cpufreq: schedutil: Disable pl by default

* We already done smoothen WALT predicted load boosting To prevent spikes in CPU frequency selection, let's make sure that instead of immediately switching to WALT's predicted load (if > *util), we average it out and produce a smoother, more realistic load boost. Signed-off-by: Shashank Patil <[email protected]>
itsshashanksp · Jun 5, 2024 · 5c75f53 · 5c75f53
1 parent b988c0e
commit 5c75f53
Showing 1 changed file with 0 additions and 2 deletions.
diff --git a/kernel/sched/cpufreq_schedutil.c b/kernel/sched/cpufreq_schedutil.c
@@ -1118,8 +1118,6 @@ static int sugov_init(struct cpufreq_policy *policy)
 
 	tunables->iowait_boost_enable = false;
 
-        tunables->pl = 1;
-
 	switch (policy->cpu) {
 	default:
 	case 0: