Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754475AbZGEKjI (ORCPT ); Sun, 5 Jul 2009 06:39:08 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1752644AbZGEKiz (ORCPT ); Sun, 5 Jul 2009 06:38:55 -0400 Received: from fgwmail5.fujitsu.co.jp ([192.51.44.35]:54836 "EHLO fgwmail5.fujitsu.co.jp" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751065AbZGEKiz (ORCPT ); Sun, 5 Jul 2009 06:38:55 -0400 X-SecurityPolicyCheck-FJ: OK by FujitsuOutboundMailChecker v1.3.1 From: KOSAKI Motohiro To: Wu Fengguang Subject: Re: Found the commit that causes the OOMs Cc: kosaki.motohiro@jp.fujitsu.com, Rik van Riel , David Woodhouse , David Howells , Minchan Kim , Mel Gorman , Johannes Weiner , Andrew Morton , LKML , Christoph Lameter , "peterz@infradead.org" , "tytso@mit.edu" , "linux-mm@kvack.org" , "elladan@eskimo.com" , "npiggin@suse.de" , "Barnes, Jesse" In-Reply-To: <20090705095520.GA31587@localhost> References: <4A4AD07E.2040508@redhat.com> <20090705095520.GA31587@localhost> Message-Id: <20090705193551.090E.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: Sun, 5 Jul 2009 19:38:54 +0900 (JST) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1304 Lines: 38 > >> OK. thanks. > >> I plan to submit this patch after small more tests. it is useful for OOM analysis. > > > > It is also useful for throttling page reclaim. > > > > If more than half of the inactive pages in a zone are > > isolated, we are probably beyond the point where adding > > additional reclaim processes will do more harm than good. > > Maybe we can try limiting the isolation phase of direct reclaims to > one per CPU? > > mutex_lock(per_cpu_lock); > isolate_pages(); > shrink_page_list(); > put_back_pages(); > mutex_unlock(per_cpu_lock); > > This way the isolated pages as well as major parts of direct reclaims > will be bounded by CPU numbers. The added overheads should be trivial > comparing to the reclaim costs. hm, this idea makes performance degression on few CPU machine, I think. e.g. if system have only one cpu and sysmtem makes lumpy reclaim, lumpy reclaim makes synchronous pageout and it makes very long waiting time. I suspect per-cpu decision is not useful in this area. thanks. -- 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/