG_DEBUG, "msg_len is %ld\n", msg_len); + BMA_LOG(DLOG_DEBUG, "msg_len is %zu\n", msg_len); if (copy_from_user(hdr->data, msg, msg_len)) { BMA_LOG(DLOG_ERROR, "copy_from_user error\n"); diff --git a/drivers/net/ethernet/huawei/bma/edma_drv/edma_host.c b/drivers/net/ethernet/huawei/bma/edma_drv/edma_host.c index 2d5f4ffd79d9..3525d41c865f 100644 --- a/drivers/net/ethernet/huawei/bma/edma_drv/edma_host.c +++ b/drivers/net/ethernet/huawei/bma/edma_drv/edma_host.c @@ -789,7 +789,7 @@ static int edma_host_send_msg(struct edma_host_s *edma_host) if (edma_host->msg_send_write > HOST_MAX_SEND_MBX_LEN - SIZE_OF_MBX_HDR) { BMA_LOG(DLOG_ERROR, - "Length of send message %u is larger than %lu\n", + "Length of send message %u is larger than %zu\n", edma_host->msg_send_write, HOST_MAX_SEND_MBX_LEN - SIZE_OF_MBX_HDR); edma_host->msg_send_write = 0; diff --git a/drivers/net/ethernet/huawei/bma/veth_drv/veth_hb.c b/drivers/net/ethernet/huawei/bma/veth_drv/veth_hb.c index 9681ce3bfc7b..240db31d7178 100644 --- a/drivers/net/ethernet/huawei/bma/veth_drv/veth_hb.c +++ b/drivers/net/ethernet/huawei/bma/veth_drv/veth_hb.c @@ -1488,8 +1488,8 @@ s32 veth_recv_pkt(struct bspveth_rxtx_q *prx_queue, int queue) skb->len, skb->protocol); VETH_LOG(DLOG_DEBUG, - "dma_p=0x%llx,dma_map=0x%llx,", - pbd_v->dma_p, dma_map); + "dma_p=0x%llx,dma_map=%pad,", + pbd_v->dma_p, &dma_map); VETH_LOG(DLOG_DEBUG, "skb=%p,skb->data=%p,skb->len=%d,tail=%d,shm_off=%d\n", -- 2.20.1

1 0

[PATCH openEuler-5.10 01/21] RDMA/hns: Enable stash feature of HIP09
by Zheng Zengkai 08 Jan '22

08 Jan '22

From: Yixing Liu <liuyixing1(a)huawei.com> mainline inclusion from mainline-v5.15-rc1 commit 260f64a40198 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4O23M CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… ---------------------------------------------------------------------------- The stash feature is enabled by default on HIP09. Fixes: f93c39bc9547 ("RDMA/hns: Add support for QP stash") Fixes: bfefae9f108d ("RDMA/hns: Add support for CQ stash") Link: https://lore.kernel.org/r/1629539607-33217-3-git-send-email-liangwenpeng@hu… Signed-off-by: Yixing Liu <liuyixing1(a)huawei.com> Signed-off-by: Wenpeng Liang <liangwenpeng(a)huawei.com> Signed-off-by: Jason Gunthorpe <jgg(a)nvidia.com> sigend-off-by: Guofeng Yue <yueguofeng(a)hisilicon.com> Reviewed-by: Yangyang Li <liyangyang20(a)huawei.com> Acked-by: Xie XiuQi <xiexiuqi(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> --- drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c index aea6a6a7b238..4f8738bda0a2 100644 --- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c +++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c @@ -2004,6 +2004,7 @@ static void set_default_caps(struct hns_roce_dev *hr_dev) caps->gid_table_len[0] = HNS_ROCE_V2_GID_INDEX_NUM; if (hr_dev->pci_dev->revision >= PCI_REVISION_ID_HIP09) { + caps->flags |= HNS_ROCE_CAP_FLAG_STASH; caps->max_sq_inline = HNS_ROCE_V3_MAX_SQ_INLINE; } else { caps->max_sq_inline = HNS_ROCE_V2_MAX_SQ_INLINE; -- 2.20.1

1 20

[PATCH OLK-5.10] BMA: Fix format string compile warning in arm32 builds
by Zheng Zengkai 08 Jan '22

08 Jan '22

driver inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4ETXO CVE: NA ----------------------------------------- Fix following build warnings in arm32 builds: drivers/net/ethernet/huawei/bma/edma_drv/bma_devintf.c: In function ‘bma_cdev_add_msg’: drivers/net/ethernet/huawei/bma/edma_drv/bma_pci.h:92:20: warning: format ‘%ld’ expects argument of type ‘long int’, but argument 5 has type ‘size_t {aka unsigned int}’ [-Wformat=] Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> --- .../bma/cdev_veth_drv/virtual_cdev_eth_net.c | 14 +++++++------- .../net/ethernet/huawei/bma/edma_drv/bma_devintf.c | 2 +- .../net/ethernet/huawei/bma/edma_drv/edma_host.c | 2 +- drivers/net/ethernet/huawei/bma/veth_drv/veth_hb.c | 2 +- 4 files changed, 10 insertions(+), 10 deletions(-) diff --git a/drivers/net/ethernet/huawei/bma/cdev_veth_drv/virtual_cdev_eth_net.c b/drivers/net/ethernet/huawei/bma/cdev_veth_drv/virtual_cdev_eth_net.c index 04ea55f4e0a2..99466848c034 100644 --- a/drivers/net/ethernet/huawei/bma/cdev_veth_drv/virtual_cdev_eth_net.c +++ b/drivers/net/ethernet/huawei/bma/cdev_veth_drv/virtual_cdev_eth_net.c @@ -605,7 +605,7 @@ static int edma_veth_cut_tx_packet_send(struct edma_eth_dev_s *eth_dev, do_queue_rate_limit(eth_dev->ptx_queue); while (length > 0) { - LOG(DLOG_DEBUG, "length: %u/%lu", length, len); + LOG(DLOG_DEBUG, "length: %u/%u", length, len); if (length > BSPPACKET_MTU_MAX) { /* fragment. */ @@ -1689,7 +1689,7 @@ static ssize_t cdev_copy_packet_to_user(struct edma_eth_dev_s *dev, start = dev->rx_packet[dev->rx_packet_head].packet + g_read_pos; LOG(DLOG_DEBUG, - "User needs %ld bytes, pos: %u, total len: %u, left: %ld.", + "User needs %d bytes, pos: %u, total len: %u, left: %d.", count, g_read_pos, dev->rx_packet[dev->rx_packet_head].len, left); if (left <= 0) { /* No more data in this message, retry. */ @@ -1721,7 +1721,7 @@ static ssize_t cdev_copy_packet_to_user(struct edma_eth_dev_s *dev, } LOG(DLOG_DEBUG, - "Copied bytes: %ld, pos: %d, buf len: %lu, free_packet: %d.", + "Copied bytes: %d, pos: %d, buf len: %u, free_packet: %d.", length, g_read_pos, count, free_packet); if (packet) { @@ -1807,11 +1807,11 @@ ssize_t cdev_read(struct file *filp, char __user *data, if (!data || count >= MAX_PACKET_LEN) return -EFAULT; - LOG(DLOG_DEBUG, "read begin, count: %ld, pos: %u.", count, g_read_pos); + LOG(DLOG_DEBUG, "read begin, count: %d, pos: %u.", count, g_read_pos); length = cdev_copy_packet_to_user(dev, data, count); - LOG(DLOG_DEBUG, "read done, length: %ld, pos: %u.", length, g_read_pos); + LOG(DLOG_DEBUG, "read done, length: %d, pos: %u.", length, g_read_pos); return length; } @@ -1837,7 +1837,7 @@ ssize_t cdev_write(struct file *filp, const char __user *data, g_peer_not_ready = 0; } - LOG(DLOG_DEBUG, "data length is %lu, pos: %u (%u/%u)", + LOG(DLOG_DEBUG, "data length is %u, pos: %u (%u/%u)", count, g_read_pos, pdev->ptx_queue->pshmqhd_v->count, pdev->ptx_queue->pshmqhd_v->total); @@ -1859,4 +1859,4 @@ MODULE_DESCRIPTION("HUAWEI CDEV DRIVER"); MODULE_LICENSE("GPL"); module_init(edma_cdev_init); -module_exit(edma_cdev_exit); \ No newline at end of file +module_exit(edma_cdev_exit); diff --git a/drivers/net/ethernet/huawei/bma/edma_drv/bma_devintf.c b/drivers/net/ethernet/huawei/bma/edma_drv/bma_devintf.c index 7817f58f8635..2d63e44f0ed2 100644 --- a/drivers/net/ethernet/huawei/bma/edma_drv/bma_devintf.c +++ b/drivers/net/ethernet/huawei/bma/edma_drv/bma_devintf.c @@ -497,7 +497,7 @@ int bma_cdev_add_msg(void *handle, const char __user *msg, size_t msg_len) hdr->sub_type = priv->user.sub_type; hdr->user_id = priv->user.user_id; hdr->datalen = msg_len; - BMA_LOG(DLOG_DEBUG, "msg_len is %ld\n", msg_len); + BMA_LOG(DLOG_DEBUG, "msg_len is %d\n", msg_len); if (copy_from_user(hdr->data, msg, msg_len)) { BMA_LOG(DLOG_ERROR, "copy_from_user error\n"); diff --git a/drivers/net/ethernet/huawei/bma/edma_drv/edma_host.c b/drivers/net/ethernet/huawei/bma/edma_drv/edma_host.c index 2d5f4ffd79d9..a89a4160dedd 100644 --- a/drivers/net/ethernet/huawei/bma/edma_drv/edma_host.c +++ b/drivers/net/ethernet/huawei/bma/edma_drv/edma_host.c @@ -789,7 +789,7 @@ static int edma_host_send_msg(struct edma_host_s *edma_host) if (edma_host->msg_send_write > HOST_MAX_SEND_MBX_LEN - SIZE_OF_MBX_HDR) { BMA_LOG(DLOG_ERROR, - "Length of send message %u is larger than %lu\n", + "Length of send message %u is larger than %u\n", edma_host->msg_send_write, HOST_MAX_SEND_MBX_LEN - SIZE_OF_MBX_HDR); edma_host->msg_send_write = 0; diff --git a/drivers/net/ethernet/huawei/bma/veth_drv/veth_hb.c b/drivers/net/ethernet/huawei/bma/veth_drv/veth_hb.c index 9681ce3bfc7b..7b511c05c2e9 100644 --- a/drivers/net/ethernet/huawei/bma/veth_drv/veth_hb.c +++ b/drivers/net/ethernet/huawei/bma/veth_drv/veth_hb.c @@ -1488,7 +1488,7 @@ s32 veth_recv_pkt(struct bspveth_rxtx_q *prx_queue, int queue) skb->len, skb->protocol); VETH_LOG(DLOG_DEBUG, - "dma_p=0x%llx,dma_map=0x%llx,", + "dma_p=0x%llx,dma_map=0x%x,", pbd_v->dma_p, dma_map); VETH_LOG(DLOG_DEBUG, -- 2.20.1

2 1

[PATCH openEuler-5.10 001/125] kabi: Generalize naming of kabi helper macros
by Zheng Zengkai 07 Jan '22

07 Jan '22

hulk inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4K3S5 -------------------------- Generalize naming of some kabi helper macros. Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> Reviewed-by: Xie XiuQi <xiexiuqi(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> --- include/linux/kabi.h | 38 +++++++++++++++++++------------------- 1 file changed, 19 insertions(+), 19 deletions(-) diff --git a/include/linux/kabi.h b/include/linux/kabi.h index 713f63cf56cb..0bc7ca2483f4 100644 --- a/include/linux/kabi.h +++ b/include/linux/kabi.h @@ -151,7 +151,7 @@ * approaches can (and often are) combined. * * To use this for 'struct foo' (the "base structure"), define a new - * structure called 'struct foo_rh'; this new struct is called "auxiliary + * structure called 'struct foo_resvd'; this new struct is called "auxiliary * structure". Then add KABI_AUX_EMBED or KABI_AUX_PTR to the end * of the base structure. The argument is the name of the base structure, * without the 'struct' keyword. @@ -174,7 +174,7 @@ * end. Note the auxiliary structure cannot be shrunk in size later (i.e., * fields cannot be removed, only deprecated). Any code accessing fields * from the aux struct must guard the access using the KABI_AUX macro. - * The access itself is then done via a '_rh' field in the base struct. + * The access itself is then done via a '_resvd' field in the base struct. * * The auxiliary structure is not guaranteed for access by modules unless * explicitly commented as such in the declaration of the aux struct @@ -182,7 +182,7 @@ * * Example: * - * struct foo_rh { + * struct foo_resvd { * int newly_added; * }; * @@ -194,20 +194,20 @@ * void use(struct foo *f) * { * if (KABI_AUX(f, foo, newly_added)) - * f->_rh->newly_added = 123; + * f->_resvd->newly_added = 123; * else * // the field 'newly_added' is not present in the passed * // struct, fall back to old behavior * f->big_hammer = true; * } * - * static struct foo_rh my_foo_rh { + * static struct foo_resvd my_foo_resvd { * .newly_added = 0; * } * * static struct foo my_foo = { * .big_hammer = false, - * ._rh = &my_foo_rh, + * ._resvd = &my_foo_resvd, * KABI_AUX_INIT_SIZE(foo) * }; * @@ -218,7 +218,7 @@ * * Example: * - * struct foo_rh { + * struct foo_resvd { * }; * * struct foo { @@ -385,7 +385,7 @@ _Static_assert(__alignof__(struct{_new;}) <= __alignof__(struct{_orig;}), \ __FILE__ ":" __stringify(__LINE__) ": " __stringify(_orig) " is not aligned the same as " __stringify(_new) KABI_ALIGN_WARNING); \ } -# define __ABI_CHECK_SIZE(_item, _size) \ +# define __KABI_CHECK_SIZE(_item, _size) \ _Static_assert(sizeof(struct{_item;}) <= _size, \ __FILE__ ":" __stringify(__LINE__) ": " __stringify(_item) " is larger than the reserved size (" __stringify(_size) " bytes)" KABI_ALIGN_WARNING) #else @@ -451,14 +451,14 @@ }) #define _KABI_AUX_PTR(_struct) \ - size_t _struct##_size_rh; \ - _KABI_EXCLUDE(struct _struct##_rh *_rh) + size_t _struct##_size_resvd; \ + _KABI_EXCLUDE(struct _struct##_resvd *_resvd) #define KABI_AUX_PTR(_struct) \ _KABI_AUX_PTR(_struct); #define _KABI_AUX_EMBED(_struct) \ - size_t _struct##_size_rh; \ - _KABI_EXCLUDE(struct _struct##_rh _rh) + size_t _struct##_size_resvd; \ + _KABI_EXCLUDE(struct _struct##_resvd _resvd) #define KABI_AUX_EMBED(_struct) \ _KABI_AUX_EMBED(_struct); @@ -468,7 +468,7 @@ /* * KABI_AUX_SET_SIZE calculates and sets the size of the extended struct and - * stores it in the size_rh field for structs that are dynamically allocated. + * stores it in the size_resvd field for structs that are dynamically allocated. * This macro MUST be called when expanding a base struct with * KABI_SIZE_AND_EXTEND, and it MUST be called from the allocation site * regardless of being allocated in the kernel or a module. @@ -476,28 +476,28 @@ * a semicolon is necessary at the end of the line where it is invoked. */ #define KABI_AUX_SET_SIZE(_name, _struct) ({ \ - (_name)->_struct##_size_rh = sizeof(struct _struct##_rh); \ + (_name)->_struct##_size_resvd = sizeof(struct _struct##_resvd); \ }) /* * KABI_AUX_INIT_SIZE calculates and sets the size of the extended struct and - * stores it in the size_rh field for structs that are statically allocated. + * stores it in the size_resvd field for structs that are statically allocated. * This macro MUST be called when expanding a base struct with * KABI_SIZE_AND_EXTEND, and it MUST be called from the declaration site * regardless of being allocated in the kernel or a module. */ #define KABI_AUX_INIT_SIZE(_struct) \ - ._struct##_size_rh = sizeof(struct _struct##_rh), + ._struct##_size_resvd = sizeof(struct _struct##_resvd), /* * KABI_AUX verifies allocated memory exists. This MUST be called to - * verify that memory in the _rh struct is valid, and can be called + * verify that memory in the _resvd struct is valid, and can be called * regardless if KABI_SIZE_AND_EXTEND or KABI_SIZE_AND_EXTEND_PTR is * used. */ #define KABI_AUX(_ptr, _struct, _field) ({ \ - size_t __off = offsetof(struct _struct##_rh, _field); \ - (_ptr)->_struct##_size_rh > __off ? true : false; \ + size_t __off = offsetof(struct _struct##_resvd, _field); \ + (_ptr)->_struct##_size_resvd > __off ? true : false; \ }) #endif /* _LINUX_KABI_H */ -- 2.20.1

1 124

Subscribe
by Song Zhi 07 Jan '22

07 Jan '22

Subscribe <https://aka.ms/AAb9ysg>

1 0

[PATCH openEuler-5.10 1/2] arm64: Add support for memmap kernel parameters
by Zheng Zengkai 07 Jan '22

07 Jan '22

From: Peng Liu <liupeng256(a)huawei.com> hulk inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4NYPZ CVE: NA ------------------------------------------------- Add support for memmap kernel parameters for ARM64. The three below modes are supported: memmap=exactmap Enable setting of an exact memory map, as specified by the user. memmap=nn[KMG]@ss[KMG] Force usage of a specific region of memory. memmap=nn[KMG]$ss[KMG] Region of memory to be reserved is from ss to ss+nn, the region must be in the range of existed memory, otherwise will be ignored. If users set memmap=exactmap before memmap=nn[KMG]@ss[KMG], they will get the exact memory specified by memmap=nn[KMG]@ss[KMG]. For example, on one machine with 4GB memory, "memmap=exactmap memmap=1G@1G" will make kernel use the memory from 1GB to 2GB only. Signed-off-by: Peng Liu <liupeng256(a)huawei.com> Reviewed-by: Kefeng Wang <wangkefeng.wang(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> --- .../admin-guide/kernel-parameters.txt | 9 ++- arch/arm64/mm/init.c | 59 +++++++++++++++++++ 2 files changed, 65 insertions(+), 3 deletions(-) diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt index de8f7d447295..64be32ba4373 100644 --- a/Documentation/admin-guide/kernel-parameters.txt +++ b/Documentation/admin-guide/kernel-parameters.txt @@ -2794,8 +2794,8 @@ option. See Documentation/admin-guide/mm/memory-hotplug.rst. - memmap=exactmap [KNL,X86] Enable setting of an exact - E820 memory map, as specified by the user. + memmap=exactmap [KNL,X86,ARM64] Enable setting of an exact + E820 and ARM64 memory map, as specified by the user. Such memmap=exactmap lines can be constructed based on BIOS output or other requirements. See the memmap=nn@ss option description. @@ -2806,7 +2806,8 @@ If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG], which limits max address to nn[KMG]. Multiple different regions can be specified, - comma delimited. + comma delimited, example as follows is not supported to + ARM64. Example: memmap=100M@2G,100M#3G,1G!1024G @@ -2817,6 +2818,8 @@ memmap=nn[KMG]$ss[KMG] [KNL,ACPI] Mark specific memory as reserved. Region of memory to be reserved is from ss to ss+nn. + For ARM64, reserved memory must be in the range of + existed memory. Example: Exclude memory from 0x18690000-0x1869ffff memmap=64K$0x18690000 or diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c index e8d446164c76..f59546d3b0de 100644 --- a/arch/arm64/mm/init.c +++ b/arch/arm64/mm/init.c @@ -404,6 +404,65 @@ static int __init reserve_park_mem(void) } #endif +static int need_remove_real_memblock __initdata; + +static int __init parse_memmap_one(char *p) +{ + char *oldp; + u64 start_at, mem_size; + + if (!p) + return -EINVAL; + + if (!strncmp(p, "exactmap", 8)) { + need_remove_real_memblock = 1; + return -EINVAL; + } + + oldp = p; + mem_size = memparse(p, &p); + if (p == oldp) + return -EINVAL; + + if (!mem_size) + return -EINVAL; + + if (*p == '@') { + start_at = memparse(p + 1, &p); + /* + * use the exactmap defined by nn[KMG]@ss[KMG], remove + * memblock populated by DT etc. + */ + if (need_remove_real_memblock) { + need_remove_real_memblock = 0; + memblock_remove(0, ULLONG_MAX); + } + memblock_add(start_at, mem_size); + } else if (*p == '$') { + start_at = memparse(p + 1, &p); + memblock_reserve(start_at, mem_size); + } else + pr_info("Unrecognized memmap option, please check the parameter.\n"); + + return *p == '\0' ? 0 : -EINVAL; +} + +static int __init parse_memmap_opt(char *str) +{ + while (str) { + char *k = strchr(str, ','); + + if (k) + *k++ = 0; + + parse_memmap_one(str); + str = k; + } + + return 0; +} +early_param("memmap", parse_memmap_opt); + void __init arm64_memblock_init(void) { const s64 linear_region_size = BIT(vabits_actual - 1); -- 2.20.1

1 1

[PATCH openEuler-1.0-LTS] Add new module parameters:time out
by Yanling Song 07 Jan '22

07 Jan '22

Ramaxel inclusion category: features bugzilla: https://gitee.com/openeuler/kernel/issues/I4ON8F CVE: NA Changes: 1. Split scmd_tmout_nonpt into two parameters: scmd_tmout_vd/scmd_tmout_rawdisk 2. Return -ETIME instead of -EINVAL when command is timeout. 3. Add one module parameters: max_io_force. Signed-off-by: Yanling Song <songyl(a)ramaxel.com> Reviewed-by: Jiang Yu<yujiang(a)ramaxel.com> --- drivers/scsi/spraid/spraid_main.c | 96 +++++++++++++++---------------- 1 file changed, 45 insertions(+), 51 deletions(-) diff --git a/drivers/scsi/spraid/spraid_main.c b/drivers/scsi/spraid/spraid_main.c index 519b39f44e91..7069582d741a 100644 --- a/drivers/scsi/spraid/spraid_main.c +++ b/drivers/scsi/spraid/spraid_main.c @@ -42,10 +42,17 @@ static u32 admin_tmout = 60; module_param(admin_tmout, uint, 0644); MODULE_PARM_DESC(admin_tmout, "admin commands timeout (seconds)"); -static u32 scmd_tmout_nonpt = 180; -module_param(scmd_tmout_nonpt, uint, 0644); -MODULE_PARM_DESC(scmd_tmout_nonpt, - "scsi commands timeout for rawdisk&raid(seconds)"); +static u32 scmd_tmout_rawdisk = 180; +module_param(scmd_tmout_rawdisk, uint, 0644); +MODULE_PARM_DESC(scmd_tmout_rawdisk, "scsi commands timeout for rawdisk(seconds)"); + +static u32 scmd_tmout_vd = 180; +module_param(scmd_tmout_vd, uint, 0644); +MODULE_PARM_DESC(scmd_tmout_vd, "scsi commands timeout for vd(seconds)"); + +static bool max_io_force; +module_param(max_io_force, bool, 0644); +MODULE_PARM_DESC(max_io_force, "force max_hw_sectors_kb = 1024, default false(performance first)"); static int ioq_depth_set(const char *val, const struct kernel_param *kp); static const struct kernel_param_ops ioq_depth_ops = { @@ -78,7 +85,7 @@ static unsigned char log_debug_switch; module_param_cb(log_debug_switch, &log_debug_switch_ops, &log_debug_switch, 0644); MODULE_PARM_DESC(log_debug_switch, - "set log state, default non-zero for switch on"); + "set log state, default zero for switch off"); static int small_pool_num_set(const char *val, const struct kernel_param *kp) { @@ -132,7 +139,6 @@ static struct workqueue_struct *spraid_wq; #define SPRAID_DRV_VERSION "1.0.0.0" #define ADMIN_TIMEOUT (admin_tmout * HZ) -#define ADMIN_ERR_TIMEOUT 32757 #define SPRAID_WAIT_ABNL_CMD_TIMEOUT (3 * 2) @@ -242,13 +248,6 @@ static int spraid_pci_enable(struct spraid_dev *hdev) goto disable; } - ret = pci_alloc_irq_vectors(pdev, 1, 1, PCI_IRQ_ALL_TYPES); - if (ret < 0) { - dev_err(hdev->dev, - "Allocate one IRQ for setup admin channel failed\n"); - goto disable; - } - hdev->cap = lo_hi_readq(hdev->bar + SPRAID_REG_CAP); hdev->ioq_depth = min_t(u32, SPRAID_CAP_MQES(hdev->cap) + 1, io_queue_depth); @@ -261,13 +260,21 @@ static int spraid_pci_enable(struct spraid_dev *hdev) maskbit); maskbit = SPRAID_DMA_MSK_BIT_MAX; } - if (dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(maskbit))) { + if (dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(maskbit)) && + dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(32))) { dev_err(hdev->dev, "set dma mask and coherent failed\n"); goto disable; } dev_info(hdev->dev, "set dma mask[%llu] success\n", maskbit); + ret = pci_alloc_irq_vectors(pdev, 1, 1, PCI_IRQ_ALL_TYPES); + + if (ret < 0) { + dev_err(hdev->dev, "Allocate one IRQ for setup admin channel failed\n"); + goto disable; + } + pci_enable_pcie_error_reporting(pdev); pci_save_state(pdev); @@ -840,6 +847,9 @@ static void spraid_map_status(struct spraid_iod *iod, struct scsi_cmnd *scmd, break; default: set_host_byte(scmd, DID_BAD_TARGET); + dev_warn(iod->spraidq->hdev->dev, "[%s] cid[%d] qid[%d];" + "bad status[0x%x]\n", __func__, cqe->cmd_id, + le16_to_cpu(cqe->sq_id), le16_to_cpu(cqe->status)); break; } } @@ -981,32 +991,17 @@ static void spraid_slave_destroy(struct scsi_device *sdev) static int spraid_slave_configure(struct scsi_device *sdev) { - u16 idx; - unsigned int timeout = scmd_tmout_nonpt * HZ; + unsigned int timeout = scmd_tmout_rawdisk * HZ; struct spraid_dev *hdev = shost_priv(sdev->host); struct spraid_sdev_hostdata *hostdata = sdev->hostdata; u32 max_sec = sdev->host->max_sectors; if (hostdata) { - idx = hostdata->hdid - 1; - if (sdev->channel == hdev->devices[idx].channel && - sdev->id == le16_to_cpu(hdev->devices[idx].target) && - sdev->lun < hdev->devices[idx].lun) { - if (SPRAID_DEV_INFO_ATTR_PT(hdev->devices[idx].attr)) - timeout = 30 * HZ; - else - timeout = scmd_tmout_nonpt * HZ; - max_sec = le16_to_cpu(hdev->devices[idx].max_io_kb) - << 1; - } else { - dev_err(hdev->dev, - "[%s] err, sdev->channel:id:lun[%d:%d:%lld];" - "devices[%d], channel:target:lun[%d:%d:%d]\n", - __func__, sdev->channel, sdev->id, sdev->lun, - idx, hdev->devices[idx].channel, - hdev->devices[idx].target, - hdev->devices[idx].lun); - } + if (SPRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + timeout = scmd_tmout_vd * HZ; + else if (SPRAID_DEV_INFO_ATTR_RAWDISK(hostdata->attr)) + timeout = scmd_tmout_rawdisk * HZ; + max_sec = hostdata->max_io_kb << 1; } else { dev_err(hdev->dev, "[%s] err, sdev->hostdata is null\n", __func__); @@ -1018,11 +1013,12 @@ static int spraid_slave_configure(struct scsi_device *sdev) if ((max_sec == 0) || (max_sec > sdev->host->max_sectors)) max_sec = sdev->host->max_sectors; - dev_info(hdev->dev, - "[%s] sdev->channel:id:lun[%d:%d:%lld];" - " scmd_timeout[%d]s, maxsec[%d]\n", - __func__, sdev->channel, sdev->id, - sdev->lun, timeout / HZ, max_sec); + if (!max_io_force) + blk_queue_max_hw_sectors(sdev->request_queue, max_sec); + + dev_info(hdev->dev, "[%s] sdev->channel:id:lun[%d:%d:%lld];" + "scmd_timeout[%d]s, maxsec[%d]\n", __func__, sdev->channel, + sdev->id, sdev->lun, timeout / HZ, max_sec); return 0; } @@ -1677,7 +1673,7 @@ static int spraid_submit_admin_sync_cmd(struct spraid_dev *hdev, cmd->usr_cmd.opcode, cmd->usr_cmd.info_0.subopcode); WRITE_ONCE(adm_cmd->state, SPRAID_CMD_TIMEOUT); spraid_put_cmd(hdev, adm_cmd, SPRAID_CMD_ADM); - return -EINVAL; + return -ETIME; } if (result0) @@ -2630,9 +2626,6 @@ static int spraid_user_admin_cmd(struct spraid_dev *hdev, struct bsg_job *job) return -EBUSY; } - dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x] init\n", - __func__, cmd->opcode, cmd->info_0.subopcode); - memset(&admin_cmd, 0, sizeof(admin_cmd)); admin_cmd.common.opcode = cmd->opcode; admin_cmd.common.flags = cmd->flags; @@ -2659,10 +2652,11 @@ static int spraid_user_admin_cmd(struct spraid_dev *hdev, struct bsg_job *job) memcpy(job->reply, result, sizeof(result)); } - dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x];" - " status[0x%x] result0[0x%x] result1[0x%x]\n", - __func__, cmd->opcode, cmd->info_0.subopcode, - status, result[0], result[1]); + if (status) + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x];" + " status[0x%x] result0[0x%x] result1[0x%x]\n", + __func__, cmd->opcode, cmd->info_0.subopcode, + status, result[0], result[1]); spraid_bsg_unmap_data(hdev, job); @@ -2742,7 +2736,7 @@ static int spraid_submit_ioq_sync_cmd(struct spraid_dev *hdev, (le32_to_cpu(cmd->common.cdw3[0]) & 0xffff)); WRITE_ONCE(pt_cmd->state, SPRAID_CMD_TIMEOUT); spraid_put_cmd(hdev, pt_cmd, SPRAID_CMD_IOPT); - return -EINVAL; + return -ETIME; } if (result && reslen) { @@ -3140,7 +3134,7 @@ static int spraid_abort_handler(struct scsi_cmnd *scmd) dev_warn(hdev->dev, "cid[%d] qid[%d] timeout, aborting\n", cid, hwq); ret = spraid_send_abort_cmd(hdev, hostdata->hdid, hwq, cid); - if (ret != ADMIN_ERR_TIMEOUT) { + if (ret != -ETIME) { ret = spraid_wait_abnl_cmd_done(iod); if (ret) { dev_warn(hdev->dev, "cid[%d] qid[%d] abort failed;" @@ -3623,7 +3617,7 @@ static int spraid_bsg_host_dispatch(struct bsg_job *job) struct spraid_bsg_request *bsg_req = job->request; int ret = 0; - dev_info(hdev->dev, "[%s] msgcode[%d], msglen[%d], timeout[%d];" + dev_log_dbg(hdev->dev, "[%s] msgcode[%d], msglen[%d], timeout[%d];" " req_nsge[%d], req_len[%d]\n", __func__, bsg_req->msgcode, job->request_len, rq->timeout, job->request_payload.sg_cnt, -- 2.27.0

1 0

[PATCH openEuler-1.0-LTS 1/2] arm64/mpam: fix mpam probe error for wrong init order
by Yang Yingliang 06 Jan '22

06 Jan '22

From: Xingang Wang <wangxingang5(a)huawei.com> ascend inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I49RB2 CVE: NA --------------------------------------------------- The mpam init procedure failed when probe with ACPI: [ 1.148657 ] ACPI MPAM: No CPU has cache with PPTT reference 0x72 [ 1.148658 ] ACPI MPAM: All CPUs must be online to probe mpam. [ 1.148660 ] ACPI MPAM: discovery failed: -19 This is because mpam need to be probed after all cpus be online, the arm_mpam_driver_init must be called after cacheinfo_sysfs_init, so the device_initcall should be replaced with device_initcall_sync. Fixes: b45bdb5a8604 ("arm64/mpam: add device tree support for mpam initialization") Signed-off-by: Xingang Wang <wangxingang5(a)huawei.com> Reviewed-by: Wang ShaoBo <bobo.shaobowang(a)huawei.com> Reviewed-by: Cheng Jian <cj.chengjian(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- arch/arm64/kernel/mpam/mpam_device.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm64/kernel/mpam/mpam_device.c b/arch/arm64/kernel/mpam/mpam_device.c index 4910e99348798..c7b5c50d431bd 100644 --- a/arch/arm64/kernel/mpam/mpam_device.c +++ b/arch/arm64/kernel/mpam/mpam_device.c @@ -1876,4 +1876,4 @@ static int __init arm_mpam_driver_init(void) * We want to run after cacheinfo_sysfs_init() has caused the cacheinfo * structures to be populated. That runs as a device_initcall. */ -device_initcall(arm_mpam_driver_init); +device_initcall_sync(arm_mpam_driver_init); -- 2.25.1

1 1

[PATCH openEuler-5.10 0/2] Add CONFIG_KABI_RESERVE to control KABI
by Zheng Zengkai 06 Jan '22

06 Jan '22

Add CONFIG_KABI_RESERVE to control KABI padding reserve and enable it by default. Zheng Zengkai (2): KABI: Add CONFIG_KABI_RESERVE to control KABI padding reserve openeuler_defconfig: Enable CONFIG_KABI_RESERVE for x86 and arm64 Kconfig | 8 ++++++++ arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + include/linux/kabi.h | 6 +++++- 4 files changed, 15 insertions(+), 1 deletion(-) -- 2.20.1

1 2

[PATCH OLK-5.10] KABI: Fix allmodconfig build error
by Zheng Zengkai 06 Jan '22

06 Jan '22

hulk inclusion bugzilla: https://gitee.com/openeuler/kernel/issues/I4MZU1 CVE: NA --------------------------- For x86 platform, make allmodconfig & make -j64, following two build errors are reported. First error: - build failed: - In file included from <command-line>:32:0: ./usr/include/asm/bootparam.h:45:10: fatal error: linux/kabi.h: No such file or directory #include <linux/kabi.h> ^~~~~~~~~~~~~~ compilation terminated. make[2]: *** [usr/include/asm/bootparam.hdrtest] Error 1 Second error: ./arch/x86/include/asm/paravirt_types.h:198:2: error: expected specifier-qualifier-list before ‘KABI_RESERVE’ KABI_RESERVE(1) ^~~~~~~~~~~~ ./arch/x86/include/asm/paravirt_types.h:286:2: error: expected specifier-qualifier-list before ‘KABI_RESERVE’ KABI_RESERVE(1) ^~~~~~~~~~~~ ./arch/x86/include/asm/paravirt_types.h:309:2: error: expected specifier-qualifier-list before ‘KABI_RESERVE’ KABI_RESERVE(1) ^~~~~~~~~~~~ make[1]: *** [scripts/Makefile.build:117: arch/x86/kernel/asm-offsets.s] Error 1 To fix first error, reverts commit 3ba63bacfc bootparam: Add kabi_reserve in bootparam. To fix second error, add include file kabi.h to arch/x86/include/asm/paravirt_types.h. Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> --- arch/x86/include/asm/paravirt_types.h | 2 ++ arch/x86/include/uapi/asm/bootparam.h | 2 -- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/x86/include/asm/paravirt_types.h b/arch/x86/include/asm/paravirt_types.h index 98eb135b1888..17a2308ef3af 100644 --- a/arch/x86/include/asm/paravirt_types.h +++ b/arch/x86/include/asm/paravirt_types.h @@ -2,6 +2,8 @@ #ifndef _ASM_X86_PARAVIRT_TYPES_H #define _ASM_X86_PARAVIRT_TYPES_H +#include <linux/kabi.h> + /* Bitmask of what can be clobbered: usually at least eax. */ #define CLBR_NONE 0 #define CLBR_EAX (1 << 0) diff --git a/arch/x86/include/uapi/asm/bootparam.h b/arch/x86/include/uapi/asm/bootparam.h index b1dce768d073..600a141c8805 100644 --- a/arch/x86/include/uapi/asm/bootparam.h +++ b/arch/x86/include/uapi/asm/bootparam.h @@ -42,7 +42,6 @@ #include <linux/types.h> #include <linux/screen_info.h> #include <linux/apm_bios.h> -#include <linux/kabi.h> #include <linux/edd.h> #include <asm/ist.h> #include <video/edid.h> @@ -103,7 +102,6 @@ struct setup_header { __u32 init_size; __u32 handover_offset; __u32 kernel_info_offset; - KABI_RESERVE(1) } __attribute__((packed)); struct sys_desc_table { -- 2.20.1

1 0

[PATCH openEuler-1.0-LTS 1/3] bpf: Use dedicated bpf_trace_printk event instead of trace_printk()
by Liu Xinpeng 06 Jan '22

06 Jan '22

From: Alan Maguire <alan.maguire(a)oracle.com> mainline inclusion from mainline-v5.10.78 commit ac5a72ea5c8989871e61f6bb0852e0f91de51ebe category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I4PKDW CVE: NA -------------------------------- The bpf helper bpf_trace_printk() uses trace_printk() under the hood. This leads to an alarming warning message originating from trace buffer allocation which occurs the first time a program using bpf_trace_printk() is loaded. We can instead create a trace event for bpf_trace_printk() and enable it in-kernel when/if we encounter a program using the bpf_trace_printk() helper. With this approach, trace_printk() is not used directly and no warning message appears. This work was started by Steven (see Link) and finished by Alan; added Steven's Signed-off-by with his permission. Signed-off-by: Steven Rostedt (VMware) <rostedt(a)goodmis.org> Signed-off-by: Alan Maguire <alan.maguire(a)oracle.com> Signed-off-by: Alexei Starovoitov <ast(a)kernel.org> Acked-by: Andrii Nakryiko <andriin(a)fb.com> Link: https://lore.kernel.org/r/20200628194334.6238b933@oasis.local.home Link: https://lore.kernel.org/bpf/1594641154-18897-2-git-send-email-alan.maguire@… Signed-off-by: Liu Xinpeng <liuxp11(a)chinatelecom.cn> # openEuler_contributor Signed-off-by: Ctyun Kernel <ctyuncommiter01(a)chinatelecom.cn> # openEuler_contributor --- kernel/trace/Makefile | 2 ++ kernel/trace/bpf_trace.c | 42 +++++++++++++++++++++++++++++++++++++----- kernel/trace/bpf_trace.h | 34 ++++++++++++++++++++++++++++++++++ 3 files changed, 73 insertions(+), 5 deletions(-) create mode 100644 kernel/trace/bpf_trace.h diff --git a/kernel/trace/Makefile b/kernel/trace/Makefile index f81dadbc7..846b6b0 100644 --- a/kernel/trace/Makefile +++ b/kernel/trace/Makefile @@ -28,6 +28,8 @@ ifdef CONFIG_GCOV_PROFILE_FTRACE GCOV_PROFILE := y endif +CFLAGS_bpf_trace.o := -I$(src) + CFLAGS_trace_benchmark.o := -I$(src) CFLAGS_trace_events_filter.o := -I$(src) diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c index e0b0e92..f359f79 100644 --- a/kernel/trace/bpf_trace.c +++ b/kernel/trace/bpf_trace.c @@ -11,12 +11,16 @@ #include <linux/uaccess.h> #include <linux/ctype.h> #include <linux/kprobes.h> +#include <linux/spinlock.h> #include <linux/syscalls.h> #include <linux/error-injection.h> #include "trace_probe.h" #include "trace.h" +#define CREATE_TRACE_POINTS +#include "bpf_trace.h" + #ifdef CONFIG_MODULES struct bpf_trace_module { struct module *module; @@ -318,6 +322,30 @@ static const struct bpf_func_proto *bpf_get_probe_write_proto(void) return &bpf_probe_write_user_proto; } +static DEFINE_RAW_SPINLOCK(trace_printk_lock); + +#define BPF_TRACE_PRINTK_SIZE 1024 + +static inline __printf(1, 0) int bpf_do_trace_printk(const char *fmt, ...) +{ + static char buf[BPF_TRACE_PRINTK_SIZE]; + unsigned long flags; + va_list ap; + int ret; + + raw_spin_lock_irqsave(&trace_printk_lock, flags); + va_start(ap, fmt); + ret = vsnprintf(buf, sizeof(buf), fmt, ap); + va_end(ap); + /* vsnprintf() will not append null for zero-length strings */ + if (ret == 0) + buf[0] = '\0'; + trace_bpf_trace_printk(buf); + raw_spin_unlock_irqrestore(&trace_printk_lock, flags); + + return ret; +} + /* * Only limited trace_printk() conversion specifiers allowed: * %d %i %u %x %ld %li %lu %lx %lld %lli %llu %llx %p %s @@ -408,8 +436,7 @@ static const struct bpf_func_proto *bpf_get_probe_write_proto(void) */ #define __BPF_TP_EMIT() __BPF_ARG3_TP() #define __BPF_TP(...) \ - __trace_printk(0 /* Fake ip */, \ - fmt, ##__VA_ARGS__) + bpf_do_trace_printk(fmt, ##__VA_ARGS__) #define __BPF_ARG1_TP(...) \ ((mod[0] == 2 || (mod[0] == 1 && __BITS_PER_LONG == 64)) \ @@ -446,10 +473,15 @@ static const struct bpf_func_proto *bpf_get_probe_write_proto(void) const struct bpf_func_proto *bpf_get_trace_printk_proto(void) { /* - * this program might be calling bpf_trace_printk, - * so allocate per-cpu printk buffers + * This program might be calling bpf_trace_printk, + * so enable the associated bpf_trace/bpf_trace_printk event. + * Repeat this each time as it is possible a user has + * disabled bpf_trace_printk events. By loading a program + * calling bpf_trace_printk() however the user has expressed + * the intent to see such events. */ - trace_printk_init_buffers(); + if (trace_set_clr_event("bpf_trace", "bpf_trace_printk", 1)) + pr_warn_ratelimited("could not enable bpf_trace_printk events"); return &bpf_trace_printk_proto; } diff --git a/kernel/trace/bpf_trace.h b/kernel/trace/bpf_trace.h new file mode 100644 index 00000000..9acbc11 --- /dev/null +++ b/kernel/trace/bpf_trace.h @@ -0,0 +1,34 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM bpf_trace + +#if !defined(_TRACE_BPF_TRACE_H) || defined(TRACE_HEADER_MULTI_READ) + +#define _TRACE_BPF_TRACE_H + +#include <linux/tracepoint.h> + +TRACE_EVENT(bpf_trace_printk, + + TP_PROTO(const char *bpf_string), + + TP_ARGS(bpf_string), + + TP_STRUCT__entry( + __string(bpf_string, bpf_string) + ), + + TP_fast_assign( + __assign_str(bpf_string, bpf_string); + ), + + TP_printk("%s", __get_str(bpf_string)) +); + +#endif /* _TRACE_BPF_TRACE_H */ + +#undef TRACE_INCLUDE_PATH +#define TRACE_INCLUDE_PATH . +#define TRACE_INCLUDE_FILE bpf_trace + +#include <trace/define_trace.h> -- 1.8.3.1

1 2

[PATCH OLK-5.10] scsi:spraid: use bsg module to replace with ioctrl
by Yanling Song 06 Jan '22

06 Jan '22

Ramaxel inclusion category: features bugzilla: https://gitee.com/openeuler/kernel/issues/I4JXCG CVE: NA Changes: 1. Use bsg module to replace with ioctrl 2. Split scmd_tmout_nonpt into two parameters: scmd_tmout_vd/scmd_tmout_rawdisk 3. Return -ETIME instead of -EINVAL when command is timeout. 4. Add one module parameters: max_io_force. 5. Remove some unnecessary module parameters. 6. Report disks by the order of channel/target id. 7. Add host_reset handler. 8. Use get_unaligned_be24. Signed-off-by: Yanling Song <songyl(a)ramaxel.com> Reviewed-by: Jiang Yu<yujiang(a)ramaxel.com> --- drivers/scsi/spraid/Kconfig | 6 +- drivers/scsi/spraid/spraid.h | 142 ++- drivers/scsi/spraid/spraid_main.c | 1633 +++++++++++++++-------------- 3 files changed, 968 insertions(+), 813 deletions(-) diff --git a/drivers/scsi/spraid/Kconfig b/drivers/scsi/spraid/Kconfig index 83962efaab07..bfbba3db8db0 100644 --- a/drivers/scsi/spraid/Kconfig +++ b/drivers/scsi/spraid/Kconfig @@ -5,7 +5,9 @@ config RAMAXEL_SPRAID tristate "Ramaxel spraid Adapter" depends on PCI && SCSI + select BLK_DEV_BSGLIB depends on ARM64 || X86_64 - default m help - This driver supports Ramaxel spraid driver. + This driver supports Ramaxel SPRxxx serial + raid controller, which has PCIE Gen4 interface + with host and supports SAS/SATA Hdd/ssd. diff --git a/drivers/scsi/spraid/spraid.h b/drivers/scsi/spraid/spraid.h index da46d8e1b4b6..983d7af2faa8 100644 --- a/drivers/scsi/spraid/spraid.h +++ b/drivers/scsi/spraid/spraid.h @@ -1,4 +1,5 @@ /* SPDX-License-Identifier: GPL-2.0 */ +/* Copyright(c) 2021 Ramaxel Memory Technology, Ltd */ #ifndef __SPRAID_H_ #define __SPRAID_H_ @@ -24,7 +25,7 @@ #define SENSE_SIZE(depth) ((depth) * SCSI_SENSE_BUFFERSIZE) #define SPRAID_AQ_DEPTH 128 -#define SPRAID_NR_AEN_COMMANDS 1 +#define SPRAID_NR_AEN_COMMANDS 16 #define SPRAID_AQ_BLK_MQ_DEPTH (SPRAID_AQ_DEPTH - SPRAID_NR_AEN_COMMANDS) #define SPRAID_AQ_MQ_TAG_DEPTH (SPRAID_AQ_BLK_MQ_DEPTH - 1) @@ -44,7 +45,7 @@ #define SMALL_POOL_SIZE 256 #define MAX_SMALL_POOL_NUM 16 -#define MAX_CMD_PER_DEV 32 +#define MAX_CMD_PER_DEV 64 #define MAX_CDB_LEN 32 #define SPRAID_UP_TO_MULTY4(x) (((x) + 4) & (~0x03)) @@ -53,7 +54,7 @@ #define PCI_VENDOR_ID_RAMAXEL_LOGIC 0x1E81 -#define SPRAID_SERVER_DEVICE_HAB_DID 0x2100 +#define SPRAID_SERVER_DEVICE_HBA_DID 0x2100 #define SPRAID_SERVER_DEVICE_RAID_DID 0x2200 #define IO_6_DEFAULT_TX_LEN 256 @@ -142,11 +143,15 @@ enum { enum { SPRAID_AEN_DEV_CHANGED = 0x00, + SPRAID_AEN_FW_ACT_START = 0x01, SPRAID_AEN_HOST_PROBING = 0x10, }; enum { - SPRAID_AEN_TIMESYN = 0x07 + SPRAID_AEN_TIMESYN = 0x00, + SPRAID_AEN_FW_ACT_FINISH = 0x02, + SPRAID_AEN_EVENT_MIN = 0x80, + SPRAID_AEN_EVENT_MAX = 0xff, }; enum { @@ -175,6 +180,16 @@ enum spraid_state { SPRAID_DEAD, }; +enum { + SPRAID_CARD_HBA, + SPRAID_CARD_RAID, +}; + +enum spraid_cmd_type { + SPRAID_CMD_ADM, + SPRAID_CMD_IOPT, +}; + struct spraid_completion { __le32 result; union { @@ -217,8 +232,6 @@ struct spraid_dev { struct dma_pool *prp_page_pool; struct dma_pool *prp_small_pool[MAX_SMALL_POOL_NUM]; mempool_t *iod_mempool; - struct blk_mq_tag_set admin_tagset; - struct request_queue *admin_q; void __iomem *bar; u32 max_qid; u32 num_vecs; @@ -232,23 +245,27 @@ struct spraid_dev { u32 ctrl_config; u32 online_queues; u64 cap; - struct device ctrl_device; - struct cdev cdev; int instance; struct spraid_ctrl_info *ctrl_info; struct spraid_dev_info *devices; - struct spraid_ioq_ptcmd *ioq_ptcmds; + struct spraid_cmd *adm_cmds; + struct list_head adm_cmd_list; + spinlock_t adm_cmd_lock; + + struct spraid_cmd *ioq_ptcmds; struct list_head ioq_pt_list; spinlock_t ioq_pt_lock; - struct work_struct aen_work; struct work_struct scan_work; struct work_struct timesyn_work; struct work_struct reset_work; + struct work_struct fw_act_work; enum spraid_state state; spinlock_t state_lock; + + struct request_queue *bsg_queue; }; struct spraid_sgl_desc { @@ -347,6 +364,35 @@ struct spraid_get_info { __u32 rsvd12[4]; }; +struct spraid_usr_cmd { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + union { + struct { + __le16 subopcode; + __le16 rsvd1; + } info_0; + __le32 cdw2; + }; + union { + struct { + __le16 data_len; + __le16 param_len; + } info_1; + __le32 cdw3; + }; + __u64 metadata; + union spraid_data_ptr dptr; + __le32 cdw10; + __le32 cdw11; + __le32 cdw12; + __le32 cdw13; + __le32 cdw14; + __le32 cdw15; +}; + enum { SPRAID_CMD_FLAG_SGL_METABUF = (1 << 6), SPRAID_CMD_FLAG_SGL_METASEG = (1 << 7), @@ -393,6 +439,7 @@ struct spraid_admin_command { struct spraid_get_info get_info; struct spraid_abort_cmd abort; struct spraid_reset_cmd reset; + struct spraid_usr_cmd usr_cmd; }; }; @@ -456,9 +503,6 @@ struct spraid_ioq_command { }; }; -#define SPRAID_IOCTL_RESET_CMD _IOWR('N', 0x80, struct spraid_passthru_common_cmd) -#define SPRAID_IOCTL_ADMIN_CMD _IOWR('N', 0x41, struct spraid_passthru_common_cmd) - struct spraid_passthru_common_cmd { __u8 opcode; __u8 flags; @@ -494,8 +538,6 @@ struct spraid_passthru_common_cmd { __u32 result1; }; -#define SPRAID_IOCTL_IOQ_CMD _IOWR('N', 0x42, struct spraid_ioq_passthru_cmd) - struct spraid_ioq_passthru_cmd { __u8 opcode; __u8 flags; @@ -560,7 +602,21 @@ struct spraid_ioq_passthru_cmd { __u32 result1; }; -struct spraid_ioq_ptcmd { +struct spraid_bsg_request { + u32 msgcode; + u32 control; + union { + struct spraid_passthru_common_cmd admcmd; + struct spraid_ioq_passthru_cmd ioqcmd; + }; +}; + +enum { + SPRAID_BSG_ADM, + SPRAID_BSG_IOQ, +}; + +struct spraid_cmd { int qid; int cid; u32 result0; @@ -572,14 +628,6 @@ struct spraid_ioq_ptcmd { struct list_head list; }; -struct spraid_admin_request { - struct spraid_admin_command *cmd; - u32 result0; - u32 result1; - u16 flags; - u16 status; -}; - struct spraid_queue { struct spraid_dev *hdev; spinlock_t sq_lock; /* spinlock for lock handling */ @@ -607,7 +655,6 @@ struct spraid_queue { }; struct spraid_iod { - struct spraid_admin_request req; struct spraid_queue *spraidq; enum spraid_cmd_state state; int npages; @@ -623,13 +670,51 @@ struct spraid_iod { }; #define SPRAID_DEV_INFO_ATTR_BOOT(attr) ((attr) & 0x01) -#define SPRAID_DEV_INFO_ATTR_HDD(attr) ((attr) & 0x02) +#define SPRAID_DEV_INFO_ATTR_VD(attr) (((attr) & 0x02) == 0x0) #define SPRAID_DEV_INFO_ATTR_PT(attr) (((attr) & 0x22) == 0x02) #define SPRAID_DEV_INFO_ATTR_RAWDISK(attr) ((attr) & 0x20) #define SPRAID_DEV_INFO_FLAG_VALID(flag) ((flag) & 0x01) #define SPRAID_DEV_INFO_FLAG_CHANGE(flag) ((flag) & 0x02) +#define BGTASK_TYPE_REBUILD 4 +#define USR_CMD_READ 0xc2 +#define USR_CMD_RDLEN 0x1000 +#define USR_CMD_VDINFO 0x704 +#define USR_CMD_BGTASK 0x504 +#define VDINFO_PARAM_LEN 0x04 + +struct spraid_vd_info { + __u8 name[32]; + __le16 id; + __u8 rg_id; + __u8 rg_level; + __u8 sg_num; + __u8 sg_disk_num; + __u8 vd_status; + __u8 vd_type; + __u8 rsvd1[4056]; +}; + +#define MAX_REALTIME_BGTASK_NUM 32 + +struct bgtask_info { + __u8 type; + __u8 progress; + __u8 rate; + __u8 rsvd0; + __le16 vd_id; + __le16 time_left; + __u8 rsvd1[4]; +}; + +struct spraid_bgtask { + __u8 sw; + __u8 task_num; + __u8 rsvd[6]; + struct bgtask_info bgtask[MAX_REALTIME_BGTASK_NUM]; +}; + struct spraid_dev_info { __le32 hdid; __le16 target; @@ -649,6 +734,11 @@ struct spraid_dev_list { struct spraid_sdev_hostdata { u32 hdid; + u16 max_io_kb; + u8 attr; + u8 flag; + u8 rg_id; + u8 rsvd[3]; }; #endif diff --git a/drivers/scsi/spraid/spraid_main.c b/drivers/scsi/spraid/spraid_main.c index a0a75ecb0027..c6d2a0b8e35e 100644 --- a/drivers/scsi/spraid/spraid_main.c +++ b/drivers/scsi/spraid/spraid_main.c @@ -1,8 +1,8 @@ // SPDX-License-Identifier: GPL-2.0 -/* - * Linux spraid device driver - * Copyright(c) 2021 Ramaxel Memory Technology, Ltd - */ +/* Copyright(c) 2021 Ramaxel Memory Technology, Ltd */ + +/* Ramaxel Raid SPXXX Series Linux Driver */ + #define pr_fmt(fmt) "spraid: " fmt #include <linux/sched/signal.h> @@ -23,6 +23,9 @@ #include <linux/debugfs.h> #include <linux/io-64-nonatomic-lo-hi.h> #include <linux/blkdev.h> +#include <linux/bsg-lib.h> +#include <asm/unaligned.h> +#include <linux/sort.h> #include <scsi/scsi.h> #include <scsi/scsi_cmnd.h> @@ -31,27 +34,24 @@ #include <scsi/scsi_transport.h> #include <scsi/scsi_dbg.h> + #include "spraid.h" static u32 admin_tmout = 60; module_param(admin_tmout, uint, 0644); MODULE_PARM_DESC(admin_tmout, "admin commands timeout (seconds)"); -static u32 scmd_tmout_pt = 30; -module_param(scmd_tmout_pt, uint, 0644); -MODULE_PARM_DESC(scmd_tmout_pt, "scsi commands timeout for passthrough(seconds)"); +static u32 scmd_tmout_rawdisk = 180; +module_param(scmd_tmout_rawdisk, uint, 0644); +MODULE_PARM_DESC(scmd_tmout_rawdisk, "scsi commands timeout for rawdisk(seconds)"); -static u32 scmd_tmout_nonpt = 180; -module_param(scmd_tmout_nonpt, uint, 0644); -MODULE_PARM_DESC(scmd_tmout_nonpt, "scsi commands timeout for rawdisk&raid(seconds)"); +static u32 scmd_tmout_vd = 180; +module_param(scmd_tmout_vd, uint, 0644); +MODULE_PARM_DESC(scmd_tmout_vd, "scsi commands timeout for vd(seconds)"); -static u32 wait_abl_tmout = 3; -module_param(wait_abl_tmout, uint, 0644); -MODULE_PARM_DESC(wait_abl_tmout, "wait abnormal io timeout(seconds)"); - -static bool use_sgl_force; -module_param(use_sgl_force, bool, 0644); -MODULE_PARM_DESC(use_sgl_force, "force IO use sgl format, default false"); +static bool max_io_force; +module_param(max_io_force, bool, 0644); +MODULE_PARM_DESC(max_io_force, "force max_hw_sectors_kb = 1024, default false(performance first)"); static int ioq_depth_set(const char *val, const struct kernel_param *kp); static const struct kernel_param_ops ioq_depth_ops = { @@ -106,16 +106,20 @@ static const struct kernel_param_ops small_pool_num_ops = { .get = param_get_byte, }; +/* It was found that the spindlock of a single pool conflicts + * a lot with multiple CPUs.So multiple pools are introduced + * to reduce the conflictions. + */ static unsigned char small_pool_num = 4; module_param_cb(small_pool_num, &small_pool_num_ops, &small_pool_num, 0644); MODULE_PARM_DESC(small_pool_num, "set prp small pool num, default 4, MAX 16"); static void spraid_free_queue(struct spraid_queue *spraidq); static void spraid_handle_aen_notice(struct spraid_dev *hdev, u32 result); -static void spraid_handle_aen_vs(struct spraid_dev *hdev, u32 result); +static void spraid_handle_aen_vs(struct spraid_dev *hdev, u32 result, u32 result1); static DEFINE_IDA(spraid_instance_ida); -static dev_t spraid_chr_devt; + static struct class *spraid_class; #define SPRAID_CAP_TIMEOUT_UNIT_MS (HZ / 2) @@ -131,9 +135,8 @@ static struct workqueue_struct *spraid_wq; #define SPRAID_DRV_VERSION "1.0.0.0" #define ADMIN_TIMEOUT (admin_tmout * HZ) -#define ADMIN_ERR_TIMEOUT 32757 -#define SPRAID_WAIT_ABNL_CMD_TIMEOUT (wait_abl_tmout * 2) +#define SPRAID_WAIT_ABNL_CMD_TIMEOUT (3 * 2) #define SPRAID_DMA_MSK_BIT_MAX 64 @@ -147,6 +150,13 @@ enum FW_STAT_CODE { FW_STAT_NEED_RETRY }; +static const char * const raid_levels[] = {"0", "1", "5", "6", "10", "50", "60", "NA"}; + +static const char * const raid_states[] = { + "NA", "NORMAL", "FAULT", "DEGRADE", "NOT_FORMATTED", "FORMATTING", "SANITIZING", + "INITIALIZING", "INITIALIZE_FAIL", "DELETING", "DELETE_FAIL", "WRITE_PROTECT" +}; + static int ioq_depth_set(const char *val, const struct kernel_param *kp) { int n = 0; @@ -231,12 +241,6 @@ static int spraid_pci_enable(struct spraid_dev *hdev) goto disable; } - ret = pci_alloc_irq_vectors(pdev, 1, 1, PCI_IRQ_ALL_TYPES); - if (ret < 0) { - dev_err(hdev->dev, "Allocate one IRQ for setup admin channel failed\n"); - goto disable; - } - hdev->cap = lo_hi_readq(hdev->bar + SPRAID_REG_CAP); hdev->ioq_depth = min_t(u32, SPRAID_CAP_MQES(hdev->cap) + 1, io_queue_depth); hdev->db_stride = 1 << SPRAID_CAP_STRIDE(hdev->cap); @@ -246,13 +250,20 @@ static int spraid_pci_enable(struct spraid_dev *hdev) dev_err(hdev->dev, "err, dma mask invalid[%llu], set to default\n", maskbit); maskbit = SPRAID_DMA_MSK_BIT_MAX; } - if (dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(maskbit))) { + if (dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(maskbit)) && + dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(32))) { dev_err(hdev->dev, "set dma mask and coherent failed\n"); goto disable; } dev_info(hdev->dev, "set dma mask[%llu] success\n", maskbit); + ret = pci_alloc_irq_vectors(pdev, 1, 1, PCI_IRQ_ALL_TYPES); + if (ret < 0) { + dev_err(hdev->dev, "Allocate one IRQ for setup admin channel failed\n"); + goto disable; + } + pci_enable_pcie_error_reporting(pdev); pci_save_state(pdev); @@ -263,12 +274,6 @@ static int spraid_pci_enable(struct spraid_dev *hdev) return ret; } -static inline -struct spraid_admin_request *spraid_admin_req(struct request *req) -{ - return blk_mq_rq_to_pdu(req); -} - static int spraid_npages_prp(u32 size, struct spraid_dev *hdev) { u32 nprps = DIV_ROUND_UP(size + hdev->page_size, hdev->page_size); @@ -419,7 +424,7 @@ static void spraid_submit_cmd(struct spraid_queue *spraidq, const void *cmd) writel(spraidq->sq_tail, spraidq->q_db); spin_unlock_irqrestore(&spraidq->sq_lock, flags); - dev_log_dbg(spraidq->hdev->dev, "cid[%d], qid[%d], opcode[0x%x], flags[0x%x], hdid[%u]\n", + dev_log_dbg(spraidq->hdev->dev, "cid[%d] qid[%d], opcode[0x%x], flags[0x%x], hdid[%u]\n", acd->command_id, spraidq->qid, acd->opcode, acd->flags, le32_to_cpu(acd->hdid)); } @@ -605,18 +610,15 @@ static void spraid_setup_rw_cmd(struct spraid_dev *hdev, if (scmd->cmd_len == 6) { datalength = (u32)(scmd->cmnd[4] == 0 ? IO_6_DEFAULT_TX_LEN : scmd->cmnd[4]); - start_lba_lo = ((u32)scmd->cmnd[1] << 16) | - ((u32)scmd->cmnd[2] << 8) | (u32)scmd->cmnd[3]; + start_lba_lo = (u32)get_unaligned_be24(&scmd->cmnd[1]); start_lba_lo &= 0x1FFFFF; } /* 10-byte READ(0x28) or WRITE(0x2A) cdb */ else if (scmd->cmd_len == 10) { - datalength = (u32)scmd->cmnd[8] | ((u32)scmd->cmnd[7] << 8); - start_lba_lo = ((u32)scmd->cmnd[2] << 24) | - ((u32)scmd->cmnd[3] << 16) | - ((u32)scmd->cmnd[4] << 8) | (u32)scmd->cmnd[5]; + datalength = (u32)get_unaligned_be16(&scmd->cmnd[7]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[2]); if (scmd->cmnd[1] & FUA_MASK) control |= SPRAID_RW_FUA; @@ -624,42 +626,26 @@ static void spraid_setup_rw_cmd(struct spraid_dev *hdev, /* 12-byte READ(0xA8) or WRITE(0xAA) cdb */ else if (scmd->cmd_len == 12) { - datalength = ((u32)scmd->cmnd[6] << 24) | - ((u32)scmd->cmnd[7] << 16) | - ((u32)scmd->cmnd[8] << 8) | (u32)scmd->cmnd[9]; - start_lba_lo = ((u32)scmd->cmnd[2] << 24) | - ((u32)scmd->cmnd[3] << 16) | - ((u32)scmd->cmnd[4] << 8) | (u32)scmd->cmnd[5]; + datalength = get_unaligned_be32(&scmd->cmnd[6]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[2]); if (scmd->cmnd[1] & FUA_MASK) control |= SPRAID_RW_FUA; } /* 16-byte READ(0x88) or WRITE(0x8A) cdb */ else if (scmd->cmd_len == 16) { - datalength = ((u32)scmd->cmnd[10] << 24) | - ((u32)scmd->cmnd[11] << 16) | - ((u32)scmd->cmnd[12] << 8) | (u32)scmd->cmnd[13]; - start_lba_lo = ((u32)scmd->cmnd[6] << 24) | - ((u32)scmd->cmnd[7] << 16) | - ((u32)scmd->cmnd[8] << 8) | (u32)scmd->cmnd[9]; - start_lba_hi = ((u32)scmd->cmnd[2] << 24) | - ((u32)scmd->cmnd[3] << 16) | - ((u32)scmd->cmnd[4] << 8) | (u32)scmd->cmnd[5]; + datalength = get_unaligned_be32(&scmd->cmnd[10]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[6]); + start_lba_hi = get_unaligned_be32(&scmd->cmnd[2]); if (scmd->cmnd[1] & FUA_MASK) control |= SPRAID_RW_FUA; } /* 32-byte READ(0x88) or WRITE(0x8A) cdb */ else if (scmd->cmd_len == 32) { - datalength = ((u32)scmd->cmnd[28] << 24) | - ((u32)scmd->cmnd[29] << 16) | - ((u32)scmd->cmnd[30] << 8) | (u32)scmd->cmnd[31]; - start_lba_lo = ((u32)scmd->cmnd[16] << 24) | - ((u32)scmd->cmnd[17] << 16) | - ((u32)scmd->cmnd[18] << 8) | (u32)scmd->cmnd[19]; - start_lba_hi = ((u32)scmd->cmnd[12] << 24) | - ((u32)scmd->cmnd[13] << 16) | - ((u32)scmd->cmnd[14] << 8) | (u32)scmd->cmnd[15]; + datalength = get_unaligned_be32(&scmd->cmnd[28]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[16]); + start_lba_hi = get_unaligned_be32(&scmd->cmnd[12]); if (scmd->cmnd[10] & FUA_MASK) control |= SPRAID_RW_FUA; @@ -814,7 +800,7 @@ static void spraid_map_status(struct spraid_iod *iod, struct scsi_cmnd *scmd, if (scmd->result & SAM_STAT_CHECK_CONDITION) { memset(scmd->sense_buffer, 0, SCSI_SENSE_BUFFERSIZE); memcpy(scmd->sense_buffer, iod->sense, SCSI_SENSE_BUFFERSIZE); - set_driver_byte(scmd, DRIVER_SENSE); + scmd->result = (scmd->result & 0x00ffffff) | (DRIVER_SENSE << 24); } break; case FW_STAT_ABORTED: @@ -825,6 +811,8 @@ static void spraid_map_status(struct spraid_iod *iod, struct scsi_cmnd *scmd, break; default: set_host_byte(scmd, DID_BAD_TARGET); + dev_warn(iod->spraidq->hdev->dev, "[%s] cid[%d] qid[%d] bad status[0x%x]\n", + __func__, cqe->cmd_id, le16_to_cpu(cqe->sq_id), le16_to_cpu(cqe->status)); break; } } @@ -850,14 +838,13 @@ static int spraid_queue_command(struct Scsi_Host *shost, struct scsi_cmnd *scmd) int ret; if (unlikely(!scmd)) { - dev_err(hdev->dev, "err, scmd is null, return 0\n"); + dev_err(hdev->dev, "err, scmd is null\n"); return 0; } if (unlikely(hdev->state != SPRAID_LIVE)) { set_host_byte(scmd, DID_NO_CONNECT); scmd->scsi_done(scmd); - dev_err(hdev->dev, "[%s] err, hdev state is not live\n", __func__); return 0; } @@ -894,7 +881,7 @@ static int spraid_queue_command(struct Scsi_Host *shost, struct scsi_cmnd *scmd) WRITE_ONCE(iod->state, SPRAID_CMD_IN_FLIGHT); spraid_submit_cmd(ioq, &ioq_cmd); elapsed = jiffies - scmd->jiffies_at_alloc; - dev_log_dbg(hdev->dev, "cid[%d], qid[%d] submit IO cost %3ld.%3ld seconds\n", + dev_log_dbg(hdev->dev, "cid[%d] qid[%d] submit IO cost %3ld.%3ld seconds\n", cid, hwq, elapsed / HZ, elapsed % HZ); return 0; @@ -945,6 +932,10 @@ static int spraid_slave_alloc(struct scsi_device *sdev) scan_host: hostdata->hdid = le32_to_cpu(hdev->devices[idx].hdid); + hostdata->max_io_kb = le16_to_cpu(hdev->devices[idx].max_io_kb); + hostdata->attr = hdev->devices[idx].attr; + hostdata->flag = hdev->devices[idx].flag; + hostdata->rg_id = 0xff; sdev->hostdata = hostdata; up_read(&hdev->devices_rwsem); return 0; @@ -958,30 +949,17 @@ static void spraid_slave_destroy(struct scsi_device *sdev) static int spraid_slave_configure(struct scsi_device *sdev) { - u16 idx; - unsigned int timeout = scmd_tmout_nonpt * HZ; + unsigned int timeout = scmd_tmout_rawdisk * HZ; struct spraid_dev *hdev = shost_priv(sdev->host); struct spraid_sdev_hostdata *hostdata = sdev->hostdata; u32 max_sec = sdev->host->max_sectors; - if (!hostdata) { - idx = hostdata->hdid - 1; - if (sdev->channel == hdev->devices[idx].channel && - sdev->id == le16_to_cpu(hdev->devices[idx].target) && - sdev->lun < hdev->devices[idx].lun) { - if (SPRAID_DEV_INFO_ATTR_PT(hdev->devices[idx].attr)) - timeout = scmd_tmout_pt * HZ; - else - timeout = scmd_tmout_nonpt * HZ; - max_sec = le16_to_cpu(hdev->devices[idx].max_io_kb) << 1; - } else { - dev_err(hdev->dev, "[%s] err, sdev->channel:id:lun[%d:%d:%lld];" - "devices[%d], channel:target:lun[%d:%d:%d]\n", - __func__, sdev->channel, sdev->id, sdev->lun, - idx, hdev->devices[idx].channel, - hdev->devices[idx].target, - hdev->devices[idx].lun); - } + if (hostdata) { + if (SPRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + timeout = scmd_tmout_vd * HZ; + else if (SPRAID_DEV_INFO_ATTR_RAWDISK(hostdata->attr)) + timeout = scmd_tmout_rawdisk * HZ; + max_sec = hostdata->max_io_kb << 1; } else { dev_err(hdev->dev, "[%s] err, sdev->hostdata is null\n", __func__); } @@ -991,7 +969,9 @@ static int spraid_slave_configure(struct scsi_device *sdev) if ((max_sec == 0) || (max_sec > sdev->host->max_sectors)) max_sec = sdev->host->max_sectors; - blk_queue_max_hw_sectors(sdev->request_queue, max_sec); + + if (!max_io_force) + blk_queue_max_hw_sectors(sdev->request_queue, max_sec); dev_info(hdev->dev, "[%s] sdev->channel:id:lun[%d:%d:%lld], scmd_timeout[%d]s, maxsec[%d]\n", __func__, sdev->channel, sdev->id, sdev->lun, timeout / HZ, max_sec); @@ -1176,6 +1156,75 @@ static inline bool spraid_cqe_pending(struct spraid_queue *spraidq) spraidq->cq_phase; } +static void spraid_sata_report_zone_handle(struct scsi_cmnd *scmd, struct spraid_iod *iod) +{ + int i = 0; + unsigned int bytes = 0; + struct scatterlist *sg = scsi_sglist(scmd); + + scsi_for_each_sg(scmd, sg, iod->nsge, i) { + unsigned int offset = 0; + + if (bytes == 0) { + char *hdr; + u32 list_length; + u64 max_lba, opt_lba; + u16 same; + + hdr = sg_virt(sg); + + list_length = get_unaligned_le32(&hdr[0]); + same = get_unaligned_le16(&hdr[4]); + max_lba = get_unaligned_le64(&hdr[8]); + opt_lba = get_unaligned_le64(&hdr[16]); + put_unaligned_be32(list_length, &hdr[0]); + hdr[4] = same & 0xf; + put_unaligned_be64(max_lba, &hdr[8]); + put_unaligned_be64(opt_lba, &hdr[16]); + offset += 64; + bytes += 64; + } + while (offset < sg_dma_len(sg)) { + char *rec; + u8 cond, type, non_seq, reset; + u64 size, start, wp; + + rec = sg_virt(sg) + offset; + type = rec[0] & 0xf; + cond = (rec[1] >> 4) & 0xf; + non_seq = (rec[1] & 2); + reset = (rec[1] & 1); + size = get_unaligned_le64(&rec[8]); + start = get_unaligned_le64(&rec[16]); + wp = get_unaligned_le64(&rec[24]); + rec[0] = type; + rec[1] = (cond << 4) | non_seq | reset; + put_unaligned_be64(size, &rec[8]); + put_unaligned_be64(start, &rec[16]); + put_unaligned_be64(wp, &rec[24]); + WARN_ON(offset + 64 > sg_dma_len(sg)); + offset += 64; + bytes += 64; + } + } +} + +static inline void spraid_handle_ata_cmd(struct spraid_dev *hdev, struct scsi_cmnd *scmd, + struct spraid_iod *iod) +{ + if (hdev->ctrl_info->card_type != SPRAID_CARD_HBA) + return; + + switch (scmd->cmnd[0]) { + case ZBC_IN: + dev_info(hdev->dev, "[%s] process report zone\n", __func__); + spraid_sata_report_zone_handle(scmd, iod); + break; + default: + break; + } +} + static void spraid_complete_ioq_cmnd(struct spraid_queue *ioq, struct spraid_completion *cqe) { struct spraid_dev *hdev = ioq->hdev; @@ -1197,12 +1246,12 @@ static void spraid_complete_ioq_cmnd(struct spraid_queue *ioq, struct spraid_com iod = scsi_cmd_priv(scmd); elapsed = jiffies - scmd->jiffies_at_alloc; - dev_log_dbg(hdev->dev, "cid[%d], qid[%d] finish IO cost %3ld.%3ld seconds\n", + dev_log_dbg(hdev->dev, "cid[%d] qid[%d] finish IO cost %3ld.%3ld seconds\n", cqe->cmd_id, ioq->qid, elapsed / HZ, elapsed % HZ); if (cmpxchg(&iod->state, SPRAID_CMD_IN_FLIGHT, SPRAID_CMD_COMPLETE) != SPRAID_CMD_IN_FLIGHT) { - dev_warn(hdev->dev, "cid[%d], qid[%d] enters abnormal handler, cost %3ld.%3ld seconds\n", + dev_warn(hdev->dev, "cid[%d] qid[%d] enters abnormal handler, cost %3ld.%3ld seconds\n", cqe->cmd_id, ioq->qid, elapsed / HZ, elapsed % HZ); WRITE_ONCE(iod->state, SPRAID_CMD_TMO_COMPLETE); @@ -1215,6 +1264,8 @@ static void spraid_complete_ioq_cmnd(struct spraid_queue *ioq, struct spraid_com return; } + spraid_handle_ata_cmd(hdev, scmd, iod); + spraid_map_status(iod, scmd, cqe); if (iod->nsge) { iod->nsge = 0; @@ -1224,38 +1275,36 @@ static void spraid_complete_ioq_cmnd(struct spraid_queue *ioq, struct spraid_com scmd->scsi_done(scmd); } -static inline void spraid_end_admin_request(struct request *req, __le16 status, - __le32 result0, __le32 result1) -{ - struct spraid_admin_request *rq = spraid_admin_req(req); - - rq->status = le16_to_cpu(status) >> 1; - rq->result0 = le32_to_cpu(result0); - rq->result1 = le32_to_cpu(result1); - blk_mq_complete_request(req); -} - static void spraid_complete_adminq_cmnd(struct spraid_queue *adminq, struct spraid_completion *cqe) { - struct blk_mq_tags *tags = adminq->hdev->admin_tagset.tags[0]; - struct request *req; + struct spraid_dev *hdev = adminq->hdev; + struct spraid_cmd *adm_cmd; - req = blk_mq_tag_to_rq(tags, cqe->cmd_id); - if (unlikely(!req)) { + adm_cmd = hdev->adm_cmds + cqe->cmd_id; + if (unlikely(adm_cmd->state == SPRAID_CMD_IDLE)) { dev_warn(adminq->hdev->dev, "Invalid id %d completed on queue %d\n", cqe->cmd_id, le16_to_cpu(cqe->sq_id)); return; } - spraid_end_admin_request(req, cqe->status, cqe->result, cqe->result1); + + adm_cmd->status = le16_to_cpu(cqe->status) >> 1; + adm_cmd->result0 = le32_to_cpu(cqe->result); + adm_cmd->result1 = le32_to_cpu(cqe->result1); + + complete(&adm_cmd->cmd_done); } +static void spraid_send_aen(struct spraid_dev *hdev, u16 cid); + static void spraid_complete_aen(struct spraid_queue *spraidq, struct spraid_completion *cqe) { struct spraid_dev *hdev = spraidq->hdev; u32 result = le32_to_cpu(cqe->result); - dev_info(hdev->dev, "rcv aen, status[%x], result[%x]\n", - le16_to_cpu(cqe->status) >> 1, result); + dev_info(hdev->dev, "rcv aen, cid[%d], status[0x%x], result[0x%x]\n", + cqe->cmd_id, le16_to_cpu(cqe->status) >> 1, result); + + spraid_send_aen(hdev, cqe->cmd_id); if ((le16_to_cpu(cqe->status) >> 1) != SPRAID_SC_SUCCESS) return; @@ -1264,22 +1313,19 @@ static void spraid_complete_aen(struct spraid_queue *spraidq, struct spraid_comp spraid_handle_aen_notice(hdev, result); break; case SPRAID_AEN_VS: - spraid_handle_aen_vs(hdev, result); + spraid_handle_aen_vs(hdev, result, le32_to_cpu(cqe->result1)); break; default: dev_warn(hdev->dev, "Unsupported async event type: %u\n", result & 0x7); break; } - queue_work(spraid_wq, &hdev->aen_work); } -static void spraid_put_ioq_ptcmd(struct spraid_dev *hdev, struct spraid_ioq_ptcmd *cmd); - static void spraid_complete_ioq_sync_cmnd(struct spraid_queue *ioq, struct spraid_completion *cqe) { struct spraid_dev *hdev = ioq->hdev; - struct spraid_ioq_ptcmd *ptcmd; + struct spraid_cmd *ptcmd; ptcmd = hdev->ioq_ptcmds + (ioq->qid - 1) * SPRAID_PTCMDS_PERQ + cqe->cmd_id - SPRAID_IO_BLK_MQ_DEPTH; @@ -1289,8 +1335,6 @@ static void spraid_complete_ioq_sync_cmnd(struct spraid_queue *ioq, struct sprai ptcmd->result1 = le32_to_cpu(cqe->result1); complete(&ptcmd->cmd_done); - - spraid_put_ioq_ptcmd(hdev, ptcmd); } static inline void spraid_handle_cqe(struct spraid_queue *spraidq, u16 idx) @@ -1304,7 +1348,7 @@ static inline void spraid_handle_cqe(struct spraid_queue *spraidq, u16 idx) return; } - dev_log_dbg(hdev->dev, "cid[%d], qid[%d], result[0x%x], sq_id[%d], status[0x%x]\n", + dev_log_dbg(hdev->dev, "cid[%d] qid[%d], result[0x%x], sq_id[%d], status[0x%x]\n", cqe->cmd_id, spraidq->qid, le32_to_cpu(cqe->result), le16_to_cpu(cqe->sq_id), le16_to_cpu(cqe->status)); @@ -1452,62 +1496,119 @@ static u32 spraid_bar_size(struct spraid_dev *hdev, u32 nr_ioqs) return (SPRAID_REG_DBS + ((nr_ioqs + 1) * 8 * hdev->db_stride)); } -static inline void spraid_clear_spraid_request(struct request *req) +static int spraid_alloc_admin_cmds(struct spraid_dev *hdev) { - if (!(req->rq_flags & RQF_DONTPREP)) { - spraid_admin_req(req)->flags = 0; - req->rq_flags |= RQF_DONTPREP; + int i; + + INIT_LIST_HEAD(&hdev->adm_cmd_list); + spin_lock_init(&hdev->adm_cmd_lock); + + hdev->adm_cmds = kcalloc_node(SPRAID_AQ_BLK_MQ_DEPTH, sizeof(struct spraid_cmd), + GFP_KERNEL, hdev->numa_node); + + if (!hdev->adm_cmds) { + dev_err(hdev->dev, "Alloc admin cmds failed\n"); + return -ENOMEM; + } + + for (i = 0; i < SPRAID_AQ_BLK_MQ_DEPTH; i++) { + hdev->adm_cmds[i].qid = 0; + hdev->adm_cmds[i].cid = i; + list_add_tail(&(hdev->adm_cmds[i].list), &hdev->adm_cmd_list); } + + dev_info(hdev->dev, "Alloc admin cmds success, num[%d]\n", SPRAID_AQ_BLK_MQ_DEPTH); + + return 0; } -static struct request *spraid_alloc_admin_request(struct request_queue *q, - struct spraid_admin_command *cmd, - blk_mq_req_flags_t flags) +static void spraid_free_admin_cmds(struct spraid_dev *hdev) { - u32 op = COMMAND_IS_WRITE(cmd) ? REQ_OP_DRV_OUT : REQ_OP_DRV_IN; - struct request *req; + kfree(hdev->adm_cmds); + hdev->adm_cmds = NULL; + INIT_LIST_HEAD(&hdev->adm_cmd_list); +} + +static struct spraid_cmd *spraid_get_cmd(struct spraid_dev *hdev, enum spraid_cmd_type type) +{ + struct spraid_cmd *cmd = NULL; + unsigned long flags; + struct list_head *head = &hdev->adm_cmd_list; + spinlock_t *slock = &hdev->adm_cmd_lock; + + if (type == SPRAID_CMD_IOPT) { + head = &hdev->ioq_pt_list; + slock = &hdev->ioq_pt_lock; + } + + spin_lock_irqsave(slock, flags); + if (list_empty(head)) { + spin_unlock_irqrestore(slock, flags); + dev_err(hdev->dev, "err, cmd[%d] list empty\n", type); + return NULL; + } + cmd = list_entry(head->next, struct spraid_cmd, list); + list_del_init(&cmd->list); + spin_unlock_irqrestore(slock, flags); - req = blk_mq_alloc_request(q, op, flags); - if (IS_ERR(req)) - return req; - req->cmd_flags |= REQ_FAILFAST_DRIVER; - spraid_clear_spraid_request(req); - spraid_admin_req(req)->cmd = cmd; + WRITE_ONCE(cmd->state, SPRAID_CMD_IN_FLIGHT); - return req; + return cmd; } -static int spraid_submit_admin_sync_cmd(struct request_queue *q, - struct spraid_admin_command *cmd, - u32 *result, void *buffer, - u32 bufflen, u32 timeout, int at_head, blk_mq_req_flags_t flags) +static void spraid_put_cmd(struct spraid_dev *hdev, struct spraid_cmd *cmd, + enum spraid_cmd_type type) { - struct request *req; - int ret; + unsigned long flags; + struct list_head *head = &hdev->adm_cmd_list; + spinlock_t *slock = &hdev->adm_cmd_lock; - req = spraid_alloc_admin_request(q, cmd, flags); - if (IS_ERR(req)) - return PTR_ERR(req); + if (type == SPRAID_CMD_IOPT) { + head = &hdev->ioq_pt_list; + slock = &hdev->ioq_pt_lock; + } - req->timeout = timeout ? timeout : ADMIN_TIMEOUT; - if (buffer && bufflen) { - ret = blk_rq_map_kern(q, req, buffer, bufflen, GFP_KERNEL); - if (ret) - goto out; + spin_lock_irqsave(slock, flags); + WRITE_ONCE(cmd->state, SPRAID_CMD_IDLE); + list_add_tail(&cmd->list, head); + spin_unlock_irqrestore(slock, flags); +} + + +static int spraid_submit_admin_sync_cmd(struct spraid_dev *hdev, struct spraid_admin_command *cmd, + u32 *result0, u32 *result1, u32 timeout) +{ + struct spraid_cmd *adm_cmd = spraid_get_cmd(hdev, SPRAID_CMD_ADM); + + if (!adm_cmd) { + dev_err(hdev->dev, "err, get admin cmd failed\n"); + return -EFAULT; } - blk_execute_rq(req->q, NULL, req, at_head); - if (result) - *result = spraid_admin_req(req)->result0; + timeout = timeout ? timeout : ADMIN_TIMEOUT; - if (spraid_admin_req(req)->flags & SPRAID_REQ_CANCELLED) - ret = -EINTR; - else - ret = spraid_admin_req(req)->status; + init_completion(&adm_cmd->cmd_done); -out: - blk_mq_free_request(req); - return ret; + cmd->common.command_id = adm_cmd->cid; + spraid_submit_cmd(&hdev->queues[0], cmd); + + if (!wait_for_completion_timeout(&adm_cmd->cmd_done, timeout)) { + dev_err(hdev->dev, "[%s] cid[%d] qid[%d] timeout, opcode[0x%x] subopcode[0x%x]\n", + __func__, adm_cmd->cid, adm_cmd->qid, cmd->usr_cmd.opcode, + cmd->usr_cmd.info_0.subopcode); + WRITE_ONCE(adm_cmd->state, SPRAID_CMD_TIMEOUT); + spraid_put_cmd(hdev, adm_cmd, SPRAID_CMD_ADM); + return -ETIME; + } + + if (result0) + *result0 = adm_cmd->result0; + if (result1) + *result1 = adm_cmd->result1; + + spraid_put_cmd(hdev, adm_cmd, SPRAID_CMD_ADM); + + return adm_cmd->status; } static int spraid_create_cq(struct spraid_dev *hdev, u16 qid, @@ -1524,8 +1625,7 @@ static int spraid_create_cq(struct spraid_dev *hdev, u16 qid, admin_cmd.create_cq.cq_flags = cpu_to_le16(flags); admin_cmd.create_cq.irq_vector = cpu_to_le16(cq_vector); - return spraid_submit_admin_sync_cmd(hdev->admin_q, &admin_cmd, NULL, - NULL, 0, 0, 0, 0); + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); } static int spraid_create_sq(struct spraid_dev *hdev, u16 qid, @@ -1542,8 +1642,7 @@ static int spraid_create_sq(struct spraid_dev *hdev, u16 qid, admin_cmd.create_sq.sq_flags = cpu_to_le16(flags); admin_cmd.create_sq.cqid = cpu_to_le16(qid); - return spraid_submit_admin_sync_cmd(hdev->admin_q, &admin_cmd, NULL, - NULL, 0, 0, 0, 0); + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); } static void spraid_free_queue(struct spraid_queue *spraidq) @@ -1581,8 +1680,7 @@ static int spraid_delete_queue(struct spraid_dev *hdev, u8 op, u16 id) admin_cmd.delete_queue.opcode = op; admin_cmd.delete_queue.qid = cpu_to_le16(id); - ret = spraid_submit_admin_sync_cmd(hdev->admin_q, &admin_cmd, NULL, - NULL, 0, 0, 0, 0); + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); if (ret) dev_err(hdev->dev, "Delete %s:[%d] failed\n", @@ -1663,19 +1761,28 @@ static int spraid_set_features(struct spraid_dev *hdev, u32 fid, u32 dword11, vo size_t buflen, u32 *result) { struct spraid_admin_command admin_cmd; - u32 res; int ret; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + + if (buffer && buflen) { + data_ptr = dma_alloc_coherent(hdev->dev, buflen, &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memcpy(data_ptr, buffer, buflen); + } memset(&admin_cmd, 0, sizeof(admin_cmd)); admin_cmd.features.opcode = SPRAID_ADMIN_SET_FEATURES; admin_cmd.features.fid = cpu_to_le32(fid); admin_cmd.features.dword11 = cpu_to_le32(dword11); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); - ret = spraid_submit_admin_sync_cmd(hdev->admin_q, &admin_cmd, &res, - buffer, buflen, 0, 0, 0); + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, result, NULL, 0); - if (!ret && result) - *result = res; + if (data_ptr) + dma_free_coherent(hdev->dev, buflen, data_ptr, data_dma); return ret; } @@ -1764,8 +1871,7 @@ static int spraid_setup_io_queues(struct spraid_dev *hdev) break; } dev_info(hdev->dev, "[%s] max_qid: %d, queue_count: %d, online_queue: %d, ioq_depth: %d\n", - __func__, hdev->max_qid, hdev->queue_count, - hdev->online_queues, hdev->ioq_depth); + __func__, hdev->max_qid, hdev->queue_count, hdev->online_queues, hdev->ioq_depth); return spraid_create_io_queues(hdev); } @@ -1889,10 +1995,11 @@ static int spraid_get_dev_list(struct spraid_dev *hdev, struct spraid_dev_info * u32 nd = le32_to_cpu(hdev->ctrl_info->nd); struct spraid_admin_command admin_cmd; struct spraid_dev_list *list_buf; + dma_addr_t data_dma = 0; u32 i, idx, hdid, ndev; int ret = 0; - list_buf = kmalloc(sizeof(*list_buf), GFP_KERNEL); + list_buf = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &data_dma, GFP_KERNEL); if (!list_buf) return -ENOMEM; @@ -1901,9 +2008,9 @@ static int spraid_get_dev_list(struct spraid_dev *hdev, struct spraid_dev_info * admin_cmd.get_info.opcode = SPRAID_ADMIN_GET_INFO; admin_cmd.get_info.type = SPRAID_GET_INFO_DEV_LIST; admin_cmd.get_info.cdw11 = cpu_to_le32(idx); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); - ret = spraid_submit_admin_sync_cmd(hdev->admin_q, &admin_cmd, NULL, list_buf, - sizeof(*list_buf), 0, 0, 0); + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); if (ret) { dev_err(hdev->dev, "Get device list failed, nd: %u, idx: %u, ret: %d\n", @@ -1916,12 +2023,11 @@ static int spraid_get_dev_list(struct spraid_dev *hdev, struct spraid_dev_info * for (i = 0; i < ndev; i++) { hdid = le32_to_cpu(list_buf->devices[i].hdid); - dev_info(hdev->dev, "list_buf->devices[%d], hdid: %u target: %d, channel: %d, lun: %d, attr[%x]\n", - i, hdid, - le16_to_cpu(list_buf->devices[i].target), - list_buf->devices[i].channel, - list_buf->devices[i].lun, - list_buf->devices[i].attr); + dev_info(hdev->dev, "list_buf->devices[%d], hdid: %u target: %d, channel: %d, lun: %d, attr[0x%x]\n", + i, hdid, le16_to_cpu(list_buf->devices[i].target), + list_buf->devices[i].channel, + list_buf->devices[i].lun, + list_buf->devices[i].attr); if (hdid > nd || hdid == 0) { dev_err(hdev->dev, "err, hdid[%d] invalid\n", hdid); continue; @@ -1936,21 +2042,29 @@ static int spraid_get_dev_list(struct spraid_dev *hdev, struct spraid_dev_info * } out: - kfree(list_buf); + dma_free_coherent(hdev->dev, PAGE_SIZE, list_buf, data_dma); return ret; } -static void spraid_send_aen(struct spraid_dev *hdev) +static void spraid_send_aen(struct spraid_dev *hdev, u16 cid) { struct spraid_queue *adminq = &hdev->queues[0]; struct spraid_admin_command admin_cmd; memset(&admin_cmd, 0, sizeof(admin_cmd)); admin_cmd.common.opcode = SPRAID_ADMIN_ASYNC_EVENT; - admin_cmd.common.command_id = SPRAID_AQ_BLK_MQ_DEPTH; + admin_cmd.common.command_id = cid; spraid_submit_cmd(adminq, &admin_cmd); - dev_info(hdev->dev, "send aen, cid[%d]\n", SPRAID_AQ_BLK_MQ_DEPTH); + dev_info(hdev->dev, "send aen, cid[%d]\n", cid); +} + +static inline void spraid_send_all_aen(struct spraid_dev *hdev) +{ + u16 i; + + for (i = 0; i < hdev->ctrl_info->aerl; i++) + spraid_send_aen(hdev, i + SPRAID_AQ_BLK_MQ_DEPTH); } static int spraid_add_device(struct spraid_dev *hdev, struct spraid_dev_info *device) @@ -1958,6 +2072,10 @@ static int spraid_add_device(struct spraid_dev *hdev, struct spraid_dev_info *de struct Scsi_Host *shost = hdev->shost; struct scsi_device *sdev; + dev_info(hdev->dev, "add device, hdid: %u target: %d, channel: %d, lun: %d, attr[0x%x]\n", + le32_to_cpu(device->hdid), le16_to_cpu(device->target), + device->channel, device->lun, device->attr); + sdev = scsi_device_lookup(shost, device->channel, le16_to_cpu(device->target), 0); if (sdev) { dev_warn(hdev->dev, "Device is already exist, channel: %d, target_id: %d, lun: %d\n", @@ -1974,9 +2092,13 @@ static int spraid_rescan_device(struct spraid_dev *hdev, struct spraid_dev_info struct Scsi_Host *shost = hdev->shost; struct scsi_device *sdev; + dev_info(hdev->dev, "rescan device, hdid: %u target: %d, channel: %d, lun: %d, attr[0x%x]\n", + le32_to_cpu(device->hdid), le16_to_cpu(device->target), + device->channel, device->lun, device->attr); + sdev = scsi_device_lookup(shost, device->channel, le16_to_cpu(device->target), 0); if (!sdev) { - dev_warn(hdev->dev, "Device is not exit, channel: %d, target_id: %d, lun: %d\n", + dev_warn(hdev->dev, "device is not exit rescan it, channel: %d, target_id: %d, lun: %d\n", device->channel, le16_to_cpu(device->target), 0); return -ENODEV; } @@ -1991,9 +2113,13 @@ static int spraid_remove_device(struct spraid_dev *hdev, struct spraid_dev_info struct Scsi_Host *shost = hdev->shost; struct scsi_device *sdev; + dev_info(hdev->dev, "remove device, hdid: %u target: %d, channel: %d, lun: %d, attr[0x%x]\n", + le32_to_cpu(org_device->hdid), le16_to_cpu(org_device->target), + org_device->channel, org_device->lun, org_device->attr); + sdev = scsi_device_lookup(shost, org_device->channel, le16_to_cpu(org_device->target), 0); if (!sdev) { - dev_warn(hdev->dev, "Device is not exit, channel: %d, target_id: %d, lun: %d\n", + dev_warn(hdev->dev, "device is not exit remove it, channel: %d, target_id: %d, lun: %d\n", org_device->channel, le16_to_cpu(org_device->target), 0); return -ENODEV; } @@ -2029,36 +2155,54 @@ static int spraid_dev_list_init(struct spraid_dev *hdev) return 0; } +static int luntarget_cmp_func(const void *l, const void *r) +{ + const struct spraid_dev_info *ln = l; + const struct spraid_dev_info *rn = r; + + if (ln->channel == rn->channel) + return le16_to_cpu(ln->target) - le16_to_cpu(rn->target); + + return ln->channel - rn->channel; +} + static void spraid_scan_work(struct work_struct *work) { struct spraid_dev *hdev = container_of(work, struct spraid_dev, scan_work); struct spraid_dev_info *devices, *org_devices; + struct spraid_dev_info *sortdevice; u32 nd = le32_to_cpu(hdev->ctrl_info->nd); u8 flag, org_flag; int i, ret; + int count = 0; devices = kcalloc(nd, sizeof(struct spraid_dev_info), GFP_KERNEL); if (!devices) return; + + sortdevice = kcalloc(nd, sizeof(struct spraid_dev_info), GFP_KERNEL); + if (!sortdevice) + goto free_list; + ret = spraid_get_dev_list(hdev, devices); if (ret) - goto free_list; + goto free_all; org_devices = hdev->devices; for (i = 0; i < nd; i++) { org_flag = org_devices[i].flag; flag = devices[i].flag; - dev_log_dbg(hdev->dev, "i: %d, org_flag: 0x%x, flag: 0x%x\n", - i, org_flag, flag); + dev_log_dbg(hdev->dev, "i: %d, org_flag: 0x%x, flag: 0x%x\n", i, org_flag, flag); if (SPRAID_DEV_INFO_FLAG_VALID(flag)) { if (!SPRAID_DEV_INFO_FLAG_VALID(org_flag)) { down_write(&hdev->devices_rwsem); memcpy(&org_devices[i], &devices[i], - sizeof(struct spraid_dev_info)); + sizeof(struct spraid_dev_info)); + memcpy(&sortdevice[count++], &devices[i], + sizeof(struct spraid_dev_info)); up_write(&hdev->devices_rwsem); - spraid_add_device(hdev, &devices[i]); } else if (SPRAID_DEV_INFO_FLAG_CHANGE(flag)) { spraid_rescan_device(hdev, &devices[i]); } @@ -2071,6 +2215,16 @@ static void spraid_scan_work(struct work_struct *work) } } } + + dev_info(hdev->dev, "scan work add device count = %d\n", count); + + sort(sortdevice, count, sizeof(sortdevice[0]), luntarget_cmp_func, NULL); + + for (i = 0; i < count; i++) + spraid_add_device(hdev, &sortdevice[i]); + +free_all: + kfree(sortdevice); free_list: kfree(devices); } @@ -2083,6 +2237,15 @@ static void spraid_timesyn_work(struct work_struct *work) spraid_configure_timestamp(hdev); } +static int spraid_init_ctrl_info(struct spraid_dev *hdev); +static void spraid_fw_act_work(struct work_struct *work) +{ + struct spraid_dev *hdev = container_of(work, struct spraid_dev, fw_act_work); + + if (spraid_init_ctrl_info(hdev)) + dev_err(hdev->dev, "get ctrl info failed after fw act\n"); +} + static void spraid_queue_scan(struct spraid_dev *hdev) { queue_work(spraid_wq, &hdev->scan_work); @@ -2094,6 +2257,9 @@ static void spraid_handle_aen_notice(struct spraid_dev *hdev, u32 result) case SPRAID_AEN_DEV_CHANGED: spraid_queue_scan(hdev); break; + case SPRAID_AEN_FW_ACT_START: + dev_info(hdev->dev, "fw activation starting\n"); + break; case SPRAID_AEN_HOST_PROBING: break; default: @@ -2101,25 +2267,25 @@ static void spraid_handle_aen_notice(struct spraid_dev *hdev, u32 result) } } -static void spraid_handle_aen_vs(struct spraid_dev *hdev, u32 result) +static void spraid_handle_aen_vs(struct spraid_dev *hdev, u32 result, u32 result1) { - switch (result) { + switch ((result & 0xff00) >> 8) { case SPRAID_AEN_TIMESYN: queue_work(spraid_wq, &hdev->timesyn_work); break; + case SPRAID_AEN_FW_ACT_FINISH: + dev_info(hdev->dev, "fw activation finish\n"); + queue_work(spraid_wq, &hdev->fw_act_work); + break; + case SPRAID_AEN_EVENT_MIN ... SPRAID_AEN_EVENT_MAX: + dev_info(hdev->dev, "rcv card event[%d], param1[0x%x] param2[0x%x]\n", + (result & 0xff00) >> 8, result, result1); + break; default: - dev_warn(hdev->dev, "async event result: %x\n", result); + dev_warn(hdev->dev, "async event result: 0x%x\n", result); } } -static void spraid_async_event_work(struct work_struct *work) -{ - struct spraid_dev *hdev = - container_of(work, struct spraid_dev, aen_work); - - spraid_send_aen(hdev); -} - static int spraid_alloc_resources(struct spraid_dev *hdev) { int ret, nqueue; @@ -2149,10 +2315,16 @@ static int spraid_alloc_resources(struct spraid_dev *hdev) goto destroy_dma_pools; } + ret = spraid_alloc_admin_cmds(hdev); + if (ret) + goto free_queues; + dev_info(hdev->dev, "[%s] queues num: %d\n", __func__, nqueue); return 0; +free_queues: + kfree(hdev->queues); destroy_dma_pools: spraid_destroy_dma_pools(hdev); free_ctrl_info: @@ -2164,50 +2336,18 @@ static int spraid_alloc_resources(struct spraid_dev *hdev) static void spraid_free_resources(struct spraid_dev *hdev) { + spraid_free_admin_cmds(hdev); kfree(hdev->queues); spraid_destroy_dma_pools(hdev); kfree(hdev->ctrl_info); ida_free(&spraid_instance_ida, hdev->instance); } -static void spraid_setup_passthrough(struct request *req, struct spraid_admin_command *cmd) -{ - memcpy(cmd, spraid_admin_req(req)->cmd, sizeof(*cmd)); - cmd->common.flags &= ~SPRAID_CMD_FLAG_SGL_ALL; -} - -static inline void spraid_clear_hreq(struct request *req) -{ - if (!(req->rq_flags & RQF_DONTPREP)) { - spraid_admin_req(req)->flags = 0; - req->rq_flags |= RQF_DONTPREP; - } -} - -static blk_status_t spraid_setup_admin_cmd(struct request *req, struct spraid_admin_command *cmd) +static void spraid_bsg_unmap_data(struct spraid_dev *hdev, struct bsg_job *job) { - spraid_clear_hreq(req); - - memset(cmd, 0, sizeof(*cmd)); - switch (req_op(req)) { - case REQ_OP_DRV_IN: - case REQ_OP_DRV_OUT: - spraid_setup_passthrough(req, cmd); - break; - default: - WARN_ON_ONCE(1); - return BLK_STS_IOERR; - } - - cmd->common.command_id = req->tag; - return BLK_STS_OK; -} - -static void spraid_unmap_data(struct spraid_dev *hdev, struct request *req) -{ - struct spraid_iod *iod = blk_mq_rq_to_pdu(req); - enum dma_data_direction dma_dir = rq_data_dir(req) ? - DMA_TO_DEVICE : DMA_FROM_DEVICE; + struct request *rq = blk_mq_rq_from_pdu(job); + struct spraid_iod *iod = job->dd_data; + enum dma_data_direction dma_dir = rq_data_dir(rq) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; if (iod->nsge) dma_unmap_sg(hdev->dev, iod->sg, iod->nsge, dma_dir); @@ -2215,36 +2355,36 @@ static void spraid_unmap_data(struct spraid_dev *hdev, struct request *req) spraid_free_iod_res(hdev, iod); } -static blk_status_t spraid_admin_map_data(struct spraid_dev *hdev, struct request *req, - struct spraid_admin_command *cmd) +static int spraid_bsg_map_data(struct spraid_dev *hdev, struct bsg_job *job, + struct spraid_admin_command *cmd) { - struct spraid_iod *iod = blk_mq_rq_to_pdu(req); - struct request_queue *admin_q = req->q; - enum dma_data_direction dma_dir = rq_data_dir(req) ? - DMA_TO_DEVICE : DMA_FROM_DEVICE; - blk_status_t ret = BLK_STS_IOERR; - int nr_mapped; - int res; + struct request *rq = blk_mq_rq_from_pdu(job); + struct spraid_iod *iod = job->dd_data; + enum dma_data_direction dma_dir = rq_data_dir(rq) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; + int ret = 0; + + iod->sg = job->request_payload.sg_list; + iod->nsge = job->request_payload.sg_cnt; + iod->length = job->request_payload.payload_len; + iod->use_sgl = false; + iod->npages = -1; + iod->sg_drv_mgmt = false; - sg_init_table(iod->sg, blk_rq_nr_phys_segments(req)); - iod->nsge = blk_rq_map_sg(admin_q, req, iod->sg); if (!iod->nsge) goto out; - dev_info(hdev->dev, "nseg: %u, nsge: %u\n", - blk_rq_nr_phys_segments(req), iod->nsge); - - ret = BLK_STS_RESOURCE; - nr_mapped = dma_map_sg_attrs(hdev->dev, iod->sg, iod->nsge, dma_dir, DMA_ATTR_NO_WARN); - if (!nr_mapped) + ret = dma_map_sg_attrs(hdev->dev, iod->sg, iod->nsge, dma_dir, DMA_ATTR_NO_WARN); + if (!ret) goto out; - res = spraid_setup_prps(hdev, iod); - if (res) + ret = spraid_setup_prps(hdev, iod); + if (ret) goto unmap; + cmd->common.dptr.prp1 = cpu_to_le64(sg_dma_address(iod->sg)); cmd->common.dptr.prp2 = cpu_to_le64(iod->first_dma); - return BLK_STS_OK; + + return 0; unmap: dma_unmap_sg(hdev->dev, iod->sg, iod->nsge, dma_dir); @@ -2252,137 +2392,29 @@ static blk_status_t spraid_admin_map_data(struct spraid_dev *hdev, struct reques return ret; } -static blk_status_t spraid_init_admin_iod(struct request *rq, struct spraid_dev *hdev) -{ - struct spraid_iod *iod = blk_mq_rq_to_pdu(rq); - int nents = blk_rq_nr_phys_segments(rq); - unsigned int size = blk_rq_payload_bytes(rq); - - if (nents > SPRAID_INT_PAGES || size > SPRAID_INT_BYTES(hdev)) { - iod->sg = mempool_alloc(hdev->iod_mempool, GFP_ATOMIC); - if (!iod->sg) - return BLK_STS_RESOURCE; - } else { - iod->sg = iod->inline_sg; - } - - iod->nsge = 0; - iod->use_sgl = false; - iod->npages = -1; - iod->length = size; - iod->sg_drv_mgmt = true; - - return BLK_STS_OK; -} - -static blk_status_t spraid_queue_admin_rq(struct blk_mq_hw_ctx *hctx, - const struct blk_mq_queue_data *bd) -{ - struct spraid_queue *adminq = hctx->driver_data; - struct spraid_dev *hdev = adminq->hdev; - struct request *req = bd->rq; - struct spraid_iod *iod = blk_mq_rq_to_pdu(req); - struct spraid_admin_command cmd; - blk_status_t ret; - - ret = spraid_setup_admin_cmd(req, &cmd); - if (ret) - goto out; - - ret = spraid_init_admin_iod(req, hdev); - if (ret) - goto out; - - if (blk_rq_nr_phys_segments(req)) { - ret = spraid_admin_map_data(hdev, req, &cmd); - if (ret) - goto cleanup_iod; - } - - blk_mq_start_request(req); - spraid_submit_cmd(adminq, &cmd); - return BLK_STS_OK; - -cleanup_iod: - spraid_free_iod_res(hdev, iod); -out: - return ret; -} - -static blk_status_t spraid_error_status(struct request *req) -{ - switch (spraid_admin_req(req)->status & 0x7ff) { - case SPRAID_SC_SUCCESS: - return BLK_STS_OK; - default: - return BLK_STS_IOERR; - } -} - -static void spraid_complete_admin_rq(struct request *req) -{ - struct spraid_iod *iod = blk_mq_rq_to_pdu(req); - struct spraid_dev *hdev = iod->spraidq->hdev; - - if (blk_rq_nr_phys_segments(req)) - spraid_unmap_data(hdev, req); - blk_mq_end_request(req, spraid_error_status(req)); -} - -static int spraid_admin_init_hctx(struct blk_mq_hw_ctx *hctx, void *data, unsigned int hctx_idx) -{ - struct spraid_dev *hdev = data; - struct spraid_queue *adminq = &hdev->queues[0]; - - WARN_ON(hctx_idx != 0); - WARN_ON(hdev->admin_tagset.tags[0] != hctx->tags); - - hctx->driver_data = adminq; - return 0; -} - -static int spraid_admin_init_request(struct blk_mq_tag_set *set, struct request *req, - unsigned int hctx_idx, unsigned int numa_node) -{ - struct spraid_dev *hdev = set->driver_data; - struct spraid_iod *iod = blk_mq_rq_to_pdu(req); - struct spraid_queue *adminq = &hdev->queues[0]; - - WARN_ON(!adminq); - iod->spraidq = adminq; - return 0; -} - -static enum blk_eh_timer_return -spraid_admin_timeout(struct request *req, bool reserved) -{ - struct spraid_iod *iod = blk_mq_rq_to_pdu(req); - struct spraid_queue *spraidq = iod->spraidq; - struct spraid_dev *hdev = spraidq->hdev; - - dev_err(hdev->dev, "Admin cid[%d] qid[%d] timeout\n", - req->tag, spraidq->qid); - - if (spraid_poll_cq(spraidq, req->tag)) { - dev_warn(hdev->dev, "cid[%d] qid[%d] timeout, completion polled\n", - req->tag, spraidq->qid); - return BLK_EH_DONE; - } - - spraid_end_admin_request(req, cpu_to_le16(-EINVAL), 0, 0); - return BLK_EH_DONE; -} - static int spraid_get_ctrl_info(struct spraid_dev *hdev, struct spraid_ctrl_info *ctrl_info) { struct spraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + int ret; + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; memset(&admin_cmd, 0, sizeof(admin_cmd)); admin_cmd.get_info.opcode = SPRAID_ADMIN_GET_INFO; admin_cmd.get_info.type = SPRAID_GET_INFO_CTRL; + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + if (!ret) + memcpy(ctrl_info, data_ptr, sizeof(struct spraid_ctrl_info)); - return spraid_submit_admin_sync_cmd(hdev->admin_q, &admin_cmd, NULL, - ctrl_info, sizeof(struct spraid_ctrl_info), 0, 0, 0); + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, data_dma); + + return ret; } static int spraid_init_ctrl_info(struct spraid_dev *hdev) @@ -2416,6 +2448,11 @@ static int spraid_init_ctrl_info(struct spraid_dev *hdev) dev_info(hdev->dev, "[%s]sn = %s\n", __func__, hdev->ctrl_info->sn); dev_info(hdev->dev, "[%s]fr = %s\n", __func__, hdev->ctrl_info->fr); + if (!hdev->ctrl_info->aerl) + hdev->ctrl_info->aerl = 1; + if (hdev->ctrl_info->aerl > SPRAID_NR_AEN_COMMANDS) + hdev->ctrl_info->aerl = SPRAID_NR_AEN_COMMANDS; + return 0; } @@ -2444,98 +2481,51 @@ static void spraid_free_iod_ext_mem_pool(struct spraid_dev *hdev) mempool_destroy(hdev->iod_mempool); } -static int spraid_submit_user_cmd(struct request_queue *q, struct spraid_admin_command *cmd, - void __user *ubuffer, unsigned int bufflen, u32 *result, - unsigned int timeout) +static int spraid_user_admin_cmd(struct spraid_dev *hdev, struct bsg_job *job) { - struct request *req; - struct bio *bio = NULL; - int ret; - - req = spraid_alloc_admin_request(q, cmd, 0); - if (IS_ERR(req)) - return PTR_ERR(req); + struct spraid_bsg_request *bsg_req = job->request; + struct spraid_passthru_common_cmd *cmd = &(bsg_req->admcmd); + struct spraid_admin_command admin_cmd; + u32 timeout = msecs_to_jiffies(cmd->timeout_ms); + u32 result[2] = {0}; + int status; - req->timeout = timeout ? timeout : ADMIN_TIMEOUT; - spraid_admin_req(req)->flags |= SPRAID_REQ_USERCMD; + if (hdev->state >= SPRAID_RESETTING) { + dev_err(hdev->dev, "[%s] err, host state:[%d] is not right\n", + __func__, hdev->state); + return -EBUSY; + } - if (ubuffer && bufflen) { - ret = blk_rq_map_user(q, req, NULL, ubuffer, bufflen, GFP_KERNEL); - if (ret) - goto out; - bio = req->bio; + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.common.opcode = cmd->opcode; + admin_cmd.common.flags = cmd->flags; + admin_cmd.common.hdid = cpu_to_le32(cmd->nsid); + admin_cmd.common.cdw2[0] = cpu_to_le32(cmd->cdw2); + admin_cmd.common.cdw2[1] = cpu_to_le32(cmd->cdw3); + admin_cmd.common.cdw10 = cpu_to_le32(cmd->cdw10); + admin_cmd.common.cdw11 = cpu_to_le32(cmd->cdw11); + admin_cmd.common.cdw12 = cpu_to_le32(cmd->cdw12); + admin_cmd.common.cdw13 = cpu_to_le32(cmd->cdw13); + admin_cmd.common.cdw14 = cpu_to_le32(cmd->cdw14); + admin_cmd.common.cdw15 = cpu_to_le32(cmd->cdw15); + + status = spraid_bsg_map_data(hdev, job, &admin_cmd); + if (status) { + dev_err(hdev->dev, "[%s] err, map data failed\n", __func__); + return status; } - blk_execute_rq(req->q, NULL, req, 0); - if (spraid_admin_req(req)->flags & SPRAID_REQ_CANCELLED) - ret = -EINTR; - else - ret = spraid_admin_req(req)->status; - if (result) { - result[0] = spraid_admin_req(req)->result0; - result[1] = spraid_admin_req(req)->result1; - } - if (bio) - blk_rq_unmap_user(bio); -out: - blk_mq_free_request(req); - return ret; -} -static int spraid_user_admin_cmd(struct spraid_dev *hdev, - struct spraid_passthru_common_cmd __user *ucmd) -{ - struct spraid_passthru_common_cmd cmd; - struct spraid_admin_command admin_cmd; - u32 timeout = 0; - int status; - - if (!capable(CAP_SYS_ADMIN)) { - dev_err(hdev->dev, "Current user hasn't administrator right, reject service\n"); - return -EACCES; - } - - if (copy_from_user(&cmd, ucmd, sizeof(cmd))) { - dev_err(hdev->dev, "Copy command from user space to kernel space failed\n"); - return -EFAULT; - } - - if (cmd.flags) { - dev_err(hdev->dev, "Invalid flags in user command\n"); - return -EINVAL; + status = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, &result[0], &result[1], timeout); + if (status >= 0) { + job->reply_len = sizeof(result); + memcpy(job->reply, result, sizeof(result)); } - dev_info(hdev->dev, "user_admin_cmd opcode: 0x%x, subopcode: 0x%x\n", - cmd.opcode, cmd.cdw2 & 0x7ff); + if (status) + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x], status[0x%x] result0[0x%x] result1[0x%x]\n", + __func__, cmd->opcode, cmd->info_0.subopcode, status, result[0], result[1]); - memset(&admin_cmd, 0, sizeof(admin_cmd)); - admin_cmd.common.opcode = cmd.opcode; - admin_cmd.common.flags = cmd.flags; - admin_cmd.common.hdid = cpu_to_le32(cmd.nsid); - admin_cmd.common.cdw2[0] = cpu_to_le32(cmd.cdw2); - admin_cmd.common.cdw2[1] = cpu_to_le32(cmd.cdw3); - admin_cmd.common.cdw10 = cpu_to_le32(cmd.cdw10); - admin_cmd.common.cdw11 = cpu_to_le32(cmd.cdw11); - admin_cmd.common.cdw12 = cpu_to_le32(cmd.cdw12); - admin_cmd.common.cdw13 = cpu_to_le32(cmd.cdw13); - admin_cmd.common.cdw14 = cpu_to_le32(cmd.cdw14); - admin_cmd.common.cdw15 = cpu_to_le32(cmd.cdw15); - - if (cmd.timeout_ms) - timeout = msecs_to_jiffies(cmd.timeout_ms); - - status = spraid_submit_user_cmd(hdev->admin_q, &admin_cmd, - (void __user *)(uintptr_t)cmd.addr, cmd.info_1.data_len, - &cmd.result0, timeout); - - dev_info(hdev->dev, "user_admin_cmd status: 0x%x, result0: 0x%x, result1: 0x%x\n", - status, cmd.result0, cmd.result1); - - if (status >= 0) { - if (put_user(cmd.result0, &ucmd->result0)) - return -EFAULT; - if (put_user(cmd.result1, &ucmd->result1)) - return -EFAULT; - } + spraid_bsg_unmap_data(hdev, job); return status; } @@ -2548,8 +2538,8 @@ static int spraid_alloc_ioq_ptcmds(struct spraid_dev *hdev) INIT_LIST_HEAD(&hdev->ioq_pt_list); spin_lock_init(&hdev->ioq_pt_lock); - hdev->ioq_ptcmds = kcalloc_node(ptnum, sizeof(struct spraid_ioq_ptcmd), - GFP_KERNEL, hdev->numa_node); + hdev->ioq_ptcmds = kcalloc_node(ptnum, sizeof(struct spraid_cmd), + GFP_KERNEL, hdev->numa_node); if (!hdev->ioq_ptcmds) { dev_err(hdev->dev, "Alloc ioq_ptcmds failed\n"); @@ -2567,55 +2557,35 @@ static int spraid_alloc_ioq_ptcmds(struct spraid_dev *hdev) return 0; } -static struct spraid_ioq_ptcmd *spraid_get_ioq_ptcmd(struct spraid_dev *hdev) -{ - struct spraid_ioq_ptcmd *cmd = NULL; - unsigned long flags; - - spin_lock_irqsave(&hdev->ioq_pt_lock, flags); - if (list_empty(&hdev->ioq_pt_list)) { - spin_unlock_irqrestore(&hdev->ioq_pt_lock, flags); - dev_err(hdev->dev, "err, ioq ptcmd list empty\n"); - return NULL; - } - cmd = list_entry((&hdev->ioq_pt_list)->next, struct spraid_ioq_ptcmd, list); - list_del_init(&cmd->list); - spin_unlock_irqrestore(&hdev->ioq_pt_lock, flags); - - WRITE_ONCE(cmd->state, SPRAID_CMD_IDLE); - - return cmd; -} - -static void spraid_put_ioq_ptcmd(struct spraid_dev *hdev, struct spraid_ioq_ptcmd *cmd) +static void spraid_free_ioq_ptcmds(struct spraid_dev *hdev) { - unsigned long flags; + kfree(hdev->ioq_ptcmds); + hdev->ioq_ptcmds = NULL; - spin_lock_irqsave(&hdev->ioq_pt_lock, flags); - list_add(&cmd->list, (&hdev->ioq_pt_list)->next); - spin_unlock_irqrestore(&hdev->ioq_pt_lock, flags); + INIT_LIST_HEAD(&hdev->ioq_pt_list); } static int spraid_submit_ioq_sync_cmd(struct spraid_dev *hdev, struct spraid_ioq_command *cmd, - u32 *result, void **sense, u32 timeout) + u32 *result, u32 *reslen, u32 timeout) { - struct spraid_queue *ioq; int ret; dma_addr_t sense_dma; - struct spraid_ioq_ptcmd *pt_cmd = spraid_get_ioq_ptcmd(hdev); - - *sense = NULL; + struct spraid_queue *ioq; + void *sense_addr = NULL; + struct spraid_cmd *pt_cmd = spraid_get_cmd(hdev, SPRAID_CMD_IOPT); - if (!pt_cmd) + if (!pt_cmd) { + dev_err(hdev->dev, "err, get ioq cmd failed\n"); return -EFAULT; + } - dev_info(hdev->dev, "[%s] ptcmd, cid[%d], qid[%d]\n", __func__, pt_cmd->cid, pt_cmd->qid); + timeout = timeout ? timeout : ADMIN_TIMEOUT; init_completion(&pt_cmd->cmd_done); ioq = &hdev->queues[pt_cmd->qid]; ret = pt_cmd->cid * SCSI_SENSE_BUFFERSIZE; - pt_cmd->priv = ioq->sense + ret; + sense_addr = ioq->sense + ret; sense_dma = ioq->sense_dma_addr + ret; cmd->common.sense_addr = cpu_to_le64(sense_dma); @@ -2625,260 +2595,87 @@ static int spraid_submit_ioq_sync_cmd(struct spraid_dev *hdev, struct spraid_ioq spraid_submit_cmd(ioq, cmd); if (!wait_for_completion_timeout(&pt_cmd->cmd_done, timeout)) { - dev_err(hdev->dev, "[%s] cid[%d], qid[%d] timeout\n", - __func__, pt_cmd->cid, pt_cmd->qid); + dev_err(hdev->dev, "[%s] cid[%d] qid[%d] timeout, opcode[0x%x] subopcode[0x%x]\n", + __func__, pt_cmd->cid, pt_cmd->qid, cmd->common.opcode, + (le32_to_cpu(cmd->common.cdw3[0]) & 0xffff)); WRITE_ONCE(pt_cmd->state, SPRAID_CMD_TIMEOUT); - return -EINVAL; + spraid_put_cmd(hdev, pt_cmd, SPRAID_CMD_IOPT); + return -ETIME; } - if (result) { - result[0] = pt_cmd->result0; - result[1] = pt_cmd->result1; + if (result && reslen) { + if ((pt_cmd->status & 0x17f) == 0x101) { + memcpy(result, sense_addr, SCSI_SENSE_BUFFERSIZE); + *reslen = SCSI_SENSE_BUFFERSIZE; + } } - if ((pt_cmd->status & 0x17f) == 0x101) - *sense = pt_cmd->priv; + spraid_put_cmd(hdev, pt_cmd, SPRAID_CMD_IOPT); return pt_cmd->status; } -static int spraid_user_ioq_cmd(struct spraid_dev *hdev, - struct spraid_ioq_passthru_cmd __user *ucmd) +static int spraid_user_ioq_cmd(struct spraid_dev *hdev, struct bsg_job *job) { - struct spraid_ioq_passthru_cmd cmd; + struct spraid_bsg_request *bsg_req = (struct spraid_bsg_request *)(job->request); + struct spraid_ioq_passthru_cmd *cmd = &(bsg_req->ioqcmd); struct spraid_ioq_command ioq_cmd; - u32 timeout = 0; int status = 0; - u8 *data_ptr = NULL; - dma_addr_t data_dma; - enum dma_data_direction dma_dir = DMA_NONE; - void *sense = NULL; - - if (!capable(CAP_SYS_ADMIN)) { - dev_err(hdev->dev, "Current user hasn't administrator right, reject service\n"); - return -EACCES; - } + u32 timeout = msecs_to_jiffies(cmd->timeout_ms); - if (copy_from_user(&cmd, ucmd, sizeof(cmd))) { - dev_err(hdev->dev, "Copy command from user space to kernel space failed\n"); - return -EFAULT; - } - - if (cmd.data_len > PAGE_SIZE) { + if (cmd->data_len > PAGE_SIZE) { dev_err(hdev->dev, "[%s] data len bigger than 4k\n", __func__); return -EFAULT; } - dev_info(hdev->dev, "[%s] opcode: 0x%x, subopcode: 0x%x, datalen: %d\n", - __func__, cmd.opcode, cmd.info_1.subopcode, cmd.data_len); - - if (cmd.addr && cmd.data_len) { - data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &data_dma, GFP_KERNEL); - if (!data_ptr) - return -ENOMEM; - - dma_dir = (cmd.opcode & 1) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; + if (hdev->state != SPRAID_LIVE) { + dev_err(hdev->dev, "[%s] err, host state:[%d] is not live\n", + __func__, hdev->state); + return -EBUSY; } - if (dma_dir == DMA_TO_DEVICE) { - if (copy_from_user(data_ptr, (void __user *)(uintptr_t)cmd.addr, cmd.data_len)) { - dev_err(hdev->dev, "[%s] copy user data failed\n", __func__); - status = -EFAULT; - goto free_dma_mem; - } - } + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x] init, datalen[%d]\n", + __func__, cmd->opcode, cmd->info_1.subopcode, cmd->data_len); memset(&ioq_cmd, 0, sizeof(ioq_cmd)); - ioq_cmd.common.opcode = cmd.opcode; - ioq_cmd.common.flags = cmd.flags; - ioq_cmd.common.hdid = cpu_to_le32(cmd.nsid); - ioq_cmd.common.sense_len = cpu_to_le16(cmd.info_0.res_sense_len); - ioq_cmd.common.cdb_len = cmd.info_0.cdb_len; - ioq_cmd.common.rsvd2 = cmd.info_0.rsvd0; - ioq_cmd.common.cdw3[0] = cpu_to_le32(cmd.cdw3); - ioq_cmd.common.cdw3[1] = cpu_to_le32(cmd.cdw4); - ioq_cmd.common.cdw3[2] = cpu_to_le32(cmd.cdw5); - ioq_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); - - ioq_cmd.common.cdw10[0] = cpu_to_le32(cmd.cdw10); - ioq_cmd.common.cdw10[1] = cpu_to_le32(cmd.cdw11); - ioq_cmd.common.cdw10[2] = cpu_to_le32(cmd.cdw12); - ioq_cmd.common.cdw10[3] = cpu_to_le32(cmd.cdw13); - ioq_cmd.common.cdw10[4] = cpu_to_le32(cmd.cdw14); - ioq_cmd.common.cdw10[5] = cpu_to_le32(cmd.data_len); - - memcpy(ioq_cmd.common.cdb, &cmd.cdw16, cmd.info_0.cdb_len); - - ioq_cmd.common.cdw26[0] = cpu_to_le32(cmd.cdw26[0]); - ioq_cmd.common.cdw26[1] = cpu_to_le32(cmd.cdw26[1]); - ioq_cmd.common.cdw26[2] = cpu_to_le32(cmd.cdw26[2]); - ioq_cmd.common.cdw26[3] = cpu_to_le32(cmd.cdw26[3]); - - if (cmd.timeout_ms) - timeout = msecs_to_jiffies(cmd.timeout_ms); - timeout = timeout ? timeout : ADMIN_TIMEOUT; - - status = spraid_submit_ioq_sync_cmd(hdev, &ioq_cmd, &cmd.result0, &sense, timeout); - - if (status >= 0) { - if (put_user(cmd.result0, &ucmd->result0)) { - status = -EFAULT; - goto free_dma_mem; - } - if (put_user(cmd.result1, &ucmd->result1)) { - status = -EFAULT; - goto free_dma_mem; - } - if (dma_dir == DMA_FROM_DEVICE && - copy_to_user((void __user *)(uintptr_t)cmd.addr, data_ptr, cmd.data_len)) { - status = -EFAULT; - goto free_dma_mem; - } - } - - if (sense) { - if (copy_to_user((void *__user *)(uintptr_t)cmd.sense_addr, - sense, cmd.info_0.res_sense_len)) { - status = -EFAULT; - goto free_dma_mem; - } - } - -free_dma_mem: - if (data_ptr) - dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, data_dma); - - return status; - -} - -static int spraid_reset_work_sync(struct spraid_dev *hdev); - -static int spraid_user_reset_cmd(struct spraid_dev *hdev) -{ - int ret; - - dev_info(hdev->dev, "[%s] start user reset cmd\n", __func__); - ret = spraid_reset_work_sync(hdev); - dev_info(hdev->dev, "[%s] stop user reset cmd[%d]\n", __func__, ret); - - return ret; -} - -static int hdev_open(struct inode *inode, struct file *file) -{ - struct spraid_dev *hdev = - container_of(inode->i_cdev, struct spraid_dev, cdev); - file->private_data = hdev; - return 0; -} - -static long hdev_ioctl(struct file *file, u32 cmd, unsigned long arg) -{ - struct spraid_dev *hdev = file->private_data; - void __user *argp = (void __user *)arg; - - switch (cmd) { - case SPRAID_IOCTL_ADMIN_CMD: - return spraid_user_admin_cmd(hdev, argp); - case SPRAID_IOCTL_IOQ_CMD: - return spraid_user_ioq_cmd(hdev, argp); - case SPRAID_IOCTL_RESET_CMD: - return spraid_user_reset_cmd(hdev); - default: - return -ENOTTY; - } -} - -static const struct file_operations spraid_dev_fops = { - .owner = THIS_MODULE, - .open = hdev_open, - .unlocked_ioctl = hdev_ioctl, - .compat_ioctl = hdev_ioctl, -}; - -static int spraid_create_cdev(struct spraid_dev *hdev) -{ - int ret; - - device_initialize(&hdev->ctrl_device); - hdev->ctrl_device.devt = MKDEV(MAJOR(spraid_chr_devt), hdev->instance); - hdev->ctrl_device.class = spraid_class; - hdev->ctrl_device.parent = hdev->dev; - dev_set_drvdata(&hdev->ctrl_device, hdev); - ret = dev_set_name(&hdev->ctrl_device, "spraid%d", hdev->instance); - if (ret) - return ret; - cdev_init(&hdev->cdev, &spraid_dev_fops); - hdev->cdev.owner = THIS_MODULE; - ret = cdev_device_add(&hdev->cdev, &hdev->ctrl_device); - if (ret) { - dev_err(hdev->dev, "Add cdev failed, ret: %d", ret); - put_device(&hdev->ctrl_device); - kfree_const(hdev->ctrl_device.kobj.name); - return ret; - } - - return 0; -} - -static inline void spraid_remove_cdev(struct spraid_dev *hdev) -{ - cdev_device_del(&hdev->cdev, &hdev->ctrl_device); -} - -static const struct blk_mq_ops spraid_admin_mq_ops = { - .queue_rq = spraid_queue_admin_rq, - .complete = spraid_complete_admin_rq, - .init_hctx = spraid_admin_init_hctx, - .init_request = spraid_admin_init_request, - .timeout = spraid_admin_timeout, -}; - -static void spraid_remove_admin_tagset(struct spraid_dev *hdev) -{ - if (hdev->admin_q && !blk_queue_dying(hdev->admin_q)) { - blk_mq_unquiesce_queue(hdev->admin_q); - blk_cleanup_queue(hdev->admin_q); - blk_mq_free_tag_set(&hdev->admin_tagset); + ioq_cmd.common.opcode = cmd->opcode; + ioq_cmd.common.flags = cmd->flags; + ioq_cmd.common.hdid = cpu_to_le32(cmd->nsid); + ioq_cmd.common.sense_len = cpu_to_le16(cmd->info_0.res_sense_len); + ioq_cmd.common.cdb_len = cmd->info_0.cdb_len; + ioq_cmd.common.rsvd2 = cmd->info_0.rsvd0; + ioq_cmd.common.cdw3[0] = cpu_to_le32(cmd->cdw3); + ioq_cmd.common.cdw3[1] = cpu_to_le32(cmd->cdw4); + ioq_cmd.common.cdw3[2] = cpu_to_le32(cmd->cdw5); + + ioq_cmd.common.cdw10[0] = cpu_to_le32(cmd->cdw10); + ioq_cmd.common.cdw10[1] = cpu_to_le32(cmd->cdw11); + ioq_cmd.common.cdw10[2] = cpu_to_le32(cmd->cdw12); + ioq_cmd.common.cdw10[3] = cpu_to_le32(cmd->cdw13); + ioq_cmd.common.cdw10[4] = cpu_to_le32(cmd->cdw14); + ioq_cmd.common.cdw10[5] = cpu_to_le32(cmd->data_len); + + memcpy(ioq_cmd.common.cdb, &cmd->cdw16, cmd->info_0.cdb_len); + + ioq_cmd.common.cdw26[0] = cpu_to_le32(cmd->cdw26[0]); + ioq_cmd.common.cdw26[1] = cpu_to_le32(cmd->cdw26[1]); + ioq_cmd.common.cdw26[2] = cpu_to_le32(cmd->cdw26[2]); + ioq_cmd.common.cdw26[3] = cpu_to_le32(cmd->cdw26[3]); + + status = spraid_bsg_map_data(hdev, job, (struct spraid_admin_command *)&ioq_cmd); + if (status) { + dev_err(hdev->dev, "[%s] err, map data failed\n", __func__); + return status; } -} -static int spraid_alloc_admin_tags(struct spraid_dev *hdev) -{ - if (!hdev->admin_q) { - hdev->admin_tagset.ops = &spraid_admin_mq_ops; - hdev->admin_tagset.nr_hw_queues = 1; + status = spraid_submit_ioq_sync_cmd(hdev, &ioq_cmd, job->reply, &job->reply_len, timeout); - hdev->admin_tagset.queue_depth = SPRAID_AQ_MQ_TAG_DEPTH; - hdev->admin_tagset.timeout = ADMIN_TIMEOUT; - hdev->admin_tagset.numa_node = hdev->numa_node; - hdev->admin_tagset.cmd_size = - spraid_cmd_size(hdev, true, false); - hdev->admin_tagset.flags = BLK_MQ_F_NO_SCHED; - hdev->admin_tagset.driver_data = hdev; + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x], status[0x%x], reply_len[%d]\n", + __func__, cmd->opcode, cmd->info_1.subopcode, status, job->reply_len); - if (blk_mq_alloc_tag_set(&hdev->admin_tagset)) { - dev_err(hdev->dev, "Allocate admin tagset failed\n"); - return -ENOMEM; - } + spraid_bsg_unmap_data(hdev, job); - hdev->admin_q = blk_mq_init_queue(&hdev->admin_tagset); - if (IS_ERR(hdev->admin_q)) { - dev_err(hdev->dev, "Initialize admin request queue failed\n"); - blk_mq_free_tag_set(&hdev->admin_tagset); - return -ENOMEM; - } - if (!blk_get_queue(hdev->admin_q)) { - dev_err(hdev->dev, "Get admin request queue failed\n"); - spraid_remove_admin_tagset(hdev); - hdev->admin_q = NULL; - return -ENODEV; - } - } else { - blk_mq_unquiesce_queue(hdev->admin_q); - } - return 0; + return status; } static bool spraid_check_scmd_completed(struct scsi_cmnd *scmd) @@ -2891,7 +2688,7 @@ static bool spraid_check_scmd_completed(struct scsi_cmnd *scmd) spraid_get_tag_from_scmd(scmd, &hwq, &cid); spraidq = &hdev->queues[hwq]; if (READ_ONCE(iod->state) == SPRAID_CMD_COMPLETE || spraid_poll_cq(spraidq, cid)) { - dev_warn(hdev->dev, "cid[%d], qid[%d] has been completed\n", + dev_warn(hdev->dev, "cid[%d] qid[%d] has been completed\n", cid, spraidq->qid); return true; } @@ -2927,8 +2724,7 @@ static int spraid_send_abort_cmd(struct spraid_dev *hdev, u32 hdid, u16 qid, u16 admin_cmd.abort.sqid = cpu_to_le16(qid); admin_cmd.abort.cid = cpu_to_le16(cid); - return spraid_submit_admin_sync_cmd(hdev->admin_q, &admin_cmd, NULL, - NULL, 0, 0, 0, 0); + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); } /* send reset command by admin quueue temporary */ @@ -2941,8 +2737,7 @@ static int spraid_send_reset_cmd(struct spraid_dev *hdev, int type, u32 hdid) admin_cmd.reset.hdid = cpu_to_le32(hdid); admin_cmd.reset.type = type; - return spraid_submit_admin_sync_cmd(hdev->admin_q, &admin_cmd, NULL, - NULL, 0, 0, 0, 0); + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); } static bool spraid_change_host_state(struct spraid_dev *hdev, enum spraid_state newstate) @@ -3022,7 +2817,7 @@ static void spraid_back_fault_cqe(struct spraid_queue *ioq, struct spraid_comple scsi_dma_unmap(scmd); spraid_free_iod_res(hdev, iod); scmd->scsi_done(scmd); - dev_warn(hdev->dev, "Back fault CQE, cid[%d], qid[%d]\n", + dev_warn(hdev->dev, "Back fault CQE, cid[%d] qid[%d]\n", cqe->cmd_id, ioq->qid); } @@ -3032,6 +2827,8 @@ static void spraid_back_all_io(struct spraid_dev *hdev) struct spraid_queue *ioq; struct spraid_completion cqe = { 0 }; + scsi_block_requests(hdev->shost); + for (i = 1; i <= hdev->shost->nr_hw_queues; i++) { ioq = &hdev->queues[i]; for (j = 0; j < hdev->shost->can_queue; j++) { @@ -3039,6 +2836,8 @@ static void spraid_back_all_io(struct spraid_dev *hdev) spraid_back_fault_cqe(ioq, &cqe); } } + + scsi_unblock_requests(hdev->shost); } static void spraid_dev_disable(struct spraid_dev *hdev, bool shutdown) @@ -3106,17 +2905,13 @@ static void spraid_reset_work(struct work_struct *work) if (ret) goto pci_disable; - ret = spraid_alloc_admin_tags(hdev); - if (ret) - goto pci_disable; - ret = spraid_setup_io_queues(hdev); if (ret || hdev->online_queues <= hdev->shost->nr_hw_queues) goto pci_disable; spraid_change_host_state(hdev, SPRAID_LIVE); - spraid_send_aen(hdev); + spraid_send_all_aen(hdev); return; @@ -3174,8 +2969,8 @@ static int spraid_abort_handler(struct scsi_cmnd *scmd) scsi_print_command(scmd); - if (!spraid_wait_abnl_cmd_done(iod) || spraid_check_scmd_completed(scmd) || - hdev->state != SPRAID_LIVE) + if (hdev->state != SPRAID_LIVE || !spraid_wait_abnl_cmd_done(iod) || + spraid_check_scmd_completed(scmd)) return SUCCESS; hostdata = scmd->device->hostdata; @@ -3183,7 +2978,7 @@ static int spraid_abort_handler(struct scsi_cmnd *scmd) dev_warn(hdev->dev, "cid[%d] qid[%d] timeout, aborting\n", cid, hwq); ret = spraid_send_abort_cmd(hdev, hostdata->hdid, hwq, cid); - if (ret != ADMIN_ERR_TIMEOUT) { + if (ret != -ETIME) { ret = spraid_wait_abnl_cmd_done(iod); if (ret) { dev_warn(hdev->dev, "cid[%d] qid[%d] abort failed, not found\n", cid, hwq); @@ -3206,8 +3001,8 @@ static int spraid_tgt_reset_handler(struct scsi_cmnd *scmd) scsi_print_command(scmd); - if (!spraid_wait_abnl_cmd_done(iod) || spraid_check_scmd_completed(scmd) || - hdev->state != SPRAID_LIVE) + if (hdev->state != SPRAID_LIVE || !spraid_wait_abnl_cmd_done(iod) || + spraid_check_scmd_completed(scmd)) return SUCCESS; hostdata = scmd->device->hostdata; @@ -3241,8 +3036,8 @@ static int spraid_bus_reset_handler(struct scsi_cmnd *scmd) scsi_print_command(scmd); - if (!spraid_wait_abnl_cmd_done(iod) || spraid_check_scmd_completed(scmd) || - hdev->state != SPRAID_LIVE) + if (hdev->state != SPRAID_LIVE || !spraid_wait_abnl_cmd_done(iod) || + spraid_check_scmd_completed(scmd)) return SUCCESS; hostdata = scmd->device->hostdata; @@ -3272,7 +3067,7 @@ static int spraid_shost_reset_handler(struct scsi_cmnd *scmd) struct spraid_dev *hdev = shost_priv(scmd->device->host); scsi_print_command(scmd); - if (spraid_check_scmd_completed(scmd) || hdev->state != SPRAID_LIVE) + if (hdev->state != SPRAID_LIVE || spraid_check_scmd_completed(scmd)) return SUCCESS; spraid_get_tag_from_scmd(scmd, &hwq, &cid); @@ -3288,6 +3083,62 @@ static int spraid_shost_reset_handler(struct scsi_cmnd *scmd) return SUCCESS; } +static pci_ers_result_t spraid_pci_error_detected(struct pci_dev *pdev, + pci_channel_state_t state) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "enter pci error detect, state:%d\n", state); + + switch (state) { + case pci_channel_io_normal: + dev_warn(hdev->dev, "channel is normal, do nothing\n"); + + return PCI_ERS_RESULT_CAN_RECOVER; + case pci_channel_io_frozen: + dev_warn(hdev->dev, "channel io frozen, need reset controller\n"); + + scsi_block_requests(hdev->shost); + + spraid_change_host_state(hdev, SPRAID_RESETTING); + + return PCI_ERS_RESULT_NEED_RESET; + case pci_channel_io_perm_failure: + dev_warn(hdev->dev, "channel io failure, request disconnect\n"); + + return PCI_ERS_RESULT_DISCONNECT; + } + + return PCI_ERS_RESULT_NEED_RESET; +} + +static pci_ers_result_t spraid_pci_slot_reset(struct pci_dev *pdev) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "restart after slot reset\n"); + + pci_restore_state(pdev); + + if (!queue_work(spraid_wq, &hdev->reset_work)) { + dev_err(hdev->dev, "[%s] err, the device is resetting state\n", __func__); + return PCI_ERS_RESULT_NONE; + } + + flush_work(&hdev->reset_work); + + scsi_unblock_requests(hdev->shost); + + return PCI_ERS_RESULT_RECOVERED; +} + +static void spraid_reset_done(struct pci_dev *pdev) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "enter spraid reset done\n"); +} + static ssize_t csts_pp_show(struct device *cdev, struct device_attribute *attr, char *buf) { struct Scsi_Host *shost = class_to_shost(cdev); @@ -3347,7 +3198,7 @@ static ssize_t fw_version_show(struct device *cdev, struct device_attribute *att struct Scsi_Host *shost = class_to_shost(cdev); struct spraid_dev *hdev = shost_priv(shost); - return snprintf(buf, sizeof(hdev->ctrl_info->fr), "%s\n", hdev->ctrl_info->fr); + return snprintf(buf, PAGE_SIZE, "%s\n", hdev->ctrl_info->fr); } static DEVICE_ATTR_RO(csts_pp); @@ -3365,6 +3216,185 @@ static struct device_attribute *spraid_host_attrs[] = { NULL, }; +static int spraid_get_vd_info(struct spraid_dev *hdev, struct spraid_vd_info *vd_info, u16 vid) +{ + struct spraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + int ret; + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.usr_cmd.opcode = USR_CMD_READ; + admin_cmd.usr_cmd.info_0.subopcode = cpu_to_le16(USR_CMD_VDINFO); + admin_cmd.usr_cmd.info_1.data_len = cpu_to_le16(USR_CMD_RDLEN); + admin_cmd.usr_cmd.info_1.param_len = cpu_to_le16(VDINFO_PARAM_LEN); + admin_cmd.usr_cmd.cdw10 = cpu_to_le32(vid); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + if (!ret) + memcpy(vd_info, data_ptr, sizeof(struct spraid_vd_info)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, data_dma); + + return ret; +} + +static int spraid_get_bgtask(struct spraid_dev *hdev, struct spraid_bgtask *bgtask) +{ + struct spraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + int ret; + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.usr_cmd.opcode = USR_CMD_READ; + admin_cmd.usr_cmd.info_0.subopcode = cpu_to_le16(USR_CMD_BGTASK); + admin_cmd.usr_cmd.info_1.data_len = cpu_to_le16(USR_CMD_RDLEN); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + if (!ret) + memcpy(bgtask, data_ptr, sizeof(struct spraid_bgtask)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, data_dma); + + return ret; +} + +static ssize_t raid_level_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct spraid_dev *hdev; + struct spraid_vd_info *vd_info; + struct spraid_sdev_hostdata *hostdata; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !SPRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = spraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) + vd_info->rg_level = ARRAY_SIZE(raid_levels) - 1; + + ret = (vd_info->rg_level < ARRAY_SIZE(raid_levels)) ? + vd_info->rg_level : (ARRAY_SIZE(raid_levels) - 1); + + kfree(vd_info); + + return snprintf(buf, PAGE_SIZE, "RAID-%s\n", raid_levels[ret]); +} + +static ssize_t raid_state_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct spraid_dev *hdev; + struct spraid_vd_info *vd_info; + struct spraid_sdev_hostdata *hostdata; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !SPRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = spraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) { + vd_info->vd_status = 0; + vd_info->rg_id = 0xff; + } + + ret = (vd_info->vd_status < ARRAY_SIZE(raid_states)) ? vd_info->vd_status : 0; + + kfree(vd_info); + + return snprintf(buf, PAGE_SIZE, "%s\n", raid_states[ret]); +} + +static ssize_t raid_resync_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct spraid_dev *hdev; + struct spraid_vd_info *vd_info; + struct spraid_bgtask *bgtask; + struct spraid_sdev_hostdata *hostdata; + u8 rg_id, i, progress = 0; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !SPRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = spraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) + goto out; + + rg_id = vd_info->rg_id; + + bgtask = (struct spraid_bgtask *)vd_info; + ret = spraid_get_bgtask(hdev, bgtask); + if (ret) + goto out; + for (i = 0; i < bgtask->task_num; i++) { + if ((bgtask->bgtask[i].type == BGTASK_TYPE_REBUILD) && + (le16_to_cpu(bgtask->bgtask[i].vd_id) == rg_id)) + progress = bgtask->bgtask[i].progress; + } + +out: + kfree(vd_info); + return snprintf(buf, PAGE_SIZE, "%d\n", progress); +} + +static DEVICE_ATTR_RO(raid_level); +static DEVICE_ATTR_RO(raid_state); +static DEVICE_ATTR_RO(raid_resync); + +static struct device_attribute *spraid_dev_attrs[] = { + &dev_attr_raid_level, + &dev_attr_raid_state, + &dev_attr_raid_resync, + NULL, +}; + +static struct pci_error_handlers spraid_err_handler = { + .error_detected = spraid_pci_error_detected, + .slot_reset = spraid_pci_slot_reset, + .reset_done = spraid_reset_done, +}; + +static int spraid_sysfs_host_reset(struct Scsi_Host *shost, int reset_type) +{ + int ret; + struct spraid_dev *hdev = shost_priv(shost); + + dev_info(hdev->dev, "[%s] start sysfs host reset cmd\n", __func__); + ret = spraid_reset_work_sync(hdev); + dev_info(hdev->dev, "[%s] stop sysfs host reset cmd[%d]\n", __func__, ret); + + return ret; +} + static struct scsi_host_template spraid_driver_template = { .module = THIS_MODULE, .name = "Ramaxel Logic spraid driver", @@ -3379,9 +3409,11 @@ static struct scsi_host_template spraid_driver_template = { .eh_bus_reset_handler = spraid_bus_reset_handler, .eh_host_reset_handler = spraid_shost_reset_handler, .change_queue_depth = scsi_change_queue_depth, - .host_tagset = 1, + .host_tagset = 0, .this_id = -1, .shost_attrs = spraid_host_attrs, + .sdev_attrs = spraid_dev_attrs, + .host_reset = spraid_sysfs_host_reset, }; static void spraid_shutdown(struct pci_dev *pdev) @@ -3392,11 +3424,53 @@ static void spraid_shutdown(struct pci_dev *pdev) spraid_disable_admin_queue(hdev, true); } +/* bsg dispatch user command */ +static int spraid_bsg_host_dispatch(struct bsg_job *job) +{ + struct Scsi_Host *shost = dev_to_shost(job->dev); + struct spraid_dev *hdev = shost_priv(shost); + struct request *rq = blk_mq_rq_from_pdu(job); + struct spraid_bsg_request *bsg_req = job->request; + int ret = 0; + + dev_log_dbg(hdev->dev, "[%s] msgcode[%d], msglen[%d], timeout[%d], req_nsge[%d], req_len[%d]\n", + __func__, bsg_req->msgcode, job->request_len, rq->timeout, + job->request_payload.sg_cnt, job->request_payload.payload_len); + + job->reply_len = 0; + + switch (bsg_req->msgcode) { + case SPRAID_BSG_ADM: + ret = spraid_user_admin_cmd(hdev, job); + break; + case SPRAID_BSG_IOQ: + ret = spraid_user_ioq_cmd(hdev, job); + break; + default: + dev_info(hdev->dev, "[%s] unsupport msgcode[%d]\n", __func__, bsg_req->msgcode); + break; + } + + if (ret > 0) + ret = ret | (ret << 8); + + bsg_job_done(job, ret, 0); + return 0; +} + +static inline void spraid_remove_bsg(struct spraid_dev *hdev) +{ + if (hdev->bsg_queue) { + bsg_unregister_queue(hdev->bsg_queue); + blk_cleanup_queue(hdev->bsg_queue); + } +} static int spraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) { struct spraid_dev *hdev; struct Scsi_Host *shost; int node, ret; + char bsg_name[15]; shost = scsi_host_alloc(&spraid_driver_template, sizeof(*hdev)); if (!shost) { @@ -3421,10 +3495,10 @@ static int spraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) goto put_dev; init_rwsem(&hdev->devices_rwsem); - INIT_WORK(&hdev->aen_work, spraid_async_event_work); INIT_WORK(&hdev->scan_work, spraid_scan_work); INIT_WORK(&hdev->timesyn_work, spraid_timesyn_work); INIT_WORK(&hdev->reset_work, spraid_reset_work); + INIT_WORK(&hdev->fw_act_work, spraid_fw_act_work); spin_lock_init(&hdev->state_lock); ret = spraid_alloc_resources(hdev); @@ -3439,17 +3513,13 @@ static int spraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) if (ret) goto pci_disable; - ret = spraid_alloc_admin_tags(hdev); - if (ret) - goto disable_admin_q; - ret = spraid_init_ctrl_info(hdev); if (ret) - goto free_admin_tagset; + goto disable_admin_q; ret = spraid_alloc_iod_ext_mem_pool(hdev); if (ret) - goto free_admin_tagset; + goto disable_admin_q; ret = spraid_setup_io_queues(hdev); if (ret) @@ -3464,9 +3534,15 @@ static int spraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) goto remove_io_queues; } - ret = spraid_create_cdev(hdev); - if (ret) + snprintf(bsg_name, sizeof(bsg_name), "spraid%d", shost->host_no); + hdev->bsg_queue = bsg_setup_queue(&shost->shost_gendev, bsg_name, + spraid_bsg_host_dispatch, NULL, + spraid_cmd_size(hdev, true, false)); + if (IS_ERR(hdev->bsg_queue)) { + dev_err(hdev->dev, "err, setup bsg failed\n"); + hdev->bsg_queue = NULL; goto remove_io_queues; + } if (hdev->online_queues == SPRAID_ADMIN_QUEUE_NUM) { dev_warn(hdev->dev, "warn only admin queue can be used\n"); @@ -3475,11 +3551,11 @@ static int spraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) hdev->state = SPRAID_LIVE; - spraid_send_aen(hdev); + spraid_send_all_aen(hdev); ret = spraid_dev_list_init(hdev); if (ret) - goto remove_cdev; + goto remove_bsg; ret = spraid_configure_timestamp(hdev); if (ret) @@ -3487,20 +3563,18 @@ static int spraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) ret = spraid_alloc_ioq_ptcmds(hdev); if (ret) - goto remove_cdev; + goto remove_bsg; scsi_scan_host(hdev->shost); return 0; -remove_cdev: - spraid_remove_cdev(hdev); +remove_bsg: + spraid_remove_bsg(hdev); remove_io_queues: spraid_remove_io_queues(hdev); free_iod_mempool: spraid_free_iod_ext_mem_pool(hdev); -free_admin_tagset: - spraid_remove_admin_tagset(hdev); disable_admin_q: spraid_disable_admin_queue(hdev, false); pci_disable: @@ -3524,22 +3598,17 @@ static void spraid_remove(struct pci_dev *pdev) dev_info(hdev->dev, "enter spraid remove\n"); spraid_change_host_state(hdev, SPRAID_DELETING); + flush_work(&hdev->reset_work); - if (!pci_device_is_present(pdev)) { - scsi_block_requests(shost); + if (!pci_device_is_present(pdev)) spraid_back_all_io(hdev); - scsi_unblock_requests(shost); - } - flush_work(&hdev->reset_work); + spraid_remove_bsg(hdev); scsi_remove_host(shost); - - kfree(hdev->ioq_ptcmds); + spraid_free_ioq_ptcmds(hdev); kfree(hdev->devices); - spraid_remove_cdev(hdev); spraid_remove_io_queues(hdev); spraid_free_iod_ext_mem_pool(hdev); - spraid_remove_admin_tagset(hdev); spraid_disable_admin_queue(hdev, false); spraid_pci_disable(hdev); spraid_free_resources(hdev); @@ -3551,7 +3620,7 @@ static void spraid_remove(struct pci_dev *pdev) } static const struct pci_device_id spraid_id_table[] = { - { PCI_DEVICE(PCI_VENDOR_ID_RAMAXEL_LOGIC, SPRAID_SERVER_DEVICE_HAB_DID) }, + { PCI_DEVICE(PCI_VENDOR_ID_RAMAXEL_LOGIC, SPRAID_SERVER_DEVICE_HBA_DID) }, { PCI_DEVICE(PCI_VENDOR_ID_RAMAXEL_LOGIC, SPRAID_SERVER_DEVICE_RAID_DID) }, { 0, } }; @@ -3563,6 +3632,7 @@ static struct pci_driver spraid_driver = { .probe = spraid_probe, .remove = spraid_remove, .shutdown = spraid_shutdown, + .err_handler = &spraid_err_handler, }; static int __init spraid_init(void) @@ -3573,14 +3643,10 @@ static int __init spraid_init(void) if (!spraid_wq) return -ENOMEM; - ret = alloc_chrdev_region(&spraid_chr_devt, 0, SPRAID_MINORS, "spraid"); - if (ret < 0) - goto destroy_wq; - spraid_class = class_create(THIS_MODULE, "spraid"); if (IS_ERR(spraid_class)) { ret = PTR_ERR(spraid_class); - goto unregister_chrdev; + goto destroy_wq; } ret = pci_register_driver(&spraid_driver); @@ -3591,8 +3657,6 @@ static int __init spraid_init(void) destroy_class: class_destroy(spraid_class); -unregister_chrdev: - unregister_chrdev_region(spraid_chr_devt, SPRAID_MINORS); destroy_wq: destroy_workqueue(spraid_wq); @@ -3603,12 +3667,11 @@ static void __exit spraid_exit(void) { pci_unregister_driver(&spraid_driver); class_destroy(spraid_class); - unregister_chrdev_region(spraid_chr_devt, SPRAID_MINORS); destroy_workqueue(spraid_wq); ida_destroy(&spraid_instance_ida); } -MODULE_AUTHOR("Ramaxel Memory Technology"); +MODULE_AUTHOR("songyl(a)ramaxel.com"); MODULE_DESCRIPTION("Ramaxel Memory Technology SPraid Driver"); MODULE_LICENSE("GPL"); MODULE_VERSION(SPRAID_DRV_VERSION); -- 2.32.0

2 1

[PATCH 1.0-LTS ] Add new module parameters:time out
by Yanling Song 05 Jan '22

05 Jan '22

Ramaxel inclusion category: features bugzilla: https://gitee.com/openeuler/kernel/issues/I4ON8F CVE: NA Changes: 1. Split scmd_tmout_nonpt into two parameters: scmd_tmout_vd/scmd_tmout_rawdisk 2. Return -ETIME instead of -EINVAL when command is timeout. 3. Add one module parameters: max_io_force. Signed-off-by: Yanling Song <songyl(a)ramaxel.com> Reviewed-by: Jiang Yu<yujiang(a)ramaxel.com> --- drivers/scsi/spraid/spraid_main.c | 96 +++++++++++++++---------------- 1 file changed, 45 insertions(+), 51 deletions(-) diff --git a/drivers/scsi/spraid/spraid_main.c b/drivers/scsi/spraid/spraid_main.c index 519b39f44e91..7069582d741a 100644 --- a/drivers/scsi/spraid/spraid_main.c +++ b/drivers/scsi/spraid/spraid_main.c @@ -42,10 +42,17 @@ static u32 admin_tmout = 60; module_param(admin_tmout, uint, 0644); MODULE_PARM_DESC(admin_tmout, "admin commands timeout (seconds)"); -static u32 scmd_tmout_nonpt = 180; -module_param(scmd_tmout_nonpt, uint, 0644); -MODULE_PARM_DESC(scmd_tmout_nonpt, - "scsi commands timeout for rawdisk&raid(seconds)"); +static u32 scmd_tmout_rawdisk = 180; +module_param(scmd_tmout_rawdisk, uint, 0644); +MODULE_PARM_DESC(scmd_tmout_rawdisk, "scsi commands timeout for rawdisk(seconds)"); + +static u32 scmd_tmout_vd = 180; +module_param(scmd_tmout_vd, uint, 0644); +MODULE_PARM_DESC(scmd_tmout_vd, "scsi commands timeout for vd(seconds)"); + +static bool max_io_force; +module_param(max_io_force, bool, 0644); +MODULE_PARM_DESC(max_io_force, "force max_hw_sectors_kb = 1024, default false(performance first)"); static int ioq_depth_set(const char *val, const struct kernel_param *kp); static const struct kernel_param_ops ioq_depth_ops = { @@ -78,7 +85,7 @@ static unsigned char log_debug_switch; module_param_cb(log_debug_switch, &log_debug_switch_ops, &log_debug_switch, 0644); MODULE_PARM_DESC(log_debug_switch, - "set log state, default non-zero for switch on"); + "set log state, default zero for switch off"); static int small_pool_num_set(const char *val, const struct kernel_param *kp) { @@ -132,7 +139,6 @@ static struct workqueue_struct *spraid_wq; #define SPRAID_DRV_VERSION "1.0.0.0" #define ADMIN_TIMEOUT (admin_tmout * HZ) -#define ADMIN_ERR_TIMEOUT 32757 #define SPRAID_WAIT_ABNL_CMD_TIMEOUT (3 * 2) @@ -242,13 +248,6 @@ static int spraid_pci_enable(struct spraid_dev *hdev) goto disable; } - ret = pci_alloc_irq_vectors(pdev, 1, 1, PCI_IRQ_ALL_TYPES); - if (ret < 0) { - dev_err(hdev->dev, - "Allocate one IRQ for setup admin channel failed\n"); - goto disable; - } - hdev->cap = lo_hi_readq(hdev->bar + SPRAID_REG_CAP); hdev->ioq_depth = min_t(u32, SPRAID_CAP_MQES(hdev->cap) + 1, io_queue_depth); @@ -261,13 +260,21 @@ static int spraid_pci_enable(struct spraid_dev *hdev) maskbit); maskbit = SPRAID_DMA_MSK_BIT_MAX; } - if (dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(maskbit))) { + if (dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(maskbit)) && + dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(32))) { dev_err(hdev->dev, "set dma mask and coherent failed\n"); goto disable; } dev_info(hdev->dev, "set dma mask[%llu] success\n", maskbit); + ret = pci_alloc_irq_vectors(pdev, 1, 1, PCI_IRQ_ALL_TYPES); + + if (ret < 0) { + dev_err(hdev->dev, "Allocate one IRQ for setup admin channel failed\n"); + goto disable; + } + pci_enable_pcie_error_reporting(pdev); pci_save_state(pdev); @@ -840,6 +847,9 @@ static void spraid_map_status(struct spraid_iod *iod, struct scsi_cmnd *scmd, break; default: set_host_byte(scmd, DID_BAD_TARGET); + dev_warn(iod->spraidq->hdev->dev, "[%s] cid[%d] qid[%d];" + "bad status[0x%x]\n", __func__, cqe->cmd_id, + le16_to_cpu(cqe->sq_id), le16_to_cpu(cqe->status)); break; } } @@ -981,32 +991,17 @@ static void spraid_slave_destroy(struct scsi_device *sdev) static int spraid_slave_configure(struct scsi_device *sdev) { - u16 idx; - unsigned int timeout = scmd_tmout_nonpt * HZ; + unsigned int timeout = scmd_tmout_rawdisk * HZ; struct spraid_dev *hdev = shost_priv(sdev->host); struct spraid_sdev_hostdata *hostdata = sdev->hostdata; u32 max_sec = sdev->host->max_sectors; if (hostdata) { - idx = hostdata->hdid - 1; - if (sdev->channel == hdev->devices[idx].channel && - sdev->id == le16_to_cpu(hdev->devices[idx].target) && - sdev->lun < hdev->devices[idx].lun) { - if (SPRAID_DEV_INFO_ATTR_PT(hdev->devices[idx].attr)) - timeout = 30 * HZ; - else - timeout = scmd_tmout_nonpt * HZ; - max_sec = le16_to_cpu(hdev->devices[idx].max_io_kb) - << 1; - } else { - dev_err(hdev->dev, - "[%s] err, sdev->channel:id:lun[%d:%d:%lld];" - "devices[%d], channel:target:lun[%d:%d:%d]\n", - __func__, sdev->channel, sdev->id, sdev->lun, - idx, hdev->devices[idx].channel, - hdev->devices[idx].target, - hdev->devices[idx].lun); - } + if (SPRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + timeout = scmd_tmout_vd * HZ; + else if (SPRAID_DEV_INFO_ATTR_RAWDISK(hostdata->attr)) + timeout = scmd_tmout_rawdisk * HZ; + max_sec = hostdata->max_io_kb << 1; } else { dev_err(hdev->dev, "[%s] err, sdev->hostdata is null\n", __func__); @@ -1018,11 +1013,12 @@ static int spraid_slave_configure(struct scsi_device *sdev) if ((max_sec == 0) || (max_sec > sdev->host->max_sectors)) max_sec = sdev->host->max_sectors; - dev_info(hdev->dev, - "[%s] sdev->channel:id:lun[%d:%d:%lld];" - " scmd_timeout[%d]s, maxsec[%d]\n", - __func__, sdev->channel, sdev->id, - sdev->lun, timeout / HZ, max_sec); + if (!max_io_force) + blk_queue_max_hw_sectors(sdev->request_queue, max_sec); + + dev_info(hdev->dev, "[%s] sdev->channel:id:lun[%d:%d:%lld];" + "scmd_timeout[%d]s, maxsec[%d]\n", __func__, sdev->channel, + sdev->id, sdev->lun, timeout / HZ, max_sec); return 0; } @@ -1677,7 +1673,7 @@ static int spraid_submit_admin_sync_cmd(struct spraid_dev *hdev, cmd->usr_cmd.opcode, cmd->usr_cmd.info_0.subopcode); WRITE_ONCE(adm_cmd->state, SPRAID_CMD_TIMEOUT); spraid_put_cmd(hdev, adm_cmd, SPRAID_CMD_ADM); - return -EINVAL; + return -ETIME; } if (result0) @@ -2630,9 +2626,6 @@ static int spraid_user_admin_cmd(struct spraid_dev *hdev, struct bsg_job *job) return -EBUSY; } - dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x] init\n", - __func__, cmd->opcode, cmd->info_0.subopcode); - memset(&admin_cmd, 0, sizeof(admin_cmd)); admin_cmd.common.opcode = cmd->opcode; admin_cmd.common.flags = cmd->flags; @@ -2659,10 +2652,11 @@ static int spraid_user_admin_cmd(struct spraid_dev *hdev, struct bsg_job *job) memcpy(job->reply, result, sizeof(result)); } - dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x];" - " status[0x%x] result0[0x%x] result1[0x%x]\n", - __func__, cmd->opcode, cmd->info_0.subopcode, - status, result[0], result[1]); + if (status) + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x];" + " status[0x%x] result0[0x%x] result1[0x%x]\n", + __func__, cmd->opcode, cmd->info_0.subopcode, + status, result[0], result[1]); spraid_bsg_unmap_data(hdev, job); @@ -2742,7 +2736,7 @@ static int spraid_submit_ioq_sync_cmd(struct spraid_dev *hdev, (le32_to_cpu(cmd->common.cdw3[0]) & 0xffff)); WRITE_ONCE(pt_cmd->state, SPRAID_CMD_TIMEOUT); spraid_put_cmd(hdev, pt_cmd, SPRAID_CMD_IOPT); - return -EINVAL; + return -ETIME; } if (result && reslen) { @@ -3140,7 +3134,7 @@ static int spraid_abort_handler(struct scsi_cmnd *scmd) dev_warn(hdev->dev, "cid[%d] qid[%d] timeout, aborting\n", cid, hwq); ret = spraid_send_abort_cmd(hdev, hostdata->hdid, hwq, cid); - if (ret != ADMIN_ERR_TIMEOUT) { + if (ret != -ETIME) { ret = spraid_wait_abnl_cmd_done(iod); if (ret) { dev_warn(hdev->dev, "cid[%d] qid[%d] abort failed;" @@ -3623,7 +3617,7 @@ static int spraid_bsg_host_dispatch(struct bsg_job *job) struct spraid_bsg_request *bsg_req = job->request; int ret = 0; - dev_info(hdev->dev, "[%s] msgcode[%d], msglen[%d], timeout[%d];" + dev_log_dbg(hdev->dev, "[%s] msgcode[%d], msglen[%d], timeout[%d];" " req_nsge[%d], req_len[%d]\n", __func__, bsg_req->msgcode, job->request_len, rq->timeout, job->request_payload.sg_cnt, -- 2.27.0

1 0

[PATCH openEuler-5.10 1/2] kabi: reserve space for arm64 cpufeature related structure
by Zheng Zengkai 04 Jan '22

04 Jan '22

From: Jialin Zhang <zhangjialin11(a)huawei.com> hulk inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4JBL0 CVE: NA ------------------------------- Reserve space for the structure cpu_hwcap_keys and cpu_hwcaps. Signed-off-by: Jialin Zhang <zhangjialin11(a)huawei.com> Reviewed-by: wangxiongfeng <wangxiongfeng2(a)huawei.com> Reviewed-by: Yang Yingliang <yangyingliang(a)huawei.com> Reviewed-by: Hanjun Guo <guohanjun(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> --- arch/arm64/include/asm/cpucaps.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm64/include/asm/cpucaps.h b/arch/arm64/include/asm/cpucaps.h index 93bae3795165..c58a1919bfc9 100644 --- a/arch/arm64/include/asm/cpucaps.h +++ b/arch/arm64/include/asm/cpucaps.h @@ -72,6 +72,6 @@ #define ARM64_HAS_TWED 62 #define ARM64_WORKAROUND_HISILICON_1980005 63 -#define ARM64_NCAPS 64 +#define ARM64_NCAPS 80 #endif /* __ASM_CPUCAPS_H */ -- 2.20.1

1 1

[PATCH openEuler-1.0-LTS] drm/atomic-helper: Bump vblank timeout to 100 ms
by Gou Hao 04 Jan '22

04 Jan '22

From: Linus Walleij <linus.walleij(a)linaro.org> mainline inclusion from mainline-v5.3-rc1 commit b3198c38f02d54a5e964258a2180d502abe6eaf0 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I4P7YB CVE: N/A --- The 50 ms default timeout waiting for vblanks is too small for the first vblank from the ST-Ericsson MCDE display controller over DSI. Presumably this is because the DSI display is command-mode only and the state machine will take some time setting up its state for the first display update, and we hit a timeout. 100 ms makes it pass without problems. Signed-off-by: Linus Walleij <linus.walleij(a)linaro.org> Reviewed-by: Ville Syrjälä <ville.syrjala(a)linux.intel.com> Reviewed-by: Daniel Vetter <daniel.vetter(a)ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20190430093746.26485-1-linus.… Signed-off-by: Gou Hao <gouhao(a)uniontech.com> --- drivers/gpu/drm/drm_atomic_helper.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/drm_atomic_helper.c b/drivers/gpu/drm/drm_atomic_helper.c index 86efd2da37f9..553415fe8ede 100644 --- a/drivers/gpu/drm/drm_atomic_helper.c +++ b/drivers/gpu/drm/drm_atomic_helper.c @@ -1424,7 +1424,7 @@ drm_atomic_helper_wait_for_vblanks(struct drm_device *dev, ret = wait_event_timeout(dev->vblank[i].queue, old_state->crtcs[i].last_vblank_count != drm_crtc_vblank_count(crtc), - msecs_to_jiffies(50)); + msecs_to_jiffies(100)); WARN(!ret, "[CRTC:%d:%s] vblank wait timed out\n", crtc->base.id, crtc->name); -- 2.20.1

1 0

[PATCH openEuler-1.0-LTS] drm/atomic-helper: Bump vblank timeout to 100 ms
by Gou Hao 04 Jan '22

04 Jan '22

From: Linus Walleij <linus.walleij(a)linaro.org> mainline inclusion from mainline-v5.3-rc1 commit b3198c38f02d54a5e964258a2180d502abe6eaf0 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I4P7YB CVE: N/A --- The 50 ms default timeout waiting for vblanks is too small for the first vblank from the ST-Ericsson MCDE display controller over DSI. Presumably this is because the DSI display is command-mode only and the state machine will take some time setting up its state for the first display update, and we hit a timeout. 100 ms makes it pass without problems. Signed-off-by: Linus Walleij <linus.walleij(a)linaro.org> Reviewed-by: Ville Syrjälä <ville.syrjala(a)linux.intel.com> Reviewed-by: Daniel Vetter <daniel.vetter(a)ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20190430093746.26485-1-linus.… Signed-off-by: Gou Hao <gouhao(a)uniontech.com> --- drivers/gpu/drm/drm_atomic_helper.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/drm_atomic_helper.c b/drivers/gpu/drm/drm_atomic_helper.c index 86efd2da37f9..553415fe8ede 100644 --- a/drivers/gpu/drm/drm_atomic_helper.c +++ b/drivers/gpu/drm/drm_atomic_helper.c @@ -1424,7 +1424,7 @@ drm_atomic_helper_wait_for_vblanks(struct drm_device *dev, ret = wait_event_timeout(dev->vblank[i].queue, old_state->crtcs[i].last_vblank_count != drm_crtc_vblank_count(crtc), - msecs_to_jiffies(50)); + msecs_to_jiffies(100)); WARN(!ret, "[CRTC:%d:%s] vblank wait timed out\n", crtc->base.id, crtc->name); -- 2.20.1

1 0

[PATCH openEuler-1.0-LTS] mm: export collect_procs()
by Yang Yingliang 31 Dec '21

31 Dec '21

From: Zhang Jian <zhangjian210(a)huawei.com> ascend inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4OXH9 CVE: NA ------------------------------------------------- Collect the processes who have the page mapped via collect_procs(). @page if the page is a part of the hugepages/compound-page, we must using compound_head() to find it's head page to prevent the kernel panic, and make the page be locked. @to_kill the function will return a linked list, when we have used this list, we must kfree the list. @force_early if we want to find all process, we must make it be true, if it's false, the function will only return the process who have PF_MCE_PROCESS or PF_MCE_EARLY mark. limits: if force_early is true, sysctl_memory_failure_early_kill is useless. If it's false, no process have PF_MCE_PROCESS and PF_MCE_EARLY flag, and the sysctl_memory_failure_early_kill is enabled, function will return all tasks whether the task have the PF_MCE_PROCESS and PF_MCE_EARLY flag. Signed-off-by: Zhang Jian <zhangjian210(a)huawei.com> Reviewed-by: Weilong Chen <chenweilong(a)huawei.com> Reviewed-by: Kefeng Wang<wangkefeng.wang(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- include/linux/mm.h | 2 ++ mm/memory-failure.c | 3 ++- 2 files changed, 4 insertions(+), 1 deletion(-) diff --git a/include/linux/mm.h b/include/linux/mm.h index f801fa5e60289..7b724d39e6ee0 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -2866,6 +2866,8 @@ extern int sysctl_memory_failure_recovery; extern void shake_page(struct page *p, int access); extern atomic_long_t num_poisoned_pages __read_mostly; extern int soft_offline_page(struct page *page, int flags); +extern void collect_procs(struct page *page, struct list_head *tokill, + int force_early); /* diff --git a/mm/memory-failure.c b/mm/memory-failure.c index 52619a9bc3b0e..72e1746e5386f 100644 --- a/mm/memory-failure.c +++ b/mm/memory-failure.c @@ -512,7 +512,7 @@ static void collect_procs_file(struct page *page, struct list_head *to_kill, * First preallocate one tokill structure outside the spin locks, * so that we can kill at least one process reasonably reliable. */ -static void collect_procs(struct page *page, struct list_head *tokill, +void collect_procs(struct page *page, struct list_head *tokill, int force_early) { struct to_kill *tk; @@ -529,6 +529,7 @@ static void collect_procs(struct page *page, struct list_head *tokill, collect_procs_file(page, tokill, &tk, force_early); kfree(tk); } +EXPORT_SYMBOL_GPL(collect_procs); static const char *action_name[] = { [MF_IGNORED] = "Ignored", -- 2.25.1

1 0

[PATCH v4 ] scsi:spraid: initial commit of Ramaxel spraid driver
by Yanling Song 31 Dec '21

31 Dec '21

This initial commit contains Ramaxel's spraid module. Signed-off-by: Yanling Song <songyl(a)ramaxel.com> Changes from V3: 1. Use macro to repalce scmd_tmout_nonpt module parameter. Changes from V2: 1. Change "depends BLK_DEV_BSGLIB" to "select BLK_DEV_BSGLIB". 2. Add more descriptions in help. 3. Use pr_debug() instead of introducing dev_log_dbg(). 4. Use get_unaligned_be*() in spraid_setup_rw_cmd(); 5. Remove some unncessarry module parameters: scmd_tmout_pt, wait_abl_tmout, use_sgl_force 6. Some module parameters will be changed in the next version: admin_tmout, scmd_tmout_nonpt Changes from V1: 1. Use BSG module to replace with ioctrl 2. Use author's email as MODULE_AUTHOR 3. Remove "default=m" in drivers/scsi/spraid/Kconfig 4. To be changed in the next version: a. Use get_unaligned_be*() in spraid_setup_rw_cmd(); b. Use pr_debug() instead of introducing dev_log_dbg(). --- Documentation/scsi/spraid.rst | 16 + MAINTAINERS | 7 + drivers/scsi/Kconfig | 1 + drivers/scsi/Makefile | 1 + drivers/scsi/spraid/Kconfig | 13 + drivers/scsi/spraid/Makefile | 7 + drivers/scsi/spraid/spraid.h | 695 ++++++ drivers/scsi/spraid/spraid_main.c | 3266 +++++++++++++++++++++++++++++ 8 files changed, 4006 insertions(+) create mode 100644 Documentation/scsi/spraid.rst create mode 100644 drivers/scsi/spraid/Kconfig create mode 100644 drivers/scsi/spraid/Makefile create mode 100644 drivers/scsi/spraid/spraid.h create mode 100644 drivers/scsi/spraid/spraid_main.c diff --git a/Documentation/scsi/spraid.rst b/Documentation/scsi/spraid.rst new file mode 100644 index 000000000000..f87b02a6907b --- /dev/null +++ b/Documentation/scsi/spraid.rst @@ -0,0 +1,16 @@ +.. SPDX-License-Identifier: GPL-2.0 + +============================================== +SPRAID - Ramaxel SCSI Raid Controller driver +============================================== + +This file describes the spraid SCSI driver for Ramaxel +raid controllers. The spraid driver is the first generation raid driver for +Ramaxel Corp. + +For Ramaxel spraid controller support, enable the spraid driver +when configuring the kernel. + +Supported devices +================= +<Controller names to be added as they become publicly available.> diff --git a/MAINTAINERS b/MAINTAINERS index 57f7656fe513..68560008b259 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -17973,6 +17973,13 @@ F: include/dt-bindings/spmi/spmi.h F: include/linux/spmi.h F: include/trace/events/spmi.h +SPRAID SCSI/Raid DRIVERS +M: Yanling Song <yanling.song(a)ramaxel.com> +L: linux-scsi(a)vger.kernel.org +S: Maintained +F: Documentation/scsi/spraid.rst +F: drivers/scsi/spraid/ + SPU FILE SYSTEM M: Jeremy Kerr <jk(a)ozlabs.org> L: linuxppc-dev(a)lists.ozlabs.org diff --git a/drivers/scsi/Kconfig b/drivers/scsi/Kconfig index 6e3a04107bb6..3da5d26e1e11 100644 --- a/drivers/scsi/Kconfig +++ b/drivers/scsi/Kconfig @@ -501,6 +501,7 @@ source "drivers/scsi/mpt3sas/Kconfig" source "drivers/scsi/mpi3mr/Kconfig" source "drivers/scsi/smartpqi/Kconfig" source "drivers/scsi/ufs/Kconfig" +source "drivers/scsi/spraid/Kconfig" config SCSI_HPTIOP tristate "HighPoint RocketRAID 3xxx/4xxx Controller support" diff --git a/drivers/scsi/Makefile b/drivers/scsi/Makefile index 19814c26c908..192f8fb10a19 100644 --- a/drivers/scsi/Makefile +++ b/drivers/scsi/Makefile @@ -96,6 +96,7 @@ obj-$(CONFIG_SCSI_ZALON) += zalon7xx.o obj-$(CONFIG_SCSI_DC395x) += dc395x.o obj-$(CONFIG_SCSI_AM53C974) += esp_scsi.o am53c974.o obj-$(CONFIG_CXLFLASH) += cxlflash/ +obj-$(CONFIG_RAMAXEL_SPRAID) += spraid/ obj-$(CONFIG_MEGARAID_LEGACY) += megaraid.o obj-$(CONFIG_MEGARAID_NEWGEN) += megaraid/ obj-$(CONFIG_MEGARAID_SAS) += megaraid/ diff --git a/drivers/scsi/spraid/Kconfig b/drivers/scsi/spraid/Kconfig new file mode 100644 index 000000000000..bfbba3db8db0 --- /dev/null +++ b/drivers/scsi/spraid/Kconfig @@ -0,0 +1,13 @@ +# +# Ramaxel driver configuration +# + +config RAMAXEL_SPRAID + tristate "Ramaxel spraid Adapter" + depends on PCI && SCSI + select BLK_DEV_BSGLIB + depends on ARM64 || X86_64 + help + This driver supports Ramaxel SPRxxx serial + raid controller, which has PCIE Gen4 interface + with host and supports SAS/SATA Hdd/ssd. diff --git a/drivers/scsi/spraid/Makefile b/drivers/scsi/spraid/Makefile new file mode 100644 index 000000000000..ad2c2a84ddaf --- /dev/null +++ b/drivers/scsi/spraid/Makefile @@ -0,0 +1,7 @@ +# +# Makefile for the Ramaxel device drivers. +# + +obj-$(CONFIG_RAMAXEL_SPRAID) += spraid.o + +spraid-objs := spraid_main.o diff --git a/drivers/scsi/spraid/spraid.h b/drivers/scsi/spraid/spraid.h new file mode 100644 index 000000000000..617f5c2cdb82 --- /dev/null +++ b/drivers/scsi/spraid/spraid.h @@ -0,0 +1,695 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* Copyright(c) 2021 Ramaxel Memory Technology, Ltd */ + +#ifndef __SPRAID_H_ +#define __SPRAID_H_ + +#define SPRAID_CAP_MQES(cap) ((cap) & 0xffff) +#define SPRAID_CAP_STRIDE(cap) (((cap) >> 32) & 0xf) +#define SPRAID_CAP_MPSMIN(cap) (((cap) >> 48) & 0xf) +#define SPRAID_CAP_MPSMAX(cap) (((cap) >> 52) & 0xf) +#define SPRAID_CAP_TIMEOUT(cap) (((cap) >> 24) & 0xff) +#define SPRAID_CAP_DMAMASK(cap) (((cap) >> 37) & 0xff) + +#define SPRAID_DEFAULT_MAX_CHANNEL 4 +#define SPRAID_DEFAULT_MAX_ID 240 +#define SPRAID_DEFAULT_MAX_LUN_PER_HOST 8 +#define MAX_SECTORS 2048 + +#define IO_SQE_SIZE sizeof(struct spraid_ioq_command) +#define ADMIN_SQE_SIZE sizeof(struct spraid_admin_command) +#define SQE_SIZE(qid) (((qid) > 0) ? IO_SQE_SIZE : ADMIN_SQE_SIZE) +#define CQ_SIZE(depth) ((depth) * sizeof(struct spraid_completion)) +#define SQ_SIZE(qid, depth) ((depth) * SQE_SIZE(qid)) + +#define SENSE_SIZE(depth) ((depth) * SCSI_SENSE_BUFFERSIZE) + +#define SPRAID_AQ_DEPTH 128 +#define SPRAID_NR_AEN_COMMANDS 16 +#define SPRAID_AQ_BLK_MQ_DEPTH (SPRAID_AQ_DEPTH - SPRAID_NR_AEN_COMMANDS) +#define SPRAID_AQ_MQ_TAG_DEPTH (SPRAID_AQ_BLK_MQ_DEPTH - 1) + +#define SPRAID_ADMIN_QUEUE_NUM 1 +#define SPRAID_PTCMDS_PERQ 1 +#define SPRAID_IO_BLK_MQ_DEPTH (hdev->shost->can_queue) +#define SPRAID_NR_IOQ_PTCMDS (SPRAID_PTCMDS_PERQ * hdev->shost->nr_hw_queues) + +#define FUA_MASK 0x08 +#define SPRAID_MINORS BIT(MINORBITS) + +#define COMMAND_IS_WRITE(cmd) ((cmd)->common.opcode & 1) + +#define SPRAID_IO_IOSQES 7 +#define SPRAID_IO_IOCQES 4 +#define PRP_ENTRY_SIZE 8 + +#define SMALL_POOL_SIZE 256 +#define MAX_SMALL_POOL_NUM 16 +#define MAX_CMD_PER_DEV 64 +#define MAX_CDB_LEN 32 + +#define SPRAID_UP_TO_MULTY4(x) (((x) + 4) & (~0x03)) + +#define CQE_STATUS_SUCCESS (0x0) + +#define PCI_VENDOR_ID_RAMAXEL_LOGIC 0x1E81 + +#define SPRAID_SERVER_DEVICE_HBA_DID 0x2100 +#define SPRAID_SERVER_DEVICE_RAID_DID 0x2200 + +#define IO_6_DEFAULT_TX_LEN 256 + +#define SPRAID_INT_PAGES 2 +#define SPRAID_INT_BYTES(hdev) (SPRAID_INT_PAGES * (hdev)->page_size) + +enum { + SPRAID_REQ_CANCELLED = (1 << 0), + SPRAID_REQ_USERCMD = (1 << 1), +}; + +enum { + SPRAID_SC_SUCCESS = 0x0, + SPRAID_SC_INVALID_OPCODE = 0x1, + SPRAID_SC_INVALID_FIELD = 0x2, + + SPRAID_SC_ABORT_LIMIT = 0x103, + SPRAID_SC_ABORT_MISSING = 0x104, + SPRAID_SC_ASYNC_LIMIT = 0x105, + + SPRAID_SC_DNR = 0x4000, +}; + +enum { + SPRAID_REG_CAP = 0x0000, + SPRAID_REG_CC = 0x0014, + SPRAID_REG_CSTS = 0x001c, + SPRAID_REG_AQA = 0x0024, + SPRAID_REG_ASQ = 0x0028, + SPRAID_REG_ACQ = 0x0030, + SPRAID_REG_DBS = 0x1000, +}; + +enum { + SPRAID_CC_ENABLE = 1 << 0, + SPRAID_CC_CSS_NVM = 0 << 4, + SPRAID_CC_MPS_SHIFT = 7, + SPRAID_CC_AMS_SHIFT = 11, + SPRAID_CC_SHN_SHIFT = 14, + SPRAID_CC_IOSQES_SHIFT = 16, + SPRAID_CC_IOCQES_SHIFT = 20, + SPRAID_CC_AMS_RR = 0 << SPRAID_CC_AMS_SHIFT, + SPRAID_CC_SHN_NONE = 0 << SPRAID_CC_SHN_SHIFT, + SPRAID_CC_IOSQES = SPRAID_IO_IOSQES << SPRAID_CC_IOSQES_SHIFT, + SPRAID_CC_IOCQES = SPRAID_IO_IOCQES << SPRAID_CC_IOCQES_SHIFT, + SPRAID_CC_SHN_NORMAL = 1 << SPRAID_CC_SHN_SHIFT, + SPRAID_CC_SHN_MASK = 3 << SPRAID_CC_SHN_SHIFT, + SPRAID_CSTS_CFS_SHIFT = 1, + SPRAID_CSTS_SHST_SHIFT = 2, + SPRAID_CSTS_PP_SHIFT = 5, + SPRAID_CSTS_RDY = 1 << 0, + SPRAID_CSTS_SHST_CMPLT = 2 << 2, + SPRAID_CSTS_SHST_MASK = 3 << 2, + SPRAID_CSTS_CFS_MASK = 1 << SPRAID_CSTS_CFS_SHIFT, + SPRAID_CSTS_PP_MASK = 1 << SPRAID_CSTS_PP_SHIFT, +}; + +enum { + SPRAID_ADMIN_DELETE_SQ = 0x00, + SPRAID_ADMIN_CREATE_SQ = 0x01, + SPRAID_ADMIN_DELETE_CQ = 0x04, + SPRAID_ADMIN_CREATE_CQ = 0x05, + SPRAID_ADMIN_ABORT_CMD = 0x08, + SPRAID_ADMIN_SET_FEATURES = 0x09, + SPRAID_ADMIN_ASYNC_EVENT = 0x0c, + SPRAID_ADMIN_GET_INFO = 0xc6, + SPRAID_ADMIN_RESET = 0xc8, +}; + +enum { + SPRAID_GET_INFO_CTRL = 0, + SPRAID_GET_INFO_DEV_LIST = 1, +}; + +enum { + SPRAID_RESET_TARGET = 0, + SPRAID_RESET_BUS = 1, +}; + +enum { + SPRAID_AEN_ERROR = 0, + SPRAID_AEN_NOTICE = 2, + SPRAID_AEN_VS = 7, +}; + +enum { + SPRAID_AEN_DEV_CHANGED = 0x00, + SPRAID_AEN_HOST_PROBING = 0x10, +}; + +enum { + SPRAID_AEN_TIMESYN = 0x07 +}; + +enum { + SPRAID_CMD_WRITE = 0x01, + SPRAID_CMD_READ = 0x02, + + SPRAID_CMD_NONIO_NONE = 0x80, + SPRAID_CMD_NONIO_TODEV = 0x81, + SPRAID_CMD_NONIO_FROMDEV = 0x82, +}; + +enum { + SPRAID_QUEUE_PHYS_CONTIG = (1 << 0), + SPRAID_CQ_IRQ_ENABLED = (1 << 1), + + SPRAID_FEAT_NUM_QUEUES = 0x07, + SPRAID_FEAT_ASYNC_EVENT = 0x0b, + SPRAID_FEAT_TIMESTAMP = 0x0e, +}; + +enum spraid_state { + SPRAID_NEW, + SPRAID_LIVE, + SPRAID_RESETTING, + SPRAID_DELETING, + SPRAID_DEAD, +}; + +enum { + SPRAID_CARD_HBA, + SPRAID_CARD_RAID, +}; + +enum spraid_cmd_type { + SPRAID_CMD_ADM, + SPRAID_CMD_IOPT, +}; + +struct spraid_completion { + __le32 result; + union { + struct { + __u8 sense_len; + __u8 resv[3]; + }; + __le32 result1; + }; + __le16 sq_head; + __le16 sq_id; + __u16 cmd_id; + __le16 status; +}; + +struct spraid_ctrl_info { + __le32 nd; + __le16 max_cmds; + __le16 max_channel; + __le32 max_tgt_id; + __le16 max_lun; + __le16 max_num_sge; + __le16 lun_num_in_boot; + __u8 mdts; + __u8 acl; + __u8 aerl; + __u8 card_type; + __u16 rsvd; + __u32 rtd3e; + __u8 sn[32]; + __u8 fr[16]; + __u8 rsvd1[4020]; +}; + +struct spraid_dev { + struct pci_dev *pdev; + struct device *dev; + struct Scsi_Host *shost; + struct spraid_queue *queues; + struct dma_pool *prp_page_pool; + struct dma_pool *prp_small_pool[MAX_SMALL_POOL_NUM]; + mempool_t *iod_mempool; + void __iomem *bar; + u32 max_qid; + u32 num_vecs; + u32 queue_count; + u32 ioq_depth; + int db_stride; + u32 __iomem *dbs; + struct rw_semaphore devices_rwsem; + int numa_node; + u32 page_size; + u32 ctrl_config; + u32 online_queues; + u64 cap; + int instance; + struct spraid_ctrl_info *ctrl_info; + struct spraid_dev_info *devices; + + struct spraid_cmd *adm_cmds; + struct list_head adm_cmd_list; + spinlock_t adm_cmd_lock; + + struct spraid_cmd *ioq_ptcmds; + struct list_head ioq_pt_list; + spinlock_t ioq_pt_lock; + + struct work_struct scan_work; + struct work_struct timesyn_work; + struct work_struct reset_work; + + enum spraid_state state; + spinlock_t state_lock; + struct request_queue *bsg_queue; +}; + +struct spraid_sgl_desc { + __le64 addr; + __le32 length; + __u8 rsvd[3]; + __u8 type; +}; + +union spraid_data_ptr { + struct { + __le64 prp1; + __le64 prp2; + }; + struct spraid_sgl_desc sgl; +}; + +struct spraid_admin_common_command { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __le32 cdw2[4]; + union spraid_data_ptr dptr; + __le32 cdw10; + __le32 cdw11; + __le32 cdw12; + __le32 cdw13; + __le32 cdw14; + __le32 cdw15; +}; + +struct spraid_features { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __u64 rsvd2[2]; + union spraid_data_ptr dptr; + __le32 fid; + __le32 dword11; + __le32 dword12; + __le32 dword13; + __le32 dword14; + __le32 dword15; +}; + +struct spraid_create_cq { + __u8 opcode; + __u8 flags; + __u16 command_id; + __u32 rsvd1[5]; + __le64 prp1; + __u64 rsvd8; + __le16 cqid; + __le16 qsize; + __le16 cq_flags; + __le16 irq_vector; + __u32 rsvd12[4]; +}; + +struct spraid_create_sq { + __u8 opcode; + __u8 flags; + __u16 command_id; + __u32 rsvd1[5]; + __le64 prp1; + __u64 rsvd8; + __le16 sqid; + __le16 qsize; + __le16 sq_flags; + __le16 cqid; + __u32 rsvd12[4]; +}; + +struct spraid_delete_queue { + __u8 opcode; + __u8 flags; + __u16 command_id; + __u32 rsvd1[9]; + __le16 qid; + __u16 rsvd10; + __u32 rsvd11[5]; +}; + +struct spraid_get_info { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __u32 rsvd2[4]; + union spraid_data_ptr dptr; + __u8 type; + __u8 rsvd10[3]; + __le32 cdw11; + __u32 rsvd12[4]; +}; + +struct spraid_usr_cmd { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + union { + struct { + __le16 subopcode; + __le16 rsvd1; + } info_0; + __le32 cdw2; + }; + union { + struct { + __le16 data_len; + __le16 param_len; + } info_1; + __le32 cdw3; + }; + __u64 metadata; + union spraid_data_ptr dptr; + __le32 cdw10; + __le32 cdw11; + __le32 cdw12; + __le32 cdw13; + __le32 cdw14; + __le32 cdw15; +}; + +enum { + SPRAID_CMD_FLAG_SGL_METABUF = (1 << 6), + SPRAID_CMD_FLAG_SGL_METASEG = (1 << 7), + SPRAID_CMD_FLAG_SGL_ALL = SPRAID_CMD_FLAG_SGL_METABUF | SPRAID_CMD_FLAG_SGL_METASEG, +}; + +enum spraid_cmd_state { + SPRAID_CMD_IDLE = 0, + SPRAID_CMD_IN_FLIGHT = 1, + SPRAID_CMD_COMPLETE = 2, + SPRAID_CMD_TIMEOUT = 3, + SPRAID_CMD_TMO_COMPLETE = 4, +}; + +struct spraid_abort_cmd { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __u64 rsvd2[4]; + __le16 sqid; + __le16 cid; + __u32 rsvd11[5]; +}; + +struct spraid_reset_cmd { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __u64 rsvd2[4]; + __u8 type; + __u8 rsvd10[3]; + __u32 rsvd11[5]; +}; + +struct spraid_admin_command { + union { + struct spraid_admin_common_command common; + struct spraid_features features; + struct spraid_create_cq create_cq; + struct spraid_create_sq create_sq; + struct spraid_delete_queue delete_queue; + struct spraid_get_info get_info; + struct spraid_abort_cmd abort; + struct spraid_reset_cmd reset; + struct spraid_usr_cmd usr_cmd; + }; +}; + +struct spraid_ioq_common_command { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __le16 sense_len; + __u8 cdb_len; + __u8 rsvd2; + __le32 cdw3[3]; + union spraid_data_ptr dptr; + __le32 cdw10[6]; + __u8 cdb[32]; + __le64 sense_addr; + __le32 cdw26[6]; +}; + +struct spraid_rw_command { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __le16 sense_len; + __u8 cdb_len; + __u8 rsvd2; + __u32 rsvd3[3]; + union spraid_data_ptr dptr; + __le64 slba; + __le16 nlb; + __le16 control; + __u32 rsvd13[3]; + __u8 cdb[32]; + __le64 sense_addr; + __u32 rsvd26[6]; +}; + +struct spraid_scsi_nonio { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __le16 sense_len; + __u8 cdb_length; + __u8 rsvd2; + __u32 rsvd3[3]; + union spraid_data_ptr dptr; + __u32 rsvd10[5]; + __le32 buffer_len; + __u8 cdb[32]; + __le64 sense_addr; + __u32 rsvd26[6]; +}; + +struct spraid_ioq_command { + union { + struct spraid_ioq_common_command common; + struct spraid_rw_command rw; + struct spraid_scsi_nonio scsi_nonio; + }; +}; + +struct spraid_passthru_common_cmd { + __u8 opcode; + __u8 flags; + __u16 rsvd0; + __u32 nsid; + union { + struct { + __u16 subopcode; + __u16 rsvd1; + } info_0; + __u32 cdw2; + }; + union { + struct { + __u16 data_len; + __u16 param_len; + } info_1; + __u32 cdw3; + }; + __u64 metadata; + + __u64 addr; + __u64 prp2; + + __u32 cdw10; + __u32 cdw11; + __u32 cdw12; + __u32 cdw13; + __u32 cdw14; + __u32 cdw15; + __u32 timeout_ms; + __u32 result0; + __u32 result1; +}; + +struct spraid_ioq_passthru_cmd { + __u8 opcode; + __u8 flags; + __u16 rsvd0; + __u32 nsid; + union { + struct { + __u16 res_sense_len; + __u8 cdb_len; + __u8 rsvd0; + } info_0; + __u32 cdw2; + }; + union { + struct { + __u16 subopcode; + __u16 rsvd1; + } info_1; + __u32 cdw3; + }; + union { + struct { + __u16 rsvd; + __u16 param_len; + } info_2; + __u32 cdw4; + }; + __u32 cdw5; + __u64 addr; + __u64 prp2; + union { + struct { + __u16 eid; + __u16 sid; + } info_3; + __u32 cdw10; + }; + union { + struct { + __u16 did; + __u8 did_flag; + __u8 rsvd2; + } info_4; + __u32 cdw11; + }; + __u32 cdw12; + __u32 cdw13; + __u32 cdw14; + __u32 data_len; + __u32 cdw16; + __u32 cdw17; + __u32 cdw18; + __u32 cdw19; + __u32 cdw20; + __u32 cdw21; + __u32 cdw22; + __u32 cdw23; + __u64 sense_addr; + __u32 cdw26[4]; + __u32 timeout_ms; + __u32 result0; + __u32 result1; +}; + +struct spraid_bsg_request { + u32 msgcode; + u32 control; + union { + struct spraid_passthru_common_cmd admcmd; + struct spraid_ioq_passthru_cmd ioqcmd; + }; +}; + +enum { + SPRAID_BSG_ADM, + SPRAID_BSG_IOQ, +}; + +struct spraid_cmd { + int qid; + int cid; + u32 result0; + u32 result1; + u16 status; + void *priv; + enum spraid_cmd_state state; + struct completion cmd_done; + struct list_head list; +}; + +struct spraid_queue { + struct spraid_dev *hdev; + spinlock_t sq_lock; /* spinlock for lock handling */ + + spinlock_t cq_lock ____cacheline_aligned_in_smp; /* spinlock for lock handling */ + + void *sq_cmds; + + struct spraid_completion *cqes; + + dma_addr_t sq_dma_addr; + dma_addr_t cq_dma_addr; + u32 __iomem *q_db; + u8 cq_phase; + u8 sqes; + u16 qid; + u16 sq_tail; + u16 cq_head; + u16 last_cq_head; + u16 q_depth; + s16 cq_vector; + void *sense; + dma_addr_t sense_dma_addr; + struct dma_pool *prp_small_pool; +}; + +struct spraid_iod { + struct spraid_queue *spraidq; + enum spraid_cmd_state state; + int npages; + u32 nsge; + u32 length; + bool use_sgl; + bool sg_drv_mgmt; + dma_addr_t first_dma; + void *sense; + dma_addr_t sense_dma; + struct scatterlist *sg; + struct scatterlist inline_sg[0]; +}; + +#define SPRAID_DEV_INFO_ATTR_BOOT(attr) ((attr) & 0x01) +#define SPRAID_DEV_INFO_ATTR_VD(attr) (((attr) & 0x02) == 0x0) +#define SPRAID_DEV_INFO_ATTR_PT(attr) (((attr) & 0x22) == 0x02) +#define SPRAID_DEV_INFO_ATTR_RAWDISK(attr) ((attr) & 0x20) + +#define SPRAID_DEV_INFO_FLAG_VALID(flag) ((flag) & 0x01) +#define SPRAID_DEV_INFO_FLAG_CHANGE(flag) ((flag) & 0x02) + +struct spraid_dev_info { + __le32 hdid; + __le16 target; + __u8 channel; + __u8 lun; + __u8 attr; + __u8 flag; + __le16 max_io_kb; +}; + +#define MAX_DEV_ENTRY_PER_PAGE_4K 340 +struct spraid_dev_list { + __le32 dev_num; + __u32 rsvd0[3]; + struct spraid_dev_info devices[MAX_DEV_ENTRY_PER_PAGE_4K]; +}; + +struct spraid_sdev_hostdata { + u32 hdid; +}; + +#endif diff --git a/drivers/scsi/spraid/spraid_main.c b/drivers/scsi/spraid/spraid_main.c new file mode 100644 index 000000000000..7edce06b62a4 --- /dev/null +++ b/drivers/scsi/spraid/spraid_main.c @@ -0,0 +1,3266 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright(c) 2021 Ramaxel Memory Technology, Ltd */ + +/* Ramaxel Raid SPXXX Series Linux Driver */ + +#define pr_fmt(fmt) "spraid: " fmt + +#include <linux/sched/signal.h> +#include <linux/pci.h> +#include <linux/aer.h> +#include <linux/module.h> +#include <linux/ioport.h> +#include <linux/device.h> +#include <linux/delay.h> +#include <linux/interrupt.h> +#include <linux/cdev.h> +#include <linux/sysfs.h> +#include <linux/gfp.h> +#include <linux/types.h> +#include <linux/ratelimit.h> +#include <linux/once.h> +#include <linux/debugfs.h> +#include <linux/io-64-nonatomic-lo-hi.h> +#include <linux/blkdev.h> +#include <linux/bsg-lib.h> +#include <asm/unaligned.h> + +#include <scsi/scsi.h> +#include <scsi/scsi_cmnd.h> +#include <scsi/scsi_device.h> +#include <scsi/scsi_host.h> +#include <scsi/scsi_transport.h> +#include <scsi/scsi_dbg.h> +#include <scsi/sg.h> + + +#include "spraid.h" + +#define SCMD_TMOUT_RAWDISK 180 + +#define SCMD_TMOUT_VD 180 + +static int ioq_depth_set(const char *val, const struct kernel_param *kp); +static const struct kernel_param_ops ioq_depth_ops = { + .set = ioq_depth_set, + .get = param_get_uint, +}; + +static u32 io_queue_depth = 1024; +module_param_cb(io_queue_depth, &ioq_depth_ops, &io_queue_depth, 0644); +MODULE_PARM_DESC(io_queue_depth, "set io queue depth, should >= 2"); + +static int small_pool_num_set(const char *val, const struct kernel_param *kp) +{ + u8 n = 0; + int ret; + + ret = kstrtou8(val, 10, &n); + if (ret != 0) + return -EINVAL; + if (n > MAX_SMALL_POOL_NUM) + n = MAX_SMALL_POOL_NUM; + if (n < 1) + n = 1; + *((u8 *)kp->arg) = n; + + return 0; +} + +static const struct kernel_param_ops small_pool_num_ops = { + .set = small_pool_num_set, + .get = param_get_byte, +}; + +/* It was found that the spindlock of a single pool conflicts + * a lot with multiple CPUs.So multiple pools are introduced + * to reduce the conflictions. + */ +static unsigned char small_pool_num = 4; +module_param_cb(small_pool_num, &small_pool_num_ops, &small_pool_num, 0644); +MODULE_PARM_DESC(small_pool_num, "set prp small pool num, default 4, MAX 16"); + +static void spraid_free_queue(struct spraid_queue *spraidq); +static void spraid_handle_aen_notice(struct spraid_dev *hdev, u32 result); +static void spraid_handle_aen_vs(struct spraid_dev *hdev, u32 result, u32 result1); + +static DEFINE_IDA(spraid_instance_ida); + +static struct class *spraid_class; + +#define SPRAID_CAP_TIMEOUT_UNIT_MS (HZ / 2) + +static struct workqueue_struct *spraid_wq; + +#define SPRAID_DRV_VERSION "1.0.0.0" + +#define ADMIN_TIMEOUT (60 * HZ) +#define ADMIN_ERR_TIMEOUT 32757 + +#define SPRAID_WAIT_ABNL_CMD_TIMEOUT (3 * 2) + +#define SPRAID_DMA_MSK_BIT_MAX 64 + +enum FW_STAT_CODE { + FW_STAT_OK = 0, + FW_STAT_NEED_CHECK, + FW_STAT_ERROR, + FW_STAT_EP_PCIE_ERROR, + FW_STAT_NAC_DMA_ERROR, + FW_STAT_ABORTED, + FW_STAT_NEED_RETRY +}; + +static int ioq_depth_set(const char *val, const struct kernel_param *kp) +{ + int n = 0; + int ret; + + ret = kstrtoint(val, 10, &n); + if (ret != 0 || n < 2) + return -EINVAL; + + return param_set_int(val, kp); +} + +static int spraid_remap_bar(struct spraid_dev *hdev, u32 size) +{ + struct pci_dev *pdev = hdev->pdev; + + if (size > pci_resource_len(pdev, 0)) { + dev_err(hdev->dev, "Input size[%u] exceed bar0 length[%llu]\n", + size, pci_resource_len(pdev, 0)); + return -ENOMEM; + } + + if (hdev->bar) + iounmap(hdev->bar); + + hdev->bar = ioremap(pci_resource_start(pdev, 0), size); + if (!hdev->bar) { + dev_err(hdev->dev, "ioremap for bar0 failed\n"); + return -ENOMEM; + } + hdev->dbs = hdev->bar + SPRAID_REG_DBS; + + return 0; +} + +static int spraid_dev_map(struct spraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + int ret; + + ret = pci_request_mem_regions(pdev, "spraid"); + if (ret) { + dev_err(hdev->dev, "fail to request memory regions\n"); + return ret; + } + + ret = spraid_remap_bar(hdev, SPRAID_REG_DBS + 4096); + if (ret) { + pci_release_mem_regions(pdev); + return ret; + } + + return 0; +} + +static void spraid_dev_unmap(struct spraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + + if (hdev->bar) { + iounmap(hdev->bar); + hdev->bar = NULL; + } + pci_release_mem_regions(pdev); +} + +static int spraid_pci_enable(struct spraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + int ret = -ENOMEM; + u64 maskbit = SPRAID_DMA_MSK_BIT_MAX; + + if (pci_enable_device_mem(pdev)) { + dev_err(hdev->dev, "Enable pci device memory resources failed\n"); + return ret; + } + pci_set_master(pdev); + + if (readl(hdev->bar + SPRAID_REG_CSTS) == U32_MAX) { + ret = -ENODEV; + dev_err(hdev->dev, "Read csts register failed\n"); + goto disable; + } + + ret = pci_alloc_irq_vectors(pdev, 1, 1, PCI_IRQ_ALL_TYPES); + if (ret < 0) { + dev_err(hdev->dev, "Allocate one IRQ for setup admin channel failed\n"); + goto disable; + } + + hdev->cap = lo_hi_readq(hdev->bar + SPRAID_REG_CAP); + hdev->ioq_depth = min_t(u32, SPRAID_CAP_MQES(hdev->cap) + 1, io_queue_depth); + hdev->db_stride = 1 << SPRAID_CAP_STRIDE(hdev->cap); + + maskbit = SPRAID_CAP_DMAMASK(hdev->cap); + if (maskbit < 32 || maskbit > SPRAID_DMA_MSK_BIT_MAX) { + dev_err(hdev->dev, "err, dma mask invalid[%llu], set to default\n", maskbit); + maskbit = SPRAID_DMA_MSK_BIT_MAX; + } + if (dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(maskbit))) { + dev_err(hdev->dev, "set dma mask and coherent failed\n"); + goto disable; + } + + dev_info(hdev->dev, "set dma mask[%llu] success\n", maskbit); + + pci_enable_pcie_error_reporting(pdev); + pci_save_state(pdev); + + return 0; + +disable: + pci_disable_device(pdev); + return ret; +} + +static int spraid_npages_prp(u32 size, struct spraid_dev *hdev) +{ + u32 nprps = DIV_ROUND_UP(size + hdev->page_size, hdev->page_size); + + return DIV_ROUND_UP(PRP_ENTRY_SIZE * nprps, PAGE_SIZE - PRP_ENTRY_SIZE); +} + +static int spraid_npages_sgl(u32 nseg) +{ + return DIV_ROUND_UP(nseg * sizeof(struct spraid_sgl_desc), PAGE_SIZE); +} + +static void **spraid_iod_list(struct spraid_iod *iod) +{ + return (void **)(iod->inline_sg + (iod->sg_drv_mgmt ? iod->nsge : 0)); +} + +static u32 spraid_iod_ext_size(struct spraid_dev *hdev, u32 size, u32 nsge, + bool sg_drv_mgmt, bool use_sgl) +{ + size_t alloc_size, sg_size; + + if (use_sgl) + alloc_size = sizeof(__le64 *) * spraid_npages_sgl(nsge); + else + alloc_size = sizeof(__le64 *) * spraid_npages_prp(size, hdev); + + sg_size = sg_drv_mgmt ? (sizeof(struct scatterlist) * nsge) : 0; + return sg_size + alloc_size; +} + +static u32 spraid_cmd_size(struct spraid_dev *hdev, bool sg_drv_mgmt, bool use_sgl) +{ + u32 alloc_size = spraid_iod_ext_size(hdev, SPRAID_INT_BYTES(hdev), + SPRAID_INT_PAGES, sg_drv_mgmt, use_sgl); + + dev_info(hdev->dev, "sg_drv_mgmt: %s, use_sgl: %s, iod size: %lu, alloc_size: %u\n", + sg_drv_mgmt ? "true" : "false", use_sgl ? "true" : "false", + sizeof(struct spraid_iod), alloc_size); + + return sizeof(struct spraid_iod) + alloc_size; +} + +static int spraid_setup_prps(struct spraid_dev *hdev, struct spraid_iod *iod) +{ + struct scatterlist *sg = iod->sg; + u64 dma_addr = sg_dma_address(sg); + int dma_len = sg_dma_len(sg); + __le64 *prp_list, *old_prp_list; + u32 page_size = hdev->page_size; + int offset = dma_addr & (page_size - 1); + void **list = spraid_iod_list(iod); + int length = iod->length; + struct dma_pool *pool; + dma_addr_t prp_dma; + int nprps, i; + + length -= (page_size - offset); + if (length <= 0) { + iod->first_dma = 0; + return 0; + } + + dma_len -= (page_size - offset); + if (dma_len) { + dma_addr += (page_size - offset); + } else { + sg = sg_next(sg); + dma_addr = sg_dma_address(sg); + dma_len = sg_dma_len(sg); + } + + if (length <= page_size) { + iod->first_dma = dma_addr; + return 0; + } + + nprps = DIV_ROUND_UP(length, page_size); + if (nprps <= (SMALL_POOL_SIZE / PRP_ENTRY_SIZE)) { + pool = iod->spraidq->prp_small_pool; + iod->npages = 0; + } else { + pool = hdev->prp_page_pool; + iod->npages = 1; + } + + prp_list = dma_pool_alloc(pool, GFP_ATOMIC, &prp_dma); + if (!prp_list) { + dev_err_ratelimited(hdev->dev, "Allocate first prp_list memory failed\n"); + iod->first_dma = dma_addr; + iod->npages = -1; + return -ENOMEM; + } + list[0] = prp_list; + iod->first_dma = prp_dma; + i = 0; + for (;;) { + if (i == page_size / PRP_ENTRY_SIZE) { + old_prp_list = prp_list; + + prp_list = dma_pool_alloc(pool, GFP_ATOMIC, &prp_dma); + if (!prp_list) { + dev_err_ratelimited(hdev->dev, "Allocate %dth prp_list memory failed\n", + iod->npages + 1); + return -ENOMEM; + } + list[iod->npages++] = prp_list; + prp_list[0] = old_prp_list[i - 1]; + old_prp_list[i - 1] = cpu_to_le64(prp_dma); + i = 1; + } + prp_list[i++] = cpu_to_le64(dma_addr); + dma_len -= page_size; + dma_addr += page_size; + length -= page_size; + if (length <= 0) + break; + if (dma_len > 0) + continue; + if (unlikely(dma_len < 0)) + goto bad_sgl; + sg = sg_next(sg); + dma_addr = sg_dma_address(sg); + dma_len = sg_dma_len(sg); + } + + return 0; + +bad_sgl: + dev_err(hdev->dev, "Setup prps, invalid SGL for payload: %d nents: %d\n", + iod->length, iod->nsge); + return -EIO; +} + +#define SGES_PER_PAGE (PAGE_SIZE / sizeof(struct spraid_sgl_desc)) + +static void spraid_submit_cmd(struct spraid_queue *spraidq, const void *cmd) +{ + u32 sqes = SQE_SIZE(spraidq->qid); + unsigned long flags; + struct spraid_admin_common_command *acd = (struct spraid_admin_common_command *)cmd; + + spin_lock_irqsave(&spraidq->sq_lock, flags); + memcpy((spraidq->sq_cmds + sqes * spraidq->sq_tail), cmd, sqes); + if (++spraidq->sq_tail == spraidq->q_depth) + spraidq->sq_tail = 0; + + writel(spraidq->sq_tail, spraidq->q_db); + spin_unlock_irqrestore(&spraidq->sq_lock, flags); + + pr_debug("cid[%d], qid[%d], opcode[0x%x], flags[0x%x], hdid[%u]\n", + acd->command_id, spraidq->qid, acd->opcode, acd->flags, le32_to_cpu(acd->hdid)); +} + +static u32 spraid_mod64(u64 dividend, u32 divisor) +{ + u64 d; + u32 remainder; + + if (!divisor) + pr_err("DIVISOR is zero, in div fn\n"); + + d = dividend; + remainder = do_div(d, divisor); + return remainder; +} + +static inline bool spraid_is_rw_scmd(struct scsi_cmnd *scmd) +{ + switch (scmd->cmnd[0]) { + case READ_6: + case READ_10: + case READ_12: + case READ_16: + case READ_32: + case WRITE_6: + case WRITE_10: + case WRITE_12: + case WRITE_16: + case WRITE_32: + return true; + default: + return false; + } +} + +static bool spraid_is_prp(struct spraid_dev *hdev, struct scsi_cmnd *scmd, u32 nsge) +{ + struct scatterlist *sg = scsi_sglist(scmd); + u32 page_size = hdev->page_size; + bool is_prp = true; + int i = 0; + + scsi_for_each_sg(scmd, sg, nsge, i) { + if (i != 0 && i != nsge - 1) { + if (spraid_mod64(sg_dma_len(sg), page_size) || + spraid_mod64(sg_dma_address(sg), page_size)) { + is_prp = false; + break; + } + } + + if (nsge > 1 && i == 0) { + if ((spraid_mod64((sg_dma_address(sg) + sg_dma_len(sg)), page_size))) { + is_prp = false; + break; + } + } + + if (nsge > 1 && i == (nsge - 1)) { + if (spraid_mod64(sg_dma_address(sg), page_size)) { + is_prp = false; + break; + } + } + } + + return is_prp; +} + +enum { + SPRAID_SGL_FMT_DATA_DESC = 0x00, + SPRAID_SGL_FMT_SEG_DESC = 0x02, + SPRAID_SGL_FMT_LAST_SEG_DESC = 0x03, + SPRAID_KEY_SGL_FMT_DATA_DESC = 0x04, + SPRAID_TRANSPORT_SGL_DATA_DESC = 0x05 +}; + +static void spraid_sgl_set_data(struct spraid_sgl_desc *sge, struct scatterlist *sg) +{ + sge->addr = cpu_to_le64(sg_dma_address(sg)); + sge->length = cpu_to_le32(sg_dma_len(sg)); + sge->type = SPRAID_SGL_FMT_DATA_DESC << 4; +} + +static void spraid_sgl_set_seg(struct spraid_sgl_desc *sge, dma_addr_t dma_addr, int entries) +{ + sge->addr = cpu_to_le64(dma_addr); + if (entries <= SGES_PER_PAGE) { + sge->length = cpu_to_le32(entries * sizeof(*sge)); + sge->type = SPRAID_SGL_FMT_LAST_SEG_DESC << 4; + } else { + sge->length = cpu_to_le32(PAGE_SIZE); + sge->type = SPRAID_SGL_FMT_SEG_DESC << 4; + } +} + +static int spraid_setup_ioq_cmd_sgl(struct spraid_dev *hdev, + struct scsi_cmnd *scmd, struct spraid_ioq_command *ioq_cmd, + struct spraid_iod *iod) +{ + struct spraid_sgl_desc *sg_list, *link, *old_sg_list; + struct scatterlist *sg = scsi_sglist(scmd); + void **list = spraid_iod_list(iod); + struct dma_pool *pool; + int nsge = iod->nsge; + dma_addr_t sgl_dma; + int i = 0; + + ioq_cmd->common.flags |= SPRAID_CMD_FLAG_SGL_METABUF; + + if (nsge == 1) { + spraid_sgl_set_data(&ioq_cmd->common.dptr.sgl, sg); + return 0; + } + + if (nsge <= (SMALL_POOL_SIZE / sizeof(struct spraid_sgl_desc))) { + pool = iod->spraidq->prp_small_pool; + iod->npages = 0; + } else { + pool = hdev->prp_page_pool; + iod->npages = 1; + } + + sg_list = dma_pool_alloc(pool, GFP_ATOMIC, &sgl_dma); + if (!sg_list) { + dev_err_ratelimited(hdev->dev, "Allocate first sgl_list failed\n"); + iod->npages = -1; + return -ENOMEM; + } + + list[0] = sg_list; + iod->first_dma = sgl_dma; + spraid_sgl_set_seg(&ioq_cmd->common.dptr.sgl, sgl_dma, nsge); + do { + if (i == SGES_PER_PAGE) { + old_sg_list = sg_list; + link = &old_sg_list[SGES_PER_PAGE - 1]; + + sg_list = dma_pool_alloc(pool, GFP_ATOMIC, &sgl_dma); + if (!sg_list) { + dev_err_ratelimited(hdev->dev, "Allocate %dth sgl_list failed\n", + iod->npages + 1); + return -ENOMEM; + } + list[iod->npages++] = sg_list; + + i = 0; + memcpy(&sg_list[i++], link, sizeof(*link)); + spraid_sgl_set_seg(link, sgl_dma, nsge); + } + + spraid_sgl_set_data(&sg_list[i++], sg); + sg = sg_next(sg); + } while (--nsge > 0); + + return 0; +} + +#define SPRAID_RW_FUA BIT(14) + +static void spraid_setup_rw_cmd(struct spraid_dev *hdev, + struct spraid_rw_command *rw, + struct scsi_cmnd *scmd) +{ + u32 start_lba_lo, start_lba_hi; + u32 datalength = 0; + u16 control = 0; + + start_lba_lo = 0; + start_lba_hi = 0; + + if (scmd->sc_data_direction == DMA_TO_DEVICE) { + rw->opcode = SPRAID_CMD_WRITE; + } else if (scmd->sc_data_direction == DMA_FROM_DEVICE) { + rw->opcode = SPRAID_CMD_READ; + } else { + dev_err(hdev->dev, "Invalid IO for unsupported data direction: %d\n", + scmd->sc_data_direction); + WARN_ON(1); + } + + /* 6-byte READ(0x08) or WRITE(0x0A) cdb */ + if (scmd->cmd_len == 6) { + datalength = (u32)(scmd->cmnd[4] == 0 ? + IO_6_DEFAULT_TX_LEN : scmd->cmnd[4]); + start_lba_lo = (u32)get_unaligned_be24(&scmd->cmnd[1]); + + start_lba_lo &= 0x1FFFFF; + } + + /* 10-byte READ(0x28) or WRITE(0x2A) cdb */ + else if (scmd->cmd_len == 10) { + datalength = (u32)get_unaligned_be16(&scmd->cmnd[7]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[2]); + + if (scmd->cmnd[1] & FUA_MASK) + control |= SPRAID_RW_FUA; + } + + /* 12-byte READ(0xA8) or WRITE(0xAA) cdb */ + else if (scmd->cmd_len == 12) { + datalength = get_unaligned_be32(&scmd->cmnd[6]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[2]); + + if (scmd->cmnd[1] & FUA_MASK) + control |= SPRAID_RW_FUA; + } + /* 16-byte READ(0x88) or WRITE(0x8A) cdb */ + else if (scmd->cmd_len == 16) { + datalength = get_unaligned_be32(&scmd->cmnd[10]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[6]); + start_lba_hi = get_unaligned_be32(&scmd->cmnd[2]); + + if (scmd->cmnd[1] & FUA_MASK) + control |= SPRAID_RW_FUA; + } + /* 32-byte READ(0x88) or WRITE(0x8A) cdb */ + else if (scmd->cmd_len == 32) { + datalength = get_unaligned_be32(&scmd->cmnd[28]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[16]); + start_lba_hi = get_unaligned_be32(&scmd->cmnd[12]); + + if (scmd->cmnd[10] & FUA_MASK) + control |= SPRAID_RW_FUA; + } + + if (unlikely(datalength > U16_MAX || datalength == 0)) { + dev_err(hdev->dev, "Invalid IO for illegal transfer data length: %u\n", + datalength); + WARN_ON(1); + } + + rw->slba = cpu_to_le64(((u64)start_lba_hi << 32) | start_lba_lo); + /* 0base for nlb */ + rw->nlb = cpu_to_le16((u16)(datalength - 1)); + rw->control = cpu_to_le16(control); +} + +static void spraid_setup_nonio_cmd(struct spraid_dev *hdev, + struct spraid_scsi_nonio *scsi_nonio, struct scsi_cmnd *scmd) +{ + scsi_nonio->buffer_len = cpu_to_le32(scsi_bufflen(scmd)); + + switch (scmd->sc_data_direction) { + case DMA_NONE: + scsi_nonio->opcode = SPRAID_CMD_NONIO_NONE; + break; + case DMA_TO_DEVICE: + scsi_nonio->opcode = SPRAID_CMD_NONIO_TODEV; + break; + case DMA_FROM_DEVICE: + scsi_nonio->opcode = SPRAID_CMD_NONIO_FROMDEV; + break; + default: + dev_err(hdev->dev, "Invalid IO for unsupported data direction: %d\n", + scmd->sc_data_direction); + WARN_ON(1); + } +} + +static void spraid_setup_ioq_cmd(struct spraid_dev *hdev, + struct spraid_ioq_command *ioq_cmd, struct scsi_cmnd *scmd) +{ + memcpy(ioq_cmd->common.cdb, scmd->cmnd, scmd->cmd_len); + ioq_cmd->common.cdb_len = scmd->cmd_len; + + if (spraid_is_rw_scmd(scmd)) + spraid_setup_rw_cmd(hdev, &ioq_cmd->rw, scmd); + else + spraid_setup_nonio_cmd(hdev, &ioq_cmd->scsi_nonio, scmd); +} + +static int spraid_init_iod(struct spraid_dev *hdev, + struct spraid_iod *iod, struct spraid_ioq_command *ioq_cmd, + struct scsi_cmnd *scmd) +{ + if (unlikely(!iod->sense)) { + dev_err(hdev->dev, "Allocate sense data buffer failed\n"); + return -ENOMEM; + } + ioq_cmd->common.sense_addr = cpu_to_le64(iod->sense_dma); + ioq_cmd->common.sense_len = cpu_to_le16(SCSI_SENSE_BUFFERSIZE); + + iod->nsge = 0; + iod->npages = -1; + iod->use_sgl = 0; + iod->sg_drv_mgmt = false; + WRITE_ONCE(iod->state, SPRAID_CMD_IDLE); + + return 0; +} + +static void spraid_free_iod_res(struct spraid_dev *hdev, struct spraid_iod *iod) +{ + const int last_prp = hdev->page_size / sizeof(__le64) - 1; + dma_addr_t dma_addr, next_dma_addr; + struct spraid_sgl_desc *sg_list; + __le64 *prp_list; + void *addr; + int i; + + dma_addr = iod->first_dma; + if (iod->npages == 0) + dma_pool_free(iod->spraidq->prp_small_pool, spraid_iod_list(iod)[0], dma_addr); + + for (i = 0; i < iod->npages; i++) { + addr = spraid_iod_list(iod)[i]; + + if (iod->use_sgl) { + sg_list = addr; + next_dma_addr = + le64_to_cpu((sg_list[SGES_PER_PAGE - 1]).addr); + } else { + prp_list = addr; + next_dma_addr = le64_to_cpu(prp_list[last_prp]); + } + + dma_pool_free(hdev->prp_page_pool, addr, dma_addr); + dma_addr = next_dma_addr; + } + + if (iod->sg_drv_mgmt && iod->sg != iod->inline_sg) { + iod->sg_drv_mgmt = false; + mempool_free(iod->sg, hdev->iod_mempool); + } + + iod->sense = NULL; + iod->npages = -1; +} + +static int spraid_io_map_data(struct spraid_dev *hdev, struct spraid_iod *iod, + struct scsi_cmnd *scmd, struct spraid_ioq_command *ioq_cmd) +{ + int ret; + + iod->nsge = scsi_dma_map(scmd); + + /* No data to DMA, it may be scsi no-rw command */ + if (unlikely(iod->nsge == 0)) + return 0; + + iod->length = scsi_bufflen(scmd); + iod->sg = scsi_sglist(scmd); + iod->use_sgl = !spraid_is_prp(hdev, scmd, iod->nsge); + + if (iod->use_sgl) { + ret = spraid_setup_ioq_cmd_sgl(hdev, scmd, ioq_cmd, iod); + } else { + ret = spraid_setup_prps(hdev, iod); + ioq_cmd->common.dptr.prp1 = + cpu_to_le64(sg_dma_address(iod->sg)); + ioq_cmd->common.dptr.prp2 = cpu_to_le64(iod->first_dma); + } + + if (ret) + scsi_dma_unmap(scmd); + + return ret; +} + +static void spraid_map_status(struct spraid_iod *iod, struct scsi_cmnd *scmd, + struct spraid_completion *cqe) +{ + scsi_set_resid(scmd, 0); + + switch ((le16_to_cpu(cqe->status) >> 1) & 0x7f) { + case FW_STAT_OK: + set_host_byte(scmd, DID_OK); + break; + case FW_STAT_NEED_CHECK: + set_host_byte(scmd, DID_OK); + scmd->result |= le16_to_cpu(cqe->status) >> 8; + if (scmd->result & SAM_STAT_CHECK_CONDITION) { + memset(scmd->sense_buffer, 0, SCSI_SENSE_BUFFERSIZE); + memcpy(scmd->sense_buffer, iod->sense, SCSI_SENSE_BUFFERSIZE); + scmd->result = (scmd->result & 0x00ffffff) | (DRIVER_SENSE << 24); + } + break; + case FW_STAT_ABORTED: + set_host_byte(scmd, DID_ABORT); + break; + case FW_STAT_NEED_RETRY: + set_host_byte(scmd, DID_REQUEUE); + break; + default: + set_host_byte(scmd, DID_BAD_TARGET); + break; + } +} + +static inline void spraid_get_tag_from_scmd(struct scsi_cmnd *scmd, u16 *qid, u16 *cid) +{ + u32 tag = blk_mq_unique_tag(scsi_cmd_to_rq(scmd)); + + *qid = blk_mq_unique_tag_to_hwq(tag) + 1; + *cid = blk_mq_unique_tag_to_tag(tag); +} + +static int spraid_queue_command(struct Scsi_Host *shost, struct scsi_cmnd *scmd) +{ + struct spraid_iod *iod = scsi_cmd_priv(scmd); + struct spraid_dev *hdev = shost_priv(shost); + struct scsi_device *sdev = scmd->device; + struct spraid_sdev_hostdata *hostdata; + struct spraid_ioq_command ioq_cmd; + struct spraid_queue *ioq; + unsigned long elapsed; + u16 hwq, cid; + int ret; + + if (unlikely(!scmd)) { + dev_err(hdev->dev, "err, scmd is null, return 0\n"); + return 0; + } + + if (unlikely(hdev->state != SPRAID_LIVE)) { + set_host_byte(scmd, DID_NO_CONNECT); + scsi_done(scmd); + dev_err(hdev->dev, "[%s] err, hdev state is not live\n", __func__); + return 0; + } + + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + hostdata = sdev->hostdata; + ioq = &hdev->queues[hwq]; + memset(&ioq_cmd, 0, sizeof(ioq_cmd)); + ioq_cmd.rw.hdid = cpu_to_le32(hostdata->hdid); + ioq_cmd.rw.command_id = cid; + + spraid_setup_ioq_cmd(hdev, &ioq_cmd, scmd); + + ret = cid * SCSI_SENSE_BUFFERSIZE; + iod->sense = ioq->sense + ret; + iod->sense_dma = ioq->sense_dma_addr + ret; + + ret = spraid_init_iod(hdev, iod, &ioq_cmd, scmd); + if (unlikely(ret)) + return SCSI_MLQUEUE_HOST_BUSY; + + iod->spraidq = ioq; + ret = spraid_io_map_data(hdev, iod, scmd, &ioq_cmd); + if (unlikely(ret)) { + dev_err(hdev->dev, "spraid_io_map_data Err.\n"); + set_host_byte(scmd, DID_ERROR); + scsi_done(scmd); + ret = 0; + goto deinit_iod; + } + + WRITE_ONCE(iod->state, SPRAID_CMD_IN_FLIGHT); + spraid_submit_cmd(ioq, &ioq_cmd); + elapsed = jiffies - scmd->jiffies_at_alloc; + pr_debug("cid[%d], qid[%d] submit IO cost %3ld.%3ld seconds\n", + cid, hwq, elapsed / HZ, elapsed % HZ); + return 0; + +deinit_iod: + spraid_free_iod_res(hdev, iod); + return ret; +} + +static int spraid_match_dev(struct spraid_dev *hdev, u16 idx, struct scsi_device *sdev) +{ + if (SPRAID_DEV_INFO_FLAG_VALID(hdev->devices[idx].flag)) { + if (sdev->channel == hdev->devices[idx].channel && + sdev->id == le16_to_cpu(hdev->devices[idx].target) && + sdev->lun < hdev->devices[idx].lun) { + dev_info(hdev->dev, "Match device success, channel:target:lun[%d:%d:%d]\n", + hdev->devices[idx].channel, + hdev->devices[idx].target, + hdev->devices[idx].lun); + return 1; + } + } + + return 0; +} + +static int spraid_slave_alloc(struct scsi_device *sdev) +{ + struct spraid_sdev_hostdata *hostdata; + struct spraid_dev *hdev; + u16 idx; + + hdev = shost_priv(sdev->host); + hostdata = kzalloc(sizeof(*hostdata), GFP_KERNEL); + if (!hostdata) { + dev_err(hdev->dev, "Alloc scsi host data memory failed\n"); + return -ENOMEM; + } + + down_read(&hdev->devices_rwsem); + for (idx = 0; idx < le32_to_cpu(hdev->ctrl_info->nd); idx++) { + if (spraid_match_dev(hdev, idx, sdev)) + goto scan_host; + } + up_read(&hdev->devices_rwsem); + + kfree(hostdata); + return -ENXIO; + +scan_host: + hostdata->hdid = le32_to_cpu(hdev->devices[idx].hdid); + sdev->hostdata = hostdata; + up_read(&hdev->devices_rwsem); + return 0; +} + +static void spraid_slave_destroy(struct scsi_device *sdev) +{ + kfree(sdev->hostdata); + sdev->hostdata = NULL; +} + +static int spraid_slave_configure(struct scsi_device *sdev) +{ + u16 idx; + unsigned int timeout = SCMD_TMOUT_RAWDISK * HZ; + struct spraid_dev *hdev = shost_priv(sdev->host); + struct spraid_sdev_hostdata *hostdata = sdev->hostdata; + u32 max_sec = sdev->host->max_sectors; + + if (hostdata) { + idx = hostdata->hdid - 1; + if (sdev->channel == hdev->devices[idx].channel && + sdev->id == le16_to_cpu(hdev->devices[idx].target) && + sdev->lun < hdev->devices[idx].lun) { + if (SPRAID_DEV_INFO_ATTR_VD(hdev->devices[idx].attr)) + timeout = SCMD_TMOUT_VD * HZ; + else if (SPRAID_DEV_INFO_ATTR_RAWDISK(hdev->devices[idx].attr)) + timeout = SCMD_TMOUT_RAWDISK * HZ; + max_sec = le16_to_cpu(hdev->devices[idx].max_io_kb) << 1; + } else { + dev_err(hdev->dev, "[%s] err, sdev->channel:id:lun[%d:%d:%lld];" + "devices[%d], channel:target:lun[%d:%d:%d]\n", + __func__, sdev->channel, sdev->id, sdev->lun, + idx, hdev->devices[idx].channel, + hdev->devices[idx].target, + hdev->devices[idx].lun); + } + } else { + dev_err(hdev->dev, "[%s] err, sdev->hostdata is null\n", __func__); + } + + blk_queue_rq_timeout(sdev->request_queue, timeout); + sdev->eh_timeout = timeout; + + if ((max_sec == 0) || (max_sec > sdev->host->max_sectors)) + max_sec = sdev->host->max_sectors; + blk_queue_max_hw_sectors(sdev->request_queue, max_sec); + + dev_info(hdev->dev, "[%s] sdev->channel:id:lun[%d:%d:%lld], scmd_timeout[%d]s, maxsec[%d]\n", + __func__, sdev->channel, sdev->id, sdev->lun, timeout / HZ, max_sec); + + return 0; +} + +static void spraid_shost_init(struct spraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + u8 domain, bus; + u32 dev_func; + + domain = pci_domain_nr(pdev->bus); + bus = pdev->bus->number; + dev_func = pdev->devfn; + + hdev->shost->nr_hw_queues = hdev->online_queues - 1; + hdev->shost->can_queue = (hdev->ioq_depth - SPRAID_PTCMDS_PERQ); + + hdev->shost->sg_tablesize = le16_to_cpu(hdev->ctrl_info->max_num_sge); + /* 512B per sector */ + hdev->shost->max_sectors = (1U << ((hdev->ctrl_info->mdts) * 1U) << 12) / 512; + hdev->shost->cmd_per_lun = MAX_CMD_PER_DEV; + hdev->shost->max_channel = le16_to_cpu(hdev->ctrl_info->max_channel) - 1; + hdev->shost->max_id = le32_to_cpu(hdev->ctrl_info->max_tgt_id); + hdev->shost->max_lun = le16_to_cpu(hdev->ctrl_info->max_lun); + + hdev->shost->this_id = -1; + hdev->shost->unique_id = (domain << 16) | (bus << 8) | dev_func; + hdev->shost->max_cmd_len = MAX_CDB_LEN; + hdev->shost->hostt->cmd_size = max(spraid_cmd_size(hdev, false, true), + spraid_cmd_size(hdev, false, false)); +} + +static inline void spraid_host_deinit(struct spraid_dev *hdev) +{ + ida_free(&spraid_instance_ida, hdev->instance); +} + +static int spraid_alloc_queue(struct spraid_dev *hdev, u16 qid, u16 depth) +{ + struct spraid_queue *spraidq = &hdev->queues[qid]; + int ret = 0; + + if (hdev->queue_count > qid) { + dev_info(hdev->dev, "[%s] warn: queue[%d] is exist\n", __func__, qid); + return 0; + } + + spraidq->cqes = dma_alloc_coherent(hdev->dev, CQ_SIZE(depth), + &spraidq->cq_dma_addr, GFP_KERNEL | __GFP_ZERO); + if (!spraidq->cqes) + return -ENOMEM; + + spraidq->sq_cmds = dma_alloc_coherent(hdev->dev, SQ_SIZE(qid, depth), + &spraidq->sq_dma_addr, GFP_KERNEL); + if (!spraidq->sq_cmds) { + ret = -ENOMEM; + goto free_cqes; + } + + spin_lock_init(&spraidq->sq_lock); + spin_lock_init(&spraidq->cq_lock); + spraidq->hdev = hdev; + spraidq->q_depth = depth; + spraidq->qid = qid; + spraidq->cq_vector = -1; + hdev->queue_count++; + + /* alloc sense buffer */ + spraidq->sense = dma_alloc_coherent(hdev->dev, SENSE_SIZE(depth), + &spraidq->sense_dma_addr, GFP_KERNEL | __GFP_ZERO); + if (!spraidq->sense) { + ret = -ENOMEM; + goto free_sq_cmds; + } + + return 0; + +free_sq_cmds: + dma_free_coherent(hdev->dev, SQ_SIZE(qid, depth), (void *)spraidq->sq_cmds, + spraidq->sq_dma_addr); +free_cqes: + dma_free_coherent(hdev->dev, CQ_SIZE(depth), (void *)spraidq->cqes, + spraidq->cq_dma_addr); + return ret; +} + +static int spraid_wait_ready(struct spraid_dev *hdev, u64 cap, bool enabled) +{ + unsigned long timeout = + ((SPRAID_CAP_TIMEOUT(cap) + 1) * SPRAID_CAP_TIMEOUT_UNIT_MS) + jiffies; + u32 bit = enabled ? SPRAID_CSTS_RDY : 0; + + while ((readl(hdev->bar + SPRAID_REG_CSTS) & SPRAID_CSTS_RDY) != bit) { + usleep_range(1000, 2000); + if (fatal_signal_pending(current)) + return -EINTR; + + if (time_after(jiffies, timeout)) { + dev_err(hdev->dev, "Device not ready; aborting %s\n", + enabled ? "initialisation" : "reset"); + return -ENODEV; + } + } + return 0; +} + +static int spraid_shutdown_ctrl(struct spraid_dev *hdev) +{ + unsigned long timeout = hdev->ctrl_info->rtd3e + jiffies; + + hdev->ctrl_config &= ~SPRAID_CC_SHN_MASK; + hdev->ctrl_config |= SPRAID_CC_SHN_NORMAL; + writel(hdev->ctrl_config, hdev->bar + SPRAID_REG_CC); + + while ((readl(hdev->bar + SPRAID_REG_CSTS) & SPRAID_CSTS_SHST_MASK) != + SPRAID_CSTS_SHST_CMPLT) { + msleep(100); + if (fatal_signal_pending(current)) + return -EINTR; + if (time_after(jiffies, timeout)) { + dev_err(hdev->dev, "Device shutdown incomplete; abort shutdown\n"); + return -ENODEV; + } + } + return 0; +} + +static int spraid_disable_ctrl(struct spraid_dev *hdev) +{ + hdev->ctrl_config &= ~SPRAID_CC_SHN_MASK; + hdev->ctrl_config &= ~SPRAID_CC_ENABLE; + writel(hdev->ctrl_config, hdev->bar + SPRAID_REG_CC); + + return spraid_wait_ready(hdev, hdev->cap, false); +} + +static int spraid_enable_ctrl(struct spraid_dev *hdev) +{ + u64 cap = hdev->cap; + u32 dev_page_min = SPRAID_CAP_MPSMIN(cap) + 12; + u32 page_shift = PAGE_SHIFT; + + if (page_shift < dev_page_min) { + dev_err(hdev->dev, "Minimum device page size[%u], too large for host[%u]\n", + 1U << dev_page_min, 1U << page_shift); + return -ENODEV; + } + + page_shift = min_t(unsigned int, SPRAID_CAP_MPSMAX(cap) + 12, PAGE_SHIFT); + hdev->page_size = 1U << page_shift; + + hdev->ctrl_config = SPRAID_CC_CSS_NVM; + hdev->ctrl_config |= (page_shift - 12) << SPRAID_CC_MPS_SHIFT; + hdev->ctrl_config |= SPRAID_CC_AMS_RR | SPRAID_CC_SHN_NONE; + hdev->ctrl_config |= SPRAID_CC_IOSQES | SPRAID_CC_IOCQES; + hdev->ctrl_config |= SPRAID_CC_ENABLE; + writel(hdev->ctrl_config, hdev->bar + SPRAID_REG_CC); + + return spraid_wait_ready(hdev, cap, true); +} + +static void spraid_init_queue(struct spraid_queue *spraidq, u16 qid) +{ + struct spraid_dev *hdev = spraidq->hdev; + + memset((void *)spraidq->cqes, 0, CQ_SIZE(spraidq->q_depth)); + + spraidq->sq_tail = 0; + spraidq->cq_head = 0; + spraidq->cq_phase = 1; + spraidq->q_db = &hdev->dbs[qid * 2 * hdev->db_stride]; + spraidq->prp_small_pool = hdev->prp_small_pool[qid % small_pool_num]; + hdev->online_queues++; +} + +static inline bool spraid_cqe_pending(struct spraid_queue *spraidq) +{ + return (le16_to_cpu(spraidq->cqes[spraidq->cq_head].status) & 1) == + spraidq->cq_phase; +} + +static void spraid_complete_ioq_cmnd(struct spraid_queue *ioq, struct spraid_completion *cqe) +{ + struct spraid_dev *hdev = ioq->hdev; + struct blk_mq_tags *tags; + struct scsi_cmnd *scmd; + struct spraid_iod *iod; + struct request *req; + unsigned long elapsed; + + tags = hdev->shost->tag_set.tags[ioq->qid - 1]; + req = blk_mq_tag_to_rq(tags, cqe->cmd_id); + if (unlikely(!req || !blk_mq_request_started(req))) { + dev_warn(hdev->dev, "Invalid id %d completed on queue %d\n", + cqe->cmd_id, ioq->qid); + return; + } + + scmd = blk_mq_rq_to_pdu(req); + iod = scsi_cmd_priv(scmd); + + elapsed = jiffies - scmd->jiffies_at_alloc; + pr_debug("cid[%d], qid[%d] finish IO cost %3ld.%3ld seconds\n", + cqe->cmd_id, ioq->qid, elapsed / HZ, elapsed % HZ); + + if (cmpxchg(&iod->state, SPRAID_CMD_IN_FLIGHT, SPRAID_CMD_COMPLETE) != + SPRAID_CMD_IN_FLIGHT) { + dev_warn(hdev->dev, "cid[%d], qid[%d] enters abnormal handler, cost %3ld.%3ld seconds\n", + cqe->cmd_id, ioq->qid, elapsed / HZ, elapsed % HZ); + WRITE_ONCE(iod->state, SPRAID_CMD_TMO_COMPLETE); + + if (iod->nsge) { + iod->nsge = 0; + scsi_dma_unmap(scmd); + } + spraid_free_iod_res(hdev, iod); + + return; + } + + spraid_map_status(iod, scmd, cqe); + if (iod->nsge) { + iod->nsge = 0; + scsi_dma_unmap(scmd); + } + spraid_free_iod_res(hdev, iod); + scsi_done(scmd); +} + + +static void spraid_complete_adminq_cmnd(struct spraid_queue *adminq, struct spraid_completion *cqe) +{ + struct spraid_dev *hdev = adminq->hdev; + struct spraid_cmd *adm_cmd; + + adm_cmd = hdev->adm_cmds + cqe->cmd_id; + if (unlikely(adm_cmd->state == SPRAID_CMD_IDLE)) { + dev_warn(adminq->hdev->dev, "Invalid id %d completed on queue %d\n", + cqe->cmd_id, le16_to_cpu(cqe->sq_id)); + return; + } + + adm_cmd->status = le16_to_cpu(cqe->status) >> 1; + adm_cmd->result0 = le32_to_cpu(cqe->result); + adm_cmd->result1 = le32_to_cpu(cqe->result1); + + complete(&adm_cmd->cmd_done); +} + +static void spraid_send_aen(struct spraid_dev *hdev, u16 cid); + +static void spraid_complete_aen(struct spraid_queue *spraidq, struct spraid_completion *cqe) +{ + struct spraid_dev *hdev = spraidq->hdev; + u32 result = le32_to_cpu(cqe->result); + + dev_info(hdev->dev, "rcv aen, status[%x], result[%x]\n", + le16_to_cpu(cqe->status) >> 1, result); + + spraid_send_aen(hdev, cqe->cmd_id); + + if ((le16_to_cpu(cqe->status) >> 1) != SPRAID_SC_SUCCESS) + return; + switch (result & 0x7) { + case SPRAID_AEN_NOTICE: + spraid_handle_aen_notice(hdev, result); + break; + case SPRAID_AEN_VS: + spraid_handle_aen_vs(hdev, result, le32_to_cpu(cqe->result1)); + break; + default: + dev_warn(hdev->dev, "Unsupported async event type: %u\n", + result & 0x7); + break; + } +} + +static void spraid_complete_ioq_sync_cmnd(struct spraid_queue *ioq, struct spraid_completion *cqe) +{ + struct spraid_dev *hdev = ioq->hdev; + struct spraid_cmd *ptcmd; + + ptcmd = hdev->ioq_ptcmds + (ioq->qid - 1) * SPRAID_PTCMDS_PERQ + + cqe->cmd_id - SPRAID_IO_BLK_MQ_DEPTH; + + ptcmd->status = le16_to_cpu(cqe->status) >> 1; + ptcmd->result0 = le32_to_cpu(cqe->result); + ptcmd->result1 = le32_to_cpu(cqe->result1); + + complete(&ptcmd->cmd_done); +} + +static inline void spraid_handle_cqe(struct spraid_queue *spraidq, u16 idx) +{ + struct spraid_completion *cqe = &spraidq->cqes[idx]; + struct spraid_dev *hdev = spraidq->hdev; + + if (unlikely(cqe->cmd_id >= spraidq->q_depth)) { + dev_err(hdev->dev, "Invalid command id[%d] completed on queue %d\n", + cqe->cmd_id, cqe->sq_id); + return; + } + + pr_debug("cid[%d], qid[%d], result[0x%x], sq_id[%d], status[0x%x]\n", + cqe->cmd_id, spraidq->qid, le32_to_cpu(cqe->result), + le16_to_cpu(cqe->sq_id), le16_to_cpu(cqe->status)); + + if (unlikely(spraidq->qid == 0 && cqe->cmd_id >= SPRAID_AQ_BLK_MQ_DEPTH)) { + spraid_complete_aen(spraidq, cqe); + return; + } + + if (unlikely(spraidq->qid && cqe->cmd_id >= SPRAID_IO_BLK_MQ_DEPTH)) { + spraid_complete_ioq_sync_cmnd(spraidq, cqe); + return; + } + + if (spraidq->qid) + spraid_complete_ioq_cmnd(spraidq, cqe); + else + spraid_complete_adminq_cmnd(spraidq, cqe); +} + +static void spraid_complete_cqes(struct spraid_queue *spraidq, u16 start, u16 end) +{ + while (start != end) { + spraid_handle_cqe(spraidq, start); + if (++start == spraidq->q_depth) + start = 0; + } +} + +static inline void spraid_update_cq_head(struct spraid_queue *spraidq) +{ + if (++spraidq->cq_head == spraidq->q_depth) { + spraidq->cq_head = 0; + spraidq->cq_phase = !spraidq->cq_phase; + } +} + +static inline bool spraid_process_cq(struct spraid_queue *spraidq, u16 *start, u16 *end, int tag) +{ + bool found = false; + + *start = spraidq->cq_head; + while (!found && spraid_cqe_pending(spraidq)) { + if (spraidq->cqes[spraidq->cq_head].cmd_id == tag) + found = true; + spraid_update_cq_head(spraidq); + } + *end = spraidq->cq_head; + + if (*start != *end) + writel(spraidq->cq_head, spraidq->q_db + spraidq->hdev->db_stride); + + return found; +} + +static bool spraid_poll_cq(struct spraid_queue *spraidq, int cid) +{ + u16 start, end; + bool found; + + if (!spraid_cqe_pending(spraidq)) + return 0; + + spin_lock_irq(&spraidq->cq_lock); + found = spraid_process_cq(spraidq, &start, &end, cid); + spin_unlock_irq(&spraidq->cq_lock); + + spraid_complete_cqes(spraidq, start, end); + return found; +} + +static irqreturn_t spraid_irq(int irq, void *data) +{ + struct spraid_queue *spraidq = data; + irqreturn_t ret = IRQ_NONE; + u16 start, end; + + spin_lock(&spraidq->cq_lock); + if (spraidq->cq_head != spraidq->last_cq_head) + ret = IRQ_HANDLED; + + spraid_process_cq(spraidq, &start, &end, -1); + spraidq->last_cq_head = spraidq->cq_head; + spin_unlock(&spraidq->cq_lock); + + if (start != end) { + spraid_complete_cqes(spraidq, start, end); + ret = IRQ_HANDLED; + } + return ret; +} + +static int spraid_setup_admin_queue(struct spraid_dev *hdev) +{ + struct spraid_queue *adminq = &hdev->queues[0]; + u32 aqa; + int ret; + + dev_info(hdev->dev, "[%s] start disable ctrl\n", __func__); + + ret = spraid_disable_ctrl(hdev); + if (ret) + return ret; + + ret = spraid_alloc_queue(hdev, 0, SPRAID_AQ_DEPTH); + if (ret) + return ret; + + aqa = adminq->q_depth - 1; + aqa |= aqa << 16; + writel(aqa, hdev->bar + SPRAID_REG_AQA); + lo_hi_writeq(adminq->sq_dma_addr, hdev->bar + SPRAID_REG_ASQ); + lo_hi_writeq(adminq->cq_dma_addr, hdev->bar + SPRAID_REG_ACQ); + + dev_info(hdev->dev, "[%s] start enable ctrl\n", __func__); + + ret = spraid_enable_ctrl(hdev); + if (ret) { + ret = -ENODEV; + goto free_queue; + } + + adminq->cq_vector = 0; + spraid_init_queue(adminq, 0); + ret = pci_request_irq(hdev->pdev, adminq->cq_vector, spraid_irq, NULL, + adminq, "spraid%d_q%d", hdev->instance, adminq->qid); + + if (ret) { + adminq->cq_vector = -1; + hdev->online_queues--; + goto free_queue; + } + + dev_info(hdev->dev, "[%s] success, queuecount:[%d], onlinequeue:[%d]\n", + __func__, hdev->queue_count, hdev->online_queues); + + return 0; + +free_queue: + spraid_free_queue(adminq); + return ret; +} + +static u32 spraid_bar_size(struct spraid_dev *hdev, u32 nr_ioqs) +{ + return (SPRAID_REG_DBS + ((nr_ioqs + 1) * 8 * hdev->db_stride)); +} + +static int spraid_alloc_admin_cmds(struct spraid_dev *hdev) +{ + int i; + + INIT_LIST_HEAD(&hdev->adm_cmd_list); + spin_lock_init(&hdev->adm_cmd_lock); + + hdev->adm_cmds = kcalloc_node(SPRAID_AQ_BLK_MQ_DEPTH, sizeof(struct spraid_cmd), + GFP_KERNEL, hdev->numa_node); + + if (!hdev->adm_cmds) { + dev_err(hdev->dev, "Alloc admin cmds failed\n"); + return -ENOMEM; + } + + for (i = 0; i < SPRAID_AQ_BLK_MQ_DEPTH; i++) { + hdev->adm_cmds[i].qid = 0; + hdev->adm_cmds[i].cid = i; + list_add_tail(&(hdev->adm_cmds[i].list), &hdev->adm_cmd_list); + } + + dev_info(hdev->dev, "Alloc admin cmds success, num[%d]\n", SPRAID_AQ_BLK_MQ_DEPTH); + + return 0; +} + +static void spraid_free_admin_cmds(struct spraid_dev *hdev) +{ + kfree(hdev->adm_cmds); + INIT_LIST_HEAD(&hdev->adm_cmd_list); +} + +static struct spraid_cmd *spraid_get_cmd(struct spraid_dev *hdev, enum spraid_cmd_type type) +{ + struct spraid_cmd *cmd = NULL; + unsigned long flags; + struct list_head *head = &hdev->adm_cmd_list; + spinlock_t *slock = &hdev->adm_cmd_lock; + + if (type == SPRAID_CMD_IOPT) { + head = &hdev->ioq_pt_list; + slock = &hdev->ioq_pt_lock; + } + + spin_lock_irqsave(slock, flags); + if (list_empty(head)) { + spin_unlock_irqrestore(slock, flags); + dev_err(hdev->dev, "err, cmd[%d] list empty\n", type); + return NULL; + } + cmd = list_entry(head->next, struct spraid_cmd, list); + list_del_init(&cmd->list); + spin_unlock_irqrestore(slock, flags); + + WRITE_ONCE(cmd->state, SPRAID_CMD_IN_FLIGHT); + + return cmd; +} + +static void spraid_put_cmd(struct spraid_dev *hdev, struct spraid_cmd *cmd, + enum spraid_cmd_type type) +{ + unsigned long flags; + struct list_head *head = &hdev->adm_cmd_list; + spinlock_t *slock = &hdev->adm_cmd_lock; + + if (type == SPRAID_CMD_IOPT) { + head = &hdev->ioq_pt_list; + slock = &hdev->ioq_pt_lock; + } + + spin_lock_irqsave(slock, flags); + WRITE_ONCE(cmd->state, SPRAID_CMD_IDLE); + list_add_tail(&cmd->list, head); + spin_unlock_irqrestore(slock, flags); +} + + +static int spraid_submit_admin_sync_cmd(struct spraid_dev *hdev, struct spraid_admin_command *cmd, + u32 *result0, u32 *result1, u32 timeout) +{ + struct spraid_cmd *adm_cmd = spraid_get_cmd(hdev, SPRAID_CMD_ADM); + + if (!adm_cmd) { + dev_err(hdev->dev, "err, get admin cmd failed\n"); + return -EFAULT; + } + + timeout = timeout ? timeout : ADMIN_TIMEOUT; + + init_completion(&adm_cmd->cmd_done); + + cmd->common.command_id = adm_cmd->cid; + spraid_submit_cmd(&hdev->queues[0], cmd); + + if (!wait_for_completion_timeout(&adm_cmd->cmd_done, timeout)) { + dev_err(hdev->dev, "[%s] cid[%d] qid[%d] timeout, opcode[0x%x] subopcode[0x%x]\n", + __func__, adm_cmd->cid, adm_cmd->qid, cmd->usr_cmd.opcode, + cmd->usr_cmd.info_0.subopcode); + WRITE_ONCE(adm_cmd->state, SPRAID_CMD_TIMEOUT); + return -EINVAL; + } + + if (result0) + *result0 = adm_cmd->result0; + if (result1) + *result1 = adm_cmd->result1; + + spraid_put_cmd(hdev, adm_cmd, SPRAID_CMD_ADM); + + return adm_cmd->status; +} + +static int spraid_create_cq(struct spraid_dev *hdev, u16 qid, + struct spraid_queue *spraidq, u16 cq_vector) +{ + struct spraid_admin_command admin_cmd; + int flags = SPRAID_QUEUE_PHYS_CONTIG | SPRAID_CQ_IRQ_ENABLED; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.create_cq.opcode = SPRAID_ADMIN_CREATE_CQ; + admin_cmd.create_cq.prp1 = cpu_to_le64(spraidq->cq_dma_addr); + admin_cmd.create_cq.cqid = cpu_to_le16(qid); + admin_cmd.create_cq.qsize = cpu_to_le16(spraidq->q_depth - 1); + admin_cmd.create_cq.cq_flags = cpu_to_le16(flags); + admin_cmd.create_cq.irq_vector = cpu_to_le16(cq_vector); + + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); +} + +static int spraid_create_sq(struct spraid_dev *hdev, u16 qid, + struct spraid_queue *spraidq) +{ + struct spraid_admin_command admin_cmd; + int flags = SPRAID_QUEUE_PHYS_CONTIG; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.create_sq.opcode = SPRAID_ADMIN_CREATE_SQ; + admin_cmd.create_sq.prp1 = cpu_to_le64(spraidq->sq_dma_addr); + admin_cmd.create_sq.sqid = cpu_to_le16(qid); + admin_cmd.create_sq.qsize = cpu_to_le16(spraidq->q_depth - 1); + admin_cmd.create_sq.sq_flags = cpu_to_le16(flags); + admin_cmd.create_sq.cqid = cpu_to_le16(qid); + + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); +} + +static void spraid_free_queue(struct spraid_queue *spraidq) +{ + struct spraid_dev *hdev = spraidq->hdev; + + hdev->queue_count--; + dma_free_coherent(hdev->dev, CQ_SIZE(spraidq->q_depth), + (void *)spraidq->cqes, spraidq->cq_dma_addr); + dma_free_coherent(hdev->dev, SQ_SIZE(spraidq->qid, spraidq->q_depth), + spraidq->sq_cmds, spraidq->sq_dma_addr); + dma_free_coherent(hdev->dev, SENSE_SIZE(spraidq->q_depth), + spraidq->sense, spraidq->sense_dma_addr); +} + +static void spraid_free_admin_queue(struct spraid_dev *hdev) +{ + spraid_free_queue(&hdev->queues[0]); +} + +static void spraid_free_io_queues(struct spraid_dev *hdev) +{ + int i; + + for (i = hdev->queue_count - 1; i >= 1; i--) + spraid_free_queue(&hdev->queues[i]); +} + +static int spraid_delete_queue(struct spraid_dev *hdev, u8 op, u16 id) +{ + struct spraid_admin_command admin_cmd; + int ret; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.delete_queue.opcode = op; + admin_cmd.delete_queue.qid = cpu_to_le16(id); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + + if (ret) + dev_err(hdev->dev, "Delete %s:[%d] failed\n", + (op == SPRAID_ADMIN_DELETE_CQ) ? "cq" : "sq", id); + + return ret; +} + +static int spraid_delete_cq(struct spraid_dev *hdev, u16 cqid) +{ + return spraid_delete_queue(hdev, SPRAID_ADMIN_DELETE_CQ, cqid); +} + +static int spraid_delete_sq(struct spraid_dev *hdev, u16 sqid) +{ + return spraid_delete_queue(hdev, SPRAID_ADMIN_DELETE_SQ, sqid); +} + +static int spraid_create_queue(struct spraid_queue *spraidq, u16 qid) +{ + struct spraid_dev *hdev = spraidq->hdev; + u16 cq_vector; + int ret; + + cq_vector = (hdev->num_vecs == 1) ? 0 : qid; + ret = spraid_create_cq(hdev, qid, spraidq, cq_vector); + if (ret) + return ret; + + ret = spraid_create_sq(hdev, qid, spraidq); + if (ret) + goto delete_cq; + + spraid_init_queue(spraidq, qid); + spraidq->cq_vector = cq_vector; + + ret = pci_request_irq(hdev->pdev, cq_vector, spraid_irq, NULL, + spraidq, "spraid%d_q%d", hdev->instance, qid); + + if (ret) { + dev_err(hdev->dev, "Request queue[%d] irq failed\n", qid); + goto delete_sq; + } + + return 0; + +delete_sq: + spraidq->cq_vector = -1; + hdev->online_queues--; + spraid_delete_sq(hdev, qid); +delete_cq: + spraid_delete_cq(hdev, qid); + + return ret; +} + +static int spraid_create_io_queues(struct spraid_dev *hdev) +{ + u32 i, max; + int ret = 0; + + max = min(hdev->max_qid, hdev->queue_count - 1); + for (i = hdev->online_queues; i <= max; i++) { + ret = spraid_create_queue(&hdev->queues[i], i); + if (ret) { + dev_err(hdev->dev, "Create queue[%d] failed\n", i); + break; + } + } + + dev_info(hdev->dev, "[%s] queue_count[%d], online_queue[%d]", + __func__, hdev->queue_count, hdev->online_queues); + + return ret >= 0 ? 0 : ret; +} + +static int spraid_set_features(struct spraid_dev *hdev, u32 fid, u32 dword11, void *buffer, + size_t buflen, u32 *result) +{ + struct spraid_admin_command admin_cmd; + int ret; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + + if (buffer && buflen) { + data_ptr = dma_alloc_coherent(hdev->dev, buflen, &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memcpy(data_ptr, buffer, buflen); + } + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.features.opcode = SPRAID_ADMIN_SET_FEATURES; + admin_cmd.features.fid = cpu_to_le32(fid); + admin_cmd.features.dword11 = cpu_to_le32(dword11); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, result, NULL, 0); + + if (data_ptr) + dma_free_coherent(hdev->dev, buflen, data_ptr, data_dma); + + return ret; +} + +static int spraid_configure_timestamp(struct spraid_dev *hdev) +{ + __le64 ts; + int ret; + + ts = cpu_to_le64(ktime_to_ms(ktime_get_real())); + ret = spraid_set_features(hdev, SPRAID_FEAT_TIMESTAMP, 0, &ts, sizeof(ts), NULL); + + if (ret) + dev_err(hdev->dev, "set timestamp failed: %d\n", ret); + return ret; +} + +static int spraid_set_queue_cnt(struct spraid_dev *hdev, u32 *cnt) +{ + u32 q_cnt = (*cnt - 1) | ((*cnt - 1) << 16); + u32 nr_ioqs, result; + int status; + + status = spraid_set_features(hdev, SPRAID_FEAT_NUM_QUEUES, q_cnt, NULL, 0, &result); + if (status) { + dev_err(hdev->dev, "Set queue count failed, status: %d\n", + status); + return -EIO; + } + + nr_ioqs = min(result & 0xffff, result >> 16) + 1; + *cnt = min(*cnt, nr_ioqs); + if (*cnt == 0) { + dev_err(hdev->dev, "Illegal queue count: zero\n"); + return -EIO; + } + return 0; +} + +static int spraid_setup_io_queues(struct spraid_dev *hdev) +{ + struct spraid_queue *adminq = &hdev->queues[0]; + struct pci_dev *pdev = hdev->pdev; + u32 nr_ioqs = num_online_cpus(); + u32 i, size; + int ret; + + struct irq_affinity affd = { + .pre_vectors = 1 + }; + + ret = spraid_set_queue_cnt(hdev, &nr_ioqs); + if (ret < 0) + return ret; + + size = spraid_bar_size(hdev, nr_ioqs); + ret = spraid_remap_bar(hdev, size); + if (ret) + return -ENOMEM; + + adminq->q_db = hdev->dbs; + + pci_free_irq(pdev, 0, adminq); + pci_free_irq_vectors(pdev); + + ret = pci_alloc_irq_vectors_affinity(pdev, 1, (nr_ioqs + 1), + PCI_IRQ_ALL_TYPES | PCI_IRQ_AFFINITY, &affd); + if (ret <= 0) + return -EIO; + + hdev->num_vecs = ret; + + hdev->max_qid = max(ret - 1, 1); + + ret = pci_request_irq(pdev, adminq->cq_vector, spraid_irq, NULL, + adminq, "spraid%d_q%d", hdev->instance, adminq->qid); + if (ret) { + dev_err(hdev->dev, "Request admin irq failed\n"); + adminq->cq_vector = -1; + return ret; + } + + for (i = hdev->queue_count; i <= hdev->max_qid; i++) { + ret = spraid_alloc_queue(hdev, i, hdev->ioq_depth); + if (ret) + break; + } + dev_info(hdev->dev, "[%s] max_qid: %d, queue_count: %d, online_queue: %d, ioq_depth: %d\n", + __func__, hdev->max_qid, hdev->queue_count, hdev->online_queues, hdev->ioq_depth); + + return spraid_create_io_queues(hdev); +} + +static void spraid_delete_io_queues(struct spraid_dev *hdev) +{ + u16 queues = hdev->online_queues - 1; + u8 opcode = SPRAID_ADMIN_DELETE_SQ; + u16 i, pass; + + if (!pci_device_is_present(hdev->pdev)) { + dev_err(hdev->dev, "pci_device is not present, skip disable io queues\n"); + return; + } + + if (hdev->online_queues < 2) { + dev_err(hdev->dev, "[%s] err, io queue has been delete\n", __func__); + return; + } + + for (pass = 0; pass < 2; pass++) { + for (i = queues; i > 0; i--) + if (spraid_delete_queue(hdev, opcode, i)) + break; + + opcode = SPRAID_ADMIN_DELETE_CQ; + } +} + +static void spraid_remove_io_queues(struct spraid_dev *hdev) +{ + spraid_delete_io_queues(hdev); + spraid_free_io_queues(hdev); +} + +static void spraid_pci_disable(struct spraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + u32 i; + + for (i = 0; i < hdev->online_queues; i++) + pci_free_irq(pdev, hdev->queues[i].cq_vector, &hdev->queues[i]); + pci_free_irq_vectors(pdev); + if (pci_is_enabled(pdev)) { + pci_disable_pcie_error_reporting(pdev); + pci_disable_device(pdev); + } + hdev->online_queues = 0; +} + +static void spraid_disable_admin_queue(struct spraid_dev *hdev, bool shutdown) +{ + struct spraid_queue *adminq = &hdev->queues[0]; + u16 start, end; + + if (pci_device_is_present(hdev->pdev)) { + if (shutdown) + spraid_shutdown_ctrl(hdev); + else + spraid_disable_ctrl(hdev); + } + + if (hdev->queue_count == 0) { + dev_err(hdev->dev, "[%s] err, admin queue has been delete\n", __func__); + return; + } + + spin_lock_irq(&adminq->cq_lock); + spraid_process_cq(adminq, &start, &end, -1); + spin_unlock_irq(&adminq->cq_lock); + + spraid_complete_cqes(adminq, start, end); + spraid_free_admin_queue(hdev); +} + +static int spraid_create_dma_pools(struct spraid_dev *hdev) +{ + int i; + char poolname[20] = { 0 }; + + hdev->prp_page_pool = dma_pool_create("prp list page", hdev->dev, + PAGE_SIZE, PAGE_SIZE, 0); + + if (!hdev->prp_page_pool) { + dev_err(hdev->dev, "create prp_page_pool failed\n"); + return -ENOMEM; + } + + for (i = 0; i < small_pool_num; i++) { + sprintf(poolname, "prp_list_256_%d", i); + hdev->prp_small_pool[i] = dma_pool_create(poolname, hdev->dev, SMALL_POOL_SIZE, + SMALL_POOL_SIZE, 0); + + if (!hdev->prp_small_pool[i]) { + dev_err(hdev->dev, "create prp_small_pool %d failed\n", i); + goto destroy_prp_small_pool; + } + } + + return 0; + +destroy_prp_small_pool: + while (i > 0) + dma_pool_destroy(hdev->prp_small_pool[--i]); + dma_pool_destroy(hdev->prp_page_pool); + + return -ENOMEM; +} + +static void spraid_destroy_dma_pools(struct spraid_dev *hdev) +{ + int i; + + for (i = 0; i < small_pool_num; i++) + dma_pool_destroy(hdev->prp_small_pool[i]); + dma_pool_destroy(hdev->prp_page_pool); +} + +static int spraid_get_dev_list(struct spraid_dev *hdev, struct spraid_dev_info *devices) +{ + u32 nd = le32_to_cpu(hdev->ctrl_info->nd); + struct spraid_admin_command admin_cmd; + struct spraid_dev_list *list_buf; + dma_addr_t data_dma = 0; + u32 i, idx, hdid, ndev; + int ret = 0; + + list_buf = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &data_dma, GFP_KERNEL); + if (!list_buf) + return -ENOMEM; + + for (idx = 0; idx < nd;) { + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.get_info.opcode = SPRAID_ADMIN_GET_INFO; + admin_cmd.get_info.type = SPRAID_GET_INFO_DEV_LIST; + admin_cmd.get_info.cdw11 = cpu_to_le32(idx); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + + if (ret) { + dev_err(hdev->dev, "Get device list failed, nd: %u, idx: %u, ret: %d\n", + nd, idx, ret); + goto out; + } + ndev = le32_to_cpu(list_buf->dev_num); + + dev_info(hdev->dev, "ndev numbers: %u\n", ndev); + + for (i = 0; i < ndev; i++) { + hdid = le32_to_cpu(list_buf->devices[i].hdid); + dev_info(hdev->dev, "list_buf->devices[%d], hdid: %u target: %d, channel: %d, lun: %d, attr[0x%x]\n", + i, hdid, le16_to_cpu(list_buf->devices[i].target), + list_buf->devices[i].channel, + list_buf->devices[i].lun, + list_buf->devices[i].attr); + if (hdid > nd || hdid == 0) { + dev_err(hdev->dev, "err, hdid[%d] invalid\n", hdid); + continue; + } + memcpy(&devices[hdid - 1], &list_buf->devices[i], + sizeof(struct spraid_dev_info)); + } + idx += ndev; + + if (idx < MAX_DEV_ENTRY_PER_PAGE_4K) + break; + } + +out: + dma_free_coherent(hdev->dev, PAGE_SIZE, list_buf, data_dma); + return ret; +} + +static void spraid_send_aen(struct spraid_dev *hdev, u16 cid) +{ + struct spraid_queue *adminq = &hdev->queues[0]; + struct spraid_admin_command admin_cmd; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.common.opcode = SPRAID_ADMIN_ASYNC_EVENT; + admin_cmd.common.command_id = cid; + + spraid_submit_cmd(adminq, &admin_cmd); + dev_info(hdev->dev, "send aen, cid[%d]\n", cid); +} +static inline void spraid_send_all_aen(struct spraid_dev *hdev) +{ + u16 i; + + for (i = 0; i < hdev->ctrl_info->aerl; i++) + spraid_send_aen(hdev, i + SPRAID_AQ_BLK_MQ_DEPTH); +} + +static int spraid_add_device(struct spraid_dev *hdev, struct spraid_dev_info *device) +{ + struct Scsi_Host *shost = hdev->shost; + struct scsi_device *sdev; + + sdev = scsi_device_lookup(shost, device->channel, le16_to_cpu(device->target), 0); + if (sdev) { + dev_warn(hdev->dev, "Device is already exist, channel: %d, target_id: %d, lun: %d\n", + device->channel, le16_to_cpu(device->target), 0); + scsi_device_put(sdev); + return -EEXIST; + } + scsi_add_device(shost, device->channel, le16_to_cpu(device->target), 0); + return 0; +} + +static int spraid_rescan_device(struct spraid_dev *hdev, struct spraid_dev_info *device) +{ + struct Scsi_Host *shost = hdev->shost; + struct scsi_device *sdev; + + sdev = scsi_device_lookup(shost, device->channel, le16_to_cpu(device->target), 0); + if (!sdev) { + dev_warn(hdev->dev, "Device is not exit, channel: %d, target_id: %d, lun: %d\n", + device->channel, le16_to_cpu(device->target), 0); + return -ENODEV; + } + + scsi_rescan_device(&sdev->sdev_gendev); + scsi_device_put(sdev); + return 0; +} + +static int spraid_remove_device(struct spraid_dev *hdev, struct spraid_dev_info *org_device) +{ + struct Scsi_Host *shost = hdev->shost; + struct scsi_device *sdev; + + sdev = scsi_device_lookup(shost, org_device->channel, le16_to_cpu(org_device->target), 0); + if (!sdev) { + dev_warn(hdev->dev, "Device is not exit, channel: %d, target_id: %d, lun: %d\n", + org_device->channel, le16_to_cpu(org_device->target), 0); + return -ENODEV; + } + + scsi_remove_device(sdev); + scsi_device_put(sdev); + return 0; +} + +static int spraid_dev_list_init(struct spraid_dev *hdev) +{ + u32 nd = le32_to_cpu(hdev->ctrl_info->nd); + int i, ret; + + hdev->devices = kzalloc_node(nd * sizeof(struct spraid_dev_info), + GFP_KERNEL, hdev->numa_node); + if (!hdev->devices) + return -ENOMEM; + + ret = spraid_get_dev_list(hdev, hdev->devices); + if (ret) { + dev_err(hdev->dev, "Ignore failure of getting device list within initialization\n"); + return 0; + } + + for (i = 0; i < nd; i++) { + if (SPRAID_DEV_INFO_FLAG_VALID(hdev->devices[i].flag) && + SPRAID_DEV_INFO_ATTR_BOOT(hdev->devices[i].attr)) { + spraid_add_device(hdev, &hdev->devices[i]); + break; + } + } + return 0; +} + +static void spraid_scan_work(struct work_struct *work) +{ + struct spraid_dev *hdev = + container_of(work, struct spraid_dev, scan_work); + struct spraid_dev_info *devices, *org_devices; + u32 nd = le32_to_cpu(hdev->ctrl_info->nd); + u8 flag, org_flag; + int i, ret; + + devices = kcalloc(nd, sizeof(struct spraid_dev_info), GFP_KERNEL); + if (!devices) + return; + ret = spraid_get_dev_list(hdev, devices); + if (ret) + goto free_list; + org_devices = hdev->devices; + for (i = 0; i < nd; i++) { + org_flag = org_devices[i].flag; + flag = devices[i].flag; + + pr_debug("i: %d, org_flag: 0x%x, flag: 0x%x\n", + i, org_flag, flag); + + if (SPRAID_DEV_INFO_FLAG_VALID(flag)) { + if (!SPRAID_DEV_INFO_FLAG_VALID(org_flag)) { + down_write(&hdev->devices_rwsem); + memcpy(&org_devices[i], &devices[i], + sizeof(struct spraid_dev_info)); + up_write(&hdev->devices_rwsem); + spraid_add_device(hdev, &devices[i]); + } else if (SPRAID_DEV_INFO_FLAG_CHANGE(flag)) { + spraid_rescan_device(hdev, &devices[i]); + } + } else { + if (SPRAID_DEV_INFO_FLAG_VALID(org_flag)) { + down_write(&hdev->devices_rwsem); + org_devices[i].flag &= 0xfe; + up_write(&hdev->devices_rwsem); + spraid_remove_device(hdev, &org_devices[i]); + } + } + } +free_list: + kfree(devices); +} + +static void spraid_timesyn_work(struct work_struct *work) +{ + struct spraid_dev *hdev = + container_of(work, struct spraid_dev, timesyn_work); + + spraid_configure_timestamp(hdev); +} + +static void spraid_queue_scan(struct spraid_dev *hdev) +{ + queue_work(spraid_wq, &hdev->scan_work); +} + +static void spraid_handle_aen_notice(struct spraid_dev *hdev, u32 result) +{ + switch ((result & 0xff00) >> 8) { + case SPRAID_AEN_DEV_CHANGED: + spraid_queue_scan(hdev); + break; + case SPRAID_AEN_HOST_PROBING: + break; + default: + dev_warn(hdev->dev, "async event result %08x\n", result); + } +} + +static void spraid_handle_aen_vs(struct spraid_dev *hdev, u32 result, u32 result1) +{ + switch ((result & 0xff00) >> 8) { + case SPRAID_AEN_TIMESYN: + queue_work(spraid_wq, &hdev->timesyn_work); + break; + default: + dev_warn(hdev->dev, "async event result: 0x%x\n", result); + } +} + +static int spraid_alloc_resources(struct spraid_dev *hdev) +{ + int ret, nqueue; + + ret = ida_alloc(&spraid_instance_ida, GFP_KERNEL); + if (ret < 0) { + dev_err(hdev->dev, "Get instance id failed\n"); + return ret; + } + hdev->instance = ret; + + hdev->ctrl_info = kzalloc_node(sizeof(*hdev->ctrl_info), + GFP_KERNEL, hdev->numa_node); + if (!hdev->ctrl_info) { + ret = -ENOMEM; + goto release_instance; + } + + ret = spraid_create_dma_pools(hdev); + if (ret) + goto free_ctrl_info; + nqueue = num_possible_cpus() + 1; + hdev->queues = kcalloc_node(nqueue, sizeof(struct spraid_queue), + GFP_KERNEL, hdev->numa_node); + if (!hdev->queues) { + ret = -ENOMEM; + goto destroy_dma_pools; + } + + ret = spraid_alloc_admin_cmds(hdev); + if (ret) + goto free_queues; + + dev_info(hdev->dev, "[%s] queues num: %d\n", __func__, nqueue); + + return 0; + +free_queues: + kfree(hdev->queues); +destroy_dma_pools: + spraid_destroy_dma_pools(hdev); +free_ctrl_info: + kfree(hdev->ctrl_info); +release_instance: + ida_free(&spraid_instance_ida, hdev->instance); + return ret; +} + +static void spraid_free_resources(struct spraid_dev *hdev) +{ + spraid_free_admin_cmds(hdev); + kfree(hdev->queues); + spraid_destroy_dma_pools(hdev); + kfree(hdev->ctrl_info); + ida_free(&spraid_instance_ida, hdev->instance); +} + + + +static void spraid_bsg_unmap_data(struct spraid_dev *hdev, struct bsg_job *job) +{ + struct request *rq = blk_mq_rq_from_pdu(job); + struct spraid_iod *iod = job->dd_data; + enum dma_data_direction dma_dir = rq_data_dir(rq) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; + + if (iod->nsge) + dma_unmap_sg(hdev->dev, iod->sg, iod->nsge, dma_dir); + + spraid_free_iod_res(hdev, iod); +} + +static int spraid_bsg_map_data(struct spraid_dev *hdev, struct bsg_job *job, + struct spraid_admin_command *cmd) +{ + struct request *rq = blk_mq_rq_from_pdu(job); + struct spraid_iod *iod = job->dd_data; + enum dma_data_direction dma_dir = rq_data_dir(rq) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; + int ret = 0; + + iod->sg = job->request_payload.sg_list; + iod->nsge = job->request_payload.sg_cnt; + iod->length = job->request_payload.payload_len; + iod->use_sgl = false; + iod->npages = -1; + iod->sg_drv_mgmt = false; + + if (!iod->nsge) + goto out; + + ret = dma_map_sg_attrs(hdev->dev, iod->sg, iod->nsge, dma_dir, DMA_ATTR_NO_WARN); + if (!ret) + goto out; + + ret = spraid_setup_prps(hdev, iod); + if (ret) + goto unmap; + + cmd->common.dptr.prp1 = cpu_to_le64(sg_dma_address(iod->sg)); + cmd->common.dptr.prp2 = cpu_to_le64(iod->first_dma); + + return 0; + +unmap: + dma_unmap_sg(hdev->dev, iod->sg, iod->nsge, dma_dir); +out: + return ret; +} + +static int spraid_get_ctrl_info(struct spraid_dev *hdev, struct spraid_ctrl_info *ctrl_info) +{ + struct spraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + int ret; + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.get_info.opcode = SPRAID_ADMIN_GET_INFO; + admin_cmd.get_info.type = SPRAID_GET_INFO_CTRL; + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + if (!ret) + memcpy(ctrl_info, data_ptr, sizeof(struct spraid_ctrl_info)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, data_dma); + + return ret; +} + +static int spraid_init_ctrl_info(struct spraid_dev *hdev) +{ + int ret; + + hdev->ctrl_info->nd = cpu_to_le32(240); + hdev->ctrl_info->mdts = 8; + hdev->ctrl_info->max_cmds = cpu_to_le16(4096); + hdev->ctrl_info->max_num_sge = cpu_to_le16(128); + hdev->ctrl_info->max_channel = cpu_to_le16(4); + hdev->ctrl_info->max_tgt_id = cpu_to_le32(3239); + hdev->ctrl_info->max_lun = cpu_to_le16(2); + + ret = spraid_get_ctrl_info(hdev, hdev->ctrl_info); + if (ret) + dev_err(hdev->dev, "get controller info failed: %d\n", ret); + + dev_info(hdev->dev, "[%s]nd = %d\n", __func__, hdev->ctrl_info->nd); + dev_info(hdev->dev, "[%s]max_cmd = %d\n", __func__, hdev->ctrl_info->max_cmds); + dev_info(hdev->dev, "[%s]max_channel = %d\n", __func__, hdev->ctrl_info->max_channel); + dev_info(hdev->dev, "[%s]max_tgt_id = %d\n", __func__, hdev->ctrl_info->max_tgt_id); + dev_info(hdev->dev, "[%s]max_lun = %d\n", __func__, hdev->ctrl_info->max_lun); + dev_info(hdev->dev, "[%s]max_num_sge = %d\n", __func__, hdev->ctrl_info->max_num_sge); + dev_info(hdev->dev, "[%s]lun_num_boot = %d\n", __func__, hdev->ctrl_info->lun_num_in_boot); + dev_info(hdev->dev, "[%s]mdts = %d\n", __func__, hdev->ctrl_info->mdts); + dev_info(hdev->dev, "[%s]acl = %d\n", __func__, hdev->ctrl_info->acl); + dev_info(hdev->dev, "[%s]aer1 = %d\n", __func__, hdev->ctrl_info->aerl); + dev_info(hdev->dev, "[%s]card_type = %d\n", __func__, hdev->ctrl_info->card_type); + dev_info(hdev->dev, "[%s]rtd3e = %d\n", __func__, hdev->ctrl_info->rtd3e); + dev_info(hdev->dev, "[%s]sn = %s\n", __func__, hdev->ctrl_info->sn); + dev_info(hdev->dev, "[%s]fr = %s\n", __func__, hdev->ctrl_info->fr); + + return 0; +} + +#define SPRAID_MAX_ADMIN_PAYLOAD_SIZE BIT(16) +static int spraid_alloc_iod_ext_mem_pool(struct spraid_dev *hdev) +{ + u16 max_sge = le16_to_cpu(hdev->ctrl_info->max_num_sge); + size_t alloc_size; + + alloc_size = spraid_iod_ext_size(hdev, SPRAID_MAX_ADMIN_PAYLOAD_SIZE, + max_sge, true, false); + if (alloc_size > PAGE_SIZE) + dev_warn(hdev->dev, "It is unreasonable for sg allocation more than one page\n"); + hdev->iod_mempool = mempool_create_node(1, mempool_kmalloc, mempool_kfree, + (void *)alloc_size, GFP_KERNEL, hdev->numa_node); + if (!hdev->iod_mempool) { + dev_err(hdev->dev, "Create iod extension memory pool failed\n"); + return -ENOMEM; + } + + return 0; +} + +static void spraid_free_iod_ext_mem_pool(struct spraid_dev *hdev) +{ + mempool_destroy(hdev->iod_mempool); +} + +static int spraid_user_admin_cmd(struct spraid_dev *hdev, struct bsg_job *job) +{ + struct spraid_bsg_request *bsg_req = (struct spraid_bsg_request *)(job->request); + struct spraid_passthru_common_cmd *cmd = &(bsg_req->admcmd); + struct spraid_admin_command admin_cmd; + u32 timeout = msecs_to_jiffies(cmd->timeout_ms); + u32 result[2] = {0}; + int status; + + if (hdev->state >= SPRAID_RESETTING) { + dev_err(hdev->dev, "[%s] err, host state:[%d] is not right\n", + __func__, hdev->state); + return -EBUSY; + } + + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x] init\n", + __func__, cmd->opcode, cmd->info_0.subopcode); + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.common.opcode = cmd->opcode; + admin_cmd.common.flags = cmd->flags; + admin_cmd.common.hdid = cpu_to_le32(cmd->nsid); + admin_cmd.common.cdw2[0] = cpu_to_le32(cmd->cdw2); + admin_cmd.common.cdw2[1] = cpu_to_le32(cmd->cdw3); + admin_cmd.common.cdw10 = cpu_to_le32(cmd->cdw10); + admin_cmd.common.cdw11 = cpu_to_le32(cmd->cdw11); + admin_cmd.common.cdw12 = cpu_to_le32(cmd->cdw12); + admin_cmd.common.cdw13 = cpu_to_le32(cmd->cdw13); + admin_cmd.common.cdw14 = cpu_to_le32(cmd->cdw14); + admin_cmd.common.cdw15 = cpu_to_le32(cmd->cdw15); + + status = spraid_bsg_map_data(hdev, job, &admin_cmd); + if (status) { + dev_err(hdev->dev, "[%s] err, map data failed\n", __func__); + return status; + } + + status = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, &result[0], &result[1], timeout); + if (status >= 0) { + job->reply_len = sizeof(result); + memcpy(job->reply, result, sizeof(result)); + } + + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x], status[0x%x] result0[0x%x] result1[0x%x]\n", + __func__, cmd->opcode, cmd->info_0.subopcode, status, result[0], result[1]); + + spraid_bsg_unmap_data(hdev, job); + + return status; +} + +static int spraid_alloc_ioq_ptcmds(struct spraid_dev *hdev) +{ + int i; + int ptnum = SPRAID_NR_IOQ_PTCMDS; + + INIT_LIST_HEAD(&hdev->ioq_pt_list); + spin_lock_init(&hdev->ioq_pt_lock); + + hdev->ioq_ptcmds = kcalloc_node(ptnum, sizeof(struct spraid_cmd), + GFP_KERNEL, hdev->numa_node); + + if (!hdev->ioq_ptcmds) { + dev_err(hdev->dev, "Alloc ioq_ptcmds failed\n"); + return -ENOMEM; + } + + for (i = 0; i < ptnum; i++) { + hdev->ioq_ptcmds[i].qid = i / SPRAID_PTCMDS_PERQ + 1; + hdev->ioq_ptcmds[i].cid = i % SPRAID_PTCMDS_PERQ + SPRAID_IO_BLK_MQ_DEPTH; + list_add_tail(&(hdev->ioq_ptcmds[i].list), &hdev->ioq_pt_list); + } + + dev_info(hdev->dev, "Alloc ioq_ptcmds success, ptnum[%d]\n", ptnum); + + return 0; +} + +static void spraid_free_ioq_ptcmds(struct spraid_dev *hdev) +{ + kfree(hdev->ioq_ptcmds); + INIT_LIST_HEAD(&hdev->ioq_pt_list); +} + +static int spraid_submit_ioq_sync_cmd(struct spraid_dev *hdev, struct spraid_ioq_command *cmd, + u32 *result, u32 *reslen, u32 timeout) +{ + int ret; + dma_addr_t sense_dma; + struct spraid_queue *ioq; + void *sense_addr = NULL; + struct spraid_cmd *pt_cmd = spraid_get_cmd(hdev, SPRAID_CMD_IOPT); + + if (!pt_cmd) { + dev_err(hdev->dev, "err, get ioq cmd failed\n"); + return -EFAULT; + } + + timeout = timeout ? timeout : ADMIN_TIMEOUT; + + init_completion(&pt_cmd->cmd_done); + + ioq = &hdev->queues[pt_cmd->qid]; + ret = pt_cmd->cid * SCSI_SENSE_BUFFERSIZE; + sense_addr = ioq->sense + ret; + sense_dma = ioq->sense_dma_addr + ret; + + cmd->common.sense_addr = cpu_to_le64(sense_dma); + cmd->common.sense_len = cpu_to_le16(SCSI_SENSE_BUFFERSIZE); + cmd->common.command_id = pt_cmd->cid; + + spraid_submit_cmd(ioq, cmd); + + if (!wait_for_completion_timeout(&pt_cmd->cmd_done, timeout)) { + dev_err(hdev->dev, "[%s] cid[%d] qid[%d] timeout, opcode[0x%x] subopcode[0x%x]\n", + __func__, pt_cmd->cid, pt_cmd->qid, cmd->common.opcode, + (le32_to_cpu(cmd->common.cdw3[0]) & 0xffff)); + WRITE_ONCE(pt_cmd->state, SPRAID_CMD_TIMEOUT); + return -EINVAL; + } + + if (result && reslen) { + if ((pt_cmd->status & 0x17f) == 0x101) { + memcpy(result, sense_addr, SCSI_SENSE_BUFFERSIZE); + *reslen = SCSI_SENSE_BUFFERSIZE; + } + } + + spraid_put_cmd(hdev, pt_cmd, SPRAID_CMD_IOPT); + + return pt_cmd->status; +} + +static int spraid_user_ioq_cmd(struct spraid_dev *hdev, struct bsg_job *job) +{ + struct spraid_bsg_request *bsg_req = (struct spraid_bsg_request *)(job->request); + struct spraid_ioq_passthru_cmd *cmd = &(bsg_req->ioqcmd); + struct spraid_ioq_command ioq_cmd; + int status = 0; + u32 timeout = msecs_to_jiffies(cmd->timeout_ms); + + if (cmd->data_len > PAGE_SIZE) { + dev_err(hdev->dev, "[%s] data len bigger than 4k\n", __func__); + return -EFAULT; + } + + if (hdev->state != SPRAID_LIVE) { + dev_err(hdev->dev, "[%s] err, host state:[%d] is not live\n", + __func__, hdev->state); + return -EBUSY; + } + + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x] init, datalen[%d]\n", + __func__, cmd->opcode, cmd->info_1.subopcode, cmd->data_len); + + memset(&ioq_cmd, 0, sizeof(ioq_cmd)); + ioq_cmd.common.opcode = cmd->opcode; + ioq_cmd.common.flags = cmd->flags; + ioq_cmd.common.hdid = cpu_to_le32(cmd->nsid); + ioq_cmd.common.sense_len = cpu_to_le16(cmd->info_0.res_sense_len); + ioq_cmd.common.cdb_len = cmd->info_0.cdb_len; + ioq_cmd.common.rsvd2 = cmd->info_0.rsvd0; + ioq_cmd.common.cdw3[0] = cpu_to_le32(cmd->cdw3); + ioq_cmd.common.cdw3[1] = cpu_to_le32(cmd->cdw4); + ioq_cmd.common.cdw3[2] = cpu_to_le32(cmd->cdw5); + + ioq_cmd.common.cdw10[0] = cpu_to_le32(cmd->cdw10); + ioq_cmd.common.cdw10[1] = cpu_to_le32(cmd->cdw11); + ioq_cmd.common.cdw10[2] = cpu_to_le32(cmd->cdw12); + ioq_cmd.common.cdw10[3] = cpu_to_le32(cmd->cdw13); + ioq_cmd.common.cdw10[4] = cpu_to_le32(cmd->cdw14); + ioq_cmd.common.cdw10[5] = cpu_to_le32(cmd->data_len); + + memcpy(ioq_cmd.common.cdb, &cmd->cdw16, cmd->info_0.cdb_len); + + ioq_cmd.common.cdw26[0] = cpu_to_le32(cmd->cdw26[0]); + ioq_cmd.common.cdw26[1] = cpu_to_le32(cmd->cdw26[1]); + ioq_cmd.common.cdw26[2] = cpu_to_le32(cmd->cdw26[2]); + ioq_cmd.common.cdw26[3] = cpu_to_le32(cmd->cdw26[3]); + + status = spraid_bsg_map_data(hdev, job, (struct spraid_admin_command *)&ioq_cmd); + if (status) { + dev_err(hdev->dev, "[%s] err, map data failed\n", __func__); + return status; + } + + status = spraid_submit_ioq_sync_cmd(hdev, &ioq_cmd, job->reply, &job->reply_len, timeout); + + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x], status[0x%x] reply_len[%d]\n", + __func__, cmd->opcode, cmd->info_1.subopcode, status, job->reply_len); + + spraid_bsg_unmap_data(hdev, job); + + return status; +} + +static int spraid_reset_work_sync(struct spraid_dev *hdev); + +static bool spraid_check_scmd_completed(struct scsi_cmnd *scmd) +{ + struct spraid_dev *hdev = shost_priv(scmd->device->host); + struct spraid_iod *iod = scsi_cmd_priv(scmd); + struct spraid_queue *spraidq; + u16 hwq, cid; + + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + spraidq = &hdev->queues[hwq]; + if (READ_ONCE(iod->state) == SPRAID_CMD_COMPLETE || spraid_poll_cq(spraidq, cid)) { + dev_warn(hdev->dev, "cid[%d], qid[%d] has been completed\n", + cid, spraidq->qid); + return true; + } + return false; +} + +static enum blk_eh_timer_return spraid_scmd_timeout(struct scsi_cmnd *scmd) +{ + struct spraid_iod *iod = scsi_cmd_priv(scmd); + unsigned int timeout = scmd->device->request_queue->rq_timeout; + + if (spraid_check_scmd_completed(scmd)) + goto out; + + if (time_after(jiffies, scmd->jiffies_at_alloc + timeout)) { + if (cmpxchg(&iod->state, SPRAID_CMD_IN_FLIGHT, SPRAID_CMD_TIMEOUT) == + SPRAID_CMD_IN_FLIGHT) { + return BLK_EH_DONE; + } + } +out: + return BLK_EH_RESET_TIMER; +} + +/* send abort command by admin queue temporary */ +static int spraid_send_abort_cmd(struct spraid_dev *hdev, u32 hdid, u16 qid, u16 cid) +{ + struct spraid_admin_command admin_cmd; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.abort.opcode = SPRAID_ADMIN_ABORT_CMD; + admin_cmd.abort.hdid = cpu_to_le32(hdid); + admin_cmd.abort.sqid = cpu_to_le16(qid); + admin_cmd.abort.cid = cpu_to_le16(cid); + + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); +} + +/* send reset command by admin quueue temporary */ +static int spraid_send_reset_cmd(struct spraid_dev *hdev, int type, u32 hdid) +{ + struct spraid_admin_command admin_cmd; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.reset.opcode = SPRAID_ADMIN_RESET; + admin_cmd.reset.hdid = cpu_to_le32(hdid); + admin_cmd.reset.type = type; + + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); +} + +static bool spraid_change_host_state(struct spraid_dev *hdev, enum spraid_state newstate) +{ + unsigned long flags; + enum spraid_state oldstate; + bool change = false; + + spin_lock_irqsave(&hdev->state_lock, flags); + + oldstate = hdev->state; + switch (newstate) { + case SPRAID_LIVE: + switch (oldstate) { + case SPRAID_NEW: + case SPRAID_RESETTING: + change = true; + break; + default: + break; + } + break; + case SPRAID_RESETTING: + switch (oldstate) { + case SPRAID_LIVE: + change = true; + break; + default: + break; + } + break; + case SPRAID_DELETING: + if (oldstate != SPRAID_DELETING) + change = true; + break; + case SPRAID_DEAD: + switch (oldstate) { + case SPRAID_NEW: + case SPRAID_LIVE: + case SPRAID_RESETTING: + change = true; + break; + default: + break; + } + break; + default: + break; + } + if (change) + hdev->state = newstate; + spin_unlock_irqrestore(&hdev->state_lock, flags); + + dev_info(hdev->dev, "[%s][%d]->[%d], change[%d]\n", __func__, oldstate, newstate, change); + + return change; +} + +static void spraid_back_fault_cqe(struct spraid_queue *ioq, struct spraid_completion *cqe) +{ + struct spraid_dev *hdev = ioq->hdev; + struct blk_mq_tags *tags; + struct scsi_cmnd *scmd; + struct spraid_iod *iod; + struct request *req; + + tags = hdev->shost->tag_set.tags[ioq->qid - 1]; + req = blk_mq_tag_to_rq(tags, cqe->cmd_id); + if (unlikely(!req || !blk_mq_request_started(req))) + return; + + scmd = blk_mq_rq_to_pdu(req); + iod = scsi_cmd_priv(scmd); + + set_host_byte(scmd, DID_NO_CONNECT); + if (iod->nsge) + scsi_dma_unmap(scmd); + spraid_free_iod_res(hdev, iod); + scsi_done(scmd); + dev_warn(hdev->dev, "Back fault CQE, cid[%d], qid[%d]\n", + cqe->cmd_id, ioq->qid); +} + +static void spraid_back_all_io(struct spraid_dev *hdev) +{ + int i, j; + struct spraid_queue *ioq; + struct spraid_completion cqe = { 0 }; + + for (i = 1; i <= hdev->shost->nr_hw_queues; i++) { + ioq = &hdev->queues[i]; + for (j = 0; j < hdev->shost->can_queue; j++) { + cqe.cmd_id = j; + spraid_back_fault_cqe(ioq, &cqe); + } + } +} + +static void spraid_dev_disable(struct spraid_dev *hdev, bool shutdown) +{ + struct spraid_queue *adminq = &hdev->queues[0]; + u16 start, end; + unsigned long timeout = jiffies + 600 * HZ; + + if (pci_device_is_present(hdev->pdev)) { + if (shutdown) + spraid_shutdown_ctrl(hdev); + else + spraid_disable_ctrl(hdev); + } + + while (!time_after(jiffies, timeout)) { + if (!pci_device_is_present(hdev->pdev)) { + dev_info(hdev->dev, "[%s] pci_device not present, skip wait\n", __func__); + break; + } + if (!spraid_wait_ready(hdev, hdev->cap, false)) { + dev_info(hdev->dev, "[%s] wait ready success after reset\n", __func__); + break; + } + dev_info(hdev->dev, "[%s] waiting csts_rdy ready\n", __func__); + } + + if (hdev->queue_count == 0) { + dev_err(hdev->dev, "[%s] warn, queue has been delete\n", __func__); + return; + } + + spin_lock_irq(&adminq->cq_lock); + spraid_process_cq(adminq, &start, &end, -1); + spin_unlock_irq(&adminq->cq_lock); + spraid_complete_cqes(adminq, start, end); + + spraid_pci_disable(hdev); + + spraid_back_all_io(hdev); +} + +static void spraid_reset_work(struct work_struct *work) +{ + int ret; + struct spraid_dev *hdev = container_of(work, struct spraid_dev, reset_work); + + if (hdev->state != SPRAID_RESETTING) { + dev_err(hdev->dev, "[%s] err, host is not reset state\n", __func__); + return; + } + + dev_info(hdev->dev, "[%s] enter host reset\n", __func__); + + if (hdev->ctrl_config & SPRAID_CC_ENABLE) { + dev_info(hdev->dev, "[%s] start dev_disable\n", __func__); + spraid_dev_disable(hdev, false); + } + + ret = spraid_pci_enable(hdev); + if (ret) + goto out; + + ret = spraid_setup_admin_queue(hdev); + if (ret) + goto pci_disable; + + ret = spraid_setup_io_queues(hdev); + if (ret || hdev->online_queues <= hdev->shost->nr_hw_queues) + goto pci_disable; + + spraid_change_host_state(hdev, SPRAID_LIVE); + + spraid_send_all_aen(hdev); + + return; + +pci_disable: + spraid_pci_disable(hdev); +out: + spraid_change_host_state(hdev, SPRAID_DEAD); + dev_err(hdev->dev, "[%s] err, host reset failed\n", __func__); +} + +static int spraid_reset_work_sync(struct spraid_dev *hdev) +{ + if (!spraid_change_host_state(hdev, SPRAID_RESETTING)) { + dev_info(hdev->dev, "[%s] can't change to reset state\n", __func__); + return -EBUSY; + } + + if (!queue_work(spraid_wq, &hdev->reset_work)) { + dev_err(hdev->dev, "[%s] err, host is already in reset state\n", __func__); + return -EBUSY; + } + + flush_work(&hdev->reset_work); + if (hdev->state != SPRAID_LIVE) + return -ENODEV; + + return 0; +} + +static int spraid_wait_abnl_cmd_done(struct spraid_iod *iod) +{ + u16 times = 0; + + do { + if (READ_ONCE(iod->state) == SPRAID_CMD_TMO_COMPLETE) + break; + msleep(500); + times++; + } while (times <= SPRAID_WAIT_ABNL_CMD_TIMEOUT); + + /* wait command completion timeout after abort/reset success */ + if (times >= SPRAID_WAIT_ABNL_CMD_TIMEOUT) + return -ETIMEDOUT; + + return 0; +} + +static int spraid_abort_handler(struct scsi_cmnd *scmd) +{ + struct spraid_dev *hdev = shost_priv(scmd->device->host); + struct spraid_iod *iod = scsi_cmd_priv(scmd); + struct spraid_sdev_hostdata *hostdata; + u16 hwq, cid; + int ret; + + scsi_print_command(scmd); + + if (!spraid_wait_abnl_cmd_done(iod) || spraid_check_scmd_completed(scmd) || + hdev->state != SPRAID_LIVE) + return SUCCESS; + + hostdata = scmd->device->hostdata; + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + + dev_warn(hdev->dev, "cid[%d] qid[%d] timeout, aborting\n", cid, hwq); + ret = spraid_send_abort_cmd(hdev, hostdata->hdid, hwq, cid); + if (ret != ADMIN_ERR_TIMEOUT) { + ret = spraid_wait_abnl_cmd_done(iod); + if (ret) { + dev_warn(hdev->dev, "cid[%d] qid[%d] abort failed, not found\n", cid, hwq); + return FAILED; + } + dev_warn(hdev->dev, "cid[%d] qid[%d] abort succ\n", cid, hwq); + return SUCCESS; + } + dev_warn(hdev->dev, "cid[%d] qid[%d] abort failed, timeout\n", cid, hwq); + return FAILED; +} + +static int spraid_tgt_reset_handler(struct scsi_cmnd *scmd) +{ + struct spraid_dev *hdev = shost_priv(scmd->device->host); + struct spraid_iod *iod = scsi_cmd_priv(scmd); + struct spraid_sdev_hostdata *hostdata; + u16 hwq, cid; + int ret; + + scsi_print_command(scmd); + + if (!spraid_wait_abnl_cmd_done(iod) || spraid_check_scmd_completed(scmd) || + hdev->state != SPRAID_LIVE) + return SUCCESS; + + hostdata = scmd->device->hostdata; + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + + dev_warn(hdev->dev, "cid[%d] qid[%d] timeout, target reset\n", cid, hwq); + ret = spraid_send_reset_cmd(hdev, SPRAID_RESET_TARGET, hostdata->hdid); + if (ret == 0) { + ret = spraid_wait_abnl_cmd_done(iod); + if (ret) { + dev_warn(hdev->dev, "cid[%d] qid[%d]target reset failed, not found\n", + cid, hwq); + return FAILED; + } + + dev_warn(hdev->dev, "cid[%d] qid[%d] target reset success\n", cid, hwq); + return SUCCESS; + } + + dev_warn(hdev->dev, "cid[%d] qid[%d] ret[%d] target reset failed\n", cid, hwq, ret); + return FAILED; +} + +static int spraid_bus_reset_handler(struct scsi_cmnd *scmd) +{ + struct spraid_dev *hdev = shost_priv(scmd->device->host); + struct spraid_iod *iod = scsi_cmd_priv(scmd); + struct spraid_sdev_hostdata *hostdata; + u16 hwq, cid; + int ret; + + scsi_print_command(scmd); + + if (!spraid_wait_abnl_cmd_done(iod) || spraid_check_scmd_completed(scmd) || + hdev->state != SPRAID_LIVE) + return SUCCESS; + + hostdata = scmd->device->hostdata; + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + + dev_warn(hdev->dev, "cid[%d] qid[%d] timeout, bus reset\n", cid, hwq); + ret = spraid_send_reset_cmd(hdev, SPRAID_RESET_BUS, hostdata->hdid); + if (ret == 0) { + ret = spraid_wait_abnl_cmd_done(iod); + if (ret) { + dev_warn(hdev->dev, "cid[%d] qid[%d] bus reset failed, not found\n", + cid, hwq); + return FAILED; + } + + dev_warn(hdev->dev, "cid[%d] qid[%d] bus reset succ\n", cid, hwq); + return SUCCESS; + } + + dev_warn(hdev->dev, "cid[%d] qid[%d] ret[%d] bus reset failed\n", cid, hwq, ret); + return FAILED; +} + +static int spraid_shost_reset_handler(struct scsi_cmnd *scmd) +{ + u16 hwq, cid; + struct spraid_dev *hdev = shost_priv(scmd->device->host); + + scsi_print_command(scmd); + if (spraid_check_scmd_completed(scmd) || hdev->state != SPRAID_LIVE) + return SUCCESS; + + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + dev_warn(hdev->dev, "cid[%d] qid[%d] host reset\n", cid, hwq); + + if (spraid_reset_work_sync(hdev)) { + dev_warn(hdev->dev, "cid[%d] qid[%d] host reset failed\n", cid, hwq); + return FAILED; + } + + dev_warn(hdev->dev, "cid[%d] qid[%d] host reset success\n", cid, hwq); + + return SUCCESS; +} + +static ssize_t csts_pp_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct spraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) { + ret = (readl(hdev->bar + SPRAID_REG_CSTS) & SPRAID_CSTS_PP_MASK); + ret >>= SPRAID_CSTS_PP_SHIFT; + } + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t csts_shst_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct spraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) { + ret = (readl(hdev->bar + SPRAID_REG_CSTS) & SPRAID_CSTS_SHST_MASK); + ret >>= SPRAID_CSTS_SHST_SHIFT; + } + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t csts_cfs_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct spraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) { + ret = (readl(hdev->bar + SPRAID_REG_CSTS) & SPRAID_CSTS_CFS_MASK); + ret >>= SPRAID_CSTS_CFS_SHIFT; + } + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t csts_rdy_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct spraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) + ret = (readl(hdev->bar + SPRAID_REG_CSTS) & SPRAID_CSTS_RDY); + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t fw_version_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct spraid_dev *hdev = shost_priv(shost); + + return snprintf(buf, PAGE_SIZE, "%s\n", hdev->ctrl_info->fr); +} + +static DEVICE_ATTR_RO(csts_pp); +static DEVICE_ATTR_RO(csts_shst); +static DEVICE_ATTR_RO(csts_cfs); +static DEVICE_ATTR_RO(csts_rdy); +static DEVICE_ATTR_RO(fw_version); + +static struct attribute *spraid_host_attrs[] = { + &dev_attr_csts_pp.attr, + &dev_attr_csts_shst.attr, + &dev_attr_csts_cfs.attr, + &dev_attr_csts_rdy.attr, + &dev_attr_fw_version.attr, + NULL, +}; + +static const struct attribute_group spraid_host_attr_group = { + .attrs = spraid_host_attrs +}; + +static const struct attribute_group *spraid_host_attr_groups[] = { + &spraid_host_attr_group, + NULL, +}; + +static struct scsi_host_template spraid_driver_template = { + .module = THIS_MODULE, + .name = "Ramaxel Logic spraid driver", + .proc_name = "spraid", + .queuecommand = spraid_queue_command, + .slave_alloc = spraid_slave_alloc, + .slave_destroy = spraid_slave_destroy, + .slave_configure = spraid_slave_configure, + .eh_timed_out = spraid_scmd_timeout, + .eh_abort_handler = spraid_abort_handler, + .eh_target_reset_handler = spraid_tgt_reset_handler, + .eh_bus_reset_handler = spraid_bus_reset_handler, + .eh_host_reset_handler = spraid_shost_reset_handler, + .change_queue_depth = scsi_change_queue_depth, + .host_tagset = 0, + .this_id = -1, + .shost_groups = spraid_host_attr_groups, +}; + +static void spraid_shutdown(struct pci_dev *pdev) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + + spraid_remove_io_queues(hdev); + spraid_disable_admin_queue(hdev, true); +} + +/* bsg dispatch user command */ +static int spraid_bsg_host_dispatch(struct bsg_job *job) +{ + struct Scsi_Host *shost = dev_to_shost(job->dev); + struct spraid_dev *hdev = shost_priv(shost); + struct request *rq = blk_mq_rq_from_pdu(job); + struct spraid_bsg_request *bsg_req = job->request; + int ret = 0; + + dev_info(hdev->dev, "[%s]msgcode[%d], msglen[%d], timeout[%d], req_nsge[%d], req_len[%d]\n", + __func__, bsg_req->msgcode, job->request_len, rq->timeout, + job->request_payload.sg_cnt, job->request_payload.payload_len); + + job->reply_len = 0; + + switch (bsg_req->msgcode) { + case SPRAID_BSG_ADM: + ret = spraid_user_admin_cmd(hdev, job); + break; + case SPRAID_BSG_IOQ: + ret = spraid_user_ioq_cmd(hdev, job); + break; + default: + dev_info(hdev->dev, "[%s] unsupport msgcode[%d]\n", __func__, bsg_req->msgcode); + break; + } + + bsg_job_done(job, ret, 0); + return 0; +} + +static inline void spraid_remove_bsg(struct spraid_dev *hdev) +{ + struct request_queue *q = hdev->bsg_queue; + + if (q) + bsg_remove_queue(q); +} +static int spraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) +{ + struct spraid_dev *hdev; + struct Scsi_Host *shost; + int node, ret; + char bsg_name[15]; + + shost = scsi_host_alloc(&spraid_driver_template, sizeof(*hdev)); + if (!shost) { + dev_err(&pdev->dev, "Failed to allocate scsi host\n"); + return -ENOMEM; + } + hdev = shost_priv(shost); + hdev->pdev = pdev; + hdev->dev = get_device(&pdev->dev); + + node = dev_to_node(hdev->dev); + if (node == NUMA_NO_NODE) { + node = first_memory_node; + set_dev_node(hdev->dev, node); + } + hdev->numa_node = node; + hdev->shost = shost; + pci_set_drvdata(pdev, hdev); + + ret = spraid_dev_map(hdev); + if (ret) + goto put_dev; + + init_rwsem(&hdev->devices_rwsem); + INIT_WORK(&hdev->scan_work, spraid_scan_work); + INIT_WORK(&hdev->timesyn_work, spraid_timesyn_work); + INIT_WORK(&hdev->reset_work, spraid_reset_work); + spin_lock_init(&hdev->state_lock); + + ret = spraid_alloc_resources(hdev); + if (ret) + goto dev_unmap; + + ret = spraid_pci_enable(hdev); + if (ret) + goto resources_free; + + ret = spraid_setup_admin_queue(hdev); + if (ret) + goto pci_disable; + + ret = spraid_init_ctrl_info(hdev); + if (ret) + goto disable_admin_q; + + ret = spraid_alloc_iod_ext_mem_pool(hdev); + if (ret) + goto disable_admin_q; + + ret = spraid_setup_io_queues(hdev); + if (ret) + goto free_iod_mempool; + + spraid_shost_init(hdev); + + ret = scsi_add_host(hdev->shost, hdev->dev); + if (ret) { + dev_err(hdev->dev, "Add shost to system failed, ret: %d\n", + ret); + goto remove_io_queues; + } + + snprintf(bsg_name, sizeof(bsg_name), "spraid%d", shost->host_no); + hdev->bsg_queue = bsg_setup_queue(&shost->shost_gendev, bsg_name, + spraid_bsg_host_dispatch, NULL, + spraid_cmd_size(hdev, true, false)); + if (IS_ERR(hdev->bsg_queue)) { + dev_err(hdev->dev, "err, setup bsg failed\n"); + hdev->bsg_queue = NULL; + goto remove_io_queues; + } + + if (hdev->online_queues == SPRAID_ADMIN_QUEUE_NUM) { + dev_warn(hdev->dev, "warn only admin queue can be used\n"); + return 0; + } + + hdev->state = SPRAID_LIVE; + + spraid_send_all_aen(hdev); + + ret = spraid_dev_list_init(hdev); + if (ret) + goto remove_bsg; + + ret = spraid_configure_timestamp(hdev); + if (ret) + dev_warn(hdev->dev, "init set timestamp failed\n"); + + ret = spraid_alloc_ioq_ptcmds(hdev); + if (ret) + goto remove_bsg; + + scsi_scan_host(hdev->shost); + + return 0; + +remove_bsg: + spraid_remove_bsg(hdev); +remove_io_queues: + spraid_remove_io_queues(hdev); +free_iod_mempool: + spraid_free_iod_ext_mem_pool(hdev); +disable_admin_q: + spraid_disable_admin_queue(hdev, false); +pci_disable: + spraid_pci_disable(hdev); +resources_free: + spraid_free_resources(hdev); +dev_unmap: + spraid_dev_unmap(hdev); +put_dev: + put_device(hdev->dev); + scsi_host_put(shost); + + return -ENODEV; +} + +static void spraid_remove(struct pci_dev *pdev) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + struct Scsi_Host *shost = hdev->shost; + + dev_info(hdev->dev, "enter spraid remove\n"); + + spraid_change_host_state(hdev, SPRAID_DELETING); + + if (!pci_device_is_present(pdev)) { + scsi_block_requests(shost); + spraid_back_all_io(hdev); + scsi_unblock_requests(shost); + } + + flush_work(&hdev->reset_work); + spraid_remove_bsg(hdev); + scsi_remove_host(shost); + spraid_free_ioq_ptcmds(hdev); + kfree(hdev->devices); + spraid_remove_io_queues(hdev); + spraid_free_iod_ext_mem_pool(hdev); + spraid_disable_admin_queue(hdev, false); + spraid_pci_disable(hdev); + spraid_free_resources(hdev); + spraid_dev_unmap(hdev); + put_device(hdev->dev); + scsi_host_put(shost); + + dev_info(hdev->dev, "exit spraid remove\n"); +} + +static const struct pci_device_id spraid_id_table[] = { + { PCI_DEVICE(PCI_VENDOR_ID_RAMAXEL_LOGIC, SPRAID_SERVER_DEVICE_HBA_DID) }, + { PCI_DEVICE(PCI_VENDOR_ID_RAMAXEL_LOGIC, SPRAID_SERVER_DEVICE_RAID_DID) }, + { 0, } +}; +MODULE_DEVICE_TABLE(pci, spraid_id_table); + +static struct pci_driver spraid_driver = { + .name = "spraid", + .id_table = spraid_id_table, + .probe = spraid_probe, + .remove = spraid_remove, + .shutdown = spraid_shutdown, +}; + +static int __init spraid_init(void) +{ + int ret; + + spraid_wq = alloc_workqueue("spraid-wq", WQ_UNBOUND | WQ_MEM_RECLAIM | WQ_SYSFS, 0); + if (!spraid_wq) + return -ENOMEM; + + spraid_class = class_create(THIS_MODULE, "spraid"); + if (IS_ERR(spraid_class)) { + ret = PTR_ERR(spraid_class); + goto destroy_wq; + } + + ret = pci_register_driver(&spraid_driver); + if (ret < 0) + goto destroy_class; + + return 0; + +destroy_class: + class_destroy(spraid_class); +destroy_wq: + destroy_workqueue(spraid_wq); + + return ret; +} + +static void __exit spraid_exit(void) +{ + pci_unregister_driver(&spraid_driver); + class_destroy(spraid_class); + destroy_workqueue(spraid_wq); + ida_destroy(&spraid_instance_ida); +} + +MODULE_AUTHOR("songyl(a)ramaxel.com"); +MODULE_DESCRIPTION("Ramaxel Memory Technology SPraid Driver"); +MODULE_LICENSE("GPL"); +MODULE_VERSION(SPRAID_DRV_VERSION); +module_init(spraid_init); +module_exit(spraid_exit); -- 2.27.0

1 0

[PATCH openEuler-5.10 00/49] KABI: Add KABI paddings for base
by Zheng Zengkai 31 Dec '21

31 Dec '21

KABI: Add KABI paddings for base structures Alexey Gladkov (1): Increase size of ucounts to atomic_long_t Cui GaoSheng (1): kabi: reserve space for net_namespace GONG, Ruiqi (1): kabi: reserve space for cred and user_namespace Guan Jing (1): KABI:reserve space for sched structures Guo Zihua (1): KABI: reserve space for IMA IPE Gustavo A. R. Silva (1): UAPI: nfsfh.h: Replace one-element array with flexible-array member Jialin Zhang (2): kabi: reserve space for posix clock related structure kabi: reserve space for power management related structure Lin Ruizhe (6): bootparam: Add kabi_reserve in bootparam interrupt: Add kabi_reserve in interrupt.h irq: Add kabi_reserve in irq irq_desc: Add kabi_reserve in irq_desc irqdomain: Add kabi_reserve in irqdomain msi: Add kabi_reserve in msi.h Lu Jialin (4): kabi: reserve space for cgroup framework related structures kabi: reserve space for memcg related structures kabi: reserve space for cpu cgroup and cpuset cgroup related structures kabi: reserve space for cgroup bpf structures Tan Xiaojun (1): kabi: reserve space for pci subsystem related structure Wang Hai (6): kabi: net: reserve space for net base subsystem related structure kabi: net: reserve space for net can subsystem related structure kabi: net: reserve space for net sunrpc subsystem related structure kabi: net: reserve space for net rdma subsystem related structure kabi: net: reserve space for net bpf subsystem related structure kabi: net: reserve space for net netfilter subsystem related structure Wang ShaoBo (8): kabi: reserve space for io subsystem related structures kabi: reserve space for kobject related structures kabi: reserve space for struct module kabi: reserve space for struct ptp_clock kabi: reserve space for struct ptp_clock_info kabi: reserve space for ptp_clock.h kabi: reserve space for iommu.h kabi: reserve space for fwnode.h Xie XiuQi (6): kabi: add kabi helper macros kabi: add KABI_SIZE_ALIGN_CHECKS for more stringent kabi checks kabi: enables more stringent kabi checks kabi: add script tools to check kabi symbol kabi: add a tool to generate the kabi reference relationship kabi: add kABI reference checking tool Yang Jihong (1): kabi: reserve space for perf subsystem related structures Yang Yingliang (2): kabi: reserve space for struct cpu_stop_work kabi: reserve space for struct dma_map_ops Yongqiang Liu (1): kabi: mm: reserve space for memory subsystem related Yu Liao (3): kabi: reserve space for struct worker kabi: reserve space for time and workqueue subsystem related structure kabi: reserve space for hrtimer related structures Zheng Zengkai (2): KABI: add KABI padding to cpuidle structures KABI: add KABI padding to x86/paravirt ops structures Zhihao Cheng (1): kabi: Add kabi reservation for storage module Kconfig | 7 + arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + arch/x86/include/asm/paravirt_types.h | 9 + arch/x86/include/uapi/asm/bootparam.h | 2 + block/blk-mq-tag.h | 7 + drivers/nvme/host/nvme.h | 6 + drivers/pci/pci.h | 9 + drivers/ptp/ptp_private.h | 4 + include/linux/backing-dev-defs.h | 11 + include/linux/bio.h | 10 + include/linux/blk-cgroup.h | 11 + include/linux/blk-mq.h | 30 ++ include/linux/blk_types.h | 12 + include/linux/blkdev.h | 16 + include/linux/bpf-cgroup.h | 18 + include/linux/can/core.h | 4 + include/linux/can/dev.h | 6 + include/linux/can/rx-offload.h | 4 + include/linux/can/skb.h | 1 + include/linux/cgroup-defs.h | 29 ++ include/linux/cpuidle.h | 15 + include/linux/cred.h | 12 + include/linux/dcache.h | 9 + include/linux/delayacct.h | 4 + include/linux/device.h | 19 + include/linux/device/class.h | 7 + include/linux/device/driver.h | 6 + include/linux/dma-map-ops.h | 7 + include/linux/elevator.h | 15 + include/linux/ethtool.h | 6 + include/linux/exportfs.h | 4 + include/linux/fs.h | 48 +++ include/linux/fsnotify_backend.h | 3 + include/linux/fwnode.h | 7 + include/linux/genhd.h | 14 + include/linux/hrtimer.h | 10 + include/linux/interrupt.h | 3 + include/linux/iomap.h | 6 + include/linux/iommu.h | 12 + include/linux/ioport.h | 6 + include/linux/ipv6.h | 6 + include/linux/irq.h | 6 + include/linux/irqdesc.h | 3 +- include/linux/irqdomain.h | 2 + include/linux/jbd2.h | 6 + include/linux/kabi.h | 494 ++++++++++++++++++++++++- include/linux/kernfs.h | 9 + include/linux/kobject.h | 13 + include/linux/lsm_hooks.h | 2 + include/linux/memcontrol.h | 20 + include/linux/mm.h | 6 + include/linux/mm_types.h | 15 + include/linux/mmu_notifier.h | 9 + include/linux/mmzone.h | 9 + include/linux/module.h | 4 + include/linux/mount.h | 3 + include/linux/msi.h | 3 + include/linux/net.h | 6 + include/linux/netdevice.h | 47 +++ include/linux/netfilter.h | 9 + include/linux/netfilter/ipset/ip_set.h | 7 + include/linux/netfilter/nfnetlink.h | 5 + include/linux/netfilter_ipv6.h | 3 + include/linux/ns_common.h | 4 + include/linux/pci.h | 34 ++ include/linux/pci_hotplug.h | 18 + include/linux/perf_event.h | 8 + include/linux/pm.h | 14 + include/linux/pm_domain.h | 9 + include/linux/pm_qos.h | 7 + include/linux/pm_wakeup.h | 4 + include/linux/posix-clock.h | 17 + include/linux/ptp_clock_kernel.h | 4 + include/linux/quota.h | 7 + include/linux/sbitmap.h | 3 + include/linux/sched.h | 30 ++ include/linux/sched/signal.h | 6 + include/linux/sched/topology.h | 4 + include/linux/sched/user.h | 4 + include/linux/skbuff.h | 6 + include/linux/skmsg.h | 11 + include/linux/stop_machine.h | 2 + include/linux/sunrpc/cache.h | 4 + include/linux/sunrpc/clnt.h | 10 + include/linux/sunrpc/sched.h | 14 + include/linux/sunrpc/stats.h | 5 + include/linux/sunrpc/xprt.h | 17 + include/linux/swap.h | 4 + include/linux/sysfs.h | 3 + include/linux/timer.h | 6 + include/linux/user_namespace.h | 36 +- include/linux/workqueue.h | 11 + include/linux/writeback.h | 4 + include/linux/xattr.h | 3 + include/net/dcbnl.h | 9 + include/net/dst.h | 10 + include/net/dst_ops.h | 10 + include/net/fib_rules.h | 10 + include/net/flow.h | 10 + include/net/genetlink.h | 10 + include/net/inet_connection_sock.h | 5 + include/net/ip6_fib.h | 9 + include/net/l3mdev.h | 6 + include/net/lwtunnel.h | 7 + include/net/neighbour.h | 8 + include/net/net_namespace.h | 6 + include/net/netfilter/nf_conntrack.h | 4 + include/net/netlink.h | 5 + include/net/netns/can.h | 3 + include/net/netns/ipv4.h | 8 + include/net/netns/ipv6.h | 3 + include/net/netns/netfilter.h | 3 + include/net/netns/nftables.h | 3 + include/net/netns/xfrm.h | 3 + include/net/page_pool.h | 3 + include/net/rtnetlink.h | 10 + include/net/sch_generic.h | 7 + include/net/sock.h | 19 + include/net/switchdev.h | 6 + include/net/tcp.h | 6 + include/net/tls.h | 5 + include/net/xdp.h | 6 + include/net/xfrm.h | 6 + include/rdma/ib_addr.h | 3 + include/rdma/rdma_cm.h | 8 + include/rdma/rdma_counter.h | 3 + include/rdma/rdma_netlink.h | 4 + include/rdma/rdma_vt.h | 6 + include/rdma/rdmavt_qp.h | 5 + include/scsi/scsi_cmnd.h | 6 + include/scsi/scsi_device.h | 15 + include/scsi/scsi_host.h | 13 + include/scsi/scsi_transport_fc.h | 25 ++ include/target/target_core_base.h | 7 + include/uapi/linux/nfsd/nfsfh.h | 27 +- include/uapi/linux/ptp_clock.h | 4 + kernel/cgroup/cpuset.c | 6 + kernel/sched/sched.h | 23 ++ kernel/ucount.c | 32 +- kernel/workqueue_internal.h | 6 + scripts/check-kabi | 147 ++++++++ scripts/kabideps | 161 ++++++++ scripts/kabisyms | 141 +++++++ 144 files changed, 2220 insertions(+), 29 deletions(-) create mode 100755 scripts/check-kabi create mode 100755 scripts/kabideps create mode 100755 scripts/kabisyms -- 2.20.1

1 49

[PATCH openEuler-5.10 01/71] perf mem: Search event name with more flexible path
by Zheng Zengkai 31 Dec '21

31 Dec '21

From: Leo Yan <leo.yan(a)linaro.org> mainline inclusion from mainline-v5.11-rc1 commit f9f16dfbe76e category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4NGPV CVE: NA ------------------------------------------------- The perf tool searches a memory event name under the folder '/sys/devices/cpu/events/', this leads to the limitation for the selection of a memory profiling event which must be under this folder. Thus it's impossible to use any other event as memory event which is not under this specific folder, e.g. Arm SPE hardware event is not located in '/sys/devices/cpu/events/' so it cannot be enabled for memory profiling. This patch changes to search folder from '/sys/devices/cpu/events/' to '/sys/devices', so it give flexibility to find events which can be used for memory profiling. Signed-off-by: Leo Yan <leo.yan(a)linaro.org> Acked-by: Jiri Olsa <jolsa(a)redhat.com> Link: https://lore.kernel.org/r/20201106094853.21082-2-leo.yan@linaro.org Signed-off-by: Arnaldo Carvalho de Melo <acme(a)redhat.com> Signed-off-by: Wei Li <liwei391(a)huawei.com> Reviewed-by: Yang Jihong <yangjihong1(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> --- tools/perf/util/mem-events.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/tools/perf/util/mem-events.c b/tools/perf/util/mem-events.c index ea0af0bc4314..35c8d175a9d2 100644 --- a/tools/perf/util/mem-events.c +++ b/tools/perf/util/mem-events.c @@ -18,8 +18,8 @@ unsigned int perf_mem_events__loads_ldlat = 30; #define E(t, n, s) { .tag = t, .name = n, .sysfs_name = s } struct perf_mem_event perf_mem_events[PERF_MEM_EVENTS__MAX] = { - E("ldlat-loads", "cpu/mem-loads,ldlat=%u/P", "mem-loads"), - E("ldlat-stores", "cpu/mem-stores/P", "mem-stores"), + E("ldlat-loads", "cpu/mem-loads,ldlat=%u/P", "cpu/events/mem-loads"), + E("ldlat-stores", "cpu/mem-stores/P", "cpu/events/mem-stores"), }; #undef E @@ -93,7 +93,7 @@ int perf_mem_events__init(void) struct perf_mem_event *e = &perf_mem_events[j]; struct stat st; - scnprintf(path, PATH_MAX, "%s/devices/cpu/events/%s", + scnprintf(path, PATH_MAX, "%s/devices/%s", mnt, e->sysfs_name); if (!stat(path, &st)) -- 2.20.1

1 70

[PATCH openEuler-1.0-LTS 1/4] net: hns3: fix the concurrency between functions reading debugfs
by Yang Yingliang 31 Dec '21

31 Dec '21

From: Yufeng Mo <moyufeng(a)huawei.com> driver inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I4OSRU CVE: NA ---------------------------- [1298504.847848] Call trace: [1298504.847859] [<ffff000008089e14>] dump_backtrace+0x0/0x23c [1298504.847865] [<ffff00000808a074>] show_stack+0x24/0x2c [1298504.847870] [<ffff0000088568a8>] dump_stack+0x84/0xa8 [1298504.847878] [<ffff0000082122fc>] bad_page+0xec/0x14c [1298504.847883] [<ffff000008219384>] free_pages_check_bad+0x90/0x9c [1298504.847888] [<ffff00000821307c>] __free_pages_ok+0x2b8/0x2ec [1298504.847894] [<ffff0000082153ec>] __free_pages+0x44/0x64 [1298504.847900] [<ffff000008288788>] kfree+0x198/0x1a0 [1298504.847905] [<ffff00000823432c>] kvfree+0x3c/0x58 [1298504.847937] [<ffff0000014fabf4>] hns3_dbg_read+0xf4/0x278 [hns3] [1298504.847944] [<ffff000008359550>] full_proxy_read+0x60/0x90 [1298504.847949] [<ffff0000082b22a4>] __vfs_read+0x58/0x178 [1298504.847952] [<ffff0000082b2454>] vfs_read+0x90/0x14c [1298504.847956] [<ffff0000082b2b70>] SyS_read+0x60/0xc0 When different functions reading the same debugfs node, it will cause double free problem, because different functions shared the same node buffer. This patch make different functions have their own buffer to fix the problem. Fixes: 319ba0a4d154 ("net: hns3: fix race condition in debugfs") Fixes: c91910efc03a ("net: hns3: refactor the debugfs process") Signed-off-by: Yufeng Mo <moyufeng(a)huawei.com> Signed-off-by: Yonglong Liu <liuyonglong(a)huawei.com> Reviewed-by: Jian Shen <shenjian15(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/net/ethernet/hisilicon/hns3/hnae3.h | 1 + .../net/ethernet/hisilicon/hns3/hns3_debugfs.c | 15 +++++++++++---- .../net/ethernet/hisilicon/hns3/hns3_debugfs.h | 1 - 3 files changed, 12 insertions(+), 5 deletions(-) diff --git a/drivers/net/ethernet/hisilicon/hns3/hnae3.h b/drivers/net/ethernet/hisilicon/hns3/hnae3.h index 0d849c13f22f5..61b92430e56db 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hnae3.h +++ b/drivers/net/ethernet/hisilicon/hns3/hnae3.h @@ -765,6 +765,7 @@ struct hnae3_handle { u8 netdev_flags; struct dentry *hnae3_dbgfs; struct mutex dbgfs_lock; + char **dbgfs_buf; /* Network interface message level enabled bits */ u32 msg_enable; diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c b/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c index c68e5f3d0ba52..8dbbf597d8a2c 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c @@ -808,7 +808,7 @@ static ssize_t hns3_dbg_read(struct file *filp, char __user *buffer, return ret; mutex_lock(&handle->dbgfs_lock); - save_buf = &hns3_dbg_cmd[index].buf; + save_buf = &handle->dbgfs_buf[index]; if (!test_bit(HNS3_NIC_STATE_INITED, &priv->state) || test_bit(HNS3_NIC_STATE_RESETTING, &priv->state)) { @@ -911,6 +911,13 @@ int hns3_dbg_init(struct hnae3_handle *handle) int ret; u32 i; + handle->dbgfs_buf = devm_kcalloc(&handle->pdev->dev, + ARRAY_SIZE(hns3_dbg_cmd), + sizeof(*handle->dbgfs_buf), + GFP_KERNEL); + if (!handle->dbgfs_buf) + return -ENOMEM; + hns3_dbg_dentry[HNS3_DBG_DENTRY_COMMON].dentry = debugfs_create_dir(name, hns3_dbgfs_root); handle->hnae3_dbgfs = hns3_dbg_dentry[HNS3_DBG_DENTRY_COMMON].dentry; @@ -952,9 +959,9 @@ void hns3_dbg_uninit(struct hnae3_handle *handle) u32 i; for (i = 0; i < ARRAY_SIZE(hns3_dbg_cmd); i++) - if (hns3_dbg_cmd[i].buf) { - kvfree(hns3_dbg_cmd[i].buf); - hns3_dbg_cmd[i].buf = NULL; + if (handle->dbgfs_buf[i]) { + kvfree(handle->dbgfs_buf[i]); + handle->dbgfs_buf[i] = NULL; } mutex_destroy(&handle->dbgfs_lock); diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.h b/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.h index ea65e5174ea98..902e16d99fb7c 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.h +++ b/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.h @@ -49,7 +49,6 @@ struct hns3_dbg_cmd_info { enum hnae3_dbg_cmd cmd; enum hns3_dbg_dentry_type dentry; u32 buf_len; - char *buf; int (*init)(struct hnae3_handle *handle, unsigned int cmd); }; -- 2.25.1

1 3

[PATCH openEuler-1.0-LTS v2] f2fs: fix to do sanity check on last xattr entry in __f2fs_setxattr()
by Yang Yingliang 31 Dec '21

31 Dec '21

From: Chao Yu <chao(a)kernel.org> mainline inclusion from mainline-v5.17 commit 5598b24efaf4892741c798b425d543e4bed357a1 category: bugfix CVE: CVE-2021-45469 -------------------------------- As Wenqing Liu reported in bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=215235 - Overview page fault in f2fs_setxattr() when mount and operate on corrupted image - Reproduce tested on kernel 5.16-rc3, 5.15.X under root 1. unzip tmp7.zip 2. ./single.sh f2fs 7 Sometimes need to run the script several times - Kernel dump loop0: detected capacity change from 0 to 131072 F2FS-fs (loop0): Found nat_bits in checkpoint F2FS-fs (loop0): Mounted with checkpoint version = 7548c2ee BUG: unable to handle page fault for address: ffffe47bc7123f48 RIP: 0010:kfree+0x66/0x320 Call Trace: __f2fs_setxattr+0x2aa/0xc00 [f2fs] f2fs_setxattr+0xfa/0x480 [f2fs] __f2fs_set_acl+0x19b/0x330 [f2fs] __vfs_removexattr+0x52/0x70 __vfs_removexattr_locked+0xb1/0x140 vfs_removexattr+0x56/0x100 removexattr+0x57/0x80 path_removexattr+0xa3/0xc0 __x64_sys_removexattr+0x17/0x20 do_syscall_64+0x37/0xb0 entry_SYSCALL_64_after_hwframe+0x44/0xae The root cause is in __f2fs_setxattr(), we missed to do sanity check on last xattr entry, result in out-of-bound memory access during updating inconsistent xattr data of target inode. After the fix, it can detect such xattr inconsistency as below: F2FS-fs (loop11): inode (7) has invalid last xattr entry, entry_size: 60676 F2FS-fs (loop11): inode (8) has corrupted xattr F2FS-fs (loop11): inode (8) has corrupted xattr F2FS-fs (loop11): inode (8) has invalid last xattr entry, entry_size: 47736 Cc: stable(a)vger.kernel.org Reported-by: Wenqing Liu <wenqingliu0120(a)gmail.com> Signed-off-by: Chao Yu <chao(a)kernel.org> Signed-off-by: Jaegeuk Kim <jaegeuk(a)kernel.org> Conflicts: fs/f2fs/xattr.c [yyl: replace f2fs_err() with f2fs_msg(KERN_ERR)] Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> Reviewed-by: Xiu Jianfeng <xiujianfeng(a)huawei.com> Reviewed-by: fang wei <fangwei1(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- fs/f2fs/xattr.c | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-) diff --git a/fs/f2fs/xattr.c b/fs/f2fs/xattr.c index 90d0d305f6a36..1858c66d871d1 100644 --- a/fs/f2fs/xattr.c +++ b/fs/f2fs/xattr.c @@ -657,8 +657,17 @@ static int __f2fs_setxattr(struct inode *inode, int index, } last = here; - while (!IS_XATTR_LAST_ENTRY(last)) + while (!IS_XATTR_LAST_ENTRY(last)) { + if ((void *)(last) + sizeof(__u32) > last_base_addr || + (void *)XATTR_NEXT_ENTRY(last) > last_base_addr) { + f2fs_msg(F2FS_I_SB(inode)->sb, KERN_ERR, "inode (%lu) has invalid last xattr entry, entry_size: %zu", + inode->i_ino, ENTRY_SIZE(last)); + set_sbi_flag(F2FS_I_SB(inode), SBI_NEED_FSCK); + error = -EFSCORRUPTED; + goto exit; + } last = XATTR_NEXT_ENTRY(last); + } newsize = XATTR_ALIGN(sizeof(struct f2fs_xattr_entry) + len + size); -- 2.25.1

1 0

[PATCH openEuler-1.0-LTS 0/3] fixes CVE-2021-44733
by Yang Yingliang 30 Dec '21

30 Dec '21

patch #1 ~ #2 prepare for backporting fix patch patch #3 fix CVE-2021-44733 Jens Wiklander (3): tee: remove linked list of struct tee_shm tee: don't assign shm id for private shms tee: handle lookup of shm with reference count 0 drivers/tee/tee_core.c | 1 - drivers/tee/tee_private.h | 3 +- drivers/tee/tee_shm.c | 196 ++++++++++++++------------------------ include/linux/tee_drv.h | 7 +- 4 files changed, 77 insertions(+), 130 deletions(-) -- 2.25.1

1 3

[PATCH] f2fs: fix to do sanity check on last xattr entry in __f2fs_setxattr()
by Yang Yingliang 30 Dec '21

30 Dec '21

From: Chao Yu <chao(a)kernel.org> mainline inclusion from mainline-v5.17 commit 5598b24efaf4892741c798b425d543e4bed357a1 category: bugfix CVE: CVE-2021-45469 -------------------------------- As Wenqing Liu reported in bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=215235 - Overview page fault in f2fs_setxattr() when mount and operate on corrupted image - Reproduce tested on kernel 5.16-rc3, 5.15.X under root 1. unzip tmp7.zip 2. ./single.sh f2fs 7 Sometimes need to run the script several times - Kernel dump loop0: detected capacity change from 0 to 131072 F2FS-fs (loop0): Found nat_bits in checkpoint F2FS-fs (loop0): Mounted with checkpoint version = 7548c2ee BUG: unable to handle page fault for address: ffffe47bc7123f48 RIP: 0010:kfree+0x66/0x320 Call Trace: __f2fs_setxattr+0x2aa/0xc00 [f2fs] f2fs_setxattr+0xfa/0x480 [f2fs] __f2fs_set_acl+0x19b/0x330 [f2fs] __vfs_removexattr+0x52/0x70 __vfs_removexattr_locked+0xb1/0x140 vfs_removexattr+0x56/0x100 removexattr+0x57/0x80 path_removexattr+0xa3/0xc0 __x64_sys_removexattr+0x17/0x20 do_syscall_64+0x37/0xb0 entry_SYSCALL_64_after_hwframe+0x44/0xae The root cause is in __f2fs_setxattr(), we missed to do sanity check on last xattr entry, result in out-of-bound memory access during updating inconsistent xattr data of target inode. After the fix, it can detect such xattr inconsistency as below: F2FS-fs (loop11): inode (7) has invalid last xattr entry, entry_size: 60676 F2FS-fs (loop11): inode (8) has corrupted xattr F2FS-fs (loop11): inode (8) has corrupted xattr F2FS-fs (loop11): inode (8) has invalid last xattr entry, entry_size: 47736 Cc: stable(a)vger.kernel.org Reported-by: Wenqing Liu <wenqingliu0120(a)gmail.com> Signed-off-by: Chao Yu <chao(a)kernel.org> Signed-off-by: Jaegeuk Kim <jaegeuk(a)kernel.org> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- fs/f2fs/xattr.c | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-) diff --git a/fs/f2fs/xattr.c b/fs/f2fs/xattr.c index 90d0d305f6a36..b55d0920e777c 100644 --- a/fs/f2fs/xattr.c +++ b/fs/f2fs/xattr.c @@ -657,8 +657,17 @@ static int __f2fs_setxattr(struct inode *inode, int index, } last = here; - while (!IS_XATTR_LAST_ENTRY(last)) + while (!IS_XATTR_LAST_ENTRY(last)) { + if ((void *)(last) + sizeof(__u32) > last_base_addr || + (void *)XATTR_NEXT_ENTRY(last) > last_base_addr) { + f2fs_err(F2FS_I_SB(inode), "inode (%lu) has invalid last xattr entry, entry_size: %zu", + inode->i_ino, ENTRY_SIZE(last)); + set_sbi_flag(F2FS_I_SB(inode), SBI_NEED_FSCK); + error = -EFSCORRUPTED; + goto exit; + } last = XATTR_NEXT_ENTRY(last); + } newsize = XATTR_ALIGN(sizeof(struct f2fs_xattr_entry) + len + size); -- 2.25.1

1 0

[PATCH] mwifiex: Fix skb_over_panic in mwifiex_usb_recv()
by Yang Yingliang 30 Dec '21

30 Dec '21

From: Zekun Shen <bruceshenzk(a)gmail.com> mainline inclusion from mainline-v5.17 commit 04d80663f67ccef893061b49ec8a42ff7045ae84 category: bugfix CVE: CVE-2021-43976 -------------------------------- Currently, with an unknown recv_type, mwifiex_usb_recv just return -1 without restoring the skb. Next time mwifiex_usb_rx_complete is invoked with the same skb, calling skb_put causes skb_over_panic. The bug is triggerable with a compromised/malfunctioning usb device. After applying the patch, skb_over_panic no longer shows up with the same input. Attached is the panic report from fuzzing. skbuff: skb_over_panic: text:000000003bf1b5fa len:2048 put:4 head:00000000dd6a115b data:000000000a9445d8 tail:0x844 end:0x840 dev:<NULL> kernel BUG at net/core/skbuff.c:109! invalid opcode: 0000 [#1] SMP KASAN NOPTI CPU: 0 PID: 198 Comm: in:imklog Not tainted 5.6.0 #60 RIP: 0010:skb_panic+0x15f/0x161 Call Trace: <IRQ> ? mwifiex_usb_rx_complete+0x26b/0xfcd [mwifiex_usb] skb_put.cold+0x24/0x24 mwifiex_usb_rx_complete+0x26b/0xfcd [mwifiex_usb] __usb_hcd_giveback_urb+0x1e4/0x380 usb_giveback_urb_bh+0x241/0x4f0 ? __hrtimer_run_queues+0x316/0x740 ? __usb_hcd_giveback_urb+0x380/0x380 tasklet_action_common.isra.0+0x135/0x330 __do_softirq+0x18c/0x634 irq_exit+0x114/0x140 smp_apic_timer_interrupt+0xde/0x380 apic_timer_interrupt+0xf/0x20 </IRQ> Reported-by: Brendan Dolan-Gavitt <brendandg(a)nyu.edu> Signed-off-by: Zekun Shen <bruceshenzk(a)gmail.com> Signed-off-by: Kalle Valo <kvalo(a)codeaurora.org> Link: https://lore.kernel.org/r/YX4CqjfRcTa6bVL+@Zekuns-MBP-16.fios-router.home Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/net/wireless/marvell/mwifiex/usb.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/wireless/marvell/mwifiex/usb.c b/drivers/net/wireless/marvell/mwifiex/usb.c index d445acc4786b7..22a42eb0cc0ac 100644 --- a/drivers/net/wireless/marvell/mwifiex/usb.c +++ b/drivers/net/wireless/marvell/mwifiex/usb.c @@ -130,7 +130,8 @@ static int mwifiex_usb_recv(struct mwifiex_adapter *adapter, default: mwifiex_dbg(adapter, ERROR, "unknown recv_type %#x\n", recv_type); - return -1; + ret = -1; + goto exit_restore_skb; } break; case MWIFIEX_USB_EP_DATA: -- 2.25.1

1 0

[PATCH openEuler-5.10 01/87] kvm: SMM: fix losing SMI problem
by Zheng Zengkai 30 Dec '21

30 Dec '21

From: xulei <stone.xulei(a)huawei.com> virt inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I4NP0K CVE: NA ------------------- fix losing SMI problem Signed-off-by: xulei <stone.xulei(a)huawei.com> Signed-off-by: Jingyi Wang <wangjingyi11(a)huawei.com> Reviewed-by: Zenghui Yu <yuzenghui(a)huawei.com> Reviewed-by: Wei Li <liwei391(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> --- arch/x86/kvm/x86.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index 12db47c8bd3f..e33414f36dba 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -4227,6 +4227,11 @@ static int kvm_vcpu_ioctl_nmi(struct kvm_vcpu *vcpu) static int kvm_vcpu_ioctl_smi(struct kvm_vcpu *vcpu) { + if (is_smm(vcpu)) { + vcpu->arch.hflags &= ~HF_SMM_MASK; + vcpu->arch.smi_pending = 0; + } + kvm_make_request(KVM_REQ_SMI, vcpu); return 0; -- 2.20.1

1 86

[PATCH openEuler-1.0-LTS] mm/page_alloc: Use cmdline to disable "place pages to tail"
by Yang Yingliang 30 Dec '21

30 Dec '21

From: Peng Liu <liupeng256(a)huawei.com> hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I4EG0R CVE: NA ----------------------------------------------- Add cmdline to disable "place pages to tail" when online memory. Signed-off-by: Peng Liu <liupeng256(a)huawei.com> Reviewed-by: Kefeng Wang <wangkefeng.wang(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- .../admin-guide/kernel-parameters.txt | 9 +++++++ mm/page_alloc.c | 27 ++++++++++++++----- 2 files changed, 30 insertions(+), 6 deletions(-) diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt index 425015359b5ce..420f7317aa6af 100644 --- a/Documentation/admin-guide/kernel-parameters.txt +++ b/Documentation/admin-guide/kernel-parameters.txt @@ -1254,6 +1254,15 @@ Warning: use of this parameter will taint the kernel and may cause unknown problems. + fpi_to_tail=off + [KNL] Place pages to front in __free_pages_core(). + This kernel start-up parameter can be just used as + "fpi_to_tail=off", which means memory is online to + the front of freelist. Dynamic open is not supportted, + in other words, "fpi_to_tail=on" will be ignored by + kernel. This is just an ugly solution for circumvention + for some latent bugs revealed by "place pages to tail". + ftrace=[tracer] [FTRACE] will set and start the specified tracer as early as possible in order to facilitate early diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 935435e03cdd4..4cad86f1e3a91 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -102,6 +102,17 @@ typedef int __bitwise fpi_t; */ #define FPI_TO_TAIL ((__force fpi_t)BIT(1)) +static bool fpi_to_tail = true; + +static int __init early_fpi_to_tail(char *str) +{ + if (str && !strcmp(str, "off")) + fpi_to_tail = false; + + return 0; +} +early_param("fpi_to_tail", early_fpi_to_tail); + /* prevent >1 _updater_ of zone percpu pageset ->high and ->batch fields */ static DEFINE_MUTEX(pcp_batch_high_lock); #define MIN_PERCPU_PAGELIST_FRACTION (8) @@ -1342,12 +1353,16 @@ void __free_pages_core(struct page *page, unsigned int order) } __ClearPageReserved(p); set_page_count(p, 0); - - /* - * Bypass PCP and place fresh pages right to the tail, primarily - * relevant for memory onlining. - */ - __free_pages_ok(page, order, FPI_TO_TAIL); + if (fpi_to_tail) { + /* + * Bypass PCP and place fresh pages right to the tail, primarily + * relevant for memory onlining. + */ + __free_pages_ok(page, order, FPI_TO_TAIL); + } else { + set_page_refcounted(page); + __free_pages(page, order); + } } #if defined(CONFIG_HAVE_ARCH_EARLY_PFN_TO_NID) || \ -- 2.25.1

1 0

[PATCH OLK-5.10 0/3] Add three modifications
by Yanling Song 30 Dec '21

30 Dec '21

Add three modifications. 1.RSS initialization process optimization 2.Attribute negotiation and optimization. 3.The reset command flags modification. Yanling Song (3): net/spnic:RSS initialization process optimization net/spnic:Attribute negotiation and optimization. net/spnic:The reset command flags modification. .../net/ethernet/ramaxel/spnic/hw/sphw_hw.h | 28 ++++----- .../ethernet/ramaxel/spnic/hw/sphw_hwdev.c | 9 ++- .../ethernet/ramaxel/spnic/hw/sphw_hwdev.h | 2 + .../net/ethernet/ramaxel/spnic/spnic_main.c | 22 +++---- .../ramaxel/spnic/spnic_mgmt_interface.h | 12 ---- .../ethernet/ramaxel/spnic/spnic_nic_cfg.c | 57 ++++++++++++------- .../ethernet/ramaxel/spnic/spnic_nic_cfg.h | 19 +------ .../ethernet/ramaxel/spnic/spnic_nic_dev.h | 2 + .../net/ethernet/ramaxel/spnic/spnic_rss.c | 16 +----- .../ethernet/ramaxel/spnic/spnic_rss_cfg.c | 57 ------------------- 10 files changed, 70 insertions(+), 154 deletions(-) -- 2.23.0

1 3

[PATCH openEuler-1.0-LTS] scsi: ufs: Correct the LUN used in eh_device_reset_handler() callback
by Yang Yingliang 30 Dec '21

30 Dec '21

From: Can Guo <cang(a)codeaurora.org> mainline inclusion from mainline-v5.11-rc4 commit 35fc4cd34426c242ab015ef280853b7bff101f48 category: bugfix bugzilla: NA CVE: CVE-2021-39657 ----------------------------------------------- Users can initiate resets to specific SCSI device/target/host through IOCTL. When this happens, the SCSI cmd passed to eh_device/target/host _reset_handler() callbacks is initialized with a request whose tag is -1. In this case it is not right for eh_device_reset_handler() callback to count on the LUN get from hba->lrb[-1]. Fix it by getting LUN from the SCSI device associated with the SCSI cmd. Link: https://lore.kernel.org/r/1609157080-26283-1-git-send-email-cang@codeaurora… Reviewed-by: Avri Altman <avri.altman(a)wdc.com> Reviewed-by: Stanley Chu <stanley.chu(a)mediatek.com> Signed-off-by: Can Guo <cang(a)codeaurora.org> Signed-off-by: Martin K. Petersen <martin.petersen(a)oracle.com> Signed-off-by: Ye Bin <yebin10(a)huawei.com> Reviewed-by: Xiu Jianfeng <xiujianfeng(a)huawei.com> Reviewed-by: Jason Yan <yanaijie(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/scsi/ufs/ufshcd.c | 11 ++++------- 1 file changed, 4 insertions(+), 7 deletions(-) diff --git a/drivers/scsi/ufs/ufshcd.c b/drivers/scsi/ufs/ufshcd.c index 3601e770da162..3d9bfa5207689 100644 --- a/drivers/scsi/ufs/ufshcd.c +++ b/drivers/scsi/ufs/ufshcd.c @@ -5752,19 +5752,16 @@ static int ufshcd_eh_device_reset_handler(struct scsi_cmnd *cmd) { struct Scsi_Host *host; struct ufs_hba *hba; - unsigned int tag; u32 pos; int err; - u8 resp = 0xF; - struct ufshcd_lrb *lrbp; + u8 resp = 0xF, lun; unsigned long flags; host = cmd->device->host; hba = shost_priv(host); - tag = cmd->request->tag; - lrbp = &hba->lrb[tag]; - err = ufshcd_issue_tm_cmd(hba, lrbp->lun, 0, UFS_LOGICAL_RESET, &resp); + lun = ufshcd_scsi_to_upiu_lun(cmd->device->lun); + err = ufshcd_issue_tm_cmd(hba, lun, 0, UFS_LOGICAL_RESET, &resp); if (err || resp != UPIU_TASK_MANAGEMENT_FUNC_COMPL) { if (!err) err = resp; @@ -5773,7 +5770,7 @@ static int ufshcd_eh_device_reset_handler(struct scsi_cmnd *cmd) /* clear the commands that were pending for corresponding LUN */ for_each_set_bit(pos, &hba->outstanding_reqs, hba->nutrs) { - if (hba->lrb[pos].lun == lrbp->lun) { + if (hba->lrb[pos].lun == lun) { err = ufshcd_clear_cmd(hba, pos); if (err) break; -- 2.25.1

1 0

[PATCH openEuler-1.0-LTS] ext4: fix an use-after-free issue about data=journal writeback mode
by Yang Yingliang 30 Dec '21

30 Dec '21

From: Zhang Yi <yi.zhang(a)huawei.com> hulk inclusion category: bugfix bugzilla: 185944, https://gitee.com/openeuler/kernel/issues/I4OP92 CVE: NA --------------------------- Our syzkaller report an use-after-free issue that accessing the freed buffer_head on the writeback page in __ext4_journalled_writepage(). The problem is that if there was a truncate racing with the data=journalled writeback procedure, the writeback length could become zero and bget_one() refuse to get buffer_head's refcount, then the truncate procedure release buffer once we drop page lock, finally, the last ext4_walk_page_buffers() trigger the use-after-free problem. sync truncate ext4_sync_file() file_write_and_wait_range() ext4_setattr(0) inode->i_size = 0 ext4_writepage() len = 0 __ext4_journalled_writepage() page_bufs = page_buffers(page) ext4_walk_page_buffers(bget_one) <- does not get refcount do_invalidatepage() free_buffer_head() ext4_walk_page_buffers(page_bufs) <- trigger use-after-free After commit bdf96838aea6 ("ext4: fix race between truncate and __ext4_journalled_writepage()"), we have already handled the racing case, so the bget_one() and bput_one() are not needed. So this patch simply remove these hunk, and recheck the i_size to make it safe. Fixes: bdf96838aea6 ("ext4: fix race between truncate and __ext4_journalled_writepage()") Signed-off-by: Zhang Yi <yi.zhang(a)huawei.com> Reviewed-by: yangerkun <yangerkun(a)huawei.com> Reviewed-by: Jason Yan <yanaijie(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- fs/ext4/inode.c | 35 ++++++++++------------------------- 1 file changed, 10 insertions(+), 25 deletions(-) diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 6213816056cd6..e106a9317b429 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -1962,28 +1962,16 @@ int ext4_da_get_block_prep(struct inode *inode, sector_t iblock, return 0; } -static int bget_one(handle_t *handle, struct buffer_head *bh) -{ - get_bh(bh); - return 0; -} - -static int bput_one(handle_t *handle, struct buffer_head *bh) -{ - put_bh(bh); - return 0; -} - static int __ext4_journalled_writepage(struct page *page, unsigned int len) { struct address_space *mapping = page->mapping; struct inode *inode = mapping->host; - struct buffer_head *page_bufs = NULL; handle_t *handle = NULL; int ret = 0, err = 0; int inline_data = ext4_has_inline_data(inode); struct buffer_head *inode_bh = NULL; + loff_t size; ClearPageChecked(page); @@ -1993,14 +1981,6 @@ static int __ext4_journalled_writepage(struct page *page, inode_bh = ext4_journalled_write_inline_data(inode, len, page); if (inode_bh == NULL) goto out; - } else { - page_bufs = page_buffers(page); - if (!page_bufs) { - BUG(); - goto out; - } - ext4_walk_page_buffers(handle, page_bufs, 0, len, - NULL, bget_one); } /* * We need to release the page lock before we start the @@ -2021,7 +2001,8 @@ static int __ext4_journalled_writepage(struct page *page, lock_page(page); put_page(page); - if (page->mapping != mapping) { + size = i_size_read(inode); + if (page->mapping != mapping || page_offset(page) > size) { /* The page got truncated from under us */ ext4_journal_stop(handle); ret = 0; @@ -2031,6 +2012,13 @@ static int __ext4_journalled_writepage(struct page *page, if (inline_data) { ret = ext4_mark_inode_dirty(handle, inode); } else { + struct buffer_head *page_bufs = page_buffers(page); + + if (page->index == size >> PAGE_SHIFT) + len = size & ~PAGE_MASK; + else + len = PAGE_SIZE; + ret = ext4_walk_page_buffers(handle, page_bufs, 0, len, NULL, do_journal_get_write_access); @@ -2048,9 +2036,6 @@ static int __ext4_journalled_writepage(struct page *page, out: unlock_page(page); out_no_pagelock: - if (!inline_data && page_bufs) - ext4_walk_page_buffers(NULL, page_bufs, 0, len, - NULL, bput_one); brelse(inode_bh); return ret; } -- 2.25.1

1 0

[PATCH openEuler-1.0-LTS] netdevsim: Zero-initialize memory for new map's value in function nsim_bpf_map_alloc
by Yang Yingliang 30 Dec '21

30 Dec '21

From: Haimin Zhang <tcs.kernel(a)gmail.com> mainline inclusion from mainline-5.16-rc6 commit 481221775d53d6215a6e5e9ce1cce6d2b4ab9a46 category: bugfix CVE: CVE-2021-4135 -------------------------------- Zero-initialize memory for new map's value in function nsim_bpf_map_alloc since it may cause a potential kernel information leak issue, as follows: 1. nsim_bpf_map_alloc calls nsim_map_alloc_elem to allocate elements for a new map. 2. nsim_map_alloc_elem uses kmalloc to allocate map's value, but doesn't zero it. 3. A user application can use IOCTL BPF_MAP_LOOKUP_ELEM to get specific element's information in the map. 4. The kernel function map_lookup_elem will call bpf_map_copy_value to get the information allocated at step-2, then use copy_to_user to copy to the user buffer. This can only leak information for an array map. Fixes: 395cacb5f1a0 ("netdevsim: bpf: support fake map offload") Suggested-by: Jakub Kicinski <kuba(a)kernel.org> Acked-by: Jakub Kicinski <kuba(a)kernel.org> Signed-off-by: Haimin Zhang <tcs.kernel(a)gmail.com> Link: https://lore.kernel.org/r/20211215111530.72103-1-tcs.kernel@gmail.com Signed-off-by: Jakub Kicinski <kuba(a)kernel.org> Signed-off-by: Zhengchao Shao <shaozhengchao(a)huawei.com> Reviewed-by: Wei Yongjun <weiyongjun1(a)huawei.com> Reviewed-by: Yue Haibing <yuehaibing(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/net/netdevsim/bpf.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/netdevsim/bpf.c b/drivers/net/netdevsim/bpf.c index 81444208b2162..12f100392ed11 100644 --- a/drivers/net/netdevsim/bpf.c +++ b/drivers/net/netdevsim/bpf.c @@ -493,6 +493,7 @@ nsim_bpf_map_alloc(struct netdevsim *ns, struct bpf_offloaded_map *offmap) goto err_free; key = nmap->entry[i].key; *key = i; + memset(nmap->entry[i].value, 0, offmap->map.value_size); } } -- 2.25.1

1 0

[PATCH openEuler-1.0-LTS 1/7] bpf: Fix up register-based shifts in interpreter to silence KUBSAN
by Yang Yingliang 30 Dec '21

30 Dec '21

From: Daniel Borkmann <daniel(a)iogearbox.net> mainline inclusion from mainline-v5.12 commit 28131e9d933339a92f78e7ab6429f4aaaa07061c category: bugfix bugzilla: NA CVE: NA ----------------------------------------------------------------------- syzbot reported a shift-out-of-bounds that KUBSAN observed in the interpreter: [...] UBSAN: shift-out-of-bounds in kernel/bpf/core.c:1420:2 shift exponent 255 is too large for 64-bit type 'long long unsigned int' CPU: 1 PID: 11097 Comm: syz-executor.4 Not tainted 5.12.0-rc2-syzkaller #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:79 [inline] dump_stack+0x141/0x1d7 lib/dump_stack.c:120 ubsan_epilogue+0xb/0x5a lib/ubsan.c:148 __ubsan_handle_shift_out_of_bounds.cold+0xb1/0x181 lib/ubsan.c:327 ___bpf_prog_run.cold+0x19/0x56c kernel/bpf/core.c:1420 __bpf_prog_run32+0x8f/0xd0 kernel/bpf/core.c:1735 bpf_dispatcher_nop_func include/linux/bpf.h:644 [inline] bpf_prog_run_pin_on_cpu include/linux/filter.h:624 [inline] bpf_prog_run_clear_cb include/linux/filter.h:755 [inline] run_filter+0x1a1/0x470 net/packet/af_packet.c:2031 packet_rcv+0x313/0x13e0 net/packet/af_packet.c:2104 dev_queue_xmit_nit+0x7c2/0xa90 net/core/dev.c:2387 xmit_one net/core/dev.c:3588 [inline] dev_hard_start_xmit+0xad/0x920 net/core/dev.c:3609 __dev_queue_xmit+0x2121/0x2e00 net/core/dev.c:4182 __bpf_tx_skb net/core/filter.c:2116 [inline] __bpf_redirect_no_mac net/core/filter.c:2141 [inline] __bpf_redirect+0x548/0xc80 net/core/filter.c:2164 ____bpf_clone_redirect net/core/filter.c:2448 [inline] bpf_clone_redirect+0x2ae/0x420 net/core/filter.c:2420 ___bpf_prog_run+0x34e1/0x77d0 kernel/bpf/core.c:1523 __bpf_prog_run512+0x99/0xe0 kernel/bpf/core.c:1737 bpf_dispatcher_nop_func include/linux/bpf.h:644 [inline] bpf_test_run+0x3ed/0xc50 net/bpf/test_run.c:50 bpf_prog_test_run_skb+0xabc/0x1c50 net/bpf/test_run.c:582 bpf_prog_test_run kernel/bpf/syscall.c:3127 [inline] __do_sys_bpf+0x1ea9/0x4f00 kernel/bpf/syscall.c:4406 do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46 entry_SYSCALL_64_after_hwframe+0x44/0xae [...] Generally speaking, KUBSAN reports from the kernel should be fixed. However, in case of BPF, this particular report caused concerns since the large shift is not wrong from BPF point of view, just undefined. In the verifier, K-based shifts that are >= {64,32} (depending on the bitwidth of the instruction) are already rejected. The register-based cases were not given their content might not be known at verification time. Ideas such as verifier instruction rewrite with an additional AND instruction for the source register were brought up, but regularly rejected due to the additional runtime overhead they incur. As Edward Cree rightly put it: Shifts by more than insn bitness are legal in the BPF ISA; they are implementation-defined behaviour [of the underlying architecture], rather than UB, and have been made legal for performance reasons. Each of the JIT backends compiles the BPF shift operations to machine instructions which produce implementation-defined results in such a case; the resulting contents of the register may be arbitrary but program behaviour as a whole remains defined. Guard checks in the fast path (i.e. affecting JITted code) will thus not be accepted. The case of division by zero is not truly analogous here, as division instructions on many of the JIT-targeted architectures will raise a machine exception / fault on division by zero, whereas (to the best of my knowledge) none will do so on an out-of-bounds shift. Given the KUBSAN report only affects the BPF interpreter, but not JITs, one solution is to add the ANDs with 63 or 31 into ___bpf_prog_run(). That would make the shifts defined, and thus shuts up KUBSAN, and the compiler would optimize out the AND on any CPU that interprets the shift amounts modulo the width anyway (e.g., confirmed from disassembly that on x86-64 and arm64 the generated interpreter code is the same before and after this fix). The BPF interpreter is slow path, and most likely compiled out anyway as distros select BPF_JIT_ALWAYS_ON to avoid speculative execution of BPF instructions by the interpreter. Given the main argument was to avoid sacrificing performance, the fact that the AND is optimized away from compiler for mainstream archs helps as well as a solution moving forward. Also add a comment on LSH/RSH/ARSH translation for JIT authors to provide guidance when they see the ___bpf_prog_run() interpreter code and use it as a model for a new JIT backend. Reported-by: syzbot+bed360704c521841c85d(a)syzkaller.appspotmail.com Reported-by: Kurt Manucredo <fuzzybritches0(a)gmail.com> Signed-off-by: Eric Biggers <ebiggers(a)kernel.org> Co-developed-by: Eric Biggers <ebiggers(a)kernel.org> Signed-off-by: Daniel Borkmann <daniel(a)iogearbox.net> Acked-by: Alexei Starovoitov <ast(a)kernel.org> Acked-by: Andrii Nakryiko <andrii(a)kernel.org> Tested-by: syzbot+bed360704c521841c85d(a)syzkaller.appspotmail.com Cc: Edward Cree <ecree.xilinx(a)gmail.com> Link: https://lore.kernel.org/bpf/0000000000008f912605bd30d5d7@google.com Link: https://lore.kernel.org/bpf/bac16d8d-c174-bdc4-91bd-bfa62b410190@gmail.com conflicts: kernel/bpf/core.c Signed-off-by: He Fengqing <hefengqing(a)huawei.com> Reviewed-by: Kuohai Xu <xukuohai(a)huawei.com> Reviewed-by: Xiu Jianfeng <xiujianfeng(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- kernel/bpf/core.c | 59 +++++++++++++++++++++++++++++++++-------------- 1 file changed, 42 insertions(+), 17 deletions(-) diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c index 641622b65de44..625edbc8360de 100644 --- a/kernel/bpf/core.c +++ b/kernel/bpf/core.c @@ -1100,29 +1100,54 @@ static u64 ___bpf_prog_run(u64 *regs, const struct bpf_insn *insn, u64 *stack) select_insn: goto *jumptable[insn->code]; - /* ALU */ -#define ALU(OPCODE, OP) \ - ALU64_##OPCODE##_X: \ - DST = DST OP SRC; \ - CONT; \ - ALU_##OPCODE##_X: \ - DST = (u32) DST OP (u32) SRC; \ - CONT; \ - ALU64_##OPCODE##_K: \ - DST = DST OP IMM; \ - CONT; \ - ALU_##OPCODE##_K: \ - DST = (u32) DST OP (u32) IMM; \ + /* Explicitly mask the register-based shift amounts with 63 or 31 + * to avoid undefined behavior. Normally this won't affect the + * generated code, for example, in case of native 64 bit archs such + * as x86-64 or arm64, the compiler is optimizing the AND away for + * the interpreter. In case of JITs, each of the JIT backends compiles + * the BPF shift operations to machine instructions which produce + * implementation-defined results in such a case; the resulting + * contents of the register may be arbitrary, but program behaviour + * as a whole remains defined. In other words, in case of JIT backends, + * the AND must /not/ be added to the emitted LSH/RSH/ARSH translation. + */ + /* ALU (shifts) */ +#define SHT(OPCODE, OP) \ + ALU64_##OPCODE##_X: \ + DST = DST OP (SRC & 63); \ + CONT; \ + ALU_##OPCODE##_X: \ + DST = (u32) DST OP ((u32) SRC & 31); \ + CONT; \ + ALU64_##OPCODE##_K: \ + DST = DST OP IMM; \ + CONT; \ + ALU_##OPCODE##_K: \ + DST = (u32) DST OP (u32) IMM; \ + CONT; + /* ALU (rest) */ +#define ALU(OPCODE, OP) \ + ALU64_##OPCODE##_X: \ + DST = DST OP SRC; \ + CONT; \ + ALU_##OPCODE##_X: \ + DST = (u32) DST OP (u32) SRC; \ + CONT; \ + ALU64_##OPCODE##_K: \ + DST = DST OP IMM; \ + CONT; \ + ALU_##OPCODE##_K: \ + DST = (u32) DST OP (u32) IMM; \ CONT; - ALU(ADD, +) ALU(SUB, -) ALU(AND, &) ALU(OR, |) - ALU(LSH, <<) - ALU(RSH, >>) ALU(XOR, ^) ALU(MUL, *) + SHT(LSH, <<) + SHT(RSH, >>) +#undef SHT #undef ALU ALU_NEG: DST = (u32) -DST; @@ -1147,7 +1172,7 @@ static u64 ___bpf_prog_run(u64 *regs, const struct bpf_insn *insn, u64 *stack) insn++; CONT; ALU64_ARSH_X: - (*(s64 *) &DST) >>= SRC; + (*(s64 *) &DST) >>= (SRC & 63); CONT; ALU64_ARSH_K: (*(s64 *) &DST) >>= IMM; -- 2.25.1

1 6

[PATCH openEuler-1.0-LTS 00/14] fixes CVEs for xen
by Yang Yingliang 30 Dec '21

30 Dec '21

patch #1 ~ #9 prepare for fix CVEs patch #10 ~ #14 fixes: CVE-2021-28711 CVE-2021-28712 CVE-2021-28713 CVE-2021-28714 CVE-2021-28715 Juergen Gross (14): xen/netback: avoid race in xenvif_rx_ring_slots_available() xen: sync include/xen/interface/io/ring.h with Xen's newest version xen/blkfront: read response from backend only once xen/blkfront: don't take local copy of a request from the ring page xen/blkfront: don't trust the backend response data blindly xen/netfront: read response from backend only once xen/netfront: don't read data from request on the ring page xen/netfront: disentangle tx_skb_freelist xen/netfront: don't trust the backend response data blindly xen/blkfront: harden blkfront against event channel storms xen/netfront: harden netfront against event channel storms xen/console: harden hvc_xen against event channel storms xen/netback: fix rx queue stall detection xen/netback: don't queue unlimited number of packages drivers/block/xen-blkfront.c | 141 ++++++++---- drivers/net/xen-netback/common.h | 1 + drivers/net/xen-netback/rx.c | 70 ++++-- drivers/net/xen-netfront.c | 372 ++++++++++++++++++++----------- drivers/tty/hvc/hvc_xen.c | 30 ++- include/xen/interface/io/ring.h | 293 +++++++++++++----------- 6 files changed, 575 insertions(+), 332 deletions(-) -- 2.25.1

1 14

[PATCH 1.0-LTS ] Add new module parameters:time out
by Yanling Song 29 Dec '21

29 Dec '21

Ramaxel inclusion category: features bugzilla: https://gitee.com/openeuler/kernel/issues/I4ON8F CVE: NA Changes: 1. Split scmd_tmout_nonpt into two parameters: scmd_tmout_vd/scmd_tmout_rawdisk 2. Return -ETIME instead of -EINVAL when command is timeout. 3. Add one module parameters: max_io_force. Signed-off-by: Yanling Song <songyl(a)ramaxel.com> Reviewed-by: Jiang Yu<yujiang(a)ramaxel.com> --- drivers/scsi/spraid/spraid_main.c | 55 ++++++++++++++----------------- 1 file changed, 24 insertions(+), 31 deletions(-) diff --git a/drivers/scsi/spraid/spraid_main.c b/drivers/scsi/spraid/spraid_main.c index 519b39f44e91..df15b1cea426 100644 --- a/drivers/scsi/spraid/spraid_main.c +++ b/drivers/scsi/spraid/spraid_main.c @@ -42,10 +42,17 @@ static u32 admin_tmout = 60; module_param(admin_tmout, uint, 0644); MODULE_PARM_DESC(admin_tmout, "admin commands timeout (seconds)"); -static u32 scmd_tmout_nonpt = 180; -module_param(scmd_tmout_nonpt, uint, 0644); -MODULE_PARM_DESC(scmd_tmout_nonpt, - "scsi commands timeout for rawdisk&raid(seconds)"); +static u32 scmd_tmout_rawdisk = 180; +module_param(scmd_tmout_rawdisk, uint, 0644); +MODULE_PARM_DESC(scmd_tmout_rawdisk, "scsi commands timeout for rawdisk(seconds)"); + +static u32 scmd_tmout_vd = 180; +module_param(scmd_tmout_vd, uint, 0644); +MODULE_PARM_DESC(scmd_tmout_vd, "scsi commands timeout for vd(seconds)"); + +static bool max_io_force; +module_param(max_io_force, bool, 0644); +MODULE_PARM_DESC(max_io_force, "force max_hw_sectors_kb = 1024, default false(performance first)"); static int ioq_depth_set(const char *val, const struct kernel_param *kp); static const struct kernel_param_ops ioq_depth_ops = { @@ -78,7 +85,7 @@ static unsigned char log_debug_switch; module_param_cb(log_debug_switch, &log_debug_switch_ops, &log_debug_switch, 0644); MODULE_PARM_DESC(log_debug_switch, - "set log state, default non-zero for switch on"); + "set log state, default zero for switch off"); static int small_pool_num_set(const char *val, const struct kernel_param *kp) { @@ -981,32 +988,17 @@ static void spraid_slave_destroy(struct scsi_device *sdev) static int spraid_slave_configure(struct scsi_device *sdev) { - u16 idx; - unsigned int timeout = scmd_tmout_nonpt * HZ; + unsigned int timeout = scmd_tmout_rawdisk * HZ; struct spraid_dev *hdev = shost_priv(sdev->host); struct spraid_sdev_hostdata *hostdata = sdev->hostdata; u32 max_sec = sdev->host->max_sectors; if (hostdata) { - idx = hostdata->hdid - 1; - if (sdev->channel == hdev->devices[idx].channel && - sdev->id == le16_to_cpu(hdev->devices[idx].target) && - sdev->lun < hdev->devices[idx].lun) { - if (SPRAID_DEV_INFO_ATTR_PT(hdev->devices[idx].attr)) - timeout = 30 * HZ; - else - timeout = scmd_tmout_nonpt * HZ; - max_sec = le16_to_cpu(hdev->devices[idx].max_io_kb) - << 1; - } else { - dev_err(hdev->dev, - "[%s] err, sdev->channel:id:lun[%d:%d:%lld];" - "devices[%d], channel:target:lun[%d:%d:%d]\n", - __func__, sdev->channel, sdev->id, sdev->lun, - idx, hdev->devices[idx].channel, - hdev->devices[idx].target, - hdev->devices[idx].lun); - } + if (SPRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + timeout = scmd_tmout_vd * HZ; + else if (SPRAID_DEV_INFO_ATTR_RAWDISK(hostdata->attr)) + timeout = scmd_tmout_rawdisk * HZ; + max_sec = hostdata->max_io_kb << 1; } else { dev_err(hdev->dev, "[%s] err, sdev->hostdata is null\n", __func__); @@ -1018,11 +1010,12 @@ static int spraid_slave_configure(struct scsi_device *sdev) if ((max_sec == 0) || (max_sec > sdev->host->max_sectors)) max_sec = sdev->host->max_sectors; - dev_info(hdev->dev, - "[%s] sdev->channel:id:lun[%d:%d:%lld];" - " scmd_timeout[%d]s, maxsec[%d]\n", - __func__, sdev->channel, sdev->id, - sdev->lun, timeout / HZ, max_sec); + if (!max_io_force) + blk_queue_max_hw_sectors(sdev->request_queue, max_sec); + + dev_info(hdev->dev, "[%s] sdev->channel:id:lun[%d:%d:%lld];" + "scmd_timeout[%d]s, maxsec[%d]\n", __func__, sdev->channel, + sdev->id, sdev->lun, timeout / HZ, max_sec); return 0; } -- 2.27.0

1 0

[PATCH openEuler-1.0-LTS 1/2] sctp: account stream padding length for reconf chunk
by Yang Yingliang 29 Dec '21

29 Dec '21

From: Eiichi Tsukata <eiichi.tsukata(a)nutanix.com> stable inclusion from linux-4.19.213 commit c57fdeff69b152185fafabd37e6bfecfce51efda category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I4OJ3D CVE: NA -------------------------------- commit a2d859e3fc97e79d907761550dbc03ff1b36479c upstream. sctp_make_strreset_req() makes repeated calls to sctp_addto_chunk() which will automatically account for padding on each call. inreq and outreq are already 4 bytes aligned, but the payload is not and doing SCTP_PAD4(a + b) (which _sctp_make_chunk() did implicitly here) is different from SCTP_PAD4(a) + SCTP_PAD4(b) and not enough. It led to possible attempt to use more buffer than it was allocated and triggered a BUG_ON. Cc: Vlad Yasevich <vyasevich(a)gmail.com> Cc: Neil Horman <nhorman(a)tuxdriver.com> Cc: Greg KH <gregkh(a)linuxfoundation.org> Fixes: cc16f00f6529 ("sctp: add support for generating stream reconf ssn reset request chunk") Reported-by: Eiichi Tsukata <eiichi.tsukata(a)nutanix.com> Signed-off-by: Eiichi Tsukata <eiichi.tsukata(a)nutanix.com> Signed-off-by: Marcelo Ricardo Leitner <marcelo.leitner(a)gmail.com> Signed-off-by: Marcelo Ricardo Leitner <mleitner(a)redhat.com> Reviewed-by: Xin Long <lucien.xin(a)gmail.com> Link: https://lore.kernel.org/r/b97c1f8b0c7ff79ac4ed206fc2c49d3612e0850c.16341568… Signed-off-by: Jakub Kicinski <kuba(a)kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Wang Hai <wanghai38(a)huawei.com> Reviewed-by: Wei Yongjun <weiyongjun1(a)huawei.com> Reviewed-by: Yue Haibing <yuehaibing(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- net/sctp/sm_make_chunk.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/sctp/sm_make_chunk.c b/net/sctp/sm_make_chunk.c index 0789109c2d093..35e1fb708f4a7 100644 --- a/net/sctp/sm_make_chunk.c +++ b/net/sctp/sm_make_chunk.c @@ -3673,7 +3673,7 @@ struct sctp_chunk *sctp_make_strreset_req( outlen = (sizeof(outreq) + stream_len) * out; inlen = (sizeof(inreq) + stream_len) * in; - retval = sctp_make_reconf(asoc, outlen + inlen); + retval = sctp_make_reconf(asoc, SCTP_PAD4(outlen) + SCTP_PAD4(inlen)); if (!retval) return NULL; -- 2.25.1

1 1

[PATCH openEuler-5.10 01/64] arm64: Revert feature: Add memmap parameter and register pmem
by Zheng Zengkai 29 Dec '21

29 Dec '21

From: Zhuling <zhuling8(a)huawei.com> euleros inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4O31I?from=project-issue CVE: NA -------------------------------- The reserve memory of PMEM conflicts with "add memmap interface to reserved memory for mremap syscall usage", need to rollback, resubmit after adaptation. Feature related commit： 1.PMEM function commit: 94dc364f5eda10f49449ba573dc3322e1ea92280 2.PMEM feature config commit: 36d7a831e15ceb84e937122c87d01c14242dc377 Signed-off-by: Zhuling <zhuling8(a)huawei.com> Reviewed-by: Sang Yan <sangyan(a)huawei.com> Reviewed-by: Kefeng Wang <wangkefeng.wang(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> --- arch/arm64/Kconfig | 21 ------ arch/arm64/configs/openeuler_defconfig | 3 - arch/arm64/kernel/Makefile | 1 - arch/arm64/kernel/pmem.c | 35 ---------- arch/arm64/kernel/setup.c | 10 --- arch/arm64/mm/init.c | 94 -------------------------- drivers/nvdimm/Kconfig | 5 -- drivers/nvdimm/Makefile | 1 - 8 files changed, 170 deletions(-) delete mode 100644 arch/arm64/kernel/pmem.c diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig index df90a6e05ad2..2df4b310eb23 100644 --- a/arch/arm64/Kconfig +++ b/arch/arm64/Kconfig @@ -1321,27 +1321,6 @@ config RODATA_FULL_DEFAULT_ENABLED This requires the linear region to be mapped down to pages, which may adversely affect performance in some cases. -config ARM64_PMEM_RESERVE - bool "Reserve memory for persistent storage" - default n - help - Use memmap=nn[KMG]!ss[KMG](memmap=100K!0x1a0000000) reserve - memory for persistent storage. - - Say y here to enable this feature. - -config ARM64_PMEM_LEGACY_DEVICE - bool "Create persistent storage" - depends on BLK_DEV - depends on LIBNVDIMM - select ARM64_PMEM_RESERVE - help - Use reserved memory for persistent storage when the kernel - restart or update. the data in PMEM will not be lost and - can be loaded faster. - - Say y if unsure. - config ARM64_SW_TTBR0_PAN bool "Emulate Privileged Access Never using TTBR0_EL1 switching" help diff --git a/arch/arm64/configs/openeuler_defconfig b/arch/arm64/configs/openeuler_defconfig index 17bc8750bba7..b5fc851f1949 100644 --- a/arch/arm64/configs/openeuler_defconfig +++ b/arch/arm64/configs/openeuler_defconfig @@ -416,8 +416,6 @@ CONFIG_ARM64_CPU_PARK=y CONFIG_FORCE_MAX_ZONEORDER=11 CONFIG_UNMAP_KERNEL_AT_EL0=y CONFIG_RODATA_FULL_DEFAULT_ENABLED=y -CONFIG_ARM64_PMEM_RESERVE=y -CONFIG_ARM64_PMEM_LEGACY_DEVICE=y # CONFIG_ARM64_SW_TTBR0_PAN is not set CONFIG_ARM64_TAGGED_ADDR_ABI=y CONFIG_ARM64_ILP32=y @@ -6026,7 +6024,6 @@ CONFIG_ND_BTT=m CONFIG_BTT=y CONFIG_OF_PMEM=m CONFIG_NVDIMM_KEYS=y -CONFIG_PMEM_LEGACY=m CONFIG_DAX_DRIVER=y CONFIG_DAX=y CONFIG_DEV_DAX=m diff --git a/arch/arm64/kernel/Makefile b/arch/arm64/kernel/Makefile index f6153250b631..169d90f11cf5 100644 --- a/arch/arm64/kernel/Makefile +++ b/arch/arm64/kernel/Makefile @@ -68,7 +68,6 @@ obj-$(CONFIG_ARM64_PTR_AUTH) += pointer_auth.o obj-$(CONFIG_SHADOW_CALL_STACK) += scs.o obj-$(CONFIG_ARM64_MTE) += mte.o obj-$(CONFIG_MPAM) += mpam/ -obj-$(CONFIG_ARM64_PMEM_LEGACY_DEVICE) += pmem.o obj-y += vdso/ probes/ obj-$(CONFIG_COMPAT_VDSO) += vdso32/ diff --git a/arch/arm64/kernel/pmem.c b/arch/arm64/kernel/pmem.c deleted file mode 100644 index 16eaf706f671..000000000000 --- a/arch/arm64/kernel/pmem.c +++ /dev/null @@ -1,35 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -/* - * Copyright(c) 2021 Huawei Technologies Co., Ltd - * - * Derived from x86 and arm64 implement PMEM. - */ -#include <linux/platform_device.h> -#include <linux/init.h> -#include <linux/ioport.h> -#include <linux/module.h> - -static int found(struct resource *res, void *data) -{ - return 1; -} - -static int __init register_e820_pmem(void) -{ - struct platform_device *pdev; - int rc; - - rc = walk_iomem_res_desc(IORES_DESC_PERSISTENT_MEMORY_LEGACY, - IORESOURCE_MEM, 0, -1, NULL, found); - if (rc <= 0) - return 0; - - /* - * See drivers/nvdimm/e820.c for the implementation, this is - * simply here to trigger the module to load on demand. - */ - pdev = platform_device_alloc("e820_pmem", -1); - - return platform_device_add(pdev); -} -device_initcall(register_e820_pmem); diff --git a/arch/arm64/kernel/setup.c b/arch/arm64/kernel/setup.c index 88ec49504001..7cd042536d3b 100644 --- a/arch/arm64/kernel/setup.c +++ b/arch/arm64/kernel/setup.c @@ -70,10 +70,6 @@ static int __init arm64_enable_cpu0_hotplug(char *str) __setup("arm64_cpu0_hotplug", arm64_enable_cpu0_hotplug); #endif -#ifdef CONFIG_ARM64_PMEM_RESERVE -extern struct resource pmem_res; -#endif - phys_addr_t __fdt_pointer __initdata; /* @@ -288,12 +284,6 @@ static void __init request_standard_resources(void) request_resource(res, &pin_memory_resource); #endif } - -#ifdef CONFIG_ARM64_PMEM_RESERVE - if (pmem_res.end && pmem_res.start) - request_resource(&iomem_resource, &pmem_res); -#endif - } static int __init reserve_memblock_reserved_regions(void) diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c index 3b9401ee9c58..e8d446164c76 100644 --- a/arch/arm64/mm/init.c +++ b/arch/arm64/mm/init.c @@ -55,7 +55,6 @@ */ s64 memstart_addr __ro_after_init = -1; EXPORT_SYMBOL(memstart_addr); -phys_addr_t start_at, mem_size; #ifdef CONFIG_PIN_MEMORY struct resource pin_memory_resource = { @@ -112,18 +111,6 @@ static void __init reserve_pin_memory_res(void) */ phys_addr_t arm64_dma_phys_limit __ro_after_init; -static unsigned long long pmem_size, pmem_start; - -#ifdef CONFIG_ARM64_PMEM_RESERVE -struct resource pmem_res = { - .name = "Persistent Memory (legacy)", - .start = 0, - .end = 0, - .flags = IORESOURCE_MEM, - .desc = IORES_DESC_PERSISTENT_MEMORY_LEGACY -}; -#endif - #ifndef CONFIG_KEXEC_CORE static void __init reserve_crashkernel(void) { @@ -417,83 +404,6 @@ static int __init reserve_park_mem(void) } #endif -static bool __init is_mem_valid(unsigned long long mem_size, unsigned long long mem_start) -{ - if (!memblock_is_region_memory(mem_start, mem_size)) { - pr_warn("cannot reserve mem: region is not memory!\n"); - return false; - } - - if (memblock_is_region_reserved(mem_start, mem_size)) { - pr_warn("cannot reserve mem: region overlaps reserved memory!\n"); - return false; - } - - if (!IS_ALIGNED(mem_start, SZ_2M)) { - pr_warn("cannot reserve mem: base address is not 2MB aligned!\n"); - return false; - } - - return true; -} - -static int __init parse_memmap_one(char *p) -{ - char *oldp; - - if (!p) - return -EINVAL; - - oldp = p; - mem_size = memparse(p, &p); - if (p == oldp) - return -EINVAL; - - if (!mem_size) - return -EINVAL; - - mem_size = PAGE_ALIGN(mem_size); - - if (*p == '!') { - start_at = memparse(p+1, &p); - - pmem_start = start_at; - pmem_size = mem_size; - } else - pr_info("Unrecognized memmap option, please check the parameter.\n"); - - return *p == '\0' ? 0 : -EINVAL; -} - -static int __init parse_memmap_opt(char *str) -{ - while (str) { - char *k = strchr(str, ','); - - if (k) - *k++ = 0; - parse_memmap_one(str); - str = k; - } - - return 0; -} -early_param("memmap", parse_memmap_opt); - -#ifdef CONFIG_ARM64_PMEM_RESERVE -static void __init reserve_pmem(void) -{ - if (!is_mem_valid(mem_size, start_at)) - return; - - memblock_remove(pmem_start, pmem_size); - pr_info("pmem reserved: 0x%016llx - 0x%016llx (%lld MB)\n", - pmem_start, pmem_start + pmem_size, pmem_size >> 20); - pmem_res.start = pmem_start; - pmem_res.end = pmem_start + pmem_size - 1; -} -#endif - void __init arm64_memblock_init(void) { const s64 linear_region_size = BIT(vabits_actual - 1); @@ -668,10 +578,6 @@ void __init bootmem_init(void) reserve_quick_kexec(); #endif -#ifdef CONFIG_ARM64_PMEM_RESERVE - reserve_pmem(); -#endif - reserve_pin_memory_res(); memblock_dump_all(); diff --git a/drivers/nvdimm/Kconfig b/drivers/nvdimm/Kconfig index ce4de75262b9..b7d1eb38b27d 100644 --- a/drivers/nvdimm/Kconfig +++ b/drivers/nvdimm/Kconfig @@ -132,8 +132,3 @@ config NVDIMM_TEST_BUILD infrastructure. endif - -config PMEM_LEGACY - tristate "Pmem_legacy" - select X86_PMEM_LEGACY if X86 - select ARM64_PMEM_LEGACY_DEVICE if ARM64 diff --git a/drivers/nvdimm/Makefile b/drivers/nvdimm/Makefile index 6f8dc9242a81..04077532f7ed 100644 --- a/drivers/nvdimm/Makefile +++ b/drivers/nvdimm/Makefile @@ -3,7 +3,6 @@ obj-$(CONFIG_LIBNVDIMM) += libnvdimm.o obj-$(CONFIG_BLK_DEV_PMEM) += nd_pmem.o obj-$(CONFIG_ND_BTT) += nd_btt.o obj-$(CONFIG_ND_BLK) += nd_blk.o -obj-$(CONFIG_PMEM_LEGACY) += nd_e820.o obj-$(CONFIG_OF_PMEM) += of_pmem.o obj-$(CONFIG_VIRTIO_PMEM) += virtio_pmem.o nd_virtio.o -- 2.20.1

1 63

[PATCH openEuler-1.0-LTS 00/93] Enabling AMD Milan series processor support
by Laibin Qiu 28 Dec '21

28 Dec '21

bugzilla: https://gitee.com/openeuler/kernel/issues/I4MKP4 Babu Moger (1): KVM: SVM: Clear the CR4 register on reset David Edmondson (1): KVM: x86: clflushopt should be treated as a no-op by emulation Fenghua Yu (2): x86/cpufeatures: Enumerate MOVDIRI instruction x86/cpufeatures: Enumerate MOVDIR64B instruction Haiyan Song (2): perf vendor events intel: Add Icelake V1.00 event file perf vendor events intel: Add Tremontx event file v1.02 Isaac Vaughn (1): EDAC/amd64: Add PCI device IDs for family 17h, model 70h Jan H. Schönherr (1): x86/mce: Fix use of uninitialized MCE message string Jim Mattson (1): kvm: x86: Expose RDPID in KVM_GET_SUPPORTED_CPUID John Allen (2): kvm/svm: PKU not currently supported x86/microcode/AMD: Increase microcode PATCH_MAX_SIZE John Garry (4): perf jevents: Add support for Hisi hip08 DDRC PMU aliasing perf jevents: Add support for Hisi hip08 HHA PMU aliasing perf jevents: Add support for Hisi hip08 L3C PMU aliasing perf vendor events arm64: Fix Hisi hip08 DDRC PMU eventname Kai Huang (3): kvm: x86: Move kvm_set_mmio_spte_mask() from x86.c to mmu.c kvm: x86: Fix reserved bits related calculation errors caused by MKTME kvm: x86: Fix L1TF mitigation for shadow MMU Kan Liang (1): perf vendor events intel: Add uncore_upi JSON support Kim Phillips (17): perf/amd/uncore: Prepare L3 thread mask code for Family 19h perf/amd/uncore: Make L3 thread mask code more readable perf/amd/uncore: Add support for Family 19h L3 PMU arch/x86/amd/ibs: Fix re-arming IBS Fetch perf/x86/amd/ibs: Fix raw sample data accumulation perf/amd/uncore: Set all slices and threads to restore perf stat -a behaviour perf/x86/amd/ibs: Don't include randomized bits in get_ibs_op_count() perf/amd/uncore: Prepare to scale for more attributes that vary per family perf/amd/uncore: Allow F17h user threadmask and slicemask specification perf/amd/uncore: Allow F19h user coreid, threadmask, and sliceid specification perf vendor events amd: Add L3 cache events for Family 17h perf vendor events amd: Remove redundant '[' perf vendor events amd: Enable Family 19h users by matching Zen2 events x86/cpu/amd: Call init_amd_zn() om Family 19h processors too perf/x86/amd/ibs: Support 27-bit extended Op/cycle counter tools/power turbostat: Support AMD Family 19h perf vendor events amd: Add L2 Prefetch events for zen1 Krish Sadhukhan (1): KVM: SVM: Replace hard-coded value with #define Like Xu (1): perf/x86/amd: Don't touch the AMD64_EVENTSEL_HOSTONLY bit inside the guest Liu Jingqi (2): KVM: x86: expose MOVDIRI CPU feature into VM. KVM: x86: expose MOVDIR64B CPU feature into VM. Maciej S. Szmigiero (1): KVM: mmu: Fix SPTE encoding of MMIO generation upper half Marcel Bocu (1): x86/amd_nb: Add PCI device IDs for family 17h, model 70h Martin Liška (1): perf vendor events amd: perf PMU events for AMD Family 17h Nathan Chancellor (2): crypto: ccp - Remove forward declaration perf/amd/uncore: Fix sysfs type mismatch Paolo Bonzini (3): KVM: x86: only do L1TF workaround on affected processors KVM: x86: assign two bits to track SPTE kinds KVM: x86: fix overlap between SPTE_MMIO_MASK and generation Rasmus Villemoes (1): build_bug.h: add wrapper for _Static_assert Sean Christopherson (13): KVM: nVMX: Allocate and configure VM{READ,WRITE} bitmaps iff enable_shadow_vmcs KVM: x86: Add requisite includes to kvm_cache_regs.h KVM: x86: Add requisite includes to hyperv.h KVM: x86: Use a u64 when passing the MMIO gen around KVM: Explicitly define the "memslot update in-progress" bit KVM: x86: Refactor the MMIO SPTE generation handling KVM: x86: Rename access permissions cache member in struct kvm_vcpu_arch KVM: x86/mmu: Add explicit access mask for MMIO SPTEs KVM: x86/mmu: Consolidate "is MMIO SPTE" code KVM: x86/mmu: Apply max PA check for MMIO sptes to 32-bit KVM KVM: x86/mmu: Set mmio_value to '0' if reserved #PF can't be generated KVM: Remove the hack to trigger memslot generation wraparound KVM: Move the memslot update in-progress flag to bit 63 Sebastian Andrzej Siewior (1): x86/pkeys: Don't check if PKRU is zero before writing it Tom Lendacky (1): KVM: SVM: Override default MMIO mask if memory encryption is enabled Vijay Thakkar (3): perf vendor events amd: Restrict model detection for zen1 based processors perf vendor events amd: Add Zen2 events perf vendor events amd: Update Zen1 events to V2 Woods, Brian (2): hwmon/k10temp, x86/amd_nb: Consolidate shared device IDs x86/amd_nb: Add PCI device IDs for family 17h, model 30h Yazen Ghannam (24): EDAC/amd64: Drop some family checks for newer systems x86/amd_nb: Add Family 19h PCI IDs EDAC/mce_amd: Always load on SMCA systems x86/MCE/AMD, EDAC/mce_amd: Add new MP5, NBIO, and PCIE SMCA bank types x86/MCE/AMD, EDAC/mce_amd: Add new McaTypes for CS, PSP, and SMU units x86/MCE/AMD, EDAC/mce_amd: Add new error descriptions for some SMCA bank types x86/MCE/AMD, EDAC/mce_amd: Add new Load Store unit McaType EDAC/amd64: Use a macro for iterating over Unified Memory Controllers EDAC/amd64: Support more than two controllers for chip selects handling EDAC/amd64: Initialize DIMM info for systems with more than two channels EDAC/amd64: Add Family 17h Model 30h PCI IDs EDAC/amd64: Support more than two Unified Memory Controllers EDAC/amd64: Set maximum channel layer size depending on family EDAC/amd64: Recognize x16 symbol size EDAC/amd64: Adjust printed chip select sizes when interleaved EDAC/amd64: Find Chip Select memory size using Address Mask EDAC/amd64: Cache secondary Chip Select registers EDAC/amd64: Support asymmetric dual-rank DIMMs EDAC/amd64: Set grain per DIMM EDAC/amd64: Make struct amd64_family_type global EDAC/amd64: Gather hardware information early EDAC/amd64: Save max number of controllers to family type EDAC/amd64: Add family ops for Family 19h Models 00h-0Fh EDAC/amd64: Handle three rank interleaving mode Documentation/virtual/kvm/mmu.txt | 13 +- arch/x86/events/amd/ibs.c | 93 +- arch/x86/events/amd/uncore.c | 179 ++-- arch/x86/events/perf_event.h | 3 +- arch/x86/include/asm/cpufeatures.h | 4 +- arch/x86/include/asm/kvm_host.h | 10 +- arch/x86/include/asm/mce.h | 7 + arch/x86/include/asm/microcode_amd.h | 2 +- arch/x86/include/asm/msr-index.h | 1 + arch/x86/include/asm/perf_event.h | 16 +- arch/x86/kernel/amd_nb.c | 15 +- arch/x86/kernel/cpu/amd.c | 3 +- arch/x86/kernel/cpu/mce/amd.c | 28 +- arch/x86/kernel/cpu/mce/core.c | 4 +- arch/x86/kvm/cpuid.c | 6 +- arch/x86/kvm/emulate.c | 8 +- arch/x86/kvm/hyperv.h | 2 + arch/x86/kvm/kvm_cache_regs.h | 2 + arch/x86/kvm/mmu.c | 215 +++-- arch/x86/kvm/mmu.h | 2 +- arch/x86/kvm/svm.c | 53 +- arch/x86/kvm/vmx.c | 51 +- arch/x86/kvm/x86.c | 33 +- arch/x86/kvm/x86.h | 4 +- arch/x86/mm/pkeys.c | 7 - drivers/crypto/ccp/sp-platform.c | 53 +- drivers/edac/amd64_edac.c | 609 ++++++++---- drivers/edac/amd64_edac.h | 31 +- drivers/edac/mce_amd.c | 134 ++- drivers/hwmon/k10temp.c | 9 +- include/linux/build_bug.h | 19 + include/linux/kvm_host.h | 21 + include/linux/pci_ids.h | 5 + .../arm64/hisilicon/hip08/uncore-ddrc.json | 44 + .../arm64/hisilicon/hip08/uncore-hha.json | 51 + .../arm64/hisilicon/hip08/uncore-l3c.json | 37 + .../pmu-events/arch/x86/amdzen1/branch.json | 23 + .../pmu-events/arch/x86/amdzen1/cache.json | 312 ++++++ .../pmu-events/arch/x86/amdzen1/core.json | 125 +++ .../arch/x86/amdzen1/floating-point.json | 224 +++++ .../pmu-events/arch/x86/amdzen1/memory.json | 184 ++++ .../pmu-events/arch/x86/amdzen1/other.json | 56 ++ .../pmu-events/arch/x86/amdzen2/branch.json | 52 + .../pmu-events/arch/x86/amdzen2/cache.json | 338 +++++++ .../pmu-events/arch/x86/amdzen2/core.json | 130 +++ .../arch/x86/amdzen2/floating-point.json | 140 +++ .../pmu-events/arch/x86/amdzen2/memory.json | 341 +++++++ .../pmu-events/arch/x86/amdzen2/other.json | 115 +++ .../pmu-events/arch/x86/icelake/cache.json | 552 +++++++++++ .../arch/x86/icelake/floating-point.json | 102 ++ .../pmu-events/arch/x86/icelake/frontend.json | 424 +++++++++ .../pmu-events/arch/x86/icelake/memory.json | 410 ++++++++ .../pmu-events/arch/x86/icelake/other.json | 121 +++ .../pmu-events/arch/x86/icelake/pipeline.json | 892 ++++++++++++++++++ .../arch/x86/icelake/virtual-memory.json | 236 +++++ tools/perf/pmu-events/arch/x86/mapfile.csv | 6 + .../pmu-events/arch/x86/tremontx/cache.json | 111 +++ .../arch/x86/tremontx/frontend.json | 26 + .../pmu-events/arch/x86/tremontx/memory.json | 26 + .../pmu-events/arch/x86/tremontx/other.json | 26 + .../arch/x86/tremontx/pipeline.json | 111 +++ .../arch/x86/tremontx/uncore-memory.json | 73 ++ .../arch/x86/tremontx/uncore-other.json | 431 +++++++++ .../arch/x86/tremontx/uncore-power.json | 11 + .../arch/x86/tremontx/virtual-memory.json | 86 ++ tools/perf/pmu-events/jevents.c | 5 + tools/power/x86/turbostat/turbostat.c | 34 +- virt/kvm/kvm_main.c | 36 +- 68 files changed, 6981 insertions(+), 552 deletions(-) create mode 100644 tools/perf/pmu-events/arch/arm64/hisilicon/hip08/uncore-ddrc.json create mode 100644 tools/perf/pmu-events/arch/arm64/hisilicon/hip08/uncore-hha.json create mode 100644 tools/perf/pmu-events/arch/arm64/hisilicon/hip08/uncore-l3c.json create mode 100644 tools/perf/pmu-events/arch/x86/amdzen1/branch.json create mode 100644 tools/perf/pmu-events/arch/x86/amdzen1/cache.json create mode 100644 tools/perf/pmu-events/arch/x86/amdzen1/core.json create mode 100644 tools/perf/pmu-events/arch/x86/amdzen1/floating-point.json create mode 100644 tools/perf/pmu-events/arch/x86/amdzen1/memory.json create mode 100644 tools/perf/pmu-events/arch/x86/amdzen1/other.json create mode 100644 tools/perf/pmu-events/arch/x86/amdzen2/branch.json create mode 100644 tools/perf/pmu-events/arch/x86/amdzen2/cache.json create mode 100644 tools/perf/pmu-events/arch/x86/amdzen2/core.json create mode 100644 tools/perf/pmu-events/arch/x86/amdzen2/floating-point.json create mode 100644 tools/perf/pmu-events/arch/x86/amdzen2/memory.json create mode 100644 tools/perf/pmu-events/arch/x86/amdzen2/other.json create mode 100644 tools/perf/pmu-events/arch/x86/icelake/cache.json create mode 100644 tools/perf/pmu-events/arch/x86/icelake/floating-point.json create mode 100644 tools/perf/pmu-events/arch/x86/icelake/frontend.json create mode 100644 tools/perf/pmu-events/arch/x86/icelake/memory.json create mode 100644 tools/perf/pmu-events/arch/x86/icelake/other.json create mode 100644 tools/perf/pmu-events/arch/x86/icelake/pipeline.json create mode 100644 tools/perf/pmu-events/arch/x86/icelake/virtual-memory.json create mode 100644 tools/perf/pmu-events/arch/x86/tremontx/cache.json create mode 100644 tools/perf/pmu-events/arch/x86/tremontx/frontend.json create mode 100644 tools/perf/pmu-events/arch/x86/tremontx/memory.json create mode 100644 tools/perf/pmu-events/arch/x86/tremontx/other.json create mode 100644 tools/perf/pmu-events/arch/x86/tremontx/pipeline.json create mode 100644 tools/perf/pmu-events/arch/x86/tremontx/uncore-memory.json create mode 100644 tools/perf/pmu-events/arch/x86/tremontx/uncore-other.json create mode 100644 tools/perf/pmu-events/arch/x86/tremontx/uncore-power.json create mode 100644 tools/perf/pmu-events/arch/x86/tremontx/virtual-memory.json -- 2.22.0

1 93

[PATCH openEuler-1.0-LTS] netfilter: fix regression in looped (broad|multi)cast's MAC handling
by Yang Yingliang 28 Dec '21

28 Dec '21

From: Ignacy Gawędzki <ignacy.gawedzki(a)green-communications.fr> mainline inclusion from mainline-v5.16-rc7 commit ebb966d3bdfed581ecccbb4a7432341baf7619b4 category: bugfix bugzilla: NA CVE: NA -------------------------------- In commit 5648b5e1169f ("netfilter: nfnetlink_queue: fix OOB when mac header was cleared"), the test for non-empty MAC header introduced in commit 2c38de4c1f8da7 ("netfilter: fix looped (broad|multi)cast's MAC handling") has been replaced with a test for a set MAC header. This breaks the case when the MAC header has been reset (using skb_reset_mac_header), as is the case with looped-back multicast packets. As a result, the packets ending up in NFQUEUE get a bogus hwaddr interpreted from the first bytes of the IP header. This patch adds a test for a non-empty MAC header in addition to the test for a set MAC header. The same two tests are also implemented in nfnetlink_log.c, where the initial code of commit 2c38de4c1f8da7 ("netfilter: fix looped (broad|multi)cast's MAC handling") has not been touched, but where supposedly the same situation may happen. Fixes: 5648b5e1169f ("netfilter: nfnetlink_queue: fix OOB when mac header was cleared") Signed-off-by: Ignacy Gawędzki <ignacy.gawedzki(a)green-communications.fr> Reviewed-by: Florian Westphal <fw(a)strlen.de> Signed-off-by: Pablo Neira Ayuso <pablo(a)netfilter.org> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> Reviewed-by: Yue Haibing <yuehaibing(a)huawei.com> Reviewed-by: Wei Yongjun <weiyongjun1(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- net/netfilter/nfnetlink_log.c | 3 ++- net/netfilter/nfnetlink_queue.c | 3 ++- 2 files changed, 4 insertions(+), 2 deletions(-) diff --git a/net/netfilter/nfnetlink_log.c b/net/netfilter/nfnetlink_log.c index 25298b3eb8546..17ca9a681d47b 100644 --- a/net/netfilter/nfnetlink_log.c +++ b/net/netfilter/nfnetlink_log.c @@ -509,7 +509,8 @@ __build_packet_message(struct nfnl_log_net *log, goto nla_put_failure; if (indev && skb->dev && - skb->mac_header != skb->network_header) { + skb_mac_header_was_set(skb) && + skb_mac_header_len(skb) != 0) { struct nfulnl_msg_packet_hw phw; int len; diff --git a/net/netfilter/nfnetlink_queue.c b/net/netfilter/nfnetlink_queue.c index eb5a052d3b252..8955431f2ab26 100644 --- a/net/netfilter/nfnetlink_queue.c +++ b/net/netfilter/nfnetlink_queue.c @@ -566,7 +566,8 @@ nfqnl_build_packet_message(struct net *net, struct nfqnl_instance *queue, goto nla_put_failure; if (indev && entskb->dev && - skb_mac_header_was_set(entskb)) { + skb_mac_header_was_set(entskb) && + skb_mac_header_len(entskb) != 0) { struct nfqnl_msg_packet_hw phw; int len; -- 2.25.1

1 0

[PATCH openEuler-5.10 v2] Disable legacy DRM drivers
by Liu Zixian 28 Dec '21

28 Dec '21

euleros inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4OBDE CVE: NA ------------------------------------------------- This config enables dangerous old drivers. It is selected by CONFIG_NOUVEAU_LEGACY_CTX_SUPPORT which is suggested by upstream community that "modern distros should consider turning it off". CONFIG_DRM_VM was selected by CONFIG_DRM_LEGACY and can be turned off now. Other distros has disabled this config: https://src.fedoraproject.org/rpms/kernel/blob/rawhide/f/kernel-aarch64-rhe… Signed-off-by: Liu Zixian <liuzixian4(a)huawei.com> --- arch/arm64/configs/openeuler_defconfig | 10 ++-------- arch/x86/configs/openeuler_defconfig | 10 ++-------- 2 files changed, 4 insertions(+), 16 deletions(-) diff --git a/arch/arm64/configs/openeuler_defconfig b/arch/arm64/configs/openeuler_defconfig index 17bc8750b..6949824e8 100644 --- a/arch/arm64/configs/openeuler_defconfig +++ b/arch/arm64/configs/openeuler_defconfig @@ -4673,7 +4673,6 @@ CONFIG_DRM_TTM_DMA_PAGE_POOL=y CONFIG_DRM_VRAM_HELPER=y CONFIG_DRM_TTM_HELPER=y CONFIG_DRM_GEM_SHMEM_HELPER=y -CONFIG_DRM_VM=y CONFIG_DRM_SCHED=m # @@ -4719,7 +4718,7 @@ CONFIG_DRM_AMD_DC_DCN=y # CONFIG_HSA_AMD is not set CONFIG_DRM_NOUVEAU=m -CONFIG_NOUVEAU_LEGACY_CTX_SUPPORT=y +# CONFIG_NOUVEAU_LEGACY_CTX_SUPPORT is not set CONFIG_NOUVEAU_DEBUG=5 CONFIG_NOUVEAU_DEBUG_DEFAULT=3 # CONFIG_NOUVEAU_DEBUG_MMU is not set @@ -4817,12 +4816,7 @@ CONFIG_DRM_CIRRUS_QEMU=m # CONFIG_DRM_LIMA is not set # CONFIG_DRM_PANFROST is not set # CONFIG_DRM_TIDSS is not set -CONFIG_DRM_LEGACY=y -# CONFIG_DRM_TDFX is not set -# CONFIG_DRM_R128 is not set -# CONFIG_DRM_MGA is not set -# CONFIG_DRM_VIA is not set -# CONFIG_DRM_SAVAGE is not set +# CONFIG_DRM_LEGACY is not set CONFIG_DRM_PANEL_ORIENTATION_QUIRKS=y # diff --git a/arch/x86/configs/openeuler_defconfig b/arch/x86/configs/openeuler_defconfig index 7b6083018..dbae04231 100644 --- a/arch/x86/configs/openeuler_defconfig +++ b/arch/x86/configs/openeuler_defconfig @@ -5040,7 +5040,6 @@ CONFIG_DRM_TTM_DMA_PAGE_POOL=y CONFIG_DRM_VRAM_HELPER=m CONFIG_DRM_TTM_HELPER=m CONFIG_DRM_GEM_SHMEM_HELPER=y -CONFIG_DRM_VM=y CONFIG_DRM_SCHED=m # @@ -5085,7 +5084,7 @@ CONFIG_DRM_AMD_DC_DCN=y # CONFIG_HSA_AMD is not set CONFIG_DRM_NOUVEAU=m -CONFIG_NOUVEAU_LEGACY_CTX_SUPPORT=y +# CONFIG_NOUVEAU_LEGACY_CTX_SUPPORT is not set CONFIG_NOUVEAU_DEBUG=5 CONFIG_NOUVEAU_DEBUG_DEFAULT=3 # CONFIG_NOUVEAU_DEBUG_MMU is not set @@ -5226,12 +5225,7 @@ CONFIG_DRM_CIRRUS_QEMU=m # CONFIG_TINYDRM_ST7735R is not set # CONFIG_DRM_XEN is not set # CONFIG_DRM_VBOXVIDEO is not set -CONFIG_DRM_LEGACY=y -# CONFIG_DRM_TDFX is not set -# CONFIG_DRM_R128 is not set -# CONFIG_DRM_MGA is not set -# CONFIG_DRM_VIA is not set -# CONFIG_DRM_SAVAGE is not set +# CONFIG_DRM_LEGACY is not set CONFIG_DRM_PANEL_ORIENTATION_QUIRKS=y # -- 2.27.0

1 0

【邀请函】诚邀您参与2021年华为生态健康度调研
by Ecosurvey 28 Dec '21

28 Dec '21

尊敬的女士/先生：您好！感谢您选择与华为合作及在双方合作中给予的支持！为了更好地实现共同成长、共同发展，我们目前在开展“2021年华为生态健康度调研”。为此，我们委托独立的第三方调研机构广州尼尔森市场研究有限公司（以下简称尼尔森），来开展相关调研，以了解您对华为生态系统的评价和建议，并分析改进。本次调研内容仅供项目研究使用，对外会严格保密。问卷答题时长约5-10分钟。参与答题请点击问卷链接： https://csurveys.nielseniq.cn/wix/p1611532.aspx 您对本次调查有任何疑问或需要任何帮助，可以通过Ecosurvey(a)huawei.xn--com,-os0g9z677h027f 为工作人员将会解答您的问题。感谢您的参与！ 2021华为生态健康度调研项目组

1 0

[PATCH openEuler-5.10 RESEND] Disable legacy DRM drivers
by Liu Zixian 28 Dec '21

28 Dec '21

From: sdlzx <hdu_sdlzx(a)163.com> redhat inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4OBDE CVE: NA ------------------------------------------------- This config enables dangerous old drivers. It is selected by CONFIG_NOUVEAU_LEGACY_CTX_SUPPORT which is suggested by upstream community that "modern distros should consider turning it off". CONFIG_DRM_VM was selected by CONFIG_DRM_LEGACY and can be turned off now. Other distros has disabled this config: https://src.fedoraproject.org/rpms/kernel/blob/rawhide/f/kernel-aarch64-rhe… Signed-off-by: sdlzx <hdu_sdlzx(a)163.com> --- arch/arm64/configs/openeuler_defconfig | 10 ++-------- arch/x86/configs/openeuler_defconfig | 10 ++-------- 2 files changed, 4 insertions(+), 16 deletions(-) diff --git a/arch/arm64/configs/openeuler_defconfig b/arch/arm64/configs/openeuler_defconfig index 17bc8750b..6949824e8 100644 --- a/arch/arm64/configs/openeuler_defconfig +++ b/arch/arm64/configs/openeuler_defconfig @@ -4673,7 +4673,6 @@ CONFIG_DRM_TTM_DMA_PAGE_POOL=y CONFIG_DRM_VRAM_HELPER=y CONFIG_DRM_TTM_HELPER=y CONFIG_DRM_GEM_SHMEM_HELPER=y -CONFIG_DRM_VM=y CONFIG_DRM_SCHED=m # @@ -4719,7 +4718,7 @@ CONFIG_DRM_AMD_DC_DCN=y # CONFIG_HSA_AMD is not set CONFIG_DRM_NOUVEAU=m -CONFIG_NOUVEAU_LEGACY_CTX_SUPPORT=y +# CONFIG_NOUVEAU_LEGACY_CTX_SUPPORT is not set CONFIG_NOUVEAU_DEBUG=5 CONFIG_NOUVEAU_DEBUG_DEFAULT=3 # CONFIG_NOUVEAU_DEBUG_MMU is not set @@ -4817,12 +4816,7 @@ CONFIG_DRM_CIRRUS_QEMU=m # CONFIG_DRM_LIMA is not set # CONFIG_DRM_PANFROST is not set # CONFIG_DRM_TIDSS is not set -CONFIG_DRM_LEGACY=y -# CONFIG_DRM_TDFX is not set -# CONFIG_DRM_R128 is not set -# CONFIG_DRM_MGA is not set -# CONFIG_DRM_VIA is not set -# CONFIG_DRM_SAVAGE is not set +# CONFIG_DRM_LEGACY is not set CONFIG_DRM_PANEL_ORIENTATION_QUIRKS=y # diff --git a/arch/x86/configs/openeuler_defconfig b/arch/x86/configs/openeuler_defconfig index 7b6083018..dbae04231 100644 --- a/arch/x86/configs/openeuler_defconfig +++ b/arch/x86/configs/openeuler_defconfig @@ -5040,7 +5040,6 @@ CONFIG_DRM_TTM_DMA_PAGE_POOL=y CONFIG_DRM_VRAM_HELPER=m CONFIG_DRM_TTM_HELPER=m CONFIG_DRM_GEM_SHMEM_HELPER=y -CONFIG_DRM_VM=y CONFIG_DRM_SCHED=m # @@ -5085,7 +5084,7 @@ CONFIG_DRM_AMD_DC_DCN=y # CONFIG_HSA_AMD is not set CONFIG_DRM_NOUVEAU=m -CONFIG_NOUVEAU_LEGACY_CTX_SUPPORT=y +# CONFIG_NOUVEAU_LEGACY_CTX_SUPPORT is not set CONFIG_NOUVEAU_DEBUG=5 CONFIG_NOUVEAU_DEBUG_DEFAULT=3 # CONFIG_NOUVEAU_DEBUG_MMU is not set @@ -5226,12 +5225,7 @@ CONFIG_DRM_CIRRUS_QEMU=m # CONFIG_TINYDRM_ST7735R is not set # CONFIG_DRM_XEN is not set # CONFIG_DRM_VBOXVIDEO is not set -CONFIG_DRM_LEGACY=y -# CONFIG_DRM_TDFX is not set -# CONFIG_DRM_R128 is not set -# CONFIG_DRM_MGA is not set -# CONFIG_DRM_VIA is not set -# CONFIG_DRM_SAVAGE is not set +# CONFIG_DRM_LEGACY is not set CONFIG_DRM_PANEL_ORIENTATION_QUIRKS=y # -- 2.27.0

2 1

【邀请函】诚邀您参与2021年华为生态健康度调研
by Ecosurvey 28 Dec '21

28 Dec '21

尊敬的女士/先生：您好！感谢您选择与华为合作及在双方合作中给予的支持！为了更好地实现共同成长、共同发展，我们目前在开展“2021年华为生态健康度调研”。为此，我们委托独立的第三方调研机构广州尼尔森市场研究有限公司（以下简称尼尔森），来开展相关调研，以了解您对华为生态系统的评价和建议，并分析改进。本次调研内容仅供项目研究使用，对外会严格保密。问卷答题时长约5-10分钟。参与答题请点击问卷链接： https://csurveys.nielseniq.cn/wix/p1611532.aspx 您对本次调查有任何疑问或需要任何帮助，可以通过Ecosurvey(a)huawei.xn--com,-os0g9z677h027f 为工作人员将会解答您的问题。感谢您的参与！ 2021华为生态健康度调研项目组

1 0

[PATCH OLK-5.10 00/13] Introduce AESNI/AVX and AESNI/AVX2 accelerated implementation for SM4 algorithm
by shenzijun 28 Dec '21

28 Dec '21

From: 沈子俊 <shenzijun(a)kylinos.cn> This patchset support test GCM/CCM mode for SM4. The GCM/CCM mode of SM4 is defined in the RFC 8998 specification: https://datatracker.ietf.org/doc/html/rfc8998 This patchset extracts the public SM4 algorithm as a separate library, At the same time, the acceleration implementation of SM4 in arm64 was adjusted to adapt to this SM4 library. Then introduces an accelerated implementation of the AESNI/AVX and AESNI/AVX2 on x86_64. The AESNI/AVX2 implementation reuses the function of AESNI/AVX. The optimization supports the four modes of SM4, ECB, CBC, CFB, and CTR. Since CBC and CFB do not support multiple block parallel encryption, the optimization effect is not obvious. And all selftests have passed already. The main algorithm implementation comes from SM4 AES-NI work by libgcrypt and Markku-Juhani O. Saarinen at: https://github.com/mjosaarinen/sm4ni Finally, add the configuration for accelerated of SM4. And introduce some bugfix about AESNI/AVX2 accelerated implementation. Benchmark on Intel i5-6200U 2.30GHz, performance data of three implementation methods, pure software sm4-generic, aesni/avx acceleration, and aesni/avx2 acceleration, the data comes from the 218 mode and 518 mode of tcrypt. The abscissas are blocks of different lengths. The data is tabulated and the unit is Mb/s: block-size | 16 64 128 256 1024 1420 4096 sm4-generic ECB enc | 60.94 70.41 72.27 73.02 73.87 73.58 73.59 ECB dec | 61.87 70.53 72.15 73.09 73.89 73.92 73.86 CBC enc | 56.71 66.31 68.05 69.84 70.02 70.12 70.24 CBC dec | 54.54 65.91 68.22 69.51 70.63 70.79 70.82 CFB enc | 57.21 67.24 69.10 70.25 70.73 70.52 71.42 CFB dec | 57.22 64.74 66.31 67.24 67.40 67.64 67.58 CTR enc | 59.47 68.64 69.91 71.02 71.86 71.61 71.95 CTR dec | 59.94 68.77 69.95 71.00 71.84 71.55 71.95 sm4-aesni-avx ECB enc | 44.95 177.35 292.06 316.98 339.48 322.27 330.59 ECB dec | 45.28 178.66 292.31 317.52 339.59 322.52 331.16 CBC enc | 57.75 67.68 69.72 70.60 71.48 71.63 71.74 CBC dec | 44.32 176.83 284.32 307.24 328.61 312.61 325.82 CFB enc | 57.81 67.64 69.63 70.55 71.40 71.35 71.70 CFB dec | 43.14 167.78 282.03 307.20 328.35 318.24 325.95 CTR enc | 42.35 163.32 279.11 302.93 320.86 310.56 317.93 CTR dec | 42.39 162.81 278.49 302.37 321.11 310.33 318.37 sm4-aesni-avx2 ECB enc | 45.19 177.41 292.42 316.12 339.90 322.53 330.54 ECB dec | 44.83 178.90 291.45 317.31 339.85 322.55 331.07 CBC enc | 57.66 67.62 69.73 70.55 71.58 71.66 71.77 CBC dec | 44.34 176.86 286.10 501.68 559.58 483.87 527.46 CFB enc | 57.43 67.60 69.61 70.52 71.43 71.28 71.65 CFB dec | 43.12 167.75 268.09 499.33 558.35 490.36 524.73 CTR enc | 42.42 163.39 256.17 493.95 552.45 481.58 517.19 CTR dec | 42.49 163.11 256.36 493.34 552.62 481.49 516.83 From the benchmark data, it can be seen that when the block size is 1024, compared to AVX acceleration, the performance achieved by AVX2 has increased by about 70%, it is also 7.7 times of the pure software implementation of sm4-generic. 沈子俊 (13): crypto: tcrypt - Fix missing return value check crypto: testmgr - Add GCM/CCM mode test of SM4 algorithm crypto: tcrypt - add GCM/CCM mode test for SM4 algorithm crypto: sm4 - create SM4 library based on sm4 generic code crypto: arm64/sm4-ce - Make dependent on sm4 library instead of sm4-generic crypto: x86/sm4 - add AES-NI/AVX/x86_64 implementation crypto: tcrypt - add the asynchronous speed test for SM4 crypto: x86/sm4 - export reusable AESNI/AVX functions crypto: x86/sm4 - add AES-NI/AVX2/x86_64 implementation Add the configuration for accelerated of SM4 crypto: x86/sm4 - Fix frame pointer stack corruption crypto: sm4 - Do not change section of ck and sbox crypto: x86/sm4 - Fix invalid section entry size arch/arm64/crypto/Kconfig | 2 +- arch/arm64/crypto/sm4-ce-glue.c | 20 +- arch/x86/configs/openeuler_defconfig | 2 + arch/x86/crypto/Makefile | 6 + arch/x86/crypto/sm4-aesni-avx-asm_64.S | 594 ++++++++++++++++++++++++ arch/x86/crypto/sm4-aesni-avx2-asm_64.S | 501 ++++++++++++++++++++ arch/x86/crypto/sm4-avx.h | 24 + arch/x86/crypto/sm4_aesni_avx2_glue.c | 169 +++++++ arch/x86/crypto/sm4_aesni_avx_glue.c | 487 +++++++++++++++++++ crypto/Kconfig | 44 ++ crypto/sm4_generic.c | 180 +------ crypto/tcrypt.c | 99 +++- crypto/testmgr.c | 29 ++ crypto/testmgr.h | 148 ++++++ include/crypto/sm4.h | 25 +- lib/crypto/Kconfig | 3 + lib/crypto/Makefile | 3 + lib/crypto/sm4.c | 176 +++++++ 18 files changed, 2325 insertions(+), 187 deletions(-) create mode 100644 arch/x86/crypto/sm4-aesni-avx-asm_64.S create mode 100644 arch/x86/crypto/sm4-aesni-avx2-asm_64.S create mode 100644 arch/x86/crypto/sm4-avx.h create mode 100644 arch/x86/crypto/sm4_aesni_avx2_glue.c create mode 100644 arch/x86/crypto/sm4_aesni_avx_glue.c create mode 100644 lib/crypto/sm4.c -- 2.30.0

3 16

[PATCH OLK5.10 v3] scsi:spraid: use bsg module to replace with ioctrl
by Yanling Song 28 Dec '21

28 Dec '21

Ramaxel inclusion category: features bugzilla: https://gitee.com/openeuler/kernel/issues/I4JXCG CVE: NA Changes from v2: 1. Split scmd_tmout_nonpt into two parameters: scmd_tmout_vd/scmd_tmout_rawdisk 2. Return -ETIME instead of -EINVAL when command is timeout. 3. Add one module parameters: max_io_force. Changes from v1: 1. Add more debug infor. 2. Remove some unnecessary module parameters. 3. Report disks by the order of channel/target id. 4. Add host_reset handler. 5. Use get_unaligned_be24. Signed-off-by: Yanling Song <songyl(a)ramaxel.com> Reviewed-by: Jiang Yu<yujiang(a)ramaxel.com> --- drivers/scsi/spraid/Kconfig | 6 +- drivers/scsi/spraid/spraid.h | 142 ++- drivers/scsi/spraid/spraid_main.c | 1633 +++++++++++++++-------------- 3 files changed, 968 insertions(+), 813 deletions(-) diff --git a/drivers/scsi/spraid/Kconfig b/drivers/scsi/spraid/Kconfig index 83962efaab07..bfbba3db8db0 100644 --- a/drivers/scsi/spraid/Kconfig +++ b/drivers/scsi/spraid/Kconfig @@ -5,7 +5,9 @@ config RAMAXEL_SPRAID tristate "Ramaxel spraid Adapter" depends on PCI && SCSI + select BLK_DEV_BSGLIB depends on ARM64 || X86_64 - default m help - This driver supports Ramaxel spraid driver. + This driver supports Ramaxel SPRxxx serial + raid controller, which has PCIE Gen4 interface + with host and supports SAS/SATA Hdd/ssd. diff --git a/drivers/scsi/spraid/spraid.h b/drivers/scsi/spraid/spraid.h index da46d8e1b4b6..983d7af2faa8 100644 --- a/drivers/scsi/spraid/spraid.h +++ b/drivers/scsi/spraid/spraid.h @@ -1,4 +1,5 @@ /* SPDX-License-Identifier: GPL-2.0 */ +/* Copyright(c) 2021 Ramaxel Memory Technology, Ltd */ #ifndef __SPRAID_H_ #define __SPRAID_H_ @@ -24,7 +25,7 @@ #define SENSE_SIZE(depth) ((depth) * SCSI_SENSE_BUFFERSIZE) #define SPRAID_AQ_DEPTH 128 -#define SPRAID_NR_AEN_COMMANDS 1 +#define SPRAID_NR_AEN_COMMANDS 16 #define SPRAID_AQ_BLK_MQ_DEPTH (SPRAID_AQ_DEPTH - SPRAID_NR_AEN_COMMANDS) #define SPRAID_AQ_MQ_TAG_DEPTH (SPRAID_AQ_BLK_MQ_DEPTH - 1) @@ -44,7 +45,7 @@ #define SMALL_POOL_SIZE 256 #define MAX_SMALL_POOL_NUM 16 -#define MAX_CMD_PER_DEV 32 +#define MAX_CMD_PER_DEV 64 #define MAX_CDB_LEN 32 #define SPRAID_UP_TO_MULTY4(x) (((x) + 4) & (~0x03)) @@ -53,7 +54,7 @@ #define PCI_VENDOR_ID_RAMAXEL_LOGIC 0x1E81 -#define SPRAID_SERVER_DEVICE_HAB_DID 0x2100 +#define SPRAID_SERVER_DEVICE_HBA_DID 0x2100 #define SPRAID_SERVER_DEVICE_RAID_DID 0x2200 #define IO_6_DEFAULT_TX_LEN 256 @@ -142,11 +143,15 @@ enum { enum { SPRAID_AEN_DEV_CHANGED = 0x00, + SPRAID_AEN_FW_ACT_START = 0x01, SPRAID_AEN_HOST_PROBING = 0x10, }; enum { - SPRAID_AEN_TIMESYN = 0x07 + SPRAID_AEN_TIMESYN = 0x00, + SPRAID_AEN_FW_ACT_FINISH = 0x02, + SPRAID_AEN_EVENT_MIN = 0x80, + SPRAID_AEN_EVENT_MAX = 0xff, }; enum { @@ -175,6 +180,16 @@ enum spraid_state { SPRAID_DEAD, }; +enum { + SPRAID_CARD_HBA, + SPRAID_CARD_RAID, +}; + +enum spraid_cmd_type { + SPRAID_CMD_ADM, + SPRAID_CMD_IOPT, +}; + struct spraid_completion { __le32 result; union { @@ -217,8 +232,6 @@ struct spraid_dev { struct dma_pool *prp_page_pool; struct dma_pool *prp_small_pool[MAX_SMALL_POOL_NUM]; mempool_t *iod_mempool; - struct blk_mq_tag_set admin_tagset; - struct request_queue *admin_q; void __iomem *bar; u32 max_qid; u32 num_vecs; @@ -232,23 +245,27 @@ struct spraid_dev { u32 ctrl_config; u32 online_queues; u64 cap; - struct device ctrl_device; - struct cdev cdev; int instance; struct spraid_ctrl_info *ctrl_info; struct spraid_dev_info *devices; - struct spraid_ioq_ptcmd *ioq_ptcmds; + struct spraid_cmd *adm_cmds; + struct list_head adm_cmd_list; + spinlock_t adm_cmd_lock; + + struct spraid_cmd *ioq_ptcmds; struct list_head ioq_pt_list; spinlock_t ioq_pt_lock; - struct work_struct aen_work; struct work_struct scan_work; struct work_struct timesyn_work; struct work_struct reset_work; + struct work_struct fw_act_work; enum spraid_state state; spinlock_t state_lock; + + struct request_queue *bsg_queue; }; struct spraid_sgl_desc { @@ -347,6 +364,35 @@ struct spraid_get_info { __u32 rsvd12[4]; }; +struct spraid_usr_cmd { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + union { + struct { + __le16 subopcode; + __le16 rsvd1; + } info_0; + __le32 cdw2; + }; + union { + struct { + __le16 data_len; + __le16 param_len; + } info_1; + __le32 cdw3; + }; + __u64 metadata; + union spraid_data_ptr dptr; + __le32 cdw10; + __le32 cdw11; + __le32 cdw12; + __le32 cdw13; + __le32 cdw14; + __le32 cdw15; +}; + enum { SPRAID_CMD_FLAG_SGL_METABUF = (1 << 6), SPRAID_CMD_FLAG_SGL_METASEG = (1 << 7), @@ -393,6 +439,7 @@ struct spraid_admin_command { struct spraid_get_info get_info; struct spraid_abort_cmd abort; struct spraid_reset_cmd reset; + struct spraid_usr_cmd usr_cmd; }; }; @@ -456,9 +503,6 @@ struct spraid_ioq_command { }; }; -#define SPRAID_IOCTL_RESET_CMD _IOWR('N', 0x80, struct spraid_passthru_common_cmd) -#define SPRAID_IOCTL_ADMIN_CMD _IOWR('N', 0x41, struct spraid_passthru_common_cmd) - struct spraid_passthru_common_cmd { __u8 opcode; __u8 flags; @@ -494,8 +538,6 @@ struct spraid_passthru_common_cmd { __u32 result1; }; -#define SPRAID_IOCTL_IOQ_CMD _IOWR('N', 0x42, struct spraid_ioq_passthru_cmd) - struct spraid_ioq_passthru_cmd { __u8 opcode; __u8 flags; @@ -560,7 +602,21 @@ struct spraid_ioq_passthru_cmd { __u32 result1; }; -struct spraid_ioq_ptcmd { +struct spraid_bsg_request { + u32 msgcode; + u32 control; + union { + struct spraid_passthru_common_cmd admcmd; + struct spraid_ioq_passthru_cmd ioqcmd; + }; +}; + +enum { + SPRAID_BSG_ADM, + SPRAID_BSG_IOQ, +}; + +struct spraid_cmd { int qid; int cid; u32 result0; @@ -572,14 +628,6 @@ struct spraid_ioq_ptcmd { struct list_head list; }; -struct spraid_admin_request { - struct spraid_admin_command *cmd; - u32 result0; - u32 result1; - u16 flags; - u16 status; -}; - struct spraid_queue { struct spraid_dev *hdev; spinlock_t sq_lock; /* spinlock for lock handling */ @@ -607,7 +655,6 @@ struct spraid_queue { }; struct spraid_iod { - struct spraid_admin_request req; struct spraid_queue *spraidq; enum spraid_cmd_state state; int npages; @@ -623,13 +670,51 @@ struct spraid_iod { }; #define SPRAID_DEV_INFO_ATTR_BOOT(attr) ((attr) & 0x01) -#define SPRAID_DEV_INFO_ATTR_HDD(attr) ((attr) & 0x02) +#define SPRAID_DEV_INFO_ATTR_VD(attr) (((attr) & 0x02) == 0x0) #define SPRAID_DEV_INFO_ATTR_PT(attr) (((attr) & 0x22) == 0x02) #define SPRAID_DEV_INFO_ATTR_RAWDISK(attr) ((attr) & 0x20) #define SPRAID_DEV_INFO_FLAG_VALID(flag) ((flag) & 0x01) #define SPRAID_DEV_INFO_FLAG_CHANGE(flag) ((flag) & 0x02) +#define BGTASK_TYPE_REBUILD 4 +#define USR_CMD_READ 0xc2 +#define USR_CMD_RDLEN 0x1000 +#define USR_CMD_VDINFO 0x704 +#define USR_CMD_BGTASK 0x504 +#define VDINFO_PARAM_LEN 0x04 + +struct spraid_vd_info { + __u8 name[32]; + __le16 id; + __u8 rg_id; + __u8 rg_level; + __u8 sg_num; + __u8 sg_disk_num; + __u8 vd_status; + __u8 vd_type; + __u8 rsvd1[4056]; +}; + +#define MAX_REALTIME_BGTASK_NUM 32 + +struct bgtask_info { + __u8 type; + __u8 progress; + __u8 rate; + __u8 rsvd0; + __le16 vd_id; + __le16 time_left; + __u8 rsvd1[4]; +}; + +struct spraid_bgtask { + __u8 sw; + __u8 task_num; + __u8 rsvd[6]; + struct bgtask_info bgtask[MAX_REALTIME_BGTASK_NUM]; +}; + struct spraid_dev_info { __le32 hdid; __le16 target; @@ -649,6 +734,11 @@ struct spraid_dev_list { struct spraid_sdev_hostdata { u32 hdid; + u16 max_io_kb; + u8 attr; + u8 flag; + u8 rg_id; + u8 rsvd[3]; }; #endif diff --git a/drivers/scsi/spraid/spraid_main.c b/drivers/scsi/spraid/spraid_main.c index a0a75ecb0027..c6d2a0b8e35e 100644 --- a/drivers/scsi/spraid/spraid_main.c +++ b/drivers/scsi/spraid/spraid_main.c @@ -1,8 +1,8 @@ // SPDX-License-Identifier: GPL-2.0 -/* - * Linux spraid device driver - * Copyright(c) 2021 Ramaxel Memory Technology, Ltd - */ +/* Copyright(c) 2021 Ramaxel Memory Technology, Ltd */ + +/* Ramaxel Raid SPXXX Series Linux Driver */ + #define pr_fmt(fmt) "spraid: " fmt #include <linux/sched/signal.h> @@ -23,6 +23,9 @@ #include <linux/debugfs.h> #include <linux/io-64-nonatomic-lo-hi.h> #include <linux/blkdev.h> +#include <linux/bsg-lib.h> +#include <asm/unaligned.h> +#include <linux/sort.h> #include <scsi/scsi.h> #include <scsi/scsi_cmnd.h> @@ -31,27 +34,24 @@ #include <scsi/scsi_transport.h> #include <scsi/scsi_dbg.h> + #include "spraid.h" static u32 admin_tmout = 60; module_param(admin_tmout, uint, 0644); MODULE_PARM_DESC(admin_tmout, "admin commands timeout (seconds)"); -static u32 scmd_tmout_pt = 30; -module_param(scmd_tmout_pt, uint, 0644); -MODULE_PARM_DESC(scmd_tmout_pt, "scsi commands timeout for passthrough(seconds)"); +static u32 scmd_tmout_rawdisk = 180; +module_param(scmd_tmout_rawdisk, uint, 0644); +MODULE_PARM_DESC(scmd_tmout_rawdisk, "scsi commands timeout for rawdisk(seconds)"); -static u32 scmd_tmout_nonpt = 180; -module_param(scmd_tmout_nonpt, uint, 0644); -MODULE_PARM_DESC(scmd_tmout_nonpt, "scsi commands timeout for rawdisk&raid(seconds)"); +static u32 scmd_tmout_vd = 180; +module_param(scmd_tmout_vd, uint, 0644); +MODULE_PARM_DESC(scmd_tmout_vd, "scsi commands timeout for vd(seconds)"); -static u32 wait_abl_tmout = 3; -module_param(wait_abl_tmout, uint, 0644); -MODULE_PARM_DESC(wait_abl_tmout, "wait abnormal io timeout(seconds)"); - -static bool use_sgl_force; -module_param(use_sgl_force, bool, 0644); -MODULE_PARM_DESC(use_sgl_force, "force IO use sgl format, default false"); +static bool max_io_force; +module_param(max_io_force, bool, 0644); +MODULE_PARM_DESC(max_io_force, "force max_hw_sectors_kb = 1024, default false(performance first)"); static int ioq_depth_set(const char *val, const struct kernel_param *kp); static const struct kernel_param_ops ioq_depth_ops = { @@ -106,16 +106,20 @@ static const struct kernel_param_ops small_pool_num_ops = { .get = param_get_byte, }; +/* It was found that the spindlock of a single pool conflicts + * a lot with multiple CPUs.So multiple pools are introduced + * to reduce the conflictions. + */ static unsigned char small_pool_num = 4; module_param_cb(small_pool_num, &small_pool_num_ops, &small_pool_num, 0644); MODULE_PARM_DESC(small_pool_num, "set prp small pool num, default 4, MAX 16"); static void spraid_free_queue(struct spraid_queue *spraidq); static void spraid_handle_aen_notice(struct spraid_dev *hdev, u32 result); -static void spraid_handle_aen_vs(struct spraid_dev *hdev, u32 result); +static void spraid_handle_aen_vs(struct spraid_dev *hdev, u32 result, u32 result1); static DEFINE_IDA(spraid_instance_ida); -static dev_t spraid_chr_devt; + static struct class *spraid_class; #define SPRAID_CAP_TIMEOUT_UNIT_MS (HZ / 2) @@ -131,9 +135,8 @@ static struct workqueue_struct *spraid_wq; #define SPRAID_DRV_VERSION "1.0.0.0" #define ADMIN_TIMEOUT (admin_tmout * HZ) -#define ADMIN_ERR_TIMEOUT 32757 -#define SPRAID_WAIT_ABNL_CMD_TIMEOUT (wait_abl_tmout * 2) +#define SPRAID_WAIT_ABNL_CMD_TIMEOUT (3 * 2) #define SPRAID_DMA_MSK_BIT_MAX 64 @@ -147,6 +150,13 @@ enum FW_STAT_CODE { FW_STAT_NEED_RETRY }; +static const char * const raid_levels[] = {"0", "1", "5", "6", "10", "50", "60", "NA"}; + +static const char * const raid_states[] = { + "NA", "NORMAL", "FAULT", "DEGRADE", "NOT_FORMATTED", "FORMATTING", "SANITIZING", + "INITIALIZING", "INITIALIZE_FAIL", "DELETING", "DELETE_FAIL", "WRITE_PROTECT" +}; + static int ioq_depth_set(const char *val, const struct kernel_param *kp) { int n = 0; @@ -231,12 +241,6 @@ static int spraid_pci_enable(struct spraid_dev *hdev) goto disable; } - ret = pci_alloc_irq_vectors(pdev, 1, 1, PCI_IRQ_ALL_TYPES); - if (ret < 0) { - dev_err(hdev->dev, "Allocate one IRQ for setup admin channel failed\n"); - goto disable; - } - hdev->cap = lo_hi_readq(hdev->bar + SPRAID_REG_CAP); hdev->ioq_depth = min_t(u32, SPRAID_CAP_MQES(hdev->cap) + 1, io_queue_depth); hdev->db_stride = 1 << SPRAID_CAP_STRIDE(hdev->cap); @@ -246,13 +250,20 @@ static int spraid_pci_enable(struct spraid_dev *hdev) dev_err(hdev->dev, "err, dma mask invalid[%llu], set to default\n", maskbit); maskbit = SPRAID_DMA_MSK_BIT_MAX; } - if (dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(maskbit))) { + if (dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(maskbit)) && + dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(32))) { dev_err(hdev->dev, "set dma mask and coherent failed\n"); goto disable; } dev_info(hdev->dev, "set dma mask[%llu] success\n", maskbit); + ret = pci_alloc_irq_vectors(pdev, 1, 1, PCI_IRQ_ALL_TYPES); + if (ret < 0) { + dev_err(hdev->dev, "Allocate one IRQ for setup admin channel failed\n"); + goto disable; + } + pci_enable_pcie_error_reporting(pdev); pci_save_state(pdev); @@ -263,12 +274,6 @@ static int spraid_pci_enable(struct spraid_dev *hdev) return ret; } -static inline -struct spraid_admin_request *spraid_admin_req(struct request *req) -{ - return blk_mq_rq_to_pdu(req); -} - static int spraid_npages_prp(u32 size, struct spraid_dev *hdev) { u32 nprps = DIV_ROUND_UP(size + hdev->page_size, hdev->page_size); @@ -419,7 +424,7 @@ static void spraid_submit_cmd(struct spraid_queue *spraidq, const void *cmd) writel(spraidq->sq_tail, spraidq->q_db); spin_unlock_irqrestore(&spraidq->sq_lock, flags); - dev_log_dbg(spraidq->hdev->dev, "cid[%d], qid[%d], opcode[0x%x], flags[0x%x], hdid[%u]\n", + dev_log_dbg(spraidq->hdev->dev, "cid[%d] qid[%d], opcode[0x%x], flags[0x%x], hdid[%u]\n", acd->command_id, spraidq->qid, acd->opcode, acd->flags, le32_to_cpu(acd->hdid)); } @@ -605,18 +610,15 @@ static void spraid_setup_rw_cmd(struct spraid_dev *hdev, if (scmd->cmd_len == 6) { datalength = (u32)(scmd->cmnd[4] == 0 ? IO_6_DEFAULT_TX_LEN : scmd->cmnd[4]); - start_lba_lo = ((u32)scmd->cmnd[1] << 16) | - ((u32)scmd->cmnd[2] << 8) | (u32)scmd->cmnd[3]; + start_lba_lo = (u32)get_unaligned_be24(&scmd->cmnd[1]); start_lba_lo &= 0x1FFFFF; } /* 10-byte READ(0x28) or WRITE(0x2A) cdb */ else if (scmd->cmd_len == 10) { - datalength = (u32)scmd->cmnd[8] | ((u32)scmd->cmnd[7] << 8); - start_lba_lo = ((u32)scmd->cmnd[2] << 24) | - ((u32)scmd->cmnd[3] << 16) | - ((u32)scmd->cmnd[4] << 8) | (u32)scmd->cmnd[5]; + datalength = (u32)get_unaligned_be16(&scmd->cmnd[7]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[2]); if (scmd->cmnd[1] & FUA_MASK) control |= SPRAID_RW_FUA; @@ -624,42 +626,26 @@ static void spraid_setup_rw_cmd(struct spraid_dev *hdev, /* 12-byte READ(0xA8) or WRITE(0xAA) cdb */ else if (scmd->cmd_len == 12) { - datalength = ((u32)scmd->cmnd[6] << 24) | - ((u32)scmd->cmnd[7] << 16) | - ((u32)scmd->cmnd[8] << 8) | (u32)scmd->cmnd[9]; - start_lba_lo = ((u32)scmd->cmnd[2] << 24) | - ((u32)scmd->cmnd[3] << 16) | - ((u32)scmd->cmnd[4] << 8) | (u32)scmd->cmnd[5]; + datalength = get_unaligned_be32(&scmd->cmnd[6]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[2]); if (scmd->cmnd[1] & FUA_MASK) control |= SPRAID_RW_FUA; } /* 16-byte READ(0x88) or WRITE(0x8A) cdb */ else if (scmd->cmd_len == 16) { - datalength = ((u32)scmd->cmnd[10] << 24) | - ((u32)scmd->cmnd[11] << 16) | - ((u32)scmd->cmnd[12] << 8) | (u32)scmd->cmnd[13]; - start_lba_lo = ((u32)scmd->cmnd[6] << 24) | - ((u32)scmd->cmnd[7] << 16) | - ((u32)scmd->cmnd[8] << 8) | (u32)scmd->cmnd[9]; - start_lba_hi = ((u32)scmd->cmnd[2] << 24) | - ((u32)scmd->cmnd[3] << 16) | - ((u32)scmd->cmnd[4] << 8) | (u32)scmd->cmnd[5]; + datalength = get_unaligned_be32(&scmd->cmnd[10]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[6]); + start_lba_hi = get_unaligned_be32(&scmd->cmnd[2]); if (scmd->cmnd[1] & FUA_MASK) control |= SPRAID_RW_FUA; } /* 32-byte READ(0x88) or WRITE(0x8A) cdb */ else if (scmd->cmd_len == 32) { - datalength = ((u32)scmd->cmnd[28] << 24) | - ((u32)scmd->cmnd[29] << 16) | - ((u32)scmd->cmnd[30] << 8) | (u32)scmd->cmnd[31]; - start_lba_lo = ((u32)scmd->cmnd[16] << 24) | - ((u32)scmd->cmnd[17] << 16) | - ((u32)scmd->cmnd[18] << 8) | (u32)scmd->cmnd[19]; - start_lba_hi = ((u32)scmd->cmnd[12] << 24) | - ((u32)scmd->cmnd[13] << 16) | - ((u32)scmd->cmnd[14] << 8) | (u32)scmd->cmnd[15]; + datalength = get_unaligned_be32(&scmd->cmnd[28]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[16]); + start_lba_hi = get_unaligned_be32(&scmd->cmnd[12]); if (scmd->cmnd[10] & FUA_MASK) control |= SPRAID_RW_FUA; @@ -814,7 +800,7 @@ static void spraid_map_status(struct spraid_iod *iod, struct scsi_cmnd *scmd, if (scmd->result & SAM_STAT_CHECK_CONDITION) { memset(scmd->sense_buffer, 0, SCSI_SENSE_BUFFERSIZE); memcpy(scmd->sense_buffer, iod->sense, SCSI_SENSE_BUFFERSIZE); - set_driver_byte(scmd, DRIVER_SENSE); + scmd->result = (scmd->result & 0x00ffffff) | (DRIVER_SENSE << 24); } break; case FW_STAT_ABORTED: @@ -825,6 +811,8 @@ static void spraid_map_status(struct spraid_iod *iod, struct scsi_cmnd *scmd, break; default: set_host_byte(scmd, DID_BAD_TARGET); + dev_warn(iod->spraidq->hdev->dev, "[%s] cid[%d] qid[%d] bad status[0x%x]\n", + __func__, cqe->cmd_id, le16_to_cpu(cqe->sq_id), le16_to_cpu(cqe->status)); break; } } @@ -850,14 +838,13 @@ static int spraid_queue_command(struct Scsi_Host *shost, struct scsi_cmnd *scmd) int ret; if (unlikely(!scmd)) { - dev_err(hdev->dev, "err, scmd is null, return 0\n"); + dev_err(hdev->dev, "err, scmd is null\n"); return 0; } if (unlikely(hdev->state != SPRAID_LIVE)) { set_host_byte(scmd, DID_NO_CONNECT); scmd->scsi_done(scmd); - dev_err(hdev->dev, "[%s] err, hdev state is not live\n", __func__); return 0; } @@ -894,7 +881,7 @@ static int spraid_queue_command(struct Scsi_Host *shost, struct scsi_cmnd *scmd) WRITE_ONCE(iod->state, SPRAID_CMD_IN_FLIGHT); spraid_submit_cmd(ioq, &ioq_cmd); elapsed = jiffies - scmd->jiffies_at_alloc; - dev_log_dbg(hdev->dev, "cid[%d], qid[%d] submit IO cost %3ld.%3ld seconds\n", + dev_log_dbg(hdev->dev, "cid[%d] qid[%d] submit IO cost %3ld.%3ld seconds\n", cid, hwq, elapsed / HZ, elapsed % HZ); return 0; @@ -945,6 +932,10 @@ static int spraid_slave_alloc(struct scsi_device *sdev) scan_host: hostdata->hdid = le32_to_cpu(hdev->devices[idx].hdid); + hostdata->max_io_kb = le16_to_cpu(hdev->devices[idx].max_io_kb); + hostdata->attr = hdev->devices[idx].attr; + hostdata->flag = hdev->devices[idx].flag; + hostdata->rg_id = 0xff; sdev->hostdata = hostdata; up_read(&hdev->devices_rwsem); return 0; @@ -958,30 +949,17 @@ static void spraid_slave_destroy(struct scsi_device *sdev) static int spraid_slave_configure(struct scsi_device *sdev) { - u16 idx; - unsigned int timeout = scmd_tmout_nonpt * HZ; + unsigned int timeout = scmd_tmout_rawdisk * HZ; struct spraid_dev *hdev = shost_priv(sdev->host); struct spraid_sdev_hostdata *hostdata = sdev->hostdata; u32 max_sec = sdev->host->max_sectors; - if (!hostdata) { - idx = hostdata->hdid - 1; - if (sdev->channel == hdev->devices[idx].channel && - sdev->id == le16_to_cpu(hdev->devices[idx].target) && - sdev->lun < hdev->devices[idx].lun) { - if (SPRAID_DEV_INFO_ATTR_PT(hdev->devices[idx].attr)) - timeout = scmd_tmout_pt * HZ; - else - timeout = scmd_tmout_nonpt * HZ; - max_sec = le16_to_cpu(hdev->devices[idx].max_io_kb) << 1; - } else { - dev_err(hdev->dev, "[%s] err, sdev->channel:id:lun[%d:%d:%lld];" - "devices[%d], channel:target:lun[%d:%d:%d]\n", - __func__, sdev->channel, sdev->id, sdev->lun, - idx, hdev->devices[idx].channel, - hdev->devices[idx].target, - hdev->devices[idx].lun); - } + if (hostdata) { + if (SPRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + timeout = scmd_tmout_vd * HZ; + else if (SPRAID_DEV_INFO_ATTR_RAWDISK(hostdata->attr)) + timeout = scmd_tmout_rawdisk * HZ; + max_sec = hostdata->max_io_kb << 1; } else { dev_err(hdev->dev, "[%s] err, sdev->hostdata is null\n", __func__); } @@ -991,7 +969,9 @@ static int spraid_slave_configure(struct scsi_device *sdev) if ((max_sec == 0) || (max_sec > sdev->host->max_sectors)) max_sec = sdev->host->max_sectors; - blk_queue_max_hw_sectors(sdev->request_queue, max_sec); + + if (!max_io_force) + blk_queue_max_hw_sectors(sdev->request_queue, max_sec); dev_info(hdev->dev, "[%s] sdev->channel:id:lun[%d:%d:%lld], scmd_timeout[%d]s, maxsec[%d]\n", __func__, sdev->channel, sdev->id, sdev->lun, timeout / HZ, max_sec); @@ -1176,6 +1156,75 @@ static inline bool spraid_cqe_pending(struct spraid_queue *spraidq) spraidq->cq_phase; } +static void spraid_sata_report_zone_handle(struct scsi_cmnd *scmd, struct spraid_iod *iod) +{ + int i = 0; + unsigned int bytes = 0; + struct scatterlist *sg = scsi_sglist(scmd); + + scsi_for_each_sg(scmd, sg, iod->nsge, i) { + unsigned int offset = 0; + + if (bytes == 0) { + char *hdr; + u32 list_length; + u64 max_lba, opt_lba; + u16 same; + + hdr = sg_virt(sg); + + list_length = get_unaligned_le32(&hdr[0]); + same = get_unaligned_le16(&hdr[4]); + max_lba = get_unaligned_le64(&hdr[8]); + opt_lba = get_unaligned_le64(&hdr[16]); + put_unaligned_be32(list_length, &hdr[0]); + hdr[4] = same & 0xf; + put_unaligned_be64(max_lba, &hdr[8]); + put_unaligned_be64(opt_lba, &hdr[16]); + offset += 64; + bytes += 64; + } + while (offset < sg_dma_len(sg)) { + char *rec; + u8 cond, type, non_seq, reset; + u64 size, start, wp; + + rec = sg_virt(sg) + offset; + type = rec[0] & 0xf; + cond = (rec[1] >> 4) & 0xf; + non_seq = (rec[1] & 2); + reset = (rec[1] & 1); + size = get_unaligned_le64(&rec[8]); + start = get_unaligned_le64(&rec[16]); + wp = get_unaligned_le64(&rec[24]); + rec[0] = type; + rec[1] = (cond << 4) | non_seq | reset; + put_unaligned_be64(size, &rec[8]); + put_unaligned_be64(start, &rec[16]); + put_unaligned_be64(wp, &rec[24]); + WARN_ON(offset + 64 > sg_dma_len(sg)); + offset += 64; + bytes += 64; + } + } +} + +static inline void spraid_handle_ata_cmd(struct spraid_dev *hdev, struct scsi_cmnd *scmd, + struct spraid_iod *iod) +{ + if (hdev->ctrl_info->card_type != SPRAID_CARD_HBA) + return; + + switch (scmd->cmnd[0]) { + case ZBC_IN: + dev_info(hdev->dev, "[%s] process report zone\n", __func__); + spraid_sata_report_zone_handle(scmd, iod); + break; + default: + break; + } +} + static void spraid_complete_ioq_cmnd(struct spraid_queue *ioq, struct spraid_completion *cqe) { struct spraid_dev *hdev = ioq->hdev; @@ -1197,12 +1246,12 @@ static void spraid_complete_ioq_cmnd(struct spraid_queue *ioq, struct spraid_com iod = scsi_cmd_priv(scmd); elapsed = jiffies - scmd->jiffies_at_alloc; - dev_log_dbg(hdev->dev, "cid[%d], qid[%d] finish IO cost %3ld.%3ld seconds\n", + dev_log_dbg(hdev->dev, "cid[%d] qid[%d] finish IO cost %3ld.%3ld seconds\n", cqe->cmd_id, ioq->qid, elapsed / HZ, elapsed % HZ); if (cmpxchg(&iod->state, SPRAID_CMD_IN_FLIGHT, SPRAID_CMD_COMPLETE) != SPRAID_CMD_IN_FLIGHT) { - dev_warn(hdev->dev, "cid[%d], qid[%d] enters abnormal handler, cost %3ld.%3ld seconds\n", + dev_warn(hdev->dev, "cid[%d] qid[%d] enters abnormal handler, cost %3ld.%3ld seconds\n", cqe->cmd_id, ioq->qid, elapsed / HZ, elapsed % HZ); WRITE_ONCE(iod->state, SPRAID_CMD_TMO_COMPLETE); @@ -1215,6 +1264,8 @@ static void spraid_complete_ioq_cmnd(struct spraid_queue *ioq, struct spraid_com return; } + spraid_handle_ata_cmd(hdev, scmd, iod); + spraid_map_status(iod, scmd, cqe); if (iod->nsge) { iod->nsge = 0; @@ -1224,38 +1275,36 @@ static void spraid_complete_ioq_cmnd(struct spraid_queue *ioq, struct spraid_com scmd->scsi_done(scmd); } -static inline void spraid_end_admin_request(struct request *req, __le16 status, - __le32 result0, __le32 result1) -{ - struct spraid_admin_request *rq = spraid_admin_req(req); - - rq->status = le16_to_cpu(status) >> 1; - rq->result0 = le32_to_cpu(result0); - rq->result1 = le32_to_cpu(result1); - blk_mq_complete_request(req); -} - static void spraid_complete_adminq_cmnd(struct spraid_queue *adminq, struct spraid_completion *cqe) { - struct blk_mq_tags *tags = adminq->hdev->admin_tagset.tags[0]; - struct request *req; + struct spraid_dev *hdev = adminq->hdev; + struct spraid_cmd *adm_cmd; - req = blk_mq_tag_to_rq(tags, cqe->cmd_id); - if (unlikely(!req)) { + adm_cmd = hdev->adm_cmds + cqe->cmd_id; + if (unlikely(adm_cmd->state == SPRAID_CMD_IDLE)) { dev_warn(adminq->hdev->dev, "Invalid id %d completed on queue %d\n", cqe->cmd_id, le16_to_cpu(cqe->sq_id)); return; } - spraid_end_admin_request(req, cqe->status, cqe->result, cqe->result1); + + adm_cmd->status = le16_to_cpu(cqe->status) >> 1; + adm_cmd->result0 = le32_to_cpu(cqe->result); + adm_cmd->result1 = le32_to_cpu(cqe->result1); + + complete(&adm_cmd->cmd_done); } +static void spraid_send_aen(struct spraid_dev *hdev, u16 cid); + static void spraid_complete_aen(struct spraid_queue *spraidq, struct spraid_completion *cqe) { struct spraid_dev *hdev = spraidq->hdev; u32 result = le32_to_cpu(cqe->result); - dev_info(hdev->dev, "rcv aen, status[%x], result[%x]\n", - le16_to_cpu(cqe->status) >> 1, result); + dev_info(hdev->dev, "rcv aen, cid[%d], status[0x%x], result[0x%x]\n", + cqe->cmd_id, le16_to_cpu(cqe->status) >> 1, result); + + spraid_send_aen(hdev, cqe->cmd_id); if ((le16_to_cpu(cqe->status) >> 1) != SPRAID_SC_SUCCESS) return; @@ -1264,22 +1313,19 @@ static void spraid_complete_aen(struct spraid_queue *spraidq, struct spraid_comp spraid_handle_aen_notice(hdev, result); break; case SPRAID_AEN_VS: - spraid_handle_aen_vs(hdev, result); + spraid_handle_aen_vs(hdev, result, le32_to_cpu(cqe->result1)); break; default: dev_warn(hdev->dev, "Unsupported async event type: %u\n", result & 0x7); break; } - queue_work(spraid_wq, &hdev->aen_work); } -static void spraid_put_ioq_ptcmd(struct spraid_dev *hdev, struct spraid_ioq_ptcmd *cmd); - static void spraid_complete_ioq_sync_cmnd(struct spraid_queue *ioq, struct spraid_completion *cqe) { struct spraid_dev *hdev = ioq->hdev; - struct spraid_ioq_ptcmd *ptcmd; + struct spraid_cmd *ptcmd; ptcmd = hdev->ioq_ptcmds + (ioq->qid - 1) * SPRAID_PTCMDS_PERQ + cqe->cmd_id - SPRAID_IO_BLK_MQ_DEPTH; @@ -1289,8 +1335,6 @@ static void spraid_complete_ioq_sync_cmnd(struct spraid_queue *ioq, struct sprai ptcmd->result1 = le32_to_cpu(cqe->result1); complete(&ptcmd->cmd_done); - - spraid_put_ioq_ptcmd(hdev, ptcmd); } static inline void spraid_handle_cqe(struct spraid_queue *spraidq, u16 idx) @@ -1304,7 +1348,7 @@ static inline void spraid_handle_cqe(struct spraid_queue *spraidq, u16 idx) return; } - dev_log_dbg(hdev->dev, "cid[%d], qid[%d], result[0x%x], sq_id[%d], status[0x%x]\n", + dev_log_dbg(hdev->dev, "cid[%d] qid[%d], result[0x%x], sq_id[%d], status[0x%x]\n", cqe->cmd_id, spraidq->qid, le32_to_cpu(cqe->result), le16_to_cpu(cqe->sq_id), le16_to_cpu(cqe->status)); @@ -1452,62 +1496,119 @@ static u32 spraid_bar_size(struct spraid_dev *hdev, u32 nr_ioqs) return (SPRAID_REG_DBS + ((nr_ioqs + 1) * 8 * hdev->db_stride)); } -static inline void spraid_clear_spraid_request(struct request *req) +static int spraid_alloc_admin_cmds(struct spraid_dev *hdev) { - if (!(req->rq_flags & RQF_DONTPREP)) { - spraid_admin_req(req)->flags = 0; - req->rq_flags |= RQF_DONTPREP; + int i; + + INIT_LIST_HEAD(&hdev->adm_cmd_list); + spin_lock_init(&hdev->adm_cmd_lock); + + hdev->adm_cmds = kcalloc_node(SPRAID_AQ_BLK_MQ_DEPTH, sizeof(struct spraid_cmd), + GFP_KERNEL, hdev->numa_node); + + if (!hdev->adm_cmds) { + dev_err(hdev->dev, "Alloc admin cmds failed\n"); + return -ENOMEM; + } + + for (i = 0; i < SPRAID_AQ_BLK_MQ_DEPTH; i++) { + hdev->adm_cmds[i].qid = 0; + hdev->adm_cmds[i].cid = i; + list_add_tail(&(hdev->adm_cmds[i].list), &hdev->adm_cmd_list); } + + dev_info(hdev->dev, "Alloc admin cmds success, num[%d]\n", SPRAID_AQ_BLK_MQ_DEPTH); + + return 0; } -static struct request *spraid_alloc_admin_request(struct request_queue *q, - struct spraid_admin_command *cmd, - blk_mq_req_flags_t flags) +static void spraid_free_admin_cmds(struct spraid_dev *hdev) { - u32 op = COMMAND_IS_WRITE(cmd) ? REQ_OP_DRV_OUT : REQ_OP_DRV_IN; - struct request *req; + kfree(hdev->adm_cmds); + hdev->adm_cmds = NULL; + INIT_LIST_HEAD(&hdev->adm_cmd_list); +} + +static struct spraid_cmd *spraid_get_cmd(struct spraid_dev *hdev, enum spraid_cmd_type type) +{ + struct spraid_cmd *cmd = NULL; + unsigned long flags; + struct list_head *head = &hdev->adm_cmd_list; + spinlock_t *slock = &hdev->adm_cmd_lock; + + if (type == SPRAID_CMD_IOPT) { + head = &hdev->ioq_pt_list; + slock = &hdev->ioq_pt_lock; + } + + spin_lock_irqsave(slock, flags); + if (list_empty(head)) { + spin_unlock_irqrestore(slock, flags); + dev_err(hdev->dev, "err, cmd[%d] list empty\n", type); + return NULL; + } + cmd = list_entry(head->next, struct spraid_cmd, list); + list_del_init(&cmd->list); + spin_unlock_irqrestore(slock, flags); - req = blk_mq_alloc_request(q, op, flags); - if (IS_ERR(req)) - return req; - req->cmd_flags |= REQ_FAILFAST_DRIVER; - spraid_clear_spraid_request(req); - spraid_admin_req(req)->cmd = cmd; + WRITE_ONCE(cmd->state, SPRAID_CMD_IN_FLIGHT); - return req; + return cmd; } -static int spraid_submit_admin_sync_cmd(struct request_queue *q, - struct spraid_admin_command *cmd, - u32 *result, void *buffer, - u32 bufflen, u32 timeout, int at_head, blk_mq_req_flags_t flags) +static void spraid_put_cmd(struct spraid_dev *hdev, struct spraid_cmd *cmd, + enum spraid_cmd_type type) { - struct request *req; - int ret; + unsigned long flags; + struct list_head *head = &hdev->adm_cmd_list; + spinlock_t *slock = &hdev->adm_cmd_lock; - req = spraid_alloc_admin_request(q, cmd, flags); - if (IS_ERR(req)) - return PTR_ERR(req); + if (type == SPRAID_CMD_IOPT) { + head = &hdev->ioq_pt_list; + slock = &hdev->ioq_pt_lock; + } - req->timeout = timeout ? timeout : ADMIN_TIMEOUT; - if (buffer && bufflen) { - ret = blk_rq_map_kern(q, req, buffer, bufflen, GFP_KERNEL); - if (ret) - goto out; + spin_lock_irqsave(slock, flags); + WRITE_ONCE(cmd->state, SPRAID_CMD_IDLE); + list_add_tail(&cmd->list, head); + spin_unlock_irqrestore(slock, flags); +} + + +static int spraid_submit_admin_sync_cmd(struct spraid_dev *hdev, struct spraid_admin_command *cmd, + u32 *result0, u32 *result1, u32 timeout) +{ + struct spraid_cmd *adm_cmd = spraid_get_cmd(hdev, SPRAID_CMD_ADM); + + if (!adm_cmd) { + dev_err(hdev->dev, "err, get admin cmd failed\n"); + return -EFAULT; } - blk_execute_rq(req->q, NULL, req, at_head); - if (result) - *result = spraid_admin_req(req)->result0; + timeout = timeout ? timeout : ADMIN_TIMEOUT; - if (spraid_admin_req(req)->flags & SPRAID_REQ_CANCELLED) - ret = -EINTR; - else - ret = spraid_admin_req(req)->status; + init_completion(&adm_cmd->cmd_done); -out: - blk_mq_free_request(req); - return ret; + cmd->common.command_id = adm_cmd->cid; + spraid_submit_cmd(&hdev->queues[0], cmd); + + if (!wait_for_completion_timeout(&adm_cmd->cmd_done, timeout)) { + dev_err(hdev->dev, "[%s] cid[%d] qid[%d] timeout, opcode[0x%x] subopcode[0x%x]\n", + __func__, adm_cmd->cid, adm_cmd->qid, cmd->usr_cmd.opcode, + cmd->usr_cmd.info_0.subopcode); + WRITE_ONCE(adm_cmd->state, SPRAID_CMD_TIMEOUT); + spraid_put_cmd(hdev, adm_cmd, SPRAID_CMD_ADM); + return -ETIME; + } + + if (result0) + *result0 = adm_cmd->result0; + if (result1) + *result1 = adm_cmd->result1; + + spraid_put_cmd(hdev, adm_cmd, SPRAID_CMD_ADM); + + return adm_cmd->status; } static int spraid_create_cq(struct spraid_dev *hdev, u16 qid, @@ -1524,8 +1625,7 @@ static int spraid_create_cq(struct spraid_dev *hdev, u16 qid, admin_cmd.create_cq.cq_flags = cpu_to_le16(flags); admin_cmd.create_cq.irq_vector = cpu_to_le16(cq_vector); - return spraid_submit_admin_sync_cmd(hdev->admin_q, &admin_cmd, NULL, - NULL, 0, 0, 0, 0); + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); } static int spraid_create_sq(struct spraid_dev *hdev, u16 qid, @@ -1542,8 +1642,7 @@ static int spraid_create_sq(struct spraid_dev *hdev, u16 qid, admin_cmd.create_sq.sq_flags = cpu_to_le16(flags); admin_cmd.create_sq.cqid = cpu_to_le16(qid); - return spraid_submit_admin_sync_cmd(hdev->admin_q, &admin_cmd, NULL, - NULL, 0, 0, 0, 0); + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); } static void spraid_free_queue(struct spraid_queue *spraidq) @@ -1581,8 +1680,7 @@ static int spraid_delete_queue(struct spraid_dev *hdev, u8 op, u16 id) admin_cmd.delete_queue.opcode = op; admin_cmd.delete_queue.qid = cpu_to_le16(id); - ret = spraid_submit_admin_sync_cmd(hdev->admin_q, &admin_cmd, NULL, - NULL, 0, 0, 0, 0); + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); if (ret) dev_err(hdev->dev, "Delete %s:[%d] failed\n", @@ -1663,19 +1761,28 @@ static int spraid_set_features(struct spraid_dev *hdev, u32 fid, u32 dword11, vo size_t buflen, u32 *result) { struct spraid_admin_command admin_cmd; - u32 res; int ret; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + + if (buffer && buflen) { + data_ptr = dma_alloc_coherent(hdev->dev, buflen, &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memcpy(data_ptr, buffer, buflen); + } memset(&admin_cmd, 0, sizeof(admin_cmd)); admin_cmd.features.opcode = SPRAID_ADMIN_SET_FEATURES; admin_cmd.features.fid = cpu_to_le32(fid); admin_cmd.features.dword11 = cpu_to_le32(dword11); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); - ret = spraid_submit_admin_sync_cmd(hdev->admin_q, &admin_cmd, &res, - buffer, buflen, 0, 0, 0); + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, result, NULL, 0); - if (!ret && result) - *result = res; + if (data_ptr) + dma_free_coherent(hdev->dev, buflen, data_ptr, data_dma); return ret; } @@ -1764,8 +1871,7 @@ static int spraid_setup_io_queues(struct spraid_dev *hdev) break; } dev_info(hdev->dev, "[%s] max_qid: %d, queue_count: %d, online_queue: %d, ioq_depth: %d\n", - __func__, hdev->max_qid, hdev->queue_count, - hdev->online_queues, hdev->ioq_depth); + __func__, hdev->max_qid, hdev->queue_count, hdev->online_queues, hdev->ioq_depth); return spraid_create_io_queues(hdev); } @@ -1889,10 +1995,11 @@ static int spraid_get_dev_list(struct spraid_dev *hdev, struct spraid_dev_info * u32 nd = le32_to_cpu(hdev->ctrl_info->nd); struct spraid_admin_command admin_cmd; struct spraid_dev_list *list_buf; + dma_addr_t data_dma = 0; u32 i, idx, hdid, ndev; int ret = 0; - list_buf = kmalloc(sizeof(*list_buf), GFP_KERNEL); + list_buf = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &data_dma, GFP_KERNEL); if (!list_buf) return -ENOMEM; @@ -1901,9 +2008,9 @@ static int spraid_get_dev_list(struct spraid_dev *hdev, struct spraid_dev_info * admin_cmd.get_info.opcode = SPRAID_ADMIN_GET_INFO; admin_cmd.get_info.type = SPRAID_GET_INFO_DEV_LIST; admin_cmd.get_info.cdw11 = cpu_to_le32(idx); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); - ret = spraid_submit_admin_sync_cmd(hdev->admin_q, &admin_cmd, NULL, list_buf, - sizeof(*list_buf), 0, 0, 0); + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); if (ret) { dev_err(hdev->dev, "Get device list failed, nd: %u, idx: %u, ret: %d\n", @@ -1916,12 +2023,11 @@ static int spraid_get_dev_list(struct spraid_dev *hdev, struct spraid_dev_info * for (i = 0; i < ndev; i++) { hdid = le32_to_cpu(list_buf->devices[i].hdid); - dev_info(hdev->dev, "list_buf->devices[%d], hdid: %u target: %d, channel: %d, lun: %d, attr[%x]\n", - i, hdid, - le16_to_cpu(list_buf->devices[i].target), - list_buf->devices[i].channel, - list_buf->devices[i].lun, - list_buf->devices[i].attr); + dev_info(hdev->dev, "list_buf->devices[%d], hdid: %u target: %d, channel: %d, lun: %d, attr[0x%x]\n", + i, hdid, le16_to_cpu(list_buf->devices[i].target), + list_buf->devices[i].channel, + list_buf->devices[i].lun, + list_buf->devices[i].attr); if (hdid > nd || hdid == 0) { dev_err(hdev->dev, "err, hdid[%d] invalid\n", hdid); continue; @@ -1936,21 +2042,29 @@ static int spraid_get_dev_list(struct spraid_dev *hdev, struct spraid_dev_info * } out: - kfree(list_buf); + dma_free_coherent(hdev->dev, PAGE_SIZE, list_buf, data_dma); return ret; } -static void spraid_send_aen(struct spraid_dev *hdev) +static void spraid_send_aen(struct spraid_dev *hdev, u16 cid) { struct spraid_queue *adminq = &hdev->queues[0]; struct spraid_admin_command admin_cmd; memset(&admin_cmd, 0, sizeof(admin_cmd)); admin_cmd.common.opcode = SPRAID_ADMIN_ASYNC_EVENT; - admin_cmd.common.command_id = SPRAID_AQ_BLK_MQ_DEPTH; + admin_cmd.common.command_id = cid; spraid_submit_cmd(adminq, &admin_cmd); - dev_info(hdev->dev, "send aen, cid[%d]\n", SPRAID_AQ_BLK_MQ_DEPTH); + dev_info(hdev->dev, "send aen, cid[%d]\n", cid); +} + +static inline void spraid_send_all_aen(struct spraid_dev *hdev) +{ + u16 i; + + for (i = 0; i < hdev->ctrl_info->aerl; i++) + spraid_send_aen(hdev, i + SPRAID_AQ_BLK_MQ_DEPTH); } static int spraid_add_device(struct spraid_dev *hdev, struct spraid_dev_info *device) @@ -1958,6 +2072,10 @@ static int spraid_add_device(struct spraid_dev *hdev, struct spraid_dev_info *de struct Scsi_Host *shost = hdev->shost; struct scsi_device *sdev; + dev_info(hdev->dev, "add device, hdid: %u target: %d, channel: %d, lun: %d, attr[0x%x]\n", + le32_to_cpu(device->hdid), le16_to_cpu(device->target), + device->channel, device->lun, device->attr); + sdev = scsi_device_lookup(shost, device->channel, le16_to_cpu(device->target), 0); if (sdev) { dev_warn(hdev->dev, "Device is already exist, channel: %d, target_id: %d, lun: %d\n", @@ -1974,9 +2092,13 @@ static int spraid_rescan_device(struct spraid_dev *hdev, struct spraid_dev_info struct Scsi_Host *shost = hdev->shost; struct scsi_device *sdev; + dev_info(hdev->dev, "rescan device, hdid: %u target: %d, channel: %d, lun: %d, attr[0x%x]\n", + le32_to_cpu(device->hdid), le16_to_cpu(device->target), + device->channel, device->lun, device->attr); + sdev = scsi_device_lookup(shost, device->channel, le16_to_cpu(device->target), 0); if (!sdev) { - dev_warn(hdev->dev, "Device is not exit, channel: %d, target_id: %d, lun: %d\n", + dev_warn(hdev->dev, "device is not exit rescan it, channel: %d, target_id: %d, lun: %d\n", device->channel, le16_to_cpu(device->target), 0); return -ENODEV; } @@ -1991,9 +2113,13 @@ static int spraid_remove_device(struct spraid_dev *hdev, struct spraid_dev_info struct Scsi_Host *shost = hdev->shost; struct scsi_device *sdev; + dev_info(hdev->dev, "remove device, hdid: %u target: %d, channel: %d, lun: %d, attr[0x%x]\n", + le32_to_cpu(org_device->hdid), le16_to_cpu(org_device->target), + org_device->channel, org_device->lun, org_device->attr); + sdev = scsi_device_lookup(shost, org_device->channel, le16_to_cpu(org_device->target), 0); if (!sdev) { - dev_warn(hdev->dev, "Device is not exit, channel: %d, target_id: %d, lun: %d\n", + dev_warn(hdev->dev, "device is not exit remove it, channel: %d, target_id: %d, lun: %d\n", org_device->channel, le16_to_cpu(org_device->target), 0); return -ENODEV; } @@ -2029,36 +2155,54 @@ static int spraid_dev_list_init(struct spraid_dev *hdev) return 0; } +static int luntarget_cmp_func(const void *l, const void *r) +{ + const struct spraid_dev_info *ln = l; + const struct spraid_dev_info *rn = r; + + if (ln->channel == rn->channel) + return le16_to_cpu(ln->target) - le16_to_cpu(rn->target); + + return ln->channel - rn->channel; +} + static void spraid_scan_work(struct work_struct *work) { struct spraid_dev *hdev = container_of(work, struct spraid_dev, scan_work); struct spraid_dev_info *devices, *org_devices; + struct spraid_dev_info *sortdevice; u32 nd = le32_to_cpu(hdev->ctrl_info->nd); u8 flag, org_flag; int i, ret; + int count = 0; devices = kcalloc(nd, sizeof(struct spraid_dev_info), GFP_KERNEL); if (!devices) return; + + sortdevice = kcalloc(nd, sizeof(struct spraid_dev_info), GFP_KERNEL); + if (!sortdevice) + goto free_list; + ret = spraid_get_dev_list(hdev, devices); if (ret) - goto free_list; + goto free_all; org_devices = hdev->devices; for (i = 0; i < nd; i++) { org_flag = org_devices[i].flag; flag = devices[i].flag; - dev_log_dbg(hdev->dev, "i: %d, org_flag: 0x%x, flag: 0x%x\n", - i, org_flag, flag); + dev_log_dbg(hdev->dev, "i: %d, org_flag: 0x%x, flag: 0x%x\n", i, org_flag, flag); if (SPRAID_DEV_INFO_FLAG_VALID(flag)) { if (!SPRAID_DEV_INFO_FLAG_VALID(org_flag)) { down_write(&hdev->devices_rwsem); memcpy(&org_devices[i], &devices[i], - sizeof(struct spraid_dev_info)); + sizeof(struct spraid_dev_info)); + memcpy(&sortdevice[count++], &devices[i], + sizeof(struct spraid_dev_info)); up_write(&hdev->devices_rwsem); - spraid_add_device(hdev, &devices[i]); } else if (SPRAID_DEV_INFO_FLAG_CHANGE(flag)) { spraid_rescan_device(hdev, &devices[i]); } @@ -2071,6 +2215,16 @@ static void spraid_scan_work(struct work_struct *work) } } } + + dev_info(hdev->dev, "scan work add device count = %d\n", count); + + sort(sortdevice, count, sizeof(sortdevice[0]), luntarget_cmp_func, NULL); + + for (i = 0; i < count; i++) + spraid_add_device(hdev, &sortdevice[i]); + +free_all: + kfree(sortdevice); free_list: kfree(devices); } @@ -2083,6 +2237,15 @@ static void spraid_timesyn_work(struct work_struct *work) spraid_configure_timestamp(hdev); } +static int spraid_init_ctrl_info(struct spraid_dev *hdev); +static void spraid_fw_act_work(struct work_struct *work) +{ + struct spraid_dev *hdev = container_of(work, struct spraid_dev, fw_act_work); + + if (spraid_init_ctrl_info(hdev)) + dev_err(hdev->dev, "get ctrl info failed after fw act\n"); +} + static void spraid_queue_scan(struct spraid_dev *hdev) { queue_work(spraid_wq, &hdev->scan_work); @@ -2094,6 +2257,9 @@ static void spraid_handle_aen_notice(struct spraid_dev *hdev, u32 result) case SPRAID_AEN_DEV_CHANGED: spraid_queue_scan(hdev); break; + case SPRAID_AEN_FW_ACT_START: + dev_info(hdev->dev, "fw activation starting\n"); + break; case SPRAID_AEN_HOST_PROBING: break; default: @@ -2101,25 +2267,25 @@ static void spraid_handle_aen_notice(struct spraid_dev *hdev, u32 result) } } -static void spraid_handle_aen_vs(struct spraid_dev *hdev, u32 result) +static void spraid_handle_aen_vs(struct spraid_dev *hdev, u32 result, u32 result1) { - switch (result) { + switch ((result & 0xff00) >> 8) { case SPRAID_AEN_TIMESYN: queue_work(spraid_wq, &hdev->timesyn_work); break; + case SPRAID_AEN_FW_ACT_FINISH: + dev_info(hdev->dev, "fw activation finish\n"); + queue_work(spraid_wq, &hdev->fw_act_work); + break; + case SPRAID_AEN_EVENT_MIN ... SPRAID_AEN_EVENT_MAX: + dev_info(hdev->dev, "rcv card event[%d], param1[0x%x] param2[0x%x]\n", + (result & 0xff00) >> 8, result, result1); + break; default: - dev_warn(hdev->dev, "async event result: %x\n", result); + dev_warn(hdev->dev, "async event result: 0x%x\n", result); } } -static void spraid_async_event_work(struct work_struct *work) -{ - struct spraid_dev *hdev = - container_of(work, struct spraid_dev, aen_work); - - spraid_send_aen(hdev); -} - static int spraid_alloc_resources(struct spraid_dev *hdev) { int ret, nqueue; @@ -2149,10 +2315,16 @@ static int spraid_alloc_resources(struct spraid_dev *hdev) goto destroy_dma_pools; } + ret = spraid_alloc_admin_cmds(hdev); + if (ret) + goto free_queues; + dev_info(hdev->dev, "[%s] queues num: %d\n", __func__, nqueue); return 0; +free_queues: + kfree(hdev->queues); destroy_dma_pools: spraid_destroy_dma_pools(hdev); free_ctrl_info: @@ -2164,50 +2336,18 @@ static int spraid_alloc_resources(struct spraid_dev *hdev) static void spraid_free_resources(struct spraid_dev *hdev) { + spraid_free_admin_cmds(hdev); kfree(hdev->queues); spraid_destroy_dma_pools(hdev); kfree(hdev->ctrl_info); ida_free(&spraid_instance_ida, hdev->instance); } -static void spraid_setup_passthrough(struct request *req, struct spraid_admin_command *cmd) -{ - memcpy(cmd, spraid_admin_req(req)->cmd, sizeof(*cmd)); - cmd->common.flags &= ~SPRAID_CMD_FLAG_SGL_ALL; -} - -static inline void spraid_clear_hreq(struct request *req) -{ - if (!(req->rq_flags & RQF_DONTPREP)) { - spraid_admin_req(req)->flags = 0; - req->rq_flags |= RQF_DONTPREP; - } -} - -static blk_status_t spraid_setup_admin_cmd(struct request *req, struct spraid_admin_command *cmd) +static void spraid_bsg_unmap_data(struct spraid_dev *hdev, struct bsg_job *job) { - spraid_clear_hreq(req); - - memset(cmd, 0, sizeof(*cmd)); - switch (req_op(req)) { - case REQ_OP_DRV_IN: - case REQ_OP_DRV_OUT: - spraid_setup_passthrough(req, cmd); - break; - default: - WARN_ON_ONCE(1); - return BLK_STS_IOERR; - } - - cmd->common.command_id = req->tag; - return BLK_STS_OK; -} - -static void spraid_unmap_data(struct spraid_dev *hdev, struct request *req) -{ - struct spraid_iod *iod = blk_mq_rq_to_pdu(req); - enum dma_data_direction dma_dir = rq_data_dir(req) ? - DMA_TO_DEVICE : DMA_FROM_DEVICE; + struct request *rq = blk_mq_rq_from_pdu(job); + struct spraid_iod *iod = job->dd_data; + enum dma_data_direction dma_dir = rq_data_dir(rq) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; if (iod->nsge) dma_unmap_sg(hdev->dev, iod->sg, iod->nsge, dma_dir); @@ -2215,36 +2355,36 @@ static void spraid_unmap_data(struct spraid_dev *hdev, struct request *req) spraid_free_iod_res(hdev, iod); } -static blk_status_t spraid_admin_map_data(struct spraid_dev *hdev, struct request *req, - struct spraid_admin_command *cmd) +static int spraid_bsg_map_data(struct spraid_dev *hdev, struct bsg_job *job, + struct spraid_admin_command *cmd) { - struct spraid_iod *iod = blk_mq_rq_to_pdu(req); - struct request_queue *admin_q = req->q; - enum dma_data_direction dma_dir = rq_data_dir(req) ? - DMA_TO_DEVICE : DMA_FROM_DEVICE; - blk_status_t ret = BLK_STS_IOERR; - int nr_mapped; - int res; + struct request *rq = blk_mq_rq_from_pdu(job); + struct spraid_iod *iod = job->dd_data; + enum dma_data_direction dma_dir = rq_data_dir(rq) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; + int ret = 0; + + iod->sg = job->request_payload.sg_list; + iod->nsge = job->request_payload.sg_cnt; + iod->length = job->request_payload.payload_len; + iod->use_sgl = false; + iod->npages = -1; + iod->sg_drv_mgmt = false; - sg_init_table(iod->sg, blk_rq_nr_phys_segments(req)); - iod->nsge = blk_rq_map_sg(admin_q, req, iod->sg); if (!iod->nsge) goto out; - dev_info(hdev->dev, "nseg: %u, nsge: %u\n", - blk_rq_nr_phys_segments(req), iod->nsge); - - ret = BLK_STS_RESOURCE; - nr_mapped = dma_map_sg_attrs(hdev->dev, iod->sg, iod->nsge, dma_dir, DMA_ATTR_NO_WARN); - if (!nr_mapped) + ret = dma_map_sg_attrs(hdev->dev, iod->sg, iod->nsge, dma_dir, DMA_ATTR_NO_WARN); + if (!ret) goto out; - res = spraid_setup_prps(hdev, iod); - if (res) + ret = spraid_setup_prps(hdev, iod); + if (ret) goto unmap; + cmd->common.dptr.prp1 = cpu_to_le64(sg_dma_address(iod->sg)); cmd->common.dptr.prp2 = cpu_to_le64(iod->first_dma); - return BLK_STS_OK; + + return 0; unmap: dma_unmap_sg(hdev->dev, iod->sg, iod->nsge, dma_dir); @@ -2252,137 +2392,29 @@ static blk_status_t spraid_admin_map_data(struct spraid_dev *hdev, struct reques return ret; } -static blk_status_t spraid_init_admin_iod(struct request *rq, struct spraid_dev *hdev) -{ - struct spraid_iod *iod = blk_mq_rq_to_pdu(rq); - int nents = blk_rq_nr_phys_segments(rq); - unsigned int size = blk_rq_payload_bytes(rq); - - if (nents > SPRAID_INT_PAGES || size > SPRAID_INT_BYTES(hdev)) { - iod->sg = mempool_alloc(hdev->iod_mempool, GFP_ATOMIC); - if (!iod->sg) - return BLK_STS_RESOURCE; - } else { - iod->sg = iod->inline_sg; - } - - iod->nsge = 0; - iod->use_sgl = false; - iod->npages = -1; - iod->length = size; - iod->sg_drv_mgmt = true; - - return BLK_STS_OK; -} - -static blk_status_t spraid_queue_admin_rq(struct blk_mq_hw_ctx *hctx, - const struct blk_mq_queue_data *bd) -{ - struct spraid_queue *adminq = hctx->driver_data; - struct spraid_dev *hdev = adminq->hdev; - struct request *req = bd->rq; - struct spraid_iod *iod = blk_mq_rq_to_pdu(req); - struct spraid_admin_command cmd; - blk_status_t ret; - - ret = spraid_setup_admin_cmd(req, &cmd); - if (ret) - goto out; - - ret = spraid_init_admin_iod(req, hdev); - if (ret) - goto out; - - if (blk_rq_nr_phys_segments(req)) { - ret = spraid_admin_map_data(hdev, req, &cmd); - if (ret) - goto cleanup_iod; - } - - blk_mq_start_request(req); - spraid_submit_cmd(adminq, &cmd); - return BLK_STS_OK; - -cleanup_iod: - spraid_free_iod_res(hdev, iod); -out: - return ret; -} - -static blk_status_t spraid_error_status(struct request *req) -{ - switch (spraid_admin_req(req)->status & 0x7ff) { - case SPRAID_SC_SUCCESS: - return BLK_STS_OK; - default: - return BLK_STS_IOERR; - } -} - -static void spraid_complete_admin_rq(struct request *req) -{ - struct spraid_iod *iod = blk_mq_rq_to_pdu(req); - struct spraid_dev *hdev = iod->spraidq->hdev; - - if (blk_rq_nr_phys_segments(req)) - spraid_unmap_data(hdev, req); - blk_mq_end_request(req, spraid_error_status(req)); -} - -static int spraid_admin_init_hctx(struct blk_mq_hw_ctx *hctx, void *data, unsigned int hctx_idx) -{ - struct spraid_dev *hdev = data; - struct spraid_queue *adminq = &hdev->queues[0]; - - WARN_ON(hctx_idx != 0); - WARN_ON(hdev->admin_tagset.tags[0] != hctx->tags); - - hctx->driver_data = adminq; - return 0; -} - -static int spraid_admin_init_request(struct blk_mq_tag_set *set, struct request *req, - unsigned int hctx_idx, unsigned int numa_node) -{ - struct spraid_dev *hdev = set->driver_data; - struct spraid_iod *iod = blk_mq_rq_to_pdu(req); - struct spraid_queue *adminq = &hdev->queues[0]; - - WARN_ON(!adminq); - iod->spraidq = adminq; - return 0; -} - -static enum blk_eh_timer_return -spraid_admin_timeout(struct request *req, bool reserved) -{ - struct spraid_iod *iod = blk_mq_rq_to_pdu(req); - struct spraid_queue *spraidq = iod->spraidq; - struct spraid_dev *hdev = spraidq->hdev; - - dev_err(hdev->dev, "Admin cid[%d] qid[%d] timeout\n", - req->tag, spraidq->qid); - - if (spraid_poll_cq(spraidq, req->tag)) { - dev_warn(hdev->dev, "cid[%d] qid[%d] timeout, completion polled\n", - req->tag, spraidq->qid); - return BLK_EH_DONE; - } - - spraid_end_admin_request(req, cpu_to_le16(-EINVAL), 0, 0); - return BLK_EH_DONE; -} - static int spraid_get_ctrl_info(struct spraid_dev *hdev, struct spraid_ctrl_info *ctrl_info) { struct spraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + int ret; + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; memset(&admin_cmd, 0, sizeof(admin_cmd)); admin_cmd.get_info.opcode = SPRAID_ADMIN_GET_INFO; admin_cmd.get_info.type = SPRAID_GET_INFO_CTRL; + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + if (!ret) + memcpy(ctrl_info, data_ptr, sizeof(struct spraid_ctrl_info)); - return spraid_submit_admin_sync_cmd(hdev->admin_q, &admin_cmd, NULL, - ctrl_info, sizeof(struct spraid_ctrl_info), 0, 0, 0); + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, data_dma); + + return ret; } static int spraid_init_ctrl_info(struct spraid_dev *hdev) @@ -2416,6 +2448,11 @@ static int spraid_init_ctrl_info(struct spraid_dev *hdev) dev_info(hdev->dev, "[%s]sn = %s\n", __func__, hdev->ctrl_info->sn); dev_info(hdev->dev, "[%s]fr = %s\n", __func__, hdev->ctrl_info->fr); + if (!hdev->ctrl_info->aerl) + hdev->ctrl_info->aerl = 1; + if (hdev->ctrl_info->aerl > SPRAID_NR_AEN_COMMANDS) + hdev->ctrl_info->aerl = SPRAID_NR_AEN_COMMANDS; + return 0; } @@ -2444,98 +2481,51 @@ static void spraid_free_iod_ext_mem_pool(struct spraid_dev *hdev) mempool_destroy(hdev->iod_mempool); } -static int spraid_submit_user_cmd(struct request_queue *q, struct spraid_admin_command *cmd, - void __user *ubuffer, unsigned int bufflen, u32 *result, - unsigned int timeout) +static int spraid_user_admin_cmd(struct spraid_dev *hdev, struct bsg_job *job) { - struct request *req; - struct bio *bio = NULL; - int ret; - - req = spraid_alloc_admin_request(q, cmd, 0); - if (IS_ERR(req)) - return PTR_ERR(req); + struct spraid_bsg_request *bsg_req = job->request; + struct spraid_passthru_common_cmd *cmd = &(bsg_req->admcmd); + struct spraid_admin_command admin_cmd; + u32 timeout = msecs_to_jiffies(cmd->timeout_ms); + u32 result[2] = {0}; + int status; - req->timeout = timeout ? timeout : ADMIN_TIMEOUT; - spraid_admin_req(req)->flags |= SPRAID_REQ_USERCMD; + if (hdev->state >= SPRAID_RESETTING) { + dev_err(hdev->dev, "[%s] err, host state:[%d] is not right\n", + __func__, hdev->state); + return -EBUSY; + } - if (ubuffer && bufflen) { - ret = blk_rq_map_user(q, req, NULL, ubuffer, bufflen, GFP_KERNEL); - if (ret) - goto out; - bio = req->bio; + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.common.opcode = cmd->opcode; + admin_cmd.common.flags = cmd->flags; + admin_cmd.common.hdid = cpu_to_le32(cmd->nsid); + admin_cmd.common.cdw2[0] = cpu_to_le32(cmd->cdw2); + admin_cmd.common.cdw2[1] = cpu_to_le32(cmd->cdw3); + admin_cmd.common.cdw10 = cpu_to_le32(cmd->cdw10); + admin_cmd.common.cdw11 = cpu_to_le32(cmd->cdw11); + admin_cmd.common.cdw12 = cpu_to_le32(cmd->cdw12); + admin_cmd.common.cdw13 = cpu_to_le32(cmd->cdw13); + admin_cmd.common.cdw14 = cpu_to_le32(cmd->cdw14); + admin_cmd.common.cdw15 = cpu_to_le32(cmd->cdw15); + + status = spraid_bsg_map_data(hdev, job, &admin_cmd); + if (status) { + dev_err(hdev->dev, "[%s] err, map data failed\n", __func__); + return status; } - blk_execute_rq(req->q, NULL, req, 0); - if (spraid_admin_req(req)->flags & SPRAID_REQ_CANCELLED) - ret = -EINTR; - else - ret = spraid_admin_req(req)->status; - if (result) { - result[0] = spraid_admin_req(req)->result0; - result[1] = spraid_admin_req(req)->result1; - } - if (bio) - blk_rq_unmap_user(bio); -out: - blk_mq_free_request(req); - return ret; -} -static int spraid_user_admin_cmd(struct spraid_dev *hdev, - struct spraid_passthru_common_cmd __user *ucmd) -{ - struct spraid_passthru_common_cmd cmd; - struct spraid_admin_command admin_cmd; - u32 timeout = 0; - int status; - - if (!capable(CAP_SYS_ADMIN)) { - dev_err(hdev->dev, "Current user hasn't administrator right, reject service\n"); - return -EACCES; - } - - if (copy_from_user(&cmd, ucmd, sizeof(cmd))) { - dev_err(hdev->dev, "Copy command from user space to kernel space failed\n"); - return -EFAULT; - } - - if (cmd.flags) { - dev_err(hdev->dev, "Invalid flags in user command\n"); - return -EINVAL; + status = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, &result[0], &result[1], timeout); + if (status >= 0) { + job->reply_len = sizeof(result); + memcpy(job->reply, result, sizeof(result)); } - dev_info(hdev->dev, "user_admin_cmd opcode: 0x%x, subopcode: 0x%x\n", - cmd.opcode, cmd.cdw2 & 0x7ff); + if (status) + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x], status[0x%x] result0[0x%x] result1[0x%x]\n", + __func__, cmd->opcode, cmd->info_0.subopcode, status, result[0], result[1]); - memset(&admin_cmd, 0, sizeof(admin_cmd)); - admin_cmd.common.opcode = cmd.opcode; - admin_cmd.common.flags = cmd.flags; - admin_cmd.common.hdid = cpu_to_le32(cmd.nsid); - admin_cmd.common.cdw2[0] = cpu_to_le32(cmd.cdw2); - admin_cmd.common.cdw2[1] = cpu_to_le32(cmd.cdw3); - admin_cmd.common.cdw10 = cpu_to_le32(cmd.cdw10); - admin_cmd.common.cdw11 = cpu_to_le32(cmd.cdw11); - admin_cmd.common.cdw12 = cpu_to_le32(cmd.cdw12); - admin_cmd.common.cdw13 = cpu_to_le32(cmd.cdw13); - admin_cmd.common.cdw14 = cpu_to_le32(cmd.cdw14); - admin_cmd.common.cdw15 = cpu_to_le32(cmd.cdw15); - - if (cmd.timeout_ms) - timeout = msecs_to_jiffies(cmd.timeout_ms); - - status = spraid_submit_user_cmd(hdev->admin_q, &admin_cmd, - (void __user *)(uintptr_t)cmd.addr, cmd.info_1.data_len, - &cmd.result0, timeout); - - dev_info(hdev->dev, "user_admin_cmd status: 0x%x, result0: 0x%x, result1: 0x%x\n", - status, cmd.result0, cmd.result1); - - if (status >= 0) { - if (put_user(cmd.result0, &ucmd->result0)) - return -EFAULT; - if (put_user(cmd.result1, &ucmd->result1)) - return -EFAULT; - } + spraid_bsg_unmap_data(hdev, job); return status; } @@ -2548,8 +2538,8 @@ static int spraid_alloc_ioq_ptcmds(struct spraid_dev *hdev) INIT_LIST_HEAD(&hdev->ioq_pt_list); spin_lock_init(&hdev->ioq_pt_lock); - hdev->ioq_ptcmds = kcalloc_node(ptnum, sizeof(struct spraid_ioq_ptcmd), - GFP_KERNEL, hdev->numa_node); + hdev->ioq_ptcmds = kcalloc_node(ptnum, sizeof(struct spraid_cmd), + GFP_KERNEL, hdev->numa_node); if (!hdev->ioq_ptcmds) { dev_err(hdev->dev, "Alloc ioq_ptcmds failed\n"); @@ -2567,55 +2557,35 @@ static int spraid_alloc_ioq_ptcmds(struct spraid_dev *hdev) return 0; } -static struct spraid_ioq_ptcmd *spraid_get_ioq_ptcmd(struct spraid_dev *hdev) -{ - struct spraid_ioq_ptcmd *cmd = NULL; - unsigned long flags; - - spin_lock_irqsave(&hdev->ioq_pt_lock, flags); - if (list_empty(&hdev->ioq_pt_list)) { - spin_unlock_irqrestore(&hdev->ioq_pt_lock, flags); - dev_err(hdev->dev, "err, ioq ptcmd list empty\n"); - return NULL; - } - cmd = list_entry((&hdev->ioq_pt_list)->next, struct spraid_ioq_ptcmd, list); - list_del_init(&cmd->list); - spin_unlock_irqrestore(&hdev->ioq_pt_lock, flags); - - WRITE_ONCE(cmd->state, SPRAID_CMD_IDLE); - - return cmd; -} - -static void spraid_put_ioq_ptcmd(struct spraid_dev *hdev, struct spraid_ioq_ptcmd *cmd) +static void spraid_free_ioq_ptcmds(struct spraid_dev *hdev) { - unsigned long flags; + kfree(hdev->ioq_ptcmds); + hdev->ioq_ptcmds = NULL; - spin_lock_irqsave(&hdev->ioq_pt_lock, flags); - list_add(&cmd->list, (&hdev->ioq_pt_list)->next); - spin_unlock_irqrestore(&hdev->ioq_pt_lock, flags); + INIT_LIST_HEAD(&hdev->ioq_pt_list); } static int spraid_submit_ioq_sync_cmd(struct spraid_dev *hdev, struct spraid_ioq_command *cmd, - u32 *result, void **sense, u32 timeout) + u32 *result, u32 *reslen, u32 timeout) { - struct spraid_queue *ioq; int ret; dma_addr_t sense_dma; - struct spraid_ioq_ptcmd *pt_cmd = spraid_get_ioq_ptcmd(hdev); - - *sense = NULL; + struct spraid_queue *ioq; + void *sense_addr = NULL; + struct spraid_cmd *pt_cmd = spraid_get_cmd(hdev, SPRAID_CMD_IOPT); - if (!pt_cmd) + if (!pt_cmd) { + dev_err(hdev->dev, "err, get ioq cmd failed\n"); return -EFAULT; + } - dev_info(hdev->dev, "[%s] ptcmd, cid[%d], qid[%d]\n", __func__, pt_cmd->cid, pt_cmd->qid); + timeout = timeout ? timeout : ADMIN_TIMEOUT; init_completion(&pt_cmd->cmd_done); ioq = &hdev->queues[pt_cmd->qid]; ret = pt_cmd->cid * SCSI_SENSE_BUFFERSIZE; - pt_cmd->priv = ioq->sense + ret; + sense_addr = ioq->sense + ret; sense_dma = ioq->sense_dma_addr + ret; cmd->common.sense_addr = cpu_to_le64(sense_dma); @@ -2625,260 +2595,87 @@ static int spraid_submit_ioq_sync_cmd(struct spraid_dev *hdev, struct spraid_ioq spraid_submit_cmd(ioq, cmd); if (!wait_for_completion_timeout(&pt_cmd->cmd_done, timeout)) { - dev_err(hdev->dev, "[%s] cid[%d], qid[%d] timeout\n", - __func__, pt_cmd->cid, pt_cmd->qid); + dev_err(hdev->dev, "[%s] cid[%d] qid[%d] timeout, opcode[0x%x] subopcode[0x%x]\n", + __func__, pt_cmd->cid, pt_cmd->qid, cmd->common.opcode, + (le32_to_cpu(cmd->common.cdw3[0]) & 0xffff)); WRITE_ONCE(pt_cmd->state, SPRAID_CMD_TIMEOUT); - return -EINVAL; + spraid_put_cmd(hdev, pt_cmd, SPRAID_CMD_IOPT); + return -ETIME; } - if (result) { - result[0] = pt_cmd->result0; - result[1] = pt_cmd->result1; + if (result && reslen) { + if ((pt_cmd->status & 0x17f) == 0x101) { + memcpy(result, sense_addr, SCSI_SENSE_BUFFERSIZE); + *reslen = SCSI_SENSE_BUFFERSIZE; + } } - if ((pt_cmd->status & 0x17f) == 0x101) - *sense = pt_cmd->priv; + spraid_put_cmd(hdev, pt_cmd, SPRAID_CMD_IOPT); return pt_cmd->status; } -static int spraid_user_ioq_cmd(struct spraid_dev *hdev, - struct spraid_ioq_passthru_cmd __user *ucmd) +static int spraid_user_ioq_cmd(struct spraid_dev *hdev, struct bsg_job *job) { - struct spraid_ioq_passthru_cmd cmd; + struct spraid_bsg_request *bsg_req = (struct spraid_bsg_request *)(job->request); + struct spraid_ioq_passthru_cmd *cmd = &(bsg_req->ioqcmd); struct spraid_ioq_command ioq_cmd; - u32 timeout = 0; int status = 0; - u8 *data_ptr = NULL; - dma_addr_t data_dma; - enum dma_data_direction dma_dir = DMA_NONE; - void *sense = NULL; - - if (!capable(CAP_SYS_ADMIN)) { - dev_err(hdev->dev, "Current user hasn't administrator right, reject service\n"); - return -EACCES; - } + u32 timeout = msecs_to_jiffies(cmd->timeout_ms); - if (copy_from_user(&cmd, ucmd, sizeof(cmd))) { - dev_err(hdev->dev, "Copy command from user space to kernel space failed\n"); - return -EFAULT; - } - - if (cmd.data_len > PAGE_SIZE) { + if (cmd->data_len > PAGE_SIZE) { dev_err(hdev->dev, "[%s] data len bigger than 4k\n", __func__); return -EFAULT; } - dev_info(hdev->dev, "[%s] opcode: 0x%x, subopcode: 0x%x, datalen: %d\n", - __func__, cmd.opcode, cmd.info_1.subopcode, cmd.data_len); - - if (cmd.addr && cmd.data_len) { - data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &data_dma, GFP_KERNEL); - if (!data_ptr) - return -ENOMEM; - - dma_dir = (cmd.opcode & 1) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; + if (hdev->state != SPRAID_LIVE) { + dev_err(hdev->dev, "[%s] err, host state:[%d] is not live\n", + __func__, hdev->state); + return -EBUSY; } - if (dma_dir == DMA_TO_DEVICE) { - if (copy_from_user(data_ptr, (void __user *)(uintptr_t)cmd.addr, cmd.data_len)) { - dev_err(hdev->dev, "[%s] copy user data failed\n", __func__); - status = -EFAULT; - goto free_dma_mem; - } - } + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x] init, datalen[%d]\n", + __func__, cmd->opcode, cmd->info_1.subopcode, cmd->data_len); memset(&ioq_cmd, 0, sizeof(ioq_cmd)); - ioq_cmd.common.opcode = cmd.opcode; - ioq_cmd.common.flags = cmd.flags; - ioq_cmd.common.hdid = cpu_to_le32(cmd.nsid); - ioq_cmd.common.sense_len = cpu_to_le16(cmd.info_0.res_sense_len); - ioq_cmd.common.cdb_len = cmd.info_0.cdb_len; - ioq_cmd.common.rsvd2 = cmd.info_0.rsvd0; - ioq_cmd.common.cdw3[0] = cpu_to_le32(cmd.cdw3); - ioq_cmd.common.cdw3[1] = cpu_to_le32(cmd.cdw4); - ioq_cmd.common.cdw3[2] = cpu_to_le32(cmd.cdw5); - ioq_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); - - ioq_cmd.common.cdw10[0] = cpu_to_le32(cmd.cdw10); - ioq_cmd.common.cdw10[1] = cpu_to_le32(cmd.cdw11); - ioq_cmd.common.cdw10[2] = cpu_to_le32(cmd.cdw12); - ioq_cmd.common.cdw10[3] = cpu_to_le32(cmd.cdw13); - ioq_cmd.common.cdw10[4] = cpu_to_le32(cmd.cdw14); - ioq_cmd.common.cdw10[5] = cpu_to_le32(cmd.data_len); - - memcpy(ioq_cmd.common.cdb, &cmd.cdw16, cmd.info_0.cdb_len); - - ioq_cmd.common.cdw26[0] = cpu_to_le32(cmd.cdw26[0]); - ioq_cmd.common.cdw26[1] = cpu_to_le32(cmd.cdw26[1]); - ioq_cmd.common.cdw26[2] = cpu_to_le32(cmd.cdw26[2]); - ioq_cmd.common.cdw26[3] = cpu_to_le32(cmd.cdw26[3]); - - if (cmd.timeout_ms) - timeout = msecs_to_jiffies(cmd.timeout_ms); - timeout = timeout ? timeout : ADMIN_TIMEOUT; - - status = spraid_submit_ioq_sync_cmd(hdev, &ioq_cmd, &cmd.result0, &sense, timeout); - - if (status >= 0) { - if (put_user(cmd.result0, &ucmd->result0)) { - status = -EFAULT; - goto free_dma_mem; - } - if (put_user(cmd.result1, &ucmd->result1)) { - status = -EFAULT; - goto free_dma_mem; - } - if (dma_dir == DMA_FROM_DEVICE && - copy_to_user((void __user *)(uintptr_t)cmd.addr, data_ptr, cmd.data_len)) { - status = -EFAULT; - goto free_dma_mem; - } - } - - if (sense) { - if (copy_to_user((void *__user *)(uintptr_t)cmd.sense_addr, - sense, cmd.info_0.res_sense_len)) { - status = -EFAULT; - goto free_dma_mem; - } - } - -free_dma_mem: - if (data_ptr) - dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, data_dma); - - return status; - -} - -static int spraid_reset_work_sync(struct spraid_dev *hdev); - -static int spraid_user_reset_cmd(struct spraid_dev *hdev) -{ - int ret; - - dev_info(hdev->dev, "[%s] start user reset cmd\n", __func__); - ret = spraid_reset_work_sync(hdev); - dev_info(hdev->dev, "[%s] stop user reset cmd[%d]\n", __func__, ret); - - return ret; -} - -static int hdev_open(struct inode *inode, struct file *file) -{ - struct spraid_dev *hdev = - container_of(inode->i_cdev, struct spraid_dev, cdev); - file->private_data = hdev; - return 0; -} - -static long hdev_ioctl(struct file *file, u32 cmd, unsigned long arg) -{ - struct spraid_dev *hdev = file->private_data; - void __user *argp = (void __user *)arg; - - switch (cmd) { - case SPRAID_IOCTL_ADMIN_CMD: - return spraid_user_admin_cmd(hdev, argp); - case SPRAID_IOCTL_IOQ_CMD: - return spraid_user_ioq_cmd(hdev, argp); - case SPRAID_IOCTL_RESET_CMD: - return spraid_user_reset_cmd(hdev); - default: - return -ENOTTY; - } -} - -static const struct file_operations spraid_dev_fops = { - .owner = THIS_MODULE, - .open = hdev_open, - .unlocked_ioctl = hdev_ioctl, - .compat_ioctl = hdev_ioctl, -}; - -static int spraid_create_cdev(struct spraid_dev *hdev) -{ - int ret; - - device_initialize(&hdev->ctrl_device); - hdev->ctrl_device.devt = MKDEV(MAJOR(spraid_chr_devt), hdev->instance); - hdev->ctrl_device.class = spraid_class; - hdev->ctrl_device.parent = hdev->dev; - dev_set_drvdata(&hdev->ctrl_device, hdev); - ret = dev_set_name(&hdev->ctrl_device, "spraid%d", hdev->instance); - if (ret) - return ret; - cdev_init(&hdev->cdev, &spraid_dev_fops); - hdev->cdev.owner = THIS_MODULE; - ret = cdev_device_add(&hdev->cdev, &hdev->ctrl_device); - if (ret) { - dev_err(hdev->dev, "Add cdev failed, ret: %d", ret); - put_device(&hdev->ctrl_device); - kfree_const(hdev->ctrl_device.kobj.name); - return ret; - } - - return 0; -} - -static inline void spraid_remove_cdev(struct spraid_dev *hdev) -{ - cdev_device_del(&hdev->cdev, &hdev->ctrl_device); -} - -static const struct blk_mq_ops spraid_admin_mq_ops = { - .queue_rq = spraid_queue_admin_rq, - .complete = spraid_complete_admin_rq, - .init_hctx = spraid_admin_init_hctx, - .init_request = spraid_admin_init_request, - .timeout = spraid_admin_timeout, -}; - -static void spraid_remove_admin_tagset(struct spraid_dev *hdev) -{ - if (hdev->admin_q && !blk_queue_dying(hdev->admin_q)) { - blk_mq_unquiesce_queue(hdev->admin_q); - blk_cleanup_queue(hdev->admin_q); - blk_mq_free_tag_set(&hdev->admin_tagset); + ioq_cmd.common.opcode = cmd->opcode; + ioq_cmd.common.flags = cmd->flags; + ioq_cmd.common.hdid = cpu_to_le32(cmd->nsid); + ioq_cmd.common.sense_len = cpu_to_le16(cmd->info_0.res_sense_len); + ioq_cmd.common.cdb_len = cmd->info_0.cdb_len; + ioq_cmd.common.rsvd2 = cmd->info_0.rsvd0; + ioq_cmd.common.cdw3[0] = cpu_to_le32(cmd->cdw3); + ioq_cmd.common.cdw3[1] = cpu_to_le32(cmd->cdw4); + ioq_cmd.common.cdw3[2] = cpu_to_le32(cmd->cdw5); + + ioq_cmd.common.cdw10[0] = cpu_to_le32(cmd->cdw10); + ioq_cmd.common.cdw10[1] = cpu_to_le32(cmd->cdw11); + ioq_cmd.common.cdw10[2] = cpu_to_le32(cmd->cdw12); + ioq_cmd.common.cdw10[3] = cpu_to_le32(cmd->cdw13); + ioq_cmd.common.cdw10[4] = cpu_to_le32(cmd->cdw14); + ioq_cmd.common.cdw10[5] = cpu_to_le32(cmd->data_len); + + memcpy(ioq_cmd.common.cdb, &cmd->cdw16, cmd->info_0.cdb_len); + + ioq_cmd.common.cdw26[0] = cpu_to_le32(cmd->cdw26[0]); + ioq_cmd.common.cdw26[1] = cpu_to_le32(cmd->cdw26[1]); + ioq_cmd.common.cdw26[2] = cpu_to_le32(cmd->cdw26[2]); + ioq_cmd.common.cdw26[3] = cpu_to_le32(cmd->cdw26[3]); + + status = spraid_bsg_map_data(hdev, job, (struct spraid_admin_command *)&ioq_cmd); + if (status) { + dev_err(hdev->dev, "[%s] err, map data failed\n", __func__); + return status; } -} -static int spraid_alloc_admin_tags(struct spraid_dev *hdev) -{ - if (!hdev->admin_q) { - hdev->admin_tagset.ops = &spraid_admin_mq_ops; - hdev->admin_tagset.nr_hw_queues = 1; + status = spraid_submit_ioq_sync_cmd(hdev, &ioq_cmd, job->reply, &job->reply_len, timeout); - hdev->admin_tagset.queue_depth = SPRAID_AQ_MQ_TAG_DEPTH; - hdev->admin_tagset.timeout = ADMIN_TIMEOUT; - hdev->admin_tagset.numa_node = hdev->numa_node; - hdev->admin_tagset.cmd_size = - spraid_cmd_size(hdev, true, false); - hdev->admin_tagset.flags = BLK_MQ_F_NO_SCHED; - hdev->admin_tagset.driver_data = hdev; + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x], status[0x%x], reply_len[%d]\n", + __func__, cmd->opcode, cmd->info_1.subopcode, status, job->reply_len); - if (blk_mq_alloc_tag_set(&hdev->admin_tagset)) { - dev_err(hdev->dev, "Allocate admin tagset failed\n"); - return -ENOMEM; - } + spraid_bsg_unmap_data(hdev, job); - hdev->admin_q = blk_mq_init_queue(&hdev->admin_tagset); - if (IS_ERR(hdev->admin_q)) { - dev_err(hdev->dev, "Initialize admin request queue failed\n"); - blk_mq_free_tag_set(&hdev->admin_tagset); - return -ENOMEM; - } - if (!blk_get_queue(hdev->admin_q)) { - dev_err(hdev->dev, "Get admin request queue failed\n"); - spraid_remove_admin_tagset(hdev); - hdev->admin_q = NULL; - return -ENODEV; - } - } else { - blk_mq_unquiesce_queue(hdev->admin_q); - } - return 0; + return status; } static bool spraid_check_scmd_completed(struct scsi_cmnd *scmd) @@ -2891,7 +2688,7 @@ static bool spraid_check_scmd_completed(struct scsi_cmnd *scmd) spraid_get_tag_from_scmd(scmd, &hwq, &cid); spraidq = &hdev->queues[hwq]; if (READ_ONCE(iod->state) == SPRAID_CMD_COMPLETE || spraid_poll_cq(spraidq, cid)) { - dev_warn(hdev->dev, "cid[%d], qid[%d] has been completed\n", + dev_warn(hdev->dev, "cid[%d] qid[%d] has been completed\n", cid, spraidq->qid); return true; } @@ -2927,8 +2724,7 @@ static int spraid_send_abort_cmd(struct spraid_dev *hdev, u32 hdid, u16 qid, u16 admin_cmd.abort.sqid = cpu_to_le16(qid); admin_cmd.abort.cid = cpu_to_le16(cid); - return spraid_submit_admin_sync_cmd(hdev->admin_q, &admin_cmd, NULL, - NULL, 0, 0, 0, 0); + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); } /* send reset command by admin quueue temporary */ @@ -2941,8 +2737,7 @@ static int spraid_send_reset_cmd(struct spraid_dev *hdev, int type, u32 hdid) admin_cmd.reset.hdid = cpu_to_le32(hdid); admin_cmd.reset.type = type; - return spraid_submit_admin_sync_cmd(hdev->admin_q, &admin_cmd, NULL, - NULL, 0, 0, 0, 0); + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); } static bool spraid_change_host_state(struct spraid_dev *hdev, enum spraid_state newstate) @@ -3022,7 +2817,7 @@ static void spraid_back_fault_cqe(struct spraid_queue *ioq, struct spraid_comple scsi_dma_unmap(scmd); spraid_free_iod_res(hdev, iod); scmd->scsi_done(scmd); - dev_warn(hdev->dev, "Back fault CQE, cid[%d], qid[%d]\n", + dev_warn(hdev->dev, "Back fault CQE, cid[%d] qid[%d]\n", cqe->cmd_id, ioq->qid); } @@ -3032,6 +2827,8 @@ static void spraid_back_all_io(struct spraid_dev *hdev) struct spraid_queue *ioq; struct spraid_completion cqe = { 0 }; + scsi_block_requests(hdev->shost); + for (i = 1; i <= hdev->shost->nr_hw_queues; i++) { ioq = &hdev->queues[i]; for (j = 0; j < hdev->shost->can_queue; j++) { @@ -3039,6 +2836,8 @@ static void spraid_back_all_io(struct spraid_dev *hdev) spraid_back_fault_cqe(ioq, &cqe); } } + + scsi_unblock_requests(hdev->shost); } static void spraid_dev_disable(struct spraid_dev *hdev, bool shutdown) @@ -3106,17 +2905,13 @@ static void spraid_reset_work(struct work_struct *work) if (ret) goto pci_disable; - ret = spraid_alloc_admin_tags(hdev); - if (ret) - goto pci_disable; - ret = spraid_setup_io_queues(hdev); if (ret || hdev->online_queues <= hdev->shost->nr_hw_queues) goto pci_disable; spraid_change_host_state(hdev, SPRAID_LIVE); - spraid_send_aen(hdev); + spraid_send_all_aen(hdev); return; @@ -3174,8 +2969,8 @@ static int spraid_abort_handler(struct scsi_cmnd *scmd) scsi_print_command(scmd); - if (!spraid_wait_abnl_cmd_done(iod) || spraid_check_scmd_completed(scmd) || - hdev->state != SPRAID_LIVE) + if (hdev->state != SPRAID_LIVE || !spraid_wait_abnl_cmd_done(iod) || + spraid_check_scmd_completed(scmd)) return SUCCESS; hostdata = scmd->device->hostdata; @@ -3183,7 +2978,7 @@ static int spraid_abort_handler(struct scsi_cmnd *scmd) dev_warn(hdev->dev, "cid[%d] qid[%d] timeout, aborting\n", cid, hwq); ret = spraid_send_abort_cmd(hdev, hostdata->hdid, hwq, cid); - if (ret != ADMIN_ERR_TIMEOUT) { + if (ret != -ETIME) { ret = spraid_wait_abnl_cmd_done(iod); if (ret) { dev_warn(hdev->dev, "cid[%d] qid[%d] abort failed, not found\n", cid, hwq); @@ -3206,8 +3001,8 @@ static int spraid_tgt_reset_handler(struct scsi_cmnd *scmd) scsi_print_command(scmd); - if (!spraid_wait_abnl_cmd_done(iod) || spraid_check_scmd_completed(scmd) || - hdev->state != SPRAID_LIVE) + if (hdev->state != SPRAID_LIVE || !spraid_wait_abnl_cmd_done(iod) || + spraid_check_scmd_completed(scmd)) return SUCCESS; hostdata = scmd->device->hostdata; @@ -3241,8 +3036,8 @@ static int spraid_bus_reset_handler(struct scsi_cmnd *scmd) scsi_print_command(scmd); - if (!spraid_wait_abnl_cmd_done(iod) || spraid_check_scmd_completed(scmd) || - hdev->state != SPRAID_LIVE) + if (hdev->state != SPRAID_LIVE || !spraid_wait_abnl_cmd_done(iod) || + spraid_check_scmd_completed(scmd)) return SUCCESS; hostdata = scmd->device->hostdata; @@ -3272,7 +3067,7 @@ static int spraid_shost_reset_handler(struct scsi_cmnd *scmd) struct spraid_dev *hdev = shost_priv(scmd->device->host); scsi_print_command(scmd); - if (spraid_check_scmd_completed(scmd) || hdev->state != SPRAID_LIVE) + if (hdev->state != SPRAID_LIVE || spraid_check_scmd_completed(scmd)) return SUCCESS; spraid_get_tag_from_scmd(scmd, &hwq, &cid); @@ -3288,6 +3083,62 @@ static int spraid_shost_reset_handler(struct scsi_cmnd *scmd) return SUCCESS; } +static pci_ers_result_t spraid_pci_error_detected(struct pci_dev *pdev, + pci_channel_state_t state) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "enter pci error detect, state:%d\n", state); + + switch (state) { + case pci_channel_io_normal: + dev_warn(hdev->dev, "channel is normal, do nothing\n"); + + return PCI_ERS_RESULT_CAN_RECOVER; + case pci_channel_io_frozen: + dev_warn(hdev->dev, "channel io frozen, need reset controller\n"); + + scsi_block_requests(hdev->shost); + + spraid_change_host_state(hdev, SPRAID_RESETTING); + + return PCI_ERS_RESULT_NEED_RESET; + case pci_channel_io_perm_failure: + dev_warn(hdev->dev, "channel io failure, request disconnect\n"); + + return PCI_ERS_RESULT_DISCONNECT; + } + + return PCI_ERS_RESULT_NEED_RESET; +} + +static pci_ers_result_t spraid_pci_slot_reset(struct pci_dev *pdev) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "restart after slot reset\n"); + + pci_restore_state(pdev); + + if (!queue_work(spraid_wq, &hdev->reset_work)) { + dev_err(hdev->dev, "[%s] err, the device is resetting state\n", __func__); + return PCI_ERS_RESULT_NONE; + } + + flush_work(&hdev->reset_work); + + scsi_unblock_requests(hdev->shost); + + return PCI_ERS_RESULT_RECOVERED; +} + +static void spraid_reset_done(struct pci_dev *pdev) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "enter spraid reset done\n"); +} + static ssize_t csts_pp_show(struct device *cdev, struct device_attribute *attr, char *buf) { struct Scsi_Host *shost = class_to_shost(cdev); @@ -3347,7 +3198,7 @@ static ssize_t fw_version_show(struct device *cdev, struct device_attribute *att struct Scsi_Host *shost = class_to_shost(cdev); struct spraid_dev *hdev = shost_priv(shost); - return snprintf(buf, sizeof(hdev->ctrl_info->fr), "%s\n", hdev->ctrl_info->fr); + return snprintf(buf, PAGE_SIZE, "%s\n", hdev->ctrl_info->fr); } static DEVICE_ATTR_RO(csts_pp); @@ -3365,6 +3216,185 @@ static struct device_attribute *spraid_host_attrs[] = { NULL, }; +static int spraid_get_vd_info(struct spraid_dev *hdev, struct spraid_vd_info *vd_info, u16 vid) +{ + struct spraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + int ret; + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.usr_cmd.opcode = USR_CMD_READ; + admin_cmd.usr_cmd.info_0.subopcode = cpu_to_le16(USR_CMD_VDINFO); + admin_cmd.usr_cmd.info_1.data_len = cpu_to_le16(USR_CMD_RDLEN); + admin_cmd.usr_cmd.info_1.param_len = cpu_to_le16(VDINFO_PARAM_LEN); + admin_cmd.usr_cmd.cdw10 = cpu_to_le32(vid); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + if (!ret) + memcpy(vd_info, data_ptr, sizeof(struct spraid_vd_info)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, data_dma); + + return ret; +} + +static int spraid_get_bgtask(struct spraid_dev *hdev, struct spraid_bgtask *bgtask) +{ + struct spraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + int ret; + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.usr_cmd.opcode = USR_CMD_READ; + admin_cmd.usr_cmd.info_0.subopcode = cpu_to_le16(USR_CMD_BGTASK); + admin_cmd.usr_cmd.info_1.data_len = cpu_to_le16(USR_CMD_RDLEN); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + if (!ret) + memcpy(bgtask, data_ptr, sizeof(struct spraid_bgtask)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, data_dma); + + return ret; +} + +static ssize_t raid_level_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct spraid_dev *hdev; + struct spraid_vd_info *vd_info; + struct spraid_sdev_hostdata *hostdata; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !SPRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = spraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) + vd_info->rg_level = ARRAY_SIZE(raid_levels) - 1; + + ret = (vd_info->rg_level < ARRAY_SIZE(raid_levels)) ? + vd_info->rg_level : (ARRAY_SIZE(raid_levels) - 1); + + kfree(vd_info); + + return snprintf(buf, PAGE_SIZE, "RAID-%s\n", raid_levels[ret]); +} + +static ssize_t raid_state_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct spraid_dev *hdev; + struct spraid_vd_info *vd_info; + struct spraid_sdev_hostdata *hostdata; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !SPRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = spraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) { + vd_info->vd_status = 0; + vd_info->rg_id = 0xff; + } + + ret = (vd_info->vd_status < ARRAY_SIZE(raid_states)) ? vd_info->vd_status : 0; + + kfree(vd_info); + + return snprintf(buf, PAGE_SIZE, "%s\n", raid_states[ret]); +} + +static ssize_t raid_resync_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct spraid_dev *hdev; + struct spraid_vd_info *vd_info; + struct spraid_bgtask *bgtask; + struct spraid_sdev_hostdata *hostdata; + u8 rg_id, i, progress = 0; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !SPRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = spraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) + goto out; + + rg_id = vd_info->rg_id; + + bgtask = (struct spraid_bgtask *)vd_info; + ret = spraid_get_bgtask(hdev, bgtask); + if (ret) + goto out; + for (i = 0; i < bgtask->task_num; i++) { + if ((bgtask->bgtask[i].type == BGTASK_TYPE_REBUILD) && + (le16_to_cpu(bgtask->bgtask[i].vd_id) == rg_id)) + progress = bgtask->bgtask[i].progress; + } + +out: + kfree(vd_info); + return snprintf(buf, PAGE_SIZE, "%d\n", progress); +} + +static DEVICE_ATTR_RO(raid_level); +static DEVICE_ATTR_RO(raid_state); +static DEVICE_ATTR_RO(raid_resync); + +static struct device_attribute *spraid_dev_attrs[] = { + &dev_attr_raid_level, + &dev_attr_raid_state, + &dev_attr_raid_resync, + NULL, +}; + +static struct pci_error_handlers spraid_err_handler = { + .error_detected = spraid_pci_error_detected, + .slot_reset = spraid_pci_slot_reset, + .reset_done = spraid_reset_done, +}; + +static int spraid_sysfs_host_reset(struct Scsi_Host *shost, int reset_type) +{ + int ret; + struct spraid_dev *hdev = shost_priv(shost); + + dev_info(hdev->dev, "[%s] start sysfs host reset cmd\n", __func__); + ret = spraid_reset_work_sync(hdev); + dev_info(hdev->dev, "[%s] stop sysfs host reset cmd[%d]\n", __func__, ret); + + return ret; +} + static struct scsi_host_template spraid_driver_template = { .module = THIS_MODULE, .name = "Ramaxel Logic spraid driver", @@ -3379,9 +3409,11 @@ static struct scsi_host_template spraid_driver_template = { .eh_bus_reset_handler = spraid_bus_reset_handler, .eh_host_reset_handler = spraid_shost_reset_handler, .change_queue_depth = scsi_change_queue_depth, - .host_tagset = 1, + .host_tagset = 0, .this_id = -1, .shost_attrs = spraid_host_attrs, + .sdev_attrs = spraid_dev_attrs, + .host_reset = spraid_sysfs_host_reset, }; static void spraid_shutdown(struct pci_dev *pdev) @@ -3392,11 +3424,53 @@ static void spraid_shutdown(struct pci_dev *pdev) spraid_disable_admin_queue(hdev, true); } +/* bsg dispatch user command */ +static int spraid_bsg_host_dispatch(struct bsg_job *job) +{ + struct Scsi_Host *shost = dev_to_shost(job->dev); + struct spraid_dev *hdev = shost_priv(shost); + struct request *rq = blk_mq_rq_from_pdu(job); + struct spraid_bsg_request *bsg_req = job->request; + int ret = 0; + + dev_log_dbg(hdev->dev, "[%s] msgcode[%d], msglen[%d], timeout[%d], req_nsge[%d], req_len[%d]\n", + __func__, bsg_req->msgcode, job->request_len, rq->timeout, + job->request_payload.sg_cnt, job->request_payload.payload_len); + + job->reply_len = 0; + + switch (bsg_req->msgcode) { + case SPRAID_BSG_ADM: + ret = spraid_user_admin_cmd(hdev, job); + break; + case SPRAID_BSG_IOQ: + ret = spraid_user_ioq_cmd(hdev, job); + break; + default: + dev_info(hdev->dev, "[%s] unsupport msgcode[%d]\n", __func__, bsg_req->msgcode); + break; + } + + if (ret > 0) + ret = ret | (ret << 8); + + bsg_job_done(job, ret, 0); + return 0; +} + +static inline void spraid_remove_bsg(struct spraid_dev *hdev) +{ + if (hdev->bsg_queue) { + bsg_unregister_queue(hdev->bsg_queue); + blk_cleanup_queue(hdev->bsg_queue); + } +} static int spraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) { struct spraid_dev *hdev; struct Scsi_Host *shost; int node, ret; + char bsg_name[15]; shost = scsi_host_alloc(&spraid_driver_template, sizeof(*hdev)); if (!shost) { @@ -3421,10 +3495,10 @@ static int spraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) goto put_dev; init_rwsem(&hdev->devices_rwsem); - INIT_WORK(&hdev->aen_work, spraid_async_event_work); INIT_WORK(&hdev->scan_work, spraid_scan_work); INIT_WORK(&hdev->timesyn_work, spraid_timesyn_work); INIT_WORK(&hdev->reset_work, spraid_reset_work); + INIT_WORK(&hdev->fw_act_work, spraid_fw_act_work); spin_lock_init(&hdev->state_lock); ret = spraid_alloc_resources(hdev); @@ -3439,17 +3513,13 @@ static int spraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) if (ret) goto pci_disable; - ret = spraid_alloc_admin_tags(hdev); - if (ret) - goto disable_admin_q; - ret = spraid_init_ctrl_info(hdev); if (ret) - goto free_admin_tagset; + goto disable_admin_q; ret = spraid_alloc_iod_ext_mem_pool(hdev); if (ret) - goto free_admin_tagset; + goto disable_admin_q; ret = spraid_setup_io_queues(hdev); if (ret) @@ -3464,9 +3534,15 @@ static int spraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) goto remove_io_queues; } - ret = spraid_create_cdev(hdev); - if (ret) + snprintf(bsg_name, sizeof(bsg_name), "spraid%d", shost->host_no); + hdev->bsg_queue = bsg_setup_queue(&shost->shost_gendev, bsg_name, + spraid_bsg_host_dispatch, NULL, + spraid_cmd_size(hdev, true, false)); + if (IS_ERR(hdev->bsg_queue)) { + dev_err(hdev->dev, "err, setup bsg failed\n"); + hdev->bsg_queue = NULL; goto remove_io_queues; + } if (hdev->online_queues == SPRAID_ADMIN_QUEUE_NUM) { dev_warn(hdev->dev, "warn only admin queue can be used\n"); @@ -3475,11 +3551,11 @@ static int spraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) hdev->state = SPRAID_LIVE; - spraid_send_aen(hdev); + spraid_send_all_aen(hdev); ret = spraid_dev_list_init(hdev); if (ret) - goto remove_cdev; + goto remove_bsg; ret = spraid_configure_timestamp(hdev); if (ret) @@ -3487,20 +3563,18 @@ static int spraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) ret = spraid_alloc_ioq_ptcmds(hdev); if (ret) - goto remove_cdev; + goto remove_bsg; scsi_scan_host(hdev->shost); return 0; -remove_cdev: - spraid_remove_cdev(hdev); +remove_bsg: + spraid_remove_bsg(hdev); remove_io_queues: spraid_remove_io_queues(hdev); free_iod_mempool: spraid_free_iod_ext_mem_pool(hdev); -free_admin_tagset: - spraid_remove_admin_tagset(hdev); disable_admin_q: spraid_disable_admin_queue(hdev, false); pci_disable: @@ -3524,22 +3598,17 @@ static void spraid_remove(struct pci_dev *pdev) dev_info(hdev->dev, "enter spraid remove\n"); spraid_change_host_state(hdev, SPRAID_DELETING); + flush_work(&hdev->reset_work); - if (!pci_device_is_present(pdev)) { - scsi_block_requests(shost); + if (!pci_device_is_present(pdev)) spraid_back_all_io(hdev); - scsi_unblock_requests(shost); - } - flush_work(&hdev->reset_work); + spraid_remove_bsg(hdev); scsi_remove_host(shost); - - kfree(hdev->ioq_ptcmds); + spraid_free_ioq_ptcmds(hdev); kfree(hdev->devices); - spraid_remove_cdev(hdev); spraid_remove_io_queues(hdev); spraid_free_iod_ext_mem_pool(hdev); - spraid_remove_admin_tagset(hdev); spraid_disable_admin_queue(hdev, false); spraid_pci_disable(hdev); spraid_free_resources(hdev); @@ -3551,7 +3620,7 @@ static void spraid_remove(struct pci_dev *pdev) } static const struct pci_device_id spraid_id_table[] = { - { PCI_DEVICE(PCI_VENDOR_ID_RAMAXEL_LOGIC, SPRAID_SERVER_DEVICE_HAB_DID) }, + { PCI_DEVICE(PCI_VENDOR_ID_RAMAXEL_LOGIC, SPRAID_SERVER_DEVICE_HBA_DID) }, { PCI_DEVICE(PCI_VENDOR_ID_RAMAXEL_LOGIC, SPRAID_SERVER_DEVICE_RAID_DID) }, { 0, } }; @@ -3563,6 +3632,7 @@ static struct pci_driver spraid_driver = { .probe = spraid_probe, .remove = spraid_remove, .shutdown = spraid_shutdown, + .err_handler = &spraid_err_handler, }; static int __init spraid_init(void) @@ -3573,14 +3643,10 @@ static int __init spraid_init(void) if (!spraid_wq) return -ENOMEM; - ret = alloc_chrdev_region(&spraid_chr_devt, 0, SPRAID_MINORS, "spraid"); - if (ret < 0) - goto destroy_wq; - spraid_class = class_create(THIS_MODULE, "spraid"); if (IS_ERR(spraid_class)) { ret = PTR_ERR(spraid_class); - goto unregister_chrdev; + goto destroy_wq; } ret = pci_register_driver(&spraid_driver); @@ -3591,8 +3657,6 @@ static int __init spraid_init(void) destroy_class: class_destroy(spraid_class); -unregister_chrdev: - unregister_chrdev_region(spraid_chr_devt, SPRAID_MINORS); destroy_wq: destroy_workqueue(spraid_wq); @@ -3603,12 +3667,11 @@ static void __exit spraid_exit(void) { pci_unregister_driver(&spraid_driver); class_destroy(spraid_class); - unregister_chrdev_region(spraid_chr_devt, SPRAID_MINORS); destroy_workqueue(spraid_wq); ida_destroy(&spraid_instance_ida); } -MODULE_AUTHOR("Ramaxel Memory Technology"); +MODULE_AUTHOR("songyl(a)ramaxel.com"); MODULE_DESCRIPTION("Ramaxel Memory Technology SPraid Driver"); MODULE_LICENSE("GPL"); MODULE_VERSION(SPRAID_DRV_VERSION); -- 2.32.0

1 0

(no subject)
by Liu Zixian 27 Dec '21

27 Dec '21

From: sdlzx <hdu_sdlzx(a)163.com> Subject: [PATCH openEuler-5.10] Disable legacy DRM drivers euleros inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4OBDE CVE: NA ------------------------------------------------- This config enables dangerous old drivers. It is selected by CONFIG_NOUVEAU_LEGACY_CTX_SUPPORT which is suggested by upstream community that "modern distros should consider turning it off". CONFIG_DRM_VM was selected by CONFIG_DRM_LEGACY and can be turned off now. Signed-off-by: sdlzx <hdu_sdlzx(a)163.com> --- arch/arm64/configs/openeuler_defconfig | 10 ++-------- arch/x86/configs/openeuler_defconfig | 10 ++-------- 2 files changed, 4 insertions(+), 16 deletions(-) diff --git a/arch/arm64/configs/openeuler_defconfig b/arch/arm64/configs/openeuler_defconfig index 17bc8750b..6949824e8 100644 --- a/arch/arm64/configs/openeuler_defconfig +++ b/arch/arm64/configs/openeuler_defconfig @@ -4673,7 +4673,6 @@ CONFIG_DRM_TTM_DMA_PAGE_POOL=y CONFIG_DRM_VRAM_HELPER=y CONFIG_DRM_TTM_HELPER=y CONFIG_DRM_GEM_SHMEM_HELPER=y -CONFIG_DRM_VM=y CONFIG_DRM_SCHED=m # @@ -4719,7 +4718,7 @@ CONFIG_DRM_AMD_DC_DCN=y # CONFIG_HSA_AMD is not set CONFIG_DRM_NOUVEAU=m -CONFIG_NOUVEAU_LEGACY_CTX_SUPPORT=y +# CONFIG_NOUVEAU_LEGACY_CTX_SUPPORT is not set CONFIG_NOUVEAU_DEBUG=5 CONFIG_NOUVEAU_DEBUG_DEFAULT=3 # CONFIG_NOUVEAU_DEBUG_MMU is not set @@ -4817,12 +4816,7 @@ CONFIG_DRM_CIRRUS_QEMU=m # CONFIG_DRM_LIMA is not set # CONFIG_DRM_PANFROST is not set # CONFIG_DRM_TIDSS is not set -CONFIG_DRM_LEGACY=y -# CONFIG_DRM_TDFX is not set -# CONFIG_DRM_R128 is not set -# CONFIG_DRM_MGA is not set -# CONFIG_DRM_VIA is not set -# CONFIG_DRM_SAVAGE is not set +# CONFIG_DRM_LEGACY is not set CONFIG_DRM_PANEL_ORIENTATION_QUIRKS=y # diff --git a/arch/x86/configs/openeuler_defconfig b/arch/x86/configs/openeuler_defconfig index 7b6083018..dbae04231 100644 --- a/arch/x86/configs/openeuler_defconfig +++ b/arch/x86/configs/openeuler_defconfig @@ -5040,7 +5040,6 @@ CONFIG_DRM_TTM_DMA_PAGE_POOL=y CONFIG_DRM_VRAM_HELPER=m CONFIG_DRM_TTM_HELPER=m CONFIG_DRM_GEM_SHMEM_HELPER=y -CONFIG_DRM_VM=y CONFIG_DRM_SCHED=m # @@ -5085,7 +5084,7 @@ CONFIG_DRM_AMD_DC_DCN=y # CONFIG_HSA_AMD is not set CONFIG_DRM_NOUVEAU=m -CONFIG_NOUVEAU_LEGACY_CTX_SUPPORT=y +# CONFIG_NOUVEAU_LEGACY_CTX_SUPPORT is not set CONFIG_NOUVEAU_DEBUG=5 CONFIG_NOUVEAU_DEBUG_DEFAULT=3 # CONFIG_NOUVEAU_DEBUG_MMU is not set @@ -5226,12 +5225,7 @@ CONFIG_DRM_CIRRUS_QEMU=m # CONFIG_TINYDRM_ST7735R is not set # CONFIG_DRM_XEN is not set # CONFIG_DRM_VBOXVIDEO is not set -CONFIG_DRM_LEGACY=y -# CONFIG_DRM_TDFX is not set -# CONFIG_DRM_R128 is not set -# CONFIG_DRM_MGA is not set -# CONFIG_DRM_VIA is not set -# CONFIG_DRM_SAVAGE is not set +# CONFIG_DRM_LEGACY is not set CONFIG_DRM_PANEL_ORIENTATION_QUIRKS=y # -- 2.27.0

1 0

[PATCH openEuler-5.10 001/129] ipvlan: Introduce l2e mode
by Zheng Zengkai 27 Dec '21

27 Dec '21

From: Keefe LIU <liuqifa(a)huawei.com> hulk inclusion category: feature bugzilla: 9511, https://gitee.com/openeuler/kernel/issues/I4IHL1 CVE: NA ------------------------------------------------- In a typical IPvlan L2 setup where master is in default-ns and each slave is into different (slave) ns. In this setup, if master and slaves in different net, egress packet processing for traffic originating from slave-ns can't be forwarded to master or other machine whose ip in the same net with master, and they can't be forwarded to other interface in default-ns. This patch introuce a new mode l2e for ipvlan to realize above goals, and it won't affect the original l2, l3, l3s mode. As the ip tool doesn't support l2e mode, We use module param "ipvlan_default_mode" to set the default work mode. 0 for l2 mode, 1 for l3, 2 for l2e, 3 for l3s, others invalid now. Attention, when we create ipvlan devices by "ip" commond, if we assign the mode, ipvlan will work in the mode we assigned other then the "ipvlan_default_mode". Signed-off-by: Keefe LIU <liuqifa(a)huawei.com> Reviewed-by: Wei Yongjun <weiyongjun1(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> Modified: when insmod the ipvlan module, the origin version(4.19) is 0 for l2, 1 for l3, 2 for l2e, 3 for l3s mode, but in 5.10 version is 0 for l2, 1 for l3, 2 for l3s, 3 for l2e mode. Signed-off-by: Lu Wei <luwei32(a)huawei.com> Reviewed-by: Yue Haibing <yuehaibing(a)huawei.com> Reviewed-by: Wei Yongjun <weiyongjun1(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> --- drivers/net/ipvlan/ipvlan_core.c | 190 +++++++++++++++++++++++++++++++ drivers/net/ipvlan/ipvlan_main.c | 6 +- include/uapi/linux/if_link.h | 1 + 3 files changed, 196 insertions(+), 1 deletion(-) diff --git a/drivers/net/ipvlan/ipvlan_core.c b/drivers/net/ipvlan/ipvlan_core.c index 8801d093135c..5b695ec5c650 100644 --- a/drivers/net/ipvlan/ipvlan_core.c +++ b/drivers/net/ipvlan/ipvlan_core.c @@ -532,6 +532,122 @@ static int ipvlan_process_outbound(struct sk_buff *skb) return ret; } +static int ipvlan_process_v4_forward(struct sk_buff *skb) +{ + const struct iphdr *ip4h = ip_hdr(skb); + struct net_device *dev = skb->dev; + struct net *net = dev_net(dev); + struct rtable *rt; + int err, ret = NET_XMIT_DROP; + struct flowi4 fl4 = { + .flowi4_tos = RT_TOS(ip4h->tos), + .flowi4_flags = FLOWI_FLAG_ANYSRC, + .flowi4_mark = skb->mark, + .daddr = ip4h->daddr, + .saddr = ip4h->saddr, + }; + + rt = ip_route_output_flow(net, &fl4, NULL); + if (IS_ERR(rt)) + goto err; + + if (rt->rt_type != RTN_UNICAST && rt->rt_type != RTN_LOCAL) { + ip_rt_put(rt); + goto err; + } + skb_dst_set(skb, &rt->dst); + err = ip_local_out(net, skb->sk, skb); + if (unlikely(net_xmit_eval(err))) + dev->stats.tx_errors++; + else + ret = NET_XMIT_SUCCESS; + goto out; +err: + dev->stats.tx_errors++; + kfree_skb(skb); +out: + return ret; +} + +#if IS_ENABLED(CONFIG_IPV6) +static int ipvlan_process_v6_forward(struct sk_buff *skb) +{ + const struct ipv6hdr *ip6h = ipv6_hdr(skb); + struct net_device *dev = skb->dev; + struct net *net = dev_net(dev); + struct dst_entry *dst; + int err, ret = NET_XMIT_DROP; + struct flowi6 fl6 = { + .daddr = ip6h->daddr, + .saddr = ip6h->saddr, + .flowi6_flags = FLOWI_FLAG_ANYSRC, + .flowlabel = ip6_flowinfo(ip6h), + .flowi6_mark = skb->mark, + .flowi6_proto = ip6h->nexthdr, + }; + + dst = ip6_route_output(net, NULL, &fl6); + if (dst->error) { + ret = dst->error; + dst_release(dst); + goto err; + } + skb_dst_set(skb, dst); + err = ip6_local_out(net, skb->sk, skb); + if (unlikely(net_xmit_eval(err))) + dev->stats.tx_errors++; + else + ret = NET_XMIT_SUCCESS; + goto out; +err: + dev->stats.tx_errors++; + kfree_skb(skb); +out: + return ret; +} +#else +static int ipvlan_process_v6_forward(struct sk_buff *skb) +{ + return NET_XMIT_DROP; +} +#endif + +static int ipvlan_process_forward(struct sk_buff *skb) +{ + struct ethhdr *ethh = eth_hdr(skb); + int ret = NET_XMIT_DROP; + + /* In this mode we dont care about multicast and broadcast traffic */ + if (is_multicast_ether_addr(ethh->h_dest)) { + pr_debug_ratelimited("Dropped {multi|broad}cast of type=[%x]\n", + ntohs(skb->protocol)); + kfree_skb(skb); + goto out; + } + + /* The ipvlan is a pseudo-L2 device, so the packets that we receive + * will have L2; which need to discarded and processed further + * in the net-ns of the main-device. + */ + if (skb_mac_header_was_set(skb)) { + skb_pull(skb, sizeof(*ethh)); + skb->mac_header = (typeof(skb->mac_header))~0U; + skb_reset_network_header(skb); + } + + if (skb->protocol == htons(ETH_P_IPV6)) { + ret = ipvlan_process_v6_forward(skb); + } else if (skb->protocol == htons(ETH_P_IP)) { + ret = ipvlan_process_v4_forward(skb); + } else { + pr_warn_ratelimited("Dropped outbound packet type=%x\n", + ntohs(skb->protocol)); + kfree_skb(skb); + } +out: + return ret; +} + static void ipvlan_multicast_enqueue(struct ipvl_port *port, struct sk_buff *skb, bool tx_pkt) { @@ -629,6 +745,46 @@ static int ipvlan_xmit_mode_l2(struct sk_buff *skb, struct net_device *dev) return dev_queue_xmit(skb); } +static int ipvlan_xmit_mode_l2e(struct sk_buff *skb, struct net_device *dev) +{ + const struct ipvl_dev *ipvlan = netdev_priv(dev); + struct ethhdr *eth = eth_hdr(skb); + struct ipvl_addr *addr; + void *lyr3h; + int addr_type; + + if (!ipvlan_is_vepa(ipvlan->port) && + ether_addr_equal(eth->h_dest, eth->h_source)) { + lyr3h = ipvlan_get_L3_hdr(ipvlan->port, skb, &addr_type); + if (lyr3h) { + addr = ipvlan_addr_lookup(ipvlan->port, lyr3h, + addr_type, true); + if (addr) { + if (ipvlan_is_private(ipvlan->port)) { + consume_skb(skb); + return NET_XMIT_DROP; + } + return ipvlan_rcv_frame(addr, &skb, true); + } + } + skb = skb_share_check(skb, GFP_ATOMIC); + if (!skb) + return NET_XMIT_DROP; + + /* maybe the packet need been forward */ + skb->dev = ipvlan->phy_dev; + ipvlan_skb_crossing_ns(skb, ipvlan->phy_dev); + return ipvlan_process_forward(skb); + } else if (is_multicast_ether_addr(eth->h_dest)) { + ipvlan_skb_crossing_ns(skb, NULL); + ipvlan_multicast_enqueue(ipvlan->port, skb, true); + return NET_XMIT_SUCCESS; + } + + skb->dev = ipvlan->phy_dev; + return dev_queue_xmit(skb); +} + int ipvlan_queue_xmit(struct sk_buff *skb, struct net_device *dev) { struct ipvl_dev *ipvlan = netdev_priv(dev); @@ -648,6 +804,8 @@ int ipvlan_queue_xmit(struct sk_buff *skb, struct net_device *dev) case IPVLAN_MODE_L3S: #endif return ipvlan_xmit_mode_l3(skb, dev); + case IPVLAN_MODE_L2E: + return ipvlan_xmit_mode_l2e(skb, dev); } /* Should not reach here */ @@ -729,6 +887,36 @@ static rx_handler_result_t ipvlan_handle_mode_l2(struct sk_buff **pskb, return ret; } +static rx_handler_result_t ipvlan_handle_mode_l2e(struct sk_buff **pskb, + struct ipvl_port *port) +{ + struct sk_buff *skb = *pskb; + struct ethhdr *eth = eth_hdr(skb); + rx_handler_result_t ret = RX_HANDLER_PASS; + + if (is_multicast_ether_addr(eth->h_dest)) { + if (ipvlan_external_frame(skb, port)) { + struct sk_buff *nskb = skb_clone(skb, GFP_ATOMIC); + + /* External frames are queued for device local + * distribution, but a copy is given to master + * straight away to avoid sending duplicates later + * when work-queue processes this frame. This is + * achieved by returning RX_HANDLER_PASS. + */ + if (nskb) { + ipvlan_skb_crossing_ns(nskb, NULL); + ipvlan_multicast_enqueue(port, nskb, false); + } + } + } else { + /* Perform like l3 mode for non-multicast packet */ + ret = ipvlan_handle_mode_l3(pskb, port); + } + + return ret; +} + rx_handler_result_t ipvlan_handle_frame(struct sk_buff **pskb) { struct sk_buff *skb = *pskb; @@ -746,6 +934,8 @@ rx_handler_result_t ipvlan_handle_frame(struct sk_buff **pskb) case IPVLAN_MODE_L3S: return RX_HANDLER_PASS; #endif + case IPVLAN_MODE_L2E: + return ipvlan_handle_mode_l2e(pskb, port); } /* Should not reach here */ diff --git a/drivers/net/ipvlan/ipvlan_main.c b/drivers/net/ipvlan/ipvlan_main.c index 60b7d93bb834..efd1452cd929 100644 --- a/drivers/net/ipvlan/ipvlan_main.c +++ b/drivers/net/ipvlan/ipvlan_main.c @@ -4,6 +4,10 @@ #include "ipvlan.h" +static int ipvlan_default_mode = IPVLAN_MODE_L3; +module_param(ipvlan_default_mode, int, 0400); +MODULE_PARM_DESC(ipvlan_default_mode, "set ipvlan default mode: 0 for l2, 1 for l3, 2 for l3s, 3 for l2e, others invalid now"); + static int ipvlan_set_port_mode(struct ipvl_port *port, u16 nval, struct netlink_ext_ack *extack) { @@ -535,7 +539,7 @@ int ipvlan_link_new(struct net *src_net, struct net_device *dev, struct ipvl_port *port; struct net_device *phy_dev; int err; - u16 mode = IPVLAN_MODE_L3; + u16 mode = ipvlan_default_mode; if (!tb[IFLA_LINK]) return -EINVAL; diff --git a/include/uapi/linux/if_link.h b/include/uapi/linux/if_link.h index c4b23f06f69e..50d4705e1cbc 100644 --- a/include/uapi/linux/if_link.h +++ b/include/uapi/linux/if_link.h @@ -690,6 +690,7 @@ enum ipvlan_mode { IPVLAN_MODE_L2 = 0, IPVLAN_MODE_L3, IPVLAN_MODE_L3S, + IPVLAN_MODE_L2E, IPVLAN_MODE_MAX }; -- 2.20.1

1 128

[PATCH OLK-5.10 v2] arm64: Revert feature: Add memmap parameter and register pmem
by Zhuling 27 Dec '21

27 Dec '21

euleros inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4O31I?from=project-issue CVE: NA The reserve memory of PMEM conflicts with "add memmap interface to reserved memory for mremap syscall usage", need to rollback, resubmit after adaptation. Feature related commit： 1.PMEM function commit:94dc364f5eda10f49449ba573dc3322e1ea92280 2.PMEM feature config commit: 36d7a831e15ceb84e937122c87d01c14242dc377 Signed-off-by: Zhuling <zhuling8(a)huawei.com> --- arch/arm64/Kconfig | 21 -------- arch/arm64/configs/openeuler_defconfig | 3 -- arch/arm64/kernel/Makefile | 1 - arch/arm64/kernel/pmem.c | 35 ------------- arch/arm64/kernel/setup.c | 10 ---- arch/arm64/mm/init.c | 94 ---------------------------------- drivers/nvdimm/Kconfig | 5 -- drivers/nvdimm/Makefile | 1 - 8 files changed, 170 deletions(-) delete mode 100644 arch/arm64/kernel/pmem.c diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig index df90a6e..2df4b31 100644 --- a/arch/arm64/Kconfig +++ b/arch/arm64/Kconfig @@ -1321,27 +1321,6 @@ config RODATA_FULL_DEFAULT_ENABLED This requires the linear region to be mapped down to pages, which may adversely affect performance in some cases. -config ARM64_PMEM_RESERVE - bool "Reserve memory for persistent storage" - default n - help - Use memmap=nn[KMG]!ss[KMG](memmap=100K!0x1a0000000) reserve - memory for persistent storage. - - Say y here to enable this feature. - -config ARM64_PMEM_LEGACY_DEVICE - bool "Create persistent storage" - depends on BLK_DEV - depends on LIBNVDIMM - select ARM64_PMEM_RESERVE - help - Use reserved memory for persistent storage when the kernel - restart or update. the data in PMEM will not be lost and - can be loaded faster. - - Say y if unsure. - config ARM64_SW_TTBR0_PAN bool "Emulate Privileged Access Never using TTBR0_EL1 switching" help diff --git a/arch/arm64/configs/openeuler_defconfig b/arch/arm64/configs/openeuler_defconfig index 17bc875..b5fc851 100644 --- a/arch/arm64/configs/openeuler_defconfig +++ b/arch/arm64/configs/openeuler_defconfig @@ -416,8 +416,6 @@ CONFIG_ARM64_CPU_PARK=y CONFIG_FORCE_MAX_ZONEORDER=11 CONFIG_UNMAP_KERNEL_AT_EL0=y CONFIG_RODATA_FULL_DEFAULT_ENABLED=y -CONFIG_ARM64_PMEM_RESERVE=y -CONFIG_ARM64_PMEM_LEGACY_DEVICE=y # CONFIG_ARM64_SW_TTBR0_PAN is not set CONFIG_ARM64_TAGGED_ADDR_ABI=y CONFIG_ARM64_ILP32=y @@ -6026,7 +6024,6 @@ CONFIG_ND_BTT=m CONFIG_BTT=y CONFIG_OF_PMEM=m CONFIG_NVDIMM_KEYS=y -CONFIG_PMEM_LEGACY=m CONFIG_DAX_DRIVER=y CONFIG_DAX=y CONFIG_DEV_DAX=m diff --git a/arch/arm64/kernel/Makefile b/arch/arm64/kernel/Makefile index f615325..169d90f 100644 --- a/arch/arm64/kernel/Makefile +++ b/arch/arm64/kernel/Makefile @@ -68,7 +68,6 @@ obj-$(CONFIG_ARM64_PTR_AUTH) += pointer_auth.o obj-$(CONFIG_SHADOW_CALL_STACK) += scs.o obj-$(CONFIG_ARM64_MTE) += mte.o obj-$(CONFIG_MPAM) += mpam/ -obj-$(CONFIG_ARM64_PMEM_LEGACY_DEVICE) += pmem.o obj-y += vdso/ probes/ obj-$(CONFIG_COMPAT_VDSO) += vdso32/ diff --git a/arch/arm64/kernel/pmem.c b/arch/arm64/kernel/pmem.c deleted file mode 100644 index 16eaf70..0000000 --- a/arch/arm64/kernel/pmem.c +++ /dev/null @@ -1,35 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -/* - * Copyright(c) 2021 Huawei Technologies Co., Ltd - * - * Derived from x86 and arm64 implement PMEM. - */ -#include <linux/platform_device.h> -#include <linux/init.h> -#include <linux/ioport.h> -#include <linux/module.h> - -static int found(struct resource *res, void *data) -{ - return 1; -} - -static int __init register_e820_pmem(void) -{ - struct platform_device *pdev; - int rc; - - rc = walk_iomem_res_desc(IORES_DESC_PERSISTENT_MEMORY_LEGACY, - IORESOURCE_MEM, 0, -1, NULL, found); - if (rc <= 0) - return 0; - - /* - * See drivers/nvdimm/e820.c for the implementation, this is - * simply here to trigger the module to load on demand. - */ - pdev = platform_device_alloc("e820_pmem", -1); - - return platform_device_add(pdev); -} -device_initcall(register_e820_pmem); diff --git a/arch/arm64/kernel/setup.c b/arch/arm64/kernel/setup.c index 88ec495..7cd0425 100644 --- a/arch/arm64/kernel/setup.c +++ b/arch/arm64/kernel/setup.c @@ -70,10 +70,6 @@ static int __init arm64_enable_cpu0_hotplug(char *str) __setup("arm64_cpu0_hotplug", arm64_enable_cpu0_hotplug); #endif -#ifdef CONFIG_ARM64_PMEM_RESERVE -extern struct resource pmem_res; -#endif - phys_addr_t __fdt_pointer __initdata; /* @@ -288,12 +284,6 @@ static void __init request_standard_resources(void) request_resource(res, &pin_memory_resource); #endif } - -#ifdef CONFIG_ARM64_PMEM_RESERVE - if (pmem_res.end && pmem_res.start) - request_resource(&iomem_resource, &pmem_res); -#endif - } static int __init reserve_memblock_reserved_regions(void) diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c index 3b9401e..e8d4461 100644 --- a/arch/arm64/mm/init.c +++ b/arch/arm64/mm/init.c @@ -55,7 +55,6 @@ */ s64 memstart_addr __ro_after_init = -1; EXPORT_SYMBOL(memstart_addr); -phys_addr_t start_at, mem_size; #ifdef CONFIG_PIN_MEMORY struct resource pin_memory_resource = { @@ -112,18 +111,6 @@ static void __init reserve_pin_memory_res(void) */ phys_addr_t arm64_dma_phys_limit __ro_after_init; -static unsigned long long pmem_size, pmem_start; - -#ifdef CONFIG_ARM64_PMEM_RESERVE -struct resource pmem_res = { - .name = "Persistent Memory (legacy)", - .start = 0, - .end = 0, - .flags = IORESOURCE_MEM, - .desc = IORES_DESC_PERSISTENT_MEMORY_LEGACY -}; -#endif - #ifndef CONFIG_KEXEC_CORE static void __init reserve_crashkernel(void) { @@ -417,83 +404,6 @@ static int __init reserve_park_mem(void) } #endif -static bool __init is_mem_valid(unsigned long long mem_size, unsigned long long mem_start) -{ - if (!memblock_is_region_memory(mem_start, mem_size)) { - pr_warn("cannot reserve mem: region is not memory!\n"); - return false; - } - - if (memblock_is_region_reserved(mem_start, mem_size)) { - pr_warn("cannot reserve mem: region overlaps reserved memory!\n"); - return false; - } - - if (!IS_ALIGNED(mem_start, SZ_2M)) { - pr_warn("cannot reserve mem: base address is not 2MB aligned!\n"); - return false; - } - - return true; -} - -static int __init parse_memmap_one(char *p) -{ - char *oldp; - - if (!p) - return -EINVAL; - - oldp = p; - mem_size = memparse(p, &p); - if (p == oldp) - return -EINVAL; - - if (!mem_size) - return -EINVAL; - - mem_size = PAGE_ALIGN(mem_size); - - if (*p == '!') { - start_at = memparse(p+1, &p); - - pmem_start = start_at; - pmem_size = mem_size; - } else - pr_info("Unrecognized memmap option, please check the parameter.\n"); - - return *p == '\0' ? 0 : -EINVAL; -} - -static int __init parse_memmap_opt(char *str) -{ - while (str) { - char *k = strchr(str, ','); - - if (k) - *k++ = 0; - parse_memmap_one(str); - str = k; - } - - return 0; -} -early_param("memmap", parse_memmap_opt); - -#ifdef CONFIG_ARM64_PMEM_RESERVE -static void __init reserve_pmem(void) -{ - if (!is_mem_valid(mem_size, start_at)) - return; - - memblock_remove(pmem_start, pmem_size); - pr_info("pmem reserved: 0x%016llx - 0x%016llx (%lld MB)\n", - pmem_start, pmem_start + pmem_size, pmem_size >> 20); - pmem_res.start = pmem_start; - pmem_res.end = pmem_start + pmem_size - 1; -} -#endif - void __init arm64_memblock_init(void) { const s64 linear_region_size = BIT(vabits_actual - 1); @@ -668,10 +578,6 @@ void __init bootmem_init(void) reserve_quick_kexec(); #endif -#ifdef CONFIG_ARM64_PMEM_RESERVE - reserve_pmem(); -#endif - reserve_pin_memory_res(); memblock_dump_all(); diff --git a/drivers/nvdimm/Kconfig b/drivers/nvdimm/Kconfig index ce4de75..b7d1eb3 100644 --- a/drivers/nvdimm/Kconfig +++ b/drivers/nvdimm/Kconfig @@ -132,8 +132,3 @@ config NVDIMM_TEST_BUILD infrastructure. endif - -config PMEM_LEGACY - tristate "Pmem_legacy" - select X86_PMEM_LEGACY if X86 - select ARM64_PMEM_LEGACY_DEVICE if ARM64 diff --git a/drivers/nvdimm/Makefile b/drivers/nvdimm/Makefile index 6f8dc92..0407753 100644 --- a/drivers/nvdimm/Makefile +++ b/drivers/nvdimm/Makefile @@ -3,7 +3,6 @@ obj-$(CONFIG_LIBNVDIMM) += libnvdimm.o obj-$(CONFIG_BLK_DEV_PMEM) += nd_pmem.o obj-$(CONFIG_ND_BTT) += nd_btt.o obj-$(CONFIG_ND_BLK) += nd_blk.o -obj-$(CONFIG_PMEM_LEGACY) += nd_e820.o obj-$(CONFIG_OF_PMEM) += of_pmem.o obj-$(CONFIG_VIRTIO_PMEM) += virtio_pmem.o nd_virtio.o -- 2.9.5

2 1

[PATCH openEuler-1.0-LTS 01/36] net: Prevent infinite while loop in skb_tx_hash()
by Yang Yingliang 27 Dec '21

27 Dec '21

From: Michael Chan <michael.chan(a)broadcom.com> stable inclusion from linux-4.19.215 commit 02302cbd52264337630a32848ac03648648e9685 -------------------------------- commit 0c57eeecc559ca6bc18b8c4e2808bc78dbe769b0 upstream. Drivers call netdev_set_num_tc() and then netdev_set_tc_queue() to set the queue count and offset for each TC. So the queue count and offset for the TCs may be zero for a short period after dev->num_tc has been set. If a TX packet is being transmitted at this time in the code path netdev_pick_tx() -> skb_tx_hash(), skb_tx_hash() may see nonzero dev->num_tc but zero qcount for the TC. The while loop that keeps looping while hash >= qcount will not end. Fix it by checking the TC's qcount to be nonzero before using it. Fixes: eadec877ce9c ("net: Add support for subordinate traffic classes to netdev_pick_tx") Reviewed-by: Andy Gospodarek <gospo(a)broadcom.com> Signed-off-by: Michael Chan <michael.chan(a)broadcom.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- net/core/dev.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/net/core/dev.c b/net/core/dev.c index 5d9800804d4a4..6ed7810ed7bd4 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -2846,6 +2846,12 @@ static u16 skb_tx_hash(const struct net_device *dev, qoffset = sb_dev->tc_to_txq[tc].offset; qcount = sb_dev->tc_to_txq[tc].count; + if (unlikely(!qcount)) { + net_warn_ratelimited("%s: invalid qcount, qoffset %u for tc %u\n", + sb_dev->name, qoffset, tc); + qoffset = 0; + qcount = dev->real_num_tx_queues; + } } if (skb_rx_queue_recorded(skb)) { -- 2.25.1

1 35

[[PATCH OLK-5.10] Revert feature — arm64: Add memmap parameter and register pmem] revert feature — [arm64: Add memmap parameter and register pmem]
by Zhuling 27 Dec '21

27 Dec '21

euleros inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4O31I?from=project-issue CVE: NA The reserve memory of PMEM conflicts with "add memmap interface to reserved memory for mremap syscall usage", need to rollback, resubmit after adaptation. Feature related commit： 1.PMEM function commit:94dc364f5eda10f49449ba573dc3322e1ea92280 2.PMEM feature config commit: 36d7a831e15ceb84e937122c87d01c14242dc377 Signed-off-by: Zhuling <zhuling8(a)huawei.com> --- arch/arm64/Kconfig | 21 -------- arch/arm64/configs/openeuler_defconfig | 3 -- arch/arm64/kernel/Makefile | 1 - arch/arm64/kernel/pmem.c | 35 ------------- arch/arm64/kernel/setup.c | 10 ---- arch/arm64/mm/init.c | 94 ---------------------------------- drivers/nvdimm/Kconfig | 5 -- drivers/nvdimm/Makefile | 1 - 8 files changed, 170 deletions(-) delete mode 100644 arch/arm64/kernel/pmem.c diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig index df90a6e..2df4b31 100644 --- a/arch/arm64/Kconfig +++ b/arch/arm64/Kconfig @@ -1321,27 +1321,6 @@ config RODATA_FULL_DEFAULT_ENABLED This requires the linear region to be mapped down to pages, which may adversely affect performance in some cases. -config ARM64_PMEM_RESERVE - bool "Reserve memory for persistent storage" - default n - help - Use memmap=nn[KMG]!ss[KMG](memmap=100K!0x1a0000000) reserve - memory for persistent storage. - - Say y here to enable this feature. - -config ARM64_PMEM_LEGACY_DEVICE - bool "Create persistent storage" - depends on BLK_DEV - depends on LIBNVDIMM - select ARM64_PMEM_RESERVE - help - Use reserved memory for persistent storage when the kernel - restart or update. the data in PMEM will not be lost and - can be loaded faster. - - Say y if unsure. - config ARM64_SW_TTBR0_PAN bool "Emulate Privileged Access Never using TTBR0_EL1 switching" help diff --git a/arch/arm64/configs/openeuler_defconfig b/arch/arm64/configs/openeuler_defconfig index 17bc875..b5fc851 100644 --- a/arch/arm64/configs/openeuler_defconfig +++ b/arch/arm64/configs/openeuler_defconfig @@ -416,8 +416,6 @@ CONFIG_ARM64_CPU_PARK=y CONFIG_FORCE_MAX_ZONEORDER=11 CONFIG_UNMAP_KERNEL_AT_EL0=y CONFIG_RODATA_FULL_DEFAULT_ENABLED=y -CONFIG_ARM64_PMEM_RESERVE=y -CONFIG_ARM64_PMEM_LEGACY_DEVICE=y # CONFIG_ARM64_SW_TTBR0_PAN is not set CONFIG_ARM64_TAGGED_ADDR_ABI=y CONFIG_ARM64_ILP32=y @@ -6026,7 +6024,6 @@ CONFIG_ND_BTT=m CONFIG_BTT=y CONFIG_OF_PMEM=m CONFIG_NVDIMM_KEYS=y -CONFIG_PMEM_LEGACY=m CONFIG_DAX_DRIVER=y CONFIG_DAX=y CONFIG_DEV_DAX=m diff --git a/arch/arm64/kernel/Makefile b/arch/arm64/kernel/Makefile index f615325..169d90f 100644 --- a/arch/arm64/kernel/Makefile +++ b/arch/arm64/kernel/Makefile @@ -68,7 +68,6 @@ obj-$(CONFIG_ARM64_PTR_AUTH) += pointer_auth.o obj-$(CONFIG_SHADOW_CALL_STACK) += scs.o obj-$(CONFIG_ARM64_MTE) += mte.o obj-$(CONFIG_MPAM) += mpam/ -obj-$(CONFIG_ARM64_PMEM_LEGACY_DEVICE) += pmem.o obj-y += vdso/ probes/ obj-$(CONFIG_COMPAT_VDSO) += vdso32/ diff --git a/arch/arm64/kernel/pmem.c b/arch/arm64/kernel/pmem.c deleted file mode 100644 index 16eaf70..0000000 --- a/arch/arm64/kernel/pmem.c +++ /dev/null @@ -1,35 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -/* - * Copyright(c) 2021 Huawei Technologies Co., Ltd - * - * Derived from x86 and arm64 implement PMEM. - */ -#include <linux/platform_device.h> -#include <linux/init.h> -#include <linux/ioport.h> -#include <linux/module.h> - -static int found(struct resource *res, void *data) -{ - return 1; -} - -static int __init register_e820_pmem(void) -{ - struct platform_device *pdev; - int rc; - - rc = walk_iomem_res_desc(IORES_DESC_PERSISTENT_MEMORY_LEGACY, - IORESOURCE_MEM, 0, -1, NULL, found); - if (rc <= 0) - return 0; - - /* - * See drivers/nvdimm/e820.c for the implementation, this is - * simply here to trigger the module to load on demand. - */ - pdev = platform_device_alloc("e820_pmem", -1); - - return platform_device_add(pdev); -} -device_initcall(register_e820_pmem); diff --git a/arch/arm64/kernel/setup.c b/arch/arm64/kernel/setup.c index 88ec495..7cd0425 100644 --- a/arch/arm64/kernel/setup.c +++ b/arch/arm64/kernel/setup.c @@ -70,10 +70,6 @@ static int __init arm64_enable_cpu0_hotplug(char *str) __setup("arm64_cpu0_hotplug", arm64_enable_cpu0_hotplug); #endif -#ifdef CONFIG_ARM64_PMEM_RESERVE -extern struct resource pmem_res; -#endif - phys_addr_t __fdt_pointer __initdata; /* @@ -288,12 +284,6 @@ static void __init request_standard_resources(void) request_resource(res, &pin_memory_resource); #endif } - -#ifdef CONFIG_ARM64_PMEM_RESERVE - if (pmem_res.end && pmem_res.start) - request_resource(&iomem_resource, &pmem_res); -#endif - } static int __init reserve_memblock_reserved_regions(void) diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c index 3b9401e..e8d4461 100644 --- a/arch/arm64/mm/init.c +++ b/arch/arm64/mm/init.c @@ -55,7 +55,6 @@ */ s64 memstart_addr __ro_after_init = -1; EXPORT_SYMBOL(memstart_addr); -phys_addr_t start_at, mem_size; #ifdef CONFIG_PIN_MEMORY struct resource pin_memory_resource = { @@ -112,18 +111,6 @@ static void __init reserve_pin_memory_res(void) */ phys_addr_t arm64_dma_phys_limit __ro_after_init; -static unsigned long long pmem_size, pmem_start; - -#ifdef CONFIG_ARM64_PMEM_RESERVE -struct resource pmem_res = { - .name = "Persistent Memory (legacy)", - .start = 0, - .end = 0, - .flags = IORESOURCE_MEM, - .desc = IORES_DESC_PERSISTENT_MEMORY_LEGACY -}; -#endif - #ifndef CONFIG_KEXEC_CORE static void __init reserve_crashkernel(void) { @@ -417,83 +404,6 @@ static int __init reserve_park_mem(void) } #endif -static bool __init is_mem_valid(unsigned long long mem_size, unsigned long long mem_start) -{ - if (!memblock_is_region_memory(mem_start, mem_size)) { - pr_warn("cannot reserve mem: region is not memory!\n"); - return false; - } - - if (memblock_is_region_reserved(mem_start, mem_size)) { - pr_warn("cannot reserve mem: region overlaps reserved memory!\n"); - return false; - } - - if (!IS_ALIGNED(mem_start, SZ_2M)) { - pr_warn("cannot reserve mem: base address is not 2MB aligned!\n"); - return false; - } - - return true; -} - -static int __init parse_memmap_one(char *p) -{ - char *oldp; - - if (!p) - return -EINVAL; - - oldp = p; - mem_size = memparse(p, &p); - if (p == oldp) - return -EINVAL; - - if (!mem_size) - return -EINVAL; - - mem_size = PAGE_ALIGN(mem_size); - - if (*p == '!') { - start_at = memparse(p+1, &p); - - pmem_start = start_at; - pmem_size = mem_size; - } else - pr_info("Unrecognized memmap option, please check the parameter.\n"); - - return *p == '\0' ? 0 : -EINVAL; -} - -static int __init parse_memmap_opt(char *str) -{ - while (str) { - char *k = strchr(str, ','); - - if (k) - *k++ = 0; - parse_memmap_one(str); - str = k; - } - - return 0; -} -early_param("memmap", parse_memmap_opt); - -#ifdef CONFIG_ARM64_PMEM_RESERVE -static void __init reserve_pmem(void) -{ - if (!is_mem_valid(mem_size, start_at)) - return; - - memblock_remove(pmem_start, pmem_size); - pr_info("pmem reserved: 0x%016llx - 0x%016llx (%lld MB)\n", - pmem_start, pmem_start + pmem_size, pmem_size >> 20); - pmem_res.start = pmem_start; - pmem_res.end = pmem_start + pmem_size - 1; -} -#endif - void __init arm64_memblock_init(void) { const s64 linear_region_size = BIT(vabits_actual - 1); @@ -668,10 +578,6 @@ void __init bootmem_init(void) reserve_quick_kexec(); #endif -#ifdef CONFIG_ARM64_PMEM_RESERVE - reserve_pmem(); -#endif - reserve_pin_memory_res(); memblock_dump_all(); diff --git a/drivers/nvdimm/Kconfig b/drivers/nvdimm/Kconfig index ce4de75..b7d1eb3 100644 --- a/drivers/nvdimm/Kconfig +++ b/drivers/nvdimm/Kconfig @@ -132,8 +132,3 @@ config NVDIMM_TEST_BUILD infrastructure. endif - -config PMEM_LEGACY - tristate "Pmem_legacy" - select X86_PMEM_LEGACY if X86 - select ARM64_PMEM_LEGACY_DEVICE if ARM64 diff --git a/drivers/nvdimm/Makefile b/drivers/nvdimm/Makefile index 6f8dc92..0407753 100644 --- a/drivers/nvdimm/Makefile +++ b/drivers/nvdimm/Makefile @@ -3,7 +3,6 @@ obj-$(CONFIG_LIBNVDIMM) += libnvdimm.o obj-$(CONFIG_BLK_DEV_PMEM) += nd_pmem.o obj-$(CONFIG_ND_BTT) += nd_btt.o obj-$(CONFIG_ND_BLK) += nd_blk.o -obj-$(CONFIG_PMEM_LEGACY) += nd_e820.o obj-$(CONFIG_OF_PMEM) += of_pmem.o obj-$(CONFIG_VIRTIO_PMEM) += virtio_pmem.o nd_virtio.o -- 2.9.5

4 4

[PATCH openEuler-1.0-LTS] watchdog: Fix check_preemption_disabled() error
by Yang Yingliang 27 Dec '21

27 Dec '21

From: Wei Li <liwei391(a)huawei.com> hulk inclusion category: bugfix bugzilla: 173968, https://gitee.com/openeuler/kernel/issues/I3J87Y CVE: NA ------------------------------------------------- When enabling CONFIG_DEBUG_PREEMPT and CONFIG_PREEMPT, it triggers a 'BUG' in the pmu based nmi_watchdog initializaion: [ 3.341853] BUG: using smp_processor_id() in preemptible [00000000] code: swapper/0/1 [ 3.344392] caller is debug_smp_processor_id+0x17/0x20 [ 3.344395] CPU: 1 PID: 1 Comm: swapper/0 Not tainted 5.10.0+ #398 [ 3.344397] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.10.2-0-g5f4c7b1-prebuilt.qemu-project.org 04/01/2014 [ 3.344399] Call Trace: [ 3.344410] dump_stack+0x60/0x76 [ 3.344412] check_preemption_disabled+0xba/0xc0 [ 3.344415] debug_smp_processor_id+0x17/0x20 [ 3.344422] hardlockup_detector_event_create+0xf/0x60 [ 3.344427] hardlockup_detector_perf_init+0xf/0x41 [ 3.344430] watchdog_nmi_probe+0xe/0x10 [ 3.344432] lockup_detector_init+0x22/0x5b [ 3.344437] kernel_init_freeable+0x20c/0x245 [ 3.344439] ? rest_init+0xd0/0xd0 [ 3.344441] kernel_init+0xe/0x110 [ 3.344446] ret_from_fork+0x22/0x30 This issue was introduced by commit 141482cb4b01, which move down lockup_detector_init() after do_basic_setup(), after sched_init_smp() too. hardlockup_detector_event_create |- hardlockup_detector_perf_init (unsafe) |- watchdog_nmi_probe |- lockup_detector_init |- hardlockup_detector_perf_enable |- watchdog_nmi_enable |- watchdog_enable |- lockup_detector_online_cpu |- softlockup_start_fn |- softlockup_start_all |- lockup_detector_reconfigure |- lockup_detector_setup |- lockup_detector_init After analysing the calling context, it's only unsafe to use smp_processor_id() in hardlockup_detector_perf_init() as the thread 'kernel_init' is preemptible after sched_init_smp(). While it is just a test if we can enable the pmu based nmi_watchdog, the real enabling process is in softlockup_start_fn() later which ensures that watchdog_enable() is called on all cores. So it's free to disable preempt to fix this 'BUG'. Fixes: 141482cb4b01 ("lockup_detector: init lockup detector after all the init_calls") Signed-off-by: Wei Li <liwei391(a)huawei.com> Reviewed-by: Xiongfeng Wang <wangxiongfeng2(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- kernel/watchdog_hld.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/kernel/watchdog_hld.c b/kernel/watchdog_hld.c index 43832b1023693..a5aff8ffa48ce 100644 --- a/kernel/watchdog_hld.c +++ b/kernel/watchdog_hld.c @@ -508,14 +508,17 @@ void __init hardlockup_detector_perf_restart(void) */ int __init hardlockup_detector_perf_init(void) { - int ret = hardlockup_detector_event_create(); + int ret; + preempt_disable(); + ret = hardlockup_detector_event_create(); if (ret) { pr_info("Perf NMI watchdog permanently disabled\n"); } else { perf_event_release_kernel(this_cpu_read(watchdog_ev)); this_cpu_write(watchdog_ev, NULL); } + preempt_enable(); return ret; } #endif /* CONFIG_HARDLOCKUP_DETECTOR_PERF */ -- 2.25.1

1 0

[PATCH openEuler-1.0-LTS] btrfs: unlock newly allocated extent buffer after error
by Yang Yingliang 27 Dec '21

27 Dec '21

From: Qu Wenruo <wqu(a)suse.com> mainline inclusion from mainline-v5.15-rc6 commit 19ea40dddf1833db868533958ca066f368862211 category: bugfix bugzilla: NA CVE: CVE-2021-4149 -------------------------------- [BUG] There is a bug report that injected ENOMEM error could leave a tree block locked while we return to user-space: BTRFS info (device loop0): enabling ssd optimizations FAULT_INJECTION: forcing a failure. name failslab, interval 1, probability 0, space 0, times 0 CPU: 0 PID: 7579 Comm: syz-executor Not tainted 5.15.0-rc1 #16 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.12.0-59-gc9ba5276e321-prebuilt.qemu.org 04/01/2014 Call Trace: __dump_stack lib/dump_stack.c:88 [inline] dump_stack_lvl+0x8d/0xcf lib/dump_stack.c:106 fail_dump lib/fault-inject.c:52 [inline] should_fail+0x13c/0x160 lib/fault-inject.c:146 should_failslab+0x5/0x10 mm/slab_common.c:1328 slab_pre_alloc_hook.constprop.99+0x4e/0xc0 mm/slab.h:494 slab_alloc_node mm/slub.c:3120 [inline] slab_alloc mm/slub.c:3214 [inline] kmem_cache_alloc+0x44/0x280 mm/slub.c:3219 btrfs_alloc_delayed_extent_op fs/btrfs/delayed-ref.h:299 [inline] btrfs_alloc_tree_block+0x38c/0x670 fs/btrfs/extent-tree.c:4833 __btrfs_cow_block+0x16f/0x7d0 fs/btrfs/ctree.c:415 btrfs_cow_block+0x12a/0x300 fs/btrfs/ctree.c:570 btrfs_search_slot+0x6b0/0xee0 fs/btrfs/ctree.c:1768 btrfs_insert_empty_items+0x80/0xf0 fs/btrfs/ctree.c:3905 btrfs_new_inode+0x311/0xa60 fs/btrfs/inode.c:6530 btrfs_create+0x12b/0x270 fs/btrfs/inode.c:6783 lookup_open+0x660/0x780 fs/namei.c:3282 open_last_lookups fs/namei.c:3352 [inline] path_openat+0x465/0xe20 fs/namei.c:3557 do_filp_open+0xe3/0x170 fs/namei.c:3588 do_sys_openat2+0x357/0x4a0 fs/open.c:1200 do_sys_open+0x87/0xd0 fs/open.c:1216 do_syscall_x64 arch/x86/entry/common.c:50 [inline] do_syscall_64+0x34/0xb0 arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x44/0xae RIP: 0033:0x46ae99 Code: f7 d8 64 89 02 b8 ff ff ff ff c3 66 0f 1f 44 00 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 bc ff ff ff f7 d8 64 89 01 48 RSP: 002b:00007f46711b9c48 EFLAGS: 00000246 ORIG_RAX: 0000000000000055 RAX: ffffffffffffffda RBX: 000000000078c0a0 RCX: 000000000046ae99 RDX: 0000000000000000 RSI: 00000000000000a1 RDI: 0000000020005800 RBP: 00007f46711b9c80 R08: 0000000000000000 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000017 R13: 0000000000000000 R14: 000000000078c0a0 R15: 00007ffc129da6e0 ================================================ WARNING: lock held when returning to user space! 5.15.0-rc1 #16 Not tainted ------------------------------------------------ syz-executor/7579 is leaving the kernel with locks still held! 1 lock held by syz-executor/7579: #0: ffff888104b73da8 (btrfs-tree-01/1){+.+.}-{3:3}, at: __btrfs_tree_lock+0x2e/0x1a0 fs/btrfs/locking.c:112 [CAUSE] In btrfs_alloc_tree_block(), after btrfs_init_new_buffer(), the new extent buffer @buf is locked, but if later operations like adding delayed tree ref fail, we just free @buf without unlocking it, resulting above warning. [FIX] Unlock @buf in out_free_buf: label. Reported-by: Hao Sun <sunhao.th(a)gmail.com> Link: https://lore.kernel.org/linux-btrfs/CACkBjsZ9O6Zr0KK1yGn=1rQi6Crh1yeCRdTSBx… CC: stable(a)vger.kernel.org # 5.4+ Signed-off-by: Qu Wenruo <wqu(a)suse.com> Reviewed-by: David Sterba <dsterba(a)suse.com> Signed-off-by: David Sterba <dsterba(a)suse.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> Reviewed-by: Xiu Jianfeng <xiujianfeng(a)huawei.com> Reviewed-by: Jason Yan <yanaijie(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- fs/btrfs/extent-tree.c | 1 + 1 file changed, 1 insertion(+) diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 4bf93184362ec..dcbb76b62a0b0 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -8324,6 +8324,7 @@ struct extent_buffer *btrfs_alloc_tree_block(struct btrfs_trans_handle *trans, out_free_delayed: btrfs_free_delayed_extent_op(extent_op); out_free_buf: + btrfs_tree_unlock(buf); free_extent_buffer(buf); out_free_reserved: btrfs_free_reserved_extent(fs_info, ins.objectid, ins.offset, 0); -- 2.25.1

1 0

[PATCH openEuler-1.0-LTS] ext4: Fix null-ptr-deref in '__ext4_journal_ensure_credits'
by Yang Yingliang 25 Dec '21

25 Dec '21

From: Ye Bin <yebin10(a)huawei.com> hulk inclusion category: bugfix bugzilla: 185945, https://gitee.com/openeuler/kernel/issues/I4O33F CVE: NA ----------------------------------------------- We got issue as follows when run syzkaller test: [ 1901.130043] EXT4-fs error (device vda): ext4_remount:5624: comm syz-executor.5: Abort forced by user [ 1901.130901] Aborting journal on device vda-8. [ 1901.131437] EXT4-fs error (device vda): ext4_journal_check_start:61: comm syz-executor.16: Detected aborted journal [ 1901.131566] EXT4-fs error (device vda): ext4_journal_check_start:61: comm syz-executor.11: Detected aborted journal [ 1901.132586] EXT4-fs error (device vda): ext4_journal_check_start:61: comm syz-executor.18: Detected aborted journal [ 1901.132751] EXT4-fs error (device vda): ext4_journal_check_start:61: comm syz-executor.9: Detected aborted journal [ 1901.136149] EXT4-fs error (device vda) in ext4_reserve_inode_write:6035: Journal has aborted [ 1901.136837] EXT4-fs error (device vda): ext4_journal_check_start:61: comm syz-fuzzer: Detected aborted journal [ 1901.136915] ================================================================== [ 1901.138175] BUG: KASAN: null-ptr-deref in __ext4_journal_ensure_credits+0x74/0x140 [ext4] [ 1901.138343] EXT4-fs error (device vda): ext4_journal_check_start:61: comm syz-executor.13: Detected aborted journal [ 1901.138398] EXT4-fs error (device vda): ext4_journal_check_start:61: comm syz-executor.1: Detected aborted journal [ 1901.138808] Read of size 8 at addr 0000000000000000 by task syz-executor.17/968 [ 1901.138817] [ 1901.138852] EXT4-fs error (device vda): ext4_journal_check_start:61: comm syz-executor.30: Detected aborted journal [ 1901.144779] CPU: 1 PID: 968 Comm: syz-executor.17 Not tainted 4.19.90-vhulk2111.1.0.h893.eulerosv2r10.aarch64+ #1 [ 1901.146479] Hardware name: linux,dummy-virt (DT) [ 1901.147317] Call trace: [ 1901.147552] dump_backtrace+0x0/0x2d8 [ 1901.147898] show_stack+0x28/0x38 [ 1901.148215] dump_stack+0xec/0x15c [ 1901.148746] kasan_report+0x108/0x338 [ 1901.149207] __asan_load8+0x58/0xb0 [ 1901.149753] __ext4_journal_ensure_credits+0x74/0x140 [ext4] [ 1901.150579] ext4_xattr_delete_inode+0xe4/0x700 [ext4] [ 1901.151316] ext4_evict_inode+0x524/0xba8 [ext4] [ 1901.151985] evict+0x1a4/0x378 [ 1901.152353] iput+0x310/0x428 [ 1901.152733] do_unlinkat+0x260/0x428 [ 1901.153056] __arm64_sys_unlinkat+0x6c/0xc0 [ 1901.153455] el0_svc_common+0xc8/0x320 [ 1901.153799] el0_svc_handler+0xf8/0x160 [ 1901.154265] el0_svc+0x10/0x218 [ 1901.154682] ================================================================== This issue may happens like this: Process1 Process2 ext4_evict_inode ext4_journal_start ext4_truncate ext4_ind_truncate ext4_free_branches ext4_ind_truncate_ensure_credits ext4_journal_ensure_credits_fn ext4_journal_restart handle->h_transaction = NULL; mount -o remount,abort /mnt -> trigger JBD abort start_this_handle -> will return failed ext4_xattr_delete_inode ext4_journal_ensure_credits ext4_journal_ensure_credits_fn __ext4_journal_ensure_credits jbd2_handle_buffer_credits journal = handle->h_transaction->t_journal; ->null-ptr-deref Now, indirect truncate process didn't handle error. To solve this issue maybe simply add check handle is abort in '__ext4_journal_ensure_credits' is enough, and i also think this is necessary. Signed-off-by: Ye Bin <yebin10(a)huawei.com> Reviewed-by: Zhang Yi <yi.zhang(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- fs/ext4/ext4_jbd2.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c index 022ab8afd9499..2c6dc99292fdc 100644 --- a/fs/ext4/ext4_jbd2.c +++ b/fs/ext4/ext4_jbd2.c @@ -140,6 +140,8 @@ int __ext4_journal_ensure_credits(handle_t *handle, int check_cred, { if (!ext4_handle_valid(handle)) return 0; + if (is_handle_aborted(handle)) + return -EROFS; if (jbd2_handle_buffer_credits(handle) >= check_cred && handle->h_revoke_credits >= revoke_cred) return 0; -- 2.25.1

1 0

[[PATCH OLK-5.10] Revert feature — [arm64: Add memmap parameter and register pmem]] revert feature — [arm64: Add memmap parameter and register pmem]
by Zhuling 25 Dec '21

25 Dec '21

euleros inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4O31I?from=project-issue CVE: NA The reserve memory of PMEM conflicts with "add memmap interface to reserved memory for mremap syscall usage", need to rollback, resubmit after adaptation. Feature related commit： 1.PMEM function commit:94dc364f5eda10f49449ba573dc3322e1ea92280 2.PMEM feature config commit: 36d7a831e15ceb84e937122c87d01c14242dc377 Signed-off-by: Zhuling <zhuling8(a)huawei.com> --- arch/arm64/Kconfig | 21 -------- arch/arm64/configs/openeuler_defconfig | 3 -- arch/arm64/kernel/Makefile | 1 - arch/arm64/kernel/pmem.c | 35 ------------- arch/arm64/kernel/setup.c | 10 ---- arch/arm64/mm/init.c | 94 ---------------------------------- drivers/nvdimm/Kconfig | 5 -- drivers/nvdimm/Makefile | 1 - 8 files changed, 170 deletions(-) delete mode 100644 arch/arm64/kernel/pmem.c diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig index df90a6e..2df4b31 100644 --- a/arch/arm64/Kconfig +++ b/arch/arm64/Kconfig @@ -1321,27 +1321,6 @@ config RODATA_FULL_DEFAULT_ENABLED This requires the linear region to be mapped down to pages, which may adversely affect performance in some cases. -config ARM64_PMEM_RESERVE - bool "Reserve memory for persistent storage" - default n - help - Use memmap=nn[KMG]!ss[KMG](memmap=100K!0x1a0000000) reserve - memory for persistent storage. - - Say y here to enable this feature. - -config ARM64_PMEM_LEGACY_DEVICE - bool "Create persistent storage" - depends on BLK_DEV - depends on LIBNVDIMM - select ARM64_PMEM_RESERVE - help - Use reserved memory for persistent storage when the kernel - restart or update. the data in PMEM will not be lost and - can be loaded faster. - - Say y if unsure. - config ARM64_SW_TTBR0_PAN bool "Emulate Privileged Access Never using TTBR0_EL1 switching" help diff --git a/arch/arm64/configs/openeuler_defconfig b/arch/arm64/configs/openeuler_defconfig index 17bc875..b5fc851 100644 --- a/arch/arm64/configs/openeuler_defconfig +++ b/arch/arm64/configs/openeuler_defconfig @@ -416,8 +416,6 @@ CONFIG_ARM64_CPU_PARK=y CONFIG_FORCE_MAX_ZONEORDER=11 CONFIG_UNMAP_KERNEL_AT_EL0=y CONFIG_RODATA_FULL_DEFAULT_ENABLED=y -CONFIG_ARM64_PMEM_RESERVE=y -CONFIG_ARM64_PMEM_LEGACY_DEVICE=y # CONFIG_ARM64_SW_TTBR0_PAN is not set CONFIG_ARM64_TAGGED_ADDR_ABI=y CONFIG_ARM64_ILP32=y @@ -6026,7 +6024,6 @@ CONFIG_ND_BTT=m CONFIG_BTT=y CONFIG_OF_PMEM=m CONFIG_NVDIMM_KEYS=y -CONFIG_PMEM_LEGACY=m CONFIG_DAX_DRIVER=y CONFIG_DAX=y CONFIG_DEV_DAX=m diff --git a/arch/arm64/kernel/Makefile b/arch/arm64/kernel/Makefile index f615325..169d90f 100644 --- a/arch/arm64/kernel/Makefile +++ b/arch/arm64/kernel/Makefile @@ -68,7 +68,6 @@ obj-$(CONFIG_ARM64_PTR_AUTH) += pointer_auth.o obj-$(CONFIG_SHADOW_CALL_STACK) += scs.o obj-$(CONFIG_ARM64_MTE) += mte.o obj-$(CONFIG_MPAM) += mpam/ -obj-$(CONFIG_ARM64_PMEM_LEGACY_DEVICE) += pmem.o obj-y += vdso/ probes/ obj-$(CONFIG_COMPAT_VDSO) += vdso32/ diff --git a/arch/arm64/kernel/pmem.c b/arch/arm64/kernel/pmem.c deleted file mode 100644 index 16eaf70..0000000 --- a/arch/arm64/kernel/pmem.c +++ /dev/null @@ -1,35 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -/* - * Copyright(c) 2021 Huawei Technologies Co., Ltd - * - * Derived from x86 and arm64 implement PMEM. - */ -#include <linux/platform_device.h> -#include <linux/init.h> -#include <linux/ioport.h> -#include <linux/module.h> - -static int found(struct resource *res, void *data) -{ - return 1; -} - -static int __init register_e820_pmem(void) -{ - struct platform_device *pdev; - int rc; - - rc = walk_iomem_res_desc(IORES_DESC_PERSISTENT_MEMORY_LEGACY, - IORESOURCE_MEM, 0, -1, NULL, found); - if (rc <= 0) - return 0; - - /* - * See drivers/nvdimm/e820.c for the implementation, this is - * simply here to trigger the module to load on demand. - */ - pdev = platform_device_alloc("e820_pmem", -1); - - return platform_device_add(pdev); -} -device_initcall(register_e820_pmem); diff --git a/arch/arm64/kernel/setup.c b/arch/arm64/kernel/setup.c index 88ec495..7cd0425 100644 --- a/arch/arm64/kernel/setup.c +++ b/arch/arm64/kernel/setup.c @@ -70,10 +70,6 @@ static int __init arm64_enable_cpu0_hotplug(char *str) __setup("arm64_cpu0_hotplug", arm64_enable_cpu0_hotplug); #endif -#ifdef CONFIG_ARM64_PMEM_RESERVE -extern struct resource pmem_res; -#endif - phys_addr_t __fdt_pointer __initdata; /* @@ -288,12 +284,6 @@ static void __init request_standard_resources(void) request_resource(res, &pin_memory_resource); #endif } - -#ifdef CONFIG_ARM64_PMEM_RESERVE - if (pmem_res.end && pmem_res.start) - request_resource(&iomem_resource, &pmem_res); -#endif - } static int __init reserve_memblock_reserved_regions(void) diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c index 3b9401e..e8d4461 100644 --- a/arch/arm64/mm/init.c +++ b/arch/arm64/mm/init.c @@ -55,7 +55,6 @@ */ s64 memstart_addr __ro_after_init = -1; EXPORT_SYMBOL(memstart_addr); -phys_addr_t start_at, mem_size; #ifdef CONFIG_PIN_MEMORY struct resource pin_memory_resource = { @@ -112,18 +111,6 @@ static void __init reserve_pin_memory_res(void) */ phys_addr_t arm64_dma_phys_limit __ro_after_init; -static unsigned long long pmem_size, pmem_start; - -#ifdef CONFIG_ARM64_PMEM_RESERVE -struct resource pmem_res = { - .name = "Persistent Memory (legacy)", - .start = 0, - .end = 0, - .flags = IORESOURCE_MEM, - .desc = IORES_DESC_PERSISTENT_MEMORY_LEGACY -}; -#endif - #ifndef CONFIG_KEXEC_CORE static void __init reserve_crashkernel(void) { @@ -417,83 +404,6 @@ static int __init reserve_park_mem(void) } #endif -static bool __init is_mem_valid(unsigned long long mem_size, unsigned long long mem_start) -{ - if (!memblock_is_region_memory(mem_start, mem_size)) { - pr_warn("cannot reserve mem: region is not memory!\n"); - return false; - } - - if (memblock_is_region_reserved(mem_start, mem_size)) { - pr_warn("cannot reserve mem: region overlaps reserved memory!\n"); - return false; - } - - if (!IS_ALIGNED(mem_start, SZ_2M)) { - pr_warn("cannot reserve mem: base address is not 2MB aligned!\n"); - return false; - } - - return true; -} - -static int __init parse_memmap_one(char *p) -{ - char *oldp; - - if (!p) - return -EINVAL; - - oldp = p; - mem_size = memparse(p, &p); - if (p == oldp) - return -EINVAL; - - if (!mem_size) - return -EINVAL; - - mem_size = PAGE_ALIGN(mem_size); - - if (*p == '!') { - start_at = memparse(p+1, &p); - - pmem_start = start_at; - pmem_size = mem_size; - } else - pr_info("Unrecognized memmap option, please check the parameter.\n"); - - return *p == '\0' ? 0 : -EINVAL; -} - -static int __init parse_memmap_opt(char *str) -{ - while (str) { - char *k = strchr(str, ','); - - if (k) - *k++ = 0; - parse_memmap_one(str); - str = k; - } - - return 0; -} -early_param("memmap", parse_memmap_opt); - -#ifdef CONFIG_ARM64_PMEM_RESERVE -static void __init reserve_pmem(void) -{ - if (!is_mem_valid(mem_size, start_at)) - return; - - memblock_remove(pmem_start, pmem_size); - pr_info("pmem reserved: 0x%016llx - 0x%016llx (%lld MB)\n", - pmem_start, pmem_start + pmem_size, pmem_size >> 20); - pmem_res.start = pmem_start; - pmem_res.end = pmem_start + pmem_size - 1; -} -#endif - void __init arm64_memblock_init(void) { const s64 linear_region_size = BIT(vabits_actual - 1); @@ -668,10 +578,6 @@ void __init bootmem_init(void) reserve_quick_kexec(); #endif -#ifdef CONFIG_ARM64_PMEM_RESERVE - reserve_pmem(); -#endif - reserve_pin_memory_res(); memblock_dump_all(); diff --git a/drivers/nvdimm/Kconfig b/drivers/nvdimm/Kconfig index ce4de75..b7d1eb3 100644 --- a/drivers/nvdimm/Kconfig +++ b/drivers/nvdimm/Kconfig @@ -132,8 +132,3 @@ config NVDIMM_TEST_BUILD infrastructure. endif - -config PMEM_LEGACY - tristate "Pmem_legacy" - select X86_PMEM_LEGACY if X86 - select ARM64_PMEM_LEGACY_DEVICE if ARM64 diff --git a/drivers/nvdimm/Makefile b/drivers/nvdimm/Makefile index 6f8dc92..0407753 100644 --- a/drivers/nvdimm/Makefile +++ b/drivers/nvdimm/Makefile @@ -3,7 +3,6 @@ obj-$(CONFIG_LIBNVDIMM) += libnvdimm.o obj-$(CONFIG_BLK_DEV_PMEM) += nd_pmem.o obj-$(CONFIG_ND_BTT) += nd_btt.o obj-$(CONFIG_ND_BLK) += nd_blk.o -obj-$(CONFIG_PMEM_LEGACY) += nd_e820.o obj-$(CONFIG_OF_PMEM) += of_pmem.o obj-$(CONFIG_VIRTIO_PMEM) += virtio_pmem.o nd_virtio.o -- 2.9.5

1 0

[PATCH openEuler-1.0-LTS] net/hinic: Fix call trace when the rx_buff module parameter is grater than 2
by Yang Yingliang 25 Dec '21

25 Dec '21

From: Chiqijun <chiqijun(a)huawei.com> driver inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I4O2ZZ ----------------------------------------------------------------------- When rx_buff is greater than 2, the driver will alloc for more than 1 page of memory for network rx, but the __GFP_COMP gfp flag is not set, resulting in the following call trace: CPU: 3 PID: 494041 Comm: ping Kdump: loaded Tainted: G W OE 4.19.90-2106.3.0.0095.oe1.x86_64 #1 Hardware name: Huawei Technologies Co., Ltd. RH2288H V3/BC11HGSA0, BIOS 5.15 05/21/2019 RIP: 0010:copy_page_to_iter+0x154/0x310 Code: 31 b8 00 10 00 00 f7 c6 00 80 00 00 74 07 0f b6 49 51 48 d3 e0 48 39 c2 0f 86 ed fe ff ff 48 c7 c7 30 RSP: 0018:ffffbd6907d03bd8 EFLAGS: 00010286 RAX: 0000000000000024 RBX: ffffe0ffee5b3000 RCX: 0000000000000000 RDX: 0000000000000000 RSI: ffff9edbbfcd6858 RDI: ffff9edbbfcd6858 RBP: 0000000000000001 R08: 000000000001574a R09: 0000000000000004 R10: 000000000000004e R11: 0000000000000001 R12: ffffbd6907d03ed0 R13: 0000000000002100 R14: 0000000000000030 R15: 0000000000000000 FS: 00007f9d37244dc0(0000) GS:ffff9edbbfcc0000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00007ffe0e715f80 CR3: 000000203c018005 CR4: 00000000001606e0 Call Trace: skb_copy_datagram_iter+0x16c/0x2a0 raw_recvmsg+0xd0/0x1f0 inet_recvmsg+0x5b/0xd0 ____sys_recvmsg+0x95/0x160 ? import_iovec+0x37/0xd0 ? copy_msghdr_from_user+0x5c/0x90 ___sys_recvmsg+0x8c/0xd0 ? __audit_syscall_exit+0x228/0x290 ? kretprobe_trampoline+0x25/0x50 ? __sys_recvmsg+0x5b/0xa0 __sys_recvmsg+0x5b/0xa0 do_syscall_64+0x5f/0x240 entry_SYSCALL_64_after_hwframe+0x44/0xa9 Use 'dev_alloc_pages' instead of calling ’alloc_pages_node‘ directly. Signed-off-by: Chiqijun <chiqijun(a)huawei.com> Reviewed-by: Wangxiaoyun <cloud.wangxiaoyun(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/net/ethernet/huawei/hinic/hinic_rx.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/huawei/hinic/hinic_rx.c b/drivers/net/ethernet/huawei/hinic/hinic_rx.c index a9047432f3053..019cd439ce696 100644 --- a/drivers/net/ethernet/huawei/hinic/hinic_rx.c +++ b/drivers/net/ethernet/huawei/hinic/hinic_rx.c @@ -67,7 +67,7 @@ static bool rx_alloc_mapped_page(struct hinic_rxq *rxq, return true; /* alloc new page for storage */ - page = alloc_pages_node(NUMA_NO_NODE, GFP_ATOMIC, nic_dev->page_order); + page = dev_alloc_pages(nic_dev->page_order); if (unlikely(!page)) { RXQ_STATS_INC(rxq, alloc_rx_buf_err); return false; -- 2.25.1

2 1

[PATCH openEuler-1.0-LTS 1/3] arm64/mpam: remove __init macro to support driver probe
by Yang Yingliang 24 Dec '21

24 Dec '21

From: Xingang Wang <wangxingang5(a)huawei.com> ascend inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I49RB2 CVE: NA --------------------------------------------------- To support device tree boot for arm64 mpam, the __init macro might be used by the dts driver. This remove the necessary __init macro for the related functions. Signed-off-by: Xingang Wang <wangxingang5(a)huawei.com> Reviewed-by: Wang ShaoBo <bobo.shaobowang(a)huawei.com> Acked-by: Xie XiuQi <xiexiuqi(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- arch/arm64/include/asm/resctrl.h | 2 +- arch/arm64/kernel/mpam/mpam_device.c | 14 +++++++------- arch/arm64/kernel/mpam/mpam_internal.h | 2 +- arch/arm64/kernel/mpam/mpam_resctrl.c | 2 +- fs/resctrlfs.c | 2 +- include/linux/arm_mpam.h | 14 +++++++------- 6 files changed, 18 insertions(+), 18 deletions(-) diff --git a/arch/arm64/include/asm/resctrl.h b/arch/arm64/include/asm/resctrl.h index ff4ce56430d1f..3fceaa482e5a3 100644 --- a/arch/arm64/include/asm/resctrl.h +++ b/arch/arm64/include/asm/resctrl.h @@ -424,7 +424,7 @@ int resctrl_update_groups_config(struct rdtgroup *rdtgrp); #define RESCTRL_MAX_CLOSID 32 -int __init resctrl_group_init(void); +int resctrl_group_init(void); void post_resctrl_mount(void); diff --git a/arch/arm64/kernel/mpam/mpam_device.c b/arch/arm64/kernel/mpam/mpam_device.c index c0615f6947a1f..4139b4476a836 100644 --- a/arch/arm64/kernel/mpam/mpam_device.c +++ b/arch/arm64/kernel/mpam/mpam_device.c @@ -534,7 +534,7 @@ static void mpam_disable_irqs(void) * Scheduled by mpam_discovery_complete() once all devices have been created. * Also scheduled when new devices are probed when new CPUs come online. */ -static void __init mpam_enable(struct work_struct *work) +static void mpam_enable(struct work_struct *work) { int err; unsigned long flags; @@ -761,7 +761,7 @@ static struct mpam_class * __init mpam_class_get(u8 level_idx, * class/component structures may be allocated. * Returns the new device, or an ERR_PTR(). */ -struct mpam_device * __init +struct mpam_device * __mpam_device_create(u8 level_idx, enum mpam_class_types type, int component_id, const struct cpumask *fw_affinity, phys_addr_t hwpage_address) @@ -810,7 +810,7 @@ __mpam_device_create(u8 level_idx, enum mpam_class_types type, return dev; } -void __init mpam_device_set_error_irq(struct mpam_device *dev, u32 irq, +void mpam_device_set_error_irq(struct mpam_device *dev, u32 irq, u32 flags) { unsigned long irq_save_flags; @@ -821,7 +821,7 @@ void __init mpam_device_set_error_irq(struct mpam_device *dev, u32 irq, spin_unlock_irqrestore(&dev->lock, irq_save_flags); } -void __init mpam_device_set_overflow_irq(struct mpam_device *dev, u32 irq, +void mpam_device_set_overflow_irq(struct mpam_device *dev, u32 irq, u32 flags) { unsigned long irq_save_flags; @@ -864,7 +864,7 @@ static inline u16 mpam_cpu_max_pmg(void) /* * prepare for initializing devices. */ -int __init mpam_discovery_start(void) +int mpam_discovery_start(void) { if (!mpam_cpus_have_feature()) return -EOPNOTSUPP; @@ -1104,7 +1104,7 @@ static int mpam_cpu_offline(unsigned int cpu) return 0; } -int __init mpam_discovery_complete(void) +int mpam_discovery_complete(void) { int ret = 0; @@ -1121,7 +1121,7 @@ int __init mpam_discovery_complete(void) return ret; } -void __init mpam_discovery_failed(void) +void mpam_discovery_failed(void) { struct mpam_class *class, *tmp; diff --git a/arch/arm64/kernel/mpam/mpam_internal.h b/arch/arm64/kernel/mpam/mpam_internal.h index cfaef82428aa5..7b84ea54975aa 100644 --- a/arch/arm64/kernel/mpam/mpam_internal.h +++ b/arch/arm64/kernel/mpam/mpam_internal.h @@ -329,7 +329,7 @@ int mpam_resctrl_setup(void); struct raw_resctrl_resource * mpam_get_raw_resctrl_resource(u32 level); -int __init mpam_resctrl_init(void); +int mpam_resctrl_init(void); int mpam_resctrl_set_default_cpu(unsigned int cpu); void mpam_resctrl_clear_default_cpu(unsigned int cpu); diff --git a/arch/arm64/kernel/mpam/mpam_resctrl.c b/arch/arm64/kernel/mpam/mpam_resctrl.c index 51cdefebaeba8..1ad4cd49762a3 100644 --- a/arch/arm64/kernel/mpam/mpam_resctrl.c +++ b/arch/arm64/kernel/mpam/mpam_resctrl.c @@ -2209,7 +2209,7 @@ static int __init mpam_setup(char *str) } __setup("mpam", mpam_setup); -int __init mpam_resctrl_init(void) +int mpam_resctrl_init(void) { mpam_init_padding(); diff --git a/fs/resctrlfs.c b/fs/resctrlfs.c index 11741c87eb0af..ea9df7d77b95a 100644 --- a/fs/resctrlfs.c +++ b/fs/resctrlfs.c @@ -1029,7 +1029,7 @@ static int __init resctrl_group_setup_root(void) * * Return: 0 on success or -errno */ -int __init resctrl_group_init(void) +int resctrl_group_init(void) { int ret = 0; diff --git a/include/linux/arm_mpam.h b/include/linux/arm_mpam.h index 44d0690ae8c4b..a242d4fdff8b3 100644 --- a/include/linux/arm_mpam.h +++ b/include/linux/arm_mpam.h @@ -16,7 +16,7 @@ enum mpam_class_types { MPAM_CLASS_UNKNOWN, /* Everything else, e.g. TLBs etc */ }; -struct mpam_device * __init +struct mpam_device * __mpam_device_create(u8 level_idx, enum mpam_class_types type, int component_id, const struct cpumask *fw_affinity, phys_addr_t hwpage_address); @@ -54,9 +54,9 @@ mpam_device_create_memory(int nid, phys_addr_t hwpage_address) return __mpam_device_create(~0, MPAM_CLASS_MEMORY, nid, &dev_affinity, hwpage_address); } -int __init mpam_discovery_start(void); -int __init mpam_discovery_complete(void); -void __init mpam_discovery_failed(void); +int mpam_discovery_start(void); +int mpam_discovery_complete(void); +void mpam_discovery_failed(void); enum mpam_enable_type { MPAM_ENABLE_DENIED = 0, @@ -71,12 +71,12 @@ extern enum mpam_enable_type mpam_enabled; #define mpam_irq_flags_to_acpi(x) ((x & MPAM_IRQ_MODE_LEVEL) ? \ ACPI_LEVEL_SENSITIVE : ACPI_EDGE_SENSITIVE) -void __init mpam_device_set_error_irq(struct mpam_device *dev, u32 irq, +void mpam_device_set_error_irq(struct mpam_device *dev, u32 irq, u32 flags); -void __init mpam_device_set_overflow_irq(struct mpam_device *dev, u32 irq, +void mpam_device_set_overflow_irq(struct mpam_device *dev, u32 irq, u32 flags); -static inline int __init mpam_register_device_irq(struct mpam_device *dev, +static inline int mpam_register_device_irq(struct mpam_device *dev, u32 overflow_interrupt, u32 overflow_flags, u32 error_interrupt, u32 error_flags) { -- 2.25.1

1 2

[openEuler-5.10 01/19] drm/hisilicon: Support i2c driver algorithms for bit-shift adapters
by Zheng Zengkai 23 Dec '21

23 Dec '21

From: Ma Hai <mahai1(a)huawei.com> mainline inclusion from mainline-v5.12-rc2 commit 4eb4d99dfe3018d86f4529112aa7082f43b6996a category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I4M6CD?from=project-issue CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/d… ---------------------------------------------------------------------- Adding driver implementation to support i2c driver algorithms for bit-shift adapters, so hibmc will using the interface provided by drm to read edid. Signed-off-by: Ma Hai <mahai1(a)huawei.com> Signed-off-by: Tian Tao <tiantao6(a)hisilicon.com> Reviewed-by: Thomas Zimmermann <tzimmermann(a)suse.de> Reviewed-by: Li Dongming <lidongming5(a)huawei.com> Link: https://patchwork.freedesktop.org/patch/msgid/1600778670-60370-2-git-send-e… Acked-by: Xie XiuQi <xiexiuqi(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> --- drivers/gpu/drm/hisilicon/hibmc/Makefile | 2 +- .../gpu/drm/hisilicon/hibmc/hibmc_drm_drv.h | 25 ++++- .../gpu/drm/hisilicon/hibmc/hibmc_drm_i2c.c | 99 +++++++++++++++++++ .../gpu/drm/hisilicon/hibmc/hibmc_drm_vdac.c | 2 +- 4 files changed, 125 insertions(+), 3 deletions(-) create mode 100644 drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_i2c.c diff --git a/drivers/gpu/drm/hisilicon/hibmc/Makefile b/drivers/gpu/drm/hisilicon/hibmc/Makefile index f99132715597..684ef794eb7c 100644 --- a/drivers/gpu/drm/hisilicon/hibmc/Makefile +++ b/drivers/gpu/drm/hisilicon/hibmc/Makefile @@ -1,4 +1,4 @@ # SPDX-License-Identifier: GPL-2.0-only -hibmc-drm-y := hibmc_drm_drv.o hibmc_drm_de.o hibmc_drm_vdac.o hibmc_ttm.o +hibmc-drm-y := hibmc_drm_drv.o hibmc_drm_de.o hibmc_drm_vdac.o hibmc_ttm.o hibmc_drm_i2c.o obj-$(CONFIG_DRM_HISI_HIBMC) += hibmc-drm.o diff --git a/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_drv.h b/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_drv.h index 197485e2fe0b..87d2aad0bb5e 100644 --- a/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_drv.h +++ b/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_drv.h @@ -14,11 +14,23 @@ #ifndef HIBMC_DRM_DRV_H #define HIBMC_DRM_DRV_H +#include <linux/gpio/consumer.h> +#include <linux/i2c-algo-bit.h> +#include <linux/i2c.h> + +#include <drm/drm_edid.h> #include <drm/drm_fb_helper.h> #include <drm/drm_framebuffer.h> struct drm_device; +struct hibmc_connector { + struct drm_connector base; + + struct i2c_adapter adapter; + struct i2c_algo_bit_data bit_data; +}; + struct hibmc_drm_private { /* hw */ void __iomem *mmio; @@ -31,10 +43,20 @@ struct hibmc_drm_private { struct drm_plane primary_plane; struct drm_crtc crtc; struct drm_encoder encoder; - struct drm_connector connector; + struct hibmc_connector connector; bool mode_config_initialized; }; +static inline struct hibmc_connector *to_hibmc_connector(struct drm_connector *connector) +{ + return container_of(connector, struct hibmc_connector, base); +} + +static inline struct hibmc_drm_private *to_hibmc_drm_private(struct drm_device *dev) +{ + return dev->dev_private; +} + void hibmc_set_power_mode(struct hibmc_drm_private *priv, unsigned int power_mode); void hibmc_set_current_gate(struct hibmc_drm_private *priv, @@ -47,6 +69,7 @@ int hibmc_mm_init(struct hibmc_drm_private *hibmc); void hibmc_mm_fini(struct hibmc_drm_private *hibmc); int hibmc_dumb_create(struct drm_file *file, struct drm_device *dev, struct drm_mode_create_dumb *args); +int hibmc_ddc_create(struct drm_device *drm_dev, struct hibmc_connector *connector); extern const struct drm_mode_config_funcs hibmc_mode_funcs; diff --git a/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_i2c.c b/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_i2c.c new file mode 100644 index 000000000000..86d712090d87 --- /dev/null +++ b/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_i2c.c @@ -0,0 +1,99 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* Hisilicon Hibmc SoC drm driver + * + * Based on the bochs drm driver. + * + * Copyright (c) 2016 Huawei Limited. + * + * Author: + * Tian Tao <tiantao6(a)hisilicon.com> + */ + +#include <linux/delay.h> +#include <linux/pci.h> + +#include <drm/drm_atomic_helper.h> +#include <drm/drm_probe_helper.h> + +#include "hibmc_drm_drv.h" + +#define GPIO_DATA 0x0802A0 +#define GPIO_DATA_DIRECTION 0x0802A4 + +#define I2C_SCL_MASK BIT(0) +#define I2C_SDA_MASK BIT(1) + +static void hibmc_set_i2c_signal(void *data, u32 mask, int value) +{ + struct hibmc_connector *hibmc_connector = data; + struct hibmc_drm_private *priv = to_hibmc_drm_private(hibmc_connector->base.dev); + u32 tmp_dir = readl(priv->mmio + GPIO_DATA_DIRECTION); + + if (value) { + tmp_dir &= ~mask; + writel(tmp_dir, priv->mmio + GPIO_DATA_DIRECTION); + } else { + u32 tmp_data = readl(priv->mmio + GPIO_DATA); + + tmp_data &= ~mask; + writel(tmp_data, priv->mmio + GPIO_DATA); + + tmp_dir |= mask; + writel(tmp_dir, priv->mmio + GPIO_DATA_DIRECTION); + } +} + +static int hibmc_get_i2c_signal(void *data, u32 mask) +{ + struct hibmc_connector *hibmc_connector = data; + struct hibmc_drm_private *priv = to_hibmc_drm_private(hibmc_connector->base.dev); + u32 tmp_dir = readl(priv->mmio + GPIO_DATA_DIRECTION); + + if ((tmp_dir & mask) != mask) { + tmp_dir &= ~mask; + writel(tmp_dir, priv->mmio + GPIO_DATA_DIRECTION); + } + + return (readl(priv->mmio + GPIO_DATA) & mask) ? 1 : 0; +} + +static void hibmc_ddc_setsda(void *data, int state) +{ + hibmc_set_i2c_signal(data, I2C_SDA_MASK, state); +} + +static void hibmc_ddc_setscl(void *data, int state) +{ + hibmc_set_i2c_signal(data, I2C_SCL_MASK, state); +} + +static int hibmc_ddc_getsda(void *data) +{ + return hibmc_get_i2c_signal(data, I2C_SDA_MASK); +} + +static int hibmc_ddc_getscl(void *data) +{ + return hibmc_get_i2c_signal(data, I2C_SCL_MASK); +} + +int hibmc_ddc_create(struct drm_device *drm_dev, + struct hibmc_connector *connector) +{ + connector->adapter.owner = THIS_MODULE; + connector->adapter.class = I2C_CLASS_DDC; + snprintf(connector->adapter.name, I2C_NAME_SIZE, "HIS i2c bit bus"); + connector->adapter.dev.parent = &drm_dev->pdev->dev; + i2c_set_adapdata(&connector->adapter, connector); + connector->adapter.algo_data = &connector->bit_data; + + connector->bit_data.udelay = 20; + connector->bit_data.timeout = usecs_to_jiffies(2000); + connector->bit_data.data = connector; + connector->bit_data.setsda = hibmc_ddc_setsda; + connector->bit_data.setscl = hibmc_ddc_setscl; + connector->bit_data.getsda = hibmc_ddc_getsda; + connector->bit_data.getscl = hibmc_ddc_getscl; + + return i2c_bit_add_bus(&connector->adapter); +} diff --git a/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_vdac.c b/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_vdac.c index 376a05ddbc2f..c8b14afbcbed 100644 --- a/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_vdac.c +++ b/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_vdac.c @@ -78,7 +78,7 @@ int hibmc_vdac_init(struct hibmc_drm_private *priv) { struct drm_device *dev = priv->dev; struct drm_encoder *encoder = &priv->encoder; - struct drm_connector *connector = &priv->connector; + struct drm_connector *connector = &priv->connector.base; int ret; encoder->possible_crtcs = 0x1; -- 2.20.1

1 18

[PATCH openEuler-1.0-LTS 1/3] arm64/mpam: Fix mpam corrupt when cpu online
by Yang Yingliang 23 Dec '21

23 Dec '21

From: Wang ShaoBo <bobo.shaobowang(a)huawei.com> hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I3YAI3 CVE: NA ------------------------------------------------- The following error occurred occasionally on a machine that supports MPAM: [ 13.321386][ T658] Unable to handle kernel paging request at virtual address ffff80001115816c [ 13.326013][ T684] hid-generic 0003:12D1:0003.0002: input,hidraw1: USB HID v1.10 Mouse [Keyboard/Mouse KVM 1.1.0] on usb-0000:7a:01.0-1.1/input1 [ 13.340558][ T658] Mem abort info: [ 13.340563][ T658] ESR = 0x86000007 [ 13.352567][ T5] hub 6-1:1.0: USB hub found [ 13.364750][ T658] EC = 0x21: IABT (current EL), IL = 32 bits [ 13.369891][ T5] hub 6-1:1.0: 4 ports detected [ 13.373871][ T658] SET = 0, FnV = 0 [ 13.396107][ T658] EA = 0, S1PTW = 0 [ 13.400599][ T658] swapper pgtable: 64k pages, 48-bit VAs, pgdp=0000000029540000 [ 13.408726][ T658] [ffff80001115816c] pgd=0000205fffff0003, p4d=0000205fffff0003, pud=0000205fffff0003, pmd=0000205ffffe0003, pte=0000000000000000 [ 13.423346][ T658] Internal error: Oops: 86000007 [#1] SMP [ 13.429720][ T658] Modules linked in: [ 13.434243][ T658] CPU: 72 PID: 658 Comm: kworker/72:1 Not tainted 5.10.0-4.17.0.28.oe1.aarch64 #1 [ 13.443966][ T658] Hardware name: Huawei TaiShan 200 (Model 2280)/BC82AMDDA, BIOS 1.70 01/07/2021 [ 13.453683][ T658] Workqueue: events mpam_enable [ 13.459206][ T658] pstate: 20c00009 (nzCv daif +PAN +UAO -TCO BTYPE=--) [ 13.466625][ T658] pc : mpam_enable+0x194/0x1d8 [ 13.472019][ T658] lr : mpam_enable+0x194/0x1d8 [ 13.477301][ T658] sp : ffff80004664fd70 [ 13.481937][ T658] x29: ffff80004664fd70 x28: 0000000000000000 [ 13.488578][ T658] x27: ffff00400484a648 x26: ffff800011b71080 [ 13.495306][ T658] x25: 0000000000000000 x24: ffff800011b6cda0 [ 13.502001][ T658] x23: ffff800011646f18 x22: ffff800011b6cd80 [ 13.508684][ T658] x21: ffff800011b6c000 x20: ffff800011646f08 [ 13.515425][ T658] x19: ffff800011646f70 x18: 0000000000000020 [ 13.522075][ T658] x17: 000000001790b332 x16: 0000000000000001 [ 13.528785][ T658] x15: ffffffffffffffff x14: ff00000000000000 [ 13.535464][ T658] x13: ffffffffffffffff x12: 0000000000000006 [ 13.542045][ T658] x11: 00000091cea718e2 x10: 0000000000000b90 [ 13.548735][ T658] x9 : ffff80001009ebac x8 : ffff2040061aabf0 [ 13.555383][ T658] x7 : ffffa05f8dca0000 x6 : 000000000000000f [ 13.561924][ T658] x5 : 0000000000000000 x4 : ffff2040061aa000 [ 13.568613][ T658] x3 : ffff80001164dfa0 x2 : 00000000ffffffff [ 13.575267][ T658] x1 : ffffa05f8dca0000 x0 : 00000000000000c1 [ 13.581813][ T658] Call trace: [ 13.585600][ T658] mpam_enable+0x194/0x1d8 [ 13.590450][ T658] process_one_work+0x1cc/0x390 [ 13.595654][ T658] worker_thread+0x70/0x2f0 [ 13.600499][ T658] kthread+0x118/0x120 [ 13.604935][ T658] ret_from_fork+0x10/0x18 [ 13.609717][ T658] Code: bad PC value [ 13.613944][ T658] ---[ end trace f1e305d2c339f67f ]--- [ 13.753818][ T658] Kernel panic - not syncing: Oops: Fatal exception [ 13.760885][ T658] SMP: stopping secondary CPUs [ 13.765933][ T658] Kernel Offset: disabled [ 13.770516][ T658] CPU features: 0x8040002,22208a38 [ 13.775862][ T658] Memory Limit: none [ 13.913929][ T658] ---[ end Kernel panic - not syncing: The process of MPAM devices initialization is like this: mpam_discovery_start() ... // discover devices mpam_discovery_complete() // hang up the mpam_online/offline_cpu callbacks -=> mpam_cpu_online() // probe all devices -=> mpam_enable() // prepare for resctrl (1) -=> cpuhp_remove_state() // clean resctrl internal structure (2) -=> cpuhp_setup_state() // rehang mpam_online/offline_cpu callbacks -=> mpam_cpu_online() // it does not call mpam_enable again -=> mpam_resctrl_cpu_online() // pull up resctrl Re-hang process of mpam_cpu_online/offline callbacks should not be disturbed by irqs, to ensure that CPU context is reliable before re-entering mpam_cpu_online(), which always happens between (1) and (2). Fixes: 2ab89c893faf ("arm64/mpam: resctrl: Re-synchronise resctrl's view of online CPUs") Signed-off-by: Wang ShaoBo <bobo.shaobowang(a)huawei.com> Reviewed-by: Cheng Jian <cj.chengjian(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- arch/arm64/kernel/mpam/mpam_device.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/arm64/kernel/mpam/mpam_device.c b/arch/arm64/kernel/mpam/mpam_device.c index f8840274b902f..c0615f6947a1f 100644 --- a/arch/arm64/kernel/mpam/mpam_device.c +++ b/arch/arm64/kernel/mpam/mpam_device.c @@ -593,9 +593,11 @@ static void __init mpam_enable(struct work_struct *work) pr_err("Failed to setup/init resctrl\n"); mutex_unlock(&mpam_devices_lock); + local_irq_disable(); mpam_cpuhp_state = cpuhp_setup_state(CPUHP_AP_ONLINE_DYN, "mpam:online", mpam_cpu_online, mpam_cpu_offline); + local_irq_enable(); if (mpam_cpuhp_state <= 0) pr_err("Failed to re-register 'dyn' cpuhp callbacks"); mutex_unlock(&mpam_cpuhp_lock); -- 2.25.1

1 2

[PATCH openEuler-1.0-LTS 1/2] kprobes: Set unoptimized flag after unoptimizing code
by Yang Yingliang 22 Dec '21

22 Dec '21

From: Masami Hiramatsu <mhiramat(a)kernel.org> stable inclusion from stable-v4.19.108 commit 39af044d1ccb207da43af5c060bb0fb0dd548b3e category: bugfix bugzilla: NA CVE: NA ------------------------------------------------- commit f66c0447cca1281116224d474cdb37d6a18e4b5b upstream. Set the unoptimized flag after confirming the code is completely unoptimized. Without this fix, when a kprobe hits the intermediate modified instruction (the first byte is replaced by an INT3, but later bytes can still be a jump address operand) while unoptimizing, it can return to the middle byte of the modified code, which causes an invalid instruction exception in the kernel. Usually, this is a rare case, but if we put a probe on the function call while text patching, it always causes a kernel panic as below: # echo p text_poke+5 > kprobe_events # echo 1 > events/kprobes/enable # echo 0 > events/kprobes/enable invalid opcode: 0000 [#1] PREEMPT SMP PTI RIP: 0010:text_poke+0x9/0x50 Call Trace: arch_unoptimize_kprobe+0x22/0x28 arch_unoptimize_kprobes+0x39/0x87 kprobe_optimizer+0x6e/0x290 process_one_work+0x2a0/0x610 worker_thread+0x28/0x3d0 ? process_one_work+0x610/0x610 kthread+0x10d/0x130 ? kthread_park+0x80/0x80 ret_from_fork+0x3a/0x50 text_poke() is used for patching the code in optprobes. This can happen even if we blacklist text_poke() and other functions, because there is a small time window during which we show the intermediate code to other CPUs. [ mingo: Edited the changelog. ] Tested-by: Alexei Starovoitov <ast(a)kernel.org> Signed-off-by: Masami Hiramatsu <mhiramat(a)kernel.org> Cc: Andy Lutomirski <luto(a)kernel.org> Cc: Borislav Petkov <bp(a)alien8.de> Cc: Linus Torvalds <torvalds(a)linux-foundation.org> Cc: Peter Zijlstra <peterz(a)infradead.org> Cc: Steven Rostedt <rostedt(a)goodmis.org> Cc: Thomas Gleixner <tglx(a)linutronix.de> Cc: bristot(a)redhat.com Fixes: 6274de4984a6 ("kprobes: Support delayed unoptimizing") Link: https://lkml.kernel.org/r/157483422375.25881.13508326028469515760.stgit@dev… Signed-off-by: Ingo Molnar <mingo(a)kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Yang Jihong <yangjihong1(a)huawei.com> Reviewed-by: Kuohai Xu <xukuohai(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- kernel/kprobes.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/kernel/kprobes.c b/kernel/kprobes.c index f6e6edaf964c1..666243fff573d 100644 --- a/kernel/kprobes.c +++ b/kernel/kprobes.c @@ -523,6 +523,8 @@ static void do_unoptimize_kprobes(void) arch_unoptimize_kprobes(&unoptimizing_list, &freeing_list); /* Loop free_list for disarming */ list_for_each_entry_safe(op, tmp, &freeing_list, list) { + /* Switching from detour code to origin */ + op->kp.flags &= ~KPROBE_FLAG_OPTIMIZED; /* Disarm probes if marked disabled */ if (kprobe_disabled(&op->kp)) arch_disarm_kprobe(&op->kp); @@ -662,6 +664,7 @@ static void force_unoptimize_kprobe(struct optimized_kprobe *op) { lockdep_assert_cpus_held(); arch_unoptimize_kprobe(op); + op->kp.flags &= ~KPROBE_FLAG_OPTIMIZED; if (kprobe_disabled(&op->kp)) arch_disarm_kprobe(&op->kp); } @@ -689,7 +692,6 @@ static void unoptimize_kprobe(struct kprobe *p, bool force) return; } - op->kp.flags &= ~KPROBE_FLAG_OPTIMIZED; if (!list_empty(&op->list)) { /* Dequeue from the optimization queue */ list_del_init(&op->list); -- 2.25.1

1 1

[PATCH openEuler-1.0-LTS] cpufreq: schedutil: Destroy mutex before kobject_put() frees the memory
by Yang Yingliang 22 Dec '21

22 Dec '21

From: James Morse <james.morse(a)arm.com> stable inclusion from linux-4.19.209 commit 3cb9595e23d879824ebcefeabdf5156bc288000c -------------------------------- [ Upstream commit cdef1196608892b9a46caa5f2b64095a7f0be60c ] Since commit e5c6b312ce3c ("cpufreq: schedutil: Use kobject release() method to free sugov_tunables") kobject_put() has kfree()d the attr_set before gov_attr_set_put() returns. kobject_put() isn't the last user of attr_set in gov_attr_set_put(), the subsequent mutex_destroy() triggers a use-after-free: | BUG: KASAN: use-after-free in mutex_is_locked+0x20/0x60 | Read of size 8 at addr ffff000800ca4250 by task cpuhp/2/20 | | CPU: 2 PID: 20 Comm: cpuhp/2 Not tainted 5.15.0-rc1 #12369 | Hardware name: ARM LTD ARM Juno Development Platform/ARM Juno Development | Platform, BIOS EDK II Jul 30 2018 | Call trace: | dump_backtrace+0x0/0x380 | show_stack+0x1c/0x30 | dump_stack_lvl+0x8c/0xb8 | print_address_description.constprop.0+0x74/0x2b8 | kasan_report+0x1f4/0x210 | kasan_check_range+0xfc/0x1a4 | __kasan_check_read+0x38/0x60 | mutex_is_locked+0x20/0x60 | mutex_destroy+0x80/0x100 | gov_attr_set_put+0xfc/0x150 | sugov_exit+0x78/0x190 | cpufreq_offline.isra.0+0x2c0/0x660 | cpuhp_cpufreq_offline+0x14/0x24 | cpuhp_invoke_callback+0x430/0x6d0 | cpuhp_thread_fun+0x1b0/0x624 | smpboot_thread_fn+0x5e0/0xa6c | kthread+0x3a0/0x450 | ret_from_fork+0x10/0x20 Swap the order of the calls. Fixes: e5c6b312ce3c ("cpufreq: schedutil: Use kobject release() method to free sugov_tunables") Cc: 4.7+ <stable(a)vger.kernel.org> # 4.7+ Signed-off-by: James Morse <james.morse(a)arm.com> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki(a)intel.com> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/cpufreq/cpufreq_governor_attr_set.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/cpufreq/cpufreq_governor_attr_set.c b/drivers/cpufreq/cpufreq_governor_attr_set.c index 52841f807a7eb..45fdf30cade39 100644 --- a/drivers/cpufreq/cpufreq_governor_attr_set.c +++ b/drivers/cpufreq/cpufreq_governor_attr_set.c @@ -77,8 +77,8 @@ unsigned int gov_attr_set_put(struct gov_attr_set *attr_set, struct list_head *l if (count) return count; - kobject_put(&attr_set->kobj); mutex_destroy(&attr_set->update_lock); + kobject_put(&attr_set->kobj); return 0; } EXPORT_SYMBOL_GPL(gov_attr_set_put); -- 2.25.1

1 0

[PATCH openEuler-1.0-LTS v3] scsi:spraid: support Ramaxel's spraid driver
by Yanling Song 22 Dec '21

