)]}'
{
  "log": [
    {
      "commit": "4405c18c95d2d0f99f6c581ef0ee8591a1aa6ccd",
      "tree": "f07c049f80f98f71047c0a29a357de05e240e0ad",
      "parents": [
        "d866d5bafc5e1417032255002e95f8ec3121754b",
        "a001666afbf9d2a6096fea95686549b9b1c9d968"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Nov 18 09:51:04 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Nov 18 09:51:04 2026 -0800"
      },
      "message": "Merge tag \u0027mm-nonmm-hotfixes-stable-2026-11-18-00-06\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm\n\nPull non-MM updates from Andrew Morton:\n\n - \"get_maintainer: stop sending every patch to LKML\" (Konstantin\n   Ryabitsev):\n\n   Make THE REST a fallback when no usable subsystem list matches and\n   update the submission guide. This avoids routine copies to LKML while\n   retaining a mailing list for otherwise unmatched patches.\n\n - On local OCFS2 mounts, a failed shared mmap write fault now delivers\n   SIGBUS instead of hitting BUG_ON(). Journal shutdown is serialized\n   with inode cleanup, and checkpoint waiters no longer use a queue\n   embedded in the journal being freed.\n\n - Queued dlmfs lock work now keeps the superblock active through its\n   final iput(), avoiding a use-after-free during unmount. The OCFS2\n   locking_state debugfs iterator follows its live cursor instead of\n   list links copied from a lock resource which may have been freed\n   during a read-only remount.\n\n - Fault-injection boot parameters no longer call code which allocates\n   memory and schedules work before slab and workqueues are initialized,\n   fixing a boot failure.\n\n* tag \u0027mm-nonmm-hotfixes-stable-2026-11-18-00-06\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm:\n  MAINTAINERS: DRM DRIVERS AND MISC GPU PATCHES\n  MAINTAINERS: update Andrea Cervesato\u0027s email for LTP\n  mailmap: update email address for Sai Teja Aluvala\n  mailmap: add an alias for Alex Elder\n  mailmap: update entry for Guodong Xu\n  mailmap: add entries for Ridong Chen\n  mailmap: add Wang Yibo\u0027s work email address\n  mailmap: add entry for Beat M. Liechti\n  mailmap: update email address for Naman Jain\n  mailmap: add entries for Luiz Capitulino\n  mailmap: add entries for Miaoqing Pan\n  mailmap: update Mika Westerberg\u0027s email address\n  mailmap: update email address for Bradley Morgan\n  docs: submitting-patches: make LKML a list of last resort again\n  get_maintainer: only use THE REST as a fallback\n  ocfs2: fix kernel BUG in ocfs2_page_mkwrite() when write_end fails\n  fault-inject: fix boot failure with fault injection boot parameters\n  ocfs2: advance locking_state through the live cursor\n  ocfs2/dlmfs: keep the superblock alive during lock work\n  ocfs2: serialize inode journal access with shutdown\n"
    },
    {
      "commit": "d866d5bafc5e1417032255002e95f8ec3121754b",
      "tree": "36f2e232340d4f343b32adc8a292fa502d20247d",
      "parents": [
        "fb52ae0e9afb44d5f06bcf855b490584dad66687",
        "a4e667346a8d91e629a2ca9affc70dc28978f4a1"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Nov 18 09:01:16 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Nov 18 09:01:16 2026 -0800"
      },
      "message": "Merge tag \u0027mm-hotfixes-stable-2026-11-17-19-55\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm\n\nPull MM fixes from Andrew Morton:\n\n - \"maple_tree fixes for 7.3\" (Liam R. Howlett):\n\n   Fix an RCU reader NULL dereference when the tree is emptied during a\n   walk, including /proc/PID/maps racing with munmap(). Add tests and\n   make debug WARN and BUG checks report failures properly\n\n - \"mm/kasan: Detect double-free involving kfree_rcu_sheaf()\" (Kuniyuki\n   Iwashima):\n\n   Detect double-free when an object enters an RCU sheaf. KASAN could\n   previously miss it until the object was reused, leading to a later\n   free of a live object. Add KUnit tests for the different free orders\n\n - \"mm: Fix page-table teardown during memory hot-remove\" (Muchun Song):\n\n   Balance page-table constructors and destructors when hot-remove frees\n   tables on x86-64, RISC-V and arm64. This fixes NR_PAGETABLE\n   accounting for tables with and without constructors\n\n - \"mm/damon: fix watermark state, kunit pid leak and selftest fake\n   pass\" (SJ Park):\n\n   Recompute watermark wait state when DAMON schemes are removed, so\n   kdamond resumes monitoring. Fix a pid reference leak in KUnit and a\n   selftest that reports a skipped run as a pass\n\n - An XArray iterator could step backwards when a large page-cache entry\n   was split, sending filemap_map_pages() outside the PTE page it\n   locked. The XArray correction and a filemap bound check prevent\n   out-of-bounds accesses and use-after-free. Another XArray fix\n   prevents lockless walkers from following a parent pointer into a\n   freed node\n\n - Userfaultfd now validates the whole write-protection range before\n   changing any VMA, avoiding partial unprotect without waking blocked\n   faults. Failed unmaps now abort queued UFFD_EVENT_UNMAPs; previously\n   they could report a nonexistent unmap, leak the userfaultfd context\n   and leave later operations returning -EAGAIN. A fork fix clears stale\n   uffd PTE bits if the new context is released before its event is\n   read\n\n - Moving the hugetlbfs inode trace event before RCU freeing avoids a\n   use-after-free of its superblock. Shmem swapin no longer retries\n   forever on reused swap slots containing a mismatched large swap-cache\n   folio. Swapoff also avoids a cluster-list loop that can soft-lock\n   reclaim and leave swapoff hung\n\n - Weighted interleave recalculation now uses one node set throughout,\n   avoiding a divide-by-zero during memory hotplug. Hwpoison holds the\n   folio lock when inspecting its mapping and no longer leaks a folio\n   reference; it also updates poison statistics before notifying\n   rasdaemon through a trace event\n\n - Cross-node migration now moves NR_FILE_THPS and NR_KERNEL_FILE_PAGES\n   with page-cache folios, preventing per-node and memcg-node accounting\n   from retaining stale counts or underflowing after the folios are\n   freed\n\n* tag \u0027mm-hotfixes-stable-2026-11-17-19-55\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm: (27 commits)\n  xarray: fix index jumping backwards in xas_find()\n  userfaultfd: validate the whole range before changing write protection\n  maple_tree: update internal warning and bug_on code\n  maple_tree: test emptying tree while RCU reading\n  maple_tree: fix rcu reader NULL dereference after the tree is emptied\n  mm/vma: read anon_vma carefully in validate_mm()\n  hugetlbfs: move hugetlbfs_free_inode tracepoint to -\u003edestroy_inode()\n  selftests/damon/sysfs_memcg_path_leak: define ksft_skip\n  mm/damon/tests/sysfs-kunit: fix leaked pid reference in add_targets test\n  mm/damon/core: reset watermark wait state on each iteration\n  mm/filemap: stop fault-around when the walk steps back\n  mm: memory-failure: update statistics before emitting the trace event\n  mm/vma: undo userfaultfd unmap preparation when unmapping fails\n  userfaultfd: clear uffd state in child PTEs if uffd is released during fork\n  mm/memory-failure: fix folio refcount leak and min_order_for_split() locking\n  mm/shmem: fix swapin livelock on a mismatched large swap cache folio\n  mm/mempolicy: use one node set for interleave weights\n  xarray: clear the new head\u0027s parent before freeing the old head\n  arm64/mm: fix destructor for unconstructed hotplug page tables\n  riscv/mm: fix missing destructor for hotplug PUD tables\n  ...\n"
    },
    {
      "commit": "fb52ae0e9afb44d5f06bcf855b490584dad66687",
      "tree": "a01c85f755303667c04a8c4d7f3f33cca43c61af",
      "parents": [
        "cea5f91daa3ac86c98b63b3a2e89da622670ccf6",
        "2de253924266eeedb798fb662173726c90b0cf04"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Nov 18 08:48:22 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Nov 18 08:48:22 2026 -0800"
      },
      "message": "Merge tag \u0027fscrypt-for-linus\u0027 of git://git.kernel.org/pub/scm/fs/fscrypt/linux\n\nPull fscrypt fix from Eric Biggers:\n \"Fix a NULL dereference that was theoretically possible if\n  FS_IOC_SETFLAGS was called to set the casefold flag on an\n  encrypted directory concurrently with its encryption key\n  being set up\"\n\n* tag \u0027fscrypt-for-linus\u0027 of git://git.kernel.org/pub/scm/fs/fscrypt/linux:\n  fscrypt: Initialize -\u003eci_master_key before publication\n"
    },
    {
      "commit": "cea5f91daa3ac86c98b63b3a2e89da622670ccf6",
      "tree": "a023f4e285309c0d6e948db41f89b1becf92918d",
      "parents": [
        "20eef29b7c8465ac98d7dd6d6f2ac07a8d8056ca"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Tue Nov 17 11:42:03 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Wed Nov 18 07:54:26 2026 -0800"
      },
      "message": "wait-event: clean up the internal helper arguments\n\nThe ___wait_event() helper for historical reasons took a \"0/1\" argument\nfor whether the wait-queue entry should be marked WQ_FLAG_EXCLUSIVE or\nnot.  That didn\u0027t really help readability, but it also limited the\nwq_flags for that helper to only that one flag.\n\nJust make that macro argument be the actual \u0027wq_flags\u0027 value, which\nsimplifies the helper, and makes it more generic.  It also happens to\nmake the users of this macro more obvious when an unexplained \u00271\u0027\nargument becomes a more self-explanatory \u0027WQ_FLAG_EXCLUSIVE\u0027.\n\nAlso change the existing users that use that raw ___wait_event()\ninterface with just a plain schedule() to use the pre-existing helpers\nmacros: __wait_event_state() and __wait_event_state_exclusive().\n\n[ This is not a bug-fix: it was prep-work for a bug report that turned\n  out to be just a \"don\u0027t do that\" situation instead. But I also didn\u0027t\n  want to just throw it away or leave it pending, since it just made\n  some of that grotty code much clearer - Linus]\n\nReviewed-by: Oleg Nesterov \u003coleg@redhat.com\u003e\nLink: https://lore.kernel.org/all/20261111-pipe-epollexclusive-wakeup-v1-1-0af6283e9fda@columbia.edu/\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n"
    },
    {
      "commit": "a001666afbf9d2a6096fea95686549b9b1c9d968",
      "tree": "47d84cfff8bdf3b01f364cd474349cc92bd9dde7",
      "parents": [
        "59233c1123c2e800055ec685407a004351ae3f62"
      ],
      "author": {
        "name": "Joe Perches",
        "email": "joe@perches.com",
        "time": "Sun Nov 01 00:41:12 2026 -0700"
      },
      "committer": {
        "name": "Andrew Morton",
        "email": "akpm@linux-foundation.org",
        "time": "Tue Nov 17 21:40:05 2026 -0800"
      },
      "message": "MAINTAINERS: DRM DRIVERS AND MISC GPU PATCHES\n\nThere\u0027s a small defect in get_maintainers.pl that allows\nincomplete file pattern matching.\n\nUpdate the \"F: include/drm/drm pattern\" to specify only the actual files\ncovered instead of a prefixed partial path to these .h files.\n\nLink: https://lore.kernel.org/576e04e1d8f0d5ac0805d450b9c87e69f7f0d2c8.camel@perches.com\nSigned-off-by: Joe Perches \u003cjoe@perches.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nCc: Maarten Lankhorst \u003cmaarten.lankhorst@linux.intel.com\u003e\nCc: Maxime Ripard \u003cmripard@kernel.org\u003e\nCc: Thomas Zimmermann \u003ctzimmermann@suse.de\u003e\nCc: Borislav Petkov \u003cbp@alien8.de\u003e\n\n MAINTAINERS |    2 +-\n 1 file changed, 1 insertion(+), 1 deletion(-)\n\n"
    },
    {
      "commit": "59233c1123c2e800055ec685407a004351ae3f62",
      "tree": "bfb6116590a9301e3ff73043b54644f17d685b3f",
      "parents": [
        "971da83016eb6c6d110580d9b2881f7d2bcfe503"
      ],
      "author": {
        "name": "Andrea Cervesato",
        "email": "andrea.cervesato@linux.dev",
        "time": "Mon Oct 26 11:47:17 2026 +0100"
      },
      "committer": {
        "name": "Andrew Morton",
        "email": "akpm@linux-foundation.org",
        "time": "Tue Nov 17 21:40:05 2026 -0800"
      },
      "message": "MAINTAINERS: update Andrea Cervesato\u0027s email for LTP\n\nI\u0027m not working for SUSE anymore and I will use linux.dev email to\ndevelop and maintain Linux Test Project. Also update .mailmap.\n\nLink: https://lore.kernel.org/20261026-acer_maintainers-v2-1-4cf0f350fdf1@linux.dev\nSigned-off-by: Andrea Cervesato \u003candrea.cervesato@linux.dev\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nAcked-by: Petr Vorel \u003cpvorel@suse.cz\u003e\nReviewed-by: Li Wang (Hieco) \u003cli.wang@linux.dev\u003e\nReviewed-by: Cyril Hrubis \u003cchrubis@suse.cz\u003e\n\n .mailmap    |    1 +\n MAINTAINERS |    2 +-\n 2 files changed, 2 insertions(+), 1 deletion(-)\n\n"
    },
    {
      "commit": "971da83016eb6c6d110580d9b2881f7d2bcfe503",
      "tree": "952aea023b4bd205731de4781b19e274ff890eb1",
      "parents": [
        "099ed44ed4fb83e5eae85db900aa704142c5d3b9"
      ],
      "author": {
        "name": "Sai Teja Aluvala",
        "email": "aluvala.sai.teja@intel.com",
        "time": "Tue Nov 17 19:00:08 2026 +0530"
      },
      "committer": {
        "name": "Andrew Morton",
        "email": "akpm@linux-foundation.org",
        "time": "Tue Nov 17 21:40:04 2026 -0800"
      },
      "message": "mailmap: update email address for Sai Teja Aluvala\n\nMap my former Qualcomm email address to my current Intel email address.\n\nLink: https://lore.kernel.org/20261117133008.2560155-1-aluvala.sai.teja@intel.com\nSigned-off-by: Sai Teja Aluvala \u003caluvala.sai.teja@intel.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nCc: Jeff Johnson \u003cjeff.johnson@oss.qualcomm.com\u003e\n\n .mailmap |    1 +\n 1 file changed, 1 insertion(+)\n\n"
    },
    {
      "commit": "099ed44ed4fb83e5eae85db900aa704142c5d3b9",
      "tree": "12b075b449977085d42218fcfde4b5fec6eae167",
      "parents": [
        "b2ddfc1c1afdd087f5d8bb3a59bdba8cb562ad83"
      ],
      "author": {
        "name": "Alex Elder",
        "email": "elder@riscstar.com",
        "time": "Tue Nov 10 08:42:47 2026 -0600"
      },
      "committer": {
        "name": "Andrew Morton",
        "email": "akpm@linux-foundation.org",
        "time": "Tue Nov 17 21:40:04 2026 -0800"
      },
      "message": "mailmap: add an alias for Alex Elder\n\nAdd an entry to .mailmap to indicate that \u003celder@riscstar.com\u003e represents\nthe same person as \u003celder@kernel.org\u003e.\n\nLink: https://lore.kernel.org/20261110144250.656840-1-elder@riscstar.com\nSigned-off-by: Alex Elder \u003celder@riscstar.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nCc: Jakub Kicinski \u003ckuba@kernel.org\u003e\nCc: Alex Elder \u003celder@kernel.org\u003e\n\n .mailmap |    1 +\n 1 file changed, 1 insertion(+)\n\n"
    },
    {
      "commit": "b2ddfc1c1afdd087f5d8bb3a59bdba8cb562ad83",
      "tree": "b3c60ac48118309e982ee23ddb2679feface41b4",
      "parents": [
        "3a1ed9b7ecdcbfadd3529b59938cd706b74149c4"
      ],
      "author": {
        "name": "Guodong Xu",
        "email": "guodong.xu@linux.dev",
        "time": "Fri Nov 06 16:30:48 2026 -0500"
      },
      "committer": {
        "name": "Andrew Morton",
        "email": "akpm@linux-foundation.org",
        "time": "Tue Nov 17 21:40:04 2026 -0800"
      },
      "message": "mailmap: update entry for Guodong Xu\n\nUse my linux.dev address for kernel correspondence, including for\ncontributions made with my other email addresses. It is not tied to\na personal or employer mailbox.\n\nLink: https://lore.kernel.org/7e9369cf1473539bafe9cb90207ba6fe98b534ca.1794001419.git.guodong.xu@linux.dev\nSigned-off-by: Guodong Xu \u003cguodong.xu@linux.dev\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\n\n .mailmap |    6 ++++--\n 1 file changed, 4 insertions(+), 2 deletions(-)\n\n"
    },
    {
      "commit": "3a1ed9b7ecdcbfadd3529b59938cd706b74149c4",
      "tree": "6392a510b66fa4e23343c101ed1c125bf83105e0",
      "parents": [
        "6f7fd2b4c2f1fad4394bae20c59192723a0704fc"
      ],
      "author": {
        "name": "Ridong Chen",
        "email": "chenridong@xiaomi.com",
        "time": "Sun Nov 08 20:35:09 2026 +0800"
      },
      "committer": {
        "name": "Andrew Morton",
        "email": "akpm@linux-foundation.org",
        "time": "Tue Nov 17 21:40:04 2026 -0800"
      },
      "message": "mailmap: add entries for Ridong Chen\n\nMap my former Huawei address and my current Xiaomi address to my\npersonal address, so that all my contributions are attributed to a\nsingle identity.\n\nLink: https://lore.kernel.org/20261108123509.214127-1-ridong.chen@linux.dev\nSigned-off-by: Ridong Chen \u003cchenridong@xiaomi.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\n\n .mailmap |    2 ++\n 1 file changed, 2 insertions(+)\n\n"
    },
    {
      "commit": "6f7fd2b4c2f1fad4394bae20c59192723a0704fc",
      "tree": "0ab171bdfcc0dd2de226fdb350724c4e6cdc6215",
      "parents": [
        "b5267747764169008da1d5cb88e19ecf0a289b3e"
      ],
      "author": {
        "name": "Wang Yibo",
        "email": "wangyibo@kylinos.cn",
        "time": "Mon Nov 09 11:06:51 2026 +0800"
      },
      "committer": {
        "name": "Andrew Morton",
        "email": "akpm@linux-foundation.org",
        "time": "Tue Nov 17 21:40:03 2026 -0800"
      },
      "message": "mailmap: add Wang Yibo\u0027s work email address\n\nMy kernel work will be signed off with my company address from now on.\nAdd it to .mailmap and map the personal address I used before to it, so\nthat both resolve to the same author in git log and git shortlog.\n\nLink: https://lore.kernel.org/20261109030651.2289573-1-wangyibo@kylinos.cn\nSigned-off-by: Wang Yibo \u003cwangyibo@kylinos.cn\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\n\n .mailmap |    1 +\n 1 file changed, 1 insertion(+)\n\n"
    },
    {
      "commit": "b5267747764169008da1d5cb88e19ecf0a289b3e",
      "tree": "aa4207d057e0a989677e2a8ff7316acf79e259d2",
      "parents": [
        "7aa8cba37cfd56e3d08f0391a99b312f9aed4df6"
      ],
      "author": {
        "name": "Beat M. Liechti",
        "email": "beat@softbase.tech",
        "time": "Sun Nov 01 13:37:29 2026 +0100"
      },
      "committer": {
        "name": "Andrew Morton",
        "email": "akpm@linux-foundation.org",
        "time": "Tue Nov 17 21:40:03 2026 -0800"
      },
      "message": "mailmap: add entry for Beat M. Liechti\n\nMap my old address, used for the firedtv fixes in 2009, to my current\none.\n\nLink: https://lore.kernel.org/20261101123733.10578-1-beat@softbase.tech\nSigned-off-by: Beat M. Liechti \u003cbeat@softbase.tech\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\n\n .mailmap |    1 +\n 1 file changed, 1 insertion(+)\n\n"
    },
    {
      "commit": "7aa8cba37cfd56e3d08f0391a99b312f9aed4df6",
      "tree": "c498476c14beda3548f907033cb5feb346153d57",
      "parents": [
        "a357471e83422c49d1bb7fb8a666f34aa5e52433"
      ],
      "author": {
        "name": "Naman Jain",
        "email": "namjain@linux.microsoft.com",
        "time": "Tue Nov 17 08:45:42 2026 +0000"
      },
      "committer": {
        "name": "Andrew Morton",
        "email": "akpm@linux-foundation.org",
        "time": "Tue Nov 17 21:40:03 2026 -0800"
      },
      "message": "mailmap: update email address for Naman Jain\n\nMy Qualcomm email address is no longer accessible. Map it to my current\nMicrosoft email address.\n\nLink: https://lore.kernel.org/20261117084542.2076622-1-namjain@linux.microsoft.com\nSigned-off-by: Naman Jain \u003cnamjain@linux.microsoft.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nCc: Jakub Kicinski \u003ckuba@kernel.org\u003e\nCc: Bastien Curutchet (Schneider Electric) \u003cbastien.curutchet@bootlin.com\u003e\nCc: Douglas Anderson \u003cdianders@chromium.org\u003e\nCc: Jesper Dangaard Brouer \u003chawk@kernel.org\u003e\nCc: Alexander Sverdlin \u003calex@sverdlin.org\u003e\nCc: Chuck Lever \u003ccel@kernel.org\u003e\nCc: Jeff Johnson \u003cjeff.johnson@oss.qualcomm.com\u003e\n\n .mailmap |    1 +\n 1 file changed, 1 insertion(+)\n\n"
    },
    {
      "commit": "a357471e83422c49d1bb7fb8a666f34aa5e52433",
      "tree": "b1671d476980c0039e3637aee54d592500e3c378",
      "parents": [
        "f0bff4871f91501b8d30e787f15865c3f1f08870"
      ],
      "author": {
        "name": "Luiz Capitulino (Red Hat)",
        "email": "luiz.capitulino@linux.dev",
        "time": "Mon Nov 16 16:57:56 2026 -0500"
      },
      "committer": {
        "name": "Andrew Morton",
        "email": "akpm@linux-foundation.org",
        "time": "Tue Nov 17 21:40:03 2026 -0800"
      },
      "message": "mailmap: add entries for Luiz Capitulino\n\nMap my previous addresses to my linux.dev address, which I\u0027ll be using for\nkernel work going forward.\n\nLink: https://lore.kernel.org/20261116215756.133593-1-luiz.capitulino@linux.dev\nSigned-off-by: Luiz Capitulino (Red Hat) \u003cluiz.capitulino@linux.dev\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nAcked-by: Nico Pache \u003cnico.pache@linux.dev\u003e\n\n .mailmap |    3 +++\n 1 file changed, 3 insertions(+)\n\n"
    },
    {
      "commit": "f0bff4871f91501b8d30e787f15865c3f1f08870",
      "tree": "e392a0df4a670315b37f040cc92b2c50618be419",
      "parents": [
        "d61d78af3286548cef8e31ac578c0ed0df090f94"
      ],
      "author": {
        "name": "Miaoqing Pan",
        "email": "miaoqing.pan@oss.qualcomm.com",
        "time": "Mon Nov 16 10:32:26 2026 +0800"
      },
      "committer": {
        "name": "Andrew Morton",
        "email": "akpm@linux-foundation.org",
        "time": "Tue Nov 17 21:40:02 2026 -0800"
      },
      "message": "mailmap: add entries for Miaoqing Pan\n\nMap past iterations of my e-mail addresses to the current one.\n\nLink: https://lore.kernel.org/20261116023226.17265-1-miaoqing.pan@oss.qualcomm.com\nSigned-off-by: Miaoqing Pan \u003cmiaoqing.pan@oss.qualcomm.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nCc: Jeff Johnson \u003cjeff.johnson@oss.qualcomm.com\u003e\nCc: Jakub Kicinski \u003ckuba@kernel.org\u003e\nCc: Alexander Sverdlin \u003calex@sverdlin.org\u003e\nCc: Chuck Lever \u003ccel@kernel.org\u003e\nCc: Eugen Hristev \u003cehristev@kernel.org\u003e\nCc: Stéphane Grosjean \u003cs.grosjean@peak-system.fr\u003e\nCc: Alice Mikityanska \u003calice.kernel@fastmail.im\u003e\n\n .mailmap |    3 ++-\n 1 file changed, 2 insertions(+), 1 deletion(-)\n\n"
    },
    {
      "commit": "d61d78af3286548cef8e31ac578c0ed0df090f94",
      "tree": "d3a83c61f8557b7e68503b4176c9d7a7a7427db6",
      "parents": [
        "f5d6f3184eacca5cdf04b39db1e8fc4b244e029e"
      ],
      "author": {
        "name": "Mika Westerberg",
        "email": "westeri@kernel.org",
        "time": "Wed Nov 11 13:39:09 2026 +0100"
      },
      "committer": {
        "name": "Andrew Morton",
        "email": "akpm@linux-foundation.org",
        "time": "Tue Nov 17 21:40:02 2026 -0800"
      },
      "message": "mailmap: update Mika Westerberg\u0027s email address\n\nMy @linux.intel.com address will be going away soon so switch to use my\n@kernel.org address.\n\nLink: https://lore.kernel.org/20261111123910.1324579-1-mika.westerberg@linux.intel.com\nSigned-off-by: Mika Westerberg \u003cwesteri@kernel.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nCc: Yehezkel Bernat \u003cyehezkelshb@gmail.com\u003e\nCc: Lukas Wunner \u003clukas@wunner.de\u003e\nCc: Andreas Noever \u003candreas.noever@gmail.com\u003e\n\n .mailmap |    1 +\n 1 file changed, 1 insertion(+)\n\n"
    },
    {
      "commit": "f5d6f3184eacca5cdf04b39db1e8fc4b244e029e",
      "tree": "5d125dceb3e8f48ee3da6a38e213b2b49c2c0305",
      "parents": [
        "3a198413bb7137907a351f43dbf1c3bf5e956900"
      ],
      "author": {
        "name": "Bradley Morgan",
        "email": "bradley.morgan@linux.dev",
        "time": "Mon Nov 09 19:19:28 2026 +0000"
      },
      "committer": {
        "name": "Andrew Morton",
        "email": "akpm@linux-foundation.org",
        "time": "Tue Nov 17 21:40:02 2026 -0800"
      },
      "message": "mailmap: update email address for Bradley Morgan\n\nI switched to bradley.morgan@linux.dev for kernel work, so map the\nbrads@mainlining.org and include@grrlz.net addresses over to keep shortlog\nand blame from splitting commits between them.\n\nLink: https://lore.kernel.org/20261109191928.3222846-1-bradley.morgan@linux.dev\nSigned-off-by: Bradley Morgan \u003cbradley.morgan@linux.dev\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nCc: Bradley Morgan \u003cbrads@mainlining.org\u003e\n\n .mailmap |    3 ++-\n 1 file changed, 2 insertions(+), 1 deletion(-)\n\n"
    },
    {
      "commit": "3a198413bb7137907a351f43dbf1c3bf5e956900",
      "tree": "633b7b7d0e8e4bd69532e1356594dc7d53f51bde",
      "parents": [
        "e00dcd6e54b774a810d28119bad740f317df377c"
      ],
      "author": {
        "name": "Konstantin Ryabitsev",
        "email": "konstantin@linuxfoundation.org",
        "time": "Fri Nov 13 18:02:46 2026 -0500"
      },
      "committer": {
        "name": "Andrew Morton",
        "email": "akpm@linux-foundation.org",
        "time": "Tue Nov 17 21:40:02 2026 -0800"
      },
