From: Andrew Morton <akpm@linux-foundation.org>
To: Eric St-Laurent <ericstl34@sympatico.ca>
Cc: linux-kernel@vger.kernel.org
Subject: Re: 2.6.23-rc1-mm2 (vm-dont-run-touch_buffer-during-buffercache-lookups.patch)
Date: Wed, 1 Aug 2007 00:46:25 -0700 [thread overview]
Message-ID: <20070801004625.46a24a26.akpm@linux-foundation.org> (raw)
In-Reply-To: <1185953790.6393.20.camel@perkele>
On Wed, 01 Aug 2007 03:36:30 -0400 Eric St-Laurent <ericstl34@sympatico.ca> wrote:
> On Tue, 2007-31-07 at 23:09 -0700, Andrew Morton wrote:
>
> > +vm-dont-run-touch_buffer-during-buffercache-lookups.patch
> >
> > A little VM experiment. See changelog for details.
>
> > We don't have any tests to determine the effects of this, and nobody will
> > bother setting one up, so ho hum, this remains in -mm for ever.
>
> > I don't think there's any point in doing this until we have some decent
> > testcases.
>
>
> Hi Andrew,
>
>
> For which problem this patch was coded?
Good question. I think the current behaviour is just wrong. What the
effect of changing it will be is hard to predict - probably little.
> Is it a potential fix to the
> updatedb problem?
That's one workload which is particularly susceptible to the problemn which
that patch addresses, yes. But in my (brief) testing it didn't make musch
difference.
> Is the patch effective without the filesystem dependant change you talk
> about? (I use reiserfs)
Yes, it'll work as designed with reiserfs.
> I've been thinking about a test case for the updatedb problem:
>
> 1. Script or program that create a large number of directories and zero
> sized files. Same setup for everyone to have reproducible results.
>
> 2. Run updatedb on those.
>
> 3. Observe the effects (with vmstat, slabinfo and meminfo) before,
> during and after the updatedb run.
>
> 4. Do something to trigger some reclaim like copying a large file.
>
> 5. See the effects.
>
>
> What do you think? What would be the ideal test case for the problem in
> your opinion?
Sounds good, yes.
Or you could do something more real-worldly like start up OO, firefox and
friends, then run /etc/cron.daily/everything and see what the
before-and-after effects are. The aggregate info we're looking for is
captured in /proc/meminfo: swapped, Mapped, Cached, Buffers.
next prev parent reply other threads:[~2007-08-01 7:47 UTC|newest]
Thread overview: 96+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-08-01 6:09 2.6.23-rc1-mm2 Andrew Morton
2007-08-01 6:25 ` 2.6.23-rc1-mm2 Paul Mundt
2007-08-01 7:58 ` 2.6.23-rc1-mm2 Mike Frysinger
2007-08-01 8:10 ` 2.6.23-rc1-mm2 Andrew Morton
2007-08-01 13:45 ` 2.6.23-rc1-mm2 Christoph Hellwig
2007-08-01 13:57 ` 2.6.23-rc1-mm2 Jason Wessel
2007-08-01 8:15 ` 2.6.23-rc1-mm2 Paul Mundt
2007-08-01 9:04 ` 2.6.23-rc1-mm2 Mike Frysinger
2007-08-01 12:22 ` 2.6.23-rc1-mm2 Jason Wessel
2007-08-01 7:36 ` 2.6.23-rc1-mm2 (vm-dont-run-touch_buffer-during-buffercache-lookups.patch) Eric St-Laurent
2007-08-01 7:46 ` Andrew Morton [this message]
2007-08-01 8:04 ` Eric St-Laurent
2007-08-01 8:30 ` Andrew Morton
2007-08-01 8:02 ` 2.6.23-rc1-mm2 Mariusz Kozlowski
2007-08-01 8:13 ` 2.6.23-rc1-mm2 Andrew Morton
2007-08-01 8:16 ` 2.6.23-rc1-mm2 Ingo Molnar
2007-08-01 10:23 ` 2.6.23-rc1-mm2 Jiri Kosina
2007-08-02 9:47 ` 2.6.23-rc1-mm2 Mariusz Kozlowski
2007-08-02 14:20 ` [linux-usb-devel] 2.6.23-rc1-mm2 Alan Stern
2007-08-02 14:26 ` Jiri Kosina
2007-08-02 14:32 ` Mariusz Kozlowski
2007-08-01 10:32 ` 2.6.23-rc1-mm2 Paul Mackerras
2007-08-02 10:14 ` 2.6.23-rc1-mm2 Mariusz Kozlowski
2007-08-03 9:39 ` 2.6.23-rc1-mm2 Kumar Gala
2007-08-06 19:12 ` 2.6.23-rc1-mm2 Segher Boessenkool
2007-08-06 19:10 ` 2.6.23-rc1-mm2 Segher Boessenkool
