March 2021 - Kernel - mailweb.openeuler.org

[PATCH openEuler-1.0-LTS] sysctl: control if check validity of freelist and connector debug by sysctl
by Cheng Jian 11 Mar '21

11 Mar '21

From: Yang Yingliang <yangyingliang(a)huawei.com> hulk inclusion category: bugfix bugzilla: 47452 CVE: NA ------------------------------------------------- Change control methods of check validity of freelist and connector debug to sysctl. Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> Reviewed-by: Kefeng Wang <wangkefeng.wang(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> Signed-off-by: Cheng Jian <cj.chengjian(a)huawei.com> --- drivers/connector/connector.c | 1 + include/linux/slab.h | 4 --- kernel/sysctl.c | 20 +++++++++++ mm/slub.c | 63 ++++++++++++++--------------------- 4 files changed, 46 insertions(+), 42 deletions(-) diff --git a/drivers/connector/connector.c b/drivers/connector/connector.c index a8df9ecbf42b..b5cdd0b99736 100644 --- a/drivers/connector/connector.c +++ b/drivers/connector/connector.c @@ -42,6 +42,7 @@ MODULE_ALIAS_NET_PF_PROTO(PF_NETLINK, NETLINK_CONNECTOR); static struct cn_dev cdev; static int cn_already_initialized; +int sysctl_connector_debug = 0; /* * Sends mult (multiple) cn_msg at a time. diff --git a/include/linux/slab.h b/include/linux/slab.h index 1ab5fd97dec8..d6393413ef09 100644 --- a/include/linux/slab.h +++ b/include/linux/slab.h @@ -24,8 +24,6 @@ */ /* DEBUG: Perform (expensive) checks on alloc/free */ #define SLAB_CONSISTENCY_CHECKS ((slab_flags_t __force)0x00000100U) -/* Check if freelist is valid */ -#define SLAB_NO_FREELIST_CHECKS ((slab_flags_t __force)0x00000200U) /* DEBUG: Red zone objs in a cache */ #define SLAB_RED_ZONE ((slab_flags_t __force)0x00000400U) /* DEBUG: Poison objects */ @@ -113,8 +111,6 @@ #define SLAB_KASAN 0 #endif -#define SLAB_CONNECTOR_DEBUG ((slab_flags_t __force)0x10000000U) - /* The following flags affect the page allocator grouping pages by mobility */ /* Objects are reclaimable */ #define SLAB_RECLAIM_ACCOUNT ((slab_flags_t __force)0x00020000U) diff --git a/kernel/sysctl.c b/kernel/sysctl.c index 665c9e2a8802..c921ee10615a 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -1245,6 +1245,7 @@ static struct ctl_table kern_table[] = { { } }; +extern int sysctl_isolate_corrupted_freelist; static struct ctl_table vm_table[] = { { .procname = "overcommit_memory", @@ -1714,6 +1715,15 @@ static struct ctl_table vm_table[] = { .extra2 = (void *)&mmap_rnd_compat_bits_max, }, #endif + { + .procname = "isolate_corrupted_freelist", + .data = &sysctl_isolate_corrupted_freelist, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec_minmax, + .extra1 = &zero, + .extra2 = &one, + }, { } }; @@ -1925,6 +1935,7 @@ static struct ctl_table fs_table[] = { { } }; +extern int sysctl_connector_debug; static struct ctl_table debug_table[] = { #ifdef CONFIG_SYSCTL_EXCEPTION_TRACE { @@ -1946,6 +1957,15 @@ static struct ctl_table debug_table[] = { .extra2 = &one, }, #endif + { + .procname = "connector-debug", + .data = &sysctl_connector_debug, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec_minmax, + .extra1 = &zero, + .extra2 = &one, + }, { } }; diff --git a/mm/slub.c b/mm/slub.c index ee0d18699bde..2301c04353d0 100644 --- a/mm/slub.c +++ b/mm/slub.c @@ -1221,30 +1221,6 @@ static noinline int free_debug_processing( return ret; } -/* - * Check if the next object in freechain is - * valid, then isolate the corrupted freelist. - */ -int isolate_cnt = 0; -static bool isolate_corrupted_freelist(struct kmem_cache *s, struct page *page, - void **freelist, void *next) -{ - if (!(slub_debug & SLAB_NO_FREELIST_CHECKS) && - (!check_valid_pointer(s, page, next))) { - /* Need caller make sure freelist is not NULL */ - *freelist = NULL; - isolate_cnt++; - - slab_fix(s, "Freelist corrupt,isolate corrupted freechain."); - pr_info("freelist=%lx object=%lx, page->base=%lx, page->objects=%u, objsize=%u, s->offset=%d page:%llx\n", - (long)freelist, (long)next, (long)page_address(page), page->objects, s->size, s->offset, (u64)page); - - return true; - } - - return false; -} - static int __init setup_slub_debug(char *str) { slub_debug = DEBUG_DEFAULT_FLAGS; @@ -1291,14 +1267,6 @@ static int __init setup_slub_debug(char *str) case 'a': slub_debug |= SLAB_FAILSLAB; break; - case 'n': - slub_debug |= SLAB_NO_FREELIST_CHECKS; - pr_info("Freelist pointer check disabled."); - break; - case 'd': - slub_debug |= SLAB_CONNECTOR_DEBUG; - pr_info("Connector debug enabled.\n"); - break; case 'o': /* * Avoid enabling debugging on caches if its minimum @@ -1311,8 +1279,6 @@ static int __init setup_slub_debug(char *str) *str); } } - if (!(slub_debug & SLAB_NO_FREELIST_CHECKS)) - pr_info("Freelist pointer check enabled."); check_slabs: if (*str == ',') @@ -1380,12 +1346,33 @@ static bool freelist_corrupted(struct kmem_cache *s, struct page *page, { return false; } +#endif /* CONFIG_SLUB_DEBUG */ + +extern int sysctl_connector_debug; +/* + * Check if the next object in freechain is + * valid, then isolate the corrupted freelist. + */ +int isolate_cnt = 0; +int sysctl_isolate_corrupted_freelist = 1; static bool isolate_corrupted_freelist(struct kmem_cache *s, struct page *page, - void **freelist, void *nextfree) + void **freelist, void *next) { + if (sysctl_isolate_corrupted_freelist && + (!check_valid_pointer(s, page, next))) { + /* Need caller make sure freelist is not NULL */ + *freelist = NULL; + isolate_cnt++; + + slab_fix(s, "Freelist corrupt,isolate corrupted freechain."); + pr_info("freelist=%lx object=%lx, page->base=%lx, page->objects=%u, objsize=%u, s->offset=%d page:%llx\n", + (long)freelist, (long)next, (long)page_address(page), page->objects, s->size, s->offset, (u64)page); + + return true; + } + return false; } -#endif /* CONFIG_SLUB_DEBUG */ /* * Hooks for other subsystems that check memory allocations. In a typical @@ -2798,7 +2785,7 @@ static __always_inline void *slab_alloc_node(struct kmem_cache *s, if (unlikely(gfpflags & __GFP_ZERO) && object) memset(object, 0, s->object_size); - if ((slub_debug & SLAB_CONNECTOR_DEBUG) && + if (sysctl_connector_debug && unlikely(gfpflags & __GFP_CONNECTOR) && object) { if (s->object_size == 512) memset(object + 504, 0xad, 8); @@ -4015,7 +4002,7 @@ void kfree(const void *x) __free_pages(page, compound_order(page)); return; } - if ((slub_debug & SLAB_CONNECTOR_DEBUG) && + if (sysctl_connector_debug && unlikely(page->slab_cache->object_size == 512)) { u64 *tail = (u64 *)(x + 504); -- 2.25.1

