- Kernel - mailweb.openeuler.org

[PATCH OLK-6.6 0/3] arm64: Support hardware xcall and xint
by Jinjie Ruan 26 May '25

26 May '25

Support hardware xcall and xint. Jinjie Ruan (2): arm64: entry: Support hardware xcall and xint arm64: idle: Handle idle state for xcall/xint Liao Chang (1): arm64/Kconfig: Add turbo features submenuconfig arch/Kconfig | 65 --------------------- arch/arm64/Kconfig | 4 +- arch/arm64/Kconfig.turbo | 77 ++++++++++++++++++++++++ arch/arm64/configs/openeuler_defconfig | 14 +++-- arch/arm64/include/asm/actlr.h | 15 +++++ arch/arm64/include/asm/cpufeature.h | 8 +++ arch/arm64/include/asm/cpuidle.h | 32 ++++++++++ arch/arm64/include/asm/vectors.h | 1 + arch/arm64/kernel/cpufeature.c | 81 ++++++++++++++++++++++++++ arch/arm64/kernel/entry-common.c | 21 ++++++- arch/arm64/kernel/entry.S | 52 ++++++++++++++++- arch/arm64/kernel/idle.c | 31 ++++++++++ arch/arm64/tools/cpucaps | 1 + 13 files changed, 326 insertions(+), 76 deletions(-) create mode 100644 arch/arm64/Kconfig.turbo create mode 100644 arch/arm64/include/asm/actlr.h -- 2.34.1

2 4

[PATCH OLK-6.6] perf/x86/intel: KVM: Mask PEBS_ENABLE loaded for guest with vCPU's value.
by Luo Gengkun 26 May '25

26 May '25

From: Sean Christopherson <seanjc(a)google.com> stable inclusion from stable-v6.6.90 commit 34b6fa11431aef71045ae5a00d90a7d630597eda category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/IC9929 CVE: CVE-2025-37936 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- commit 58f6217e5d0132a9f14e401e62796916aa055c1b upstream. When generating the MSR_IA32_PEBS_ENABLE value that will be loaded on VM-Entry to a KVM guest, mask the value with the vCPU's desired PEBS_ENABLE value. Consulting only the host kernel's host vs. guest masks results in running the guest with PEBS enabled even when the guest doesn't want to use PEBS. Because KVM uses perf events to proxy the guest virtual PMU, simply looking at exclude_host can't differentiate between events created by host userspace, and events created by KVM on behalf of the guest. Running the guest with PEBS unexpectedly enabled typically manifests as crashes due to a near-infinite stream of #PFs. E.g. if the guest hasn't written MSR_IA32_DS_AREA, the CPU will hit page faults on address '0' when trying to record PEBS events. The issue is most easily reproduced by running `perf kvm top` from before commit 7b100989b4f6 ("perf evlist: Remove __evlist__add_default") (after which, `perf kvm top` effectively stopped using PEBS). The userspace side of perf creates a guest-only PEBS event, which intel_guest_get_msrs() misconstrues a guest-*owned* PEBS event. Arguably, this is a userspace bug, as enabling PEBS on guest-only events simply cannot work, and userspace can kill VMs in many other ways (there is no danger to the host). However, even if this is considered to be bad userspace behavior, there's zero downside to perf/KVM restricting PEBS to guest-owned events. Note, commit 854250329c02 ("KVM: x86/pmu: Disable guest PEBS temporarily in two rare situations") fixed the case where host userspace is profiling KVM *and* userspace, but missed the case where userspace is profiling only KVM. Fixes: c59a1f106f5c ("KVM: x86/pmu: Add IA32_PEBS_ENABLE MSR emulation for extended PEBS") Closes: https://lore.kernel.org/all/Z_VUswFkWiTYI0eD@do-x1carbon Reported-by: Seth Forshee <sforshee(a)kernel.org> Signed-off-by: Sean Christopherson <seanjc(a)google.com> Signed-off-by: Peter Zijlstra (Intel) <peterz(a)infradead.org> Reviewed-by: Dapeng Mi <dapeng1.mi(a)linux.intel.com> Tested-by: "Seth Forshee (DigitalOcean)" <sforshee(a)kernel.org> Cc: stable(a)vger.kernel.org Link: https://lore.kernel.org/r/20250426001355.1026530-1-seanjc@google.com Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Luo Gengkun <luogengkun2(a)huawei.com> --- arch/x86/events/intel/core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c index 9605e2a03c57..ca9ed9a32632 100644 --- a/arch/x86/events/intel/core.c +++ b/arch/x86/events/intel/core.c @@ -4166,7 +4166,7 @@ static struct perf_guest_switch_msr *intel_guest_get_msrs(int *nr, void *data) arr[pebs_enable] = (struct perf_guest_switch_msr){ .msr = MSR_IA32_PEBS_ENABLE, .host = cpuc->pebs_enabled & ~cpuc->intel_ctrl_guest_mask, - .guest = pebs_mask & ~cpuc->intel_ctrl_host_mask, + .guest = pebs_mask & ~cpuc->intel_ctrl_host_mask & kvm_pmu->pebs_enable, }; if (arr[pebs_enable].host) { -- 2.34.1

