Re: [PATCH 1/2] memcg: some modification to softlimit under hierarchical memory reclaim.

[KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>]

On Tue, 29 Sep 2009 11:41:32 +0530
Balbir Singh <balbir@linux.vnet.ibm.com> wrote:

> * KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> [2009-09-29 15:01:41]:
> 
> > No major changes in this patch for 3 weeks.
> > While testing, I found a few css->refcnt bug in softlimit.(and posted patches)
> > But it seems no more (easy) ones.
> >
> 
> Kamezawa-San, this worries me, could you please confirm if you are
> able to see this behaviour without your patches applied as well?

will try just with BUG_ON() for css->refcnt patch.
But it happend only once even with my patch set.
I found the potential bug by review. I checked all css_get/put/tryget
and it's an only candidates.

> I am doing some more stress tests on my side.
>  
"a few" includes my patch and Nishimura-san's patch for refcnt
which fixes css->refcnt leak. Which are already in -rc.

As I said (in merge plan), I have some concerns on softlimit. But it's
not from softlimit code, but from memcg's nature and complication especially
with hierarchy. These 2 years history shows there are tons of race condition.

Thanks,
-Kame

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>