      "message": "docs: submitting-patches: make LKML a list of last resort again\n\nCommit 77167b966b7e (\"docs: submitting-patches: clarify the role of LKML\")\nchanged the documentation to say that all patches should be sent to LKML\nby default, because get_maintainer.pl always suggested it.\n\nNow that get_maintainer.pl only falls back to THE REST when no other\nmatching section provides a mailing list, update the documentation to\nmatch: send patches to the subsystem-specific list, and use LKML only when\nthere isn\u0027t one.\n\nLink: https://lore.kernel.org/20261113-get-maintainer-the-rest-fallback-v3-2-36ee5a3e8d13@linuxfoundation.org\nSigned-off-by: Konstantin Ryabitsev \u003ckonstantin@linuxfoundation.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSuggested-by: Petr Pavlu \u003cpetr.pavlu@suse.com\u003e\nAcked-by: Randy Dunlap \u003crdunlap@infradead.org\u003e\nAcked-by: Joe Perches \u003cjoe@perches.com\u003e\nCc: Jonathan Corbet \u003ccorbet@lwn.net\u003e\nCc: Shuah Khan \u003cskhan@linuxfoundation.org\u003e\nCc: Federico Vaga \u003cfederico.vaga@vaga.pv.it\u003e\nCc: Akira Yokosawa \u003cakiyks@gmail.com\u003e\nCc: Carlos Bilbao \u003ccarlos.bilbao@kernel.org\u003e\nCc: Avadhut Naik \u003cavadhut.naik@amd.com\u003e\nCc: Alex Shi \u003calexs@kernel.org\u003e\nCc: Yanteng Si \u003csi.yanteng@linux.dev\u003e\nCc: Dongliang Mu \u003cdzm91@hust.edu.cn\u003e\nCc: Chen-Yu Yeh \u003cchenyou910331@gmail.com\u003e\nCc: Weijie Yuan \u003cwy@wyuan.org\u003e\nCc: \u003cusers@kernel.org\u003e\n\n Documentation/process/submitting-patches.rst |    9 ++++++---\n 1 file changed, 6 insertions(+), 3 deletions(-)\n\n"
    },
    {
      "commit": "e00dcd6e54b774a810d28119bad740f317df377c",
      "tree": "e7f8cf7b09b8a6a4409c2d9798c23ec023213d45",
      "parents": [
        "5158f371faceb185235ba42882e67a840cfd7f48"
      ],
      "author": {
        "name": "Konstantin Ryabitsev",
        "email": "konstantin@linuxfoundation.org",
        "time": "Fri Nov 13 18:02:45 2026 -0500"
      },
      "committer": {
        "name": "Andrew Morton",
        "email": "akpm@linux-foundation.org",
        "time": "Tue Nov 17 21:40:01 2026 -0800"
      },
      "message": "get_maintainer: only use THE REST as a fallback\n\nPatch series \"get_maintainer: stop sending every patch to LKML\", v3.\n\nLKML has become mostly a firehose of patches that are already going to a\nsubsystem list.  The reason is THE REST: it matches every file in the\ntree, so get_maintainer.pl adds linux-kernel@vger.kernel.org to every\nsingle patch.  During the week of Oct 1-7, 2026:\n\n  - LKML carried 13,641 messages (~1,950/day), 90% of which were also\n    sent to at least one other list;\n  - with ~2,100 subscribers, that is about 29 million deliveries per\n    week for LKML alone.\n\nThis problem is becoming urgent and untenable.  The increased volume of\npatches sent using AI tooling has resulted in dramatically higher delivery\nqueues and if this change is not implemented, we will have to start\nforce-unsubscribing addresses that do not show any developer activity or\nuse similar drastic measures.\n\nThis issue came up during the 2026 Maintainer Summit and the consensus was\nthat \"always include linux-kernel for every patch\" was never the intent.\n\nThis series makes THE REST a fallback as initially intended: it is only\nadded for a file when no other matching MAINTAINERS section provides a\nmailing list that get_maintainer.pl would actually output for it.  Patch 2\nupdates submitting-patches.rst, which currently tells submitters to send\nevery patch to LKML.\n\nSeveral v1 reviewers asked for a catch-all list that would keep carrying\nevery patch for those who want to follow all kernel development.  That is\na worthwhile discussion, but it is separate from this series, which only\nrestores THE REST to its original role.\n\nTranslation maintainers are Cc\u0027d because the it_IT, ja_JP, sp_SP, zh_CN\nand zh_TW versions of submitting-patches.rst carry the same \"send every\npatch to LKML\" advice and will need a follow-up update once the English\ntext is settled.\n\n\nThis patch (of 2):\n\nTHE REST matches every file in the tree via \"F: *\" and \"F: */\", so\nget_maintainer.pl currently adds linux-kernel@vger.kernel.org to every\npatch, regardless of whether the touched files already belong to a\nsubsystem with its own mailing list.\n\nOnly add THE REST for a file when no other matching section provides a\nmailing list that get_maintainer.pl would actually output. Lists marked\n\"subscribers-only\" are omitted unless --s is given, and lists marked\n\"moderated\" are omitted with --no-moderated, so such lists don\u0027t count\nas a destination in those cases. Files that match no section, or only\nsections without such a list, still fall back to THE REST, so every\npatch keeps reaching at least one known-public mailing list.\n\nLink: https://lore.kernel.org/20261113-get-maintainer-the-rest-fallback-v3-0-36ee5a3e8d13@linuxfoundation.org\nLink: https://lore.kernel.org/20261113-get-maintainer-the-rest-fallback-v3-1-36ee5a3e8d13@linuxfoundation.org\nSigned-off-by: Konstantin Ryabitsev \u003ckonstantin@linuxfoundation.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nAcked-by: Joe Perches \u003cjoe@perches.com\u003e\nCc: Jonathan Corbet \u003ccorbet@lwn.net\u003e\nCc: Shuah Khan \u003cskhan@linuxfoundation.org\u003e\nCc: Randy Dunlap \u003crdunlap@infradead.org\u003e\nCc: Federico Vaga \u003cfederico.vaga@vaga.pv.it\u003e\nCc: Akira Yokosawa \u003cakiyks@gmail.com\u003e\nCc: Carlos Bilbao \u003ccarlos.bilbao@kernel.org\u003e\nCc: Avadhut Naik \u003cavadhut.naik@amd.com\u003e\nCc: Alex Shi \u003calexs@kernel.org\u003e\nCc: Yanteng Si \u003csi.yanteng@linux.dev\u003e\nCc: Dongliang Mu \u003cdzm91@hust.edu.cn\u003e\nCc: Chen-Yu Yeh \u003cchenyou910331@gmail.com\u003e\nCc: Weijie Yuan \u003cwy@wyuan.org\u003e\nCc: \u003cusers@kernel.org\u003e\nCc: Petr Pavlu \u003cpetr.pavlu@suse.com\u003e\n\n scripts/get_maintainer.pl |   43 ++++++++++++++++++++++++++++++++++++\n 1 file changed, 43 insertions(+)\n\n"
    },
    {
      "commit": "5158f371faceb185235ba42882e67a840cfd7f48",
      "tree": "dc342e1025b499af112d3b29e002171a66d0f0bb",
      "parents": [
        "27538caf7f72ce93be9c352148dba4f2274cf300"
      ],
      "author": {
        "name": "Zhiyuan Zhang",
        "email": "zhiyuanzhang912@gmail.com",
        "time": "Wed Nov 11 02:34:44 2026 +0000"
      },
      "committer": {
        "name": "Andrew Morton",
        "email": "akpm@linux-foundation.org",
        "time": "Tue Nov 17 21:40:01 2026 -0800"
      },
      "message": "ocfs2: fix kernel BUG in ocfs2_page_mkwrite() when write_end fails\n\nsyzbot reported a kernel BUG in the ocfs2 write fault path:\n\n  (syz.0.17,5578,0):ocfs2_write_end_nolock:1945 ERROR: status \u003d -30\n  kernel BUG at fs/ocfs2/mmap.c:106!\n  Call Trace:\n   do_page_mkwrite+0x147/0x310 mm/memory.c:3879\n   do_shared_fault mm/memory.c:6145 [inline]\n   handle_mm_fault+0x1a53/0x3750 mm/memory.c:6943\n\n__ocfs2_page_mkwrite() calls ocfs2_write_begin_nolock() and then\nocfs2_write_end_nolock(), and uses BUG_ON() to assert that the latter\nreturns the full length.  But ocfs2_write_end_nolock() can fail: since\ncommit 7f27ec978b0e (\"ocfs2: call ocfs2_journal_access_di() before\nocfs2_journal_dirty() in ocfs2_write_end_nolock()\") it gets journal\naccess to the inode buffer again, because the handle may have been\nrestarted since ocfs2_write_begin_nolock().  That fails with -EROFS or\n-EIO once the journal has been aborted or jbd2 has seen write errors\non the device.  The syzbot reproducer gets there by shrinking the\nbacking loop device to zero with LOOP_SET_STATUS while writing to a\nshared file mapping.\n\nHandle the failure instead of crashing.  ocfs2_write_end_nolock()\nleaves the mmap folio locked, because ocfs2_unlock_folios() skips it\nwhen w_target_locked is set so that it can be returned with\nVM_FAULT_LOCKED, and do_page_mkwrite() does not unlock it either when\n-\u003epage_mkwrite() returns an error.  So unlock it and fail the fault;\nthe faulting task gets SIGBUS, like for the other errors in this\nfunction.\n\nFor an mmap write ocfs2_write_end_nolock() returns either len or a\nnegative errno: inline data has already been converted to extents by\nocfs2_write_begin_nolock(), and copied is equal to len.\n\nThe BUG matters for local mounts, where the filesystem is expected to\ngo read-only and keep running after a journal abort.  On clustered\nmounts ocfs2_abort() panics the node anyway as soon as it notices the\naborted journal.\n\nLink: https://lore.kernel.org/20261111023444.22688-1-zhiyuanzhang912@gmail.com\nFixes: 7f27ec978b0e (\"ocfs2: call ocfs2_journal_access_di() before ocfs2_journal_dirty() in ocfs2_write_end_nolock()\")\nSigned-off-by: Zhiyuan Zhang \u003czhiyuanzhang912@gmail.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nReported-by: \u003csyzbot+42984127d45ad96219de@syzkaller.appspotmail.com\u003e\nCloses: https://syzkaller.appspot.com/bug?extid\u003d42984127d45ad96219de\nReviewed-by: Joseph Qi \u003cjoseph.qi@linux.alibaba.com\u003e\nReviewed-by: Heming Zhao \u003cheming.zhao@suse.com\u003e\nCc: Mark Fasheh \u003cmark@fasheh.com\u003e\nCc: Joel Becker \u003cjlbec@evilplan.org\u003e\nCc: yangwenfang \u003cvicky.yangwenfang@huawei.com\u003e\nCc: \u003cstable@vger.kernel.org\u003e\n\n fs/ocfs2/mmap.c |   12 +++++++++++-\n 1 file changed, 11 insertions(+), 1 deletion(-)\n\n"
    },
    {
      "commit": "27538caf7f72ce93be9c352148dba4f2274cf300",
      "tree": "8351e00fccdae7db688ca3472397a4fdbd425a26",
      "parents": [
        "fffc86cbe3ebfd7fea3937f8698e4322b3137984"
      ],
      "author": {
        "name": "Sang-Heon Jeon",
        "email": "ekffu200098@gmail.com",
        "time": "Fri Nov 06 01:01:58 2026 +0900"
      },
      "committer": {
        "name": "Andrew Morton",
        "email": "akpm@linux-foundation.org",
        "time": "Tue Nov 17 21:40:01 2026 -0800"
      },
      "message": "fault-inject: fix boot failure with fault injection boot parameters\n\nSince commit e9bc360b105c (\"fault-inject: use prandom where\ncryptographically secure randomness is not needed\"), setup_fault_attr()\ncalls prandom_init_once(), which calls kmalloc() and schedule_work().\n\nBut setup_fault_attr() can be called from __setup handlers before slab and\nworkqueues are initialized, so any fault injection boot parameter leads to\nboot failure.\n\nSo use DO_ONCE_LITE() instead.\n\nLink: https://lore.kernel.org/20261105160209.1504867-1-ekffu200098@gmail.com\nFixes: e9bc360b105c (\"fault-inject: use prandom where cryptographically secure randomness is not needed\")\nSigned-off-by: Sang-Heon Jeon \u003cekffu200098@gmail.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nReviewed-by: Akinobu Mita \u003cakinobu.mita@gmail.com\u003e\nCc: \u003cstable@vger.kernel.org\u003e\n\n lib/fault-inject.c |    8 +++++++-\n 1 file changed, 7 insertions(+), 1 deletion(-)\n\n"
    },
    {
      "commit": "fffc86cbe3ebfd7fea3937f8698e4322b3137984",
      "tree": "af374a2c2b72e885d764fa7164821d38d61bd186",
      "parents": [
        "e9e66aa5ef1a76ec92fe41c5eb16fda7a39db279"
      ],
      "author": {
        "name": "Cen Zhang",
        "email": "zzzccc427@gmail.com",
        "time": "Thu Oct 29 14:56:48 2026 +0800"
      },
      "committer": {
        "name": "Andrew Morton",
        "email": "akpm@linux-foundation.org",
        "time": "Tue Nov 17 21:40:01 2026 -0800"
      },
      "message": "ocfs2: advance locking_state through the live cursor\n\nlocking_state copies each lock resource under ocfs2_dlm_tracking_lock so\nits containing inode or quota object can be freed during the dump.  The\ncopied l_debug_list pointers do not have that lifetime protection.  Yet\nocfs2_dlm_seq_next() follows them to find the next record.\n\nRemounting read-only can free a quota-info object while a debugfs read is\nbetween the snapshot and the next iterator callback:\n\n  locking_state reader                Quota suspension\n  --------------------                ----------------\n  snapshot current lock resource\n    copy l_debug_list.next\n                                      remove next resource from live list\n                                      free quota-info object\n  seq_next(): follow copied link\n    read freed lock resource\n\nAdvance through the live dummy cursor instead.  List removal updates its\nlinks under the tracking lock.  Position it after the resource being\ncopied before releasing the lock, and never walk the snapshot\u0027s links. \nReuse the per-open snapshot for a nonzero seq_start() position, including\nread continuation and buffer retries.  Rewind the live cursor only when\nseq_file starts at index zero; seq_lseek() traverses from there through\nseq_next().  Clear the snapshot\u0027s l_ops marker at end of list so later\nreads remain at EOF.  This avoids replaying the list prefix under the\ntracking lock on each read, without adding a new cursor or position field.\n\nKASAN report as below:\n\n    BUG: KASAN: slab-use-after-free in ocfs2_dlm_next_res.isra.0+0x1b1/0x1d0\n    Read of size 8 at addr ffff888117ea9838 by task python3/489\n    \n    CPU: 3 UID: 0 PID: 489 Comm: python3 Not tainted 7.3.0-rc4-next-20260921-pmb-bt-functional-v1+ #1 PREEMPT(lazy) \n    Hardware name: QEMU Ubuntu 24.04 PC v2 (i440FX + PIIX, arch_caps fix, 1996), BIOS 1.16.3-debian-1.16.3-2 04/01/2014\n    Call Trace:\n     \u003cTASK\u003e\n     dump_stack_lvl+0x93/0xd0\n     print_report+0xce/0x630\n     ? ocfs2_dlm_next_res.isra.0+0x1b1/0x1d0\n     ? srso_alias_return_thunk+0x5/0xfbef5\n     ? __virt_addr_valid+0x20e/0x420\n     ? ocfs2_dlm_next_res.isra.0+0x1b1/0x1d0\n     kasan_report+0xe0/0x110\n     ? ocfs2_dlm_next_res.isra.0+0x1b1/0x1d0\n     ocfs2_dlm_next_res.isra.0+0x1b1/0x1d0\n     ? __pfx_ocfs2_dlm_next_res.isra.0+0x10/0x10\n     ? __pfx_do_raw_spin_lock+0x10/0x10\n     ocfs2_dlm_seq_next+0xb8/0x4c0\n     ? srso_alias_return_thunk+0x5/0xfbef5\n     seq_read_iter+0x92f/0x1110\n     ? srso_alias_return_thunk+0x5/0xfbef5\n     seq_read+0x362/0x5a0\n     ? __pfx_seq_read+0x10/0x10\n     ? srso_alias_return_thunk+0x5/0xfbef5\n     ? srso_alias_return_thunk+0x5/0xfbef5\n     ? selinux_file_permission+0x11d/0x540\n     full_proxy_read+0x11f/0x180\n     ? __pfx_full_proxy_read+0x10/0x10\n     vfs_read+0x1a0/0xc20\n     ? srso_alias_return_thunk+0x5/0xfbef5\n     ? kmem_cache_free+0x242/0x580\n     ? __pfx_vfs_read+0x10/0x10\n     ? srso_alias_return_thunk+0x5/0xfbef5\n     ? putname+0x93/0xe0\n     ? srso_alias_return_thunk+0x5/0xfbef5\n     ? do_sys_openat2+0x12a/0x1b0\n     ? __pfx_do_sys_openat2+0x10/0x10\n     ksys_read+0x111/0x200\n     ? __pfx_ksys_read+0x10/0x10\n     ? irqentry_exit+0xbe/0x820\n     ? srso_alias_return_thunk+0x5/0xfbef5\n     ? trace_hardirqs_on_prepare+0x11d/0x150\n     do_syscall_64+0x114/0x620\n     entry_SYSCALL_64_after_hwframe+0x77/0x7f\n    RIP: 0033:0x7f0266dbb687\n    Code: 48 89 fa 4c 89 df e8 58 b3 00 00 8b 93 08 03 00 00 59 5e 48 83 f8 fc 74 1a 5b c3 0f 1f 84 00 00 00 00 00 48 8b 44 24 10 0f 05 \u003c5b\u003e c3 0f 1f 80 00 00 00 00 83 e2 39 83 fa 08 75 de e8 23 ff ff ff\n    RSP: 002b:00007ffccf76fc40 EFLAGS: 00000202 ORIG_RAX: 0000000000000000\n    RAX: ffffffffffffffda RBX: 00007f0266d27780 RCX: 00007f0266dbb687\n    RDX: 0000000000000001 RSI: 00007f0266a53200 RDI: 0000000000000004\n    RBP: 00007f0266d27700 R08: 0000000000000000 R09: 0000000000000000\n    R10: 0000000000000000 R11: 0000000000000202 R12: 0000000000000001\n    R13: 00007f0266a53200 R14: 0000000000000004 R15: 00007f0266b651e0\n     \u003c/TASK\u003e\n    \n    Allocated by task 547:\n     kasan_save_stack+0x33/0x60\n     kasan_save_track+0x14/0x30\n     __kasan_kmalloc+0xaa/0xb0\n     __kmalloc_cache_noprof+0x28d/0x610\n     ocfs2_local_read_info+0x181/0x1580\n     dquot_load_quota_sb+0x68b/0x1000\n     dquot_load_quota_inode+0x227/0x3e0\n     ocfs2_enable_quotas+0x27c/0x520\n     ocfs2_fill_super+0x21f9/0x38f0\n     get_tree_bdev_flags+0x35d/0x5f0\n     vfs_get_tree+0x93/0x320\n     fc_mount+0x1d/0x200\n     path_mount+0x592/0x1e00\n     __x64_sys_mount+0x24b/0x2d0\n     do_syscall_64+0x114/0x620\n     entry_SYSCALL_64_after_hwframe+0x77/0x7f\n    \n    Freed by task 577:\n     kasan_save_stack+0x33/0x60\n     kasan_save_track+0x14/0x30\n     kasan_save_free_info+0x3b/0x60\n     __kasan_slab_free+0x5f/0x80\n     kfree+0x308/0x580\n     ocfs2_local_free_info+0x47b/0x800\n     dquot_disable+0x5d9/0x17a0\n     ocfs2_susp_quotas.isra.0+0x22c/0x2b0\n     ocfs2_reconfigure+0x284/0xd00\n     reconfigure_super+0x3ca/0xa70\n     path_mount+0x16ae/0x1e00\n     __x64_sys_mount+0x24b/0x2d0\n     do_syscall_64+0x114/0x620\n     entry_SYSCALL_64_after_hwframe+0x77/0x7f\n    \n    The buggy address belongs to the object at ffff888117ea9800\n     which belongs to the cache kmalloc-1k of size 1024\n    The buggy address is located 56 bytes inside of\n     freed 1024-byte region [ffff888117ea9800, ffff888117ea9c00)\n    \n    The buggy address belongs to the physical page:\n    page: refcount:0 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x117ea8\n    head: order:3 mapcount:0 entire_mapcount:0 nr_pages_mapped:0 pincount:0\n    flags: 0x200000000000040(head|node\u003d0|zone\u003d2)\n    page_type: f5(slab)\n    raw: 0200000000000040 ffff888100042dc0 dead000000000100 dead000000000122\n    raw: 0000000000000000 0000000000100010 00000000f5000000 0000000000000000\n    head: 0200000000000040 ffff888100042dc0 dead000000000100 dead000000000122\n    head: 0000000000000000 0000000000100010 00000000f5000000 0000000000000000\n    head: 0200000000000003 fffffffffffffe01 00000000ffffffff 00000000ffffffff\n    head: 0000000000000000 0000000000000000 00000000ffffffff 0000000000000000\n    page dumped because: kasan: bad access detected\n    \n    Memory state around the buggy address:\n     ffff888117ea9700: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc\n     ffff888117ea9780: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc\n    \u003effff888117ea9800: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb\n                                            ^\n     ffff888117ea9880: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb\n     ffff888117ea9900: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb\n    \u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nLink: https://lore.kernel.org/pm-ocfs2-objects-candidate-0053-v7-ec41d4bbd38bb4d795a8@gmail.com\nFixes: ccd979bdbce9 (\"[PATCH] OCFS2: The Second Oracle Cluster Filesystem\")\nSigned-off-by: Cen Zhang \u003czzzccc427@gmail.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSuggested-by: Joseph Qi \u003cjoseph.qi@linux.alibaba.com\u003e\nReviewed-by: Joseph Qi \u003cjoseph.qi@linux.alibaba.com\u003e\nCc: Mark Fasheh \u003cmark@fasheh.com\u003e\nCc: Joel Becker \u003cjlbec@evilplan.org\u003e\nCc: Heming Zhao \u003cheming.zhao@suse.com\u003e\nCc: Jia-Ju Bai \u003cbaijiaju1990@gmail.com\u003e\nCc: \u003cjjzuming@gmail.com\u003e\nCc: \u003cstable@vger.kernel.org\u003e\n\n fs/ocfs2/dlmglue.c |   19 ++++++++++++++++---\n 1 file changed, 16 insertions(+), 3 deletions(-)\n\n"
    },
    {
      "commit": "e9e66aa5ef1a76ec92fe41c5eb16fda7a39db279",
      "tree": "58fb2bf19a48a385d34236dd68310f577d14d3de",
      "parents": [
        "f277230d002d03ed98b814b735f9d3d01b42020a"
      ],
      "author": {
        "name": "Cen Zhang",
        "email": "zzzccc427@gmail.com",
        "time": "Thu Oct 29 15:00:13 2026 +0800"
      },
      "committer": {
        "name": "Andrew Morton",
        "email": "akpm@linux-foundation.org",
        "time": "Tue Nov 17 21:40:00 2026 -0800"
      },
      "message": "ocfs2/dlmfs: keep the superblock alive during lock work\n\nQueued dlmfs lock work holds an inode reference, but no active reference\nto its superblock.  The workqueue is drained on module exit rather than on\neach dlmfs unmount, and an inode reference does not stop superblock\nshutdown: generic_shutdown_super() reports the busy inode and goes on.\n\n  DLM callback / worker               Unmount\n  ---------------------               -------\n  queue lock work; igrab(inode)\n  worker delayed\n                                      drop last active superblock ref\n                                      kill_anon_super()\n                                        report the busy inode\n                                        free the superblock\n  finish lock work\n  iput(inode)\n    access the inode\u0027s freed superblock\n\nFlushing the workqueue from a dlmfs kill_sb would not be enough: a\nblocking AST can queue new work for an inode that is still alive after the\nflush and before eviction, and that work again holds only an inode\nreference.\n\nTake an active superblock reference before acquiring the queued inode\nreference, with the atomic_inc_not_zero() on s_active that grab_super()\nand the NFS and procfs superblock walkers use.  Decline to queue if the\nsuperblock is already shutting down; the inode is then being evicted and\nuser_dlm_destroy_lock() tears the lock down.  The existing l_lock and\nUSER_LOCK_QUEUED flag serialize this with other queue attempts.  Each\nexecution drops its inode reference before releasing the active superblock\nreference, outside l_lock.  If that releases the last active reference,\nthe unmount completes on the worker; it only waits for DLM replies and\nnever for this workqueue.\n\nOnly the relevant frames of the report are kept:\n\nOops: general protection fault, probably for non-canonical address 0xfbd59c0000000026: 0000 [#1] SMP KASAN NOPTI\nKASAN: maybe wild-memory-access in range [0xdead000000000130-0xdead000000000137]\nWorkqueue: user_dlm user_dlm_unblock_lock\nRIP: 0010:iput+0x246/0xc10\nCall Trace:\n \u003cTASK\u003e\n user_dlm_unblock_lock+0x503/0xa50\n process_one_work+0x9a8/0x1b40\n worker_thread+0x65c/0xe40\n \u003c/TASK\u003e\n\nLink: https://lore.kernel.org/pm-ocfs2-objects-candidate-0054-v6-c3887cceae3b039e591e@gmail.com\nFixes: 8df08c89c668 (\"[PATCH] OCFS2: The Second Oracle Cluster Filesystem\")\nSigned-off-by: Cen Zhang \u003czzzccc427@gmail.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nReviewed-by: Joseph Qi \u003cjoseph.qi@linux.alibaba.com\u003e\nCc: Mark Fasheh \u003cmark@fasheh.com\u003e\nCc: Joel Becker \u003cjlbec@evilplan.org\u003e\nCc: Heming Zhao \u003cheming.zhao@suse.com\u003e\nCc: Jia-Ju Bai \u003cbaijiaju1990@gmail.com\u003e\nCc: \u003cjjzuming@gmail.com\u003e\nCc: \u003cstable@vger.kernel.org\u003e\n\n fs/ocfs2/dlmfs/userdlm.c |   12 ++++++++++--\n 1 file changed, 10 insertions(+), 2 deletions(-)\n\n"
    },
    {
      "commit": "f277230d002d03ed98b814b735f9d3d01b42020a",
      "tree": "2aa4093bb315dc001af2f7f3508a539291d3f500",
      "parents": [
        "fa4ad798f6f681031775266044ba1b0cf3c670bd"
      ],
      "author": {
        "name": "Cen Zhang",
        "email": "zzzccc427@gmail.com",
        "time": "Fri Oct 30 12:48:56 2026 +0800"
      },
      "committer": {
        "name": "Andrew Morton",
        "email": "akpm@linux-foundation.org",
        "time": "Tue Nov 17 21:40:00 2026 -0800"
      },
      "message": "ocfs2: serialize inode journal access with shutdown\n\nThe downconvert thread can run inode cleanup until ocfs2_dlm_shutdown()\nstops it, but unmount shuts down the journal earlier. A final dentry\npost-unlock put can therefore enter ocfs2_clear_inode() while the journal\nis being destroyed.\n\nocfs2_ci_fully_checkpointed() loads osb-\u003ejournal before taking\ntrans_inc_lock. Journal shutdown neither takes that lock for removal\nnor waits for the downconvert thread before freeing the journal:\n\n  Downconvert thread                    Unmount\n  ------------------                    -------\n  dentry post-unlock; final iput()\n    ocfs2_checkpoint_inode()\n      load osb-\u003ejournal\n                                        ocfs2_journal_shutdown()\n                                          free journal; clear pointer\n      read the transaction ID\n\nThe checkpoint wait queue is embedded in that journal, and the final\njbd2_journal_release_jbd_inode() also needs its JBD2 journal to remain\nalive for the whole call. Protecting only the pointer load would leave\nthose longer uses unprotected.\n\nRead and clear osb-\u003ejournal under trans_inc_lock when checking checkpoint\nprogress. Treat a removed journal as checkpointed. Move checkpoint\nwaiters to the existing mount-owned checkpoint_event and wake it after\ncheckpoint progress and journal removal, so the queue outlives the\njournal. The commit thread also waits there and rechecks its condition\nwhen woken.\n\nSerialize the final JBD2 inode release against JBD2 destruction and the\njournal free with a mount-owned rw_semaphore. Take the write side only\nafter stopping the commit thread, and drop it before the journal inode\u0027s\nfinal iput(), whose cleanup can take the read side. Inode cleanup holds\nthe read side only around the JBD2 release, outside checkpoint waits and\nDLM lock teardown.\n\nKASAN report as below:\n\n    Oops: general protection fault, probably for non-canonical address 0xdffffc000000000e: 0000 [#1] SMP KASAN NOPTI\n    KASAN: null-ptr-deref in range [0x0000000000000070-0x0000000000000077]\n    CPU: 0 UID: 0 PID: 587 Comm: ocfs2dc-33CAF04 Not tainted 7.3.0-rc4-next-20260921-pmb-ocfs2-functional-v1+ #1 PREEMPT(lazy) \n    Hardware name: QEMU Ubuntu 24.04 PC v2 (i440FX + PIIX, arch_caps fix, 1996), BIOS 1.16.3-debian-1.16.3-2 04/01/2014\n    RIP: 0010:ocfs2_clear_inode+0x8ae/0x1630\n    Code: fc ff df 48 c1 ea 03 80 3c 02 00 0f 85 27 0c 00 00 49 8d 7d 70 48 8b 43 c8 48 ba 00 00 00 00 00 fc ff df 48 89 f9 48 c1 e9 03 \u003c80\u003e 3c 11 00 0f 85 f0 0b 00 00 49 3b 45 70 0f 88 62 03 00 00 e8 49\n    RSP: 0018:ffff888118d3f878 EFLAGS: 00010212\n    RAX: 0000000000000001 RBX: ffff88811a5edc60 RCX: 000000000000000e\n    RDX: dffffc0000000000 RSI: 0000000000000001 RDI: 0000000000000070\n    RBP: ffff888116134000 R08: 0000000000000001 R09: 0000000000000001\n    R10: ffffffff8965765f R11: 0000000000000001 R12: ffff88811a5ed400\n    R13: 0000000000000000 R14: ffff88811a5edbf8 R15: ffff88811a5ed408\n    FS:  0000000000000000(0000) GS:ffff8881fd4dc000(0000) knlGS:0000000000000000\n    CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033\n    CR2: 00007ff35b66f2c0 CR3: 00000001092ae005 CR4: 0000000000770ef0\n    PKRU: 55555554\n    Call Trace:\n     \u003cTASK\u003e\n     ? __pfx_ocfs2_clear_inode+0x10/0x10\n     ? __pfx_ocfs2_inode_is_valid_to_delete+0x10/0x10\n     ? srso_alias_return_thunk+0x5/0xfbef5\n     ? __lock_acquire+0x5a3/0x2140\n     ocfs2_evict_inode+0x1ed/0x11e0\n     ? inode_wait_for_writeback+0x14c/0x330\n     ? __pfx_ocfs2_evict_inode+0x10/0x10\n     ? lock_acquire+0x190/0x300\n     ? srso_alias_return_thunk+0x5/0xfbef5\n     ? find_held_lock+0x2b/0x80\n     ? evict+0x347/0x9f0\n     ? srso_alias_return_thunk+0x5/0xfbef5\n     ? lock_release+0xc8/0x290\n     ? _raw_spin_unlock+0x23/0x40\n     ? __pfx_ocfs2_evict_inode+0x10/0x10\n     evict+0x38b/0x9f0\n     ? __pfx_evict+0x10/0x10\n     ? lock_acquire+0x190/0x300\n     ? srso_alias_return_thunk+0x5/0xfbef5\n     ? lock_release+0xc8/0x290\n     ? __pfx_inode_just_drop+0x10/0x10\n     iput+0x680/0xb80\n     ocfs2_dentry_lock_put+0xc2/0x110\n     ? __pfx_ocfs2_dentry_post_unlock+0x10/0x10\n     ocfs2_process_blocked_lock+0x1fb/0x420\n     ? __pfx_ocfs2_process_blocked_lock+0x10/0x10\n     ? __pfx_pmbd_probe_hit_cookie+0x10/0x10\n     ? lock_release+0xc8/0x290\n     ? srso_alias_return_thunk+0x5/0xfbef5\n     ? lockdep_hardirqs_on_prepare+0xeb/0x1a0\n     ? srso_alias_return_thunk+0x5/0xfbef5\n     ocfs2_downconvert_thread+0x32b/0x780\n     ? do_raw_spin_lock+0x130/0x270\n     ? __pfx_ocfs2_downconvert_thread+0x10/0x10\n     ? __kthread_parkme+0xc1/0x220\n     ? srso_alias_return_thunk+0x5/0xfbef5\n     ? lock_release+0xc8/0x290\n     ? __pfx_autoremove_wake_function+0x10/0x10\n     ? srso_alias_return_thunk+0x5/0xfbef5\n     ? srso_alias_return_thunk+0x5/0xfbef5\n     ? __kthread_parkme+0x177/0x220\n     ? __pfx_ocfs2_downconvert_thread+0x10/0x10\n     kthread+0x351/0x460\n     ? srso_alias_return_thunk+0x5/0xfbef5\n     ? __pfx_kthread+0x10/0x10\n     ret_from_fork+0x659/0x940\n     ? __pfx_ret_from_fork+0x10/0x10\n     ? srso_alias_return_thunk+0x5/0xfbef5\n     ? __switch_to+0x74f/0xf70\n     ? __pfx_kthread+0x10/0x10\n     ret_from_fork_asm+0x1a/0x30\n     \u003c/TASK\u003e\n    Modules linked in:\n    ---[ end trace 0000000000000000 ]---\n\nLink: https://lore.kernel.org/pm-ocfs2-objects-candidate-0004-v8-dbcb8f80b8df476084d6@gmail.com\nFixes: da5e7c87827e (\"ocfs2: cleanup journal init and shutdown\")\nSigned-off-by: Cen Zhang \u003czzzccc427@gmail.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nReviewed-by: Joseph Qi \u003cjoseph.qi@linux.alibaba.com\u003e \nCc: Mark Fasheh \u003cmark@fasheh.com\u003e\nCc: Joel Becker \u003cjlbec@evilplan.org\u003e\nCc: Heming Zhao \u003cheming.zhao@suse.com\u003e\nCc: Jia-Ju Bai \u003cbaijiaju1990@gmail.com\u003e\nCc: \u003cjjzuming@gmail.com\u003e\nCc: \u003cstable@vger.kernel.org\u003e\n\n fs/ocfs2/inode.c   |   17 +++++++++--------\n fs/ocfs2/journal.c |   30 ++++++++++++++++++++++--------\n fs/ocfs2/journal.h |   17 +++++++++--------\n fs/ocfs2/ocfs2.h   |    1 +\n fs/ocfs2/super.c   |    1 +\n 5 files changed, 42 insertions(+), 24 deletions(-)\n\n"