22 Dec '21

Ramaxel inclusion category: features bugzilla: https://gitee.com/openeuler/kernel/issues/I4NJIB CVE: NA v2->v3: 1. Add deconfig for hulk in arm64 and x86 v1->v2: 1. Add deconfig for arm64 and x86 Support Ramaxel's SPRxxx Raid controller Signed-off-by: Yanling Song <songyl(a)ramaxel.com> Reviewed-by: Jiang Yu<yujiang(a)ramaxel.com> --- arch/arm64/configs/hulk_defconfig | 1 + arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/hulk_defconfig | 3 +- arch/x86/configs/openeuler_defconfig | 1 + drivers/scsi/Kconfig | 1 + drivers/scsi/Makefile | 1 + drivers/scsi/spraid/Kconfig | 13 + drivers/scsi/spraid/Makefile | 7 + drivers/scsi/spraid/spraid.h | 746 +++++ drivers/scsi/spraid/spraid_main.c | 3875 ++++++++++++++++++++++++ 10 files changed, 4648 insertions(+), 1 deletion(-) create mode 100644 drivers/scsi/spraid/Kconfig create mode 100644 drivers/scsi/spraid/Makefile create mode 100644 drivers/scsi/spraid/spraid.h create mode 100644 drivers/scsi/spraid/spraid_main.c diff --git a/arch/arm64/configs/hulk_defconfig b/arch/arm64/configs/hulk_defconfig index bc350c41b333..71517c091c43 100644 --- a/arch/arm64/configs/hulk_defconfig +++ b/arch/arm64/configs/hulk_defconfig @@ -2149,6 +2149,7 @@ CONFIG_SCSI_MPT2SAS_MAX_SGE=128 CONFIG_SCSI_MPT3SAS_MAX_SGE=128 CONFIG_SCSI_MPT2SAS=m CONFIG_SCSI_SMARTPQI=m +CONFIG_RAMAXEL_SPRAID=m # CONFIG_SCSI_UFSHCD is not set # CONFIG_SCSI_HPTIOP is not set CONFIG_LIBFC=m diff --git a/arch/arm64/configs/openeuler_defconfig b/arch/arm64/configs/openeuler_defconfig index d2167c80757f..47b0931b91bf 100644 --- a/arch/arm64/configs/openeuler_defconfig +++ b/arch/arm64/configs/openeuler_defconfig @@ -2143,6 +2143,7 @@ CONFIG_SCSI_MPT2SAS_MAX_SGE=128 CONFIG_SCSI_MPT3SAS_MAX_SGE=128 CONFIG_SCSI_MPT2SAS=m CONFIG_SCSI_SMARTPQI=m +CONFIG_RAMAXEL_SPRAID=m # CONFIG_SCSI_UFSHCD is not set # CONFIG_SCSI_HPTIOP is not set CONFIG_LIBFC=m diff --git a/arch/x86/configs/hulk_defconfig b/arch/x86/configs/hulk_defconfig index 983d52ccad82..536ffb38c00c 100644 --- a/arch/x86/configs/hulk_defconfig +++ b/arch/x86/configs/hulk_defconfig @@ -1,4 +1,4 @@ -# + # Automatically generated file; DO NOT EDIT. # Linux/x86 4.19.21 Kernel Configuration # @@ -2177,6 +2177,7 @@ CONFIG_SCSI_MPT2SAS_MAX_SGE=128 CONFIG_SCSI_MPT3SAS_MAX_SGE=128 CONFIG_SCSI_MPT2SAS=m CONFIG_SCSI_SMARTPQI=m +CONFIG_RAMAXEL_SPRAID=m # CONFIG_SCSI_UFSHCD is not set # CONFIG_SCSI_HPTIOP is not set # CONFIG_SCSI_BUSLOGIC is not set diff --git a/arch/x86/configs/openeuler_defconfig b/arch/x86/configs/openeuler_defconfig index 7043426d9780..786186288fea 100644 --- a/arch/x86/configs/openeuler_defconfig +++ b/arch/x86/configs/openeuler_defconfig @@ -2182,6 +2182,7 @@ CONFIG_SCSI_MPT2SAS_MAX_SGE=128 CONFIG_SCSI_MPT3SAS_MAX_SGE=128 CONFIG_SCSI_MPT2SAS=m CONFIG_SCSI_SMARTPQI=m +CONFIG_RAMAXEL_SPRAID=m # CONFIG_SCSI_UFSHCD is not set # CONFIG_SCSI_HPTIOP is not set # CONFIG_SCSI_BUSLOGIC is not set diff --git a/drivers/scsi/Kconfig b/drivers/scsi/Kconfig index 8450484184e3..63d2aaa22834 100644 --- a/drivers/scsi/Kconfig +++ b/drivers/scsi/Kconfig @@ -522,6 +522,7 @@ source "drivers/scsi/megaraid/Kconfig.megaraid" source "drivers/scsi/mpt3sas/Kconfig" source "drivers/scsi/smartpqi/Kconfig" source "drivers/scsi/ufs/Kconfig" +source "drivers/scsi/spraid/Kconfig" config SCSI_HPTIOP tristate "HighPoint RocketRAID 3xxx/4xxx Controller support" diff --git a/drivers/scsi/Makefile b/drivers/scsi/Makefile index 2973693f6dcc..4056cf26e09e 100644 --- a/drivers/scsi/Makefile +++ b/drivers/scsi/Makefile @@ -94,6 +94,7 @@ obj-$(CONFIG_SCSI_ZALON) += zalon7xx.o obj-$(CONFIG_SCSI_DC395x) += dc395x.o obj-$(CONFIG_SCSI_AM53C974) += esp_scsi.o am53c974.o obj-$(CONFIG_CXLFLASH) += cxlflash/ +obj-$(CONFIG_RAMAXEL_SPRAID) += spraid/ obj-$(CONFIG_MEGARAID_LEGACY) += megaraid.o obj-$(CONFIG_MEGARAID_NEWGEN) += megaraid/ obj-$(CONFIG_MEGARAID_SAS) += megaraid/ diff --git a/drivers/scsi/spraid/Kconfig b/drivers/scsi/spraid/Kconfig new file mode 100644 index 000000000000..bfbba3db8db0 --- /dev/null +++ b/drivers/scsi/spraid/Kconfig @@ -0,0 +1,13 @@ +# +# Ramaxel driver configuration +# + +config RAMAXEL_SPRAID + tristate "Ramaxel spraid Adapter" + depends on PCI && SCSI + select BLK_DEV_BSGLIB + depends on ARM64 || X86_64 + help + This driver supports Ramaxel SPRxxx serial + raid controller, which has PCIE Gen4 interface + with host and supports SAS/SATA Hdd/ssd. diff --git a/drivers/scsi/spraid/Makefile b/drivers/scsi/spraid/Makefile new file mode 100644 index 000000000000..aadc2ffd37eb --- /dev/null +++ b/drivers/scsi/spraid/Makefile @@ -0,0 +1,7 @@ +# +# Makefile for the Ramaxel device drivers. +# + +obj-$(CONFIG_RAMAXEL_SPRAID) += spraid.o + +spraid-objs := spraid_main.o \ No newline at end of file diff --git a/drivers/scsi/spraid/spraid.h b/drivers/scsi/spraid/spraid.h new file mode 100644 index 000000000000..c1e4980e18e5 --- /dev/null +++ b/drivers/scsi/spraid/spraid.h @@ -0,0 +1,746 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* Copyright(c) 2021 Ramaxel Memory Technology, Ltd */ + +#ifndef __SPRAID_H_ +#define __SPRAID_H_ + +#define SPRAID_CAP_MQES(cap) ((cap) & 0xffff) +#define SPRAID_CAP_STRIDE(cap) (((cap) >> 32) & 0xf) +#define SPRAID_CAP_MPSMIN(cap) (((cap) >> 48) & 0xf) +#define SPRAID_CAP_MPSMAX(cap) (((cap) >> 52) & 0xf) +#define SPRAID_CAP_TIMEOUT(cap) (((cap) >> 24) & 0xff) +#define SPRAID_CAP_DMAMASK(cap) (((cap) >> 37) & 0xff) + +#define SPRAID_DEFAULT_MAX_CHANNEL 4 +#define SPRAID_DEFAULT_MAX_ID 240 +#define SPRAID_DEFAULT_MAX_LUN_PER_HOST 8 +#define MAX_SECTORS 2048 + +#define IO_SQE_SIZE sizeof(struct spraid_ioq_command) +#define ADMIN_SQE_SIZE sizeof(struct spraid_admin_command) +#define SQE_SIZE(qid) (((qid) > 0) ? IO_SQE_SIZE : ADMIN_SQE_SIZE) +#define CQ_SIZE(depth) ((depth) * sizeof(struct spraid_completion)) +#define SQ_SIZE(qid, depth) ((depth) * SQE_SIZE(qid)) + +#define SENSE_SIZE(depth) ((depth) * SCSI_SENSE_BUFFERSIZE) + +#define SPRAID_AQ_DEPTH 128 +#define SPRAID_NR_AEN_COMMANDS 16 +#define SPRAID_AQ_BLK_MQ_DEPTH (SPRAID_AQ_DEPTH - SPRAID_NR_AEN_COMMANDS) +#define SPRAID_AQ_MQ_TAG_DEPTH (SPRAID_AQ_BLK_MQ_DEPTH - 1) + +#define SPRAID_ADMIN_QUEUE_NUM 1 +#define SPRAID_PTCMDS_PERQ 1 +#define SPRAID_IO_BLK_MQ_DEPTH (hdev->shost->can_queue) +#define SPRAID_NR_IOQ_PTCMDS (SPRAID_PTCMDS_PERQ * hdev->shost->nr_hw_queues) + +#define FUA_MASK 0x08 +#define SPRAID_MINORS BIT(MINORBITS) + +#define COMMAND_IS_WRITE(cmd) ((cmd)->common.opcode & 1) + +#define SPRAID_IO_IOSQES 7 +#define SPRAID_IO_IOCQES 4 +#define PRP_ENTRY_SIZE 8 + +#define SMALL_POOL_SIZE 256 +#define MAX_SMALL_POOL_NUM 16 +#define MAX_CMD_PER_DEV 64 +#define MAX_CDB_LEN 32 + +#define SPRAID_UP_TO_MULTY4(x) (((x) + 4) & (~0x03)) + +#define CQE_STATUS_SUCCESS (0x0) + +#define PCI_VENDOR_ID_RAMAXEL_LOGIC 0x1E81 + +#define SPRAID_SERVER_DEVICE_HBA_DID 0x2100 +#define SPRAID_SERVER_DEVICE_RAID_DID 0x2200 + +#define IO_6_DEFAULT_TX_LEN 256 + +#define SPRAID_INT_PAGES 2 +#define SPRAID_INT_BYTES(hdev) (SPRAID_INT_PAGES * (hdev)->page_size) + +enum { + SPRAID_REQ_CANCELLED = (1 << 0), + SPRAID_REQ_USERCMD = (1 << 1), +}; + +enum { + SPRAID_SC_SUCCESS = 0x0, + SPRAID_SC_INVALID_OPCODE = 0x1, + SPRAID_SC_INVALID_FIELD = 0x2, + + SPRAID_SC_ABORT_LIMIT = 0x103, + SPRAID_SC_ABORT_MISSING = 0x104, + SPRAID_SC_ASYNC_LIMIT = 0x105, + + SPRAID_SC_DNR = 0x4000, +}; + +enum { + SPRAID_REG_CAP = 0x0000, + SPRAID_REG_CC = 0x0014, + SPRAID_REG_CSTS = 0x001c, + SPRAID_REG_AQA = 0x0024, + SPRAID_REG_ASQ = 0x0028, + SPRAID_REG_ACQ = 0x0030, + SPRAID_REG_DBS = 0x1000, +}; + +enum { + SPRAID_CC_ENABLE = 1 << 0, + SPRAID_CC_CSS_NVM = 0 << 4, + SPRAID_CC_MPS_SHIFT = 7, + SPRAID_CC_AMS_SHIFT = 11, + SPRAID_CC_SHN_SHIFT = 14, + SPRAID_CC_IOSQES_SHIFT = 16, + SPRAID_CC_IOCQES_SHIFT = 20, + SPRAID_CC_AMS_RR = 0 << SPRAID_CC_AMS_SHIFT, + SPRAID_CC_SHN_NONE = 0 << SPRAID_CC_SHN_SHIFT, + SPRAID_CC_IOSQES = SPRAID_IO_IOSQES << SPRAID_CC_IOSQES_SHIFT, + SPRAID_CC_IOCQES = SPRAID_IO_IOCQES << SPRAID_CC_IOCQES_SHIFT, + SPRAID_CC_SHN_NORMAL = 1 << SPRAID_CC_SHN_SHIFT, + SPRAID_CC_SHN_MASK = 3 << SPRAID_CC_SHN_SHIFT, + SPRAID_CSTS_CFS_SHIFT = 1, + SPRAID_CSTS_SHST_SHIFT = 2, + SPRAID_CSTS_PP_SHIFT = 5, + SPRAID_CSTS_RDY = 1 << 0, + SPRAID_CSTS_SHST_CMPLT = 2 << 2, + SPRAID_CSTS_SHST_MASK = 3 << 2, + SPRAID_CSTS_CFS_MASK = 1 << SPRAID_CSTS_CFS_SHIFT, + SPRAID_CSTS_PP_MASK = 1 << SPRAID_CSTS_PP_SHIFT, +}; + +enum { + SPRAID_ADMIN_DELETE_SQ = 0x00, + SPRAID_ADMIN_CREATE_SQ = 0x01, + SPRAID_ADMIN_DELETE_CQ = 0x04, + SPRAID_ADMIN_CREATE_CQ = 0x05, + SPRAID_ADMIN_ABORT_CMD = 0x08, + SPRAID_ADMIN_SET_FEATURES = 0x09, + SPRAID_ADMIN_ASYNC_EVENT = 0x0c, + SPRAID_ADMIN_GET_INFO = 0xc6, + SPRAID_ADMIN_RESET = 0xc8, +}; + +enum { + SPRAID_GET_INFO_CTRL = 0, + SPRAID_GET_INFO_DEV_LIST = 1, +}; + +enum { + SPRAID_RESET_TARGET = 0, + SPRAID_RESET_BUS = 1, +}; + +enum { + SPRAID_AEN_ERROR = 0, + SPRAID_AEN_NOTICE = 2, + SPRAID_AEN_VS = 7, +}; + +enum { + SPRAID_AEN_DEV_CHANGED = 0x00, + SPRAID_AEN_FW_ACT_START = 0x01, + SPRAID_AEN_HOST_PROBING = 0x10, +}; + +enum { + SPRAID_AEN_TIMESYN = 0x00, + SPRAID_AEN_FW_ACT_FINISH = 0x02, + SPRAID_AEN_EVENT_MIN = 0x80, + SPRAID_AEN_EVENT_MAX = 0xff, +}; + +enum { + SPRAID_CMD_WRITE = 0x01, + SPRAID_CMD_READ = 0x02, + + SPRAID_CMD_NONIO_NONE = 0x80, + SPRAID_CMD_NONIO_TODEV = 0x81, + SPRAID_CMD_NONIO_FROMDEV = 0x82, +}; + +enum { + SPRAID_QUEUE_PHYS_CONTIG = (1 << 0), + SPRAID_CQ_IRQ_ENABLED = (1 << 1), + + SPRAID_FEAT_NUM_QUEUES = 0x07, + SPRAID_FEAT_ASYNC_EVENT = 0x0b, + SPRAID_FEAT_TIMESTAMP = 0x0e, +}; + +enum spraid_state { + SPRAID_NEW, + SPRAID_LIVE, + SPRAID_RESETTING, + SPRAID_DELETING, + SPRAID_DEAD, +}; + +enum { + SPRAID_CARD_HBA, + SPRAID_CARD_RAID, +}; + +enum spraid_cmd_type { + SPRAID_CMD_ADM, + SPRAID_CMD_IOPT, +}; + +struct spraid_completion { + __le32 result; + union { + struct { + __u8 sense_len; + __u8 resv[3]; + }; + __le32 result1; + }; + __le16 sq_head; + __le16 sq_id; + __u16 cmd_id; + __le16 status; +}; + +struct spraid_ctrl_info { + __le32 nd; + __le16 max_cmds; + __le16 max_channel; + __le32 max_tgt_id; + __le16 max_lun; + __le16 max_num_sge; + __le16 lun_num_in_boot; + __u8 mdts; + __u8 acl; + __u8 aerl; + __u8 card_type; + __u16 rsvd; + __u32 rtd3e; + __u8 sn[32]; + __u8 fr[16]; + __u8 rsvd1[4020]; +}; + +struct spraid_dev { + struct pci_dev *pdev; + struct device *dev; + struct Scsi_Host *shost; + struct spraid_queue *queues; + struct dma_pool *prp_page_pool; + struct dma_pool *prp_small_pool[MAX_SMALL_POOL_NUM]; + mempool_t *iod_mempool; + void __iomem *bar; + u32 max_qid; + u32 num_vecs; + u32 queue_count; + u32 ioq_depth; + int db_stride; + u32 __iomem *dbs; + struct rw_semaphore devices_rwsem; + int numa_node; + u32 page_size; + u32 ctrl_config; + u32 online_queues; + u64 cap; + int instance; + struct spraid_ctrl_info *ctrl_info; + struct spraid_dev_info *devices; + + struct spraid_cmd *adm_cmds; + struct list_head adm_cmd_list; + spinlock_t adm_cmd_lock; + + struct spraid_cmd *ioq_ptcmds; + struct list_head ioq_pt_list; + spinlock_t ioq_pt_lock; + + struct work_struct scan_work; + struct work_struct timesyn_work; + struct work_struct reset_work; + struct work_struct fw_act_work; + + enum spraid_state state; + spinlock_t state_lock; + + struct request_queue *bsg_queue; +}; + +struct spraid_sgl_desc { + __le64 addr; + __le32 length; + __u8 rsvd[3]; + __u8 type; +}; + +union spraid_data_ptr { + struct { + __le64 prp1; + __le64 prp2; + }; + struct spraid_sgl_desc sgl; +}; + +struct spraid_admin_common_command { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __le32 cdw2[4]; + union spraid_data_ptr dptr; + __le32 cdw10; + __le32 cdw11; + __le32 cdw12; + __le32 cdw13; + __le32 cdw14; + __le32 cdw15; +}; + +struct spraid_features { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __u64 rsvd2[2]; + union spraid_data_ptr dptr; + __le32 fid; + __le32 dword11; + __le32 dword12; + __le32 dword13; + __le32 dword14; + __le32 dword15; +}; + +struct spraid_create_cq { + __u8 opcode; + __u8 flags; + __u16 command_id; + __u32 rsvd1[5]; + __le64 prp1; + __u64 rsvd8; + __le16 cqid; + __le16 qsize; + __le16 cq_flags; + __le16 irq_vector; + __u32 rsvd12[4]; +}; + +struct spraid_create_sq { + __u8 opcode; + __u8 flags; + __u16 command_id; + __u32 rsvd1[5]; + __le64 prp1; + __u64 rsvd8; + __le16 sqid; + __le16 qsize; + __le16 sq_flags; + __le16 cqid; + __u32 rsvd12[4]; +}; + +struct spraid_delete_queue { + __u8 opcode; + __u8 flags; + __u16 command_id; + __u32 rsvd1[9]; + __le16 qid; + __u16 rsvd10; + __u32 rsvd11[5]; +}; + +struct spraid_get_info { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __u32 rsvd2[4]; + union spraid_data_ptr dptr; + __u8 type; + __u8 rsvd10[3]; + __le32 cdw11; + __u32 rsvd12[4]; +}; + +struct spraid_usr_cmd { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + union { + struct { + __le16 subopcode; + __le16 rsvd1; + } info_0; + __le32 cdw2; + }; + union { + struct { + __le16 data_len; + __le16 param_len; + } info_1; + __le32 cdw3; + }; + __u64 metadata; + union spraid_data_ptr dptr; + __le32 cdw10; + __le32 cdw11; + __le32 cdw12; + __le32 cdw13; + __le32 cdw14; + __le32 cdw15; +}; + +enum { + SPRAID_CMD_FLAG_SGL_METABUF = (1 << 6), + SPRAID_CMD_FLAG_SGL_METASEG = (1 << 7), + SPRAID_CMD_FLAG_SGL_ALL = SPRAID_CMD_FLAG_SGL_METABUF | + SPRAID_CMD_FLAG_SGL_METASEG, +}; + +enum spraid_cmd_state { + SPRAID_CMD_IDLE = 0, + SPRAID_CMD_IN_FLIGHT = 1, + SPRAID_CMD_COMPLETE = 2, + SPRAID_CMD_TIMEOUT = 3, + SPRAID_CMD_TMO_COMPLETE = 4, +}; + +struct spraid_abort_cmd { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __u64 rsvd2[4]; + __le16 sqid; + __le16 cid; + __u32 rsvd11[5]; +}; + +struct spraid_reset_cmd { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __u64 rsvd2[4]; + __u8 type; + __u8 rsvd10[3]; + __u32 rsvd11[5]; +}; + +struct spraid_admin_command { + union { + struct spraid_admin_common_command common; + struct spraid_features features; + struct spraid_create_cq create_cq; + struct spraid_create_sq create_sq; + struct spraid_delete_queue delete_queue; + struct spraid_get_info get_info; + struct spraid_abort_cmd abort; + struct spraid_reset_cmd reset; + struct spraid_usr_cmd usr_cmd; + }; +}; + +struct spraid_ioq_common_command { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __le16 sense_len; + __u8 cdb_len; + __u8 rsvd2; + __le32 cdw3[3]; + union spraid_data_ptr dptr; + __le32 cdw10[6]; + __u8 cdb[32]; + __le64 sense_addr; + __le32 cdw26[6]; +}; + +struct spraid_rw_command { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __le16 sense_len; + __u8 cdb_len; + __u8 rsvd2; + __u32 rsvd3[3]; + union spraid_data_ptr dptr; + __le64 slba; + __le16 nlb; + __le16 control; + __u32 rsvd13[3]; + __u8 cdb[32]; + __le64 sense_addr; + __u32 rsvd26[6]; +}; + +struct spraid_scsi_nonio { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __le16 sense_len; + __u8 cdb_length; + __u8 rsvd2; + __u32 rsvd3[3]; + union spraid_data_ptr dptr; + __u32 rsvd10[5]; + __le32 buffer_len; + __u8 cdb[32]; + __le64 sense_addr; + __u32 rsvd26[6]; +}; + +struct spraid_ioq_command { + union { + struct spraid_ioq_common_command common; + struct spraid_rw_command rw; + struct spraid_scsi_nonio scsi_nonio; + }; +}; + +struct spraid_passthru_common_cmd { + __u8 opcode; + __u8 flags; + __u16 rsvd0; + __u32 nsid; + union { + struct { + __u16 subopcode; + __u16 rsvd1; + } info_0; + __u32 cdw2; + }; + union { + struct { + __u16 data_len; + __u16 param_len; + } info_1; + __u32 cdw3; + }; + __u64 metadata; + + __u64 addr; + __u64 prp2; + + __u32 cdw10; + __u32 cdw11; + __u32 cdw12; + __u32 cdw13; + __u32 cdw14; + __u32 cdw15; + __u32 timeout_ms; + __u32 result0; + __u32 result1; +}; + +struct spraid_ioq_passthru_cmd { + __u8 opcode; + __u8 flags; + __u16 rsvd0; + __u32 nsid; + union { + struct { + __u16 res_sense_len; + __u8 cdb_len; + __u8 rsvd0; + } info_0; + __u32 cdw2; + }; + union { + struct { + __u16 subopcode; + __u16 rsvd1; + } info_1; + __u32 cdw3; + }; + union { + struct { + __u16 rsvd; + __u16 param_len; + } info_2; + __u32 cdw4; + }; + __u32 cdw5; + __u64 addr; + __u64 prp2; + union { + struct { + __u16 eid; + __u16 sid; + } info_3; + __u32 cdw10; + }; + union { + struct { + __u16 did; + __u8 did_flag; + __u8 rsvd2; + } info_4; + __u32 cdw11; + }; + __u32 cdw12; + __u32 cdw13; + __u32 cdw14; + __u32 data_len; + __u32 cdw16; + __u32 cdw17; + __u32 cdw18; + __u32 cdw19; + __u32 cdw20; + __u32 cdw21; + __u32 cdw22; + __u32 cdw23; + __u64 sense_addr; + __u32 cdw26[4]; + __u32 timeout_ms; + __u32 result0; + __u32 result1; +}; + +struct spraid_bsg_request { + u32 msgcode; + u32 control; + union { + struct spraid_passthru_common_cmd admcmd; + struct spraid_ioq_passthru_cmd ioqcmd; + }; +}; + +enum { + SPRAID_BSG_ADM, + SPRAID_BSG_IOQ, +}; + +struct spraid_cmd { + int qid; + int cid; + u32 result0; + u32 result1; + u16 status; + void *priv; + enum spraid_cmd_state state; + struct completion cmd_done; + struct list_head list; +}; + +struct spraid_queue { + struct spraid_dev *hdev; + spinlock_t sq_lock; + + spinlock_t cq_lock ____cacheline_aligned_in_smp; + + void *sq_cmds; + + struct spraid_completion *cqes; + + dma_addr_t sq_dma_addr; + dma_addr_t cq_dma_addr; + u32 __iomem *q_db; + u8 cq_phase; + u8 sqes; + u16 qid; + u16 sq_tail; + u16 cq_head; + u16 last_cq_head; + u16 q_depth; + s16 cq_vector; + void *sense; + dma_addr_t sense_dma_addr; + struct dma_pool *prp_small_pool; +}; + +struct spraid_iod { + struct spraid_queue *spraidq; + enum spraid_cmd_state state; + int npages; + u32 nsge; + u32 length; + bool use_sgl; + bool sg_drv_mgmt; + dma_addr_t first_dma; + void *sense; + dma_addr_t sense_dma; + struct scatterlist *sg; + struct scatterlist inline_sg[0]; +}; + +#define SPRAID_DEV_INFO_ATTR_BOOT(attr) ((attr) & 0x01) +#define SPRAID_DEV_INFO_ATTR_VD(attr) (((attr) & 0x02) == 0x0) +#define SPRAID_DEV_INFO_ATTR_PT(attr) (((attr) & 0x22) == 0x02) +#define SPRAID_DEV_INFO_ATTR_RAWDISK(attr) ((attr) & 0x20) + +#define SPRAID_DEV_INFO_FLAG_VALID(flag) ((flag) & 0x01) +#define SPRAID_DEV_INFO_FLAG_CHANGE(flag) ((flag) & 0x02) + +#define BGTASK_TYPE_REBUILD 4 +#define USR_CMD_READ 0xc2 +#define USR_CMD_RDLEN 0x1000 +#define USR_CMD_VDINFO 0x704 +#define USR_CMD_BGTASK 0x504 +#define VDINFO_PARAM_LEN 0x04 + +struct spraid_vd_info { + __u8 name[32]; + __le16 id; + __u8 rg_id; + __u8 rg_level; + __u8 sg_num; + __u8 sg_disk_num; + __u8 vd_status; + __u8 vd_type; + __u8 rsvd1[4056]; +}; + +#define MAX_REALTIME_BGTASK_NUM 32 + +struct bgtask_info { + __u8 type; + __u8 progress; + __u8 rate; + __u8 rsvd0; + __le16 vd_id; + __le16 time_left; + __u8 rsvd1[4]; +}; + +struct spraid_bgtask { + __u8 sw; + __u8 task_num; + __u8 rsvd[6]; + struct bgtask_info bgtask[MAX_REALTIME_BGTASK_NUM]; +}; + +struct spraid_dev_info { + __le32 hdid; + __le16 target; + __u8 channel; + __u8 lun; + __u8 attr; + __u8 flag; + __le16 max_io_kb; +}; + +#define MAX_DEV_ENTRY_PER_PAGE_4K 340 +struct spraid_dev_list { + __le32 dev_num; + __u32 rsvd0[3]; + struct spraid_dev_info devices[MAX_DEV_ENTRY_PER_PAGE_4K]; +}; + +struct spraid_sdev_hostdata { + u32 hdid; + u16 max_io_kb; + u8 attr; + u8 flag; + u8 rg_id; + u8 rsvd[3]; +}; + +#endif + diff --git a/drivers/scsi/spraid/spraid_main.c b/drivers/scsi/spraid/spraid_main.c new file mode 100644 index 000000000000..519b39f44e91 --- /dev/null +++ b/drivers/scsi/spraid/spraid_main.c @@ -0,0 +1,3875 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright(c) 2021 Ramaxel Memory Technology, Ltd */ + +/* Ramaxel Raid SPXXX Series Linux Driver */ + +#define pr_fmt(fmt) "spraid: " fmt + +#include <linux/sched/signal.h> +#include <linux/version.h> +#include <linux/pci.h> +#include <linux/aer.h> +#include <linux/module.h> +#include <linux/ioport.h> +#include <linux/device.h> +#include <linux/delay.h> +#include <linux/interrupt.h> +#include <linux/cdev.h> +#include <linux/sysfs.h> +#include <linux/gfp.h> +#include <linux/types.h> +#include <linux/ratelimit.h> +#include <linux/once.h> +#include <linux/debugfs.h> +#include <linux/io-64-nonatomic-lo-hi.h> +#include <linux/blkdev.h> +#include <linux/bsg-lib.h> +#include <asm/unaligned.h> +#include <linux/sort.h> +#include <target/target_core_backend.h> + +#include <scsi/scsi.h> +#include <scsi/scsi_cmnd.h> +#include <scsi/scsi_device.h> +#include <scsi/scsi_host.h> +#include <scsi/scsi_transport.h> +#include <scsi/scsi_dbg.h> + + +#include "spraid.h" + +static u32 admin_tmout = 60; +module_param(admin_tmout, uint, 0644); +MODULE_PARM_DESC(admin_tmout, "admin commands timeout (seconds)"); + +static u32 scmd_tmout_nonpt = 180; +module_param(scmd_tmout_nonpt, uint, 0644); +MODULE_PARM_DESC(scmd_tmout_nonpt, + "scsi commands timeout for rawdisk&raid(seconds)"); + +static int ioq_depth_set(const char *val, const struct kernel_param *kp); +static const struct kernel_param_ops ioq_depth_ops = { + .set = ioq_depth_set, + .get = param_get_uint, +}; + +static u32 io_queue_depth = 1024; +module_param_cb(io_queue_depth, &ioq_depth_ops, &io_queue_depth, 0644); +MODULE_PARM_DESC(io_queue_depth, "set io queue depth, should >= 2"); + +static int log_debug_switch_set(const char *val, const struct kernel_param *kp) +{ + u8 n = 0; + int ret; + + ret = kstrtou8(val, 10, &n); + if (ret != 0) + return -EINVAL; + + return param_set_byte(val, kp); +} + +static const struct kernel_param_ops log_debug_switch_ops = { + .set = log_debug_switch_set, + .get = param_get_byte, +}; + +static unsigned char log_debug_switch; +module_param_cb(log_debug_switch, &log_debug_switch_ops, + &log_debug_switch, 0644); +MODULE_PARM_DESC(log_debug_switch, + "set log state, default non-zero for switch on"); + +static int small_pool_num_set(const char *val, const struct kernel_param *kp) +{ + u8 n = 0; + int ret; + + ret = kstrtou8(val, 10, &n); + if (ret != 0) + return -EINVAL; + if (n > MAX_SMALL_POOL_NUM) + n = MAX_SMALL_POOL_NUM; + if (n < 1) + n = 1; + *((u8 *)kp->arg) = n; + + return 0; +} + +static const struct kernel_param_ops small_pool_num_ops = { + .set = small_pool_num_set, + .get = param_get_byte, +}; + +/* It was found that the spindlock of a single pool conflicts + * a lot with multiple CPUs.So multiple pools are introduced + * to reduce the conflictions. + */ +static unsigned char small_pool_num = 4; +module_param_cb(small_pool_num, &small_pool_num_ops, &small_pool_num, 0644); +MODULE_PARM_DESC(small_pool_num, "set prp small pool num, default 4, MAX 16"); + +static void spraid_free_queue(struct spraid_queue *spraidq); +static void spraid_handle_aen_notice(struct spraid_dev *hdev, u32 result); +static void spraid_handle_aen_vs(struct spraid_dev *hdev, + u32 result, u32 result1); + +static DEFINE_IDA(spraid_instance_ida); + +static struct class *spraid_class; + +#define SPRAID_CAP_TIMEOUT_UNIT_MS (HZ / 2) + +static struct workqueue_struct *spraid_wq; + +#define dev_log_dbg(dev, fmt, ...) do { \ + if (unlikely(log_debug_switch)) \ + dev_info(dev, "[%s] [%d] " fmt, \ + __func__, __LINE__, ##__VA_ARGS__); \ +} while (0) + +#define SPRAID_DRV_VERSION "1.0.0.0" + +#define ADMIN_TIMEOUT (admin_tmout * HZ) +#define ADMIN_ERR_TIMEOUT 32757 + +#define SPRAID_WAIT_ABNL_CMD_TIMEOUT (3 * 2) + +#define SPRAID_DMA_MSK_BIT_MAX 64 + +enum FW_STAT_CODE { + FW_STAT_OK = 0, + FW_STAT_NEED_CHECK, + FW_STAT_ERROR, + FW_STAT_EP_PCIE_ERROR, + FW_STAT_NAC_DMA_ERROR, + FW_STAT_ABORTED, + FW_STAT_NEED_RETRY +}; + +static const char * const raid_levels[] = {"0", "1", "5", "6", "10", "50", "60", + "NA"}; + +static const char * const raid_states[] = { + "NA", "NORMAL", "FAULT", "DEGRADE", "NOT_FORMATTED", + "FORMATTING", "SANITIZING", "INITIALIZING", "INITIALIZE_FAIL", + "DELETING", "DELETE_FAIL", "WRITE_PROTECT" +}; + +static int ioq_depth_set(const char *val, const struct kernel_param *kp) +{ + int n = 0; + int ret; + + ret = kstrtoint(val, 10, &n); + if (ret != 0 || n < 2) + return -EINVAL; + + return param_set_int(val, kp); +} + +static int spraid_remap_bar(struct spraid_dev *hdev, u32 size) +{ + struct pci_dev *pdev = hdev->pdev; + + if (size > pci_resource_len(pdev, 0)) { + dev_err(hdev->dev, "Input size[%u] exceed bar0 length[%llu]\n", + size, pci_resource_len(pdev, 0)); + return -ENOMEM; + } + + if (hdev->bar) + iounmap(hdev->bar); + + hdev->bar = ioremap(pci_resource_start(pdev, 0), size); + if (!hdev->bar) { + dev_err(hdev->dev, "ioremap for bar0 failed\n"); + return -ENOMEM; + } + hdev->dbs = hdev->bar + SPRAID_REG_DBS; + + return 0; +} + +static int spraid_dev_map(struct spraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + int ret; + + ret = pci_request_mem_regions(pdev, "spraid"); + if (ret) { + dev_err(hdev->dev, "fail to request memory regions\n"); + return ret; + } + + ret = spraid_remap_bar(hdev, SPRAID_REG_DBS + 4096); + if (ret) { + pci_release_mem_regions(pdev); + return ret; + } + + return 0; +} + +static void spraid_dev_unmap(struct spraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + + if (hdev->bar) { + iounmap(hdev->bar); + hdev->bar = NULL; + } + pci_release_mem_regions(pdev); +} + +static int spraid_pci_enable(struct spraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + int ret = -ENOMEM; + u64 maskbit = SPRAID_DMA_MSK_BIT_MAX; + + if (pci_enable_device_mem(pdev)) { + dev_err(hdev->dev, + "Enable pci device memory resources failed\n"); + return ret; + } + pci_set_master(pdev); + + if (readl(hdev->bar + SPRAID_REG_CSTS) == U32_MAX) { + ret = -ENODEV; + dev_err(hdev->dev, "Read csts register failed\n"); + goto disable; + } + + ret = pci_alloc_irq_vectors(pdev, 1, 1, PCI_IRQ_ALL_TYPES); + if (ret < 0) { + dev_err(hdev->dev, + "Allocate one IRQ for setup admin channel failed\n"); + goto disable; + } + + hdev->cap = lo_hi_readq(hdev->bar + SPRAID_REG_CAP); + hdev->ioq_depth = min_t(u32, SPRAID_CAP_MQES(hdev->cap) + 1, + io_queue_depth); + hdev->db_stride = 1 << SPRAID_CAP_STRIDE(hdev->cap); + + maskbit = SPRAID_CAP_DMAMASK(hdev->cap); + if (maskbit < 32 || maskbit > SPRAID_DMA_MSK_BIT_MAX) { + dev_err(hdev->dev, + "err, dma mask invalid[%llu], set to default\n", + maskbit); + maskbit = SPRAID_DMA_MSK_BIT_MAX; + } + if (dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(maskbit))) { + dev_err(hdev->dev, "set dma mask and coherent failed\n"); + goto disable; + } + + dev_info(hdev->dev, "set dma mask[%llu] success\n", maskbit); + + pci_enable_pcie_error_reporting(pdev); + pci_save_state(pdev); + + return 0; + +disable: + pci_disable_device(pdev); + return ret; +} + +static int spraid_npages_prp(u32 size, struct spraid_dev *hdev) +{ + u32 nprps = DIV_ROUND_UP(size + hdev->page_size, hdev->page_size); + + return DIV_ROUND_UP(PRP_ENTRY_SIZE * nprps, PAGE_SIZE - PRP_ENTRY_SIZE); +} + +static int spraid_npages_sgl(u32 nseg) +{ + return DIV_ROUND_UP(nseg * sizeof(struct spraid_sgl_desc), PAGE_SIZE); +} + +static void **spraid_iod_list(struct spraid_iod *iod) +{ + return (void **)(iod->inline_sg + (iod->sg_drv_mgmt ? iod->nsge : 0)); +} + +static u32 spraid_iod_ext_size(struct spraid_dev *hdev, u32 size, u32 nsge, + bool sg_drv_mgmt, bool use_sgl) +{ + size_t alloc_size, sg_size; + + if (use_sgl) + alloc_size = sizeof(__le64 *) * spraid_npages_sgl(nsge); + else + alloc_size = sizeof(__le64 *) * spraid_npages_prp(size, hdev); + + sg_size = sg_drv_mgmt ? (sizeof(struct scatterlist) * nsge) : 0; + return sg_size + alloc_size; +} + +static u32 spraid_cmd_size(struct spraid_dev *hdev, + bool sg_drv_mgmt, bool use_sgl) +{ + u32 alloc_size = spraid_iod_ext_size(hdev, SPRAID_INT_BYTES(hdev), + SPRAID_INT_PAGES, sg_drv_mgmt, use_sgl); + + dev_info(hdev->dev, "sg_drv_mgmt: %s, use_sgl: %s, iod size: %lu;" + " alloc_size: %u\n", sg_drv_mgmt ? "true" : "false", + use_sgl ? "true" : "false", + sizeof(struct spraid_iod), alloc_size); + + return sizeof(struct spraid_iod) + alloc_size; +} + +static int spraid_setup_prps(struct spraid_dev *hdev, struct spraid_iod *iod) +{ + struct scatterlist *sg = iod->sg; + u64 dma_addr = sg_dma_address(sg); + int dma_len = sg_dma_len(sg); + __le64 *prp_list, *old_prp_list; + u32 page_size = hdev->page_size; + int offset = dma_addr & (page_size - 1); + void **list = spraid_iod_list(iod); + int length = iod->length; + struct dma_pool *pool; + dma_addr_t prp_dma; + int nprps, i; + + length -= (page_size - offset); + if (length <= 0) { + iod->first_dma = 0; + return 0; + } + + dma_len -= (page_size - offset); + if (dma_len) { + dma_addr += (page_size - offset); + } else { + sg = sg_next(sg); + dma_addr = sg_dma_address(sg); + dma_len = sg_dma_len(sg); + } + + if (length <= page_size) { + iod->first_dma = dma_addr; + return 0; + } + + nprps = DIV_ROUND_UP(length, page_size); + if (nprps <= (SMALL_POOL_SIZE / PRP_ENTRY_SIZE)) { + pool = iod->spraidq->prp_small_pool; + iod->npages = 0; + } else { + pool = hdev->prp_page_pool; + iod->npages = 1; + } + + prp_list = dma_pool_alloc(pool, GFP_ATOMIC, &prp_dma); + if (!prp_list) { + dev_err_ratelimited(hdev->dev, + "Allocate first prp_list memory failed\n"); + iod->first_dma = dma_addr; + iod->npages = -1; + return -ENOMEM; + } + list[0] = prp_list; + iod->first_dma = prp_dma; + i = 0; + for (;;) { + if (i == page_size / PRP_ENTRY_SIZE) { + old_prp_list = prp_list; + + prp_list = dma_pool_alloc(pool, GFP_ATOMIC, &prp_dma); + if (!prp_list) { + dev_err_ratelimited(hdev->dev, "Allocate %dth;" + " prp_list memory failed\n", + iod->npages + 1); + return -ENOMEM; + } + list[iod->npages++] = prp_list; + prp_list[0] = old_prp_list[i - 1]; + old_prp_list[i - 1] = cpu_to_le64(prp_dma); + i = 1; + } + prp_list[i++] = cpu_to_le64(dma_addr); + dma_len -= page_size; + dma_addr += page_size; + length -= page_size; + if (length <= 0) + break; + if (dma_len > 0) + continue; + if (unlikely(dma_len < 0)) + goto bad_sgl; + sg = sg_next(sg); + dma_addr = sg_dma_address(sg); + dma_len = sg_dma_len(sg); + } + + return 0; + +bad_sgl: + dev_err(hdev->dev, + "Setup prps, invalid SGL for payload: %d nents: %d\n", + iod->length, iod->nsge); + return -EIO; +} + +#define SGES_PER_PAGE (PAGE_SIZE / sizeof(struct spraid_sgl_desc)) + +static void spraid_submit_cmd(struct spraid_queue *spraidq, const void *cmd) +{ + u32 sqes = SQE_SIZE(spraidq->qid); + unsigned long flags; + struct spraid_admin_common_command *acd = + (struct spraid_admin_common_command *)cmd; + + spin_lock_irqsave(&spraidq->sq_lock, flags); + memcpy((spraidq->sq_cmds + sqes * spraidq->sq_tail), cmd, sqes); + if (++spraidq->sq_tail == spraidq->q_depth) + spraidq->sq_tail = 0; + + writel(spraidq->sq_tail, spraidq->q_db); + spin_unlock_irqrestore(&spraidq->sq_lock, flags); + + dev_log_dbg(spraidq->hdev->dev, + "cid[%d] qid[%d], opcode[0x%x], flags[0x%x], hdid[%u]\n", + acd->command_id, spraidq->qid, acd->opcode, + acd->flags, le32_to_cpu(acd->hdid)); +} + +static u32 spraid_mod64(u64 dividend, u32 divisor) +{ + u64 d; + u32 remainder; + + if (!divisor) + pr_err("DIVISOR is zero, in div fn\n"); + + d = dividend; + remainder = do_div(d, divisor); + return remainder; +} + +static inline bool spraid_is_rw_scmd(struct scsi_cmnd *scmd) +{ + switch (scmd->cmnd[0]) { + case READ_6: + case READ_10: + case READ_12: + case READ_16: + case READ_32: + case WRITE_6: + case WRITE_10: + case WRITE_12: + case WRITE_16: + case WRITE_32: + return true; + default: + return false; + } +} + +static bool spraid_is_prp(struct spraid_dev *hdev, + struct scsi_cmnd *scmd, u32 nsge) +{ + struct scatterlist *sg = scsi_sglist(scmd); + u32 page_size = hdev->page_size; + bool is_prp = true; + int i = 0; + + scsi_for_each_sg(scmd, sg, nsge, i) { + if (i != 0 && i != nsge - 1) { + if (spraid_mod64(sg_dma_len(sg), page_size) || + spraid_mod64(sg_dma_address(sg), page_size)) { + is_prp = false; + break; + } + } + + if (nsge > 1 && i == 0) { + if ((spraid_mod64((sg_dma_address(sg) + sg_dma_len(sg)), + page_size))) { + is_prp = false; + break; + } + } + + if (nsge > 1 && i == (nsge - 1)) { + if (spraid_mod64(sg_dma_address(sg), page_size)) { + is_prp = false; + break; + } + } + } + + return is_prp; +} + +enum { + SPRAID_SGL_FMT_DATA_DESC = 0x00, + SPRAID_SGL_FMT_SEG_DESC = 0x02, + SPRAID_SGL_FMT_LAST_SEG_DESC = 0x03, + SPRAID_KEY_SGL_FMT_DATA_DESC = 0x04, + SPRAID_TRANSPORT_SGL_DATA_DESC = 0x05 +}; + +static void spraid_sgl_set_data(struct spraid_sgl_desc *sge, + struct scatterlist *sg) +{ + sge->addr = cpu_to_le64(sg_dma_address(sg)); + sge->length = cpu_to_le32(sg_dma_len(sg)); + sge->type = SPRAID_SGL_FMT_DATA_DESC << 4; +} + +static void spraid_sgl_set_seg(struct spraid_sgl_desc *sge, + dma_addr_t dma_addr, int entries) +{ + sge->addr = cpu_to_le64(dma_addr); + if (entries <= SGES_PER_PAGE) { + sge->length = cpu_to_le32(entries * sizeof(*sge)); + sge->type = SPRAID_SGL_FMT_LAST_SEG_DESC << 4; + } else { + sge->length = cpu_to_le32(PAGE_SIZE); + sge->type = SPRAID_SGL_FMT_SEG_DESC << 4; + } +} + +static int spraid_setup_ioq_cmd_sgl(struct spraid_dev *hdev, + struct scsi_cmnd *scmd, + struct spraid_ioq_command *ioq_cmd, + struct spraid_iod *iod) +{ + struct spraid_sgl_desc *sg_list, *link, *old_sg_list; + struct scatterlist *sg = scsi_sglist(scmd); + void **list = spraid_iod_list(iod); + struct dma_pool *pool; + int nsge = iod->nsge; + dma_addr_t sgl_dma; + int i = 0; + + ioq_cmd->common.flags |= SPRAID_CMD_FLAG_SGL_METABUF; + + if (nsge == 1) { + spraid_sgl_set_data(&ioq_cmd->common.dptr.sgl, sg); + return 0; + } + + if (nsge <= (SMALL_POOL_SIZE / sizeof(struct spraid_sgl_desc))) { + pool = iod->spraidq->prp_small_pool; + iod->npages = 0; + } else { + pool = hdev->prp_page_pool; + iod->npages = 1; + } + + sg_list = dma_pool_alloc(pool, GFP_ATOMIC, &sgl_dma); + if (!sg_list) { + dev_err_ratelimited(hdev->dev, + "Allocate first sgl_list failed\n"); + iod->npages = -1; + return -ENOMEM; + } + + list[0] = sg_list; + iod->first_dma = sgl_dma; + spraid_sgl_set_seg(&ioq_cmd->common.dptr.sgl, sgl_dma, nsge); + do { + if (i == SGES_PER_PAGE) { + old_sg_list = sg_list; + link = &old_sg_list[SGES_PER_PAGE - 1]; + + sg_list = dma_pool_alloc(pool, GFP_ATOMIC, &sgl_dma); + if (!sg_list) { + dev_err_ratelimited(hdev->dev, + "Allocate %dth sgl_list;" + " failed\n", + iod->npages + 1); + return -ENOMEM; + } + list[iod->npages++] = sg_list; + + i = 0; + memcpy(&sg_list[i++], link, sizeof(*link)); + spraid_sgl_set_seg(link, sgl_dma, nsge); + } + + spraid_sgl_set_data(&sg_list[i++], sg); + sg = sg_next(sg); + } while (--nsge > 0); + + return 0; +} + +#define SPRAID_RW_FUA BIT(14) + +static void spraid_setup_rw_cmd(struct spraid_dev *hdev, + struct spraid_rw_command *rw, + struct scsi_cmnd *scmd) +{ + u32 start_lba_lo, start_lba_hi; + u32 datalength = 0; + u16 control = 0; + + start_lba_lo = 0; + start_lba_hi = 0; + + if (scmd->sc_data_direction == DMA_TO_DEVICE) { + rw->opcode = SPRAID_CMD_WRITE; + } else if (scmd->sc_data_direction == DMA_FROM_DEVICE) { + rw->opcode = SPRAID_CMD_READ; + } else { + dev_err(hdev->dev, + "Invalid IO for unsupported data direction: %d\n", + scmd->sc_data_direction); + WARN_ON(1); + } + + /* 6-byte READ(0x08) or WRITE(0x0A) cdb */ + if (scmd->cmd_len == 6) { + datalength = (u32)(scmd->cmnd[4] == 0 ? + IO_6_DEFAULT_TX_LEN : scmd->cmnd[4]); + start_lba_lo = (u32)get_unaligned_be24(&scmd->cmnd[1]); + + start_lba_lo &= 0x1FFFFF; + } + + /* 10-byte READ(0x28) or WRITE(0x2A) cdb */ + else if (scmd->cmd_len == 10) { + datalength = (u32)get_unaligned_be16(&scmd->cmnd[7]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[2]); + + if (scmd->cmnd[1] & FUA_MASK) + control |= SPRAID_RW_FUA; + } + + /* 12-byte READ(0xA8) or WRITE(0xAA) cdb */ + else if (scmd->cmd_len == 12) { + datalength = get_unaligned_be32(&scmd->cmnd[6]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[2]); + + if (scmd->cmnd[1] & FUA_MASK) + control |= SPRAID_RW_FUA; + } + /* 16-byte READ(0x88) or WRITE(0x8A) cdb */ + else if (scmd->cmd_len == 16) { + datalength = get_unaligned_be32(&scmd->cmnd[10]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[6]); + start_lba_hi = get_unaligned_be32(&scmd->cmnd[2]); + + if (scmd->cmnd[1] & FUA_MASK) + control |= SPRAID_RW_FUA; + } + /* 32-byte READ(0x88) or WRITE(0x8A) cdb */ + else if (scmd->cmd_len == 32) { + datalength = get_unaligned_be32(&scmd->cmnd[28]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[16]); + start_lba_hi = get_unaligned_be32(&scmd->cmnd[12]); + + if (scmd->cmnd[10] & FUA_MASK) + control |= SPRAID_RW_FUA; + } + + if (unlikely(datalength > U16_MAX || datalength == 0)) { + dev_err(hdev->dev, + "Invalid IO for illegal transfer data length: %u\n", + datalength); + WARN_ON(1); + } + + rw->slba = cpu_to_le64(((u64)start_lba_hi << 32) | start_lba_lo); + /* 0base for nlb */ + rw->nlb = cpu_to_le16((u16)(datalength - 1)); + rw->control = cpu_to_le16(control); +} + +static void spraid_setup_nonio_cmd(struct spraid_dev *hdev, + struct spraid_scsi_nonio *scsi_nonio, + struct scsi_cmnd *scmd) +{ + scsi_nonio->buffer_len = cpu_to_le32(scsi_bufflen(scmd)); + + switch (scmd->sc_data_direction) { + case DMA_NONE: + scsi_nonio->opcode = SPRAID_CMD_NONIO_NONE; + break; + case DMA_TO_DEVICE: + scsi_nonio->opcode = SPRAID_CMD_NONIO_TODEV; + break; + case DMA_FROM_DEVICE: + scsi_nonio->opcode = SPRAID_CMD_NONIO_FROMDEV; + break; + default: + dev_err(hdev->dev, + "Invalid IO for unsupported data direction: %d\n", + scmd->sc_data_direction); + WARN_ON(1); + } +} + +static void spraid_setup_ioq_cmd(struct spraid_dev *hdev, + struct spraid_ioq_command *ioq_cmd, + struct scsi_cmnd *scmd) +{ + memcpy(ioq_cmd->common.cdb, scmd->cmnd, scmd->cmd_len); + ioq_cmd->common.cdb_len = scmd->cmd_len; + + if (spraid_is_rw_scmd(scmd)) + spraid_setup_rw_cmd(hdev, &ioq_cmd->rw, scmd); + else + spraid_setup_nonio_cmd(hdev, &ioq_cmd->scsi_nonio, scmd); +} + +static int spraid_init_iod(struct spraid_dev *hdev, struct spraid_iod *iod, + struct spraid_ioq_command *ioq_cmd, + struct scsi_cmnd *scmd) +{ + if (unlikely(!iod->sense)) { + dev_err(hdev->dev, "Allocate sense data buffer failed\n"); + return -ENOMEM; + } + ioq_cmd->common.sense_addr = cpu_to_le64(iod->sense_dma); + ioq_cmd->common.sense_len = cpu_to_le16(SCSI_SENSE_BUFFERSIZE); + + iod->nsge = 0; + iod->npages = -1; + iod->use_sgl = 0; + iod->sg_drv_mgmt = false; + WRITE_ONCE(iod->state, SPRAID_CMD_IDLE); + + return 0; +} + +static void spraid_free_iod_res(struct spraid_dev *hdev, struct spraid_iod *iod) +{ + const int last_prp = hdev->page_size / sizeof(__le64) - 1; + dma_addr_t dma_addr, next_dma_addr; + struct spraid_sgl_desc *sg_list; + __le64 *prp_list; + void *addr; + int i; + + dma_addr = iod->first_dma; + if (iod->npages == 0) + dma_pool_free(iod->spraidq->prp_small_pool, + spraid_iod_list(iod)[0], dma_addr); + + for (i = 0; i < iod->npages; i++) { + addr = spraid_iod_list(iod)[i]; + + if (iod->use_sgl) { + sg_list = addr; + next_dma_addr = + le64_to_cpu((sg_list[SGES_PER_PAGE - 1]).addr); + } else { + prp_list = addr; + next_dma_addr = le64_to_cpu(prp_list[last_prp]); + } + + dma_pool_free(hdev->prp_page_pool, addr, dma_addr); + dma_addr = next_dma_addr; + } + + if (iod->sg_drv_mgmt && iod->sg != iod->inline_sg) { + iod->sg_drv_mgmt = false; + mempool_free(iod->sg, hdev->iod_mempool); + } + + iod->sense = NULL; + iod->npages = -1; +} + +static int spraid_io_map_data(struct spraid_dev *hdev, struct spraid_iod *iod, + struct scsi_cmnd *scmd, + struct spraid_ioq_command *ioq_cmd) +{ + int ret; + + iod->nsge = scsi_dma_map(scmd); + + /* No data to DMA, it may be scsi no-rw command */ + if (unlikely(iod->nsge == 0)) + return 0; + + iod->length = scsi_bufflen(scmd); + iod->sg = scsi_sglist(scmd); + iod->use_sgl = !spraid_is_prp(hdev, scmd, iod->nsge); + + if (iod->use_sgl) { + ret = spraid_setup_ioq_cmd_sgl(hdev, scmd, ioq_cmd, iod); + } else { + ret = spraid_setup_prps(hdev, iod); + ioq_cmd->common.dptr.prp1 = + cpu_to_le64(sg_dma_address(iod->sg)); + ioq_cmd->common.dptr.prp2 = cpu_to_le64(iod->first_dma); + } + + if (ret) + scsi_dma_unmap(scmd); + + return ret; +} + +static void spraid_map_status(struct spraid_iod *iod, struct scsi_cmnd *scmd, + struct spraid_completion *cqe) +{ + scsi_set_resid(scmd, 0); + + switch ((le16_to_cpu(cqe->status) >> 1) & 0x7f) { + case FW_STAT_OK: + set_host_byte(scmd, DID_OK); + break; + case FW_STAT_NEED_CHECK: + set_host_byte(scmd, DID_OK); + scmd->result |= le16_to_cpu(cqe->status) >> 8; + if (scmd->result & SAM_STAT_CHECK_CONDITION) { + memset(scmd->sense_buffer, 0, SCSI_SENSE_BUFFERSIZE); + memcpy(scmd->sense_buffer, iod->sense, + SCSI_SENSE_BUFFERSIZE); + scmd->result = + (scmd->result & 0x00ffffff) | (DRIVER_SENSE << 24); + } + break; + case FW_STAT_ABORTED: + set_host_byte(scmd, DID_ABORT); + break; + case FW_STAT_NEED_RETRY: + set_host_byte(scmd, DID_REQUEUE); + break; + default: + set_host_byte(scmd, DID_BAD_TARGET); + break; + } +} + +static inline void spraid_get_tag_from_scmd(struct scsi_cmnd *scmd, + u16 *qid, u16 *cid) +{ + u32 tag = blk_mq_unique_tag(scmd->request); + + *qid = blk_mq_unique_tag_to_hwq(tag) + 1; + *cid = blk_mq_unique_tag_to_tag(tag); +} + +static int spraid_queue_command(struct Scsi_Host *shost, struct scsi_cmnd *scmd) +{ + struct spraid_iod *iod = scsi_cmd_priv(scmd); + struct spraid_dev *hdev = shost_priv(shost); + struct scsi_device *sdev = scmd->device; + struct spraid_sdev_hostdata *hostdata; + struct spraid_ioq_command ioq_cmd; + struct spraid_queue *ioq; + unsigned long elapsed; + u16 hwq, cid; + int ret; + + if (unlikely(!scmd)) { + dev_err(hdev->dev, "err, scmd is null\n"); + return 0; + } + + if (unlikely(hdev->state != SPRAID_LIVE)) { + set_host_byte(scmd, DID_NO_CONNECT); + scmd->scsi_done(scmd); + return 0; + } + + if (log_debug_switch) + scsi_print_command(scmd); + + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + hostdata = sdev->hostdata; + ioq = &hdev->queues[hwq]; + memset(&ioq_cmd, 0, sizeof(ioq_cmd)); + ioq_cmd.rw.hdid = cpu_to_le32(hostdata->hdid); + ioq_cmd.rw.command_id = cid; + + spraid_setup_ioq_cmd(hdev, &ioq_cmd, scmd); + + ret = cid * SCSI_SENSE_BUFFERSIZE; + iod->sense = ioq->sense + ret; + iod->sense_dma = ioq->sense_dma_addr + ret; + + ret = spraid_init_iod(hdev, iod, &ioq_cmd, scmd); + if (unlikely(ret)) + return SCSI_MLQUEUE_HOST_BUSY; + + iod->spraidq = ioq; + ret = spraid_io_map_data(hdev, iod, scmd, &ioq_cmd); + if (unlikely(ret)) { + dev_err(hdev->dev, "spraid_io_map_data Err.\n"); + set_host_byte(scmd, DID_ERROR); + scmd->scsi_done(scmd); + ret = 0; + goto deinit_iod; + } + + WRITE_ONCE(iod->state, SPRAID_CMD_IN_FLIGHT); + spraid_submit_cmd(ioq, &ioq_cmd); + elapsed = jiffies - scmd->jiffies_at_alloc; + dev_log_dbg(hdev->dev, + "cid[%d] qid[%d] submit IO cost %3ld.%3ld seconds\n", + cid, hwq, elapsed / HZ, elapsed % HZ); + return 0; + +deinit_iod: + spraid_free_iod_res(hdev, iod); + return ret; +} + +static int spraid_match_dev(struct spraid_dev *hdev, u16 idx, + struct scsi_device *sdev) +{ + if (SPRAID_DEV_INFO_FLAG_VALID(hdev->devices[idx].flag)) { + if (sdev->channel == hdev->devices[idx].channel && + sdev->id == le16_to_cpu(hdev->devices[idx].target) && + sdev->lun < hdev->devices[idx].lun) { + dev_info(hdev->dev, + "Match device success, channel;" + "target:lun[%d:%d:%d]\n", + hdev->devices[idx].channel, + hdev->devices[idx].target, + hdev->devices[idx].lun); + return 1; + } + } + + return 0; +} + +static int spraid_slave_alloc(struct scsi_device *sdev) +{ + struct spraid_sdev_hostdata *hostdata; + struct spraid_dev *hdev; + u16 idx; + + hdev = shost_priv(sdev->host); + hostdata = kzalloc(sizeof(*hostdata), GFP_KERNEL); + if (!hostdata) { + dev_err(hdev->dev, "Alloc scsi host data memory failed\n"); + return -ENOMEM; + } + + down_read(&hdev->devices_rwsem); + for (idx = 0; idx < le32_to_cpu(hdev->ctrl_info->nd); idx++) { + if (spraid_match_dev(hdev, idx, sdev)) + goto scan_host; + } + up_read(&hdev->devices_rwsem); + + kfree(hostdata); + return -ENXIO; + +scan_host: + hostdata->hdid = le32_to_cpu(hdev->devices[idx].hdid); + hostdata->max_io_kb = le16_to_cpu(hdev->devices[idx].max_io_kb); + hostdata->attr = hdev->devices[idx].attr; + hostdata->flag = hdev->devices[idx].flag; + hostdata->rg_id = 0xff; + sdev->hostdata = hostdata; + up_read(&hdev->devices_rwsem); + return 0; +} + +static void spraid_slave_destroy(struct scsi_device *sdev) +{ + kfree(sdev->hostdata); + sdev->hostdata = NULL; +} + +static int spraid_slave_configure(struct scsi_device *sdev) +{ + u16 idx; + unsigned int timeout = scmd_tmout_nonpt * HZ; + struct spraid_dev *hdev = shost_priv(sdev->host); + struct spraid_sdev_hostdata *hostdata = sdev->hostdata; + u32 max_sec = sdev->host->max_sectors; + + if (hostdata) { + idx = hostdata->hdid - 1; + if (sdev->channel == hdev->devices[idx].channel && + sdev->id == le16_to_cpu(hdev->devices[idx].target) && + sdev->lun < hdev->devices[idx].lun) { + if (SPRAID_DEV_INFO_ATTR_PT(hdev->devices[idx].attr)) + timeout = 30 * HZ; + else + timeout = scmd_tmout_nonpt * HZ; + max_sec = le16_to_cpu(hdev->devices[idx].max_io_kb) + << 1; + } else { + dev_err(hdev->dev, + "[%s] err, sdev->channel:id:lun[%d:%d:%lld];" + "devices[%d], channel:target:lun[%d:%d:%d]\n", + __func__, sdev->channel, sdev->id, sdev->lun, + idx, hdev->devices[idx].channel, + hdev->devices[idx].target, + hdev->devices[idx].lun); + } + } else { + dev_err(hdev->dev, "[%s] err, sdev->hostdata is null\n", + __func__); + } + + blk_queue_rq_timeout(sdev->request_queue, timeout); + sdev->eh_timeout = timeout; + + if ((max_sec == 0) || (max_sec > sdev->host->max_sectors)) + max_sec = sdev->host->max_sectors; + + dev_info(hdev->dev, + "[%s] sdev->channel:id:lun[%d:%d:%lld];" + " scmd_timeout[%d]s, maxsec[%d]\n", + __func__, sdev->channel, sdev->id, + sdev->lun, timeout / HZ, max_sec); + + return 0; +} + +static void spraid_shost_init(struct spraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + u8 domain, bus; + u32 dev_func; + + domain = pci_domain_nr(pdev->bus); + bus = pdev->bus->number; + dev_func = pdev->devfn; + + hdev->shost->nr_hw_queues = hdev->online_queues - 1; + hdev->shost->can_queue = (hdev->ioq_depth - SPRAID_PTCMDS_PERQ); + + hdev->shost->sg_tablesize = le16_to_cpu(hdev->ctrl_info->max_num_sge); + /* 512B per sector */ + hdev->shost->max_sectors = + (1U << ((hdev->ctrl_info->mdts) * 1U) << 12) / 512; + hdev->shost->cmd_per_lun = MAX_CMD_PER_DEV; + hdev->shost->max_channel = + le16_to_cpu(hdev->ctrl_info->max_channel) - 1; + hdev->shost->max_id = le32_to_cpu(hdev->ctrl_info->max_tgt_id); + hdev->shost->max_lun = le16_to_cpu(hdev->ctrl_info->max_lun); + + hdev->shost->this_id = -1; + hdev->shost->unique_id = (domain << 16) | (bus << 8) | dev_func; + hdev->shost->max_cmd_len = MAX_CDB_LEN; + hdev->shost->hostt->cmd_size = max(spraid_cmd_size(hdev, false, true), + spraid_cmd_size(hdev, false, false)); +} + +static inline void spraid_host_deinit(struct spraid_dev *hdev) +{ + ida_free(&spraid_instance_ida, hdev->instance); +} + +static int spraid_alloc_queue(struct spraid_dev *hdev, u16 qid, u16 depth) +{ + struct spraid_queue *spraidq = &hdev->queues[qid]; + int ret = 0; + + if (hdev->queue_count > qid) { + dev_info(hdev->dev, "[%s] warn: queue[%d] is exist\n", + __func__, qid); + return 0; + } + + spraidq->cqes = dma_alloc_coherent(hdev->dev, CQ_SIZE(depth), + &spraidq->cq_dma_addr, + GFP_KERNEL | __GFP_ZERO); + if (!spraidq->cqes) + return -ENOMEM; + + spraidq->sq_cmds = dma_alloc_coherent(hdev->dev, SQ_SIZE(qid, depth), + &spraidq->sq_dma_addr, + GFP_KERNEL); + if (!spraidq->sq_cmds) { + ret = -ENOMEM; + goto free_cqes; + } + + spin_lock_init(&spraidq->sq_lock); + spin_lock_init(&spraidq->cq_lock); + spraidq->hdev = hdev; + spraidq->q_depth = depth; + spraidq->qid = qid; + spraidq->cq_vector = -1; + hdev->queue_count++; + + /* alloc sense buffer */ + spraidq->sense = dma_alloc_coherent(hdev->dev, SENSE_SIZE(depth), + &spraidq->sense_dma_addr, + GFP_KERNEL | __GFP_ZERO); + if (!spraidq->sense) { + ret = -ENOMEM; + goto free_sq_cmds; + } + + return 0; + +free_sq_cmds: + dma_free_coherent(hdev->dev, SQ_SIZE(qid, depth), + (void *)spraidq->sq_cmds, spraidq->sq_dma_addr); +free_cqes: + dma_free_coherent(hdev->dev, CQ_SIZE(depth), (void *)spraidq->cqes, + spraidq->cq_dma_addr); + return ret; +} + +static int spraid_wait_ready(struct spraid_dev *hdev, u64 cap, bool enabled) +{ + unsigned long timeout = + ((SPRAID_CAP_TIMEOUT(cap) + 1) * SPRAID_CAP_TIMEOUT_UNIT_MS) + jiffies; + u32 bit = enabled ? SPRAID_CSTS_RDY : 0; + + while ((readl(hdev->bar + SPRAID_REG_CSTS) & SPRAID_CSTS_RDY) != bit) { + usleep_range(1000, 2000); + if (fatal_signal_pending(current)) + return -EINTR; + + if (time_after(jiffies, timeout)) { + dev_err(hdev->dev, "Device not ready; aborting %s\n", + enabled ? "initialisation" : "reset"); + return -ENODEV; + } + } + return 0; +} + +static int spraid_shutdown_ctrl(struct spraid_dev *hdev) +{ + unsigned long timeout = hdev->ctrl_info->rtd3e + jiffies; + + hdev->ctrl_config &= ~SPRAID_CC_SHN_MASK; + hdev->ctrl_config |= SPRAID_CC_SHN_NORMAL; + writel(hdev->ctrl_config, hdev->bar + SPRAID_REG_CC); + + while ((readl(hdev->bar + SPRAID_REG_CSTS) & SPRAID_CSTS_SHST_MASK) != + SPRAID_CSTS_SHST_CMPLT) { + msleep(100); + if (fatal_signal_pending(current)) + return -EINTR; + if (time_after(jiffies, timeout)) { + dev_err(hdev->dev, + "Device shutdown incomplete; abort shutdown\n"); + return -ENODEV; + } + } + return 0; +} + +static int spraid_disable_ctrl(struct spraid_dev *hdev) +{ + hdev->ctrl_config &= ~SPRAID_CC_SHN_MASK; + hdev->ctrl_config &= ~SPRAID_CC_ENABLE; + writel(hdev->ctrl_config, hdev->bar + SPRAID_REG_CC); + + return spraid_wait_ready(hdev, hdev->cap, false); +} + +static int spraid_enable_ctrl(struct spraid_dev *hdev) +{ + u64 cap = hdev->cap; + u32 dev_page_min = SPRAID_CAP_MPSMIN(cap) + 12; + u32 page_shift = PAGE_SHIFT; + + if (page_shift < dev_page_min) { + dev_err(hdev->dev, + "Minimum device page size[%u], too large for host[%u]\n", + 1U << dev_page_min, 1U << page_shift); + return -ENODEV; + } + + page_shift = min_t(unsigned int, SPRAID_CAP_MPSMAX(cap) + 12, + PAGE_SHIFT); + hdev->page_size = 1U << page_shift; + + hdev->ctrl_config = SPRAID_CC_CSS_NVM; + hdev->ctrl_config |= (page_shift - 12) << SPRAID_CC_MPS_SHIFT; + hdev->ctrl_config |= SPRAID_CC_AMS_RR | SPRAID_CC_SHN_NONE; + hdev->ctrl_config |= SPRAID_CC_IOSQES | SPRAID_CC_IOCQES; + hdev->ctrl_config |= SPRAID_CC_ENABLE; + writel(hdev->ctrl_config, hdev->bar + SPRAID_REG_CC); + + return spraid_wait_ready(hdev, cap, true); +} + +static void spraid_init_queue(struct spraid_queue *spraidq, u16 qid) +{ + struct spraid_dev *hdev = spraidq->hdev; + + memset((void *)spraidq->cqes, 0, CQ_SIZE(spraidq->q_depth)); + + spraidq->sq_tail = 0; + spraidq->cq_head = 0; + spraidq->cq_phase = 1; + spraidq->q_db = &hdev->dbs[qid * 2 * hdev->db_stride]; + spraidq->prp_small_pool = hdev->prp_small_pool[qid % small_pool_num]; + hdev->online_queues++; +} + +static inline bool spraid_cqe_pending(struct spraid_queue *spraidq) +{ + return (le16_to_cpu(spraidq->cqes[spraidq->cq_head].status) & 1) == + spraidq->cq_phase; +} + +static void spraid_sata_report_zone_handle(struct scsi_cmnd *scmd, + struct spraid_iod *iod) +{ + int i = 0; + unsigned int bytes = 0; + struct scatterlist *sg = scsi_sglist(scmd); + + scsi_for_each_sg(scmd, sg, iod->nsge, i) { + unsigned int offset = 0; + + if (bytes == 0) { + char *hdr; + u32 list_length; + u64 max_lba, opt_lba; + u16 same; + + hdr = sg_virt(sg); + + list_length = get_unaligned_le32(&hdr[0]); + same = get_unaligned_le16(&hdr[4]); + max_lba = get_unaligned_le64(&hdr[8]); + opt_lba = get_unaligned_le64(&hdr[16]); + put_unaligned_be32(list_length, &hdr[0]); + hdr[4] = same & 0xf; + put_unaligned_be64(max_lba, &hdr[8]); + put_unaligned_be64(opt_lba, &hdr[16]); + offset += 64; + bytes += 64; + } + while (offset < sg_dma_len(sg)) { + char *rec; + u8 cond, type, non_seq, reset; + u64 size, start, wp; + + rec = sg_virt(sg) + offset; + type = rec[0] & 0xf; + cond = (rec[1] >> 4) & 0xf; + non_seq = (rec[1] & 2); + reset = (rec[1] & 1); + size = get_unaligned_le64(&rec[8]); + start = get_unaligned_le64(&rec[16]); + wp = get_unaligned_le64(&rec[24]); + rec[0] = type; + rec[1] = (cond << 4) | non_seq | reset; + put_unaligned_be64(size, &rec[8]); + put_unaligned_be64(start, &rec[16]); + put_unaligned_be64(wp, &rec[24]); + WARN_ON(offset + 64 > sg_dma_len(sg)); + offset += 64; + bytes += 64; + } + } +} + +static inline void spraid_handle_ata_cmd(struct spraid_dev *hdev, + struct scsi_cmnd *scmd, + struct spraid_iod *iod) +{ + if (hdev->ctrl_info->card_type != SPRAID_CARD_HBA) + return; + + switch (scmd->cmnd[0]) { + case ZBC_IN: + dev_info(hdev->dev, "[%s] process report zone\n", __func__); + spraid_sata_report_zone_handle(scmd, iod); + break; + default: + break; + } +} + +static void spraid_complete_ioq_cmnd(struct spraid_queue *ioq, + struct spraid_completion *cqe) +{ + struct spraid_dev *hdev = ioq->hdev; + struct blk_mq_tags *tags; + struct scsi_cmnd *scmd; + struct spraid_iod *iod; + struct request *req; + unsigned long elapsed; + + tags = hdev->shost->tag_set.tags[ioq->qid - 1]; + req = blk_mq_tag_to_rq(tags, cqe->cmd_id); + if (unlikely(!req || !blk_mq_request_started(req))) { + dev_warn(hdev->dev, "Invalid id %d completed on queue %d\n", + cqe->cmd_id, ioq->qid); + return; + } + + scmd = blk_mq_rq_to_pdu(req); + iod = scsi_cmd_priv(scmd); + + elapsed = jiffies - scmd->jiffies_at_alloc; + dev_log_dbg(hdev->dev, + "cid[%d] qid[%d] finish IO cost %3ld.%3ld seconds\n", + cqe->cmd_id, ioq->qid, elapsed / HZ, elapsed % HZ); + + if (cmpxchg(&iod->state, SPRAID_CMD_IN_FLIGHT, SPRAID_CMD_COMPLETE) != + SPRAID_CMD_IN_FLIGHT) { + dev_warn(hdev->dev, + "cid[%d] qid[%d] enters abnormal handler;" + " cost %3ld.%3ld seconds\n", + cqe->cmd_id, ioq->qid, elapsed / HZ, elapsed % HZ); + WRITE_ONCE(iod->state, SPRAID_CMD_TMO_COMPLETE); + + if (iod->nsge) { + iod->nsge = 0; + scsi_dma_unmap(scmd); + } + spraid_free_iod_res(hdev, iod); + + return; + } + + spraid_handle_ata_cmd(hdev, scmd, iod); + + spraid_map_status(iod, scmd, cqe); + if (iod->nsge) { + iod->nsge = 0; + scsi_dma_unmap(scmd); + } + spraid_free_iod_res(hdev, iod); + scmd->scsi_done(scmd); +} + +static void spraid_complete_adminq_cmnd(struct spraid_queue *adminq, + struct spraid_completion *cqe) +{ + struct spraid_dev *hdev = adminq->hdev; + struct spraid_cmd *adm_cmd; + + adm_cmd = hdev->adm_cmds + cqe->cmd_id; + if (unlikely(adm_cmd->state == SPRAID_CMD_IDLE)) { + dev_warn(adminq->hdev->dev, + "Invalid id %d completed on queue %d\n", + cqe->cmd_id, le16_to_cpu(cqe->sq_id)); + return; + } + + adm_cmd->status = le16_to_cpu(cqe->status) >> 1; + adm_cmd->result0 = le32_to_cpu(cqe->result); + adm_cmd->result1 = le32_to_cpu(cqe->result1); + + complete(&adm_cmd->cmd_done); +} + +static void spraid_send_aen(struct spraid_dev *hdev, u16 cid); + +static void spraid_complete_aen(struct spraid_queue *spraidq, + struct spraid_completion *cqe) +{ + struct spraid_dev *hdev = spraidq->hdev; + u32 result = le32_to_cpu(cqe->result); + + dev_info(hdev->dev, "rcv aen, cid[%d], status[0x%x], result[0x%x]\n", + cqe->cmd_id, le16_to_cpu(cqe->status) >> 1, result); + + spraid_send_aen(hdev, cqe->cmd_id); + + if ((le16_to_cpu(cqe->status) >> 1) != SPRAID_SC_SUCCESS) + return; + switch (result & 0x7) { + case SPRAID_AEN_NOTICE: + spraid_handle_aen_notice(hdev, result); + break; + case SPRAID_AEN_VS: + spraid_handle_aen_vs(hdev, result, le32_to_cpu(cqe->result1)); + break; + default: + dev_warn(hdev->dev, "Unsupported async event type: %u\n", + result & 0x7); + break; + } +} + +static void spraid_complete_ioq_sync_cmnd(struct spraid_queue *ioq, + struct spraid_completion *cqe) +{ + struct spraid_dev *hdev = ioq->hdev; + struct spraid_cmd *ptcmd; + + ptcmd = hdev->ioq_ptcmds + (ioq->qid - 1) * SPRAID_PTCMDS_PERQ + + cqe->cmd_id - SPRAID_IO_BLK_MQ_DEPTH; + + ptcmd->status = le16_to_cpu(cqe->status) >> 1; + ptcmd->result0 = le32_to_cpu(cqe->result); + ptcmd->result1 = le32_to_cpu(cqe->result1); + + complete(&ptcmd->cmd_done); +} + +static inline void spraid_handle_cqe(struct spraid_queue *spraidq, u16 idx) +{ + struct spraid_completion *cqe = &spraidq->cqes[idx]; + struct spraid_dev *hdev = spraidq->hdev; + + if (unlikely(cqe->cmd_id >= spraidq->q_depth)) { + dev_err(hdev->dev, + "Invalid command id[%d] completed on queue %d\n", + cqe->cmd_id, cqe->sq_id); + return; + } + + dev_log_dbg(hdev->dev, "cid[%d] qid[%d];" + " result[0x%x], sq_id[%d], status[0x%x]\n", + cqe->cmd_id, spraidq->qid, le32_to_cpu(cqe->result), + le16_to_cpu(cqe->sq_id), le16_to_cpu(cqe->status)); + + if (unlikely(spraidq->qid == 0 + && cqe->cmd_id >= SPRAID_AQ_BLK_MQ_DEPTH)) { + spraid_complete_aen(spraidq, cqe); + return; + } + + if (unlikely(spraidq->qid && cqe->cmd_id >= SPRAID_IO_BLK_MQ_DEPTH)) { + spraid_complete_ioq_sync_cmnd(spraidq, cqe); + return; + } + + if (spraidq->qid) + spraid_complete_ioq_cmnd(spraidq, cqe); + else + spraid_complete_adminq_cmnd(spraidq, cqe); +} + +static void spraid_complete_cqes(struct spraid_queue *spraidq, + u16 start, u16 end) +{ + while (start != end) { + spraid_handle_cqe(spraidq, start); + if (++start == spraidq->q_depth) + start = 0; + } +} + +static inline void spraid_update_cq_head(struct spraid_queue *spraidq) +{ + if (++spraidq->cq_head == spraidq->q_depth) { + spraidq->cq_head = 0; + spraidq->cq_phase = !spraidq->cq_phase; + } +} + +static inline bool spraid_process_cq(struct spraid_queue *spraidq, + u16 *start, u16 *end, int tag) +{ + bool found = false; + + *start = spraidq->cq_head; + while (!found && spraid_cqe_pending(spraidq)) { + if (spraidq->cqes[spraidq->cq_head].cmd_id == tag) + found = true; + spraid_update_cq_head(spraidq); + } + *end = spraidq->cq_head; + + if (*start != *end) + writel(spraidq->cq_head, + spraidq->q_db + spraidq->hdev->db_stride); + + return found; +} + +static bool spraid_poll_cq(struct spraid_queue *spraidq, int cid) +{ + u16 start, end; + bool found; + + if (!spraid_cqe_pending(spraidq)) + return 0; + + spin_lock_irq(&spraidq->cq_lock); + found = spraid_process_cq(spraidq, &start, &end, cid); + spin_unlock_irq(&spraidq->cq_lock); + + spraid_complete_cqes(spraidq, start, end); + return found; +} + +static irqreturn_t spraid_irq(int irq, void *data) +{ + struct spraid_queue *spraidq = data; + irqreturn_t ret = IRQ_NONE; + u16 start, end; + + spin_lock(&spraidq->cq_lock); + if (spraidq->cq_head != spraidq->last_cq_head) + ret = IRQ_HANDLED; + + spraid_process_cq(spraidq, &start, &end, -1); + spraidq->last_cq_head = spraidq->cq_head; + spin_unlock(&spraidq->cq_lock); + + if (start != end) { + spraid_complete_cqes(spraidq, start, end); + ret = IRQ_HANDLED; + } + return ret; +} + +static int spraid_setup_admin_queue(struct spraid_dev *hdev) +{ + struct spraid_queue *adminq = &hdev->queues[0]; + u32 aqa; + int ret; + + dev_info(hdev->dev, "[%s] start disable ctrl\n", __func__); + + ret = spraid_disable_ctrl(hdev); + if (ret) + return ret; + + ret = spraid_alloc_queue(hdev, 0, SPRAID_AQ_DEPTH); + if (ret) + return ret; + + aqa = adminq->q_depth - 1; + aqa |= aqa << 16; + writel(aqa, hdev->bar + SPRAID_REG_AQA); + lo_hi_writeq(adminq->sq_dma_addr, hdev->bar + SPRAID_REG_ASQ); + lo_hi_writeq(adminq->cq_dma_addr, hdev->bar + SPRAID_REG_ACQ); + + dev_info(hdev->dev, "[%s] start enable ctrl\n", __func__); + + ret = spraid_enable_ctrl(hdev); + if (ret) { + ret = -ENODEV; + goto free_queue; + } + + adminq->cq_vector = 0; + spraid_init_queue(adminq, 0); + ret = pci_request_irq(hdev->pdev, adminq->cq_vector, spraid_irq, NULL, + adminq, "spraid%d_q%d", + hdev->instance, adminq->qid); + + if (ret) { + adminq->cq_vector = -1; + hdev->online_queues--; + goto free_queue; + } + + dev_info(hdev->dev, "[%s] success, queuecount:[%d], onlinequeue:[%d]\n", + __func__, hdev->queue_count, hdev->online_queues); + + return 0; + +free_queue: + spraid_free_queue(adminq); + return ret; +} + +static u32 spraid_bar_size(struct spraid_dev *hdev, u32 nr_ioqs) +{ + return (SPRAID_REG_DBS + ((nr_ioqs + 1) * 8 * hdev->db_stride)); +} + +static int spraid_alloc_admin_cmds(struct spraid_dev *hdev) +{ + int i; + + INIT_LIST_HEAD(&hdev->adm_cmd_list); + spin_lock_init(&hdev->adm_cmd_lock); + + hdev->adm_cmds = kcalloc_node(SPRAID_AQ_BLK_MQ_DEPTH, + sizeof(struct spraid_cmd), + GFP_KERNEL, hdev->numa_node); + + if (!hdev->adm_cmds) { + dev_err(hdev->dev, "Alloc admin cmds failed\n"); + return -ENOMEM; + } + + for (i = 0; i < SPRAID_AQ_BLK_MQ_DEPTH; i++) { + hdev->adm_cmds[i].qid = 0; + hdev->adm_cmds[i].cid = i; + list_add_tail(&(hdev->adm_cmds[i].list), &hdev->adm_cmd_list); + } + + dev_info(hdev->dev, "Alloc admin cmds success, num[%d]\n", + SPRAID_AQ_BLK_MQ_DEPTH); + + return 0; +} + +static void spraid_free_admin_cmds(struct spraid_dev *hdev) +{ + kfree(hdev->adm_cmds); + hdev->adm_cmds = NULL; + INIT_LIST_HEAD(&hdev->adm_cmd_list); +} + +static struct spraid_cmd *spraid_get_cmd(struct spraid_dev *hdev, + enum spraid_cmd_type type) +{ + struct spraid_cmd *cmd = NULL; + unsigned long flags; + struct list_head *head = &hdev->adm_cmd_list; + spinlock_t *slock = &hdev->adm_cmd_lock; + + if (type == SPRAID_CMD_IOPT) { + head = &hdev->ioq_pt_list; + slock = &hdev->ioq_pt_lock; + } + + spin_lock_irqsave(slock, flags); + if (list_empty(head)) { + spin_unlock_irqrestore(slock, flags); + dev_err(hdev->dev, "err, cmd[%d] list empty\n", type); + return NULL; + } + cmd = list_entry(head->next, struct spraid_cmd, list); + list_del_init(&cmd->list); + spin_unlock_irqrestore(slock, flags); + + WRITE_ONCE(cmd->state, SPRAID_CMD_IN_FLIGHT); + + return cmd; +} + +static void spraid_put_cmd(struct spraid_dev *hdev, struct spraid_cmd *cmd, + enum spraid_cmd_type type) +{ + unsigned long flags; + struct list_head *head = &hdev->adm_cmd_list; + spinlock_t *slock = &hdev->adm_cmd_lock; + + if (type == SPRAID_CMD_IOPT) { + head = &hdev->ioq_pt_list; + slock = &hdev->ioq_pt_lock; + } + + spin_lock_irqsave(slock, flags); + WRITE_ONCE(cmd->state, SPRAID_CMD_IDLE); + list_add_tail(&cmd->list, head); + spin_unlock_irqrestore(slock, flags); +} + + +static int spraid_submit_admin_sync_cmd(struct spraid_dev *hdev, + struct spraid_admin_command *cmd, + u32 *result0, u32 *result1, u32 timeout) +{ + struct spraid_cmd *adm_cmd = spraid_get_cmd(hdev, SPRAID_CMD_ADM); + + if (!adm_cmd) { + dev_err(hdev->dev, "err, get admin cmd failed\n"); + return -EFAULT; + } + + timeout = timeout ? timeout : ADMIN_TIMEOUT; + + init_completion(&adm_cmd->cmd_done); + + cmd->common.command_id = adm_cmd->cid; + spraid_submit_cmd(&hdev->queues[0], cmd); + + if (!wait_for_completion_timeout(&adm_cmd->cmd_done, timeout)) { + dev_err(hdev->dev, "[%s] cid[%d] qid[%d] timeout;" + " opcode[0x%x] subopcode[0x%x]\n", + __func__, adm_cmd->cid, adm_cmd->qid, + cmd->usr_cmd.opcode, cmd->usr_cmd.info_0.subopcode); + WRITE_ONCE(adm_cmd->state, SPRAID_CMD_TIMEOUT); + spraid_put_cmd(hdev, adm_cmd, SPRAID_CMD_ADM); + return -EINVAL; + } + + if (result0) + *result0 = adm_cmd->result0; + if (result1) + *result1 = adm_cmd->result1; + + spraid_put_cmd(hdev, adm_cmd, SPRAID_CMD_ADM); + + return adm_cmd->status; +} + +static int spraid_create_cq(struct spraid_dev *hdev, u16 qid, + struct spraid_queue *spraidq, u16 cq_vector) +{ + struct spraid_admin_command admin_cmd; + int flags = SPRAID_QUEUE_PHYS_CONTIG | SPRAID_CQ_IRQ_ENABLED; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.create_cq.opcode = SPRAID_ADMIN_CREATE_CQ; + admin_cmd.create_cq.prp1 = cpu_to_le64(spraidq->cq_dma_addr); + admin_cmd.create_cq.cqid = cpu_to_le16(qid); + admin_cmd.create_cq.qsize = cpu_to_le16(spraidq->q_depth - 1); + admin_cmd.create_cq.cq_flags = cpu_to_le16(flags); + admin_cmd.create_cq.irq_vector = cpu_to_le16(cq_vector); + + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); +} + +static int spraid_create_sq(struct spraid_dev *hdev, u16 qid, + struct spraid_queue *spraidq) +{ + struct spraid_admin_command admin_cmd; + int flags = SPRAID_QUEUE_PHYS_CONTIG; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.create_sq.opcode = SPRAID_ADMIN_CREATE_SQ; + admin_cmd.create_sq.prp1 = cpu_to_le64(spraidq->sq_dma_addr); + admin_cmd.create_sq.sqid = cpu_to_le16(qid); + admin_cmd.create_sq.qsize = cpu_to_le16(spraidq->q_depth - 1); + admin_cmd.create_sq.sq_flags = cpu_to_le16(flags); + admin_cmd.create_sq.cqid = cpu_to_le16(qid); + + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); +} + +static void spraid_free_queue(struct spraid_queue *spraidq) +{ + struct spraid_dev *hdev = spraidq->hdev; + + hdev->queue_count--; + dma_free_coherent(hdev->dev, CQ_SIZE(spraidq->q_depth), + (void *)spraidq->cqes, spraidq->cq_dma_addr); + dma_free_coherent(hdev->dev, SQ_SIZE(spraidq->qid, spraidq->q_depth), + spraidq->sq_cmds, spraidq->sq_dma_addr); + dma_free_coherent(hdev->dev, SENSE_SIZE(spraidq->q_depth), + spraidq->sense, spraidq->sense_dma_addr); +} + +static void spraid_free_admin_queue(struct spraid_dev *hdev) +{ + spraid_free_queue(&hdev->queues[0]); +} + +static void spraid_free_io_queues(struct spraid_dev *hdev) +{ + int i; + + for (i = hdev->queue_count - 1; i >= 1; i--) + spraid_free_queue(&hdev->queues[i]); +} + +static int spraid_delete_queue(struct spraid_dev *hdev, u8 op, u16 id) +{ + struct spraid_admin_command admin_cmd; + int ret; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.delete_queue.opcode = op; + admin_cmd.delete_queue.qid = cpu_to_le16(id); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + + if (ret) + dev_err(hdev->dev, "Delete %s:[%d] failed\n", + (op == SPRAID_ADMIN_DELETE_CQ) ? "cq" : "sq", id); + + return ret; +} + +static int spraid_delete_cq(struct spraid_dev *hdev, u16 cqid) +{ + return spraid_delete_queue(hdev, SPRAID_ADMIN_DELETE_CQ, cqid); +} + +static int spraid_delete_sq(struct spraid_dev *hdev, u16 sqid) +{ + return spraid_delete_queue(hdev, SPRAID_ADMIN_DELETE_SQ, sqid); +} + +static int spraid_create_queue(struct spraid_queue *spraidq, u16 qid) +{ + struct spraid_dev *hdev = spraidq->hdev; + u16 cq_vector; + int ret; + + cq_vector = (hdev->num_vecs == 1) ? 0 : qid; + ret = spraid_create_cq(hdev, qid, spraidq, cq_vector); + if (ret) + return ret; + + ret = spraid_create_sq(hdev, qid, spraidq); + if (ret) + goto delete_cq; + + spraid_init_queue(spraidq, qid); + spraidq->cq_vector = cq_vector; + + ret = pci_request_irq(hdev->pdev, cq_vector, spraid_irq, NULL, + spraidq, "spraid%d_q%d", hdev->instance, qid); + + if (ret) { + dev_err(hdev->dev, "Request queue[%d] irq failed\n", qid); + goto delete_sq; + } + + return 0; + +delete_sq: + spraidq->cq_vector = -1; + hdev->online_queues--; + spraid_delete_sq(hdev, qid); +delete_cq: + spraid_delete_cq(hdev, qid); + + return ret; +} + +static int spraid_create_io_queues(struct spraid_dev *hdev) +{ + u32 i, max; + int ret = 0; + + max = min(hdev->max_qid, hdev->queue_count - 1); + for (i = hdev->online_queues; i <= max; i++) { + ret = spraid_create_queue(&hdev->queues[i], i); + if (ret) { + dev_err(hdev->dev, "Create queue[%d] failed\n", i); + break; + } + } + + dev_info(hdev->dev, "[%s] queue_count[%d], online_queue[%d]", + __func__, hdev->queue_count, hdev->online_queues); + + return ret >= 0 ? 0 : ret; +} + +static int spraid_set_features(struct spraid_dev *hdev, u32 fid, + u32 dword11, void *buffer, + size_t buflen, u32 *result) +{ + struct spraid_admin_command admin_cmd; + int ret; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + + if (buffer && buflen) { + data_ptr = dma_alloc_coherent(hdev->dev, buflen, + &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memcpy(data_ptr, buffer, buflen); + } + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.features.opcode = SPRAID_ADMIN_SET_FEATURES; + admin_cmd.features.fid = cpu_to_le32(fid); + admin_cmd.features.dword11 = cpu_to_le32(dword11); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, result, NULL, 0); + + if (data_ptr) + dma_free_coherent(hdev->dev, buflen, data_ptr, data_dma); + + return ret; +} + +static int spraid_configure_timestamp(struct spraid_dev *hdev) +{ + __le64 ts; + int ret; + + ts = cpu_to_le64(ktime_to_ms(ktime_get_real())); + ret = spraid_set_features(hdev, SPRAID_FEAT_TIMESTAMP, + 0, &ts, sizeof(ts), NULL); + + if (ret) + dev_err(hdev->dev, "set timestamp failed: %d\n", ret); + return ret; +} + +static int spraid_set_queue_cnt(struct spraid_dev *hdev, u32 *cnt) +{ + u32 q_cnt = (*cnt - 1) | ((*cnt - 1) << 16); + u32 nr_ioqs, result; + int status; + + status = spraid_set_features(hdev, SPRAID_FEAT_NUM_QUEUES, + q_cnt, NULL, 0, &result); + if (status) { + dev_err(hdev->dev, "Set queue count failed, status: %d\n", + status); + return -EIO; + } + + nr_ioqs = min(result & 0xffff, result >> 16) + 1; + *cnt = min(*cnt, nr_ioqs); + if (*cnt == 0) { + dev_err(hdev->dev, "Illegal queue count: zero\n"); + return -EIO; + } + return 0; +} + +static int spraid_setup_io_queues(struct spraid_dev *hdev) +{ + struct spraid_queue *adminq = &hdev->queues[0]; + struct pci_dev *pdev = hdev->pdev; + u32 nr_ioqs = num_online_cpus(); + u32 i, size; + int ret; + + struct irq_affinity affd = { + .pre_vectors = 1 + }; + + ret = spraid_set_queue_cnt(hdev, &nr_ioqs); + if (ret < 0) + return ret; + + size = spraid_bar_size(hdev, nr_ioqs); + ret = spraid_remap_bar(hdev, size); + if (ret) + return -ENOMEM; + + adminq->q_db = hdev->dbs; + + pci_free_irq(pdev, 0, adminq); + pci_free_irq_vectors(pdev); + + ret = pci_alloc_irq_vectors_affinity(pdev, 1, (nr_ioqs + 1), + PCI_IRQ_ALL_TYPES | PCI_IRQ_AFFINITY, &affd); + if (ret <= 0) + return -EIO; + + hdev->num_vecs = ret; + + hdev->max_qid = max(ret - 1, 1); + + ret = pci_request_irq(pdev, adminq->cq_vector, spraid_irq, NULL, + adminq, "spraid%d_q%d", + hdev->instance, adminq->qid); + if (ret) { + dev_err(hdev->dev, "Request admin irq failed\n"); + adminq->cq_vector = -1; + return ret; + } + + for (i = hdev->queue_count; i <= hdev->max_qid; i++) { + ret = spraid_alloc_queue(hdev, i, hdev->ioq_depth); + if (ret) + break; + } + dev_info(hdev->dev, "[%s] max_qid: %d, queue_count: %d;" + " online_queue: %d, ioq_depth: %d\n", + __func__, hdev->max_qid, hdev->queue_count, + hdev->online_queues, hdev->ioq_depth); + + return spraid_create_io_queues(hdev); +} + +static void spraid_delete_io_queues(struct spraid_dev *hdev) +{ + u16 queues = hdev->online_queues - 1; + u8 opcode = SPRAID_ADMIN_DELETE_SQ; + u16 i, pass; + + if (!pci_device_is_present(hdev->pdev)) { + dev_err(hdev->dev, + "pci_device is not present, skip disable io queues\n"); + return; + } + + if (hdev->online_queues < 2) { + dev_err(hdev->dev, "[%s] err, io queue has been delete\n", + __func__); + return; + } + + for (pass = 0; pass < 2; pass++) { + for (i = queues; i > 0; i--) + if (spraid_delete_queue(hdev, opcode, i)) + break; + + opcode = SPRAID_ADMIN_DELETE_CQ; + } +} + +static void spraid_remove_io_queues(struct spraid_dev *hdev) +{ + spraid_delete_io_queues(hdev); + spraid_free_io_queues(hdev); +} + +static void spraid_pci_disable(struct spraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + u32 i; + + for (i = 0; i < hdev->online_queues; i++) + pci_free_irq(pdev, hdev->queues[i].cq_vector, &hdev->queues[i]); + pci_free_irq_vectors(pdev); + if (pci_is_enabled(pdev)) { + pci_disable_pcie_error_reporting(pdev); + pci_disable_device(pdev); + } + hdev->online_queues = 0; +} + +static void spraid_disable_admin_queue(struct spraid_dev *hdev, bool shutdown) +{ + struct spraid_queue *adminq = &hdev->queues[0]; + u16 start, end; + + if (pci_device_is_present(hdev->pdev)) { + if (shutdown) + spraid_shutdown_ctrl(hdev); + else + spraid_disable_ctrl(hdev); + } + + if (hdev->queue_count == 0) { + dev_err(hdev->dev, "[%s] err, admin queue has been delete\n", + __func__); + return; + } + + spin_lock_irq(&adminq->cq_lock); + spraid_process_cq(adminq, &start, &end, -1); + spin_unlock_irq(&adminq->cq_lock); + + spraid_complete_cqes(adminq, start, end); + spraid_free_admin_queue(hdev); +} + +static int spraid_create_dma_pools(struct spraid_dev *hdev) +{ + int i; + char poolname[20] = { 0 }; + + hdev->prp_page_pool = dma_pool_create("prp list page", hdev->dev, + PAGE_SIZE, PAGE_SIZE, 0); + + if (!hdev->prp_page_pool) { + dev_err(hdev->dev, "create prp_page_pool failed\n"); + return -ENOMEM; + } + + for (i = 0; i < small_pool_num; i++) { + sprintf(poolname, "prp_list_256_%d", i); + hdev->prp_small_pool[i] = + dma_pool_create(poolname, hdev->dev, SMALL_POOL_SIZE, + SMALL_POOL_SIZE, 0); + + if (!hdev->prp_small_pool[i]) { + dev_err(hdev->dev, "create prp_small_pool %d failed\n", + i); + goto destroy_prp_small_pool; + } + } + + return 0; + +destroy_prp_small_pool: + while (i > 0) + dma_pool_destroy(hdev->prp_small_pool[--i]); + dma_pool_destroy(hdev->prp_page_pool); + + return -ENOMEM; +} + +static void spraid_destroy_dma_pools(struct spraid_dev *hdev) +{ + int i; + + for (i = 0; i < small_pool_num; i++) + dma_pool_destroy(hdev->prp_small_pool[i]); + dma_pool_destroy(hdev->prp_page_pool); +} + +static int spraid_get_dev_list(struct spraid_dev *hdev, + struct spraid_dev_info *devices) +{ + u32 nd = le32_to_cpu(hdev->ctrl_info->nd); + struct spraid_admin_command admin_cmd; + struct spraid_dev_list *list_buf; + dma_addr_t data_dma = 0; + u32 i, idx, hdid, ndev; + int ret = 0; + + list_buf = dma_alloc_coherent(hdev->dev, PAGE_SIZE, + &data_dma, GFP_KERNEL); + if (!list_buf) + return -ENOMEM; + + for (idx = 0; idx < nd;) { + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.get_info.opcode = SPRAID_ADMIN_GET_INFO; + admin_cmd.get_info.type = SPRAID_GET_INFO_DEV_LIST; + admin_cmd.get_info.cdw11 = cpu_to_le32(idx); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, + NULL, NULL, 0); + + if (ret) { + dev_err(hdev->dev, "Get device list failed, nd: %u;" + "idx: %u, ret: %d\n", + nd, idx, ret); + goto out; + } + ndev = le32_to_cpu(list_buf->dev_num); + + dev_info(hdev->dev, "ndev numbers: %u\n", ndev); + + for (i = 0; i < ndev; i++) { + hdid = le32_to_cpu(list_buf->devices[i].hdid); + dev_info(hdev->dev, "list_buf->devices[%d], hdid: %u;" + "target: %d, channel: %d, lun: %d, attr[0x%x]\n", + i, hdid, + le16_to_cpu(list_buf->devices[i].target), + list_buf->devices[i].channel, + list_buf->devices[i].lun, + list_buf->devices[i].attr); + if (hdid > nd || hdid == 0) { + dev_err(hdev->dev, "err, hdid[%d] invalid\n", + hdid); + continue; + } + memcpy(&devices[hdid - 1], &list_buf->devices[i], + sizeof(struct spraid_dev_info)); + } + idx += ndev; + + if (idx < MAX_DEV_ENTRY_PER_PAGE_4K) + break; + } + +out: + dma_free_coherent(hdev->dev, PAGE_SIZE, list_buf, data_dma); + return ret; +} + +static void spraid_send_aen(struct spraid_dev *hdev, u16 cid) +{ + struct spraid_queue *adminq = &hdev->queues[0]; + struct spraid_admin_command admin_cmd; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.common.opcode = SPRAID_ADMIN_ASYNC_EVENT; + admin_cmd.common.command_id = cid; + + spraid_submit_cmd(adminq, &admin_cmd); + dev_info(hdev->dev, "send aen, cid[%d]\n", cid); +} + +static inline void spraid_send_all_aen(struct spraid_dev *hdev) +{ + u16 i; + + for (i = 0; i < hdev->ctrl_info->aerl; i++) + spraid_send_aen(hdev, i + SPRAID_AQ_BLK_MQ_DEPTH); +} + +static int spraid_add_device(struct spraid_dev *hdev, + struct spraid_dev_info *device) +{ + struct Scsi_Host *shost = hdev->shost; + struct scsi_device *sdev; + + dev_info(hdev->dev, "add device, hdid: %u target: %d, channel: %d;" + " lun: %d, attr[0x%x]\n", + le32_to_cpu(device->hdid), le16_to_cpu(device->target), + device->channel, device->lun, device->attr); + + sdev = scsi_device_lookup(shost, device->channel, + le16_to_cpu(device->target), 0); + if (sdev) { + dev_warn(hdev->dev, "Device is already exist, channel: %d;" + " target_id: %d, lun: %d\n", + device->channel, le16_to_cpu(device->target), 0); + scsi_device_put(sdev); + return -EEXIST; + } + scsi_add_device(shost, device->channel, le16_to_cpu(device->target), 0); + return 0; +} + +static int spraid_rescan_device(struct spraid_dev *hdev, + struct spraid_dev_info *device) +{ + struct Scsi_Host *shost = hdev->shost; + struct scsi_device *sdev; + + dev_info(hdev->dev, "rescan device, hdid: %u target: %d, channel: %d;" + " lun: %d, attr[0x%x]\n", + le32_to_cpu(device->hdid), le16_to_cpu(device->target), + device->channel, device->lun, device->attr); + + sdev = scsi_device_lookup(shost, device->channel, + le16_to_cpu(device->target), 0); + if (!sdev) { + dev_warn(hdev->dev, "device is not exit rescan it, channel: %d;" + " target_id: %d, lun: %d\n", + device->channel, le16_to_cpu(device->target), 0); + return -ENODEV; + } + + scsi_rescan_device(&sdev->sdev_gendev); + scsi_device_put(sdev); + return 0; +} + +static int spraid_remove_device(struct spraid_dev *hdev, + struct spraid_dev_info *org_device) +{ + struct Scsi_Host *shost = hdev->shost; + struct scsi_device *sdev; + + dev_info(hdev->dev, "remove device, hdid: %u target: %d, channel: %d;" + " lun: %d, attr[0x%x]\n", + le32_to_cpu(org_device->hdid), le16_to_cpu(org_device->target), + org_device->channel, org_device->lun, org_device->attr); + + sdev = scsi_device_lookup(shost, org_device->channel, + le16_to_cpu(org_device->target), 0); + if (!sdev) { + dev_warn(hdev->dev, "device is not exit remove it, channel: %d;" + " target_id: %d, lun: %d\n", + org_device->channel, + le16_to_cpu(org_device->target), 0); + return -ENODEV; + } + + scsi_remove_device(sdev); + scsi_device_put(sdev); + return 0; +} + +static int spraid_dev_list_init(struct spraid_dev *hdev) +{ + u32 nd = le32_to_cpu(hdev->ctrl_info->nd); + int i, ret; + + hdev->devices = kzalloc_node(nd * sizeof(struct spraid_dev_info), + GFP_KERNEL, hdev->numa_node); + if (!hdev->devices) + return -ENOMEM; + + ret = spraid_get_dev_list(hdev, hdev->devices); + if (ret) { + dev_err(hdev->dev, + "Ignore failure of getting device list;" + " within initialization\n"); + return 0; + } + + for (i = 0; i < nd; i++) { + if (SPRAID_DEV_INFO_FLAG_VALID(hdev->devices[i].flag) && + SPRAID_DEV_INFO_ATTR_BOOT(hdev->devices[i].attr)) { + spraid_add_device(hdev, &hdev->devices[i]); + break; + } + } + return 0; +} + +static int luntarget_cmp_func(const void *l, const void *r) +{ + const struct spraid_dev_info *ln = l; + const struct spraid_dev_info *rn = r; + + if (ln->channel == rn->channel) + return le16_to_cpu(ln->target) - le16_to_cpu(rn->target); + + return ln->channel - rn->channel; +} + +static void spraid_scan_work(struct work_struct *work) +{ + struct spraid_dev *hdev = + container_of(work, struct spraid_dev, scan_work); + struct spraid_dev_info *devices, *org_devices; + struct spraid_dev_info *sortdevice; + u32 nd = le32_to_cpu(hdev->ctrl_info->nd); + u8 flag, org_flag; + int i, ret; + int count = 0; + + devices = kcalloc(nd, sizeof(struct spraid_dev_info), GFP_KERNEL); + if (!devices) + return; + + sortdevice = kcalloc(nd, sizeof(struct spraid_dev_info), GFP_KERNEL); + if (!sortdevice) + goto free_list; + + ret = spraid_get_dev_list(hdev, devices); + if (ret) + goto free_all; + org_devices = hdev->devices; + for (i = 0; i < nd; i++) { + org_flag = org_devices[i].flag; + flag = devices[i].flag; + + dev_log_dbg(hdev->dev, "i: %d, org_flag: 0x%x, flag: 0x%x\n", + i, org_flag, flag); + + if (SPRAID_DEV_INFO_FLAG_VALID(flag)) { + if (!SPRAID_DEV_INFO_FLAG_VALID(org_flag)) { + down_write(&hdev->devices_rwsem); + memcpy(&org_devices[i], &devices[i], + sizeof(struct spraid_dev_info)); + memcpy(&sortdevice[count++], &devices[i], + sizeof(struct spraid_dev_info)); + up_write(&hdev->devices_rwsem); + } else if (SPRAID_DEV_INFO_FLAG_CHANGE(flag)) { + spraid_rescan_device(hdev, &devices[i]); + } + } else { + if (SPRAID_DEV_INFO_FLAG_VALID(org_flag)) { + down_write(&hdev->devices_rwsem); + org_devices[i].flag &= 0xfe; + up_write(&hdev->devices_rwsem); + spraid_remove_device(hdev, &org_devices[i]); + } + } + } + + dev_info(hdev->dev, "scan work add device count = %d\n", count); + + sort(sortdevice, count, sizeof(sortdevice[0]), + luntarget_cmp_func, NULL); + + for (i = 0; i < count; i++) + spraid_add_device(hdev, &sortdevice[i]); + +free_all: + kfree(sortdevice); +free_list: + kfree(devices); +} + +static void spraid_timesyn_work(struct work_struct *work) +{ + struct spraid_dev *hdev = + container_of(work, struct spraid_dev, timesyn_work); + + spraid_configure_timestamp(hdev); +} + +static int spraid_init_ctrl_info(struct spraid_dev *hdev); +static void spraid_fw_act_work(struct work_struct *work) +{ + struct spraid_dev *hdev = + container_of(work, struct spraid_dev, fw_act_work); + + if (spraid_init_ctrl_info(hdev)) + dev_err(hdev->dev, "get ctrl info failed after fw act\n"); +} + +static void spraid_queue_scan(struct spraid_dev *hdev) +{ + queue_work(spraid_wq, &hdev->scan_work); +} + +static void spraid_handle_aen_notice(struct spraid_dev *hdev, u32 result) +{ + switch ((result & 0xff00) >> 8) { + case SPRAID_AEN_DEV_CHANGED: + spraid_queue_scan(hdev); + break; + case SPRAID_AEN_FW_ACT_START: + dev_info(hdev->dev, "fw activation starting\n"); + break; + case SPRAID_AEN_HOST_PROBING: + break; + default: + dev_warn(hdev->dev, "async event result %08x\n", result); + } +} + +static void spraid_handle_aen_vs(struct spraid_dev *hdev, + u32 result, u32 result1) +{ + switch ((result & 0xff00) >> 8) { + case SPRAID_AEN_TIMESYN: + queue_work(spraid_wq, &hdev->timesyn_work); + break; + case SPRAID_AEN_FW_ACT_FINISH: + dev_info(hdev->dev, "fw activation finish\n"); + queue_work(spraid_wq, &hdev->fw_act_work); + break; + case SPRAID_AEN_EVENT_MIN ... SPRAID_AEN_EVENT_MAX: + dev_info(hdev->dev, "rcv card event[%d];" + " param1[0x%x] param2[0x%x]\n", + (result & 0xff00) >> 8, result, result1); + break; + default: + dev_warn(hdev->dev, "async event result: 0x%x\n", result); + } +} + +static int spraid_alloc_resources(struct spraid_dev *hdev) +{ + int ret, nqueue; + + ret = ida_alloc(&spraid_instance_ida, GFP_KERNEL); + if (ret < 0) { + dev_err(hdev->dev, "Get instance id failed\n"); + return ret; + } + hdev->instance = ret; + + hdev->ctrl_info = kzalloc_node(sizeof(*hdev->ctrl_info), + GFP_KERNEL, hdev->numa_node); + if (!hdev->ctrl_info) { + ret = -ENOMEM; + goto release_instance; + } + + ret = spraid_create_dma_pools(hdev); + if (ret) + goto free_ctrl_info; + nqueue = num_possible_cpus() + 1; + hdev->queues = kcalloc_node(nqueue, sizeof(struct spraid_queue), + GFP_KERNEL, hdev->numa_node); + if (!hdev->queues) { + ret = -ENOMEM; + goto destroy_dma_pools; + } + + ret = spraid_alloc_admin_cmds(hdev); + if (ret) + goto free_queues; + + dev_info(hdev->dev, "[%s] queues num: %d\n", __func__, nqueue); + + return 0; + +free_queues: + kfree(hdev->queues); +destroy_dma_pools: + spraid_destroy_dma_pools(hdev); +free_ctrl_info: + kfree(hdev->ctrl_info); +release_instance: + ida_free(&spraid_instance_ida, hdev->instance); + return ret; +} + +static void spraid_free_resources(struct spraid_dev *hdev) +{ + spraid_free_admin_cmds(hdev); + kfree(hdev->queues); + spraid_destroy_dma_pools(hdev); + kfree(hdev->ctrl_info); + ida_free(&spraid_instance_ida, hdev->instance); +} + +static void spraid_bsg_unmap_data(struct spraid_dev *hdev, struct bsg_job *job) +{ + struct request *rq = blk_mq_rq_from_pdu(job); + struct spraid_iod *iod = job->dd_data; + enum dma_data_direction dma_dir = + rq_data_dir(rq) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; + + if (iod->nsge) + dma_unmap_sg(hdev->dev, iod->sg, iod->nsge, dma_dir); + + spraid_free_iod_res(hdev, iod); +} + +static int spraid_bsg_map_data(struct spraid_dev *hdev, struct bsg_job *job, + struct spraid_admin_command *cmd) +{ + struct request *rq = blk_mq_rq_from_pdu(job); + struct spraid_iod *iod = job->dd_data; + enum dma_data_direction dma_dir = + rq_data_dir(rq) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; + int ret = 0; + + iod->sg = job->request_payload.sg_list; + iod->nsge = job->request_payload.sg_cnt; + iod->length = job->request_payload.payload_len; + iod->use_sgl = false; + iod->npages = -1; + iod->sg_drv_mgmt = false; + + if (!iod->nsge) + goto out; + + ret = dma_map_sg_attrs(hdev->dev, iod->sg, iod->nsge, + dma_dir, DMA_ATTR_NO_WARN); + if (!ret) + goto out; + + ret = spraid_setup_prps(hdev, iod); + if (ret) + goto unmap; + + cmd->common.dptr.prp1 = cpu_to_le64(sg_dma_address(iod->sg)); + cmd->common.dptr.prp2 = cpu_to_le64(iod->first_dma); + + return 0; + +unmap: + dma_unmap_sg(hdev->dev, iod->sg, iod->nsge, dma_dir); +out: + return ret; +} + +static int spraid_get_ctrl_info(struct spraid_dev *hdev, + struct spraid_ctrl_info *ctrl_info) +{ + struct spraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + int ret; + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, + &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.get_info.opcode = SPRAID_ADMIN_GET_INFO; + admin_cmd.get_info.type = SPRAID_GET_INFO_CTRL; + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + if (!ret) + memcpy(ctrl_info, data_ptr, sizeof(struct spraid_ctrl_info)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, data_dma); + + return ret; +} + +static int spraid_init_ctrl_info(struct spraid_dev *hdev) +{ + int ret; + + hdev->ctrl_info->nd = cpu_to_le32(240); + hdev->ctrl_info->mdts = 8; + hdev->ctrl_info->max_cmds = cpu_to_le16(4096); + hdev->ctrl_info->max_num_sge = cpu_to_le16(128); + hdev->ctrl_info->max_channel = cpu_to_le16(4); + hdev->ctrl_info->max_tgt_id = cpu_to_le32(3239); + hdev->ctrl_info->max_lun = cpu_to_le16(2); + + ret = spraid_get_ctrl_info(hdev, hdev->ctrl_info); + if (ret) + dev_err(hdev->dev, "get controller info failed: %d\n", ret); + + dev_info(hdev->dev, "[%s]nd = %d\n", __func__, hdev->ctrl_info->nd); + dev_info(hdev->dev, "[%s]max_cmd = %d\n", + __func__, hdev->ctrl_info->max_cmds); + dev_info(hdev->dev, "[%s]max_channel = %d\n", + __func__, hdev->ctrl_info->max_channel); + dev_info(hdev->dev, "[%s]max_tgt_id = %d\n", + __func__, hdev->ctrl_info->max_tgt_id); + dev_info(hdev->dev, "[%s]max_lun = %d\n", + __func__, hdev->ctrl_info->max_lun); + dev_info(hdev->dev, "[%s]max_num_sge = %d\n", + __func__, hdev->ctrl_info->max_num_sge); + dev_info(hdev->dev, "[%s]lun_num_boot = %d\n", + __func__, hdev->ctrl_info->lun_num_in_boot); + dev_info(hdev->dev, "[%s]mdts = %d\n", __func__, hdev->ctrl_info->mdts); + dev_info(hdev->dev, "[%s]acl = %d\n", __func__, hdev->ctrl_info->acl); + dev_info(hdev->dev, "[%s]aer1 = %d\n", __func__, hdev->ctrl_info->aerl); + dev_info(hdev->dev, "[%s]card_type = %d\n", + __func__, hdev->ctrl_info->card_type); + dev_info(hdev->dev, "[%s]rtd3e = %d\n", + __func__, hdev->ctrl_info->rtd3e); + dev_info(hdev->dev, "[%s]sn = %s\n", __func__, hdev->ctrl_info->sn); + dev_info(hdev->dev, "[%s]fr = %s\n", __func__, hdev->ctrl_info->fr); + + if (!hdev->ctrl_info->aerl) + hdev->ctrl_info->aerl = 1; + if (hdev->ctrl_info->aerl > SPRAID_NR_AEN_COMMANDS) + hdev->ctrl_info->aerl = SPRAID_NR_AEN_COMMANDS; + + return 0; +} + +#define SPRAID_MAX_ADMIN_PAYLOAD_SIZE BIT(16) +static int spraid_alloc_iod_ext_mem_pool(struct spraid_dev *hdev) +{ + u16 max_sge = le16_to_cpu(hdev->ctrl_info->max_num_sge); + size_t alloc_size; + + alloc_size = spraid_iod_ext_size(hdev, SPRAID_MAX_ADMIN_PAYLOAD_SIZE, + max_sge, true, false); + if (alloc_size > PAGE_SIZE) + dev_warn(hdev->dev, "It is unreasonable ;" + " sg allocation more than one page\n"); + hdev->iod_mempool = mempool_create_node(1, mempool_kmalloc, + mempool_kfree, + (void *)alloc_size, GFP_KERNEL, + hdev->numa_node); + if (!hdev->iod_mempool) { + dev_err(hdev->dev, "Create iod extension memory pool failed\n"); + return -ENOMEM; + } + + return 0; +} + +static void spraid_free_iod_ext_mem_pool(struct spraid_dev *hdev) +{ + mempool_destroy(hdev->iod_mempool); +} + +static int spraid_user_admin_cmd(struct spraid_dev *hdev, struct bsg_job *job) +{ + struct spraid_bsg_request *bsg_req = job->request; + struct spraid_passthru_common_cmd *cmd = &(bsg_req->admcmd); + struct spraid_admin_command admin_cmd; + u32 timeout = msecs_to_jiffies(cmd->timeout_ms); + u32 result[2] = {0}; + int status; + + if (hdev->state >= SPRAID_RESETTING) { + dev_err(hdev->dev, "[%s] err, host state:[%d] is not right\n", + __func__, hdev->state); + return -EBUSY; + } + + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x] init\n", + __func__, cmd->opcode, cmd->info_0.subopcode); + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.common.opcode = cmd->opcode; + admin_cmd.common.flags = cmd->flags; + admin_cmd.common.hdid = cpu_to_le32(cmd->nsid); + admin_cmd.common.cdw2[0] = cpu_to_le32(cmd->cdw2); + admin_cmd.common.cdw2[1] = cpu_to_le32(cmd->cdw3); + admin_cmd.common.cdw10 = cpu_to_le32(cmd->cdw10); + admin_cmd.common.cdw11 = cpu_to_le32(cmd->cdw11); + admin_cmd.common.cdw12 = cpu_to_le32(cmd->cdw12); + admin_cmd.common.cdw13 = cpu_to_le32(cmd->cdw13); + admin_cmd.common.cdw14 = cpu_to_le32(cmd->cdw14); + admin_cmd.common.cdw15 = cpu_to_le32(cmd->cdw15); + + status = spraid_bsg_map_data(hdev, job, &admin_cmd); + if (status) { + dev_err(hdev->dev, "[%s] err, map data failed\n", __func__); + return status; + } + + status = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, &result[0], + &result[1], timeout); + if (status >= 0) { + job->reply_len = sizeof(result); + memcpy(job->reply, result, sizeof(result)); + } + + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x];" + " status[0x%x] result0[0x%x] result1[0x%x]\n", + __func__, cmd->opcode, cmd->info_0.subopcode, + status, result[0], result[1]); + + spraid_bsg_unmap_data(hdev, job); + + return status; +} + +static int spraid_alloc_ioq_ptcmds(struct spraid_dev *hdev) +{ + int i; + int ptnum = SPRAID_NR_IOQ_PTCMDS; + + INIT_LIST_HEAD(&hdev->ioq_pt_list); + spin_lock_init(&hdev->ioq_pt_lock); + + hdev->ioq_ptcmds = kcalloc_node(ptnum, sizeof(struct spraid_cmd), + GFP_KERNEL, hdev->numa_node); + + if (!hdev->ioq_ptcmds) { + dev_err(hdev->dev, "Alloc ioq_ptcmds failed\n"); + return -ENOMEM; + } + + for (i = 0; i < ptnum; i++) { + hdev->ioq_ptcmds[i].qid = i / SPRAID_PTCMDS_PERQ + 1; + hdev->ioq_ptcmds[i].cid = i % SPRAID_PTCMDS_PERQ + + SPRAID_IO_BLK_MQ_DEPTH; + list_add_tail(&(hdev->ioq_ptcmds[i].list), &hdev->ioq_pt_list); + } + + dev_info(hdev->dev, "Alloc ioq_ptcmds success, ptnum[%d]\n", ptnum); + + return 0; +} + +static void spraid_free_ioq_ptcmds(struct spraid_dev *hdev) +{ + kfree(hdev->ioq_ptcmds); + hdev->ioq_ptcmds = NULL; + + INIT_LIST_HEAD(&hdev->ioq_pt_list); +} + +static int spraid_submit_ioq_sync_cmd(struct spraid_dev *hdev, + struct spraid_ioq_command *cmd, + u32 *result, u32 *reslen, u32 timeout) +{ + int ret; + dma_addr_t sense_dma; + struct spraid_queue *ioq; + void *sense_addr = NULL; + struct spraid_cmd *pt_cmd = spraid_get_cmd(hdev, SPRAID_CMD_IOPT); + + if (!pt_cmd) { + dev_err(hdev->dev, "err, get ioq cmd failed\n"); + return -EFAULT; + } + + timeout = timeout ? timeout : ADMIN_TIMEOUT; + + init_completion(&pt_cmd->cmd_done); + + ioq = &hdev->queues[pt_cmd->qid]; + ret = pt_cmd->cid * SCSI_SENSE_BUFFERSIZE; + sense_addr = ioq->sense + ret; + sense_dma = ioq->sense_dma_addr + ret; + + cmd->common.sense_addr = cpu_to_le64(sense_dma); + cmd->common.sense_len = cpu_to_le16(SCSI_SENSE_BUFFERSIZE); + cmd->common.command_id = pt_cmd->cid; + + spraid_submit_cmd(ioq, cmd); + + if (!wait_for_completion_timeout(&pt_cmd->cmd_done, timeout)) { + dev_err(hdev->dev, "[%s] cid[%d] qid[%d] timeout;" + " opcode[0x%x] subopcode[0x%x]\n", + __func__, pt_cmd->cid, pt_cmd->qid, cmd->common.opcode, + (le32_to_cpu(cmd->common.cdw3[0]) & 0xffff)); + WRITE_ONCE(pt_cmd->state, SPRAID_CMD_TIMEOUT); + spraid_put_cmd(hdev, pt_cmd, SPRAID_CMD_IOPT); + return -EINVAL; + } + + if (result && reslen) { + if ((pt_cmd->status & 0x17f) == 0x101) { + memcpy(result, sense_addr, SCSI_SENSE_BUFFERSIZE); + *reslen = SCSI_SENSE_BUFFERSIZE; + } + } + + spraid_put_cmd(hdev, pt_cmd, SPRAID_CMD_IOPT); + + return pt_cmd->status; +} + +static int spraid_user_ioq_cmd(struct spraid_dev *hdev, struct bsg_job *job) +{ + struct spraid_bsg_request *bsg_req = + (struct spraid_bsg_request *)(job->request); + struct spraid_ioq_passthru_cmd *cmd = &(bsg_req->ioqcmd); + struct spraid_ioq_command ioq_cmd; + int status = 0; + u32 timeout = msecs_to_jiffies(cmd->timeout_ms); + + if (cmd->data_len > PAGE_SIZE) { + dev_err(hdev->dev, "[%s] data len bigger than 4k\n", __func__); + return -EFAULT; + } + + if (hdev->state != SPRAID_LIVE) { + dev_err(hdev->dev, "[%s] err, host state:[%d] is not live\n", + __func__, hdev->state); + return -EBUSY; + } + + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x] init;" + " datalen[%d]\n", + __func__, cmd->opcode, cmd->info_1.subopcode, cmd->data_len); + + memset(&ioq_cmd, 0, sizeof(ioq_cmd)); + ioq_cmd.common.opcode = cmd->opcode; + ioq_cmd.common.flags = cmd->flags; + ioq_cmd.common.hdid = cpu_to_le32(cmd->nsid); + ioq_cmd.common.sense_len = cpu_to_le16(cmd->info_0.res_sense_len); + ioq_cmd.common.cdb_len = cmd->info_0.cdb_len; + ioq_cmd.common.rsvd2 = cmd->info_0.rsvd0; + ioq_cmd.common.cdw3[0] = cpu_to_le32(cmd->cdw3); + ioq_cmd.common.cdw3[1] = cpu_to_le32(cmd->cdw4); + ioq_cmd.common.cdw3[2] = cpu_to_le32(cmd->cdw5); + + ioq_cmd.common.cdw10[0] = cpu_to_le32(cmd->cdw10); + ioq_cmd.common.cdw10[1] = cpu_to_le32(cmd->cdw11); + ioq_cmd.common.cdw10[2] = cpu_to_le32(cmd->cdw12); + ioq_cmd.common.cdw10[3] = cpu_to_le32(cmd->cdw13); + ioq_cmd.common.cdw10[4] = cpu_to_le32(cmd->cdw14); + ioq_cmd.common.cdw10[5] = cpu_to_le32(cmd->data_len); + + memcpy(ioq_cmd.common.cdb, &cmd->cdw16, cmd->info_0.cdb_len); + + ioq_cmd.common.cdw26[0] = cpu_to_le32(cmd->cdw26[0]); + ioq_cmd.common.cdw26[1] = cpu_to_le32(cmd->cdw26[1]); + ioq_cmd.common.cdw26[2] = cpu_to_le32(cmd->cdw26[2]); + ioq_cmd.common.cdw26[3] = cpu_to_le32(cmd->cdw26[3]); + + status = spraid_bsg_map_data(hdev, job, + (struct spraid_admin_command *)&ioq_cmd); + if (status) { + dev_err(hdev->dev, "[%s] err, map data failed\n", __func__); + return status; + } + + status = spraid_submit_ioq_sync_cmd(hdev, &ioq_cmd, job->reply, + &job->reply_len, timeout); + + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x], status[0x%x];" + " reply_len[%d]\n", + __func__, cmd->opcode, cmd->info_1.subopcode, + status, job->reply_len); + + spraid_bsg_unmap_data(hdev, job); + + return status; +} + +static bool spraid_check_scmd_completed(struct scsi_cmnd *scmd) +{ + struct spraid_dev *hdev = shost_priv(scmd->device->host); + struct spraid_iod *iod = scsi_cmd_priv(scmd); + struct spraid_queue *spraidq; + u16 hwq, cid; + + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + spraidq = &hdev->queues[hwq]; + if (READ_ONCE(iod->state) == SPRAID_CMD_COMPLETE + || spraid_poll_cq(spraidq, cid)) { + dev_warn(hdev->dev, "cid[%d] qid[%d] has been completed\n", + cid, spraidq->qid); + return true; + } + return false; +} + +static enum blk_eh_timer_return spraid_scmd_timeout(struct scsi_cmnd *scmd) +{ + struct spraid_iod *iod = scsi_cmd_priv(scmd); + unsigned int timeout = scmd->device->request_queue->rq_timeout; + + if (spraid_check_scmd_completed(scmd)) + goto out; + + if (time_after(jiffies, scmd->jiffies_at_alloc + timeout)) { + if (cmpxchg(&iod->state, + SPRAID_CMD_IN_FLIGHT, + SPRAID_CMD_TIMEOUT) == SPRAID_CMD_IN_FLIGHT) { + return BLK_EH_DONE; + } + } +out: + return BLK_EH_RESET_TIMER; +} + +/* send abort command by admin queue temporary */ +static int spraid_send_abort_cmd(struct spraid_dev *hdev, + u32 hdid, u16 qid, u16 cid) +{ + struct spraid_admin_command admin_cmd; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.abort.opcode = SPRAID_ADMIN_ABORT_CMD; + admin_cmd.abort.hdid = cpu_to_le32(hdid); + admin_cmd.abort.sqid = cpu_to_le16(qid); + admin_cmd.abort.cid = cpu_to_le16(cid); + + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); +} + +/* send reset command by admin quueue temporary */ +static int spraid_send_reset_cmd(struct spraid_dev *hdev, int type, u32 hdid) +{ + struct spraid_admin_command admin_cmd; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.reset.opcode = SPRAID_ADMIN_RESET; + admin_cmd.reset.hdid = cpu_to_le32(hdid); + admin_cmd.reset.type = type; + + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); +} + +static bool spraid_change_host_state(struct spraid_dev *hdev, + enum spraid_state newstate) +{ + unsigned long flags; + enum spraid_state oldstate; + bool change = false; + + spin_lock_irqsave(&hdev->state_lock, flags); + + oldstate = hdev->state; + switch (newstate) { + case SPRAID_LIVE: + switch (oldstate) { + case SPRAID_NEW: + case SPRAID_RESETTING: + change = true; + break; + default: + break; + } + break; + case SPRAID_RESETTING: + switch (oldstate) { + case SPRAID_LIVE: + change = true; + break; + default: + break; + } + break; + case SPRAID_DELETING: + if (oldstate != SPRAID_DELETING) + change = true; + break; + case SPRAID_DEAD: + switch (oldstate) { + case SPRAID_NEW: + case SPRAID_LIVE: + case SPRAID_RESETTING: + change = true; + break; + default: + break; + } + break; + default: + break; + } + if (change) + hdev->state = newstate; + spin_unlock_irqrestore(&hdev->state_lock, flags); + + dev_info(hdev->dev, "[%s][%d]->[%d], change[%d]\n", + __func__, oldstate, newstate, change); + + return change; +} + +static void spraid_back_fault_cqe(struct spraid_queue *ioq, + struct spraid_completion *cqe) +{ + struct spraid_dev *hdev = ioq->hdev; + struct blk_mq_tags *tags; + struct scsi_cmnd *scmd; + struct spraid_iod *iod; + struct request *req; + + tags = hdev->shost->tag_set.tags[ioq->qid - 1]; + req = blk_mq_tag_to_rq(tags, cqe->cmd_id); + if (unlikely(!req || !blk_mq_request_started(req))) + return; + + scmd = blk_mq_rq_to_pdu(req); + iod = scsi_cmd_priv(scmd); + + set_host_byte(scmd, DID_NO_CONNECT); + if (iod->nsge) + scsi_dma_unmap(scmd); + spraid_free_iod_res(hdev, iod); + scmd->scsi_done(scmd); + dev_warn(hdev->dev, "Back fault CQE, cid[%d] qid[%d]\n", + cqe->cmd_id, ioq->qid); +} + +static void spraid_back_all_io(struct spraid_dev *hdev) +{ + int i, j; + struct spraid_queue *ioq; + struct spraid_completion cqe = { 0 }; + + scsi_block_requests(hdev->shost); + + for (i = 1; i <= hdev->shost->nr_hw_queues; i++) { + ioq = &hdev->queues[i]; + for (j = 0; j < hdev->shost->can_queue; j++) { + cqe.cmd_id = j; + spraid_back_fault_cqe(ioq, &cqe); + } + } + + scsi_unblock_requests(hdev->shost); +} + +static void spraid_dev_disable(struct spraid_dev *hdev, bool shutdown) +{ + struct spraid_queue *adminq = &hdev->queues[0]; + u16 start, end; + unsigned long timeout = jiffies + 600 * HZ; + + if (pci_device_is_present(hdev->pdev)) { + if (shutdown) + spraid_shutdown_ctrl(hdev); + else + spraid_disable_ctrl(hdev); + } + + while (!time_after(jiffies, timeout)) { + if (!pci_device_is_present(hdev->pdev)) { + dev_info(hdev->dev, "[%s] pci_device not present;" + " skip wait\n", __func__); + break; + } + if (!spraid_wait_ready(hdev, hdev->cap, false)) { + dev_info(hdev->dev, + "[%s] wait ready success after reset\n", + __func__); + break; + } + dev_info(hdev->dev, "[%s] waiting csts_rdy ready\n", __func__); + } + + if (hdev->queue_count == 0) { + dev_err(hdev->dev, "[%s] warn, queue has been delete\n", + __func__); + return; + } + + spin_lock_irq(&adminq->cq_lock); + spraid_process_cq(adminq, &start, &end, -1); + spin_unlock_irq(&adminq->cq_lock); + spraid_complete_cqes(adminq, start, end); + + spraid_pci_disable(hdev); + + spraid_back_all_io(hdev); +} + +static void spraid_reset_work(struct work_struct *work) +{ + int ret; + struct spraid_dev *hdev = + container_of(work, struct spraid_dev, reset_work); + + if (hdev->state != SPRAID_RESETTING) { + dev_err(hdev->dev, "[%s] err, host is not reset state\n", + __func__); + return; + } + + dev_info(hdev->dev, "[%s] enter host reset\n", __func__); + + if (hdev->ctrl_config & SPRAID_CC_ENABLE) { + dev_info(hdev->dev, "[%s] start dev_disable\n", __func__); + spraid_dev_disable(hdev, false); + } + + ret = spraid_pci_enable(hdev); + if (ret) + goto out; + + ret = spraid_setup_admin_queue(hdev); + if (ret) + goto pci_disable; + + ret = spraid_setup_io_queues(hdev); + if (ret || hdev->online_queues <= hdev->shost->nr_hw_queues) + goto pci_disable; + + spraid_change_host_state(hdev, SPRAID_LIVE); + + spraid_send_all_aen(hdev); + + return; + +pci_disable: + spraid_pci_disable(hdev); +out: + spraid_change_host_state(hdev, SPRAID_DEAD); + dev_err(hdev->dev, "[%s] err, host reset failed\n", __func__); +} + +static int spraid_reset_work_sync(struct spraid_dev *hdev) +{ + if (!spraid_change_host_state(hdev, SPRAID_RESETTING)) { + dev_info(hdev->dev, "[%s] can't change to reset state\n", + __func__); + return -EBUSY; + } + + if (!queue_work(spraid_wq, &hdev->reset_work)) { + dev_err(hdev->dev, "[%s] err, host is already in reset state\n", + __func__); + return -EBUSY; + } + + flush_work(&hdev->reset_work); + if (hdev->state != SPRAID_LIVE) + return -ENODEV; + + return 0; +} + +static int spraid_wait_abnl_cmd_done(struct spraid_iod *iod) +{ + u16 times = 0; + + do { + if (READ_ONCE(iod->state) == SPRAID_CMD_TMO_COMPLETE) + break; + msleep(500); + times++; + } while (times <= SPRAID_WAIT_ABNL_CMD_TIMEOUT); + + /* wait command completion timeout after abort/reset success */ + if (times >= SPRAID_WAIT_ABNL_CMD_TIMEOUT) + return -ETIMEDOUT; + + return 0; +} + +static int spraid_abort_handler(struct scsi_cmnd *scmd) +{ + struct spraid_dev *hdev = shost_priv(scmd->device->host); + struct spraid_iod *iod = scsi_cmd_priv(scmd); + struct spraid_sdev_hostdata *hostdata; + u16 hwq, cid; + int ret; + + scsi_print_command(scmd); + + if (hdev->state != SPRAID_LIVE || !spraid_wait_abnl_cmd_done(iod) || + spraid_check_scmd_completed(scmd)) + return SUCCESS; + + hostdata = scmd->device->hostdata; + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + + dev_warn(hdev->dev, "cid[%d] qid[%d] timeout, aborting\n", cid, hwq); + ret = spraid_send_abort_cmd(hdev, hostdata->hdid, hwq, cid); + if (ret != ADMIN_ERR_TIMEOUT) { + ret = spraid_wait_abnl_cmd_done(iod); + if (ret) { + dev_warn(hdev->dev, "cid[%d] qid[%d] abort failed;" + " not found\n", cid, hwq); + return FAILED; + } + dev_warn(hdev->dev, "cid[%d] qid[%d] abort succ\n", cid, hwq); + return SUCCESS; + } + dev_warn(hdev->dev, "cid[%d] qid[%d] abort failed, timeout\n", + cid, hwq); + return FAILED; +} + +static int spraid_tgt_reset_handler(struct scsi_cmnd *scmd) +{ + struct spraid_dev *hdev = shost_priv(scmd->device->host); + struct spraid_iod *iod = scsi_cmd_priv(scmd); + struct spraid_sdev_hostdata *hostdata; + u16 hwq, cid; + int ret; + + scsi_print_command(scmd); + + if (hdev->state != SPRAID_LIVE || !spraid_wait_abnl_cmd_done(iod) || + spraid_check_scmd_completed(scmd)) + return SUCCESS; + + hostdata = scmd->device->hostdata; + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + + dev_warn(hdev->dev, "cid[%d] qid[%d] timeout, target reset\n", + cid, hwq); + ret = spraid_send_reset_cmd(hdev, SPRAID_RESET_TARGET, hostdata->hdid); + if (ret == 0) { + ret = spraid_wait_abnl_cmd_done(iod); + if (ret) { + dev_warn(hdev->dev, + "cid[%d] qid[%d]target reset failed;" + " not found\n", cid, hwq); + return FAILED; + } + + dev_warn(hdev->dev, "cid[%d] qid[%d] target reset success\n", + cid, hwq); + return SUCCESS; + } + + dev_warn(hdev->dev, "cid[%d] qid[%d] ret[%d] target reset failed\n", + cid, hwq, ret); + return FAILED; +} + +static int spraid_bus_reset_handler(struct scsi_cmnd *scmd) +{ + struct spraid_dev *hdev = shost_priv(scmd->device->host); + struct spraid_iod *iod = scsi_cmd_priv(scmd); + struct spraid_sdev_hostdata *hostdata; + u16 hwq, cid; + int ret; + + scsi_print_command(scmd); + + if (hdev->state != SPRAID_LIVE || !spraid_wait_abnl_cmd_done(iod) || + spraid_check_scmd_completed(scmd)) + return SUCCESS; + + hostdata = scmd->device->hostdata; + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + + dev_warn(hdev->dev, "cid[%d] qid[%d] timeout, bus reset\n", cid, hwq); + ret = spraid_send_reset_cmd(hdev, SPRAID_RESET_BUS, hostdata->hdid); + if (ret == 0) { + ret = spraid_wait_abnl_cmd_done(iod); + if (ret) { + dev_warn(hdev->dev, + "cid[%d] qid[%d] bus reset failed;" + " not found\n", cid, hwq); + return FAILED; + } + + dev_warn(hdev->dev, "cid[%d] qid[%d] bus reset succ\n", + cid, hwq); + return SUCCESS; + } + + dev_warn(hdev->dev, "cid[%d] qid[%d] ret[%d] bus reset failed\n", + cid, hwq, ret); + return FAILED; +} + +static int spraid_shost_reset_handler(struct scsi_cmnd *scmd) +{ + u16 hwq, cid; + struct spraid_dev *hdev = shost_priv(scmd->device->host); + + scsi_print_command(scmd); + if (hdev->state != SPRAID_LIVE || spraid_check_scmd_completed(scmd)) + return SUCCESS; + + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + dev_warn(hdev->dev, "cid[%d] qid[%d] host reset\n", cid, hwq); + + if (spraid_reset_work_sync(hdev)) { + dev_warn(hdev->dev, "cid[%d] qid[%d] host reset failed\n", + cid, hwq); + return FAILED; + } + + dev_warn(hdev->dev, "cid[%d] qid[%d] host reset success\n", cid, hwq); + + return SUCCESS; +} + +static pci_ers_result_t spraid_pci_error_detected(struct pci_dev *pdev, + pci_channel_state_t state) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "enter pci error detect, state:%d\n", state); + + switch (state) { + case pci_channel_io_normal: + dev_warn(hdev->dev, "channel is normal, do nothing\n"); + + return PCI_ERS_RESULT_CAN_RECOVER; + case pci_channel_io_frozen: + dev_warn(hdev->dev, + "channel io frozen, need reset controller\n"); + + scsi_block_requests(hdev->shost); + + spraid_change_host_state(hdev, SPRAID_RESETTING); + + return PCI_ERS_RESULT_NEED_RESET; + case pci_channel_io_perm_failure: + dev_warn(hdev->dev, "channel io failure, request disconnect\n"); + + return PCI_ERS_RESULT_DISCONNECT; + } + + return PCI_ERS_RESULT_NEED_RESET; +} + +static pci_ers_result_t spraid_pci_slot_reset(struct pci_dev *pdev) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "restart after slot reset\n"); + + pci_restore_state(pdev); + + if (!queue_work(spraid_wq, &hdev->reset_work)) { + dev_err(hdev->dev, "[%s] err, the device is resetting state\n", + __func__); + return PCI_ERS_RESULT_NONE; + } + + flush_work(&hdev->reset_work); + + scsi_unblock_requests(hdev->shost); + + return PCI_ERS_RESULT_RECOVERED; +} + +static void spraid_reset_done(struct pci_dev *pdev) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "enter spraid reset done\n"); +} + +static ssize_t csts_pp_show(struct device *cdev, + struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct spraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) { + ret = (readl(hdev->bar + SPRAID_REG_CSTS) + & SPRAID_CSTS_PP_MASK); + ret >>= SPRAID_CSTS_PP_SHIFT; + } + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t csts_shst_show(struct device *cdev, + struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct spraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) { + ret = (readl(hdev->bar + SPRAID_REG_CSTS) + & SPRAID_CSTS_SHST_MASK); + ret >>= SPRAID_CSTS_SHST_SHIFT; + } + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t csts_cfs_show(struct device *cdev, + struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct spraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) { + ret = (readl(hdev->bar + SPRAID_REG_CSTS) + & SPRAID_CSTS_CFS_MASK); + ret >>= SPRAID_CSTS_CFS_SHIFT; + } + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t csts_rdy_show(struct device *cdev, + struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct spraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) + ret = (readl(hdev->bar + SPRAID_REG_CSTS) & SPRAID_CSTS_RDY); + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t fw_version_show(struct device *cdev, + struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct spraid_dev *hdev = shost_priv(shost); + + return snprintf(buf, PAGE_SIZE, "%s\n", hdev->ctrl_info->fr); +} + +static DEVICE_ATTR_RO(csts_pp); +static DEVICE_ATTR_RO(csts_shst); +static DEVICE_ATTR_RO(csts_cfs); +static DEVICE_ATTR_RO(csts_rdy); +static DEVICE_ATTR_RO(fw_version); + +static struct device_attribute *spraid_host_attrs[] = { + &dev_attr_csts_pp, + &dev_attr_csts_shst, + &dev_attr_csts_cfs, + &dev_attr_csts_rdy, + &dev_attr_fw_version, + NULL, +}; + +static int spraid_get_vd_info(struct spraid_dev *hdev, + struct spraid_vd_info *vd_info, u16 vid) +{ + struct spraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + int ret; + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, + &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.usr_cmd.opcode = USR_CMD_READ; + admin_cmd.usr_cmd.info_0.subopcode = cpu_to_le16(USR_CMD_VDINFO); + admin_cmd.usr_cmd.info_1.data_len = cpu_to_le16(USR_CMD_RDLEN); + admin_cmd.usr_cmd.info_1.param_len = cpu_to_le16(VDINFO_PARAM_LEN); + admin_cmd.usr_cmd.cdw10 = cpu_to_le32(vid); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + if (!ret) + memcpy(vd_info, data_ptr, sizeof(struct spraid_vd_info)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, data_dma); + + return ret; +} + +static int spraid_get_bgtask(struct spraid_dev *hdev, + struct spraid_bgtask *bgtask) +{ + struct spraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + int ret; + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, + &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.usr_cmd.opcode = USR_CMD_READ; + admin_cmd.usr_cmd.info_0.subopcode = cpu_to_le16(USR_CMD_BGTASK); + admin_cmd.usr_cmd.info_1.data_len = cpu_to_le16(USR_CMD_RDLEN); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + if (!ret) + memcpy(bgtask, data_ptr, sizeof(struct spraid_bgtask)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, data_dma); + + return ret; +} + +static ssize_t raid_level_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct spraid_dev *hdev; + struct spraid_vd_info *vd_info; + struct spraid_sdev_hostdata *hostdata; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !SPRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = spraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) + vd_info->rg_level = ARRAY_SIZE(raid_levels) - 1; + + ret = (vd_info->rg_level < ARRAY_SIZE(raid_levels)) ? + vd_info->rg_level : (ARRAY_SIZE(raid_levels) - 1); + + kfree(vd_info); + + return snprintf(buf, PAGE_SIZE, "RAID-%s\n", raid_levels[ret]); +} + +static ssize_t raid_state_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct spraid_dev *hdev; + struct spraid_vd_info *vd_info; + struct spraid_sdev_hostdata *hostdata; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !SPRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = spraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) { + vd_info->vd_status = 0; + vd_info->rg_id = 0xff; + } + + ret = (vd_info->vd_status < ARRAY_SIZE(raid_states)) ? + vd_info->vd_status : 0; + + kfree(vd_info); + + return snprintf(buf, PAGE_SIZE, "%s\n", raid_states[ret]); +} + +static ssize_t raid_resync_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct spraid_dev *hdev; + struct spraid_vd_info *vd_info; + struct spraid_bgtask *bgtask; + struct spraid_sdev_hostdata *hostdata; + u8 rg_id, i, progress = 0; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !SPRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = spraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) + goto out; + + rg_id = vd_info->rg_id; + + bgtask = (struct spraid_bgtask *)vd_info; + ret = spraid_get_bgtask(hdev, bgtask); + if (ret) + goto out; + for (i = 0; i < bgtask->task_num; i++) { + if ((bgtask->bgtask[i].type == BGTASK_TYPE_REBUILD) && + (le16_to_cpu(bgtask->bgtask[i].vd_id) == rg_id)) + progress = bgtask->bgtask[i].progress; + } + +out: + kfree(vd_info); + return snprintf(buf, PAGE_SIZE, "%d\n", progress); +} + +static DEVICE_ATTR_RO(raid_level); +static DEVICE_ATTR_RO(raid_state); +static DEVICE_ATTR_RO(raid_resync); + +static struct device_attribute *spraid_dev_attrs[] = { + &dev_attr_raid_level, + &dev_attr_raid_state, + &dev_attr_raid_resync, + NULL, +}; + +static struct pci_error_handlers spraid_err_handler = { + .error_detected = spraid_pci_error_detected, + .slot_reset = spraid_pci_slot_reset, + .reset_done = spraid_reset_done, +}; + +static int spraid_sysfs_host_reset(struct Scsi_Host *shost, int reset_type) +{ + int ret; + struct spraid_dev *hdev = shost_priv(shost); + + dev_info(hdev->dev, "[%s] start sysfs host reset cmd\n", __func__); + ret = spraid_reset_work_sync(hdev); + dev_info(hdev->dev, "[%s] stop sysfs host reset cmd[%d]\n", + __func__, ret); + + return ret; +} + +static struct scsi_host_template spraid_driver_template = { + .module = THIS_MODULE, + .name = "Ramaxel Logic spraid driver", + .proc_name = "spraid", + .queuecommand = spraid_queue_command, + .slave_alloc = spraid_slave_alloc, + .slave_destroy = spraid_slave_destroy, + .slave_configure = spraid_slave_configure, + .eh_timed_out = spraid_scmd_timeout, + .eh_abort_handler = spraid_abort_handler, + .eh_target_reset_handler = spraid_tgt_reset_handler, + .eh_bus_reset_handler = spraid_bus_reset_handler, + .eh_host_reset_handler = spraid_shost_reset_handler, + .change_queue_depth = scsi_change_queue_depth, + .this_id = -1, + .shost_attrs = spraid_host_attrs, + .sdev_attrs = spraid_dev_attrs, + .host_reset = spraid_sysfs_host_reset, +}; + +static void spraid_shutdown(struct pci_dev *pdev) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + + spraid_remove_io_queues(hdev); + spraid_disable_admin_queue(hdev, true); +} + +/* bsg dispatch user command */ +static int spraid_bsg_host_dispatch(struct bsg_job *job) +{ + struct Scsi_Host *shost = dev_to_shost(job->dev); + struct spraid_dev *hdev = shost_priv(shost); + struct request *rq = blk_mq_rq_from_pdu(job); + struct spraid_bsg_request *bsg_req = job->request; + int ret = 0; + + dev_info(hdev->dev, "[%s] msgcode[%d], msglen[%d], timeout[%d];" + " req_nsge[%d], req_len[%d]\n", + __func__, bsg_req->msgcode, job->request_len, + rq->timeout, job->request_payload.sg_cnt, + job->request_payload.payload_len); + + job->reply_len = 0; + + switch (bsg_req->msgcode) { + case SPRAID_BSG_ADM: + ret = spraid_user_admin_cmd(hdev, job); + break; + case SPRAID_BSG_IOQ: + ret = spraid_user_ioq_cmd(hdev, job); + break; + default: + dev_info(hdev->dev, "[%s] unsupport msgcode[%d]\n", + __func__, bsg_req->msgcode); + break; + } + + if (ret > 0) + ret = ret | (ret << 8); + + bsg_job_done(job, ret, 0); + return 0; +} + +static inline void spraid_remove_bsg(struct spraid_dev *hdev) +{ + if (hdev->bsg_queue) { + bsg_unregister_queue(hdev->bsg_queue); + blk_cleanup_queue(hdev->bsg_queue); + } +} +static int spraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) +{ + struct spraid_dev *hdev; + struct Scsi_Host *shost; + int node, ret; + char bsg_name[15]; + + shost = scsi_host_alloc(&spraid_driver_template, sizeof(*hdev)); + if (!shost) { + dev_err(&pdev->dev, "Failed to allocate scsi host\n"); + return -ENOMEM; + } + hdev = shost_priv(shost); + hdev->pdev = pdev; + hdev->dev = get_device(&pdev->dev); + + node = dev_to_node(hdev->dev); + if (node == NUMA_NO_NODE) { + node = first_memory_node; + set_dev_node(hdev->dev, node); + } + hdev->numa_node = node; + hdev->shost = shost; + pci_set_drvdata(pdev, hdev); + + ret = spraid_dev_map(hdev); + if (ret) + goto put_dev; + + init_rwsem(&hdev->devices_rwsem); + INIT_WORK(&hdev->scan_work, spraid_scan_work); + INIT_WORK(&hdev->timesyn_work, spraid_timesyn_work); + INIT_WORK(&hdev->reset_work, spraid_reset_work); + INIT_WORK(&hdev->fw_act_work, spraid_fw_act_work); + spin_lock_init(&hdev->state_lock); + + ret = spraid_alloc_resources(hdev); + if (ret) + goto dev_unmap; + + ret = spraid_pci_enable(hdev); + if (ret) + goto resources_free; + + ret = spraid_setup_admin_queue(hdev); + if (ret) + goto pci_disable; + + ret = spraid_init_ctrl_info(hdev); + if (ret) + goto disable_admin_q; + + ret = spraid_alloc_iod_ext_mem_pool(hdev); + if (ret) + goto disable_admin_q; + + ret = spraid_setup_io_queues(hdev); + if (ret) + goto free_iod_mempool; + + spraid_shost_init(hdev); + + ret = scsi_add_host(hdev->shost, hdev->dev); + if (ret) { + dev_err(hdev->dev, "Add shost to system failed, ret: %d\n", + ret); + goto remove_io_queues; + } + + snprintf(bsg_name, sizeof(bsg_name), "spraid%d", shost->host_no); + hdev->bsg_queue = bsg_setup_queue(&shost->shost_gendev, bsg_name, + spraid_bsg_host_dispatch, + spraid_cmd_size(hdev, true, false)); + if (IS_ERR(hdev->bsg_queue)) { + dev_err(hdev->dev, "err, setup bsg failed\n"); + hdev->bsg_queue = NULL; + goto remove_io_queues; + } + + if (hdev->online_queues == SPRAID_ADMIN_QUEUE_NUM) { + dev_warn(hdev->dev, "warn only admin queue can be used\n"); + return 0; + } + + hdev->state = SPRAID_LIVE; + + spraid_send_all_aen(hdev); + + ret = spraid_dev_list_init(hdev); + if (ret) + goto remove_bsg; + + ret = spraid_configure_timestamp(hdev); + if (ret) + dev_warn(hdev->dev, "init set timestamp failed\n"); + + ret = spraid_alloc_ioq_ptcmds(hdev); + if (ret) + goto remove_bsg; + + scsi_scan_host(hdev->shost); + + return 0; + +remove_bsg: + spraid_remove_bsg(hdev); +remove_io_queues: + spraid_remove_io_queues(hdev); +free_iod_mempool: + spraid_free_iod_ext_mem_pool(hdev); +disable_admin_q: + spraid_disable_admin_queue(hdev, false); +pci_disable: + spraid_pci_disable(hdev); +resources_free: + spraid_free_resources(hdev); +dev_unmap: + spraid_dev_unmap(hdev); +put_dev: + put_device(hdev->dev); + scsi_host_put(shost); + + return -ENODEV; +} + +static void spraid_remove(struct pci_dev *pdev) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + struct Scsi_Host *shost = hdev->shost; + + dev_info(hdev->dev, "enter spraid remove\n"); + + spraid_change_host_state(hdev, SPRAID_DELETING); + flush_work(&hdev->reset_work); + + if (!pci_device_is_present(pdev)) + spraid_back_all_io(hdev); + + spraid_remove_bsg(hdev); + scsi_remove_host(shost); + spraid_free_ioq_ptcmds(hdev); + kfree(hdev->devices); + spraid_remove_io_queues(hdev); + spraid_free_iod_ext_mem_pool(hdev); + spraid_disable_admin_queue(hdev, false); + spraid_pci_disable(hdev); + spraid_free_resources(hdev); + spraid_dev_unmap(hdev); + put_device(hdev->dev); + scsi_host_put(shost); + + dev_info(hdev->dev, "exit spraid remove\n"); +} + +static const struct pci_device_id spraid_id_table[] = { + { PCI_DEVICE(PCI_VENDOR_ID_RAMAXEL_LOGIC, + SPRAID_SERVER_DEVICE_HBA_DID) }, + { PCI_DEVICE(PCI_VENDOR_ID_RAMAXEL_LOGIC, + SPRAID_SERVER_DEVICE_RAID_DID) }, + { 0, } +}; +MODULE_DEVICE_TABLE(pci, spraid_id_table); + +static struct pci_driver spraid_driver = { + .name = "spraid", + .id_table = spraid_id_table, + .probe = spraid_probe, + .remove = spraid_remove, + .shutdown = spraid_shutdown, + .err_handler = &spraid_err_handler, +}; + +static int __init spraid_init(void) +{ + int ret; + + spraid_wq = alloc_workqueue("spraid-wq", WQ_UNBOUND | WQ_MEM_RECLAIM | + WQ_SYSFS, 0); + if (!spraid_wq) + return -ENOMEM; + + spraid_class = class_create(THIS_MODULE, "spraid"); + if (IS_ERR(spraid_class)) { + ret = PTR_ERR(spraid_class); + goto destroy_wq; + } + + ret = pci_register_driver(&spraid_driver); + if (ret < 0) + goto destroy_class; + + return 0; + +destroy_class: + class_destroy(spraid_class); +destroy_wq: + destroy_workqueue(spraid_wq); + + return ret; +} + +static void __exit spraid_exit(void) +{ + pci_unregister_driver(&spraid_driver); + class_destroy(spraid_class); + destroy_workqueue(spraid_wq); + ida_destroy(&spraid_instance_ida); +} + +MODULE_AUTHOR("songyl(a)ramaxel.com"); +MODULE_DESCRIPTION("Ramaxel Memory Technology SPraid Driver"); +MODULE_LICENSE("GPL"); +MODULE_VERSION(SPRAID_DRV_VERSION); +module_init(spraid_init); +module_exit(spraid_exit); -- 2.27.0

2 1

[PATCH openEuler-1.0-LTS] USB: gadget: bRequestType is a bitfield, not a enum
by Yang Yingliang 22 Dec '21

22 Dec '21

From: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> mainline inclusion from mainline-v5.16-rc6 commit f08adf5add9a071160c68bb2a61d697f39ab0758 category: bugfix bugzilla: NA CVE: CVE-2021-39685 -------------------------------- Szymon rightly pointed out that the previous check for the endpoint direction in bRequestType was not looking at only the bit involved, but rather the whole value. Normally this is ok, but for some request types, bits other than bit 8 could be set and the check for the endpoint length could not stall correctly. Fix that up by only checking the single bit. Fixes: 153a2d7e3350 ("USB: gadget: detect too-big endpoint 0 requests") Cc: Felipe Balbi <balbi(a)kernel.org> Reported-by: Szymon Heidrich <szymon.heidrich(a)gmail.com> Link: https://lore.kernel.org/r/20211214184621.385828-1-gregkh@linuxfoundation.org Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> Reviewed-by: Xiu Jianfeng <xiujianfeng(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/usb/gadget/composite.c | 6 +++--- drivers/usb/gadget/legacy/dbgp.c | 6 +++--- drivers/usb/gadget/legacy/inode.c | 6 +++--- 3 files changed, 9 insertions(+), 9 deletions(-) diff --git a/drivers/usb/gadget/composite.c b/drivers/usb/gadget/composite.c index 3813249387b3d..d39ebf5bdbdd5 100644 --- a/drivers/usb/gadget/composite.c +++ b/drivers/usb/gadget/composite.c @@ -1570,14 +1570,14 @@ composite_setup(struct usb_gadget *gadget, const struct usb_ctrlrequest *ctrl) u8 endp; if (w_length > USB_COMP_EP0_BUFSIZ) { - if (ctrl->bRequestType == USB_DIR_OUT) { - goto done; - } else { + if (ctrl->bRequestType & USB_DIR_IN) { /* Cast away the const, we are going to overwrite on purpose. */ __le16 *temp = (__le16 *)&ctrl->wLength; *temp = cpu_to_le16(USB_COMP_EP0_BUFSIZ); w_length = USB_COMP_EP0_BUFSIZ; + } else { + goto done; } } diff --git a/drivers/usb/gadget/legacy/dbgp.c b/drivers/usb/gadget/legacy/dbgp.c index e567afcb2794c..ffe58d44b2cb1 100644 --- a/drivers/usb/gadget/legacy/dbgp.c +++ b/drivers/usb/gadget/legacy/dbgp.c @@ -346,14 +346,14 @@ static int dbgp_setup(struct usb_gadget *gadget, u16 len = 0; if (length > DBGP_REQ_LEN) { - if (ctrl->bRequestType == USB_DIR_OUT) { - return err; - } else { + if (ctrl->bRequestType & USB_DIR_IN) { /* Cast away the const, we are going to overwrite on purpose. */ __le16 *temp = (__le16 *)&ctrl->wLength; *temp = cpu_to_le16(DBGP_REQ_LEN); length = DBGP_REQ_LEN; + } else { + return err; } } diff --git a/drivers/usb/gadget/legacy/inode.c b/drivers/usb/gadget/legacy/inode.c index 83fd1b55d497a..f91d403da3141 100644 --- a/drivers/usb/gadget/legacy/inode.c +++ b/drivers/usb/gadget/legacy/inode.c @@ -1335,14 +1335,14 @@ gadgetfs_setup (struct usb_gadget *gadget, const struct usb_ctrlrequest *ctrl) u16 w_length = le16_to_cpu(ctrl->wLength); if (w_length > RBUF_SIZE) { - if (ctrl->bRequestType == USB_DIR_OUT) { - return value; - } else { + if (ctrl->bRequestType & USB_DIR_IN) { /* Cast away the const, we are going to overwrite on purpose. */ __le16 *temp = (__le16 *)&ctrl->wLength; *temp = cpu_to_le16(RBUF_SIZE); w_length = RBUF_SIZE; + } else { + return value; } } -- 2.25.1

1 0

[PATCH openEuler-1.0-LTS 1/2] scsi:spraid: support Ramaxel's spraid driver
by Yang Yingliang 22 Dec '21

22 Dec '21

From: Yanling Song <songyl(a)ramaxel.com> Ramaxel inclusion category: features bugzilla: https://gitee.com/openeuler/kernel/issues/I4NJIB CVE: NA Support Ramaxel's SPRxxx Raid controller Signed-off-by: Yanling Song <songyl(a)ramaxel.com> Reviewed-by: Jiang Yu<yujiang(a)ramaxel.com> Acked-by: Xie XiuQi <xiexiuqi(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/scsi/Kconfig | 1 + drivers/scsi/Makefile | 1 + drivers/scsi/spraid/Kconfig | 13 + drivers/scsi/spraid/Makefile | 7 + drivers/scsi/spraid/spraid.h | 746 ++++++ drivers/scsi/spraid/spraid_main.c | 3875 +++++++++++++++++++++++++++++ 6 files changed, 4643 insertions(+) create mode 100644 drivers/scsi/spraid/Kconfig create mode 100644 drivers/scsi/spraid/Makefile create mode 100644 drivers/scsi/spraid/spraid.h create mode 100644 drivers/scsi/spraid/spraid_main.c diff --git a/drivers/scsi/Kconfig b/drivers/scsi/Kconfig index 8450484184e36..63d2aaa228349 100644 --- a/drivers/scsi/Kconfig +++ b/drivers/scsi/Kconfig @@ -522,6 +522,7 @@ source "drivers/scsi/megaraid/Kconfig.megaraid" source "drivers/scsi/mpt3sas/Kconfig" source "drivers/scsi/smartpqi/Kconfig" source "drivers/scsi/ufs/Kconfig" +source "drivers/scsi/spraid/Kconfig" config SCSI_HPTIOP tristate "HighPoint RocketRAID 3xxx/4xxx Controller support" diff --git a/drivers/scsi/Makefile b/drivers/scsi/Makefile index 2973693f6dcc7..4056cf26e09e6 100644 --- a/drivers/scsi/Makefile +++ b/drivers/scsi/Makefile @@ -94,6 +94,7 @@ obj-$(CONFIG_SCSI_ZALON) += zalon7xx.o obj-$(CONFIG_SCSI_DC395x) += dc395x.o obj-$(CONFIG_SCSI_AM53C974) += esp_scsi.o am53c974.o obj-$(CONFIG_CXLFLASH) += cxlflash/ +obj-$(CONFIG_RAMAXEL_SPRAID) += spraid/ obj-$(CONFIG_MEGARAID_LEGACY) += megaraid.o obj-$(CONFIG_MEGARAID_NEWGEN) += megaraid/ obj-$(CONFIG_MEGARAID_SAS) += megaraid/ diff --git a/drivers/scsi/spraid/Kconfig b/drivers/scsi/spraid/Kconfig new file mode 100644 index 0000000000000..bfbba3db8db03 --- /dev/null +++ b/drivers/scsi/spraid/Kconfig @@ -0,0 +1,13 @@ +# +# Ramaxel driver configuration +# + +config RAMAXEL_SPRAID + tristate "Ramaxel spraid Adapter" + depends on PCI && SCSI + select BLK_DEV_BSGLIB + depends on ARM64 || X86_64 + help + This driver supports Ramaxel SPRxxx serial + raid controller, which has PCIE Gen4 interface + with host and supports SAS/SATA Hdd/ssd. diff --git a/drivers/scsi/spraid/Makefile b/drivers/scsi/spraid/Makefile new file mode 100644 index 0000000000000..aadc2ffd37ebd --- /dev/null +++ b/drivers/scsi/spraid/Makefile @@ -0,0 +1,7 @@ +# +# Makefile for the Ramaxel device drivers. +# + +obj-$(CONFIG_RAMAXEL_SPRAID) += spraid.o + +spraid-objs := spraid_main.o \ No newline at end of file diff --git a/drivers/scsi/spraid/spraid.h b/drivers/scsi/spraid/spraid.h new file mode 100644 index 0000000000000..c1e4980e18e5d --- /dev/null +++ b/drivers/scsi/spraid/spraid.h @@ -0,0 +1,746 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* Copyright(c) 2021 Ramaxel Memory Technology, Ltd */ + +#ifndef __SPRAID_H_ +#define __SPRAID_H_ + +#define SPRAID_CAP_MQES(cap) ((cap) & 0xffff) +#define SPRAID_CAP_STRIDE(cap) (((cap) >> 32) & 0xf) +#define SPRAID_CAP_MPSMIN(cap) (((cap) >> 48) & 0xf) +#define SPRAID_CAP_MPSMAX(cap) (((cap) >> 52) & 0xf) +#define SPRAID_CAP_TIMEOUT(cap) (((cap) >> 24) & 0xff) +#define SPRAID_CAP_DMAMASK(cap) (((cap) >> 37) & 0xff) + +#define SPRAID_DEFAULT_MAX_CHANNEL 4 +#define SPRAID_DEFAULT_MAX_ID 240 +#define SPRAID_DEFAULT_MAX_LUN_PER_HOST 8 +#define MAX_SECTORS 2048 + +#define IO_SQE_SIZE sizeof(struct spraid_ioq_command) +#define ADMIN_SQE_SIZE sizeof(struct spraid_admin_command) +#define SQE_SIZE(qid) (((qid) > 0) ? IO_SQE_SIZE : ADMIN_SQE_SIZE) +#define CQ_SIZE(depth) ((depth) * sizeof(struct spraid_completion)) +#define SQ_SIZE(qid, depth) ((depth) * SQE_SIZE(qid)) + +#define SENSE_SIZE(depth) ((depth) * SCSI_SENSE_BUFFERSIZE) + +#define SPRAID_AQ_DEPTH 128 +#define SPRAID_NR_AEN_COMMANDS 16 +#define SPRAID_AQ_BLK_MQ_DEPTH (SPRAID_AQ_DEPTH - SPRAID_NR_AEN_COMMANDS) +#define SPRAID_AQ_MQ_TAG_DEPTH (SPRAID_AQ_BLK_MQ_DEPTH - 1) + +#define SPRAID_ADMIN_QUEUE_NUM 1 +#define SPRAID_PTCMDS_PERQ 1 +#define SPRAID_IO_BLK_MQ_DEPTH (hdev->shost->can_queue) +#define SPRAID_NR_IOQ_PTCMDS (SPRAID_PTCMDS_PERQ * hdev->shost->nr_hw_queues) + +#define FUA_MASK 0x08 +#define SPRAID_MINORS BIT(MINORBITS) + +#define COMMAND_IS_WRITE(cmd) ((cmd)->common.opcode & 1) + +#define SPRAID_IO_IOSQES 7 +#define SPRAID_IO_IOCQES 4 +#define PRP_ENTRY_SIZE 8 + +#define SMALL_POOL_SIZE 256 +#define MAX_SMALL_POOL_NUM 16 +#define MAX_CMD_PER_DEV 64 +#define MAX_CDB_LEN 32 + +#define SPRAID_UP_TO_MULTY4(x) (((x) + 4) & (~0x03)) + +#define CQE_STATUS_SUCCESS (0x0) + +#define PCI_VENDOR_ID_RAMAXEL_LOGIC 0x1E81 + +#define SPRAID_SERVER_DEVICE_HBA_DID 0x2100 +#define SPRAID_SERVER_DEVICE_RAID_DID 0x2200 + +#define IO_6_DEFAULT_TX_LEN 256 + +#define SPRAID_INT_PAGES 2 +#define SPRAID_INT_BYTES(hdev) (SPRAID_INT_PAGES * (hdev)->page_size) + +enum { + SPRAID_REQ_CANCELLED = (1 << 0), + SPRAID_REQ_USERCMD = (1 << 1), +}; + +enum { + SPRAID_SC_SUCCESS = 0x0, + SPRAID_SC_INVALID_OPCODE = 0x1, + SPRAID_SC_INVALID_FIELD = 0x2, + + SPRAID_SC_ABORT_LIMIT = 0x103, + SPRAID_SC_ABORT_MISSING = 0x104, + SPRAID_SC_ASYNC_LIMIT = 0x105, + + SPRAID_SC_DNR = 0x4000, +}; + +enum { + SPRAID_REG_CAP = 0x0000, + SPRAID_REG_CC = 0x0014, + SPRAID_REG_CSTS = 0x001c, + SPRAID_REG_AQA = 0x0024, + SPRAID_REG_ASQ = 0x0028, + SPRAID_REG_ACQ = 0x0030, + SPRAID_REG_DBS = 0x1000, +}; + +enum { + SPRAID_CC_ENABLE = 1 << 0, + SPRAID_CC_CSS_NVM = 0 << 4, + SPRAID_CC_MPS_SHIFT = 7, + SPRAID_CC_AMS_SHIFT = 11, + SPRAID_CC_SHN_SHIFT = 14, + SPRAID_CC_IOSQES_SHIFT = 16, + SPRAID_CC_IOCQES_SHIFT = 20, + SPRAID_CC_AMS_RR = 0 << SPRAID_CC_AMS_SHIFT, + SPRAID_CC_SHN_NONE = 0 << SPRAID_CC_SHN_SHIFT, + SPRAID_CC_IOSQES = SPRAID_IO_IOSQES << SPRAID_CC_IOSQES_SHIFT, + SPRAID_CC_IOCQES = SPRAID_IO_IOCQES << SPRAID_CC_IOCQES_SHIFT, + SPRAID_CC_SHN_NORMAL = 1 << SPRAID_CC_SHN_SHIFT, + SPRAID_CC_SHN_MASK = 3 << SPRAID_CC_SHN_SHIFT, + SPRAID_CSTS_CFS_SHIFT = 1, + SPRAID_CSTS_SHST_SHIFT = 2, + SPRAID_CSTS_PP_SHIFT = 5, + SPRAID_CSTS_RDY = 1 << 0, + SPRAID_CSTS_SHST_CMPLT = 2 << 2, + SPRAID_CSTS_SHST_MASK = 3 << 2, + SPRAID_CSTS_CFS_MASK = 1 << SPRAID_CSTS_CFS_SHIFT, + SPRAID_CSTS_PP_MASK = 1 << SPRAID_CSTS_PP_SHIFT, +}; + +enum { + SPRAID_ADMIN_DELETE_SQ = 0x00, + SPRAID_ADMIN_CREATE_SQ = 0x01, + SPRAID_ADMIN_DELETE_CQ = 0x04, + SPRAID_ADMIN_CREATE_CQ = 0x05, + SPRAID_ADMIN_ABORT_CMD = 0x08, + SPRAID_ADMIN_SET_FEATURES = 0x09, + SPRAID_ADMIN_ASYNC_EVENT = 0x0c, + SPRAID_ADMIN_GET_INFO = 0xc6, + SPRAID_ADMIN_RESET = 0xc8, +}; + +enum { + SPRAID_GET_INFO_CTRL = 0, + SPRAID_GET_INFO_DEV_LIST = 1, +}; + +enum { + SPRAID_RESET_TARGET = 0, + SPRAID_RESET_BUS = 1, +}; + +enum { + SPRAID_AEN_ERROR = 0, + SPRAID_AEN_NOTICE = 2, + SPRAID_AEN_VS = 7, +}; + +enum { + SPRAID_AEN_DEV_CHANGED = 0x00, + SPRAID_AEN_FW_ACT_START = 0x01, + SPRAID_AEN_HOST_PROBING = 0x10, +}; + +enum { + SPRAID_AEN_TIMESYN = 0x00, + SPRAID_AEN_FW_ACT_FINISH = 0x02, + SPRAID_AEN_EVENT_MIN = 0x80, + SPRAID_AEN_EVENT_MAX = 0xff, +}; + +enum { + SPRAID_CMD_WRITE = 0x01, + SPRAID_CMD_READ = 0x02, + + SPRAID_CMD_NONIO_NONE = 0x80, + SPRAID_CMD_NONIO_TODEV = 0x81, + SPRAID_CMD_NONIO_FROMDEV = 0x82, +}; + +enum { + SPRAID_QUEUE_PHYS_CONTIG = (1 << 0), + SPRAID_CQ_IRQ_ENABLED = (1 << 1), + + SPRAID_FEAT_NUM_QUEUES = 0x07, + SPRAID_FEAT_ASYNC_EVENT = 0x0b, + SPRAID_FEAT_TIMESTAMP = 0x0e, +}; + +enum spraid_state { + SPRAID_NEW, + SPRAID_LIVE, + SPRAID_RESETTING, + SPRAID_DELETING, + SPRAID_DEAD, +}; + +enum { + SPRAID_CARD_HBA, + SPRAID_CARD_RAID, +}; + +enum spraid_cmd_type { + SPRAID_CMD_ADM, + SPRAID_CMD_IOPT, +}; + +struct spraid_completion { + __le32 result; + union { + struct { + __u8 sense_len; + __u8 resv[3]; + }; + __le32 result1; + }; + __le16 sq_head; + __le16 sq_id; + __u16 cmd_id; + __le16 status; +}; + +struct spraid_ctrl_info { + __le32 nd; + __le16 max_cmds; + __le16 max_channel; + __le32 max_tgt_id; + __le16 max_lun; + __le16 max_num_sge; + __le16 lun_num_in_boot; + __u8 mdts; + __u8 acl; + __u8 aerl; + __u8 card_type; + __u16 rsvd; + __u32 rtd3e; + __u8 sn[32]; + __u8 fr[16]; + __u8 rsvd1[4020]; +}; + +struct spraid_dev { + struct pci_dev *pdev; + struct device *dev; + struct Scsi_Host *shost; + struct spraid_queue *queues; + struct dma_pool *prp_page_pool; + struct dma_pool *prp_small_pool[MAX_SMALL_POOL_NUM]; + mempool_t *iod_mempool; + void __iomem *bar; + u32 max_qid; + u32 num_vecs; + u32 queue_count; + u32 ioq_depth; + int db_stride; + u32 __iomem *dbs; + struct rw_semaphore devices_rwsem; + int numa_node; + u32 page_size; + u32 ctrl_config; + u32 online_queues; + u64 cap; + int instance; + struct spraid_ctrl_info *ctrl_info; + struct spraid_dev_info *devices; + + struct spraid_cmd *adm_cmds; + struct list_head adm_cmd_list; + spinlock_t adm_cmd_lock; + + struct spraid_cmd *ioq_ptcmds; + struct list_head ioq_pt_list; + spinlock_t ioq_pt_lock; + + struct work_struct scan_work; + struct work_struct timesyn_work; + struct work_struct reset_work; + struct work_struct fw_act_work; + + enum spraid_state state; + spinlock_t state_lock; + + struct request_queue *bsg_queue; +}; + +struct spraid_sgl_desc { + __le64 addr; + __le32 length; + __u8 rsvd[3]; + __u8 type; +}; + +union spraid_data_ptr { + struct { + __le64 prp1; + __le64 prp2; + }; + struct spraid_sgl_desc sgl; +}; + +struct spraid_admin_common_command { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __le32 cdw2[4]; + union spraid_data_ptr dptr; + __le32 cdw10; + __le32 cdw11; + __le32 cdw12; + __le32 cdw13; + __le32 cdw14; + __le32 cdw15; +}; + +struct spraid_features { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __u64 rsvd2[2]; + union spraid_data_ptr dptr; + __le32 fid; + __le32 dword11; + __le32 dword12; + __le32 dword13; + __le32 dword14; + __le32 dword15; +}; + +struct spraid_create_cq { + __u8 opcode; + __u8 flags; + __u16 command_id; + __u32 rsvd1[5]; + __le64 prp1; + __u64 rsvd8; + __le16 cqid; + __le16 qsize; + __le16 cq_flags; + __le16 irq_vector; + __u32 rsvd12[4]; +}; + +struct spraid_create_sq { + __u8 opcode; + __u8 flags; + __u16 command_id; + __u32 rsvd1[5]; + __le64 prp1; + __u64 rsvd8; + __le16 sqid; + __le16 qsize; + __le16 sq_flags; + __le16 cqid; + __u32 rsvd12[4]; +}; + +struct spraid_delete_queue { + __u8 opcode; + __u8 flags; + __u16 command_id; + __u32 rsvd1[9]; + __le16 qid; + __u16 rsvd10; + __u32 rsvd11[5]; +}; + +struct spraid_get_info { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __u32 rsvd2[4]; + union spraid_data_ptr dptr; + __u8 type; + __u8 rsvd10[3]; + __le32 cdw11; + __u32 rsvd12[4]; +}; + +struct spraid_usr_cmd { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + union { + struct { + __le16 subopcode; + __le16 rsvd1; + } info_0; + __le32 cdw2; + }; + union { + struct { + __le16 data_len; + __le16 param_len; + } info_1; + __le32 cdw3; + }; + __u64 metadata; + union spraid_data_ptr dptr; + __le32 cdw10; + __le32 cdw11; + __le32 cdw12; + __le32 cdw13; + __le32 cdw14; + __le32 cdw15; +}; + +enum { + SPRAID_CMD_FLAG_SGL_METABUF = (1 << 6), + SPRAID_CMD_FLAG_SGL_METASEG = (1 << 7), + SPRAID_CMD_FLAG_SGL_ALL = SPRAID_CMD_FLAG_SGL_METABUF | + SPRAID_CMD_FLAG_SGL_METASEG, +}; + +enum spraid_cmd_state { + SPRAID_CMD_IDLE = 0, + SPRAID_CMD_IN_FLIGHT = 1, + SPRAID_CMD_COMPLETE = 2, + SPRAID_CMD_TIMEOUT = 3, + SPRAID_CMD_TMO_COMPLETE = 4, +}; + +struct spraid_abort_cmd { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __u64 rsvd2[4]; + __le16 sqid; + __le16 cid; + __u32 rsvd11[5]; +}; + +struct spraid_reset_cmd { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __u64 rsvd2[4]; + __u8 type; + __u8 rsvd10[3]; + __u32 rsvd11[5]; +}; + +struct spraid_admin_command { + union { + struct spraid_admin_common_command common; + struct spraid_features features; + struct spraid_create_cq create_cq; + struct spraid_create_sq create_sq; + struct spraid_delete_queue delete_queue; + struct spraid_get_info get_info; + struct spraid_abort_cmd abort; + struct spraid_reset_cmd reset; + struct spraid_usr_cmd usr_cmd; + }; +}; + +struct spraid_ioq_common_command { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __le16 sense_len; + __u8 cdb_len; + __u8 rsvd2; + __le32 cdw3[3]; + union spraid_data_ptr dptr; + __le32 cdw10[6]; + __u8 cdb[32]; + __le64 sense_addr; + __le32 cdw26[6]; +}; + +struct spraid_rw_command { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __le16 sense_len; + __u8 cdb_len; + __u8 rsvd2; + __u32 rsvd3[3]; + union spraid_data_ptr dptr; + __le64 slba; + __le16 nlb; + __le16 control; + __u32 rsvd13[3]; + __u8 cdb[32]; + __le64 sense_addr; + __u32 rsvd26[6]; +}; + +struct spraid_scsi_nonio { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __le16 sense_len; + __u8 cdb_length; + __u8 rsvd2; + __u32 rsvd3[3]; + union spraid_data_ptr dptr; + __u32 rsvd10[5]; + __le32 buffer_len; + __u8 cdb[32]; + __le64 sense_addr; + __u32 rsvd26[6]; +}; + +struct spraid_ioq_command { + union { + struct spraid_ioq_common_command common; + struct spraid_rw_command rw; + struct spraid_scsi_nonio scsi_nonio; + }; +}; + +struct spraid_passthru_common_cmd { + __u8 opcode; + __u8 flags; + __u16 rsvd0; + __u32 nsid; + union { + struct { + __u16 subopcode; + __u16 rsvd1; + } info_0; + __u32 cdw2; + }; + union { + struct { + __u16 data_len; + __u16 param_len; + } info_1; + __u32 cdw3; + }; + __u64 metadata; + + __u64 addr; + __u64 prp2; + + __u32 cdw10; + __u32 cdw11; + __u32 cdw12; + __u32 cdw13; + __u32 cdw14; + __u32 cdw15; + __u32 timeout_ms; + __u32 result0; + __u32 result1; +}; + +struct spraid_ioq_passthru_cmd { + __u8 opcode; + __u8 flags; + __u16 rsvd0; + __u32 nsid; + union { + struct { + __u16 res_sense_len; + __u8 cdb_len; + __u8 rsvd0; + } info_0; + __u32 cdw2; + }; + union { + struct { + __u16 subopcode; + __u16 rsvd1; + } info_1; + __u32 cdw3; + }; + union { + struct { + __u16 rsvd; + __u16 param_len; + } info_2; + __u32 cdw4; + }; + __u32 cdw5; + __u64 addr; + __u64 prp2; + union { + struct { + __u16 eid; + __u16 sid; + } info_3; + __u32 cdw10; + }; + union { + struct { + __u16 did; + __u8 did_flag; + __u8 rsvd2; + } info_4; + __u32 cdw11; + }; + __u32 cdw12; + __u32 cdw13; + __u32 cdw14; + __u32 data_len; + __u32 cdw16; + __u32 cdw17; + __u32 cdw18; + __u32 cdw19; + __u32 cdw20; + __u32 cdw21; + __u32 cdw22; + __u32 cdw23; + __u64 sense_addr; + __u32 cdw26[4]; + __u32 timeout_ms; + __u32 result0; + __u32 result1; +}; + +struct spraid_bsg_request { + u32 msgcode; + u32 control; + union { + struct spraid_passthru_common_cmd admcmd; + struct spraid_ioq_passthru_cmd ioqcmd; + }; +}; + +enum { + SPRAID_BSG_ADM, + SPRAID_BSG_IOQ, +}; + +struct spraid_cmd { + int qid; + int cid; + u32 result0; + u32 result1; + u16 status; + void *priv; + enum spraid_cmd_state state; + struct completion cmd_done; + struct list_head list; +}; + +struct spraid_queue { + struct spraid_dev *hdev; + spinlock_t sq_lock; + + spinlock_t cq_lock ____cacheline_aligned_in_smp; + + void *sq_cmds; + + struct spraid_completion *cqes; + + dma_addr_t sq_dma_addr; + dma_addr_t cq_dma_addr; + u32 __iomem *q_db; + u8 cq_phase; + u8 sqes; + u16 qid; + u16 sq_tail; + u16 cq_head; + u16 last_cq_head; + u16 q_depth; + s16 cq_vector; + void *sense; + dma_addr_t sense_dma_addr; + struct dma_pool *prp_small_pool; +}; + +struct spraid_iod { + struct spraid_queue *spraidq; + enum spraid_cmd_state state; + int npages; + u32 nsge; + u32 length; + bool use_sgl; + bool sg_drv_mgmt; + dma_addr_t first_dma; + void *sense; + dma_addr_t sense_dma; + struct scatterlist *sg; + struct scatterlist inline_sg[0]; +}; + +#define SPRAID_DEV_INFO_ATTR_BOOT(attr) ((attr) & 0x01) +#define SPRAID_DEV_INFO_ATTR_VD(attr) (((attr) & 0x02) == 0x0) +#define SPRAID_DEV_INFO_ATTR_PT(attr) (((attr) & 0x22) == 0x02) +#define SPRAID_DEV_INFO_ATTR_RAWDISK(attr) ((attr) & 0x20) + +#define SPRAID_DEV_INFO_FLAG_VALID(flag) ((flag) & 0x01) +#define SPRAID_DEV_INFO_FLAG_CHANGE(flag) ((flag) & 0x02) + +#define BGTASK_TYPE_REBUILD 4 +#define USR_CMD_READ 0xc2 +#define USR_CMD_RDLEN 0x1000 +#define USR_CMD_VDINFO 0x704 +#define USR_CMD_BGTASK 0x504 +#define VDINFO_PARAM_LEN 0x04 + +struct spraid_vd_info { + __u8 name[32]; + __le16 id; + __u8 rg_id; + __u8 rg_level; + __u8 sg_num; + __u8 sg_disk_num; + __u8 vd_status; + __u8 vd_type; + __u8 rsvd1[4056]; +}; + +#define MAX_REALTIME_BGTASK_NUM 32 + +struct bgtask_info { + __u8 type; + __u8 progress; + __u8 rate; + __u8 rsvd0; + __le16 vd_id; + __le16 time_left; + __u8 rsvd1[4]; +}; + +struct spraid_bgtask { + __u8 sw; + __u8 task_num; + __u8 rsvd[6]; + struct bgtask_info bgtask[MAX_REALTIME_BGTASK_NUM]; +}; + +struct spraid_dev_info { + __le32 hdid; + __le16 target; + __u8 channel; + __u8 lun; + __u8 attr; + __u8 flag; + __le16 max_io_kb; +}; + +#define MAX_DEV_ENTRY_PER_PAGE_4K 340 +struct spraid_dev_list { + __le32 dev_num; + __u32 rsvd0[3]; + struct spraid_dev_info devices[MAX_DEV_ENTRY_PER_PAGE_4K]; +}; + +struct spraid_sdev_hostdata { + u32 hdid; + u16 max_io_kb; + u8 attr; + u8 flag; + u8 rg_id; + u8 rsvd[3]; +}; + +#endif + diff --git a/drivers/scsi/spraid/spraid_main.c b/drivers/scsi/spraid/spraid_main.c new file mode 100644 index 0000000000000..519b39f44e914 --- /dev/null +++ b/drivers/scsi/spraid/spraid_main.c @@ -0,0 +1,3875 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright(c) 2021 Ramaxel Memory Technology, Ltd */ + +/* Ramaxel Raid SPXXX Series Linux Driver */ + +#define pr_fmt(fmt) "spraid: " fmt + +#include <linux/sched/signal.h> +#include <linux/version.h> +#include <linux/pci.h> +#include <linux/aer.h> +#include <linux/module.h> +#include <linux/ioport.h> +#include <linux/device.h> +#include <linux/delay.h> +#include <linux/interrupt.h> +#include <linux/cdev.h> +#include <linux/sysfs.h> +#include <linux/gfp.h> +#include <linux/types.h> +#include <linux/ratelimit.h> +#include <linux/once.h> +#include <linux/debugfs.h> +#include <linux/io-64-nonatomic-lo-hi.h> +#include <linux/blkdev.h> +#include <linux/bsg-lib.h> +#include <asm/unaligned.h> +#include <linux/sort.h> +#include <target/target_core_backend.h> + +#include <scsi/scsi.h> +#include <scsi/scsi_cmnd.h> +#include <scsi/scsi_device.h> +#include <scsi/scsi_host.h> +#include <scsi/scsi_transport.h> +#include <scsi/scsi_dbg.h> + + +#include "spraid.h" + +static u32 admin_tmout = 60; +module_param(admin_tmout, uint, 0644); +MODULE_PARM_DESC(admin_tmout, "admin commands timeout (seconds)"); + +static u32 scmd_tmout_nonpt = 180; +module_param(scmd_tmout_nonpt, uint, 0644); +MODULE_PARM_DESC(scmd_tmout_nonpt, + "scsi commands timeout for rawdisk&raid(seconds)"); + +static int ioq_depth_set(const char *val, const struct kernel_param *kp); +static const struct kernel_param_ops ioq_depth_ops = { + .set = ioq_depth_set, + .get = param_get_uint, +}; + +static u32 io_queue_depth = 1024; +module_param_cb(io_queue_depth, &ioq_depth_ops, &io_queue_depth, 0644); +MODULE_PARM_DESC(io_queue_depth, "set io queue depth, should >= 2"); + +static int log_debug_switch_set(const char *val, const struct kernel_param *kp) +{ + u8 n = 0; + int ret; + + ret = kstrtou8(val, 10, &n); + if (ret != 0) + return -EINVAL; + + return param_set_byte(val, kp); +} + +static const struct kernel_param_ops log_debug_switch_ops = { + .set = log_debug_switch_set, + .get = param_get_byte, +}; + +static unsigned char log_debug_switch; +module_param_cb(log_debug_switch, &log_debug_switch_ops, + &log_debug_switch, 0644); +MODULE_PARM_DESC(log_debug_switch, + "set log state, default non-zero for switch on"); + +static int small_pool_num_set(const char *val, const struct kernel_param *kp) +{ + u8 n = 0; + int ret; + + ret = kstrtou8(val, 10, &n); + if (ret != 0) + return -EINVAL; + if (n > MAX_SMALL_POOL_NUM) + n = MAX_SMALL_POOL_NUM; + if (n < 1) + n = 1; + *((u8 *)kp->arg) = n; + + return 0; +} + +static const struct kernel_param_ops small_pool_num_ops = { + .set = small_pool_num_set, + .get = param_get_byte, +}; + +/* It was found that the spindlock of a single pool conflicts + * a lot with multiple CPUs.So multiple pools are introduced + * to reduce the conflictions. + */ +static unsigned char small_pool_num = 4; +module_param_cb(small_pool_num, &small_pool_num_ops, &small_pool_num, 0644); +MODULE_PARM_DESC(small_pool_num, "set prp small pool num, default 4, MAX 16"); + +static void spraid_free_queue(struct spraid_queue *spraidq); +static void spraid_handle_aen_notice(struct spraid_dev *hdev, u32 result); +static void spraid_handle_aen_vs(struct spraid_dev *hdev, + u32 result, u32 result1); + +static DEFINE_IDA(spraid_instance_ida); + +static struct class *spraid_class; + +#define SPRAID_CAP_TIMEOUT_UNIT_MS (HZ / 2) + +static struct workqueue_struct *spraid_wq; + +#define dev_log_dbg(dev, fmt, ...) do { \ + if (unlikely(log_debug_switch)) \ + dev_info(dev, "[%s] [%d] " fmt, \ + __func__, __LINE__, ##__VA_ARGS__); \ +} while (0) + +#define SPRAID_DRV_VERSION "1.0.0.0" + +#define ADMIN_TIMEOUT (admin_tmout * HZ) +#define ADMIN_ERR_TIMEOUT 32757 + +#define SPRAID_WAIT_ABNL_CMD_TIMEOUT (3 * 2) + +#define SPRAID_DMA_MSK_BIT_MAX 64 + +enum FW_STAT_CODE { + FW_STAT_OK = 0, + FW_STAT_NEED_CHECK, + FW_STAT_ERROR, + FW_STAT_EP_PCIE_ERROR, + FW_STAT_NAC_DMA_ERROR, + FW_STAT_ABORTED, + FW_STAT_NEED_RETRY +}; + +static const char * const raid_levels[] = {"0", "1", "5", "6", "10", "50", "60", + "NA"}; + +static const char * const raid_states[] = { + "NA", "NORMAL", "FAULT", "DEGRADE", "NOT_FORMATTED", + "FORMATTING", "SANITIZING", "INITIALIZING", "INITIALIZE_FAIL", + "DELETING", "DELETE_FAIL", "WRITE_PROTECT" +}; + +static int ioq_depth_set(const char *val, const struct kernel_param *kp) +{ + int n = 0; + int ret; + + ret = kstrtoint(val, 10, &n); + if (ret != 0 || n < 2) + return -EINVAL; + + return param_set_int(val, kp); +} + +static int spraid_remap_bar(struct spraid_dev *hdev, u32 size) +{ + struct pci_dev *pdev = hdev->pdev; + + if (size > pci_resource_len(pdev, 0)) { + dev_err(hdev->dev, "Input size[%u] exceed bar0 length[%llu]\n", + size, pci_resource_len(pdev, 0)); + return -ENOMEM; + } + + if (hdev->bar) + iounmap(hdev->bar); + + hdev->bar = ioremap(pci_resource_start(pdev, 0), size); + if (!hdev->bar) { + dev_err(hdev->dev, "ioremap for bar0 failed\n"); + return -ENOMEM; + } + hdev->dbs = hdev->bar + SPRAID_REG_DBS; + + return 0; +} + +static int spraid_dev_map(struct spraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + int ret; + + ret = pci_request_mem_regions(pdev, "spraid"); + if (ret) { + dev_err(hdev->dev, "fail to request memory regions\n"); + return ret; + } + + ret = spraid_remap_bar(hdev, SPRAID_REG_DBS + 4096); + if (ret) { + pci_release_mem_regions(pdev); + return ret; + } + + return 0; +} + +static void spraid_dev_unmap(struct spraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + + if (hdev->bar) { + iounmap(hdev->bar); + hdev->bar = NULL; + } + pci_release_mem_regions(pdev); +} + +static int spraid_pci_enable(struct spraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + int ret = -ENOMEM; + u64 maskbit = SPRAID_DMA_MSK_BIT_MAX; + + if (pci_enable_device_mem(pdev)) { + dev_err(hdev->dev, + "Enable pci device memory resources failed\n"); + return ret; + } + pci_set_master(pdev); + + if (readl(hdev->bar + SPRAID_REG_CSTS) == U32_MAX) { + ret = -ENODEV; + dev_err(hdev->dev, "Read csts register failed\n"); + goto disable; + } + + ret = pci_alloc_irq_vectors(pdev, 1, 1, PCI_IRQ_ALL_TYPES); + if (ret < 0) { + dev_err(hdev->dev, + "Allocate one IRQ for setup admin channel failed\n"); + goto disable; + } + + hdev->cap = lo_hi_readq(hdev->bar + SPRAID_REG_CAP); + hdev->ioq_depth = min_t(u32, SPRAID_CAP_MQES(hdev->cap) + 1, + io_queue_depth); + hdev->db_stride = 1 << SPRAID_CAP_STRIDE(hdev->cap); + + maskbit = SPRAID_CAP_DMAMASK(hdev->cap); + if (maskbit < 32 || maskbit > SPRAID_DMA_MSK_BIT_MAX) { + dev_err(hdev->dev, + "err, dma mask invalid[%llu], set to default\n", + maskbit); + maskbit = SPRAID_DMA_MSK_BIT_MAX; + } + if (dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(maskbit))) { + dev_err(hdev->dev, "set dma mask and coherent failed\n"); + goto disable; + } + + dev_info(hdev->dev, "set dma mask[%llu] success\n", maskbit); + + pci_enable_pcie_error_reporting(pdev); + pci_save_state(pdev); + + return 0; + +disable: + pci_disable_device(pdev); + return ret; +} + +static int spraid_npages_prp(u32 size, struct spraid_dev *hdev) +{ + u32 nprps = DIV_ROUND_UP(size + hdev->page_size, hdev->page_size); + + return DIV_ROUND_UP(PRP_ENTRY_SIZE * nprps, PAGE_SIZE - PRP_ENTRY_SIZE); +} + +static int spraid_npages_sgl(u32 nseg) +{ + return DIV_ROUND_UP(nseg * sizeof(struct spraid_sgl_desc), PAGE_SIZE); +} + +static void **spraid_iod_list(struct spraid_iod *iod) +{ + return (void **)(iod->inline_sg + (iod->sg_drv_mgmt ? iod->nsge : 0)); +} + +static u32 spraid_iod_ext_size(struct spraid_dev *hdev, u32 size, u32 nsge, + bool sg_drv_mgmt, bool use_sgl) +{ + size_t alloc_size, sg_size; + + if (use_sgl) + alloc_size = sizeof(__le64 *) * spraid_npages_sgl(nsge); + else + alloc_size = sizeof(__le64 *) * spraid_npages_prp(size, hdev); + + sg_size = sg_drv_mgmt ? (sizeof(struct scatterlist) * nsge) : 0; + return sg_size + alloc_size; +} + +static u32 spraid_cmd_size(struct spraid_dev *hdev, + bool sg_drv_mgmt, bool use_sgl) +{ + u32 alloc_size = spraid_iod_ext_size(hdev, SPRAID_INT_BYTES(hdev), + SPRAID_INT_PAGES, sg_drv_mgmt, use_sgl); + + dev_info(hdev->dev, "sg_drv_mgmt: %s, use_sgl: %s, iod size: %lu;" + " alloc_size: %u\n", sg_drv_mgmt ? "true" : "false", + use_sgl ? "true" : "false", + sizeof(struct spraid_iod), alloc_size); + + return sizeof(struct spraid_iod) + alloc_size; +} + +static int spraid_setup_prps(struct spraid_dev *hdev, struct spraid_iod *iod) +{ + struct scatterlist *sg = iod->sg; + u64 dma_addr = sg_dma_address(sg); + int dma_len = sg_dma_len(sg); + __le64 *prp_list, *old_prp_list; + u32 page_size = hdev->page_size; + int offset = dma_addr & (page_size - 1); + void **list = spraid_iod_list(iod); + int length = iod->length; + struct dma_pool *pool; + dma_addr_t prp_dma; + int nprps, i; + + length -= (page_size - offset); + if (length <= 0) { + iod->first_dma = 0; + return 0; + } + + dma_len -= (page_size - offset); + if (dma_len) { + dma_addr += (page_size - offset); + } else { + sg = sg_next(sg); + dma_addr = sg_dma_address(sg); + dma_len = sg_dma_len(sg); + } + + if (length <= page_size) { + iod->first_dma = dma_addr; + return 0; + } + + nprps = DIV_ROUND_UP(length, page_size); + if (nprps <= (SMALL_POOL_SIZE / PRP_ENTRY_SIZE)) { + pool = iod->spraidq->prp_small_pool; + iod->npages = 0; + } else { + pool = hdev->prp_page_pool; + iod->npages = 1; + } + + prp_list = dma_pool_alloc(pool, GFP_ATOMIC, &prp_dma); + if (!prp_list) { + dev_err_ratelimited(hdev->dev, + "Allocate first prp_list memory failed\n"); + iod->first_dma = dma_addr; + iod->npages = -1; + return -ENOMEM; + } + list[0] = prp_list; + iod->first_dma = prp_dma; + i = 0; + for (;;) { + if (i == page_size / PRP_ENTRY_SIZE) { + old_prp_list = prp_list; + + prp_list = dma_pool_alloc(pool, GFP_ATOMIC, &prp_dma); + if (!prp_list) { + dev_err_ratelimited(hdev->dev, "Allocate %dth;" + " prp_list memory failed\n", + iod->npages + 1); + return -ENOMEM; + } + list[iod->npages++] = prp_list; + prp_list[0] = old_prp_list[i - 1]; + old_prp_list[i - 1] = cpu_to_le64(prp_dma); + i = 1; + } + prp_list[i++] = cpu_to_le64(dma_addr); + dma_len -= page_size; + dma_addr += page_size; + length -= page_size; + if (length <= 0) + break; + if (dma_len > 0) + continue; + if (unlikely(dma_len < 0)) + goto bad_sgl; + sg = sg_next(sg); + dma_addr = sg_dma_address(sg); + dma_len = sg_dma_len(sg); + } + + return 0; + +bad_sgl: + dev_err(hdev->dev, + "Setup prps, invalid SGL for payload: %d nents: %d\n", + iod->length, iod->nsge); + return -EIO; +} + +#define SGES_PER_PAGE (PAGE_SIZE / sizeof(struct spraid_sgl_desc)) + +static void spraid_submit_cmd(struct spraid_queue *spraidq, const void *cmd) +{ + u32 sqes = SQE_SIZE(spraidq->qid); + unsigned long flags; + struct spraid_admin_common_command *acd = + (struct spraid_admin_common_command *)cmd; + + spin_lock_irqsave(&spraidq->sq_lock, flags); + memcpy((spraidq->sq_cmds + sqes * spraidq->sq_tail), cmd, sqes); + if (++spraidq->sq_tail == spraidq->q_depth) + spraidq->sq_tail = 0; + + writel(spraidq->sq_tail, spraidq->q_db); + spin_unlock_irqrestore(&spraidq->sq_lock, flags); + + dev_log_dbg(spraidq->hdev->dev, + "cid[%d] qid[%d], opcode[0x%x], flags[0x%x], hdid[%u]\n", + acd->command_id, spraidq->qid, acd->opcode, + acd->flags, le32_to_cpu(acd->hdid)); +} + +static u32 spraid_mod64(u64 dividend, u32 divisor) +{ + u64 d; + u32 remainder; + + if (!divisor) + pr_err("DIVISOR is zero, in div fn\n"); + + d = dividend; + remainder = do_div(d, divisor); + return remainder; +} + +static inline bool spraid_is_rw_scmd(struct scsi_cmnd *scmd) +{ + switch (scmd->cmnd[0]) { + case READ_6: + case READ_10: + case READ_12: + case READ_16: + case READ_32: + case WRITE_6: + case WRITE_10: + case WRITE_12: + case WRITE_16: + case WRITE_32: + return true; + default: + return false; + } +} + +static bool spraid_is_prp(struct spraid_dev *hdev, + struct scsi_cmnd *scmd, u32 nsge) +{ + struct scatterlist *sg = scsi_sglist(scmd); + u32 page_size = hdev->page_size; + bool is_prp = true; + int i = 0; + + scsi_for_each_sg(scmd, sg, nsge, i) { + if (i != 0 && i != nsge - 1) { + if (spraid_mod64(sg_dma_len(sg), page_size) || + spraid_mod64(sg_dma_address(sg), page_size)) { + is_prp = false; + break; + } + } + + if (nsge > 1 && i == 0) { + if ((spraid_mod64((sg_dma_address(sg) + sg_dma_len(sg)), + page_size))) { + is_prp = false; + break; + } + } + + if (nsge > 1 && i == (nsge - 1)) { + if (spraid_mod64(sg_dma_address(sg), page_size)) { + is_prp = false; + break; + } + } + } + + return is_prp; +} + +enum { + SPRAID_SGL_FMT_DATA_DESC = 0x00, + SPRAID_SGL_FMT_SEG_DESC = 0x02, + SPRAID_SGL_FMT_LAST_SEG_DESC = 0x03, + SPRAID_KEY_SGL_FMT_DATA_DESC = 0x04, + SPRAID_TRANSPORT_SGL_DATA_DESC = 0x05 +}; + +static void spraid_sgl_set_data(struct spraid_sgl_desc *sge, + struct scatterlist *sg) +{ + sge->addr = cpu_to_le64(sg_dma_address(sg)); + sge->length = cpu_to_le32(sg_dma_len(sg)); + sge->type = SPRAID_SGL_FMT_DATA_DESC << 4; +} + +static void spraid_sgl_set_seg(struct spraid_sgl_desc *sge, + dma_addr_t dma_addr, int entries) +{ + sge->addr = cpu_to_le64(dma_addr); + if (entries <= SGES_PER_PAGE) { + sge->length = cpu_to_le32(entries * sizeof(*sge)); + sge->type = SPRAID_SGL_FMT_LAST_SEG_DESC << 4; + } else { + sge->length = cpu_to_le32(PAGE_SIZE); + sge->type = SPRAID_SGL_FMT_SEG_DESC << 4; + } +} + +static int spraid_setup_ioq_cmd_sgl(struct spraid_dev *hdev, + struct scsi_cmnd *scmd, + struct spraid_ioq_command *ioq_cmd, + struct spraid_iod *iod) +{ + struct spraid_sgl_desc *sg_list, *link, *old_sg_list; + struct scatterlist *sg = scsi_sglist(scmd); + void **list = spraid_iod_list(iod); + struct dma_pool *pool; + int nsge = iod->nsge; + dma_addr_t sgl_dma; + int i = 0; + + ioq_cmd->common.flags |= SPRAID_CMD_FLAG_SGL_METABUF; + + if (nsge == 1) { + spraid_sgl_set_data(&ioq_cmd->common.dptr.sgl, sg); + return 0; + } + + if (nsge <= (SMALL_POOL_SIZE / sizeof(struct spraid_sgl_desc))) { + pool = iod->spraidq->prp_small_pool; + iod->npages = 0; + } else { + pool = hdev->prp_page_pool; + iod->npages = 1; + } + + sg_list = dma_pool_alloc(pool, GFP_ATOMIC, &sgl_dma); + if (!sg_list) { + dev_err_ratelimited(hdev->dev, + "Allocate first sgl_list failed\n"); + iod->npages = -1; + return -ENOMEM; + } + + list[0] = sg_list; + iod->first_dma = sgl_dma; + spraid_sgl_set_seg(&ioq_cmd->common.dptr.sgl, sgl_dma, nsge); + do { + if (i == SGES_PER_PAGE) { + old_sg_list = sg_list; + link = &old_sg_list[SGES_PER_PAGE - 1]; + + sg_list = dma_pool_alloc(pool, GFP_ATOMIC, &sgl_dma); + if (!sg_list) { + dev_err_ratelimited(hdev->dev, + "Allocate %dth sgl_list;" + " failed\n", + iod->npages + 1); + return -ENOMEM; + } + list[iod->npages++] = sg_list; + + i = 0; + memcpy(&sg_list[i++], link, sizeof(*link)); + spraid_sgl_set_seg(link, sgl_dma, nsge); + } + + spraid_sgl_set_data(&sg_list[i++], sg); + sg = sg_next(sg); + } while (--nsge > 0); + + return 0; +} + +#define SPRAID_RW_FUA BIT(14) + +static void spraid_setup_rw_cmd(struct spraid_dev *hdev, + struct spraid_rw_command *rw, + struct scsi_cmnd *scmd) +{ + u32 start_lba_lo, start_lba_hi; + u32 datalength = 0; + u16 control = 0; + + start_lba_lo = 0; + start_lba_hi = 0; + + if (scmd->sc_data_direction == DMA_TO_DEVICE) { + rw->opcode = SPRAID_CMD_WRITE; + } else if (scmd->sc_data_direction == DMA_FROM_DEVICE) { + rw->opcode = SPRAID_CMD_READ; + } else { + dev_err(hdev->dev, + "Invalid IO for unsupported data direction: %d\n", + scmd->sc_data_direction); + WARN_ON(1); + } + + /* 6-byte READ(0x08) or WRITE(0x0A) cdb */ + if (scmd->cmd_len == 6) { + datalength = (u32)(scmd->cmnd[4] == 0 ? + IO_6_DEFAULT_TX_LEN : scmd->cmnd[4]); + start_lba_lo = (u32)get_unaligned_be24(&scmd->cmnd[1]); + + start_lba_lo &= 0x1FFFFF; + } + + /* 10-byte READ(0x28) or WRITE(0x2A) cdb */ + else if (scmd->cmd_len == 10) { + datalength = (u32)get_unaligned_be16(&scmd->cmnd[7]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[2]); + + if (scmd->cmnd[1] & FUA_MASK) + control |= SPRAID_RW_FUA; + } + + /* 12-byte READ(0xA8) or WRITE(0xAA) cdb */ + else if (scmd->cmd_len == 12) { + datalength = get_unaligned_be32(&scmd->cmnd[6]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[2]); + + if (scmd->cmnd[1] & FUA_MASK) + control |= SPRAID_RW_FUA; + } + /* 16-byte READ(0x88) or WRITE(0x8A) cdb */ + else if (scmd->cmd_len == 16) { + datalength = get_unaligned_be32(&scmd->cmnd[10]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[6]); + start_lba_hi = get_unaligned_be32(&scmd->cmnd[2]); + + if (scmd->cmnd[1] & FUA_MASK) + control |= SPRAID_RW_FUA; + } + /* 32-byte READ(0x88) or WRITE(0x8A) cdb */ + else if (scmd->cmd_len == 32) { + datalength = get_unaligned_be32(&scmd->cmnd[28]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[16]); + start_lba_hi = get_unaligned_be32(&scmd->cmnd[12]); + + if (scmd->cmnd[10] & FUA_MASK) + control |= SPRAID_RW_FUA; + } + + if (unlikely(datalength > U16_MAX || datalength == 0)) { + dev_err(hdev->dev, + "Invalid IO for illegal transfer data length: %u\n", + datalength); + WARN_ON(1); + } + + rw->slba = cpu_to_le64(((u64)start_lba_hi << 32) | start_lba_lo); + /* 0base for nlb */ + rw->nlb = cpu_to_le16((u16)(datalength - 1)); + rw->control = cpu_to_le16(control); +} + +static void spraid_setup_nonio_cmd(struct spraid_dev *hdev, + struct spraid_scsi_nonio *scsi_nonio, + struct scsi_cmnd *scmd) +{ + scsi_nonio->buffer_len = cpu_to_le32(scsi_bufflen(scmd)); + + switch (scmd->sc_data_direction) { + case DMA_NONE: + scsi_nonio->opcode = SPRAID_CMD_NONIO_NONE; + break; + case DMA_TO_DEVICE: + scsi_nonio->opcode = SPRAID_CMD_NONIO_TODEV; + break; + case DMA_FROM_DEVICE: + scsi_nonio->opcode = SPRAID_CMD_NONIO_FROMDEV; + break; + default: + dev_err(hdev->dev, + "Invalid IO for unsupported data direction: %d\n", + scmd->sc_data_direction); + WARN_ON(1); + } +} + +static void spraid_setup_ioq_cmd(struct spraid_dev *hdev, + struct spraid_ioq_command *ioq_cmd, + struct scsi_cmnd *scmd) +{ + memcpy(ioq_cmd->common.cdb, scmd->cmnd, scmd->cmd_len); + ioq_cmd->common.cdb_len = scmd->cmd_len; + + if (spraid_is_rw_scmd(scmd)) + spraid_setup_rw_cmd(hdev, &ioq_cmd->rw, scmd); + else + spraid_setup_nonio_cmd(hdev, &ioq_cmd->scsi_nonio, scmd); +} + +static int spraid_init_iod(struct spraid_dev *hdev, struct spraid_iod *iod, + struct spraid_ioq_command *ioq_cmd, + struct scsi_cmnd *scmd) +{ + if (unlikely(!iod->sense)) { + dev_err(hdev->dev, "Allocate sense data buffer failed\n"); + return -ENOMEM; + } + ioq_cmd->common.sense_addr = cpu_to_le64(iod->sense_dma); + ioq_cmd->common.sense_len = cpu_to_le16(SCSI_SENSE_BUFFERSIZE); + + iod->nsge = 0; + iod->npages = -1; + iod->use_sgl = 0; + iod->sg_drv_mgmt = false; + WRITE_ONCE(iod->state, SPRAID_CMD_IDLE); + + return 0; +} + +static void spraid_free_iod_res(struct spraid_dev *hdev, struct spraid_iod *iod) +{ + const int last_prp = hdev->page_size / sizeof(__le64) - 1; + dma_addr_t dma_addr, next_dma_addr; + struct spraid_sgl_desc *sg_list; + __le64 *prp_list; + void *addr; + int i; + + dma_addr = iod->first_dma; + if (iod->npages == 0) + dma_pool_free(iod->spraidq->prp_small_pool, + spraid_iod_list(iod)[0], dma_addr); + + for (i = 0; i < iod->npages; i++) { + addr = spraid_iod_list(iod)[i]; + + if (iod->use_sgl) { + sg_list = addr; + next_dma_addr = + le64_to_cpu((sg_list[SGES_PER_PAGE - 1]).addr); + } else { + prp_list = addr; + next_dma_addr = le64_to_cpu(prp_list[last_prp]); + } + + dma_pool_free(hdev->prp_page_pool, addr, dma_addr); + dma_addr = next_dma_addr; + } + + if (iod->sg_drv_mgmt && iod->sg != iod->inline_sg) { + iod->sg_drv_mgmt = false; + mempool_free(iod->sg, hdev->iod_mempool); + } + + iod->sense = NULL; + iod->npages = -1; +} + +static int spraid_io_map_data(struct spraid_dev *hdev, struct spraid_iod *iod, + struct scsi_cmnd *scmd, + struct spraid_ioq_command *ioq_cmd) +{ + int ret; + + iod->nsge = scsi_dma_map(scmd); + + /* No data to DMA, it may be scsi no-rw command */ + if (unlikely(iod->nsge == 0)) + return 0; + + iod->length = scsi_bufflen(scmd); + iod->sg = scsi_sglist(scmd); + iod->use_sgl = !spraid_is_prp(hdev, scmd, iod->nsge); + + if (iod->use_sgl) { + ret = spraid_setup_ioq_cmd_sgl(hdev, scmd, ioq_cmd, iod); + } else { + ret = spraid_setup_prps(hdev, iod); + ioq_cmd->common.dptr.prp1 = + cpu_to_le64(sg_dma_address(iod->sg)); + ioq_cmd->common.dptr.prp2 = cpu_to_le64(iod->first_dma); + } + + if (ret) + scsi_dma_unmap(scmd); + + return ret; +} + +static void spraid_map_status(struct spraid_iod *iod, struct scsi_cmnd *scmd, + struct spraid_completion *cqe) +{ + scsi_set_resid(scmd, 0); + + switch ((le16_to_cpu(cqe->status) >> 1) & 0x7f) { + case FW_STAT_OK: + set_host_byte(scmd, DID_OK); + break; + case FW_STAT_NEED_CHECK: + set_host_byte(scmd, DID_OK); + scmd->result |= le16_to_cpu(cqe->status) >> 8; + if (scmd->result & SAM_STAT_CHECK_CONDITION) { + memset(scmd->sense_buffer, 0, SCSI_SENSE_BUFFERSIZE); + memcpy(scmd->sense_buffer, iod->sense, + SCSI_SENSE_BUFFERSIZE); + scmd->result = + (scmd->result & 0x00ffffff) | (DRIVER_SENSE << 24); + } + break; + case FW_STAT_ABORTED: + set_host_byte(scmd, DID_ABORT); + break; + case FW_STAT_NEED_RETRY: + set_host_byte(scmd, DID_REQUEUE); + break; + default: + set_host_byte(scmd, DID_BAD_TARGET); + break; + } +} + +static inline void spraid_get_tag_from_scmd(struct scsi_cmnd *scmd, + u16 *qid, u16 *cid) +{ + u32 tag = blk_mq_unique_tag(scmd->request); + + *qid = blk_mq_unique_tag_to_hwq(tag) + 1; + *cid = blk_mq_unique_tag_to_tag(tag); +} + +static int spraid_queue_command(struct Scsi_Host *shost, struct scsi_cmnd *scmd) +{ + struct spraid_iod *iod = scsi_cmd_priv(scmd); + struct spraid_dev *hdev = shost_priv(shost); + struct scsi_device *sdev = scmd->device; + struct spraid_sdev_hostdata *hostdata; + struct spraid_ioq_command ioq_cmd; + struct spraid_queue *ioq; + unsigned long elapsed; + u16 hwq, cid; + int ret; + + if (unlikely(!scmd)) { + dev_err(hdev->dev, "err, scmd is null\n"); + return 0; + } + + if (unlikely(hdev->state != SPRAID_LIVE)) { + set_host_byte(scmd, DID_NO_CONNECT); + scmd->scsi_done(scmd); + return 0; + } + + if (log_debug_switch) + scsi_print_command(scmd); + + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + hostdata = sdev->hostdata; + ioq = &hdev->queues[hwq]; + memset(&ioq_cmd, 0, sizeof(ioq_cmd)); + ioq_cmd.rw.hdid = cpu_to_le32(hostdata->hdid); + ioq_cmd.rw.command_id = cid; + + spraid_setup_ioq_cmd(hdev, &ioq_cmd, scmd); + + ret = cid * SCSI_SENSE_BUFFERSIZE; + iod->sense = ioq->sense + ret; + iod->sense_dma = ioq->sense_dma_addr + ret; + + ret = spraid_init_iod(hdev, iod, &ioq_cmd, scmd); + if (unlikely(ret)) + return SCSI_MLQUEUE_HOST_BUSY; + + iod->spraidq = ioq; + ret = spraid_io_map_data(hdev, iod, scmd, &ioq_cmd); + if (unlikely(ret)) { + dev_err(hdev->dev, "spraid_io_map_data Err.\n"); + set_host_byte(scmd, DID_ERROR); + scmd->scsi_done(scmd); + ret = 0; + goto deinit_iod; + } + + WRITE_ONCE(iod->state, SPRAID_CMD_IN_FLIGHT); + spraid_submit_cmd(ioq, &ioq_cmd); + elapsed = jiffies - scmd->jiffies_at_alloc; + dev_log_dbg(hdev->dev, + "cid[%d] qid[%d] submit IO cost %3ld.%3ld seconds\n", + cid, hwq, elapsed / HZ, elapsed % HZ); + return 0; + +deinit_iod: + spraid_free_iod_res(hdev, iod); + return ret; +} + +static int spraid_match_dev(struct spraid_dev *hdev, u16 idx, + struct scsi_device *sdev) +{ + if (SPRAID_DEV_INFO_FLAG_VALID(hdev->devices[idx].flag)) { + if (sdev->channel == hdev->devices[idx].channel && + sdev->id == le16_to_cpu(hdev->devices[idx].target) && + sdev->lun < hdev->devices[idx].lun) { + dev_info(hdev->dev, + "Match device success, channel;" + "target:lun[%d:%d:%d]\n", + hdev->devices[idx].channel, + hdev->devices[idx].target, + hdev->devices[idx].lun); + return 1; + } + } + + return 0; +} + +static int spraid_slave_alloc(struct scsi_device *sdev) +{ + struct spraid_sdev_hostdata *hostdata; + struct spraid_dev *hdev; + u16 idx; + + hdev = shost_priv(sdev->host); + hostdata = kzalloc(sizeof(*hostdata), GFP_KERNEL); + if (!hostdata) { + dev_err(hdev->dev, "Alloc scsi host data memory failed\n"); + return -ENOMEM; + } + + down_read(&hdev->devices_rwsem); + for (idx = 0; idx < le32_to_cpu(hdev->ctrl_info->nd); idx++) { + if (spraid_match_dev(hdev, idx, sdev)) + goto scan_host; + } + up_read(&hdev->devices_rwsem); + + kfree(hostdata); + return -ENXIO; + +scan_host: + hostdata->hdid = le32_to_cpu(hdev->devices[idx].hdid); + hostdata->max_io_kb = le16_to_cpu(hdev->devices[idx].max_io_kb); + hostdata->attr = hdev->devices[idx].attr; + hostdata->flag = hdev->devices[idx].flag; + hostdata->rg_id = 0xff; + sdev->hostdata = hostdata; + up_read(&hdev->devices_rwsem); + return 0; +} + +static void spraid_slave_destroy(struct scsi_device *sdev) +{ + kfree(sdev->hostdata); + sdev->hostdata = NULL; +} + +static int spraid_slave_configure(struct scsi_device *sdev) +{ + u16 idx; + unsigned int timeout = scmd_tmout_nonpt * HZ; + struct spraid_dev *hdev = shost_priv(sdev->host); + struct spraid_sdev_hostdata *hostdata = sdev->hostdata; + u32 max_sec = sdev->host->max_sectors; + + if (hostdata) { + idx = hostdata->hdid - 1; + if (sdev->channel == hdev->devices[idx].channel && + sdev->id == le16_to_cpu(hdev->devices[idx].target) && + sdev->lun < hdev->devices[idx].lun) { + if (SPRAID_DEV_INFO_ATTR_PT(hdev->devices[idx].attr)) + timeout = 30 * HZ; + else + timeout = scmd_tmout_nonpt * HZ; + max_sec = le16_to_cpu(hdev->devices[idx].max_io_kb) + << 1; + } else { + dev_err(hdev->dev, + "[%s] err, sdev->channel:id:lun[%d:%d:%lld];" + "devices[%d], channel:target:lun[%d:%d:%d]\n", + __func__, sdev->channel, sdev->id, sdev->lun, + idx, hdev->devices[idx].channel, + hdev->devices[idx].target, + hdev->devices[idx].lun); + } + } else { + dev_err(hdev->dev, "[%s] err, sdev->hostdata is null\n", + __func__); + } + + blk_queue_rq_timeout(sdev->request_queue, timeout); + sdev->eh_timeout = timeout; + + if ((max_sec == 0) || (max_sec > sdev->host->max_sectors)) + max_sec = sdev->host->max_sectors; + + dev_info(hdev->dev, + "[%s] sdev->channel:id:lun[%d:%d:%lld];" + " scmd_timeout[%d]s, maxsec[%d]\n", + __func__, sdev->channel, sdev->id, + sdev->lun, timeout / HZ, max_sec); + + return 0; +} + +static void spraid_shost_init(struct spraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + u8 domain, bus; + u32 dev_func; + + domain = pci_domain_nr(pdev->bus); + bus = pdev->bus->number; + dev_func = pdev->devfn; + + hdev->shost->nr_hw_queues = hdev->online_queues - 1; + hdev->shost->can_queue = (hdev->ioq_depth - SPRAID_PTCMDS_PERQ); + + hdev->shost->sg_tablesize = le16_to_cpu(hdev->ctrl_info->max_num_sge); + /* 512B per sector */ + hdev->shost->max_sectors = + (1U << ((hdev->ctrl_info->mdts) * 1U) << 12) / 512; + hdev->shost->cmd_per_lun = MAX_CMD_PER_DEV; + hdev->shost->max_channel = + le16_to_cpu(hdev->ctrl_info->max_channel) - 1; + hdev->shost->max_id = le32_to_cpu(hdev->ctrl_info->max_tgt_id); + hdev->shost->max_lun = le16_to_cpu(hdev->ctrl_info->max_lun); + + hdev->shost->this_id = -1; + hdev->shost->unique_id = (domain << 16) | (bus << 8) | dev_func; + hdev->shost->max_cmd_len = MAX_CDB_LEN; + hdev->shost->hostt->cmd_size = max(spraid_cmd_size(hdev, false, true), + spraid_cmd_size(hdev, false, false)); +} + +static inline void spraid_host_deinit(struct spraid_dev *hdev) +{ + ida_free(&spraid_instance_ida, hdev->instance); +} + +static int spraid_alloc_queue(struct spraid_dev *hdev, u16 qid, u16 depth) +{ + struct spraid_queue *spraidq = &hdev->queues[qid]; + int ret = 0; + + if (hdev->queue_count > qid) { + dev_info(hdev->dev, "[%s] warn: queue[%d] is exist\n", + __func__, qid); + return 0; + } + + spraidq->cqes = dma_alloc_coherent(hdev->dev, CQ_SIZE(depth), + &spraidq->cq_dma_addr, + GFP_KERNEL | __GFP_ZERO); + if (!spraidq->cqes) + return -ENOMEM; + + spraidq->sq_cmds = dma_alloc_coherent(hdev->dev, SQ_SIZE(qid, depth), + &spraidq->sq_dma_addr, + GFP_KERNEL); + if (!spraidq->sq_cmds) { + ret = -ENOMEM; + goto free_cqes; + } + + spin_lock_init(&spraidq->sq_lock); + spin_lock_init(&spraidq->cq_lock); + spraidq->hdev = hdev; + spraidq->q_depth = depth; + spraidq->qid = qid; + spraidq->cq_vector = -1; + hdev->queue_count++; + + /* alloc sense buffer */ + spraidq->sense = dma_alloc_coherent(hdev->dev, SENSE_SIZE(depth), + &spraidq->sense_dma_addr, + GFP_KERNEL | __GFP_ZERO); + if (!spraidq->sense) { + ret = -ENOMEM; + goto free_sq_cmds; + } + + return 0; + +free_sq_cmds: + dma_free_coherent(hdev->dev, SQ_SIZE(qid, depth), + (void *)spraidq->sq_cmds, spraidq->sq_dma_addr); +free_cqes: + dma_free_coherent(hdev->dev, CQ_SIZE(depth), (void *)spraidq->cqes, + spraidq->cq_dma_addr); + return ret; +} + +static int spraid_wait_ready(struct spraid_dev *hdev, u64 cap, bool enabled) +{ + unsigned long timeout = + ((SPRAID_CAP_TIMEOUT(cap) + 1) * SPRAID_CAP_TIMEOUT_UNIT_MS) + jiffies; + u32 bit = enabled ? SPRAID_CSTS_RDY : 0; + + while ((readl(hdev->bar + SPRAID_REG_CSTS) & SPRAID_CSTS_RDY) != bit) { + usleep_range(1000, 2000); + if (fatal_signal_pending(current)) + return -EINTR; + + if (time_after(jiffies, timeout)) { + dev_err(hdev->dev, "Device not ready; aborting %s\n", + enabled ? "initialisation" : "reset"); + return -ENODEV; + } + } + return 0; +} + +static int spraid_shutdown_ctrl(struct spraid_dev *hdev) +{ + unsigned long timeout = hdev->ctrl_info->rtd3e + jiffies; + + hdev->ctrl_config &= ~SPRAID_CC_SHN_MASK; + hdev->ctrl_config |= SPRAID_CC_SHN_NORMAL; + writel(hdev->ctrl_config, hdev->bar + SPRAID_REG_CC); + + while ((readl(hdev->bar + SPRAID_REG_CSTS) & SPRAID_CSTS_SHST_MASK) != + SPRAID_CSTS_SHST_CMPLT) { + msleep(100); + if (fatal_signal_pending(current)) + return -EINTR; + if (time_after(jiffies, timeout)) { + dev_err(hdev->dev, + "Device shutdown incomplete; abort shutdown\n"); + return -ENODEV; + } + } + return 0; +} + +static int spraid_disable_ctrl(struct spraid_dev *hdev) +{ + hdev->ctrl_config &= ~SPRAID_CC_SHN_MASK; + hdev->ctrl_config &= ~SPRAID_CC_ENABLE; + writel(hdev->ctrl_config, hdev->bar + SPRAID_REG_CC); + + return spraid_wait_ready(hdev, hdev->cap, false); +} + +static int spraid_enable_ctrl(struct spraid_dev *hdev) +{ + u64 cap = hdev->cap; + u32 dev_page_min = SPRAID_CAP_MPSMIN(cap) + 12; + u32 page_shift = PAGE_SHIFT; + + if (page_shift < dev_page_min) { + dev_err(hdev->dev, + "Minimum device page size[%u], too large for host[%u]\n", + 1U << dev_page_min, 1U << page_shift); + return -ENODEV; + } + + page_shift = min_t(unsigned int, SPRAID_CAP_MPSMAX(cap) + 12, + PAGE_SHIFT); + hdev->page_size = 1U << page_shift; + + hdev->ctrl_config = SPRAID_CC_CSS_NVM; + hdev->ctrl_config |= (page_shift - 12) << SPRAID_CC_MPS_SHIFT; + hdev->ctrl_config |= SPRAID_CC_AMS_RR | SPRAID_CC_SHN_NONE; + hdev->ctrl_config |= SPRAID_CC_IOSQES | SPRAID_CC_IOCQES; + hdev->ctrl_config |= SPRAID_CC_ENABLE; + writel(hdev->ctrl_config, hdev->bar + SPRAID_REG_CC); + + return spraid_wait_ready(hdev, cap, true); +} + +static void spraid_init_queue(struct spraid_queue *spraidq, u16 qid) +{ + struct spraid_dev *hdev = spraidq->hdev; + + memset((void *)spraidq->cqes, 0, CQ_SIZE(spraidq->q_depth)); + + spraidq->sq_tail = 0; + spraidq->cq_head = 0; + spraidq->cq_phase = 1; + spraidq->q_db = &hdev->dbs[qid * 2 * hdev->db_stride]; + spraidq->prp_small_pool = hdev->prp_small_pool[qid % small_pool_num]; + hdev->online_queues++; +} + +static inline bool spraid_cqe_pending(struct spraid_queue *spraidq) +{ + return (le16_to_cpu(spraidq->cqes[spraidq->cq_head].status) & 1) == + spraidq->cq_phase; +} + +static void spraid_sata_report_zone_handle(struct scsi_cmnd *scmd, + struct spraid_iod *iod) +{ + int i = 0; + unsigned int bytes = 0; + struct scatterlist *sg = scsi_sglist(scmd); + + scsi_for_each_sg(scmd, sg, iod->nsge, i) { + unsigned int offset = 0; + + if (bytes == 0) { + char *hdr; + u32 list_length; + u64 max_lba, opt_lba; + u16 same; + + hdr = sg_virt(sg); + + list_length = get_unaligned_le32(&hdr[0]); + same = get_unaligned_le16(&hdr[4]); + max_lba = get_unaligned_le64(&hdr[8]); + opt_lba = get_unaligned_le64(&hdr[16]); + put_unaligned_be32(list_length, &hdr[0]); + hdr[4] = same & 0xf; + put_unaligned_be64(max_lba, &hdr[8]); + put_unaligned_be64(opt_lba, &hdr[16]); + offset += 64; + bytes += 64; + } + while (offset < sg_dma_len(sg)) { + char *rec; + u8 cond, type, non_seq, reset; + u64 size, start, wp; + + rec = sg_virt(sg) + offset; + type = rec[0] & 0xf; + cond = (rec[1] >> 4) & 0xf; + non_seq = (rec[1] & 2); + reset = (rec[1] & 1); + size = get_unaligned_le64(&rec[8]); + start = get_unaligned_le64(&rec[16]); + wp = get_unaligned_le64(&rec[24]); + rec[0] = type; + rec[1] = (cond << 4) | non_seq | reset; + put_unaligned_be64(size, &rec[8]); + put_unaligned_be64(start, &rec[16]); + put_unaligned_be64(wp, &rec[24]); + WARN_ON(offset + 64 > sg_dma_len(sg)); + offset += 64; + bytes += 64; + } + } +} + +static inline void spraid_handle_ata_cmd(struct spraid_dev *hdev, + struct scsi_cmnd *scmd, + struct spraid_iod *iod) +{ + if (hdev->ctrl_info->card_type != SPRAID_CARD_HBA) + return; + + switch (scmd->cmnd[0]) { + case ZBC_IN: + dev_info(hdev->dev, "[%s] process report zone\n", __func__); + spraid_sata_report_zone_handle(scmd, iod); + break; + default: + break; + } +} + +static void spraid_complete_ioq_cmnd(struct spraid_queue *ioq, + struct spraid_completion *cqe) +{ + struct spraid_dev *hdev = ioq->hdev; + struct blk_mq_tags *tags; + struct scsi_cmnd *scmd; + struct spraid_iod *iod; + struct request *req; + unsigned long elapsed; + + tags = hdev->shost->tag_set.tags[ioq->qid - 1]; + req = blk_mq_tag_to_rq(tags, cqe->cmd_id); + if (unlikely(!req || !blk_mq_request_started(req))) { + dev_warn(hdev->dev, "Invalid id %d completed on queue %d\n", + cqe->cmd_id, ioq->qid); + return; + } + + scmd = blk_mq_rq_to_pdu(req); + iod = scsi_cmd_priv(scmd); + + elapsed = jiffies - scmd->jiffies_at_alloc; + dev_log_dbg(hdev->dev, + "cid[%d] qid[%d] finish IO cost %3ld.%3ld seconds\n", + cqe->cmd_id, ioq->qid, elapsed / HZ, elapsed % HZ); + + if (cmpxchg(&iod->state, SPRAID_CMD_IN_FLIGHT, SPRAID_CMD_COMPLETE) != + SPRAID_CMD_IN_FLIGHT) { + dev_warn(hdev->dev, + "cid[%d] qid[%d] enters abnormal handler;" + " cost %3ld.%3ld seconds\n", + cqe->cmd_id, ioq->qid, elapsed / HZ, elapsed % HZ); + WRITE_ONCE(iod->state, SPRAID_CMD_TMO_COMPLETE); + + if (iod->nsge) { + iod->nsge = 0; + scsi_dma_unmap(scmd); + } + spraid_free_iod_res(hdev, iod); + + return; + } + + spraid_handle_ata_cmd(hdev, scmd, iod); + + spraid_map_status(iod, scmd, cqe); + if (iod->nsge) { + iod->nsge = 0; + scsi_dma_unmap(scmd); + } + spraid_free_iod_res(hdev, iod); + scmd->scsi_done(scmd); +} + +static void spraid_complete_adminq_cmnd(struct spraid_queue *adminq, + struct spraid_completion *cqe) +{ + struct spraid_dev *hdev = adminq->hdev; + struct spraid_cmd *adm_cmd; + + adm_cmd = hdev->adm_cmds + cqe->cmd_id; + if (unlikely(adm_cmd->state == SPRAID_CMD_IDLE)) { + dev_warn(adminq->hdev->dev, + "Invalid id %d completed on queue %d\n", + cqe->cmd_id, le16_to_cpu(cqe->sq_id)); + return; + } + + adm_cmd->status = le16_to_cpu(cqe->status) >> 1; + adm_cmd->result0 = le32_to_cpu(cqe->result); + adm_cmd->result1 = le32_to_cpu(cqe->result1); + + complete(&adm_cmd->cmd_done); +} + +static void spraid_send_aen(struct spraid_dev *hdev, u16 cid); + +static void spraid_complete_aen(struct spraid_queue *spraidq, + struct spraid_completion *cqe) +{ + struct spraid_dev *hdev = spraidq->hdev; + u32 result = le32_to_cpu(cqe->result); + + dev_info(hdev->dev, "rcv aen, cid[%d], status[0x%x], result[0x%x]\n", + cqe->cmd_id, le16_to_cpu(cqe->status) >> 1, result); + + spraid_send_aen(hdev, cqe->cmd_id); + + if ((le16_to_cpu(cqe->status) >> 1) != SPRAID_SC_SUCCESS) + return; + switch (result & 0x7) { + case SPRAID_AEN_NOTICE: + spraid_handle_aen_notice(hdev, result); + break; + case SPRAID_AEN_VS: + spraid_handle_aen_vs(hdev, result, le32_to_cpu(cqe->result1)); + break; + default: + dev_warn(hdev->dev, "Unsupported async event type: %u\n", + result & 0x7); + break; + } +} + +static void spraid_complete_ioq_sync_cmnd(struct spraid_queue *ioq, + struct spraid_completion *cqe) +{ + struct spraid_dev *hdev = ioq->hdev; + struct spraid_cmd *ptcmd; + + ptcmd = hdev->ioq_ptcmds + (ioq->qid - 1) * SPRAID_PTCMDS_PERQ + + cqe->cmd_id - SPRAID_IO_BLK_MQ_DEPTH; + + ptcmd->status = le16_to_cpu(cqe->status) >> 1; + ptcmd->result0 = le32_to_cpu(cqe->result); + ptcmd->result1 = le32_to_cpu(cqe->result1); + + complete(&ptcmd->cmd_done); +} + +static inline void spraid_handle_cqe(struct spraid_queue *spraidq, u16 idx) +{ + struct spraid_completion *cqe = &spraidq->cqes[idx]; + struct spraid_dev *hdev = spraidq->hdev; + + if (unlikely(cqe->cmd_id >= spraidq->q_depth)) { + dev_err(hdev->dev, + "Invalid command id[%d] completed on queue %d\n", + cqe->cmd_id, cqe->sq_id); + return; + } + + dev_log_dbg(hdev->dev, "cid[%d] qid[%d];" + " result[0x%x], sq_id[%d], status[0x%x]\n", + cqe->cmd_id, spraidq->qid, le32_to_cpu(cqe->result), + le16_to_cpu(cqe->sq_id), le16_to_cpu(cqe->status)); + + if (unlikely(spraidq->qid == 0 + && cqe->cmd_id >= SPRAID_AQ_BLK_MQ_DEPTH)) { + spraid_complete_aen(spraidq, cqe); + return; + } + + if (unlikely(spraidq->qid && cqe->cmd_id >= SPRAID_IO_BLK_MQ_DEPTH)) { + spraid_complete_ioq_sync_cmnd(spraidq, cqe); + return; + } + + if (spraidq->qid) + spraid_complete_ioq_cmnd(spraidq, cqe); + else + spraid_complete_adminq_cmnd(spraidq, cqe); +} + +static void spraid_complete_cqes(struct spraid_queue *spraidq, + u16 start, u16 end) +{ + while (start != end) { + spraid_handle_cqe(spraidq, start); + if (++start == spraidq->q_depth) + start = 0; + } +} + +static inline void spraid_update_cq_head(struct spraid_queue *spraidq) +{ + if (++spraidq->cq_head == spraidq->q_depth) { + spraidq->cq_head = 0; + spraidq->cq_phase = !spraidq->cq_phase; + } +} + +static inline bool spraid_process_cq(struct spraid_queue *spraidq, + u16 *start, u16 *end, int tag) +{ + bool found = false; + + *start = spraidq->cq_head; + while (!found && spraid_cqe_pending(spraidq)) { + if (spraidq->cqes[spraidq->cq_head].cmd_id == tag) + found = true; + spraid_update_cq_head(spraidq); + } + *end = spraidq->cq_head; + + if (*start != *end) + writel(spraidq->cq_head, + spraidq->q_db + spraidq->hdev->db_stride); + + return found; +} + +static bool spraid_poll_cq(struct spraid_queue *spraidq, int cid) +{ + u16 start, end; + bool found; + + if (!spraid_cqe_pending(spraidq)) + return 0; + + spin_lock_irq(&spraidq->cq_lock); + found = spraid_process_cq(spraidq, &start, &end, cid); + spin_unlock_irq(&spraidq->cq_lock); + + spraid_complete_cqes(spraidq, start, end); + return found; +} + +static irqreturn_t spraid_irq(int irq, void *data) +{ + struct spraid_queue *spraidq = data; + irqreturn_t ret = IRQ_NONE; + u16 start, end; + + spin_lock(&spraidq->cq_lock); + if (spraidq->cq_head != spraidq->last_cq_head) + ret = IRQ_HANDLED; + + spraid_process_cq(spraidq, &start, &end, -1); + spraidq->last_cq_head = spraidq->cq_head; + spin_unlock(&spraidq->cq_lock); + + if (start != end) { + spraid_complete_cqes(spraidq, start, end); + ret = IRQ_HANDLED; + } + return ret; +} + +static int spraid_setup_admin_queue(struct spraid_dev *hdev) +{ + struct spraid_queue *adminq = &hdev->queues[0]; + u32 aqa; + int ret; + + dev_info(hdev->dev, "[%s] start disable ctrl\n", __func__); + + ret = spraid_disable_ctrl(hdev); + if (ret) + return ret; + + ret = spraid_alloc_queue(hdev, 0, SPRAID_AQ_DEPTH); + if (ret) + return ret; + + aqa = adminq->q_depth - 1; + aqa |= aqa << 16; + writel(aqa, hdev->bar + SPRAID_REG_AQA); + lo_hi_writeq(adminq->sq_dma_addr, hdev->bar + SPRAID_REG_ASQ); + lo_hi_writeq(adminq->cq_dma_addr, hdev->bar + SPRAID_REG_ACQ); + + dev_info(hdev->dev, "[%s] start enable ctrl\n", __func__); + + ret = spraid_enable_ctrl(hdev); + if (ret) { + ret = -ENODEV; + goto free_queue; + } + + adminq->cq_vector = 0; + spraid_init_queue(adminq, 0); + ret = pci_request_irq(hdev->pdev, adminq->cq_vector, spraid_irq, NULL, + adminq, "spraid%d_q%d", + hdev->instance, adminq->qid); + + if (ret) { + adminq->cq_vector = -1; + hdev->online_queues--; + goto free_queue; + } + + dev_info(hdev->dev, "[%s] success, queuecount:[%d], onlinequeue:[%d]\n", + __func__, hdev->queue_count, hdev->online_queues); + + return 0; + +free_queue: + spraid_free_queue(adminq); + return ret; +} + +static u32 spraid_bar_size(struct spraid_dev *hdev, u32 nr_ioqs) +{ + return (SPRAID_REG_DBS + ((nr_ioqs + 1) * 8 * hdev->db_stride)); +} + +static int spraid_alloc_admin_cmds(struct spraid_dev *hdev) +{ + int i; + + INIT_LIST_HEAD(&hdev->adm_cmd_list); + spin_lock_init(&hdev->adm_cmd_lock); + + hdev->adm_cmds = kcalloc_node(SPRAID_AQ_BLK_MQ_DEPTH, + sizeof(struct spraid_cmd), + GFP_KERNEL, hdev->numa_node); + + if (!hdev->adm_cmds) { + dev_err(hdev->dev, "Alloc admin cmds failed\n"); + return -ENOMEM; + } + + for (i = 0; i < SPRAID_AQ_BLK_MQ_DEPTH; i++) { + hdev->adm_cmds[i].qid = 0; + hdev->adm_cmds[i].cid = i; + list_add_tail(&(hdev->adm_cmds[i].list), &hdev->adm_cmd_list); + } + + dev_info(hdev->dev, "Alloc admin cmds success, num[%d]\n", + SPRAID_AQ_BLK_MQ_DEPTH); + + return 0; +} + +static void spraid_free_admin_cmds(struct spraid_dev *hdev) +{ + kfree(hdev->adm_cmds); + hdev->adm_cmds = NULL; + INIT_LIST_HEAD(&hdev->adm_cmd_list); +} + +static struct spraid_cmd *spraid_get_cmd(struct spraid_dev *hdev, + enum spraid_cmd_type type) +{ + struct spraid_cmd *cmd = NULL; + unsigned long flags; + struct list_head *head = &hdev->adm_cmd_list; + spinlock_t *slock = &hdev->adm_cmd_lock; + + if (type == SPRAID_CMD_IOPT) { + head = &hdev->ioq_pt_list; + slock = &hdev->ioq_pt_lock; + } + + spin_lock_irqsave(slock, flags); + if (list_empty(head)) { + spin_unlock_irqrestore(slock, flags); + dev_err(hdev->dev, "err, cmd[%d] list empty\n", type); + return NULL; + } + cmd = list_entry(head->next, struct spraid_cmd, list); + list_del_init(&cmd->list); + spin_unlock_irqrestore(slock, flags); + + WRITE_ONCE(cmd->state, SPRAID_CMD_IN_FLIGHT); + + return cmd; +} + +static void spraid_put_cmd(struct spraid_dev *hdev, struct spraid_cmd *cmd, + enum spraid_cmd_type type) +{ + unsigned long flags; + struct list_head *head = &hdev->adm_cmd_list; + spinlock_t *slock = &hdev->adm_cmd_lock; + + if (type == SPRAID_CMD_IOPT) { + head = &hdev->ioq_pt_list; + slock = &hdev->ioq_pt_lock; + } + + spin_lock_irqsave(slock, flags); + WRITE_ONCE(cmd->state, SPRAID_CMD_IDLE); + list_add_tail(&cmd->list, head); + spin_unlock_irqrestore(slock, flags); +} + + +static int spraid_submit_admin_sync_cmd(struct spraid_dev *hdev, + struct spraid_admin_command *cmd, + u32 *result0, u32 *result1, u32 timeout) +{ + struct spraid_cmd *adm_cmd = spraid_get_cmd(hdev, SPRAID_CMD_ADM); + + if (!adm_cmd) { + dev_err(hdev->dev, "err, get admin cmd failed\n"); + return -EFAULT; + } + + timeout = timeout ? timeout : ADMIN_TIMEOUT; + + init_completion(&adm_cmd->cmd_done); + + cmd->common.command_id = adm_cmd->cid; + spraid_submit_cmd(&hdev->queues[0], cmd); + + if (!wait_for_completion_timeout(&adm_cmd->cmd_done, timeout)) { + dev_err(hdev->dev, "[%s] cid[%d] qid[%d] timeout;" + " opcode[0x%x] subopcode[0x%x]\n", + __func__, adm_cmd->cid, adm_cmd->qid, + cmd->usr_cmd.opcode, cmd->usr_cmd.info_0.subopcode); + WRITE_ONCE(adm_cmd->state, SPRAID_CMD_TIMEOUT); + spraid_put_cmd(hdev, adm_cmd, SPRAID_CMD_ADM); + return -EINVAL; + } + + if (result0) + *result0 = adm_cmd->result0; + if (result1) + *result1 = adm_cmd->result1; + + spraid_put_cmd(hdev, adm_cmd, SPRAID_CMD_ADM); + + return adm_cmd->status; +} + +static int spraid_create_cq(struct spraid_dev *hdev, u16 qid, + struct spraid_queue *spraidq, u16 cq_vector) +{ + struct spraid_admin_command admin_cmd; + int flags = SPRAID_QUEUE_PHYS_CONTIG | SPRAID_CQ_IRQ_ENABLED; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.create_cq.opcode = SPRAID_ADMIN_CREATE_CQ; + admin_cmd.create_cq.prp1 = cpu_to_le64(spraidq->cq_dma_addr); + admin_cmd.create_cq.cqid = cpu_to_le16(qid); + admin_cmd.create_cq.qsize = cpu_to_le16(spraidq->q_depth - 1); + admin_cmd.create_cq.cq_flags = cpu_to_le16(flags); + admin_cmd.create_cq.irq_vector = cpu_to_le16(cq_vector); + + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); +} + +static int spraid_create_sq(struct spraid_dev *hdev, u16 qid, + struct spraid_queue *spraidq) +{ + struct spraid_admin_command admin_cmd; + int flags = SPRAID_QUEUE_PHYS_CONTIG; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.create_sq.opcode = SPRAID_ADMIN_CREATE_SQ; + admin_cmd.create_sq.prp1 = cpu_to_le64(spraidq->sq_dma_addr); + admin_cmd.create_sq.sqid = cpu_to_le16(qid); + admin_cmd.create_sq.qsize = cpu_to_le16(spraidq->q_depth - 1); + admin_cmd.create_sq.sq_flags = cpu_to_le16(flags); + admin_cmd.create_sq.cqid = cpu_to_le16(qid); + + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); +} + +static void spraid_free_queue(struct spraid_queue *spraidq) +{ + struct spraid_dev *hdev = spraidq->hdev; + + hdev->queue_count--; + dma_free_coherent(hdev->dev, CQ_SIZE(spraidq->q_depth), + (void *)spraidq->cqes, spraidq->cq_dma_addr); + dma_free_coherent(hdev->dev, SQ_SIZE(spraidq->qid, spraidq->q_depth), + spraidq->sq_cmds, spraidq->sq_dma_addr); + dma_free_coherent(hdev->dev, SENSE_SIZE(spraidq->q_depth), + spraidq->sense, spraidq->sense_dma_addr); +} + +static void spraid_free_admin_queue(struct spraid_dev *hdev) +{ + spraid_free_queue(&hdev->queues[0]); +} + +static void spraid_free_io_queues(struct spraid_dev *hdev) +{ + int i; + + for (i = hdev->queue_count - 1; i >= 1; i--) + spraid_free_queue(&hdev->queues[i]); +} + +static int spraid_delete_queue(struct spraid_dev *hdev, u8 op, u16 id) +{ + struct spraid_admin_command admin_cmd; + int ret; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.delete_queue.opcode = op; + admin_cmd.delete_queue.qid = cpu_to_le16(id); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + + if (ret) + dev_err(hdev->dev, "Delete %s:[%d] failed\n", + (op == SPRAID_ADMIN_DELETE_CQ) ? "cq" : "sq", id); + + return ret; +} + +static int spraid_delete_cq(struct spraid_dev *hdev, u16 cqid) +{ + return spraid_delete_queue(hdev, SPRAID_ADMIN_DELETE_CQ, cqid); +} + +static int spraid_delete_sq(struct spraid_dev *hdev, u16 sqid) +{ + return spraid_delete_queue(hdev, SPRAID_ADMIN_DELETE_SQ, sqid); +} + +static int spraid_create_queue(struct spraid_queue *spraidq, u16 qid) +{ + struct spraid_dev *hdev = spraidq->hdev; + u16 cq_vector; + int ret; + + cq_vector = (hdev->num_vecs == 1) ? 0 : qid; + ret = spraid_create_cq(hdev, qid, spraidq, cq_vector); + if (ret) + return ret; + + ret = spraid_create_sq(hdev, qid, spraidq); + if (ret) + goto delete_cq; + + spraid_init_queue(spraidq, qid); + spraidq->cq_vector = cq_vector; + + ret = pci_request_irq(hdev->pdev, cq_vector, spraid_irq, NULL, + spraidq, "spraid%d_q%d", hdev->instance, qid); + + if (ret) { + dev_err(hdev->dev, "Request queue[%d] irq failed\n", qid); + goto delete_sq; + } + + return 0; + +delete_sq: + spraidq->cq_vector = -1; + hdev->online_queues--; + spraid_delete_sq(hdev, qid); +delete_cq: + spraid_delete_cq(hdev, qid); + + return ret; +} + +static int spraid_create_io_queues(struct spraid_dev *hdev) +{ + u32 i, max; + int ret = 0; + + max = min(hdev->max_qid, hdev->queue_count - 1); + for (i = hdev->online_queues; i <= max; i++) { + ret = spraid_create_queue(&hdev->queues[i], i); + if (ret) { + dev_err(hdev->dev, "Create queue[%d] failed\n", i); + break; + } + } + + dev_info(hdev->dev, "[%s] queue_count[%d], online_queue[%d]", + __func__, hdev->queue_count, hdev->online_queues); + + return ret >= 0 ? 0 : ret; +} + +static int spraid_set_features(struct spraid_dev *hdev, u32 fid, + u32 dword11, void *buffer, + size_t buflen, u32 *result) +{ + struct spraid_admin_command admin_cmd; + int ret; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + + if (buffer && buflen) { + data_ptr = dma_alloc_coherent(hdev->dev, buflen, + &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memcpy(data_ptr, buffer, buflen); + } + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.features.opcode = SPRAID_ADMIN_SET_FEATURES; + admin_cmd.features.fid = cpu_to_le32(fid); + admin_cmd.features.dword11 = cpu_to_le32(dword11); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, result, NULL, 0); + + if (data_ptr) + dma_free_coherent(hdev->dev, buflen, data_ptr, data_dma); + + return ret; +} + +static int spraid_configure_timestamp(struct spraid_dev *hdev) +{ + __le64 ts; + int ret; + + ts = cpu_to_le64(ktime_to_ms(ktime_get_real())); + ret = spraid_set_features(hdev, SPRAID_FEAT_TIMESTAMP, + 0, &ts, sizeof(ts), NULL); + + if (ret) + dev_err(hdev->dev, "set timestamp failed: %d\n", ret); + return ret; +} + +static int spraid_set_queue_cnt(struct spraid_dev *hdev, u32 *cnt) +{ + u32 q_cnt = (*cnt - 1) | ((*cnt - 1) << 16); + u32 nr_ioqs, result; + int status; + + status = spraid_set_features(hdev, SPRAID_FEAT_NUM_QUEUES, + q_cnt, NULL, 0, &result); + if (status) { + dev_err(hdev->dev, "Set queue count failed, status: %d\n", + status); + return -EIO; + } + + nr_ioqs = min(result & 0xffff, result >> 16) + 1; + *cnt = min(*cnt, nr_ioqs); + if (*cnt == 0) { + dev_err(hdev->dev, "Illegal queue count: zero\n"); + return -EIO; + } + return 0; +} + +static int spraid_setup_io_queues(struct spraid_dev *hdev) +{ + struct spraid_queue *adminq = &hdev->queues[0]; + struct pci_dev *pdev = hdev->pdev; + u32 nr_ioqs = num_online_cpus(); + u32 i, size; + int ret; + + struct irq_affinity affd = { + .pre_vectors = 1 + }; + + ret = spraid_set_queue_cnt(hdev, &nr_ioqs); + if (ret < 0) + return ret; + + size = spraid_bar_size(hdev, nr_ioqs); + ret = spraid_remap_bar(hdev, size); + if (ret) + return -ENOMEM; + + adminq->q_db = hdev->dbs; + + pci_free_irq(pdev, 0, adminq); + pci_free_irq_vectors(pdev); + + ret = pci_alloc_irq_vectors_affinity(pdev, 1, (nr_ioqs + 1), + PCI_IRQ_ALL_TYPES | PCI_IRQ_AFFINITY, &affd); + if (ret <= 0) + return -EIO; + + hdev->num_vecs = ret; + + hdev->max_qid = max(ret - 1, 1); + + ret = pci_request_irq(pdev, adminq->cq_vector, spraid_irq, NULL, + adminq, "spraid%d_q%d", + hdev->instance, adminq->qid); + if (ret) { + dev_err(hdev->dev, "Request admin irq failed\n"); + adminq->cq_vector = -1; + return ret; + } + + for (i = hdev->queue_count; i <= hdev->max_qid; i++) { + ret = spraid_alloc_queue(hdev, i, hdev->ioq_depth); + if (ret) + break; + } + dev_info(hdev->dev, "[%s] max_qid: %d, queue_count: %d;" + " online_queue: %d, ioq_depth: %d\n", + __func__, hdev->max_qid, hdev->queue_count, + hdev->online_queues, hdev->ioq_depth); + + return spraid_create_io_queues(hdev); +} + +static void spraid_delete_io_queues(struct spraid_dev *hdev) +{ + u16 queues = hdev->online_queues - 1; + u8 opcode = SPRAID_ADMIN_DELETE_SQ; + u16 i, pass; + + if (!pci_device_is_present(hdev->pdev)) { + dev_err(hdev->dev, + "pci_device is not present, skip disable io queues\n"); + return; + } + + if (hdev->online_queues < 2) { + dev_err(hdev->dev, "[%s] err, io queue has been delete\n", + __func__); + return; + } + + for (pass = 0; pass < 2; pass++) { + for (i = queues; i > 0; i--) + if (spraid_delete_queue(hdev, opcode, i)) + break; + + opcode = SPRAID_ADMIN_DELETE_CQ; + } +} + +static void spraid_remove_io_queues(struct spraid_dev *hdev) +{ + spraid_delete_io_queues(hdev); + spraid_free_io_queues(hdev); +} + +static void spraid_pci_disable(struct spraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + u32 i; + + for (i = 0; i < hdev->online_queues; i++) + pci_free_irq(pdev, hdev->queues[i].cq_vector, &hdev->queues[i]); + pci_free_irq_vectors(pdev); + if (pci_is_enabled(pdev)) { + pci_disable_pcie_error_reporting(pdev); + pci_disable_device(pdev); + } + hdev->online_queues = 0; +} + +static void spraid_disable_admin_queue(struct spraid_dev *hdev, bool shutdown) +{ + struct spraid_queue *adminq = &hdev->queues[0]; + u16 start, end; + + if (pci_device_is_present(hdev->pdev)) { + if (shutdown) + spraid_shutdown_ctrl(hdev); + else + spraid_disable_ctrl(hdev); + } + + if (hdev->queue_count == 0) { + dev_err(hdev->dev, "[%s] err, admin queue has been delete\n", + __func__); + return; + } + + spin_lock_irq(&adminq->cq_lock); + spraid_process_cq(adminq, &start, &end, -1); + spin_unlock_irq(&adminq->cq_lock); + + spraid_complete_cqes(adminq, start, end); + spraid_free_admin_queue(hdev); +} + +static int spraid_create_dma_pools(struct spraid_dev *hdev) +{ + int i; + char poolname[20] = { 0 }; + + hdev->prp_page_pool = dma_pool_create("prp list page", hdev->dev, + PAGE_SIZE, PAGE_SIZE, 0); + + if (!hdev->prp_page_pool) { + dev_err(hdev->dev, "create prp_page_pool failed\n"); + return -ENOMEM; + } + + for (i = 0; i < small_pool_num; i++) { + sprintf(poolname, "prp_list_256_%d", i); + hdev->prp_small_pool[i] = + dma_pool_create(poolname, hdev->dev, SMALL_POOL_SIZE, + SMALL_POOL_SIZE, 0); + + if (!hdev->prp_small_pool[i]) { + dev_err(hdev->dev, "create prp_small_pool %d failed\n", + i); + goto destroy_prp_small_pool; + } + } + + return 0; + +destroy_prp_small_pool: + while (i > 0) + dma_pool_destroy(hdev->prp_small_pool[--i]); + dma_pool_destroy(hdev->prp_page_pool); + + return -ENOMEM; +} + +static void spraid_destroy_dma_pools(struct spraid_dev *hdev) +{ + int i; + + for (i = 0; i < small_pool_num; i++) + dma_pool_destroy(hdev->prp_small_pool[i]); + dma_pool_destroy(hdev->prp_page_pool); +} + +static int spraid_get_dev_list(struct spraid_dev *hdev, + struct spraid_dev_info *devices) +{ + u32 nd = le32_to_cpu(hdev->ctrl_info->nd); + struct spraid_admin_command admin_cmd; + struct spraid_dev_list *list_buf; + dma_addr_t data_dma = 0; + u32 i, idx, hdid, ndev; + int ret = 0; + + list_buf = dma_alloc_coherent(hdev->dev, PAGE_SIZE, + &data_dma, GFP_KERNEL); + if (!list_buf) + return -ENOMEM; + + for (idx = 0; idx < nd;) { + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.get_info.opcode = SPRAID_ADMIN_GET_INFO; + admin_cmd.get_info.type = SPRAID_GET_INFO_DEV_LIST; + admin_cmd.get_info.cdw11 = cpu_to_le32(idx); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, + NULL, NULL, 0); + + if (ret) { + dev_err(hdev->dev, "Get device list failed, nd: %u;" + "idx: %u, ret: %d\n", + nd, idx, ret); + goto out; + } + ndev = le32_to_cpu(list_buf->dev_num); + + dev_info(hdev->dev, "ndev numbers: %u\n", ndev); + + for (i = 0; i < ndev; i++) { + hdid = le32_to_cpu(list_buf->devices[i].hdid); + dev_info(hdev->dev, "list_buf->devices[%d], hdid: %u;" + "target: %d, channel: %d, lun: %d, attr[0x%x]\n", + i, hdid, + le16_to_cpu(list_buf->devices[i].target), + list_buf->devices[i].channel, + list_buf->devices[i].lun, + list_buf->devices[i].attr); + if (hdid > nd || hdid == 0) { + dev_err(hdev->dev, "err, hdid[%d] invalid\n", + hdid); + continue; + } + memcpy(&devices[hdid - 1], &list_buf->devices[i], + sizeof(struct spraid_dev_info)); + } + idx += ndev; + + if (idx < MAX_DEV_ENTRY_PER_PAGE_4K) + break; + } + +out: + dma_free_coherent(hdev->dev, PAGE_SIZE, list_buf, data_dma); + return ret; +} + +static void spraid_send_aen(struct spraid_dev *hdev, u16 cid) +{ + struct spraid_queue *adminq = &hdev->queues[0]; + struct spraid_admin_command admin_cmd; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.common.opcode = SPRAID_ADMIN_ASYNC_EVENT; + admin_cmd.common.command_id = cid; + + spraid_submit_cmd(adminq, &admin_cmd); + dev_info(hdev->dev, "send aen, cid[%d]\n", cid); +} + +static inline void spraid_send_all_aen(struct spraid_dev *hdev) +{ + u16 i; + + for (i = 0; i < hdev->ctrl_info->aerl; i++) + spraid_send_aen(hdev, i + SPRAID_AQ_BLK_MQ_DEPTH); +} + +static int spraid_add_device(struct spraid_dev *hdev, + struct spraid_dev_info *device) +{ + struct Scsi_Host *shost = hdev->shost; + struct scsi_device *sdev; + + dev_info(hdev->dev, "add device, hdid: %u target: %d, channel: %d;" + " lun: %d, attr[0x%x]\n", + le32_to_cpu(device->hdid), le16_to_cpu(device->target), + device->channel, device->lun, device->attr); + + sdev = scsi_device_lookup(shost, device->channel, + le16_to_cpu(device->target), 0); + if (sdev) { + dev_warn(hdev->dev, "Device is already exist, channel: %d;" + " target_id: %d, lun: %d\n", + device->channel, le16_to_cpu(device->target), 0); + scsi_device_put(sdev); + return -EEXIST; + } + scsi_add_device(shost, device->channel, le16_to_cpu(device->target), 0); + return 0; +} + +static int spraid_rescan_device(struct spraid_dev *hdev, + struct spraid_dev_info *device) +{ + struct Scsi_Host *shost = hdev->shost; + struct scsi_device *sdev; + + dev_info(hdev->dev, "rescan device, hdid: %u target: %d, channel: %d;" + " lun: %d, attr[0x%x]\n", + le32_to_cpu(device->hdid), le16_to_cpu(device->target), + device->channel, device->lun, device->attr); + + sdev = scsi_device_lookup(shost, device->channel, + le16_to_cpu(device->target), 0); + if (!sdev) { + dev_warn(hdev->dev, "device is not exit rescan it, channel: %d;" + " target_id: %d, lun: %d\n", + device->channel, le16_to_cpu(device->target), 0); + return -ENODEV; + } + + scsi_rescan_device(&sdev->sdev_gendev); + scsi_device_put(sdev); + return 0; +} + +static int spraid_remove_device(struct spraid_dev *hdev, + struct spraid_dev_info *org_device) +{ + struct Scsi_Host *shost = hdev->shost; + struct scsi_device *sdev; + + dev_info(hdev->dev, "remove device, hdid: %u target: %d, channel: %d;" + " lun: %d, attr[0x%x]\n", + le32_to_cpu(org_device->hdid), le16_to_cpu(org_device->target), + org_device->channel, org_device->lun, org_device->attr); + + sdev = scsi_device_lookup(shost, org_device->channel, + le16_to_cpu(org_device->target), 0); + if (!sdev) { + dev_warn(hdev->dev, "device is not exit remove it, channel: %d;" + " target_id: %d, lun: %d\n", + org_device->channel, + le16_to_cpu(org_device->target), 0); + return -ENODEV; + } + + scsi_remove_device(sdev); + scsi_device_put(sdev); + return 0; +} + +static int spraid_dev_list_init(struct spraid_dev *hdev) +{ + u32 nd = le32_to_cpu(hdev->ctrl_info->nd); + int i, ret; + + hdev->devices = kzalloc_node(nd * sizeof(struct spraid_dev_info), + GFP_KERNEL, hdev->numa_node); + if (!hdev->devices) + return -ENOMEM; + + ret = spraid_get_dev_list(hdev, hdev->devices); + if (ret) { + dev_err(hdev->dev, + "Ignore failure of getting device list;" + " within initialization\n"); + return 0; + } + + for (i = 0; i < nd; i++) { + if (SPRAID_DEV_INFO_FLAG_VALID(hdev->devices[i].flag) && + SPRAID_DEV_INFO_ATTR_BOOT(hdev->devices[i].attr)) { + spraid_add_device(hdev, &hdev->devices[i]); + break; + } + } + return 0; +} + +static int luntarget_cmp_func(const void *l, const void *r) +{ + const struct spraid_dev_info *ln = l; + const struct spraid_dev_info *rn = r; + + if (ln->channel == rn->channel) + return le16_to_cpu(ln->target) - le16_to_cpu(rn->target); + + return ln->channel - rn->channel; +} + +static void spraid_scan_work(struct work_struct *work) +{ + struct spraid_dev *hdev = + container_of(work, struct spraid_dev, scan_work); + struct spraid_dev_info *devices, *org_devices; + struct spraid_dev_info *sortdevice; + u32 nd = le32_to_cpu(hdev->ctrl_info->nd); + u8 flag, org_flag; + int i, ret; + int count = 0; + + devices = kcalloc(nd, sizeof(struct spraid_dev_info), GFP_KERNEL); + if (!devices) + return; + + sortdevice = kcalloc(nd, sizeof(struct spraid_dev_info), GFP_KERNEL); + if (!sortdevice) + goto free_list; + + ret = spraid_get_dev_list(hdev, devices); + if (ret) + goto free_all; + org_devices = hdev->devices; + for (i = 0; i < nd; i++) { + org_flag = org_devices[i].flag; + flag = devices[i].flag; + + dev_log_dbg(hdev->dev, "i: %d, org_flag: 0x%x, flag: 0x%x\n", + i, org_flag, flag); + + if (SPRAID_DEV_INFO_FLAG_VALID(flag)) { + if (!SPRAID_DEV_INFO_FLAG_VALID(org_flag)) { + down_write(&hdev->devices_rwsem); + memcpy(&org_devices[i], &devices[i], + sizeof(struct spraid_dev_info)); + memcpy(&sortdevice[count++], &devices[i], + sizeof(struct spraid_dev_info)); + up_write(&hdev->devices_rwsem); + } else if (SPRAID_DEV_INFO_FLAG_CHANGE(flag)) { + spraid_rescan_device(hdev, &devices[i]); + } + } else { + if (SPRAID_DEV_INFO_FLAG_VALID(org_flag)) { + down_write(&hdev->devices_rwsem); + org_devices[i].flag &= 0xfe; + up_write(&hdev->devices_rwsem); + spraid_remove_device(hdev, &org_devices[i]); + } + } + } + + dev_info(hdev->dev, "scan work add device count = %d\n", count); + + sort(sortdevice, count, sizeof(sortdevice[0]), + luntarget_cmp_func, NULL); + + for (i = 0; i < count; i++) + spraid_add_device(hdev, &sortdevice[i]); + +free_all: + kfree(sortdevice); +free_list: + kfree(devices); +} + +static void spraid_timesyn_work(struct work_struct *work) +{ + struct spraid_dev *hdev = + container_of(work, struct spraid_dev, timesyn_work); + + spraid_configure_timestamp(hdev); +} + +static int spraid_init_ctrl_info(struct spraid_dev *hdev); +static void spraid_fw_act_work(struct work_struct *work) +{ + struct spraid_dev *hdev = + container_of(work, struct spraid_dev, fw_act_work); + + if (spraid_init_ctrl_info(hdev)) + dev_err(hdev->dev, "get ctrl info failed after fw act\n"); +} + +static void spraid_queue_scan(struct spraid_dev *hdev) +{ + queue_work(spraid_wq, &hdev->scan_work); +} + +static void spraid_handle_aen_notice(struct spraid_dev *hdev, u32 result) +{ + switch ((result & 0xff00) >> 8) { + case SPRAID_AEN_DEV_CHANGED: + spraid_queue_scan(hdev); + break; + case SPRAID_AEN_FW_ACT_START: + dev_info(hdev->dev, "fw activation starting\n"); + break; + case SPRAID_AEN_HOST_PROBING: + break; + default: + dev_warn(hdev->dev, "async event result %08x\n", result); + } +} + +static void spraid_handle_aen_vs(struct spraid_dev *hdev, + u32 result, u32 result1) +{ + switch ((result & 0xff00) >> 8) { + case SPRAID_AEN_TIMESYN: + queue_work(spraid_wq, &hdev->timesyn_work); + break; + case SPRAID_AEN_FW_ACT_FINISH: + dev_info(hdev->dev, "fw activation finish\n"); + queue_work(spraid_wq, &hdev->fw_act_work); + break; + case SPRAID_AEN_EVENT_MIN ... SPRAID_AEN_EVENT_MAX: + dev_info(hdev->dev, "rcv card event[%d];" + " param1[0x%x] param2[0x%x]\n", + (result & 0xff00) >> 8, result, result1); + break; + default: + dev_warn(hdev->dev, "async event result: 0x%x\n", result); + } +} + +static int spraid_alloc_resources(struct spraid_dev *hdev) +{ + int ret, nqueue; + + ret = ida_alloc(&spraid_instance_ida, GFP_KERNEL); + if (ret < 0) { + dev_err(hdev->dev, "Get instance id failed\n"); + return ret; + } + hdev->instance = ret; + + hdev->ctrl_info = kzalloc_node(sizeof(*hdev->ctrl_info), + GFP_KERNEL, hdev->numa_node); + if (!hdev->ctrl_info) { + ret = -ENOMEM; + goto release_instance; + } + + ret = spraid_create_dma_pools(hdev); + if (ret) + goto free_ctrl_info; + nqueue = num_possible_cpus() + 1; + hdev->queues = kcalloc_node(nqueue, sizeof(struct spraid_queue), + GFP_KERNEL, hdev->numa_node); + if (!hdev->queues) { + ret = -ENOMEM; + goto destroy_dma_pools; + } + + ret = spraid_alloc_admin_cmds(hdev); + if (ret) + goto free_queues; + + dev_info(hdev->dev, "[%s] queues num: %d\n", __func__, nqueue); + + return 0; + +free_queues: + kfree(hdev->queues); +destroy_dma_pools: + spraid_destroy_dma_pools(hdev); +free_ctrl_info: + kfree(hdev->ctrl_info); +release_instance: + ida_free(&spraid_instance_ida, hdev->instance); + return ret; +} + +static void spraid_free_resources(struct spraid_dev *hdev) +{ + spraid_free_admin_cmds(hdev); + kfree(hdev->queues); + spraid_destroy_dma_pools(hdev); + kfree(hdev->ctrl_info); + ida_free(&spraid_instance_ida, hdev->instance); +} + +static void spraid_bsg_unmap_data(struct spraid_dev *hdev, struct bsg_job *job) +{ + struct request *rq = blk_mq_rq_from_pdu(job); + struct spraid_iod *iod = job->dd_data; + enum dma_data_direction dma_dir = + rq_data_dir(rq) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; + + if (iod->nsge) + dma_unmap_sg(hdev->dev, iod->sg, iod->nsge, dma_dir); + + spraid_free_iod_res(hdev, iod); +} + +static int spraid_bsg_map_data(struct spraid_dev *hdev, struct bsg_job *job, + struct spraid_admin_command *cmd) +{ + struct request *rq = blk_mq_rq_from_pdu(job); + struct spraid_iod *iod = job->dd_data; + enum dma_data_direction dma_dir = + rq_data_dir(rq) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; + int ret = 0; + + iod->sg = job->request_payload.sg_list; + iod->nsge = job->request_payload.sg_cnt; + iod->length = job->request_payload.payload_len; + iod->use_sgl = false; + iod->npages = -1; + iod->sg_drv_mgmt = false; + + if (!iod->nsge) + goto out; + + ret = dma_map_sg_attrs(hdev->dev, iod->sg, iod->nsge, + dma_dir, DMA_ATTR_NO_WARN); + if (!ret) + goto out; + + ret = spraid_setup_prps(hdev, iod); + if (ret) + goto unmap; + + cmd->common.dptr.prp1 = cpu_to_le64(sg_dma_address(iod->sg)); + cmd->common.dptr.prp2 = cpu_to_le64(iod->first_dma); + + return 0; + +unmap: + dma_unmap_sg(hdev->dev, iod->sg, iod->nsge, dma_dir); +out: + return ret; +} + +static int spraid_get_ctrl_info(struct spraid_dev *hdev, + struct spraid_ctrl_info *ctrl_info) +{ + struct spraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + int ret; + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, + &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.get_info.opcode = SPRAID_ADMIN_GET_INFO; + admin_cmd.get_info.type = SPRAID_GET_INFO_CTRL; + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + if (!ret) + memcpy(ctrl_info, data_ptr, sizeof(struct spraid_ctrl_info)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, data_dma); + + return ret; +} + +static int spraid_init_ctrl_info(struct spraid_dev *hdev) +{ + int ret; + + hdev->ctrl_info->nd = cpu_to_le32(240); + hdev->ctrl_info->mdts = 8; + hdev->ctrl_info->max_cmds = cpu_to_le16(4096); + hdev->ctrl_info->max_num_sge = cpu_to_le16(128); + hdev->ctrl_info->max_channel = cpu_to_le16(4); + hdev->ctrl_info->max_tgt_id = cpu_to_le32(3239); + hdev->ctrl_info->max_lun = cpu_to_le16(2); + + ret = spraid_get_ctrl_info(hdev, hdev->ctrl_info); + if (ret) + dev_err(hdev->dev, "get controller info failed: %d\n", ret); + + dev_info(hdev->dev, "[%s]nd = %d\n", __func__, hdev->ctrl_info->nd); + dev_info(hdev->dev, "[%s]max_cmd = %d\n", + __func__, hdev->ctrl_info->max_cmds); + dev_info(hdev->dev, "[%s]max_channel = %d\n", + __func__, hdev->ctrl_info->max_channel); + dev_info(hdev->dev, "[%s]max_tgt_id = %d\n", + __func__, hdev->ctrl_info->max_tgt_id); + dev_info(hdev->dev, "[%s]max_lun = %d\n", + __func__, hdev->ctrl_info->max_lun); + dev_info(hdev->dev, "[%s]max_num_sge = %d\n", + __func__, hdev->ctrl_info->max_num_sge); + dev_info(hdev->dev, "[%s]lun_num_boot = %d\n", + __func__, hdev->ctrl_info->lun_num_in_boot); + dev_info(hdev->dev, "[%s]mdts = %d\n", __func__, hdev->ctrl_info->mdts); + dev_info(hdev->dev, "[%s]acl = %d\n", __func__, hdev->ctrl_info->acl); + dev_info(hdev->dev, "[%s]aer1 = %d\n", __func__, hdev->ctrl_info->aerl); + dev_info(hdev->dev, "[%s]card_type = %d\n", + __func__, hdev->ctrl_info->card_type); + dev_info(hdev->dev, "[%s]rtd3e = %d\n", + __func__, hdev->ctrl_info->rtd3e); + dev_info(hdev->dev, "[%s]sn = %s\n", __func__, hdev->ctrl_info->sn); + dev_info(hdev->dev, "[%s]fr = %s\n", __func__, hdev->ctrl_info->fr); + + if (!hdev->ctrl_info->aerl) + hdev->ctrl_info->aerl = 1; + if (hdev->ctrl_info->aerl > SPRAID_NR_AEN_COMMANDS) + hdev->ctrl_info->aerl = SPRAID_NR_AEN_COMMANDS; + + return 0; +} + +#define SPRAID_MAX_ADMIN_PAYLOAD_SIZE BIT(16) +static int spraid_alloc_iod_ext_mem_pool(struct spraid_dev *hdev) +{ + u16 max_sge = le16_to_cpu(hdev->ctrl_info->max_num_sge); + size_t alloc_size; + + alloc_size = spraid_iod_ext_size(hdev, SPRAID_MAX_ADMIN_PAYLOAD_SIZE, + max_sge, true, false); + if (alloc_size > PAGE_SIZE) + dev_warn(hdev->dev, "It is unreasonable ;" + " sg allocation more than one page\n"); + hdev->iod_mempool = mempool_create_node(1, mempool_kmalloc, + mempool_kfree, + (void *)alloc_size, GFP_KERNEL, + hdev->numa_node); + if (!hdev->iod_mempool) { + dev_err(hdev->dev, "Create iod extension memory pool failed\n"); + return -ENOMEM; + } + + return 0; +} + +static void spraid_free_iod_ext_mem_pool(struct spraid_dev *hdev) +{ + mempool_destroy(hdev->iod_mempool); +} + +static int spraid_user_admin_cmd(struct spraid_dev *hdev, struct bsg_job *job) +{ + struct spraid_bsg_request *bsg_req = job->request; + struct spraid_passthru_common_cmd *cmd = &(bsg_req->admcmd); + struct spraid_admin_command admin_cmd; + u32 timeout = msecs_to_jiffies(cmd->timeout_ms); + u32 result[2] = {0}; + int status; + + if (hdev->state >= SPRAID_RESETTING) { + dev_err(hdev->dev, "[%s] err, host state:[%d] is not right\n", + __func__, hdev->state); + return -EBUSY; + } + + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x] init\n", + __func__, cmd->opcode, cmd->info_0.subopcode); + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.common.opcode = cmd->opcode; + admin_cmd.common.flags = cmd->flags; + admin_cmd.common.hdid = cpu_to_le32(cmd->nsid); + admin_cmd.common.cdw2[0] = cpu_to_le32(cmd->cdw2); + admin_cmd.common.cdw2[1] = cpu_to_le32(cmd->cdw3); + admin_cmd.common.cdw10 = cpu_to_le32(cmd->cdw10); + admin_cmd.common.cdw11 = cpu_to_le32(cmd->cdw11); + admin_cmd.common.cdw12 = cpu_to_le32(cmd->cdw12); + admin_cmd.common.cdw13 = cpu_to_le32(cmd->cdw13); + admin_cmd.common.cdw14 = cpu_to_le32(cmd->cdw14); + admin_cmd.common.cdw15 = cpu_to_le32(cmd->cdw15); + + status = spraid_bsg_map_data(hdev, job, &admin_cmd); + if (status) { + dev_err(hdev->dev, "[%s] err, map data failed\n", __func__); + return status; + } + + status = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, &result[0], + &result[1], timeout); + if (status >= 0) { + job->reply_len = sizeof(result); + memcpy(job->reply, result, sizeof(result)); + } + + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x];" + " status[0x%x] result0[0x%x] result1[0x%x]\n", + __func__, cmd->opcode, cmd->info_0.subopcode, + status, result[0], result[1]); + + spraid_bsg_unmap_data(hdev, job); + + return status; +} + +static int spraid_alloc_ioq_ptcmds(struct spraid_dev *hdev) +{ + int i; + int ptnum = SPRAID_NR_IOQ_PTCMDS; + + INIT_LIST_HEAD(&hdev->ioq_pt_list); + spin_lock_init(&hdev->ioq_pt_lock); + + hdev->ioq_ptcmds = kcalloc_node(ptnum, sizeof(struct spraid_cmd), + GFP_KERNEL, hdev->numa_node); + + if (!hdev->ioq_ptcmds) { + dev_err(hdev->dev, "Alloc ioq_ptcmds failed\n"); + return -ENOMEM; + } + + for (i = 0; i < ptnum; i++) { + hdev->ioq_ptcmds[i].qid = i / SPRAID_PTCMDS_PERQ + 1; + hdev->ioq_ptcmds[i].cid = i % SPRAID_PTCMDS_PERQ + + SPRAID_IO_BLK_MQ_DEPTH; + list_add_tail(&(hdev->ioq_ptcmds[i].list), &hdev->ioq_pt_list); + } + + dev_info(hdev->dev, "Alloc ioq_ptcmds success, ptnum[%d]\n", ptnum); + + return 0; +} + +static void spraid_free_ioq_ptcmds(struct spraid_dev *hdev) +{ + kfree(hdev->ioq_ptcmds); + hdev->ioq_ptcmds = NULL; + + INIT_LIST_HEAD(&hdev->ioq_pt_list); +} + +static int spraid_submit_ioq_sync_cmd(struct spraid_dev *hdev, + struct spraid_ioq_command *cmd, + u32 *result, u32 *reslen, u32 timeout) +{ + int ret; + dma_addr_t sense_dma; + struct spraid_queue *ioq; + void *sense_addr = NULL; + struct spraid_cmd *pt_cmd = spraid_get_cmd(hdev, SPRAID_CMD_IOPT); + + if (!pt_cmd) { + dev_err(hdev->dev, "err, get ioq cmd failed\n"); + return -EFAULT; + } + + timeout = timeout ? timeout : ADMIN_TIMEOUT; + + init_completion(&pt_cmd->cmd_done); + + ioq = &hdev->queues[pt_cmd->qid]; + ret = pt_cmd->cid * SCSI_SENSE_BUFFERSIZE; + sense_addr = ioq->sense + ret; + sense_dma = ioq->sense_dma_addr + ret; + + cmd->common.sense_addr = cpu_to_le64(sense_dma); + cmd->common.sense_len = cpu_to_le16(SCSI_SENSE_BUFFERSIZE); + cmd->common.command_id = pt_cmd->cid; + + spraid_submit_cmd(ioq, cmd); + + if (!wait_for_completion_timeout(&pt_cmd->cmd_done, timeout)) { + dev_err(hdev->dev, "[%s] cid[%d] qid[%d] timeout;" + " opcode[0x%x] subopcode[0x%x]\n", + __func__, pt_cmd->cid, pt_cmd->qid, cmd->common.opcode, + (le32_to_cpu(cmd->common.cdw3[0]) & 0xffff)); + WRITE_ONCE(pt_cmd->state, SPRAID_CMD_TIMEOUT); + spraid_put_cmd(hdev, pt_cmd, SPRAID_CMD_IOPT); + return -EINVAL; + } + + if (result && reslen) { + if ((pt_cmd->status & 0x17f) == 0x101) { + memcpy(result, sense_addr, SCSI_SENSE_BUFFERSIZE); + *reslen = SCSI_SENSE_BUFFERSIZE; + } + } + + spraid_put_cmd(hdev, pt_cmd, SPRAID_CMD_IOPT); + + return pt_cmd->status; +} + +static int spraid_user_ioq_cmd(struct spraid_dev *hdev, struct bsg_job *job) +{ + struct spraid_bsg_request *bsg_req = + (struct spraid_bsg_request *)(job->request); + struct spraid_ioq_passthru_cmd *cmd = &(bsg_req->ioqcmd); + struct spraid_ioq_command ioq_cmd; + int status = 0; + u32 timeout = msecs_to_jiffies(cmd->timeout_ms); + + if (cmd->data_len > PAGE_SIZE) { + dev_err(hdev->dev, "[%s] data len bigger than 4k\n", __func__); + return -EFAULT; + } + + if (hdev->state != SPRAID_LIVE) { + dev_err(hdev->dev, "[%s] err, host state:[%d] is not live\n", + __func__, hdev->state); + return -EBUSY; + } + + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x] init;" + " datalen[%d]\n", + __func__, cmd->opcode, cmd->info_1.subopcode, cmd->data_len); + + memset(&ioq_cmd, 0, sizeof(ioq_cmd)); + ioq_cmd.common.opcode = cmd->opcode; + ioq_cmd.common.flags = cmd->flags; + ioq_cmd.common.hdid = cpu_to_le32(cmd->nsid); + ioq_cmd.common.sense_len = cpu_to_le16(cmd->info_0.res_sense_len); + ioq_cmd.common.cdb_len = cmd->info_0.cdb_len; + ioq_cmd.common.rsvd2 = cmd->info_0.rsvd0; + ioq_cmd.common.cdw3[0] = cpu_to_le32(cmd->cdw3); + ioq_cmd.common.cdw3[1] = cpu_to_le32(cmd->cdw4); + ioq_cmd.common.cdw3[2] = cpu_to_le32(cmd->cdw5); + + ioq_cmd.common.cdw10[0] = cpu_to_le32(cmd->cdw10); + ioq_cmd.common.cdw10[1] = cpu_to_le32(cmd->cdw11); + ioq_cmd.common.cdw10[2] = cpu_to_le32(cmd->cdw12); + ioq_cmd.common.cdw10[3] = cpu_to_le32(cmd->cdw13); + ioq_cmd.common.cdw10[4] = cpu_to_le32(cmd->cdw14); + ioq_cmd.common.cdw10[5] = cpu_to_le32(cmd->data_len); + + memcpy(ioq_cmd.common.cdb, &cmd->cdw16, cmd->info_0.cdb_len); + + ioq_cmd.common.cdw26[0] = cpu_to_le32(cmd->cdw26[0]); + ioq_cmd.common.cdw26[1] = cpu_to_le32(cmd->cdw26[1]); + ioq_cmd.common.cdw26[2] = cpu_to_le32(cmd->cdw26[2]); + ioq_cmd.common.cdw26[3] = cpu_to_le32(cmd->cdw26[3]); + + status = spraid_bsg_map_data(hdev, job, + (struct spraid_admin_command *)&ioq_cmd); + if (status) { + dev_err(hdev->dev, "[%s] err, map data failed\n", __func__); + return status; + } + + status = spraid_submit_ioq_sync_cmd(hdev, &ioq_cmd, job->reply, + &job->reply_len, timeout); + + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x], status[0x%x];" + " reply_len[%d]\n", + __func__, cmd->opcode, cmd->info_1.subopcode, + status, job->reply_len); + + spraid_bsg_unmap_data(hdev, job); + + return status; +} + +static bool spraid_check_scmd_completed(struct scsi_cmnd *scmd) +{ + struct spraid_dev *hdev = shost_priv(scmd->device->host); + struct spraid_iod *iod = scsi_cmd_priv(scmd); + struct spraid_queue *spraidq; + u16 hwq, cid; + + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + spraidq = &hdev->queues[hwq]; + if (READ_ONCE(iod->state) == SPRAID_CMD_COMPLETE + || spraid_poll_cq(spraidq, cid)) { + dev_warn(hdev->dev, "cid[%d] qid[%d] has been completed\n", + cid, spraidq->qid); + return true; + } + return false; +} + +static enum blk_eh_timer_return spraid_scmd_timeout(struct scsi_cmnd *scmd) +{ + struct spraid_iod *iod = scsi_cmd_priv(scmd); + unsigned int timeout = scmd->device->request_queue->rq_timeout; + + if (spraid_check_scmd_completed(scmd)) + goto out; + + if (time_after(jiffies, scmd->jiffies_at_alloc + timeout)) { + if (cmpxchg(&iod->state, + SPRAID_CMD_IN_FLIGHT, + SPRAID_CMD_TIMEOUT) == SPRAID_CMD_IN_FLIGHT) { + return BLK_EH_DONE; + } + } +out: + return BLK_EH_RESET_TIMER; +} + +/* send abort command by admin queue temporary */ +static int spraid_send_abort_cmd(struct spraid_dev *hdev, + u32 hdid, u16 qid, u16 cid) +{ + struct spraid_admin_command admin_cmd; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.abort.opcode = SPRAID_ADMIN_ABORT_CMD; + admin_cmd.abort.hdid = cpu_to_le32(hdid); + admin_cmd.abort.sqid = cpu_to_le16(qid); + admin_cmd.abort.cid = cpu_to_le16(cid); + + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); +} + +/* send reset command by admin quueue temporary */ +static int spraid_send_reset_cmd(struct spraid_dev *hdev, int type, u32 hdid) +{ + struct spraid_admin_command admin_cmd; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.reset.opcode = SPRAID_ADMIN_RESET; + admin_cmd.reset.hdid = cpu_to_le32(hdid); + admin_cmd.reset.type = type; + + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); +} + +static bool spraid_change_host_state(struct spraid_dev *hdev, + enum spraid_state newstate) +{ + unsigned long flags; + enum spraid_state oldstate; + bool change = false; + + spin_lock_irqsave(&hdev->state_lock, flags); + + oldstate = hdev->state; + switch (newstate) { + case SPRAID_LIVE: + switch (oldstate) { + case SPRAID_NEW: + case SPRAID_RESETTING: + change = true; + break; + default: + break; + } + break; + case SPRAID_RESETTING: + switch (oldstate) { + case SPRAID_LIVE: + change = true; + break; + default: + break; + } + break; + case SPRAID_DELETING: + if (oldstate != SPRAID_DELETING) + change = true; + break; + case SPRAID_DEAD: + switch (oldstate) { + case SPRAID_NEW: + case SPRAID_LIVE: + case SPRAID_RESETTING: + change = true; + break; + default: + break; + } + break; + default: + break; + } + if (change) + hdev->state = newstate; + spin_unlock_irqrestore(&hdev->state_lock, flags); + + dev_info(hdev->dev, "[%s][%d]->[%d], change[%d]\n", + __func__, oldstate, newstate, change); + + return change; +} + +static void spraid_back_fault_cqe(struct spraid_queue *ioq, + struct spraid_completion *cqe) +{ + struct spraid_dev *hdev = ioq->hdev; + struct blk_mq_tags *tags; + struct scsi_cmnd *scmd; + struct spraid_iod *iod; + struct request *req; + + tags = hdev->shost->tag_set.tags[ioq->qid - 1]; + req = blk_mq_tag_to_rq(tags, cqe->cmd_id); + if (unlikely(!req || !blk_mq_request_started(req))) + return; + + scmd = blk_mq_rq_to_pdu(req); + iod = scsi_cmd_priv(scmd); + + set_host_byte(scmd, DID_NO_CONNECT); + if (iod->nsge) + scsi_dma_unmap(scmd); + spraid_free_iod_res(hdev, iod); + scmd->scsi_done(scmd); + dev_warn(hdev->dev, "Back fault CQE, cid[%d] qid[%d]\n", + cqe->cmd_id, ioq->qid); +} + +static void spraid_back_all_io(struct spraid_dev *hdev) +{ + int i, j; + struct spraid_queue *ioq; + struct spraid_completion cqe = { 0 }; + + scsi_block_requests(hdev->shost); + + for (i = 1; i <= hdev->shost->nr_hw_queues; i++) { + ioq = &hdev->queues[i]; + for (j = 0; j < hdev->shost->can_queue; j++) { + cqe.cmd_id = j; + spraid_back_fault_cqe(ioq, &cqe); + } + } + + scsi_unblock_requests(hdev->shost); +} + +static void spraid_dev_disable(struct spraid_dev *hdev, bool shutdown) +{ + struct spraid_queue *adminq = &hdev->queues[0]; + u16 start, end; + unsigned long timeout = jiffies + 600 * HZ; + + if (pci_device_is_present(hdev->pdev)) { + if (shutdown) + spraid_shutdown_ctrl(hdev); + else + spraid_disable_ctrl(hdev); + } + + while (!time_after(jiffies, timeout)) { + if (!pci_device_is_present(hdev->pdev)) { + dev_info(hdev->dev, "[%s] pci_device not present;" + " skip wait\n", __func__); + break; + } + if (!spraid_wait_ready(hdev, hdev->cap, false)) { + dev_info(hdev->dev, + "[%s] wait ready success after reset\n", + __func__); + break; + } + dev_info(hdev->dev, "[%s] waiting csts_rdy ready\n", __func__); + } + + if (hdev->queue_count == 0) { + dev_err(hdev->dev, "[%s] warn, queue has been delete\n", + __func__); + return; + } + + spin_lock_irq(&adminq->cq_lock); + spraid_process_cq(adminq, &start, &end, -1); + spin_unlock_irq(&adminq->cq_lock); + spraid_complete_cqes(adminq, start, end); + + spraid_pci_disable(hdev); + + spraid_back_all_io(hdev); +} + +static void spraid_reset_work(struct work_struct *work) +{ + int ret; + struct spraid_dev *hdev = + container_of(work, struct spraid_dev, reset_work); + + if (hdev->state != SPRAID_RESETTING) { + dev_err(hdev->dev, "[%s] err, host is not reset state\n", + __func__); + return; + } + + dev_info(hdev->dev, "[%s] enter host reset\n", __func__); + + if (hdev->ctrl_config & SPRAID_CC_ENABLE) { + dev_info(hdev->dev, "[%s] start dev_disable\n", __func__); + spraid_dev_disable(hdev, false); + } + + ret = spraid_pci_enable(hdev); + if (ret) + goto out; + + ret = spraid_setup_admin_queue(hdev); + if (ret) + goto pci_disable; + + ret = spraid_setup_io_queues(hdev); + if (ret || hdev->online_queues <= hdev->shost->nr_hw_queues) + goto pci_disable; + + spraid_change_host_state(hdev, SPRAID_LIVE); + + spraid_send_all_aen(hdev); + + return; + +pci_disable: + spraid_pci_disable(hdev); +out: + spraid_change_host_state(hdev, SPRAID_DEAD); + dev_err(hdev->dev, "[%s] err, host reset failed\n", __func__); +} + +static int spraid_reset_work_sync(struct spraid_dev *hdev) +{ + if (!spraid_change_host_state(hdev, SPRAID_RESETTING)) { + dev_info(hdev->dev, "[%s] can't change to reset state\n", + __func__); + return -EBUSY; + } + + if (!queue_work(spraid_wq, &hdev->reset_work)) { + dev_err(hdev->dev, "[%s] err, host is already in reset state\n", + __func__); + return -EBUSY; + } + + flush_work(&hdev->reset_work); + if (hdev->state != SPRAID_LIVE) + return -ENODEV; + + return 0; +} + +static int spraid_wait_abnl_cmd_done(struct spraid_iod *iod) +{ + u16 times = 0; + + do { + if (READ_ONCE(iod->state) == SPRAID_CMD_TMO_COMPLETE) + break; + msleep(500); + times++; + } while (times <= SPRAID_WAIT_ABNL_CMD_TIMEOUT); + + /* wait command completion timeout after abort/reset success */ + if (times >= SPRAID_WAIT_ABNL_CMD_TIMEOUT) + return -ETIMEDOUT; + + return 0; +} + +static int spraid_abort_handler(struct scsi_cmnd *scmd) +{ + struct spraid_dev *hdev = shost_priv(scmd->device->host); + struct spraid_iod *iod = scsi_cmd_priv(scmd); + struct spraid_sdev_hostdata *hostdata; + u16 hwq, cid; + int ret; + + scsi_print_command(scmd); + + if (hdev->state != SPRAID_LIVE || !spraid_wait_abnl_cmd_done(iod) || + spraid_check_scmd_completed(scmd)) + return SUCCESS; + + hostdata = scmd->device->hostdata; + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + + dev_warn(hdev->dev, "cid[%d] qid[%d] timeout, aborting\n", cid, hwq); + ret = spraid_send_abort_cmd(hdev, hostdata->hdid, hwq, cid); + if (ret != ADMIN_ERR_TIMEOUT) { + ret = spraid_wait_abnl_cmd_done(iod); + if (ret) { + dev_warn(hdev->dev, "cid[%d] qid[%d] abort failed;" + " not found\n", cid, hwq); + return FAILED; + } + dev_warn(hdev->dev, "cid[%d] qid[%d] abort succ\n", cid, hwq); + return SUCCESS; + } + dev_warn(hdev->dev, "cid[%d] qid[%d] abort failed, timeout\n", + cid, hwq); + return FAILED; +} + +static int spraid_tgt_reset_handler(struct scsi_cmnd *scmd) +{ + struct spraid_dev *hdev = shost_priv(scmd->device->host); + struct spraid_iod *iod = scsi_cmd_priv(scmd); + struct spraid_sdev_hostdata *hostdata; + u16 hwq, cid; + int ret; + + scsi_print_command(scmd); + + if (hdev->state != SPRAID_LIVE || !spraid_wait_abnl_cmd_done(iod) || + spraid_check_scmd_completed(scmd)) + return SUCCESS; + + hostdata = scmd->device->hostdata; + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + + dev_warn(hdev->dev, "cid[%d] qid[%d] timeout, target reset\n", + cid, hwq); + ret = spraid_send_reset_cmd(hdev, SPRAID_RESET_TARGET, hostdata->hdid); + if (ret == 0) { + ret = spraid_wait_abnl_cmd_done(iod); + if (ret) { + dev_warn(hdev->dev, + "cid[%d] qid[%d]target reset failed;" + " not found\n", cid, hwq); + return FAILED; + } + + dev_warn(hdev->dev, "cid[%d] qid[%d] target reset success\n", + cid, hwq); + return SUCCESS; + } + + dev_warn(hdev->dev, "cid[%d] qid[%d] ret[%d] target reset failed\n", + cid, hwq, ret); + return FAILED; +} + +static int spraid_bus_reset_handler(struct scsi_cmnd *scmd) +{ + struct spraid_dev *hdev = shost_priv(scmd->device->host); + struct spraid_iod *iod = scsi_cmd_priv(scmd); + struct spraid_sdev_hostdata *hostdata; + u16 hwq, cid; + int ret; + + scsi_print_command(scmd); + + if (hdev->state != SPRAID_LIVE || !spraid_wait_abnl_cmd_done(iod) || + spraid_check_scmd_completed(scmd)) + return SUCCESS; + + hostdata = scmd->device->hostdata; + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + + dev_warn(hdev->dev, "cid[%d] qid[%d] timeout, bus reset\n", cid, hwq); + ret = spraid_send_reset_cmd(hdev, SPRAID_RESET_BUS, hostdata->hdid); + if (ret == 0) { + ret = spraid_wait_abnl_cmd_done(iod); + if (ret) { + dev_warn(hdev->dev, + "cid[%d] qid[%d] bus reset failed;" + " not found\n", cid, hwq); + return FAILED; + } + + dev_warn(hdev->dev, "cid[%d] qid[%d] bus reset succ\n", + cid, hwq); + return SUCCESS; + } + + dev_warn(hdev->dev, "cid[%d] qid[%d] ret[%d] bus reset failed\n", + cid, hwq, ret); + return FAILED; +} + +static int spraid_shost_reset_handler(struct scsi_cmnd *scmd) +{ + u16 hwq, cid; + struct spraid_dev *hdev = shost_priv(scmd->device->host); + + scsi_print_command(scmd); + if (hdev->state != SPRAID_LIVE || spraid_check_scmd_completed(scmd)) + return SUCCESS; + + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + dev_warn(hdev->dev, "cid[%d] qid[%d] host reset\n", cid, hwq); + + if (spraid_reset_work_sync(hdev)) { + dev_warn(hdev->dev, "cid[%d] qid[%d] host reset failed\n", + cid, hwq); + return FAILED; + } + + dev_warn(hdev->dev, "cid[%d] qid[%d] host reset success\n", cid, hwq); + + return SUCCESS; +} + +static pci_ers_result_t spraid_pci_error_detected(struct pci_dev *pdev, + pci_channel_state_t state) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "enter pci error detect, state:%d\n", state); + + switch (state) { + case pci_channel_io_normal: + dev_warn(hdev->dev, "channel is normal, do nothing\n"); + + return PCI_ERS_RESULT_CAN_RECOVER; + case pci_channel_io_frozen: + dev_warn(hdev->dev, + "channel io frozen, need reset controller\n"); + + scsi_block_requests(hdev->shost); + + spraid_change_host_state(hdev, SPRAID_RESETTING); + + return PCI_ERS_RESULT_NEED_RESET; + case pci_channel_io_perm_failure: + dev_warn(hdev->dev, "channel io failure, request disconnect\n"); + + return PCI_ERS_RESULT_DISCONNECT; + } + + return PCI_ERS_RESULT_NEED_RESET; +} + +static pci_ers_result_t spraid_pci_slot_reset(struct pci_dev *pdev) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "restart after slot reset\n"); + + pci_restore_state(pdev); + + if (!queue_work(spraid_wq, &hdev->reset_work)) { + dev_err(hdev->dev, "[%s] err, the device is resetting state\n", + __func__); + return PCI_ERS_RESULT_NONE; + } + + flush_work(&hdev->reset_work); + + scsi_unblock_requests(hdev->shost); + + return PCI_ERS_RESULT_RECOVERED; +} + +static void spraid_reset_done(struct pci_dev *pdev) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "enter spraid reset done\n"); +} + +static ssize_t csts_pp_show(struct device *cdev, + struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct spraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) { + ret = (readl(hdev->bar + SPRAID_REG_CSTS) + & SPRAID_CSTS_PP_MASK); + ret >>= SPRAID_CSTS_PP_SHIFT; + } + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t csts_shst_show(struct device *cdev, + struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct spraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) { + ret = (readl(hdev->bar + SPRAID_REG_CSTS) + & SPRAID_CSTS_SHST_MASK); + ret >>= SPRAID_CSTS_SHST_SHIFT; + } + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t csts_cfs_show(struct device *cdev, + struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct spraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) { + ret = (readl(hdev->bar + SPRAID_REG_CSTS) + & SPRAID_CSTS_CFS_MASK); + ret >>= SPRAID_CSTS_CFS_SHIFT; + } + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t csts_rdy_show(struct device *cdev, + struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct spraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) + ret = (readl(hdev->bar + SPRAID_REG_CSTS) & SPRAID_CSTS_RDY); + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t fw_version_show(struct device *cdev, + struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct spraid_dev *hdev = shost_priv(shost); + + return snprintf(buf, PAGE_SIZE, "%s\n", hdev->ctrl_info->fr); +} + +static DEVICE_ATTR_RO(csts_pp); +static DEVICE_ATTR_RO(csts_shst); +static DEVICE_ATTR_RO(csts_cfs); +static DEVICE_ATTR_RO(csts_rdy); +static DEVICE_ATTR_RO(fw_version); + +static struct device_attribute *spraid_host_attrs[] = { + &dev_attr_csts_pp, + &dev_attr_csts_shst, + &dev_attr_csts_cfs, + &dev_attr_csts_rdy, + &dev_attr_fw_version, + NULL, +}; + +static int spraid_get_vd_info(struct spraid_dev *hdev, + struct spraid_vd_info *vd_info, u16 vid) +{ + struct spraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + int ret; + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, + &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.usr_cmd.opcode = USR_CMD_READ; + admin_cmd.usr_cmd.info_0.subopcode = cpu_to_le16(USR_CMD_VDINFO); + admin_cmd.usr_cmd.info_1.data_len = cpu_to_le16(USR_CMD_RDLEN); + admin_cmd.usr_cmd.info_1.param_len = cpu_to_le16(VDINFO_PARAM_LEN); + admin_cmd.usr_cmd.cdw10 = cpu_to_le32(vid); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + if (!ret) + memcpy(vd_info, data_ptr, sizeof(struct spraid_vd_info)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, data_dma); + + return ret; +} + +static int spraid_get_bgtask(struct spraid_dev *hdev, + struct spraid_bgtask *bgtask) +{ + struct spraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + int ret; + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, + &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.usr_cmd.opcode = USR_CMD_READ; + admin_cmd.usr_cmd.info_0.subopcode = cpu_to_le16(USR_CMD_BGTASK); + admin_cmd.usr_cmd.info_1.data_len = cpu_to_le16(USR_CMD_RDLEN); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + if (!ret) + memcpy(bgtask, data_ptr, sizeof(struct spraid_bgtask)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, data_dma); + + return ret; +} + +static ssize_t raid_level_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct spraid_dev *hdev; + struct spraid_vd_info *vd_info; + struct spraid_sdev_hostdata *hostdata; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !SPRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = spraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) + vd_info->rg_level = ARRAY_SIZE(raid_levels) - 1; + + ret = (vd_info->rg_level < ARRAY_SIZE(raid_levels)) ? + vd_info->rg_level : (ARRAY_SIZE(raid_levels) - 1); + + kfree(vd_info); + + return snprintf(buf, PAGE_SIZE, "RAID-%s\n", raid_levels[ret]); +} + +static ssize_t raid_state_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct spraid_dev *hdev; + struct spraid_vd_info *vd_info; + struct spraid_sdev_hostdata *hostdata; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !SPRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = spraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) { + vd_info->vd_status = 0; + vd_info->rg_id = 0xff; + } + + ret = (vd_info->vd_status < ARRAY_SIZE(raid_states)) ? + vd_info->vd_status : 0; + + kfree(vd_info); + + return snprintf(buf, PAGE_SIZE, "%s\n", raid_states[ret]); +} + +static ssize_t raid_resync_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct spraid_dev *hdev; + struct spraid_vd_info *vd_info; + struct spraid_bgtask *bgtask; + struct spraid_sdev_hostdata *hostdata; + u8 rg_id, i, progress = 0; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !SPRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = spraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) + goto out; + + rg_id = vd_info->rg_id; + + bgtask = (struct spraid_bgtask *)vd_info; + ret = spraid_get_bgtask(hdev, bgtask); + if (ret) + goto out; + for (i = 0; i < bgtask->task_num; i++) { + if ((bgtask->bgtask[i].type == BGTASK_TYPE_REBUILD) && + (le16_to_cpu(bgtask->bgtask[i].vd_id) == rg_id)) + progress = bgtask->bgtask[i].progress; + } + +out: + kfree(vd_info); + return snprintf(buf, PAGE_SIZE, "%d\n", progress); +} + +static DEVICE_ATTR_RO(raid_level); +static DEVICE_ATTR_RO(raid_state); +static DEVICE_ATTR_RO(raid_resync); + +static struct device_attribute *spraid_dev_attrs[] = { + &dev_attr_raid_level, + &dev_attr_raid_state, + &dev_attr_raid_resync, + NULL, +}; + +static struct pci_error_handlers spraid_err_handler = { + .error_detected = spraid_pci_error_detected, + .slot_reset = spraid_pci_slot_reset, + .reset_done = spraid_reset_done, +}; + +static int spraid_sysfs_host_reset(struct Scsi_Host *shost, int reset_type) +{ + int ret; + struct spraid_dev *hdev = shost_priv(shost); + + dev_info(hdev->dev, "[%s] start sysfs host reset cmd\n", __func__); + ret = spraid_reset_work_sync(hdev); + dev_info(hdev->dev, "[%s] stop sysfs host reset cmd[%d]\n", + __func__, ret); + + return ret; +} + +static struct scsi_host_template spraid_driver_template = { + .module = THIS_MODULE, + .name = "Ramaxel Logic spraid driver", + .proc_name = "spraid", + .queuecommand = spraid_queue_command, + .slave_alloc = spraid_slave_alloc, + .slave_destroy = spraid_slave_destroy, + .slave_configure = spraid_slave_configure, + .eh_timed_out = spraid_scmd_timeout, + .eh_abort_handler = spraid_abort_handler, + .eh_target_reset_handler = spraid_tgt_reset_handler, + .eh_bus_reset_handler = spraid_bus_reset_handler, + .eh_host_reset_handler = spraid_shost_reset_handler, + .change_queue_depth = scsi_change_queue_depth, + .this_id = -1, + .shost_attrs = spraid_host_attrs, + .sdev_attrs = spraid_dev_attrs, + .host_reset = spraid_sysfs_host_reset, +}; + +static void spraid_shutdown(struct pci_dev *pdev) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + + spraid_remove_io_queues(hdev); + spraid_disable_admin_queue(hdev, true); +} + +/* bsg dispatch user command */ +static int spraid_bsg_host_dispatch(struct bsg_job *job) +{ + struct Scsi_Host *shost = dev_to_shost(job->dev); + struct spraid_dev *hdev = shost_priv(shost); + struct request *rq = blk_mq_rq_from_pdu(job); + struct spraid_bsg_request *bsg_req = job->request; + int ret = 0; + + dev_info(hdev->dev, "[%s] msgcode[%d], msglen[%d], timeout[%d];" + " req_nsge[%d], req_len[%d]\n", + __func__, bsg_req->msgcode, job->request_len, + rq->timeout, job->request_payload.sg_cnt, + job->request_payload.payload_len); + + job->reply_len = 0; + + switch (bsg_req->msgcode) { + case SPRAID_BSG_ADM: + ret = spraid_user_admin_cmd(hdev, job); + break; + case SPRAID_BSG_IOQ: + ret = spraid_user_ioq_cmd(hdev, job); + break; + default: + dev_info(hdev->dev, "[%s] unsupport msgcode[%d]\n", + __func__, bsg_req->msgcode); + break; + } + + if (ret > 0) + ret = ret | (ret << 8); + + bsg_job_done(job, ret, 0); + return 0; +} + +static inline void spraid_remove_bsg(struct spraid_dev *hdev) +{ + if (hdev->bsg_queue) { + bsg_unregister_queue(hdev->bsg_queue); + blk_cleanup_queue(hdev->bsg_queue); + } +} +static int spraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) +{ + struct spraid_dev *hdev; + struct Scsi_Host *shost; + int node, ret; + char bsg_name[15]; + + shost = scsi_host_alloc(&spraid_driver_template, sizeof(*hdev)); + if (!shost) { + dev_err(&pdev->dev, "Failed to allocate scsi host\n"); + return -ENOMEM; + } + hdev = shost_priv(shost); + hdev->pdev = pdev; + hdev->dev = get_device(&pdev->dev); + + node = dev_to_node(hdev->dev); + if (node == NUMA_NO_NODE) { + node = first_memory_node; + set_dev_node(hdev->dev, node); + } + hdev->numa_node = node; + hdev->shost = shost; + pci_set_drvdata(pdev, hdev); + + ret = spraid_dev_map(hdev); + if (ret) + goto put_dev; + + init_rwsem(&hdev->devices_rwsem); + INIT_WORK(&hdev->scan_work, spraid_scan_work); + INIT_WORK(&hdev->timesyn_work, spraid_timesyn_work); + INIT_WORK(&hdev->reset_work, spraid_reset_work); + INIT_WORK(&hdev->fw_act_work, spraid_fw_act_work); + spin_lock_init(&hdev->state_lock); + + ret = spraid_alloc_resources(hdev); + if (ret) + goto dev_unmap; + + ret = spraid_pci_enable(hdev); + if (ret) + goto resources_free; + + ret = spraid_setup_admin_queue(hdev); + if (ret) + goto pci_disable; + + ret = spraid_init_ctrl_info(hdev); + if (ret) + goto disable_admin_q; + + ret = spraid_alloc_iod_ext_mem_pool(hdev); + if (ret) + goto disable_admin_q; + + ret = spraid_setup_io_queues(hdev); + if (ret) + goto free_iod_mempool; + + spraid_shost_init(hdev); + + ret = scsi_add_host(hdev->shost, hdev->dev); + if (ret) { + dev_err(hdev->dev, "Add shost to system failed, ret: %d\n", + ret); + goto remove_io_queues; + } + + snprintf(bsg_name, sizeof(bsg_name), "spraid%d", shost->host_no); + hdev->bsg_queue = bsg_setup_queue(&shost->shost_gendev, bsg_name, + spraid_bsg_host_dispatch, + spraid_cmd_size(hdev, true, false)); + if (IS_ERR(hdev->bsg_queue)) { + dev_err(hdev->dev, "err, setup bsg failed\n"); + hdev->bsg_queue = NULL; + goto remove_io_queues; + } + + if (hdev->online_queues == SPRAID_ADMIN_QUEUE_NUM) { + dev_warn(hdev->dev, "warn only admin queue can be used\n"); + return 0; + } + + hdev->state = SPRAID_LIVE; + + spraid_send_all_aen(hdev); + + ret = spraid_dev_list_init(hdev); + if (ret) + goto remove_bsg; + + ret = spraid_configure_timestamp(hdev); + if (ret) + dev_warn(hdev->dev, "init set timestamp failed\n"); + + ret = spraid_alloc_ioq_ptcmds(hdev); + if (ret) + goto remove_bsg; + + scsi_scan_host(hdev->shost); + + return 0; + +remove_bsg: + spraid_remove_bsg(hdev); +remove_io_queues: + spraid_remove_io_queues(hdev); +free_iod_mempool: + spraid_free_iod_ext_mem_pool(hdev); +disable_admin_q: + spraid_disable_admin_queue(hdev, false); +pci_disable: + spraid_pci_disable(hdev); +resources_free: + spraid_free_resources(hdev); +dev_unmap: + spraid_dev_unmap(hdev); +put_dev: + put_device(hdev->dev); + scsi_host_put(shost); + + return -ENODEV; +} + +static void spraid_remove(struct pci_dev *pdev) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + struct Scsi_Host *shost = hdev->shost; + + dev_info(hdev->dev, "enter spraid remove\n"); + + spraid_change_host_state(hdev, SPRAID_DELETING); + flush_work(&hdev->reset_work); + + if (!pci_device_is_present(pdev)) + spraid_back_all_io(hdev); + + spraid_remove_bsg(hdev); + scsi_remove_host(shost); + spraid_free_ioq_ptcmds(hdev); + kfree(hdev->devices); + spraid_remove_io_queues(hdev); + spraid_free_iod_ext_mem_pool(hdev); + spraid_disable_admin_queue(hdev, false); + spraid_pci_disable(hdev); + spraid_free_resources(hdev); + spraid_dev_unmap(hdev); + put_device(hdev->dev); + scsi_host_put(shost); + + dev_info(hdev->dev, "exit spraid remove\n"); +} + +static const struct pci_device_id spraid_id_table[] = { + { PCI_DEVICE(PCI_VENDOR_ID_RAMAXEL_LOGIC, + SPRAID_SERVER_DEVICE_HBA_DID) }, + { PCI_DEVICE(PCI_VENDOR_ID_RAMAXEL_LOGIC, + SPRAID_SERVER_DEVICE_RAID_DID) }, + { 0, } +}; +MODULE_DEVICE_TABLE(pci, spraid_id_table); + +static struct pci_driver spraid_driver = { + .name = "spraid", + .id_table = spraid_id_table, + .probe = spraid_probe, + .remove = spraid_remove, + .shutdown = spraid_shutdown, + .err_handler = &spraid_err_handler, +}; + +static int __init spraid_init(void) +{ + int ret; + + spraid_wq = alloc_workqueue("spraid-wq", WQ_UNBOUND | WQ_MEM_RECLAIM | + WQ_SYSFS, 0); + if (!spraid_wq) + return -ENOMEM; + + spraid_class = class_create(THIS_MODULE, "spraid"); + if (IS_ERR(spraid_class)) { + ret = PTR_ERR(spraid_class); + goto destroy_wq; + } + + ret = pci_register_driver(&spraid_driver); + if (ret < 0) + goto destroy_class; + + return 0; + +destroy_class: + class_destroy(spraid_class); +destroy_wq: + destroy_workqueue(spraid_wq); + + return ret; +} + +static void __exit spraid_exit(void) +{ + pci_unregister_driver(&spraid_driver); + class_destroy(spraid_class); + destroy_workqueue(spraid_wq); + ida_destroy(&spraid_instance_ida); +} + +MODULE_AUTHOR("songyl(a)ramaxel.com"); +MODULE_DESCRIPTION("Ramaxel Memory Technology SPraid Driver"); +MODULE_LICENSE("GPL"); +MODULE_VERSION(SPRAID_DRV_VERSION); +module_init(spraid_init); +module_exit(spraid_exit); -- 2.25.1

1 1

[PATCH OLK-5.10 1/2] reboot: allow to specify reboot mode via
by Hongyu Li 22 Dec '21

22 Dec '21

From: Matteo Croce <mcroce(a)microsoft.com> maillist inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4BULZ CVE: NA Reference: https://lore.kernel.org/all/20201110202746.9690-1-mcroce@linux.microsoft.co… ---------------------------------------------------------------------- The kernel cmdline reboot= option offers some sort of control on how the reboot is issued. Add handles in sysfs to allow setting these reboot options, so they can be changed when the system is booted, other than at boot time. The handlers are under <sysfs>/kernel/reboot, can be read to get the current configuration and written to alter it. # cd /sys/kernel/reboot/ # grep . * cpu:0 force:0 mode:cold type:acpi # echo 2 >cpu # echo yes >force # echo soft >mode # echo bios >type # grep . * cpu:2 force:1 mode:soft type:bios Before setting anything, check for CAP_SYS_BOOT capability, so it's possible to allow an unpriviledged process to change these settings simply by relaxing the handles permissions, without opening them to the world. Signed-off-by: Matteo Croce <mcroce(a)microsoft.com> Signed-off-by: Hongyu Li <543306408(a)qq.com> --- Documentation/ABI/testing/sysfs-kernel-reboot | 32 +++ kernel/reboot.c | 206 ++++++++++++++++++ 2 files changed, 238 insertions(+) create mode 100644 Documentation/ABI/testing/sysfs-kernel-reboot diff --git a/Documentation/ABI/testing/sysfs-kernel-reboot b/Documentation/ABI/testing/sysfs-kernel-reboot new file mode 100644 index 000000000000..837330fb2511 --- /dev/null +++ b/Documentation/ABI/testing/sysfs-kernel-reboot @@ -0,0 +1,32 @@ +What: /sys/kernel/reboot +Date: November 2020 +KernelVersion: 5.11 +Contact: Matteo Croce <mcroce(a)microsoft.com> +Description: Interface to set the kernel reboot behavior, similarly to + what can be done via the reboot= cmdline option. + (see Documentation/admin-guide/kernel-parameters.txt) + +What: /sys/kernel/reboot/mode +Date: November 2020 +KernelVersion: 5.11 +Contact: Matteo Croce <mcroce(a)microsoft.com> +Description: Reboot mode. Valid values are: cold warm hard soft gpio + +What: /sys/kernel/reboot/type +Date: November 2020 +KernelVersion: 5.11 +Contact: Matteo Croce <mcroce(a)microsoft.com> +Description: Reboot type. Valid values are: bios acpi kbd triple efi pci + +What: /sys/kernel/reboot/cpu +Date: November 2020 +KernelVersion: 5.11 +Contact: Matteo Croce <mcroce(a)microsoft.com> +Description: CPU number to use to reboot. + +What: /sys/kernel/reboot/force +Date: November 2020 +KernelVersion: 5.11 +Contact: Matteo Croce <mcroce(a)microsoft.com> +Description: Don't wait for any other CPUs on reboot and + avoid anything that could hang. diff --git a/kernel/reboot.c b/kernel/reboot.c index af6f23d8bea1..778979d92dc7 100644 --- a/kernel/reboot.c +++ b/kernel/reboot.c @@ -594,3 +594,209 @@ static int __init reboot_setup(char *str) return 1; } __setup("reboot=", reboot_setup); + +#ifdef CONFIG_SYSFS + +#define REBOOT_COLD_STR "cold" +#define REBOOT_WARM_STR "warm" +#define REBOOT_HARD_STR "hard" +#define REBOOT_SOFT_STR "soft" +#define REBOOT_GPIO_STR "gpio" +#define REBOOT_UNDEFINED_STR "undefined" + +#define BOOT_TRIPLE_STR "triple" +#define BOOT_KBD_STR "kbd" +#define BOOT_BIOS_STR "bios" +#define BOOT_ACPI_STR "acpi" +#define BOOT_EFI_STR "efi" +#define BOOT_CF9_FORCE_STR "cf9_force" +#define BOOT_CF9_SAFE_STR "cf9_safe" + +static ssize_t mode_show(struct kobject *kobj, struct kobj_attribute *attr, char *buf) +{ + const char *val; + + switch (reboot_mode) { + case REBOOT_COLD: + val = REBOOT_COLD_STR; + break; + case REBOOT_WARM: + val = REBOOT_WARM_STR; + break; + case REBOOT_HARD: + val = REBOOT_HARD_STR; + break; + case REBOOT_SOFT: + val = REBOOT_SOFT_STR; + break; + case REBOOT_GPIO: + val = REBOOT_GPIO_STR; + break; + default: + val = REBOOT_UNDEFINED_STR; + } + + return sprintf(buf, "%s\n", val); +} +static ssize_t mode_store(struct kobject *kobj, struct kobj_attribute *attr, + const char *buf, size_t count) +{ + if (!capable(CAP_SYS_BOOT)) + return -EPERM; + + if (!strncmp(buf, REBOOT_COLD_STR, strlen(REBOOT_COLD_STR))) + reboot_mode = REBOOT_COLD; + else if (!strncmp(buf, REBOOT_WARM_STR, strlen(REBOOT_WARM_STR))) + reboot_mode = REBOOT_WARM; + else if (!strncmp(buf, REBOOT_HARD_STR, strlen(REBOOT_HARD_STR))) + reboot_mode = REBOOT_HARD; + else if (!strncmp(buf, REBOOT_SOFT_STR, strlen(REBOOT_SOFT_STR))) + reboot_mode = REBOOT_SOFT; + else if (!strncmp(buf, REBOOT_GPIO_STR, strlen(REBOOT_GPIO_STR))) + reboot_mode = REBOOT_GPIO; + else + return -EINVAL; + + return count; +} +static struct kobj_attribute reboot_mode_attr = __ATTR_RW(mode); + +static ssize_t type_show(struct kobject *kobj, struct kobj_attribute *attr, char *buf) +{ + const char *val; + + switch (reboot_type) { + case BOOT_TRIPLE: + val = BOOT_TRIPLE_STR; + break; + case BOOT_KBD: + val = BOOT_KBD_STR; + break; + case BOOT_BIOS: + val = BOOT_BIOS_STR; + break; + case BOOT_ACPI: + val = BOOT_ACPI_STR; + break; + case BOOT_EFI: + val = BOOT_EFI_STR; + break; + case BOOT_CF9_FORCE: + val = BOOT_CF9_FORCE_STR; + break; + case BOOT_CF9_SAFE: + val = BOOT_CF9_SAFE_STR; + break; + default: + val = REBOOT_UNDEFINED_STR; + } + + return sprintf(buf, "%s\n", val); +} +static ssize_t type_store(struct kobject *kobj, struct kobj_attribute *attr, + const char *buf, size_t count) +{ + if (!capable(CAP_SYS_BOOT)) + return -EPERM; + + if (!strncmp(buf, BOOT_TRIPLE_STR, strlen(BOOT_TRIPLE_STR))) + reboot_mode = BOOT_TRIPLE; + else if (!strncmp(buf, BOOT_KBD_STR, strlen(BOOT_KBD_STR))) + reboot_mode = BOOT_KBD; + else if (!strncmp(buf, BOOT_BIOS_STR, strlen(BOOT_BIOS_STR))) + reboot_mode = BOOT_BIOS; + else if (!strncmp(buf, BOOT_ACPI_STR, strlen(BOOT_ACPI_STR))) + reboot_mode = BOOT_ACPI; + else if (!strncmp(buf, BOOT_EFI_STR, strlen(BOOT_EFI_STR))) + reboot_mode = BOOT_EFI; + else if (!strncmp(buf, BOOT_CF9_FORCE_STR, strlen(BOOT_CF9_FORCE_STR))) + reboot_mode = BOOT_CF9_FORCE; + else if (!strncmp(buf, BOOT_CF9_SAFE_STR, strlen(BOOT_CF9_SAFE_STR))) + reboot_mode = BOOT_CF9_SAFE; + else + return -EINVAL; + + return count; +} +static struct kobj_attribute reboot_type_attr = __ATTR_RW(type); + +static ssize_t cpu_show(struct kobject *kobj, struct kobj_attribute *attr, char *buf) +{ + return sprintf(buf, "%d\n", reboot_cpu); +} +static ssize_t cpu_store(struct kobject *kobj, struct kobj_attribute *attr, + const char *buf, size_t count) +{ + unsigned int cpunum; + int rc; + + if (!capable(CAP_SYS_BOOT)) + return -EPERM; + + rc = kstrtouint(buf, 0, &cpunum); + + if (rc) + return rc; + + if (cpunum >= num_possible_cpus()) + return -ERANGE; + + reboot_cpu = cpunum; + + return count; +} +static struct kobj_attribute reboot_cpu_attr = __ATTR_RW(cpu); + +static ssize_t force_show(struct kobject *kobj, struct kobj_attribute *attr, char *buf) +{ + return sprintf(buf, "%d\n", reboot_force); +} +static ssize_t force_store(struct kobject *kobj, struct kobj_attribute *attr, + const char *buf, size_t count) +{ + bool res; + + if (!capable(CAP_SYS_BOOT)) + return -EPERM; + + if (kstrtobool(buf, &res)) + return -EINVAL; + + reboot_force = res; + + return count; +} +static struct kobj_attribute reboot_force_attr = __ATTR_RW(force); + +static struct attribute *reboot_attrs[] = { + &reboot_mode_attr.attr, + &reboot_type_attr.attr, + &reboot_cpu_attr.attr, + &reboot_force_attr.attr, + NULL, +}; + +static const struct attribute_group reboot_attr_group = { + .attrs = reboot_attrs, +}; + +static int __init reboot_ksysfs_init(void) +{ + struct kobject *reboot_kobj; + int ret; + + reboot_kobj = kobject_create_and_add("reboot", kernel_kobj); + if (!reboot_kobj) + return -ENOMEM; + + ret = sysfs_create_group(reboot_kobj, &reboot_attr_group); + if (ret) { + kobject_put(reboot_kobj); + return ret; + } + + return 0; +} +late_initcall(reboot_ksysfs_init); + +#endif -- 2.17.1

