2020-02-27 15:43:59

by Vincent Guittot

Subject: [PATCH v2] sched/fair: fix runnable_avg for throttled cfs

When a cfs_rq is throttled, its group entity is dequeued and its running
tasks are removed. We must update runnable_avg with the old h_nr_running
and update group_se->runnable_weight with the new h_nr_running at each
level of the hierarchy.

Fixes: 9f68395333ad ("sched/pelt: Add a new runnable average signal")
Signed-off-by: Vincent Guittot <[email protected]>
---
This patch applies on top of tip/sched/core

Changes since v1:
- update commit message
- add missing {}

kernel/sched/fair.c | 14 ++++++++++++--
1 file changed, 12 insertions(+), 2 deletions(-)

diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index fcc968669aea..22d067279269 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -4701,8 +4701,13 @@ static void throttle_cfs_rq(struct cfs_rq *cfs_rq)
if (!se->on_rq)
break;

- if (dequeue)
+ if (dequeue) {
dequeue_entity(qcfs_rq, se, DEQUEUE_SLEEP);
+ } else {
+ update_load_avg(qcfs_rq, se, 0);
+ se_update_runnable(se);
+ }
+
qcfs_rq->h_nr_running -= task_delta;
qcfs_rq->idle_h_nr_running -= idle_task_delta;

@@ -4770,8 +4775,13 @@ void unthrottle_cfs_rq(struct cfs_rq *cfs_rq)
enqueue = 0;

cfs_rq = cfs_rq_of(se);
- if (enqueue)
+ if (enqueue) {
enqueue_entity(cfs_rq, se, ENQUEUE_WAKEUP);
+ } else {
+ update_load_avg(cfs_rq, se, 0);
+ se_update_runnable(se);
+ }
+
cfs_rq->h_nr_running += task_delta;
cfs_rq->idle_h_nr_running += idle_task_delta;

--
2.17.1

2020-02-27 19:19:13

by Benjamin Segall

Subject: Re: [PATCH v2] sched/fair: fix runnable_avg for throttled cfs

Vincent Guittot <[email protected]> writes:

> When a cfs_rq is throttled, its group entity is dequeued and its running
> tasks are removed. We must update runnable_avg with the old h_nr_running
> and update group_se->runnable_weight with the new h_nr_running at each
> level of the hierarchy.

Looks good to me.

Reviewed-by: Ben Segall <[email protected]>

>
> Fixes: 9f68395333ad ("sched/pelt: Add a new runnable average signal")
> Signed-off-by: Vincent Guittot <[email protected]>
> ---
> This patch applies on top of tip/sched/core
>
> Changes since v1:
> - update commit message
> - add missing {}
>
> kernel/sched/fair.c | 14 ++++++++++++--
> 1 file changed, 12 insertions(+), 2 deletions(-)
>
> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
> index fcc968669aea..22d067279269 100644
> --- a/kernel/sched/fair.c
> +++ b/kernel/sched/fair.c
> @@ -4701,8 +4701,13 @@ static void throttle_cfs_rq(struct cfs_rq *cfs_rq)
> if (!se->on_rq)
> break;
>
> - if (dequeue)
> + if (dequeue) {
> dequeue_entity(qcfs_rq, se, DEQUEUE_SLEEP);
> + } else {
> + update_load_avg(qcfs_rq, se, 0);
> + se_update_runnable(se);
> + }
> +
> qcfs_rq->h_nr_running -= task_delta;
> qcfs_rq->idle_h_nr_running -= idle_task_delta;
>
> @@ -4770,8 +4775,13 @@ void unthrottle_cfs_rq(struct cfs_rq *cfs_rq)
> enqueue = 0;
>
> cfs_rq = cfs_rq_of(se);
> - if (enqueue)
> + if (enqueue) {
> enqueue_entity(cfs_rq, se, ENQUEUE_WAKEUP);
> + } else {
> + update_load_avg(cfs_rq, se, 0);
> + se_update_runnable(se);
> + }
> +
> cfs_rq->h_nr_running += task_delta;
> cfs_rq->idle_h_nr_running += idle_task_delta;

2020-03-06 14:44:22

by tip-bot2 for Jacob Pan

Subject: [tip: sched/core] sched/fair: Fix runnable_avg for throttled cfs

The following commit has been merged into the sched/core branch of tip:

Commit-ID: 6212437f0f6043e825e021e4afc5cd63e248a2b4
Gitweb: https://git.kernel.org/tip/6212437f0f6043e825e021e4afc5cd63e248a2b4
Author: Vincent Guittot <[email protected]>
AuthorDate: Thu, 27 Feb 2020 16:41:15 +01:00
Committer: Ingo Molnar <[email protected]>
CommitterDate: Fri, 06 Mar 2020 12:57:25 +01:00

sched/fair: Fix runnable_avg for throttled cfs

When a cfs_rq is throttled, its group entity is dequeued and its running
tasks are removed. We must update runnable_avg with the old h_nr_running
and update group_se->runnable_weight with the new h_nr_running at each
level of the hierarchy.

Reviewed-by: Ben Segall <[email protected]>
Signed-off-by: Vincent Guittot <[email protected]>
Signed-off-by: Peter Zijlstra (Intel) <[email protected]>
Signed-off-by: Ingo Molnar <[email protected]>
Fixes: 9f68395333ad ("sched/pelt: Add a new runnable average signal")
Link: https://lkml.kernel.org/r/[email protected]
---
kernel/sched/fair.c | 14 ++++++++++++--
1 file changed, 12 insertions(+), 2 deletions(-)

diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 3887b73..54bd628 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -4720,8 +4720,13 @@ static void throttle_cfs_rq(struct cfs_rq *cfs_rq)
if (!se->on_rq)
break;

- if (dequeue)
+ if (dequeue) {
dequeue_entity(qcfs_rq, se, DEQUEUE_SLEEP);
+ } else {
+ update_load_avg(qcfs_rq, se, 0);
+ se_update_runnable(se);
+ }
+
qcfs_rq->h_nr_running -= task_delta;
qcfs_rq->idle_h_nr_running -= idle_task_delta;

@@ -4789,8 +4794,13 @@ void unthrottle_cfs_rq(struct cfs_rq *cfs_rq)
enqueue = 0;

cfs_rq = cfs_rq_of(se);
- if (enqueue)
+ if (enqueue) {
enqueue_entity(cfs_rq, se, ENQUEUE_WAKEUP);
+ } else {
+ update_load_avg(cfs_rq, se, 0);
+ se_update_runnable(se);
+ }
+
cfs_rq->h_nr_running += task_delta;
cfs_rq->idle_h_nr_running += idle_task_delta;