    },
    {
      "commit": "20eef29b7c8465ac98d7dd6d6f2ac07a8d8056ca",
      "tree": "037fa1e8a4f19e14b7dd84c966d795000b332481",
      "parents": [
        "7ccb044c27720d48175577080b851f047ddcfd1c",
        "603b4fd69b4e77770761ea03451e2164186f8590"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Tue Nov 17 18:55:10 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Tue Nov 17 18:55:10 2026 -0800"
      },
      "message": "Merge tag \u0027arc-fixes-7.4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/vgupta/arc\n\nPull ARC fix from Vineet Gupta:\n\n - workaround a GCC bug which has been causing upstream build fails\n\n* tag \u0027arc-fixes-7.4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/vgupta/arc:\n  ARC: io: work around GCC PR127718 in __raw_write{w,l}()\n"
    },
    {
      "commit": "7ccb044c27720d48175577080b851f047ddcfd1c",
      "tree": "6da4276d0f64b391105d0459e68ebc02017a5198",
      "parents": [
        "63fdb511c27afcebce797e3db580b3534ddfa639",
        "9500416c8aa7c271b85be6a0d5856da62b60223a"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Tue Nov 17 18:27:24 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Tue Nov 17 18:27:24 2026 -0800"
      },
      "message": "Merge tag \u0027powerpc-7.4-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux\n\nPull powerpc fixes from Madhavan Srinivasan:\n\n - Handle of_property_count_u32_elems() error in parse_thread_groups()\n\n - Defer Nemo IRQ initialization\n\n - disable watchdog when watchdog_thresh\u003d0\n\n - Reserve a PLT entry for ftrace_regs_caller\n\n - powerpc KVM fixes\n\n - misc fixes and cleanups\n\nThanks to Amit Machhiwal, Bradley Morgan, Christian Zigotzky, Christophe\nLeroy, Gautam Menghani, Jiale Yao, Luiz Capitulino, Michael Ellerman,\nMohammad Shahid, Nathan Chancellor, Ritesh Harjani, Sasha Levin,\nShrikanth Hegde, and Timothy Pearson\n\n* tag \u0027powerpc-7.4-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:\n  powerpc/perf/vpa-dtl: use GFP_ATOMIC for DTL buffer allocation\n  powerpc32/modules: Reserve a PLT entry for ftrace_regs_caller\n  powerpc/watchdog: disable watchdog when watchdog_thresh\u003d0\n  selftests/powerpc: override EMIT_TESTS instead of the emit_tests rule\n  powerpc/smp: Handle of_property_count_u32_elems() error in parse_thread_groups()\n  powerpc/pasemi: Defer Nemo IRQ initialization\n  KVM: PPC: Book3S HV: Avoid spurious interrupts caused by LPCR_MER bit\n  KVM: PPC: Book3S HV: Take SRCU read lock around kvmppc_h_page_init() call\n  KVM: PPC: Book3S HV: XICS: Avoid NULL ptr deref in kvmppc_xive_cleanup_vcpu()\n  KVM: PPC: Book3S HV: XICS: Update irq_type for cleanup to work\n  KVM: PPC: Book3S HV: Ensure that calls to idr_alloc are synchronized\n  KVM: PPC: Book3S HV: Disable preemption around host-side HPT lock users\n  powerpc/eeh: NUL-terminate force recovery input\n"
    },
    {
      "commit": "63fdb511c27afcebce797e3db580b3534ddfa639",
      "tree": "0cbd5e51ca9cef10e99632878a6002e4ea078096",
      "parents": [
        "666539eecdf4db623f5e3c3e55cf9f2dfad24aa4",
        "25f5c92e1ac06496679ea229e5c825aa88968efd"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Tue Nov 17 11:51:40 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Tue Nov 17 11:51:40 2026 -0800"
      },
      "message": "Merge tag \u0027mm-core-7.4-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mm/core\n\nPull core memory management fixes from David Hildenbrand:\n \"The main changes are:\n\n   - Prevent folio refcount overflows with pagecache folios, mostly\n     relevant for large folios today. The networking changes for this\n     will go through the net tree.\n\n   - Fix out-of-bounds page table accesses with folded PMDs in 3\n     scenarios.\n\n   - Fix write-sealing of memfd to detect GUP pins with concurrent\n     (un)mapping more reliably, preventing writes through GUP pins after\n     write-sealing.\n\n   - Fix lost wakeups when splitting large folios.\n\n   - Fix exposing uninitialized data through MADV_COLLAPSE to user\n     space.\n\n  The refcount changes also include a related iommufd fix for properly\n  cleaning up when pinning memfd folios fails, which is now easier to\n  trigger\"\n\n* tag \u0027mm-core-7.4-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mm/core:\n  mm/gup: limit pagecache folio refcount when grabbing references\n  iommufd: fix cleanup when adding memfd folio pins fails\n  mm/mmu_gather: flush pending references early before the folio limit\n  mm: limit pagecache folio refcount when splicing into pipes\n  mm: limit pagecache folio refcount when mapping pages\n  mm/huge_memory: prevent losing PG_waiters in __split_folio_to_order()\n  mm/rmap: fix PTE table overruns during batched reference checks\n  mm/rmap: fix PTE table overruns during batched unmap\n  mm/userfaultfd: fix PTE table overruns in UFFDIO_MOVE\n  mm/memfd: handle GUP pins more reliably when write-sealing\n  mm/khugepaged: don\u0027t map folios that are not uptodate in MADV_COLLAPSE\n"
    },
    {
      "commit": "666539eecdf4db623f5e3c3e55cf9f2dfad24aa4",
      "tree": "88d260f400bac3469d55cce5ab2774eef35bf903",
      "parents": [
        "141b164bbe186423e295b139e40e8e2d11cf1341",
        "bbab563c51315a1c6dbf6e801e4237b6d2b8b81e"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Mon Nov 16 09:35:38 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Mon Nov 16 09:35:38 2026 -0800"
      },
      "message": "Merge tag \u0027assoc-array-fixes-v2\u0027\n\nClean up assoc_array library code\n\nI had already gotten rid of this tag locally, because we then decided\nto just remove the assoc_array library entirely since it has only one\nsingle user - the key handling code - and that user is getting rewritten\nto use rhashtable instead.\n\nBut that removal is scheduled for 7.5, and this fixes real bugs and does\nclean up some of the code.  So while it\u0027s all slated for removal, some\nof the fixes need back-porting and we might as well remove the\ncleaned-up version.  Maybe somebody decides they want to resurrect it\nfor another project - it really is fairly generic library code - they\nshouldn\u0027t get the messy end state that we decided to jettison.\n\nBackground: Nils Ollrogge reported a bug in the assoc_array library and\nhad a patch to fix it too, but that code was too obscure for me to\nfollow, so I used some LLM tools to help make sense of it all and found\nmore issues in the process.\n\n* tag \u0027assoc-array-fixes-v2\u0027:\n  assoc_array: use assoc_array_parent_link() when splitting a shortcut\n  assoc_array: find and collapse in deletion without gotos\n  assoc_array: move deletion\u0027s collapse into a function of its own\n  assoc_array: replace the labels in gc\u0027s tree walk with a loop\n  assoc_array: move gc\u0027s folding into a function of its own\n  assoc_array: keep gc\u0027s cleanup out of the live tree on allocation failure\n  assoc_array: don\u0027t chain a new shortcut onto an existing one\n  assoc_array: choose between splitting a node and a shortcut without gotos\n  assoc_array: move shortcut creation into a function of its own\n  assoc_array: move node splitting into a function of its own\n  assoc_array: fill the segment cache in one place\n  assoc_array: move in-place insertion into a function of its own\n  assoc_array: don\u0027t lose objects when a root shortcut is split at level 4\n  assoc_array: don\u0027t lose objects when a shortcut is split on a word boundary\n  assoc_array: correct the documentation of get_key_chunk()\u0027s level argument\n"
    },
    {
      "commit": "141b164bbe186423e295b139e40e8e2d11cf1341",
      "tree": "6a5ae31558c8823927d8129582f947d713603e04",
      "parents": [
        "b6793a9e2a3c280bb7c6c7d39be271e2f57aa2bb",
        "4dbe5ea929ac5a5b8dc21787d0e79fc55a0ea9ac"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Mon Nov 16 08:02:58 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Mon Nov 16 08:02:58 2026 -0800"
      },
      "message": "Merge tag \u0027char-misc-7.4-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc\n\nPull char/misc/iio driver fixes from Greg KH:\n \"Here are a bunch of char, misc, and iio driver fixes for reported\n  problems for 7.4-rc4. Included in here are:\n\n   - usual collection of lots of IIO driver fixes for issues\n\n   - binder driver fix\n\n   - iio, w1, and comedi driver fixes found and fixed by Josef\n\n   - comedi driver fix for a reported problem\n\n   - other misc driver fixes\n\n  All of these have been in linux-next for a while with no reported\n  issues\"\n\n* tag \u0027char-misc-7.4-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc: (68 commits)\n  iio: hid-sensors: fix fractional part for unit exponents -7 and -8\n  iio: trigger: free the trigger ID after device_del()\n  iio: hid-sensors: fix a division by zero\n  iio: humidity: hid-sensor-humidity: Use per-instance callbacks\n  iio: temperature: hid-sensor-temperature: Use per-instance callbacks\n  comedi: comedi_bond: Fix leak of opened device references\n  mei: me: add ice lake point LP iTouch DID\n  mei: lb: read the MTU once\n  rust_binder: push BR_TRANSACTION_COMPLETE before BR_FAILED_REPLY\n  misc: ibmvmc: initialize HMC state before device registration\n  w1: w1_therm: leave the slaves of other families alone\n  w1: w1_therm: lock the list of slaves for the bulk read\n  w1: don\u0027t hold list_mutex while a slave is torn down\n  w1: netlink: drop the slave after the bus is unlocked\n  w1: ds2805: use the bus mutex to talk on the bus\n  w1: ds28e04: use the bus mutex to talk on the bus\n  w1: ds2482: set the strong pull-up with its byte\n  w1: remove the master\u0027s files before taking its locks\n  w1: use the bus mutex for the \"rw\" file\n  misc: bcm-vk: fix DMA allocation memory leak\n  ...\n"
    },
    {
      "commit": "b6793a9e2a3c280bb7c6c7d39be271e2f57aa2bb",
      "tree": "7eecd2cd6c00f32c13325e57c410214074463c07",
      "parents": [
        "f11fe7603df07fbb9fa02b0d385c1696c87a6dab",
        "eedf2cf3505c1562a2fb24f8b997056f023d242b"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Mon Nov 16 07:59:28 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Mon Nov 16 07:59:28 2026 -0800"
      },
      "message": "Merge tag \u0027tty-7.4-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty\n\nPull tty/serial driver fixes from Greg KH:\n \"Here are some small tty and serial driver fixes for reported problems\n  for 7.4-rc4. Included in here are:\n\n   - tty driver and vt fixes reported and fixed by Josef Bacik and Chris\n     Mason\n\n   - some serial driver fixes by Jiri for found problems\n\n  All of these have been in linux-next for almost 2 weeks now, with no\n  reported issues\"\n\n* tag \u0027tty-7.4-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty:\n  serial: samsung: put the previous baud clock\n  serial: samsung: don\u0027t leak the tty on an RX IRQ\n  serial: rda: free the interrupt on shutdown\n  serial: imx: fix the tty reference of the wakeup event\n  serial: fsl_lpuart: fix tty ref of wakeup event\n  serial: 8250_mxpcie: handle hangup during break\n  serial: 8250_lpss: don\u0027t divide by 0 above 6.25 Mbaud\n  serial: core: respect runtime PM when sending x_char\n  vt: serialise the producers of the flip buffer\n  tty: fix misapplied compat_ptr() in tty_compat_ioctl()\n  serial: 8250_hp300: reorder HPDCA code\n  serial: 8250_exar: check priv-\u003egpio properly\n"
    },
    {
      "commit": "f11fe7603df07fbb9fa02b0d385c1696c87a6dab",
      "tree": "30889aba8d0655c2ef7de555daa86051bf5d134f",
      "parents": [
        "7f6fccf6ee5a345b778706b22159c4e3b547c808",
        "3149e8b3237f58d4dd440fc64c45c93404c4b064"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Mon Nov 16 07:55:02 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Mon Nov 16 07:55:02 2026 -0800"
      },
      "message": "Merge tag \u0027usb-7.3-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb\n\nPull USB/Thunderbolt fixes from Greg KH:\n \"Here are a bunch of USB (and one Thunderbolt) fixes. Sorry for the\n  delay on them, been stuck in travel hell for a bit and these have been\n  sitting all ready to be merged for too long. Included in here are:\n\n   - lots of tiny USB fixes found by anthropic scans\n\n   - usb-serial bugfixes\n\n   - xhci driver fixes for issues regressions by many people\n\n   - dwc3 driver fixes for reported problems\n\n   - some usb gadget driver fixes for reported problems\n\n   - other minor USB driver fixes\n\n  Most of these have been in linux-next for weeks with no reported\n  issues, the others have been in linux-next for 11 days, also with no\n  reported problems\"\n\n* tag \u0027usb-7.3-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb: (45 commits)\n  usb: xhci: fix \u0027num_active_eps\u0027 accounting on allocation failure\n  usb: xhci: clear stale bandwidth data after hibernation\n  usb: pci-quirks: always assert xHCI OS ownership\n  usb: xhci: disable the correct slot on alloc_dev failure\n  xhci: keep port link_inactive and connected flags in sync when polled\n  usb: typec: ucsi: retry init only while the PPM is not ready\n  xhci: give back TDs and clean up ep0 after disconnect\n  USB: serial: garmin_gps: fix deadlock on reset request\n  USB: serial: oti6858: fix the teardown order in close\n  USB: serial: io_edgeport: serialise MCR accesses\n  USB: serial: io_edgeport: fix potential use-after-free on disconnect\n  USB: serial: io_edgeport: fix potential fifo corruption\n  USB: serial: io_edgeport: move initial termios update to open()\n  USB: serial: io_edgeport: fix potential NULL-deref at open\n  USB: serial: io_edgeport: fix potential close() use-after-free\n  USB: serial: io_edgeport: fix NULL-deref on early status\n  USB: serial: upd78f0730: fix NULL-deref on hangup\n  USB: serial: sierra: validate URB length in instat callback\n  USB: serial: garmin_gps: validate packet data length in nat_receive()\n  usb: gadget: f_ncm: restart the TX timer if the TX freelist is empty\n  ...\n"
    },
    {
      "commit": "7f6fccf6ee5a345b778706b22159c4e3b547c808",
      "tree": "33732fcb6f8ed711ad47955a95da3aff154f96f3",
      "parents": [
        "aebb7274476403040d582c61478df8f4a7bf9cd3",
        "c3f98ff6309a29979e1867bc3c9d4172755ca122"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Mon Nov 16 07:26:07 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Mon Nov 16 07:26:07 2026 -0800"
      },
      "message": "Merge tag \u0027rcu-fixes.v7.4-20261115a\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rcu/linux\n\nPull RCU fixes from Joel Fernandes:\n \"Fix a build break with recent clang, which warns about unsequenced\n  accesses to the volatile jiffies variable in show_rcu_nocb_state() and\n  fails the build with CONFIG_WERROR\u003dy / W\u003de. Read jiffies once into a\n  local variable instead.\n\n  Also replace the rcupdate.rcu_task_collapse_debug module parameter,\n  added in the v7.4 merge window, with pr_debug() for the RCU Tasks\n  callback contend/collapse messages. The messages stay off by default\n  and can be enabled with dynamic debug, and no official kernel carries\n  the extra parameter\"\n\n* tag \u0027rcu-fixes.v7.4-20261115a\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rcu/linux:\n  rcu/nocb: Avoid unsequenced accesses to jiffies in show_rcu_nocb_state()\n  rcu-tasks: Use pr_debug() for callback contend/collapse messages\n"
    },
    {
      "commit": "c3f98ff6309a29979e1867bc3c9d4172755ca122",
      "tree": "33732fcb6f8ed711ad47955a95da3aff154f96f3",
      "parents": [
        "38828d44f31e4366b2963d5786d06d3358e8ae58"
      ],
      "author": {
        "name": "Nathan Chancellor",
        "email": "nathan@kernel.org",
        "time": "Sat Nov 14 21:26:10 2026 -0700"
      },
      "committer": {
        "name": "Joel Fernandes",
        "email": "jfern@kernel.org",
        "time": "Sun Nov 15 18:52:04 2026 -0500"
      },
      "message": "rcu/nocb: Avoid unsequenced accesses to jiffies in show_rcu_nocb_state()\n\nAfter a recent improvement in clang to warn on unsequenced accesses to\nvolatile variables [1], an instance appears in show_rcu_nocb_state(),\nwhich can break the build with CONFIG_WERROR\u003dy / W\u003de:\n\n  In file included from kernel/rcu/tree.c:5123:\n  kernel/rcu/tree_nocb.h:1651:3: error: unsequenced accesses to volatile qualified \u0027jiffies\u0027 [-Werror,-Wunsequenced]\n   1650 |                 jiffies - rdp-\u003enocb_bypass_first,\n        |                 ~~~~~~~\n   1651 |                 jiffies - rdp-\u003enocb_nobypass_last,\n        |                 ^\n\nPer C11 5.1.2.3.2, accessing a volatile object is a side effect. Per C11\n6.5.2, unsequenced side effects on the same scalar object are undefined\nbehavior. Per C11 6.5.2.2.10/12, the evaluation of arguments in a\nfunction call is indeterminately sequenced, meaning any side effects in\nthe arguments are unsequenced.\n\nAdd a local variable to hold the value of jiffies before the function\ncall so that there are no unsequenced side effects, clearing up the\nwarning.\n\nCc: stable@vger.kernel.org\nFixes: f7a81b12d6af (\"rcu/nocb: Print no-CBs diagnostics when rcutorture writer unduly delayed\")\nCloses: https://github.com/ClangBuiltLinux/linux/issues/2192\nLink: https://github.com/llvm/llvm-project/commit/955715460eaf9188c241c74750637fc82dfcc423 [1]\nSigned-off-by: Nathan Chancellor \u003cnathan@kernel.org\u003e\nReviewed-by: Joel Fernandes \u003cjfern@kernel.org\u003e\nSigned-off-by: Joel Fernandes \u003cjfern@kernel.org\u003e\n"
    },
    {
      "commit": "38828d44f31e4366b2963d5786d06d3358e8ae58",
      "tree": "ba9d52489ed2d94246fd7aa8b47d5af21ddac00f",
      "parents": [
        "aebb7274476403040d582c61478df8f4a7bf9cd3"
      ],
      "author": {
        "name": "Song Liu",
        "email": "song@kernel.org",
        "time": "Tue Nov 10 21:35:51 2026 -0800"
      },
      "committer": {
        "name": "Joel Fernandes",
        "email": "jfern@kernel.org",
        "time": "Sun Nov 15 18:51:16 2026 -0500"
      },
      "message": "rcu-tasks: Use pr_debug() for callback contend/collapse messages\n\nCommit 5de907d24919 (\"rcu-tasks: Disable callback contend/collapse\nmessages by default\") added the rcupdate.rcu_task_collapse_debug module\nparameter to gate the three pr_info() messages that report RCU Tasks\nswitching between CPU-0 and per-CPU callback queuing.\n\nThe kernel already has a mechanism for debug-only messages that are off\nby default: pr_debug().  Use pr_debug() for these messages, and remove\nthe rcu_task_collapse_debug module parameter, its documentation, and\nits boot-time announcement.\n\nThe messages stay off by default.  On CONFIG_DYNAMIC_DEBUG\u003dy kernels,\nthey can be enabled at boot with:\n\n  dyndbg\u003d\"file kernel/rcu/tasks.h +p\"\n\nor at runtime with:\n\n  echo \u0027file kernel/rcu/tasks.h +p\u0027 \u003e /proc/dynamic_debug/control\n\nAssisted-by: Claude:claude-opus-5\nSigned-off-by: Song Liu \u003csong@kernel.org\u003e\nReviewed-by: Joel Fernandes \u003cjfern@kernel.org\u003e\nReviewed-by: Paul E. McKenney \u003cpaulmck@kernel.org\u003e\nTested-by: Breno Leitao \u003cleitao@debian.org\u003e\nSigned-off-by: Joel Fernandes \u003cjfern@kernel.org\u003e\n"
    },
    {
      "commit": "aebb7274476403040d582c61478df8f4a7bf9cd3",
      "tree": "e5a64382f7926f5f6023e74c62216f1f9117bf84",
      "parents": [
        "a9daf327f3ff90625204876981f88cfe2b6f6fff"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 14:30:11 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 14:30:11 2026 -0800"
      },
      "message": "Linux 7.4-rc3\n"
    },
    {
      "commit": "a9daf327f3ff90625204876981f88cfe2b6f6fff",
      "tree": "eec61ac3be07a8e0553472704240f9a5b50112ab",
      "parents": [
        "4b15991b3be43e9581aca489ec30a351310bb49d",
        "700a332c3fe33014bbd5ae852d71ba679fc29e0a"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 14:20:40 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 14:20:40 2026 -0800"
      },
      "message": "Merge tag \u0027sched_ext-for-7.4-rc2-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tj/sched_ext\n\nPull sched_ext fixes from Tejun Heo:\n\n - With proxy execution or core scheduling, a CPU could run a task that\n   had been requeued on another CPU while dispatch had the rq lock\n   dropped.\n\n - Unregistering a scheduler could return while its teardown was still\n   running, letting the teardown call into BPF code that was already\n   freed.\n\n - A CPU coming online could call into the BPF scheduler before the\n   scheduler was told that the CPU is online.\n\n - A parent scheduler could be asked to attach a sub-scheduler to a\n   cgroup that was already removed.\n\n - Hardening: more kfuncs that take a queue lock are rejected in tracing\n   programs, where they could deadlock the CPU, and a queue iterator no\n   longer restarts after reaching the end, which the BPF verifier\n   assumes can\u0027t happen.\n\n - Smaller fixes: tasks needlessly sent back to the BPF scheduler on\n   idle CPUs, stale scheduling state carried into a task\u0027s next\n   scheduler, and uneven CPU shards with partially populated SMT cores.\n\n* tag \u0027sched_ext-for-7.4-rc2-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tj/sched_ext:\n  sched_ext: Reject DSQ iterator kfuncs in tracing programs\n  sched_ext: make DSQ iterator stay drained after NULL\n  sched_ext: Fix partial-SMT core counting in calc_shard_layout()\n  sched_ext: Reset task fields on exit of an SCX_TASK_READY task\n  sched_ext: Read exit_kind before flushing in scx_flush_disable_work()\n  sched_ext: Reject scx_bpf_destroy_dsq() in tracing programs\n  sched_ext: Fix ops.enqueue/dispatch() calls before ops.cpu_online()\n  sched_ext: Factor out drain_offline_rq()\n  sched_ext: Fix stale pick when the donor changes during dispatch\n  sched_ext: Call ops.sub_attach() only while the cgroup is online\n  sched_ext: Fix scx_bpf_dsq_reenq() kerneldoc claiming NMI calls are allowed\n  sched_ext: Avoid unnecessary IMMED reenqueues on idle CPUs\n"
    },
    {
      "commit": "4b15991b3be43e9581aca489ec30a351310bb49d",
      "tree": "f201b9e4d298e0058f99105e0a8c96787ff816e8",
      "parents": [
        "f70886e5a510cd7fc2c3aef91c46a14f4485fe63",
        "acec9637e03ca54bd3c3a0c1a6ad5c6dfa07c0df"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 14:17:31 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 14:17:31 2026 -0800"
      },
