[PATCH] epoll: add a reschedule point in ep_free()

[PATCH] epoll: add a reschedule point in ep_free()

[Eric Dumazet]

From: Eric Dumazet <edumazet@google.com>

ep_free() might iterate on a huge set of epitems and hold
cpu too long. Add two cond_resched() in order to yield cpu
to other tasks. This is safe as we only hold mutexes in
this function.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Cc: Theodore Ts'o <tytso@mit.edu>
---
Google-Bug-Id: 8940884

 fs/eventpoll.c |    2 ++
 1 file changed, 2 insertions(+)

diff --git a/fs/eventpoll.c b/fs/eventpoll.c
index 9ad17b15..4ff595d 100644
--- a/fs/eventpoll.c
+++ b/fs/eventpoll.c
@@ -740,6 +740,7 @@ static void ep_free(struct eventpoll *ep)
 		epi = rb_entry(rbp, struct epitem, rbn);
 
 		ep_unregister_pollwait(ep, epi);
+		cond_resched();
 	}
 
 	/*
@@ -754,6 +755,7 @@ static void ep_free(struct eventpoll *ep)
 	while ((rbp = rb_first(&ep->rbr)) != NULL) {
 		epi = rb_entry(rbp, struct epitem, rbn);
 		ep_remove(ep, epi);
+		cond_resched();
 	}
 	mutex_unlock(&ep->mtx);
 

Re: [PATCH] epoll: add a reschedule point in ep_free()

[Eric Wong]

Eric Dumazet <eric.dumazet@gmail.com> wrote:
> From: Eric Dumazet <edumazet@google.com>
> 
> ep_free() might iterate on a huge set of epitems and hold
> cpu too long. Add two cond_resched() in order to yield cpu
> to other tasks. This is safe as we only hold mutexes in
> this function.
> 
> Signed-off-by: Eric Dumazet <edumazet@google.com>
> Cc: Al Viro <viro@ZenIV.linux.org.uk>
> Cc: Theodore Ts'o <tytso@mit.edu>

Seems reasonable.  This isn't a remotely hot path and f_op->release
runs in a workqueue nowadays (right?).

Acked-by: Eric Wong <normalperson@yhbt.net>