Donate to e Foundation | Murena handsets with /e/OS | Own a part of Murena! Learn more

Commit 9a2dd585 authored by Peter Zijlstra's avatar Peter Zijlstra Committed by Ingo Molnar
Browse files

sched/fair: Implement more accurate async detach 



The problem with the overestimate is that it will subtract too big a
value from the load_sum, thereby pushing it down further than it ought
to go. Since runnable_load_avg is not subject to a similar 'force',
this results in the occasional 'runnable_load > load' situation.

Signed-off-by: default avatarPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-kernel@vger.kernel.org
Signed-off-by: default avatarIngo Molnar <mingo@kernel.org>
parent f207934f

kernel/sched/fair.c

+3 −6
Original line number Original line Diff line number Diff line
@@ -3574,6 +3574,7 @@ update_cfs_rq_load_avg(u64 now, struct cfs_rq *cfs_rq) 




	if (cfs_rq->removed.nr) {

	if (cfs_rq->removed.nr) {

		unsigned long r;

		unsigned long r;

		u32 divider = LOAD_AVG_MAX - 1024 + sa->period_contrib;





		raw_spin_lock(&cfs_rq->removed.lock);

		raw_spin_lock(&cfs_rq->removed.lock);

		swap(cfs_rq->removed.util_avg, removed_util);

		swap(cfs_rq->removed.util_avg, removed_util);
@@ -3582,17 +3583,13 @@ update_cfs_rq_load_avg(u64 now, struct cfs_rq *cfs_rq) 
		cfs_rq->removed.nr = 0;

		cfs_rq->removed.nr = 0;

		raw_spin_unlock(&cfs_rq->removed.lock);

		raw_spin_unlock(&cfs_rq->removed.lock);





		/*

		 * The LOAD_AVG_MAX for _sum is a slight over-estimate,

		 * which is safe due to sub_positive() clipping at 0.

		 */

		r = removed_load;

		r = removed_load;

		sub_positive(&sa->load_avg, r);

		sub_positive(&sa->load_avg, r);

		sub_positive(&sa->load_sum, r * LOAD_AVG_MAX);

		sub_positive(&sa->load_sum, r * divider);





		r = removed_util;

		r = removed_util;

		sub_positive(&sa->util_avg, r);

		sub_positive(&sa->util_avg, r);

		sub_positive(&sa->util_sum, r * LOAD_AVG_MAX);

		sub_positive(&sa->util_sum, r * divider);





		add_tg_cfs_propagate(cfs_rq, -(long)removed_runnable_sum);

		add_tg_cfs_propagate(cfs_rq, -(long)removed_runnable_sum);