fs/epoll: avoid barrier after an epoll_wait(2) timeout

Upon timeout, we can just exit out of the loop, without the cost of the
changing the task's state with an smp_store_mb call.  Just exit out of
the loop and be done - setting the task state afterwards will be, of
course, redundant.

[dave@stgolabs.net: forgotten fixlets]
  Link: http://lkml.kernel.org/r/20181109155258.jxcr4t2pnz6zqct3@linux-r8p5
Link: http://lkml.kernel.org/r/20181108051006.18751-7-dave@stgolabs.net
Signed-off-by: Davidlohr Bueso <dave@stgolabs.net>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Davidlohr Bueso <dbueso@suse.de>
Cc: Jason Baron <jbaron@akamai.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

fs/eventpoll.c

@@ -1820,15 +1820,19 @@ static int ep_poll(struct eventpoll *ep, struct epoll_event __user *events,
 			res = -EINTR;
 			break;
 		}
-		if (ep_events_available(ep) || timed_out)
+
+		eavail = ep_events_available(ep);
+		if (eavail)
 			break;
 		if (signal_pending(current)) {
 			res = -EINTR;
 			break;
 		}
 
-		if (!schedule_hrtimeout_range(to, slack, HRTIMER_MODE_ABS))
+		if (!schedule_hrtimeout_range(to, slack, HRTIMER_MODE_ABS)) {
 			timed_out = 1;
+			break;
+		}
 	}
 
 	__set_current_state(TASK_RUNNING);