From: Wu Fengguang <fengguang.wu@gmail.com>
To: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: David Woodhouse <dwmw2@infradead.org>,
David Howells <dhowells@redhat.com>,
Minchan Kim <minchan.kim@gmail.com>, Mel Gorman <mel@csn.ul.ie>,
Johannes Weiner <hannes@cmpxchg.org>,
"riel@redhat.com" <riel@redhat.com>,
Andrew Morton <akpm@linux-foundation.org>,
LKML <linux-kernel@vger.kernel.org>,
Christoph Lameter <cl@linux-foundation.org>,
"peterz@infradead.org" <peterz@infradead.org>,
"tytso@mit.edu" <tytso@mit.edu>,
"linux-mm@kvack.org" <linux-mm@kvack.org>,
"elladan@eskimo.com" <elladan@eskimo.com>,
"npiggin@suse.de" <npiggin@suse.de>,
"Barnes, Jesse" <jesse.barnes@intel.com>
Subject: Re: Found the commit that causes the OOMs
Date: Wed, 1 Jul 2009 11:54:15 +0800 [thread overview]
Message-ID: <20090701035415.GA22364@localhost> (raw)
In-Reply-To: <20090701114959.85D3.A69D9226@jp.fujitsu.com>
On Wed, Jul 01, 2009 at 11:51:54AM +0900, KOSAKI Motohiro wrote:
> > > > What is "hidden" anon pages?
> > > > each shrink_{in}active_list isolate 32 pages from lru. it mean anon or file lru
> > > > accounting decrease temporary.
> > > >
> > > > if system have plenty thread or process, heavy memory pressure makes
> > > > #-of-thread x 32pages isolation.
> > > >
> > > > msgctl11 makes >10K processes.
> > >
> > > More exactly, ~16K processes:
> > >
> > > msgctl11 0 INFO : Using upto 16298 pids
> > >
> > > So the maximum number of isolated pages is 16K * 32 = 512K, or 2GiB.
> > >
> > > > I have debugging patch for this case.
> > > > Wu, Can you please try this patch?
> > >
> > > OK. But the OOM is not quite reproducible. Sometimes it produces these
> > > messages:
> >
> > This time I got the OOM: there are 69817 isolated pages (just as expected)!
> >
> (snip)
>
> > [ 1522.019259] Active_anon:11 active_file:6 inactive_anon:0
> > [ 1522.019260] inactive_file:0 unevictable:0 dirty:0 writeback:0 unstable:0
> > [ 1522.019261] free:1985 slab:44399 mapped:132 pagetables:61830 bounce:0
> > [ 1522.019262] isolate:69817
>
> OK. thanks.
> I plan to submit this patch after small more tests. it is useful for OOM analysis.
Other counters to consider are NR_ANON_PAGES/NR_FILE_PAGES.
If they were showed in the oom message, this problem could be found
much earlier. In this case, we'll find that the total file+anon pages
outnumbered the active+inactive file/anon pages.
Thanks,
Fengguang
next prev parent reply other threads:[~2009-07-01 3:54 UTC|newest]
Thread overview: 92+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-05-17 2:23 [PATCH 0/3] make mapped executable pages the first class citizen Wu Fengguang
2009-05-17 2:23 ` [PATCH 1/3] vmscan: report vm_flags in page_referenced() Wu Fengguang
2009-05-17 2:23 ` [PATCH 2/3] vmscan: make mapped executable pages the first class citizen Wu Fengguang
2009-05-19 8:59 ` Wu Fengguang
2009-05-17 2:23 ` [PATCH 3/3] vmscan: merge duplicate code in shrink_active_list() Wu Fengguang
2009-05-18 9:16 ` Wu Fengguang
2009-05-19 2:43 ` Wu Fengguang
2009-05-19 10:18 ` Johannes Weiner
2009-05-19 10:32 ` Wu Fengguang
2009-06-18 14:46 ` [PATCH 0/3] make mapped executable pages the first class citizen David Howells
2009-06-19 5:24 ` Wu Fengguang
2009-06-19 5:58 ` Wu Fengguang
2009-06-19 8:06 ` David Howells