3 2

[PATCH OLK-5.10 2/2] reboot: Fix variable assignments in
by Hongyu Li 22 Dec '21

22 Dec '21

From: Nathan Chancellor <natechancellor(a)gmail.com> maillist inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4BULZ CVE: NA Reference: https://lore.kernel.org/all/20201112035023.974748-1-natechancellor@gmail.co… ---------------------------------------------------------------------- Clang warns: kernel/reboot.c:707:17: warning: implicit conversion from enumeration type 'enum reboot_type' to different enumeration type 'enum reboot_mode' [-Wenum-conversion] reboot_mode = BOOT_TRIPLE; ~ ^~~~~~~~~~~ kernel/reboot.c:709:17: warning: implicit conversion from enumeration type 'enum reboot_type' to different enumeration type 'enum reboot_mode' [-Wenum-conversion] reboot_mode = BOOT_KBD; ~ ^~~~~~~~ kernel/reboot.c:711:17: warning: implicit conversion from enumeration type 'enum reboot_type' to different enumeration type 'enum reboot_mode' [-Wenum-conversion] reboot_mode = BOOT_BIOS; ~ ^~~~~~~~~ kernel/reboot.c:713:17: warning: implicit conversion from enumeration type 'enum reboot_type' to different enumeration type 'enum reboot_mode' [-Wenum-conversion] reboot_mode = BOOT_ACPI; ~ ^~~~~~~~~ kernel/reboot.c:715:17: warning: implicit conversion from enumeration type 'enum reboot_type' to different enumeration type 'enum reboot_mode' [-Wenum-conversion] reboot_mode = BOOT_EFI; ~ ^~~~~~~~ kernel/reboot.c:717:17: warning: implicit conversion from enumeration type 'enum reboot_type' to different enumeration type 'enum reboot_mode' [-Wenum-conversion] reboot_mode = BOOT_CF9_FORCE; ~ ^~~~~~~~~~~~~~ kernel/reboot.c:719:17: warning: implicit conversion from enumeration type 'enum reboot_type' to different enumeration type 'enum reboot_mode' [-Wenum-conversion] reboot_mode = BOOT_CF9_SAFE; ~ ^~~~~~~~~~~~~ 7 warnings generated. It seems that these assignment should be to reboot_type, not reboot_mode. Fix it so there are no more warnings and the code works properly. Fixes: eab8da48579d ("reboot: allow to specify reboot mode via sysfs") Link: https://github.com/ClangBuiltLinux/linux/issues/1197 Signed-off-by: Nathan Chancellor <natechancellor(a)gmail.com> Reviewed-and-tested-by: Matteo Croce <mcroce(a)microsoft.com> Reviewed-by: Petr Mladek <pmladek(a)suse.com> Signed-off-by: Hongyu Li <543306408(a)qq.com> --- kernel/reboot.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/kernel/reboot.c b/kernel/reboot.c index 778979d92dc7..7a445c710f9a 100644 --- a/kernel/reboot.c +++ b/kernel/reboot.c @@ -700,19 +700,19 @@ static ssize_t type_store(struct kobject *kobj, struct kobj_attribute *attr, return -EPERM; if (!strncmp(buf, BOOT_TRIPLE_STR, strlen(BOOT_TRIPLE_STR))) - reboot_mode = BOOT_TRIPLE; + reboot_type = BOOT_TRIPLE; else if (!strncmp(buf, BOOT_KBD_STR, strlen(BOOT_KBD_STR))) - reboot_mode = BOOT_KBD; + reboot_type = BOOT_KBD; else if (!strncmp(buf, BOOT_BIOS_STR, strlen(BOOT_BIOS_STR))) - reboot_mode = BOOT_BIOS; + reboot_type = BOOT_BIOS; else if (!strncmp(buf, BOOT_ACPI_STR, strlen(BOOT_ACPI_STR))) - reboot_mode = BOOT_ACPI; + reboot_type = BOOT_ACPI; else if (!strncmp(buf, BOOT_EFI_STR, strlen(BOOT_EFI_STR))) - reboot_mode = BOOT_EFI; + reboot_type = BOOT_EFI; else if (!strncmp(buf, BOOT_CF9_FORCE_STR, strlen(BOOT_CF9_FORCE_STR))) - reboot_mode = BOOT_CF9_FORCE; + reboot_type = BOOT_CF9_FORCE; else if (!strncmp(buf, BOOT_CF9_SAFE_STR, strlen(BOOT_CF9_SAFE_STR))) - reboot_mode = BOOT_CF9_SAFE; + reboot_type = BOOT_CF9_SAFE; else return -EINVAL; -- 2.17.1

2 1

[PATCH openEuler-1.0-LTS v2] scsi:spraid: support Ramaxel's spraid driver
by Yanling Song 22 Dec '21

22 Dec '21

Ramaxel inclusion category: features bugzilla: https://gitee.com/openeuler/kernel/issues/I4NJIB CVE: NA v1->v2: 1. Add deconfig for arm64 and x86 Support Ramaxel's SPRxxx Raid controller Signed-off-by: Yanling Song <songyl(a)ramaxel.com> Reviewed-by: Jiang Yu<yujiang(a)ramaxel.com> --- arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + drivers/scsi/Kconfig | 1 + drivers/scsi/Makefile | 1 + drivers/scsi/spraid/Kconfig | 13 + drivers/scsi/spraid/Makefile | 7 + drivers/scsi/spraid/spraid.h | 746 +++++ drivers/scsi/spraid/spraid_main.c | 3875 ++++++++++++++++++++++++ 8 files changed, 4645 insertions(+) create mode 100644 drivers/scsi/spraid/Kconfig create mode 100644 drivers/scsi/spraid/Makefile create mode 100644 drivers/scsi/spraid/spraid.h create mode 100644 drivers/scsi/spraid/spraid_main.c diff --git a/arch/arm64/configs/openeuler_defconfig b/arch/arm64/configs/openeuler_defconfig index d2167c80757f..47b0931b91bf 100644 --- a/arch/arm64/configs/openeuler_defconfig +++ b/arch/arm64/configs/openeuler_defconfig @@ -2143,6 +2143,7 @@ CONFIG_SCSI_MPT2SAS_MAX_SGE=128 CONFIG_SCSI_MPT3SAS_MAX_SGE=128 CONFIG_SCSI_MPT2SAS=m CONFIG_SCSI_SMARTPQI=m +CONFIG_RAMAXEL_SPRAID=m # CONFIG_SCSI_UFSHCD is not set # CONFIG_SCSI_HPTIOP is not set CONFIG_LIBFC=m diff --git a/arch/x86/configs/openeuler_defconfig b/arch/x86/configs/openeuler_defconfig index 7043426d9780..786186288fea 100644 --- a/arch/x86/configs/openeuler_defconfig +++ b/arch/x86/configs/openeuler_defconfig @@ -2182,6 +2182,7 @@ CONFIG_SCSI_MPT2SAS_MAX_SGE=128 CONFIG_SCSI_MPT3SAS_MAX_SGE=128 CONFIG_SCSI_MPT2SAS=m CONFIG_SCSI_SMARTPQI=m +CONFIG_RAMAXEL_SPRAID=m # CONFIG_SCSI_UFSHCD is not set # CONFIG_SCSI_HPTIOP is not set # CONFIG_SCSI_BUSLOGIC is not set diff --git a/drivers/scsi/Kconfig b/drivers/scsi/Kconfig index 8450484184e3..63d2aaa22834 100644 --- a/drivers/scsi/Kconfig +++ b/drivers/scsi/Kconfig @@ -522,6 +522,7 @@ source "drivers/scsi/megaraid/Kconfig.megaraid" source "drivers/scsi/mpt3sas/Kconfig" source "drivers/scsi/smartpqi/Kconfig" source "drivers/scsi/ufs/Kconfig" +source "drivers/scsi/spraid/Kconfig" config SCSI_HPTIOP tristate "HighPoint RocketRAID 3xxx/4xxx Controller support" diff --git a/drivers/scsi/Makefile b/drivers/scsi/Makefile index 2973693f6dcc..4056cf26e09e 100644 --- a/drivers/scsi/Makefile +++ b/drivers/scsi/Makefile @@ -94,6 +94,7 @@ obj-$(CONFIG_SCSI_ZALON) += zalon7xx.o obj-$(CONFIG_SCSI_DC395x) += dc395x.o obj-$(CONFIG_SCSI_AM53C974) += esp_scsi.o am53c974.o obj-$(CONFIG_CXLFLASH) += cxlflash/ +obj-$(CONFIG_RAMAXEL_SPRAID) += spraid/ obj-$(CONFIG_MEGARAID_LEGACY) += megaraid.o obj-$(CONFIG_MEGARAID_NEWGEN) += megaraid/ obj-$(CONFIG_MEGARAID_SAS) += megaraid/ diff --git a/drivers/scsi/spraid/Kconfig b/drivers/scsi/spraid/Kconfig new file mode 100644 index 000000000000..bfbba3db8db0 --- /dev/null +++ b/drivers/scsi/spraid/Kconfig @@ -0,0 +1,13 @@ +# +# Ramaxel driver configuration +# + +config RAMAXEL_SPRAID + tristate "Ramaxel spraid Adapter" + depends on PCI && SCSI + select BLK_DEV_BSGLIB + depends on ARM64 || X86_64 + help + This driver supports Ramaxel SPRxxx serial + raid controller, which has PCIE Gen4 interface + with host and supports SAS/SATA Hdd/ssd. diff --git a/drivers/scsi/spraid/Makefile b/drivers/scsi/spraid/Makefile new file mode 100644 index 000000000000..aadc2ffd37eb --- /dev/null +++ b/drivers/scsi/spraid/Makefile @@ -0,0 +1,7 @@ +# +# Makefile for the Ramaxel device drivers. +# + +obj-$(CONFIG_RAMAXEL_SPRAID) += spraid.o + +spraid-objs := spraid_main.o \ No newline at end of file diff --git a/drivers/scsi/spraid/spraid.h b/drivers/scsi/spraid/spraid.h new file mode 100644 index 000000000000..c1e4980e18e5 --- /dev/null +++ b/drivers/scsi/spraid/spraid.h @@ -0,0 +1,746 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* Copyright(c) 2021 Ramaxel Memory Technology, Ltd */ + +#ifndef __SPRAID_H_ +#define __SPRAID_H_ + +#define SPRAID_CAP_MQES(cap) ((cap) & 0xffff) +#define SPRAID_CAP_STRIDE(cap) (((cap) >> 32) & 0xf) +#define SPRAID_CAP_MPSMIN(cap) (((cap) >> 48) & 0xf) +#define SPRAID_CAP_MPSMAX(cap) (((cap) >> 52) & 0xf) +#define SPRAID_CAP_TIMEOUT(cap) (((cap) >> 24) & 0xff) +#define SPRAID_CAP_DMAMASK(cap) (((cap) >> 37) & 0xff) + +#define SPRAID_DEFAULT_MAX_CHANNEL 4 +#define SPRAID_DEFAULT_MAX_ID 240 +#define SPRAID_DEFAULT_MAX_LUN_PER_HOST 8 +#define MAX_SECTORS 2048 + +#define IO_SQE_SIZE sizeof(struct spraid_ioq_command) +#define ADMIN_SQE_SIZE sizeof(struct spraid_admin_command) +#define SQE_SIZE(qid) (((qid) > 0) ? IO_SQE_SIZE : ADMIN_SQE_SIZE) +#define CQ_SIZE(depth) ((depth) * sizeof(struct spraid_completion)) +#define SQ_SIZE(qid, depth) ((depth) * SQE_SIZE(qid)) + +#define SENSE_SIZE(depth) ((depth) * SCSI_SENSE_BUFFERSIZE) + +#define SPRAID_AQ_DEPTH 128 +#define SPRAID_NR_AEN_COMMANDS 16 +#define SPRAID_AQ_BLK_MQ_DEPTH (SPRAID_AQ_DEPTH - SPRAID_NR_AEN_COMMANDS) +#define SPRAID_AQ_MQ_TAG_DEPTH (SPRAID_AQ_BLK_MQ_DEPTH - 1) + +#define SPRAID_ADMIN_QUEUE_NUM 1 +#define SPRAID_PTCMDS_PERQ 1 +#define SPRAID_IO_BLK_MQ_DEPTH (hdev->shost->can_queue) +#define SPRAID_NR_IOQ_PTCMDS (SPRAID_PTCMDS_PERQ * hdev->shost->nr_hw_queues) + +#define FUA_MASK 0x08 +#define SPRAID_MINORS BIT(MINORBITS) + +#define COMMAND_IS_WRITE(cmd) ((cmd)->common.opcode & 1) + +#define SPRAID_IO_IOSQES 7 +#define SPRAID_IO_IOCQES 4 +#define PRP_ENTRY_SIZE 8 + +#define SMALL_POOL_SIZE 256 +#define MAX_SMALL_POOL_NUM 16 +#define MAX_CMD_PER_DEV 64 +#define MAX_CDB_LEN 32 + +#define SPRAID_UP_TO_MULTY4(x) (((x) + 4) & (~0x03)) + +#define CQE_STATUS_SUCCESS (0x0) + +#define PCI_VENDOR_ID_RAMAXEL_LOGIC 0x1E81 + +#define SPRAID_SERVER_DEVICE_HBA_DID 0x2100 +#define SPRAID_SERVER_DEVICE_RAID_DID 0x2200 + +#define IO_6_DEFAULT_TX_LEN 256 + +#define SPRAID_INT_PAGES 2 +#define SPRAID_INT_BYTES(hdev) (SPRAID_INT_PAGES * (hdev)->page_size) + +enum { + SPRAID_REQ_CANCELLED = (1 << 0), + SPRAID_REQ_USERCMD = (1 << 1), +}; + +enum { + SPRAID_SC_SUCCESS = 0x0, + SPRAID_SC_INVALID_OPCODE = 0x1, + SPRAID_SC_INVALID_FIELD = 0x2, + + SPRAID_SC_ABORT_LIMIT = 0x103, + SPRAID_SC_ABORT_MISSING = 0x104, + SPRAID_SC_ASYNC_LIMIT = 0x105, + + SPRAID_SC_DNR = 0x4000, +}; + +enum { + SPRAID_REG_CAP = 0x0000, + SPRAID_REG_CC = 0x0014, + SPRAID_REG_CSTS = 0x001c, + SPRAID_REG_AQA = 0x0024, + SPRAID_REG_ASQ = 0x0028, + SPRAID_REG_ACQ = 0x0030, + SPRAID_REG_DBS = 0x1000, +}; + +enum { + SPRAID_CC_ENABLE = 1 << 0, + SPRAID_CC_CSS_NVM = 0 << 4, + SPRAID_CC_MPS_SHIFT = 7, + SPRAID_CC_AMS_SHIFT = 11, + SPRAID_CC_SHN_SHIFT = 14, + SPRAID_CC_IOSQES_SHIFT = 16, + SPRAID_CC_IOCQES_SHIFT = 20, + SPRAID_CC_AMS_RR = 0 << SPRAID_CC_AMS_SHIFT, + SPRAID_CC_SHN_NONE = 0 << SPRAID_CC_SHN_SHIFT, + SPRAID_CC_IOSQES = SPRAID_IO_IOSQES << SPRAID_CC_IOSQES_SHIFT, + SPRAID_CC_IOCQES = SPRAID_IO_IOCQES << SPRAID_CC_IOCQES_SHIFT, + SPRAID_CC_SHN_NORMAL = 1 << SPRAID_CC_SHN_SHIFT, + SPRAID_CC_SHN_MASK = 3 << SPRAID_CC_SHN_SHIFT, + SPRAID_CSTS_CFS_SHIFT = 1, + SPRAID_CSTS_SHST_SHIFT = 2, + SPRAID_CSTS_PP_SHIFT = 5, + SPRAID_CSTS_RDY = 1 << 0, + SPRAID_CSTS_SHST_CMPLT = 2 << 2, + SPRAID_CSTS_SHST_MASK = 3 << 2, + SPRAID_CSTS_CFS_MASK = 1 << SPRAID_CSTS_CFS_SHIFT, + SPRAID_CSTS_PP_MASK = 1 << SPRAID_CSTS_PP_SHIFT, +}; + +enum { + SPRAID_ADMIN_DELETE_SQ = 0x00, + SPRAID_ADMIN_CREATE_SQ = 0x01, + SPRAID_ADMIN_DELETE_CQ = 0x04, + SPRAID_ADMIN_CREATE_CQ = 0x05, + SPRAID_ADMIN_ABORT_CMD = 0x08, + SPRAID_ADMIN_SET_FEATURES = 0x09, + SPRAID_ADMIN_ASYNC_EVENT = 0x0c, + SPRAID_ADMIN_GET_INFO = 0xc6, + SPRAID_ADMIN_RESET = 0xc8, +}; + +enum { + SPRAID_GET_INFO_CTRL = 0, + SPRAID_GET_INFO_DEV_LIST = 1, +}; + +enum { + SPRAID_RESET_TARGET = 0, + SPRAID_RESET_BUS = 1, +}; + +enum { + SPRAID_AEN_ERROR = 0, + SPRAID_AEN_NOTICE = 2, + SPRAID_AEN_VS = 7, +}; + +enum { + SPRAID_AEN_DEV_CHANGED = 0x00, + SPRAID_AEN_FW_ACT_START = 0x01, + SPRAID_AEN_HOST_PROBING = 0x10, +}; + +enum { + SPRAID_AEN_TIMESYN = 0x00, + SPRAID_AEN_FW_ACT_FINISH = 0x02, + SPRAID_AEN_EVENT_MIN = 0x80, + SPRAID_AEN_EVENT_MAX = 0xff, +}; + +enum { + SPRAID_CMD_WRITE = 0x01, + SPRAID_CMD_READ = 0x02, + + SPRAID_CMD_NONIO_NONE = 0x80, + SPRAID_CMD_NONIO_TODEV = 0x81, + SPRAID_CMD_NONIO_FROMDEV = 0x82, +}; + +enum { + SPRAID_QUEUE_PHYS_CONTIG = (1 << 0), + SPRAID_CQ_IRQ_ENABLED = (1 << 1), + + SPRAID_FEAT_NUM_QUEUES = 0x07, + SPRAID_FEAT_ASYNC_EVENT = 0x0b, + SPRAID_FEAT_TIMESTAMP = 0x0e, +}; + +enum spraid_state { + SPRAID_NEW, + SPRAID_LIVE, + SPRAID_RESETTING, + SPRAID_DELETING, + SPRAID_DEAD, +}; + +enum { + SPRAID_CARD_HBA, + SPRAID_CARD_RAID, +}; + +enum spraid_cmd_type { + SPRAID_CMD_ADM, + SPRAID_CMD_IOPT, +}; + +struct spraid_completion { + __le32 result; + union { + struct { + __u8 sense_len; + __u8 resv[3]; + }; + __le32 result1; + }; + __le16 sq_head; + __le16 sq_id; + __u16 cmd_id; + __le16 status; +}; + +struct spraid_ctrl_info { + __le32 nd; + __le16 max_cmds; + __le16 max_channel; + __le32 max_tgt_id; + __le16 max_lun; + __le16 max_num_sge; + __le16 lun_num_in_boot; + __u8 mdts; + __u8 acl; + __u8 aerl; + __u8 card_type; + __u16 rsvd; + __u32 rtd3e; + __u8 sn[32]; + __u8 fr[16]; + __u8 rsvd1[4020]; +}; + +struct spraid_dev { + struct pci_dev *pdev; + struct device *dev; + struct Scsi_Host *shost; + struct spraid_queue *queues; + struct dma_pool *prp_page_pool; + struct dma_pool *prp_small_pool[MAX_SMALL_POOL_NUM]; + mempool_t *iod_mempool; + void __iomem *bar; + u32 max_qid; + u32 num_vecs; + u32 queue_count; + u32 ioq_depth; + int db_stride; + u32 __iomem *dbs; + struct rw_semaphore devices_rwsem; + int numa_node; + u32 page_size; + u32 ctrl_config; + u32 online_queues; + u64 cap; + int instance; + struct spraid_ctrl_info *ctrl_info; + struct spraid_dev_info *devices; + + struct spraid_cmd *adm_cmds; + struct list_head adm_cmd_list; + spinlock_t adm_cmd_lock; + + struct spraid_cmd *ioq_ptcmds; + struct list_head ioq_pt_list; + spinlock_t ioq_pt_lock; + + struct work_struct scan_work; + struct work_struct timesyn_work; + struct work_struct reset_work; + struct work_struct fw_act_work; + + enum spraid_state state; + spinlock_t state_lock; + + struct request_queue *bsg_queue; +}; + +struct spraid_sgl_desc { + __le64 addr; + __le32 length; + __u8 rsvd[3]; + __u8 type; +}; + +union spraid_data_ptr { + struct { + __le64 prp1; + __le64 prp2; + }; + struct spraid_sgl_desc sgl; +}; + +struct spraid_admin_common_command { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __le32 cdw2[4]; + union spraid_data_ptr dptr; + __le32 cdw10; + __le32 cdw11; + __le32 cdw12; + __le32 cdw13; + __le32 cdw14; + __le32 cdw15; +}; + +struct spraid_features { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __u64 rsvd2[2]; + union spraid_data_ptr dptr; + __le32 fid; + __le32 dword11; + __le32 dword12; + __le32 dword13; + __le32 dword14; + __le32 dword15; +}; + +struct spraid_create_cq { + __u8 opcode; + __u8 flags; + __u16 command_id; + __u32 rsvd1[5]; + __le64 prp1; + __u64 rsvd8; + __le16 cqid; + __le16 qsize; + __le16 cq_flags; + __le16 irq_vector; + __u32 rsvd12[4]; +}; + +struct spraid_create_sq { + __u8 opcode; + __u8 flags; + __u16 command_id; + __u32 rsvd1[5]; + __le64 prp1; + __u64 rsvd8; + __le16 sqid; + __le16 qsize; + __le16 sq_flags; + __le16 cqid; + __u32 rsvd12[4]; +}; + +struct spraid_delete_queue { + __u8 opcode; + __u8 flags; + __u16 command_id; + __u32 rsvd1[9]; + __le16 qid; + __u16 rsvd10; + __u32 rsvd11[5]; +}; + +struct spraid_get_info { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __u32 rsvd2[4]; + union spraid_data_ptr dptr; + __u8 type; + __u8 rsvd10[3]; + __le32 cdw11; + __u32 rsvd12[4]; +}; + +struct spraid_usr_cmd { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + union { + struct { + __le16 subopcode; + __le16 rsvd1; + } info_0; + __le32 cdw2; + }; + union { + struct { + __le16 data_len; + __le16 param_len; + } info_1; + __le32 cdw3; + }; + __u64 metadata; + union spraid_data_ptr dptr; + __le32 cdw10; + __le32 cdw11; + __le32 cdw12; + __le32 cdw13; + __le32 cdw14; + __le32 cdw15; +}; + +enum { + SPRAID_CMD_FLAG_SGL_METABUF = (1 << 6), + SPRAID_CMD_FLAG_SGL_METASEG = (1 << 7), + SPRAID_CMD_FLAG_SGL_ALL = SPRAID_CMD_FLAG_SGL_METABUF | + SPRAID_CMD_FLAG_SGL_METASEG, +}; + +enum spraid_cmd_state { + SPRAID_CMD_IDLE = 0, + SPRAID_CMD_IN_FLIGHT = 1, + SPRAID_CMD_COMPLETE = 2, + SPRAID_CMD_TIMEOUT = 3, + SPRAID_CMD_TMO_COMPLETE = 4, +}; + +struct spraid_abort_cmd { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __u64 rsvd2[4]; + __le16 sqid; + __le16 cid; + __u32 rsvd11[5]; +}; + +struct spraid_reset_cmd { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __u64 rsvd2[4]; + __u8 type; + __u8 rsvd10[3]; + __u32 rsvd11[5]; +}; + +struct spraid_admin_command { + union { + struct spraid_admin_common_command common; + struct spraid_features features; + struct spraid_create_cq create_cq; + struct spraid_create_sq create_sq; + struct spraid_delete_queue delete_queue; + struct spraid_get_info get_info; + struct spraid_abort_cmd abort; + struct spraid_reset_cmd reset; + struct spraid_usr_cmd usr_cmd; + }; +}; + +struct spraid_ioq_common_command { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __le16 sense_len; + __u8 cdb_len; + __u8 rsvd2; + __le32 cdw3[3]; + union spraid_data_ptr dptr; + __le32 cdw10[6]; + __u8 cdb[32]; + __le64 sense_addr; + __le32 cdw26[6]; +}; + +struct spraid_rw_command { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __le16 sense_len; + __u8 cdb_len; + __u8 rsvd2; + __u32 rsvd3[3]; + union spraid_data_ptr dptr; + __le64 slba; + __le16 nlb; + __le16 control; + __u32 rsvd13[3]; + __u8 cdb[32]; + __le64 sense_addr; + __u32 rsvd26[6]; +}; + +struct spraid_scsi_nonio { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __le16 sense_len; + __u8 cdb_length; + __u8 rsvd2; + __u32 rsvd3[3]; + union spraid_data_ptr dptr; + __u32 rsvd10[5]; + __le32 buffer_len; + __u8 cdb[32]; + __le64 sense_addr; + __u32 rsvd26[6]; +}; + +struct spraid_ioq_command { + union { + struct spraid_ioq_common_command common; + struct spraid_rw_command rw; + struct spraid_scsi_nonio scsi_nonio; + }; +}; + +struct spraid_passthru_common_cmd { + __u8 opcode; + __u8 flags; + __u16 rsvd0; + __u32 nsid; + union { + struct { + __u16 subopcode; + __u16 rsvd1; + } info_0; + __u32 cdw2; + }; + union { + struct { + __u16 data_len; + __u16 param_len; + } info_1; + __u32 cdw3; + }; + __u64 metadata; + + __u64 addr; + __u64 prp2; + + __u32 cdw10; + __u32 cdw11; + __u32 cdw12; + __u32 cdw13; + __u32 cdw14; + __u32 cdw15; + __u32 timeout_ms; + __u32 result0; + __u32 result1; +}; + +struct spraid_ioq_passthru_cmd { + __u8 opcode; + __u8 flags; + __u16 rsvd0; + __u32 nsid; + union { + struct { + __u16 res_sense_len; + __u8 cdb_len; + __u8 rsvd0; + } info_0; + __u32 cdw2; + }; + union { + struct { + __u16 subopcode; + __u16 rsvd1; + } info_1; + __u32 cdw3; + }; + union { + struct { + __u16 rsvd; + __u16 param_len; + } info_2; + __u32 cdw4; + }; + __u32 cdw5; + __u64 addr; + __u64 prp2; + union { + struct { + __u16 eid; + __u16 sid; + } info_3; + __u32 cdw10; + }; + union { + struct { + __u16 did; + __u8 did_flag; + __u8 rsvd2; + } info_4; + __u32 cdw11; + }; + __u32 cdw12; + __u32 cdw13; + __u32 cdw14; + __u32 data_len; + __u32 cdw16; + __u32 cdw17; + __u32 cdw18; + __u32 cdw19; + __u32 cdw20; + __u32 cdw21; + __u32 cdw22; + __u32 cdw23; + __u64 sense_addr; + __u32 cdw26[4]; + __u32 timeout_ms; + __u32 result0; + __u32 result1; +}; + +struct spraid_bsg_request { + u32 msgcode; + u32 control; + union { + struct spraid_passthru_common_cmd admcmd; + struct spraid_ioq_passthru_cmd ioqcmd; + }; +}; + +enum { + SPRAID_BSG_ADM, + SPRAID_BSG_IOQ, +}; + +struct spraid_cmd { + int qid; + int cid; + u32 result0; + u32 result1; + u16 status; + void *priv; + enum spraid_cmd_state state; + struct completion cmd_done; + struct list_head list; +}; + +struct spraid_queue { + struct spraid_dev *hdev; + spinlock_t sq_lock; + + spinlock_t cq_lock ____cacheline_aligned_in_smp; + + void *sq_cmds; + + struct spraid_completion *cqes; + + dma_addr_t sq_dma_addr; + dma_addr_t cq_dma_addr; + u32 __iomem *q_db; + u8 cq_phase; + u8 sqes; + u16 qid; + u16 sq_tail; + u16 cq_head; + u16 last_cq_head; + u16 q_depth; + s16 cq_vector; + void *sense; + dma_addr_t sense_dma_addr; + struct dma_pool *prp_small_pool; +}; + +struct spraid_iod { + struct spraid_queue *spraidq; + enum spraid_cmd_state state; + int npages; + u32 nsge; + u32 length; + bool use_sgl; + bool sg_drv_mgmt; + dma_addr_t first_dma; + void *sense; + dma_addr_t sense_dma; + struct scatterlist *sg; + struct scatterlist inline_sg[0]; +}; + +#define SPRAID_DEV_INFO_ATTR_BOOT(attr) ((attr) & 0x01) +#define SPRAID_DEV_INFO_ATTR_VD(attr) (((attr) & 0x02) == 0x0) +#define SPRAID_DEV_INFO_ATTR_PT(attr) (((attr) & 0x22) == 0x02) +#define SPRAID_DEV_INFO_ATTR_RAWDISK(attr) ((attr) & 0x20) + +#define SPRAID_DEV_INFO_FLAG_VALID(flag) ((flag) & 0x01) +#define SPRAID_DEV_INFO_FLAG_CHANGE(flag) ((flag) & 0x02) + +#define BGTASK_TYPE_REBUILD 4 +#define USR_CMD_READ 0xc2 +#define USR_CMD_RDLEN 0x1000 +#define USR_CMD_VDINFO 0x704 +#define USR_CMD_BGTASK 0x504 +#define VDINFO_PARAM_LEN 0x04 + +struct spraid_vd_info { + __u8 name[32]; + __le16 id; + __u8 rg_id; + __u8 rg_level; + __u8 sg_num; + __u8 sg_disk_num; + __u8 vd_status; + __u8 vd_type; + __u8 rsvd1[4056]; +}; + +#define MAX_REALTIME_BGTASK_NUM 32 + +struct bgtask_info { + __u8 type; + __u8 progress; + __u8 rate; + __u8 rsvd0; + __le16 vd_id; + __le16 time_left; + __u8 rsvd1[4]; +}; + +struct spraid_bgtask { + __u8 sw; + __u8 task_num; + __u8 rsvd[6]; + struct bgtask_info bgtask[MAX_REALTIME_BGTASK_NUM]; +}; + +struct spraid_dev_info { + __le32 hdid; + __le16 target; + __u8 channel; + __u8 lun; + __u8 attr; + __u8 flag; + __le16 max_io_kb; +}; + +#define MAX_DEV_ENTRY_PER_PAGE_4K 340 +struct spraid_dev_list { + __le32 dev_num; + __u32 rsvd0[3]; + struct spraid_dev_info devices[MAX_DEV_ENTRY_PER_PAGE_4K]; +}; + +struct spraid_sdev_hostdata { + u32 hdid; + u16 max_io_kb; + u8 attr; + u8 flag; + u8 rg_id; + u8 rsvd[3]; +}; + +#endif + diff --git a/drivers/scsi/spraid/spraid_main.c b/drivers/scsi/spraid/spraid_main.c new file mode 100644 index 000000000000..519b39f44e91 --- /dev/null +++ b/drivers/scsi/spraid/spraid_main.c @@ -0,0 +1,3875 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright(c) 2021 Ramaxel Memory Technology, Ltd */ + +/* Ramaxel Raid SPXXX Series Linux Driver */ + +#define pr_fmt(fmt) "spraid: " fmt + +#include <linux/sched/signal.h> +#include <linux/version.h> +#include <linux/pci.h> +#include <linux/aer.h> +#include <linux/module.h> +#include <linux/ioport.h> +#include <linux/device.h> +#include <linux/delay.h> +#include <linux/interrupt.h> +#include <linux/cdev.h> +#include <linux/sysfs.h> +#include <linux/gfp.h> +#include <linux/types.h> +#include <linux/ratelimit.h> +#include <linux/once.h> +#include <linux/debugfs.h> +#include <linux/io-64-nonatomic-lo-hi.h> +#include <linux/blkdev.h> +#include <linux/bsg-lib.h> +#include <asm/unaligned.h> +#include <linux/sort.h> +#include <target/target_core_backend.h> + +#include <scsi/scsi.h> +#include <scsi/scsi_cmnd.h> +#include <scsi/scsi_device.h> +#include <scsi/scsi_host.h> +#include <scsi/scsi_transport.h> +#include <scsi/scsi_dbg.h> + + +#include "spraid.h" + +static u32 admin_tmout = 60; +module_param(admin_tmout, uint, 0644); +MODULE_PARM_DESC(admin_tmout, "admin commands timeout (seconds)"); + +static u32 scmd_tmout_nonpt = 180; +module_param(scmd_tmout_nonpt, uint, 0644); +MODULE_PARM_DESC(scmd_tmout_nonpt, + "scsi commands timeout for rawdisk&raid(seconds)"); + +static int ioq_depth_set(const char *val, const struct kernel_param *kp); +static const struct kernel_param_ops ioq_depth_ops = { + .set = ioq_depth_set, + .get = param_get_uint, +}; + +static u32 io_queue_depth = 1024; +module_param_cb(io_queue_depth, &ioq_depth_ops, &io_queue_depth, 0644); +MODULE_PARM_DESC(io_queue_depth, "set io queue depth, should >= 2"); + +static int log_debug_switch_set(const char *val, const struct kernel_param *kp) +{ + u8 n = 0; + int ret; + + ret = kstrtou8(val, 10, &n); + if (ret != 0) + return -EINVAL; + + return param_set_byte(val, kp); +} + +static const struct kernel_param_ops log_debug_switch_ops = { + .set = log_debug_switch_set, + .get = param_get_byte, +}; + +static unsigned char log_debug_switch; +module_param_cb(log_debug_switch, &log_debug_switch_ops, + &log_debug_switch, 0644); +MODULE_PARM_DESC(log_debug_switch, + "set log state, default non-zero for switch on"); + +static int small_pool_num_set(const char *val, const struct kernel_param *kp) +{ + u8 n = 0; + int ret; + + ret = kstrtou8(val, 10, &n); + if (ret != 0) + return -EINVAL; + if (n > MAX_SMALL_POOL_NUM) + n = MAX_SMALL_POOL_NUM; + if (n < 1) + n = 1; + *((u8 *)kp->arg) = n; + + return 0; +} + +static const struct kernel_param_ops small_pool_num_ops = { + .set = small_pool_num_set, + .get = param_get_byte, +}; + +/* It was found that the spindlock of a single pool conflicts + * a lot with multiple CPUs.So multiple pools are introduced + * to reduce the conflictions. + */ +static unsigned char small_pool_num = 4; +module_param_cb(small_pool_num, &small_pool_num_ops, &small_pool_num, 0644); +MODULE_PARM_DESC(small_pool_num, "set prp small pool num, default 4, MAX 16"); + +static void spraid_free_queue(struct spraid_queue *spraidq); +static void spraid_handle_aen_notice(struct spraid_dev *hdev, u32 result); +static void spraid_handle_aen_vs(struct spraid_dev *hdev, + u32 result, u32 result1); + +static DEFINE_IDA(spraid_instance_ida); + +static struct class *spraid_class; + +#define SPRAID_CAP_TIMEOUT_UNIT_MS (HZ / 2) + +static struct workqueue_struct *spraid_wq; + +#define dev_log_dbg(dev, fmt, ...) do { \ + if (unlikely(log_debug_switch)) \ + dev_info(dev, "[%s] [%d] " fmt, \ + __func__, __LINE__, ##__VA_ARGS__); \ +} while (0) + +#define SPRAID_DRV_VERSION "1.0.0.0" + +#define ADMIN_TIMEOUT (admin_tmout * HZ) +#define ADMIN_ERR_TIMEOUT 32757 + +#define SPRAID_WAIT_ABNL_CMD_TIMEOUT (3 * 2) + +#define SPRAID_DMA_MSK_BIT_MAX 64 + +enum FW_STAT_CODE { + FW_STAT_OK = 0, + FW_STAT_NEED_CHECK, + FW_STAT_ERROR, + FW_STAT_EP_PCIE_ERROR, + FW_STAT_NAC_DMA_ERROR, + FW_STAT_ABORTED, + FW_STAT_NEED_RETRY +}; + +static const char * const raid_levels[] = {"0", "1", "5", "6", "10", "50", "60", + "NA"}; + +static const char * const raid_states[] = { + "NA", "NORMAL", "FAULT", "DEGRADE", "NOT_FORMATTED", + "FORMATTING", "SANITIZING", "INITIALIZING", "INITIALIZE_FAIL", + "DELETING", "DELETE_FAIL", "WRITE_PROTECT" +}; + +static int ioq_depth_set(const char *val, const struct kernel_param *kp) +{ + int n = 0; + int ret; + + ret = kstrtoint(val, 10, &n); + if (ret != 0 || n < 2) + return -EINVAL; + + return param_set_int(val, kp); +} + +static int spraid_remap_bar(struct spraid_dev *hdev, u32 size) +{ + struct pci_dev *pdev = hdev->pdev; + + if (size > pci_resource_len(pdev, 0)) { + dev_err(hdev->dev, "Input size[%u] exceed bar0 length[%llu]\n", + size, pci_resource_len(pdev, 0)); + return -ENOMEM; + } + + if (hdev->bar) + iounmap(hdev->bar); + + hdev->bar = ioremap(pci_resource_start(pdev, 0), size); + if (!hdev->bar) { + dev_err(hdev->dev, "ioremap for bar0 failed\n"); + return -ENOMEM; + } + hdev->dbs = hdev->bar + SPRAID_REG_DBS; + + return 0; +} + +static int spraid_dev_map(struct spraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + int ret; + + ret = pci_request_mem_regions(pdev, "spraid"); + if (ret) { + dev_err(hdev->dev, "fail to request memory regions\n"); + return ret; + } + + ret = spraid_remap_bar(hdev, SPRAID_REG_DBS + 4096); + if (ret) { + pci_release_mem_regions(pdev); + return ret; + } + + return 0; +} + +static void spraid_dev_unmap(struct spraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + + if (hdev->bar) { + iounmap(hdev->bar); + hdev->bar = NULL; + } + pci_release_mem_regions(pdev); +} + +static int spraid_pci_enable(struct spraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + int ret = -ENOMEM; + u64 maskbit = SPRAID_DMA_MSK_BIT_MAX; + + if (pci_enable_device_mem(pdev)) { + dev_err(hdev->dev, + "Enable pci device memory resources failed\n"); + return ret; + } + pci_set_master(pdev); + + if (readl(hdev->bar + SPRAID_REG_CSTS) == U32_MAX) { + ret = -ENODEV; + dev_err(hdev->dev, "Read csts register failed\n"); + goto disable; + } + + ret = pci_alloc_irq_vectors(pdev, 1, 1, PCI_IRQ_ALL_TYPES); + if (ret < 0) { + dev_err(hdev->dev, + "Allocate one IRQ for setup admin channel failed\n"); + goto disable; + } + + hdev->cap = lo_hi_readq(hdev->bar + SPRAID_REG_CAP); + hdev->ioq_depth = min_t(u32, SPRAID_CAP_MQES(hdev->cap) + 1, + io_queue_depth); + hdev->db_stride = 1 << SPRAID_CAP_STRIDE(hdev->cap); + + maskbit = SPRAID_CAP_DMAMASK(hdev->cap); + if (maskbit < 32 || maskbit > SPRAID_DMA_MSK_BIT_MAX) { + dev_err(hdev->dev, + "err, dma mask invalid[%llu], set to default\n", + maskbit); + maskbit = SPRAID_DMA_MSK_BIT_MAX; + } + if (dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(maskbit))) { + dev_err(hdev->dev, "set dma mask and coherent failed\n"); + goto disable; + } + + dev_info(hdev->dev, "set dma mask[%llu] success\n", maskbit); + + pci_enable_pcie_error_reporting(pdev); + pci_save_state(pdev); + + return 0; + +disable: + pci_disable_device(pdev); + return ret; +} + +static int spraid_npages_prp(u32 size, struct spraid_dev *hdev) +{ + u32 nprps = DIV_ROUND_UP(size + hdev->page_size, hdev->page_size); + + return DIV_ROUND_UP(PRP_ENTRY_SIZE * nprps, PAGE_SIZE - PRP_ENTRY_SIZE); +} + +static int spraid_npages_sgl(u32 nseg) +{ + return DIV_ROUND_UP(nseg * sizeof(struct spraid_sgl_desc), PAGE_SIZE); +} + +static void **spraid_iod_list(struct spraid_iod *iod) +{ + return (void **)(iod->inline_sg + (iod->sg_drv_mgmt ? iod->nsge : 0)); +} + +static u32 spraid_iod_ext_size(struct spraid_dev *hdev, u32 size, u32 nsge, + bool sg_drv_mgmt, bool use_sgl) +{ + size_t alloc_size, sg_size; + + if (use_sgl) + alloc_size = sizeof(__le64 *) * spraid_npages_sgl(nsge); + else + alloc_size = sizeof(__le64 *) * spraid_npages_prp(size, hdev); + + sg_size = sg_drv_mgmt ? (sizeof(struct scatterlist) * nsge) : 0; + return sg_size + alloc_size; +} + +static u32 spraid_cmd_size(struct spraid_dev *hdev, + bool sg_drv_mgmt, bool use_sgl) +{ + u32 alloc_size = spraid_iod_ext_size(hdev, SPRAID_INT_BYTES(hdev), + SPRAID_INT_PAGES, sg_drv_mgmt, use_sgl); + + dev_info(hdev->dev, "sg_drv_mgmt: %s, use_sgl: %s, iod size: %lu;" + " alloc_size: %u\n", sg_drv_mgmt ? "true" : "false", + use_sgl ? "true" : "false", + sizeof(struct spraid_iod), alloc_size); + + return sizeof(struct spraid_iod) + alloc_size; +} + +static int spraid_setup_prps(struct spraid_dev *hdev, struct spraid_iod *iod) +{ + struct scatterlist *sg = iod->sg; + u64 dma_addr = sg_dma_address(sg); + int dma_len = sg_dma_len(sg); + __le64 *prp_list, *old_prp_list; + u32 page_size = hdev->page_size; + int offset = dma_addr & (page_size - 1); + void **list = spraid_iod_list(iod); + int length = iod->length; + struct dma_pool *pool; + dma_addr_t prp_dma; + int nprps, i; + + length -= (page_size - offset); + if (length <= 0) { + iod->first_dma = 0; + return 0; + } + + dma_len -= (page_size - offset); + if (dma_len) { + dma_addr += (page_size - offset); + } else { + sg = sg_next(sg); + dma_addr = sg_dma_address(sg); + dma_len = sg_dma_len(sg); + } + + if (length <= page_size) { + iod->first_dma = dma_addr; + return 0; + } + + nprps = DIV_ROUND_UP(length, page_size); + if (nprps <= (SMALL_POOL_SIZE / PRP_ENTRY_SIZE)) { + pool = iod->spraidq->prp_small_pool; + iod->npages = 0; + } else { + pool = hdev->prp_page_pool; + iod->npages = 1; + } + + prp_list = dma_pool_alloc(pool, GFP_ATOMIC, &prp_dma); + if (!prp_list) { + dev_err_ratelimited(hdev->dev, + "Allocate first prp_list memory failed\n"); + iod->first_dma = dma_addr; + iod->npages = -1; + return -ENOMEM; + } + list[0] = prp_list; + iod->first_dma = prp_dma; + i = 0; + for (;;) { + if (i == page_size / PRP_ENTRY_SIZE) { + old_prp_list = prp_list; + + prp_list = dma_pool_alloc(pool, GFP_ATOMIC, &prp_dma); + if (!prp_list) { + dev_err_ratelimited(hdev->dev, "Allocate %dth;" + " prp_list memory failed\n", + iod->npages + 1); + return -ENOMEM; + } + list[iod->npages++] = prp_list; + prp_list[0] = old_prp_list[i - 1]; + old_prp_list[i - 1] = cpu_to_le64(prp_dma); + i = 1; + } + prp_list[i++] = cpu_to_le64(dma_addr); + dma_len -= page_size; + dma_addr += page_size; + length -= page_size; + if (length <= 0) + break; + if (dma_len > 0) + continue; + if (unlikely(dma_len < 0)) + goto bad_sgl; + sg = sg_next(sg); + dma_addr = sg_dma_address(sg); + dma_len = sg_dma_len(sg); + } + + return 0; + +bad_sgl: + dev_err(hdev->dev, + "Setup prps, invalid SGL for payload: %d nents: %d\n", + iod->length, iod->nsge); + return -EIO; +} + +#define SGES_PER_PAGE (PAGE_SIZE / sizeof(struct spraid_sgl_desc)) + +static void spraid_submit_cmd(struct spraid_queue *spraidq, const void *cmd) +{ + u32 sqes = SQE_SIZE(spraidq->qid); + unsigned long flags; + struct spraid_admin_common_command *acd = + (struct spraid_admin_common_command *)cmd; + + spin_lock_irqsave(&spraidq->sq_lock, flags); + memcpy((spraidq->sq_cmds + sqes * spraidq->sq_tail), cmd, sqes); + if (++spraidq->sq_tail == spraidq->q_depth) + spraidq->sq_tail = 0; + + writel(spraidq->sq_tail, spraidq->q_db); + spin_unlock_irqrestore(&spraidq->sq_lock, flags); + + dev_log_dbg(spraidq->hdev->dev, + "cid[%d] qid[%d], opcode[0x%x], flags[0x%x], hdid[%u]\n", + acd->command_id, spraidq->qid, acd->opcode, + acd->flags, le32_to_cpu(acd->hdid)); +} + +static u32 spraid_mod64(u64 dividend, u32 divisor) +{ + u64 d; + u32 remainder; + + if (!divisor) + pr_err("DIVISOR is zero, in div fn\n"); + + d = dividend; + remainder = do_div(d, divisor); + return remainder; +} + +static inline bool spraid_is_rw_scmd(struct scsi_cmnd *scmd) +{ + switch (scmd->cmnd[0]) { + case READ_6: + case READ_10: + case READ_12: + case READ_16: + case READ_32: + case WRITE_6: + case WRITE_10: + case WRITE_12: + case WRITE_16: + case WRITE_32: + return true; + default: + return false; + } +} + +static bool spraid_is_prp(struct spraid_dev *hdev, + struct scsi_cmnd *scmd, u32 nsge) +{ + struct scatterlist *sg = scsi_sglist(scmd); + u32 page_size = hdev->page_size; + bool is_prp = true; + int i = 0; + + scsi_for_each_sg(scmd, sg, nsge, i) { + if (i != 0 && i != nsge - 1) { + if (spraid_mod64(sg_dma_len(sg), page_size) || + spraid_mod64(sg_dma_address(sg), page_size)) { + is_prp = false; + break; + } + } + + if (nsge > 1 && i == 0) { + if ((spraid_mod64((sg_dma_address(sg) + sg_dma_len(sg)), + page_size))) { + is_prp = false; + break; + } + } + + if (nsge > 1 && i == (nsge - 1)) { + if (spraid_mod64(sg_dma_address(sg), page_size)) { + is_prp = false; + break; + } + } + } + + return is_prp; +} + +enum { + SPRAID_SGL_FMT_DATA_DESC = 0x00, + SPRAID_SGL_FMT_SEG_DESC = 0x02, + SPRAID_SGL_FMT_LAST_SEG_DESC = 0x03, + SPRAID_KEY_SGL_FMT_DATA_DESC = 0x04, + SPRAID_TRANSPORT_SGL_DATA_DESC = 0x05 +}; + +static void spraid_sgl_set_data(struct spraid_sgl_desc *sge, + struct scatterlist *sg) +{ + sge->addr = cpu_to_le64(sg_dma_address(sg)); + sge->length = cpu_to_le32(sg_dma_len(sg)); + sge->type = SPRAID_SGL_FMT_DATA_DESC << 4; +} + +static void spraid_sgl_set_seg(struct spraid_sgl_desc *sge, + dma_addr_t dma_addr, int entries) +{ + sge->addr = cpu_to_le64(dma_addr); + if (entries <= SGES_PER_PAGE) { + sge->length = cpu_to_le32(entries * sizeof(*sge)); + sge->type = SPRAID_SGL_FMT_LAST_SEG_DESC << 4; + } else { + sge->length = cpu_to_le32(PAGE_SIZE); + sge->type = SPRAID_SGL_FMT_SEG_DESC << 4; + } +} + +static int spraid_setup_ioq_cmd_sgl(struct spraid_dev *hdev, + struct scsi_cmnd *scmd, + struct spraid_ioq_command *ioq_cmd, + struct spraid_iod *iod) +{ + struct spraid_sgl_desc *sg_list, *link, *old_sg_list; + struct scatterlist *sg = scsi_sglist(scmd); + void **list = spraid_iod_list(iod); + struct dma_pool *pool; + int nsge = iod->nsge; + dma_addr_t sgl_dma; + int i = 0; + + ioq_cmd->common.flags |= SPRAID_CMD_FLAG_SGL_METABUF; + + if (nsge == 1) { + spraid_sgl_set_data(&ioq_cmd->common.dptr.sgl, sg); + return 0; + } + + if (nsge <= (SMALL_POOL_SIZE / sizeof(struct spraid_sgl_desc))) { + pool = iod->spraidq->prp_small_pool; + iod->npages = 0; + } else { + pool = hdev->prp_page_pool; + iod->npages = 1; + } + + sg_list = dma_pool_alloc(pool, GFP_ATOMIC, &sgl_dma); + if (!sg_list) { + dev_err_ratelimited(hdev->dev, + "Allocate first sgl_list failed\n"); + iod->npages = -1; + return -ENOMEM; + } + + list[0] = sg_list; + iod->first_dma = sgl_dma; + spraid_sgl_set_seg(&ioq_cmd->common.dptr.sgl, sgl_dma, nsge); + do { + if (i == SGES_PER_PAGE) { + old_sg_list = sg_list; + link = &old_sg_list[SGES_PER_PAGE - 1]; + + sg_list = dma_pool_alloc(pool, GFP_ATOMIC, &sgl_dma); + if (!sg_list) { + dev_err_ratelimited(hdev->dev, + "Allocate %dth sgl_list;" + " failed\n", + iod->npages + 1); + return -ENOMEM; + } + list[iod->npages++] = sg_list; + + i = 0; + memcpy(&sg_list[i++], link, sizeof(*link)); + spraid_sgl_set_seg(link, sgl_dma, nsge); + } + + spraid_sgl_set_data(&sg_list[i++], sg); + sg = sg_next(sg); + } while (--nsge > 0); + + return 0; +} + +#define SPRAID_RW_FUA BIT(14) + +static void spraid_setup_rw_cmd(struct spraid_dev *hdev, + struct spraid_rw_command *rw, + struct scsi_cmnd *scmd) +{ + u32 start_lba_lo, start_lba_hi; + u32 datalength = 0; + u16 control = 0; + + start_lba_lo = 0; + start_lba_hi = 0; + + if (scmd->sc_data_direction == DMA_TO_DEVICE) { + rw->opcode = SPRAID_CMD_WRITE; + } else if (scmd->sc_data_direction == DMA_FROM_DEVICE) { + rw->opcode = SPRAID_CMD_READ; + } else { + dev_err(hdev->dev, + "Invalid IO for unsupported data direction: %d\n", + scmd->sc_data_direction); + WARN_ON(1); + } + + /* 6-byte READ(0x08) or WRITE(0x0A) cdb */ + if (scmd->cmd_len == 6) { + datalength = (u32)(scmd->cmnd[4] == 0 ? + IO_6_DEFAULT_TX_LEN : scmd->cmnd[4]); + start_lba_lo = (u32)get_unaligned_be24(&scmd->cmnd[1]); + + start_lba_lo &= 0x1FFFFF; + } + + /* 10-byte READ(0x28) or WRITE(0x2A) cdb */ + else if (scmd->cmd_len == 10) { + datalength = (u32)get_unaligned_be16(&scmd->cmnd[7]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[2]); + + if (scmd->cmnd[1] & FUA_MASK) + control |= SPRAID_RW_FUA; + } + + /* 12-byte READ(0xA8) or WRITE(0xAA) cdb */ + else if (scmd->cmd_len == 12) { + datalength = get_unaligned_be32(&scmd->cmnd[6]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[2]); + + if (scmd->cmnd[1] & FUA_MASK) + control |= SPRAID_RW_FUA; + } + /* 16-byte READ(0x88) or WRITE(0x8A) cdb */ + else if (scmd->cmd_len == 16) { + datalength = get_unaligned_be32(&scmd->cmnd[10]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[6]); + start_lba_hi = get_unaligned_be32(&scmd->cmnd[2]); + + if (scmd->cmnd[1] & FUA_MASK) + control |= SPRAID_RW_FUA; + } + /* 32-byte READ(0x88) or WRITE(0x8A) cdb */ + else if (scmd->cmd_len == 32) { + datalength = get_unaligned_be32(&scmd->cmnd[28]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[16]); + start_lba_hi = get_unaligned_be32(&scmd->cmnd[12]); + + if (scmd->cmnd[10] & FUA_MASK) + control |= SPRAID_RW_FUA; + } + + if (unlikely(datalength > U16_MAX || datalength == 0)) { + dev_err(hdev->dev, + "Invalid IO for illegal transfer data length: %u\n", + datalength); + WARN_ON(1); + } + + rw->slba = cpu_to_le64(((u64)start_lba_hi << 32) | start_lba_lo); + /* 0base for nlb */ + rw->nlb = cpu_to_le16((u16)(datalength - 1)); + rw->control = cpu_to_le16(control); +} + +static void spraid_setup_nonio_cmd(struct spraid_dev *hdev, + struct spraid_scsi_nonio *scsi_nonio, + struct scsi_cmnd *scmd) +{ + scsi_nonio->buffer_len = cpu_to_le32(scsi_bufflen(scmd)); + + switch (scmd->sc_data_direction) { + case DMA_NONE: + scsi_nonio->opcode = SPRAID_CMD_NONIO_NONE; + break; + case DMA_TO_DEVICE: + scsi_nonio->opcode = SPRAID_CMD_NONIO_TODEV; + break; + case DMA_FROM_DEVICE: + scsi_nonio->opcode = SPRAID_CMD_NONIO_FROMDEV; + break; + default: + dev_err(hdev->dev, + "Invalid IO for unsupported data direction: %d\n", + scmd->sc_data_direction); + WARN_ON(1); + } +} + +static void spraid_setup_ioq_cmd(struct spraid_dev *hdev, + struct spraid_ioq_command *ioq_cmd, + struct scsi_cmnd *scmd) +{ + memcpy(ioq_cmd->common.cdb, scmd->cmnd, scmd->cmd_len); + ioq_cmd->common.cdb_len = scmd->cmd_len; + + if (spraid_is_rw_scmd(scmd)) + spraid_setup_rw_cmd(hdev, &ioq_cmd->rw, scmd); + else + spraid_setup_nonio_cmd(hdev, &ioq_cmd->scsi_nonio, scmd); +} + +static int spraid_init_iod(struct spraid_dev *hdev, struct spraid_iod *iod, + struct spraid_ioq_command *ioq_cmd, + struct scsi_cmnd *scmd) +{ + if (unlikely(!iod->sense)) { + dev_err(hdev->dev, "Allocate sense data buffer failed\n"); + return -ENOMEM; + } + ioq_cmd->common.sense_addr = cpu_to_le64(iod->sense_dma); + ioq_cmd->common.sense_len = cpu_to_le16(SCSI_SENSE_BUFFERSIZE); + + iod->nsge = 0; + iod->npages = -1; + iod->use_sgl = 0; + iod->sg_drv_mgmt = false; + WRITE_ONCE(iod->state, SPRAID_CMD_IDLE); + + return 0; +} + +static void spraid_free_iod_res(struct spraid_dev *hdev, struct spraid_iod *iod) +{ + const int last_prp = hdev->page_size / sizeof(__le64) - 1; + dma_addr_t dma_addr, next_dma_addr; + struct spraid_sgl_desc *sg_list; + __le64 *prp_list; + void *addr; + int i; + + dma_addr = iod->first_dma; + if (iod->npages == 0) + dma_pool_free(iod->spraidq->prp_small_pool, + spraid_iod_list(iod)[0], dma_addr); + + for (i = 0; i < iod->npages; i++) { + addr = spraid_iod_list(iod)[i]; + + if (iod->use_sgl) { + sg_list = addr; + next_dma_addr = + le64_to_cpu((sg_list[SGES_PER_PAGE - 1]).addr); + } else { + prp_list = addr; + next_dma_addr = le64_to_cpu(prp_list[last_prp]); + } + + dma_pool_free(hdev->prp_page_pool, addr, dma_addr); + dma_addr = next_dma_addr; + } + + if (iod->sg_drv_mgmt && iod->sg != iod->inline_sg) { + iod->sg_drv_mgmt = false; + mempool_free(iod->sg, hdev->iod_mempool); + } + + iod->sense = NULL; + iod->npages = -1; +} + +static int spraid_io_map_data(struct spraid_dev *hdev, struct spraid_iod *iod, + struct scsi_cmnd *scmd, + struct spraid_ioq_command *ioq_cmd) +{ + int ret; + + iod->nsge = scsi_dma_map(scmd); + + /* No data to DMA, it may be scsi no-rw command */ + if (unlikely(iod->nsge == 0)) + return 0; + + iod->length = scsi_bufflen(scmd); + iod->sg = scsi_sglist(scmd); + iod->use_sgl = !spraid_is_prp(hdev, scmd, iod->nsge); + + if (iod->use_sgl) { + ret = spraid_setup_ioq_cmd_sgl(hdev, scmd, ioq_cmd, iod); + } else { + ret = spraid_setup_prps(hdev, iod); + ioq_cmd->common.dptr.prp1 = + cpu_to_le64(sg_dma_address(iod->sg)); + ioq_cmd->common.dptr.prp2 = cpu_to_le64(iod->first_dma); + } + + if (ret) + scsi_dma_unmap(scmd); + + return ret; +} + +static void spraid_map_status(struct spraid_iod *iod, struct scsi_cmnd *scmd, + struct spraid_completion *cqe) +{ + scsi_set_resid(scmd, 0); + + switch ((le16_to_cpu(cqe->status) >> 1) & 0x7f) { + case FW_STAT_OK: + set_host_byte(scmd, DID_OK); + break; + case FW_STAT_NEED_CHECK: + set_host_byte(scmd, DID_OK); + scmd->result |= le16_to_cpu(cqe->status) >> 8; + if (scmd->result & SAM_STAT_CHECK_CONDITION) { + memset(scmd->sense_buffer, 0, SCSI_SENSE_BUFFERSIZE); + memcpy(scmd->sense_buffer, iod->sense, + SCSI_SENSE_BUFFERSIZE); + scmd->result = + (scmd->result & 0x00ffffff) | (DRIVER_SENSE << 24); + } + break; + case FW_STAT_ABORTED: + set_host_byte(scmd, DID_ABORT); + break; + case FW_STAT_NEED_RETRY: + set_host_byte(scmd, DID_REQUEUE); + break; + default: + set_host_byte(scmd, DID_BAD_TARGET); + break; + } +} + +static inline void spraid_get_tag_from_scmd(struct scsi_cmnd *scmd, + u16 *qid, u16 *cid) +{ + u32 tag = blk_mq_unique_tag(scmd->request); + + *qid = blk_mq_unique_tag_to_hwq(tag) + 1; + *cid = blk_mq_unique_tag_to_tag(tag); +} + +static int spraid_queue_command(struct Scsi_Host *shost, struct scsi_cmnd *scmd) +{ + struct spraid_iod *iod = scsi_cmd_priv(scmd); + struct spraid_dev *hdev = shost_priv(shost); + struct scsi_device *sdev = scmd->device; + struct spraid_sdev_hostdata *hostdata; + struct spraid_ioq_command ioq_cmd; + struct spraid_queue *ioq; + unsigned long elapsed; + u16 hwq, cid; + int ret; + + if (unlikely(!scmd)) { + dev_err(hdev->dev, "err, scmd is null\n"); + return 0; + } + + if (unlikely(hdev->state != SPRAID_LIVE)) { + set_host_byte(scmd, DID_NO_CONNECT); + scmd->scsi_done(scmd); + return 0; + } + + if (log_debug_switch) + scsi_print_command(scmd); + + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + hostdata = sdev->hostdata; + ioq = &hdev->queues[hwq]; + memset(&ioq_cmd, 0, sizeof(ioq_cmd)); + ioq_cmd.rw.hdid = cpu_to_le32(hostdata->hdid); + ioq_cmd.rw.command_id = cid; + + spraid_setup_ioq_cmd(hdev, &ioq_cmd, scmd); + + ret = cid * SCSI_SENSE_BUFFERSIZE; + iod->sense = ioq->sense + ret; + iod->sense_dma = ioq->sense_dma_addr + ret; + + ret = spraid_init_iod(hdev, iod, &ioq_cmd, scmd); + if (unlikely(ret)) + return SCSI_MLQUEUE_HOST_BUSY; + + iod->spraidq = ioq; + ret = spraid_io_map_data(hdev, iod, scmd, &ioq_cmd); + if (unlikely(ret)) { + dev_err(hdev->dev, "spraid_io_map_data Err.\n"); + set_host_byte(scmd, DID_ERROR); + scmd->scsi_done(scmd); + ret = 0; + goto deinit_iod; + } + + WRITE_ONCE(iod->state, SPRAID_CMD_IN_FLIGHT); + spraid_submit_cmd(ioq, &ioq_cmd); + elapsed = jiffies - scmd->jiffies_at_alloc; + dev_log_dbg(hdev->dev, + "cid[%d] qid[%d] submit IO cost %3ld.%3ld seconds\n", + cid, hwq, elapsed / HZ, elapsed % HZ); + return 0; + +deinit_iod: + spraid_free_iod_res(hdev, iod); + return ret; +} + +static int spraid_match_dev(struct spraid_dev *hdev, u16 idx, + struct scsi_device *sdev) +{ + if (SPRAID_DEV_INFO_FLAG_VALID(hdev->devices[idx].flag)) { + if (sdev->channel == hdev->devices[idx].channel && + sdev->id == le16_to_cpu(hdev->devices[idx].target) && + sdev->lun < hdev->devices[idx].lun) { + dev_info(hdev->dev, + "Match device success, channel;" + "target:lun[%d:%d:%d]\n", + hdev->devices[idx].channel, + hdev->devices[idx].target, + hdev->devices[idx].lun); + return 1; + } + } + + return 0; +} + +static int spraid_slave_alloc(struct scsi_device *sdev) +{ + struct spraid_sdev_hostdata *hostdata; + struct spraid_dev *hdev; + u16 idx; + + hdev = shost_priv(sdev->host); + hostdata = kzalloc(sizeof(*hostdata), GFP_KERNEL); + if (!hostdata) { + dev_err(hdev->dev, "Alloc scsi host data memory failed\n"); + return -ENOMEM; + } + + down_read(&hdev->devices_rwsem); + for (idx = 0; idx < le32_to_cpu(hdev->ctrl_info->nd); idx++) { + if (spraid_match_dev(hdev, idx, sdev)) + goto scan_host; + } + up_read(&hdev->devices_rwsem); + + kfree(hostdata); + return -ENXIO; + +scan_host: + hostdata->hdid = le32_to_cpu(hdev->devices[idx].hdid); + hostdata->max_io_kb = le16_to_cpu(hdev->devices[idx].max_io_kb); + hostdata->attr = hdev->devices[idx].attr; + hostdata->flag = hdev->devices[idx].flag; + hostdata->rg_id = 0xff; + sdev->hostdata = hostdata; + up_read(&hdev->devices_rwsem); + return 0; +} + +static void spraid_slave_destroy(struct scsi_device *sdev) +{ + kfree(sdev->hostdata); + sdev->hostdata = NULL; +} + +static int spraid_slave_configure(struct scsi_device *sdev) +{ + u16 idx; + unsigned int timeout = scmd_tmout_nonpt * HZ; + struct spraid_dev *hdev = shost_priv(sdev->host); + struct spraid_sdev_hostdata *hostdata = sdev->hostdata; + u32 max_sec = sdev->host->max_sectors; + + if (hostdata) { + idx = hostdata->hdid - 1; + if (sdev->channel == hdev->devices[idx].channel && + sdev->id == le16_to_cpu(hdev->devices[idx].target) && + sdev->lun < hdev->devices[idx].lun) { + if (SPRAID_DEV_INFO_ATTR_PT(hdev->devices[idx].attr)) + timeout = 30 * HZ; + else + timeout = scmd_tmout_nonpt * HZ; + max_sec = le16_to_cpu(hdev->devices[idx].max_io_kb) + << 1; + } else { + dev_err(hdev->dev, + "[%s] err, sdev->channel:id:lun[%d:%d:%lld];" + "devices[%d], channel:target:lun[%d:%d:%d]\n", + __func__, sdev->channel, sdev->id, sdev->lun, + idx, hdev->devices[idx].channel, + hdev->devices[idx].target, + hdev->devices[idx].lun); + } + } else { + dev_err(hdev->dev, "[%s] err, sdev->hostdata is null\n", + __func__); + } + + blk_queue_rq_timeout(sdev->request_queue, timeout); + sdev->eh_timeout = timeout; + + if ((max_sec == 0) || (max_sec > sdev->host->max_sectors)) + max_sec = sdev->host->max_sectors; + + dev_info(hdev->dev, + "[%s] sdev->channel:id:lun[%d:%d:%lld];" + " scmd_timeout[%d]s, maxsec[%d]\n", + __func__, sdev->channel, sdev->id, + sdev->lun, timeout / HZ, max_sec); + + return 0; +} + +static void spraid_shost_init(struct spraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + u8 domain, bus; + u32 dev_func; + + domain = pci_domain_nr(pdev->bus); + bus = pdev->bus->number; + dev_func = pdev->devfn; + + hdev->shost->nr_hw_queues = hdev->online_queues - 1; + hdev->shost->can_queue = (hdev->ioq_depth - SPRAID_PTCMDS_PERQ); + + hdev->shost->sg_tablesize = le16_to_cpu(hdev->ctrl_info->max_num_sge); + /* 512B per sector */ + hdev->shost->max_sectors = + (1U << ((hdev->ctrl_info->mdts) * 1U) << 12) / 512; + hdev->shost->cmd_per_lun = MAX_CMD_PER_DEV; + hdev->shost->max_channel = + le16_to_cpu(hdev->ctrl_info->max_channel) - 1; + hdev->shost->max_id = le32_to_cpu(hdev->ctrl_info->max_tgt_id); + hdev->shost->max_lun = le16_to_cpu(hdev->ctrl_info->max_lun); + + hdev->shost->this_id = -1; + hdev->shost->unique_id = (domain << 16) | (bus << 8) | dev_func; + hdev->shost->max_cmd_len = MAX_CDB_LEN; + hdev->shost->hostt->cmd_size = max(spraid_cmd_size(hdev, false, true), + spraid_cmd_size(hdev, false, false)); +} + +static inline void spraid_host_deinit(struct spraid_dev *hdev) +{ + ida_free(&spraid_instance_ida, hdev->instance); +} + +static int spraid_alloc_queue(struct spraid_dev *hdev, u16 qid, u16 depth) +{ + struct spraid_queue *spraidq = &hdev->queues[qid]; + int ret = 0; + + if (hdev->queue_count > qid) { + dev_info(hdev->dev, "[%s] warn: queue[%d] is exist\n", + __func__, qid); + return 0; + } + + spraidq->cqes = dma_alloc_coherent(hdev->dev, CQ_SIZE(depth), + &spraidq->cq_dma_addr, + GFP_KERNEL | __GFP_ZERO); + if (!spraidq->cqes) + return -ENOMEM; + + spraidq->sq_cmds = dma_alloc_coherent(hdev->dev, SQ_SIZE(qid, depth), + &spraidq->sq_dma_addr, + GFP_KERNEL); + if (!spraidq->sq_cmds) { + ret = -ENOMEM; + goto free_cqes; + } + + spin_lock_init(&spraidq->sq_lock); + spin_lock_init(&spraidq->cq_lock); + spraidq->hdev = hdev; + spraidq->q_depth = depth; + spraidq->qid = qid; + spraidq->cq_vector = -1; + hdev->queue_count++; + + /* alloc sense buffer */ + spraidq->sense = dma_alloc_coherent(hdev->dev, SENSE_SIZE(depth), + &spraidq->sense_dma_addr, + GFP_KERNEL | __GFP_ZERO); + if (!spraidq->sense) { + ret = -ENOMEM; + goto free_sq_cmds; + } + + return 0; + +free_sq_cmds: + dma_free_coherent(hdev->dev, SQ_SIZE(qid, depth), + (void *)spraidq->sq_cmds, spraidq->sq_dma_addr); +free_cqes: + dma_free_coherent(hdev->dev, CQ_SIZE(depth), (void *)spraidq->cqes, + spraidq->cq_dma_addr); + return ret; +} + +static int spraid_wait_ready(struct spraid_dev *hdev, u64 cap, bool enabled) +{ + unsigned long timeout = + ((SPRAID_CAP_TIMEOUT(cap) + 1) * SPRAID_CAP_TIMEOUT_UNIT_MS) + jiffies; + u32 bit = enabled ? SPRAID_CSTS_RDY : 0; + + while ((readl(hdev->bar + SPRAID_REG_CSTS) & SPRAID_CSTS_RDY) != bit) { + usleep_range(1000, 2000); + if (fatal_signal_pending(current)) + return -EINTR; + + if (time_after(jiffies, timeout)) { + dev_err(hdev->dev, "Device not ready; aborting %s\n", + enabled ? "initialisation" : "reset"); + return -ENODEV; + } + } + return 0; +} + +static int spraid_shutdown_ctrl(struct spraid_dev *hdev) +{ + unsigned long timeout = hdev->ctrl_info->rtd3e + jiffies; + + hdev->ctrl_config &= ~SPRAID_CC_SHN_MASK; + hdev->ctrl_config |= SPRAID_CC_SHN_NORMAL; + writel(hdev->ctrl_config, hdev->bar + SPRAID_REG_CC); + + while ((readl(hdev->bar + SPRAID_REG_CSTS) & SPRAID_CSTS_SHST_MASK) != + SPRAID_CSTS_SHST_CMPLT) { + msleep(100); + if (fatal_signal_pending(current)) + return -EINTR; + if (time_after(jiffies, timeout)) { + dev_err(hdev->dev, + "Device shutdown incomplete; abort shutdown\n"); + return -ENODEV; + } + } + return 0; +} + +static int spraid_disable_ctrl(struct spraid_dev *hdev) +{ + hdev->ctrl_config &= ~SPRAID_CC_SHN_MASK; + hdev->ctrl_config &= ~SPRAID_CC_ENABLE; + writel(hdev->ctrl_config, hdev->bar + SPRAID_REG_CC); + + return spraid_wait_ready(hdev, hdev->cap, false); +} + +static int spraid_enable_ctrl(struct spraid_dev *hdev) +{ + u64 cap = hdev->cap; + u32 dev_page_min = SPRAID_CAP_MPSMIN(cap) + 12; + u32 page_shift = PAGE_SHIFT; + + if (page_shift < dev_page_min) { + dev_err(hdev->dev, + "Minimum device page size[%u], too large for host[%u]\n", + 1U << dev_page_min, 1U << page_shift); + return -ENODEV; + } + + page_shift = min_t(unsigned int, SPRAID_CAP_MPSMAX(cap) + 12, + PAGE_SHIFT); + hdev->page_size = 1U << page_shift; + + hdev->ctrl_config = SPRAID_CC_CSS_NVM; + hdev->ctrl_config |= (page_shift - 12) << SPRAID_CC_MPS_SHIFT; + hdev->ctrl_config |= SPRAID_CC_AMS_RR | SPRAID_CC_SHN_NONE; + hdev->ctrl_config |= SPRAID_CC_IOSQES | SPRAID_CC_IOCQES; + hdev->ctrl_config |= SPRAID_CC_ENABLE; + writel(hdev->ctrl_config, hdev->bar + SPRAID_REG_CC); + + return spraid_wait_ready(hdev, cap, true); +} + +static void spraid_init_queue(struct spraid_queue *spraidq, u16 qid) +{ + struct spraid_dev *hdev = spraidq->hdev; + + memset((void *)spraidq->cqes, 0, CQ_SIZE(spraidq->q_depth)); + + spraidq->sq_tail = 0; + spraidq->cq_head = 0; + spraidq->cq_phase = 1; + spraidq->q_db = &hdev->dbs[qid * 2 * hdev->db_stride]; + spraidq->prp_small_pool = hdev->prp_small_pool[qid % small_pool_num]; + hdev->online_queues++; +} + +static inline bool spraid_cqe_pending(struct spraid_queue *spraidq) +{ + return (le16_to_cpu(spraidq->cqes[spraidq->cq_head].status) & 1) == + spraidq->cq_phase; +} + +static void spraid_sata_report_zone_handle(struct scsi_cmnd *scmd, + struct spraid_iod *iod) +{ + int i = 0; + unsigned int bytes = 0; + struct scatterlist *sg = scsi_sglist(scmd); + + scsi_for_each_sg(scmd, sg, iod->nsge, i) { + unsigned int offset = 0; + + if (bytes == 0) { + char *hdr; + u32 list_length; + u64 max_lba, opt_lba; + u16 same; + + hdr = sg_virt(sg); + + list_length = get_unaligned_le32(&hdr[0]); + same = get_unaligned_le16(&hdr[4]); + max_lba = get_unaligned_le64(&hdr[8]); + opt_lba = get_unaligned_le64(&hdr[16]); + put_unaligned_be32(list_length, &hdr[0]); + hdr[4] = same & 0xf; + put_unaligned_be64(max_lba, &hdr[8]); + put_unaligned_be64(opt_lba, &hdr[16]); + offset += 64; + bytes += 64; + } + while (offset < sg_dma_len(sg)) { + char *rec; + u8 cond, type, non_seq, reset; + u64 size, start, wp; + + rec = sg_virt(sg) + offset; + type = rec[0] & 0xf; + cond = (rec[1] >> 4) & 0xf; + non_seq = (rec[1] & 2); + reset = (rec[1] & 1); + size = get_unaligned_le64(&rec[8]); + start = get_unaligned_le64(&rec[16]); + wp = get_unaligned_le64(&rec[24]); + rec[0] = type; + rec[1] = (cond << 4) | non_seq | reset; + put_unaligned_be64(size, &rec[8]); + put_unaligned_be64(start, &rec[16]); + put_unaligned_be64(wp, &rec[24]); + WARN_ON(offset + 64 > sg_dma_len(sg)); + offset += 64; + bytes += 64; + } + } +} + +static inline void spraid_handle_ata_cmd(struct spraid_dev *hdev, + struct scsi_cmnd *scmd, + struct spraid_iod *iod) +{ + if (hdev->ctrl_info->card_type != SPRAID_CARD_HBA) + return; + + switch (scmd->cmnd[0]) { + case ZBC_IN: + dev_info(hdev->dev, "[%s] process report zone\n", __func__); + spraid_sata_report_zone_handle(scmd, iod); + break; + default: + break; + } +} + +static void spraid_complete_ioq_cmnd(struct spraid_queue *ioq, + struct spraid_completion *cqe) +{ + struct spraid_dev *hdev = ioq->hdev; + struct blk_mq_tags *tags; + struct scsi_cmnd *scmd; + struct spraid_iod *iod; + struct request *req; + unsigned long elapsed; + + tags = hdev->shost->tag_set.tags[ioq->qid - 1]; + req = blk_mq_tag_to_rq(tags, cqe->cmd_id); + if (unlikely(!req || !blk_mq_request_started(req))) { + dev_warn(hdev->dev, "Invalid id %d completed on queue %d\n", + cqe->cmd_id, ioq->qid); + return; + } + + scmd = blk_mq_rq_to_pdu(req); + iod = scsi_cmd_priv(scmd); + + elapsed = jiffies - scmd->jiffies_at_alloc; + dev_log_dbg(hdev->dev, + "cid[%d] qid[%d] finish IO cost %3ld.%3ld seconds\n", + cqe->cmd_id, ioq->qid, elapsed / HZ, elapsed % HZ); + + if (cmpxchg(&iod->state, SPRAID_CMD_IN_FLIGHT, SPRAID_CMD_COMPLETE) != + SPRAID_CMD_IN_FLIGHT) { + dev_warn(hdev->dev, + "cid[%d] qid[%d] enters abnormal handler;" + " cost %3ld.%3ld seconds\n", + cqe->cmd_id, ioq->qid, elapsed / HZ, elapsed % HZ); + WRITE_ONCE(iod->state, SPRAID_CMD_TMO_COMPLETE); + + if (iod->nsge) { + iod->nsge = 0; + scsi_dma_unmap(scmd); + } + spraid_free_iod_res(hdev, iod); + + return; + } + + spraid_handle_ata_cmd(hdev, scmd, iod); + + spraid_map_status(iod, scmd, cqe); + if (iod->nsge) { + iod->nsge = 0; + scsi_dma_unmap(scmd); + } + spraid_free_iod_res(hdev, iod); + scmd->scsi_done(scmd); +} + +static void spraid_complete_adminq_cmnd(struct spraid_queue *adminq, + struct spraid_completion *cqe) +{ + struct spraid_dev *hdev = adminq->hdev; + struct spraid_cmd *adm_cmd; + + adm_cmd = hdev->adm_cmds + cqe->cmd_id; + if (unlikely(adm_cmd->state == SPRAID_CMD_IDLE)) { + dev_warn(adminq->hdev->dev, + "Invalid id %d completed on queue %d\n", + cqe->cmd_id, le16_to_cpu(cqe->sq_id)); + return; + } + + adm_cmd->status = le16_to_cpu(cqe->status) >> 1; + adm_cmd->result0 = le32_to_cpu(cqe->result); + adm_cmd->result1 = le32_to_cpu(cqe->result1); + + complete(&adm_cmd->cmd_done); +} + +static void spraid_send_aen(struct spraid_dev *hdev, u16 cid); + +static void spraid_complete_aen(struct spraid_queue *spraidq, + struct spraid_completion *cqe) +{ + struct spraid_dev *hdev = spraidq->hdev; + u32 result = le32_to_cpu(cqe->result); + + dev_info(hdev->dev, "rcv aen, cid[%d], status[0x%x], result[0x%x]\n", + cqe->cmd_id, le16_to_cpu(cqe->status) >> 1, result); + + spraid_send_aen(hdev, cqe->cmd_id); + + if ((le16_to_cpu(cqe->status) >> 1) != SPRAID_SC_SUCCESS) + return; + switch (result & 0x7) { + case SPRAID_AEN_NOTICE: + spraid_handle_aen_notice(hdev, result); + break; + case SPRAID_AEN_VS: + spraid_handle_aen_vs(hdev, result, le32_to_cpu(cqe->result1)); + break; + default: + dev_warn(hdev->dev, "Unsupported async event type: %u\n", + result & 0x7); + break; + } +} + +static void spraid_complete_ioq_sync_cmnd(struct spraid_queue *ioq, + struct spraid_completion *cqe) +{ + struct spraid_dev *hdev = ioq->hdev; + struct spraid_cmd *ptcmd; + + ptcmd = hdev->ioq_ptcmds + (ioq->qid - 1) * SPRAID_PTCMDS_PERQ + + cqe->cmd_id - SPRAID_IO_BLK_MQ_DEPTH; + + ptcmd->status = le16_to_cpu(cqe->status) >> 1; + ptcmd->result0 = le32_to_cpu(cqe->result); + ptcmd->result1 = le32_to_cpu(cqe->result1); + + complete(&ptcmd->cmd_done); +} + +static inline void spraid_handle_cqe(struct spraid_queue *spraidq, u16 idx) +{ + struct spraid_completion *cqe = &spraidq->cqes[idx]; + struct spraid_dev *hdev = spraidq->hdev; + + if (unlikely(cqe->cmd_id >= spraidq->q_depth)) { + dev_err(hdev->dev, + "Invalid command id[%d] completed on queue %d\n", + cqe->cmd_id, cqe->sq_id); + return; + } + + dev_log_dbg(hdev->dev, "cid[%d] qid[%d];" + " result[0x%x], sq_id[%d], status[0x%x]\n", + cqe->cmd_id, spraidq->qid, le32_to_cpu(cqe->result), + le16_to_cpu(cqe->sq_id), le16_to_cpu(cqe->status)); + + if (unlikely(spraidq->qid == 0 + && cqe->cmd_id >= SPRAID_AQ_BLK_MQ_DEPTH)) { + spraid_complete_aen(spraidq, cqe); + return; + } + + if (unlikely(spraidq->qid && cqe->cmd_id >= SPRAID_IO_BLK_MQ_DEPTH)) { + spraid_complete_ioq_sync_cmnd(spraidq, cqe); + return; + } + + if (spraidq->qid) + spraid_complete_ioq_cmnd(spraidq, cqe); + else + spraid_complete_adminq_cmnd(spraidq, cqe); +} + +static void spraid_complete_cqes(struct spraid_queue *spraidq, + u16 start, u16 end) +{ + while (start != end) { + spraid_handle_cqe(spraidq, start); + if (++start == spraidq->q_depth) + start = 0; + } +} + +static inline void spraid_update_cq_head(struct spraid_queue *spraidq) +{ + if (++spraidq->cq_head == spraidq->q_depth) { + spraidq->cq_head = 0; + spraidq->cq_phase = !spraidq->cq_phase; + } +} + +static inline bool spraid_process_cq(struct spraid_queue *spraidq, + u16 *start, u16 *end, int tag) +{ + bool found = false; + + *start = spraidq->cq_head; + while (!found && spraid_cqe_pending(spraidq)) { + if (spraidq->cqes[spraidq->cq_head].cmd_id == tag) + found = true; + spraid_update_cq_head(spraidq); + } + *end = spraidq->cq_head; + + if (*start != *end) + writel(spraidq->cq_head, + spraidq->q_db + spraidq->hdev->db_stride); + + return found; +} + +static bool spraid_poll_cq(struct spraid_queue *spraidq, int cid) +{ + u16 start, end; + bool found; + + if (!spraid_cqe_pending(spraidq)) + return 0; + + spin_lock_irq(&spraidq->cq_lock); + found = spraid_process_cq(spraidq, &start, &end, cid); + spin_unlock_irq(&spraidq->cq_lock); + + spraid_complete_cqes(spraidq, start, end); + return found; +} + +static irqreturn_t spraid_irq(int irq, void *data) +{ + struct spraid_queue *spraidq = data; + irqreturn_t ret = IRQ_NONE; + u16 start, end; + + spin_lock(&spraidq->cq_lock); + if (spraidq->cq_head != spraidq->last_cq_head) + ret = IRQ_HANDLED; + + spraid_process_cq(spraidq, &start, &end, -1); + spraidq->last_cq_head = spraidq->cq_head; + spin_unlock(&spraidq->cq_lock); + + if (start != end) { + spraid_complete_cqes(spraidq, start, end); + ret = IRQ_HANDLED; + } + return ret; +} + +static int spraid_setup_admin_queue(struct spraid_dev *hdev) +{ + struct spraid_queue *adminq = &hdev->queues[0]; + u32 aqa; + int ret; + + dev_info(hdev->dev, "[%s] start disable ctrl\n", __func__); + + ret = spraid_disable_ctrl(hdev); + if (ret) + return ret; + + ret = spraid_alloc_queue(hdev, 0, SPRAID_AQ_DEPTH); + if (ret) + return ret; + + aqa = adminq->q_depth - 1; + aqa |= aqa << 16; + writel(aqa, hdev->bar + SPRAID_REG_AQA); + lo_hi_writeq(adminq->sq_dma_addr, hdev->bar + SPRAID_REG_ASQ); + lo_hi_writeq(adminq->cq_dma_addr, hdev->bar + SPRAID_REG_ACQ); + + dev_info(hdev->dev, "[%s] start enable ctrl\n", __func__); + + ret = spraid_enable_ctrl(hdev); + if (ret) { + ret = -ENODEV; + goto free_queue; + } + + adminq->cq_vector = 0; + spraid_init_queue(adminq, 0); + ret = pci_request_irq(hdev->pdev, adminq->cq_vector, spraid_irq, NULL, + adminq, "spraid%d_q%d", + hdev->instance, adminq->qid); + + if (ret) { + adminq->cq_vector = -1; + hdev->online_queues--; + goto free_queue; + } + + dev_info(hdev->dev, "[%s] success, queuecount:[%d], onlinequeue:[%d]\n", + __func__, hdev->queue_count, hdev->online_queues); + + return 0; + +free_queue: + spraid_free_queue(adminq); + return ret; +} + +static u32 spraid_bar_size(struct spraid_dev *hdev, u32 nr_ioqs) +{ + return (SPRAID_REG_DBS + ((nr_ioqs + 1) * 8 * hdev->db_stride)); +} + +static int spraid_alloc_admin_cmds(struct spraid_dev *hdev) +{ + int i; + + INIT_LIST_HEAD(&hdev->adm_cmd_list); + spin_lock_init(&hdev->adm_cmd_lock); + + hdev->adm_cmds = kcalloc_node(SPRAID_AQ_BLK_MQ_DEPTH, + sizeof(struct spraid_cmd), + GFP_KERNEL, hdev->numa_node); + + if (!hdev->adm_cmds) { + dev_err(hdev->dev, "Alloc admin cmds failed\n"); + return -ENOMEM; + } + + for (i = 0; i < SPRAID_AQ_BLK_MQ_DEPTH; i++) { + hdev->adm_cmds[i].qid = 0; + hdev->adm_cmds[i].cid = i; + list_add_tail(&(hdev->adm_cmds[i].list), &hdev->adm_cmd_list); + } + + dev_info(hdev->dev, "Alloc admin cmds success, num[%d]\n", + SPRAID_AQ_BLK_MQ_DEPTH); + + return 0; +} + +static void spraid_free_admin_cmds(struct spraid_dev *hdev) +{ + kfree(hdev->adm_cmds); + hdev->adm_cmds = NULL; + INIT_LIST_HEAD(&hdev->adm_cmd_list); +} + +static struct spraid_cmd *spraid_get_cmd(struct spraid_dev *hdev, + enum spraid_cmd_type type) +{ + struct spraid_cmd *cmd = NULL; + unsigned long flags; + struct list_head *head = &hdev->adm_cmd_list; + spinlock_t *slock = &hdev->adm_cmd_lock; + + if (type == SPRAID_CMD_IOPT) { + head = &hdev->ioq_pt_list; + slock = &hdev->ioq_pt_lock; + } + + spin_lock_irqsave(slock, flags); + if (list_empty(head)) { + spin_unlock_irqrestore(slock, flags); + dev_err(hdev->dev, "err, cmd[%d] list empty\n", type); + return NULL; + } + cmd = list_entry(head->next, struct spraid_cmd, list); + list_del_init(&cmd->list); + spin_unlock_irqrestore(slock, flags); + + WRITE_ONCE(cmd->state, SPRAID_CMD_IN_FLIGHT); + + return cmd; +} + +static void spraid_put_cmd(struct spraid_dev *hdev, struct spraid_cmd *cmd, + enum spraid_cmd_type type) +{ + unsigned long flags; + struct list_head *head = &hdev->adm_cmd_list; + spinlock_t *slock = &hdev->adm_cmd_lock; + + if (type == SPRAID_CMD_IOPT) { + head = &hdev->ioq_pt_list; + slock = &hdev->ioq_pt_lock; + } + + spin_lock_irqsave(slock, flags); + WRITE_ONCE(cmd->state, SPRAID_CMD_IDLE); + list_add_tail(&cmd->list, head); + spin_unlock_irqrestore(slock, flags); +} + + +static int spraid_submit_admin_sync_cmd(struct spraid_dev *hdev, + struct spraid_admin_command *cmd, + u32 *result0, u32 *result1, u32 timeout) +{ + struct spraid_cmd *adm_cmd = spraid_get_cmd(hdev, SPRAID_CMD_ADM); + + if (!adm_cmd) { + dev_err(hdev->dev, "err, get admin cmd failed\n"); + return -EFAULT; + } + + timeout = timeout ? timeout : ADMIN_TIMEOUT; + + init_completion(&adm_cmd->cmd_done); + + cmd->common.command_id = adm_cmd->cid; + spraid_submit_cmd(&hdev->queues[0], cmd); + + if (!wait_for_completion_timeout(&adm_cmd->cmd_done, timeout)) { + dev_err(hdev->dev, "[%s] cid[%d] qid[%d] timeout;" + " opcode[0x%x] subopcode[0x%x]\n", + __func__, adm_cmd->cid, adm_cmd->qid, + cmd->usr_cmd.opcode, cmd->usr_cmd.info_0.subopcode); + WRITE_ONCE(adm_cmd->state, SPRAID_CMD_TIMEOUT); + spraid_put_cmd(hdev, adm_cmd, SPRAID_CMD_ADM); + return -EINVAL; + } + + if (result0) + *result0 = adm_cmd->result0; + if (result1) + *result1 = adm_cmd->result1; + + spraid_put_cmd(hdev, adm_cmd, SPRAID_CMD_ADM); + + return adm_cmd->status; +} + +static int spraid_create_cq(struct spraid_dev *hdev, u16 qid, + struct spraid_queue *spraidq, u16 cq_vector) +{ + struct spraid_admin_command admin_cmd; + int flags = SPRAID_QUEUE_PHYS_CONTIG | SPRAID_CQ_IRQ_ENABLED; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.create_cq.opcode = SPRAID_ADMIN_CREATE_CQ; + admin_cmd.create_cq.prp1 = cpu_to_le64(spraidq->cq_dma_addr); + admin_cmd.create_cq.cqid = cpu_to_le16(qid); + admin_cmd.create_cq.qsize = cpu_to_le16(spraidq->q_depth - 1); + admin_cmd.create_cq.cq_flags = cpu_to_le16(flags); + admin_cmd.create_cq.irq_vector = cpu_to_le16(cq_vector); + + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); +} + +static int spraid_create_sq(struct spraid_dev *hdev, u16 qid, + struct spraid_queue *spraidq) +{ + struct spraid_admin_command admin_cmd; + int flags = SPRAID_QUEUE_PHYS_CONTIG; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.create_sq.opcode = SPRAID_ADMIN_CREATE_SQ; + admin_cmd.create_sq.prp1 = cpu_to_le64(spraidq->sq_dma_addr); + admin_cmd.create_sq.sqid = cpu_to_le16(qid); + admin_cmd.create_sq.qsize = cpu_to_le16(spraidq->q_depth - 1); + admin_cmd.create_sq.sq_flags = cpu_to_le16(flags); + admin_cmd.create_sq.cqid = cpu_to_le16(qid); + + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); +} + +static void spraid_free_queue(struct spraid_queue *spraidq) +{ + struct spraid_dev *hdev = spraidq->hdev; + + hdev->queue_count--; + dma_free_coherent(hdev->dev, CQ_SIZE(spraidq->q_depth), + (void *)spraidq->cqes, spraidq->cq_dma_addr); + dma_free_coherent(hdev->dev, SQ_SIZE(spraidq->qid, spraidq->q_depth), + spraidq->sq_cmds, spraidq->sq_dma_addr); + dma_free_coherent(hdev->dev, SENSE_SIZE(spraidq->q_depth), + spraidq->sense, spraidq->sense_dma_addr); +} + +static void spraid_free_admin_queue(struct spraid_dev *hdev) +{ + spraid_free_queue(&hdev->queues[0]); +} + +static void spraid_free_io_queues(struct spraid_dev *hdev) +{ + int i; + + for (i = hdev->queue_count - 1; i >= 1; i--) + spraid_free_queue(&hdev->queues[i]); +} + +static int spraid_delete_queue(struct spraid_dev *hdev, u8 op, u16 id) +{ + struct spraid_admin_command admin_cmd; + int ret; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.delete_queue.opcode = op; + admin_cmd.delete_queue.qid = cpu_to_le16(id); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + + if (ret) + dev_err(hdev->dev, "Delete %s:[%d] failed\n", + (op == SPRAID_ADMIN_DELETE_CQ) ? "cq" : "sq", id); + + return ret; +} + +static int spraid_delete_cq(struct spraid_dev *hdev, u16 cqid) +{ + return spraid_delete_queue(hdev, SPRAID_ADMIN_DELETE_CQ, cqid); +} + +static int spraid_delete_sq(struct spraid_dev *hdev, u16 sqid) +{ + return spraid_delete_queue(hdev, SPRAID_ADMIN_DELETE_SQ, sqid); +} + +static int spraid_create_queue(struct spraid_queue *spraidq, u16 qid) +{ + struct spraid_dev *hdev = spraidq->hdev; + u16 cq_vector; + int ret; + + cq_vector = (hdev->num_vecs == 1) ? 0 : qid; + ret = spraid_create_cq(hdev, qid, spraidq, cq_vector); + if (ret) + return ret; + + ret = spraid_create_sq(hdev, qid, spraidq); + if (ret) + goto delete_cq; + + spraid_init_queue(spraidq, qid); + spraidq->cq_vector = cq_vector; + + ret = pci_request_irq(hdev->pdev, cq_vector, spraid_irq, NULL, + spraidq, "spraid%d_q%d", hdev->instance, qid); + + if (ret) { + dev_err(hdev->dev, "Request queue[%d] irq failed\n", qid); + goto delete_sq; + } + + return 0; + +delete_sq: + spraidq->cq_vector = -1; + hdev->online_queues--; + spraid_delete_sq(hdev, qid); +delete_cq: + spraid_delete_cq(hdev, qid); + + return ret; +} + +static int spraid_create_io_queues(struct spraid_dev *hdev) +{ + u32 i, max; + int ret = 0; + + max = min(hdev->max_qid, hdev->queue_count - 1); + for (i = hdev->online_queues; i <= max; i++) { + ret = spraid_create_queue(&hdev->queues[i], i); + if (ret) { + dev_err(hdev->dev, "Create queue[%d] failed\n", i); + break; + } + } + + dev_info(hdev->dev, "[%s] queue_count[%d], online_queue[%d]", + __func__, hdev->queue_count, hdev->online_queues); + + return ret >= 0 ? 0 : ret; +} + +static int spraid_set_features(struct spraid_dev *hdev, u32 fid, + u32 dword11, void *buffer, + size_t buflen, u32 *result) +{ + struct spraid_admin_command admin_cmd; + int ret; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + + if (buffer && buflen) { + data_ptr = dma_alloc_coherent(hdev->dev, buflen, + &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memcpy(data_ptr, buffer, buflen); + } + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.features.opcode = SPRAID_ADMIN_SET_FEATURES; + admin_cmd.features.fid = cpu_to_le32(fid); + admin_cmd.features.dword11 = cpu_to_le32(dword11); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, result, NULL, 0); + + if (data_ptr) + dma_free_coherent(hdev->dev, buflen, data_ptr, data_dma); + + return ret; +} + +static int spraid_configure_timestamp(struct spraid_dev *hdev) +{ + __le64 ts; + int ret; + + ts = cpu_to_le64(ktime_to_ms(ktime_get_real())); + ret = spraid_set_features(hdev, SPRAID_FEAT_TIMESTAMP, + 0, &ts, sizeof(ts), NULL); + + if (ret) + dev_err(hdev->dev, "set timestamp failed: %d\n", ret); + return ret; +} + +static int spraid_set_queue_cnt(struct spraid_dev *hdev, u32 *cnt) +{ + u32 q_cnt = (*cnt - 1) | ((*cnt - 1) << 16); + u32 nr_ioqs, result; + int status; + + status = spraid_set_features(hdev, SPRAID_FEAT_NUM_QUEUES, + q_cnt, NULL, 0, &result); + if (status) { + dev_err(hdev->dev, "Set queue count failed, status: %d\n", + status); + return -EIO; + } + + nr_ioqs = min(result & 0xffff, result >> 16) + 1; + *cnt = min(*cnt, nr_ioqs); + if (*cnt == 0) { + dev_err(hdev->dev, "Illegal queue count: zero\n"); + return -EIO; + } + return 0; +} + +static int spraid_setup_io_queues(struct spraid_dev *hdev) +{ + struct spraid_queue *adminq = &hdev->queues[0]; + struct pci_dev *pdev = hdev->pdev; + u32 nr_ioqs = num_online_cpus(); + u32 i, size; + int ret; + + struct irq_affinity affd = { + .pre_vectors = 1 + }; + + ret = spraid_set_queue_cnt(hdev, &nr_ioqs); + if (ret < 0) + return ret; + + size = spraid_bar_size(hdev, nr_ioqs); + ret = spraid_remap_bar(hdev, size); + if (ret) + return -ENOMEM; + + adminq->q_db = hdev->dbs; + + pci_free_irq(pdev, 0, adminq); + pci_free_irq_vectors(pdev); + + ret = pci_alloc_irq_vectors_affinity(pdev, 1, (nr_ioqs + 1), + PCI_IRQ_ALL_TYPES | PCI_IRQ_AFFINITY, &affd); + if (ret <= 0) + return -EIO; + + hdev->num_vecs = ret; + + hdev->max_qid = max(ret - 1, 1); + + ret = pci_request_irq(pdev, adminq->cq_vector, spraid_irq, NULL, + adminq, "spraid%d_q%d", + hdev->instance, adminq->qid); + if (ret) { + dev_err(hdev->dev, "Request admin irq failed\n"); + adminq->cq_vector = -1; + return ret; + } + + for (i = hdev->queue_count; i <= hdev->max_qid; i++) { + ret = spraid_alloc_queue(hdev, i, hdev->ioq_depth); + if (ret) + break; + } + dev_info(hdev->dev, "[%s] max_qid: %d, queue_count: %d;" + " online_queue: %d, ioq_depth: %d\n", + __func__, hdev->max_qid, hdev->queue_count, + hdev->online_queues, hdev->ioq_depth); + + return spraid_create_io_queues(hdev); +} + +static void spraid_delete_io_queues(struct spraid_dev *hdev) +{ + u16 queues = hdev->online_queues - 1; + u8 opcode = SPRAID_ADMIN_DELETE_SQ; + u16 i, pass; + + if (!pci_device_is_present(hdev->pdev)) { + dev_err(hdev->dev, + "pci_device is not present, skip disable io queues\n"); + return; + } + + if (hdev->online_queues < 2) { + dev_err(hdev->dev, "[%s] err, io queue has been delete\n", + __func__); + return; + } + + for (pass = 0; pass < 2; pass++) { + for (i = queues; i > 0; i--) + if (spraid_delete_queue(hdev, opcode, i)) + break; + + opcode = SPRAID_ADMIN_DELETE_CQ; + } +} + +static void spraid_remove_io_queues(struct spraid_dev *hdev) +{ + spraid_delete_io_queues(hdev); + spraid_free_io_queues(hdev); +} + +static void spraid_pci_disable(struct spraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + u32 i; + + for (i = 0; i < hdev->online_queues; i++) + pci_free_irq(pdev, hdev->queues[i].cq_vector, &hdev->queues[i]); + pci_free_irq_vectors(pdev); + if (pci_is_enabled(pdev)) { + pci_disable_pcie_error_reporting(pdev); + pci_disable_device(pdev); + } + hdev->online_queues = 0; +} + +static void spraid_disable_admin_queue(struct spraid_dev *hdev, bool shutdown) +{ + struct spraid_queue *adminq = &hdev->queues[0]; + u16 start, end; + + if (pci_device_is_present(hdev->pdev)) { + if (shutdown) + spraid_shutdown_ctrl(hdev); + else + spraid_disable_ctrl(hdev); + } + + if (hdev->queue_count == 0) { + dev_err(hdev->dev, "[%s] err, admin queue has been delete\n", + __func__); + return; + } + + spin_lock_irq(&adminq->cq_lock); + spraid_process_cq(adminq, &start, &end, -1); + spin_unlock_irq(&adminq->cq_lock); + + spraid_complete_cqes(adminq, start, end); + spraid_free_admin_queue(hdev); +} + +static int spraid_create_dma_pools(struct spraid_dev *hdev) +{ + int i; + char poolname[20] = { 0 }; + + hdev->prp_page_pool = dma_pool_create("prp list page", hdev->dev, + PAGE_SIZE, PAGE_SIZE, 0); + + if (!hdev->prp_page_pool) { + dev_err(hdev->dev, "create prp_page_pool failed\n"); + return -ENOMEM; + } + + for (i = 0; i < small_pool_num; i++) { + sprintf(poolname, "prp_list_256_%d", i); + hdev->prp_small_pool[i] = + dma_pool_create(poolname, hdev->dev, SMALL_POOL_SIZE, + SMALL_POOL_SIZE, 0); + + if (!hdev->prp_small_pool[i]) { + dev_err(hdev->dev, "create prp_small_pool %d failed\n", + i); + goto destroy_prp_small_pool; + } + } + + return 0; + +destroy_prp_small_pool: + while (i > 0) + dma_pool_destroy(hdev->prp_small_pool[--i]); + dma_pool_destroy(hdev->prp_page_pool); + + return -ENOMEM; +} + +static void spraid_destroy_dma_pools(struct spraid_dev *hdev) +{ + int i; + + for (i = 0; i < small_pool_num; i++) + dma_pool_destroy(hdev->prp_small_pool[i]); + dma_pool_destroy(hdev->prp_page_pool); +} + +static int spraid_get_dev_list(struct spraid_dev *hdev, + struct spraid_dev_info *devices) +{ + u32 nd = le32_to_cpu(hdev->ctrl_info->nd); + struct spraid_admin_command admin_cmd; + struct spraid_dev_list *list_buf; + dma_addr_t data_dma = 0; + u32 i, idx, hdid, ndev; + int ret = 0; + + list_buf = dma_alloc_coherent(hdev->dev, PAGE_SIZE, + &data_dma, GFP_KERNEL); + if (!list_buf) + return -ENOMEM; + + for (idx = 0; idx < nd;) { + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.get_info.opcode = SPRAID_ADMIN_GET_INFO; + admin_cmd.get_info.type = SPRAID_GET_INFO_DEV_LIST; + admin_cmd.get_info.cdw11 = cpu_to_le32(idx); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, + NULL, NULL, 0); + + if (ret) { + dev_err(hdev->dev, "Get device list failed, nd: %u;" + "idx: %u, ret: %d\n", + nd, idx, ret); + goto out; + } + ndev = le32_to_cpu(list_buf->dev_num); + + dev_info(hdev->dev, "ndev numbers: %u\n", ndev); + + for (i = 0; i < ndev; i++) { + hdid = le32_to_cpu(list_buf->devices[i].hdid); + dev_info(hdev->dev, "list_buf->devices[%d], hdid: %u;" + "target: %d, channel: %d, lun: %d, attr[0x%x]\n", + i, hdid, + le16_to_cpu(list_buf->devices[i].target), + list_buf->devices[i].channel, + list_buf->devices[i].lun, + list_buf->devices[i].attr); + if (hdid > nd || hdid == 0) { + dev_err(hdev->dev, "err, hdid[%d] invalid\n", + hdid); + continue; + } + memcpy(&devices[hdid - 1], &list_buf->devices[i], + sizeof(struct spraid_dev_info)); + } + idx += ndev; + + if (idx < MAX_DEV_ENTRY_PER_PAGE_4K) + break; + } + +out: + dma_free_coherent(hdev->dev, PAGE_SIZE, list_buf, data_dma); + return ret; +} + +static void spraid_send_aen(struct spraid_dev *hdev, u16 cid) +{ + struct spraid_queue *adminq = &hdev->queues[0]; + struct spraid_admin_command admin_cmd; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.common.opcode = SPRAID_ADMIN_ASYNC_EVENT; + admin_cmd.common.command_id = cid; + + spraid_submit_cmd(adminq, &admin_cmd); + dev_info(hdev->dev, "send aen, cid[%d]\n", cid); +} + +static inline void spraid_send_all_aen(struct spraid_dev *hdev) +{ + u16 i; + + for (i = 0; i < hdev->ctrl_info->aerl; i++) + spraid_send_aen(hdev, i + SPRAID_AQ_BLK_MQ_DEPTH); +} + +static int spraid_add_device(struct spraid_dev *hdev, + struct spraid_dev_info *device) +{ + struct Scsi_Host *shost = hdev->shost; + struct scsi_device *sdev; + + dev_info(hdev->dev, "add device, hdid: %u target: %d, channel: %d;" + " lun: %d, attr[0x%x]\n", + le32_to_cpu(device->hdid), le16_to_cpu(device->target), + device->channel, device->lun, device->attr); + + sdev = scsi_device_lookup(shost, device->channel, + le16_to_cpu(device->target), 0); + if (sdev) { + dev_warn(hdev->dev, "Device is already exist, channel: %d;" + " target_id: %d, lun: %d\n", + device->channel, le16_to_cpu(device->target), 0); + scsi_device_put(sdev); + return -EEXIST; + } + scsi_add_device(shost, device->channel, le16_to_cpu(device->target), 0); + return 0; +} + +static int spraid_rescan_device(struct spraid_dev *hdev, + struct spraid_dev_info *device) +{ + struct Scsi_Host *shost = hdev->shost; + struct scsi_device *sdev; + + dev_info(hdev->dev, "rescan device, hdid: %u target: %d, channel: %d;" + " lun: %d, attr[0x%x]\n", + le32_to_cpu(device->hdid), le16_to_cpu(device->target), + device->channel, device->lun, device->attr); + + sdev = scsi_device_lookup(shost, device->channel, + le16_to_cpu(device->target), 0); + if (!sdev) { + dev_warn(hdev->dev, "device is not exit rescan it, channel: %d;" + " target_id: %d, lun: %d\n", + device->channel, le16_to_cpu(device->target), 0); + return -ENODEV; + } + + scsi_rescan_device(&sdev->sdev_gendev); + scsi_device_put(sdev); + return 0; +} + +static int spraid_remove_device(struct spraid_dev *hdev, + struct spraid_dev_info *org_device) +{ + struct Scsi_Host *shost = hdev->shost; + struct scsi_device *sdev; + + dev_info(hdev->dev, "remove device, hdid: %u target: %d, channel: %d;" + " lun: %d, attr[0x%x]\n", + le32_to_cpu(org_device->hdid), le16_to_cpu(org_device->target), + org_device->channel, org_device->lun, org_device->attr); + + sdev = scsi_device_lookup(shost, org_device->channel, + le16_to_cpu(org_device->target), 0); + if (!sdev) { + dev_warn(hdev->dev, "device is not exit remove it, channel: %d;" + " target_id: %d, lun: %d\n", + org_device->channel, + le16_to_cpu(org_device->target), 0); + return -ENODEV; + } + + scsi_remove_device(sdev); + scsi_device_put(sdev); + return 0; +} + +static int spraid_dev_list_init(struct spraid_dev *hdev) +{ + u32 nd = le32_to_cpu(hdev->ctrl_info->nd); + int i, ret; + + hdev->devices = kzalloc_node(nd * sizeof(struct spraid_dev_info), + GFP_KERNEL, hdev->numa_node); + if (!hdev->devices) + return -ENOMEM; + + ret = spraid_get_dev_list(hdev, hdev->devices); + if (ret) { + dev_err(hdev->dev, + "Ignore failure of getting device list;" + " within initialization\n"); + return 0; + } + + for (i = 0; i < nd; i++) { + if (SPRAID_DEV_INFO_FLAG_VALID(hdev->devices[i].flag) && + SPRAID_DEV_INFO_ATTR_BOOT(hdev->devices[i].attr)) { + spraid_add_device(hdev, &hdev->devices[i]); + break; + } + } + return 0; +} + +static int luntarget_cmp_func(const void *l, const void *r) +{ + const struct spraid_dev_info *ln = l; + const struct spraid_dev_info *rn = r; + + if (ln->channel == rn->channel) + return le16_to_cpu(ln->target) - le16_to_cpu(rn->target); + + return ln->channel - rn->channel; +} + +static void spraid_scan_work(struct work_struct *work) +{ + struct spraid_dev *hdev = + container_of(work, struct spraid_dev, scan_work); + struct spraid_dev_info *devices, *org_devices; + struct spraid_dev_info *sortdevice; + u32 nd = le32_to_cpu(hdev->ctrl_info->nd); + u8 flag, org_flag; + int i, ret; + int count = 0; + + devices = kcalloc(nd, sizeof(struct spraid_dev_info), GFP_KERNEL); + if (!devices) + return; + + sortdevice = kcalloc(nd, sizeof(struct spraid_dev_info), GFP_KERNEL); + if (!sortdevice) + goto free_list; + + ret = spraid_get_dev_list(hdev, devices); + if (ret) + goto free_all; + org_devices = hdev->devices; + for (i = 0; i < nd; i++) { + org_flag = org_devices[i].flag; + flag = devices[i].flag; + + dev_log_dbg(hdev->dev, "i: %d, org_flag: 0x%x, flag: 0x%x\n", + i, org_flag, flag); + + if (SPRAID_DEV_INFO_FLAG_VALID(flag)) { + if (!SPRAID_DEV_INFO_FLAG_VALID(org_flag)) { + down_write(&hdev->devices_rwsem); + memcpy(&org_devices[i], &devices[i], + sizeof(struct spraid_dev_info)); + memcpy(&sortdevice[count++], &devices[i], + sizeof(struct spraid_dev_info)); + up_write(&hdev->devices_rwsem); + } else if (SPRAID_DEV_INFO_FLAG_CHANGE(flag)) { + spraid_rescan_device(hdev, &devices[i]); + } + } else { + if (SPRAID_DEV_INFO_FLAG_VALID(org_flag)) { + down_write(&hdev->devices_rwsem); + org_devices[i].flag &= 0xfe; + up_write(&hdev->devices_rwsem); + spraid_remove_device(hdev, &org_devices[i]); + } + } + } + + dev_info(hdev->dev, "scan work add device count = %d\n", count); + + sort(sortdevice, count, sizeof(sortdevice[0]), + luntarget_cmp_func, NULL); + + for (i = 0; i < count; i++) + spraid_add_device(hdev, &sortdevice[i]); + +free_all: + kfree(sortdevice); +free_list: + kfree(devices); +} + +static void spraid_timesyn_work(struct work_struct *work) +{ + struct spraid_dev *hdev = + container_of(work, struct spraid_dev, timesyn_work); + + spraid_configure_timestamp(hdev); +} + +static int spraid_init_ctrl_info(struct spraid_dev *hdev); +static void spraid_fw_act_work(struct work_struct *work) +{ + struct spraid_dev *hdev = + container_of(work, struct spraid_dev, fw_act_work); + + if (spraid_init_ctrl_info(hdev)) + dev_err(hdev->dev, "get ctrl info failed after fw act\n"); +} + +static void spraid_queue_scan(struct spraid_dev *hdev) +{ + queue_work(spraid_wq, &hdev->scan_work); +} + +static void spraid_handle_aen_notice(struct spraid_dev *hdev, u32 result) +{ + switch ((result & 0xff00) >> 8) { + case SPRAID_AEN_DEV_CHANGED: + spraid_queue_scan(hdev); + break; + case SPRAID_AEN_FW_ACT_START: + dev_info(hdev->dev, "fw activation starting\n"); + break; + case SPRAID_AEN_HOST_PROBING: + break; + default: + dev_warn(hdev->dev, "async event result %08x\n", result); + } +} + +static void spraid_handle_aen_vs(struct spraid_dev *hdev, + u32 result, u32 result1) +{ + switch ((result & 0xff00) >> 8) { + case SPRAID_AEN_TIMESYN: + queue_work(spraid_wq, &hdev->timesyn_work); + break; + case SPRAID_AEN_FW_ACT_FINISH: + dev_info(hdev->dev, "fw activation finish\n"); + queue_work(spraid_wq, &hdev->fw_act_work); + break; + case SPRAID_AEN_EVENT_MIN ... SPRAID_AEN_EVENT_MAX: + dev_info(hdev->dev, "rcv card event[%d];" + " param1[0x%x] param2[0x%x]\n", + (result & 0xff00) >> 8, result, result1); + break; + default: + dev_warn(hdev->dev, "async event result: 0x%x\n", result); + } +} + +static int spraid_alloc_resources(struct spraid_dev *hdev) +{ + int ret, nqueue; + + ret = ida_alloc(&spraid_instance_ida, GFP_KERNEL); + if (ret < 0) { + dev_err(hdev->dev, "Get instance id failed\n"); + return ret; + } + hdev->instance = ret; + + hdev->ctrl_info = kzalloc_node(sizeof(*hdev->ctrl_info), + GFP_KERNEL, hdev->numa_node); + if (!hdev->ctrl_info) { + ret = -ENOMEM; + goto release_instance; + } + + ret = spraid_create_dma_pools(hdev); + if (ret) + goto free_ctrl_info; + nqueue = num_possible_cpus() + 1; + hdev->queues = kcalloc_node(nqueue, sizeof(struct spraid_queue), + GFP_KERNEL, hdev->numa_node); + if (!hdev->queues) { + ret = -ENOMEM; + goto destroy_dma_pools; + } + + ret = spraid_alloc_admin_cmds(hdev); + if (ret) + goto free_queues; + + dev_info(hdev->dev, "[%s] queues num: %d\n", __func__, nqueue); + + return 0; + +free_queues: + kfree(hdev->queues); +destroy_dma_pools: + spraid_destroy_dma_pools(hdev); +free_ctrl_info: + kfree(hdev->ctrl_info); +release_instance: + ida_free(&spraid_instance_ida, hdev->instance); + return ret; +} + +static void spraid_free_resources(struct spraid_dev *hdev) +{ + spraid_free_admin_cmds(hdev); + kfree(hdev->queues); + spraid_destroy_dma_pools(hdev); + kfree(hdev->ctrl_info); + ida_free(&spraid_instance_ida, hdev->instance); +} + +static void spraid_bsg_unmap_data(struct spraid_dev *hdev, struct bsg_job *job) +{ + struct request *rq = blk_mq_rq_from_pdu(job); + struct spraid_iod *iod = job->dd_data; + enum dma_data_direction dma_dir = + rq_data_dir(rq) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; + + if (iod->nsge) + dma_unmap_sg(hdev->dev, iod->sg, iod->nsge, dma_dir); + + spraid_free_iod_res(hdev, iod); +} + +static int spraid_bsg_map_data(struct spraid_dev *hdev, struct bsg_job *job, + struct spraid_admin_command *cmd) +{ + struct request *rq = blk_mq_rq_from_pdu(job); + struct spraid_iod *iod = job->dd_data; + enum dma_data_direction dma_dir = + rq_data_dir(rq) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; + int ret = 0; + + iod->sg = job->request_payload.sg_list; + iod->nsge = job->request_payload.sg_cnt; + iod->length = job->request_payload.payload_len; + iod->use_sgl = false; + iod->npages = -1; + iod->sg_drv_mgmt = false; + + if (!iod->nsge) + goto out; + + ret = dma_map_sg_attrs(hdev->dev, iod->sg, iod->nsge, + dma_dir, DMA_ATTR_NO_WARN); + if (!ret) + goto out; + + ret = spraid_setup_prps(hdev, iod); + if (ret) + goto unmap; + + cmd->common.dptr.prp1 = cpu_to_le64(sg_dma_address(iod->sg)); + cmd->common.dptr.prp2 = cpu_to_le64(iod->first_dma); + + return 0; + +unmap: + dma_unmap_sg(hdev->dev, iod->sg, iod->nsge, dma_dir); +out: + return ret; +} + +static int spraid_get_ctrl_info(struct spraid_dev *hdev, + struct spraid_ctrl_info *ctrl_info) +{ + struct spraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + int ret; + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, + &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.get_info.opcode = SPRAID_ADMIN_GET_INFO; + admin_cmd.get_info.type = SPRAID_GET_INFO_CTRL; + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + if (!ret) + memcpy(ctrl_info, data_ptr, sizeof(struct spraid_ctrl_info)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, data_dma); + + return ret; +} + +static int spraid_init_ctrl_info(struct spraid_dev *hdev) +{ + int ret; + + hdev->ctrl_info->nd = cpu_to_le32(240); + hdev->ctrl_info->mdts = 8; + hdev->ctrl_info->max_cmds = cpu_to_le16(4096); + hdev->ctrl_info->max_num_sge = cpu_to_le16(128); + hdev->ctrl_info->max_channel = cpu_to_le16(4); + hdev->ctrl_info->max_tgt_id = cpu_to_le32(3239); + hdev->ctrl_info->max_lun = cpu_to_le16(2); + + ret = spraid_get_ctrl_info(hdev, hdev->ctrl_info); + if (ret) + dev_err(hdev->dev, "get controller info failed: %d\n", ret); + + dev_info(hdev->dev, "[%s]nd = %d\n", __func__, hdev->ctrl_info->nd); + dev_info(hdev->dev, "[%s]max_cmd = %d\n", + __func__, hdev->ctrl_info->max_cmds); + dev_info(hdev->dev, "[%s]max_channel = %d\n", + __func__, hdev->ctrl_info->max_channel); + dev_info(hdev->dev, "[%s]max_tgt_id = %d\n", + __func__, hdev->ctrl_info->max_tgt_id); + dev_info(hdev->dev, "[%s]max_lun = %d\n", + __func__, hdev->ctrl_info->max_lun); + dev_info(hdev->dev, "[%s]max_num_sge = %d\n", + __func__, hdev->ctrl_info->max_num_sge); + dev_info(hdev->dev, "[%s]lun_num_boot = %d\n", + __func__, hdev->ctrl_info->lun_num_in_boot); + dev_info(hdev->dev, "[%s]mdts = %d\n", __func__, hdev->ctrl_info->mdts); + dev_info(hdev->dev, "[%s]acl = %d\n", __func__, hdev->ctrl_info->acl); + dev_info(hdev->dev, "[%s]aer1 = %d\n", __func__, hdev->ctrl_info->aerl); + dev_info(hdev->dev, "[%s]card_type = %d\n", + __func__, hdev->ctrl_info->card_type); + dev_info(hdev->dev, "[%s]rtd3e = %d\n", + __func__, hdev->ctrl_info->rtd3e); + dev_info(hdev->dev, "[%s]sn = %s\n", __func__, hdev->ctrl_info->sn); + dev_info(hdev->dev, "[%s]fr = %s\n", __func__, hdev->ctrl_info->fr); + + if (!hdev->ctrl_info->aerl) + hdev->ctrl_info->aerl = 1; + if (hdev->ctrl_info->aerl > SPRAID_NR_AEN_COMMANDS) + hdev->ctrl_info->aerl = SPRAID_NR_AEN_COMMANDS; + + return 0; +} + +#define SPRAID_MAX_ADMIN_PAYLOAD_SIZE BIT(16) +static int spraid_alloc_iod_ext_mem_pool(struct spraid_dev *hdev) +{ + u16 max_sge = le16_to_cpu(hdev->ctrl_info->max_num_sge); + size_t alloc_size; + + alloc_size = spraid_iod_ext_size(hdev, SPRAID_MAX_ADMIN_PAYLOAD_SIZE, + max_sge, true, false); + if (alloc_size > PAGE_SIZE) + dev_warn(hdev->dev, "It is unreasonable ;" + " sg allocation more than one page\n"); + hdev->iod_mempool = mempool_create_node(1, mempool_kmalloc, + mempool_kfree, + (void *)alloc_size, GFP_KERNEL, + hdev->numa_node); + if (!hdev->iod_mempool) { + dev_err(hdev->dev, "Create iod extension memory pool failed\n"); + return -ENOMEM; + } + + return 0; +} + +static void spraid_free_iod_ext_mem_pool(struct spraid_dev *hdev) +{ + mempool_destroy(hdev->iod_mempool); +} + +static int spraid_user_admin_cmd(struct spraid_dev *hdev, struct bsg_job *job) +{ + struct spraid_bsg_request *bsg_req = job->request; + struct spraid_passthru_common_cmd *cmd = &(bsg_req->admcmd); + struct spraid_admin_command admin_cmd; + u32 timeout = msecs_to_jiffies(cmd->timeout_ms); + u32 result[2] = {0}; + int status; + + if (hdev->state >= SPRAID_RESETTING) { + dev_err(hdev->dev, "[%s] err, host state:[%d] is not right\n", + __func__, hdev->state); + return -EBUSY; + } + + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x] init\n", + __func__, cmd->opcode, cmd->info_0.subopcode); + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.common.opcode = cmd->opcode; + admin_cmd.common.flags = cmd->flags; + admin_cmd.common.hdid = cpu_to_le32(cmd->nsid); + admin_cmd.common.cdw2[0] = cpu_to_le32(cmd->cdw2); + admin_cmd.common.cdw2[1] = cpu_to_le32(cmd->cdw3); + admin_cmd.common.cdw10 = cpu_to_le32(cmd->cdw10); + admin_cmd.common.cdw11 = cpu_to_le32(cmd->cdw11); + admin_cmd.common.cdw12 = cpu_to_le32(cmd->cdw12); + admin_cmd.common.cdw13 = cpu_to_le32(cmd->cdw13); + admin_cmd.common.cdw14 = cpu_to_le32(cmd->cdw14); + admin_cmd.common.cdw15 = cpu_to_le32(cmd->cdw15); + + status = spraid_bsg_map_data(hdev, job, &admin_cmd); + if (status) { + dev_err(hdev->dev, "[%s] err, map data failed\n", __func__); + return status; + } + + status = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, &result[0], + &result[1], timeout); + if (status >= 0) { + job->reply_len = sizeof(result); + memcpy(job->reply, result, sizeof(result)); + } + + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x];" + " status[0x%x] result0[0x%x] result1[0x%x]\n", + __func__, cmd->opcode, cmd->info_0.subopcode, + status, result[0], result[1]); + + spraid_bsg_unmap_data(hdev, job); + + return status; +} + +static int spraid_alloc_ioq_ptcmds(struct spraid_dev *hdev) +{ + int i; + int ptnum = SPRAID_NR_IOQ_PTCMDS; + + INIT_LIST_HEAD(&hdev->ioq_pt_list); + spin_lock_init(&hdev->ioq_pt_lock); + + hdev->ioq_ptcmds = kcalloc_node(ptnum, sizeof(struct spraid_cmd), + GFP_KERNEL, hdev->numa_node); + + if (!hdev->ioq_ptcmds) { + dev_err(hdev->dev, "Alloc ioq_ptcmds failed\n"); + return -ENOMEM; + } + + for (i = 0; i < ptnum; i++) { + hdev->ioq_ptcmds[i].qid = i / SPRAID_PTCMDS_PERQ + 1; + hdev->ioq_ptcmds[i].cid = i % SPRAID_PTCMDS_PERQ + + SPRAID_IO_BLK_MQ_DEPTH; + list_add_tail(&(hdev->ioq_ptcmds[i].list), &hdev->ioq_pt_list); + } + + dev_info(hdev->dev, "Alloc ioq_ptcmds success, ptnum[%d]\n", ptnum); + + return 0; +} + +static void spraid_free_ioq_ptcmds(struct spraid_dev *hdev) +{ + kfree(hdev->ioq_ptcmds); + hdev->ioq_ptcmds = NULL; + + INIT_LIST_HEAD(&hdev->ioq_pt_list); +} + +static int spraid_submit_ioq_sync_cmd(struct spraid_dev *hdev, + struct spraid_ioq_command *cmd, + u32 *result, u32 *reslen, u32 timeout) +{ + int ret; + dma_addr_t sense_dma; + struct spraid_queue *ioq; + void *sense_addr = NULL; + struct spraid_cmd *pt_cmd = spraid_get_cmd(hdev, SPRAID_CMD_IOPT); + + if (!pt_cmd) { + dev_err(hdev->dev, "err, get ioq cmd failed\n"); + return -EFAULT; + } + + timeout = timeout ? timeout : ADMIN_TIMEOUT; + + init_completion(&pt_cmd->cmd_done); + + ioq = &hdev->queues[pt_cmd->qid]; + ret = pt_cmd->cid * SCSI_SENSE_BUFFERSIZE; + sense_addr = ioq->sense + ret; + sense_dma = ioq->sense_dma_addr + ret; + + cmd->common.sense_addr = cpu_to_le64(sense_dma); + cmd->common.sense_len = cpu_to_le16(SCSI_SENSE_BUFFERSIZE); + cmd->common.command_id = pt_cmd->cid; + + spraid_submit_cmd(ioq, cmd); + + if (!wait_for_completion_timeout(&pt_cmd->cmd_done, timeout)) { + dev_err(hdev->dev, "[%s] cid[%d] qid[%d] timeout;" + " opcode[0x%x] subopcode[0x%x]\n", + __func__, pt_cmd->cid, pt_cmd->qid, cmd->common.opcode, + (le32_to_cpu(cmd->common.cdw3[0]) & 0xffff)); + WRITE_ONCE(pt_cmd->state, SPRAID_CMD_TIMEOUT); + spraid_put_cmd(hdev, pt_cmd, SPRAID_CMD_IOPT); + return -EINVAL; + } + + if (result && reslen) { + if ((pt_cmd->status & 0x17f) == 0x101) { + memcpy(result, sense_addr, SCSI_SENSE_BUFFERSIZE); + *reslen = SCSI_SENSE_BUFFERSIZE; + } + } + + spraid_put_cmd(hdev, pt_cmd, SPRAID_CMD_IOPT); + + return pt_cmd->status; +} + +static int spraid_user_ioq_cmd(struct spraid_dev *hdev, struct bsg_job *job) +{ + struct spraid_bsg_request *bsg_req = + (struct spraid_bsg_request *)(job->request); + struct spraid_ioq_passthru_cmd *cmd = &(bsg_req->ioqcmd); + struct spraid_ioq_command ioq_cmd; + int status = 0; + u32 timeout = msecs_to_jiffies(cmd->timeout_ms); + + if (cmd->data_len > PAGE_SIZE) { + dev_err(hdev->dev, "[%s] data len bigger than 4k\n", __func__); + return -EFAULT; + } + + if (hdev->state != SPRAID_LIVE) { + dev_err(hdev->dev, "[%s] err, host state:[%d] is not live\n", + __func__, hdev->state); + return -EBUSY; + } + + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x] init;" + " datalen[%d]\n", + __func__, cmd->opcode, cmd->info_1.subopcode, cmd->data_len); + + memset(&ioq_cmd, 0, sizeof(ioq_cmd)); + ioq_cmd.common.opcode = cmd->opcode; + ioq_cmd.common.flags = cmd->flags; + ioq_cmd.common.hdid = cpu_to_le32(cmd->nsid); + ioq_cmd.common.sense_len = cpu_to_le16(cmd->info_0.res_sense_len); + ioq_cmd.common.cdb_len = cmd->info_0.cdb_len; + ioq_cmd.common.rsvd2 = cmd->info_0.rsvd0; + ioq_cmd.common.cdw3[0] = cpu_to_le32(cmd->cdw3); + ioq_cmd.common.cdw3[1] = cpu_to_le32(cmd->cdw4); + ioq_cmd.common.cdw3[2] = cpu_to_le32(cmd->cdw5); + + ioq_cmd.common.cdw10[0] = cpu_to_le32(cmd->cdw10); + ioq_cmd.common.cdw10[1] = cpu_to_le32(cmd->cdw11); + ioq_cmd.common.cdw10[2] = cpu_to_le32(cmd->cdw12); + ioq_cmd.common.cdw10[3] = cpu_to_le32(cmd->cdw13); + ioq_cmd.common.cdw10[4] = cpu_to_le32(cmd->cdw14); + ioq_cmd.common.cdw10[5] = cpu_to_le32(cmd->data_len); + + memcpy(ioq_cmd.common.cdb, &cmd->cdw16, cmd->info_0.cdb_len); + + ioq_cmd.common.cdw26[0] = cpu_to_le32(cmd->cdw26[0]); + ioq_cmd.common.cdw26[1] = cpu_to_le32(cmd->cdw26[1]); + ioq_cmd.common.cdw26[2] = cpu_to_le32(cmd->cdw26[2]); + ioq_cmd.common.cdw26[3] = cpu_to_le32(cmd->cdw26[3]); + + status = spraid_bsg_map_data(hdev, job, + (struct spraid_admin_command *)&ioq_cmd); + if (status) { + dev_err(hdev->dev, "[%s] err, map data failed\n", __func__); + return status; + } + + status = spraid_submit_ioq_sync_cmd(hdev, &ioq_cmd, job->reply, + &job->reply_len, timeout); + + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x], status[0x%x];" + " reply_len[%d]\n", + __func__, cmd->opcode, cmd->info_1.subopcode, + status, job->reply_len); + + spraid_bsg_unmap_data(hdev, job); + + return status; +} + +static bool spraid_check_scmd_completed(struct scsi_cmnd *scmd) +{ + struct spraid_dev *hdev = shost_priv(scmd->device->host); + struct spraid_iod *iod = scsi_cmd_priv(scmd); + struct spraid_queue *spraidq; + u16 hwq, cid; + + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + spraidq = &hdev->queues[hwq]; + if (READ_ONCE(iod->state) == SPRAID_CMD_COMPLETE + || spraid_poll_cq(spraidq, cid)) { + dev_warn(hdev->dev, "cid[%d] qid[%d] has been completed\n", + cid, spraidq->qid); + return true; + } + return false; +} + +static enum blk_eh_timer_return spraid_scmd_timeout(struct scsi_cmnd *scmd) +{ + struct spraid_iod *iod = scsi_cmd_priv(scmd); + unsigned int timeout = scmd->device->request_queue->rq_timeout; + + if (spraid_check_scmd_completed(scmd)) + goto out; + + if (time_after(jiffies, scmd->jiffies_at_alloc + timeout)) { + if (cmpxchg(&iod->state, + SPRAID_CMD_IN_FLIGHT, + SPRAID_CMD_TIMEOUT) == SPRAID_CMD_IN_FLIGHT) { + return BLK_EH_DONE; + } + } +out: + return BLK_EH_RESET_TIMER; +} + +/* send abort command by admin queue temporary */ +static int spraid_send_abort_cmd(struct spraid_dev *hdev, + u32 hdid, u16 qid, u16 cid) +{ + struct spraid_admin_command admin_cmd; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.abort.opcode = SPRAID_ADMIN_ABORT_CMD; + admin_cmd.abort.hdid = cpu_to_le32(hdid); + admin_cmd.abort.sqid = cpu_to_le16(qid); + admin_cmd.abort.cid = cpu_to_le16(cid); + + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); +} + +/* send reset command by admin quueue temporary */ +static int spraid_send_reset_cmd(struct spraid_dev *hdev, int type, u32 hdid) +{ + struct spraid_admin_command admin_cmd; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.reset.opcode = SPRAID_ADMIN_RESET; + admin_cmd.reset.hdid = cpu_to_le32(hdid); + admin_cmd.reset.type = type; + + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); +} + +static bool spraid_change_host_state(struct spraid_dev *hdev, + enum spraid_state newstate) +{ + unsigned long flags; + enum spraid_state oldstate; + bool change = false; + + spin_lock_irqsave(&hdev->state_lock, flags); + + oldstate = hdev->state; + switch (newstate) { + case SPRAID_LIVE: + switch (oldstate) { + case SPRAID_NEW: + case SPRAID_RESETTING: + change = true; + break; + default: + break; + } + break; + case SPRAID_RESETTING: + switch (oldstate) { + case SPRAID_LIVE: + change = true; + break; + default: + break; + } + break; + case SPRAID_DELETING: + if (oldstate != SPRAID_DELETING) + change = true; + break; + case SPRAID_DEAD: + switch (oldstate) { + case SPRAID_NEW: + case SPRAID_LIVE: + case SPRAID_RESETTING: + change = true; + break; + default: + break; + } + break; + default: + break; + } + if (change) + hdev->state = newstate; + spin_unlock_irqrestore(&hdev->state_lock, flags); + + dev_info(hdev->dev, "[%s][%d]->[%d], change[%d]\n", + __func__, oldstate, newstate, change); + + return change; +} + +static void spraid_back_fault_cqe(struct spraid_queue *ioq, + struct spraid_completion *cqe) +{ + struct spraid_dev *hdev = ioq->hdev; + struct blk_mq_tags *tags; + struct scsi_cmnd *scmd; + struct spraid_iod *iod; + struct request *req; + + tags = hdev->shost->tag_set.tags[ioq->qid - 1]; + req = blk_mq_tag_to_rq(tags, cqe->cmd_id); + if (unlikely(!req || !blk_mq_request_started(req))) + return; + + scmd = blk_mq_rq_to_pdu(req); + iod = scsi_cmd_priv(scmd); + + set_host_byte(scmd, DID_NO_CONNECT); + if (iod->nsge) + scsi_dma_unmap(scmd); + spraid_free_iod_res(hdev, iod); + scmd->scsi_done(scmd); + dev_warn(hdev->dev, "Back fault CQE, cid[%d] qid[%d]\n", + cqe->cmd_id, ioq->qid); +} + +static void spraid_back_all_io(struct spraid_dev *hdev) +{ + int i, j; + struct spraid_queue *ioq; + struct spraid_completion cqe = { 0 }; + + scsi_block_requests(hdev->shost); + + for (i = 1; i <= hdev->shost->nr_hw_queues; i++) { + ioq = &hdev->queues[i]; + for (j = 0; j < hdev->shost->can_queue; j++) { + cqe.cmd_id = j; + spraid_back_fault_cqe(ioq, &cqe); + } + } + + scsi_unblock_requests(hdev->shost); +} + +static void spraid_dev_disable(struct spraid_dev *hdev, bool shutdown) +{ + struct spraid_queue *adminq = &hdev->queues[0]; + u16 start, end; + unsigned long timeout = jiffies + 600 * HZ; + + if (pci_device_is_present(hdev->pdev)) { + if (shutdown) + spraid_shutdown_ctrl(hdev); + else + spraid_disable_ctrl(hdev); + } + + while (!time_after(jiffies, timeout)) { + if (!pci_device_is_present(hdev->pdev)) { + dev_info(hdev->dev, "[%s] pci_device not present;" + " skip wait\n", __func__); + break; + } + if (!spraid_wait_ready(hdev, hdev->cap, false)) { + dev_info(hdev->dev, + "[%s] wait ready success after reset\n", + __func__); + break; + } + dev_info(hdev->dev, "[%s] waiting csts_rdy ready\n", __func__); + } + + if (hdev->queue_count == 0) { + dev_err(hdev->dev, "[%s] warn, queue has been delete\n", + __func__); + return; + } + + spin_lock_irq(&adminq->cq_lock); + spraid_process_cq(adminq, &start, &end, -1); + spin_unlock_irq(&adminq->cq_lock); + spraid_complete_cqes(adminq, start, end); + + spraid_pci_disable(hdev); + + spraid_back_all_io(hdev); +} + +static void spraid_reset_work(struct work_struct *work) +{ + int ret; + struct spraid_dev *hdev = + container_of(work, struct spraid_dev, reset_work); + + if (hdev->state != SPRAID_RESETTING) { + dev_err(hdev->dev, "[%s] err, host is not reset state\n", + __func__); + return; + } + + dev_info(hdev->dev, "[%s] enter host reset\n", __func__); + + if (hdev->ctrl_config & SPRAID_CC_ENABLE) { + dev_info(hdev->dev, "[%s] start dev_disable\n", __func__); + spraid_dev_disable(hdev, false); + } + + ret = spraid_pci_enable(hdev); + if (ret) + goto out; + + ret = spraid_setup_admin_queue(hdev); + if (ret) + goto pci_disable; + + ret = spraid_setup_io_queues(hdev); + if (ret || hdev->online_queues <= hdev->shost->nr_hw_queues) + goto pci_disable; + + spraid_change_host_state(hdev, SPRAID_LIVE); + + spraid_send_all_aen(hdev); + + return; + +pci_disable: + spraid_pci_disable(hdev); +out: + spraid_change_host_state(hdev, SPRAID_DEAD); + dev_err(hdev->dev, "[%s] err, host reset failed\n", __func__); +} + +static int spraid_reset_work_sync(struct spraid_dev *hdev) +{ + if (!spraid_change_host_state(hdev, SPRAID_RESETTING)) { + dev_info(hdev->dev, "[%s] can't change to reset state\n", + __func__); + return -EBUSY; + } + + if (!queue_work(spraid_wq, &hdev->reset_work)) { + dev_err(hdev->dev, "[%s] err, host is already in reset state\n", + __func__); + return -EBUSY; + } + + flush_work(&hdev->reset_work); + if (hdev->state != SPRAID_LIVE) + return -ENODEV; + + return 0; +} + +static int spraid_wait_abnl_cmd_done(struct spraid_iod *iod) +{ + u16 times = 0; + + do { + if (READ_ONCE(iod->state) == SPRAID_CMD_TMO_COMPLETE) + break; + msleep(500); + times++; + } while (times <= SPRAID_WAIT_ABNL_CMD_TIMEOUT); + + /* wait command completion timeout after abort/reset success */ + if (times >= SPRAID_WAIT_ABNL_CMD_TIMEOUT) + return -ETIMEDOUT; + + return 0; +} + +static int spraid_abort_handler(struct scsi_cmnd *scmd) +{ + struct spraid_dev *hdev = shost_priv(scmd->device->host); + struct spraid_iod *iod = scsi_cmd_priv(scmd); + struct spraid_sdev_hostdata *hostdata; + u16 hwq, cid; + int ret; + + scsi_print_command(scmd); + + if (hdev->state != SPRAID_LIVE || !spraid_wait_abnl_cmd_done(iod) || + spraid_check_scmd_completed(scmd)) + return SUCCESS; + + hostdata = scmd->device->hostdata; + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + + dev_warn(hdev->dev, "cid[%d] qid[%d] timeout, aborting\n", cid, hwq); + ret = spraid_send_abort_cmd(hdev, hostdata->hdid, hwq, cid); + if (ret != ADMIN_ERR_TIMEOUT) { + ret = spraid_wait_abnl_cmd_done(iod); + if (ret) { + dev_warn(hdev->dev, "cid[%d] qid[%d] abort failed;" + " not found\n", cid, hwq); + return FAILED; + } + dev_warn(hdev->dev, "cid[%d] qid[%d] abort succ\n", cid, hwq); + return SUCCESS; + } + dev_warn(hdev->dev, "cid[%d] qid[%d] abort failed, timeout\n", + cid, hwq); + return FAILED; +} + +static int spraid_tgt_reset_handler(struct scsi_cmnd *scmd) +{ + struct spraid_dev *hdev = shost_priv(scmd->device->host); + struct spraid_iod *iod = scsi_cmd_priv(scmd); + struct spraid_sdev_hostdata *hostdata; + u16 hwq, cid; + int ret; + + scsi_print_command(scmd); + + if (hdev->state != SPRAID_LIVE || !spraid_wait_abnl_cmd_done(iod) || + spraid_check_scmd_completed(scmd)) + return SUCCESS; + + hostdata = scmd->device->hostdata; + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + + dev_warn(hdev->dev, "cid[%d] qid[%d] timeout, target reset\n", + cid, hwq); + ret = spraid_send_reset_cmd(hdev, SPRAID_RESET_TARGET, hostdata->hdid); + if (ret == 0) { + ret = spraid_wait_abnl_cmd_done(iod); + if (ret) { + dev_warn(hdev->dev, + "cid[%d] qid[%d]target reset failed;" + " not found\n", cid, hwq); + return FAILED; + } + + dev_warn(hdev->dev, "cid[%d] qid[%d] target reset success\n", + cid, hwq); + return SUCCESS; + } + + dev_warn(hdev->dev, "cid[%d] qid[%d] ret[%d] target reset failed\n", + cid, hwq, ret); + return FAILED; +} + +static int spraid_bus_reset_handler(struct scsi_cmnd *scmd) +{ + struct spraid_dev *hdev = shost_priv(scmd->device->host); + struct spraid_iod *iod = scsi_cmd_priv(scmd); + struct spraid_sdev_hostdata *hostdata; + u16 hwq, cid; + int ret; + + scsi_print_command(scmd); + + if (hdev->state != SPRAID_LIVE || !spraid_wait_abnl_cmd_done(iod) || + spraid_check_scmd_completed(scmd)) + return SUCCESS; + + hostdata = scmd->device->hostdata; + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + + dev_warn(hdev->dev, "cid[%d] qid[%d] timeout, bus reset\n", cid, hwq); + ret = spraid_send_reset_cmd(hdev, SPRAID_RESET_BUS, hostdata->hdid); + if (ret == 0) { + ret = spraid_wait_abnl_cmd_done(iod); + if (ret) { + dev_warn(hdev->dev, + "cid[%d] qid[%d] bus reset failed;" + " not found\n", cid, hwq); + return FAILED; + } + + dev_warn(hdev->dev, "cid[%d] qid[%d] bus reset succ\n", + cid, hwq); + return SUCCESS; + } + + dev_warn(hdev->dev, "cid[%d] qid[%d] ret[%d] bus reset failed\n", + cid, hwq, ret); + return FAILED; +} + +static int spraid_shost_reset_handler(struct scsi_cmnd *scmd) +{ + u16 hwq, cid; + struct spraid_dev *hdev = shost_priv(scmd->device->host); + + scsi_print_command(scmd); + if (hdev->state != SPRAID_LIVE || spraid_check_scmd_completed(scmd)) + return SUCCESS; + + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + dev_warn(hdev->dev, "cid[%d] qid[%d] host reset\n", cid, hwq); + + if (spraid_reset_work_sync(hdev)) { + dev_warn(hdev->dev, "cid[%d] qid[%d] host reset failed\n", + cid, hwq); + return FAILED; + } + + dev_warn(hdev->dev, "cid[%d] qid[%d] host reset success\n", cid, hwq); + + return SUCCESS; +} + +static pci_ers_result_t spraid_pci_error_detected(struct pci_dev *pdev, + pci_channel_state_t state) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "enter pci error detect, state:%d\n", state); + + switch (state) { + case pci_channel_io_normal: + dev_warn(hdev->dev, "channel is normal, do nothing\n"); + + return PCI_ERS_RESULT_CAN_RECOVER; + case pci_channel_io_frozen: + dev_warn(hdev->dev, + "channel io frozen, need reset controller\n"); + + scsi_block_requests(hdev->shost); + + spraid_change_host_state(hdev, SPRAID_RESETTING); + + return PCI_ERS_RESULT_NEED_RESET; + case pci_channel_io_perm_failure: + dev_warn(hdev->dev, "channel io failure, request disconnect\n"); + + return PCI_ERS_RESULT_DISCONNECT; + } + + return PCI_ERS_RESULT_NEED_RESET; +} + +static pci_ers_result_t spraid_pci_slot_reset(struct pci_dev *pdev) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "restart after slot reset\n"); + + pci_restore_state(pdev); + + if (!queue_work(spraid_wq, &hdev->reset_work)) { + dev_err(hdev->dev, "[%s] err, the device is resetting state\n", + __func__); + return PCI_ERS_RESULT_NONE; + } + + flush_work(&hdev->reset_work); + + scsi_unblock_requests(hdev->shost); + + return PCI_ERS_RESULT_RECOVERED; +} + +static void spraid_reset_done(struct pci_dev *pdev) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "enter spraid reset done\n"); +} + +static ssize_t csts_pp_show(struct device *cdev, + struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct spraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) { + ret = (readl(hdev->bar + SPRAID_REG_CSTS) + & SPRAID_CSTS_PP_MASK); + ret >>= SPRAID_CSTS_PP_SHIFT; + } + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t csts_shst_show(struct device *cdev, + struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct spraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) { + ret = (readl(hdev->bar + SPRAID_REG_CSTS) + & SPRAID_CSTS_SHST_MASK); + ret >>= SPRAID_CSTS_SHST_SHIFT; + } + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t csts_cfs_show(struct device *cdev, + struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct spraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) { + ret = (readl(hdev->bar + SPRAID_REG_CSTS) + & SPRAID_CSTS_CFS_MASK); + ret >>= SPRAID_CSTS_CFS_SHIFT; + } + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t csts_rdy_show(struct device *cdev, + struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct spraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) + ret = (readl(hdev->bar + SPRAID_REG_CSTS) & SPRAID_CSTS_RDY); + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t fw_version_show(struct device *cdev, + struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct spraid_dev *hdev = shost_priv(shost); + + return snprintf(buf, PAGE_SIZE, "%s\n", hdev->ctrl_info->fr); +} + +static DEVICE_ATTR_RO(csts_pp); +static DEVICE_ATTR_RO(csts_shst); +static DEVICE_ATTR_RO(csts_cfs); +static DEVICE_ATTR_RO(csts_rdy); +static DEVICE_ATTR_RO(fw_version); + +static struct device_attribute *spraid_host_attrs[] = { + &dev_attr_csts_pp, + &dev_attr_csts_shst, + &dev_attr_csts_cfs, + &dev_attr_csts_rdy, + &dev_attr_fw_version, + NULL, +}; + +static int spraid_get_vd_info(struct spraid_dev *hdev, + struct spraid_vd_info *vd_info, u16 vid) +{ + struct spraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + int ret; + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, + &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.usr_cmd.opcode = USR_CMD_READ; + admin_cmd.usr_cmd.info_0.subopcode = cpu_to_le16(USR_CMD_VDINFO); + admin_cmd.usr_cmd.info_1.data_len = cpu_to_le16(USR_CMD_RDLEN); + admin_cmd.usr_cmd.info_1.param_len = cpu_to_le16(VDINFO_PARAM_LEN); + admin_cmd.usr_cmd.cdw10 = cpu_to_le32(vid); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + if (!ret) + memcpy(vd_info, data_ptr, sizeof(struct spraid_vd_info)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, data_dma); + + return ret; +} + +static int spraid_get_bgtask(struct spraid_dev *hdev, + struct spraid_bgtask *bgtask) +{ + struct spraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + int ret; + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, + &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.usr_cmd.opcode = USR_CMD_READ; + admin_cmd.usr_cmd.info_0.subopcode = cpu_to_le16(USR_CMD_BGTASK); + admin_cmd.usr_cmd.info_1.data_len = cpu_to_le16(USR_CMD_RDLEN); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + if (!ret) + memcpy(bgtask, data_ptr, sizeof(struct spraid_bgtask)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, data_dma); + + return ret; +} + +static ssize_t raid_level_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct spraid_dev *hdev; + struct spraid_vd_info *vd_info; + struct spraid_sdev_hostdata *hostdata; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !SPRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = spraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) + vd_info->rg_level = ARRAY_SIZE(raid_levels) - 1; + + ret = (vd_info->rg_level < ARRAY_SIZE(raid_levels)) ? + vd_info->rg_level : (ARRAY_SIZE(raid_levels) - 1); + + kfree(vd_info); + + return snprintf(buf, PAGE_SIZE, "RAID-%s\n", raid_levels[ret]); +} + +static ssize_t raid_state_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct spraid_dev *hdev; + struct spraid_vd_info *vd_info; + struct spraid_sdev_hostdata *hostdata; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !SPRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = spraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) { + vd_info->vd_status = 0; + vd_info->rg_id = 0xff; + } + + ret = (vd_info->vd_status < ARRAY_SIZE(raid_states)) ? + vd_info->vd_status : 0; + + kfree(vd_info); + + return snprintf(buf, PAGE_SIZE, "%s\n", raid_states[ret]); +} + +static ssize_t raid_resync_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct spraid_dev *hdev; + struct spraid_vd_info *vd_info; + struct spraid_bgtask *bgtask; + struct spraid_sdev_hostdata *hostdata; + u8 rg_id, i, progress = 0; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !SPRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = spraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) + goto out; + + rg_id = vd_info->rg_id; + + bgtask = (struct spraid_bgtask *)vd_info; + ret = spraid_get_bgtask(hdev, bgtask); + if (ret) + goto out; + for (i = 0; i < bgtask->task_num; i++) { + if ((bgtask->bgtask[i].type == BGTASK_TYPE_REBUILD) && + (le16_to_cpu(bgtask->bgtask[i].vd_id) == rg_id)) + progress = bgtask->bgtask[i].progress; + } + +out: + kfree(vd_info); + return snprintf(buf, PAGE_SIZE, "%d\n", progress); +} + +static DEVICE_ATTR_RO(raid_level); +static DEVICE_ATTR_RO(raid_state); +static DEVICE_ATTR_RO(raid_resync); + +static struct device_attribute *spraid_dev_attrs[] = { + &dev_attr_raid_level, + &dev_attr_raid_state, + &dev_attr_raid_resync, + NULL, +}; + +static struct pci_error_handlers spraid_err_handler = { + .error_detected = spraid_pci_error_detected, + .slot_reset = spraid_pci_slot_reset, + .reset_done = spraid_reset_done, +}; + +static int spraid_sysfs_host_reset(struct Scsi_Host *shost, int reset_type) +{ + int ret; + struct spraid_dev *hdev = shost_priv(shost); + + dev_info(hdev->dev, "[%s] start sysfs host reset cmd\n", __func__); + ret = spraid_reset_work_sync(hdev); + dev_info(hdev->dev, "[%s] stop sysfs host reset cmd[%d]\n", + __func__, ret); + + return ret; +} + +static struct scsi_host_template spraid_driver_template = { + .module = THIS_MODULE, + .name = "Ramaxel Logic spraid driver", + .proc_name = "spraid", + .queuecommand = spraid_queue_command, + .slave_alloc = spraid_slave_alloc, + .slave_destroy = spraid_slave_destroy, + .slave_configure = spraid_slave_configure, + .eh_timed_out = spraid_scmd_timeout, + .eh_abort_handler = spraid_abort_handler, + .eh_target_reset_handler = spraid_tgt_reset_handler, + .eh_bus_reset_handler = spraid_bus_reset_handler, + .eh_host_reset_handler = spraid_shost_reset_handler, + .change_queue_depth = scsi_change_queue_depth, + .this_id = -1, + .shost_attrs = spraid_host_attrs, + .sdev_attrs = spraid_dev_attrs, + .host_reset = spraid_sysfs_host_reset, +}; + +static void spraid_shutdown(struct pci_dev *pdev) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + + spraid_remove_io_queues(hdev); + spraid_disable_admin_queue(hdev, true); +} + +/* bsg dispatch user command */ +static int spraid_bsg_host_dispatch(struct bsg_job *job) +{ + struct Scsi_Host *shost = dev_to_shost(job->dev); + struct spraid_dev *hdev = shost_priv(shost); + struct request *rq = blk_mq_rq_from_pdu(job); + struct spraid_bsg_request *bsg_req = job->request; + int ret = 0; + + dev_info(hdev->dev, "[%s] msgcode[%d], msglen[%d], timeout[%d];" + " req_nsge[%d], req_len[%d]\n", + __func__, bsg_req->msgcode, job->request_len, + rq->timeout, job->request_payload.sg_cnt, + job->request_payload.payload_len); + + job->reply_len = 0; + + switch (bsg_req->msgcode) { + case SPRAID_BSG_ADM: + ret = spraid_user_admin_cmd(hdev, job); + break; + case SPRAID_BSG_IOQ: + ret = spraid_user_ioq_cmd(hdev, job); + break; + default: + dev_info(hdev->dev, "[%s] unsupport msgcode[%d]\n", + __func__, bsg_req->msgcode); + break; + } + + if (ret > 0) + ret = ret | (ret << 8); + + bsg_job_done(job, ret, 0); + return 0; +} + +static inline void spraid_remove_bsg(struct spraid_dev *hdev) +{ + if (hdev->bsg_queue) { + bsg_unregister_queue(hdev->bsg_queue); + blk_cleanup_queue(hdev->bsg_queue); + } +} +static int spraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) +{ + struct spraid_dev *hdev; + struct Scsi_Host *shost; + int node, ret; + char bsg_name[15]; + + shost = scsi_host_alloc(&spraid_driver_template, sizeof(*hdev)); + if (!shost) { + dev_err(&pdev->dev, "Failed to allocate scsi host\n"); + return -ENOMEM; + } + hdev = shost_priv(shost); + hdev->pdev = pdev; + hdev->dev = get_device(&pdev->dev); + + node = dev_to_node(hdev->dev); + if (node == NUMA_NO_NODE) { + node = first_memory_node; + set_dev_node(hdev->dev, node); + } + hdev->numa_node = node; + hdev->shost = shost; + pci_set_drvdata(pdev, hdev); + + ret = spraid_dev_map(hdev); + if (ret) + goto put_dev; + + init_rwsem(&hdev->devices_rwsem); + INIT_WORK(&hdev->scan_work, spraid_scan_work); + INIT_WORK(&hdev->timesyn_work, spraid_timesyn_work); + INIT_WORK(&hdev->reset_work, spraid_reset_work); + INIT_WORK(&hdev->fw_act_work, spraid_fw_act_work); + spin_lock_init(&hdev->state_lock); + + ret = spraid_alloc_resources(hdev); + if (ret) + goto dev_unmap; + + ret = spraid_pci_enable(hdev); + if (ret) + goto resources_free; + + ret = spraid_setup_admin_queue(hdev); + if (ret) + goto pci_disable; + + ret = spraid_init_ctrl_info(hdev); + if (ret) + goto disable_admin_q; + + ret = spraid_alloc_iod_ext_mem_pool(hdev); + if (ret) + goto disable_admin_q; + + ret = spraid_setup_io_queues(hdev); + if (ret) + goto free_iod_mempool; + + spraid_shost_init(hdev); + + ret = scsi_add_host(hdev->shost, hdev->dev); + if (ret) { + dev_err(hdev->dev, "Add shost to system failed, ret: %d\n", + ret); + goto remove_io_queues; + } + + snprintf(bsg_name, sizeof(bsg_name), "spraid%d", shost->host_no); + hdev->bsg_queue = bsg_setup_queue(&shost->shost_gendev, bsg_name, + spraid_bsg_host_dispatch, + spraid_cmd_size(hdev, true, false)); + if (IS_ERR(hdev->bsg_queue)) { + dev_err(hdev->dev, "err, setup bsg failed\n"); + hdev->bsg_queue = NULL; + goto remove_io_queues; + } + + if (hdev->online_queues == SPRAID_ADMIN_QUEUE_NUM) { + dev_warn(hdev->dev, "warn only admin queue can be used\n"); + return 0; + } + + hdev->state = SPRAID_LIVE; + + spraid_send_all_aen(hdev); + + ret = spraid_dev_list_init(hdev); + if (ret) + goto remove_bsg; + + ret = spraid_configure_timestamp(hdev); + if (ret) + dev_warn(hdev->dev, "init set timestamp failed\n"); + + ret = spraid_alloc_ioq_ptcmds(hdev); + if (ret) + goto remove_bsg; + + scsi_scan_host(hdev->shost); + + return 0; + +remove_bsg: + spraid_remove_bsg(hdev); +remove_io_queues: + spraid_remove_io_queues(hdev); +free_iod_mempool: + spraid_free_iod_ext_mem_pool(hdev); +disable_admin_q: + spraid_disable_admin_queue(hdev, false); +pci_disable: + spraid_pci_disable(hdev); +resources_free: + spraid_free_resources(hdev); +dev_unmap: + spraid_dev_unmap(hdev); +put_dev: + put_device(hdev->dev); + scsi_host_put(shost); + + return -ENODEV; +} + +static void spraid_remove(struct pci_dev *pdev) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + struct Scsi_Host *shost = hdev->shost; + + dev_info(hdev->dev, "enter spraid remove\n"); + + spraid_change_host_state(hdev, SPRAID_DELETING); + flush_work(&hdev->reset_work); + + if (!pci_device_is_present(pdev)) + spraid_back_all_io(hdev); + + spraid_remove_bsg(hdev); + scsi_remove_host(shost); + spraid_free_ioq_ptcmds(hdev); + kfree(hdev->devices); + spraid_remove_io_queues(hdev); + spraid_free_iod_ext_mem_pool(hdev); + spraid_disable_admin_queue(hdev, false); + spraid_pci_disable(hdev); + spraid_free_resources(hdev); + spraid_dev_unmap(hdev); + put_device(hdev->dev); + scsi_host_put(shost); + + dev_info(hdev->dev, "exit spraid remove\n"); +} + +static const struct pci_device_id spraid_id_table[] = { + { PCI_DEVICE(PCI_VENDOR_ID_RAMAXEL_LOGIC, + SPRAID_SERVER_DEVICE_HBA_DID) }, + { PCI_DEVICE(PCI_VENDOR_ID_RAMAXEL_LOGIC, + SPRAID_SERVER_DEVICE_RAID_DID) }, + { 0, } +}; +MODULE_DEVICE_TABLE(pci, spraid_id_table); + +static struct pci_driver spraid_driver = { + .name = "spraid", + .id_table = spraid_id_table, + .probe = spraid_probe, + .remove = spraid_remove, + .shutdown = spraid_shutdown, + .err_handler = &spraid_err_handler, +}; + +static int __init spraid_init(void) +{ + int ret; + + spraid_wq = alloc_workqueue("spraid-wq", WQ_UNBOUND | WQ_MEM_RECLAIM | + WQ_SYSFS, 0); + if (!spraid_wq) + return -ENOMEM; + + spraid_class = class_create(THIS_MODULE, "spraid"); + if (IS_ERR(spraid_class)) { + ret = PTR_ERR(spraid_class); + goto destroy_wq; + } + + ret = pci_register_driver(&spraid_driver); + if (ret < 0) + goto destroy_class; + + return 0; + +destroy_class: + class_destroy(spraid_class); +destroy_wq: + destroy_workqueue(spraid_wq); + + return ret; +} + +static void __exit spraid_exit(void) +{ + pci_unregister_driver(&spraid_driver); + class_destroy(spraid_class); + destroy_workqueue(spraid_wq); + ida_destroy(&spraid_instance_ida); +} + +MODULE_AUTHOR("songyl(a)ramaxel.com"); +MODULE_DESCRIPTION("Ramaxel Memory Technology SPraid Driver"); +MODULE_LICENSE("GPL"); +MODULE_VERSION(SPRAID_DRV_VERSION); +module_init(spraid_init); +module_exit(spraid_exit); -- 2.27.0

1 0

[PATCH openEuler-1.0-LTS 1/2] USB: gadget: detect too-big endpoint 0 requests
by Yang Yingliang 22 Dec '21

22 Dec '21

From: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> mainline inclusion from mainline-v5.16-rc5 commit 153a2d7e3350cc89d406ba2d35be8793a64c2038 category: bugfix bugzilla: NA CVE: CVE-2021-39685 -------------------------------- Sometimes USB hosts can ask for buffers that are too large from endpoint 0, which should not be allowed. If this happens for OUT requests, stall the endpoint, but for IN requests, trim the request size to the endpoint buffer size. Co-developed-by: Szymon Heidrich <szymon.heidrich(a)gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> Reviewed-by: Xiu Jianfeng <xiujianfeng(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/usb/gadget/composite.c | 12 ++++++++++++ drivers/usb/gadget/legacy/dbgp.c | 13 +++++++++++++ drivers/usb/gadget/legacy/inode.c | 16 +++++++++++++++- 3 files changed, 40 insertions(+), 1 deletion(-) diff --git a/drivers/usb/gadget/composite.c b/drivers/usb/gadget/composite.c index af0702cfa92ff..3813249387b3d 100644 --- a/drivers/usb/gadget/composite.c +++ b/drivers/usb/gadget/composite.c @@ -1569,6 +1569,18 @@ composite_setup(struct usb_gadget *gadget, const struct usb_ctrlrequest *ctrl) struct usb_function *f = NULL; u8 endp; + if (w_length > USB_COMP_EP0_BUFSIZ) { + if (ctrl->bRequestType == USB_DIR_OUT) { + goto done; + } else { + /* Cast away the const, we are going to overwrite on purpose. */ + __le16 *temp = (__le16 *)&ctrl->wLength; + + *temp = cpu_to_le16(USB_COMP_EP0_BUFSIZ); + w_length = USB_COMP_EP0_BUFSIZ; + } + } + /* partial re-init of the response message; the function or the * gadget might need to intercept e.g. a control-OUT completion * when we delegate to it. diff --git a/drivers/usb/gadget/legacy/dbgp.c b/drivers/usb/gadget/legacy/dbgp.c index e1d566c9918ae..e567afcb2794c 100644 --- a/drivers/usb/gadget/legacy/dbgp.c +++ b/drivers/usb/gadget/legacy/dbgp.c @@ -345,6 +345,19 @@ static int dbgp_setup(struct usb_gadget *gadget, void *data = NULL; u16 len = 0; + if (length > DBGP_REQ_LEN) { + if (ctrl->bRequestType == USB_DIR_OUT) { + return err; + } else { + /* Cast away the const, we are going to overwrite on purpose. */ + __le16 *temp = (__le16 *)&ctrl->wLength; + + *temp = cpu_to_le16(DBGP_REQ_LEN); + length = DBGP_REQ_LEN; + } + } + + if (request == USB_REQ_GET_DESCRIPTOR) { switch (value>>8) { case USB_DT_DEVICE: diff --git a/drivers/usb/gadget/legacy/inode.c b/drivers/usb/gadget/legacy/inode.c index 37ca0e669bd85..83fd1b55d497a 100644 --- a/drivers/usb/gadget/legacy/inode.c +++ b/drivers/usb/gadget/legacy/inode.c @@ -109,6 +109,8 @@ enum ep0_state { /* enough for the whole queue: most events invalidate others */ #define N_EVENT 5 +#define RBUF_SIZE 256 + struct dev_data { spinlock_t lock; refcount_t count; @@ -143,7 +145,7 @@ struct dev_data { struct dentry *dentry; /* except this scratch i/o buffer for ep0 */ - u8 rbuf [256]; + u8 rbuf[RBUF_SIZE]; }; static inline void get_dev (struct dev_data *data) @@ -1332,6 +1334,18 @@ gadgetfs_setup (struct usb_gadget *gadget, const struct usb_ctrlrequest *ctrl) u16 w_value = le16_to_cpu(ctrl->wValue); u16 w_length = le16_to_cpu(ctrl->wLength); + if (w_length > RBUF_SIZE) { + if (ctrl->bRequestType == USB_DIR_OUT) { + return value; + } else { + /* Cast away the const, we are going to overwrite on purpose. */ + __le16 *temp = (__le16 *)&ctrl->wLength; + + *temp = cpu_to_le16(RBUF_SIZE); + w_length = RBUF_SIZE; + } + } + spin_lock (&dev->lock); dev->setup_abort = 0; if (dev->state == STATE_DEV_UNCONNECTED) { -- 2.25.1

1 1

[PATCH openEuler-1.0-LTS] scsi:spraid: support Ramaxel's spraid driver
by Yanling Song 21 Dec '21

21 Dec '21

Ramaxel inclusion category: features bugzilla: https://gitee.com/openeuler/kernel/issues/I4NJIB CVE: NA Support Ramaxel's SPRxxx Raid controller Signed-off-by: Yanling Song <songyl(a)ramaxel.com> Reviewed-by: Jiang Yu<yujiang(a)ramaxel.com> --- drivers/scsi/Kconfig | 1 + drivers/scsi/Makefile | 1 + drivers/scsi/spraid/Kconfig | 13 + drivers/scsi/spraid/Makefile | 7 + drivers/scsi/spraid/spraid.h | 746 ++++++ drivers/scsi/spraid/spraid_main.c | 3875 +++++++++++++++++++++++++++++ 6 files changed, 4643 insertions(+) create mode 100644 drivers/scsi/spraid/Kconfig create mode 100644 drivers/scsi/spraid/Makefile create mode 100644 drivers/scsi/spraid/spraid.h create mode 100644 drivers/scsi/spraid/spraid_main.c diff --git a/drivers/scsi/Kconfig b/drivers/scsi/Kconfig index 8450484184e3..63d2aaa22834 100644 --- a/drivers/scsi/Kconfig +++ b/drivers/scsi/Kconfig @@ -522,6 +522,7 @@ source "drivers/scsi/megaraid/Kconfig.megaraid" source "drivers/scsi/mpt3sas/Kconfig" source "drivers/scsi/smartpqi/Kconfig" source "drivers/scsi/ufs/Kconfig" +source "drivers/scsi/spraid/Kconfig" config SCSI_HPTIOP tristate "HighPoint RocketRAID 3xxx/4xxx Controller support" diff --git a/drivers/scsi/Makefile b/drivers/scsi/Makefile index 2973693f6dcc..4056cf26e09e 100644 --- a/drivers/scsi/Makefile +++ b/drivers/scsi/Makefile @@ -94,6 +94,7 @@ obj-$(CONFIG_SCSI_ZALON) += zalon7xx.o obj-$(CONFIG_SCSI_DC395x) += dc395x.o obj-$(CONFIG_SCSI_AM53C974) += esp_scsi.o am53c974.o obj-$(CONFIG_CXLFLASH) += cxlflash/ +obj-$(CONFIG_RAMAXEL_SPRAID) += spraid/ obj-$(CONFIG_MEGARAID_LEGACY) += megaraid.o obj-$(CONFIG_MEGARAID_NEWGEN) += megaraid/ obj-$(CONFIG_MEGARAID_SAS) += megaraid/ diff --git a/drivers/scsi/spraid/Kconfig b/drivers/scsi/spraid/Kconfig new file mode 100644 index 000000000000..bfbba3db8db0 --- /dev/null +++ b/drivers/scsi/spraid/Kconfig @@ -0,0 +1,13 @@ +# +# Ramaxel driver configuration +# + +config RAMAXEL_SPRAID + tristate "Ramaxel spraid Adapter" + depends on PCI && SCSI + select BLK_DEV_BSGLIB + depends on ARM64 || X86_64 + help + This driver supports Ramaxel SPRxxx serial + raid controller, which has PCIE Gen4 interface + with host and supports SAS/SATA Hdd/ssd. diff --git a/drivers/scsi/spraid/Makefile b/drivers/scsi/spraid/Makefile new file mode 100644 index 000000000000..aadc2ffd37eb --- /dev/null +++ b/drivers/scsi/spraid/Makefile @@ -0,0 +1,7 @@ +# +# Makefile for the Ramaxel device drivers. +# + +obj-$(CONFIG_RAMAXEL_SPRAID) += spraid.o + +spraid-objs := spraid_main.o \ No newline at end of file diff --git a/drivers/scsi/spraid/spraid.h b/drivers/scsi/spraid/spraid.h new file mode 100644 index 000000000000..c1e4980e18e5 --- /dev/null +++ b/drivers/scsi/spraid/spraid.h @@ -0,0 +1,746 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* Copyright(c) 2021 Ramaxel Memory Technology, Ltd */ + +#ifndef __SPRAID_H_ +#define __SPRAID_H_ + +#define SPRAID_CAP_MQES(cap) ((cap) & 0xffff) +#define SPRAID_CAP_STRIDE(cap) (((cap) >> 32) & 0xf) +#define SPRAID_CAP_MPSMIN(cap) (((cap) >> 48) & 0xf) +#define SPRAID_CAP_MPSMAX(cap) (((cap) >> 52) & 0xf) +#define SPRAID_CAP_TIMEOUT(cap) (((cap) >> 24) & 0xff) +#define SPRAID_CAP_DMAMASK(cap) (((cap) >> 37) & 0xff) + +#define SPRAID_DEFAULT_MAX_CHANNEL 4 +#define SPRAID_DEFAULT_MAX_ID 240 +#define SPRAID_DEFAULT_MAX_LUN_PER_HOST 8 +#define MAX_SECTORS 2048 + +#define IO_SQE_SIZE sizeof(struct spraid_ioq_command) +#define ADMIN_SQE_SIZE sizeof(struct spraid_admin_command) +#define SQE_SIZE(qid) (((qid) > 0) ? IO_SQE_SIZE : ADMIN_SQE_SIZE) +#define CQ_SIZE(depth) ((depth) * sizeof(struct spraid_completion)) +#define SQ_SIZE(qid, depth) ((depth) * SQE_SIZE(qid)) + +#define SENSE_SIZE(depth) ((depth) * SCSI_SENSE_BUFFERSIZE) + +#define SPRAID_AQ_DEPTH 128 +#define SPRAID_NR_AEN_COMMANDS 16 +#define SPRAID_AQ_BLK_MQ_DEPTH (SPRAID_AQ_DEPTH - SPRAID_NR_AEN_COMMANDS) +#define SPRAID_AQ_MQ_TAG_DEPTH (SPRAID_AQ_BLK_MQ_DEPTH - 1) + +#define SPRAID_ADMIN_QUEUE_NUM 1 +#define SPRAID_PTCMDS_PERQ 1 +#define SPRAID_IO_BLK_MQ_DEPTH (hdev->shost->can_queue) +#define SPRAID_NR_IOQ_PTCMDS (SPRAID_PTCMDS_PERQ * hdev->shost->nr_hw_queues) + +#define FUA_MASK 0x08 +#define SPRAID_MINORS BIT(MINORBITS) + +#define COMMAND_IS_WRITE(cmd) ((cmd)->common.opcode & 1) + +#define SPRAID_IO_IOSQES 7 +#define SPRAID_IO_IOCQES 4 +#define PRP_ENTRY_SIZE 8 + +#define SMALL_POOL_SIZE 256 +#define MAX_SMALL_POOL_NUM 16 +#define MAX_CMD_PER_DEV 64 +#define MAX_CDB_LEN 32 + +#define SPRAID_UP_TO_MULTY4(x) (((x) + 4) & (~0x03)) + +#define CQE_STATUS_SUCCESS (0x0) + +#define PCI_VENDOR_ID_RAMAXEL_LOGIC 0x1E81 + +#define SPRAID_SERVER_DEVICE_HBA_DID 0x2100 +#define SPRAID_SERVER_DEVICE_RAID_DID 0x2200 + +#define IO_6_DEFAULT_TX_LEN 256 + +#define SPRAID_INT_PAGES 2 +#define SPRAID_INT_BYTES(hdev) (SPRAID_INT_PAGES * (hdev)->page_size) + +enum { + SPRAID_REQ_CANCELLED = (1 << 0), + SPRAID_REQ_USERCMD = (1 << 1), +}; + +enum { + SPRAID_SC_SUCCESS = 0x0, + SPRAID_SC_INVALID_OPCODE = 0x1, + SPRAID_SC_INVALID_FIELD = 0x2, + + SPRAID_SC_ABORT_LIMIT = 0x103, + SPRAID_SC_ABORT_MISSING = 0x104, + SPRAID_SC_ASYNC_LIMIT = 0x105, + + SPRAID_SC_DNR = 0x4000, +}; + +enum { + SPRAID_REG_CAP = 0x0000, + SPRAID_REG_CC = 0x0014, + SPRAID_REG_CSTS = 0x001c, + SPRAID_REG_AQA = 0x0024, + SPRAID_REG_ASQ = 0x0028, + SPRAID_REG_ACQ = 0x0030, + SPRAID_REG_DBS = 0x1000, +}; + +enum { + SPRAID_CC_ENABLE = 1 << 0, + SPRAID_CC_CSS_NVM = 0 << 4, + SPRAID_CC_MPS_SHIFT = 7, + SPRAID_CC_AMS_SHIFT = 11, + SPRAID_CC_SHN_SHIFT = 14, + SPRAID_CC_IOSQES_SHIFT = 16, + SPRAID_CC_IOCQES_SHIFT = 20, + SPRAID_CC_AMS_RR = 0 << SPRAID_CC_AMS_SHIFT, + SPRAID_CC_SHN_NONE = 0 << SPRAID_CC_SHN_SHIFT, + SPRAID_CC_IOSQES = SPRAID_IO_IOSQES << SPRAID_CC_IOSQES_SHIFT, + SPRAID_CC_IOCQES = SPRAID_IO_IOCQES << SPRAID_CC_IOCQES_SHIFT, + SPRAID_CC_SHN_NORMAL = 1 << SPRAID_CC_SHN_SHIFT, + SPRAID_CC_SHN_MASK = 3 << SPRAID_CC_SHN_SHIFT, + SPRAID_CSTS_CFS_SHIFT = 1, + SPRAID_CSTS_SHST_SHIFT = 2, + SPRAID_CSTS_PP_SHIFT = 5, + SPRAID_CSTS_RDY = 1 << 0, + SPRAID_CSTS_SHST_CMPLT = 2 << 2, + SPRAID_CSTS_SHST_MASK = 3 << 2, + SPRAID_CSTS_CFS_MASK = 1 << SPRAID_CSTS_CFS_SHIFT, + SPRAID_CSTS_PP_MASK = 1 << SPRAID_CSTS_PP_SHIFT, +}; + +enum { + SPRAID_ADMIN_DELETE_SQ = 0x00, + SPRAID_ADMIN_CREATE_SQ = 0x01, + SPRAID_ADMIN_DELETE_CQ = 0x04, + SPRAID_ADMIN_CREATE_CQ = 0x05, + SPRAID_ADMIN_ABORT_CMD = 0x08, + SPRAID_ADMIN_SET_FEATURES = 0x09, + SPRAID_ADMIN_ASYNC_EVENT = 0x0c, + SPRAID_ADMIN_GET_INFO = 0xc6, + SPRAID_ADMIN_RESET = 0xc8, +}; + +enum { + SPRAID_GET_INFO_CTRL = 0, + SPRAID_GET_INFO_DEV_LIST = 1, +}; + +enum { + SPRAID_RESET_TARGET = 0, + SPRAID_RESET_BUS = 1, +}; + +enum { + SPRAID_AEN_ERROR = 0, + SPRAID_AEN_NOTICE = 2, + SPRAID_AEN_VS = 7, +}; + +enum { + SPRAID_AEN_DEV_CHANGED = 0x00, + SPRAID_AEN_FW_ACT_START = 0x01, + SPRAID_AEN_HOST_PROBING = 0x10, +}; + +enum { + SPRAID_AEN_TIMESYN = 0x00, + SPRAID_AEN_FW_ACT_FINISH = 0x02, + SPRAID_AEN_EVENT_MIN = 0x80, + SPRAID_AEN_EVENT_MAX = 0xff, +}; + +enum { + SPRAID_CMD_WRITE = 0x01, + SPRAID_CMD_READ = 0x02, + + SPRAID_CMD_NONIO_NONE = 0x80, + SPRAID_CMD_NONIO_TODEV = 0x81, + SPRAID_CMD_NONIO_FROMDEV = 0x82, +}; + +enum { + SPRAID_QUEUE_PHYS_CONTIG = (1 << 0), + SPRAID_CQ_IRQ_ENABLED = (1 << 1), + + SPRAID_FEAT_NUM_QUEUES = 0x07, + SPRAID_FEAT_ASYNC_EVENT = 0x0b, + SPRAID_FEAT_TIMESTAMP = 0x0e, +}; + +enum spraid_state { + SPRAID_NEW, + SPRAID_LIVE, + SPRAID_RESETTING, + SPRAID_DELETING, + SPRAID_DEAD, +}; + +enum { + SPRAID_CARD_HBA, + SPRAID_CARD_RAID, +}; + +enum spraid_cmd_type { + SPRAID_CMD_ADM, + SPRAID_CMD_IOPT, +}; + +struct spraid_completion { + __le32 result; + union { + struct { + __u8 sense_len; + __u8 resv[3]; + }; + __le32 result1; + }; + __le16 sq_head; + __le16 sq_id; + __u16 cmd_id; + __le16 status; +}; + +struct spraid_ctrl_info { + __le32 nd; + __le16 max_cmds; + __le16 max_channel; + __le32 max_tgt_id; + __le16 max_lun; + __le16 max_num_sge; + __le16 lun_num_in_boot; + __u8 mdts; + __u8 acl; + __u8 aerl; + __u8 card_type; + __u16 rsvd; + __u32 rtd3e; + __u8 sn[32]; + __u8 fr[16]; + __u8 rsvd1[4020]; +}; + +struct spraid_dev { + struct pci_dev *pdev; + struct device *dev; + struct Scsi_Host *shost; + struct spraid_queue *queues; + struct dma_pool *prp_page_pool; + struct dma_pool *prp_small_pool[MAX_SMALL_POOL_NUM]; + mempool_t *iod_mempool; + void __iomem *bar; + u32 max_qid; + u32 num_vecs; + u32 queue_count; + u32 ioq_depth; + int db_stride; + u32 __iomem *dbs; + struct rw_semaphore devices_rwsem; + int numa_node; + u32 page_size; + u32 ctrl_config; + u32 online_queues; + u64 cap; + int instance; + struct spraid_ctrl_info *ctrl_info; + struct spraid_dev_info *devices; + + struct spraid_cmd *adm_cmds; + struct list_head adm_cmd_list; + spinlock_t adm_cmd_lock; + + struct spraid_cmd *ioq_ptcmds; + struct list_head ioq_pt_list; + spinlock_t ioq_pt_lock; + + struct work_struct scan_work; + struct work_struct timesyn_work; + struct work_struct reset_work; + struct work_struct fw_act_work; + + enum spraid_state state; + spinlock_t state_lock; + + struct request_queue *bsg_queue; +}; + +struct spraid_sgl_desc { + __le64 addr; + __le32 length; + __u8 rsvd[3]; + __u8 type; +}; + +union spraid_data_ptr { + struct { + __le64 prp1; + __le64 prp2; + }; + struct spraid_sgl_desc sgl; +}; + +struct spraid_admin_common_command { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __le32 cdw2[4]; + union spraid_data_ptr dptr; + __le32 cdw10; + __le32 cdw11; + __le32 cdw12; + __le32 cdw13; + __le32 cdw14; + __le32 cdw15; +}; + +struct spraid_features { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __u64 rsvd2[2]; + union spraid_data_ptr dptr; + __le32 fid; + __le32 dword11; + __le32 dword12; + __le32 dword13; + __le32 dword14; + __le32 dword15; +}; + +struct spraid_create_cq { + __u8 opcode; + __u8 flags; + __u16 command_id; + __u32 rsvd1[5]; + __le64 prp1; + __u64 rsvd8; + __le16 cqid; + __le16 qsize; + __le16 cq_flags; + __le16 irq_vector; + __u32 rsvd12[4]; +}; + +struct spraid_create_sq { + __u8 opcode; + __u8 flags; + __u16 command_id; + __u32 rsvd1[5]; + __le64 prp1; + __u64 rsvd8; + __le16 sqid; + __le16 qsize; + __le16 sq_flags; + __le16 cqid; + __u32 rsvd12[4]; +}; + +struct spraid_delete_queue { + __u8 opcode; + __u8 flags; + __u16 command_id; + __u32 rsvd1[9]; + __le16 qid; + __u16 rsvd10; + __u32 rsvd11[5]; +}; + +struct spraid_get_info { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __u32 rsvd2[4]; + union spraid_data_ptr dptr; + __u8 type; + __u8 rsvd10[3]; + __le32 cdw11; + __u32 rsvd12[4]; +}; + +struct spraid_usr_cmd { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + union { + struct { + __le16 subopcode; + __le16 rsvd1; + } info_0; + __le32 cdw2; + }; + union { + struct { + __le16 data_len; + __le16 param_len; + } info_1; + __le32 cdw3; + }; + __u64 metadata; + union spraid_data_ptr dptr; + __le32 cdw10; + __le32 cdw11; + __le32 cdw12; + __le32 cdw13; + __le32 cdw14; + __le32 cdw15; +}; + +enum { + SPRAID_CMD_FLAG_SGL_METABUF = (1 << 6), + SPRAID_CMD_FLAG_SGL_METASEG = (1 << 7), + SPRAID_CMD_FLAG_SGL_ALL = SPRAID_CMD_FLAG_SGL_METABUF | + SPRAID_CMD_FLAG_SGL_METASEG, +}; + +enum spraid_cmd_state { + SPRAID_CMD_IDLE = 0, + SPRAID_CMD_IN_FLIGHT = 1, + SPRAID_CMD_COMPLETE = 2, + SPRAID_CMD_TIMEOUT = 3, + SPRAID_CMD_TMO_COMPLETE = 4, +}; + +struct spraid_abort_cmd { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __u64 rsvd2[4]; + __le16 sqid; + __le16 cid; + __u32 rsvd11[5]; +}; + +struct spraid_reset_cmd { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __u64 rsvd2[4]; + __u8 type; + __u8 rsvd10[3]; + __u32 rsvd11[5]; +}; + +struct spraid_admin_command { + union { + struct spraid_admin_common_command common; + struct spraid_features features; + struct spraid_create_cq create_cq; + struct spraid_create_sq create_sq; + struct spraid_delete_queue delete_queue; + struct spraid_get_info get_info; + struct spraid_abort_cmd abort; + struct spraid_reset_cmd reset; + struct spraid_usr_cmd usr_cmd; + }; +}; + +struct spraid_ioq_common_command { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __le16 sense_len; + __u8 cdb_len; + __u8 rsvd2; + __le32 cdw3[3]; + union spraid_data_ptr dptr; + __le32 cdw10[6]; + __u8 cdb[32]; + __le64 sense_addr; + __le32 cdw26[6]; +}; + +struct spraid_rw_command { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __le16 sense_len; + __u8 cdb_len; + __u8 rsvd2; + __u32 rsvd3[3]; + union spraid_data_ptr dptr; + __le64 slba; + __le16 nlb; + __le16 control; + __u32 rsvd13[3]; + __u8 cdb[32]; + __le64 sense_addr; + __u32 rsvd26[6]; +}; + +struct spraid_scsi_nonio { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + __le16 sense_len; + __u8 cdb_length; + __u8 rsvd2; + __u32 rsvd3[3]; + union spraid_data_ptr dptr; + __u32 rsvd10[5]; + __le32 buffer_len; + __u8 cdb[32]; + __le64 sense_addr; + __u32 rsvd26[6]; +}; + +struct spraid_ioq_command { + union { + struct spraid_ioq_common_command common; + struct spraid_rw_command rw; + struct spraid_scsi_nonio scsi_nonio; + }; +}; + +struct spraid_passthru_common_cmd { + __u8 opcode; + __u8 flags; + __u16 rsvd0; + __u32 nsid; + union { + struct { + __u16 subopcode; + __u16 rsvd1; + } info_0; + __u32 cdw2; + }; + union { + struct { + __u16 data_len; + __u16 param_len; + } info_1; + __u32 cdw3; + }; + __u64 metadata; + + __u64 addr; + __u64 prp2; + + __u32 cdw10; + __u32 cdw11; + __u32 cdw12; + __u32 cdw13; + __u32 cdw14; + __u32 cdw15; + __u32 timeout_ms; + __u32 result0; + __u32 result1; +}; + +struct spraid_ioq_passthru_cmd { + __u8 opcode; + __u8 flags; + __u16 rsvd0; + __u32 nsid; + union { + struct { + __u16 res_sense_len; + __u8 cdb_len; + __u8 rsvd0; + } info_0; + __u32 cdw2; + }; + union { + struct { + __u16 subopcode; + __u16 rsvd1; + } info_1; + __u32 cdw3; + }; + union { + struct { + __u16 rsvd; + __u16 param_len; + } info_2; + __u32 cdw4; + }; + __u32 cdw5; + __u64 addr; + __u64 prp2; + union { + struct { + __u16 eid; + __u16 sid; + } info_3; + __u32 cdw10; + }; + union { + struct { + __u16 did; + __u8 did_flag; + __u8 rsvd2; + } info_4; + __u32 cdw11; + }; + __u32 cdw12; + __u32 cdw13; + __u32 cdw14; + __u32 data_len; + __u32 cdw16; + __u32 cdw17; + __u32 cdw18; + __u32 cdw19; + __u32 cdw20; + __u32 cdw21; + __u32 cdw22; + __u32 cdw23; + __u64 sense_addr; + __u32 cdw26[4]; + __u32 timeout_ms; + __u32 result0; + __u32 result1; +}; + +struct spraid_bsg_request { + u32 msgcode; + u32 control; + union { + struct spraid_passthru_common_cmd admcmd; + struct spraid_ioq_passthru_cmd ioqcmd; + }; +}; + +enum { + SPRAID_BSG_ADM, + SPRAID_BSG_IOQ, +}; + +struct spraid_cmd { + int qid; + int cid; + u32 result0; + u32 result1; + u16 status; + void *priv; + enum spraid_cmd_state state; + struct completion cmd_done; + struct list_head list; +}; + +struct spraid_queue { + struct spraid_dev *hdev; + spinlock_t sq_lock; + + spinlock_t cq_lock ____cacheline_aligned_in_smp; + + void *sq_cmds; + + struct spraid_completion *cqes; + + dma_addr_t sq_dma_addr; + dma_addr_t cq_dma_addr; + u32 __iomem *q_db; + u8 cq_phase; + u8 sqes; + u16 qid; + u16 sq_tail; + u16 cq_head; + u16 last_cq_head; + u16 q_depth; + s16 cq_vector; + void *sense; + dma_addr_t sense_dma_addr; + struct dma_pool *prp_small_pool; +}; + +struct spraid_iod { + struct spraid_queue *spraidq; + enum spraid_cmd_state state; + int npages; + u32 nsge; + u32 length; + bool use_sgl; + bool sg_drv_mgmt; + dma_addr_t first_dma; + void *sense; + dma_addr_t sense_dma; + struct scatterlist *sg; + struct scatterlist inline_sg[0]; +}; + +#define SPRAID_DEV_INFO_ATTR_BOOT(attr) ((attr) & 0x01) +#define SPRAID_DEV_INFO_ATTR_VD(attr) (((attr) & 0x02) == 0x0) +#define SPRAID_DEV_INFO_ATTR_PT(attr) (((attr) & 0x22) == 0x02) +#define SPRAID_DEV_INFO_ATTR_RAWDISK(attr) ((attr) & 0x20) + +#define SPRAID_DEV_INFO_FLAG_VALID(flag) ((flag) & 0x01) +#define SPRAID_DEV_INFO_FLAG_CHANGE(flag) ((flag) & 0x02) + +#define BGTASK_TYPE_REBUILD 4 +#define USR_CMD_READ 0xc2 +#define USR_CMD_RDLEN 0x1000 +#define USR_CMD_VDINFO 0x704 +#define USR_CMD_BGTASK 0x504 +#define VDINFO_PARAM_LEN 0x04 + +struct spraid_vd_info { + __u8 name[32]; + __le16 id; + __u8 rg_id; + __u8 rg_level; + __u8 sg_num; + __u8 sg_disk_num; + __u8 vd_status; + __u8 vd_type; + __u8 rsvd1[4056]; +}; + +#define MAX_REALTIME_BGTASK_NUM 32 + +struct bgtask_info { + __u8 type; + __u8 progress; + __u8 rate; + __u8 rsvd0; + __le16 vd_id; + __le16 time_left; + __u8 rsvd1[4]; +}; + +struct spraid_bgtask { + __u8 sw; + __u8 task_num; + __u8 rsvd[6]; + struct bgtask_info bgtask[MAX_REALTIME_BGTASK_NUM]; +}; + +struct spraid_dev_info { + __le32 hdid; + __le16 target; + __u8 channel; + __u8 lun; + __u8 attr; + __u8 flag; + __le16 max_io_kb; +}; + +#define MAX_DEV_ENTRY_PER_PAGE_4K 340 +struct spraid_dev_list { + __le32 dev_num; + __u32 rsvd0[3]; + struct spraid_dev_info devices[MAX_DEV_ENTRY_PER_PAGE_4K]; +}; + +struct spraid_sdev_hostdata { + u32 hdid; + u16 max_io_kb; + u8 attr; + u8 flag; + u8 rg_id; + u8 rsvd[3]; +}; + +#endif + diff --git a/drivers/scsi/spraid/spraid_main.c b/drivers/scsi/spraid/spraid_main.c new file mode 100644 index 000000000000..519b39f44e91 --- /dev/null +++ b/drivers/scsi/spraid/spraid_main.c @@ -0,0 +1,3875 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright(c) 2021 Ramaxel Memory Technology, Ltd */ + +/* Ramaxel Raid SPXXX Series Linux Driver */ + +#define pr_fmt(fmt) "spraid: " fmt + +#include <linux/sched/signal.h> +#include <linux/version.h> +#include <linux/pci.h> +#include <linux/aer.h> +#include <linux/module.h> +#include <linux/ioport.h> +#include <linux/device.h> +#include <linux/delay.h> +#include <linux/interrupt.h> +#include <linux/cdev.h> +#include <linux/sysfs.h> +#include <linux/gfp.h> +#include <linux/types.h> +#include <linux/ratelimit.h> +#include <linux/once.h> +#include <linux/debugfs.h> +#include <linux/io-64-nonatomic-lo-hi.h> +#include <linux/blkdev.h> +#include <linux/bsg-lib.h> +#include <asm/unaligned.h> +#include <linux/sort.h> +#include <target/target_core_backend.h> + +#include <scsi/scsi.h> +#include <scsi/scsi_cmnd.h> +#include <scsi/scsi_device.h> +#include <scsi/scsi_host.h> +#include <scsi/scsi_transport.h> +#include <scsi/scsi_dbg.h> + + +#include "spraid.h" + +static u32 admin_tmout = 60; +module_param(admin_tmout, uint, 0644); +MODULE_PARM_DESC(admin_tmout, "admin commands timeout (seconds)"); + +static u32 scmd_tmout_nonpt = 180; +module_param(scmd_tmout_nonpt, uint, 0644); +MODULE_PARM_DESC(scmd_tmout_nonpt, + "scsi commands timeout for rawdisk&raid(seconds)"); + +static int ioq_depth_set(const char *val, const struct kernel_param *kp); +static const struct kernel_param_ops ioq_depth_ops = { + .set = ioq_depth_set, + .get = param_get_uint, +}; + +static u32 io_queue_depth = 1024; +module_param_cb(io_queue_depth, &ioq_depth_ops, &io_queue_depth, 0644); +MODULE_PARM_DESC(io_queue_depth, "set io queue depth, should >= 2"); + +static int log_debug_switch_set(const char *val, const struct kernel_param *kp) +{ + u8 n = 0; + int ret; + + ret = kstrtou8(val, 10, &n); + if (ret != 0) + return -EINVAL; + + return param_set_byte(val, kp); +} + +static const struct kernel_param_ops log_debug_switch_ops = { + .set = log_debug_switch_set, + .get = param_get_byte, +}; + +static unsigned char log_debug_switch; +module_param_cb(log_debug_switch, &log_debug_switch_ops, + &log_debug_switch, 0644); +MODULE_PARM_DESC(log_debug_switch, + "set log state, default non-zero for switch on"); + +static int small_pool_num_set(const char *val, const struct kernel_param *kp) +{ + u8 n = 0; + int ret; + + ret = kstrtou8(val, 10, &n); + if (ret != 0) + return -EINVAL; + if (n > MAX_SMALL_POOL_NUM) + n = MAX_SMALL_POOL_NUM; + if (n < 1) + n = 1; + *((u8 *)kp->arg) = n; + + return 0; +} + +static const struct kernel_param_ops small_pool_num_ops = { + .set = small_pool_num_set, + .get = param_get_byte, +}; + +/* It was found that the spindlock of a single pool conflicts + * a lot with multiple CPUs.So multiple pools are introduced + * to reduce the conflictions. + */ +static unsigned char small_pool_num = 4; +module_param_cb(small_pool_num, &small_pool_num_ops, &small_pool_num, 0644); +MODULE_PARM_DESC(small_pool_num, "set prp small pool num, default 4, MAX 16"); + +static void spraid_free_queue(struct spraid_queue *spraidq); +static void spraid_handle_aen_notice(struct spraid_dev *hdev, u32 result); +static void spraid_handle_aen_vs(struct spraid_dev *hdev, + u32 result, u32 result1); + +static DEFINE_IDA(spraid_instance_ida); + +static struct class *spraid_class; + +#define SPRAID_CAP_TIMEOUT_UNIT_MS (HZ / 2) + +static struct workqueue_struct *spraid_wq; + +#define dev_log_dbg(dev, fmt, ...) do { \ + if (unlikely(log_debug_switch)) \ + dev_info(dev, "[%s] [%d] " fmt, \ + __func__, __LINE__, ##__VA_ARGS__); \ +} while (0) + +#define SPRAID_DRV_VERSION "1.0.0.0" + +#define ADMIN_TIMEOUT (admin_tmout * HZ) +#define ADMIN_ERR_TIMEOUT 32757 + +#define SPRAID_WAIT_ABNL_CMD_TIMEOUT (3 * 2) + +#define SPRAID_DMA_MSK_BIT_MAX 64 + +enum FW_STAT_CODE { + FW_STAT_OK = 0, + FW_STAT_NEED_CHECK, + FW_STAT_ERROR, + FW_STAT_EP_PCIE_ERROR, + FW_STAT_NAC_DMA_ERROR, + FW_STAT_ABORTED, + FW_STAT_NEED_RETRY +}; + +static const char * const raid_levels[] = {"0", "1", "5", "6", "10", "50", "60", + "NA"}; + +static const char * const raid_states[] = { + "NA", "NORMAL", "FAULT", "DEGRADE", "NOT_FORMATTED", + "FORMATTING", "SANITIZING", "INITIALIZING", "INITIALIZE_FAIL", + "DELETING", "DELETE_FAIL", "WRITE_PROTECT" +}; + +static int ioq_depth_set(const char *val, const struct kernel_param *kp) +{ + int n = 0; + int ret; + + ret = kstrtoint(val, 10, &n); + if (ret != 0 || n < 2) + return -EINVAL; + + return param_set_int(val, kp); +} + +static int spraid_remap_bar(struct spraid_dev *hdev, u32 size) +{ + struct pci_dev *pdev = hdev->pdev; + + if (size > pci_resource_len(pdev, 0)) { + dev_err(hdev->dev, "Input size[%u] exceed bar0 length[%llu]\n", + size, pci_resource_len(pdev, 0)); + return -ENOMEM; + } + + if (hdev->bar) + iounmap(hdev->bar); + + hdev->bar = ioremap(pci_resource_start(pdev, 0), size); + if (!hdev->bar) { + dev_err(hdev->dev, "ioremap for bar0 failed\n"); + return -ENOMEM; + } + hdev->dbs = hdev->bar + SPRAID_REG_DBS; + + return 0; +} + +static int spraid_dev_map(struct spraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + int ret; + + ret = pci_request_mem_regions(pdev, "spraid"); + if (ret) { + dev_err(hdev->dev, "fail to request memory regions\n"); + return ret; + } + + ret = spraid_remap_bar(hdev, SPRAID_REG_DBS + 4096); + if (ret) { + pci_release_mem_regions(pdev); + return ret; + } + + return 0; +} + +static void spraid_dev_unmap(struct spraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + + if (hdev->bar) { + iounmap(hdev->bar); + hdev->bar = NULL; + } + pci_release_mem_regions(pdev); +} + +static int spraid_pci_enable(struct spraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + int ret = -ENOMEM; + u64 maskbit = SPRAID_DMA_MSK_BIT_MAX; + + if (pci_enable_device_mem(pdev)) { + dev_err(hdev->dev, + "Enable pci device memory resources failed\n"); + return ret; + } + pci_set_master(pdev); + + if (readl(hdev->bar + SPRAID_REG_CSTS) == U32_MAX) { + ret = -ENODEV; + dev_err(hdev->dev, "Read csts register failed\n"); + goto disable; + } + + ret = pci_alloc_irq_vectors(pdev, 1, 1, PCI_IRQ_ALL_TYPES); + if (ret < 0) { + dev_err(hdev->dev, + "Allocate one IRQ for setup admin channel failed\n"); + goto disable; + } + + hdev->cap = lo_hi_readq(hdev->bar + SPRAID_REG_CAP); + hdev->ioq_depth = min_t(u32, SPRAID_CAP_MQES(hdev->cap) + 1, + io_queue_depth); + hdev->db_stride = 1 << SPRAID_CAP_STRIDE(hdev->cap); + + maskbit = SPRAID_CAP_DMAMASK(hdev->cap); + if (maskbit < 32 || maskbit > SPRAID_DMA_MSK_BIT_MAX) { + dev_err(hdev->dev, + "err, dma mask invalid[%llu], set to default\n", + maskbit); + maskbit = SPRAID_DMA_MSK_BIT_MAX; + } + if (dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(maskbit))) { + dev_err(hdev->dev, "set dma mask and coherent failed\n"); + goto disable; + } + + dev_info(hdev->dev, "set dma mask[%llu] success\n", maskbit); + + pci_enable_pcie_error_reporting(pdev); + pci_save_state(pdev); + + return 0; + +disable: + pci_disable_device(pdev); + return ret; +} + +static int spraid_npages_prp(u32 size, struct spraid_dev *hdev) +{ + u32 nprps = DIV_ROUND_UP(size + hdev->page_size, hdev->page_size); + + return DIV_ROUND_UP(PRP_ENTRY_SIZE * nprps, PAGE_SIZE - PRP_ENTRY_SIZE); +} + +static int spraid_npages_sgl(u32 nseg) +{ + return DIV_ROUND_UP(nseg * sizeof(struct spraid_sgl_desc), PAGE_SIZE); +} + +static void **spraid_iod_list(struct spraid_iod *iod) +{ + return (void **)(iod->inline_sg + (iod->sg_drv_mgmt ? iod->nsge : 0)); +} + +static u32 spraid_iod_ext_size(struct spraid_dev *hdev, u32 size, u32 nsge, + bool sg_drv_mgmt, bool use_sgl) +{ + size_t alloc_size, sg_size; + + if (use_sgl) + alloc_size = sizeof(__le64 *) * spraid_npages_sgl(nsge); + else + alloc_size = sizeof(__le64 *) * spraid_npages_prp(size, hdev); + + sg_size = sg_drv_mgmt ? (sizeof(struct scatterlist) * nsge) : 0; + return sg_size + alloc_size; +} + +static u32 spraid_cmd_size(struct spraid_dev *hdev, + bool sg_drv_mgmt, bool use_sgl) +{ + u32 alloc_size = spraid_iod_ext_size(hdev, SPRAID_INT_BYTES(hdev), + SPRAID_INT_PAGES, sg_drv_mgmt, use_sgl); + + dev_info(hdev->dev, "sg_drv_mgmt: %s, use_sgl: %s, iod size: %lu;" + " alloc_size: %u\n", sg_drv_mgmt ? "true" : "false", + use_sgl ? "true" : "false", + sizeof(struct spraid_iod), alloc_size); + + return sizeof(struct spraid_iod) + alloc_size; +} + +static int spraid_setup_prps(struct spraid_dev *hdev, struct spraid_iod *iod) +{ + struct scatterlist *sg = iod->sg; + u64 dma_addr = sg_dma_address(sg); + int dma_len = sg_dma_len(sg); + __le64 *prp_list, *old_prp_list; + u32 page_size = hdev->page_size; + int offset = dma_addr & (page_size - 1); + void **list = spraid_iod_list(iod); + int length = iod->length; + struct dma_pool *pool; + dma_addr_t prp_dma; + int nprps, i; + + length -= (page_size - offset); + if (length <= 0) { + iod->first_dma = 0; + return 0; + } + + dma_len -= (page_size - offset); + if (dma_len) { + dma_addr += (page_size - offset); + } else { + sg = sg_next(sg); + dma_addr = sg_dma_address(sg); + dma_len = sg_dma_len(sg); + } + + if (length <= page_size) { + iod->first_dma = dma_addr; + return 0; + } + + nprps = DIV_ROUND_UP(length, page_size); + if (nprps <= (SMALL_POOL_SIZE / PRP_ENTRY_SIZE)) { + pool = iod->spraidq->prp_small_pool; + iod->npages = 0; + } else { + pool = hdev->prp_page_pool; + iod->npages = 1; + } + + prp_list = dma_pool_alloc(pool, GFP_ATOMIC, &prp_dma); + if (!prp_list) { + dev_err_ratelimited(hdev->dev, + "Allocate first prp_list memory failed\n"); + iod->first_dma = dma_addr; + iod->npages = -1; + return -ENOMEM; + } + list[0] = prp_list; + iod->first_dma = prp_dma; + i = 0; + for (;;) { + if (i == page_size / PRP_ENTRY_SIZE) { + old_prp_list = prp_list; + + prp_list = dma_pool_alloc(pool, GFP_ATOMIC, &prp_dma); + if (!prp_list) { + dev_err_ratelimited(hdev->dev, "Allocate %dth;" + " prp_list memory failed\n", + iod->npages + 1); + return -ENOMEM; + } + list[iod->npages++] = prp_list; + prp_list[0] = old_prp_list[i - 1]; + old_prp_list[i - 1] = cpu_to_le64(prp_dma); + i = 1; + } + prp_list[i++] = cpu_to_le64(dma_addr); + dma_len -= page_size; + dma_addr += page_size; + length -= page_size; + if (length <= 0) + break; + if (dma_len > 0) + continue; + if (unlikely(dma_len < 0)) + goto bad_sgl; + sg = sg_next(sg); + dma_addr = sg_dma_address(sg); + dma_len = sg_dma_len(sg); + } + + return 0; + +bad_sgl: + dev_err(hdev->dev, + "Setup prps, invalid SGL for payload: %d nents: %d\n", + iod->length, iod->nsge); + return -EIO; +} + +#define SGES_PER_PAGE (PAGE_SIZE / sizeof(struct spraid_sgl_desc)) + +static void spraid_submit_cmd(struct spraid_queue *spraidq, const void *cmd) +{ + u32 sqes = SQE_SIZE(spraidq->qid); + unsigned long flags; + struct spraid_admin_common_command *acd = + (struct spraid_admin_common_command *)cmd; + + spin_lock_irqsave(&spraidq->sq_lock, flags); + memcpy((spraidq->sq_cmds + sqes * spraidq->sq_tail), cmd, sqes); + if (++spraidq->sq_tail == spraidq->q_depth) + spraidq->sq_tail = 0; + + writel(spraidq->sq_tail, spraidq->q_db); + spin_unlock_irqrestore(&spraidq->sq_lock, flags); + + dev_log_dbg(spraidq->hdev->dev, + "cid[%d] qid[%d], opcode[0x%x], flags[0x%x], hdid[%u]\n", + acd->command_id, spraidq->qid, acd->opcode, + acd->flags, le32_to_cpu(acd->hdid)); +} + +static u32 spraid_mod64(u64 dividend, u32 divisor) +{ + u64 d; + u32 remainder; + + if (!divisor) + pr_err("DIVISOR is zero, in div fn\n"); + + d = dividend; + remainder = do_div(d, divisor); + return remainder; +} + +static inline bool spraid_is_rw_scmd(struct scsi_cmnd *scmd) +{ + switch (scmd->cmnd[0]) { + case READ_6: + case READ_10: + case READ_12: + case READ_16: + case READ_32: + case WRITE_6: + case WRITE_10: + case WRITE_12: + case WRITE_16: + case WRITE_32: + return true; + default: + return false; + } +} + +static bool spraid_is_prp(struct spraid_dev *hdev, + struct scsi_cmnd *scmd, u32 nsge) +{ + struct scatterlist *sg = scsi_sglist(scmd); + u32 page_size = hdev->page_size; + bool is_prp = true; + int i = 0; + + scsi_for_each_sg(scmd, sg, nsge, i) { + if (i != 0 && i != nsge - 1) { + if (spraid_mod64(sg_dma_len(sg), page_size) || + spraid_mod64(sg_dma_address(sg), page_size)) { + is_prp = false; + break; + } + } + + if (nsge > 1 && i == 0) { + if ((spraid_mod64((sg_dma_address(sg) + sg_dma_len(sg)), + page_size))) { + is_prp = false; + break; + } + } + + if (nsge > 1 && i == (nsge - 1)) { + if (spraid_mod64(sg_dma_address(sg), page_size)) { + is_prp = false; + break; + } + } + } + + return is_prp; +} + +enum { + SPRAID_SGL_FMT_DATA_DESC = 0x00, + SPRAID_SGL_FMT_SEG_DESC = 0x02, + SPRAID_SGL_FMT_LAST_SEG_DESC = 0x03, + SPRAID_KEY_SGL_FMT_DATA_DESC = 0x04, + SPRAID_TRANSPORT_SGL_DATA_DESC = 0x05 +}; + +static void spraid_sgl_set_data(struct spraid_sgl_desc *sge, + struct scatterlist *sg) +{ + sge->addr = cpu_to_le64(sg_dma_address(sg)); + sge->length = cpu_to_le32(sg_dma_len(sg)); + sge->type = SPRAID_SGL_FMT_DATA_DESC << 4; +} + +static void spraid_sgl_set_seg(struct spraid_sgl_desc *sge, + dma_addr_t dma_addr, int entries) +{ + sge->addr = cpu_to_le64(dma_addr); + if (entries <= SGES_PER_PAGE) { + sge->length = cpu_to_le32(entries * sizeof(*sge)); + sge->type = SPRAID_SGL_FMT_LAST_SEG_DESC << 4; + } else { + sge->length = cpu_to_le32(PAGE_SIZE); + sge->type = SPRAID_SGL_FMT_SEG_DESC << 4; + } +} + +static int spraid_setup_ioq_cmd_sgl(struct spraid_dev *hdev, + struct scsi_cmnd *scmd, + struct spraid_ioq_command *ioq_cmd, + struct spraid_iod *iod) +{ + struct spraid_sgl_desc *sg_list, *link, *old_sg_list; + struct scatterlist *sg = scsi_sglist(scmd); + void **list = spraid_iod_list(iod); + struct dma_pool *pool; + int nsge = iod->nsge; + dma_addr_t sgl_dma; + int i = 0; + + ioq_cmd->common.flags |= SPRAID_CMD_FLAG_SGL_METABUF; + + if (nsge == 1) { + spraid_sgl_set_data(&ioq_cmd->common.dptr.sgl, sg); + return 0; + } + + if (nsge <= (SMALL_POOL_SIZE / sizeof(struct spraid_sgl_desc))) { + pool = iod->spraidq->prp_small_pool; + iod->npages = 0; + } else { + pool = hdev->prp_page_pool; + iod->npages = 1; + } + + sg_list = dma_pool_alloc(pool, GFP_ATOMIC, &sgl_dma); + if (!sg_list) { + dev_err_ratelimited(hdev->dev, + "Allocate first sgl_list failed\n"); + iod->npages = -1; + return -ENOMEM; + } + + list[0] = sg_list; + iod->first_dma = sgl_dma; + spraid_sgl_set_seg(&ioq_cmd->common.dptr.sgl, sgl_dma, nsge); + do { + if (i == SGES_PER_PAGE) { + old_sg_list = sg_list; + link = &old_sg_list[SGES_PER_PAGE - 1]; + + sg_list = dma_pool_alloc(pool, GFP_ATOMIC, &sgl_dma); + if (!sg_list) { + dev_err_ratelimited(hdev->dev, + "Allocate %dth sgl_list;" + " failed\n", + iod->npages + 1); + return -ENOMEM; + } + list[iod->npages++] = sg_list; + + i = 0; + memcpy(&sg_list[i++], link, sizeof(*link)); + spraid_sgl_set_seg(link, sgl_dma, nsge); + } + + spraid_sgl_set_data(&sg_list[i++], sg); + sg = sg_next(sg); + } while (--nsge > 0); + + return 0; +} + +#define SPRAID_RW_FUA BIT(14) + +static void spraid_setup_rw_cmd(struct spraid_dev *hdev, + struct spraid_rw_command *rw, + struct scsi_cmnd *scmd) +{ + u32 start_lba_lo, start_lba_hi; + u32 datalength = 0; + u16 control = 0; + + start_lba_lo = 0; + start_lba_hi = 0; + + if (scmd->sc_data_direction == DMA_TO_DEVICE) { + rw->opcode = SPRAID_CMD_WRITE; + } else if (scmd->sc_data_direction == DMA_FROM_DEVICE) { + rw->opcode = SPRAID_CMD_READ; + } else { + dev_err(hdev->dev, + "Invalid IO for unsupported data direction: %d\n", + scmd->sc_data_direction); + WARN_ON(1); + } + + /* 6-byte READ(0x08) or WRITE(0x0A) cdb */ + if (scmd->cmd_len == 6) { + datalength = (u32)(scmd->cmnd[4] == 0 ? + IO_6_DEFAULT_TX_LEN : scmd->cmnd[4]); + start_lba_lo = (u32)get_unaligned_be24(&scmd->cmnd[1]); + + start_lba_lo &= 0x1FFFFF; + } + + /* 10-byte READ(0x28) or WRITE(0x2A) cdb */ + else if (scmd->cmd_len == 10) { + datalength = (u32)get_unaligned_be16(&scmd->cmnd[7]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[2]); + + if (scmd->cmnd[1] & FUA_MASK) + control |= SPRAID_RW_FUA; + } + + /* 12-byte READ(0xA8) or WRITE(0xAA) cdb */ + else if (scmd->cmd_len == 12) { + datalength = get_unaligned_be32(&scmd->cmnd[6]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[2]); + + if (scmd->cmnd[1] & FUA_MASK) + control |= SPRAID_RW_FUA; + } + /* 16-byte READ(0x88) or WRITE(0x8A) cdb */ + else if (scmd->cmd_len == 16) { + datalength = get_unaligned_be32(&scmd->cmnd[10]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[6]); + start_lba_hi = get_unaligned_be32(&scmd->cmnd[2]); + + if (scmd->cmnd[1] & FUA_MASK) + control |= SPRAID_RW_FUA; + } + /* 32-byte READ(0x88) or WRITE(0x8A) cdb */ + else if (scmd->cmd_len == 32) { + datalength = get_unaligned_be32(&scmd->cmnd[28]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[16]); + start_lba_hi = get_unaligned_be32(&scmd->cmnd[12]); + + if (scmd->cmnd[10] & FUA_MASK) + control |= SPRAID_RW_FUA; + } + + if (unlikely(datalength > U16_MAX || datalength == 0)) { + dev_err(hdev->dev, + "Invalid IO for illegal transfer data length: %u\n", + datalength); + WARN_ON(1); + } + + rw->slba = cpu_to_le64(((u64)start_lba_hi << 32) | start_lba_lo); + /* 0base for nlb */ + rw->nlb = cpu_to_le16((u16)(datalength - 1)); + rw->control = cpu_to_le16(control); +} + +static void spraid_setup_nonio_cmd(struct spraid_dev *hdev, + struct spraid_scsi_nonio *scsi_nonio, + struct scsi_cmnd *scmd) +{ + scsi_nonio->buffer_len = cpu_to_le32(scsi_bufflen(scmd)); + + switch (scmd->sc_data_direction) { + case DMA_NONE: + scsi_nonio->opcode = SPRAID_CMD_NONIO_NONE; + break; + case DMA_TO_DEVICE: + scsi_nonio->opcode = SPRAID_CMD_NONIO_TODEV; + break; + case DMA_FROM_DEVICE: + scsi_nonio->opcode = SPRAID_CMD_NONIO_FROMDEV; + break; + default: + dev_err(hdev->dev, + "Invalid IO for unsupported data direction: %d\n", + scmd->sc_data_direction); + WARN_ON(1); + } +} + +static void spraid_setup_ioq_cmd(struct spraid_dev *hdev, + struct spraid_ioq_command *ioq_cmd, + struct scsi_cmnd *scmd) +{ + memcpy(ioq_cmd->common.cdb, scmd->cmnd, scmd->cmd_len); + ioq_cmd->common.cdb_len = scmd->cmd_len; + + if (spraid_is_rw_scmd(scmd)) + spraid_setup_rw_cmd(hdev, &ioq_cmd->rw, scmd); + else + spraid_setup_nonio_cmd(hdev, &ioq_cmd->scsi_nonio, scmd); +} + +static int spraid_init_iod(struct spraid_dev *hdev, struct spraid_iod *iod, + struct spraid_ioq_command *ioq_cmd, + struct scsi_cmnd *scmd) +{ + if (unlikely(!iod->sense)) { + dev_err(hdev->dev, "Allocate sense data buffer failed\n"); + return -ENOMEM; + } + ioq_cmd->common.sense_addr = cpu_to_le64(iod->sense_dma); + ioq_cmd->common.sense_len = cpu_to_le16(SCSI_SENSE_BUFFERSIZE); + + iod->nsge = 0; + iod->npages = -1; + iod->use_sgl = 0; + iod->sg_drv_mgmt = false; + WRITE_ONCE(iod->state, SPRAID_CMD_IDLE); + + return 0; +} + +static void spraid_free_iod_res(struct spraid_dev *hdev, struct spraid_iod *iod) +{ + const int last_prp = hdev->page_size / sizeof(__le64) - 1; + dma_addr_t dma_addr, next_dma_addr; + struct spraid_sgl_desc *sg_list; + __le64 *prp_list; + void *addr; + int i; + + dma_addr = iod->first_dma; + if (iod->npages == 0) + dma_pool_free(iod->spraidq->prp_small_pool, + spraid_iod_list(iod)[0], dma_addr); + + for (i = 0; i < iod->npages; i++) { + addr = spraid_iod_list(iod)[i]; + + if (iod->use_sgl) { + sg_list = addr; + next_dma_addr = + le64_to_cpu((sg_list[SGES_PER_PAGE - 1]).addr); + } else { + prp_list = addr; + next_dma_addr = le64_to_cpu(prp_list[last_prp]); + } + + dma_pool_free(hdev->prp_page_pool, addr, dma_addr); + dma_addr = next_dma_addr; + } + + if (iod->sg_drv_mgmt && iod->sg != iod->inline_sg) { + iod->sg_drv_mgmt = false; + mempool_free(iod->sg, hdev->iod_mempool); + } + + iod->sense = NULL; + iod->npages = -1; +} + +static int spraid_io_map_data(struct spraid_dev *hdev, struct spraid_iod *iod, + struct scsi_cmnd *scmd, + struct spraid_ioq_command *ioq_cmd) +{ + int ret; + + iod->nsge = scsi_dma_map(scmd); + + /* No data to DMA, it may be scsi no-rw command */ + if (unlikely(iod->nsge == 0)) + return 0; + + iod->length = scsi_bufflen(scmd); + iod->sg = scsi_sglist(scmd); + iod->use_sgl = !spraid_is_prp(hdev, scmd, iod->nsge); + + if (iod->use_sgl) { + ret = spraid_setup_ioq_cmd_sgl(hdev, scmd, ioq_cmd, iod); + } else { + ret = spraid_setup_prps(hdev, iod); + ioq_cmd->common.dptr.prp1 = + cpu_to_le64(sg_dma_address(iod->sg)); + ioq_cmd->common.dptr.prp2 = cpu_to_le64(iod->first_dma); + } + + if (ret) + scsi_dma_unmap(scmd); + + return ret; +} + +static void spraid_map_status(struct spraid_iod *iod, struct scsi_cmnd *scmd, + struct spraid_completion *cqe) +{ + scsi_set_resid(scmd, 0); + + switch ((le16_to_cpu(cqe->status) >> 1) & 0x7f) { + case FW_STAT_OK: + set_host_byte(scmd, DID_OK); + break; + case FW_STAT_NEED_CHECK: + set_host_byte(scmd, DID_OK); + scmd->result |= le16_to_cpu(cqe->status) >> 8; + if (scmd->result & SAM_STAT_CHECK_CONDITION) { + memset(scmd->sense_buffer, 0, SCSI_SENSE_BUFFERSIZE); + memcpy(scmd->sense_buffer, iod->sense, + SCSI_SENSE_BUFFERSIZE); + scmd->result = + (scmd->result & 0x00ffffff) | (DRIVER_SENSE << 24); + } + break; + case FW_STAT_ABORTED: + set_host_byte(scmd, DID_ABORT); + break; + case FW_STAT_NEED_RETRY: + set_host_byte(scmd, DID_REQUEUE); + break; + default: + set_host_byte(scmd, DID_BAD_TARGET); + break; + } +} + +static inline void spraid_get_tag_from_scmd(struct scsi_cmnd *scmd, + u16 *qid, u16 *cid) +{ + u32 tag = blk_mq_unique_tag(scmd->request); + + *qid = blk_mq_unique_tag_to_hwq(tag) + 1; + *cid = blk_mq_unique_tag_to_tag(tag); +} + +static int spraid_queue_command(struct Scsi_Host *shost, struct scsi_cmnd *scmd) +{ + struct spraid_iod *iod = scsi_cmd_priv(scmd); + struct spraid_dev *hdev = shost_priv(shost); + struct scsi_device *sdev = scmd->device; + struct spraid_sdev_hostdata *hostdata; + struct spraid_ioq_command ioq_cmd; + struct spraid_queue *ioq; + unsigned long elapsed; + u16 hwq, cid; + int ret; + + if (unlikely(!scmd)) { + dev_err(hdev->dev, "err, scmd is null\n"); + return 0; + } + + if (unlikely(hdev->state != SPRAID_LIVE)) { + set_host_byte(scmd, DID_NO_CONNECT); + scmd->scsi_done(scmd); + return 0; + } + + if (log_debug_switch) + scsi_print_command(scmd); + + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + hostdata = sdev->hostdata; + ioq = &hdev->queues[hwq]; + memset(&ioq_cmd, 0, sizeof(ioq_cmd)); + ioq_cmd.rw.hdid = cpu_to_le32(hostdata->hdid); + ioq_cmd.rw.command_id = cid; + + spraid_setup_ioq_cmd(hdev, &ioq_cmd, scmd); + + ret = cid * SCSI_SENSE_BUFFERSIZE; + iod->sense = ioq->sense + ret; + iod->sense_dma = ioq->sense_dma_addr + ret; + + ret = spraid_init_iod(hdev, iod, &ioq_cmd, scmd); + if (unlikely(ret)) + return SCSI_MLQUEUE_HOST_BUSY; + + iod->spraidq = ioq; + ret = spraid_io_map_data(hdev, iod, scmd, &ioq_cmd); + if (unlikely(ret)) { + dev_err(hdev->dev, "spraid_io_map_data Err.\n"); + set_host_byte(scmd, DID_ERROR); + scmd->scsi_done(scmd); + ret = 0; + goto deinit_iod; + } + + WRITE_ONCE(iod->state, SPRAID_CMD_IN_FLIGHT); + spraid_submit_cmd(ioq, &ioq_cmd); + elapsed = jiffies - scmd->jiffies_at_alloc; + dev_log_dbg(hdev->dev, + "cid[%d] qid[%d] submit IO cost %3ld.%3ld seconds\n", + cid, hwq, elapsed / HZ, elapsed % HZ); + return 0; + +deinit_iod: + spraid_free_iod_res(hdev, iod); + return ret; +} + +static int spraid_match_dev(struct spraid_dev *hdev, u16 idx, + struct scsi_device *sdev) +{ + if (SPRAID_DEV_INFO_FLAG_VALID(hdev->devices[idx].flag)) { + if (sdev->channel == hdev->devices[idx].channel && + sdev->id == le16_to_cpu(hdev->devices[idx].target) && + sdev->lun < hdev->devices[idx].lun) { + dev_info(hdev->dev, + "Match device success, channel;" + "target:lun[%d:%d:%d]\n", + hdev->devices[idx].channel, + hdev->devices[idx].target, + hdev->devices[idx].lun); + return 1; + } + } + + return 0; +} + +static int spraid_slave_alloc(struct scsi_device *sdev) +{ + struct spraid_sdev_hostdata *hostdata; + struct spraid_dev *hdev; + u16 idx; + + hdev = shost_priv(sdev->host); + hostdata = kzalloc(sizeof(*hostdata), GFP_KERNEL); + if (!hostdata) { + dev_err(hdev->dev, "Alloc scsi host data memory failed\n"); + return -ENOMEM; + } + + down_read(&hdev->devices_rwsem); + for (idx = 0; idx < le32_to_cpu(hdev->ctrl_info->nd); idx++) { + if (spraid_match_dev(hdev, idx, sdev)) + goto scan_host; + } + up_read(&hdev->devices_rwsem); + + kfree(hostdata); + return -ENXIO; + +scan_host: + hostdata->hdid = le32_to_cpu(hdev->devices[idx].hdid); + hostdata->max_io_kb = le16_to_cpu(hdev->devices[idx].max_io_kb); + hostdata->attr = hdev->devices[idx].attr; + hostdata->flag = hdev->devices[idx].flag; + hostdata->rg_id = 0xff; + sdev->hostdata = hostdata; + up_read(&hdev->devices_rwsem); + return 0; +} + +static void spraid_slave_destroy(struct scsi_device *sdev) +{ + kfree(sdev->hostdata); + sdev->hostdata = NULL; +} + +static int spraid_slave_configure(struct scsi_device *sdev) +{ + u16 idx; + unsigned int timeout = scmd_tmout_nonpt * HZ; + struct spraid_dev *hdev = shost_priv(sdev->host); + struct spraid_sdev_hostdata *hostdata = sdev->hostdata; + u32 max_sec = sdev->host->max_sectors; + + if (hostdata) { + idx = hostdata->hdid - 1; + if (sdev->channel == hdev->devices[idx].channel && + sdev->id == le16_to_cpu(hdev->devices[idx].target) && + sdev->lun < hdev->devices[idx].lun) { + if (SPRAID_DEV_INFO_ATTR_PT(hdev->devices[idx].attr)) + timeout = 30 * HZ; + else + timeout = scmd_tmout_nonpt * HZ; + max_sec = le16_to_cpu(hdev->devices[idx].max_io_kb) + << 1; + } else { + dev_err(hdev->dev, + "[%s] err, sdev->channel:id:lun[%d:%d:%lld];" + "devices[%d], channel:target:lun[%d:%d:%d]\n", + __func__, sdev->channel, sdev->id, sdev->lun, + idx, hdev->devices[idx].channel, + hdev->devices[idx].target, + hdev->devices[idx].lun); + } + } else { + dev_err(hdev->dev, "[%s] err, sdev->hostdata is null\n", + __func__); + } + + blk_queue_rq_timeout(sdev->request_queue, timeout); + sdev->eh_timeout = timeout; + + if ((max_sec == 0) || (max_sec > sdev->host->max_sectors)) + max_sec = sdev->host->max_sectors; + + dev_info(hdev->dev, + "[%s] sdev->channel:id:lun[%d:%d:%lld];" + " scmd_timeout[%d]s, maxsec[%d]\n", + __func__, sdev->channel, sdev->id, + sdev->lun, timeout / HZ, max_sec); + + return 0; +} + +static void spraid_shost_init(struct spraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + u8 domain, bus; + u32 dev_func; + + domain = pci_domain_nr(pdev->bus); + bus = pdev->bus->number; + dev_func = pdev->devfn; + + hdev->shost->nr_hw_queues = hdev->online_queues - 1; + hdev->shost->can_queue = (hdev->ioq_depth - SPRAID_PTCMDS_PERQ); + + hdev->shost->sg_tablesize = le16_to_cpu(hdev->ctrl_info->max_num_sge); + /* 512B per sector */ + hdev->shost->max_sectors = + (1U << ((hdev->ctrl_info->mdts) * 1U) << 12) / 512; + hdev->shost->cmd_per_lun = MAX_CMD_PER_DEV; + hdev->shost->max_channel = + le16_to_cpu(hdev->ctrl_info->max_channel) - 1; + hdev->shost->max_id = le32_to_cpu(hdev->ctrl_info->max_tgt_id); + hdev->shost->max_lun = le16_to_cpu(hdev->ctrl_info->max_lun); + + hdev->shost->this_id = -1; + hdev->shost->unique_id = (domain << 16) | (bus << 8) | dev_func; + hdev->shost->max_cmd_len = MAX_CDB_LEN; + hdev->shost->hostt->cmd_size = max(spraid_cmd_size(hdev, false, true), + spraid_cmd_size(hdev, false, false)); +} + +static inline void spraid_host_deinit(struct spraid_dev *hdev) +{ + ida_free(&spraid_instance_ida, hdev->instance); +} + +static int spraid_alloc_queue(struct spraid_dev *hdev, u16 qid, u16 depth) +{ + struct spraid_queue *spraidq = &hdev->queues[qid]; + int ret = 0; + + if (hdev->queue_count > qid) { + dev_info(hdev->dev, "[%s] warn: queue[%d] is exist\n", + __func__, qid); + return 0; + } + + spraidq->cqes = dma_alloc_coherent(hdev->dev, CQ_SIZE(depth), + &spraidq->cq_dma_addr, + GFP_KERNEL | __GFP_ZERO); + if (!spraidq->cqes) + return -ENOMEM; + + spraidq->sq_cmds = dma_alloc_coherent(hdev->dev, SQ_SIZE(qid, depth), + &spraidq->sq_dma_addr, + GFP_KERNEL); + if (!spraidq->sq_cmds) { + ret = -ENOMEM; + goto free_cqes; + } + + spin_lock_init(&spraidq->sq_lock); + spin_lock_init(&spraidq->cq_lock); + spraidq->hdev = hdev; + spraidq->q_depth = depth; + spraidq->qid = qid; + spraidq->cq_vector = -1; + hdev->queue_count++; + + /* alloc sense buffer */ + spraidq->sense = dma_alloc_coherent(hdev->dev, SENSE_SIZE(depth), + &spraidq->sense_dma_addr, + GFP_KERNEL | __GFP_ZERO); + if (!spraidq->sense) { + ret = -ENOMEM; + goto free_sq_cmds; + } + + return 0; + +free_sq_cmds: + dma_free_coherent(hdev->dev, SQ_SIZE(qid, depth), + (void *)spraidq->sq_cmds, spraidq->sq_dma_addr); +free_cqes: + dma_free_coherent(hdev->dev, CQ_SIZE(depth), (void *)spraidq->cqes, + spraidq->cq_dma_addr); + return ret; +} + +static int spraid_wait_ready(struct spraid_dev *hdev, u64 cap, bool enabled) +{ + unsigned long timeout = + ((SPRAID_CAP_TIMEOUT(cap) + 1) * SPRAID_CAP_TIMEOUT_UNIT_MS) + jiffies; + u32 bit = enabled ? SPRAID_CSTS_RDY : 0; + + while ((readl(hdev->bar + SPRAID_REG_CSTS) & SPRAID_CSTS_RDY) != bit) { + usleep_range(1000, 2000); + if (fatal_signal_pending(current)) + return -EINTR; + + if (time_after(jiffies, timeout)) { + dev_err(hdev->dev, "Device not ready; aborting %s\n", + enabled ? "initialisation" : "reset"); + return -ENODEV; + } + } + return 0; +} + +static int spraid_shutdown_ctrl(struct spraid_dev *hdev) +{ + unsigned long timeout = hdev->ctrl_info->rtd3e + jiffies; + + hdev->ctrl_config &= ~SPRAID_CC_SHN_MASK; + hdev->ctrl_config |= SPRAID_CC_SHN_NORMAL; + writel(hdev->ctrl_config, hdev->bar + SPRAID_REG_CC); + + while ((readl(hdev->bar + SPRAID_REG_CSTS) & SPRAID_CSTS_SHST_MASK) != + SPRAID_CSTS_SHST_CMPLT) { + msleep(100); + if (fatal_signal_pending(current)) + return -EINTR; + if (time_after(jiffies, timeout)) { + dev_err(hdev->dev, + "Device shutdown incomplete; abort shutdown\n"); + return -ENODEV; + } + } + return 0; +} + +static int spraid_disable_ctrl(struct spraid_dev *hdev) +{ + hdev->ctrl_config &= ~SPRAID_CC_SHN_MASK; + hdev->ctrl_config &= ~SPRAID_CC_ENABLE; + writel(hdev->ctrl_config, hdev->bar + SPRAID_REG_CC); + + return spraid_wait_ready(hdev, hdev->cap, false); +} + +static int spraid_enable_ctrl(struct spraid_dev *hdev) +{ + u64 cap = hdev->cap; + u32 dev_page_min = SPRAID_CAP_MPSMIN(cap) + 12; + u32 page_shift = PAGE_SHIFT; + + if (page_shift < dev_page_min) { + dev_err(hdev->dev, + "Minimum device page size[%u], too large for host[%u]\n", + 1U << dev_page_min, 1U << page_shift); + return -ENODEV; + } + + page_shift = min_t(unsigned int, SPRAID_CAP_MPSMAX(cap) + 12, + PAGE_SHIFT); + hdev->page_size = 1U << page_shift; + + hdev->ctrl_config = SPRAID_CC_CSS_NVM; + hdev->ctrl_config |= (page_shift - 12) << SPRAID_CC_MPS_SHIFT; + hdev->ctrl_config |= SPRAID_CC_AMS_RR | SPRAID_CC_SHN_NONE; + hdev->ctrl_config |= SPRAID_CC_IOSQES | SPRAID_CC_IOCQES; + hdev->ctrl_config |= SPRAID_CC_ENABLE; + writel(hdev->ctrl_config, hdev->bar + SPRAID_REG_CC); + + return spraid_wait_ready(hdev, cap, true); +} + +static void spraid_init_queue(struct spraid_queue *spraidq, u16 qid) +{ + struct spraid_dev *hdev = spraidq->hdev; + + memset((void *)spraidq->cqes, 0, CQ_SIZE(spraidq->q_depth)); + + spraidq->sq_tail = 0; + spraidq->cq_head = 0; + spraidq->cq_phase = 1; + spraidq->q_db = &hdev->dbs[qid * 2 * hdev->db_stride]; + spraidq->prp_small_pool = hdev->prp_small_pool[qid % small_pool_num]; + hdev->online_queues++; +} + +static inline bool spraid_cqe_pending(struct spraid_queue *spraidq) +{ + return (le16_to_cpu(spraidq->cqes[spraidq->cq_head].status) & 1) == + spraidq->cq_phase; +} + +static void spraid_sata_report_zone_handle(struct scsi_cmnd *scmd, + struct spraid_iod *iod) +{ + int i = 0; + unsigned int bytes = 0; + struct scatterlist *sg = scsi_sglist(scmd); + + scsi_for_each_sg(scmd, sg, iod->nsge, i) { + unsigned int offset = 0; + + if (bytes == 0) { + char *hdr; + u32 list_length; + u64 max_lba, opt_lba; + u16 same; + + hdr = sg_virt(sg); + + list_length = get_unaligned_le32(&hdr[0]); + same = get_unaligned_le16(&hdr[4]); + max_lba = get_unaligned_le64(&hdr[8]); + opt_lba = get_unaligned_le64(&hdr[16]); + put_unaligned_be32(list_length, &hdr[0]); + hdr[4] = same & 0xf; + put_unaligned_be64(max_lba, &hdr[8]); + put_unaligned_be64(opt_lba, &hdr[16]); + offset += 64; + bytes += 64; + } + while (offset < sg_dma_len(sg)) { + char *rec; + u8 cond, type, non_seq, reset; + u64 size, start, wp; + + rec = sg_virt(sg) + offset; + type = rec[0] & 0xf; + cond = (rec[1] >> 4) & 0xf; + non_seq = (rec[1] & 2); + reset = (rec[1] & 1); + size = get_unaligned_le64(&rec[8]); + start = get_unaligned_le64(&rec[16]); + wp = get_unaligned_le64(&rec[24]); + rec[0] = type; + rec[1] = (cond << 4) | non_seq | reset; + put_unaligned_be64(size, &rec[8]); + put_unaligned_be64(start, &rec[16]); + put_unaligned_be64(wp, &rec[24]); + WARN_ON(offset + 64 > sg_dma_len(sg)); + offset += 64; + bytes += 64; + } + } +} + +static inline void spraid_handle_ata_cmd(struct spraid_dev *hdev, + struct scsi_cmnd *scmd, + struct spraid_iod *iod) +{ + if (hdev->ctrl_info->card_type != SPRAID_CARD_HBA) + return; + + switch (scmd->cmnd[0]) { + case ZBC_IN: + dev_info(hdev->dev, "[%s] process report zone\n", __func__); + spraid_sata_report_zone_handle(scmd, iod); + break; + default: + break; + } +} + +static void spraid_complete_ioq_cmnd(struct spraid_queue *ioq, + struct spraid_completion *cqe) +{ + struct spraid_dev *hdev = ioq->hdev; + struct blk_mq_tags *tags; + struct scsi_cmnd *scmd; + struct spraid_iod *iod; + struct request *req; + unsigned long elapsed; + + tags = hdev->shost->tag_set.tags[ioq->qid - 1]; + req = blk_mq_tag_to_rq(tags, cqe->cmd_id); + if (unlikely(!req || !blk_mq_request_started(req))) { + dev_warn(hdev->dev, "Invalid id %d completed on queue %d\n", + cqe->cmd_id, ioq->qid); + return; + } + + scmd = blk_mq_rq_to_pdu(req); + iod = scsi_cmd_priv(scmd); + + elapsed = jiffies - scmd->jiffies_at_alloc; + dev_log_dbg(hdev->dev, + "cid[%d] qid[%d] finish IO cost %3ld.%3ld seconds\n", + cqe->cmd_id, ioq->qid, elapsed / HZ, elapsed % HZ); + + if (cmpxchg(&iod->state, SPRAID_CMD_IN_FLIGHT, SPRAID_CMD_COMPLETE) != + SPRAID_CMD_IN_FLIGHT) { + dev_warn(hdev->dev, + "cid[%d] qid[%d] enters abnormal handler;" + " cost %3ld.%3ld seconds\n", + cqe->cmd_id, ioq->qid, elapsed / HZ, elapsed % HZ); + WRITE_ONCE(iod->state, SPRAID_CMD_TMO_COMPLETE); + + if (iod->nsge) { + iod->nsge = 0; + scsi_dma_unmap(scmd); + } + spraid_free_iod_res(hdev, iod); + + return; + } + + spraid_handle_ata_cmd(hdev, scmd, iod); + + spraid_map_status(iod, scmd, cqe); + if (iod->nsge) { + iod->nsge = 0; + scsi_dma_unmap(scmd); + } + spraid_free_iod_res(hdev, iod); + scmd->scsi_done(scmd); +} + +static void spraid_complete_adminq_cmnd(struct spraid_queue *adminq, + struct spraid_completion *cqe) +{ + struct spraid_dev *hdev = adminq->hdev; + struct spraid_cmd *adm_cmd; + + adm_cmd = hdev->adm_cmds + cqe->cmd_id; + if (unlikely(adm_cmd->state == SPRAID_CMD_IDLE)) { + dev_warn(adminq->hdev->dev, + "Invalid id %d completed on queue %d\n", + cqe->cmd_id, le16_to_cpu(cqe->sq_id)); + return; + } + + adm_cmd->status = le16_to_cpu(cqe->status) >> 1; + adm_cmd->result0 = le32_to_cpu(cqe->result); + adm_cmd->result1 = le32_to_cpu(cqe->result1); + + complete(&adm_cmd->cmd_done); +} + +static void spraid_send_aen(struct spraid_dev *hdev, u16 cid); + +static void spraid_complete_aen(struct spraid_queue *spraidq, + struct spraid_completion *cqe) +{ + struct spraid_dev *hdev = spraidq->hdev; + u32 result = le32_to_cpu(cqe->result); + + dev_info(hdev->dev, "rcv aen, cid[%d], status[0x%x], result[0x%x]\n", + cqe->cmd_id, le16_to_cpu(cqe->status) >> 1, result); + + spraid_send_aen(hdev, cqe->cmd_id); + + if ((le16_to_cpu(cqe->status) >> 1) != SPRAID_SC_SUCCESS) + return; + switch (result & 0x7) { + case SPRAID_AEN_NOTICE: + spraid_handle_aen_notice(hdev, result); + break; + case SPRAID_AEN_VS: + spraid_handle_aen_vs(hdev, result, le32_to_cpu(cqe->result1)); + break; + default: + dev_warn(hdev->dev, "Unsupported async event type: %u\n", + result & 0x7); + break; + } +} + +static void spraid_complete_ioq_sync_cmnd(struct spraid_queue *ioq, + struct spraid_completion *cqe) +{ + struct spraid_dev *hdev = ioq->hdev; + struct spraid_cmd *ptcmd; + + ptcmd = hdev->ioq_ptcmds + (ioq->qid - 1) * SPRAID_PTCMDS_PERQ + + cqe->cmd_id - SPRAID_IO_BLK_MQ_DEPTH; + + ptcmd->status = le16_to_cpu(cqe->status) >> 1; + ptcmd->result0 = le32_to_cpu(cqe->result); + ptcmd->result1 = le32_to_cpu(cqe->result1); + + complete(&ptcmd->cmd_done); +} + +static inline void spraid_handle_cqe(struct spraid_queue *spraidq, u16 idx) +{ + struct spraid_completion *cqe = &spraidq->cqes[idx]; + struct spraid_dev *hdev = spraidq->hdev; + + if (unlikely(cqe->cmd_id >= spraidq->q_depth)) { + dev_err(hdev->dev, + "Invalid command id[%d] completed on queue %d\n", + cqe->cmd_id, cqe->sq_id); + return; + } + + dev_log_dbg(hdev->dev, "cid[%d] qid[%d];" + " result[0x%x], sq_id[%d], status[0x%x]\n", + cqe->cmd_id, spraidq->qid, le32_to_cpu(cqe->result), + le16_to_cpu(cqe->sq_id), le16_to_cpu(cqe->status)); + + if (unlikely(spraidq->qid == 0 + && cqe->cmd_id >= SPRAID_AQ_BLK_MQ_DEPTH)) { + spraid_complete_aen(spraidq, cqe); + return; + } + + if (unlikely(spraidq->qid && cqe->cmd_id >= SPRAID_IO_BLK_MQ_DEPTH)) { + spraid_complete_ioq_sync_cmnd(spraidq, cqe); + return; + } + + if (spraidq->qid) + spraid_complete_ioq_cmnd(spraidq, cqe); + else + spraid_complete_adminq_cmnd(spraidq, cqe); +} + +static void spraid_complete_cqes(struct spraid_queue *spraidq, + u16 start, u16 end) +{ + while (start != end) { + spraid_handle_cqe(spraidq, start); + if (++start == spraidq->q_depth) + start = 0; + } +} + +static inline void spraid_update_cq_head(struct spraid_queue *spraidq) +{ + if (++spraidq->cq_head == spraidq->q_depth) { + spraidq->cq_head = 0; + spraidq->cq_phase = !spraidq->cq_phase; + } +} + +static inline bool spraid_process_cq(struct spraid_queue *spraidq, + u16 *start, u16 *end, int tag) +{ + bool found = false; + + *start = spraidq->cq_head; + while (!found && spraid_cqe_pending(spraidq)) { + if (spraidq->cqes[spraidq->cq_head].cmd_id == tag) + found = true; + spraid_update_cq_head(spraidq); + } + *end = spraidq->cq_head; + + if (*start != *end) + writel(spraidq->cq_head, + spraidq->q_db + spraidq->hdev->db_stride); + + return found; +} + +static bool spraid_poll_cq(struct spraid_queue *spraidq, int cid) +{ + u16 start, end; + bool found; + + if (!spraid_cqe_pending(spraidq)) + return 0; + + spin_lock_irq(&spraidq->cq_lock); + found = spraid_process_cq(spraidq, &start, &end, cid); + spin_unlock_irq(&spraidq->cq_lock); + + spraid_complete_cqes(spraidq, start, end); + return found; +} + +static irqreturn_t spraid_irq(int irq, void *data) +{ + struct spraid_queue *spraidq = data; + irqreturn_t ret = IRQ_NONE; + u16 start, end; + + spin_lock(&spraidq->cq_lock); + if (spraidq->cq_head != spraidq->last_cq_head) + ret = IRQ_HANDLED; + + spraid_process_cq(spraidq, &start, &end, -1); + spraidq->last_cq_head = spraidq->cq_head; + spin_unlock(&spraidq->cq_lock); + + if (start != end) { + spraid_complete_cqes(spraidq, start, end); + ret = IRQ_HANDLED; + } + return ret; +} + +static int spraid_setup_admin_queue(struct spraid_dev *hdev) +{ + struct spraid_queue *adminq = &hdev->queues[0]; + u32 aqa; + int ret; + + dev_info(hdev->dev, "[%s] start disable ctrl\n", __func__); + + ret = spraid_disable_ctrl(hdev); + if (ret) + return ret; + + ret = spraid_alloc_queue(hdev, 0, SPRAID_AQ_DEPTH); + if (ret) + return ret; + + aqa = adminq->q_depth - 1; + aqa |= aqa << 16; + writel(aqa, hdev->bar + SPRAID_REG_AQA); + lo_hi_writeq(adminq->sq_dma_addr, hdev->bar + SPRAID_REG_ASQ); + lo_hi_writeq(adminq->cq_dma_addr, hdev->bar + SPRAID_REG_ACQ); + + dev_info(hdev->dev, "[%s] start enable ctrl\n", __func__); + + ret = spraid_enable_ctrl(hdev); + if (ret) { + ret = -ENODEV; + goto free_queue; + } + + adminq->cq_vector = 0; + spraid_init_queue(adminq, 0); + ret = pci_request_irq(hdev->pdev, adminq->cq_vector, spraid_irq, NULL, + adminq, "spraid%d_q%d", + hdev->instance, adminq->qid); + + if (ret) { + adminq->cq_vector = -1; + hdev->online_queues--; + goto free_queue; + } + + dev_info(hdev->dev, "[%s] success, queuecount:[%d], onlinequeue:[%d]\n", + __func__, hdev->queue_count, hdev->online_queues); + + return 0; + +free_queue: + spraid_free_queue(adminq); + return ret; +} + +static u32 spraid_bar_size(struct spraid_dev *hdev, u32 nr_ioqs) +{ + return (SPRAID_REG_DBS + ((nr_ioqs + 1) * 8 * hdev->db_stride)); +} + +static int spraid_alloc_admin_cmds(struct spraid_dev *hdev) +{ + int i; + + INIT_LIST_HEAD(&hdev->adm_cmd_list); + spin_lock_init(&hdev->adm_cmd_lock); + + hdev->adm_cmds = kcalloc_node(SPRAID_AQ_BLK_MQ_DEPTH, + sizeof(struct spraid_cmd), + GFP_KERNEL, hdev->numa_node); + + if (!hdev->adm_cmds) { + dev_err(hdev->dev, "Alloc admin cmds failed\n"); + return -ENOMEM; + } + + for (i = 0; i < SPRAID_AQ_BLK_MQ_DEPTH; i++) { + hdev->adm_cmds[i].qid = 0; + hdev->adm_cmds[i].cid = i; + list_add_tail(&(hdev->adm_cmds[i].list), &hdev->adm_cmd_list); + } + + dev_info(hdev->dev, "Alloc admin cmds success, num[%d]\n", + SPRAID_AQ_BLK_MQ_DEPTH); + + return 0; +} + +static void spraid_free_admin_cmds(struct spraid_dev *hdev) +{ + kfree(hdev->adm_cmds); + hdev->adm_cmds = NULL; + INIT_LIST_HEAD(&hdev->adm_cmd_list); +} + +static struct spraid_cmd *spraid_get_cmd(struct spraid_dev *hdev, + enum spraid_cmd_type type) +{ + struct spraid_cmd *cmd = NULL; + unsigned long flags; + struct list_head *head = &hdev->adm_cmd_list; + spinlock_t *slock = &hdev->adm_cmd_lock; + + if (type == SPRAID_CMD_IOPT) { + head = &hdev->ioq_pt_list; + slock = &hdev->ioq_pt_lock; + } + + spin_lock_irqsave(slock, flags); + if (list_empty(head)) { + spin_unlock_irqrestore(slock, flags); + dev_err(hdev->dev, "err, cmd[%d] list empty\n", type); + return NULL; + } + cmd = list_entry(head->next, struct spraid_cmd, list); + list_del_init(&cmd->list); + spin_unlock_irqrestore(slock, flags); + + WRITE_ONCE(cmd->state, SPRAID_CMD_IN_FLIGHT); + + return cmd; +} + +static void spraid_put_cmd(struct spraid_dev *hdev, struct spraid_cmd *cmd, + enum spraid_cmd_type type) +{ + unsigned long flags; + struct list_head *head = &hdev->adm_cmd_list; + spinlock_t *slock = &hdev->adm_cmd_lock; + + if (type == SPRAID_CMD_IOPT) { + head = &hdev->ioq_pt_list; + slock = &hdev->ioq_pt_lock; + } + + spin_lock_irqsave(slock, flags); + WRITE_ONCE(cmd->state, SPRAID_CMD_IDLE); + list_add_tail(&cmd->list, head); + spin_unlock_irqrestore(slock, flags); +} + + +static int spraid_submit_admin_sync_cmd(struct spraid_dev *hdev, + struct spraid_admin_command *cmd, + u32 *result0, u32 *result1, u32 timeout) +{ + struct spraid_cmd *adm_cmd = spraid_get_cmd(hdev, SPRAID_CMD_ADM); + + if (!adm_cmd) { + dev_err(hdev->dev, "err, get admin cmd failed\n"); + return -EFAULT; + } + + timeout = timeout ? timeout : ADMIN_TIMEOUT; + + init_completion(&adm_cmd->cmd_done); + + cmd->common.command_id = adm_cmd->cid; + spraid_submit_cmd(&hdev->queues[0], cmd); + + if (!wait_for_completion_timeout(&adm_cmd->cmd_done, timeout)) { + dev_err(hdev->dev, "[%s] cid[%d] qid[%d] timeout;" + " opcode[0x%x] subopcode[0x%x]\n", + __func__, adm_cmd->cid, adm_cmd->qid, + cmd->usr_cmd.opcode, cmd->usr_cmd.info_0.subopcode); + WRITE_ONCE(adm_cmd->state, SPRAID_CMD_TIMEOUT); + spraid_put_cmd(hdev, adm_cmd, SPRAID_CMD_ADM); + return -EINVAL; + } + + if (result0) + *result0 = adm_cmd->result0; + if (result1) + *result1 = adm_cmd->result1; + + spraid_put_cmd(hdev, adm_cmd, SPRAID_CMD_ADM); + + return adm_cmd->status; +} + +static int spraid_create_cq(struct spraid_dev *hdev, u16 qid, + struct spraid_queue *spraidq, u16 cq_vector) +{ + struct spraid_admin_command admin_cmd; + int flags = SPRAID_QUEUE_PHYS_CONTIG | SPRAID_CQ_IRQ_ENABLED; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.create_cq.opcode = SPRAID_ADMIN_CREATE_CQ; + admin_cmd.create_cq.prp1 = cpu_to_le64(spraidq->cq_dma_addr); + admin_cmd.create_cq.cqid = cpu_to_le16(qid); + admin_cmd.create_cq.qsize = cpu_to_le16(spraidq->q_depth - 1); + admin_cmd.create_cq.cq_flags = cpu_to_le16(flags); + admin_cmd.create_cq.irq_vector = cpu_to_le16(cq_vector); + + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); +} + +static int spraid_create_sq(struct spraid_dev *hdev, u16 qid, + struct spraid_queue *spraidq) +{ + struct spraid_admin_command admin_cmd; + int flags = SPRAID_QUEUE_PHYS_CONTIG; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.create_sq.opcode = SPRAID_ADMIN_CREATE_SQ; + admin_cmd.create_sq.prp1 = cpu_to_le64(spraidq->sq_dma_addr); + admin_cmd.create_sq.sqid = cpu_to_le16(qid); + admin_cmd.create_sq.qsize = cpu_to_le16(spraidq->q_depth - 1); + admin_cmd.create_sq.sq_flags = cpu_to_le16(flags); + admin_cmd.create_sq.cqid = cpu_to_le16(qid); + + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); +} + +static void spraid_free_queue(struct spraid_queue *spraidq) +{ + struct spraid_dev *hdev = spraidq->hdev; + + hdev->queue_count--; + dma_free_coherent(hdev->dev, CQ_SIZE(spraidq->q_depth), + (void *)spraidq->cqes, spraidq->cq_dma_addr); + dma_free_coherent(hdev->dev, SQ_SIZE(spraidq->qid, spraidq->q_depth), + spraidq->sq_cmds, spraidq->sq_dma_addr); + dma_free_coherent(hdev->dev, SENSE_SIZE(spraidq->q_depth), + spraidq->sense, spraidq->sense_dma_addr); +} + +static void spraid_free_admin_queue(struct spraid_dev *hdev) +{ + spraid_free_queue(&hdev->queues[0]); +} + +static void spraid_free_io_queues(struct spraid_dev *hdev) +{ + int i; + + for (i = hdev->queue_count - 1; i >= 1; i--) + spraid_free_queue(&hdev->queues[i]); +} + +static int spraid_delete_queue(struct spraid_dev *hdev, u8 op, u16 id) +{ + struct spraid_admin_command admin_cmd; + int ret; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.delete_queue.opcode = op; + admin_cmd.delete_queue.qid = cpu_to_le16(id); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + + if (ret) + dev_err(hdev->dev, "Delete %s:[%d] failed\n", + (op == SPRAID_ADMIN_DELETE_CQ) ? "cq" : "sq", id); + + return ret; +} + +static int spraid_delete_cq(struct spraid_dev *hdev, u16 cqid) +{ + return spraid_delete_queue(hdev, SPRAID_ADMIN_DELETE_CQ, cqid); +} + +static int spraid_delete_sq(struct spraid_dev *hdev, u16 sqid) +{ + return spraid_delete_queue(hdev, SPRAID_ADMIN_DELETE_SQ, sqid); +} + +static int spraid_create_queue(struct spraid_queue *spraidq, u16 qid) +{ + struct spraid_dev *hdev = spraidq->hdev; + u16 cq_vector; + int ret; + + cq_vector = (hdev->num_vecs == 1) ? 0 : qid; + ret = spraid_create_cq(hdev, qid, spraidq, cq_vector); + if (ret) + return ret; + + ret = spraid_create_sq(hdev, qid, spraidq); + if (ret) + goto delete_cq; + + spraid_init_queue(spraidq, qid); + spraidq->cq_vector = cq_vector; + + ret = pci_request_irq(hdev->pdev, cq_vector, spraid_irq, NULL, + spraidq, "spraid%d_q%d", hdev->instance, qid); + + if (ret) { + dev_err(hdev->dev, "Request queue[%d] irq failed\n", qid); + goto delete_sq; + } + + return 0; + +delete_sq: + spraidq->cq_vector = -1; + hdev->online_queues--; + spraid_delete_sq(hdev, qid); +delete_cq: + spraid_delete_cq(hdev, qid); + + return ret; +} + +static int spraid_create_io_queues(struct spraid_dev *hdev) +{ + u32 i, max; + int ret = 0; + + max = min(hdev->max_qid, hdev->queue_count - 1); + for (i = hdev->online_queues; i <= max; i++) { + ret = spraid_create_queue(&hdev->queues[i], i); + if (ret) { + dev_err(hdev->dev, "Create queue[%d] failed\n", i); + break; + } + } + + dev_info(hdev->dev, "[%s] queue_count[%d], online_queue[%d]", + __func__, hdev->queue_count, hdev->online_queues); + + return ret >= 0 ? 0 : ret; +} + +static int spraid_set_features(struct spraid_dev *hdev, u32 fid, + u32 dword11, void *buffer, + size_t buflen, u32 *result) +{ + struct spraid_admin_command admin_cmd; + int ret; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + + if (buffer && buflen) { + data_ptr = dma_alloc_coherent(hdev->dev, buflen, + &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memcpy(data_ptr, buffer, buflen); + } + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.features.opcode = SPRAID_ADMIN_SET_FEATURES; + admin_cmd.features.fid = cpu_to_le32(fid); + admin_cmd.features.dword11 = cpu_to_le32(dword11); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, result, NULL, 0); + + if (data_ptr) + dma_free_coherent(hdev->dev, buflen, data_ptr, data_dma); + + return ret; +} + +static int spraid_configure_timestamp(struct spraid_dev *hdev) +{ + __le64 ts; + int ret; + + ts = cpu_to_le64(ktime_to_ms(ktime_get_real())); + ret = spraid_set_features(hdev, SPRAID_FEAT_TIMESTAMP, + 0, &ts, sizeof(ts), NULL); + + if (ret) + dev_err(hdev->dev, "set timestamp failed: %d\n", ret); + return ret; +} + +static int spraid_set_queue_cnt(struct spraid_dev *hdev, u32 *cnt) +{ + u32 q_cnt = (*cnt - 1) | ((*cnt - 1) << 16); + u32 nr_ioqs, result; + int status; + + status = spraid_set_features(hdev, SPRAID_FEAT_NUM_QUEUES, + q_cnt, NULL, 0, &result); + if (status) { + dev_err(hdev->dev, "Set queue count failed, status: %d\n", + status); + return -EIO; + } + + nr_ioqs = min(result & 0xffff, result >> 16) + 1; + *cnt = min(*cnt, nr_ioqs); + if (*cnt == 0) { + dev_err(hdev->dev, "Illegal queue count: zero\n"); + return -EIO; + } + return 0; +} + +static int spraid_setup_io_queues(struct spraid_dev *hdev) +{ + struct spraid_queue *adminq = &hdev->queues[0]; + struct pci_dev *pdev = hdev->pdev; + u32 nr_ioqs = num_online_cpus(); + u32 i, size; + int ret; + + struct irq_affinity affd = { + .pre_vectors = 1 + }; + + ret = spraid_set_queue_cnt(hdev, &nr_ioqs); + if (ret < 0) + return ret; + + size = spraid_bar_size(hdev, nr_ioqs); + ret = spraid_remap_bar(hdev, size); + if (ret) + return -ENOMEM; + + adminq->q_db = hdev->dbs; + + pci_free_irq(pdev, 0, adminq); + pci_free_irq_vectors(pdev); + + ret = pci_alloc_irq_vectors_affinity(pdev, 1, (nr_ioqs + 1), + PCI_IRQ_ALL_TYPES | PCI_IRQ_AFFINITY, &affd); + if (ret <= 0) + return -EIO; + + hdev->num_vecs = ret; + + hdev->max_qid = max(ret - 1, 1); + + ret = pci_request_irq(pdev, adminq->cq_vector, spraid_irq, NULL, + adminq, "spraid%d_q%d", + hdev->instance, adminq->qid); + if (ret) { + dev_err(hdev->dev, "Request admin irq failed\n"); + adminq->cq_vector = -1; + return ret; + } + + for (i = hdev->queue_count; i <= hdev->max_qid; i++) { + ret = spraid_alloc_queue(hdev, i, hdev->ioq_depth); + if (ret) + break; + } + dev_info(hdev->dev, "[%s] max_qid: %d, queue_count: %d;" + " online_queue: %d, ioq_depth: %d\n", + __func__, hdev->max_qid, hdev->queue_count, + hdev->online_queues, hdev->ioq_depth); + + return spraid_create_io_queues(hdev); +} + +static void spraid_delete_io_queues(struct spraid_dev *hdev) +{ + u16 queues = hdev->online_queues - 1; + u8 opcode = SPRAID_ADMIN_DELETE_SQ; + u16 i, pass; + + if (!pci_device_is_present(hdev->pdev)) { + dev_err(hdev->dev, + "pci_device is not present, skip disable io queues\n"); + return; + } + + if (hdev->online_queues < 2) { + dev_err(hdev->dev, "[%s] err, io queue has been delete\n", + __func__); + return; + } + + for (pass = 0; pass < 2; pass++) { + for (i = queues; i > 0; i--) + if (spraid_delete_queue(hdev, opcode, i)) + break; + + opcode = SPRAID_ADMIN_DELETE_CQ; + } +} + +static void spraid_remove_io_queues(struct spraid_dev *hdev) +{ + spraid_delete_io_queues(hdev); + spraid_free_io_queues(hdev); +} + +static void spraid_pci_disable(struct spraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + u32 i; + + for (i = 0; i < hdev->online_queues; i++) + pci_free_irq(pdev, hdev->queues[i].cq_vector, &hdev->queues[i]); + pci_free_irq_vectors(pdev); + if (pci_is_enabled(pdev)) { + pci_disable_pcie_error_reporting(pdev); + pci_disable_device(pdev); + } + hdev->online_queues = 0; +} + +static void spraid_disable_admin_queue(struct spraid_dev *hdev, bool shutdown) +{ + struct spraid_queue *adminq = &hdev->queues[0]; + u16 start, end; + + if (pci_device_is_present(hdev->pdev)) { + if (shutdown) + spraid_shutdown_ctrl(hdev); + else + spraid_disable_ctrl(hdev); + } + + if (hdev->queue_count == 0) { + dev_err(hdev->dev, "[%s] err, admin queue has been delete\n", + __func__); + return; + } + + spin_lock_irq(&adminq->cq_lock); + spraid_process_cq(adminq, &start, &end, -1); + spin_unlock_irq(&adminq->cq_lock); + + spraid_complete_cqes(adminq, start, end); + spraid_free_admin_queue(hdev); +} + +static int spraid_create_dma_pools(struct spraid_dev *hdev) +{ + int i; + char poolname[20] = { 0 }; + + hdev->prp_page_pool = dma_pool_create("prp list page", hdev->dev, + PAGE_SIZE, PAGE_SIZE, 0); + + if (!hdev->prp_page_pool) { + dev_err(hdev->dev, "create prp_page_pool failed\n"); + return -ENOMEM; + } + + for (i = 0; i < small_pool_num; i++) { + sprintf(poolname, "prp_list_256_%d", i); + hdev->prp_small_pool[i] = + dma_pool_create(poolname, hdev->dev, SMALL_POOL_SIZE, + SMALL_POOL_SIZE, 0); + + if (!hdev->prp_small_pool[i]) { + dev_err(hdev->dev, "create prp_small_pool %d failed\n", + i); + goto destroy_prp_small_pool; + } + } + + return 0; + +destroy_prp_small_pool: + while (i > 0) + dma_pool_destroy(hdev->prp_small_pool[--i]); + dma_pool_destroy(hdev->prp_page_pool); + + return -ENOMEM; +} + +static void spraid_destroy_dma_pools(struct spraid_dev *hdev) +{ + int i; + + for (i = 0; i < small_pool_num; i++) + dma_pool_destroy(hdev->prp_small_pool[i]); + dma_pool_destroy(hdev->prp_page_pool); +} + +static int spraid_get_dev_list(struct spraid_dev *hdev, + struct spraid_dev_info *devices) +{ + u32 nd = le32_to_cpu(hdev->ctrl_info->nd); + struct spraid_admin_command admin_cmd; + struct spraid_dev_list *list_buf; + dma_addr_t data_dma = 0; + u32 i, idx, hdid, ndev; + int ret = 0; + + list_buf = dma_alloc_coherent(hdev->dev, PAGE_SIZE, + &data_dma, GFP_KERNEL); + if (!list_buf) + return -ENOMEM; + + for (idx = 0; idx < nd;) { + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.get_info.opcode = SPRAID_ADMIN_GET_INFO; + admin_cmd.get_info.type = SPRAID_GET_INFO_DEV_LIST; + admin_cmd.get_info.cdw11 = cpu_to_le32(idx); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, + NULL, NULL, 0); + + if (ret) { + dev_err(hdev->dev, "Get device list failed, nd: %u;" + "idx: %u, ret: %d\n", + nd, idx, ret); + goto out; + } + ndev = le32_to_cpu(list_buf->dev_num); + + dev_info(hdev->dev, "ndev numbers: %u\n", ndev); + + for (i = 0; i < ndev; i++) { + hdid = le32_to_cpu(list_buf->devices[i].hdid); + dev_info(hdev->dev, "list_buf->devices[%d], hdid: %u;" + "target: %d, channel: %d, lun: %d, attr[0x%x]\n", + i, hdid, + le16_to_cpu(list_buf->devices[i].target), + list_buf->devices[i].channel, + list_buf->devices[i].lun, + list_buf->devices[i].attr); + if (hdid > nd || hdid == 0) { + dev_err(hdev->dev, "err, hdid[%d] invalid\n", + hdid); + continue; + } + memcpy(&devices[hdid - 1], &list_buf->devices[i], + sizeof(struct spraid_dev_info)); + } + idx += ndev; + + if (idx < MAX_DEV_ENTRY_PER_PAGE_4K) + break; + } + +out: + dma_free_coherent(hdev->dev, PAGE_SIZE, list_buf, data_dma); + return ret; +} + +static void spraid_send_aen(struct spraid_dev *hdev, u16 cid) +{ + struct spraid_queue *adminq = &hdev->queues[0]; + struct spraid_admin_command admin_cmd; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.common.opcode = SPRAID_ADMIN_ASYNC_EVENT; + admin_cmd.common.command_id = cid; + + spraid_submit_cmd(adminq, &admin_cmd); + dev_info(hdev->dev, "send aen, cid[%d]\n", cid); +} + +static inline void spraid_send_all_aen(struct spraid_dev *hdev) +{ + u16 i; + + for (i = 0; i < hdev->ctrl_info->aerl; i++) + spraid_send_aen(hdev, i + SPRAID_AQ_BLK_MQ_DEPTH); +} + +static int spraid_add_device(struct spraid_dev *hdev, + struct spraid_dev_info *device) +{ + struct Scsi_Host *shost = hdev->shost; + struct scsi_device *sdev; + + dev_info(hdev->dev, "add device, hdid: %u target: %d, channel: %d;" + " lun: %d, attr[0x%x]\n", + le32_to_cpu(device->hdid), le16_to_cpu(device->target), + device->channel, device->lun, device->attr); + + sdev = scsi_device_lookup(shost, device->channel, + le16_to_cpu(device->target), 0); + if (sdev) { + dev_warn(hdev->dev, "Device is already exist, channel: %d;" + " target_id: %d, lun: %d\n", + device->channel, le16_to_cpu(device->target), 0); + scsi_device_put(sdev); + return -EEXIST; + } + scsi_add_device(shost, device->channel, le16_to_cpu(device->target), 0); + return 0; +} + +static int spraid_rescan_device(struct spraid_dev *hdev, + struct spraid_dev_info *device) +{ + struct Scsi_Host *shost = hdev->shost; + struct scsi_device *sdev; + + dev_info(hdev->dev, "rescan device, hdid: %u target: %d, channel: %d;" + " lun: %d, attr[0x%x]\n", + le32_to_cpu(device->hdid), le16_to_cpu(device->target), + device->channel, device->lun, device->attr); + + sdev = scsi_device_lookup(shost, device->channel, + le16_to_cpu(device->target), 0); + if (!sdev) { + dev_warn(hdev->dev, "device is not exit rescan it, channel: %d;" + " target_id: %d, lun: %d\n", + device->channel, le16_to_cpu(device->target), 0); + return -ENODEV; + } + + scsi_rescan_device(&sdev->sdev_gendev); + scsi_device_put(sdev); + return 0; +} + +static int spraid_remove_device(struct spraid_dev *hdev, + struct spraid_dev_info *org_device) +{ + struct Scsi_Host *shost = hdev->shost; + struct scsi_device *sdev; + + dev_info(hdev->dev, "remove device, hdid: %u target: %d, channel: %d;" + " lun: %d, attr[0x%x]\n", + le32_to_cpu(org_device->hdid), le16_to_cpu(org_device->target), + org_device->channel, org_device->lun, org_device->attr); + + sdev = scsi_device_lookup(shost, org_device->channel, + le16_to_cpu(org_device->target), 0); + if (!sdev) { + dev_warn(hdev->dev, "device is not exit remove it, channel: %d;" + " target_id: %d, lun: %d\n", + org_device->channel, + le16_to_cpu(org_device->target), 0); + return -ENODEV; + } + + scsi_remove_device(sdev); + scsi_device_put(sdev); + return 0; +} + +static int spraid_dev_list_init(struct spraid_dev *hdev) +{ + u32 nd = le32_to_cpu(hdev->ctrl_info->nd); + int i, ret; + + hdev->devices = kzalloc_node(nd * sizeof(struct spraid_dev_info), + GFP_KERNEL, hdev->numa_node); + if (!hdev->devices) + return -ENOMEM; + + ret = spraid_get_dev_list(hdev, hdev->devices); + if (ret) { + dev_err(hdev->dev, + "Ignore failure of getting device list;" + " within initialization\n"); + return 0; + } + + for (i = 0; i < nd; i++) { + if (SPRAID_DEV_INFO_FLAG_VALID(hdev->devices[i].flag) && + SPRAID_DEV_INFO_ATTR_BOOT(hdev->devices[i].attr)) { + spraid_add_device(hdev, &hdev->devices[i]); + break; + } + } + return 0; +} + +static int luntarget_cmp_func(const void *l, const void *r) +{ + const struct spraid_dev_info *ln = l; + const struct spraid_dev_info *rn = r; + + if (ln->channel == rn->channel) + return le16_to_cpu(ln->target) - le16_to_cpu(rn->target); + + return ln->channel - rn->channel; +} + +static void spraid_scan_work(struct work_struct *work) +{ + struct spraid_dev *hdev = + container_of(work, struct spraid_dev, scan_work); + struct spraid_dev_info *devices, *org_devices; + struct spraid_dev_info *sortdevice; + u32 nd = le32_to_cpu(hdev->ctrl_info->nd); + u8 flag, org_flag; + int i, ret; + int count = 0; + + devices = kcalloc(nd, sizeof(struct spraid_dev_info), GFP_KERNEL); + if (!devices) + return; + + sortdevice = kcalloc(nd, sizeof(struct spraid_dev_info), GFP_KERNEL); + if (!sortdevice) + goto free_list; + + ret = spraid_get_dev_list(hdev, devices); + if (ret) + goto free_all; + org_devices = hdev->devices; + for (i = 0; i < nd; i++) { + org_flag = org_devices[i].flag; + flag = devices[i].flag; + + dev_log_dbg(hdev->dev, "i: %d, org_flag: 0x%x, flag: 0x%x\n", + i, org_flag, flag); + + if (SPRAID_DEV_INFO_FLAG_VALID(flag)) { + if (!SPRAID_DEV_INFO_FLAG_VALID(org_flag)) { + down_write(&hdev->devices_rwsem); + memcpy(&org_devices[i], &devices[i], + sizeof(struct spraid_dev_info)); + memcpy(&sortdevice[count++], &devices[i], + sizeof(struct spraid_dev_info)); + up_write(&hdev->devices_rwsem); + } else if (SPRAID_DEV_INFO_FLAG_CHANGE(flag)) { + spraid_rescan_device(hdev, &devices[i]); + } + } else { + if (SPRAID_DEV_INFO_FLAG_VALID(org_flag)) { + down_write(&hdev->devices_rwsem); + org_devices[i].flag &= 0xfe; + up_write(&hdev->devices_rwsem); + spraid_remove_device(hdev, &org_devices[i]); + } + } + } + + dev_info(hdev->dev, "scan work add device count = %d\n", count); + + sort(sortdevice, count, sizeof(sortdevice[0]), + luntarget_cmp_func, NULL); + + for (i = 0; i < count; i++) + spraid_add_device(hdev, &sortdevice[i]); + +free_all: + kfree(sortdevice); +free_list: + kfree(devices); +} + +static void spraid_timesyn_work(struct work_struct *work) +{ + struct spraid_dev *hdev = + container_of(work, struct spraid_dev, timesyn_work); + + spraid_configure_timestamp(hdev); +} + +static int spraid_init_ctrl_info(struct spraid_dev *hdev); +static void spraid_fw_act_work(struct work_struct *work) +{ + struct spraid_dev *hdev = + container_of(work, struct spraid_dev, fw_act_work); + + if (spraid_init_ctrl_info(hdev)) + dev_err(hdev->dev, "get ctrl info failed after fw act\n"); +} + +static void spraid_queue_scan(struct spraid_dev *hdev) +{ + queue_work(spraid_wq, &hdev->scan_work); +} + +static void spraid_handle_aen_notice(struct spraid_dev *hdev, u32 result) +{ + switch ((result & 0xff00) >> 8) { + case SPRAID_AEN_DEV_CHANGED: + spraid_queue_scan(hdev); + break; + case SPRAID_AEN_FW_ACT_START: + dev_info(hdev->dev, "fw activation starting\n"); + break; + case SPRAID_AEN_HOST_PROBING: + break; + default: + dev_warn(hdev->dev, "async event result %08x\n", result); + } +} + +static void spraid_handle_aen_vs(struct spraid_dev *hdev, + u32 result, u32 result1) +{ + switch ((result & 0xff00) >> 8) { + case SPRAID_AEN_TIMESYN: + queue_work(spraid_wq, &hdev->timesyn_work); + break; + case SPRAID_AEN_FW_ACT_FINISH: + dev_info(hdev->dev, "fw activation finish\n"); + queue_work(spraid_wq, &hdev->fw_act_work); + break; + case SPRAID_AEN_EVENT_MIN ... SPRAID_AEN_EVENT_MAX: + dev_info(hdev->dev, "rcv card event[%d];" + " param1[0x%x] param2[0x%x]\n", + (result & 0xff00) >> 8, result, result1); + break; + default: + dev_warn(hdev->dev, "async event result: 0x%x\n", result); + } +} + +static int spraid_alloc_resources(struct spraid_dev *hdev) +{ + int ret, nqueue; + + ret = ida_alloc(&spraid_instance_ida, GFP_KERNEL); + if (ret < 0) { + dev_err(hdev->dev, "Get instance id failed\n"); + return ret; + } + hdev->instance = ret; + + hdev->ctrl_info = kzalloc_node(sizeof(*hdev->ctrl_info), + GFP_KERNEL, hdev->numa_node); + if (!hdev->ctrl_info) { + ret = -ENOMEM; + goto release_instance; + } + + ret = spraid_create_dma_pools(hdev); + if (ret) + goto free_ctrl_info; + nqueue = num_possible_cpus() + 1; + hdev->queues = kcalloc_node(nqueue, sizeof(struct spraid_queue), + GFP_KERNEL, hdev->numa_node); + if (!hdev->queues) { + ret = -ENOMEM; + goto destroy_dma_pools; + } + + ret = spraid_alloc_admin_cmds(hdev); + if (ret) + goto free_queues; + + dev_info(hdev->dev, "[%s] queues num: %d\n", __func__, nqueue); + + return 0; + +free_queues: + kfree(hdev->queues); +destroy_dma_pools: + spraid_destroy_dma_pools(hdev); +free_ctrl_info: + kfree(hdev->ctrl_info); +release_instance: + ida_free(&spraid_instance_ida, hdev->instance); + return ret; +} + +static void spraid_free_resources(struct spraid_dev *hdev) +{ + spraid_free_admin_cmds(hdev); + kfree(hdev->queues); + spraid_destroy_dma_pools(hdev); + kfree(hdev->ctrl_info); + ida_free(&spraid_instance_ida, hdev->instance); +} + +static void spraid_bsg_unmap_data(struct spraid_dev *hdev, struct bsg_job *job) +{ + struct request *rq = blk_mq_rq_from_pdu(job); + struct spraid_iod *iod = job->dd_data; + enum dma_data_direction dma_dir = + rq_data_dir(rq) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; + + if (iod->nsge) + dma_unmap_sg(hdev->dev, iod->sg, iod->nsge, dma_dir); + + spraid_free_iod_res(hdev, iod); +} + +static int spraid_bsg_map_data(struct spraid_dev *hdev, struct bsg_job *job, + struct spraid_admin_command *cmd) +{ + struct request *rq = blk_mq_rq_from_pdu(job); + struct spraid_iod *iod = job->dd_data; + enum dma_data_direction dma_dir = + rq_data_dir(rq) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; + int ret = 0; + + iod->sg = job->request_payload.sg_list; + iod->nsge = job->request_payload.sg_cnt; + iod->length = job->request_payload.payload_len; + iod->use_sgl = false; + iod->npages = -1; + iod->sg_drv_mgmt = false; + + if (!iod->nsge) + goto out; + + ret = dma_map_sg_attrs(hdev->dev, iod->sg, iod->nsge, + dma_dir, DMA_ATTR_NO_WARN); + if (!ret) + goto out; + + ret = spraid_setup_prps(hdev, iod); + if (ret) + goto unmap; + + cmd->common.dptr.prp1 = cpu_to_le64(sg_dma_address(iod->sg)); + cmd->common.dptr.prp2 = cpu_to_le64(iod->first_dma); + + return 0; + +unmap: + dma_unmap_sg(hdev->dev, iod->sg, iod->nsge, dma_dir); +out: + return ret; +} + +static int spraid_get_ctrl_info(struct spraid_dev *hdev, + struct spraid_ctrl_info *ctrl_info) +{ + struct spraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + int ret; + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, + &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.get_info.opcode = SPRAID_ADMIN_GET_INFO; + admin_cmd.get_info.type = SPRAID_GET_INFO_CTRL; + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + if (!ret) + memcpy(ctrl_info, data_ptr, sizeof(struct spraid_ctrl_info)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, data_dma); + + return ret; +} + +static int spraid_init_ctrl_info(struct spraid_dev *hdev) +{ + int ret; + + hdev->ctrl_info->nd = cpu_to_le32(240); + hdev->ctrl_info->mdts = 8; + hdev->ctrl_info->max_cmds = cpu_to_le16(4096); + hdev->ctrl_info->max_num_sge = cpu_to_le16(128); + hdev->ctrl_info->max_channel = cpu_to_le16(4); + hdev->ctrl_info->max_tgt_id = cpu_to_le32(3239); + hdev->ctrl_info->max_lun = cpu_to_le16(2); + + ret = spraid_get_ctrl_info(hdev, hdev->ctrl_info); + if (ret) + dev_err(hdev->dev, "get controller info failed: %d\n", ret); + + dev_info(hdev->dev, "[%s]nd = %d\n", __func__, hdev->ctrl_info->nd); + dev_info(hdev->dev, "[%s]max_cmd = %d\n", + __func__, hdev->ctrl_info->max_cmds); + dev_info(hdev->dev, "[%s]max_channel = %d\n", + __func__, hdev->ctrl_info->max_channel); + dev_info(hdev->dev, "[%s]max_tgt_id = %d\n", + __func__, hdev->ctrl_info->max_tgt_id); + dev_info(hdev->dev, "[%s]max_lun = %d\n", + __func__, hdev->ctrl_info->max_lun); + dev_info(hdev->dev, "[%s]max_num_sge = %d\n", + __func__, hdev->ctrl_info->max_num_sge); + dev_info(hdev->dev, "[%s]lun_num_boot = %d\n", + __func__, hdev->ctrl_info->lun_num_in_boot); + dev_info(hdev->dev, "[%s]mdts = %d\n", __func__, hdev->ctrl_info->mdts); + dev_info(hdev->dev, "[%s]acl = %d\n", __func__, hdev->ctrl_info->acl); + dev_info(hdev->dev, "[%s]aer1 = %d\n", __func__, hdev->ctrl_info->aerl); + dev_info(hdev->dev, "[%s]card_type = %d\n", + __func__, hdev->ctrl_info->card_type); + dev_info(hdev->dev, "[%s]rtd3e = %d\n", + __func__, hdev->ctrl_info->rtd3e); + dev_info(hdev->dev, "[%s]sn = %s\n", __func__, hdev->ctrl_info->sn); + dev_info(hdev->dev, "[%s]fr = %s\n", __func__, hdev->ctrl_info->fr); + + if (!hdev->ctrl_info->aerl) + hdev->ctrl_info->aerl = 1; + if (hdev->ctrl_info->aerl > SPRAID_NR_AEN_COMMANDS) + hdev->ctrl_info->aerl = SPRAID_NR_AEN_COMMANDS; + + return 0; +} + +#define SPRAID_MAX_ADMIN_PAYLOAD_SIZE BIT(16) +static int spraid_alloc_iod_ext_mem_pool(struct spraid_dev *hdev) +{ + u16 max_sge = le16_to_cpu(hdev->ctrl_info->max_num_sge); + size_t alloc_size; + + alloc_size = spraid_iod_ext_size(hdev, SPRAID_MAX_ADMIN_PAYLOAD_SIZE, + max_sge, true, false); + if (alloc_size > PAGE_SIZE) + dev_warn(hdev->dev, "It is unreasonable ;" + " sg allocation more than one page\n"); + hdev->iod_mempool = mempool_create_node(1, mempool_kmalloc, + mempool_kfree, + (void *)alloc_size, GFP_KERNEL, + hdev->numa_node); + if (!hdev->iod_mempool) { + dev_err(hdev->dev, "Create iod extension memory pool failed\n"); + return -ENOMEM; + } + + return 0; +} + +static void spraid_free_iod_ext_mem_pool(struct spraid_dev *hdev) +{ + mempool_destroy(hdev->iod_mempool); +} + +static int spraid_user_admin_cmd(struct spraid_dev *hdev, struct bsg_job *job) +{ + struct spraid_bsg_request *bsg_req = job->request; + struct spraid_passthru_common_cmd *cmd = &(bsg_req->admcmd); + struct spraid_admin_command admin_cmd; + u32 timeout = msecs_to_jiffies(cmd->timeout_ms); + u32 result[2] = {0}; + int status; + + if (hdev->state >= SPRAID_RESETTING) { + dev_err(hdev->dev, "[%s] err, host state:[%d] is not right\n", + __func__, hdev->state); + return -EBUSY; + } + + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x] init\n", + __func__, cmd->opcode, cmd->info_0.subopcode); + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.common.opcode = cmd->opcode; + admin_cmd.common.flags = cmd->flags; + admin_cmd.common.hdid = cpu_to_le32(cmd->nsid); + admin_cmd.common.cdw2[0] = cpu_to_le32(cmd->cdw2); + admin_cmd.common.cdw2[1] = cpu_to_le32(cmd->cdw3); + admin_cmd.common.cdw10 = cpu_to_le32(cmd->cdw10); + admin_cmd.common.cdw11 = cpu_to_le32(cmd->cdw11); + admin_cmd.common.cdw12 = cpu_to_le32(cmd->cdw12); + admin_cmd.common.cdw13 = cpu_to_le32(cmd->cdw13); + admin_cmd.common.cdw14 = cpu_to_le32(cmd->cdw14); + admin_cmd.common.cdw15 = cpu_to_le32(cmd->cdw15); + + status = spraid_bsg_map_data(hdev, job, &admin_cmd); + if (status) { + dev_err(hdev->dev, "[%s] err, map data failed\n", __func__); + return status; + } + + status = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, &result[0], + &result[1], timeout); + if (status >= 0) { + job->reply_len = sizeof(result); + memcpy(job->reply, result, sizeof(result)); + } + + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x];" + " status[0x%x] result0[0x%x] result1[0x%x]\n", + __func__, cmd->opcode, cmd->info_0.subopcode, + status, result[0], result[1]); + + spraid_bsg_unmap_data(hdev, job); + + return status; +} + +static int spraid_alloc_ioq_ptcmds(struct spraid_dev *hdev) +{ + int i; + int ptnum = SPRAID_NR_IOQ_PTCMDS; + + INIT_LIST_HEAD(&hdev->ioq_pt_list); + spin_lock_init(&hdev->ioq_pt_lock); + + hdev->ioq_ptcmds = kcalloc_node(ptnum, sizeof(struct spraid_cmd), + GFP_KERNEL, hdev->numa_node); + + if (!hdev->ioq_ptcmds) { + dev_err(hdev->dev, "Alloc ioq_ptcmds failed\n"); + return -ENOMEM; + } + + for (i = 0; i < ptnum; i++) { + hdev->ioq_ptcmds[i].qid = i / SPRAID_PTCMDS_PERQ + 1; + hdev->ioq_ptcmds[i].cid = i % SPRAID_PTCMDS_PERQ + + SPRAID_IO_BLK_MQ_DEPTH; + list_add_tail(&(hdev->ioq_ptcmds[i].list), &hdev->ioq_pt_list); + } + + dev_info(hdev->dev, "Alloc ioq_ptcmds success, ptnum[%d]\n", ptnum); + + return 0; +} + +static void spraid_free_ioq_ptcmds(struct spraid_dev *hdev) +{ + kfree(hdev->ioq_ptcmds); + hdev->ioq_ptcmds = NULL; + + INIT_LIST_HEAD(&hdev->ioq_pt_list); +} + +static int spraid_submit_ioq_sync_cmd(struct spraid_dev *hdev, + struct spraid_ioq_command *cmd, + u32 *result, u32 *reslen, u32 timeout) +{ + int ret; + dma_addr_t sense_dma; + struct spraid_queue *ioq; + void *sense_addr = NULL; + struct spraid_cmd *pt_cmd = spraid_get_cmd(hdev, SPRAID_CMD_IOPT); + + if (!pt_cmd) { + dev_err(hdev->dev, "err, get ioq cmd failed\n"); + return -EFAULT; + } + + timeout = timeout ? timeout : ADMIN_TIMEOUT; + + init_completion(&pt_cmd->cmd_done); + + ioq = &hdev->queues[pt_cmd->qid]; + ret = pt_cmd->cid * SCSI_SENSE_BUFFERSIZE; + sense_addr = ioq->sense + ret; + sense_dma = ioq->sense_dma_addr + ret; + + cmd->common.sense_addr = cpu_to_le64(sense_dma); + cmd->common.sense_len = cpu_to_le16(SCSI_SENSE_BUFFERSIZE); + cmd->common.command_id = pt_cmd->cid; + + spraid_submit_cmd(ioq, cmd); + + if (!wait_for_completion_timeout(&pt_cmd->cmd_done, timeout)) { + dev_err(hdev->dev, "[%s] cid[%d] qid[%d] timeout;" + " opcode[0x%x] subopcode[0x%x]\n", + __func__, pt_cmd->cid, pt_cmd->qid, cmd->common.opcode, + (le32_to_cpu(cmd->common.cdw3[0]) & 0xffff)); + WRITE_ONCE(pt_cmd->state, SPRAID_CMD_TIMEOUT); + spraid_put_cmd(hdev, pt_cmd, SPRAID_CMD_IOPT); + return -EINVAL; + } + + if (result && reslen) { + if ((pt_cmd->status & 0x17f) == 0x101) { + memcpy(result, sense_addr, SCSI_SENSE_BUFFERSIZE); + *reslen = SCSI_SENSE_BUFFERSIZE; + } + } + + spraid_put_cmd(hdev, pt_cmd, SPRAID_CMD_IOPT); + + return pt_cmd->status; +} + +static int spraid_user_ioq_cmd(struct spraid_dev *hdev, struct bsg_job *job) +{ + struct spraid_bsg_request *bsg_req = + (struct spraid_bsg_request *)(job->request); + struct spraid_ioq_passthru_cmd *cmd = &(bsg_req->ioqcmd); + struct spraid_ioq_command ioq_cmd; + int status = 0; + u32 timeout = msecs_to_jiffies(cmd->timeout_ms); + + if (cmd->data_len > PAGE_SIZE) { + dev_err(hdev->dev, "[%s] data len bigger than 4k\n", __func__); + return -EFAULT; + } + + if (hdev->state != SPRAID_LIVE) { + dev_err(hdev->dev, "[%s] err, host state:[%d] is not live\n", + __func__, hdev->state); + return -EBUSY; + } + + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x] init;" + " datalen[%d]\n", + __func__, cmd->opcode, cmd->info_1.subopcode, cmd->data_len); + + memset(&ioq_cmd, 0, sizeof(ioq_cmd)); + ioq_cmd.common.opcode = cmd->opcode; + ioq_cmd.common.flags = cmd->flags; + ioq_cmd.common.hdid = cpu_to_le32(cmd->nsid); + ioq_cmd.common.sense_len = cpu_to_le16(cmd->info_0.res_sense_len); + ioq_cmd.common.cdb_len = cmd->info_0.cdb_len; + ioq_cmd.common.rsvd2 = cmd->info_0.rsvd0; + ioq_cmd.common.cdw3[0] = cpu_to_le32(cmd->cdw3); + ioq_cmd.common.cdw3[1] = cpu_to_le32(cmd->cdw4); + ioq_cmd.common.cdw3[2] = cpu_to_le32(cmd->cdw5); + + ioq_cmd.common.cdw10[0] = cpu_to_le32(cmd->cdw10); + ioq_cmd.common.cdw10[1] = cpu_to_le32(cmd->cdw11); + ioq_cmd.common.cdw10[2] = cpu_to_le32(cmd->cdw12); + ioq_cmd.common.cdw10[3] = cpu_to_le32(cmd->cdw13); + ioq_cmd.common.cdw10[4] = cpu_to_le32(cmd->cdw14); + ioq_cmd.common.cdw10[5] = cpu_to_le32(cmd->data_len); + + memcpy(ioq_cmd.common.cdb, &cmd->cdw16, cmd->info_0.cdb_len); + + ioq_cmd.common.cdw26[0] = cpu_to_le32(cmd->cdw26[0]); + ioq_cmd.common.cdw26[1] = cpu_to_le32(cmd->cdw26[1]); + ioq_cmd.common.cdw26[2] = cpu_to_le32(cmd->cdw26[2]); + ioq_cmd.common.cdw26[3] = cpu_to_le32(cmd->cdw26[3]); + + status = spraid_bsg_map_data(hdev, job, + (struct spraid_admin_command *)&ioq_cmd); + if (status) { + dev_err(hdev->dev, "[%s] err, map data failed\n", __func__); + return status; + } + + status = spraid_submit_ioq_sync_cmd(hdev, &ioq_cmd, job->reply, + &job->reply_len, timeout); + + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x], status[0x%x];" + " reply_len[%d]\n", + __func__, cmd->opcode, cmd->info_1.subopcode, + status, job->reply_len); + + spraid_bsg_unmap_data(hdev, job); + + return status; +} + +static bool spraid_check_scmd_completed(struct scsi_cmnd *scmd) +{ + struct spraid_dev *hdev = shost_priv(scmd->device->host); + struct spraid_iod *iod = scsi_cmd_priv(scmd); + struct spraid_queue *spraidq; + u16 hwq, cid; + + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + spraidq = &hdev->queues[hwq]; + if (READ_ONCE(iod->state) == SPRAID_CMD_COMPLETE + || spraid_poll_cq(spraidq, cid)) { + dev_warn(hdev->dev, "cid[%d] qid[%d] has been completed\n", + cid, spraidq->qid); + return true; + } + return false; +} + +static enum blk_eh_timer_return spraid_scmd_timeout(struct scsi_cmnd *scmd) +{ + struct spraid_iod *iod = scsi_cmd_priv(scmd); + unsigned int timeout = scmd->device->request_queue->rq_timeout; + + if (spraid_check_scmd_completed(scmd)) + goto out; + + if (time_after(jiffies, scmd->jiffies_at_alloc + timeout)) { + if (cmpxchg(&iod->state, + SPRAID_CMD_IN_FLIGHT, + SPRAID_CMD_TIMEOUT) == SPRAID_CMD_IN_FLIGHT) { + return BLK_EH_DONE; + } + } +out: + return BLK_EH_RESET_TIMER; +} + +/* send abort command by admin queue temporary */ +static int spraid_send_abort_cmd(struct spraid_dev *hdev, + u32 hdid, u16 qid, u16 cid) +{ + struct spraid_admin_command admin_cmd; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.abort.opcode = SPRAID_ADMIN_ABORT_CMD; + admin_cmd.abort.hdid = cpu_to_le32(hdid); + admin_cmd.abort.sqid = cpu_to_le16(qid); + admin_cmd.abort.cid = cpu_to_le16(cid); + + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); +} + +/* send reset command by admin quueue temporary */ +static int spraid_send_reset_cmd(struct spraid_dev *hdev, int type, u32 hdid) +{ + struct spraid_admin_command admin_cmd; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.reset.opcode = SPRAID_ADMIN_RESET; + admin_cmd.reset.hdid = cpu_to_le32(hdid); + admin_cmd.reset.type = type; + + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); +} + +static bool spraid_change_host_state(struct spraid_dev *hdev, + enum spraid_state newstate) +{ + unsigned long flags; + enum spraid_state oldstate; + bool change = false; + + spin_lock_irqsave(&hdev->state_lock, flags); + + oldstate = hdev->state; + switch (newstate) { + case SPRAID_LIVE: + switch (oldstate) { + case SPRAID_NEW: + case SPRAID_RESETTING: + change = true; + break; + default: + break; + } + break; + case SPRAID_RESETTING: + switch (oldstate) { + case SPRAID_LIVE: + change = true; + break; + default: + break; + } + break; + case SPRAID_DELETING: + if (oldstate != SPRAID_DELETING) + change = true; + break; + case SPRAID_DEAD: + switch (oldstate) { + case SPRAID_NEW: + case SPRAID_LIVE: + case SPRAID_RESETTING: + change = true; + break; + default: + break; + } + break; + default: + break; + } + if (change) + hdev->state = newstate; + spin_unlock_irqrestore(&hdev->state_lock, flags); + + dev_info(hdev->dev, "[%s][%d]->[%d], change[%d]\n", + __func__, oldstate, newstate, change); + + return change; +} + +static void spraid_back_fault_cqe(struct spraid_queue *ioq, + struct spraid_completion *cqe) +{ + struct spraid_dev *hdev = ioq->hdev; + struct blk_mq_tags *tags; + struct scsi_cmnd *scmd; + struct spraid_iod *iod; + struct request *req; + + tags = hdev->shost->tag_set.tags[ioq->qid - 1]; + req = blk_mq_tag_to_rq(tags, cqe->cmd_id); + if (unlikely(!req || !blk_mq_request_started(req))) + return; + + scmd = blk_mq_rq_to_pdu(req); + iod = scsi_cmd_priv(scmd); + + set_host_byte(scmd, DID_NO_CONNECT); + if (iod->nsge) + scsi_dma_unmap(scmd); + spraid_free_iod_res(hdev, iod); + scmd->scsi_done(scmd); + dev_warn(hdev->dev, "Back fault CQE, cid[%d] qid[%d]\n", + cqe->cmd_id, ioq->qid); +} + +static void spraid_back_all_io(struct spraid_dev *hdev) +{ + int i, j; + struct spraid_queue *ioq; + struct spraid_completion cqe = { 0 }; + + scsi_block_requests(hdev->shost); + + for (i = 1; i <= hdev->shost->nr_hw_queues; i++) { + ioq = &hdev->queues[i]; + for (j = 0; j < hdev->shost->can_queue; j++) { + cqe.cmd_id = j; + spraid_back_fault_cqe(ioq, &cqe); + } + } + + scsi_unblock_requests(hdev->shost); +} + +static void spraid_dev_disable(struct spraid_dev *hdev, bool shutdown) +{ + struct spraid_queue *adminq = &hdev->queues[0]; + u16 start, end; + unsigned long timeout = jiffies + 600 * HZ; + + if (pci_device_is_present(hdev->pdev)) { + if (shutdown) + spraid_shutdown_ctrl(hdev); + else + spraid_disable_ctrl(hdev); + } + + while (!time_after(jiffies, timeout)) { + if (!pci_device_is_present(hdev->pdev)) { + dev_info(hdev->dev, "[%s] pci_device not present;" + " skip wait\n", __func__); + break; + } + if (!spraid_wait_ready(hdev, hdev->cap, false)) { + dev_info(hdev->dev, + "[%s] wait ready success after reset\n", + __func__); + break; + } + dev_info(hdev->dev, "[%s] waiting csts_rdy ready\n", __func__); + } + + if (hdev->queue_count == 0) { + dev_err(hdev->dev, "[%s] warn, queue has been delete\n", + __func__); + return; + } + + spin_lock_irq(&adminq->cq_lock); + spraid_process_cq(adminq, &start, &end, -1); + spin_unlock_irq(&adminq->cq_lock); + spraid_complete_cqes(adminq, start, end); + + spraid_pci_disable(hdev); + + spraid_back_all_io(hdev); +} + +static void spraid_reset_work(struct work_struct *work) +{ + int ret; + struct spraid_dev *hdev = + container_of(work, struct spraid_dev, reset_work); + + if (hdev->state != SPRAID_RESETTING) { + dev_err(hdev->dev, "[%s] err, host is not reset state\n", + __func__); + return; + } + + dev_info(hdev->dev, "[%s] enter host reset\n", __func__); + + if (hdev->ctrl_config & SPRAID_CC_ENABLE) { + dev_info(hdev->dev, "[%s] start dev_disable\n", __func__); + spraid_dev_disable(hdev, false); + } + + ret = spraid_pci_enable(hdev); + if (ret) + goto out; + + ret = spraid_setup_admin_queue(hdev); + if (ret) + goto pci_disable; + + ret = spraid_setup_io_queues(hdev); + if (ret || hdev->online_queues <= hdev->shost->nr_hw_queues) + goto pci_disable; + + spraid_change_host_state(hdev, SPRAID_LIVE); + + spraid_send_all_aen(hdev); + + return; + +pci_disable: + spraid_pci_disable(hdev); +out: + spraid_change_host_state(hdev, SPRAID_DEAD); + dev_err(hdev->dev, "[%s] err, host reset failed\n", __func__); +} + +static int spraid_reset_work_sync(struct spraid_dev *hdev) +{ + if (!spraid_change_host_state(hdev, SPRAID_RESETTING)) { + dev_info(hdev->dev, "[%s] can't change to reset state\n", + __func__); + return -EBUSY; + } + + if (!queue_work(spraid_wq, &hdev->reset_work)) { + dev_err(hdev->dev, "[%s] err, host is already in reset state\n", + __func__); + return -EBUSY; + } + + flush_work(&hdev->reset_work); + if (hdev->state != SPRAID_LIVE) + return -ENODEV; + + return 0; +} + +static int spraid_wait_abnl_cmd_done(struct spraid_iod *iod) +{ + u16 times = 0; + + do { + if (READ_ONCE(iod->state) == SPRAID_CMD_TMO_COMPLETE) + break; + msleep(500); + times++; + } while (times <= SPRAID_WAIT_ABNL_CMD_TIMEOUT); + + /* wait command completion timeout after abort/reset success */ + if (times >= SPRAID_WAIT_ABNL_CMD_TIMEOUT) + return -ETIMEDOUT; + + return 0; +} + +static int spraid_abort_handler(struct scsi_cmnd *scmd) +{ + struct spraid_dev *hdev = shost_priv(scmd->device->host); + struct spraid_iod *iod = scsi_cmd_priv(scmd); + struct spraid_sdev_hostdata *hostdata; + u16 hwq, cid; + int ret; + + scsi_print_command(scmd); + + if (hdev->state != SPRAID_LIVE || !spraid_wait_abnl_cmd_done(iod) || + spraid_check_scmd_completed(scmd)) + return SUCCESS; + + hostdata = scmd->device->hostdata; + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + + dev_warn(hdev->dev, "cid[%d] qid[%d] timeout, aborting\n", cid, hwq); + ret = spraid_send_abort_cmd(hdev, hostdata->hdid, hwq, cid); + if (ret != ADMIN_ERR_TIMEOUT) { + ret = spraid_wait_abnl_cmd_done(iod); + if (ret) { + dev_warn(hdev->dev, "cid[%d] qid[%d] abort failed;" + " not found\n", cid, hwq); + return FAILED; + } + dev_warn(hdev->dev, "cid[%d] qid[%d] abort succ\n", cid, hwq); + return SUCCESS; + } + dev_warn(hdev->dev, "cid[%d] qid[%d] abort failed, timeout\n", + cid, hwq); + return FAILED; +} + +static int spraid_tgt_reset_handler(struct scsi_cmnd *scmd) +{ + struct spraid_dev *hdev = shost_priv(scmd->device->host); + struct spraid_iod *iod = scsi_cmd_priv(scmd); + struct spraid_sdev_hostdata *hostdata; + u16 hwq, cid; + int ret; + + scsi_print_command(scmd); + + if (hdev->state != SPRAID_LIVE || !spraid_wait_abnl_cmd_done(iod) || + spraid_check_scmd_completed(scmd)) + return SUCCESS; + + hostdata = scmd->device->hostdata; + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + + dev_warn(hdev->dev, "cid[%d] qid[%d] timeout, target reset\n", + cid, hwq); + ret = spraid_send_reset_cmd(hdev, SPRAID_RESET_TARGET, hostdata->hdid); + if (ret == 0) { + ret = spraid_wait_abnl_cmd_done(iod); + if (ret) { + dev_warn(hdev->dev, + "cid[%d] qid[%d]target reset failed;" + " not found\n", cid, hwq); + return FAILED; + } + + dev_warn(hdev->dev, "cid[%d] qid[%d] target reset success\n", + cid, hwq); + return SUCCESS; + } + + dev_warn(hdev->dev, "cid[%d] qid[%d] ret[%d] target reset failed\n", + cid, hwq, ret); + return FAILED; +} + +static int spraid_bus_reset_handler(struct scsi_cmnd *scmd) +{ + struct spraid_dev *hdev = shost_priv(scmd->device->host); + struct spraid_iod *iod = scsi_cmd_priv(scmd); + struct spraid_sdev_hostdata *hostdata; + u16 hwq, cid; + int ret; + + scsi_print_command(scmd); + + if (hdev->state != SPRAID_LIVE || !spraid_wait_abnl_cmd_done(iod) || + spraid_check_scmd_completed(scmd)) + return SUCCESS; + + hostdata = scmd->device->hostdata; + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + + dev_warn(hdev->dev, "cid[%d] qid[%d] timeout, bus reset\n", cid, hwq); + ret = spraid_send_reset_cmd(hdev, SPRAID_RESET_BUS, hostdata->hdid); + if (ret == 0) { + ret = spraid_wait_abnl_cmd_done(iod); + if (ret) { + dev_warn(hdev->dev, + "cid[%d] qid[%d] bus reset failed;" + " not found\n", cid, hwq); + return FAILED; + } + + dev_warn(hdev->dev, "cid[%d] qid[%d] bus reset succ\n", + cid, hwq); + return SUCCESS; + } + + dev_warn(hdev->dev, "cid[%d] qid[%d] ret[%d] bus reset failed\n", + cid, hwq, ret); + return FAILED; +} + +static int spraid_shost_reset_handler(struct scsi_cmnd *scmd) +{ + u16 hwq, cid; + struct spraid_dev *hdev = shost_priv(scmd->device->host); + + scsi_print_command(scmd); + if (hdev->state != SPRAID_LIVE || spraid_check_scmd_completed(scmd)) + return SUCCESS; + + spraid_get_tag_from_scmd(scmd, &hwq, &cid); + dev_warn(hdev->dev, "cid[%d] qid[%d] host reset\n", cid, hwq); + + if (spraid_reset_work_sync(hdev)) { + dev_warn(hdev->dev, "cid[%d] qid[%d] host reset failed\n", + cid, hwq); + return FAILED; + } + + dev_warn(hdev->dev, "cid[%d] qid[%d] host reset success\n", cid, hwq); + + return SUCCESS; +} + +static pci_ers_result_t spraid_pci_error_detected(struct pci_dev *pdev, + pci_channel_state_t state) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "enter pci error detect, state:%d\n", state); + + switch (state) { + case pci_channel_io_normal: + dev_warn(hdev->dev, "channel is normal, do nothing\n"); + + return PCI_ERS_RESULT_CAN_RECOVER; + case pci_channel_io_frozen: + dev_warn(hdev->dev, + "channel io frozen, need reset controller\n"); + + scsi_block_requests(hdev->shost); + + spraid_change_host_state(hdev, SPRAID_RESETTING); + + return PCI_ERS_RESULT_NEED_RESET; + case pci_channel_io_perm_failure: + dev_warn(hdev->dev, "channel io failure, request disconnect\n"); + + return PCI_ERS_RESULT_DISCONNECT; + } + + return PCI_ERS_RESULT_NEED_RESET; +} + +static pci_ers_result_t spraid_pci_slot_reset(struct pci_dev *pdev) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "restart after slot reset\n"); + + pci_restore_state(pdev); + + if (!queue_work(spraid_wq, &hdev->reset_work)) { + dev_err(hdev->dev, "[%s] err, the device is resetting state\n", + __func__); + return PCI_ERS_RESULT_NONE; + } + + flush_work(&hdev->reset_work); + + scsi_unblock_requests(hdev->shost); + + return PCI_ERS_RESULT_RECOVERED; +} + +static void spraid_reset_done(struct pci_dev *pdev) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "enter spraid reset done\n"); +} + +static ssize_t csts_pp_show(struct device *cdev, + struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct spraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) { + ret = (readl(hdev->bar + SPRAID_REG_CSTS) + & SPRAID_CSTS_PP_MASK); + ret >>= SPRAID_CSTS_PP_SHIFT; + } + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t csts_shst_show(struct device *cdev, + struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct spraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) { + ret = (readl(hdev->bar + SPRAID_REG_CSTS) + & SPRAID_CSTS_SHST_MASK); + ret >>= SPRAID_CSTS_SHST_SHIFT; + } + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t csts_cfs_show(struct device *cdev, + struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct spraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) { + ret = (readl(hdev->bar + SPRAID_REG_CSTS) + & SPRAID_CSTS_CFS_MASK); + ret >>= SPRAID_CSTS_CFS_SHIFT; + } + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t csts_rdy_show(struct device *cdev, + struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct spraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) + ret = (readl(hdev->bar + SPRAID_REG_CSTS) & SPRAID_CSTS_RDY); + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t fw_version_show(struct device *cdev, + struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct spraid_dev *hdev = shost_priv(shost); + + return snprintf(buf, PAGE_SIZE, "%s\n", hdev->ctrl_info->fr); +} + +static DEVICE_ATTR_RO(csts_pp); +static DEVICE_ATTR_RO(csts_shst); +static DEVICE_ATTR_RO(csts_cfs); +static DEVICE_ATTR_RO(csts_rdy); +static DEVICE_ATTR_RO(fw_version); + +static struct device_attribute *spraid_host_attrs[] = { + &dev_attr_csts_pp, + &dev_attr_csts_shst, + &dev_attr_csts_cfs, + &dev_attr_csts_rdy, + &dev_attr_fw_version, + NULL, +}; + +static int spraid_get_vd_info(struct spraid_dev *hdev, + struct spraid_vd_info *vd_info, u16 vid) +{ + struct spraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + int ret; + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, + &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.usr_cmd.opcode = USR_CMD_READ; + admin_cmd.usr_cmd.info_0.subopcode = cpu_to_le16(USR_CMD_VDINFO); + admin_cmd.usr_cmd.info_1.data_len = cpu_to_le16(USR_CMD_RDLEN); + admin_cmd.usr_cmd.info_1.param_len = cpu_to_le16(VDINFO_PARAM_LEN); + admin_cmd.usr_cmd.cdw10 = cpu_to_le32(vid); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + if (!ret) + memcpy(vd_info, data_ptr, sizeof(struct spraid_vd_info)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, data_dma); + + return ret; +} + +static int spraid_get_bgtask(struct spraid_dev *hdev, + struct spraid_bgtask *bgtask) +{ + struct spraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + int ret; + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, + &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.usr_cmd.opcode = USR_CMD_READ; + admin_cmd.usr_cmd.info_0.subopcode = cpu_to_le16(USR_CMD_BGTASK); + admin_cmd.usr_cmd.info_1.data_len = cpu_to_le16(USR_CMD_RDLEN); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + if (!ret) + memcpy(bgtask, data_ptr, sizeof(struct spraid_bgtask)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, data_dma); + + return ret; +} + +static ssize_t raid_level_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct spraid_dev *hdev; + struct spraid_vd_info *vd_info; + struct spraid_sdev_hostdata *hostdata; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !SPRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = spraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) + vd_info->rg_level = ARRAY_SIZE(raid_levels) - 1; + + ret = (vd_info->rg_level < ARRAY_SIZE(raid_levels)) ? + vd_info->rg_level : (ARRAY_SIZE(raid_levels) - 1); + + kfree(vd_info); + + return snprintf(buf, PAGE_SIZE, "RAID-%s\n", raid_levels[ret]); +} + +static ssize_t raid_state_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct spraid_dev *hdev; + struct spraid_vd_info *vd_info; + struct spraid_sdev_hostdata *hostdata; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !SPRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = spraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) { + vd_info->vd_status = 0; + vd_info->rg_id = 0xff; + } + + ret = (vd_info->vd_status < ARRAY_SIZE(raid_states)) ? + vd_info->vd_status : 0; + + kfree(vd_info); + + return snprintf(buf, PAGE_SIZE, "%s\n", raid_states[ret]); +} + +static ssize_t raid_resync_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct spraid_dev *hdev; + struct spraid_vd_info *vd_info; + struct spraid_bgtask *bgtask; + struct spraid_sdev_hostdata *hostdata; + u8 rg_id, i, progress = 0; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !SPRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = spraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) + goto out; + + rg_id = vd_info->rg_id; + + bgtask = (struct spraid_bgtask *)vd_info; + ret = spraid_get_bgtask(hdev, bgtask); + if (ret) + goto out; + for (i = 0; i < bgtask->task_num; i++) { + if ((bgtask->bgtask[i].type == BGTASK_TYPE_REBUILD) && + (le16_to_cpu(bgtask->bgtask[i].vd_id) == rg_id)) + progress = bgtask->bgtask[i].progress; + } + +out: + kfree(vd_info); + return snprintf(buf, PAGE_SIZE, "%d\n", progress); +} + +static DEVICE_ATTR_RO(raid_level); +static DEVICE_ATTR_RO(raid_state); +static DEVICE_ATTR_RO(raid_resync); + +static struct device_attribute *spraid_dev_attrs[] = { + &dev_attr_raid_level, + &dev_attr_raid_state, + &dev_attr_raid_resync, + NULL, +}; + +static struct pci_error_handlers spraid_err_handler = { + .error_detected = spraid_pci_error_detected, + .slot_reset = spraid_pci_slot_reset, + .reset_done = spraid_reset_done, +}; + +static int spraid_sysfs_host_reset(struct Scsi_Host *shost, int reset_type) +{ + int ret; + struct spraid_dev *hdev = shost_priv(shost); + + dev_info(hdev->dev, "[%s] start sysfs host reset cmd\n", __func__); + ret = spraid_reset_work_sync(hdev); + dev_info(hdev->dev, "[%s] stop sysfs host reset cmd[%d]\n", + __func__, ret); + + return ret; +} + +static struct scsi_host_template spraid_driver_template = { + .module = THIS_MODULE, + .name = "Ramaxel Logic spraid driver", + .proc_name = "spraid", + .queuecommand = spraid_queue_command, + .slave_alloc = spraid_slave_alloc, + .slave_destroy = spraid_slave_destroy, + .slave_configure = spraid_slave_configure, + .eh_timed_out = spraid_scmd_timeout, + .eh_abort_handler = spraid_abort_handler, + .eh_target_reset_handler = spraid_tgt_reset_handler, + .eh_bus_reset_handler = spraid_bus_reset_handler, + .eh_host_reset_handler = spraid_shost_reset_handler, + .change_queue_depth = scsi_change_queue_depth, + .this_id = -1, + .shost_attrs = spraid_host_attrs, + .sdev_attrs = spraid_dev_attrs, + .host_reset = spraid_sysfs_host_reset, +}; + +static void spraid_shutdown(struct pci_dev *pdev) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + + spraid_remove_io_queues(hdev); + spraid_disable_admin_queue(hdev, true); +} + +/* bsg dispatch user command */ +static int spraid_bsg_host_dispatch(struct bsg_job *job) +{ + struct Scsi_Host *shost = dev_to_shost(job->dev); + struct spraid_dev *hdev = shost_priv(shost); + struct request *rq = blk_mq_rq_from_pdu(job); + struct spraid_bsg_request *bsg_req = job->request; + int ret = 0; + + dev_info(hdev->dev, "[%s] msgcode[%d], msglen[%d], timeout[%d];" + " req_nsge[%d], req_len[%d]\n", + __func__, bsg_req->msgcode, job->request_len, + rq->timeout, job->request_payload.sg_cnt, + job->request_payload.payload_len); + + job->reply_len = 0; + + switch (bsg_req->msgcode) { + case SPRAID_BSG_ADM: + ret = spraid_user_admin_cmd(hdev, job); + break; + case SPRAID_BSG_IOQ: + ret = spraid_user_ioq_cmd(hdev, job); + break; + default: + dev_info(hdev->dev, "[%s] unsupport msgcode[%d]\n", + __func__, bsg_req->msgcode); + break; + } + + if (ret > 0) + ret = ret | (ret << 8); + + bsg_job_done(job, ret, 0); + return 0; +} + +static inline void spraid_remove_bsg(struct spraid_dev *hdev) +{ + if (hdev->bsg_queue) { + bsg_unregister_queue(hdev->bsg_queue); + blk_cleanup_queue(hdev->bsg_queue); + } +} +static int spraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) +{ + struct spraid_dev *hdev; + struct Scsi_Host *shost; + int node, ret; + char bsg_name[15]; + + shost = scsi_host_alloc(&spraid_driver_template, sizeof(*hdev)); + if (!shost) { + dev_err(&pdev->dev, "Failed to allocate scsi host\n"); + return -ENOMEM; + } + hdev = shost_priv(shost); + hdev->pdev = pdev; + hdev->dev = get_device(&pdev->dev); + + node = dev_to_node(hdev->dev); + if (node == NUMA_NO_NODE) { + node = first_memory_node; + set_dev_node(hdev->dev, node); + } + hdev->numa_node = node; + hdev->shost = shost; + pci_set_drvdata(pdev, hdev); + + ret = spraid_dev_map(hdev); + if (ret) + goto put_dev; + + init_rwsem(&hdev->devices_rwsem); + INIT_WORK(&hdev->scan_work, spraid_scan_work); + INIT_WORK(&hdev->timesyn_work, spraid_timesyn_work); + INIT_WORK(&hdev->reset_work, spraid_reset_work); + INIT_WORK(&hdev->fw_act_work, spraid_fw_act_work); + spin_lock_init(&hdev->state_lock); + + ret = spraid_alloc_resources(hdev); + if (ret) + goto dev_unmap; + + ret = spraid_pci_enable(hdev); + if (ret) + goto resources_free; + + ret = spraid_setup_admin_queue(hdev); + if (ret) + goto pci_disable; + + ret = spraid_init_ctrl_info(hdev); + if (ret) + goto disable_admin_q; + + ret = spraid_alloc_iod_ext_mem_pool(hdev); + if (ret) + goto disable_admin_q; + + ret = spraid_setup_io_queues(hdev); + if (ret) + goto free_iod_mempool; + + spraid_shost_init(hdev); + + ret = scsi_add_host(hdev->shost, hdev->dev); + if (ret) { + dev_err(hdev->dev, "Add shost to system failed, ret: %d\n", + ret); + goto remove_io_queues; + } + + snprintf(bsg_name, sizeof(bsg_name), "spraid%d", shost->host_no); + hdev->bsg_queue = bsg_setup_queue(&shost->shost_gendev, bsg_name, + spraid_bsg_host_dispatch, + spraid_cmd_size(hdev, true, false)); + if (IS_ERR(hdev->bsg_queue)) { + dev_err(hdev->dev, "err, setup bsg failed\n"); + hdev->bsg_queue = NULL; + goto remove_io_queues; + } + + if (hdev->online_queues == SPRAID_ADMIN_QUEUE_NUM) { + dev_warn(hdev->dev, "warn only admin queue can be used\n"); + return 0; + } + + hdev->state = SPRAID_LIVE; + + spraid_send_all_aen(hdev); + + ret = spraid_dev_list_init(hdev); + if (ret) + goto remove_bsg; + + ret = spraid_configure_timestamp(hdev); + if (ret) + dev_warn(hdev->dev, "init set timestamp failed\n"); + + ret = spraid_alloc_ioq_ptcmds(hdev); + if (ret) + goto remove_bsg; + + scsi_scan_host(hdev->shost); + + return 0; + +remove_bsg: + spraid_remove_bsg(hdev); +remove_io_queues: + spraid_remove_io_queues(hdev); +free_iod_mempool: + spraid_free_iod_ext_mem_pool(hdev); +disable_admin_q: + spraid_disable_admin_queue(hdev, false); +pci_disable: + spraid_pci_disable(hdev); +resources_free: + spraid_free_resources(hdev); +dev_unmap: + spraid_dev_unmap(hdev); +put_dev: + put_device(hdev->dev); + scsi_host_put(shost); + + return -ENODEV; +} + +static void spraid_remove(struct pci_dev *pdev) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + struct Scsi_Host *shost = hdev->shost; + + dev_info(hdev->dev, "enter spraid remove\n"); + + spraid_change_host_state(hdev, SPRAID_DELETING); + flush_work(&hdev->reset_work); + + if (!pci_device_is_present(pdev)) + spraid_back_all_io(hdev); + + spraid_remove_bsg(hdev); + scsi_remove_host(shost); + spraid_free_ioq_ptcmds(hdev); + kfree(hdev->devices); + spraid_remove_io_queues(hdev); + spraid_free_iod_ext_mem_pool(hdev); + spraid_disable_admin_queue(hdev, false); + spraid_pci_disable(hdev); + spraid_free_resources(hdev); + spraid_dev_unmap(hdev); + put_device(hdev->dev); + scsi_host_put(shost); + + dev_info(hdev->dev, "exit spraid remove\n"); +} + +static const struct pci_device_id spraid_id_table[] = { + { PCI_DEVICE(PCI_VENDOR_ID_RAMAXEL_LOGIC, + SPRAID_SERVER_DEVICE_HBA_DID) }, + { PCI_DEVICE(PCI_VENDOR_ID_RAMAXEL_LOGIC, + SPRAID_SERVER_DEVICE_RAID_DID) }, + { 0, } +}; +MODULE_DEVICE_TABLE(pci, spraid_id_table); + +static struct pci_driver spraid_driver = { + .name = "spraid", + .id_table = spraid_id_table, + .probe = spraid_probe, + .remove = spraid_remove, + .shutdown = spraid_shutdown, + .err_handler = &spraid_err_handler, +}; + +static int __init spraid_init(void) +{ + int ret; + + spraid_wq = alloc_workqueue("spraid-wq", WQ_UNBOUND | WQ_MEM_RECLAIM | + WQ_SYSFS, 0); + if (!spraid_wq) + return -ENOMEM; + + spraid_class = class_create(THIS_MODULE, "spraid"); + if (IS_ERR(spraid_class)) { + ret = PTR_ERR(spraid_class); + goto destroy_wq; + } + + ret = pci_register_driver(&spraid_driver); + if (ret < 0) + goto destroy_class; + + return 0; + +destroy_class: + class_destroy(spraid_class); +destroy_wq: + destroy_workqueue(spraid_wq); + + return ret; +} + +static void __exit spraid_exit(void) +{ + pci_unregister_driver(&spraid_driver); + class_destroy(spraid_class); + destroy_workqueue(spraid_wq); + ida_destroy(&spraid_instance_ida); +} + +MODULE_AUTHOR("songyl(a)ramaxel.com"); +MODULE_DESCRIPTION("Ramaxel Memory Technology SPraid Driver"); +MODULE_LICENSE("GPL"); +MODULE_VERSION(SPRAID_DRV_VERSION); +module_init(spraid_init); +module_exit(spraid_exit); -- 2.27.0

1 0

[PATCH openEuler-1.0-LTS 1/2] blk-mq: fix abnormal free in single queue process
by Yang Yingliang 20 Dec '21

20 Dec '21

From: Zhang Wensheng <zhangwensheng5(a)huawei.com> hulk inclusion category: bugfix bugzilla: 185922, https://gitee.com/openeuler/kernel/issues/I4NDPF CVE: NA -------------------------------- fix the problem introduced by commit 80086651acbd ("blk: Fix lock inversion between ioc lock and bfqd lock"), which lead to blk_queue_bio using blk_mq_free_request to free abnormally. fix the problem using __blk_put_request to replace it. Fixes: 80086651acbdb ("blk: Fix lock inversion between ioc lock and bfqd lock") Signed-off-by: Zhang Wensheng <zhangwensheng5(a)huawei.com> Reviewed-by: Laibin Qiu <qiulaibin(a)huawei.com> Reviewed-by: Jason Yan <yanaijie(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- block/blk-mq.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/block/blk-mq.h b/block/blk-mq.h index bb8aabbc5c579..b3540d62c541e 100644 --- a/block/blk-mq.h +++ b/block/blk-mq.h @@ -228,7 +228,7 @@ static inline void blk_mq_free_requests(struct list_head *list) struct request *rq = list_entry_rq(list->next); list_del_init(&rq->queuelist); - blk_mq_free_request(rq); + __blk_put_request(rq->q, rq); } } -- 2.25.1

1 1

[PATCH openEuler-1.0-LTS] scsi: hisi_sas: Add support for sata disk I/O errors report to libsas
by Yang Yingliang 20 Dec '21