2 1

[PATCH OLK-5.10] xfs: read only mounts with fsopen mount API are busted
by Zizhi Wo 26 May '25

26 May '25

From: Dave Chinner <dchinner(a)redhat.com> stable inclusion from stable-v5.15.149 commit e44e240d537eb5df567ac80f8d5540aa6c5f0d04 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/IC95MC Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- commit d8d222e09dab84a17bb65dda4b94d01c565f5327 upstream. Recently xfs/513 started failing on my test machines testing "-o ro,norecovery" mount options. This was being emitted in dmesg: [ 9906.932724] XFS (pmem0): no-recovery mounts must be read-only. Turns out, readonly mounts with the fsopen()/fsconfig() mount API have been busted since day zero. It's only taken 5 years for debian unstable to start using this "new" mount API, and shortly after this I noticed xfs/513 had started to fail as per above. The syscall trace is: fsopen("xfs", FSOPEN_CLOEXEC) = 3 mount_setattr(-1, NULL, 0, NULL, 0) = -1 EINVAL (Invalid argument) ..... fsconfig(3, FSCONFIG_SET_STRING, "source", "/dev/pmem0", 0) = 0 fsconfig(3, FSCONFIG_SET_FLAG, "ro", NULL, 0) = 0 fsconfig(3, FSCONFIG_SET_FLAG, "norecovery", NULL, 0) = 0 fsconfig(3, FSCONFIG_CMD_CREATE, NULL, NULL, 0) = -1 EINVAL (Invalid argument) close(3) = 0 Showing that the actual mount instantiation (FSCONFIG_CMD_CREATE) is what threw out the error. During mount instantiation, we call xfs_fs_validate_params() which does: /* No recovery flag requires a read-only mount */ if (xfs_has_norecovery(mp) && !xfs_is_readonly(mp)) { xfs_warn(mp, "no-recovery mounts must be read-only."); return -EINVAL; } and xfs_is_readonly() checks internal mount flags for read only state. This state is set in xfs_init_fs_context() from the context superblock flag state: /* * Copy binary VFS mount flags we are interested in. */ if (fc->sb_flags & SB_RDONLY) set_bit(XFS_OPSTATE_READONLY, &mp->m_opstate); With the old mount API, all of the VFS specific superblock flags had already been parsed and set before xfs_init_fs_context() is called, so this all works fine. However, in the brave new fsopen/fsconfig world, xfs_init_fs_context() is called from fsopen() context, before any VFS superblock have been set or parsed. Hence if we use fsopen(), the internal XFS readonly state is *never set*. Hence anything that depends on xfs_is_readonly() actually returning true for read only mounts is broken if fsopen() has been used to mount the filesystem. Fix this by moving this internal state initialisation to xfs_fs_fill_super() before we attempt to validate the parameters that have been set prior to the FSCONFIG_CMD_CREATE call being made. Signed-off-by: Dave Chinner <dchinner(a)redhat.com> Fixes: 73e5fff98b64 ("xfs: switch to use the new mount-api") cc: stable(a)vger.kernel.org Reviewed-by: Christoph Hellwig <hch(a)lst.de> Signed-off-by: Chandan Babu R <chandanbabu(a)kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Conflicts: fs/xfs/xfs_super.c [xxx_fc_xxx->xxx_fs_xxx: 1e5c39dfd3a4 (xfs: rename xfs_fc_* back to xfs_fs_*), does not affect the patch.] Signed-off-by: Zizhi Wo <wozizhi(a)huawei.com> --- fs/xfs/xfs_super.c | 27 +++++++++++++++++---------- 1 file changed, 17 insertions(+), 10 deletions(-) diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c index 451386e21262..4178e5b463c8 100644 --- a/fs/xfs/xfs_super.c +++ b/fs/xfs/xfs_super.c @@ -1460,6 +1460,18 @@ xfs_fc_fill_super( mp->m_super = sb; + /* + * Copy VFS mount flags from the context now that all parameter parsing + * is guaranteed to have been completed by either the old mount API or + * the newer fsopen/fsconfig API. + */ + if (fc->sb_flags & SB_RDONLY) + set_bit(XFS_OPSTATE_READONLY, &mp->m_opstate); + if (fc->sb_flags & SB_DIRSYNC) + mp->m_features |= XFS_FEAT_DIRSYNC; + if (fc->sb_flags & SB_SYNCHRONOUS) + mp->m_features |= XFS_FEAT_WSYNC; + error = xfs_fc_validate_params(mp); if (error) goto out_free_names; @@ -1986,6 +1998,11 @@ static const struct fs_context_operations xfs_context_ops = { .free = xfs_fc_free, }; +/* + * WARNING: do not initialise any parameters in this function that depend on + * mount option parsing having already been performed as this can be called from + * fsopen() before any parameters have been set. + */ static int xfs_init_fs_context( struct fs_context *fc) { @@ -2018,16 +2035,6 @@ static int xfs_init_fs_context( mp->m_logbsize = -1; mp->m_allocsize_log = 16; /* 64k */ - /* - * Copy binary VFS mount flags we are interested in. - */ - if (fc->sb_flags & SB_RDONLY) - set_bit(XFS_OPSTATE_READONLY, &mp->m_opstate); - if (fc->sb_flags & SB_DIRSYNC) - mp->m_features |= XFS_FEAT_DIRSYNC; - if (fc->sb_flags & SB_SYNCHRONOUS) - mp->m_features |= XFS_FEAT_WSYNC; - fc->s_fs_info = mp; fc->ops = &xfs_context_ops; -- 2.39.2

