Re: [RFC][PATCH 1/6] memcg: free all at rmdir

From: Balbir Singh
Date: Wed Nov 12 2008 - 22:27:25 EST


KAMEZAWA Hiroyuki wrote:
> On Thu, 13 Nov 2008 08:44:59 +0530
> Balbir Singh <balbir@xxxxxxxxxxxxxxxxxx> wrote:
>
>> KAMEZAWA Hiroyuki wrote:
>>> On Wed, 12 Nov 2008 16:07:58 -0800
>>> Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> wrote:
>>>> If we do this then we can make the above "keep" behaviour non-optional,
>>>> and the operator gets to choose whether or not to drop the caches
>>>> before doing the rmdir.
>>>>
>>>> Plus, we get a new per-memcg drop_caches capability. And it's a nicer
>>>> interface, and it doesn't have the obvious races which on_rmdir has,
>>>> etc.
>>>>
>>>> hm?
>>>>
>>> Balbir, how would you want to do ?
>>>
>>> I planned to post shrink_uage patch later (it's easy to be implemented) regardless
>>> of acceptance of this patch.
>>>
>>> So, I think we should add shrink_usage now and drop this is a way to go.
>> I am a bit concerned about dropping stuff at will later. Ubuntu 8.10 has memory
>> controller enabled and we exposed memory.force_empty interface there and now
>> we've dropped it (bad on our part). I think we should have deprecated it and
>> dropped it later.
>>
> I *was* documented as "for debug only".
>

I know, but I suspect users won't. I am not blaming you in person, just trying
the find way to do such things. I think I've seen sysfs add a deprecated option
and put all deprecated files there, they are only created if someone cares about
them.

> Hmm, adding force_empty again to do "shrink usage to 0" is another choice.

Yes, sounds reasonable.

--
Balbir
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/