[Devel] Re: [RFC/PATCH] cgroup swap subsystem

[Daisuke Nishimura]

Hi.

kamezawa.hiroyu at jp.fujitsu.com wrote:
>>> At first look, remembering mm struct is not very good.
>>> Remembering swap controller itself is better.
>> The swap_cgroup when the page(and page_cgroup) is allocated and
>> the swap_cgroup when the page is going to be swapped out may be
>> different by swap_cgroup_move_task(), so I think swap_cgroup
>> to be charged should be determined at the point of swapout.
>>
> Accounting swap against an entity which allocs anon memory is
> not strange. Problem here is move_task itself.
> Now, charges against anon is not moved when a task which uses it
> is moved. please fix this behavior first if you think this is
> problematic.
> 
> But, finally, a daemon driven by process event connector
> determines the group before process starts using anon. It's
> doubtful that it's worth to add complicated/costly ones.
> 

I agree with you.

I think the current behavior of move_task is problematic,
and should fix it.
But fixing it would be difficult and add a costly process,
so I should consider more.


Thanks,
Daisuke Nishimura.

_______________________________________________
Containers mailing list
Containers at lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/containers