From: ebiederm@xmission.com (Eric W. Biederman)
To: "Martin J. Bligh" <mbligh@aracnet.com>
Cc: Bill Davidsen <davidsen@tmr.com>,
Dave McCracken <dmccr@us.ibm.com>, Andrew Morton <akpm@digeo.com>,
Linux Kernel <linux-kernel@vger.kernel.org>,
Linux Memory Management <linux-mm@kvack.org>
Subject: Re: [PATCH 2.5.43-mm2] New shared page table patch
Date: 21 Oct 2002 08:55:24 -0600 [thread overview]
Message-ID: <m1bs5nvo2r.fsf@frodo.biederman.org> (raw)
In-Reply-To: <2458064740.1035069495@[10.10.2.3]>
"Martin J. Bligh" <mbligh@aracnet.com> writes:
> >> For reference, one of the tests was TPC-H. My code reduced the number of
> >> allocated pte_chains from 5 million to 50 thousand.
> >
> > Don't tease, what did that do for performance? I see that someone has
> > already posted a possible problem, and the code would pass for complex for
> > most people, so is the gain worth the pain?
>
> In many cases, this will stop the box from falling over flat on it's
> face due to ZONE_NORMAL exhaustion (from pte-chains), or even total
> RAM exhaustion (from PTEs). Thus the performance gain is infinite ;-)
So why has no one written a pte_chain reaper? It is perfectly sane
to allocate a swap entry and move an entire pte_chain to the swap
cache.
Eric
WARNING: multiple messages have this Message-ID (diff)
From: ebiederm@xmission.com (Eric W. Biederman)
To: "Martin J. Bligh" <mbligh@aracnet.com>
Cc: Bill Davidsen <davidsen@tmr.com>,
Dave McCracken <dmccr@us.ibm.com>, Andrew Morton <akpm@digeo.com>,
Linux Kernel <linux-kernel@vger.kernel.org>,
Linux Memory Management <linux-mm@kvack.org>
Subject: Re: [PATCH 2.5.43-mm2] New shared page table patch
Date: 21 Oct 2002 08:55:24 -0600 [thread overview]
Message-ID: <m1bs5nvo2r.fsf@frodo.biederman.org> (raw)
In-Reply-To: <2458064740.1035069495@[10.10.2.3]>
"Martin J. Bligh" <mbligh@aracnet.com> writes:
> >> For reference, one of the tests was TPC-H. My code reduced the number of
> >> allocated pte_chains from 5 million to 50 thousand.
> >
> > Don't tease, what did that do for performance? I see that someone has
> > already posted a possible problem, and the code would pass for complex for
> > most people, so is the gain worth the pain?
>
> In many cases, this will stop the box from falling over flat on it's
> face due to ZONE_NORMAL exhaustion (from pte-chains), or even total
> RAM exhaustion (from PTEs). Thus the performance gain is infinite ;-)
So why has no one written a pte_chain reaper? It is perfectly sane
to allocate a swap entry and move an entire pte_chain to the swap
cache.
Eric
--
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/
next prev parent reply other threads:[~2002-10-21 14:52 UTC|newest]
Thread overview: 92+ messages / expand[flat|nested] mbox.gz Atom feed top
2002-10-18 23:06 [PATCH 2.5.43-mm2] New shared page table patch Dave McCracken
2002-10-19 1:32 ` Ed Tomlinson
2002-10-19 1:32 ` Ed Tomlinson
2002-10-19 19:17 ` Bill Davidsen
2002-10-19 19:36 ` Dave McCracken
2002-10-19 19:36 ` Dave McCracken
2002-10-20 4:20 ` Eric W. Biederman
2002-10-20 4:20 ` Eric W. Biederman
2002-10-20 6:18 ` Martin J. Bligh
2002-10-20 6:18 ` Martin J. Bligh
2002-10-21 14:55 ` Eric W. Biederman [this message]
