Commit 8dd77156 authored Mar 16, 2015 by Syed Rameez Mustafa Committed by Matt Wagantall May 30, 2015

sched: Ensure attempting load balance when HMP active balance flags are set



find_busiest_group() can end up returning a NULL group due to load based
checks even though there are tasks that can be migrated to higher capacity
CPUs (LBF_BIG_TASK_ACTIVE_BALANCE) or EA core rotation is possible
(LBF_EA_ACTIVE_BALANCE). To get best power and performance ensure that load
balance does attempt to pull tasks when HMP_ACTIVE_BALANCE flag is set.
Since sched boost also falls under the same category club it into the same
generic condition.

Change-Id: I3db7ec200d2a038917b1f2341602eb87b5aed289
Signed-off-by: Syed Rameez Mustafa <rameezmustafa@codeaurora.org>

parent 692a5361

kernel/sched/fair.c

+1 −2

Original line number	Diff line number	Diff line
		@@ -8666,8 +8666,7 @@ static struct sched_group find_busiest_group(struct lb_env env)
		if (!sds.busiest \|\| busiest->sum_nr_running == 0)
		goto out_balanced;

		if (sched_boost() && (capacity(env->dst_rq) >
		group_rq_capacity(sds.busiest)))
		if (env->flags & LBF_HMP_ACTIVE_BALANCE)
		goto force_balance;

		if (bail_inter_cluster_balance(env, &sds))