1 0

[PATCH openEuler-1.0-LTS 00/82] bugfix for 20.03 at 202102
by Cheng Jian 11 Mar '21

11 Mar '21

bugfix for 20.03 at 202102 Alexey Gladkov (1): moduleparam: Save information about built-in modules in separate file Amir Goldstein (1): ovl: skip getxattr of security labels Aurelien Aptel (1): cifs: report error instead of invalid when revalidating a dentry fails Baptiste Lepers (1): udp: Prevent reuseport_select_sock from reading uninitialized socks Byron Stanoszek (1): tmpfs: restore functionality of nr_inodes=0 Chris Down (2): tmpfs: per-superblock i_ino support tmpfs: support 64-bit inums per-sb Christian Brauner (1): sysctl: handle overflow in proc_get_long Cong Wang (1): af_key: relax availability checks for skb size calculation Dave Wysochanski (2): SUNRPC: Move simple_get_bytes and simple_get_netobj into private header SUNRPC: Handle 0 length opaque XDR object data properly Dong Kai (1): livepatch/core: Fix jump_label_apply_nops called multi times Edwin Peer (1): net: watchdog: hold device global xmit lock during tx disable Eric Biggers (1): fs: fix lazytime expiration handling in __writeback_single_inode() Eric Dumazet (1): net_sched: gen_estimator: support large ewma log Eyal Birger (1): xfrm: fix disable_xfrm sysctl when used on xfrm interfaces Florian Westphal (1): netfilter: conntrack: skip identical origin tuple in same zone only Gaurav Kohli (1): tracing: Fix race in trace_open and buffer resize call Gustavo A. R. Silva (1): smb3: Fix out-of-bounds bug in SMB2_negotiate() Hugh Dickins (1): mm: thp: fix MADV_REMOVE deadlock on shmem THP Jakub Kicinski (1): net: sit: unregister_netdevice on newlink's error path Jan Kara (1): writeback: Drop I_DIRTY_TIME_EXPIRE Jozsef Kadlecsik (1): netfilter: xt_recent: Fix attempt to update deleted entry Liangyan (1): ovl: fix dentry leak in ovl_get_redirect Lin Feng (1): bfq-iosched: Revert "bfq: Fix computation of shallow depth" Marc Zyngier (1): genirq/msi: Activate Multi-MSI early when MSI_FLAG_ACTIVATE_EARLY is set Martin K. Petersen (2): scsi: sd: block: Fix regressions in read-only block device handling scsi: sd: block: Fix kabi change by 'scsi: sd: block: Fix regressions in read-only block device handling' Martin Willi (1): vrf: Fix fast path output packet handling with async Netfilter rules Masami Hiramatsu (1): tracing/kprobe: Fix to support kretprobe events on unloaded modules Miklos Szeredi (4): proc/mounts: add cursor ovl: perform vfs_getxattr() with mounter creds cap: fix conversions on getxattr ovl: expand warning in ovl_d_real() Mikulas Patocka (1): dm integrity: conditionally disable "recalculate" feature Ming Lei (4): scsi: core: Run queue in case of I/O resource contention failure scsi: core: Only re-run queue in scsi_end_request() if device queue is busy block: don't hold q->sysfs_lock in elevator_init_mq blk-mq: don't hold q->sysfs_lock in blk_mq_map_swqueue Muchun Song (4): mm: hugetlbfs: fix cannot migrate the fallocated HugeTLB page mm: hugetlb: fix a race between freeing and dissolving the page mm: hugetlb: fix a race between isolating and freeing page mm: hugetlb: remove VM_BUG_ON_PAGE from page_huge_active NeilBrown (1): net: fix iteration for sctp transport seq_files Ondrej Jirman (1): brcmfmac: Loading the correct firmware for brcm43456 Pablo Neira Ayuso (1): netfilter: nft_dynset: add timeout extension to template Pavel Begunkov (1): list: introduce list_for_each_continue() Pengcheng Yang (1): tcp: fix TLP timer not set when CA_STATE changes from DISORDER to OPEN Peter Zijlstra (2): kthread: Extract KTHREAD_IS_PER_CPU workqueue: Restrict affinity change to rescuer Roi Dayan (1): net/mlx5: Fix memory leak on flow table creation error flow Roman Gushchin (1): memblock: do not start bottom-up allocations with kernel_end Sabyrzhan Tasbolatov (1): net/rds: restrict iovecs length for RDS_CMSG_RDMA_ARGS Shmulik Ladkani (1): xfrm: Fix oops in xfrm_replay_advance_bmp Steven Rostedt (VMware) (3): fgraph: Initialize tracing_graph_pause at task creation tracing: Do not count ftrace events in top level enable output tracing: Check length before giving out the filter buffer Sven Auhagen (1): netfilter: flowtable: fix tcp and udp header checksum update Vadim Fedorenko (1): net: ip_tunnel: fix mtu calculation Wang Hai (1): Revert "mm/slub: fix a memory leak in sysfs_slab_add()" Wang ShaoBo (1): kretprobe: Avoid re-registration of the same kretprobe earlier Willem de Bruijn (1): esp: avoid unneeded kmap_atomic call Xiao Ni (1): md: Set prev_flush_start and flush_bio in an atomic way Yang Yingliang (3): sysctl/mm: fix compile error when CONFIG_SLUB is disabled config: disable config TMPFS_INODE64 by default connector: change GFP_CONNECTOR to bit-31 Ye Bin (3): scsi: sd: block: Fix read-only flag residuals when partition table change Revert "scsi: sd: block: Fix read-only flag residuals when partition table change" Revert "scsi: sg: fix memory leak in sg_build_indirect" Yonglong Liu (4): net: hns3: adds