Re: [BUGFIX][PATCH 5/5] memcg: fix percpu cached charge drainingfrequency

[KAMEZAWA Hiroyuki]

On Wed, 15 Jun 2011 09:12:45 +0900
KAMEZAWA Hiroyuki <kamezawa.hiroyu@xxxxxxxxxxxxxx> wrote:

> On Tue, 14 Jun 2011 09:36:51 +0200
> Michal Hocko <mhocko@xxxxxxx> wrote:
> 
> > On Mon 13-06-11 12:16:48, KAMEZAWA Hiroyuki wrote:
> > > From 18b12e53f1cdf6d7feed1f9226c189c34866338c Mon Sep 17 00:00:00 2001
> > > From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@xxxxxxxxxxxxxx>
> > > Date: Mon, 13 Jun 2011 11:25:43 +0900
> > > Subject: [PATCH 5/5] memcg: fix percpu cached charge draining frequency
> > > 
> > >  For performance, memory cgroup caches some "charge" from res_counter
> > >  into per cpu cache. This works well but because it's cache,
> > >  it needs to be flushed in some cases. Typical cases are
> > >          1. when someone hit limit.
> > >          2. when rmdir() is called and need to charges to be 0.
> > > 
> > > But "1" has problem.
> > > 
> > > Recently, with large SMP machines, we see many kworker runs because
> > > of flushing memcg's cache. Bad things in implementation are
> > > that even if a cpu contains a cache for memcg not related to
> > > a memcg which hits limit, drain code is called.
> > > 
> > > This patch does
> > > 	D) don't call at softlimit reclaim.
> > 
> > I think this needs some justification. The decision is not that
> > obvious IMO. I would say that this is a good decision because cached
> > charges will not help to free any memory (at least not directly) during
> > background reclaim. What about something like:
> > "
> > We are not draining per cpu cached charges during soft limit reclaim 
> > because background reclaim doesn't care about charges. It tries to free
> > some memory and charges will not give any.
> > Cached charges might influence only selection of the biggest soft limit
> > offender but as the call is done only after the selection has been
> > already done it makes no change.
> > "
> > 
> > Anyway, wouldn't it be better to have this change separate from the
> > async draining logic change?
> 
> Hmm. I think calling "draining" at softlimit is just a bug.
> 
I'll divide patches.

Thanks,
-kame

--
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/