Re: [RFC PATCH] memcg: oom: fix totalpages calculation for swappiness==0

[Michal Hocko <mhocko@suse.cz>]

On Mon 15-10-12 10:25:14, KOSAKI Motohiro wrote:
> > diff --git a/Documentation/sysctl/vm.txt b/Documentation/sysctl/vm.txt
> > index 078701f..308fd77 100644
> > --- a/Documentation/sysctl/vm.txt
> > +++ b/Documentation/sysctl/vm.txt
> > @@ -640,6 +640,9 @@ swappiness
> >  This control is used to define how aggressive the kernel will swap
> >  memory pages.  Higher values will increase agressiveness, lower values
> >  decrease the amount of swap.
> > +The value can be used from the [0, 100] range, where 0 means no swapping
> > +at all (even if there is a swap storage enabled) while 100 means that
> > +anonymous pages are reclaimed in the same rate as file pages.
> 
> I think this only correct when memcg. Even if swappiness==0, global reclaim swap
> out anon pages before oom.

Right you are (we really do swap when the file pages are really
low)! Sorry about the confusion. I kind of became if(global_reclaim)
block blind...

Then this really needs a memcg specific documentation fix. What about
the following?
---
>From 59a60705abd2faf9e266a4270bbf302001845588 Mon Sep 17 00:00:00 2001
From: Michal Hocko <mhocko@suse.cz>
Date: Mon, 15 Oct 2012 11:43:56 +0200
Subject: [PATCH] doc: describe memcg swappiness more precisely

since fe35004f (mm: avoid swapping out with swappiness==0) memcg reclaim
stopped swapping out anon pages completely when 0 value is used.
Although this is somehow expected it hasn't been done for a really long
time this way and so it is probably better to be explicit about the
effect. Moreover global reclaim swapps out even when swappiness is 0
to prevent from OOM killer.

Signed-off-by: Michal Hocko <mhocko@suse.cz>
---
 Documentation/cgroups/memory.txt |    4 ++++
 1 file changed, 4 insertions(+)

diff --git a/Documentation/cgroups/memory.txt b/Documentation/cgroups/memory.txt
index c07f7b4..71c4da4 100644
--- a/Documentation/cgroups/memory.txt
+++ b/Documentation/cgroups/memory.txt
@@ -466,6 +466,10 @@ Note:
 5.3 swappiness
 
 Similar to /proc/sys/vm/swappiness, but affecting a hierarchy of groups only.
+Please note that unlike the global swappiness, memcg knob set to 0
+really prevents from any swapping even if there is a swap storage
+available. This might lead to memcg OOM killer if there are no file
+pages to reclaim.
 
 Following cgroups' swappiness can't be changed.
 - root cgroup (uses /proc/sys/vm/swappiness).
-- 
1.7.10.4

-- 
Michal Hocko
SUSE Labs