mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Linus Torvalds <torvalds@linux-foundation.org>
To: Jeff Garzik <jeff@garzik.org>
Cc: Robin Holt <holt@sgi.com>,
	"Eric W. Biederman" <ebiederm@xmission.com>,
	Ingo Molnar <mingo@elte.hu>,
	linux-kernel@vger.kernel.org,
	Jack Steiner <steiner@americas.sgi.com>
Subject: Re: init's children list is long and slows reaping children.
Date: Fri, 6 Apr 2007 15:51:40 -0700 (PDT)	[thread overview]
Message-ID: <Pine.LNX.4.64.0704061546590.6730@woody.linux-foundation.org> (raw)
In-Reply-To: <4616CBF0.7090606@garzik.org>



On Fri, 6 Apr 2007, Jeff Garzik wrote:
> 
> I would rather change the implementation under the hood to start per-CPU
> threads on demand, similar to a thread-pool implementation.
> 
> Boxes with $BigNum CPUs probably won't ever use half of those threads.

The counter-argument is that boxes with $BigNum CPU's really don't hurt 
from it either, and having per-process data structures is often simpler 
and more efficient than trying to have some thread pool.

IOW, once we get the processes off the global list, there just isn't any 
downside from them. Sure, they use some memory, but people who buy 
1024-cpu machines won't care about a few kB per CPU..

So the *only* downside is literally the process list, and one suggested 
patch already just removes kernel threads entirely from the parenthood 
lists.

The other potential downside could be "ps is slow", but on the other hand, 
having the things stick around and have things like CPU-time accumulate is 
probably worth it - if there are some issues, they'd show up properly 
accounted for in a way that process pools would have a hard time doing.

So I really don't think this is worth changing things over, apart from 
literally removing them from process lists, which I think everybody agrees 
we should just do - it just never even came up before!

			Linus

  reply	other threads:[~2007-04-06 22:51 UTC|newest]