support for setting pf max tx rate via sysfs net: hns3: update hns3 version to 1.9.38.10 net: hns3: fix 'ret' may be used uninitialized problem net: hns3: update hns3 version to 1.9.38.11 Yufen Yu (1): scsi: fix kabi for scsi_device Zhang Xiaoxu (1): proc/mounts: Fix kabi broken liubo (3): etmem: add etmem-scan feature etmem: add etmem-swap feature config: Enable the config option of the etmem feature shiyongbang (3): gpu: hibmc: Fix erratic display during startup stage. gpu: hibmc: Use drm get pci dev api. gpu: hibmc: Fix stuck when switch GUI to text. zhangyi (F) (1): ext4: find old entry again if failed to rename whiteout .gitignore | 1 + Documentation/device-mapper/dm-integrity.txt | 7 + Documentation/dontdiff | 1 + Documentation/filesystems/tmpfs.txt | 17 + Documentation/kbuild/kbuild.txt | 5 + Makefile | 2 + arch/arm64/configs/hulk_defconfig | 3 + arch/arm64/configs/openeuler_defconfig | 2 + arch/x86/configs/hulk_defconfig | 2 + arch/x86/configs/openeuler_defconfig | 2 + block/bfq-iosched.c | 8 +- block/blk-mq.c | 7 - block/elevator.c | 14 +- block/genhd.c | 33 +- block/ioctl.c | 4 + block/partition-generic.c | 7 +- drivers/connector/connector.c | 6 +- .../gpu/drm/hisilicon/hibmc/hibmc_drm_drv.c | 68 +- .../gpu/drm/hisilicon/hibmc/hibmc_drm_drv.h | 2 + .../gpu/drm/hisilicon/hibmc/hibmc_drm_vdac.c | 1 + drivers/md/dm-integrity.c | 24 +- drivers/md/md.c | 2 + drivers/net/ethernet/hisilicon/hns3/Makefile | 3 +- drivers/net/ethernet/hisilicon/hns3/hnae3.h | 10 +- .../hns3/hns3_cae/hns3_cae_version.h | 2 +- .../net/ethernet/hisilicon/hns3/hns3_enet.h | 2 +- .../hns3_extension/hns3pf/hclge_main_it.c | 33 + .../hns3/hns3_extension/hns3pf/hclge_sysfs.c | 140 +++ .../hns3/hns3_extension/hns3pf/hclge_sysfs.h | 14 + .../hisilicon/hns3/hns3pf/hclge_main.c | 17 + .../hisilicon/hns3/hns3pf/hclge_main.h | 7 +- .../hisilicon/hns3/hns3vf/hclgevf_main.h | 2 +- .../net/ethernet/mellanox/mlx5/core/fs_core.c | 1 + drivers/net/vrf.c | 92 +- .../broadcom/brcm80211/brcmfmac/sdio.c | 4 +- drivers/scsi/scsi_lib.c | 59 +- drivers/scsi/sg.c | 6 +- fs/Kconfig | 21 + fs/cifs/dir.c | 22 +- fs/cifs/smb2pdu.h | 2 +- fs/ext4/inode.c | 2 +- fs/ext4/namei.c | 29 +- fs/fs-writeback.c | 36 +- fs/hugetlbfs/inode.c | 3 +- fs/mount.h | 17 +- fs/namespace.c | 119 +- fs/overlayfs/copy_up.c | 15 +- fs/overlayfs/dir.c | 2 +- fs/overlayfs/inode.c | 2 + fs/overlayfs/super.c | 13 +- fs/proc/Makefile | 2 + fs/proc/base.c | 4 + fs/proc/etmem_scan.c | 1046 +++++++++++++++++ fs/proc/etmem_scan.h | 132 +++ fs/proc/etmem_swap.c | 102 ++ fs/proc/internal.h | 2 + fs/proc/task_mmu.c | 117 ++ fs/proc_namespace.c | 4 +- fs/xfs/xfs_trans_inode.c | 4 +- include/asm-generic/vmlinux.lds.h | 1 + include/linux/fs.h | 16 +- include/linux/genhd.h | 5 + include/linux/gfp.h | 8 +- include/linux/hugetlb.h | 3 + include/linux/kprobes.h | 2 +- include/linux/kthread.h | 3 + include/linux/list.h | 10 + include/linux/mm_types.h | 18 + include/linux/module.h | 1 + include/linux/moduleparam.h | 12 +- include/linux/mount.h | 4 +- include/linux/msi.h | 6 + include/linux/netdevice.h | 2 + include/linux/shmem_fs.h | 3 + include/linux/sunrpc/xdr.h | 3 +- include/linux/swap.h | 5 + include/net/tcp.h | 2 +- include/scsi/scsi_device.h | 3 +- include/trace/events/writeback.h | 1 - init/init_task.c | 3 +- kernel/irq/msi.c | 44 +- kernel/kprobes.c | 36 +- kernel/kthread.c | 27 +- kernel/livepatch/core.c | 20 +- kernel/smpboot.c | 1 + kernel/sysctl.c | 6 + kernel/trace/ftrace.c | 2 - kernel/trace/ring_buffer.c | 4 + kernel/trace/trace.c | 2 +- kernel/trace/trace_events.c | 3 +- kernel/trace/trace_kprobe.c | 4 +- kernel/workqueue.c | 9 +- lib/Kconfig | 11 + mm/huge_memory.c | 37 +- mm/hugetlb.c | 48 +- mm/memblock.c | 49 +- mm/pagewalk.c | 1 + mm/shmem.c | 133 ++- mm/slub.c | 14 +- mm/vmscan.c | 112 ++ mm/vmstat.c | 4 + net/core/gen_estimator.c | 11 +- net/core/sock_reuseport.c | 2 +- net/ipv4/esp4.c | 7 +- net/ipv4/ip_tunnel.c | 16 +- net/ipv4/tcp_input.c | 10 +- net/ipv4/tcp_recovery.c | 5 +- net/ipv6/esp6.c | 7 +- net/ipv6/sit.c | 5 +- net/key/af_key.c | 6 +- net/netfilter/nf_conntrack_core.c | 3 +- net/netfilter/nf_flow_table_core.c | 4 +- net/netfilter/nft_dynset.c | 4 +- net/netfilter/xt_recent.c | 12 +- net/rds/rdma.c | 3 + net/sctp/proc.c | 16 +- net/sunrpc/auth_gss/auth_gss.c | 30 +- net/sunrpc/auth_gss/auth_gss_internal.h | 45 + net/sunrpc/auth_gss/gss_krb5_mech.c | 31 +- net/xfrm/xfrm_input.c | 2 +- net/xfrm/xfrm_policy.c | 4 +- scripts/link-vmlinux.sh | 3 + security/commoncap.c | 67 +- virt/kvm/kvm_main.c | 6 + 124 files changed, 2817 insertions(+), 466 deletions(-) create mode 100644 drivers/net/ethernet/hisilicon/hns3/hns3_extension/hns3pf/hclge_sysfs.c create mode 100644 drivers/net/ethernet/hisilicon/hns3/hns3_extension/hns3pf/hclge_sysfs.h create mode 100644 fs/proc/etmem_scan.c create mode 100644 fs/proc/etmem_scan.h create mode 100644 fs/proc/etmem_swap.c create mode 100644 net/sunrpc/auth_gss/auth_gss_internal.h -- 2.25.1

