diff options
| author | Jens Axboe <axboe@kernel.dk> | 2026-01-23 13:58:03 -0700 |
|---|---|---|
| committer | Jens Axboe <axboe@kernel.dk> | 2026-01-23 13:58:03 -0700 |
| commit | 816095894c0f44aaba4372d92c874121af687596 (patch) | |
| tree | debffb5bb09639711cee0774627b7535db89bb6b /io_uring | |
| parent | 1edf0891d0f4a6c186721e41323c9a3b86ceceda (diff) | |
io_uring/io-wq: handle !sysctl_hung_task_timeout_secs
If the hung_task_timeout sysctl is set to 0, then we'll end up busy
looping inside io_wq_exit_workers() after an earlier commit switched to
using wait_for_completion_timeout(). Use the maximum schedule timeout
value for that case.
Fixes: 1f293098a313 ("io_uring/io-wq: don't trigger hung task for syzbot craziness")
Reported-by: Chris Mason <clm@fb.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Diffstat (limited to 'io_uring')
| -rw-r--r-- | io_uring/io-wq.c | 2 |
1 files changed, 2 insertions, 0 deletions
diff --git a/io_uring/io-wq.c b/io_uring/io-wq.c index aaf1dfc21763..3b55feb620d9 100644 --- a/io_uring/io-wq.c +++ b/io_uring/io-wq.c @@ -1340,6 +1340,8 @@ static void io_wq_exit_workers(struct io_wq *wq) * up waiting more than IO_URING_EXIT_WAIT_MAX. */ timeout = sysctl_hung_task_timeout_secs * HZ / 2; + if (!timeout) + timeout = MAX_SCHEDULE_TIMEOUT; warn_timeout = jiffies + IO_URING_EXIT_WAIT_MAX; do { if (wait_for_completion_timeout(&wq->worker_done, timeout)) |