sched: fix rt-bandwidth hotplug race

When we hot-unplug a cpu and rebuild the sched-domain, all cpus will be
detatched. Alex observed the case where a runqueue was stealing bandwidth
from an already disabled runqueue to satisfy its own needs.

Stop this by skipping over already disabled runqueues.

Reported-by: Alex Nixon <alex.nixon@citrix.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Tested-by: Alex Nixon <alex.nixon@citrix.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

diff --git a/kernel/sched_rt.c b/kernel/sched_rt.c
index 6163e4cf885b90075fd2734427fac4f3f24eae1f..998ba54b4543d876a6fa82f4a41debb981923319 100644 (file)
--- a/kernel/sched_rt.c
+++ b/kernel/sched_rt.c
@@ -298,7 +298,7 @@ static void __disable_runtime(struct rq *rq)
                        struct rt_rq *iter = sched_rt_period_rt_rq(rt_b, i);
                        s64 diff;
 
-                       if (iter == rt_rq)
+                       if (iter == rt_rq || iter->rt_runtime == RUNTIME_INF)
                                continue;
 
                        spin_lock(&iter->rt_runtime_lock);