public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Robert Love <rlove@rlove.org>
To: "J.A. Magallon" <jamagallon@able.es>
Cc: linux-kernel@vger.kernel.org
Subject: Re: 2.6.11-mm2
Date: Tue, 08 Mar 2005 18:44:02 -0500	[thread overview]
Message-ID: <1110325442.30255.8.camel@localhost> (raw)
In-Reply-To: <1110325018l.6106l.0l@werewolf.able.es>

On Tue, 2005-03-08 at 23:36 +0000, J.A. Magallon wrote:

> Can cpu affinity really be changed for a running process ?

Yes.

> Does it need something like io or yielding to take effect ?

No.

> I am playin with Robert Love's taskset (symlinked to runon, it is easier
> to type and I'm more used to it), because I want to play with hyperthreading
> and wanted a method to force two threads on the same physical package.
> It works fine to bound a new process to a cpu set, but I does not change
> anything for a running process.
> 
> I try runon -c -p 0 <pid> for my numbercruncher and it does nothing, top
> shows it is in the same cpus where it started:
> 
> werewolf:~# runon -c -p 0 8277
> pid 8277's current affinity list: 0-3
> pid 8277's new affinity list: 0
> werewolf:~# runon -c -p 8277
> pid 8277's current affinity list: 0

This looks fine.  As expected.

Although, you have the syntax wrong.  It should be

	taskset -c 0 -p 8277

and

	taskset -p 8277

> The program uses posix threads, 2 in this case. The two threads change from
> cpu sometimes (not too often), but do not go into the same processor
> immediately as when I start the program directly with runon/taskset.

You have to bind all of the threads individually.

	Robert Love



  reply	other threads:[~2005-03-08 23:51 UTC|newest]

Thread overview: 30+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2005-03-08 11:38 2.6.11-mm2 Andrew Morton
2005-03-08 13:58 ` 2.6.11-mm2 Paul Mundt
2005-03-08 19:40   ` 2.6.11-mm2 Andrew Morton
2005-03-08 16:00 ` 2.6.11-mm2 (compile stats) John Cherry
2005-03-08 18:54 ` 2.6.11-mm2 fremap.c compile error Jurriaan
2005-03-11 22:50   ` Adrian Bunk
2005-03-08 19:29 ` inconsistent kallsyms data [2.6.11-mm2] Dominik Brodowski
2005-03-08 20:35   ` Andrew Morton
2005-03-08 20:45     ` Dominik Brodowski
2005-03-09 12:57       ` Paulo Marques
2005-03-09 20:16         ` Paulo Marques
2005-03-10 12:12           ` Paulo Marques
2005-03-13  8:54             ` Sam Ravnborg
2005-03-14 13:33               ` Paulo Marques
2005-03-14 22:17               ` Dominik Brodowski
2005-03-14 22:14             ` Dominik Brodowski
2005-03-08 23:20 ` 2.6.11-mm2 Christoph Hellwig
2005-03-08 23:29   ` 2.6.11-mm2 Andrew Morton
2005-03-08 23:36 ` 2.6.11-mm2 J.A. Magallon
2005-03-08 23:44   ` Robert Love [this message]
2005-03-08 23:51     ` 2.6.11-mm2 J.A. Magallon
2005-03-09  0:02       ` 2.6.11-mm2 Robert Love
2005-03-09  0:16 ` 2.6.11-mm2 Adrian Bunk
2005-03-09  0:53   ` 2.6.11-mm2 Andrew Morton
2005-03-09  1:39     ` 2.6.11-mm2 Jeff Garzik
2005-03-09  0:20 ` 2.6.11-mm2 Adrian Bunk
2005-03-09  1:50   ` 2.6.11-mm2 Karsten Keil
2005-03-10  7:57 ` 2.6.11-mm2 Stefano Rivoir
2005-03-10  8:09   ` 2.6.11-mm2 Andrew Morton
2005-03-21 23:45   ` 2.6.11-mm2 Andrew Morton

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=1110325442.30255.8.camel@localhost \
    --to=rlove@rlove.org \
    --cc=jamagallon@able.es \
    --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