public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: linux@horizon.com
To: clameter@engr.sgi.com
Cc: linux-kernel@vger.kernel.org
Subject: Re: Terminate process that fails on a constrained allocation
Date: 8 Feb 2006 15:14:36 -0500	[thread overview]
Message-ID: <20060208201436.14693.qmail@science.horizon.com> (raw)

Would perhaps a less drastic solution, that at least supports the common
partitioned-system configuration, be to limit the oom killer to processes
whose nodes are a *subset* of ours?

That way, a limited process won't kill any unlimited processes, but it
can fight with other processes with the same limits.  However, an
unlimited process discovering oom can kill anything on the system if
necessary.

(This requires a modified version of cpuset_excl_nodes_overlap that
calls nodes_subset() instead of nodes_intersects(), but it's pretty
straightforward.)

             reply	other threads:[~2006-02-08 20:14 UTC|newest]

Thread overview: 28+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2006-02-08 20:14 linux [this message]
  -- strict thread matches above, loose matches on Subject: below --
2006-02-08 18:05 Terminate process that fails on a constrained allocation Christoph Lameter
2006-02-08 18:13 ` Andi Kleen
2006-02-08 18:34   ` Paul Jackson
2006-02-08 18:54     ` Christoph Lameter
2006-02-08 19:01       ` Andi Kleen
2006-02-08 19:15         ` Christoph Lameter
2006-02-08 18:33 ` Paul Jackson
2006-02-08 18:42   ` Christoph Lameter
2006-02-08 18:57     ` Paul Jackson
2006-02-08 19:02       ` Christoph Lameter
2006-02-08 19:05       ` Andi Kleen
2006-02-08 20:22         ` Paul Jackson
2006-02-08 20:36           ` Christoph Lameter
2006-02-08 20:55             ` Paul Jackson
2006-02-08 21:01               ` Andi Kleen
2006-02-08 21:03                 ` Paul Jackson
2006-02-08 21:21                   ` Andi Kleen
2006-02-08 21:39                     ` Andrew Morton
2006-02-08 22:11                       ` Christoph Lameter
2006-02-08 22:41                         ` Andi Kleen
2006-02-08 23:29                           ` Christoph Lameter
2006-02-08 23:35                           ` Andrew Morton
2006-02-08 22:48                         ` Christoph Lameter
2006-02-08 23:28                       ` Christoph Lameter
2006-02-08 23:43                         ` Andrew Morton
2006-02-08 23:54                           ` Christoph Lameter
2006-02-08 23:57                             ` Andi Kleen

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=20060208201436.14693.qmail@science.horizon.com \
    --to=linux@horizon.com \
    --cc=clameter@engr.sgi.com \
    --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