Skip to contentSkip to conversation

sched/fair: Update min_vruntime for reweight_entity() correctly

PATCH Applied Yiwei Lin 4 participants 4 messages Nov 17, 2023 – Nov 29, 2023 Scheduler

lore
Share & tools
lore

Yiwei Lin

author
#0 lore raw

PATCH 1/1 sched/fair: Update min_vruntime for reweight_entity() correctly

Since reweight_entity() may have chance to change the weight of cfs_rq->curr entity, we should also update_min_vruntime() if this is the case

Fixes eab03c23c2a1 ("sched/eevdf: Fix vruntime adjustment on reweight")
---
 kernel/sched/fair.c | 20 ++++++++++----------
 1 file changed, 10 insertions(+), 10 deletions(-)
kernel/sched/fair.c +10 -10
index 07f555857..6fb89f4a3 100644
@@ -3815,17 +3815,17 @@ static void reweight_entity(struct cfs_rq *cfs_rq, struct sched_entity *se,
 	enqueue_load_avg(cfs_rq, se);
 	if (se->on_rq) {
 		update_load_add(&cfs_rq->load, se->load.weight);
-		if (!curr) {
-			/*
-			 * The entity's vruntime has been adjusted, so let's check
-			 * whether the rq-wide min_vruntime needs updated too. Since
-			 * the calculations above require stable min_vruntime rather
-			 * than up-to-date one, we do the update at the end of the
-			 * reweight process.
-			 */
+		if (!curr)
 			__enqueue_entity(cfs_rq, se);
-			update_min_vruntime(cfs_rq);
-		}
+
+		/*
+		 * The entity's vruntime has been adjusted, so let's check
+		 * whether the rq-wide min_vruntime needs updated too. Since
+		 * the calculations above require stable min_vruntime rather
+		 * than up-to-date one, we do the update at the end of the
+		 * reweight process.
+		 */
+		update_min_vruntime(cfs_rq);
 	}
 }
 
Signature · 1 line
2.34.1
Reviewed-by on v1 1/1
#1 lore raw

On 11/17/23 4:01 PM, Yiwei Lin Wrote:

Quoted 32 lines

Since reweight_entity() may have chance to change the weight of cfs_rq->curr entity, we should also update_min_vruntime() if this is the case

Fixes eab03c23c2a1 ("sched/eevdf: Fix vruntime adjustment on reweight")
---
  kernel/sched/fair.c | 20 ++++++++++----------
  1 file changed, 10 insertions(+), 10 deletions(-)
kernel/sched/fair.c +3 -10
index 07f555857..6fb89f4a3 100644
@@ -3815,17 +3815,17 @@ static void reweight_entity(struct cfs_rq *cfs_rq, struct sched_entity *se,
  	enqueue_load_avg(cfs_rq, se);
  	if (se->on_rq) {
  		update_load_add(&cfs_rq->load, se->load.weight);
-		if (!curr) {
-			/*
-			 * The entity's vruntime has been adjusted, so let's check
-			 * whether the rq-wide min_vruntime needs updated too. Since
-			 * the calculations above require stable min_vruntime rather
-			 * than up-to-date one, we do the update at the end of the
-			 * reweight process.
-			 */
+		if (!curr)
  			__enqueue_entity(cfs_rq, se);
-			update_min_vruntime(cfs_rq);
-		}
+
+		/*

kernel/sched/fair.c static void reweight_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, · line 3831

Quoted 10 lines
kernel/sched/fair.c +7 -0
+		 * The entity's vruntime has been adjusted, so let's check
+		 * whether the rq-wide min_vruntime needs updated too. Since
+		 * the calculations above require stable min_vruntime rather
+		 * than up-to-date one, we do the update at the end of the
+		 * reweight process.
+		 */
+		update_min_vruntime(cfs_rq);
  	}
  }
 

Thanks!

maintainer
#2 lore raw
In reply to Yiwei Lin
Quoted 6 lines · Yiwei Lin

Since reweight_entity() may have chance to change the weight of cfs_rq->curr entity, we should also update_min_vruntime() if this is the case

Fixes eab03c23c2a1 ("sched/eevdf: Fix vruntime adjustment on reweight")

Right, but as you wrote in your other patch, min_vruntime isn't all that critical anymore. I'll go queue this in sched/core, I don't think this warrants /urgent.

