On 08/18/2016 09:28 PM, Jason Low wrote:
On Thu, 2016-08-18 at 21:15 -0400, Waiman Long wrote:
Both waiter_spinning and yield_to_waiter are used for somewhat similar
purpose. The waiter_spinning flag is used in CONFIG_MUTEX_SPIN_ON_OWNER
to make optimistic spinner yield to spinning waiter; whereas the
yield_to_waiter is used in !CONFIG_MUTEX_SPIN_ON_OWNER to make new
incoming mutex locker to yield to the sleeping waiter.
This patch unifies these 2 flags into a single yield_to_waiter
flag that is used in both CONFIG_MUTEX_SPIN_ON_OWNER and
!CONFIG_MUTEX_SPIN_ON_OWNER codes.
Signed-off-by: Waiman Long<[email protected]>
---
include/linux/mutex.h | 8 ++++++--
kernel/locking/mutex.c | 12 ++++++------
2 files changed, 12 insertions(+), 8 deletions(-)
diff --git a/include/linux/mutex.h b/include/linux/mutex.h
index 988c020..2b3dcdb 100644
--- a/include/linux/mutex.h
+++ b/include/linux/mutex.h
@@ -57,8 +57,12 @@ struct mutex {
#endif
#ifdef CONFIG_MUTEX_SPIN_ON_OWNER
struct optimistic_spin_queue osq; /* Spinner MCS lock */
- int waiter_spinning;
-#elif defined(CONFIG_SMP)
+#endif
+#ifdef CONFIG_SMP
+ /*
+ * Used by both CONFIG_MUTEX_SPIN_ON_OWNER (depends on CONFIG_SMP)&
+ * !CONFIG_MUTEX_SPIN_ON_OWNER codes.
+ */
int yield_to_waiter;
#endif
#ifdef CONFIG_DEBUG_MUTEXES
diff --git a/kernel/locking/mutex.c b/kernel/locking/mutex.c
index c4261fa..7218835 100644
--- a/kernel/locking/mutex.c
+++ b/kernel/locking/mutex.c
@@ -55,8 +55,8 @@ __mutex_init(struct mutex *lock, const char *name, struct
lock_class_key *key)
mutex_clear_owner(lock);
#ifdef CONFIG_MUTEX_SPIN_ON_OWNER
osq_lock_init(&lock->osq);
- lock->waiter_spinning = false;
-#elif defined(CONFIG_SMP)
+#endif
+#ifdef CONFIG_SMP
lock->yield_to_waiter = false;
#endif
@@ -351,7 +351,7 @@ static bool mutex_optimistic_spin(struct mutex *lock,
* Turn on the waiter spinning flag to discourage the spinner
* from getting the lock.
Might want to update this comment to "Turn on the yield to waiter flag
to discourage optimistic spinners from stealing the lock."
Besides that:
Acked-by: Jason Low<[email protected]>
You are right. I should have made that comment change. As you have a new
patch outstanding, I will wait a bit until it stabilizes before I update
my patch.
Cheers,
Longman