      "message": "Merge tag \u0027cgroup-for-7.4-rc2-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup\n\nPull cgroup fixes from Tejun Heo:\n\n - A file notification racing the file\u0027s removal could re-arm the\n   notification timer after it was shut down, corrupting the timer wheel\n   or firing on freed memory.\n\n - The v1 freezer could leave an exiting task frozen forever after its\n   cgroup was thawed or removed.\n\n - With memcg built without the block cgroup controller, memcg reclaim\n   waited on writeback and could deadlock against a filesystem freeze.\n   memcg now selects the controller.\n\n - cpuset partition state could go wrong when CPUs move between nested\n   partitions or go offline: isolated CPUs leaked or went missing,\n   invalid partitions kept exclusive CPUs or came back with no usable\n   CPU, and expanding a remote partition could fail. Separately, a\n   forked child could inherit a corrupted memory node mask.\n\n* tag \u0027cgroup-for-7.4-rc2-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup:\n  cgroup/cpuset: Don\u0027t revive invalid partition without active CPU on hotplug\n  cgroup/cpuset: Remove redundant WARN_ON_ONCE() in remote_cpus_update()\n  memcg: make MEMCG select BLK_CGROUP\n  cpuset: generalize the validate_partition() interface\n  cpuset: settle isolation state when CPUs are given back to parent\n  cgroup/cpuset: Correctly set isolated_cpus when switching partition state\n  cgroup/cpuset: Protect current-\u003emems_allowed read in cpuset_fork()\n  selftests/cgroup: Fix \"test_cpuset_prs.sh -v\" failures\n  cpuset: add early empty cpumask check in partition_xcpus_add/del\n  cgroup/cpuset: Reset partition data when an invalid partition stays invalid\n  cgroup/freezer: Fix exiting tasks left frozen forever after rmdir\n  cgroup/freezer: walk exiting tasks in the v1 freezer\n  cgroup: arm the notify timer only while the file exists\n  cgroup/cpuset: Do housekeeping check when isolated child with root parent may get more xcpus\n  cgroup/cpuset: Recompute effective cpumask in update_cpumasks_hier() after calling update_parent_effective_cpumask()\n  cgroup/cpuset: Fix incorrect CPU accounting with parent\u0027s effective_xcpus expansion\n"
    },
    {
      "commit": "f70886e5a510cd7fc2c3aef91c46a14f4485fe63",
      "tree": "ffc6034fb2d6fa15cabe1e8ee8a2b93684c583ff",
      "parents": [
        "87aecd24f58e15256870befed777bbffc5d4a9d5",
        "151d95a1e696979a7d54b23c96338754be2fcc1f"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 13:48:31 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 13:48:31 2026 -0800"
      },
      "message": "Merge tag \u0027nfs-for-7.4-2\u0027 of git://git.linux-nfs.org/projects/anna/linux-nfs\n\nPull NFS client fixes from Anna Schumaker:\n\n - Various NFS and SunRPC fixes from Avi Weiss\n\n - Various localio fixes from Mike Snitzer\n\n - Delegated timestamp fixes from Jeff Layton\n\n - CB_NOTIFY_LOCK fixes from Olga Kornievskaia\n\n - Various sunrpc, flexfiles, and xprtrdma fixes from Tim Menninger\n\n* tag \u0027nfs-for-7.4-2\u0027 of git://git.linux-nfs.org/projects/anna/linux-nfs: (26 commits)\n  NFSv4: Prevent indefinite waits for out-of-order OPEN stateids\n  NFSv4/flexfiles: limit transport disconnects when cancelling I/O\n  SUNRPC: add request-scoped disconnects when cancelling tasks\n  pNFS: avoid repeatedly cancelling I/O for the same lseg\n  xprtrdma: serialize unmap_sync with xprt_disconnect\n  NFSv4: remember that the server may send CB_NOTIFY_LOCK\n  NFSv4: wake blocking lock waiters when a delegated lock is released\n  NFS: stamp after the server\u0027s ctime following a CLONE\n  NFS: stamp after the server\u0027s ctime when truncating under delegated times\n  NFS: don\u0027t let delegated c/mtime stamps go backward\n  NFS/localio: give the LOCALIO probe a short timeout of its own\n  NFS/localio: probe again for LOCALIO when localio_enabled is set to Y\n  NFS/localio: hold the client across an asynchronous LOCALIO probe\n  SUNRPC: Balance queue accounting when relocating a task\n  SUNRPC: Balance switch queue length on transport removal\n  SUNRPC: Fix TCP-with-TLS transport class identification\n  SUNRPC: preserve transport security policy when cloning clients\n  SUNRPC: preserve timeout policy when cloning clients\n  SUNRPC: preserve fatal network errors during RPC binding\n  sunrpc: instantiate rpc_pipefs dentry before publishing it\n  ...\n"
    },
    {
      "commit": "87aecd24f58e15256870befed777bbffc5d4a9d5",
      "tree": "39308e80b4caa25d82d7dc5f35197a090bdd46cd",
      "parents": [
        "481f9afb337b7096aa177cde0c077dee0c649ae5",
        "e92cd41e8e5ee5616c6c7cd4378661d942f0a7c0"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 13:45:14 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 13:45:14 2026 -0800"
      },
      "message": "Merge tag \u0027for-7.4/dm-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm\n\nPull device mapper fixes from Mikulas Patocka:\n \"dm:\n   - fix infinite loop when removing a zoned device\n   - fix misbehavior if kobject_init_and_add fails\n   - fix refcount underflow if zone reset failed\n   - fix misbehavior if we attempt to load a table after previous table\n     load failed\n   - fix use-after-free if we open a device read-only and then re-open\n     it read-write\n   - fix error being ignored if flush failed\n\n  dm-pcache:\n   - fix misbehavior on flush failure\n\n  dm-raid:\n   - validate array_position after it is read from the superblock\n\n  dm-cache:\n   - fix a crash if dm_kcopyd_client_create fails\n   - fix divide by zero on extremely large device\n   - fix NULL pointer dereference on open failure\n   - fix misbehavior if dm_bitset_cursor_begin fails\n\n  dm-snapshot:\n   - properly advertise discard_granularity in the queue limits\n   - fix data corruption on unaligned discards\n\n  dm-crypt:\n   - fix a crash on invalid table line\n   - fix scheduling in atomic context when using no_write_workqueue and\n     asynchronous completion cipher\n   - fix misbehavior if percpu_counter_init fails\n\n  dm-thin:\n   - fix integer overflow if no_space_timeout is too high\n   - fix data corruption if discard races with regular I/O\n   - fix leak if __set_snapshot_details fails\n   - fix parsing of feature arguments\n\n  dm-mpath:\n   - fix buffer overflow when parsing the table string\n   - fix integer overflow in queue_if_no_path_timeout_secs\n\n  dm-log-userspace:\n   - fix open device leak on initialization error\n\n  dm-zoned:\n   - back off when reclaim cannot make progress\n   - do not reclaim a random zone into another random zone\n\n  dm-integrity:\n   - don\u0027t recalculate with HMAC because of a dirty bitmap\n   - don\u0027t attempt to commit journal on disabled device\n\n  dm-persistent-data:\n   - don\u0027t ignore the return value of node_copy\n   - fix buffer leak if find_leaf fails\n\n  dm-verity:\n   - don\u0027t restart or panic on errors during system shutdown\n\n  dm-raid:\n   - flush a work item in raid_dtr, fix use-after-free\"\n\n* tag \u0027for-7.4/dm-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm: (34 commits)\n  dm: fail a REQ_PREFLUSH bio with data when its preflush fails\n  dm-raid: fix use-after-free in raid_dtr()\n  dm-integrity: don\u0027t attempt to commit journal on disabled device\n  dm-verity: don\u0027t restart or panic on errors during system shutdown\n  dm crypt: fix use of uninitialized percpu_counter in ctr error path\n  dm: fix use-after-free when dm_get_device() widens the open mode\n  dm persistent data: don\u0027t leak bufio buffer if find_leaf fails\n  dm persistent data: don\u0027t ignore writeback errors\n  dm persistent data: test return value of node_copy\n  dm-thin: fix parsing of feature arguments\n  dm-thin: fix leak if __set_snapshot_details fails\n  dm-thin: fix data corruption if discard races with regular I/O\n  dm-cache: fix ending a cursor that never began\n  dm-cache: fix NULL pointer dereference on open failure\n  dm-crypt: defer async write submission to write thread\n  dm mpath: cap queue_if_no_path_timeout_secs\n  dm cache: validate the discard block size before narrowing\n  dm: Reject table setup after queue or disk teardown\n  dm: avoid double-dropping the zone reset submission reference\n  dm-sysfs: fix misbehavior if kobject_init_and_add fails\n  ...\n"
    },
    {
      "commit": "481f9afb337b7096aa177cde0c077dee0c649ae5",
      "tree": "65c79198fe6e95106aa8371ca7df43a79b3af229",
      "parents": [
        "e147c68fea1594901975ad0c9777a6e42077b39e",
        "2a490521d74ac76e5eb3c9a69d928a4c76dd3f78"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 13:41:14 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 13:41:14 2026 -0800"
      },
      "message": "Merge tag \u0027spi-fix-v7.4-rc2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi\n\nPull spi fixes from Mark Brown:\n \"There\u0027s a few small collection of driver fixes here, the usual story\n  where none of them are super remarkable if you don\u0027t have the have the\n  hardware. The largest batch is the Tegra fixes which are mainly around\n  error handling\"\n\n* tag \u0027spi-fix-v7.4-rc2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi:\n  spi: axi-spi-engine: Start the program length at zero\n  spi: virtio: move teardown to .remove()\n  spi: dt-bindings: pl022: Allow a power domain\n  spi: tegra210-quad: Unmap DMA buffers on transfer timeout\n  spi: tegra210-quad: Track DMA mapping state\n  spi: tegra210-quad: don\u0027t overwrite refreshed status with lost-IRQ snapshot\n"
    },
    {
      "commit": "e147c68fea1594901975ad0c9777a6e42077b39e",
      "tree": "7e52a1d6bafe349e31d9e0007189f237eafbd392",
      "parents": [
        "7893d906220694e2c8a13dfb52ec85b853cbbf7b",
        "fd89c52811a4c0754795199f86e0889bee7a70a3"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 13:31:55 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 13:31:55 2026 -0800"
      },
      "message": "Merge tag \u0027regulator-fix-v7.4-rc2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regulator\n\nPull regulator fixes from Mark Brown:\n \"A small change adding support for the Silergy SYM827 which is a\n  variation on the FAN53555 so needs some quirk data and a new DT\n  compatible\"\n\n* tag \u0027regulator-fix-v7.4-rc2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regulator:\n  regulator: fan53555: Add support for Silergy SYM827\n  regulator: dt-bindings: fan53555: Add Silergy SYM827\n"
    },
    {
      "commit": "7893d906220694e2c8a13dfb52ec85b853cbbf7b",
      "tree": "d474f287fe210e0223b22f9a5475bb3b1ba2e17a",
      "parents": [
        "fa2933aac170dd19c6a7f0d7b81a8f537e40915c",
        "f8623ce596492d2e96ca57eeec992938d3d79f00"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 13:30:32 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 13:30:32 2026 -0800"
      },
      "message": "Merge tag \u0027driver-core-7.4-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/driver-core/driver-core\n\nPull driver core fix from Danilo Krummrich:\n\n - Fix a use-after-free during device removal when a child device reuses\n   its parent\u0027s managed software node (affecting Xe)\n\n* tag \u0027driver-core-7.4-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/driver-core/driver-core:\n  software node: only let the owner release a managed node\n"
    },
    {
      "commit": "acec9637e03ca54bd3c3a0c1a6ad5c6dfa07c0df",
      "tree": "ed1bf085a62834fb05c34dc61f5acfdedca8e7f0",
      "parents": [
        "42ef815ff6b282e75248e8e084508fca4a694b70"
      ],
      "author": {
        "name": "Ridong Chen",
        "email": "chenridong@xiaomi.com",
        "time": "Sat Nov 14 20:05:32 2026 +0800"
      },
      "committer": {
        "name": "Tejun Heo",
        "email": "tj@kernel.org",
        "time": "Sun Nov 15 10:15:12 2026 -1000"
      },
      "message": "cgroup/cpuset: Don\u0027t revive invalid partition without active CPU on hotplug\n\nWhen a CPU hotplug leaves no CPU outside of the partitions for the top\ncpuset, cpuset_handle_hotplug() clears subpartitions_cpus and gives all\nthe active CPUs back to the top cpuset. cpuset_hotplug_update_tasks()\nthen walks the hierarchy and invalidates the partitions, as\nsubpartitions_cpus is empty.\n\nHowever, an invalid partition visited in the same walk may be revived\nif its effective_xcpus is not empty. tasks_nocpu_error() doesn\u0027t catch\nit when its CPUs are all offline but it has no task. Its CPUs are put\nback into subpartitions_cpus, and the valid partitions visited after it\nare no longer invalidated, as subpartitions_cpus is not empty any more.\nThe top cpuset is left with CPUs overlapping with these partitions.\n\nIt can be reproduced on a system with 8 CPUs:\n\n  # cd /sys/fs/cgroup\n  # echo +cpuset \u003e cgroup.subtree_control\n  # mkdir A1 B1\n  # echo 1 \u003e A1/cpuset.cpus\n  # echo 1 \u003e A1/cpuset.cpus.exclusive\n  # echo 0,2-7 \u003e B1/cpuset.cpus\n  # echo root \u003e B1/cpuset.cpus.partition\n  # echo root \u003e A1/cpuset.cpus.partition\n  # cat A1/cpuset.cpus.partition\n  root invalid (Parent unable to distribute cpu downstream)\n  # echo 0 \u003e /sys/devices/system/cpu/cpu1/online\n  # cat A1/cpuset.cpus.partition B1/cpuset.cpus.partition\n  root\n  root\n  # cat cpuset.cpus.effective\n  0,2-7\n\nA1 can\u0027t take CPU 1, the last CPU of the root cgroup which has tasks.\nWhen CPU 1 is offlined, A1 is revived without any active CPU, and B1\nstays valid while its CPUs are also in the effective CPUs of the root\ncgroup, which triggers\n\n  WARNING: kernel/cgroup/cpuset.c:919 at rebuild_sched_domains_locked\n\nOnly revive an invalid partition in hotplug if its effective_xcpus has\nan active CPU, the same as what is required to create a partition with\ntasks. With that, A1 and B1 are both invalidated in the above case.\n\nFixes: 0c7f293efc87 (\"cgroup/cpuset: Add cpuset.cpus.exclusive.effective for v2\")\nCc: stable@vger.kernel.org # v6.7+\nReported-by: Sashiko \u003csashiko-bot@kernel.org\u003e\nLink: https://sashiko.dev/#/patchset/20261114083124.1340140-1-ridong.chen%40linux.dev?part\u003d11\nSigned-off-by: Ridong Chen \u003cchenridong@xiaomi.com\u003e\nReviewed-by: Waiman Long \u003clongman@redhat.com\u003e\nSigned-off-by: Tejun Heo \u003ctj@kernel.org\u003e\n"
    },
    {
      "commit": "42ef815ff6b282e75248e8e084508fca4a694b70",
      "tree": "13d10705080df3573b69dd16c57e36f7cace234e",
      "parents": [
        "7693e1350b06191143724da742559fb47211348e"
      ],
      "author": {
        "name": "Ridong Chen",
        "email": "chenridong@xiaomi.com",
        "time": "Sat Nov 14 18:06:12 2026 +0800"
      },
      "committer": {
        "name": "Tejun Heo",
        "email": "tj@kernel.org",
        "time": "Sun Nov 15 10:15:02 2026 -1000"
      },
      "message": "cgroup/cpuset: Remove redundant WARN_ON_ONCE() in remote_cpus_update()\n\nremote_cpus_update() warns if the CPUs to be added to a remote partition\nintersect subpartitions_cpus, and then checks the very same condition\nright below to set PERR_NOCPUS and invalidate the partition:\n\n\tWARN_ON_ONCE(cpumask_intersects(tmp-\u003eaddmask, subpartitions_cpus));\n\tif (!capable(CAP_SYS_ADMIN))\n\t\tWRITE_ONCE(cs-\u003eprs_err, PERR_ACCESS);\n\telse if (cpumask_intersects(tmp-\u003eaddmask, subpartitions_cpus) ||\n\t\t cpumask_subset(top_cpuset.effective_cpus, tmp-\u003eaddmask))\n\t\tWRITE_ONCE(cs-\u003eprs_err, PERR_NOCPUS);\n\nSashiko reported that the warning could be triggered by a write to a\ncpuset control file, before the CAP_SYS_ADMIN check, which would panic a\nsystem with panic_on_warn set.\n\nThe warning was added by commit 86888c7bd117 (\"cgroup/cpuset: Add\nwarnings to catch inconsistency in exclusive CPUs\") as an assertion that\nsuch CPUs have been filtered out before. No way to trigger it with the\ncurrent interfaces has been found, as these CPUs are rejected when\ncpuset.cpus.exclusive is written. But the warning is redundant, as the\nconflict is handled right below anyway.\n\nRemove the warning and rely on the PERR_NOCPUS check, the same as\nremote_partition_enable() already does for this conflict.\n\nReported-by: Sashiko \u003csashiko-bot@kernel.org\u003e\nLink: https://sashiko.dev/#/patchset/20261114083124.1340140-1-ridong.chen%40linux.dev?part\u003d6\nSigned-off-by: Ridong Chen \u003cchenridong@xiaomi.com\u003e\nReviewed-by: Waiman Long \u003clongman@redhat.com\u003e\nSigned-off-by: Tejun Heo \u003ctj@kernel.org\u003e\n"
    },
    {
      "commit": "fa2933aac170dd19c6a7f0d7b81a8f537e40915c",
      "tree": "4c1542e3a0fc0e265383336a564a54ac18ca5751",
      "parents": [
        "0feaf5349e0a0521de3406343df32954801784d9",
        "0dc2a4dac3d04e9e4479aeb0b46172feab1cca0a"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 09:18:05 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 09:18:05 2026 -0800"
      },
      "message": "Merge tag \u0027x86-urgent-2026-11-15\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull x86 fixes from Ingo Molnar:\n\n - Don\u0027t print reset reason in VMs, to fix Xen PV domU boot\n   regression on AMD Zen CPUs (Jason Andryuk)\n\n - Preserve the default VGA fallback for legacy modes, to fix\n   regression where a SeaBIOS QEMU guest with bochs and VKMS\n   the boot VGA (bochs) loses its DRM boot_display attribute\n   (Hengyu Liang)\n\n - Ensure write visibility during LDT installation (Dave Hansen)\n\n* tag \u0027x86-urgent-2026-11-15\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n  x86/mm: Ensure write visibility during LDT installation\n  x86/video: Preserve the default VGA fallback for legacy modes\n  x86/CPU/AMD: Don\u0027t print reset reason in VMs\n"
    },
    {
      "commit": "0feaf5349e0a0521de3406343df32954801784d9",
      "tree": "f292e3d410f8abeaac28a12d8d19ad5c2e6272a5",
      "parents": [
        "fc28723faec6c5eda0263b46e23dd89d5c29f464",
        "c1b96ba87e3caaa701d6dba254e58db7cc3c99be"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 09:09:04 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 09:09:04 2026 -0800"
      },
      "message": "Merge tag \u0027timers-urgent-2026-11-15\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull timer fixes from Ingo Molnar::\n\n - Add VDSO_UNALIGNED_PACKED config option for targets without efficient\n   unaligned memory accesses (such as ARMv5), and turn it on for\n   !CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS platforms.\n\n   Propagate these changes into tooling headers as well (Ian Rogers)\n\n - Read getboottime64() offsets under the tk_core.seq sequence counter\n   (Hungyu Lin)\n\n - Keep tsacct task begin times stable across suspend (Hungyu Lin)\n\n - Reject cross timestamps for invalid AUX clocks (Avi Weiss)\n\n - Refresh auxiliary timekeeper base before offset validation (Avi\n   Weiss)\n\n* tag \u0027timers-urgent-2026-11-15\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n  timekeeping: Refresh auxiliary base before offset validation\n  timekeeping: Reject cross timestamps for invalid AUX clocks\n  tsacct: Keep task begin times stable across suspend\n  timekeeping: Read getboottime64() offsets under the sequence counter\n  tools headers: Update the vdso/unaligned.h copy with the kernel sources\n  vdso: Add VDSO_UNALIGNED_PACKED config option\n"
    },
    {
      "commit": "fc28723faec6c5eda0263b46e23dd89d5c29f464",
      "tree": "acf3cb8974eddd14f97034e792497a9de584da67",
      "parents": [
        "ca6709295ae5420d1a391f32e6a7f23b7fda14b3",
        "ec6f6731538e0d45872eb279fdeb477bed152d6f"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 09:04:14 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 09:04:14 2026 -0800"
      },
      "message": "Merge tag \u0027sched-urgent-2026-11-15\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull scheduler fixes from Ingo Molnar:\n\n - Fix spurious SYS_sched_setattr() -EPERM for migration-disabled\n   tasks (Andrea Parri)\n\n - Reject clamp values above SCHED_CAPACITY_SCALE (Andrea Parri)\n\n - Fix cpuset_mutex double unlock in __sched_setscheduler()\n   (Andrea Parri)\n\n - Don\u0027t use fits_capacity() to compare thread and CPU counts in\n   in invalid_llc_nr() (Zhan Xusheng)\n\n - Do not touch SMT siblings that have not joined the core yet\n   during CPU onlining (Seiji Nishikawa)\n\n* tag \u0027sched-urgent-2026-11-15\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n  sched/core: Do not touch SMT siblings that have not joined the core yet\n  sched/cache: Don\u0027t use fits_capacity() to compare thread and CPU counts\n  sched/core: Fix cpuset_mutex double unlock in __sched_setscheduler()\n  sched/uclamp: Reject clamp values above SCHED_CAPACITY_SCALE\n  sched/deadline: Fix spurious -EPERM for migration-disabled tasks\n"
    },
    {
      "commit": "ca6709295ae5420d1a391f32e6a7f23b7fda14b3",
      "tree": "2690ef98bcf97764e6a4ed0f3349e5b03e9e00e6",
      "parents": [
        "d56cd25dc0dc3fdb79e1f1df8dc366dabff8ec54",
        "59f1398140869432c7171b9a2a279ca315467c4e"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 09:01:48 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 09:01:48 2026 -0800"
      },
      "message": "Merge tag \u0027perf-urgent-2026-11-15\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull perf events fixes from Ingo Molnar:\n\n - Avoid AMD uncore context migration from non-owners (Sandipan Das)\n\n - Clear stop state in perf_trace_add(), to fix self-profilers such as\n   async-profiler. Also fix a similar bug in hw_breakpoint_add() (Paolo\n   Bertinetti)\n\n - Fix racy update of cpc-\u003ehrtimer_interval (Benjamin Segall)\n\n - Do not resurrect cpuctx-\u003ecgrp after last cgroup event exit, to fix\n   UAF bug (Phat Dinh)\n\n* tag \u0027perf-urgent-2026-11-15\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n  perf/core: Do not resurrect cpuctx-\u003ecgrp after last cgroup event exit\n  perf: Fix racy update of cpc-\u003ehrtimer_interval\n  perf/hw_breakpoint: Clear stop state in hw_breakpoint_add()\n  tracing/perf: Clear stop state in perf_trace_add()\n  perf/x86/amd/uncore: Avoid context migration from non-owners\n"
    },
    {
      "commit": "d56cd25dc0dc3fdb79e1f1df8dc366dabff8ec54",
      "tree": "8193905210d6d7a43d801916675d453d5a69c282",
      "parents": [
        "04da1a14b9eae1dbc2ccc007b54f46771f9cb529",
        "efe191f1605a149b06eb0fe1ff975e7d3137c9bb"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 08:59:45 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 08:59:45 2026 -0800"
      },
      "message": "Merge tag \u0027core-urgent-2026-11-15\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull core syscall fixes from Ingo Molnar:\n\n - Two generic syscall entry code fixes for return code\n   data type definitions (Florian Bezdeka)\n\n* tag \u0027core-urgent-2026-11-15\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n  entry: Complete syscall_enter_from_user_mode_work() to bool migration\n  entry: Fix return type of syscall_trace_enter()\n"
    },
    {
      "commit": "04da1a14b9eae1dbc2ccc007b54f46771f9cb529",
      "tree": "66bc298c4c6ec73b40d71fb11418538bc239b518",
      "parents": [
        "bfac855a820c3a32219779e4fb44804e8f81977d",
        "58c5852ca96c58cdfdb9646870b217fc54d06a97"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 08:55:09 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 08:55:09 2026 -0800"
      },
      "message": "Merge tag \u0027fixes-2026-11-15\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/liveupdate/linux\n\nPull liveupdate fixes from Mike Rapoport:\n\n - avoid address translation in is_kho_boot() before the linear map is\n   initialized, fixing early boot BUG_ONs on arm64\n\n - check for duplicate session names before growing the sessions\n   preservations block set to avoid preserving an empty block that\n   causes restoration to panic()\n\n* tag \u0027fixes-2026-11-15\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/liveupdate/linux:\n  kho: check fdt_phys in is_kho_boot()\n  liveupdate: check for duplicate session before growing block set\n"
    },
    {
      "commit": "bfac855a820c3a32219779e4fb44804e8f81977d",
      "tree": "8fb5f211384f0bf5df27cae545f38acc4b236002",
      "parents": [
        "97f4f3805fd22bfc9cbdede5a2cfb0c48ee7df1e",
        "ab75132e99eb0b7dcd3a1882e35519f8229a669e"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 08:54:17 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 08:54:17 2026 -0800"
      },
      "message": "Merge tag \u0027kexec-fixes-2026-11-15\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/liveupdate/linux\n\nPull kexec MAINTAINERS update from Mike Rapoport:\n \"Remove Dave Young as kdump reviewer\"\n\n* tag \u0027kexec-fixes-2026-11-15\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/liveupdate/linux:\n  Remove myself as a kdump reviewer\n"
    },
    {
      "commit": "97f4f3805fd22bfc9cbdede5a2cfb0c48ee7df1e",
      "tree": "553ec0cc06bb06686680af2727c076ef0bf2a078",
      "parents": [
        "442fbdb90e120c39bc0b27ffdcc6d7733472ddb9",
        "c15120853eebbda98f3fe1f1e59370439c6c4447"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 08:34:09 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 08:34:09 2026 -0800"
      },