20 Dec '21

From: Xingui Yang <yangxingui(a)huawei.com> driver inclusion category: feature bugzilla: NA CVE: NA --------------------------- If the response frame has been written back to the memory and carries the disk error and status when sata I/O completion abnormally, then set task stat to SAS_PROTO_RESPONSE and let libsas to handle it. Signed-off-by: Xingui Yang <yangxingui(a)huawei.com> Reviewed-by: Kangfenglong <kangfenglong(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/scsi/hisi_sas/hisi_sas_v3_hw.c | 17 +++++++++++++---- 1 file changed, 13 insertions(+), 4 deletions(-) diff --git a/drivers/scsi/hisi_sas/hisi_sas_v3_hw.c b/drivers/scsi/hisi_sas/hisi_sas_v3_hw.c index 4508c4a2f02fc..3c662fa5f6c0c 100644 --- a/drivers/scsi/hisi_sas/hisi_sas_v3_hw.c +++ b/drivers/scsi/hisi_sas/hisi_sas_v3_hw.c @@ -470,6 +470,9 @@ struct hisi_sas_err_record_v3 { #define RX_DATA_LEN_UNDERFLOW_OFF 6 #define RX_DATA_LEN_UNDERFLOW_MSK (1 << RX_DATA_LEN_UNDERFLOW_OFF) +#define RX_FIS_STATUS_ERR_OFF 0 +#define RX_FIS_STATUS_ERR_MSK (1 << RX_FIS_STATUS_ERR_OFF) + #define HISI_SAS_COMMAND_ENTRIES_V3_HW 4096 #define HISI_SAS_MSI_COUNT_V3_HW 32 @@ -2241,17 +2244,20 @@ slot_err_v3_hw(struct hisi_hba *hisi_hba, struct sas_task *task, hisi_sas_status_buf_addr_mem(slot); u32 dma_rx_err_type = record->dma_rx_err_type; u32 trans_tx_fail_type = record->trans_tx_fail_type; + u16 sipc_rx_err_type = le16_to_cpu(record->sipc_rx_err_type); + u32 dw0 = le32_to_cpu(complete_hdr->dw0); + u32 dw3 = le32_to_cpu(complete_hdr->dw3); switch (task->task_proto) { case SAS_PROTOCOL_SSP: if (dma_rx_err_type & RX_DATA_LEN_UNDERFLOW_MSK) { ts->residual = trans_tx_fail_type; ts->stat = SAS_DATA_UNDERRUN; - if ((!(complete_hdr->dw0 & CMPLT_HDR_RSPNS_GOOD_MSK)) && - (complete_hdr->dw0 & CMPLT_HDR_RSPNS_XFRD_MSK)) { + if (!(dw0 & CMPLT_HDR_RSPNS_GOOD_MSK) && + (dw0 & CMPLT_HDR_RSPNS_XFRD_MSK)) { hisi_sas_set_sense_data(task, slot); } - } else if (complete_hdr->dw3 & CMPLT_HDR_IO_IN_TARGET_MSK) { + } else if (dw3 & CMPLT_HDR_IO_IN_TARGET_MSK) { ts->stat = SAS_QUEUE_FULL; slot->abort = 1; } else { @@ -2262,7 +2268,10 @@ slot_err_v3_hw(struct hisi_hba *hisi_hba, struct sas_task *task, case SAS_PROTOCOL_SATA: case SAS_PROTOCOL_STP: case SAS_PROTOCOL_SATA | SAS_PROTOCOL_STP: - if (complete_hdr->dw3 & CMPLT_HDR_IO_IN_TARGET_MSK) { + if ((dw0 & CMPLT_HDR_RSPNS_XFRD_MSK) && + (sipc_rx_err_type & RX_FIS_STATUS_ERR_MSK)) { + ts->stat = SAS_PROTO_RESPONSE; + } else if (dw3 & CMPLT_HDR_IO_IN_TARGET_MSK) { ts->stat = SAS_PHY_DOWN; slot->abort = 1; } else if (dma_rx_err_type & RX_DATA_LEN_UNDERFLOW_MSK) { -- 2.25.1

1 0

[PATCH OLK5.10 v2 ] scsi:spraid: use bsg module to replace with ioctrl
by Yanling Song 20 Dec '21

20 Dec '21

Ramaxel inclusion category: features bugzilla: https://gitee.com/openeuler/kernel/issues/I4JXCG CVE: NA Changes from v1: 1. Add more debug infor. 2. Remove some unnecessary module parameters. 3. Report disks by the order of channel/target id. 4. Add host_reset handler. 5. Use get_unaligned_be24. Signed-off-by: Yanling Song <songyl(a)ramaxel.com> Reviewed-by: Jiang Yu<yujiang(a)ramaxel.com> --- drivers/scsi/spraid/Kconfig | 6 +- drivers/scsi/spraid/spraid.h | 142 ++- drivers/scsi/spraid/spraid_main.c | 1571 +++++++++++++++-------------- 3 files changed, 938 insertions(+), 781 deletions(-) diff --git a/drivers/scsi/spraid/Kconfig b/drivers/scsi/spraid/Kconfig index 83962efaab07..bfbba3db8db0 100644 --- a/drivers/scsi/spraid/Kconfig +++ b/drivers/scsi/spraid/Kconfig @@ -5,7 +5,9 @@ config RAMAXEL_SPRAID tristate "Ramaxel spraid Adapter" depends on PCI && SCSI + select BLK_DEV_BSGLIB depends on ARM64 || X86_64 - default m help - This driver supports Ramaxel spraid driver. + This driver supports Ramaxel SPRxxx serial + raid controller, which has PCIE Gen4 interface + with host and supports SAS/SATA Hdd/ssd. diff --git a/drivers/scsi/spraid/spraid.h b/drivers/scsi/spraid/spraid.h index da46d8e1b4b6..983d7af2faa8 100644 --- a/drivers/scsi/spraid/spraid.h +++ b/drivers/scsi/spraid/spraid.h @@ -1,4 +1,5 @@ /* SPDX-License-Identifier: GPL-2.0 */ +/* Copyright(c) 2021 Ramaxel Memory Technology, Ltd */ #ifndef __SPRAID_H_ #define __SPRAID_H_ @@ -24,7 +25,7 @@ #define SENSE_SIZE(depth) ((depth) * SCSI_SENSE_BUFFERSIZE) #define SPRAID_AQ_DEPTH 128 -#define SPRAID_NR_AEN_COMMANDS 1 +#define SPRAID_NR_AEN_COMMANDS 16 #define SPRAID_AQ_BLK_MQ_DEPTH (SPRAID_AQ_DEPTH - SPRAID_NR_AEN_COMMANDS) #define SPRAID_AQ_MQ_TAG_DEPTH (SPRAID_AQ_BLK_MQ_DEPTH - 1) @@ -44,7 +45,7 @@ #define SMALL_POOL_SIZE 256 #define MAX_SMALL_POOL_NUM 16 -#define MAX_CMD_PER_DEV 32 +#define MAX_CMD_PER_DEV 64 #define MAX_CDB_LEN 32 #define SPRAID_UP_TO_MULTY4(x) (((x) + 4) & (~0x03)) @@ -53,7 +54,7 @@ #define PCI_VENDOR_ID_RAMAXEL_LOGIC 0x1E81 -#define SPRAID_SERVER_DEVICE_HAB_DID 0x2100 +#define SPRAID_SERVER_DEVICE_HBA_DID 0x2100 #define SPRAID_SERVER_DEVICE_RAID_DID 0x2200 #define IO_6_DEFAULT_TX_LEN 256 @@ -142,11 +143,15 @@ enum { enum { SPRAID_AEN_DEV_CHANGED = 0x00, + SPRAID_AEN_FW_ACT_START = 0x01, SPRAID_AEN_HOST_PROBING = 0x10, }; enum { - SPRAID_AEN_TIMESYN = 0x07 + SPRAID_AEN_TIMESYN = 0x00, + SPRAID_AEN_FW_ACT_FINISH = 0x02, + SPRAID_AEN_EVENT_MIN = 0x80, + SPRAID_AEN_EVENT_MAX = 0xff, }; enum { @@ -175,6 +180,16 @@ enum spraid_state { SPRAID_DEAD, }; +enum { + SPRAID_CARD_HBA, + SPRAID_CARD_RAID, +}; + +enum spraid_cmd_type { + SPRAID_CMD_ADM, + SPRAID_CMD_IOPT, +}; + struct spraid_completion { __le32 result; union { @@ -217,8 +232,6 @@ struct spraid_dev { struct dma_pool *prp_page_pool; struct dma_pool *prp_small_pool[MAX_SMALL_POOL_NUM]; mempool_t *iod_mempool; - struct blk_mq_tag_set admin_tagset; - struct request_queue *admin_q; void __iomem *bar; u32 max_qid; u32 num_vecs; @@ -232,23 +245,27 @@ struct spraid_dev { u32 ctrl_config; u32 online_queues; u64 cap; - struct device ctrl_device; - struct cdev cdev; int instance; struct spraid_ctrl_info *ctrl_info; struct spraid_dev_info *devices; - struct spraid_ioq_ptcmd *ioq_ptcmds; + struct spraid_cmd *adm_cmds; + struct list_head adm_cmd_list; + spinlock_t adm_cmd_lock; + + struct spraid_cmd *ioq_ptcmds; struct list_head ioq_pt_list; spinlock_t ioq_pt_lock; - struct work_struct aen_work; struct work_struct scan_work; struct work_struct timesyn_work; struct work_struct reset_work; + struct work_struct fw_act_work; enum spraid_state state; spinlock_t state_lock; + + struct request_queue *bsg_queue; }; struct spraid_sgl_desc { @@ -347,6 +364,35 @@ struct spraid_get_info { __u32 rsvd12[4]; }; +struct spraid_usr_cmd { + __u8 opcode; + __u8 flags; + __u16 command_id; + __le32 hdid; + union { + struct { + __le16 subopcode; + __le16 rsvd1; + } info_0; + __le32 cdw2; + }; + union { + struct { + __le16 data_len; + __le16 param_len; + } info_1; + __le32 cdw3; + }; + __u64 metadata; + union spraid_data_ptr dptr; + __le32 cdw10; + __le32 cdw11; + __le32 cdw12; + __le32 cdw13; + __le32 cdw14; + __le32 cdw15; +}; + enum { SPRAID_CMD_FLAG_SGL_METABUF = (1 << 6), SPRAID_CMD_FLAG_SGL_METASEG = (1 << 7), @@ -393,6 +439,7 @@ struct spraid_admin_command { struct spraid_get_info get_info; struct spraid_abort_cmd abort; struct spraid_reset_cmd reset; + struct spraid_usr_cmd usr_cmd; }; }; @@ -456,9 +503,6 @@ struct spraid_ioq_command { }; }; -#define SPRAID_IOCTL_RESET_CMD _IOWR('N', 0x80, struct spraid_passthru_common_cmd) -#define SPRAID_IOCTL_ADMIN_CMD _IOWR('N', 0x41, struct spraid_passthru_common_cmd) - struct spraid_passthru_common_cmd { __u8 opcode; __u8 flags; @@ -494,8 +538,6 @@ struct spraid_passthru_common_cmd { __u32 result1; }; -#define SPRAID_IOCTL_IOQ_CMD _IOWR('N', 0x42, struct spraid_ioq_passthru_cmd) - struct spraid_ioq_passthru_cmd { __u8 opcode; __u8 flags; @@ -560,7 +602,21 @@ struct spraid_ioq_passthru_cmd { __u32 result1; }; -struct spraid_ioq_ptcmd { +struct spraid_bsg_request { + u32 msgcode; + u32 control; + union { + struct spraid_passthru_common_cmd admcmd; + struct spraid_ioq_passthru_cmd ioqcmd; + }; +}; + +enum { + SPRAID_BSG_ADM, + SPRAID_BSG_IOQ, +}; + +struct spraid_cmd { int qid; int cid; u32 result0; @@ -572,14 +628,6 @@ struct spraid_ioq_ptcmd { struct list_head list; }; -struct spraid_admin_request { - struct spraid_admin_command *cmd; - u32 result0; - u32 result1; - u16 flags; - u16 status; -}; - struct spraid_queue { struct spraid_dev *hdev; spinlock_t sq_lock; /* spinlock for lock handling */ @@ -607,7 +655,6 @@ struct spraid_queue { }; struct spraid_iod { - struct spraid_admin_request req; struct spraid_queue *spraidq; enum spraid_cmd_state state; int npages; @@ -623,13 +670,51 @@ struct spraid_iod { }; #define SPRAID_DEV_INFO_ATTR_BOOT(attr) ((attr) & 0x01) -#define SPRAID_DEV_INFO_ATTR_HDD(attr) ((attr) & 0x02) +#define SPRAID_DEV_INFO_ATTR_VD(attr) (((attr) & 0x02) == 0x0) #define SPRAID_DEV_INFO_ATTR_PT(attr) (((attr) & 0x22) == 0x02) #define SPRAID_DEV_INFO_ATTR_RAWDISK(attr) ((attr) & 0x20) #define SPRAID_DEV_INFO_FLAG_VALID(flag) ((flag) & 0x01) #define SPRAID_DEV_INFO_FLAG_CHANGE(flag) ((flag) & 0x02) +#define BGTASK_TYPE_REBUILD 4 +#define USR_CMD_READ 0xc2 +#define USR_CMD_RDLEN 0x1000 +#define USR_CMD_VDINFO 0x704 +#define USR_CMD_BGTASK 0x504 +#define VDINFO_PARAM_LEN 0x04 + +struct spraid_vd_info { + __u8 name[32]; + __le16 id; + __u8 rg_id; + __u8 rg_level; + __u8 sg_num; + __u8 sg_disk_num; + __u8 vd_status; + __u8 vd_type; + __u8 rsvd1[4056]; +}; + +#define MAX_REALTIME_BGTASK_NUM 32 + +struct bgtask_info { + __u8 type; + __u8 progress; + __u8 rate; + __u8 rsvd0; + __le16 vd_id; + __le16 time_left; + __u8 rsvd1[4]; +}; + +struct spraid_bgtask { + __u8 sw; + __u8 task_num; + __u8 rsvd[6]; + struct bgtask_info bgtask[MAX_REALTIME_BGTASK_NUM]; +}; + struct spraid_dev_info { __le32 hdid; __le16 target; @@ -649,6 +734,11 @@ struct spraid_dev_list { struct spraid_sdev_hostdata { u32 hdid; + u16 max_io_kb; + u8 attr; + u8 flag; + u8 rg_id; + u8 rsvd[3]; }; #endif diff --git a/drivers/scsi/spraid/spraid_main.c b/drivers/scsi/spraid/spraid_main.c index a0a75ecb0027..976b7ded336a 100644 --- a/drivers/scsi/spraid/spraid_main.c +++ b/drivers/scsi/spraid/spraid_main.c @@ -1,8 +1,8 @@ // SPDX-License-Identifier: GPL-2.0 -/* - * Linux spraid device driver - * Copyright(c) 2021 Ramaxel Memory Technology, Ltd - */ +/* Copyright(c) 2021 Ramaxel Memory Technology, Ltd */ + +/* Ramaxel Raid SPXXX Series Linux Driver */ + #define pr_fmt(fmt) "spraid: " fmt #include <linux/sched/signal.h> @@ -23,6 +23,9 @@ #include <linux/debugfs.h> #include <linux/io-64-nonatomic-lo-hi.h> #include <linux/blkdev.h> +#include <linux/bsg-lib.h> +#include <asm/unaligned.h> +#include <linux/sort.h> #include <scsi/scsi.h> #include <scsi/scsi_cmnd.h> @@ -31,28 +34,17 @@ #include <scsi/scsi_transport.h> #include <scsi/scsi_dbg.h> + #include "spraid.h" static u32 admin_tmout = 60; module_param(admin_tmout, uint, 0644); MODULE_PARM_DESC(admin_tmout, "admin commands timeout (seconds)"); -static u32 scmd_tmout_pt = 30; -module_param(scmd_tmout_pt, uint, 0644); -MODULE_PARM_DESC(scmd_tmout_pt, "scsi commands timeout for passthrough(seconds)"); - static u32 scmd_tmout_nonpt = 180; module_param(scmd_tmout_nonpt, uint, 0644); MODULE_PARM_DESC(scmd_tmout_nonpt, "scsi commands timeout for rawdisk&raid(seconds)"); -static u32 wait_abl_tmout = 3; -module_param(wait_abl_tmout, uint, 0644); -MODULE_PARM_DESC(wait_abl_tmout, "wait abnormal io timeout(seconds)"); - -static bool use_sgl_force; -module_param(use_sgl_force, bool, 0644); -MODULE_PARM_DESC(use_sgl_force, "force IO use sgl format, default false"); - static int ioq_depth_set(const char *val, const struct kernel_param *kp); static const struct kernel_param_ops ioq_depth_ops = { .set = ioq_depth_set, @@ -106,16 +98,20 @@ static const struct kernel_param_ops small_pool_num_ops = { .get = param_get_byte, }; +/* It was found that the spindlock of a single pool conflicts + * a lot with multiple CPUs.So multiple pools are introduced + * to reduce the conflictions. + */ static unsigned char small_pool_num = 4; module_param_cb(small_pool_num, &small_pool_num_ops, &small_pool_num, 0644); MODULE_PARM_DESC(small_pool_num, "set prp small pool num, default 4, MAX 16"); static void spraid_free_queue(struct spraid_queue *spraidq); static void spraid_handle_aen_notice(struct spraid_dev *hdev, u32 result); -static void spraid_handle_aen_vs(struct spraid_dev *hdev, u32 result); +static void spraid_handle_aen_vs(struct spraid_dev *hdev, u32 result, u32 result1); static DEFINE_IDA(spraid_instance_ida); -static dev_t spraid_chr_devt; + static struct class *spraid_class; #define SPRAID_CAP_TIMEOUT_UNIT_MS (HZ / 2) @@ -133,7 +129,7 @@ static struct workqueue_struct *spraid_wq; #define ADMIN_TIMEOUT (admin_tmout * HZ) #define ADMIN_ERR_TIMEOUT 32757 -#define SPRAID_WAIT_ABNL_CMD_TIMEOUT (wait_abl_tmout * 2) +#define SPRAID_WAIT_ABNL_CMD_TIMEOUT (3 * 2) #define SPRAID_DMA_MSK_BIT_MAX 64 @@ -147,6 +143,13 @@ enum FW_STAT_CODE { FW_STAT_NEED_RETRY }; +static const char * const raid_levels[] = {"0", "1", "5", "6", "10", "50", "60", "NA"}; + +static const char * const raid_states[] = { + "NA", "NORMAL", "FAULT", "DEGRADE", "NOT_FORMATTED", "FORMATTING", "SANITIZING", + "INITIALIZING", "INITIALIZE_FAIL", "DELETING", "DELETE_FAIL", "WRITE_PROTECT" +}; + static int ioq_depth_set(const char *val, const struct kernel_param *kp) { int n = 0; @@ -263,12 +266,6 @@ static int spraid_pci_enable(struct spraid_dev *hdev) return ret; } -static inline -struct spraid_admin_request *spraid_admin_req(struct request *req) -{ - return blk_mq_rq_to_pdu(req); -} - static int spraid_npages_prp(u32 size, struct spraid_dev *hdev) { u32 nprps = DIV_ROUND_UP(size + hdev->page_size, hdev->page_size); @@ -419,7 +416,7 @@ static void spraid_submit_cmd(struct spraid_queue *spraidq, const void *cmd) writel(spraidq->sq_tail, spraidq->q_db); spin_unlock_irqrestore(&spraidq->sq_lock, flags); - dev_log_dbg(spraidq->hdev->dev, "cid[%d], qid[%d], opcode[0x%x], flags[0x%x], hdid[%u]\n", + dev_log_dbg(spraidq->hdev->dev, "cid[%d] qid[%d], opcode[0x%x], flags[0x%x], hdid[%u]\n", acd->command_id, spraidq->qid, acd->opcode, acd->flags, le32_to_cpu(acd->hdid)); } @@ -605,18 +602,15 @@ static void spraid_setup_rw_cmd(struct spraid_dev *hdev, if (scmd->cmd_len == 6) { datalength = (u32)(scmd->cmnd[4] == 0 ? IO_6_DEFAULT_TX_LEN : scmd->cmnd[4]); - start_lba_lo = ((u32)scmd->cmnd[1] << 16) | - ((u32)scmd->cmnd[2] << 8) | (u32)scmd->cmnd[3]; + start_lba_lo = (u32)get_unaligned_be24(&scmd->cmnd[1]); start_lba_lo &= 0x1FFFFF; } /* 10-byte READ(0x28) or WRITE(0x2A) cdb */ else if (scmd->cmd_len == 10) { - datalength = (u32)scmd->cmnd[8] | ((u32)scmd->cmnd[7] << 8); - start_lba_lo = ((u32)scmd->cmnd[2] << 24) | - ((u32)scmd->cmnd[3] << 16) | - ((u32)scmd->cmnd[4] << 8) | (u32)scmd->cmnd[5]; + datalength = (u32)get_unaligned_be16(&scmd->cmnd[7]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[2]); if (scmd->cmnd[1] & FUA_MASK) control |= SPRAID_RW_FUA; @@ -624,42 +618,26 @@ static void spraid_setup_rw_cmd(struct spraid_dev *hdev, /* 12-byte READ(0xA8) or WRITE(0xAA) cdb */ else if (scmd->cmd_len == 12) { - datalength = ((u32)scmd->cmnd[6] << 24) | - ((u32)scmd->cmnd[7] << 16) | - ((u32)scmd->cmnd[8] << 8) | (u32)scmd->cmnd[9]; - start_lba_lo = ((u32)scmd->cmnd[2] << 24) | - ((u32)scmd->cmnd[3] << 16) | - ((u32)scmd->cmnd[4] << 8) | (u32)scmd->cmnd[5]; + datalength = get_unaligned_be32(&scmd->cmnd[6]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[2]); if (scmd->cmnd[1] & FUA_MASK) control |= SPRAID_RW_FUA; } /* 16-byte READ(0x88) or WRITE(0x8A) cdb */ else if (scmd->cmd_len == 16) { - datalength = ((u32)scmd->cmnd[10] << 24) | - ((u32)scmd->cmnd[11] << 16) | - ((u32)scmd->cmnd[12] << 8) | (u32)scmd->cmnd[13]; - start_lba_lo = ((u32)scmd->cmnd[6] << 24) | - ((u32)scmd->cmnd[7] << 16) | - ((u32)scmd->cmnd[8] << 8) | (u32)scmd->cmnd[9]; - start_lba_hi = ((u32)scmd->cmnd[2] << 24) | - ((u32)scmd->cmnd[3] << 16) | - ((u32)scmd->cmnd[4] << 8) | (u32)scmd->cmnd[5]; + datalength = get_unaligned_be32(&scmd->cmnd[10]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[6]); + start_lba_hi = get_unaligned_be32(&scmd->cmnd[2]); if (scmd->cmnd[1] & FUA_MASK) control |= SPRAID_RW_FUA; } /* 32-byte READ(0x88) or WRITE(0x8A) cdb */ else if (scmd->cmd_len == 32) { - datalength = ((u32)scmd->cmnd[28] << 24) | - ((u32)scmd->cmnd[29] << 16) | - ((u32)scmd->cmnd[30] << 8) | (u32)scmd->cmnd[31]; - start_lba_lo = ((u32)scmd->cmnd[16] << 24) | - ((u32)scmd->cmnd[17] << 16) | - ((u32)scmd->cmnd[18] << 8) | (u32)scmd->cmnd[19]; - start_lba_hi = ((u32)scmd->cmnd[12] << 24) | - ((u32)scmd->cmnd[13] << 16) | - ((u32)scmd->cmnd[14] << 8) | (u32)scmd->cmnd[15]; + datalength = get_unaligned_be32(&scmd->cmnd[28]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[16]); + start_lba_hi = get_unaligned_be32(&scmd->cmnd[12]); if (scmd->cmnd[10] & FUA_MASK) control |= SPRAID_RW_FUA; @@ -814,7 +792,7 @@ static void spraid_map_status(struct spraid_iod *iod, struct scsi_cmnd *scmd, if (scmd->result & SAM_STAT_CHECK_CONDITION) { memset(scmd->sense_buffer, 0, SCSI_SENSE_BUFFERSIZE); memcpy(scmd->sense_buffer, iod->sense, SCSI_SENSE_BUFFERSIZE); - set_driver_byte(scmd, DRIVER_SENSE); + scmd->result = (scmd->result & 0x00ffffff) | (DRIVER_SENSE << 24); } break; case FW_STAT_ABORTED: @@ -850,14 +828,13 @@ static int spraid_queue_command(struct Scsi_Host *shost, struct scsi_cmnd *scmd) int ret; if (unlikely(!scmd)) { - dev_err(hdev->dev, "err, scmd is null, return 0\n"); + dev_err(hdev->dev, "err, scmd is null\n"); return 0; } if (unlikely(hdev->state != SPRAID_LIVE)) { set_host_byte(scmd, DID_NO_CONNECT); scmd->scsi_done(scmd); - dev_err(hdev->dev, "[%s] err, hdev state is not live\n", __func__); return 0; } @@ -894,7 +871,7 @@ static int spraid_queue_command(struct Scsi_Host *shost, struct scsi_cmnd *scmd) WRITE_ONCE(iod->state, SPRAID_CMD_IN_FLIGHT); spraid_submit_cmd(ioq, &ioq_cmd); elapsed = jiffies - scmd->jiffies_at_alloc; - dev_log_dbg(hdev->dev, "cid[%d], qid[%d] submit IO cost %3ld.%3ld seconds\n", + dev_log_dbg(hdev->dev, "cid[%d] qid[%d] submit IO cost %3ld.%3ld seconds\n", cid, hwq, elapsed / HZ, elapsed % HZ); return 0; @@ -945,6 +922,10 @@ static int spraid_slave_alloc(struct scsi_device *sdev) scan_host: hostdata->hdid = le32_to_cpu(hdev->devices[idx].hdid); + hostdata->max_io_kb = le16_to_cpu(hdev->devices[idx].max_io_kb); + hostdata->attr = hdev->devices[idx].attr; + hostdata->flag = hdev->devices[idx].flag; + hostdata->rg_id = 0xff; sdev->hostdata = hostdata; up_read(&hdev->devices_rwsem); return 0; @@ -964,13 +945,13 @@ static int spraid_slave_configure(struct scsi_device *sdev) struct spraid_sdev_hostdata *hostdata = sdev->hostdata; u32 max_sec = sdev->host->max_sectors; - if (!hostdata) { + if (hostdata) { idx = hostdata->hdid - 1; if (sdev->channel == hdev->devices[idx].channel && sdev->id == le16_to_cpu(hdev->devices[idx].target) && sdev->lun < hdev->devices[idx].lun) { if (SPRAID_DEV_INFO_ATTR_PT(hdev->devices[idx].attr)) - timeout = scmd_tmout_pt * HZ; + timeout = 30 * HZ; else timeout = scmd_tmout_nonpt * HZ; max_sec = le16_to_cpu(hdev->devices[idx].max_io_kb) << 1; @@ -991,7 +972,6 @@ static int spraid_slave_configure(struct scsi_device *sdev) if ((max_sec == 0) || (max_sec > sdev->host->max_sectors)) max_sec = sdev->host->max_sectors; - blk_queue_max_hw_sectors(sdev->request_queue, max_sec); dev_info(hdev->dev, "[%s] sdev->channel:id:lun[%d:%d:%lld], scmd_timeout[%d]s, maxsec[%d]\n", __func__, sdev->channel, sdev->id, sdev->lun, timeout / HZ, max_sec); @@ -1176,6 +1156,75 @@ static inline bool spraid_cqe_pending(struct spraid_queue *spraidq) spraidq->cq_phase; } +static void spraid_sata_report_zone_handle(struct scsi_cmnd *scmd, struct spraid_iod *iod) +{ + int i = 0; + unsigned int bytes = 0; + struct scatterlist *sg = scsi_sglist(scmd); + + scsi_for_each_sg(scmd, sg, iod->nsge, i) { + unsigned int offset = 0; + + if (bytes == 0) { + char *hdr; + u32 list_length; + u64 max_lba, opt_lba; + u16 same; + + hdr = sg_virt(sg); + + list_length = get_unaligned_le32(&hdr[0]); + same = get_unaligned_le16(&hdr[4]); + max_lba = get_unaligned_le64(&hdr[8]); + opt_lba = get_unaligned_le64(&hdr[16]); + put_unaligned_be32(list_length, &hdr[0]); + hdr[4] = same & 0xf; + put_unaligned_be64(max_lba, &hdr[8]); + put_unaligned_be64(opt_lba, &hdr[16]); + offset += 64; + bytes += 64; + } + while (offset < sg_dma_len(sg)) { + char *rec; + u8 cond, type, non_seq, reset; + u64 size, start, wp; + + rec = sg_virt(sg) + offset; + type = rec[0] & 0xf; + cond = (rec[1] >> 4) & 0xf; + non_seq = (rec[1] & 2); + reset = (rec[1] & 1); + size = get_unaligned_le64(&rec[8]); + start = get_unaligned_le64(&rec[16]); + wp = get_unaligned_le64(&rec[24]); + rec[0] = type; + rec[1] = (cond << 4) | non_seq | reset; + put_unaligned_be64(size, &rec[8]); + put_unaligned_be64(start, &rec[16]); + put_unaligned_be64(wp, &rec[24]); + WARN_ON(offset + 64 > sg_dma_len(sg)); + offset += 64; + bytes += 64; + } + } +} + +static inline void spraid_handle_ata_cmd(struct spraid_dev *hdev, struct scsi_cmnd *scmd, + struct spraid_iod *iod) +{ + if (hdev->ctrl_info->card_type != SPRAID_CARD_HBA) + return; + + switch (scmd->cmnd[0]) { + case ZBC_IN: + dev_info(hdev->dev, "[%s] process report zone\n", __func__); + spraid_sata_report_zone_handle(scmd, iod); + break; + default: + break; + } +} + static void spraid_complete_ioq_cmnd(struct spraid_queue *ioq, struct spraid_completion *cqe) { struct spraid_dev *hdev = ioq->hdev; @@ -1197,12 +1246,12 @@ static void spraid_complete_ioq_cmnd(struct spraid_queue *ioq, struct spraid_com iod = scsi_cmd_priv(scmd); elapsed = jiffies - scmd->jiffies_at_alloc; - dev_log_dbg(hdev->dev, "cid[%d], qid[%d] finish IO cost %3ld.%3ld seconds\n", + dev_log_dbg(hdev->dev, "cid[%d] qid[%d] finish IO cost %3ld.%3ld seconds\n", cqe->cmd_id, ioq->qid, elapsed / HZ, elapsed % HZ); if (cmpxchg(&iod->state, SPRAID_CMD_IN_FLIGHT, SPRAID_CMD_COMPLETE) != SPRAID_CMD_IN_FLIGHT) { - dev_warn(hdev->dev, "cid[%d], qid[%d] enters abnormal handler, cost %3ld.%3ld seconds\n", + dev_warn(hdev->dev, "cid[%d] qid[%d] enters abnormal handler, cost %3ld.%3ld seconds\n", cqe->cmd_id, ioq->qid, elapsed / HZ, elapsed % HZ); WRITE_ONCE(iod->state, SPRAID_CMD_TMO_COMPLETE); @@ -1215,6 +1264,8 @@ static void spraid_complete_ioq_cmnd(struct spraid_queue *ioq, struct spraid_com return; } + spraid_handle_ata_cmd(hdev, scmd, iod); + spraid_map_status(iod, scmd, cqe); if (iod->nsge) { iod->nsge = 0; @@ -1224,38 +1275,36 @@ static void spraid_complete_ioq_cmnd(struct spraid_queue *ioq, struct spraid_com scmd->scsi_done(scmd); } -static inline void spraid_end_admin_request(struct request *req, __le16 status, - __le32 result0, __le32 result1) -{ - struct spraid_admin_request *rq = spraid_admin_req(req); - - rq->status = le16_to_cpu(status) >> 1; - rq->result0 = le32_to_cpu(result0); - rq->result1 = le32_to_cpu(result1); - blk_mq_complete_request(req); -} - static void spraid_complete_adminq_cmnd(struct spraid_queue *adminq, struct spraid_completion *cqe) { - struct blk_mq_tags *tags = adminq->hdev->admin_tagset.tags[0]; - struct request *req; + struct spraid_dev *hdev = adminq->hdev; + struct spraid_cmd *adm_cmd; - req = blk_mq_tag_to_rq(tags, cqe->cmd_id); - if (unlikely(!req)) { + adm_cmd = hdev->adm_cmds + cqe->cmd_id; + if (unlikely(adm_cmd->state == SPRAID_CMD_IDLE)) { dev_warn(adminq->hdev->dev, "Invalid id %d completed on queue %d\n", cqe->cmd_id, le16_to_cpu(cqe->sq_id)); return; } - spraid_end_admin_request(req, cqe->status, cqe->result, cqe->result1); + + adm_cmd->status = le16_to_cpu(cqe->status) >> 1; + adm_cmd->result0 = le32_to_cpu(cqe->result); + adm_cmd->result1 = le32_to_cpu(cqe->result1); + + complete(&adm_cmd->cmd_done); } +static void spraid_send_aen(struct spraid_dev *hdev, u16 cid); + static void spraid_complete_aen(struct spraid_queue *spraidq, struct spraid_completion *cqe) { struct spraid_dev *hdev = spraidq->hdev; u32 result = le32_to_cpu(cqe->result); - dev_info(hdev->dev, "rcv aen, status[%x], result[%x]\n", - le16_to_cpu(cqe->status) >> 1, result); + dev_info(hdev->dev, "rcv aen, cid[%d], status[0x%x], result[0x%x]\n", + cqe->cmd_id, le16_to_cpu(cqe->status) >> 1, result); + + spraid_send_aen(hdev, cqe->cmd_id); if ((le16_to_cpu(cqe->status) >> 1) != SPRAID_SC_SUCCESS) return; @@ -1264,22 +1313,19 @@ static void spraid_complete_aen(struct spraid_queue *spraidq, struct spraid_comp spraid_handle_aen_notice(hdev, result); break; case SPRAID_AEN_VS: - spraid_handle_aen_vs(hdev, result); + spraid_handle_aen_vs(hdev, result, le32_to_cpu(cqe->result1)); break; default: dev_warn(hdev->dev, "Unsupported async event type: %u\n", result & 0x7); break; } - queue_work(spraid_wq, &hdev->aen_work); } -static void spraid_put_ioq_ptcmd(struct spraid_dev *hdev, struct spraid_ioq_ptcmd *cmd); - static void spraid_complete_ioq_sync_cmnd(struct spraid_queue *ioq, struct spraid_completion *cqe) { struct spraid_dev *hdev = ioq->hdev; - struct spraid_ioq_ptcmd *ptcmd; + struct spraid_cmd *ptcmd; ptcmd = hdev->ioq_ptcmds + (ioq->qid - 1) * SPRAID_PTCMDS_PERQ + cqe->cmd_id - SPRAID_IO_BLK_MQ_DEPTH; @@ -1289,8 +1335,6 @@ static void spraid_complete_ioq_sync_cmnd(struct spraid_queue *ioq, struct sprai ptcmd->result1 = le32_to_cpu(cqe->result1); complete(&ptcmd->cmd_done); - - spraid_put_ioq_ptcmd(hdev, ptcmd); } static inline void spraid_handle_cqe(struct spraid_queue *spraidq, u16 idx) @@ -1304,7 +1348,7 @@ static inline void spraid_handle_cqe(struct spraid_queue *spraidq, u16 idx) return; } - dev_log_dbg(hdev->dev, "cid[%d], qid[%d], result[0x%x], sq_id[%d], status[0x%x]\n", + dev_log_dbg(hdev->dev, "cid[%d] qid[%d], result[0x%x], sq_id[%d], status[0x%x]\n", cqe->cmd_id, spraidq->qid, le32_to_cpu(cqe->result), le16_to_cpu(cqe->sq_id), le16_to_cpu(cqe->status)); @@ -1452,62 +1496,119 @@ static u32 spraid_bar_size(struct spraid_dev *hdev, u32 nr_ioqs) return (SPRAID_REG_DBS + ((nr_ioqs + 1) * 8 * hdev->db_stride)); } -static inline void spraid_clear_spraid_request(struct request *req) +static int spraid_alloc_admin_cmds(struct spraid_dev *hdev) { - if (!(req->rq_flags & RQF_DONTPREP)) { - spraid_admin_req(req)->flags = 0; - req->rq_flags |= RQF_DONTPREP; + int i; + + INIT_LIST_HEAD(&hdev->adm_cmd_list); + spin_lock_init(&hdev->adm_cmd_lock); + + hdev->adm_cmds = kcalloc_node(SPRAID_AQ_BLK_MQ_DEPTH, sizeof(struct spraid_cmd), + GFP_KERNEL, hdev->numa_node); + + if (!hdev->adm_cmds) { + dev_err(hdev->dev, "Alloc admin cmds failed\n"); + return -ENOMEM; + } + + for (i = 0; i < SPRAID_AQ_BLK_MQ_DEPTH; i++) { + hdev->adm_cmds[i].qid = 0; + hdev->adm_cmds[i].cid = i; + list_add_tail(&(hdev->adm_cmds[i].list), &hdev->adm_cmd_list); } + + dev_info(hdev->dev, "Alloc admin cmds success, num[%d]\n", SPRAID_AQ_BLK_MQ_DEPTH); + + return 0; } -static struct request *spraid_alloc_admin_request(struct request_queue *q, - struct spraid_admin_command *cmd, - blk_mq_req_flags_t flags) +static void spraid_free_admin_cmds(struct spraid_dev *hdev) { - u32 op = COMMAND_IS_WRITE(cmd) ? REQ_OP_DRV_OUT : REQ_OP_DRV_IN; - struct request *req; + kfree(hdev->adm_cmds); + hdev->adm_cmds = NULL; + INIT_LIST_HEAD(&hdev->adm_cmd_list); +} + +static struct spraid_cmd *spraid_get_cmd(struct spraid_dev *hdev, enum spraid_cmd_type type) +{ + struct spraid_cmd *cmd = NULL; + unsigned long flags; + struct list_head *head = &hdev->adm_cmd_list; + spinlock_t *slock = &hdev->adm_cmd_lock; - req = blk_mq_alloc_request(q, op, flags); - if (IS_ERR(req)) - return req; - req->cmd_flags |= REQ_FAILFAST_DRIVER; - spraid_clear_spraid_request(req); - spraid_admin_req(req)->cmd = cmd; + if (type == SPRAID_CMD_IOPT) { + head = &hdev->ioq_pt_list; + slock = &hdev->ioq_pt_lock; + } + + spin_lock_irqsave(slock, flags); + if (list_empty(head)) { + spin_unlock_irqrestore(slock, flags); + dev_err(hdev->dev, "err, cmd[%d] list empty\n", type); + return NULL; + } + cmd = list_entry(head->next, struct spraid_cmd, list); + list_del_init(&cmd->list); + spin_unlock_irqrestore(slock, flags); + + WRITE_ONCE(cmd->state, SPRAID_CMD_IN_FLIGHT); - return req; + return cmd; } -static int spraid_submit_admin_sync_cmd(struct request_queue *q, - struct spraid_admin_command *cmd, - u32 *result, void *buffer, - u32 bufflen, u32 timeout, int at_head, blk_mq_req_flags_t flags) +static void spraid_put_cmd(struct spraid_dev *hdev, struct spraid_cmd *cmd, + enum spraid_cmd_type type) { - struct request *req; - int ret; + unsigned long flags; + struct list_head *head = &hdev->adm_cmd_list; + spinlock_t *slock = &hdev->adm_cmd_lock; - req = spraid_alloc_admin_request(q, cmd, flags); - if (IS_ERR(req)) - return PTR_ERR(req); + if (type == SPRAID_CMD_IOPT) { + head = &hdev->ioq_pt_list; + slock = &hdev->ioq_pt_lock; + } - req->timeout = timeout ? timeout : ADMIN_TIMEOUT; - if (buffer && bufflen) { - ret = blk_rq_map_kern(q, req, buffer, bufflen, GFP_KERNEL); - if (ret) - goto out; + spin_lock_irqsave(slock, flags); + WRITE_ONCE(cmd->state, SPRAID_CMD_IDLE); + list_add_tail(&cmd->list, head); + spin_unlock_irqrestore(slock, flags); +} + + +static int spraid_submit_admin_sync_cmd(struct spraid_dev *hdev, struct spraid_admin_command *cmd, + u32 *result0, u32 *result1, u32 timeout) +{ + struct spraid_cmd *adm_cmd = spraid_get_cmd(hdev, SPRAID_CMD_ADM); + + if (!adm_cmd) { + dev_err(hdev->dev, "err, get admin cmd failed\n"); + return -EFAULT; } - blk_execute_rq(req->q, NULL, req, at_head); - if (result) - *result = spraid_admin_req(req)->result0; + timeout = timeout ? timeout : ADMIN_TIMEOUT; - if (spraid_admin_req(req)->flags & SPRAID_REQ_CANCELLED) - ret = -EINTR; - else - ret = spraid_admin_req(req)->status; + init_completion(&adm_cmd->cmd_done); -out: - blk_mq_free_request(req); - return ret; + cmd->common.command_id = adm_cmd->cid; + spraid_submit_cmd(&hdev->queues[0], cmd); + + if (!wait_for_completion_timeout(&adm_cmd->cmd_done, timeout)) { + dev_err(hdev->dev, "[%s] cid[%d] qid[%d] timeout, opcode[0x%x] subopcode[0x%x]\n", + __func__, adm_cmd->cid, adm_cmd->qid, cmd->usr_cmd.opcode, + cmd->usr_cmd.info_0.subopcode); + WRITE_ONCE(adm_cmd->state, SPRAID_CMD_TIMEOUT); + spraid_put_cmd(hdev, adm_cmd, SPRAID_CMD_ADM); + return -EINVAL; + } + + if (result0) + *result0 = adm_cmd->result0; + if (result1) + *result1 = adm_cmd->result1; + + spraid_put_cmd(hdev, adm_cmd, SPRAID_CMD_ADM); + + return adm_cmd->status; } static int spraid_create_cq(struct spraid_dev *hdev, u16 qid, @@ -1524,8 +1625,7 @@ static int spraid_create_cq(struct spraid_dev *hdev, u16 qid, admin_cmd.create_cq.cq_flags = cpu_to_le16(flags); admin_cmd.create_cq.irq_vector = cpu_to_le16(cq_vector); - return spraid_submit_admin_sync_cmd(hdev->admin_q, &admin_cmd, NULL, - NULL, 0, 0, 0, 0); + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); } static int spraid_create_sq(struct spraid_dev *hdev, u16 qid, @@ -1542,8 +1642,7 @@ static int spraid_create_sq(struct spraid_dev *hdev, u16 qid, admin_cmd.create_sq.sq_flags = cpu_to_le16(flags); admin_cmd.create_sq.cqid = cpu_to_le16(qid); - return spraid_submit_admin_sync_cmd(hdev->admin_q, &admin_cmd, NULL, - NULL, 0, 0, 0, 0); + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); } static void spraid_free_queue(struct spraid_queue *spraidq) @@ -1581,8 +1680,7 @@ static int spraid_delete_queue(struct spraid_dev *hdev, u8 op, u16 id) admin_cmd.delete_queue.opcode = op; admin_cmd.delete_queue.qid = cpu_to_le16(id); - ret = spraid_submit_admin_sync_cmd(hdev->admin_q, &admin_cmd, NULL, - NULL, 0, 0, 0, 0); + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); if (ret) dev_err(hdev->dev, "Delete %s:[%d] failed\n", @@ -1663,19 +1761,28 @@ static int spraid_set_features(struct spraid_dev *hdev, u32 fid, u32 dword11, vo size_t buflen, u32 *result) { struct spraid_admin_command admin_cmd; - u32 res; int ret; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + + if (buffer && buflen) { + data_ptr = dma_alloc_coherent(hdev->dev, buflen, &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memcpy(data_ptr, buffer, buflen); + } memset(&admin_cmd, 0, sizeof(admin_cmd)); admin_cmd.features.opcode = SPRAID_ADMIN_SET_FEATURES; admin_cmd.features.fid = cpu_to_le32(fid); admin_cmd.features.dword11 = cpu_to_le32(dword11); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); - ret = spraid_submit_admin_sync_cmd(hdev->admin_q, &admin_cmd, &res, - buffer, buflen, 0, 0, 0); + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, result, NULL, 0); - if (!ret && result) - *result = res; + if (data_ptr) + dma_free_coherent(hdev->dev, buflen, data_ptr, data_dma); return ret; } @@ -1764,8 +1871,7 @@ static int spraid_setup_io_queues(struct spraid_dev *hdev) break; } dev_info(hdev->dev, "[%s] max_qid: %d, queue_count: %d, online_queue: %d, ioq_depth: %d\n", - __func__, hdev->max_qid, hdev->queue_count, - hdev->online_queues, hdev->ioq_depth); + __func__, hdev->max_qid, hdev->queue_count, hdev->online_queues, hdev->ioq_depth); return spraid_create_io_queues(hdev); } @@ -1889,10 +1995,11 @@ static int spraid_get_dev_list(struct spraid_dev *hdev, struct spraid_dev_info * u32 nd = le32_to_cpu(hdev->ctrl_info->nd); struct spraid_admin_command admin_cmd; struct spraid_dev_list *list_buf; + dma_addr_t data_dma = 0; u32 i, idx, hdid, ndev; int ret = 0; - list_buf = kmalloc(sizeof(*list_buf), GFP_KERNEL); + list_buf = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &data_dma, GFP_KERNEL); if (!list_buf) return -ENOMEM; @@ -1901,9 +2008,9 @@ static int spraid_get_dev_list(struct spraid_dev *hdev, struct spraid_dev_info * admin_cmd.get_info.opcode = SPRAID_ADMIN_GET_INFO; admin_cmd.get_info.type = SPRAID_GET_INFO_DEV_LIST; admin_cmd.get_info.cdw11 = cpu_to_le32(idx); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); - ret = spraid_submit_admin_sync_cmd(hdev->admin_q, &admin_cmd, NULL, list_buf, - sizeof(*list_buf), 0, 0, 0); + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); if (ret) { dev_err(hdev->dev, "Get device list failed, nd: %u, idx: %u, ret: %d\n", @@ -1916,12 +2023,11 @@ static int spraid_get_dev_list(struct spraid_dev *hdev, struct spraid_dev_info * for (i = 0; i < ndev; i++) { hdid = le32_to_cpu(list_buf->devices[i].hdid); - dev_info(hdev->dev, "list_buf->devices[%d], hdid: %u target: %d, channel: %d, lun: %d, attr[%x]\n", - i, hdid, - le16_to_cpu(list_buf->devices[i].target), - list_buf->devices[i].channel, - list_buf->devices[i].lun, - list_buf->devices[i].attr); + dev_info(hdev->dev, "list_buf->devices[%d], hdid: %u target: %d, channel: %d, lun: %d, attr[0x%x]\n", + i, hdid, le16_to_cpu(list_buf->devices[i].target), + list_buf->devices[i].channel, + list_buf->devices[i].lun, + list_buf->devices[i].attr); if (hdid > nd || hdid == 0) { dev_err(hdev->dev, "err, hdid[%d] invalid\n", hdid); continue; @@ -1936,21 +2042,29 @@ static int spraid_get_dev_list(struct spraid_dev *hdev, struct spraid_dev_info * } out: - kfree(list_buf); + dma_free_coherent(hdev->dev, PAGE_SIZE, list_buf, data_dma); return ret; } -static void spraid_send_aen(struct spraid_dev *hdev) +static void spraid_send_aen(struct spraid_dev *hdev, u16 cid) { struct spraid_queue *adminq = &hdev->queues[0]; struct spraid_admin_command admin_cmd; memset(&admin_cmd, 0, sizeof(admin_cmd)); admin_cmd.common.opcode = SPRAID_ADMIN_ASYNC_EVENT; - admin_cmd.common.command_id = SPRAID_AQ_BLK_MQ_DEPTH; + admin_cmd.common.command_id = cid; spraid_submit_cmd(adminq, &admin_cmd); - dev_info(hdev->dev, "send aen, cid[%d]\n", SPRAID_AQ_BLK_MQ_DEPTH); + dev_info(hdev->dev, "send aen, cid[%d]\n", cid); +} + +static inline void spraid_send_all_aen(struct spraid_dev *hdev) +{ + u16 i; + + for (i = 0; i < hdev->ctrl_info->aerl; i++) + spraid_send_aen(hdev, i + SPRAID_AQ_BLK_MQ_DEPTH); } static int spraid_add_device(struct spraid_dev *hdev, struct spraid_dev_info *device) @@ -1958,6 +2072,10 @@ static int spraid_add_device(struct spraid_dev *hdev, struct spraid_dev_info *de struct Scsi_Host *shost = hdev->shost; struct scsi_device *sdev; + dev_info(hdev->dev, "add device, hdid: %u target: %d, channel: %d, lun: %d, attr[0x%x]\n", + le32_to_cpu(device->hdid), le16_to_cpu(device->target), + device->channel, device->lun, device->attr); + sdev = scsi_device_lookup(shost, device->channel, le16_to_cpu(device->target), 0); if (sdev) { dev_warn(hdev->dev, "Device is already exist, channel: %d, target_id: %d, lun: %d\n", @@ -1974,9 +2092,13 @@ static int spraid_rescan_device(struct spraid_dev *hdev, struct spraid_dev_info struct Scsi_Host *shost = hdev->shost; struct scsi_device *sdev; + dev_info(hdev->dev, "rescan device, hdid: %u target: %d, channel: %d, lun: %d, attr[0x%x]\n", + le32_to_cpu(device->hdid), le16_to_cpu(device->target), + device->channel, device->lun, device->attr); + sdev = scsi_device_lookup(shost, device->channel, le16_to_cpu(device->target), 0); if (!sdev) { - dev_warn(hdev->dev, "Device is not exit, channel: %d, target_id: %d, lun: %d\n", + dev_warn(hdev->dev, "device is not exit rescan it, channel: %d, target_id: %d, lun: %d\n", device->channel, le16_to_cpu(device->target), 0); return -ENODEV; } @@ -1991,9 +2113,13 @@ static int spraid_remove_device(struct spraid_dev *hdev, struct spraid_dev_info struct Scsi_Host *shost = hdev->shost; struct scsi_device *sdev; + dev_info(hdev->dev, "remove device, hdid: %u target: %d, channel: %d, lun: %d, attr[0x%x]\n", + le32_to_cpu(org_device->hdid), le16_to_cpu(org_device->target), + org_device->channel, org_device->lun, org_device->attr); + sdev = scsi_device_lookup(shost, org_device->channel, le16_to_cpu(org_device->target), 0); if (!sdev) { - dev_warn(hdev->dev, "Device is not exit, channel: %d, target_id: %d, lun: %d\n", + dev_warn(hdev->dev, "device is not exit remove it, channel: %d, target_id: %d, lun: %d\n", org_device->channel, le16_to_cpu(org_device->target), 0); return -ENODEV; } @@ -2029,36 +2155,54 @@ static int spraid_dev_list_init(struct spraid_dev *hdev) return 0; } +static int luntarget_cmp_func(const void *l, const void *r) +{ + const struct spraid_dev_info *ln = l; + const struct spraid_dev_info *rn = r; + + if (ln->channel == rn->channel) + return ln->target - rn->target; + + return ln->channel - rn->channel; +} + static void spraid_scan_work(struct work_struct *work) { struct spraid_dev *hdev = container_of(work, struct spraid_dev, scan_work); struct spraid_dev_info *devices, *org_devices; + struct spraid_dev_info *sortdevice; u32 nd = le32_to_cpu(hdev->ctrl_info->nd); u8 flag, org_flag; int i, ret; + int count = 0; devices = kcalloc(nd, sizeof(struct spraid_dev_info), GFP_KERNEL); if (!devices) return; + + sortdevice = kcalloc(nd, sizeof(struct spraid_dev_info), GFP_KERNEL); + if (!sortdevice) + goto free_list; + ret = spraid_get_dev_list(hdev, devices); if (ret) - goto free_list; + goto free_all; org_devices = hdev->devices; for (i = 0; i < nd; i++) { org_flag = org_devices[i].flag; flag = devices[i].flag; - dev_log_dbg(hdev->dev, "i: %d, org_flag: 0x%x, flag: 0x%x\n", - i, org_flag, flag); + dev_log_dbg(hdev->dev, "i: %d, org_flag: 0x%x, flag: 0x%x\n", i, org_flag, flag); if (SPRAID_DEV_INFO_FLAG_VALID(flag)) { if (!SPRAID_DEV_INFO_FLAG_VALID(org_flag)) { down_write(&hdev->devices_rwsem); memcpy(&org_devices[i], &devices[i], - sizeof(struct spraid_dev_info)); + sizeof(struct spraid_dev_info)); + memcpy(&sortdevice[count++], &devices[i], + sizeof(struct spraid_dev_info)); up_write(&hdev->devices_rwsem); - spraid_add_device(hdev, &devices[i]); } else if (SPRAID_DEV_INFO_FLAG_CHANGE(flag)) { spraid_rescan_device(hdev, &devices[i]); } @@ -2071,6 +2215,16 @@ static void spraid_scan_work(struct work_struct *work) } } } + + dev_info(hdev->dev, "scan work add device count = %d\n", count); + + sort(sortdevice, count, sizeof(sortdevice[0]), luntarget_cmp_func, NULL); + + for (i = 0; i < count; i++) + spraid_add_device(hdev, &sortdevice[i]); + +free_all: + kfree(sortdevice); free_list: kfree(devices); } @@ -2083,6 +2237,15 @@ static void spraid_timesyn_work(struct work_struct *work) spraid_configure_timestamp(hdev); } +static int spraid_init_ctrl_info(struct spraid_dev *hdev); +static void spraid_fw_act_work(struct work_struct *work) +{ + struct spraid_dev *hdev = container_of(work, struct spraid_dev, fw_act_work); + + if (spraid_init_ctrl_info(hdev)) + dev_err(hdev->dev, "get ctrl info failed after fw act\n"); +} + static void spraid_queue_scan(struct spraid_dev *hdev) { queue_work(spraid_wq, &hdev->scan_work); @@ -2094,6 +2257,9 @@ static void spraid_handle_aen_notice(struct spraid_dev *hdev, u32 result) case SPRAID_AEN_DEV_CHANGED: spraid_queue_scan(hdev); break; + case SPRAID_AEN_FW_ACT_START: + dev_info(hdev->dev, "fw activation starting\n"); + break; case SPRAID_AEN_HOST_PROBING: break; default: @@ -2101,25 +2267,25 @@ static void spraid_handle_aen_notice(struct spraid_dev *hdev, u32 result) } } -static void spraid_handle_aen_vs(struct spraid_dev *hdev, u32 result) +static void spraid_handle_aen_vs(struct spraid_dev *hdev, u32 result, u32 result1) { - switch (result) { + switch ((result & 0xff00) >> 8) { case SPRAID_AEN_TIMESYN: queue_work(spraid_wq, &hdev->timesyn_work); break; + case SPRAID_AEN_FW_ACT_FINISH: + dev_info(hdev->dev, "fw activation finish\n"); + queue_work(spraid_wq, &hdev->fw_act_work); + break; + case SPRAID_AEN_EVENT_MIN ... SPRAID_AEN_EVENT_MAX: + dev_info(hdev->dev, "rcv card event[%d], param1[0x%x] param2[0x%x]\n", + (result & 0xff00) >> 8, result, result1); + break; default: - dev_warn(hdev->dev, "async event result: %x\n", result); + dev_warn(hdev->dev, "async event result: 0x%x\n", result); } } -static void spraid_async_event_work(struct work_struct *work) -{ - struct spraid_dev *hdev = - container_of(work, struct spraid_dev, aen_work); - - spraid_send_aen(hdev); -} - static int spraid_alloc_resources(struct spraid_dev *hdev) { int ret, nqueue; @@ -2149,10 +2315,16 @@ static int spraid_alloc_resources(struct spraid_dev *hdev) goto destroy_dma_pools; } + ret = spraid_alloc_admin_cmds(hdev); + if (ret) + goto free_queues; + dev_info(hdev->dev, "[%s] queues num: %d\n", __func__, nqueue); return 0; +free_queues: + kfree(hdev->queues); destroy_dma_pools: spraid_destroy_dma_pools(hdev); free_ctrl_info: @@ -2164,50 +2336,18 @@ static int spraid_alloc_resources(struct spraid_dev *hdev) static void spraid_free_resources(struct spraid_dev *hdev) { + spraid_free_admin_cmds(hdev); kfree(hdev->queues); spraid_destroy_dma_pools(hdev); kfree(hdev->ctrl_info); ida_free(&spraid_instance_ida, hdev->instance); } -static void spraid_setup_passthrough(struct request *req, struct spraid_admin_command *cmd) -{ - memcpy(cmd, spraid_admin_req(req)->cmd, sizeof(*cmd)); - cmd->common.flags &= ~SPRAID_CMD_FLAG_SGL_ALL; -} - -static inline void spraid_clear_hreq(struct request *req) -{ - if (!(req->rq_flags & RQF_DONTPREP)) { - spraid_admin_req(req)->flags = 0; - req->rq_flags |= RQF_DONTPREP; - } -} - -static blk_status_t spraid_setup_admin_cmd(struct request *req, struct spraid_admin_command *cmd) +static void spraid_bsg_unmap_data(struct spraid_dev *hdev, struct bsg_job *job) { - spraid_clear_hreq(req); - - memset(cmd, 0, sizeof(*cmd)); - switch (req_op(req)) { - case REQ_OP_DRV_IN: - case REQ_OP_DRV_OUT: - spraid_setup_passthrough(req, cmd); - break; - default: - WARN_ON_ONCE(1); - return BLK_STS_IOERR; - } - - cmd->common.command_id = req->tag; - return BLK_STS_OK; -} - -static void spraid_unmap_data(struct spraid_dev *hdev, struct request *req) -{ - struct spraid_iod *iod = blk_mq_rq_to_pdu(req); - enum dma_data_direction dma_dir = rq_data_dir(req) ? - DMA_TO_DEVICE : DMA_FROM_DEVICE; + struct request *rq = blk_mq_rq_from_pdu(job); + struct spraid_iod *iod = job->dd_data; + enum dma_data_direction dma_dir = rq_data_dir(rq) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; if (iod->nsge) dma_unmap_sg(hdev->dev, iod->sg, iod->nsge, dma_dir); @@ -2215,36 +2355,36 @@ static void spraid_unmap_data(struct spraid_dev *hdev, struct request *req) spraid_free_iod_res(hdev, iod); } -static blk_status_t spraid_admin_map_data(struct spraid_dev *hdev, struct request *req, - struct spraid_admin_command *cmd) +static int spraid_bsg_map_data(struct spraid_dev *hdev, struct bsg_job *job, + struct spraid_admin_command *cmd) { - struct spraid_iod *iod = blk_mq_rq_to_pdu(req); - struct request_queue *admin_q = req->q; - enum dma_data_direction dma_dir = rq_data_dir(req) ? - DMA_TO_DEVICE : DMA_FROM_DEVICE; - blk_status_t ret = BLK_STS_IOERR; - int nr_mapped; - int res; + struct request *rq = blk_mq_rq_from_pdu(job); + struct spraid_iod *iod = job->dd_data; + enum dma_data_direction dma_dir = rq_data_dir(rq) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; + int ret = 0; + + iod->sg = job->request_payload.sg_list; + iod->nsge = job->request_payload.sg_cnt; + iod->length = job->request_payload.payload_len; + iod->use_sgl = false; + iod->npages = -1; + iod->sg_drv_mgmt = false; - sg_init_table(iod->sg, blk_rq_nr_phys_segments(req)); - iod->nsge = blk_rq_map_sg(admin_q, req, iod->sg); if (!iod->nsge) goto out; - dev_info(hdev->dev, "nseg: %u, nsge: %u\n", - blk_rq_nr_phys_segments(req), iod->nsge); - - ret = BLK_STS_RESOURCE; - nr_mapped = dma_map_sg_attrs(hdev->dev, iod->sg, iod->nsge, dma_dir, DMA_ATTR_NO_WARN); - if (!nr_mapped) + ret = dma_map_sg_attrs(hdev->dev, iod->sg, iod->nsge, dma_dir, DMA_ATTR_NO_WARN); + if (!ret) goto out; - res = spraid_setup_prps(hdev, iod); - if (res) + ret = spraid_setup_prps(hdev, iod); + if (ret) goto unmap; + cmd->common.dptr.prp1 = cpu_to_le64(sg_dma_address(iod->sg)); cmd->common.dptr.prp2 = cpu_to_le64(iod->first_dma); - return BLK_STS_OK; + + return 0; unmap: dma_unmap_sg(hdev->dev, iod->sg, iod->nsge, dma_dir); @@ -2252,137 +2392,29 @@ static blk_status_t spraid_admin_map_data(struct spraid_dev *hdev, struct reques return ret; } -static blk_status_t spraid_init_admin_iod(struct request *rq, struct spraid_dev *hdev) -{ - struct spraid_iod *iod = blk_mq_rq_to_pdu(rq); - int nents = blk_rq_nr_phys_segments(rq); - unsigned int size = blk_rq_payload_bytes(rq); - - if (nents > SPRAID_INT_PAGES || size > SPRAID_INT_BYTES(hdev)) { - iod->sg = mempool_alloc(hdev->iod_mempool, GFP_ATOMIC); - if (!iod->sg) - return BLK_STS_RESOURCE; - } else { - iod->sg = iod->inline_sg; - } - - iod->nsge = 0; - iod->use_sgl = false; - iod->npages = -1; - iod->length = size; - iod->sg_drv_mgmt = true; - - return BLK_STS_OK; -} - -static blk_status_t spraid_queue_admin_rq(struct blk_mq_hw_ctx *hctx, - const struct blk_mq_queue_data *bd) -{ - struct spraid_queue *adminq = hctx->driver_data; - struct spraid_dev *hdev = adminq->hdev; - struct request *req = bd->rq; - struct spraid_iod *iod = blk_mq_rq_to_pdu(req); - struct spraid_admin_command cmd; - blk_status_t ret; - - ret = spraid_setup_admin_cmd(req, &cmd); - if (ret) - goto out; - - ret = spraid_init_admin_iod(req, hdev); - if (ret) - goto out; - - if (blk_rq_nr_phys_segments(req)) { - ret = spraid_admin_map_data(hdev, req, &cmd); - if (ret) - goto cleanup_iod; - } - - blk_mq_start_request(req); - spraid_submit_cmd(adminq, &cmd); - return BLK_STS_OK; - -cleanup_iod: - spraid_free_iod_res(hdev, iod); -out: - return ret; -} - -static blk_status_t spraid_error_status(struct request *req) -{ - switch (spraid_admin_req(req)->status & 0x7ff) { - case SPRAID_SC_SUCCESS: - return BLK_STS_OK; - default: - return BLK_STS_IOERR; - } -} - -static void spraid_complete_admin_rq(struct request *req) -{ - struct spraid_iod *iod = blk_mq_rq_to_pdu(req); - struct spraid_dev *hdev = iod->spraidq->hdev; - - if (blk_rq_nr_phys_segments(req)) - spraid_unmap_data(hdev, req); - blk_mq_end_request(req, spraid_error_status(req)); -} - -static int spraid_admin_init_hctx(struct blk_mq_hw_ctx *hctx, void *data, unsigned int hctx_idx) -{ - struct spraid_dev *hdev = data; - struct spraid_queue *adminq = &hdev->queues[0]; - - WARN_ON(hctx_idx != 0); - WARN_ON(hdev->admin_tagset.tags[0] != hctx->tags); - - hctx->driver_data = adminq; - return 0; -} - -static int spraid_admin_init_request(struct blk_mq_tag_set *set, struct request *req, - unsigned int hctx_idx, unsigned int numa_node) -{ - struct spraid_dev *hdev = set->driver_data; - struct spraid_iod *iod = blk_mq_rq_to_pdu(req); - struct spraid_queue *adminq = &hdev->queues[0]; - - WARN_ON(!adminq); - iod->spraidq = adminq; - return 0; -} - -static enum blk_eh_timer_return -spraid_admin_timeout(struct request *req, bool reserved) -{ - struct spraid_iod *iod = blk_mq_rq_to_pdu(req); - struct spraid_queue *spraidq = iod->spraidq; - struct spraid_dev *hdev = spraidq->hdev; - - dev_err(hdev->dev, "Admin cid[%d] qid[%d] timeout\n", - req->tag, spraidq->qid); - - if (spraid_poll_cq(spraidq, req->tag)) { - dev_warn(hdev->dev, "cid[%d] qid[%d] timeout, completion polled\n", - req->tag, spraidq->qid); - return BLK_EH_DONE; - } - - spraid_end_admin_request(req, cpu_to_le16(-EINVAL), 0, 0); - return BLK_EH_DONE; -} - static int spraid_get_ctrl_info(struct spraid_dev *hdev, struct spraid_ctrl_info *ctrl_info) { struct spraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + int ret; + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; memset(&admin_cmd, 0, sizeof(admin_cmd)); admin_cmd.get_info.opcode = SPRAID_ADMIN_GET_INFO; admin_cmd.get_info.type = SPRAID_GET_INFO_CTRL; + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + if (!ret) + memcpy(ctrl_info, data_ptr, sizeof(struct spraid_ctrl_info)); - return spraid_submit_admin_sync_cmd(hdev->admin_q, &admin_cmd, NULL, - ctrl_info, sizeof(struct spraid_ctrl_info), 0, 0, 0); + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, data_dma); + + return ret; } static int spraid_init_ctrl_info(struct spraid_dev *hdev) @@ -2416,6 +2448,11 @@ static int spraid_init_ctrl_info(struct spraid_dev *hdev) dev_info(hdev->dev, "[%s]sn = %s\n", __func__, hdev->ctrl_info->sn); dev_info(hdev->dev, "[%s]fr = %s\n", __func__, hdev->ctrl_info->fr); + if (!hdev->ctrl_info->aerl) + hdev->ctrl_info->aerl = 1; + if (hdev->ctrl_info->aerl > SPRAID_NR_AEN_COMMANDS) + hdev->ctrl_info->aerl = SPRAID_NR_AEN_COMMANDS; + return 0; } @@ -2444,99 +2481,54 @@ static void spraid_free_iod_ext_mem_pool(struct spraid_dev *hdev) mempool_destroy(hdev->iod_mempool); } -static int spraid_submit_user_cmd(struct request_queue *q, struct spraid_admin_command *cmd, - void __user *ubuffer, unsigned int bufflen, u32 *result, - unsigned int timeout) -{ - struct request *req; - struct bio *bio = NULL; - int ret; - - req = spraid_alloc_admin_request(q, cmd, 0); - if (IS_ERR(req)) - return PTR_ERR(req); - - req->timeout = timeout ? timeout : ADMIN_TIMEOUT; - spraid_admin_req(req)->flags |= SPRAID_REQ_USERCMD; - - if (ubuffer && bufflen) { - ret = blk_rq_map_user(q, req, NULL, ubuffer, bufflen, GFP_KERNEL); - if (ret) - goto out; - bio = req->bio; - } - blk_execute_rq(req->q, NULL, req, 0); - if (spraid_admin_req(req)->flags & SPRAID_REQ_CANCELLED) - ret = -EINTR; - else - ret = spraid_admin_req(req)->status; - if (result) { - result[0] = spraid_admin_req(req)->result0; - result[1] = spraid_admin_req(req)->result1; - } - if (bio) - blk_rq_unmap_user(bio); -out: - blk_mq_free_request(req); - return ret; -} - -static int spraid_user_admin_cmd(struct spraid_dev *hdev, - struct spraid_passthru_common_cmd __user *ucmd) +static int spraid_user_admin_cmd(struct spraid_dev *hdev, struct bsg_job *job) { - struct spraid_passthru_common_cmd cmd; + struct spraid_bsg_request *bsg_req = job->request; + struct spraid_passthru_common_cmd *cmd = &(bsg_req->admcmd); struct spraid_admin_command admin_cmd; - u32 timeout = 0; + u32 timeout = msecs_to_jiffies(cmd->timeout_ms); + u32 result[2] = {0}; int status; - if (!capable(CAP_SYS_ADMIN)) { - dev_err(hdev->dev, "Current user hasn't administrator right, reject service\n"); - return -EACCES; - } - - if (copy_from_user(&cmd, ucmd, sizeof(cmd))) { - dev_err(hdev->dev, "Copy command from user space to kernel space failed\n"); - return -EFAULT; - } - - if (cmd.flags) { - dev_err(hdev->dev, "Invalid flags in user command\n"); - return -EINVAL; + if (hdev->state >= SPRAID_RESETTING) { + dev_err(hdev->dev, "[%s] err, host state:[%d] is not right\n", + __func__, hdev->state); + return -EBUSY; } - dev_info(hdev->dev, "user_admin_cmd opcode: 0x%x, subopcode: 0x%x\n", - cmd.opcode, cmd.cdw2 & 0x7ff); + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x] init\n", + __func__, cmd->opcode, cmd->info_0.subopcode); memset(&admin_cmd, 0, sizeof(admin_cmd)); - admin_cmd.common.opcode = cmd.opcode; - admin_cmd.common.flags = cmd.flags; - admin_cmd.common.hdid = cpu_to_le32(cmd.nsid); - admin_cmd.common.cdw2[0] = cpu_to_le32(cmd.cdw2); - admin_cmd.common.cdw2[1] = cpu_to_le32(cmd.cdw3); - admin_cmd.common.cdw10 = cpu_to_le32(cmd.cdw10); - admin_cmd.common.cdw11 = cpu_to_le32(cmd.cdw11); - admin_cmd.common.cdw12 = cpu_to_le32(cmd.cdw12); - admin_cmd.common.cdw13 = cpu_to_le32(cmd.cdw13); - admin_cmd.common.cdw14 = cpu_to_le32(cmd.cdw14); - admin_cmd.common.cdw15 = cpu_to_le32(cmd.cdw15); - - if (cmd.timeout_ms) - timeout = msecs_to_jiffies(cmd.timeout_ms); - - status = spraid_submit_user_cmd(hdev->admin_q, &admin_cmd, - (void __user *)(uintptr_t)cmd.addr, cmd.info_1.data_len, - &cmd.result0, timeout); - - dev_info(hdev->dev, "user_admin_cmd status: 0x%x, result0: 0x%x, result1: 0x%x\n", - status, cmd.result0, cmd.result1); + admin_cmd.common.opcode = cmd->opcode; + admin_cmd.common.flags = cmd->flags; + admin_cmd.common.hdid = cpu_to_le32(cmd->nsid); + admin_cmd.common.cdw2[0] = cpu_to_le32(cmd->cdw2); + admin_cmd.common.cdw2[1] = cpu_to_le32(cmd->cdw3); + admin_cmd.common.cdw10 = cpu_to_le32(cmd->cdw10); + admin_cmd.common.cdw11 = cpu_to_le32(cmd->cdw11); + admin_cmd.common.cdw12 = cpu_to_le32(cmd->cdw12); + admin_cmd.common.cdw13 = cpu_to_le32(cmd->cdw13); + admin_cmd.common.cdw14 = cpu_to_le32(cmd->cdw14); + admin_cmd.common.cdw15 = cpu_to_le32(cmd->cdw15); + + status = spraid_bsg_map_data(hdev, job, &admin_cmd); + if (status) { + dev_err(hdev->dev, "[%s] err, map data failed\n", __func__); + return status; + } + status = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, &result[0], &result[1], timeout); if (status >= 0) { - if (put_user(cmd.result0, &ucmd->result0)) - return -EFAULT; - if (put_user(cmd.result1, &ucmd->result1)) - return -EFAULT; + job->reply_len = sizeof(result); + memcpy(job->reply, result, sizeof(result)); } + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x], status[0x%x] result0[0x%x] result1[0x%x]\n", + __func__, cmd->opcode, cmd->info_0.subopcode, status, result[0], result[1]); + + spraid_bsg_unmap_data(hdev, job); + return status; } @@ -2548,8 +2540,8 @@ static int spraid_alloc_ioq_ptcmds(struct spraid_dev *hdev) INIT_LIST_HEAD(&hdev->ioq_pt_list); spin_lock_init(&hdev->ioq_pt_lock); - hdev->ioq_ptcmds = kcalloc_node(ptnum, sizeof(struct spraid_ioq_ptcmd), - GFP_KERNEL, hdev->numa_node); + hdev->ioq_ptcmds = kcalloc_node(ptnum, sizeof(struct spraid_cmd), + GFP_KERNEL, hdev->numa_node); if (!hdev->ioq_ptcmds) { dev_err(hdev->dev, "Alloc ioq_ptcmds failed\n"); @@ -2567,55 +2559,35 @@ static int spraid_alloc_ioq_ptcmds(struct spraid_dev *hdev) return 0; } -static struct spraid_ioq_ptcmd *spraid_get_ioq_ptcmd(struct spraid_dev *hdev) +static void spraid_free_ioq_ptcmds(struct spraid_dev *hdev) { - struct spraid_ioq_ptcmd *cmd = NULL; - unsigned long flags; - - spin_lock_irqsave(&hdev->ioq_pt_lock, flags); - if (list_empty(&hdev->ioq_pt_list)) { - spin_unlock_irqrestore(&hdev->ioq_pt_lock, flags); - dev_err(hdev->dev, "err, ioq ptcmd list empty\n"); - return NULL; - } - cmd = list_entry((&hdev->ioq_pt_list)->next, struct spraid_ioq_ptcmd, list); - list_del_init(&cmd->list); - spin_unlock_irqrestore(&hdev->ioq_pt_lock, flags); - - WRITE_ONCE(cmd->state, SPRAID_CMD_IDLE); - - return cmd; -} - -static void spraid_put_ioq_ptcmd(struct spraid_dev *hdev, struct spraid_ioq_ptcmd *cmd) -{ - unsigned long flags; + kfree(hdev->ioq_ptcmds); + hdev->ioq_ptcmds = NULL; - spin_lock_irqsave(&hdev->ioq_pt_lock, flags); - list_add(&cmd->list, (&hdev->ioq_pt_list)->next); - spin_unlock_irqrestore(&hdev->ioq_pt_lock, flags); + INIT_LIST_HEAD(&hdev->ioq_pt_list); } static int spraid_submit_ioq_sync_cmd(struct spraid_dev *hdev, struct spraid_ioq_command *cmd, - u32 *result, void **sense, u32 timeout) + u32 *result, u32 *reslen, u32 timeout) { - struct spraid_queue *ioq; int ret; dma_addr_t sense_dma; - struct spraid_ioq_ptcmd *pt_cmd = spraid_get_ioq_ptcmd(hdev); - - *sense = NULL; + struct spraid_queue *ioq; + void *sense_addr = NULL; + struct spraid_cmd *pt_cmd = spraid_get_cmd(hdev, SPRAID_CMD_IOPT); - if (!pt_cmd) + if (!pt_cmd) { + dev_err(hdev->dev, "err, get ioq cmd failed\n"); return -EFAULT; + } - dev_info(hdev->dev, "[%s] ptcmd, cid[%d], qid[%d]\n", __func__, pt_cmd->cid, pt_cmd->qid); + timeout = timeout ? timeout : ADMIN_TIMEOUT; init_completion(&pt_cmd->cmd_done); ioq = &hdev->queues[pt_cmd->qid]; ret = pt_cmd->cid * SCSI_SENSE_BUFFERSIZE; - pt_cmd->priv = ioq->sense + ret; + sense_addr = ioq->sense + ret; sense_dma = ioq->sense_dma_addr + ret; cmd->common.sense_addr = cpu_to_le64(sense_dma); @@ -2625,260 +2597,87 @@ static int spraid_submit_ioq_sync_cmd(struct spraid_dev *hdev, struct spraid_ioq spraid_submit_cmd(ioq, cmd); if (!wait_for_completion_timeout(&pt_cmd->cmd_done, timeout)) { - dev_err(hdev->dev, "[%s] cid[%d], qid[%d] timeout\n", - __func__, pt_cmd->cid, pt_cmd->qid); + dev_err(hdev->dev, "[%s] cid[%d] qid[%d] timeout, opcode[0x%x] subopcode[0x%x]\n", + __func__, pt_cmd->cid, pt_cmd->qid, cmd->common.opcode, + (le32_to_cpu(cmd->common.cdw3[0]) & 0xffff)); WRITE_ONCE(pt_cmd->state, SPRAID_CMD_TIMEOUT); + spraid_put_cmd(hdev, pt_cmd, SPRAID_CMD_IOPT); return -EINVAL; } - if (result) { - result[0] = pt_cmd->result0; - result[1] = pt_cmd->result1; + if (result && reslen) { + if ((pt_cmd->status & 0x17f) == 0x101) { + memcpy(result, sense_addr, SCSI_SENSE_BUFFERSIZE); + *reslen = SCSI_SENSE_BUFFERSIZE; + } } - if ((pt_cmd->status & 0x17f) == 0x101) - *sense = pt_cmd->priv; + spraid_put_cmd(hdev, pt_cmd, SPRAID_CMD_IOPT); return pt_cmd->status; } -static int spraid_user_ioq_cmd(struct spraid_dev *hdev, - struct spraid_ioq_passthru_cmd __user *ucmd) +static int spraid_user_ioq_cmd(struct spraid_dev *hdev, struct bsg_job *job) { - struct spraid_ioq_passthru_cmd cmd; + struct spraid_bsg_request *bsg_req = (struct spraid_bsg_request *)(job->request); + struct spraid_ioq_passthru_cmd *cmd = &(bsg_req->ioqcmd); struct spraid_ioq_command ioq_cmd; - u32 timeout = 0; int status = 0; - u8 *data_ptr = NULL; - dma_addr_t data_dma; - enum dma_data_direction dma_dir = DMA_NONE; - void *sense = NULL; - - if (!capable(CAP_SYS_ADMIN)) { - dev_err(hdev->dev, "Current user hasn't administrator right, reject service\n"); - return -EACCES; - } - - if (copy_from_user(&cmd, ucmd, sizeof(cmd))) { - dev_err(hdev->dev, "Copy command from user space to kernel space failed\n"); - return -EFAULT; - } + u32 timeout = msecs_to_jiffies(cmd->timeout_ms); - if (cmd.data_len > PAGE_SIZE) { + if (cmd->data_len > PAGE_SIZE) { dev_err(hdev->dev, "[%s] data len bigger than 4k\n", __func__); return -EFAULT; } - dev_info(hdev->dev, "[%s] opcode: 0x%x, subopcode: 0x%x, datalen: %d\n", - __func__, cmd.opcode, cmd.info_1.subopcode, cmd.data_len); - - if (cmd.addr && cmd.data_len) { - data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &data_dma, GFP_KERNEL); - if (!data_ptr) - return -ENOMEM; - - dma_dir = (cmd.opcode & 1) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; + if (hdev->state != SPRAID_LIVE) { + dev_err(hdev->dev, "[%s] err, host state:[%d] is not live\n", + __func__, hdev->state); + return -EBUSY; } - if (dma_dir == DMA_TO_DEVICE) { - if (copy_from_user(data_ptr, (void __user *)(uintptr_t)cmd.addr, cmd.data_len)) { - dev_err(hdev->dev, "[%s] copy user data failed\n", __func__); - status = -EFAULT; - goto free_dma_mem; - } - } + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x] init, datalen[%d]\n", + __func__, cmd->opcode, cmd->info_1.subopcode, cmd->data_len); memset(&ioq_cmd, 0, sizeof(ioq_cmd)); - ioq_cmd.common.opcode = cmd.opcode; - ioq_cmd.common.flags = cmd.flags; - ioq_cmd.common.hdid = cpu_to_le32(cmd.nsid); - ioq_cmd.common.sense_len = cpu_to_le16(cmd.info_0.res_sense_len); - ioq_cmd.common.cdb_len = cmd.info_0.cdb_len; - ioq_cmd.common.rsvd2 = cmd.info_0.rsvd0; - ioq_cmd.common.cdw3[0] = cpu_to_le32(cmd.cdw3); - ioq_cmd.common.cdw3[1] = cpu_to_le32(cmd.cdw4); - ioq_cmd.common.cdw3[2] = cpu_to_le32(cmd.cdw5); - ioq_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); - - ioq_cmd.common.cdw10[0] = cpu_to_le32(cmd.cdw10); - ioq_cmd.common.cdw10[1] = cpu_to_le32(cmd.cdw11); - ioq_cmd.common.cdw10[2] = cpu_to_le32(cmd.cdw12); - ioq_cmd.common.cdw10[3] = cpu_to_le32(cmd.cdw13); - ioq_cmd.common.cdw10[4] = cpu_to_le32(cmd.cdw14); - ioq_cmd.common.cdw10[5] = cpu_to_le32(cmd.data_len); - - memcpy(ioq_cmd.common.cdb, &cmd.cdw16, cmd.info_0.cdb_len); - - ioq_cmd.common.cdw26[0] = cpu_to_le32(cmd.cdw26[0]); - ioq_cmd.common.cdw26[1] = cpu_to_le32(cmd.cdw26[1]); - ioq_cmd.common.cdw26[2] = cpu_to_le32(cmd.cdw26[2]); - ioq_cmd.common.cdw26[3] = cpu_to_le32(cmd.cdw26[3]); - - if (cmd.timeout_ms) - timeout = msecs_to_jiffies(cmd.timeout_ms); - timeout = timeout ? timeout : ADMIN_TIMEOUT; - - status = spraid_submit_ioq_sync_cmd(hdev, &ioq_cmd, &cmd.result0, &sense, timeout); - - if (status >= 0) { - if (put_user(cmd.result0, &ucmd->result0)) { - status = -EFAULT; - goto free_dma_mem; - } - if (put_user(cmd.result1, &ucmd->result1)) { - status = -EFAULT; - goto free_dma_mem; - } - if (dma_dir == DMA_FROM_DEVICE && - copy_to_user((void __user *)(uintptr_t)cmd.addr, data_ptr, cmd.data_len)) { - status = -EFAULT; - goto free_dma_mem; - } - } - - if (sense) { - if (copy_to_user((void *__user *)(uintptr_t)cmd.sense_addr, - sense, cmd.info_0.res_sense_len)) { - status = -EFAULT; - goto free_dma_mem; - } - } - -free_dma_mem: - if (data_ptr) - dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, data_dma); - - return status; - -} - -static int spraid_reset_work_sync(struct spraid_dev *hdev); - -static int spraid_user_reset_cmd(struct spraid_dev *hdev) -{ - int ret; - - dev_info(hdev->dev, "[%s] start user reset cmd\n", __func__); - ret = spraid_reset_work_sync(hdev); - dev_info(hdev->dev, "[%s] stop user reset cmd[%d]\n", __func__, ret); - - return ret; -} - -static int hdev_open(struct inode *inode, struct file *file) -{ - struct spraid_dev *hdev = - container_of(inode->i_cdev, struct spraid_dev, cdev); - file->private_data = hdev; - return 0; -} - -static long hdev_ioctl(struct file *file, u32 cmd, unsigned long arg) -{ - struct spraid_dev *hdev = file->private_data; - void __user *argp = (void __user *)arg; - - switch (cmd) { - case SPRAID_IOCTL_ADMIN_CMD: - return spraid_user_admin_cmd(hdev, argp); - case SPRAID_IOCTL_IOQ_CMD: - return spraid_user_ioq_cmd(hdev, argp); - case SPRAID_IOCTL_RESET_CMD: - return spraid_user_reset_cmd(hdev); - default: - return -ENOTTY; - } -} - -static const struct file_operations spraid_dev_fops = { - .owner = THIS_MODULE, - .open = hdev_open, - .unlocked_ioctl = hdev_ioctl, - .compat_ioctl = hdev_ioctl, -}; - -static int spraid_create_cdev(struct spraid_dev *hdev) -{ - int ret; - - device_initialize(&hdev->ctrl_device); - hdev->ctrl_device.devt = MKDEV(MAJOR(spraid_chr_devt), hdev->instance); - hdev->ctrl_device.class = spraid_class; - hdev->ctrl_device.parent = hdev->dev; - dev_set_drvdata(&hdev->ctrl_device, hdev); - ret = dev_set_name(&hdev->ctrl_device, "spraid%d", hdev->instance); - if (ret) - return ret; - cdev_init(&hdev->cdev, &spraid_dev_fops); - hdev->cdev.owner = THIS_MODULE; - ret = cdev_device_add(&hdev->cdev, &hdev->ctrl_device); - if (ret) { - dev_err(hdev->dev, "Add cdev failed, ret: %d", ret); - put_device(&hdev->ctrl_device); - kfree_const(hdev->ctrl_device.kobj.name); - return ret; - } - - return 0; -} - -static inline void spraid_remove_cdev(struct spraid_dev *hdev) -{ - cdev_device_del(&hdev->cdev, &hdev->ctrl_device); -} - -static const struct blk_mq_ops spraid_admin_mq_ops = { - .queue_rq = spraid_queue_admin_rq, - .complete = spraid_complete_admin_rq, - .init_hctx = spraid_admin_init_hctx, - .init_request = spraid_admin_init_request, - .timeout = spraid_admin_timeout, -}; - -static void spraid_remove_admin_tagset(struct spraid_dev *hdev) -{ - if (hdev->admin_q && !blk_queue_dying(hdev->admin_q)) { - blk_mq_unquiesce_queue(hdev->admin_q); - blk_cleanup_queue(hdev->admin_q); - blk_mq_free_tag_set(&hdev->admin_tagset); + ioq_cmd.common.opcode = cmd->opcode; + ioq_cmd.common.flags = cmd->flags; + ioq_cmd.common.hdid = cpu_to_le32(cmd->nsid); + ioq_cmd.common.sense_len = cpu_to_le16(cmd->info_0.res_sense_len); + ioq_cmd.common.cdb_len = cmd->info_0.cdb_len; + ioq_cmd.common.rsvd2 = cmd->info_0.rsvd0; + ioq_cmd.common.cdw3[0] = cpu_to_le32(cmd->cdw3); + ioq_cmd.common.cdw3[1] = cpu_to_le32(cmd->cdw4); + ioq_cmd.common.cdw3[2] = cpu_to_le32(cmd->cdw5); + + ioq_cmd.common.cdw10[0] = cpu_to_le32(cmd->cdw10); + ioq_cmd.common.cdw10[1] = cpu_to_le32(cmd->cdw11); + ioq_cmd.common.cdw10[2] = cpu_to_le32(cmd->cdw12); + ioq_cmd.common.cdw10[3] = cpu_to_le32(cmd->cdw13); + ioq_cmd.common.cdw10[4] = cpu_to_le32(cmd->cdw14); + ioq_cmd.common.cdw10[5] = cpu_to_le32(cmd->data_len); + + memcpy(ioq_cmd.common.cdb, &cmd->cdw16, cmd->info_0.cdb_len); + + ioq_cmd.common.cdw26[0] = cpu_to_le32(cmd->cdw26[0]); + ioq_cmd.common.cdw26[1] = cpu_to_le32(cmd->cdw26[1]); + ioq_cmd.common.cdw26[2] = cpu_to_le32(cmd->cdw26[2]); + ioq_cmd.common.cdw26[3] = cpu_to_le32(cmd->cdw26[3]); + + status = spraid_bsg_map_data(hdev, job, (struct spraid_admin_command *)&ioq_cmd); + if (status) { + dev_err(hdev->dev, "[%s] err, map data failed\n", __func__); + return status; } -} -static int spraid_alloc_admin_tags(struct spraid_dev *hdev) -{ - if (!hdev->admin_q) { - hdev->admin_tagset.ops = &spraid_admin_mq_ops; - hdev->admin_tagset.nr_hw_queues = 1; + status = spraid_submit_ioq_sync_cmd(hdev, &ioq_cmd, job->reply, &job->reply_len, timeout); - hdev->admin_tagset.queue_depth = SPRAID_AQ_MQ_TAG_DEPTH; - hdev->admin_tagset.timeout = ADMIN_TIMEOUT; - hdev->admin_tagset.numa_node = hdev->numa_node; - hdev->admin_tagset.cmd_size = - spraid_cmd_size(hdev, true, false); - hdev->admin_tagset.flags = BLK_MQ_F_NO_SCHED; - hdev->admin_tagset.driver_data = hdev; + dev_info(hdev->dev, "[%s] opcode[0x%x] subopcode[0x%x], status[0x%x], reply_len[%d]\n", + __func__, cmd->opcode, cmd->info_1.subopcode, status, job->reply_len); - if (blk_mq_alloc_tag_set(&hdev->admin_tagset)) { - dev_err(hdev->dev, "Allocate admin tagset failed\n"); - return -ENOMEM; - } + spraid_bsg_unmap_data(hdev, job); - hdev->admin_q = blk_mq_init_queue(&hdev->admin_tagset); - if (IS_ERR(hdev->admin_q)) { - dev_err(hdev->dev, "Initialize admin request queue failed\n"); - blk_mq_free_tag_set(&hdev->admin_tagset); - return -ENOMEM; - } - if (!blk_get_queue(hdev->admin_q)) { - dev_err(hdev->dev, "Get admin request queue failed\n"); - spraid_remove_admin_tagset(hdev); - hdev->admin_q = NULL; - return -ENODEV; - } - } else { - blk_mq_unquiesce_queue(hdev->admin_q); - } - return 0; + return status; } static bool spraid_check_scmd_completed(struct scsi_cmnd *scmd) @@ -2891,7 +2690,7 @@ static bool spraid_check_scmd_completed(struct scsi_cmnd *scmd) spraid_get_tag_from_scmd(scmd, &hwq, &cid); spraidq = &hdev->queues[hwq]; if (READ_ONCE(iod->state) == SPRAID_CMD_COMPLETE || spraid_poll_cq(spraidq, cid)) { - dev_warn(hdev->dev, "cid[%d], qid[%d] has been completed\n", + dev_warn(hdev->dev, "cid[%d] qid[%d] has been completed\n", cid, spraidq->qid); return true; } @@ -2927,8 +2726,7 @@ static int spraid_send_abort_cmd(struct spraid_dev *hdev, u32 hdid, u16 qid, u16 admin_cmd.abort.sqid = cpu_to_le16(qid); admin_cmd.abort.cid = cpu_to_le16(cid); - return spraid_submit_admin_sync_cmd(hdev->admin_q, &admin_cmd, NULL, - NULL, 0, 0, 0, 0); + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); } /* send reset command by admin quueue temporary */ @@ -2941,8 +2739,7 @@ static int spraid_send_reset_cmd(struct spraid_dev *hdev, int type, u32 hdid) admin_cmd.reset.hdid = cpu_to_le32(hdid); admin_cmd.reset.type = type; - return spraid_submit_admin_sync_cmd(hdev->admin_q, &admin_cmd, NULL, - NULL, 0, 0, 0, 0); + return spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); } static bool spraid_change_host_state(struct spraid_dev *hdev, enum spraid_state newstate) @@ -3022,7 +2819,7 @@ static void spraid_back_fault_cqe(struct spraid_queue *ioq, struct spraid_comple scsi_dma_unmap(scmd); spraid_free_iod_res(hdev, iod); scmd->scsi_done(scmd); - dev_warn(hdev->dev, "Back fault CQE, cid[%d], qid[%d]\n", + dev_warn(hdev->dev, "Back fault CQE, cid[%d] qid[%d]\n", cqe->cmd_id, ioq->qid); } @@ -3032,6 +2829,8 @@ static void spraid_back_all_io(struct spraid_dev *hdev) struct spraid_queue *ioq; struct spraid_completion cqe = { 0 }; + scsi_block_requests(hdev->shost); + for (i = 1; i <= hdev->shost->nr_hw_queues; i++) { ioq = &hdev->queues[i]; for (j = 0; j < hdev->shost->can_queue; j++) { @@ -3039,6 +2838,8 @@ static void spraid_back_all_io(struct spraid_dev *hdev) spraid_back_fault_cqe(ioq, &cqe); } } + + scsi_unblock_requests(hdev->shost); } static void spraid_dev_disable(struct spraid_dev *hdev, bool shutdown) @@ -3106,17 +2907,13 @@ static void spraid_reset_work(struct work_struct *work) if (ret) goto pci_disable; - ret = spraid_alloc_admin_tags(hdev); - if (ret) - goto pci_disable; - ret = spraid_setup_io_queues(hdev); if (ret || hdev->online_queues <= hdev->shost->nr_hw_queues) goto pci_disable; spraid_change_host_state(hdev, SPRAID_LIVE); - spraid_send_aen(hdev); + spraid_send_all_aen(hdev); return; @@ -3174,8 +2971,8 @@ static int spraid_abort_handler(struct scsi_cmnd *scmd) scsi_print_command(scmd); - if (!spraid_wait_abnl_cmd_done(iod) || spraid_check_scmd_completed(scmd) || - hdev->state != SPRAID_LIVE) + if (hdev->state != SPRAID_LIVE || !spraid_wait_abnl_cmd_done(iod) || + spraid_check_scmd_completed(scmd)) return SUCCESS; hostdata = scmd->device->hostdata; @@ -3206,8 +3003,8 @@ static int spraid_tgt_reset_handler(struct scsi_cmnd *scmd) scsi_print_command(scmd); - if (!spraid_wait_abnl_cmd_done(iod) || spraid_check_scmd_completed(scmd) || - hdev->state != SPRAID_LIVE) + if (hdev->state != SPRAID_LIVE || !spraid_wait_abnl_cmd_done(iod) || + spraid_check_scmd_completed(scmd)) return SUCCESS; hostdata = scmd->device->hostdata; @@ -3241,8 +3038,8 @@ static int spraid_bus_reset_handler(struct scsi_cmnd *scmd) scsi_print_command(scmd); - if (!spraid_wait_abnl_cmd_done(iod) || spraid_check_scmd_completed(scmd) || - hdev->state != SPRAID_LIVE) + if (hdev->state != SPRAID_LIVE || !spraid_wait_abnl_cmd_done(iod) || + spraid_check_scmd_completed(scmd)) return SUCCESS; hostdata = scmd->device->hostdata; @@ -3272,7 +3069,7 @@ static int spraid_shost_reset_handler(struct scsi_cmnd *scmd) struct spraid_dev *hdev = shost_priv(scmd->device->host); scsi_print_command(scmd); - if (spraid_check_scmd_completed(scmd) || hdev->state != SPRAID_LIVE) + if (hdev->state != SPRAID_LIVE || spraid_check_scmd_completed(scmd)) return SUCCESS; spraid_get_tag_from_scmd(scmd, &hwq, &cid); @@ -3288,6 +3085,62 @@ static int spraid_shost_reset_handler(struct scsi_cmnd *scmd) return SUCCESS; } +static pci_ers_result_t spraid_pci_error_detected(struct pci_dev *pdev, + pci_channel_state_t state) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "enter pci error detect, state:%d\n", state); + + switch (state) { + case pci_channel_io_normal: + dev_warn(hdev->dev, "channel is normal, do nothing\n"); + + return PCI_ERS_RESULT_CAN_RECOVER; + case pci_channel_io_frozen: + dev_warn(hdev->dev, "channel io frozen, need reset controller\n"); + + scsi_block_requests(hdev->shost); + + spraid_change_host_state(hdev, SPRAID_RESETTING); + + return PCI_ERS_RESULT_NEED_RESET; + case pci_channel_io_perm_failure: + dev_warn(hdev->dev, "channel io failure, request disconnect\n"); + + return PCI_ERS_RESULT_DISCONNECT; + } + + return PCI_ERS_RESULT_NEED_RESET; +} + +static pci_ers_result_t spraid_pci_slot_reset(struct pci_dev *pdev) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "restart after slot reset\n"); + + pci_restore_state(pdev); + + if (!queue_work(spraid_wq, &hdev->reset_work)) { + dev_err(hdev->dev, "[%s] err, the device is resetting state\n", __func__); + return PCI_ERS_RESULT_NONE; + } + + flush_work(&hdev->reset_work); + + scsi_unblock_requests(hdev->shost); + + return PCI_ERS_RESULT_RECOVERED; +} + +static void spraid_reset_done(struct pci_dev *pdev) +{ + struct spraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "enter spraid reset done\n"); +} + static ssize_t csts_pp_show(struct device *cdev, struct device_attribute *attr, char *buf) { struct Scsi_Host *shost = class_to_shost(cdev); @@ -3347,7 +3200,7 @@ static ssize_t fw_version_show(struct device *cdev, struct device_attribute *att struct Scsi_Host *shost = class_to_shost(cdev); struct spraid_dev *hdev = shost_priv(shost); - return snprintf(buf, sizeof(hdev->ctrl_info->fr), "%s\n", hdev->ctrl_info->fr); + return snprintf(buf, PAGE_SIZE, "%s\n", hdev->ctrl_info->fr); } static DEVICE_ATTR_RO(csts_pp); @@ -3365,6 +3218,185 @@ static struct device_attribute *spraid_host_attrs[] = { NULL, }; +static int spraid_get_vd_info(struct spraid_dev *hdev, struct spraid_vd_info *vd_info, u16 vid) +{ + struct spraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + int ret; + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.usr_cmd.opcode = USR_CMD_READ; + admin_cmd.usr_cmd.info_0.subopcode = cpu_to_le16(USR_CMD_VDINFO); + admin_cmd.usr_cmd.info_1.data_len = cpu_to_le16(USR_CMD_RDLEN); + admin_cmd.usr_cmd.info_1.param_len = cpu_to_le16(VDINFO_PARAM_LEN); + admin_cmd.usr_cmd.cdw10 = cpu_to_le32(vid); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + if (!ret) + memcpy(vd_info, data_ptr, sizeof(struct spraid_vd_info)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, data_dma); + + return ret; +} + +static int spraid_get_bgtask(struct spraid_dev *hdev, struct spraid_bgtask *bgtask) +{ + struct spraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t data_dma = 0; + int ret; + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &data_dma, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.usr_cmd.opcode = USR_CMD_READ; + admin_cmd.usr_cmd.info_0.subopcode = cpu_to_le16(USR_CMD_BGTASK); + admin_cmd.usr_cmd.info_1.data_len = cpu_to_le16(USR_CMD_RDLEN); + admin_cmd.common.dptr.prp1 = cpu_to_le64(data_dma); + + ret = spraid_submit_admin_sync_cmd(hdev, &admin_cmd, NULL, NULL, 0); + if (!ret) + memcpy(bgtask, data_ptr, sizeof(struct spraid_bgtask)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, data_dma); + + return ret; +} + +static ssize_t raid_level_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct spraid_dev *hdev; + struct spraid_vd_info *vd_info; + struct spraid_sdev_hostdata *hostdata; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !SPRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = spraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) + vd_info->rg_level = ARRAY_SIZE(raid_levels) - 1; + + ret = (vd_info->rg_level < ARRAY_SIZE(raid_levels)) ? + vd_info->rg_level : (ARRAY_SIZE(raid_levels) - 1); + + kfree(vd_info); + + return snprintf(buf, PAGE_SIZE, "RAID-%s\n", raid_levels[ret]); +} + +static ssize_t raid_state_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct spraid_dev *hdev; + struct spraid_vd_info *vd_info; + struct spraid_sdev_hostdata *hostdata; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !SPRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = spraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) { + vd_info->vd_status = 0; + vd_info->rg_id = 0xff; + } + + ret = (vd_info->vd_status < ARRAY_SIZE(raid_states)) ? vd_info->vd_status : 0; + + kfree(vd_info); + + return snprintf(buf, PAGE_SIZE, "%s\n", raid_states[ret]); +} + +static ssize_t raid_resync_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct spraid_dev *hdev; + struct spraid_vd_info *vd_info; + struct spraid_bgtask *bgtask; + struct spraid_sdev_hostdata *hostdata; + u8 rg_id, i, progress = 0; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !SPRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = spraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) + goto out; + + rg_id = vd_info->rg_id; + + bgtask = (struct spraid_bgtask *)vd_info; + ret = spraid_get_bgtask(hdev, bgtask); + if (ret) + goto out; + for (i = 0; i < bgtask->task_num; i++) { + if ((bgtask->bgtask[i].type == BGTASK_TYPE_REBUILD) && + (le16_to_cpu(bgtask->bgtask[i].vd_id) == rg_id)) + progress = bgtask->bgtask[i].progress; + } + +out: + kfree(vd_info); + return snprintf(buf, PAGE_SIZE, "%d\n", progress); +} + +static DEVICE_ATTR_RO(raid_level); +static DEVICE_ATTR_RO(raid_state); +static DEVICE_ATTR_RO(raid_resync); + +static struct device_attribute *spraid_dev_attrs[] = { + &dev_attr_raid_level, + &dev_attr_raid_state, + &dev_attr_raid_resync, + NULL, +}; + +static struct pci_error_handlers spraid_err_handler = { + .error_detected = spraid_pci_error_detected, + .slot_reset = spraid_pci_slot_reset, + .reset_done = spraid_reset_done, +}; + +static int spraid_sysfs_host_reset(struct Scsi_Host *shost, int reset_type) +{ + int ret; + struct spraid_dev *hdev = shost_priv(shost); + + dev_info(hdev->dev, "[%s] start sysfs host reset cmd\n", __func__); + ret = spraid_reset_work_sync(hdev); + dev_info(hdev->dev, "[%s] stop sysfs host reset cmd[%d]\n", __func__, ret); + + return ret; +} + static struct scsi_host_template spraid_driver_template = { .module = THIS_MODULE, .name = "Ramaxel Logic spraid driver", @@ -3379,9 +3411,11 @@ static struct scsi_host_template spraid_driver_template = { .eh_bus_reset_handler = spraid_bus_reset_handler, .eh_host_reset_handler = spraid_shost_reset_handler, .change_queue_depth = scsi_change_queue_depth, - .host_tagset = 1, + .host_tagset = 0, .this_id = -1, .shost_attrs = spraid_host_attrs, + .sdev_attrs = spraid_dev_attrs, + .host_reset = spraid_sysfs_host_reset, }; static void spraid_shutdown(struct pci_dev *pdev) @@ -3392,11 +3426,53 @@ static void spraid_shutdown(struct pci_dev *pdev) spraid_disable_admin_queue(hdev, true); } +/* bsg dispatch user command */ +static int spraid_bsg_host_dispatch(struct bsg_job *job) +{ + struct Scsi_Host *shost = dev_to_shost(job->dev); + struct spraid_dev *hdev = shost_priv(shost); + struct request *rq = blk_mq_rq_from_pdu(job); + struct spraid_bsg_request *bsg_req = job->request; + int ret = 0; + + dev_info(hdev->dev, "[%s] msgcode[%d], msglen[%d], timeout[%d], req_nsge[%d], req_len[%d]\n", + __func__, bsg_req->msgcode, job->request_len, rq->timeout, + job->request_payload.sg_cnt, job->request_payload.payload_len); + + job->reply_len = 0; + + switch (bsg_req->msgcode) { + case SPRAID_BSG_ADM: + ret = spraid_user_admin_cmd(hdev, job); + break; + case SPRAID_BSG_IOQ: + ret = spraid_user_ioq_cmd(hdev, job); + break; + default: + dev_info(hdev->dev, "[%s] unsupport msgcode[%d]\n", __func__, bsg_req->msgcode); + break; + } + + if (ret > 0) + ret = ret | (ret << 8); + + bsg_job_done(job, ret, 0); + return 0; +} + +static inline void spraid_remove_bsg(struct spraid_dev *hdev) +{ + if (hdev->bsg_queue) { + bsg_unregister_queue(hdev->bsg_queue); + blk_cleanup_queue(hdev->bsg_queue); + } +} static int spraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) { struct spraid_dev *hdev; struct Scsi_Host *shost; int node, ret; + char bsg_name[15]; shost = scsi_host_alloc(&spraid_driver_template, sizeof(*hdev)); if (!shost) { @@ -3421,10 +3497,10 @@ static int spraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) goto put_dev; init_rwsem(&hdev->devices_rwsem); - INIT_WORK(&hdev->aen_work, spraid_async_event_work); INIT_WORK(&hdev->scan_work, spraid_scan_work); INIT_WORK(&hdev->timesyn_work, spraid_timesyn_work); INIT_WORK(&hdev->reset_work, spraid_reset_work); + INIT_WORK(&hdev->fw_act_work, spraid_fw_act_work); spin_lock_init(&hdev->state_lock); ret = spraid_alloc_resources(hdev); @@ -3439,17 +3515,13 @@ static int spraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) if (ret) goto pci_disable; - ret = spraid_alloc_admin_tags(hdev); - if (ret) - goto disable_admin_q; - ret = spraid_init_ctrl_info(hdev); if (ret) - goto free_admin_tagset; + goto disable_admin_q; ret = spraid_alloc_iod_ext_mem_pool(hdev); if (ret) - goto free_admin_tagset; + goto disable_admin_q; ret = spraid_setup_io_queues(hdev); if (ret) @@ -3464,9 +3536,15 @@ static int spraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) goto remove_io_queues; } - ret = spraid_create_cdev(hdev); - if (ret) + snprintf(bsg_name, sizeof(bsg_name), "spraid%d", shost->host_no); + hdev->bsg_queue = bsg_setup_queue(&shost->shost_gendev, bsg_name, + spraid_bsg_host_dispatch, NULL, + spraid_cmd_size(hdev, true, false)); + if (IS_ERR(hdev->bsg_queue)) { + dev_err(hdev->dev, "err, setup bsg failed\n"); + hdev->bsg_queue = NULL; goto remove_io_queues; + } if (hdev->online_queues == SPRAID_ADMIN_QUEUE_NUM) { dev_warn(hdev->dev, "warn only admin queue can be used\n"); @@ -3475,11 +3553,11 @@ static int spraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) hdev->state = SPRAID_LIVE; - spraid_send_aen(hdev); + spraid_send_all_aen(hdev); ret = spraid_dev_list_init(hdev); if (ret) - goto remove_cdev; + goto remove_bsg; ret = spraid_configure_timestamp(hdev); if (ret) @@ -3487,20 +3565,18 @@ static int spraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) ret = spraid_alloc_ioq_ptcmds(hdev); if (ret) - goto remove_cdev; + goto remove_bsg; scsi_scan_host(hdev->shost); return 0; -remove_cdev: - spraid_remove_cdev(hdev); +remove_bsg: + spraid_remove_bsg(hdev); remove_io_queues: spraid_remove_io_queues(hdev); free_iod_mempool: spraid_free_iod_ext_mem_pool(hdev); -free_admin_tagset: - spraid_remove_admin_tagset(hdev); disable_admin_q: spraid_disable_admin_queue(hdev, false); pci_disable: @@ -3524,22 +3600,17 @@ static void spraid_remove(struct pci_dev *pdev) dev_info(hdev->dev, "enter spraid remove\n"); spraid_change_host_state(hdev, SPRAID_DELETING); + flush_work(&hdev->reset_work); - if (!pci_device_is_present(pdev)) { - scsi_block_requests(shost); + if (!pci_device_is_present(pdev)) spraid_back_all_io(hdev); - scsi_unblock_requests(shost); - } - flush_work(&hdev->reset_work); + spraid_remove_bsg(hdev); scsi_remove_host(shost); - - kfree(hdev->ioq_ptcmds); + spraid_free_ioq_ptcmds(hdev); kfree(hdev->devices); - spraid_remove_cdev(hdev); spraid_remove_io_queues(hdev); spraid_free_iod_ext_mem_pool(hdev); - spraid_remove_admin_tagset(hdev); spraid_disable_admin_queue(hdev, false); spraid_pci_disable(hdev); spraid_free_resources(hdev); @@ -3551,7 +3622,7 @@ static void spraid_remove(struct pci_dev *pdev) } static const struct pci_device_id spraid_id_table[] = { - { PCI_DEVICE(PCI_VENDOR_ID_RAMAXEL_LOGIC, SPRAID_SERVER_DEVICE_HAB_DID) }, + { PCI_DEVICE(PCI_VENDOR_ID_RAMAXEL_LOGIC, SPRAID_SERVER_DEVICE_HBA_DID) }, { PCI_DEVICE(PCI_VENDOR_ID_RAMAXEL_LOGIC, SPRAID_SERVER_DEVICE_RAID_DID) }, { 0, } }; @@ -3563,6 +3634,7 @@ static struct pci_driver spraid_driver = { .probe = spraid_probe, .remove = spraid_remove, .shutdown = spraid_shutdown, + .err_handler = &spraid_err_handler, }; static int __init spraid_init(void) @@ -3573,14 +3645,10 @@ static int __init spraid_init(void) if (!spraid_wq) return -ENOMEM; - ret = alloc_chrdev_region(&spraid_chr_devt, 0, SPRAID_MINORS, "spraid"); - if (ret < 0) - goto destroy_wq; - spraid_class = class_create(THIS_MODULE, "spraid"); if (IS_ERR(spraid_class)) { ret = PTR_ERR(spraid_class); - goto unregister_chrdev; + goto destroy_wq; } ret = pci_register_driver(&spraid_driver); @@ -3591,8 +3659,6 @@ static int __init spraid_init(void) destroy_class: class_destroy(spraid_class); -unregister_chrdev: - unregister_chrdev_region(spraid_chr_devt, SPRAID_MINORS); destroy_wq: destroy_workqueue(spraid_wq); @@ -3603,12 +3669,11 @@ static void __exit spraid_exit(void) { pci_unregister_driver(&spraid_driver); class_destroy(spraid_class); - unregister_chrdev_region(spraid_chr_devt, SPRAID_MINORS); destroy_workqueue(spraid_wq); ida_destroy(&spraid_instance_ida); } -MODULE_AUTHOR("Ramaxel Memory Technology"); +MODULE_AUTHOR("songyl(a)ramaxel.com"); MODULE_DESCRIPTION("Ramaxel Memory Technology SPraid Driver"); MODULE_LICENSE("GPL"); MODULE_VERSION(SPRAID_DRV_VERSION); -- 2.27.0

1 0

[PATCH openEuler-1.0-LTS v2 01/10] arm64: mm: Restore mm_cpumask (revert commit 38d96287504a ("arm64: mm: kill mm_cpumask usage"))
by Yang Yingliang 20 Dec '21

20 Dec '21

From: Takao Indoh <indou.takao(a)fujitsu.com> hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I4BLL0 CVE: NA --------------------------- mm_cpumask was deleted by the commit 38d96287504a ("arm64: mm: kill mm_cpumask usage") because it was not used at that time. Now this is needed to find appropriate CPUs for TLB flush, so this patch reverts this commit. Signed-off-by: QI Fuli <qi.fuli(a)fujitsu.com> Signed-off-by: Takao Indoh <indou.takao(a)fujitsu.com> Signed-off-by: Cheng Jian <cj.chengjian(a)huawei.com> Reviewed-by: Xie XiuQi <xiexiuqi(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- arch/arm64/kernel/smp.c | 6 ++++++ arch/arm64/mm/context.c | 2 ++ 2 files changed, 8 insertions(+) diff --git a/arch/arm64/kernel/smp.c b/arch/arm64/kernel/smp.c index fe562778de352..e86940d353a3e 100644 --- a/arch/arm64/kernel/smp.c +++ b/arch/arm64/kernel/smp.c @@ -387,6 +387,7 @@ asmlinkage notrace void secondary_start_kernel(void) */ mmgrab(mm); current->active_mm = mm; + cpumask_set_cpu(cpu, mm_cpumask(mm)); /* * TTBR0 is only used for the identity mapping at this stage. Make it @@ -489,6 +490,11 @@ int __cpu_disable(void) */ irq_migrate_all_off_this_cpu(); + /* + * Remove this CPU from the vm mask set of all processes. + */ + clear_tasks_mm_cpumask(cpu); + return 0; } diff --git a/arch/arm64/mm/context.c b/arch/arm64/mm/context.c index 2b80ceff5d6c2..27d1f3fec1cc9 100644 --- a/arch/arm64/mm/context.c +++ b/arch/arm64/mm/context.c @@ -207,6 +207,7 @@ static u64 new_context(struct mm_struct *mm, unsigned int cpu) set_asid: __set_bit(asid, asid_map); cur_idx = asid; + cpumask_clear(mm_cpumask(mm)); return idx2asid(asid) | generation; } @@ -254,6 +255,7 @@ void check_and_switch_context(struct mm_struct *mm, unsigned int cpu) switch_mm_fastpath: arm64_apply_bp_hardening(); + cpumask_set_cpu(cpu, mm_cpumask(mm)); /* * Defer TTBR0_EL1 setting for user threads to uaccess_enable() when -- 2.25.1

1 9

[PATCH openEuler-1.0-LTS 1/2] audit: improve robustness of the audit queue handling
by Yang Yingliang 20 Dec '21

20 Dec '21

From: Paul Moore <paul(a)paul-moore.com> hulk inclusion category: bugfix bugzilla: 185904, https://gitee.com/openeuler/kernel/issues/I4NAZN CVE: NA Reference: https://patchwork.kernel.org/project/linux-audit/patch/163942029335.62691.7… ------------------------------------------------------------------- If the audit daemon were ever to get stuck in a stopped state the kernel's kauditd_thread() could get blocked attempting to send audit records to the userspace audit daemon. With the kernel thread blocked it is possible that the audit queue could grow unbounded as certain audit record generating events must be exempt from the queue limits else the system enter a deadlock state. This patch resolves this problem by lowering the kernel thread's socket sending timeout from MAX_SCHEDULE_TIMEOUT to HZ/10 and tweaks the kauditd_send_queue() function to better manage the various audit queues when connection problems occur between the kernel and the audit daemon. With this patch, the backlog may temporarily grow beyond the defined limits when the audit daemon is stopped and the system is under heavy audit pressure, but kauditd_thread() will continue to make progress and drain the queues as it would for other connection problems. For example, with the audit daemon put into a stopped state and the system configured to audit every syscall it was still possible to shutdown the system without a kernel panic, deadlock, etc.; granted, the system was slow to shutdown but that is to be expected given the extreme pressure of recording every syscall. The timeout value of HZ/10 was chosen primarily through experimentation and this developer's "gut feeling". There is likely no one perfect value, but as this scenario is limited in scope (root privileges would be needed to send SIGSTOP to the audit daemon), it is likely not worth exposing this as a tunable at present. This can always be done at a later date if it proves necessary. Cc: stable(a)vger.kernel.org Fixes: 5b52330bbfe63 ("audit: fix auditd/kernel connection state tracking") Reported-by: Gaosheng Cui <cuigaosheng1(a)huawei.com> Tested-by: Gaosheng Cui <cuigaosheng1(a)huawei.com> Reviewed-by: Richard Guy Briggs <rgb(a)redhat.com> Signed-off-by: Paul Moore <paul(a)paul-moore.com> Signed-off-by: Cui GaoSheng <cuigaosheng1(a)huawei.com> Reviewed-by: Xiu Jianfeng <xiujianfeng(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- kernel/audit.c | 21 ++++++++++----------- 1 file changed, 10 insertions(+), 11 deletions(-) diff --git a/kernel/audit.c b/kernel/audit.c index 45741c3c48a47..968921d376b98 100644 --- a/kernel/audit.c +++ b/kernel/audit.c @@ -726,7 +726,7 @@ static int kauditd_send_queue(struct sock *sk, u32 portid, { int rc = 0; struct sk_buff *skb; - static unsigned int failed = 0; + unsigned int failed = 0; /* NOTE: kauditd_thread takes care of all our locking, we just use * the netlink info passed to us (e.g. sk and portid) */ @@ -743,32 +743,30 @@ static int kauditd_send_queue(struct sock *sk, u32 portid, continue; } +retry: /* grab an extra skb reference in case of error */ skb_get(skb); rc = netlink_unicast(sk, skb, portid, 0); if (rc < 0) { - /* fatal failure for our queue flush attempt? */ + /* send failed - try a few times unless fatal error */ if (++failed >= retry_limit || rc == -ECONNREFUSED || rc == -EPERM) { - /* yes - error processing for the queue */ sk = NULL; if (err_hook) (*err_hook)(skb); - if (!skb_hook) - goto out; - /* keep processing with the skb_hook */ + if (rc == -EAGAIN) + rc = 0; + /* continue to drain the queue */ continue; } else - /* no - requeue to preserve ordering */ - skb_queue_head(queue, skb); + goto retry; } else { - /* it worked - drop the extra reference and continue */ + /* skb sent - drop the extra reference and continue */ consume_skb(skb); failed = 0; } } -out: return (rc >= 0 ? 0 : rc); } @@ -1557,7 +1555,8 @@ static int __net_init audit_net_init(struct net *net) audit_panic("cannot initialize netlink socket in namespace"); return -ENOMEM; } - aunet->sk->sk_sndtimeo = MAX_SCHEDULE_TIMEOUT; + /* limit the timeout in case auditd is blocked/stopped */ + aunet->sk->sk_sndtimeo = HZ / 10; return 0; } -- 2.25.1

1 1

[PATCH openEuler-1.0-LTS v11 00/35] support psi feature
by Liu Xinpeng 19 Dec '21

19 Dec '21

backport psi feature and avoid kabi change bugzilla: https://gitee.com/openeuler/kernel/issues/I47QS2 Changes since v9: 1.Third-party kernel modules should not use struct cgroup, to avoid kabi changes by __GENKSYMS__, from Xie Xiuqi. Baruch Siach (1): psi: fix reference to kernel commandline enable Dan Schatzberg (1): kernel/sched/psi.c: expose pressure metrics on root cgroup Jason Xing (1): psi: get poll_work to run when calling poll syscall next time Johannes Weiner (14): mm: workingset: tell cache transitions from workingset thrashing sched: loadavg: consolidate LOAD_INT, LOAD_FRAC, CALC_LOAD sched: loadavg: make calc_load_n() public sched: sched.h: make rq locking and clock functions available in stats.h sched: introduce this_rq_lock_irq() psi: pressure stall information for CPU, memory, and IO psi: cgroup support psi: make disabling/enabling easier for vendor kernels psi: fix aggregation idle shut-off psi: avoid divide-by-zero crash inside virtual machines fs: kernfs: add poll file operation kernel: cgroup: add poll file operation sched/psi: Fix sampling error and rare div0 crashes with cgroups and high uptime psi: Fix a division error in psi poll() Josef Bacik (1): blk-iolatency: use a percentile approache for ssd's Liu Xinpeng (2): psi:enable psi in config psi:avoid kabi change Miklos Szeredi (1): fuse: ignore PG_workingset after stealing Miles Chen (1): sched/psi: Correct overly pessimistic size calculation Odin Ugedal (1): cgroup: fix psi monitor for root cgroup Olof Johansson (1): kernel/sched/psi.c: simplify cgroup_move_task() Peter Zijlstra (1): sched/psi: Reduce psimon FIFO priority Suren Baghdasaryan (9): psi: introduce state_mask to represent stalled psi states psi: make psi_enable static psi: rename psi fields in preparation for psi trigger addition psi: split update_stats into parts psi: track changed states include/: refactor headers to allow kthread.h inclusion in psi_types.h psi: introduce psi monitor sched/psi: Do not require setsched permission from the sched/psi: Fix OOB write when writing 0 bytes to PSI files Yafang Shao (1): mm, memcg: add workingset_restore in memory.stat Documentation/accounting/psi.txt | 180 ++++ Documentation/admin-guide/cgroup-v2.rst | 22 + Documentation/admin-guide/kernel-parameters.txt | 4 + arch/arm64/configs/openeuler_defconfig | 2 + arch/powerpc/platforms/cell/cpufreq_spudemand.c | 2 +- arch/powerpc/platforms/cell/spufs/sched.c | 9 +- arch/s390/appldata/appldata_os.c | 4 - arch/x86/configs/openeuler_defconfig | 2 + block/blk-iolatency.c | 183 +++- drivers/cpuidle/governors/menu.c | 4 - drivers/spi/spi-rockchip.c | 1 + fs/fuse/dev.c | 1 + fs/kernfs/file.c | 31 +- fs/proc/loadavg.c | 3 - include/linux/cgroup-defs.h | 11 + include/linux/cgroup.h | 15 + include/linux/kernfs.h | 8 + include/linux/kthread.h | 4 + include/linux/page-flags.h | 5 + include/linux/psi.h | 65 ++ include/linux/psi_types.h | 173 +++ include/linux/sched.h | 13 + include/linux/sched/loadavg.h | 24 +- include/linux/swap.h | 1 + include/trace/events/mmflags.h | 1 + init/Kconfig | 28 + kernel/cgroup/cgroup.c | 131 ++- kernel/debug/kdb/kdb_main.c | 7 +- kernel/fork.c | 4 + kernel/kthread.c | 1 + kernel/sched/Makefile | 1 + kernel/sched/core.c | 16 +- kernel/sched/loadavg.c | 139 ++- kernel/sched/psi.c | 1292 +++++++++++++++++++++++ kernel/sched/sched.h | 178 ++-- kernel/sched/stats.h | 86 ++ kernel/workqueue.c | 23 + kernel/workqueue_internal.h | 6 +- mm/compaction.c | 5 + mm/filemap.c | 20 +- mm/huge_memory.c | 1 + mm/migrate.c | 2 + mm/page_alloc.c | 9 + mm/swap_state.c | 1 + mm/vmscan.c | 10 + mm/workingset.c | 113 +- 46 files changed, 2562 insertions(+), 279 deletions(-) create mode 100644 Documentation/accounting/psi.txt create mode 100644 include/linux/psi.h create mode 100644 include/linux/psi_types.h create mode 100644 kernel/sched/psi.c -- 1.8.3.1

1 35

[PATCH openEuler-1.0-LTS] block/wbt: fix negative inflight counter when remove scsi device
by Yang Yingliang 18 Dec '21

18 Dec '21

From: Laibin Qiu <qiulaibin(a)huawei.com> hulk inclusion category: bugfix bugzilla: 185857 CVE: NA -------------------------------- Now that we disable wbt by set WBT_STATE_OFF_DEFAULT in wbt_disable_default() when switch elevator to bfq. And when we remove scsi device, wbt will be enabled by wbt_enable_default. If it become false positive between wbt_wait() and wbt_track() when submit write request. The following is the scenario that triggered the problem. T1 T2 T3 elevator_switch_mq bfq_init_queue wbt_disable_default <= Set rwb->enable_state (OFF) Submit_bio blk_mq_make_request rq_qos_throttle <= rwb->enable_state (OFF) scsi_remove_device sd_remove del_gendisk blk_unregister_queue elv_unregister_queue wbt_enable_default <= Set rwb->enable_state (ON) q_qos_track <= rwb->enable_state (ON) ^^^^^^ this request will mark WBT_TRACKED without inflight add and will lead to drop rqw->inflight to -1 in wbt_done() which will trigger IO hung. Fix this by move wbt_enable_default() from elv_unregister to bfq_exit_queue(). Only re-enable wbt when bfq exit. Fixes: 76a8040817b4b ("blk-wbt: make sure throttle is enabled properly") confilct: block/elevator.c Signed-off-by: Laibin Qiu <qiulaibin(a)huawei.com> Reviewed-by: Jason Yan <yanaijie(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- block/bfq-iosched.c | 4 ++++ block/elevator.c | 2 -- 2 files changed, 4 insertions(+), 2 deletions(-) diff --git a/block/bfq-iosched.c b/block/bfq-iosched.c index 9078299ffe3da..fe4255698d810 100644 --- a/block/bfq-iosched.c +++ b/block/bfq-iosched.c @@ -5381,6 +5381,7 @@ static void bfq_exit_queue(struct elevator_queue *e) { struct bfq_data *bfqd = e->elevator_data; struct bfq_queue *bfqq, *n; + struct request_queue *q = bfqd->queue; hrtimer_cancel(&bfqd->idle_slice_timer); @@ -5404,6 +5405,9 @@ static void bfq_exit_queue(struct elevator_queue *e) #endif kfree(bfqd); + + /* Re-enable throttling in case elevator disabled it */ + wbt_enable_default(q); } static void bfq_init_root_group(struct bfq_group *root_group, diff --git a/block/elevator.c b/block/elevator.c index 70ea9f3e64d7e..9621a812319e7 100644 --- a/block/elevator.c +++ b/block/elevator.c @@ -876,8 +876,6 @@ void elv_unregister_queue(struct request_queue *q) kobject_uevent(&e->kobj, KOBJ_REMOVE); kobject_del(&e->kobj); e->registered = 0; - /* Re-enable throttling in case elevator disabled it */ - wbt_enable_default(q); } } -- 2.25.1

1 0

[PATCH openEuler-1.0-LTS] nbd: Fix use-after-free in blk_mq_free_rqs
by Yang Yingliang 18 Dec '21

18 Dec '21

From: Ye Bin <yebin10(a)huawei.com> hulk inclusion category: bugfix bugzilla: 185875 CVE: NA ----------------------------------------------- We got issue as follows: [ 833.786542] nbd: failed to add new device [ 833.791613] ================================================================== [ 833.794918] BUG: KASAN: use-after-free in blk_mq_free_rqs+0x558/0x6c0 [ 833.798108] Read of size 8 at addr ffff800109b7c288 by task kworker/0:3/113 [ 833.804216] CPU: 0 PID: 113 Comm: kworker/0:3 Kdump: loaded Not tainted 4.19.90 #1 [ 833.807635] Hardware name: QEMU KVM Virtual Machine, BIOS 0.0.0 02/06/2015 [ 833.811798] Workqueue: events __blk_release_queue [ 833.815035] Call trace: [ 833.817964] dump_backtrace+0x0/0x3c0 [ 833.821070] show_stack+0x28/0x38 [ 833.824091] dump_stack+0xfc/0x154 [ 833.827042] print_address_description+0x68/0x278 [ 833.830147] kasan_report+0x204/0x330 [ 833.833121] __asan_report_load8_noabort+0x30/0x40 [ 833.836180] blk_mq_free_rqs+0x558/0x6c0 [ 833.839089] blk_mq_sched_tags_teardown+0xf4/0x1c0 [ 833.842035] blk_mq_exit_sched+0x1b8/0x260 [ 833.844878] elevator_exit+0x114/0x148 [ 833.847634] blk_exit_queue+0x68/0xe8 [ 833.850352] __blk_release_queue+0xd0/0x408 [ 833.853113] process_one_work+0x55c/0x10d0 [ 833.855864] worker_thread+0x3d4/0xe30 [ 833.858558] kthread+0x2c8/0x348 [ 833.861139] ret_from_fork+0x10/0x18 [ 833.863714] [ 833.866000] Allocated by task 186531: [ 833.868467] kasan_kmalloc+0xe0/0x190 [ 833.870936] kmem_cache_alloc_trace+0x104/0x218 [ 833.873483] nbd_dev_add+0x54/0x760 [nbd] [ 833.875988] nbd_genl_connect+0x3c4/0x1348 [nbd] [ 833.878591] genl_family_rcv_msg+0x798/0xa10 [ 833.881113] genl_rcv_msg+0xc0/0x170 [ 833.883489] netlink_rcv_skb+0x1b4/0x370 [ 833.885897] genl_rcv+0x40/0x58 [ 833.888225] netlink_unicast+0x4bc/0x660 [ 833.890661] netlink_sendmsg+0x880/0xa60 [ 833.893112] sock_sendmsg+0xb8/0x110 [ 833.895513] ____sys_sendmsg+0x570/0x698 [ 833.897927] ___sys_sendmsg+0x108/0x188 [ 833.900350] __sys_sendmsg+0xe8/0x198 [ 833.900360] __arm64_sys_sendmsg+0x78/0xa8 [ 833.906911] el0_svc_common+0x10c/0x330 [ 833.909289] el0_svc_handler+0x60/0xd0 [ 833.911660] el0_svc+0x8/0x1b0 [ 833.913963] [ 833.916117] Freed by task 186531: [ 833.918445] __kasan_slab_free+0x120/0x228 [ 833.920860] kasan_slab_free+0x10/0x18 [ 833.923193] kfree+0x80/0x1f0 [ 833.925392] nbd_dev_add+0xf0/0x760 [nbd] [ 833.927686] nbd_genl_connect+0x3c4/0x1348 [nbd] [ 833.929989] genl_family_rcv_msg+0x798/0xa10 [ 833.932231] genl_rcv_msg+0xc0/0x170 [ 833.934335] netlink_rcv_skb+0x1b4/0x370 [ 833.936444] genl_rcv+0x40/0x58 [ 833.938460] netlink_unicast+0x4bc/0x660 [ 833.940570] netlink_sendmsg+0x880/0xa60 [ 833.942682] sock_sendmsg+0xb8/0x110 [ 833.944745] ____sys_sendmsg+0x570/0x698 [ 833.946849] ___sys_sendmsg+0x108/0x188 [ 833.948924] __sys_sendmsg+0xe8/0x198 [ 833.950980] __arm64_sys_sendmsg+0x78/0xa8 [ 833.953059] el0_svc_common+0x10c/0x330 [ 833.955121] el0_svc_handler+0x60/0xd0 [ 833.957143] el0_svc+0x8/0x1b0 [ 833.959088] [ 833.960846] The buggy address belongs to the object at ffff800109b7c280 [ 833.960846] which belongs to the cache kmalloc-512 of size 512 [ 833.965502] The buggy address is located 8 bytes inside of [ 833.965502] 512-byte region [ffff800109b7c280, ffff800109b7c480) [ 833.970108] The buggy address belongs to the page: [ 833.972390] page:ffff7e000426df00 count:1 mapcount:0 mapping:ffff8000c0003800 index:0x0 compound_mapcount: 0 [ 833.975269] flags: 0x7ffff0000008100(slab|head) [ 833.977640] raw: 07ffff0000008100 ffff7e00035d1600 0000000200000002 ffff8000c0003800 [ 833.980426] raw: 0000000000000000 00000000000c000c 00000001ffffffff 0000000000000000 [ 833.983212] page dumped because: kasan: bad access detected [ 833.985778] [ 833.987935] Memory state around the buggy address: [ 833.990448] ffff800109b7c180: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb [ 833.993265] ffff800109b7c200: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc [ 833.996097] >ffff800109b7c280: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb [ 833.998930] ^ [ 834.001384] ffff800109b7c300: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb [ 834.004329] ffff800109b7c380: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb [ 834.007264] ================================================================== As 98fd847a495f commit add check "disk->first_minor", if failed will free tags and finally call put_disk will free request_queue. blk_mq_free_tag_set blk_mq_free_map_and_requests blk_mq_free_rqs put_disk: __blk_release_queue blk_exit_queue elevator_exit blk_mq_exit_sched blk_mq_sched_tags_teardown blk_mq_free_rqs -->will trigger UAF To address this issue, just move 'disk->first_minor' check at the first in nbd_dev_add. Fixes:98fd847a495f("nbd: add sanity check for first_minor") Signed-off-by: Ye Bin <yebin10(a)huawei.com> Reviewed-by: Yu Kuai <yukuai3(a)huawei.com> Reviewed-by: Jason Yan <yanaijie(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/block/nbd.c | 22 ++++++++++------------ 1 file changed, 10 insertions(+), 12 deletions(-) diff --git a/drivers/block/nbd.c b/drivers/block/nbd.c index 775cbb4c1bbcd..a35189098581b 100644 --- a/drivers/block/nbd.c +++ b/drivers/block/nbd.c @@ -1702,6 +1702,15 @@ static int nbd_dev_add(int index) struct gendisk *disk; struct request_queue *q; int err = -ENOMEM; + int first_minor = index << part_shift; + + /* + * Too big index can cause duplicate creation of sysfs files/links, + * because MKDEV() expect that the max first minor is MINORMASK, or + * index << part_shift can overflow. + */ + if (first_minor < index || first_minor > MINORMASK) + return -EINVAL; nbd = kzalloc(sizeof(struct nbd_device), GFP_KERNEL); if (!nbd) @@ -1765,18 +1774,7 @@ static int nbd_dev_add(int index) refcount_set(&nbd->refs, 1); INIT_LIST_HEAD(&nbd->list); disk->major = NBD_MAJOR; - - /* - * Too big index can cause duplicate creation of sysfs files/links, - * because MKDEV() expect that the max first minor is MINORMASK, or - * index << part_shift can overflow. - */ - disk->first_minor = index << part_shift; - if (disk->first_minor < index || disk->first_minor > MINORMASK) { - err = -EINVAL; - goto out_free_tags; - } - + disk->first_minor = first_minor; disk->fops = &nbd_fops; disk->private_data = nbd; sprintf(disk->disk_name, "nbd%d", index); -- 2.25.1

1 0

[PATCH openEuler-1.0-LTS 1/5] block, bfq: improve asymmetric scenarios detection
by Yang Yingliang 17 Dec '21

17 Dec '21

From: Federico Motta <federico(a)willer.it> mainline inclusion from mainline-v4.20-rc1 commit 2d29c9f89fcd9bf408fcdaaf515c90a169f22ecd category: bugfix bugzilla: 185905 CVE: NA --------------------------- bfq defines as asymmetric a scenario where an active entity, say E (representing either a single bfq_queue or a group of other entities), has a higher weight than some other entities. If the entity E does sync I/O in such a scenario, then bfq plugs the dispatch of the I/O of the other entities in the following situation: E is in service but temporarily has no pending I/O request. In fact, without this plugging, all the times that E stops being temporarily idle, it may find the internal queues of the storage device already filled with an out-of-control number of extra requests, from other entities. So E may have to wait for the service of these extra requests, before finally having its own requests served. This may easily break service guarantees, with E getting less than its fair share of the device throughput. Usually, the end result is that E gets the same fraction of the throughput as the other entities, instead of getting more, according to its higher weight. Yet there are two other more subtle cases where E, even if its weight is actually equal to or even lower than the weight of any other active entities, may get less than its fair share of the throughput in case the above I/O plugging is not performed: 1. other entities issue larger requests than E; 2. other entities contain more active child entities than E (or in general tend to have more backlog than E). In the first case, other entities may get more service than E because they get larger requests, than those of E, served during the temporary idle periods of E. In the second case, other entities get more service because, by having many child entities, they have many requests ready for dispatching while E is temporarily idle. This commit addresses this issue by extending the definition of asymmetric scenario: a scenario is asymmetric when - active entities representing bfq_queues have differentiated weights, as in the original definition or (inclusive) - one or more entities representing groups of entities are active. This broader definition makes sure that I/O plugging will be performed in all the above cases, provided that there is at least one active group. Of course, this definition is very coarse, so it will trigger I/O plugging also in cases where it is not needed, such as, e.g., multiple active entities with just one child each, and all with the same I/O-request size. The reason for this coarse definition is just that a finer-grained definition would be rather heavy to compute. On the opposite end, even this new definition does not trigger I/O plugging in all cases where there is no active group, and all bfq_queues have the same weight. So, in these cases some unfairness may occur if there are asymmetries in I/O-request sizes. We made this choice because I/O plugging may lower throughput, and probably a user that has not created any group cares more about throughput than about perfect fairness. At any rate, as for possible applications that may care about service guarantees, bfq already guarantees a high responsiveness and a low latency to soft real-time applications automatically. Signed-off-by: Federico Motta <federico(a)willer.it> Signed-off-by: Paolo Valente <paolo.valente(a)linaro.org> Signed-off-by: Jens Axboe <axboe(a)kernel.dk> Signed-off-by: Yu Kuai <yukuai3(a)huawei.com> Reviewed-by: Hou Tao <houtao1(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- block/bfq-iosched.c | 223 ++++++++++++++++++++++++-------------------- block/bfq-iosched.h | 27 +++--- block/bfq-wf2q.c | 35 +++---- 3 files changed, 154 insertions(+), 131 deletions(-) diff --git a/block/bfq-iosched.c b/block/bfq-iosched.c index 4686ca1e93afe..e2b5c3ccd5502 100644 --- a/block/bfq-iosched.c +++ b/block/bfq-iosched.c @@ -625,12 +625,13 @@ void bfq_pos_tree_add_move(struct bfq_data *bfqd, struct bfq_queue *bfqq) } /* - * Tell whether there are active queues or groups with differentiated weights. + * Tell whether there are active queues with different weights or + * active groups. */ -static bool bfq_differentiated_weights(struct bfq_data *bfqd) +static bool bfq_varied_queue_weights_or_active_groups(struct bfq_data *bfqd) { /* - * For weights to differ, at least one of the trees must contain + * For queue weights to differ, queue_weights_tree must contain * at least two nodes. */ return (!RB_EMPTY_ROOT(&bfqd->queue_weights_tree) && @@ -638,9 +639,7 @@ static bool bfq_differentiated_weights(struct bfq_data *bfqd) bfqd->queue_weights_tree.rb_node->rb_right) #ifdef CONFIG_BFQ_GROUP_IOSCHED ) || - (!RB_EMPTY_ROOT(&bfqd->group_weights_tree) && - (bfqd->group_weights_tree.rb_node->rb_left || - bfqd->group_weights_tree.rb_node->rb_right) + (bfqd->num_active_groups > 0 #endif ); } @@ -658,26 +657,25 @@ static bool bfq_differentiated_weights(struct bfq_data *bfqd) * 3) all active groups at the same level in the groups tree have the same * number of children. * - * Unfortunately, keeping the necessary state for evaluating exactly the - * above symmetry conditions would be quite complex and time-consuming. - * Therefore this function evaluates, instead, the following stronger - * sub-conditions, for which it is much easier to maintain the needed - * state: + * Unfortunately, keeping the necessary state for evaluating exactly + * the last two symmetry sub-conditions above would be quite complex + * and time consuming. Therefore this function evaluates, instead, + * only the following stronger two sub-conditions, for which it is + * much easier to maintain the needed state: * 1) all active queues have the same weight, - * 2) all active groups have the same weight, - * 3) all active groups have at most one active child each. - * In particular, the last two conditions are always true if hierarchical - * support and the cgroups interface are not enabled, thus no state needs - * to be maintained in this case. + * 2) there are no active groups. + * In particular, the last condition is always true if hierarchical + * support or the cgroups interface are not enabled, thus no state + * needs to be maintained in this case. */ static bool bfq_symmetric_scenario(struct bfq_data *bfqd) { - return !bfq_differentiated_weights(bfqd); + return !bfq_varied_queue_weights_or_active_groups(bfqd); } /* * If the weight-counter tree passed as input contains no counter for - * the weight of the input entity, then add that counter; otherwise just + * the weight of the input queue, then add that counter; otherwise just * increment the existing counter. * * Note that weight-counter trees contain few nodes in mostly symmetric @@ -688,25 +686,25 @@ static bool bfq_symmetric_scenario(struct bfq_data *bfqd) * In most scenarios, the rate at which nodes are created/destroyed * should be low too. */ -void bfq_weights_tree_add(struct bfq_data *bfqd, struct bfq_entity *entity, +void bfq_weights_tree_add(struct bfq_data *bfqd, struct bfq_queue *bfqq, struct rb_root *root) { + struct bfq_entity *entity = &bfqq->entity; struct rb_node **new = &(root->rb_node), *parent = NULL; /* - * Do not insert if the entity is already associated with a + * Do not insert if the queue is already associated with a * counter, which happens if: - * 1) the entity is associated with a queue, - * 2) a request arrival has caused the queue to become both + * 1) a request arrival has caused the queue to become both * non-weight-raised, and hence change its weight, and * backlogged; in this respect, each of the two events * causes an invocation of this function, - * 3) this is the invocation of this function caused by the + * 2) this is the invocation of this function caused by the * second event. This second invocation is actually useless, * and we handle this fact by exiting immediately. More * efficient or clearer solutions might possibly be adopted. */ - if (entity->weight_counter) + if (bfqq->weight_counter) return; while (*new) { @@ -716,7 +714,7 @@ void bfq_weights_tree_add(struct bfq_data *bfqd, struct bfq_entity *entity, parent = *new; if (entity->weight == __counter->weight) { - entity->weight_counter = __counter; + bfqq->weight_counter = __counter; goto inc_counter; } if (entity->weight < __counter->weight) @@ -725,66 +723,67 @@ void bfq_weights_tree_add(struct bfq_data *bfqd, struct bfq_entity *entity, new = &((*new)->rb_right); } - entity->weight_counter = kzalloc(sizeof(struct bfq_weight_counter), - GFP_ATOMIC); + bfqq->weight_counter = kzalloc(sizeof(struct bfq_weight_counter), + GFP_ATOMIC); /* * In the unlucky event of an allocation failure, we just - * exit. This will cause the weight of entity to not be - * considered in bfq_differentiated_weights, which, in its - * turn, causes the scenario to be deemed wrongly symmetric in - * case entity's weight would have been the only weight making - * the scenario asymmetric. On the bright side, no unbalance - * will however occur when entity becomes inactive again (the - * invocation of this function is triggered by an activation - * of entity). In fact, bfq_weights_tree_remove does nothing - * if !entity->weight_counter. + * exit. This will cause the weight of queue to not be + * considered in bfq_varied_queue_weights_or_active_groups, + * which, in its turn, causes the scenario to be deemed + * wrongly symmetric in case bfqq's weight would have been + * the only weight making the scenario asymmetric. On the + * bright side, no unbalance will however occur when bfqq + * becomes inactive again (the invocation of this function + * is triggered by an activation of queue). In fact, + * bfq_weights_tree_remove does nothing if + * !bfqq->weight_counter. */ - if (unlikely(!entity->weight_counter)) + if (unlikely(!bfqq->weight_counter)) return; - entity->weight_counter->weight = entity->weight; - rb_link_node(&entity->weight_counter->weights_node, parent, new); - rb_insert_color(&entity->weight_counter->weights_node, root); + bfqq->weight_counter->weight = entity->weight; + rb_link_node(&bfqq->weight_counter->weights_node, parent, new); + rb_insert_color(&bfqq->weight_counter->weights_node, root); inc_counter: - entity->weight_counter->num_active++; + bfqq->weight_counter->num_active++; } /* - * Decrement the weight counter associated with the entity, and, if the + * Decrement the weight counter associated with the queue, and, if the * counter reaches 0, remove the counter from the tree. * See the comments to the function bfq_weights_tree_add() for considerations * about overhead. */ void __bfq_weights_tree_remove(struct bfq_data *bfqd, - struct bfq_entity *entity, + struct bfq_queue *bfqq, struct rb_root *root) { - if (!entity->weight_counter) + if (!bfqq->weight_counter) return; - entity->weight_counter->num_active--; - if (entity->weight_counter->num_active > 0) + bfqq->weight_counter->num_active--; + if (bfqq->weight_counter->num_active > 0) goto reset_entity_pointer; - rb_erase(&entity->weight_counter->weights_node, root); - kfree(entity->weight_counter); + rb_erase(&bfqq->weight_counter->weights_node, root); + kfree(bfqq->weight_counter); reset_entity_pointer: - entity->weight_counter = NULL; + bfqq->weight_counter = NULL; } /* - * Invoke __bfq_weights_tree_remove on bfqq and all its inactive - * parent entities. + * Invoke __bfq_weights_tree_remove on bfqq and decrement the number + * of active groups for each queue's inactive parent entity. */ void bfq_weights_tree_remove(struct bfq_data *bfqd, struct bfq_queue *bfqq) { struct bfq_entity *entity = bfqq->entity.parent; - __bfq_weights_tree_remove(bfqd, &bfqq->entity, + __bfq_weights_tree_remove(bfqd, bfqq, &bfqd->queue_weights_tree); for_each_entity(entity) { @@ -798,17 +797,13 @@ void bfq_weights_tree_remove(struct bfq_data *bfqd, * next_in_service for details on why * in_service_entity must be checked too). * - * As a consequence, the weight of entity is - * not to be removed. In addition, if entity - * is active, then its parent entities are - * active as well, and thus their weights are - * not to be removed either. In the end, this - * loop must stop here. + * As a consequence, its parent entities are + * active as well, and thus this loop must + * stop here. */ break; } - __bfq_weights_tree_remove(bfqd, entity, - &bfqd->group_weights_tree); + bfqd->num_active_groups--; } } @@ -3539,9 +3534,11 @@ static bool bfq_better_to_idle(struct bfq_queue *bfqq) * symmetric scenario where: * (i) each of these processes must get the same throughput as * the others; - * (ii) all these processes have the same I/O pattern - (either sequential or random). - * In fact, in such a scenario, the drive will tend to treat + * (ii) the I/O of each process has the same properties, in + * terms of locality (sequential or random), direction + * (reads or writes), request sizes, greediness + * (from I/O-bound to sporadic), and so on. + * In fact, in such a scenario, the drive tends to treat * the requests of each of these processes in about the same * way as the requests of the others, and thus to provide * each of these processes with about the same throughput @@ -3550,18 +3547,50 @@ static bool bfq_better_to_idle(struct bfq_queue *bfqq) * certainly needed to guarantee that bfqq receives its * assigned fraction of the device throughput (see [1] for * details). + * The problem is that idling may significantly reduce + * throughput with certain combinations of types of I/O and + * devices. An important example is sync random I/O, on flash + * storage with command queueing. So, unless bfqq falls in the + * above cases where idling also boosts throughput, it would + * be important to check conditions (i) and (ii) accurately, + * so as to avoid idling when not strictly needed for service + * guarantees. + * + * Unfortunately, it is extremely difficult to thoroughly + * check condition (ii). And, in case there are active groups, + * it becomes very difficult to check condition (i) too. In + * fact, if there are active groups, then, for condition (i) + * to become false, it is enough that an active group contains + * more active processes or sub-groups than some other active + * group. We address this issue with the following bi-modal + * behavior, implemented in the function + * bfq_symmetric_scenario(). * - * We address this issue by controlling, actually, only the - * symmetry sub-condition (i), i.e., provided that - * sub-condition (i) holds, idling is not performed, - * regardless of whether sub-condition (ii) holds. In other - * words, only if sub-condition (i) holds, then idling is + * If there are active groups, then the scenario is tagged as + * asymmetric, conservatively, without checking any of the + * conditions (i) and (ii). So the device is idled for bfqq. + * This behavior matches also the fact that groups are created + * exactly if controlling I/O (to preserve bandwidth and + * latency guarantees) is a primary concern. + * + * On the opposite end, if there are no active groups, then + * only condition (i) is actually controlled, i.e., provided + * that condition (i) holds, idling is not performed, + * regardless of whether condition (ii) holds. In other words, + * only if condition (i) does not hold, then idling is * allowed, and the device tends to be prevented from queueing - * many requests, possibly of several processes. The reason - * for not controlling also sub-condition (ii) is that we - * exploit preemption to preserve guarantees in case of - * symmetric scenarios, even if (ii) does not hold, as - * explained in the next two paragraphs. + * many requests, possibly of several processes. Since there + * are no active groups, then, to control condition (i) it is + * enough to check whether all active queues have the same + * weight. + * + * Not checking condition (ii) evidently exposes bfqq to the + * risk of getting less throughput than its fair share. + * However, for queues with the same weight, a further + * mechanism, preemption, mitigates or even eliminates this + * problem. And it does so without consequences on overall + * throughput. This mechanism and its benefits are explained + * in the next three paragraphs. * * Even if a queue, say Q, is expired when it remains idle, Q * can still preempt the new in-service queue if the next @@ -3575,11 +3604,7 @@ static bool bfq_better_to_idle(struct bfq_queue *bfqq) * idling allows the internal queues of the device to contain * many requests, and thus to reorder requests, we can rather * safely assume that the internal scheduler still preserves a - * minimum of mid-term fairness. The motivation for using - * preemption instead of idling is that, by not idling, - * service guarantees are preserved without minimally - * sacrificing throughput. In other words, both a high - * throughput and its desired distribution are obtained. + * minimum of mid-term fairness. * * More precisely, this preemption-based, idleless approach * provides fairness in terms of IOPS, and not sectors per @@ -3598,27 +3623,27 @@ static bool bfq_better_to_idle(struct bfq_queue *bfqq) * 1024/8 times as high as the service received by the other * queue. * - * On the other hand, device idling is performed, and thus - * pure sector-domain guarantees are provided, for the - * following queues, which are likely to need stronger - * throughput guarantees: weight-raised queues, and queues - * with a higher weight than other queues. When such queues - * are active, sub-condition (i) is false, which triggers - * device idling. + * The motivation for using preemption instead of idling (for + * queues with the same weight) is that, by not idling, + * service guarantees are preserved (completely or at least in + * part) without minimally sacrificing throughput. And, if + * there is no active group, then the primary expectation for + * this device is probably a high throughput. * - * According to the above considerations, the next variable is - * true (only) if sub-condition (i) holds. To compute the - * value of this variable, we not only use the return value of - * the function bfq_symmetric_scenario(), but also check - * whether bfqq is being weight-raised, because - * bfq_symmetric_scenario() does not take into account also - * weight-raised queues (see comments on - * bfq_weights_tree_add()). In particular, if bfqq is being - * weight-raised, it is important to idle only if there are - * other, non-weight-raised queues that may steal throughput - * to bfqq. Actually, we should be even more precise, and - * differentiate between interactive weight raising and - * soft real-time weight raising. + * We are now left only with explaining the additional + * compound condition that is checked below for deciding + * whether the scenario is asymmetric. To explain this + * compound condition, we need to add that the function + * bfq_symmetric_scenario checks the weights of only + * non-weight-raised queues, for efficiency reasons (see + * comments on bfq_weights_tree_add()). Then the fact that + * bfqq is weight-raised is checked explicitly here. More + * precisely, the compound condition below takes into account + * also the fact that, even if bfqq is being weight-raised, + * the scenario is still symmetric if all active queues happen + * to be weight-raised. Actually, we should be even more + * precise here, and differentiate between interactive weight + * raising and soft real-time weight raising. * * As a side note, it is worth considering that the above * device-idling countermeasures may however fail in the @@ -5408,7 +5433,7 @@ static int bfq_init_queue(struct request_queue *q, struct elevator_type *e) bfqd->idle_slice_timer.function = bfq_idle_slice_timer; bfqd->queue_weights_tree = RB_ROOT; - bfqd->group_weights_tree = RB_ROOT; + bfqd->num_active_groups = 0; INIT_LIST_HEAD(&bfqd->active_list); INIT_LIST_HEAD(&bfqd->idle_list); diff --git a/block/bfq-iosched.h b/block/bfq-iosched.h index a41e9884f2dd2..e9960727cb565 100644 --- a/block/bfq-iosched.h +++ b/block/bfq-iosched.h @@ -108,15 +108,14 @@ struct bfq_sched_data { }; /** - * struct bfq_weight_counter - counter of the number of all active entities + * struct bfq_weight_counter - counter of the number of all active queues * with a given weight. */ struct bfq_weight_counter { - unsigned int weight; /* weight of the entities this counter refers to */ - unsigned int num_active; /* nr of active entities with this weight */ + unsigned int weight; /* weight of the queues this counter refers to */ + unsigned int num_active; /* nr of active queues with this weight */ /* - * Weights tree member (see bfq_data's @queue_weights_tree and - * @group_weights_tree) + * Weights tree member (see bfq_data's @queue_weights_tree) */ struct rb_node weights_node; }; @@ -151,8 +150,6 @@ struct bfq_weight_counter { struct bfq_entity { /* service_tree member */ struct rb_node rb_node; - /* pointer to the weight counter associated with this entity */ - struct bfq_weight_counter *weight_counter; /* * Flag, true if the entity is on a tree (either the active or @@ -266,6 +263,9 @@ struct bfq_queue { /* entity representing this queue in the scheduler */ struct bfq_entity entity; + /* pointer to the weight counter associated with this entity */ + struct bfq_weight_counter *weight_counter; + /* maximum budget allowed from the feedback mechanism */ int max_budget; /* budget expiration (in jiffies) */ @@ -449,14 +449,9 @@ struct bfq_data { */ struct rb_root queue_weights_tree; /* - * rbtree of non-queue @bfq_entity weight counters, sorted by - * weight. Used to keep track of whether all @bfq_groups have - * the same weight. The tree contains one counter for each - * distinct weight associated to some active @bfq_group (see - * the comments to the functions bfq_weights_tree_[add|remove] - * for further details). + * number of groups with requests still waiting for completion */ - struct rb_root group_weights_tree; + unsigned int num_active_groups; /* * Number of bfq_queues containing requests (including the @@ -854,10 +849,10 @@ struct bfq_queue *bic_to_bfqq(struct bfq_io_cq *bic, bool is_sync); void bic_set_bfqq(struct bfq_io_cq *bic, struct bfq_queue *bfqq, bool is_sync); struct bfq_data *bic_to_bfqd(struct bfq_io_cq *bic); void bfq_pos_tree_add_move(struct bfq_data *bfqd, struct bfq_queue *bfqq); -void bfq_weights_tree_add(struct bfq_data *bfqd, struct bfq_entity *entity, +void bfq_weights_tree_add(struct bfq_data *bfqd, struct bfq_queue *bfqq, struct rb_root *root); void __bfq_weights_tree_remove(struct bfq_data *bfqd, - struct bfq_entity *entity, + struct bfq_queue *bfqq, struct rb_root *root); void bfq_weights_tree_remove(struct bfq_data *bfqd, struct bfq_queue *bfqq); diff --git a/block/bfq-wf2q.c b/block/bfq-wf2q.c index e830715fe15d6..cdd164c82398a 100644 --- a/block/bfq-wf2q.c +++ b/block/bfq-wf2q.c @@ -788,25 +788,29 @@ __bfq_entity_update_weight_prio(struct bfq_service_tree *old_st, new_weight = entity->orig_weight * (bfqq ? bfqq->wr_coeff : 1); /* - * If the weight of the entity changes, remove the entity - * from its old weight counter (if there is a counter - * associated with the entity), and add it to the counter - * associated with its new weight. + * If the weight of the entity changes, and the entity is a + * queue, remove the entity from its old weight counter (if + * there is a counter associated with the entity). */ if (prev_weight != new_weight) { - root = bfqq ? &bfqd->queue_weights_tree : - &bfqd->group_weights_tree; - __bfq_weights_tree_remove(bfqd, entity, root); + if (bfqq) { + root = &bfqd->queue_weights_tree; + __bfq_weights_tree_remove(bfqd, bfqq, root); + } else + bfqd->num_active_groups--; } entity->weight = new_weight; /* - * Add the entity to its weights tree only if it is - * not associated with a weight-raised queue. + * Add the entity, if it is not a weight-raised queue, + * to the counter associated with its new weight. */ - if (prev_weight != new_weight && - (bfqq ? bfqq->wr_coeff == 1 : 1)) - /* If we get here, root has been initialized. */ - bfq_weights_tree_add(bfqd, entity, root); + if (prev_weight != new_weight) { + if (bfqq && bfqq->wr_coeff == 1) { + /* If we get here, root has been initialized. */ + bfq_weights_tree_add(bfqd, bfqq, root); + } else + bfqd->num_active_groups++; + } new_st->wsum += entity->weight; @@ -1014,8 +1018,7 @@ static void __bfq_activate_entity(struct bfq_entity *entity, container_of(entity, struct bfq_group, entity); struct bfq_data *bfqd = bfqg->bfqd; - bfq_weights_tree_add(bfqg->bfqd, entity, - &bfqd->group_weights_tree); + bfqd->num_active_groups++; } #endif @@ -1702,7 +1705,7 @@ void bfq_add_bfqq_busy(struct bfq_data *bfqd, struct bfq_queue *bfqq) if (!bfqq->dispatched) if (bfqq->wr_coeff == 1) - bfq_weights_tree_add(bfqd, &bfqq->entity, + bfq_weights_tree_add(bfqd, bfqq, &bfqd->queue_weights_tree); if (bfqq->wr_coeff > 1) -- 2.25.1

1 4

[PATCH openEuler-1.0-LTS 01/10] arm64: mm: Restore mm_cpumask (revert commit 38d96287504a ("arm64: mm: kill mm_cpumask usage"))
by Yang Yingliang 17 Dec '21

17 Dec '21

From: Takao Indoh <indou.takao(a)fujitsu.com> hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I4BLL0 CVE: NA --------------------------- mm_cpumask was deleted by the commit 38d96287504a ("arm64: mm: kill mm_cpumask usage") because it was not used at that time. Now this is needed to find appropriate CPUs for TLB flush, so this patch reverts this commit. Signed-off-by: QI Fuli <qi.fuli(a)fujitsu.com> Signed-off-by: Takao Indoh <indou.takao(a)fujitsu.com> Signed-off-by: Cheng Jian <cj.chengjian(a)huawei.com> Reviewed-by: Xie XiuQi <xiexiuqi(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- arch/arm64/kernel/smp.c | 6 ++++++ arch/arm64/mm/context.c | 2 ++ 2 files changed, 8 insertions(+) diff --git a/arch/arm64/kernel/smp.c b/arch/arm64/kernel/smp.c index fe562778de352..e86940d353a3e 100644 --- a/arch/arm64/kernel/smp.c +++ b/arch/arm64/kernel/smp.c @@ -387,6 +387,7 @@ asmlinkage notrace void secondary_start_kernel(void) */ mmgrab(mm); current->active_mm = mm; + cpumask_set_cpu(cpu, mm_cpumask(mm)); /* * TTBR0 is only used for the identity mapping at this stage. Make it @@ -489,6 +490,11 @@ int __cpu_disable(void) */ irq_migrate_all_off_this_cpu(); + /* + * Remove this CPU from the vm mask set of all processes. + */ + clear_tasks_mm_cpumask(cpu); + return 0; } diff --git a/arch/arm64/mm/context.c b/arch/arm64/mm/context.c index 2b80ceff5d6c2..27d1f3fec1cc9 100644 --- a/arch/arm64/mm/context.c +++ b/arch/arm64/mm/context.c @@ -207,6 +207,7 @@ static u64 new_context(struct mm_struct *mm, unsigned int cpu) set_asid: __set_bit(asid, asid_map); cur_idx = asid; + cpumask_clear(mm_cpumask(mm)); return idx2asid(asid) | generation; } @@ -254,6 +255,7 @@ void check_and_switch_context(struct mm_struct *mm, unsigned int cpu) switch_mm_fastpath: arm64_apply_bp_hardening(); + cpumask_set_cpu(cpu, mm_cpumask(mm)); /* * Defer TTBR0_EL1 setting for user threads to uaccess_enable() when -- 2.25.1

1 9

[PATCH openEuler-1.0-LTS] fget: check that the fd still exists after getting a ref to it
by Yang Yingliang 17 Dec '21

17 Dec '21

From: Linus Torvalds <torvalds(a)linux-foundation.org> stable inclusion from stable-4.19.220 commit 8bf31f9d9395b71af3ed33166a057cd3ec0c59da category: bugfix bugzilla: 185907 CVE: CVE-2021-4083 ------------------------------------------------- commit 054aa8d439b9185d4f5eb9a90282d1ce74772969 upstream. Jann Horn points out that there is another possible race wrt Unix domain socket garbage collection, somewhat reminiscent of the one fixed in commit cbcf01128d0a ("af_unix: fix garbage collect vs MSG_PEEK"). See the extended comment about the garbage collection requirements added to unix_peek_fds() by that commit for details. The race comes from how we can locklessly look up a file descriptor just as it is in the process of being closed, and with the right artificial timing (Jann added a few strategic 'mdelay(500)' calls to do that), the Unix domain socket garbage collector could see the reference count decrement of the close() happen before fget() took its reference to the file and the file was attached onto a new file descriptor. This is all (intentionally) correct on the 'struct file *' side, with RCU lookups and lockless reference counting very much part of the design. Getting that reference count out of order isn't a problem per se. But the garbage collector can get confused by seeing this situation of having seen a file not having any remaining external references and then seeing it being attached to an fd. In commit cbcf01128d0a ("af_unix: fix garbage collect vs MSG_PEEK") the fix was to serialize the file descriptor install with the garbage collector by taking and releasing the unix_gc_lock. That's not really an option here, but since this all happens when we are in the process of looking up a file descriptor, we can instead simply just re-check that the file hasn't been closed in the meantime, and just re-do the lookup if we raced with a concurrent close() of the same file descriptor. Reported-and-tested-by: Jann Horn <jannh(a)google.com> Acked-by: Miklos Szeredi <mszeredi(a)redhat.com> Signed-off-by: Linus Torvalds <torvalds(a)linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Baokun Li <libaokun1(a)huawei.com> Reviewed-by: Xiu Jianfeng <xiujianfeng(a)huawei.com> Reviewed-by: Zhang Yi <yi.zhang(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- fs/file.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/fs/file.c b/fs/file.c index 0caadedd3cc43..20a44d0ac400b 100644 --- a/fs/file.c +++ b/fs/file.c @@ -779,6 +779,10 @@ static struct file *__fget(unsigned int fd, fmode_t mask, unsigned int refs) file = NULL; else if (!get_file_rcu_many(file, refs)) goto loop; + else if (__fcheck_files(files, fd) != file) { + fput_many(file, refs); + goto loop; + } } rcu_read_unlock(); -- 2.25.1

1 0

the patches of ntfs3
by 尹秀江 17 Dec '21

17 Dec '21

1 0

[PATCH openEuler-1.0-LTS 1/4] Revert "Revert "ext4: Allow parallel DIO reads""
by Yang Yingliang 17 Dec '21

17 Dec '21

From: Zhang Xiaoxu <zhangxiaoxu5(a)huawei.com> hulk inclusion category: feature bugzilla: 185871, https://gitee.com/openeuler/kernel/issues/I4MIQ0 CVE: NA --------------------------- This reverts commit bb24e7e6309ae1bbc80062366b8f9bb2e8fcc025. In order to improve the dio reads performance, we enable the parallel dio reads. And make it configuratable Conflict: fs/ext4/Kconfig fs/ext4/inode.c Signed-off-by: Zhang Xiaoxu <zhangxiaoxu5(a)huawei.com> Reviewed-by: Zhang Yi <yi.zhang(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- fs/ext4/Kconfig | 5 +++++ fs/ext4/inode.c | 34 ++++++++++++++++++++++++++++------ 2 files changed, 33 insertions(+), 6 deletions(-) diff --git a/fs/ext4/Kconfig b/fs/ext4/Kconfig index a453cc87082b5..de836b75c6f2b 100644 --- a/fs/ext4/Kconfig +++ b/fs/ext4/Kconfig @@ -120,3 +120,8 @@ config EXT4_DEBUG If you select Y here, then you will be able to turn on debugging with a command such as: echo 1 > /sys/module/ext4/parameters/mballoc_debug +config EXT4_PARALLEL_DIO_READ + bool "EXT4 parallel dio reads" + depends on EXT4_FS + help + Enable parallel dio read. diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index fb4de707cd9ad..2f5231899d342 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -3795,15 +3795,31 @@ static ssize_t ext4_direct_IO_write(struct kiocb *iocb, struct iov_iter *iter) static ssize_t ext4_direct_IO_read(struct kiocb *iocb, struct iov_iter *iter) { + struct address_space *mapping = iocb->ki_filp->f_mapping; + struct inode *inode = mapping->host; +#ifdef CONFIG_EXT4_PARALLEL_DIO_READ + size_t count = iov_iter_count(iter); +#else int unlocked = 0; - struct inode *inode = iocb->ki_filp->f_mapping->host; +#endif ssize_t ret; loff_t offset = iocb->ki_pos; loff_t size = i_size_read(inode); if (offset >= size) return 0; - +#ifdef CONFIG_EXT4_PARALLEL_DIO_READ + /* + * Shared inode_lock is enough for us - it protects against concurrent + * writes & truncates and since we take care of writing back page cache, + * we are protected against page writeback as well. + */ + inode_lock_shared(inode); + ret = filemap_write_and_wait_range(mapping, iocb->ki_pos, + iocb->ki_pos + count - 1); + if (ret) + goto out_unlock; +#else if (ext4_should_dioread_nolock(inode)) { /* * Nolock dioread optimization may be dynamically disabled @@ -3813,18 +3829,24 @@ static ssize_t ext4_direct_IO_read(struct kiocb *iocb, struct iov_iter *iter) inode_dio_begin(inode); smp_mb(); if (unlikely(ext4_test_inode_state(inode, - EXT4_STATE_DIOREAD_LOCK))) + EXT4_STATE_DIOREAD_LOCK))) inode_dio_end(inode); else unlocked = 1; } +#endif + ret = __blockdev_direct_IO(iocb, inode, inode->i_sb->s_bdev, - iter, ext4_dio_get_block, - NULL, NULL, + iter, ext4_dio_get_block, NULL, NULL, +#ifdef CONFIG_EXT4_PARALLEL_DIO_READ + 0); +out_unlock: + inode_unlock_shared(inode); +#else unlocked ? 0 : DIO_LOCKING); - if (unlocked) inode_dio_end(inode); +#endif return ret; } -- 2.25.1

1 3

[PATCH openEuler-1.0-LTS 1/4] Revert "Revert "ext4: Allow parallel DIO reads""
by Yang Yingliang 17 Dec '21

17 Dec '21

From: Zhang Xiaoxu <zhangxiaoxu5(a)huawei.com> hulk inclusion category: feature bugzilla: 185871, https://gitee.com/openeuler/kernel/issues/I4MIQ0 CVE: NA --------------------------- This reverts commit bb24e7e6309ae1bbc80062366b8f9bb2e8fcc025. In order to improve the dio reads performance, we enable the parallel dio reads. And make it configuratable Conflict: fs/ext4/Kconfig fs/ext4/inode.c Signed-off-by: Zhang Xiaoxu <zhangxiaoxu5(a)huawei.com> Reviewed-by: Zhang Yi <yi.zhang(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- fs/ext4/Kconfig | 5 +++++ fs/ext4/inode.c | 34 ++++++++++++++++++++++++++++------ 2 files changed, 33 insertions(+), 6 deletions(-) diff --git a/fs/ext4/Kconfig b/fs/ext4/Kconfig index a453cc87082b5..de836b75c6f2b 100644 --- a/fs/ext4/Kconfig +++ b/fs/ext4/Kconfig @@ -120,3 +120,8 @@ config EXT4_DEBUG If you select Y here, then you will be able to turn on debugging with a command such as: echo 1 > /sys/module/ext4/parameters/mballoc_debug +config EXT4_PARALLEL_DIO_READ + bool "EXT4 parallel dio reads" + depends on EXT4_FS + help + Enable parallel dio read. diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index fb4de707cd9ad..2f5231899d342 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -3795,15 +3795,31 @@ static ssize_t ext4_direct_IO_write(struct kiocb *iocb, struct iov_iter *iter) static ssize_t ext4_direct_IO_read(struct kiocb *iocb, struct iov_iter *iter) { + struct address_space *mapping = iocb->ki_filp->f_mapping; + struct inode *inode = mapping->host; +#ifdef CONFIG_EXT4_PARALLEL_DIO_READ + size_t count = iov_iter_count(iter); +#else int unlocked = 0; - struct inode *inode = iocb->ki_filp->f_mapping->host; +#endif ssize_t ret; loff_t offset = iocb->ki_pos; loff_t size = i_size_read(inode); if (offset >= size) return 0; - +#ifdef CONFIG_EXT4_PARALLEL_DIO_READ + /* + * Shared inode_lock is enough for us - it protects against concurrent + * writes & truncates and since we take care of writing back page cache, + * we are protected against page writeback as well. + */ + inode_lock_shared(inode); + ret = filemap_write_and_wait_range(mapping, iocb->ki_pos, + iocb->ki_pos + count - 1); + if (ret) + goto out_unlock; +#else if (ext4_should_dioread_nolock(inode)) { /* * Nolock dioread optimization may be dynamically disabled @@ -3813,18 +3829,24 @@ static ssize_t ext4_direct_IO_read(struct kiocb *iocb, struct iov_iter *iter) inode_dio_begin(inode); smp_mb(); if (unlikely(ext4_test_inode_state(inode, - EXT4_STATE_DIOREAD_LOCK))) + EXT4_STATE_DIOREAD_LOCK))) inode_dio_end(inode); else unlocked = 1; } +#endif + ret = __blockdev_direct_IO(iocb, inode, inode->i_sb->s_bdev, - iter, ext4_dio_get_block, - NULL, NULL, + iter, ext4_dio_get_block, NULL, NULL, +#ifdef CONFIG_EXT4_PARALLEL_DIO_READ + 0); +out_unlock: + inode_unlock_shared(inode); +#else unlocked ? 0 : DIO_LOCKING); - if (unlocked) inode_dio_end(inode); +#endif return ret; } -- 2.25.1

1 3

[PATCH openEuler-1.0-LTS 1/4] net: hns3: fix VF RSS failed problem after PF enable multi-TCs
by Yang Yingliang 16 Dec '21

16 Dec '21

From: Guangbin Huang <huangguangbin2(a)huawei.com> mainline inclusion from mainline-v5.16-rc3 commit 8d2ad993aa05c0768f00c886c9d369cd97a337ac category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I4MWF6 CVE: NA ---------------------------- When PF is set to multi-TCs and configured mapping relationship between priorities and TCs, the hardware will active these settings for this PF and its VFs. In this case when VF just uses one TC and its rx packets contain priority, and if the priority is not mapped to TC0, as other TCs of VF is not valid, hardware always put this kind of packets to the queue 0. It cause this kind of packets of VF can not be used RSS function. To fix this problem, set tc mode of all unused TCs of VF to the setting of TC0, then rx packet with priority which map to unused TC will be direct to TC0. Fixes: e2cb1dec9779 ("net: hns3: Add HNS3 VF HCL(Hardware Compatibility Layer) Support") Signed-off-by: Guangbin Huang <huangguangbin2(a)huawei.com> Signed-off-by: Jakub Kicinski <kuba(a)kernel.org> Signed-off-by: Yonglong Liu <liuyonglong(a)huawei.com> Reviewed-by: li yongxin <liyongxin1(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c index 8e79818f149d2..ebe09a304aeda 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c @@ -686,9 +686,9 @@ static int hclgevf_set_rss_tc_mode(struct hclgevf_dev *hdev, u16 rss_size) roundup_size = ilog2(roundup_size); for (i = 0; i < HCLGEVF_MAX_TC_NUM; i++) { - tc_valid[i] = !!(hdev->hw_tc_map & BIT(i)); + tc_valid[i] = 1; tc_size[i] = roundup_size; - tc_offset[i] = rss_size * i; + tc_offset[i] = (hdev->hw_tc_map & BIT(i)) ? rss_size * i : 0; } hclgevf_cmd_setup_basic_desc(&desc, HCLGEVF_OPC_RSS_TC_MODE, false); -- 2.25.1

1 3

[PATCH openEuler-1.0-LTS 1/2] hugetlbfs: flush TLBs correctly after huge_pmd_unshare
by Yang Yingliang 16 Dec '21

16 Dec '21

From: Nadav Amit <namit(a)vmware.com> stable inclusion from linux-v4.19.219 commit b0313bc7f5fbb6beee327af39d818ffdc921821a category: bugfix bugzilla: 185854 CVE: CVE-2021-4002 ----------------------------------------------- commit a4a118f2eead1d6c49e00765de89878288d4b890 upstream. When __unmap_hugepage_range() calls to huge_pmd_unshare() succeed, a TLB flush is missing. This TLB flush must be performed before releasing the i_mmap_rwsem, in order to prevent an unshared PMDs page from being released and reused before the TLB flush took place. Arguably, a comprehensive solution would use mmu_gather interface to batch the TLB flushes and the PMDs page release, however it is not an easy solution: (1) try_to_unmap_one() and try_to_migrate_one() also call huge_pmd_unshare() and they cannot use the mmu_gather interface; and (2) deferring the release of the page reference for the PMDs page until after i_mmap_rwsem is dropeed can confuse huge_pmd_unshare() into thinking PMDs are shared when they are not. Fix __unmap_hugepage_range() by adding the missing TLB flush, and forcing a flush when unshare is successful. Fixes: 24669e58477e ("hugetlb: use mmu_gather instead of a temporary linked list for accumulating pages)" # 3.6 Signed-off-by: Nadav Amit <namit(a)vmware.com> Reviewed-by: Mike Kravetz <mike.kravetz(a)oracle.com> Cc: Aneesh Kumar K.V <aneesh.kumar(a)linux.vnet.ibm.com> Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu(a)jp.fujitsu.com> Cc: Andrew Morton <akpm(a)linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds(a)linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Conflicts: include/asm-generic/tlb.h mm/mmu_gather.c Signed-off-by: Liu Shixin <liushixin2(a)huawei.com> Reviewed-by: tong tiangen <tongtiangen(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- arch/arm/include/asm/tlb.h | 8 ++++++++ arch/ia64/include/asm/tlb.h | 10 ++++++++++ arch/s390/include/asm/tlb.h | 16 ++++++++++++++++ arch/sh/include/asm/tlb.h | 10 ++++++++++ arch/um/include/asm/tlb.h | 12 ++++++++++++ include/asm-generic/tlb.h | 2 ++ mm/hugetlb.c | 23 +++++++++++++++++++---- mm/mmu_gather.c | 10 ++++++++++ 8 files changed, 87 insertions(+), 4 deletions(-) diff --git a/arch/arm/include/asm/tlb.h b/arch/arm/include/asm/tlb.h index f854148c8d7c2..00baa13c158d7 100644 --- a/arch/arm/include/asm/tlb.h +++ b/arch/arm/include/asm/tlb.h @@ -280,6 +280,14 @@ tlb_remove_pmd_tlb_entry(struct mmu_gather *tlb, pmd_t *pmdp, unsigned long addr tlb_add_flush(tlb, addr); } +static inline void +tlb_flush_pmd_range(struct mmu_gather *tlb, unsigned long address, + unsigned long size) +{ + tlb_add_flush(tlb, address); + tlb_add_flush(tlb, address + size - PMD_SIZE); +} + #define pte_free_tlb(tlb, ptep, addr) __pte_free_tlb(tlb, ptep, addr) #define pmd_free_tlb(tlb, pmdp, addr) __pmd_free_tlb(tlb, pmdp, addr) #define pud_free_tlb(tlb, pudp, addr) pud_free((tlb)->mm, pudp) diff --git a/arch/ia64/include/asm/tlb.h b/arch/ia64/include/asm/tlb.h index 516355a774bfe..5d032d97c254e 100644 --- a/arch/ia64/include/asm/tlb.h +++ b/arch/ia64/include/asm/tlb.h @@ -268,6 +268,16 @@ __tlb_remove_tlb_entry (struct mmu_gather *tlb, pte_t *ptep, unsigned long addre tlb->end_addr = address + PAGE_SIZE; } +static inline void +tlb_flush_pmd_range(struct mmu_gather *tlb, unsigned long address, + unsigned long size) +{ + if (tlb->start_addr > address) + tlb->start_addr = address; + if (tlb->end_addr < address + size) + tlb->end_addr = address + size; +} + #define tlb_migrate_finish(mm) platform_tlb_migrate_finish(mm) #define tlb_start_vma(tlb, vma) do { } while (0) diff --git a/arch/s390/include/asm/tlb.h b/arch/s390/include/asm/tlb.h index b31c779cf5817..1df28a8e2f19e 100644 --- a/arch/s390/include/asm/tlb.h +++ b/arch/s390/include/asm/tlb.h @@ -116,6 +116,20 @@ static inline void tlb_remove_page_size(struct mmu_gather *tlb, return tlb_remove_page(tlb, page); } +static inline void tlb_flush_pmd_range(struct mmu_gather *tlb, + unsigned long address, unsigned long size) +{ + /* + * the range might exceed the original range that was provided to + * tlb_gather_mmu(), so we need to update it despite the fact it is + * usually not updated. + */ + if (tlb->start > address) + tlb->start = address; + if (tlb->end < address + size) + tlb->end = address + size; +} + /* * pte_free_tlb frees a pte table and clears the CRSTE for the * page table from the tlb. @@ -177,6 +191,8 @@ static inline void pud_free_tlb(struct mmu_gather *tlb, pud_t *pud, #define tlb_remove_tlb_entry(tlb, ptep, addr) do { } while (0) #define tlb_remove_pmd_tlb_entry(tlb, pmdp, addr) do { } while (0) #define tlb_migrate_finish(mm) do { } while (0) +#define tlb_flush_pmd_range(tlb, addr, sz) do { } while (0) + #define tlb_remove_huge_tlb_entry(h, tlb, ptep, address) \ tlb_remove_tlb_entry(tlb, ptep, address) diff --git a/arch/sh/include/asm/tlb.h b/arch/sh/include/asm/tlb.h index 77abe192fb43d..adcb0bfe238e3 100644 --- a/arch/sh/include/asm/tlb.h +++ b/arch/sh/include/asm/tlb.h @@ -127,6 +127,16 @@ static inline void tlb_remove_page_size(struct mmu_gather *tlb, return tlb_remove_page(tlb, page); } +static inline void +tlb_flush_pmd_range(struct mmu_gather *tlb, unsigned long address, + unsigned long size) +{ + if (tlb->start > address) + tlb->start = address; + if (tlb->end < address + size) + tlb->end = address + size; +} + #define tlb_remove_check_page_size_change tlb_remove_check_page_size_change static inline void tlb_remove_check_page_size_change(struct mmu_gather *tlb, unsigned int page_size) diff --git a/arch/um/include/asm/tlb.h b/arch/um/include/asm/tlb.h index dce6db147f245..02e61f6abfcab 100644 --- a/arch/um/include/asm/tlb.h +++ b/arch/um/include/asm/tlb.h @@ -130,6 +130,18 @@ static inline void tlb_remove_page_size(struct mmu_gather *tlb, return tlb_remove_page(tlb, page); } +static inline void +tlb_flush_pmd_range(struct mmu_gather *tlb, unsigned long address, + unsigned long size) +{ + tlb->need_flush = 1; + + if (tlb->start > address) + tlb->start = address; + if (tlb->end < address + size) + tlb->end = address + size; +} + /** * tlb_remove_tlb_entry - remember a pte unmapping for later tlb invalidation. * diff --git a/include/asm-generic/tlb.h b/include/asm-generic/tlb.h index 6be86c1c5c583..cfc86e3ba4606 100644 --- a/include/asm-generic/tlb.h +++ b/include/asm-generic/tlb.h @@ -139,6 +139,8 @@ void tlb_flush_mmu(struct mmu_gather *tlb); void arch_tlb_finish_mmu(struct mmu_gather *tlb, unsigned long start, unsigned long end, bool force); void tlb_flush_mmu_free(struct mmu_gather *tlb); +void tlb_flush_pmd_range(struct mmu_gather *tlb, unsigned long address, + unsigned long size); extern bool __tlb_remove_page_size(struct mmu_gather *tlb, struct page *page, int page_size); diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 2f65dad443ab3..b2ed9174bfd73 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -3595,6 +3595,7 @@ void __unmap_hugepage_range(struct mmu_gather *tlb, struct vm_area_struct *vma, unsigned long sz = huge_page_size(h); unsigned long mmun_start = start; /* For mmu_notifiers */ unsigned long mmun_end = end; /* For mmu_notifiers */ + bool force_flush = false; WARN_ON(!is_vm_hugetlb_page(vma)); BUG_ON(start & ~huge_page_mask(h)); @@ -3621,10 +3622,8 @@ void __unmap_hugepage_range(struct mmu_gather *tlb, struct vm_area_struct *vma, ptl = huge_pte_lock(h, mm, ptep); if (huge_pmd_unshare(mm, &address, ptep)) { spin_unlock(ptl); - /* - * We just unmapped a page of PMDs by clearing a PUD. - * The caller's TLB flush range should cover this area. - */ + tlb_flush_pmd_range(tlb, address & PUD_MASK, PUD_SIZE); + force_flush = true; continue; } @@ -3688,6 +3687,22 @@ void __unmap_hugepage_range(struct mmu_gather *tlb, struct vm_area_struct *vma, } mmu_notifier_invalidate_range_end(mm, mmun_start, mmun_end); tlb_end_vma(tlb, vma); + + /* + * If we unshared PMDs, the TLB flush was not recorded in mmu_gather. We + * could defer the flush until now, since by holding i_mmap_rwsem we + * guaranteed that the last refernece would not be dropped. But we must + * do the flushing before we return, as otherwise i_mmap_rwsem will be + * dropped and the last reference to the shared PMDs page might be + * dropped as well. + * + * In theory we could defer the freeing of the PMD pages as well, but + * huge_pmd_unshare() relies on the exact page_count for the PMD page to + * detect sharing, so we cannot defer the release of the page either. + * Instead, do flush now. + */ + if (force_flush) + tlb_flush_mmu_tlbonly(tlb); } void __unmap_hugepage_range_final(struct mmu_gather *tlb, diff --git a/mm/mmu_gather.c b/mm/mmu_gather.c index 2a9fbc4a37d59..c147a5aacfa96 100644 --- a/mm/mmu_gather.c +++ b/mm/mmu_gather.c @@ -139,6 +139,16 @@ bool __tlb_remove_page_size(struct mmu_gather *tlb, struct page *page, int page_ return false; } +void tlb_flush_pmd_range(struct mmu_gather *tlb, unsigned long address, + unsigned long size) +{ + if (tlb->page_size != 0 && tlb->page_size != PMD_SIZE) + tlb_flush_mmu(tlb); + + tlb->page_size = PMD_SIZE; + tlb->start = min(tlb->start, address); + tlb->end = max(tlb->end, address + size); +} #endif /* HAVE_GENERIC_MMU_GATHER */ #ifdef CONFIG_HAVE_RCU_TABLE_FREE -- 2.25.1

1 1

[PATCH openEuler-1.0-LTS 1/2] mm: share_pool: adjust sp_make_share_k2u behavior when coredump
by Yang Yingliang 16 Dec '21

16 Dec '21

From: Guo Mengqi <guomengqi3(a)huawei.com> ascend inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I4MUV2 CVE: NA when k2u is being executed ont the whole sharepool group, and one process coredumps, k2u will skip the coredumped process and continue on the rest processes in the group. Signed-off-by: Guo Mengqi <guomengqi3(a)huawei.com> Reviewed-by: Weilong Chen <chenweilong(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- mm/share_pool.c | 50 +++++++++++++++++++++++++++++++------------------ 1 file changed, 32 insertions(+), 18 deletions(-) diff --git a/mm/share_pool.c b/mm/share_pool.c index c9079e6a40b1a..ac97a68a70264 100644 --- a/mm/share_pool.c +++ b/mm/share_pool.c @@ -678,8 +678,25 @@ static unsigned long sp_mmap(struct mm_struct *mm, struct file *file, struct sp_area *spa, unsigned long *populate, unsigned long prot); static void sp_munmap(struct mm_struct *mm, unsigned long addr, unsigned long size); + +#define K2U_NORMAL 0 +#define K2U_COREDUMP 1 + +struct sp_k2u_context { + unsigned long kva; + unsigned long kva_aligned; + unsigned long size; + unsigned long size_aligned; + unsigned long sp_flags; + int state; + int spg_id; + bool to_task; + struct timespec64 start; + struct timespec64 end; +}; + static unsigned long sp_remap_kva_to_vma(unsigned long kva, struct sp_area *spa, - struct mm_struct *mm, unsigned long prot); + struct mm_struct *mm, unsigned long prot, struct sp_k2u_context *kc); static void free_sp_group_id(int spg_id) { @@ -1334,7 +1351,7 @@ int mg_sp_group_add_task(int pid, unsigned long prot, int spg_id) spin_unlock(&sp_area_lock); if (spa->type == SPA_TYPE_K2SPG && spa->kva) { - addr = sp_remap_kva_to_vma(spa->kva, spa, mm, prot); + addr = sp_remap_kva_to_vma(spa->kva, spa, mm, prot, NULL); if (IS_ERR_VALUE(addr)) pr_warn("add group remap k2u failed %ld\n", addr); @@ -2606,7 +2623,7 @@ static unsigned long __sp_remap_get_pfn(unsigned long kva) /* when called by k2u to group, always make sure rw_lock of spg is down */ static unsigned long sp_remap_kva_to_vma(unsigned long kva, struct sp_area *spa, - struct mm_struct *mm, unsigned long prot) + struct mm_struct *mm, unsigned long prot, struct sp_k2u_context *kc) { struct vm_area_struct *vma; unsigned long ret_addr; @@ -2618,6 +2635,8 @@ static unsigned long sp_remap_kva_to_vma(unsigned long kva, struct sp_area *spa, if (unlikely(mm->core_state)) { pr_err("k2u mmap: encountered coredump, abort\n"); ret_addr = -EBUSY; + if (kc) + kc->state = K2U_COREDUMP; goto put_mm; } @@ -2703,7 +2722,7 @@ static void *sp_make_share_kva_to_task(unsigned long kva, unsigned long size, un spa->kva = kva; - uva = (void *)sp_remap_kva_to_vma(kva, spa, current->mm, prot); + uva = (void *)sp_remap_kva_to_vma(kva, spa, current->mm, prot, NULL); __sp_area_drop(spa); if (IS_ERR(uva)) pr_err("remap k2u to task failed %ld\n", PTR_ERR(uva)); @@ -2731,6 +2750,8 @@ static void *sp_make_share_kva_to_spg(unsigned long kva, unsigned long size, struct mm_struct *mm; struct sp_group_node *spg_node; void *uva = ERR_PTR(-ENODEV); + struct sp_k2u_context kc; + unsigned long ret_addr = -ENODEV; down_read(&spg->rw_lock); spa = sp_alloc_area(size, sp_flags, spg, SPA_TYPE_K2SPG, current->tgid); @@ -2745,12 +2766,17 @@ static void *sp_make_share_kva_to_spg(unsigned long kva, unsigned long size, list_for_each_entry(spg_node, &spg->procs, proc_node) { mm = spg_node->master->mm; - uva = (void *)sp_remap_kva_to_vma(kva, spa, mm, spg_node->prot); - if (IS_ERR(uva)) { + kc.state = K2U_NORMAL; + ret_addr = sp_remap_kva_to_vma(kva, spa, mm, spg_node->prot, &kc); + if (IS_ERR_VALUE(ret_addr)) { + if (kc.state == K2U_COREDUMP) + continue; + uva = (void *)ret_addr; pr_err("remap k2u to spg failed %ld\n", PTR_ERR(uva)); __sp_free(spg, spa->va_start, spa_size(spa), mm); goto out; } + uva = (void *)ret_addr; } out: @@ -2775,18 +2801,6 @@ static bool vmalloc_area_set_flag(unsigned long kva, unsigned long flags) return false; } -struct sp_k2u_context { - unsigned long kva; - unsigned long kva_aligned; - unsigned long size; - unsigned long size_aligned; - unsigned long sp_flags; - int spg_id; - bool to_task; - struct timespec64 start; - struct timespec64 end; -}; - static void trace_sp_k2u_begin(struct sp_k2u_context *kc) { if (!sysctl_sp_perf_k2u) -- 2.25.1

1 1

[PATCH OLK-5.10 0/2]drm/hisilicon: Support hibmc get EDID
by Gou Hao 16 Dec '21

16 Dec '21

bugzilla: https://gitee.com/openeuler/kernel/issues/I469VQ Tian Tao (2): drm/hisilicon: Support i2c driver algorithms for bit-shift adapters drm/hisilicon: Features to support reading resolutions from EDID drivers/gpu/drm/hisilicon/hibmc/Makefile | 2 +- .../gpu/drm/hisilicon/hibmc/hibmc_drm_drv.h | 25 ++++- .../gpu/drm/hisilicon/hibmc/hibmc_drm_i2c.c | 99 +++++++++++++++++++ .../gpu/drm/hisilicon/hibmc/hibmc_drm_vdac.c | 38 ++++++- 4 files changed, 158 insertions(+), 6 deletions(-) create mode 100644 drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_i2c.c -- 2.20.1

1 2

[PATCH OLK-5.10 0/2]drm/hisilicon: Support hibmc get EDID
by Gou Hao 16 Dec '21

16 Dec '21

bugzilla: https://gitee.com/openeuler/kernel/issues/I469VQ Tian Tao (2): drm/hisilicon: Support i2c driver algorithms for bit-shift adapters drm/hisilicon: Features to support reading resolutions from EDID drivers/gpu/drm/hisilicon/hibmc/Makefile | 2 +- .../gpu/drm/hisilicon/hibmc/hibmc_drm_drv.h | 25 ++++- .../gpu/drm/hisilicon/hibmc/hibmc_drm_i2c.c | 99 +++++++++++++++++++ .../gpu/drm/hisilicon/hibmc/hibmc_drm_vdac.c | 38 ++++++- 4 files changed, 158 insertions(+), 6 deletions(-) create mode 100644 drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_i2c.c -- 2.20.1

1 1

[PATCH openEuler-1.0-LTS v10 00/26] support psi feature
by Liu Xinpeng 16 Dec '21

16 Dec '21

backport psi feature and avoid kabi change bugzilla: https://gitee.com/openeuler/kernel/issues/I47QS2 Changes since v9: 1.There are some problems when allocating cgroups, from Cheng Jian. First. We use `sizeof(struct cgroup_psi)` to allocate more space. This is inappropriate. Second, variable length arrays isn't the last member of cgroups_psi, so the offset of the psi will not be a compile-determined value. It's dangerous to use cgroup_psi(). 2.free path should change too,from Xie Xiuqi. Baruch Siach (1): psi: fix reference to kernel commandline enable Dan Schatzberg (1): kernel/sched/psi.c: expose pressure metrics on root cgroup Johannes Weiner (12): mm: workingset: tell cache transitions from workingset thrashing sched: loadavg: consolidate LOAD_INT, LOAD_FRAC, CALC_LOAD sched: loadavg: make calc_load_n() public sched: sched.h: make rq locking and clock functions available in stats.h sched: introduce this_rq_lock_irq() psi: pressure stall information for CPU, memory, and IO psi: cgroup support psi: make disabling/enabling easier for vendor kernels psi: fix aggregation idle shut-off psi: avoid divide-by-zero crash inside virtual machines fs: kernfs: add poll file operation sched/psi: Fix sampling error and rare div0 crashes with cgroups and high uptime Josef Bacik (1): blk-iolatency: use a percentile approache for ssd's Liu Xinpeng (2): psi:enable psi in config psi:avoid kabi change Miklos Szeredi (1): fuse: ignore PG_workingset after stealing Olof Johansson (1): kernel/sched/psi.c: simplify cgroup_move_task() Suren Baghdasaryan (6): psi: introduce state_mask to represent stalled psi states psi: make psi_enable static psi: rename psi fields in preparation for psi trigger addition psi: split update_stats into parts psi: track changed states include/: refactor headers to allow kthread.h inclusion in psi_types.h Yafang Shao (1): mm, memcg: add workingset_restore in memory.stat Documentation/accounting/psi.txt | 73 +++ Documentation/admin-guide/cgroup-v2.rst | 22 + Documentation/admin-guide/kernel-parameters.txt | 4 + arch/arm64/configs/openeuler_defconfig | 2 + arch/powerpc/platforms/cell/cpufreq_spudemand.c | 2 +- arch/powerpc/platforms/cell/spufs/sched.c | 9 +- arch/s390/appldata/appldata_os.c | 4 - arch/x86/configs/openeuler_defconfig | 2 + block/blk-iolatency.c | 183 +++++- drivers/cpuidle/governors/menu.c | 4 - drivers/spi/spi-rockchip.c | 1 + fs/fuse/dev.c | 1 + fs/kernfs/file.c | 31 +- fs/proc/loadavg.c | 3 - include/linux/cgroup-defs.h | 12 + include/linux/cgroup.h | 17 + include/linux/kernfs.h | 8 + include/linux/kthread.h | 4 + include/linux/page-flags.h | 5 + include/linux/psi.h | 55 ++ include/linux/psi_types.h | 95 +++ include/linux/sched.h | 13 + include/linux/sched/loadavg.h | 24 +- include/linux/swap.h | 1 + include/trace/events/mmflags.h | 1 + init/Kconfig | 28 + kernel/cgroup/cgroup.c | 69 +- kernel/debug/kdb/kdb_main.c | 7 +- kernel/fork.c | 4 + kernel/kthread.c | 3 + kernel/sched/Makefile | 1 + kernel/sched/core.c | 16 +- kernel/sched/loadavg.c | 139 ++-- kernel/sched/psi.c | 823 ++++++++++++++++++++++++ kernel/sched/sched.h | 178 ++--- kernel/sched/stats.h | 86 +++ kernel/workqueue.c | 23 + kernel/workqueue_internal.h | 6 +- mm/compaction.c | 5 + mm/filemap.c | 20 +- mm/huge_memory.c | 1 + mm/migrate.c | 2 + mm/page_alloc.c | 9 + mm/swap_state.c | 1 + mm/vmscan.c | 10 + mm/workingset.c | 113 +++- 46 files changed, 1841 insertions(+), 279 deletions(-) create mode 100644 Documentation/accounting/psi.txt create mode 100644 include/linux/psi.h create mode 100644 include/linux/psi_types.h create mode 100644 kernel/sched/psi.c -- 1.8.3.1

3 28

[PATCH openEuler-1.0-LTS 1/4] time: Normalize timespec64 before timespec64_compare()
by Yang Yingliang 15 Dec '21

15 Dec '21

From: Yu Liao <liaoyu15(a)huawei.com> hulk inclusion category: bugfix bugzilla: 185831, https://gitee.com/openeuler/kernel/issues/I4MO2G CVE: NA ------------------------------------------------- Passing unnormalized timespec64 to timespec64_compare() may cause incorrect results. For example: wall_to_monotonic = {tv_sec = -10, tv_nsec = 900000000} ts_delta = {tv_sec = -9, tv_nsec = -900000000} timespec64_compare() returns -1, but actually wall_to_monotonic > ts_delta. This will cause wall_to_monotonic to become a positive number. Use timespec64_sub() instead of direct subtraction to avoid this. Signed-off-by: Yu Liao <liaoyu15(a)huawei.com> Reviewed-by: Xiongfeng Wang <wangxiongfeng2(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- include/linux/time64.h | 2 ++ kernel/time/timekeeping.c | 3 +-- 2 files changed, 3 insertions(+), 2 deletions(-) diff --git a/include/linux/time64.h b/include/linux/time64.h index b58142aa0b1dd..e58f59674ad6d 100644 --- a/include/linux/time64.h +++ b/include/linux/time64.h @@ -62,6 +62,8 @@ static inline int timespec64_equal(const struct timespec64 *a, * lhs < rhs: return <0 * lhs == rhs: return 0 * lhs > rhs: return >0 + * + * Note: Both lhs and rhs must be normalized. */ static inline int timespec64_compare(const struct timespec64 *lhs, const struct timespec64 *rhs) { diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c index 64aa492f79422..17ba0a2f5f5d1 100644 --- a/kernel/time/timekeeping.c +++ b/kernel/time/timekeeping.c @@ -1236,8 +1236,7 @@ int do_settimeofday64(const struct timespec64 *ts) timekeeping_forward_now(tk); xt = tk_xtime(tk); - ts_delta.tv_sec = ts->tv_sec - xt.tv_sec; - ts_delta.tv_nsec = ts->tv_nsec - xt.tv_nsec; + ts_delta = timespec64_sub(*ts, xt); if (timespec64_compare(&tk->wall_to_monotonic, &ts_delta) > 0) { ret = -EINVAL; -- 2.25.1

1 3

[PATCH] scsi: pm8001: Fix memory leak in pm8001_send_abort_all()
by Jianglei Nie 15 Dec '21

15 Dec '21

openEuler inclusion category: bugfix bugzilla: NA CVE: NA ------------------------ In line 1767, sas_alloc_slow_task() allocates and initializes a sas_task structure. When some errors occur, line 1778 and line 1795 forget to free this structure, which will lead to a memory leak. There is a similar snippet of code in the same file (in function pm8001_send_read_log) as allocating and initializing in line 1812 as well as releasing the memory in line 1822 and line 1867. We can fix it by calling sas_free_task() when the res and ret is true and before the function returns. Signed-off-by: Jianglei Nie <niejianglei2021(a)163.com> --- drivers/scsi/pm8001/pm8001_hwi.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/drivers/scsi/pm8001/pm8001_hwi.c b/drivers/scsi/pm8001/pm8001_hwi.c index 124cb69740c6..25045a91620e 100644 --- a/drivers/scsi/pm8001/pm8001_hwi.c +++ b/drivers/scsi/pm8001/pm8001_hwi.c @@ -1774,8 +1774,10 @@ static void pm8001_send_abort_all(struct pm8001_hba_info *pm8001_ha, task->task_done = pm8001_task_done; res = pm8001_tag_alloc(pm8001_ha, &ccb_tag); - if (res) + if (res) { + sas_free_task(task); return; + } ccb = &pm8001_ha->ccb_info[ccb_tag]; ccb->device = pm8001_ha_dev; @@ -1791,8 +1793,10 @@ static void pm8001_send_abort_all(struct pm8001_hba_info *pm8001_ha, ret = pm8001_mpi_build_cmd(pm8001_ha, circularQ, opc, &task_abort, sizeof(task_abort), 0); - if (ret) + if (ret) { + sas_free_task(task); pm8001_tag_free(pm8001_ha, ccb_tag); + } } -- 2.25.1

1 0

perf spe-c2c record失败的问题
by haoxin 15 Dec '21

15 Dec '21

您好，我采用最细的openEuler-1.0测试perf spe-c2c的功能，但是发现/sys/devices下没有arm_spe_0这个目录执行perf spe-c2c record时显示： event syntax error: 'arm_spe_0/ts_enable=1,pct_enable=1,pa_enable=1,load_filter=1,jitter=1,store_filter=1,min_latency=0/' \___ Cannot find PMU `arm_spe_0'. Missing kernel support? 想问下您这是为什么？需要哪里开启spe功能吗？

3 2

[PATCH openEuler-1.0-LTS] fix kabi effect by change in md_rdev
by Yang Yingliang 14 Dec '21

14 Dec '21

From: zhangwensheng <zhangwensheng5(a)huawei.com> hulk inclusion category: bugfix bugzilla: 185891 CVE: NA -------------------------------- because enlarge the last_events in md_rdev, to fix the kabi change; Signed-off-by: zhangwensheng <zhangwensheng5(a)huawei.com> Reviewed-by: Zhang Yi <yi.zhang(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/md/md.h | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/drivers/md/md.h b/drivers/md/md.h index e745f701e1a82..422af63f1e1ee 100644 --- a/drivers/md/md.h +++ b/drivers/md/md.h @@ -47,8 +47,16 @@ struct md_rdev { sector_t sectors; /* Device size (in 512bytes sectors) */ struct mddev *mddev; /* RAID array if running */ - long long last_events; /* IO event timestamp */ + /* + * IO event timestamp, this pos has 64 bit space, + * can enlarge it solving its small size. + */ +#ifndef __GENKSYMS__ + long long last_events; +#else + int last_events; +#endif /* * If meta_bdev is non-NULL, it means that a separate device is * being used to store the metadata (superblock/bitmap) which -- 2.25.1

1 0

[PATCH openEuler-1.0-LTS] Revert "dm space maps: don't reset space map allocation cursor when committing"
by Yang Yingliang 14 Dec '21

14 Dec '21

From: Luo Meng <luomeng12(a)huawei.com> hulk inclusion category: bugfix bugzilla: 185894, https://gitee.com/openeuler/kernel/issues/I4MITL CVE: NA ----------------------------------------------- This reverts commit 38cb5d45845b9d8962eb26660c14b7f05abc602f. Signed-off-by: Luo Meng <luomeng12(a)huawei.com> Reviewed-by: Jason Yan <yanaijie(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/md/persistent-data/dm-space-map-disk.c | 9 +-------- drivers/md/persistent-data/dm-space-map-metadata.c | 9 +-------- 2 files changed, 2 insertions(+), 16 deletions(-) diff --git a/drivers/md/persistent-data/dm-space-map-disk.c b/drivers/md/persistent-data/dm-space-map-disk.c index e0acae7a3815d..bf4c5e2ccb6ff 100644 --- a/drivers/md/persistent-data/dm-space-map-disk.c +++ b/drivers/md/persistent-data/dm-space-map-disk.c @@ -171,14 +171,6 @@ static int sm_disk_new_block(struct dm_space_map *sm, dm_block_t *b) * Any block we allocate has to be free in both the old and current ll. */ r = sm_ll_find_common_free_block(&smd->old_ll, &smd->ll, smd->begin, smd->ll.nr_blocks, b); - if (r == -ENOSPC) { - /* - * There's no free block between smd->begin and the end of the metadata device. - * We search before smd->begin in case something has been freed. - */ - r = sm_ll_find_common_free_block(&smd->old_ll, &smd->ll, 0, smd->begin, b); - } - if (r) return r; @@ -207,6 +199,7 @@ static int sm_disk_commit(struct dm_space_map *sm) return r; memcpy(&smd->old_ll, &smd->ll, sizeof(smd->old_ll)); + smd->begin = 0; smd->nr_allocated_this_transaction = 0; r = sm_disk_get_nr_free(sm, &nr_free); diff --git a/drivers/md/persistent-data/dm-space-map-metadata.c b/drivers/md/persistent-data/dm-space-map-metadata.c index da439ac857963..9e3c64ec2026f 100644 --- a/drivers/md/persistent-data/dm-space-map-metadata.c +++ b/drivers/md/persistent-data/dm-space-map-metadata.c @@ -452,14 +452,6 @@ static int sm_metadata_new_block_(struct dm_space_map *sm, dm_block_t *b) * Any block we allocate has to be free in both the old and current ll. */ r = sm_ll_find_common_free_block(&smm->old_ll, &smm->ll, smm->begin, smm->ll.nr_blocks, b); - if (r == -ENOSPC) { - /* - * There's no free block between smm->begin and the end of the metadata device. - * We search before smm->begin in case something has been freed. - */ - r = sm_ll_find_common_free_block(&smm->old_ll, &smm->ll, 0, smm->begin, b); - } - if (r) return r; @@ -511,6 +503,7 @@ static int sm_metadata_commit(struct dm_space_map *sm) return r; memcpy(&smm->old_ll, &smm->ll, sizeof(smm->old_ll)); + smm->begin = 0; smm->allocated_this_transaction = 0; return 0; -- 2.25.1

1 0

[PATCH openEuler-1.0-LTS 1/4] nvme-fabrics: reject I/O to offline device
by Yang Yingliang 14 Dec '21

14 Dec '21

From: Victor Gladkov <victor.gladkov(a)kioxia.com> mainline inclusion from mainline-v5.11-rc1 commit 8c4dfea97f15b80097b3f882ca428fb2751ec30c category: bugfix bugzilla: NA CVE: NA Link: https://gitee.com/openeuler/kernel/issues/I4JFPM?from=project-issue ------------------------------------------------- Commands get stuck while Host NVMe-oF controller is in reconnect state. The controller enters into reconnect state when it loses connection with the target. It tries to reconnect every 10 seconds (default) until a successful reconnect or until the reconnect time-out is reached. The default reconnect time out is 10 minutes. Applications are expecting commands to complete with success or error within a certain timeout (30 seconds by default). The NVMe host is enforcing that timeout while it is connected, but during reconnect the timeout is not enforced and commands may get stuck for a long period or even forever. To fix this long delay due to the default timeout, introduce new "fast_io_fail_tmo" session parameter. The timeout is measured in seconds from the controller reconnect and any command beyond that timeout is rejected. The new parameter value may be passed during 'connect'. The default value of -1 means no timeout (similar to current behavior). Signed-off-by: Victor Gladkov <victor.gladkov(a)kioxia.com> Signed-off-by: Chaitanya Kulkarni <chaitanya.kulkarni(a)wdc.com> Reviewed-by: Hannes Reinecke <hare(a)suse.de> Reviewed-by: Sagi Grimberg <sagi(a)grimberg.me> Reviewed-by: Chao Leng <lengchao(a)huawei.com> Signed-off-by: Christoph Hellwig <hch(a)lst.de> conflicts:: drivers/nvme/host/core.c [adjust context] Signed-off-by: jiangtao <jiangtao62(a)huawei.com> Reviewed-by: chengjike <chengjike.cheng(a)huawei.com> Reviewed-by: Ao Sun <sunao.sun(a)huawei.com> Reviewed-by: Zhenwei Yang <yangzhenwei(a)huawei.com> Reviewed-by: Hou Tao <houtao1(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/nvme/host/core.c | 51 +++++++++++++++++++++++++++++++++-- drivers/nvme/host/fabrics.c | 28 ++++++++++++++++--- drivers/nvme/host/fabrics.h | 5 ++++ drivers/nvme/host/multipath.c | 2 ++ drivers/nvme/host/nvme.h | 3 +++ 5 files changed, 83 insertions(+), 6 deletions(-) diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c index b45ae13aa5fe6..c9ca375a5ddb9 100644 --- a/drivers/nvme/host/core.c +++ b/drivers/nvme/host/core.c @@ -130,6 +130,37 @@ static void nvme_queue_scan(struct nvme_ctrl *ctrl) queue_work(nvme_wq, &ctrl->scan_work); } +static void nvme_failfast_work(struct work_struct *work) +{ + struct nvme_ctrl *ctrl = container_of(to_delayed_work(work), + struct nvme_ctrl, failfast_work); + + if (ctrl->state != NVME_CTRL_CONNECTING) + return; + + set_bit(NVME_CTRL_FAILFAST_EXPIRED, &ctrl->flags); + dev_info(ctrl->device, "failfast expired\n"); + nvme_kick_requeue_lists(ctrl); +} + +static inline void nvme_start_failfast_work(struct nvme_ctrl *ctrl) +{ + if (!ctrl->opts || ctrl->opts->fast_io_fail_tmo == -1) + return; + + schedule_delayed_work(&ctrl->failfast_work, + ctrl->opts->fast_io_fail_tmo * HZ); +} + +static inline void nvme_stop_failfast_work(struct nvme_ctrl *ctrl) +{ + if (!ctrl->opts) + return; + + cancel_delayed_work_sync(&ctrl->failfast_work); + clear_bit(NVME_CTRL_FAILFAST_EXPIRED, &ctrl->flags); +} + int nvme_reset_ctrl(struct nvme_ctrl *ctrl) { if (!nvme_change_ctrl_state(ctrl, NVME_CTRL_RESETTING)) @@ -385,8 +416,21 @@ bool nvme_change_ctrl_state(struct nvme_ctrl *ctrl, ctrl->state = new_state; spin_unlock_irqrestore(&ctrl->lock, flags); - if (changed && ctrl->state == NVME_CTRL_LIVE) - nvme_kick_requeue_lists(ctrl); + if (changed) { + switch (ctrl->state) { + case NVME_CTRL_LIVE: + if (old_state == NVME_CTRL_CONNECTING) + nvme_stop_failfast_work(ctrl); + nvme_kick_requeue_lists(ctrl); + break; + case NVME_CTRL_CONNECTING: + if (old_state == NVME_CTRL_RESETTING) + nvme_start_failfast_work(ctrl); + break; + default: + break; + } + } return changed; } EXPORT_SYMBOL_GPL(nvme_change_ctrl_state); @@ -3712,6 +3756,7 @@ void nvme_stop_ctrl(struct nvme_ctrl *ctrl) { nvme_mpath_stop(ctrl); nvme_stop_keep_alive(ctrl); + nvme_stop_failfast_work(ctrl); flush_work(&ctrl->async_event_work); cancel_work_sync(&ctrl->fw_act_work); if (ctrl->ops->stop_ctrl) @@ -3776,6 +3821,7 @@ int nvme_init_ctrl(struct nvme_ctrl *ctrl, struct device *dev, int ret; ctrl->state = NVME_CTRL_NEW; + clear_bit(NVME_CTRL_FAILFAST_EXPIRED, &ctrl->flags); spin_lock_init(&ctrl->lock); mutex_init(&ctrl->scan_lock); INIT_LIST_HEAD(&ctrl->namespaces); @@ -3791,6 +3837,7 @@ int nvme_init_ctrl(struct nvme_ctrl *ctrl, struct device *dev, INIT_DELAYED_WORK(&ctrl->ka_work, nvme_keep_alive_work); memset(&ctrl->ka_cmd, 0, sizeof(ctrl->ka_cmd)); ctrl->ka_cmd.common.opcode = nvme_admin_keep_alive; + INIT_DELAYED_WORK(&ctrl->failfast_work, nvme_failfast_work); BUILD_BUG_ON(NVME_DSM_MAX_RANGES * sizeof(struct nvme_dsm_range) > PAGE_SIZE); diff --git a/drivers/nvme/host/fabrics.c b/drivers/nvme/host/fabrics.c index 738794af3f389..650b3bd89968f 100644 --- a/drivers/nvme/host/fabrics.c +++ b/drivers/nvme/host/fabrics.c @@ -550,6 +550,7 @@ blk_status_t nvmf_fail_nonready_command(struct nvme_ctrl *ctrl, { if (ctrl->state != NVME_CTRL_DELETING && ctrl->state != NVME_CTRL_DEAD && + !test_bit(NVME_CTRL_FAILFAST_EXPIRED, &ctrl->flags) && !blk_noretry_request(rq) && !(rq->cmd_flags & REQ_NVME_MPATH)) return BLK_STS_RESOURCE; @@ -607,6 +608,7 @@ static const match_table_t opt_tokens = { { NVMF_OPT_HOST_TRADDR, "host_traddr=%s" }, { NVMF_OPT_HOST_ID, "hostid=%s" }, { NVMF_OPT_DUP_CONNECT, "duplicate_connect" }, + { NVMF_OPT_FAIL_FAST_TMO, "fast_io_fail_tmo=%d" }, { NVMF_OPT_ERR, NULL } }; @@ -626,6 +628,7 @@ static int nvmf_parse_options(struct nvmf_ctrl_options *opts, opts->reconnect_delay = NVMF_DEF_RECONNECT_DELAY; opts->kato = NVME_DEFAULT_KATO; opts->duplicate_connect = false; + opts->fast_io_fail_tmo = NVMF_DEF_FAIL_FAST_TMO; options = o = kstrdup(buf, GFP_KERNEL); if (!options) @@ -750,6 +753,17 @@ static int nvmf_parse_options(struct nvmf_ctrl_options *opts, pr_warn("ctrl_loss_tmo < 0 will reconnect forever\n"); ctrl_loss_tmo = token; break; + case NVMF_OPT_FAIL_FAST_TMO: + if (match_int(args, &token)) { + ret = -EINVAL; + goto out; + } + + if (token >= 0) + pr_warn("I/O will fail on after %d sec reconnect\n", + token); + opts->fast_io_fail_tmo = token; + break; case NVMF_OPT_HOSTNQN: if (opts->host) { pr_err("hostnqn already user-assigned: %s\n", @@ -830,11 +844,17 @@ static int nvmf_parse_options(struct nvmf_ctrl_options *opts, opts->nr_io_queues = 0; opts->duplicate_connect = true; } - if (ctrl_loss_tmo < 0) + + if (ctrl_loss_tmo < 0) { opts->max_reconnects = -1; - else + } else { opts->max_reconnects = DIV_ROUND_UP(ctrl_loss_tmo, opts->reconnect_delay); + if (ctrl_loss_tmo < opts->fast_io_fail_tmo) + pr_warn("failfast tmo (%d) > ctrl_loss_tmo (%d)\n", + opts->fast_io_fail_tmo, + ctrl_loss_tmo); + } if (!opts->host) { kref_get(&nvmf_default_host->ref); @@ -903,8 +923,8 @@ EXPORT_SYMBOL_GPL(nvmf_free_options); #define NVMF_REQUIRED_OPTS (NVMF_OPT_TRANSPORT | NVMF_OPT_NQN) #define NVMF_ALLOWED_OPTS (NVMF_OPT_QUEUE_SIZE | NVMF_OPT_NR_IO_QUEUES | \ NVMF_OPT_KATO | NVMF_OPT_HOSTNQN | \ - NVMF_OPT_HOST_ID | NVMF_OPT_DUP_CONNECT) - + NVMF_OPT_HOST_ID | NVMF_OPT_DUP_CONNECT |\ + NVMF_OPT_FAIL_FAST_TMO) static struct nvme_ctrl * nvmf_create_ctrl(struct device *dev, const char *buf, size_t count) { diff --git a/drivers/nvme/host/fabrics.h b/drivers/nvme/host/fabrics.h index 188ebbeec32ce..645ed75f22980 100644 --- a/drivers/nvme/host/fabrics.h +++ b/drivers/nvme/host/fabrics.h @@ -24,6 +24,8 @@ /* default to 600 seconds of reconnect attempts before giving up */ #define NVMF_DEF_CTRL_LOSS_TMO 600 #define NVMF_DEF_RECONNECT_FOREVER -1 +/* default is -1: the fail fast mechanism is disabled */ +#define NVMF_DEF_FAIL_FAST_TMO -1 /* * Define a host as seen by the target. We allocate one at boot, but also @@ -59,6 +61,7 @@ enum { NVMF_OPT_CTRL_LOSS_TMO = 1 << 11, NVMF_OPT_HOST_ID = 1 << 12, NVMF_OPT_DUP_CONNECT = 1 << 13, + NVMF_OPT_FAIL_FAST_TMO = 1 << 20, }; /** @@ -86,6 +89,7 @@ enum { * @max_reconnects: maximum number of allowed reconnect attempts before removing * the controller, (-1) means reconnect forever, zero means remove * immediately; + * @fast_io_fail_tmo: Fast I/O fail timeout in seconds */ struct nvmf_ctrl_options { unsigned mask; @@ -102,6 +106,7 @@ struct nvmf_ctrl_options { unsigned int kato; struct nvmf_host *host; int max_reconnects; + int fast_io_fail_tmo; }; /* diff --git a/drivers/nvme/host/multipath.c b/drivers/nvme/host/multipath.c index f4616df1f7f15..9201946568630 100644 --- a/drivers/nvme/host/multipath.c +++ b/drivers/nvme/host/multipath.c @@ -198,6 +198,8 @@ static bool nvme_available_path(struct nvme_ns_head *head) struct nvme_ns *ns; list_for_each_entry_rcu(ns, &head->list, siblings) { + if (test_bit(NVME_CTRL_FAILFAST_EXPIRED, &ns->ctrl->flags)) + continue; switch (ns->ctrl->state) { case NVME_CTRL_LIVE: case NVME_CTRL_RESETTING: diff --git a/drivers/nvme/host/nvme.h b/drivers/nvme/host/nvme.h index 85f814a91ce9c..cd66364eb5429 100644 --- a/drivers/nvme/host/nvme.h +++ b/drivers/nvme/host/nvme.h @@ -211,6 +211,7 @@ struct nvme_ctrl { struct work_struct scan_work; struct work_struct async_event_work; struct delayed_work ka_work; + struct delayed_work failfast_work; struct nvme_command ka_cmd; struct work_struct fw_act_work; unsigned long events; @@ -245,6 +246,8 @@ struct nvme_ctrl { u16 icdoff; u16 maxcmd; int nr_reconnects; + unsigned long flags; +#define NVME_CTRL_FAILFAST_EXPIRED 0 struct nvmf_ctrl_options *opts; struct page *discard_page; -- 2.25.1

1 3

[PATCH openEuler-1.0-LTS 01/12] nvme: add proper discard setup for the multipath device
by Yang Yingliang 14 Dec '21

14 Dec '21

From: Christoph Hellwig <hch(a)lst.de> mainline inclusion from mainline-v5.1-rc1 commit 2631857160ecbea04e54423f5053133fe2b6ea45 category: bugfix bugzilla: NA CVE: NA Link: https://gitee.com/openeuler/kernel/issues/I4JFBE?from=project-issue ------------------------------------------------- Add a gendisk argument to nvme_config_discard so that the call to nvme_update_disk_info for the multipath device node updates the proper request_queue. Signed-off-by: Christoph Hellwig <hch(a)lst.de> Reported-by: Sagi Grimberg <sagi(a)grimberg.me> Reviewed-by: Keith Busch <keith.busch(a)intel.com> Reviewed-by: Max Gurtovoy <maxg(a)mellanox.com> Tested-by: Sagi Grimberg <sagi(a)grimberg.me> Signed-off-by: Jens Axboe <axboe(a)kernel.dk> conflicts: drivers/nvme/host/core.c [adjust context] Signed-off-by: chengjike <chengjike.cheng(a)huawei.com> Reviewed-by: Ao Sun <sunao.sun(a)huawei.com> Reviewed-by: Zhenwei Yang <yangzhenwei(a)huawei.com> Reviewed-by: Hou Tao <houtao1(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/nvme/host/core.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c index 522ab9c0133ca..7c4a97e2cd93a 100644 --- a/drivers/nvme/host/core.c +++ b/drivers/nvme/host/core.c @@ -1558,10 +1558,10 @@ static void nvme_set_chunk_size(struct nvme_ns *ns) blk_queue_chunk_sectors(ns->queue, rounddown_pow_of_two(chunk_size)); } -static void nvme_config_discard(struct nvme_ns *ns) +static void nvme_config_discard(struct gendisk *disk, struct nvme_ns *ns) { struct nvme_ctrl *ctrl = ns->ctrl; - struct request_queue *queue = ns->queue; + struct request_queue *queue = disk->queue; u32 size = queue_logical_block_size(queue); if (!(ctrl->oncs & NVME_CTRL_ONCS_DSM)) { @@ -1647,7 +1647,7 @@ static void nvme_update_disk_info(struct gendisk *disk, capacity = 0; set_capacity(disk, capacity); - nvme_config_discard(ns); + nvme_config_discard(disk, ns); if (id->nsattr & (1 << 0)) set_disk_ro(disk, true); -- 2.25.1

1 11

[PATCH openEuler-1.0-LTS 1/2] md: Fix undefined behaviour in is_mddev_idle
by Yang Yingliang 14 Dec '21

14 Dec '21

From: zhangwensheng <zhangwensheng5(a)huawei.com> hulk inclusion category: bugfix bugzilla: 185891 CVE: NA -------------------------------- UBSAN reports this problem: [ 5984.281385] UBSAN: Undefined behaviour in drivers/md/md.c:8175:15 [ 5984.281390] signed integer overflow: [ 5984.281393] -2147483291 - 2072033152 cannot be represented in type 'int' [ 5984.281400] CPU: 25 PID: 1854 Comm: md101_resync Kdump: loaded Not tainted 4.19.90 [ 5984.281404] Hardware name: Huawei TaiShan 200 (Model 5280)/BC82AMDDA [ 5984.281406] Call trace: [ 5984.281415] dump_backtrace+0x0/0x310 [ 5984.281418] show_stack+0x28/0x38 [ 5984.281425] dump_stack+0xec/0x15c [ 5984.281430] ubsan_epilogue+0x18/0x84 [ 5984.281434] handle_overflow+0x14c/0x19c [ 5984.281439] __ubsan_handle_sub_overflow+0x34/0x44 [ 5984.281445] is_mddev_idle+0x338/0x3d8 [ 5984.281449] md_do_sync+0x1bb8/0x1cf8 [ 5984.281452] md_thread+0x220/0x288 [ 5984.281457] kthread+0x1d8/0x1e0 [ 5984.281461] ret_from_fork+0x10/0x18 When the stat aacum of the disk is greater than INT_MAX, its value becomes negative after casting to 'int', which may lead to overflow after subtracting a positive number. In the same way, when the value of sync_io is greater than INT_MAX,overflow may also occur. These situations will lead to undefined behavior. Otherwise, if the stat accum of the disk is close to INT_MAX when creating raid arrays, the initial value of last_events would be set close to INT_MAX when mddev initializes IO event counters. 'curr_events - rdev->last_events > 64' will always false during synchronization. If all the disks of mddev are in this case, is_mddev_idle() will always return 1, which may cause non-sync IO is very slow. To address these problems, need to use 64bit signed integer type for sync_io,last_events, and curr_events. Signed-off-by: zhangwensheng <zhangwensheng5(a)huawei.com> Reviewed-by: Hou Tao <houtao1(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/md/md.c | 7 ++++--- drivers/md/md.h | 6 +++--- include/linux/genhd.h | 2 +- 3 files changed, 8 insertions(+), 7 deletions(-) diff --git a/drivers/md/md.c b/drivers/md/md.c index 409ec5ffd28d3..8a68ff71fbbf8 100644 --- a/drivers/md/md.c +++ b/drivers/md/md.c @@ -8161,14 +8161,15 @@ static int is_mddev_idle(struct mddev *mddev, int init) { struct md_rdev *rdev; int idle; - int curr_events; + long long curr_events; idle = 1; rcu_read_lock(); rdev_for_each_rcu(rdev, mddev) { struct gendisk *disk = rdev->bdev->bd_contains->bd_disk; - curr_events = (int)part_stat_read_accum(&disk->part0, sectors) - - atomic_read(&disk->sync_io); + curr_events = + (long long)part_stat_read_accum(&disk->part0, sectors) - + atomic64_read(&disk->sync_io_sectors); /* sync IO will cause sync_io to increase before the disk_stats * as sync_io is counted when a request starts, and * disk_stats is counted when it completes. diff --git a/drivers/md/md.h b/drivers/md/md.h index ee3c9768ff9e7..e745f701e1a82 100644 --- a/drivers/md/md.h +++ b/drivers/md/md.h @@ -47,7 +47,7 @@ struct md_rdev { sector_t sectors; /* Device size (in 512bytes sectors) */ struct mddev *mddev; /* RAID array if running */ - int last_events; /* IO event timestamp */ + long long last_events; /* IO event timestamp */ /* * If meta_bdev is non-NULL, it means that a separate device is @@ -528,12 +528,12 @@ extern void mddev_unlock(struct mddev *mddev); static inline void md_sync_acct(struct block_device *bdev, unsigned long nr_sectors) { - atomic_add(nr_sectors, &bdev->bd_contains->bd_disk->sync_io); + atomic64_add(nr_sectors, &bdev->bd_contains->bd_disk->sync_io_sectors); } static inline void md_sync_acct_bio(struct bio *bio, unsigned long nr_sectors) { - atomic_add(nr_sectors, &bio->bi_disk->sync_io); + atomic64_add(nr_sectors, &bio->bi_disk->sync_io_sectors); } struct md_personality diff --git a/include/linux/genhd.h b/include/linux/genhd.h index 9c398294b6279..38737c853b28d 100644 --- a/include/linux/genhd.h +++ b/include/linux/genhd.h @@ -212,7 +212,6 @@ struct gendisk { struct kobject *slave_dir; struct timer_rand_state *random; - atomic_t sync_io; /* RAID */ struct disk_events *ev; #ifdef CONFIG_BLK_DEV_INTEGRITY struct kobject integrity_kobj; @@ -220,6 +219,7 @@ struct gendisk { int node_id; struct badblocks *bb; struct lockdep_map lockdep_map; + atomic64_t sync_io_sectors; /* RAID */ #ifndef __GENKSYMS__ unsigned long *user_ro_bitmap; -- 2.25.1

1 1

[PATCH openEuler-1.0-LTS 1/4] xfs: ensure that the inode uid/gid match values match the icdinode ones
by Yang Yingliang 14 Dec '21

14 Dec '21

From: Christoph Hellwig <hch(a)lst.de> mainline inclusion from mainline-v5.6-rc4 commit 3d8f2821502d0b60bac2789d0bea951fda61de0c category: bugfix bugzilla: 185881 CVE: CVE-2021-4037 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… ------------------------------------------------- Instead of only synchronizing the uid/gid values in xfs_setup_inode, ensure that they always match to prepare for removing the icdinode fields. Signed-off-by: Christoph Hellwig <hch(a)lst.de> Reviewed-by: Darrick J. Wong <darrick.wong(a)oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong(a)oracle.com> Signed-off-by: Guo Xuenan <guoxuenan(a)huawei.com> Reviewed-by: Zhang Yi <yi.zhang(a)huawei.com> Reviewed-by: Xiu Jianfeng <xiujianfeng(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- fs/xfs/libxfs/xfs_inode_buf.c | 2 ++ fs/xfs/xfs_icache.c | 4 ++++ fs/xfs/xfs_inode.c | 8 ++++++-- fs/xfs/xfs_iops.c | 3 --- 4 files changed, 12 insertions(+), 5 deletions(-) diff --git a/fs/xfs/libxfs/xfs_inode_buf.c b/fs/xfs/libxfs/xfs_inode_buf.c index 82c4374acb4c6..ed58741e09ea8 100644 --- a/fs/xfs/libxfs/xfs_inode_buf.c +++ b/fs/xfs/libxfs/xfs_inode_buf.c @@ -225,7 +225,9 @@ xfs_inode_from_disk( to->di_format = from->di_format; to->di_uid = be32_to_cpu(from->di_uid); + inode->i_uid = xfs_uid_to_kuid(to->di_uid); to->di_gid = be32_to_cpu(from->di_gid); + inode->i_gid = xfs_gid_to_kgid(to->di_gid); to->di_flushiter = be16_to_cpu(from->di_flushiter); /* diff --git a/fs/xfs/xfs_icache.c b/fs/xfs/xfs_icache.c index 56e9043bddc71..ceee27b703842 100644 --- a/fs/xfs/xfs_icache.c +++ b/fs/xfs/xfs_icache.c @@ -286,6 +286,8 @@ xfs_reinit_inode( uint64_t version = inode_peek_iversion(inode); umode_t mode = inode->i_mode; dev_t dev = inode->i_rdev; + kuid_t uid = inode->i_uid; + kgid_t gid = inode->i_gid; error = inode_init_always(mp->m_super, inode); @@ -294,6 +296,8 @@ xfs_reinit_inode( inode_set_iversion_queried(inode, version); inode->i_mode = mode; inode->i_rdev = dev; + inode->i_uid = uid; + inode->i_gid = gid; return error; } diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c index c3226c48079da..70d5f0e2aa129 100644 --- a/fs/xfs/xfs_inode.c +++ b/fs/xfs/xfs_inode.c @@ -812,15 +812,19 @@ xfs_ialloc( inode->i_mode = mode; set_nlink(inode, nlink); - ip->i_d.di_uid = xfs_kuid_to_uid(current_fsuid()); - ip->i_d.di_gid = xfs_kgid_to_gid(current_fsgid()); + inode->i_uid = current_fsuid(); + ip->i_d.di_uid = xfs_kuid_to_uid(inode->i_uid); inode->i_rdev = rdev; xfs_set_projid(ip, prid); if (pip && XFS_INHERIT_GID(pip)) { + inode->i_gid = VFS_I(pip)->i_gid; ip->i_d.di_gid = pip->i_d.di_gid; if ((VFS_I(pip)->i_mode & S_ISGID) && S_ISDIR(mode)) inode->i_mode |= S_ISGID; + } else { + inode->i_gid = current_fsgid(); + ip->i_d.di_gid = xfs_kgid_to_gid(inode->i_gid); } /* diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c index 0ac63cafb32a1..2c9d2d2f92983 100644 --- a/fs/xfs/xfs_iops.c +++ b/fs/xfs/xfs_iops.c @@ -1305,9 +1305,6 @@ xfs_setup_inode( /* make the inode look hashed for the writeback code */ inode_fake_hash(inode); - inode->i_uid = xfs_uid_to_kuid(ip->i_d.di_uid); - inode->i_gid = xfs_gid_to_kgid(ip->i_d.di_gid); - i_size_write(inode, ip->i_d.di_size); xfs_diflags_to_iflags(inode, ip); -- 2.25.1

1 3

[PATCH openEuler-1.0-LTS] configfs: fix a use-after-free in __configfs_open_file
by Yang Yingliang 14 Dec '21

14 Dec '21

From: Daiyue Zhang <zhangdaiyue1(a)huawei.com> mainline inclusion from mainline-v5.12-rc3 commit 14fbbc8297728e880070f7b077b3301a8c698ef9 category: bugfix bugzilla: NA CVE: CVE-2021-39656 ----------------------------------------------- Commit b0841eefd969 ("configfs: provide exclusion between IO and removals") uses ->frag_dead to mark the fragment state, thus no bothering with extra refcount on config_item when opening a file. The configfs_get_config_item was removed in __configfs_open_file, but not with config_item_put. So the refcount on config_item will lost its balance, causing use-after-free issues in some occasions like this: Test: 1. Mount configfs on /config with read-only items: drwxrwx--- 289 root root 0 2021-04-01 11:55 /config drwxr-xr-x 2 root root 0 2021-04-01 11:54 /config/a --w--w--w- 1 root root 4096 2021-04-01 11:53 /config/a/1.txt ...... 2. Then run: for file in /config do echo $file grep -R 'key' $file done 3. __configfs_open_file will be called in parallel, the first one got called will do: if (file->f_mode & FMODE_READ) { if (!(inode->i_mode & S_IRUGO)) goto out_put_module; config_item_put(buffer->item); kref_put() package_details_release() kfree() the other one will run into use-after-free issues like this: BUG: KASAN: use-after-free in __configfs_open_file+0x1bc/0x3b0 Read of size 8 at addr fffffff155f02480 by task grep/13096 CPU: 0 PID: 13096 Comm: grep VIP: 00 Tainted: G W 4.14.116-kasan #1 TGID: 13096 Comm: grep Call trace: dump_stack+0x118/0x160 kasan_report+0x22c/0x294 __asan_load8+0x80/0x88 __configfs_open_file+0x1bc/0x3b0 configfs_open_file+0x28/0x34 do_dentry_open+0x2cc/0x5c0 vfs_open+0x80/0xe0 path_openat+0xd8c/0x2988 do_filp_open+0x1c4/0x2fc do_sys_open+0x23c/0x404 SyS_openat+0x38/0x48 Allocated by task 2138: kasan_kmalloc+0xe0/0x1ac kmem_cache_alloc_trace+0x334/0x394 packages_make_item+0x4c/0x180 configfs_mkdir+0x358/0x740 vfs_mkdir2+0x1bc/0x2e8 SyS_mkdirat+0x154/0x23c el0_svc_naked+0x34/0x38 Freed by task 13096: kasan_slab_free+0xb8/0x194 kfree+0x13c/0x910 package_details_release+0x524/0x56c kref_put+0xc4/0x104 config_item_put+0x24/0x34 __configfs_open_file+0x35c/0x3b0 configfs_open_file+0x28/0x34 do_dentry_open+0x2cc/0x5c0 vfs_open+0x80/0xe0 path_openat+0xd8c/0x2988 do_filp_open+0x1c4/0x2fc do_sys_open+0x23c/0x404 SyS_openat+0x38/0x48 el0_svc_naked+0x34/0x38 To fix this issue, remove the config_item_put in __configfs_open_file to balance the refcount of config_item. Fixes: b0841eefd969 ("configfs: provide exclusion between IO and removals") Signed-off-by: Daiyue Zhang <zhangdaiyue1(a)huawei.com> Signed-off-by: Yi Chen <chenyi77(a)huawei.com> Signed-off-by: Ge Qiu <qiuge(a)huawei.com> Reviewed-by: Chao Yu <yuchao0(a)huawei.com> Acked-by: Al Viro <viro(a)zeniv.linux.org.uk> Signed-off-by: Christoph Hellwig <hch(a)lst.de> Signed-off-by: Ye Bin <yebin10(a)huawei.com> Reviewed-by: Zhang Yi <yi.zhang(a)huawei.com> Reviewed-by: weiyang wang <wangweiyang2(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- fs/configfs/file.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/fs/configfs/file.c b/fs/configfs/file.c index bb0a427517e92..50b7c4c4310e0 100644 --- a/fs/configfs/file.c +++ b/fs/configfs/file.c @@ -392,7 +392,7 @@ static int __configfs_open_file(struct inode *inode, struct file *file, int type attr = to_attr(dentry); if (!attr) - goto out_put_item; + goto out_free_buffer; if (type & CONFIGFS_ITEM_BIN_ATTR) { buffer->bin_attr = to_bin_attr(dentry); @@ -405,7 +405,7 @@ static int __configfs_open_file(struct inode *inode, struct file *file, int type /* Grab the module reference for this attribute if we have one */ error = -ENODEV; if (!try_module_get(buffer->owner)) - goto out_put_item; + goto out_free_buffer; error = -EACCES; if (!buffer->item->ci_type) @@ -449,8 +449,6 @@ static int __configfs_open_file(struct inode *inode, struct file *file, int type out_put_module: module_put(buffer->owner); -out_put_item: - config_item_put(buffer->item); out_free_buffer: up_read(&frag->frag_sem); kfree(buffer); -- 2.25.1

1 0

[PATCH] vfio/iommu_type1: replace kfree with kvfree
by Jiacheng Shi 13 Dec '21

13 Dec '21

Variables allocated by kvmalloc should not be freed by kfree. Because they may be allocated by vmalloc. So we replace kfree with kvfree here. Fixes: 738fe155f58b ("vfio/iommu_type1: Add support for manual dirty log clear") Signed-off-by: Jiacheng Shi <billsjc(a)sjtu.edu.cn> --- drivers/vfio/vfio_iommu_type1.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/vfio/vfio_iommu_type1.c b/drivers/vfio/vfio_iommu_type1.c index 5daceec48811..6811a85109aa 100644 --- a/drivers/vfio/vfio_iommu_type1.c +++ b/drivers/vfio/vfio_iommu_type1.c @@ -1150,7 +1150,7 @@ static int vfio_iova_dirty_log_clear(u64 __user *bitmap, } out: - kfree(bitmap_buffer); + kvfree(bitmap_buffer); return ret; } -- 2.17.1

3 4

[PATCH openEuler-1.0-LTS 1/2] share_pool: Remove the redundant warning message
by Yang Yingliang 13 Dec '21

13 Dec '21

From: Ding Tianhong <dingtianhong(a)huawei.com> ascend inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I4M253 CVE: NA ------------------------------------------------- No need to record the redundant message for out of memory, the kernel already handle it when oom. Signed-off-by: Ding Tianhong <dingtianhong(a)huawei.com> Reviewed-by: Weilong Chen <chenweilong(a)huawei.com> Reviewed-by: Hanjun Guo <guohanjun(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- mm/share_pool.c | 20 +++++--------------- 1 file changed, 5 insertions(+), 15 deletions(-) diff --git a/mm/share_pool.c b/mm/share_pool.c index 2cc15f894cf84..ea61f69cdd6ad 100644 --- a/mm/share_pool.c +++ b/mm/share_pool.c @@ -137,10 +137,8 @@ static struct sp_group_master *sp_init_group_master_locked( } master = kmalloc(sizeof(struct sp_group_master), GFP_KERNEL); - if (master == NULL) { - pr_err_ratelimited("no memory for spg master\n"); + if (master == NULL) return ERR_PTR(-ENOMEM); - } INIT_LIST_HEAD(&master->node_list); master->count = 0; @@ -192,10 +190,8 @@ static struct sp_proc_stat *create_proc_stat(struct mm_struct *mm, struct sp_proc_stat *stat; stat = kmalloc(sizeof(*stat), GFP_KERNEL); - if (stat == NULL) { - pr_err_ratelimited("no memory for proc stat\n"); + if (stat == NULL) return ERR_PTR(-ENOMEM); - } atomic_set(&stat->use_count, 1); atomic64_set(&stat->alloc_size, 0); @@ -342,10 +338,8 @@ static struct spg_proc_stat *create_spg_proc_stat(int tgid, int spg_id) struct spg_proc_stat *stat; stat = kmalloc(sizeof(struct spg_proc_stat), GFP_KERNEL); - if (stat == NULL) { - pr_err_ratelimited("no memory for spg proc stat\n"); + if (stat == NULL) return ERR_PTR(-ENOMEM); - } stat->tgid = tgid; stat->spg_id = spg_id; @@ -413,10 +407,8 @@ static struct sp_spg_stat *create_spg_stat(int spg_id) struct sp_spg_stat *stat; stat = kmalloc(sizeof(*stat), GFP_KERNEL); - if (stat == NULL) { - pr_err_ratelimited("no memory for spg stat\n"); + if (stat == NULL) return ERR_PTR(-ENOMEM); - } stat->spg_id = spg_id; atomic_set(&stat->hugepage_failures, 0); @@ -1624,10 +1616,8 @@ static struct sp_area *sp_alloc_area(unsigned long size, unsigned long flags, } spa = __kmalloc_node(sizeof(struct sp_area), GFP_KERNEL, node_id); - if (unlikely(!spa)) { - pr_err_ratelimited("no memory for spa\n"); + if (unlikely(!spa)) return ERR_PTR(-ENOMEM); - } spin_lock(&sp_area_lock); -- 2.25.1

1 1

[PATCH OLK-5.10 000/107] NTFS read-write driver GPL implementation by Paragon Software
by Yin Xiujiang 13 Dec '21

13 Dec '21

This patch adds NTFS Read-Write driver to fs/ntfs3. Having decades of expertise in commercial file systems development and huge test coverage, we at Paragon Software GmbH want to make our contribution to the Open Source Community by providing implementation of NTFS Read-Write driver for the Linux Kernel. This is fully functional NTFS Read-Write driver. Current version works with NTFS(including v3.1) and normal/compressed/sparse files and supports journal replaying. We plan to support this version after the codebase once merged, and add new features and fix bugs. For example, full journaling support over JBD will be added in later updates. v2: - patch splitted to chunks (file-wise) - build issues fixed - sparse and checkpatch.pl errors fixed - NULL pointer dereference on mkfs.ntfs-formatted volume mount fixed - cosmetics + code cleanup v3: - added acl, noatime, no_acs_rules, prealloc mount options - added fiemap support - fixed encodings support - removed typedefs - adapted Kernel-way logging mechanisms - fixed typos and corner-case issues v4: - atomic_open() refactored - code style updated - bugfixes v5: - nls/nls_alt mount options added - Unicode conversion fixes - Improved very fragmented files operations - logging cosmetics v6: - Security Descriptors processing changed added system.ntfs_security xattr to set SD - atomic_open() optimized - cosmetics Christophe JAILLET (2): fs/ntfs3: Remove a useless test in 'indx_find()' fs/ntfs3: Remove a useless shadowing variable Colin Ian King (4): fs/ntfs3: Fix various spelling mistakes fs/ntfs3: Fix integer overflow in multiplication fs/ntfs3: Remove redundant initialization of variable err fs/ntfs3: Fix a memory leak on object opts Dan Carpenter (5): fs/ntfs3: add checks for allocation failure fs/ntfs3: fix an error code in ntfs_get_acl_ex() fs/ntfs3: Fix error code in indx_add_allocate() fs/ntfs3: Potential NULL dereference in hdr_find_split() fs/ntfs3: Fix error handling in indx_insert_into_root() Gustavo A. R. Silva (1): fs/ntfs3: Fix fall-through warnings for Clang Jiapeng Chong (1): fs/ntfs3: Remove unused including <linux/version.h> Kari Argillander (54): fs/ntfs3: Use linux/log2 is_power_of_2 function fs/ntfs3: Add ifndef + define to all header files fs/ntfs3: Fix one none utf8 char in source file fs/ntfs3: Restyle comment block in ni_parse_reparse() fs/ntfs3: Use kernel ALIGN macros over driver specific fs/ntfs3: Do not use driver own alloc wrappers fs/ntfs3: Use kcalloc/kmalloc_array over kzalloc/kmalloc fs/ntfs3: Restyle comments to better align with kernel-doc fs/ntfs3: Remove fat ioctl's from ntfs3 driver for now fs/ntfs3: Fix integer overflow in ni_fiemap with fiemap_prep() fs/ntfs3: Remove unnecessary condition checking from ntfs_file_read_iter fs/ntfs3: Remove GPL boilerplates from decompress lib files fs/ntfs3: Change how module init/info messages are displayed fs/ntfs3: Remove unnecesarry mount option noatime fs/ntfs3: Remove unnecesarry remount flag handling fs/ntfs3: Convert mount options to pointer in sbi fs/ntfs3: Use new api for mounting fs/ntfs3: Init spi more in init_fs_context than fill_super fs/ntfs3: Make mount option nohidden more universal fs/ntfs3: Add iocharset= mount option as alias for nls= fs/ntfs3: Rename mount option no_acs_rules > (no)acsrules fs/ntfs3: Show uid/gid always in show_options() fs/ntfs3. Add forward declarations for structs to debug.h fs/ntfs3: Add missing header files to ntfs.h fs/ntfs3: Add missing headers and forward declarations to ntfs_fs.h fs/ntfs3: Add missing header and guards to lib/ headers fs/ntfs3: Change right headers to bitfunc.c fs/ntfs3: Change right headers to upcase.c fs/ntfs3: Change right headers to lznt.c fs/ntfs3: Remove unneeded header files from c files fs/ntfs3: Limit binary search table size fs/ntfs3: Make binary search to search smaller chunks in beginning fs/ntfs3: Always use binary search with entry search fs/ntfs3: Remove '+' before constant in ni_insert_resident() fs/ntfs3: Place Comparisons constant right side of the test fs/ntfs3: Remove braces from single statment block fs/ntfs3: Remove tabs before spaces from comment fs/ntfs3: Fix ntfs_look_for_free_space() does only report -ENOSPC fs/ntfs3: Remove always false condition check fs/ntfs3: Use clamp/max macros instead of comparisons fs/ntfs3: Use min/max macros instated of ternary operators fs/ntfs3: Fix wrong error message $Logfile -> $UpCase fs/ntfs3: Change EINVAL to ENOMEM when d_make_root fails fs/ntfs3: Remove impossible fault condition in fill_super fs/ntfs3: Return straight without goto in fill_super fs/ntfs3: Remove unnecessary variable loading in fill_super fs/ntfs3: Use sb instead of sbi->sb in fill_super fs/ntfs3: Remove tmp var is_ro in ntfs_fill_super fs/ntfs3: Remove tmp pointer bd_inode in fill_super fs/ntfs3: Remove tmp pointer upcase in fill_super fs/ntfs3: Initialize pointer before use place in fill_super fs/ntfs3: Initiliaze sb blocksize only in one place + refactor Doc/fs/ntfs3: Fix rst format and make it cleaner fs/ntfs3: Remove deprecated mount options nls Konstantin Komarov (37): fs/ntfs3: Add headers and misc files fs/ntfs3: Add initialization of super block fs/ntfs3: Add bitmap fs/ntfs3: Add file operations and implementation fs/ntfs3: Add attrib operations fs/ntfs3: Add compression fs/ntfs3: Add NTFS journal fs/ntfs3: Add Kconfig, Makefile and doc fs/ntfs3: Add NTFS3 in fs/Kconfig and fs/Makefile fs/ntfs3: Rework file operations fs/ntfs3: Restyle comments to better align with kernel-doc fs/ntfs3: Fix insertion of attr in ni_ins_attr_ext fs/ntfs3: Change max hardlinks limit to 4000 fs/ntfs3: Add sync flag to ntfs_sb_write_run and al_update fs/ntfs3: Fix logical error in ntfs_create_inode fs/ntfs3: Move ni_lock_dir and ni_unlock into ntfs_create_inode fs/ntfs3: Refactor ntfs_get_acl_ex for better readability fs/ntfs3: Pass flags to ntfs_set_ea in ntfs_set_acl_ex fs/ntfs3: Change posix_acl_equiv_mode to posix_acl_update_mode fs/ntfs3: Refactoring lock in ntfs_init_acl fs/ntfs3: Reject mount if boot's cluster size < media sector size fs/ntfs3: Refactoring of ntfs_init_from_boot fs/ntfs3: Check for NULL if ATTR_EA_INFO is incorrect fs/ntfs3: Use available posix_acl_release instead of ntfs_posix_acl_release fs/ntfs3: Remove locked argument in ntfs_set_ea fs/ntfs3: Refactoring of ntfs_set_ea fs/ntfs3: Forbid FALLOC_FL_PUNCH_HOLE for normal files fs/ntfs3: Remove unnecessary functions fs/ntfs3: Keep prealloc for all types of files fs/ntfs3: Fix memory leak if fill_super failed fs/ntfs3: Rework ntfs_utf16_to_nls fs/ntfs3: Refactor ntfs_readlink_hlp fs/ntfs3: Refactor ntfs_create_inode fs/ntfs3: Refactor ni_parse_reparse fs/ntfs3: Refactor ntfs_read_mft fs/ntfs3: Check for NULL pointers in ni_try_remove_attr_list fs/ntfs3: Add MAINTAINERS Nathan Chancellor (1): fs/ntfs3: Remove unused variable cnt in ntfs_security_init() Yin Xiujiang (2): fs/ntfs3: Fix the issue from backport 5.15 to 5.10 fs/ntfs3: Add ntfs3 module in openeuler_defconfig Documentation/filesystems/index.rst | 1 + Documentation/filesystems/ntfs3.rst | 115 + MAINTAINERS | 9 + arch/arm64/configs/openeuler_defconfig | 4 + arch/x86/configs/openeuler_defconfig | 4 + fs/Kconfig | 1 + fs/Makefile | 1 + fs/ntfs3/Kconfig | 46 + fs/ntfs3/Makefile | 36 + fs/ntfs3/attrib.c | 2083 ++++++++++ fs/ntfs3/attrlist.c | 457 +++ fs/ntfs3/bitfunc.c | 128 + fs/ntfs3/bitmap.c | 1491 +++++++ fs/ntfs3/debug.h | 55 + fs/ntfs3/dir.c | 593 +++ fs/ntfs3/file.c | 1254 ++++++ fs/ntfs3/frecord.c | 3284 +++++++++++++++ fs/ntfs3/fslog.c | 5213 ++++++++++++++++++++++++ fs/ntfs3/fsntfs.c | 2506 ++++++++++++ fs/ntfs3/index.c | 2584 ++++++++++++ fs/ntfs3/inode.c | 1960 +++++++++ fs/ntfs3/lib/decompress_common.c | 319 ++ fs/ntfs3/lib/decompress_common.h | 343 ++ fs/ntfs3/lib/lib.h | 32 + fs/ntfs3/lib/lzx_decompress.c | 670 +++ fs/ntfs3/lib/xpress_decompress.c | 142 + fs/ntfs3/lznt.c | 453 ++ fs/ntfs3/namei.c | 387 ++ fs/ntfs3/ntfs.h | 1224 ++++++ fs/ntfs3/ntfs_fs.h | 1138 ++++++ fs/ntfs3/record.c | 602 +++ fs/ntfs3/run.c | 1111 +++++ fs/ntfs3/super.c | 1507 +++++++ fs/ntfs3/upcase.c | 104 + fs/ntfs3/xattr.c | 991 +++++ 35 files changed, 30848 insertions(+) create mode 100644 Documentation/filesystems/ntfs3.rst create mode 100644 fs/ntfs3/Kconfig create mode 100644 fs/ntfs3/Makefile create mode 100644 fs/ntfs3/attrib.c create mode 100644 fs/ntfs3/attrlist.c create mode 100644 fs/ntfs3/bitfunc.c create mode 100644 fs/ntfs3/bitmap.c create mode 100644 fs/ntfs3/debug.h create mode 100644 fs/ntfs3/dir.c create mode 100644 fs/ntfs3/file.c create mode 100644 fs/ntfs3/frecord.c create mode 100644 fs/ntfs3/fslog.c create mode 100644 fs/ntfs3/fsntfs.c create mode 100644 fs/ntfs3/index.c create mode 100644 fs/ntfs3/inode.c create mode 100644 fs/ntfs3/lib/decompress_common.c create mode 100644 fs/ntfs3/lib/decompress_common.h create mode 100644 fs/ntfs3/lib/lib.h create mode 100644 fs/ntfs3/lib/lzx_decompress.c create mode 100644 fs/ntfs3/lib/xpress_decompress.c create mode 100644 fs/ntfs3/lznt.c create mode 100644 fs/ntfs3/namei.c create mode 100644 fs/ntfs3/ntfs.h create mode 100644 fs/ntfs3/ntfs_fs.h create mode 100644 fs/ntfs3/record.c create mode 100644 fs/ntfs3/run.c create mode 100644 fs/ntfs3/super.c create mode 100644 fs/ntfs3/upcase.c create mode 100644 fs/ntfs3/xattr.c -- 2.30.0

1 34

[PATCH OLK-5.10 001/107] fs/ntfs3: Add headers and misc files
by Yin Xiujiang 13 Dec '21

13 Dec '21

From: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> mainline inclusion from mainline-v5.15 commit 4534a70b7056fd4b9a1c6db5a4ce3c98546b291e category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- This adds headers and misc files Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/debug.h | 64 +++ fs/ntfs3/ntfs.h | 1238 ++++++++++++++++++++++++++++++++++++++++++++ fs/ntfs3/ntfs_fs.h | 1092 ++++++++++++++++++++++++++++++++++++++ fs/ntfs3/upcase.c | 105 ++++ 4 files changed, 2499 insertions(+) create mode 100644 fs/ntfs3/debug.h create mode 100644 fs/ntfs3/ntfs.h create mode 100644 fs/ntfs3/ntfs_fs.h create mode 100644 fs/ntfs3/upcase.c diff --git a/fs/ntfs3/debug.h b/fs/ntfs3/debug.h new file mode 100644 index 000000000000..dfaa4c79dc6d --- /dev/null +++ b/fs/ntfs3/debug.h @@ -0,0 +1,64 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * + * Copyright (C) 2019-2021 Paragon Software GmbH, All rights reserved. + * + * useful functions for debuging + */ + +// clang-format off +#ifndef Add2Ptr +#define Add2Ptr(P, I) ((void *)((u8 *)(P) + (I))) +#define PtrOffset(B, O) ((size_t)((size_t)(O) - (size_t)(B))) +#endif + +#define QuadAlign(n) (((n) + 7u) & (~7u)) +#define IsQuadAligned(n) (!((size_t)(n)&7u)) +#define Quad2Align(n) (((n) + 15u) & (~15u)) +#define IsQuad2Aligned(n) (!((size_t)(n)&15u)) +#define Quad4Align(n) (((n) + 31u) & (~31u)) +#define IsSizeTAligned(n) (!((size_t)(n) & (sizeof(size_t) - 1))) +#define DwordAlign(n) (((n) + 3u) & (~3u)) +#define IsDwordAligned(n) (!((size_t)(n)&3u)) +#define WordAlign(n) (((n) + 1u) & (~1u)) +#define IsWordAligned(n) (!((size_t)(n)&1u)) + +#ifdef CONFIG_PRINTK +__printf(2, 3) +void ntfs_printk(const struct super_block *sb, const char *fmt, ...); +__printf(2, 3) +void ntfs_inode_printk(struct inode *inode, const char *fmt, ...); +#else +static inline __printf(2, 3) +void ntfs_printk(const struct super_block *sb, const char *fmt, ...) +{ +} + +static inline __printf(2, 3) +void ntfs_inode_printk(struct inode *inode, const char *fmt, ...) +{ +} +#endif + +/* + * Logging macros ( thanks Joe Perches <joe(a)perches.com> for implementation ) + */ + +#define ntfs_err(sb, fmt, ...) ntfs_printk(sb, KERN_ERR fmt, ##__VA_ARGS__) +#define ntfs_warn(sb, fmt, ...) ntfs_printk(sb, KERN_WARNING fmt, ##__VA_ARGS__) +#define ntfs_info(sb, fmt, ...) ntfs_printk(sb, KERN_INFO fmt, ##__VA_ARGS__) +#define ntfs_notice(sb, fmt, ...) \ + ntfs_printk(sb, KERN_NOTICE fmt, ##__VA_ARGS__) + +#define ntfs_inode_err(inode, fmt, ...) \ + ntfs_inode_printk(inode, KERN_ERR fmt, ##__VA_ARGS__) +#define ntfs_inode_warn(inode, fmt, ...) \ + ntfs_inode_printk(inode, KERN_WARNING fmt, ##__VA_ARGS__) + +#define ntfs_malloc(s) kmalloc(s, GFP_NOFS) +#define ntfs_zalloc(s) kzalloc(s, GFP_NOFS) +#define ntfs_vmalloc(s) kvmalloc(s, GFP_KERNEL) +#define ntfs_free(p) kfree(p) +#define ntfs_vfree(p) kvfree(p) +#define ntfs_memdup(src, len) kmemdup(src, len, GFP_NOFS) +// clang-format on diff --git a/fs/ntfs3/ntfs.h b/fs/ntfs3/ntfs.h new file mode 100644 index 000000000000..40398e6c39c9 --- /dev/null +++ b/fs/ntfs3/ntfs.h @@ -0,0 +1,1238 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * + * Copyright (C) 2019-2021 Paragon Software GmbH, All rights reserved. + * + * on-disk ntfs structs + */ + +// clang-format off + +/* TODO: + * - Check 4K mft record and 512 bytes cluster + */ + +/* + * Activate this define to use binary search in indexes + */ +#define NTFS3_INDEX_BINARY_SEARCH + +/* + * Check each run for marked clusters + */ +#define NTFS3_CHECK_FREE_CLST + +#define NTFS_NAME_LEN 255 + +/* + * ntfs.sys used 500 maximum links + * on-disk struct allows up to 0xffff + */ +#define NTFS_LINK_MAX 0x400 +//#define NTFS_LINK_MAX 0xffff + +/* + * Activate to use 64 bit clusters instead of 32 bits in ntfs.sys + * Logical and virtual cluster number + * If needed, may be redefined to use 64 bit value + */ +//#define CONFIG_NTFS3_64BIT_CLUSTER + +#define NTFS_LZNT_MAX_CLUSTER 4096 +#define NTFS_LZNT_CUNIT 4 +#define NTFS_LZNT_CLUSTERS (1u<<NTFS_LZNT_CUNIT) + +struct GUID { + __le32 Data1; + __le16 Data2; + __le16 Data3; + u8 Data4[8]; +}; + +/* + * this struct repeats layout of ATTR_FILE_NAME + * at offset 0x40 + * it used to store global constants NAME_MFT/NAME_MIRROR... + * most constant names are shorter than 10 + */ +struct cpu_str { + u8 len; + u8 unused; + u16 name[10]; +}; + +struct le_str { + u8 len; + u8 unused; + __le16 name[]; +}; + +static_assert(SECTOR_SHIFT == 9); + +#ifdef CONFIG_NTFS3_64BIT_CLUSTER +typedef u64 CLST; +static_assert(sizeof(size_t) == 8); +#else +typedef u32 CLST; +#endif + +#define SPARSE_LCN64 ((u64)-1) +#define SPARSE_LCN ((CLST)-1) +#define RESIDENT_LCN ((CLST)-2) +#define COMPRESSED_LCN ((CLST)-3) + +#define COMPRESSION_UNIT 4 +#define COMPRESS_MAX_CLUSTER 0x1000 +#define MFT_INCREASE_CHUNK 1024 + +enum RECORD_NUM { + MFT_REC_MFT = 0, + MFT_REC_MIRR = 1, + MFT_REC_LOG = 2, + MFT_REC_VOL = 3, + MFT_REC_ATTR = 4, + MFT_REC_ROOT = 5, + MFT_REC_BITMAP = 6, + MFT_REC_BOOT = 7, + MFT_REC_BADCLUST = 8, + //MFT_REC_QUOTA = 9, + MFT_REC_SECURE = 9, // NTFS 3.0 + MFT_REC_UPCASE = 10, + MFT_REC_EXTEND = 11, // NTFS 3.0 + MFT_REC_RESERVED = 11, + MFT_REC_FREE = 16, + MFT_REC_USER = 24, +}; + +enum ATTR_TYPE { + ATTR_ZERO = cpu_to_le32(0x00), + ATTR_STD = cpu_to_le32(0x10), + ATTR_LIST = cpu_to_le32(0x20), + ATTR_NAME = cpu_to_le32(0x30), + // ATTR_VOLUME_VERSION on Nt4 + ATTR_ID = cpu_to_le32(0x40), + ATTR_SECURE = cpu_to_le32(0x50), + ATTR_LABEL = cpu_to_le32(0x60), + ATTR_VOL_INFO = cpu_to_le32(0x70), + ATTR_DATA = cpu_to_le32(0x80), + ATTR_ROOT = cpu_to_le32(0x90), + ATTR_ALLOC = cpu_to_le32(0xA0), + ATTR_BITMAP = cpu_to_le32(0xB0), + // ATTR_SYMLINK on Nt4 + ATTR_REPARSE = cpu_to_le32(0xC0), + ATTR_EA_INFO = cpu_to_le32(0xD0), + ATTR_EA = cpu_to_le32(0xE0), + ATTR_PROPERTYSET = cpu_to_le32(0xF0), + ATTR_LOGGED_UTILITY_STREAM = cpu_to_le32(0x100), + ATTR_END = cpu_to_le32(0xFFFFFFFF) +}; + +static_assert(sizeof(enum ATTR_TYPE) == 4); + +enum FILE_ATTRIBUTE { + FILE_ATTRIBUTE_READONLY = cpu_to_le32(0x00000001), + FILE_ATTRIBUTE_HIDDEN = cpu_to_le32(0x00000002), + FILE_ATTRIBUTE_SYSTEM = cpu_to_le32(0x00000004), + FILE_ATTRIBUTE_ARCHIVE = cpu_to_le32(0x00000020), + FILE_ATTRIBUTE_DEVICE = cpu_to_le32(0x00000040), + FILE_ATTRIBUTE_TEMPORARY = cpu_to_le32(0x00000100), + FILE_ATTRIBUTE_SPARSE_FILE = cpu_to_le32(0x00000200), + FILE_ATTRIBUTE_REPARSE_POINT = cpu_to_le32(0x00000400), + FILE_ATTRIBUTE_COMPRESSED = cpu_to_le32(0x00000800), + FILE_ATTRIBUTE_OFFLINE = cpu_to_le32(0x00001000), + FILE_ATTRIBUTE_NOT_CONTENT_INDEXED = cpu_to_le32(0x00002000), + FILE_ATTRIBUTE_ENCRYPTED = cpu_to_le32(0x00004000), + FILE_ATTRIBUTE_VALID_FLAGS = cpu_to_le32(0x00007fb7), + FILE_ATTRIBUTE_DIRECTORY = cpu_to_le32(0x10000000), +}; + +static_assert(sizeof(enum FILE_ATTRIBUTE) == 4); + +extern const struct cpu_str NAME_MFT; +extern const struct cpu_str NAME_MIRROR; +extern const struct cpu_str NAME_LOGFILE; +extern const struct cpu_str NAME_VOLUME; +extern const struct cpu_str NAME_ATTRDEF; +extern const struct cpu_str NAME_ROOT; +extern const struct cpu_str NAME_BITMAP; +extern const struct cpu_str NAME_BOOT; +extern const struct cpu_str NAME_BADCLUS; +extern const struct cpu_str NAME_QUOTA; +extern const struct cpu_str NAME_SECURE; +extern const struct cpu_str NAME_UPCASE; +extern const struct cpu_str NAME_EXTEND; +extern const struct cpu_str NAME_OBJID; +extern const struct cpu_str NAME_REPARSE; +extern const struct cpu_str NAME_USNJRNL; + +extern const __le16 I30_NAME[4]; +extern const __le16 SII_NAME[4]; +extern const __le16 SDH_NAME[4]; +extern const __le16 SO_NAME[2]; +extern const __le16 SQ_NAME[2]; +extern const __le16 SR_NAME[2]; + +extern const __le16 BAD_NAME[4]; +extern const __le16 SDS_NAME[4]; +extern const __le16 WOF_NAME[17]; /* WofCompressedData */ + +/* MFT record number structure */ +struct MFT_REF { + __le32 low; // The low part of the number + __le16 high; // The high part of the number + __le16 seq; // The sequence number of MFT record +}; + +static_assert(sizeof(__le64) == sizeof(struct MFT_REF)); + +static inline CLST ino_get(const struct MFT_REF *ref) +{ +#ifdef CONFIG_NTFS3_64BIT_CLUSTER + return le32_to_cpu(ref->low) | ((u64)le16_to_cpu(ref->high) << 32); +#else + return le32_to_cpu(ref->low); +#endif +} + +struct NTFS_BOOT { + u8 jump_code[3]; // 0x00: Jump to boot code + u8 system_id[8]; // 0x03: System ID, equals "NTFS " + + // NOTE: this member is not aligned(!) + // bytes_per_sector[0] must be 0 + // bytes_per_sector[1] must be multiplied by 256 + u8 bytes_per_sector[2]; // 0x0B: Bytes per sector + + u8 sectors_per_clusters;// 0x0D: Sectors per cluster + u8 unused1[7]; + u8 media_type; // 0x15: Media type (0xF8 - harddisk) + u8 unused2[2]; + __le16 sct_per_track; // 0x18: number of sectors per track + __le16 heads; // 0x1A: number of heads per cylinder + __le32 hidden_sectors; // 0x1C: number of 'hidden' sectors + u8 unused3[4]; + u8 bios_drive_num; // 0x24: BIOS drive number =0x80 + u8 unused4; + u8 signature_ex; // 0x26: Extended BOOT signature =0x80 + u8 unused5; + __le64 sectors_per_volume;// 0x28: size of volume in sectors + __le64 mft_clst; // 0x30: first cluster of $MFT + __le64 mft2_clst; // 0x38: first cluster of $MFTMirr + s8 record_size; // 0x40: size of MFT record in clusters(sectors) + u8 unused6[3]; + s8 index_size; // 0x44: size of INDX record in clusters(sectors) + u8 unused7[3]; + __le64 serial_num; // 0x48: Volume serial number + __le32 check_sum; // 0x50: Simple additive checksum of all + // of the u32's which precede the 'check_sum' + + u8 boot_code[0x200 - 0x50 - 2 - 4]; // 0x54: + u8 boot_magic[2]; // 0x1FE: Boot signature =0x55 + 0xAA +}; + +static_assert(sizeof(struct NTFS_BOOT) == 0x200); + +enum NTFS_SIGNATURE { + NTFS_FILE_SIGNATURE = cpu_to_le32(0x454C4946), // 'FILE' + NTFS_INDX_SIGNATURE = cpu_to_le32(0x58444E49), // 'INDX' + NTFS_CHKD_SIGNATURE = cpu_to_le32(0x444B4843), // 'CHKD' + NTFS_RSTR_SIGNATURE = cpu_to_le32(0x52545352), // 'RSTR' + NTFS_RCRD_SIGNATURE = cpu_to_le32(0x44524352), // 'RCRD' + NTFS_BAAD_SIGNATURE = cpu_to_le32(0x44414142), // 'BAAD' + NTFS_HOLE_SIGNATURE = cpu_to_le32(0x454C4F48), // 'HOLE' + NTFS_FFFF_SIGNATURE = cpu_to_le32(0xffffffff), +}; + +static_assert(sizeof(enum NTFS_SIGNATURE) == 4); + +/* MFT Record header structure */ +struct NTFS_RECORD_HEADER { + /* Record magic number, equals 'FILE'/'INDX'/'RSTR'/'RCRD' */ + enum NTFS_SIGNATURE sign; // 0x00: + __le16 fix_off; // 0x04: + __le16 fix_num; // 0x06: + __le64 lsn; // 0x08: Log file sequence number +}; + +static_assert(sizeof(struct NTFS_RECORD_HEADER) == 0x10); + +static inline int is_baad(const struct NTFS_RECORD_HEADER *hdr) +{ + return hdr->sign == NTFS_BAAD_SIGNATURE; +} + +/* Possible bits in struct MFT_REC.flags */ +enum RECORD_FLAG { + RECORD_FLAG_IN_USE = cpu_to_le16(0x0001), + RECORD_FLAG_DIR = cpu_to_le16(0x0002), + RECORD_FLAG_SYSTEM = cpu_to_le16(0x0004), + RECORD_FLAG_UNKNOWN = cpu_to_le16(0x0008), +}; + +/* MFT Record structure */ +struct MFT_REC { + struct NTFS_RECORD_HEADER rhdr; // 'FILE' + + __le16 seq; // 0x10: Sequence number for this record + __le16 hard_links; // 0x12: The number of hard links to record + __le16 attr_off; // 0x14: Offset to attributes + __le16 flags; // 0x16: See RECORD_FLAG + __le32 used; // 0x18: The size of used part + __le32 total; // 0x1C: Total record size + + struct MFT_REF parent_ref; // 0x20: Parent MFT record + __le16 next_attr_id; // 0x28: The next attribute Id + + __le16 res; // 0x2A: High part of mft record? + __le32 mft_record; // 0x2C: Current mft record number + __le16 fixups[]; // 0x30: +}; + +#define MFTRECORD_FIXUP_OFFSET_1 offsetof(struct MFT_REC, res) +#define MFTRECORD_FIXUP_OFFSET_3 offsetof(struct MFT_REC, fixups) + +static_assert(MFTRECORD_FIXUP_OFFSET_1 == 0x2A); +static_assert(MFTRECORD_FIXUP_OFFSET_3 == 0x30); + +static inline bool is_rec_base(const struct MFT_REC *rec) +{ + const struct MFT_REF *r = &rec->parent_ref; + + return !r->low && !r->high && !r->seq; +} + +static inline bool is_mft_rec5(const struct MFT_REC *rec) +{ + return le16_to_cpu(rec->rhdr.fix_off) >= + offsetof(struct MFT_REC, fixups); +} + +static inline bool is_rec_inuse(const struct MFT_REC *rec) +{ + return rec->flags & RECORD_FLAG_IN_USE; +} + +static inline bool clear_rec_inuse(struct MFT_REC *rec) +{ + return rec->flags &= ~RECORD_FLAG_IN_USE; +} + +/* Possible values of ATTR_RESIDENT.flags */ +#define RESIDENT_FLAG_INDEXED 0x01 + +struct ATTR_RESIDENT { + __le32 data_size; // 0x10: The size of data + __le16 data_off; // 0x14: Offset to data + u8 flags; // 0x16: resident flags ( 1 - indexed ) + u8 res; // 0x17: +}; // sizeof() = 0x18 + +struct ATTR_NONRESIDENT { + __le64 svcn; // 0x10: Starting VCN of this segment + __le64 evcn; // 0x18: End VCN of this segment + __le16 run_off; // 0x20: Offset to packed runs + // Unit of Compression size for this stream, expressed + // as a log of the cluster size. + // + // 0 means file is not compressed + // 1, 2, 3, and 4 are potentially legal values if the + // stream is compressed, however the implementation + // may only choose to use 4, or possibly 3. Note + // that 4 means cluster size time 16. If convenient + // the implementation may wish to accept a + // reasonable range of legal values here (1-5?), + // even if the implementation only generates + // a smaller set of values itself. + u8 c_unit; // 0x22 + u8 res1[5]; // 0x23: + __le64 alloc_size; // 0x28: The allocated size of attribute in bytes + // (multiple of cluster size) + __le64 data_size; // 0x30: The size of attribute in bytes <= alloc_size + __le64 valid_size; // 0x38: The size of valid part in bytes <= data_size + __le64 total_size; // 0x40: The sum of the allocated clusters for a file + // (present only for the first segment (0 == vcn) + // of compressed attribute) + +}; // sizeof()=0x40 or 0x48 (if compressed) + +/* Possible values of ATTRIB.flags: */ +#define ATTR_FLAG_COMPRESSED cpu_to_le16(0x0001) +#define ATTR_FLAG_COMPRESSED_MASK cpu_to_le16(0x00FF) +#define ATTR_FLAG_ENCRYPTED cpu_to_le16(0x4000) +#define ATTR_FLAG_SPARSED cpu_to_le16(0x8000) + +struct ATTRIB { + enum ATTR_TYPE type; // 0x00: The type of this attribute + __le32 size; // 0x04: The size of this attribute + u8 non_res; // 0x08: Is this attribute non-resident ? + u8 name_len; // 0x09: This attribute name length + __le16 name_off; // 0x0A: Offset to the attribute name + __le16 flags; // 0x0C: See ATTR_FLAG_XXX + __le16 id; // 0x0E: unique id (per record) + + union { + struct ATTR_RESIDENT res; // 0x10 + struct ATTR_NONRESIDENT nres; // 0x10 + }; +}; + +/* Define attribute sizes */ +#define SIZEOF_RESIDENT 0x18 +#define SIZEOF_NONRESIDENT_EX 0x48 +#define SIZEOF_NONRESIDENT 0x40 + +#define SIZEOF_RESIDENT_LE cpu_to_le16(0x18) +#define SIZEOF_NONRESIDENT_EX_LE cpu_to_le16(0x48) +#define SIZEOF_NONRESIDENT_LE cpu_to_le16(0x40) + +static inline u64 attr_ondisk_size(const struct ATTRIB *attr) +{ + return attr->non_res ? ((attr->flags & + (ATTR_FLAG_COMPRESSED | ATTR_FLAG_SPARSED)) ? + le64_to_cpu(attr->nres.total_size) : + le64_to_cpu(attr->nres.alloc_size)) : + QuadAlign(le32_to_cpu(attr->res.data_size)); +} + +static inline u64 attr_size(const struct ATTRIB *attr) +{ + return attr->non_res ? le64_to_cpu(attr->nres.data_size) : + le32_to_cpu(attr->res.data_size); +} + +static inline bool is_attr_encrypted(const struct ATTRIB *attr) +{ + return attr->flags & ATTR_FLAG_ENCRYPTED; +} + +static inline bool is_attr_sparsed(const struct ATTRIB *attr) +{ + return attr->flags & ATTR_FLAG_SPARSED; +} + +static inline bool is_attr_compressed(const struct ATTRIB *attr) +{ + return attr->flags & ATTR_FLAG_COMPRESSED; +} + +static inline bool is_attr_ext(const struct ATTRIB *attr) +{ + return attr->flags & (ATTR_FLAG_SPARSED | ATTR_FLAG_COMPRESSED); +} + +static inline bool is_attr_indexed(const struct ATTRIB *attr) +{ + return !attr->non_res && (attr->res.flags & RESIDENT_FLAG_INDEXED); +} + +static inline __le16 const *attr_name(const struct ATTRIB *attr) +{ + return Add2Ptr(attr, le16_to_cpu(attr->name_off)); +} + +static inline u64 attr_svcn(const struct ATTRIB *attr) +{ + return attr->non_res ? le64_to_cpu(attr->nres.svcn) : 0; +} + +/* the size of resident attribute by its resident size */ +#define BYTES_PER_RESIDENT(b) (0x18 + (b)) + +static_assert(sizeof(struct ATTRIB) == 0x48); +static_assert(sizeof(((struct ATTRIB *)NULL)->res) == 0x08); +static_assert(sizeof(((struct ATTRIB *)NULL)->nres) == 0x38); + +static inline void *resident_data_ex(const struct ATTRIB *attr, u32 datasize) +{ + u32 asize, rsize; + u16 off; + + if (attr->non_res) + return NULL; + + asize = le32_to_cpu(attr->size); + off = le16_to_cpu(attr->res.data_off); + + if (asize < datasize + off) + return NULL; + + rsize = le32_to_cpu(attr->res.data_size); + if (rsize < datasize) + return NULL; + + return Add2Ptr(attr, off); +} + +static inline void *resident_data(const struct ATTRIB *attr) +{ + return Add2Ptr(attr, le16_to_cpu(attr->res.data_off)); +} + +static inline void *attr_run(const struct ATTRIB *attr) +{ + return Add2Ptr(attr, le16_to_cpu(attr->nres.run_off)); +} + +/* Standard information attribute (0x10) */ +struct ATTR_STD_INFO { + __le64 cr_time; // 0x00: File creation file + __le64 m_time; // 0x08: File modification time + __le64 c_time; // 0x10: Last time any attribute was modified + __le64 a_time; // 0x18: File last access time + enum FILE_ATTRIBUTE fa; // 0x20: Standard DOS attributes & more + __le32 max_ver_num; // 0x24: Maximum Number of Versions + __le32 ver_num; // 0x28: Version Number + __le32 class_id; // 0x2C: Class Id from bidirectional Class Id index +}; + +static_assert(sizeof(struct ATTR_STD_INFO) == 0x30); + +#define SECURITY_ID_INVALID 0x00000000 +#define SECURITY_ID_FIRST 0x00000100 + +struct ATTR_STD_INFO5 { + __le64 cr_time; // 0x00: File creation file + __le64 m_time; // 0x08: File modification time + __le64 c_time; // 0x10: Last time any attribute was modified + __le64 a_time; // 0x18: File last access time + enum FILE_ATTRIBUTE fa; // 0x20: Standard DOS attributes & more + __le32 max_ver_num; // 0x24: Maximum Number of Versions + __le32 ver_num; // 0x28: Version Number + __le32 class_id; // 0x2C: Class Id from bidirectional Class Id index + + __le32 owner_id; // 0x30: Owner Id of the user owning the file. + __le32 security_id; // 0x34: The Security Id is a key in the $SII Index and $SDS + __le64 quota_charge; // 0x38: + __le64 usn; // 0x40: Last Update Sequence Number of the file. This is a direct + // index into the file $UsnJrnl. If zero, the USN Journal is + // disabled. +}; + +static_assert(sizeof(struct ATTR_STD_INFO5) == 0x48); + +/* attribute list entry structure (0x20) */ +struct ATTR_LIST_ENTRY { + enum ATTR_TYPE type; // 0x00: The type of attribute + __le16 size; // 0x04: The size of this record + u8 name_len; // 0x06: The length of attribute name + u8 name_off; // 0x07: The offset to attribute name + __le64 vcn; // 0x08: Starting VCN of this attribute + struct MFT_REF ref; // 0x10: MFT record number with attribute + __le16 id; // 0x18: struct ATTRIB ID + __le16 name[3]; // 0x1A: Just to align. To get real name can use bNameOffset + +}; // sizeof(0x20) + +static_assert(sizeof(struct ATTR_LIST_ENTRY) == 0x20); + +static inline u32 le_size(u8 name_len) +{ + return QuadAlign(offsetof(struct ATTR_LIST_ENTRY, name) + + name_len * sizeof(short)); +} + +/* returns 0 if 'attr' has the same type and name */ +static inline int le_cmp(const struct ATTR_LIST_ENTRY *le, + const struct ATTRIB *attr) +{ + return le->type != attr->type || le->name_len != attr->name_len || + (!le->name_len && + memcmp(Add2Ptr(le, le->name_off), + Add2Ptr(attr, le16_to_cpu(attr->name_off)), + le->name_len * sizeof(short))); +} + +static inline __le16 const *le_name(const struct ATTR_LIST_ENTRY *le) +{ + return Add2Ptr(le, le->name_off); +} + +/* File name types (the field type in struct ATTR_FILE_NAME ) */ +#define FILE_NAME_POSIX 0 +#define FILE_NAME_UNICODE 1 +#define FILE_NAME_DOS 2 +#define FILE_NAME_UNICODE_AND_DOS (FILE_NAME_DOS | FILE_NAME_UNICODE) + +/* Filename attribute structure (0x30) */ +struct NTFS_DUP_INFO { + __le64 cr_time; // 0x00: File creation file + __le64 m_time; // 0x08: File modification time + __le64 c_time; // 0x10: Last time any attribute was modified + __le64 a_time; // 0x18: File last access time + __le64 alloc_size; // 0x20: Data attribute allocated size, multiple of cluster size + __le64 data_size; // 0x28: Data attribute size <= Dataalloc_size + enum FILE_ATTRIBUTE fa; // 0x30: Standard DOS attributes & more + __le16 ea_size; // 0x34: Packed EAs + __le16 reparse; // 0x36: Used by Reparse + +}; // 0x38 + +struct ATTR_FILE_NAME { + struct MFT_REF home; // 0x00: MFT record for directory + struct NTFS_DUP_INFO dup;// 0x08 + u8 name_len; // 0x40: File name length in words + u8 type; // 0x41: File name type + __le16 name[]; // 0x42: File name +}; + +static_assert(sizeof(((struct ATTR_FILE_NAME *)NULL)->dup) == 0x38); +static_assert(offsetof(struct ATTR_FILE_NAME, name) == 0x42); +#define SIZEOF_ATTRIBUTE_FILENAME 0x44 +#define SIZEOF_ATTRIBUTE_FILENAME_MAX (0x42 + 255 * 2) + +static inline struct ATTRIB *attr_from_name(struct ATTR_FILE_NAME *fname) +{ + return (struct ATTRIB *)((char *)fname - SIZEOF_RESIDENT); +} + +static inline u16 fname_full_size(const struct ATTR_FILE_NAME *fname) +{ + // don't return struct_size(fname, name, fname->name_len); + return offsetof(struct ATTR_FILE_NAME, name) + + fname->name_len * sizeof(short); +} + +static inline u8 paired_name(u8 type) +{ + if (type == FILE_NAME_UNICODE) + return FILE_NAME_DOS; + if (type == FILE_NAME_DOS) + return FILE_NAME_UNICODE; + return FILE_NAME_POSIX; +} + +/* Index entry defines ( the field flags in NtfsDirEntry ) */ +#define NTFS_IE_HAS_SUBNODES cpu_to_le16(1) +#define NTFS_IE_LAST cpu_to_le16(2) + +/* Directory entry structure */ +struct NTFS_DE { + union { + struct MFT_REF ref; // 0x00: MFT record number with this file + struct { + __le16 data_off; // 0x00: + __le16 data_size; // 0x02: + __le32 res; // 0x04: must be 0 + } view; + }; + __le16 size; // 0x08: The size of this entry + __le16 key_size; // 0x0A: The size of File name length in bytes + 0x42 + __le16 flags; // 0x0C: Entry flags: NTFS_IE_XXX + __le16 res; // 0x0E: + + // Here any indexed attribute can be placed + // One of them is: + // struct ATTR_FILE_NAME AttrFileName; + // + + // The last 8 bytes of this structure contains + // the VBN of subnode + // !!! Note !!! + // This field is presented only if (flags & NTFS_IE_HAS_SUBNODES) + // __le64 vbn; +}; + +static_assert(sizeof(struct NTFS_DE) == 0x10); + +static inline void de_set_vbn_le(struct NTFS_DE *e, __le64 vcn) +{ + __le64 *v = Add2Ptr(e, le16_to_cpu(e->size) - sizeof(__le64)); + + *v = vcn; +} + +static inline void de_set_vbn(struct NTFS_DE *e, CLST vcn) +{ + __le64 *v = Add2Ptr(e, le16_to_cpu(e->size) - sizeof(__le64)); + + *v = cpu_to_le64(vcn); +} + +static inline __le64 de_get_vbn_le(const struct NTFS_DE *e) +{ + return *(__le64 *)Add2Ptr(e, le16_to_cpu(e->size) - sizeof(__le64)); +} + +static inline CLST de_get_vbn(const struct NTFS_DE *e) +{ + __le64 *v = Add2Ptr(e, le16_to_cpu(e->size) - sizeof(__le64)); + + return le64_to_cpu(*v); +} + +static inline struct NTFS_DE *de_get_next(const struct NTFS_DE *e) +{ + return Add2Ptr(e, le16_to_cpu(e->size)); +} + +static inline struct ATTR_FILE_NAME *de_get_fname(const struct NTFS_DE *e) +{ + return le16_to_cpu(e->key_size) >= SIZEOF_ATTRIBUTE_FILENAME ? + Add2Ptr(e, sizeof(struct NTFS_DE)) : + NULL; +} + +static inline bool de_is_last(const struct NTFS_DE *e) +{ + return e->flags & NTFS_IE_LAST; +} + +static inline bool de_has_vcn(const struct NTFS_DE *e) +{ + return e->flags & NTFS_IE_HAS_SUBNODES; +} + +static inline bool de_has_vcn_ex(const struct NTFS_DE *e) +{ + return (e->flags & NTFS_IE_HAS_SUBNODES) && + (u64)(-1) != *((u64 *)Add2Ptr(e, le16_to_cpu(e->size) - + sizeof(__le64))); +} + +#define MAX_BYTES_PER_NAME_ENTRY \ + QuadAlign(sizeof(struct NTFS_DE) + \ + offsetof(struct ATTR_FILE_NAME, name) + \ + NTFS_NAME_LEN * sizeof(short)) + +struct INDEX_HDR { + __le32 de_off; // 0x00: The offset from the start of this structure + // to the first NTFS_DE + __le32 used; // 0x04: The size of this structure plus all + // entries (quad-word aligned) + __le32 total; // 0x08: The allocated size of for this structure plus all entries + u8 flags; // 0x0C: 0x00 = Small directory, 0x01 = Large directory + u8 res[3]; + + // + // de_off + used <= total + // +}; + +static_assert(sizeof(struct INDEX_HDR) == 0x10); + +static inline struct NTFS_DE *hdr_first_de(const struct INDEX_HDR *hdr) +{ + u32 de_off = le32_to_cpu(hdr->de_off); + u32 used = le32_to_cpu(hdr->used); + struct NTFS_DE *e = Add2Ptr(hdr, de_off); + u16 esize; + + if (de_off >= used || de_off >= le32_to_cpu(hdr->total)) + return NULL; + + esize = le16_to_cpu(e->size); + if (esize < sizeof(struct NTFS_DE) || de_off + esize > used) + return NULL; + + return e; +} + +static inline struct NTFS_DE *hdr_next_de(const struct INDEX_HDR *hdr, + const struct NTFS_DE *e) +{ + size_t off = PtrOffset(hdr, e); + u32 used = le32_to_cpu(hdr->used); + u16 esize; + + if (off >= used) + return NULL; + + esize = le16_to_cpu(e->size); + + if (esize < sizeof(struct NTFS_DE) || + off + esize + sizeof(struct NTFS_DE) > used) + return NULL; + + return Add2Ptr(e, esize); +} + +static inline bool hdr_has_subnode(const struct INDEX_HDR *hdr) +{ + return hdr->flags & 1; +} + +struct INDEX_BUFFER { + struct NTFS_RECORD_HEADER rhdr; // 'INDX' + __le64 vbn; // 0x10: vcn if index >= cluster or vsn id index < cluster + struct INDEX_HDR ihdr; // 0x18: +}; + +static_assert(sizeof(struct INDEX_BUFFER) == 0x28); + +static inline bool ib_is_empty(const struct INDEX_BUFFER *ib) +{ + const struct NTFS_DE *first = hdr_first_de(&ib->ihdr); + + return !first || de_is_last(first); +} + +static inline bool ib_is_leaf(const struct INDEX_BUFFER *ib) +{ + return !(ib->ihdr.flags & 1); +} + +/* Index root structure ( 0x90 ) */ +enum COLLATION_RULE { + NTFS_COLLATION_TYPE_BINARY = cpu_to_le32(0), + // $I30 + NTFS_COLLATION_TYPE_FILENAME = cpu_to_le32(0x01), + // $SII of $Secure and $Q of Quota + NTFS_COLLATION_TYPE_UINT = cpu_to_le32(0x10), + // $O of Quota + NTFS_COLLATION_TYPE_SID = cpu_to_le32(0x11), + // $SDH of $Secure + NTFS_COLLATION_TYPE_SECURITY_HASH = cpu_to_le32(0x12), + // $O of ObjId and "$R" for Reparse + NTFS_COLLATION_TYPE_UINTS = cpu_to_le32(0x13) +}; + +static_assert(sizeof(enum COLLATION_RULE) == 4); + +// +struct INDEX_ROOT { + enum ATTR_TYPE type; // 0x00: The type of attribute to index on + enum COLLATION_RULE rule; // 0x04: The rule + __le32 index_block_size;// 0x08: The size of index record + u8 index_block_clst; // 0x0C: The number of clusters or sectors per index + u8 res[3]; + struct INDEX_HDR ihdr; // 0x10: +}; + +static_assert(sizeof(struct INDEX_ROOT) == 0x20); +static_assert(offsetof(struct INDEX_ROOT, ihdr) == 0x10); + +#define VOLUME_FLAG_DIRTY cpu_to_le16(0x0001) +#define VOLUME_FLAG_RESIZE_LOG_FILE cpu_to_le16(0x0002) + +struct VOLUME_INFO { + __le64 res1; // 0x00 + u8 major_ver; // 0x08: NTFS major version number (before .) + u8 minor_ver; // 0x09: NTFS minor version number (after .) + __le16 flags; // 0x0A: Volume flags, see VOLUME_FLAG_XXX + +}; // sizeof=0xC + +#define SIZEOF_ATTRIBUTE_VOLUME_INFO 0xc + +#define NTFS_LABEL_MAX_LENGTH (0x100 / sizeof(short)) +#define NTFS_ATTR_INDEXABLE cpu_to_le32(0x00000002) +#define NTFS_ATTR_DUPALLOWED cpu_to_le32(0x00000004) +#define NTFS_ATTR_MUST_BE_INDEXED cpu_to_le32(0x00000010) +#define NTFS_ATTR_MUST_BE_NAMED cpu_to_le32(0x00000020) +#define NTFS_ATTR_MUST_BE_RESIDENT cpu_to_le32(0x00000040) +#define NTFS_ATTR_LOG_ALWAYS cpu_to_le32(0x00000080) + +/* $AttrDef file entry */ +struct ATTR_DEF_ENTRY { + __le16 name[0x40]; // 0x00: Attr name + enum ATTR_TYPE type; // 0x80: struct ATTRIB type + __le32 res; // 0x84: + enum COLLATION_RULE rule; // 0x88: + __le32 flags; // 0x8C: NTFS_ATTR_XXX (see above) + __le64 min_sz; // 0x90: Minimum attribute data size + __le64 max_sz; // 0x98: Maximum attribute data size +}; + +static_assert(sizeof(struct ATTR_DEF_ENTRY) == 0xa0); + +/* Object ID (0x40) */ +struct OBJECT_ID { + struct GUID ObjId; // 0x00: Unique Id assigned to file + struct GUID BirthVolumeId;// 0x10: Birth Volume Id is the Object Id of the Volume on + // which the Object Id was allocated. It never changes + struct GUID BirthObjectId; // 0x20: Birth Object Id is the first Object Id that was + // ever assigned to this MFT Record. I.e. If the Object Id + // is changed for some reason, this field will reflect the + // original value of the Object Id. + struct GUID DomainId; // 0x30: Domain Id is currently unused but it is intended to be + // used in a network environment where the local machine is + // part of a Windows 2000 Domain. This may be used in a Windows + // 2000 Advanced Server managed domain. +}; + +static_assert(sizeof(struct OBJECT_ID) == 0x40); + +/* O Directory entry structure ( rule = 0x13 ) */ +struct NTFS_DE_O { + struct NTFS_DE de; + struct GUID ObjId; // 0x10: Unique Id assigned to file + struct MFT_REF ref; // 0x20: MFT record number with this file + struct GUID BirthVolumeId; // 0x28: Birth Volume Id is the Object Id of the Volume on + // which the Object Id was allocated. It never changes + struct GUID BirthObjectId; // 0x38: Birth Object Id is the first Object Id that was + // ever assigned to this MFT Record. I.e. If the Object Id + // is changed for some reason, this field will reflect the + // original value of the Object Id. + // This field is valid if data_size == 0x48 + struct GUID BirthDomainId; // 0x48: Domain Id is currently unused but it is intended + // to be used in a network environment where the local + // machine is part of a Windows 2000 Domain. This may be + // used in a Windows 2000 Advanced Server managed domain. +}; + +static_assert(sizeof(struct NTFS_DE_O) == 0x58); + +#define NTFS_OBJECT_ENTRY_DATA_SIZE1 \ + 0x38 // struct NTFS_DE_O.BirthDomainId is not used +#define NTFS_OBJECT_ENTRY_DATA_SIZE2 \ + 0x48 // struct NTFS_DE_O.BirthDomainId is used + +/* Q Directory entry structure ( rule = 0x11 ) */ +struct NTFS_DE_Q { + struct NTFS_DE de; + __le32 owner_id; // 0x10: Unique Id assigned to file + __le32 Version; // 0x14: 0x02 + __le32 flags2; // 0x18: Quota flags, see above + __le64 BytesUsed; // 0x1C: + __le64 ChangeTime; // 0x24: + __le64 WarningLimit; // 0x28: + __le64 HardLimit; // 0x34: + __le64 ExceededTime; // 0x3C: + + // SID is placed here +}; // sizeof() = 0x44 + +#define SIZEOF_NTFS_DE_Q 0x44 + +#define SecurityDescriptorsBlockSize 0x40000 // 256K +#define SecurityDescriptorMaxSize 0x20000 // 128K +#define Log2OfSecurityDescriptorsBlockSize 18 + +struct SECURITY_KEY { + __le32 hash; // Hash value for descriptor + __le32 sec_id; // Security Id (guaranteed unique) +}; + +/* Security descriptors (the content of $Secure::SDS data stream) */ +struct SECURITY_HDR { + struct SECURITY_KEY key; // 0x00: Security Key + __le64 off; // 0x08: Offset of this entry in the file + __le32 size; // 0x10: Size of this entry, 8 byte aligned + // + // Security descriptor itself is placed here + // Total size is 16 byte aligned + // +} __packed; + +#define SIZEOF_SECURITY_HDR 0x14 + +/* SII Directory entry structure */ +struct NTFS_DE_SII { + struct NTFS_DE de; + __le32 sec_id; // 0x10: Key: sizeof(security_id) = wKeySize + struct SECURITY_HDR sec_hdr; // 0x14: +} __packed; + +#define SIZEOF_SII_DIRENTRY 0x28 + +/* SDH Directory entry structure */ +struct NTFS_DE_SDH { + struct NTFS_DE de; + struct SECURITY_KEY key; // 0x10: Key + struct SECURITY_HDR sec_hdr; // 0x18: Data + __le16 magic[2]; // 0x2C: 0x00490049 "I I" +}; + +#define SIZEOF_SDH_DIRENTRY 0x30 + +struct REPARSE_KEY { + __le32 ReparseTag; // 0x00: Reparse Tag + struct MFT_REF ref; // 0x04: MFT record number with this file +}; // sizeof() = 0x0C + +static_assert(offsetof(struct REPARSE_KEY, ref) == 0x04); +#define SIZEOF_REPARSE_KEY 0x0C + +/* Reparse Directory entry structure */ +struct NTFS_DE_R { + struct NTFS_DE de; + struct REPARSE_KEY key; // 0x10: Reparse Key + u32 zero; // 0x1c +}; // sizeof() = 0x20 + +static_assert(sizeof(struct NTFS_DE_R) == 0x20); + +/* CompressReparseBuffer.WofVersion */ +#define WOF_CURRENT_VERSION cpu_to_le32(1) +/* CompressReparseBuffer.WofProvider */ +#define WOF_PROVIDER_WIM cpu_to_le32(1) +/* CompressReparseBuffer.WofProvider */ +#define WOF_PROVIDER_SYSTEM cpu_to_le32(2) +/* CompressReparseBuffer.ProviderVer */ +#define WOF_PROVIDER_CURRENT_VERSION cpu_to_le32(1) + +#define WOF_COMPRESSION_XPRESS4K cpu_to_le32(0) // 4k +#define WOF_COMPRESSION_LZX32K cpu_to_le32(1) // 32k +#define WOF_COMPRESSION_XPRESS8K cpu_to_le32(2) // 8k +#define WOF_COMPRESSION_XPRESS16K cpu_to_le32(3) // 16k + +/* + * ATTR_REPARSE (0xC0) + * + * The reparse struct GUID structure is used by all 3rd party layered drivers to + * store data in a reparse point. For non-Microsoft tags, The struct GUID field + * cannot be GUID_NULL. + * The constraints on reparse tags are defined below. + * Microsoft tags can also be used with this format of the reparse point buffer. + */ +struct REPARSE_POINT { + __le32 ReparseTag; // 0x00: + __le16 ReparseDataLength;// 0x04: + __le16 Reserved; + + struct GUID Guid; // 0x08: + + // + // Here GenericReparseBuffer is placed + // +}; + +static_assert(sizeof(struct REPARSE_POINT) == 0x18); + +// +// Maximum allowed size of the reparse data. +// +#define MAXIMUM_REPARSE_DATA_BUFFER_SIZE (16 * 1024) + +// +// The value of the following constant needs to satisfy the following +// conditions: +// (1) Be at least as large as the largest of the reserved tags. +// (2) Be strictly smaller than all the tags in use. +// +#define IO_REPARSE_TAG_RESERVED_RANGE 1 + +// +// The reparse tags are a ULONG. The 32 bits are laid out as follows: +// +// 3 3 2 2 2 2 2 2 2 2 2 2 1 1 1 1 1 1 1 1 1 1 +// 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 1 0 +// +-+-+-+-+-----------------------+-------------------------------+ +// |M|R|N|R| Reserved bits | Reparse Tag Value | +// +-+-+-+-+-----------------------+-------------------------------+ +// +// M is the Microsoft bit. When set to 1, it denotes a tag owned by Microsoft. +// All ISVs must use a tag with a 0 in this position. +// Note: If a Microsoft tag is used by non-Microsoft software, the +// behavior is not defined. +// +// R is reserved. Must be zero for non-Microsoft tags. +// +// N is name surrogate. When set to 1, the file represents another named +// entity in the system. +// +// The M and N bits are OR-able. +// The following macros check for the M and N bit values: +// + +// +// Macro to determine whether a reparse point tag corresponds to a tag +// owned by Microsoft. +// +#define IsReparseTagMicrosoft(_tag) (((_tag)&IO_REPARSE_TAG_MICROSOFT)) + +// +// Macro to determine whether a reparse point tag is a name surrogate +// +#define IsReparseTagNameSurrogate(_tag) (((_tag)&IO_REPARSE_TAG_NAME_SURROGATE)) + +// +// The following constant represents the bits that are valid to use in +// reparse tags. +// +#define IO_REPARSE_TAG_VALID_VALUES 0xF000FFFF + +// +// Macro to determine whether a reparse tag is a valid tag. +// +#define IsReparseTagValid(_tag) \ + (!((_tag) & ~IO_REPARSE_TAG_VALID_VALUES) && \ + ((_tag) > IO_REPARSE_TAG_RESERVED_RANGE)) + +// +// Microsoft tags for reparse points. +// + +enum IO_REPARSE_TAG { + IO_REPARSE_TAG_SYMBOLIC_LINK = cpu_to_le32(0), + IO_REPARSE_TAG_NAME_SURROGATE = cpu_to_le32(0x20000000), + IO_REPARSE_TAG_MICROSOFT = cpu_to_le32(0x80000000), + IO_REPARSE_TAG_MOUNT_POINT = cpu_to_le32(0xA0000003), + IO_REPARSE_TAG_SYMLINK = cpu_to_le32(0xA000000C), + IO_REPARSE_TAG_HSM = cpu_to_le32(0xC0000004), + IO_REPARSE_TAG_SIS = cpu_to_le32(0x80000007), + IO_REPARSE_TAG_DEDUP = cpu_to_le32(0x80000013), + IO_REPARSE_TAG_COMPRESS = cpu_to_le32(0x80000017), + + // + // The reparse tag 0x80000008 is reserved for Microsoft internal use + // (may be published in the future) + // + + // + // Microsoft reparse tag reserved for DFS + // + IO_REPARSE_TAG_DFS = cpu_to_le32(0x8000000A), + + // + // Microsoft reparse tag reserved for the file system filter manager + // + IO_REPARSE_TAG_FILTER_MANAGER = cpu_to_le32(0x8000000B), + + // + // Non-Microsoft tags for reparse points + // + + // + // Tag allocated to CONGRUENT, May 2000. Used by IFSTEST + // + IO_REPARSE_TAG_IFSTEST_CONGRUENT = cpu_to_le32(0x00000009), + + // + // Tag allocated to ARKIVIO + // + IO_REPARSE_TAG_ARKIVIO = cpu_to_le32(0x0000000C), + + // + // Tag allocated to SOLUTIONSOFT + // + IO_REPARSE_TAG_SOLUTIONSOFT = cpu_to_le32(0x2000000D), + + // + // Tag allocated to COMMVAULT + // + IO_REPARSE_TAG_COMMVAULT = cpu_to_le32(0x0000000E), + + // OneDrive?? + IO_REPARSE_TAG_CLOUD = cpu_to_le32(0x9000001A), + IO_REPARSE_TAG_CLOUD_1 = cpu_to_le32(0x9000101A), + IO_REPARSE_TAG_CLOUD_2 = cpu_to_le32(0x9000201A), + IO_REPARSE_TAG_CLOUD_3 = cpu_to_le32(0x9000301A), + IO_REPARSE_TAG_CLOUD_4 = cpu_to_le32(0x9000401A), + IO_REPARSE_TAG_CLOUD_5 = cpu_to_le32(0x9000501A), + IO_REPARSE_TAG_CLOUD_6 = cpu_to_le32(0x9000601A), + IO_REPARSE_TAG_CLOUD_7 = cpu_to_le32(0x9000701A), + IO_REPARSE_TAG_CLOUD_8 = cpu_to_le32(0x9000801A), + IO_REPARSE_TAG_CLOUD_9 = cpu_to_le32(0x9000901A), + IO_REPARSE_TAG_CLOUD_A = cpu_to_le32(0x9000A01A), + IO_REPARSE_TAG_CLOUD_B = cpu_to_le32(0x9000B01A), + IO_REPARSE_TAG_CLOUD_C = cpu_to_le32(0x9000C01A), + IO_REPARSE_TAG_CLOUD_D = cpu_to_le32(0x9000D01A), + IO_REPARSE_TAG_CLOUD_E = cpu_to_le32(0x9000E01A), + IO_REPARSE_TAG_CLOUD_F = cpu_to_le32(0x9000F01A), + +}; + +#define SYMLINK_FLAG_RELATIVE 1 + +/* Microsoft reparse buffer. (see DDK for details) */ +struct REPARSE_DATA_BUFFER { + __le32 ReparseTag; // 0x00: + __le16 ReparseDataLength; // 0x04: + __le16 Reserved; + + union { + // If ReparseTag == 0xA0000003 (IO_REPARSE_TAG_MOUNT_POINT) + struct { + __le16 SubstituteNameOffset; // 0x08 + __le16 SubstituteNameLength; // 0x0A + __le16 PrintNameOffset; // 0x0C + __le16 PrintNameLength; // 0x0E + __le16 PathBuffer[]; // 0x10 + } MountPointReparseBuffer; + + // If ReparseTag == 0xA000000C (IO_REPARSE_TAG_SYMLINK) + // https://msdn.microsoft.com/en-us/library/cc232006.aspx + struct { + __le16 SubstituteNameOffset; // 0x08 + __le16 SubstituteNameLength; // 0x0A + __le16 PrintNameOffset; // 0x0C + __le16 PrintNameLength; // 0x0E + // 0-absolute path 1- relative path, SYMLINK_FLAG_RELATIVE + __le32 Flags; // 0x10 + __le16 PathBuffer[]; // 0x14 + } SymbolicLinkReparseBuffer; + + // If ReparseTag == 0x80000017U + struct { + __le32 WofVersion; // 0x08 == 1 + /* 1 - WIM backing provider ("WIMBoot"), + * 2 - System compressed file provider + */ + __le32 WofProvider; // 0x0C + __le32 ProviderVer; // 0x10: == 1 WOF_FILE_PROVIDER_CURRENT_VERSION == 1 + __le32 CompressionFormat; // 0x14: 0, 1, 2, 3. See WOF_COMPRESSION_XXX + } CompressReparseBuffer; + + struct { + u8 DataBuffer[1]; // 0x08 + } GenericReparseBuffer; + }; +}; + +/* ATTR_EA_INFO (0xD0) */ + +#define FILE_NEED_EA 0x80 // See ntifs.h +/* FILE_NEED_EA, indicates that the file to which the EA belongs cannot be + * interpreted without understanding the associated extended attributes. + */ +struct EA_INFO { + __le16 size_pack; // 0x00: Size of buffer to hold in packed form + __le16 count; // 0x02: Count of EA's with FILE_NEED_EA bit set + __le32 size; // 0x04: Size of buffer to hold in unpacked form +}; + +static_assert(sizeof(struct EA_INFO) == 8); + +/* ATTR_EA (0xE0) */ +struct EA_FULL { + __le32 size; // 0x00: (not in packed) + u8 flags; // 0x04 + u8 name_len; // 0x05 + __le16 elength; // 0x06 + u8 name[]; // 0x08 +}; + +static_assert(offsetof(struct EA_FULL, name) == 8); + +#define ACL_REVISION 2 +#define ACL_REVISION_DS 4 + +#define SE_SELF_RELATIVE cpu_to_le16(0x8000) + +struct SECURITY_DESCRIPTOR_RELATIVE { + u8 Revision; + u8 Sbz1; + __le16 Control; + __le32 Owner; + __le32 Group; + __le32 Sacl; + __le32 Dacl; +}; +static_assert(sizeof(struct SECURITY_DESCRIPTOR_RELATIVE) == 0x14); + +struct ACE_HEADER { + u8 AceType; + u8 AceFlags; + __le16 AceSize; +}; +static_assert(sizeof(struct ACE_HEADER) == 4); + +struct ACL { + u8 AclRevision; + u8 Sbz1; + __le16 AclSize; + __le16 AceCount; + __le16 Sbz2; +}; +static_assert(sizeof(struct ACL) == 8); + +struct SID { + u8 Revision; + u8 SubAuthorityCount; + u8 IdentifierAuthority[6]; + __le32 SubAuthority[]; +}; +static_assert(offsetof(struct SID, SubAuthority) == 8); + +// clang-format on diff --git a/fs/ntfs3/ntfs_fs.h b/fs/ntfs3/ntfs_fs.h new file mode 100644 index 000000000000..0c3ac89c3115 --- /dev/null +++ b/fs/ntfs3/ntfs_fs.h @@ -0,0 +1,1092 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * + * Copyright (C) 2019-2021 Paragon Software GmbH, All rights reserved. + * + */ + +// clang-format off +#define MINUS_ONE_T ((size_t)(-1)) +/* Biggest MFT / smallest cluster */ +#define MAXIMUM_BYTES_PER_MFT 4096 +#define NTFS_BLOCKS_PER_MFT_RECORD (MAXIMUM_BYTES_PER_MFT / 512) + +#define MAXIMUM_BYTES_PER_INDEX 4096 +#define NTFS_BLOCKS_PER_INODE (MAXIMUM_BYTES_PER_INDEX / 512) + +/* ntfs specific error code when fixup failed*/ +#define E_NTFS_FIXUP 555 +/* ntfs specific error code about resident->nonresident*/ +#define E_NTFS_NONRESIDENT 556 +/* ntfs specific error code about punch hole*/ +#define E_NTFS_NOTALIGNED 557 + + +/* sbi->flags */ +#define NTFS_FLAGS_NODISCARD 0x00000001 +/* Set when LogFile is replaying */ +#define NTFS_FLAGS_LOG_REPLAYING 0x00000008 +/* Set when we changed first MFT's which copy must be updated in $MftMirr */ +#define NTFS_FLAGS_MFTMIRR 0x00001000 +#define NTFS_FLAGS_NEED_REPLAY 0x04000000 + + +/* ni->ni_flags */ +/* + * Data attribute is external compressed (lzx/xpress) + * 1 - WOF_COMPRESSION_XPRESS4K + * 2 - WOF_COMPRESSION_XPRESS8K + * 3 - WOF_COMPRESSION_XPRESS16K + * 4 - WOF_COMPRESSION_LZX32K + */ +#define NI_FLAG_COMPRESSED_MASK 0x0000000f +/* Data attribute is deduplicated */ +#define NI_FLAG_DEDUPLICATED 0x00000010 +#define NI_FLAG_EA 0x00000020 +#define NI_FLAG_DIR 0x00000040 +#define NI_FLAG_RESIDENT 0x00000080 +#define NI_FLAG_UPDATE_PARENT 0x00000100 +// clang-format on + +struct ntfs_mount_options { + struct nls_table *nls; + + kuid_t fs_uid; + kgid_t fs_gid; + u16 fs_fmask_inv; + u16 fs_dmask_inv; + + unsigned uid : 1, /* uid was set */ + gid : 1, /* gid was set */ + fmask : 1, /* fmask was set */ + dmask : 1, /*dmask was set*/ + sys_immutable : 1, /* immutable system files */ + discard : 1, /* issue discard requests on deletions */ + sparse : 1, /*create sparse files*/ + showmeta : 1, /*show meta files*/ + nohidden : 1, /*do not show hidden files*/ + force : 1, /*rw mount dirty volume*/ + no_acs_rules : 1, /*exclude acs rules*/ + prealloc : 1 /*preallocate space when file is growing*/ + ; +}; + +/* special value to unpack and deallocate*/ +#define RUN_DEALLOCATE ((struct runs_tree *)(size_t)1) + +/* TODO: use rb tree instead of array */ +struct runs_tree { + struct ntfs_run *runs; + size_t count; // Currently used size a ntfs_run storage. + size_t allocated; // Currently allocated ntfs_run storage size. +}; + +struct ntfs_buffers { + /* Biggest MFT / smallest cluster = 4096 / 512 = 8 */ + /* Biggest index / smallest cluster = 4096 / 512 = 8 */ + struct buffer_head *bh[PAGE_SIZE >> SECTOR_SHIFT]; + u32 bytes; + u32 nbufs; + u32 off; +}; + +enum ALLOCATE_OPT { + ALLOCATE_DEF = 0, // Allocate all clusters + ALLOCATE_MFT = 1, // Allocate for MFT +}; + +enum bitmap_mutex_classes { + BITMAP_MUTEX_CLUSTERS = 0, + BITMAP_MUTEX_MFT = 1, +}; + +struct wnd_bitmap { + struct super_block *sb; + struct rw_semaphore rw_lock; + + struct runs_tree run; + size_t nbits; + + size_t total_zeroes; // total number of free bits + u16 *free_bits; // free bits in each window + size_t nwnd; + u32 bits_last; // bits in last window + + struct rb_root start_tree; // extents, sorted by 'start' + struct rb_root count_tree; // extents, sorted by 'count + start' + size_t count; // extents count + + /* + * -1 Tree is activated but not updated (too many fragments) + * 0 - Tree is not activated + * 1 - Tree is activated and updated + */ + int uptodated; + size_t extent_min; // Minimal extent used while building + size_t extent_max; // Upper estimate of biggest free block + + /* Zone [bit, end) */ + size_t zone_bit; + size_t zone_end; + + bool set_tail; // not necessary in driver + bool inited; +}; + +typedef int (*NTFS_CMP_FUNC)(const void *key1, size_t len1, const void *key2, + size_t len2, const void *param); + +enum index_mutex_classed { + INDEX_MUTEX_I30 = 0, + INDEX_MUTEX_SII = 1, + INDEX_MUTEX_SDH = 2, + INDEX_MUTEX_SO = 3, + INDEX_MUTEX_SQ = 4, + INDEX_MUTEX_SR = 5, + INDEX_MUTEX_TOTAL +}; + +/* ntfs_index - allocation unit inside directory */ +struct ntfs_index { + struct runs_tree bitmap_run; + struct runs_tree alloc_run; + /* read/write access to 'bitmap_run'/'alloc_run' while ntfs_readdir */ + struct rw_semaphore run_lock; + + /*TODO: remove 'cmp'*/ + NTFS_CMP_FUNC cmp; + + u8 index_bits; // log2(root->index_block_size) + u8 idx2vbn_bits; // log2(root->index_block_clst) + u8 vbn2vbo_bits; // index_block_size < cluster? 9 : cluster_bits + u8 type; // index_mutex_classed +}; + +/* Minimum mft zone */ +#define NTFS_MIN_MFT_ZONE 100 + +/* ntfs file system in-core superblock data */ +struct ntfs_sb_info { + struct super_block *sb; + + u32 discard_granularity; + u64 discard_granularity_mask_inv; // ~(discard_granularity_mask_inv-1) + + u32 cluster_size; // bytes per cluster + u32 cluster_mask; // == cluster_size - 1 + u64 cluster_mask_inv; // ~(cluster_size - 1) + u32 block_mask; // sb->s_blocksize - 1 + u32 blocks_per_cluster; // cluster_size / sb->s_blocksize + + u32 record_size; + u32 sector_size; + u32 index_size; + + u8 sector_bits; + u8 cluster_bits; + u8 record_bits; + + u64 maxbytes; // Maximum size for normal files + u64 maxbytes_sparse; // Maximum size for sparse file + + u32 flags; // See NTFS_FLAGS_XXX + + CLST bad_clusters; // The count of marked bad clusters + + u16 max_bytes_per_attr; // maximum attribute size in record + u16 attr_size_tr; // attribute size threshold (320 bytes) + + /* Records in $Extend */ + CLST objid_no; + CLST quota_no; + CLST reparse_no; + CLST usn_jrnl_no; + + struct ATTR_DEF_ENTRY *def_table; // attribute definition table + u32 def_entries; + u32 ea_max_size; + + struct MFT_REC *new_rec; + + u16 *upcase; + + struct { + u64 lbo, lbo2; + struct ntfs_inode *ni; + struct wnd_bitmap bitmap; // $MFT::Bitmap + /* + * MFT records [11-24) used to expand MFT itself + * They always marked as used in $MFT::Bitmap + * 'reserved_bitmap' contains real bitmap of these records + */ + ulong reserved_bitmap; // bitmap of used records [11 - 24) + size_t next_free; // The next record to allocate from + size_t used; // mft valid size in records + u32 recs_mirr; // Number of records in MFTMirr + u8 next_reserved; + u8 reserved_bitmap_inited; + } mft; + + struct { + struct wnd_bitmap bitmap; // $Bitmap::Data + CLST next_free_lcn; + } used; + + struct { + u64 size; // in bytes + u64 blocks; // in blocks + u64 ser_num; + struct ntfs_inode *ni; + __le16 flags; // cached current VOLUME_INFO::flags, VOLUME_FLAG_DIRTY + u8 major_ver; + u8 minor_ver; + char label[65]; + bool real_dirty; /* real fs state*/ + } volume; + + struct { + struct ntfs_index index_sii; + struct ntfs_index index_sdh; + struct ntfs_inode *ni; + u32 next_id; + u64 next_off; + + __le32 def_security_id; + } security; + + struct { + struct ntfs_index index_r; + struct ntfs_inode *ni; + u64 max_size; // 16K + } reparse; + + struct { + struct ntfs_index index_o; + struct ntfs_inode *ni; + } objid; + + struct { + struct mutex mtx_lznt; + struct lznt *lznt; +#ifdef CONFIG_NTFS3_LZX_XPRESS + struct mutex mtx_xpress; + struct xpress_decompressor *xpress; + struct mutex mtx_lzx; + struct lzx_decompressor *lzx; +#endif + } compress; + + struct ntfs_mount_options options; + struct ratelimit_state msg_ratelimit; +}; + +/* + * one MFT record(usually 1024 bytes), consists of attributes + */ +struct mft_inode { + struct rb_node node; + struct ntfs_sb_info *sbi; + + struct MFT_REC *mrec; + struct ntfs_buffers nb; + + CLST rno; + bool dirty; +}; + +/* nested class for ntfs_inode::ni_lock */ +enum ntfs_inode_mutex_lock_class { + NTFS_INODE_MUTEX_DIRTY, + NTFS_INODE_MUTEX_SECURITY, + NTFS_INODE_MUTEX_OBJID, + NTFS_INODE_MUTEX_REPARSE, + NTFS_INODE_MUTEX_NORMAL, + NTFS_INODE_MUTEX_PARENT, +}; + +/* + * ntfs inode - extends linux inode. consists of one or more mft inodes + */ +struct ntfs_inode { + struct mft_inode mi; // base record + + /* + * Valid size: [0 - i_valid) - these range in file contains valid data + * Range [i_valid - inode->i_size) - contains 0 + * Usually i_valid <= inode->i_size + */ + u64 i_valid; + struct timespec64 i_crtime; + + struct mutex ni_lock; + + /* file attributes from std */ + enum FILE_ATTRIBUTE std_fa; + __le32 std_security_id; + + /* + * tree of mft_inode + * not empty when primary MFT record (usually 1024 bytes) can't save all attributes + * e.g. file becomes too fragmented or contains a lot of names + */ + struct rb_root mi_tree; + + /* + * This member is used in ntfs_readdir to ensure that all subrecords are loaded + */ + u8 mi_loaded; + + union { + struct ntfs_index dir; + struct { + struct rw_semaphore run_lock; + struct runs_tree run; +#ifdef CONFIG_NTFS3_LZX_XPRESS + struct page *offs_page; +#endif + } file; + }; + + struct { + struct runs_tree run; + struct ATTR_LIST_ENTRY *le; // 1K aligned memory + size_t size; + bool dirty; + } attr_list; + + size_t ni_flags; // NI_FLAG_XXX + + struct inode vfs_inode; +}; + +struct indx_node { + struct ntfs_buffers nb; + struct INDEX_BUFFER *index; +}; + +struct ntfs_fnd { + int level; + struct indx_node *nodes[20]; + struct NTFS_DE *de[20]; + struct NTFS_DE *root_de; +}; + +enum REPARSE_SIGN { + REPARSE_NONE = 0, + REPARSE_COMPRESSED = 1, + REPARSE_DEDUPLICATED = 2, + REPARSE_LINK = 3 +}; + +/* functions from attrib.c*/ +int attr_load_runs(struct ATTRIB *attr, struct ntfs_inode *ni, + struct runs_tree *run, const CLST *vcn); +int attr_allocate_clusters(struct ntfs_sb_info *sbi, struct runs_tree *run, + CLST vcn, CLST lcn, CLST len, CLST *pre_alloc, + enum ALLOCATE_OPT opt, CLST *alen, const size_t fr, + CLST *new_lcn); +int attr_make_nonresident(struct ntfs_inode *ni, struct ATTRIB *attr, + struct ATTR_LIST_ENTRY *le, struct mft_inode *mi, + u64 new_size, struct runs_tree *run, + struct ATTRIB **ins_attr, struct page *page); +int attr_set_size(struct ntfs_inode *ni, enum ATTR_TYPE type, + const __le16 *name, u8 name_len, struct runs_tree *run, + u64 new_size, const u64 *new_valid, bool keep_prealloc, + struct ATTRIB **ret); +int attr_data_get_block(struct ntfs_inode *ni, CLST vcn, CLST clen, CLST *lcn, + CLST *len, bool *new); +int attr_data_read_resident(struct ntfs_inode *ni, struct page *page); +int attr_data_write_resident(struct ntfs_inode *ni, struct page *page); +int attr_load_runs_vcn(struct ntfs_inode *ni, enum ATTR_TYPE type, + const __le16 *name, u8 name_len, struct runs_tree *run, + CLST vcn); +int attr_load_runs_range(struct ntfs_inode *ni, enum ATTR_TYPE type, + const __le16 *name, u8 name_len, struct runs_tree *run, + u64 from, u64 to); +int attr_wof_frame_info(struct ntfs_inode *ni, struct ATTRIB *attr, + struct runs_tree *run, u64 frame, u64 frames, + u8 frame_bits, u32 *ondisk_size, u64 *vbo_data); +int attr_is_frame_compressed(struct ntfs_inode *ni, struct ATTRIB *attr, + CLST frame, CLST *clst_data); +int attr_allocate_frame(struct ntfs_inode *ni, CLST frame, size_t compr_size, + u64 new_valid); +int attr_collapse_range(struct ntfs_inode *ni, u64 vbo, u64 bytes); +int attr_punch_hole(struct ntfs_inode *ni, u64 vbo, u64 bytes, u32 *frame_size); + +/* functions from attrlist.c*/ +void al_destroy(struct ntfs_inode *ni); +bool al_verify(struct ntfs_inode *ni); +int ntfs_load_attr_list(struct ntfs_inode *ni, struct ATTRIB *attr); +struct ATTR_LIST_ENTRY *al_enumerate(struct ntfs_inode *ni, + struct ATTR_LIST_ENTRY *le); +struct ATTR_LIST_ENTRY *al_find_le(struct ntfs_inode *ni, + struct ATTR_LIST_ENTRY *le, + const struct ATTRIB *attr); +struct ATTR_LIST_ENTRY *al_find_ex(struct ntfs_inode *ni, + struct ATTR_LIST_ENTRY *le, + enum ATTR_TYPE type, const __le16 *name, + u8 name_len, const CLST *vcn); +int al_add_le(struct ntfs_inode *ni, enum ATTR_TYPE type, const __le16 *name, + u8 name_len, CLST svcn, __le16 id, const struct MFT_REF *ref, + struct ATTR_LIST_ENTRY **new_le); +bool al_remove_le(struct ntfs_inode *ni, struct ATTR_LIST_ENTRY *le); +bool al_delete_le(struct ntfs_inode *ni, enum ATTR_TYPE type, CLST vcn, + const __le16 *name, size_t name_len, + const struct MFT_REF *ref); +int al_update(struct ntfs_inode *ni); +static inline size_t al_aligned(size_t size) +{ + return (size + 1023) & ~(size_t)1023; +} + +/* globals from bitfunc.c */ +bool are_bits_clear(const ulong *map, size_t bit, size_t nbits); +bool are_bits_set(const ulong *map, size_t bit, size_t nbits); +size_t get_set_bits_ex(const ulong *map, size_t bit, size_t nbits); + +/* globals from dir.c */ +int ntfs_utf16_to_nls(struct ntfs_sb_info *sbi, const struct le_str *uni, + u8 *buf, int buf_len); +int ntfs_nls_to_utf16(struct ntfs_sb_info *sbi, const u8 *name, u32 name_len, + struct cpu_str *uni, u32 max_ulen, + enum utf16_endian endian); +struct inode *dir_search_u(struct inode *dir, const struct cpu_str *uni, + struct ntfs_fnd *fnd); +bool dir_is_empty(struct inode *dir); +extern const struct file_operations ntfs_dir_operations; + +/* globals from file.c*/ +int ntfs_getattr(struct user_namespace *mnt_userns, const struct path *path, + struct kstat *stat, u32 request_mask, u32 flags); +void ntfs_sparse_cluster(struct inode *inode, struct page *page0, CLST vcn, + CLST len); +int ntfs3_setattr(struct user_namespace *mnt_userns, struct dentry *dentry, + struct iattr *attr); +int ntfs_file_open(struct inode *inode, struct file *file); +int ntfs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, + __u64 start, __u64 len); +extern const struct inode_operations ntfs_special_inode_operations; +extern const struct inode_operations ntfs_file_inode_operations; +extern const struct file_operations ntfs_file_operations; + +/* globals from frecord.c */ +void ni_remove_mi(struct ntfs_inode *ni, struct mft_inode *mi); +struct ATTR_STD_INFO *ni_std(struct ntfs_inode *ni); +struct ATTR_STD_INFO5 *ni_std5(struct ntfs_inode *ni); +void ni_clear(struct ntfs_inode *ni); +int ni_load_mi_ex(struct ntfs_inode *ni, CLST rno, struct mft_inode **mi); +int ni_load_mi(struct ntfs_inode *ni, struct ATTR_LIST_ENTRY *le, + struct mft_inode **mi); +struct ATTRIB *ni_find_attr(struct ntfs_inode *ni, struct ATTRIB *attr, + struct ATTR_LIST_ENTRY **entry_o, + enum ATTR_TYPE type, const __le16 *name, + u8 name_len, const CLST *vcn, + struct mft_inode **mi); +struct ATTRIB *ni_enum_attr_ex(struct ntfs_inode *ni, struct ATTRIB *attr, + struct ATTR_LIST_ENTRY **le, + struct mft_inode **mi); +struct ATTRIB *ni_load_attr(struct ntfs_inode *ni, enum ATTR_TYPE type, + const __le16 *name, u8 name_len, CLST vcn, + struct mft_inode **pmi); +int ni_load_all_mi(struct ntfs_inode *ni); +bool ni_add_subrecord(struct ntfs_inode *ni, CLST rno, struct mft_inode **mi); +int ni_remove_attr(struct ntfs_inode *ni, enum ATTR_TYPE type, + const __le16 *name, size_t name_len, bool base_only, + const __le16 *id); +int ni_create_attr_list(struct ntfs_inode *ni); +int ni_expand_list(struct ntfs_inode *ni); +int ni_insert_nonresident(struct ntfs_inode *ni, enum ATTR_TYPE type, + const __le16 *name, u8 name_len, + const struct runs_tree *run, CLST svcn, CLST len, + __le16 flags, struct ATTRIB **new_attr, + struct mft_inode **mi); +int ni_insert_resident(struct ntfs_inode *ni, u32 data_size, + enum ATTR_TYPE type, const __le16 *name, u8 name_len, + struct ATTRIB **new_attr, struct mft_inode **mi); +int ni_remove_attr_le(struct ntfs_inode *ni, struct ATTRIB *attr, + struct ATTR_LIST_ENTRY *le); +int ni_delete_all(struct ntfs_inode *ni); +struct ATTR_FILE_NAME *ni_fname_name(struct ntfs_inode *ni, + const struct cpu_str *uni, + const struct MFT_REF *home, + struct ATTR_LIST_ENTRY **entry); +struct ATTR_FILE_NAME *ni_fname_type(struct ntfs_inode *ni, u8 name_type, + struct ATTR_LIST_ENTRY **entry); +int ni_new_attr_flags(struct ntfs_inode *ni, enum FILE_ATTRIBUTE new_fa); +enum REPARSE_SIGN ni_parse_reparse(struct ntfs_inode *ni, struct ATTRIB *attr, + void *buffer); +int ni_write_inode(struct inode *inode, int sync, const char *hint); +#define _ni_write_inode(i, w) ni_write_inode(i, w, __func__) +int ni_fiemap(struct ntfs_inode *ni, struct fiemap_extent_info *fieinfo, + __u64 vbo, __u64 len); +int ni_readpage_cmpr(struct ntfs_inode *ni, struct page *page); +int ni_decompress_file(struct ntfs_inode *ni); +int ni_read_frame(struct ntfs_inode *ni, u64 frame_vbo, struct page **pages, + u32 pages_per_frame); +int ni_write_frame(struct ntfs_inode *ni, struct page **pages, + u32 pages_per_frame); + +/* globals from fslog.c */ +int log_replay(struct ntfs_inode *ni, bool *initialized); + +/* globals from fsntfs.c */ +bool ntfs_fix_pre_write(struct NTFS_RECORD_HEADER *rhdr, size_t bytes); +int ntfs_fix_post_read(struct NTFS_RECORD_HEADER *rhdr, size_t bytes, + bool simple); +int ntfs_extend_init(struct ntfs_sb_info *sbi); +int ntfs_loadlog_and_replay(struct ntfs_inode *ni, struct ntfs_sb_info *sbi); +const struct ATTR_DEF_ENTRY *ntfs_query_def(struct ntfs_sb_info *sbi, + enum ATTR_TYPE Type); +int ntfs_look_for_free_space(struct ntfs_sb_info *sbi, CLST lcn, CLST len, + CLST *new_lcn, CLST *new_len, + enum ALLOCATE_OPT opt); +int ntfs_look_free_mft(struct ntfs_sb_info *sbi, CLST *rno, bool mft, + struct ntfs_inode *ni, struct mft_inode **mi); +void ntfs_mark_rec_free(struct ntfs_sb_info *sbi, CLST rno); +int ntfs_clear_mft_tail(struct ntfs_sb_info *sbi, size_t from, size_t to); +int ntfs_refresh_zone(struct ntfs_sb_info *sbi); +int ntfs_update_mftmirr(struct ntfs_sb_info *sbi, int wait); +enum NTFS_DIRTY_FLAGS { + NTFS_DIRTY_CLEAR = 0, + NTFS_DIRTY_DIRTY = 1, + NTFS_DIRTY_ERROR = 2, +}; +int ntfs_set_state(struct ntfs_sb_info *sbi, enum NTFS_DIRTY_FLAGS dirty); +int ntfs_sb_read(struct super_block *sb, u64 lbo, size_t bytes, void *buffer); +int ntfs_sb_write(struct super_block *sb, u64 lbo, size_t bytes, + const void *buffer, int wait); +int ntfs_sb_write_run(struct ntfs_sb_info *sbi, const struct runs_tree *run, + u64 vbo, const void *buf, size_t bytes); +struct buffer_head *ntfs_bread_run(struct ntfs_sb_info *sbi, + const struct runs_tree *run, u64 vbo); +int ntfs_read_run_nb(struct ntfs_sb_info *sbi, const struct runs_tree *run, + u64 vbo, void *buf, u32 bytes, struct ntfs_buffers *nb); +int ntfs_read_bh(struct ntfs_sb_info *sbi, const struct runs_tree *run, u64 vbo, + struct NTFS_RECORD_HEADER *rhdr, u32 bytes, + struct ntfs_buffers *nb); +int ntfs_get_bh(struct ntfs_sb_info *sbi, const struct runs_tree *run, u64 vbo, + u32 bytes, struct ntfs_buffers *nb); +int ntfs_write_bh(struct ntfs_sb_info *sbi, struct NTFS_RECORD_HEADER *rhdr, + struct ntfs_buffers *nb, int sync); +int ntfs_bio_pages(struct ntfs_sb_info *sbi, const struct runs_tree *run, + struct page **pages, u32 nr_pages, u64 vbo, u32 bytes, + u32 op); +int ntfs_bio_fill_1(struct ntfs_sb_info *sbi, const struct runs_tree *run); +int ntfs_vbo_to_lbo(struct ntfs_sb_info *sbi, const struct runs_tree *run, + u64 vbo, u64 *lbo, u64 *bytes); +struct ntfs_inode *ntfs_new_inode(struct ntfs_sb_info *sbi, CLST nRec, + bool dir); +extern const u8 s_default_security[0x50]; +bool is_sd_valid(const struct SECURITY_DESCRIPTOR_RELATIVE *sd, u32 len); +int ntfs_security_init(struct ntfs_sb_info *sbi); +int ntfs_get_security_by_id(struct ntfs_sb_info *sbi, __le32 security_id, + struct SECURITY_DESCRIPTOR_RELATIVE **sd, + size_t *size); +int ntfs_insert_security(struct ntfs_sb_info *sbi, + const struct SECURITY_DESCRIPTOR_RELATIVE *sd, + u32 size, __le32 *security_id, bool *inserted); +int ntfs_reparse_init(struct ntfs_sb_info *sbi); +int ntfs_objid_init(struct ntfs_sb_info *sbi); +int ntfs_objid_remove(struct ntfs_sb_info *sbi, struct GUID *guid); +int ntfs_insert_reparse(struct ntfs_sb_info *sbi, __le32 rtag, + const struct MFT_REF *ref); +int ntfs_remove_reparse(struct ntfs_sb_info *sbi, __le32 rtag, + const struct MFT_REF *ref); +void mark_as_free_ex(struct ntfs_sb_info *sbi, CLST lcn, CLST len, bool trim); +int run_deallocate(struct ntfs_sb_info *sbi, struct runs_tree *run, bool trim); + +/* globals from index.c */ +int indx_used_bit(struct ntfs_index *indx, struct ntfs_inode *ni, size_t *bit); +void fnd_clear(struct ntfs_fnd *fnd); +static inline struct ntfs_fnd *fnd_get(void) +{ + return ntfs_zalloc(sizeof(struct ntfs_fnd)); +} +static inline void fnd_put(struct ntfs_fnd *fnd) +{ + if (fnd) { + fnd_clear(fnd); + ntfs_free(fnd); + } +} +void indx_clear(struct ntfs_index *idx); +int indx_init(struct ntfs_index *indx, struct ntfs_sb_info *sbi, + const struct ATTRIB *attr, enum index_mutex_classed type); +struct INDEX_ROOT *indx_get_root(struct ntfs_index *indx, struct ntfs_inode *ni, + struct ATTRIB **attr, struct mft_inode **mi); +int indx_read(struct ntfs_index *idx, struct ntfs_inode *ni, CLST vbn, + struct indx_node **node); +int indx_find(struct ntfs_index *indx, struct ntfs_inode *dir, + const struct INDEX_ROOT *root, const void *Key, size_t KeyLen, + const void *param, int *diff, struct NTFS_DE **entry, + struct ntfs_fnd *fnd); +int indx_find_sort(struct ntfs_index *indx, struct ntfs_inode *ni, + const struct INDEX_ROOT *root, struct NTFS_DE **entry, + struct ntfs_fnd *fnd); +int indx_find_raw(struct ntfs_index *indx, struct ntfs_inode *ni, + const struct INDEX_ROOT *root, struct NTFS_DE **entry, + size_t *off, struct ntfs_fnd *fnd); +int indx_insert_entry(struct ntfs_index *indx, struct ntfs_inode *ni, + const struct NTFS_DE *new_de, const void *param, + struct ntfs_fnd *fnd); +int indx_delete_entry(struct ntfs_index *indx, struct ntfs_inode *ni, + const void *key, u32 key_len, const void *param); +int indx_update_dup(struct ntfs_inode *ni, struct ntfs_sb_info *sbi, + const struct ATTR_FILE_NAME *fname, + const struct NTFS_DUP_INFO *dup, int sync); + +/* globals from inode.c */ +struct inode *ntfs_iget5(struct super_block *sb, const struct MFT_REF *ref, + const struct cpu_str *name); +int ntfs_set_size(struct inode *inode, u64 new_size); +int reset_log_file(struct inode *inode); +int ntfs_get_block(struct inode *inode, sector_t vbn, + struct buffer_head *bh_result, int create); +int ntfs3_write_inode(struct inode *inode, struct writeback_control *wbc); +int ntfs_sync_inode(struct inode *inode); +int ntfs_flush_inodes(struct super_block *sb, struct inode *i1, + struct inode *i2); +int inode_write_data(struct inode *inode, const void *data, size_t bytes); +struct inode *ntfs_create_inode(struct user_namespace *mnt_userns, + struct inode *dir, struct dentry *dentry, + const struct cpu_str *uni, umode_t mode, + dev_t dev, const char *symname, u32 size, + struct ntfs_fnd *fnd); +int ntfs_link_inode(struct inode *inode, struct dentry *dentry); +int ntfs_unlink_inode(struct inode *dir, const struct dentry *dentry); +void ntfs_evict_inode(struct inode *inode); +extern const struct inode_operations ntfs_link_inode_operations; +extern const struct address_space_operations ntfs_aops; +extern const struct address_space_operations ntfs_aops_cmpr; + +/* globals from name_i.c*/ +int fill_name_de(struct ntfs_sb_info *sbi, void *buf, const struct qstr *name, + const struct cpu_str *uni); +struct dentry *ntfs3_get_parent(struct dentry *child); + +extern const struct inode_operations ntfs_dir_inode_operations; +extern const struct inode_operations ntfs_special_inode_operations; + +/* globals from record.c */ +int mi_get(struct ntfs_sb_info *sbi, CLST rno, struct mft_inode **mi); +void mi_put(struct mft_inode *mi); +int mi_init(struct mft_inode *mi, struct ntfs_sb_info *sbi, CLST rno); +int mi_read(struct mft_inode *mi, bool is_mft); +struct ATTRIB *mi_enum_attr(struct mft_inode *mi, struct ATTRIB *attr); +// TODO: id? +struct ATTRIB *mi_find_attr(struct mft_inode *mi, struct ATTRIB *attr, + enum ATTR_TYPE type, const __le16 *name, + size_t name_len, const __le16 *id); +static inline struct ATTRIB *rec_find_attr_le(struct mft_inode *rec, + struct ATTR_LIST_ENTRY *le) +{ + return mi_find_attr(rec, NULL, le->type, le_name(le), le->name_len, + &le->id); +} +int mi_write(struct mft_inode *mi, int wait); +int mi_format_new(struct mft_inode *mi, struct ntfs_sb_info *sbi, CLST rno, + __le16 flags, bool is_mft); +void mi_mark_free(struct mft_inode *mi); +struct ATTRIB *mi_insert_attr(struct mft_inode *mi, enum ATTR_TYPE type, + const __le16 *name, u8 name_len, u32 asize, + u16 name_off); + +bool mi_remove_attr(struct mft_inode *mi, struct ATTRIB *attr); +bool mi_resize_attr(struct mft_inode *mi, struct ATTRIB *attr, int bytes); +int mi_pack_runs(struct mft_inode *mi, struct ATTRIB *attr, + struct runs_tree *run, CLST len); +static inline bool mi_is_ref(const struct mft_inode *mi, + const struct MFT_REF *ref) +{ + if (le32_to_cpu(ref->low) != mi->rno) + return false; + if (ref->seq != mi->mrec->seq) + return false; + +#ifdef CONFIG_NTFS3_64BIT_CLUSTER + return le16_to_cpu(ref->high) == (mi->rno >> 32); +#else + return !ref->high; +#endif +} + +static inline void mi_get_ref(const struct mft_inode *mi, struct MFT_REF *ref) +{ + ref->low = cpu_to_le32(mi->rno); +#ifdef CONFIG_NTFS3_64BIT_CLUSTER + ref->high = cpu_to_le16(mi->rno >> 32); +#else + ref->high = 0; +#endif + ref->seq = mi->mrec->seq; +} + +/* globals from run.c */ +bool run_lookup_entry(const struct runs_tree *run, CLST vcn, CLST *lcn, + CLST *len, size_t *index); +void run_truncate(struct runs_tree *run, CLST vcn); +void run_truncate_head(struct runs_tree *run, CLST vcn); +void run_truncate_around(struct runs_tree *run, CLST vcn); +bool run_lookup(const struct runs_tree *run, CLST vcn, size_t *Index); +bool run_add_entry(struct runs_tree *run, CLST vcn, CLST lcn, CLST len, + bool is_mft); +bool run_collapse_range(struct runs_tree *run, CLST vcn, CLST len); +bool run_get_entry(const struct runs_tree *run, size_t index, CLST *vcn, + CLST *lcn, CLST *len); +bool run_is_mapped_full(const struct runs_tree *run, CLST svcn, CLST evcn); + +int run_pack(const struct runs_tree *run, CLST svcn, CLST len, u8 *run_buf, + u32 run_buf_size, CLST *packed_vcns); +int run_unpack(struct runs_tree *run, struct ntfs_sb_info *sbi, CLST ino, + CLST svcn, CLST evcn, CLST vcn, const u8 *run_buf, + u32 run_buf_size); + +#ifdef NTFS3_CHECK_FREE_CLST +int run_unpack_ex(struct runs_tree *run, struct ntfs_sb_info *sbi, CLST ino, + CLST svcn, CLST evcn, CLST vcn, const u8 *run_buf, + u32 run_buf_size); +#else +#define run_unpack_ex run_unpack +#endif +int run_get_highest_vcn(CLST vcn, const u8 *run_buf, u64 *highest_vcn); + +/* globals from super.c */ +void *ntfs_set_shared(void *ptr, u32 bytes); +void *ntfs_put_shared(void *ptr); +void ntfs_unmap_meta(struct super_block *sb, CLST lcn, CLST len); +int ntfs_discard(struct ntfs_sb_info *sbi, CLST Lcn, CLST Len); + +/* globals from bitmap.c*/ +int __init ntfs3_init_bitmap(void); +void ntfs3_exit_bitmap(void); +void wnd_close(struct wnd_bitmap *wnd); +static inline size_t wnd_zeroes(const struct wnd_bitmap *wnd) +{ + return wnd->total_zeroes; +} +int wnd_init(struct wnd_bitmap *wnd, struct super_block *sb, size_t nbits); +int wnd_set_free(struct wnd_bitmap *wnd, size_t bit, size_t bits); +int wnd_set_used(struct wnd_bitmap *wnd, size_t bit, size_t bits); +bool wnd_is_free(struct wnd_bitmap *wnd, size_t bit, size_t bits); +bool wnd_is_used(struct wnd_bitmap *wnd, size_t bit, size_t bits); + +/* Possible values for 'flags' 'wnd_find' */ +#define BITMAP_FIND_MARK_AS_USED 0x01 +#define BITMAP_FIND_FULL 0x02 +size_t wnd_find(struct wnd_bitmap *wnd, size_t to_alloc, size_t hint, + size_t flags, size_t *allocated); +int wnd_extend(struct wnd_bitmap *wnd, size_t new_bits); +void wnd_zone_set(struct wnd_bitmap *wnd, size_t Lcn, size_t Len); +int ntfs_trim_fs(struct ntfs_sb_info *sbi, struct fstrim_range *range); + +/* globals from upcase.c */ +int ntfs_cmp_names(const __le16 *s1, size_t l1, const __le16 *s2, size_t l2, + const u16 *upcase, bool bothcase); +int ntfs_cmp_names_cpu(const struct cpu_str *uni1, const struct le_str *uni2, + const u16 *upcase, bool bothcase); + +/* globals from xattr.c */ +#ifdef CONFIG_NTFS3_FS_POSIX_ACL +struct posix_acl *ntfs_get_acl(struct inode *inode, int type); +int ntfs_set_acl(struct user_namespace *mnt_userns, struct inode *inode, + struct posix_acl *acl, int type); +int ntfs_init_acl(struct user_namespace *mnt_userns, struct inode *inode, + struct inode *dir); +#else +#define ntfs_get_acl NULL +#define ntfs_set_acl NULL +#endif + +int ntfs_acl_chmod(struct user_namespace *mnt_userns, struct inode *inode); +int ntfs_permission(struct user_namespace *mnt_userns, struct inode *inode, + int mask); +ssize_t ntfs_listxattr(struct dentry *dentry, char *buffer, size_t size); +extern const struct xattr_handler *ntfs_xattr_handlers[]; + +int ntfs_save_wsl_perm(struct inode *inode); +void ntfs_get_wsl_perm(struct inode *inode); + +/* globals from lznt.c */ +struct lznt *get_lznt_ctx(int level); +size_t compress_lznt(const void *uncompressed, size_t uncompressed_size, + void *compressed, size_t compressed_size, + struct lznt *ctx); +ssize_t decompress_lznt(const void *compressed, size_t compressed_size, + void *uncompressed, size_t uncompressed_size); + +static inline bool is_ntfs3(struct ntfs_sb_info *sbi) +{ + return sbi->volume.major_ver >= 3; +} + +/*(sb->s_flags & SB_ACTIVE)*/ +static inline bool is_mounted(struct ntfs_sb_info *sbi) +{ + return !!sbi->sb->s_root; +} + +static inline bool ntfs_is_meta_file(struct ntfs_sb_info *sbi, CLST rno) +{ + return rno < MFT_REC_FREE || rno == sbi->objid_no || + rno == sbi->quota_no || rno == sbi->reparse_no || + rno == sbi->usn_jrnl_no; +} + +static inline void ntfs_unmap_page(struct page *page) +{ + kunmap(page); + put_page(page); +} + +static inline struct page *ntfs_map_page(struct address_space *mapping, + unsigned long index) +{ + struct page *page = read_mapping_page(mapping, index, NULL); + + if (!IS_ERR(page)) { + kmap(page); + if (!PageError(page)) + return page; + ntfs_unmap_page(page); + return ERR_PTR(-EIO); + } + return page; +} + +static inline size_t wnd_zone_bit(const struct wnd_bitmap *wnd) +{ + return wnd->zone_bit; +} + +static inline size_t wnd_zone_len(const struct wnd_bitmap *wnd) +{ + return wnd->zone_end - wnd->zone_bit; +} + +static inline void run_init(struct runs_tree *run) +{ + run->runs = NULL; + run->count = 0; + run->allocated = 0; +} + +static inline struct runs_tree *run_alloc(void) +{ + return ntfs_zalloc(sizeof(struct runs_tree)); +} + +static inline void run_close(struct runs_tree *run) +{ + ntfs_vfree(run->runs); + memset(run, 0, sizeof(*run)); +} + +static inline void run_free(struct runs_tree *run) +{ + if (run) { + ntfs_vfree(run->runs); + ntfs_free(run); + } +} + +static inline bool run_is_empty(struct runs_tree *run) +{ + return !run->count; +} + +/* NTFS uses quad aligned bitmaps */ +static inline size_t bitmap_size(size_t bits) +{ + return QuadAlign((bits + 7) >> 3); +} + +#define _100ns2seconds 10000000 +#define SecondsToStartOf1970 0x00000002B6109100 + +#define NTFS_TIME_GRAN 100 + +/* + * kernel2nt + * + * converts in-memory kernel timestamp into nt time + */ +static inline __le64 kernel2nt(const struct timespec64 *ts) +{ + // 10^7 units of 100 nanoseconds one second + return cpu_to_le64(_100ns2seconds * + (ts->tv_sec + SecondsToStartOf1970) + + ts->tv_nsec / NTFS_TIME_GRAN); +} + +/* + * nt2kernel + * + * converts on-disk nt time into kernel timestamp + */ +static inline void nt2kernel(const __le64 tm, struct timespec64 *ts) +{ + u64 t = le64_to_cpu(tm) - _100ns2seconds * SecondsToStartOf1970; + + // WARNING: do_div changes its first argument(!) + ts->tv_nsec = do_div(t, _100ns2seconds) * 100; + ts->tv_sec = t; +} + +static inline struct ntfs_sb_info *ntfs_sb(struct super_block *sb) +{ + return sb->s_fs_info; +} + +/* Align up on cluster boundary */ +static inline u64 ntfs_up_cluster(const struct ntfs_sb_info *sbi, u64 size) +{ + return (size + sbi->cluster_mask) & sbi->cluster_mask_inv; +} + +/* Align up on cluster boundary */ +static inline u64 ntfs_up_block(const struct super_block *sb, u64 size) +{ + return (size + sb->s_blocksize - 1) & ~(u64)(sb->s_blocksize - 1); +} + +static inline CLST bytes_to_cluster(const struct ntfs_sb_info *sbi, u64 size) +{ + return (size + sbi->cluster_mask) >> sbi->cluster_bits; +} + +static inline u64 bytes_to_block(const struct super_block *sb, u64 size) +{ + return (size + sb->s_blocksize - 1) >> sb->s_blocksize_bits; +} + +static inline struct buffer_head *ntfs_bread(struct super_block *sb, + sector_t block) +{ + struct buffer_head *bh = sb_bread(sb, block); + + if (bh) + return bh; + + ntfs_err(sb, "failed to read volume at offset 0x%llx", + (u64)block << sb->s_blocksize_bits); + return NULL; +} + +static inline bool is_power_of2(size_t v) +{ + return v && !(v & (v - 1)); +} + +static inline struct ntfs_inode *ntfs_i(struct inode *inode) +{ + return container_of(inode, struct ntfs_inode, vfs_inode); +} + +static inline bool is_compressed(const struct ntfs_inode *ni) +{ + return (ni->std_fa & FILE_ATTRIBUTE_COMPRESSED) || + (ni->ni_flags & NI_FLAG_COMPRESSED_MASK); +} + +static inline int ni_ext_compress_bits(const struct ntfs_inode *ni) +{ + return 0xb + (ni->ni_flags & NI_FLAG_COMPRESSED_MASK); +} + +/* bits - 0xc, 0xd, 0xe, 0xf, 0x10 */ +static inline void ni_set_ext_compress_bits(struct ntfs_inode *ni, u8 bits) +{ + ni->ni_flags |= (bits - 0xb) & NI_FLAG_COMPRESSED_MASK; +} + +static inline bool is_dedup(const struct ntfs_inode *ni) +{ + return ni->ni_flags & NI_FLAG_DEDUPLICATED; +} + +static inline bool is_encrypted(const struct ntfs_inode *ni) +{ + return ni->std_fa & FILE_ATTRIBUTE_ENCRYPTED; +} + +static inline bool is_sparsed(const struct ntfs_inode *ni) +{ + return ni->std_fa & FILE_ATTRIBUTE_SPARSE_FILE; +} + +static inline int is_resident(struct ntfs_inode *ni) +{ + return ni->ni_flags & NI_FLAG_RESIDENT; +} + +static inline void le16_sub_cpu(__le16 *var, u16 val) +{ + *var = cpu_to_le16(le16_to_cpu(*var) - val); +} + +static inline void le32_sub_cpu(__le32 *var, u32 val) +{ + *var = cpu_to_le32(le32_to_cpu(*var) - val); +} + +static inline void nb_put(struct ntfs_buffers *nb) +{ + u32 i, nbufs = nb->nbufs; + + if (!nbufs) + return; + + for (i = 0; i < nbufs; i++) + put_bh(nb->bh[i]); + nb->nbufs = 0; +} + +static inline void put_indx_node(struct indx_node *in) +{ + if (!in) + return; + + ntfs_free(in->index); + nb_put(&in->nb); + ntfs_free(in); +} + +static inline void mi_clear(struct mft_inode *mi) +{ + nb_put(&mi->nb); + ntfs_free(mi->mrec); + mi->mrec = NULL; +} + +static inline void ni_lock(struct ntfs_inode *ni) +{ + mutex_lock_nested(&ni->ni_lock, NTFS_INODE_MUTEX_NORMAL); +} + +static inline void ni_lock_dir(struct ntfs_inode *ni) +{ + mutex_lock_nested(&ni->ni_lock, NTFS_INODE_MUTEX_PARENT); +} + +static inline void ni_unlock(struct ntfs_inode *ni) +{ + mutex_unlock(&ni->ni_lock); +} + +static inline int ni_trylock(struct ntfs_inode *ni) +{ + return mutex_trylock(&ni->ni_lock); +} + +static inline int attr_load_runs_attr(struct ntfs_inode *ni, + struct ATTRIB *attr, + struct runs_tree *run, CLST vcn) +{ + return attr_load_runs_vcn(ni, attr->type, attr_name(attr), + attr->name_len, run, vcn); +} + +static inline void le64_sub_cpu(__le64 *var, u64 val) +{ + *var = cpu_to_le64(le64_to_cpu(*var) - val); +} diff --git a/fs/ntfs3/upcase.c b/fs/ntfs3/upcase.c new file mode 100644 index 000000000000..9617382aca64 --- /dev/null +++ b/fs/ntfs3/upcase.c @@ -0,0 +1,105 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * + * Copyright (C) 2019-2021 Paragon Software GmbH, All rights reserved. + * + */ +#include <linux/blkdev.h> +#include <linux/buffer_head.h> +#include <linux/module.h> +#include <linux/nls.h> + +#include "debug.h" +#include "ntfs.h" +#include "ntfs_fs.h" + +static inline u16 upcase_unicode_char(const u16 *upcase, u16 chr) +{ + if (chr < 'a') + return chr; + + if (chr <= 'z') + return chr - ('a' - 'A'); + + return upcase[chr]; +} + +/* + * Thanks Kari Argillander <kari.argillander(a)gmail.com> for idea and implementation 'bothcase' + * + * Straigth way to compare names: + * - case insensitive + * - if name equals and 'bothcases' then + * - case sensitive + * 'Straigth way' code scans input names twice in worst case + * Optimized code scans input names only once + */ +int ntfs_cmp_names(const __le16 *s1, size_t l1, const __le16 *s2, size_t l2, + const u16 *upcase, bool bothcase) +{ + int diff1 = 0; + int diff2; + size_t len = min(l1, l2); + + if (!bothcase && upcase) + goto case_insentive; + + for (; len; s1++, s2++, len--) { + diff1 = le16_to_cpu(*s1) - le16_to_cpu(*s2); + if (diff1) { + if (bothcase && upcase) + goto case_insentive; + + return diff1; + } + } + return l1 - l2; + +case_insentive: + for (; len; s1++, s2++, len--) { + diff2 = upcase_unicode_char(upcase, le16_to_cpu(*s1)) - + upcase_unicode_char(upcase, le16_to_cpu(*s2)); + if (diff2) + return diff2; + } + + diff2 = l1 - l2; + return diff2 ? diff2 : diff1; +} + +int ntfs_cmp_names_cpu(const struct cpu_str *uni1, const struct le_str *uni2, + const u16 *upcase, bool bothcase) +{ + const u16 *s1 = uni1->name; + const __le16 *s2 = uni2->name; + size_t l1 = uni1->len; + size_t l2 = uni2->len; + size_t len = min(l1, l2); + int diff1 = 0; + int diff2; + + if (!bothcase && upcase) + goto case_insentive; + + for (; len; s1++, s2++, len--) { + diff1 = *s1 - le16_to_cpu(*s2); + if (diff1) { + if (bothcase && upcase) + goto case_insentive; + + return diff1; + } + } + return l1 - l2; + +case_insentive: + for (; len; s1++, s2++, len--) { + diff2 = upcase_unicode_char(upcase, *s1) - + upcase_unicode_char(upcase, le16_to_cpu(*s2)); + if (diff2) + return diff2; + } + + diff2 = l1 - l2; + return diff2 ? diff2 : diff1; +} -- 2.30.0

