From: Mel Gorman <mel@csn.ul.ie>
To: Andrew Morton <akpm@linux-foundation.org>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>,
linux-mm@kvack.org, Nick Piggin <npiggin@suse.de>,
Martin Schwidefsky <schwidefsky@de.ibm.com>,
Andreas Krebbel <krebbel@linux.vnet.ibm.com>
Subject: Re: oom-killer killing even if memory is available?
Date: Mon, 23 Mar 2009 11:55:32 +0000 [thread overview]
Message-ID: <20090323115531.GA15416@csn.ul.ie> (raw)
In-Reply-To: <20090320140255.e0c01a59.akpm@linux-foundation.org>
On Fri, Mar 20, 2009 at 02:02:55PM -0700, Andrew Morton wrote:
> On Fri, 20 Mar 2009 15:27:00 +0000 Mel Gorman <mel@csn.ul.ie> wrote:
>
> > >
> > > Something must have allocated (and possibly leaked) it.
> > >
> >
> > This looks like a memory leak all right. There used to be a patch that
> > recorded a stack trace for every page allocation but it was dropped from
> > -mm ages ago because of a merge conflict. I didn't revive it at the time
> > because it wasn't of immediate concern.
> >
> > Should I revive the patch or do we have preferred ways of tracking down
> > memory leaks these days?
>
> We know that a dentry is getting leaked but afaik we don't know which one
> or why.
>
> We could get more info via the page-owner-tracking-leak-detector.patch
> approach, or by dumping the info in the cached dentries - I think Wu
> Fengguang prepared a patch which does that.
>
Looks like it
> I'm not sure why I dropped page-owner-tracking-leak-detector.patch actually
> - it was pretty useful sometimes and afaik we still haven't merged any tool
> which duplicates it.
>
The note I got at the time was "This patch was dropped because procfs
changes broke it".
> Here's the latest version which I have:
>
That matches what I have. I'll check and see can I figure out what broke
with it.
> From: Alexander Nyberg <alexn@dsv.su.se>
>
> Introduces CONFIG_PAGE_OWNER that keeps track of the call chain under which a
> page was allocated. Includes a user-space helper in
> Documentation/page_owner.c to sort the enormous amount of output that this may
> give (thanks tridge).
>
> <SNIP>
--
Mel Gorman
Part-time Phd Student Linux Technology Center
University of Limerick IBM Dublin Software Lab
--
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>
next prev parent reply other threads:[~2009-03-23 10:51 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-03-17 9:00 oom-killer killing even if memory is available? Heiko Carstens
2009-03-17 9:46 ` Andrew Morton
2009-03-17 10:17 ` Heiko Carstens
2009-03-17 10:28 ` Heiko Carstens
2009-03-17 10:49 ` Nick Piggin
2009-03-17 11:39 ` Heiko Carstens
2009-03-20 5:08 ` Wu Fengguang
2009-03-20 15:27 ` Mel Gorman
2009-03-20 21:02 ` Andrew Morton
2009-03-23 11:55 ` Mel Gorman [this message]
2009-03-23 14:58 ` Mel Gorman
2009-03-17 9:51 ` Nick Piggin
2009-03-17 10:11 ` Heiko Carstens
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=20090323115531.GA15416@csn.ul.ie \
--to=mel@csn.ul.ie \
--cc=akpm@linux-foundation.org \
--cc=heiko.carstens@de.ibm.com \
--cc=krebbel@linux.vnet.ibm.com \
--cc=linux-mm@kvack.org \
--cc=npiggin@suse.de \
--cc=schwidefsky@de.ibm.com \
/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).