stable.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH 6.6.y 00/19] io_uring fixes for 6.6-stable
@ 2026-11-21 21:30 Jens Axboe
  2026-11-21 21:30 ` [PATCH 6.6.y 01/19] io_uring: enable toggle of iowait usage when waiting on CQEs Jens Axboe
                   ` (18 more replies)
  0 siblings, 19 replies; 20+ messages in thread
From: Jens Axboe @ 2026-11-21 21:30 UTC (permalink / raw)
  To: stable

Hi Greg,

Here are io_uring fixes for 6.6-stable. All 19 patches are in Linus's
tree. Patches 2-5 came with the io_uring-7.4 pull, merged as commit
78d9aedf2342 (tag io_uring-7.4-20261119); the others were merged
earlier.

They apply on top of v6.6.162 plus queue-6.6 at stable-queue commit
3dd4b2d3c1f7. They do not apply to v6.6.162 alone: patches 1-7 do, but
8/19 depends on the queued io_uring sqpoll patches.

Patches 1-8, 11, 15 and 18 needed changes for 6.6; the [axboe: ...] note
at the end of each says what was changed and why. The rest are clean
backports.

Please take patches 6-19 together: they are the SQPOLL private task_work
handling from 6.9 (8/19, with 6/19 and 7/19 as prep) and the fixes that
followed it upstream, whose Fixes: tags name it. Patches 1-5 are
independent of that group.

Patch 1/19 (io_uring: enable toggle of iowait usage when waiting on
CQEs) is a feature, not a fix. It is in stable because the QEMU
developers asked for it: QEMU uses IORING_ENTER_NO_IOWAIT and wants it
supported on all current stable kernels. Nothing changes unless the flag
is set.

Testing: ran the liburing-2.15-75-ga10946dc test suite in KVM guests
(PROVE_LOCKING on) on the baseline v6.6.162 kernel and on the patched
kernel, which is v6.6.162 plus these patches and the other io_uring
commits already queued. The patched kernel failed the same tests as the
baseline, except connect-backlog, which failed only on the baseline (a
flaky test). The failures that remain are on both kernels (napi-test
timeout, send-zerocopy, timestamp and similar). No new dmesg splat
(WARNING, BUG, lockdep, KASAN, Oops). io_uring/ builds without warnings
on the release plus queue plus patches.

CVE assessment, by the io_uring maintainer:

Should not get a CVE:
  1/19 io_uring: enable toggle of iowait usage when waiting on CQEs
      Feature, not a fix. IORING_ENTER_NO_IOWAIT only skips setting
      in_iowait; nothing changes unless the flag is set.
  2/19 io_uring: fail soft links on negative IOU_COMPLETE results
      io_issue_sqe() now calls req_set_fail() for negative IOU_COMPLETE
      results, so IOSQE_IO_LINK dependents of a failed request are
      cancelled. This only changes link semantics of the caller's own
      requests.
  3/19 io_uring/rw: fail soft links on IOPOLL short or error IO
      The IOPOLL completion path now calls req_set_fail() when the
      result changed, so linked requests are cancelled after a short or
      failed read/write. Only the caller's own link chain is affected.
  4/19 io_uring/timeout: don't rearm a multishot timeout on cancel
      A cancelled multishot timeout was re-armed from the cancel task
      work. A zero interval plus moving the CQ head far ahead only
      delays teardown of the caller's own ring; the loop goes through
      hrtimer and workqueue, so it reschedules and holds no spinlock.
  5/19 io_uring/rsrc: keep the old file when a file update fails
      The old file was dropped from the registered slot before the new
      fd was checked, so a failed update left the slot empty and posted
      its tag CQE. Only the caller's own file table, no refcount or
      lifetime problem.
  6/19 io_uring: cleanup handle_tw_list() calling convention
      prep for 8/19 (handle_tw_list() calling convention).
  7/19 io_uring: pass in counter to handle_tw_list() rather than return
      it
      prep for 8/19 (handle_tw_list() counter).
  8/19 io_uring/sqpoll: manage task_work privately
      Rework, not a fix: SQPOLL runs its own task_work list. Prep for
      the follow-up fixes 9-19, which are fixes for this patch.
  9/19 io_uring: wake SQPOLL task when task_work is added to an empty
      queue
      Fixes a stall introduced by 8/19; 6.6 releases don't have it.
      Missed wakeup.
  10/19 io_uring/sqpoll: use the correct check for pending task_work
      Fixes 8/19: wrong pending check, a stall. Not present in any 6.6
      release.
  11/19 io_uring/sqpoll: early exit thread if task_context wasn't
      allocated
      Fixes 8/19 (a missing tctx early exit). Part of the same rework,
      no 6.6 release has the bug.
  12/19 io_uring/sqpoll: ensure that normal task_work is also run timely
      Fixes 8/19: normal task_work not run timely, a latency bug.
  13/19 io_uring: signal SQPOLL task_work with TWA_SIGNAL_NO_IPI
      Fixes 8/19: signal method for SQPOLL task_work, a latency bug.
  14/19 io_uring/sqpoll: ensure task state is TASK_RUNNING when running
      task_work
      Fixes a might_sleep WARN that 8/19 introduces; the old 6.6 code
      ran task_work through io_run_task_work() and never had it. Not in
      any 6.6 release.
  15/19 io_uring: don't touch sqd->thread off tw add
      Fixes 9/19: a theoretical sqd->thread reload, the author says
      unlikely.
  16/19 io_uring/sqpoll: zero sqd->thread on tctx errors
      KASAN UAF in fdinfo, but only with fault injection and only on the
      early exit added by 11/19; no 6.6 release has it.
  17/19 io_uring/sqpoll: Increase task_work submission batch size
      Batch size tuning for 8/19, performance.
  18/19 io_uring: fix task_work add use-after-free with SQPOLL MIME-
      Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-
      Transfer-Encoding: 8bit
      KASAN UAF in the task_work add, but only through the private
      SQPOLL list added by 8/19; no 6.6 release has it.
  19/19 io_uring/sqpoll: do the return-to-userspace work in the SQPOLL
      thread
      Fixes 8/19: memory.high reclaim not done in the SQPOLL thread,
      accounting/throttling only.

