[PATCH 1/2] cgroup: rstat: explicitly put loop variant in while

[PATCH 1/2] cgroup: rstat: explicitly put loop variant in while

[Wei Yang]

Instead of do while unconditionally, let's put the loop variant in
while.

Signed-off-by: Wei Yang <richard.weiyang-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
---
 kernel/cgroup/rstat.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/kernel/cgroup/rstat.c b/kernel/cgroup/rstat.c
index 1486768f2318..a9d344e0521d 100644
--- a/kernel/cgroup/rstat.c
+++ b/kernel/cgroup/rstat.c
@@ -124,12 +124,10 @@ static struct cgroup *cgroup_rstat_cpu_pop_updated(struct cgroup *pos,
 
 			prstatc = cgroup_rstat_cpu(parent, cpu);
 			nextp = &prstatc->updated_children;
-			while (true) {
+			while (*nextp != pos) {
 				struct cgroup_rstat_cpu *nrstatc;
 
 				nrstatc = cgroup_rstat_cpu(*nextp, cpu);
-				if (*nextp == pos)
-					break;
 				WARN_ON_ONCE(*nextp == parent);
 				nextp = &nrstatc->updated_next;
 			}
-- 
2.33.1

[PATCH 2/2] cgroup/rstat: check updated_next only for root

[Wei Yang]

After commit dc26532aed0a ("cgroup: rstat: punt root-level optimization to
individual controllers"), each rstat on updated_children list has its
->updated_next not NULL.

This means we can remove the check on ->updated_next, if we make sure
the subtree from @root is on list, which could be done by checking
updated_next for root.

Signed-off-by: Wei Yang <richard.weiyang-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
---
 kernel/cgroup/rstat.c | 46 +++++++++++++++++++++----------------------
 1 file changed, 22 insertions(+), 24 deletions(-)

diff --git a/kernel/cgroup/rstat.c b/kernel/cgroup/rstat.c
index a9d344e0521d..a00875375f7d 100644
--- a/kernel/cgroup/rstat.c
+++ b/kernel/cgroup/rstat.c
@@ -88,6 +88,7 @@ static struct cgroup *cgroup_rstat_cpu_pop_updated(struct cgroup *pos,
 						   struct cgroup *root, int cpu)
 {
 	struct cgroup_rstat_cpu *rstatc;
+	struct cgroup *parent;
 
 	if (pos == root)
 		return NULL;
@@ -96,9 +97,12 @@ static struct cgroup *cgroup_rstat_cpu_pop_updated(struct cgroup *pos,
 	 * We're gonna walk down to the first leaf and visit/remove it.  We
 	 * can pick whatever unvisited node as the starting point.
 	 */
-	if (!pos)
+	if (!pos) {
 		pos = root;
-	else
+		// return NULL if this subtree is not on-list
+		if (!cgroup_rstat_cpu(pos, cpu)->updated_next)
+			return NULL;
+	} else
 		pos = cgroup_parent(pos);
 
 	/* walk down to the first leaf */
@@ -115,31 +119,25 @@ static struct cgroup *cgroup_rstat_cpu_pop_updated(struct cgroup *pos,
 	 * However, due to the way we traverse, @pos will be the first
 	 * child in most cases. The only exception is @root.
 	 */
-	if (rstatc->updated_next) {
-		struct cgroup *parent = cgroup_parent(pos);
-
-		if (parent) {
-			struct cgroup_rstat_cpu *prstatc;
-			struct cgroup **nextp;
-
-			prstatc = cgroup_rstat_cpu(parent, cpu);
-			nextp = &prstatc->updated_children;
-			while (*nextp != pos) {
-				struct cgroup_rstat_cpu *nrstatc;
-
-				nrstatc = cgroup_rstat_cpu(*nextp, cpu);
-				WARN_ON_ONCE(*nextp == parent);
-				nextp = &nrstatc->updated_next;
-			}
-			*nextp = rstatc->updated_next;
-		}
+	parent = cgroup_parent(pos);
+	if (parent) {
+		struct cgroup_rstat_cpu *prstatc;
+		struct cgroup **nextp;
 
-		rstatc->updated_next = NULL;
-		return pos;
+		prstatc = cgroup_rstat_cpu(parent, cpu);
+		nextp = &prstatc->updated_children;
+		while (*nextp != pos) {
+			struct cgroup_rstat_cpu *nrstatc;
+
+			nrstatc = cgroup_rstat_cpu(*nextp, cpu);
+			WARN_ON_ONCE(*nextp == parent);
+			nextp = &nrstatc->updated_next;
+		}
+		*nextp = rstatc->updated_next;
 	}
 
-	/* only happens for @root */
-	return NULL;
+	rstatc->updated_next = NULL;
+	return pos;
 }
 
 /* see cgroup_rstat_flush() */
-- 
2.33.1

Re: [PATCH 2/2] cgroup/rstat: check updated_next only for root

[Michal Koutný]

On Sat, Dec 25, 2021 at 12:09:32AM +0000, Wei Yang <richard.weiyang-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
> This means we can remove the check on ->updated_next, if we make sure
> the subtree from @root is on list, which could be done by checking
> updated_next for root.

