[Devel] Re: [PATCH v3 16/28] memcg: kmem controller charge/uncharge infrastructure

[Glauber Costa]

On 05/30/2012 04:17 PM, Frederic Weisbecker wrote:
> On Fri, May 25, 2012 at 05:03:36PM +0400, Glauber Costa wrote:
>>   #endif /* CONFIG_CGROUP_MEM_RES_CTLR_KMEM */
>> +static __always_inline struct kmem_cache *
>> +mem_cgroup_get_kmem_cache(struct kmem_cache *cachep, gfp_t gfp)
>> +{
>> +	if (!mem_cgroup_kmem_on)
>> +		return cachep;
>> +	if (!current->mm)
>> +		return cachep;
>> +	if (in_interrupt())
>> +		return cachep;
>
> Does that mean interrupts are kept out of accounting?

Well, since interrupts have no process context, if you are in an 
interrupt I can't think of any sane thing to do than relay it to the 
root memcg. That's what happen when I return cachep: I return the 
original parent cache, and we allocate from that.