From: Soheil Hassas Yeganeh <soheil@xxxxxxxxxx>
After abc610e01c66, we break out of the ep_poll loop upon timeout,
without checking whether there is any new events available. Prior to
that patch-series we always called ep_events_available() after
exiting the loop.
This can cause races and missed wakeups. For example, consider
the following scenario reported by Guantao Liu:
Suppose we have an eventfd added using EPOLLET to an epollfd.
Thread 1: Sleeps for just below 5ms and then writes to an eventfd.
Thread 2: Calls epoll_wait with a timeout of 5 ms. If it sees an
event of the eventfd, it will write back on that fd.
Thread 3: Calls epoll_wait with a negative timeout.
Prior to abc610e01c66, it is guaranteed that Thread 3 will wake up
either by Thread 1 or Thread 2. After abc610e01c66, Thread 3 can
be blocked indefinitely if Thread 2 sees a timeout right before
the write to the eventfd by Thread 1. Thread 2 will be woken up from
schedule_hrtimeout_range and, with evail 0, it will not call
ep_send_events().
To fix this issue, while holding the lock, try to remove the thread that
timed out the wait queue and check whether it was woken up or not.
Fixes: abc610e01c66 ("fs/epoll: avoid barrier after an epoll_wait(2) timeout")
Reported-by: Guantao Liu <guantaol@xxxxxxxxxx>
Tested-by: Guantao Liu <guantaol@xxxxxxxxxx>
Signed-off-by: Soheil Hassas Yeganeh <soheil@xxxxxxxxxx>
Reviewed-by: Eric Dumazet <edumazet@xxxxxxxxxx>
Acked-by: Willem de Bruijn <willemb@xxxxxxxxxx>
Reviewed-by: Khazhismel Kumykov <khazhy@xxxxxxxxxx>
Cc: Davidlohr Bueso <dave@xxxxxxxxxxxx>
---
fs/eventpoll.c | 16 +++++++++++++++-
1 file changed, 15 insertions(+), 1 deletion(-)
diff --git a/fs/eventpoll.c b/fs/eventpoll.c
index 4df61129566d..11388436b85a 100644
--- a/fs/eventpoll.c
+++ b/fs/eventpoll.c
@@ -1907,7 +1907,21 @@ static int ep_poll(struct eventpoll *ep, struct epoll_event __user *events,
if (!schedule_hrtimeout_range(to, slack, HRTIMER_MODE_ABS)) {
timed_out = 1;
- break;
+ __set_current_state(TASK_RUNNING);
+ /*
+ * Acquire the lock and try to remove this thread from
+ * the wait queue. If this thread is not on the wait
+ * queue, it has woken up after its timeout ended
+ * before it could re-acquire the lock. In that case,
+ * try to harvest some events.
+ */
+ write_lock_irq(&ep->lock);
+ if (!list_empty(&wait.entry))
+ __remove_wait_queue(&ep->wq, &wait);
+ else
+ eavail = 1;
+ write_unlock_irq(&ep->lock);
+ goto send_events;
}
/* We were woken up, thus go and try to harvest some events */
--
2.29.0.rc2.309.g374f81d7ae-goog
