[v4,3/5] sched/core: Streamlining calls to task_rq_unlock()

Message ID 20180613121711.5018-4-juri.lelli@redhat.com
State New
Headers show
Series
  • [v4,1/5] sched/topology: Add check to backup comment about hotplug lock
Related show

Commit Message

Juri Lelli June 13, 2018, 12:17 p.m.
From: Mathieu Poirier <mathieu.poirier@linaro.org>


Calls to task_rq_unlock() are done several times in function
__sched_setscheduler().  This is fine when only the rq lock needs to be
handled but not so much when other locks come into play.

This patch streamlines the release of the rq lock so that only one
location need to be modified when dealing with more than one lock.

No change of functionality is introduced by this patch.

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>

---
 kernel/sched/core.c | 24 ++++++++++++++----------
 1 file changed, 14 insertions(+), 10 deletions(-)

-- 
2.14.3

Comments

Steven Rostedt June 14, 2018, 1:42 p.m. | #1
On Wed, 13 Jun 2018 14:17:09 +0200
Juri Lelli <juri.lelli@redhat.com> wrote:

> From: Mathieu Poirier <mathieu.poirier@linaro.org>

> 

> Calls to task_rq_unlock() are done several times in function

> __sched_setscheduler().  This is fine when only the rq lock needs to be

> handled but not so much when other locks come into play.

> 

> This patch streamlines the release of the rq lock so that only one

> location need to be modified when dealing with more than one lock.

> 

> No change of functionality is introduced by this patch.

> 

> Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>


Reviewed-by: Steven Rostedt (VMware) <rostedt@goodmis.org>


-- Steve

Patch

diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index d1555185c054..ca788f74259d 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -4237,8 +4237,8 @@  static int __sched_setscheduler(struct task_struct *p,
 	 * Changing the policy of the stop threads its a very bad idea:
 	 */
 	if (p == rq->stop) {
-		task_rq_unlock(rq, p, &rf);
-		return -EINVAL;
+		retval = -EINVAL;
+		goto unlock;
 	}
 
 	/*
@@ -4254,8 +4254,8 @@  static int __sched_setscheduler(struct task_struct *p,
 			goto change;
 
 		p->sched_reset_on_fork = reset_on_fork;
-		task_rq_unlock(rq, p, &rf);
-		return 0;
+		retval = 0;
+		goto unlock;
 	}
 change:
 
@@ -4268,8 +4268,8 @@  static int __sched_setscheduler(struct task_struct *p,
 		if (rt_bandwidth_enabled() && rt_policy(policy) &&
 				task_group(p)->rt_bandwidth.rt_runtime == 0 &&
 				!task_group_is_autogroup(task_group(p))) {
-			task_rq_unlock(rq, p, &rf);
-			return -EPERM;
+			retval = -EPERM;
+			goto unlock;
 		}
 #endif
 #ifdef CONFIG_SMP
@@ -4284,8 +4284,8 @@  static int __sched_setscheduler(struct task_struct *p,
 			 */
 			if (!cpumask_subset(span, &p->cpus_allowed) ||
 			    rq->rd->dl_bw.bw == 0) {
-				task_rq_unlock(rq, p, &rf);
-				return -EPERM;
+				retval = -EPERM;
+				goto unlock;
 			}
 		}
 #endif
@@ -4304,8 +4304,8 @@  static int __sched_setscheduler(struct task_struct *p,
 	 * is available.
 	 */
 	if ((dl_policy(policy) || dl_task(p)) && sched_dl_overflow(p, policy, attr)) {
-		task_rq_unlock(rq, p, &rf);
-		return -EBUSY;
+		retval = -EBUSY;
+		goto unlock;
 	}
 
 	p->sched_reset_on_fork = reset_on_fork;
@@ -4361,6 +4361,10 @@  static int __sched_setscheduler(struct task_struct *p,
 	preempt_enable();
 
 	return 0;
+
+unlock:
+	task_rq_unlock(rq, p, &rf);
+	return retval;
 }
 
 static int _sched_setscheduler(struct task_struct *p, int policy,