2 1

Re: [PATCH OLK-5.10] x86/config: enable btrfs feature
by Xie XiuQi 13 Dec '21

13 Dec '21

Hi, Thanks for your patch. Please subscribe the kernel(a)openeuler.org and kernel-discuss(a)openeuler.org mailing list before sending patches, or your patches would be blocked. Others could not receive your email. Please resend this patch. On 2021/12/13 15:11, Jiao Fenfang wrote: > From: JiaoFenfang <jiaofenfang(a)uniontech.com> > > openEuler inclusion > category: feature > bugzilla: https://gitee.com/openeuler/kernel/issues/I4KVP7?from=project-issue > Add commit message here. > Signed-off-by: Jiao Fenfang <jiaofenfang(a)uniontech.com> > --- > arch/x86/configs/openeuler_defconfig | 8 +++++++- > 1 file changed, 7 insertions(+), 1 deletion(-) > Also enable it on arm64 platform? > diff --git a/arch/x86/configs/openeuler_defconfig b/arch/x86/configs/openeuler_defconfig > index 7b608301823c..01a902eba3b9 100644 > --- a/arch/x86/configs/openeuler_defconfig > +++ b/arch/x86/configs/openeuler_defconfig > @@ -7451,7 +7451,13 @@ CONFIG_XFS_POSIX_ACL=y > CONFIG_GFS2_FS=m > CONFIG_GFS2_FS_LOCKING_DLM=y > # CONFIG_OCFS2_FS is not set > -# CONFIG_BTRFS_FS is not set > +CONFIG_BTRFS_FS=m > +# CONFIG_BTRFS_FS_POSIX_ACL is not set > +# CONFIG_BTRFS_FS_CHECK_INTEGRITY is not set > +# CONFIG_BTRFS_FS_RUN_SANITY_TESTS is not set > +# CONFIG_BTRFS_DEBUG is not set > +# CONFIG_BTRFS_ASSERT is not set > +# CONFIG_BTRFS_FS_REF_VERIFY is not set > # CONFIG_NILFS2_FS is not set > # CONFIG_F2FS_FS is not set > # CONFIG_ZONEFS_FS is not set >

1 0

[PATCH openEuler-1.0-LTS] ixgbe: fix large MTU request from VF
by Yang Yingliang 13 Dec '21

13 Dec '21

From: Jesse Brandeburg <jesse.brandeburg(a)intel.com> mainline inclusion from mainline-v5.13-rc4 commit 63e39d29b3da02e901349f6cd71159818a4737a6 category: bugfix bugzilla: NA CVE: CVE-2021-33098 ----------------------------------------------- Check that the MTU value requested by the VF is in the supported range of MTUs before attempting to set the VF large packet enable, otherwise reject the request. This also avoids unnecessary register updates in the case of the 82599 controller. Fixes: 872844ddb9e4 ("ixgbe: Enable jumbo frames support w/ SR-IOV") Co-developed-by: Piotr Skajewski <piotrx.skajewski(a)intel.com> Signed-off-by: Piotr Skajewski <piotrx.skajewski(a)intel.com> Signed-off-by: Jesse Brandeburg <jesse.brandeburg(a)intel.com> Co-developed-by: Mateusz Palczewski <mateusz.palczewski(a)intel.com> Signed-off-by: Mateusz Palczewski <mateusz.palczewski(a)intel.com> Tested-by: Konrad Jankowski <konrad0.jankowski(a)intel.com> Signed-off-by: Tony Nguyen <anthony.l.nguyen(a)intel.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Huang Guobin <huangguobin4(a)huawei.com> Reviewed-by: Wei Yongjun <weiyongjun1(a)huawei.com> Reviewed-by: Xiu Jianfeng <xiujianfeng(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.c | 16 +++++++--------- 1 file changed, 7 insertions(+), 9 deletions(-) diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.c index f6ffd9fb20793..8aaf856771d7b 100644 --- a/drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.c +++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_sriov.c @@ -467,12 +467,16 @@ static int ixgbe_set_vf_vlan(struct ixgbe_adapter *adapter, int add, int vid, return err; } -static s32 ixgbe_set_vf_lpe(struct ixgbe_adapter *adapter, u32 *msgbuf, u32 vf) +static int ixgbe_set_vf_lpe(struct ixgbe_adapter *adapter, u32 max_frame, u32 vf) { struct ixgbe_hw *hw = &adapter->hw; - int max_frame = msgbuf[1]; u32 max_frs; + if (max_frame < ETH_MIN_MTU || max_frame > IXGBE_MAX_JUMBO_FRAME_SIZE) { + e_err(drv, "VF max_frame %d out of range\n", max_frame); + return -EINVAL; + } + /* * For 82599EB we have to keep all PFs and VFs operating with * the same max_frame value in order to avoid sending an oversize @@ -532,12 +536,6 @@ static s32 ixgbe_set_vf_lpe(struct ixgbe_adapter *adapter, u32 *msgbuf, u32 vf) } } - /* MTU < 68 is an error and causes problems on some kernels */ - if (max_frame > IXGBE_MAX_JUMBO_FRAME_SIZE) { - e_err(drv, "VF max_frame %d out of range\n", max_frame); - return -EINVAL; - } - /* pull current max frame size from hardware */ max_frs = IXGBE_READ_REG(hw, IXGBE_MAXFRS); max_frs &= IXGBE_MHADD_MFS_MASK; @@ -1240,7 +1238,7 @@ static int ixgbe_rcv_msg_from_vf(struct ixgbe_adapter *adapter, u32 vf) retval = ixgbe_set_vf_vlan_msg(adapter, msgbuf, vf); break; case IXGBE_VF_SET_LPE: - retval = ixgbe_set_vf_lpe(adapter, msgbuf, vf); + retval = ixgbe_set_vf_lpe(adapter, msgbuf[1], vf); break; case IXGBE_VF_SET_MACVLAN: retval = ixgbe_set_vf_macvlan_msg(adapter, msgbuf, vf); -- 2.25.1

1 0

Re: [PATCH] vfio/iommu_type1: replace kfree with kvfree
by Xie XiuQi 13 Dec '21

13 Dec '21

Hi, Thanks for your patch. Please subscribe the kernel(a)openeuler.org and kernel-discuss(a)openeuler.org mailing list before sending patches, or your patches would be blocked. Please resend this patch, and specific the right branch. Welcome to start a topic at openEuler kernel sig meeting via seeding mail to kernel-discuss(a)openeuler.org. --- 可以通过openEuler kernel 小助手，参加openEuler kernel SIG 微信群。 https://openeuler.gitee.io/kernel-portal/img/wechat/openEuler_kernel_helper… On 2021/12/12 18:14, Jiacheng Shi wrote: > From: billsjchw <billsjc(a)126.com> > > Variables allocated by kvmalloc should not be freed by kfree. > Because they may be allocated by vmalloc. > So we replace kfree with kvfree here. > > Signed-off-by: Jiacheng Shi <billsjc(a)sjtu.edu.cn> > --- > drivers/vfio/vfio_iommu_type1.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/drivers/vfio/vfio_iommu_type1.c b/drivers/vfio/vfio_iommu_type1.c > index 5daceec48811..6811a85109aa 100644 > --- a/drivers/vfio/vfio_iommu_type1.c > +++ b/drivers/vfio/vfio_iommu_type1.c > @@ -1150,7 +1150,7 @@ static int vfio_iova_dirty_log_clear(u64 __user *bitmap, > } > > out: > - kfree(bitmap_buffer); > + kvfree(bitmap_buffer); > return ret; > } > >

1 0

[PATCH master] staging: r8188eu: fix a memory leak in rtw_mp_QueryDrv()
by Jianglei Nie 12 Dec '21

12 Dec '21

stable inclusion category: bugfix bugzilla: NA CVE: NA Line 6191 (#1) allocates a memory chunk for input by kmalloc(). Line 6213 (#3) frees the input before the function returns while line 6199 (#2) forget to free it, which will lead to a memory leak. This bug influences all stable versions from 5.15.1 to 5.15.7. We should kfree() input in line 6199 (#2). 6186 static int rtw_mp_QueryDrv(struct net_device *dev, 6187 struct iw_request_info *info, 6188 union iwreq_data *wrqu, char *extra) 6189 { 6191 char *input = kmalloc(wrqu->data.length, GFP_KERNEL); // #1: kmalloc space 6195 if (!input) 6196 return -ENOMEM; 6198 if (copy_from_user(input, wrqu->data.pointer, wrqu->data.length)) 6199 return -EFAULT; // #2: missing kfree 6213 kfree(input); // #3: kfree space 6214 return 0; 6215 } Signed-off-by: Jianglei Nie <niejianglei2021(a)163.com> --- drivers/staging/r8188eu/os_dep/ioctl_linux.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/staging/r8188eu/os_dep/ioctl_linux.c b/drivers/staging/r8188eu/os_dep/ioctl_linux.c index 906a57eae1af..edc660f15436 100644 --- a/drivers/staging/r8188eu/os_dep/ioctl_linux.c +++ b/drivers/staging/r8188eu/os_dep/ioctl_linux.c @@ -6195,8 +6195,11 @@ static int rtw_mp_QueryDrv(struct net_device *dev, if (!input) return -ENOMEM; - if (copy_from_user(input, wrqu->data.pointer, wrqu->data.length)) - return -EFAULT; + if (copy_from_user(input, wrqu->data.pointer, wrqu->data.length)) { + kfree(input); + return -EFAULT; + } + DBG_88E("%s:iwpriv in =%s\n", __func__, input); qAutoLoad = strncmp(input, "autoload", 8); /* strncmp true is 0 */ -- 2.25.1

1 0

[PATCH master] staging: r8188eu: fix a memory leak in rtw_mp_pwrtrk()
by Jianglei Nie 12 Dec '21

12 Dec '21

stable inclusion category: bugfix bugzilla: NA CVE: NA Line 5968 (#1) allocates a memory chunk for input by kmalloc(). Line 5973 (#2), line 5989 (#4) and line 5994 (#5) free the input before the function returns while line 5986 (#3) forget to free it, which will lead to a memory leak. This bug influences all stable versions from 5.15 to 5.15.7. We should kfree() input in line 5986 (#3). 5960 static int rtw_mp_pwrtrk(struct net_device *dev, 5961 struct iw_request_info *info, 5962 struct iw_point *wrqu, char *extra) 5963 { 5968 char *input = kmalloc(wrqu->length, GFP_KERNEL); // #1: kmalloc space 5970 if (!input) 5971 return -ENOMEM; 5972 if (copy_from_user(input, wrqu->pointer, wrqu->length)) { 5973 kfree(input); // #2: kfree space 5974 return -EFAULT; 5975 } 5980 if (strncmp(input, "stop", 4) == 0) { 5981 enable = 0; 5982 sprintf(extra, "mp tx power tracking stop"); 5983 } else if (sscanf(input, "ther =%d", &thermal)) { 5984 ret = Hal_SetThermalMeter(padapter, (u8)thermal); 5985 if (ret == _FAIL) 5986 return -EPERM; // #3: missing kfree 5987 sprintf(extra, "mp tx power tracking start, target value =%d ok ", thermal); 5988 } else { 5989 kfree(input); // #4: kfree space 5990 return -EINVAL; 5991 } 5994 kfree(input); // #5: kfree space 6000 return 0; 6001 } Signed-off-by: Jianglei Nie <niejianglei2021(a)163.com> --- drivers/staging/r8188eu/os_dep/ioctl_linux.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/staging/r8188eu/os_dep/ioctl_linux.c b/drivers/staging/r8188eu/os_dep/ioctl_linux.c index 0eccce57c63a..906a57eae1af 100644 --- a/drivers/staging/r8188eu/os_dep/ioctl_linux.c +++ b/drivers/staging/r8188eu/os_dep/ioctl_linux.c @@ -5982,8 +5982,10 @@ static int rtw_mp_pwrtrk(struct net_device *dev, sprintf(extra, "mp tx power tracking stop"); } else if (sscanf(input, "ther =%d", &thermal)) { ret = Hal_SetThermalMeter(padapter, (u8)thermal); - if (ret == _FAIL) + if (ret == _FAIL) { + kfree(input); return -EPERM; + } sprintf(extra, "mp tx power tracking start, target value =%d ok ", thermal); } else { kfree(input); -- 2.25.1

1 0

[PATCH master] powercap: DTPM: Fix reference leak in cpuhp_dtpm_cpu_offline()
by Jianglei Nie 12 Dec '21

12 Dec '21

stable inclusion category: bugfix bugzilla: NA CVE: NA In line 153 (#1), cpufreq_cpu_get() increments the kobject reference counter of the policy it returned on success. According to the document, the policy returned by cpufreq_cpu_get() has to be released with the help of cpufreq_cpu_put() to balance its kobject reference counter properly. Forgetting the cpufreq_cpu_put() operation will result in reference leak.This bug influences all stable versions from v5.15 to v5.15.7. We can fix it by calling cpufreq_cpu_put() before the function returns (#2, #3 and #4). 147 static int cpuhp_dtpm_cpu_offline(unsigned int cpu) 148 { 153 policy = cpufreq_cpu_get(cpu); // #1: reference increment 155 if (!policy) 156 return 0; 158 pd = em_cpu_get(cpu); 159 if (!pd) 160 return -EINVAL; // #2: missing reference decrement 166 if (cpumask_weight(policy->cpus) != 1) 167 return 0; // #3: missing reference decrement 174 return 0; // #4: missing reference decrement 175 } Signed-off-by: Jianglei Nie <niejianglei2021(a)163.com> --- drivers/powercap/dtpm_cpu.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/drivers/powercap/dtpm_cpu.c b/drivers/powercap/dtpm_cpu.c index 51c366938acd..6c94515b21ef 100644 --- a/drivers/powercap/dtpm_cpu.c +++ b/drivers/powercap/dtpm_cpu.c @@ -156,21 +156,25 @@ static int cpuhp_dtpm_cpu_offline(unsigned int cpu) return 0; pd = em_cpu_get(cpu); - if (!pd) + if (!pd) { + cpufreq_cpu_put(policy); return -EINVAL; + } dtpm = per_cpu(dtpm_per_cpu, cpu); power_sub(dtpm, pd); - if (cpumask_weight(policy->cpus) != 1) + if (cpumask_weight(policy->cpus) != 1) { + cpufreq_cpu_put(policy); return 0; + } for_each_cpu(cpu, policy->related_cpus) per_cpu(dtpm_per_cpu, cpu) = NULL; dtpm_unregister(dtpm); - + cpufreq_cpu_put(policy); return 0; } -- 2.25.1

1 0

[PATCH master] security:trusted_tpm2: Fix memory leak in tpm2_key_encode()
by Jianglei Nie 12 Dec '21

12 Dec '21

openEuler inclusion category: bugfix bugzilla: NA CVE: NA Line 36 (#1) allocates a memory chunk for scratch by kmalloc(), but it is never freed through the function, which will lead to a memory leak. We should kfree() scratch before the function returns (#2, #3 and #4). 31 static int tpm2_key_encode(struct trusted_key_payload *payload, 32 struct trusted_key_options *options, 33 u8 *src, u32 len) 34 { 36 u8 *scratch = kmalloc(SCRATCH_SIZE, GFP_KERNEL); // #1: kmalloc space 37 u8 *work = scratch, *work1; 50 if (!scratch) 51 return -ENOMEM; 56 if (options->blobauth_len == 0) { 60 if (WARN(IS_ERR(w), "BUG: Boolean failed to encode")) 61 return PTR_ERR(w); // #2: missing kfree 63 } 71 if (WARN(work - scratch + pub_len + priv_len + 14 > SCRATCH_SIZE, 72 "BUG: scratch buffer is too small")) 73 return -EINVAL; // #3: missing kfree // #4: missing kfree: scratch is never used afterwards. 82 if (WARN(IS_ERR(work1), "BUG: ASN.1 encoder failed")) 83 return PTR_ERR(work1); 85 return work1 - payload->blob; 86 } Signed-off-by: Jianglei Nie <niejianglei2021(a)163.com> --- security/keys/trusted-keys/trusted_tpm2.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) diff --git a/security/keys/trusted-keys/trusted_tpm2.c b/security/keys/trusted-keys/trusted_tpm2.c index 0165da386289..3408a74c855f 100644 --- a/security/keys/trusted-keys/trusted_tpm2.c +++ b/security/keys/trusted-keys/trusted_tpm2.c @@ -57,8 +57,10 @@ static int tpm2_key_encode(struct trusted_key_payload *payload, unsigned char bool[3], *w = bool; /* tag 0 is emptyAuth */ w = asn1_encode_boolean(w, w + sizeof(bool), true); - if (WARN(IS_ERR(w), "BUG: Boolean failed to encode")) + if (WARN(IS_ERR(w), "BUG: Boolean failed to encode")) { + kfree(scratch); return PTR_ERR(w); + } work = asn1_encode_tag(work, end_work, 0, bool, w - bool); } @@ -69,9 +71,12 @@ static int tpm2_key_encode(struct trusted_key_payload *payload, * trigger, so if it does there's something nefarious going on */ if (WARN(work - scratch + pub_len + priv_len + 14 > SCRATCH_SIZE, - "BUG: scratch buffer is too small")) + "BUG: scratch buffer is too small")) { + kfree(scratch); return -EINVAL; + } + kfree(scratch); work = asn1_encode_integer(work, end_work, options->keyhandle); work = asn1_encode_octet_string(work, end_work, pub, pub_len); work = asn1_encode_octet_string(work, end_work, priv, priv_len); -- 2.25.1

1 0

[PATCH master] btrfs: Fix memory leak in __add_inode_ref()
by Jianglei Nie 12 Dec '21

12 Dec '21

openEuler inclusion category: bugfix bugzilla: NA CVE: NA Line 1169 (#3) allocates a memory chunk for victim_name by kmalloc(), but when the function returns in line 1184 (#4) victim_name allcoated by line 1169 (#3) is not freed, which will lead to a memory leak. There is a similar snippet of code in this function as allocating a memory chunk for victim_name in line 1104 (#1) as well as releasing the memory in line 1116 (#2). We should kfree() victim_name when the return value of backref_in_log() is less than zero and before the function returns in line 1184 (#4). 1057 static inline int __add_inode_ref(struct btrfs_trans_handle *trans, 1058 struct btrfs_root *root, 1059 struct btrfs_path *path, 1060 struct btrfs_root *log_root, 1061 struct btrfs_inode *dir, 1062 struct btrfs_inode *inode, 1063 u64 inode_objectid, u64 parent_objectid, 1064 u64 ref_index, char *name, int namelen, 1065 int *search_done) 1066 { 1104 victim_name = kmalloc(victim_name_len, GFP_NOFS); // #1: kmalloc (victim_name-1) 1105 if (!victim_name) 1106 return -ENOMEM; 1112 ret = backref_in_log(log_root, &search_key, 1113 parent_objectid, victim_name, 1114 victim_name_len); 1115 if (ret < 0) { 1116 kfree(victim_name); // #2: kfree (victim_name-1) 1117 return ret; 1118 } else if (!ret) { 1169 victim_name = kmalloc(victim_name_len, GFP_NOFS); // #3: kmalloc (victim_name-2) 1170 if (!victim_name) 1171 return -ENOMEM; 1180 ret = backref_in_log(log_root, &search_key, 1181 parent_objectid, victim_name, 1182 victim_name_len); 1183 if (ret < 0) { 1184 return ret; // #4: missing kfree (victim_name-2) 1185 } else if (!ret) { 1241 return 0; 1242 } Signed-off-by: Jianglei Nie <niejianglei2021(a)163.com> --- fs/btrfs/tree-log.c | 1 + 1 file changed, 1 insertion(+) diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index 8ab33caf016f..d373fec55521 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -1181,6 +1181,7 @@ static inline int __add_inode_ref(struct btrfs_trans_handle *trans, parent_objectid, victim_name, victim_name_len); if (ret < 0) { + kfree(victim_name); return ret; } else if (!ret) { ret = -ENOENT; -- 2.25.1

1 0

[PATCH master] scsi: csiostor: Fix memory leak in csio_wr_eq_destroy()
by Jianglei Nie 12 Dec '21

12 Dec '21

openEuler inclusion category: bugfix bugzilla: NA CVE: NA Line 715 (#1) calls mempool_alloc() to allocate an element from a specific memory pool. When some errors occur, line 727 (#2) frees this memory pool but line 731 (#3) does not free it, which will lead to a memory leak. We can fix it by calling mempool_free() when the cbfn is not equal to NULL and before the function returns 0 in line 732 (#3). 705 static int 706 csio_wr_eq_destroy(struct csio_hw *hw, void *priv, int eq_idx, 707 void (*cbfn) (struct csio_hw *, struct csio_mb *)) 708 { 710 struct csio_mb *mbp; 715 mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC); // #1: allocate memory pool 716 if (!mbp) 717 return -ENOMEM; 725 rv = csio_mb_issue(hw, mbp); 726 if (rv != 0) { 727 mempool_free(mbp, hw->mb_mempool); // #2: free memory pool 728 return rv; 729 } 731 if (cbfn != NULL) 732 return 0; // #3: missing free 734 return csio_wr_eq_destroy_rsp(hw, mbp, eq_idx); 735 } Signed-off-by: Jianglei Nie <niejianglei2021(a)163.com> --- drivers/scsi/csiostor/csio_wr.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/scsi/csiostor/csio_wr.c b/drivers/scsi/csiostor/csio_wr.c index fe0355c964bc..7dcc4fda0483 100644 --- a/drivers/scsi/csiostor/csio_wr.c +++ b/drivers/scsi/csiostor/csio_wr.c @@ -728,8 +728,10 @@ csio_wr_eq_destroy(struct csio_hw *hw, void *priv, int eq_idx, return rv; } - if (cbfn != NULL) + if (cbfn != NULL) { + mempool_free(mbp, hw->mb_mempool); return 0; + } return csio_wr_eq_destroy_rsp(hw, mbp, eq_idx); } -- 2.25.1

1 0

[PATCH master] drm/amdgpu: Fix reference leak in psp_xgmi_reflect_topology_info()
by Jianglei Nie 12 Dec '21

12 Dec '21

openEuler inclusion category: bugfix bugzilla: NA CVE: NA In line 1138 (#1), amdgpu_get_xgmi_hive() increases the kobject reference counter of the hive it returned. The hive returned by amdgpu_get_xgmi_hive() should be released with the help of amdgpu_put_xgmi_hive() to balance its kobject reference counter properly. Forgetting the amdgpu_put_xgmi_hive() operation will result in reference leak. We can fix it by calling amdgpu_put_xgmi_hive() before the end of the function (#2). 1128 static void psp_xgmi_reflect_topology_info(struct psp_context *psp, 1129 struct psp_xgmi_node_info node_info) 1130 { 1138 hive = amdgpu_get_xgmi_hive(psp->adev); // #1: kzalloc space reference increment 1139 list_for_each_entry(mirror_adev, &hive->device_list, gmc.xgmi.head) { 1140 struct psp_xgmi_topology_info *mirror_top_info; 1141 int j; 1143 if (mirror_adev->gmc.xgmi.node_id != dst_node_id) 1144 continue; 1146 mirror_top_info = &mirror_adev->psp.xgmi_context.top_info; 1147 for (j = 0; j < mirror_top_info->num_nodes; j++) { 1148 if (mirror_top_info->nodes[j].node_id != src_node_id) 1149 continue; 1151 mirror_top_info->nodes[j].num_hops = dst_num_hops; 1157 if (dst_num_links) 1158 mirror_top_info->nodes[j].num_links = dst_num_links; 1160 break; 1161 } 1163 break; 1164 } // #2: missing reference decrement 1165 } Signed-off-by: Jianglei Nie <niejianglei2021(a)163.com> --- drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c index c641f84649d6..f6362047ed71 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c @@ -1162,6 +1162,7 @@ static void psp_xgmi_reflect_topology_info(struct psp_context *psp, break; } + amdgpu_put_xgmi_hive(hive); } int psp_xgmi_get_topology_info(struct psp_context *psp, -- 2.25.1

1 0

[PATCH master] nfp: Fix memory leak in nfp_cpp_area_cache_add()
by Jianglei Nie 12 Dec '21

12 Dec '21

mainline inclusion from mainline-v5.16-rc7 category: bugfix commit: c56c96303e9289cc34716b1179597b6f470833de bugzilla: NA CVE: NA ------------------------ In line 800 (#1), nfp_cpp_area_alloc() allocates and initializes a CPP area structure. But in line 807 (#2), when the cache is allocated failed, this CPP area structure is not freed, which will result in memory leak. We can fix it by freeing the CPP area when the cache is allocated failed (#2). 792 int nfp_cpp_area_cache_add(struct nfp_cpp *cpp, size_t size) 793 { 794 struct nfp_cpp_area_cache *cache; 795 struct nfp_cpp_area *area; 800 area = nfp_cpp_area_alloc(cpp, NFP_CPP_ID(7, NFP_CPP_ACTION_RW, 0), 801 0, size); // #1: allocates and initializes 802 if (!area) 803 return -ENOMEM; 805 cache = kzalloc(sizeof(*cache), GFP_KERNEL); 806 if (!cache) 807 return -ENOMEM; // #2: missing free 817 return 0; 818 } Fixes: 4cb584e0ee7d ("nfp: add CPP access core") Signed-off-by: Jianglei Nie <niejianglei2021(a)163.com> Acked-by: Simon Horman <simon.horman(a)corigine.com> Link: https://lore.kernel.org/r/20211209061511.122535-1-niejianglei2021@163.com Signed-off-by: Jakub Kicinski <kuba(a)kernel.org> Signed-off-by: Jianglei Nie <niejianglei2021(a)163.com> --- drivers/net/ethernet/netronome/nfp/nfpcore/nfp_cppcore.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/netronome/nfp/nfpcore/nfp_cppcore.c b/drivers/net/ethernet/netronome/nfp/nfpcore/nfp_cppcore.c index d7ac0307797f..34c0d2ddf9ef 100644 --- a/drivers/net/ethernet/netronome/nfp/nfpcore/nfp_cppcore.c +++ b/drivers/net/ethernet/netronome/nfp/nfpcore/nfp_cppcore.c @@ -803,8 +803,10 @@ int nfp_cpp_area_cache_add(struct nfp_cpp *cpp, size_t size) return -ENOMEM; cache = kzalloc(sizeof(*cache), GFP_KERNEL); - if (!cache) + if (!cache) { + nfp_cpp_area_free(area); return -ENOMEM; + } cache->id = 0; cache->addr = 0; -- 2.25.1

1 0

Re: [PATCH] btrfs: Fix memory leak in __add_inode_ref()
by Xie XiuQi 10 Dec '21

10 Dec '21

Hi, Thanks for your patchset. Please subscribe the kernel(a)openeuler.org and kernel-discuss(a)openeuler.org mailing list before sending patchs, or your patchs would be blocked. Welcome to start a topic at openEuler kernel sig meeting via seeding mail to kernel-discuss(a)openeuler.org. --- 可以通过openEuler kernel 小助手，参加openEuler kernel SIG 微信群。 https://openeuler.gitee.io/kernel-portal/img/wechat/openEuler_kernel_helper… On 2021/12/10 16:09, Jianglei Nie wrote: > Line 1169 (#3) allocates a memory chunk for victim_name by kmalloc(), > but when the function returns in line 1184 (#4) victim_name allcoated > by line 1169 (#3) is not freed, which will lead to a memory leak. > There is a similar snippet of code in this function as allocating a memory > chunk for victim_name in line 1104 (#1) as well as releasing the memory > in line 1116 (#2). > > We should kfree() victim_name when the return value of backref_in_log() > is less than zero and before the function returns in line 1184 (#4). > > 1057 static inline int __add_inode_ref(struct btrfs_trans_handle *trans, > 1058 struct btrfs_root *root, > 1059 struct btrfs_path *path, > 1060 struct btrfs_root *log_root, > 1061 struct btrfs_inode *dir, > 1062 struct btrfs_inode *inode, > 1063 u64 inode_objectid, u64 parent_objectid, > 1064 u64 ref_index, char *name, int namelen, > 1065 int *search_done) > 1066 { > > 1104 victim_name = kmalloc(victim_name_len, GFP_NOFS); > // #1: kmalloc (victim_name-1) > 1105 if (!victim_name) > 1106 return -ENOMEM; > > 1112 ret = backref_in_log(log_root, &search_key, > 1113 parent_objectid, victim_name, > 1114 victim_name_len); > 1115 if (ret < 0) { > 1116 kfree(victim_name); // #2: kfree (victim_name-1) > 1117 return ret; > 1118 } else if (!ret) { > > 1169 victim_name = kmalloc(victim_name_len, GFP_NOFS); > // #3: kmalloc (victim_name-2) > 1170 if (!victim_name) > 1171 return -ENOMEM; > > 1180 ret = backref_in_log(log_root, &search_key, > 1181 parent_objectid, victim_name, > 1182 victim_name_len); > 1183 if (ret < 0) { > 1184 return ret; // #4: missing kfree (victim_name-2) > 1185 } else if (!ret) { > > 1241 return 0; > 1242 } > > Signed-off-by: Jianglei Nie <niejianglei2021(a)163.com> > --- > fs/btrfs/tree-log.c | 1 + > 1 file changed, 1 insertion(+) > > diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c > index 8ab33caf016f..d373fec55521 100644 > --- a/fs/btrfs/tree-log.c > +++ b/fs/btrfs/tree-log.c > @@ -1181,6 +1181,7 @@ static inline int __add_inode_ref(struct btrfs_trans_handle *trans, > parent_objectid, victim_name, > victim_name_len); > if (ret < 0) { > + kfree(victim_name); > return ret; > } else if (!ret) { > ret = -ENOENT; >

1 0

[PATCH openEuler-5.10 1/7] bfq: Remove merged request already in bfq_requests_merged()
by Zheng Zengkai 10 Dec '21

10 Dec '21

From: Jan Kara <jack(a)suse.cz> mainline inclusion from mainline-v5.14-rc1 commit a921c655f2033dd1ce1379128efe881dda23ea37 category: bugfix bugzilla: 185777 https://gitee.com/openeuler/kernel/issues/I4LM14 CVE: NA --------------------------- Currently, bfq does very little in bfq_requests_merged() and handles all the request cleanup in bfq_finish_requeue_request() called from blk_mq_free_request(). That is currently safe only because blk_mq_free_request() is called shortly after bfq_requests_merged() while bfqd->lock is still held. However to fix a lock inversion between bfqd->lock and ioc->lock, we need to call blk_mq_free_request() after dropping bfqd->lock. That would mean that already merged request could be seen by other processes inside bfq queues and possibly dispatched to the device which is wrong. So move cleanup of the request from bfq_finish_requeue_request() to bfq_requests_merged(). Acked-by: Paolo Valente <paolo.valente(a)linaro.org> Signed-off-by: Jan Kara <jack(a)suse.cz> Link: https://lore.kernel.org/r/20210623093634.27879-2-jack@suse.cz Signed-off-by: Jens Axboe <axboe(a)kernel.dk> conflict: in bfq_finish_requeue_request, code in hulk have the line atomic_dec(&rq->mq_hctx->elevator_queued); that is conflicted; Signed-off-by: zhangwensheng <zhangwensheng5(a)huawei.com> Reviewed-by: qiulaibin <qiulaibin(a)huawei.com> Reviewed-by: Jason Yan <yanaijie(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> --- block/bfq-iosched.c | 41 +++++++++++++---------------------------- 1 file changed, 13 insertions(+), 28 deletions(-) diff --git a/block/bfq-iosched.c b/block/bfq-iosched.c index fd3c23d516b8..27e01b4cd528 100644 --- a/block/bfq-iosched.c +++ b/block/bfq-iosched.c @@ -2326,7 +2326,7 @@ static void bfq_requests_merged(struct request_queue *q, struct request *rq, *next_bfqq = bfq_init_rq(next); if (!bfqq) - return; + goto remove; /* * If next and rq belong to the same bfq_queue and next is older @@ -2349,6 +2349,14 @@ static void bfq_requests_merged(struct request_queue *q, struct request *rq, bfqq->next_rq = rq; bfqg_stats_update_io_merged(bfqq_group(bfqq), next->cmd_flags); +remove: + /* Merged request may be in the IO scheduler. Remove it. */ + if (!RB_EMPTY_NODE(&next->rb_node)) { + bfq_remove_request(next->q, next); + if (next_bfqq) + bfqg_stats_update_io_remove(bfqq_group(next_bfqq), + next->cmd_flags); + } } /* Must be called with bfqq != NULL */ @@ -5901,6 +5909,7 @@ static void bfq_finish_requeue_request(struct request *rq) { struct bfq_queue *bfqq = RQ_BFQQ(rq); struct bfq_data *bfqd; + unsigned long flags; /* * rq either is not associated with any icq, or is an already @@ -5918,40 +5927,16 @@ static void bfq_finish_requeue_request(struct request *rq) rq->io_start_time_ns, rq->cmd_flags); + spin_lock_irqsave(&bfqd->lock, flags); if (likely(rq->rq_flags & RQF_STARTED)) { - unsigned long flags; - - spin_lock_irqsave(&bfqd->lock, flags); - if (rq == bfqd->waited_rq) bfq_update_inject_limit(bfqd, bfqq); bfq_completed_request(bfqq, bfqd); - bfq_finish_requeue_request_body(bfqq); atomic_dec(&rq->mq_hctx->elevator_queued); - - spin_unlock_irqrestore(&bfqd->lock, flags); - } else { - /* - * Request rq may be still/already in the scheduler, - * in which case we need to remove it (this should - * never happen in case of requeue). And we cannot - * defer such a check and removal, to avoid - * inconsistencies in the time interval from the end - * of this function to the start of the deferred work. - * This situation seems to occur only in process - * context, as a consequence of a merge. In the - * current version of the code, this implies that the - * lock is held. - */ - - if (!RB_EMPTY_NODE(&rq->rb_node)) { - bfq_remove_request(rq->q, rq); - bfqg_stats_update_io_remove(bfqq_group(bfqq), - rq->cmd_flags); - } - bfq_finish_requeue_request_body(bfqq); } + bfq_finish_requeue_request_body(bfqq); + spin_unlock_irqrestore(&bfqd->lock, flags); /* * Reset private fields. In case of a requeue, this allows -- 2.20.1

1 6

[PATCH openEuler-1.0-LTS v2 0/6] Add the no_hash_pointers startup parameter
by He Jinjin 10 Dec '21

10 Dec '21

The address printed by %p in the kernel will expose the kernel address information, which is extremely unsafe. So Linux v4.15 limited the information printed by %p which will print a hashed value. This patchset add no_hash_pointers startup parameter which can disable the restriction that %P only prints hashed values, so that %P can print the actual address in the kernel. I patched this function and the test modules associated with this and passed these tests after recompiling. Tobin C. Harding (3): lib/test_printf: Add empty module_exit function kselftest: Add test module framework header lib: Use new kselftest header Timur Tabi(3): kselftest: add support for skipped tests lib/vsprintf: no_hash_pointers prints all addresses as unhashed lib: use KSTM_MODULE_GLOBALS macro in kselftest drivers .../admin-guide/kernel-parameters.txt | 15 +++ Documentation/dev-tools/kselftest.rst | 94 +++++++++++++++++- lib/test_bitmap.c | 23 +---- lib/test_printf.c | 29 +++--- lib/vsprintf.c | 36 ++++++- tools/testing/selftests/kselftest_module.h | 54 ++++++++++ 6 files changed, 215 insertions(+), 36 deletions(-) create mode 100644 tools/testing/selftests/kselftest_module.h -- 2.30.0

1 6

[PATCH openEuler-1.0-LTS] block, bfq: move bfqq to root_group if parent group is offlined
by Yang Yingliang 10 Dec '21

10 Dec '21

From: Yu Kuai <yukuai3(a)huawei.com> hulk inclusion category: bugfix bugzilla: 185863 CVE: NA --------------------------- Our test report a uaf problem: [ 154.237639] ================================================================== [ 154.239896] BUG: KASAN: use-after-free in __bfq_deactivate_entity+0x25/0x290 [ 154.241910] Read of size 1 at addr ffff88824501f7b8 by task rmmod/2447 [ 154.244248] CPU: 7 PID: 2447 Comm: rmmod Not tainted 4.19.90+ #1 [ 154.245962] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.13.0-0-gf21b5a4aeb02-prebuilt.qemu.org 04/01/2014 [ 154.248184] Call Trace: [ 154.248532] dump_stack+0x7a/0xac [ 154.248995] print_address_description+0x6c/0x237 [ 154.249649] ? __bfq_deactivate_entity+0x25/0x290 [ 154.250297] kasan_report.cold+0x88/0x29c [ 154.250853] __bfq_deactivate_entity+0x25/0x290 [ 154.251483] bfq_pd_offline+0x13e/0x790 [ 154.252017] ? blk_mq_freeze_queue_wait+0x165/0x180 [ 154.252687] ? bfq_reparent_leaf_entity+0xa0/0xa0 [ 154.253333] ? bfq_put_queue+0x12c/0x1e0 [ 154.253877] ? kmem_cache_free+0x8e/0x1e0 [ 154.254433] ? hrtimer_active+0x53/0xa0 [ 154.254966] ? hrtimer_try_to_cancel+0x6d/0x1c0 [ 154.255576] ? __hrtimer_get_remaining+0xf0/0xf0 [ 154.256197] ? __bfq_deactivate_entity+0x11b/0x290 [ 154.256843] blkcg_deactivate_policy+0x106/0x1f0 [ 154.257464] bfq_exit_queue+0xf1/0x110 [ 154.257975] blk_mq_exit_sched+0x114/0x140 [ 154.258530] elevator_exit+0x9a/0xa0 [ 154.259023] blk_exit_queue+0x3d/0x70 [ 154.259523] blk_cleanup_queue+0x160/0x1e0 [ 154.260099] null_del_dev+0xda/0x1f0 [null_blk] [ 154.260723] null_exit+0x5f/0xab [null_blk] [ 154.261298] __x64_sys_delete_module+0x20e/0x2f0 [ 154.261931] ? __ia32_sys_delete_module+0x2f0/0x2f0 [ 154.262597] ? exit_to_usermode_loop+0x45/0xe0 [ 154.263219] do_syscall_64+0x73/0x280 [ 154.263731] ? page_fault+0x8/0x30 [ 154.264197] entry_SYSCALL_64_after_hwframe+0x44/0xa9 [ 154.264882] RIP: 0033:0x7f033bf63acb [ 154.265370] Code: 73 01 c3 48 8b 0d bd 33 0c 00 f7 d8 64 89 01 48 83 c8 ff c3 66 2e 0f 1f 84 00 00 00 00 00 90 f3 0f 1e fa b8 b0 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 8d 33 0c 00 f7 d8 64 89 01 48 [ 154.267880] RSP: 002b:00007ffc7fe52548 EFLAGS: 00000206 ORIG_RAX: 00000000000000b0 [ 154.268900] RAX: ffffffffffffffda RBX: 00005583e2b8e530 RCX: 00007f033bf63acb [ 154.269865] RDX: 000000000000000a RSI: 0000000000000800 RDI: 00005583e2b8e598 [ 154.270837] RBP: 00007ffc7fe525a8 R08: 0000000000000000 R09: 0000000000000000 [ 154.271802] R10: 00007f033bfd7ac0 R11: 0000000000000206 R12: 00007ffc7fe52770 [ 154.272763] R13: 00007ffc7fe536f8 R14: 00005583e2b8d2a0 R15: 00005583e2b8e530 [ 154.273939] Allocated by task 2350: [ 154.274419] kasan_kmalloc+0xc6/0xe0 [ 154.274916] kmem_cache_alloc_node_trace+0x119/0x240 [ 154.275594] bfq_pd_alloc+0x50/0x510 [ 154.276081] blkg_alloc+0x237/0x310 [ 154.276557] blkg_create+0x48a/0x5e0 [ 154.277044] blkg_lookup_create+0x144/0x1c0 [ 154.277614] generic_make_request_checks+0x5cf/0xad0 [ 154.278290] generic_make_request+0xdd/0x6c0 [ 154.278877] submit_bio+0xaa/0x250 [ 154.279342] mpage_readpages+0x2a2/0x3b0 [ 154.279878] read_pages+0xdf/0x3a0 [ 154.280343] __do_page_cache_readahead+0x27c/0x2a0 [ 154.280989] ondemand_readahead+0x275/0x460 [ 154.281556] generic_file_read_iter+0xc4e/0x1790 [ 154.282182] aio_read+0x174/0x260 [ 154.282635] io_submit_one+0x7d4/0x14b0 [ 154.283164] __x64_sys_io_submit+0x102/0x230 [ 154.283749] do_syscall_64+0x73/0x280 [ 154.284250] entry_SYSCALL_64_after_hwframe+0x44/0xa9 [ 154.285159] Freed by task 2315: [ 154.285588] __kasan_slab_free+0x12f/0x180 [ 154.286150] kfree+0xab/0x1d0 [ 154.286561] blkg_free.part.0+0x4a/0xe0 [ 154.287089] rcu_process_callbacks+0x424/0x6d0 [ 154.287689] __do_softirq+0x10d/0x370 [ 154.288395] The buggy address belongs to the object at ffff88824501f700 which belongs to the cache kmalloc-2048 of size 2048 [ 154.290083] The buggy address is located 184 bytes inside of 2048-byte region [ffff88824501f700, ffff88824501ff00) [ 154.291661] The buggy address belongs to the page: [ 154.292306] page:ffffea0009140600 count:1 mapcount:0 mapping:ffff88824bc0e800 index:0x0 compound_mapcount: 0 [ 154.293610] flags: 0x17ffffc0008100(slab|head) [ 154.294211] raw: 0017ffffc0008100 ffffea000896da00 0000000200000002 ffff88824bc0e800 [ 154.295247] raw: 0000000000000000 00000000800f000f 00000001ffffffff 0000000000000000 [ 154.296294] page dumped because: kasan: bad access detected [ 154.297261] Memory state around the buggy address: [ 154.297913] ffff88824501f680: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc [ 154.298884] ffff88824501f700: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb [ 154.299858] >ffff88824501f780: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb [ 154.300824] ^ [ 154.301505] ffff88824501f800: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb [ 154.302479] ffff88824501f880: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb [ 154.303459] ================================================================== This is because when bfq_group is offlined, if the bfq_queues are not in active tree, their parents(bfqq->entity.partent) are still point to the offlined bfq_group. And after some ios are issued to such bfq_queues, the offlined bfq_group is reinserted to service tree. Fix the problem by move bfq_queue to root_group if we found it's parent is offlined. Fixes: e21b7a0b9887 ("block, bfq: add full hierarchical scheduling and cgroups support") Signed-off-by: Yu Kuai <yukuai3(a)huawei.com> Reviewed-by: Hou Tao <houtao1(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- block/bfq-cgroup.c | 14 ++++++++++---- block/bfq-wf2q.c | 9 +++++++++ 2 files changed, 19 insertions(+), 4 deletions(-) diff --git a/block/bfq-cgroup.c b/block/bfq-cgroup.c index 78cfd008b89d7..73b82a5c03717 100644 --- a/block/bfq-cgroup.c +++ b/block/bfq-cgroup.c @@ -556,6 +556,7 @@ void bfq_bfqq_move(struct bfq_data *bfqd, struct bfq_queue *bfqq, struct bfq_group *bfqg) { struct bfq_entity *entity = &bfqq->entity; + struct bfq_group *old_parent = bfqq_group(bfqq); /* * Get extra reference to prevent bfqq from being freed in @@ -577,17 +578,21 @@ void bfq_bfqq_move(struct bfq_data *bfqd, struct bfq_queue *bfqq, bfq_deactivate_bfqq(bfqd, bfqq, false, false); else if (entity->on_st) bfq_put_idle_entity(bfq_entity_service_tree(entity), entity); - bfqg_and_blkg_put(bfqq_group(bfqq)); entity->parent = bfqg->my_entity; entity->sched_data = &bfqg->sched_data; /* pin down bfqg and its associated blkg */ bfqg_and_blkg_get(bfqg); - if (bfq_bfqq_busy(bfqq)) { - bfq_pos_tree_add_move(bfqd, bfqq); + /* + * Don't leave the bfqq->pos_root to old bfqg, since the ref to old + * bfqg will be released and the bfqg might be freed. + */ + bfq_pos_tree_add_move(bfqd, bfqq); + bfqg_and_blkg_put(old_parent); + + if (bfq_bfqq_busy(bfqq)) bfq_activate_bfqq(bfqd, bfqq); - } if (!bfqd->in_service_queue && !bfqd->rq_in_driver) bfq_schedule_dispatch(bfqd); @@ -860,6 +865,7 @@ static void bfq_pd_offline(struct blkg_policy_data *pd) put_async_queues: bfq_put_async_queues(bfqd, bfqg); + pd->plid = BLKCG_MAX_POLS; spin_unlock_irqrestore(&bfqd->lock, flags); /* diff --git a/block/bfq-wf2q.c b/block/bfq-wf2q.c index 316a1c2d1b610..e830715fe15d6 100644 --- a/block/bfq-wf2q.c +++ b/block/bfq-wf2q.c @@ -1684,6 +1684,15 @@ void bfq_del_bfqq_busy(struct bfq_data *bfqd, struct bfq_queue *bfqq, */ void bfq_add_bfqq_busy(struct bfq_data *bfqd, struct bfq_queue *bfqq) { +#ifdef CONFIG_BFQ_GROUP_IOSCHED + /* If parent group is offlined, move the bfqq to root group */ + if (bfqq->entity.parent) { + struct bfq_group *bfqg = bfq_bfqq_to_bfqg(bfqq); + + if (bfqg->pd.plid >= BLKCG_MAX_POLS) + bfq_bfqq_move(bfqd, bfqq, bfqd->root_group); + } +#endif bfq_log_bfqq(bfqd, bfqq, "add to busy"); bfq_activate_bfqq(bfqd, bfqq); -- 2.25.1

1 0

openEuler kernel技术分享第十六期 & 双周例会
by openEuler conference 10 Dec '21

10 Dec '21

您好！ Kernel SIG 邀请您参加 2021-12-17 14:00 召开的ZOOM会议(自动录制) 会议主题：openEuler kernel技术分享第十六期 & 双周例会会议内容： 14:10-15:30 openEuler kernel 技术分享第16期-openEuler内核热补丁介绍会议链接：https://us06web.zoom.us/j/88618073121?pwd=V3JyZG5IOWRVWENieDJhNndpV2x5UT09 温馨提醒：建议接入会议后修改参会人的姓名，也可以使用您在gitee.com的ID 更多资讯尽在：https://openeuler.org/zh/ Hello! openEuler Kernel SIG invites you to attend the ZOOM conference(auto recording) will be held at 2021-12-17 14:00, The subject of the conference is openEuler kernel技术分享第十六期 & 双周例会, Summary: 14:10-15:30 openEuler kernel 技术分享第16期-openEuler内核热补丁介绍 You can join the meeting at https://us06web.zoom.us/j/88618073121?pwd=V3JyZG5IOWRVWENieDJhNndpV2x5UT09. Note: You are advised to change the participant name after joining the conference or use your ID at gitee.com. More information: https://openeuler.org/en/

1 0

[PATCH openEuler-1.0-LTS 1/2] io_uring: fix soft lockup when call __io_remove_buffers
by Yang Yingliang 09 Dec '21

09 Dec '21

From: Ye Bin <yebin10(a)huawei.com> mainline inclusion from mainline-v5.16-rc3 commit 1d0254e6b47e73222fd3d6ae95cccbaafe5b3ecf category: bugfix bugzilla: 185746 CVE: NA ----------------------------------------------- I got issue as follows: [ 567.094140] __io_remove_buffers: [1]start ctx=0xffff8881067bf000 bgid=65533 buf=0xffff8881fefe1680 [ 594.360799] watchdog: BUG: soft lockup - CPU#2 stuck for 26s! [kworker/u32:5:108] [ 594.364987] Modules linked in: [ 594.365405] irq event stamp: 604180238 [ 594.365906] hardirqs last enabled at (604180237): [<ffffffff93fec9bd>] _raw_spin_unlock_irqrestore+0x2d/0x50 [ 594.367181] hardirqs last disabled at (604180238): [<ffffffff93fbbadb>] sysvec_apic_timer_interrupt+0xb/0xc0 [ 594.368420] softirqs last enabled at (569080666): [<ffffffff94200654>] __do_softirq+0x654/0xa9e [ 594.369551] softirqs last disabled at (569080575): [<ffffffff913e1d6a>] irq_exit_rcu+0x1ca/0x250 [ 594.370692] CPU: 2 PID: 108 Comm: kworker/u32:5 Tainted: G L 5.15.0-next-20211112+ #88 [ 594.371891] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS ?-20190727_073836-buildvm-ppc64le-16.ppc.fedoraproject.org-3.fc31 04/01/2014 [ 594.373604] Workqueue: events_unbound io_ring_exit_work [ 594.374303] RIP: 0010:_raw_spin_unlock_irqrestore+0x33/0x50 [ 594.375037] Code: 48 83 c7 18 53 48 89 f3 48 8b 74 24 10 e8 55 f5 55 fd 48 89 ef e8 ed a7 56 fd 80 e7 02 74 06 e8 43 13 7b fd fb bf 01 00 00 00 <e8> f8 78 474 [ 594.377433] RSP: 0018:ffff888101587a70 EFLAGS: 00000202 [ 594.378120] RAX: 0000000024030f0d RBX: 0000000000000246 RCX: 1ffffffff2f09106 [ 594.379053] RDX: 0000000000000000 RSI: ffffffff9449f0e0 RDI: 0000000000000001 [ 594.379991] RBP: ffffffff9586cdc0 R08: 0000000000000001 R09: fffffbfff2effcab [ 594.380923] R10: ffffffff977fe557 R11: fffffbfff2effcaa R12: ffff8881b8f3def0 [ 594.381858] R13: 0000000000000246 R14: ffff888153a8b070 R15: 0000000000000000 [ 594.382787] FS: 0000000000000000(0000) GS:ffff888399c00000(0000) knlGS:0000000000000000 [ 594.383851] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 594.384602] CR2: 00007fcbe71d2000 CR3: 00000000b4216000 CR4: 00000000000006e0 [ 594.385540] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 594.386474] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 594.387403] Call Trace: [ 594.387738] <TASK> [ 594.388042] find_and_remove_object+0x118/0x160 [ 594.389321] delete_object_full+0xc/0x20 [ 594.389852] kfree+0x193/0x470 [ 594.390275] __io_remove_buffers.part.0+0xed/0x147 [ 594.390931] io_ring_ctx_free+0x342/0x6a2 [ 594.392159] io_ring_exit_work+0x41e/0x486 [ 594.396419] process_one_work+0x906/0x15a0 [ 594.399185] worker_thread+0x8b/0xd80 [ 594.400259] kthread+0x3bf/0x4a0 [ 594.401847] ret_from_fork+0x22/0x30 [ 594.402343] </TASK> Message from syslogd@localhost at Nov 13 09:09:54 ... kernel:watchdog: BUG: soft lockup - CPU#2 stuck for 26s! [kworker/u32:5:108] [ 596.793660] __io_remove_buffers: [2099199]start ctx=0xffff8881067bf000 bgid=65533 buf=0xffff8881fefe1680 We can reproduce this issue by follow syzkaller log: r0 = syz_io_uring_setup(0x401, &(0x7f0000000300), &(0x7f0000003000/0x2000)=nil, &(0x7f0000ff8000/0x4000)=nil, &(0x7f0000000280)=<r1=>0x0, &(0x7f0000000380)=<r2=>0x0) sendmsg$ETHTOOL_MSG_FEATURES_SET(0xffffffffffffffff, &(0x7f0000003080)={0x0, 0x0, &(0x7f0000003040)={&(0x7f0000000040)=ANY=[], 0x18}}, 0x0) syz_io_uring_submit(r1, r2, &(0x7f0000000240)=@IORING_OP_PROVIDE_BUFFERS={0x1f, 0x5, 0x0, 0x401, 0x1, 0x0, 0x100, 0x0, 0x1, {0xfffd}}, 0x0) io_uring_enter(r0, 0x3a2d, 0x0, 0x0, 0x0, 0x0) The reason above issue is 'buf->list' has 2,100,000 nodes, occupied cpu lead to soft lockup. To solve this issue, we need add schedule point when do while loop in '__io_remove_buffers'. After add schedule point we do regression, get follow data. [ 240.141864] __io_remove_buffers: [1]start ctx=0xffff888170603000 bgid=65533 buf=0xffff8881116fcb00 [ 268.408260] __io_remove_buffers: [1]start ctx=0xffff8881b92d2000 bgid=65533 buf=0xffff888130c83180 [ 275.899234] __io_remove_buffers: [2099199]start ctx=0xffff888170603000 bgid=65533 buf=0xffff8881116fcb00 [ 296.741404] __io_remove_buffers: [1]start ctx=0xffff8881b659c000 bgid=65533 buf=0xffff8881010fe380 [ 305.090059] __io_remove_buffers: [2099199]start ctx=0xffff8881b92d2000 bgid=65533 buf=0xffff888130c83180 [ 325.415746] __io_remove_buffers: [1]start ctx=0xffff8881b92d1000 bgid=65533 buf=0xffff8881a17d8f00 [ 333.160318] __io_remove_buffers: [2099199]start ctx=0xffff8881b659c000 bgid=65533 buf=0xffff8881010fe380 ... Fixes:8bab4c09f24e("io_uring: allow conditional reschedule for intensive iterators") Signed-off-by: Ye Bin <yebin10(a)huawei.com> Link: https://lore.kernel.org/r/20211122024737.2198530-1-yebin10@huawei.com Signed-off-by: Jens Axboe <axboe(a)kernel.dk> conflicts: fs/io_uring.c Signed-off-by: Ye Bin <yebin10(a)huawei.com> Reviewed-by: Zhang Yi <yi.zhang(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- fs/io_uring.c | 1 + 1 file changed, 1 insertion(+) diff --git a/fs/io_uring.c b/fs/io_uring.c index d07388600bbed..f3b5f9d670df3 100644 --- a/fs/io_uring.c +++ b/fs/io_uring.c @@ -3462,6 +3462,7 @@ static int __io_remove_buffers(struct io_ring_ctx *ctx, struct io_buffer *buf, kfree(nxt); if (++i == nbufs) return i; + cond_resched(); } i++; kfree(buf); -- 2.25.1

1 1

[PATCH openEuler-1.0-LTS] block: Fix fsync always failed if once failed
by Yang Yingliang 09 Dec '21

09 Dec '21

From: Ye Bin <yebin10(a)huawei.com> mainline inclusion from mainline-v5.17 commit 8a7518931baa8ea023700987f3db31cb0a80610b category: bugfix bugzilla: 185819 CVE: NA ----------------------------------------------- We do test with inject error fault base on v4.19, after test some time we found sync /dev/sda always failed. [root@localhost] sync /dev/sda sync: error syncing '/dev/sda': Input/output error scsi log as follows: [19069.812296] sd 0:0:0:0: [sda] tag#64 Send: scmd 0x00000000d03a0b6b [19069.812302] sd 0:0:0:0: [sda] tag#64 CDB: Synchronize Cache(10) 35 00 00 00 00 00 00 00 00 00 [19069.812533] sd 0:0:0:0: [sda] tag#64 Done: SUCCESS Result: hostbyte=DID_OK driverbyte=DRIVER_OK [19069.812536] sd 0:0:0:0: [sda] tag#64 CDB: Synchronize Cache(10) 35 00 00 00 00 00 00 00 00 00 [19069.812539] sd 0:0:0:0: [sda] tag#64 scsi host busy 1 failed 0 [19069.812542] sd 0:0:0:0: Notifying upper driver of completion (result 0) [19069.812546] sd 0:0:0:0: [sda] tag#64 sd_done: completed 0 of 0 bytes [19069.812549] sd 0:0:0:0: [sda] tag#64 0 sectors total, 0 bytes done. [19069.812564] print_req_error: I/O error, dev sda, sector 0 ftrace log as follows: rep-306069 [007] .... 19654.923315: block_bio_queue: 8,0 FWS 0 + 0 [rep] rep-306069 [007] .... 19654.923333: block_getrq: 8,0 FWS 0 + 0 [rep] kworker/7:1H-250 [007] .... 19654.923352: block_rq_issue: 8,0 FF 0 () 0 + 0 [kworker/7:1H] <idle>-0 [007] ..s. 19654.923562: block_rq_complete: 8,0 FF () 18446744073709551615 + 0 [0] <idle>-0 [007] d.s. 19654.923576: block_rq_complete: 8,0 WS () 0 + 0 [-5] As 8d6996630c03 introduce 'fq->rq_status', this data only update when 'flush_rq' reference count isn't zero. If flush request once failed and record error code in 'fq->rq_status'. If there is no chance to update 'fq->rq_status',then do fsync will always failed. To address this issue reset 'fq->rq_status' after return error code to upper layer. Fixes: 8d6996630c03("block: fix null pointer dereference in blk_mq_rq_timed_out()") Signed-off-by: Ye Bin <yebin10(a)huawei.com> Reviewed-by: Ming Lei <ming.lei(a)redhat.com> Link: https://lore.kernel.org/r/20211129012659.1553733-1-yebin10@huawei.com Signed-off-by: Jens Axboe <axboe(a)kernel.dk> conflicts: block/blk-flush.c Signed-off-by: Ye Bin <yebin10(a)huawei.com> Reviewed-by: Jason Yan <yanaijie(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- block/blk-flush.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/block/blk-flush.c b/block/blk-flush.c index 022a2bbb012d7..c1bfcde165af5 100644 --- a/block/blk-flush.c +++ b/block/blk-flush.c @@ -245,8 +245,10 @@ static void flush_end_io(struct request *flush_rq, blk_status_t error) * avoiding use-after-free. */ WRITE_ONCE(flush_rq->state, MQ_RQ_IDLE); - if (fq->rq_status != BLK_STS_OK) + if (fq->rq_status != BLK_STS_OK) { error = fq->rq_status; + fq->rq_status = BLK_STS_OK; + } hctx = blk_mq_map_queue(q, flush_rq->mq_ctx->cpu); if (!q->elevator) { -- 2.25.1

1 0

[PATCH openEuler-1.0-LTS 1/3] blk-mq: factor out some helps to quiesce/unquiesce queue
by Yang Yingliang 09 Dec '21

09 Dec '21

From: Yu Kuai <yukuai3(a)huawei.com> hulk inclusion category: bugfix bugzilla: 173974 CVE: NA --------------------------- Prepare to support concurrent quiesce queue between drivers and block layer, no functional changes. Signed-off-by: Yu Kuai <yukuai3(a)huawei.com> Reviewed-by: Hou Tao <houtao1(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- block/blk-mq.c | 58 ++++++++++++++++++++++++++++++++++---------------- 1 file changed, 40 insertions(+), 18 deletions(-) diff --git a/block/blk-mq.c b/block/blk-mq.c index ef62a83314a5d..f9b4b73a2f38d 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -211,32 +211,29 @@ void blk_mq_unfreeze_queue(struct request_queue *q) } EXPORT_SYMBOL_GPL(blk_mq_unfreeze_queue); +static void __blk_mq_quiesce_queue_nowait(struct request_queue *q, + unsigned int flag) +{ + blk_queue_flag_set(flag, q); +} + /* * FIXME: replace the scsi_internal_device_*block_nowait() calls in the * mpt3sas driver such that this function can be removed. */ void blk_mq_quiesce_queue_nowait(struct request_queue *q) { - blk_queue_flag_set(QUEUE_FLAG_QUIESCED, q); + __blk_mq_quiesce_queue_nowait(q, QUEUE_FLAG_QUIESCED); } EXPORT_SYMBOL_GPL(blk_mq_quiesce_queue_nowait); -/** - * blk_mq_quiesce_queue() - wait until all ongoing dispatches have finished - * @q: request queue. - * - * Note: this function does not prevent that the struct request end_io() - * callback function is invoked. Once this function is returned, we make - * sure no dispatch can happen until the queue is unquiesced via - * blk_mq_unquiesce_queue(). - */ -void blk_mq_quiesce_queue(struct request_queue *q) +static void __blk_mq_quiesce_queue(struct request_queue *q, unsigned int flag) { struct blk_mq_hw_ctx *hctx; unsigned int i; bool rcu = false; - blk_mq_quiesce_queue_nowait(q); + __blk_mq_quiesce_queue_nowait(q, flag); queue_for_each_hw_ctx(q, hctx, i) { if (hctx->flags & BLK_MQ_F_BLOCKING) @@ -247,15 +244,30 @@ void blk_mq_quiesce_queue(struct request_queue *q) if (rcu) synchronize_rcu(); } + +/** + * blk_mq_quiesce_queue() - wait until all ongoing dispatches have finished + * @q: request queue. + * + * Note: this function does not prevent that the struct request end_io() + * callback function is invoked. Once this function is returned, we make + * sure no dispatch can happen until the queue is unquiesced via + * blk_mq_unquiesce_queue(). + */ +void blk_mq_quiesce_queue(struct request_queue *q) +{ + __blk_mq_quiesce_queue(q, QUEUE_FLAG_QUIESCED); +} EXPORT_SYMBOL_GPL(blk_mq_quiesce_queue); -bool blk_mq_quiesce_queue_without_rcu(struct request_queue *q) +static bool __blk_mq_quiesce_queue_without_rcu(struct request_queue *q, + unsigned int flag) { struct blk_mq_hw_ctx *hctx; unsigned int i; bool rcu = false; - blk_mq_quiesce_queue_nowait(q); + __blk_mq_quiesce_queue_nowait(q, flag); queue_for_each_hw_ctx(q, hctx, i) { if (hctx->flags & BLK_MQ_F_BLOCKING) @@ -265,8 +277,21 @@ bool blk_mq_quiesce_queue_without_rcu(struct request_queue *q) } return rcu; } + +bool blk_mq_quiesce_queue_without_rcu(struct request_queue *q) +{ + return __blk_mq_quiesce_queue_without_rcu(q, QUEUE_FLAG_QUIESCED); +} EXPORT_SYMBOL_GPL(blk_mq_quiesce_queue_without_rcu); +static void __blk_mq_unquiesce_queue(struct request_queue *q, unsigned int flag) +{ + blk_queue_flag_clear(flag, q); + + /* dispatch requests which are inserted during quiescing */ + blk_mq_run_hw_queues(q, true); +} + /* * blk_mq_unquiesce_queue() - counterpart of blk_mq_quiesce_queue() * @q: request queue. @@ -276,10 +301,7 @@ EXPORT_SYMBOL_GPL(blk_mq_quiesce_queue_without_rcu); */ void blk_mq_unquiesce_queue(struct request_queue *q) { - blk_queue_flag_clear(QUEUE_FLAG_QUIESCED, q); - - /* dispatch requests which are inserted during quiescing */ - blk_mq_run_hw_queues(q, true); + __blk_mq_unquiesce_queue(q, QUEUE_FLAG_QUIESCED); } EXPORT_SYMBOL_GPL(blk_mq_unquiesce_queue); -- 2.25.1

1 2

[PATCH openEuler-1.0-LTS 0/6] Add the no_hash_pointers startup parameter
by He Jinjin 09 Dec '21

09 Dec '21

The address printed by %p in the kernel will expose the kernel address information, which is extremely unsafe. So Linux v4.15 limited the information printed by %p which will print a hashed value. This patchset add no_hash_pointers startup parameter which can disable the restriction that %P only prints hashed values, so that %P can print the actual address in the kernel. I patched this function and the test modules associated with this and passed these tests after recompiling. Tobin C. Harding (3): lib/test_printf: Add empty module_exit function kselftest: Add test module framework header lib: Use new kselftest header Timur Tabi(3): kselftest: add support for skipped tests lib: use KSTM_MODULE_GLOBALS macro in kselftest drivers lib/vsprintf: no_hash_pointers prints all addresses as unhashed .../admin-guide/kernel-parameters.txt | 15 +++ Documentation/dev-tools/kselftest.rst | 94 +++++++++++++++++- lib/test_bitmap.c | 23 +---- lib/test_printf.c | 29 +++--- lib/vsprintf.c | 36 ++++++- tools/testing/selftests/kselftest_module.h | 54 ++++++++++ 6 files changed, 215 insertions(+), 36 deletions(-) create mode 100644 tools/testing/selftests/kselftest_module.h -- 2.30.0

1 6

[PATCH openEuler-1.0-LTS 0/6] Add the no_hash_pointers startup parameter
by He Jinjin 09 Dec '21

09 Dec '21

The address printed by %p in the kernel will expose the kernel address information, which is extremely unsafe. So Linux v4.15 limited the information printed by %p which will print a hashed value. This patchset add no_hash_pointers startup parameter which can disable the restriction that %P only prints hashed values, so that %P can print the actual address in the kernel. I patched this function and the test modules associated with this and passed these tests after recompiling. Tobin C. Harding (3): lib/test_printf: Add empty module_exit function kselftest: Add test module framework header lib: Use new kselftest header Timur Tabi(3): kselftest: add support for skipped tests lib: use KSTM_MODULE_GLOBALS macro in kselftest drivers lib/vsprintf: no_hash_pointers prints all addresses as unhashed .../admin-guide/kernel-parameters.txt | 15 +++ Documentation/dev-tools/kselftest.rst | 94 +++++++++++++++++- lib/test_bitmap.c | 23 +---- lib/test_printf.c | 29 +++--- lib/vsprintf.c | 36 ++++++- tools/testing/selftests/kselftest_module.h | 54 ++++++++++ 6 files changed, 215 insertions(+), 36 deletions(-) create mode 100644 tools/testing/selftests/kselftest_module.h -- 2.30.0

1 6

[PATCH OLK-5.10 0/2] Add the function to modify the reboot mode
by Hongyu Li 09 Dec '21

09 Dec '21

The current kernel needs to provide the reboot mode in the boot parameter. However, we can not know this in advance. The kernel should have a method to set the reboot options when the system is booted. This patchset add handles in the <sysfs>/kernel/reboot to read reboot configuration and rewrite it. Therefore users can change the reboot mode as they want. Matteo Croce (1): reboot: allow to specify reboot mode via sysfs Nathan Chancellor (1): reboot: Fix variable assignments in type_store Documentation/ABI/testing/sysfs-kernel-reboot | 32 +++ kernel/reboot.c | 206 ++++++++++++++++++ 2 files changed, 238 insertions(+) create mode 100644 Documentation/ABI/testing/sysfs-kernel-reboot -- 2.17.1

1 0

[PATCH OLK-5.10 0/2] Add the function to modify the reboot mode
by Hongyu Li 09 Dec '21

09 Dec '21

The current kernel needs to provide the reboot mode in the boot parameter. However, we can not know this in advance. The kernel should have a method to set the reboot options when the system is booted. This patchset add handles in the <sysfs>/kernel/reboot to read reboot configuration and rewrite it. Therefore users can change the reboot mode as they want. Matteo Croce (1): reboot: allow to specify reboot mode via sysfs Nathan Chancellor (1): reboot: Fix variable assignments in type_store Documentation/ABI/testing/sysfs-kernel-reboot | 32 +++ kernel/reboot.c | 206 ++++++++++++++++++ 2 files changed, 238 insertions(+) create mode 100644 Documentation/ABI/testing/sysfs-kernel-reboot -- 2.17.1

1 0

[PATCH openEuler-1.0-LTS 1/2] bfq: Remove merged request already in bfq_requests_merged()
by Yang Yingliang 09 Dec '21

09 Dec '21

From: Jan Kara <jack(a)suse.cz> mainline inclusion from mainline-v5.14-rc1 commit a921c655f2033dd1ce1379128efe881dda23ea37 category: bugfix bugzilla: 185777, 185811 CVE: NA Currently, bfq does very little in bfq_requests_merged() and handles all the request cleanup in bfq_finish_requeue_request() called from blk_mq_free_request(). That is currently safe only because blk_mq_free_request() is called shortly after bfq_requests_merged() while bfqd->lock is still held. However to fix a lock inversion between bfqd->lock and ioc->lock, we need to call blk_mq_free_request() after dropping bfqd->lock. That would mean that already merged request could be seen by other processes inside bfq queues and possibly dispatched to the device which is wrong. So move cleanup of the request from bfq_finish_requeue_request() to bfq_requests_merged(). Acked-by: Paolo Valente <paolo.valente(a)linaro.org> Signed-off-by: Jan Kara <jack(a)suse.cz> Link: https://lore.kernel.org/r/20210623093634.27879-2-jack@suse.cz Signed-off-by: Jens Axboe <axboe(a)kernel.dk> conflict: in bfq_finish_requeue_request, 4.19 not have bfq_update_inject_limit branch; Signed-off-by: zhangwensheng <zhangwensheng5(a)huawei.com> Reviewed-by: Jason Yan <yanaijie(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- block/bfq-iosched.c | 41 +++++++++++++---------------------------- 1 file changed, 13 insertions(+), 28 deletions(-) diff --git a/block/bfq-iosched.c b/block/bfq-iosched.c index 7d77de9a0f5c0..5452d892480ba 100644 --- a/block/bfq-iosched.c +++ b/block/bfq-iosched.c @@ -1937,7 +1937,7 @@ static void bfq_requests_merged(struct request_queue *q, struct request *rq, *next_bfqq = bfq_init_rq(next); if (!bfqq) - return; + goto remove; /* * If next and rq belong to the same bfq_queue and next is older @@ -1960,6 +1960,14 @@ static void bfq_requests_merged(struct request_queue *q, struct request *rq, bfqq->next_rq = rq; bfqg_stats_update_io_merged(bfqq_group(bfqq), next->cmd_flags); +remove: + /* Merged request may be in the IO scheduler. Remove it. */ + if (!RB_EMPTY_NODE(&next->rb_node)) { + bfq_remove_request(next->q, next); + if (next_bfqq) + bfqg_stats_update_io_remove(bfqq_group(next_bfqq), + next->cmd_flags); + } } /* Must be called with bfqq != NULL */ @@ -4876,6 +4884,7 @@ static void bfq_finish_requeue_request(struct request *rq) { struct bfq_queue *bfqq = RQ_BFQQ(rq); struct bfq_data *bfqd; + unsigned long flags; /* * rq either is not associated with any icq, or is an already @@ -4893,36 +4902,12 @@ static void bfq_finish_requeue_request(struct request *rq) rq->io_start_time_ns, rq->cmd_flags); + spin_lock_irqsave(&bfqd->lock, flags); if (likely(rq->rq_flags & RQF_STARTED)) { - unsigned long flags; - - spin_lock_irqsave(&bfqd->lock, flags); - bfq_completed_request(bfqq, bfqd); - bfq_finish_requeue_request_body(bfqq); - - spin_unlock_irqrestore(&bfqd->lock, flags); - } else { - /* - * Request rq may be still/already in the scheduler, - * in which case we need to remove it (this should - * never happen in case of requeue). And we cannot - * defer such a check and removal, to avoid - * inconsistencies in the time interval from the end - * of this function to the start of the deferred work. - * This situation seems to occur only in process - * context, as a consequence of a merge. In the - * current version of the code, this implies that the - * lock is held. - */ - - if (!RB_EMPTY_NODE(&rq->rb_node)) { - bfq_remove_request(rq->q, rq); - bfqg_stats_update_io_remove(bfqq_group(bfqq), - rq->cmd_flags); - } - bfq_finish_requeue_request_body(bfqq); } + bfq_finish_requeue_request_body(bfqq); + spin_unlock_irqrestore(&bfqd->lock, flags); /* * Reset private fields. In case of a requeue, this allows -- 2.25.1

1 1

[PATCH openEuler-1.0-LTS] md: fix a warning caused by a race between concurrent md_ioctl()s
by Yang Yingliang 09 Dec '21

09 Dec '21

From: "Dae R. Jeong" <dae.r.jeong(a)kaist.ac.kr> mainline inclusion from mainline-v5.11-rc1 commit c731b84b51bf7fe83448bea8f56a6d55006b0615 category: bugfix bugzilla: 185833 CVE: NA ----------------------------------------------- Syzkaller reports a warning as belows. WARNING: CPU: 0 PID: 9647 at drivers/md/md.c:7169 ... Call Trace: ... RIP: 0010:md_ioctl+0x4017/0x5980 drivers/md/md.c:7169 RSP: 0018:ffff888096027950 EFLAGS: 00010293 RAX: ffff88809322c380 RBX: 0000000000000932 RCX: ffffffff84e266f2 RDX: 0000000000000000 RSI: ffffffff84e299f7 RDI: 0000000000000007 RBP: ffff888096027bc0 R08: ffff88809322c380 R09: ffffed101341a482 R10: ffff888096027940 R11: ffff88809a0d240f R12: 0000000000000932 R13: ffff8880a2c14100 R14: ffff88809a0d2268 R15: ffff88809a0d2408 __blkdev_driver_ioctl block/ioctl.c:304 [inline] blkdev_ioctl+0xece/0x1c10 block/ioctl.c:606 block_ioctl+0xee/0x130 fs/block_dev.c:1930 vfs_ioctl fs/ioctl.c:46 [inline] file_ioctl fs/ioctl.c:509 [inline] do_vfs_ioctl+0xd5f/0x1380 fs/ioctl.c:696 ksys_ioctl+0xab/0xd0 fs/ioctl.c:713 __do_sys_ioctl fs/ioctl.c:720 [inline] __se_sys_ioctl fs/ioctl.c:718 [inline] __x64_sys_ioctl+0x73/0xb0 fs/ioctl.c:718 do_syscall_64+0xfd/0x680 arch/x86/entry/common.c:301 entry_SYSCALL_64_after_hwframe+0x49/0xbe This is caused by a race between two concurrenct md_ioctl()s closing the array. CPU1 (md_ioctl()) CPU2 (md_ioctl()) ------ ------ set_bit(MD_CLOSING, &mddev->flags); did_set_md_closing = true; WARN_ON_ONCE(test_bit(MD_CLOSING, &mddev->flags)); if(did_set_md_closing) clear_bit(MD_CLOSING, &mddev->flags); Fix the warning by returning immediately if the MD_CLOSING bit is set in &mddev->flags which indicates that the array is being closed. Fixes: 065e519e71b2 ("md: MD_CLOSING needs to be cleared after called md_set_readonly or do_md_stop") Reported-by: syzbot+1e46a0864c1a6e9bd3d8(a)syzkaller.appspotmail.com Cc: stable(a)vger.kernel.org Signed-off-by: Dae R. Jeong <dae.r.jeong(a)kaist.ac.kr> Signed-off-by: Song Liu <songliubraving(a)fb.com> Signed-off-by: Ye Bin <yebin10(a)huawei.com> Reviewed-by: Jason Yan <yanaijie(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/md/md.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/md/md.c b/drivers/md/md.c index 8a2656cf7127d..409ec5ffd28d3 100644 --- a/drivers/md/md.c +++ b/drivers/md/md.c @@ -7306,8 +7306,11 @@ static int md_ioctl(struct block_device *bdev, fmode_t mode, err = -EBUSY; goto out; } - WARN_ON_ONCE(test_bit(MD_CLOSING, &mddev->flags)); - set_bit(MD_CLOSING, &mddev->flags); + if (test_and_set_bit(MD_CLOSING, &mddev->flags)) { + mutex_unlock(&mddev->open_mutex); + err = -EBUSY; + goto out; + } did_set_md_closing = true; mutex_unlock(&mddev->open_mutex); sync_blockdev(bdev); -- 2.25.1

1 0

[PATCH openEuler-1.0-LTS 1/2] net: hns3: fix race condition in debugfs
by Yang Yingliang 09 Dec '21

09 Dec '21

From: Yonglong Liu <liuyonglong(a)huawei.com> driver inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I4LD5U CVE: NA ---------------------------- When multiple users access debugfs at the same time, the process of alloc and release memory becomes disordered, causing the kernel crash like this: [763845.759089] PC is at kfree+0x19c/0x1a0 [763845.759100] LR is at kvfree+0x3c/0x58 [763845.759103] pc : [<ffff00000828878c>] lr : [<ffff00000823432c>] pstate: 60400009 [763845.759105] sp : ffff00003744fc90 [763845.759108] x29: ffff00003744fc90 x28: ffff8027dc87b800 [763845.759115] x27: ffff0000088a1000 x26: ffff000002970f48 [763845.759121] x25: ffff802502600000 x24: 00000000000000af [763845.759127] x23: 0000000000010000 x22: 0000000013dc0000 [763845.759133] x21: ffff00000823432c x20: ffff802502600000 [763845.759139] x19: ffff802502600000 x18: 0000ffffdaa06b10 [763845.759145] x17: 00000000004201c8 x16: ffff0000082b2b10 [763845.759151] x15: 000000000003013f x14: 0000ffffa462ffe0 [763845.759157] x13: ffffffffffffffff x12: 0433526ae61f3300 [763845.759163] x11: ffff000009694b30 x10: 0000000000000001 [763845.759169] x9 : 000000000007b224 x8 : ffff000009719edc [763845.759175] x7 : ffff7fe009409800 x6 : 00000045757af8cf [763845.759181] x5 : ffff8027fced69f0 x4 : 0000000000000000 [763845.759187] x3 : 0000000000000000 x2 : 0433526ae61f3300 [763845.759192] x1 : 0000000000000000 x0 : dead000000000100 [763845.759200] Process cat (pid: 57988, stack limit = 0xffff000037440000) [763845.759203] Call trace: [763845.759207] Exception stack(0xffff00003744fb50 to 0xffff00003744fc90) [763845.759211] fb40: dead000000000100 0000000000000000 [768745.759215] fb60: 0433526ae61f3300 0000000000000000 0000000000000000 ffff8027fced69f0 [763845.759219] fb80: 00000045757af8cf ffff7fe009409800 ffff000009719edc 000000000007b224 [763845.759222] fba0: 0000000000000001 ffff000009694b30 0433526ae61f3300 ffffffffffffffff [763845.759226] fbc0: 0000ffffa462ffe0 000000000003013f ffff0000082b2b10 00000000004201c8 [763845.759231] fbe0: 0000ffffdaa06b10 ffff802502600000 ffff802502600000 ffff00000823432c [763845.759235] fc00: 0000000013dc0000 0000000000010000 00000000000000af ffff802502600000 [763845.759238] fc20: ffff000002970f48 ffff0000088a1000 ffff8027dc87b800 ffff00003744fc90 [763845.759243] fc40: ffff00000823432c ffff00003744fc90 ffff00000828878c 0000000060400009 [763845.759247] fc60: ffff00003744feb0 0000000013dc0000 0000ffffffffffff 0000000000000023 [763845.759250] fc80: ffff00003744fc90 ffff00000828878c [763845.759259] [<ffff00000828878c>] kfree+0x19c/0x1a0 [763845.759263] [<ffff00000823432c>] kvfree+0x3c/0x58 [763845.759306] [<ffff00000295ab94>] hns3_dbg_read+0x94/0x240 [hns3] [763845.759318] [<ffff000008359550>] full_proxy_read+0x60/0x90 [763845.759324] [<ffff0000082b22a4>] __vfs_read+0x58/0x178 [763845.759327] [<ffff0000082b2454>] vfs_read+0x90/0x14c [763845.759332] [<ffff0000082b2b70>] SyS_read+0x60/0xc0 This patch adds a mutex lock to fix the race condition, and need to call hns3_dbg_read_cmd() function when buffer is NULL to avoid reading empty data. Fixes: c91910efc03a ("net: hns3: refactor the debugfs process") Signed-off-by: Yonglong Liu <liuyonglong(a)huawei.com> Reviewed-by: li yongxin <liyongxin1(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/net/ethernet/hisilicon/hns3/hnae3.h | 1 + .../ethernet/hisilicon/hns3/hns3_debugfs.c | 19 +++++++++++++------ 2 files changed, 14 insertions(+), 6 deletions(-) diff --git a/drivers/net/ethernet/hisilicon/hns3/hnae3.h b/drivers/net/ethernet/hisilicon/hns3/hnae3.h index c9ac1e7cf4492..048de5b367c19 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hnae3.h +++ b/drivers/net/ethernet/hisilicon/hns3/hnae3.h @@ -764,6 +764,7 @@ struct hnae3_handle { u8 netdev_flags; struct dentry *hnae3_dbgfs; + struct mutex dbgfs_lock; /* Network interface message level enabled bits */ u32 msg_enable; diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c b/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c index 7f3b7084e382f..c68e5f3d0ba52 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c @@ -807,6 +807,7 @@ static ssize_t hns3_dbg_read(struct file *filp, char __user *buffer, if (ret) return ret; + mutex_lock(&handle->dbgfs_lock); save_buf = &hns3_dbg_cmd[index].buf; if (!test_bit(HNS3_NIC_STATE_INITED, &priv->state) || @@ -819,15 +820,15 @@ static ssize_t hns3_dbg_read(struct file *filp, char __user *buffer, read_buf = *save_buf; } else { read_buf = kvzalloc(hns3_dbg_cmd[index].buf_len, GFP_KERNEL); - if (!read_buf) - return -ENOMEM; + if (!read_buf) { + ret = -ENOMEM; + goto out; + } /* save the buffer addr until the last read operation */ *save_buf = read_buf; - } - /* get data ready for the first time to read */ - if (!*ppos) { + /* get data ready for the first time to read */ ret = hns3_dbg_read_cmd(dbg_data, hns3_dbg_cmd[index].cmd, read_buf, hns3_dbg_cmd[index].buf_len); if (ret) @@ -836,8 +837,10 @@ static ssize_t hns3_dbg_read(struct file *filp, char __user *buffer, size = simple_read_from_buffer(buffer, count, ppos, read_buf, strlen(read_buf)); - if (size > 0) + if (size > 0) { + mutex_unlock(&handle->dbgfs_lock); return size; + } out: /* free the buffer for the last read operation */ @@ -846,6 +849,7 @@ static ssize_t hns3_dbg_read(struct file *filp, char __user *buffer, *save_buf = NULL; } + mutex_unlock(&handle->dbgfs_lock); return ret; } @@ -916,6 +920,7 @@ int hns3_dbg_init(struct hnae3_handle *handle) debugfs_create_dir(hns3_dbg_dentry[i].name, handle->hnae3_dbgfs); + mutex_init(&handle->dbgfs_lock); for (i = 0; i < ARRAY_SIZE(hns3_dbg_cmd); i++) { if (!hns3_dbg_cmd[i].init) { dev_err(&handle->pdev->dev, @@ -936,6 +941,7 @@ int hns3_dbg_init(struct hnae3_handle *handle) return 0; out: + mutex_destroy(&handle->dbgfs_lock); debugfs_remove_recursive(handle->hnae3_dbgfs); handle->hnae3_dbgfs = NULL; return ret; @@ -951,6 +957,7 @@ void hns3_dbg_uninit(struct hnae3_handle *handle) hns3_dbg_cmd[i].buf = NULL; } + mutex_destroy(&handle->dbgfs_lock); debugfs_remove_recursive(handle->hnae3_dbgfs); handle->hnae3_dbgfs = NULL; } -- 2.25.1

1 1

[PATCH openEuler-1.0-LTS] kabi: fix kabi broken in struct sock
by Yang Yingliang 09 Dec '21

09 Dec '21

hulk inclusion category: bugfix bugzilla: NA CVE: NA --------------------------- If config DEBUG_SPINLOCK or DEBUG_LOCK_ALLOC is enabled, don't need fix the kabi broken. It's introduced by 93c5c1d15abcd ("af_unix: fix races in sk_peer_pid and sk_peer_cred accesses"). Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> Reviewed-by: Cheng Jian <cj.chengjian(a)huawei.com> Reviewed-by: Yue Haibing <yuehaibing(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- include/net/sock.h | 13 +++++++++++++ 1 file changed, 13 insertions(+) diff --git a/include/net/sock.h b/include/net/sock.h index b90b92882b3b8..803464e66e02c 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -472,7 +472,9 @@ struct sock { u32 sk_ack_backlog; u32 sk_max_ack_backlog; kuid_t sk_uid; +#if defined(CONFIG_DEBUG_SPINLOCK) || defined(CONFIG_DEBUG_LOCK_ALLOC) spinlock_t sk_peer_lock; +#endif struct pid *sk_peer_pid; const struct cred *sk_peer_cred; @@ -513,7 +515,18 @@ struct sock { struct sock_reuseport __rcu *sk_reuseport_cb; struct rcu_head sk_rcu; +#if !defined(CONFIG_DEBUG_SPINLOCK) && !defined(CONFIG_DEBUG_LOCK_ALLOC) +#ifndef __GENKSYMS__ + union { + spinlock_t sk_peer_lock; + unsigned long kabi_reserve1; + }; +#else KABI_RESERVE(1) +#endif +#else + KABI_RESERVE(1) +#endif KABI_RESERVE(2) KABI_RESERVE(3) KABI_RESERVE(4) -- 2.25.1

1 0

[PATCH openEuler-1.0-LTS 01/31] cifs: fix incorrect check for null pointer in header_assemble
by Yang Yingliang 08 Dec '21

08 Dec '21

From: Steve French <stfrench(a)microsoft.com> stable inclusion from linux-4.19.209 commit 43d2e0fbc67f8bcfb069130f4028a04887ae76b6 -------------------------------- commit 9ed38fd4a15417cac83967360cf20b853bfab9b6 upstream. Although very unlikely that the tlink pointer would be null in this case, get_next_mid function can in theory return null (but not an error) so need to check for null (not for IS_ERR, which can not be returned here). Address warning: fs/smbfs_client/connect.c:2392 cifs_match_super() warn: 'tlink' isn't an ERR_PTR Pointed out by Dan Carpenter via smatch code analysis tool CC: stable(a)vger.kernel.org Reported-by: Dan Carpenter <dan.carpenter(a)oracle.com> Acked-by: Ronnie Sahlberg <lsahlber(a)redhat.com> Signed-off-by: Steve French <stfrench(a)microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Acked-by: Jason Yan <yanaijie(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- fs/cifs/connect.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 907be252c5d47..36104dd8eb4dd 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -3373,9 +3373,10 @@ cifs_match_super(struct super_block *sb, void *data) spin_lock(&cifs_tcp_ses_lock); cifs_sb = CIFS_SB(sb); tlink = cifs_get_tlink(cifs_sb_master_tlink(cifs_sb)); - if (IS_ERR(tlink)) { + if (tlink == NULL) { + /* can not match superblock if tlink were ever null */ spin_unlock(&cifs_tcp_ses_lock); - return rc; + return 0; } tcon = tlink_tcon(tlink); ses = tcon->ses; -- 2.25.1

1 30

[PATCH OLK-5.10 107/107] fs/ntfs3: Add ntfs3 module in openeuler_defconfig
by Yin Xiujiang 08 Dec '21

08 Dec '21

kylin inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ------------------------------------------------- This adds ntfs3 module in openeuler_defconfig Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- arch/arm64/configs/openeuler_defconfig | 4 ++++ arch/x86/configs/openeuler_defconfig | 4 ++++ 2 files changed, 8 insertions(+) diff --git a/arch/arm64/configs/openeuler_defconfig b/arch/arm64/configs/openeuler_defconfig index 76d6a118330d..b1e8524eb5e6 100644 --- a/arch/arm64/configs/openeuler_defconfig +++ b/arch/arm64/configs/openeuler_defconfig @@ -6163,6 +6163,10 @@ CONFIG_EXFAT_DEFAULT_IOCHARSET="utf8" CONFIG_NTFS_FS=m # CONFIG_NTFS_DEBUG is not set # CONFIG_NTFS_RW is not set +CONFIG_NTFS3_FS=m +CONFIG_NTFS3_64BIT_CLUSTER=y +CONFIG_NTFS3_LZX_XPRESS=y +CONFIG_NTFS3_FS_POSIX_ACL=y # end of DOS/FAT/EXFAT/NT Filesystems # diff --git a/arch/x86/configs/openeuler_defconfig b/arch/x86/configs/openeuler_defconfig index b25d908dc7a1..83e143d139fa 100644 --- a/arch/x86/configs/openeuler_defconfig +++ b/arch/x86/configs/openeuler_defconfig @@ -7515,6 +7515,10 @@ CONFIG_EXFAT_DEFAULT_IOCHARSET="utf8" CONFIG_NTFS_FS=m # CONFIG_NTFS_DEBUG is not set # CONFIG_NTFS_RW is not set +CONFIG_NTFS3_FS=m +CONFIG_NTFS3_64BIT_CLUSTER=y +CONFIG_NTFS3_LZX_XPRESS=y +CONFIG_NTFS3_FS_POSIX_ACL=y # end of DOS/FAT/EXFAT/NT Filesystems # -- 2.30.0

1 0

[PATCH OLK-5.10 081/107] fs/ntfs3: Remove a useless shadowing variable
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Christophe JAILLET <christophe.jaillet(a)wanadoo.fr> mainline inclusion from mainline-v5.15 commit 808bc0a82bcd2cbe32a139613325b1a3e03f35f1 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- There is already a 'u8 mask' defined at the top of the function. There is no need to define a new one here. Remove the useless and shadowing new 'mask' variable. Signed-off-by: Christophe JAILLET <christophe.jaillet(a)wanadoo.fr> Reviewed-by: Kari Argillander <kari.argillander(a)gmail.com> Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/bitfunc.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/fs/ntfs3/bitfunc.c b/fs/ntfs3/bitfunc.c index bf10e2da5c6e..50d838093790 100644 --- a/fs/ntfs3/bitfunc.c +++ b/fs/ntfs3/bitfunc.c @@ -119,8 +119,7 @@ bool are_bits_set(const ulong *lmap, size_t bit, size_t nbits) pos = nbits & 7; if (pos) { - u8 mask = fill_mask[pos]; - + mask = fill_mask[pos]; if ((*map & mask) != mask) return false; } -- 2.30.0

1 0

[PATCH OLK-5.10 080/107] fs/ntfs3: Remove a useless test in 'indx_find()'
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Christophe JAILLET <christophe.jaillet(a)wanadoo.fr> mainline inclusion from mainline-v5.15 commit d2846bf33c1423ff872c7a7c2afde292ad502c04 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- 'fnd' has been dereferenced several time before, so testing it here is pointless. Moreover, all callers of 'indx_find()' already have some error handling code that makes sure that no NULL 'fnd' is passed. So, remove the useless test. Signed-off-by: Christophe JAILLET <christophe.jaillet(a)wanadoo.fr> Reviewed-by: Kari Argillander <kari.argillander(a)gmail.com> Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/index.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/fs/ntfs3/index.c b/fs/ntfs3/index.c index 4f71a91f07d9..6f81e3a49abf 100644 --- a/fs/ntfs3/index.c +++ b/fs/ntfs3/index.c @@ -1072,9 +1072,7 @@ int indx_find(struct ntfs_index *indx, struct ntfs_inode *ni, if (!e) return -EINVAL; - if (fnd) - fnd->root_de = e; - + fnd->root_de = e; err = 0; for (;;) { -- 2.30.0

1 0

[PATCH OLK-5.10 078/107] fs/ntfs3: Change max hardlinks limit to 4000
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> mainline inclusion from mainline-v5.15 commit 56eaeb10e2619081cc383febf6740a4c3e806777 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- xfstest generic/041 works with 3003 hardlinks. Because of this we raise hardlinks limit to 4000. There are no drawbacks or regressions. Theoretically we can raise all the way up to ffff, but there is no practical use for this. Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/ntfs.h | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/fs/ntfs3/ntfs.h b/fs/ntfs3/ntfs.h index 303a162c3158..9cc396b117bf 100644 --- a/fs/ntfs3/ntfs.h +++ b/fs/ntfs3/ntfs.h @@ -26,9 +26,11 @@ #define NTFS_NAME_LEN 255 -/* ntfs.sys used 500 maximum links on-disk struct allows up to 0xffff. */ -#define NTFS_LINK_MAX 0x400 -//#define NTFS_LINK_MAX 0xffff +/* + * ntfs.sys used 500 maximum links on-disk struct allows up to 0xffff. + * xfstest generic/041 creates 3003 hardlinks. + */ +#define NTFS_LINK_MAX 4000 /* * Activate to use 64 bit clusters instead of 32 bits in ntfs.sys. -- 2.30.0

1 0

[PATCH OLK-5.10 076/107] fs/ntfs3: Fix a memory leak on object opts
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Colin Ian King <colin.king(a)canonical.com> mainline inclusion from mainline-v5.15 commit 880301bb313295a65523e79bc5666f5cf49eb3ed category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- Currently a failed allocation on sbi->upcase will cause an exit via the label free_sbi causing a memory leak on object opts. Fix this by re-ordering the exit paths free_opts and free_sbi so that kfree's occur in the reverse allocation order. Addresses-Coverity: ("Resource leak") Fixes: 27fac77707a1 ("fs/ntfs3: Init spi more in init_fs_context than fill_super") Signed-off-by: Colin Ian King <colin.king(a)canonical.com> Reviewed-by: Kari Argillander <kari.argillander(a)gmail.com> Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/super.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/ntfs3/super.c b/fs/ntfs3/super.c index cefb9ddaf4db..6a535b144ff9 100644 --- a/fs/ntfs3/super.c +++ b/fs/ntfs3/super.c @@ -1393,10 +1393,10 @@ static int ntfs_init_fs_context(struct fs_context *fc) fc->ops = &ntfs_context_ops; return 0; -free_opts: - kfree(opts); free_sbi: kfree(sbi); +free_opts: + kfree(opts); return -ENOMEM; } -- 2.30.0

1 0

[PATCH OLK-5.10 048/107] fs/ntfs3: Add missing header and guards to lib/ headers
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Kari Argillander <kari.argillander(a)gmail.com> mainline inclusion from mainline-v5.15 commit b6ba81034b1b74cf426abcece4becda2611504a4 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- size_t needs header. Add missing header guards so that compiler will only include these ones. Signed-off-by: Kari Argillander <kari.argillander(a)gmail.com> Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/lib/decompress_common.h | 5 +++++ fs/ntfs3/lib/lib.h | 6 ++++++ 2 files changed, 11 insertions(+) diff --git a/fs/ntfs3/lib/decompress_common.h b/fs/ntfs3/lib/decompress_common.h index 2d70ae42f1b5..dd7ced000d0e 100644 --- a/fs/ntfs3/lib/decompress_common.h +++ b/fs/ntfs3/lib/decompress_common.h @@ -5,6 +5,9 @@ * Copyright (C) 2015 Eric Biggers */ +#ifndef _LINUX_NTFS3_LIB_DECOMPRESS_COMMON_H +#define _LINUX_NTFS3_LIB_DECOMPRESS_COMMON_H + #include <linux/string.h> #include <linux/compiler.h> #include <linux/types.h> @@ -336,3 +339,5 @@ static forceinline u8 *lz_copy(u8 *dst, u32 length, u32 offset, const u8 *bufend return dst; } + +#endif /* _LINUX_NTFS3_LIB_DECOMPRESS_COMMON_H */ diff --git a/fs/ntfs3/lib/lib.h b/fs/ntfs3/lib/lib.h index f508fbad2e71..90309a5ae59c 100644 --- a/fs/ntfs3/lib/lib.h +++ b/fs/ntfs3/lib/lib.h @@ -7,6 +7,10 @@ * - linux kernel code style */ +#ifndef _LINUX_NTFS3_LIB_LIB_H +#define _LINUX_NTFS3_LIB_LIB_H + +#include <linux/types.h> /* globals from xpress_decompress.c */ struct xpress_decompressor *xpress_allocate_decompressor(void); @@ -24,3 +28,5 @@ int lzx_decompress(struct lzx_decompressor *__restrict d, const void *__restrict compressed_data, size_t compressed_size, void *__restrict uncompressed_data, size_t uncompressed_size); + +#endif /* _LINUX_NTFS3_LIB_LIB_H */ -- 2.30.0

1 0

[PATCH OLK-5.10 047/107] fs/ntfs3: Add missing headers and forward declarations to ntfs_fs.h
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Kari Argillander <kari.argillander(a)gmail.com> mainline inclusion from mainline-v5.15 commit f239b3a95dd4f7daba26ea17f339a5b19a7d40a1 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- We do not have headers at all in this file. We should have them so that not every .c file needs to include all of the stuff which this file need for building. This way we can remove some headers from other files and get better picture what is needed. This can save some compilation time. And this can help if we sometimes want to separate this one big header. Also use forward declarations for structs and enums when it not included straight with include and it is used in function declarations input. This will prevent possible compiler warning: xxx declared inside parameter list will not be visible outside of this definition or declaration Here is list which I made when parsing this. There is not necessarily all example from this header file, but this just proofs we need it. <linux/blkdev.h> SECTOR_SHIFT <linux/buffer_head.h> sb_bread(), put_bh <linux/cleancache.h> put_page() <linux/fs.h> struct inode (Just struct ntfs_inode need it) <linux/highmem.h> kunmap(), kmap() <linux/kernel.h> cpu_to_leXX() ALIGN <linux/mm.h> kvfree() <linux/mutex.h> struct mutex, mutex_(un/try)lock() <linux/page-flags.h> PageError() <linux/pagemap.h> read_mapping_page() <linux/rbtree.h> struct rb_root <linux/rwsem.h> struct rw_semaphore <linux/slab.h> krfree(), kzalloc() <linux/string.h> memset() <linux/time64.h> struct timespec64 <linux/types.h> uXX, __leXX <linux/uidgid.h> kuid_t, kgid_t <asm/div64.h> do_div() <asm/page.h> PAGE_SIZE "debug.h" ntfs_err() (Just one entry. Maybe we can drop this) "ntfs.h" Do you even ask? Signed-off-by: Kari Argillander <kari.argillander(a)gmail.com> Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/ntfs_fs.h | 31 +++++++++++++++++++++++++++++++ 1 file changed, 31 insertions(+) diff --git a/fs/ntfs3/ntfs_fs.h b/fs/ntfs3/ntfs_fs.h index 372cda697dd4..dae6dd4ac619 100644 --- a/fs/ntfs3/ntfs_fs.h +++ b/fs/ntfs3/ntfs_fs.h @@ -9,6 +9,37 @@ #ifndef _LINUX_NTFS3_NTFS_FS_H #define _LINUX_NTFS3_NTFS_FS_H +#include <linux/blkdev.h> +#include <linux/buffer_head.h> +#include <linux/cleancache.h> +#include <linux/fs.h> +#include <linux/highmem.h> +#include <linux/kernel.h> +#include <linux/mm.h> +#include <linux/mutex.h> +#include <linux/page-flags.h> +#include <linux/pagemap.h> +#include <linux/rbtree.h> +#include <linux/rwsem.h> +#include <linux/slab.h> +#include <linux/string.h> +#include <linux/time64.h> +#include <linux/types.h> +#include <linux/uidgid.h> +#include <asm/div64.h> +#include <asm/page.h> + +#include "debug.h" +#include "ntfs.h" + +struct dentry; +struct fiemap_extent_info; +struct user_namespace; +struct page; +struct writeback_control; +enum utf16_endian; + + #define MINUS_ONE_T ((size_t)(-1)) /* Biggest MFT / smallest cluster */ #define MAXIMUM_BYTES_PER_MFT 4096 -- 2.30.0

1 0

[PATCH OLK-5.10 046/107] fs/ntfs3: Add missing header files to ntfs.h
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Kari Argillander <kari.argillander(a)gmail.com> mainline inclusion from mainline-v5.15 commit 4dfe83320e1e9665b986840b426742ea764e08d7 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- We do not have header files at all in this file. Add following headers and there is also explanation which for it was added. Note that explanation might not be complete, but it just proofs it is needed. <linux/blkdev.h> // SECTOR_SHIFT <linux/build_bug.h> // static_assert() <linux/kernel.h> // cpu_to_le64, cpu_to_le32, ALIGN <linux/stddef.h> // offsetof() <linux/string.h> // memcmp() <linux/types.h> //__le32, __le16 "debug.h" // PtrOffset(), Add2Ptr() Signed-off-by: Kari Argillander <kari.argillander(a)gmail.com> Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/ntfs.h | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/fs/ntfs3/ntfs.h b/fs/ntfs3/ntfs.h index 6bb3e595263b..695b684bce20 100644 --- a/fs/ntfs3/ntfs.h +++ b/fs/ntfs3/ntfs.h @@ -10,6 +10,15 @@ #ifndef _LINUX_NTFS3_NTFS_H #define _LINUX_NTFS3_NTFS_H +#include <linux/blkdev.h> +#include <linux/build_bug.h> +#include <linux/kernel.h> +#include <linux/stddef.h> +#include <linux/string.h> +#include <linux/types.h> + +#include "debug.h" + /* TODO: Check 4K MFT record and 512 bytes cluster. */ /* Activate this define to use binary search in indexes. */ -- 2.30.0

1 0

[PATCH OLK-5.10 044/107] fs/ntfs3: Remove redundant initialization of variable err
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Colin Ian King <colin.king(a)canonical.com> mainline inclusion from mainline-v5.15 commit 0327c6d01a97a3242cf10717819994aa6e095a1d category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- The variable err is being initialized with a value that is never read, it is being updated later on. The assignment is redundant and can be removed. Addresses-Coverity: ("Unused value") Signed-off-by: Colin Ian King <colin.king(a)canonical.com> Reviewed-by: Kari Argillander <kari.argillander(a)gmail.com> Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/index.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/ntfs3/index.c b/fs/ntfs3/index.c index 0daca9adc54c..b1175542d854 100644 --- a/fs/ntfs3/index.c +++ b/fs/ntfs3/index.c @@ -1401,7 +1401,7 @@ int indx_find_raw(struct ntfs_index *indx, struct ntfs_inode *ni, static int indx_create_allocate(struct ntfs_index *indx, struct ntfs_inode *ni, CLST *vbn) { - int err = -ENOMEM; + int err; struct ntfs_sb_info *sbi = ni->mi.sbi; struct ATTRIB *bitmap; struct ATTRIB *alloc; -- 2.30.0

1 0

[PATCH OLK-5.10 042/107] fs/ntfs3: Rename mount option no_acs_rules > (no)acsrules
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Kari Argillander <kari.argillander(a)gmail.com> mainline inclusion from mainline-v5.15 commit 28a941ffc1404b66d67228cbe8392bbadb94af0d category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- Rename mount option no_acs_rules to (no)acsrules. This allow us to use possibility to mount with options noaclrules or aclrules. Acked-by: Christian Brauner <christian.brauner(a)ubuntu.com> Reviewed-by: Christoph Hellwig <hch(a)lst.de> Signed-off-by: Kari Argillander <kari.argillander(a)gmail.com> Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- Documentation/filesystems/ntfs3.rst | 2 +- fs/ntfs3/file.c | 2 +- fs/ntfs3/ntfs_fs.h | 2 +- fs/ntfs3/super.c | 12 ++++++------ fs/ntfs3/xattr.c | 2 +- 5 files changed, 10 insertions(+), 10 deletions(-) diff --git a/Documentation/filesystems/ntfs3.rst b/Documentation/filesystems/ntfs3.rst index ded706474825..7b6afe452197 100644 --- a/Documentation/filesystems/ntfs3.rst +++ b/Documentation/filesystems/ntfs3.rst @@ -73,7 +73,7 @@ prealloc Preallocate space for files excessively when file size is increasing on writes. Decreases fragmentation in case of parallel write operations to different files. -no_acs_rules "No access rules" mount option sets access rights for +noacsrules "No access rules" mount option sets access rights for files/folders to 777 and owner/group to root. This mount option absorbs all other permissions: - permissions change for files/folders will be reported diff --git a/fs/ntfs3/file.c b/fs/ntfs3/file.c index fef57141b161..0743d806c567 100644 --- a/fs/ntfs3/file.c +++ b/fs/ntfs3/file.c @@ -737,7 +737,7 @@ int ntfs3_setattr(struct user_namespace *mnt_userns, struct dentry *dentry, umode_t mode = inode->i_mode; int err; - if (sbi->options->no_acs_rules) { + if (sbi->options->noacsrules) { /* "No access rules" - Force any changes of time etc. */ attr->ia_valid |= ATTR_FORCE; /* and disable for editing some attributes. */ diff --git a/fs/ntfs3/ntfs_fs.h b/fs/ntfs3/ntfs_fs.h index aa18f12b7096..15bab48bc1ad 100644 --- a/fs/ntfs3/ntfs_fs.h +++ b/fs/ntfs3/ntfs_fs.h @@ -70,7 +70,7 @@ struct ntfs_mount_options { showmeta : 1, /* Show meta files. */ nohidden : 1, /* Do not show hidden files. */ force : 1, /* Rw mount dirty volume. */ - no_acs_rules : 1, /*Exclude acs rules. */ + noacsrules : 1, /*Exclude acs rules. */ prealloc : 1 /* Preallocate space when file is growing. */ ; }; diff --git a/fs/ntfs3/super.c b/fs/ntfs3/super.c index 503e2e23f711..0690e7e4f00d 100644 --- a/fs/ntfs3/super.c +++ b/fs/ntfs3/super.c @@ -228,7 +228,7 @@ enum Opt { Opt_acl, Opt_iocharset, Opt_prealloc, - Opt_no_acs_rules, + Opt_noacsrules, Opt_err, }; @@ -246,7 +246,7 @@ static const struct fs_parameter_spec ntfs_fs_parameters[] = { fsparam_flag_no("acl", Opt_acl), fsparam_flag_no("showmeta", Opt_showmeta), fsparam_flag_no("prealloc", Opt_prealloc), - fsparam_flag("no_acs_rules", Opt_no_acs_rules), + fsparam_flag_no("acsrules", Opt_noacsrules), fsparam_string("iocharset", Opt_iocharset), __fsparam(fs_param_is_string, @@ -358,8 +358,8 @@ static int ntfs_fs_parse_param(struct fs_context *fc, case Opt_prealloc: opts->prealloc = result.negated ? 0 : 1; break; - case Opt_no_acs_rules: - opts->no_acs_rules = 1; + case Opt_noacsrules: + opts->noacsrules = result.negated ? 1 : 0; break; default: /* Should not be here unless we forget add case. */ @@ -547,8 +547,8 @@ static int ntfs_show_options(struct seq_file *m, struct dentry *root) seq_puts(m, ",nohidden"); if (opts->force) seq_puts(m, ",force"); - if (opts->no_acs_rules) - seq_puts(m, ",no_acs_rules"); + if (opts->noacsrules) + seq_puts(m, ",noacsrules"); if (opts->prealloc) seq_puts(m, ",prealloc"); if (sb->s_flags & SB_POSIXACL) diff --git a/fs/ntfs3/xattr.c b/fs/ntfs3/xattr.c index ac4b37bf8832..6f88cb77a17f 100644 --- a/fs/ntfs3/xattr.c +++ b/fs/ntfs3/xattr.c @@ -769,7 +769,7 @@ int ntfs_acl_chmod(struct user_namespace *mnt_userns, struct inode *inode) int ntfs_permission(struct user_namespace *mnt_userns, struct inode *inode, int mask) { - if (ntfs_sb(inode->i_sb)->options->no_acs_rules) { + if (ntfs_sb(inode->i_sb)->options->noacsrules) { /* "No access rules" mode - Allow all changes. */ return 0; } -- 2.30.0

1 0

[PATCH OLK-5.10 041/107] fs/ntfs3: Add iocharset= mount option as alias for nls=
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Kari Argillander <kari.argillander(a)gmail.com> mainline inclusion from mainline-v5.15 commit e274cde8c7550cac46eb7aba3a77aff44ae0b301 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- Other fs drivers are using iocharset= mount option for specifying charset. So add it also for ntfs3 and mark old nls= mount option as deprecated. Reviewed-by: Pali Rohár <pali(a)kernel.org> Signed-off-by: Kari Argillander <kari.argillander(a)gmail.com> Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- Documentation/filesystems/ntfs3.rst | 4 ++-- fs/ntfs3/super.c | 18 +++++++++++------- 2 files changed, 13 insertions(+), 9 deletions(-) diff --git a/Documentation/filesystems/ntfs3.rst b/Documentation/filesystems/ntfs3.rst index af7158de6fde..ded706474825 100644 --- a/Documentation/filesystems/ntfs3.rst +++ b/Documentation/filesystems/ntfs3.rst @@ -32,12 +32,12 @@ generic ones. =============================================================================== -nls=name This option informs the driver how to interpret path +iocharset=name This option informs the driver how to interpret path strings and translate them to Unicode and back. If this option is not set, the default codepage will be used (CONFIG_NLS_DEFAULT). Examples: - 'nls=utf8' + 'iocharset=utf8' uid= gid= diff --git a/fs/ntfs3/super.c b/fs/ntfs3/super.c index 729ead6f2fac..503e2e23f711 100644 --- a/fs/ntfs3/super.c +++ b/fs/ntfs3/super.c @@ -226,7 +226,7 @@ enum Opt { Opt_nohidden, Opt_showmeta, Opt_acl, - Opt_nls, + Opt_iocharset, Opt_prealloc, Opt_no_acs_rules, Opt_err, @@ -245,9 +245,13 @@ static const struct fs_parameter_spec ntfs_fs_parameters[] = { fsparam_flag_no("hidden", Opt_nohidden), fsparam_flag_no("acl", Opt_acl), fsparam_flag_no("showmeta", Opt_showmeta), - fsparam_string("nls", Opt_nls), fsparam_flag_no("prealloc", Opt_prealloc), fsparam_flag("no_acs_rules", Opt_no_acs_rules), + fsparam_string("iocharset", Opt_iocharset), + + __fsparam(fs_param_is_string, + "nls", Opt_iocharset, + fs_param_deprecated, NULL), {} }; @@ -346,7 +350,7 @@ static int ntfs_fs_parse_param(struct fs_context *fc, case Opt_showmeta: opts->showmeta = result.negated ? 0 : 1; break; - case Opt_nls: + case Opt_iocharset: kfree(opts->nls_name); opts->nls_name = param->string; param->string = NULL; @@ -380,11 +384,11 @@ static int ntfs_fs_reconfigure(struct fs_context *fc) new_opts->nls = ntfs_load_nls(new_opts->nls_name); if (IS_ERR(new_opts->nls)) { new_opts->nls = NULL; - errorf(fc, "ntfs3: Cannot load nls %s", new_opts->nls_name); + errorf(fc, "ntfs3: Cannot load iocharset %s", new_opts->nls_name); return -EINVAL; } if (new_opts->nls != sbi->options->nls) - return invalf(fc, "ntfs3: Cannot use different nls when remounting!"); + return invalf(fc, "ntfs3: Cannot use different iocharset when remounting!"); sync_filesystem(sb); @@ -528,9 +532,9 @@ static int ntfs_show_options(struct seq_file *m, struct dentry *root) if (opts->dmask) seq_printf(m, ",dmask=%04o", ~opts->fs_dmask_inv); if (opts->nls) - seq_printf(m, ",nls=%s", opts->nls->charset); + seq_printf(m, ",iocharset=%s", opts->nls->charset); else - seq_puts(m, ",nls=utf8"); + seq_puts(m, ",iocharset=utf8"); if (opts->sys_immutable) seq_puts(m, ",sys_immutable"); if (opts->discard) -- 2.30.0

1 0

[PATCH OLK-5.10 040/107] fs/ntfs3: Make mount option nohidden more universal
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Kari Argillander <kari.argillander(a)gmail.com> mainline inclusion from mainline-v5.15 commit 9d1939f4575f3fda70dd94542dbd4d775e104132 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- If we call Opt_nohidden with just keyword hidden, then we can use hidden/nohidden when mounting. We already use this method for almoust all other parameters so it is just logical that this will use same method. Acked-by: Christian Brauner <christian.brauner(a)ubuntu.com> Reviewed-by: Christoph Hellwig <hch(a)lst.de> Reviewed-by: Pali Rohár <pali(a)kernel.org> Signed-off-by: Kari Argillander <kari.argillander(a)gmail.com> Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/super.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/ntfs3/super.c b/fs/ntfs3/super.c index 420cd1409170..729ead6f2fac 100644 --- a/fs/ntfs3/super.c +++ b/fs/ntfs3/super.c @@ -242,7 +242,7 @@ static const struct fs_parameter_spec ntfs_fs_parameters[] = { fsparam_flag_no("discard", Opt_discard), fsparam_flag_no("force", Opt_force), fsparam_flag_no("sparse", Opt_sparse), - fsparam_flag("nohidden", Opt_nohidden), + fsparam_flag_no("hidden", Opt_nohidden), fsparam_flag_no("acl", Opt_acl), fsparam_flag_no("showmeta", Opt_showmeta), fsparam_string("nls", Opt_nls), @@ -331,7 +331,7 @@ static int ntfs_fs_parse_param(struct fs_context *fc, opts->sparse = result.negated ? 0 : 1; break; case Opt_nohidden: - opts->nohidden = 1; + opts->nohidden = result.negated ? 1 : 0; break; case Opt_acl: if (!result.negated) -- 2.30.0

1 0

[PATCH OLK-5.10 035/107] fs/ntfs3: Remove unnecesarry mount option noatime
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Kari Argillander <kari.argillander(a)gmail.com> mainline inclusion from mainline-v5.15 commit b8a30b4171b9a3c22ef0605ed74a21544d00c680 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- Remove unnecesarry mount option noatime because this will be handled by VFS. Our option parser will never get opt like this. Acked-by: Christian Brauner <christian.brauner(a)ubuntu.com> Reviewed-by: Christoph Hellwig <hch(a)lst.de> Reviewed-by: Pali Rohár <pali(a)kernel.org> Signed-off-by: Kari Argillander <kari.argillander(a)gmail.com> Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- Documentation/filesystems/ntfs3.rst | 4 ---- fs/ntfs3/super.c | 7 ------- 2 files changed, 11 deletions(-) diff --git a/Documentation/filesystems/ntfs3.rst b/Documentation/filesystems/ntfs3.rst index ffe9ea0c1499..af7158de6fde 100644 --- a/Documentation/filesystems/ntfs3.rst +++ b/Documentation/filesystems/ntfs3.rst @@ -85,10 +85,6 @@ acl Support POSIX ACLs (Access Control Lists). Effective if supported by Kernel. Not to be confused with NTFS ACLs. The option specified as acl enables support for POSIX ACLs. -noatime All files and directories will not update their last access - time attribute if a partition is mounted with this parameter. - This option can speed up file system operation. - =============================================================================== ToDo list diff --git a/fs/ntfs3/super.c b/fs/ntfs3/super.c index 55bbc9200a10..a18b99a3e3b5 100644 --- a/fs/ntfs3/super.c +++ b/fs/ntfs3/super.c @@ -223,7 +223,6 @@ enum Opt { Opt_nohidden, Opt_showmeta, Opt_acl, - Opt_noatime, Opt_nls, Opt_prealloc, Opt_no_acs_rules, @@ -242,7 +241,6 @@ static const match_table_t ntfs_tokens = { { Opt_sparse, "sparse" }, { Opt_nohidden, "nohidden" }, { Opt_acl, "acl" }, - { Opt_noatime, "noatime" }, { Opt_showmeta, "showmeta" }, { Opt_nls, "nls=%s" }, { Opt_prealloc, "prealloc" }, @@ -333,9 +331,6 @@ static noinline int ntfs_parse_options(struct super_block *sb, char *options, ntfs_err(sb, "support for ACL not compiled in!"); return -EINVAL; #endif - case Opt_noatime: - sb->s_flags |= SB_NOATIME; - break; case Opt_showmeta: opts->showmeta = 1; break; @@ -587,8 +582,6 @@ static int ntfs_show_options(struct seq_file *m, struct dentry *root) seq_puts(m, ",prealloc"); if (sb->s_flags & SB_POSIXACL) seq_puts(m, ",acl"); - if (sb->s_flags & SB_NOATIME) - seq_puts(m, ",noatime"); return 0; } -- 2.30.0

1 0

[PATCH OLK-5.10 031/107] fs/ntfs3: Fix integer overflow in ni_fiemap with fiemap_prep()
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Kari Argillander <kari.argillander(a)gmail.com> mainline inclusion from mainline-v5.15 commit d4e8e135a9af7d8d939bba1874ab314322fc2dc2 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- Use fiemap_prep() to check valid flags. It also shrink request scope (@len) to what the fs can actually handle. This address following Smatch static checker warning: fs/ntfs3/frecord.c:1894 ni_fiemap() warn: potential integer overflow from user 'vbo + len' Because fiemap_prep() shrinks @len this cannot happened anymore. Reported-by: Dan Carpenter <dan.carpenter(a)oracle.com> Link: lore.kernel.org/ntfs3/20210825080440.GA17407@kili/ Fixes: 4342306f0f0d ("fs/ntfs3: Add file operations and implementation") Signed-off-by: Kari Argillander <kari.argillander(a)gmail.com> Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/file.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/fs/ntfs3/file.c b/fs/ntfs3/file.c index 89557d60a9b0..f9c9a8c91b46 100644 --- a/fs/ntfs3/file.c +++ b/fs/ntfs3/file.c @@ -1212,8 +1212,9 @@ int ntfs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, int err; struct ntfs_inode *ni = ntfs_i(inode); - if (fieinfo->fi_flags & FIEMAP_FLAG_XATTR) - return -EOPNOTSUPP; + err = fiemap_prep(inode, fieinfo, start, &len, ~FIEMAP_FLAG_XATTR); + if (err) + return err; ni_lock(ni); -- 2.30.0

1 0

[PATCH OLK-5.10 030/107] fs/ntfs3: Restyle comments to better align with kernel-doc
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> mainline inclusion from mainline-v5.15 commit d3624466b56dd5b1886c1dff500525b544c19c83 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/attrib.c | 18 +++++++++++------- fs/ntfs3/bitmap.c | 5 ++--- fs/ntfs3/file.c | 18 +++++++++--------- fs/ntfs3/frecord.c | 27 +++++++++++++-------------- fs/ntfs3/fslog.c | 11 +++++++---- fs/ntfs3/fsntfs.c | 8 ++++---- fs/ntfs3/index.c | 8 +++++--- fs/ntfs3/inode.c | 20 ++++++++++---------- fs/ntfs3/lznt.c | 5 +++-- fs/ntfs3/ntfs.h | 2 +- fs/ntfs3/ntfs_fs.h | 24 ++++++++++++------------ fs/ntfs3/record.c | 2 +- fs/ntfs3/super.c | 2 +- fs/ntfs3/upcase.c | 2 +- fs/ntfs3/xattr.c | 7 ++++--- 15 files changed, 84 insertions(+), 75 deletions(-) diff --git a/fs/ntfs3/attrib.c b/fs/ntfs3/attrib.c index ffc323bacc9f..34c4cbf7e29b 100644 --- a/fs/ntfs3/attrib.c +++ b/fs/ntfs3/attrib.c @@ -199,6 +199,7 @@ int attr_allocate_clusters(struct ntfs_sb_info *sbi, struct runs_tree *run, /* Add new fragment into run storage. */ if (!run_add_entry(run, vcn, lcn, flen, opt == ALLOCATE_MFT)) { + /* Undo last 'ntfs_look_for_free_space' */ down_write_nested(&wnd->rw_lock, BITMAP_MUTEX_CLUSTERS); wnd_set_free(wnd, lcn, flen); up_write(&wnd->rw_lock); @@ -351,7 +352,6 @@ int attr_make_nonresident(struct ntfs_inode *ni, struct ATTRIB *attr, run_close(run); out1: kfree(attr_s); - /* Reinsert le. */ out: return err; } @@ -1153,14 +1153,18 @@ int attr_load_runs_vcn(struct ntfs_inode *ni, enum ATTR_TYPE type, u16 ro; attr = ni_find_attr(ni, NULL, NULL, type, name, name_len, &vcn, NULL); - if (!attr) + if (!attr) { + /* Is record corrupted? */ return -ENOENT; + } svcn = le64_to_cpu(attr->nres.svcn); evcn = le64_to_cpu(attr->nres.evcn); - if (evcn < vcn || vcn < svcn) + if (evcn < vcn || vcn < svcn) { + /* Is record corrupted? */ return -EINVAL; + } ro = le16_to_cpu(attr->nres.run_off); err = run_unpack_ex(run, ni->mi.sbi, ni->mi.rno, svcn, evcn, svcn, @@ -1171,7 +1175,7 @@ int attr_load_runs_vcn(struct ntfs_inode *ni, enum ATTR_TYPE type, } /* - * attr_wof_load_runs_range - Load runs for given range [from to). + * attr_load_runs_range - Load runs for given range [from to). */ int attr_load_runs_range(struct ntfs_inode *ni, enum ATTR_TYPE type, const __le16 *name, u8 name_len, struct runs_tree *run, @@ -1974,7 +1978,7 @@ int attr_punch_hole(struct ntfs_inode *ni, u64 vbo, u64 bytes, u32 *frame_size) total_size = le64_to_cpu(attr_b->nres.total_size); if (vbo >= alloc_size) { - // NOTE: It is allowed. + /* NOTE: It is allowed. */ return 0; } @@ -1986,9 +1990,9 @@ int attr_punch_hole(struct ntfs_inode *ni, u64 vbo, u64 bytes, u32 *frame_size) bytes -= vbo; if ((vbo & mask) || (bytes & mask)) { - /* We have to zero a range(s)*/ + /* We have to zero a range(s). */ if (frame_size == NULL) { - /* Caller insists range is aligned */ + /* Caller insists range is aligned. */ return -EINVAL; } *frame_size = mask + 1; diff --git a/fs/ntfs3/bitmap.c b/fs/ntfs3/bitmap.c index 06ae38adb8ad..831501555009 100644 --- a/fs/ntfs3/bitmap.c +++ b/fs/ntfs3/bitmap.c @@ -29,7 +29,6 @@ struct rb_node_key { size_t key; }; -/* Tree is sorted by start (key). */ struct e_node { struct rb_node_key start; /* Tree sorted by start. */ struct rb_node_key count; /* Tree sorted by len. */ @@ -1117,7 +1116,7 @@ size_t wnd_find(struct wnd_bitmap *wnd, size_t to_alloc, size_t hint, sb = wnd->sb; log2_bits = sb->s_blocksize_bits + 3; - /* At most two ranges [hint, max_alloc) + [0, hint) */ + /* At most two ranges [hint, max_alloc) + [0, hint). */ Again: /* TODO: Optimize request for case nbits > wbits. */ @@ -1241,7 +1240,7 @@ size_t wnd_find(struct wnd_bitmap *wnd, size_t to_alloc, size_t hint, continue; } - /* Read window */ + /* Read window. */ bh = wnd_map(wnd, iw); if (IS_ERR(bh)) { // TODO: Error. diff --git a/fs/ntfs3/file.c b/fs/ntfs3/file.c index 62ebfa324bff..89557d60a9b0 100644 --- a/fs/ntfs3/file.c +++ b/fs/ntfs3/file.c @@ -190,7 +190,8 @@ static int ntfs_extend_initialized_size(struct file *file, /* * ntfs_zero_range - Helper function for punch_hole. - * It zeroes a range [vbo, vbo_to) + * + * It zeroes a range [vbo, vbo_to). */ static int ntfs_zero_range(struct inode *inode, u64 vbo, u64 vbo_to) { @@ -231,12 +232,12 @@ static int ntfs_zero_range(struct inode *inode, u64 vbo, u64 vbo_to) if (!buffer_mapped(bh)) { ntfs_get_block(inode, iblock, bh, 0); - /* unmapped? It's a hole - nothing to do */ + /* Unmapped? It's a hole - nothing to do. */ if (!buffer_mapped(bh)) continue; } - /* Ok, it's mapped. Make sure it's up-to-date */ + /* Ok, it's mapped. Make sure it's up-to-date. */ if (PageUptodate(page)) set_buffer_uptodate(bh); @@ -272,9 +273,8 @@ static int ntfs_zero_range(struct inode *inode, u64 vbo, u64 vbo_to) } /* - * ntfs_sparse_cluster + * ntfs_sparse_cluster - Helper function to zero a new allocated clusters. * - * Helper function to zero a new allocated clusters * NOTE: 512 <= cluster size <= 2M */ void ntfs_sparse_cluster(struct inode *inode, struct page *page0, CLST vcn, @@ -588,7 +588,7 @@ static long ntfs_fallocate(struct file *file, int mode, loff_t vbo, loff_t len) truncate_pagecache(inode, vbo_down); if (!is_sparsed(ni) && !is_compressed(ni)) { - /* normal file */ + /* Normal file. */ err = ntfs_zero_range(inode, vbo, end); goto out; } @@ -599,7 +599,7 @@ static long ntfs_fallocate(struct file *file, int mode, loff_t vbo, loff_t len) if (err != E_NTFS_NOTALIGNED) goto out; - /* process not aligned punch */ + /* Process not aligned punch. */ mask = frame_size - 1; vbo_a = (vbo + mask) & ~mask; end_a = end & ~mask; @@ -647,7 +647,7 @@ static long ntfs_fallocate(struct file *file, int mode, loff_t vbo, loff_t len) if (err) goto out; - /* Wait for existing dio to complete */ + /* Wait for existing dio to complete. */ inode_dio_wait(inode); truncate_pagecache(inode, vbo_down); @@ -1127,7 +1127,7 @@ static ssize_t ntfs_file_write_iter(struct kiocb *iocb, struct iov_iter *from) goto out; if (WARN_ON(ni->ni_flags & NI_FLAG_COMPRESSED_MASK)) { - /* Should never be here, see ntfs_file_open() */ + /* Should never be here, see ntfs_file_open(). */ ret = -EOPNOTSUPP; goto out; } diff --git a/fs/ntfs3/frecord.c b/fs/ntfs3/frecord.c index 3f48b612ec96..938b12d56ca6 100644 --- a/fs/ntfs3/frecord.c +++ b/fs/ntfs3/frecord.c @@ -56,7 +56,7 @@ static struct mft_inode *ni_find_mi(struct ntfs_inode *ni, CLST rno) /* * ni_add_mi - Add new mft_inode into ntfs_inode. -*/ + */ static void ni_add_mi(struct ntfs_inode *ni, struct mft_inode *mi) { ni_ins_mi(ni, &ni->mi_tree, mi->rno, &mi->node); @@ -70,9 +70,8 @@ void ni_remove_mi(struct ntfs_inode *ni, struct mft_inode *mi) rb_erase(&mi->node, &ni->mi_tree); } -/* ni_std - * - * Return: Pointer into std_info from primary record. +/* + * ni_std - Return: Pointer into std_info from primary record. */ struct ATTR_STD_INFO *ni_std(struct ntfs_inode *ni) { @@ -385,7 +384,7 @@ bool ni_add_subrecord(struct ntfs_inode *ni, CLST rno, struct mft_inode **mi) /* * ni_remove_attr - Remove all attributes for the given type/name/id. -*/ + */ int ni_remove_attr(struct ntfs_inode *ni, enum ATTR_TYPE type, const __le16 *name, size_t name_len, bool base_only, const __le16 *id) @@ -740,7 +739,7 @@ static int ni_try_remove_attr_list(struct ntfs_inode *ni) /* * ni_create_attr_list - Generates an attribute list for this primary record. -*/ + */ int ni_create_attr_list(struct ntfs_inode *ni) { struct ntfs_sb_info *sbi = ni->mi.sbi; @@ -939,7 +938,7 @@ static int ni_ins_attr_ext(struct ntfs_inode *ni, struct ATTR_LIST_ENTRY *le, if (is_mft_data && (mi_enum_attr(mi, NULL) || vbo <= ((u64)mi->rno << sbi->record_bits))) { - /* We can't accept this record 'case MFT's bootstrapping. */ + /* We can't accept this record 'cause MFT's bootstrapping. */ continue; } if (is_mft && @@ -1078,7 +1077,7 @@ static int ni_insert_attr(struct ntfs_inode *ni, enum ATTR_TYPE type, */ max_free = free; - /* Estimate the result of moving all possible attributes away.*/ + /* Estimate the result of moving all possible attributes away. */ attr = NULL; while ((attr = mi_enum_attr(&ni->mi, attr))) { @@ -1095,7 +1094,7 @@ static int ni_insert_attr(struct ntfs_inode *ni, enum ATTR_TYPE type, goto out; } - /* Start real attribute moving */ + /* Start real attribute moving. */ attr = NULL; for (;;) { @@ -1542,7 +1541,7 @@ int ni_delete_all(struct ntfs_inode *ni) node = next; } - /* Free base record */ + /* Free base record. */ clear_rec_inuse(ni->mi.mrec); ni->mi.dirty = true; err = mi_write(&ni->mi, 0); @@ -2243,7 +2242,7 @@ int ni_decompress_file(struct ntfs_inode *ni) } if (attr->non_res && is_attr_sparsed(attr)) { - /* Sarsed attribute header is 8 bytes bigger than normal. */ + /* Sparsed attribute header is 8 bytes bigger than normal. */ struct MFT_REC *rec = mi->mrec; u32 used = le32_to_cpu(rec->used); u32 asize = le32_to_cpu(attr->size); @@ -2324,7 +2323,7 @@ static int decompress_lzx_xpress(struct ntfs_sb_info *sbi, const char *cmpr, mutex_lock(&sbi->compress.mtx_xpress); ctx = sbi->compress.xpress; if (!ctx) { - /* Lazy initialize Xpress decompress context */ + /* Lazy initialize Xpress decompress context. */ ctx = xpress_allocate_decompressor(); if (!ctx) { err = -ENOMEM; @@ -2348,7 +2347,7 @@ static int decompress_lzx_xpress(struct ntfs_sb_info *sbi, const char *cmpr, /* * ni_read_frame * - * Pages - array of locked pages. + * Pages - Array of locked pages. */ int ni_read_frame(struct ntfs_inode *ni, u64 frame_vbo, struct page **pages, u32 pages_per_frame) @@ -2740,7 +2739,7 @@ int ni_write_frame(struct ntfs_inode *ni, struct page **pages, lznt = NULL; } - /* Compress: frame_mem -> frame_ondisk. */ + /* Compress: frame_mem -> frame_ondisk */ compr_size = compress_lznt(frame_mem, frame_size, frame_ondisk, frame_size, sbi->compress.lznt); mutex_unlock(&sbi->compress.mtx_lznt); diff --git a/fs/ntfs3/fslog.c b/fs/ntfs3/fslog.c index 6f6057129fdd..b5853aed0e25 100644 --- a/fs/ntfs3/fslog.c +++ b/fs/ntfs3/fslog.c @@ -1362,7 +1362,8 @@ static void log_create(struct ntfs_log *log, u32 l_size, const u64 last_lsn, /* Compute the log page values. */ log->data_off = ALIGN( offsetof(struct RECORD_PAGE_HDR, fixups) + - sizeof(short) * ((log->page_size >> SECTOR_SHIFT) + 1), 8); + sizeof(short) * ((log->page_size >> SECTOR_SHIFT) + 1), + 8); log->data_size = log->page_size - log->data_off; log->record_header_len = sizeof(struct LFS_RECORD_HDR); @@ -1372,7 +1373,9 @@ static void log_create(struct ntfs_log *log, u32 l_size, const u64 last_lsn, /* Compute the restart page values. */ log->ra_off = ALIGN( offsetof(struct RESTART_HDR, fixups) + - sizeof(short) * ((log->sys_page_size >> SECTOR_SHIFT) + 1), 8); + sizeof(short) * + ((log->sys_page_size >> SECTOR_SHIFT) + 1), + 8); log->restart_size = log->sys_page_size - log->ra_off; log->ra_size = struct_size(log->ra, clients, 1); log->current_openlog_count = open_log_count; @@ -5132,8 +5135,8 @@ int log_replay(struct ntfs_inode *ni, bool *initialized) rh->sys_page_size = cpu_to_le32(log->page_size); rh->page_size = cpu_to_le32(log->page_size); - t16 = ALIGN(offsetof(struct RESTART_HDR, fixups) + - sizeof(short) * t16, 8); + t16 = ALIGN(offsetof(struct RESTART_HDR, fixups) + sizeof(short) * t16, + 8); rh->ra_off = cpu_to_le16(t16); rh->minor_ver = cpu_to_le16(1); // 0x1A: rh->major_ver = cpu_to_le16(1); // 0x1C: diff --git a/fs/ntfs3/fsntfs.c b/fs/ntfs3/fsntfs.c index 669249439217..91e3743e1442 100644 --- a/fs/ntfs3/fsntfs.c +++ b/fs/ntfs3/fsntfs.c @@ -312,7 +312,7 @@ int ntfs_loadlog_and_replay(struct ntfs_inode *ni, struct ntfs_sb_info *sbi) if (sb_rdonly(sb) || !initialized) goto out; - /* Fill LogFile by '-1' if it is initialized.ssss */ + /* Fill LogFile by '-1' if it is initialized. */ err = ntfs_bio_fill_1(sbi, &ni->file.run); out: @@ -960,10 +960,10 @@ int ntfs_set_state(struct ntfs_sb_info *sbi, enum NTFS_DIRTY_FLAGS dirty) /* verify(!ntfs_update_mftmirr()); */ /* - * if we used wait=1, sync_inode_metadata waits for the io for the + * If we used wait=1, sync_inode_metadata waits for the io for the * inode to finish. It hangs when media is removed. * So wait=0 is sent down to sync_inode_metadata - * and filemap_fdatawrite is used for the data blocks + * and filemap_fdatawrite is used for the data blocks. */ err = sync_inode_metadata(&ni->vfs_inode, 0); if (!err) @@ -1917,7 +1917,7 @@ int ntfs_security_init(struct ntfs_sb_info *sbi) sbi->security.next_id = SECURITY_ID_FIRST; /* Always write new security at the end of bucket. */ sbi->security.next_off = - ALIGN(sds_size - SecurityDescriptorsBlockSize, 16); + ALIGN(sds_size - SecurityDescriptorsBlockSize, 16); off = 0; ne = NULL; diff --git a/fs/ntfs3/index.c b/fs/ntfs3/index.c index 1224b8e42b3e..0daca9adc54c 100644 --- a/fs/ntfs3/index.c +++ b/fs/ntfs3/index.c @@ -2624,17 +2624,19 @@ int indx_update_dup(struct ntfs_inode *ni, struct ntfs_sb_info *sbi, e_fname = (struct ATTR_FILE_NAME *)(e + 1); if (!memcmp(&e_fname->dup, dup, sizeof(*dup))) { - /* Nothing to update in index! Try to avoid this call. */ + /* + * Nothing to update in index! Try to avoid this call. + */ goto out; } memcpy(&e_fname->dup, dup, sizeof(*dup)); if (fnd->level) { - /* directory entry in index */ + /* Directory entry in index. */ err = indx_write(indx, ni, fnd->nodes[fnd->level - 1], sync); } else { - /* directory entry in directory MFT record */ + /* Directory entry in directory MFT record. */ mi->dirty = true; if (sync) err = mi_write(mi, 1); diff --git a/fs/ntfs3/inode.c b/fs/ntfs3/inode.c index 8f72066b3229..db2a5a4c38e4 100644 --- a/fs/ntfs3/inode.c +++ b/fs/ntfs3/inode.c @@ -89,7 +89,7 @@ static struct inode *ntfs_read_mft(struct inode *inode, } if (le32_to_cpu(rec->total) != sbi->record_size) { - // Bad inode? + /* Bad inode? */ err = -EINVAL; goto out; } @@ -605,7 +605,7 @@ static noinline int ntfs_get_block_vbo(struct inode *inode, u64 vbo, if (vbo >= valid) set_buffer_new(bh); } else if (create) { - /*normal write*/ + /* Normal write. */ if (bytes > bh->b_size) bytes = bh->b_size; @@ -1091,7 +1091,7 @@ int inode_write_data(struct inode *inode, const void *data, size_t bytes) /* * ntfs_reparse_bytes * - * Number of bytes to for REPARSE_DATA_BUFFER(IO_REPARSE_TAG_SYMLINK) + * Number of bytes for REPARSE_DATA_BUFFER(IO_REPARSE_TAG_SYMLINK) * for unicode string of @uni_len length. */ static inline u32 ntfs_reparse_bytes(u32 uni_len) @@ -1205,13 +1205,13 @@ struct inode *ntfs_create_inode(struct user_namespace *mnt_userns, return ERR_PTR(-EINVAL); if (S_ISDIR(mode)) { - /* use parent's directory attributes */ + /* Use parent's directory attributes. */ fa = dir_ni->std_fa | FILE_ATTRIBUTE_DIRECTORY | FILE_ATTRIBUTE_ARCHIVE; /* - * By default child directory inherits parent attributes - * root directory is hidden + system - * Make an exception for children in root + * By default child directory inherits parent attributes. + * Root directory is hidden + system. + * Make an exception for children in root. */ if (dir->i_ino == MFT_REC_ROOT) fa &= ~(FILE_ATTRIBUTE_HIDDEN | FILE_ATTRIBUTE_SYSTEM); @@ -1220,8 +1220,8 @@ struct inode *ntfs_create_inode(struct user_namespace *mnt_userns, fa = FILE_ATTRIBUTE_REPARSE_POINT; /* - * linux: there are dir/file/symlink and so on. - * NTFS: symlinks are "dir + reparse" or "file + reparse". + * Linux: there are dir/file/symlink and so on. + * NTFS: symlinks are "dir + reparse" or "file + reparse" * It is good idea to create: * dir + reparse if 'symname' points to directory * or @@ -1860,7 +1860,7 @@ static noinline int ntfs_readlink_hlp(struct inode *inode, char *buffer, default: if (IsReparseTagMicrosoft(rp->ReparseTag)) { - /* unknown Microsoft Tag */ + /* Unknown Microsoft Tag. */ goto out; } if (!IsReparseTagNameSurrogate(rp->ReparseTag) || diff --git a/fs/ntfs3/lznt.c b/fs/ntfs3/lznt.c index 3acf0d9f0b15..f1f691a67cc4 100644 --- a/fs/ntfs3/lznt.c +++ b/fs/ntfs3/lznt.c @@ -296,8 +296,9 @@ static inline ssize_t decompress_chunk(u8 *unc, u8 *unc_end, const u8 *cmpr, */ struct lznt *get_lznt_ctx(int level) { - struct lznt *r = kzalloc(level ? offsetof(struct lznt, hash) : - sizeof(struct lznt), GFP_NOFS); + struct lznt *r = kzalloc(level ? offsetof(struct lznt, hash) + : sizeof(struct lznt), + GFP_NOFS); if (r) r->std = !level; diff --git a/fs/ntfs3/ntfs.h b/fs/ntfs3/ntfs.h index 0fd7bffb98d4..6bb3e595263b 100644 --- a/fs/ntfs3/ntfs.h +++ b/fs/ntfs3/ntfs.h @@ -262,7 +262,7 @@ enum RECORD_FLAG { RECORD_FLAG_UNKNOWN = cpu_to_le16(0x0008), }; -/* MFT Record structure, */ +/* MFT Record structure. */ struct MFT_REC { struct NTFS_RECORD_HEADER rhdr; // 'FILE' diff --git a/fs/ntfs3/ntfs_fs.h b/fs/ntfs3/ntfs_fs.h index f9436cbbc347..97e682ebcfb9 100644 --- a/fs/ntfs3/ntfs_fs.h +++ b/fs/ntfs3/ntfs_fs.h @@ -59,18 +59,18 @@ struct ntfs_mount_options { u16 fs_fmask_inv; u16 fs_dmask_inv; - unsigned uid : 1, /* uid was set. */ - gid : 1, /* gid was set. */ - fmask : 1, /* fmask was set. */ - dmask : 1, /* dmask was set. */ - sys_immutable : 1,/* Immutable system files. */ - discard : 1, /* Issue discard requests on deletions. */ - sparse : 1, /* Create sparse files. */ - showmeta : 1, /* Show meta files. */ - nohidden : 1, /* Do not show hidden files. */ - force : 1, /* Rw mount dirty volume. */ - no_acs_rules : 1,/*Exclude acs rules. */ - prealloc : 1 /* Preallocate space when file is growing. */ + unsigned uid : 1, /* uid was set. */ + gid : 1, /* gid was set. */ + fmask : 1, /* fmask was set. */ + dmask : 1, /* dmask was set. */ + sys_immutable : 1, /* Immutable system files. */ + discard : 1, /* Issue discard requests on deletions. */ + sparse : 1, /* Create sparse files. */ + showmeta : 1, /* Show meta files. */ + nohidden : 1, /* Do not show hidden files. */ + force : 1, /* Rw mount dirty volume. */ + no_acs_rules : 1, /*Exclude acs rules. */ + prealloc : 1 /* Preallocate space when file is growing. */ ; }; diff --git a/fs/ntfs3/record.c b/fs/ntfs3/record.c index 61e3f2fb619f..103705c86772 100644 --- a/fs/ntfs3/record.c +++ b/fs/ntfs3/record.c @@ -219,7 +219,7 @@ struct ATTRIB *mi_enum_attr(struct mft_inode *mi, struct ATTRIB *attr) asize = le32_to_cpu(attr->size); if (asize < SIZEOF_RESIDENT) { - /* Impossible 'cause we should not return such attribute */ + /* Impossible 'cause we should not return such attribute. */ return NULL; } diff --git a/fs/ntfs3/super.c b/fs/ntfs3/super.c index 2fbab8a931ee..dbecf095da59 100644 --- a/fs/ntfs3/super.c +++ b/fs/ntfs3/super.c @@ -1053,7 +1053,7 @@ static int ntfs_fill_super(struct super_block *sb, void *data, int silent) iput(inode); - /* Load $LogFile to replay. */ + /* Load LogFile to replay. */ ref.low = cpu_to_le32(MFT_REC_LOG); ref.seq = cpu_to_le16(MFT_REC_LOG); inode = ntfs_iget5(sb, &ref, &NAME_LOGFILE); diff --git a/fs/ntfs3/upcase.c b/fs/ntfs3/upcase.c index eb65bbd939e8..bbeba778237e 100644 --- a/fs/ntfs3/upcase.c +++ b/fs/ntfs3/upcase.c @@ -34,7 +34,7 @@ static inline u16 upcase_unicode_char(const u16 *upcase, u16 chr) * - Case insensitive * - If name equals and 'bothcases' then * - Case sensitive - * 'Straigth way' code scans input names twice in worst case. + * 'Straight way' code scans input names twice in worst case. * Optimized code scans input names only once. */ int ntfs_cmp_names(const __le16 *s1, size_t l1, const __le16 *s2, size_t l2, diff --git a/fs/ntfs3/xattr.c b/fs/ntfs3/xattr.c index 22fd5eb32c5b..b15d532e4a17 100644 --- a/fs/ntfs3/xattr.c +++ b/fs/ntfs3/xattr.c @@ -26,9 +26,10 @@ static inline size_t unpacked_ea_size(const struct EA_FULL *ea) { return ea->size ? le32_to_cpu(ea->size) - : ALIGN(struct_size( - ea, name, - 1 + ea->name_len + le16_to_cpu(ea->elength)), 4); + : ALIGN(struct_size(ea, name, + 1 + ea->name_len + + le16_to_cpu(ea->elength)), + 4); } static inline size_t packed_ea_size(const struct EA_FULL *ea) -- 2.30.0

1 0

[PATCH OLK-5.10 026/107] fs/ntfs3: Fix error handling in indx_insert_into_root()
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Dan Carpenter <dan.carpenter(a)oracle.com> mainline inclusion from mainline-v5.15 commit b8155e95de38b25a69dfb03e4731fd6c5a28531e category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- There are three bugs in this code: 1) If indx_get_root() fails, then return -EINVAL instead of success. 2) On the "/* make root external */" -EOPNOTSUPP; error path it should free "re" but it has a memory leak. 3) If indx_new() fails then it will lead to an error pointer dereference when we call put_indx_node(). I've re-written the error handling to be more clear. Fixes: 82cae269cfa9 ("fs/ntfs3: Add initialization of super block") Signed-off-by: Dan Carpenter <dan.carpenter(a)oracle.com> Reviewed-by: Kari Argillander <kari.argillander(a)gmail.com> Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/index.c | 36 ++++++++++++++++-------------------- 1 file changed, 16 insertions(+), 20 deletions(-) diff --git a/fs/ntfs3/index.c b/fs/ntfs3/index.c index f4729aa50671..69c6c4e0b4d9 100644 --- a/fs/ntfs3/index.c +++ b/fs/ntfs3/index.c @@ -1555,12 +1555,12 @@ static int indx_insert_into_root(struct ntfs_index *indx, struct ntfs_inode *ni, u32 root_size, new_root_size; struct ntfs_sb_info *sbi; int ds_root; - struct INDEX_ROOT *root, *a_root = NULL; + struct INDEX_ROOT *root, *a_root; /* Get the record this root placed in */ root = indx_get_root(indx, ni, &attr, &mi); if (!root) - goto out; + return -EINVAL; /* * Try easy case: @@ -1592,10 +1592,8 @@ static int indx_insert_into_root(struct ntfs_index *indx, struct ntfs_inode *ni, /* Make a copy of root attribute to restore if error */ a_root = kmemdup(attr, asize, GFP_NOFS); - if (!a_root) { - err = -ENOMEM; - goto out; - } + if (!a_root) + return -ENOMEM; /* copy all the non-end entries from the index root to the new buffer.*/ to_move = 0; @@ -1605,7 +1603,7 @@ static int indx_insert_into_root(struct ntfs_index *indx, struct ntfs_inode *ni, for (e = e0;; e = hdr_next_de(hdr, e)) { if (!e) { err = -EINVAL; - goto out; + goto out_free_root; } if (de_is_last(e)) @@ -1613,14 +1611,13 @@ static int indx_insert_into_root(struct ntfs_index *indx, struct ntfs_inode *ni, to_move += le16_to_cpu(e->size); } - n = NULL; if (!to_move) { re = NULL; } else { re = kmemdup(e0, to_move, GFP_NOFS); if (!re) { err = -ENOMEM; - goto out; + goto out_free_root; } } @@ -1637,7 +1634,7 @@ static int indx_insert_into_root(struct ntfs_index *indx, struct ntfs_inode *ni, if (ds_root > 0 && used + ds_root > sbi->max_bytes_per_attr) { /* make root external */ err = -EOPNOTSUPP; - goto out; + goto out_free_re; } if (ds_root) @@ -1667,7 +1664,7 @@ static int indx_insert_into_root(struct ntfs_index *indx, struct ntfs_inode *ni, /* bug? */ ntfs_set_state(sbi, NTFS_DIRTY_ERROR); err = -EINVAL; - goto out1; + goto out_free_re; } if (err) { @@ -1678,7 +1675,7 @@ static int indx_insert_into_root(struct ntfs_index *indx, struct ntfs_inode *ni, /* bug? */ ntfs_set_state(sbi, NTFS_DIRTY_ERROR); } - goto out1; + goto out_free_re; } e = (struct NTFS_DE *)(root + 1); @@ -1689,7 +1686,7 @@ static int indx_insert_into_root(struct ntfs_index *indx, struct ntfs_inode *ni, n = indx_new(indx, ni, new_vbn, sub_vbn); if (IS_ERR(n)) { err = PTR_ERR(n); - goto out1; + goto out_free_re; } hdr = &n->index->ihdr; @@ -1716,7 +1713,7 @@ static int indx_insert_into_root(struct ntfs_index *indx, struct ntfs_inode *ni, put_indx_node(n); fnd_clear(fnd); err = indx_insert_entry(indx, ni, new_de, ctx, fnd); - goto out; + goto out_free_root; } /* @@ -1726,7 +1723,7 @@ static int indx_insert_into_root(struct ntfs_index *indx, struct ntfs_inode *ni, e = hdr_insert_de(indx, hdr, new_de, NULL, ctx); if (!e) { err = -EINVAL; - goto out1; + goto out_put_n; } fnd_push(fnd, n, e); @@ -1735,12 +1732,11 @@ static int indx_insert_into_root(struct ntfs_index *indx, struct ntfs_inode *ni, n = NULL; -out1: +out_put_n: + put_indx_node(n); +out_free_re: kfree(re); - if (n) - put_indx_node(n); - -out: +out_free_root: kfree(a_root); return err; } -- 2.30.0

1 0

[PATCH OLK-5.10 025/107] fs/ntfs3: Potential NULL dereference in hdr_find_split()
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Dan Carpenter <dan.carpenter(a)oracle.com> mainline inclusion from mainline-v5.15 commit 8c83a4851da1c7eda83098ade238665b15774da3 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- The "e" pointer is dereferenced before it has been checked for NULL. Move the dereference after the NULL check to prevent an Oops. Fixes: 82cae269cfa9 ("fs/ntfs3: Add initialization of super block") Signed-off-by: Dan Carpenter <dan.carpenter(a)oracle.com> Reviewed-by: Kari Argillander <kari.argillander(a)gmail.com> Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/index.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/fs/ntfs3/index.c b/fs/ntfs3/index.c index 5fb41c9c8910..f4729aa50671 100644 --- a/fs/ntfs3/index.c +++ b/fs/ntfs3/index.c @@ -557,11 +557,12 @@ static const struct NTFS_DE *hdr_find_split(const struct INDEX_HDR *hdr) size_t o; const struct NTFS_DE *e = hdr_first_de(hdr); u32 used_2 = le32_to_cpu(hdr->used) >> 1; - u16 esize = le16_to_cpu(e->size); + u16 esize; if (!e || de_is_last(e)) return NULL; + esize = le16_to_cpu(e->size); for (o = le32_to_cpu(hdr->de_off) + esize; o < used_2; o += esize) { const struct NTFS_DE *p = e; -- 2.30.0

1 0

[PATCH OLK-5.10 024/107] fs/ntfs3: Fix error code in indx_add_allocate()
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Dan Carpenter <dan.carpenter(a)oracle.com> mainline inclusion from mainline-v5.15 commit 04810f000afdbdd37825ca7f563f036119422cb7 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- Return -EINVAL if ni_find_attr() fails. Don't return success. Fixes: 82cae269cfa9 ("fs/ntfs3: Add initialization of super block") Signed-off-by: Dan Carpenter <dan.carpenter(a)oracle.com> Reviewed-by: Kari Argillander <kari.argillander(a)gmail.com> Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/index.c | 1 + 1 file changed, 1 insertion(+) diff --git a/fs/ntfs3/index.c b/fs/ntfs3/index.c index 334a3cef714b..5fb41c9c8910 100644 --- a/fs/ntfs3/index.c +++ b/fs/ntfs3/index.c @@ -1500,6 +1500,7 @@ static int indx_add_allocate(struct ntfs_index *indx, struct ntfs_inode *ni, alloc = ni_find_attr(ni, NULL, NULL, ATTR_ALLOC, in->name, in->name_len, NULL, &mi); if (!alloc) { + err = -EINVAL; if (bmp) goto out2; goto out1; -- 2.30.0

1 0

[PATCH OLK-5.10 023/107] fs/ntfs3: fix an error code in ntfs_get_acl_ex()
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Dan Carpenter <dan.carpenter(a)oracle.com> mainline inclusion from mainline-v5.15 commit 2926e4297053c735ab65450192dfba32a4f47fa9 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- The ntfs_get_ea() function returns negative error codes or on success it returns the length. In the original code a zero length return was treated as -ENODATA and results in a NULL return. But it should be treated as an invalid length and result in an PTR_ERR(-EINVAL) return. Fixes: be71b5cba2e6 ("fs/ntfs3: Add attrib operations") Signed-off-by: Dan Carpenter <dan.carpenter(a)oracle.com> Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/xattr.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/ntfs3/xattr.c b/fs/ntfs3/xattr.c index af89e50f7b9f..d3d5b9d331d1 100644 --- a/fs/ntfs3/xattr.c +++ b/fs/ntfs3/xattr.c @@ -521,7 +521,7 @@ static struct posix_acl *ntfs_get_acl_ex(struct user_namespace *mnt_userns, ni_unlock(ni); /* Translate extended attribute to acl */ - if (err > 0) { + if (err >= 0) { acl = posix_acl_from_xattr(mnt_userns, buf, err); if (!IS_ERR(acl)) set_cached_acl(inode, type, acl); -- 2.30.0

1 0

[PATCH OLK-5.10 022/107] fs/ntfs3: add checks for allocation failure
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Dan Carpenter <dan.carpenter(a)oracle.com> mainline inclusion from mainline-v5.15 commit a1b04d380ab64790a7b4a8eb52e14679e47065ab category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- Add a check for when the kzalloc() in init_rsttbl() fails. Some of the callers checked for NULL and some did not. I went down the call tree and added NULL checks where ever they were missing. Fixes: b46acd6a6a62 ("fs/ntfs3: Add NTFS journal") Signed-off-by: Dan Carpenter <dan.carpenter(a)oracle.com> Reviewed-by: Kari Argillander <kari.argillander(a)gmail.com> Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/fslog.c | 18 ++++++++++++++++-- 1 file changed, 16 insertions(+), 2 deletions(-) diff --git a/fs/ntfs3/fslog.c b/fs/ntfs3/fslog.c index 2c213b55979e..7144ea8a9ab8 100644 --- a/fs/ntfs3/fslog.c +++ b/fs/ntfs3/fslog.c @@ -809,6 +809,9 @@ static inline struct RESTART_TABLE *init_rsttbl(u16 esize, u16 used) u32 lf = sizeof(struct RESTART_TABLE) + (used - 1) * esize; struct RESTART_TABLE *t = kzalloc(bytes, GFP_NOFS); + if (!t) + return NULL; + t->size = cpu_to_le16(esize); t->used = cpu_to_le16(used); t->free_goal = cpu_to_le32(~0u); @@ -831,7 +834,11 @@ static inline struct RESTART_TABLE *extend_rsttbl(struct RESTART_TABLE *tbl, u16 esize = le16_to_cpu(tbl->size); __le32 osize = cpu_to_le32(bytes_per_rt(tbl)); u32 used = le16_to_cpu(tbl->used); - struct RESTART_TABLE *rt = init_rsttbl(esize, used + add); + struct RESTART_TABLE *rt; + + rt = init_rsttbl(esize, used + add); + if (!rt) + return NULL; memcpy(rt + 1, tbl + 1, esize * used); @@ -864,8 +871,11 @@ static inline void *alloc_rsttbl_idx(struct RESTART_TABLE **tbl) __le32 *e; struct RESTART_TABLE *t = *tbl; - if (!t->first_free) + if (!t->first_free) { *tbl = t = extend_rsttbl(t, 16, ~0u); + if (!t) + return NULL; + } off = le32_to_cpu(t->first_free); @@ -4482,6 +4492,10 @@ int log_replay(struct ntfs_inode *ni, bool *initialized) } dp = alloc_rsttbl_idx(&dptbl); + if (!dp) { + err = -ENOMEM; + goto out; + } dp->target_attr = cpu_to_le32(t16); dp->transfer_len = cpu_to_le32(t32 << sbi->cluster_bits); dp->lcns_follow = cpu_to_le32(t32); -- 2.30.0

1 0

[PATCH OLK-5.10 021/107] fs/ntfs3: Use kcalloc/kmalloc_array over kzalloc/kmalloc
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Kari Argillander <kari.argillander(a)gmail.com> mainline inclusion from mainline-v5.15 commit 345482bc431f6492beb464696341626057f67771 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- Use kcalloc/kmalloc_array over kzalloc/kmalloc when we allocate array. Checkpatch found these after we did not use our own defined allocation wrappers. Reviewed-by: Christoph Hellwig <hch(a)lst.de> Signed-off-by: Kari Argillander <kari.argillander(a)gmail.com> Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/bitmap.c | 2 +- fs/ntfs3/file.c | 2 +- fs/ntfs3/frecord.c | 7 +++---- 3 files changed, 5 insertions(+), 6 deletions(-) diff --git a/fs/ntfs3/bitmap.c b/fs/ntfs3/bitmap.c index d502bba323d0..2de05062c78b 100644 --- a/fs/ntfs3/bitmap.c +++ b/fs/ntfs3/bitmap.c @@ -683,7 +683,7 @@ int wnd_init(struct wnd_bitmap *wnd, struct super_block *sb, size_t nbits) if (!wnd->bits_last) wnd->bits_last = wbits; - wnd->free_bits = kzalloc(wnd->nwnd * sizeof(u16), GFP_NOFS); + wnd->free_bits = kcalloc(wnd->nwnd, sizeof(u16), GFP_NOFS); if (!wnd->free_bits) return -ENOMEM; diff --git a/fs/ntfs3/file.c b/fs/ntfs3/file.c index 8d27c520bec5..a959f6197c99 100644 --- a/fs/ntfs3/file.c +++ b/fs/ntfs3/file.c @@ -900,7 +900,7 @@ static ssize_t ntfs_compress_write(struct kiocb *iocb, struct iov_iter *from) return -EOPNOTSUPP; } - pages = kmalloc(pages_per_frame * sizeof(struct page *), GFP_NOFS); + pages = kmalloc_array(pages_per_frame, sizeof(struct page *), GFP_NOFS); if (!pages) return -ENOMEM; diff --git a/fs/ntfs3/frecord.c b/fs/ntfs3/frecord.c index 2f7d16543530..329bc76dfb09 100644 --- a/fs/ntfs3/frecord.c +++ b/fs/ntfs3/frecord.c @@ -2054,7 +2054,7 @@ int ni_readpage_cmpr(struct ntfs_inode *ni, struct page *page) idx = (vbo - frame_vbo) >> PAGE_SHIFT; pages_per_frame = frame_size >> PAGE_SHIFT; - pages = kzalloc(pages_per_frame * sizeof(struct page *), GFP_NOFS); + pages = kcalloc(pages_per_frame, sizeof(struct page *), GFP_NOFS); if (!pages) { err = -ENOMEM; goto out; @@ -2137,7 +2137,7 @@ int ni_decompress_file(struct ntfs_inode *ni) frame_bits = ni_ext_compress_bits(ni); frame_size = 1u << frame_bits; pages_per_frame = frame_size >> PAGE_SHIFT; - pages = kzalloc(pages_per_frame * sizeof(struct page *), GFP_NOFS); + pages = kcalloc(pages_per_frame, sizeof(struct page *), GFP_NOFS); if (!pages) { err = -ENOMEM; goto out; @@ -2709,8 +2709,7 @@ int ni_write_frame(struct ntfs_inode *ni, struct page **pages, goto out; } - pages_disk = kzalloc(pages_per_frame * sizeof(struct page *), - GFP_NOFS); + pages_disk = kcalloc(pages_per_frame, sizeof(struct page *), GFP_NOFS); if (!pages_disk) { err = -ENOMEM; goto out; -- 2.30.0

1 0

[PATCH OLK-5.10 017/107] fs/ntfs3: Remove unused including <linux/version.h>
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Jiapeng Chong <jiapeng.chong(a)linux.alibaba.com> mainline inclusion from mainline-v5.15 commit 1263eddfea9988125a4b9608efecc8aff2c721f9 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- Eliminate the follow versioncheck warning: ./fs/ntfs3/inode.c: 16 linux/version.h not needed. Reported-by: Abaci Robot <abaci(a)linux.alibaba.com> Fixes: 82cae269cfa9 ("fs/ntfs3: Add initialization of super block") Signed-off-by: Jiapeng Chong <jiapeng.chong(a)linux.alibaba.com> Reviewed-by: Kari Argillander <kari.argillander(a)gmail.com> Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/inode.c | 1 - 1 file changed, 1 deletion(-) diff --git a/fs/ntfs3/inode.c b/fs/ntfs3/inode.c index a573c6e98cb8..ed64489edf73 100644 --- a/fs/ntfs3/inode.c +++ b/fs/ntfs3/inode.c @@ -13,7 +13,6 @@ #include <linux/namei.h> #include <linux/nls.h> #include <linux/uio.h> -#include <linux/version.h> #include <linux/writeback.h> #include "debug.h" -- 2.30.0

1 0

[PATCH OLK-5.10 016/107] fs/ntfs3: Fix fall-through warnings for Clang
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: "Gustavo A. R. Silva" <gustavoars(a)kernel.org> mainline inclusion from mainline-v5.15 commit abfeb2ee2103f07dd93b9d7b32317e26d1c8ef79 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- Fix the following fallthrough warnings: fs/ntfs3/inode.c:1792:2: warning: unannotated fall-through between switch labels [-Wimplicit-fallthrough] fs/ntfs3/index.c:178:2: warning: unannotated fall-through between switch labels [-Wimplicit-fallthrough] This helps with the ongoing efforts to globally enable -Wimplicit-fallthrough for Clang. Link: https://github.com/KSPP/linux/issues/115 Signed-off-by: Gustavo A. R. Silva <gustavoars(a)kernel.org> Reviewed-by: Nathan Chancellor <nathan(a)kernel.org> Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/index.c | 1 + fs/ntfs3/inode.c | 1 + 2 files changed, 2 insertions(+) diff --git a/fs/ntfs3/index.c b/fs/ntfs3/index.c index 9386c551e208..189d46e2c38d 100644 --- a/fs/ntfs3/index.c +++ b/fs/ntfs3/index.c @@ -175,6 +175,7 @@ static inline NTFS_CMP_FUNC get_cmp_func(const struct INDEX_ROOT *root) default: break; } + break; default: break; } diff --git a/fs/ntfs3/inode.c b/fs/ntfs3/inode.c index bf51e294432e..a573c6e98cb8 100644 --- a/fs/ntfs3/inode.c +++ b/fs/ntfs3/inode.c @@ -1789,6 +1789,7 @@ int ntfs_unlink_inode(struct inode *dir, const struct dentry *dentry) switch (err) { case 0: drop_nlink(inode); + break; case -ENOTEMPTY: case -ENOSPC: case -EROFS: -- 2.30.0

1 0

[PATCH OLK-5.10 015/107] fs/ntfs3: Fix one none utf8 char in source file
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Kari Argillander <kari.argillander(a)gmail.com> mainline inclusion from mainline-v5.15 commit be87e821fdb5ec8c6d404f29e118130c7879ce5b category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- In one source file there is for some reason non utf8 char. But hey this is fs development so this kind of thing might happen. Signed-off-by: Kari Argillander <kari.argillander(a)gmail.com> Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/frecord.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/ntfs3/frecord.c b/fs/ntfs3/frecord.c index c3121bf9c62f..e377d72477df 100644 --- a/fs/ntfs3/frecord.c +++ b/fs/ntfs3/frecord.c @@ -1784,7 +1784,7 @@ enum REPARSE_SIGN ni_parse_reparse(struct ntfs_inode *ni, struct ATTRIB *attr, /* * WOF - Windows Overlay Filter - used to compress files with lzx/xpress * Unlike native NTFS file compression, the Windows Overlay Filter supports - * only read operations. This means that it doesn�t need to sector-align each + * only read operations. This means that it doesn't need to sector-align each * compressed chunk, so the compressed data can be packed more tightly together. * If you open the file for writing, the Windows Overlay Filter just decompresses * the entire file, turning it back into a plain file. -- 2.30.0

1 0

[PATCH OLK-5.10 014/107] fs/ntfs3: Remove unused variable cnt in ntfs_security_init()
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Nathan Chancellor <nathan(a)kernel.org> mainline inclusion from mainline-v5.15 commit 8c01308b6d6b2bc8e9163c6a3400856fb782dee6 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- Clang warns: fs/ntfs3/fsntfs.c:1874:9: warning: variable 'cnt' set but not used [-Wunused-but-set-variable] size_t cnt, off; ^ 1 warning generated. It is indeed unused so remove it. Fixes: 82cae269cfa9 ("fs/ntfs3: Add initialization of super block") Signed-off-by: Nathan Chancellor <nathan(a)kernel.org> Reviewed-by: Nick Desaulniers <ndesaulniers(a)google.com> Reviewed-by: Kari Argillander <kari.argillander(a)gmail.com> Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/fsntfs.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/fs/ntfs3/fsntfs.c b/fs/ntfs3/fsntfs.c index 92140050fb6c..c6599c514acf 100644 --- a/fs/ntfs3/fsntfs.c +++ b/fs/ntfs3/fsntfs.c @@ -1871,7 +1871,7 @@ int ntfs_security_init(struct ntfs_sb_info *sbi) struct ATTRIB *attr; struct ATTR_LIST_ENTRY *le; u64 sds_size; - size_t cnt, off; + size_t off; struct NTFS_DE *ne; struct NTFS_DE_SII *sii_e; struct ntfs_fnd *fnd_sii = NULL; @@ -1946,7 +1946,6 @@ int ntfs_security_init(struct ntfs_sb_info *sbi) sbi->security.next_off = Quad2Align(sds_size - SecurityDescriptorsBlockSize); - cnt = 0; off = 0; ne = NULL; @@ -1964,8 +1963,6 @@ int ntfs_security_init(struct ntfs_sb_info *sbi) next_id = le32_to_cpu(sii_e->sec_id) + 1; if (next_id >= sbi->security.next_id) sbi->security.next_id = next_id; - - cnt += 1; } sbi->security.ni = ni; -- 2.30.0

1 0

[PATCH OLK-5.10 013/107] fs/ntfs3: Fix integer overflow in multiplication
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Colin Ian King <colin.king(a)canonical.com> mainline inclusion from mainline-v5.15 commit 71eeb6ace80be7389d942b9647765417e5b039f7 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- The multiplication of the u32 data_size with a int is being performed using 32 bit arithmetic however the results is being assigned to the variable nbits that is a size_t (64 bit) value. Fix a potential integer overflow by casting the u32 value to a size_t before the multiply to use a size_t sized bit multiply operation. Addresses-Coverity: ("Unintentional integer overflow") Fixes: 82cae269cfa9 ("fs/ntfs3: Add initialization of super block") Signed-off-by: Colin Ian King <colin.king(a)canonical.com> Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/index.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/ntfs3/index.c b/fs/ntfs3/index.c index 6aa9540ece47..9386c551e208 100644 --- a/fs/ntfs3/index.c +++ b/fs/ntfs3/index.c @@ -2012,7 +2012,7 @@ static int indx_shrink(struct ntfs_index *indx, struct ntfs_inode *ni, unsigned long pos; const unsigned long *bm = resident_data(b); - nbits = le32_to_cpu(b->res.data_size) * 8; + nbits = (size_t)le32_to_cpu(b->res.data_size) * 8; if (bit >= nbits) return 0; -- 2.30.0

1 0

[PATCH OLK-5.10 012/107] fs/ntfs3: Add ifndef + define to all header files
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Kari Argillander <kari.argillander(a)gmail.com> mainline inclusion from mainline-v5.15 commit 87790b65343932411af43bc9b218f086ecebd6a5 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- Add guards so that compiler will only include header files once. Signed-off-by: Kari Argillander <kari.argillander(a)gmail.com> Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/debug.h | 5 +++++ fs/ntfs3/ntfs.h | 3 +++ fs/ntfs3/ntfs_fs.h | 5 +++++ 3 files changed, 13 insertions(+) diff --git a/fs/ntfs3/debug.h b/fs/ntfs3/debug.h index 15ac42185e5b..357d9f4dfba7 100644 --- a/fs/ntfs3/debug.h +++ b/fs/ntfs3/debug.h @@ -7,6 +7,9 @@ */ // clang-format off +#ifndef _LINUX_NTFS3_DEBUG_H +#define _LINUX_NTFS3_DEBUG_H + #ifndef Add2Ptr #define Add2Ptr(P, I) ((void *)((u8 *)(P) + (I))) #define PtrOffset(B, O) ((size_t)((size_t)(O) - (size_t)(B))) @@ -61,4 +64,6 @@ void ntfs_inode_printk(struct inode *inode, const char *fmt, ...) #define ntfs_free(p) kfree(p) #define ntfs_vfree(p) kvfree(p) #define ntfs_memdup(src, len) kmemdup(src, len, GFP_NOFS) + +#endif /* _LINUX_NTFS3_DEBUG_H */ // clang-format on diff --git a/fs/ntfs3/ntfs.h b/fs/ntfs3/ntfs.h index 40398e6c39c9..16da514af124 100644 --- a/fs/ntfs3/ntfs.h +++ b/fs/ntfs3/ntfs.h @@ -7,6 +7,8 @@ */ // clang-format off +#ifndef _LINUX_NTFS3_NTFS_H +#define _LINUX_NTFS3_NTFS_H /* TODO: * - Check 4K mft record and 512 bytes cluster @@ -1235,4 +1237,5 @@ struct SID { }; static_assert(offsetof(struct SID, SubAuthority) == 8); +#endif /* _LINUX_NTFS3_NTFS_H */ // clang-format on diff --git a/fs/ntfs3/ntfs_fs.h b/fs/ntfs3/ntfs_fs.h index c8ea6dd38c21..b5da2f06f7cb 100644 --- a/fs/ntfs3/ntfs_fs.h +++ b/fs/ntfs3/ntfs_fs.h @@ -6,6 +6,9 @@ */ // clang-format off +#ifndef _LINUX_NTFS3_NTFS_FS_H +#define _LINUX_NTFS3_NTFS_FS_H + #define MINUS_ONE_T ((size_t)(-1)) /* Biggest MFT / smallest cluster */ #define MAXIMUM_BYTES_PER_MFT 4096 @@ -1085,3 +1088,5 @@ static inline void le64_sub_cpu(__le64 *var, u64 val) { *var = cpu_to_le64(le64_to_cpu(*var) - val); } + +#endif /* _LINUX_NTFS3_NTFS_FS_H */ -- 2.30.0

1 0

[PATCH OLK-5.10 011/107] fs/ntfs3: Use linux/log2 is_power_of_2 function
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Kari Argillander <kari.argillander(a)gmail.com> mainline inclusion from mainline-v5.15 commit 528c9b3d1edf291685151afecd741d176f527ddf category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- We do not need our own implementation for this function in this driver. It is much better to use generic one. Signed-off-by: Kari Argillander <kari.argillander(a)gmail.com> Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/ntfs_fs.h | 5 ----- fs/ntfs3/run.c | 3 ++- fs/ntfs3/super.c | 9 +++++---- 3 files changed, 7 insertions(+), 10 deletions(-) diff --git a/fs/ntfs3/ntfs_fs.h b/fs/ntfs3/ntfs_fs.h index 0c3ac89c3115..c8ea6dd38c21 100644 --- a/fs/ntfs3/ntfs_fs.h +++ b/fs/ntfs3/ntfs_fs.h @@ -972,11 +972,6 @@ static inline struct buffer_head *ntfs_bread(struct super_block *sb, return NULL; } -static inline bool is_power_of2(size_t v) -{ - return v && !(v & (v - 1)); -} - static inline struct ntfs_inode *ntfs_i(struct inode *inode) { return container_of(inode, struct ntfs_inode, vfs_inode); diff --git a/fs/ntfs3/run.c b/fs/ntfs3/run.c index f9c362ac672e..60c64deab738 100644 --- a/fs/ntfs3/run.c +++ b/fs/ntfs3/run.c @@ -9,6 +9,7 @@ #include <linux/blkdev.h> #include <linux/buffer_head.h> #include <linux/fs.h> +#include <linux/log2.h> #include <linux/nls.h> #include "debug.h" @@ -376,7 +377,7 @@ bool run_add_entry(struct runs_tree *run, CLST vcn, CLST lcn, CLST len, if (!used) { bytes = 64; } else if (used <= 16 * PAGE_SIZE) { - if (is_power_of2(run->allocated)) + if (is_power_of_2(run->allocated)) bytes = run->allocated << 1; else bytes = (size_t)1 diff --git a/fs/ntfs3/super.c b/fs/ntfs3/super.c index 84d4f389f685..903975b7e832 100644 --- a/fs/ntfs3/super.c +++ b/fs/ntfs3/super.c @@ -29,6 +29,7 @@ #include <linux/exportfs.h> #include <linux/fs.h> #include <linux/iversion.h> +#include <linux/log2.h> #include <linux/module.h> #include <linux/nls.h> #include <linux/parser.h> @@ -735,13 +736,13 @@ static int ntfs_init_from_boot(struct super_block *sb, u32 sector_size, boot_sector_size = (u32)boot->bytes_per_sector[1] << 8; if (boot->bytes_per_sector[0] || boot_sector_size < SECTOR_SIZE || - !is_power_of2(boot_sector_size)) { + !is_power_of_2(boot_sector_size)) { goto out; } /* cluster size: 512, 1K, 2K, 4K, ... 2M */ sct_per_clst = true_sectors_per_clst(boot); - if (!is_power_of2(sct_per_clst)) + if (!is_power_of_2(sct_per_clst)) goto out; mlcn = le64_to_cpu(boot->mft_clst); @@ -757,14 +758,14 @@ static int ntfs_init_from_boot(struct super_block *sb, u32 sector_size, /* Check MFT record size */ if ((boot->record_size < 0 && SECTOR_SIZE > (2U << (-boot->record_size))) || - (boot->record_size >= 0 && !is_power_of2(boot->record_size))) { + (boot->record_size >= 0 && !is_power_of_2(boot->record_size))) { goto out; } /* Check index record size */ if ((boot->index_size < 0 && SECTOR_SIZE > (2U << (-boot->index_size))) || - (boot->index_size >= 0 && !is_power_of2(boot->index_size))) { + (boot->index_size >= 0 && !is_power_of_2(boot->index_size))) { goto out; } -- 2.30.0

1 0

[PATCH OLK-5.10 010/107] fs/ntfs3: Fix various spelling mistakes
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Colin Ian King <colin.king(a)canonical.com> mainline inclusion from mainline-v5.15 commit f8d87ed9f0d546ac5b05e8e7d2b148d4b77599fa category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- There is a spelling mistake in a ntfs_err error message. Also fix various spelling mistakes in comments. Signed-off-by: Colin Ian King <colin.king(a)canonical.com> Reviewed-by: Kari Argillander <kari.argillander(a)gmail.com> Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/debug.h | 2 +- fs/ntfs3/lib/decompress_common.c | 2 +- fs/ntfs3/run.c | 2 +- fs/ntfs3/super.c | 4 ++-- fs/ntfs3/upcase.c | 2 +- 5 files changed, 6 insertions(+), 6 deletions(-) diff --git a/fs/ntfs3/debug.h b/fs/ntfs3/debug.h index dfaa4c79dc6d..15ac42185e5b 100644 --- a/fs/ntfs3/debug.h +++ b/fs/ntfs3/debug.h @@ -3,7 +3,7 @@ * * Copyright (C) 2019-2021 Paragon Software GmbH, All rights reserved. * - * useful functions for debuging + * useful functions for debugging */ // clang-format off diff --git a/fs/ntfs3/lib/decompress_common.c b/fs/ntfs3/lib/decompress_common.c index 83c9e93aea77..850d8e8c8f1f 100644 --- a/fs/ntfs3/lib/decompress_common.c +++ b/fs/ntfs3/lib/decompress_common.c @@ -292,7 +292,7 @@ int make_huffman_decode_table(u16 decode_table[], const u32 num_syms, * of as simply the root of the tree. The * representation of these internal nodes is * simply the index of the left child combined - * with the special bits 0xC000 to distingush + * with the special bits 0xC000 to distinguish * the entry from direct mapping and leaf node * entries. */ diff --git a/fs/ntfs3/run.c b/fs/ntfs3/run.c index 5cdf6efe67e0..f9c362ac672e 100644 --- a/fs/ntfs3/run.c +++ b/fs/ntfs3/run.c @@ -949,7 +949,7 @@ int run_unpack(struct runs_tree *run, struct ntfs_sb_info *sbi, CLST ino, if (next_vcn > 0x100000000ull || (lcn + len) > 0x100000000ull) { ntfs_err( sbi->sb, - "This driver is compiled whitout CONFIG_NTFS3_64BIT_CLUSTER (like windows driver).\n" + "This driver is compiled without CONFIG_NTFS3_64BIT_CLUSTER (like windows driver).\n" "Volume contains 64 bits run: vcn %llx, lcn %llx, len %llx.\n" "Activate CONFIG_NTFS3_64BIT_CLUSTER to process this case", vcn64, lcn, len); diff --git a/fs/ntfs3/super.c b/fs/ntfs3/super.c index 6be13e256c1a..84d4f389f685 100644 --- a/fs/ntfs3/super.c +++ b/fs/ntfs3/super.c @@ -124,7 +124,7 @@ void ntfs_inode_printk(struct inode *inode, const char *fmt, ...) /* * Shared memory struct. * - * on-disk ntfs's upcase table is created by ntfs formater + * on-disk ntfs's upcase table is created by ntfs formatter * 'upcase' table is 128K bytes of memory * we should read it into memory when mounting * Several ntfs volumes likely use the same 'upcase' table @@ -1208,7 +1208,7 @@ static int ntfs_fill_super(struct super_block *sb, void *data, int silent) sbi->def_entries = 1; done = sizeof(struct ATTR_DEF_ENTRY); sbi->reparse.max_size = MAXIMUM_REPARSE_DATA_BUFFER_SIZE; - sbi->ea_max_size = 0x10000; /* default formater value */ + sbi->ea_max_size = 0x10000; /* default formatter value */ while (done + sizeof(struct ATTR_DEF_ENTRY) <= bytes) { u32 t32 = le32_to_cpu(t->type); diff --git a/fs/ntfs3/upcase.c b/fs/ntfs3/upcase.c index 9617382aca64..b53943538f9f 100644 --- a/fs/ntfs3/upcase.c +++ b/fs/ntfs3/upcase.c @@ -27,7 +27,7 @@ static inline u16 upcase_unicode_char(const u16 *upcase, u16 chr) /* * Thanks Kari Argillander <kari.argillander(a)gmail.com> for idea and implementation 'bothcase' * - * Straigth way to compare names: + * Straight way to compare names: * - case insensitive * - if name equals and 'bothcases' then * - case sensitive -- 2.30.0

1 0

[PATCH OLK-5.10 009/107] fs/ntfs3: Add NTFS3 in fs/Kconfig and fs/Makefile
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> mainline inclusion from mainline-v5.15 commit 6e5be40d32fb1907285277c02e74493ed43d77fe category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- This adds NTFS3 in fs/Kconfig and fs/Makefile Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/Kconfig | 1 + fs/Makefile | 1 + 2 files changed, 2 insertions(+) diff --git a/fs/Kconfig b/fs/Kconfig index 3cc647e00f3c..225088d505f4 100644 --- a/fs/Kconfig +++ b/fs/Kconfig @@ -145,6 +145,7 @@ menu "DOS/FAT/EXFAT/NT Filesystems" source "fs/fat/Kconfig" source "fs/exfat/Kconfig" source "fs/ntfs/Kconfig" +source "fs/ntfs3/Kconfig" endmenu endif # BLOCK diff --git a/fs/Makefile b/fs/Makefile index fec76c1b4e06..73acb48ce6bc 100644 --- a/fs/Makefile +++ b/fs/Makefile @@ -102,6 +102,7 @@ obj-$(CONFIG_SYSV_FS) += sysv/ obj-$(CONFIG_CIFS) += cifs/ obj-$(CONFIG_HPFS_FS) += hpfs/ obj-$(CONFIG_NTFS_FS) += ntfs/ +obj-$(CONFIG_NTFS3_FS) += ntfs3/ obj-$(CONFIG_UFS_FS) += ufs/ obj-$(CONFIG_EFS_FS) += efs/ obj-$(CONFIG_JFFS2_FS) += jffs2/ -- 2.30.0

1 0

[PATCH OLK-5.10 008/107] fs/ntfs3: Add Kconfig, Makefile and doc
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> mainline inclusion from mainline-v5.15 commit 12dad495eaab95e0bb784c43869073617c513ea4 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- This adds Kconfig, Makefile and doc Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- Documentation/filesystems/index.rst | 1 + Documentation/filesystems/ntfs3.rst | 106 ++++++++++++++++++++++++++++ fs/ntfs3/Kconfig | 46 ++++++++++++ fs/ntfs3/Makefile | 36 ++++++++++ 4 files changed, 189 insertions(+) create mode 100644 Documentation/filesystems/ntfs3.rst create mode 100644 fs/ntfs3/Kconfig create mode 100644 fs/ntfs3/Makefile diff --git a/Documentation/filesystems/index.rst b/Documentation/filesystems/index.rst index 98f59a864242..757684537248 100644 --- a/Documentation/filesystems/index.rst +++ b/Documentation/filesystems/index.rst @@ -97,6 +97,7 @@ Documentation for filesystem implementations. nilfs2 nfs/index ntfs + ntfs3 ocfs2 ocfs2-online-filecheck omfs diff --git a/Documentation/filesystems/ntfs3.rst b/Documentation/filesystems/ntfs3.rst new file mode 100644 index 000000000000..ffe9ea0c1499 --- /dev/null +++ b/Documentation/filesystems/ntfs3.rst @@ -0,0 +1,106 @@ +.. SPDX-License-Identifier: GPL-2.0 + +===== +NTFS3 +===== + + +Summary and Features +==================== + +NTFS3 is fully functional NTFS Read-Write driver. The driver works with +NTFS versions up to 3.1, normal/compressed/sparse files +and journal replaying. File system type to use on mount is 'ntfs3'. + +- This driver implements NTFS read/write support for normal, sparse and + compressed files. +- Supports native journal replaying; +- Supports extended attributes + Predefined extended attributes: + - 'system.ntfs_security' gets/sets security + descriptor (SECURITY_DESCRIPTOR_RELATIVE) + - 'system.ntfs_attrib' gets/sets ntfs file/dir attributes. + Note: applied to empty files, this allows to switch type between + sparse(0x200), compressed(0x800) and normal; +- Supports NFS export of mounted NTFS volumes. + +Mount Options +============= + +The list below describes mount options supported by NTFS3 driver in addition to +generic ones. + +=============================================================================== + +nls=name This option informs the driver how to interpret path + strings and translate them to Unicode and back. If + this option is not set, the default codepage will be + used (CONFIG_NLS_DEFAULT). + Examples: + 'nls=utf8' + +uid= +gid= +umask= Controls the default permissions for files/directories created + after the NTFS volume is mounted. + +fmask= +dmask= Instead of specifying umask which applies both to + files and directories, fmask applies only to files and + dmask only to directories. + +nohidden Files with the Windows-specific HIDDEN (FILE_ATTRIBUTE_HIDDEN) + attribute will not be shown under Linux. + +sys_immutable Files with the Windows-specific SYSTEM + (FILE_ATTRIBUTE_SYSTEM) attribute will be marked as system + immutable files. + +discard Enable support of the TRIM command for improved performance + on delete operations, which is recommended for use with the + solid-state drives (SSD). + +force Forces the driver to mount partitions even if 'dirty' flag + (volume dirty) is set. Not recommended for use. + +sparse Create new files as "sparse". + +showmeta Use this parameter to show all meta-files (System Files) on + a mounted NTFS partition. + By default, all meta-files are hidden. + +prealloc Preallocate space for files excessively when file size is + increasing on writes. Decreases fragmentation in case of + parallel write operations to different files. + +no_acs_rules "No access rules" mount option sets access rights for + files/folders to 777 and owner/group to root. This mount + option absorbs all other permissions: + - permissions change for files/folders will be reported + as successful, but they will remain 777; + - owner/group change will be reported as successful, but + they will stay as root + +acl Support POSIX ACLs (Access Control Lists). Effective if + supported by Kernel. Not to be confused with NTFS ACLs. + The option specified as acl enables support for POSIX ACLs. + +noatime All files and directories will not update their last access + time attribute if a partition is mounted with this parameter. + This option can speed up file system operation. + +=============================================================================== + +ToDo list +========= + +- Full journaling support (currently journal replaying is supported) over JBD. + + +References +========== +https://www.paragon-software.com/home/ntfs-linux-professional/ + - Commercial version of the NTFS driver for Linux. + +almaz.alexandrovich(a)paragon-software.com + - Direct e-mail address for feedback and requests on the NTFS3 implementation. diff --git a/fs/ntfs3/Kconfig b/fs/ntfs3/Kconfig new file mode 100644 index 000000000000..6e4cbc48ab8e --- /dev/null +++ b/fs/ntfs3/Kconfig @@ -0,0 +1,46 @@ +# SPDX-License-Identifier: GPL-2.0-only +config NTFS3_FS + tristate "NTFS Read-Write file system support" + select NLS + help + Windows OS native file system (NTFS) support up to NTFS version 3.1. + + Y or M enables the NTFS3 driver with full features enabled (read, + write, journal replaying, sparse/compressed files support). + File system type to use on mount is "ntfs3". Module name (M option) + is also "ntfs3". + + Documentation: <file:Documentation/filesystems/ntfs3.rst> + +config NTFS3_64BIT_CLUSTER + bool "64 bits per NTFS clusters" + depends on NTFS3_FS && 64BIT + help + Windows implementation of ntfs.sys uses 32 bits per clusters. + If activated 64 bits per clusters you will be able to use 4k cluster + for 16T+ volumes. Windows will not be able to mount such volumes. + + It is recommended to say N here. + +config NTFS3_LZX_XPRESS + bool "activate support of external compressions lzx/xpress" + depends on NTFS3_FS + help + In Windows 10 one can use command "compact" to compress any files. + 4 possible variants of compression are: xpress4k, xpress8k, xpress16k and lzx. + If activated you will be able to read such files correctly. + + It is recommended to say Y here. + +config NTFS3_FS_POSIX_ACL + bool "NTFS POSIX Access Control Lists" + depends on NTFS3_FS + select FS_POSIX_ACL + help + POSIX Access Control Lists (ACLs) support additional access rights + for users and groups beyond the standard owner/group/world scheme, + and this option selects support for ACLs specifically for ntfs + filesystems. + NOTE: this is linux only feature. Windows will ignore these ACLs. + + If you don't know what Access Control Lists are, say N. diff --git a/fs/ntfs3/Makefile b/fs/ntfs3/Makefile new file mode 100644 index 000000000000..279701b62bbe --- /dev/null +++ b/fs/ntfs3/Makefile @@ -0,0 +1,36 @@ +# SPDX-License-Identifier: GPL-2.0 +# +# Makefile for the ntfs3 filesystem support. +# + +# to check robot warnings +ccflags-y += -Wint-to-pointer-cast \ + $(call cc-option,-Wunused-but-set-variable,-Wunused-const-variable) \ + $(call cc-option,-Wold-style-declaration,-Wout-of-line-declaration) + +obj-$(CONFIG_NTFS3_FS) += ntfs3.o + +ntfs3-y := attrib.o \ + attrlist.o \ + bitfunc.o \ + bitmap.o \ + dir.o \ + fsntfs.o \ + frecord.o \ + file.o \ + fslog.o \ + inode.o \ + index.o \ + lznt.o \ + namei.o \ + record.o \ + run.o \ + super.o \ + upcase.o \ + xattr.o + +ntfs3-$(CONFIG_NTFS3_LZX_XPRESS) += $(addprefix lib/,\ + decompress_common.o \ + lzx_decompress.o \ + xpress_decompress.o \ + ) \ No newline at end of file -- 2.30.0

1 0

[PATCH OLK-5.10 007/107] fs/ntfs3: Add NTFS journal
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> mainline inclusion from mainline-v5.15 commit b46acd6a6a627d876898e1c84d3f84902264b445 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- This adds NTFS journal Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/fslog.c | 5182 ++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 5182 insertions(+) create mode 100644 fs/ntfs3/fslog.c diff --git a/fs/ntfs3/fslog.c b/fs/ntfs3/fslog.c new file mode 100644 index 000000000000..397ba6a956e7 --- /dev/null +++ b/fs/ntfs3/fslog.c @@ -0,0 +1,5182 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * + * Copyright (C) 2019-2021 Paragon Software GmbH, All rights reserved. + * + */ + +#include <linux/blkdev.h> +#include <linux/buffer_head.h> +#include <linux/fs.h> +#include <linux/hash.h> +#include <linux/nls.h> +#include <linux/random.h> +#include <linux/ratelimit.h> +#include <linux/slab.h> + +#include "debug.h" +#include "ntfs.h" +#include "ntfs_fs.h" + +/* + * LOG FILE structs + */ + +// clang-format off + +#define MaxLogFileSize 0x100000000ull +#define DefaultLogPageSize 4096 +#define MinLogRecordPages 0x30 + +struct RESTART_HDR { + struct NTFS_RECORD_HEADER rhdr; // 'RSTR' + __le32 sys_page_size; // 0x10: Page size of the system which initialized the log + __le32 page_size; // 0x14: Log page size used for this log file + __le16 ra_off; // 0x18: + __le16 minor_ver; // 0x1A: + __le16 major_ver; // 0x1C: + __le16 fixups[]; +}; + +#define LFS_NO_CLIENT 0xffff +#define LFS_NO_CLIENT_LE cpu_to_le16(0xffff) + +struct CLIENT_REC { + __le64 oldest_lsn; + __le64 restart_lsn; // 0x08: + __le16 prev_client; // 0x10: + __le16 next_client; // 0x12: + __le16 seq_num; // 0x14: + u8 align[6]; // 0x16 + __le32 name_bytes; // 0x1C: in bytes + __le16 name[32]; // 0x20: name of client +}; + +static_assert(sizeof(struct CLIENT_REC) == 0x60); + +/* Two copies of these will exist at the beginning of the log file */ +struct RESTART_AREA { + __le64 current_lsn; // 0x00: Current logical end of log file + __le16 log_clients; // 0x08: Maximum number of clients + __le16 client_idx[2]; // 0x0A: free/use index into the client record arrays + __le16 flags; // 0x0E: See RESTART_SINGLE_PAGE_IO + __le32 seq_num_bits; // 0x10: the number of bits in sequence number. + __le16 ra_len; // 0x14: + __le16 client_off; // 0x16: + __le64 l_size; // 0x18: Usable log file size. + __le32 last_lsn_data_len; // 0x20: + __le16 rec_hdr_len; // 0x24: log page data offset + __le16 data_off; // 0x26: log page data length + __le32 open_log_count; // 0x28: + __le32 align[5]; // 0x2C: + struct CLIENT_REC clients[]; // 0x40: +}; + +struct LOG_REC_HDR { + __le16 redo_op; // 0x00: NTFS_LOG_OPERATION + __le16 undo_op; // 0x02: NTFS_LOG_OPERATION + __le16 redo_off; // 0x04: Offset to Redo record + __le16 redo_len; // 0x06: Redo length + __le16 undo_off; // 0x08: Offset to Undo record + __le16 undo_len; // 0x0A: Undo length + __le16 target_attr; // 0x0C: + __le16 lcns_follow; // 0x0E: + __le16 record_off; // 0x10: + __le16 attr_off; // 0x12: + __le16 cluster_off; // 0x14: + __le16 reserved; // 0x16: + __le64 target_vcn; // 0x18: + __le64 page_lcns[]; // 0x20: +}; + +static_assert(sizeof(struct LOG_REC_HDR) == 0x20); + +#define RESTART_ENTRY_ALLOCATED 0xFFFFFFFF +#define RESTART_ENTRY_ALLOCATED_LE cpu_to_le32(0xFFFFFFFF) + +struct RESTART_TABLE { + __le16 size; // 0x00: In bytes + __le16 used; // 0x02: entries + __le16 total; // 0x04: entries + __le16 res[3]; // 0x06: + __le32 free_goal; // 0x0C: + __le32 first_free; // 0x10 + __le32 last_free; // 0x14 + +}; + +static_assert(sizeof(struct RESTART_TABLE) == 0x18); + +struct ATTR_NAME_ENTRY { + __le16 off; // offset in the Open attribute Table + __le16 name_bytes; + __le16 name[]; +}; + +struct OPEN_ATTR_ENRTY { + __le32 next; // 0x00: RESTART_ENTRY_ALLOCATED if allocated + __le32 bytes_per_index; // 0x04: + enum ATTR_TYPE type; // 0x08: + u8 is_dirty_pages; // 0x0C: + u8 is_attr_name; // 0x0B: Faked field to manage 'ptr' + u8 name_len; // 0x0C: Faked field to manage 'ptr' + u8 res; + struct MFT_REF ref; // 0x10: File Reference of file containing attribute + __le64 open_record_lsn; // 0x18: + void *ptr; // 0x20: +}; + +/* 32 bit version of 'struct OPEN_ATTR_ENRTY' */ +struct OPEN_ATTR_ENRTY_32 { + __le32 next; // 0x00: RESTART_ENTRY_ALLOCATED if allocated + __le32 ptr; // 0x04: + struct MFT_REF ref; // 0x08: + __le64 open_record_lsn; // 0x10: + u8 is_dirty_pages; // 0x18: + u8 is_attr_name; // 0x19 + u8 res1[2]; + enum ATTR_TYPE type; // 0x1C: + u8 name_len; // 0x20: in wchar + u8 res2[3]; + __le32 AttributeName; // 0x24: + __le32 bytes_per_index; // 0x28: +}; + +#define SIZEOF_OPENATTRIBUTEENTRY0 0x2c +// static_assert( 0x2C == sizeof(struct OPEN_ATTR_ENRTY_32) ); +static_assert(sizeof(struct OPEN_ATTR_ENRTY) < SIZEOF_OPENATTRIBUTEENTRY0); + +/* + * One entry exists in the Dirty Pages Table for each page which is dirty at the + * time the Restart Area is written + */ +struct DIR_PAGE_ENTRY { + __le32 next; // 0x00: RESTART_ENTRY_ALLOCATED if allocated + __le32 target_attr; // 0x04: Index into the Open attribute Table + __le32 transfer_len; // 0x08: + __le32 lcns_follow; // 0x0C: + __le64 vcn; // 0x10: Vcn of dirty page + __le64 oldest_lsn; // 0x18: + __le64 page_lcns[]; // 0x20: +}; + +static_assert(sizeof(struct DIR_PAGE_ENTRY) == 0x20); + +/* 32 bit version of 'struct DIR_PAGE_ENTRY' */ +struct DIR_PAGE_ENTRY_32 { + __le32 next; // 0x00: RESTART_ENTRY_ALLOCATED if allocated + __le32 target_attr; // 0x04: Index into the Open attribute Table + __le32 transfer_len; // 0x08: + __le32 lcns_follow; // 0x0C: + __le32 reserved; // 0x10: + __le32 vcn_low; // 0x14: Vcn of dirty page + __le32 vcn_hi; // 0x18: Vcn of dirty page + __le32 oldest_lsn_low; // 0x1C: + __le32 oldest_lsn_hi; // 0x1C: + __le32 page_lcns_low; // 0x24: + __le32 page_lcns_hi; // 0x24: +}; + +static_assert(offsetof(struct DIR_PAGE_ENTRY_32, vcn_low) == 0x14); +static_assert(sizeof(struct DIR_PAGE_ENTRY_32) == 0x2c); + +enum transact_state { + TransactionUninitialized = 0, + TransactionActive, + TransactionPrepared, + TransactionCommitted +}; + +struct TRANSACTION_ENTRY { + __le32 next; // 0x00: RESTART_ENTRY_ALLOCATED if allocated + u8 transact_state; // 0x04: + u8 reserved[3]; // 0x05: + __le64 first_lsn; // 0x08: + __le64 prev_lsn; // 0x10: + __le64 undo_next_lsn; // 0x18: + __le32 undo_records; // 0x20: Number of undo log records pending abort + __le32 undo_len; // 0x24: Total undo size +}; + +static_assert(sizeof(struct TRANSACTION_ENTRY) == 0x28); + +struct NTFS_RESTART { + __le32 major_ver; // 0x00: + __le32 minor_ver; // 0x04: + __le64 check_point_start; // 0x08: + __le64 open_attr_table_lsn; // 0x10: + __le64 attr_names_lsn; // 0x18: + __le64 dirty_pages_table_lsn; // 0x20: + __le64 transact_table_lsn; // 0x28: + __le32 open_attr_len; // 0x30: In bytes + __le32 attr_names_len; // 0x34: In bytes + __le32 dirty_pages_len; // 0x38: In bytes + __le32 transact_table_len; // 0x3C: In bytes +}; + +static_assert(sizeof(struct NTFS_RESTART) == 0x40); + +struct NEW_ATTRIBUTE_SIZES { + __le64 alloc_size; + __le64 valid_size; + __le64 data_size; + __le64 total_size; +}; + +struct BITMAP_RANGE { + __le32 bitmap_off; + __le32 bits; +}; + +struct LCN_RANGE { + __le64 lcn; + __le64 len; +}; + +/* The following type defines the different log record types */ +#define LfsClientRecord cpu_to_le32(1) +#define LfsClientRestart cpu_to_le32(2) + +/* This is used to uniquely identify a client for a particular log file */ +struct CLIENT_ID { + __le16 seq_num; + __le16 client_idx; +}; + +/* This is the header that begins every Log Record in the log file */ +struct LFS_RECORD_HDR { + __le64 this_lsn; // 0x00: + __le64 client_prev_lsn; // 0x08: + __le64 client_undo_next_lsn; // 0x10: + __le32 client_data_len; // 0x18: + struct CLIENT_ID client; // 0x1C: Owner of this log record + __le32 record_type; // 0x20: LfsClientRecord or LfsClientRestart + __le32 transact_id; // 0x24: + __le16 flags; // 0x28: LOG_RECORD_MULTI_PAGE + u8 align[6]; // 0x2A: +}; + +#define LOG_RECORD_MULTI_PAGE cpu_to_le16(1) + +static_assert(sizeof(struct LFS_RECORD_HDR) == 0x30); + +struct LFS_RECORD { + __le16 next_record_off; // 0x00: Offset of the free space in the page + u8 align[6]; // 0x02: + __le64 last_end_lsn; // 0x08: lsn for the last log record which ends on the page +}; + +static_assert(sizeof(struct LFS_RECORD) == 0x10); + +struct RECORD_PAGE_HDR { + struct NTFS_RECORD_HEADER rhdr; // 'RCRD' + __le32 rflags; // 0x10: See LOG_PAGE_LOG_RECORD_END + __le16 page_count; // 0x14: + __le16 page_pos; // 0x16: + struct LFS_RECORD record_hdr; // 0x18 + __le16 fixups[10]; // 0x28 + __le32 file_off; // 0x3c: used when major version >= 2 +}; + +// clang-format on + +// Page contains the end of a log record +#define LOG_PAGE_LOG_RECORD_END cpu_to_le32(0x00000001) + +static inline bool is_log_record_end(const struct RECORD_PAGE_HDR *hdr) +{ + return hdr->rflags & LOG_PAGE_LOG_RECORD_END; +} + +static_assert(offsetof(struct RECORD_PAGE_HDR, file_off) == 0x3c); + +/* + * END of NTFS LOG structures + */ + +/* Define some tuning parameters to keep the restart tables a reasonable size */ +#define INITIAL_NUMBER_TRANSACTIONS 5 + +enum NTFS_LOG_OPERATION { + + Noop = 0x00, + CompensationLogRecord = 0x01, + InitializeFileRecordSegment = 0x02, + DeallocateFileRecordSegment = 0x03, + WriteEndOfFileRecordSegment = 0x04, + CreateAttribute = 0x05, + DeleteAttribute = 0x06, + UpdateResidentValue = 0x07, + UpdateNonresidentValue = 0x08, + UpdateMappingPairs = 0x09, + DeleteDirtyClusters = 0x0A, + SetNewAttributeSizes = 0x0B, + AddIndexEntryRoot = 0x0C, + DeleteIndexEntryRoot = 0x0D, + AddIndexEntryAllocation = 0x0E, + DeleteIndexEntryAllocation = 0x0F, + WriteEndOfIndexBuffer = 0x10, + SetIndexEntryVcnRoot = 0x11, + SetIndexEntryVcnAllocation = 0x12, + UpdateFileNameRoot = 0x13, + UpdateFileNameAllocation = 0x14, + SetBitsInNonresidentBitMap = 0x15, + ClearBitsInNonresidentBitMap = 0x16, + HotFix = 0x17, + EndTopLevelAction = 0x18, + PrepareTransaction = 0x19, + CommitTransaction = 0x1A, + ForgetTransaction = 0x1B, + OpenNonresidentAttribute = 0x1C, + OpenAttributeTableDump = 0x1D, + AttributeNamesDump = 0x1E, + DirtyPageTableDump = 0x1F, + TransactionTableDump = 0x20, + UpdateRecordDataRoot = 0x21, + UpdateRecordDataAllocation = 0x22, + + UpdateRelativeDataInIndex = + 0x23, // NtOfsRestartUpdateRelativeDataInIndex + UpdateRelativeDataInIndex2 = 0x24, + ZeroEndOfFileRecord = 0x25, +}; + +/* + * Array for log records which require a target attribute + * A true indicates that the corresponding restart operation requires a target attribute + */ +static const u8 AttributeRequired[] = { + 0xFC, 0xFB, 0xFF, 0x10, 0x06, +}; + +static inline bool is_target_required(u16 op) +{ + bool ret = op <= UpdateRecordDataAllocation && + (AttributeRequired[op >> 3] >> (op & 7) & 1); + return ret; +} + +static inline bool can_skip_action(enum NTFS_LOG_OPERATION op) +{ + switch (op) { + case Noop: + case DeleteDirtyClusters: + case HotFix: + case EndTopLevelAction: + case PrepareTransaction: + case CommitTransaction: + case ForgetTransaction: + case CompensationLogRecord: + case OpenNonresidentAttribute: + case OpenAttributeTableDump: + case AttributeNamesDump: + case DirtyPageTableDump: + case TransactionTableDump: + return true; + default: + return false; + } +} + +enum { lcb_ctx_undo_next, lcb_ctx_prev, lcb_ctx_next }; + +/* bytes per restart table */ +static inline u32 bytes_per_rt(const struct RESTART_TABLE *rt) +{ + return le16_to_cpu(rt->used) * le16_to_cpu(rt->size) + + sizeof(struct RESTART_TABLE); +} + +/* log record length */ +static inline u32 lrh_length(const struct LOG_REC_HDR *lr) +{ + u16 t16 = le16_to_cpu(lr->lcns_follow); + + return struct_size(lr, page_lcns, max_t(u16, 1, t16)); +} + +struct lcb { + struct LFS_RECORD_HDR *lrh; // Log record header of the current lsn + struct LOG_REC_HDR *log_rec; + u32 ctx_mode; // lcb_ctx_undo_next/lcb_ctx_prev/lcb_ctx_next + struct CLIENT_ID client; + bool alloc; // if true the we should deallocate 'log_rec' +}; + +static void lcb_put(struct lcb *lcb) +{ + if (lcb->alloc) + ntfs_free(lcb->log_rec); + ntfs_free(lcb->lrh); + ntfs_free(lcb); +} + +/* + * oldest_client_lsn + * + * find the oldest lsn from active clients. + */ +static inline void oldest_client_lsn(const struct CLIENT_REC *ca, + __le16 next_client, u64 *oldest_lsn) +{ + while (next_client != LFS_NO_CLIENT_LE) { + const struct CLIENT_REC *cr = ca + le16_to_cpu(next_client); + u64 lsn = le64_to_cpu(cr->oldest_lsn); + + /* ignore this block if it's oldest lsn is 0 */ + if (lsn && lsn < *oldest_lsn) + *oldest_lsn = lsn; + + next_client = cr->next_client; + } +} + +static inline bool is_rst_page_hdr_valid(u32 file_off, + const struct RESTART_HDR *rhdr) +{ + u32 sys_page = le32_to_cpu(rhdr->sys_page_size); + u32 page_size = le32_to_cpu(rhdr->page_size); + u32 end_usa; + u16 ro; + + if (sys_page < SECTOR_SIZE || page_size < SECTOR_SIZE || + sys_page & (sys_page - 1) || page_size & (page_size - 1)) { + return false; + } + + /* Check that if the file offset isn't 0, it is the system page size */ + if (file_off && file_off != sys_page) + return false; + + /* Check support version 1.1+ */ + if (le16_to_cpu(rhdr->major_ver) <= 1 && !rhdr->minor_ver) + return false; + + if (le16_to_cpu(rhdr->major_ver) > 2) + return false; + + ro = le16_to_cpu(rhdr->ra_off); + if (!IsQuadAligned(ro) || ro > sys_page) + return false; + + end_usa = ((sys_page >> SECTOR_SHIFT) + 1) * sizeof(short); + end_usa += le16_to_cpu(rhdr->rhdr.fix_off); + + if (ro < end_usa) + return false; + + return true; +} + +static inline bool is_rst_area_valid(const struct RESTART_HDR *rhdr) +{ + const struct RESTART_AREA *ra; + u16 cl, fl, ul; + u32 off, l_size, file_dat_bits, file_size_round; + u16 ro = le16_to_cpu(rhdr->ra_off); + u32 sys_page = le32_to_cpu(rhdr->sys_page_size); + + if (ro + offsetof(struct RESTART_AREA, l_size) > + SECTOR_SIZE - sizeof(short)) + return false; + + ra = Add2Ptr(rhdr, ro); + cl = le16_to_cpu(ra->log_clients); + + if (cl > 1) + return false; + + off = le16_to_cpu(ra->client_off); + + if (!IsQuadAligned(off) || ro + off > SECTOR_SIZE - sizeof(short)) + return false; + + off += cl * sizeof(struct CLIENT_REC); + + if (off > sys_page) + return false; + + /* + * Check the restart length field and whether the entire + * restart area is contained that length + */ + if (le16_to_cpu(rhdr->ra_off) + le16_to_cpu(ra->ra_len) > sys_page || + off > le16_to_cpu(ra->ra_len)) { + return false; + } + + /* + * As a final check make sure that the use list and the free list + * are either empty or point to a valid client + */ + fl = le16_to_cpu(ra->client_idx[0]); + ul = le16_to_cpu(ra->client_idx[1]); + if ((fl != LFS_NO_CLIENT && fl >= cl) || + (ul != LFS_NO_CLIENT && ul >= cl)) + return false; + + /* Make sure the sequence number bits match the log file size */ + l_size = le64_to_cpu(ra->l_size); + + file_dat_bits = sizeof(u64) * 8 - le32_to_cpu(ra->seq_num_bits); + file_size_round = 1u << (file_dat_bits + 3); + if (file_size_round != l_size && + (file_size_round < l_size || (file_size_round / 2) > l_size)) { + return false; + } + + /* The log page data offset and record header length must be quad-aligned */ + if (!IsQuadAligned(le16_to_cpu(ra->data_off)) || + !IsQuadAligned(le16_to_cpu(ra->rec_hdr_len))) + return false; + + return true; +} + +static inline bool is_client_area_valid(const struct RESTART_HDR *rhdr, + bool usa_error) +{ + u16 ro = le16_to_cpu(rhdr->ra_off); + const struct RESTART_AREA *ra = Add2Ptr(rhdr, ro); + u16 ra_len = le16_to_cpu(ra->ra_len); + const struct CLIENT_REC *ca; + u32 i; + + if (usa_error && ra_len + ro > SECTOR_SIZE - sizeof(short)) + return false; + + /* Find the start of the client array */ + ca = Add2Ptr(ra, le16_to_cpu(ra->client_off)); + + /* + * Start with the free list + * Check that all the clients are valid and that there isn't a cycle + * Do the in-use list on the second pass + */ + for (i = 0; i < 2; i++) { + u16 client_idx = le16_to_cpu(ra->client_idx[i]); + bool first_client = true; + u16 clients = le16_to_cpu(ra->log_clients); + + while (client_idx != LFS_NO_CLIENT) { + const struct CLIENT_REC *cr; + + if (!clients || + client_idx >= le16_to_cpu(ra->log_clients)) + return false; + + clients -= 1; + cr = ca + client_idx; + + client_idx = le16_to_cpu(cr->next_client); + + if (first_client) { + first_client = false; + if (cr->prev_client != LFS_NO_CLIENT_LE) + return false; + } + } + } + + return true; +} + +/* + * remove_client + * + * remove a client record from a client record list an restart area + */ +static inline void remove_client(struct CLIENT_REC *ca, + const struct CLIENT_REC *cr, __le16 *head) +{ + if (cr->prev_client == LFS_NO_CLIENT_LE) + *head = cr->next_client; + else + ca[le16_to_cpu(cr->prev_client)].next_client = cr->next_client; + + if (cr->next_client != LFS_NO_CLIENT_LE) + ca[le16_to_cpu(cr->next_client)].prev_client = cr->prev_client; +} + +/* + * add_client + * + * add a client record to the start of a list + */ +static inline void add_client(struct CLIENT_REC *ca, u16 index, __le16 *head) +{ + struct CLIENT_REC *cr = ca + index; + + cr->prev_client = LFS_NO_CLIENT_LE; + cr->next_client = *head; + + if (*head != LFS_NO_CLIENT_LE) + ca[le16_to_cpu(*head)].prev_client = cpu_to_le16(index); + + *head = cpu_to_le16(index); +} + +/* + * enum_rstbl + * + */ +static inline void *enum_rstbl(struct RESTART_TABLE *t, void *c) +{ + __le32 *e; + u32 bprt; + u16 rsize = t ? le16_to_cpu(t->size) : 0; + + if (!c) { + if (!t || !t->total) + return NULL; + e = Add2Ptr(t, sizeof(struct RESTART_TABLE)); + } else { + e = Add2Ptr(c, rsize); + } + + /* Loop until we hit the first one allocated, or the end of the list */ + for (bprt = bytes_per_rt(t); PtrOffset(t, e) < bprt; + e = Add2Ptr(e, rsize)) { + if (*e == RESTART_ENTRY_ALLOCATED_LE) + return e; + } + return NULL; +} + +/* + * find_dp + * + * searches for a 'vcn' in Dirty Page Table, + */ +static inline struct DIR_PAGE_ENTRY *find_dp(struct RESTART_TABLE *dptbl, + u32 target_attr, u64 vcn) +{ + __le32 ta = cpu_to_le32(target_attr); + struct DIR_PAGE_ENTRY *dp = NULL; + + while ((dp = enum_rstbl(dptbl, dp))) { + u64 dp_vcn = le64_to_cpu(dp->vcn); + + if (dp->target_attr == ta && vcn >= dp_vcn && + vcn < dp_vcn + le32_to_cpu(dp->lcns_follow)) { + return dp; + } + } + return NULL; +} + +static inline u32 norm_file_page(u32 page_size, u32 *l_size, bool use_default) +{ + if (use_default) + page_size = DefaultLogPageSize; + + /* Round the file size down to a system page boundary */ + *l_size &= ~(page_size - 1); + + /* File should contain at least 2 restart pages and MinLogRecordPages pages */ + if (*l_size < (MinLogRecordPages + 2) * page_size) + return 0; + + return page_size; +} + +static bool check_log_rec(const struct LOG_REC_HDR *lr, u32 bytes, u32 tr, + u32 bytes_per_attr_entry) +{ + u16 t16; + + if (bytes < sizeof(struct LOG_REC_HDR)) + return false; + if (!tr) + return false; + + if ((tr - sizeof(struct RESTART_TABLE)) % + sizeof(struct TRANSACTION_ENTRY)) + return false; + + if (le16_to_cpu(lr->redo_off) & 7) + return false; + + if (le16_to_cpu(lr->undo_off) & 7) + return false; + + if (lr->target_attr) + goto check_lcns; + + if (is_target_required(le16_to_cpu(lr->redo_op))) + return false; + + if (is_target_required(le16_to_cpu(lr->undo_op))) + return false; + +check_lcns: + if (!lr->lcns_follow) + goto check_length; + + t16 = le16_to_cpu(lr->target_attr); + if ((t16 - sizeof(struct RESTART_TABLE)) % bytes_per_attr_entry) + return false; + +check_length: + if (bytes < lrh_length(lr)) + return false; + + return true; +} + +static bool check_rstbl(const struct RESTART_TABLE *rt, size_t bytes) +{ + u32 ts; + u32 i, off; + u16 rsize = le16_to_cpu(rt->size); + u16 ne = le16_to_cpu(rt->used); + u32 ff = le32_to_cpu(rt->first_free); + u32 lf = le32_to_cpu(rt->last_free); + + ts = rsize * ne + sizeof(struct RESTART_TABLE); + + if (!rsize || rsize > bytes || + rsize + sizeof(struct RESTART_TABLE) > bytes || bytes < ts || + le16_to_cpu(rt->total) > ne || ff > ts || lf > ts || + (ff && ff < sizeof(struct RESTART_TABLE)) || + (lf && lf < sizeof(struct RESTART_TABLE))) { + return false; + } + + /* Verify each entry is either allocated or points + * to a valid offset the table + */ + for (i = 0; i < ne; i++) { + off = le32_to_cpu(*(__le32 *)Add2Ptr( + rt, i * rsize + sizeof(struct RESTART_TABLE))); + + if (off != RESTART_ENTRY_ALLOCATED && off && + (off < sizeof(struct RESTART_TABLE) || + ((off - sizeof(struct RESTART_TABLE)) % rsize))) { + return false; + } + } + + /* Walk through the list headed by the first entry to make + * sure none of the entries are currently being used + */ + for (off = ff; off;) { + if (off == RESTART_ENTRY_ALLOCATED) + return false; + + off = le32_to_cpu(*(__le32 *)Add2Ptr(rt, off)); + } + + return true; +} + +/* + * free_rsttbl_idx + * + * frees a previously allocated index a Restart Table. + */ +static inline void free_rsttbl_idx(struct RESTART_TABLE *rt, u32 off) +{ + __le32 *e; + u32 lf = le32_to_cpu(rt->last_free); + __le32 off_le = cpu_to_le32(off); + + e = Add2Ptr(rt, off); + + if (off < le32_to_cpu(rt->free_goal)) { + *e = rt->first_free; + rt->first_free = off_le; + if (!lf) + rt->last_free = off_le; + } else { + if (lf) + *(__le32 *)Add2Ptr(rt, lf) = off_le; + else + rt->first_free = off_le; + + rt->last_free = off_le; + *e = 0; + } + + le16_sub_cpu(&rt->total, 1); +} + +static inline struct RESTART_TABLE *init_rsttbl(u16 esize, u16 used) +{ + __le32 *e, *last_free; + u32 off; + u32 bytes = esize * used + sizeof(struct RESTART_TABLE); + u32 lf = sizeof(struct RESTART_TABLE) + (used - 1) * esize; + struct RESTART_TABLE *t = ntfs_zalloc(bytes); + + t->size = cpu_to_le16(esize); + t->used = cpu_to_le16(used); + t->free_goal = cpu_to_le32(~0u); + t->first_free = cpu_to_le32(sizeof(struct RESTART_TABLE)); + t->last_free = cpu_to_le32(lf); + + e = (__le32 *)(t + 1); + last_free = Add2Ptr(t, lf); + + for (off = sizeof(struct RESTART_TABLE) + esize; e < last_free; + e = Add2Ptr(e, esize), off += esize) { + *e = cpu_to_le32(off); + } + return t; +} + +static inline struct RESTART_TABLE *extend_rsttbl(struct RESTART_TABLE *tbl, + u32 add, u32 free_goal) +{ + u16 esize = le16_to_cpu(tbl->size); + __le32 osize = cpu_to_le32(bytes_per_rt(tbl)); + u32 used = le16_to_cpu(tbl->used); + struct RESTART_TABLE *rt = init_rsttbl(esize, used + add); + + memcpy(rt + 1, tbl + 1, esize * used); + + rt->free_goal = free_goal == ~0u + ? cpu_to_le32(~0u) + : cpu_to_le32(sizeof(struct RESTART_TABLE) + + free_goal * esize); + + if (tbl->first_free) { + rt->first_free = tbl->first_free; + *(__le32 *)Add2Ptr(rt, le32_to_cpu(tbl->last_free)) = osize; + } else { + rt->first_free = osize; + } + + rt->total = tbl->total; + + ntfs_free(tbl); + return rt; +} + +/* + * alloc_rsttbl_idx + * + * allocates an index from within a previously initialized Restart Table + */ +static inline void *alloc_rsttbl_idx(struct RESTART_TABLE **tbl) +{ + u32 off; + __le32 *e; + struct RESTART_TABLE *t = *tbl; + + if (!t->first_free) + *tbl = t = extend_rsttbl(t, 16, ~0u); + + off = le32_to_cpu(t->first_free); + + /* Dequeue this entry and zero it. */ + e = Add2Ptr(t, off); + + t->first_free = *e; + + memset(e, 0, le16_to_cpu(t->size)); + + *e = RESTART_ENTRY_ALLOCATED_LE; + + /* If list is going empty, then we fix the last_free as well. */ + if (!t->first_free) + t->last_free = 0; + + le16_add_cpu(&t->total, 1); + + return Add2Ptr(t, off); +} + +/* + * alloc_rsttbl_from_idx + * + * allocates a specific index from within a previously initialized Restart Table + */ +static inline void *alloc_rsttbl_from_idx(struct RESTART_TABLE **tbl, u32 vbo) +{ + u32 off; + __le32 *e; + struct RESTART_TABLE *rt = *tbl; + u32 bytes = bytes_per_rt(rt); + u16 esize = le16_to_cpu(rt->size); + + /* If the entry is not the table, we will have to extend the table */ + if (vbo >= bytes) { + /* + * extend the size by computing the number of entries between + * the existing size and the desired index and adding + * 1 to that + */ + u32 bytes2idx = vbo - bytes; + + /* There should always be an integral number of entries being added */ + /* Now extend the table */ + *tbl = rt = extend_rsttbl(rt, bytes2idx / esize + 1, bytes); + if (!rt) + return NULL; + } + + /* see if the entry is already allocated, and just return if it is. */ + e = Add2Ptr(rt, vbo); + + if (*e == RESTART_ENTRY_ALLOCATED_LE) + return e; + + /* + * Walk through the table, looking for the entry we're + * interested and the previous entry + */ + off = le32_to_cpu(rt->first_free); + e = Add2Ptr(rt, off); + + if (off == vbo) { + /* this is a match */ + rt->first_free = *e; + goto skip_looking; + } + + /* + * need to walk through the list looking for the predecessor of our entry + */ + for (;;) { + /* Remember the entry just found */ + u32 last_off = off; + __le32 *last_e = e; + + /* should never run of entries. */ + + /* Lookup up the next entry the list */ + off = le32_to_cpu(*last_e); + e = Add2Ptr(rt, off); + + /* If this is our match we are done */ + if (off == vbo) { + *last_e = *e; + + /* If this was the last entry, we update that the table as well */ + if (le32_to_cpu(rt->last_free) == off) + rt->last_free = cpu_to_le32(last_off); + break; + } + } + +skip_looking: + /* If the list is now empty, we fix the last_free as well */ + if (!rt->first_free) + rt->last_free = 0; + + /* Zero this entry */ + memset(e, 0, esize); + *e = RESTART_ENTRY_ALLOCATED_LE; + + le16_add_cpu(&rt->total, 1); + + return e; +} + +#define RESTART_SINGLE_PAGE_IO cpu_to_le16(0x0001) + +#define NTFSLOG_WRAPPED 0x00000001 +#define NTFSLOG_MULTIPLE_PAGE_IO 0x00000002 +#define NTFSLOG_NO_LAST_LSN 0x00000004 +#define NTFSLOG_REUSE_TAIL 0x00000010 +#define NTFSLOG_NO_OLDEST_LSN 0x00000020 + +/* + * Helper struct to work with NTFS LogFile + */ +struct ntfs_log { + struct ntfs_inode *ni; + + u32 l_size; + u32 sys_page_size; + u32 sys_page_mask; + u32 page_size; + u32 page_mask; // page_size - 1 + u8 page_bits; + struct RECORD_PAGE_HDR *one_page_buf; + + struct RESTART_TABLE *open_attr_tbl; + u32 transaction_id; + u32 clst_per_page; + + u32 first_page; + u32 next_page; + u32 ra_off; + u32 data_off; + u32 restart_size; + u32 data_size; + u16 record_header_len; + u64 seq_num; + u32 seq_num_bits; + u32 file_data_bits; + u32 seq_num_mask; /* (1 << file_data_bits) - 1 */ + + struct RESTART_AREA *ra; /* in-memory image of the next restart area */ + u32 ra_size; /* the usable size of the restart area */ + + /* + * If true, then the in-memory restart area is to be written + * to the first position on the disk + */ + bool init_ra; + bool set_dirty; /* true if we need to set dirty flag */ + + u64 oldest_lsn; + + u32 oldest_lsn_off; + u64 last_lsn; + + u32 total_avail; + u32 total_avail_pages; + u32 total_undo_commit; + u32 max_current_avail; + u32 current_avail; + u32 reserved; + + short major_ver; + short minor_ver; + + u32 l_flags; /* See NTFSLOG_XXX */ + u32 current_openlog_count; /* On-disk value for open_log_count */ + + struct CLIENT_ID client_id; + u32 client_undo_commit; +}; + +static inline u32 lsn_to_vbo(struct ntfs_log *log, const u64 lsn) +{ + u32 vbo = (lsn << log->seq_num_bits) >> (log->seq_num_bits - 3); + + return vbo; +} + +/* compute the offset in the log file of the next log page */ +static inline u32 next_page_off(struct ntfs_log *log, u32 off) +{ + off = (off & ~log->sys_page_mask) + log->page_size; + return off >= log->l_size ? log->first_page : off; +} + +static inline u32 lsn_to_page_off(struct ntfs_log *log, u64 lsn) +{ + return (((u32)lsn) << 3) & log->page_mask; +} + +static inline u64 vbo_to_lsn(struct ntfs_log *log, u32 off, u64 Seq) +{ + return (off >> 3) + (Seq << log->file_data_bits); +} + +static inline bool is_lsn_in_file(struct ntfs_log *log, u64 lsn) +{ + return lsn >= log->oldest_lsn && + lsn <= le64_to_cpu(log->ra->current_lsn); +} + +static inline u32 hdr_file_off(struct ntfs_log *log, + struct RECORD_PAGE_HDR *hdr) +{ + if (log->major_ver < 2) + return le64_to_cpu(hdr->rhdr.lsn); + + return le32_to_cpu(hdr->file_off); +} + +static inline u64 base_lsn(struct ntfs_log *log, + const struct RECORD_PAGE_HDR *hdr, u64 lsn) +{ + u64 h_lsn = le64_to_cpu(hdr->rhdr.lsn); + u64 ret = (((h_lsn >> log->file_data_bits) + + (lsn < (lsn_to_vbo(log, h_lsn) & ~log->page_mask) ? 1 : 0)) + << log->file_data_bits) + + ((((is_log_record_end(hdr) && + h_lsn <= le64_to_cpu(hdr->record_hdr.last_end_lsn)) + ? le16_to_cpu(hdr->record_hdr.next_record_off) + : log->page_size) + + lsn) >> + 3); + + return ret; +} + +static inline bool verify_client_lsn(struct ntfs_log *log, + const struct CLIENT_REC *client, u64 lsn) +{ + return lsn >= le64_to_cpu(client->oldest_lsn) && + lsn <= le64_to_cpu(log->ra->current_lsn) && lsn; +} + +struct restart_info { + u64 last_lsn; + struct RESTART_HDR *r_page; + u32 vbo; + bool chkdsk_was_run; + bool valid_page; + bool initialized; + bool restart; +}; + +static int read_log_page(struct ntfs_log *log, u32 vbo, + struct RECORD_PAGE_HDR **buffer, bool *usa_error) +{ + int err = 0; + u32 page_idx = vbo >> log->page_bits; + u32 page_off = vbo & log->page_mask; + u32 bytes = log->page_size - page_off; + void *to_free = NULL; + u32 page_vbo = page_idx << log->page_bits; + struct RECORD_PAGE_HDR *page_buf; + struct ntfs_inode *ni = log->ni; + bool bBAAD; + + if (vbo >= log->l_size) + return -EINVAL; + + if (!*buffer) { + to_free = ntfs_malloc(bytes); + if (!to_free) + return -ENOMEM; + *buffer = to_free; + } + + page_buf = page_off ? log->one_page_buf : *buffer; + + err = ntfs_read_run_nb(ni->mi.sbi, &ni->file.run, page_vbo, page_buf, + log->page_size, NULL); + if (err) + goto out; + + if (page_buf->rhdr.sign != NTFS_FFFF_SIGNATURE) + ntfs_fix_post_read(&page_buf->rhdr, PAGE_SIZE, false); + + if (page_buf != *buffer) + memcpy(*buffer, Add2Ptr(page_buf, page_off), bytes); + + bBAAD = page_buf->rhdr.sign == NTFS_BAAD_SIGNATURE; + + if (usa_error) + *usa_error = bBAAD; + /* Check that the update sequence array for this page is valid */ + /* If we don't allow errors, raise an error status */ + else if (bBAAD) + err = -EINVAL; + +out: + if (err && to_free) { + ntfs_free(to_free); + *buffer = NULL; + } + + return err; +} + +/* + * log_read_rst + * + * it walks through 512 blocks of the file looking for a valid restart page header + * It will stop the first time we find a valid page header + */ +static int log_read_rst(struct ntfs_log *log, u32 l_size, bool first, + struct restart_info *info) +{ + u32 skip, vbo; + struct RESTART_HDR *r_page = ntfs_malloc(DefaultLogPageSize); + + if (!r_page) + return -ENOMEM; + + memset(info, 0, sizeof(struct restart_info)); + + /* Determine which restart area we are looking for */ + if (first) { + vbo = 0; + skip = 512; + } else { + vbo = 512; + skip = 0; + } + + /* loop continuously until we succeed */ + for (; vbo < l_size; vbo = 2 * vbo + skip, skip = 0) { + bool usa_error; + u32 sys_page_size; + bool brst, bchk; + struct RESTART_AREA *ra; + + /* Read a page header at the current offset */ + if (read_log_page(log, vbo, (struct RECORD_PAGE_HDR **)&r_page, + &usa_error)) { + /* ignore any errors */ + continue; + } + + /* exit if the signature is a log record page */ + if (r_page->rhdr.sign == NTFS_RCRD_SIGNATURE) { + info->initialized = true; + break; + } + + brst = r_page->rhdr.sign == NTFS_RSTR_SIGNATURE; + bchk = r_page->rhdr.sign == NTFS_CHKD_SIGNATURE; + + if (!bchk && !brst) { + if (r_page->rhdr.sign != NTFS_FFFF_SIGNATURE) { + /* + * Remember if the signature does not + * indicate uninitialized file + */ + info->initialized = true; + } + continue; + } + + ra = NULL; + info->valid_page = false; + info->initialized = true; + info->vbo = vbo; + + /* Let's check the restart area if this is a valid page */ + if (!is_rst_page_hdr_valid(vbo, r_page)) + goto check_result; + ra = Add2Ptr(r_page, le16_to_cpu(r_page->ra_off)); + + if (!is_rst_area_valid(r_page)) + goto check_result; + + /* + * We have a valid restart page header and restart area. + * If chkdsk was run or we have no clients then we have + * no more checking to do + */ + if (bchk || ra->client_idx[1] == LFS_NO_CLIENT_LE) { + info->valid_page = true; + goto check_result; + } + + /* Read the entire restart area */ + sys_page_size = le32_to_cpu(r_page->sys_page_size); + if (DefaultLogPageSize != sys_page_size) { + ntfs_free(r_page); + r_page = ntfs_zalloc(sys_page_size); + if (!r_page) + return -ENOMEM; + + if (read_log_page(log, vbo, + (struct RECORD_PAGE_HDR **)&r_page, + &usa_error)) { + /* ignore any errors */ + ntfs_free(r_page); + r_page = NULL; + continue; + } + } + + if (is_client_area_valid(r_page, usa_error)) { + info->valid_page = true; + ra = Add2Ptr(r_page, le16_to_cpu(r_page->ra_off)); + } + +check_result: + /* If chkdsk was run then update the caller's values and return */ + if (r_page->rhdr.sign == NTFS_CHKD_SIGNATURE) { + info->chkdsk_was_run = true; + info->last_lsn = le64_to_cpu(r_page->rhdr.lsn); + info->restart = true; + info->r_page = r_page; + return 0; + } + + /* If we have a valid page then copy the values we need from it */ + if (info->valid_page) { + info->last_lsn = le64_to_cpu(ra->current_lsn); + info->restart = true; + info->r_page = r_page; + return 0; + } + } + + ntfs_free(r_page); + + return 0; +} + +/* + * log_init_pg_hdr + * + * init "log' from restart page header + */ +static void log_init_pg_hdr(struct ntfs_log *log, u32 sys_page_size, + u32 page_size, u16 major_ver, u16 minor_ver) +{ + log->sys_page_size = sys_page_size; + log->sys_page_mask = sys_page_size - 1; + log->page_size = page_size; + log->page_mask = page_size - 1; + log->page_bits = blksize_bits(page_size); + + log->clst_per_page = log->page_size >> log->ni->mi.sbi->cluster_bits; + if (!log->clst_per_page) + log->clst_per_page = 1; + + log->first_page = major_ver >= 2 + ? 0x22 * page_size + : ((sys_page_size << 1) + (page_size << 1)); + log->major_ver = major_ver; + log->minor_ver = minor_ver; +} + +/* + * log_create + * + * init "log" in cases when we don't have a restart area to use + */ +static void log_create(struct ntfs_log *log, u32 l_size, const u64 last_lsn, + u32 open_log_count, bool wrapped, bool use_multi_page) +{ + log->l_size = l_size; + /* All file offsets must be quadword aligned */ + log->file_data_bits = blksize_bits(l_size) - 3; + log->seq_num_mask = (8 << log->file_data_bits) - 1; + log->seq_num_bits = sizeof(u64) * 8 - log->file_data_bits; + log->seq_num = (last_lsn >> log->file_data_bits) + 2; + log->next_page = log->first_page; + log->oldest_lsn = log->seq_num << log->file_data_bits; + log->oldest_lsn_off = 0; + log->last_lsn = log->oldest_lsn; + + log->l_flags |= NTFSLOG_NO_LAST_LSN | NTFSLOG_NO_OLDEST_LSN; + + /* Set the correct flags for the I/O and indicate if we have wrapped */ + if (wrapped) + log->l_flags |= NTFSLOG_WRAPPED; + + if (use_multi_page) + log->l_flags |= NTFSLOG_MULTIPLE_PAGE_IO; + + /* Compute the log page values */ + log->data_off = QuadAlign( + offsetof(struct RECORD_PAGE_HDR, fixups) + + sizeof(short) * ((log->page_size >> SECTOR_SHIFT) + 1)); + log->data_size = log->page_size - log->data_off; + log->record_header_len = sizeof(struct LFS_RECORD_HDR); + + /* Remember the different page sizes for reservation */ + log->reserved = log->data_size - log->record_header_len; + + /* Compute the restart page values. */ + log->ra_off = QuadAlign( + offsetof(struct RESTART_HDR, fixups) + + sizeof(short) * ((log->sys_page_size >> SECTOR_SHIFT) + 1)); + log->restart_size = log->sys_page_size - log->ra_off; + log->ra_size = struct_size(log->ra, clients, 1); + log->current_openlog_count = open_log_count; + + /* + * The total available log file space is the number of + * log file pages times the space available on each page + */ + log->total_avail_pages = log->l_size - log->first_page; + log->total_avail = log->total_avail_pages >> log->page_bits; + + /* + * We assume that we can't use the end of the page less than + * the file record size + * Then we won't need to reserve more than the caller asks for + */ + log->max_current_avail = log->total_avail * log->reserved; + log->total_avail = log->total_avail * log->data_size; + log->current_avail = log->max_current_avail; +} + +/* + * log_create_ra + * + * This routine is called to fill a restart area from the values stored in 'log' + */ +static struct RESTART_AREA *log_create_ra(struct ntfs_log *log) +{ + struct CLIENT_REC *cr; + struct RESTART_AREA *ra = ntfs_zalloc(log->restart_size); + + if (!ra) + return NULL; + + ra->current_lsn = cpu_to_le64(log->last_lsn); + ra->log_clients = cpu_to_le16(1); + ra->client_idx[1] = LFS_NO_CLIENT_LE; + if (log->l_flags & NTFSLOG_MULTIPLE_PAGE_IO) + ra->flags = RESTART_SINGLE_PAGE_IO; + ra->seq_num_bits = cpu_to_le32(log->seq_num_bits); + ra->ra_len = cpu_to_le16(log->ra_size); + ra->client_off = cpu_to_le16(offsetof(struct RESTART_AREA, clients)); + ra->l_size = cpu_to_le64(log->l_size); + ra->rec_hdr_len = cpu_to_le16(log->record_header_len); + ra->data_off = cpu_to_le16(log->data_off); + ra->open_log_count = cpu_to_le32(log->current_openlog_count + 1); + + cr = ra->clients; + + cr->prev_client = LFS_NO_CLIENT_LE; + cr->next_client = LFS_NO_CLIENT_LE; + + return ra; +} + +static u32 final_log_off(struct ntfs_log *log, u64 lsn, u32 data_len) +{ + u32 base_vbo = lsn << 3; + u32 final_log_off = (base_vbo & log->seq_num_mask) & ~log->page_mask; + u32 page_off = base_vbo & log->page_mask; + u32 tail = log->page_size - page_off; + + page_off -= 1; + + /* Add the length of the header */ + data_len += log->record_header_len; + + /* + * If this lsn is contained this log page we are done + * Otherwise we need to walk through several log pages + */ + if (data_len > tail) { + data_len -= tail; + tail = log->data_size; + page_off = log->data_off - 1; + + for (;;) { + final_log_off = next_page_off(log, final_log_off); + + /* We are done if the remaining bytes fit on this page */ + if (data_len <= tail) + break; + data_len -= tail; + } + } + + /* + * We add the remaining bytes to our starting position on this page + * and then add that value to the file offset of this log page + */ + return final_log_off + data_len + page_off; +} + +static int next_log_lsn(struct ntfs_log *log, const struct LFS_RECORD_HDR *rh, + u64 *lsn) +{ + int err; + u64 this_lsn = le64_to_cpu(rh->this_lsn); + u32 vbo = lsn_to_vbo(log, this_lsn); + u32 end = + final_log_off(log, this_lsn, le32_to_cpu(rh->client_data_len)); + u32 hdr_off = end & ~log->sys_page_mask; + u64 seq = this_lsn >> log->file_data_bits; + struct RECORD_PAGE_HDR *page = NULL; + + /* Remember if we wrapped */ + if (end <= vbo) + seq += 1; + + /* log page header for this page */ + err = read_log_page(log, hdr_off, &page, NULL); + if (err) + return err; + + /* + * If the lsn we were given was not the last lsn on this page, + * then the starting offset for the next lsn is on a quad word + * boundary following the last file offset for the current lsn + * Otherwise the file offset is the start of the data on the next page + */ + if (this_lsn == le64_to_cpu(page->rhdr.lsn)) { + /* If we wrapped, we need to increment the sequence number */ + hdr_off = next_page_off(log, hdr_off); + if (hdr_off == log->first_page) + seq += 1; + + vbo = hdr_off + log->data_off; + } else { + vbo = QuadAlign(end); + } + + /* Compute the lsn based on the file offset and the sequence count */ + *lsn = vbo_to_lsn(log, vbo, seq); + + /* + * If this lsn is within the legal range for the file, we return true + * Otherwise false indicates that there are no more lsn's + */ + if (!is_lsn_in_file(log, *lsn)) + *lsn = 0; + + ntfs_free(page); + + return 0; +} + +/* + * current_log_avail + * + * calculate the number of bytes available for log records + */ +static u32 current_log_avail(struct ntfs_log *log) +{ + u32 oldest_off, next_free_off, free_bytes; + + if (log->l_flags & NTFSLOG_NO_LAST_LSN) { + /* The entire file is available */ + return log->max_current_avail; + } + + /* + * If there is a last lsn the restart area then we know that we will + * have to compute the free range + * If there is no oldest lsn then start at the first page of the file + */ + oldest_off = (log->l_flags & NTFSLOG_NO_OLDEST_LSN) + ? log->first_page + : (log->oldest_lsn_off & ~log->sys_page_mask); + + /* + * We will use the next log page offset to compute the next free page\ + * If we are going to reuse this page go to the next page + * If we are at the first page then use the end of the file + */ + next_free_off = (log->l_flags & NTFSLOG_REUSE_TAIL) + ? log->next_page + log->page_size + : log->next_page == log->first_page + ? log->l_size + : log->next_page; + + /* If the two offsets are the same then there is no available space */ + if (oldest_off == next_free_off) + return 0; + /* + * If the free offset follows the oldest offset then subtract + * this range from the total available pages + */ + free_bytes = + oldest_off < next_free_off + ? log->total_avail_pages - (next_free_off - oldest_off) + : oldest_off - next_free_off; + + free_bytes >>= log->page_bits; + return free_bytes * log->reserved; +} + +static bool check_subseq_log_page(struct ntfs_log *log, + const struct RECORD_PAGE_HDR *rp, u32 vbo, + u64 seq) +{ + u64 lsn_seq; + const struct NTFS_RECORD_HEADER *rhdr = &rp->rhdr; + u64 lsn = le64_to_cpu(rhdr->lsn); + + if (rhdr->sign == NTFS_FFFF_SIGNATURE || !rhdr->sign) + return false; + + /* + * If the last lsn on the page occurs was written after the page + * that caused the original error then we have a fatal error + */ + lsn_seq = lsn >> log->file_data_bits; + + /* + * If the sequence number for the lsn the page is equal or greater + * than lsn we expect, then this is a subsequent write + */ + return lsn_seq >= seq || + (lsn_seq == seq - 1 && log->first_page == vbo && + vbo != (lsn_to_vbo(log, lsn) & ~log->page_mask)); +} + +/* + * last_log_lsn + * + * This routine walks through the log pages for a file, searching for the + * last log page written to the file + */ +static int last_log_lsn(struct ntfs_log *log) +{ + int err; + bool usa_error = false; + bool replace_page = false; + bool reuse_page = log->l_flags & NTFSLOG_REUSE_TAIL; + bool wrapped_file, wrapped; + + u32 page_cnt = 1, page_pos = 1; + u32 page_off = 0, page_off1 = 0, saved_off = 0; + u32 final_off, second_off, final_off_prev = 0, second_off_prev = 0; + u32 first_file_off = 0, second_file_off = 0; + u32 part_io_count = 0; + u32 tails = 0; + u32 this_off, curpage_off, nextpage_off, remain_pages; + + u64 expected_seq, seq_base = 0, lsn_base = 0; + u64 best_lsn, best_lsn1, best_lsn2; + u64 lsn_cur, lsn1, lsn2; + u64 last_ok_lsn = reuse_page ? log->last_lsn : 0; + + u16 cur_pos, best_page_pos; + + struct RECORD_PAGE_HDR *page = NULL; + struct RECORD_PAGE_HDR *tst_page = NULL; + struct RECORD_PAGE_HDR *first_tail = NULL; + struct RECORD_PAGE_HDR *second_tail = NULL; + struct RECORD_PAGE_HDR *tail_page = NULL; + struct RECORD_PAGE_HDR *second_tail_prev = NULL; + struct RECORD_PAGE_HDR *first_tail_prev = NULL; + struct RECORD_PAGE_HDR *page_bufs = NULL; + struct RECORD_PAGE_HDR *best_page; + + if (log->major_ver >= 2) { + final_off = 0x02 * log->page_size; + second_off = 0x12 * log->page_size; + + // 0x10 == 0x12 - 0x2 + page_bufs = ntfs_malloc(log->page_size * 0x10); + if (!page_bufs) + return -ENOMEM; + } else { + second_off = log->first_page - log->page_size; + final_off = second_off - log->page_size; + } + +next_tail: + /* Read second tail page (at pos 3/0x12000) */ + if (read_log_page(log, second_off, &second_tail, &usa_error) || + usa_error || second_tail->rhdr.sign != NTFS_RCRD_SIGNATURE) { + ntfs_free(second_tail); + second_tail = NULL; + second_file_off = 0; + lsn2 = 0; + } else { + second_file_off = hdr_file_off(log, second_tail); + lsn2 = le64_to_cpu(second_tail->record_hdr.last_end_lsn); + } + + /* Read first tail page (at pos 2/0x2000 ) */ + if (read_log_page(log, final_off, &first_tail, &usa_error) || + usa_error || first_tail->rhdr.sign != NTFS_RCRD_SIGNATURE) { + ntfs_free(first_tail); + first_tail = NULL; + first_file_off = 0; + lsn1 = 0; + } else { + first_file_off = hdr_file_off(log, first_tail); + lsn1 = le64_to_cpu(first_tail->record_hdr.last_end_lsn); + } + + if (log->major_ver < 2) { + int best_page; + + first_tail_prev = first_tail; + final_off_prev = first_file_off; + second_tail_prev = second_tail; + second_off_prev = second_file_off; + tails = 1; + + if (!first_tail && !second_tail) + goto tail_read; + + if (first_tail && second_tail) + best_page = lsn1 < lsn2 ? 1 : 0; + else if (first_tail) + best_page = 0; + else + best_page = 1; + + page_off = best_page ? second_file_off : first_file_off; + seq_base = (best_page ? lsn2 : lsn1) >> log->file_data_bits; + goto tail_read; + } + + best_lsn1 = first_tail ? base_lsn(log, first_tail, first_file_off) : 0; + best_lsn2 = + second_tail ? base_lsn(log, second_tail, second_file_off) : 0; + + if (first_tail && second_tail) { + if (best_lsn1 > best_lsn2) { + best_lsn = best_lsn1; + best_page = first_tail; + this_off = first_file_off; + } else { + best_lsn = best_lsn2; + best_page = second_tail; + this_off = second_file_off; + } + } else if (first_tail) { + best_lsn = best_lsn1; + best_page = first_tail; + this_off = first_file_off; + } else if (second_tail) { + best_lsn = best_lsn2; + best_page = second_tail; + this_off = second_file_off; + } else { + goto tail_read; + } + + best_page_pos = le16_to_cpu(best_page->page_pos); + + if (!tails) { + if (best_page_pos == page_pos) { + seq_base = best_lsn >> log->file_data_bits; + saved_off = page_off = le32_to_cpu(best_page->file_off); + lsn_base = best_lsn; + + memmove(page_bufs, best_page, log->page_size); + + page_cnt = le16_to_cpu(best_page->page_count); + if (page_cnt > 1) + page_pos += 1; + + tails = 1; + } + } else if (seq_base == (best_lsn >> log->file_data_bits) && + saved_off + log->page_size == this_off && + lsn_base < best_lsn && + (page_pos != page_cnt || best_page_pos == page_pos || + best_page_pos == 1) && + (page_pos >= page_cnt || best_page_pos == page_pos)) { + u16 bppc = le16_to_cpu(best_page->page_count); + + saved_off += log->page_size; + lsn_base = best_lsn; + + memmove(Add2Ptr(page_bufs, tails * log->page_size), best_page, + log->page_size); + + tails += 1; + + if (best_page_pos != bppc) { + page_cnt = bppc; + page_pos = best_page_pos; + + if (page_cnt > 1) + page_pos += 1; + } else { + page_pos = page_cnt = 1; + } + } else { + ntfs_free(first_tail); + ntfs_free(second_tail); + goto tail_read; + } + + ntfs_free(first_tail_prev); + first_tail_prev = first_tail; + final_off_prev = first_file_off; + first_tail = NULL; + + ntfs_free(second_tail_prev); + second_tail_prev = second_tail; + second_off_prev = second_file_off; + second_tail = NULL; + + final_off += log->page_size; + second_off += log->page_size; + + if (tails < 0x10) + goto next_tail; +tail_read: + first_tail = first_tail_prev; + final_off = final_off_prev; + + second_tail = second_tail_prev; + second_off = second_off_prev; + + page_cnt = page_pos = 1; + + curpage_off = seq_base == log->seq_num ? min(log->next_page, page_off) + : log->next_page; + + wrapped_file = + curpage_off == log->first_page && + !(log->l_flags & (NTFSLOG_NO_LAST_LSN | NTFSLOG_REUSE_TAIL)); + + expected_seq = wrapped_file ? (log->seq_num + 1) : log->seq_num; + + nextpage_off = curpage_off; + +next_page: + tail_page = NULL; + /* Read the next log page */ + err = read_log_page(log, curpage_off, &page, &usa_error); + + /* Compute the next log page offset the file */ + nextpage_off = next_page_off(log, curpage_off); + wrapped = nextpage_off == log->first_page; + + if (tails > 1) { + struct RECORD_PAGE_HDR *cur_page = + Add2Ptr(page_bufs, curpage_off - page_off); + + if (curpage_off == saved_off) { + tail_page = cur_page; + goto use_tail_page; + } + + if (page_off > curpage_off || curpage_off >= saved_off) + goto use_tail_page; + + if (page_off1) + goto use_cur_page; + + if (!err && !usa_error && + page->rhdr.sign == NTFS_RCRD_SIGNATURE && + cur_page->rhdr.lsn == page->rhdr.lsn && + cur_page->record_hdr.next_record_off == + page->record_hdr.next_record_off && + ((page_pos == page_cnt && + le16_to_cpu(page->page_pos) == 1) || + (page_pos != page_cnt && + le16_to_cpu(page->page_pos) == page_pos + 1 && + le16_to_cpu(page->page_count) == page_cnt))) { + cur_page = NULL; + goto use_tail_page; + } + + page_off1 = page_off; + +use_cur_page: + + lsn_cur = le64_to_cpu(cur_page->rhdr.lsn); + + if (last_ok_lsn != + le64_to_cpu(cur_page->record_hdr.last_end_lsn) && + ((lsn_cur >> log->file_data_bits) + + ((curpage_off < + (lsn_to_vbo(log, lsn_cur) & ~log->page_mask)) + ? 1 + : 0)) != expected_seq) { + goto check_tail; + } + + if (!is_log_record_end(cur_page)) { + tail_page = NULL; + last_ok_lsn = lsn_cur; + goto next_page_1; + } + + log->seq_num = expected_seq; + log->l_flags &= ~NTFSLOG_NO_LAST_LSN; + log->last_lsn = le64_to_cpu(cur_page->record_hdr.last_end_lsn); + log->ra->current_lsn = cur_page->record_hdr.last_end_lsn; + + if (log->record_header_len <= + log->page_size - + le16_to_cpu(cur_page->record_hdr.next_record_off)) { + log->l_flags |= NTFSLOG_REUSE_TAIL; + log->next_page = curpage_off; + } else { + log->l_flags &= ~NTFSLOG_REUSE_TAIL; + log->next_page = nextpage_off; + } + + if (wrapped_file) + log->l_flags |= NTFSLOG_WRAPPED; + + last_ok_lsn = le64_to_cpu(cur_page->record_hdr.last_end_lsn); + goto next_page_1; + } + + /* + * If we are at the expected first page of a transfer check to see + * if either tail copy is at this offset + * If this page is the last page of a transfer, check if we wrote + * a subsequent tail copy + */ + if (page_cnt == page_pos || page_cnt == page_pos + 1) { + /* + * Check if the offset matches either the first or second + * tail copy. It is possible it will match both + */ + if (curpage_off == final_off) + tail_page = first_tail; + + /* + * If we already matched on the first page then + * check the ending lsn's. + */ + if (curpage_off == second_off) { + if (!tail_page || + (second_tail && + le64_to_cpu(second_tail->record_hdr.last_end_lsn) > + le64_to_cpu(first_tail->record_hdr + .last_end_lsn))) { + tail_page = second_tail; + } + } + } + +use_tail_page: + if (tail_page) { + /* we have a candidate for a tail copy */ + lsn_cur = le64_to_cpu(tail_page->record_hdr.last_end_lsn); + + if (last_ok_lsn < lsn_cur) { + /* + * If the sequence number is not expected, + * then don't use the tail copy + */ + if (expected_seq != (lsn_cur >> log->file_data_bits)) + tail_page = NULL; + } else if (last_ok_lsn > lsn_cur) { + /* + * If the last lsn is greater than the one on + * this page then forget this tail + */ + tail_page = NULL; + } + } + + /* If we have an error on the current page, we will break of this loop */ + if (err || usa_error) + goto check_tail; + + /* + * Done if the last lsn on this page doesn't match the previous known + * last lsn or the sequence number is not expected + */ + lsn_cur = le64_to_cpu(page->rhdr.lsn); + if (last_ok_lsn != lsn_cur && + expected_seq != (lsn_cur >> log->file_data_bits)) { + goto check_tail; + } + + /* + * Check that the page position and page count values are correct + * If this is the first page of a transfer the position must be 1 + * and the count will be unknown + */ + if (page_cnt == page_pos) { + if (page->page_pos != cpu_to_le16(1) && + (!reuse_page || page->page_pos != page->page_count)) { + /* + * If the current page is the first page we are + * looking at and we are reusing this page then + * it can be either the first or last page of a + * transfer. Otherwise it can only be the first. + */ + goto check_tail; + } + } else if (le16_to_cpu(page->page_count) != page_cnt || + le16_to_cpu(page->page_pos) != page_pos + 1) { + /* + * The page position better be 1 more than the last page + * position and the page count better match + */ + goto check_tail; + } + + /* + * We have a valid page the file and may have a valid page + * the tail copy area + * If the tail page was written after the page the file then + * break of the loop + */ + if (tail_page && + le64_to_cpu(tail_page->record_hdr.last_end_lsn) > lsn_cur) { + /* Remember if we will replace the page */ + replace_page = true; + goto check_tail; + } + + tail_page = NULL; + + if (is_log_record_end(page)) { + /* + * Since we have read this page we know the sequence number + * is the same as our expected value + */ + log->seq_num = expected_seq; + log->last_lsn = le64_to_cpu(page->record_hdr.last_end_lsn); + log->ra->current_lsn = page->record_hdr.last_end_lsn; + log->l_flags &= ~NTFSLOG_NO_LAST_LSN; + + /* + * If there is room on this page for another header then + * remember we want to reuse the page + */ + if (log->record_header_len <= + log->page_size - + le16_to_cpu(page->record_hdr.next_record_off)) { + log->l_flags |= NTFSLOG_REUSE_TAIL; + log->next_page = curpage_off; + } else { + log->l_flags &= ~NTFSLOG_REUSE_TAIL; + log->next_page = nextpage_off; + } + + /* Remember if we wrapped the log file */ + if (wrapped_file) + log->l_flags |= NTFSLOG_WRAPPED; + } + + /* + * Remember the last page count and position. + * Also remember the last known lsn + */ + page_cnt = le16_to_cpu(page->page_count); + page_pos = le16_to_cpu(page->page_pos); + last_ok_lsn = le64_to_cpu(page->rhdr.lsn); + +next_page_1: + + if (wrapped) { + expected_seq += 1; + wrapped_file = 1; + } + + curpage_off = nextpage_off; + ntfs_free(page); + page = NULL; + reuse_page = 0; + goto next_page; + +check_tail: + if (tail_page) { + log->seq_num = expected_seq; + log->last_lsn = le64_to_cpu(tail_page->record_hdr.last_end_lsn); + log->ra->current_lsn = tail_page->record_hdr.last_end_lsn; + log->l_flags &= ~NTFSLOG_NO_LAST_LSN; + + if (log->page_size - + le16_to_cpu( + tail_page->record_hdr.next_record_off) >= + log->record_header_len) { + log->l_flags |= NTFSLOG_REUSE_TAIL; + log->next_page = curpage_off; + } else { + log->l_flags &= ~NTFSLOG_REUSE_TAIL; + log->next_page = nextpage_off; + } + + if (wrapped) + log->l_flags |= NTFSLOG_WRAPPED; + } + + /* Remember that the partial IO will start at the next page */ + second_off = nextpage_off; + + /* + * If the next page is the first page of the file then update + * the sequence number for log records which begon the next page + */ + if (wrapped) + expected_seq += 1; + + /* + * If we have a tail copy or are performing single page I/O we can + * immediately look at the next page + */ + if (replace_page || (log->ra->flags & RESTART_SINGLE_PAGE_IO)) { + page_cnt = 2; + page_pos = 1; + goto check_valid; + } + + if (page_pos != page_cnt) + goto check_valid; + /* + * If the next page causes us to wrap to the beginning of the log + * file then we know which page to check next. + */ + if (wrapped) { + page_cnt = 2; + page_pos = 1; + goto check_valid; + } + + cur_pos = 2; + +next_test_page: + ntfs_free(tst_page); + tst_page = NULL; + + /* Walk through the file, reading log pages */ + err = read_log_page(log, nextpage_off, &tst_page, &usa_error); + + /* + * If we get a USA error then assume that we correctly found + * the end of the original transfer + */ + if (usa_error) + goto file_is_valid; + + /* + * If we were able to read the page, we examine it to see if it + * is the same or different Io block + */ + if (err) + goto next_test_page_1; + + if (le16_to_cpu(tst_page->page_pos) == cur_pos && + check_subseq_log_page(log, tst_page, nextpage_off, expected_seq)) { + page_cnt = le16_to_cpu(tst_page->page_count) + 1; + page_pos = le16_to_cpu(tst_page->page_pos); + goto check_valid; + } else { + goto file_is_valid; + } + +next_test_page_1: + + nextpage_off = next_page_off(log, curpage_off); + wrapped = nextpage_off == log->first_page; + + if (wrapped) { + expected_seq += 1; + page_cnt = 2; + page_pos = 1; + } + + cur_pos += 1; + part_io_count += 1; + if (!wrapped) + goto next_test_page; + +check_valid: + /* Skip over the remaining pages this transfer */ + remain_pages = page_cnt - page_pos - 1; + part_io_count += remain_pages; + + while (remain_pages--) { + nextpage_off = next_page_off(log, curpage_off); + wrapped = nextpage_off == log->first_page; + + if (wrapped) + expected_seq += 1; + } + + /* Call our routine to check this log page */ + ntfs_free(tst_page); + tst_page = NULL; + + err = read_log_page(log, nextpage_off, &tst_page, &usa_error); + if (!err && !usa_error && + check_subseq_log_page(log, tst_page, nextpage_off, expected_seq)) { + err = -EINVAL; + goto out; + } + +file_is_valid: + + /* We have a valid file */ + if (page_off1 || tail_page) { + struct RECORD_PAGE_HDR *tmp_page; + + if (sb_rdonly(log->ni->mi.sbi->sb)) { + err = -EROFS; + goto out; + } + + if (page_off1) { + tmp_page = Add2Ptr(page_bufs, page_off1 - page_off); + tails -= (page_off1 - page_off) / log->page_size; + if (!tail_page) + tails -= 1; + } else { + tmp_page = tail_page; + tails = 1; + } + + while (tails--) { + u64 off = hdr_file_off(log, tmp_page); + + if (!page) { + page = ntfs_malloc(log->page_size); + if (!page) + return -ENOMEM; + } + + /* + * Correct page and copy the data from this page + * into it and flush it to disk + */ + memcpy(page, tmp_page, log->page_size); + + /* Fill last flushed lsn value flush the page */ + if (log->major_ver < 2) + page->rhdr.lsn = page->record_hdr.last_end_lsn; + else + page->file_off = 0; + + page->page_pos = page->page_count = cpu_to_le16(1); + + ntfs_fix_pre_write(&page->rhdr, log->page_size); + + err = ntfs_sb_write_run(log->ni->mi.sbi, + &log->ni->file.run, off, page, + log->page_size); + + if (err) + goto out; + + if (part_io_count && second_off == off) { + second_off += log->page_size; + part_io_count -= 1; + } + + tmp_page = Add2Ptr(tmp_page, log->page_size); + } + } + + if (part_io_count) { + if (sb_rdonly(log->ni->mi.sbi->sb)) { + err = -EROFS; + goto out; + } + } + +out: + ntfs_free(second_tail); + ntfs_free(first_tail); + ntfs_free(page); + ntfs_free(tst_page); + ntfs_free(page_bufs); + + return err; +} + +/* + * read_log_rec_buf + * + * copies a log record from the file to a buffer + * The log record may span several log pages and may even wrap the file + */ +static int read_log_rec_buf(struct ntfs_log *log, + const struct LFS_RECORD_HDR *rh, void *buffer) +{ + int err; + struct RECORD_PAGE_HDR *ph = NULL; + u64 lsn = le64_to_cpu(rh->this_lsn); + u32 vbo = lsn_to_vbo(log, lsn) & ~log->page_mask; + u32 off = lsn_to_page_off(log, lsn) + log->record_header_len; + u32 data_len = le32_to_cpu(rh->client_data_len); + + /* + * While there are more bytes to transfer, + * we continue to attempt to perform the read + */ + for (;;) { + bool usa_error; + u32 tail = log->page_size - off; + + if (tail >= data_len) + tail = data_len; + + data_len -= tail; + + err = read_log_page(log, vbo, &ph, &usa_error); + if (err) + goto out; + + /* + * The last lsn on this page better be greater or equal + * to the lsn we are copying + */ + if (lsn > le64_to_cpu(ph->rhdr.lsn)) { + err = -EINVAL; + goto out; + } + + memcpy(buffer, Add2Ptr(ph, off), tail); + + /* If there are no more bytes to transfer, we exit the loop */ + if (!data_len) { + if (!is_log_record_end(ph) || + lsn > le64_to_cpu(ph->record_hdr.last_end_lsn)) { + err = -EINVAL; + goto out; + } + break; + } + + if (ph->rhdr.lsn == ph->record_hdr.last_end_lsn || + lsn > le64_to_cpu(ph->rhdr.lsn)) { + err = -EINVAL; + goto out; + } + + vbo = next_page_off(log, vbo); + off = log->data_off; + + /* + * adjust our pointer the user's buffer to transfer + * the next block to + */ + buffer = Add2Ptr(buffer, tail); + } + +out: + ntfs_free(ph); + return err; +} + +static int read_rst_area(struct ntfs_log *log, struct NTFS_RESTART **rst_, + u64 *lsn) +{ + int err; + struct LFS_RECORD_HDR *rh = NULL; + const struct CLIENT_REC *cr = + Add2Ptr(log->ra, le16_to_cpu(log->ra->client_off)); + u64 lsnr, lsnc = le64_to_cpu(cr->restart_lsn); + u32 len; + struct NTFS_RESTART *rst; + + *lsn = 0; + *rst_ = NULL; + + /* If the client doesn't have a restart area, go ahead and exit now */ + if (!lsnc) + return 0; + + err = read_log_page(log, lsn_to_vbo(log, lsnc), + (struct RECORD_PAGE_HDR **)&rh, NULL); + if (err) + return err; + + rst = NULL; + lsnr = le64_to_cpu(rh->this_lsn); + + if (lsnc != lsnr) { + /* If the lsn values don't match, then the disk is corrupt */ + err = -EINVAL; + goto out; + } + + *lsn = lsnr; + len = le32_to_cpu(rh->client_data_len); + + if (!len) { + err = 0; + goto out; + } + + if (len < sizeof(struct NTFS_RESTART)) { + err = -EINVAL; + goto out; + } + + rst = ntfs_malloc(len); + if (!rst) { + err = -ENOMEM; + goto out; + } + + /* Copy the data into the 'rst' buffer */ + err = read_log_rec_buf(log, rh, rst); + if (err) + goto out; + + *rst_ = rst; + rst = NULL; + +out: + ntfs_free(rh); + ntfs_free(rst); + + return err; +} + +static int find_log_rec(struct ntfs_log *log, u64 lsn, struct lcb *lcb) +{ + int err; + struct LFS_RECORD_HDR *rh = lcb->lrh; + u32 rec_len, len; + + /* Read the record header for this lsn */ + if (!rh) { + err = read_log_page(log, lsn_to_vbo(log, lsn), + (struct RECORD_PAGE_HDR **)&rh, NULL); + + lcb->lrh = rh; + if (err) + return err; + } + + /* + * If the lsn the log record doesn't match the desired + * lsn then the disk is corrupt + */ + if (lsn != le64_to_cpu(rh->this_lsn)) + return -EINVAL; + + len = le32_to_cpu(rh->client_data_len); + + /* + * check that the length field isn't greater than the total + * available space the log file + */ + rec_len = len + log->record_header_len; + if (rec_len >= log->total_avail) + return -EINVAL; + + /* + * If the entire log record is on this log page, + * put a pointer to the log record the context block + */ + if (rh->flags & LOG_RECORD_MULTI_PAGE) { + void *lr = ntfs_malloc(len); + + if (!lr) + return -ENOMEM; + + lcb->log_rec = lr; + lcb->alloc = true; + + /* Copy the data into the buffer returned */ + err = read_log_rec_buf(log, rh, lr); + if (err) + return err; + } else { + /* If beyond the end of the current page -> an error */ + u32 page_off = lsn_to_page_off(log, lsn); + + if (page_off + len + log->record_header_len > log->page_size) + return -EINVAL; + + lcb->log_rec = Add2Ptr(rh, sizeof(struct LFS_RECORD_HDR)); + lcb->alloc = false; + } + + return 0; +} + +/* + * read_log_rec_lcb + * + * initiates the query operation. + */ +static int read_log_rec_lcb(struct ntfs_log *log, u64 lsn, u32 ctx_mode, + struct lcb **lcb_) +{ + int err; + const struct CLIENT_REC *cr; + struct lcb *lcb; + + switch (ctx_mode) { + case lcb_ctx_undo_next: + case lcb_ctx_prev: + case lcb_ctx_next: + break; + default: + return -EINVAL; + } + + /* check that the given lsn is the legal range for this client */ + cr = Add2Ptr(log->ra, le16_to_cpu(log->ra->client_off)); + + if (!verify_client_lsn(log, cr, lsn)) + return -EINVAL; + + lcb = ntfs_zalloc(sizeof(struct lcb)); + if (!lcb) + return -ENOMEM; + lcb->client = log->client_id; + lcb->ctx_mode = ctx_mode; + + /* Find the log record indicated by the given lsn */ + err = find_log_rec(log, lsn, lcb); + if (err) + goto out; + + *lcb_ = lcb; + return 0; + +out: + lcb_put(lcb); + *lcb_ = NULL; + return err; +} + +/* + * find_client_next_lsn + * + * attempt to find the next lsn to return to a client based on the context mode. + */ +static int find_client_next_lsn(struct ntfs_log *log, struct lcb *lcb, u64 *lsn) +{ + int err; + u64 next_lsn; + struct LFS_RECORD_HDR *hdr; + + hdr = lcb->lrh; + *lsn = 0; + + if (lcb_ctx_next != lcb->ctx_mode) + goto check_undo_next; + + /* Loop as long as another lsn can be found */ + for (;;) { + u64 current_lsn; + + err = next_log_lsn(log, hdr, &current_lsn); + if (err) + goto out; + + if (!current_lsn) + break; + + if (hdr != lcb->lrh) + ntfs_free(hdr); + + hdr = NULL; + err = read_log_page(log, lsn_to_vbo(log, current_lsn), + (struct RECORD_PAGE_HDR **)&hdr, NULL); + if (err) + goto out; + + if (memcmp(&hdr->client, &lcb->client, + sizeof(struct CLIENT_ID))) { + /*err = -EINVAL; */ + } else if (LfsClientRecord == hdr->record_type) { + ntfs_free(lcb->lrh); + lcb->lrh = hdr; + *lsn = current_lsn; + return 0; + } + } + +out: + if (hdr != lcb->lrh) + ntfs_free(hdr); + return err; + +check_undo_next: + if (lcb_ctx_undo_next == lcb->ctx_mode) + next_lsn = le64_to_cpu(hdr->client_undo_next_lsn); + else if (lcb_ctx_prev == lcb->ctx_mode) + next_lsn = le64_to_cpu(hdr->client_prev_lsn); + else + return 0; + + if (!next_lsn) + return 0; + + if (!verify_client_lsn( + log, Add2Ptr(log->ra, le16_to_cpu(log->ra->client_off)), + next_lsn)) + return 0; + + hdr = NULL; + err = read_log_page(log, lsn_to_vbo(log, next_lsn), + (struct RECORD_PAGE_HDR **)&hdr, NULL); + if (err) + return err; + ntfs_free(lcb->lrh); + lcb->lrh = hdr; + + *lsn = next_lsn; + + return 0; +} + +static int read_next_log_rec(struct ntfs_log *log, struct lcb *lcb, u64 *lsn) +{ + int err; + + err = find_client_next_lsn(log, lcb, lsn); + if (err) + return err; + + if (!*lsn) + return 0; + + if (lcb->alloc) + ntfs_free(lcb->log_rec); + + lcb->log_rec = NULL; + lcb->alloc = false; + ntfs_free(lcb->lrh); + lcb->lrh = NULL; + + return find_log_rec(log, *lsn, lcb); +} + +static inline bool check_index_header(const struct INDEX_HDR *hdr, size_t bytes) +{ + __le16 mask; + u32 min_de, de_off, used, total; + const struct NTFS_DE *e; + + if (hdr_has_subnode(hdr)) { + min_de = sizeof(struct NTFS_DE) + sizeof(u64); + mask = NTFS_IE_HAS_SUBNODES; + } else { + min_de = sizeof(struct NTFS_DE); + mask = 0; + } + + de_off = le32_to_cpu(hdr->de_off); + used = le32_to_cpu(hdr->used); + total = le32_to_cpu(hdr->total); + + if (de_off > bytes - min_de || used > bytes || total > bytes || + de_off + min_de > used || used > total) { + return false; + } + + e = Add2Ptr(hdr, de_off); + for (;;) { + u16 esize = le16_to_cpu(e->size); + struct NTFS_DE *next = Add2Ptr(e, esize); + + if (esize < min_de || PtrOffset(hdr, next) > used || + (e->flags & NTFS_IE_HAS_SUBNODES) != mask) { + return false; + } + + if (de_is_last(e)) + break; + + e = next; + } + + return true; +} + +static inline bool check_index_buffer(const struct INDEX_BUFFER *ib, u32 bytes) +{ + u16 fo; + const struct NTFS_RECORD_HEADER *r = &ib->rhdr; + + if (r->sign != NTFS_INDX_SIGNATURE) + return false; + + fo = (SECTOR_SIZE - ((bytes >> SECTOR_SHIFT) + 1) * sizeof(short)); + + if (le16_to_cpu(r->fix_off) > fo) + return false; + + if ((le16_to_cpu(r->fix_num) - 1) * SECTOR_SIZE != bytes) + return false; + + return check_index_header(&ib->ihdr, + bytes - offsetof(struct INDEX_BUFFER, ihdr)); +} + +static inline bool check_index_root(const struct ATTRIB *attr, + struct ntfs_sb_info *sbi) +{ + bool ret; + const struct INDEX_ROOT *root = resident_data(attr); + u8 index_bits = le32_to_cpu(root->index_block_size) >= sbi->cluster_size + ? sbi->cluster_bits + : SECTOR_SHIFT; + u8 block_clst = root->index_block_clst; + + if (le32_to_cpu(attr->res.data_size) < sizeof(struct INDEX_ROOT) || + (root->type != ATTR_NAME && root->type != ATTR_ZERO) || + (root->type == ATTR_NAME && + root->rule != NTFS_COLLATION_TYPE_FILENAME) || + (le32_to_cpu(root->index_block_size) != + (block_clst << index_bits)) || + (block_clst != 1 && block_clst != 2 && block_clst != 4 && + block_clst != 8 && block_clst != 0x10 && block_clst != 0x20 && + block_clst != 0x40 && block_clst != 0x80)) { + return false; + } + + ret = check_index_header(&root->ihdr, + le32_to_cpu(attr->res.data_size) - + offsetof(struct INDEX_ROOT, ihdr)); + return ret; +} + +static inline bool check_attr(const struct MFT_REC *rec, + const struct ATTRIB *attr, + struct ntfs_sb_info *sbi) +{ + u32 asize = le32_to_cpu(attr->size); + u32 rsize = 0; + u64 dsize, svcn, evcn; + u16 run_off; + + /* Check the fixed part of the attribute record header */ + if (asize >= sbi->record_size || + asize + PtrOffset(rec, attr) >= sbi->record_size || + (attr->name_len && + le16_to_cpu(attr->name_off) + attr->name_len * sizeof(short) > + asize)) { + return false; + } + + /* Check the attribute fields */ + switch (attr->non_res) { + case 0: + rsize = le32_to_cpu(attr->res.data_size); + if (rsize >= asize || + le16_to_cpu(attr->res.data_off) + rsize > asize) { + return false; + } + break; + + case 1: + dsize = le64_to_cpu(attr->nres.data_size); + svcn = le64_to_cpu(attr->nres.svcn); + evcn = le64_to_cpu(attr->nres.evcn); + run_off = le16_to_cpu(attr->nres.run_off); + + if (svcn > evcn + 1 || run_off >= asize || + le64_to_cpu(attr->nres.valid_size) > dsize || + dsize > le64_to_cpu(attr->nres.alloc_size)) { + return false; + } + + if (run_unpack(NULL, sbi, 0, svcn, evcn, svcn, + Add2Ptr(attr, run_off), asize - run_off) < 0) { + return false; + } + + return true; + + default: + return false; + } + + switch (attr->type) { + case ATTR_NAME: + if (fname_full_size(Add2Ptr( + attr, le16_to_cpu(attr->res.data_off))) > asize) { + return false; + } + break; + + case ATTR_ROOT: + return check_index_root(attr, sbi); + + case ATTR_STD: + if (rsize < sizeof(struct ATTR_STD_INFO5) && + rsize != sizeof(struct ATTR_STD_INFO)) { + return false; + } + break; + + case ATTR_LIST: + case ATTR_ID: + case ATTR_SECURE: + case ATTR_LABEL: + case ATTR_VOL_INFO: + case ATTR_DATA: + case ATTR_ALLOC: + case ATTR_BITMAP: + case ATTR_REPARSE: + case ATTR_EA_INFO: + case ATTR_EA: + case ATTR_PROPERTYSET: + case ATTR_LOGGED_UTILITY_STREAM: + break; + + default: + return false; + } + + return true; +} + +static inline bool check_file_record(const struct MFT_REC *rec, + const struct MFT_REC *rec2, + struct ntfs_sb_info *sbi) +{ + const struct ATTRIB *attr; + u16 fo = le16_to_cpu(rec->rhdr.fix_off); + u16 fn = le16_to_cpu(rec->rhdr.fix_num); + u16 ao = le16_to_cpu(rec->attr_off); + u32 rs = sbi->record_size; + + /* check the file record header for consistency */ + if (rec->rhdr.sign != NTFS_FILE_SIGNATURE || + fo > (SECTOR_SIZE - ((rs >> SECTOR_SHIFT) + 1) * sizeof(short)) || + (fn - 1) * SECTOR_SIZE != rs || ao < MFTRECORD_FIXUP_OFFSET_1 || + ao > sbi->record_size - SIZEOF_RESIDENT || !is_rec_inuse(rec) || + le32_to_cpu(rec->total) != rs) { + return false; + } + + /* Loop to check all of the attributes */ + for (attr = Add2Ptr(rec, ao); attr->type != ATTR_END; + attr = Add2Ptr(attr, le32_to_cpu(attr->size))) { + if (check_attr(rec, attr, sbi)) + continue; + return false; + } + + return true; +} + +static inline int check_lsn(const struct NTFS_RECORD_HEADER *hdr, + const u64 *rlsn) +{ + u64 lsn; + + if (!rlsn) + return true; + + lsn = le64_to_cpu(hdr->lsn); + + if (hdr->sign == NTFS_HOLE_SIGNATURE) + return false; + + if (*rlsn > lsn) + return true; + + return false; +} + +static inline bool check_if_attr(const struct MFT_REC *rec, + const struct LOG_REC_HDR *lrh) +{ + u16 ro = le16_to_cpu(lrh->record_off); + u16 o = le16_to_cpu(rec->attr_off); + const struct ATTRIB *attr = Add2Ptr(rec, o); + + while (o < ro) { + u32 asize; + + if (attr->type == ATTR_END) + break; + + asize = le32_to_cpu(attr->size); + if (!asize) + break; + + o += asize; + attr = Add2Ptr(attr, asize); + } + + return o == ro; +} + +static inline bool check_if_index_root(const struct MFT_REC *rec, + const struct LOG_REC_HDR *lrh) +{ + u16 ro = le16_to_cpu(lrh->record_off); + u16 o = le16_to_cpu(rec->attr_off); + const struct ATTRIB *attr = Add2Ptr(rec, o); + + while (o < ro) { + u32 asize; + + if (attr->type == ATTR_END) + break; + + asize = le32_to_cpu(attr->size); + if (!asize) + break; + + o += asize; + attr = Add2Ptr(attr, asize); + } + + return o == ro && attr->type == ATTR_ROOT; +} + +static inline bool check_if_root_index(const struct ATTRIB *attr, + const struct INDEX_HDR *hdr, + const struct LOG_REC_HDR *lrh) +{ + u16 ao = le16_to_cpu(lrh->attr_off); + u32 de_off = le32_to_cpu(hdr->de_off); + u32 o = PtrOffset(attr, hdr) + de_off; + const struct NTFS_DE *e = Add2Ptr(hdr, de_off); + u32 asize = le32_to_cpu(attr->size); + + while (o < ao) { + u16 esize; + + if (o >= asize) + break; + + esize = le16_to_cpu(e->size); + if (!esize) + break; + + o += esize; + e = Add2Ptr(e, esize); + } + + return o == ao; +} + +static inline bool check_if_alloc_index(const struct INDEX_HDR *hdr, + u32 attr_off) +{ + u32 de_off = le32_to_cpu(hdr->de_off); + u32 o = offsetof(struct INDEX_BUFFER, ihdr) + de_off; + const struct NTFS_DE *e = Add2Ptr(hdr, de_off); + u32 used = le32_to_cpu(hdr->used); + + while (o < attr_off) { + u16 esize; + + if (de_off >= used) + break; + + esize = le16_to_cpu(e->size); + if (!esize) + break; + + o += esize; + de_off += esize; + e = Add2Ptr(e, esize); + } + + return o == attr_off; +} + +static inline void change_attr_size(struct MFT_REC *rec, struct ATTRIB *attr, + u32 nsize) +{ + u32 asize = le32_to_cpu(attr->size); + int dsize = nsize - asize; + u8 *next = Add2Ptr(attr, asize); + u32 used = le32_to_cpu(rec->used); + + memmove(Add2Ptr(attr, nsize), next, used - PtrOffset(rec, next)); + + rec->used = cpu_to_le32(used + dsize); + attr->size = cpu_to_le32(nsize); +} + +struct OpenAttr { + struct ATTRIB *attr; + struct runs_tree *run1; + struct runs_tree run0; + struct ntfs_inode *ni; + // CLST rno; +}; + +/* Returns 0 if 'attr' has the same type and name */ +static inline int cmp_type_and_name(const struct ATTRIB *a1, + const struct ATTRIB *a2) +{ + return a1->type != a2->type || a1->name_len != a2->name_len || + (a1->name_len && memcmp(attr_name(a1), attr_name(a2), + a1->name_len * sizeof(short))); +} + +static struct OpenAttr *find_loaded_attr(struct ntfs_log *log, + const struct ATTRIB *attr, CLST rno) +{ + struct OPEN_ATTR_ENRTY *oe = NULL; + + while ((oe = enum_rstbl(log->open_attr_tbl, oe))) { + struct OpenAttr *op_attr; + + if (ino_get(&oe->ref) != rno) + continue; + + op_attr = (struct OpenAttr *)oe->ptr; + if (!cmp_type_and_name(op_attr->attr, attr)) + return op_attr; + } + return NULL; +} + +static struct ATTRIB *attr_create_nonres_log(struct ntfs_sb_info *sbi, + enum ATTR_TYPE type, u64 size, + const u16 *name, size_t name_len, + __le16 flags) +{ + struct ATTRIB *attr; + u32 name_size = QuadAlign(name_len * sizeof(short)); + bool is_ext = flags & (ATTR_FLAG_COMPRESSED | ATTR_FLAG_SPARSED); + u32 asize = name_size + + (is_ext ? SIZEOF_NONRESIDENT_EX : SIZEOF_NONRESIDENT); + + attr = ntfs_zalloc(asize); + if (!attr) + return NULL; + + attr->type = type; + attr->size = cpu_to_le32(asize); + attr->flags = flags; + attr->non_res = 1; + attr->name_len = name_len; + + attr->nres.evcn = cpu_to_le64((u64)bytes_to_cluster(sbi, size) - 1); + attr->nres.alloc_size = cpu_to_le64(ntfs_up_cluster(sbi, size)); + attr->nres.data_size = cpu_to_le64(size); + attr->nres.valid_size = attr->nres.data_size; + if (is_ext) { + attr->name_off = SIZEOF_NONRESIDENT_EX_LE; + if (is_attr_compressed(attr)) + attr->nres.c_unit = COMPRESSION_UNIT; + + attr->nres.run_off = + cpu_to_le16(SIZEOF_NONRESIDENT_EX + name_size); + memcpy(Add2Ptr(attr, SIZEOF_NONRESIDENT_EX), name, + name_len * sizeof(short)); + } else { + attr->name_off = SIZEOF_NONRESIDENT_LE; + attr->nres.run_off = + cpu_to_le16(SIZEOF_NONRESIDENT + name_size); + memcpy(Add2Ptr(attr, SIZEOF_NONRESIDENT), name, + name_len * sizeof(short)); + } + + return attr; +} + +/* + * do_action + * + * common routine for the Redo and Undo Passes + * If rlsn is NULL then undo + */ +static int do_action(struct ntfs_log *log, struct OPEN_ATTR_ENRTY *oe, + const struct LOG_REC_HDR *lrh, u32 op, void *data, + u32 dlen, u32 rec_len, const u64 *rlsn) +{ + int err = 0; + struct ntfs_sb_info *sbi = log->ni->mi.sbi; + struct inode *inode = NULL, *inode_parent; + struct mft_inode *mi = NULL, *mi2_child = NULL; + CLST rno = 0, rno_base = 0; + struct INDEX_BUFFER *ib = NULL; + struct MFT_REC *rec = NULL; + struct ATTRIB *attr = NULL, *attr2; + struct INDEX_HDR *hdr; + struct INDEX_ROOT *root; + struct NTFS_DE *e, *e1, *e2; + struct NEW_ATTRIBUTE_SIZES *new_sz; + struct ATTR_FILE_NAME *fname; + struct OpenAttr *oa, *oa2; + u32 nsize, t32, asize, used, esize, bmp_off, bmp_bits; + u16 id, id2; + u32 record_size = sbi->record_size; + u64 t64; + u16 roff = le16_to_cpu(lrh->record_off); + u16 aoff = le16_to_cpu(lrh->attr_off); + u64 lco = 0; + u64 cbo = (u64)le16_to_cpu(lrh->cluster_off) << SECTOR_SHIFT; + u64 tvo = le64_to_cpu(lrh->target_vcn) << sbi->cluster_bits; + u64 vbo = cbo + tvo; + void *buffer_le = NULL; + u32 bytes = 0; + bool a_dirty = false; + u16 data_off; + + oa = oe->ptr; + + /* Big switch to prepare */ + switch (op) { + /* ============================================================ + * Process MFT records, as described by the current log record + * ============================================================ + */ + case InitializeFileRecordSegment: + case DeallocateFileRecordSegment: + case WriteEndOfFileRecordSegment: + case CreateAttribute: + case DeleteAttribute: + case UpdateResidentValue: + case UpdateMappingPairs: + case SetNewAttributeSizes: + case AddIndexEntryRoot: + case DeleteIndexEntryRoot: + case SetIndexEntryVcnRoot: + case UpdateFileNameRoot: + case UpdateRecordDataRoot: + case ZeroEndOfFileRecord: + rno = vbo >> sbi->record_bits; + inode = ilookup(sbi->sb, rno); + if (inode) { + mi = &ntfs_i(inode)->mi; + } else if (op == InitializeFileRecordSegment) { + mi = ntfs_zalloc(sizeof(struct mft_inode)); + if (!mi) + return -ENOMEM; + err = mi_format_new(mi, sbi, rno, 0, false); + if (err) + goto out; + } else { + /* read from disk */ + err = mi_get(sbi, rno, &mi); + if (err) + return err; + } + rec = mi->mrec; + + if (op == DeallocateFileRecordSegment) + goto skip_load_parent; + + if (InitializeFileRecordSegment != op) { + if (rec->rhdr.sign == NTFS_BAAD_SIGNATURE) + goto dirty_vol; + if (!check_lsn(&rec->rhdr, rlsn)) + goto out; + if (!check_file_record(rec, NULL, sbi)) + goto dirty_vol; + attr = Add2Ptr(rec, roff); + } + + if (is_rec_base(rec) || InitializeFileRecordSegment == op) { + rno_base = rno; + goto skip_load_parent; + } + + rno_base = ino_get(&rec->parent_ref); + inode_parent = ntfs_iget5(sbi->sb, &rec->parent_ref, NULL); + if (IS_ERR(inode_parent)) + goto skip_load_parent; + + if (is_bad_inode(inode_parent)) { + iput(inode_parent); + goto skip_load_parent; + } + + if (ni_load_mi_ex(ntfs_i(inode_parent), rno, &mi2_child)) { + iput(inode_parent); + } else { + if (mi2_child->mrec != mi->mrec) + memcpy(mi2_child->mrec, mi->mrec, + sbi->record_size); + + if (inode) + iput(inode); + else if (mi) + mi_put(mi); + + inode = inode_parent; + mi = mi2_child; + rec = mi2_child->mrec; + attr = Add2Ptr(rec, roff); + } + +skip_load_parent: + inode_parent = NULL; + break; + + /* ============================================================ + * Process attributes, as described by the current log record + * ============================================================ + */ + case UpdateNonresidentValue: + case AddIndexEntryAllocation: + case DeleteIndexEntryAllocation: + case WriteEndOfIndexBuffer: + case SetIndexEntryVcnAllocation: + case UpdateFileNameAllocation: + case SetBitsInNonresidentBitMap: + case ClearBitsInNonresidentBitMap: + case UpdateRecordDataAllocation: + attr = oa->attr; + bytes = UpdateNonresidentValue == op ? dlen : 0; + lco = (u64)le16_to_cpu(lrh->lcns_follow) << sbi->cluster_bits; + + if (attr->type == ATTR_ALLOC) { + t32 = le32_to_cpu(oe->bytes_per_index); + if (bytes < t32) + bytes = t32; + } + + if (!bytes) + bytes = lco - cbo; + + bytes += roff; + if (attr->type == ATTR_ALLOC) + bytes = (bytes + 511) & ~511; // align + + buffer_le = ntfs_malloc(bytes); + if (!buffer_le) + return -ENOMEM; + + err = ntfs_read_run_nb(sbi, oa->run1, vbo, buffer_le, bytes, + NULL); + if (err) + goto out; + + if (attr->type == ATTR_ALLOC && *(int *)buffer_le) + ntfs_fix_post_read(buffer_le, bytes, false); + break; + + default: + WARN_ON(1); + } + + /* Big switch to do operation */ + switch (op) { + case InitializeFileRecordSegment: + if (roff + dlen > record_size) + goto dirty_vol; + + memcpy(Add2Ptr(rec, roff), data, dlen); + mi->dirty = true; + break; + + case DeallocateFileRecordSegment: + clear_rec_inuse(rec); + le16_add_cpu(&rec->seq, 1); + mi->dirty = true; + break; + + case WriteEndOfFileRecordSegment: + attr2 = (struct ATTRIB *)data; + if (!check_if_attr(rec, lrh) || roff + dlen > record_size) + goto dirty_vol; + + memmove(attr, attr2, dlen); + rec->used = cpu_to_le32(QuadAlign(roff + dlen)); + + mi->dirty = true; + break; + + case CreateAttribute: + attr2 = (struct ATTRIB *)data; + asize = le32_to_cpu(attr2->size); + used = le32_to_cpu(rec->used); + + if (!check_if_attr(rec, lrh) || dlen < SIZEOF_RESIDENT || + !IsQuadAligned(asize) || + Add2Ptr(attr2, asize) > Add2Ptr(lrh, rec_len) || + dlen > record_size - used) { + goto dirty_vol; + } + + memmove(Add2Ptr(attr, asize), attr, used - roff); + memcpy(attr, attr2, asize); + + rec->used = cpu_to_le32(used + asize); + id = le16_to_cpu(rec->next_attr_id); + id2 = le16_to_cpu(attr2->id); + if (id <= id2) + rec->next_attr_id = cpu_to_le16(id2 + 1); + if (is_attr_indexed(attr)) + le16_add_cpu(&rec->hard_links, 1); + + oa2 = find_loaded_attr(log, attr, rno_base); + if (oa2) { + void *p2 = ntfs_memdup(attr, le32_to_cpu(attr->size)); + + if (p2) { + // run_close(oa2->run1); + ntfs_free(oa2->attr); + oa2->attr = p2; + } + } + + mi->dirty = true; + break; + + case DeleteAttribute: + asize = le32_to_cpu(attr->size); + used = le32_to_cpu(rec->used); + + if (!check_if_attr(rec, lrh)) + goto dirty_vol; + + rec->used = cpu_to_le32(used - asize); + if (is_attr_indexed(attr)) + le16_add_cpu(&rec->hard_links, -1); + + memmove(attr, Add2Ptr(attr, asize), used - asize - roff); + + mi->dirty = true; + break; + + case UpdateResidentValue: + nsize = aoff + dlen; + + if (!check_if_attr(rec, lrh)) + goto dirty_vol; + + asize = le32_to_cpu(attr->size); + used = le32_to_cpu(rec->used); + + if (lrh->redo_len == lrh->undo_len) { + if (nsize > asize) + goto dirty_vol; + goto move_data; + } + + if (nsize > asize && nsize - asize > record_size - used) + goto dirty_vol; + + nsize = QuadAlign(nsize); + data_off = le16_to_cpu(attr->res.data_off); + + if (nsize < asize) { + memmove(Add2Ptr(attr, aoff), data, dlen); + data = NULL; // To skip below memmove + } + + memmove(Add2Ptr(attr, nsize), Add2Ptr(attr, asize), + used - le16_to_cpu(lrh->record_off) - asize); + + rec->used = cpu_to_le32(used + nsize - asize); + attr->size = cpu_to_le32(nsize); + attr->res.data_size = cpu_to_le32(aoff + dlen - data_off); + +move_data: + if (data) + memmove(Add2Ptr(attr, aoff), data, dlen); + + oa2 = find_loaded_attr(log, attr, rno_base); + if (oa2) { + void *p2 = ntfs_memdup(attr, le32_to_cpu(attr->size)); + + if (p2) { + // run_close(&oa2->run0); + oa2->run1 = &oa2->run0; + ntfs_free(oa2->attr); + oa2->attr = p2; + } + } + + mi->dirty = true; + break; + + case UpdateMappingPairs: + nsize = aoff + dlen; + asize = le32_to_cpu(attr->size); + used = le32_to_cpu(rec->used); + + if (!check_if_attr(rec, lrh) || !attr->non_res || + aoff < le16_to_cpu(attr->nres.run_off) || aoff > asize || + (nsize > asize && nsize - asize > record_size - used)) { + goto dirty_vol; + } + + nsize = QuadAlign(nsize); + + memmove(Add2Ptr(attr, nsize), Add2Ptr(attr, asize), + used - le16_to_cpu(lrh->record_off) - asize); + rec->used = cpu_to_le32(used + nsize - asize); + attr->size = cpu_to_le32(nsize); + memmove(Add2Ptr(attr, aoff), data, dlen); + + if (run_get_highest_vcn(le64_to_cpu(attr->nres.svcn), + attr_run(attr), &t64)) { + goto dirty_vol; + } + + attr->nres.evcn = cpu_to_le64(t64); + oa2 = find_loaded_attr(log, attr, rno_base); + if (oa2 && oa2->attr->non_res) + oa2->attr->nres.evcn = attr->nres.evcn; + + mi->dirty = true; + break; + + case SetNewAttributeSizes: + new_sz = data; + if (!check_if_attr(rec, lrh) || !attr->non_res) + goto dirty_vol; + + attr->nres.alloc_size = new_sz->alloc_size; + attr->nres.data_size = new_sz->data_size; + attr->nres.valid_size = new_sz->valid_size; + + if (dlen >= sizeof(struct NEW_ATTRIBUTE_SIZES)) + attr->nres.total_size = new_sz->total_size; + + oa2 = find_loaded_attr(log, attr, rno_base); + if (oa2) { + void *p2 = ntfs_memdup(attr, le32_to_cpu(attr->size)); + + if (p2) { + ntfs_free(oa2->attr); + oa2->attr = p2; + } + } + mi->dirty = true; + break; + + case AddIndexEntryRoot: + e = (struct NTFS_DE *)data; + esize = le16_to_cpu(e->size); + root = resident_data(attr); + hdr = &root->ihdr; + used = le32_to_cpu(hdr->used); + + if (!check_if_index_root(rec, lrh) || + !check_if_root_index(attr, hdr, lrh) || + Add2Ptr(data, esize) > Add2Ptr(lrh, rec_len) || + esize > le32_to_cpu(rec->total) - le32_to_cpu(rec->used)) { + goto dirty_vol; + } + + e1 = Add2Ptr(attr, le16_to_cpu(lrh->attr_off)); + + change_attr_size(rec, attr, le32_to_cpu(attr->size) + esize); + + memmove(Add2Ptr(e1, esize), e1, + PtrOffset(e1, Add2Ptr(hdr, used))); + memmove(e1, e, esize); + + le32_add_cpu(&attr->res.data_size, esize); + hdr->used = cpu_to_le32(used + esize); + le32_add_cpu(&hdr->total, esize); + + mi->dirty = true; + break; + + case DeleteIndexEntryRoot: + root = resident_data(attr); + hdr = &root->ihdr; + used = le32_to_cpu(hdr->used); + + if (!check_if_index_root(rec, lrh) || + !check_if_root_index(attr, hdr, lrh)) { + goto dirty_vol; + } + + e1 = Add2Ptr(attr, le16_to_cpu(lrh->attr_off)); + esize = le16_to_cpu(e1->size); + e2 = Add2Ptr(e1, esize); + + memmove(e1, e2, PtrOffset(e2, Add2Ptr(hdr, used))); + + le32_sub_cpu(&attr->res.data_size, esize); + hdr->used = cpu_to_le32(used - esize); + le32_sub_cpu(&hdr->total, esize); + + change_attr_size(rec, attr, le32_to_cpu(attr->size) - esize); + + mi->dirty = true; + break; + + case SetIndexEntryVcnRoot: + root = resident_data(attr); + hdr = &root->ihdr; + + if (!check_if_index_root(rec, lrh) || + !check_if_root_index(attr, hdr, lrh)) { + goto dirty_vol; + } + + e = Add2Ptr(attr, le16_to_cpu(lrh->attr_off)); + + de_set_vbn_le(e, *(__le64 *)data); + mi->dirty = true; + break; + + case UpdateFileNameRoot: + root = resident_data(attr); + hdr = &root->ihdr; + + if (!check_if_index_root(rec, lrh) || + !check_if_root_index(attr, hdr, lrh)) { + goto dirty_vol; + } + + e = Add2Ptr(attr, le16_to_cpu(lrh->attr_off)); + fname = (struct ATTR_FILE_NAME *)(e + 1); + memmove(&fname->dup, data, sizeof(fname->dup)); // + mi->dirty = true; + break; + + case UpdateRecordDataRoot: + root = resident_data(attr); + hdr = &root->ihdr; + + if (!check_if_index_root(rec, lrh) || + !check_if_root_index(attr, hdr, lrh)) { + goto dirty_vol; + } + + e = Add2Ptr(attr, le16_to_cpu(lrh->attr_off)); + + memmove(Add2Ptr(e, le16_to_cpu(e->view.data_off)), data, dlen); + + mi->dirty = true; + break; + + case ZeroEndOfFileRecord: + if (roff + dlen > record_size) + goto dirty_vol; + + memset(attr, 0, dlen); + mi->dirty = true; + break; + + case UpdateNonresidentValue: + if (lco < cbo + roff + dlen) + goto dirty_vol; + + memcpy(Add2Ptr(buffer_le, roff), data, dlen); + + a_dirty = true; + if (attr->type == ATTR_ALLOC) + ntfs_fix_pre_write(buffer_le, bytes); + break; + + case AddIndexEntryAllocation: + ib = Add2Ptr(buffer_le, roff); + hdr = &ib->ihdr; + e = data; + esize = le16_to_cpu(e->size); + e1 = Add2Ptr(ib, aoff); + + if (is_baad(&ib->rhdr)) + goto dirty_vol; + if (!check_lsn(&ib->rhdr, rlsn)) + goto out; + + used = le32_to_cpu(hdr->used); + + if (!check_index_buffer(ib, bytes) || + !check_if_alloc_index(hdr, aoff) || + Add2Ptr(e, esize) > Add2Ptr(lrh, rec_len) || + used + esize > le32_to_cpu(hdr->total)) { + goto dirty_vol; + } + + memmove(Add2Ptr(e1, esize), e1, + PtrOffset(e1, Add2Ptr(hdr, used))); + memcpy(e1, e, esize); + + hdr->used = cpu_to_le32(used + esize); + + a_dirty = true; + + ntfs_fix_pre_write(&ib->rhdr, bytes); + break; + + case DeleteIndexEntryAllocation: + ib = Add2Ptr(buffer_le, roff); + hdr = &ib->ihdr; + e = Add2Ptr(ib, aoff); + esize = le16_to_cpu(e->size); + + if (is_baad(&ib->rhdr)) + goto dirty_vol; + if (!check_lsn(&ib->rhdr, rlsn)) + goto out; + + if (!check_index_buffer(ib, bytes) || + !check_if_alloc_index(hdr, aoff)) { + goto dirty_vol; + } + + e1 = Add2Ptr(e, esize); + nsize = esize; + used = le32_to_cpu(hdr->used); + + memmove(e, e1, PtrOffset(e1, Add2Ptr(hdr, used))); + + hdr->used = cpu_to_le32(used - nsize); + + a_dirty = true; + + ntfs_fix_pre_write(&ib->rhdr, bytes); + break; + + case WriteEndOfIndexBuffer: + ib = Add2Ptr(buffer_le, roff); + hdr = &ib->ihdr; + e = Add2Ptr(ib, aoff); + + if (is_baad(&ib->rhdr)) + goto dirty_vol; + if (!check_lsn(&ib->rhdr, rlsn)) + goto out; + if (!check_index_buffer(ib, bytes) || + !check_if_alloc_index(hdr, aoff) || + aoff + dlen > offsetof(struct INDEX_BUFFER, ihdr) + + le32_to_cpu(hdr->total)) { + goto dirty_vol; + } + + hdr->used = cpu_to_le32(dlen + PtrOffset(hdr, e)); + memmove(e, data, dlen); + + a_dirty = true; + ntfs_fix_pre_write(&ib->rhdr, bytes); + break; + + case SetIndexEntryVcnAllocation: + ib = Add2Ptr(buffer_le, roff); + hdr = &ib->ihdr; + e = Add2Ptr(ib, aoff); + + if (is_baad(&ib->rhdr)) + goto dirty_vol; + + if (!check_lsn(&ib->rhdr, rlsn)) + goto out; + if (!check_index_buffer(ib, bytes) || + !check_if_alloc_index(hdr, aoff)) { + goto dirty_vol; + } + + de_set_vbn_le(e, *(__le64 *)data); + + a_dirty = true; + ntfs_fix_pre_write(&ib->rhdr, bytes); + break; + + case UpdateFileNameAllocation: + ib = Add2Ptr(buffer_le, roff); + hdr = &ib->ihdr; + e = Add2Ptr(ib, aoff); + + if (is_baad(&ib->rhdr)) + goto dirty_vol; + + if (!check_lsn(&ib->rhdr, rlsn)) + goto out; + if (!check_index_buffer(ib, bytes) || + !check_if_alloc_index(hdr, aoff)) { + goto dirty_vol; + } + + fname = (struct ATTR_FILE_NAME *)(e + 1); + memmove(&fname->dup, data, sizeof(fname->dup)); + + a_dirty = true; + ntfs_fix_pre_write(&ib->rhdr, bytes); + break; + + case SetBitsInNonresidentBitMap: + bmp_off = + le32_to_cpu(((struct BITMAP_RANGE *)data)->bitmap_off); + bmp_bits = le32_to_cpu(((struct BITMAP_RANGE *)data)->bits); + + if (cbo + (bmp_off + 7) / 8 > lco || + cbo + ((bmp_off + bmp_bits + 7) / 8) > lco) { + goto dirty_vol; + } + + __bitmap_set(Add2Ptr(buffer_le, roff), bmp_off, bmp_bits); + a_dirty = true; + break; + + case ClearBitsInNonresidentBitMap: + bmp_off = + le32_to_cpu(((struct BITMAP_RANGE *)data)->bitmap_off); + bmp_bits = le32_to_cpu(((struct BITMAP_RANGE *)data)->bits); + + if (cbo + (bmp_off + 7) / 8 > lco || + cbo + ((bmp_off + bmp_bits + 7) / 8) > lco) { + goto dirty_vol; + } + + __bitmap_clear(Add2Ptr(buffer_le, roff), bmp_off, bmp_bits); + a_dirty = true; + break; + + case UpdateRecordDataAllocation: + ib = Add2Ptr(buffer_le, roff); + hdr = &ib->ihdr; + e = Add2Ptr(ib, aoff); + + if (is_baad(&ib->rhdr)) + goto dirty_vol; + + if (!check_lsn(&ib->rhdr, rlsn)) + goto out; + if (!check_index_buffer(ib, bytes) || + !check_if_alloc_index(hdr, aoff)) { + goto dirty_vol; + } + + memmove(Add2Ptr(e, le16_to_cpu(e->view.data_off)), data, dlen); + + a_dirty = true; + ntfs_fix_pre_write(&ib->rhdr, bytes); + break; + + default: + WARN_ON(1); + } + + if (rlsn) { + __le64 t64 = cpu_to_le64(*rlsn); + + if (rec) + rec->rhdr.lsn = t64; + if (ib) + ib->rhdr.lsn = t64; + } + + if (mi && mi->dirty) { + err = mi_write(mi, 0); + if (err) + goto out; + } + + if (a_dirty) { + attr = oa->attr; + err = ntfs_sb_write_run(sbi, oa->run1, vbo, buffer_le, bytes); + if (err) + goto out; + } + +out: + + if (inode) + iput(inode); + else if (mi != mi2_child) + mi_put(mi); + + ntfs_free(buffer_le); + + return err; + +dirty_vol: + log->set_dirty = true; + goto out; +} + +/* + * log_replay + * + * this function is called during mount operation + * it replays log and empties it + * initialized is set false if logfile contains '-1' + */ +int log_replay(struct ntfs_inode *ni, bool *initialized) +{ + int err; + struct ntfs_sb_info *sbi = ni->mi.sbi; + struct ntfs_log *log; + + struct restart_info rst_info, rst_info2; + u64 rec_lsn, ra_lsn, checkpt_lsn = 0, rlsn = 0; + struct ATTR_NAME_ENTRY *attr_names = NULL; + struct ATTR_NAME_ENTRY *ane; + struct RESTART_TABLE *dptbl = NULL; + struct RESTART_TABLE *trtbl = NULL; + const struct RESTART_TABLE *rt; + struct RESTART_TABLE *oatbl = NULL; + struct inode *inode; + struct OpenAttr *oa; + struct ntfs_inode *ni_oe; + struct ATTRIB *attr = NULL; + u64 size, vcn, undo_next_lsn; + CLST rno, lcn, lcn0, len0, clen; + void *data; + struct NTFS_RESTART *rst = NULL; + struct lcb *lcb = NULL; + struct OPEN_ATTR_ENRTY *oe; + struct TRANSACTION_ENTRY *tr; + struct DIR_PAGE_ENTRY *dp; + u32 i, bytes_per_attr_entry; + u32 l_size = ni->vfs_inode.i_size; + u32 orig_file_size = l_size; + u32 page_size, vbo, tail, off, dlen; + u32 saved_len, rec_len, transact_id; + bool use_second_page; + struct RESTART_AREA *ra2, *ra = NULL; + struct CLIENT_REC *ca, *cr; + __le16 client; + struct RESTART_HDR *rh; + const struct LFS_RECORD_HDR *frh; + const struct LOG_REC_HDR *lrh; + bool is_mapped; + bool is_ro = sb_rdonly(sbi->sb); + u64 t64; + u16 t16; + u32 t32; + + /* Get the size of page. NOTE: To replay we can use default page */ +#if PAGE_SIZE >= DefaultLogPageSize && PAGE_SIZE <= DefaultLogPageSize * 2 + page_size = norm_file_page(PAGE_SIZE, &l_size, true); +#else + page_size = norm_file_page(PAGE_SIZE, &l_size, false); +#endif + if (!page_size) + return -EINVAL; + + log = ntfs_zalloc(sizeof(struct ntfs_log)); + if (!log) + return -ENOMEM; + + log->ni = ni; + log->l_size = l_size; + log->one_page_buf = ntfs_malloc(page_size); + + if (!log->one_page_buf) { + err = -ENOMEM; + goto out; + } + + log->page_size = page_size; + log->page_mask = page_size - 1; + log->page_bits = blksize_bits(page_size); + + /* Look for a restart area on the disk */ + err = log_read_rst(log, l_size, true, &rst_info); + if (err) + goto out; + + /* remember 'initialized' */ + *initialized = rst_info.initialized; + + if (!rst_info.restart) { + if (rst_info.initialized) { + /* no restart area but the file is not initialized */ + err = -EINVAL; + goto out; + } + + log_init_pg_hdr(log, page_size, page_size, 1, 1); + log_create(log, l_size, 0, get_random_int(), false, false); + + log->ra = ra; + + ra = log_create_ra(log); + if (!ra) { + err = -ENOMEM; + goto out; + } + log->ra = ra; + log->init_ra = true; + + goto process_log; + } + + /* + * If the restart offset above wasn't zero then we won't + * look for a second restart + */ + if (rst_info.vbo) + goto check_restart_area; + + err = log_read_rst(log, l_size, false, &rst_info2); + + /* Determine which restart area to use */ + if (!rst_info2.restart || rst_info2.last_lsn <= rst_info.last_lsn) + goto use_first_page; + + use_second_page = true; + + if (rst_info.chkdsk_was_run && page_size != rst_info.vbo) { + struct RECORD_PAGE_HDR *sp = NULL; + bool usa_error; + + if (!read_log_page(log, page_size, &sp, &usa_error) && + sp->rhdr.sign == NTFS_CHKD_SIGNATURE) { + use_second_page = false; + } + ntfs_free(sp); + } + + if (use_second_page) { + ntfs_free(rst_info.r_page); + memcpy(&rst_info, &rst_info2, sizeof(struct restart_info)); + rst_info2.r_page = NULL; + } + +use_first_page: + ntfs_free(rst_info2.r_page); + +check_restart_area: + /* If the restart area is at offset 0, we want to write the second restart area first */ + log->init_ra = !!rst_info.vbo; + + /* If we have a valid page then grab a pointer to the restart area */ + ra2 = rst_info.valid_page + ? Add2Ptr(rst_info.r_page, + le16_to_cpu(rst_info.r_page->ra_off)) + : NULL; + + if (rst_info.chkdsk_was_run || + (ra2 && ra2->client_idx[1] == LFS_NO_CLIENT_LE)) { + bool wrapped = false; + bool use_multi_page = false; + u32 open_log_count; + + /* Do some checks based on whether we have a valid log page */ + if (!rst_info.valid_page) { + open_log_count = get_random_int(); + goto init_log_instance; + } + open_log_count = le32_to_cpu(ra2->open_log_count); + + /* + * If the restart page size isn't changing then we want to + * check how much work we need to do + */ + if (page_size != le32_to_cpu(rst_info.r_page->sys_page_size)) + goto init_log_instance; + +init_log_instance: + log_init_pg_hdr(log, page_size, page_size, 1, 1); + + log_create(log, l_size, rst_info.last_lsn, open_log_count, + wrapped, use_multi_page); + + ra = log_create_ra(log); + if (!ra) { + err = -ENOMEM; + goto out; + } + log->ra = ra; + + /* Put the restart areas and initialize the log file as required */ + goto process_log; + } + + if (!ra2) { + err = -EINVAL; + goto out; + } + + /* + * If the log page or the system page sizes have changed, we can't use the log file + * We must use the system page size instead of the default size + * if there is not a clean shutdown + */ + t32 = le32_to_cpu(rst_info.r_page->sys_page_size); + if (page_size != t32) { + l_size = orig_file_size; + page_size = + norm_file_page(t32, &l_size, t32 == DefaultLogPageSize); + } + + if (page_size != t32 || + page_size != le32_to_cpu(rst_info.r_page->page_size)) { + err = -EINVAL; + goto out; + } + + /* If the file size has shrunk then we won't mount it */ + if (l_size < le64_to_cpu(ra2->l_size)) { + err = -EINVAL; + goto out; + } + + log_init_pg_hdr(log, page_size, page_size, + le16_to_cpu(rst_info.r_page->major_ver), + le16_to_cpu(rst_info.r_page->minor_ver)); + + log->l_size = le64_to_cpu(ra2->l_size); + log->seq_num_bits = le32_to_cpu(ra2->seq_num_bits); + log->file_data_bits = sizeof(u64) * 8 - log->seq_num_bits; + log->seq_num_mask = (8 << log->file_data_bits) - 1; + log->last_lsn = le64_to_cpu(ra2->current_lsn); + log->seq_num = log->last_lsn >> log->file_data_bits; + log->ra_off = le16_to_cpu(rst_info.r_page->ra_off); + log->restart_size = log->sys_page_size - log->ra_off; + log->record_header_len = le16_to_cpu(ra2->rec_hdr_len); + log->ra_size = le16_to_cpu(ra2->ra_len); + log->data_off = le16_to_cpu(ra2->data_off); + log->data_size = log->page_size - log->data_off; + log->reserved = log->data_size - log->record_header_len; + + vbo = lsn_to_vbo(log, log->last_lsn); + + if (vbo < log->first_page) { + /* This is a pseudo lsn */ + log->l_flags |= NTFSLOG_NO_LAST_LSN; + log->next_page = log->first_page; + goto find_oldest; + } + + /* Find the end of this log record */ + off = final_log_off(log, log->last_lsn, + le32_to_cpu(ra2->last_lsn_data_len)); + + /* If we wrapped the file then increment the sequence number */ + if (off <= vbo) { + log->seq_num += 1; + log->l_flags |= NTFSLOG_WRAPPED; + } + + /* Now compute the next log page to use */ + vbo &= ~log->sys_page_mask; + tail = log->page_size - (off & log->page_mask) - 1; + + /* If we can fit another log record on the page, move back a page the log file */ + if (tail >= log->record_header_len) { + log->l_flags |= NTFSLOG_REUSE_TAIL; + log->next_page = vbo; + } else { + log->next_page = next_page_off(log, vbo); + } + +find_oldest: + /* Find the oldest client lsn. Use the last flushed lsn as a starting point */ + log->oldest_lsn = log->last_lsn; + oldest_client_lsn(Add2Ptr(ra2, le16_to_cpu(ra2->client_off)), + ra2->client_idx[1], &log->oldest_lsn); + log->oldest_lsn_off = lsn_to_vbo(log, log->oldest_lsn); + + if (log->oldest_lsn_off < log->first_page) + log->l_flags |= NTFSLOG_NO_OLDEST_LSN; + + if (!(ra2->flags & RESTART_SINGLE_PAGE_IO)) + log->l_flags |= NTFSLOG_WRAPPED | NTFSLOG_MULTIPLE_PAGE_IO; + + log->current_openlog_count = le32_to_cpu(ra2->open_log_count); + log->total_avail_pages = log->l_size - log->first_page; + log->total_avail = log->total_avail_pages >> log->page_bits; + log->max_current_avail = log->total_avail * log->reserved; + log->total_avail = log->total_avail * log->data_size; + + log->current_avail = current_log_avail(log); + + ra = ntfs_zalloc(log->restart_size); + if (!ra) { + err = -ENOMEM; + goto out; + } + log->ra = ra; + + t16 = le16_to_cpu(ra2->client_off); + if (t16 == offsetof(struct RESTART_AREA, clients)) { + memcpy(ra, ra2, log->ra_size); + } else { + memcpy(ra, ra2, offsetof(struct RESTART_AREA, clients)); + memcpy(ra->clients, Add2Ptr(ra2, t16), + le16_to_cpu(ra2->ra_len) - t16); + + log->current_openlog_count = get_random_int(); + ra->open_log_count = cpu_to_le32(log->current_openlog_count); + log->ra_size = offsetof(struct RESTART_AREA, clients) + + sizeof(struct CLIENT_REC); + ra->client_off = + cpu_to_le16(offsetof(struct RESTART_AREA, clients)); + ra->ra_len = cpu_to_le16(log->ra_size); + } + + le32_add_cpu(&ra->open_log_count, 1); + + /* Now we need to walk through looking for the last lsn */ + err = last_log_lsn(log); + if (err) + goto out; + + log->current_avail = current_log_avail(log); + + /* Remember which restart area to write first */ + log->init_ra = rst_info.vbo; + +process_log: + /* 1.0, 1.1, 2.0 log->major_ver/minor_ver - short values */ + switch ((log->major_ver << 16) + log->minor_ver) { + case 0x10000: + case 0x10001: + case 0x20000: + break; + default: + ntfs_warn(sbi->sb, "\x24LogFile version %d.%d is not supported", + log->major_ver, log->minor_ver); + err = -EOPNOTSUPP; + log->set_dirty = true; + goto out; + } + + /* One client "NTFS" per logfile */ + ca = Add2Ptr(ra, le16_to_cpu(ra->client_off)); + + for (client = ra->client_idx[1];; client = cr->next_client) { + if (client == LFS_NO_CLIENT_LE) { + /* Insert "NTFS" client LogFile */ + client = ra->client_idx[0]; + if (client == LFS_NO_CLIENT_LE) + return -EINVAL; + + t16 = le16_to_cpu(client); + cr = ca + t16; + + remove_client(ca, cr, &ra->client_idx[0]); + + cr->restart_lsn = 0; + cr->oldest_lsn = cpu_to_le64(log->oldest_lsn); + cr->name_bytes = cpu_to_le32(8); + cr->name[0] = cpu_to_le16('N'); + cr->name[1] = cpu_to_le16('T'); + cr->name[2] = cpu_to_le16('F'); + cr->name[3] = cpu_to_le16('S'); + + add_client(ca, t16, &ra->client_idx[1]); + break; + } + + cr = ca + le16_to_cpu(client); + + if (cpu_to_le32(8) == cr->name_bytes && + cpu_to_le16('N') == cr->name[0] && + cpu_to_le16('T') == cr->name[1] && + cpu_to_le16('F') == cr->name[2] && + cpu_to_le16('S') == cr->name[3]) + break; + } + + /* Update the client handle with the client block information */ + log->client_id.seq_num = cr->seq_num; + log->client_id.client_idx = client; + + err = read_rst_area(log, &rst, &ra_lsn); + if (err) + goto out; + + if (!rst) + goto out; + + bytes_per_attr_entry = !rst->major_ver ? 0x2C : 0x28; + + checkpt_lsn = le64_to_cpu(rst->check_point_start); + if (!checkpt_lsn) + checkpt_lsn = ra_lsn; + + /* Allocate and Read the Transaction Table */ + if (!rst->transact_table_len) + goto check_dirty_page_table; + + t64 = le64_to_cpu(rst->transact_table_lsn); + err = read_log_rec_lcb(log, t64, lcb_ctx_prev, &lcb); + if (err) + goto out; + + lrh = lcb->log_rec; + frh = lcb->lrh; + rec_len = le32_to_cpu(frh->client_data_len); + + if (!check_log_rec(lrh, rec_len, le32_to_cpu(frh->transact_id), + bytes_per_attr_entry)) { + err = -EINVAL; + goto out; + } + + t16 = le16_to_cpu(lrh->redo_off); + + rt = Add2Ptr(lrh, t16); + t32 = rec_len - t16; + + /* Now check that this is a valid restart table */ + if (!check_rstbl(rt, t32)) { + err = -EINVAL; + goto out; + } + + trtbl = ntfs_memdup(rt, t32); + if (!trtbl) { + err = -ENOMEM; + goto out; + } + + lcb_put(lcb); + lcb = NULL; + +check_dirty_page_table: + /* The next record back should be the Dirty Pages Table */ + if (!rst->dirty_pages_len) + goto check_attribute_names; + + t64 = le64_to_cpu(rst->dirty_pages_table_lsn); + err = read_log_rec_lcb(log, t64, lcb_ctx_prev, &lcb); + if (err) + goto out; + + lrh = lcb->log_rec; + frh = lcb->lrh; + rec_len = le32_to_cpu(frh->client_data_len); + + if (!check_log_rec(lrh, rec_len, le32_to_cpu(frh->transact_id), + bytes_per_attr_entry)) { + err = -EINVAL; + goto out; + } + + t16 = le16_to_cpu(lrh->redo_off); + + rt = Add2Ptr(lrh, t16); + t32 = rec_len - t16; + + /* Now check that this is a valid restart table */ + if (!check_rstbl(rt, t32)) { + err = -EINVAL; + goto out; + } + + dptbl = ntfs_memdup(rt, t32); + if (!dptbl) { + err = -ENOMEM; + goto out; + } + + /* Convert Ra version '0' into version '1' */ + if (rst->major_ver) + goto end_conv_1; + + dp = NULL; + while ((dp = enum_rstbl(dptbl, dp))) { + struct DIR_PAGE_ENTRY_32 *dp0 = (struct DIR_PAGE_ENTRY_32 *)dp; + // NOTE: Danger. Check for of boundary + memmove(&dp->vcn, &dp0->vcn_low, + 2 * sizeof(u64) + + le32_to_cpu(dp->lcns_follow) * sizeof(u64)); + } + +end_conv_1: + lcb_put(lcb); + lcb = NULL; + + /* Go through the table and remove the duplicates, remembering the oldest lsn values */ + if (sbi->cluster_size <= log->page_size) + goto trace_dp_table; + + dp = NULL; + while ((dp = enum_rstbl(dptbl, dp))) { + struct DIR_PAGE_ENTRY *next = dp; + + while ((next = enum_rstbl(dptbl, next))) { + if (next->target_attr == dp->target_attr && + next->vcn == dp->vcn) { + if (le64_to_cpu(next->oldest_lsn) < + le64_to_cpu(dp->oldest_lsn)) { + dp->oldest_lsn = next->oldest_lsn; + } + + free_rsttbl_idx(dptbl, PtrOffset(dptbl, next)); + } + } + } +trace_dp_table: +check_attribute_names: + /* The next record should be the Attribute Names */ + if (!rst->attr_names_len) + goto check_attr_table; + + t64 = le64_to_cpu(rst->attr_names_lsn); + err = read_log_rec_lcb(log, t64, lcb_ctx_prev, &lcb); + if (err) + goto out; + + lrh = lcb->log_rec; + frh = lcb->lrh; + rec_len = le32_to_cpu(frh->client_data_len); + + if (!check_log_rec(lrh, rec_len, le32_to_cpu(frh->transact_id), + bytes_per_attr_entry)) { + err = -EINVAL; + goto out; + } + + t32 = lrh_length(lrh); + rec_len -= t32; + + attr_names = ntfs_memdup(Add2Ptr(lrh, t32), rec_len); + + lcb_put(lcb); + lcb = NULL; + +check_attr_table: + /* The next record should be the attribute Table */ + if (!rst->open_attr_len) + goto check_attribute_names2; + + t64 = le64_to_cpu(rst->open_attr_table_lsn); + err = read_log_rec_lcb(log, t64, lcb_ctx_prev, &lcb); + if (err) + goto out; + + lrh = lcb->log_rec; + frh = lcb->lrh; + rec_len = le32_to_cpu(frh->client_data_len); + + if (!check_log_rec(lrh, rec_len, le32_to_cpu(frh->transact_id), + bytes_per_attr_entry)) { + err = -EINVAL; + goto out; + } + + t16 = le16_to_cpu(lrh->redo_off); + + rt = Add2Ptr(lrh, t16); + t32 = rec_len - t16; + + if (!check_rstbl(rt, t32)) { + err = -EINVAL; + goto out; + } + + oatbl = ntfs_memdup(rt, t32); + if (!oatbl) { + err = -ENOMEM; + goto out; + } + + log->open_attr_tbl = oatbl; + + /* Clear all of the Attr pointers */ + oe = NULL; + while ((oe = enum_rstbl(oatbl, oe))) { + if (!rst->major_ver) { + struct OPEN_ATTR_ENRTY_32 oe0; + + /* Really 'oe' points to OPEN_ATTR_ENRTY_32 */ + memcpy(&oe0, oe, SIZEOF_OPENATTRIBUTEENTRY0); + + oe->bytes_per_index = oe0.bytes_per_index; + oe->type = oe0.type; + oe->is_dirty_pages = oe0.is_dirty_pages; + oe->name_len = 0; + oe->ref = oe0.ref; + oe->open_record_lsn = oe0.open_record_lsn; + } + + oe->is_attr_name = 0; + oe->ptr = NULL; + } + + lcb_put(lcb); + lcb = NULL; + +check_attribute_names2: + if (!rst->attr_names_len) + goto trace_attribute_table; + + ane = attr_names; + if (!oatbl) + goto trace_attribute_table; + while (ane->off) { + /* TODO: Clear table on exit! */ + oe = Add2Ptr(oatbl, le16_to_cpu(ane->off)); + t16 = le16_to_cpu(ane->name_bytes); + oe->name_len = t16 / sizeof(short); + oe->ptr = ane->name; + oe->is_attr_name = 2; + ane = Add2Ptr(ane, sizeof(struct ATTR_NAME_ENTRY) + t16); + } + +trace_attribute_table: + /* + * If the checkpt_lsn is zero, then this is a freshly + * formatted disk and we have no work to do + */ + if (!checkpt_lsn) { + err = 0; + goto out; + } + + if (!oatbl) { + oatbl = init_rsttbl(bytes_per_attr_entry, 8); + if (!oatbl) { + err = -ENOMEM; + goto out; + } + } + + log->open_attr_tbl = oatbl; + + /* Start the analysis pass from the Checkpoint lsn. */ + rec_lsn = checkpt_lsn; + + /* Read the first lsn */ + err = read_log_rec_lcb(log, checkpt_lsn, lcb_ctx_next, &lcb); + if (err) + goto out; + + /* Loop to read all subsequent records to the end of the log file */ +next_log_record_analyze: + err = read_next_log_rec(log, lcb, &rec_lsn); + if (err) + goto out; + + if (!rec_lsn) + goto end_log_records_enumerate; + + frh = lcb->lrh; + transact_id = le32_to_cpu(frh->transact_id); + rec_len = le32_to_cpu(frh->client_data_len); + lrh = lcb->log_rec; + + if (!check_log_rec(lrh, rec_len, transact_id, bytes_per_attr_entry)) { + err = -EINVAL; + goto out; + } + + /* + * The first lsn after the previous lsn remembered + * the checkpoint is the first candidate for the rlsn + */ + if (!rlsn) + rlsn = rec_lsn; + + if (LfsClientRecord != frh->record_type) + goto next_log_record_analyze; + + /* + * Now update the Transaction Table for this transaction + * If there is no entry present or it is unallocated we allocate the entry + */ + if (!trtbl) { + trtbl = init_rsttbl(sizeof(struct TRANSACTION_ENTRY), + INITIAL_NUMBER_TRANSACTIONS); + if (!trtbl) { + err = -ENOMEM; + goto out; + } + } + + tr = Add2Ptr(trtbl, transact_id); + + if (transact_id >= bytes_per_rt(trtbl) || + tr->next != RESTART_ENTRY_ALLOCATED_LE) { + tr = alloc_rsttbl_from_idx(&trtbl, transact_id); + if (!tr) { + err = -ENOMEM; + goto out; + } + tr->transact_state = TransactionActive; + tr->first_lsn = cpu_to_le64(rec_lsn); + } + + tr->prev_lsn = tr->undo_next_lsn = cpu_to_le64(rec_lsn); + + /* + * If this is a compensation log record, then change + * the undo_next_lsn to be the undo_next_lsn of this record + */ + if (lrh->undo_op == cpu_to_le16(CompensationLogRecord)) + tr->undo_next_lsn = frh->client_undo_next_lsn; + + /* Dispatch to handle log record depending on type */ + switch (le16_to_cpu(lrh->redo_op)) { + case InitializeFileRecordSegment: + case DeallocateFileRecordSegment: + case WriteEndOfFileRecordSegment: + case CreateAttribute: + case DeleteAttribute: + case UpdateResidentValue: + case UpdateNonresidentValue: + case UpdateMappingPairs: + case SetNewAttributeSizes: + case AddIndexEntryRoot: + case DeleteIndexEntryRoot: + case AddIndexEntryAllocation: + case DeleteIndexEntryAllocation: + case WriteEndOfIndexBuffer: + case SetIndexEntryVcnRoot: + case SetIndexEntryVcnAllocation: + case UpdateFileNameRoot: + case UpdateFileNameAllocation: + case SetBitsInNonresidentBitMap: + case ClearBitsInNonresidentBitMap: + case UpdateRecordDataRoot: + case UpdateRecordDataAllocation: + case ZeroEndOfFileRecord: + t16 = le16_to_cpu(lrh->target_attr); + t64 = le64_to_cpu(lrh->target_vcn); + dp = find_dp(dptbl, t16, t64); + + if (dp) + goto copy_lcns; + + /* + * Calculate the number of clusters per page the system + * which wrote the checkpoint, possibly creating the table + */ + if (dptbl) { + t32 = (le16_to_cpu(dptbl->size) - + sizeof(struct DIR_PAGE_ENTRY)) / + sizeof(u64); + } else { + t32 = log->clst_per_page; + ntfs_free(dptbl); + dptbl = init_rsttbl(struct_size(dp, page_lcns, t32), + 32); + if (!dptbl) { + err = -ENOMEM; + goto out; + } + } + + dp = alloc_rsttbl_idx(&dptbl); + dp->target_attr = cpu_to_le32(t16); + dp->transfer_len = cpu_to_le32(t32 << sbi->cluster_bits); + dp->lcns_follow = cpu_to_le32(t32); + dp->vcn = cpu_to_le64(t64 & ~((u64)t32 - 1)); + dp->oldest_lsn = cpu_to_le64(rec_lsn); + +copy_lcns: + /* + * Copy the Lcns from the log record into the Dirty Page Entry + * TODO: for different page size support, must somehow make + * whole routine a loop, case Lcns do not fit below + */ + t16 = le16_to_cpu(lrh->lcns_follow); + for (i = 0; i < t16; i++) { + size_t j = (size_t)(le64_to_cpu(lrh->target_vcn) - + le64_to_cpu(dp->vcn)); + dp->page_lcns[j + i] = lrh->page_lcns[i]; + } + + goto next_log_record_analyze; + + case DeleteDirtyClusters: { + u32 range_count = + le16_to_cpu(lrh->redo_len) / sizeof(struct LCN_RANGE); + const struct LCN_RANGE *r = + Add2Ptr(lrh, le16_to_cpu(lrh->redo_off)); + + /* Loop through all of the Lcn ranges this log record */ + for (i = 0; i < range_count; i++, r++) { + u64 lcn0 = le64_to_cpu(r->lcn); + u64 lcn_e = lcn0 + le64_to_cpu(r->len) - 1; + + dp = NULL; + while ((dp = enum_rstbl(dptbl, dp))) { + u32 j; + + t32 = le32_to_cpu(dp->lcns_follow); + for (j = 0; j < t32; j++) { + t64 = le64_to_cpu(dp->page_lcns[j]); + if (t64 >= lcn0 && t64 <= lcn_e) + dp->page_lcns[j] = 0; + } + } + } + goto next_log_record_analyze; + ; + } + + case OpenNonresidentAttribute: + t16 = le16_to_cpu(lrh->target_attr); + if (t16 >= bytes_per_rt(oatbl)) { + /* + * Compute how big the table needs to be. + * Add 10 extra entries for some cushion + */ + u32 new_e = t16 / le16_to_cpu(oatbl->size); + + new_e += 10 - le16_to_cpu(oatbl->used); + + oatbl = extend_rsttbl(oatbl, new_e, ~0u); + log->open_attr_tbl = oatbl; + if (!oatbl) { + err = -ENOMEM; + goto out; + } + } + + /* Point to the entry being opened */ + oe = alloc_rsttbl_from_idx(&oatbl, t16); + log->open_attr_tbl = oatbl; + if (!oe) { + err = -ENOMEM; + goto out; + } + + /* Initialize this entry from the log record */ + t16 = le16_to_cpu(lrh->redo_off); + if (!rst->major_ver) { + /* Convert version '0' into version '1' */ + struct OPEN_ATTR_ENRTY_32 *oe0 = Add2Ptr(lrh, t16); + + oe->bytes_per_index = oe0->bytes_per_index; + oe->type = oe0->type; + oe->is_dirty_pages = oe0->is_dirty_pages; + oe->name_len = 0; //oe0.name_len; + oe->ref = oe0->ref; + oe->open_record_lsn = oe0->open_record_lsn; + } else { + memcpy(oe, Add2Ptr(lrh, t16), bytes_per_attr_entry); + } + + t16 = le16_to_cpu(lrh->undo_len); + if (t16) { + oe->ptr = ntfs_malloc(t16); + if (!oe->ptr) { + err = -ENOMEM; + goto out; + } + oe->name_len = t16 / sizeof(short); + memcpy(oe->ptr, + Add2Ptr(lrh, le16_to_cpu(lrh->undo_off)), t16); + oe->is_attr_name = 1; + } else { + oe->ptr = NULL; + oe->is_attr_name = 0; + } + + goto next_log_record_analyze; + + case HotFix: + t16 = le16_to_cpu(lrh->target_attr); + t64 = le64_to_cpu(lrh->target_vcn); + dp = find_dp(dptbl, t16, t64); + if (dp) { + size_t j = le64_to_cpu(lrh->target_vcn) - + le64_to_cpu(dp->vcn); + if (dp->page_lcns[j]) + dp->page_lcns[j] = lrh->page_lcns[0]; + } + goto next_log_record_analyze; + + case EndTopLevelAction: + tr = Add2Ptr(trtbl, transact_id); + tr->prev_lsn = cpu_to_le64(rec_lsn); + tr->undo_next_lsn = frh->client_undo_next_lsn; + goto next_log_record_analyze; + + case PrepareTransaction: + tr = Add2Ptr(trtbl, transact_id); + tr->transact_state = TransactionPrepared; + goto next_log_record_analyze; + + case CommitTransaction: + tr = Add2Ptr(trtbl, transact_id); + tr->transact_state = TransactionCommitted; + goto next_log_record_analyze; + + case ForgetTransaction: + free_rsttbl_idx(trtbl, transact_id); + goto next_log_record_analyze; + + case Noop: + case OpenAttributeTableDump: + case AttributeNamesDump: + case DirtyPageTableDump: + case TransactionTableDump: + /* The following cases require no action the Analysis Pass */ + goto next_log_record_analyze; + + default: + /* + * All codes will be explicitly handled. + * If we see a code we do not expect, then we are trouble + */ + goto next_log_record_analyze; + } + +end_log_records_enumerate: + lcb_put(lcb); + lcb = NULL; + + /* + * Scan the Dirty Page Table and Transaction Table for + * the lowest lsn, and return it as the Redo lsn + */ + dp = NULL; + while ((dp = enum_rstbl(dptbl, dp))) { + t64 = le64_to_cpu(dp->oldest_lsn); + if (t64 && t64 < rlsn) + rlsn = t64; + } + + tr = NULL; + while ((tr = enum_rstbl(trtbl, tr))) { + t64 = le64_to_cpu(tr->first_lsn); + if (t64 && t64 < rlsn) + rlsn = t64; + } + + /* Only proceed if the Dirty Page Table or Transaction table are not empty */ + if ((!dptbl || !dptbl->total) && (!trtbl || !trtbl->total)) + goto end_reply; + + sbi->flags |= NTFS_FLAGS_NEED_REPLAY; + if (is_ro) + goto out; + + /* Reopen all of the attributes with dirty pages */ + oe = NULL; +next_open_attribute: + + oe = enum_rstbl(oatbl, oe); + if (!oe) { + err = 0; + dp = NULL; + goto next_dirty_page; + } + + oa = ntfs_zalloc(sizeof(struct OpenAttr)); + if (!oa) { + err = -ENOMEM; + goto out; + } + + inode = ntfs_iget5(sbi->sb, &oe->ref, NULL); + if (IS_ERR(inode)) + goto fake_attr; + + if (is_bad_inode(inode)) { + iput(inode); +fake_attr: + if (oa->ni) { + iput(&oa->ni->vfs_inode); + oa->ni = NULL; + } + + attr = attr_create_nonres_log(sbi, oe->type, 0, oe->ptr, + oe->name_len, 0); + if (!attr) { + ntfs_free(oa); + err = -ENOMEM; + goto out; + } + oa->attr = attr; + oa->run1 = &oa->run0; + goto final_oe; + } + + ni_oe = ntfs_i(inode); + oa->ni = ni_oe; + + attr = ni_find_attr(ni_oe, NULL, NULL, oe->type, oe->ptr, oe->name_len, + NULL, NULL); + + if (!attr) + goto fake_attr; + + t32 = le32_to_cpu(attr->size); + oa->attr = ntfs_memdup(attr, t32); + if (!oa->attr) + goto fake_attr; + + if (!S_ISDIR(inode->i_mode)) { + if (attr->type == ATTR_DATA && !attr->name_len) { + oa->run1 = &ni_oe->file.run; + goto final_oe; + } + } else { + if (attr->type == ATTR_ALLOC && + attr->name_len == ARRAY_SIZE(I30_NAME) && + !memcmp(attr_name(attr), I30_NAME, sizeof(I30_NAME))) { + oa->run1 = &ni_oe->dir.alloc_run; + goto final_oe; + } + } + + if (attr->non_res) { + u16 roff = le16_to_cpu(attr->nres.run_off); + CLST svcn = le64_to_cpu(attr->nres.svcn); + + err = run_unpack(&oa->run0, sbi, inode->i_ino, svcn, + le64_to_cpu(attr->nres.evcn), svcn, + Add2Ptr(attr, roff), t32 - roff); + if (err < 0) { + ntfs_free(oa->attr); + oa->attr = NULL; + goto fake_attr; + } + err = 0; + } + oa->run1 = &oa->run0; + attr = oa->attr; + +final_oe: + if (oe->is_attr_name == 1) + ntfs_free(oe->ptr); + oe->is_attr_name = 0; + oe->ptr = oa; + oe->name_len = attr->name_len; + + goto next_open_attribute; + + /* + * Now loop through the dirty page table to extract all of the Vcn/Lcn + * Mapping that we have, and insert it into the appropriate run + */ +next_dirty_page: + dp = enum_rstbl(dptbl, dp); + if (!dp) + goto do_redo_1; + + oe = Add2Ptr(oatbl, le32_to_cpu(dp->target_attr)); + + if (oe->next != RESTART_ENTRY_ALLOCATED_LE) + goto next_dirty_page; + + oa = oe->ptr; + if (!oa) + goto next_dirty_page; + + i = -1; +next_dirty_page_vcn: + i += 1; + if (i >= le32_to_cpu(dp->lcns_follow)) + goto next_dirty_page; + + vcn = le64_to_cpu(dp->vcn) + i; + size = (vcn + 1) << sbi->cluster_bits; + + if (!dp->page_lcns[i]) + goto next_dirty_page_vcn; + + rno = ino_get(&oe->ref); + if (rno <= MFT_REC_MIRR && + size < (MFT_REC_VOL + 1) * sbi->record_size && + oe->type == ATTR_DATA) { + goto next_dirty_page_vcn; + } + + lcn = le64_to_cpu(dp->page_lcns[i]); + + if ((!run_lookup_entry(oa->run1, vcn, &lcn0, &len0, NULL) || + lcn0 != lcn) && + !run_add_entry(oa->run1, vcn, lcn, 1, false)) { + err = -ENOMEM; + goto out; + } + attr = oa->attr; + t64 = le64_to_cpu(attr->nres.alloc_size); + if (size > t64) { + attr->nres.valid_size = attr->nres.data_size = + attr->nres.alloc_size = cpu_to_le64(size); + } + goto next_dirty_page_vcn; + +do_redo_1: + /* + * Perform the Redo Pass, to restore all of the dirty pages to the same + * contents that they had immediately before the crash + * If the dirty page table is empty, then we can skip the entire Redo Pass + */ + if (!dptbl || !dptbl->total) + goto do_undo_action; + + rec_lsn = rlsn; + + /* + * Read the record at the Redo lsn, before falling + * into common code to handle each record + */ + err = read_log_rec_lcb(log, rlsn, lcb_ctx_next, &lcb); + if (err) + goto out; + + /* + * Now loop to read all of our log records forwards, + * until we hit the end of the file, cleaning up at the end + */ +do_action_next: + frh = lcb->lrh; + + if (LfsClientRecord != frh->record_type) + goto read_next_log_do_action; + + transact_id = le32_to_cpu(frh->transact_id); + rec_len = le32_to_cpu(frh->client_data_len); + lrh = lcb->log_rec; + + if (!check_log_rec(lrh, rec_len, transact_id, bytes_per_attr_entry)) { + err = -EINVAL; + goto out; + } + + /* Ignore log records that do not update pages */ + if (lrh->lcns_follow) + goto find_dirty_page; + + goto read_next_log_do_action; + +find_dirty_page: + t16 = le16_to_cpu(lrh->target_attr); + t64 = le64_to_cpu(lrh->target_vcn); + dp = find_dp(dptbl, t16, t64); + + if (!dp) + goto read_next_log_do_action; + + if (rec_lsn < le64_to_cpu(dp->oldest_lsn)) + goto read_next_log_do_action; + + t16 = le16_to_cpu(lrh->target_attr); + if (t16 >= bytes_per_rt(oatbl)) { + err = -EINVAL; + goto out; + } + + oe = Add2Ptr(oatbl, t16); + + if (oe->next != RESTART_ENTRY_ALLOCATED_LE) { + err = -EINVAL; + goto out; + } + + oa = oe->ptr; + + if (!oa) { + err = -EINVAL; + goto out; + } + attr = oa->attr; + + vcn = le64_to_cpu(lrh->target_vcn); + + if (!run_lookup_entry(oa->run1, vcn, &lcn, NULL, NULL) || + lcn == SPARSE_LCN) { + goto read_next_log_do_action; + } + + /* Point to the Redo data and get its length */ + data = Add2Ptr(lrh, le16_to_cpu(lrh->redo_off)); + dlen = le16_to_cpu(lrh->redo_len); + + /* Shorten length by any Lcns which were deleted */ + saved_len = dlen; + + for (i = le16_to_cpu(lrh->lcns_follow); i; i--) { + size_t j; + u32 alen, voff; + + voff = le16_to_cpu(lrh->record_off) + + le16_to_cpu(lrh->attr_off); + voff += le16_to_cpu(lrh->cluster_off) << SECTOR_SHIFT; + + /* If the Vcn question is allocated, we can just get out.*/ + j = le64_to_cpu(lrh->target_vcn) - le64_to_cpu(dp->vcn); + if (dp->page_lcns[j + i - 1]) + break; + + if (!saved_len) + saved_len = 1; + + /* + * Calculate the allocated space left relative to the + * log record Vcn, after removing this unallocated Vcn + */ + alen = (i - 1) << sbi->cluster_bits; + + /* + * If the update described this log record goes beyond + * the allocated space, then we will have to reduce the length + */ + if (voff >= alen) + dlen = 0; + else if (voff + dlen > alen) + dlen = alen - voff; + } + + /* If the resulting dlen from above is now zero, we can skip this log record */ + if (!dlen && saved_len) + goto read_next_log_do_action; + + t16 = le16_to_cpu(lrh->redo_op); + if (can_skip_action(t16)) + goto read_next_log_do_action; + + /* Apply the Redo operation a common routine */ + err = do_action(log, oe, lrh, t16, data, dlen, rec_len, &rec_lsn); + if (err) + goto out; + + /* Keep reading and looping back until end of file */ +read_next_log_do_action: + err = read_next_log_rec(log, lcb, &rec_lsn); + if (!err && rec_lsn) + goto do_action_next; + + lcb_put(lcb); + lcb = NULL; + +do_undo_action: + /* Scan Transaction Table */ + tr = NULL; +transaction_table_next: + tr = enum_rstbl(trtbl, tr); + if (!tr) + goto undo_action_done; + + if (TransactionActive != tr->transact_state || !tr->undo_next_lsn) { + free_rsttbl_idx(trtbl, PtrOffset(trtbl, tr)); + goto transaction_table_next; + } + + log->transaction_id = PtrOffset(trtbl, tr); + undo_next_lsn = le64_to_cpu(tr->undo_next_lsn); + + /* + * We only have to do anything if the transaction has + * something its undo_next_lsn field + */ + if (!undo_next_lsn) + goto commit_undo; + + /* Read the first record to be undone by this transaction */ + err = read_log_rec_lcb(log, undo_next_lsn, lcb_ctx_undo_next, &lcb); + if (err) + goto out; + + /* + * Now loop to read all of our log records forwards, + * until we hit the end of the file, cleaning up at the end + */ +undo_action_next: + + lrh = lcb->log_rec; + frh = lcb->lrh; + transact_id = le32_to_cpu(frh->transact_id); + rec_len = le32_to_cpu(frh->client_data_len); + + if (!check_log_rec(lrh, rec_len, transact_id, bytes_per_attr_entry)) { + err = -EINVAL; + goto out; + } + + if (lrh->undo_op == cpu_to_le16(Noop)) + goto read_next_log_undo_action; + + oe = Add2Ptr(oatbl, le16_to_cpu(lrh->target_attr)); + oa = oe->ptr; + + t16 = le16_to_cpu(lrh->lcns_follow); + if (!t16) + goto add_allocated_vcns; + + is_mapped = run_lookup_entry(oa->run1, le64_to_cpu(lrh->target_vcn), + &lcn, &clen, NULL); + + /* + * If the mapping isn't already the table or the mapping + * corresponds to a hole the mapping, we need to make sure + * there is no partial page already memory + */ + if (is_mapped && lcn != SPARSE_LCN && clen >= t16) + goto add_allocated_vcns; + + vcn = le64_to_cpu(lrh->target_vcn); + vcn &= ~(log->clst_per_page - 1); + +add_allocated_vcns: + for (i = 0, vcn = le64_to_cpu(lrh->target_vcn), + size = (vcn + 1) << sbi->cluster_bits; + i < t16; i++, vcn += 1, size += sbi->cluster_size) { + attr = oa->attr; + if (!attr->non_res) { + if (size > le32_to_cpu(attr->res.data_size)) + attr->res.data_size = cpu_to_le32(size); + } else { + if (size > le64_to_cpu(attr->nres.data_size)) + attr->nres.valid_size = attr->nres.data_size = + attr->nres.alloc_size = + cpu_to_le64(size); + } + } + + t16 = le16_to_cpu(lrh->undo_op); + if (can_skip_action(t16)) + goto read_next_log_undo_action; + + /* Point to the Redo data and get its length */ + data = Add2Ptr(lrh, le16_to_cpu(lrh->undo_off)); + dlen = le16_to_cpu(lrh->undo_len); + + /* it is time to apply the undo action */ + err = do_action(log, oe, lrh, t16, data, dlen, rec_len, NULL); + +read_next_log_undo_action: + /* + * Keep reading and looping back until we have read the + * last record for this transaction + */ + err = read_next_log_rec(log, lcb, &rec_lsn); + if (err) + goto out; + + if (rec_lsn) + goto undo_action_next; + + lcb_put(lcb); + lcb = NULL; + +commit_undo: + free_rsttbl_idx(trtbl, log->transaction_id); + + log->transaction_id = 0; + + goto transaction_table_next; + +undo_action_done: + + ntfs_update_mftmirr(sbi, 0); + + sbi->flags &= ~NTFS_FLAGS_NEED_REPLAY; + +end_reply: + + err = 0; + if (is_ro) + goto out; + + rh = ntfs_zalloc(log->page_size); + if (!rh) { + err = -ENOMEM; + goto out; + } + + rh->rhdr.sign = NTFS_RSTR_SIGNATURE; + rh->rhdr.fix_off = cpu_to_le16(offsetof(struct RESTART_HDR, fixups)); + t16 = (log->page_size >> SECTOR_SHIFT) + 1; + rh->rhdr.fix_num = cpu_to_le16(t16); + rh->sys_page_size = cpu_to_le32(log->page_size); + rh->page_size = cpu_to_le32(log->page_size); + + t16 = QuadAlign(offsetof(struct RESTART_HDR, fixups) + + sizeof(short) * t16); + rh->ra_off = cpu_to_le16(t16); + rh->minor_ver = cpu_to_le16(1); // 0x1A: + rh->major_ver = cpu_to_le16(1); // 0x1C: + + ra2 = Add2Ptr(rh, t16); + memcpy(ra2, ra, sizeof(struct RESTART_AREA)); + + ra2->client_idx[0] = 0; + ra2->client_idx[1] = LFS_NO_CLIENT_LE; + ra2->flags = cpu_to_le16(2); + + le32_add_cpu(&ra2->open_log_count, 1); + + ntfs_fix_pre_write(&rh->rhdr, log->page_size); + + err = ntfs_sb_write_run(sbi, &ni->file.run, 0, rh, log->page_size); + if (!err) + err = ntfs_sb_write_run(sbi, &log->ni->file.run, log->page_size, + rh, log->page_size); + + ntfs_free(rh); + if (err) + goto out; + +out: + ntfs_free(rst); + if (lcb) + lcb_put(lcb); + + /* Scan the Open Attribute Table to close all of the open attributes */ + oe = NULL; + while ((oe = enum_rstbl(oatbl, oe))) { + rno = ino_get(&oe->ref); + + if (oe->is_attr_name == 1) { + ntfs_free(oe->ptr); + oe->ptr = NULL; + continue; + } + + if (oe->is_attr_name) + continue; + + oa = oe->ptr; + if (!oa) + continue; + + run_close(&oa->run0); + ntfs_free(oa->attr); + if (oa->ni) + iput(&oa->ni->vfs_inode); + ntfs_free(oa); + } + + ntfs_free(trtbl); + ntfs_free(oatbl); + ntfs_free(dptbl); + ntfs_free(attr_names); + ntfs_free(rst_info.r_page); + + ntfs_free(ra); + ntfs_free(log->one_page_buf); + + if (err) + sbi->flags |= NTFS_FLAGS_NEED_REPLAY; + + if (err == -EROFS) + err = 0; + else if (log->set_dirty) + ntfs_set_state(sbi, NTFS_DIRTY_ERROR); + + ntfs_free(log); + + return err; +} -- 2.30.0