      "message": "Merge tag \u0027probes-fixes-v7.4-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace\n\nPull probe fixes from Masami Hiramatsu:\n \"Fix ftrace ops switching and optprobe state transitions:\n\n   - Switch an armed ftrace-based kprobe from kprobe_ftrace_ops to\n     kprobe_ipmodify_ops in register_aggr_kprobe() when adding a probe\n     with a post_handler, so later disarming does not fail with -ENOENT\n     and leave the probe armed\n\n   - Queue forcibly unoptimized probes onto the generation\u0027s\n     cooling_list in force_unoptimize_kprobe() so unregister_kprobes()\n     waits for Tasks RCU before freeing the optinsn slot, preventing a\n     use-after-free\n\n   - Skip arch_arm_kprobe() in __arm_kprobe() when a probe is queued for\n     unoptimization so the active relative jump instruction is not\n     partially overwritten with a breakpoint before the optimizer runs\n\n   - Re-queue any jump-optimized sibling probe covering a disabled\n     probe\u0027s address when that probe finishes cooling in\n     optprobe_finalize_generation(), matching the freeing_list path\n\n  Fix multiprobe lifecycle and boot kprobe setup\n\n   - Return early in setup_boot_kprobe_events() when the kprobe_event\u003d\n     boot buffer is empty to avoid contending on event_mutex with\n     asynchronous eval-map work during normal boot\n\n   - Verify that a multiprobe event is neither enabled nor busy before\n     unlinking any sibling kprobe, fprobe, uprobe, or eprobe so active\n     probes cannot be unlinked and freed via dynamic_events\n\n   - Register and enable a newly appended fprobe or uprobe immediately\n     when the target multiprobe event is already enabled so the probe\n     fires and disable or perf cleanup does not act on an unregistered\n     probe\n\n   - Free the temporary event name string with __free(kfree) in\n     __trace_uprobe_create() to prevent a memory leak if a subsequent\n     allocation fails\n\n  Fix BTF parsing and string array fetch handling:\n\n   - Compute BTF bitfield byte and bit offsets consistently, select the\n     smallest unsigned fetch type covering the bitfield, and size\n     FETCH_OP_MOD_BF from that type to prevent an 8-byte out-of-bounds\n     write on smaller ring buffer slots\n\n   - Validate BTF_KIND_PTR and non-NULL BTF_KIND_INT types in\n     btf_type_is_char_ptr() and btf_type_is_char_array() before reading\n     inner type metadata to avoid misinterpreting byte sizes as type IDs\n     or reading out of bounds\n\n   - Check for FETCH_OP_UDEREF before advancing the fetch instruction\n     slot in finalize_fetch_insn() so user-space dereference string\n     arrays emit FETCH_OP_ST_USTRING instead of reading kernel memory\n\n   - Reject string array counts when the preceding fetch operation is\n     not a memory dereference so non-dereference string sources ($comm,\n     immediate strings, eprobe fields) do not emit invalid\n     FETCH_OP_LP_ARRAY loops\n\n  Fix signed 1-byte field fetch and fprobe target check:\n\n   - Use explicit fixed-width integer types (s8/u8, s16/u16, s32/u32) in\n     get_event_field() so 1-byte signed event fields are sign-extended\n     instead of zero-extended under -funsigned-char\n\n   - Exclude TRACE_EVENT_FL_FPROBE events in find_and_get_event() so\n     eprobes cannot attach to fprobe or tprobe events\n\n  Fix typo in kprobes documentation:\n\n   - Remove a stray closing parenthesis in Documentation/trace/kprobes.rst\"\n\n* tag \u0027probes-fixes-v7.4-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace:\n  tracing/uprobes: Fix memory leak of tail in __trace_uprobe_create()\n  tracing/probes: Reject array count on non-dereference string sources\n  tracing/probes: Fix user-space dereference string array store type\n  tracing/eprobe: Disallow attaching eprobes to fprobe events\n  tracing/eprobe: Fix 1-byte signed field fetch in get_event_field()\n  tracing/probes: Fix missing BTF kind checks in char pointer/array helpers\n  tracing/probes: Fix BTF bitfield parsing and out-of-bounds buffer write\n  tracing/probes: Enable newly appended probe when target event is enabled\n  tracing/probes: Check enabled and busy state before unlinking sibling probes\n  kprobes: Re-optimize blocked sibling probe when cooling finishes\n  kprobes: Do not overwrite jump instruction when re-arming unoptimizing probe\n  kprobes: Fix UAF and queue cooling probes on forced unoptimization\n  kprobes: Switch ftrace ops when adding probe with post_handler\n  docs, kprobes: Fix bracket\n  tracing: skip boot kprobe setup when no events were requested\n"
    },
    {
      "commit": "442fbdb90e120c39bc0b27ffdcc6d7733472ddb9",
      "tree": "32e17131059b549a7cc28f49aa7ff3bf6daae19e",
      "parents": [
        "23c62304dde29abe92013d85910d03a77dada35d",
        "381b391bfaa3f10d52da4740529804b0b6b8160d"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 08:28:08 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sun Nov 15 08:28:08 2026 -0800"
      },
      "message": "Merge tag \u0027bootconfig-fixes-v7.4-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace\n\nPull bootconfig fixes from Masami Hiramatsu:\n\n - bootconfig: Allow keys with the maximum supported depth\n\n   Check the array capacity before storing the next key component in\n   xbc_node_compose_key_after() so keys at XBC_DEPTH_MAX (16) are\n   composed instead of failing with -ERANGE when listing keys or\n   creating /proc/bootconfig.\n\n - bootconfig: Trim unquoted values at end of file\n\n   Trim trailing whitespace from unquoted values in __xbc_parse_value()\n   after scanning finishes so values at end of input without a trailing\n   newline do not retain extra whitespace.\n\n - tools/bootconfig: Close only opened braces in compact output\n\n   Check key children while skipping value nodes when backtracking in\n   xbc_show_compact_tree() so a brace is not closed when a single subkey\n   under a valued key was printed in dotted notation.\n\n - tools/bootconfig: Test listing keys at the maximum depth\n\n   Add regression tests verifying that a 16-component key can be\n   attached and listed in full while a 17-component key is rejected.\n\n* tag \u0027bootconfig-fixes-v7.4-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace:\n  bootconfig: Trim unquoted values at end of file\n  tools/bootconfig: Close only opened braces in compact output\n  tools/bootconfig: Test listing keys at the maximum depth\n  bootconfig: Allow keys with the maximum supported depth\n"
    },
    {
      "commit": "7693e1350b06191143724da742559fb47211348e",
      "tree": "5bb9fa17b98e11dbe055d9abeb8e1c46a4ef315c",
      "parents": [
        "e234ed1bb82c96a5643fd70784f16cf3d72eeed3"
      ],
      "author": {
        "name": "Christian Brauner",
        "email": "brauner@kernel.org",
        "time": "Sat Nov 14 22:40:28 2026 +0100"
      },
      "committer": {
        "name": "Tejun Heo",
        "email": "tj@kernel.org",
        "time": "Sun Nov 15 05:34:45 2026 -1000"
      },
      "message": "memcg: make MEMCG select BLK_CGROUP\n\nMemcg reclaim only relies on writeback throttling if\nwriteback_throttling_sane() does. Since commit 97c9341f7271 (\"mm:\nvmscan: disable memcg direct reclaim stalling if cgroup writeback\nsupport is in use\") that\u0027s the case if the memory controller is on the\ndefault hierarchy and CONFIG_CGROUP_WRITEBACK is set. Otherwise\nshrink_folio_list() sleeps in folio_wait_writeback() for a folio that\nis under writeback and already marked for reclaim if the allocation may\nenter the filesystem.\n\nCONFIG_CGROUP_WRITEBACK is enabled whenever CONFIG_MEMCG and\nCONFIG_BLK_CGROUP are. So with CONFIG_MEMCG\u003dy and CONFIG_BLK_CGROUP\u003dn\nmemcg reclaim on cgroup v2 waits for writeback just like it does on\ncgroup v1.\n\nThat wait can deadlock against freezing a filesystem. Loop takes freeze\nprotection on the backing filesystem for its writes. A task in the\nmiddle of a write() to the backing filesystem holds freeze protection\nas well.\n\nIf that task enters memcg reclaim it can end up waiting for writeback of\na folio that belongs to a filesystem on the loop device. If a freeze of\nthe backing filesystem is pending the loop worker blocks when it tries\nto write that folio out. And the freeze waits for the task in reclaim to\ndrop its freeze protection. The same applies to other drivers that serve\nblock I/O by writing to a file on a local filesystem and take freeze\nprotection for it, e.g., the scsi target\u0027s fileio backend.\n\nMake CONFIG_MEMCG select CONFIG_BLK_CGROUP if CONFIG_BLOCK is set. Then\nCONFIG_CGROUP_WRITEBACK is always enabled together with memcg and memcg\nreclaim on cgroup v2 doesn\u0027t sleep in folio_wait_writeback() anymore.\na\nWithout CONFIG_BLOCK there are no block devices that could deadlock\nlike this.\n\nSigned-off-by: Christian Brauner (Amutable) \u003cbrauner@kernel.org\u003e\nAcked-by: Shakeel Butt \u003cshakeel.butt@linux.dev\u003e\nSigned-off-by: Tejun Heo \u003ctj@kernel.org\u003e\n"
    },
    {
      "commit": "0dc2a4dac3d04e9e4479aeb0b46172feab1cca0a",
      "tree": "e5052544214683cc67874802544b37e2844a2605",
      "parents": [
        "3b4c98b47ba7c3c57bcf54fb85f54a854622041c"
      ],
      "author": {
        "name": "Dave Hansen",
        "email": "dave.hansen@linux.intel.com",
        "time": "Tue Nov 10 09:27:39 2026 -0800"
      },
      "committer": {
        "name": "Ingo Molnar",
        "email": "mingo@kernel.org",
        "time": "Sun Nov 15 12:53:02 2026 +0100"
      },
      "message": "x86/mm: Ensure write visibility during LDT installation\n\nDuring LDT installation, the intent is to overwrite the old per-mm\nLDT value, then send an IPI to the mm user CPUs to have each load\nthe new value.\n\nBut there is an ordering problem: a new mm user CPU sets itself in\nmm_cpumask() and then reads mm-\u003econtext.ldt. The LDT-installing CPU\nwrites mm-\u003econtext.ldt and then reads mm_cpumask(). But, there is\nnothing which orders the LDT installation side actions. The CPU\ncan read mm_cpumask() before its write becomes visible.\n\nEnsure that new mm user CPUs see the new mm-\u003econtext.ldt. Make\nmm-\u003econtext.ldt visible before reading mm_cpumask().\n\nAlso convert the smp_store_release() over to a WRITE_ONCE().\nsmp_store_release() has an internal barrier which is weaker than\nthe new smp_mb(). Avoid an extra barrier by converting it to\nWRITE_ONCE().\n\nAlso refresh/correct misleading comments. READ/WRITE_ONCE() do not\n\"synchronize\" anything themselves.\n\nReported-by: Jaeyoung Chung \u003cjjy600901@snu.ac.kr\u003e\nSigned-off-by: Dave Hansen \u003cdave.hansen@linux.intel.com\u003e\nSigned-off-by: Ingo Molnar \u003cmingo@kernel.org\u003e\nTested-by: Jaeyoung Chung \u003cjjy600901@snu.ac.kr\u003e\nAcked-by: Linus Torvalds \u003ctorvalds@linuxfoundation.org\u003e\nAcked-by: Peter Zijlstra (Intel) \u003cpeterz@infradead.org\u003e\nCc: \u003cstable@vger.kernel.org\u003e\n"
    },
    {
      "commit": "23c62304dde29abe92013d85910d03a77dada35d",
      "tree": "78c11a69c8b8226b0cefc40082c0cdc8030b2f8b",
      "parents": [
        "54975c6523dc77862c25deeb5fa395990b8d3fc4",
        "2b66f2cc753243a6adb3e9a5c038b0fc59ba7ec9"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sat Nov 14 10:59:13 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sat Nov 14 10:59:13 2026 -0800"
      },
      "message": "Merge tag \u0027for-linus-7.4a-rc3-tag\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip\n\nPull xen fixes from Juergen Gross:\n \"A bunch of isolated fixes of Xen related drivers\"\n\n* tag \u0027for-linus-7.4a-rc3-tag\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip:\n  Xen/gnttab: improve error handling in gnttab_dma_{alloc,free}_pages()\n  xen: don\u0027t allow gnttab_dma_[alloc|free]_pages() in PV guest\n  xen/pvcalls: destroy the per-socket workqueue on release\n  xen/pvcalls: unwind pvcalls_new_active_socket() step by step\n  xen/privcmd: fix double free when an irqfd is shut down twice\n  xen/gntdev: fix deadlock between map ioctl and mmap\n  xen/platform-pci: Disable device before releasing PCI regions\n  xen/grant-table: annotate on-stack work in gnttab_unmap_refs_sync()\n"
    },
    {
      "commit": "54975c6523dc77862c25deeb5fa395990b8d3fc4",
      "tree": "f05ea1e31d8da28607ee3d0e0e191b95fd37270c",
      "parents": [
        "aacb7797504183a90d66b19b1efb555cfe9b3eec",
        "b5c5b67e02406fb48acf57cf2c30a570230fd6aa"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sat Nov 14 10:57:26 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sat Nov 14 10:57:26 2026 -0800"
      },
      "message": "Merge tag \u0027hwmon-for-v7.4-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging\n\nPull hwmon fixes from Guenter Roeck:\n\n - Read tracking-mode voltage fault limits dynamically in the mp2975\n   PMBus driver instead of caching probe-time snapshots, avoiding false\n   over-voltage alarms when CPU voltage changes at runtime.\n\n - Fix SCMI sensor probe by enabling supported sensors via firmware IDs\n   before exposing channels and checking channel config allocations.\n\n - Fix runtime PM enable and reference count imbalances during probe and\n   teardown in the gpio-fan driver.\n\n - Fix max31760 fan controller initialization and reporting, including\n   leaving standby at probe, initial 0 RPM reporting, and swapped\n   temperature limit registers.\n\n - Assorted minor sensor reading, interval clamping, and register\n   handling fixes in other drivers.\n\n* tag \u0027hwmon-for-v7.4-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging:\n  hwmon: (aht10) Fix update_interval overflow on 64-bit\n  hwmon: (scmi) Configure sensors before exposing channels\n  hwmon: (scmi) Check channel config allocation\n  hwmon: (nct6775) Handle dual use registers\n  hwmon: (scmi) Enable temperature sensors using their sensor IDs\n  hwmon: (pmbus/mp2975) Get vout fault limits dynamically\n  hwmon: (sy7636a) fix negative temperatures\n  hwmon: (max31760) Report 0 RPM before the first fan speed measurement\n  hwmon: (max31760) Leave standby mode at probe\n  hwmon: (max31760) Fix the local temperature limit registers\n  hwmon: (gpio-fan) Fix runtime PM imbalance in probe and teardown\n"
    },
    {
      "commit": "aacb7797504183a90d66b19b1efb555cfe9b3eec",
      "tree": "eb3ca886236bdeb1e33f7f2b2ee793512873949d",
      "parents": [
        "68271fbe7d14b6397e90df730c3967d3a481902b",
        "748739f3171ed1c44a30a98b0e0341b1a8d62d31"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sat Nov 14 08:32:19 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sat Nov 14 08:32:19 2026 -0800"
      },
      "message": "Merge tag \u0027ovl-fixes-7.4-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/overlayfs/vfs\n\nPull overlayfs fixes from Amir Goldstein:\n\n - Fixes for data-only layers (composefs), some for regressions from\n   v6.16, which relaxed the requirements for redirects to data-only\n   layers\n\n - Fix a use-after-free in lookup of a hardlink with a FUSE upper,\n   reported by Sashiko\n\n - Fix a mount failure right after unmount of an overlay with the same\n   upperdir\n\n - Reject lower layers on the same fs with different idmaps, which could\n   decode file handles through the wrong layer\n\n - Minor fixes for a memory leak, an error message and readdir d_ino\n\n* tag \u0027ovl-fixes-7.4-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/overlayfs/vfs:\n  selftests: overlayfs: reject different lower idmaps on same fs\n  ovl: reject different lower idmaps on the same fs\n  ovl: bump the impure dir version when linking a file with origin\n  ovl: return EIO if the lowerdata is missing\n  ovl: follow lower to lowerdata redirects from any lower layer\n  ovl: report the requested redirect_dir mode in conflicting options\n  ovl: do not allow nfs_export with data-only layers\n  ovl: do not check the deferred lowerdata redirect on index lookup\n  ovl: fix leak of lowerdata redirect on ovl_get_inode() failure\n  ovl: do not always clear absolute_redirect\n  ovl: release the upperdir/workdir in-use locks before the umount sync\n  ovl: do not read the flags of a put upper dentry in lookup\n"
    },
    {
      "commit": "68271fbe7d14b6397e90df730c3967d3a481902b",
      "tree": "dcf01ef60e0df7fbeb50dfad0b6d2cc306865028",
      "parents": [
        "cba88960471ac67635080708b67ed5c2a2e4e9fd",
        "bada7bbb11aaa6feba8904cf38ff833453f5b5d9"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sat Nov 14 08:27:30 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sat Nov 14 08:27:30 2026 -0800"
      },
      "message": "Merge tag \u0027ntfs-for-7.4-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/linkinjeon/ntfs\n\nPull ntfs fixes from Namjae Jeon:\n\n - Validate that $INDEX_ALLOCATION is open and non-resident before\n   reading or writing index blocks, preventing NULL pointer dereferences\n   when an earlier index operation has not opened it\n\n - Check that index allocation inodes are non-resident and child VCNs\n   are non-negative and within bounds before directory lookup reads,\n   preventing a deadlock when $INDEX_ALLOCATION is missing or damaged\n\n - Validate file names against the index entry\u0027s key length, preventing\n   out-of-bounds reads when directory enumeration resumes using a copied\n   key\n\n - Return -ENOENT for NTFS_IOC_LIST_STREAMS on deleted base files and\n   fix unsigned underflow in its buffer bounds check, preventing\n   out-of-bounds writes when the sizing and filling passes disagree\n\n - Use kfree() for the temporary runlist allocated with kmalloc() on\n   merge failure, resolving the Coccinelle allocation/free mismatch\n   warning\n\n* tag \u0027ntfs-for-7.4-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/linkinjeon/ntfs:\n  ntfs: open the index allocation before reading or writing an index block\n  ntfs: do not read an index block outside $INDEX_ALLOCATION\n  ntfs: silence kmalloc/kvfree mismatch warning\n  ntfs: check the file name against the key length of an index entry\n  ntfs: fix out-of-bounds write race in NTFS_IOC_LIST_STREAMS\n"
    },
    {
      "commit": "cba88960471ac67635080708b67ed5c2a2e4e9fd",
      "tree": "f07a56d009c171eb4bc9d14927496f46448d8c79",
      "parents": [
        "860b3016e6d6fd6ae83eb8f05d1861a1876db060",
        "b0af9a3e6b1534d353ca6ac23caaf46127f62e4a"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sat Nov 14 08:23:39 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Sat Nov 14 08:23:39 2026 -0800"
      },
      "message": "Merge tag \u0027ksmbd-for-7.4-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/linkinjeon/smb\n\nPull smb server fixes from Namjae Jeon:\n\n - Evaluate explicit-access requests in DACL order with exact SID\n   matching, and prevent guest and anonymous sessions from matching\n   Authenticated Users\n\n - Apply Unix group and OWNER RIGHTS ACEs, and keep their denials when\n   POSIX ACL fallback determines access\n\n - Allow the parent directory’s DACL to grant explicitly requested\n   DELETE and FILE_READ_ATTRIBUTES through FILE_DELETE_CHILD and\n   FILE_LIST_DIRECTORY\n\n - Exclude metadata-only file and stream opens from sharing conflicts\n   and stream delete-sharing checks, while accounting for writes implied\n   by overwrites and named-stream creation\n\n - Return STATUS_ACCESS_DENIED for reads on files lacking FMODE_READ,\n   preventing warnings in __kernel_read() and potential panics\n\n - Pin parent dentries during SMB2 CREATE permission, status, and ACL\n   inheritance operations to prevent use-after-free races with\n   concurrent rename and removal\n\n - Keep expired sessions bound to their original user during NTLM\n   reauthentication, preventing identity changes and a user-object leak\n\n* tag \u0027ksmbd-for-7.4-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/linkinjeon/smb:\n  ksmbd: keep group denials when a POSIX ACL entry decides access\n  ksmbd: refuse reauthentication of an expired session as another user\n  ksmbd: pin parent dentries in SMB2 CREATE ACL path\n  ksmbd: let the parent\u0027s DACL grant DELETE and FILE_READ_ATTRIBUTES\n  ksmbd: evaluate the DACL in order for explicit access requests\n  ksmbd: skip metadata-only stream opens in the delete check\n  ksmbd: skip metadata-only opens in sharing checks\n  ksmbd: refuse to read a handle that is not open for reading\n  ksmbd: grant a caller what its groups and OWNER RIGHTS ACEs allow\n"
    },
    {
      "commit": "25f5c92e1ac06496679ea229e5c825aa88968efd",
      "tree": "eafdb35910f2917d010f02d36b3cc4b0706d4e25",
      "parents": [
        "7183af7f01c839ecef87840a90c0d8c5843f37d7",
        "2d9bcd380585abc436f055fbb618b34f98882004"
      ],
      "author": {
        "name": "David Hildenbrand (Arm)",
        "email": "david@kernel.org",
        "time": "Fri Nov 13 11:18:23 2026 +0100"
      },
      "committer": {
        "name": "David Hildenbrand (Arm)",
        "email": "david@kernel.org",
        "time": "Sat Nov 14 16:36:05 2026 +0100"
      },
      "message": "Merge patch series \"mm: limit pagecache folio refcount\"\n\nDavid Hildenbrand (Arm) \u003cdavid@kernel.org\u003e says:\n\nUsers can create a lot of folio references to pagecache folios, making\nit possible to overflow the refcount of large folios.\n\nWhile we knew that one could in theory overflow a folio\u0027s refcount, we\nmostly assumed that bad things would happen and it wouldn\u0027t go\nunnoticed.\n\nNot so sure if that is really always the case, and leaving that problem\nunsolved is irresponsible, so let\u0027s limit pagecache folio refcounts.\n\nThis is mostly relevant for large folios / thps in the page cache, where\na single VMA can hold multiple folio references, allowing for many folio\nreferences. But with sufficient (memory+process) resources, it is also\npossible with small folios, just not that relevant in practice today.\n\nWith anon folios (including THPs), we should not be able to trigger folio\nrefcount overflows, as one would need so many processes that it\nbecomes impossible in practice. It was likely a problem for anon folios as\nwell before we had a dedicated pincount.\n\nLet\u0027s limit the refcount before it turns negative, giving admin processes\nsome extra reserves (to recover a problematic system). We still leave some\nheadroom to allow for some slight limit overshooting, to make the\nimplementation easier.\n\nBut to make that work reliably, we have to make anybody that can take\nan unbounded number of folio references obey these limits.\n\nPatch #1 handles mapping references from page tables. We also have to take\ncare of mmu_gather, to not hold too many references when zapping (whereby\nnew page faults could remap pages concurrently), which we do in patch #3.\nPatch #4 fixes iommufd cleanup when adding folio pins fails, which these\nlimits make easier to trigger.\n\nSplicing pagecache folios to a pipe (patch #2) also needs care. After\nfeedback from Linus, these users really should be using try_get_page(),\nwhere we now also implement the new limit check. We also teach GUP in\npatch #5 to stay below the limit, to avoid negative refcounts that cause\ntrouble.\n\nDoing some digging after getting a hint from Andrew, there are various\npaths in net that can also take an unbounded number of page refs and are\nproblematic. Let\u0027s use try_get_page() in all affected paths.\n\nNet paths that take page refs that I don\u0027t touch because the references\nare either bounded or the paths are not relevant in practice:\n\n - skb_split(): can add one reference per split, but each split\n   allocates a new skb that is charged to the socket, and the references\n   go away once the data was acked. Reworking that to be able to fail\n   requires more plumbing.\n - IP-TFS reassembly: needs one new inner skb per additional reference.\n - io_uring zerocopy send (managed frag references): each zerocopy piece\n   is charged at least 4 KiB of TCP memory, so not relevant in practice.\n - kTLS software splicing: only transfers references from the pipe, so\n   not a problem.\n - AF_XDP zerocopy TX, nfsd splice reads, sockmap / sk_msg, espintcp:\n   bounded, or require privileges / special setups, so not a problem in\n   practice.\n - Driver-specific paths (e.g., kTLS device offload): require special\n   hardware, didn\u0027t look into these.\n - page_pool and net_iov frags: not pagecache pages, so not a problem.\n\nFor page table mappings, this is only a temporary solution for\nupstream+stable, until we have something better, which will be harder to\nbackport:\n\n - Letting the mapcount of large folios be a long\n - Holding a single refcount if the folio is mapped, across all mappings\n - Reverting the limits introduced in this series\n\nAfter that, most of patch #1 will be reverted, everything else will\nstay.\n\nTested on x86-64 by lowering the folio refcount limits such that limits\ncan easily be exceeded, to then test fault (SIGBUS), fork(), splice and GUP\nbehavior briefly when we run into refcount limits on a shmem THP.\nThe net patches were tested similarly with AF_PACKET, nft dup, NFQUEUE,\nsoftware GSO segmentation, TCP receive coalescing with dropped ACKs and\nsocket splicing.\n\n* patches from https://patch.msgid.link/20261112-refcount_overflow_stable-v4-0-ad3deef0bd23@kernel.org:\n  mm/gup: limit pagecache folio refcount when grabbing references\n  iommufd: fix cleanup when adding memfd folio pins fails\n  mm/mmu_gather: flush pending references early before the folio limit\n  mm: limit pagecache folio refcount when splicing into pipes\n  mm: limit pagecache folio refcount when mapping pages\n\nThe -net patches part of the series will be dealt with separately.\n\nLink: https://patch.msgid.link/20261112-refcount_overflow_stable-v4-0-ad3deef0bd23@kernel.org\nSigned-off-by: David Hildenbrand (Arm) \u003cdavid@kernel.org\u003e\n"
    },
    {
      "commit": "2d9bcd380585abc436f055fbb618b34f98882004",
      "tree": "eafdb35910f2917d010f02d36b3cc4b0706d4e25",
      "parents": [
        "b8ba6cc9943222676d46ac77d3d3de2685a55bf8"
      ],
      "author": {
        "name": "David Hildenbrand (Arm)",
        "email": "david@kernel.org",
        "time": "Thu Nov 12 23:53:24 2026 +0100"
      },
      "committer": {
        "name": "David Hildenbrand (Arm)",
        "email": "david@kernel.org",
        "time": "Sat Nov 14 16:36:05 2026 +0100"
      },
      "message": "mm/gup: limit pagecache folio refcount when grabbing references\n\nLet\u0027s add __folio_ref_count_over_limit() checks in fast and slow GUP\nbefore acquiring additional references or pins.\n\nKeep it simple and use the check for all folio types.\n\nAlso propagate try_grab_folio() errors when grabbing additional\nreferences in __get_user_pages() and in memfd_pin_folios(), so hitting\nthe limit returns -ENOMEM instead of -EFAULT and -EINVAL, respectively.\n\nKeep the existing invalid-refcount warnings.\n\nFixes: 800d8c63b2e9 (\"shmem: add huge pages support\")\nFixes: 99cb0dbd47a1 (\"mm,thp: add read-only THP support for (non-shmem) FS\")\nFixes: 793917d997df (\"mm/readahead: Add large folio readahead\")\nCc: stable@vger.kernel.org\nReviewed-by: Jose A. Perez de Azpillaga \u003cazpijr@gmail.com\u003e\nReviewed-by: Pedro Falcato \u003cpfalcato@suse.de\u003e\nLink: https://patch.msgid.link/20261112-refcount_overflow_stable-v4-5-ad3deef0bd23@kernel.org\nSigned-off-by: David Hildenbrand (Arm) \u003cdavid@kernel.org\u003e\n"