Nice refactoring.

> @@ -96,9 +97,12 @@ static struct cgroup *cgroup_rstat_cpu_pop_updated(struct cgroup *pos,
>  	 * We're gonna walk down to the first leaf and visit/remove it.  We
>  	 * can pick whatever unvisited node as the starting point.
>  	 */
> -	if (!pos)
> +	if (!pos) {
>  		pos = root;
> -	else
> +		// return NULL if this subtree is not on-list
> +		if (!cgroup_rstat_cpu(pos, cpu)->updated_next)
> +			return NULL;
> +	} else
+		/* return NULL if this subtree is not on-list */

Just a coding style nitpick.

The patch is otherwise
Reviewed-by: Michal Koutný <mkoutny-IBi9RG/b67k@public.gmane.org>

Re: [PATCH 2/2] cgroup/rstat: check updated_next only for root

[Wei Yang]

On Wed, Jan 05, 2022 at 08:35:04PM +0100, Michal Koutn? wrote:
>On Sat, Dec 25, 2021 at 12:09:32AM +0000, Wei Yang <richard.weiyang-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
>> This means we can remove the check on ->updated_next, if we make sure
>> the subtree from @root is on list, which could be done by checking
>> updated_next for root.
>
>Nice refactoring.
>
>> @@ -96,9 +97,12 @@ static struct cgroup *cgroup_rstat_cpu_pop_updated(struct cgroup *pos,
>>  	 * We're gonna walk down to the first leaf and visit/remove it.  We
>>  	 * can pick whatever unvisited node as the starting point.
>>  	 */
>> -	if (!pos)
>> +	if (!pos) {
>>  		pos = root;
>> -	else
>> +		// return NULL if this subtree is not on-list
>> +		if (!cgroup_rstat_cpu(pos, cpu)->updated_next)
>> +			return NULL;
>> +	} else
>+		/* return NULL if this subtree is not on-list */
>
>Just a coding style nitpick.

Thanks for comment. Would you like me to send a v2?

>
>The patch is otherwise
>Reviewed-by: Michal Koutn? <mkoutny-IBi9RG/b67k@public.gmane.org>

-- 
Wei Yang
Help you, Help me

Re: [PATCH 2/2] cgroup/rstat: check updated_next only for root

[Tejun Heo]

On Sat, Dec 25, 2021 at 12:09:32AM +0000, Wei Yang wrote:
> After commit dc26532aed0a ("cgroup: rstat: punt root-level optimization to
> individual controllers"), each rstat on updated_children list has its
> ->updated_next not NULL.
> 
> This means we can remove the check on ->updated_next, if we make sure
> the subtree from @root is on list, which could be done by checking
> updated_next for root.
> 
> Signed-off-by: Wei Yang <richard.weiyang-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>

Applied to cgroup/for-5.17 w/ coding style fixes.

Thanks.

-- 
tejun

Re: [PATCH 1/2] cgroup: rstat: explicitly put loop variant in while

[Michal Koutný]

On Sat, Dec 25, 2021 at 12:09:31AM +0000, Wei Yang <richard.weiyang-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
> Instead of do while unconditionally, let's put the loop variant in
> while.
> 
> Signed-off-by: Wei Yang <richard.weiyang-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
> ---
>  kernel/cgroup/rstat.c | 4 +---
>  1 file changed, 1 insertion(+), 3 deletions(-)

Makes sense,
Reviewed-by: Michal Koutný <mkoutny-IBi9RG/b67k@public.gmane.org>

Re: [PATCH 1/2] cgroup: rstat: explicitly put loop variant in while

[Tejun Heo]

On Sat, Dec 25, 2021 at 12:09:31AM +0000, Wei Yang wrote:
> Instead of do while unconditionally, let's put the loop variant in
> while.
> 
> Signed-off-by: Wei Yang <richard.weiyang-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>

Applied to cgorup/for-5.17.

Thanks.

-- 
tejun