From: Paul Jackson <pj@sgi.com>
To: David Rientjes <rientjes@google.com>
Cc: clameter@sgi.com, Lee.Schermerhorn@hp.com,
akpm@linux-foundation.org, ak@suse.de,
linux-kernel@vger.kernel.org
Subject: Re: [patch 2/2] cpusets: add interleave_over_allowed option
Date: Tue, 30 Oct 2007 17:29:23 -0700 [thread overview]
Message-ID: <20071030172923.d5d8dac3.pj@sgi.com> (raw)
In-Reply-To: <alpine.DEB.0.9999.0710301648540.28111@chino.kir.corp.google.com>
David wrote:
> Of course they have specific affinity needs, that's why they used
> mempolicies.
No. Good grief. If they are just looking for some set of memory
banks, not to other node-specific hardware, then they might not need
a specific node.
Consider for example a multi-threaded, compute bound, long running
scientific computation that has a substantial and fussy memory layout.
Remapping it from one cpuset to another having the same NUMA topology
may well work fine, once its memory caches recover. Reverting it to
the lowest common denominator MPOL_DEFAULT policy because (Choice C) it
no longer has access to its initial nodes might devastate its
performance.
pj wrote:
> I sure wish I knew what real world, actual, not hypothetical, situations
> were motivating this.
I'm still wishing ...
--
I won't rest till it's the best ...
Programmer, Linux Scalability
Paul Jackson <pj@sgi.com> 1.925.600.0401
next prev parent reply other threads:[~2007-10-31 0:29 UTC|newest]
Thread overview: 98+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-10-25 22:54 [patch 1/2] cpusets: extract mmarray loading from update_nodemask David Rientjes
2007-10-25 22:54 ` [patch 2/2] cpusets: add interleave_over_allowed option David Rientjes
2007-10-25 23:37 ` Christoph Lameter
2007-10-25 23:56 ` David Rientjes
2007-10-26 0:28 ` Christoph Lameter
2007-10-26 1:55 ` Paul Jackson
2007-10-26 2:11 ` David Rientjes
2007-10-26 2:29 ` Paul Jackson
2007-10-26 2:45 ` David Rientjes
2007-10-26 3:14 ` Paul Jackson
2007-10-26 3:58 ` David Rientjes
2007-10-26 4:34 ` Paul Jackson
2007-10-26 15:37 ` Lee Schermerhorn
2007-10-26 17:04 ` Paul Jackson
2007-10-26 17:28 ` Lee Schermerhorn
2007-10-26 20:21 ` Michael Kerrisk
2007-10-26 20:25 ` Paul Jackson
2007-10-26 20:33 ` Michael Kerrisk
2007-10-26 15:30 ` Lee Schermerhorn
2007-10-26 18:46 ` David Rientjes
2007-10-26 19:00 ` Paul Jackson
2007-10-26 20:45 ` David Rientjes
2007-10-26 21:05 ` Christoph Lameter
2007-10-26 21:08 ` David Rientjes
2007-10-26 21:12 ` Christoph Lameter
2007-10-26 21:15 ` David Rientjes
2007-10-26 21:13 ` Lee Schermerhorn
2007-10-26 21:17 ` Christoph Lameter
2007-10-26 21:26 ` Lee Schermerhorn
2007-10-26 21:37 ` Christoph Lameter
2007-10-29 15:00 ` Lee Schermerhorn
2007-10-29 17:33 ` Paul Jackson
2007-10-29 17:46 ` Lee Schermerhorn
2007-10-29 20:35 ` Christoph Lameter
2007-10-26 21:18 ` David Rientjes
2007-10-26 21:31 ` Lee Schermerhorn
2007-10-26 21:39 ` David Rientjes
2007-10-27 1:07 ` Paul Jackson
2007-10-27 1:26 ` Christoph Lameter
2007-10-27 2:41 ` Paul Jackson
2007-10-27 2:50 ` Christoph Lameter
2007-10-27 5:16 ` Paul Jackson
2007-10-27 6:07 ` Christoph Lameter
2007-10-27 8:36 ` Paul Jackson
2007-10-27 17:47 ` Christoph Lameter
2007-10-27 20:59 ` Paul Jackson
2007-10-27 17:50 ` David Rientjes
2007-10-27 23:19 ` Paul Jackson
2007-10-28 18:19 ` David Rientjes
2007-10-28 23:46 ` Paul Jackson
2007-10-29 1:04 ` David Rientjes
2007-10-29 4:27 ` Paul Jackson
2007-10-29 4:47 ` David Rientjes
2007-10-29 5:45 ` Paul Jackson
2007-10-29 7:00 ` David Rientjes
2007-10-29 7:26 ` Paul Jackson
2007-10-30 22:53 ` David Rientjes
2007-10-30 23:17 ` Paul Jackson
2007-10-30 23:25 ` David Rientjes
2007-10-31 0:03 ` Paul Jackson
2007-10-31 0:05 ` Paul Jackson
2007-10-29 7:15 ` Paul Jackson
2007-10-30 23:12 ` David Rientjes
2007-10-30 23:44 ` Paul Jackson
2007-10-30 23:53 ` David Rientjes
2007-10-31 0:29 ` Paul Jackson [this message]
2007-10-29 16:54 ` Lee Schermerhorn
2007-10-29 19:40 ` Paul Jackson
2007-10-29 19:45 ` Paul Jackson
2007-10-29 19:57 ` Paul Jackson
2007-10-29 20:02 ` Paul Jackson
2007-10-27 17:45 ` David Rientjes
2007-10-27 21:22 ` Paul Jackson
2007-10-29 15:10 ` Lee Schermerhorn
2007-10-29 18:41 ` Paul Jackson
2007-10-29 19:01 ` Lee Schermerhorn
2007-10-30 23:17 ` David Rientjes
2007-10-31 0:03 ` Paul Jackson
2007-10-30 22:57 ` David Rientjes
2007-10-30 23:46 ` Paul Jackson
2007-10-26 20:43 ` Lee Schermerhorn
2007-10-26 15:18 ` Lee Schermerhorn
2007-10-26 17:36 ` Christoph Lameter
2007-10-26 18:45 ` David Rientjes
2007-10-26 19:02 ` Paul Jackson
2007-10-27 19:16 ` David Rientjes
2007-10-29 16:23 ` Lee Schermerhorn
2007-10-29 17:35 ` Andi Kleen
2007-10-29 19:35 ` Paul Jackson
2007-10-29 20:36 ` Christoph Lameter
2007-10-29 21:08 ` Andi Kleen
2007-10-29 22:48 ` Paul Jackson
2007-10-30 19:47 ` Paul Jackson
2007-10-30 20:20 ` Lee Schermerhorn
2007-10-30 20:26 ` Paul Jackson
2007-10-30 20:27 ` Andi Kleen
2007-10-26 1:13 ` Paul Jackson
2007-10-26 1:30 ` David Rientjes
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=20071030172923.d5d8dac3.pj@sgi.com \
--to=pj@sgi.com \
--cc=Lee.Schermerhorn@hp.com \
--cc=ak@suse.de \
--cc=akpm@linux-foundation.org \
--cc=clameter@sgi.com \
--cc=linux-kernel@vger.kernel.org \
--cc=rientjes@google.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