sched: Make sched_class::get_rr_interval() optional
authorPeter Zijlstra <peterz@infradead.org>
Mon, 27 Jan 2014 10:54:13 +0000 (11:54 +0100)
committerIngo Molnar <mingo@kernel.org>
Tue, 28 Jan 2014 12:08:41 +0000 (13:08 +0100)
Not all classes implement (or can implement) a useful get_rr_interval()
function, default to a 0 time-slice for them.

This fixes a crash reported by Tommi Rantala.

Reported-by: Tommi Rantala <tt.rantala@gmail.com>
Cc: Dave Jones <davej@redhat.com>
Cc: Tommi Rantala <tt.rantala@gmail.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20140127105413.GC11314@laptop.programming.kicks-ass.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
kernel/sched/core.c

index 36c951b7eef87d88056b484814c379245d39c1fb..81343d6bd9cbc27c1eae24382c339bf1f8913711 100644 (file)
@@ -4324,7 +4324,9 @@ SYSCALL_DEFINE2(sched_rr_get_interval, pid_t, pid,
                goto out_unlock;
 
        rq = task_rq_lock(p, &flags);
-       time_slice = p->sched_class->get_rr_interval(rq, p);
+       time_slice = 0;
+       if (p->sched_class->get_rr_interval)
+               time_slice = p->sched_class->get_rr_interval(rq, p);
        task_rq_unlock(rq, p, &flags);
 
        rcu_read_unlock();