1 82

[PATCH 01/82] list: introduce list_for_each_continue()
by Cheng Jian 11 Mar '21

11 Mar '21

From: Pavel Begunkov <asml.silence(a)gmail.com> mainline inclusion from mainline-v5.6-rc1 commit 28ca0d6d39ab1d01c86762c82a585b7cedd2920c category: bugfix bugzilla: 35619 CVE: NA -------------------------------- As other *continue() helpers, this continues iteration from a given position. Signed-off-by: Pavel Begunkov <asml.silence(a)gmail.com> Signed-off-by: Jens Axboe <axboe(a)kernel.dk> Signed-off-by: Zhang Xiaoxu <zhangxiaoxu5(a)huawei.com> Reviewed-by: zhangyi (F) <yi.zhang(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- include/linux/list.h | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/include/linux/list.h b/include/linux/list.h index de04cc5ed536..0e540581d52c 100644 --- a/include/linux/list.h +++ b/include/linux/list.h @@ -455,6 +455,16 @@ static inline void list_splice_tail_init(struct list_head *list, #define list_for_each(pos, head) \ for (pos = (head)->next; pos != (head); pos = pos->next) +/** + * list_for_each_continue - continue iteration over a list + * @pos: the &struct list_head to use as a loop cursor. + * @head: the head for your list. + * + * Continue to iterate over a list, continuing after the current position. + */ +#define list_for_each_continue(pos, head) \ + for (pos = pos->next; pos != (head); pos = pos->next) + /** * list_for_each_prev - iterate over a list backwards * @pos: the &struct list_head to use as a loop cursor. -- 2.25.1