2009-06-18 14:51 ` David Howells
2009-06-18 16:18 ` David Howells
2009-06-18 16:57 ` Andrew Morton
2009-06-20 4:33 ` Wu Fengguang
2009-06-20 8:24 ` David Howells
2009-06-23 14:43 ` David Howells
2009-06-24 1:43 ` KOSAKI Motohiro
2009-06-24 2:32 ` Wu Fengguang
2009-06-24 2:43 ` KOSAKI Motohiro
2009-06-24 2:49 ` Wu Fengguang
2009-06-27 11:53 ` Johannes Weiner
2009-06-27 18:40 ` David Howells
2009-06-24 13:07 ` David Howells
2009-06-27 7:12 ` Found the commit that causes the OOMs David Howells
2009-06-27 12:07 ` Minchan Kim
2009-06-27 12:54 ` Johannes Weiner
2009-06-27 13:50 ` Minchan Kim
2009-06-27 15:36 ` Johannes Weiner
2009-06-28 16:53 ` Minchan Kim
2009-06-27 15:52 ` KOSAKI Motohiro
2009-06-28 11:32 ` Wu Fengguang
2009-06-28 13:30 ` Minchan Kim
2009-06-28 13:36 ` Minchan Kim
2009-06-28 14:22 ` Wu Fengguang
2009-06-28 15:01 ` KOSAKI Motohiro
2009-06-28 15:10 ` Wu Fengguang
2009-06-28 16:50 ` Minchan Kim
2009-06-29 0:17 ` Minchan Kim
2009-06-29 7:34 ` Wu Fengguang
2009-06-29 10:10 ` David Howells
2009-06-29 12:55 ` Wu Fengguang
2009-06-29 14:21 ` David Howells
2009-06-29 15:00 ` Minchan Kim
2009-06-29 15:14 ` Wu Fengguang
2009-06-29 15:54 ` David Howells
2009-06-29 15:56 ` David Woodhouse
2009-06-30 14:05 ` Wu Fengguang
2009-06-30 15:50 ` Minchan Kim
2009-07-01 2:30 ` Wu Fengguang
2009-07-01 1:18 ` KOSAKI Motohiro
2009-07-01 2:13 ` Rik van Riel
2009-07-01 2:16 ` Wu Fengguang
2009-07-01 2:26 ` Wu Fengguang
2009-07-01 2:51 ` KOSAKI Motohiro
2009-07-01 2:57 ` Rik van Riel
2009-07-01 4:06 ` Wu Fengguang
2009-07-01 4:18 ` KOSAKI Motohiro
2009-07-01 4:25 ` Wu Fengguang
2009-07-01 4:30 ` KOSAKI Motohiro
2009-07-01 11:27 ` Wu Fengguang
2009-07-05 9:55 ` Wu Fengguang
2009-07-05 10:38 ` KOSAKI Motohiro
2009-07-05 10:51 ` Wu Fengguang
2009-07-01 3:54 ` Wu Fengguang [this message]
2009-06-29 16:07 ` Mel Gorman
2009-06-30 4:07 ` Minchan Kim
2009-06-30 9:22 ` Mel Gorman
2009-06-30 9:30 ` Minchan Kim
2009-06-30 14:00 ` Minchan Kim
2009-06-30 19:57 ` David Howells
2009-07-02 7:41 ` Minchan Kim
2009-07-02 7:44 ` Minchan Kim
2009-07-02 12:43 ` Wu Fengguang
2009-07-02 14:08 ` Minchan Kim
2009-06-29 15:27 ` David Howells
2009-06-28 14:49 ` KOSAKI Motohiro
2009-06-28 15:04 ` Wu Fengguang
2009-06-28 16:47 ` Minchan Kim
2009-06-29 7:48 ` KOSAKI Motohiro
2009-06-29 9:32 ` Minchan Kim
2009-06-29 12:43 ` David Howells
2009-06-29 12:59 ` Wu Fengguang
2009-06-29 16:57 ` Andrew Morton
2009-06-29 18:54 ` KOSAKI Motohiro
2009-06-29 19:08 ` KOSAKI Motohiro
2009-06-27 18:35 ` David Howells
2009-06-27 18:58 ` David Howells
2009-06-28 7:55 ` David Howells
2009-06-19 5:27 ` [PATCH 0/3] make mapped executable pages the first class citizen Wu Fengguang
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20090701035415.GA22364@localhost \
--to=fengguang.wu@gmail.com \
--cc=akpm@linux-foundation.org \
--cc=cl@linux-foundation.org \
--cc=dhowells@redhat.com \
--cc=dwmw2@infradead.org \
--cc=elladan@eskimo.com \
--cc=hannes@cmpxchg.org \
--cc=jesse.barnes@intel.com \
--cc=kosaki.motohiro@jp.fujitsu.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mel@csn.ul.ie \
--cc=minchan.kim@gmail.com \
--cc=npiggin@suse.de \
--cc=peterz@infradead.org \
--cc=riel@redhat.com \
--cc=tytso@mit.edu \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).