sched/deadline: Avoid a superfluous check
authorWanpeng Li <wanpeng.li@linux.intel.com>
Tue, 17 Mar 2015 11:15:30 +0000 (19:15 +0800)
committerIngo Molnar <mingo@kernel.org>
Fri, 27 Mar 2015 08:36:12 +0000 (09:36 +0100)
Since commit 40767b0dc768 ("sched/deadline: Fix deadline parameter
modification handling") we clear the thottled state when switching
from a dl task, therefore we should never find it set in switching to
a dl task.

Signed-off-by: Wanpeng Li <wanpeng.li@linux.intel.com>
[ Improved the changelog. ]
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Juri Lelli <juri.lelli@arm.com>
Link: http://lkml.kernel.org/r/1426590931-4639-1-git-send-email-wanpeng.li@linux.intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
kernel/sched/deadline.c

index 0a81a954c0410595b03c331bdf9c88bc6a6d1852..24c18dc10fd7ade5656c6868f1c587dd1e3f507d 100644 (file)
@@ -1665,14 +1665,6 @@ static void switched_to_dl(struct rq *rq, struct task_struct *p)
 {
        int check_resched = 1;
 
-       /*
-        * If p is throttled, don't consider the possibility
-        * of preempting rq->curr, the check will be done right
-        * after its runtime will get replenished.
-        */
-       if (unlikely(p->dl.dl_throttled))
-               return;
-
        if (task_on_rq_queued(p) && rq->curr != p) {
 #ifdef CONFIG_SMP
                if (p->nr_cpus_allowed > 1 && rq->dl.overloaded &&