1 11

[PATCH openEuler-21.03 v1] park: Reserve park mem before kexec reserved
by sangyan＠huawei.com 10 Mar '21

10 Mar '21

From: Sang Yan <sangyan(a)huawei.com> hulk inclusion category: feature bugzilla: 48159 CVE: N/A ------------------------------ reserve_crashkernel or reserve_quick_kexec may find one sutiable memory region and reserves it, which address of the region is not fixed. As a result, cpu park reserves memory could be failed while specified address used by crashkernel or quickkexec. So, move reserve_park_mem before reserve_crashkernel and reserve_quick_kexec. Signed-off-by: Sang Yan <sangyan(a)huawei.com> --- arch/arm64/mm/init.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c index b343744..043a981 100644 --- a/arch/arm64/mm/init.c +++ b/arch/arm64/mm/init.c @@ -497,16 +497,16 @@ void __init arm64_memblock_init(void) else arm64_dma32_phys_limit = PHYS_MASK + 1; +#ifdef CONFIG_ARM64_CPU_PARK + reserve_park_mem(); +#endif + reserve_crashkernel(); #ifdef CONFIG_QUICK_KEXEC reserve_quick_kexec(); #endif -#ifdef CONFIG_ARM64_CPU_PARK - reserve_park_mem(); -#endif - reserve_pin_memory_res(); reserve_elfcorehdr(); -- 2.9.5

1 0

[PATCH openEuler-21.03 v4 1/2] arm64: Add memmap parameter and register pmem
by Zhuling 10 Mar '21

10 Mar '21

From: ZhuLing <zhuling8(a)huawei.com> hulk inclusion category: feature bugzilla: 48159 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 | 21 +++++++++ arch/arm64/kernel/Makefile | 1 + arch/arm64/kernel/pmem.c | 35 ++++++++++++++ arch/arm64/kernel/setup.c | 10 ++++ arch/arm64/mm/init.c | 97 ++++++++++++++++++++++++++++++++++++++ drivers/nvdimm/Kconfig | 5 ++ drivers/nvdimm/Makefile | 2 +- 7 files changed, 170 insertions(+), 1 deletion(-) create mode 100644 arch/arm64/kernel/pmem.c diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig index c451137ab..326f26d40 100644 --- a/arch/arm64/Kconfig +++ b/arch/arm64/Kconfig @@ -360,6 +360,27 @@ config ARM64_CPU_PARK config ARCH_HAS_CPU_RELAX def_bool y +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. + source "arch/arm64/Kconfig.platforms" menu "Kernel Features" diff --git a/arch/arm64/kernel/Makefile b/arch/arm64/kernel/Makefile index 967cb3c6d..be996f3c1 100644 --- a/arch/arm64/kernel/Makefile +++ b/arch/arm64/kernel/Makefile @@ -67,6 +67,7 @@ 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 000000000..16eaf706f --- /dev/null +++ b/arch/arm64/kernel/pmem.c @@ -0,0 +1,35 @@ +// 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 5e282d31a..84c71c88d 100644 --- a/arch/arm64/kernel/setup.c +++ b/arch/arm64/kernel/setup.c @@ -57,6 +57,10 @@ static int num_standard_resources; static struct resource *standard_resources; +#ifdef CONFIG_ARM64_PMEM_RESERVE +extern struct resource pmem_res; +#endif + phys_addr_t __fdt_pointer __initdata; /* @@ -270,6 +274,12 @@ 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 b3437440d..f22faea1a 100644 --- a/arch/arm64/mm/init.c +++ b/arch/arm64/mm/init.c @@ -66,6 +66,18 @@ EXPORT_SYMBOL(memstart_addr); phys_addr_t arm64_dma_phys_limit __ro_after_init; phys_addr_t arm64_dma32_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) { @@ -378,6 +390,87 @@ static int __init reserve_park_mem(void) } #endif +static int __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 -EINVAL; + } + + if (memblock_is_region_reserved(mem_start, mem_size)) { + pr_warn("cannot reserve mem: region overlaps reserved memory!\n"); + return -EINVAL; + } + + if (!IS_ALIGNED(mem_start, SZ_2M)) { + pr_warn("cannot reserve mem: base address is not 2MB aligned!\n"); + return -EINVAL; + } + + return 0; +} + +static int __init parse_memmap_one(char *p) +{ + char *oldp; + phys_addr_t start_at, mem_size; + + 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); + + if (is_mem_valid(mem_size, start_at) != 0) + return -EINVAL; + + pr_info("pmem reserved: 0x%016llx - 0x%016llx (%lld MB)\n", + start_at, start_at + mem_size, mem_size >> 20); + 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) +{ + 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); @@ -511,6 +604,10 @@ void __init arm64_memblock_init(void) reserve_elfcorehdr(); +#ifdef CONFIG_ARM64_PMEM_RESERVE + reserve_pmem(); +#endif + high_memory = __va(memblock_end_of_DRAM() - 1) + 1; dma_contiguous_reserve(arm64_dma32_phys_limit); diff --git a/drivers/nvdimm/Kconfig b/drivers/nvdimm/Kconfig index b7d1eb38b..ce4de7526 100644 --- a/drivers/nvdimm/Kconfig +++ b/drivers/nvdimm/Kconfig @@ -132,3 +132,8 @@ 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 29203f3d3..6f8dc9242 100644 --- a/drivers/nvdimm/Makefile +++ b/drivers/nvdimm/Makefile @@ -3,7 +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-$(CONFIG_X86_PMEM_LEGACY) += nd_e820.o +obj-$(CONFIG_PMEM_LEGACY) += nd_e820.o obj-$(CONFIG_OF_PMEM) += of_pmem.o obj-$(CONFIG_VIRTIO_PMEM) += virtio_pmem.o nd_virtio.o -- 2.19.1

