Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1750890AbZLOFDQ (ORCPT ); Tue, 15 Dec 2009 00:03:16 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1750771AbZLOFDN (ORCPT ); Tue, 15 Dec 2009 00:03:13 -0500 Received: from smtp-out.google.com ([216.239.33.17]:25726 "EHLO smtp-out.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750760AbZLOFDM (ORCPT ); Tue, 15 Dec 2009 00:03:12 -0500 DomainKey-Signature: a=rsa-sha1; s=beta; d=google.com; c=nofws; q=dns; h=date:from:x-x-sender:to:cc:subject:in-reply-to:message-id: references:user-agent:mime-version:content-type:x-system-of-record; b=xmPghrEkS7ZYuVHAisLCYE/aQNYawqDODtmqWQewuPhh/NSbAEFwomjrmGVGTOj35 R7ojtR+jIi1XXnXqTL/HA== Date: Mon, 14 Dec 2009 21:03:07 -0800 (PST) From: David Rientjes X-X-Sender: rientjes@chino.kir.corp.google.com To: KOSAKI Motohiro cc: KAMEZAWA Hiroyuki , Andrew Morton , Daisuke Nishimura , linux-mm@kvack.org, linux-kernel@vger.kernel.org, Christoph Lameter Subject: Re: [BUGFIX][PATCH] oom-kill: fix NUMA consraint check with nodemask v4.2 In-Reply-To: <20091215134319.CDD3.A69D9226@jp.fujitsu.com> Message-ID: References: <20091215103202.eacfd64e.kamezawa.hiroyu@jp.fujitsu.com> <20091215134319.CDD3.A69D9226@jp.fujitsu.com> User-Agent: Alpine 2.00 (DEB 1167 2008-08-23) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-System-Of-Record: true Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1028 Lines: 22 On Tue, 15 Dec 2009, KOSAKI Motohiro wrote: > To compare vsz is only meaningful when the same program are compared. > But oom killer don't. To compare vsz between another program DONT detect > any memory leak. > You're losing the ability to detect that memory leak because you'd be using a baseline that userspace cannot possibly know at the time of oom. You cannot possibly insist that users understand the amount of resident memory for all applications when tuning the heuristic adjuster from userspace. In other words, how do you plan on userspace being able to identify tasks that are memory leakers if you change the baseline to rss? Unless you have an answer to this question, you're not admitting the problem that the oom killer is primarily designed to address. -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/