    },
    {
      "commit": "b8ba6cc9943222676d46ac77d3d3de2685a55bf8",
      "tree": "67014ae9a4c6b79dbdcca584119489edebb459c6",
      "parents": [
        "317bd7d28b248fd4d4e14614d7c30b17a36ef4d7"
      ],
      "author": {
        "name": "David Hildenbrand (Arm)",
        "email": "david@kernel.org",
        "time": "Thu Nov 12 23:53:23 2026 +0100"
      },
      "committer": {
        "name": "David Hildenbrand (Arm)",
        "email": "david@kernel.org",
        "time": "Sat Nov 14 16:36:05 2026 +0100"
      },
      "message": "iommufd: fix cleanup when adding memfd folio pins fails\n\nfolio_add_pins() can fail after memfd_pin_folios() succeeded pinning a\nfolio. Upcoming folio refcount limit checks will make this easier to\ntrigger.\n\nLet\u0027s fix three problems in the error handling:\n\n1. pfn_reader_release_pins() uses the allocated array size instead of the\n   number of folios returned by memfd_pin_folios(). With large folios,\n   cleanup can unpin uninitialized entries. Remember the number of returned\n   folios and use that during cleanup.\n\n2. pin_memfd_pages() returns on failure without releasing the pins acquired\n   by the current call. Release the per-page pins for folios already\n   expanded into the page array, and the single pins for the remaining\n   folios.\n\n3. pfn_reader_next() returns on failure before updating batch_end_index,\n   even if the batch already contains successfully expanded folios. Update\n   it before propagating the error so cleanup releases those pins as well.\n\nFixes: 92687c793644 (\"iommufd: pfn_reader for file mappings\")\nCc: stable@vger.kernel.org\nReviewed-by: Lance Yang \u003clance.yang@linux.dev\u003e\nReviewed-by: Jason Gunthorpe \u003cjgg@nvidia.com\u003e\nLink: https://patch.msgid.link/20261112-refcount_overflow_stable-v4-4-ad3deef0bd23@kernel.org\nSigned-off-by: David Hildenbrand (Arm) \u003cdavid@kernel.org\u003e\n"
    },
    {
      "commit": "317bd7d28b248fd4d4e14614d7c30b17a36ef4d7",
      "tree": "15d20129a75c334c773754627a83fb4e5d2f6f15",
      "parents": [
        "c6c9a4c1c9a1669693330205cb9d86fafd3cf235"
      ],
      "author": {
        "name": "David Hildenbrand (Arm)",
        "email": "david@kernel.org",
        "time": "Thu Nov 12 23:53:22 2026 +0100"
      },
      "committer": {
        "name": "David Hildenbrand (Arm)",
        "email": "david@kernel.org",
        "time": "Sat Nov 14 16:36:05 2026 +0100"
      },
      "message": "mm/mmu_gather: flush pending references early before the folio limit\n\nWhen a task zaps mappings (e.g., MADV_DONTNEED), it can drop the PT lock\nand defer dropping the folio references. In the meantime, page faults\ncan refault pagecache folios, creating more references for the same\nfolios. These deferred references can build up over multiple VMAs.\n\nLet\u0027s just request a flush after enqueueing folios whose refcount reached\na certain threshold. Start flushing well before new references are refused,\nregardless of the caller\u0027s privileges.\n\nWhen returning \"true\" from __tlb_remove_folio_pages_size(), PTE zapping\nwill flush the TLB before dropping the PT lock, but drop the references\nafterwards. Refaults can still race with that, but flushing early reduces\nthe risk of letting others hit the limit.\n\nThis is a temporary measure until we have something better.\n\nReported-by: Chris Mason \u003cmason@kernel.org\u003e\nFixes: 800d8c63b2e9 (\"shmem: add huge pages support\")\nFixes: 99cb0dbd47a1 (\"mm,thp: add read-only THP support for (non-shmem) FS\")\nFixes: 793917d997df (\"mm/readahead: Add large folio readahead\")\nCc: stable@vger.kernel.org\nReviewed-by: Jose A. Perez de Azpillaga \u003cazpijr@gmail.com\u003e\nReviewed-by: Lance Yang \u003clance.yang@linux.dev\u003e\nReviewed-by: Pedro Falcato \u003cpfalcato@suse.de\u003e\nLink: https://patch.msgid.link/20261112-refcount_overflow_stable-v4-3-ad3deef0bd23@kernel.org\nSigned-off-by: David Hildenbrand (Arm) \u003cdavid@kernel.org\u003e\n"
    },
    {
      "commit": "c6c9a4c1c9a1669693330205cb9d86fafd3cf235",
      "tree": "ad18043f8fc4e3feab1073be551c557ba1629e44",
      "parents": [
        "26b8f029ce5224dd5c5f45a4bc964a209c259914"
      ],
      "author": {
        "name": "David Hildenbrand (Arm)",
        "email": "david@kernel.org",
        "time": "Thu Nov 12 23:53:21 2026 +0100"
      },
      "committer": {
        "name": "David Hildenbrand (Arm)",
        "email": "david@kernel.org",
        "time": "Sat Nov 14 16:36:05 2026 +0100"
      },
      "message": "mm: limit pagecache folio refcount when splicing into pipes\n\nSimilar to mapping pagecache folios into page tables, let\u0027s reject\nadditional references when splicing folios into a pipe if the refcount\nis at its limit.\n\nThis will prevent more cases of possible folio refcount overflows with\nlarge pagecache folios.\n\nIntroduce folio_get_checked() for checked reference acquisition when\nthe caller already holds a reference. It warns and fails on a non-positive\nrefcount and rejects additional references at the limit.\n\nAdd __folio_ref_count_over_limit() so folio_get_checked() can reuse the\nrefcount instead of having to read it a second time.\n\nUse folio_get_checked() in splice_folio_into_pipe() and\nfuse_ref_folio(), returning -EFAULT on failure, like we already do for\npipe-buffer duplication. Similarly, make try_get_page() use it instead.\n\nLikely try_get_page() should be renamed (or removed) later, as it doesn\u0027t\nmatch the folio_try_get() semantics and is confusing.\n\nIf we hit the limit after splicing some data, return the number of bytes\nalready spliced. Teach the callers to handle errors, and stop\nfilemap_splice_read() if we only spliced part of a folio, so we don\u0027t\nskip ahead to the next folio: note that we wouldn\u0027t run into this\nscenario before as we always spliced the requested range.\n\nThe check in try_get_page() also covers pipe-buffer duplication by tee()\nand partial pipe-to-pipe splicing.\n\nNote that we deliberately will not fall back to the slow path of copying\nthe folio, particularly as the folio might have an invalid refcount. If\nit is ever really required, we can revisit.\n\nFixes: 800d8c63b2e9 (\"shmem: add huge pages support\")\nFixes: 99cb0dbd47a1 (\"mm,thp: add read-only THP support for (non-shmem) FS\")\nFixes: 793917d997df (\"mm/readahead: Add large folio readahead\")\nCc: stable@vger.kernel.org\nReviewed-by: Jose A. Perez de Azpillaga \u003cazpijr@gmail.com\u003e\nReviewed-by: Pedro Falcato \u003cpfalcato@suse.de\u003e\nLink: https://patch.msgid.link/20261112-refcount_overflow_stable-v4-2-ad3deef0bd23@kernel.org\nSigned-off-by: David Hildenbrand (Arm) \u003cdavid@kernel.org\u003e\n"
    },
    {
      "commit": "26b8f029ce5224dd5c5f45a4bc964a209c259914",
      "tree": "ac4d83a1ede33a7d46744f50dc26443d0048dd0a",
      "parents": [
        "7183af7f01c839ecef87840a90c0d8c5843f37d7"
      ],
      "author": {
        "name": "David Hildenbrand (Arm)",
        "email": "david@kernel.org",
        "time": "Thu Nov 12 23:53:20 2026 +0100"
      },
      "committer": {
        "name": "David Hildenbrand (Arm)",
        "email": "david@kernel.org",
        "time": "Sat Nov 14 16:35:59 2026 +0100"
      },
      "message": "mm: limit pagecache folio refcount when mapping pages\n\nWith large file folios, overflowing the folio mapcount and refcount\nbecame possible more easily. For anon folios it is currently not really\nan issue: each large anon folio can only be mapped once into a process; on\n4k kernels it would be more processes than usually allowed on a system,\nand on 16k or 64k kernels the processes alone would consume hundreds of\nGiB. Something to be sorted out at some point, but it\u0027s not as urgent.\n\nBeing able to overflow the refcount is not really big news, but it\u0027s about\ntime that we do something about it because some modern technology might\nfind ways to abuse it :)\n\nOne of the enablers for that was commit ddc58f27f9ee (\"mm: drop tail page\nrefcounting\"), however, file thp support was added later.\n\nFile folios can be mapped multiple times into the same process, allowing\nfor processes to hold many references to the same folio.\n\nAs a short-term solution, let\u0027s introduce two folio refcount limits:\none for ordinary consumers and one for admins. Leave some headroom\nbefore the refcount goes negative, which we really don\u0027t want (having\nthe mapcount go negative is also not expected).\n\nStart reporting SIGBUS and a ratelimited warning if we run into limits.\nWhen restoring mappings from migration entries, instead zap the pte/pmd\nentry. We allow for slightly exceeding the limit to keep things simpler.\n\nDuring fork(), prefer not copying page table entries over failing fork.\nFor fault-around, check in next_uptodate_folio() before trying to lock\nthe folio.\n\nExport folio_ref_count_over_limit_slowpath() using EXPORT_SYMBOL_GPL\nright away so modules can use it through try_get_page(), where we will\nuse it next.\n\nUse has_capability_noaudit(current, CAP_SYS_RESOURCE),\nso filesystem credential overrides do not affect access to the admin\nreserve, and to not generate auditing messages simply because we access\noptional reserves.\n\nDon\u0027t grant access to the admin reserve from kernel threads, from\ninterrupt context, where \"current\" is just the interrupted task, or with\nIRQs disabled, as even unaudited capability checks can take locks.\n\nThis is far from ideal, but better than all easy alternatives we can\nbackport comfortably. For now it fixes the big problem: letting\nthe refcount (+mapcount) go negative and cause trouble.\n\nThis is only a temporary fix. Keep the mapping checks explicit and\nseparate from reference acquisition: most are expected to disappear when\nwe rework the underlying refcounting.\n\nCheck order 0 folios as well for consistency: each\nfolio_add_file_rmap*() and folio_dup_file_rmap*() needs a prior check.\n\nAdjust the check in folio_get() to just check for invalid refcounts and\ndrop folio_ref_zero_or_close_to_overflow(): we really don\u0027t want\nnegative refcounts or negative mapcounts. Convert to VM_WARN_ON_FOLIO as\nwe touch it.\n\nAlso, just document -EHWPOISON for copy_present_ptes() while at it.\n\nNote: When zapping file migration entries, uffd-wp is preserved using PTE\nmarkers, but uffd-rwp protection is lost because there is no corresponding\nmarker. try_to_unmap() already has the same problem when unmapping file\nfolios, so we\u0027ll happily ignore it.\n\nReported-by: Chris Mason \u003cmason@kernel.org\u003e\nFixes: 800d8c63b2e9 (\"shmem: add huge pages support\")\nFixes: 99cb0dbd47a1 (\"mm,thp: add read-only THP support for (non-shmem) FS\")\nFixes: 793917d997df (\"mm/readahead: Add large folio readahead\")\nCc: stable@vger.kernel.org\nReviewed-by: Jose A. Perez de Azpillaga \u003cazpijr@gmail.com\u003e\nAcked-by: Zi Yan \u003cziy@nvidia.com\u003e\nReviewed-by: Pedro Falcato \u003cpfalcato@suse.de\u003e\nLink: https://patch.msgid.link/20261112-refcount_overflow_stable-v4-1-ad3deef0bd23@kernel.org\nSigned-off-by: David Hildenbrand (Arm) \u003cdavid@kernel.org\u003e\n"
    },
    {
      "commit": "603b4fd69b4e77770761ea03451e2164186f8590",
      "tree": "4c0925ecd1ef1c7fb1e688c27eb087fbee512927",
      "parents": [
        "7475ea8b58616b33c04afb976cf0fbaad9dd63fb"
      ],
      "author": {
        "name": "Vineet Gupta",
        "email": "vgupta@kernel.org",
        "time": "Thu Sep 24 21:16:00 2026 -0700"
      },
      "committer": {
        "name": "Vineet Gupta",
        "email": "vgupta@kernel.org",
        "time": "Fri Nov 13 21:21:13 2026 -0800"
      },
      "message": "ARC: io: work around GCC PR127718 in __raw_write{w,l}()\n\nThe __raw_write{w,l}() accessors use inline asm with an \"m\" operand for\nthe store destination. GCC\u0027s ARC backend lets \"m\" match any legitimate\naddress, including the register-indexed [b,c] forms that only loads\nsupport: stores only have [reg], [reg,s9] and [limm] addressing. With\nnewer GCC (late-combine in GCC 15, combine in GCC 16), a writew() with a\nscaled, non-constant offset gets\n\n\tstw.as r2,[r0,r1]\n\nwhich the assembler rejects:\n\n  Error: inappropriate arguments for opcode \u0027stw\u0027\n\nWith GCC 16 this breaks even hsdk_defconfig (drivers/net/pcs/\npcs-xpcs-plat.c) and axs101_defconfig. allyesconfig additionally trips\nin drivers/pinctrl/renesas/pinctrl-rza2.c and drivers/video/fbdev/\ns3c-fb.c.\n\nWork around it by passing the address in a register and using a fixed\n[%1] address in the asm template, so the compiler can\u0027t pick an\naddressing mode that stores don\u0027t support. This is the minimal fix.\nThe next patch drops the inline asm altogether.\n\nLink: https://gcc.gnu.org/bugzilla/show_bug.cgi?id\u003d127718\nCc: stable@vger.kernel.org\nSigned-off-by: Vineet Gupta \u003cvgupta@kernel.org\u003e\n"
    },
    {
      "commit": "860b3016e6d6fd6ae83eb8f05d1861a1876db060",
      "tree": "d89e76fbb508c8df7b8c5cbd6eea3f7c11211427",
      "parents": [
        "b21c45d2151a2ce78806b4234513a1793525611b",
        "bab48a1044475f62d6e910c6d7267ce4bae847f9"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Fri Nov 13 18:14:32 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Fri Nov 13 18:14:32 2026 -0800"
      },
      "message": "Merge tag \u0027keys-v7.4-rc3-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jarkko/linux-tpmdd\n\nPull key fixes from Jarkko Sakkinen:\n\n - Invoke the match callback before the state checks in\n   keyring_search_iterator(), so that a non-matching revoked,\n   expired or invalidated key cannot decide the search result\n\n - Take key-\u003esem for reading around the -\u003easym_*() method calls\n   in the keyctl pkey operations, as a key reference does not\n   pin the module providing the key type\n\n* tag \u0027keys-v7.4-rc3-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jarkko/linux-tpmdd:\n  KEYS: Protect pkey method checks against type removal\n  KEYS: do not let an unmatched key decide the search result\n"
    },
    {
      "commit": "b21c45d2151a2ce78806b4234513a1793525611b",
      "tree": "4151ab8c1965cb0b4a8c477c3e96ae4ce174647a",
      "parents": [
        "0716b41b472ed6ec9c9c9bae1f8836eadd149f95",
        "637c4514732baa5d19bb716c00d6fd67034f238f"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Fri Nov 13 18:08:25 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Fri Nov 13 18:08:25 2026 -0800"
      },
      "message": "Merge tag \u0027drm-fixes-2026-11-14\u0027 of https://gitlab.freedesktop.org/drm/kernel\n\nPull drm fixes from Dave Airlie:\n \"Weekly drm fixes, this is a bit bigger as we missed xe last week due\n  to a tree screw up, so it\u0027s mostly xe and amdgpu, then the usual\n  scattering of fixes across drivers.\n\n  sched:\n   - Ensure stopped entity cannot be re-added to the run-queue\n\n  bridge:\n   - ssd2825: Fix panel enable sequence\n   - sii902x: validate I2S data lane values\n\n  display:\n   - hdmi: Fixes to SCDC\n\n  i915:\n   - Fix for PTL GitLab #9196: eDP-2 on DDI B ends up in a state where\n     any enable fails\n   - PSR fixes\n\n  amdgpu:\n   - DCN 6 DML2 fix\n   - VCE fixes\n   - GC 11.0.2 fix\n   - SMU 15 fix\n   - Error handling fixes\n   - Partition fix\n   - IMU firmware clean up fix\n   - TLB fence handling fix\n\n  amdkfd:\n   - SVM fixes\n   - CRIU fixes\n\n  xe:\n   - select configfs\n   - Sync Wa_16023105232 in xe_was to match OOB rule\n   - Program LMTT directory pointer from each GT init path\n   - Demote uncorrectable local error log on CRI\n   - Apply tile workarounds after PCODE init\n   - VRAM offlining fixes\n   - ras: Fix sysctrl and drm-ras discrepancy\n   - pat: Return error from xe_pat_init_early() on unknown\n   - Prevent VF memory IRQ from accessing a torn-down engine\n   - guc: Protect timeout handling from device unplug\n   - gt: Disable GT reset worker before engine teardown\n   - Various hwmon fixes\n   - tuning: Restrict Anyhit NULL-query tuning to render engines\n   - gsc: Fix NVL-S GSC compatibility number\n   - Don\u0027t abort the shrinker walk on -EBUSY\n   - Mark xe_vma_ops_alloc() noinline for error injection\n   - lrc: Restore CTX_CS_INDIRECT_CTX_OFFSET programming for pre-Xe2\n   - Fix build with DRM_XE\u003dy and CONFIGFS_FS\u003dm\n   - madvise: use already clamped pat_index\n   - Various wedging fixes\n   - Rebind BOs evicted before the first LR queue\n   - Toggle the D3Cold enable flag at runtime resume\n   - Access GT MMIO after pcode init\n   - xe_late_bind_fw: Fix firmware leak on missing manifest entry\n   - svm: Preserve default PAT index when resetting VMA attributes\n   - Unwrap sync fences before adding them as job dependencies\n   - Schedule disable G2H fixes\n\n  gud:\n   - Avoid divide error in gud_stats_debugfs\n\n  meson:\n   - load ENCL gamma table while ENCL is disabled\n\n  nouveau:\n   - disp: Set atomic functions in struct drm_mode_config_funcs\n   - gsp: Handle only IRQs routed to host\n\n  panfrost:\n   - Fixes to perfcnt\n\n  vc4:\n   - Fix firmware reference leak\n   - Hold runtime PM reference while allocating the binner BO\n   - Rebuild V3D interrupt mask on every power transition\n   - Re-enable the V3D interrupt after a GPU reset\"\n\n* tag \u0027drm-fixes-2026-11-14\u0027 of https://gitlab.freedesktop.org/drm/kernel: (81 commits)\n  drm/vc4: Re-enable the V3D interrupt after a GPU reset\n  drm/vc4: Rebuild the V3D interrupt mask on every power transition\n  drm/vc4: Hold a runtime PM reference while allocating the binner BO\n  drm: xe: select configfs\n  drm/xe: Sync Wa_16023105232 in xe_was to match OOB rule\n  drm/xe/pf: Program LMTT directory pointer from each GT\u0027s init path\n  drm/xe/guc: Demote uncorrectable local error log on CRI\n  drm/xe/pm: Apply tile workarounds after PCODE initialization\n  drm/gud: avoid divide error in gud_stats_debugfs\n  drm/nouveau/disp: Set drm_mode_config_funcs.atomic_(check|commit) (attempt 2)\n  drm/amdgpu: fix TLB invalidation fence handling v4\n  drm/amdkfd: check CRIU SVM priv data bounds before allocating\n  drm/amdgpu: release the IMU firmware with the other GFX microcode\n  drm/amdgpu: Allow cancelling a requested NPS mode\n  drm/amdgpu: fix the error handling of amdgpu_vm_init_entities()\n  drm/amdgpu: use supported SMU clock messages on SMU 15.0.0\n  drm/amdkfd: skip GPUs without a VM when checkpointing the BO mappings\n  drm/amdkfd: Acquire cwsr buffer during CRIU resume\n  drm/amdkfd: Acquire buffers in CRIU queue restore (again)\n  drm/amdgpu: Fix VCE buffer size check for DPB and input surface\n  ...\n"
    },
    {
      "commit": "0716b41b472ed6ec9c9c9bae1f8836eadd149f95",
      "tree": "8b9685a1aa6821a8ee586b392f88042884b36527",
      "parents": [
        "97d73c8251226cf27cc6656b4c542869bb8c69fa",
        "f00afd5d8d5dc03994cb67246543d5c70571ebbe"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Fri Nov 13 17:53:27 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Fri Nov 13 17:53:27 2026 -0800"
      },
      "message": "Merge tag \u0027watchdog-fixes-for-v7.4-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging\n\nPull watchdog fixes from Guenter Roeck:\n\n - gunyah_wdt: Re-enable the watchdog on resume\n\n - mediatek: Register reset controller before watchdog device\n\n - meson_gxbb_wdt: fix suspend/resume handling of HW_RUNNING watchdog\n\n - mtk_wdt: Stop a running watchdog over system sleep even if not opened\n\n - npcm: fix clock prepare leak when setting timeout on a running\n   watchdog\n\n - omap: Fix early_enable handling in omap_wdt_probe()\n\n - orion_wdt: load the pretimeout counter on Armada 370/XP start\n\n - rzv2h: Drop runtime PM handling from WDTDCR init, and use\n   pm_runtime_put_sync()\n\n - sbsa_gwdt: Limit the MediaTek WS0 race workaround to revision 0\n\n - sc1200wdt: add missing MODULE_DEVICE_TABLE()\n\n - sp5100_tco: Use release_mem_region() for EFCH MMIO resource\n\n* tag \u0027watchdog-fixes-for-v7.4-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging:\n  watchdog: rzv2h: Drop runtime PM handling from WDTDCR init\n  watchdog: rzv2h: Use pm_runtime_put_sync()\n  watchdog: omap: Fix early_enable handling in omap_wdt_probe()\n  watchdog: npcm: fix clock prepare leak when setting timeout on a running watchdog\n  watchdog: gunyah_wdt: Re-enable the watchdog on resume\n  watchdog: mediatek: Register reset controller before watchdog device\n  watchdog: sp5100_tco: Use release_mem_region() for EFCH MMIO resource\n  watchdog: sbsa_gwdt: Limit the MediaTek WS0 race workaround to revision 0\n  watchdog: meson_gxbb_wdt: fix suspend/resume handling of HW_RUNNING watchdog\n  watchdog: mtk_wdt: Stop a running watchdog over system sleep even if not opened\n  watchdog: orion_wdt: load the pretimeout counter on Armada 370/XP start\n  watchdog: sc1200wdt: add missing MODULE_DEVICE_TABLE()\n"
    },
    {
      "commit": "637c4514732baa5d19bb716c00d6fd67034f238f",
      "tree": "67443a82ef56b32fb0f7aeb49c3aad7dc4979396",
      "parents": [
        "7364a9c7a3a00bda0fc420d40682ad1eb5a41905",
        "bafda823590ddf7031ecb765d673572fae7ab44a"
      ],
      "author": {
        "name": "Dave Airlie",
        "email": "airlied@redhat.com",
        "time": "Sat Nov 14 09:48:50 2026 +1000"
      },
      "committer": {
        "name": "Dave Airlie",
        "email": "airlied@redhat.com",
        "time": "Sat Nov 14 09:49:09 2026 +1000"
      },
      "message": "Merge tag \u0027drm-misc-fixes-2026-11-13\u0027 of https://gitlab.freedesktop.org/drm/misc/kernel into drm-fixes\n\nShort summary of fixes pull:\n\nbridge:\n- ssd2825: Fix panel enable sequence\n- sii902x: validate I2S data lane values\n\ndisplay:\n- hdmi: Fixes to SCDC\n\ngud:\n- Avoid divide error in gud_stats_debugfs\n\nmeson:\n- load ENCL gamma table while ENCL is disabled\n\nnouveau:\n- disp: Set atomic functions in struct drm_mode_config_funcs\n- gsp: Handle only IRQs routed to host\n\npanfrost:\n- Fixes to perfcnt\n\nsched:\n- Ensure stopped entity cannot be re-added to the run-queue\n\nvc4:\n- Fix firmware reference leak\n- Hold runtime PM reference while allocating the binner BO\n- Rebuild V3D interrupt mask on every power transition\n- Re-enable the V3D interrupt after a GPU reset\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n\nFrom: Thomas Zimmermann \u003ctzimmermann@suse.de\u003e\nLink: https://patch.msgid.link/20261113074438.GA14592@linux.fritz.box\n"
    },
    {
      "commit": "97d73c8251226cf27cc6656b4c542869bb8c69fa",
      "tree": "54035c76f6e4c330b72eb5c0140b46b5143e5b9f",
      "parents": [
        "2800868f63227c8237d70c41e70d441c919620d6",
        "ad3fc05c7e65009355effeea06aa629a687acef2"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Fri Nov 13 13:44:27 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Fri Nov 13 13:44:27 2026 -0800"
      },
      "message": "Merge tag \u0027arm64-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux\n\nPull arm64 fixes from Catalin Marinas:\n \"Fix hugetlb write fault livelock, permission overlay and GCS state\n  preservation across CPU suspend, cpuidle context tracking during\n  system suspend, GCS mode handling on signal return, AMU counter errata\n  workarounds, kexec loading of images with a non-zero text_offset and a\n  spurious ptrace warning:\n\n   - Fix hugetlb write fault livelocks caused by stale read-only TLB\n     entries after another CPU makes the page writable and only performs\n     a local invalidate. Unlike the generic fault path, hugetlb has no\n     fix_spurious_fault() fallback\n\n   - Save and restore POR_EL0, GCSPR_EL0 and GCSCRE0_EL1 across CPU\n     suspend\n\n   - Restrict cpuidle context tracking in cpu_suspend() to idle tasks,\n     avoiding context tracking warnings during system suspend\n\n   - Ignore attempts to change GCS mode through signal return, which\n     updated the software state without updating the hardware. Changes\n     made through prctl() in the signal handler are preserved. This\n     changes the documented ABI but no known userspace relies on it.\n     Allowing mode changes through both prctl() and signal return also\n     creates confusing interactions, with signal return potentially\n     undoing changes deliberately made by the handler\n\n   - Extend the AMU constant counter workaround to additional CPUs and\n     add a core cycle counter workaround for Cortex-A715/A720, treating\n     the affected counters as unavailable\n\n   - Avoid CMA allocations when loading a kexec Image with a non-zero\n     text_offset, fixing an incorrect entry point and a memory leak when\n     freeing the image\n\n   - Remove a spurious ptrace warning when setting the syscall number of\n     a task in a non-seized group-stop without siginfo\"\n\n* tag \u0027arm64-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:\n  arm64: suspend: Only track cpuidle context for idle tasks\n  arm64: ptrace: don\u0027t warn on missing siginfo in a group-stop\n  arm64: kexec_file: Do not use CMA for images with a non-zero text_offset\n  arm64: errata: Add Cortex-A715/A720 AMU core counter workaround\n  arm64: errata: Extend the workaround for broken AMEVCNTR01 counter\n  arm64/gcs: Ignore attempts to change GCS mode during signal return\n  arm64: hugetlb: Prevent write fault livelocks on stale TLB entries\n  arm64: suspend: Save GCS state over suspend\n  arm64: suspend: Save POR_EL0 over suspend\n"
    },
    {
      "commit": "e92cd41e8e5ee5616c6c7cd4378661d942f0a7c0",
      "tree": "b0fce2cfd1ec4872a9df68b48df763df69b475b5",
      "parents": [
        "46af2c975958119290ddcd984a98491966123a7b"
      ],
      "author": {
        "name": "Daejun Park",
        "email": "daejun7.park@samsung.com",
        "time": "Thu Nov 12 10:35:29 2026 +0900"
      },
      "committer": {
        "name": "Mikulas Patocka",
        "email": "mpatocka@redhat.com",
        "time": "Fri Nov 13 22:18:14 2026 +0100"
