]> git.ipfire.org Git - thirdparty/kernel/stable.git/commit
nvme: always punt polled uring_cmd end_io work to task_work
authorJens Axboe <axboe@kernel.dk>
Fri, 13 Jun 2025 19:37:41 +0000 (13:37 -0600)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Sun, 6 Jul 2025 09:00:16 +0000 (11:00 +0200)
commit6f27bbf3de0d26090c298a58fcd9bb99f941ea07
tree4451959666a608aa5119e55bee49e050e9e57ddf
parentcdfb20e4b34ad99b3fe122aafb4f8ee7b9856e1f
nvme: always punt polled uring_cmd end_io work to task_work

Commit 9ce6c9875f3e995be5fd720b65835291f8a609b1 upstream.

Currently NVMe uring_cmd completions will complete locally, if they are
polled. This is done because those completions are always invoked from
task context. And while that is true, there's no guarantee that it's
invoked under the right ring context, or even task. If someone does
NVMe passthrough via multiple threads and with a limited number of
poll queues, then ringA may find completions from ringB. For that case,
completing the request may not be sound.

Always just punt the passthrough completions via task_work, which will
redirect the completion, if needed.

Cc: stable@vger.kernel.org
Fixes: 585079b6e425 ("nvme: wire up async polling for io passthrough commands")
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
drivers/nvme/host/ioctl.c