1 1

(no subject)
by 点击设置昵称 09 Mar '21

09 Mar '21

ok!

1 0

openEuler kernel sig meeting
by openEuler conference 08 Mar '21

08 Mar '21

4 7

[PATCH] kbuild: fix compile error in Makefile.lib
by fangyafenqidai＠163.com 08 Mar '21

08 Mar '21

From: Fang Yafen <yafen(a)iscas.ac.cn> Fix the following compile error when using bcm2711_defconfig (for RPi). /bin/sh: -c: line 0: syntax error near unexpected token `(' /bin/sh: -c: line 0: `set -e; echo ' DTCO arch/arm64/boot/dts/overlays/act-led.dtbo'; mkdir -p arch/arm64/boot/dts/overlays/ ; gcc -E -Wp,-MMD,arch/arm64/boot/dts/overlays/.act-led.dtbo.d.pre.tmp -nostdinc -I./scripts/dtc/include-prefixes -undef -D__DTS__ -x assembler-with-cpp -o arch/arm64/boot/dts/overlays/.act-led.dtbo.dts.tmp arch/arm64/boot/dts/overlays/act-led-overlay.dts ; ./scripts/dtc/dtc -@ -H epapr -O dtb -o arch/arm64/boot/dts/overlays/act-led.dtbo -b 0 -i arch/arm64/boot/dts/overlays/ -Wno-interrupt_provider -Wno-unit_address_vs_reg -Wno-unit_address_format -Wno-gpios_property -Wno-avoid_unnecessary_addr_size -Wno-alias_paths -Wno-graph_child_address -Wno-simple_bus_reg -Wno-unique_unit_address -Wno-pci_device_reg -Wno-interrupts_property ifeq (y,y) -Wno-label_is_string -Wno-reg_format -Wno-pci_device_bus_num -Wno-i2c_bus_reg -Wno-spi_bus_reg -Wno-avoid_default_addr_size endif -d arch/arm64/boot/dts/overlays/.act-led.dtbo.d.dtc.tmp arch/arm64/boot/dts/overlays/.act-led.dtbo.dts.tmp ; cat ...; rm -f arch/arm64/boot/dts/overlays/.act-led.dtbo.d' make[2]: *** [scripts/Makefile.lib;363: arch/arm64/boot/dts/overlays/act-led.dtbo] Error 1 make[2]: *** Waiting for unfinished jobs.... Related patches: ffa2d13ccc3 BCM2708: Add core Device Tree support 4894352ec98 kbuild: Silence unavoidable dtc overlay warnings a4a4d07f0cf kbuild: keep the original function for non-RPi Signed-off-by: Fang Yafen <yafen(a)iscas.ac.cn> --- scripts/Makefile.lib | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/scripts/Makefile.lib b/scripts/Makefile.lib index c23e3ae7ef40..a0e0e2543165 100644 --- a/scripts/Makefile.lib +++ b/scripts/Makefile.lib @@ -342,25 +342,25 @@ endef $(obj)/%.dt.yaml: $(src)/%.dts $(DTC) $(DT_TMP_SCHEMA) FORCE $(call if_changed_rule,dtc,yaml) +ifeq ($(CONFIG_OPENEULER_RASPBERRYPI),y) quiet_cmd_dtco = DTCO $@ cmd_dtco = mkdir -p $(dir ${dtc-tmp}) ; \ $(CPP) $(dtc_cpp_flags) -x assembler-with-cpp -o $(dtc-tmp) $< ; \ $(DTC) -@ -H epapr -O dtb -o $@ -b 0 \ -i $(dir $<) $(DTC_FLAGS) \ -Wno-interrupts_property \ -ifeq ($(CONFIG_OPENEULER_RASPBERRYPI),y) \ -Wno-label_is_string \ -Wno-reg_format \ -Wno-pci_device_bus_num \ -Wno-i2c_bus_reg \ -Wno-spi_bus_reg \ -Wno-avoid_default_addr_size \ -endif \ -d $(depfile).dtc.tmp $(dtc-tmp) ; \ cat $(depfile).pre.tmp $(depfile).dtc.tmp > $(depfile) $(obj)/%.dtbo: $(src)/%-overlay.dts FORCE $(call if_changed_dep,dtco) +endif dtc-tmp = $(subst $(comma),_,$(dot-target).dts.tmp) -- 2.27.0

