From: Grzegorz Kulewski <kangur@polcom.net>
To: Con Kolivas <kernel@kolivas.org>
Cc: Andrew Morton <akpm@osdl.org>,
ck@vds.kolivas.org, linux-kernel@vger.kernel.org
Subject: Re: [ck] swap prefetching merge plans
Date: Thu, 23 Mar 2006 08:56:24 +0100 (CET) [thread overview]
Message-ID: <Pine.LNX.4.63.0603230837500.14361@alpha.polcom.net> (raw)
In-Reply-To: <200603231804.36334.kernel@kolivas.org>
On Thu, 23 Mar 2006, Con Kolivas wrote:
> On Thu, 23 Mar 2006 03:53 pm, Andrew Morton wrote:
>> A look at the -mm lineup for 2.6.17:
>
>> mm-implement-swap-prefetching.patch
>> mm-implement-swap-prefetching-fix.patch
>> mm-implement-swap-prefetching-tweaks.patch
>
>> Still don't have a compelling argument for this, IMO.
>
> For those users who feel they do have a compelling argument for it, please
> speak now or I'll end up maintaining this in -ck only forever.
Well, it works for me rather good. After I close some memory demanding
aplication (like game or qemu) that caused huge portion of the memory to
be swapped out to disk, the time needed to swith back to firefox or
eclipse is really really shorter. Also the GUI is more responsive (because
there are nearly none disk accesses needed when I am entering some dialog
box or menu).
Without the patch system is taking up to several seconds to come back and
even after that is feels jerky because it still must get some page from
disk while I do some work.
Also I would really be glad to see the second part of the idea: mirroring
some pages from memory to disk while system is relatively idle. This could
make swapout nearly a no-op and this could really help some workloads if
you are short on RAM. And software engeneerers and programmers are really
working hard to eat all your RAM even if you think that "now you have
really much".
I have Athlon XP 3000+ system with 1GB of RAM and 4GB of swap. It works
pretty well on that configuration. I have so much swap because I am
compiling I my packages (Gentoo) on tmpfs. So during (or after) some big
compilation (like qt, gcc or something) swap prefetching is really helpful
in getting my desktop apps back. (Also Con's scheduler is helpfull
in allowing me to compile things in the background without nearly any
problems and bad interaction with eclipse, openoffice and firefox.
It does not work excelent in all cases but it usually works very well.)
So, not that I have any vote here, I would really glad to have it merged.
More people will be able to read the code and maybe it will become even
better.
Thanks for your work,
Grzegorz Kulewski
next prev parent reply other threads:[~2006-03-23 7:56 UTC|newest]
Thread overview: 43+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-03-23 4:53 -mm " Andrew Morton
2006-03-23 7:04 ` swap prefetching " Con Kolivas
2006-03-23 7:56 ` Grzegorz Kulewski [this message]
2006-03-23 8:01 ` [ck] " jos poortvliet
2006-03-23 8:22 ` Nick Piggin
2006-03-23 9:09 ` Con Kolivas
2006-03-23 9:24 ` Con Kolivas
2006-03-23 13:23 ` Martin J. Bligh
2006-03-23 8:33 ` Thomas Kuther
2006-03-23 11:19 ` Jesper Juhl
2006-03-23 11:24 ` [ck] " André Goddard Rosa
2006-03-23 11:57 ` Spike
2006-03-23 12:03 ` Spike
2006-03-23 12:16 ` Spike
2006-03-23 11:58 ` Lenar Lõhmus
2006-03-23 13:58 ` Francesco Biscani
2006-03-23 15:37 ` James Rayner
2006-03-23 15:34 ` [ck] " Radoslaw Szkodzinski
2006-03-23 18:35 ` Mattia Dongili
2006-03-23 18:40 ` Rafael J. Wysocki
2006-03-25 14:37 ` Jan Engelhardt
2006-03-25 15:24 ` Nick Piggin
2006-03-25 23:44 ` Con Kolivas
2006-03-26 5:54 ` Mike Galbraith
2006-03-26 8:08 ` Con Kolivas
2006-03-26 7:42 ` Mike Galbraith
2006-03-26 9:34 ` Con Kolivas
2006-03-27 15:37 ` [ck] " André Goddard Rosa
2006-03-27 22:19 ` Con Kolivas
2006-03-26 9:20 ` Jan Engelhardt
2006-03-28 6:16 ` [ck] " Chase Venters
2006-03-23 7:59 ` -mm " Ryan M.
2006-03-23 8:09 ` [ck] " Jens Axboe
2006-03-23 9:07 ` Con Kolivas
2006-03-23 12:12 ` [ck] " Jens Axboe
2006-03-23 10:39 ` Takashi Iwai
2006-03-23 12:33 ` Christoph Hellwig
2006-03-23 12:49 ` [PATCH] remove pps support Roman Zippel
2006-03-23 13:43 ` -mm merge plans Michal Piotrowski
2006-03-24 23:33 ` Eric W. Biederman
2006-03-24 23:53 ` Andrew Morton
2006-03-25 0:50 ` Eric W. Biederman
2006-03-25 1:53 ` Arthur Othieno
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=Pine.LNX.4.63.0603230837500.14361@alpha.polcom.net \
--to=kangur@polcom.net \
--cc=akpm@osdl.org \
--cc=ck@vds.kolivas.org \
--cc=kernel@kolivas.org \
--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
all inboxes | Powered by JetHome®