      },
      "message": "dm: fail a REQ_PREFLUSH bio with data when its preflush fails\n\nFor a bio with REQ_PREFLUSH and data, dm core may send an empty flush\nfirst. If the flush fails, __dm_io_complete() still reissues the bio\nwithout REQ_PREFLUSH and drops the error. When that write succeeds,\nthe XFS log keeps running and fsync() can return 0.\n\nEnd the bio with the flush error instead of writing the data, as\nblk-mq does.\n\nFixes: 6a8736d10cb4 (\"dm: relax ordering of bio-based flush implementation\")\nCc: stable@vger.kernel.org\nSigned-off-by: Daejun Park \u003cdaejun7.park@samsung.com\u003e\nAcked-by: Tejun Heo \u003ctj@kernel.org\u003e\nSigned-off-by: Mikulas Patocka \u003cmpatocka@redhat.com\u003e\n"
    },
    {
      "commit": "b5c5b67e02406fb48acf57cf2c30a570230fd6aa",
      "tree": "8f287da9b2afe9677bc63d74c797fbf558367c61",
      "parents": [
        "68af85b730f9a9e5f66c96ba861a66da08cfa5a9"
      ],
      "author": {
        "name": "Rupesh Majhi",
        "email": "zoone.rupert@gmail.com",
        "time": "Sun Nov 08 19:26:51 2026 +0200"
      },
      "committer": {
        "name": "Guenter Roeck",
        "email": "linux@roeck-us.net",
        "time": "Fri Nov 13 12:08:12 2026 -0800"
      },
      "message": "hwmon: (aht10) Fix update_interval overflow on 64-bit\n\nOn 64-bit, update_interval writes are clamped to LONG_MAX ms, but\nms_to_ktime() overflows above 9223372036854 ms and the interval wraps.\nLONG_MAX becomes -1 ms, so every read starts a new measurement and\nupdate_interval reads back -1.\n\nClamp to INT_MAX like sht4x, so the limit is the same on all platforms.\n\nFixes: 8c78f0dee437 (\"hwmon: Add AHT10 Temperature and Humidity Sensor Driver\")\nCc: stable@vger.kernel.org\nSigned-off-by: Rupesh Majhi \u003czoone.rupert@gmail.com\u003e\nLink: https://patch.msgid.link/20261108172651.220121-1-zoone.rupert@gmail.com\nSigned-off-by: Guenter Roeck \u003clinux@roeck-us.net\u003e\n"
    },
    {
      "commit": "68af85b730f9a9e5f66c96ba861a66da08cfa5a9",
      "tree": "994228c1718c0c70aa0e0dc6eb48aa3206ebb6d5",
      "parents": [
        "a2996d892d0150d059d39c683d95d6fd7a37630a"
      ],
      "author": {
        "name": "Sudeep Holla",
        "email": "sudeep.holla@kernel.org",
        "time": "Mon Nov 09 15:19:27 2026 +0000"
      },
      "committer": {
        "name": "Guenter Roeck",
        "email": "linux@roeck-us.net",
        "time": "Fri Nov 13 12:08:12 2026 -0800"
      },
      "message": "hwmon: (scmi) Configure sensors before exposing channels\n\nSCMI hwmon enables sensors with SENSOR_CONFIG_SET, but only after\nregistering the hwmon device and only for temperature sensors. This\nleaves other sensor classes disabled on firmware that defaults sensors\nto off, and exposes temperature channels before they are ready.\n\nConfigure each supported sensor while building the channel list, before\nregistering the hwmon device. Leave failed sensors out of the list so\ntheir channels remain hidden, while continuing with other sensors.\n\nCc: Marek Vasut \u003cmarek.vasut@mailbox.org\u003e\nReported-by: Sashiko \u003csashiko-bot@kernel.org\u003e\nCloses: https://sashiko.dev/#/bug/linux-4bdf9a68-22d5-4305-b684-f11fe6d1dbcc\nFixes: 72804d1631b3 (\"hwmon: (scmi) Enable sensors to assure they can be read\")\nCc: stable@vger.kernel.org\nSigned-off-by: Sudeep Holla \u003csudeep.holla@kernel.org\u003e\nLink: https://patch.msgid.link/20261109151927.1672956-2-sudeep.holla@kernel.org\nSigned-off-by: Guenter Roeck \u003clinux@roeck-us.net\u003e\n"
    },
    {
      "commit": "a2996d892d0150d059d39c683d95d6fd7a37630a",
      "tree": "340676b08fd1befdaf801fcc9c6df9ee4b148d34",
      "parents": [
        "ca1abdd1a9b9ce51713a18c8905bf10579533c3b"
      ],
      "author": {
        "name": "Sudeep Holla",
        "email": "sudeep.holla@kernel.org",
        "time": "Mon Nov 09 15:19:26 2026 +0000"
      },
      "committer": {
        "name": "Guenter Roeck",
        "email": "linux@roeck-us.net",
        "time": "Fri Nov 13 12:08:12 2026 -0800"
      },
      "message": "hwmon: (scmi) Check channel config allocation\n\nscmi_hwmon_add_chan_info() can fail to allocate a channel\u0027s config\narray. Ignoring that error leaves a channel with a NULL config pointer\nin the hwmon chip info, which hwmon dereferences while registering the\ndevice.\n\nCheck the helper\u0027s return value and abort probe on failure.\n\nReported-by: Sashiko \u003csashiko-bot@kernel.org\u003e\nCloses: https://sashiko.dev/#/bug/linux-d10ced6b-a6f0-403b-9619-f5a6573f9e7f\nFixes: b23688aefb8b (\"hwmon: add support for sensors exported via ARM SCMI\")\nCc: stable@vger.kernel.org\nSigned-off-by: Sudeep Holla \u003csudeep.holla@kernel.org\u003e\nLink: https://patch.msgid.link/20261109151927.1672956-1-sudeep.holla@kernel.org\nSigned-off-by: Guenter Roeck \u003clinux@roeck-us.net\u003e\n"
    },
    {
      "commit": "ca1abdd1a9b9ce51713a18c8905bf10579533c3b",
      "tree": "e122633b646032e7c3e348489d967a076dc8f950",
      "parents": [
        "3ae1af2faad93276645d7620261643457daec73b"
      ],
      "author": {
        "name": "Anton Lundin",
        "email": "glance@ac2.se",
        "time": "Mon Nov 09 09:25:09 2026 +0100"
      },
      "committer": {
        "name": "Guenter Roeck",
        "email": "linux@roeck-us.net",
        "time": "Fri Nov 13 12:08:12 2026 -0800"
      },
      "message": "hwmon: (nct6775) Handle dual use registers\n\nThe REG_TARGET register has different meaning, depending on if pwm is in\nthermal_cruise or speed_cruise.\n\nThis means we can\u0027t assume it\u0027s always word sized, so we need to handle\nit differently depending on which mode we\u0027re in.\n\nThe code already does this, but the register was configured as word\nsized for the speed_cruise mode, which would do the wrong thing in\nthermal_cruise mode.\n\nFixes: 6c009501ff20 (\"hwmon: (nct6775) Add support for NCT6102D/6106D\")\nFixes: 29c7cb485b32 (\"hwmon: (nct6775) Integrate new model nct6116\")\nFixes: 437b3c2a4bdc (\"hwmon: (nct6775) Add support for NCT6126D\")\nCc: stable@vger.kernel.org\nSigned-off-by: Anton Lundin \u003cglance@ac2.se\u003e\nLink: https://patch.msgid.link/20261109082511.851377-2-glance@ac2.se\nSigned-off-by: Guenter Roeck \u003clinux@roeck-us.net\u003e\n"
    },
    {
      "commit": "3ae1af2faad93276645d7620261643457daec73b",
      "tree": "c1559433fdbae35d80630ff02cc2ae1709379335",
      "parents": [
        "00944852275efd39706fe7b9ce75c28edde983eb"
      ],
      "author": {
        "name": "Xu Rao",
        "email": "raoxu@uniontech.com",
        "time": "Mon Nov 09 16:28:54 2026 +0800"
      },
      "committer": {
        "name": "Guenter Roeck",
        "email": "linux@roeck-us.net",
        "time": "Fri Nov 13 12:08:12 2026 -0800"
      },
      "message": "hwmon: (scmi) Enable temperature sensors using their sensor IDs\n\nscmi_hwmon_probe() groups sensors by hwmon type, so the index in\nscmi_sensors-\u003einfo[hwmon_temp] is not necessarily the SCMI sensor ID.\nPassing that index to config_set() targets a different sensor if the\nplatform exposes non-temperature sensors ahead of temperature sensors.\nThe intended temperature sensor can remain disabled, causing subsequent\nreadings to fail. If configuring the wrong sensor fails, thermal\nregistration of the intended sensor is also skipped.\n\nUse the firmware sensor ID, as the read and thermal registration paths\nalready do, so temperature sensors are enabled correctly.\n\nThe patch was reviewed and the commit message expanded with OpenCode\nAI assistance.\n\nFixes: 72804d1631b3 (\"hwmon: (scmi) Enable sensors to assure they can be read\")\nCc: stable@vger.kernel.org\nSigned-off-by: Xu Rao \u003craoxu@uniontech.com\u003e\nReviewed-by: Sudeep Holla \u003csudeep.holla@kernel.org\u003e\nLink: https://patch.msgid.link/AED18609B941BCFD+20261109082854.3739085-1-raoxu@uniontech.com\nSigned-off-by: Guenter Roeck \u003clinux@roeck-us.net\u003e\n"
    },
    {
      "commit": "00944852275efd39706fe7b9ce75c28edde983eb",
      "tree": "6420a2da95b3841fe59da5179a1a030101d5a859",
      "parents": [
        "fc8519fab34e162b0c29dacccf73031820878917"
      ],
      "author": {
        "name": "Felix Radensky",
        "email": "fradensky@nvidia.com",
        "time": "Thu Nov 05 21:32:00 2026 +0200"
      },
      "committer": {
        "name": "Guenter Roeck",
        "email": "linux@roeck-us.net",
        "time": "Fri Nov 13 12:08:12 2026 -0800"
      },
      "message": "hwmon: (pmbus/mp2975) Get vout fault limits dynamically\n\nThe over- and under-voltage fault limits of the tracking mode are\nrelative to the voltage reference, which follows the VID/AVS requests\nissued to the device and therefore changes at runtime.\n\nThe driver reads the voltage reference once during probe and derives the\nlimits from that snapshot, which the PMBus core caches in addition. As a\nresult in2_crit and in2_lcrit reflect the operating point at probe, rather\nthan the thresholds the device enforces. On systems where the device\nsupplies the CPU core voltage, in2_crit was reported anywhere between\n1.15V and 1.70V across identical systems, causing false over-voltage\nreports once the CPU voltage exceeded the value latched at probe.\n\nRead the voltage reference and the over-voltage offset selector whenever\na limit is requested, and mark both limits with pmbus_set_update(), so\nthat the core does not cache them. The voltage reference is available on\npage 3 only. Let the core change the page, so that the page it caches\nmatches the device, except for a single rail configuration, for which the\ncore does not issue PAGE commands, and the page is changed directly.\n\nWith the selector read at runtime, a device configured with an\nunsupported selector no longer fails to probe. The OVP1 threshold, which\ndoes not depend on the selector, is reported for it instead, with a\nwarning.\n\nClamp the under-voltage limit to zero, since the offset can exceed a low\nvoltage reference, while a negative value would be reported as a read\nerror.\n\nFixes: 2c6fcbb21149 (\"hwmon: (pmbus) Add support for MPS Multi-phase mp2975 controller\")\nCc: stable@vger.kernel.org\nSigned-off-by: Felix Radensky \u003cfradensky@nvidia.com\u003e\nReviewed-by: Vadim Pasternak \u003cvadimp@nvidia.com\u003e\nLink: https://patch.msgid.link/20261105193200.1570996-1-fradensky@nvidia.com\nSigned-off-by: Guenter Roeck \u003clinux@roeck-us.net\u003e\n"
    },
    {
      "commit": "fc8519fab34e162b0c29dacccf73031820878917",
      "tree": "5de65026ebd91a3c84ca3edb01c20bd31589c6f2",
      "parents": [
        "33d58cbf2ab26bda81bc3052158d8b9867d5b9a6"
      ],
      "author": {
        "name": "Pablo Mazzini",
        "email": "pmazzini@gmail.com",
        "time": "Sun Nov 08 09:35:59 2026 +0000"
      },
      "committer": {
        "name": "Guenter Roeck",
        "email": "linux@roeck-us.net",
        "time": "Fri Nov 13 12:08:07 2026 -0800"
      },
      "message": "hwmon: (sy7636a) fix negative temperatures\n\nThe thermistor readout is a signed 8-bit value in degrees Celsius.\n\nFixes: de34a4053250 (\"hwmon: sy7636a: Add temperature driver for sy7636a\")\nCc: stable@vger.kernel.org\nSigned-off-by: Pablo Mazzini \u003cpmazzini@gmail.com\u003e\nLink: https://patch.msgid.link/20261108093602.329554-3-pmazzini@gmail.com\nSigned-off-by: Guenter Roeck \u003clinux@roeck-us.net\u003e\n"
    },
    {
      "commit": "e234ed1bb82c96a5643fd70784f16cf3d72eeed3",
      "tree": "4bda09c25f9793ba5537bc9b61f71b8a8826db77",
      "parents": [
        "c1ad01173511e8aec378af9f5285f98b767b1381"
      ],
      "author": {
        "name": "Ridong Chen",
        "email": "chenridong@xiaomi.com",
        "time": "Thu Nov 12 11:19:31 2026 +0800"
      },
      "committer": {
        "name": "Tejun Heo",
        "email": "tj@kernel.org",
        "time": "Fri Nov 13 09:23:33 2026 -1000"
      },
      "message": "cpuset: generalize the validate_partition() interface\n\nRefactor validate_partition() to take the parent partition, the new\npartition root state, the new effective exclusive CPUs and the CPUs to\nbe deleted, instead of a trial cpuset. A NULL parent stands for a\nremote partition, whose CPUs come from top_cpuset. This makes the\ninterface flexible enough to handle both local and remote partition\nvalidation.\n\nThis also fixes a false PERR_NOCPUS error when expanding a remote\npartition. The no CPU left for tasks check in partition_cpus_change()\nis done against the cgroup parent, but the CPUs of a remote partition\ncome from top_cpuset, not from its parent. For example, on a 4-CPU\nsystem:\n\n  echo 2-3 \u003e P/cpuset.cpus\n  echo 2-3 \u003e P/cpuset.cpus.exclusive\n  echo 2-3 \u003e P/C/cpuset.cpus\n  echo 2 \u003e P/C/cpuset.cpus.exclusive\n  echo root \u003e P/C/cpuset.cpus.partition\n  (a task in P)\n  echo 2-3 \u003e P/C/cpuset.cpus.exclusive\n  cat P/C/cpuset.cpus.partition\n  root invalid (Parent is not a partition root)\n\nThe effective_cpus of P is 3, a subset of the new exclusive CPUs 2-3 of\nC, and P has a task, so the expansion is rejected. However, P is not a\npartition root and does not give any CPU to C, and top_cpuset still has\nCPUs 0-1 for the task. Enabling C directly with exclusive CPUs 2-3\nsucceeds, as remote_partition_enable() checks top_cpuset instead.\n\nThe no CPU left for tasks check now uses the given parent instead of\nparent_cs(cs), and top_cpuset for a remote partition, as that is where\nits CPUs are taken from.\n\nThe housekeeping conflict check is folded into validate_partition(). It\nis skipped when the parent is not a valid partition root, as @cs can\u0027t\nbe a valid local partition in that case.\n\nUse the new interface for local partition enablement in\nupdate_parent_effective_cpumask() and in partition_cpus_change(),\nreplacing the open-coded checks there. The order of the checks is now\nthe same for both paths: empty CPUs, housekeeping conflict, then no CPU\nleft for tasks.\n\nAdd a test case to test_cpuset_prs.sh for the remote partition\nexpansion above.\n\nFixes: 181c8e091aae (\"cgroup/cpuset: Introduce remote partition\")\nSigned-off-by: Ridong Chen \u003cchenridong@xiaomi.com\u003e\nReviewed-by: Guopeng Zhang \u003czhangguopeng@kylinos.cn\u003e\nReviewed-by: Waiman Long \u003clongman@redhat.com\u003e\nSigned-off-by: Tejun Heo \u003ctj@kernel.org\u003e\n"
    },
    {
      "commit": "c1ad01173511e8aec378af9f5285f98b767b1381",
      "tree": "40e2dd32aac51f31dcaed20b0b7eebc51da4631d",
      "parents": [
        "56681bc448a80bb2473026b41c44c133766bc38d"
      ],
      "author": {
        "name": "Ridong Chen",
        "email": "chenridong@xiaomi.com",
        "time": "Thu Nov 12 11:19:30 2026 +0800"
      },
      "committer": {
        "name": "Tejun Heo",
        "email": "tj@kernel.org",
        "time": "Fri Nov 13 09:23:28 2026 -1000"
      },
      "message": "cpuset: settle isolation state when CPUs are given back to parent\n\npartition_xcpus_del() only updates isolated_cpus when the partition that\ngives back the CPUs has a different state than its parent. CPUs that were\nfurther distributed to descendant isolated partitions are not taken into\naccount, so isolated_cpus_release() has to be called afterwards in\nseveral places to walk up the ancestors and fix isolated_cpus.\n\nThis misses some cases and leaks isolated CPUs, as reported by Guopeng.\nAn isolated partition can be invalidated after its parent has already\nbecome invalid, and its CPUs are left in isolated_cpus. For example, on a\n32-CPU system with all CPUs online and without isolcpus\u003d or nohz_full\u003d\nboot parameters:\n\n  cd /sys/fs/cgroup\n  echo +cpuset \u003e cgroup.subtree_control\n  mkdir A\n  echo 0,2-31 \u003e A/cpuset.cpus\n  echo root \u003e A/cpuset.cpus.partition\n  echo +cpuset \u003e A/cgroup.subtree_control\n  mkdir A/B\n  echo 2-3 \u003e A/B/cpuset.cpus\n  echo isolated \u003e A/B/cpuset.cpus.partition\n  cat cpuset.cpus.isolated\n  2-3\n  echo 0 \u003e /sys/devices/system/cpu/cpu1/online\n  cat A/B/cpuset.cpus.partition\n  isolated invalid\n  cat cpuset.cpus.isolated\n  2-3\n\nWhen A is invalidated, partition_xcpus_del() gives CPUs 0,2-31 back to\ntop cpuset. Both A and top cpuset are root partitions, so isolated_cpus\nis left untouched although CPUs 2-3 were distributed to the isolated B.\nB is then invalidated with an invalid parent, and nothing releases CPUs\n2-3 either.\n\nMake partition_xcpus_del() always set the isolation state of the given\nback CPUs to the state of the parent that receives them. As the\neffective_xcpus of a partition covers the CPUs distributed to its\ndescendants, the isolation state of those CPUs is settled at the same\ntime. Boot-time isolated CPUs are still preserved by isolated_cpu_update().\n\ncompute_partition_effective_cpumask() invalidates a child partition\nwithout going through partition_xcpus_del(), so do the same there: the\nCPUs of the invalidated child take the isolation state of its parent.\n\nAs the parent may be shrunk and its effective_xcpus not updated yet,\nonly do so for the CPUs the parent still owns. This is to keep the\nisolation state correct for now, the in-place invalidation here will be\nremoved later in the partition rework.\n\nisolated_cpus_update() assumes the old and new states are different,\nwhich no longer holds now that partition_xcpus_del() proceeds regardless\nof the states. Drop its @old_prs argument and the WARN_ON_ONCE() so that\nit simply sets the isolation state of the given CPUs, and use it\neverywhere. The state check in partition_xcpus_add() is dropped as well,\nas setting the state is a no-op when the states are the same.\n\nisolated_cpus_release() is no longer needed and is removed.\n\nFixes: 11e5f407b64a (\"cgroup/cpuset: Keep track of CPUs in isolated partitions\")\nReported-by: Guopeng Zhang \u003czhangguopeng@kylinos.cn\u003e\nCloses: https://lore.kernel.org/all/20261105130344.664336-4-guopeng.zhang@linux.dev/\nSigned-off-by: Ridong Chen \u003cchenridong@xiaomi.com\u003e\nReviewed-by: Guopeng Zhang \u003czhangguopeng@kylinos.cn\u003e\nReviewed-by: Waiman Long \u003clongman@redhat.com\u003e\nSigned-off-by: Tejun Heo \u003ctj@kernel.org\u003e\n"
    },
    {
      "commit": "2800868f63227c8237d70c41e70d441c919620d6",
      "tree": "12e0c9a0acc4940db21180fd35608c41ef26f328",
      "parents": [
        "e93fd809fb5692d1571bd008988ba734fd7fc976",
        "bca568589ad118a9b0bac086760baee2a41976a5"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Fri Nov 13 10:36:22 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Fri Nov 13 10:36:22 2026 -0800"
      },
      "message": "Merge tag \u0027sound-7.4-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound\n\nPull sound fixes from Takashi Iwai:\n \"A collection of small fixes as usual. A couple of regression fixes are\n  included in addition to the various quirks and device-specific fixes.\n\n  ALSA Core:\n   - PCM: Fix potential inconsistent PCM state reporting after XRUN\n\n  HD-audio:\n   - Fix sysfs hint parser\n   - Revert CS4208 fixup for MacBookAir 7,2 due to regressions\n   - ACPI _DSD fixes for CS35L41 side-codec on various HP laptops\n   - Device quirks for Acer (Predator, TravelLite) and Dell (Pro QC1255,\n     XPS 8940)\n\n  USB-audio:\n   - Fix for USB controller bugs reporting oversized packets\n   - Fix UAF on disconnect in caiaq\n   - Quirks for NiceHCK Octave DAC and RYDZ Technology Piva AC20\n\n  ASoC:\n   - Fix a regression due to the conversion to devm_kzalloc_flex() that\n     initialized the counter unexpectedly\n   - Copy DAI link names into card device to survive component rebind\n   - SOF: Fix pin format memory leak in IPC4 process widget setup,\n   - SOF: Support alternatively signed firmware loading on AMD platforms\n   - AMD: Guard SoundWire context with NULL checks in IRQ handlers\n   - AMD: SoundWire support on LG gram 16Z95U, fix speaker pop noise on\n     ACP3x with ES83xx codecs, quirks for Acer, ASUS, Huawei, and Lenovo\n   - Qualcomm: Fix headset jack at card rebind, PCM position fix on\n     q6asm\n   - Apple: Fix clock selection for clock consumer FEs in MCA driver\n   - Intel: SoundWire jack-detection quirk for Dell Latitude 9420\n   - Codecs: Fixes for CS43130 (power down on close), ES9356 (gain dB\n     scale), and MAX98090 (preserve mic boost)\"\n\n* tag \u0027sound-7.4-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound: (34 commits)\n  ALSA: aw2-alsa: select SND_PCM\n  ALSA: usb-audio: Add native DSD support for NiceHCK Octave\n  Revert \"ALSA: hda/cs420x: Add CS4208 fixup for MacBookAir 7,2\"\n  ALSA: hda/realtek: Update quirk for Acer Predator PHN16-72\n  ALSA: hda/realtek - remove headphone MIC support for Dell Pro QC1255\n  ALSA: caiaq: Kill EP1 URB before unregistering the input device\n  ALSA: pcm: update state after xrun\n  ASoC: max98090: keep the MIC boost across mic input power-up\n  ALSA: hda/realtek: Fix the Dell XPS 8940 headset microphone\n  ASoC: amd: acp70: Add rt713 + rt1320 SoundWire machine for LG gram 16Z95U\n  ASoC: amd: acp: Add ACP70 ACPI flag override for LG gram 16Z95U\n  ASoC: apple: mca: Fix clock selection for clock consumer FEs\n  ASoC: soc-core: reset rtd-\u003enum_components after allocation\n  ASoC: SOF: ipc4-topology: Fix pin format leak in process widget setup\n  ALSA: usb-audio: Drop capture packets with bogus actual_length\n  ALSA: usb-audio: Add iface reset and delay quirk for RYDZ Technology Piva AC20\n  ASoC: amd: acp: Add Lenovo Legion 5 15AGP11 to ACP70 override table\n  ASoC: SOF: amd: Load alternatively signed firmware with an \"_l\" suffix\n  ASoC: es9356: fix the capture gain dB scale\n  ASoC: amd: acp: es83xx: Enable speaker amplifier after the codec output\n  ...\n"
    },
    {
      "commit": "e93fd809fb5692d1571bd008988ba734fd7fc976",
      "tree": "f90dfe2773a029319cef7a5039ed4748d3ff1220",
      "parents": [
        "831615b085fa5b9c85c5a52dadd71ebcfb8d5985",
        "7cc2dda4f7aaf596c1796184649007151ff98f8d"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Fri Nov 13 10:31:05 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Fri Nov 13 10:31:05 2026 -0800"
      },
      "message": "Merge tag \u0027block-7.4-20261112\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux\n\nPull block fixes from Jens Axboe:\n\n - ublk fixes: don\u0027t publish a new device in the idr until it\u0027s fully\n   set up, and don\u0027t hold ublk_ctl_mutex across device removal, as a\n   stuck filesystem on one device otherwise blocked ADD_DEV and DEL_DEV\n   for every other ublk device.\n\n - s390/dasd fixes: re-initialize the RCD buffer magic before every\n   request start, fix a wrong free and refcount leak in the check_tcw\n   error path, don\u0027t trust the userspace-supplied result field in the\n   format check, and a typo fix.\n\n* tag \u0027block-7.4-20261112\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux:\n  s390/dasd: Don\u0027t trust userspace-supplied result field in format check\n  s390/dasd: Fix wrong free and refcount leak in build_check_tcw error path\n  s390/dasd: Re-initialize RCD buffer with \"V1.0\" before every start\n  s390/dasd: Fix typo \u0027immediatly\u0027 -\u003e \u0027immediately\u0027\n  ublk: don\u0027t take ublk_ctl_mutex in DEL_DEV\n  ublk: publish a new device by its id only once it is set up\n"
    },
    {
      "commit": "831615b085fa5b9c85c5a52dadd71ebcfb8d5985",
      "tree": "4a18b61e2d43d40c75fdb977d3e04bf42f33c05d",
      "parents": [
        "3db1c74a20f0552fadc2c6886a43feee8894fa19",
        "39c7210e970200f8b17de3767c73efe64173a95f"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Fri Nov 13 10:20:23 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Fri Nov 13 10:20:23 2026 -0800"
      },
      "message": "Merge tag \u0027io_uring-7.4-20261112\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux\n\nPull io_uring fix from Jens Axboe:\n\n - Don\u0027t end the NAPI busy poll early on need_resched() when a single\n   NAPI is registered, as that left the rest of the wait timeout\n   sleeping in schedule_hrtimeout() rather than busy polling.\n\n* tag \u0027io_uring-7.4-20261112\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux:\n  io_uring/napi: don\u0027t end the single NAPI busy poll early on need_resched()\n"
    },
    {
      "commit": "3db1c74a20f0552fadc2c6886a43feee8894fa19",
      "tree": "20be7a91a108ea58b0854a388d43fc703f6ac4ca",
      "parents": [
        "b970ee43be13fc45851e957469807cac049c10ce",
        "25912ed3b75dcf5ff623dcb549b925dd0d0948f7"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Fri Nov 13 09:56:22 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Fri Nov 13 09:56:22 2026 -0800"
      },
