rt-mutex: fix stale return value

c0d1d2bf · Thomas Gleixner <[email protected]> · 2007-06-08 13:46:57 -07:00

Alexey Kuznetsov found some problems in the pi-futex code.

The major problem is a stale return value in rt_mutex_slowlock():

When the pi chain walk returns -EDEADLK, but the waiter was woken up during
the phases where the locks were dropped, the rtmutex could be acquired, but
due to the stale return value -EDEADLK returned to the caller.

Reset the return value in the retry path.

Signed-off-by: Thomas Gleixner <[email protected]>
Acked-by: Ingo Molnar <[email protected]>
Cc: Steven Rostedt <[email protected]>
Cc: Alexey Kuznetsov <[email protected]>
Cc: Ulrich Drepper <[email protected]>
Cc: Eric Dumazet <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>

Commit: c0d1d2bf5a28f78def7b68ca1eb5ba31aafd43a1
Parents: 51b94d2a

Changes

1 files changed · +9 −2

File	Status	Add	Del	Content
kernel/rtmutex.c	modified	+9	−2	View file

kernel/rtmutex.c

View file

diff --git a/kernel/rtmutex.c b/kernel/rtmutex.c
index 12879f6..a273183 100644
--- a/kernel/rtmutex.c
+++ b/kernel/rtmutex.c
@@ -636,9 +636,16 @@ rt_mutex_slowlock(struct rt_mutex *lock, int state,
 			 * all over without going into schedule to try
 			 * to get the lock now:
 			 */
-			if (unlikely(!waiter.task))
+			if (unlikely(!waiter.task)) {
+				/*
+				 * Reset the return value. We might
+				 * have returned with -EDEADLK and the
+				 * owner released the lock while we
+				 * were walking the pi chain.
+				 */
+				ret = 0;
 				continue;
-
+			}
 			if (unlikely(ret))
 				break;
 		}