2 1

[PATCH OLK-5.10 000/124] Backport SMC-D feature
by Wang Liang 26 May '25

26 May '25

Albert Huang (1): net/smc: fix smc clc failed issue when netdevice not in init_net Allen Pais (1): net: smc: convert tasklets to use new tasklet_setup() API Christophe JAILLET (1): net/smc: Use the bitmap API when applicable Colin Ian King (1): net/smc: remove redundant re-assignment of pointer link D. Wythe (5): net/smc: Limit backlog connections net/smc: reduce unnecessary blocking in smcr_lgr_reg_rmbs() net/smc: fix potential panic dues to unprotected smc_llc_srv_add_link() net/smc: fix dangling sock under state SMC_APPFINCLOSEWAIT net/smc: return ETIMEDOUT when smc_connect_clc() timeout Dan Carpenter (2): net/smc: Fix an error code in smc_lgr_create() net/smc: delete pointless divide by one David Howells (1): smc: Drop smc_sendpage() in favour of smc_sendmsg() + MSG_SPLICE_PAGES Dmitry Antipov (1): net: smc: fix spurious error message from __sock_release() Dust Li (2): net/smc: add comments for smc_link_{usable|sendable} net/smc: return the right falback reason when prefix checks fail Eric Dumazet (2): net: inline sock_prot_inuse_add() net: make sock_inuse_add() available Gerd Bayer (2): net/smc: Fix dependency of SMC on ISM s390/ism: ism driver implies smc protocol Guangguan Wang (14): net/smc: support smc release version negotiation in clc handshake net/smc: add vendor unique experimental options area in clc handshake net/smc: support smc v2.x features validate net/smc: support max connections per lgr negotiation net/smc: support max links per lgr negotiation in clc handshake net/smc: Extend SMCR v2 linkgroup netlink attribute net/smc: set rmb's SG_MAX_SINGLE_ALLOC limitation only when CONFIG_ARCH_NO_SG_CHAIN is defined net/smc: protect link down work from execute after lgr freed net/smc: check sndbuf_space again after NOSPACE flag is set in smc_poll net/smc: check v2_ext_offset/eid_cnt/ism_gid_cnt when receiving proposal msg net/smc: check smcd_v2_ext_offset when receiving proposal msg net/smc: check return value of sock_recvmsg when draining clc data net/smc: fix data error when recvmsg with MSG_PEEK flag net/smc: use the correct ndev to find pnetid by pnetid table Gustavo A. R. Silva (1): net/smc: Avoid -Wflex-array-member-not-at-end warnings Guvenc Gulce (1): net/smc: Use active link of the connection Heiko Carstens (1): net/smc: Eliminate struct smc_ism_position Jiapeng Chong (1): net/smc: Remove redundant assignment to rc Julian Ruess (2): s390/ism: Fix trying to free already-freed IRQ by repeated ism_dev_exit() s390/ism: add release function for struct device Julian Wiedmann (1): net/smc: no need to flush smcd_dev's event_wq before destroying it Kai Shen (2): net/smc: Use percpu ref for wr tx reference net/smc: Fix memory leak when using percpu refs Karsten Graul (16): net/smc: improve return codes for SMC-Dv2 net/smc: use helper smc_conn_abort() in listen processing