[tip:sched/core] sched/balancing: Fix 'local->avg_load > sds->avg_load' case in calculate_imbalance()

From: tip-bot for Vladimir Davydov
Date: Fri Sep 20 2013 - 09:47:24 EST


Commit-ID: b18855500fc40da050512d9df82d2f1471e59642
Gitweb: http://git.kernel.org/tip/b18855500fc40da050512d9df82d2f1471e59642
Author: Vladimir Davydov <vdavydov@xxxxxxxxxxxxx>
AuthorDate: Sun, 15 Sep 2013 17:49:13 +0400
Committer: Ingo Molnar <mingo@xxxxxxxxxx>
CommitDate: Fri, 20 Sep 2013 11:59:36 +0200

sched/balancing: Fix 'local->avg_load > sds->avg_load' case in calculate_imbalance()

In busiest->group_imb case we can come to calculate_imbalance() with
local->avg_load >= busiest->avg_load >= sds->avg_load. This can result
in imbalance overflow, because it is calculated as follows

env->imbalance = min(
max_pull * busiest->group_power,
(sds->avg_load - local->avg_load) * local->group_power) / SCHED_POWER_SCALE;

As a result we can end up constantly bouncing tasks from one cpu to
another if there are pinned tasks.

Fix this by skipping the assignment and assuming imbalance=0 in case
local->avg_load > sds->avg_load.

[ The bug can be caught by running 2*N cpuhogs pinned to two logical cpus
belonging to different cores on an HT-enabled machine with N logical
cpus: just look at se.nr_migrations growth. ]

Signed-off-by: Vladimir Davydov <vdavydov@xxxxxxxxxxxxx>
Signed-off-by: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
Link: http://lkml.kernel.org/r/8f596cc6bc0e5e655119dc892c9bfcad26e971f4.1379252740.git.vdavydov@xxxxxxxxxxxxx
Signed-off-by: Ingo Molnar <mingo@xxxxxxxxxx>
---
kernel/sched/fair.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 11cd136..0b99aae 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -4896,7 +4896,8 @@ static inline void calculate_imbalance(struct lb_env *env, struct sd_lb_stats *s
* max load less than avg load(as we skip the groups at or below
* its cpu_power, while calculating max_load..)
*/
- if (busiest->avg_load < sds->avg_load) {
+ if (busiest->avg_load <= sds->avg_load ||
+ local->avg_load >= sds->avg_load) {
env->imbalance = 0;
return fix_small_imbalance(env, sds);
}
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/