2 1

[PATCH kernel-4.19 01/32] ascend: sharepool: don't enable the vmalloc to use hugepage default
by Yang Yingliang 08 Mar '21

08 Mar '21

From: Ding Tianhong <dingtianhong(a)huawei.com> ascend inclusion category: feature bugzilla: NA CVE: NA ------------------------------------------------- The commit 59a57a82fb2a ("mm/vmalloc: Hugepage vmalloc mappings") would enable the vmalloc for hugepage default when the alloc size is bigger than the PMD_SIZE, it looks like the transparent hugepage for mmap, the driver could not control the hugepage accurately and be break the logic, now the share pool already export the vmalloc_hugepage_xxx function to control the vmalloc hugepage allocation, it looks like the static hugepage for vmalloc, so disable the transparent hugepage function. This patch also fix the problem of breaking the kabi of vm_struct, the user could applied it for commercial version. Fixes: 59a57a82fb2a ("mm/vmalloc: Hugepage vmalloc mappings") Signed-off-by: Ding Tianhong <dingtianhong(a)huawei.com> Reviewed-by: Kefeng Wang <wangkefeng.wang(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- include/linux/share_pool.h | 51 ++++++++++++++++++++++++++++---------- include/linux/vmalloc.h | 1 - mm/vmalloc.c | 47 ++++++++++++----------------------- 3 files changed, 54 insertions(+), 45 deletions(-) diff --git a/include/linux/share_pool.h b/include/linux/share_pool.h index c3120b7b24948..4a18c88d5a10e 100644 --- a/include/linux/share_pool.h +++ b/include/linux/share_pool.h @@ -211,15 +211,6 @@ static inline void sp_area_work_around(struct vm_unmapped_area_info *info) extern struct page *sp_alloc_pages(struct vm_struct *area, gfp_t mask, unsigned int page_order, int node); - -static inline void sp_free_pages(struct page *page, struct vm_struct *area) -{ - if (PageHuge(page)) - put_page(page); - else - __free_pages(page, area->page_order); -} - static inline bool sp_check_vm_share_pool(unsigned long vm_flags) { if (enable_ascend_share_pool && (vm_flags & VM_SHARE_POOL)) @@ -264,6 +255,30 @@ extern void *buff_vzalloc_hugepage_user(unsigned long size); void sp_exit_mm(struct mm_struct *mm); +static inline bool is_vmalloc_huge(unsigned long vm_flags) +{ + if (enable_ascend_share_pool && (vm_flags & VM_HUGE_PAGES)) + return true; + + return false; +} + +static inline bool is_vmalloc_sharepool(unsigned long vm_flags) +{ + if (enable_ascend_share_pool && (vm_flags & VM_SHAREPOOL)) + return true; + + return false; +} + +static inline void sp_free_pages(struct page *page, struct vm_struct *area) +{ + if (PageHuge(page)) + put_page(page); + else + __free_pages(page, is_vmalloc_huge(area->flags) ? PMD_SHIFT - PAGE_SHIFT : 0); +} + #else static inline int sp_group_add_task(int pid, int spg_id) @@ -400,10 +415,6 @@ static inline struct page *sp_alloc_pages(void *area, gfp_t mask, return NULL; } -static inline void sp_free_pages(struct page *page, struct vm_struct *area) -{ -} - static inline bool sp_check_vm_share_pool(unsigned long vm_flags) { return false; @@ -448,6 +459,20 @@ static inline void *buff_vzalloc_hugepage_user(unsigned long size) return NULL; } +static inline bool is_vmalloc_huge(struct vm_struct *vm) +{ + return NULL; +} + +static inline bool is_vmalloc_sharepool(struct vm_struct *vm) +{ + return NULL; +} + +static inline void sp_free_pages(struct page *page, struct vm_struct *area) +{ +} + #endif #endif /* LINUX_SHARE_POOL_H */ diff --git a/include/linux/vmalloc.h b/include/linux/vmalloc.h index bb814f6418fd9..298eff5579b21 100644 --- a/include/linux/vmalloc.h +++ b/include/linux/vmalloc.h @@ -43,7 +43,6 @@ struct vm_struct { unsigned long size; unsigned long flags; struct page **pages; - unsigned int page_order; unsigned int nr_pages; phys_addr_t phys_addr; const void *caller; diff --git a/mm/vmalloc.c b/mm/vmalloc.c index 37b4762871142..8c70131e0b078 100644 --- a/mm/vmalloc.c +++ b/mm/vmalloc.c @@ -2354,6 +2354,7 @@ struct vm_struct *remove_vm_area(const void *addr) static void __vunmap(const void *addr, int deallocate_pages) { struct vm_struct *area; + unsigned int page_order = 0; if (!addr) return; @@ -2369,13 +2370,14 @@ static void __vunmap(const void *addr, int deallocate_pages) return; } -#ifdef CONFIG_ASCEND_SHARE_POOL /* unmap a sharepool vm area will cause meamleak! */ - if (area->flags & VM_SHAREPOOL) { + if (is_vmalloc_sharepool(area->flags)) { WARN(1, KERN_ERR "Memory leak due to vfree() sharepool vm area (%p) !\n", addr); return; } -#endif + + if (is_vmalloc_huge(area->flags)) + page_order = PMD_SHIFT - PAGE_SHIFT; debug_check_no_locks_freed(area->addr, get_vm_area_size(area)); debug_check_no_obj_freed(area->addr, get_vm_area_size(area)); @@ -2384,14 +2386,14 @@ static void __vunmap(const void *addr, int deallocate_pages) if (deallocate_pages) { int i; - for (i = 0; i < area->nr_pages; i += 1U << area->page_order) { + for (i = 0; i < area->nr_pages; i += 1U << page_order) { struct page *page = area->pages[i]; BUG_ON(!page); if (sp_is_enabled()) sp_free_pages(page, area); else - __free_pages(page, area->page_order); + __free_pages(page, page_order); } kvfree(area->pages); @@ -2589,7 +2591,6 @@ static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask, area->pages = pages; area->nr_pages = nr_pages; - area->page_order = page_order; for (i = 0; i < area->nr_pages; i += 1U << page_order) { struct page *page; @@ -2657,27 +2658,17 @@ void *__vmalloc_node_range(unsigned long size, unsigned long align, if (!size || (size >> PAGE_SHIFT) > totalram_pages) goto fail; - if (vmap_allow_huge && (pgprot_val(prot) == pgprot_val(PAGE_KERNEL))) { - unsigned long size_per_node; - + if (vmap_allow_huge && (pgprot_val(prot) == pgprot_val(PAGE_KERNEL)) && is_vmalloc_huge(vm_flags)) { /* - * Try huge pages. Only try for PAGE_KERNEL allocations, - * others like modules don't yet expect huge pages in - * their allocations due to apply_to_page_range not - * supporting them. + * Alloc huge pages. Only valid for PAGE_KERNEL allocations and + * VM_HUGE_PAGES flags. */ - size_per_node = size; - if (node == NUMA_NO_NODE && !sp_is_enabled()) - size_per_node /= num_online_nodes(); - if (size_per_node >= PMD_SIZE) { - shift = PMD_SHIFT; - align = max(real_align, 1UL << shift); - size = ALIGN(real_size, 1UL << shift); - } + shift = PMD_SHIFT; + align = max(real_align, 1UL << shift); + size = ALIGN(real_size, 1UL << shift); } -again: size = PAGE_ALIGN(size); area = __get_vm_area_node(size, align, VM_ALLOC | VM_UNINITIALIZED | vm_flags, start, end, node, gfp_mask, caller); @@ -2706,12 +2697,6 @@ void *__vmalloc_node_range(unsigned long size, unsigned long align, return addr; fail: - if (shift > PAGE_SHIFT) { - shift = PAGE_SHIFT; - align = real_align; - size = real_size; - goto again; - } if (!area) { /* Warn for area allocation, page allocations already warn */ @@ -3776,7 +3761,7 @@ static int s_show(struct seq_file *m, void *p) seq_printf(m, " %pS", v->caller); if (v->nr_pages) - seq_printf(m, " pages=%d order=%d", v->nr_pages, v->page_order); + seq_printf(m, " pages=%d", v->nr_pages); if (v->phys_addr) seq_printf(m, " phys=%pa", &v->phys_addr); @@ -3796,8 +3781,8 @@ static int s_show(struct seq_file *m, void *p) if (is_vmalloc_addr(v->pages)) seq_puts(m, " vpages"); - if (sp_is_enabled()) - seq_printf(m, " order=%d", v->page_order); + if (is_vmalloc_huge(v->flags)) + seq_printf(m, " order=%d", PMD_SHIFT - PAGE_SHIFT); show_numa_info(m, v); seq_putc(m, '\n'); -- 2.25.1

