[PATCH] zsmalloc: do not take class lock in zs_shrinker_count()

[PATCH] zsmalloc: do not take class lock in zs_shrinker_count()

[Sergey Senozhatsky]

We can avoid taking class ->lock around zs_can_compact() in
zs_pages_to_compact(), because the number that we return back
is outdated in general case, by design. We have different
sources that are able to change class's state right after we
return from zs_can_compact() -- ongoing I/O operations, manually
triggered compaction, or two of them happening simultaneously.

We re-do this calculations during compaction on a per class basis
anyway.

zs_unregister_shrinker() will not return until we have an active
shrinker, so classes won't unexpectedly disappear while
zs_pages_to_compact(), invoked by zs_shrinker_count(), iterates
them.

Signed-off-by: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
---
 mm/zsmalloc.c | 2 --
 1 file changed, 2 deletions(-)

diff --git a/mm/zsmalloc.c b/mm/zsmalloc.c
index 1edd8a0..ed64cf5 100644
--- a/mm/zsmalloc.c
+++ b/mm/zsmalloc.c
@@ -1836,9 +1836,7 @@ static unsigned long zs_shrinker_count(struct shrinker *shrinker,
 		if (class->index != i)
 			continue;
 
-		spin_lock(&class->lock);
 		pages_to_free += zs_can_compact(class);
-		spin_unlock(&class->lock);
 	}
 
 	return pages_to_free;
-- 
2.4.6

[PATCH] zsmalloc: do not take class lock in zs_shrinker_count()

[Sergey Senozhatsky]

We can avoid taking class ->lock around zs_can_compact() in
zs_pages_to_compact(), because the number that we return back
is outdated in general case, by design. We have different
sources that are able to change class's state right after we
return from zs_can_compact() -- ongoing I/O operations, manually
triggered compaction, or two of them happening simultaneously.

We re-do this calculations during compaction on a per class basis
anyway.

zs_unregister_shrinker() will not return until we have an active
shrinker, so classes won't unexpectedly disappear while
zs_pages_to_compact(), invoked by zs_shrinker_count(), iterates
them.

Signed-off-by: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
---
 mm/zsmalloc.c | 2 --
 1 file changed, 2 deletions(-)

diff --git a/mm/zsmalloc.c b/mm/zsmalloc.c
index 1edd8a0..ed64cf5 100644
--- a/mm/zsmalloc.c
+++ b/mm/zsmalloc.c
@@ -1836,9 +1836,7 @@ static unsigned long zs_shrinker_count(struct shrinker *shrinker,
 		if (class->index != i)
 			continue;
 
-		spin_lock(&class->lock);
 		pages_to_free += zs_can_compact(class);
-		spin_unlock(&class->lock);
 	}
 
 	return pages_to_free;
-- 
2.4.6

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [PATCH] zsmalloc: do not take class lock in zs_shrinker_count()

[Sergey Senozhatsky]

On (07/16/15 21:00), Sergey Senozhatsky wrote:
> Subject: [PATCH] zsmalloc: do not take class lock in zs_shrinker_count()
> X-Mailer: git-send-email 2.4.6
> 
> We can avoid taking class ->lock around zs_can_compact() in
> zs_pages_to_compact(), because the number that we return back

eek... a leftover

s/zs_pages_to_compact/zs_shrinker_count/

I'll resend the patch later today.

	-ss

> is outdated in general case, by design. We have different
> sources that are able to change class's state right after we
> return from zs_can_compact() -- ongoing I/O operations, manually
> triggered compaction, or two of them happening simultaneously.
> 
> We re-do this calculations during compaction on a per class basis
> anyway.
> 
> zs_unregister_shrinker() will not return until we have an active
> shrinker, so classes won't unexpectedly disappear while
> zs_pages_to_compact(), invoked by zs_shrinker_count(), iterates
> them.
> 
> Signed-off-by: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
> ---
>  mm/zsmalloc.c | 2 --
>  1 file changed, 2 deletions(-)
> 
> diff --git a/mm/zsmalloc.c b/mm/zsmalloc.c
> index 1edd8a0..ed64cf5 100644
> --- a/mm/zsmalloc.c
> +++ b/mm/zsmalloc.c
> @@ -1836,9 +1836,7 @@ static unsigned long zs_shrinker_count(struct shrinker *shrinker,
>  		if (class->index != i)
>  			continue;
>  
> -		spin_lock(&class->lock);
>  		pages_to_free += zs_can_compact(class);
> -		spin_unlock(&class->lock);
>  	}
>  
>  	return pages_to_free;
> -- 
> 2.4.6
> 

Re: [PATCH] zsmalloc: do not take class lock in zs_shrinker_count()

[Sergey Senozhatsky]

On (07/16/15 21:00), Sergey Senozhatsky wrote:
> Subject: [PATCH] zsmalloc: do not take class lock in zs_shrinker_count()
> X-Mailer: git-send-email 2.4.6
> 
> We can avoid taking class ->lock around zs_can_compact() in
> zs_pages_to_compact(), because the number that we return back

eek... a leftover

s/zs_pages_to_compact/zs_shrinker_count/

I'll resend the patch later today.

	-ss

> is outdated in general case, by design. We have different
> sources that are able to change class's state right after we
> return from zs_can_compact() -- ongoing I/O operations, manually
> triggered compaction, or two of them happening simultaneously.
> 
> We re-do this calculations during compaction on a per class basis
> anyway.
> 
> zs_unregister_shrinker() will not return until we have an active
> shrinker, so classes won't unexpectedly disappear while
> zs_pages_to_compact(), invoked by zs_shrinker_count(), iterates
> them.
> 
> Signed-off-by: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
> ---
>  mm/zsmalloc.c | 2 --
>  1 file changed, 2 deletions(-)
> 
> diff --git a/mm/zsmalloc.c b/mm/zsmalloc.c
> index 1edd8a0..ed64cf5 100644
> --- a/mm/zsmalloc.c
> +++ b/mm/zsmalloc.c
> @@ -1836,9 +1836,7 @@ static unsigned long zs_shrinker_count(struct shrinker *shrinker,
>  		if (class->index != i)
>  			continue;
>  
> -		spin_lock(&class->lock);
>  		pages_to_free += zs_can_compact(class);
> -		spin_unlock(&class->lock);
>  	}
>  
>  	return pages_to_free;
> -- 
> 2.4.6
> 

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>