public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Andrew Morton <akpm@osdl.org>
To: Rik van Riel <riel@redhat.com>
Cc: cw@f00f.org, piggin@cyberone.com.au, linux-kernel@vger.kernel.org
Subject: Re: [PATCH] vm-fix-all_zones_ok (was Re: 2.6.3-mm3)
Date: Tue, 24 Feb 2004 14:36:18 -0800	[thread overview]
Message-ID: <20040224143618.368dfdca.akpm@osdl.org> (raw)
In-Reply-To: <Pine.LNX.4.44.0402241553550.21522-100000@chimarrao.boston.redhat.com>

Rik van Riel <riel@redhat.com> wrote:
>
> On Tue, 24 Feb 2004, Andrew Morton wrote:
> > Chris Wedgwood <cw@f00f.org> wrote:
> > > On Tue, Feb 24, 2004 at 03:11:40PM +1100, Nick Piggin wrote:
> > > 
> > > > Out of interest, what is the worst you can make it do with
> > > > contrived cases?
> > > 
> > > 700MB slab used.
> > 
> > Sigh.  There is absolutely nothing wrong with having a large slab cache. 
> > And there is nothing necessarily right about having a small one.
> 
> Could it be that the lower zone protection stuff simply means
> that Chris's system only ever allocates page cache and anonymous
> memory from his 600 MB highmem, leaving the 900 MB lowmem for
> the slab cache to roam freely ?

The lower-zone protection code will only preserve an extra megabyte or so
of the normal zone in response to __GFP_HIGHMEM allocations, so it won't be
coming into play here.

I'd prefer to replace the lower-zone/incremental-min code with 2.4's
watermark code.  It's pretty much equivalent, but makes the calculations
once-only and it is easier to observe and understand its effects.  But
there's too much work going on in there to make this change at present.


> I guess highmem allocations really should put some pressure on
> lowmem, even when there is enough lowmem free, because otherwise
> you end up effectively not using half of the memory on 1.5-2 GB
> systems for paging ...

Yup.  With the current -mm patches the reclaim rate from lowmem and highmem
is nicely proportional to each zone's size for pagecache-heavy workloads. 
For lowmem-intensive workloads the reclaim rate from lowmem is higher, as
one would expect.  It seems to be working OK now.


  reply	other threads:[~2004-02-24 22:38 UTC|newest]