1 0

[PATCH OLK-5.10 006/107] fs/ntfs3: Add compression
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> mainline inclusion from mainline-v5.15-rc1 commit 522e010b58379fbe19b38fdef5016bca0c3cf405 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- This patch adds different types of NTFS-applicable compressions: - lznt - lzx - xpress Latter two (lzx, xpress) implement Windows Compact OS feature and were taken from ntfs-3g system comression plugin authored by Eric Biggers (https://github.com/ebiggers/ntfs-3g-system-compression) which were ported to ntfs3 and adapted to Linux Kernel environment. Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/lib/decompress_common.c | 332 +++++++++++++++ fs/ntfs3/lib/decompress_common.h | 352 ++++++++++++++++ fs/ntfs3/lib/lib.h | 26 ++ fs/ntfs3/lib/lzx_decompress.c | 683 +++++++++++++++++++++++++++++++ fs/ntfs3/lib/xpress_decompress.c | 155 +++++++ fs/ntfs3/lznt.c | 452 ++++++++++++++++++++ 6 files changed, 2000 insertions(+) create mode 100644 fs/ntfs3/lib/decompress_common.c create mode 100644 fs/ntfs3/lib/decompress_common.h create mode 100644 fs/ntfs3/lib/lib.h create mode 100644 fs/ntfs3/lib/lzx_decompress.c create mode 100644 fs/ntfs3/lib/xpress_decompress.c create mode 100644 fs/ntfs3/lznt.c diff --git a/fs/ntfs3/lib/decompress_common.c b/fs/ntfs3/lib/decompress_common.c new file mode 100644 index 000000000000..83c9e93aea77 --- /dev/null +++ b/fs/ntfs3/lib/decompress_common.c @@ -0,0 +1,332 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * decompress_common.c - Code shared by the XPRESS and LZX decompressors + * + * Copyright (C) 2015 Eric Biggers + * + * This program is free software: you can redistribute it and/or modify it under + * the terms of the GNU General Public License as published by the Free Software + * Foundation, either version 2 of the License, or (at your option) any later + * version. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS + * FOR A PARTICULAR PURPOSE. See the GNU General Public License for more + * details. + * + * You should have received a copy of the GNU General Public License along with + * this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#include "decompress_common.h" + +/* + * make_huffman_decode_table() - + * + * Build a decoding table for a canonical prefix code, or "Huffman code". + * + * This is an internal function, not part of the library API! + * + * This takes as input the length of the codeword for each symbol in the + * alphabet and produces as output a table that can be used for fast + * decoding of prefix-encoded symbols using read_huffsym(). + * + * Strictly speaking, a canonical prefix code might not be a Huffman + * code. But this algorithm will work either way; and in fact, since + * Huffman codes are defined in terms of symbol frequencies, there is no + * way for the decompressor to know whether the code is a true Huffman + * code or not until all symbols have been decoded. + * + * Because the prefix code is assumed to be "canonical", it can be + * reconstructed directly from the codeword lengths. A prefix code is + * canonical if and only if a longer codeword never lexicographically + * precedes a shorter codeword, and the lexicographic ordering of + * codewords of the same length is the same as the lexicographic ordering + * of the corresponding symbols. Consequently, we can sort the symbols + * primarily by codeword length and secondarily by symbol value, then + * reconstruct the prefix code by generating codewords lexicographically + * in that order. + * + * This function does not, however, generate the prefix code explicitly. + * Instead, it directly builds a table for decoding symbols using the + * code. The basic idea is this: given the next 'max_codeword_len' bits + * in the input, we can look up the decoded symbol by indexing a table + * containing 2**max_codeword_len entries. A codeword with length + * 'max_codeword_len' will have exactly one entry in this table, whereas + * a codeword shorter than 'max_codeword_len' will have multiple entries + * in this table. Precisely, a codeword of length n will be represented + * by 2**(max_codeword_len - n) entries in this table. The 0-based index + * of each such entry will contain the corresponding codeword as a prefix + * when zero-padded on the left to 'max_codeword_len' binary digits. + * + * That's the basic idea, but we implement two optimizations regarding + * the format of the decode table itself: + * + * - For many compression formats, the maximum codeword length is too + * long for it to be efficient to build the full decoding table + * whenever a new prefix code is used. Instead, we can build the table + * using only 2**table_bits entries, where 'table_bits' is some number + * less than or equal to 'max_codeword_len'. Then, only codewords of + * length 'table_bits' and shorter can be directly looked up. For + * longer codewords, the direct lookup instead produces the root of a + * binary tree. Using this tree, the decoder can do traditional + * bit-by-bit decoding of the remainder of the codeword. Child nodes + * are allocated in extra entries at the end of the table; leaf nodes + * contain symbols. Note that the long-codeword case is, in general, + * not performance critical, since in Huffman codes the most frequently + * used symbols are assigned the shortest codeword lengths. + * + * - When we decode a symbol using a direct lookup of the table, we still + * need to know its length so that the bitstream can be advanced by the + * appropriate number of bits. The simple solution is to simply retain + * the 'lens' array and use the decoded symbol as an index into it. + * However, this requires two separate array accesses in the fast path. + * The optimization is to store the length directly in the decode + * table. We use the bottom 11 bits for the symbol and the top 5 bits + * for the length. In addition, to combine this optimization with the + * previous one, we introduce a special case where the top 2 bits of + * the length are both set if the entry is actually the root of a + * binary tree. + * + * @decode_table: + * The array in which to create the decoding table. This must have + * a length of at least ((2**table_bits) + 2 * num_syms) entries. + * + * @num_syms: + * The number of symbols in the alphabet; also, the length of the + * 'lens' array. Must be less than or equal to 2048. + * + * @table_bits: + * The order of the decode table size, as explained above. Must be + * less than or equal to 13. + * + * @lens: + * An array of length @num_syms, indexable by symbol, that gives the + * length of the codeword, in bits, for that symbol. The length can + * be 0, which means that the symbol does not have a codeword + * assigned. + * + * @max_codeword_len: + * The longest codeword length allowed in the compression format. + * All entries in 'lens' must be less than or equal to this value. + * This must be less than or equal to 23. + * + * @working_space + * A temporary array of length '2 * (max_codeword_len + 1) + + * num_syms'. + * + * Returns 0 on success, or -1 if the lengths do not form a valid prefix + * code. + */ +int make_huffman_decode_table(u16 decode_table[], const u32 num_syms, + const u32 table_bits, const u8 lens[], + const u32 max_codeword_len, + u16 working_space[]) +{ + const u32 table_num_entries = 1 << table_bits; + u16 * const len_counts = &working_space[0]; + u16 * const offsets = &working_space[1 * (max_codeword_len + 1)]; + u16 * const sorted_syms = &working_space[2 * (max_codeword_len + 1)]; + int left; + void *decode_table_ptr; + u32 sym_idx; + u32 codeword_len; + u32 stores_per_loop; + u32 decode_table_pos; + u32 len; + u32 sym; + + /* Count how many symbols have each possible codeword length. + * Note that a length of 0 indicates the corresponding symbol is not + * used in the code and therefore does not have a codeword. + */ + for (len = 0; len <= max_codeword_len; len++) + len_counts[len] = 0; + for (sym = 0; sym < num_syms; sym++) + len_counts[lens[sym]]++; + + /* We can assume all lengths are <= max_codeword_len, but we + * cannot assume they form a valid prefix code. A codeword of + * length n should require a proportion of the codespace equaling + * (1/2)^n. The code is valid if and only if the codespace is + * exactly filled by the lengths, by this measure. + */ + left = 1; + for (len = 1; len <= max_codeword_len; len++) { + left <<= 1; + left -= len_counts[len]; + if (left < 0) { + /* The lengths overflow the codespace; that is, the code + * is over-subscribed. + */ + return -1; + } + } + + if (left) { + /* The lengths do not fill the codespace; that is, they form an + * incomplete set. + */ + if (left == (1 << max_codeword_len)) { + /* The code is completely empty. This is arguably + * invalid, but in fact it is valid in LZX and XPRESS, + * so we must allow it. By definition, no symbols can + * be decoded with an empty code. Consequently, we + * technically don't even need to fill in the decode + * table. However, to avoid accessing uninitialized + * memory if the algorithm nevertheless attempts to + * decode symbols using such a code, we zero out the + * decode table. + */ + memset(decode_table, 0, + table_num_entries * sizeof(decode_table[0])); + return 0; + } + return -1; + } + + /* Sort the symbols primarily by length and secondarily by symbol order. + */ + + /* Initialize 'offsets' so that offsets[len] for 1 <= len <= + * max_codeword_len is the number of codewords shorter than 'len' bits. + */ + offsets[1] = 0; + for (len = 1; len < max_codeword_len; len++) + offsets[len + 1] = offsets[len] + len_counts[len]; + + /* Use the 'offsets' array to sort the symbols. Note that we do not + * include symbols that are not used in the code. Consequently, fewer + * than 'num_syms' entries in 'sorted_syms' may be filled. + */ + for (sym = 0; sym < num_syms; sym++) + if (lens[sym]) + sorted_syms[offsets[lens[sym]]++] = sym; + + /* Fill entries for codewords with length <= table_bits + * --- that is, those short enough for a direct mapping. + * + * The table will start with entries for the shortest codeword(s), which + * have the most entries. From there, the number of entries per + * codeword will decrease. + */ + decode_table_ptr = decode_table; + sym_idx = 0; + codeword_len = 1; + stores_per_loop = (1 << (table_bits - codeword_len)); + for (; stores_per_loop != 0; codeword_len++, stores_per_loop >>= 1) { + u32 end_sym_idx = sym_idx + len_counts[codeword_len]; + + for (; sym_idx < end_sym_idx; sym_idx++) { + u16 entry; + u16 *p; + u32 n; + + entry = ((u32)codeword_len << 11) | sorted_syms[sym_idx]; + p = (u16 *)decode_table_ptr; + n = stores_per_loop; + + do { + *p++ = entry; + } while (--n); + + decode_table_ptr = p; + } + } + + /* If we've filled in the entire table, we are done. Otherwise, + * there are codewords longer than table_bits for which we must + * generate binary trees. + */ + decode_table_pos = (u16 *)decode_table_ptr - decode_table; + if (decode_table_pos != table_num_entries) { + u32 j; + u32 next_free_tree_slot; + u32 cur_codeword; + + /* First, zero out the remaining entries. This is + * necessary so that these entries appear as + * "unallocated" in the next part. Each of these entries + * will eventually be filled with the representation of + * the root node of a binary tree. + */ + j = decode_table_pos; + do { + decode_table[j] = 0; + } while (++j != table_num_entries); + + /* We allocate child nodes starting at the end of the + * direct lookup table. Note that there should be + * 2*num_syms extra entries for this purpose, although + * fewer than this may actually be needed. + */ + next_free_tree_slot = table_num_entries; + + /* Iterate through each codeword with length greater than + * 'table_bits', primarily in order of codeword length + * and secondarily in order of symbol. + */ + for (cur_codeword = decode_table_pos << 1; + codeword_len <= max_codeword_len; + codeword_len++, cur_codeword <<= 1) { + u32 end_sym_idx = sym_idx + len_counts[codeword_len]; + + for (; sym_idx < end_sym_idx; sym_idx++, cur_codeword++) { + /* 'sorted_sym' is the symbol represented by the + * codeword. + */ + u32 sorted_sym = sorted_syms[sym_idx]; + u32 extra_bits = codeword_len - table_bits; + u32 node_idx = cur_codeword >> extra_bits; + + /* Go through each bit of the current codeword + * beyond the prefix of length @table_bits and + * walk the appropriate binary tree, allocating + * any slots that have not yet been allocated. + * + * Note that the 'pointer' entry to the binary + * tree, which is stored in the direct lookup + * portion of the table, is represented + * identically to other internal (non-leaf) + * nodes of the binary tree; it can be thought + * of as simply the root of the tree. The + * representation of these internal nodes is + * simply the index of the left child combined + * with the special bits 0xC000 to distingush + * the entry from direct mapping and leaf node + * entries. + */ + do { + /* At least one bit remains in the + * codeword, but the current node is an + * unallocated leaf. Change it to an + * internal node. + */ + if (decode_table[node_idx] == 0) { + decode_table[node_idx] = + next_free_tree_slot | 0xC000; + decode_table[next_free_tree_slot++] = 0; + decode_table[next_free_tree_slot++] = 0; + } + + /* Go to the left child if the next bit + * in the codeword is 0; otherwise go to + * the right child. + */ + node_idx = decode_table[node_idx] & 0x3FFF; + --extra_bits; + node_idx += (cur_codeword >> extra_bits) & 1; + } while (extra_bits != 0); + + /* We've traversed the tree using the entire + * codeword, and we're now at the entry where + * the actual symbol will be stored. This is + * distinguished from internal nodes by not + * having its high two bits set. + */ + decode_table[node_idx] = sorted_sym; + } + } + } + return 0; +} diff --git a/fs/ntfs3/lib/decompress_common.h b/fs/ntfs3/lib/decompress_common.h new file mode 100644 index 000000000000..66297f398403 --- /dev/null +++ b/fs/ntfs3/lib/decompress_common.h @@ -0,0 +1,352 @@ +/* SPDX-License-Identifier: GPL-2.0-or-later */ + +/* + * decompress_common.h - Code shared by the XPRESS and LZX decompressors + * + * Copyright (C) 2015 Eric Biggers + * + * This program is free software: you can redistribute it and/or modify it under + * the terms of the GNU General Public License as published by the Free Software + * Foundation, either version 2 of the License, or (at your option) any later + * version. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS + * FOR A PARTICULAR PURPOSE. See the GNU General Public License for more + * details. + * + * You should have received a copy of the GNU General Public License along with + * this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#include <linux/string.h> +#include <linux/compiler.h> +#include <linux/types.h> +#include <linux/slab.h> +#include <asm/unaligned.h> + + +/* "Force inline" macro (not required, but helpful for performance) */ +#define forceinline __always_inline + +/* Enable whole-word match copying on selected architectures */ +#if defined(__i386__) || defined(__x86_64__) || defined(__ARM_FEATURE_UNALIGNED) +# define FAST_UNALIGNED_ACCESS +#endif + +/* Size of a machine word */ +#define WORDBYTES (sizeof(size_t)) + +static forceinline void +copy_unaligned_word(const void *src, void *dst) +{ + put_unaligned(get_unaligned((const size_t *)src), (size_t *)dst); +} + + +/* Generate a "word" with platform-dependent size whose bytes all contain the + * value 'b'. + */ +static forceinline size_t repeat_byte(u8 b) +{ + size_t v; + + v = b; + v |= v << 8; + v |= v << 16; + v |= v << ((WORDBYTES == 8) ? 32 : 0); + return v; +} + +/* Structure that encapsulates a block of in-memory data being interpreted as a + * stream of bits, optionally with interwoven literal bytes. Bits are assumed + * to be stored in little endian 16-bit coding units, with the bits ordered high + * to low. + */ +struct input_bitstream { + + /* Bits that have been read from the input buffer. The bits are + * left-justified; the next bit is always bit 31. + */ + u32 bitbuf; + + /* Number of bits currently held in @bitbuf. */ + u32 bitsleft; + + /* Pointer to the next byte to be retrieved from the input buffer. */ + const u8 *next; + + /* Pointer to just past the end of the input buffer. */ + const u8 *end; +}; + +/* Initialize a bitstream to read from the specified input buffer. */ +static forceinline void init_input_bitstream(struct input_bitstream *is, + const void *buffer, u32 size) +{ + is->bitbuf = 0; + is->bitsleft = 0; + is->next = buffer; + is->end = is->next + size; +} + +/* Ensure the bit buffer variable for the bitstream contains at least @num_bits + * bits. Following this, bitstream_peek_bits() and/or bitstream_remove_bits() + * may be called on the bitstream to peek or remove up to @num_bits bits. Note + * that @num_bits must be <= 16. + */ +static forceinline void bitstream_ensure_bits(struct input_bitstream *is, + u32 num_bits) +{ + if (is->bitsleft < num_bits) { + if (is->end - is->next >= 2) { + is->bitbuf |= (u32)get_unaligned_le16(is->next) + << (16 - is->bitsleft); + is->next += 2; + } + is->bitsleft += 16; + } +} + +/* Return the next @num_bits bits from the bitstream, without removing them. + * There must be at least @num_bits remaining in the buffer variable, from a + * previous call to bitstream_ensure_bits(). + */ +static forceinline u32 +bitstream_peek_bits(const struct input_bitstream *is, const u32 num_bits) +{ + return (is->bitbuf >> 1) >> (sizeof(is->bitbuf) * 8 - num_bits - 1); +} + +/* Remove @num_bits from the bitstream. There must be at least @num_bits + * remaining in the buffer variable, from a previous call to + * bitstream_ensure_bits(). + */ +static forceinline void +bitstream_remove_bits(struct input_bitstream *is, u32 num_bits) +{ + is->bitbuf <<= num_bits; + is->bitsleft -= num_bits; +} + +/* Remove and return @num_bits bits from the bitstream. There must be at least + * @num_bits remaining in the buffer variable, from a previous call to + * bitstream_ensure_bits(). + */ +static forceinline u32 +bitstream_pop_bits(struct input_bitstream *is, u32 num_bits) +{ + u32 bits = bitstream_peek_bits(is, num_bits); + + bitstream_remove_bits(is, num_bits); + return bits; +} + +/* Read and return the next @num_bits bits from the bitstream. */ +static forceinline u32 +bitstream_read_bits(struct input_bitstream *is, u32 num_bits) +{ + bitstream_ensure_bits(is, num_bits); + return bitstream_pop_bits(is, num_bits); +} + +/* Read and return the next literal byte embedded in the bitstream. */ +static forceinline u8 +bitstream_read_byte(struct input_bitstream *is) +{ + if (unlikely(is->end == is->next)) + return 0; + return *is->next++; +} + +/* Read and return the next 16-bit integer embedded in the bitstream. */ +static forceinline u16 +bitstream_read_u16(struct input_bitstream *is) +{ + u16 v; + + if (unlikely(is->end - is->next < 2)) + return 0; + v = get_unaligned_le16(is->next); + is->next += 2; + return v; +} + +/* Read and return the next 32-bit integer embedded in the bitstream. */ +static forceinline u32 +bitstream_read_u32(struct input_bitstream *is) +{ + u32 v; + + if (unlikely(is->end - is->next < 4)) + return 0; + v = get_unaligned_le32(is->next); + is->next += 4; + return v; +} + +/* Read into @dst_buffer an array of literal bytes embedded in the bitstream. + * Return either a pointer to the byte past the last written, or NULL if the + * read overflows the input buffer. + */ +static forceinline void *bitstream_read_bytes(struct input_bitstream *is, + void *dst_buffer, size_t count) +{ + if ((size_t)(is->end - is->next) < count) + return NULL; + memcpy(dst_buffer, is->next, count); + is->next += count; + return (u8 *)dst_buffer + count; +} + +/* Align the input bitstream on a coding-unit boundary. */ +static forceinline void bitstream_align(struct input_bitstream *is) +{ + is->bitsleft = 0; + is->bitbuf = 0; +} + +extern int make_huffman_decode_table(u16 decode_table[], const u32 num_syms, + const u32 num_bits, const u8 lens[], + const u32 max_codeword_len, + u16 working_space[]); + + +/* Reads and returns the next Huffman-encoded symbol from a bitstream. If the + * input data is exhausted, the Huffman symbol is decoded as if the missing bits + * are all zeroes. + */ +static forceinline u32 read_huffsym(struct input_bitstream *istream, + const u16 decode_table[], + u32 table_bits, + u32 max_codeword_len) +{ + u32 entry; + u32 key_bits; + + bitstream_ensure_bits(istream, max_codeword_len); + + /* Index the decode table by the next table_bits bits of the input. */ + key_bits = bitstream_peek_bits(istream, table_bits); + entry = decode_table[key_bits]; + if (entry < 0xC000) { + /* Fast case: The decode table directly provided the + * symbol and codeword length. The low 11 bits are the + * symbol, and the high 5 bits are the codeword length. + */ + bitstream_remove_bits(istream, entry >> 11); + return entry & 0x7FF; + } + /* Slow case: The codeword for the symbol is longer than + * table_bits, so the symbol does not have an entry + * directly in the first (1 << table_bits) entries of the + * decode table. Traverse the appropriate binary tree + * bit-by-bit to decode the symbol. + */ + bitstream_remove_bits(istream, table_bits); + do { + key_bits = (entry & 0x3FFF) + bitstream_pop_bits(istream, 1); + } while ((entry = decode_table[key_bits]) >= 0xC000); + return entry; +} + +/* + * Copy an LZ77 match at (dst - offset) to dst. + * + * The length and offset must be already validated --- that is, (dst - offset) + * can't underrun the output buffer, and (dst + length) can't overrun the output + * buffer. Also, the length cannot be 0. + * + * @bufend points to the byte past the end of the output buffer. This function + * won't write any data beyond this position. + * + * Returns dst + length. + */ +static forceinline u8 *lz_copy(u8 *dst, u32 length, u32 offset, const u8 *bufend, + u32 min_length) +{ + const u8 *src = dst - offset; + + /* + * Try to copy one machine word at a time. On i386 and x86_64 this is + * faster than copying one byte at a time, unless the data is + * near-random and all the matches have very short lengths. Note that + * since this requires unaligned memory accesses, it won't necessarily + * be faster on every architecture. + * + * Also note that we might copy more than the length of the match. For + * example, if a word is 8 bytes and the match is of length 5, then + * we'll simply copy 8 bytes. This is okay as long as we don't write + * beyond the end of the output buffer, hence the check for (bufend - + * end >= WORDBYTES - 1). + */ +#ifdef FAST_UNALIGNED_ACCESS + u8 * const end = dst + length; + + if (bufend - end >= (ptrdiff_t)(WORDBYTES - 1)) { + + if (offset >= WORDBYTES) { + /* The source and destination words don't overlap. */ + + /* To improve branch prediction, one iteration of this + * loop is unrolled. Most matches are short and will + * fail the first check. But if that check passes, then + * it becomes increasing likely that the match is long + * and we'll need to continue copying. + */ + + copy_unaligned_word(src, dst); + src += WORDBYTES; + dst += WORDBYTES; + + if (dst < end) { + do { + copy_unaligned_word(src, dst); + src += WORDBYTES; + dst += WORDBYTES; + } while (dst < end); + } + return end; + } else if (offset == 1) { + + /* Offset 1 matches are equivalent to run-length + * encoding of the previous byte. This case is common + * if the data contains many repeated bytes. + */ + size_t v = repeat_byte(*(dst - 1)); + + do { + put_unaligned(v, (size_t *)dst); + src += WORDBYTES; + dst += WORDBYTES; + } while (dst < end); + return end; + } + /* + * We don't bother with special cases for other 'offset < + * WORDBYTES', which are usually rarer than 'offset == 1'. Extra + * checks will just slow things down. Actually, it's possible + * to handle all the 'offset < WORDBYTES' cases using the same + * code, but it still becomes more complicated doesn't seem any + * faster overall; it definitely slows down the more common + * 'offset == 1' case. + */ + } +#endif /* FAST_UNALIGNED_ACCESS */ + + /* Fall back to a bytewise copy. */ + + if (min_length >= 2) { + *dst++ = *src++; + length--; + } + if (min_length >= 3) { + *dst++ = *src++; + length--; + } + do { + *dst++ = *src++; + } while (--length); + + return dst; +} diff --git a/fs/ntfs3/lib/lib.h b/fs/ntfs3/lib/lib.h new file mode 100644 index 000000000000..f508fbad2e71 --- /dev/null +++ b/fs/ntfs3/lib/lib.h @@ -0,0 +1,26 @@ +/* SPDX-License-Identifier: GPL-2.0-or-later */ +/* + * Adapted for linux kernel by Alexander Mamaev: + * - remove implementations of get_unaligned_ + * - assume GCC is always defined + * - ISO C90 + * - linux kernel code style + */ + + +/* globals from xpress_decompress.c */ +struct xpress_decompressor *xpress_allocate_decompressor(void); +void xpress_free_decompressor(struct xpress_decompressor *d); +int xpress_decompress(struct xpress_decompressor *__restrict d, + const void *__restrict compressed_data, + size_t compressed_size, + void *__restrict uncompressed_data, + size_t uncompressed_size); + +/* globals from lzx_decompress.c */ +struct lzx_decompressor *lzx_allocate_decompressor(void); +void lzx_free_decompressor(struct lzx_decompressor *d); +int lzx_decompress(struct lzx_decompressor *__restrict d, + const void *__restrict compressed_data, + size_t compressed_size, void *__restrict uncompressed_data, + size_t uncompressed_size); diff --git a/fs/ntfs3/lib/lzx_decompress.c b/fs/ntfs3/lib/lzx_decompress.c new file mode 100644 index 000000000000..77a381a693d1 --- /dev/null +++ b/fs/ntfs3/lib/lzx_decompress.c @@ -0,0 +1,683 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * lzx_decompress.c - A decompressor for the LZX compression format, which can + * be used in "System Compressed" files. This is based on the code from wimlib. + * This code only supports a window size (dictionary size) of 32768 bytes, since + * this is the only size used in System Compression. + * + * Copyright (C) 2015 Eric Biggers + * + * This program is free software: you can redistribute it and/or modify it under + * the terms of the GNU General Public License as published by the Free Software + * Foundation, either version 2 of the License, or (at your option) any later + * version. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS + * FOR A PARTICULAR PURPOSE. See the GNU General Public License for more + * details. + * + * You should have received a copy of the GNU General Public License along with + * this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#include "decompress_common.h" +#include "lib.h" + +/* Number of literal byte values */ +#define LZX_NUM_CHARS 256 + +/* The smallest and largest allowed match lengths */ +#define LZX_MIN_MATCH_LEN 2 +#define LZX_MAX_MATCH_LEN 257 + +/* Number of distinct match lengths that can be represented */ +#define LZX_NUM_LENS (LZX_MAX_MATCH_LEN - LZX_MIN_MATCH_LEN + 1) + +/* Number of match lengths for which no length symbol is required */ +#define LZX_NUM_PRIMARY_LENS 7 +#define LZX_NUM_LEN_HEADERS (LZX_NUM_PRIMARY_LENS + 1) + +/* Valid values of the 3-bit block type field */ +#define LZX_BLOCKTYPE_VERBATIM 1 +#define LZX_BLOCKTYPE_ALIGNED 2 +#define LZX_BLOCKTYPE_UNCOMPRESSED 3 + +/* Number of offset slots for a window size of 32768 */ +#define LZX_NUM_OFFSET_SLOTS 30 + +/* Number of symbols in the main code for a window size of 32768 */ +#define LZX_MAINCODE_NUM_SYMBOLS \ + (LZX_NUM_CHARS + (LZX_NUM_OFFSET_SLOTS * LZX_NUM_LEN_HEADERS)) + +/* Number of symbols in the length code */ +#define LZX_LENCODE_NUM_SYMBOLS (LZX_NUM_LENS - LZX_NUM_PRIMARY_LENS) + +/* Number of symbols in the precode */ +#define LZX_PRECODE_NUM_SYMBOLS 20 + +/* Number of bits in which each precode codeword length is represented */ +#define LZX_PRECODE_ELEMENT_SIZE 4 + +/* Number of low-order bits of each match offset that are entropy-encoded in + * aligned offset blocks + */ +#define LZX_NUM_ALIGNED_OFFSET_BITS 3 + +/* Number of symbols in the aligned offset code */ +#define LZX_ALIGNEDCODE_NUM_SYMBOLS (1 << LZX_NUM_ALIGNED_OFFSET_BITS) + +/* Mask for the match offset bits that are entropy-encoded in aligned offset + * blocks + */ +#define LZX_ALIGNED_OFFSET_BITMASK ((1 << LZX_NUM_ALIGNED_OFFSET_BITS) - 1) + +/* Number of bits in which each aligned offset codeword length is represented */ +#define LZX_ALIGNEDCODE_ELEMENT_SIZE 3 + +/* Maximum lengths (in bits) of the codewords in each Huffman code */ +#define LZX_MAX_MAIN_CODEWORD_LEN 16 +#define LZX_MAX_LEN_CODEWORD_LEN 16 +#define LZX_MAX_PRE_CODEWORD_LEN ((1 << LZX_PRECODE_ELEMENT_SIZE) - 1) +#define LZX_MAX_ALIGNED_CODEWORD_LEN ((1 << LZX_ALIGNEDCODE_ELEMENT_SIZE) - 1) + +/* The default "filesize" value used in pre/post-processing. In the LZX format + * used in cabinet files this value must be given to the decompressor, whereas + * in the LZX format used in WIM files and system-compressed files this value is + * fixed at 12000000. + */ +#define LZX_DEFAULT_FILESIZE 12000000 + +/* Assumed block size when the encoded block size begins with a 0 bit. */ +#define LZX_DEFAULT_BLOCK_SIZE 32768 + +/* Number of offsets in the recent (or "repeat") offsets queue. */ +#define LZX_NUM_RECENT_OFFSETS 3 + +/* These values are chosen for fast decompression. */ +#define LZX_MAINCODE_TABLEBITS 11 +#define LZX_LENCODE_TABLEBITS 10 +#define LZX_PRECODE_TABLEBITS 6 +#define LZX_ALIGNEDCODE_TABLEBITS 7 + +#define LZX_READ_LENS_MAX_OVERRUN 50 + +/* Mapping: offset slot => first match offset that uses that offset slot. + */ +static const u32 lzx_offset_slot_base[LZX_NUM_OFFSET_SLOTS + 1] = { + 0, 1, 2, 3, 4, /* 0 --- 4 */ + 6, 8, 12, 16, 24, /* 5 --- 9 */ + 32, 48, 64, 96, 128, /* 10 --- 14 */ + 192, 256, 384, 512, 768, /* 15 --- 19 */ + 1024, 1536, 2048, 3072, 4096, /* 20 --- 24 */ + 6144, 8192, 12288, 16384, 24576, /* 25 --- 29 */ + 32768, /* extra */ +}; + +/* Mapping: offset slot => how many extra bits must be read and added to the + * corresponding offset slot base to decode the match offset. + */ +static const u8 lzx_extra_offset_bits[LZX_NUM_OFFSET_SLOTS] = { + 0, 0, 0, 0, 1, + 1, 2, 2, 3, 3, + 4, 4, 5, 5, 6, + 6, 7, 7, 8, 8, + 9, 9, 10, 10, 11, + 11, 12, 12, 13, 13, +}; + +/* Reusable heap-allocated memory for LZX decompression */ +struct lzx_decompressor { + + /* Huffman decoding tables, and arrays that map symbols to codeword + * lengths + */ + + u16 maincode_decode_table[(1 << LZX_MAINCODE_TABLEBITS) + + (LZX_MAINCODE_NUM_SYMBOLS * 2)]; + u8 maincode_lens[LZX_MAINCODE_NUM_SYMBOLS + LZX_READ_LENS_MAX_OVERRUN]; + + + u16 lencode_decode_table[(1 << LZX_LENCODE_TABLEBITS) + + (LZX_LENCODE_NUM_SYMBOLS * 2)]; + u8 lencode_lens[LZX_LENCODE_NUM_SYMBOLS + LZX_READ_LENS_MAX_OVERRUN]; + + + u16 alignedcode_decode_table[(1 << LZX_ALIGNEDCODE_TABLEBITS) + + (LZX_ALIGNEDCODE_NUM_SYMBOLS * 2)]; + u8 alignedcode_lens[LZX_ALIGNEDCODE_NUM_SYMBOLS]; + + u16 precode_decode_table[(1 << LZX_PRECODE_TABLEBITS) + + (LZX_PRECODE_NUM_SYMBOLS * 2)]; + u8 precode_lens[LZX_PRECODE_NUM_SYMBOLS]; + + /* Temporary space for make_huffman_decode_table() */ + u16 working_space[2 * (1 + LZX_MAX_MAIN_CODEWORD_LEN) + + LZX_MAINCODE_NUM_SYMBOLS]; +}; + +static void undo_e8_translation(void *target, s32 input_pos) +{ + s32 abs_offset, rel_offset; + + abs_offset = get_unaligned_le32(target); + if (abs_offset >= 0) { + if (abs_offset < LZX_DEFAULT_FILESIZE) { + /* "good translation" */ + rel_offset = abs_offset - input_pos; + put_unaligned_le32(rel_offset, target); + } + } else { + if (abs_offset >= -input_pos) { + /* "compensating translation" */ + rel_offset = abs_offset + LZX_DEFAULT_FILESIZE; + put_unaligned_le32(rel_offset, target); + } + } +} + +/* + * Undo the 'E8' preprocessing used in LZX. Before compression, the + * uncompressed data was preprocessed by changing the targets of suspected x86 + * CALL instructions from relative offsets to absolute offsets. After + * match/literal decoding, the decompressor must undo the translation. + */ +static void lzx_postprocess(u8 *data, u32 size) +{ + /* + * A worthwhile optimization is to push the end-of-buffer check into the + * relatively rare E8 case. This is possible if we replace the last six + * bytes of data with E8 bytes; then we are guaranteed to hit an E8 byte + * before reaching end-of-buffer. In addition, this scheme guarantees + * that no translation can begin following an E8 byte in the last 10 + * bytes because a 4-byte offset containing E8 as its high byte is a + * large negative number that is not valid for translation. That is + * exactly what we need. + */ + u8 *tail; + u8 saved_bytes[6]; + u8 *p; + + if (size <= 10) + return; + + tail = &data[size - 6]; + memcpy(saved_bytes, tail, 6); + memset(tail, 0xE8, 6); + p = data; + for (;;) { + while (*p != 0xE8) + p++; + if (p >= tail) + break; + undo_e8_translation(p + 1, p - data); + p += 5; + } + memcpy(tail, saved_bytes, 6); +} + +/* Read a Huffman-encoded symbol using the precode. */ +static forceinline u32 read_presym(const struct lzx_decompressor *d, + struct input_bitstream *is) +{ + return read_huffsym(is, d->precode_decode_table, + LZX_PRECODE_TABLEBITS, LZX_MAX_PRE_CODEWORD_LEN); +} + +/* Read a Huffman-encoded symbol using the main code. */ +static forceinline u32 read_mainsym(const struct lzx_decompressor *d, + struct input_bitstream *is) +{ + return read_huffsym(is, d->maincode_decode_table, + LZX_MAINCODE_TABLEBITS, LZX_MAX_MAIN_CODEWORD_LEN); +} + +/* Read a Huffman-encoded symbol using the length code. */ +static forceinline u32 read_lensym(const struct lzx_decompressor *d, + struct input_bitstream *is) +{ + return read_huffsym(is, d->lencode_decode_table, + LZX_LENCODE_TABLEBITS, LZX_MAX_LEN_CODEWORD_LEN); +} + +/* Read a Huffman-encoded symbol using the aligned offset code. */ +static forceinline u32 read_alignedsym(const struct lzx_decompressor *d, + struct input_bitstream *is) +{ + return read_huffsym(is, d->alignedcode_decode_table, + LZX_ALIGNEDCODE_TABLEBITS, + LZX_MAX_ALIGNED_CODEWORD_LEN); +} + +/* + * Read the precode from the compressed input bitstream, then use it to decode + * @num_lens codeword length values. + * + * @is: The input bitstream. + * + * @lens: An array that contains the length values from the previous time + * the codeword lengths for this Huffman code were read, or all 0's + * if this is the first time. This array must have at least + * (@num_lens + LZX_READ_LENS_MAX_OVERRUN) entries. + * + * @num_lens: Number of length values to decode. + * + * Returns 0 on success, or -1 if the data was invalid. + */ +static int lzx_read_codeword_lens(struct lzx_decompressor *d, + struct input_bitstream *is, + u8 *lens, u32 num_lens) +{ + u8 *len_ptr = lens; + u8 *lens_end = lens + num_lens; + int i; + + /* Read the lengths of the precode codewords. These are given + * explicitly. + */ + for (i = 0; i < LZX_PRECODE_NUM_SYMBOLS; i++) { + d->precode_lens[i] = + bitstream_read_bits(is, LZX_PRECODE_ELEMENT_SIZE); + } + + /* Make the decoding table for the precode. */ + if (make_huffman_decode_table(d->precode_decode_table, + LZX_PRECODE_NUM_SYMBOLS, + LZX_PRECODE_TABLEBITS, + d->precode_lens, + LZX_MAX_PRE_CODEWORD_LEN, + d->working_space)) + return -1; + + /* Decode the codeword lengths. */ + do { + u32 presym; + u8 len; + + /* Read the next precode symbol. */ + presym = read_presym(d, is); + if (presym < 17) { + /* Difference from old length */ + len = *len_ptr - presym; + if ((s8)len < 0) + len += 17; + *len_ptr++ = len; + } else { + /* Special RLE values */ + + u32 run_len; + + if (presym == 17) { + /* Run of 0's */ + run_len = 4 + bitstream_read_bits(is, 4); + len = 0; + } else if (presym == 18) { + /* Longer run of 0's */ + run_len = 20 + bitstream_read_bits(is, 5); + len = 0; + } else { + /* Run of identical lengths */ + run_len = 4 + bitstream_read_bits(is, 1); + presym = read_presym(d, is); + if (presym > 17) + return -1; + len = *len_ptr - presym; + if ((s8)len < 0) + len += 17; + } + + do { + *len_ptr++ = len; + } while (--run_len); + /* Worst case overrun is when presym == 18, + * run_len == 20 + 31, and only 1 length was remaining. + * So LZX_READ_LENS_MAX_OVERRUN == 50. + * + * Overrun while reading the first half of maincode_lens + * can corrupt the previous values in the second half. + * This doesn't really matter because the resulting + * lengths will still be in range, and data that + * generates overruns is invalid anyway. + */ + } + } while (len_ptr < lens_end); + + return 0; +} + +/* + * Read the header of an LZX block and save the block type and (uncompressed) + * size in *block_type_ret and *block_size_ret, respectively. + * + * If the block is compressed, also update the Huffman decode @tables with the + * new Huffman codes. If the block is uncompressed, also update the match + * offset @queue with the new match offsets. + * + * Return 0 on success, or -1 if the data was invalid. + */ +static int lzx_read_block_header(struct lzx_decompressor *d, + struct input_bitstream *is, + int *block_type_ret, + u32 *block_size_ret, + u32 recent_offsets[]) +{ + int block_type; + u32 block_size; + int i; + + bitstream_ensure_bits(is, 4); + + /* The first three bits tell us what kind of block it is, and should be + * one of the LZX_BLOCKTYPE_* values. + */ + block_type = bitstream_pop_bits(is, 3); + + /* Read the block size. */ + if (bitstream_pop_bits(is, 1)) { + block_size = LZX_DEFAULT_BLOCK_SIZE; + } else { + block_size = 0; + block_size |= bitstream_read_bits(is, 8); + block_size <<= 8; + block_size |= bitstream_read_bits(is, 8); + } + + switch (block_type) { + + case LZX_BLOCKTYPE_ALIGNED: + + /* Read the aligned offset code and prepare its decode table. + */ + + for (i = 0; i < LZX_ALIGNEDCODE_NUM_SYMBOLS; i++) { + d->alignedcode_lens[i] = + bitstream_read_bits(is, + LZX_ALIGNEDCODE_ELEMENT_SIZE); + } + + if (make_huffman_decode_table(d->alignedcode_decode_table, + LZX_ALIGNEDCODE_NUM_SYMBOLS, + LZX_ALIGNEDCODE_TABLEBITS, + d->alignedcode_lens, + LZX_MAX_ALIGNED_CODEWORD_LEN, + d->working_space)) + return -1; + + /* Fall though, since the rest of the header for aligned offset + * blocks is the same as that for verbatim blocks. + */ + fallthrough; + + case LZX_BLOCKTYPE_VERBATIM: + + /* Read the main code and prepare its decode table. + * + * Note that the codeword lengths in the main code are encoded + * in two parts: one part for literal symbols, and one part for + * match symbols. + */ + + if (lzx_read_codeword_lens(d, is, d->maincode_lens, + LZX_NUM_CHARS)) + return -1; + + if (lzx_read_codeword_lens(d, is, + d->maincode_lens + LZX_NUM_CHARS, + LZX_MAINCODE_NUM_SYMBOLS - LZX_NUM_CHARS)) + return -1; + + if (make_huffman_decode_table(d->maincode_decode_table, + LZX_MAINCODE_NUM_SYMBOLS, + LZX_MAINCODE_TABLEBITS, + d->maincode_lens, + LZX_MAX_MAIN_CODEWORD_LEN, + d->working_space)) + return -1; + + /* Read the length code and prepare its decode table. */ + + if (lzx_read_codeword_lens(d, is, d->lencode_lens, + LZX_LENCODE_NUM_SYMBOLS)) + return -1; + + if (make_huffman_decode_table(d->lencode_decode_table, + LZX_LENCODE_NUM_SYMBOLS, + LZX_LENCODE_TABLEBITS, + d->lencode_lens, + LZX_MAX_LEN_CODEWORD_LEN, + d->working_space)) + return -1; + + break; + + case LZX_BLOCKTYPE_UNCOMPRESSED: + + /* Before reading the three recent offsets from the uncompressed + * block header, the stream must be aligned on a 16-bit + * boundary. But if the stream is *already* aligned, then the + * next 16 bits must be discarded. + */ + bitstream_ensure_bits(is, 1); + bitstream_align(is); + + recent_offsets[0] = bitstream_read_u32(is); + recent_offsets[1] = bitstream_read_u32(is); + recent_offsets[2] = bitstream_read_u32(is); + + /* Offsets of 0 are invalid. */ + if (recent_offsets[0] == 0 || recent_offsets[1] == 0 || + recent_offsets[2] == 0) + return -1; + break; + + default: + /* Unrecognized block type. */ + return -1; + } + + *block_type_ret = block_type; + *block_size_ret = block_size; + return 0; +} + +/* Decompress a block of LZX-compressed data. */ +static int lzx_decompress_block(const struct lzx_decompressor *d, + struct input_bitstream *is, + int block_type, u32 block_size, + u8 * const out_begin, u8 *out_next, + u32 recent_offsets[]) +{ + u8 * const block_end = out_next + block_size; + u32 ones_if_aligned = 0U - (block_type == LZX_BLOCKTYPE_ALIGNED); + + do { + u32 mainsym; + u32 match_len; + u32 match_offset; + u32 offset_slot; + u32 num_extra_bits; + + mainsym = read_mainsym(d, is); + if (mainsym < LZX_NUM_CHARS) { + /* Literal */ + *out_next++ = mainsym; + continue; + } + + /* Match */ + + /* Decode the length header and offset slot. */ + mainsym -= LZX_NUM_CHARS; + match_len = mainsym % LZX_NUM_LEN_HEADERS; + offset_slot = mainsym / LZX_NUM_LEN_HEADERS; + + /* If needed, read a length symbol to decode the full length. */ + if (match_len == LZX_NUM_PRIMARY_LENS) + match_len += read_lensym(d, is); + match_len += LZX_MIN_MATCH_LEN; + + if (offset_slot < LZX_NUM_RECENT_OFFSETS) { + /* Repeat offset */ + + /* Note: This isn't a real LRU queue, since using the R2 + * offset doesn't bump the R1 offset down to R2. This + * quirk allows all 3 recent offsets to be handled by + * the same code. (For R0, the swap is a no-op.) + */ + match_offset = recent_offsets[offset_slot]; + recent_offsets[offset_slot] = recent_offsets[0]; + recent_offsets[0] = match_offset; + } else { + /* Explicit offset */ + + /* Look up the number of extra bits that need to be read + * to decode offsets with this offset slot. + */ + num_extra_bits = lzx_extra_offset_bits[offset_slot]; + + /* Start with the offset slot base value. */ + match_offset = lzx_offset_slot_base[offset_slot]; + + /* In aligned offset blocks, the low-order 3 bits of + * each offset are encoded using the aligned offset + * code. Otherwise, all the extra bits are literal. + */ + + if ((num_extra_bits & ones_if_aligned) >= LZX_NUM_ALIGNED_OFFSET_BITS) { + match_offset += + bitstream_read_bits(is, num_extra_bits - + LZX_NUM_ALIGNED_OFFSET_BITS) + << LZX_NUM_ALIGNED_OFFSET_BITS; + match_offset += read_alignedsym(d, is); + } else { + match_offset += bitstream_read_bits(is, num_extra_bits); + } + + /* Adjust the offset. */ + match_offset -= (LZX_NUM_RECENT_OFFSETS - 1); + + /* Update the recent offsets. */ + recent_offsets[2] = recent_offsets[1]; + recent_offsets[1] = recent_offsets[0]; + recent_offsets[0] = match_offset; + } + + /* Validate the match, then copy it to the current position. */ + + if (match_len > (size_t)(block_end - out_next)) + return -1; + + if (match_offset > (size_t)(out_next - out_begin)) + return -1; + + out_next = lz_copy(out_next, match_len, match_offset, + block_end, LZX_MIN_MATCH_LEN); + + } while (out_next != block_end); + + return 0; +} + +/* + * lzx_allocate_decompressor - Allocate an LZX decompressor + * + * Return the pointer to the decompressor on success, or return NULL and set + * errno on failure. + */ +struct lzx_decompressor *lzx_allocate_decompressor(void) +{ + return kmalloc(sizeof(struct lzx_decompressor), GFP_NOFS); +} + +/* + * lzx_decompress - Decompress a buffer of LZX-compressed data + * + * @decompressor: A decompressor allocated with lzx_allocate_decompressor() + * @compressed_data: The buffer of data to decompress + * @compressed_size: Number of bytes of compressed data + * @uncompressed_data: The buffer in which to store the decompressed data + * @uncompressed_size: The number of bytes the data decompresses into + * + * Return 0 on success, or return -1 and set errno on failure. + */ +int lzx_decompress(struct lzx_decompressor *decompressor, + const void *compressed_data, size_t compressed_size, + void *uncompressed_data, size_t uncompressed_size) +{ + struct lzx_decompressor *d = decompressor; + u8 * const out_begin = uncompressed_data; + u8 *out_next = out_begin; + u8 * const out_end = out_begin + uncompressed_size; + struct input_bitstream is; + u32 recent_offsets[LZX_NUM_RECENT_OFFSETS] = {1, 1, 1}; + int e8_status = 0; + + init_input_bitstream(&is, compressed_data, compressed_size); + + /* Codeword lengths begin as all 0's for delta encoding purposes. */ + memset(d->maincode_lens, 0, LZX_MAINCODE_NUM_SYMBOLS); + memset(d->lencode_lens, 0, LZX_LENCODE_NUM_SYMBOLS); + + /* Decompress blocks until we have all the uncompressed data. */ + + while (out_next != out_end) { + int block_type; + u32 block_size; + + if (lzx_read_block_header(d, &is, &block_type, &block_size, + recent_offsets)) + goto invalid; + + if (block_size < 1 || block_size > (size_t)(out_end - out_next)) + goto invalid; + + if (block_type != LZX_BLOCKTYPE_UNCOMPRESSED) { + + /* Compressed block */ + + if (lzx_decompress_block(d, + &is, + block_type, + block_size, + out_begin, + out_next, + recent_offsets)) + goto invalid; + + e8_status |= d->maincode_lens[0xe8]; + out_next += block_size; + } else { + /* Uncompressed block */ + + out_next = bitstream_read_bytes(&is, out_next, + block_size); + if (!out_next) + goto invalid; + + if (block_size & 1) + bitstream_read_byte(&is); + + e8_status = 1; + } + } + + /* Postprocess the data unless it cannot possibly contain 0xe8 bytes. */ + if (e8_status) + lzx_postprocess(uncompressed_data, uncompressed_size); + + return 0; + +invalid: + return -1; +} + +/* + * lzx_free_decompressor - Free an LZX decompressor + * + * @decompressor: A decompressor that was allocated with + * lzx_allocate_decompressor(), or NULL. + */ +void lzx_free_decompressor(struct lzx_decompressor *decompressor) +{ + kfree(decompressor); +} diff --git a/fs/ntfs3/lib/xpress_decompress.c b/fs/ntfs3/lib/xpress_decompress.c new file mode 100644 index 000000000000..3d98f36a981e --- /dev/null +++ b/fs/ntfs3/lib/xpress_decompress.c @@ -0,0 +1,155 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * xpress_decompress.c - A decompressor for the XPRESS compression format + * (Huffman variant), which can be used in "System Compressed" files. This is + * based on the code from wimlib. + * + * Copyright (C) 2015 Eric Biggers + * + * This program is free software: you can redistribute it and/or modify it under + * the terms of the GNU General Public License as published by the Free Software + * Foundation, either version 2 of the License, or (at your option) any later + * version. + * + * This program is distributed in the hope that it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS + * FOR A PARTICULAR PURPOSE. See the GNU General Public License for more + * details. + * + * You should have received a copy of the GNU General Public License along with + * this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#include "decompress_common.h" +#include "lib.h" + +#define XPRESS_NUM_SYMBOLS 512 +#define XPRESS_MAX_CODEWORD_LEN 15 +#define XPRESS_MIN_MATCH_LEN 3 + +/* This value is chosen for fast decompression. */ +#define XPRESS_TABLEBITS 12 + +/* Reusable heap-allocated memory for XPRESS decompression */ +struct xpress_decompressor { + + /* The Huffman decoding table */ + u16 decode_table[(1 << XPRESS_TABLEBITS) + 2 * XPRESS_NUM_SYMBOLS]; + + /* An array that maps symbols to codeword lengths */ + u8 lens[XPRESS_NUM_SYMBOLS]; + + /* Temporary space for make_huffman_decode_table() */ + u16 working_space[2 * (1 + XPRESS_MAX_CODEWORD_LEN) + + XPRESS_NUM_SYMBOLS]; +}; + +/* + * xpress_allocate_decompressor - Allocate an XPRESS decompressor + * + * Return the pointer to the decompressor on success, or return NULL and set + * errno on failure. + */ +struct xpress_decompressor *xpress_allocate_decompressor(void) +{ + return kmalloc(sizeof(struct xpress_decompressor), GFP_NOFS); +} + +/* + * xpress_decompress - Decompress a buffer of XPRESS-compressed data + * + * @decompressor: A decompressor that was allocated with + * xpress_allocate_decompressor() + * @compressed_data: The buffer of data to decompress + * @compressed_size: Number of bytes of compressed data + * @uncompressed_data: The buffer in which to store the decompressed data + * @uncompressed_size: The number of bytes the data decompresses into + * + * Return 0 on success, or return -1 and set errno on failure. + */ +int xpress_decompress(struct xpress_decompressor *decompressor, + const void *compressed_data, size_t compressed_size, + void *uncompressed_data, size_t uncompressed_size) +{ + struct xpress_decompressor *d = decompressor; + const u8 * const in_begin = compressed_data; + u8 * const out_begin = uncompressed_data; + u8 *out_next = out_begin; + u8 * const out_end = out_begin + uncompressed_size; + struct input_bitstream is; + u32 i; + + /* Read the Huffman codeword lengths. */ + if (compressed_size < XPRESS_NUM_SYMBOLS / 2) + goto invalid; + for (i = 0; i < XPRESS_NUM_SYMBOLS / 2; i++) { + d->lens[i*2 + 0] = in_begin[i] & 0xF; + d->lens[i*2 + 1] = in_begin[i] >> 4; + } + + /* Build a decoding table for the Huffman code. */ + if (make_huffman_decode_table(d->decode_table, XPRESS_NUM_SYMBOLS, + XPRESS_TABLEBITS, d->lens, + XPRESS_MAX_CODEWORD_LEN, + d->working_space)) + goto invalid; + + /* Decode the matches and literals. */ + + init_input_bitstream(&is, in_begin + XPRESS_NUM_SYMBOLS / 2, + compressed_size - XPRESS_NUM_SYMBOLS / 2); + + while (out_next != out_end) { + u32 sym; + u32 log2_offset; + u32 length; + u32 offset; + + sym = read_huffsym(&is, d->decode_table, + XPRESS_TABLEBITS, XPRESS_MAX_CODEWORD_LEN); + if (sym < 256) { + /* Literal */ + *out_next++ = sym; + } else { + /* Match */ + length = sym & 0xf; + log2_offset = (sym >> 4) & 0xf; + + bitstream_ensure_bits(&is, 16); + + offset = ((u32)1 << log2_offset) | + bitstream_pop_bits(&is, log2_offset); + + if (length == 0xf) { + length += bitstream_read_byte(&is); + if (length == 0xf + 0xff) + length = bitstream_read_u16(&is); + } + length += XPRESS_MIN_MATCH_LEN; + + if (offset > (size_t)(out_next - out_begin)) + goto invalid; + + if (length > (size_t)(out_end - out_next)) + goto invalid; + + out_next = lz_copy(out_next, length, offset, out_end, + XPRESS_MIN_MATCH_LEN); + } + } + return 0; + +invalid: + return -1; +} + +/* + * xpress_free_decompressor - Free an XPRESS decompressor + * + * @decompressor: A decompressor that was allocated with + * xpress_allocate_decompressor(), or NULL. + */ +void xpress_free_decompressor(struct xpress_decompressor *decompressor) +{ + kfree(decompressor); +} diff --git a/fs/ntfs3/lznt.c b/fs/ntfs3/lznt.c new file mode 100644 index 000000000000..ead9ab7d69b3 --- /dev/null +++ b/fs/ntfs3/lznt.c @@ -0,0 +1,452 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * + * Copyright (C) 2019-2021 Paragon Software GmbH, All rights reserved. + * + */ +#include <linux/blkdev.h> +#include <linux/buffer_head.h> +#include <linux/fs.h> +#include <linux/nls.h> + +#include "debug.h" +#include "ntfs.h" +#include "ntfs_fs.h" + +// clang-format off +/* src buffer is zero */ +#define LZNT_ERROR_ALL_ZEROS 1 +#define LZNT_CHUNK_SIZE 0x1000 +// clang-format on + +struct lznt_hash { + const u8 *p1; + const u8 *p2; +}; + +struct lznt { + const u8 *unc; + const u8 *unc_end; + const u8 *best_match; + size_t max_len; + bool std; + + struct lznt_hash hash[LZNT_CHUNK_SIZE]; +}; + +static inline size_t get_match_len(const u8 *ptr, const u8 *end, const u8 *prev, + size_t max_len) +{ + size_t len = 0; + + while (ptr + len < end && ptr[len] == prev[len] && ++len < max_len) + ; + return len; +} + +static size_t longest_match_std(const u8 *src, struct lznt *ctx) +{ + size_t hash_index; + size_t len1 = 0, len2 = 0; + const u8 **hash; + + hash_index = + ((40543U * ((((src[0] << 4) ^ src[1]) << 4) ^ src[2])) >> 4) & + (LZNT_CHUNK_SIZE - 1); + + hash = &(ctx->hash[hash_index].p1); + + if (hash[0] >= ctx->unc && hash[0] < src && hash[0][0] == src[0] && + hash[0][1] == src[1] && hash[0][2] == src[2]) { + len1 = 3; + if (ctx->max_len > 3) + len1 += get_match_len(src + 3, ctx->unc_end, + hash[0] + 3, ctx->max_len - 3); + } + + if (hash[1] >= ctx->unc && hash[1] < src && hash[1][0] == src[0] && + hash[1][1] == src[1] && hash[1][2] == src[2]) { + len2 = 3; + if (ctx->max_len > 3) + len2 += get_match_len(src + 3, ctx->unc_end, + hash[1] + 3, ctx->max_len - 3); + } + + /* Compare two matches and select the best one */ + if (len1 < len2) { + ctx->best_match = hash[1]; + len1 = len2; + } else { + ctx->best_match = hash[0]; + } + + hash[1] = hash[0]; + hash[0] = src; + return len1; +} + +static size_t longest_match_best(const u8 *src, struct lznt *ctx) +{ + size_t max_len; + const u8 *ptr; + + if (ctx->unc >= src || !ctx->max_len) + return 0; + + max_len = 0; + for (ptr = ctx->unc; ptr < src; ++ptr) { + size_t len = + get_match_len(src, ctx->unc_end, ptr, ctx->max_len); + if (len >= max_len) { + max_len = len; + ctx->best_match = ptr; + } + } + + return max_len >= 3 ? max_len : 0; +} + +static const size_t s_max_len[] = { + 0x1002, 0x802, 0x402, 0x202, 0x102, 0x82, 0x42, 0x22, 0x12, +}; + +static const size_t s_max_off[] = { + 0x10, 0x20, 0x40, 0x80, 0x100, 0x200, 0x400, 0x800, 0x1000, +}; + +static inline u16 make_pair(size_t offset, size_t len, size_t index) +{ + return ((offset - 1) << (12 - index)) | + ((len - 3) & (((1 << (12 - index)) - 1))); +} + +static inline size_t parse_pair(u16 pair, size_t *offset, size_t index) +{ + *offset = 1 + (pair >> (12 - index)); + return 3 + (pair & ((1 << (12 - index)) - 1)); +} + +/* + * compress_chunk + * + * returns one of the three values: + * 0 - ok, 'cmpr' contains 'cmpr_chunk_size' bytes of compressed data + * 1 - input buffer is full zero + * -2 - the compressed buffer is too small to hold the compressed data + */ +static inline int compress_chunk(size_t (*match)(const u8 *, struct lznt *), + const u8 *unc, const u8 *unc_end, u8 *cmpr, + u8 *cmpr_end, size_t *cmpr_chunk_size, + struct lznt *ctx) +{ + size_t cnt = 0; + size_t idx = 0; + const u8 *up = unc; + u8 *cp = cmpr + 3; + u8 *cp2 = cmpr + 2; + u8 not_zero = 0; + /* Control byte of 8-bit values: ( 0 - means byte as is, 1 - short pair ) */ + u8 ohdr = 0; + u8 *last; + u16 t16; + + if (unc + LZNT_CHUNK_SIZE < unc_end) + unc_end = unc + LZNT_CHUNK_SIZE; + + last = min(cmpr + LZNT_CHUNK_SIZE + sizeof(short), cmpr_end); + + ctx->unc = unc; + ctx->unc_end = unc_end; + ctx->max_len = s_max_len[0]; + + while (up < unc_end) { + size_t max_len; + + while (unc + s_max_off[idx] < up) + ctx->max_len = s_max_len[++idx]; + + // Find match + max_len = up + 3 <= unc_end ? (*match)(up, ctx) : 0; + + if (!max_len) { + if (cp >= last) + goto NotCompressed; + not_zero |= *cp++ = *up++; + } else if (cp + 1 >= last) { + goto NotCompressed; + } else { + t16 = make_pair(up - ctx->best_match, max_len, idx); + *cp++ = t16; + *cp++ = t16 >> 8; + + ohdr |= 1 << cnt; + up += max_len; + } + + cnt = (cnt + 1) & 7; + if (!cnt) { + *cp2 = ohdr; + ohdr = 0; + cp2 = cp; + cp += 1; + } + } + + if (cp2 < last) + *cp2 = ohdr; + else + cp -= 1; + + *cmpr_chunk_size = cp - cmpr; + + t16 = (*cmpr_chunk_size - 3) | 0xB000; + cmpr[0] = t16; + cmpr[1] = t16 >> 8; + + return not_zero ? 0 : LZNT_ERROR_ALL_ZEROS; + +NotCompressed: + + if ((cmpr + LZNT_CHUNK_SIZE + sizeof(short)) > last) + return -2; + + /* + * Copy non cmpr data + * 0x3FFF == ((LZNT_CHUNK_SIZE + 2 - 3) | 0x3000) + */ + cmpr[0] = 0xff; + cmpr[1] = 0x3f; + + memcpy(cmpr + sizeof(short), unc, LZNT_CHUNK_SIZE); + *cmpr_chunk_size = LZNT_CHUNK_SIZE + sizeof(short); + + return 0; +} + +static inline ssize_t decompress_chunk(u8 *unc, u8 *unc_end, const u8 *cmpr, + const u8 *cmpr_end) +{ + u8 *up = unc; + u8 ch = *cmpr++; + size_t bit = 0; + size_t index = 0; + u16 pair; + size_t offset, length; + + /* Do decompression until pointers are inside range */ + while (up < unc_end && cmpr < cmpr_end) { + /* Correct index */ + while (unc + s_max_off[index] < up) + index += 1; + + /* Check the current flag for zero */ + if (!(ch & (1 << bit))) { + /* Just copy byte */ + *up++ = *cmpr++; + goto next; + } + + /* Check for boundary */ + if (cmpr + 1 >= cmpr_end) + return -EINVAL; + + /* Read a short from little endian stream */ + pair = cmpr[1]; + pair <<= 8; + pair |= cmpr[0]; + + cmpr += 2; + + /* Translate packed information into offset and length */ + length = parse_pair(pair, &offset, index); + + /* Check offset for boundary */ + if (unc + offset > up) + return -EINVAL; + + /* Truncate the length if necessary */ + if (up + length >= unc_end) + length = unc_end - up; + + /* Now we copy bytes. This is the heart of LZ algorithm. */ + for (; length > 0; length--, up++) + *up = *(up - offset); + +next: + /* Advance flag bit value */ + bit = (bit + 1) & 7; + + if (!bit) { + if (cmpr >= cmpr_end) + break; + + ch = *cmpr++; + } + } + + /* return the size of uncompressed data */ + return up - unc; +} + +/* + * 0 - standard compression + * !0 - best compression, requires a lot of cpu + */ +struct lznt *get_lznt_ctx(int level) +{ + struct lznt *r = ntfs_zalloc(level ? offsetof(struct lznt, hash) + : sizeof(struct lznt)); + + if (r) + r->std = !level; + return r; +} + +/* + * compress_lznt + * + * Compresses "unc" into "cmpr" + * +x - ok, 'cmpr' contains 'final_compressed_size' bytes of compressed data + * 0 - input buffer is full zero + */ +size_t compress_lznt(const void *unc, size_t unc_size, void *cmpr, + size_t cmpr_size, struct lznt *ctx) +{ + int err; + size_t (*match)(const u8 *src, struct lznt *ctx); + u8 *p = cmpr; + u8 *end = p + cmpr_size; + const u8 *unc_chunk = unc; + const u8 *unc_end = unc_chunk + unc_size; + bool is_zero = true; + + if (ctx->std) { + match = &longest_match_std; + memset(ctx->hash, 0, sizeof(ctx->hash)); + } else { + match = &longest_match_best; + } + + /* compression cycle */ + for (; unc_chunk < unc_end; unc_chunk += LZNT_CHUNK_SIZE) { + cmpr_size = 0; + err = compress_chunk(match, unc_chunk, unc_end, p, end, + &cmpr_size, ctx); + if (err < 0) + return unc_size; + + if (is_zero && err != LZNT_ERROR_ALL_ZEROS) + is_zero = false; + + p += cmpr_size; + } + + if (p <= end - 2) + p[0] = p[1] = 0; + + return is_zero ? 0 : PtrOffset(cmpr, p); +} + +/* + * decompress_lznt + * + * decompresses "cmpr" into "unc" + */ +ssize_t decompress_lznt(const void *cmpr, size_t cmpr_size, void *unc, + size_t unc_size) +{ + const u8 *cmpr_chunk = cmpr; + const u8 *cmpr_end = cmpr_chunk + cmpr_size; + u8 *unc_chunk = unc; + u8 *unc_end = unc_chunk + unc_size; + u16 chunk_hdr; + + if (cmpr_size < sizeof(short)) + return -EINVAL; + + /* read chunk header */ + chunk_hdr = cmpr_chunk[1]; + chunk_hdr <<= 8; + chunk_hdr |= cmpr_chunk[0]; + + /* loop through decompressing chunks */ + for (;;) { + size_t chunk_size_saved; + size_t unc_use; + size_t cmpr_use = 3 + (chunk_hdr & (LZNT_CHUNK_SIZE - 1)); + + /* Check that the chunk actually fits the supplied buffer */ + if (cmpr_chunk + cmpr_use > cmpr_end) + return -EINVAL; + + /* First make sure the chunk contains compressed data */ + if (chunk_hdr & 0x8000) { + /* Decompress a chunk and return if we get an error */ + ssize_t err = + decompress_chunk(unc_chunk, unc_end, + cmpr_chunk + sizeof(chunk_hdr), + cmpr_chunk + cmpr_use); + if (err < 0) + return err; + unc_use = err; + } else { + /* This chunk does not contain compressed data */ + unc_use = unc_chunk + LZNT_CHUNK_SIZE > unc_end + ? unc_end - unc_chunk + : LZNT_CHUNK_SIZE; + + if (cmpr_chunk + sizeof(chunk_hdr) + unc_use > + cmpr_end) { + return -EINVAL; + } + + memcpy(unc_chunk, cmpr_chunk + sizeof(chunk_hdr), + unc_use); + } + + /* Advance pointers */ + cmpr_chunk += cmpr_use; + unc_chunk += unc_use; + + /* Check for the end of unc buffer */ + if (unc_chunk >= unc_end) + break; + + /* Proceed the next chunk */ + if (cmpr_chunk > cmpr_end - 2) + break; + + chunk_size_saved = LZNT_CHUNK_SIZE; + + /* read chunk header */ + chunk_hdr = cmpr_chunk[1]; + chunk_hdr <<= 8; + chunk_hdr |= cmpr_chunk[0]; + + if (!chunk_hdr) + break; + + /* Check the size of unc buffer */ + if (unc_use < chunk_size_saved) { + size_t t1 = chunk_size_saved - unc_use; + u8 *t2 = unc_chunk + t1; + + /* 'Zero' memory */ + if (t2 >= unc_end) + break; + + memset(unc_chunk, 0, t1); + unc_chunk = t2; + } + } + + /* Check compression boundary */ + if (cmpr_chunk > cmpr_end) + return -EINVAL; + + /* + * The unc size is just a difference between current + * pointer and original one + */ + return PtrOffset(unc, unc_chunk); +} -- 2.30.0

1 0

[PATCH OLK-5.10 005/107] fs/ntfs3: Add attrib operations
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> mainline inclusion from mainline-v5.15-rc1 commit be71b5cba2e6485e8959da7a9f9a44461a1bb074 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- This adds attrib operations Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/attrib.c | 2096 +++++++++++++++++++++++++++++++++++++++++++ fs/ntfs3/attrlist.c | 456 ++++++++++ fs/ntfs3/xattr.c | 1128 +++++++++++++++++++++++ 3 files changed, 3680 insertions(+) create mode 100644 fs/ntfs3/attrib.c create mode 100644 fs/ntfs3/attrlist.c create mode 100644 fs/ntfs3/xattr.c diff --git a/fs/ntfs3/attrib.c b/fs/ntfs3/attrib.c new file mode 100644 index 000000000000..046dc57f75f2 --- /dev/null +++ b/fs/ntfs3/attrib.c @@ -0,0 +1,2096 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * + * Copyright (C) 2019-2021 Paragon Software GmbH, All rights reserved. + * + * TODO: merge attr_set_size/attr_data_get_block/attr_allocate_frame? + */ + +#include <linux/blkdev.h> +#include <linux/buffer_head.h> +#include <linux/fs.h> +#include <linux/hash.h> +#include <linux/nls.h> +#include <linux/ratelimit.h> +#include <linux/slab.h> + +#include "debug.h" +#include "ntfs.h" +#include "ntfs_fs.h" + +/* + * You can set external NTFS_MIN_LOG2_OF_CLUMP/NTFS_MAX_LOG2_OF_CLUMP to manage + * preallocate algorithm + */ +#ifndef NTFS_MIN_LOG2_OF_CLUMP +#define NTFS_MIN_LOG2_OF_CLUMP 16 +#endif + +#ifndef NTFS_MAX_LOG2_OF_CLUMP +#define NTFS_MAX_LOG2_OF_CLUMP 26 +#endif + +// 16M +#define NTFS_CLUMP_MIN (1 << (NTFS_MIN_LOG2_OF_CLUMP + 8)) +// 16G +#define NTFS_CLUMP_MAX (1ull << (NTFS_MAX_LOG2_OF_CLUMP + 8)) + +/* + * get_pre_allocated + * + */ +static inline u64 get_pre_allocated(u64 size) +{ + u32 clump; + u8 align_shift; + u64 ret; + + if (size <= NTFS_CLUMP_MIN) { + clump = 1 << NTFS_MIN_LOG2_OF_CLUMP; + align_shift = NTFS_MIN_LOG2_OF_CLUMP; + } else if (size >= NTFS_CLUMP_MAX) { + clump = 1 << NTFS_MAX_LOG2_OF_CLUMP; + align_shift = NTFS_MAX_LOG2_OF_CLUMP; + } else { + align_shift = NTFS_MIN_LOG2_OF_CLUMP - 1 + + __ffs(size >> (8 + NTFS_MIN_LOG2_OF_CLUMP)); + clump = 1u << align_shift; + } + + ret = (((size + clump - 1) >> align_shift)) << align_shift; + + return ret; +} + +/* + * attr_must_be_resident + * + * returns true if attribute must be resident + */ +static inline bool attr_must_be_resident(struct ntfs_sb_info *sbi, + enum ATTR_TYPE type) +{ + const struct ATTR_DEF_ENTRY *de; + + switch (type) { + case ATTR_STD: + case ATTR_NAME: + case ATTR_ID: + case ATTR_LABEL: + case ATTR_VOL_INFO: + case ATTR_ROOT: + case ATTR_EA_INFO: + return true; + default: + de = ntfs_query_def(sbi, type); + if (de && (de->flags & NTFS_ATTR_MUST_BE_RESIDENT)) + return true; + return false; + } +} + +/* + * attr_load_runs + * + * load all runs stored in 'attr' + */ +int attr_load_runs(struct ATTRIB *attr, struct ntfs_inode *ni, + struct runs_tree *run, const CLST *vcn) +{ + int err; + CLST svcn = le64_to_cpu(attr->nres.svcn); + CLST evcn = le64_to_cpu(attr->nres.evcn); + u32 asize; + u16 run_off; + + if (svcn >= evcn + 1 || run_is_mapped_full(run, svcn, evcn)) + return 0; + + if (vcn && (evcn < *vcn || *vcn < svcn)) + return -EINVAL; + + asize = le32_to_cpu(attr->size); + run_off = le16_to_cpu(attr->nres.run_off); + err = run_unpack_ex(run, ni->mi.sbi, ni->mi.rno, svcn, evcn, + vcn ? *vcn : svcn, Add2Ptr(attr, run_off), + asize - run_off); + if (err < 0) + return err; + + return 0; +} + +/* + * int run_deallocate_ex + * + * Deallocate clusters + */ +static int run_deallocate_ex(struct ntfs_sb_info *sbi, struct runs_tree *run, + CLST vcn, CLST len, CLST *done, bool trim) +{ + int err = 0; + CLST vcn_next, vcn0 = vcn, lcn, clen, dn = 0; + size_t idx; + + if (!len) + goto out; + + if (!run_lookup_entry(run, vcn, &lcn, &clen, &idx)) { +failed: + run_truncate(run, vcn0); + err = -EINVAL; + goto out; + } + + for (;;) { + if (clen > len) + clen = len; + + if (!clen) { + err = -EINVAL; + goto out; + } + + if (lcn != SPARSE_LCN) { + mark_as_free_ex(sbi, lcn, clen, trim); + dn += clen; + } + + len -= clen; + if (!len) + break; + + vcn_next = vcn + clen; + if (!run_get_entry(run, ++idx, &vcn, &lcn, &clen) || + vcn != vcn_next) { + // save memory - don't load entire run + goto failed; + } + } + +out: + if (done) + *done += dn; + + return err; +} + +/* + * attr_allocate_clusters + * + * find free space, mark it as used and store in 'run' + */ +int attr_allocate_clusters(struct ntfs_sb_info *sbi, struct runs_tree *run, + CLST vcn, CLST lcn, CLST len, CLST *pre_alloc, + enum ALLOCATE_OPT opt, CLST *alen, const size_t fr, + CLST *new_lcn) +{ + int err; + CLST flen, vcn0 = vcn, pre = pre_alloc ? *pre_alloc : 0; + struct wnd_bitmap *wnd = &sbi->used.bitmap; + size_t cnt = run->count; + + for (;;) { + err = ntfs_look_for_free_space(sbi, lcn, len + pre, &lcn, &flen, + opt); + + if (err == -ENOSPC && pre) { + pre = 0; + if (*pre_alloc) + *pre_alloc = 0; + continue; + } + + if (err) + goto out; + + if (new_lcn && vcn == vcn0) + *new_lcn = lcn; + + /* Add new fragment into run storage */ + if (!run_add_entry(run, vcn, lcn, flen, opt == ALLOCATE_MFT)) { + down_write_nested(&wnd->rw_lock, BITMAP_MUTEX_CLUSTERS); + wnd_set_free(wnd, lcn, flen); + up_write(&wnd->rw_lock); + err = -ENOMEM; + goto out; + } + + vcn += flen; + + if (flen >= len || opt == ALLOCATE_MFT || + (fr && run->count - cnt >= fr)) { + *alen = vcn - vcn0; + return 0; + } + + len -= flen; + } + +out: + /* undo */ + run_deallocate_ex(sbi, run, vcn0, vcn - vcn0, NULL, false); + run_truncate(run, vcn0); + + return err; +} + +/* + * if page is not NULL - it is already contains resident data + * and locked (called from ni_write_frame) + */ +int attr_make_nonresident(struct ntfs_inode *ni, struct ATTRIB *attr, + struct ATTR_LIST_ENTRY *le, struct mft_inode *mi, + u64 new_size, struct runs_tree *run, + struct ATTRIB **ins_attr, struct page *page) +{ + struct ntfs_sb_info *sbi; + struct ATTRIB *attr_s; + struct MFT_REC *rec; + u32 used, asize, rsize, aoff, align; + bool is_data; + CLST len, alen; + char *next; + int err; + + if (attr->non_res) { + *ins_attr = attr; + return 0; + } + + sbi = mi->sbi; + rec = mi->mrec; + attr_s = NULL; + used = le32_to_cpu(rec->used); + asize = le32_to_cpu(attr->size); + next = Add2Ptr(attr, asize); + aoff = PtrOffset(rec, attr); + rsize = le32_to_cpu(attr->res.data_size); + is_data = attr->type == ATTR_DATA && !attr->name_len; + + align = sbi->cluster_size; + if (is_attr_compressed(attr)) + align <<= COMPRESSION_UNIT; + len = (rsize + align - 1) >> sbi->cluster_bits; + + run_init(run); + + /* make a copy of original attribute */ + attr_s = ntfs_memdup(attr, asize); + if (!attr_s) { + err = -ENOMEM; + goto out; + } + + if (!len) { + /* empty resident -> empty nonresident */ + alen = 0; + } else { + const char *data = resident_data(attr); + + err = attr_allocate_clusters(sbi, run, 0, 0, len, NULL, + ALLOCATE_DEF, &alen, 0, NULL); + if (err) + goto out1; + + if (!rsize) { + /* empty resident -> non empty nonresident */ + } else if (!is_data) { + err = ntfs_sb_write_run(sbi, run, 0, data, rsize); + if (err) + goto out2; + } else if (!page) { + char *kaddr; + + page = grab_cache_page(ni->vfs_inode.i_mapping, 0); + if (!page) { + err = -ENOMEM; + goto out2; + } + kaddr = kmap_atomic(page); + memcpy(kaddr, data, rsize); + memset(kaddr + rsize, 0, PAGE_SIZE - rsize); + kunmap_atomic(kaddr); + flush_dcache_page(page); + SetPageUptodate(page); + set_page_dirty(page); + unlock_page(page); + put_page(page); + } + } + + /* remove original attribute */ + used -= asize; + memmove(attr, Add2Ptr(attr, asize), used - aoff); + rec->used = cpu_to_le32(used); + mi->dirty = true; + if (le) + al_remove_le(ni, le); + + err = ni_insert_nonresident(ni, attr_s->type, attr_name(attr_s), + attr_s->name_len, run, 0, alen, + attr_s->flags, &attr, NULL); + if (err) + goto out3; + + ntfs_free(attr_s); + attr->nres.data_size = cpu_to_le64(rsize); + attr->nres.valid_size = attr->nres.data_size; + + *ins_attr = attr; + + if (is_data) + ni->ni_flags &= ~NI_FLAG_RESIDENT; + + /* Resident attribute becomes non resident */ + return 0; + +out3: + attr = Add2Ptr(rec, aoff); + memmove(next, attr, used - aoff); + memcpy(attr, attr_s, asize); + rec->used = cpu_to_le32(used + asize); + mi->dirty = true; +out2: + /* undo: do not trim new allocated clusters */ + run_deallocate(sbi, run, false); + run_close(run); +out1: + ntfs_free(attr_s); + /*reinsert le*/ +out: + return err; +} + +/* + * attr_set_size_res + * + * helper for attr_set_size + */ +static int attr_set_size_res(struct ntfs_inode *ni, struct ATTRIB *attr, + struct ATTR_LIST_ENTRY *le, struct mft_inode *mi, + u64 new_size, struct runs_tree *run, + struct ATTRIB **ins_attr) +{ + struct ntfs_sb_info *sbi = mi->sbi; + struct MFT_REC *rec = mi->mrec; + u32 used = le32_to_cpu(rec->used); + u32 asize = le32_to_cpu(attr->size); + u32 aoff = PtrOffset(rec, attr); + u32 rsize = le32_to_cpu(attr->res.data_size); + u32 tail = used - aoff - asize; + char *next = Add2Ptr(attr, asize); + s64 dsize = QuadAlign(new_size) - QuadAlign(rsize); + + if (dsize < 0) { + memmove(next + dsize, next, tail); + } else if (dsize > 0) { + if (used + dsize > sbi->max_bytes_per_attr) + return attr_make_nonresident(ni, attr, le, mi, new_size, + run, ins_attr, NULL); + + memmove(next + dsize, next, tail); + memset(next, 0, dsize); + } + + if (new_size > rsize) + memset(Add2Ptr(resident_data(attr), rsize), 0, + new_size - rsize); + + rec->used = cpu_to_le32(used + dsize); + attr->size = cpu_to_le32(asize + dsize); + attr->res.data_size = cpu_to_le32(new_size); + mi->dirty = true; + *ins_attr = attr; + + return 0; +} + +/* + * attr_set_size + * + * change the size of attribute + * Extend: + * - sparse/compressed: no allocated clusters + * - normal: append allocated and preallocated new clusters + * Shrink: + * - no deallocate if keep_prealloc is set + */ +int attr_set_size(struct ntfs_inode *ni, enum ATTR_TYPE type, + const __le16 *name, u8 name_len, struct runs_tree *run, + u64 new_size, const u64 *new_valid, bool keep_prealloc, + struct ATTRIB **ret) +{ + int err = 0; + struct ntfs_sb_info *sbi = ni->mi.sbi; + u8 cluster_bits = sbi->cluster_bits; + bool is_mft = + ni->mi.rno == MFT_REC_MFT && type == ATTR_DATA && !name_len; + u64 old_valid, old_size, old_alloc, new_alloc, new_alloc_tmp; + struct ATTRIB *attr = NULL, *attr_b; + struct ATTR_LIST_ENTRY *le, *le_b; + struct mft_inode *mi, *mi_b; + CLST alen, vcn, lcn, new_alen, old_alen, svcn, evcn; + CLST next_svcn, pre_alloc = -1, done = 0; + bool is_ext; + u32 align; + struct MFT_REC *rec; + +again: + le_b = NULL; + attr_b = ni_find_attr(ni, NULL, &le_b, type, name, name_len, NULL, + &mi_b); + if (!attr_b) { + err = -ENOENT; + goto out; + } + + if (!attr_b->non_res) { + err = attr_set_size_res(ni, attr_b, le_b, mi_b, new_size, run, + &attr_b); + if (err || !attr_b->non_res) + goto out; + + /* layout of records may be changed, so do a full search */ + goto again; + } + + is_ext = is_attr_ext(attr_b); + +again_1: + align = sbi->cluster_size; + + if (is_ext) { + align <<= attr_b->nres.c_unit; + if (is_attr_sparsed(attr_b)) + keep_prealloc = false; + } + + old_valid = le64_to_cpu(attr_b->nres.valid_size); + old_size = le64_to_cpu(attr_b->nres.data_size); + old_alloc = le64_to_cpu(attr_b->nres.alloc_size); + old_alen = old_alloc >> cluster_bits; + + new_alloc = (new_size + align - 1) & ~(u64)(align - 1); + new_alen = new_alloc >> cluster_bits; + + if (keep_prealloc && is_ext) + keep_prealloc = false; + + if (keep_prealloc && new_size < old_size) { + attr_b->nres.data_size = cpu_to_le64(new_size); + mi_b->dirty = true; + goto ok; + } + + vcn = old_alen - 1; + + svcn = le64_to_cpu(attr_b->nres.svcn); + evcn = le64_to_cpu(attr_b->nres.evcn); + + if (svcn <= vcn && vcn <= evcn) { + attr = attr_b; + le = le_b; + mi = mi_b; + } else if (!le_b) { + err = -EINVAL; + goto out; + } else { + le = le_b; + attr = ni_find_attr(ni, attr_b, &le, type, name, name_len, &vcn, + &mi); + if (!attr) { + err = -EINVAL; + goto out; + } + +next_le_1: + svcn = le64_to_cpu(attr->nres.svcn); + evcn = le64_to_cpu(attr->nres.evcn); + } + +next_le: + rec = mi->mrec; + + err = attr_load_runs(attr, ni, run, NULL); + if (err) + goto out; + + if (new_size > old_size) { + CLST to_allocate; + size_t free; + + if (new_alloc <= old_alloc) { + attr_b->nres.data_size = cpu_to_le64(new_size); + mi_b->dirty = true; + goto ok; + } + + to_allocate = new_alen - old_alen; +add_alloc_in_same_attr_seg: + lcn = 0; + if (is_mft) { + /* mft allocates clusters from mftzone */ + pre_alloc = 0; + } else if (is_ext) { + /* no preallocate for sparse/compress */ + pre_alloc = 0; + } else if (pre_alloc == -1) { + pre_alloc = 0; + if (type == ATTR_DATA && !name_len && + sbi->options.prealloc) { + CLST new_alen2 = bytes_to_cluster( + sbi, get_pre_allocated(new_size)); + pre_alloc = new_alen2 - new_alen; + } + + /* Get the last lcn to allocate from */ + if (old_alen && + !run_lookup_entry(run, vcn, &lcn, NULL, NULL)) { + lcn = SPARSE_LCN; + } + + if (lcn == SPARSE_LCN) + lcn = 0; + else if (lcn) + lcn += 1; + + free = wnd_zeroes(&sbi->used.bitmap); + if (to_allocate > free) { + err = -ENOSPC; + goto out; + } + + if (pre_alloc && to_allocate + pre_alloc > free) + pre_alloc = 0; + } + + vcn = old_alen; + + if (is_ext) { + if (!run_add_entry(run, vcn, SPARSE_LCN, to_allocate, + false)) { + err = -ENOMEM; + goto out; + } + alen = to_allocate; + } else { + /* ~3 bytes per fragment */ + err = attr_allocate_clusters( + sbi, run, vcn, lcn, to_allocate, &pre_alloc, + is_mft ? ALLOCATE_MFT : 0, &alen, + is_mft ? 0 + : (sbi->record_size - + le32_to_cpu(rec->used) + 8) / + 3 + + 1, + NULL); + if (err) + goto out; + } + + done += alen; + vcn += alen; + if (to_allocate > alen) + to_allocate -= alen; + else + to_allocate = 0; + +pack_runs: + err = mi_pack_runs(mi, attr, run, vcn - svcn); + if (err) + goto out; + + next_svcn = le64_to_cpu(attr->nres.evcn) + 1; + new_alloc_tmp = (u64)next_svcn << cluster_bits; + attr_b->nres.alloc_size = cpu_to_le64(new_alloc_tmp); + mi_b->dirty = true; + + if (next_svcn >= vcn && !to_allocate) { + /* Normal way. update attribute and exit */ + attr_b->nres.data_size = cpu_to_le64(new_size); + goto ok; + } + + /* at least two mft to avoid recursive loop*/ + if (is_mft && next_svcn == vcn && + ((u64)done << sbi->cluster_bits) >= 2 * sbi->record_size) { + new_size = new_alloc_tmp; + attr_b->nres.data_size = attr_b->nres.alloc_size; + goto ok; + } + + if (le32_to_cpu(rec->used) < sbi->record_size) { + old_alen = next_svcn; + evcn = old_alen - 1; + goto add_alloc_in_same_attr_seg; + } + + attr_b->nres.data_size = attr_b->nres.alloc_size; + if (new_alloc_tmp < old_valid) + attr_b->nres.valid_size = attr_b->nres.data_size; + + if (type == ATTR_LIST) { + err = ni_expand_list(ni); + if (err) + goto out; + if (next_svcn < vcn) + goto pack_runs; + + /* layout of records is changed */ + goto again; + } + + if (!ni->attr_list.size) { + err = ni_create_attr_list(ni); + if (err) + goto out; + /* layout of records is changed */ + } + + if (next_svcn >= vcn) { + /* this is mft data, repeat */ + goto again; + } + + /* insert new attribute segment */ + err = ni_insert_nonresident(ni, type, name, name_len, run, + next_svcn, vcn - next_svcn, + attr_b->flags, &attr, &mi); + if (err) + goto out; + + if (!is_mft) + run_truncate_head(run, evcn + 1); + + svcn = le64_to_cpu(attr->nres.svcn); + evcn = le64_to_cpu(attr->nres.evcn); + + le_b = NULL; + /* layout of records maybe changed */ + /* find base attribute to update*/ + attr_b = ni_find_attr(ni, NULL, &le_b, type, name, name_len, + NULL, &mi_b); + if (!attr_b) { + err = -ENOENT; + goto out; + } + + attr_b->nres.alloc_size = cpu_to_le64((u64)vcn << cluster_bits); + attr_b->nres.data_size = attr_b->nres.alloc_size; + attr_b->nres.valid_size = attr_b->nres.alloc_size; + mi_b->dirty = true; + goto again_1; + } + + if (new_size != old_size || + (new_alloc != old_alloc && !keep_prealloc)) { + vcn = max(svcn, new_alen); + new_alloc_tmp = (u64)vcn << cluster_bits; + + alen = 0; + err = run_deallocate_ex(sbi, run, vcn, evcn - vcn + 1, &alen, + true); + if (err) + goto out; + + run_truncate(run, vcn); + + if (vcn > svcn) { + err = mi_pack_runs(mi, attr, run, vcn - svcn); + if (err) + goto out; + } else if (le && le->vcn) { + u16 le_sz = le16_to_cpu(le->size); + + /* + * NOTE: list entries for one attribute are always + * the same size. We deal with last entry (vcn==0) + * and it is not first in entries array + * (list entry for std attribute always first) + * So it is safe to step back + */ + mi_remove_attr(mi, attr); + + if (!al_remove_le(ni, le)) { + err = -EINVAL; + goto out; + } + + le = (struct ATTR_LIST_ENTRY *)((u8 *)le - le_sz); + } else { + attr->nres.evcn = cpu_to_le64((u64)vcn - 1); + mi->dirty = true; + } + + attr_b->nres.alloc_size = cpu_to_le64(new_alloc_tmp); + + if (vcn == new_alen) { + attr_b->nres.data_size = cpu_to_le64(new_size); + if (new_size < old_valid) + attr_b->nres.valid_size = + attr_b->nres.data_size; + } else { + if (new_alloc_tmp <= + le64_to_cpu(attr_b->nres.data_size)) + attr_b->nres.data_size = + attr_b->nres.alloc_size; + if (new_alloc_tmp < + le64_to_cpu(attr_b->nres.valid_size)) + attr_b->nres.valid_size = + attr_b->nres.alloc_size; + } + + if (is_ext) + le64_sub_cpu(&attr_b->nres.total_size, + ((u64)alen << cluster_bits)); + + mi_b->dirty = true; + + if (new_alloc_tmp <= new_alloc) + goto ok; + + old_size = new_alloc_tmp; + vcn = svcn - 1; + + if (le == le_b) { + attr = attr_b; + mi = mi_b; + evcn = svcn - 1; + svcn = 0; + goto next_le; + } + + if (le->type != type || le->name_len != name_len || + memcmp(le_name(le), name, name_len * sizeof(short))) { + err = -EINVAL; + goto out; + } + + err = ni_load_mi(ni, le, &mi); + if (err) + goto out; + + attr = mi_find_attr(mi, NULL, type, name, name_len, &le->id); + if (!attr) { + err = -EINVAL; + goto out; + } + goto next_le_1; + } + +ok: + if (new_valid) { + __le64 valid = cpu_to_le64(min(*new_valid, new_size)); + + if (attr_b->nres.valid_size != valid) { + attr_b->nres.valid_size = valid; + mi_b->dirty = true; + } + } + +out: + if (!err && attr_b && ret) + *ret = attr_b; + + /* update inode_set_bytes*/ + if (!err && ((type == ATTR_DATA && !name_len) || + (type == ATTR_ALLOC && name == I30_NAME))) { + bool dirty = false; + + if (ni->vfs_inode.i_size != new_size) { + ni->vfs_inode.i_size = new_size; + dirty = true; + } + + if (attr_b && attr_b->non_res) { + new_alloc = le64_to_cpu(attr_b->nres.alloc_size); + if (inode_get_bytes(&ni->vfs_inode) != new_alloc) { + inode_set_bytes(&ni->vfs_inode, new_alloc); + dirty = true; + } + } + + if (dirty) { + ni->ni_flags |= NI_FLAG_UPDATE_PARENT; + mark_inode_dirty(&ni->vfs_inode); + } + } + + return err; +} + +int attr_data_get_block(struct ntfs_inode *ni, CLST vcn, CLST clen, CLST *lcn, + CLST *len, bool *new) +{ + int err = 0; + struct runs_tree *run = &ni->file.run; + struct ntfs_sb_info *sbi; + u8 cluster_bits; + struct ATTRIB *attr = NULL, *attr_b; + struct ATTR_LIST_ENTRY *le, *le_b; + struct mft_inode *mi, *mi_b; + CLST hint, svcn, to_alloc, evcn1, next_svcn, asize, end; + u64 total_size; + u32 clst_per_frame; + bool ok; + + if (new) + *new = false; + + down_read(&ni->file.run_lock); + ok = run_lookup_entry(run, vcn, lcn, len, NULL); + up_read(&ni->file.run_lock); + + if (ok && (*lcn != SPARSE_LCN || !new)) { + /* normal way */ + return 0; + } + + if (!clen) + clen = 1; + + if (ok && clen > *len) + clen = *len; + + sbi = ni->mi.sbi; + cluster_bits = sbi->cluster_bits; + + ni_lock(ni); + down_write(&ni->file.run_lock); + + le_b = NULL; + attr_b = ni_find_attr(ni, NULL, &le_b, ATTR_DATA, NULL, 0, NULL, &mi_b); + if (!attr_b) { + err = -ENOENT; + goto out; + } + + if (!attr_b->non_res) { + *lcn = RESIDENT_LCN; + *len = 1; + goto out; + } + + asize = le64_to_cpu(attr_b->nres.alloc_size) >> sbi->cluster_bits; + if (vcn >= asize) { + err = -EINVAL; + goto out; + } + + clst_per_frame = 1u << attr_b->nres.c_unit; + to_alloc = (clen + clst_per_frame - 1) & ~(clst_per_frame - 1); + + if (vcn + to_alloc > asize) + to_alloc = asize - vcn; + + svcn = le64_to_cpu(attr_b->nres.svcn); + evcn1 = le64_to_cpu(attr_b->nres.evcn) + 1; + + attr = attr_b; + le = le_b; + mi = mi_b; + + if (le_b && (vcn < svcn || evcn1 <= vcn)) { + attr = ni_find_attr(ni, attr_b, &le, ATTR_DATA, NULL, 0, &vcn, + &mi); + if (!attr) { + err = -EINVAL; + goto out; + } + svcn = le64_to_cpu(attr->nres.svcn); + evcn1 = le64_to_cpu(attr->nres.evcn) + 1; + } + + err = attr_load_runs(attr, ni, run, NULL); + if (err) + goto out; + + if (!ok) { + ok = run_lookup_entry(run, vcn, lcn, len, NULL); + if (ok && (*lcn != SPARSE_LCN || !new)) { + /* normal way */ + err = 0; + goto ok; + } + + if (!ok && !new) { + *len = 0; + err = 0; + goto ok; + } + + if (ok && clen > *len) { + clen = *len; + to_alloc = (clen + clst_per_frame - 1) & + ~(clst_per_frame - 1); + } + } + + if (!is_attr_ext(attr_b)) { + err = -EINVAL; + goto out; + } + + /* Get the last lcn to allocate from */ + hint = 0; + + if (vcn > evcn1) { + if (!run_add_entry(run, evcn1, SPARSE_LCN, vcn - evcn1, + false)) { + err = -ENOMEM; + goto out; + } + } else if (vcn && !run_lookup_entry(run, vcn - 1, &hint, NULL, NULL)) { + hint = -1; + } + + err = attr_allocate_clusters( + sbi, run, vcn, hint + 1, to_alloc, NULL, 0, len, + (sbi->record_size - le32_to_cpu(mi->mrec->used) + 8) / 3 + 1, + lcn); + if (err) + goto out; + *new = true; + + end = vcn + *len; + + total_size = le64_to_cpu(attr_b->nres.total_size) + + ((u64)*len << cluster_bits); + +repack: + err = mi_pack_runs(mi, attr, run, max(end, evcn1) - svcn); + if (err) + goto out; + + attr_b->nres.total_size = cpu_to_le64(total_size); + inode_set_bytes(&ni->vfs_inode, total_size); + ni->ni_flags |= NI_FLAG_UPDATE_PARENT; + + mi_b->dirty = true; + mark_inode_dirty(&ni->vfs_inode); + + /* stored [vcn : next_svcn) from [vcn : end) */ + next_svcn = le64_to_cpu(attr->nres.evcn) + 1; + + if (end <= evcn1) { + if (next_svcn == evcn1) { + /* Normal way. update attribute and exit */ + goto ok; + } + /* add new segment [next_svcn : evcn1 - next_svcn )*/ + if (!ni->attr_list.size) { + err = ni_create_attr_list(ni); + if (err) + goto out; + /* layout of records is changed */ + le_b = NULL; + attr_b = ni_find_attr(ni, NULL, &le_b, ATTR_DATA, NULL, + 0, NULL, &mi_b); + if (!attr_b) { + err = -ENOENT; + goto out; + } + + attr = attr_b; + le = le_b; + mi = mi_b; + goto repack; + } + } + + svcn = evcn1; + + /* Estimate next attribute */ + attr = ni_find_attr(ni, attr, &le, ATTR_DATA, NULL, 0, &svcn, &mi); + + if (attr) { + CLST alloc = bytes_to_cluster( + sbi, le64_to_cpu(attr_b->nres.alloc_size)); + CLST evcn = le64_to_cpu(attr->nres.evcn); + + if (end < next_svcn) + end = next_svcn; + while (end > evcn) { + /* remove segment [svcn : evcn)*/ + mi_remove_attr(mi, attr); + + if (!al_remove_le(ni, le)) { + err = -EINVAL; + goto out; + } + + if (evcn + 1 >= alloc) { + /* last attribute segment */ + evcn1 = evcn + 1; + goto ins_ext; + } + + if (ni_load_mi(ni, le, &mi)) { + attr = NULL; + goto out; + } + + attr = mi_find_attr(mi, NULL, ATTR_DATA, NULL, 0, + &le->id); + if (!attr) { + err = -EINVAL; + goto out; + } + svcn = le64_to_cpu(attr->nres.svcn); + evcn = le64_to_cpu(attr->nres.evcn); + } + + if (end < svcn) + end = svcn; + + err = attr_load_runs(attr, ni, run, &end); + if (err) + goto out; + + evcn1 = evcn + 1; + attr->nres.svcn = cpu_to_le64(next_svcn); + err = mi_pack_runs(mi, attr, run, evcn1 - next_svcn); + if (err) + goto out; + + le->vcn = cpu_to_le64(next_svcn); + ni->attr_list.dirty = true; + mi->dirty = true; + + next_svcn = le64_to_cpu(attr->nres.evcn) + 1; + } +ins_ext: + if (evcn1 > next_svcn) { + err = ni_insert_nonresident(ni, ATTR_DATA, NULL, 0, run, + next_svcn, evcn1 - next_svcn, + attr_b->flags, &attr, &mi); + if (err) + goto out; + } +ok: + run_truncate_around(run, vcn); +out: + up_write(&ni->file.run_lock); + ni_unlock(ni); + + return err; +} + +int attr_data_read_resident(struct ntfs_inode *ni, struct page *page) +{ + u64 vbo; + struct ATTRIB *attr; + u32 data_size; + + attr = ni_find_attr(ni, NULL, NULL, ATTR_DATA, NULL, 0, NULL, NULL); + if (!attr) + return -EINVAL; + + if (attr->non_res) + return E_NTFS_NONRESIDENT; + + vbo = page->index << PAGE_SHIFT; + data_size = le32_to_cpu(attr->res.data_size); + if (vbo < data_size) { + const char *data = resident_data(attr); + char *kaddr = kmap_atomic(page); + u32 use = data_size - vbo; + + if (use > PAGE_SIZE) + use = PAGE_SIZE; + + memcpy(kaddr, data + vbo, use); + memset(kaddr + use, 0, PAGE_SIZE - use); + kunmap_atomic(kaddr); + flush_dcache_page(page); + SetPageUptodate(page); + } else if (!PageUptodate(page)) { + zero_user_segment(page, 0, PAGE_SIZE); + SetPageUptodate(page); + } + + return 0; +} + +int attr_data_write_resident(struct ntfs_inode *ni, struct page *page) +{ + u64 vbo; + struct mft_inode *mi; + struct ATTRIB *attr; + u32 data_size; + + attr = ni_find_attr(ni, NULL, NULL, ATTR_DATA, NULL, 0, NULL, &mi); + if (!attr) + return -EINVAL; + + if (attr->non_res) { + /*return special error code to check this case*/ + return E_NTFS_NONRESIDENT; + } + + vbo = page->index << PAGE_SHIFT; + data_size = le32_to_cpu(attr->res.data_size); + if (vbo < data_size) { + char *data = resident_data(attr); + char *kaddr = kmap_atomic(page); + u32 use = data_size - vbo; + + if (use > PAGE_SIZE) + use = PAGE_SIZE; + memcpy(data + vbo, kaddr, use); + kunmap_atomic(kaddr); + mi->dirty = true; + } + ni->i_valid = data_size; + + return 0; +} + +/* + * attr_load_runs_vcn + * + * load runs with vcn + */ +int attr_load_runs_vcn(struct ntfs_inode *ni, enum ATTR_TYPE type, + const __le16 *name, u8 name_len, struct runs_tree *run, + CLST vcn) +{ + struct ATTRIB *attr; + int err; + CLST svcn, evcn; + u16 ro; + + attr = ni_find_attr(ni, NULL, NULL, type, name, name_len, &vcn, NULL); + if (!attr) + return -ENOENT; + + svcn = le64_to_cpu(attr->nres.svcn); + evcn = le64_to_cpu(attr->nres.evcn); + + if (evcn < vcn || vcn < svcn) + return -EINVAL; + + ro = le16_to_cpu(attr->nres.run_off); + err = run_unpack_ex(run, ni->mi.sbi, ni->mi.rno, svcn, evcn, svcn, + Add2Ptr(attr, ro), le32_to_cpu(attr->size) - ro); + if (err < 0) + return err; + return 0; +} + +/* + * load runs for given range [from to) + */ +int attr_load_runs_range(struct ntfs_inode *ni, enum ATTR_TYPE type, + const __le16 *name, u8 name_len, struct runs_tree *run, + u64 from, u64 to) +{ + struct ntfs_sb_info *sbi = ni->mi.sbi; + u8 cluster_bits = sbi->cluster_bits; + CLST vcn = from >> cluster_bits; + CLST vcn_last = (to - 1) >> cluster_bits; + CLST lcn, clen; + int err; + + for (vcn = from >> cluster_bits; vcn <= vcn_last; vcn += clen) { + if (!run_lookup_entry(run, vcn, &lcn, &clen, NULL)) { + err = attr_load_runs_vcn(ni, type, name, name_len, run, + vcn); + if (err) + return err; + clen = 0; /*next run_lookup_entry(vcn) must be success*/ + } + } + + return 0; +} + +#ifdef CONFIG_NTFS3_LZX_XPRESS +/* + * attr_wof_frame_info + * + * read header of xpress/lzx file to get info about frame + */ +int attr_wof_frame_info(struct ntfs_inode *ni, struct ATTRIB *attr, + struct runs_tree *run, u64 frame, u64 frames, + u8 frame_bits, u32 *ondisk_size, u64 *vbo_data) +{ + struct ntfs_sb_info *sbi = ni->mi.sbi; + u64 vbo[2], off[2], wof_size; + u32 voff; + u8 bytes_per_off; + char *addr; + struct page *page; + int i, err; + __le32 *off32; + __le64 *off64; + + if (ni->vfs_inode.i_size < 0x100000000ull) { + /* file starts with array of 32 bit offsets */ + bytes_per_off = sizeof(__le32); + vbo[1] = frame << 2; + *vbo_data = frames << 2; + } else { + /* file starts with array of 64 bit offsets */ + bytes_per_off = sizeof(__le64); + vbo[1] = frame << 3; + *vbo_data = frames << 3; + } + + /* + * read 4/8 bytes at [vbo - 4(8)] == offset where compressed frame starts + * read 4/8 bytes at [vbo] == offset where compressed frame ends + */ + if (!attr->non_res) { + if (vbo[1] + bytes_per_off > le32_to_cpu(attr->res.data_size)) { + ntfs_inode_err(&ni->vfs_inode, "is corrupted"); + return -EINVAL; + } + addr = resident_data(attr); + + if (bytes_per_off == sizeof(__le32)) { + off32 = Add2Ptr(addr, vbo[1]); + off[0] = vbo[1] ? le32_to_cpu(off32[-1]) : 0; + off[1] = le32_to_cpu(off32[0]); + } else { + off64 = Add2Ptr(addr, vbo[1]); + off[0] = vbo[1] ? le64_to_cpu(off64[-1]) : 0; + off[1] = le64_to_cpu(off64[0]); + } + + *vbo_data += off[0]; + *ondisk_size = off[1] - off[0]; + return 0; + } + + wof_size = le64_to_cpu(attr->nres.data_size); + down_write(&ni->file.run_lock); + page = ni->file.offs_page; + if (!page) { + page = alloc_page(GFP_KERNEL); + if (!page) { + err = -ENOMEM; + goto out; + } + page->index = -1; + ni->file.offs_page = page; + } + lock_page(page); + addr = page_address(page); + + if (vbo[1]) { + voff = vbo[1] & (PAGE_SIZE - 1); + vbo[0] = vbo[1] - bytes_per_off; + i = 0; + } else { + voff = 0; + vbo[0] = 0; + off[0] = 0; + i = 1; + } + + do { + pgoff_t index = vbo[i] >> PAGE_SHIFT; + + if (index != page->index) { + u64 from = vbo[i] & ~(u64)(PAGE_SIZE - 1); + u64 to = min(from + PAGE_SIZE, wof_size); + + err = attr_load_runs_range(ni, ATTR_DATA, WOF_NAME, + ARRAY_SIZE(WOF_NAME), run, + from, to); + if (err) + goto out1; + + err = ntfs_bio_pages(sbi, run, &page, 1, from, + to - from, REQ_OP_READ); + if (err) { + page->index = -1; + goto out1; + } + page->index = index; + } + + if (i) { + if (bytes_per_off == sizeof(__le32)) { + off32 = Add2Ptr(addr, voff); + off[1] = le32_to_cpu(*off32); + } else { + off64 = Add2Ptr(addr, voff); + off[1] = le64_to_cpu(*off64); + } + } else if (!voff) { + if (bytes_per_off == sizeof(__le32)) { + off32 = Add2Ptr(addr, PAGE_SIZE - sizeof(u32)); + off[0] = le32_to_cpu(*off32); + } else { + off64 = Add2Ptr(addr, PAGE_SIZE - sizeof(u64)); + off[0] = le64_to_cpu(*off64); + } + } else { + /* two values in one page*/ + if (bytes_per_off == sizeof(__le32)) { + off32 = Add2Ptr(addr, voff); + off[0] = le32_to_cpu(off32[-1]); + off[1] = le32_to_cpu(off32[0]); + } else { + off64 = Add2Ptr(addr, voff); + off[0] = le64_to_cpu(off64[-1]); + off[1] = le64_to_cpu(off64[0]); + } + break; + } + } while (++i < 2); + + *vbo_data += off[0]; + *ondisk_size = off[1] - off[0]; + +out1: + unlock_page(page); +out: + up_write(&ni->file.run_lock); + return err; +} +#endif + +/* + * attr_is_frame_compressed + * + * This function is used to detect compressed frame + */ +int attr_is_frame_compressed(struct ntfs_inode *ni, struct ATTRIB *attr, + CLST frame, CLST *clst_data) +{ + int err; + u32 clst_frame; + CLST clen, lcn, vcn, alen, slen, vcn_next; + size_t idx; + struct runs_tree *run; + + *clst_data = 0; + + if (!is_attr_compressed(attr)) + return 0; + + if (!attr->non_res) + return 0; + + clst_frame = 1u << attr->nres.c_unit; + vcn = frame * clst_frame; + run = &ni->file.run; + + if (!run_lookup_entry(run, vcn, &lcn, &clen, &idx)) { + err = attr_load_runs_vcn(ni, attr->type, attr_name(attr), + attr->name_len, run, vcn); + if (err) + return err; + + if (!run_lookup_entry(run, vcn, &lcn, &clen, &idx)) + return -EINVAL; + } + + if (lcn == SPARSE_LCN) { + /* sparsed frame */ + return 0; + } + + if (clen >= clst_frame) { + /* + * The frame is not compressed 'cause + * it does not contain any sparse clusters + */ + *clst_data = clst_frame; + return 0; + } + + alen = bytes_to_cluster(ni->mi.sbi, le64_to_cpu(attr->nres.alloc_size)); + slen = 0; + *clst_data = clen; + + /* + * The frame is compressed if *clst_data + slen >= clst_frame + * Check next fragments + */ + while ((vcn += clen) < alen) { + vcn_next = vcn; + + if (!run_get_entry(run, ++idx, &vcn, &lcn, &clen) || + vcn_next != vcn) { + err = attr_load_runs_vcn(ni, attr->type, + attr_name(attr), + attr->name_len, run, vcn_next); + if (err) + return err; + vcn = vcn_next; + + if (!run_lookup_entry(run, vcn, &lcn, &clen, &idx)) + return -EINVAL; + } + + if (lcn == SPARSE_LCN) { + slen += clen; + } else { + if (slen) { + /* + * data_clusters + sparse_clusters = + * not enough for frame + */ + return -EINVAL; + } + *clst_data += clen; + } + + if (*clst_data + slen >= clst_frame) { + if (!slen) { + /* + * There is no sparsed clusters in this frame + * So it is not compressed + */ + *clst_data = clst_frame; + } else { + /*frame is compressed*/ + } + break; + } + } + + return 0; +} + +/* + * attr_allocate_frame + * + * allocate/free clusters for 'frame' + * assumed: down_write(&ni->file.run_lock); + */ +int attr_allocate_frame(struct ntfs_inode *ni, CLST frame, size_t compr_size, + u64 new_valid) +{ + int err = 0; + struct runs_tree *run = &ni->file.run; + struct ntfs_sb_info *sbi = ni->mi.sbi; + struct ATTRIB *attr = NULL, *attr_b; + struct ATTR_LIST_ENTRY *le, *le_b; + struct mft_inode *mi, *mi_b; + CLST svcn, evcn1, next_svcn, lcn, len; + CLST vcn, end, clst_data; + u64 total_size, valid_size, data_size; + + le_b = NULL; + attr_b = ni_find_attr(ni, NULL, &le_b, ATTR_DATA, NULL, 0, NULL, &mi_b); + if (!attr_b) + return -ENOENT; + + if (!is_attr_ext(attr_b)) + return -EINVAL; + + vcn = frame << NTFS_LZNT_CUNIT; + total_size = le64_to_cpu(attr_b->nres.total_size); + + svcn = le64_to_cpu(attr_b->nres.svcn); + evcn1 = le64_to_cpu(attr_b->nres.evcn) + 1; + data_size = le64_to_cpu(attr_b->nres.data_size); + + if (svcn <= vcn && vcn < evcn1) { + attr = attr_b; + le = le_b; + mi = mi_b; + } else if (!le_b) { + err = -EINVAL; + goto out; + } else { + le = le_b; + attr = ni_find_attr(ni, attr_b, &le, ATTR_DATA, NULL, 0, &vcn, + &mi); + if (!attr) { + err = -EINVAL; + goto out; + } + svcn = le64_to_cpu(attr->nres.svcn); + evcn1 = le64_to_cpu(attr->nres.evcn) + 1; + } + + err = attr_load_runs(attr, ni, run, NULL); + if (err) + goto out; + + err = attr_is_frame_compressed(ni, attr_b, frame, &clst_data); + if (err) + goto out; + + total_size -= (u64)clst_data << sbi->cluster_bits; + + len = bytes_to_cluster(sbi, compr_size); + + if (len == clst_data) + goto out; + + if (len < clst_data) { + err = run_deallocate_ex(sbi, run, vcn + len, clst_data - len, + NULL, true); + if (err) + goto out; + + if (!run_add_entry(run, vcn + len, SPARSE_LCN, clst_data - len, + false)) { + err = -ENOMEM; + goto out; + } + end = vcn + clst_data; + /* run contains updated range [vcn + len : end) */ + } else { + CLST alen, hint = 0; + /* Get the last lcn to allocate from */ + if (vcn + clst_data && + !run_lookup_entry(run, vcn + clst_data - 1, &hint, NULL, + NULL)) { + hint = -1; + } + + err = attr_allocate_clusters(sbi, run, vcn + clst_data, + hint + 1, len - clst_data, NULL, 0, + &alen, 0, &lcn); + if (err) + goto out; + + end = vcn + len; + /* run contains updated range [vcn + clst_data : end) */ + } + + total_size += (u64)len << sbi->cluster_bits; + +repack: + err = mi_pack_runs(mi, attr, run, max(end, evcn1) - svcn); + if (err) + goto out; + + attr_b->nres.total_size = cpu_to_le64(total_size); + inode_set_bytes(&ni->vfs_inode, total_size); + + mi_b->dirty = true; + mark_inode_dirty(&ni->vfs_inode); + + /* stored [vcn : next_svcn) from [vcn : end) */ + next_svcn = le64_to_cpu(attr->nres.evcn) + 1; + + if (end <= evcn1) { + if (next_svcn == evcn1) { + /* Normal way. update attribute and exit */ + goto ok; + } + /* add new segment [next_svcn : evcn1 - next_svcn )*/ + if (!ni->attr_list.size) { + err = ni_create_attr_list(ni); + if (err) + goto out; + /* layout of records is changed */ + le_b = NULL; + attr_b = ni_find_attr(ni, NULL, &le_b, ATTR_DATA, NULL, + 0, NULL, &mi_b); + if (!attr_b) { + err = -ENOENT; + goto out; + } + + attr = attr_b; + le = le_b; + mi = mi_b; + goto repack; + } + } + + svcn = evcn1; + + /* Estimate next attribute */ + attr = ni_find_attr(ni, attr, &le, ATTR_DATA, NULL, 0, &svcn, &mi); + + if (attr) { + CLST alloc = bytes_to_cluster( + sbi, le64_to_cpu(attr_b->nres.alloc_size)); + CLST evcn = le64_to_cpu(attr->nres.evcn); + + if (end < next_svcn) + end = next_svcn; + while (end > evcn) { + /* remove segment [svcn : evcn)*/ + mi_remove_attr(mi, attr); + + if (!al_remove_le(ni, le)) { + err = -EINVAL; + goto out; + } + + if (evcn + 1 >= alloc) { + /* last attribute segment */ + evcn1 = evcn + 1; + goto ins_ext; + } + + if (ni_load_mi(ni, le, &mi)) { + attr = NULL; + goto out; + } + + attr = mi_find_attr(mi, NULL, ATTR_DATA, NULL, 0, + &le->id); + if (!attr) { + err = -EINVAL; + goto out; + } + svcn = le64_to_cpu(attr->nres.svcn); + evcn = le64_to_cpu(attr->nres.evcn); + } + + if (end < svcn) + end = svcn; + + err = attr_load_runs(attr, ni, run, &end); + if (err) + goto out; + + evcn1 = evcn + 1; + attr->nres.svcn = cpu_to_le64(next_svcn); + err = mi_pack_runs(mi, attr, run, evcn1 - next_svcn); + if (err) + goto out; + + le->vcn = cpu_to_le64(next_svcn); + ni->attr_list.dirty = true; + mi->dirty = true; + + next_svcn = le64_to_cpu(attr->nres.evcn) + 1; + } +ins_ext: + if (evcn1 > next_svcn) { + err = ni_insert_nonresident(ni, ATTR_DATA, NULL, 0, run, + next_svcn, evcn1 - next_svcn, + attr_b->flags, &attr, &mi); + if (err) + goto out; + } +ok: + run_truncate_around(run, vcn); +out: + if (new_valid > data_size) + new_valid = data_size; + + valid_size = le64_to_cpu(attr_b->nres.valid_size); + if (new_valid != valid_size) { + attr_b->nres.valid_size = cpu_to_le64(valid_size); + mi_b->dirty = true; + } + + return err; +} + +/* Collapse range in file */ +int attr_collapse_range(struct ntfs_inode *ni, u64 vbo, u64 bytes) +{ + int err = 0; + struct runs_tree *run = &ni->file.run; + struct ntfs_sb_info *sbi = ni->mi.sbi; + struct ATTRIB *attr = NULL, *attr_b; + struct ATTR_LIST_ENTRY *le, *le_b; + struct mft_inode *mi, *mi_b; + CLST svcn, evcn1, len, dealloc, alen; + CLST vcn, end; + u64 valid_size, data_size, alloc_size, total_size; + u32 mask; + __le16 a_flags; + + if (!bytes) + return 0; + + le_b = NULL; + attr_b = ni_find_attr(ni, NULL, &le_b, ATTR_DATA, NULL, 0, NULL, &mi_b); + if (!attr_b) + return -ENOENT; + + if (!attr_b->non_res) { + /* Attribute is resident. Nothing to do? */ + return 0; + } + + data_size = le64_to_cpu(attr_b->nres.data_size); + alloc_size = le64_to_cpu(attr_b->nres.alloc_size); + a_flags = attr_b->flags; + + if (is_attr_ext(attr_b)) { + total_size = le64_to_cpu(attr_b->nres.total_size); + mask = (sbi->cluster_size << attr_b->nres.c_unit) - 1; + } else { + total_size = alloc_size; + mask = sbi->cluster_mask; + } + + if ((vbo & mask) || (bytes & mask)) { + /* allow to collapse only cluster aligned ranges */ + return -EINVAL; + } + + if (vbo > data_size) + return -EINVAL; + + down_write(&ni->file.run_lock); + + if (vbo + bytes >= data_size) { + u64 new_valid = min(ni->i_valid, vbo); + + /* Simple truncate file at 'vbo' */ + truncate_setsize(&ni->vfs_inode, vbo); + err = attr_set_size(ni, ATTR_DATA, NULL, 0, &ni->file.run, vbo, + &new_valid, true, NULL); + + if (!err && new_valid < ni->i_valid) + ni->i_valid = new_valid; + + goto out; + } + + /* + * Enumerate all attribute segments and collapse + */ + alen = alloc_size >> sbi->cluster_bits; + vcn = vbo >> sbi->cluster_bits; + len = bytes >> sbi->cluster_bits; + end = vcn + len; + dealloc = 0; + + svcn = le64_to_cpu(attr_b->nres.svcn); + evcn1 = le64_to_cpu(attr_b->nres.evcn) + 1; + + if (svcn <= vcn && vcn < evcn1) { + attr = attr_b; + le = le_b; + mi = mi_b; + } else if (!le_b) { + err = -EINVAL; + goto out; + } else { + le = le_b; + attr = ni_find_attr(ni, attr_b, &le, ATTR_DATA, NULL, 0, &vcn, + &mi); + if (!attr) { + err = -EINVAL; + goto out; + } + + svcn = le64_to_cpu(attr->nres.svcn); + evcn1 = le64_to_cpu(attr->nres.evcn) + 1; + } + + for (;;) { + if (svcn >= end) { + /* shift vcn */ + attr->nres.svcn = cpu_to_le64(svcn - len); + attr->nres.evcn = cpu_to_le64(evcn1 - 1 - len); + if (le) { + le->vcn = attr->nres.svcn; + ni->attr_list.dirty = true; + } + mi->dirty = true; + } else if (svcn < vcn || end < evcn1) { + CLST vcn1, eat, next_svcn; + + /* collapse a part of this attribute segment */ + err = attr_load_runs(attr, ni, run, &svcn); + if (err) + goto out; + vcn1 = max(vcn, svcn); + eat = min(end, evcn1) - vcn1; + + err = run_deallocate_ex(sbi, run, vcn1, eat, &dealloc, + true); + if (err) + goto out; + + if (!run_collapse_range(run, vcn1, eat)) { + err = -ENOMEM; + goto out; + } + + if (svcn >= vcn) { + /* shift vcn */ + attr->nres.svcn = cpu_to_le64(vcn); + if (le) { + le->vcn = attr->nres.svcn; + ni->attr_list.dirty = true; + } + } + + err = mi_pack_runs(mi, attr, run, evcn1 - svcn - eat); + if (err) + goto out; + + next_svcn = le64_to_cpu(attr->nres.evcn) + 1; + if (next_svcn + eat < evcn1) { + err = ni_insert_nonresident( + ni, ATTR_DATA, NULL, 0, run, next_svcn, + evcn1 - eat - next_svcn, a_flags, &attr, + &mi); + if (err) + goto out; + + /* layout of records maybe changed */ + attr_b = NULL; + le = al_find_ex(ni, NULL, ATTR_DATA, NULL, 0, + &next_svcn); + if (!le) { + err = -EINVAL; + goto out; + } + } + + /* free all allocated memory */ + run_truncate(run, 0); + } else { + u16 le_sz; + u16 roff = le16_to_cpu(attr->nres.run_off); + + /*run==1 means unpack and deallocate*/ + run_unpack_ex(RUN_DEALLOCATE, sbi, ni->mi.rno, svcn, + evcn1 - 1, svcn, Add2Ptr(attr, roff), + le32_to_cpu(attr->size) - roff); + + /* delete this attribute segment */ + mi_remove_attr(mi, attr); + if (!le) + break; + + le_sz = le16_to_cpu(le->size); + if (!al_remove_le(ni, le)) { + err = -EINVAL; + goto out; + } + + if (evcn1 >= alen) + break; + + if (!svcn) { + /* Load next record that contains this attribute */ + if (ni_load_mi(ni, le, &mi)) { + err = -EINVAL; + goto out; + } + + /* Look for required attribute */ + attr = mi_find_attr(mi, NULL, ATTR_DATA, NULL, + 0, &le->id); + if (!attr) { + err = -EINVAL; + goto out; + } + goto next_attr; + } + le = (struct ATTR_LIST_ENTRY *)((u8 *)le - le_sz); + } + + if (evcn1 >= alen) + break; + + attr = ni_enum_attr_ex(ni, attr, &le, &mi); + if (!attr) { + err = -EINVAL; + goto out; + } + +next_attr: + svcn = le64_to_cpu(attr->nres.svcn); + evcn1 = le64_to_cpu(attr->nres.evcn) + 1; + } + + if (!attr_b) { + le_b = NULL; + attr_b = ni_find_attr(ni, NULL, &le_b, ATTR_DATA, NULL, 0, NULL, + &mi_b); + if (!attr_b) { + err = -ENOENT; + goto out; + } + } + + data_size -= bytes; + valid_size = ni->i_valid; + if (vbo + bytes <= valid_size) + valid_size -= bytes; + else if (vbo < valid_size) + valid_size = vbo; + + attr_b->nres.alloc_size = cpu_to_le64(alloc_size - bytes); + attr_b->nres.data_size = cpu_to_le64(data_size); + attr_b->nres.valid_size = cpu_to_le64(min(valid_size, data_size)); + total_size -= (u64)dealloc << sbi->cluster_bits; + if (is_attr_ext(attr_b)) + attr_b->nres.total_size = cpu_to_le64(total_size); + mi_b->dirty = true; + + /*update inode size*/ + ni->i_valid = valid_size; + ni->vfs_inode.i_size = data_size; + inode_set_bytes(&ni->vfs_inode, total_size); + ni->ni_flags |= NI_FLAG_UPDATE_PARENT; + mark_inode_dirty(&ni->vfs_inode); + +out: + up_write(&ni->file.run_lock); + if (err) + make_bad_inode(&ni->vfs_inode); + + return err; +} + +/* not for normal files */ +int attr_punch_hole(struct ntfs_inode *ni, u64 vbo, u64 bytes, u32 *frame_size) +{ + int err = 0; + struct runs_tree *run = &ni->file.run; + struct ntfs_sb_info *sbi = ni->mi.sbi; + struct ATTRIB *attr = NULL, *attr_b; + struct ATTR_LIST_ENTRY *le, *le_b; + struct mft_inode *mi, *mi_b; + CLST svcn, evcn1, vcn, len, end, alen, dealloc; + u64 total_size, alloc_size; + u32 mask; + + if (!bytes) + return 0; + + le_b = NULL; + attr_b = ni_find_attr(ni, NULL, &le_b, ATTR_DATA, NULL, 0, NULL, &mi_b); + if (!attr_b) + return -ENOENT; + + if (!attr_b->non_res) { + u32 data_size = le32_to_cpu(attr->res.data_size); + u32 from, to; + + if (vbo > data_size) + return 0; + + from = vbo; + to = (vbo + bytes) < data_size ? (vbo + bytes) : data_size; + memset(Add2Ptr(resident_data(attr_b), from), 0, to - from); + return 0; + } + + if (!is_attr_ext(attr_b)) + return -EOPNOTSUPP; + + alloc_size = le64_to_cpu(attr_b->nres.alloc_size); + total_size = le64_to_cpu(attr_b->nres.total_size); + + if (vbo >= alloc_size) { + // NOTE: it is allowed + return 0; + } + + mask = (sbi->cluster_size << attr_b->nres.c_unit) - 1; + + bytes += vbo; + if (bytes > alloc_size) + bytes = alloc_size; + bytes -= vbo; + + if ((vbo & mask) || (bytes & mask)) { + /* We have to zero a range(s)*/ + if (frame_size == NULL) { + /* Caller insists range is aligned */ + return -EINVAL; + } + *frame_size = mask + 1; + return E_NTFS_NOTALIGNED; + } + + down_write(&ni->file.run_lock); + /* + * Enumerate all attribute segments and punch hole where necessary + */ + alen = alloc_size >> sbi->cluster_bits; + vcn = vbo >> sbi->cluster_bits; + len = bytes >> sbi->cluster_bits; + end = vcn + len; + dealloc = 0; + + svcn = le64_to_cpu(attr_b->nres.svcn); + evcn1 = le64_to_cpu(attr_b->nres.evcn) + 1; + + if (svcn <= vcn && vcn < evcn1) { + attr = attr_b; + le = le_b; + mi = mi_b; + } else if (!le_b) { + err = -EINVAL; + goto out; + } else { + le = le_b; + attr = ni_find_attr(ni, attr_b, &le, ATTR_DATA, NULL, 0, &vcn, + &mi); + if (!attr) { + err = -EINVAL; + goto out; + } + + svcn = le64_to_cpu(attr->nres.svcn); + evcn1 = le64_to_cpu(attr->nres.evcn) + 1; + } + + while (svcn < end) { + CLST vcn1, zero, dealloc2; + + err = attr_load_runs(attr, ni, run, &svcn); + if (err) + goto out; + vcn1 = max(vcn, svcn); + zero = min(end, evcn1) - vcn1; + + dealloc2 = dealloc; + err = run_deallocate_ex(sbi, run, vcn1, zero, &dealloc, true); + if (err) + goto out; + + if (dealloc2 == dealloc) { + /* looks like the required range is already sparsed */ + } else { + if (!run_add_entry(run, vcn1, SPARSE_LCN, zero, + false)) { + err = -ENOMEM; + goto out; + } + + err = mi_pack_runs(mi, attr, run, evcn1 - svcn); + if (err) + goto out; + } + /* free all allocated memory */ + run_truncate(run, 0); + + if (evcn1 >= alen) + break; + + attr = ni_enum_attr_ex(ni, attr, &le, &mi); + if (!attr) { + err = -EINVAL; + goto out; + } + + svcn = le64_to_cpu(attr->nres.svcn); + evcn1 = le64_to_cpu(attr->nres.evcn) + 1; + } + + total_size -= (u64)dealloc << sbi->cluster_bits; + attr_b->nres.total_size = cpu_to_le64(total_size); + mi_b->dirty = true; + + /*update inode size*/ + inode_set_bytes(&ni->vfs_inode, total_size); + ni->ni_flags |= NI_FLAG_UPDATE_PARENT; + mark_inode_dirty(&ni->vfs_inode); + +out: + up_write(&ni->file.run_lock); + if (err) + make_bad_inode(&ni->vfs_inode); + + return err; +} diff --git a/fs/ntfs3/attrlist.c b/fs/ntfs3/attrlist.c new file mode 100644 index 000000000000..ea561361b576 --- /dev/null +++ b/fs/ntfs3/attrlist.c @@ -0,0 +1,456 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * + * Copyright (C) 2019-2021 Paragon Software GmbH, All rights reserved. + * + */ + +#include <linux/blkdev.h> +#include <linux/buffer_head.h> +#include <linux/fs.h> +#include <linux/nls.h> + +#include "debug.h" +#include "ntfs.h" +#include "ntfs_fs.h" + +/* Returns true if le is valid */ +static inline bool al_is_valid_le(const struct ntfs_inode *ni, + struct ATTR_LIST_ENTRY *le) +{ + if (!le || !ni->attr_list.le || !ni->attr_list.size) + return false; + + return PtrOffset(ni->attr_list.le, le) + le16_to_cpu(le->size) <= + ni->attr_list.size; +} + +void al_destroy(struct ntfs_inode *ni) +{ + run_close(&ni->attr_list.run); + ntfs_free(ni->attr_list.le); + ni->attr_list.le = NULL; + ni->attr_list.size = 0; + ni->attr_list.dirty = false; +} + +/* + * ntfs_load_attr_list + * + * This method makes sure that the ATTRIB list, if present, + * has been properly set up. + */ +int ntfs_load_attr_list(struct ntfs_inode *ni, struct ATTRIB *attr) +{ + int err; + size_t lsize; + void *le = NULL; + + if (ni->attr_list.size) + return 0; + + if (!attr->non_res) { + lsize = le32_to_cpu(attr->res.data_size); + le = ntfs_malloc(al_aligned(lsize)); + if (!le) { + err = -ENOMEM; + goto out; + } + memcpy(le, resident_data(attr), lsize); + } else if (attr->nres.svcn) { + err = -EINVAL; + goto out; + } else { + u16 run_off = le16_to_cpu(attr->nres.run_off); + + lsize = le64_to_cpu(attr->nres.data_size); + + run_init(&ni->attr_list.run); + + err = run_unpack_ex(&ni->attr_list.run, ni->mi.sbi, ni->mi.rno, + 0, le64_to_cpu(attr->nres.evcn), 0, + Add2Ptr(attr, run_off), + le32_to_cpu(attr->size) - run_off); + if (err < 0) + goto out; + + le = ntfs_malloc(al_aligned(lsize)); + if (!le) { + err = -ENOMEM; + goto out; + } + + err = ntfs_read_run_nb(ni->mi.sbi, &ni->attr_list.run, 0, le, + lsize, NULL); + if (err) + goto out; + } + + ni->attr_list.size = lsize; + ni->attr_list.le = le; + + return 0; + +out: + ni->attr_list.le = le; + al_destroy(ni); + + return err; +} + +/* + * al_enumerate + * + * Returns the next list 'le' + * if 'le' is NULL then returns the first 'le' + */ +struct ATTR_LIST_ENTRY *al_enumerate(struct ntfs_inode *ni, + struct ATTR_LIST_ENTRY *le) +{ + size_t off; + u16 sz; + + if (!le) { + le = ni->attr_list.le; + } else { + sz = le16_to_cpu(le->size); + if (sz < sizeof(struct ATTR_LIST_ENTRY)) { + /* Impossible 'cause we should not return such 'le' */ + return NULL; + } + le = Add2Ptr(le, sz); + } + + /* Check boundary */ + off = PtrOffset(ni->attr_list.le, le); + if (off + sizeof(struct ATTR_LIST_ENTRY) > ni->attr_list.size) { + // The regular end of list + return NULL; + } + + sz = le16_to_cpu(le->size); + + /* Check 'le' for errors */ + if (sz < sizeof(struct ATTR_LIST_ENTRY) || + off + sz > ni->attr_list.size || + sz < le->name_off + le->name_len * sizeof(short)) { + return NULL; + } + + return le; +} + +/* + * al_find_le + * + * finds the first 'le' in the list which matches type, name and vcn + * Returns NULL if not found + */ +struct ATTR_LIST_ENTRY *al_find_le(struct ntfs_inode *ni, + struct ATTR_LIST_ENTRY *le, + const struct ATTRIB *attr) +{ + CLST svcn = attr_svcn(attr); + + return al_find_ex(ni, le, attr->type, attr_name(attr), attr->name_len, + &svcn); +} + +/* + * al_find_ex + * + * finds the first 'le' in the list which matches type, name and vcn + * Returns NULL if not found + */ +struct ATTR_LIST_ENTRY *al_find_ex(struct ntfs_inode *ni, + struct ATTR_LIST_ENTRY *le, + enum ATTR_TYPE type, const __le16 *name, + u8 name_len, const CLST *vcn) +{ + struct ATTR_LIST_ENTRY *ret = NULL; + u32 type_in = le32_to_cpu(type); + + while ((le = al_enumerate(ni, le))) { + u64 le_vcn; + int diff = le32_to_cpu(le->type) - type_in; + + /* List entries are sorted by type, name and vcn */ + if (diff < 0) + continue; + + if (diff > 0) + return ret; + + if (le->name_len != name_len) + continue; + + le_vcn = le64_to_cpu(le->vcn); + if (!le_vcn) { + /* + * compare entry names only for entry with vcn == 0 + */ + diff = ntfs_cmp_names(le_name(le), name_len, name, + name_len, ni->mi.sbi->upcase, + true); + if (diff < 0) + continue; + + if (diff > 0) + return ret; + } + + if (!vcn) + return le; + + if (*vcn == le_vcn) + return le; + + if (*vcn < le_vcn) + return ret; + + ret = le; + } + + return ret; +} + +/* + * al_find_le_to_insert + * + * finds the first list entry which matches type, name and vcn + */ +static struct ATTR_LIST_ENTRY *al_find_le_to_insert(struct ntfs_inode *ni, + enum ATTR_TYPE type, + const __le16 *name, + u8 name_len, CLST vcn) +{ + struct ATTR_LIST_ENTRY *le = NULL, *prev; + u32 type_in = le32_to_cpu(type); + + /* List entries are sorted by type, name, vcn */ + while ((le = al_enumerate(ni, prev = le))) { + int diff = le32_to_cpu(le->type) - type_in; + + if (diff < 0) + continue; + + if (diff > 0) + return le; + + if (!le->vcn) { + /* + * compare entry names only for entry with vcn == 0 + */ + diff = ntfs_cmp_names(le_name(le), le->name_len, name, + name_len, ni->mi.sbi->upcase, + true); + if (diff < 0) + continue; + + if (diff > 0) + return le; + } + + if (le64_to_cpu(le->vcn) >= vcn) + return le; + } + + return prev ? Add2Ptr(prev, le16_to_cpu(prev->size)) : ni->attr_list.le; +} + +/* + * al_add_le + * + * adds an "attribute list entry" to the list. + */ +int al_add_le(struct ntfs_inode *ni, enum ATTR_TYPE type, const __le16 *name, + u8 name_len, CLST svcn, __le16 id, const struct MFT_REF *ref, + struct ATTR_LIST_ENTRY **new_le) +{ + int err; + struct ATTRIB *attr; + struct ATTR_LIST_ENTRY *le; + size_t off; + u16 sz; + size_t asize, new_asize; + u64 new_size; + typeof(ni->attr_list) *al = &ni->attr_list; + + /* + * Compute the size of the new 'le' + */ + sz = le_size(name_len); + new_size = al->size + sz; + asize = al_aligned(al->size); + new_asize = al_aligned(new_size); + + /* Scan forward to the point at which the new 'le' should be inserted. */ + le = al_find_le_to_insert(ni, type, name, name_len, svcn); + off = PtrOffset(al->le, le); + + if (new_size > asize) { + void *ptr = ntfs_malloc(new_asize); + + if (!ptr) + return -ENOMEM; + + memcpy(ptr, al->le, off); + memcpy(Add2Ptr(ptr, off + sz), le, al->size - off); + le = Add2Ptr(ptr, off); + ntfs_free(al->le); + al->le = ptr; + } else { + memmove(Add2Ptr(le, sz), le, al->size - off); + } + + al->size = new_size; + + le->type = type; + le->size = cpu_to_le16(sz); + le->name_len = name_len; + le->name_off = offsetof(struct ATTR_LIST_ENTRY, name); + le->vcn = cpu_to_le64(svcn); + le->ref = *ref; + le->id = id; + memcpy(le->name, name, sizeof(short) * name_len); + + al->dirty = true; + + err = attr_set_size(ni, ATTR_LIST, NULL, 0, &al->run, new_size, + &new_size, true, &attr); + if (err) + return err; + + if (attr && attr->non_res) { + err = ntfs_sb_write_run(ni->mi.sbi, &al->run, 0, al->le, + al->size); + if (err) + return err; + } + + al->dirty = false; + *new_le = le; + + return 0; +} + +/* + * al_remove_le + * + * removes 'le' from attribute list + */ +bool al_remove_le(struct ntfs_inode *ni, struct ATTR_LIST_ENTRY *le) +{ + u16 size; + size_t off; + typeof(ni->attr_list) *al = &ni->attr_list; + + if (!al_is_valid_le(ni, le)) + return false; + + /* Save on stack the size of 'le' */ + size = le16_to_cpu(le->size); + off = PtrOffset(al->le, le); + + memmove(le, Add2Ptr(le, size), al->size - (off + size)); + + al->size -= size; + al->dirty = true; + + return true; +} + +/* + * al_delete_le + * + * deletes from the list the first 'le' which matches its parameters. + */ +bool al_delete_le(struct ntfs_inode *ni, enum ATTR_TYPE type, CLST vcn, + const __le16 *name, size_t name_len, + const struct MFT_REF *ref) +{ + u16 size; + struct ATTR_LIST_ENTRY *le; + size_t off; + typeof(ni->attr_list) *al = &ni->attr_list; + + /* Scan forward to the first 'le' that matches the input */ + le = al_find_ex(ni, NULL, type, name, name_len, &vcn); + if (!le) + return false; + + off = PtrOffset(al->le, le); + +next: + if (off >= al->size) + return false; + if (le->type != type) + return false; + if (le->name_len != name_len) + return false; + if (name_len && ntfs_cmp_names(le_name(le), name_len, name, name_len, + ni->mi.sbi->upcase, true)) + return false; + if (le64_to_cpu(le->vcn) != vcn) + return false; + + /* + * The caller specified a segment reference, so we have to + * scan through the matching entries until we find that segment + * reference or we run of matching entries. + */ + if (ref && memcmp(ref, &le->ref, sizeof(*ref))) { + off += le16_to_cpu(le->size); + le = Add2Ptr(al->le, off); + goto next; + } + + /* Save on stack the size of 'le' */ + size = le16_to_cpu(le->size); + /* Delete 'le'. */ + memmove(le, Add2Ptr(le, size), al->size - (off + size)); + + al->size -= size; + al->dirty = true; + + return true; +} + +/* + * al_update + */ +int al_update(struct ntfs_inode *ni) +{ + int err; + struct ATTRIB *attr; + typeof(ni->attr_list) *al = &ni->attr_list; + + if (!al->dirty || !al->size) + return 0; + + /* + * attribute list increased on demand in al_add_le + * attribute list decreased here + */ + err = attr_set_size(ni, ATTR_LIST, NULL, 0, &al->run, al->size, NULL, + false, &attr); + if (err) + goto out; + + if (!attr->non_res) { + memcpy(resident_data(attr), al->le, al->size); + } else { + err = ntfs_sb_write_run(ni->mi.sbi, &al->run, 0, al->le, + al->size); + if (err) + goto out; + + attr->nres.valid_size = attr->nres.data_size; + } + + ni->mi.dirty = true; + al->dirty = false; + +out: + return err; +} diff --git a/fs/ntfs3/xattr.c b/fs/ntfs3/xattr.c new file mode 100644 index 000000000000..98871c895e77 --- /dev/null +++ b/fs/ntfs3/xattr.c @@ -0,0 +1,1128 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * + * Copyright (C) 2019-2021 Paragon Software GmbH, All rights reserved. + * + */ + +#include <linux/blkdev.h> +#include <linux/buffer_head.h> +#include <linux/fs.h> +#include <linux/nls.h> +#include <linux/posix_acl.h> +#include <linux/posix_acl_xattr.h> +#include <linux/xattr.h> + +#include "debug.h" +#include "ntfs.h" +#include "ntfs_fs.h" + +// clang-format off +#define SYSTEM_DOS_ATTRIB "system.dos_attrib" +#define SYSTEM_NTFS_ATTRIB "system.ntfs_attrib" +#define SYSTEM_NTFS_SECURITY "system.ntfs_security" +// clang-format on + +static inline size_t unpacked_ea_size(const struct EA_FULL *ea) +{ + return ea->size ? le32_to_cpu(ea->size) + : DwordAlign(struct_size( + ea, name, + 1 + ea->name_len + le16_to_cpu(ea->elength))); +} + +static inline size_t packed_ea_size(const struct EA_FULL *ea) +{ + return struct_size(ea, name, + 1 + ea->name_len + le16_to_cpu(ea->elength)) - + offsetof(struct EA_FULL, flags); +} + +/* + * find_ea + * + * assume there is at least one xattr in the list + */ +static inline bool find_ea(const struct EA_FULL *ea_all, u32 bytes, + const char *name, u8 name_len, u32 *off) +{ + *off = 0; + + if (!ea_all || !bytes) + return false; + + for (;;) { + const struct EA_FULL *ea = Add2Ptr(ea_all, *off); + u32 next_off = *off + unpacked_ea_size(ea); + + if (next_off > bytes) + return false; + + if (ea->name_len == name_len && + !memcmp(ea->name, name, name_len)) + return true; + + *off = next_off; + if (next_off >= bytes) + return false; + } +} + +/* + * ntfs_read_ea + * + * reads all extended attributes + * ea - new allocated memory + * info - pointer into resident data + */ +static int ntfs_read_ea(struct ntfs_inode *ni, struct EA_FULL **ea, + size_t add_bytes, const struct EA_INFO **info) +{ + int err; + struct ATTR_LIST_ENTRY *le = NULL; + struct ATTRIB *attr_info, *attr_ea; + void *ea_p; + u32 size; + + static_assert(le32_to_cpu(ATTR_EA_INFO) < le32_to_cpu(ATTR_EA)); + + *ea = NULL; + *info = NULL; + + attr_info = + ni_find_attr(ni, NULL, &le, ATTR_EA_INFO, NULL, 0, NULL, NULL); + attr_ea = + ni_find_attr(ni, attr_info, &le, ATTR_EA, NULL, 0, NULL, NULL); + + if (!attr_ea || !attr_info) + return 0; + + *info = resident_data_ex(attr_info, sizeof(struct EA_INFO)); + if (!*info) + return -EINVAL; + + /* Check Ea limit */ + size = le32_to_cpu((*info)->size); + if (size > ni->mi.sbi->ea_max_size) + return -EFBIG; + + if (attr_size(attr_ea) > ni->mi.sbi->ea_max_size) + return -EFBIG; + + /* Allocate memory for packed Ea */ + ea_p = ntfs_malloc(size + add_bytes); + if (!ea_p) + return -ENOMEM; + + if (attr_ea->non_res) { + struct runs_tree run; + + run_init(&run); + + err = attr_load_runs(attr_ea, ni, &run, NULL); + if (!err) + err = ntfs_read_run_nb(ni->mi.sbi, &run, 0, ea_p, size, + NULL); + run_close(&run); + + if (err) + goto out; + } else { + void *p = resident_data_ex(attr_ea, size); + + if (!p) { + err = -EINVAL; + goto out; + } + memcpy(ea_p, p, size); + } + + memset(Add2Ptr(ea_p, size), 0, add_bytes); + *ea = ea_p; + return 0; + +out: + ntfs_free(ea_p); + *ea = NULL; + return err; +} + +/* + * ntfs_list_ea + * + * copy a list of xattrs names into the buffer + * provided, or compute the buffer size required + * + * Returns a negative error number on failure, or the number of bytes + * used / required on success. + */ +static ssize_t ntfs_list_ea(struct ntfs_inode *ni, char *buffer, + size_t bytes_per_buffer) +{ + const struct EA_INFO *info; + struct EA_FULL *ea_all = NULL; + const struct EA_FULL *ea; + u32 off, size; + int err; + size_t ret; + + err = ntfs_read_ea(ni, &ea_all, 0, &info); + if (err) + return err; + + if (!info || !ea_all) + return 0; + + size = le32_to_cpu(info->size); + + /* Enumerate all xattrs */ + for (ret = 0, off = 0; off < size; off += unpacked_ea_size(ea)) { + ea = Add2Ptr(ea_all, off); + + if (buffer) { + if (ret + ea->name_len + 1 > bytes_per_buffer) { + err = -ERANGE; + goto out; + } + + memcpy(buffer + ret, ea->name, ea->name_len); + buffer[ret + ea->name_len] = 0; + } + + ret += ea->name_len + 1; + } + +out: + ntfs_free(ea_all); + return err ? err : ret; +} + +static int ntfs_get_ea(struct inode *inode, const char *name, size_t name_len, + void *buffer, size_t size, size_t *required) +{ + struct ntfs_inode *ni = ntfs_i(inode); + const struct EA_INFO *info; + struct EA_FULL *ea_all = NULL; + const struct EA_FULL *ea; + u32 off, len; + int err; + + if (!(ni->ni_flags & NI_FLAG_EA)) + return -ENODATA; + + if (!required) + ni_lock(ni); + + len = 0; + + if (name_len > 255) { + err = -ENAMETOOLONG; + goto out; + } + + err = ntfs_read_ea(ni, &ea_all, 0, &info); + if (err) + goto out; + + if (!info) + goto out; + + /* Enumerate all xattrs */ + if (!find_ea(ea_all, le32_to_cpu(info->size), name, name_len, &off)) { + err = -ENODATA; + goto out; + } + ea = Add2Ptr(ea_all, off); + + len = le16_to_cpu(ea->elength); + if (!buffer) { + err = 0; + goto out; + } + + if (len > size) { + err = -ERANGE; + if (required) + *required = len; + goto out; + } + + memcpy(buffer, ea->name + ea->name_len + 1, len); + err = 0; + +out: + ntfs_free(ea_all); + if (!required) + ni_unlock(ni); + + return err ? err : len; +} + +static noinline int ntfs_set_ea(struct inode *inode, const char *name, + size_t name_len, const void *value, + size_t val_size, int flags, int locked) +{ + struct ntfs_inode *ni = ntfs_i(inode); + struct ntfs_sb_info *sbi = ni->mi.sbi; + int err; + struct EA_INFO ea_info; + const struct EA_INFO *info; + struct EA_FULL *new_ea; + struct EA_FULL *ea_all = NULL; + size_t add, new_pack; + u32 off, size; + __le16 size_pack; + struct ATTRIB *attr; + struct ATTR_LIST_ENTRY *le; + struct mft_inode *mi; + struct runs_tree ea_run; + u64 new_sz; + void *p; + + if (!locked) + ni_lock(ni); + + run_init(&ea_run); + + if (name_len > 255) { + err = -ENAMETOOLONG; + goto out; + } + + add = DwordAlign(struct_size(ea_all, name, 1 + name_len + val_size)); + + err = ntfs_read_ea(ni, &ea_all, add, &info); + if (err) + goto out; + + if (!info) { + memset(&ea_info, 0, sizeof(ea_info)); + size = 0; + size_pack = 0; + } else { + memcpy(&ea_info, info, sizeof(ea_info)); + size = le32_to_cpu(ea_info.size); + size_pack = ea_info.size_pack; + } + + if (info && find_ea(ea_all, size, name, name_len, &off)) { + struct EA_FULL *ea; + size_t ea_sz; + + if (flags & XATTR_CREATE) { + err = -EEXIST; + goto out; + } + + ea = Add2Ptr(ea_all, off); + + /* + * Check simple case when we try to insert xattr with the same value + * e.g. ntfs_save_wsl_perm + */ + if (val_size && le16_to_cpu(ea->elength) == val_size && + !memcmp(ea->name + ea->name_len + 1, value, val_size)) { + /* xattr already contains the required value */ + goto out; + } + + /* Remove current xattr */ + if (ea->flags & FILE_NEED_EA) + le16_add_cpu(&ea_info.count, -1); + + ea_sz = unpacked_ea_size(ea); + + le16_add_cpu(&ea_info.size_pack, 0 - packed_ea_size(ea)); + + memmove(ea, Add2Ptr(ea, ea_sz), size - off - ea_sz); + + size -= ea_sz; + memset(Add2Ptr(ea_all, size), 0, ea_sz); + + ea_info.size = cpu_to_le32(size); + + if ((flags & XATTR_REPLACE) && !val_size) { + /* remove xattr */ + goto update_ea; + } + } else { + if (flags & XATTR_REPLACE) { + err = -ENODATA; + goto out; + } + + if (!ea_all) { + ea_all = ntfs_zalloc(add); + if (!ea_all) { + err = -ENOMEM; + goto out; + } + } + } + + /* append new xattr */ + new_ea = Add2Ptr(ea_all, size); + new_ea->size = cpu_to_le32(add); + new_ea->flags = 0; + new_ea->name_len = name_len; + new_ea->elength = cpu_to_le16(val_size); + memcpy(new_ea->name, name, name_len); + new_ea->name[name_len] = 0; + memcpy(new_ea->name + name_len + 1, value, val_size); + new_pack = le16_to_cpu(ea_info.size_pack) + packed_ea_size(new_ea); + + /* should fit into 16 bits */ + if (new_pack > 0xffff) { + err = -EFBIG; // -EINVAL? + goto out; + } + ea_info.size_pack = cpu_to_le16(new_pack); + + /* new size of ATTR_EA */ + size += add; + if (size > sbi->ea_max_size) { + err = -EFBIG; // -EINVAL? + goto out; + } + ea_info.size = cpu_to_le32(size); + +update_ea: + + if (!info) { + /* Create xattr */ + if (!size) { + err = 0; + goto out; + } + + err = ni_insert_resident(ni, sizeof(struct EA_INFO), + ATTR_EA_INFO, NULL, 0, NULL, NULL); + if (err) + goto out; + + err = ni_insert_resident(ni, 0, ATTR_EA, NULL, 0, NULL, NULL); + if (err) + goto out; + } + + new_sz = size; + err = attr_set_size(ni, ATTR_EA, NULL, 0, &ea_run, new_sz, &new_sz, + false, NULL); + if (err) + goto out; + + le = NULL; + attr = ni_find_attr(ni, NULL, &le, ATTR_EA_INFO, NULL, 0, NULL, &mi); + if (!attr) { + err = -EINVAL; + goto out; + } + + if (!size) { + /* delete xattr, ATTR_EA_INFO */ + err = ni_remove_attr_le(ni, attr, le); + if (err) + goto out; + } else { + p = resident_data_ex(attr, sizeof(struct EA_INFO)); + if (!p) { + err = -EINVAL; + goto out; + } + memcpy(p, &ea_info, sizeof(struct EA_INFO)); + mi->dirty = true; + } + + le = NULL; + attr = ni_find_attr(ni, NULL, &le, ATTR_EA, NULL, 0, NULL, &mi); + if (!attr) { + err = -EINVAL; + goto out; + } + + if (!size) { + /* delete xattr, ATTR_EA */ + err = ni_remove_attr_le(ni, attr, le); + if (err) + goto out; + } else if (attr->non_res) { + err = ntfs_sb_write_run(sbi, &ea_run, 0, ea_all, size); + if (err) + goto out; + } else { + p = resident_data_ex(attr, size); + if (!p) { + err = -EINVAL; + goto out; + } + memcpy(p, ea_all, size); + mi->dirty = true; + } + + /* Check if we delete the last xattr */ + if (size) + ni->ni_flags |= NI_FLAG_EA; + else + ni->ni_flags &= ~NI_FLAG_EA; + + if (ea_info.size_pack != size_pack) + ni->ni_flags |= NI_FLAG_UPDATE_PARENT; + mark_inode_dirty(&ni->vfs_inode); + +out: + if (!locked) + ni_unlock(ni); + + run_close(&ea_run); + ntfs_free(ea_all); + + return err; +} + +#ifdef CONFIG_NTFS3_FS_POSIX_ACL +static inline void ntfs_posix_acl_release(struct posix_acl *acl) +{ + if (acl && refcount_dec_and_test(&acl->a_refcount)) + kfree(acl); +} + +static struct posix_acl *ntfs_get_acl_ex(struct user_namespace *mnt_userns, + struct inode *inode, int type, + int locked) +{ + struct ntfs_inode *ni = ntfs_i(inode); + const char *name; + size_t name_len; + struct posix_acl *acl; + size_t req; + int err; + void *buf; + + /* allocate PATH_MAX bytes */ + buf = __getname(); + if (!buf) + return ERR_PTR(-ENOMEM); + + /* Possible values of 'type' was already checked above */ + if (type == ACL_TYPE_ACCESS) { + name = XATTR_NAME_POSIX_ACL_ACCESS; + name_len = sizeof(XATTR_NAME_POSIX_ACL_ACCESS) - 1; + } else { + name = XATTR_NAME_POSIX_ACL_DEFAULT; + name_len = sizeof(XATTR_NAME_POSIX_ACL_DEFAULT) - 1; + } + + if (!locked) + ni_lock(ni); + + err = ntfs_get_ea(inode, name, name_len, buf, PATH_MAX, &req); + + if (!locked) + ni_unlock(ni); + + /* Translate extended attribute to acl */ + if (err > 0) { + acl = posix_acl_from_xattr(mnt_userns, buf, err); + if (!IS_ERR(acl)) + set_cached_acl(inode, type, acl); + } else { + acl = err == -ENODATA ? NULL : ERR_PTR(err); + } + + __putname(buf); + + return acl; +} + +/* + * ntfs_get_acl + * + * inode_operations::get_acl + */ +struct posix_acl *ntfs_get_acl(struct inode *inode, int type) +{ + /* TODO: init_user_ns? */ + return ntfs_get_acl_ex(&init_user_ns, inode, type, 0); +} + +static noinline int ntfs_set_acl_ex(struct user_namespace *mnt_userns, + struct inode *inode, struct posix_acl *acl, + int type, int locked) +{ + const char *name; + size_t size, name_len; + void *value = NULL; + int err = 0; + + if (S_ISLNK(inode->i_mode)) + return -EOPNOTSUPP; + + switch (type) { + case ACL_TYPE_ACCESS: + if (acl) { + umode_t mode = inode->i_mode; + + err = posix_acl_equiv_mode(acl, &mode); + if (err < 0) + return err; + + if (inode->i_mode != mode) { + inode->i_mode = mode; + mark_inode_dirty(inode); + } + + if (!err) { + /* + * acl can be exactly represented in the + * traditional file mode permission bits + */ + acl = NULL; + } + } + name = XATTR_NAME_POSIX_ACL_ACCESS; + name_len = sizeof(XATTR_NAME_POSIX_ACL_ACCESS) - 1; + break; + + case ACL_TYPE_DEFAULT: + if (!S_ISDIR(inode->i_mode)) + return acl ? -EACCES : 0; + name = XATTR_NAME_POSIX_ACL_DEFAULT; + name_len = sizeof(XATTR_NAME_POSIX_ACL_DEFAULT) - 1; + break; + + default: + return -EINVAL; + } + + if (!acl) { + size = 0; + value = NULL; + } else { + size = posix_acl_xattr_size(acl->a_count); + value = ntfs_malloc(size); + if (!value) + return -ENOMEM; + + err = posix_acl_to_xattr(mnt_userns, acl, value, size); + if (err < 0) + goto out; + } + + err = ntfs_set_ea(inode, name, name_len, value, size, + acl ? 0 : XATTR_REPLACE, locked); + if (!err) + set_cached_acl(inode, type, acl); + +out: + ntfs_free(value); + + return err; +} + +/* + * ntfs_set_acl + * + * inode_operations::set_acl + */ +int ntfs_set_acl(struct user_namespace *mnt_userns, struct inode *inode, + struct posix_acl *acl, int type) +{ + return ntfs_set_acl_ex(mnt_userns, inode, acl, type, 0); +} + +static int ntfs_xattr_get_acl(struct user_namespace *mnt_userns, + struct inode *inode, int type, void *buffer, + size_t size) +{ + struct posix_acl *acl; + int err; + + if (!(inode->i_sb->s_flags & SB_POSIXACL)) + return -EOPNOTSUPP; + + acl = ntfs_get_acl(inode, type); + if (IS_ERR(acl)) + return PTR_ERR(acl); + + if (!acl) + return -ENODATA; + + err = posix_acl_to_xattr(mnt_userns, acl, buffer, size); + ntfs_posix_acl_release(acl); + + return err; +} + +static int ntfs_xattr_set_acl(struct user_namespace *mnt_userns, + struct inode *inode, int type, const void *value, + size_t size) +{ + struct posix_acl *acl; + int err; + + if (!(inode->i_sb->s_flags & SB_POSIXACL)) + return -EOPNOTSUPP; + + if (!inode_owner_or_capable(mnt_userns, inode)) + return -EPERM; + + if (!value) { + acl = NULL; + } else { + acl = posix_acl_from_xattr(mnt_userns, value, size); + if (IS_ERR(acl)) + return PTR_ERR(acl); + + if (acl) { + err = posix_acl_valid(mnt_userns, acl); + if (err) + goto release_and_out; + } + } + + err = ntfs_set_acl(mnt_userns, inode, acl, type); + +release_and_out: + ntfs_posix_acl_release(acl); + return err; +} + +/* + * Initialize the ACLs of a new inode. Called from ntfs_create_inode. + */ +int ntfs_init_acl(struct user_namespace *mnt_userns, struct inode *inode, + struct inode *dir) +{ + struct posix_acl *default_acl, *acl; + int err; + + /* + * TODO refactoring lock + * ni_lock(dir) ... -> posix_acl_create(dir,...) -> ntfs_get_acl -> ni_lock(dir) + */ + inode->i_default_acl = NULL; + + default_acl = ntfs_get_acl_ex(mnt_userns, dir, ACL_TYPE_DEFAULT, 1); + + if (!default_acl || default_acl == ERR_PTR(-EOPNOTSUPP)) { + inode->i_mode &= ~current_umask(); + err = 0; + goto out; + } + + if (IS_ERR(default_acl)) { + err = PTR_ERR(default_acl); + goto out; + } + + acl = default_acl; + err = __posix_acl_create(&acl, GFP_NOFS, &inode->i_mode); + if (err < 0) + goto out1; + if (!err) { + posix_acl_release(acl); + acl = NULL; + } + + if (!S_ISDIR(inode->i_mode)) { + posix_acl_release(default_acl); + default_acl = NULL; + } + + if (default_acl) + err = ntfs_set_acl_ex(mnt_userns, inode, default_acl, + ACL_TYPE_DEFAULT, 1); + + if (!acl) + inode->i_acl = NULL; + else if (!err) + err = ntfs_set_acl_ex(mnt_userns, inode, acl, ACL_TYPE_ACCESS, + 1); + + posix_acl_release(acl); +out1: + posix_acl_release(default_acl); + +out: + return err; +} +#endif + +/* + * ntfs_acl_chmod + * + * helper for 'ntfs3_setattr' + */ +int ntfs_acl_chmod(struct user_namespace *mnt_userns, struct inode *inode) +{ + struct super_block *sb = inode->i_sb; + + if (!(sb->s_flags & SB_POSIXACL)) + return 0; + + if (S_ISLNK(inode->i_mode)) + return -EOPNOTSUPP; + + return posix_acl_chmod(mnt_userns, inode, inode->i_mode); +} + +/* + * ntfs_permission + * + * inode_operations::permission + */ +int ntfs_permission(struct user_namespace *mnt_userns, struct inode *inode, + int mask) +{ + if (ntfs_sb(inode->i_sb)->options.no_acs_rules) { + /* "no access rules" mode - allow all changes */ + return 0; + } + + return generic_permission(mnt_userns, inode, mask); +} + +/* + * ntfs_listxattr + * + * inode_operations::listxattr + */ +ssize_t ntfs_listxattr(struct dentry *dentry, char *buffer, size_t size) +{ + struct inode *inode = d_inode(dentry); + struct ntfs_inode *ni = ntfs_i(inode); + ssize_t ret; + + if (!(ni->ni_flags & NI_FLAG_EA)) { + /* no xattr in file */ + return 0; + } + + ni_lock(ni); + + ret = ntfs_list_ea(ni, buffer, size); + + ni_unlock(ni); + + return ret; +} + +static int ntfs_getxattr(const struct xattr_handler *handler, struct dentry *de, + struct inode *inode, const char *name, void *buffer, + size_t size) +{ + int err; + struct ntfs_inode *ni = ntfs_i(inode); + size_t name_len = strlen(name); + + /* Dispatch request */ + if (name_len == sizeof(SYSTEM_DOS_ATTRIB) - 1 && + !memcmp(name, SYSTEM_DOS_ATTRIB, sizeof(SYSTEM_DOS_ATTRIB))) { + /* system.dos_attrib */ + if (!buffer) { + err = sizeof(u8); + } else if (size < sizeof(u8)) { + err = -ENODATA; + } else { + err = sizeof(u8); + *(u8 *)buffer = le32_to_cpu(ni->std_fa); + } + goto out; + } + + if (name_len == sizeof(SYSTEM_NTFS_ATTRIB) - 1 && + !memcmp(name, SYSTEM_NTFS_ATTRIB, sizeof(SYSTEM_NTFS_ATTRIB))) { + /* system.ntfs_attrib */ + if (!buffer) { + err = sizeof(u32); + } else if (size < sizeof(u32)) { + err = -ENODATA; + } else { + err = sizeof(u32); + *(u32 *)buffer = le32_to_cpu(ni->std_fa); + } + goto out; + } + + if (name_len == sizeof(SYSTEM_NTFS_SECURITY) - 1 && + !memcmp(name, SYSTEM_NTFS_SECURITY, sizeof(SYSTEM_NTFS_SECURITY))) { + /* system.ntfs_security*/ + struct SECURITY_DESCRIPTOR_RELATIVE *sd = NULL; + size_t sd_size = 0; + + if (!is_ntfs3(ni->mi.sbi)) { + /* we should get nt4 security */ + err = -EINVAL; + goto out; + } else if (le32_to_cpu(ni->std_security_id) < + SECURITY_ID_FIRST) { + err = -ENOENT; + goto out; + } + + err = ntfs_get_security_by_id(ni->mi.sbi, ni->std_security_id, + &sd, &sd_size); + if (err) + goto out; + + if (!is_sd_valid(sd, sd_size)) { + ntfs_inode_warn( + inode, + "looks like you get incorrect security descriptor id=%u", + ni->std_security_id); + } + + if (!buffer) { + err = sd_size; + } else if (size < sd_size) { + err = -ENODATA; + } else { + err = sd_size; + memcpy(buffer, sd, sd_size); + } + ntfs_free(sd); + goto out; + } + +#ifdef CONFIG_NTFS3_FS_POSIX_ACL + if ((name_len == sizeof(XATTR_NAME_POSIX_ACL_ACCESS) - 1 && + !memcmp(name, XATTR_NAME_POSIX_ACL_ACCESS, + sizeof(XATTR_NAME_POSIX_ACL_ACCESS))) || + (name_len == sizeof(XATTR_NAME_POSIX_ACL_DEFAULT) - 1 && + !memcmp(name, XATTR_NAME_POSIX_ACL_DEFAULT, + sizeof(XATTR_NAME_POSIX_ACL_DEFAULT)))) { + /* TODO: init_user_ns? */ + err = ntfs_xattr_get_acl( + &init_user_ns, inode, + name_len == sizeof(XATTR_NAME_POSIX_ACL_ACCESS) - 1 + ? ACL_TYPE_ACCESS + : ACL_TYPE_DEFAULT, + buffer, size); + goto out; + } +#endif + /* deal with ntfs extended attribute */ + err = ntfs_get_ea(inode, name, name_len, buffer, size, NULL); + +out: + return err; +} + +/* + * ntfs_setxattr + * + * inode_operations::setxattr + */ +static noinline int ntfs_setxattr(const struct xattr_handler *handler, + struct user_namespace *mnt_userns, + struct dentry *de, struct inode *inode, + const char *name, const void *value, + size_t size, int flags) +{ + int err = -EINVAL; + struct ntfs_inode *ni = ntfs_i(inode); + size_t name_len = strlen(name); + enum FILE_ATTRIBUTE new_fa; + + /* Dispatch request */ + if (name_len == sizeof(SYSTEM_DOS_ATTRIB) - 1 && + !memcmp(name, SYSTEM_DOS_ATTRIB, sizeof(SYSTEM_DOS_ATTRIB))) { + if (sizeof(u8) != size) + goto out; + new_fa = cpu_to_le32(*(u8 *)value); + goto set_new_fa; + } + + if (name_len == sizeof(SYSTEM_NTFS_ATTRIB) - 1 && + !memcmp(name, SYSTEM_NTFS_ATTRIB, sizeof(SYSTEM_NTFS_ATTRIB))) { + if (size != sizeof(u32)) + goto out; + new_fa = cpu_to_le32(*(u32 *)value); + + if (S_ISREG(inode->i_mode)) { + /* Process compressed/sparsed in special way*/ + ni_lock(ni); + err = ni_new_attr_flags(ni, new_fa); + ni_unlock(ni); + if (err) + goto out; + } +set_new_fa: + /* + * Thanks Mark Harmstone: + * keep directory bit consistency + */ + if (S_ISDIR(inode->i_mode)) + new_fa |= FILE_ATTRIBUTE_DIRECTORY; + else + new_fa &= ~FILE_ATTRIBUTE_DIRECTORY; + + if (ni->std_fa != new_fa) { + ni->std_fa = new_fa; + if (new_fa & FILE_ATTRIBUTE_READONLY) + inode->i_mode &= ~0222; + else + inode->i_mode |= 0222; + /* std attribute always in primary record */ + ni->mi.dirty = true; + mark_inode_dirty(inode); + } + err = 0; + + goto out; + } + + if (name_len == sizeof(SYSTEM_NTFS_SECURITY) - 1 && + !memcmp(name, SYSTEM_NTFS_SECURITY, sizeof(SYSTEM_NTFS_SECURITY))) { + /* system.ntfs_security*/ + __le32 security_id; + bool inserted; + struct ATTR_STD_INFO5 *std; + + if (!is_ntfs3(ni->mi.sbi)) { + /* + * we should replace ATTR_SECURE + * Skip this way cause it is nt4 feature + */ + err = -EINVAL; + goto out; + } + + if (!is_sd_valid(value, size)) { + err = -EINVAL; + ntfs_inode_warn( + inode, + "you try to set invalid security descriptor"); + goto out; + } + + err = ntfs_insert_security(ni->mi.sbi, value, size, + &security_id, &inserted); + if (err) + goto out; + + ni_lock(ni); + std = ni_std5(ni); + if (!std) { + err = -EINVAL; + } else if (std->security_id != security_id) { + std->security_id = ni->std_security_id = security_id; + /* std attribute always in primary record */ + ni->mi.dirty = true; + mark_inode_dirty(&ni->vfs_inode); + } + ni_unlock(ni); + goto out; + } + +#ifdef CONFIG_NTFS3_FS_POSIX_ACL + if ((name_len == sizeof(XATTR_NAME_POSIX_ACL_ACCESS) - 1 && + !memcmp(name, XATTR_NAME_POSIX_ACL_ACCESS, + sizeof(XATTR_NAME_POSIX_ACL_ACCESS))) || + (name_len == sizeof(XATTR_NAME_POSIX_ACL_DEFAULT) - 1 && + !memcmp(name, XATTR_NAME_POSIX_ACL_DEFAULT, + sizeof(XATTR_NAME_POSIX_ACL_DEFAULT)))) { + err = ntfs_xattr_set_acl( + mnt_userns, inode, + name_len == sizeof(XATTR_NAME_POSIX_ACL_ACCESS) - 1 + ? ACL_TYPE_ACCESS + : ACL_TYPE_DEFAULT, + value, size); + goto out; + } +#endif + /* deal with ntfs extended attribute */ + err = ntfs_set_ea(inode, name, name_len, value, size, flags, 0); + +out: + return err; +} + +/* + * ntfs_save_wsl_perm + * + * save uid/gid/mode in xattr + */ +int ntfs_save_wsl_perm(struct inode *inode) +{ + int err; + __le32 value; + + value = cpu_to_le32(i_uid_read(inode)); + err = ntfs_set_ea(inode, "$LXUID", sizeof("$LXUID") - 1, &value, + sizeof(value), 0, 0); + if (err) + goto out; + + value = cpu_to_le32(i_gid_read(inode)); + err = ntfs_set_ea(inode, "$LXGID", sizeof("$LXGID") - 1, &value, + sizeof(value), 0, 0); + if (err) + goto out; + + value = cpu_to_le32(inode->i_mode); + err = ntfs_set_ea(inode, "$LXMOD", sizeof("$LXMOD") - 1, &value, + sizeof(value), 0, 0); + if (err) + goto out; + + if (S_ISCHR(inode->i_mode) || S_ISBLK(inode->i_mode)) { + value = cpu_to_le32(inode->i_rdev); + err = ntfs_set_ea(inode, "$LXDEV", sizeof("$LXDEV") - 1, &value, + sizeof(value), 0, 0); + if (err) + goto out; + } + +out: + /* In case of error should we delete all WSL xattr? */ + return err; +} + +/* + * ntfs_get_wsl_perm + * + * get uid/gid/mode from xattr + * it is called from ntfs_iget5->ntfs_read_mft + */ +void ntfs_get_wsl_perm(struct inode *inode) +{ + size_t sz; + __le32 value[3]; + + if (ntfs_get_ea(inode, "$LXUID", sizeof("$LXUID") - 1, &value[0], + sizeof(value[0]), &sz) == sizeof(value[0]) && + ntfs_get_ea(inode, "$LXGID", sizeof("$LXGID") - 1, &value[1], + sizeof(value[1]), &sz) == sizeof(value[1]) && + ntfs_get_ea(inode, "$LXMOD", sizeof("$LXMOD") - 1, &value[2], + sizeof(value[2]), &sz) == sizeof(value[2])) { + i_uid_write(inode, (uid_t)le32_to_cpu(value[0])); + i_gid_write(inode, (gid_t)le32_to_cpu(value[1])); + inode->i_mode = le32_to_cpu(value[2]); + + if (ntfs_get_ea(inode, "$LXDEV", sizeof("$$LXDEV") - 1, + &value[0], sizeof(value), + &sz) == sizeof(value[0])) { + inode->i_rdev = le32_to_cpu(value[0]); + } + } +} + +static bool ntfs_xattr_user_list(struct dentry *dentry) +{ + return true; +} + +// clang-format off +static const struct xattr_handler ntfs_xattr_handler = { + .prefix = "", + .get = ntfs_getxattr, + .set = ntfs_setxattr, + .list = ntfs_xattr_user_list, +}; + +const struct xattr_handler *ntfs_xattr_handlers[] = { + &ntfs_xattr_handler, + NULL, +}; +// clang-format on -- 2.30.0

1 0

[PATCH OLK-5.10 003/107] fs/ntfs3: Add bitmap
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> mainline inclusion from mainline-v5.15-rc1 commit 3f3b442b5ad2455507c9bfdacf39a3792eb3a6d0 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- This adds bitmap Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/bitfunc.c | 135 ++++ fs/ntfs3/bitmap.c | 1519 ++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 1654 insertions(+) create mode 100644 fs/ntfs3/bitfunc.c create mode 100644 fs/ntfs3/bitmap.c diff --git a/fs/ntfs3/bitfunc.c b/fs/ntfs3/bitfunc.c new file mode 100644 index 000000000000..2de5faef2721 --- /dev/null +++ b/fs/ntfs3/bitfunc.c @@ -0,0 +1,135 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * + * Copyright (C) 2019-2021 Paragon Software GmbH, All rights reserved. + * + */ +#include <linux/blkdev.h> +#include <linux/buffer_head.h> +#include <linux/fs.h> +#include <linux/nls.h> + +#include "debug.h" +#include "ntfs.h" +#include "ntfs_fs.h" + +#define BITS_IN_SIZE_T (sizeof(size_t) * 8) + +/* + * fill_mask[i] - first i bits are '1' , i = 0,1,2,3,4,5,6,7,8 + * fill_mask[i] = 0xFF >> (8-i) + */ +static const u8 fill_mask[] = { 0x00, 0x01, 0x03, 0x07, 0x0F, + 0x1F, 0x3F, 0x7F, 0xFF }; + +/* + * zero_mask[i] - first i bits are '0' , i = 0,1,2,3,4,5,6,7,8 + * zero_mask[i] = 0xFF << i + */ +static const u8 zero_mask[] = { 0xFF, 0xFE, 0xFC, 0xF8, 0xF0, + 0xE0, 0xC0, 0x80, 0x00 }; + +/* + * are_bits_clear + * + * Returns true if all bits [bit, bit+nbits) are zeros "0" + */ +bool are_bits_clear(const ulong *lmap, size_t bit, size_t nbits) +{ + size_t pos = bit & 7; + const u8 *map = (u8 *)lmap + (bit >> 3); + + if (pos) { + if (8 - pos >= nbits) + return !nbits || !(*map & fill_mask[pos + nbits] & + zero_mask[pos]); + + if (*map++ & zero_mask[pos]) + return false; + nbits -= 8 - pos; + } + + pos = ((size_t)map) & (sizeof(size_t) - 1); + if (pos) { + pos = sizeof(size_t) - pos; + if (nbits >= pos * 8) { + for (nbits -= pos * 8; pos; pos--, map++) { + if (*map) + return false; + } + } + } + + for (pos = nbits / BITS_IN_SIZE_T; pos; pos--, map += sizeof(size_t)) { + if (*((size_t *)map)) + return false; + } + + for (pos = (nbits % BITS_IN_SIZE_T) >> 3; pos; pos--, map++) { + if (*map) + return false; + } + + pos = nbits & 7; + if (pos && (*map & fill_mask[pos])) + return false; + + // All bits are zero + return true; +} + +/* + * are_bits_set + * + * Returns true if all bits [bit, bit+nbits) are ones "1" + */ +bool are_bits_set(const ulong *lmap, size_t bit, size_t nbits) +{ + u8 mask; + size_t pos = bit & 7; + const u8 *map = (u8 *)lmap + (bit >> 3); + + if (pos) { + if (8 - pos >= nbits) { + mask = fill_mask[pos + nbits] & zero_mask[pos]; + return !nbits || (*map & mask) == mask; + } + + mask = zero_mask[pos]; + if ((*map++ & mask) != mask) + return false; + nbits -= 8 - pos; + } + + pos = ((size_t)map) & (sizeof(size_t) - 1); + if (pos) { + pos = sizeof(size_t) - pos; + if (nbits >= pos * 8) { + for (nbits -= pos * 8; pos; pos--, map++) { + if (*map != 0xFF) + return false; + } + } + } + + for (pos = nbits / BITS_IN_SIZE_T; pos; pos--, map += sizeof(size_t)) { + if (*((size_t *)map) != MINUS_ONE_T) + return false; + } + + for (pos = (nbits % BITS_IN_SIZE_T) >> 3; pos; pos--, map++) { + if (*map != 0xFF) + return false; + } + + pos = nbits & 7; + if (pos) { + u8 mask = fill_mask[pos]; + + if ((*map & mask) != mask) + return false; + } + + // All bits are ones + return true; +} diff --git a/fs/ntfs3/bitmap.c b/fs/ntfs3/bitmap.c new file mode 100644 index 000000000000..32aab0031221 --- /dev/null +++ b/fs/ntfs3/bitmap.c @@ -0,0 +1,1519 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * + * Copyright (C) 2019-2021 Paragon Software GmbH, All rights reserved. + * + * This code builds two trees of free clusters extents. + * Trees are sorted by start of extent and by length of extent. + * NTFS_MAX_WND_EXTENTS defines the maximum number of elements in trees. + * In extreme case code reads on-disk bitmap to find free clusters + * + */ + +#include <linux/blkdev.h> +#include <linux/buffer_head.h> +#include <linux/fs.h> +#include <linux/nls.h> + +#include "debug.h" +#include "ntfs.h" +#include "ntfs_fs.h" + +/* + * Maximum number of extents in tree. + */ +#define NTFS_MAX_WND_EXTENTS (32u * 1024u) + +struct rb_node_key { + struct rb_node node; + size_t key; +}; + +/* + * Tree is sorted by start (key) + */ +struct e_node { + struct rb_node_key start; /* Tree sorted by start */ + struct rb_node_key count; /* Tree sorted by len*/ +}; + +static int wnd_rescan(struct wnd_bitmap *wnd); +static struct buffer_head *wnd_map(struct wnd_bitmap *wnd, size_t iw); +static bool wnd_is_free_hlp(struct wnd_bitmap *wnd, size_t bit, size_t bits); + +static struct kmem_cache *ntfs_enode_cachep; + +int __init ntfs3_init_bitmap(void) +{ + ntfs_enode_cachep = + kmem_cache_create("ntfs3_enode_cache", sizeof(struct e_node), 0, + SLAB_RECLAIM_ACCOUNT, NULL); + return ntfs_enode_cachep ? 0 : -ENOMEM; +} + +void ntfs3_exit_bitmap(void) +{ + kmem_cache_destroy(ntfs_enode_cachep); +} + +static inline u32 wnd_bits(const struct wnd_bitmap *wnd, size_t i) +{ + return i + 1 == wnd->nwnd ? wnd->bits_last : wnd->sb->s_blocksize * 8; +} + +/* + * b_pos + b_len - biggest fragment + * Scan range [wpos wbits) window 'buf' + * Returns -1 if not found + */ +static size_t wnd_scan(const ulong *buf, size_t wbit, u32 wpos, u32 wend, + size_t to_alloc, size_t *prev_tail, size_t *b_pos, + size_t *b_len) +{ + while (wpos < wend) { + size_t free_len; + u32 free_bits, end; + u32 used = find_next_zero_bit(buf, wend, wpos); + + if (used >= wend) { + if (*b_len < *prev_tail) { + *b_pos = wbit - *prev_tail; + *b_len = *prev_tail; + } + + *prev_tail = 0; + return -1; + } + + if (used > wpos) { + wpos = used; + if (*b_len < *prev_tail) { + *b_pos = wbit - *prev_tail; + *b_len = *prev_tail; + } + + *prev_tail = 0; + } + + /* + * Now we have a fragment [wpos, wend) staring with 0 + */ + end = wpos + to_alloc - *prev_tail; + free_bits = find_next_bit(buf, min(end, wend), wpos); + + free_len = *prev_tail + free_bits - wpos; + + if (*b_len < free_len) { + *b_pos = wbit + wpos - *prev_tail; + *b_len = free_len; + } + + if (free_len >= to_alloc) + return wbit + wpos - *prev_tail; + + if (free_bits >= wend) { + *prev_tail += free_bits - wpos; + return -1; + } + + wpos = free_bits + 1; + + *prev_tail = 0; + } + + return -1; +} + +/* + * wnd_close + * + * Frees all resources + */ +void wnd_close(struct wnd_bitmap *wnd) +{ + struct rb_node *node, *next; + + ntfs_free(wnd->free_bits); + run_close(&wnd->run); + + node = rb_first(&wnd->start_tree); + + while (node) { + next = rb_next(node); + rb_erase(node, &wnd->start_tree); + kmem_cache_free(ntfs_enode_cachep, + rb_entry(node, struct e_node, start.node)); + node = next; + } +} + +static struct rb_node *rb_lookup(struct rb_root *root, size_t v) +{ + struct rb_node **p = &root->rb_node; + struct rb_node *r = NULL; + + while (*p) { + struct rb_node_key *k; + + k = rb_entry(*p, struct rb_node_key, node); + if (v < k->key) { + p = &(*p)->rb_left; + } else if (v > k->key) { + r = &k->node; + p = &(*p)->rb_right; + } else { + return &k->node; + } + } + + return r; +} + +/* + * rb_insert_count + * + * Helper function to insert special kind of 'count' tree + */ +static inline bool rb_insert_count(struct rb_root *root, struct e_node *e) +{ + struct rb_node **p = &root->rb_node; + struct rb_node *parent = NULL; + size_t e_ckey = e->count.key; + size_t e_skey = e->start.key; + + while (*p) { + struct e_node *k = + rb_entry(parent = *p, struct e_node, count.node); + + if (e_ckey > k->count.key) { + p = &(*p)->rb_left; + } else if (e_ckey < k->count.key) { + p = &(*p)->rb_right; + } else if (e_skey < k->start.key) { + p = &(*p)->rb_left; + } else if (e_skey > k->start.key) { + p = &(*p)->rb_right; + } else { + WARN_ON(1); + return false; + } + } + + rb_link_node(&e->count.node, parent, p); + rb_insert_color(&e->count.node, root); + return true; +} + +/* + * inline bool rb_insert_start + * + * Helper function to insert special kind of 'start' tree + */ +static inline bool rb_insert_start(struct rb_root *root, struct e_node *e) +{ + struct rb_node **p = &root->rb_node; + struct rb_node *parent = NULL; + size_t e_skey = e->start.key; + + while (*p) { + struct e_node *k; + + parent = *p; + + k = rb_entry(parent, struct e_node, start.node); + if (e_skey < k->start.key) { + p = &(*p)->rb_left; + } else if (e_skey > k->start.key) { + p = &(*p)->rb_right; + } else { + WARN_ON(1); + return false; + } + } + + rb_link_node(&e->start.node, parent, p); + rb_insert_color(&e->start.node, root); + return true; +} + +/* + * wnd_add_free_ext + * + * adds a new extent of free space + * build = 1 when building tree + */ +static void wnd_add_free_ext(struct wnd_bitmap *wnd, size_t bit, size_t len, + bool build) +{ + struct e_node *e, *e0 = NULL; + size_t ib, end_in = bit + len; + struct rb_node *n; + + if (build) { + /* Use extent_min to filter too short extents */ + if (wnd->count >= NTFS_MAX_WND_EXTENTS && + len <= wnd->extent_min) { + wnd->uptodated = -1; + return; + } + } else { + /* Try to find extent before 'bit' */ + n = rb_lookup(&wnd->start_tree, bit); + + if (!n) { + n = rb_first(&wnd->start_tree); + } else { + e = rb_entry(n, struct e_node, start.node); + n = rb_next(n); + if (e->start.key + e->count.key == bit) { + /* Remove left */ + bit = e->start.key; + len += e->count.key; + rb_erase(&e->start.node, &wnd->start_tree); + rb_erase(&e->count.node, &wnd->count_tree); + wnd->count -= 1; + e0 = e; + } + } + + while (n) { + size_t next_end; + + e = rb_entry(n, struct e_node, start.node); + next_end = e->start.key + e->count.key; + if (e->start.key > end_in) + break; + + /* Remove right */ + n = rb_next(n); + len += next_end - end_in; + end_in = next_end; + rb_erase(&e->start.node, &wnd->start_tree); + rb_erase(&e->count.node, &wnd->count_tree); + wnd->count -= 1; + + if (!e0) + e0 = e; + else + kmem_cache_free(ntfs_enode_cachep, e); + } + + if (wnd->uptodated != 1) { + /* Check bits before 'bit' */ + ib = wnd->zone_bit == wnd->zone_end || + bit < wnd->zone_end + ? 0 + : wnd->zone_end; + + while (bit > ib && wnd_is_free_hlp(wnd, bit - 1, 1)) { + bit -= 1; + len += 1; + } + + /* Check bits after 'end_in' */ + ib = wnd->zone_bit == wnd->zone_end || + end_in > wnd->zone_bit + ? wnd->nbits + : wnd->zone_bit; + + while (end_in < ib && wnd_is_free_hlp(wnd, end_in, 1)) { + end_in += 1; + len += 1; + } + } + } + /* Insert new fragment */ + if (wnd->count >= NTFS_MAX_WND_EXTENTS) { + if (e0) + kmem_cache_free(ntfs_enode_cachep, e0); + + wnd->uptodated = -1; + + /* Compare with smallest fragment */ + n = rb_last(&wnd->count_tree); + e = rb_entry(n, struct e_node, count.node); + if (len <= e->count.key) + goto out; /* Do not insert small fragments */ + + if (build) { + struct e_node *e2; + + n = rb_prev(n); + e2 = rb_entry(n, struct e_node, count.node); + /* smallest fragment will be 'e2->count.key' */ + wnd->extent_min = e2->count.key; + } + + /* Replace smallest fragment by new one */ + rb_erase(&e->start.node, &wnd->start_tree); + rb_erase(&e->count.node, &wnd->count_tree); + wnd->count -= 1; + } else { + e = e0 ? e0 : kmem_cache_alloc(ntfs_enode_cachep, GFP_ATOMIC); + if (!e) { + wnd->uptodated = -1; + goto out; + } + + if (build && len <= wnd->extent_min) + wnd->extent_min = len; + } + e->start.key = bit; + e->count.key = len; + if (len > wnd->extent_max) + wnd->extent_max = len; + + rb_insert_start(&wnd->start_tree, e); + rb_insert_count(&wnd->count_tree, e); + wnd->count += 1; + +out:; +} + +/* + * wnd_remove_free_ext + * + * removes a run from the cached free space + */ +static void wnd_remove_free_ext(struct wnd_bitmap *wnd, size_t bit, size_t len) +{ + struct rb_node *n, *n3; + struct e_node *e, *e3; + size_t end_in = bit + len; + size_t end3, end, new_key, new_len, max_new_len; + + /* Try to find extent before 'bit' */ + n = rb_lookup(&wnd->start_tree, bit); + + if (!n) + return; + + e = rb_entry(n, struct e_node, start.node); + end = e->start.key + e->count.key; + + new_key = new_len = 0; + len = e->count.key; + + /* Range [bit,end_in) must be inside 'e' or outside 'e' and 'n' */ + if (e->start.key > bit) + ; + else if (end_in <= end) { + /* Range [bit,end_in) inside 'e' */ + new_key = end_in; + new_len = end - end_in; + len = bit - e->start.key; + } else if (bit > end) { + bool bmax = false; + + n3 = rb_next(n); + + while (n3) { + e3 = rb_entry(n3, struct e_node, start.node); + if (e3->start.key >= end_in) + break; + + if (e3->count.key == wnd->extent_max) + bmax = true; + + end3 = e3->start.key + e3->count.key; + if (end3 > end_in) { + e3->start.key = end_in; + rb_erase(&e3->count.node, &wnd->count_tree); + e3->count.key = end3 - end_in; + rb_insert_count(&wnd->count_tree, e3); + break; + } + + n3 = rb_next(n3); + rb_erase(&e3->start.node, &wnd->start_tree); + rb_erase(&e3->count.node, &wnd->count_tree); + wnd->count -= 1; + kmem_cache_free(ntfs_enode_cachep, e3); + } + if (!bmax) + return; + n3 = rb_first(&wnd->count_tree); + wnd->extent_max = + n3 ? rb_entry(n3, struct e_node, count.node)->count.key + : 0; + return; + } + + if (e->count.key != wnd->extent_max) { + ; + } else if (rb_prev(&e->count.node)) { + ; + } else { + n3 = rb_next(&e->count.node); + max_new_len = len > new_len ? len : new_len; + if (!n3) { + wnd->extent_max = max_new_len; + } else { + e3 = rb_entry(n3, struct e_node, count.node); + wnd->extent_max = max(e3->count.key, max_new_len); + } + } + + if (!len) { + if (new_len) { + e->start.key = new_key; + rb_erase(&e->count.node, &wnd->count_tree); + e->count.key = new_len; + rb_insert_count(&wnd->count_tree, e); + } else { + rb_erase(&e->start.node, &wnd->start_tree); + rb_erase(&e->count.node, &wnd->count_tree); + wnd->count -= 1; + kmem_cache_free(ntfs_enode_cachep, e); + } + goto out; + } + rb_erase(&e->count.node, &wnd->count_tree); + e->count.key = len; + rb_insert_count(&wnd->count_tree, e); + + if (!new_len) + goto out; + + if (wnd->count >= NTFS_MAX_WND_EXTENTS) { + wnd->uptodated = -1; + + /* Get minimal extent */ + e = rb_entry(rb_last(&wnd->count_tree), struct e_node, + count.node); + if (e->count.key > new_len) + goto out; + + /* Replace minimum */ + rb_erase(&e->start.node, &wnd->start_tree); + rb_erase(&e->count.node, &wnd->count_tree); + wnd->count -= 1; + } else { + e = kmem_cache_alloc(ntfs_enode_cachep, GFP_ATOMIC); + if (!e) + wnd->uptodated = -1; + } + + if (e) { + e->start.key = new_key; + e->count.key = new_len; + rb_insert_start(&wnd->start_tree, e); + rb_insert_count(&wnd->count_tree, e); + wnd->count += 1; + } + +out: + if (!wnd->count && 1 != wnd->uptodated) + wnd_rescan(wnd); +} + +/* + * wnd_rescan + * + * Scan all bitmap. used while initialization. + */ +static int wnd_rescan(struct wnd_bitmap *wnd) +{ + int err = 0; + size_t prev_tail = 0; + struct super_block *sb = wnd->sb; + struct ntfs_sb_info *sbi = sb->s_fs_info; + u64 lbo, len = 0; + u32 blocksize = sb->s_blocksize; + u8 cluster_bits = sbi->cluster_bits; + u32 wbits = 8 * sb->s_blocksize; + u32 used, frb; + const ulong *buf; + size_t wpos, wbit, iw, vbo; + struct buffer_head *bh = NULL; + CLST lcn, clen; + + wnd->uptodated = 0; + wnd->extent_max = 0; + wnd->extent_min = MINUS_ONE_T; + wnd->total_zeroes = 0; + + vbo = 0; + + for (iw = 0; iw < wnd->nwnd; iw++) { + if (iw + 1 == wnd->nwnd) + wbits = wnd->bits_last; + + if (wnd->inited) { + if (!wnd->free_bits[iw]) { + /* all ones */ + if (prev_tail) { + wnd_add_free_ext(wnd, + vbo * 8 - prev_tail, + prev_tail, true); + prev_tail = 0; + } + goto next_wnd; + } + if (wbits == wnd->free_bits[iw]) { + /* all zeroes */ + prev_tail += wbits; + wnd->total_zeroes += wbits; + goto next_wnd; + } + } + + if (!len) { + u32 off = vbo & sbi->cluster_mask; + + if (!run_lookup_entry(&wnd->run, vbo >> cluster_bits, + &lcn, &clen, NULL)) { + err = -ENOENT; + goto out; + } + + lbo = ((u64)lcn << cluster_bits) + off; + len = ((u64)clen << cluster_bits) - off; + } + + bh = ntfs_bread(sb, lbo >> sb->s_blocksize_bits); + if (!bh) { + err = -EIO; + goto out; + } + + buf = (ulong *)bh->b_data; + + used = __bitmap_weight(buf, wbits); + if (used < wbits) { + frb = wbits - used; + wnd->free_bits[iw] = frb; + wnd->total_zeroes += frb; + } + + wpos = 0; + wbit = vbo * 8; + + if (wbit + wbits > wnd->nbits) + wbits = wnd->nbits - wbit; + + do { + used = find_next_zero_bit(buf, wbits, wpos); + + if (used > wpos && prev_tail) { + wnd_add_free_ext(wnd, wbit + wpos - prev_tail, + prev_tail, true); + prev_tail = 0; + } + + wpos = used; + + if (wpos >= wbits) { + /* No free blocks */ + prev_tail = 0; + break; + } + + frb = find_next_bit(buf, wbits, wpos); + if (frb >= wbits) { + /* keep last free block */ + prev_tail += frb - wpos; + break; + } + + wnd_add_free_ext(wnd, wbit + wpos - prev_tail, + frb + prev_tail - wpos, true); + + /* Skip free block and first '1' */ + wpos = frb + 1; + /* Reset previous tail */ + prev_tail = 0; + } while (wpos < wbits); + +next_wnd: + + if (bh) + put_bh(bh); + bh = NULL; + + vbo += blocksize; + if (len) { + len -= blocksize; + lbo += blocksize; + } + } + + /* Add last block */ + if (prev_tail) + wnd_add_free_ext(wnd, wnd->nbits - prev_tail, prev_tail, true); + + /* + * Before init cycle wnd->uptodated was 0 + * If any errors or limits occurs while initialization then + * wnd->uptodated will be -1 + * If 'uptodated' is still 0 then Tree is really updated + */ + if (!wnd->uptodated) + wnd->uptodated = 1; + + if (wnd->zone_bit != wnd->zone_end) { + size_t zlen = wnd->zone_end - wnd->zone_bit; + + wnd->zone_end = wnd->zone_bit; + wnd_zone_set(wnd, wnd->zone_bit, zlen); + } + +out: + return err; +} + +/* + * wnd_init + */ +int wnd_init(struct wnd_bitmap *wnd, struct super_block *sb, size_t nbits) +{ + int err; + u32 blocksize = sb->s_blocksize; + u32 wbits = blocksize * 8; + + init_rwsem(&wnd->rw_lock); + + wnd->sb = sb; + wnd->nbits = nbits; + wnd->total_zeroes = nbits; + wnd->extent_max = MINUS_ONE_T; + wnd->zone_bit = wnd->zone_end = 0; + wnd->nwnd = bytes_to_block(sb, bitmap_size(nbits)); + wnd->bits_last = nbits & (wbits - 1); + if (!wnd->bits_last) + wnd->bits_last = wbits; + + wnd->free_bits = ntfs_zalloc(wnd->nwnd * sizeof(u16)); + if (!wnd->free_bits) + return -ENOMEM; + + err = wnd_rescan(wnd); + if (err) + return err; + + wnd->inited = true; + + return 0; +} + +/* + * wnd_map + * + * call sb_bread for requested window + */ +static struct buffer_head *wnd_map(struct wnd_bitmap *wnd, size_t iw) +{ + size_t vbo; + CLST lcn, clen; + struct super_block *sb = wnd->sb; + struct ntfs_sb_info *sbi; + struct buffer_head *bh; + u64 lbo; + + sbi = sb->s_fs_info; + vbo = (u64)iw << sb->s_blocksize_bits; + + if (!run_lookup_entry(&wnd->run, vbo >> sbi->cluster_bits, &lcn, &clen, + NULL)) { + return ERR_PTR(-ENOENT); + } + + lbo = ((u64)lcn << sbi->cluster_bits) + (vbo & sbi->cluster_mask); + + bh = ntfs_bread(wnd->sb, lbo >> sb->s_blocksize_bits); + if (!bh) + return ERR_PTR(-EIO); + + return bh; +} + +/* + * wnd_set_free + * + * Marks the bits range from bit to bit + bits as free + */ +int wnd_set_free(struct wnd_bitmap *wnd, size_t bit, size_t bits) +{ + int err = 0; + struct super_block *sb = wnd->sb; + size_t bits0 = bits; + u32 wbits = 8 * sb->s_blocksize; + size_t iw = bit >> (sb->s_blocksize_bits + 3); + u32 wbit = bit & (wbits - 1); + struct buffer_head *bh; + + while (iw < wnd->nwnd && bits) { + u32 tail, op; + ulong *buf; + + if (iw + 1 == wnd->nwnd) + wbits = wnd->bits_last; + + tail = wbits - wbit; + op = tail < bits ? tail : bits; + + bh = wnd_map(wnd, iw); + if (IS_ERR(bh)) { + err = PTR_ERR(bh); + break; + } + + buf = (ulong *)bh->b_data; + + lock_buffer(bh); + + __bitmap_clear(buf, wbit, op); + + wnd->free_bits[iw] += op; + + set_buffer_uptodate(bh); + mark_buffer_dirty(bh); + unlock_buffer(bh); + put_bh(bh); + + wnd->total_zeroes += op; + bits -= op; + wbit = 0; + iw += 1; + } + + wnd_add_free_ext(wnd, bit, bits0, false); + + return err; +} + +/* + * wnd_set_used + * + * Marks the bits range from bit to bit + bits as used + */ +int wnd_set_used(struct wnd_bitmap *wnd, size_t bit, size_t bits) +{ + int err = 0; + struct super_block *sb = wnd->sb; + size_t bits0 = bits; + size_t iw = bit >> (sb->s_blocksize_bits + 3); + u32 wbits = 8 * sb->s_blocksize; + u32 wbit = bit & (wbits - 1); + struct buffer_head *bh; + + while (iw < wnd->nwnd && bits) { + u32 tail, op; + ulong *buf; + + if (unlikely(iw + 1 == wnd->nwnd)) + wbits = wnd->bits_last; + + tail = wbits - wbit; + op = tail < bits ? tail : bits; + + bh = wnd_map(wnd, iw); + if (IS_ERR(bh)) { + err = PTR_ERR(bh); + break; + } + buf = (ulong *)bh->b_data; + + lock_buffer(bh); + + __bitmap_set(buf, wbit, op); + wnd->free_bits[iw] -= op; + + set_buffer_uptodate(bh); + mark_buffer_dirty(bh); + unlock_buffer(bh); + put_bh(bh); + + wnd->total_zeroes -= op; + bits -= op; + wbit = 0; + iw += 1; + } + + if (!RB_EMPTY_ROOT(&wnd->start_tree)) + wnd_remove_free_ext(wnd, bit, bits0); + + return err; +} + +/* + * wnd_is_free_hlp + * + * Returns true if all clusters [bit, bit+bits) are free (bitmap only) + */ +static bool wnd_is_free_hlp(struct wnd_bitmap *wnd, size_t bit, size_t bits) +{ + struct super_block *sb = wnd->sb; + size_t iw = bit >> (sb->s_blocksize_bits + 3); + u32 wbits = 8 * sb->s_blocksize; + u32 wbit = bit & (wbits - 1); + + while (iw < wnd->nwnd && bits) { + u32 tail, op; + + if (unlikely(iw + 1 == wnd->nwnd)) + wbits = wnd->bits_last; + + tail = wbits - wbit; + op = tail < bits ? tail : bits; + + if (wbits != wnd->free_bits[iw]) { + bool ret; + struct buffer_head *bh = wnd_map(wnd, iw); + + if (IS_ERR(bh)) + return false; + + ret = are_bits_clear((ulong *)bh->b_data, wbit, op); + + put_bh(bh); + if (!ret) + return false; + } + + bits -= op; + wbit = 0; + iw += 1; + } + + return true; +} + +/* + * wnd_is_free + * + * Returns true if all clusters [bit, bit+bits) are free + */ +bool wnd_is_free(struct wnd_bitmap *wnd, size_t bit, size_t bits) +{ + bool ret; + struct rb_node *n; + size_t end; + struct e_node *e; + + if (RB_EMPTY_ROOT(&wnd->start_tree)) + goto use_wnd; + + n = rb_lookup(&wnd->start_tree, bit); + if (!n) + goto use_wnd; + + e = rb_entry(n, struct e_node, start.node); + + end = e->start.key + e->count.key; + + if (bit < end && bit + bits <= end) + return true; + +use_wnd: + ret = wnd_is_free_hlp(wnd, bit, bits); + + return ret; +} + +/* + * wnd_is_used + * + * Returns true if all clusters [bit, bit+bits) are used + */ +bool wnd_is_used(struct wnd_bitmap *wnd, size_t bit, size_t bits) +{ + bool ret = false; + struct super_block *sb = wnd->sb; + size_t iw = bit >> (sb->s_blocksize_bits + 3); + u32 wbits = 8 * sb->s_blocksize; + u32 wbit = bit & (wbits - 1); + size_t end; + struct rb_node *n; + struct e_node *e; + + if (RB_EMPTY_ROOT(&wnd->start_tree)) + goto use_wnd; + + end = bit + bits; + n = rb_lookup(&wnd->start_tree, end - 1); + if (!n) + goto use_wnd; + + e = rb_entry(n, struct e_node, start.node); + if (e->start.key + e->count.key > bit) + return false; + +use_wnd: + while (iw < wnd->nwnd && bits) { + u32 tail, op; + + if (unlikely(iw + 1 == wnd->nwnd)) + wbits = wnd->bits_last; + + tail = wbits - wbit; + op = tail < bits ? tail : bits; + + if (wnd->free_bits[iw]) { + bool ret; + struct buffer_head *bh = wnd_map(wnd, iw); + + if (IS_ERR(bh)) + goto out; + + ret = are_bits_set((ulong *)bh->b_data, wbit, op); + put_bh(bh); + if (!ret) + goto out; + } + + bits -= op; + wbit = 0; + iw += 1; + } + ret = true; + +out: + return ret; +} + +/* + * wnd_find + * - flags - BITMAP_FIND_XXX flags + * + * looks for free space + * Returns 0 if not found + */ +size_t wnd_find(struct wnd_bitmap *wnd, size_t to_alloc, size_t hint, + size_t flags, size_t *allocated) +{ + struct super_block *sb; + u32 wbits, wpos, wzbit, wzend; + size_t fnd, max_alloc, b_len, b_pos; + size_t iw, prev_tail, nwnd, wbit, ebit, zbit, zend; + size_t to_alloc0 = to_alloc; + const ulong *buf; + const struct e_node *e; + const struct rb_node *pr, *cr; + u8 log2_bits; + bool fbits_valid; + struct buffer_head *bh; + + /* fast checking for available free space */ + if (flags & BITMAP_FIND_FULL) { + size_t zeroes = wnd_zeroes(wnd); + + zeroes -= wnd->zone_end - wnd->zone_bit; + if (zeroes < to_alloc0) + goto no_space; + + if (to_alloc0 > wnd->extent_max) + goto no_space; + } else { + if (to_alloc > wnd->extent_max) + to_alloc = wnd->extent_max; + } + + if (wnd->zone_bit <= hint && hint < wnd->zone_end) + hint = wnd->zone_end; + + max_alloc = wnd->nbits; + b_len = b_pos = 0; + + if (hint >= max_alloc) + hint = 0; + + if (RB_EMPTY_ROOT(&wnd->start_tree)) { + if (wnd->uptodated == 1) { + /* extents tree is updated -> no free space */ + goto no_space; + } + goto scan_bitmap; + } + + e = NULL; + if (!hint) + goto allocate_biggest; + + /* Use hint: enumerate extents by start >= hint */ + pr = NULL; + cr = wnd->start_tree.rb_node; + + for (;;) { + e = rb_entry(cr, struct e_node, start.node); + + if (e->start.key == hint) + break; + + if (e->start.key < hint) { + pr = cr; + cr = cr->rb_right; + if (!cr) + break; + continue; + } + + cr = cr->rb_left; + if (!cr) { + e = pr ? rb_entry(pr, struct e_node, start.node) : NULL; + break; + } + } + + if (!e) + goto allocate_biggest; + + if (e->start.key + e->count.key > hint) { + /* We have found extension with 'hint' inside */ + size_t len = e->start.key + e->count.key - hint; + + if (len >= to_alloc && hint + to_alloc <= max_alloc) { + fnd = hint; + goto found; + } + + if (!(flags & BITMAP_FIND_FULL)) { + if (len > to_alloc) + len = to_alloc; + + if (hint + len <= max_alloc) { + fnd = hint; + to_alloc = len; + goto found; + } + } + } + +allocate_biggest: + /* Allocate from biggest free extent */ + e = rb_entry(rb_first(&wnd->count_tree), struct e_node, count.node); + if (e->count.key != wnd->extent_max) + wnd->extent_max = e->count.key; + + if (e->count.key < max_alloc) { + if (e->count.key >= to_alloc) { + ; + } else if (flags & BITMAP_FIND_FULL) { + if (e->count.key < to_alloc0) { + /* Biggest free block is less then requested */ + goto no_space; + } + to_alloc = e->count.key; + } else if (-1 != wnd->uptodated) { + to_alloc = e->count.key; + } else { + /* Check if we can use more bits */ + size_t op, max_check; + struct rb_root start_tree; + + memcpy(&start_tree, &wnd->start_tree, + sizeof(struct rb_root)); + memset(&wnd->start_tree, 0, sizeof(struct rb_root)); + + max_check = e->start.key + to_alloc; + if (max_check > max_alloc) + max_check = max_alloc; + for (op = e->start.key + e->count.key; op < max_check; + op++) { + if (!wnd_is_free(wnd, op, 1)) + break; + } + memcpy(&wnd->start_tree, &start_tree, + sizeof(struct rb_root)); + to_alloc = op - e->start.key; + } + + /* Prepare to return */ + fnd = e->start.key; + if (e->start.key + to_alloc > max_alloc) + to_alloc = max_alloc - e->start.key; + goto found; + } + + if (wnd->uptodated == 1) { + /* extents tree is updated -> no free space */ + goto no_space; + } + + b_len = e->count.key; + b_pos = e->start.key; + +scan_bitmap: + sb = wnd->sb; + log2_bits = sb->s_blocksize_bits + 3; + + /* At most two ranges [hint, max_alloc) + [0, hint) */ +Again: + + /* TODO: optimize request for case nbits > wbits */ + iw = hint >> log2_bits; + wbits = sb->s_blocksize * 8; + wpos = hint & (wbits - 1); + prev_tail = 0; + fbits_valid = true; + + if (max_alloc == wnd->nbits) { + nwnd = wnd->nwnd; + } else { + size_t t = max_alloc + wbits - 1; + + nwnd = likely(t > max_alloc) ? (t >> log2_bits) : wnd->nwnd; + } + + /* Enumerate all windows */ + for (; iw < nwnd; iw++) { + wbit = iw << log2_bits; + + if (!wnd->free_bits[iw]) { + if (prev_tail > b_len) { + b_pos = wbit - prev_tail; + b_len = prev_tail; + } + + /* Skip full used window */ + prev_tail = 0; + wpos = 0; + continue; + } + + if (unlikely(iw + 1 == nwnd)) { + if (max_alloc == wnd->nbits) { + wbits = wnd->bits_last; + } else { + size_t t = max_alloc & (wbits - 1); + + if (t) { + wbits = t; + fbits_valid = false; + } + } + } + + if (wnd->zone_end > wnd->zone_bit) { + ebit = wbit + wbits; + zbit = max(wnd->zone_bit, wbit); + zend = min(wnd->zone_end, ebit); + + /* Here we have a window [wbit, ebit) and zone [zbit, zend) */ + if (zend <= zbit) { + /* Zone does not overlap window */ + } else { + wzbit = zbit - wbit; + wzend = zend - wbit; + + /* Zone overlaps window */ + if (wnd->free_bits[iw] == wzend - wzbit) { + prev_tail = 0; + wpos = 0; + continue; + } + + /* Scan two ranges window: [wbit, zbit) and [zend, ebit) */ + bh = wnd_map(wnd, iw); + + if (IS_ERR(bh)) { + /* TODO: error */ + prev_tail = 0; + wpos = 0; + continue; + } + + buf = (ulong *)bh->b_data; + + /* Scan range [wbit, zbit) */ + if (wpos < wzbit) { + /* Scan range [wpos, zbit) */ + fnd = wnd_scan(buf, wbit, wpos, wzbit, + to_alloc, &prev_tail, + &b_pos, &b_len); + if (fnd != MINUS_ONE_T) { + put_bh(bh); + goto found; + } + } + + prev_tail = 0; + + /* Scan range [zend, ebit) */ + if (wzend < wbits) { + fnd = wnd_scan(buf, wbit, + max(wzend, wpos), wbits, + to_alloc, &prev_tail, + &b_pos, &b_len); + if (fnd != MINUS_ONE_T) { + put_bh(bh); + goto found; + } + } + + wpos = 0; + put_bh(bh); + continue; + } + } + + /* Current window does not overlap zone */ + if (!wpos && fbits_valid && wnd->free_bits[iw] == wbits) { + /* window is empty */ + if (prev_tail + wbits >= to_alloc) { + fnd = wbit + wpos - prev_tail; + goto found; + } + + /* Increase 'prev_tail' and process next window */ + prev_tail += wbits; + wpos = 0; + continue; + } + + /* read window */ + bh = wnd_map(wnd, iw); + if (IS_ERR(bh)) { + // TODO: error + prev_tail = 0; + wpos = 0; + continue; + } + + buf = (ulong *)bh->b_data; + + /* Scan range [wpos, eBits) */ + fnd = wnd_scan(buf, wbit, wpos, wbits, to_alloc, &prev_tail, + &b_pos, &b_len); + put_bh(bh); + if (fnd != MINUS_ONE_T) + goto found; + } + + if (b_len < prev_tail) { + /* The last fragment */ + b_len = prev_tail; + b_pos = max_alloc - prev_tail; + } + + if (hint) { + /* + * We have scanned range [hint max_alloc) + * Prepare to scan range [0 hint + to_alloc) + */ + size_t nextmax = hint + to_alloc; + + if (likely(nextmax >= hint) && nextmax < max_alloc) + max_alloc = nextmax; + hint = 0; + goto Again; + } + + if (!b_len) + goto no_space; + + wnd->extent_max = b_len; + + if (flags & BITMAP_FIND_FULL) + goto no_space; + + fnd = b_pos; + to_alloc = b_len; + +found: + if (flags & BITMAP_FIND_MARK_AS_USED) { + /* TODO optimize remove extent (pass 'e'?) */ + if (wnd_set_used(wnd, fnd, to_alloc)) + goto no_space; + } else if (wnd->extent_max != MINUS_ONE_T && + to_alloc > wnd->extent_max) { + wnd->extent_max = to_alloc; + } + + *allocated = fnd; + return to_alloc; + +no_space: + return 0; +} + +/* + * wnd_extend + * + * Extend bitmap ($MFT bitmap) + */ +int wnd_extend(struct wnd_bitmap *wnd, size_t new_bits) +{ + int err; + struct super_block *sb = wnd->sb; + struct ntfs_sb_info *sbi = sb->s_fs_info; + u32 blocksize = sb->s_blocksize; + u32 wbits = blocksize * 8; + u32 b0, new_last; + size_t bits, iw, new_wnd; + size_t old_bits = wnd->nbits; + u16 *new_free; + + if (new_bits <= old_bits) + return -EINVAL; + + /* align to 8 byte boundary */ + new_wnd = bytes_to_block(sb, bitmap_size(new_bits)); + new_last = new_bits & (wbits - 1); + if (!new_last) + new_last = wbits; + + if (new_wnd != wnd->nwnd) { + new_free = ntfs_malloc(new_wnd * sizeof(u16)); + if (!new_free) + return -ENOMEM; + + if (new_free != wnd->free_bits) + memcpy(new_free, wnd->free_bits, + wnd->nwnd * sizeof(short)); + memset(new_free + wnd->nwnd, 0, + (new_wnd - wnd->nwnd) * sizeof(short)); + ntfs_free(wnd->free_bits); + wnd->free_bits = new_free; + } + + /* Zero bits [old_bits,new_bits) */ + bits = new_bits - old_bits; + b0 = old_bits & (wbits - 1); + + for (iw = old_bits >> (sb->s_blocksize_bits + 3); bits; iw += 1) { + u32 op; + size_t frb; + u64 vbo, lbo, bytes; + struct buffer_head *bh; + ulong *buf; + + if (iw + 1 == new_wnd) + wbits = new_last; + + op = b0 + bits > wbits ? wbits - b0 : bits; + vbo = (u64)iw * blocksize; + + err = ntfs_vbo_to_lbo(sbi, &wnd->run, vbo, &lbo, &bytes); + if (err) + break; + + bh = ntfs_bread(sb, lbo >> sb->s_blocksize_bits); + if (!bh) + return -EIO; + + lock_buffer(bh); + buf = (ulong *)bh->b_data; + + __bitmap_clear(buf, b0, blocksize * 8 - b0); + frb = wbits - __bitmap_weight(buf, wbits); + wnd->total_zeroes += frb - wnd->free_bits[iw]; + wnd->free_bits[iw] = frb; + + set_buffer_uptodate(bh); + mark_buffer_dirty(bh); + unlock_buffer(bh); + /*err = sync_dirty_buffer(bh);*/ + + b0 = 0; + bits -= op; + } + + wnd->nbits = new_bits; + wnd->nwnd = new_wnd; + wnd->bits_last = new_last; + + wnd_add_free_ext(wnd, old_bits, new_bits - old_bits, false); + + return 0; +} + +/* + * wnd_zone_set + */ +void wnd_zone_set(struct wnd_bitmap *wnd, size_t lcn, size_t len) +{ + size_t zlen; + + zlen = wnd->zone_end - wnd->zone_bit; + if (zlen) + wnd_add_free_ext(wnd, wnd->zone_bit, zlen, false); + + if (!RB_EMPTY_ROOT(&wnd->start_tree) && len) + wnd_remove_free_ext(wnd, lcn, len); + + wnd->zone_bit = lcn; + wnd->zone_end = lcn + len; +} + +int ntfs_trim_fs(struct ntfs_sb_info *sbi, struct fstrim_range *range) +{ + int err = 0; + struct super_block *sb = sbi->sb; + struct wnd_bitmap *wnd = &sbi->used.bitmap; + u32 wbits = 8 * sb->s_blocksize; + CLST len = 0, lcn = 0, done = 0; + CLST minlen = bytes_to_cluster(sbi, range->minlen); + CLST lcn_from = bytes_to_cluster(sbi, range->start); + size_t iw = lcn_from >> (sb->s_blocksize_bits + 3); + u32 wbit = lcn_from & (wbits - 1); + const ulong *buf; + CLST lcn_to; + + if (!minlen) + minlen = 1; + + if (range->len == (u64)-1) + lcn_to = wnd->nbits; + else + lcn_to = bytes_to_cluster(sbi, range->start + range->len); + + down_read_nested(&wnd->rw_lock, BITMAP_MUTEX_CLUSTERS); + + for (; iw < wnd->nbits; iw++, wbit = 0) { + CLST lcn_wnd = iw * wbits; + struct buffer_head *bh; + + if (lcn_wnd > lcn_to) + break; + + if (!wnd->free_bits[iw]) + continue; + + if (iw + 1 == wnd->nwnd) + wbits = wnd->bits_last; + + if (lcn_wnd + wbits > lcn_to) + wbits = lcn_to - lcn_wnd; + + bh = wnd_map(wnd, iw); + if (IS_ERR(bh)) { + err = PTR_ERR(bh); + break; + } + + buf = (ulong *)bh->b_data; + + for (; wbit < wbits; wbit++) { + if (!test_bit(wbit, buf)) { + if (!len) + lcn = lcn_wnd + wbit; + len += 1; + continue; + } + if (len >= minlen) { + err = ntfs_discard(sbi, lcn, len); + if (err) + goto out; + done += len; + } + len = 0; + } + put_bh(bh); + } + + /* Process the last fragment */ + if (len >= minlen) { + err = ntfs_discard(sbi, lcn, len); + if (err) + goto out; + done += len; + } + +out: + range->len = (u64)done << sbi->cluster_bits; + + up_read(&wnd->rw_lock); + + return err; +} -- 2.30.0

1 0

[PATCH OLK-5.10 002/107] fs/ntfs3: Add initialization of super block
by Yin Xiujiang 08 Dec '21

