Skip to contentSkip to conversation

sched/fair: simplify the work when reweighting entity

RFC PATCH Superseded Jiang Biao 3 participants 3 messages Aug 6, 2020 – Aug 11, 2020 Scheduler

v2 of 3 Superseded Latest is v3 →

lore
Share & tools
lore

Jiang Biao

author
#0 lore raw

RFC PATCH v2 1/1 sched/fair: simplify the work when reweighting entity

From: Jiang Biao <[email protected]>

If a se is on_rq when reweighting entity, all we need should be updating the load of cfs_rq, other dequeue/enqueue work could be redundant, such as,
* nr_running--/nr_running++

Even though the following dequeue/enqueue path would never be reached
* account_numa_dequeue/account_numa_enqueue
* list_del/list_add from/into cfs_tasks
but it could be a little confusing.

Simplifying the logic could be helpful to reduce a litte overhead for hot path, and make it cleaner and more readable.

---
v2<-v1:
- Amend the commit log.

 kernel/sched/fair.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
kernel/sched/fair.c +2 -2
index 04fa8dbcfa4d..18a8fc7bd0de 100644
@@ -3086,7 +3086,7 @@ static void reweight_entity(struct cfs_rq *cfs_rq, struct sched_entity *se,
 		/* commit outstanding execution time */
 		if (cfs_rq->curr == se)
 			update_curr(cfs_rq);
-		account_entity_dequeue(cfs_rq, se);
+		update_load_sub(&cfs_rq->load, se->load.weight);
 	}
 	dequeue_load_avg(cfs_rq, se);
 
@@ -3102,7 +3102,7 @@ static void reweight_entity(struct cfs_rq *cfs_rq, struct sched_entity *se,
 
 	enqueue_load_avg(cfs_rq, se);
 	if (se->on_rq)
-		account_entity_enqueue(cfs_rq, se);
+		update_load_add(&cfs_rq->load, se->load.weight);
 
 }
 
Signature · 1 line
2.21.0
#1 lore raw
Quoted 7 lines · Jiang Biao

From: Jiang Biao <[email protected]>

If a se is on_rq when reweighting entity, all we need should be updating the load of cfs_rq, other dequeue/enqueue work could be redundant, such as,
* nr_running--/nr_running++

Quoted 7 lines

Even though the following dequeue/enqueue path would never be reached
* account_numa_dequeue/account_numa_enqueue
* list_del/list_add from/into cfs_tasks
but it could be a little confusing.

Simplifying the logic could be helpful to reduce a litte overhead for hot path, and make it cleaner and more readable.

LGTM. I guess you have to resend it w/o the RFC tag.

Maybe you can rework the patch header a little?

Something like this:

The code in reweight_entity() can be simplified.

For a sched entity on the rq, the entity accounting can be replaced by cfs_rq instantaneous load updates currently called from within the entity accounting.

Even though an entity on the rq can't represent a task in reweight_entity() (a task is always dequeued before calling this function) and so the numa task accounting and the rq->cfs_tasks list management of the entity accounting are never called, the redundant cfs_rq->nr_running decrement/increment will be avoided.

Quoted 30 lines

---
v2<-v1:
- Amend the commit log.

 kernel/sched/fair.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
kernel/sched/fair.c +2 -2
index 04fa8dbcfa4d..18a8fc7bd0de 100644
@@ -3086,7 +3086,7 @@ static void reweight_entity(struct cfs_rq *cfs_rq, struct sched_entity *se,
 		/* commit outstanding execution time */
 		if (cfs_rq->curr == se)
 			update_curr(cfs_rq);
-		account_entity_dequeue(cfs_rq, se);
+		update_load_sub(&cfs_rq->load, se->load.weight);
 	}
 	dequeue_load_avg(cfs_rq, se);
 
@@ -3102,7 +3102,7 @@ static void reweight_entity(struct cfs_rq *cfs_rq, struct sched_entity *se,
 
 	enqueue_load_avg(cfs_rq, se);
 	if (se->on_rq)
-		account_entity_enqueue(cfs_rq, se);
+		update_load_add(&cfs_rq->load, se->load.weight);
 
 }
#2 lore raw

Re: [PATCH RFC v2] sched/fair: simplify the work when reweighting entity(Internet mail)

Hi,

Quoted 15 lines · Dietmar Eggemann
Quoted 12 lines · Jiang Biao

From: Jiang Biao <[email protected]>

If a se is on_rq when reweighting entity, all we need should be updating the load of cfs_rq, other dequeue/enqueue work could be redundant, such as,
* nr_running--/nr_running++

Even though the following dequeue/enqueue path would never be reached
* account_numa_dequeue/account_numa_enqueue
* list_del/list_add from/into cfs_tasks
but it could be a little confusing.

Quoted 4 lines
Quoted 2 lines

Simplifying the logic could be helpful to reduce a litte overhead for hot path, and make it cleaner and more readable.

LGTM. I guess you have to resend it w/o the RFC tag.

Will do soon.

Quoted 11 lines

Maybe you can rework the patch header a little?

Something like this:

The code in reweight_entity() can be simplified.

For a sched entity on the rq, the entity accounting can be replaced by cfs_rq instantaneous load updates currently called from within the entity accounting.

Quoted 5 lines

Even though an entity on the rq can't represent a task in reweight_entity() (a task is always dequeued before calling this function) and so the numa task accounting and the rq->cfs_tasks list management of the entity accounting are never called, the redundant cfs_rq->nr_running decrement/increment will be avoided.

That’s a much better. I’ll pick the header if you don’t mind. :)

Thanks a lot for your kindly reply and detailed explanation.

Regards,
Jiang

Quoted 31 lines
Quoted 31 lines

---
v2<-v1:
- Amend the commit log.

kernel/sched/fair.c | 4 ++--

1 file changed, 2 insertions(+), 2 deletions(-)
kernel/sched/fair.c +2 -2
index 04fa8dbcfa4d..18a8fc7bd0de 100644
@@ -3086,7 +3086,7 @@ static void reweight_entity(struct cfs_rq *cfs_rq, struct sched_entity *se,
 		/* commit outstanding execution time */
 		if (cfs_rq->curr == se)
 			update_curr(cfs_rq);
-		account_entity_dequeue(cfs_rq, se);
+		update_load_sub(&cfs_rq->load, se->load.weight);
 	}
 	dequeue_load_avg(cfs_rq, se);
 
@@ -3102,7 +3102,7 @@ static void reweight_entity(struct cfs_rq *cfs_rq, struct sched_entity *se,
 
 	enqueue_load_avg(cfs_rq, se);
 	if (se->on_rq)
-		account_entity_enqueue(cfs_rq, se);
+		update_load_add(&cfs_rq->load, se->load.weight);
 
 }