[PATCH 03/11] locking/rtmutex: Squash self-deadlock check for ww_rt_mutex.

[Sebastian Andrzej Siewior]

From: Peter Zijlstra <peterz@xxxxxxxxxxxxx>

Similar to the issues in commits:

  6467822b8cc9 ("locking/rtmutex: Prevent spurious EDEADLK return caused by ww_mutexes")
  a055fcc132d4 ("locking/rtmutex: Return success on deadlock for ww_mutex waiters")

ww_rt_mutex_lock() should not return EDEADLK without first going through
the __ww_mutex logic to set the required state. In fact, the chain-walk
can deal with the spurious cycles (per the above commits) this check
warns about and is trying to avoid.

Therefore ignore this test for ww_rt_mutex and simply let things fall
in place.

Signed-off-by: Peter Zijlstra (Intel) <peterz@xxxxxxxxxxxxx>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@xxxxxxxxxxxxx>
---
 kernel/locking/rtmutex.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/kernel/locking/rtmutex.c b/kernel/locking/rtmutex.c
index 0c6a48dfcecb3..f89620852774d 100644
--- a/kernel/locking/rtmutex.c
+++ b/kernel/locking/rtmutex.c
@@ -1103,8 +1103,11 @@ static int __sched task_blocks_on_rt_mutex(struct rt_mutex_base *lock,
 	 * the other will detect the deadlock and return -EDEADLOCK,
 	 * which is wrong, as the other waiter is not in a deadlock
 	 * situation.
+	 *
+	 * Except for ww_mutex, in that case the chain walk must already deal
+	 * with spurious cycles, see the comments at [3] and [6].
 	 */
-	if (owner == task)
+	if (owner == task && !(build_ww_mutex() && ww_ctx))
 		return -EDEADLK;
 
 	raw_spin_lock(&task->pi_lock);
-- 
2.34.0