2007-08-01 16:36 ` 2.6.23-rc1-mm2 Greg KH
2007-08-06 19:08 ` 2.6.23-rc1-mm2 Segher Boessenkool
2007-08-06 19:34 ` 2.6.23-rc1-mm2 Mariusz Kozlowski
2007-08-06 21:25 ` 2.6.23-rc1-mm2 Segher Boessenkool
2007-08-06 22:34 ` 2.6.23-rc1-mm2 Mariusz Kozlowski
2007-08-06 23:12 ` 2.6.23-rc1-mm2 Segher Boessenkool
2007-08-01 9:34 ` [PATCH] prevent SSB compilation on s390 part 2 Heiko Carstens
2007-08-01 12:24 ` John W. Linville
2007-08-01 14:43 ` Heiko Carstens
2007-08-01 14:54 ` Michael Buesch
2007-08-01 10:33 ` unionfs compile error ( Re: 2.6.23-rc1-mm2 ) Gabriel C
2007-08-01 17:22 ` Andrew Morton
2007-08-01 17:27 ` Josef Sipek
2007-08-02 16:29 ` Erez Zadok
2007-08-01 17:35 ` Gabriel C
2007-08-01 10:56 ` 2.6.23-rc1-mm2 Gabriel C
2007-08-01 17:26 ` 2.6.23-rc1-mm2 Andrew Morton
2007-08-01 17:39 ` 2.6.23-rc1-mm2 Gabriel C
2007-08-01 11:16 ` [PATCH] fix slown down printk on boot compile error Heiko Carstens
2007-08-01 16:32 ` Randy Dunlap
2007-08-01 13:01 ` drivers/scsi/advansys.c compile error ( Re: 2.6.23-rc1-mm2 ) Gabriel C
2007-08-01 13:39 ` [PATCH] drivers/scsi/advansys.c: fix advansys_board_found compile error Eugene Teo
2007-08-01 13:54 ` Gabriel C
2007-08-01 13:55 ` Matthew Wilcox
2007-08-01 14:27 ` Gabriel C
2007-08-01 14:32 ` Matthew Wilcox
2007-08-01 14:46 ` Gabriel C
2007-08-01 14:23 ` [PATCH -mm] Fix defined but not used warning in drivers/kvm/vmx.c Gabriel C
2007-08-01 18:35 ` Avi Kivity
2007-08-01 15:19 ` [PATCH -mm] Fix a section mismatch warning Gabriel C
2007-08-01 20:13 ` 2.6.23-rc1-mm2 (checks-for-80wire-cable-use-in-pata_via) Laurent Riffard
2007-08-01 20:30 ` 2.6.23-rc1-mm2 Valdis.Kletnieks
2007-08-01 20:40 ` 2.6.23-rc1-mm2 Andrew Morton
2007-08-01 20:52 ` 2.6.23-rc1-mm2 Torsten Kaiser
2007-08-01 21:17 ` 2.6.23-rc1-mm2 Andrew Morton
2007-08-01 23:40 ` 2.6.23-rc1-mm2 Mel Gorman
2007-08-02 4:38 ` 2.6.23-rc1-mm2 Torsten Kaiser
2007-08-02 14:01 ` 2.6.23-rc1-mm2 Andy Whitcroft
2007-08-02 17:44 ` 2.6.23-rc1-mm2 Torsten Kaiser
2007-08-01 23:40 ` INOTIFY=n , AUDIT*=y compile error Gabriel C
2007-08-01 23:59 ` [PATCH -mm] linux-audit list is subscribers-only Gabriel C
2007-08-02 1:30 ` Randy Dunlap
2007-08-02 13:11 ` [PATCH -mm] Fix section mismatch warnings in sound/pci/hda/ Gabriel C
2007-08-02 13:24 ` Takashi Iwai
2007-08-02 16:32 ` Sam Ravnborg
2007-08-02 17:17 ` Takashi Iwai
2007-08-02 17:31 ` 2.6.23-rc1-mm2: Fix crash in sysfs_hash_and_remove Rafael J. Wysocki
2007-08-02 17:34 ` Tejun Heo
2007-08-02 18:19 ` Eric W. Biederman
2007-08-03 11:00 ` 2.6.23-rc1-mm2 Marc Dietrich
2007-08-03 16:38 ` 2.6.23-rc1-mm2 Andrew Morton
2007-08-03 17:03 ` [NFS] 2.6.23-rc1-mm2 Trond Myklebust
2007-08-03 17:21 ` Oleg Nesterov
2007-08-06 6:24 ` Johannes Berg
2007-08-06 10:53 ` Oleg Nesterov
2007-08-06 10:58 ` Johannes Berg
2007-08-06 11:05 ` Marc Dietrich
2007-08-06 11:13 ` Johannes Berg
2007-08-06 16:24 ` Trond Myklebust
2007-08-07 12:09 ` Marc Dietrich
2007-08-07 21:08 ` Trond Myklebust
2007-08-07 21:37 ` Oleg Nesterov
2007-08-07 22:05 ` Trond Myklebust
2007-08-07 22:20 ` Oleg Nesterov
2007-08-07 23:08 ` Trond Myklebust
2007-08-07 23:14 ` Oleg Nesterov
2007-08-08 21:31 ` 2.6.23-rc1-mm2: MMC_ARMMMCI compile error Adrian Bunk
2007-08-09 12:03 ` Pierre Ossman
2007-08-14 21:21 ` [-mm patch] make pm3fb_init() static again Adrian Bunk
2007-08-14 21:21 ` [-mm patch] fs/reiser4/plugin/: make 3 functions static Adrian Bunk
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=20070801004625.46a24a26.akpm@linux-foundation.org \
--to=akpm@linux-foundation.org \
--cc=ericstl34@sympatico.ca \
--cc=linux-kernel@vger.kernel.org \
/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).