[PATCH 4/4] sched/fair: Do not migrate due to a sync wakeup on exit

From: Mel Gorman
Date: Mon Feb 12 2018 - 10:01:34 EST


When a task exits, it notifies the parent that it has exited. This is a sync
wakup and the exiting task may pull the parent towards the wakers CPU. For
even simple workloads like using a shell, it was observed that the shell
is pulled across nodes by exiting processes. This is daft as the parent
may be long-lived and properly placed. This patch special cases a sync
wakeup on exit to avoid pulling tasks across nodes. Testing on a range
of workloads and machines showed very little differences in performance
although there was a small 3% boost on some machines running a shellscript
intensive workload (git regression test suite).

Signed-off-by: Mel Gorman <mgorman@xxxxxxxxxxxxxxxxxxx>
---
kernel/sched/fair.c | 8 +++++++-
1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 28c8d9c91955..50442697b455 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -5710,8 +5710,14 @@ wake_affine_idle(int this_cpu, int prev_cpu, int sync)
if (idle_cpu(this_cpu) && cpus_share_cache(this_cpu, prev_cpu))
return idle_cpu(prev_cpu) ? prev_cpu : this_cpu;

- if (sync && cpu_rq(this_cpu)->nr_running == 1)
+ if (sync && cpu_rq(this_cpu)->nr_running == 1) {
+ /* Avoid tasks exiting pulling parents to new nodes */
+ if ((current->flags & PF_EXITING) &&
+ !cpus_share_cache(this_cpu, prev_cpu))
+ return prev_cpu;
+
return this_cpu;
+ }

return nr_cpumask_bits;
}
--
2.15.1