mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Ingo Oeser <ingo.oeser@informatik.tu-chemnitz.de>
To: Mika Liljeberg <Mika.Liljeberg@welho.com>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>,
	Benjamin LaHaise <bcrl@redhat.com>,
	Linus Torvalds <torvalds@transmeta.com>,
	linux-kernel@vger.kernel.org
Subject: Re: Context switch times
Date: Sun, 7 Oct 2001 15:03:58 +0200	[thread overview]
Message-ID: <20011007150358.G30515@nightmaster.csn.tu-chemnitz.de> (raw)
In-Reply-To: <E15pWfR-0006g5-00@the-village.bc.nu> <3BC02709.A8E6F999@welho.com>
In-Reply-To: <3BC02709.A8E6F999@welho.com>; from Mika.Liljeberg@welho.com on Sun, Oct 07, 2001 at 12:57:29PM +0300

On Sun, Oct 07, 2001 at 12:57:29PM +0300, Mika Liljeberg wrote:
> For problems 1 and 2 I propose the following solution: Insert the the
> load balancing routine itself as a (fake) task on each CPU and run it
> when the CPU gets around to it. The load balancer should behave almost
> like a CPU-bound task, scheduled on the lowest priority level with other
> runnable tasks. 

The idle-task might be (ab-)used for this, because it has perfect
data for this.

T_SystemElapsed - T_IdleTaskRun = T_CPULoaded

Balancing could be done in schedule() itself, after checking this
value for each CPU.

> The last bit is important: the load balancer should not
> be allowed to starve but should be invoked approximately once every
> "full rotation" of the scheduler.
 
If a artificial CPU-hog is used for this task, the idle task will
never be run and power savings in the CPU are impossible.

Other parts sound interesting.

Regards

Ingo Oeser
-- 
.... Our continuing mission: To seek out knowledge of C, to explore
strange UNIX commands, and to boldly code where no one has man page 4.
                        --- Mike A. Harris <mharris@meteng.on.ca>

  reply	other threads:[~2001-10-07 13:04 UTC|newest]

Thread overview: 62+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2001-10-04 21:04 Mike Kravetz
2001-10-04 21:14 ` arjan
2001-10-04 21:39   ` Richard Gooch
2001-10-09  4:55     ` Richard Gooch
2001-10-09 13:49       ` bill davidsen
2001-10-04 21:25 ` David S. Miller
2001-10-04 21:52   ` David S. Miller
2001-10-04 21:55     ` Benjamin LaHaise
2001-10-04 22:35       ` Davide Libenzi
2001-10-04 22:49         ` Mike Kravetz
2001-10-04 22:42       ` Linus Torvalds
2001-10-04 22:53         ` Benjamin LaHaise
2001-10-05 15:13           ` Alan Cox
2001-10-05 17:49             ` george anzinger
2001-10-05 22:29               ` Alan Cox
2001-10-05 22:56                 ` Davide Libenzi
2001-10-05 23:04                   ` Alan Cox
2001-10-05 23:16                     ` Davide Libenzi
2001-10-05 23:17                       ` Alan Cox
2001-10-05 23:21                         ` Davide Libenzi
2001-10-05 23:43                     ` Roger Larsson
2001-10-07  1:20                 ` george anzinger
2001-10-07  1:33                   ` Bill Davidsen
2001-10-07  9:56                   ` Alan Cox
2001-10-06  2:24             ` Albert D. Cahalan
2001-10-06  2:57               ` Davide Libenzi
2001-10-07  9:57             ` Mika Liljeberg
2001-10-07 13:03               ` Ingo Oeser [this message]
2001-10-07 13:48                 ` Mika Liljeberg
2001-10-07 14:24                   ` Ingo Oeser
2001-10-07 14:33                     ` Mika Liljeberg
2001-10-07 18:00                       ` george anzinger
2001-10-07 22:06                         ` Mika Liljeberg
2001-10-07 22:31                           ` Davide Libenzi
2001-10-07 22:33                             ` Mika Liljeberg
2001-10-07 23:49                           ` george anzinger
2001-10-08 21:07                             ` Mika Liljeberg
2001-10-08 22:54                               ` discontig physical memory Petko Manolov
2001-10-08 23:05                               ` David S. Miller
2001-10-08 23:18                                 ` Petko Manolov
2001-10-08 23:29                                 ` David S. Miller
2001-10-09  0:34                                   ` Petko Manolov
2001-10-09  0:36                                   ` Petko Manolov
2001-10-09  1:37                                   ` David S. Miller
2001-10-09  2:43                                     ` Petko Manolov
2001-10-08 15:19                       ` Context switch times bill davidsen
2001-10-10  6:07                         ` Mike Fedyk
2001-10-07 18:39               ` Davide Libenzi
2001-10-09 20:37               ` Hubertus Franke
2001-10-09 23:50                 ` george anzinger
2001-10-11 10:52                   ` Hubertus Franke
2001-10-04 23:41         ` Mike Kravetz
2001-10-04 23:50           ` Linus Torvalds
2001-10-05 15:15             ` Eric W. Biederman
2001-10-04 23:56           ` Davide Libenzi
2001-10-05  0:45           ` Andrea Arcangeli
2001-10-05  4:35             ` Mike Kravetz
2001-10-07 17:59               ` Andrea Arcangeli
2001-10-07 19:54                 ` george anzinger
2001-10-07 20:24                   ` Andrea Arcangeli
2001-10-09  5:00     ` David S. Miller
2001-10-05  6:31 Michailidis, Dimitrios

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=20011007150358.G30515@nightmaster.csn.tu-chemnitz.de \
    --to=ingo.oeser@informatik.tu-chemnitz.de \
    --cc=Mika.Liljeberg@welho.com \
    --cc=alan@lxorguk.ukuu.org.uk \
    --cc=bcrl@redhat.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=torvalds@transmeta.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

all inboxes | Powered by JetHome®