From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754738AbZGWXVP (ORCPT ); Thu, 23 Jul 2009 19:21:15 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1753400AbZGWXVO (ORCPT ); Thu, 23 Jul 2009 19:21:14 -0400 Received: from fgwmail7.fujitsu.co.jp ([192.51.44.37]:55338 "EHLO fgwmail7.fujitsu.co.jp" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753940AbZGWXVN (ORCPT ); Thu, 23 Jul 2009 19:21:13 -0400 X-SecurityPolicyCheck-FJ: OK by FujitsuOutboundMailChecker v1.3.1 From: KOSAKI Motohiro To: Paul Menage Subject: Re: [PATCH] copy over oom_adj value at fork time Cc: kosaki.motohiro@jp.fujitsu.com, David Rientjes , Rik van Riel , linux-kernel@vger.kernel.org, akpm@linux-foundation.org, mel@csn.ul.ie, npiggin@suse.de In-Reply-To: <6599ad830907211303m25141938kcb75065a74a40ee2@mail.gmail.com> References: <6599ad830907211303m25141938kcb75065a74a40ee2@mail.gmail.com> Message-Id: <20090722233347.776E.A69D9226@jp.fujitsu.com> MIME-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit X-Mailer: Becky! ver. 2.50.07 [ja] Date: Fri, 24 Jul 2009 08:21:11 +0900 (JST) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org > On Tue, Jul 21, 2009 at 1:00 PM, David Rientjes wrote: > > On Tue, 21 Jul 2009, Paul Menage wrote: > > > >> Agreed, but the same livelock can be fixed in ways that don't break > >> the API. (E.g. check for the victim being OOM_DISABLED in > >> select_bad_process() when we find a new "worst" candidate). > >> > > > > And allow /proc/pid-of-child/oom_score to represent a possible candidate > > (and, additionally, a hint at the oom killing priority) when it really > > isn't? Oops, I really don't like /proc/pid-of-child/oom_score. it is very strange. > > It's the API that's existed for years with no complaints, AFAICS. I think thead and vfork() should be separeted on this discussion. I agree vfork() regression should be fixed. but I don't think anyone hope per-thread oom score. Of cource, if simple reverting is best way, I don't oppose this.... ;-)