Thread overview: 112+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2007-04-05 19:51 Robin Holt
2007-04-05 20:57 ` Linus Torvalds
2007-04-06  0:51   ` Chris Snook
2007-04-06  1:03     ` Chris Snook
2007-04-06  1:29     ` Linus Torvalds
2007-04-06  2:15       ` Eric W. Biederman
2007-04-06 10:43         ` Robin Holt
2007-04-06 15:38           ` Eric W. Biederman
2007-04-06 16:31             ` Oleg Nesterov
2007-04-06 17:32               ` Ingo Molnar
2007-04-06 17:39                 ` Roland Dreier
2007-04-06 18:04                   ` Eric W. Biederman
2007-04-06 18:30                 ` Eric W. Biederman
2007-04-06 19:18                   ` [patch] sched: get rid of p->children use in show_task() Ingo Molnar
2007-04-06 19:22                     ` Ingo Molnar
2007-04-10 13:48                   ` init's children list is long and slows reaping children Ingo Molnar
2007-04-10 13:38                     ` Oleg Nesterov
2007-04-10 15:00                       ` Eric W. Biederman
2007-04-10 15:16                       ` [PATCH] Only send pdeath_signal when getppid changes Eric W. Biederman
2007-04-10 16:37                         ` Oleg Nesterov
2007-04-10 17:41                           ` Eric W. Biederman
2007-04-10 17:48                             ` Roland McGrath
2007-04-11  3:17                               ` Albert Cahalan
2007-04-10 14:51                     ` init's children list is long and slows reaping children Eric W. Biederman
2007-04-10 15:06                       ` Ingo Molnar
2007-04-10 15:22                         ` Eric W. Biederman
2007-04-10 15:53                           ` Ingo Molnar
2007-04-10 16:17                             ` Eric W. Biederman
2007-04-11  6:20                             ` [patch] uninline remove/add_parent() APIs Ingo Molnar
2007-04-11  7:00                               ` Eric W. Biederman
2007-04-11 22:06                                 ` Andrew Morton
2007-04-12 10:45                                   ` Eric W. Biederman
2007-04-12 22:50                                   ` Roland McGrath
2007-04-10 16:44                       ` init's children list is long and slows reaping children Oleg Nesterov
2007-04-11 19:55                         ` Bill Davidsen
2007-04-11 20:17                           ` Eric W. Biederman
2007-04-11 21:24                             ` Bill Davidsen
2007-04-11 20:19                           ` Oleg Nesterov
2007-04-06 18:02               ` Eric W. Biederman
2007-04-06 18:21               ` Davide Libenzi
2007-04-06 18:56                 ` Eric W. Biederman
2007-04-06 19:16                   ` Davide Libenzi
2007-04-06 19:19                     ` Ingo Molnar
2007-04-06 21:29                       ` Davide Libenzi
2007-04-06 21:51                         ` Linus Torvalds
2007-04-06 22:31                           ` Davide Libenzi
2007-04-06 22:46                             ` Linus Torvalds
2007-04-06 22:59                               ` Davide Libenzi
2007-04-09  8:28                           ` Ingo Molnar
2007-04-09 18:09                             ` Bill Davidsen
2007-04-09 19:28                               ` Kyle Moffett
2007-04-09 19:51                                 ` Linus Torvalds
2007-04-09 20:03                                   ` Davide Libenzi
2007-04-10 15:12                                     ` Bill Davidsen
2007-04-10 19:17                                       ` Davide Libenzi
2007-04-09 20:00                                 ` Eric W. Biederman
2007-04-06 16:41             ` Robin Holt
2007-04-09 17:37         ` Chris Snook
2007-04-06 18:05       ` Christoph Hellwig
2007-04-06 19:39         ` Eric W. Biederman
2007-04-06 22:38 ` Jeff Garzik
2007-04-06 22:51   ` Linus Torvalds [this message]
2007-04-06 23:37     ` Jeff Garzik
2007-04-11  7:28       ` Nick Piggin
2007-04-10  0:23   ` Andrew Morton
2007-04-10  0:48     ` Eric W. Biederman
2007-04-10  1:15       ` Andrew Morton
2007-04-10  6:53       ` Jeff Garzik
2007-04-10  9:42       ` Robin Holt
2007-04-10  1:59     ` Dave Jones
2007-04-10  2:30       ` Andrew Morton
2007-04-10  2:46         ` Linus Torvalds
2007-04-10  7:07           ` Jeff Garzik
2007-04-10 22:20             ` Ingo Oeser
2007-04-10  5:07         ` Alexey Dobriyan
2007-04-10  5:21           ` Dave Jones
2007-04-10  6:09         ` Torsten Kaiser
2007-04-10  7:08           ` Jeff Garzik
2007-04-10  7:05         ` Jeff Garzik
2007-04-10  7:37           ` Andrew Morton
2007-04-10  8:33             ` Jeff Garzik
2007-04-10  8:41               ` Andrew Morton
2007-04-10  8:48                 ` Jeff Garzik
2007-04-10 22:35                   ` Ingo Oeser
2007-04-10 16:35           ` Matt Mackall
2007-04-10  7:44         ` Russell King
2007-04-10  8:16           ` Jeff Garzik
2007-04-10  8:59           ` Ingo Molnar
2007-04-10  9:33             ` Jeff Garzik
2007-04-06  8:42 Oleg Nesterov
2007-04-06  9:10 ` Eric W. Biederman
2007-04-06  9:44   ` Oleg Nesterov
2007-04-06 15:45     ` Eric W. Biederman
2007-04-06 15:47       ` Oleg Nesterov
2007-04-06 17:16         ` Linus Torvalds
2007-04-06 17:27           ` Ingo Molnar
2007-04-06 17:31             ` Ingo Molnar
2007-04-06 17:34           ` Eric W. Biederman
2007-04-06 19:06             ` H. Peter Anvin
2007-04-06 19:15               ` Eric W. Biederman
2007-04-06 19:21                 ` H. Peter Anvin
2007-04-06 21:04                 ` Jeremy Fitzhardinge
2007-04-06 21:07                   ` H. Peter Anvin
2007-04-06 19:36           ` Oleg Nesterov
2007-04-06 19:43             ` Ingo Molnar
2007-04-06 20:01               ` Oleg Nesterov
2007-04-06 20:21                 ` Ingo Molnar
2007-04-06 19:47             ` Oleg Nesterov
2007-04-06 19:59               ` Eric W. Biederman
2007-04-07 20:31             ` Oleg Nesterov
2007-04-08  0:38               ` Eric W. Biederman
2007-04-08 15:46                 ` Oleg Nesterov

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.4.64.0704061546590.6730@woody.linux-foundation.org \
    --to=torvalds@linux-foundation.org \
    --cc=ebiederm@xmission.com \
    --cc=holt@sgi.com \
    --cc=jeff@garzik.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@elte.hu \
    --cc=steiner@americas.sgi.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®