[tip:core/locking] locking/mutexes/mcs: Correct barrier usage

[tip-bot for Waiman Long <tipbot@zytor.com>]

Commit-ID:  aff7385b5a16bca6b8d9243f01a9ea5a5b411e1d
Gitweb:     http://git.kernel.org/tip/aff7385b5a16bca6b8d9243f01a9ea5a5b411e1d
Author:     Waiman Long <Waiman.Long@hp.com>
AuthorDate: Tue, 21 Jan 2014 15:35:53 -0800
Committer:  Ingo Molnar <mingo@kernel.org>
CommitDate: Tue, 28 Jan 2014 13:13:26 +0100

locking/mutexes/mcs: Correct barrier usage

This patch corrects the way memory barriers are used in the MCS lock
with smp_load_acquire and smp_store_release fucnctions.  The previous
barriers could leak critical sections if mcs lock is used by itself.
It is not a problem when mcs lock is embedded in mutex but will be an
issue when the mcs_lock is used elsewhere.

The patch removes the incorrect barriers and put in correct
barriers with the pair of functions smp_load_acquire and smp_store_release.

Suggested-by: Michel Lespinasse <walken@google.com>
Reviewed-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Waiman Long <Waiman.Long@hp.com>
Signed-off-by: Jason Low <jason.low2@hp.com>
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/1390347353.3138.62.camel@schen9-DESK
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 kernel/locking/mutex.c | 18 +++++++++++++-----
 1 file changed, 13 insertions(+), 5 deletions(-)

diff --git a/kernel/locking/mutex.c b/kernel/locking/mutex.c
index 4dd6e4c..fbbd2ed 100644
--- a/kernel/locking/mutex.c
+++ b/kernel/locking/mutex.c
@@ -136,9 +136,12 @@ void mspin_lock(struct mspin_node **lock, struct mspin_node *node)
 		return;
 	}
 	ACCESS_ONCE(prev->next) = node;
-	smp_wmb();
-	/* Wait until the lock holder passes the lock down */
-	while (!ACCESS_ONCE(node->locked))
+	/*
+	 * Wait until the lock holder passes the lock down.
+	 * Using smp_load_acquire() provides a memory barrier that
+	 * ensures subsequent operations happen after the lock is acquired.
+	 */
+	while (!(smp_load_acquire(&node->locked)))
 		arch_mutex_cpu_relax();
 }
 
@@ -156,8 +159,13 @@ static void mspin_unlock(struct mspin_node **lock, struct mspin_node *node)
 		while (!(next = ACCESS_ONCE(node->next)))
 			arch_mutex_cpu_relax();
 	}
-	ACCESS_ONCE(next->locked) = 1;
-	smp_wmb();
+	/*
+	 * Pass lock to next waiter.
+	 * smp_store_release() provides a memory barrier to ensure
+	 * all operations in the critical section has been completed
+	 * before unlocking.
+	 */
+	smp_store_release(&next->locked, 1);
 }
 
 /*