Cheng Yongkang (2):
  io_uring/timeout: don't rearm a multishot timeout on cancel
  io_uring/rsrc: keep the old file when a file update fails

Gabriel Krisman Bertazi (1):
  io_uring/sqpoll: Increase task_work submission batch size

Jens Axboe (11):
  io_uring: enable toggle of iowait usage when waiting on CQEs
  io_uring: cleanup handle_tw_list() calling convention
  io_uring: pass in counter to handle_tw_list() rather than return it
  io_uring/sqpoll: manage task_work privately
  io_uring: wake SQPOLL task when task_work is added to an empty queue
  io_uring/sqpoll: use the correct check for pending task_work
  io_uring/sqpoll: early exit thread if task_context wasn't allocated
  io_uring/sqpoll: ensure that normal task_work is also run timely
  io_uring: signal SQPOLL task_work with TWA_SIGNAL_NO_IPI
  io_uring/sqpoll: ensure task state is TASK_RUNNING when running
    task_work
  io_uring: fix task_work add use-after-free with SQPOLL

Pavel Begunkov (2):
  io_uring: don't touch sqd->thread off tw add
  io_uring/sqpoll: zero sqd->thread on tctx errors

Shakeel Butt (1):
  io_uring/sqpoll: do the return-to-userspace work in the SQPOLL thread

Yang Xiuwei (2):
  io_uring: fail soft links on negative IOU_COMPLETE results
  io_uring/rw: fail soft links on IOPOLL short or error IO

 include/uapi/linux/io_uring.h |   2 +
 io_uring/io_uring.c           | 104 ++++++++++++++++++++++------------
 io_uring/io_uring.h           |  25 ++++++--
 io_uring/rsrc.c               |  25 ++++----
 io_uring/rw.c                 |   1 +
 io_uring/sqpoll.c             |  62 +++++++++++++++++++-
 io_uring/timeout.c            |   8 ++-
 7 files changed, 168 insertions(+), 59 deletions(-)

-- 
2.55.0


^ permalink raw reply	[flat|nested] 20+ messages in thread

end of thread, other threads:[~2026-11-21 21:31 UTC | newest]

Thread overview: 20+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-11-21 21:30 [PATCH 6.6.y 00/19] io_uring fixes for 6.6-stable Jens Axboe
2026-11-21 21:30 ` [PATCH 6.6.y 01/19] io_uring: enable toggle of iowait usage when waiting on CQEs Jens Axboe
2026-11-21 21:30 ` [PATCH 6.6.y 02/19] io_uring: fail soft links on negative IOU_COMPLETE results Jens Axboe
2026-11-21 21:30 ` [PATCH 6.6.y 03/19] io_uring/rw: fail soft links on IOPOLL short or error IO Jens Axboe
2026-11-21 21:30 ` [PATCH 6.6.y 04/19] io_uring/timeout: don't rearm a multishot timeout on cancel Jens Axboe
2026-11-21 21:30 ` [PATCH 6.6.y 05/19] io_uring/rsrc: keep the old file when a file update fails Jens Axboe
2026-11-21 21:30 ` [PATCH 6.6.y 06/19] io_uring: cleanup handle_tw_list() calling convention Jens Axboe
2026-11-21 21:30 ` [PATCH 6.6.y 07/19] io_uring: pass in counter to handle_tw_list() rather than return it Jens Axboe
2026-11-21 21:30 ` [PATCH 6.6.y 08/19] io_uring/sqpoll: manage task_work privately Jens Axboe
2026-11-21 21:30 ` [PATCH 6.6.y 09/19] io_uring: wake SQPOLL task when task_work is added to an empty queue Jens Axboe
2026-11-21 21:30 ` [PATCH 6.6.y 10/19] io_uring/sqpoll: use the correct check for pending task_work Jens Axboe
2026-11-21 21:30 ` [PATCH 6.6.y 11/19] io_uring/sqpoll: early exit thread if task_context wasn't allocated Jens Axboe
2026-11-21 21:30 ` [PATCH 6.6.y 12/19] io_uring/sqpoll: ensure that normal task_work is also run timely Jens Axboe
2026-11-21 21:30 ` [PATCH 6.6.y 13/19] io_uring: signal SQPOLL task_work with TWA_SIGNAL_NO_IPI Jens Axboe
2026-11-21 21:30 ` [PATCH 6.6.y 14/19] io_uring/sqpoll: ensure task state is TASK_RUNNING when running task_work Jens Axboe
2026-11-21 21:30 ` [PATCH 6.6.y 15/19] io_uring: don't touch sqd->thread off tw add Jens Axboe
2026-11-21 21:30 ` [PATCH 6.6.y 16/19] io_uring/sqpoll: zero sqd->thread on tctx errors Jens Axboe
2026-11-21 21:30 ` [PATCH 6.6.y 17/19] io_uring/sqpoll: Increase task_work submission batch size Jens Axboe
2026-11-21 21:30 ` [PATCH 6.6.y 18/19] io_uring: fix task_work add use-after-free with SQPOLL Jens Axboe
2026-11-21 21:30 ` [PATCH 6.6.y 19/19] io_uring/sqpoll: do the return-to-userspace work in the SQPOLL thread Jens Axboe

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).