1 31

[PATCH kernel-4.19] Certs: Fix a out-of-tree build Error
by Zhichang Yuan 08 Mar '21

08 Mar '21

openEuler inclusion category: bugfix bugzilla: https://bugzilla.openeuler.org/show_bug.cgi?id=6 CVE: NA When performing the out-of-tree kernel 4.19 building, some error happened saying: cannot touch 'certs/pubring.gpg'. The detail can be checked in Gitee issue #I39160 or Bug 6 of openeuler bugzilla. This patch will fix that build error. Signed-off-by: Zhichang Yuan <erik.yuan(a)arm.com> Signed-off-by: Steve Capper <Steve.Capper(a)arm.com> --- certs/Makefile | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/certs/Makefile b/certs/Makefile index 5053e3c86c97..5e56f9405758 100644 --- a/certs/Makefile +++ b/certs/Makefile @@ -5,11 +5,12 @@ obj-$(CONFIG_SYSTEM_TRUSTED_KEYRING) += system_keyring.o system_certificates.o ifdef CONFIG_PGP_PRELOAD_PUBLIC_KEYS -ifneq ($(shell ls certs/pubring.gpg 2> /dev/null), certs/pubring.gpg) -$(shell touch certs/pubring.gpg) -endif -$(obj)/system_certificates.o: certs/pubring.gpg +$(obj)/system_certificates.o: $(obj)/pubring.gpg + +$(obj)/pubring.gpg: + $(Q)touch $@ endif + obj-$(CONFIG_SYSTEM_BLACKLIST_KEYRING) += blacklist.o ifneq ($(CONFIG_SYSTEM_BLACKLIST_HASH_LIST),"") obj-$(CONFIG_SYSTEM_BLACKLIST_KEYRING) += blacklist_hashes.o -- 2.23.0 IMPORTANT NOTICE: The contents of this email and any attachments are confidential and may also be privileged. If you are not the intended recipient, please notify the sender immediately and do not disclose the contents to any other person, use it for any purpose, or store or copy the information in any medium. Thank you.

4 6