Quoted 39 lines
---
 kernel/sched/fair.c | 20 ++++++++++----------
 1 file changed, 10 insertions(+), 10 deletions(-)
kernel/sched/fair.c +10 -10
index 07f555857..6fb89f4a3 100644
@@ -3815,17 +3815,17 @@ static void reweight_entity(struct cfs_rq *cfs_rq, struct sched_entity *se,
 	enqueue_load_avg(cfs_rq, se);
 	if (se->on_rq) {
 		update_load_add(&cfs_rq->load, se->load.weight);
-		if (!curr) {
-			/*
-			 * The entity's vruntime has been adjusted, so let's check
-			 * whether the rq-wide min_vruntime needs updated too. Since
-			 * the calculations above require stable min_vruntime rather
-			 * than up-to-date one, we do the update at the end of the
-			 * reweight process.
-			 */
+		if (!curr)
 			__enqueue_entity(cfs_rq, se);
-			update_min_vruntime(cfs_rq);
-		}
+
+		/*
+		 * The entity's vruntime has been adjusted, so let's check
+		 * whether the rq-wide min_vruntime needs updated too. Since
+		 * the calculations above require stable min_vruntime rather
+		 * than up-to-date one, we do the update at the end of the
+		 * reweight process.
+		 */
+		update_min_vruntime(cfs_rq);
 	}
 }
 
Signature · 1 line
2.34.1
merge bot Applied
#3 lore raw
In reply to Yiwei Lin
The following commit has been merged into the sched/core branch of tip:

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

Commit-ID:     5068d84054b766efe7c6202fc71b2350d1c326f1
Gitweb:        https://git.kernel.org/tip/5068d84054b766efe7c6202fc71b2350d1c326f1
Author:        Yiwei Lin <[email protected]>
AuthorDate:    Fri, 17 Nov 2023 16:01:06 +08:00
Committer:     Peter Zijlstra <[email protected]>
CommitterDate: Wed, 29 Nov 2023 15:43:52 +01:00

sched/fair: Update min_vruntime for reweight_entity() correctly

Since reweight_entity() may have chance to change the weight of cfs_rq->curr entity, we should also update_min_vruntime() if this is the case

Fixes eab03c23c2a1 ("sched/eevdf: Fix vruntime adjustment on reweight")
---
 kernel/sched/fair.c | 20 ++++++++++----------
 1 file changed, 10 insertions(+), 10 deletions(-)
kernel/sched/fair.c +10 -10
index 34fe6e9..bcea3d5 100644
@@ -3815,17 +3815,17 @@ static void reweight_entity(struct cfs_rq *cfs_rq, struct sched_entity *se,
 	enqueue_load_avg(cfs_rq, se);
 	if (se->on_rq) {
 		update_load_add(&cfs_rq->load, se->load.weight);
-		if (!curr) {
-			/*
-			 * The entity's vruntime has been adjusted, so let's check
-			 * whether the rq-wide min_vruntime needs updated too. Since
-			 * the calculations above require stable min_vruntime rather
-			 * than up-to-date one, we do the update at the end of the
-			 * reweight process.
-			 */
+		if (!curr)
 			__enqueue_entity(cfs_rq, se);
-			update_min_vruntime(cfs_rq);
-		}
+
+		/*
+		 * The entity's vruntime has been adjusted, so let's check
+		 * whether the rq-wide min_vruntime needs updated too. Since
+		 * the calculations above require stable min_vruntime rather
+		 * than up-to-date one, we do the update at the end of the
+		 * reweight process.
+		 */
+		update_min_vruntime(cfs_rq);
 	}
 }