08 Dec '21

From: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> mainline inclusion from mainline-v5.15-rc1 commit 82cae269cfa953032fbb8980a7d554d60fb00b17 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4G67J?from=project-issue CVE: NA ---------------------------------------------------------------------- This adds initialization of super block Signed-off-by: Konstantin Komarov <almaz.alexandrovich(a)paragon-software.com> Signed-off-by: Yin Xiujiang <yinxiujiang(a)kylinos.cn> --- fs/ntfs3/fsntfs.c | 2551 +++++++++++++++++++++++++++++++++++++++++++ fs/ntfs3/index.c | 2647 +++++++++++++++++++++++++++++++++++++++++++++ fs/ntfs3/inode.c | 2029 ++++++++++++++++++++++++++++++++++ fs/ntfs3/super.c | 1504 ++++++++++++++++++++++++++ 4 files changed, 8731 insertions(+) create mode 100644 fs/ntfs3/fsntfs.c create mode 100644 fs/ntfs3/index.c create mode 100644 fs/ntfs3/inode.c create mode 100644 fs/ntfs3/super.c diff --git a/fs/ntfs3/fsntfs.c b/fs/ntfs3/fsntfs.c new file mode 100644 index 000000000000..92140050fb6c --- /dev/null +++ b/fs/ntfs3/fsntfs.c @@ -0,0 +1,2551 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * + * Copyright (C) 2019-2021 Paragon Software GmbH, All rights reserved. + * + */ + +#include <linux/blkdev.h> +#include <linux/buffer_head.h> +#include <linux/fs.h> +#include <linux/nls.h> + +#include "debug.h" +#include "ntfs.h" +#include "ntfs_fs.h" + +// clang-format off +const struct cpu_str NAME_MFT = { + 4, 0, { '$', 'M', 'F', 'T' }, +}; +const struct cpu_str NAME_MIRROR = { + 8, 0, { '$', 'M', 'F', 'T', 'M', 'i', 'r', 'r' }, +}; +const struct cpu_str NAME_LOGFILE = { + 8, 0, { '$', 'L', 'o', 'g', 'F', 'i', 'l', 'e' }, +}; +const struct cpu_str NAME_VOLUME = { + 7, 0, { '$', 'V', 'o', 'l', 'u', 'm', 'e' }, +}; +const struct cpu_str NAME_ATTRDEF = { + 8, 0, { '$', 'A', 't', 't', 'r', 'D', 'e', 'f' }, +}; +const struct cpu_str NAME_ROOT = { + 1, 0, { '.' }, +}; +const struct cpu_str NAME_BITMAP = { + 7, 0, { '$', 'B', 'i', 't', 'm', 'a', 'p' }, +}; +const struct cpu_str NAME_BOOT = { + 5, 0, { '$', 'B', 'o', 'o', 't' }, +}; +const struct cpu_str NAME_BADCLUS = { + 8, 0, { '$', 'B', 'a', 'd', 'C', 'l', 'u', 's' }, +}; +const struct cpu_str NAME_QUOTA = { + 6, 0, { '$', 'Q', 'u', 'o', 't', 'a' }, +}; +const struct cpu_str NAME_SECURE = { + 7, 0, { '$', 'S', 'e', 'c', 'u', 'r', 'e' }, +}; +const struct cpu_str NAME_UPCASE = { + 7, 0, { '$', 'U', 'p', 'C', 'a', 's', 'e' }, +}; +const struct cpu_str NAME_EXTEND = { + 7, 0, { '$', 'E', 'x', 't', 'e', 'n', 'd' }, +}; +const struct cpu_str NAME_OBJID = { + 6, 0, { '$', 'O', 'b', 'j', 'I', 'd' }, +}; +const struct cpu_str NAME_REPARSE = { + 8, 0, { '$', 'R', 'e', 'p', 'a', 'r', 's', 'e' }, +}; +const struct cpu_str NAME_USNJRNL = { + 8, 0, { '$', 'U', 's', 'n', 'J', 'r', 'n', 'l' }, +}; +const __le16 BAD_NAME[4] = { + cpu_to_le16('$'), cpu_to_le16('B'), cpu_to_le16('a'), cpu_to_le16('d'), +}; +const __le16 I30_NAME[4] = { + cpu_to_le16('$'), cpu_to_le16('I'), cpu_to_le16('3'), cpu_to_le16('0'), +}; +const __le16 SII_NAME[4] = { + cpu_to_le16('$'), cpu_to_le16('S'), cpu_to_le16('I'), cpu_to_le16('I'), +}; +const __le16 SDH_NAME[4] = { + cpu_to_le16('$'), cpu_to_le16('S'), cpu_to_le16('D'), cpu_to_le16('H'), +}; +const __le16 SDS_NAME[4] = { + cpu_to_le16('$'), cpu_to_le16('S'), cpu_to_le16('D'), cpu_to_le16('S'), +}; +const __le16 SO_NAME[2] = { + cpu_to_le16('$'), cpu_to_le16('O'), +}; +const __le16 SQ_NAME[2] = { + cpu_to_le16('$'), cpu_to_le16('Q'), +}; +const __le16 SR_NAME[2] = { + cpu_to_le16('$'), cpu_to_le16('R'), +}; + +#ifdef CONFIG_NTFS3_LZX_XPRESS +const __le16 WOF_NAME[17] = { + cpu_to_le16('W'), cpu_to_le16('o'), cpu_to_le16('f'), cpu_to_le16('C'), + cpu_to_le16('o'), cpu_to_le16('m'), cpu_to_le16('p'), cpu_to_le16('r'), + cpu_to_le16('e'), cpu_to_le16('s'), cpu_to_le16('s'), cpu_to_le16('e'), + cpu_to_le16('d'), cpu_to_le16('D'), cpu_to_le16('a'), cpu_to_le16('t'), + cpu_to_le16('a'), +}; +#endif + +// clang-format on + +/* + * ntfs_fix_pre_write + * + * inserts fixups into 'rhdr' before writing to disk + */ +bool ntfs_fix_pre_write(struct NTFS_RECORD_HEADER *rhdr, size_t bytes) +{ + u16 *fixup, *ptr; + u16 sample; + u16 fo = le16_to_cpu(rhdr->fix_off); + u16 fn = le16_to_cpu(rhdr->fix_num); + + if ((fo & 1) || fo + fn * sizeof(short) > SECTOR_SIZE || !fn-- || + fn * SECTOR_SIZE > bytes) { + return false; + } + + /* Get fixup pointer */ + fixup = Add2Ptr(rhdr, fo); + + if (*fixup >= 0x7FFF) + *fixup = 1; + else + *fixup += 1; + + sample = *fixup; + + ptr = Add2Ptr(rhdr, SECTOR_SIZE - sizeof(short)); + + while (fn--) { + *++fixup = *ptr; + *ptr = sample; + ptr += SECTOR_SIZE / sizeof(short); + } + return true; +} + +/* + * ntfs_fix_post_read + * + * remove fixups after reading from disk + * Returns < 0 if error, 0 if ok, 1 if need to update fixups + */ +int ntfs_fix_post_read(struct NTFS_RECORD_HEADER *rhdr, size_t bytes, + bool simple) +{ + int ret; + u16 *fixup, *ptr; + u16 sample, fo, fn; + + fo = le16_to_cpu(rhdr->fix_off); + fn = simple ? ((bytes >> SECTOR_SHIFT) + 1) + : le16_to_cpu(rhdr->fix_num); + + /* Check errors */ + if ((fo & 1) || fo + fn * sizeof(short) > SECTOR_SIZE || !fn-- || + fn * SECTOR_SIZE > bytes) { + return -EINVAL; /* native chkntfs returns ok! */ + } + + /* Get fixup pointer */ + fixup = Add2Ptr(rhdr, fo); + sample = *fixup; + ptr = Add2Ptr(rhdr, SECTOR_SIZE - sizeof(short)); + ret = 0; + + while (fn--) { + /* Test current word */ + if (*ptr != sample) { + /* Fixup does not match! Is it serious error? */ + ret = -E_NTFS_FIXUP; + } + + /* Replace fixup */ + *ptr = *++fixup; + ptr += SECTOR_SIZE / sizeof(short); + } + + return ret; +} + +/* + * ntfs_extend_init + * + * loads $Extend file + */ +int ntfs_extend_init(struct ntfs_sb_info *sbi) +{ + int err; + struct super_block *sb = sbi->sb; + struct inode *inode, *inode2; + struct MFT_REF ref; + + if (sbi->volume.major_ver < 3) { + ntfs_notice(sb, "Skip $Extend 'cause NTFS version"); + return 0; + } + + ref.low = cpu_to_le32(MFT_REC_EXTEND); + ref.high = 0; + ref.seq = cpu_to_le16(MFT_REC_EXTEND); + inode = ntfs_iget5(sb, &ref, &NAME_EXTEND); + if (IS_ERR(inode)) { + err = PTR_ERR(inode); + ntfs_err(sb, "Failed to load $Extend."); + inode = NULL; + goto out; + } + + /* if ntfs_iget5 reads from disk it never returns bad inode */ + if (!S_ISDIR(inode->i_mode)) { + err = -EINVAL; + goto out; + } + + /* Try to find $ObjId */ + inode2 = dir_search_u(inode, &NAME_OBJID, NULL); + if (inode2 && !IS_ERR(inode2)) { + if (is_bad_inode(inode2)) { + iput(inode2); + } else { + sbi->objid.ni = ntfs_i(inode2); + sbi->objid_no = inode2->i_ino; + } + } + + /* Try to find $Quota */ + inode2 = dir_search_u(inode, &NAME_QUOTA, NULL); + if (inode2 && !IS_ERR(inode2)) { + sbi->quota_no = inode2->i_ino; + iput(inode2); + } + + /* Try to find $Reparse */ + inode2 = dir_search_u(inode, &NAME_REPARSE, NULL); + if (inode2 && !IS_ERR(inode2)) { + sbi->reparse.ni = ntfs_i(inode2); + sbi->reparse_no = inode2->i_ino; + } + + /* Try to find $UsnJrnl */ + inode2 = dir_search_u(inode, &NAME_USNJRNL, NULL); + if (inode2 && !IS_ERR(inode2)) { + sbi->usn_jrnl_no = inode2->i_ino; + iput(inode2); + } + + err = 0; +out: + iput(inode); + return err; +} + +int ntfs_loadlog_and_replay(struct ntfs_inode *ni, struct ntfs_sb_info *sbi) +{ + int err = 0; + struct super_block *sb = sbi->sb; + bool initialized = false; + struct MFT_REF ref; + struct inode *inode; + + /* Check for 4GB */ + if (ni->vfs_inode.i_size >= 0x100000000ull) { + ntfs_err(sb, "\x24LogFile is too big"); + err = -EINVAL; + goto out; + } + + sbi->flags |= NTFS_FLAGS_LOG_REPLAYING; + + ref.low = cpu_to_le32(MFT_REC_MFT); + ref.high = 0; + ref.seq = cpu_to_le16(1); + + inode = ntfs_iget5(sb, &ref, NULL); + + if (IS_ERR(inode)) + inode = NULL; + + if (!inode) { + /* Try to use mft copy */ + u64 t64 = sbi->mft.lbo; + + sbi->mft.lbo = sbi->mft.lbo2; + inode = ntfs_iget5(sb, &ref, NULL); + sbi->mft.lbo = t64; + if (IS_ERR(inode)) + inode = NULL; + } + + if (!inode) { + err = -EINVAL; + ntfs_err(sb, "Failed to load $MFT."); + goto out; + } + + sbi->mft.ni = ntfs_i(inode); + + /* LogFile should not contains attribute list */ + err = ni_load_all_mi(sbi->mft.ni); + if (!err) + err = log_replay(ni, &initialized); + + iput(inode); + sbi->mft.ni = NULL; + + sync_blockdev(sb->s_bdev); + invalidate_bdev(sb->s_bdev); + + if (sbi->flags & NTFS_FLAGS_NEED_REPLAY) { + err = 0; + goto out; + } + + if (sb_rdonly(sb) || !initialized) + goto out; + + /* fill LogFile by '-1' if it is initialized */ + err = ntfs_bio_fill_1(sbi, &ni->file.run); + +out: + sbi->flags &= ~NTFS_FLAGS_LOG_REPLAYING; + + return err; +} + +/* + * ntfs_query_def + * + * returns current ATTR_DEF_ENTRY for given attribute type + */ +const struct ATTR_DEF_ENTRY *ntfs_query_def(struct ntfs_sb_info *sbi, + enum ATTR_TYPE type) +{ + int type_in = le32_to_cpu(type); + size_t min_idx = 0; + size_t max_idx = sbi->def_entries - 1; + + while (min_idx <= max_idx) { + size_t i = min_idx + ((max_idx - min_idx) >> 1); + const struct ATTR_DEF_ENTRY *entry = sbi->def_table + i; + int diff = le32_to_cpu(entry->type) - type_in; + + if (!diff) + return entry; + if (diff < 0) + min_idx = i + 1; + else if (i) + max_idx = i - 1; + else + return NULL; + } + return NULL; +} + +/* + * ntfs_look_for_free_space + * + * looks for a free space in bitmap + */ +int ntfs_look_for_free_space(struct ntfs_sb_info *sbi, CLST lcn, CLST len, + CLST *new_lcn, CLST *new_len, + enum ALLOCATE_OPT opt) +{ + int err; + struct super_block *sb = sbi->sb; + size_t a_lcn, zlen, zeroes, zlcn, zlen2, ztrim, new_zlen; + struct wnd_bitmap *wnd = &sbi->used.bitmap; + + down_write_nested(&wnd->rw_lock, BITMAP_MUTEX_CLUSTERS); + if (opt & ALLOCATE_MFT) { + CLST alen; + + zlen = wnd_zone_len(wnd); + + if (!zlen) { + err = ntfs_refresh_zone(sbi); + if (err) + goto out; + + zlen = wnd_zone_len(wnd); + + if (!zlen) { + ntfs_err(sbi->sb, + "no free space to extend mft"); + err = -ENOSPC; + goto out; + } + } + + lcn = wnd_zone_bit(wnd); + alen = zlen > len ? len : zlen; + + wnd_zone_set(wnd, lcn + alen, zlen - alen); + + err = wnd_set_used(wnd, lcn, alen); + if (err) + goto out; + + *new_lcn = lcn; + *new_len = alen; + goto ok; + } + + /* + * 'Cause cluster 0 is always used this value means that we should use + * cached value of 'next_free_lcn' to improve performance + */ + if (!lcn) + lcn = sbi->used.next_free_lcn; + + if (lcn >= wnd->nbits) + lcn = 0; + + *new_len = wnd_find(wnd, len, lcn, BITMAP_FIND_MARK_AS_USED, &a_lcn); + if (*new_len) { + *new_lcn = a_lcn; + goto ok; + } + + /* Try to use clusters from MftZone */ + zlen = wnd_zone_len(wnd); + zeroes = wnd_zeroes(wnd); + + /* Check too big request */ + if (len > zeroes + zlen) + goto no_space; + + if (zlen <= NTFS_MIN_MFT_ZONE) + goto no_space; + + /* How many clusters to cat from zone */ + zlcn = wnd_zone_bit(wnd); + zlen2 = zlen >> 1; + ztrim = len > zlen ? zlen : (len > zlen2 ? len : zlen2); + new_zlen = zlen - ztrim; + + if (new_zlen < NTFS_MIN_MFT_ZONE) { + new_zlen = NTFS_MIN_MFT_ZONE; + if (new_zlen > zlen) + new_zlen = zlen; + } + + wnd_zone_set(wnd, zlcn, new_zlen); + + /* allocate continues clusters */ + *new_len = + wnd_find(wnd, len, 0, + BITMAP_FIND_MARK_AS_USED | BITMAP_FIND_FULL, &a_lcn); + if (*new_len) { + *new_lcn = a_lcn; + goto ok; + } + +no_space: + up_write(&wnd->rw_lock); + + return -ENOSPC; + +ok: + err = 0; + + ntfs_unmap_meta(sb, *new_lcn, *new_len); + + if (opt & ALLOCATE_MFT) + goto out; + + /* Set hint for next requests */ + sbi->used.next_free_lcn = *new_lcn + *new_len; + +out: + up_write(&wnd->rw_lock); + return err; +} + +/* + * ntfs_extend_mft + * + * allocates additional MFT records + * sbi->mft.bitmap is locked for write + * + * NOTE: recursive: + * ntfs_look_free_mft -> + * ntfs_extend_mft -> + * attr_set_size -> + * ni_insert_nonresident -> + * ni_insert_attr -> + * ni_ins_attr_ext -> + * ntfs_look_free_mft -> + * ntfs_extend_mft + * To avoid recursive always allocate space for two new mft records + * see attrib.c: "at least two mft to avoid recursive loop" + */ +static int ntfs_extend_mft(struct ntfs_sb_info *sbi) +{ + int err; + struct ntfs_inode *ni = sbi->mft.ni; + size_t new_mft_total; + u64 new_mft_bytes, new_bitmap_bytes; + struct ATTRIB *attr; + struct wnd_bitmap *wnd = &sbi->mft.bitmap; + + new_mft_total = (wnd->nbits + MFT_INCREASE_CHUNK + 127) & (CLST)~127; + new_mft_bytes = (u64)new_mft_total << sbi->record_bits; + + /* Step 1: Resize $MFT::DATA */ + down_write(&ni->file.run_lock); + err = attr_set_size(ni, ATTR_DATA, NULL, 0, &ni->file.run, + new_mft_bytes, NULL, false, &attr); + + if (err) { + up_write(&ni->file.run_lock); + goto out; + } + + attr->nres.valid_size = attr->nres.data_size; + new_mft_total = le64_to_cpu(attr->nres.alloc_size) >> sbi->record_bits; + ni->mi.dirty = true; + + /* Step 2: Resize $MFT::BITMAP */ + new_bitmap_bytes = bitmap_size(new_mft_total); + + err = attr_set_size(ni, ATTR_BITMAP, NULL, 0, &sbi->mft.bitmap.run, + new_bitmap_bytes, &new_bitmap_bytes, true, NULL); + + /* Refresh Mft Zone if necessary */ + down_write_nested(&sbi->used.bitmap.rw_lock, BITMAP_MUTEX_CLUSTERS); + + ntfs_refresh_zone(sbi); + + up_write(&sbi->used.bitmap.rw_lock); + up_write(&ni->file.run_lock); + + if (err) + goto out; + + err = wnd_extend(wnd, new_mft_total); + + if (err) + goto out; + + ntfs_clear_mft_tail(sbi, sbi->mft.used, new_mft_total); + + err = _ni_write_inode(&ni->vfs_inode, 0); +out: + return err; +} + +/* + * ntfs_look_free_mft + * + * looks for a free MFT record + */ +int ntfs_look_free_mft(struct ntfs_sb_info *sbi, CLST *rno, bool mft, + struct ntfs_inode *ni, struct mft_inode **mi) +{ + int err = 0; + size_t zbit, zlen, from, to, fr; + size_t mft_total; + struct MFT_REF ref; + struct super_block *sb = sbi->sb; + struct wnd_bitmap *wnd = &sbi->mft.bitmap; + u32 ir; + + static_assert(sizeof(sbi->mft.reserved_bitmap) * 8 >= + MFT_REC_FREE - MFT_REC_RESERVED); + + if (!mft) + down_write_nested(&wnd->rw_lock, BITMAP_MUTEX_MFT); + + zlen = wnd_zone_len(wnd); + + /* Always reserve space for MFT */ + if (zlen) { + if (mft) { + zbit = wnd_zone_bit(wnd); + *rno = zbit; + wnd_zone_set(wnd, zbit + 1, zlen - 1); + } + goto found; + } + + /* No MFT zone. find the nearest to '0' free MFT */ + if (!wnd_find(wnd, 1, MFT_REC_FREE, 0, &zbit)) { + /* Resize MFT */ + mft_total = wnd->nbits; + + err = ntfs_extend_mft(sbi); + if (!err) { + zbit = mft_total; + goto reserve_mft; + } + + if (!mft || MFT_REC_FREE == sbi->mft.next_reserved) + goto out; + + err = 0; + + /* + * Look for free record reserved area [11-16) == + * [MFT_REC_RESERVED, MFT_REC_FREE ) MFT bitmap always + * marks it as used + */ + if (!sbi->mft.reserved_bitmap) { + /* Once per session create internal bitmap for 5 bits */ + sbi->mft.reserved_bitmap = 0xFF; + + ref.high = 0; + for (ir = MFT_REC_RESERVED; ir < MFT_REC_FREE; ir++) { + struct inode *i; + struct ntfs_inode *ni; + struct MFT_REC *mrec; + + ref.low = cpu_to_le32(ir); + ref.seq = cpu_to_le16(ir); + + i = ntfs_iget5(sb, &ref, NULL); + if (IS_ERR(i)) { +next: + ntfs_notice( + sb, + "Invalid reserved record %x", + ref.low); + continue; + } + if (is_bad_inode(i)) { + iput(i); + goto next; + } + + ni = ntfs_i(i); + + mrec = ni->mi.mrec; + + if (!is_rec_base(mrec)) + goto next; + + if (mrec->hard_links) + goto next; + + if (!ni_std(ni)) + goto next; + + if (ni_find_attr(ni, NULL, NULL, ATTR_NAME, + NULL, 0, NULL, NULL)) + goto next; + + __clear_bit(ir - MFT_REC_RESERVED, + &sbi->mft.reserved_bitmap); + } + } + + /* Scan 5 bits for zero. Bit 0 == MFT_REC_RESERVED */ + zbit = find_next_zero_bit(&sbi->mft.reserved_bitmap, + MFT_REC_FREE, MFT_REC_RESERVED); + if (zbit >= MFT_REC_FREE) { + sbi->mft.next_reserved = MFT_REC_FREE; + goto out; + } + + zlen = 1; + sbi->mft.next_reserved = zbit; + } else { +reserve_mft: + zlen = zbit == MFT_REC_FREE ? (MFT_REC_USER - MFT_REC_FREE) : 4; + if (zbit + zlen > wnd->nbits) + zlen = wnd->nbits - zbit; + + while (zlen > 1 && !wnd_is_free(wnd, zbit, zlen)) + zlen -= 1; + + /* [zbit, zbit + zlen) will be used for Mft itself */ + from = sbi->mft.used; + if (from < zbit) + from = zbit; + to = zbit + zlen; + if (from < to) { + ntfs_clear_mft_tail(sbi, from, to); + sbi->mft.used = to; + } + } + + if (mft) { + *rno = zbit; + zbit += 1; + zlen -= 1; + } + + wnd_zone_set(wnd, zbit, zlen); + +found: + if (!mft) { + /* The request to get record for general purpose */ + if (sbi->mft.next_free < MFT_REC_USER) + sbi->mft.next_free = MFT_REC_USER; + + for (;;) { + if (sbi->mft.next_free >= sbi->mft.bitmap.nbits) { + } else if (!wnd_find(wnd, 1, MFT_REC_USER, 0, &fr)) { + sbi->mft.next_free = sbi->mft.bitmap.nbits; + } else { + *rno = fr; + sbi->mft.next_free = *rno + 1; + break; + } + + err = ntfs_extend_mft(sbi); + if (err) + goto out; + } + } + + if (ni && !ni_add_subrecord(ni, *rno, mi)) { + err = -ENOMEM; + goto out; + } + + /* We have found a record that are not reserved for next MFT */ + if (*rno >= MFT_REC_FREE) + wnd_set_used(wnd, *rno, 1); + else if (*rno >= MFT_REC_RESERVED && sbi->mft.reserved_bitmap_inited) + __set_bit(*rno - MFT_REC_RESERVED, &sbi->mft.reserved_bitmap); + +out: + if (!mft) + up_write(&wnd->rw_lock); + + return err; +} + +/* + * ntfs_mark_rec_free + * + * marks record as free + */ +void ntfs_mark_rec_free(struct ntfs_sb_info *sbi, CLST rno) +{ + struct wnd_bitmap *wnd = &sbi->mft.bitmap; + + down_write_nested(&wnd->rw_lock, BITMAP_MUTEX_MFT); + if (rno >= wnd->nbits) + goto out; + + if (rno >= MFT_REC_FREE) { + if (!wnd_is_used(wnd, rno, 1)) + ntfs_set_state(sbi, NTFS_DIRTY_ERROR); + else + wnd_set_free(wnd, rno, 1); + } else if (rno >= MFT_REC_RESERVED && sbi->mft.reserved_bitmap_inited) { + __clear_bit(rno - MFT_REC_RESERVED, &sbi->mft.reserved_bitmap); + } + + if (rno < wnd_zone_bit(wnd)) + wnd_zone_set(wnd, rno, 1); + else if (rno < sbi->mft.next_free && rno >= MFT_REC_USER) + sbi->mft.next_free = rno; + +out: + up_write(&wnd->rw_lock); +} + +/* + * ntfs_clear_mft_tail + * + * formats empty records [from, to) + * sbi->mft.bitmap is locked for write + */ +int ntfs_clear_mft_tail(struct ntfs_sb_info *sbi, size_t from, size_t to) +{ + int err; + u32 rs; + u64 vbo; + struct runs_tree *run; + struct ntfs_inode *ni; + + if (from >= to) + return 0; + + rs = sbi->record_size; + ni = sbi->mft.ni; + run = &ni->file.run; + + down_read(&ni->file.run_lock); + vbo = (u64)from * rs; + for (; from < to; from++, vbo += rs) { + struct ntfs_buffers nb; + + err = ntfs_get_bh(sbi, run, vbo, rs, &nb); + if (err) + goto out; + + err = ntfs_write_bh(sbi, &sbi->new_rec->rhdr, &nb, 0); + nb_put(&nb); + if (err) + goto out; + } + +out: + sbi->mft.used = from; + up_read(&ni->file.run_lock); + return err; +} + +/* + * ntfs_refresh_zone + * + * refreshes Mft zone + * sbi->used.bitmap is locked for rw + * sbi->mft.bitmap is locked for write + * sbi->mft.ni->file.run_lock for write + */ +int ntfs_refresh_zone(struct ntfs_sb_info *sbi) +{ + CLST zone_limit, zone_max, lcn, vcn, len; + size_t lcn_s, zlen; + struct wnd_bitmap *wnd = &sbi->used.bitmap; + struct ntfs_inode *ni = sbi->mft.ni; + + /* Do not change anything unless we have non empty Mft zone */ + if (wnd_zone_len(wnd)) + return 0; + + /* + * Compute the mft zone at two steps + * It would be nice if we are able to allocate + * 1/8 of total clusters for MFT but not more then 512 MB + */ + zone_limit = (512 * 1024 * 1024) >> sbi->cluster_bits; + zone_max = wnd->nbits >> 3; + if (zone_max > zone_limit) + zone_max = zone_limit; + + vcn = bytes_to_cluster(sbi, + (u64)sbi->mft.bitmap.nbits << sbi->record_bits); + + if (!run_lookup_entry(&ni->file.run, vcn - 1, &lcn, &len, NULL)) + lcn = SPARSE_LCN; + + /* We should always find Last Lcn for MFT */ + if (lcn == SPARSE_LCN) + return -EINVAL; + + lcn_s = lcn + 1; + + /* Try to allocate clusters after last MFT run */ + zlen = wnd_find(wnd, zone_max, lcn_s, 0, &lcn_s); + if (!zlen) { + ntfs_notice(sbi->sb, "MftZone: unavailable"); + return 0; + } + + /* Truncate too large zone */ + wnd_zone_set(wnd, lcn_s, zlen); + + return 0; +} + +/* + * ntfs_update_mftmirr + * + * updates $MFTMirr data + */ +int ntfs_update_mftmirr(struct ntfs_sb_info *sbi, int wait) +{ + int err; + struct super_block *sb = sbi->sb; + u32 blocksize = sb->s_blocksize; + sector_t block1, block2; + u32 bytes; + + if (!(sbi->flags & NTFS_FLAGS_MFTMIRR)) + return 0; + + err = 0; + bytes = sbi->mft.recs_mirr << sbi->record_bits; + block1 = sbi->mft.lbo >> sb->s_blocksize_bits; + block2 = sbi->mft.lbo2 >> sb->s_blocksize_bits; + + for (; bytes >= blocksize; bytes -= blocksize) { + struct buffer_head *bh1, *bh2; + + bh1 = sb_bread(sb, block1++); + if (!bh1) { + err = -EIO; + goto out; + } + + bh2 = sb_getblk(sb, block2++); + if (!bh2) { + put_bh(bh1); + err = -EIO; + goto out; + } + + if (buffer_locked(bh2)) + __wait_on_buffer(bh2); + + lock_buffer(bh2); + memcpy(bh2->b_data, bh1->b_data, blocksize); + set_buffer_uptodate(bh2); + mark_buffer_dirty(bh2); + unlock_buffer(bh2); + + put_bh(bh1); + bh1 = NULL; + + if (wait) + err = sync_dirty_buffer(bh2); + + put_bh(bh2); + if (err) + goto out; + } + + sbi->flags &= ~NTFS_FLAGS_MFTMIRR; + +out: + return err; +} + +/* + * ntfs_set_state + * + * mount: ntfs_set_state(NTFS_DIRTY_DIRTY) + * umount: ntfs_set_state(NTFS_DIRTY_CLEAR) + * ntfs error: ntfs_set_state(NTFS_DIRTY_ERROR) + */ +int ntfs_set_state(struct ntfs_sb_info *sbi, enum NTFS_DIRTY_FLAGS dirty) +{ + int err; + struct ATTRIB *attr; + struct VOLUME_INFO *info; + struct mft_inode *mi; + struct ntfs_inode *ni; + + /* + * do not change state if fs was real_dirty + * do not change state if fs already dirty(clear) + * do not change any thing if mounted read only + */ + if (sbi->volume.real_dirty || sb_rdonly(sbi->sb)) + return 0; + + /* Check cached value */ + if ((dirty == NTFS_DIRTY_CLEAR ? 0 : VOLUME_FLAG_DIRTY) == + (sbi->volume.flags & VOLUME_FLAG_DIRTY)) + return 0; + + ni = sbi->volume.ni; + if (!ni) + return -EINVAL; + + mutex_lock_nested(&ni->ni_lock, NTFS_INODE_MUTEX_DIRTY); + + attr = ni_find_attr(ni, NULL, NULL, ATTR_VOL_INFO, NULL, 0, NULL, &mi); + if (!attr) { + err = -EINVAL; + goto out; + } + + info = resident_data_ex(attr, SIZEOF_ATTRIBUTE_VOLUME_INFO); + if (!info) { + err = -EINVAL; + goto out; + } + + switch (dirty) { + case NTFS_DIRTY_ERROR: + ntfs_notice(sbi->sb, "Mark volume as dirty due to NTFS errors"); + sbi->volume.real_dirty = true; + fallthrough; + case NTFS_DIRTY_DIRTY: + info->flags |= VOLUME_FLAG_DIRTY; + break; + case NTFS_DIRTY_CLEAR: + info->flags &= ~VOLUME_FLAG_DIRTY; + break; + } + /* cache current volume flags*/ + sbi->volume.flags = info->flags; + mi->dirty = true; + err = 0; + +out: + ni_unlock(ni); + if (err) + return err; + + mark_inode_dirty(&ni->vfs_inode); + /*verify(!ntfs_update_mftmirr()); */ + + /* + * if we used wait=1, sync_inode_metadata waits for the io for the + * inode to finish. It hangs when media is removed. + * So wait=0 is sent down to sync_inode_metadata + * and filemap_fdatawrite is used for the data blocks + */ + err = sync_inode_metadata(&ni->vfs_inode, 0); + if (!err) + err = filemap_fdatawrite(ni->vfs_inode.i_mapping); + + return err; +} + +/* + * security_hash + * + * calculates a hash of security descriptor + */ +static inline __le32 security_hash(const void *sd, size_t bytes) +{ + u32 hash = 0; + const __le32 *ptr = sd; + + bytes >>= 2; + while (bytes--) + hash = ((hash >> 0x1D) | (hash << 3)) + le32_to_cpu(*ptr++); + return cpu_to_le32(hash); +} + +int ntfs_sb_read(struct super_block *sb, u64 lbo, size_t bytes, void *buffer) +{ + struct block_device *bdev = sb->s_bdev; + u32 blocksize = sb->s_blocksize; + u64 block = lbo >> sb->s_blocksize_bits; + u32 off = lbo & (blocksize - 1); + u32 op = blocksize - off; + + for (; bytes; block += 1, off = 0, op = blocksize) { + struct buffer_head *bh = __bread(bdev, block, blocksize); + + if (!bh) + return -EIO; + + if (op > bytes) + op = bytes; + + memcpy(buffer, bh->b_data + off, op); + + put_bh(bh); + + bytes -= op; + buffer = Add2Ptr(buffer, op); + } + + return 0; +} + +int ntfs_sb_write(struct super_block *sb, u64 lbo, size_t bytes, + const void *buf, int wait) +{ + u32 blocksize = sb->s_blocksize; + struct block_device *bdev = sb->s_bdev; + sector_t block = lbo >> sb->s_blocksize_bits; + u32 off = lbo & (blocksize - 1); + u32 op = blocksize - off; + struct buffer_head *bh; + + if (!wait && (sb->s_flags & SB_SYNCHRONOUS)) + wait = 1; + + for (; bytes; block += 1, off = 0, op = blocksize) { + if (op > bytes) + op = bytes; + + if (op < blocksize) { + bh = __bread(bdev, block, blocksize); + if (!bh) { + ntfs_err(sb, "failed to read block %llx", + (u64)block); + return -EIO; + } + } else { + bh = __getblk(bdev, block, blocksize); + if (!bh) + return -ENOMEM; + } + + if (buffer_locked(bh)) + __wait_on_buffer(bh); + + lock_buffer(bh); + if (buf) { + memcpy(bh->b_data + off, buf, op); + buf = Add2Ptr(buf, op); + } else { + memset(bh->b_data + off, -1, op); + } + + set_buffer_uptodate(bh); + mark_buffer_dirty(bh); + unlock_buffer(bh); + + if (wait) { + int err = sync_dirty_buffer(bh); + + if (err) { + ntfs_err( + sb, + "failed to sync buffer at block %llx, error %d", + (u64)block, err); + put_bh(bh); + return err; + } + } + + put_bh(bh); + + bytes -= op; + } + return 0; +} + +int ntfs_sb_write_run(struct ntfs_sb_info *sbi, const struct runs_tree *run, + u64 vbo, const void *buf, size_t bytes) +{ + struct super_block *sb = sbi->sb; + u8 cluster_bits = sbi->cluster_bits; + u32 off = vbo & sbi->cluster_mask; + CLST lcn, clen, vcn = vbo >> cluster_bits, vcn_next; + u64 lbo, len; + size_t idx; + + if (!run_lookup_entry(run, vcn, &lcn, &clen, &idx)) + return -ENOENT; + + if (lcn == SPARSE_LCN) + return -EINVAL; + + lbo = ((u64)lcn << cluster_bits) + off; + len = ((u64)clen << cluster_bits) - off; + + for (;;) { + u32 op = len < bytes ? len : bytes; + int err = ntfs_sb_write(sb, lbo, op, buf, 0); + + if (err) + return err; + + bytes -= op; + if (!bytes) + break; + + vcn_next = vcn + clen; + if (!run_get_entry(run, ++idx, &vcn, &lcn, &clen) || + vcn != vcn_next) + return -ENOENT; + + if (lcn == SPARSE_LCN) + return -EINVAL; + + if (buf) + buf = Add2Ptr(buf, op); + + lbo = ((u64)lcn << cluster_bits); + len = ((u64)clen << cluster_bits); + } + + return 0; +} + +struct buffer_head *ntfs_bread_run(struct ntfs_sb_info *sbi, + const struct runs_tree *run, u64 vbo) +{ + struct super_block *sb = sbi->sb; + u8 cluster_bits = sbi->cluster_bits; + CLST lcn; + u64 lbo; + + if (!run_lookup_entry(run, vbo >> cluster_bits, &lcn, NULL, NULL)) + return ERR_PTR(-ENOENT); + + lbo = ((u64)lcn << cluster_bits) + (vbo & sbi->cluster_mask); + + return ntfs_bread(sb, lbo >> sb->s_blocksize_bits); +} + +int ntfs_read_run_nb(struct ntfs_sb_info *sbi, const struct runs_tree *run, + u64 vbo, void *buf, u32 bytes, struct ntfs_buffers *nb) +{ + int err; + struct super_block *sb = sbi->sb; + u32 blocksize = sb->s_blocksize; + u8 cluster_bits = sbi->cluster_bits; + u32 off = vbo & sbi->cluster_mask; + u32 nbh = 0; + CLST vcn_next, vcn = vbo >> cluster_bits; + CLST lcn, clen; + u64 lbo, len; + size_t idx; + struct buffer_head *bh; + + if (!run) { + /* first reading of $Volume + $MFTMirr + LogFile goes here*/ + if (vbo > MFT_REC_VOL * sbi->record_size) { + err = -ENOENT; + goto out; + } + + /* use absolute boot's 'MFTCluster' to read record */ + lbo = vbo + sbi->mft.lbo; + len = sbi->record_size; + } else if (!run_lookup_entry(run, vcn, &lcn, &clen, &idx)) { + err = -ENOENT; + goto out; + } else { + if (lcn == SPARSE_LCN) { + err = -EINVAL; + goto out; + } + + lbo = ((u64)lcn << cluster_bits) + off; + len = ((u64)clen << cluster_bits) - off; + } + + off = lbo & (blocksize - 1); + if (nb) { + nb->off = off; + nb->bytes = bytes; + } + + for (;;) { + u32 len32 = len >= bytes ? bytes : len; + sector_t block = lbo >> sb->s_blocksize_bits; + + do { + u32 op = blocksize - off; + + if (op > len32) + op = len32; + + bh = ntfs_bread(sb, block); + if (!bh) { + err = -EIO; + goto out; + } + + if (buf) { + memcpy(buf, bh->b_data + off, op); + buf = Add2Ptr(buf, op); + } + + if (!nb) { + put_bh(bh); + } else if (nbh >= ARRAY_SIZE(nb->bh)) { + err = -EINVAL; + goto out; + } else { + nb->bh[nbh++] = bh; + nb->nbufs = nbh; + } + + bytes -= op; + if (!bytes) + return 0; + len32 -= op; + block += 1; + off = 0; + + } while (len32); + + vcn_next = vcn + clen; + if (!run_get_entry(run, ++idx, &vcn, &lcn, &clen) || + vcn != vcn_next) { + err = -ENOENT; + goto out; + } + + if (lcn == SPARSE_LCN) { + err = -EINVAL; + goto out; + } + + lbo = ((u64)lcn << cluster_bits); + len = ((u64)clen << cluster_bits); + } + +out: + if (!nbh) + return err; + + while (nbh) { + put_bh(nb->bh[--nbh]); + nb->bh[nbh] = NULL; + } + + nb->nbufs = 0; + return err; +} + +/* Returns < 0 if error, 0 if ok, '-E_NTFS_FIXUP' if need to update fixups */ +int ntfs_read_bh(struct ntfs_sb_info *sbi, const struct runs_tree *run, u64 vbo, + struct NTFS_RECORD_HEADER *rhdr, u32 bytes, + struct ntfs_buffers *nb) +{ + int err = ntfs_read_run_nb(sbi, run, vbo, rhdr, bytes, nb); + + if (err) + return err; + return ntfs_fix_post_read(rhdr, nb->bytes, true); +} + +int ntfs_get_bh(struct ntfs_sb_info *sbi, const struct runs_tree *run, u64 vbo, + u32 bytes, struct ntfs_buffers *nb) +{ + int err = 0; + struct super_block *sb = sbi->sb; + u32 blocksize = sb->s_blocksize; + u8 cluster_bits = sbi->cluster_bits; + CLST vcn_next, vcn = vbo >> cluster_bits; + u32 off; + u32 nbh = 0; + CLST lcn, clen; + u64 lbo, len; + size_t idx; + + nb->bytes = bytes; + + if (!run_lookup_entry(run, vcn, &lcn, &clen, &idx)) { + err = -ENOENT; + goto out; + } + + off = vbo & sbi->cluster_mask; + lbo = ((u64)lcn << cluster_bits) + off; + len = ((u64)clen << cluster_bits) - off; + + nb->off = off = lbo & (blocksize - 1); + + for (;;) { + u32 len32 = len < bytes ? len : bytes; + sector_t block = lbo >> sb->s_blocksize_bits; + + do { + u32 op; + struct buffer_head *bh; + + if (nbh >= ARRAY_SIZE(nb->bh)) { + err = -EINVAL; + goto out; + } + + op = blocksize - off; + if (op > len32) + op = len32; + + if (op == blocksize) { + bh = sb_getblk(sb, block); + if (!bh) { + err = -ENOMEM; + goto out; + } + if (buffer_locked(bh)) + __wait_on_buffer(bh); + set_buffer_uptodate(bh); + } else { + bh = ntfs_bread(sb, block); + if (!bh) { + err = -EIO; + goto out; + } + } + + nb->bh[nbh++] = bh; + bytes -= op; + if (!bytes) { + nb->nbufs = nbh; + return 0; + } + + block += 1; + len32 -= op; + off = 0; + } while (len32); + + vcn_next = vcn + clen; + if (!run_get_entry(run, ++idx, &vcn, &lcn, &clen) || + vcn != vcn_next) { + err = -ENOENT; + goto out; + } + + lbo = ((u64)lcn << cluster_bits); + len = ((u64)clen << cluster_bits); + } + +out: + while (nbh) { + put_bh(nb->bh[--nbh]); + nb->bh[nbh] = NULL; + } + + nb->nbufs = 0; + + return err; +} + +int ntfs_write_bh(struct ntfs_sb_info *sbi, struct NTFS_RECORD_HEADER *rhdr, + struct ntfs_buffers *nb, int sync) +{ + int err = 0; + struct super_block *sb = sbi->sb; + u32 block_size = sb->s_blocksize; + u32 bytes = nb->bytes; + u32 off = nb->off; + u16 fo = le16_to_cpu(rhdr->fix_off); + u16 fn = le16_to_cpu(rhdr->fix_num); + u32 idx; + __le16 *fixup; + __le16 sample; + + if ((fo & 1) || fo + fn * sizeof(short) > SECTOR_SIZE || !fn-- || + fn * SECTOR_SIZE > bytes) { + return -EINVAL; + } + + for (idx = 0; bytes && idx < nb->nbufs; idx += 1, off = 0) { + u32 op = block_size - off; + char *bh_data; + struct buffer_head *bh = nb->bh[idx]; + __le16 *ptr, *end_data; + + if (op > bytes) + op = bytes; + + if (buffer_locked(bh)) + __wait_on_buffer(bh); + + lock_buffer(nb->bh[idx]); + + bh_data = bh->b_data + off; + end_data = Add2Ptr(bh_data, op); + memcpy(bh_data, rhdr, op); + + if (!idx) { + u16 t16; + + fixup = Add2Ptr(bh_data, fo); + sample = *fixup; + t16 = le16_to_cpu(sample); + if (t16 >= 0x7FFF) { + sample = *fixup = cpu_to_le16(1); + } else { + sample = cpu_to_le16(t16 + 1); + *fixup = sample; + } + + *(__le16 *)Add2Ptr(rhdr, fo) = sample; + } + + ptr = Add2Ptr(bh_data, SECTOR_SIZE - sizeof(short)); + + do { + *++fixup = *ptr; + *ptr = sample; + ptr += SECTOR_SIZE / sizeof(short); + } while (ptr < end_data); + + set_buffer_uptodate(bh); + mark_buffer_dirty(bh); + unlock_buffer(bh); + + if (sync) { + int err2 = sync_dirty_buffer(bh); + + if (!err && err2) + err = err2; + } + + bytes -= op; + rhdr = Add2Ptr(rhdr, op); + } + + return err; +} + +static inline struct bio *ntfs_alloc_bio(u32 nr_vecs) +{ + struct bio *bio = bio_alloc(GFP_NOFS | __GFP_HIGH, nr_vecs); + + if (!bio && (current->flags & PF_MEMALLOC)) { + while (!bio && (nr_vecs /= 2)) + bio = bio_alloc(GFP_NOFS | __GFP_HIGH, nr_vecs); + } + return bio; +} + +/* read/write pages from/to disk*/ +int ntfs_bio_pages(struct ntfs_sb_info *sbi, const struct runs_tree *run, + struct page **pages, u32 nr_pages, u64 vbo, u32 bytes, + u32 op) +{ + int err = 0; + struct bio *new, *bio = NULL; + struct super_block *sb = sbi->sb; + struct block_device *bdev = sb->s_bdev; + struct page *page; + u8 cluster_bits = sbi->cluster_bits; + CLST lcn, clen, vcn, vcn_next; + u32 add, off, page_idx; + u64 lbo, len; + size_t run_idx; + struct blk_plug plug; + + if (!bytes) + return 0; + + blk_start_plug(&plug); + + /* align vbo and bytes to be 512 bytes aligned */ + lbo = (vbo + bytes + 511) & ~511ull; + vbo = vbo & ~511ull; + bytes = lbo - vbo; + + vcn = vbo >> cluster_bits; + if (!run_lookup_entry(run, vcn, &lcn, &clen, &run_idx)) { + err = -ENOENT; + goto out; + } + off = vbo & sbi->cluster_mask; + page_idx = 0; + page = pages[0]; + + for (;;) { + lbo = ((u64)lcn << cluster_bits) + off; + len = ((u64)clen << cluster_bits) - off; +new_bio: + new = ntfs_alloc_bio(nr_pages - page_idx); + if (!new) { + err = -ENOMEM; + goto out; + } + if (bio) { + bio_chain(bio, new); + submit_bio(bio); + } + bio = new; + bio_set_dev(bio, bdev); + bio->bi_iter.bi_sector = lbo >> 9; + bio->bi_opf = op; + + while (len) { + off = vbo & (PAGE_SIZE - 1); + add = off + len > PAGE_SIZE ? (PAGE_SIZE - off) : len; + + if (bio_add_page(bio, page, add, off) < add) + goto new_bio; + + if (bytes <= add) + goto out; + bytes -= add; + vbo += add; + + if (add + off == PAGE_SIZE) { + page_idx += 1; + if (WARN_ON(page_idx >= nr_pages)) { + err = -EINVAL; + goto out; + } + page = pages[page_idx]; + } + + if (len <= add) + break; + len -= add; + lbo += add; + } + + vcn_next = vcn + clen; + if (!run_get_entry(run, ++run_idx, &vcn, &lcn, &clen) || + vcn != vcn_next) { + err = -ENOENT; + goto out; + } + off = 0; + } +out: + if (bio) { + if (!err) + err = submit_bio_wait(bio); + bio_put(bio); + } + blk_finish_plug(&plug); + + return err; +} + +/* + * Helper for ntfs_loadlog_and_replay + * fill on-disk logfile range by (-1) + * this means empty logfile + */ +int ntfs_bio_fill_1(struct ntfs_sb_info *sbi, const struct runs_tree *run) +{ + int err = 0; + struct super_block *sb = sbi->sb; + struct block_device *bdev = sb->s_bdev; + u8 cluster_bits = sbi->cluster_bits; + struct bio *new, *bio = NULL; + CLST lcn, clen; + u64 lbo, len; + size_t run_idx; + struct page *fill; + void *kaddr; + struct blk_plug plug; + + fill = alloc_page(GFP_KERNEL); + if (!fill) + return -ENOMEM; + + kaddr = kmap_atomic(fill); + memset(kaddr, -1, PAGE_SIZE); + kunmap_atomic(kaddr); + flush_dcache_page(fill); + lock_page(fill); + + if (!run_lookup_entry(run, 0, &lcn, &clen, &run_idx)) { + err = -ENOENT; + goto out; + } + + /* + * TODO: try blkdev_issue_write_same + */ + blk_start_plug(&plug); + do { + lbo = (u64)lcn << cluster_bits; + len = (u64)clen << cluster_bits; +new_bio: + new = ntfs_alloc_bio(BIO_MAX_VECS); + if (!new) { + err = -ENOMEM; + break; + } + if (bio) { + bio_chain(bio, new); + submit_bio(bio); + } + bio = new; + bio_set_dev(bio, bdev); + bio->bi_opf = REQ_OP_WRITE; + bio->bi_iter.bi_sector = lbo >> 9; + + for (;;) { + u32 add = len > PAGE_SIZE ? PAGE_SIZE : len; + + if (bio_add_page(bio, fill, add, 0) < add) + goto new_bio; + + lbo += add; + if (len <= add) + break; + len -= add; + } + } while (run_get_entry(run, ++run_idx, NULL, &lcn, &clen)); + + if (bio) { + if (!err) + err = submit_bio_wait(bio); + bio_put(bio); + } + blk_finish_plug(&plug); +out: + unlock_page(fill); + put_page(fill); + + return err; +} + +int ntfs_vbo_to_lbo(struct ntfs_sb_info *sbi, const struct runs_tree *run, + u64 vbo, u64 *lbo, u64 *bytes) +{ + u32 off; + CLST lcn, len; + u8 cluster_bits = sbi->cluster_bits; + + if (!run_lookup_entry(run, vbo >> cluster_bits, &lcn, &len, NULL)) + return -ENOENT; + + off = vbo & sbi->cluster_mask; + *lbo = lcn == SPARSE_LCN ? -1 : (((u64)lcn << cluster_bits) + off); + *bytes = ((u64)len << cluster_bits) - off; + + return 0; +} + +struct ntfs_inode *ntfs_new_inode(struct ntfs_sb_info *sbi, CLST rno, bool dir) +{ + int err = 0; + struct super_block *sb = sbi->sb; + struct inode *inode = new_inode(sb); + struct ntfs_inode *ni; + + if (!inode) + return ERR_PTR(-ENOMEM); + + ni = ntfs_i(inode); + + err = mi_format_new(&ni->mi, sbi, rno, dir ? RECORD_FLAG_DIR : 0, + false); + if (err) + goto out; + + inode->i_ino = rno; + if (insert_inode_locked(inode) < 0) { + err = -EIO; + goto out; + } + +out: + if (err) { + iput(inode); + ni = ERR_PTR(err); + } + return ni; +} + +/* + * O:BAG:BAD:(A;OICI;FA;;;WD) + * owner S-1-5-32-544 (Administrators) + * group S-1-5-32-544 (Administrators) + * ACE: allow S-1-1-0 (Everyone) with FILE_ALL_ACCESS + */ +const u8 s_default_security[] __aligned(8) = { + 0x01, 0x00, 0x04, 0x80, 0x30, 0x00, 0x00, 0x00, 0x40, 0x00, 0x00, 0x00, + 0x00, 0x00, 0x00, 0x00, 0x14, 0x00, 0x00, 0x00, 0x02, 0x00, 0x1C, 0x00, + 0x01, 0x00, 0x00, 0x00, 0x00, 0x03, 0x14, 0x00, 0xFF, 0x01, 0x1F, 0x00, + 0x01, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x00, + 0x01, 0x02, 0x00, 0x00, 0x00, 0x00, 0x00, 0x05, 0x20, 0x00, 0x00, 0x00, + 0x20, 0x02, 0x00, 0x00, 0x01, 0x02, 0x00, 0x00, 0x00, 0x00, 0x00, 0x05, + 0x20, 0x00, 0x00, 0x00, 0x20, 0x02, 0x00, 0x00, +}; + +static_assert(sizeof(s_default_security) == 0x50); + +static inline u32 sid_length(const struct SID *sid) +{ + return struct_size(sid, SubAuthority, sid->SubAuthorityCount); +} + +/* + * Thanks Mark Harmstone for idea + */ +static bool is_acl_valid(const struct ACL *acl, u32 len) +{ + const struct ACE_HEADER *ace; + u32 i; + u16 ace_count, ace_size; + + if (acl->AclRevision != ACL_REVISION && + acl->AclRevision != ACL_REVISION_DS) { + /* + * This value should be ACL_REVISION, unless the ACL contains an + * object-specific ACE, in which case this value must be ACL_REVISION_DS. + * All ACEs in an ACL must be at the same revision level. + */ + return false; + } + + if (acl->Sbz1) + return false; + + if (le16_to_cpu(acl->AclSize) > len) + return false; + + if (acl->Sbz2) + return false; + + len -= sizeof(struct ACL); + ace = (struct ACE_HEADER *)&acl[1]; + ace_count = le16_to_cpu(acl->AceCount); + + for (i = 0; i < ace_count; i++) { + if (len < sizeof(struct ACE_HEADER)) + return false; + + ace_size = le16_to_cpu(ace->AceSize); + if (len < ace_size) + return false; + + len -= ace_size; + ace = Add2Ptr(ace, ace_size); + } + + return true; +} + +bool is_sd_valid(const struct SECURITY_DESCRIPTOR_RELATIVE *sd, u32 len) +{ + u32 sd_owner, sd_group, sd_sacl, sd_dacl; + + if (len < sizeof(struct SECURITY_DESCRIPTOR_RELATIVE)) + return false; + + if (sd->Revision != 1) + return false; + + if (sd->Sbz1) + return false; + + if (!(sd->Control & SE_SELF_RELATIVE)) + return false; + + sd_owner = le32_to_cpu(sd->Owner); + if (sd_owner) { + const struct SID *owner = Add2Ptr(sd, sd_owner); + + if (sd_owner + offsetof(struct SID, SubAuthority) > len) + return false; + + if (owner->Revision != 1) + return false; + + if (sd_owner + sid_length(owner) > len) + return false; + } + + sd_group = le32_to_cpu(sd->Group); + if (sd_group) { + const struct SID *group = Add2Ptr(sd, sd_group); + + if (sd_group + offsetof(struct SID, SubAuthority) > len) + return false; + + if (group->Revision != 1) + return false; + + if (sd_group + sid_length(group) > len) + return false; + } + + sd_sacl = le32_to_cpu(sd->Sacl); + if (sd_sacl) { + const struct ACL *sacl = Add2Ptr(sd, sd_sacl); + + if (sd_sacl + sizeof(struct ACL) > len) + return false; + + if (!is_acl_valid(sacl, len - sd_sacl)) + return false; + } + + sd_dacl = le32_to_cpu(sd->Dacl); + if (sd_dacl) { + const struct ACL *dacl = Add2Ptr(sd, sd_dacl); + + if (sd_dacl + sizeof(struct ACL) > len) + return false; + + if (!is_acl_valid(dacl, len - sd_dacl)) + return false; + } + + return true; +} + +/* + * ntfs_security_init + * + * loads and parse $Secure + */ +int ntfs_security_init(struct ntfs_sb_info *sbi) +{ + int err; + struct super_block *sb = sbi->sb; + struct inode *inode; + struct ntfs_inode *ni; + struct MFT_REF ref; + struct ATTRIB *attr; + struct ATTR_LIST_ENTRY *le; + u64 sds_size; + size_t cnt, off; + struct NTFS_DE *ne; + struct NTFS_DE_SII *sii_e; + struct ntfs_fnd *fnd_sii = NULL; + const struct INDEX_ROOT *root_sii; + const struct INDEX_ROOT *root_sdh; + struct ntfs_index *indx_sdh = &sbi->security.index_sdh; + struct ntfs_index *indx_sii = &sbi->security.index_sii; + + ref.low = cpu_to_le32(MFT_REC_SECURE); + ref.high = 0; + ref.seq = cpu_to_le16(MFT_REC_SECURE); + + inode = ntfs_iget5(sb, &ref, &NAME_SECURE); + if (IS_ERR(inode)) { + err = PTR_ERR(inode); + ntfs_err(sb, "Failed to load $Secure."); + inode = NULL; + goto out; + } + + ni = ntfs_i(inode); + + le = NULL; + + attr = ni_find_attr(ni, NULL, &le, ATTR_ROOT, SDH_NAME, + ARRAY_SIZE(SDH_NAME), NULL, NULL); + if (!attr) { + err = -EINVAL; + goto out; + } + + root_sdh = resident_data(attr); + if (root_sdh->type != ATTR_ZERO || + root_sdh->rule != NTFS_COLLATION_TYPE_SECURITY_HASH) { + err = -EINVAL; + goto out; + } + + err = indx_init(indx_sdh, sbi, attr, INDEX_MUTEX_SDH); + if (err) + goto out; + + attr = ni_find_attr(ni, attr, &le, ATTR_ROOT, SII_NAME, + ARRAY_SIZE(SII_NAME), NULL, NULL); + if (!attr) { + err = -EINVAL; + goto out; + } + + root_sii = resident_data(attr); + if (root_sii->type != ATTR_ZERO || + root_sii->rule != NTFS_COLLATION_TYPE_UINT) { + err = -EINVAL; + goto out; + } + + err = indx_init(indx_sii, sbi, attr, INDEX_MUTEX_SII); + if (err) + goto out; + + fnd_sii = fnd_get(); + if (!fnd_sii) { + err = -ENOMEM; + goto out; + } + + sds_size = inode->i_size; + + /* Find the last valid Id */ + sbi->security.next_id = SECURITY_ID_FIRST; + /* Always write new security at the end of bucket */ + sbi->security.next_off = + Quad2Align(sds_size - SecurityDescriptorsBlockSize); + + cnt = 0; + off = 0; + ne = NULL; + + for (;;) { + u32 next_id; + + err = indx_find_raw(indx_sii, ni, root_sii, &ne, &off, fnd_sii); + if (err || !ne) + break; + + sii_e = (struct NTFS_DE_SII *)ne; + if (le16_to_cpu(ne->view.data_size) < SIZEOF_SECURITY_HDR) + continue; + + next_id = le32_to_cpu(sii_e->sec_id) + 1; + if (next_id >= sbi->security.next_id) + sbi->security.next_id = next_id; + + cnt += 1; + } + + sbi->security.ni = ni; + inode = NULL; +out: + iput(inode); + fnd_put(fnd_sii); + + return err; +} + +/* + * ntfs_get_security_by_id + * + * reads security descriptor by id + */ +int ntfs_get_security_by_id(struct ntfs_sb_info *sbi, __le32 security_id, + struct SECURITY_DESCRIPTOR_RELATIVE **sd, + size_t *size) +{ + int err; + int diff; + struct ntfs_inode *ni = sbi->security.ni; + struct ntfs_index *indx = &sbi->security.index_sii; + void *p = NULL; + struct NTFS_DE_SII *sii_e; + struct ntfs_fnd *fnd_sii; + struct SECURITY_HDR d_security; + const struct INDEX_ROOT *root_sii; + u32 t32; + + *sd = NULL; + + mutex_lock_nested(&ni->ni_lock, NTFS_INODE_MUTEX_SECURITY); + + fnd_sii = fnd_get(); + if (!fnd_sii) { + err = -ENOMEM; + goto out; + } + + root_sii = indx_get_root(indx, ni, NULL, NULL); + if (!root_sii) { + err = -EINVAL; + goto out; + } + + /* Try to find this SECURITY descriptor in SII indexes */ + err = indx_find(indx, ni, root_sii, &security_id, sizeof(security_id), + NULL, &diff, (struct NTFS_DE **)&sii_e, fnd_sii); + if (err) + goto out; + + if (diff) + goto out; + + t32 = le32_to_cpu(sii_e->sec_hdr.size); + if (t32 < SIZEOF_SECURITY_HDR) { + err = -EINVAL; + goto out; + } + + if (t32 > SIZEOF_SECURITY_HDR + 0x10000) { + /* + * looks like too big security. 0x10000 - is arbitrary big number + */ + err = -EFBIG; + goto out; + } + + *size = t32 - SIZEOF_SECURITY_HDR; + + p = ntfs_malloc(*size); + if (!p) { + err = -ENOMEM; + goto out; + } + + err = ntfs_read_run_nb(sbi, &ni->file.run, + le64_to_cpu(sii_e->sec_hdr.off), &d_security, + sizeof(d_security), NULL); + if (err) + goto out; + + if (memcmp(&d_security, &sii_e->sec_hdr, SIZEOF_SECURITY_HDR)) { + err = -EINVAL; + goto out; + } + + err = ntfs_read_run_nb(sbi, &ni->file.run, + le64_to_cpu(sii_e->sec_hdr.off) + + SIZEOF_SECURITY_HDR, + p, *size, NULL); + if (err) + goto out; + + *sd = p; + p = NULL; + +out: + ntfs_free(p); + fnd_put(fnd_sii); + ni_unlock(ni); + + return err; +} + +/* + * ntfs_insert_security + * + * inserts security descriptor into $Secure::SDS + * + * SECURITY Descriptor Stream data is organized into chunks of 256K bytes + * and it contains a mirror copy of each security descriptor. When writing + * to a security descriptor at location X, another copy will be written at + * location (X+256K). + * When writing a security descriptor that will cross the 256K boundary, + * the pointer will be advanced by 256K to skip + * over the mirror portion. + */ +int ntfs_insert_security(struct ntfs_sb_info *sbi, + const struct SECURITY_DESCRIPTOR_RELATIVE *sd, + u32 size_sd, __le32 *security_id, bool *inserted) +{ + int err, diff; + struct ntfs_inode *ni = sbi->security.ni; + struct ntfs_index *indx_sdh = &sbi->security.index_sdh; + struct ntfs_index *indx_sii = &sbi->security.index_sii; + struct NTFS_DE_SDH *e; + struct NTFS_DE_SDH sdh_e; + struct NTFS_DE_SII sii_e; + struct SECURITY_HDR *d_security; + u32 new_sec_size = size_sd + SIZEOF_SECURITY_HDR; + u32 aligned_sec_size = Quad2Align(new_sec_size); + struct SECURITY_KEY hash_key; + struct ntfs_fnd *fnd_sdh = NULL; + const struct INDEX_ROOT *root_sdh; + const struct INDEX_ROOT *root_sii; + u64 mirr_off, new_sds_size; + u32 next, left; + + static_assert((1 << Log2OfSecurityDescriptorsBlockSize) == + SecurityDescriptorsBlockSize); + + hash_key.hash = security_hash(sd, size_sd); + hash_key.sec_id = SECURITY_ID_INVALID; + + if (inserted) + *inserted = false; + *security_id = SECURITY_ID_INVALID; + + /* Allocate a temporal buffer*/ + d_security = ntfs_zalloc(aligned_sec_size); + if (!d_security) + return -ENOMEM; + + mutex_lock_nested(&ni->ni_lock, NTFS_INODE_MUTEX_SECURITY); + + fnd_sdh = fnd_get(); + if (!fnd_sdh) { + err = -ENOMEM; + goto out; + } + + root_sdh = indx_get_root(indx_sdh, ni, NULL, NULL); + if (!root_sdh) { + err = -EINVAL; + goto out; + } + + root_sii = indx_get_root(indx_sii, ni, NULL, NULL); + if (!root_sii) { + err = -EINVAL; + goto out; + } + + /* + * Check if such security already exists + * use "SDH" and hash -> to get the offset in "SDS" + */ + err = indx_find(indx_sdh, ni, root_sdh, &hash_key, sizeof(hash_key), + &d_security->key.sec_id, &diff, (struct NTFS_DE **)&e, + fnd_sdh); + if (err) + goto out; + + while (e) { + if (le32_to_cpu(e->sec_hdr.size) == new_sec_size) { + err = ntfs_read_run_nb(sbi, &ni->file.run, + le64_to_cpu(e->sec_hdr.off), + d_security, new_sec_size, NULL); + if (err) + goto out; + + if (le32_to_cpu(d_security->size) == new_sec_size && + d_security->key.hash == hash_key.hash && + !memcmp(d_security + 1, sd, size_sd)) { + *security_id = d_security->key.sec_id; + /*such security already exists*/ + err = 0; + goto out; + } + } + + err = indx_find_sort(indx_sdh, ni, root_sdh, + (struct NTFS_DE **)&e, fnd_sdh); + if (err) + goto out; + + if (!e || e->key.hash != hash_key.hash) + break; + } + + /* Zero unused space */ + next = sbi->security.next_off & (SecurityDescriptorsBlockSize - 1); + left = SecurityDescriptorsBlockSize - next; + + /* Zero gap until SecurityDescriptorsBlockSize */ + if (left < new_sec_size) { + /* zero "left" bytes from sbi->security.next_off */ + sbi->security.next_off += SecurityDescriptorsBlockSize + left; + } + + /* Zero tail of previous security */ + //used = ni->vfs_inode.i_size & (SecurityDescriptorsBlockSize - 1); + + /* + * Example: + * 0x40438 == ni->vfs_inode.i_size + * 0x00440 == sbi->security.next_off + * need to zero [0x438-0x440) + * if (next > used) { + * u32 tozero = next - used; + * zero "tozero" bytes from sbi->security.next_off - tozero + */ + + /* format new security descriptor */ + d_security->key.hash = hash_key.hash; + d_security->key.sec_id = cpu_to_le32(sbi->security.next_id); + d_security->off = cpu_to_le64(sbi->security.next_off); + d_security->size = cpu_to_le32(new_sec_size); + memcpy(d_security + 1, sd, size_sd); + + /* Write main SDS bucket */ + err = ntfs_sb_write_run(sbi, &ni->file.run, sbi->security.next_off, + d_security, aligned_sec_size); + + if (err) + goto out; + + mirr_off = sbi->security.next_off + SecurityDescriptorsBlockSize; + new_sds_size = mirr_off + aligned_sec_size; + + if (new_sds_size > ni->vfs_inode.i_size) { + err = attr_set_size(ni, ATTR_DATA, SDS_NAME, + ARRAY_SIZE(SDS_NAME), &ni->file.run, + new_sds_size, &new_sds_size, false, NULL); + if (err) + goto out; + } + + /* Write copy SDS bucket */ + err = ntfs_sb_write_run(sbi, &ni->file.run, mirr_off, d_security, + aligned_sec_size); + if (err) + goto out; + + /* Fill SII entry */ + sii_e.de.view.data_off = + cpu_to_le16(offsetof(struct NTFS_DE_SII, sec_hdr)); + sii_e.de.view.data_size = cpu_to_le16(SIZEOF_SECURITY_HDR); + sii_e.de.view.res = 0; + sii_e.de.size = cpu_to_le16(SIZEOF_SII_DIRENTRY); + sii_e.de.key_size = cpu_to_le16(sizeof(d_security->key.sec_id)); + sii_e.de.flags = 0; + sii_e.de.res = 0; + sii_e.sec_id = d_security->key.sec_id; + memcpy(&sii_e.sec_hdr, d_security, SIZEOF_SECURITY_HDR); + + err = indx_insert_entry(indx_sii, ni, &sii_e.de, NULL, NULL); + if (err) + goto out; + + /* Fill SDH entry */ + sdh_e.de.view.data_off = + cpu_to_le16(offsetof(struct NTFS_DE_SDH, sec_hdr)); + sdh_e.de.view.data_size = cpu_to_le16(SIZEOF_SECURITY_HDR); + sdh_e.de.view.res = 0; + sdh_e.de.size = cpu_to_le16(SIZEOF_SDH_DIRENTRY); + sdh_e.de.key_size = cpu_to_le16(sizeof(sdh_e.key)); + sdh_e.de.flags = 0; + sdh_e.de.res = 0; + sdh_e.key.hash = d_security->key.hash; + sdh_e.key.sec_id = d_security->key.sec_id; + memcpy(&sdh_e.sec_hdr, d_security, SIZEOF_SECURITY_HDR); + sdh_e.magic[0] = cpu_to_le16('I'); + sdh_e.magic[1] = cpu_to_le16('I'); + + fnd_clear(fnd_sdh); + err = indx_insert_entry(indx_sdh, ni, &sdh_e.de, (void *)(size_t)1, + fnd_sdh); + if (err) + goto out; + + *security_id = d_security->key.sec_id; + if (inserted) + *inserted = true; + + /* Update Id and offset for next descriptor */ + sbi->security.next_id += 1; + sbi->security.next_off += aligned_sec_size; + +out: + fnd_put(fnd_sdh); + mark_inode_dirty(&ni->vfs_inode); + ni_unlock(ni); + ntfs_free(d_security); + + return err; +} + +/* + * ntfs_reparse_init + * + * loads and parse $Extend/$Reparse + */ +int ntfs_reparse_init(struct ntfs_sb_info *sbi) +{ + int err; + struct ntfs_inode *ni = sbi->reparse.ni; + struct ntfs_index *indx = &sbi->reparse.index_r; + struct ATTRIB *attr; + struct ATTR_LIST_ENTRY *le; + const struct INDEX_ROOT *root_r; + + if (!ni) + return 0; + + le = NULL; + attr = ni_find_attr(ni, NULL, &le, ATTR_ROOT, SR_NAME, + ARRAY_SIZE(SR_NAME), NULL, NULL); + if (!attr) { + err = -EINVAL; + goto out; + } + + root_r = resident_data(attr); + if (root_r->type != ATTR_ZERO || + root_r->rule != NTFS_COLLATION_TYPE_UINTS) { + err = -EINVAL; + goto out; + } + + err = indx_init(indx, sbi, attr, INDEX_MUTEX_SR); + if (err) + goto out; + +out: + return err; +} + +/* + * ntfs_objid_init + * + * loads and parse $Extend/$ObjId + */ +int ntfs_objid_init(struct ntfs_sb_info *sbi) +{ + int err; + struct ntfs_inode *ni = sbi->objid.ni; + struct ntfs_index *indx = &sbi->objid.index_o; + struct ATTRIB *attr; + struct ATTR_LIST_ENTRY *le; + const struct INDEX_ROOT *root; + + if (!ni) + return 0; + + le = NULL; + attr = ni_find_attr(ni, NULL, &le, ATTR_ROOT, SO_NAME, + ARRAY_SIZE(SO_NAME), NULL, NULL); + if (!attr) { + err = -EINVAL; + goto out; + } + + root = resident_data(attr); + if (root->type != ATTR_ZERO || + root->rule != NTFS_COLLATION_TYPE_UINTS) { + err = -EINVAL; + goto out; + } + + err = indx_init(indx, sbi, attr, INDEX_MUTEX_SO); + if (err) + goto out; + +out: + return err; +} + +int ntfs_objid_remove(struct ntfs_sb_info *sbi, struct GUID *guid) +{ + int err; + struct ntfs_inode *ni = sbi->objid.ni; + struct ntfs_index *indx = &sbi->objid.index_o; + + if (!ni) + return -EINVAL; + + mutex_lock_nested(&ni->ni_lock, NTFS_INODE_MUTEX_OBJID); + + err = indx_delete_entry(indx, ni, guid, sizeof(*guid), NULL); + + mark_inode_dirty(&ni->vfs_inode); + ni_unlock(ni); + + return err; +} + +int ntfs_insert_reparse(struct ntfs_sb_info *sbi, __le32 rtag, + const struct MFT_REF *ref) +{ + int err; + struct ntfs_inode *ni = sbi->reparse.ni; + struct ntfs_index *indx = &sbi->reparse.index_r; + struct NTFS_DE_R re; + + if (!ni) + return -EINVAL; + + memset(&re, 0, sizeof(re)); + + re.de.view.data_off = cpu_to_le16(offsetof(struct NTFS_DE_R, zero)); + re.de.size = cpu_to_le16(sizeof(struct NTFS_DE_R)); + re.de.key_size = cpu_to_le16(sizeof(re.key)); + + re.key.ReparseTag = rtag; + memcpy(&re.key.ref, ref, sizeof(*ref)); + + mutex_lock_nested(&ni->ni_lock, NTFS_INODE_MUTEX_REPARSE); + + err = indx_insert_entry(indx, ni, &re.de, NULL, NULL); + + mark_inode_dirty(&ni->vfs_inode); + ni_unlock(ni); + + return err; +} + +int ntfs_remove_reparse(struct ntfs_sb_info *sbi, __le32 rtag, + const struct MFT_REF *ref) +{ + int err, diff; + struct ntfs_inode *ni = sbi->reparse.ni; + struct ntfs_index *indx = &sbi->reparse.index_r; + struct ntfs_fnd *fnd = NULL; + struct REPARSE_KEY rkey; + struct NTFS_DE_R *re; + struct INDEX_ROOT *root_r; + + if (!ni) + return -EINVAL; + + rkey.ReparseTag = rtag; + rkey.ref = *ref; + + mutex_lock_nested(&ni->ni_lock, NTFS_INODE_MUTEX_REPARSE); + + if (rtag) { + err = indx_delete_entry(indx, ni, &rkey, sizeof(rkey), NULL); + goto out1; + } + + fnd = fnd_get(); + if (!fnd) { + err = -ENOMEM; + goto out1; + } + + root_r = indx_get_root(indx, ni, NULL, NULL); + if (!root_r) { + err = -EINVAL; + goto out; + } + + /* 1 - forces to ignore rkey.ReparseTag when comparing keys */ + err = indx_find(indx, ni, root_r, &rkey, sizeof(rkey), (void *)1, &diff, + (struct NTFS_DE **)&re, fnd); + if (err) + goto out; + + if (memcmp(&re->key.ref, ref, sizeof(*ref))) { + /* Impossible. Looks like volume corrupt?*/ + goto out; + } + + memcpy(&rkey, &re->key, sizeof(rkey)); + + fnd_put(fnd); + fnd = NULL; + + err = indx_delete_entry(indx, ni, &rkey, sizeof(rkey), NULL); + if (err) + goto out; + +out: + fnd_put(fnd); + +out1: + mark_inode_dirty(&ni->vfs_inode); + ni_unlock(ni); + + return err; +} + +static inline void ntfs_unmap_and_discard(struct ntfs_sb_info *sbi, CLST lcn, + CLST len) +{ + ntfs_unmap_meta(sbi->sb, lcn, len); + ntfs_discard(sbi, lcn, len); +} + +void mark_as_free_ex(struct ntfs_sb_info *sbi, CLST lcn, CLST len, bool trim) +{ + CLST end, i; + struct wnd_bitmap *wnd = &sbi->used.bitmap; + + down_write_nested(&wnd->rw_lock, BITMAP_MUTEX_CLUSTERS); + if (!wnd_is_used(wnd, lcn, len)) { + ntfs_set_state(sbi, NTFS_DIRTY_ERROR); + + end = lcn + len; + len = 0; + for (i = lcn; i < end; i++) { + if (wnd_is_used(wnd, i, 1)) { + if (!len) + lcn = i; + len += 1; + continue; + } + + if (!len) + continue; + + if (trim) + ntfs_unmap_and_discard(sbi, lcn, len); + + wnd_set_free(wnd, lcn, len); + len = 0; + } + + if (!len) + goto out; + } + + if (trim) + ntfs_unmap_and_discard(sbi, lcn, len); + wnd_set_free(wnd, lcn, len); + +out: + up_write(&wnd->rw_lock); +} + +/* + * run_deallocate + * + * deallocate clusters + */ +int run_deallocate(struct ntfs_sb_info *sbi, struct runs_tree *run, bool trim) +{ + CLST lcn, len; + size_t idx = 0; + + while (run_get_entry(run, idx++, NULL, &lcn, &len)) { + if (lcn == SPARSE_LCN) + continue; + + mark_as_free_ex(sbi, lcn, len, trim); + } + + return 0; +} diff --git a/fs/ntfs3/index.c b/fs/ntfs3/index.c new file mode 100644 index 000000000000..6aa9540ece47 --- /dev/null +++ b/fs/ntfs3/index.c @@ -0,0 +1,2647 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * + * Copyright (C) 2019-2021 Paragon Software GmbH, All rights reserved. + * + */ + +#include <linux/blkdev.h> +#include <linux/buffer_head.h> +#include <linux/fs.h> +#include <linux/nls.h> + +#include "debug.h" +#include "ntfs.h" +#include "ntfs_fs.h" + +static const struct INDEX_NAMES { + const __le16 *name; + u8 name_len; +} s_index_names[INDEX_MUTEX_TOTAL] = { + { I30_NAME, ARRAY_SIZE(I30_NAME) }, { SII_NAME, ARRAY_SIZE(SII_NAME) }, + { SDH_NAME, ARRAY_SIZE(SDH_NAME) }, { SO_NAME, ARRAY_SIZE(SO_NAME) }, + { SQ_NAME, ARRAY_SIZE(SQ_NAME) }, { SR_NAME, ARRAY_SIZE(SR_NAME) }, +}; + +/* + * compare two names in index + * if l1 != 0 + * both names are little endian on-disk ATTR_FILE_NAME structs + * else + * key1 - cpu_str, key2 - ATTR_FILE_NAME + */ +static int cmp_fnames(const void *key1, size_t l1, const void *key2, size_t l2, + const void *data) +{ + const struct ATTR_FILE_NAME *f2 = key2; + const struct ntfs_sb_info *sbi = data; + const struct ATTR_FILE_NAME *f1; + u16 fsize2; + bool both_case; + + if (l2 <= offsetof(struct ATTR_FILE_NAME, name)) + return -1; + + fsize2 = fname_full_size(f2); + if (l2 < fsize2) + return -1; + + both_case = f2->type != FILE_NAME_DOS /*&& !sbi->options.nocase*/; + if (!l1) { + const struct le_str *s2 = (struct le_str *)&f2->name_len; + + /* + * If names are equal (case insensitive) + * try to compare it case sensitive + */ + return ntfs_cmp_names_cpu(key1, s2, sbi->upcase, both_case); + } + + f1 = key1; + return ntfs_cmp_names(f1->name, f1->name_len, f2->name, f2->name_len, + sbi->upcase, both_case); +} + +/* $SII of $Secure and $Q of Quota */ +static int cmp_uint(const void *key1, size_t l1, const void *key2, size_t l2, + const void *data) +{ + const u32 *k1 = key1; + const u32 *k2 = key2; + + if (l2 < sizeof(u32)) + return -1; + + if (*k1 < *k2) + return -1; + if (*k1 > *k2) + return 1; + return 0; +} + +/* $SDH of $Secure */ +static int cmp_sdh(const void *key1, size_t l1, const void *key2, size_t l2, + const void *data) +{ + const struct SECURITY_KEY *k1 = key1; + const struct SECURITY_KEY *k2 = key2; + u32 t1, t2; + + if (l2 < sizeof(struct SECURITY_KEY)) + return -1; + + t1 = le32_to_cpu(k1->hash); + t2 = le32_to_cpu(k2->hash); + + /* First value is a hash value itself */ + if (t1 < t2) + return -1; + if (t1 > t2) + return 1; + + /* Second value is security Id */ + if (data) { + t1 = le32_to_cpu(k1->sec_id); + t2 = le32_to_cpu(k2->sec_id); + if (t1 < t2) + return -1; + if (t1 > t2) + return 1; + } + + return 0; +} + +/* $O of ObjId and "$R" for Reparse */ +static int cmp_uints(const void *key1, size_t l1, const void *key2, size_t l2, + const void *data) +{ + const __le32 *k1 = key1; + const __le32 *k2 = key2; + size_t count; + + if ((size_t)data == 1) { + /* + * ni_delete_all -> ntfs_remove_reparse -> delete all with this reference + * k1, k2 - pointers to REPARSE_KEY + */ + + k1 += 1; // skip REPARSE_KEY.ReparseTag + k2 += 1; // skip REPARSE_KEY.ReparseTag + if (l2 <= sizeof(int)) + return -1; + l2 -= sizeof(int); + if (l1 <= sizeof(int)) + return 1; + l1 -= sizeof(int); + } + + if (l2 < sizeof(int)) + return -1; + + for (count = min(l1, l2) >> 2; count > 0; --count, ++k1, ++k2) { + u32 t1 = le32_to_cpu(*k1); + u32 t2 = le32_to_cpu(*k2); + + if (t1 > t2) + return 1; + if (t1 < t2) + return -1; + } + + if (l1 > l2) + return 1; + if (l1 < l2) + return -1; + + return 0; +} + +static inline NTFS_CMP_FUNC get_cmp_func(const struct INDEX_ROOT *root) +{ + switch (root->type) { + case ATTR_NAME: + if (root->rule == NTFS_COLLATION_TYPE_FILENAME) + return &cmp_fnames; + break; + case ATTR_ZERO: + switch (root->rule) { + case NTFS_COLLATION_TYPE_UINT: + return &cmp_uint; + case NTFS_COLLATION_TYPE_SECURITY_HASH: + return &cmp_sdh; + case NTFS_COLLATION_TYPE_UINTS: + return &cmp_uints; + default: + break; + } + default: + break; + } + + return NULL; +} + +struct bmp_buf { + struct ATTRIB *b; + struct mft_inode *mi; + struct buffer_head *bh; + ulong *buf; + size_t bit; + u32 nbits; + u64 new_valid; +}; + +static int bmp_buf_get(struct ntfs_index *indx, struct ntfs_inode *ni, + size_t bit, struct bmp_buf *bbuf) +{ + struct ATTRIB *b; + size_t data_size, valid_size, vbo, off = bit >> 3; + struct ntfs_sb_info *sbi = ni->mi.sbi; + CLST vcn = off >> sbi->cluster_bits; + struct ATTR_LIST_ENTRY *le = NULL; + struct buffer_head *bh; + struct super_block *sb; + u32 blocksize; + const struct INDEX_NAMES *in = &s_index_names[indx->type]; + + bbuf->bh = NULL; + + b = ni_find_attr(ni, NULL, &le, ATTR_BITMAP, in->name, in->name_len, + &vcn, &bbuf->mi); + bbuf->b = b; + if (!b) + return -EINVAL; + + if (!b->non_res) { + data_size = le32_to_cpu(b->res.data_size); + + if (off >= data_size) + return -EINVAL; + + bbuf->buf = (ulong *)resident_data(b); + bbuf->bit = 0; + bbuf->nbits = data_size * 8; + + return 0; + } + + data_size = le64_to_cpu(b->nres.data_size); + if (WARN_ON(off >= data_size)) { + /* looks like filesystem error */ + return -EINVAL; + } + + valid_size = le64_to_cpu(b->nres.valid_size); + + bh = ntfs_bread_run(sbi, &indx->bitmap_run, off); + if (!bh) + return -EIO; + + if (IS_ERR(bh)) + return PTR_ERR(bh); + + bbuf->bh = bh; + + if (buffer_locked(bh)) + __wait_on_buffer(bh); + + lock_buffer(bh); + + sb = sbi->sb; + blocksize = sb->s_blocksize; + + vbo = off & ~(size_t)sbi->block_mask; + + bbuf->new_valid = vbo + blocksize; + if (bbuf->new_valid <= valid_size) + bbuf->new_valid = 0; + else if (bbuf->new_valid > data_size) + bbuf->new_valid = data_size; + + if (vbo >= valid_size) { + memset(bh->b_data, 0, blocksize); + } else if (vbo + blocksize > valid_size) { + u32 voff = valid_size & sbi->block_mask; + + memset(bh->b_data + voff, 0, blocksize - voff); + } + + bbuf->buf = (ulong *)bh->b_data; + bbuf->bit = 8 * (off & ~(size_t)sbi->block_mask); + bbuf->nbits = 8 * blocksize; + + return 0; +} + +static void bmp_buf_put(struct bmp_buf *bbuf, bool dirty) +{ + struct buffer_head *bh = bbuf->bh; + struct ATTRIB *b = bbuf->b; + + if (!bh) { + if (b && !b->non_res && dirty) + bbuf->mi->dirty = true; + return; + } + + if (!dirty) + goto out; + + if (bbuf->new_valid) { + b->nres.valid_size = cpu_to_le64(bbuf->new_valid); + bbuf->mi->dirty = true; + } + + set_buffer_uptodate(bh); + mark_buffer_dirty(bh); + +out: + unlock_buffer(bh); + put_bh(bh); +} + +/* + * indx_mark_used + * + * marks the bit 'bit' as used + */ +static int indx_mark_used(struct ntfs_index *indx, struct ntfs_inode *ni, + size_t bit) +{ + int err; + struct bmp_buf bbuf; + + err = bmp_buf_get(indx, ni, bit, &bbuf); + if (err) + return err; + + __set_bit(bit - bbuf.bit, bbuf.buf); + + bmp_buf_put(&bbuf, true); + + return 0; +} + +/* + * indx_mark_free + * + * the bit 'bit' as free + */ +static int indx_mark_free(struct ntfs_index *indx, struct ntfs_inode *ni, + size_t bit) +{ + int err; + struct bmp_buf bbuf; + + err = bmp_buf_get(indx, ni, bit, &bbuf); + if (err) + return err; + + __clear_bit(bit - bbuf.bit, bbuf.buf); + + bmp_buf_put(&bbuf, true); + + return 0; +} + +/* + * if ntfs_readdir calls this function (indx_used_bit -> scan_nres_bitmap), + * inode is shared locked and no ni_lock + * use rw_semaphore for read/write access to bitmap_run + */ +static int scan_nres_bitmap(struct ntfs_inode *ni, struct ATTRIB *bitmap, + struct ntfs_index *indx, size_t from, + bool (*fn)(const ulong *buf, u32 bit, u32 bits, + size_t *ret), + size_t *ret) +{ + struct ntfs_sb_info *sbi = ni->mi.sbi; + struct super_block *sb = sbi->sb; + struct runs_tree *run = &indx->bitmap_run; + struct rw_semaphore *lock = &indx->run_lock; + u32 nbits = sb->s_blocksize * 8; + u32 blocksize = sb->s_blocksize; + u64 valid_size = le64_to_cpu(bitmap->nres.valid_size); + u64 data_size = le64_to_cpu(bitmap->nres.data_size); + sector_t eblock = bytes_to_block(sb, data_size); + size_t vbo = from >> 3; + sector_t blk = (vbo & sbi->cluster_mask) >> sb->s_blocksize_bits; + sector_t vblock = vbo >> sb->s_blocksize_bits; + sector_t blen, block; + CLST lcn, clen, vcn, vcn_next; + size_t idx; + struct buffer_head *bh; + bool ok; + + *ret = MINUS_ONE_T; + + if (vblock >= eblock) + return 0; + + from &= nbits - 1; + vcn = vbo >> sbi->cluster_bits; + + down_read(lock); + ok = run_lookup_entry(run, vcn, &lcn, &clen, &idx); + up_read(lock); + +next_run: + if (!ok) { + int err; + const struct INDEX_NAMES *name = &s_index_names[indx->type]; + + down_write(lock); + err = attr_load_runs_vcn(ni, ATTR_BITMAP, name->name, + name->name_len, run, vcn); + up_write(lock); + if (err) + return err; + down_read(lock); + ok = run_lookup_entry(run, vcn, &lcn, &clen, &idx); + up_read(lock); + if (!ok) + return -EINVAL; + } + + blen = (sector_t)clen * sbi->blocks_per_cluster; + block = (sector_t)lcn * sbi->blocks_per_cluster; + + for (; blk < blen; blk++, from = 0) { + bh = ntfs_bread(sb, block + blk); + if (!bh) + return -EIO; + + vbo = (u64)vblock << sb->s_blocksize_bits; + if (vbo >= valid_size) { + memset(bh->b_data, 0, blocksize); + } else if (vbo + blocksize > valid_size) { + u32 voff = valid_size & sbi->block_mask; + + memset(bh->b_data + voff, 0, blocksize - voff); + } + + if (vbo + blocksize > data_size) + nbits = 8 * (data_size - vbo); + + ok = nbits > from ? (*fn)((ulong *)bh->b_data, from, nbits, ret) + : false; + put_bh(bh); + + if (ok) { + *ret += 8 * vbo; + return 0; + } + + if (++vblock >= eblock) { + *ret = MINUS_ONE_T; + return 0; + } + } + blk = 0; + vcn_next = vcn + clen; + down_read(lock); + ok = run_get_entry(run, ++idx, &vcn, &lcn, &clen) && vcn == vcn_next; + if (!ok) + vcn = vcn_next; + up_read(lock); + goto next_run; +} + +static bool scan_for_free(const ulong *buf, u32 bit, u32 bits, size_t *ret) +{ + size_t pos = find_next_zero_bit(buf, bits, bit); + + if (pos >= bits) + return false; + *ret = pos; + return true; +} + +/* + * indx_find_free + * + * looks for free bit + * returns -1 if no free bits + */ +static int indx_find_free(struct ntfs_index *indx, struct ntfs_inode *ni, + size_t *bit, struct ATTRIB **bitmap) +{ + struct ATTRIB *b; + struct ATTR_LIST_ENTRY *le = NULL; + const struct INDEX_NAMES *in = &s_index_names[indx->type]; + int err; + + b = ni_find_attr(ni, NULL, &le, ATTR_BITMAP, in->name, in->name_len, + NULL, NULL); + + if (!b) + return -ENOENT; + + *bitmap = b; + *bit = MINUS_ONE_T; + + if (!b->non_res) { + u32 nbits = 8 * le32_to_cpu(b->res.data_size); + size_t pos = find_next_zero_bit(resident_data(b), nbits, 0); + + if (pos < nbits) + *bit = pos; + } else { + err = scan_nres_bitmap(ni, b, indx, 0, &scan_for_free, bit); + + if (err) + return err; + } + + return 0; +} + +static bool scan_for_used(const ulong *buf, u32 bit, u32 bits, size_t *ret) +{ + size_t pos = find_next_bit(buf, bits, bit); + + if (pos >= bits) + return false; + *ret = pos; + return true; +} + +/* + * indx_used_bit + * + * looks for used bit + * returns MINUS_ONE_T if no used bits + */ +int indx_used_bit(struct ntfs_index *indx, struct ntfs_inode *ni, size_t *bit) +{ + struct ATTRIB *b; + struct ATTR_LIST_ENTRY *le = NULL; + size_t from = *bit; + const struct INDEX_NAMES *in = &s_index_names[indx->type]; + int err; + + b = ni_find_attr(ni, NULL, &le, ATTR_BITMAP, in->name, in->name_len, + NULL, NULL); + + if (!b) + return -ENOENT; + + *bit = MINUS_ONE_T; + + if (!b->non_res) { + u32 nbits = le32_to_cpu(b->res.data_size) * 8; + size_t pos = find_next_bit(resident_data(b), nbits, from); + + if (pos < nbits) + *bit = pos; + } else { + err = scan_nres_bitmap(ni, b, indx, from, &scan_for_used, bit); + if (err) + return err; + } + + return 0; +} + +/* + * hdr_find_split + * + * finds a point at which the index allocation buffer would like to + * be split. + * NOTE: This function should never return 'END' entry NULL returns on error + */ +static const struct NTFS_DE *hdr_find_split(const struct INDEX_HDR *hdr) +{ + size_t o; + const struct NTFS_DE *e = hdr_first_de(hdr); + u32 used_2 = le32_to_cpu(hdr->used) >> 1; + u16 esize = le16_to_cpu(e->size); + + if (!e || de_is_last(e)) + return NULL; + + for (o = le32_to_cpu(hdr->de_off) + esize; o < used_2; o += esize) { + const struct NTFS_DE *p = e; + + e = Add2Ptr(hdr, o); + + /* We must not return END entry */ + if (de_is_last(e)) + return p; + + esize = le16_to_cpu(e->size); + } + + return e; +} + +/* + * hdr_insert_head + * + * inserts some entries at the beginning of the buffer. + * It is used to insert entries into a newly-created buffer. + */ +static const struct NTFS_DE *hdr_insert_head(struct INDEX_HDR *hdr, + const void *ins, u32 ins_bytes) +{ + u32 to_move; + struct NTFS_DE *e = hdr_first_de(hdr); + u32 used = le32_to_cpu(hdr->used); + + if (!e) + return NULL; + + /* Now we just make room for the inserted entries and jam it in. */ + to_move = used - le32_to_cpu(hdr->de_off); + memmove(Add2Ptr(e, ins_bytes), e, to_move); + memcpy(e, ins, ins_bytes); + hdr->used = cpu_to_le32(used + ins_bytes); + + return e; +} + +void fnd_clear(struct ntfs_fnd *fnd) +{ + int i; + + for (i = 0; i < fnd->level; i++) { + struct indx_node *n = fnd->nodes[i]; + + if (!n) + continue; + + put_indx_node(n); + fnd->nodes[i] = NULL; + } + fnd->level = 0; + fnd->root_de = NULL; +} + +static int fnd_push(struct ntfs_fnd *fnd, struct indx_node *n, + struct NTFS_DE *e) +{ + int i; + + i = fnd->level; + if (i < 0 || i >= ARRAY_SIZE(fnd->nodes)) + return -EINVAL; + fnd->nodes[i] = n; + fnd->de[i] = e; + fnd->level += 1; + return 0; +} + +static struct indx_node *fnd_pop(struct ntfs_fnd *fnd) +{ + struct indx_node *n; + int i = fnd->level; + + i -= 1; + n = fnd->nodes[i]; + fnd->nodes[i] = NULL; + fnd->level = i; + + return n; +} + +static bool fnd_is_empty(struct ntfs_fnd *fnd) +{ + if (!fnd->level) + return !fnd->root_de; + + return !fnd->de[fnd->level - 1]; +} + +/* + * hdr_find_e + * + * locates an entry the index buffer. + * If no matching entry is found, it returns the first entry which is greater + * than the desired entry If the search key is greater than all the entries the + * buffer, it returns the 'end' entry. This function does a binary search of the + * current index buffer, for the first entry that is <= to the search value + * Returns NULL if error + */ +static struct NTFS_DE *hdr_find_e(const struct ntfs_index *indx, + const struct INDEX_HDR *hdr, const void *key, + size_t key_len, const void *ctx, int *diff) +{ + struct NTFS_DE *e; + NTFS_CMP_FUNC cmp = indx->cmp; + u32 e_size, e_key_len; + u32 end = le32_to_cpu(hdr->used); + u32 off = le32_to_cpu(hdr->de_off); + +#ifdef NTFS3_INDEX_BINARY_SEARCH + int max_idx = 0, fnd, min_idx; + int nslots = 64; + u16 *offs; + + if (end > 0x10000) + goto next; + + offs = ntfs_malloc(sizeof(u16) * nslots); + if (!offs) + goto next; + + /* use binary search algorithm */ +next1: + if (off + sizeof(struct NTFS_DE) > end) { + e = NULL; + goto out1; + } + e = Add2Ptr(hdr, off); + e_size = le16_to_cpu(e->size); + + if (e_size < sizeof(struct NTFS_DE) || off + e_size > end) { + e = NULL; + goto out1; + } + + if (max_idx >= nslots) { + u16 *ptr; + int new_slots = QuadAlign(2 * nslots); + + ptr = ntfs_malloc(sizeof(u16) * new_slots); + if (ptr) + memcpy(ptr, offs, sizeof(u16) * max_idx); + ntfs_free(offs); + offs = ptr; + nslots = new_slots; + if (!ptr) + goto next; + } + + /* Store entry table */ + offs[max_idx] = off; + + if (!de_is_last(e)) { + off += e_size; + max_idx += 1; + goto next1; + } + + /* + * Table of pointers is created + * Use binary search to find entry that is <= to the search value + */ + fnd = -1; + min_idx = 0; + + while (min_idx <= max_idx) { + int mid_idx = min_idx + ((max_idx - min_idx) >> 1); + int diff2; + + e = Add2Ptr(hdr, offs[mid_idx]); + + e_key_len = le16_to_cpu(e->key_size); + + diff2 = (*cmp)(key, key_len, e + 1, e_key_len, ctx); + + if (!diff2) { + *diff = 0; + goto out1; + } + + if (diff2 < 0) { + max_idx = mid_idx - 1; + fnd = mid_idx; + if (!fnd) + break; + } else { + min_idx = mid_idx + 1; + } + } + + if (fnd == -1) { + e = NULL; + goto out1; + } + + *diff = -1; + e = Add2Ptr(hdr, offs[fnd]); + +out1: + ntfs_free(offs); + + return e; +#endif + +next: + /* + * Entries index are sorted + * Enumerate all entries until we find entry that is <= to the search value + */ + if (off + sizeof(struct NTFS_DE) > end) + return NULL; + + e = Add2Ptr(hdr, off); + e_size = le16_to_cpu(e->size); + + if (e_size < sizeof(struct NTFS_DE) || off + e_size > end) + return NULL; + + off += e_size; + + e_key_len = le16_to_cpu(e->key_size); + + *diff = (*cmp)(key, key_len, e + 1, e_key_len, ctx); + if (!*diff) + return e; + + if (*diff <= 0) + return e; + + if (de_is_last(e)) { + *diff = 1; + return e; + } + goto next; +} + +/* + * hdr_insert_de + * + * inserts an index entry into the buffer. + * 'before' should be a pointer previously returned from hdr_find_e + */ +static struct NTFS_DE *hdr_insert_de(const struct ntfs_index *indx, + struct INDEX_HDR *hdr, + const struct NTFS_DE *de, + struct NTFS_DE *before, const void *ctx) +{ + int diff; + size_t off = PtrOffset(hdr, before); + u32 used = le32_to_cpu(hdr->used); + u32 total = le32_to_cpu(hdr->total); + u16 de_size = le16_to_cpu(de->size); + + /* First, check to see if there's enough room */ + if (used + de_size > total) + return NULL; + + /* We know there's enough space, so we know we'll succeed. */ + if (before) { + /* Check that before is inside Index */ + if (off >= used || off < le32_to_cpu(hdr->de_off) || + off + le16_to_cpu(before->size) > total) { + return NULL; + } + goto ok; + } + /* No insert point is applied. Get it manually */ + before = hdr_find_e(indx, hdr, de + 1, le16_to_cpu(de->key_size), ctx, + &diff); + if (!before) + return NULL; + off = PtrOffset(hdr, before); + +ok: + /* Now we just make room for the entry and jam it in. */ + memmove(Add2Ptr(before, de_size), before, used - off); + + hdr->used = cpu_to_le32(used + de_size); + memcpy(before, de, de_size); + + return before; +} + +/* + * hdr_delete_de + * + * removes an entry from the index buffer + */ +static inline struct NTFS_DE *hdr_delete_de(struct INDEX_HDR *hdr, + struct NTFS_DE *re) +{ + u32 used = le32_to_cpu(hdr->used); + u16 esize = le16_to_cpu(re->size); + u32 off = PtrOffset(hdr, re); + int bytes = used - (off + esize); + + if (off >= used || esize < sizeof(struct NTFS_DE) || + bytes < sizeof(struct NTFS_DE)) + return NULL; + + hdr->used = cpu_to_le32(used - esize); + memmove(re, Add2Ptr(re, esize), bytes); + + return re; +} + +void indx_clear(struct ntfs_index *indx) +{ + run_close(&indx->alloc_run); + run_close(&indx->bitmap_run); +} + +int indx_init(struct ntfs_index *indx, struct ntfs_sb_info *sbi, + const struct ATTRIB *attr, enum index_mutex_classed type) +{ + u32 t32; + const struct INDEX_ROOT *root = resident_data(attr); + + /* Check root fields */ + if (!root->index_block_clst) + return -EINVAL; + + indx->type = type; + indx->idx2vbn_bits = __ffs(root->index_block_clst); + + t32 = le32_to_cpu(root->index_block_size); + indx->index_bits = blksize_bits(t32); + + /* Check index record size */ + if (t32 < sbi->cluster_size) { + /* index record is smaller than a cluster, use 512 blocks */ + if (t32 != root->index_block_clst * SECTOR_SIZE) + return -EINVAL; + + /* Check alignment to a cluster */ + if ((sbi->cluster_size >> SECTOR_SHIFT) & + (root->index_block_clst - 1)) { + return -EINVAL; + } + + indx->vbn2vbo_bits = SECTOR_SHIFT; + } else { + /* index record must be a multiple of cluster size */ + if (t32 != root->index_block_clst << sbi->cluster_bits) + return -EINVAL; + + indx->vbn2vbo_bits = sbi->cluster_bits; + } + + init_rwsem(&indx->run_lock); + + indx->cmp = get_cmp_func(root); + return indx->cmp ? 0 : -EINVAL; +} + +static struct indx_node *indx_new(struct ntfs_index *indx, + struct ntfs_inode *ni, CLST vbn, + const __le64 *sub_vbn) +{ + int err; + struct NTFS_DE *e; + struct indx_node *r; + struct INDEX_HDR *hdr; + struct INDEX_BUFFER *index; + u64 vbo = (u64)vbn << indx->vbn2vbo_bits; + u32 bytes = 1u << indx->index_bits; + u16 fn; + u32 eo; + + r = ntfs_zalloc(sizeof(struct indx_node)); + if (!r) + return ERR_PTR(-ENOMEM); + + index = ntfs_zalloc(bytes); + if (!index) { + ntfs_free(r); + return ERR_PTR(-ENOMEM); + } + + err = ntfs_get_bh(ni->mi.sbi, &indx->alloc_run, vbo, bytes, &r->nb); + + if (err) { + ntfs_free(index); + ntfs_free(r); + return ERR_PTR(err); + } + + /* Create header */ + index->rhdr.sign = NTFS_INDX_SIGNATURE; + index->rhdr.fix_off = cpu_to_le16(sizeof(struct INDEX_BUFFER)); // 0x28 + fn = (bytes >> SECTOR_SHIFT) + 1; // 9 + index->rhdr.fix_num = cpu_to_le16(fn); + index->vbn = cpu_to_le64(vbn); + hdr = &index->ihdr; + eo = QuadAlign(sizeof(struct INDEX_BUFFER) + fn * sizeof(short)); + hdr->de_off = cpu_to_le32(eo); + + e = Add2Ptr(hdr, eo); + + if (sub_vbn) { + e->flags = NTFS_IE_LAST | NTFS_IE_HAS_SUBNODES; + e->size = cpu_to_le16(sizeof(struct NTFS_DE) + sizeof(u64)); + hdr->used = + cpu_to_le32(eo + sizeof(struct NTFS_DE) + sizeof(u64)); + de_set_vbn_le(e, *sub_vbn); + hdr->flags = 1; + } else { + e->size = cpu_to_le16(sizeof(struct NTFS_DE)); + hdr->used = cpu_to_le32(eo + sizeof(struct NTFS_DE)); + e->flags = NTFS_IE_LAST; + } + + hdr->total = cpu_to_le32(bytes - offsetof(struct INDEX_BUFFER, ihdr)); + + r->index = index; + return r; +} + +struct INDEX_ROOT *indx_get_root(struct ntfs_index *indx, struct ntfs_inode *ni, + struct ATTRIB **attr, struct mft_inode **mi) +{ + struct ATTR_LIST_ENTRY *le = NULL; + struct ATTRIB *a; + const struct INDEX_NAMES *in = &s_index_names[indx->type]; + + a = ni_find_attr(ni, NULL, &le, ATTR_ROOT, in->name, in->name_len, NULL, + mi); + if (!a) + return NULL; + + if (attr) + *attr = a; + + return resident_data_ex(a, sizeof(struct INDEX_ROOT)); +} + +static int indx_write(struct ntfs_index *indx, struct ntfs_inode *ni, + struct indx_node *node, int sync) +{ + struct INDEX_BUFFER *ib = node->index; + + return ntfs_write_bh(ni->mi.sbi, &ib->rhdr, &node->nb, sync); +} + +/* + * if ntfs_readdir calls this function + * inode is shared locked and no ni_lock + * use rw_semaphore for read/write access to alloc_run + */ +int indx_read(struct ntfs_index *indx, struct ntfs_inode *ni, CLST vbn, + struct indx_node **node) +{ + int err; + struct INDEX_BUFFER *ib; + struct runs_tree *run = &indx->alloc_run; + struct rw_semaphore *lock = &indx->run_lock; + u64 vbo = (u64)vbn << indx->vbn2vbo_bits; + u32 bytes = 1u << indx->index_bits; + struct indx_node *in = *node; + const struct INDEX_NAMES *name; + + if (!in) { + in = ntfs_zalloc(sizeof(struct indx_node)); + if (!in) + return -ENOMEM; + } else { + nb_put(&in->nb); + } + + ib = in->index; + if (!ib) { + ib = ntfs_malloc(bytes); + if (!ib) { + err = -ENOMEM; + goto out; + } + } + + down_read(lock); + err = ntfs_read_bh(ni->mi.sbi, run, vbo, &ib->rhdr, bytes, &in->nb); + up_read(lock); + if (!err) + goto ok; + + if (err == -E_NTFS_FIXUP) + goto ok; + + if (err != -ENOENT) + goto out; + + name = &s_index_names[indx->type]; + down_write(lock); + err = attr_load_runs_range(ni, ATTR_ALLOC, name->name, name->name_len, + run, vbo, vbo + bytes); + up_write(lock); + if (err) + goto out; + + down_read(lock); + err = ntfs_read_bh(ni->mi.sbi, run, vbo, &ib->rhdr, bytes, &in->nb); + up_read(lock); + if (err == -E_NTFS_FIXUP) + goto ok; + + if (err) + goto out; + +ok: + if (err == -E_NTFS_FIXUP) { + ntfs_write_bh(ni->mi.sbi, &ib->rhdr, &in->nb, 0); + err = 0; + } + + in->index = ib; + *node = in; + +out: + if (ib != in->index) + ntfs_free(ib); + + if (*node != in) { + nb_put(&in->nb); + ntfs_free(in); + } + + return err; +} + +/* + * indx_find + * + * scans NTFS directory for given entry + */ +int indx_find(struct ntfs_index *indx, struct ntfs_inode *ni, + const struct INDEX_ROOT *root, const void *key, size_t key_len, + const void *ctx, int *diff, struct NTFS_DE **entry, + struct ntfs_fnd *fnd) +{ + int err; + struct NTFS_DE *e; + const struct INDEX_HDR *hdr; + struct indx_node *node; + + if (!root) + root = indx_get_root(&ni->dir, ni, NULL, NULL); + + if (!root) { + err = -EINVAL; + goto out; + } + + hdr = &root->ihdr; + + /* Check cache */ + e = fnd->level ? fnd->de[fnd->level - 1] : fnd->root_de; + if (e && !de_is_last(e) && + !(*indx->cmp)(key, key_len, e + 1, le16_to_cpu(e->key_size), ctx)) { + *entry = e; + *diff = 0; + return 0; + } + + /* Soft finder reset */ + fnd_clear(fnd); + + /* Lookup entry that is <= to the search value */ + e = hdr_find_e(indx, hdr, key, key_len, ctx, diff); + if (!e) + return -EINVAL; + + if (fnd) + fnd->root_de = e; + + err = 0; + + for (;;) { + node = NULL; + if (*diff >= 0 || !de_has_vcn_ex(e)) { + *entry = e; + goto out; + } + + /* Read next level. */ + err = indx_read(indx, ni, de_get_vbn(e), &node); + if (err) + goto out; + + /* Lookup entry that is <= to the search value */ + e = hdr_find_e(indx, &node->index->ihdr, key, key_len, ctx, + diff); + if (!e) { + err = -EINVAL; + put_indx_node(node); + goto out; + } + + fnd_push(fnd, node, e); + } + +out: + return err; +} + +int indx_find_sort(struct ntfs_index *indx, struct ntfs_inode *ni, + const struct INDEX_ROOT *root, struct NTFS_DE **entry, + struct ntfs_fnd *fnd) +{ + int err; + struct indx_node *n = NULL; + struct NTFS_DE *e; + size_t iter = 0; + int level = fnd->level; + + if (!*entry) { + /* Start find */ + e = hdr_first_de(&root->ihdr); + if (!e) + return 0; + fnd_clear(fnd); + fnd->root_de = e; + } else if (!level) { + if (de_is_last(fnd->root_de)) { + *entry = NULL; + return 0; + } + + e = hdr_next_de(&root->ihdr, fnd->root_de); + if (!e) + return -EINVAL; + fnd->root_de = e; + } else { + n = fnd->nodes[level - 1]; + e = fnd->de[level - 1]; + + if (de_is_last(e)) + goto pop_level; + + e = hdr_next_de(&n->index->ihdr, e); + if (!e) + return -EINVAL; + + fnd->de[level - 1] = e; + } + + /* Just to avoid tree cycle */ +next_iter: + if (iter++ >= 1000) + return -EINVAL; + + while (de_has_vcn_ex(e)) { + if (le16_to_cpu(e->size) < + sizeof(struct NTFS_DE) + sizeof(u64)) { + if (n) { + fnd_pop(fnd); + ntfs_free(n); + } + return -EINVAL; + } + + /* Read next level */ + err = indx_read(indx, ni, de_get_vbn(e), &n); + if (err) + return err; + + /* Try next level */ + e = hdr_first_de(&n->index->ihdr); + if (!e) { + ntfs_free(n); + return -EINVAL; + } + + fnd_push(fnd, n, e); + } + + if (le16_to_cpu(e->size) > sizeof(struct NTFS_DE)) { + *entry = e; + return 0; + } + +pop_level: + for (;;) { + if (!de_is_last(e)) + goto next_iter; + + /* Pop one level */ + if (n) { + fnd_pop(fnd); + ntfs_free(n); + } + + level = fnd->level; + + if (level) { + n = fnd->nodes[level - 1]; + e = fnd->de[level - 1]; + } else if (fnd->root_de) { + n = NULL; + e = fnd->root_de; + fnd->root_de = NULL; + } else { + *entry = NULL; + return 0; + } + + if (le16_to_cpu(e->size) > sizeof(struct NTFS_DE)) { + *entry = e; + if (!fnd->root_de) + fnd->root_de = e; + return 0; + } + } +} + +int indx_find_raw(struct ntfs_index *indx, struct ntfs_inode *ni, + const struct INDEX_ROOT *root, struct NTFS_DE **entry, + size_t *off, struct ntfs_fnd *fnd) +{ + int err; + struct indx_node *n = NULL; + struct NTFS_DE *e = NULL; + struct NTFS_DE *e2; + size_t bit; + CLST next_used_vbn; + CLST next_vbn; + u32 record_size = ni->mi.sbi->record_size; + + /* Use non sorted algorithm */ + if (!*entry) { + /* This is the first call */ + e = hdr_first_de(&root->ihdr); + if (!e) + return 0; + fnd_clear(fnd); + fnd->root_de = e; + + /* The first call with setup of initial element */ + if (*off >= record_size) { + next_vbn = (((*off - record_size) >> indx->index_bits)) + << indx->idx2vbn_bits; + /* jump inside cycle 'for'*/ + goto next; + } + + /* Start enumeration from root */ + *off = 0; + } else if (!fnd->root_de) + return -EINVAL; + + for (;;) { + /* Check if current entry can be used */ + if (e && le16_to_cpu(e->size) > sizeof(struct NTFS_DE)) + goto ok; + + if (!fnd->level) { + /* Continue to enumerate root */ + if (!de_is_last(fnd->root_de)) { + e = hdr_next_de(&root->ihdr, fnd->root_de); + if (!e) + return -EINVAL; + fnd->root_de = e; + continue; + } + + /* Start to enumerate indexes from 0 */ + next_vbn = 0; + } else { + /* Continue to enumerate indexes */ + e2 = fnd->de[fnd->level - 1]; + + n = fnd->nodes[fnd->level - 1]; + + if (!de_is_last(e2)) { + e = hdr_next_de(&n->index->ihdr, e2); + if (!e) + return -EINVAL; + fnd->de[fnd->level - 1] = e; + continue; + } + + /* Continue with next index */ + next_vbn = le64_to_cpu(n->index->vbn) + + root->index_block_clst; + } + +next: + /* Release current index */ + if (n) { + fnd_pop(fnd); + put_indx_node(n); + n = NULL; + } + + /* Skip all free indexes */ + bit = next_vbn >> indx->idx2vbn_bits; + err = indx_used_bit(indx, ni, &bit); + if (err == -ENOENT || bit == MINUS_ONE_T) { + /* No used indexes */ + *entry = NULL; + return 0; + } + + next_used_vbn = bit << indx->idx2vbn_bits; + + /* Read buffer into memory */ + err = indx_read(indx, ni, next_used_vbn, &n); + if (err) + return err; + + e = hdr_first_de(&n->index->ihdr); + fnd_push(fnd, n, e); + if (!e) + return -EINVAL; + } + +ok: + /* return offset to restore enumerator if necessary */ + if (!n) { + /* 'e' points in root */ + *off = PtrOffset(&root->ihdr, e); + } else { + /* 'e' points in index */ + *off = (le64_to_cpu(n->index->vbn) << indx->vbn2vbo_bits) + + record_size + PtrOffset(&n->index->ihdr, e); + } + + *entry = e; + return 0; +} + +/* + * indx_create_allocate + * + * create "Allocation + Bitmap" attributes + */ +static int indx_create_allocate(struct ntfs_index *indx, struct ntfs_inode *ni, + CLST *vbn) +{ + int err = -ENOMEM; + struct ntfs_sb_info *sbi = ni->mi.sbi; + struct ATTRIB *bitmap; + struct ATTRIB *alloc; + u32 data_size = 1u << indx->index_bits; + u32 alloc_size = ntfs_up_cluster(sbi, data_size); + CLST len = alloc_size >> sbi->cluster_bits; + const struct INDEX_NAMES *in = &s_index_names[indx->type]; + CLST alen; + struct runs_tree run; + + run_init(&run); + + err = attr_allocate_clusters(sbi, &run, 0, 0, len, NULL, 0, &alen, 0, + NULL); + if (err) + goto out; + + err = ni_insert_nonresident(ni, ATTR_ALLOC, in->name, in->name_len, + &run, 0, len, 0, &alloc, NULL); + if (err) + goto out1; + + alloc->nres.valid_size = alloc->nres.data_size = cpu_to_le64(data_size); + + err = ni_insert_resident(ni, bitmap_size(1), ATTR_BITMAP, in->name, + in->name_len, &bitmap, NULL); + if (err) + goto out2; + + if (in->name == I30_NAME) { + ni->vfs_inode.i_size = data_size; + inode_set_bytes(&ni->vfs_inode, alloc_size); + } + + memcpy(&indx->alloc_run, &run, sizeof(run)); + + *vbn = 0; + + return 0; + +out2: + mi_remove_attr(&ni->mi, alloc); + +out1: + run_deallocate(sbi, &run, false); + +out: + return err; +} + +/* + * indx_add_allocate + * + * add clusters to index + */ +static int indx_add_allocate(struct ntfs_index *indx, struct ntfs_inode *ni, + CLST *vbn) +{ + int err; + size_t bit; + u64 data_size; + u64 bmp_size, bmp_size_v; + struct ATTRIB *bmp, *alloc; + struct mft_inode *mi; + const struct INDEX_NAMES *in = &s_index_names[indx->type]; + + err = indx_find_free(indx, ni, &bit, &bmp); + if (err) + goto out1; + + if (bit != MINUS_ONE_T) { + bmp = NULL; + } else { + if (bmp->non_res) { + bmp_size = le64_to_cpu(bmp->nres.data_size); + bmp_size_v = le64_to_cpu(bmp->nres.valid_size); + } else { + bmp_size = bmp_size_v = le32_to_cpu(bmp->res.data_size); + } + + bit = bmp_size << 3; + } + + data_size = (u64)(bit + 1) << indx->index_bits; + + if (bmp) { + /* Increase bitmap */ + err = attr_set_size(ni, ATTR_BITMAP, in->name, in->name_len, + &indx->bitmap_run, bitmap_size(bit + 1), + NULL, true, NULL); + if (err) + goto out1; + } + + alloc = ni_find_attr(ni, NULL, NULL, ATTR_ALLOC, in->name, in->name_len, + NULL, &mi); + if (!alloc) { + if (bmp) + goto out2; + goto out1; + } + + /* Increase allocation */ + err = attr_set_size(ni, ATTR_ALLOC, in->name, in->name_len, + &indx->alloc_run, data_size, &data_size, true, + NULL); + if (err) { + if (bmp) + goto out2; + goto out1; + } + + *vbn = bit << indx->idx2vbn_bits; + + return 0; + +out2: + /* Ops (no space?) */ + attr_set_size(ni, ATTR_BITMAP, in->name, in->name_len, + &indx->bitmap_run, bmp_size, &bmp_size_v, false, NULL); + +out1: + return err; +} + +/* + * indx_insert_into_root + * + * attempts to insert an entry into the index root + * If necessary, it will twiddle the index b-tree. + */ +static int indx_insert_into_root(struct ntfs_index *indx, struct ntfs_inode *ni, + const struct NTFS_DE *new_de, + struct NTFS_DE *root_de, const void *ctx, + struct ntfs_fnd *fnd) +{ + int err = 0; + struct NTFS_DE *e, *e0, *re; + struct mft_inode *mi; + struct ATTRIB *attr; + struct MFT_REC *rec; + struct INDEX_HDR *hdr; + struct indx_node *n; + CLST new_vbn; + __le64 *sub_vbn, t_vbn; + u16 new_de_size; + u32 hdr_used, hdr_total, asize, used, to_move; + u32 root_size, new_root_size; + struct ntfs_sb_info *sbi; + int ds_root; + struct INDEX_ROOT *root, *a_root = NULL; + + /* Get the record this root placed in */ + root = indx_get_root(indx, ni, &attr, &mi); + if (!root) + goto out; + + /* + * Try easy case: + * hdr_insert_de will succeed if there's room the root for the new entry. + */ + hdr = &root->ihdr; + sbi = ni->mi.sbi; + rec = mi->mrec; + used = le32_to_cpu(rec->used); + new_de_size = le16_to_cpu(new_de->size); + hdr_used = le32_to_cpu(hdr->used); + hdr_total = le32_to_cpu(hdr->total); + asize = le32_to_cpu(attr->size); + root_size = le32_to_cpu(attr->res.data_size); + + ds_root = new_de_size + hdr_used - hdr_total; + + if (used + ds_root < sbi->max_bytes_per_attr) { + /* make a room for new elements */ + mi_resize_attr(mi, attr, ds_root); + hdr->total = cpu_to_le32(hdr_total + ds_root); + e = hdr_insert_de(indx, hdr, new_de, root_de, ctx); + WARN_ON(!e); + fnd_clear(fnd); + fnd->root_de = e; + + return 0; + } + + /* Make a copy of root attribute to restore if error */ + a_root = ntfs_memdup(attr, asize); + if (!a_root) { + err = -ENOMEM; + goto out; + } + + /* copy all the non-end entries from the index root to the new buffer.*/ + to_move = 0; + e0 = hdr_first_de(hdr); + + /* Calculate the size to copy */ + for (e = e0;; e = hdr_next_de(hdr, e)) { + if (!e) { + err = -EINVAL; + goto out; + } + + if (de_is_last(e)) + break; + to_move += le16_to_cpu(e->size); + } + + n = NULL; + if (!to_move) { + re = NULL; + } else { + re = ntfs_memdup(e0, to_move); + if (!re) { + err = -ENOMEM; + goto out; + } + } + + sub_vbn = NULL; + if (de_has_vcn(e)) { + t_vbn = de_get_vbn_le(e); + sub_vbn = &t_vbn; + } + + new_root_size = sizeof(struct INDEX_ROOT) + sizeof(struct NTFS_DE) + + sizeof(u64); + ds_root = new_root_size - root_size; + + if (ds_root > 0 && used + ds_root > sbi->max_bytes_per_attr) { + /* make root external */ + err = -EOPNOTSUPP; + goto out; + } + + if (ds_root) + mi_resize_attr(mi, attr, ds_root); + + /* Fill first entry (vcn will be set later) */ + e = (struct NTFS_DE *)(root + 1); + memset(e, 0, sizeof(struct NTFS_DE)); + e->size = cpu_to_le16(sizeof(struct NTFS_DE) + sizeof(u64)); + e->flags = NTFS_IE_HAS_SUBNODES | NTFS_IE_LAST; + + hdr->flags = 1; + hdr->used = hdr->total = + cpu_to_le32(new_root_size - offsetof(struct INDEX_ROOT, ihdr)); + + fnd->root_de = hdr_first_de(hdr); + mi->dirty = true; + + /* Create alloc and bitmap attributes (if not) */ + err = run_is_empty(&indx->alloc_run) + ? indx_create_allocate(indx, ni, &new_vbn) + : indx_add_allocate(indx, ni, &new_vbn); + + /* layout of record may be changed, so rescan root */ + root = indx_get_root(indx, ni, &attr, &mi); + if (!root) { + /* bug? */ + ntfs_set_state(sbi, NTFS_DIRTY_ERROR); + err = -EINVAL; + goto out1; + } + + if (err) { + /* restore root */ + if (mi_resize_attr(mi, attr, -ds_root)) + memcpy(attr, a_root, asize); + else { + /* bug? */ + ntfs_set_state(sbi, NTFS_DIRTY_ERROR); + } + goto out1; + } + + e = (struct NTFS_DE *)(root + 1); + *(__le64 *)(e + 1) = cpu_to_le64(new_vbn); + mi->dirty = true; + + /* now we can create/format the new buffer and copy the entries into */ + n = indx_new(indx, ni, new_vbn, sub_vbn); + if (IS_ERR(n)) { + err = PTR_ERR(n); + goto out1; + } + + hdr = &n->index->ihdr; + hdr_used = le32_to_cpu(hdr->used); + hdr_total = le32_to_cpu(hdr->total); + + /* Copy root entries into new buffer */ + hdr_insert_head(hdr, re, to_move); + + /* Update bitmap attribute */ + indx_mark_used(indx, ni, new_vbn >> indx->idx2vbn_bits); + + /* Check if we can insert new entry new index buffer */ + if (hdr_used + new_de_size > hdr_total) { + /* + * This occurs if mft record is the same or bigger than index + * buffer. Move all root new index and have no space to add + * new entry classic case when mft record is 1K and index + * buffer 4K the problem should not occurs + */ + ntfs_free(re); + indx_write(indx, ni, n, 0); + + put_indx_node(n); + fnd_clear(fnd); + err = indx_insert_entry(indx, ni, new_de, ctx, fnd); + goto out; + } + + /* + * Now root is a parent for new index buffer + * Insert NewEntry a new buffer + */ + e = hdr_insert_de(indx, hdr, new_de, NULL, ctx); + if (!e) { + err = -EINVAL; + goto out1; + } + fnd_push(fnd, n, e); + + /* Just write updates index into disk */ + indx_write(indx, ni, n, 0); + + n = NULL; + +out1: + ntfs_free(re); + if (n) + put_indx_node(n); + +out: + ntfs_free(a_root); + return err; +} + +/* + * indx_insert_into_buffer + * + * attempts to insert an entry into an Index Allocation Buffer. + * If necessary, it will split the buffer. + */ +static int +indx_insert_into_buffer(struct ntfs_index *indx, struct ntfs_inode *ni, + struct INDEX_ROOT *root, const struct NTFS_DE *new_de, + const void *ctx, int level, struct ntfs_fnd *fnd) +{ + int err; + const struct NTFS_DE *sp; + struct NTFS_DE *e, *de_t, *up_e = NULL; + struct indx_node *n2 = NULL; + struct indx_node *n1 = fnd->nodes[level]; + struct INDEX_HDR *hdr1 = &n1->index->ihdr; + struct INDEX_HDR *hdr2; + u32 to_copy, used; + CLST new_vbn; + __le64 t_vbn, *sub_vbn; + u16 sp_size; + + /* Try the most easy case */ + e = fnd->level - 1 == level ? fnd->de[level] : NULL; + e = hdr_insert_de(indx, hdr1, new_de, e, ctx); + fnd->de[level] = e; + if (e) { + /* Just write updated index into disk */ + indx_write(indx, ni, n1, 0); + return 0; + } + + /* + * No space to insert into buffer. Split it. + * To split we: + * - Save split point ('cause index buffers will be changed) + * - Allocate NewBuffer and copy all entries <= sp into new buffer + * - Remove all entries (sp including) from TargetBuffer + * - Insert NewEntry into left or right buffer (depending on sp <=> + * NewEntry) + * - Insert sp into parent buffer (or root) + * - Make sp a parent for new buffer + */ + sp = hdr_find_split(hdr1); + if (!sp) + return -EINVAL; + + sp_size = le16_to_cpu(sp->size); + up_e = ntfs_malloc(sp_size + sizeof(u64)); + if (!up_e) + return -ENOMEM; + memcpy(up_e, sp, sp_size); + + if (!hdr1->flags) { + up_e->flags |= NTFS_IE_HAS_SUBNODES; + up_e->size = cpu_to_le16(sp_size + sizeof(u64)); + sub_vbn = NULL; + } else { + t_vbn = de_get_vbn_le(up_e); + sub_vbn = &t_vbn; + } + + /* Allocate on disk a new index allocation buffer. */ + err = indx_add_allocate(indx, ni, &new_vbn); + if (err) + goto out; + + /* Allocate and format memory a new index buffer */ + n2 = indx_new(indx, ni, new_vbn, sub_vbn); + if (IS_ERR(n2)) { + err = PTR_ERR(n2); + goto out; + } + + hdr2 = &n2->index->ihdr; + + /* Make sp a parent for new buffer */ + de_set_vbn(up_e, new_vbn); + + /* copy all the entries <= sp into the new buffer. */ + de_t = hdr_first_de(hdr1); + to_copy = PtrOffset(de_t, sp); + hdr_insert_head(hdr2, de_t, to_copy); + + /* remove all entries (sp including) from hdr1 */ + used = le32_to_cpu(hdr1->used) - to_copy - sp_size; + memmove(de_t, Add2Ptr(sp, sp_size), used - le32_to_cpu(hdr1->de_off)); + hdr1->used = cpu_to_le32(used); + + /* Insert new entry into left or right buffer (depending on sp <=> new_de) */ + hdr_insert_de(indx, + (*indx->cmp)(new_de + 1, le16_to_cpu(new_de->key_size), + up_e + 1, le16_to_cpu(up_e->key_size), + ctx) < 0 + ? hdr2 + : hdr1, + new_de, NULL, ctx); + + indx_mark_used(indx, ni, new_vbn >> indx->idx2vbn_bits); + + indx_write(indx, ni, n1, 0); + indx_write(indx, ni, n2, 0); + + put_indx_node(n2); + + /* + * we've finished splitting everybody, so we are ready to + * insert the promoted entry into the parent. + */ + if (!level) { + /* Insert in root */ + err = indx_insert_into_root(indx, ni, up_e, NULL, ctx, fnd); + if (err) + goto out; + } else { + /* + * The target buffer's parent is another index buffer + * TODO: Remove recursion + */ + err = indx_insert_into_buffer(indx, ni, root, up_e, ctx, + level - 1, fnd); + if (err) + goto out; + } + +out: + ntfs_free(up_e); + + return err; +} + +/* + * indx_insert_entry + * + * inserts new entry into index + */ +int indx_insert_entry(struct ntfs_index *indx, struct ntfs_inode *ni, + const struct NTFS_DE *new_de, const void *ctx, + struct ntfs_fnd *fnd) +{ + int err; + int diff; + struct NTFS_DE *e; + struct ntfs_fnd *fnd_a = NULL; + struct INDEX_ROOT *root; + + if (!fnd) { + fnd_a = fnd_get(); + if (!fnd_a) { + err = -ENOMEM; + goto out1; + } + fnd = fnd_a; + } + + root = indx_get_root(indx, ni, NULL, NULL); + if (!root) { + err = -EINVAL; + goto out; + } + + if (fnd_is_empty(fnd)) { + /* Find the spot the tree where we want to insert the new entry. */ + err = indx_find(indx, ni, root, new_de + 1, + le16_to_cpu(new_de->key_size), ctx, &diff, &e, + fnd); + if (err) + goto out; + + if (!diff) { + err = -EEXIST; + goto out; + } + } + + if (!fnd->level) { + /* The root is also a leaf, so we'll insert the new entry into it. */ + err = indx_insert_into_root(indx, ni, new_de, fnd->root_de, ctx, + fnd); + if (err) + goto out; + } else { + /* found a leaf buffer, so we'll insert the new entry into it.*/ + err = indx_insert_into_buffer(indx, ni, root, new_de, ctx, + fnd->level - 1, fnd); + if (err) + goto out; + } + +out: + fnd_put(fnd_a); +out1: + return err; +} + +/* + * indx_find_buffer + * + * locates a buffer the tree. + */ +static struct indx_node *indx_find_buffer(struct ntfs_index *indx, + struct ntfs_inode *ni, + const struct INDEX_ROOT *root, + __le64 vbn, struct indx_node *n) +{ + int err; + const struct NTFS_DE *e; + struct indx_node *r; + const struct INDEX_HDR *hdr = n ? &n->index->ihdr : &root->ihdr; + + /* Step 1: Scan one level */ + for (e = hdr_first_de(hdr);; e = hdr_next_de(hdr, e)) { + if (!e) + return ERR_PTR(-EINVAL); + + if (de_has_vcn(e) && vbn == de_get_vbn_le(e)) + return n; + + if (de_is_last(e)) + break; + } + + /* Step2: Do recursion */ + e = Add2Ptr(hdr, le32_to_cpu(hdr->de_off)); + for (;;) { + if (de_has_vcn_ex(e)) { + err = indx_read(indx, ni, de_get_vbn(e), &n); + if (err) + return ERR_PTR(err); + + r = indx_find_buffer(indx, ni, root, vbn, n); + if (r) + return r; + } + + if (de_is_last(e)) + break; + + e = Add2Ptr(e, le16_to_cpu(e->size)); + } + + return NULL; +} + +/* + * indx_shrink + * + * deallocates unused tail indexes + */ +static int indx_shrink(struct ntfs_index *indx, struct ntfs_inode *ni, + size_t bit) +{ + int err = 0; + u64 bpb, new_data; + size_t nbits; + struct ATTRIB *b; + struct ATTR_LIST_ENTRY *le = NULL; + const struct INDEX_NAMES *in = &s_index_names[indx->type]; + + b = ni_find_attr(ni, NULL, &le, ATTR_BITMAP, in->name, in->name_len, + NULL, NULL); + + if (!b) + return -ENOENT; + + if (!b->non_res) { + unsigned long pos; + const unsigned long *bm = resident_data(b); + + nbits = le32_to_cpu(b->res.data_size) * 8; + + if (bit >= nbits) + return 0; + + pos = find_next_bit(bm, nbits, bit); + if (pos < nbits) + return 0; + } else { + size_t used = MINUS_ONE_T; + + nbits = le64_to_cpu(b->nres.data_size) * 8; + + if (bit >= nbits) + return 0; + + err = scan_nres_bitmap(ni, b, indx, bit, &scan_for_used, &used); + if (err) + return err; + + if (used != MINUS_ONE_T) + return 0; + } + + new_data = (u64)bit << indx->index_bits; + + err = attr_set_size(ni, ATTR_ALLOC, in->name, in->name_len, + &indx->alloc_run, new_data, &new_data, false, NULL); + if (err) + return err; + + bpb = bitmap_size(bit); + if (bpb * 8 == nbits) + return 0; + + err = attr_set_size(ni, ATTR_BITMAP, in->name, in->name_len, + &indx->bitmap_run, bpb, &bpb, false, NULL); + + return err; +} + +static int indx_free_children(struct ntfs_index *indx, struct ntfs_inode *ni, + const struct NTFS_DE *e, bool trim) +{ + int err; + struct indx_node *n; + struct INDEX_HDR *hdr; + CLST vbn = de_get_vbn(e); + size_t i; + + err = indx_read(indx, ni, vbn, &n); + if (err) + return err; + + hdr = &n->index->ihdr; + /* First, recurse into the children, if any.*/ + if (hdr_has_subnode(hdr)) { + for (e = hdr_first_de(hdr); e; e = hdr_next_de(hdr, e)) { + indx_free_children(indx, ni, e, false); + if (de_is_last(e)) + break; + } + } + + put_indx_node(n); + + i = vbn >> indx->idx2vbn_bits; + /* We've gotten rid of the children; add this buffer to the free list. */ + indx_mark_free(indx, ni, i); + + if (!trim) + return 0; + + /* + * If there are no used indexes after current free index + * then we can truncate allocation and bitmap + * Use bitmap to estimate the case + */ + indx_shrink(indx, ni, i + 1); + return 0; +} + +/* + * indx_get_entry_to_replace + * + * finds a replacement entry for a deleted entry + * always returns a node entry: + * NTFS_IE_HAS_SUBNODES is set the flags and the size includes the sub_vcn + */ +static int indx_get_entry_to_replace(struct ntfs_index *indx, + struct ntfs_inode *ni, + const struct NTFS_DE *de_next, + struct NTFS_DE **de_to_replace, + struct ntfs_fnd *fnd) +{ + int err; + int level = -1; + CLST vbn; + struct NTFS_DE *e, *te, *re; + struct indx_node *n; + struct INDEX_BUFFER *ib; + + *de_to_replace = NULL; + + /* Find first leaf entry down from de_next */ + vbn = de_get_vbn(de_next); + for (;;) { + n = NULL; + err = indx_read(indx, ni, vbn, &n); + if (err) + goto out; + + e = hdr_first_de(&n->index->ihdr); + fnd_push(fnd, n, e); + + if (!de_is_last(e)) { + /* + * This buffer is non-empty, so its first entry could be used as the + * replacement entry. + */ + level = fnd->level - 1; + } + + if (!de_has_vcn(e)) + break; + + /* This buffer is a node. Continue to go down */ + vbn = de_get_vbn(e); + } + + if (level == -1) + goto out; + + n = fnd->nodes[level]; + te = hdr_first_de(&n->index->ihdr); + /* Copy the candidate entry into the replacement entry buffer. */ + re = ntfs_malloc(le16_to_cpu(te->size) + sizeof(u64)); + if (!re) { + err = -ENOMEM; + goto out; + } + + *de_to_replace = re; + memcpy(re, te, le16_to_cpu(te->size)); + + if (!de_has_vcn(re)) { + /* + * The replacement entry we found doesn't have a sub_vcn. increase its size + * to hold one. + */ + le16_add_cpu(&re->size, sizeof(u64)); + re->flags |= NTFS_IE_HAS_SUBNODES; + } else { + /* + * The replacement entry we found was a node entry, which means that all + * its child buffers are empty. Return them to the free pool. + */ + indx_free_children(indx, ni, te, true); + } + + /* + * Expunge the replacement entry from its former location, + * and then write that buffer. + */ + ib = n->index; + e = hdr_delete_de(&ib->ihdr, te); + + fnd->de[level] = e; + indx_write(indx, ni, n, 0); + + /* Check to see if this action created an empty leaf. */ + if (ib_is_leaf(ib) && ib_is_empty(ib)) + return 0; + +out: + fnd_clear(fnd); + return err; +} + +/* + * indx_delete_entry + * + * deletes an entry from the index. + */ +int indx_delete_entry(struct ntfs_index *indx, struct ntfs_inode *ni, + const void *key, u32 key_len, const void *ctx) +{ + int err, diff; + struct INDEX_ROOT *root; + struct INDEX_HDR *hdr; + struct ntfs_fnd *fnd, *fnd2; + struct INDEX_BUFFER *ib; + struct NTFS_DE *e, *re, *next, *prev, *me; + struct indx_node *n, *n2d = NULL; + __le64 sub_vbn; + int level, level2; + struct ATTRIB *attr; + struct mft_inode *mi; + u32 e_size, root_size, new_root_size; + size_t trim_bit; + const struct INDEX_NAMES *in; + + fnd = fnd_get(); + if (!fnd) { + err = -ENOMEM; + goto out2; + } + + fnd2 = fnd_get(); + if (!fnd2) { + err = -ENOMEM; + goto out1; + } + + root = indx_get_root(indx, ni, &attr, &mi); + if (!root) { + err = -EINVAL; + goto out; + } + + /* Locate the entry to remove. */ + err = indx_find(indx, ni, root, key, key_len, ctx, &diff, &e, fnd); + if (err) + goto out; + + if (!e || diff) { + err = -ENOENT; + goto out; + } + + level = fnd->level; + + if (level) { + n = fnd->nodes[level - 1]; + e = fnd->de[level - 1]; + ib = n->index; + hdr = &ib->ihdr; + } else { + hdr = &root->ihdr; + e = fnd->root_de; + n = NULL; + } + + e_size = le16_to_cpu(e->size); + + if (!de_has_vcn_ex(e)) { + /* The entry to delete is a leaf, so we can just rip it out */ + hdr_delete_de(hdr, e); + + if (!level) { + hdr->total = hdr->used; + + /* Shrink resident root attribute */ + mi_resize_attr(mi, attr, 0 - e_size); + goto out; + } + + indx_write(indx, ni, n, 0); + + /* + * Check to see if removing that entry made + * the leaf empty. + */ + if (ib_is_leaf(ib) && ib_is_empty(ib)) { + fnd_pop(fnd); + fnd_push(fnd2, n, e); + } + } else { + /* + * The entry we wish to delete is a node buffer, so we + * have to find a replacement for it. + */ + next = de_get_next(e); + + err = indx_get_entry_to_replace(indx, ni, next, &re, fnd2); + if (err) + goto out; + + if (re) { + de_set_vbn_le(re, de_get_vbn_le(e)); + hdr_delete_de(hdr, e); + + err = level ? indx_insert_into_buffer(indx, ni, root, + re, ctx, + fnd->level - 1, + fnd) + : indx_insert_into_root(indx, ni, re, e, + ctx, fnd); + ntfs_free(re); + + if (err) + goto out; + } else { + /* + * There is no replacement for the current entry. + * This means that the subtree rooted at its node is empty, + * and can be deleted, which turn means that the node can + * just inherit the deleted entry sub_vcn + */ + indx_free_children(indx, ni, next, true); + + de_set_vbn_le(next, de_get_vbn_le(e)); + hdr_delete_de(hdr, e); + if (level) { + indx_write(indx, ni, n, 0); + } else { + hdr->total = hdr->used; + + /* Shrink resident root attribute */ + mi_resize_attr(mi, attr, 0 - e_size); + } + } + } + + /* Delete a branch of tree */ + if (!fnd2 || !fnd2->level) + goto out; + + /* Reinit root 'cause it can be changed */ + root = indx_get_root(indx, ni, &attr, &mi); + if (!root) { + err = -EINVAL; + goto out; + } + + n2d = NULL; + sub_vbn = fnd2->nodes[0]->index->vbn; + level2 = 0; + level = fnd->level; + + hdr = level ? &fnd->nodes[level - 1]->index->ihdr : &root->ihdr; + + /* Scan current level */ + for (e = hdr_first_de(hdr);; e = hdr_next_de(hdr, e)) { + if (!e) { + err = -EINVAL; + goto out; + } + + if (de_has_vcn(e) && sub_vbn == de_get_vbn_le(e)) + break; + + if (de_is_last(e)) { + e = NULL; + break; + } + } + + if (!e) { + /* Do slow search from root */ + struct indx_node *in; + + fnd_clear(fnd); + + in = indx_find_buffer(indx, ni, root, sub_vbn, NULL); + if (IS_ERR(in)) { + err = PTR_ERR(in); + goto out; + } + + if (in) + fnd_push(fnd, in, NULL); + } + + /* Merge fnd2 -> fnd */ + for (level = 0; level < fnd2->level; level++) { + fnd_push(fnd, fnd2->nodes[level], fnd2->de[level]); + fnd2->nodes[level] = NULL; + } + fnd2->level = 0; + + hdr = NULL; + for (level = fnd->level; level; level--) { + struct indx_node *in = fnd->nodes[level - 1]; + + ib = in->index; + if (ib_is_empty(ib)) { + sub_vbn = ib->vbn; + } else { + hdr = &ib->ihdr; + n2d = in; + level2 = level; + break; + } + } + + if (!hdr) + hdr = &root->ihdr; + + e = hdr_first_de(hdr); + if (!e) { + err = -EINVAL; + goto out; + } + + if (hdr != &root->ihdr || !de_is_last(e)) { + prev = NULL; + while (!de_is_last(e)) { + if (de_has_vcn(e) && sub_vbn == de_get_vbn_le(e)) + break; + prev = e; + e = hdr_next_de(hdr, e); + if (!e) { + err = -EINVAL; + goto out; + } + } + + if (sub_vbn != de_get_vbn_le(e)) { + /* + * Didn't find the parent entry, although this buffer is the parent trail. + * Something is corrupt. + */ + err = -EINVAL; + goto out; + } + + if (de_is_last(e)) { + /* + * Since we can't remove the end entry, we'll remove its + * predecessor instead. This means we have to transfer the + * predecessor's sub_vcn to the end entry. + * Note: that this index block is not empty, so the + * predecessor must exist + */ + if (!prev) { + err = -EINVAL; + goto out; + } + + if (de_has_vcn(prev)) { + de_set_vbn_le(e, de_get_vbn_le(prev)); + } else if (de_has_vcn(e)) { + le16_sub_cpu(&e->size, sizeof(u64)); + e->flags &= ~NTFS_IE_HAS_SUBNODES; + le32_sub_cpu(&hdr->used, sizeof(u64)); + } + e = prev; + } + + /* + * Copy the current entry into a temporary buffer (stripping off its + * down-pointer, if any) and delete it from the current buffer or root, + * as appropriate. + */ + e_size = le16_to_cpu(e->size); + me = ntfs_memdup(e, e_size); + if (!me) { + err = -ENOMEM; + goto out; + } + + if (de_has_vcn(me)) { + me->flags &= ~NTFS_IE_HAS_SUBNODES; + le16_sub_cpu(&me->size, sizeof(u64)); + } + + hdr_delete_de(hdr, e); + + if (hdr == &root->ihdr) { + level = 0; + hdr->total = hdr->used; + + /* Shrink resident root attribute */ + mi_resize_attr(mi, attr, 0 - e_size); + } else { + indx_write(indx, ni, n2d, 0); + level = level2; + } + + /* Mark unused buffers as free */ + trim_bit = -1; + for (; level < fnd->level; level++) { + ib = fnd->nodes[level]->index; + if (ib_is_empty(ib)) { + size_t k = le64_to_cpu(ib->vbn) >> + indx->idx2vbn_bits; + + indx_mark_free(indx, ni, k); + if (k < trim_bit) + trim_bit = k; + } + } + + fnd_clear(fnd); + /*fnd->root_de = NULL;*/ + + /* + * Re-insert the entry into the tree. + * Find the spot the tree where we want to insert the new entry. + */ + err = indx_insert_entry(indx, ni, me, ctx, fnd); + ntfs_free(me); + if (err) + goto out; + + if (trim_bit != -1) + indx_shrink(indx, ni, trim_bit); + } else { + /* + * This tree needs to be collapsed down to an empty root. + * Recreate the index root as an empty leaf and free all the bits the + * index allocation bitmap. + */ + fnd_clear(fnd); + fnd_clear(fnd2); + + in = &s_index_names[indx->type]; + + err = attr_set_size(ni, ATTR_ALLOC, in->name, in->name_len, + &indx->alloc_run, 0, NULL, false, NULL); + err = ni_remove_attr(ni, ATTR_ALLOC, in->name, in->name_len, + false, NULL); + run_close(&indx->alloc_run); + + err = attr_set_size(ni, ATTR_BITMAP, in->name, in->name_len, + &indx->bitmap_run, 0, NULL, false, NULL); + err = ni_remove_attr(ni, ATTR_BITMAP, in->name, in->name_len, + false, NULL); + run_close(&indx->bitmap_run); + + root = indx_get_root(indx, ni, &attr, &mi); + if (!root) { + err = -EINVAL; + goto out; + } + + root_size = le32_to_cpu(attr->res.data_size); + new_root_size = + sizeof(struct INDEX_ROOT) + sizeof(struct NTFS_DE); + + if (new_root_size != root_size && + !mi_resize_attr(mi, attr, new_root_size - root_size)) { + err = -EINVAL; + goto out; + } + + /* Fill first entry */ + e = (struct NTFS_DE *)(root + 1); + e->ref.low = 0; + e->ref.high = 0; + e->ref.seq = 0; + e->size = cpu_to_le16(sizeof(struct NTFS_DE)); + e->flags = NTFS_IE_LAST; // 0x02 + e->key_size = 0; + e->res = 0; + + hdr = &root->ihdr; + hdr->flags = 0; + hdr->used = hdr->total = cpu_to_le32( + new_root_size - offsetof(struct INDEX_ROOT, ihdr)); + mi->dirty = true; + } + +out: + fnd_put(fnd2); +out1: + fnd_put(fnd); +out2: + return err; +} + +/* + * Update duplicated information in directory entry + * 'dup' - info from MFT record + */ +int indx_update_dup(struct ntfs_inode *ni, struct ntfs_sb_info *sbi, + const struct ATTR_FILE_NAME *fname, + const struct NTFS_DUP_INFO *dup, int sync) +{ + int err, diff; + struct NTFS_DE *e = NULL; + struct ATTR_FILE_NAME *e_fname; + struct ntfs_fnd *fnd; + struct INDEX_ROOT *root; + struct mft_inode *mi; + struct ntfs_index *indx = &ni->dir; + + fnd = fnd_get(); + if (!fnd) { + err = -ENOMEM; + goto out1; + } + + root = indx_get_root(indx, ni, NULL, &mi); + if (!root) { + err = -EINVAL; + goto out; + } + + /* Find entry in directory */ + err = indx_find(indx, ni, root, fname, fname_full_size(fname), sbi, + &diff, &e, fnd); + if (err) + goto out; + + if (!e) { + err = -EINVAL; + goto out; + } + + if (diff) { + err = -EINVAL; + goto out; + } + + e_fname = (struct ATTR_FILE_NAME *)(e + 1); + + if (!memcmp(&e_fname->dup, dup, sizeof(*dup))) { + /* nothing to update in index! Try to avoid this call */ + goto out; + } + + memcpy(&e_fname->dup, dup, sizeof(*dup)); + + if (fnd->level) { + /* directory entry in index */ + err = indx_write(indx, ni, fnd->nodes[fnd->level - 1], sync); + } else { + /* directory entry in directory MFT record */ + mi->dirty = true; + if (sync) + err = mi_write(mi, 1); + else + mark_inode_dirty(&ni->vfs_inode); + } + +out: + fnd_put(fnd); + +out1: + return err; +} diff --git a/fs/ntfs3/inode.c b/fs/ntfs3/inode.c new file mode 100644 index 000000000000..bf51e294432e --- /dev/null +++ b/fs/ntfs3/inode.c @@ -0,0 +1,2029 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * + * Copyright (C) 2019-2021 Paragon Software GmbH, All rights reserved. + * + */ + +#include <linux/blkdev.h> +#include <linux/buffer_head.h> +#include <linux/fs.h> +#include <linux/iversion.h> +#include <linux/mpage.h> +#include <linux/namei.h> +#include <linux/nls.h> +#include <linux/uio.h> +#include <linux/version.h> +#include <linux/writeback.h> + +#include "debug.h" +#include "ntfs.h" +#include "ntfs_fs.h" + +/* + * ntfs_read_mft + * + * reads record and parses MFT + */ +static struct inode *ntfs_read_mft(struct inode *inode, + const struct cpu_str *name, + const struct MFT_REF *ref) +{ + int err = 0; + struct ntfs_inode *ni = ntfs_i(inode); + struct super_block *sb = inode->i_sb; + struct ntfs_sb_info *sbi = sb->s_fs_info; + mode_t mode = 0; + struct ATTR_STD_INFO5 *std5 = NULL; + struct ATTR_LIST_ENTRY *le; + struct ATTRIB *attr; + bool is_match = false; + bool is_root = false; + bool is_dir; + unsigned long ino = inode->i_ino; + u32 rp_fa = 0, asize, t32; + u16 roff, rsize, names = 0; + const struct ATTR_FILE_NAME *fname = NULL; + const struct INDEX_ROOT *root; + struct REPARSE_DATA_BUFFER rp; // 0x18 bytes + u64 t64; + struct MFT_REC *rec; + struct runs_tree *run; + + inode->i_op = NULL; + /* Setup 'uid' and 'gid' */ + inode->i_uid = sbi->options.fs_uid; + inode->i_gid = sbi->options.fs_gid; + + err = mi_init(&ni->mi, sbi, ino); + if (err) + goto out; + + if (!sbi->mft.ni && ino == MFT_REC_MFT && !sb->s_root) { + t64 = sbi->mft.lbo >> sbi->cluster_bits; + t32 = bytes_to_cluster(sbi, MFT_REC_VOL * sbi->record_size); + sbi->mft.ni = ni; + init_rwsem(&ni->file.run_lock); + + if (!run_add_entry(&ni->file.run, 0, t64, t32, true)) { + err = -ENOMEM; + goto out; + } + } + + err = mi_read(&ni->mi, ino == MFT_REC_MFT); + + if (err) + goto out; + + rec = ni->mi.mrec; + + if (sbi->flags & NTFS_FLAGS_LOG_REPLAYING) { + ; + } else if (ref->seq != rec->seq) { + err = -EINVAL; + ntfs_err(sb, "MFT: r=%lx, expect seq=%x instead of %x!", ino, + le16_to_cpu(ref->seq), le16_to_cpu(rec->seq)); + goto out; + } else if (!is_rec_inuse(rec)) { + err = -EINVAL; + ntfs_err(sb, "Inode r=%x is not in use!", (u32)ino); + goto out; + } + + if (le32_to_cpu(rec->total) != sbi->record_size) { + // bad inode? + err = -EINVAL; + goto out; + } + + if (!is_rec_base(rec)) + goto Ok; + + /* record should contain $I30 root */ + is_dir = rec->flags & RECORD_FLAG_DIR; + + inode->i_generation = le16_to_cpu(rec->seq); + + /* Enumerate all struct Attributes MFT */ + le = NULL; + attr = NULL; + + /* + * to reduce tab pressure use goto instead of + * while( (attr = ni_enum_attr_ex(ni, attr, &le, NULL) )) + */ +next_attr: + run = NULL; + err = -EINVAL; + attr = ni_enum_attr_ex(ni, attr, &le, NULL); + if (!attr) + goto end_enum; + + if (le && le->vcn) { + /* This is non primary attribute segment. Ignore if not MFT */ + if (ino != MFT_REC_MFT || attr->type != ATTR_DATA) + goto next_attr; + + run = &ni->file.run; + asize = le32_to_cpu(attr->size); + goto attr_unpack_run; + } + + roff = attr->non_res ? 0 : le16_to_cpu(attr->res.data_off); + rsize = attr->non_res ? 0 : le32_to_cpu(attr->res.data_size); + asize = le32_to_cpu(attr->size); + + switch (attr->type) { + case ATTR_STD: + if (attr->non_res || + asize < sizeof(struct ATTR_STD_INFO) + roff || + rsize < sizeof(struct ATTR_STD_INFO)) + goto out; + + if (std5) + goto next_attr; + + std5 = Add2Ptr(attr, roff); + +#ifdef STATX_BTIME + nt2kernel(std5->cr_time, &ni->i_crtime); +#endif + nt2kernel(std5->a_time, &inode->i_atime); + nt2kernel(std5->c_time, &inode->i_ctime); + nt2kernel(std5->m_time, &inode->i_mtime); + + ni->std_fa = std5->fa; + + if (asize >= sizeof(struct ATTR_STD_INFO5) + roff && + rsize >= sizeof(struct ATTR_STD_INFO5)) + ni->std_security_id = std5->security_id; + goto next_attr; + + case ATTR_LIST: + if (attr->name_len || le || ino == MFT_REC_LOG) + goto out; + + err = ntfs_load_attr_list(ni, attr); + if (err) + goto out; + + le = NULL; + attr = NULL; + goto next_attr; + + case ATTR_NAME: + if (attr->non_res || asize < SIZEOF_ATTRIBUTE_FILENAME + roff || + rsize < SIZEOF_ATTRIBUTE_FILENAME) + goto out; + + fname = Add2Ptr(attr, roff); + if (fname->type == FILE_NAME_DOS) + goto next_attr; + + names += 1; + if (name && name->len == fname->name_len && + !ntfs_cmp_names_cpu(name, (struct le_str *)&fname->name_len, + NULL, false)) + is_match = true; + + goto next_attr; + + case ATTR_DATA: + if (is_dir) { + /* ignore data attribute in dir record */ + goto next_attr; + } + + if (ino == MFT_REC_BADCLUST && !attr->non_res) + goto next_attr; + + if (attr->name_len && + ((ino != MFT_REC_BADCLUST || !attr->non_res || + attr->name_len != ARRAY_SIZE(BAD_NAME) || + memcmp(attr_name(attr), BAD_NAME, sizeof(BAD_NAME))) && + (ino != MFT_REC_SECURE || !attr->non_res || + attr->name_len != ARRAY_SIZE(SDS_NAME) || + memcmp(attr_name(attr), SDS_NAME, sizeof(SDS_NAME))))) { + /* file contains stream attribute. ignore it */ + goto next_attr; + } + + if (is_attr_sparsed(attr)) + ni->std_fa |= FILE_ATTRIBUTE_SPARSE_FILE; + else + ni->std_fa &= ~FILE_ATTRIBUTE_SPARSE_FILE; + + if (is_attr_compressed(attr)) + ni->std_fa |= FILE_ATTRIBUTE_COMPRESSED; + else + ni->std_fa &= ~FILE_ATTRIBUTE_COMPRESSED; + + if (is_attr_encrypted(attr)) + ni->std_fa |= FILE_ATTRIBUTE_ENCRYPTED; + else + ni->std_fa &= ~FILE_ATTRIBUTE_ENCRYPTED; + + if (!attr->non_res) { + ni->i_valid = inode->i_size = rsize; + inode_set_bytes(inode, rsize); + t32 = asize; + } else { + t32 = le16_to_cpu(attr->nres.run_off); + } + + mode = S_IFREG | (0777 & sbi->options.fs_fmask_inv); + + if (!attr->non_res) { + ni->ni_flags |= NI_FLAG_RESIDENT; + goto next_attr; + } + + inode_set_bytes(inode, attr_ondisk_size(attr)); + + ni->i_valid = le64_to_cpu(attr->nres.valid_size); + inode->i_size = le64_to_cpu(attr->nres.data_size); + if (!attr->nres.alloc_size) + goto next_attr; + + run = ino == MFT_REC_BITMAP ? &sbi->used.bitmap.run + : &ni->file.run; + break; + + case ATTR_ROOT: + if (attr->non_res) + goto out; + + root = Add2Ptr(attr, roff); + is_root = true; + + if (attr->name_len != ARRAY_SIZE(I30_NAME) || + memcmp(attr_name(attr), I30_NAME, sizeof(I30_NAME))) + goto next_attr; + + if (root->type != ATTR_NAME || + root->rule != NTFS_COLLATION_TYPE_FILENAME) + goto out; + + if (!is_dir) + goto next_attr; + + ni->ni_flags |= NI_FLAG_DIR; + + err = indx_init(&ni->dir, sbi, attr, INDEX_MUTEX_I30); + if (err) + goto out; + + mode = sb->s_root + ? (S_IFDIR | (0777 & sbi->options.fs_dmask_inv)) + : (S_IFDIR | 0777); + goto next_attr; + + case ATTR_ALLOC: + if (!is_root || attr->name_len != ARRAY_SIZE(I30_NAME) || + memcmp(attr_name(attr), I30_NAME, sizeof(I30_NAME))) + goto next_attr; + + inode->i_size = le64_to_cpu(attr->nres.data_size); + ni->i_valid = le64_to_cpu(attr->nres.valid_size); + inode_set_bytes(inode, le64_to_cpu(attr->nres.alloc_size)); + + run = &ni->dir.alloc_run; + break; + + case ATTR_BITMAP: + if (ino == MFT_REC_MFT) { + if (!attr->non_res) + goto out; +#ifndef CONFIG_NTFS3_64BIT_CLUSTER + /* 0x20000000 = 2^32 / 8 */ + if (le64_to_cpu(attr->nres.alloc_size) >= 0x20000000) + goto out; +#endif + run = &sbi->mft.bitmap.run; + break; + } else if (is_dir && attr->name_len == ARRAY_SIZE(I30_NAME) && + !memcmp(attr_name(attr), I30_NAME, + sizeof(I30_NAME)) && + attr->non_res) { + run = &ni->dir.bitmap_run; + break; + } + goto next_attr; + + case ATTR_REPARSE: + if (attr->name_len) + goto next_attr; + + rp_fa = ni_parse_reparse(ni, attr, &rp); + switch (rp_fa) { + case REPARSE_LINK: + if (!attr->non_res) { + inode->i_size = rsize; + inode_set_bytes(inode, rsize); + t32 = asize; + } else { + inode->i_size = + le64_to_cpu(attr->nres.data_size); + t32 = le16_to_cpu(attr->nres.run_off); + } + + /* Looks like normal symlink */ + ni->i_valid = inode->i_size; + + /* Clear directory bit */ + if (ni->ni_flags & NI_FLAG_DIR) { + indx_clear(&ni->dir); + memset(&ni->dir, 0, sizeof(ni->dir)); + ni->ni_flags &= ~NI_FLAG_DIR; + } else { + run_close(&ni->file.run); + } + mode = S_IFLNK | 0777; + is_dir = false; + if (attr->non_res) { + run = &ni->file.run; + goto attr_unpack_run; // double break + } + break; + + case REPARSE_COMPRESSED: + break; + + case REPARSE_DEDUPLICATED: + break; + } + goto next_attr; + + case ATTR_EA_INFO: + if (!attr->name_len && + resident_data_ex(attr, sizeof(struct EA_INFO))) { + ni->ni_flags |= NI_FLAG_EA; + /* + * ntfs_get_wsl_perm updates inode->i_uid, inode->i_gid, inode->i_mode + */ + inode->i_mode = mode; + ntfs_get_wsl_perm(inode); + mode = inode->i_mode; + } + goto next_attr; + + default: + goto next_attr; + } + +attr_unpack_run: + roff = le16_to_cpu(attr->nres.run_off); + + t64 = le64_to_cpu(attr->nres.svcn); + err = run_unpack_ex(run, sbi, ino, t64, le64_to_cpu(attr->nres.evcn), + t64, Add2Ptr(attr, roff), asize - roff); + if (err < 0) + goto out; + err = 0; + goto next_attr; + +end_enum: + + if (!std5) + goto out; + + if (!is_match && name) { + /* reuse rec as buffer for ascii name */ + err = -ENOENT; + goto out; + } + + if (std5->fa & FILE_ATTRIBUTE_READONLY) + mode &= ~0222; + + if (!names) { + err = -EINVAL; + goto out; + } + + set_nlink(inode, names); + + if (S_ISDIR(mode)) { + ni->std_fa |= FILE_ATTRIBUTE_DIRECTORY; + + /* + * dot and dot-dot should be included in count but was not + * included in enumeration. + * Usually a hard links to directories are disabled + */ + inode->i_op = &ntfs_dir_inode_operations; + inode->i_fop = &ntfs_dir_operations; + ni->i_valid = 0; + } else if (S_ISLNK(mode)) { + ni->std_fa &= ~FILE_ATTRIBUTE_DIRECTORY; + inode->i_op = &ntfs_link_inode_operations; + inode->i_fop = NULL; + inode_nohighmem(inode); // ?? + } else if (S_ISREG(mode)) { + ni->std_fa &= ~FILE_ATTRIBUTE_DIRECTORY; + inode->i_op = &ntfs_file_inode_operations; + inode->i_fop = &ntfs_file_operations; + inode->i_mapping->a_ops = + is_compressed(ni) ? &ntfs_aops_cmpr : &ntfs_aops; + if (ino != MFT_REC_MFT) + init_rwsem(&ni->file.run_lock); + } else if (S_ISCHR(mode) || S_ISBLK(mode) || S_ISFIFO(mode) || + S_ISSOCK(mode)) { + inode->i_op = &ntfs_special_inode_operations; + init_special_inode(inode, mode, inode->i_rdev); + } else if (fname && fname->home.low == cpu_to_le32(MFT_REC_EXTEND) && + fname->home.seq == cpu_to_le16(MFT_REC_EXTEND)) { + /* Records in $Extend are not a files or general directories */ + } else { + err = -EINVAL; + goto out; + } + + if ((sbi->options.sys_immutable && + (std5->fa & FILE_ATTRIBUTE_SYSTEM)) && + !S_ISFIFO(mode) && !S_ISSOCK(mode) && !S_ISLNK(mode)) { + inode->i_flags |= S_IMMUTABLE; + } else { + inode->i_flags &= ~S_IMMUTABLE; + } + + inode->i_mode = mode; + if (!(ni->ni_flags & NI_FLAG_EA)) { + /* if no xattr then no security (stored in xattr) */ + inode->i_flags |= S_NOSEC; + } + +Ok: + if (ino == MFT_REC_MFT && !sb->s_root) + sbi->mft.ni = NULL; + + unlock_new_inode(inode); + + return inode; + +out: + if (ino == MFT_REC_MFT && !sb->s_root) + sbi->mft.ni = NULL; + + iget_failed(inode); + return ERR_PTR(err); +} + +/* returns 1 if match */ +static int ntfs_test_inode(struct inode *inode, void *data) +{ + struct MFT_REF *ref = data; + + return ino_get(ref) == inode->i_ino; +} + +static int ntfs_set_inode(struct inode *inode, void *data) +{ + const struct MFT_REF *ref = data; + + inode->i_ino = ino_get(ref); + return 0; +} + +struct inode *ntfs_iget5(struct super_block *sb, const struct MFT_REF *ref, + const struct cpu_str *name) +{ + struct inode *inode; + + inode = iget5_locked(sb, ino_get(ref), ntfs_test_inode, ntfs_set_inode, + (void *)ref); + if (unlikely(!inode)) + return ERR_PTR(-ENOMEM); + + /* If this is a freshly allocated inode, need to read it now. */ + if (inode->i_state & I_NEW) + inode = ntfs_read_mft(inode, name, ref); + else if (ref->seq != ntfs_i(inode)->mi.mrec->seq) { + /* inode overlaps? */ + make_bad_inode(inode); + } + + return inode; +} + +enum get_block_ctx { + GET_BLOCK_GENERAL = 0, + GET_BLOCK_WRITE_BEGIN = 1, + GET_BLOCK_DIRECT_IO_R = 2, + GET_BLOCK_DIRECT_IO_W = 3, + GET_BLOCK_BMAP = 4, +}; + +static noinline int ntfs_get_block_vbo(struct inode *inode, u64 vbo, + struct buffer_head *bh, int create, + enum get_block_ctx ctx) +{ + struct super_block *sb = inode->i_sb; + struct ntfs_sb_info *sbi = sb->s_fs_info; + struct ntfs_inode *ni = ntfs_i(inode); + struct page *page = bh->b_page; + u8 cluster_bits = sbi->cluster_bits; + u32 block_size = sb->s_blocksize; + u64 bytes, lbo, valid; + u32 off; + int err; + CLST vcn, lcn, len; + bool new; + + /*clear previous state*/ + clear_buffer_new(bh); + clear_buffer_uptodate(bh); + + /* direct write uses 'create=0'*/ + if (!create && vbo >= ni->i_valid) { + /* out of valid */ + return 0; + } + + if (vbo >= inode->i_size) { + /* out of size */ + return 0; + } + + if (is_resident(ni)) { + ni_lock(ni); + err = attr_data_read_resident(ni, page); + ni_unlock(ni); + + if (!err) + set_buffer_uptodate(bh); + bh->b_size = block_size; + return err; + } + + vcn = vbo >> cluster_bits; + off = vbo & sbi->cluster_mask; + new = false; + + err = attr_data_get_block(ni, vcn, 1, &lcn, &len, create ? &new : NULL); + if (err) + goto out; + + if (!len) + return 0; + + bytes = ((u64)len << cluster_bits) - off; + + if (lcn == SPARSE_LCN) { + if (!create) { + if (bh->b_size > bytes) + bh->b_size = bytes; + return 0; + } + WARN_ON(1); + } + + if (new) { + set_buffer_new(bh); + if ((len << cluster_bits) > block_size) + ntfs_sparse_cluster(inode, page, vcn, len); + } + + lbo = ((u64)lcn << cluster_bits) + off; + + set_buffer_mapped(bh); + bh->b_bdev = sb->s_bdev; + bh->b_blocknr = lbo >> sb->s_blocksize_bits; + + valid = ni->i_valid; + + if (ctx == GET_BLOCK_DIRECT_IO_W) { + /*ntfs_direct_IO will update ni->i_valid */ + if (vbo >= valid) + set_buffer_new(bh); + } else if (create) { + /*normal write*/ + if (bytes > bh->b_size) + bytes = bh->b_size; + + if (vbo >= valid) + set_buffer_new(bh); + + if (vbo + bytes > valid) { + ni->i_valid = vbo + bytes; + mark_inode_dirty(inode); + } + } else if (vbo >= valid) { + /* read out of valid data*/ + /* should never be here 'cause already checked */ + clear_buffer_mapped(bh); + } else if (vbo + bytes <= valid) { + /* normal read */ + } else if (vbo + block_size <= valid) { + /* normal short read */ + bytes = block_size; + } else { + /* + * read across valid size: vbo < valid && valid < vbo + block_size + */ + bytes = block_size; + + if (page) { + u32 voff = valid - vbo; + + bh->b_size = block_size; + off = vbo & (PAGE_SIZE - 1); + set_bh_page(bh, page, off); + ll_rw_block(REQ_OP_READ, 0, 1, &bh); + wait_on_buffer(bh); + if (!buffer_uptodate(bh)) { + err = -EIO; + goto out; + } + zero_user_segment(page, off + voff, off + block_size); + } + } + + if (bh->b_size > bytes) + bh->b_size = bytes; + +#ifndef __LP64__ + if (ctx == GET_BLOCK_DIRECT_IO_W || ctx == GET_BLOCK_DIRECT_IO_R) { + static_assert(sizeof(size_t) < sizeof(loff_t)); + if (bytes > 0x40000000u) + bh->b_size = 0x40000000u; + } +#endif + + return 0; + +out: + return err; +} + +int ntfs_get_block(struct inode *inode, sector_t vbn, + struct buffer_head *bh_result, int create) +{ + return ntfs_get_block_vbo(inode, (u64)vbn << inode->i_blkbits, + bh_result, create, GET_BLOCK_GENERAL); +} + +static int ntfs_get_block_bmap(struct inode *inode, sector_t vsn, + struct buffer_head *bh_result, int create) +{ + return ntfs_get_block_vbo(inode, + (u64)vsn << inode->i_sb->s_blocksize_bits, + bh_result, create, GET_BLOCK_BMAP); +} + +static sector_t ntfs_bmap(struct address_space *mapping, sector_t block) +{ + return generic_block_bmap(mapping, block, ntfs_get_block_bmap); +} + +static int ntfs_readpage(struct file *file, struct page *page) +{ + int err; + struct address_space *mapping = page->mapping; + struct inode *inode = mapping->host; + struct ntfs_inode *ni = ntfs_i(inode); + + if (is_resident(ni)) { + ni_lock(ni); + err = attr_data_read_resident(ni, page); + ni_unlock(ni); + if (err != E_NTFS_NONRESIDENT) { + unlock_page(page); + return err; + } + } + + if (is_compressed(ni)) { + ni_lock(ni); + err = ni_readpage_cmpr(ni, page); + ni_unlock(ni); + return err; + } + + /* normal + sparse files */ + return mpage_readpage(page, ntfs_get_block); +} + +static void ntfs_readahead(struct readahead_control *rac) +{ + struct address_space *mapping = rac->mapping; + struct inode *inode = mapping->host; + struct ntfs_inode *ni = ntfs_i(inode); + u64 valid; + loff_t pos; + + if (is_resident(ni)) { + /* no readahead for resident */ + return; + } + + if (is_compressed(ni)) { + /* no readahead for compressed */ + return; + } + + valid = ni->i_valid; + pos = readahead_pos(rac); + + if (valid < i_size_read(inode) && pos <= valid && + valid < pos + readahead_length(rac)) { + /* range cross 'valid'. read it page by page */ + return; + } + + mpage_readahead(rac, ntfs_get_block); +} + +static int ntfs_get_block_direct_IO_R(struct inode *inode, sector_t iblock, + struct buffer_head *bh_result, int create) +{ + return ntfs_get_block_vbo(inode, (u64)iblock << inode->i_blkbits, + bh_result, create, GET_BLOCK_DIRECT_IO_R); +} + +static int ntfs_get_block_direct_IO_W(struct inode *inode, sector_t iblock, + struct buffer_head *bh_result, int create) +{ + return ntfs_get_block_vbo(inode, (u64)iblock << inode->i_blkbits, + bh_result, create, GET_BLOCK_DIRECT_IO_W); +} + +static ssize_t ntfs_direct_IO(struct kiocb *iocb, struct iov_iter *iter) +{ + struct file *file = iocb->ki_filp; + struct address_space *mapping = file->f_mapping; + struct inode *inode = mapping->host; + struct ntfs_inode *ni = ntfs_i(inode); + loff_t vbo = iocb->ki_pos; + loff_t end; + int wr = iov_iter_rw(iter) & WRITE; + loff_t valid; + ssize_t ret; + + if (is_resident(ni)) { + /*switch to buffered write*/ + ret = 0; + goto out; + } + + ret = blockdev_direct_IO(iocb, inode, iter, + wr ? ntfs_get_block_direct_IO_W + : ntfs_get_block_direct_IO_R); + + if (ret <= 0) + goto out; + + end = vbo + ret; + valid = ni->i_valid; + if (wr) { + if (end > valid && !S_ISBLK(inode->i_mode)) { + ni->i_valid = end; + mark_inode_dirty(inode); + } + } else if (vbo < valid && valid < end) { + /* fix page */ + iov_iter_revert(iter, end - valid); + iov_iter_zero(end - valid, iter); + } + +out: + return ret; +} + +int ntfs_set_size(struct inode *inode, u64 new_size) +{ + struct super_block *sb = inode->i_sb; + struct ntfs_sb_info *sbi = sb->s_fs_info; + struct ntfs_inode *ni = ntfs_i(inode); + int err; + + /* Check for maximum file size */ + if (is_sparsed(ni) || is_compressed(ni)) { + if (new_size > sbi->maxbytes_sparse) { + err = -EFBIG; + goto out; + } + } else if (new_size > sbi->maxbytes) { + err = -EFBIG; + goto out; + } + + ni_lock(ni); + down_write(&ni->file.run_lock); + + err = attr_set_size(ni, ATTR_DATA, NULL, 0, &ni->file.run, new_size, + &ni->i_valid, true, NULL); + + up_write(&ni->file.run_lock); + ni_unlock(ni); + + mark_inode_dirty(inode); + +out: + return err; +} + +static int ntfs_writepage(struct page *page, struct writeback_control *wbc) +{ + struct address_space *mapping = page->mapping; + struct inode *inode = mapping->host; + struct ntfs_inode *ni = ntfs_i(inode); + int err; + + if (is_resident(ni)) { + ni_lock(ni); + err = attr_data_write_resident(ni, page); + ni_unlock(ni); + if (err != E_NTFS_NONRESIDENT) { + unlock_page(page); + return err; + } + } + + return block_write_full_page(page, ntfs_get_block, wbc); +} + +static int ntfs_writepages(struct address_space *mapping, + struct writeback_control *wbc) +{ + struct inode *inode = mapping->host; + struct ntfs_inode *ni = ntfs_i(inode); + /* redirect call to 'ntfs_writepage' for resident files*/ + get_block_t *get_block = is_resident(ni) ? NULL : &ntfs_get_block; + + return mpage_writepages(mapping, wbc, get_block); +} + +static int ntfs_get_block_write_begin(struct inode *inode, sector_t vbn, + struct buffer_head *bh_result, int create) +{ + return ntfs_get_block_vbo(inode, (u64)vbn << inode->i_blkbits, + bh_result, create, GET_BLOCK_WRITE_BEGIN); +} + +static int ntfs_write_begin(struct file *file, struct address_space *mapping, + loff_t pos, u32 len, u32 flags, struct page **pagep, + void **fsdata) +{ + int err; + struct inode *inode = mapping->host; + struct ntfs_inode *ni = ntfs_i(inode); + + *pagep = NULL; + if (is_resident(ni)) { + struct page *page = grab_cache_page_write_begin( + mapping, pos >> PAGE_SHIFT, flags); + + if (!page) { + err = -ENOMEM; + goto out; + } + + ni_lock(ni); + err = attr_data_read_resident(ni, page); + ni_unlock(ni); + + if (!err) { + *pagep = page; + goto out; + } + unlock_page(page); + put_page(page); + + if (err != E_NTFS_NONRESIDENT) + goto out; + } + + err = block_write_begin(mapping, pos, len, flags, pagep, + ntfs_get_block_write_begin); + +out: + return err; +} + +/* address_space_operations::write_end */ +static int ntfs_write_end(struct file *file, struct address_space *mapping, + loff_t pos, u32 len, u32 copied, struct page *page, + void *fsdata) + +{ + struct inode *inode = mapping->host; + struct ntfs_inode *ni = ntfs_i(inode); + u64 valid = ni->i_valid; + bool dirty = false; + int err; + + if (is_resident(ni)) { + ni_lock(ni); + err = attr_data_write_resident(ni, page); + ni_unlock(ni); + if (!err) { + dirty = true; + /* clear any buffers in page*/ + if (page_has_buffers(page)) { + struct buffer_head *head, *bh; + + bh = head = page_buffers(page); + do { + clear_buffer_dirty(bh); + clear_buffer_mapped(bh); + set_buffer_uptodate(bh); + } while (head != (bh = bh->b_this_page)); + } + SetPageUptodate(page); + err = copied; + } + unlock_page(page); + put_page(page); + } else { + err = generic_write_end(file, mapping, pos, len, copied, page, + fsdata); + } + + if (err >= 0) { + if (!(ni->std_fa & FILE_ATTRIBUTE_ARCHIVE)) { + inode->i_ctime = inode->i_mtime = current_time(inode); + ni->std_fa |= FILE_ATTRIBUTE_ARCHIVE; + dirty = true; + } + + if (valid != ni->i_valid) { + /* ni->i_valid is changed in ntfs_get_block_vbo */ + dirty = true; + } + + if (dirty) + mark_inode_dirty(inode); + } + + return err; +} + +int reset_log_file(struct inode *inode) +{ + int err; + loff_t pos = 0; + u32 log_size = inode->i_size; + struct address_space *mapping = inode->i_mapping; + + for (;;) { + u32 len; + void *kaddr; + struct page *page; + + len = pos + PAGE_SIZE > log_size ? (log_size - pos) : PAGE_SIZE; + + err = block_write_begin(mapping, pos, len, 0, &page, + ntfs_get_block_write_begin); + if (err) + goto out; + + kaddr = kmap_atomic(page); + memset(kaddr, -1, len); + kunmap_atomic(kaddr); + flush_dcache_page(page); + + err = block_write_end(NULL, mapping, pos, len, len, page, NULL); + if (err < 0) + goto out; + pos += len; + + if (pos >= log_size) + break; + balance_dirty_pages_ratelimited(mapping); + } +out: + mark_inode_dirty_sync(inode); + + return err; +} + +int ntfs3_write_inode(struct inode *inode, struct writeback_control *wbc) +{ + return _ni_write_inode(inode, wbc->sync_mode == WB_SYNC_ALL); +} + +int ntfs_sync_inode(struct inode *inode) +{ + return _ni_write_inode(inode, 1); +} + +/* + * helper function for ntfs_flush_inodes. This writes both the inode + * and the file data blocks, waiting for in flight data blocks before + * the start of the call. It does not wait for any io started + * during the call + */ +static int writeback_inode(struct inode *inode) +{ + int ret = sync_inode_metadata(inode, 0); + + if (!ret) + ret = filemap_fdatawrite(inode->i_mapping); + return ret; +} + +/* + * write data and metadata corresponding to i1 and i2. The io is + * started but we do not wait for any of it to finish. + * + * filemap_flush is used for the block device, so if there is a dirty + * page for a block already in flight, we will not wait and start the + * io over again + */ +int ntfs_flush_inodes(struct super_block *sb, struct inode *i1, + struct inode *i2) +{ + int ret = 0; + + if (i1) + ret = writeback_inode(i1); + if (!ret && i2) + ret = writeback_inode(i2); + if (!ret) + ret = filemap_flush(sb->s_bdev->bd_inode->i_mapping); + return ret; +} + +int inode_write_data(struct inode *inode, const void *data, size_t bytes) +{ + pgoff_t idx; + + /* Write non resident data */ + for (idx = 0; bytes; idx++) { + size_t op = bytes > PAGE_SIZE ? PAGE_SIZE : bytes; + struct page *page = ntfs_map_page(inode->i_mapping, idx); + + if (IS_ERR(page)) + return PTR_ERR(page); + + lock_page(page); + WARN_ON(!PageUptodate(page)); + ClearPageUptodate(page); + + memcpy(page_address(page), data, op); + + flush_dcache_page(page); + SetPageUptodate(page); + unlock_page(page); + + ntfs_unmap_page(page); + + bytes -= op; + data = Add2Ptr(data, PAGE_SIZE); + } + return 0; +} + +/* + * number of bytes to for REPARSE_DATA_BUFFER(IO_REPARSE_TAG_SYMLINK) + * for unicode string of 'uni_len' length + */ +static inline u32 ntfs_reparse_bytes(u32 uni_len) +{ + /* header + unicode string + decorated unicode string */ + return sizeof(short) * (2 * uni_len + 4) + + offsetof(struct REPARSE_DATA_BUFFER, + SymbolicLinkReparseBuffer.PathBuffer); +} + +static struct REPARSE_DATA_BUFFER * +ntfs_create_reparse_buffer(struct ntfs_sb_info *sbi, const char *symname, + u32 size, u16 *nsize) +{ + int i, err; + struct REPARSE_DATA_BUFFER *rp; + __le16 *rp_name; + typeof(rp->SymbolicLinkReparseBuffer) *rs; + + rp = ntfs_zalloc(ntfs_reparse_bytes(2 * size + 2)); + if (!rp) + return ERR_PTR(-ENOMEM); + + rs = &rp->SymbolicLinkReparseBuffer; + rp_name = rs->PathBuffer; + + /* Convert link name to utf16 */ + err = ntfs_nls_to_utf16(sbi, symname, size, + (struct cpu_str *)(rp_name - 1), 2 * size, + UTF16_LITTLE_ENDIAN); + if (err < 0) + goto out; + + /* err = the length of unicode name of symlink */ + *nsize = ntfs_reparse_bytes(err); + + if (*nsize > sbi->reparse.max_size) { + err = -EFBIG; + goto out; + } + + /* translate linux '/' into windows '\' */ + for (i = 0; i < err; i++) { + if (rp_name[i] == cpu_to_le16('/')) + rp_name[i] = cpu_to_le16('\\'); + } + + rp->ReparseTag = IO_REPARSE_TAG_SYMLINK; + rp->ReparseDataLength = + cpu_to_le16(*nsize - offsetof(struct REPARSE_DATA_BUFFER, + SymbolicLinkReparseBuffer)); + + /* PrintName + SubstituteName */ + rs->SubstituteNameOffset = cpu_to_le16(sizeof(short) * err); + rs->SubstituteNameLength = cpu_to_le16(sizeof(short) * err + 8); + rs->PrintNameLength = rs->SubstituteNameOffset; + + /* + * TODO: use relative path if possible to allow windows to parse this path + * 0-absolute path 1- relative path (SYMLINK_FLAG_RELATIVE) + */ + rs->Flags = 0; + + memmove(rp_name + err + 4, rp_name, sizeof(short) * err); + + /* decorate SubstituteName */ + rp_name += err; + rp_name[0] = cpu_to_le16('\\'); + rp_name[1] = cpu_to_le16('?'); + rp_name[2] = cpu_to_le16('?'); + rp_name[3] = cpu_to_le16('\\'); + + return rp; +out: + ntfs_free(rp); + return ERR_PTR(err); +} + +struct inode *ntfs_create_inode(struct user_namespace *mnt_userns, + struct inode *dir, struct dentry *dentry, + const struct cpu_str *uni, umode_t mode, + dev_t dev, const char *symname, u32 size, + struct ntfs_fnd *fnd) +{ + int err; + struct super_block *sb = dir->i_sb; + struct ntfs_sb_info *sbi = sb->s_fs_info; + const struct qstr *name = &dentry->d_name; + CLST ino = 0; + struct ntfs_inode *dir_ni = ntfs_i(dir); + struct ntfs_inode *ni = NULL; + struct inode *inode = NULL; + struct ATTRIB *attr; + struct ATTR_STD_INFO5 *std5; + struct ATTR_FILE_NAME *fname; + struct MFT_REC *rec; + u32 asize, dsize, sd_size; + enum FILE_ATTRIBUTE fa; + __le32 security_id = SECURITY_ID_INVALID; + CLST vcn; + const void *sd; + u16 t16, nsize = 0, aid = 0; + struct INDEX_ROOT *root, *dir_root; + struct NTFS_DE *e, *new_de = NULL; + struct REPARSE_DATA_BUFFER *rp = NULL; + bool rp_inserted = false; + + dir_root = indx_get_root(&dir_ni->dir, dir_ni, NULL, NULL); + if (!dir_root) + return ERR_PTR(-EINVAL); + + if (S_ISDIR(mode)) { + /* use parent's directory attributes */ + fa = dir_ni->std_fa | FILE_ATTRIBUTE_DIRECTORY | + FILE_ATTRIBUTE_ARCHIVE; + /* + * By default child directory inherits parent attributes + * root directory is hidden + system + * Make an exception for children in root + */ + if (dir->i_ino == MFT_REC_ROOT) + fa &= ~(FILE_ATTRIBUTE_HIDDEN | FILE_ATTRIBUTE_SYSTEM); + } else if (S_ISLNK(mode)) { + /* It is good idea that link should be the same type (file/dir) as target */ + fa = FILE_ATTRIBUTE_REPARSE_POINT; + + /* + * linux: there are dir/file/symlink and so on + * NTFS: symlinks are "dir + reparse" or "file + reparse" + * It is good idea to create: + * dir + reparse if 'symname' points to directory + * or + * file + reparse if 'symname' points to file + * Unfortunately kern_path hangs if symname contains 'dir' + */ + + /* + * struct path path; + * + * if (!kern_path(symname, LOOKUP_FOLLOW, &path)){ + * struct inode *target = d_inode(path.dentry); + * + * if (S_ISDIR(target->i_mode)) + * fa |= FILE_ATTRIBUTE_DIRECTORY; + * // if ( target->i_sb == sb ){ + * // use relative path? + * // } + * path_put(&path); + * } + */ + } else if (S_ISREG(mode)) { + if (sbi->options.sparse) { + /* sparsed regular file, cause option 'sparse' */ + fa = FILE_ATTRIBUTE_SPARSE_FILE | + FILE_ATTRIBUTE_ARCHIVE; + } else if (dir_ni->std_fa & FILE_ATTRIBUTE_COMPRESSED) { + /* compressed regular file, if parent is compressed */ + fa = FILE_ATTRIBUTE_COMPRESSED | FILE_ATTRIBUTE_ARCHIVE; + } else { + /* regular file, default attributes */ + fa = FILE_ATTRIBUTE_ARCHIVE; + } + } else { + fa = FILE_ATTRIBUTE_ARCHIVE; + } + + if (!(mode & 0222)) + fa |= FILE_ATTRIBUTE_READONLY; + + /* allocate PATH_MAX bytes */ + new_de = __getname(); + if (!new_de) { + err = -ENOMEM; + goto out1; + } + + /*mark rw ntfs as dirty. it will be cleared at umount*/ + ntfs_set_state(sbi, NTFS_DIRTY_DIRTY); + + /* Step 1: allocate and fill new mft record */ + err = ntfs_look_free_mft(sbi, &ino, false, NULL, NULL); + if (err) + goto out2; + + ni = ntfs_new_inode(sbi, ino, fa & FILE_ATTRIBUTE_DIRECTORY); + if (IS_ERR(ni)) { + err = PTR_ERR(ni); + ni = NULL; + goto out3; + } + inode = &ni->vfs_inode; + inode_init_owner(mnt_userns, inode, dir, mode); + + inode->i_atime = inode->i_mtime = inode->i_ctime = ni->i_crtime = + current_time(inode); + + rec = ni->mi.mrec; + rec->hard_links = cpu_to_le16(1); + attr = Add2Ptr(rec, le16_to_cpu(rec->attr_off)); + + /* Get default security id */ + sd = s_default_security; + sd_size = sizeof(s_default_security); + + if (is_ntfs3(sbi)) { + security_id = dir_ni->std_security_id; + if (le32_to_cpu(security_id) < SECURITY_ID_FIRST) { + security_id = sbi->security.def_security_id; + + if (security_id == SECURITY_ID_INVALID && + !ntfs_insert_security(sbi, sd, sd_size, + &security_id, NULL)) + sbi->security.def_security_id = security_id; + } + } + + /* Insert standard info */ + std5 = Add2Ptr(attr, SIZEOF_RESIDENT); + + if (security_id == SECURITY_ID_INVALID) { + dsize = sizeof(struct ATTR_STD_INFO); + } else { + dsize = sizeof(struct ATTR_STD_INFO5); + std5->security_id = security_id; + ni->std_security_id = security_id; + } + asize = SIZEOF_RESIDENT + dsize; + + attr->type = ATTR_STD; + attr->size = cpu_to_le32(asize); + attr->id = cpu_to_le16(aid++); + attr->res.data_off = SIZEOF_RESIDENT_LE; + attr->res.data_size = cpu_to_le32(dsize); + + std5->cr_time = std5->m_time = std5->c_time = std5->a_time = + kernel2nt(&inode->i_atime); + + ni->std_fa = fa; + std5->fa = fa; + + attr = Add2Ptr(attr, asize); + + /* Insert file name */ + err = fill_name_de(sbi, new_de, name, uni); + if (err) + goto out4; + + mi_get_ref(&ni->mi, &new_de->ref); + + fname = (struct ATTR_FILE_NAME *)(new_de + 1); + mi_get_ref(&dir_ni->mi, &fname->home); + fname->dup.cr_time = fname->dup.m_time = fname->dup.c_time = + fname->dup.a_time = std5->cr_time; + fname->dup.alloc_size = fname->dup.data_size = 0; + fname->dup.fa = std5->fa; + fname->dup.ea_size = fname->dup.reparse = 0; + + dsize = le16_to_cpu(new_de->key_size); + asize = QuadAlign(SIZEOF_RESIDENT + dsize); + + attr->type = ATTR_NAME; + attr->size = cpu_to_le32(asize); + attr->res.data_off = SIZEOF_RESIDENT_LE; + attr->res.flags = RESIDENT_FLAG_INDEXED; + attr->id = cpu_to_le16(aid++); + attr->res.data_size = cpu_to_le32(dsize); + memcpy(Add2Ptr(attr, SIZEOF_RESIDENT), fname, dsize); + + attr = Add2Ptr(attr, asize); + + if (security_id == SECURITY_ID_INVALID) { + /* Insert security attribute */ + asize = SIZEOF_RESIDENT + QuadAlign(sd_size); + + attr->type = ATTR_SECURE; + attr->size = cpu_to_le32(asize); + attr->id = cpu_to_le16(aid++); + attr->res.data_off = SIZEOF_RESIDENT_LE; + attr->res.data_size = cpu_to_le32(sd_size); + memcpy(Add2Ptr(attr, SIZEOF_RESIDENT), sd, sd_size); + + attr = Add2Ptr(attr, asize); + } + + if (fa & FILE_ATTRIBUTE_DIRECTORY) { + /* + * regular directory or symlink to directory + * Create root attribute + */ + dsize = sizeof(struct INDEX_ROOT) + sizeof(struct NTFS_DE); + asize = sizeof(I30_NAME) + SIZEOF_RESIDENT + dsize; + + attr->type = ATTR_ROOT; + attr->size = cpu_to_le32(asize); + attr->id = cpu_to_le16(aid++); + + attr->name_len = ARRAY_SIZE(I30_NAME); + attr->name_off = SIZEOF_RESIDENT_LE; + attr->res.data_off = + cpu_to_le16(sizeof(I30_NAME) + SIZEOF_RESIDENT); + attr->res.data_size = cpu_to_le32(dsize); + memcpy(Add2Ptr(attr, SIZEOF_RESIDENT), I30_NAME, + sizeof(I30_NAME)); + + root = Add2Ptr(attr, sizeof(I30_NAME) + SIZEOF_RESIDENT); + memcpy(root, dir_root, offsetof(struct INDEX_ROOT, ihdr)); + root->ihdr.de_off = + cpu_to_le32(sizeof(struct INDEX_HDR)); // 0x10 + root->ihdr.used = cpu_to_le32(sizeof(struct INDEX_HDR) + + sizeof(struct NTFS_DE)); + root->ihdr.total = root->ihdr.used; + + e = Add2Ptr(root, sizeof(struct INDEX_ROOT)); + e->size = cpu_to_le16(sizeof(struct NTFS_DE)); + e->flags = NTFS_IE_LAST; + } else if (S_ISLNK(mode)) { + /* + * symlink to file + * Create empty resident data attribute + */ + asize = SIZEOF_RESIDENT; + + /* insert empty ATTR_DATA */ + attr->type = ATTR_DATA; + attr->size = cpu_to_le32(SIZEOF_RESIDENT); + attr->id = cpu_to_le16(aid++); + attr->name_off = SIZEOF_RESIDENT_LE; + attr->res.data_off = SIZEOF_RESIDENT_LE; + } else { + /* + * regular file or node + */ + attr->type = ATTR_DATA; + attr->id = cpu_to_le16(aid++); + + if (S_ISREG(mode)) { + /* Create empty non resident data attribute */ + attr->non_res = 1; + attr->nres.evcn = cpu_to_le64(-1ll); + if (fa & FILE_ATTRIBUTE_SPARSE_FILE) { + attr->size = + cpu_to_le32(SIZEOF_NONRESIDENT_EX + 8); + attr->name_off = SIZEOF_NONRESIDENT_EX_LE; + attr->flags = ATTR_FLAG_SPARSED; + asize = SIZEOF_NONRESIDENT_EX + 8; + } else if (fa & FILE_ATTRIBUTE_COMPRESSED) { + attr->size = + cpu_to_le32(SIZEOF_NONRESIDENT_EX + 8); + attr->name_off = SIZEOF_NONRESIDENT_EX_LE; + attr->flags = ATTR_FLAG_COMPRESSED; + attr->nres.c_unit = COMPRESSION_UNIT; + asize = SIZEOF_NONRESIDENT_EX + 8; + } else { + attr->size = + cpu_to_le32(SIZEOF_NONRESIDENT + 8); + attr->name_off = SIZEOF_NONRESIDENT_LE; + asize = SIZEOF_NONRESIDENT + 8; + } + attr->nres.run_off = attr->name_off; + } else { + /* Create empty resident data attribute */ + attr->size = cpu_to_le32(SIZEOF_RESIDENT); + attr->name_off = SIZEOF_RESIDENT_LE; + if (fa & FILE_ATTRIBUTE_SPARSE_FILE) + attr->flags = ATTR_FLAG_SPARSED; + else if (fa & FILE_ATTRIBUTE_COMPRESSED) + attr->flags = ATTR_FLAG_COMPRESSED; + attr->res.data_off = SIZEOF_RESIDENT_LE; + asize = SIZEOF_RESIDENT; + ni->ni_flags |= NI_FLAG_RESIDENT; + } + } + + if (S_ISDIR(mode)) { + ni->ni_flags |= NI_FLAG_DIR; + err = indx_init(&ni->dir, sbi, attr, INDEX_MUTEX_I30); + if (err) + goto out4; + } else if (S_ISLNK(mode)) { + rp = ntfs_create_reparse_buffer(sbi, symname, size, &nsize); + + if (IS_ERR(rp)) { + err = PTR_ERR(rp); + rp = NULL; + goto out4; + } + + /* + * Insert ATTR_REPARSE + */ + attr = Add2Ptr(attr, asize); + attr->type = ATTR_REPARSE; + attr->id = cpu_to_le16(aid++); + + /* resident or non resident? */ + asize = QuadAlign(SIZEOF_RESIDENT + nsize); + t16 = PtrOffset(rec, attr); + + if (asize + t16 + 8 > sbi->record_size) { + CLST alen; + CLST clst = bytes_to_cluster(sbi, nsize); + + /* bytes per runs */ + t16 = sbi->record_size - t16 - SIZEOF_NONRESIDENT; + + attr->non_res = 1; + attr->nres.evcn = cpu_to_le64(clst - 1); + attr->name_off = SIZEOF_NONRESIDENT_LE; + attr->nres.run_off = attr->name_off; + attr->nres.data_size = cpu_to_le64(nsize); + attr->nres.valid_size = attr->nres.data_size; + attr->nres.alloc_size = + cpu_to_le64(ntfs_up_cluster(sbi, nsize)); + + err = attr_allocate_clusters(sbi, &ni->file.run, 0, 0, + clst, NULL, 0, &alen, 0, + NULL); + if (err) + goto out5; + + err = run_pack(&ni->file.run, 0, clst, + Add2Ptr(attr, SIZEOF_NONRESIDENT), t16, + &vcn); + if (err < 0) + goto out5; + + if (vcn != clst) { + err = -EINVAL; + goto out5; + } + + asize = SIZEOF_NONRESIDENT + QuadAlign(err); + inode->i_size = nsize; + } else { + attr->res.data_off = SIZEOF_RESIDENT_LE; + attr->res.data_size = cpu_to_le32(nsize); + memcpy(Add2Ptr(attr, SIZEOF_RESIDENT), rp, nsize); + inode->i_size = nsize; + nsize = 0; + } + + attr->size = cpu_to_le32(asize); + + err = ntfs_insert_reparse(sbi, IO_REPARSE_TAG_SYMLINK, + &new_de->ref); + if (err) + goto out5; + + rp_inserted = true; + } + + attr = Add2Ptr(attr, asize); + attr->type = ATTR_END; + + rec->used = cpu_to_le32(PtrOffset(rec, attr) + 8); + rec->next_attr_id = cpu_to_le16(aid); + + /* Step 2: Add new name in index */ + err = indx_insert_entry(&dir_ni->dir, dir_ni, new_de, sbi, fnd); + if (err) + goto out6; + + /* Update current directory record */ + mark_inode_dirty(dir); + + inode->i_generation = le16_to_cpu(rec->seq); + + dir->i_mtime = dir->i_ctime = inode->i_atime; + + if (S_ISDIR(mode)) { + if (dir->i_mode & S_ISGID) + mode |= S_ISGID; + inode->i_op = &ntfs_dir_inode_operations; + inode->i_fop = &ntfs_dir_operations; + } else if (S_ISLNK(mode)) { + inode->i_op = &ntfs_link_inode_operations; + inode->i_fop = NULL; + inode->i_mapping->a_ops = &ntfs_aops; + } else if (S_ISREG(mode)) { + inode->i_op = &ntfs_file_inode_operations; + inode->i_fop = &ntfs_file_operations; + inode->i_mapping->a_ops = + is_compressed(ni) ? &ntfs_aops_cmpr : &ntfs_aops; + init_rwsem(&ni->file.run_lock); + } else { + inode->i_op = &ntfs_special_inode_operations; + init_special_inode(inode, mode, dev); + } + +#ifdef CONFIG_NTFS3_FS_POSIX_ACL + if (!S_ISLNK(mode) && (sb->s_flags & SB_POSIXACL)) { + err = ntfs_init_acl(mnt_userns, inode, dir); + if (err) + goto out6; + } else +#endif + { + inode->i_flags |= S_NOSEC; + } + + /* Write non resident data */ + if (nsize) { + err = ntfs_sb_write_run(sbi, &ni->file.run, 0, rp, nsize); + if (err) + goto out7; + } + + /* call 'd_instantiate' after inode->i_op is set but before finish_open */ + d_instantiate(dentry, inode); + + ntfs_save_wsl_perm(inode); + mark_inode_dirty(inode); + mark_inode_dirty(dir); + + /* normal exit */ + goto out2; + +out7: + + /* undo 'indx_insert_entry' */ + indx_delete_entry(&dir_ni->dir, dir_ni, new_de + 1, + le16_to_cpu(new_de->key_size), sbi); +out6: + if (rp_inserted) + ntfs_remove_reparse(sbi, IO_REPARSE_TAG_SYMLINK, &new_de->ref); + +out5: + if (S_ISDIR(mode) || run_is_empty(&ni->file.run)) + goto out4; + + run_deallocate(sbi, &ni->file.run, false); + +out4: + clear_rec_inuse(rec); + clear_nlink(inode); + ni->mi.dirty = false; + discard_new_inode(inode); +out3: + ntfs_mark_rec_free(sbi, ino); + +out2: + __putname(new_de); + ntfs_free(rp); + +out1: + if (err) + return ERR_PTR(err); + + unlock_new_inode(inode); + + return inode; +} + +int ntfs_link_inode(struct inode *inode, struct dentry *dentry) +{ + int err; + struct inode *dir = d_inode(dentry->d_parent); + struct ntfs_inode *dir_ni = ntfs_i(dir); + struct ntfs_inode *ni = ntfs_i(inode); + struct super_block *sb = inode->i_sb; + struct ntfs_sb_info *sbi = sb->s_fs_info; + const struct qstr *name = &dentry->d_name; + struct NTFS_DE *new_de = NULL; + struct ATTR_FILE_NAME *fname; + struct ATTRIB *attr; + u16 key_size; + struct INDEX_ROOT *dir_root; + + dir_root = indx_get_root(&dir_ni->dir, dir_ni, NULL, NULL); + if (!dir_root) + return -EINVAL; + + /* allocate PATH_MAX bytes */ + new_de = __getname(); + if (!new_de) + return -ENOMEM; + + /*mark rw ntfs as dirty. it will be cleared at umount*/ + ntfs_set_state(ni->mi.sbi, NTFS_DIRTY_DIRTY); + + // Insert file name + err = fill_name_de(sbi, new_de, name, NULL); + if (err) + goto out; + + key_size = le16_to_cpu(new_de->key_size); + err = ni_insert_resident(ni, key_size, ATTR_NAME, NULL, 0, &attr, NULL); + if (err) + goto out; + + mi_get_ref(&ni->mi, &new_de->ref); + + fname = (struct ATTR_FILE_NAME *)(new_de + 1); + mi_get_ref(&dir_ni->mi, &fname->home); + fname->dup.cr_time = fname->dup.m_time = fname->dup.c_time = + fname->dup.a_time = kernel2nt(&inode->i_ctime); + fname->dup.alloc_size = fname->dup.data_size = 0; + fname->dup.fa = ni->std_fa; + fname->dup.ea_size = fname->dup.reparse = 0; + + memcpy(Add2Ptr(attr, SIZEOF_RESIDENT), fname, key_size); + + err = indx_insert_entry(&dir_ni->dir, dir_ni, new_de, sbi, NULL); + if (err) + goto out; + + le16_add_cpu(&ni->mi.mrec->hard_links, 1); + ni->mi.dirty = true; + +out: + __putname(new_de); + return err; +} + +/* + * ntfs_unlink_inode + * + * inode_operations::unlink + * inode_operations::rmdir + */ +int ntfs_unlink_inode(struct inode *dir, const struct dentry *dentry) +{ + int err; + struct super_block *sb = dir->i_sb; + struct ntfs_sb_info *sbi = sb->s_fs_info; + struct inode *inode = d_inode(dentry); + struct ntfs_inode *ni = ntfs_i(inode); + const struct qstr *name = &dentry->d_name; + struct ntfs_inode *dir_ni = ntfs_i(dir); + struct ntfs_index *indx = &dir_ni->dir; + struct cpu_str *uni = NULL; + struct ATTR_FILE_NAME *fname; + u8 name_type; + struct ATTR_LIST_ENTRY *le; + struct MFT_REF ref; + bool is_dir = S_ISDIR(inode->i_mode); + struct INDEX_ROOT *dir_root; + + dir_root = indx_get_root(indx, dir_ni, NULL, NULL); + if (!dir_root) + return -EINVAL; + + ni_lock(ni); + + if (is_dir && !dir_is_empty(inode)) { + err = -ENOTEMPTY; + goto out1; + } + + if (ntfs_is_meta_file(sbi, inode->i_ino)) { + err = -EINVAL; + goto out1; + } + + /* allocate PATH_MAX bytes */ + uni = __getname(); + if (!uni) { + err = -ENOMEM; + goto out1; + } + + /* Convert input string to unicode */ + err = ntfs_nls_to_utf16(sbi, name->name, name->len, uni, NTFS_NAME_LEN, + UTF16_HOST_ENDIAN); + if (err < 0) + goto out2; + + /*mark rw ntfs as dirty. it will be cleared at umount*/ + ntfs_set_state(sbi, NTFS_DIRTY_DIRTY); + + /* find name in record */ + mi_get_ref(&dir_ni->mi, &ref); + + le = NULL; + fname = ni_fname_name(ni, uni, &ref, &le); + if (!fname) { + err = -ENOENT; + goto out3; + } + + name_type = paired_name(fname->type); + + err = indx_delete_entry(indx, dir_ni, fname, fname_full_size(fname), + sbi); + if (err) + goto out3; + + /* Then remove name from mft */ + ni_remove_attr_le(ni, attr_from_name(fname), le); + + le16_add_cpu(&ni->mi.mrec->hard_links, -1); + ni->mi.dirty = true; + + if (name_type != FILE_NAME_POSIX) { + /* Now we should delete name by type */ + fname = ni_fname_type(ni, name_type, &le); + if (fname) { + err = indx_delete_entry(indx, dir_ni, fname, + fname_full_size(fname), sbi); + if (err) + goto out3; + + ni_remove_attr_le(ni, attr_from_name(fname), le); + + le16_add_cpu(&ni->mi.mrec->hard_links, -1); + } + } +out3: + switch (err) { + case 0: + drop_nlink(inode); + case -ENOTEMPTY: + case -ENOSPC: + case -EROFS: + break; + default: + make_bad_inode(inode); + } + + dir->i_mtime = dir->i_ctime = current_time(dir); + mark_inode_dirty(dir); + inode->i_ctime = dir->i_ctime; + if (inode->i_nlink) + mark_inode_dirty(inode); + +out2: + __putname(uni); +out1: + ni_unlock(ni); + return err; +} + +void ntfs_evict_inode(struct inode *inode) +{ + truncate_inode_pages_final(&inode->i_data); + + if (inode->i_nlink) + _ni_write_inode(inode, inode_needs_sync(inode)); + + invalidate_inode_buffers(inode); + clear_inode(inode); + + ni_clear(ntfs_i(inode)); +} + +static noinline int ntfs_readlink_hlp(struct inode *inode, char *buffer, + int buflen) +{ + int i, err = 0; + struct ntfs_inode *ni = ntfs_i(inode); + struct super_block *sb = inode->i_sb; + struct ntfs_sb_info *sbi = sb->s_fs_info; + u64 i_size = inode->i_size; + u16 nlen = 0; + void *to_free = NULL; + struct REPARSE_DATA_BUFFER *rp; + struct le_str *uni; + struct ATTRIB *attr; + + /* Reparse data present. Try to parse it */ + static_assert(!offsetof(struct REPARSE_DATA_BUFFER, ReparseTag)); + static_assert(sizeof(u32) == sizeof(rp->ReparseTag)); + + *buffer = 0; + + /* Read into temporal buffer */ + if (i_size > sbi->reparse.max_size || i_size <= sizeof(u32)) { + err = -EINVAL; + goto out; + } + + attr = ni_find_attr(ni, NULL, NULL, ATTR_REPARSE, NULL, 0, NULL, NULL); + if (!attr) { + err = -EINVAL; + goto out; + } + + if (!attr->non_res) { + rp = resident_data_ex(attr, i_size); + if (!rp) { + err = -EINVAL; + goto out; + } + } else { + rp = ntfs_malloc(i_size); + if (!rp) { + err = -ENOMEM; + goto out; + } + to_free = rp; + err = ntfs_read_run_nb(sbi, &ni->file.run, 0, rp, i_size, NULL); + if (err) + goto out; + } + + err = -EINVAL; + + /* Microsoft Tag */ + switch (rp->ReparseTag) { + case IO_REPARSE_TAG_MOUNT_POINT: + /* Mount points and junctions */ + /* Can we use 'Rp->MountPointReparseBuffer.PrintNameLength'? */ + if (i_size <= offsetof(struct REPARSE_DATA_BUFFER, + MountPointReparseBuffer.PathBuffer)) + goto out; + uni = Add2Ptr(rp, + offsetof(struct REPARSE_DATA_BUFFER, + MountPointReparseBuffer.PathBuffer) + + le16_to_cpu(rp->MountPointReparseBuffer + .PrintNameOffset) - + 2); + nlen = le16_to_cpu(rp->MountPointReparseBuffer.PrintNameLength); + break; + + case IO_REPARSE_TAG_SYMLINK: + /* FolderSymbolicLink */ + /* Can we use 'Rp->SymbolicLinkReparseBuffer.PrintNameLength'? */ + if (i_size <= offsetof(struct REPARSE_DATA_BUFFER, + SymbolicLinkReparseBuffer.PathBuffer)) + goto out; + uni = Add2Ptr(rp, + offsetof(struct REPARSE_DATA_BUFFER, + SymbolicLinkReparseBuffer.PathBuffer) + + le16_to_cpu(rp->SymbolicLinkReparseBuffer + .PrintNameOffset) - + 2); + nlen = le16_to_cpu( + rp->SymbolicLinkReparseBuffer.PrintNameLength); + break; + + case IO_REPARSE_TAG_CLOUD: + case IO_REPARSE_TAG_CLOUD_1: + case IO_REPARSE_TAG_CLOUD_2: + case IO_REPARSE_TAG_CLOUD_3: + case IO_REPARSE_TAG_CLOUD_4: + case IO_REPARSE_TAG_CLOUD_5: + case IO_REPARSE_TAG_CLOUD_6: + case IO_REPARSE_TAG_CLOUD_7: + case IO_REPARSE_TAG_CLOUD_8: + case IO_REPARSE_TAG_CLOUD_9: + case IO_REPARSE_TAG_CLOUD_A: + case IO_REPARSE_TAG_CLOUD_B: + case IO_REPARSE_TAG_CLOUD_C: + case IO_REPARSE_TAG_CLOUD_D: + case IO_REPARSE_TAG_CLOUD_E: + case IO_REPARSE_TAG_CLOUD_F: + err = sizeof("OneDrive") - 1; + if (err > buflen) + err = buflen; + memcpy(buffer, "OneDrive", err); + goto out; + + default: + if (IsReparseTagMicrosoft(rp->ReparseTag)) { + /* unknown Microsoft Tag */ + goto out; + } + if (!IsReparseTagNameSurrogate(rp->ReparseTag) || + i_size <= sizeof(struct REPARSE_POINT)) { + goto out; + } + + /* Users tag */ + uni = Add2Ptr(rp, sizeof(struct REPARSE_POINT) - 2); + nlen = le16_to_cpu(rp->ReparseDataLength) - + sizeof(struct REPARSE_POINT); + } + + /* Convert nlen from bytes to UNICODE chars */ + nlen >>= 1; + + /* Check that name is available */ + if (!nlen || &uni->name[nlen] > (__le16 *)Add2Ptr(rp, i_size)) + goto out; + + /* If name is already zero terminated then truncate it now */ + if (!uni->name[nlen - 1]) + nlen -= 1; + uni->len = nlen; + + err = ntfs_utf16_to_nls(sbi, uni, buffer, buflen); + + if (err < 0) + goto out; + + /* translate windows '\' into linux '/' */ + for (i = 0; i < err; i++) { + if (buffer[i] == '\\') + buffer[i] = '/'; + } + + /* Always set last zero */ + buffer[err] = 0; +out: + ntfs_free(to_free); + return err; +} + +static const char *ntfs_get_link(struct dentry *de, struct inode *inode, + struct delayed_call *done) +{ + int err; + char *ret; + + if (!de) + return ERR_PTR(-ECHILD); + + ret = kmalloc(PAGE_SIZE, GFP_NOFS); + if (!ret) + return ERR_PTR(-ENOMEM); + + err = ntfs_readlink_hlp(inode, ret, PAGE_SIZE); + if (err < 0) { + kfree(ret); + return ERR_PTR(err); + } + + set_delayed_call(done, kfree_link, ret); + + return ret; +} + +// clang-format off +const struct inode_operations ntfs_link_inode_operations = { + .get_link = ntfs_get_link, + .setattr = ntfs3_setattr, + .listxattr = ntfs_listxattr, + .permission = ntfs_permission, + .get_acl = ntfs_get_acl, + .set_acl = ntfs_set_acl, +}; + +const struct address_space_operations ntfs_aops = { + .readpage = ntfs_readpage, + .readahead = ntfs_readahead, + .writepage = ntfs_writepage, + .writepages = ntfs_writepages, + .write_begin = ntfs_write_begin, + .write_end = ntfs_write_end, + .direct_IO = ntfs_direct_IO, + .bmap = ntfs_bmap, + .set_page_dirty = __set_page_dirty_buffers, +}; + +const struct address_space_operations ntfs_aops_cmpr = { + .readpage = ntfs_readpage, + .readahead = ntfs_readahead, +}; +// clang-format on diff --git a/fs/ntfs3/super.c b/fs/ntfs3/super.c new file mode 100644 index 000000000000..6be13e256c1a --- /dev/null +++ b/fs/ntfs3/super.c @@ -0,0 +1,1504 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * + * Copyright (C) 2019-2021 Paragon Software GmbH, All rights reserved. + * + * + * terminology + * + * cluster - allocation unit - 512,1K,2K,4K,...,2M + * vcn - virtual cluster number - offset inside the file in clusters + * vbo - virtual byte offset - offset inside the file in bytes + * lcn - logical cluster number - 0 based cluster in clusters heap + * lbo - logical byte offset - absolute position inside volume + * run - maps vcn to lcn - stored in attributes in packed form + * attr - attribute segment - std/name/data etc records inside MFT + * mi - mft inode - one MFT record(usually 1024 bytes or 4K), consists of attributes + * ni - ntfs inode - extends linux inode. consists of one or more mft inodes + * index - unit inside directory - 2K, 4K, <=page size, does not depend on cluster size + * + * WSL - Windows Subsystem for Linux + * https://docs.microsoft.com/en-us/windows/wsl/file-permissions + * It stores uid/gid/mode/dev in xattr + * + */ + +#include <linux/backing-dev.h> +#include <linux/blkdev.h> +#include <linux/buffer_head.h> +#include <linux/exportfs.h> +#include <linux/fs.h> +#include <linux/iversion.h> +#include <linux/module.h> +#include <linux/nls.h> +#include <linux/parser.h> +#include <linux/seq_file.h> +#include <linux/statfs.h> + +#include "debug.h" +#include "ntfs.h" +#include "ntfs_fs.h" +#ifdef CONFIG_NTFS3_LZX_XPRESS +#include "lib/lib.h" +#endif + +#ifdef CONFIG_PRINTK +/* + * Trace warnings/notices/errors + * Thanks Joe Perches <joe(a)perches.com> for implementation + */ +void ntfs_printk(const struct super_block *sb, const char *fmt, ...) +{ + struct va_format vaf; + va_list args; + int level; + struct ntfs_sb_info *sbi = sb->s_fs_info; + + /*should we use different ratelimits for warnings/notices/errors? */ + if (!___ratelimit(&sbi->msg_ratelimit, "ntfs3")) + return; + + va_start(args, fmt); + + level = printk_get_level(fmt); + vaf.fmt = printk_skip_level(fmt); + vaf.va = &args; + printk("%c%cntfs3: %s: %pV\n", KERN_SOH_ASCII, level, sb->s_id, &vaf); + + va_end(args); +} + +static char s_name_buf[512]; +static atomic_t s_name_buf_cnt = ATOMIC_INIT(1); // 1 means 'free s_name_buf' + +/* print warnings/notices/errors about inode using name or inode number */ +void ntfs_inode_printk(struct inode *inode, const char *fmt, ...) +{ + struct super_block *sb = inode->i_sb; + struct ntfs_sb_info *sbi = sb->s_fs_info; + char *name; + va_list args; + struct va_format vaf; + int level; + + if (!___ratelimit(&sbi->msg_ratelimit, "ntfs3")) + return; + + /* use static allocated buffer, if possible */ + name = atomic_dec_and_test(&s_name_buf_cnt) + ? s_name_buf + : kmalloc(sizeof(s_name_buf), GFP_NOFS); + + if (name) { + struct dentry *de = d_find_alias(inode); + const u32 name_len = ARRAY_SIZE(s_name_buf) - 1; + + if (de) { + spin_lock(&de->d_lock); + snprintf(name, name_len, " \"%s\"", de->d_name.name); + spin_unlock(&de->d_lock); + name[name_len] = 0; /* to be sure*/ + } else { + name[0] = 0; + } + dput(de); /* cocci warns if placed in branch "if (de)" */ + } + + va_start(args, fmt); + + level = printk_get_level(fmt); + vaf.fmt = printk_skip_level(fmt); + vaf.va = &args; + + printk("%c%cntfs3: %s: ino=%lx,%s %pV\n", KERN_SOH_ASCII, level, + sb->s_id, inode->i_ino, name ? name : "", &vaf); + + va_end(args); + + atomic_inc(&s_name_buf_cnt); + if (name != s_name_buf) + kfree(name); +} +#endif + +/* + * Shared memory struct. + * + * on-disk ntfs's upcase table is created by ntfs formater + * 'upcase' table is 128K bytes of memory + * we should read it into memory when mounting + * Several ntfs volumes likely use the same 'upcase' table + * It is good idea to share in-memory 'upcase' table between different volumes + * Unfortunately winxp/vista/win7 use different upcase tables + */ +static DEFINE_SPINLOCK(s_shared_lock); + +static struct { + void *ptr; + u32 len; + int cnt; +} s_shared[8]; + +/* + * ntfs_set_shared + * + * Returns 'ptr' if pointer was saved in shared memory + * Returns NULL if pointer was not shared + */ +void *ntfs_set_shared(void *ptr, u32 bytes) +{ + void *ret = NULL; + int i, j = -1; + + spin_lock(&s_shared_lock); + for (i = 0; i < ARRAY_SIZE(s_shared); i++) { + if (!s_shared[i].cnt) { + j = i; + } else if (bytes == s_shared[i].len && + !memcmp(s_shared[i].ptr, ptr, bytes)) { + s_shared[i].cnt += 1; + ret = s_shared[i].ptr; + break; + } + } + + if (!ret && j != -1) { + s_shared[j].ptr = ptr; + s_shared[j].len = bytes; + s_shared[j].cnt = 1; + ret = ptr; + } + spin_unlock(&s_shared_lock); + + return ret; +} + +/* + * ntfs_put_shared + * + * Returns 'ptr' if pointer is not shared anymore + * Returns NULL if pointer is still shared + */ +void *ntfs_put_shared(void *ptr) +{ + void *ret = ptr; + int i; + + spin_lock(&s_shared_lock); + for (i = 0; i < ARRAY_SIZE(s_shared); i++) { + if (s_shared[i].cnt && s_shared[i].ptr == ptr) { + if (--s_shared[i].cnt) + ret = NULL; + break; + } + } + spin_unlock(&s_shared_lock); + + return ret; +} + +static inline void clear_mount_options(struct ntfs_mount_options *options) +{ + unload_nls(options->nls); +} + +enum Opt { + Opt_uid, + Opt_gid, + Opt_umask, + Opt_dmask, + Opt_fmask, + Opt_immutable, + Opt_discard, + Opt_force, + Opt_sparse, + Opt_nohidden, + Opt_showmeta, + Opt_acl, + Opt_noatime, + Opt_nls, + Opt_prealloc, + Opt_no_acs_rules, + Opt_err, +}; + +static const match_table_t ntfs_tokens = { + { Opt_uid, "uid=%u" }, + { Opt_gid, "gid=%u" }, + { Opt_umask, "umask=%o" }, + { Opt_dmask, "dmask=%o" }, + { Opt_fmask, "fmask=%o" }, + { Opt_immutable, "sys_immutable" }, + { Opt_discard, "discard" }, + { Opt_force, "force" }, + { Opt_sparse, "sparse" }, + { Opt_nohidden, "nohidden" }, + { Opt_acl, "acl" }, + { Opt_noatime, "noatime" }, + { Opt_showmeta, "showmeta" }, + { Opt_nls, "nls=%s" }, + { Opt_prealloc, "prealloc" }, + { Opt_no_acs_rules, "no_acs_rules" }, + { Opt_err, NULL }, +}; + +static noinline int ntfs_parse_options(struct super_block *sb, char *options, + int silent, + struct ntfs_mount_options *opts) +{ + char *p; + substring_t args[MAX_OPT_ARGS]; + int option; + char nls_name[30]; + struct nls_table *nls; + + opts->fs_uid = current_uid(); + opts->fs_gid = current_gid(); + opts->fs_fmask_inv = opts->fs_dmask_inv = ~current_umask(); + nls_name[0] = 0; + + if (!options) + goto out; + + while ((p = strsep(&options, ","))) { + int token; + + if (!*p) + continue; + + token = match_token(p, ntfs_tokens, args); + switch (token) { + case Opt_immutable: + opts->sys_immutable = 1; + break; + case Opt_uid: + if (match_int(&args[0], &option)) + return -EINVAL; + opts->fs_uid = make_kuid(current_user_ns(), option); + if (!uid_valid(opts->fs_uid)) + return -EINVAL; + opts->uid = 1; + break; + case Opt_gid: + if (match_int(&args[0], &option)) + return -EINVAL; + opts->fs_gid = make_kgid(current_user_ns(), option); + if (!gid_valid(opts->fs_gid)) + return -EINVAL; + opts->gid = 1; + break; + case Opt_umask: + if (match_octal(&args[0], &option)) + return -EINVAL; + opts->fs_fmask_inv = opts->fs_dmask_inv = ~option; + opts->fmask = opts->dmask = 1; + break; + case Opt_dmask: + if (match_octal(&args[0], &option)) + return -EINVAL; + opts->fs_dmask_inv = ~option; + opts->dmask = 1; + break; + case Opt_fmask: + if (match_octal(&args[0], &option)) + return -EINVAL; + opts->fs_fmask_inv = ~option; + opts->fmask = 1; + break; + case Opt_discard: + opts->discard = 1; + break; + case Opt_force: + opts->force = 1; + break; + case Opt_sparse: + opts->sparse = 1; + break; + case Opt_nohidden: + opts->nohidden = 1; + break; + case Opt_acl: +#ifdef CONFIG_NTFS3_FS_POSIX_ACL + sb->s_flags |= SB_POSIXACL; + break; +#else + ntfs_err(sb, "support for ACL not compiled in!"); + return -EINVAL; +#endif + case Opt_noatime: + sb->s_flags |= SB_NOATIME; + break; + case Opt_showmeta: + opts->showmeta = 1; + break; + case Opt_nls: + match_strlcpy(nls_name, &args[0], sizeof(nls_name)); + break; + case Opt_prealloc: + opts->prealloc = 1; + break; + case Opt_no_acs_rules: + opts->no_acs_rules = 1; + break; + default: + if (!silent) + ntfs_err( + sb, + "Unrecognized mount option \"%s\" or missing value", + p); + //return -EINVAL; + } + } + +out: + if (!strcmp(nls_name[0] ? nls_name : CONFIG_NLS_DEFAULT, "utf8")) { + /* For UTF-8 use utf16s_to_utf8s/utf8s_to_utf16s instead of nls */ + nls = NULL; + } else if (nls_name[0]) { + nls = load_nls(nls_name); + if (!nls) { + ntfs_err(sb, "failed to load \"%s\"", nls_name); + return -EINVAL; + } + } else { + nls = load_nls_default(); + if (!nls) { + ntfs_err(sb, "failed to load default nls"); + return -EINVAL; + } + } + opts->nls = nls; + + return 0; +} + +static int ntfs_remount(struct super_block *sb, int *flags, char *data) +{ + int err, ro_rw; + struct ntfs_sb_info *sbi = sb->s_fs_info; + struct ntfs_mount_options old_opts; + char *orig_data = kstrdup(data, GFP_KERNEL); + + if (data && !orig_data) + return -ENOMEM; + + /* Store original options */ + memcpy(&old_opts, &sbi->options, sizeof(old_opts)); + clear_mount_options(&sbi->options); + memset(&sbi->options, 0, sizeof(sbi->options)); + + err = ntfs_parse_options(sb, data, 0, &sbi->options); + if (err) + goto restore_opts; + + ro_rw = sb_rdonly(sb) && !(*flags & SB_RDONLY); + if (ro_rw && (sbi->flags & NTFS_FLAGS_NEED_REPLAY)) { + ntfs_warn( + sb, + "Couldn't remount rw because journal is not replayed. Please umount/remount instead\n"); + err = -EINVAL; + goto restore_opts; + } + + sync_filesystem(sb); + + if (ro_rw && (sbi->volume.flags & VOLUME_FLAG_DIRTY) && + !sbi->options.force) { + ntfs_warn(sb, "volume is dirty and \"force\" flag is not set!"); + err = -EINVAL; + goto restore_opts; + } + + clear_mount_options(&old_opts); + + *flags = (*flags & ~SB_LAZYTIME) | (sb->s_flags & SB_LAZYTIME) | + SB_NODIRATIME | SB_NOATIME; + ntfs_info(sb, "re-mounted. Opts: %s", orig_data); + err = 0; + goto out; + +restore_opts: + clear_mount_options(&sbi->options); + memcpy(&sbi->options, &old_opts, sizeof(old_opts)); + +out: + kfree(orig_data); + return err; +} + +static struct kmem_cache *ntfs_inode_cachep; + +static struct inode *ntfs_alloc_inode(struct super_block *sb) +{ + struct ntfs_inode *ni = kmem_cache_alloc(ntfs_inode_cachep, GFP_NOFS); + + if (!ni) + return NULL; + + memset(ni, 0, offsetof(struct ntfs_inode, vfs_inode)); + + mutex_init(&ni->ni_lock); + + return &ni->vfs_inode; +} + +static void ntfs_i_callback(struct rcu_head *head) +{ + struct inode *inode = container_of(head, struct inode, i_rcu); + struct ntfs_inode *ni = ntfs_i(inode); + + mutex_destroy(&ni->ni_lock); + + kmem_cache_free(ntfs_inode_cachep, ni); +} + +static void ntfs_destroy_inode(struct inode *inode) +{ + call_rcu(&inode->i_rcu, ntfs_i_callback); +} + +static void init_once(void *foo) +{ + struct ntfs_inode *ni = foo; + + inode_init_once(&ni->vfs_inode); +} + +/* noinline to reduce binary size*/ +static noinline void put_ntfs(struct ntfs_sb_info *sbi) +{ + ntfs_free(sbi->new_rec); + ntfs_vfree(ntfs_put_shared(sbi->upcase)); + ntfs_free(sbi->def_table); + + wnd_close(&sbi->mft.bitmap); + wnd_close(&sbi->used.bitmap); + + if (sbi->mft.ni) + iput(&sbi->mft.ni->vfs_inode); + + if (sbi->security.ni) + iput(&sbi->security.ni->vfs_inode); + + if (sbi->reparse.ni) + iput(&sbi->reparse.ni->vfs_inode); + + if (sbi->objid.ni) + iput(&sbi->objid.ni->vfs_inode); + + if (sbi->volume.ni) + iput(&sbi->volume.ni->vfs_inode); + + ntfs_update_mftmirr(sbi, 0); + + indx_clear(&sbi->security.index_sii); + indx_clear(&sbi->security.index_sdh); + indx_clear(&sbi->reparse.index_r); + indx_clear(&sbi->objid.index_o); + ntfs_free(sbi->compress.lznt); +#ifdef CONFIG_NTFS3_LZX_XPRESS + xpress_free_decompressor(sbi->compress.xpress); + lzx_free_decompressor(sbi->compress.lzx); +#endif + clear_mount_options(&sbi->options); + + ntfs_free(sbi); +} + +static void ntfs_put_super(struct super_block *sb) +{ + struct ntfs_sb_info *sbi = sb->s_fs_info; + + /*mark rw ntfs as clear, if possible*/ + ntfs_set_state(sbi, NTFS_DIRTY_CLEAR); + + put_ntfs(sbi); + + sync_blockdev(sb->s_bdev); +} + +static int ntfs_statfs(struct dentry *dentry, struct kstatfs *buf) +{ + struct super_block *sb = dentry->d_sb; + struct ntfs_sb_info *sbi = sb->s_fs_info; + struct wnd_bitmap *wnd = &sbi->used.bitmap; + + buf->f_type = sb->s_magic; + buf->f_bsize = sbi->cluster_size; + buf->f_blocks = wnd->nbits; + + buf->f_bfree = buf->f_bavail = wnd_zeroes(wnd); + buf->f_fsid.val[0] = sbi->volume.ser_num; + buf->f_fsid.val[1] = (sbi->volume.ser_num >> 32); + buf->f_namelen = NTFS_NAME_LEN; + + return 0; +} + +static int ntfs_show_options(struct seq_file *m, struct dentry *root) +{ + struct super_block *sb = root->d_sb; + struct ntfs_sb_info *sbi = sb->s_fs_info; + struct ntfs_mount_options *opts = &sbi->options; + struct user_namespace *user_ns = seq_user_ns(m); + + if (opts->uid) + seq_printf(m, ",uid=%u", + from_kuid_munged(user_ns, opts->fs_uid)); + if (opts->gid) + seq_printf(m, ",gid=%u", + from_kgid_munged(user_ns, opts->fs_gid)); + if (opts->fmask) + seq_printf(m, ",fmask=%04o", ~opts->fs_fmask_inv); + if (opts->dmask) + seq_printf(m, ",dmask=%04o", ~opts->fs_dmask_inv); + if (opts->nls) + seq_printf(m, ",nls=%s", opts->nls->charset); + else + seq_puts(m, ",nls=utf8"); + if (opts->sys_immutable) + seq_puts(m, ",sys_immutable"); + if (opts->discard) + seq_puts(m, ",discard"); + if (opts->sparse) + seq_puts(m, ",sparse"); + if (opts->showmeta) + seq_puts(m, ",showmeta"); + if (opts->nohidden) + seq_puts(m, ",nohidden"); + if (opts->force) + seq_puts(m, ",force"); + if (opts->no_acs_rules) + seq_puts(m, ",no_acs_rules"); + if (opts->prealloc) + seq_puts(m, ",prealloc"); + if (sb->s_flags & SB_POSIXACL) + seq_puts(m, ",acl"); + if (sb->s_flags & SB_NOATIME) + seq_puts(m, ",noatime"); + + return 0; +} + +/*super_operations::sync_fs*/ +static int ntfs_sync_fs(struct super_block *sb, int wait) +{ + int err = 0, err2; + struct ntfs_sb_info *sbi = sb->s_fs_info; + struct ntfs_inode *ni; + struct inode *inode; + + ni = sbi->security.ni; + if (ni) { + inode = &ni->vfs_inode; + err2 = _ni_write_inode(inode, wait); + if (err2 && !err) + err = err2; + } + + ni = sbi->objid.ni; + if (ni) { + inode = &ni->vfs_inode; + err2 = _ni_write_inode(inode, wait); + if (err2 && !err) + err = err2; + } + + ni = sbi->reparse.ni; + if (ni) { + inode = &ni->vfs_inode; + err2 = _ni_write_inode(inode, wait); + if (err2 && !err) + err = err2; + } + + if (!err) + ntfs_set_state(sbi, NTFS_DIRTY_CLEAR); + + ntfs_update_mftmirr(sbi, wait); + + return err; +} + +static const struct super_operations ntfs_sops = { + .alloc_inode = ntfs_alloc_inode, + .destroy_inode = ntfs_destroy_inode, + .evict_inode = ntfs_evict_inode, + .put_super = ntfs_put_super, + .statfs = ntfs_statfs, + .show_options = ntfs_show_options, + .sync_fs = ntfs_sync_fs, + .remount_fs = ntfs_remount, + .write_inode = ntfs3_write_inode, +}; + +static struct inode *ntfs_export_get_inode(struct super_block *sb, u64 ino, + u32 generation) +{ + struct MFT_REF ref; + struct inode *inode; + + ref.low = cpu_to_le32(ino); +#ifdef CONFIG_NTFS3_64BIT_CLUSTER + ref.high = cpu_to_le16(ino >> 32); +#else + ref.high = 0; +#endif + ref.seq = cpu_to_le16(generation); + + inode = ntfs_iget5(sb, &ref, NULL); + if (!IS_ERR(inode) && is_bad_inode(inode)) { + iput(inode); + inode = ERR_PTR(-ESTALE); + } + + return inode; +} + +static struct dentry *ntfs_fh_to_dentry(struct super_block *sb, struct fid *fid, + int fh_len, int fh_type) +{ + return generic_fh_to_dentry(sb, fid, fh_len, fh_type, + ntfs_export_get_inode); +} + +static struct dentry *ntfs_fh_to_parent(struct super_block *sb, struct fid *fid, + int fh_len, int fh_type) +{ + return generic_fh_to_parent(sb, fid, fh_len, fh_type, + ntfs_export_get_inode); +} + +/* TODO: == ntfs_sync_inode */ +static int ntfs_nfs_commit_metadata(struct inode *inode) +{ + return _ni_write_inode(inode, 1); +} + +static const struct export_operations ntfs_export_ops = { + .fh_to_dentry = ntfs_fh_to_dentry, + .fh_to_parent = ntfs_fh_to_parent, + .get_parent = ntfs3_get_parent, + .commit_metadata = ntfs_nfs_commit_metadata, +}; + +/* Returns Gb,Mb to print with "%u.%02u Gb" */ +static u32 format_size_gb(const u64 bytes, u32 *mb) +{ + /* Do simple right 30 bit shift of 64 bit value */ + u64 kbytes = bytes >> 10; + u32 kbytes32 = kbytes; + + *mb = (100 * (kbytes32 & 0xfffff) + 0x7ffff) >> 20; + if (*mb >= 100) + *mb = 99; + + return (kbytes32 >> 20) | (((u32)(kbytes >> 32)) << 12); +} + +static u32 true_sectors_per_clst(const struct NTFS_BOOT *boot) +{ + return boot->sectors_per_clusters <= 0x80 + ? boot->sectors_per_clusters + : (1u << (0 - boot->sectors_per_clusters)); +} + +/* inits internal info from on-disk boot sector*/ +static int ntfs_init_from_boot(struct super_block *sb, u32 sector_size, + u64 dev_size) +{ + struct ntfs_sb_info *sbi = sb->s_fs_info; + int err; + u32 mb, gb, boot_sector_size, sct_per_clst, record_size; + u64 sectors, clusters, fs_size, mlcn, mlcn2; + struct NTFS_BOOT *boot; + struct buffer_head *bh; + struct MFT_REC *rec; + u16 fn, ao; + + sbi->volume.blocks = dev_size >> PAGE_SHIFT; + + bh = ntfs_bread(sb, 0); + if (!bh) + return -EIO; + + err = -EINVAL; + boot = (struct NTFS_BOOT *)bh->b_data; + + if (memcmp(boot->system_id, "NTFS ", sizeof("NTFS ") - 1)) + goto out; + + /* 0x55AA is not mandaroty. Thanks Maxim Suhanov*/ + /*if (0x55 != boot->boot_magic[0] || 0xAA != boot->boot_magic[1]) + * goto out; + */ + + boot_sector_size = (u32)boot->bytes_per_sector[1] << 8; + if (boot->bytes_per_sector[0] || boot_sector_size < SECTOR_SIZE || + !is_power_of2(boot_sector_size)) { + goto out; + } + + /* cluster size: 512, 1K, 2K, 4K, ... 2M */ + sct_per_clst = true_sectors_per_clst(boot); + if (!is_power_of2(sct_per_clst)) + goto out; + + mlcn = le64_to_cpu(boot->mft_clst); + mlcn2 = le64_to_cpu(boot->mft2_clst); + sectors = le64_to_cpu(boot->sectors_per_volume); + + if (mlcn * sct_per_clst >= sectors) + goto out; + + if (mlcn2 * sct_per_clst >= sectors) + goto out; + + /* Check MFT record size */ + if ((boot->record_size < 0 && + SECTOR_SIZE > (2U << (-boot->record_size))) || + (boot->record_size >= 0 && !is_power_of2(boot->record_size))) { + goto out; + } + + /* Check index record size */ + if ((boot->index_size < 0 && + SECTOR_SIZE > (2U << (-boot->index_size))) || + (boot->index_size >= 0 && !is_power_of2(boot->index_size))) { + goto out; + } + + sbi->sector_size = boot_sector_size; + sbi->sector_bits = blksize_bits(boot_sector_size); + fs_size = (sectors + 1) << sbi->sector_bits; + + gb = format_size_gb(fs_size, &mb); + + /* + * - Volume formatted and mounted with the same sector size + * - Volume formatted 4K and mounted as 512 + * - Volume formatted 512 and mounted as 4K + */ + if (sbi->sector_size != sector_size) { + ntfs_warn(sb, + "Different NTFS' sector size and media sector size"); + dev_size += sector_size - 1; + } + + sbi->cluster_size = boot_sector_size * sct_per_clst; + sbi->cluster_bits = blksize_bits(sbi->cluster_size); + + sbi->mft.lbo = mlcn << sbi->cluster_bits; + sbi->mft.lbo2 = mlcn2 << sbi->cluster_bits; + + if (sbi->cluster_size < sbi->sector_size) + goto out; + + sbi->cluster_mask = sbi->cluster_size - 1; + sbi->cluster_mask_inv = ~(u64)sbi->cluster_mask; + sbi->record_size = record_size = boot->record_size < 0 + ? 1 << (-boot->record_size) + : (u32)boot->record_size + << sbi->cluster_bits; + + if (record_size > MAXIMUM_BYTES_PER_MFT) + goto out; + + sbi->record_bits = blksize_bits(record_size); + sbi->attr_size_tr = (5 * record_size >> 4); // ~320 bytes + + sbi->max_bytes_per_attr = + record_size - QuadAlign(MFTRECORD_FIXUP_OFFSET_1) - + QuadAlign(((record_size >> SECTOR_SHIFT) * sizeof(short))) - + QuadAlign(sizeof(enum ATTR_TYPE)); + + sbi->index_size = boot->index_size < 0 + ? 1u << (-boot->index_size) + : (u32)boot->index_size << sbi->cluster_bits; + + sbi->volume.ser_num = le64_to_cpu(boot->serial_num); + sbi->volume.size = sectors << sbi->sector_bits; + + /* warning if RAW volume */ + if (dev_size < fs_size) { + u32 mb0, gb0; + + gb0 = format_size_gb(dev_size, &mb0); + ntfs_warn( + sb, + "RAW NTFS volume: Filesystem size %u.%02u Gb > volume size %u.%02u Gb. Mount in read-only", + gb, mb, gb0, mb0); + sb->s_flags |= SB_RDONLY; + } + + clusters = sbi->volume.size >> sbi->cluster_bits; +#ifndef CONFIG_NTFS3_64BIT_CLUSTER + /* 32 bits per cluster */ + if (clusters >> 32) { + ntfs_notice( + sb, + "NTFS %u.%02u Gb is too big to use 32 bits per cluster", + gb, mb); + goto out; + } +#elif BITS_PER_LONG < 64 +#error "CONFIG_NTFS3_64BIT_CLUSTER incompatible in 32 bit OS" +#endif + + sbi->used.bitmap.nbits = clusters; + + rec = ntfs_zalloc(record_size); + if (!rec) { + err = -ENOMEM; + goto out; + } + + sbi->new_rec = rec; + rec->rhdr.sign = NTFS_FILE_SIGNATURE; + rec->rhdr.fix_off = cpu_to_le16(MFTRECORD_FIXUP_OFFSET_1); + fn = (sbi->record_size >> SECTOR_SHIFT) + 1; + rec->rhdr.fix_num = cpu_to_le16(fn); + ao = QuadAlign(MFTRECORD_FIXUP_OFFSET_1 + sizeof(short) * fn); + rec->attr_off = cpu_to_le16(ao); + rec->used = cpu_to_le32(ao + QuadAlign(sizeof(enum ATTR_TYPE))); + rec->total = cpu_to_le32(sbi->record_size); + ((struct ATTRIB *)Add2Ptr(rec, ao))->type = ATTR_END; + + if (sbi->cluster_size < PAGE_SIZE) + sb_set_blocksize(sb, sbi->cluster_size); + + sbi->block_mask = sb->s_blocksize - 1; + sbi->blocks_per_cluster = sbi->cluster_size >> sb->s_blocksize_bits; + sbi->volume.blocks = sbi->volume.size >> sb->s_blocksize_bits; + + /* Maximum size for normal files */ + sbi->maxbytes = (clusters << sbi->cluster_bits) - 1; + +#ifdef CONFIG_NTFS3_64BIT_CLUSTER + if (clusters >= (1ull << (64 - sbi->cluster_bits))) + sbi->maxbytes = -1; + sbi->maxbytes_sparse = -1; +#else + /* Maximum size for sparse file */ + sbi->maxbytes_sparse = (1ull << (sbi->cluster_bits + 32)) - 1; +#endif + + err = 0; + +out: + brelse(bh); + + return err; +} + +/* try to mount*/ +static int ntfs_fill_super(struct super_block *sb, void *data, int silent) +{ + int err; + struct ntfs_sb_info *sbi; + struct block_device *bdev = sb->s_bdev; + struct inode *bd_inode = bdev->bd_inode; + struct request_queue *rq = bdev_get_queue(bdev); + struct inode *inode = NULL; + struct ntfs_inode *ni; + size_t i, tt; + CLST vcn, lcn, len; + struct ATTRIB *attr; + const struct VOLUME_INFO *info; + u32 idx, done, bytes; + struct ATTR_DEF_ENTRY *t; + u16 *upcase = NULL; + u16 *shared; + bool is_ro; + struct MFT_REF ref; + + ref.high = 0; + + sbi = ntfs_zalloc(sizeof(struct ntfs_sb_info)); + if (!sbi) + return -ENOMEM; + + sb->s_fs_info = sbi; + sbi->sb = sb; + sb->s_flags |= SB_NODIRATIME; + sb->s_magic = 0x7366746e; // "ntfs" + sb->s_op = &ntfs_sops; + sb->s_export_op = &ntfs_export_ops; + sb->s_time_gran = NTFS_TIME_GRAN; // 100 nsec + sb->s_xattr = ntfs_xattr_handlers; + + ratelimit_state_init(&sbi->msg_ratelimit, DEFAULT_RATELIMIT_INTERVAL, + DEFAULT_RATELIMIT_BURST); + + err = ntfs_parse_options(sb, data, silent, &sbi->options); + if (err) + goto out; + + if (!rq || !blk_queue_discard(rq) || !rq->limits.discard_granularity) { + ; + } else { + sbi->discard_granularity = rq->limits.discard_granularity; + sbi->discard_granularity_mask_inv = + ~(u64)(sbi->discard_granularity - 1); + } + + sb_set_blocksize(sb, PAGE_SIZE); + + /* parse boot */ + err = ntfs_init_from_boot(sb, rq ? queue_logical_block_size(rq) : 512, + bd_inode->i_size); + if (err) + goto out; + +#ifdef CONFIG_NTFS3_64BIT_CLUSTER + sb->s_maxbytes = MAX_LFS_FILESIZE; +#else + sb->s_maxbytes = 0xFFFFFFFFull << sbi->cluster_bits; +#endif + + mutex_init(&sbi->compress.mtx_lznt); +#ifdef CONFIG_NTFS3_LZX_XPRESS + mutex_init(&sbi->compress.mtx_xpress); + mutex_init(&sbi->compress.mtx_lzx); +#endif + + /* + * Load $Volume. This should be done before LogFile + * 'cause 'sbi->volume.ni' is used 'ntfs_set_state' + */ + ref.low = cpu_to_le32(MFT_REC_VOL); + ref.seq = cpu_to_le16(MFT_REC_VOL); + inode = ntfs_iget5(sb, &ref, &NAME_VOLUME); + if (IS_ERR(inode)) { + err = PTR_ERR(inode); + ntfs_err(sb, "Failed to load $Volume."); + inode = NULL; + goto out; + } + + ni = ntfs_i(inode); + + /* Load and save label (not necessary) */ + attr = ni_find_attr(ni, NULL, NULL, ATTR_LABEL, NULL, 0, NULL, NULL); + + if (!attr) { + /* It is ok if no ATTR_LABEL */ + } else if (!attr->non_res && !is_attr_ext(attr)) { + /* $AttrDef allows labels to be up to 128 symbols */ + err = utf16s_to_utf8s(resident_data(attr), + le32_to_cpu(attr->res.data_size) >> 1, + UTF16_LITTLE_ENDIAN, sbi->volume.label, + sizeof(sbi->volume.label)); + if (err < 0) + sbi->volume.label[0] = 0; + } else { + /* should we break mounting here? */ + //err = -EINVAL; + //goto out; + } + + attr = ni_find_attr(ni, attr, NULL, ATTR_VOL_INFO, NULL, 0, NULL, NULL); + if (!attr || is_attr_ext(attr)) { + err = -EINVAL; + goto out; + } + + info = resident_data_ex(attr, SIZEOF_ATTRIBUTE_VOLUME_INFO); + if (!info) { + err = -EINVAL; + goto out; + } + + sbi->volume.major_ver = info->major_ver; + sbi->volume.minor_ver = info->minor_ver; + sbi->volume.flags = info->flags; + + sbi->volume.ni = ni; + inode = NULL; + + /* Load $MFTMirr to estimate recs_mirr */ + ref.low = cpu_to_le32(MFT_REC_MIRR); + ref.seq = cpu_to_le16(MFT_REC_MIRR); + inode = ntfs_iget5(sb, &ref, &NAME_MIRROR); + if (IS_ERR(inode)) { + err = PTR_ERR(inode); + ntfs_err(sb, "Failed to load $MFTMirr."); + inode = NULL; + goto out; + } + + sbi->mft.recs_mirr = + ntfs_up_cluster(sbi, inode->i_size) >> sbi->record_bits; + + iput(inode); + + /* Load LogFile to replay */ + ref.low = cpu_to_le32(MFT_REC_LOG); + ref.seq = cpu_to_le16(MFT_REC_LOG); + inode = ntfs_iget5(sb, &ref, &NAME_LOGFILE); + if (IS_ERR(inode)) { + err = PTR_ERR(inode); + ntfs_err(sb, "Failed to load \x24LogFile."); + inode = NULL; + goto out; + } + + ni = ntfs_i(inode); + + err = ntfs_loadlog_and_replay(ni, sbi); + if (err) + goto out; + + iput(inode); + inode = NULL; + + is_ro = sb_rdonly(sbi->sb); + + if (sbi->flags & NTFS_FLAGS_NEED_REPLAY) { + if (!is_ro) { + ntfs_warn(sb, + "failed to replay log file. Can't mount rw!"); + err = -EINVAL; + goto out; + } + } else if (sbi->volume.flags & VOLUME_FLAG_DIRTY) { + if (!is_ro && !sbi->options.force) { + ntfs_warn( + sb, + "volume is dirty and \"force\" flag is not set!"); + err = -EINVAL; + goto out; + } + } + + /* Load $MFT */ + ref.low = cpu_to_le32(MFT_REC_MFT); + ref.seq = cpu_to_le16(1); + + inode = ntfs_iget5(sb, &ref, &NAME_MFT); + if (IS_ERR(inode)) { + err = PTR_ERR(inode); + ntfs_err(sb, "Failed to load $MFT."); + inode = NULL; + goto out; + } + + ni = ntfs_i(inode); + + sbi->mft.used = ni->i_valid >> sbi->record_bits; + tt = inode->i_size >> sbi->record_bits; + sbi->mft.next_free = MFT_REC_USER; + + err = wnd_init(&sbi->mft.bitmap, sb, tt); + if (err) + goto out; + + err = ni_load_all_mi(ni); + if (err) + goto out; + + sbi->mft.ni = ni; + + /* Load $BadClus */ + ref.low = cpu_to_le32(MFT_REC_BADCLUST); + ref.seq = cpu_to_le16(MFT_REC_BADCLUST); + inode = ntfs_iget5(sb, &ref, &NAME_BADCLUS); + if (IS_ERR(inode)) { + err = PTR_ERR(inode); + ntfs_err(sb, "Failed to load $BadClus."); + inode = NULL; + goto out; + } + + ni = ntfs_i(inode); + + for (i = 0; run_get_entry(&ni->file.run, i, &vcn, &lcn, &len); i++) { + if (lcn == SPARSE_LCN) + continue; + + if (!sbi->bad_clusters) + ntfs_notice(sb, "Volume contains bad blocks"); + + sbi->bad_clusters += len; + } + + iput(inode); + + /* Load $Bitmap */ + ref.low = cpu_to_le32(MFT_REC_BITMAP); + ref.seq = cpu_to_le16(MFT_REC_BITMAP); + inode = ntfs_iget5(sb, &ref, &NAME_BITMAP); + if (IS_ERR(inode)) { + err = PTR_ERR(inode); + ntfs_err(sb, "Failed to load $Bitmap."); + inode = NULL; + goto out; + } + + ni = ntfs_i(inode); + +#ifndef CONFIG_NTFS3_64BIT_CLUSTER + if (inode->i_size >> 32) { + err = -EINVAL; + goto out; + } +#endif + + /* Check bitmap boundary */ + tt = sbi->used.bitmap.nbits; + if (inode->i_size < bitmap_size(tt)) { + err = -EINVAL; + goto out; + } + + /* Not necessary */ + sbi->used.bitmap.set_tail = true; + err = wnd_init(&sbi->used.bitmap, sbi->sb, tt); + if (err) + goto out; + + iput(inode); + + /* Compute the mft zone */ + err = ntfs_refresh_zone(sbi); + if (err) + goto out; + + /* Load $AttrDef */ + ref.low = cpu_to_le32(MFT_REC_ATTR); + ref.seq = cpu_to_le16(MFT_REC_ATTR); + inode = ntfs_iget5(sbi->sb, &ref, &NAME_ATTRDEF); + if (IS_ERR(inode)) { + err = PTR_ERR(inode); + ntfs_err(sb, "Failed to load $AttrDef -> %d", err); + inode = NULL; + goto out; + } + + if (inode->i_size < sizeof(struct ATTR_DEF_ENTRY)) { + err = -EINVAL; + goto out; + } + bytes = inode->i_size; + sbi->def_table = t = ntfs_malloc(bytes); + if (!t) { + err = -ENOMEM; + goto out; + } + + for (done = idx = 0; done < bytes; done += PAGE_SIZE, idx++) { + unsigned long tail = bytes - done; + struct page *page = ntfs_map_page(inode->i_mapping, idx); + + if (IS_ERR(page)) { + err = PTR_ERR(page); + goto out; + } + memcpy(Add2Ptr(t, done), page_address(page), + min(PAGE_SIZE, tail)); + ntfs_unmap_page(page); + + if (!idx && ATTR_STD != t->type) { + err = -EINVAL; + goto out; + } + } + + t += 1; + sbi->def_entries = 1; + done = sizeof(struct ATTR_DEF_ENTRY); + sbi->reparse.max_size = MAXIMUM_REPARSE_DATA_BUFFER_SIZE; + sbi->ea_max_size = 0x10000; /* default formater value */ + + while (done + sizeof(struct ATTR_DEF_ENTRY) <= bytes) { + u32 t32 = le32_to_cpu(t->type); + u64 sz = le64_to_cpu(t->max_sz); + + if ((t32 & 0xF) || le32_to_cpu(t[-1].type) >= t32) + break; + + if (t->type == ATTR_REPARSE) + sbi->reparse.max_size = sz; + else if (t->type == ATTR_EA) + sbi->ea_max_size = sz; + + done += sizeof(struct ATTR_DEF_ENTRY); + t += 1; + sbi->def_entries += 1; + } + iput(inode); + + /* Load $UpCase */ + ref.low = cpu_to_le32(MFT_REC_UPCASE); + ref.seq = cpu_to_le16(MFT_REC_UPCASE); + inode = ntfs_iget5(sb, &ref, &NAME_UPCASE); + if (IS_ERR(inode)) { + err = PTR_ERR(inode); + ntfs_err(sb, "Failed to load \x24LogFile."); + inode = NULL; + goto out; + } + + ni = ntfs_i(inode); + + if (inode->i_size != 0x10000 * sizeof(short)) { + err = -EINVAL; + goto out; + } + + sbi->upcase = upcase = ntfs_vmalloc(0x10000 * sizeof(short)); + if (!upcase) { + err = -ENOMEM; + goto out; + } + + for (idx = 0; idx < (0x10000 * sizeof(short) >> PAGE_SHIFT); idx++) { + const __le16 *src; + u16 *dst = Add2Ptr(upcase, idx << PAGE_SHIFT); + struct page *page = ntfs_map_page(inode->i_mapping, idx); + + if (IS_ERR(page)) { + err = PTR_ERR(page); + goto out; + } + + src = page_address(page); + +#ifdef __BIG_ENDIAN + for (i = 0; i < PAGE_SIZE / sizeof(u16); i++) + *dst++ = le16_to_cpu(*src++); +#else + memcpy(dst, src, PAGE_SIZE); +#endif + ntfs_unmap_page(page); + } + + shared = ntfs_set_shared(upcase, 0x10000 * sizeof(short)); + if (shared && upcase != shared) { + sbi->upcase = shared; + ntfs_vfree(upcase); + } + + iput(inode); + inode = NULL; + + if (is_ntfs3(sbi)) { + /* Load $Secure */ + err = ntfs_security_init(sbi); + if (err) + goto out; + + /* Load $Extend */ + err = ntfs_extend_init(sbi); + if (err) + goto load_root; + + /* Load $Extend\$Reparse */ + err = ntfs_reparse_init(sbi); + if (err) + goto load_root; + + /* Load $Extend\$ObjId */ + err = ntfs_objid_init(sbi); + if (err) + goto load_root; + } + +load_root: + /* Load root */ + ref.low = cpu_to_le32(MFT_REC_ROOT); + ref.seq = cpu_to_le16(MFT_REC_ROOT); + inode = ntfs_iget5(sb, &ref, &NAME_ROOT); + if (IS_ERR(inode)) { + err = PTR_ERR(inode); + ntfs_err(sb, "Failed to load root."); + inode = NULL; + goto out; + } + + ni = ntfs_i(inode); + + sb->s_root = d_make_root(inode); + + if (!sb->s_root) { + err = -EINVAL; + goto out; + } + + return 0; + +out: + iput(inode); + + if (sb->s_root) { + d_drop(sb->s_root); + sb->s_root = NULL; + } + + put_ntfs(sbi); + + sb->s_fs_info = NULL; + return err; +} + +void ntfs_unmap_meta(struct super_block *sb, CLST lcn, CLST len) +{ + struct ntfs_sb_info *sbi = sb->s_fs_info; + struct block_device *bdev = sb->s_bdev; + sector_t devblock = (u64)lcn * sbi->blocks_per_cluster; + unsigned long blocks = (u64)len * sbi->blocks_per_cluster; + unsigned long cnt = 0; + unsigned long limit = global_zone_page_state(NR_FREE_PAGES) + << (PAGE_SHIFT - sb->s_blocksize_bits); + + if (limit >= 0x2000) + limit -= 0x1000; + else if (limit < 32) + limit = 32; + else + limit >>= 1; + + while (blocks--) { + clean_bdev_aliases(bdev, devblock++, 1); + if (cnt++ >= limit) { + sync_blockdev(bdev); + cnt = 0; + } + } +} + +/* + * ntfs_discard + * + * issue a discard request (trim for SSD) + */ +int ntfs_discard(struct ntfs_sb_info *sbi, CLST lcn, CLST len) +{ + int err; + u64 lbo, bytes, start, end; + struct super_block *sb; + + if (sbi->used.next_free_lcn == lcn + len) + sbi->used.next_free_lcn = lcn; + + if (sbi->flags & NTFS_FLAGS_NODISCARD) + return -EOPNOTSUPP; + + if (!sbi->options.discard) + return -EOPNOTSUPP; + + lbo = (u64)lcn << sbi->cluster_bits; + bytes = (u64)len << sbi->cluster_bits; + + /* Align up 'start' on discard_granularity */ + start = (lbo + sbi->discard_granularity - 1) & + sbi->discard_granularity_mask_inv; + /* Align down 'end' on discard_granularity */ + end = (lbo + bytes) & sbi->discard_granularity_mask_inv; + + sb = sbi->sb; + if (start >= end) + return 0; + + err = blkdev_issue_discard(sb->s_bdev, start >> 9, (end - start) >> 9, + GFP_NOFS, 0); + + if (err == -EOPNOTSUPP) + sbi->flags |= NTFS_FLAGS_NODISCARD; + + return err; +} + +static struct dentry *ntfs_mount(struct file_system_type *fs_type, int flags, + const char *dev_name, void *data) +{ + return mount_bdev(fs_type, flags, dev_name, data, ntfs_fill_super); +} + +// clang-format off +static struct file_system_type ntfs_fs_type = { + .owner = THIS_MODULE, + .name = "ntfs3", + .mount = ntfs_mount, + .kill_sb = kill_block_super, + .fs_flags = FS_REQUIRES_DEV | FS_ALLOW_IDMAP, +}; +// clang-format on + +static int __init init_ntfs_fs(void) +{ + int err; + + pr_notice("ntfs3: Index binary search\n"); + pr_notice("ntfs3: Hot fix free clusters\n"); + pr_notice("ntfs3: Max link count %u\n", NTFS_LINK_MAX); + +#ifdef CONFIG_NTFS3_FS_POSIX_ACL + pr_notice("ntfs3: Enabled Linux POSIX ACLs support\n"); +#endif +#ifdef CONFIG_NTFS3_64BIT_CLUSTER + pr_notice("ntfs3: Activated 64 bits per cluster\n"); +#else + pr_notice("ntfs3: Activated 32 bits per cluster\n"); +#endif +#ifdef CONFIG_NTFS3_LZX_XPRESS + pr_notice("ntfs3: Read-only lzx/xpress compression included\n"); +#endif + + err = ntfs3_init_bitmap(); + if (err) + return err; + + ntfs_inode_cachep = kmem_cache_create( + "ntfs_inode_cache", sizeof(struct ntfs_inode), 0, + (SLAB_RECLAIM_ACCOUNT | SLAB_MEM_SPREAD | SLAB_ACCOUNT), + init_once); + if (!ntfs_inode_cachep) { + err = -ENOMEM; + goto out1; + } + + err = register_filesystem(&ntfs_fs_type); + if (err) + goto out; + + return 0; +out: + kmem_cache_destroy(ntfs_inode_cachep); +out1: + ntfs3_exit_bitmap(); + return err; +} + +static void __exit exit_ntfs_fs(void) +{ + if (ntfs_inode_cachep) { + rcu_barrier(); + kmem_cache_destroy(ntfs_inode_cachep); + } + + unregister_filesystem(&ntfs_fs_type); + ntfs3_exit_bitmap(); +} + +MODULE_LICENSE("GPL"); +MODULE_DESCRIPTION("ntfs3 read/write filesystem"); +MODULE_INFO(behaviour, "Index binary search"); +MODULE_INFO(behaviour, "Hot fix free clusters"); +#ifdef CONFIG_NTFS3_FS_POSIX_ACL +MODULE_INFO(behaviour, "Enabled Linux POSIX ACLs support"); +#endif +#ifdef CONFIG_NTFS3_64BIT_CLUSTER +MODULE_INFO(cluster, "Activated 64 bits per cluster"); +#else +MODULE_INFO(cluster, "Activated 32 bits per cluster"); +#endif +#ifdef CONFIG_NTFS3_LZX_XPRESS +MODULE_INFO(compression, "Read-only lzx/xpress compression included"); +#endif + +MODULE_AUTHOR("Konstantin Komarov"); +MODULE_ALIAS_FS("ntfs3"); + +module_init(init_ntfs_fs); +module_exit(exit_ntfs_fs); -- 2.30.0

1 0

[PATCH OLK-5.10] crypto: sm4 - Do not change section of ck and sbox
by shenzijun 07 Dec '21

07 Dec '21

From: 沈子俊 <shenzijun(a)kylinos.cn> mainline inclusion from mainline-v5.16 commit 4a7e1e5fc294687a8941fa3eeb4a7e8539ca5e2f category: bugfix bugzilla: NA CVE: NA ----------------------------------------------------------------- When building with clang and GNU as, there is a warning about ignored changed section attributes: /tmp/sm4-c916c8.s: Assembler messages: /tmp/sm4-c916c8.s:677: Warning: ignoring changed section attributes for .data..cacheline_aligned "static const" places the data in .rodata but __cacheline_aligned has the section attribute to place it in .data..cacheline_aligned, in addition to the aligned attribute. To keep the alignment but avoid attempting to change sections, use the ____cacheline_aligned attribute, which is just the aligned attribute. Fixes: 2b31277af577 ("crypto: sm4 - create SM4 library based on sm4 generic code") Link: https://github.com/ClangBuiltLinux/linux/issues/1441 Signed-off-by: Nathan Chancellor <nathan(a)kernel.org> Reviewed-by: Tianjia Zhang <tianjia.zhang(a)linux.alibaba.com> Signed-off-by: Herbert Xu <herbert(a)gondor.apana.org.au> Signed-off-by: 沈子俊 <shenzijun(a)kylinos.cn> --- lib/crypto/sm4.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/lib/crypto/sm4.c b/lib/crypto/sm4.c index 633b59fed9db..284e62576d0c 100644 --- a/lib/crypto/sm4.c +++ b/lib/crypto/sm4.c @@ -15,7 +15,7 @@ static const u32 fk[4] = { 0xa3b1bac6, 0x56aa3350, 0x677d9197, 0xb27022dc }; -static const u32 __cacheline_aligned ck[32] = { +static const u32 ____cacheline_aligned ck[32] = { 0x00070e15, 0x1c232a31, 0x383f464d, 0x545b6269, 0x70777e85, 0x8c939aa1, 0xa8afb6bd, 0xc4cbd2d9, 0xe0e7eef5, 0xfc030a11, 0x181f262d, 0x343b4249, @@ -26,7 +26,7 @@ static const u32 __cacheline_aligned ck[32] = { 0x10171e25, 0x2c333a41, 0x484f565d, 0x646b7279 }; -static const u8 __cacheline_aligned sbox[256] = { +static const u8 ____cacheline_aligned sbox[256] = { 0xd6, 0x90, 0xe9, 0xfe, 0xcc, 0xe1, 0x3d, 0xb7, 0x16, 0xb6, 0x14, 0xc2, 0x28, 0xfb, 0x2c, 0x05, 0x2b, 0x67, 0x9a, 0x76, 0x2a, 0xbe, 0x04, 0xc3, -- 2.30.0

1 2

[PATCH openEuler-1.0-LTS 1/6] ext4: always panic when errors=panic is specified
by Yang Yingliang 07 Dec '21

07 Dec '21

From: Ye Bin <yebin10(a)huawei.com> mainline inclusion from mainline-v5.13-rc1 commit ac2f7ca51b0929461ea49918f27c11b680f28995 category: bugfix bugzilla: 182973 CVE: NA ------------------------------------------------- Before commit 014c9caa29d3 ("ext4: make ext4_abort() use __ext4_error()"), the following series of commands would trigger a panic: 1. mount /dev/sda -o ro,errors=panic test 2. mount /dev/sda -o remount,abort test After commit 014c9caa29d3, remounting a file system using the test mount option "abort" will no longer trigger a panic. This commit will restore the behaviour immediately before commit 014c9caa29d3. (However, note that the Linux kernel's behavior has not been consistent; some previous kernel versions, including 5.4 and 4.19 similarly did not panic after using the mount option "abort".) This also makes a change to long-standing behaviour; namely, the following series commands will now cause a panic, when previously it did not: 1. mount /dev/sda -o ro,errors=panic test 2. echo test > /sys/fs/ext4/sda/trigger_fs_error However, this makes ext4's behaviour much more consistent, so this is a good thing. Cc: stable(a)kernel.org Fixes: 014c9caa29d3 ("ext4: make ext4_abort() use __ext4_error()") Signed-off-by: Ye Bin <yebin10(a)huawei.com> Link: https://lore.kernel.org/r/20210401081903.3421208-1-yebin10@huawei.com Signed-off-by: Theodore Ts'o <tytso(a)mit.edu> Signed-off-by: Zheng Liang <zhengliang6(a)huawei.com> Reviewed-by: Zhang Yi <yi.zhang(a)huawei.com> Reviewed-by: Zhang Yi <yi.zhang(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- fs/ext4/super.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/fs/ext4/super.c b/fs/ext4/super.c index a051671f7cb89..5a58f72ac2090 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -652,12 +652,6 @@ static void ext4_handle_error(struct super_block *sb, bool force_ro, int error, ext4_commit_super(sb); } - if (sb_rdonly(sb)) - return; - - if (continue_fs) - goto out; - /* * We force ERRORS_RO behavior when system is rebooting. Otherwise we * could panic during 'reboot -f' as the underlying device got already @@ -668,6 +662,12 @@ static void ext4_handle_error(struct super_block *sb, bool force_ro, int error, sb->s_id); } + if (sb_rdonly(sb)) + return; + + if (continue_fs) + goto out; + ext4_msg(sb, KERN_CRIT, "Remounting filesystem read-only"); /* * Make sure updated value of ->s_mount_flags will be visible before -- 2.25.1

1 5

[openEuler-5.10 00/21] Backport 5.10.81 LTS
by Zheng Zengkai 06 Dec '21

06 Dec '21

Backport 5.10.81 LTS patches from upstream. Borislav Petkov (1): selftests/x86/iopl: Adjust to the faked iopl CLI/STI usage Gao Xiang (1): erofs: fix unsafe pagevec reuse of hooked pclusters Greg Thelen (1): perf/core: Avoid put_page() when GUP fails Joakim Zhang (2): net: stmmac: add clocks management for gmac driver net: stmmac: fix system hang if change mac address after interface ifdown Kees Cook (1): fortify: Explicitly disable Clang support Marc Zyngier (2): PCI/MSI: Deal with devices lying about their MSI mask capability PCI: Add MSI masking quirk for Nvidia ION AHCI Masami Hiramatsu (1): bootconfig: init: Fix memblock leak in xbc_make_cmdline() Michael Riesch (1): net: stmmac: dwmac-rk: fix unbalanced pm_runtime_enable warnings Nathan Chancellor (1): scripts/lld-version.sh: Rewrite based on upstream ld-version.sh Peter Zijlstra (1): x86/iopl: Fake iopl(3) CLI/STI usage Subbaraman Narayanamurthy (1): thermal: Fix NULL pointer dereferences in of_thermal_ functions Sven Schnelle (1): parisc/entry: fix trace test in syscall exit path Thomas Gleixner (1): PCI/MSI: Destroy sysfs before freeing entries Wei Yongjun (1): net: stmmac: platform: fix build error with !CONFIG_PM_SLEEP Wong Vee Khee (1): net: stmmac: fix issue where clk is being unprepared twice Xie Yongji (2): block: Add a helper to validate the block size loop: Use blk_validate_block_size() to validate block size Yang Yingliang (1): net: stmmac: fix missing unlock on error in stmmac_suspend() Yue Hu (1): erofs: remove the occupied parameter from z_erofs_pagevec_enqueue() arch/parisc/kernel/entry.S | 2 +- arch/x86/include/asm/insn-eval.h | 1 + arch/x86/include/asm/processor.h | 1 + arch/x86/kernel/process.c | 1 + arch/x86/kernel/traps.c | 34 ++++++ arch/x86/lib/insn-eval.c | 2 +- drivers/block/loop.c | 19 +-- .../net/ethernet/stmicro/stmmac/dwmac-rk.c | 9 -- drivers/net/ethernet/stmicro/stmmac/stmmac.h | 1 + .../net/ethernet/stmicro/stmmac/stmmac_main.c | 87 ++++++++++++-- .../net/ethernet/stmicro/stmmac/stmmac_mdio.c | 111 ++++++++++++++---- .../ethernet/stmicro/stmmac/stmmac_platform.c | 30 ++++- drivers/pci/msi.c | 27 +++-- drivers/pci/quirks.c | 6 + drivers/thermal/thermal_of.c | 9 +- fs/erofs/zdata.c | 15 ++- fs/erofs/zpvec.h | 14 ++- include/linux/blkdev.h | 8 ++ include/linux/pci.h | 2 + init/main.c | 1 + kernel/events/core.c | 10 +- scripts/lld-version.sh | 35 ++++-- security/Kconfig | 3 + tools/testing/selftests/x86/iopl.c | 78 ++++++++---- 24 files changed, 375 insertions(+), 131 deletions(-) -- 2.20.1

1 21

[openEuler-5.10 000/557] Backport 5.10.80 LTS
by Zheng Zengkai 06 Dec '21

06 Dec '21

Backport 5.10.80 LTS patches from upstream. Ahmad Fatoum (1): watchdog: f71808e_wdt: fix inaccurate report in WDIOC_GETTIMEOUT Ajay Singh (1): wilc1000: fix possible memory leak in cfg_scan_result() Alagu Sankar (1): ath10k: high latency fixes for beacon buffer Aleksander Jan Bajkowski (3): MIPS: lantiq: dma: add small delay after reset MIPS: lantiq: dma: reset correct number of channel MIPS: lantiq: dma: fix burst length for DEU Alex Bee (1): arm64: dts: rockchip: Fix GPU register width for RK3328 Alex Deucher (1): drm/amdgpu/gmc6: fix DMA mask from 44 to 40 bits Alex Xu (Hello71) (1): drm/plane-helper: fix uninitialized variable reference Alexander Tsoy (1): ALSA: usb-audio: Add registration quirk for JBL Quantum 400 Alexandru Ardelean (1): iio: st_sensors: disable regulators after device unregistration Alexei Starovoitov (2): bpf: Fix propagation of bounds from 64-bit min/max into 32-bit and var_off. bpf: Fix propagation of signed bounds from 64-bit min/max into 32-bit. Alexey Gladkov (1): Fix user namespace leak Alok Prasad (1): RDMA/qedr: Fix NULL deref for query_qp on the GSI QP Amelie Delaunay (3): usb: dwc2: drd: fix dwc2_force_mode call in dwc2_ovr_init usb: dwc2: drd: fix dwc2_drd_role_sw_set when clock could be disabled usb: dwc2: drd: reset current session before setting the new one Amit Engel (1): nvmet-tcp: fix header digest verification Anand Jain (1): btrfs: call btrfs_check_rw_degradable only if there is a missing device Anand Moon (2): arm64: dts: meson-g12a: Fix the pwm regulator supply properties arm64: dts: meson-g12b: Fix the pwm regulator supply properties Anant Thazhemadam (1): media: usb: dvd-usb: fix uninit-value bug in dibusb_read_eeprom_byte() Anders Roxell (1): PM: hibernate: fix sparse warnings Andrea Righi (1): selftests/bpf: Fix fclose/pclose mismatch in test_progs Andreas Gruenbacher (3): iov_iter: Fix iov_iter_get_pages{,_alloc} page fault return value gfs2: Cancel remote delete work asynchronously gfs2: Fix glock_hash_walk bugs Andreas Kemnade (1): arm: dts: omap3-gta04a4: accelerometer irq fix Andrej Shadura (2): HID: u2fzero: clarify error check and length calculations HID: u2fzero: properly handle timeouts in usb_submit_urb Andrey Grodzovsky (1): drm/amdgpu: Fix MMIO access page fault Andrii Nakryiko (6): selftests/bpf: Fix strobemeta selftest regression libbpf: Fix BTF data layout checks and allow empty BTF libbpf: Allow loading empty BTFs libbpf: Fix overflow in BTF sanity checks libbpf: Fix BTF header parsing checks selftests/bpf: Fix also no-alu32 strobemeta selftest André Almeida (1): ACPI: battery: Accept charges over the design capacity as full Andy Shevchenko (2): iio: st_sensors: Call st_sensors_power_enable() from bus drivers serial: 8250_dw: Drop wrong use of ACPI_PTR() Anel Orazgaliyeva (1): cpuidle: Fix kobject memory leaks in error paths Anson Jacob (1): drm/amd/display: dcn20_resource_construct reduce scope of FPU enabled Anssi Hannula (1): serial: xilinx_uartps: Fix race condition causing stuck TX Antoine Tenart (1): net-sysfs: try not to restart the syscall if it will fail eventually Arnaud Pouliquen (1): rpmsg: Fix rpmsg_create_ept return when RPMSG config is not defined Arnd Bergmann (9): hyperv/vmbus: include linux/bitops.h ifb: fix building without CONFIG_NET_CLS_ACT ARM: 9136/1: ARMv7-M uses BE-8, not BE-32 drm/amdgpu: fix warning for overflow check crypto: ecc - fix CRYPTO_DEFAULT_RNG dependency memstick: avoid out-of-range warning arm64: pgtable: make __pte_to_phys/__phys_to_pte_val inline functions ARM: 9156/1: drop cc-option fallbacks for architecture selection ath10k: fix invalid dma_addr_t token assignment Arun Easi (1): scsi: qla2xxx: Fix kernel crash when accessing port_speed sysfs file Asmaa Mnebhi (1): gpio: mlxbf2.c: Add check for bgpio_init failure Austin Kim (2): ALSA: synth: missing check for possible NULL after the call to kstrdup evm: mark evm_fixmode as __ro_after_init Baochen Qiang (2): ath11k: Change DMA_FROM_DEVICE to DMA_TO_DEVICE when map reinjected packets ath11k: Fix memory leak in ath11k_qmi_driver_event_work Baptiste Lepers (1): pnfs/flexfiles: Fix misplaced barrier in nfs4_ff_layout_prepare_ds Barnabás Pőcze (1): platform/x86: wmi: do not fail if disabling fails Bastien Roucariès (1): ARM: dts: sun7i: A20-olinuxino-lime2: Fix ethernet phy-mode Benjamin Li (2): wcn36xx: handle connection loss indication wcn36xx: add proper DMA memory barriers in rx path Bixuan Cui (1): powerpc/44x/fsp2: add missing of_node_put Bjorn Andersson (1): soc: qcom: rpmhpd: Make power_on actually enable the domain Borislav Petkov (1): x86/sev: Make the #VC exception stacks part of the default stacks storage Brett Creeley (1): ice: Fix not stopping Tx queues for VFs Bryan O'Donoghue (1): wcn36xx: Fix Antenna Diversity Switching Bryant Mairs (1): drm: panel-orientation-quirks: Add quirk for Aya Neo 2021 Can Guo (1): scsi: ufs: Refactor ufshcd_setup_clocks() to remove skip_ref_clk Charan Teja Reddy (1): dma-buf: WARN on dmabuf release with pending attachments Chen-Yu Tsai (2): media: rkvdec: Do not override sizeimage for output format media: rkvdec: Support dynamic resolution changes Chengfeng Ye (1): nfc: pn533: Fix double free when pn533_fill_fragment_skbs() fails Chenyuan Mi (1): drm/nouveau/svm: Fix refcount leak bug and missing check against null bug Christian Löhle (1): mmc: dw_mmc: Dont wait for DRTO on Write RSP error Christoph Hellwig (1): rds: stop using dmapool Christophe JAILLET (6): media: mtk-vpu: Fix a resource leak in the error handling path of 'mtk_vpu_probe()' mmc: mxs-mmc: disable regulator on error and in the remove function clk: mvebu: ap-cpu-clk: Fix a memory leak in error handling paths soc/tegra: Fix an error handling path in tegra_powergate_power_up() remoteproc: Fix a memory leak in an error handling path in 'rproc_handle_vdev()' i2c: xlr: Fix a resource leak in the error handling path of 'xlr_i2c_probe()' Christophe Leroy (1): video: fbdev: chipsfb: use memset_io() instead of memset() Claudio Imbrenda (2): KVM: s390: pv: avoid double free of sida page KVM: s390: pv: avoid stalls for kvm_s390_pv_init_vm Claudiu Beznea (2): clk: at91: sam9x60-pll: use DIV_ROUND_CLOSEST_ULL dmaengine: at_xdmac: fix AT_XDMAC_CC_PERID() macro Clément Léger (1): clk: at91: check pmc node status before registering syscore ops Colin Ian King (4): media: cxd2880-spi: Fix a null pointer dereference on error handling path media: cx23885: Fix snd_card_free call on null card pointer media: em28xx: Don't use ops->suspend if it is NULL mmc: moxart: Fix null pointer dereference on pointer host Corey Minyard (1): ipmi: Disable some operations during a panic Cyril Strejc (1): net: multicast: calculate csum of looped-back and forwarded packets Damien Le Moal (1): libata: fix read log timeout value Dan Carpenter (13): tpm: Check for integer overflow in tpm2_map_response_body() ath11k: fix some sleeping in atomic bugs b43legacy: fix a lower bounds test b43: fix a lower bounds test memstick: jmb38x_ms: use appropriate free function in jmb38x_ms_alloc_host() drm/msm: potential error pointer dereference in init() drm/msm: uninitialized variable in msm_gem_import() usb: gadget: hid: fix error code in do_config() scsi: csiostor: Uninitialized data in csio_ln_vnp_read_cbfn() phy: ti: gmii-sel: check of_get_address() for failure rtc: rv3032: fix error handling in rv3032_clkout_set_rate() zram: off by one in read_block_state() gve: Fix off by one in gve_tx_timeout() Dan Schatzberg (1): cgroup: Fix rootcg cpu.stat guest double counting Daniel Borkmann (2): net, neigh: Fix NTF_EXT_LEARNED in combination with NTF_USE net, neigh: Enable state migration between NUD_PERMANENT and NTF_USE Daniel Jordan (1): crypto: pcrypt - Delay write to padata->info Dave Jones (1): x86/mce: Add errata workaround for Skylake SKX37 David Hildenbrand (1): s390/gmap: don't unconditionally call pte_unmap_unlock() in __gmap_zap() Davide Baldo (1): ALSA: hda/realtek: Fixes HP Spectre x360 15-eb1xxx speakers Derong Liu (1): mmc: mtk-sd: Add wait dma stop done flow Desmond Cheong Zhi Xi (1): Bluetooth: fix init and cleanup of sco_conn.timeout_work Dinghao Liu (1): Bluetooth: btmtkuart: fix a memleak in mtk_hci_wmt_sync Dirk Bender (1): media: mt9p031: Fix corrupted frame after restarting stream Dmitriy Ulitin (1): media: stm32: Potential NULL pointer dereference in dcmi_irq_thread() Dmitry Bogdanov (1): scsi: qla2xxx: Fix unmap of already freed sgl Dmitry Osipenko (1): soc/tegra: pmc: Fix imbalanced clock disabling in error code path Dominique Martinet (1): 9p/net: fix missing error check in p9_check_errors Dongli Zhang (2): xen/netfront: stop tx queues during live migration vmxnet3: do not stop tx queues after netif_device_detach() Dongliang Mu (2): JFS: fix memleak in jfs_mount memory: fsl_ifc: fix leak of irq and nand_irq in fsl_ifc_ctrl_probe Dust Li (1): net/smc: fix sk_refcnt underflow on linkdown and fallback Eiichi Tsukata (1): vsock: prevent unnecessary refcnt inc for nonblocking connect Eric Badger (1): EDAC/sb_edac: Fix top-of-high-memory value for Broadwell/Haswell Eric Biggers (1): fscrypt: allow 256-bit master keys with AES-256-XTS Eric Dumazet (4): net: annotate data-race in neigh_output() tcp: switch orphan_count to bare per-cpu counters llc: fix out-of-bound array index in llc_sk_dev_hash() net/sched: sch_taprio: fix undefined behavior in ktime_mono_to_any Eric W. Biederman (3): signal: Remove the bogus sigkill_pending in ptrace_stop signal/mips: Update (_save|_restore)_fp_context to fail with -EFAULT signal/sh: Use force_sig(SIGKILL) instead of do_group_exit(SIGKILL) Erik Ekman (2): sfc: Export fibre-specific supported link modes sfc: Don't use netif_info before net_device setup Evgeny Novikov (3): media: atomisp: Fix error handling in probe media: dvb-frontends: mn88443x: Handle errors of clk_prepare_enable() mtd: spi-nor: hisi-sfc: Remove excessive clk_disable_unprepare() Fabio Estevam (1): ath10k: sdio: Add missing BH locking around napi_schdule() Filipe Manana (1): btrfs: fix lost error handling when replaying directory deletes Florian Westphal (3): fcnal-test: kill hanging ping/nettest binaries on cleanup vrf: run conntrack only in context of lower/physdev for locally generated packets netfilter: nfnetlink_queue: fix OOB when mac header was cleared Frank Rowand (1): of: unittest: fix EXPECT text for gpio hog errors Gao Xiang (1): erofs: don't trigger WARN() when decompression fails Geert Uytterhoeven (6): arm64: dts: renesas: beacon: Fix Ethernet PHY mode pinctrl: renesas: checker: Fix off-by-one bug in drive register check mips: cm: Convert to bitfield API to fix out-of-bounds access auxdisplay: img-ascii-lcd: Fix lock-up when displaying empty string auxdisplay: ht16k33: Connect backlight to fbdev auxdisplay: ht16k33: Fix frame buffer device blanking Giovanni Cabiddu (2): crypto: qat - detect PFVF collision after ACK crypto: qat - disregard spurious PFVF interrupts Guo Ren (1): irqchip/sifive-plic: Fixup EOI failed when masked Guru Das Srinagesh (1): firmware: qcom_scm: Fix error retval in __qcom_scm_is_call_available() Halil Pasic (1): s390/cio: make ccw_device_dma_* more robust Hannes Reinecke (1): nvme: drop scan_lock and always kick requeue list when removing namespaces Hans de Goede (6): drm: panel-orientation-quirks: Update the Lenovo Ideapad D330 quirk (v2) drm: panel-orientation-quirks: Add quirk for KD Kurio Smart C15200 2-in-1 drm: panel-orientation-quirks: Add quirk for the Samsung Galaxy Book 10.6 brcmfmac: Add DMI nvram filename quirk for Cyberbook T116 tablet power: supply: bq27xxx: Fix kernel crash on IRQ handler register error ACPI: PMIC: Fix intel_pmic_regs_handler() read accesses Hao Wu (1): tpm: fix Atmel TPM crash caused by too frequent queries Harald Freudenberger (1): s390/ap: Fix hanging ioctl caused by orphaned replies Helge Deller (4): parisc: Fix set_fixmap() on PA1.x CPUs parisc: Fix ptrace check on syscall return task_stack: Fix end_of_stack() for architectures with upwards-growing stack parisc: Fix backtrace to always include init funtion names Henrik Grimler (1): power: supply: max17042_battery: use VFSOC for capacity when no rsns Iago Toral Quiroga (1): drm/v3d: fix wait for TMU write combiner flush Ian Rogers (1): perf bpf: Add missing free to bpf_event__print_bpf_prog_info() Igor Pylypiv (1): scsi: pm80xx: Fix misleading log statement in pm8001_mpi_get_nvmd_resp() Ilya Leoshkevich (1): libbpf: Fix endianness detection in BPF_CORE_READ_BITFIELD_PROBED() Imre Deak (2): ALSA: hda: Release controller display power during shutdown/reboot ALSA: hda: Fix hang during shutdown due to link reset Ingmar Klein (1): PCI: Mark Atheros QCA6174 to avoid bus reset Israel Rukshin (3): nvmet: fix use-after-free when a port is removed nvmet-rdma: fix use-after-free when a port is removed nvmet-tcp: fix use-after-free when a port is removed J. Bruce Fields (1): nfsd: don't alloc under spinlock in rpc_parse_scope_id Jack Andersen (1): mfd: dln2: Add cell for initializing DLN2 ADC Jackie Liu (3): ARM: s3c: irq-s3c24xx: Fix return value check for s3c24xx_init_intc() MIPS: loongson64: make CPU_LOONGSON64 depends on MIPS_FP_SUPPORT ar7: fix kernel builds for compiler test Jaegeuk Kim (1): f2fs: should use GFP_NOFS for directory inodes Jakob Hauser (1): power: supply: rt5033_battery: Change voltage values to µV Jakub Kicinski (4): net: sched: update default qdisc visibility after Tx queue cnt changes net: stream: don't purge sk_error_queue in sk_stream_kill_queues() udp6: allow SO_MARK ctrl msg to affect routing ethtool: fix ethtool msg len calculation for pause stats Jan Kara (1): ocfs2: fix data corruption on truncate Jane Malalane (1): x86/cpu: Fix migration safety with X86_BUG_NULL_SEL Janghyub Seo (1): r8169: Add device 10ec:8162 to driver r8169 Janis Schoetterl-Glausch (1): KVM: s390: Fix handle_sske page fault handling Jaroslav Kysela (1): ALSA: hda/realtek: Add a quirk for Acer Spin SP513-54N Jason Ormes (1): ALSA: usb-audio: Line6 HX-Stomp XL USB_ID for 48k-fixed quirk Jens Axboe (2): block: bump max plugged deferred size from 16 to 32 block: remove inaccurate requeue check Jeremy Soller (1): ALSA: hda/realtek: Headset fixup for Clevo NH77HJQ Jernej Skrabec (1): drm/sun4i: Fix macros in sun8i_csc.h Jessica Zhang (1): drm/msm: Fix potential NULL dereference in DPU SSPP Jia-Ju Bai (1): fs: orangefs: fix error return code of orangefs_revalidate_lookup() Jiasheng Jiang (1): rxrpc: Fix _usecs_to_jiffies() by using usecs_to_jiffies() Jim Mattson (1): KVM: selftests: Fix nested SVM tests when built with clang Jiri Olsa (1): selftests/bpf: Fix perf_buffer test on system with offline cpus Joerg Roedel (1): x86/sev: Fix stack type check in vc_switch_off_ist() Johan Hovold (14): Input: iforce - fix control-message timeout ALSA: ua101: fix division by zero at probe ALSA: 6fire: fix control and bulk message timeouts ALSA: line6: fix control and interrupt message timeouts mwifiex: fix division by zero in fw download path ath6kl: fix division by zero in send path ath6kl: fix control-message timeout ath10k: fix control-message timeout ath10k: fix division by zero in send path rtl8187: fix control-message timeouts serial: 8250: fix racy uartclk update most: fix control-message timeouts USB: iowarrior: fix control-message timeouts USB: chipidea: fix interrupt deadlock Johannes Berg (1): iwlwifi: mvm: disable RX-diversity in powersave John Fastabend (2): bpf, sockmap: Remove unhash handler for BPF sockmap usage bpf: sockmap, strparser, and tls are reusing qdisc_skb_cb and colliding John Fraker (1): gve: Recover from queue stall due to missed IRQ Johnathon Clark (1): ALSA: hda/realtek: Fix mic mute LED for the HP Spectre x360 14 Jon Maxwell (1): tcp: don't free a FIN sk_buff in tcp_remove_empty_skb() Jonas Dreßler (5): mwifiex: Read a PCI register after writing the TX ring write pointer mwifiex: Try waking the firmware until we get an interrupt mwifiex: Run SET_BSS_MODE when changing from P2P to STATION vif-type mwifiex: Properly initialize private structure on interface type changes mwifiex: Send DELBA requests according to spec Josef Bacik (1): btrfs: do not take the uuid_mutex in btrfs_rm_device Josh Don (1): fs/proc/uptime.c: Fix idle time reporting in /proc/uptime Josh Poimboeuf (1): objtool: Add xen_start_kernel() to noreturn list Juergen Gross (1): xen/balloon: add late_initcall_sync() for initial ballooning done Junji Wei (1): RDMA/rxe: Fix wrong port_cap_flags Kai-Heng Feng (1): ALSA: hda/realtek: Add quirk for HP EliteBook 840 G7 mute LED Kalesh Singh (1): tracing/cfi: Fix cmp_entries_* functions signature mismatch Kan Liang (2): perf/x86/intel/uncore: Support extra IMC channel on Ice Lake server perf/x86/intel/uncore: Fix Intel ICX IIO event constraints Kees Cook (5): leaking_addresses: Always print a trailing newline media: radio-wl1273: Avoid card name truncation media: si470x: Avoid card name truncation media: tm6000: Avoid card name truncation clocksource/drivers/timer-ti-dm: Select TIMER_OF Kewei Xu (1): i2c: mediatek: fixing the incorrect register offset Kishon Vijay Abraham I (2): arm64: dts: ti: k3-j721e-main: Fix "max-virtual-functions" in PCIe EP nodes arm64: dts: ti: k3-j721e-main: Fix "bus-range" upto 256 bus number for PCIe Krzysztof Kozlowski (3): regulator: s5m8767: do not use reset value as DVS voltage if GPIO DVS is disabled regulator: dt-bindings: samsung,s5m8767: correct s5m8767,pmic-buck-default-dvs-idx property mfd: core: Add missing of_node_put for loop iteration Kumar Kartikeya Dwivedi (1): selftests/bpf: Fix fd cleanup in sk_lookup test Kunihiko Hayashi (1): PCI: uniphier: Serialize INTx masking/unmasking and fix the bit operation Lad Prabhakar (1): spi: spi-rpc-if: Check return value of rpcif_sw_init() Lars-Peter Clausen (1): dmaengine: dmaengine_desc_callback_valid(): Check for `callback_result` Lasse Collin (2): lib/xz: Avoid overlapping memcpy() with invalid input with in-place decompression lib/xz: Validate the value before assigning it to an enum variable Lee Jones (1): soc: qcom: rpmhpd: Provide some missing struct member descriptions Leon Romanovsky (1): RDMA/mlx4: Return missed an error if device doesn't support steering Li Chen (1): PCI: cadence: Add cdns_plat_pcie_probe() missing return Li Zhang (1): btrfs: clear MISSING device status bit in btrfs_close_one_device Linus Lüssing (1): ath9k: Fix potential interrupt storm on queue reset Linus Walleij (1): net: dsa: rtl8366rb: Fix off-by-one bug Loic Poulain (6): wcn36xx: Fix HT40 capability for 2Ghz band wcn36xx: Fix tx_status mechanism wcn36xx: Fix (QoS) null data frame bitrate/modulation wcn36xx: Correct band/freq reporting on RX ath10k: Fix missing frame timestamp for beacon/probe-resp wcn36xx: Fix discarded frames due to wrong sequence number Lorenz Bauer (3): bpf: Define bpf_jit_alloc_exec_limit for arm64 JIT bpf: Prevent increasing bpf_jit_limit above max selftests: bpf: Convert sk_lookup ctx access tests to PROG_TEST_RUN Lorenzo Bianconi (3): mt76: mt7615: fix endianness warning in mt7615_mac_write_txwi mt76: mt76x02: fix endianness warnings in mt76x02_mac.c mt76: mt7915: fix possible infinite loop release semaphore Lucas Tanure (1): ASoC: cs42l42: Disable regulators if probe fails Lukas Wunner (1): ifb: Depend on netfilter alternatively to tc Maciej W. Rozycki (1): MIPS: Fix assembly error from MIPSr2 code used within MIPS_ISA_ARCH_LEVEL Marc Kleine-Budde (1): can: mcp251xfd: mcp251xfd_chip_start(): fix error handling for mcp251xfd_chip_rx_int_enable() Marek Behún (4): PCI: pci-bridge-emul: Fix emulation of W1C bits PCI: aardvark: Fix return value of MSI domain .alloc() method PCI: aardvark: Read all 16-bits from PCIE_MSI_PAYLOAD_REG PCI: aardvark: Don't spam about PIO Response Status Marek Vasut (3): rsi: Fix module dev_oper_mode parameter description ARM: dts: stm32: Reduce DHCOR SPI NOR frequency to 50 MHz video: backlight: Drop maximum brightness override for brightness zero Marijn Suijten (1): ARM: dts: qcom: msm8974: Add xo_board reference clock to DSI0 PHY Mario (1): drm: panel-orientation-quirks: Add quirk for GPD Win3 Mark Brown (1): tpm_tis_spi: Add missing SPI ID Mark Rutland (2): KVM: arm64: Extract ESR_ELx.EC only irq: mips: avoid nested irq_enter() Markus Schneider-Pargmann (1): hwrng: mtk - Force runtime pm ops for sleep ops Martin Fuzzey (3): rsi: fix occasional initialisation failure with BT coex rsi: fix key enabled check causing unwanted encryption for vap_id > 0 rsi: fix rate mask set leading to P2P failure Martin Kepplinger (1): media: imx: set a media_device bus_info string Masami Hiramatsu (2): ia64: kprobes: Fix to pass correct trampoline address to the handler ARM: clang: Do not rely on lr register for stacktrace Mathias Nyman (1): xhci: Fix USB 3.1 enumeration issues by increasing roothub power-on-good delay Matthew Auld (1): drm/ttm: stop calling tt_swapin in vm_access Matthias Schiffer (1): net: phy: micrel: make *-skew-ps check more lenient Maurizio Lombardi (1): nvmet-tcp: fix a memory leak when releasing a queue Max Gurtovoy (1): nvme-rdma: fix error code in nvme_rdma_setup_ctrl Maxim Kiselev (1): net: davinci_emac: Fix interrupt pacing disable Meng Li (2): soc: fsl: dpio: replace smp_processor_id with raw_smp_processor_id soc: fsl: dpio: use the combined functions to protect critical zone Menglong Dong (1): workqueue: make sysfs of unbound kworker cpumask more clever Miaohe Lin (1): mm/zsmalloc.c: close race window between zs_pool_dec_isolated() and zs_unregister_migration() Michael Ellerman (1): powerpc: Fix is_kvm_guest() / kvm_para_available() Michael Pratt (1): posix-cpu-timers: Clear task::posix_cputimers_work in copy_process() Michael Tretter (1): media: allegro: ignore interrupt if mailbox is not initialized Michael Walle (1): crypto: caam - disable pkc for non-E SoCs Michal Hocko (1): mm, oom: do not trigger out_of_memory from the #PF Michał Mirosław (1): ARM: 9155/1: fix early early_iounmap() Miklos Szeredi (1): fuse: fix page stealing Miquel Raynal (9): mtd: rawnand: socrates: Keep the driver compatible with on-die ECC engines mtd: rawnand: ams-delta: Keep the driver compatible with on-die ECC engines mtd: rawnand: xway: Keep the driver compatible with on-die ECC engines mtd: rawnand: mpc5121: Keep the driver compatible with on-die ECC engines mtd: rawnand: gpio: Keep the driver compatible with on-die ECC engines mtd: rawnand: pasemi: Keep the driver compatible with on-die ECC engines mtd: rawnand: orion: Keep the driver compatible with on-die ECC engines mtd: rawnand: plat_nand: Keep the driver compatible with on-die ECC engines mtd: rawnand: au1550nd: Keep the driver compatible with on-die ECC engines Muchun Song (1): seq_file: fix passing wrong private data Nadezda Lutovinova (2): media: s5p-mfc: Add checking to s5p_mfc_probe(). media: rcar-csi2: Add checking to rcsi2_start_receiver() Naohiro Aota (1): block: schedule queue restart after BLK_STS_ZONE_RESOURCE Nathan Chancellor (1): platform/x86: thinkpad_acpi: Fix bitwise vs. logical warning Nathan Lynch (1): powerpc: fix unbalanced node refcount in check_kvm_guest() Naveen N. Rao (4): powerpc/lib: Add helper to check if offset is within conditional branch range powerpc/bpf: Validate branch ranges powerpc/security: Add a helper to query stf_barrier type powerpc/bpf: Emit stf barrier instruction sequences for BPF_NOSPEC Neeraj Upadhyay (1): rcu: Fix existing exp request check in sync_sched_exp_online_cleanup() Nehal Bakulchandra Shah (1): usb: xhci: Enable runtime-pm by default on AMD Yellow Carp platform Nikita Yushchenko (1): staging: most: dim2: do not double-register the same device Nuno Sá (2): iio: ad5770r: make devicetree property reading consistent iio: adis: do not disabe IRQs in 'adis_init()' Olivier Moysan (2): ARM: dts: stm32: fix SAI sub nodes register range ARM: dts: stm32: fix AV96 board SAI2 pin muxing on stm32mp15 Ondrej Mosnacek (1): selinux: fix race condition when computing ocontext SIDs Pablo Neira Ayuso (2): netfilter: conntrack: set on IPS_ASSURED if flows enters internal stream state netfilter: nft_dynset: relax superfluous check on set updates Pali Rohár (13): serial: core: Fix initializing and restoring termios speed PCI: aardvark: Do not clear status bits of masked interrupts PCI: aardvark: Fix checking for link up via LTSSM state PCI: aardvark: Do not unmask unused interrupts PCI: aardvark: Fix reporting Data Link Layer Link Active PCI: aardvark: Fix configuring Reference clock PCI: aardvark: Fix support for bus mastering and PCI_COMMAND on emulated bridge PCI: aardvark: Fix support for PCI_BRIDGE_CTL_BUS_RESET on emulated bridge PCI: aardvark: Set PCI Bridge Class Code to PCI Bridge PCI: aardvark: Fix support for PCI_ROM_ADDRESS1 on emulated bridge PCI: aardvark: Fix preserving PCI_EXP_RTCTL_CRSSVE flag on emulated bridge PCI: Add PCI_EXP_DEVCTL_PAYLOAD_* macros PCI: aardvark: Fix PCIe Max Payload Size setting Paul E. McKenney (1): rcu-tasks: Move RTGS_WAIT_CBS to beginning of rcu_tasks_kthread() loop Pavel Skripkin (3): ALSA: mixer: fix deadlock in snd_mixer_oss_set_volume media: em28xx: add missing em28xx_close_extension media: dvb-usb: fix ununit-value in az6027_rc_query Pawan Gupta (1): smackfs: Fix use-after-free in netlbl_catmap_walk() Paweł Anikiel (1): reset: socfpga: add empty driver allowing consumers to probe Pekka Korpinen (1): iio: dac: ad5446: Fix ad5622_write() return value Peter Rosin (1): ARM: dts: at91: tse850: the emac<->phy interface is rmii Peter Zijlstra (5): locking/lockdep: Avoid RCU-induced noinstr fail x86: Increase exception stack sizes x86/xen: Mark cpu_bringup_and_idle() as dead_end_function objtool: Fix static_call list generation rcu: Always inline rcu_dynticks_task*_{enter,exit}() Phoenix Huang (1): Input: elantench - fix misreporting trackpoint coordinates Pradeep Kumar Chitrapu (1): ath11k: fix packet drops due to incorrect 6 GHz freq value in rx status Punit Agrawal (1): kprobes: Do not use local variable when creating debugfs file Quentin Monnet (1): bpftool: Avoid leaking the JSON writer prepared for program metadata Quinn Tran (4): scsi: qla2xxx: Fix use after free in eh_abort path scsi: qla2xxx: Relogin during fabric disturbance scsi: qla2xxx: Fix gnl list corruption scsi: qla2xxx: Turn off target reset during issue_lip Rafael J. Wysocki (2): PM: sleep: Do not let "syscore" devices runtime-suspend during system transitions ACPICA: Avoid evaluating methods too early during system resume Rafał Miłecki (1): ARM: dts: BCM5301X: Fix memory nodes names Rahul Lakkireddy (1): cxgb4: fix eeprom len when diagnostics not implemented Rahul Tanwar (1): pinctrl: equilibrium: Fix function addition in multiple groups Rajat Asthana (1): media: mceusb: return without resubmitting URB in case of -EPROTO error. Randy Dunlap (5): mmc: winbond: don't build on M68K ia64: don't do IA64_CMPXCHG_DEBUG without CONFIG_PRINTK usb: musb: select GENERIC_PHY instead of depending on it usb: typec: STUSB160X should select REGMAP_I2C m68k: set a default value for MEMORY_RESERVE Ranjani Sridharan (1): ASoC: SOF: topology: do not power down primary core during topology removal Reimar Döffinger (1): libata: fix checking of DMA state Ricardo Koller (1): KVM: selftests: Add operand to vmsave/vmload/vmrun in svm.c Ricardo Ribalda (7): media: v4l2-ioctl: Fix check_ext_ctrls media: uvcvideo: Set capability in s_param media: uvcvideo: Return -EIO for control errors media: uvcvideo: Set unique vdev name based in type media: ipu3-imgu: imgu_fmt: Handle properly try media: ipu3-imgu: VIDIOC_QUERYCAP: Fix bus_info media: v4l2-ioctl: S_CTRL output the right value Richard Fitzgerald (4): ASoC: cs42l42: Correct some register default values ASoC: cs42l42: Defer probe if request_threaded_irq() returns EPROBE_DEFER ASoC: cs42l42: Use device_property API instead of of_property ASoC: cs42l42: Correct configuring of switch inversion from ts-inv Robert-Ionut Alexa (1): soc: fsl: dpaa2-console: free buffer before returning from dpaa2_console_read Russell King (Oracle) (1): net: phylink: avoid mvneta warning when setting pause parameters Ryder Lee (1): mt76: mt7915: fix an off-by-one bound check Sandeep Maheswaram (1): phy: qcom-snps: Correct the FSEL_MASK Saurav Kashyap (1): scsi: qla2xxx: Changes to support FCP2 Target Scott Wood (1): rcutorture: Avoid problematic critical section nesting on PREEMPT_RT Sean Christopherson (3): x86/irq: Ensure PI wakeup handler is unregistered before module unload KVM: VMX: Unregister posted interrupt wakeup handler on hardware unsetup KVM: nVMX: Query current VMCS when determining if MSR bitmaps are in use Sean Young (3): media: ite-cir: IR receiver stop working after receive overflow media: ir-kbd-i2c: improve responsiveness of hauppauge zilog receivers media: ir_toy: assignment to be16 should be of correct type Sebastian Andrzej Siewior (1): lockdep: Let lock_is_held_type() detect recursive read as read Sebastian Krzyszkowiak (2): power: supply: max17042_battery: Prevent int underflow in set_soc_threshold power: supply: max17042_battery: Clear status bits in interrupt handler Seevalamuthu Mariappan (1): ath11k: Align bss_chan_info structure with firmware Selvin Xavier (1): RDMA/bnxt_re: Fix query SRQ failure Shaoying Xu (1): ext4: fix lazy initialization next schedule time computation in more granular unit Shayne Chen (2): mt76: mt7915: fix sta_rec_wtbl tag len mt76: mt7915: fix muar_idx in mt7915_mcu_alloc_sta_req() Shuah Khan (2): selftests: kvm: fix mismatched fclose() after popen() selftests/core: fix conflicting types compile error for close_range() Shyam Sundar S K (1): net: amd-xgbe: Toggle PLL settings during rate change Sidong Yang (1): btrfs: reflink: initialize return value to 0 in btrfs_extent_same() Simon Ser (1): drm/panel-orientation-quirks: add Valve Steam Deck Srikar Dronamraju (3): powerpc: Refactor is_kvm_guest() declaration to new header powerpc: Rename is_kvm_guest() to check_kvm_guest() powerpc: Reintroduce is_kvm_guest() as a fast-path check Srinivas Kandagatla (2): soundwire: debugfs: use controller id and link_id for debugfs scsi: ufs: ufshcd-pltfrm: Fix memory leak due to probe defer Sriram R (2): ath11k: Avoid reg rules update during firmware recovery ath11k: Avoid race during regd updates Stafford Horne (1): openrisc: fix SMP tlb flush NULL pointer dereference Stefan Agner (2): phy: micrel: ksz8041nl: do not use power down mode serial: imx: fix detach/attach of serial console Stefan Schaeckeler (1): ACPI: AC: Quirk GK45 to skip reading _PSR Stephan Gerhold (2): arm64: dts: qcom: msm8916: Fix Secondary MI2S bit clock arm64: dts: qcom: pm8916: Remove wrong reg-names for rtc@6000 Stephen Suryaputra (1): gre/sit: Don't generate link-local addr if addr_gen_mode is IN6_ADDR_GEN_MODE_NONE Steven Rostedt (VMware) (2): ring-buffer: Protect ring_buffer_reset() from reentrancy tracefs: Have tracefs directories not set OTH permission bits by default Sudarshan Rajagopalan (1): arm64: mm: update max_pfn after memory hotplug Sukadev Bhattiprolu (2): ibmvnic: don't stop queue in xmit ibmvnic: Process crqs after enabling interrupts Sungjong Seo (1): exfat: fix incorrect loading of i_blocks for large files Sven Eckelmann (1): ath10k: fix max antenna gain unit Sven Schnelle (4): parisc: fix warning in flush_tlb_all parisc/unwind: fix unwinder when CONFIG_64BIT is enabled parisc/kgdb: add kgdb_roundup() to make kgdb work with idle polling s390/tape: fix timer initialization in tape_std_assign() Sylwester Dziedziuch (1): ice: Fix replacing VF hardware MAC to existing MAC filter Tadeusz Struk (1): scsi: core: Remove command size deduction from scsi_setup_scsi_cmnd() Takashi Iwai (8): Input: i8042 - Add quirk for Fujitsu Lifebook T725 ALSA: hda/realtek: Add a quirk for HP OMEN 15 mute LED ALSA: hda/realtek: Add quirk for ASUS UX550VE ALSA: hda: Free card instance properly at probe errors ALSA: timer: Unconditionally unlink slave instances, too ALSA: mixer: oss: Fix racy access to slots ALSA: hda: Reduce udelay() at SKL+ position reporting ALSA: hda: Use position buffer for SKL+ again Tang Bin (1): crypto: s5p-sss - Add error handling in s5p_aes_probe() Tao Zhang (1): coresight: cti: Correct the parameter for pm_runtime_put Tetsuo Handa (2): smackfs: use __GFP_NOFAIL for smk_cipso_doi() smackfs: use netlbl_cfg_cipsov4_del() for deleting cipso_v4_doi Thomas Perrot (1): spi: spl022: fix Microwire full duplex mode Tiezhu Yang (1): samples/kretprobes: Fix return value if register_kretprobe() failed Tim Crawford (1): ALSA: hda/realtek: Add quirk for Clevo PC70HS Tim Gardner (2): drm/msm: prevent NULL dereference in msm_gpu_crashstate_capture() net: enetc: unmap DMA in enetc_send_cmd() Todd Kjos (3): binder: use euid from cred instead of using task binder: use cred instead of task for selinux checks binder: use cred instead of task for getsecid Tom Lendacky (3): x86/sme: Use #define USE_EARLY_PGTABLE_L5 in mem_encrypt_identity.c arch/cc: Introduce a function to check for confidential computing features x86/sev: Add an x86 version of cc_platform_has() Tom Rix (2): media: TDA1997x: handle short reads of hdmi info frame. apparmor: fix error check Tong Zhang (1): scsi: dc395: Fix error case unwinding Tony Lindgren (3): mmc: sdhci-omap: Fix NULL pointer exception if regulator is not configured mmc: sdhci-omap: Fix context restore bus: ti-sysc: Fix timekeeping_suspended warning on resume Tony Lu (1): net/smc: Fix smc_link->llc_testlink_time overflow Trond Myklebust (6): NFS: Fix dentry verifier races NFS: Fix deadlocks in nfs_scan_commit_list() NFS: Fix up commit deadlocks NFS: Fix an Oops in pnfs_mark_request_commit() NFSv4: Fix a regression in nfs_set_open_stateid_locked() SUNRPC: Partial revert of commit 6f9f17287e78 Tuo Li (2): media: s5p-mfc: fix possible null-pointer dereference in s5p_mfc_probe() ath: dfs_pattern_detector: Fix possible null-pointer dereference in channel_detector_create() Vasant Hegde (1): powerpc/powernv/prd: Unregister OPAL_MSG_PRD2 notifier during module unload Vasily Averin (2): memcg: prohibit unconditional exceeding the limit of dying tasks mm, oom: pagefault_out_of_memory: don't force global OOM for dying tasks Vegard Nossum (1): staging: ks7010: select CRYPTO_HASH/CRYPTO_MICHAEL_MIC Vincent Donnefort (1): PM: EM: Fix inefficient states detection Vineeth Vijayan (1): s390/cio: check the subchannel validity for dev_busid Vitaly Kuznetsov (1): x86/hyperv: Protect set_hv_tscchange_cb() against getting preempted Vladimir Oltean (1): net: stmmac: allow a tc-taprio base-time of zero Vladimir Zapolskiy (1): phy: qcom-qusb2: Fix a memory leak on probe Waiman Long (1): cgroup: Make rebind_subsystems() disable v2 controllers all at once Walter Stoll (1): watchdog: Fix OMAP watchdog early handling Wan Jiabing (1): soc: qcom: apr: Add of_node_put() before return Wang Hai (3): USB: serial: keyspan: fix memleak on probe errors libertas_tf: Fix possible memory leak in probe and disconnect libertas: Fix possible memory leak in probe and disconnect Wen Gong (1): ath11k: add handler for scan event WMI_SCAN_EVENT_DEQUEUED Wen Gu (1): net/smc: Correct spelling mistake to TCPF_SYN_RECV Willem de Bruijn (1): selftests/net: udpgso_bench_rx: fix port argument Wolfram Sang (1): memory: renesas-rpc-if: Correct QSPI data transfer in Manual mode Xiao Ni (1): md: update superblock after changing rdev flags in state_store Xiaoming Ni (2): powerpc/85xx: Fix oops when mpc85xx_smp_guts_ids node cannot be found powerpc/85xx: fix timebase sync issue when CONFIG_HOTPLUG_CPU=n Xin Xiong (1): mmc: moxart: Fix reference count leaks in moxart_probe Xuan Zhuo (1): virtio_ring: check desc == NULL when using indirect with packed Yajun Deng (1): net: net_namespace: Fix undefined member in key_remove_domain() Yang Yingliang (6): ASoC: soc-core: fix null-ptr-deref in snd_soc_del_component_unlocked() pinctrl: core: fix possible memory leak in pinctrl_enable() spi: bcm-qspi: Fix missing clk_disable_unprepare() on error in bcm_qspi_probe() hwmon: Fix possible memleak in __hwmon_device_register() driver core: Fix possible memory leak in device_link_add() power: supply: max17040: fix null-ptr-deref in max17040_probe() Yazen Ghannam (1): EDAC/amd64: Handle three rank interleaving mode Yifan Zhang (1): drm/amdkfd: fix resume error when iommu disabled in Picasso Yu Xiao (1): nfp: bpf: relax prog rejection for mtu check through max_pkt_offset YueHaibing (2): opp: Fix return in _opp_add_static_v2() xen-pciback: Fix return in pm_ctrl_init() Zev Weiss (3): hwmon: (pmbus/lm25066) Add offset coefficients hwmon: (pmbus/lm25066) Let compiler determine outer dimension of lm25066_coeff mtd: core: don't remove debugfs directory if device is in use Zhang Changzhong (2): can: j1939: j1939_tp_cmd_recv(): ignore abort message in the BAM transport can: j1939: j1939_can_recv(): ignore messages with invalid source address Zhang Qiao (1): kernel/sched: Fix sched_fork() access an invalid sched_task_group Zheyu Ma (7): cavium: Return negative value when pci_alloc_irq_vectors() fails scsi: qla2xxx: Return -ENOMEM if kzalloc() fails mISDN: Fix return values of the probe function cavium: Fix return values of the probe function media: netup_unidvb: handle interrupt properly according to the firmware memstick: r592: Fix a UAF bug when removing the driver mwl8k: Fix use-after-free in mwl8k_fw_state_machine() Ziyang Xuan (2): rsi: stop thread firstly in rsi_91x_init() error handling net: vlan: fix a UAF in vlan_dev_real_dev() Zong-Zhe Yang (1): rtw88: fix RX clock gate setting while fifo dump jing yangyang (1): firmware/psci: fix application of sizeof to pointer liuyuntao (1): virtio-gpu: fix possible memory allocation failure .../admin-guide/kernel-parameters.txt | 7 + .../bindings/regulator/samsung,s5m8767.txt | 23 +- Documentation/filesystems/fscrypt.rst | 10 +- arch/Kconfig | 3 + arch/arm/Makefile | 22 +- arch/arm/boot/dts/at91-tse850-3.dts | 2 +- arch/arm/boot/dts/bcm4708-netgear-r6250.dts | 2 +- arch/arm/boot/dts/bcm4709-asus-rt-ac87u.dts | 2 +- .../boot/dts/bcm4709-buffalo-wxr-1900dhp.dts | 2 +- arch/arm/boot/dts/bcm4709-linksys-ea9200.dts | 2 +- arch/arm/boot/dts/bcm4709-netgear-r7000.dts | 2 +- arch/arm/boot/dts/bcm4709-netgear-r8000.dts | 2 +- .../boot/dts/bcm4709-tplink-archer-c9-v1.dts | 2 +- arch/arm/boot/dts/bcm47094-luxul-xwc-2000.dts | 2 +- arch/arm/boot/dts/bcm53016-meraki-mr32.dts | 2 +- arch/arm/boot/dts/bcm94708.dts | 2 +- arch/arm/boot/dts/bcm94709.dts | 2 +- arch/arm/boot/dts/omap3-gta04.dtsi | 2 +- arch/arm/boot/dts/qcom-msm8974.dtsi | 4 +- arch/arm/boot/dts/stm32mp15-pinctrl.dtsi | 8 +- arch/arm/boot/dts/stm32mp151.dtsi | 16 +- arch/arm/boot/dts/stm32mp15xx-dhcor-som.dtsi | 2 +- .../boot/dts/sun7i-a20-olinuxino-lime2.dts | 2 +- arch/arm/kernel/stacktrace.c | 3 +- arch/arm/mach-s3c/irq-s3c24xx.c | 22 +- arch/arm/mm/Kconfig | 2 +- arch/arm/mm/mmu.c | 4 +- .../boot/dts/amlogic/meson-g12a-sei510.dts | 2 +- .../boot/dts/amlogic/meson-g12a-u200.dts | 2 +- .../boot/dts/amlogic/meson-g12a-x96-max.dts | 2 +- .../dts/amlogic/meson-g12b-khadas-vim3.dtsi | 4 +- .../dts/amlogic/meson-g12b-odroid-n2.dtsi | 4 +- .../boot/dts/amlogic/meson-g12b-w400.dtsi | 4 +- arch/arm64/boot/dts/qcom/msm8916.dtsi | 8 +- arch/arm64/boot/dts/qcom/pm8916.dtsi | 1 - .../boot/dts/renesas/beacon-renesom-som.dtsi | 1 + arch/arm64/boot/dts/rockchip/rk3328.dtsi | 2 +- arch/arm64/boot/dts/ti/k3-j721e-main.dtsi | 16 +- arch/arm64/include/asm/esr.h | 1 + arch/arm64/include/asm/pgtable.h | 12 +- arch/arm64/kvm/hyp/hyp-entry.S | 2 +- arch/arm64/kvm/hyp/nvhe/host.S | 2 +- arch/arm64/mm/mmu.c | 5 + arch/arm64/net/bpf_jit_comp.c | 5 + arch/ia64/Kconfig.debug | 2 +- arch/ia64/kernel/kprobes.c | 9 +- arch/m68k/Kconfig.machine | 1 + arch/mips/Kconfig | 1 + arch/mips/include/asm/cmpxchg.h | 5 +- arch/mips/include/asm/mips-cm.h | 12 +- arch/mips/kernel/mips-cm.c | 21 +- arch/mips/kernel/r2300_fpu.S | 4 +- arch/mips/kernel/syscall.c | 9 - arch/mips/lantiq/xway/dma.c | 23 +- arch/openrisc/kernel/dma.c | 4 +- arch/openrisc/kernel/smp.c | 6 +- arch/parisc/kernel/entry.S | 2 +- arch/parisc/kernel/smp.c | 19 +- arch/parisc/kernel/unwind.c | 21 +- arch/parisc/kernel/vmlinux.lds.S | 3 +- arch/parisc/mm/fixmap.c | 5 +- arch/parisc/mm/init.c | 4 +- arch/powerpc/include/asm/code-patching.h | 1 + arch/powerpc/include/asm/firmware.h | 6 - arch/powerpc/include/asm/kvm_guest.h | 25 ++ arch/powerpc/include/asm/kvm_para.h | 2 +- arch/powerpc/include/asm/security_features.h | 5 + arch/powerpc/kernel/firmware.c | 12 +- arch/powerpc/kernel/security.c | 5 + arch/powerpc/lib/code-patching.c | 7 +- arch/powerpc/net/bpf_jit.h | 33 ++- arch/powerpc/net/bpf_jit64.h | 8 +- arch/powerpc/net/bpf_jit_comp64.c | 64 ++++- arch/powerpc/platforms/44x/fsp2.c | 2 + arch/powerpc/platforms/85xx/Makefile | 4 +- arch/powerpc/platforms/85xx/mpc85xx_pm_ops.c | 7 +- arch/powerpc/platforms/85xx/smp.c | 12 +- arch/powerpc/platforms/powernv/opal-prd.c | 12 +- arch/powerpc/platforms/pseries/smp.c | 3 + arch/s390/kvm/priv.c | 2 + arch/s390/kvm/pv.c | 21 +- arch/s390/mm/gmap.c | 5 +- arch/sh/kernel/cpu/fpu.c | 10 +- arch/x86/Kconfig | 1 + arch/x86/events/intel/uncore_snbep.c | 6 +- arch/x86/hyperv/hv_init.c | 5 +- arch/x86/include/asm/cpu_entry_area.h | 8 +- arch/x86/include/asm/mem_encrypt.h | 1 + arch/x86/include/asm/page_64_types.h | 2 +- arch/x86/kernel/Makefile | 6 + arch/x86/kernel/cc_platform.c | 69 +++++ arch/x86/kernel/cpu/amd.c | 2 + arch/x86/kernel/cpu/common.c | 44 ++- arch/x86/kernel/cpu/cpu.h | 1 + arch/x86/kernel/cpu/hygon.c | 2 + arch/x86/kernel/cpu/mce/intel.c | 5 +- arch/x86/kernel/irq.c | 4 +- arch/x86/kernel/sev-es.c | 32 --- arch/x86/kernel/traps.c | 2 +- arch/x86/kvm/vmx/vmx.c | 15 +- arch/x86/mm/cpu_entry_area.c | 7 + arch/x86/mm/mem_encrypt.c | 1 + arch/x86/mm/mem_encrypt_identity.c | 9 + block/blk-mq.c | 18 +- block/blk.h | 6 + crypto/Kconfig | 2 +- crypto/pcrypt.c | 12 +- drivers/acpi/ac.c | 19 ++ drivers/acpi/acpica/acglobal.h | 2 + drivers/acpi/acpica/hwesleep.c | 8 +- drivers/acpi/acpica/hwsleep.c | 11 +- drivers/acpi/acpica/hwxfsleep.c | 7 + drivers/acpi/battery.c | 2 +- drivers/acpi/pmic/intel_pmic.c | 51 ++-- drivers/android/binder.c | 22 +- drivers/ata/libata-core.c | 2 +- drivers/ata/libata-eh.c | 8 + drivers/auxdisplay/ht16k33.c | 66 +++-- drivers/auxdisplay/img-ascii-lcd.c | 10 + drivers/base/core.c | 4 +- drivers/base/power/main.c | 9 +- drivers/block/zram/zram_drv.c | 2 +- drivers/bluetooth/btmtkuart.c | 13 +- drivers/bus/ti-sysc.c | 65 ++++- drivers/char/hw_random/mtk-rng.c | 9 +- drivers/char/ipmi/ipmi_msghandler.c | 10 +- drivers/char/ipmi/ipmi_watchdog.c | 17 +- drivers/char/tpm/tpm2-space.c | 3 + drivers/char/tpm/tpm_tis_core.c | 26 +- drivers/char/tpm/tpm_tis_core.h | 4 + drivers/char/tpm/tpm_tis_spi_main.c | 1 + drivers/clk/at91/clk-sam9x60-pll.c | 4 +- drivers/clk/at91/pmc.c | 5 + drivers/clk/mvebu/ap-cpu-clk.c | 14 +- drivers/clocksource/Kconfig | 1 + drivers/cpuidle/sysfs.c | 5 +- drivers/crypto/caam/caampkc.c | 19 +- drivers/crypto/caam/regs.h | 3 + drivers/crypto/qat/qat_common/adf_pf2vf_msg.c | 13 + drivers/crypto/qat/qat_common/adf_vf_isr.c | 6 + drivers/crypto/s5p-sss.c | 2 + drivers/dma-buf/dma-buf.c | 1 + drivers/dma/at_xdmac.c | 2 +- drivers/dma/dmaengine.h | 2 +- drivers/edac/amd64_edac.c | 22 +- drivers/edac/sb_edac.c | 2 +- drivers/firmware/psci/psci_checker.c | 2 +- drivers/firmware/qcom_scm.c | 2 +- drivers/gpio/gpio-mlxbf2.c | 5 + drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.h | 2 +- drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c | 4 +- drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c | 8 +- drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c | 17 +- drivers/gpu/drm/amd/amdkfd/kfd_device.c | 1 + .../drm/amd/display/dc/dcn20/dcn20_resource.c | 16 +- .../gpu/drm/drm_panel_orientation_quirks.c | 47 +++- drivers/gpu/drm/drm_plane_helper.c | 1 - drivers/gpu/drm/msm/disp/dpu1/dpu_hw_sspp.c | 8 +- drivers/gpu/drm/msm/disp/dpu1/dpu_kms.c | 4 + drivers/gpu/drm/msm/msm_gem.c | 4 +- drivers/gpu/drm/msm/msm_gpu.c | 2 +- drivers/gpu/drm/nouveau/nouveau_svm.c | 4 + drivers/gpu/drm/sun4i/sun8i_csc.h | 4 +- drivers/gpu/drm/ttm/ttm_bo_vm.c | 5 - drivers/gpu/drm/v3d/v3d_gem.c | 4 +- drivers/gpu/drm/virtio/virtgpu_vq.c | 8 +- drivers/hid/hid-u2fzero.c | 10 +- drivers/hv/hyperv_vmbus.h | 1 + drivers/hwmon/hwmon.c | 6 +- drivers/hwmon/pmbus/lm25066.c | 25 +- .../hwtracing/coresight/coresight-cti-core.c | 2 +- drivers/i2c/busses/i2c-mt65xx.c | 2 +- drivers/i2c/busses/i2c-xlr.c | 6 +- drivers/iio/accel/st_accel_core.c | 21 +- drivers/iio/accel/st_accel_i2c.c | 17 +- drivers/iio/accel/st_accel_spi.c | 17 +- drivers/iio/dac/ad5446.c | 9 +- drivers/iio/dac/ad5770r.c | 2 +- drivers/iio/gyro/st_gyro_core.c | 15 +- drivers/iio/gyro/st_gyro_i2c.c | 17 +- drivers/iio/gyro/st_gyro_spi.c | 17 +- drivers/iio/imu/adis.c | 4 +- drivers/iio/magnetometer/st_magn_core.c | 15 +- drivers/iio/magnetometer/st_magn_i2c.c | 14 +- drivers/iio/magnetometer/st_magn_spi.c | 14 +- drivers/iio/pressure/st_pressure_core.c | 15 +- drivers/iio/pressure/st_pressure_i2c.c | 17 +- drivers/iio/pressure/st_pressure_spi.c | 17 +- drivers/infiniband/hw/bnxt_re/qplib_fp.c | 3 +- drivers/infiniband/hw/mlx4/qp.c | 4 +- drivers/infiniband/hw/qedr/verbs.c | 15 +- drivers/infiniband/sw/rxe/rxe_param.h | 2 +- drivers/input/joystick/iforce/iforce-usb.c | 2 +- drivers/input/mouse/elantech.c | 13 + drivers/input/serio/i8042-x86ia64io.h | 14 + drivers/irqchip/irq-bcm6345-l1.c | 2 +- drivers/irqchip/irq-sifive-plic.c | 8 +- drivers/isdn/hardware/mISDN/hfcpci.c | 8 +- drivers/md/md.c | 11 +- drivers/media/dvb-frontends/mn88443x.c | 18 +- drivers/media/i2c/ir-kbd-i2c.c | 1 + drivers/media/i2c/mt9p031.c | 28 +- drivers/media/i2c/tda1997x.c | 8 +- drivers/media/pci/cx23885/cx23885-alsa.c | 3 +- .../pci/netup_unidvb/netup_unidvb_core.c | 27 +- drivers/media/platform/mtk-vpu/mtk_vpu.c | 5 +- drivers/media/platform/rcar-vin/rcar-csi2.c | 2 + drivers/media/platform/s5p-mfc/s5p_mfc.c | 6 +- drivers/media/platform/stm32/stm32-dcmi.c | 19 +- drivers/media/radio/radio-wl1273.c | 2 +- drivers/media/radio/si470x/radio-si470x-i2c.c | 2 +- drivers/media/radio/si470x/radio-si470x-usb.c | 2 +- drivers/media/rc/ir_toy.c | 2 +- drivers/media/rc/ite-cir.c | 2 +- drivers/media/rc/mceusb.c | 1 + drivers/media/spi/cxd2880-spi.c | 2 +- drivers/media/usb/dvb-usb/az6027.c | 1 + drivers/media/usb/dvb-usb/dibusb-common.c | 2 +- drivers/media/usb/em28xx/em28xx-cards.c | 5 +- drivers/media/usb/em28xx/em28xx-core.c | 5 +- drivers/media/usb/tm6000/tm6000-video.c | 3 +- drivers/media/usb/uvc/uvc_driver.c | 7 +- drivers/media/usb/uvc/uvc_v4l2.c | 7 +- drivers/media/usb/uvc/uvc_video.c | 5 + drivers/media/v4l2-core/v4l2-ioctl.c | 67 +++-- drivers/memory/fsl_ifc.c | 13 +- drivers/memory/renesas-rpc-if.c | 113 +++++--- drivers/memstick/core/ms_block.c | 2 +- drivers/memstick/host/jmb38x_ms.c | 2 +- drivers/memstick/host/r592.c | 8 +- drivers/mfd/dln2.c | 18 ++ drivers/mfd/mfd-core.c | 2 + drivers/mmc/host/Kconfig | 2 +- drivers/mmc/host/dw_mmc.c | 3 +- drivers/mmc/host/moxart-mmc.c | 29 +- drivers/mmc/host/mtk-sd.c | 5 + drivers/mmc/host/mxs-mmc.c | 10 + drivers/mmc/host/sdhci-omap.c | 18 +- drivers/most/most_usb.c | 5 +- drivers/mtd/mtdcore.c | 4 +- drivers/mtd/nand/raw/ams-delta.c | 12 +- drivers/mtd/nand/raw/au1550nd.c | 12 +- drivers/mtd/nand/raw/gpio.c | 12 +- drivers/mtd/nand/raw/mpc5121_nfc.c | 12 +- drivers/mtd/nand/raw/orion_nand.c | 12 +- drivers/mtd/nand/raw/pasemi_nand.c | 12 +- drivers/mtd/nand/raw/plat_nand.c | 12 +- drivers/mtd/nand/raw/socrates_nand.c | 12 +- drivers/mtd/nand/raw/xway_nand.c | 12 +- drivers/mtd/spi-nor/controllers/hisi-sfc.c | 1 - drivers/net/Kconfig | 2 +- .../net/can/spi/mcp251xfd/mcp251xfd-core.c | 2 +- drivers/net/dsa/rtl8366rb.c | 2 +- drivers/net/ethernet/amd/xgbe/xgbe-common.h | 8 + drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c | 20 +- .../net/ethernet/cavium/thunder/nic_main.c | 2 +- .../net/ethernet/cavium/thunder/nicvf_main.c | 4 +- .../ethernet/chelsio/cxgb4/cxgb4_ethtool.c | 7 +- drivers/net/ethernet/chelsio/cxgb4/t4_hw.h | 2 + .../chelsio/inline_crypto/chtls/chtls_cm.c | 2 +- .../chelsio/inline_crypto/chtls/chtls_cm.h | 2 +- .../net/ethernet/freescale/enetc/enetc_qos.c | 18 +- drivers/net/ethernet/google/gve/gve.h | 4 +- drivers/net/ethernet/google/gve/gve_adminq.h | 1 + drivers/net/ethernet/google/gve/gve_main.c | 48 +++- drivers/net/ethernet/ibm/ibmvnic.c | 5 +- drivers/net/ethernet/intel/ice/ice_base.c | 2 +- .../net/ethernet/intel/ice/ice_virtchnl_pf.c | 20 +- drivers/net/ethernet/netronome/nfp/bpf/main.c | 16 +- drivers/net/ethernet/netronome/nfp/bpf/main.h | 2 + .../net/ethernet/netronome/nfp/bpf/offload.c | 17 +- drivers/net/ethernet/realtek/r8169_main.c | 1 + drivers/net/ethernet/sfc/mcdi_port_common.c | 37 ++- drivers/net/ethernet/sfc/ptp.c | 4 +- drivers/net/ethernet/sfc/siena_sriov.c | 2 +- .../net/ethernet/stmicro/stmmac/stmmac_tc.c | 2 - drivers/net/ethernet/ti/davinci_emac.c | 16 +- drivers/net/ifb.c | 2 + drivers/net/phy/micrel.c | 9 +- drivers/net/phy/phylink.c | 2 +- drivers/net/vmxnet3/vmxnet3_drv.c | 1 - drivers/net/vrf.c | 28 +- drivers/net/wireless/ath/ath10k/mac.c | 45 +++- drivers/net/wireless/ath/ath10k/sdio.c | 5 +- drivers/net/wireless/ath/ath10k/usb.c | 7 +- drivers/net/wireless/ath/ath10k/wmi.c | 4 + drivers/net/wireless/ath/ath10k/wmi.h | 3 + drivers/net/wireless/ath/ath11k/dbring.c | 16 +- drivers/net/wireless/ath/ath11k/dp_rx.c | 13 +- drivers/net/wireless/ath/ath11k/mac.c | 2 +- drivers/net/wireless/ath/ath11k/qmi.c | 4 +- drivers/net/wireless/ath/ath11k/reg.c | 11 +- drivers/net/wireless/ath/ath11k/reg.h | 2 +- drivers/net/wireless/ath/ath11k/wmi.c | 40 ++- drivers/net/wireless/ath/ath11k/wmi.h | 3 +- drivers/net/wireless/ath/ath6kl/usb.c | 7 +- drivers/net/wireless/ath/ath9k/main.c | 4 +- .../net/wireless/ath/dfs_pattern_detector.c | 10 +- drivers/net/wireless/ath/wcn36xx/dxe.c | 49 ++-- drivers/net/wireless/ath/wcn36xx/main.c | 8 +- drivers/net/wireless/ath/wcn36xx/smd.c | 44 ++- drivers/net/wireless/ath/wcn36xx/txrx.c | 64 +++-- drivers/net/wireless/ath/wcn36xx/txrx.h | 3 +- drivers/net/wireless/broadcom/b43/phy_g.c | 2 +- .../net/wireless/broadcom/b43legacy/radio.c | 2 +- .../broadcom/brcm80211/brcmfmac/dmi.c | 10 + .../net/wireless/intel/iwlwifi/mvm/utils.c | 3 + .../net/wireless/marvell/libertas/if_usb.c | 2 + .../net/wireless/marvell/libertas_tf/if_usb.c | 2 + drivers/net/wireless/marvell/mwifiex/11n.c | 5 +- .../net/wireless/marvell/mwifiex/cfg80211.c | 32 +-- drivers/net/wireless/marvell/mwifiex/pcie.c | 36 ++- drivers/net/wireless/marvell/mwifiex/usb.c | 16 ++ drivers/net/wireless/marvell/mwl8k.c | 2 +- .../net/wireless/mediatek/mt76/mt7615/mac.c | 15 +- .../net/wireless/mediatek/mt76/mt76x02_mac.c | 13 +- .../net/wireless/mediatek/mt76/mt7915/mcu.c | 8 +- .../wireless/microchip/wilc1000/cfg80211.c | 3 +- .../realtek/rtl818x/rtl8187/rtl8225.c | 14 +- drivers/net/wireless/realtek/rtw88/fw.c | 7 +- drivers/net/wireless/realtek/rtw88/reg.h | 1 + drivers/net/wireless/rsi/rsi_91x_core.c | 2 + drivers/net/wireless/rsi/rsi_91x_hal.c | 10 +- drivers/net/wireless/rsi/rsi_91x_mac80211.c | 74 ++---- drivers/net/wireless/rsi/rsi_91x_main.c | 17 +- drivers/net/wireless/rsi/rsi_91x_mgmt.c | 24 +- drivers/net/wireless/rsi/rsi_91x_sdio.c | 5 +- drivers/net/wireless/rsi/rsi_91x_usb.c | 5 +- drivers/net/wireless/rsi/rsi_hal.h | 11 + drivers/net/wireless/rsi/rsi_main.h | 15 +- drivers/net/xen-netfront.c | 8 + drivers/nfc/pn533/pn533.c | 6 +- drivers/nvme/host/multipath.c | 9 +- drivers/nvme/host/rdma.c | 2 + drivers/nvme/target/configfs.c | 2 + drivers/nvme/target/rdma.c | 24 ++ drivers/nvme/target/tcp.c | 21 +- drivers/of/unittest.c | 16 +- drivers/opp/of.c | 2 +- .../controller/cadence/pcie-cadence-plat.c | 2 + drivers/pci/controller/dwc/pcie-uniphier.c | 26 +- drivers/pci/controller/pci-aardvark.c | 251 +++++++++++++++--- drivers/pci/pci-bridge-emul.c | 13 + drivers/pci/quirks.c | 1 + drivers/phy/qualcomm/phy-qcom-qusb2.c | 16 +- drivers/phy/qualcomm/phy-qcom-snps-femto-v2.c | 2 +- drivers/phy/ti/phy-gmii-sel.c | 2 + drivers/pinctrl/core.c | 2 + drivers/pinctrl/pinctrl-equilibrium.c | 7 +- drivers/pinctrl/renesas/core.c | 2 +- drivers/platform/x86/thinkpad_acpi.c | 2 +- drivers/platform/x86/wmi.c | 9 +- drivers/power/supply/bq27xxx_battery_i2c.c | 3 +- drivers/power/supply/max17040_battery.c | 2 + drivers/power/supply/max17042_battery.c | 12 +- drivers/power/supply/rt5033_battery.c | 2 +- drivers/regulator/s5m8767.c | 21 +- drivers/remoteproc/remoteproc_core.c | 8 +- drivers/reset/reset-socfpga.c | 26 ++ drivers/rtc/rtc-rv3032.c | 4 +- drivers/s390/char/tape_std.c | 3 +- drivers/s390/cio/css.c | 4 +- drivers/s390/cio/device_ops.c | 12 +- drivers/s390/crypto/ap_queue.c | 2 + drivers/scsi/csiostor/csio_lnode.c | 2 +- drivers/scsi/dc395x.c | 1 + drivers/scsi/pm8001/pm8001_hwi.c | 2 +- drivers/scsi/qla2xxx/qla_attr.c | 24 +- drivers/scsi/qla2xxx/qla_dbg.c | 3 +- drivers/scsi/qla2xxx/qla_gbl.h | 2 - drivers/scsi/qla2xxx/qla_init.c | 54 +++- drivers/scsi/qla2xxx/qla_mr.c | 23 -- drivers/scsi/qla2xxx/qla_os.c | 47 ++-- drivers/scsi/qla2xxx/qla_target.c | 14 +- drivers/scsi/scsi_lib.c | 2 - drivers/scsi/ufs/ufshcd-pltfrm.c | 6 +- drivers/scsi/ufs/ufshcd.c | 29 +- drivers/scsi/ufs/ufshcd.h | 3 + drivers/soc/fsl/dpaa2-console.c | 1 + drivers/soc/fsl/dpio/dpio-service.c | 2 +- drivers/soc/fsl/dpio/qbman-portal.c | 9 +- drivers/soc/qcom/apr.c | 2 + drivers/soc/qcom/rpmhpd.c | 21 +- drivers/soc/tegra/pmc.c | 4 +- drivers/soundwire/debugfs.c | 2 +- drivers/spi/spi-bcm-qspi.c | 5 +- drivers/spi/spi-pl022.c | 5 +- drivers/spi/spi-rpc-if.c | 4 +- drivers/staging/ks7010/Kconfig | 3 + .../staging/media/allegro-dvt/allegro-core.c | 9 + .../media/atomisp/i2c/atomisp-lm3554.c | 37 ++- .../staging/media/imx/imx-media-dev-common.c | 2 + drivers/staging/media/ipu3/ipu3-v4l2.c | 7 +- drivers/staging/media/rkvdec/rkvdec-h264.c | 5 +- drivers/staging/media/rkvdec/rkvdec.c | 40 +-- drivers/staging/most/dim2/Makefile | 2 +- drivers/staging/most/dim2/dim2.c | 24 +- drivers/staging/most/dim2/sysfs.c | 49 ---- drivers/staging/most/dim2/sysfs.h | 11 - drivers/tty/serial/8250/8250_dw.c | 2 +- drivers/tty/serial/8250/8250_port.c | 21 +- drivers/tty/serial/imx.c | 4 +- drivers/tty/serial/serial_core.c | 16 +- drivers/tty/serial/xilinx_uartps.c | 3 +- drivers/usb/chipidea/core.c | 23 +- drivers/usb/dwc2/drd.c | 24 +- drivers/usb/gadget/legacy/hid.c | 4 +- drivers/usb/host/xhci-hub.c | 3 +- drivers/usb/host/xhci-pci.c | 16 ++ drivers/usb/misc/iowarrior.c | 8 +- drivers/usb/musb/Kconfig | 2 +- drivers/usb/serial/keyspan.c | 15 +- drivers/usb/typec/Kconfig | 4 +- drivers/video/backlight/backlight.c | 6 - drivers/video/fbdev/chipsfb.c | 2 +- drivers/virtio/virtio_ring.c | 14 +- drivers/watchdog/Kconfig | 2 +- drivers/watchdog/f71808e_wdt.c | 4 +- drivers/watchdog/omap_wdt.c | 6 +- drivers/xen/balloon.c | 86 ++++-- .../xen/xen-pciback/conf_space_capability.c | 2 +- fs/btrfs/disk-io.c | 3 +- fs/btrfs/reflink.c | 2 +- fs/btrfs/tree-log.c | 4 +- fs/btrfs/volumes.c | 14 +- fs/crypto/fscrypt_private.h | 5 +- fs/crypto/hkdf.c | 11 +- fs/crypto/keysetup.c | 57 +++- fs/erofs/decompressor.c | 1 - fs/exfat/inode.c | 2 +- fs/ext4/super.c | 9 +- fs/f2fs/inode.c | 2 +- fs/f2fs/namei.c | 2 +- fs/fuse/dev.c | 14 +- fs/gfs2/glock.c | 24 +- fs/jfs/jfs_mount.c | 51 ++-- fs/nfs/dir.c | 7 +- fs/nfs/direct.c | 2 +- fs/nfs/flexfilelayout/flexfilelayoutdev.c | 4 +- fs/nfs/nfs4idmap.c | 2 +- fs/nfs/nfs4proc.c | 15 +- fs/nfs/pnfs.h | 2 +- fs/nfs/pnfs_nfs.c | 6 +- fs/nfs/write.c | 26 +- fs/ocfs2/file.c | 8 +- fs/orangefs/dcache.c | 4 +- fs/proc/stat.c | 4 +- fs/proc/uptime.c | 14 +- fs/tracefs/inode.c | 3 +- include/linux/blkdev.h | 2 - include/linux/cc_platform.h | 88 ++++++ include/linux/console.h | 2 + include/linux/ethtool_netlink.h | 3 + include/linux/filter.h | 1 + include/linux/kernel_stat.h | 1 + include/linux/libata.h | 2 +- include/linux/lsm_hook_defs.h | 14 +- include/linux/lsm_hooks.h | 14 +- include/linux/nfs_fs.h | 1 + include/linux/posix-timers.h | 2 + include/linux/rpmsg.h | 2 +- include/linux/sched/task.h | 3 +- include/linux/sched/task_stack.h | 4 + include/linux/security.h | 33 ++- include/linux/seq_file.h | 2 +- include/linux/tpm.h | 1 + include/memory/renesas-rpc-if.h | 1 + include/net/inet_connection_sock.h | 2 +- include/net/llc.h | 4 +- include/net/neighbour.h | 12 +- include/net/sch_generic.h | 4 + include/net/sock.h | 2 +- include/net/strparser.h | 16 +- include/net/tcp.h | 17 +- include/net/udp.h | 5 +- include/uapi/linux/ethtool_netlink.h | 4 +- include/uapi/linux/pci_regs.h | 6 + kernel/bpf/core.c | 4 +- kernel/bpf/verifier.c | 4 +- kernel/cgroup/cgroup.c | 31 ++- kernel/cgroup/rstat.c | 2 - kernel/fork.c | 3 +- kernel/kprobes.c | 3 +- kernel/locking/lockdep.c | 4 +- kernel/power/energy_model.c | 23 +- kernel/power/swap.c | 2 +- kernel/rcu/rcutorture.c | 48 +++- kernel/rcu/tasks.h | 3 +- kernel/rcu/tree_exp.h | 2 +- kernel/rcu/tree_plugin.h | 8 +- kernel/sched/core.c | 43 +-- kernel/signal.c | 18 +- kernel/time/posix-cpu-timers.c | 19 +- kernel/trace/ring_buffer.c | 5 + kernel/trace/tracing_map.c | 40 +-- kernel/workqueue.c | 15 +- lib/decompress_unxz.c | 2 +- lib/iov_iter.c | 5 +- lib/xz/xz_dec_lzma2.c | 21 +- lib/xz/xz_dec_stream.c | 6 +- mm/memcontrol.c | 27 +- mm/oom_kill.c | 23 +- mm/zsmalloc.c | 7 +- net/8021q/vlan.c | 3 - net/8021q/vlan_dev.c | 3 + net/9p/client.c | 2 + net/bluetooth/sco.c | 9 +- net/can/j1939/main.c | 7 + net/can/j1939/transport.c | 6 + net/core/dev.c | 5 +- net/core/filter.c | 21 ++ net/core/neighbour.c | 48 ++-- net/core/net-sysfs.c | 55 ++++ net/core/net_namespace.c | 4 + net/core/stream.c | 3 - net/core/sysctl_net_core.c | 2 +- net/dccp/dccp.h | 2 +- net/dccp/proto.c | 14 +- net/ethtool/pause.c | 3 +- net/ipv4/inet_connection_sock.c | 4 +- net/ipv4/inet_hashtables.c | 2 +- net/ipv4/proc.c | 2 +- net/ipv4/tcp.c | 40 ++- net/ipv4/tcp_bpf.c | 1 - net/ipv6/addrconf.c | 3 + net/ipv6/udp.c | 2 +- net/netfilter/nf_conntrack_proto_udp.c | 7 +- net/netfilter/nfnetlink_queue.c | 2 +- net/netfilter/nft_dynset.c | 11 +- net/rds/ib.c | 10 - net/rds/ib.h | 6 - net/rds/ib_cm.c | 128 +++++---- net/rds/ib_recv.c | 18 +- net/rds/ib_send.c | 8 + net/rxrpc/rtt.c | 2 +- net/sched/sch_generic.c | 9 + net/sched/sch_mq.c | 24 ++ net/sched/sch_mqprio.c | 23 ++ net/sched/sch_taprio.c | 27 +- net/smc/af_smc.c | 20 +- net/smc/smc_llc.c | 2 +- net/strparser/strparser.c | 10 +- net/sunrpc/addr.c | 40 ++- net/sunrpc/xprt.c | 28 +- net/vmw_vsock/af_vsock.c | 2 + samples/kprobes/kretprobe_example.c | 2 +- scripts/leaking_addresses.pl | 3 +- security/apparmor/label.c | 4 +- security/integrity/evm/evm_main.c | 2 +- security/security.c | 14 +- security/selinux/hooks.c | 36 ++- security/selinux/ss/services.c | 162 ++++++----- security/smack/smackfs.c | 11 +- sound/core/oss/mixer_oss.c | 43 ++- sound/core/timer.c | 13 +- sound/pci/hda/hda_intel.c | 74 +++--- sound/pci/hda/patch_realtek.c | 82 ++++++ sound/soc/codecs/cs42l42.c | 88 +++--- sound/soc/soc-core.c | 1 + sound/soc/sof/topology.c | 9 + sound/synth/emux/emux.c | 2 +- sound/usb/6fire/comm.c | 2 +- sound/usb/6fire/firmware.c | 6 +- sound/usb/format.c | 1 + sound/usb/line6/driver.c | 14 +- sound/usb/line6/driver.h | 2 +- sound/usb/line6/podhd.c | 6 +- sound/usb/line6/toneport.c | 2 +- sound/usb/misc/ua101.c | 4 +- sound/usb/quirks.c | 1 + tools/bpf/bpftool/prog.c | 16 +- tools/lib/bpf/bpf_core_read.h | 2 +- tools/lib/bpf/btf.c | 25 +- tools/objtool/check.c | 19 +- tools/perf/util/bpf-event.c | 4 +- .../selftests/bpf/prog_tests/perf_buffer.c | 4 +- .../selftests/bpf/prog_tests/sk_lookup.c | 85 ++++-- .../testing/selftests/bpf/progs/strobemeta.h | 4 +- .../selftests/bpf/progs/test_sk_lookup.c | 62 +++-- tools/testing/selftests/bpf/test_progs.c | 4 +- .../selftests/bpf/verifier/array_access.c | 2 +- .../testing/selftests/core/close_range_test.c | 2 +- tools/testing/selftests/kvm/lib/x86_64/svm.c | 22 +- .../selftests/kvm/x86_64/mmio_warning_test.c | 2 +- tools/testing/selftests/net/fcnal-test.sh | 3 + tools/testing/selftests/net/udpgso_bench_rx.c | 11 +- 587 files changed, 4709 insertions(+), 2317 deletions(-) create mode 100644 arch/powerpc/include/asm/kvm_guest.h create mode 100644 arch/x86/kernel/cc_platform.c delete mode 100644 drivers/staging/most/dim2/sysfs.c create mode 100644 include/linux/cc_platform.h -- 2.20.1

1 557

[PATCH OLK-5.10] Add the configuration for accelerated of SM4
by shenzijun 06 Dec '21

06 Dec '21

From: 沈子俊 <shenzijun(a)kylinos.cn> kylin inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4A842?from=project-issue CVE: NA ------------------------------------------------------------------------- Add the configuration in arch/x86/configs/openeuler_defconfig Signed-off-by: 沈子俊 <shenzijun(a)kylinos.cn> --- arch/x86/configs/openeuler_defconfig | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/x86/configs/openeuler_defconfig b/arch/x86/configs/openeuler_defconfig index b25d908dc7a1..9b23f113f669 100644 --- a/arch/x86/configs/openeuler_defconfig +++ b/arch/x86/configs/openeuler_defconfig @@ -7991,6 +7991,8 @@ CONFIG_CRYPTO_SERPENT_SSE2_X86_64=m CONFIG_CRYPTO_SERPENT_AVX_X86_64=m CONFIG_CRYPTO_SERPENT_AVX2_X86_64=m CONFIG_CRYPTO_SM4=m +CONFIG_CRYPTO_SM4_AESNI_AVX_X86_64=m +CONFIG_CRYPTO_SM4_AESNI_AVX2_X86_64=m CONFIG_CRYPTO_TEA=m CONFIG_CRYPTO_TWOFISH=m CONFIG_CRYPTO_TWOFISH_COMMON=m -- 2.30.0

1 0

[PATCH openEuler-5.10 01/54] hugetlb: before freeing hugetlb page set dtor to appropriate value
by Zheng Zengkai 03 Dec '21

03 Dec '21

From: Mike Kravetz <mike.kravetz(a)oracle.com> mainline inclusion from mainline-v5.15-rc1 commit e32d20c0c88b1cd0a44f882c4f0eb2f536363d1b category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I4IGRQ CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… ---------------------------------------------------------------------- When removing a hugetlb page from the pool the ref count is set to one (as the free page has no ref count) and compound page destructor is set to NULL_COMPOUND_DTOR. Since a subsequent call to free the hugetlb page will call __free_pages for non-gigantic pages and free_gigantic_page for gigantic pages the destructor is not used. However, consider the following race with code taking a speculative reference on the page: Thread 0 Thread 1 -------- -------- remove_hugetlb_page set_page_refcounted(page); set_compound_page_dtor(page, NULL_COMPOUND_DTOR); get_page_unless_zero(page) __update_and_free_page __free_pages(page, huge_page_order(h)); /* Note that __free_pages() will simply drop the reference to the page. */ put_page(page) __put_compound_page() destroy_compound_page NULL_COMPOUND_DTOR BUG: kernel NULL pointer dereference, address: 0000000000000000 To address this race, set the dtor to the normal compound page dtor for non-gigantic pages. The dtor for gigantic pages does not matter as gigantic pages are changed from a compound page to 'just a group of pages' before freeing. Hence, the destructor is not used. Link: https://lkml.kernel.org/r/20210809184832.18342-4-mike.kravetz@oracle.com Signed-off-by: Mike Kravetz <mike.kravetz(a)oracle.com> Reviewed-by: Muchun Song <songmuchun(a)bytedance.com> Cc: Michal Hocko <mhocko(a)suse.com> Cc: Oscar Salvador <osalvador(a)suse.de> Cc: David Hildenbrand <david(a)redhat.com> Cc: Matthew Wilcox <willy(a)infradead.org> Cc: Naoya Horiguchi <naoya.horiguchi(a)linux.dev> Cc: Mina Almasry <almasrymina(a)google.com> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds(a)linux-foundation.org> Signed-off-by: Chen Wandun <chenwandun(a)huawei.com> Reviewed-by: Kefeng Wang <wangkefeng.wang(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> --- mm/hugetlb.c | 22 +++++++++++++++++++++- 1 file changed, 21 insertions(+), 1 deletion(-) diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 47dd6b5e0040..6ae2d2e90681 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -1364,8 +1364,28 @@ static void remove_hugetlb_page(struct hstate *h, struct page *page, h->surplus_huge_pages_node[nid]--; } + /* + * Very subtle + * + * For non-gigantic pages set the destructor to the normal compound + * page dtor. This is needed in case someone takes an additional + * temporary ref to the page, and freeing is delayed until they drop + * their reference. + * + * For gigantic pages set the destructor to the null dtor. This + * destructor will never be called. Before freeing the gigantic + * page destroy_compound_gigantic_page will turn the compound page + * into a simple group of pages. After this the destructor does not + * apply. + * + * This handles the case where more than one ref is held when and + * after update_and_free_page is called. + */ set_page_refcounted(page); - set_compound_page_dtor(page, NULL_COMPOUND_DTOR); + if (hstate_is_gigantic(h)) + set_compound_page_dtor(page, NULL_COMPOUND_DTOR); + else + set_compound_page_dtor(page, COMPOUND_PAGE_DTOR); h->nr_huge_pages--; h->nr_huge_pages_node[nid]--; -- 2.20.1

1 53

Re: 【openEuler】Netswift Giga NIC驱动合入openEuler20.03
by Xie XiuQi 03 Dec '21

03 Dec '21

On 2021/12/3 17:46, 郑振鹏 wrote: > 谢工，您好 > > 在千兆驱动合入4.19内核完成后，我打算先推万兆驱动4.19内核功能升级及同步最新bug修复的补丁，再准备5.10内核千兆驱动补丁。 5.10 会在 3 月份发 22.03 LTS 版本，不打算这个版本带进来吗 > > BR， > zhenpeng > > ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ > > **************************************************************** > > 郑振鹏(zheng zhenpeng) > > 北京网迅科技有限公司杭州分公司软件工程师 > > 浙江省杭州市西湖区文三路478号华星时代广场A座507室 310013 > > Beijing WangXun Technology Co., Ltd. Software Engineer. > > Room A507, HuaXing Times Square, No.478 West Wensan Road. > > West Lake District, Hangzhou City, 310013 ZHEJIANG, P.R.CHINA. > > > > Office: +86(0571)89807901-8014 > > Mobile: +86-13656681762 > > E-Mail: z <mailto:jianwang@trustnetic.com>henpengzheng(a)net-swift.com > > **************************************************************** > > > *发件人：* Xie XiuQi <mailto:xiexiuqi@huawei.com> > *发送时间：* 2021-12-03 16:15 > *收件人：* 郑振鹏 <mailto:zhenpengzheng@net-swift.com>; QiuLaibin <mailto:qiulaibin@huawei.com> > *抄送：* yangyingliang(a)huawei.com <mailto:yangyingliang@huawei.com>; kernel(a)openeuler.org <mailto:kernel@openeuler.org> > *主题：* Re: 【openEuler】Netswift Giga NIC驱动合入openEuler20.03 > > > On 2021/12/3 15:17, 郑振鹏 wrote: > > 您好， > > > > 目前我只在x86平台上测试我此次提交的patch，如果完成arm平台测试我再提交新的patch。 > > 感谢。 > > Netswift Giga NIC 对 22.03 （5.10内核）的支持，这个是不是也在做。 > > > > > BR， > > zhenpeng > > > > > ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ > > > > **************************************************************** > > > > 郑振鹏(zheng zhenpeng) > > > > 北京网迅科技有限公司杭州分公司软件工程师 > > > > 浙江省杭州市西湖区文三路478号华星时代广场A座507室 310013 > > > > Beijing WangXun Technology Co., Ltd. Software Engineer. > > > > Room A507, HuaXing Times Square, No.478 West Wensan Road. > > > > West Lake District, Hangzhou City, 310013 ZHEJIANG, P.R.CHINA. > > > > > > > > Office: +86(0571)89807901-8014 > > > > Mobile: +86-13656681762 > > > > E-Mail: z <mailto:jianwang@trustnetic.com>henpengzheng(a)net-swift.com > > > > **************************************************************** > > > > > > *发件人：* QiuLaibin <mailto:qiulaibin@huawei.com> > > *发送时间：* 2021-12-03 15:01 > > *收件人：* zhenpengzheng(a)net-swift.com <mailto:zhenpengzheng@net-swift.com> > > *抄送：* Xiexiuqi <mailto:xiexiuqi@huawei.com>; yangyingliang(a)huawei.com <mailto:yangyingliang@huawei.com> > > *主题：* 【openEuler】Netswift Giga NIC驱动合入openEuler20.03 > > pengzheng您好： > > > > 非常感谢您的提交！ > > > > 目前我们正在向版本合入先前您针对openEuler提交的Netswift Giga NIC驱补丁集： > > > > [openEuler-1.0-LTS,1/2] net: ngbe: Add Netswift Giga NIC driver > > [openEuler-1.0-LTS,2/2] x86/config: Enable netswift Giga NIC driver for x86 > > > > 但鉴于您在编译config只在X86的config中打开，因此我们需要确认是否该驱动只能支持X86平台；arm上是否可以打开支持，若arm相关的config误打开是否会引入问题。 > > > > best regard > > Laibin Qiu > > > > > > >

1 0

Re: [Tc] [Dev] [TC]请给位议题申报人按时与会，谢谢。RE: openEuler 技术委员会例会
by Xie XiuQi 03 Dec '21

03 Dec '21

TC 议题申报：议题：openEuler 22.03 LTS ARM64 版本(5.10 内核）内核页大小及支持CPU规格决策：决策点1：openEuler 22.03 LTS 是否默认采用 4K 页表（配合 48 BIT VA/PA) 决策点2：openEuler 22.03 LTS 如果默认采用 4K 页表，是否需要在单独出 64K 页表的 kernel 包。决策点3：openEuler 22.03 LTS 支持最大 CPU 数：NR_CPUS=4096, NODES 支持 128. （20.03 原版本配置：64K, 48 BIT VA/PA, NR_CPUS 1024, NODES 16) 之前的讨论链接： https://gitee.com/openeuler/kernel/issues/I4HDHZ 理由1：从讨论看，64K 兼容性问题比较多，且业界OS多采用4K，方便驱动及上层软件在兼容性认证，也利于支撑更多的使用场景。性能上，64K 优势场景，可以采用大页，或者单独 64K 版本支持。理由2：52 BIT 大内存支持，需要 64K 页，4K 页最大支持 48 BIT VA/PA, 256T 内存空间。如果要支持更大 PB 级内存，需要 64K。理由3：从未来几年诉求，和业界OS支持规格来看，CPU 数目增加是趋势，从兼容性和规格考虑， NR_CPUS 与业界OS最大的 4096 一致。当前4路飞腾S2500 最大 32 个 nodes，为满足未来几年的演进诉求，NODES 数拟最大支持 128. On 2021/12/3 14:35, yangcong wrote: > > > 申报议题： > 议题5：License风险识别系统引入openEuler社区 - 杨聪 <yangcong_hit(a)163.com> > 议题6：针对openEuler社区repo不同的类型（内容），选择license的策略 - 杨聪 <yangcong_hit(a)163.com> > 在2021年12月1日 09:34，Hufeng (Solar, Euler)<solar.hu(a)huawei.com> 写道： > 当前已有议题如下 > 议题1：release management sig工作计划 - 胡峰（延期到本次例会） > 议题2：openEuler 新技术领域规划讨论：Edge – 刘寿永，Embedded – 任慰（延期到本次例会） > 议题3：NestOS的社区资源支持与立项讨论：杜奕威 <duyiwei(a)kylinos.cn>（延期到本次例会） > a. 如何从openEuler官网下载NestOS镜像 > b. 提供相关资源以部署nestos所需要的自动更新环境 > c. 如何立项使更多的人参与到NestOS中 > 议题4： Greenplum白皮书评审-270162781(a)qq.com - bo zhao > 上会前请确认会议遗留问题闭环情况 > > > -----Original Message----- > From: Hufeng (Solar, Euler) > Sent: Tuesday, November 30, 2021 2:45 PM > To: tc(a)openeuler.org; 'dev(a)openeuler.org' <dev(a)openeuler.org> > Subject: [TC]议题收集，请给位议题申报人按时与会，谢谢。RE: [Dev] openEuler 技术委员会例会 > > 当前已有议题如下 > 议题1：release management sig工作计划 - 胡峰（延期到本次例会） > 议题2：openEuler 新技术领域规划讨论：Edge – 刘寿永，Embedded – 任慰（延期到本次例会） > 议题3：NestOS的社区资源支持与立项讨论：杜奕威 <duyiwei(a)kylinos.cn>（延期到本次例会） > a. 如何从openEuler官网下载NestOS镜像 > b. 提供相关资源以部署nestos所需要的自动更新环境 > c. 如何立项使更多的人参与到NestOS中 > > -----Original Message----- > From: openEuler conference [mailto:public@openeuler.org] > Sent: Tuesday, November 30, 2021 2:35 PM > Subject: [Dev] openEuler 技术委员会例会 > > 您好！ > > TC SIG 邀请您参加 2021-12-01 10:00 召开的ZOOM会议(自动录制) > > 会议主题：openEuler 技术委员会例会 > > 会议链接：https://us06web.zoom.us/j/82159612220?pwd=ZlNxWkEwY1MyUlQ3SmtFNmNEVGtwQT09 > > 温馨提醒：建议接入会议后修改参会人的姓名，也可以使用您在gitee.com的ID > > 更多资讯尽在：https://openeuler.org/zh/ > > > > > Hello! > > openEuler TC SIG invites you to attend the ZOOM conference(auto recording) will be held at 2021-12-01 10:00, > > The subject of the conference is openEuler 技术委员会例会, > > You can join the meeting at https://us06web.zoom.us/j/82159612220?pwd=ZlNxWkEwY1MyUlQ3SmtFNmNEVGtwQT09. > > Note: You are advised to change the participant name after joining the conference or use your ID at gitee.com. > > More information: https://openeuler.org/en/ > > _______________________________________________ > Dev mailing list -- dev(a)openeuler.org > To unsubscribe send an email to dev-leave(a)openeuler.org > _______________________________________________ > Dev mailing list -- dev(a)openeuler.org > To unsubscribe send an email to dev-leave(a)openeuler.org > _______________________________________________ > Tc mailing list -- tc(a)openeuler.org > To unsubscribe send an email to tc-leave(a)openeuler.org >

3 2

Re: 【openEuler】Netswift Giga NIC驱动合入openEuler20.03
by Xie XiuQi 03 Dec '21

03 Dec '21

On 2021/12/3 15:17, 郑振鹏 wrote: > 您好， > > 目前我只在x86平台上测试我此次提交的patch，如果完成arm平台测试我再提交新的patch。感谢。 Netswift Giga NIC 对 22.03 （5.10内核）的支持，这个是不是也在做。 > > BR， > zhenpeng > > ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ > > **************************************************************** > > 郑振鹏(zheng zhenpeng) > > 北京网迅科技有限公司杭州分公司软件工程师 > > 浙江省杭州市西湖区文三路478号华星时代广场A座507室 310013 > > Beijing WangXun Technology Co., Ltd. Software Engineer. > > Room A507, HuaXing Times Square, No.478 West Wensan Road. > > West Lake District, Hangzhou City, 310013 ZHEJIANG, P.R.CHINA. > > > > Office: +86(0571)89807901-8014 > > Mobile: +86-13656681762 > > E-Mail: z <mailto:jianwang@trustnetic.com>henpengzheng(a)net-swift.com > > **************************************************************** > > > *发件人：* QiuLaibin <mailto:qiulaibin@huawei.com> > *发送时间：* 2021-12-03 15:01 > *收件人：* zhenpengzheng(a)net-swift.com <mailto:zhenpengzheng@net-swift.com> > *抄送：* Xiexiuqi <mailto:xiexiuqi@huawei.com>; yangyingliang(a)huawei.com <mailto:yangyingliang@huawei.com> > *主题：* 【openEuler】Netswift Giga NIC驱动合入openEuler20.03 > pengzheng您好： > > 非常感谢您的提交！ > > 目前我们正在向版本合入先前您针对openEuler提交的Netswift Giga NIC驱补丁集： > > [openEuler-1.0-LTS,1/2] net: ngbe: Add Netswift Giga NIC driver > [openEuler-1.0-LTS,2/2] x86/config: Enable netswift Giga NIC driver for x86 > > 但鉴于您在编译config只在X86的config中打开，因此我们需要确认是否该驱动只能支持X86平台；arm上是否可以打开支持，若arm相关的config误打开是否会引入问题。 > > best regard > Laibin Qiu > > >

1 0

[PATCH openEuler-1.0-LTS] config: disable CONFIG_NGBE by default in hulk_defconfig
by Yang Yingliang 03 Dec '21

03 Dec '21

driver inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4C4XW?from=project-issue CVE: NA --------------------------------------- Disable CONFIG_NGBE by default CONFIG_NGBE on ARM64. Reviewed-by: Cheng Jian <cj.chengjian(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- arch/arm64/configs/hulk_defconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/arm64/configs/hulk_defconfig b/arch/arm64/configs/hulk_defconfig index fdf628f1fa028..e80f4b7fde56d 100644 --- a/arch/arm64/configs/hulk_defconfig +++ b/arch/arm64/configs/hulk_defconfig @@ -2491,6 +2491,7 @@ CONFIG_ICE=m CONFIG_FM10K=m CONFIG_NET_VENDOR_NETSWIFT=y CONFIG_TXGBE=m +# CONFIG_NGBE is not set # CONFIG_JME is not set # CONFIG_NET_VENDOR_MARVELL is not set CONFIG_NET_VENDOR_MELLANOX=y -- 2.25.1

1 0