linux-kernel - Re: Excessive stall times on ext4 in 3.9-rc2

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20130402151558.GI32241@suse.de>
Date:	Tue, 2 Apr 2013 16:15:58 +0100
From:	Mel Gorman <mgorman@...e.de>
To:	Zheng Liu <gnehzuil.liu@...il.com>
Cc:	linux-ext4@...r.kernel.org, LKML <linux-kernel@...r.kernel.org>,
	Linux-MM <linux-mm@...ck.org>, Jiri Slaby <jslaby@...e.cz>
Subject: Re: Excessive stall times on ext4 in 3.9-rc2

On Tue, Apr 02, 2013 at 11:03:36PM +0800, Zheng Liu wrote:
> Hi Mel,
> 
> Thanks for reporting it.
> 
> On 04/02/2013 10:27 PM, Mel Gorman wrote:
> > I'm testing a page-reclaim-related series on my laptop that is partially
> > aimed at fixing long stalls when doing metadata-intensive operations on
> > low memory such as a git checkout. I've been running 3.9-rc2 with the
> > series applied but found that the interactive performance was awful even
> > when there was plenty of free memory.
> > 
> > I activated a monitor from mmtests that logs when a process is stuck for
> > a long time in D state and found that there are a lot of stalls in ext4.
> > The report first states that processes have been stalled for a total of
> > 6498 seconds on IO which seems like a lot. Here is a breakdown of the
> > recorded events.
> 
> In this merge window, we add a status tree as a extent cache.  Meanwhile
> a es_cache shrinker is registered to try to reclaim from this cache when
> we are under a high memory pressure. 

Ok.

> So I suspect that the root cause
> is this shrinker.  Could you please tell me how to reproduce this
> problem?  If I understand correctly, I can run mmtest to reproduce this
> problem, right?
> 

This is normal desktop usage with some development thrown in, nothing
spectacular but nothing obviously reproducible either unfortuantely. I
just noticed that some git operations were taking abnormally long, mutt
was very slow opening mail, applications like mozilla were very slow to
launch etc. and dug a little further. I haven't checked if regression
tests under mmtests captured something similar yet.

-- 
Mel Gorman
SUSE Labs
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/