2002-10-21 14:55 ` Eric W. Biederman
2002-10-21 15:21 ` Martin J. Bligh
2002-10-21 15:21 ` Martin J. Bligh
2002-10-22 3:54 ` Eric W. Biederman
2002-10-22 3:54 ` Eric W. Biederman
2002-10-22 5:55 ` Martin J. Bligh
2002-10-22 5:55 ` Martin J. Bligh
2002-10-22 16:09 ` Rik van Riel
2002-10-22 16:09 ` Rik van Riel
2002-10-22 16:14 ` Martin J. Bligh
2002-10-22 16:14 ` Martin J. Bligh
2002-10-22 18:06 ` Bill Davidsen
2002-10-22 18:06 ` Bill Davidsen
2002-10-22 19:02 ` Martin J. Bligh
2002-10-22 19:02 ` Martin J. Bligh
2002-10-22 19:21 ` Dave McCracken
2002-10-22 19:21 ` Dave McCracken
2002-10-24 10:50 ` Bill Davidsen
2002-10-24 10:50 ` Bill Davidsen
2002-10-24 14:22 ` Martin J. Bligh
2002-10-24 14:22 ` Martin J. Bligh
2002-10-24 14:38 ` Dave McCracken
2002-10-24 14:38 ` Dave McCracken
2002-10-24 14:51 ` Martin J. Bligh
2002-10-24 14:51 ` Martin J. Bligh
2002-10-25 17:31 ` Bill Davidsen
2002-10-25 17:31 ` Bill Davidsen
2002-10-24 10:42 ` Bill Davidsen
2002-10-24 10:42 ` Bill Davidsen
2002-10-22 17:19 ` Benjamin LaHaise
2002-10-22 17:19 ` Benjamin LaHaise
2002-10-22 17:36 ` Martin J. Bligh
2002-10-22 17:36 ` Martin J. Bligh
2002-10-22 17:45 ` Benjamin LaHaise
2002-10-22 17:45 ` Benjamin LaHaise
2002-10-22 17:56 ` Andrew Morton
2002-10-22 17:56 ` Andrew Morton
2002-10-22 18:01 ` Benjamin LaHaise
2002-10-22 18:01 ` Benjamin LaHaise
2002-10-22 19:03 ` Martin J. Bligh
2002-10-22 19:03 ` Martin J. Bligh
2002-10-22 20:23 ` Alan Cox
2002-10-22 20:23 ` Alan Cox
2002-10-22 21:33 ` Gerrit Huizenga
2002-10-22 21:33 ` Gerrit Huizenga
2002-10-22 18:22 ` Alan Cox
2002-10-22 18:22 ` Alan Cox
2002-10-22 18:47 ` Gerrit Huizenga
2002-10-22 18:47 ` Gerrit Huizenga
2002-10-22 18:55 ` Benjamin LaHaise
2002-10-22 18:55 ` Benjamin LaHaise
2002-10-22 19:27 ` Gerrit Huizenga
2002-10-22 19:27 ` Gerrit Huizenga
2002-10-22 19:29 ` Benjamin LaHaise
2002-10-22 19:29 ` Benjamin LaHaise
2002-10-22 19:56 ` Arjan van de Ven
2002-10-22 14:26 ` Rik van Riel
2002-10-22 14:26 ` Rik van Riel
2002-10-22 17:09 ` Andrew Morton
2002-10-22 17:09 ` Andrew Morton
2002-10-22 17:15 ` Rik van Riel
2002-10-22 17:15 ` Rik van Riel
2002-10-22 18:36 ` Dave McCracken
2002-10-22 18:36 ` Dave McCracken
2002-10-22 18:49 ` Andrew Morton
2002-10-22 18:49 ` Andrew Morton
2002-10-22 19:06 ` Gerrit Huizenga
2002-10-22 19:06 ` Gerrit Huizenga
2002-10-22 19:29 ` Dave McCracken
2002-10-22 19:29 ` Dave McCracken
2002-10-23 3:03 ` Andi Kleen
2002-10-23 3:10 ` Rik van Riel
2002-10-23 17:49 ` Gerrit Huizenga
[not found] ` <Pine.LNX.4.44L.0210221514430.1648-100000@duckman.distro.conecti va>
2002-10-22 19:11 ` Ulrich Drepper
2002-10-22 19:11 ` Ulrich Drepper
2002-10-22 19:29 ` Gerrit Huizenga
2002-10-22 19:29 ` Gerrit Huizenga
2002-10-22 18:45 ` Gerrit Huizenga
2002-10-22 18:45 ` Gerrit Huizenga
2002-10-22 17:54 ` Bill Davidsen
2002-10-22 17:54 ` Bill Davidsen
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=m1bs5nvo2r.fsf@frodo.biederman.org \
--to=ebiederm@xmission.com \
--cc=akpm@digeo.com \
--cc=davidsen@tmr.com \
--cc=dmccr@us.ibm.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mbligh@aracnet.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.