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

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

[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")
Signed-off-by: Yiwei Lin <s921975628@gmail.com>
---
 kernel/sched/fair.c | 20 ++++++++++----------
 1 file changed, 10 insertions(+), 10 deletions(-)

diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 07f555857..6fb89f4a3 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -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);
 	}
 }
 
-- 
2.34.1

Re: [PATCH] sched/fair: Update min_vruntime for reweight_entity() correctly

[Abel Wu]

On 11/17/23 4:01 PM, Yiwei Lin Wrote:
> 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")
> Signed-off-by: Yiwei Lin <s921975628@gmail.com>
> ---
>   kernel/sched/fair.c | 20 ++++++++++----------
>   1 file changed, 10 insertions(+), 10 deletions(-)
> 
> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
> index 07f555857..6fb89f4a3 100644
> --- a/kernel/sched/fair.c
> +++ b/kernel/sched/fair.c
> @@ -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);
>   	}
>   }
>   

Reviewed-by: Abel Wu <wuyun.abel@bytedance.com>

Thanks!

Re: [PATCH] sched/fair: Update min_vruntime for reweight_entity() correctly

[Peter Zijlstra]

On Fri, Nov 17, 2023 at 04:01:06PM +0800, Yiwei Lin wrote:
> 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")
> Signed-off-by: Yiwei Lin <s921975628@gmail.com>


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.

> ---
>  kernel/sched/fair.c | 20 ++++++++++----------
>  1 file changed, 10 insertions(+), 10 deletions(-)
> 
> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
> index 07f555857..6fb89f4a3 100644
> --- a/kernel/sched/fair.c
> +++ b/kernel/sched/fair.c
> @@ -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);
>  	}
>  }
>  
> -- 
> 2.34.1
> 

[tip: sched/core] sched/fair: Update min_vruntime for reweight_entity() correctly

[tip-bot2 for Yiwei Lin]

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 <s921975628@gmail.com>
AuthorDate:    Fri, 17 Nov 2023 16:01:06 +08:00
Committer:     Peter Zijlstra <peterz@infradead.org>
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")
Signed-off-by: Yiwei Lin <s921975628@gmail.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Abel Wu <wuyun.abel@bytedance.com>
Link: https://lore.kernel.org/r/20231117080106.12890-1-s921975628@gmail.com
---
 kernel/sched/fair.c | 20 ++++++++++----------
 1 file changed, 10 insertions(+), 10 deletions(-)

diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 34fe6e9..bcea3d5 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -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);
 	}
 }
 

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

[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

Signed-off-by: Yiwei Lin <s921975628@gmail.com>
---
 kernel/sched/fair.c | 20 ++++++++++----------
 1 file changed, 10 insertions(+), 10 deletions(-)

diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 07f555857..6fb89f4a3 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -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);
 	}
 }
 
-- 
2.34.1

Re: [PATCH] sched/fair: Update min_vruntime for reweight_entity() correctly

[Abel Wu]

Please add a Fix tag like this:

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

On 11/16/23 7:42 PM, Yiwei Lin Wrote:
> 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
> 
> Signed-off-by: Yiwei Lin <s921975628@gmail.com>
> ---
>   kernel/sched/fair.c | 20 ++++++++++----------
>   1 file changed, 10 insertions(+), 10 deletions(-)
> 
> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
> index 07f555857..6fb89f4a3 100644
> --- a/kernel/sched/fair.c
> +++ b/kernel/sched/fair.c
> @@ -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);
>   	}
>   }
>   

LGTM, thanks!

Re: [PATCH] sched/fair: Update min_vruntime for reweight_entity() correctly

[Abel Wu]

On 11/16/23 7:52 PM, Abel Wu Wrote:
> Please add a Fix tag like this:
> 
> Fixes: eab03c23c2a1 ("sched/eevdf: Fix vruntime adjustment on reweight")

Can you resend this patch with this fix tag please?

> 
> On 11/16/23 7:42 PM, Yiwei Lin Wrote:
>> 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
>>
>> Signed-off-by: Yiwei Lin <s921975628@gmail.com>
>> ---
>>   kernel/sched/fair.c | 20 ++++++++++----------
>>   1 file changed, 10 insertions(+), 10 deletions(-)
>>
>> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
>> index 07f555857..6fb89f4a3 100644
>> --- a/kernel/sched/fair.c
>> +++ b/kernel/sched/fair.c
>> @@ -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);
>>       }
>>   }
> 
> LGTM, thanks!