net/smc: avoid possible duplicate dmb unregistration net/smc: add support for user defined EIDs net/smc: keep static copy of system EID net/smc: add generic netlink support for system EID net/smc: save stack space and allocate smc_init_info net/smc: prepare for SMC-Rv2 connection net/smc: add SMC-Rv2 connection establishment net/smc: add listen processing for SMC-Rv2 net/smc: add v2 format of CLC decline message net/smc: retrieve v2 gid from IB device net/smc: add v2 support to the work request layer net/smc: extend LLC layer for SMC-Rv2 net/smc: add netlink support for SMC-Rv2 net/smc: stop links when their GID is removed Kuniyuki Iwashima (1): smc: Fix use-after-free in tcp_write_timer_handler(). Li RongQing (2): net/smc: remove unneeded atomic operations in smc_tx_sndbuf_nonempty net/smc: Fix searching in list of known pnetids in smc_pnet_add_pnetid Niklas Schnelle (3): s390/ism: Fix locking for forwarding of IRQs and events to clients s390/ism: Fix and simplify add()/remove() callback handling s390/ism: Do not unregister clients with registered DMBs Simon Horman (1): net/smc: Address spelling errors Stefan Raspl (11): s390/ism: Cleanups net/smc: Pass on DMBE bit mask in IRQ handler net/smc: Terminate connections prior to device removal s390/ism: Introduce struct ism_dmb net/ism: Add new API for client registration net/smc: Register SMC-D as ISM client net/smc: Separate SMC-D and ISM APIs s390/ism: Consolidate SMC-D-related code net/smc: De-tangle ism and smc device initialization net/smc: Fix device de-init sequence net/smc: Introduce explicit check for v2 support Tony Lu (6): net/smc: Clean up local struct sock variables net/smc: Introduce net namespace support for linkgroup net/smc: Print net namespace in log net/smc: Introduce TCP ULP support net/smc: Support SO_REUSEPORT net/smc: Fix cleanup when register ULP fails Wan Jiabing (1): net: smc: Remove repeated struct declaration Wen Gu (32): net/smc: Reset conn->lgr when link group registration fails net/smc: Remove unused function declaration net/smc: Reset connection when trying to use SMCRv2 fails. net/smc: Scan from current RMB list when no position specified net/smc: Don't use RMBs not mapped to new link in SMCRv2 ADD LINK net/smc: fix missing byte order conversion in CLC handshake net/smc: rename some 'fce' to 'fce_v2x' for clarity net/smc: introduce sub-functions for smc_clc_send_confirm_accept() net/smc: unify the structs of accept or confirm message for v1 and v2 net/smc: support SMCv2.x supplemental features negotiation net/smc: introduce virtual ISM device support feature net/smc: define a reserved CHID range for virtual ISM devices net/smc: compatible with 128-bits extended GID of virtual ISM device net/smc: support extended GID in SMC-D lgr netlink attribute net/smc: disable SEID on non-s390 archs where virtual ISM may be used net/smc: manage system EID in SMC stack instead of ISM driver