Skip to contentSkip to conversation

locking/rtmutex: Optimize setting task running after being blocked

PATCH New Davidlohr Bueso 2 participants 2 messages Feb 2, 2015 – Feb 4, 2015 Locking primitives

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 locking/rtmutex: Optimize setting task running after being blocked

We explicitly mark the task running after returning from a __rt_mutex_slowlock() call, which does the actual sleeping via wait-wake-trylocking. As such, this patch does two things:

(1) refactors the code so that setting current to TASK_RUNNING is done by __rt_mutex_slowlock(), and not by the callers. The downside to this is that it becomes a bit unclear when at what point we block. As such I've added a comment that the task blocks when calling __rt_mutex_slowlock() so readers can figure out when it is running again.

(2) relaxes setting current's state through __set_current_state(), instead of it's more expensive barrier alternative. There was no need for the implied barrier as we're obviously not planning on blocking.

---
 kernel/locking/rtmutex.c | 7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)
kernel/locking/rtmutex.c +3 -4
index 7c98873..3059bc2f 100644
@@ -1130,6 +1130,7 @@ __rt_mutex_slowlock(struct rt_mutex *lock, int state,
 		set_current_state(state);
 	}
 
+	__set_current_state(TASK_RUNNING);
 	return ret;
 }
 
@@ -1188,10 +1189,9 @@ rt_mutex_slowlock(struct rt_mutex *lock, int state,
 	ret = task_blocks_on_rt_mutex(lock, &waiter, current, chwalk);
 
 	if (likely(!ret))
+		/* sleep on the mutex */
 		ret = __rt_mutex_slowlock(lock, state, timeout, &waiter);
 
-	set_current_state(TASK_RUNNING);
-
 	if (unlikely(ret)) {
 		remove_waiter(lock, &waiter);
 		rt_mutex_handle_deadlock(ret, chwalk, &waiter);
@@ -1626,10 +1626,9 @@ int rt_mutex_finish_proxy_lock(struct rt_mutex *lock,
 
 	set_current_state(TASK_INTERRUPTIBLE);
 
+	/* sleep on the mutex */
 	ret = __rt_mutex_slowlock(lock, TASK_INTERRUPTIBLE, to, waiter);
 
-	set_current_state(TASK_RUNNING);
-
 	if (unlikely(ret))
 		remove_waiter(lock, waiter);
 
Signature · 1 line
2.1.4
#1 lore raw
locking/rtmutex: Optimize setting task running after being blocked
Commit-ID:  afffc6c1805d98e08e778cddb644a666e78cfcfd
Gitweb:     http://git.kernel.org/tip/afffc6c1805d98e08e778cddb644a666e78cfcfd
Author:     Davidlohr Bueso <[email protected]>
AuthorDate: Sun, 1 Feb 2015 22:16:24 -0800
Committer:  Ingo Molnar <[email protected]>
CommitDate: Wed, 4 Feb 2015 07:57:42 +0100

locking/rtmutex: Optimize setting task running after being blocked

We explicitly mark the task running after returning from a __rt_mutex_slowlock() call, which does the actual sleeping via wait-wake-trylocking. As such, this patch does two things:

(1) refactors the code so that setting current to TASK_RUNNING is done by __rt_mutex_slowlock(), and not by the callers. The downside to this is that it becomes a bit unclear when at what point we block. As such I've added a comment that the task blocks when calling __rt_mutex_slowlock() so readers can figure out when it is running again.

(2) relaxes setting current's state through __set_current_state(), instead of it's more expensive barrier alternative. There was no need for the implied barrier as we're obviously not planning on blocking.

---
 kernel/locking/rtmutex.c | 7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)
kernel/locking/rtmutex.c +3 -4
index 7c98873..3059bc2f 100644
@@ -1130,6 +1130,7 @@ __rt_mutex_slowlock(struct rt_mutex *lock, int state,
 		set_current_state(state);
 	}
 
+	__set_current_state(TASK_RUNNING);
 	return ret;
 }
 
@@ -1188,10 +1189,9 @@ rt_mutex_slowlock(struct rt_mutex *lock, int state,
 	ret = task_blocks_on_rt_mutex(lock, &waiter, current, chwalk);
 
 	if (likely(!ret))
+		/* sleep on the mutex */
 		ret = __rt_mutex_slowlock(lock, state, timeout, &waiter);
 
-	set_current_state(TASK_RUNNING);
-
 	if (unlikely(ret)) {
 		remove_waiter(lock, &waiter);
 		rt_mutex_handle_deadlock(ret, chwalk, &waiter);
@@ -1626,10 +1626,9 @@ int rt_mutex_finish_proxy_lock(struct rt_mutex *lock,
 
 	set_current_state(TASK_INTERRUPTIBLE);
 
+	/* sleep on the mutex */
 	ret = __rt_mutex_slowlock(lock, TASK_INTERRUPTIBLE, to, waiter);
 
-	set_current_state(TASK_RUNNING);
-
 	if (unlikely(ret))
 		remove_waiter(lock, waiter);