Re: [PATCH v5 09/14] memcg: kmem accounting lifecycle management

[Glauber Costa <glommer-bzQdu9zFT3WakBO8gow8eQ@public.gmane.org>]

On 10/18/2012 03:28 AM, David Rientjes wrote:
> On Tue, 16 Oct 2012, Glauber Costa wrote:
> 
>> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
>> index 1182188..e24b388 100644
>> --- a/mm/memcontrol.c
>> +++ b/mm/memcontrol.c
>> @@ -344,6 +344,7 @@ struct mem_cgroup {
>>  /* internal only representation about the status of kmem accounting. */
>>  enum {
>>  	KMEM_ACCOUNTED_ACTIVE = 0, /* accounted by this cgroup itself */
>> +	KMEM_ACCOUNTED_DEAD, /* dead memcg, pending kmem charges */
> 
> "dead memcg with pending kmem charges" seems better.
> 
ok.

>>  };
>>  
>>  #define KMEM_ACCOUNTED_MASK (1 << KMEM_ACCOUNTED_ACTIVE)
>> @@ -353,6 +354,22 @@ static void memcg_kmem_set_active(struct mem_cgroup *memcg)
>>  {
>>  	set_bit(KMEM_ACCOUNTED_ACTIVE, &memcg->kmem_accounted);
>>  }
>> +
>> +static bool memcg_kmem_is_active(struct mem_cgroup *memcg)
>> +{
>> +	return test_bit(KMEM_ACCOUNTED_ACTIVE, &memcg->kmem_accounted);
>> +}
> 
> I think all of these should be inline.
> 
They end up being, to be best of my knowledge the compiler can and will
inline such simple functions regardless of their marking, unless you
explicitly mark them noinline.


>> +
>> +static void memcg_kmem_mark_dead(struct mem_cgroup *memcg)
>> +{
>> +	if (test_bit(KMEM_ACCOUNTED_ACTIVE, &memcg->kmem_accounted))
>> +		set_bit(KMEM_ACCOUNTED_DEAD, &memcg->kmem_accounted);
>> +}
> 
> The set_bit() doesn't happen atomically with the test_bit(), what 
> synchronization is required for this?
> 

I believe the explanation Michal gave in answer to this is comprehensive.