net/smc: fix incorrect SMC-D link group matching logic net/smc: change the term virtual ISM to Emulated-ISM net/smc: decouple ism_client from SMC-D DMB registration net/smc: introduce loopback-ism for SMC intra-OS shortcut net/smc: implement ID-related operations of loopback-ism net/smc: implement DMB-related operations of loopback-ism net/smc: mark optional smcd_ops and check for support when called net/smc: ignore loopback-ism when dumping SMC-D devices net/smc: register loopback-ism into SMC-D device list net/smc: add operations to merge sndbuf with peer DMB net/smc: {at|de}tach sndbuf to peer DMB if supported net/smc: adapt cursor update when sndbuf and peer DMB are merged net/smc: implement DMB-merged operations of loopback-ism net/smc: fix neighbour and rtable leak in smc_ib_find_route() net/smc: avoid overwriting when adjusting sock bufsizes net/smc: fix LGR and link use-after-free issue Wenjia Zhang (1): net/smc: Fix lookup of netdev by using ib_device_get_netdev() Yue Haibing (1): net/smc: Remove unused function declarations Zhengchao Shao (6): net/smc: make smc_hash_sk/smc_unhash_sk static net/smc: remove unreferenced header in smc_loopback.h file net/smc: remove the fallback in __smc_connect net/smc: remove redundant code in smc_connect_check_aclc net/smc: remove unused input parameters in smcr_new_buf_create net/smc: add the max value of fallback reason count liuyacan (1): net/smc: set ini->smcrv2.ib_dev_v2 to NULL if SMC-Rv2 is unavailable drivers/s390/net/Kconfig | 5 +- drivers/s390/net/ism.h | 26 +- drivers/s390/net/ism_drv.c | 489 +++++++++++++----- include/linux/ism.h | 92 ++++ include/net/smc.h | 62 +-- include/net/sock.h | 24 +- include/uapi/linux/smc.h | 48 +- include/uapi/linux/smc_diag.h | 2 + net/core/sock.c | 21 - net/mptcp/subflow.c | 4 +- net/smc/Kconfig | 14 + net/smc/Makefile | 1 + net/smc/af_smc.c | 947 +++++++++++++++++++++++++--------- net/smc/smc.h | 53 +- net/smc/smc_cdc.c | 44 +- net/smc/smc_clc.c | 844 +++++++++++++++++++++++++----- net/smc/smc_clc.h | 192 +++++-- net/smc/smc_close.c | 6 +- net/smc/smc_core.c | 390 ++++++++++---- net/smc/smc_core.h | 129 ++++- net/smc/smc_diag.c | 20 +- net/smc/smc_ib.c | 171 +++++- net/smc/smc_ib.h | 24 +- net/smc/smc_ism.c | 318 ++++++++---- net/smc/smc_ism.h | 61 ++- net/smc/smc_llc.c | 670 +++++++++++++++++++----- net/smc/smc_llc.h | 12 +- net/smc/smc_loopback.c | 427 +++++++++++++++ net/smc/smc_loopback.h | 60 +++ net/smc/smc_netlink.c | 47 +- net/smc/smc_netlink.h | 2 + net/smc/smc_pnet.c | 120 +++-- net/smc/smc_rx.c | 39 +- net/smc/smc_rx.h | 8 +- net/smc/smc_stats.c | 2 +- net/smc/smc_stats.h | 3 +- net/smc/smc_tx.c | 56 +- net/smc/smc_tx.h | 2 - net/smc/smc_wr.c | 305 +++++++++-- net/smc/smc_wr.h | 17 +- 40 files changed, 4502 insertions(+), 1255 deletions(-) create mode 100644 include/linux/ism.h create mode 100644 net/smc/smc_loopback.c create mode 100644 net/smc/smc_loopback.h -- 2.34.1

