From: Jens Axboe <ax...@kernel.dk>
commit a36da65c46565d2527eec3efdb546251e38253fd upstream.
Check the ipt.error value, it must have been either cleared to zero or
set to another error than the default -EINVAL if we don't go through the
waitqueue proc addition. Just give up on poll at that point and return
failure, this will fallback to async work.
io_poll_add() doesn't suffer from this failure case, as it returns the
error value directly.
Cc: sta...@vger.kernel.org # v5.7+
Reported-by: syzbot+a730016dc0bdce4f6...@syzkaller.appspotmail.com
Reviewed-by: Stefano Garzarella <sgarz...@redhat.com>
Signed-off-by: Jens Axboe <ax...@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gre...@linuxfoundation.org>
---
fs/io_uring.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -4483,7 +4483,7 @@ static bool io_arm_poll_handler(struct i
ret = __io_arm_poll_handler(req, &apoll->poll, &ipt, mask,
io_async_wake);
- if (ret) {
+ if (ret || ipt.error) {
io_poll_remove_double(req, apoll->double_poll);
spin_unlock_irq(&ctx->completion_lock);
if (req->flags & REQ_F_WORK_INITIALIZED)
