public inbox for io-uring@vger.kernel.org
 help / color / mirror / Atom feed
From: Jens Axboe <axboe@kernel.dk>
To: io-uring@vger.kernel.org
Cc: Kyle Zeng <kylebot@openai.com>,
	stable@vger.kernel.org, Jens Axboe <axboe@kernel.dk>
Subject: [PATCH 09/10] io_uring/io-wq: save worker type before dropping its reference
Date: Wed, 28 Oct 2026 10:46:57 -0600	[thread overview]
Message-ID: <20261028164732.1264069-10-axboe@kernel.dk> (raw)
In-Reply-To: <20261028164732.1264069-1-axboe@kernel.dk>

From: Kyle Zeng <kylebot@openai.com>

io_wq_cancel_tw_create() inspects cb->func after calling
io_worker_cancel_cb(). For an existing worker, that call can drop the
last worker reference and wake io_worker_exit(). If the canceling task
is preempted, the worker can be unlinked and freed by kfree_rcu() before
the callback function is inspected. KASAN reports a slab-use-after-free
read in this path.

Save whether the callback belongs to an unpublished continuation worker
while its reference is still held, then use the saved result for the
conditional free. This preserves the existing accounting and frees only
the worker allocated for the canceled continuation.

Fixes: e6db6f9398da ("io_uring/io-wq: only free worker if it was allocated for creation")
Cc: stable@vger.kernel.org
Signed-off-by: Kyle Zeng <kylebot@openai.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 io_uring/io-wq.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/io_uring/io-wq.c b/io_uring/io-wq.c
index 28533174d0da..92df54c6f50c 100644
--- a/io_uring/io-wq.c
+++ b/io_uring/io-wq.c
@@ -1434,14 +1434,16 @@ static void io_wq_cancel_tw_create(struct io_wq *wq)
 
 	while ((cb = task_work_cancel_match(wq->task, io_task_work_match, wq)) != NULL) {
 		struct io_worker *worker;
+		bool free_worker;
 
 		worker = container_of(cb, struct io_worker, create_work);
-		io_worker_cancel_cb(worker);
 		/*
 		 * Only the worker continuation helper has worker allocated and
-		 * hence needs freeing.
+		 * hence needs freeing. Save this before dropping its reference.
 		 */
-		if (cb->func == create_worker_cont)
+		free_worker = cb->func == create_worker_cont;
+		io_worker_cancel_cb(worker);
+		if (free_worker)
 			kfree(worker);
 	}
 }
-- 
2.56.0


  parent reply	other threads:[~2026-10-28 16:47 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-10-28 16:46 [PATCHSET 0/10] Various io_uring fixes Jens Axboe
2026-10-28 16:46 ` [PATCH 01/10] io_uring/io-wq: no new workers once the exit has begun Jens Axboe
2026-10-28 16:46 ` [PATCH 02/10] io_uring/rsrc: charge buffer arrays and nodes to memcg Jens Axboe
2026-10-28 16:46 ` [PATCH 03/10] io_uring: charge tctx nodes and personalities Jens Axboe
2026-10-28 16:46 ` [PATCH 04/10] io_uring: free XArray nodes left by a failed store Jens Axboe
2026-10-28 16:46 ` [PATCH 05/10] io_uring/bpf-ops: clear bpf_ctx while submitting SQEs Jens Axboe
2026-10-28 16:46 ` [PATCH 06/10] io_uring: free the task context after a grace period Jens Axboe
2026-10-28 16:46 ` [PATCH 07/10] io_uring/sqpoll: don't wait for the thread to exit Jens Axboe
2026-10-28 16:46 ` [PATCH 08/10] io_uring/cancel: drop sqd->lock while SQPOLL waits Jens Axboe
2026-10-28 16:46 ` Jens Axboe [this message]
2026-10-28 16:46 ` [PATCH 10/10] io_uring: defer foreign io-wq enqueues to task_work Jens Axboe

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20261028164732.1264069-10-axboe@kernel.dk \
    --to=axboe@kernel.dk \
    --cc=io-uring@vger.kernel.org \
    --cc=kylebot@openai.com \
    --cc=stable@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox