user/sven/linux.git/io_uring, branch v6.17.6

io_uring: correct __must_hold annotation in io_install_fixed_file

2025-10-29T13:10:28Z

[ Upstream commit c5efc6a0b3940381d67887302ddb87a5cf623685 ] The __must_hold annotation references &req->ctx->uring_lock, but req is not in scope in io_install_fixed_file. This change updates the annotation to reference the correct ctx->uring_lock. improving code clarity. Fixes: f110ed8498af ("io_uring: split out fixed file installation and removal") Signed-off-by: Alok Tiwari Signed-off-by: Jens Axboe Signed-off-by: Sasha Levin

io_uring: fix incorrect unlikely() usage in io_waitid_prep()

2025-10-29T13:10:28Z

[ Upstream commit 4ec703ec0c384a2199808c4eb2e9037236285a8d ] The negation operator is incorrectly placed outside the unlikely() macro: if (!unlikely(iwa)) This inverts the compiler branch prediction hint, marking the NULL case as likely instead of unlikely. The intent is to indicate that allocation failures are rare, consistent with common kernel patterns. Moving the negation inside unlikely(): if (unlikely(!iwa)) Fixes: 2b4fc4cd43f2 ("io_uring/waitid: setup async data in the prep handler") Signed-off-by: Alok Tiwari Reviewed-by: Gabriel Krisman Bertazi Reviewed-by: Caleb Sander Mateos Signed-off-by: Jens Axboe Signed-off-by: Sasha Levin

io_uring/sqpoll: be smarter on when to update the stime usage

2025-10-29T13:10:21Z

commit a94e0657269c5b8e1a90b17aa2c048b3d276e16d upstream. The current approach is a bit naive, and hence calls the time querying way too often. Only start the "doing work" timer when there's actual work to do, and then use that information to terminate (and account) the work time once done. This greatly reduces the frequency of these calls, when they cannot have changed anyway. Running a basic random reader that is setup to use SQPOLL, a profile before this change shows these as the top cycle consumers: + 32.60% iou-sqp-1074 [kernel.kallsyms] [k] thread_group_cputime_adjusted + 19.97% iou-sqp-1074 [kernel.kallsyms] [k] thread_group_cputime + 12.20% io_uring io_uring [.] submitter_uring_fn + 4.13% iou-sqp-1074 [kernel.kallsyms] [k] getrusage + 2.45% iou-sqp-1074 [kernel.kallsyms] [k] io_submit_sqes + 2.18% iou-sqp-1074 [kernel.kallsyms] [k] __pi_memset_generic + 2.09% iou-sqp-1074 [kernel.kallsyms] [k] cputime_adjust and after this change, top of profile looks as follows: + 36.23% io_uring io_uring [.] submitter_uring_fn + 23.26% iou-sqp-819 [kernel.kallsyms] [k] io_sq_thread + 10.14% iou-sqp-819 [kernel.kallsyms] [k] io_sq_tw + 6.52% iou-sqp-819 [kernel.kallsyms] [k] tctx_task_work_run + 4.82% iou-sqp-819 [kernel.kallsyms] [k] nvme_submit_cmds.part.0 + 2.91% iou-sqp-819 [kernel.kallsyms] [k] io_submit_sqes [...] 0.02% iou-sqp-819 [kernel.kallsyms] [k] cputime_adjust where it's spending the cycles on things that actually matter. Reported-by: Fengnan Chang Cc: stable@vger.kernel.org Fixes: 3fcb9d17206e ("io_uring/sqpoll: statistics of the true utilization of sq threads") Signed-off-by: Jens Axboe Signed-off-by: Greg Kroah-Hartman

io_uring/sqpoll: switch away from getrusage() for CPU accounting

2025-10-29T13:10:21Z

commit 8ac9b0d33e5c0a995338ee5f25fe1b6ff7d97f65 upstream. getrusage() does a lot more than what the SQPOLL accounting needs, the latter only cares about (and uses) the stime. Rather than do a full RUSAGE_SELF summation, just query the used stime instead. Cc: stable@vger.kernel.org Fixes: 3fcb9d17206e ("io_uring/sqpoll: statistics of the true utilization of sq threads") Reviewed-by: Gabriel Krisman Bertazi Signed-off-by: Jens Axboe Signed-off-by: Greg Kroah-Hartman

io_uring: protect mem region deregistration

2025-10-23T14:24:22Z

commit be7cab44ed099566c605a8dac686c3254db01b35 upstream. io_create_region_mmap_safe() protects publishing of a region against concurrent mmap calls, however we should also protect against it when removing a region. There is a gap io_register_mem_region() where it safely publishes a region, but then copy_to_user goes wrong and it unsafely frees the region. Cc: stable@vger.kernel.org Fixes: 087f997870a94 ("io_uring/memmap: implement mmap for regions") Signed-off-by: Pavel Begunkov Signed-off-by: Jens Axboe Signed-off-by: Greg Kroah-Hartman