2 125

[PATCH OLK-6.6 v4 00/13] xfs: support log grant track bytes
by Long Li 26 May '25

26 May '25

Support log grant track bytes, instead of lsn. Darrick J. Wong (1): xfs: revert AIL TASK_KILLABLE threshold Dave Chinner (10): xfs: move and rename xfs_trans_committed_bulk xfs: AIL doesn't need manual pushing xfs: background AIL push should target physical space xfs: ensure log tail is always up to date xfs: l_last_sync_lsn is really AIL state xfs: collapse xlog_state_set_callback in caller xfs: track log space pinned by the AIL xfs: pass the full grant head to accounting functions xfs: grant heads track byte counts, not LSNs xfs: skip flushing log items during push Long Li (2): Revert "xfs: update the last_sync_lsn with ctx start lsn" xfs: fix incorrect tail lsn tracking when AIL is empty Documentation/ABI/testing/sysfs-fs-xfs | 18 +- fs/xfs/libxfs/xfs_defer.c | 4 +- fs/xfs/xfs_inode.c | 1 + fs/xfs/xfs_inode_item.c | 6 +- fs/xfs/xfs_log.c | 514 ++++++------------------- fs/xfs/xfs_log.h | 1 - fs/xfs/xfs_log_cil.c | 179 ++++++++- fs/xfs/xfs_log_priv.h | 61 ++- fs/xfs/xfs_log_recover.c | 26 +- fs/xfs/xfs_sysfs.c | 29 +- fs/xfs/xfs_trace.c | 1 + fs/xfs/xfs_trace.h | 42 +- fs/xfs/xfs_trans.c | 129 ------- fs/xfs/xfs_trans.h | 4 +- fs/xfs/xfs_trans_ail.c | 247 ++++++------ fs/xfs/xfs_trans_priv.h | 45 ++- 16 files changed, 558 insertions(+), 749 deletions(-) -- 2.39.2

2 14

[openeuler:OLK-6.6 2284/2284] arch/arm64/kvm/../../../virt/kvm/eventfd.c:719: warning: cannot understand function prototype: 'struct eventfd_shadow '
by kernel test robot 26 May '25

26 May '25

tree: https://gitee.com/openeuler/kernel.git OLK-6.6 head: 44721e23ce684272eca9901966f55cd2780f4b7e commit: 2dcf53278783921137dbb4eb4628f2c96cac65e2 [2284/2284] eventfd:Introduce the eventfd batch processing interface config: arm64-randconfig-003-20250526 (https://download.01.org/0day-ci/archive/20250526/202505261539.RLecPVlr-lkp@…) compiler: clang version 20.1.2 (https://github.com/llvm/llvm-project 58df0ef89dd64126512e4ee27b4ac3fd8ddf6247) reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20250526/202505261539.RLecPVlr-lkp@…) If you fix the issue in a separate patch/commit (i.e. not just a new version of the same patch/commit), kindly add following tags | Reported-by: kernel test robot <lkp(a)intel.com> | Closes: https://lore.kernel.org/oe-kbuild-all/202505261539.RLecPVlr-lkp@intel.com/ All warnings (new ones prefixed by >>): >> arch/arm64/kvm/../../../virt/kvm/eventfd.c:719: warning: cannot understand function prototype: 'struct eventfd_shadow ' >> arch/arm64/kvm/../../../virt/kvm/eventfd.c:1102: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst * If an exception occurs, arch/arm64/kvm/../../../virt/kvm/eventfd.c:1131: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst * If the temporarily stored eventfd cannot be successfully updated arch/arm64/kvm/../../../virt/kvm/eventfd.c:1176: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst * Copy the current buses, vim +719 arch/arm64/kvm/../../../virt/kvm/eventfd.c 713 714 /** 715 * eventfd_shadow 716 * Save the eventfd pointer and operation. 717 * To protect Kabi, these members are placed here 718 */ > 719 struct eventfd_shadow { 720 struct list_head node; 721 struct _ioeventfd *eventfd; 722 /* Indicate the operations required for this eventfd */ 723 u8 flag; 724 }; 725 -- 0-DAY CI Kernel Test Service https://github.com/intel/lkp-tests/wiki

