Re: 3.14.0+/x86: lockdep and mutexes not getting along

[Peter Zijlstra <peterz@infradead.org>]

On Thu, Apr 10, 2014 at 11:18:24AM +0200, Peter Zijlstra wrote:
> So I managed to reproduce, and the below makes it go away. I just don't
> understand why though. will stare more.

/me kicks himself.. bloody obvious fail there :-)

Not unlocking the lock after a lockdep trigger will make things get
stuck real fast :-)

---
 kernel/locking/mutex-debug.c | 19 +++++++++----------
 1 file changed, 9 insertions(+), 10 deletions(-)

diff --git a/kernel/locking/mutex-debug.c b/kernel/locking/mutex-debug.c
index e1191c996c59..5cf6731b98e9 100644
--- a/kernel/locking/mutex-debug.c
+++ b/kernel/locking/mutex-debug.c
@@ -71,18 +71,17 @@ void mutex_remove_waiter(struct mutex *lock, struct mutex_waiter *waiter,
 
 void debug_mutex_unlock(struct mutex *lock)
 {
-	if (unlikely(!debug_locks))
-		return;
+	if (likely(debug_locks)) {
+		DEBUG_LOCKS_WARN_ON(lock->magic != lock);
 
-	DEBUG_LOCKS_WARN_ON(lock->magic != lock);
+		if (!lock->owner)
+			DEBUG_LOCKS_WARN_ON(!lock->owner);
+		else
+			DEBUG_LOCKS_WARN_ON(lock->owner != current);
 
-	if (!lock->owner)
-		DEBUG_LOCKS_WARN_ON(!lock->owner);
-	else
-		DEBUG_LOCKS_WARN_ON(lock->owner != current);
-
-	DEBUG_LOCKS_WARN_ON(!lock->wait_list.prev && !lock->wait_list.next);
-	mutex_clear_owner(lock);
+		DEBUG_LOCKS_WARN_ON(!lock->wait_list.prev && !lock->wait_list.next);
+		mutex_clear_owner(lock);
+	}
 
 	/*
 	 * __mutex_slowpath_needs_to_unlock() is explicitly 0 for debug