[PATCH v3 3/4] memcg: optimize memcg_get_hierarchical_limit

[PATCH v3 3/4] memcg: optimize memcg_get_hierarchical_limit

[Wanpeng Li]

From: Wanpeng Li <liwp-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>

Optimize memcg_get_hierarchical_limit to save cpu cycle.

Signed-off-by: Wanpeng Li <liwp.linux-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
---
 mm/memcontrol.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 2e81328..4520b57 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -3917,9 +3917,9 @@ static void memcg_get_hierarchical_limit(struct mem_cgroup *memcg,
 
 	min_limit = res_counter_read_u64(&memcg->res, RES_LIMIT);
 	min_memsw_limit = res_counter_read_u64(&memcg->memsw, RES_LIMIT);
-	cgroup = memcg->css.cgroup;
 	if (!memcg->use_hierarchy)
 		goto out;
+	cgroup = memcg->css.cgroup;
 
 	while (cgroup->parent) {
 		cgroup = cgroup->parent;
-- 
1.7.9.5

Re: [PATCH v3 3/4] memcg: optimize memcg_get_hierarchical_limit

[Michal Hocko]

On Mon 25-06-12 16:44:32, Wanpeng Li wrote:
> From: Wanpeng Li <liwp-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
> 
> Optimize memcg_get_hierarchical_limit to save cpu cycle.

How did you check that this actually saves some cycles?
There are no (compiler) barriers here so it should reordered.

> Signed-off-by: Wanpeng Li <liwp.linux-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
> ---
>  mm/memcontrol.c |    2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> index 2e81328..4520b57 100644
> --- a/mm/memcontrol.c
> +++ b/mm/memcontrol.c
> @@ -3917,9 +3917,9 @@ static void memcg_get_hierarchical_limit(struct mem_cgroup *memcg,
>  
>  	min_limit = res_counter_read_u64(&memcg->res, RES_LIMIT);
>  	min_memsw_limit = res_counter_read_u64(&memcg->memsw, RES_LIMIT);
> -	cgroup = memcg->css.cgroup;
>  	if (!memcg->use_hierarchy)
>  		goto out;
> +	cgroup = memcg->css.cgroup;
>  
>  	while (cgroup->parent) {
>  		cgroup = cgroup->parent;
> -- 
> 1.7.9.5
> 

-- 
Michal Hocko
SUSE Labs
SUSE LINUX s.r.o.
Lihovarska 1060/12
190 00 Praha 9    
Czech Republic

Re: [PATCH v3 3/4] memcg: optimize memcg_get_hierarchical_limit

[Kamezawa Hiroyuki]

(2012/06/25 17:44), Wanpeng Li wrote:
> From: Wanpeng Li <liwp-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
> 
> Optimize memcg_get_hierarchical_limit to save cpu cycle.
> 
> Signed-off-by: Wanpeng Li <liwp.linux-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>

I can't understand the benefit of this patch to the real world...

So, if you wrote the patch description as "clean up" rather than optimize,
ok, I'll agree.

Thanks,
-Kame

> ---
>   mm/memcontrol.c |    2 +-
>   1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> index 2e81328..4520b57 100644
> --- a/mm/memcontrol.c
> +++ b/mm/memcontrol.c
> @@ -3917,9 +3917,9 @@ static void memcg_get_hierarchical_limit(struct mem_cgroup *memcg,
>   
>   	min_limit = res_counter_read_u64(&memcg->res, RES_LIMIT);
>   	min_memsw_limit = res_counter_read_u64(&memcg->memsw, RES_LIMIT);
> -	cgroup = memcg->css.cgroup;
>   	if (!memcg->use_hierarchy)
>   		goto out;
> +	cgroup = memcg->css.cgroup;
>   
>   	while (cgroup->parent) {
>   		cgroup = cgroup->parent;
> 

Re: [PATCH v3 3/4] memcg: optimize memcg_get_hierarchical_limit

[Wanpeng Li]

On Mon, Jun 25, 2012 at 06:52:25PM +0900, Kamezawa Hiroyuki wrote:
>(2012/06/25 17:44), Wanpeng Li wrote:
>> From: Wanpeng Li <liwp-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
>> 
>> Optimize memcg_get_hierarchical_limit to save cpu cycle.
>> 
>> Signed-off-by: Wanpeng Li <liwp.linux-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
>
>I can't understand the benefit of this patch to the real world...
>
>So, if you wrote the patch description as "clean up" rather than optimize,
>ok, I'll agree.
OK, I will change to "Cleanup". Thank you Kame.