Revert "io_uring/rw: drop -EOPNOTSUPP check in __io_complete_rw_common()"

2025-10-23T14:24:21Z

commit 927069c4ac2cd1a37efa468596fb5b8f86db9df0 upstream. This reverts commit 90bfb28d5fa8127a113a140c9791ea0b40ab156a. Kevin reports that this commit causes an issue for him with LVM snapshots, most likely because of turning off NOWAIT support while a snapshot is being created. This makes -EOPNOTSUPP bubble back through the completion handler, where io_uring read/write handling should just retry it. Reinstate the previous check removed by the referenced commit. Cc: stable@vger.kernel.org Fixes: 90bfb28d5fa8 ("io_uring/rw: drop -EOPNOTSUPP check in __io_complete_rw_common()") Reported-by: Salvatore Bonaccorso Reported-by: Kevin Lumik Link: https://lore.kernel.org/io-uring/cceb723c-051b-4de2-9a4c-4aa82e1619ee@kernel.dk/ Signed-off-by: Jens Axboe Signed-off-by: Greg Kroah-Hartman

io_uring/zcrx: increment fallback loop src offset

2025-10-19T14:37:14Z

[ Upstream commit e9a9dcb4ccb32446165800a9d83058e95c4833d2 ] Don't forget to adjust the source offset in io_copy_page(), otherwise it'll be copying into the same location in some cases for highmem setups. Fixes: e67645bb7f3f4 ("io_uring/zcrx: prepare fallback for larger pages") Signed-off-by: Pavel Begunkov Signed-off-by: Jens Axboe Signed-off-by: Sasha Levin

io_uring/zcrx: fix overshooting recv limit

2025-10-15T10:04:15Z

commit 09cfd3c52ea76f43b3cb15e570aeddf633d65e80 upstream. It's reported that sometimes a zcrx request can receive more than was requested. It's caused by io_zcrx_recv_skb() adjusting desc->count for all received buffers including frag lists, but then doing recursive calls to process frag list skbs, which leads to desc->count double accounting and underflow. Reported-and-tested-by: Matthias Jasny Fixes: 6699ec9a23f85 ("io_uring/zcrx: add a read limit to recvzc requests") Cc: stable@vger.kernel.org Signed-off-by: Pavel Begunkov Signed-off-by: Jens Axboe Signed-off-by: Greg Kroah-Hartman

io_uring/waitid: always prune wait queue entry in io_waitid_wait()

2025-10-15T10:04:15Z

commit 2f8229d53d984c6a05b71ac9e9583d4354e3b91f upstream. For a successful return, always remove our entry from the wait queue entry list. Previously this was skipped if a cancelation was in progress, but this can race with another invocation of the wait queue entry callback. Cc: stable@vger.kernel.org Fixes: f31ecf671ddc ("io_uring: add IORING_OP_WAITID support") Reported-by: syzbot+b9e83021d9c642a33d8c@syzkaller.appspotmail.com Tested-by: syzbot+b9e83021d9c642a33d8c@syzkaller.appspotmail.com Link: https://lore.kernel.org/io-uring/68e5195e.050a0220.256323.001f.GAE@google.com/ Signed-off-by: Jens Axboe Signed-off-by: Greg Kroah-Hartman

Merge tag 'io_uring-6.17-20250919' of git://git.kernel.dk/linux

2025-09-19T19:10:49Z

Pull io_uring fixes from Jens Axboe: - Fix for a regression introduced in the io-wq worker creation logic. - Remove the allocation cache for the msg_ring io_kiocb allocations. I have a suspicion that there's a bug there, and since we just fixed one in that area, let's just yank the use of that cache entirely. It's not that important, and it kills some code. - Treat a closed ring like task exiting in that any requests that trigger post that condition should just get canceled. Doesn't fix any real issues, outside of having tasks being able to rely on that guarantee. - Fix for a bug in the network zero-copy notification mechanism, where a comparison for matching tctx/ctx for notifications was buggy in that it didn't correctly compare with the previous notification. * tag 'io_uring-6.17-20250919' of git://git.kernel.dk/linux: io_uring: fix incorrect io_kiocb reference in io_link_skb io_uring/msg_ring: kill alloc_cache for io_kiocb allocations io_uring: include dying ring in task_work "should cancel" state io_uring/io-wq: fix `max_workers` breakage and `nr_workers` underflow