Skip to contentSkip to conversation

sched: Fix task affinity for select_task_rq_fair

tip:sched/core tip-bot for Peter Zijlstra 1 participant 1 message Sep 16, 2009

lore
Share & tools
lore
#0 lore raw
Commit-ID:  d7c33c4930f569caf6b2ece597432853c4151a45
Gitweb:     http://git.kernel.org/tip/d7c33c4930f569caf6b2ece597432853c4151a45
Author:     Peter Zijlstra <[email protected]>
AuthorDate: Fri, 11 Sep 2009 12:45:38 +0200
Committer:  Ingo Molnar <[email protected]>
CommitDate: Tue, 15 Sep 2009 16:01:07 +0200

sched: Fix task affinity for select_task_rq_fair

While merging select_task_rq_fair() and sched_balance_self() I made a mistake that leads to testing the wrong task affinty.

LKML-Reference: <new-submission>

---
 kernel/sched_fair.c |    5 ++---
 1 files changed, 2 insertions(+), 3 deletions(-)
kernel/sched_fair.c +2 -3
index 43dc6d1..8b3eddb 100644
@@ -1318,7 +1318,6 @@ find_idlest_cpu(struct sched_group *group, struct task_struct *p, int this_cpu)
  */
 static int select_task_rq_fair(struct task_struct *p, int flag, int sync)
 {
-	struct task_struct *t = current;
 	struct sched_domain *tmp, *sd = NULL;
 	int cpu = smp_processor_id();
 	int prev_cpu = task_cpu(p);
@@ -1393,13 +1392,13 @@ static int select_task_rq_fair(struct task_struct *p, int flag, int sync)
 			continue;
 		}
 
-		group = find_idlest_group(sd, t, cpu);
+		group = find_idlest_group(sd, p, cpu);
 		if (!group) {
 			sd = sd->child;
 			continue;
 		}
 
-		new_cpu = find_idlest_cpu(group, t, cpu);
+		new_cpu = find_idlest_cpu(group, p, cpu);
 		if (new_cpu == -1 || new_cpu == cpu) {
 			/* Now try balancing at a lower domain level of cpu */
 			sd = sd->child;