signalfd: use wake_up_pollfree()
authorEric Biggers <ebiggers@google.com>
Thu, 9 Dec 2021 01:04:53 +0000 (17:04 -0800)
committerEric Biggers <ebiggers@google.com>
Thu, 9 Dec 2021 18:49:56 +0000 (10:49 -0800)
wake_up_poll() uses nr_exclusive=1, so it's not guaranteed to wake up
all exclusive waiters.  Yet, POLLFREE *must* wake up all waiters.  epoll
and aio poll are fortunately not affected by this, but it's very
fragile.  Thus, the new function wake_up_pollfree() has been introduced.

Convert signalfd to use wake_up_pollfree().

Reported-by: Linus Torvalds <torvalds@linux-foundation.org>
Fixes: d80e731ecab4 ("epoll: introduce POLLFREE to flush ->signalfd_wqh before kfree()")
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20211209010455.42744-4-ebiggers@kernel.org
Signed-off-by: Eric Biggers <ebiggers@google.com>
fs/signalfd.c

index 040e1cf..65ce0e7 100644 (file)
 
 void signalfd_cleanup(struct sighand_struct *sighand)
 {
-       wait_queue_head_t *wqh = &sighand->signalfd_wqh;
-       /*
-        * The lockless check can race with remove_wait_queue() in progress,
-        * but in this case its caller should run under rcu_read_lock() and
-        * sighand_cachep is SLAB_TYPESAFE_BY_RCU, we can safely return.
-        */
-       if (likely(!waitqueue_active(wqh)))
-               return;
-
-       /* wait_queue_entry_t->func(POLLFREE) should do remove_wait_queue() */
-       wake_up_poll(wqh, EPOLLHUP | POLLFREE);
+       wake_up_pollfree(&sighand->signalfd_wqh);
 }
 
 struct signalfd_ctx {