1 0

[PATCH OLK-6.6 v3 00/12] xfs: support log grant track bytes
by Long Li 26 May '25

26 May '25

Support log grant track bytes, instead of lsn. Dave Chinner (10): xfs: move and rename xfs_trans_committed_bulk xfs: AIL doesn't need manual pushing xfs: background AIL push should target physical space xfs: ensure log tail is always up to date xfs: l_last_sync_lsn is really AIL state xfs: collapse xlog_state_set_callback in caller xfs: track log space pinned by the AIL xfs: pass the full grant head to accounting functions xfs: grant heads track byte counts, not LSNs xfs: skip flushing log items during push Long Li (2): Revert "xfs: update the last_sync_lsn with ctx start lsn" xfs: fix incorrect tail lsn tracking when AIL is empty Documentation/ABI/testing/sysfs-fs-xfs | 18 +- fs/xfs/libxfs/xfs_defer.c | 4 +- fs/xfs/xfs_inode.c | 1 + fs/xfs/xfs_inode_item.c | 6 +- fs/xfs/xfs_log.c | 514 ++++++------------------- fs/xfs/xfs_log.h | 1 - fs/xfs/xfs_log_cil.c | 179 ++++++++- fs/xfs/xfs_log_priv.h | 61 ++- fs/xfs/xfs_log_recover.c | 26 +- fs/xfs/xfs_sysfs.c | 29 +- fs/xfs/xfs_trace.c | 1 + fs/xfs/xfs_trace.h | 42 +- fs/xfs/xfs_trans.c | 129 ------- fs/xfs/xfs_trans.h | 4 +- fs/xfs/xfs_trans_ail.c | 244 ++++++------ fs/xfs/xfs_trans_priv.h | 45 ++- 16 files changed, 554 insertions(+), 750 deletions(-) -- 2.39.2

2 13

[PATCH OLK-5.10 v4 00/11] io_uring patches backport
by Li Lingfeng 26 May '25

26 May '25

Changes in v2: Add 01e68ce08a30d ("io_uring/io-wq: stop setting PF_NO_SETAFFINITY on io-wq workers") and a5fc1441af77 ("io_uring/sqpoll: Do not set PF_NO_SETAFFINITY on sqpoll threads") to allow users to set the cpumask for sqpoll thread and worker thread. Changes in v3: Remove PF_NO_SETAFFINITY of worker; Add 7215469659cb ("io_uring: check for iowq alloc_workqueue failure"). Changes in v4: Modify the commit message; Adjusting the patch sequence. Al Viro (1): io_uring: kiocb_done() should *not* trust ->ki_pos if ->{read,write}_iter() failed Jeff Moyer (1): io-wq: fully initialize wqe before calling cpuhp_state_add_instance_nocalls() Jens Axboe (4): io_uring/io-wq: stop setting PF_NO_SETAFFINITY on io-wq workers io_uring/fdinfo: remove need for sqpoll lock for thread/pid retrieval io_uring: use private workqueue for exit work io_uring/sqpoll: close race on waiting for sqring entries Max Kellermann (1): io_uring/io-wq: do not use bogus hash value Michal Koutný (1): io_uring/sqpoll: Do not set PF_NO_SETAFFINITY on sqpoll threads Pavel Begunkov (3): io_uring: check for iowq alloc_workqueue failure io_uring: protect register tracing io_uring/sqpoll: fix sqpoll error handling races io_uring/io-wq.c | 29 ++++++++++++++++----------- io_uring/io_uring.c | 48 ++++++++++++++++++++++++++++++++------------- 2 files changed, 52 insertions(+), 25 deletions(-) -- 2.31.1

2 12

[PATCH OLK-6.6 0/2] ext4: better scalability for ext4 block allocation
by Baokun Li 26 May '25