Best Regards,
Wanpeng Li

>
>Thanks,
>-Kame
>
>> ---
>>   mm/memcontrol.c |    2 +-
>>   1 file changed, 1 insertion(+), 1 deletion(-)
>> 
>> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
>> index 2e81328..4520b57 100644
>> --- a/mm/memcontrol.c
>> +++ b/mm/memcontrol.c
>> @@ -3917,9 +3917,9 @@ static void memcg_get_hierarchical_limit(struct mem_cgroup *memcg,
>>   
>>   	min_limit = res_counter_read_u64(&memcg->res, RES_LIMIT);
>>   	min_memsw_limit = res_counter_read_u64(&memcg->memsw, RES_LIMIT);
>> -	cgroup = memcg->css.cgroup;
>>   	if (!memcg->use_hierarchy)
>>   		goto out;
>> +	cgroup = memcg->css.cgroup;
>>   
>>   	while (cgroup->parent) {
>>   		cgroup = cgroup->parent;
>> 
>
>
>

Re: [PATCH v3 3/4] memcg: optimize memcg_get_hierarchical_limit

[Michal Hocko]

On Mon 25-06-12 18:43:57, Wanpeng Li wrote:
> On Mon, Jun 25, 2012 at 06:52:25PM +0900, Kamezawa Hiroyuki wrote:
> >(2012/06/25 17:44), Wanpeng Li wrote:
> >> From: Wanpeng Li <liwp-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
> >> 
> >> Optimize memcg_get_hierarchical_limit to save cpu cycle.
> >> 
> >> Signed-off-by: Wanpeng Li <liwp.linux-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
> >
> >I can't understand the benefit of this patch to the real world...
> >
> >So, if you wrote the patch description as "clean up" rather than optimize,
> >ok, I'll agree.
> OK, I will change to "Cleanup". Thank you Kame.

Does this really clean something up?
-- 
Michal Hocko
SUSE Labs
SUSE LINUX s.r.o.
Lihovarska 1060/12
190 00 Praha 9    
Czech Republic

Re: [PATCH v3 3/4] memcg: optimize memcg_get_hierarchical_limit

[Wanpeng Li]

On Mon, Jun 25, 2012 at 01:09:26PM +0200, Michal Hocko wrote:
>On Mon 25-06-12 18:43:57, Wanpeng Li wrote:
>> On Mon, Jun 25, 2012 at 06:52:25PM +0900, Kamezawa Hiroyuki wrote:
>> >(2012/06/25 17:44), Wanpeng Li wrote:
>> >> From: Wanpeng Li <liwp-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
>> >> 
>> >> Optimize memcg_get_hierarchical_limit to save cpu cycle.
>> >> 
>> >> Signed-off-by: Wanpeng Li <liwp.linux-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
>> >
>> >I can't understand the benefit of this patch to the real world...
>> >
>> >So, if you wrote the patch description as "clean up" rather than optimize,
>> >ok, I'll agree.
>> OK, I will change to "Cleanup". Thank you Kame.
>
>Does this really clean something up?

OK, just drop it and resend the patch. Hopefully, you can applied! :-)

Best Regards,
Wanpeng Li

>-- 
>Michal Hocko
>SUSE Labs
>SUSE LINUX s.r.o.
>Lihovarska 1060/12
>190 00 Praha 9    
>Czech Republic

Re: [PATCH v3 3/4] memcg: optimize memcg_get_hierarchical_limit

[Kamezawa Hiroyuki]

(2012/06/25 20:09), Michal Hocko wrote:
> On Mon 25-06-12 18:43:57, Wanpeng Li wrote:
>> On Mon, Jun 25, 2012 at 06:52:25PM +0900, Kamezawa Hiroyuki wrote:
>>> (2012/06/25 17:44), Wanpeng Li wrote:
>>>> From: Wanpeng Li <liwp-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
>>>>
>>>> Optimize memcg_get_hierarchical_limit to save cpu cycle.
>>>>
>>>> Signed-off-by: Wanpeng Li <liwp.linux-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
>>>
>>> I can't understand the benefit of this patch to the real world...
>>>
>>> So, if you wrote the patch description as "clean up" rather than optimize,
>>> ok, I'll agree.
>> OK, I will change to "Cleanup". Thank you Kame.
>
> Does this really clean something up?
>

I just thought unused variable is trash even if it disappears after complie.

Thanks,
-Kame