[PATCH 2/3] sched: Skip update_blocked_averages if we are defering load balance

From: Vincent Guittot
Date: Fri May 07 2021 - 17:38:10 EST


In newidle_balance(), the scheduler skips load balance to the new idle cpu when sd is this_rq and when

this_rq->avg_idle < sd->max_newidle_lb_cost

Doing a costly call to update_blocked_averages() will
not be useful and simply adds overhead when this condition is true.

Check the condition early in newidle_balance() to skip update_blocked_averages()
when possible.

Signed-off-by: Vincent Guittot <vincent.guittot@xxxxxxxxxx>
Signed-off-by: Tim Chen <tim.c.chen@xxxxxxxxxxxxxxx>
---
kernel/sched/fair.c | 9 ++++++---
1 file changed, 6 insertions(+), 3 deletions(-)

diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index b0b5698b2184..f828b75488a0 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -10612,17 +10612,20 @@ static int newidle_balance(struct rq *this_rq, struct rq_flags *rf)
*/
rq_unpin_lock(this_rq, rf);

+ rcu_read_lock();
+ sd = rcu_dereference_check_sched_domain(this_rq->sd);
+
if (this_rq->avg_idle < sysctl_sched_migration_cost ||
- !READ_ONCE(this_rq->rd->overload)) {
+ !READ_ONCE(this_rq->rd->overload) ||
+ (sd && this_rq->avg_idle < sd->max_newidle_lb_cost)) {

- rcu_read_lock();
- sd = rcu_dereference_check_sched_domain(this_rq->sd);
if (sd)
update_next_balance(sd, &next_balance);
rcu_read_unlock();

goto out;
}
+ rcu_read_unlock();

raw_spin_unlock(&this_rq->lock);

--
2.20.1