Thread overview: 73+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2004-02-23  1:22 2.6.3-mm3 Andrew Morton
2004-02-23  1:43 ` 2.6.3-mm3 Nick Piggin
2004-02-23  1:55   ` 2.6.3-mm3 Andrew Morton
2004-02-23  2:02     ` 2.6.3-mm3 Nick Piggin
2004-02-23  2:51     ` 2.6.3-mm3 Nick Piggin
2004-02-23  3:04       ` 2.6.3-mm3 Nick Piggin
2004-02-23  8:08         ` 2.6.3-mm3 Nick Piggin
2004-02-23  8:48           ` [PATCH] vm-fix-all_zones_ok (was Re: 2.6.3-mm3) Nick Piggin
2004-02-23  8:59             ` Andrew Morton
2004-02-23  9:21               ` Nick Piggin
2004-02-23  9:24                 ` Andrew Morton
2004-02-23 16:23                 ` Bill Davidsen
2004-02-23 22:47               ` Chris Wedgwood
2004-02-24  4:11                 ` Nick Piggin
2004-02-24  9:12                   ` Chris Wedgwood
2004-02-24  9:22                     ` Andrew Morton
2004-02-24  9:30                       ` Chris Wedgwood
2004-02-24  9:37                         ` Andrew Morton
2004-02-24 20:55                       ` Rik van Riel
2004-02-24 22:36                         ` Andrew Morton [this message]
2004-02-24 22:41                           ` Rik van Riel
2004-02-23  2:19 ` 2.6.3-mm3 Christoph Hellwig
2004-02-23  2:52 ` 2.6.3-mm3 Joshua Kwan
2004-02-23  3:34 ` 2.6.3-mm3 Joshua Kwan
2004-02-23 18:09 ` [patch] 2.6.3-mm3: ALSA miXart driver doesn't compile Adrian Bunk
2004-02-23 18:58 ` 2.6.3-mm3 (compile stats) John Cherry
2004-02-23 20:00   ` Sam Ravnborg
2004-02-24 22:22 ` 2.6.3-mm3 Mike Fedyk
2004-02-24 22:30   ` 2.6.3-mm3 Andrew Morton
2004-02-25 21:27     ` 2.6.3-mm3 Mike Fedyk
2004-02-26  1:06       ` 2.6.3-mm3 Nick Piggin
2004-02-26  1:18         ` 2.6.3-mm3 Marc-Christian Petersen
2004-02-26  1:28           ` 2.6.3-mm3 Mike Fedyk
2004-02-26  1:32         ` 2.6.3-mm3 Mike Fedyk
2004-02-26  1:52           ` 2.6.3-mm3 Nick Piggin
2004-02-26  2:34             ` 2.6.3-mm3 Mike Fedyk
2004-02-26  2:40               ` 2.6.3-mm3 Nick Piggin
2004-02-26  2:48                 ` 2.6.3-mm3 Mike Fedyk
2004-02-26  3:05                   ` 2.6.3-mm3 Nick Piggin
2004-02-26  3:19                     ` 2.6.3-mm3 Mike Fedyk
2004-02-26  3:29                       ` 2.6.3-mm3 Nick Piggin
2004-02-26  4:08                         ` 2.6.3-mm3 Mike Fedyk
2004-02-26  4:56             ` 2.6.3-mm3 Mike Fedyk
2004-02-27 19:02               ` 2.6.3-mm3 Mike Fedyk
2004-02-27 21:57                 ` 2.6.3-mm3 Nick Piggin
2004-02-25  0:26 ` 2.6.3-mm3 hangs on boot x440 (scsi?) john stultz
2004-02-25  1:06   ` Andrew Morton
2004-02-25  1:22     ` john stultz
2004-02-25  1:27       ` john stultz
2004-02-25  1:48       ` Andrew Morton
2004-02-25 22:15         ` john stultz
2004-02-26 14:40     ` Go Taniguchi
2004-02-26 21:26       ` john stultz
2004-02-26 23:02         ` john stultz
2004-02-26 23:15           ` Matthew Wilcox
2004-02-27  0:14             ` john stultz
2004-02-27  0:58               ` john stultz
2004-02-27  2:25             ` john stultz
2004-02-25  1:15 ` [PATCH 2.6.3-mm3] serialize_writeback_fdatawait patch Daniel McNeil
2004-02-25  1:43   ` Andrew Morton
2004-02-25 22:56     ` Daniel McNeil
2004-02-25  2:51 ` 2.6.3-mm3 Mike Fedyk
2004-02-25  3:09   ` 2.6.3-mm3 Andrew Morton
2004-02-25  3:34     ` 2.6.3-mm3 Mike Fedyk
2004-02-25 10:47 ` 2.6.3-mm3 sometimes freeze on "sync" Helge Hafting
2004-02-25  9:39   ` Andrew Morton
2004-02-26  8:49     ` Helge Hafting
2004-02-27 14:49       ` Alexander Hoogerhuis
2004-02-27 23:34 ` 2.6.3-mm3 john stultz
2004-02-28  0:06   ` 2.6.3-mm3 Andrew Morton
2004-02-28  2:48     ` 2.6.3-mm3 (ioremap failure w/ _X86_4G and _NUMA) john stultz
2004-02-28  3:18       ` Dave Hansen
2004-02-28  4:34       ` Martin J. Bligh

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=20040224143618.368dfdca.akpm@osdl.org \
    --to=akpm@osdl.org \
    --cc=cw@f00f.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=piggin@cyberone.com.au \
    --cc=riel@redhat.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