      "message": "Merge tag \u0027cifs-fixes-7.4-rc3\u0027 of https://git.manguebit.org/linux\n\nPull smb client fixes from Paulo Alcantara:\n\n  - Fix silent data loss in writeback\n\n    Since the move to netfslib a transient failure (a reconnect, a lost\n    handle, an interrupted wait, -ENOMEM, or a task killed while close()\n    flushes its dirty pages) ended writeback with the folios clean and\n    no error, dropping the write.\n\n    Retry it instead, bounded by the hard and retrans mount options, and\n    have umount -f abort the retries and the reconnect and credit waits\n    so the unmount\u0027s own sync can\u0027t hang\n\n  - Prefer a non-O_DIRECT handle for writeback, as Windows rejects\n    unaligned I/O on O_DIRECT handles and silently dropped about 1% of\n    the writes in generic/051\n\n  - netfs: keep the prepare_read() and prepare_write() errno across a\n    reissue, and don\u0027t hang a read whose retry is abandoned (the\n    writeback fixes depend on these)\n\n  - Fix an out-of-bounds write when compounding an encrypted SMB2_IOCTL\n    with a large input, which an unprivileged user can trigger with a\n    long symlink target\n\n  - Fix an out-of-bounds read parsing a DFS target referral, and a\n    double conversion of the DFS path in setup_dfs_ref()\n\n  - Validate security descriptor headers before allocating the query\n    result, and compare owner and group SIDs by their encoded lengths\n\n  - Hold a server reference while queueing reconnect work so the server\n    can\u0027t be freed under the worker\n\n  - Hold the inode across cifs_oplock_break() so dropping the handle\n    reference can\u0027t free it before the break is acknowledged\n\n  - Flush an inode\u0027s deferred handles before opening another of its hard\n    links, so the open doesn\u0027t reuse the lease key under a mismatched\n    name and get rejected with STATUS_INVALID_PARAMETER\n\n  - Skip a NULL channel in __cifs_put_smb_ses()\n\n  - Bound the invalid NTLMSSP signature diagnostic\n\n* tag \u0027cifs-fixes-7.4-rc3\u0027 of https://git.manguebit.org/linux: (29 commits)\n  smb: client: flush deferred handles of other names before a hard link open\n  smb: client: hold the inode across cifs_oplock_break()\n  smb: client: stop reconnecting and replaying during a forced unmount\n  smb: client: fail writeback sends and reopens fast when the server is gone\n  smb: client: don\u0027t replay handle reopens inside SMB2_open()\n  smb: client: let umount -f stop waiting for reconnects on hard mounts\n  smb: client: retry writeback after losing the connection, share or handle\n  smb: client: retry writes that couldn\u0027t be issued due to a reconnect\n  smb: client: don\u0027t block dying tasks waiting for writeback on fsync\n  smb: client: retry writeback on interrupts and allocation failures\n  smb: client: don\u0027t fail writeback preparation if the task is dying\n  smb: client: issue writes from a worker if the task is dying\n  smb: client: don\u0027t block dying tasks waiting for writeback on close\n  smb: client: don\u0027t spend read retries on reads cut short by a reconnect\n  smb: client: don\u0027t retry failed writes forever regardless of mount options\n  smb: client: don\u0027t give up writeback on the first replayable error\n  smb: client: let umount -f abort retried reads and writes\n  netfs: don\u0027t hang a read whose retry is abandoned\n  netfs: keep the prepare_read errno when abandoning read retries\n  netfs: preserve prepare_write errno on reissue\n  ...\n"
    },
    {
      "commit": "b970ee43be13fc45851e957469807cac049c10ce",
      "tree": "c8a23127f28c0b9987b781fbe82b412fada31b07",
      "parents": [
        "7e1392a42ea288d70a266f6553fe06f57f8647ea",
        "7906eac042a4d6facd3b2dcebfe7179fedfcf1f2"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Fri Nov 13 09:48:11 2026 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@linux-foundation.org",
        "time": "Fri Nov 13 09:48:11 2026 -0800"
      },
      "message": "Merge tag \u0027erofs-for-7.4-rc3-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/xiang/erofs\n\nPull erofs fixes from Gao Xiang:\n\n - Fix stacked fs check for file-backed mounts with extra devices\n\n - Disallow file-backed mounts on shared backing inodes\n\n - Reset metadata offsets when fsoffset\u003d is used to fix xattr reads\n   inside and outside the metabox in ont shot\n\n - Refine on-disk format summary\n\n* tag \u0027erofs-for-7.4-rc3-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/xiang/erofs:\n  docs: erofs: refine the on-disk overview\n  erofs: reset buf-\u003eoff in erofs_init_metabuf()\n  erofs: treat files shared by inode_share as stacked\n  erofs: apply the stacked fs check to extra file-backed devices\n"
    },
    {
      "commit": "ad3fc05c7e65009355effeea06aa629a687acef2",
      "tree": "46d777110fcb557d1cde0b1841253b8b7a52898e",
      "parents": [
        "5306f124f3171289b2ef5bac1e08aad175f66d1e"
      ],
      "author": {
        "name": "Uladzislau Zhauniarovich",
        "email": "uladzislau.zhauniarovich@gmail.com",
        "time": "Thu Nov 05 20:19:45 2026 +0000"
      },
      "committer": {
        "name": "Catalin Marinas",
        "email": "catalin.marinas@arm.com",
        "time": "Fri Nov 13 17:12:34 2026 +0000"
      },
      "message": "arm64: suspend: Only track cpuidle context for idle tasks\n\nCommit 19235e472798 (\"cpuidle, arm64: Fix the ARM64 cpuidle logic\") moved\nthe calls to ct_cpuidle_enter() and ct_cpuidle_exit() into cpu_suspend() so\nthat arm64 cpuidle could keep RCU watching longer and manage context\ntracking internally. However, cpu_suspend() is called not only by cpuidle\nvia psci_cpu_suspend_enter() (which runs in the context of the idle task),\nbut also during system suspend via psci_system_suspend_enter() (which runs\nin the context of the task initiating suspend, such as a process writing to\n/sys/power/state).\n\nWhen entering system suspend, current is a non-idle task. Calling\nct_cpuidle_enter() invokes ct_kernel_exit() from kernel space (!user) on a\nnon-idle task. With CONFIG_RCU_EQS_DEBUG enabled, this violates context\ntracking invariants and triggers a warning:\n\nWARNING: kernel/context_tracking.c:120 at ct_kernel_exit+0x1b4/0x1f4\nkernel/context_tracking.c:118\nCall trace:\n ct_kernel_exit+0x1b4/0x1f4 kernel/context_tracking.c:118\n ct_idle_enter+0x20/0x34 kernel/context_tracking.c:299\n ct_cpuidle_enter include/linux/cpuidle.h:134 [inline]\n cpu_suspend+0x264/0x474 arch/arm64/kernel/suspend.c:136\n psci_system_suspend_enter+0x4c/0x80 drivers/firmware/psci/psci.c:544\n suspend_enter kernel/power/suspend.c:468 [inline]\n suspend_devices_and_enter+0x88c/0xd0c kernel/power/suspend.c:533\n enter_state kernel/power/suspend.c:618 [inline]\n pm_suspend+0x498/0x698 kernel/power/suspend.c:644\n state_store+0x14c/0x1fc kernel/power/main.c:819\n kobj_attr_store+0x6c/0x90 lib/kobject.c:842\n sysfs_kf_write+0xe0/0x108 fs/sysfs/file.c:145\n kernfs_fop_write_iter+0x260/0x3b8 fs/kernfs/file.c:345\n vfs_write+0x51c/0xa0c fs/read_write.c:687\n ksys_write+0x12c/0x228 fs/read_write.c:739\n\nA corresponding warning is subsequently triggered in ct_kernel_enter() via\nct_cpuidle_exit() upon resume.\n\nFix this by checking is_idle_task(current) in cpu_suspend() and only\ninvoking ct_cpuidle_enter() and ct_cpuidle_exit() when executing in the\ncontext of an idle task.\n\nFixes: 19235e472798 (\"cpuidle, arm64: Fix the ARM64 cpuidle logic\")\nCc: \u003cstable@vger.kernel.org\u003e\nAssisted-by: Gemini:gemini-3.8-flash Gemini:gemini-3.1-pro-preview syzbot\nReported-by: syzbot+79883955da3a62652a80@syzkaller.appspotmail.com\nCloses: https://syzkaller.appspot.com/bug?extid\u003d79883955da3a62652a80\nLink: https://syzkaller.appspot.com/ai_job?id\u003d5359e16f-013b-4850-9cbc-ddfb3bac95e5\nSigned-off-by: Uladzislau Zhauniarovich \u003culadzislau.zhauniarovich@gmail.com\u003e\nTested-by: Uladzislau Zhauniarovich \u003culadzislau.zhauniarovich@gmail.com\u003e\nLink: https://lore.kernel.org/r/3e92ccb0-2952-4317-af4c-45457c15efae@gmail.com\nSigned-off-by: Catalin Marinas \u003ccatalin.marinas@arm.com\u003e\n"
    },
    {
      "commit": "2a490521d74ac76e5eb3c9a69d928a4c76dd3f78",
      "tree": "66f8fdd9dc44ecf31f98b20639b88f022bf0ab14",
      "parents": [
        "9060f5ba658c9b0a7332be55b3d8abfc2b08618b"
      ],
      "author": {
        "name": "Kees Cook",
        "email": "kees@kernel.org",
        "time": "Thu Nov 12 22:55:12 2026 -0800"
      },
      "committer": {
        "name": "Mark Brown",
        "email": "broonie@kernel.org",
        "time": "Fri Nov 13 16:26:07 2026 +0000"
      },
      "message": "spi: axi-spi-engine: Start the program length at zero\n\nspi_engine_optimize_message() allocates the program with kzalloc_flex(),\nwhich sets the __counted_by() counter \"length\" to the allocated size when\nthe compiler supports __builtin_counted_by_ref(). The program is then\nfilled by spi_engine_program_add_cmd(), which increments \"length\" and\nstores at length - 1, expecting it to start at zero as it did when the\nprogram came from kzalloc().\n\nReset \"length\" after the allocation, since it is a \"usage\" counter, not\na \"total allocation\" counter.\n\nBuild tested ARCH\u003dx86_64 defconfig with CONFIG_SPI_AXI_SPI_ENGINE\u003dm and\nW\u003d1, with GCC 16.2.0.\n\nReported-by: Dennis Gilmore \u003cdennis@ausil.us\u003e\nCloses: https://lore.kernel.org/all/20261112025252.71375-1-dennis@ausil.us/\nFixes: 69050f8d6d075 (\"treewide: Replace kmalloc with kmalloc_obj for non-scalar types\")\nCc: stable@vger.kernel.org\nSigned-off-by: Kees Cook \u003ckees@kernel.org\u003e\nReviewed-by: David Lechner \u003cdlechner@baylibre.com\u003e\nLink: https://patch.msgid.link/20261113065508.i.606-kees@kernel.org\nSigned-off-by: Mark Brown \u003cbroonie@kernel.org\u003e\n"
    },
    {
      "commit": "5306f124f3171289b2ef5bac1e08aad175f66d1e",
      "tree": "b96bb59013a8965770aa709409a401b976c05566",
      "parents": [
        "18f2d01bfce99a885511c451cba95234eea190bc"
      ],
      "author": {
        "name": "Zhao Gongyi",
        "email": "zhaogongyi@bytedance.com",
        "time": "Thu Nov 05 19:30:51 2026 +0800"
      },
      "committer": {
        "name": "Catalin Marinas",
        "email": "catalin.marinas@arm.com",
        "time": "Fri Nov 13 14:10:36 2026 +0000"
      },
      "message": "arm64: ptrace: don\u0027t warn on missing siginfo in a group-stop\n\nPTRACE_ATTACH to a group-stopped child can enter do_jobctl_trap(),\nwhich calls ptrace_stop() without siginfo. If the tracer writes\nNT_ARM_SYSTEM_CALL, system_call_set() can change syscallno from\nNO_SYSCALL before update_syscall_orig_x0_after_ptrace() checks\nlast_siginfo. WARN_ON_ONCE(!info) then fires for a valid ptrace stop.\n\nKeep the early return when siginfo is absent, but remove the warning.\nThis leaves orig_x0 untouched at a non-seized group-stop and preserves\nthe existing syscall-exit-stop handling.\n\nFixes: 88b839ce497c (\"arm64: ptrace: Keep \u0027orig_x0\u0027 in-sync with x0 on syscall entry\")\nCc: stable@vger.kernel.org\nSigned-off-by: Zhao Gongyi \u003czhaogongyi@bytedance.com\u003e\nAcked-by: Oleg Nesterov \u003coleg@redhat.com\u003e\nAcked-by: Will Deacon \u003cwill@kernel.org\u003e\n[catalin.marinas@arm.com: modified comment as per Will\u0027s suggestion]\nSigned-off-by: Catalin Marinas \u003ccatalin.marinas@arm.com\u003e\n"
    },
    {
      "commit": "c15120853eebbda98f3fe1f1e59370439c6c4447",
      "tree": "dfecb9cede47ba1bf80b4b7875e48fe79c279c5b",
      "parents": [
        "0941144ebeaf3ce46c80711f00e935921eb28b21"
      ],
      "author": {
        "name": "Masami Hiramatsu (Google)",
        "email": "mhiramat@kernel.org",
        "time": "Fri Nov 13 17:29:22 2026 +0900"
      },
      "committer": {
        "name": "Masami Hiramatsu (Google)",
        "email": "mhiramat@kernel.org",
        "time": "Fri Nov 13 23:06:31 2026 +0900"
      },
      "message": "tracing/uprobes: Fix memory leak of tail in __trace_uprobe_create()\n\nWhen generating a default event name in __trace_uprobe_create(), tail is\nallocated with kstrdup() without __free(kfree). If the subsequent\nkmalloc() for buf fails, the function returns -ENOMEM immediately and\nleaks tail.\n\nDeclare tail with __free(kfree) so it is automatically freed on all\nexit paths.\n\nLink: https://lore.kernel.org/all/179455856197.403392.16569431894958507645.stgit@devnote2/\n\nReported-by: Sashiko \u003csashiko-bot@kernel.org\u003e\nCloses: https://lore.kernel.org/all/sashiko-outbox-204240@kernel.org/\nFixes: 97e8230f89a3 (\"tracing: uprobe-event: Allocate string buffers from heap\")\nCc: stable@vger.kernel.org\nSigned-off-by: Masami Hiramatsu (Google) \u003cmhiramat@kernel.org\u003e"
    },
    {
      "commit": "0941144ebeaf3ce46c80711f00e935921eb28b21",
      "tree": "ae71454993967e9a6b2d86fc8817cf9a5ebd6fa7",
      "parents": [
        "2721573e6c6faaba2e31f352f2c02ec5f3ecf5e9"
      ],
      "author": {
        "name": "Masami Hiramatsu (Google)",
        "email": "mhiramat@kernel.org",
        "time": "Fri Nov 13 17:29:13 2026 +0900"
      },
      "committer": {
        "name": "Masami Hiramatsu (Google)",
        "email": "mhiramat@kernel.org",
        "time": "Fri Nov 13 23:06:23 2026 +0900"
      },
      "message": "tracing/probes: Reject array count on non-dereference string sources\n\nThe runtime array loop (FETCH_OP_LP_ARRAY) for string types requires a\npreceding memory dereference instruction so it can advance the pointer\naddress on each iteration. Non-dereference string sources such as $comm,\nimmediate strings, and eprobe event fields cannot be indexed as pointer\narrays, but finalize_fetch_insn() still accepts an array count on them\nand emits an invalid FETCH_OP_LP_ARRAY sequence.\n\nReject string array counts when the preceding operation is not a memory\ndereference.\n\nLink: https://lore.kernel.org/all/179455855309.403392.4197028577158545202.stgit@devnote2/\n\nFixes: 40b53b771806 (\"tracing: probeevent: Add array type support\")\nCc: stable@vger.kernel.org\nSigned-off-by: Masami Hiramatsu (Google) \u003cmhiramat@kernel.org\u003e"
    },
    {
      "commit": "2721573e6c6faaba2e31f352f2c02ec5f3ecf5e9",
      "tree": "83fb1c9956044d9ca930a163b867c6cd3b5541f8",
      "parents": [
        "92d9fa0ed69e1f724feed908bd0a462a42df4b88"
      ],
      "author": {
        "name": "Masami Hiramatsu (Google)",
        "email": "mhiramat@kernel.org",
        "time": "Fri Nov 13 17:29:04 2026 +0900"
      },
      "committer": {
        "name": "Masami Hiramatsu (Google)",
        "email": "mhiramat@kernel.org",
        "time": "Fri Nov 13 23:05:56 2026 +0900"
      },
      "message": "tracing/probes: Fix user-space dereference string array store type\n\nWhen finalizing a string array argument (e.g. +0u(%reg):string[N]),\nfinalize_fetch_insn() advances the instruction pointer to a new slot\nfor the store operation before checking whether the preceding operation\nwas FETCH_OP_UDEREF. As a result, user-space dereference string arrays\nwith type \"string\" emit FETCH_OP_ST_STRING (kernel memory read) instead\nof FETCH_OP_ST_USTRING.\n\nCheck for FETCH_OP_UDEREF before advancing the instruction pointer.\n\nLink: https://lore.kernel.org/all/179455854396.403392.18073771811100192405.stgit@devnote2/\n\nFixes: e65f7ae7f4da (\"tracing/probe: Support user-space dereference\")\nCc: stable@vger.kernel.org\nSigned-off-by: Masami Hiramatsu (Google) \u003cmhiramat@kernel.org\u003e"
    },
    {
      "commit": "92d9fa0ed69e1f724feed908bd0a462a42df4b88",
      "tree": "0ed71c7cbc785e81adfa91224218666102a732eb",
      "parents": [
        "a0f836eea463d4c468edb58eedb03911c9c0791d"
      ],
      "author": {
        "name": "Masami Hiramatsu (Google)",
        "email": "mhiramat@kernel.org",
        "time": "Fri Nov 13 17:28:54 2026 +0900"
      },
      "committer": {
        "name": "Masami Hiramatsu (Google)",
        "email": "mhiramat@kernel.org",
        "time": "Fri Nov 13 23:05:46 2026 +0900"
      },
      "message": "tracing/eprobe: Disallow attaching eprobes to fprobe events\n\nEvent probes (eprobes) are not allowed to attach to other dynamic probe\nevents (kprobes, uprobes, and eprobes) to avoid recursive probe chains\nand lifecycle dependencies. However, when fprobe events were introduced,\nTRACE_EVENT_FL_FPROBE was omitted from the exclusion mask in\nfind_and_get_event(), allowing eprobes to attach to fprobe and tprobe\nevents.\n\nExclude TRACE_EVENT_FL_FPROBE events in find_and_get_event().\n\nLink: https://lore.kernel.org/all/179455853457.403392.14549390017750529531.stgit@devnote2/\n\nFixes: 334e5519c375 (\"tracing/probes: Add fprobe events for tracing function entry and exit.\")\nCc: stable@vger.kernel.org\nSigned-off-by: Masami Hiramatsu (Google) \u003cmhiramat@kernel.org\u003e"
    },
    {
      "commit": "a0f836eea463d4c468edb58eedb03911c9c0791d",
      "tree": "d831c34058c5f39e419ae2af7f1f728b010984dd",
      "parents": [
        "c1e2c63e65421068566db3dde44dfd4fe7adb46d"
      ],
      "author": {
        "name": "Masami Hiramatsu (Google)",
        "email": "mhiramat@kernel.org",
        "time": "Fri Nov 13 17:28:45 2026 +0900"
      },
      "committer": {
        "name": "Masami Hiramatsu (Google)",
        "email": "mhiramat@kernel.org",
        "time": "Fri Nov 13 23:05:30 2026 +0900"
      },
      "message": "tracing/eprobe: Fix 1-byte signed field fetch in get_event_field()\n\nBecause the kernel is built with -funsigned-char, plain \u0027char\u0027 is\nunsigned on all architectures. Reading 1-byte signed trace event fields\nvia *(char *)addr in get_event_field() therefore zero-extends instead of\nsign-extending negative s8 values.\n\nUse explicit fixed-width integer types (s8/u8, s16/u16, s32/u32) when\nreading event fields.\n\nLink: https://lore.kernel.org/all/179455852542.403392.15666310371702309464.stgit@devnote2/\n\nFixes: 7491e2c44278 (\"tracing: Add a probe that attaches to trace events\")\nCc: stable@vger.kernel.org\nSigned-off-by: Masami Hiramatsu (Google) \u003cmhiramat@kernel.org\u003e"
    },
    {
      "commit": "c1e2c63e65421068566db3dde44dfd4fe7adb46d",
      "tree": "1fc745b93e24da480f57613d0c99fb61c36cb54c",
      "parents": [
        "9b6e9d6a11090ac34516c45f48663b37ac94a9b1"
      ],
      "author": {
        "name": "Masami Hiramatsu (Google)",
        "email": "mhiramat@kernel.org",
        "time": "Fri Nov 13 17:28:36 2026 +0900"
      },
      "committer": {
        "name": "Masami Hiramatsu (Google)",
        "email": "mhiramat@kernel.org",
        "time": "Fri Nov 13 23:04:56 2026 +0900"
      },
      "message": "tracing/probes: Fix missing BTF kind checks in char pointer/array helpers\n\nIn struct btf_type, the type ID and byte size share a union, and only\nBTF_KIND_INT types are followed by a 32-bit integer encoding word.\nBecause btf_type_is_char_ptr() and btf_type_is_char_array() do not\nverify that the outer type is BTF_KIND_PTR or that the resolved element\ntype is a non-NULL BTF_KIND_INT before inspecting it, non-pointer BTF\ntypes can have their byte size misinterpreted as a BTF type ID, and\narrays of non-integer types read out-of-bounds or misinterpret trailing\nBTF metadata as BTF_INT_* encoding bits.\n\nValidate the BTF type kinds and NULL pointers before reading inner type\nmetadata in both helpers.\n\nLink: https://lore.kernel.org/all/179455851632.403392.258264246782960594.stgit@devnote2/\n\nFixes: 27973e5c64b9 (\"tracing/probes: Add string type check with BTF\")\nCc: stable@vger.kernel.org\nSigned-off-by: Masami Hiramatsu (Google) \u003cmhiramat@kernel.org\u003e"
    },
    {
      "commit": "9b6e9d6a11090ac34516c45f48663b37ac94a9b1",
      "tree": "1638ddf774e75ae5d34fa0df26e19d15ef28eb8e",
      "parents": [
        "72d3a5cb0cd7335553549aaffecc683e3ecc0c43"
      ],
      "author": {
        "name": "Masami Hiramatsu (Google)",
        "email": "mhiramat@kernel.org",
        "time": "Fri Nov 13 17:28:27 2026 +0900"
      },
      "committer": {
        "name": "Masami Hiramatsu (Google)",
        "email": "mhiramat@kernel.org",
        "time": "Fri Nov 13 23:04:48 2026 +0900"
      },
      "message": "tracing/probes: Fix BTF bitfield parsing and out-of-bounds buffer write\n\nBTF bitfield parsing in trace_probe.c assumes a 64-bit (8-byte) container\nwhen computing FETCH_OP_MOD_BF shift amounts and basesize, whereas the\npreceding store instruction only allocates and writes the field\u0027s fetch\ntype size (which may be 1, 2, or 4 bytes when kflag\u003d1). At runtime,\nfetch_apply_bitfield() performs an 8-byte read-modify-write on the\nsmaller ring buffer slot, corrupting adjacent ring buffer data or writing\npast the end of the event record.\n\nIn addition, signed bitfields with bit widths other than 8/16/32/64 fail\nto resolve a fetch type, bitfields whose width is a multiple of 8 inside\na larger integer or spanning a byte boundary fail to emit FETCH_OP_MOD_BF\nor select a container wide enough for bitoffs + bitsize, kflag\u003d0 bitfield\noffsets \u003e\u003d 8 bits are not folded into the byte offset, and big-endian\narchitectures use the little-endian left-shift formula in\nparse_btf_bitfield().\n\nCompute the bitfield byte and bit offsets consistently during field\nlookup, select the smallest unsigned fetch type that covers the bit\noffset and bit width, and calculate the FETCH_OP_MOD_BF parameters from\nthe selected fetch type size for both big- and little-endian bitfields.\n\nLink: https://lore.kernel.org/all/179455850711.403392.10512730891220678403.stgit@devnote2/\n\nFixes: c440adfbe302 (\"tracing/probes: Support BTF based data structure field access\")\nCc: stable@vger.kernel.org\nSigned-off-by: Masami Hiramatsu (Google) \u003cmhiramat@kernel.org\u003e"
    },
    {
      "commit": "72d3a5cb0cd7335553549aaffecc683e3ecc0c43",
      "tree": "0f712f73b993ef1f132cb2465a2c2c1998f4b540",
      "parents": [
        "f1d9b7a3fe99b4228acdbab3c744c29f7ffe86ef"
      ],
      "author": {
        "name": "Masami Hiramatsu (Google)",
        "email": "mhiramat@kernel.org",
        "time": "Fri Nov 13 17:28:18 2026 +0900"
      },
      "committer": {
        "name": "Masami Hiramatsu (Google)",
        "email": "mhiramat@kernel.org",
        "time": "Fri Nov 13 23:04:40 2026 +0900"
      },
      "message": "tracing/probes: Enable newly appended probe when target event is enabled\n\nUnlike kprobe events, fprobe and uprobe events only register their\nunderlying probes when the trace event is enabled. When a new probe is\nappended to an existing multiprobe fprobe or uprobe event that is\nalready enabled, the appended probe is linked into the event\u0027s probe\nlist without being registered.\n\nAs a result, the newly appended probe never fires until the event is\ndisabled and re-enabled. Moreover, disabling an fprobe event calls\nunregister_fprobe() on the unregistered fprobe, and opening or closing a\nperf event on an enabled uprobe event dereferences a NULL tu-\u003euprobe in\nuprobe_apply().\n\nRegister and enable the newly appended fprobe or uprobe immediately when\nthe target event is already enabled.\n\nLink: https://lore.kernel.org/all/179455849798.403392.135678786198324285.stgit@devnote2/\n\nFixes: 41af3cf587f4 (\"tracing/uprobe: Add multi-probe per uprobe event support\")\nFixes: 2db832ec9090 (\"tracing: fprobe-events: Register fprobe-events only when it is enabled\")\nCc: stable@vger.kernel.org\nSigned-off-by: Masami Hiramatsu (Google) \u003cmhiramat@kernel.org\u003e"
    },
    {
      "commit": "f1d9b7a3fe99b4228acdbab3c744c29f7ffe86ef",
      "tree": "e5b0609d1681c4761f4cbc06e07a57c2823e01eb",
      "parents": [
        "c55722233c041865d1477a58adbc52540bfad017"
      ],
      "author": {
        "name": "Masami Hiramatsu (Google)",
        "email": "mhiramat@kernel.org",
        "time": "Fri Nov 13 17:28:09 2026 +0900"
      },
      "committer": {
        "name": "Masami Hiramatsu (Google)",
        "email": "mhiramat@kernel.org",
        "time": "Fri Nov 13 23:04:31 2026 +0900"
      },
      "message": "tracing/probes: Check enabled and busy state before unlinking sibling probes\n\nWhen a multiprobe event is enabled or referenced (for example, by an\neprobe, trigger, or open tracefs file) and a user attempts to delete the\nevent or one of its probes via dynamic_events, checking\ntrace_probe_has_sibling() before the enabled and busy checks allows the\nfirst probe to be unlinked and freed while the event is still active,\nonly returning -EBUSY on the last sibling probe. For uprobe events,\nunlinking and freeing an enabled sibling probe without unregistering its\nuprobe_consumer also causes a use-after-free.\n\nEnsure all four probe types (kprobe, fprobe, uprobe, and eprobe) verify\nthat the event is neither enabled nor busy before unlinking any sibling\nprobe.\n\nLink: https://lore.kernel.org/all/179455848898.403392.17675270675769076596.stgit@devnote2/\n\nFixes: ca89bc071d5e (\"tracing/kprobe: Add multi-probe per event support\")\nFixes: 41af3cf587f4 (\"tracing/uprobe: Add multi-probe per uprobe event support\")\nFixes: 7491e2c44278 (\"tracing: Add a probe that attaches to trace events\")\nFixes: 334e5519c375 (\"tracing/probes: Add fprobe events for tracing function entry and exit.\")\nCc: stable@vger.kernel.org\nSigned-off-by: Masami Hiramatsu (Google) \u003cmhiramat@kernel.org\u003e"
    },
    {
      "commit": "2b66f2cc753243a6adb3e9a5c038b0fc59ba7ec9",
      "tree": "5b60a39e0438804096028554c5cc975eb0e37889",
      "parents": [
        "4ebb3b8d129cbb419a55924f6a75f8f58435e37b"
      ],
      "author": {
        "name": "Jan Beulich",
        "email": "jbeulich@suse.com",
        "time": "Tue Nov 10 10:56:05 2026 +0100"
      },
      "committer": {
        "name": "Juergen Gross",
        "email": "jgross@suse.com",
        "time": "Fri Nov 13 12:10:17 2026 +0100"
      },
      "message": "Xen/gnttab: improve error handling in gnttab_dma_{alloc,free}_pages()\n\nBoth decrease-reservation and populate-physmap can succeed partially. For\nthe former this may be pretty unlikely, but clearly the latter can hit\nboth out-of-memory conditions in the hypervisor or denial because of the\nallocation exceeding the domain\u0027s allowance (there\u0027s no interaction with\nthe balloon driver here either).\n\nIn gnttab_dma_free_pages() we simply can\u0027t give back to the system what\nhasn\u0027t been re-filled with backing memory. For gnttab_dma_alloc_pages()\nboth parts of the overall region need dealing with differently.\n\nWhile no present caller of gnttab_dma_free_pages() checks its return\nvalue, also don\u0027t use -EFAULT there: It really is an out-of-memory\ncondition. (In gnttab_dma_alloc_pages() -EFAULT doesn\u0027t look quite right\neither, but I couldn\u0027t think of a clearly better error code there.)\n\nFixes: 9bdc7304f536 (\"xen/grant-table: Allow allocating buffers suitable for DMA\")\nSigned-off-by: Jan Beulich \u003cjbeulich@suse.com\u003e\nReviewed-by: Jan Beulich \u003cjbeulich@suse.com\u003e\nSigned-off-by: Juergen Gross \u003cjgross@suse.com\u003e\nMessage-ID: \u003c20261110095605.369848-3-jgross@suse.com\u003e\n"
    }
  ],
  "next": "4ebb3b8d129cbb419a55924f6a75f8f58435e37b"
}