26 May '25

Since servers have more and more CPUs, and we're running more containers on them, we've been using will-it-scale to test how well ext4 scales. The fallocate2 test (append 8KB to 1MB, truncate to 0, repeat) run concurrently on 64 containers revealed significant contention in block allocation/free, leading to much lower aggregate fallocate OPS compared to a single container (see below). 1 | 2 | 4 | 8 | 16 | 32 | 64 -------|--------|--------|--------|--------|--------|------- 295287 | 70665 | 33865 | 19387 | 10104 | 5588 | 3588 The main bottleneck was the ext4_lock_group(), which both block allocation and free fought over. While the block group for block free is fixed and unoptimizable, the block group for allocation is selectable. Consequently, the ext4_try_lock_group() helper function was added to avoid contention on busy groups, and you can see more in Patch 1. After we fixed the ext4_lock_group bottleneck, another one showed up: s_md_lock. This lock protects different data when allocating and freeing blocks. We got rid of the s_md_lock call in block allocation by making stream allocation work per inode instead of globally. You can find more details in Patch 2. Performance test data follows: CPU: HUAWEI Kunpeng 920 Memory: 480GB Disk: 480GB SSD SATA 3.2 Test: Running will-it-scale/fallocate2 on 64 CPU-bound containers. Observation: Average fallocate operations per container per second. |--------|--------|--------|--------|--------|--------|--------|--------| | - | 1 | 2 | 4 | 8 | 16 | 32 | 64 | |--------|--------|--------|--------|--------|--------|--------|--------| | base | 295287 | 70665 | 33865 | 19387 | 10104 | 5588 | 3588 | |--------|--------|--------|--------|--------|--------|--------|--------| | linear | 286328 | 123102 | 119542 | 90653 | 60344 | 35302 | 23280 | | | -3.0% | 74.20% | 252.9% | 367.5% | 497.2% | 531.6% | 548.7% | |--------|--------|--------|--------|--------|--------|--------|--------| |mb_optim| 292498 | 133305 | 103069 | 61727 | 29702 | 16845 | 10430 | |ize_scan| -0.9% | 88.64% | 204.3% | 218.3% | 193.9% | 201.4% | 190.6% | |--------|--------|--------|--------|--------|--------|--------|--------| Baokun Li (2): ext4: add ext4_try_lock_group() to skip busy groups ext4: move mb_last_[group|start] to ext4_inode_info fs/ext4/ext4.h | 30 ++++++++++++++++++------------ fs/ext4/mballoc.c | 34 ++++++++++++++++++++-------------- fs/ext4/super.c | 2 ++ 3 files changed, 40 insertions(+), 26 deletions(-) -- 2.46.1

2 3

[PATCH OLK-6.6 v2 00/12] xfs: support log grant track bytes
by Long Li 26 May '25

26 May '25

Support log grant track bytes, instead of lsn. Dave Chinner (10): xfs: move and rename xfs_trans_committed_bulk xfs: AIL doesn't need manual pushing xfs: background AIL push should target physical space xfs: ensure log tail is always up to date xfs: l_last_sync_lsn is really AIL state xfs: collapse xlog_state_set_callback in caller xfs: track log space pinned by the AIL xfs: pass the full grant head to accounting functions xfs: grant heads track byte counts, not LSNs xfs: skip flushing log items during push Long Li (2): Revert "xfs: update the last_sync_lsn with ctx start lsn" xfs: fix incorrect tail lsn tracking when AIL is empty Documentation/ABI/testing/sysfs-fs-xfs | 18 +- fs/xfs/libxfs/xfs_defer.c | 4 +- fs/xfs/xfs_inode.c | 1 + fs/xfs/xfs_inode_item.c | 6 +- fs/xfs/xfs_log.c | 514 ++++++------------------- fs/xfs/xfs_log.h | 1 - fs/xfs/xfs_log_cil.c | 179 ++++++++- fs/xfs/xfs_log_priv.h | 61 ++- fs/xfs/xfs_log_recover.c | 26 +- fs/xfs/xfs_sysfs.c | 29 +- fs/xfs/xfs_trace.c | 1 + fs/xfs/xfs_trace.h | 42 +- fs/xfs/xfs_trans.c | 129 ------- fs/xfs/xfs_trans.h | 4 +- fs/xfs/xfs_trans_ail.c | 244 ++++++------ fs/xfs/xfs_trans_priv.h | 45 ++- 16 files changed, 554 insertions(+), 750 deletions(-) -- 2.39.2

2 13