On 6/23/23 3:29?AM, gregkh@xxxxxxxxxxxxxxxxxxx wrote:
>
> The patch below does not apply to the 6.1-stable tree.
> If someone wants it applied there, or to any other stable or longterm
> tree, then please email the backport, including the original git commit
> id to <stable@xxxxxxxxxxxxxxx>.
>
> To reproduce the conflict and resubmit, you may use the following commands:
>
> git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-6.1.y
> git checkout FETCH_HEAD
> git cherry-pick -x ef7dfac51d8ed961b742218f526bd589f3900a59
> # <resolve conflicts, build, test, etc.>
> git commit -s
> git send-email --to '<stable@xxxxxxxxxxxxxxx>' --in-reply-to '2023062307-taco-nurture-70a2@gregkh' --subject-prefix 'PATCH 6.1.y' HEAD^..
Here's this one for 6.1-stable, thanks.
--
Jens Axboe
From 43721de4aa349adcf785e00ceecddcc4a70ac9f2 Mon Sep 17 00:00:00 2001
From: Jens Axboe <axboe@xxxxxxxxx>
Date: Sat, 17 Jun 2023 19:50:24 -0600
Subject: [PATCH] io_uring/poll: serialize poll linked timer start with poll
removal
Commit ef7dfac51d8ed961b742218f526bd589f3900a59 upstream.
We selectively grab the ctx->uring_lock for poll update/removal, but
we really should grab it from the start to fully synchronize with
linked timeouts. Normally this is indeed the case, but if requests
are forced async by the application, we don't fully cover removal
and timer disarm within the uring_lock.
Make this simpler by having consistent locking state for poll removal.
Cc: stable@xxxxxxxxxxxxxxx # 6.1+
Reported-by: Querijn Voet <querijnqyn@xxxxxxxxx>
Signed-off-by: Jens Axboe <axboe@xxxxxxxxx>
---
io_uring/poll.c | 9 ++++-----
1 file changed, 4 insertions(+), 5 deletions(-)
diff --git a/io_uring/poll.c b/io_uring/poll.c
index 4788073ec45d..869e1d2a4413 100644
--- a/io_uring/poll.c
+++ b/io_uring/poll.c
@@ -993,8 +993,9 @@ int io_poll_remove(struct io_kiocb *req, unsigned int issue_flags)
struct io_hash_bucket *bucket;
struct io_kiocb *preq;
int ret2, ret = 0;
- bool locked;
+ bool locked = true;
+ io_ring_submit_lock(ctx, issue_flags);
preq = io_poll_find(ctx, true, &cd, &ctx->cancel_table, &bucket);
ret2 = io_poll_disarm(preq);
if (bucket)
@@ -1006,12 +1007,10 @@ int io_poll_remove(struct io_kiocb *req, unsigned int issue_flags)
goto out;
}
- io_ring_submit_lock(ctx, issue_flags);
preq = io_poll_find(ctx, true, &cd, &ctx->cancel_table_locked, &bucket);
ret2 = io_poll_disarm(preq);
if (bucket)
spin_unlock(&bucket->lock);
- io_ring_submit_unlock(ctx, issue_flags);
if (ret2) {
ret = ret2;
goto out;
@@ -1035,7 +1034,7 @@ int io_poll_remove(struct io_kiocb *req, unsigned int issue_flags)
if (poll_update->update_user_data)
preq->cqe.user_data = poll_update->new_user_data;
- ret2 = io_poll_add(preq, issue_flags);
+ ret2 = io_poll_add(preq, issue_flags & ~IO_URING_F_UNLOCKED);
/* successfully updated, don't complete poll request */
if (!ret2 || ret2 == -EIOCBQUEUED)
goto out;
@@ -1043,9 +1042,9 @@ int io_poll_remove(struct io_kiocb *req, unsigned int issue_flags)
req_set_fail(preq);
io_req_set_res(preq, -ECANCELED, 0);
- locked = !(issue_flags & IO_URING_F_UNLOCKED);
io_req_task_complete(preq, &locked);
out:
+ io_ring_submit_unlock(ctx, issue_flags);
if (ret < 0) {
req_set_fail(req);
return ret;
--
2.40.1
