Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751256Ab0FUE0U (ORCPT ); Mon, 21 Jun 2010 00:26:20 -0400 Received: from fgwmail6.fujitsu.co.jp ([192.51.44.36]:35364 "EHLO fgwmail6.fujitsu.co.jp" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750853Ab0FUE0S (ORCPT ); Mon, 21 Jun 2010 00:26:18 -0400 X-SecurityPolicyCheck-FJ: OK by FujitsuOutboundMailChecker v1.3.1 Date: Mon, 21 Jun 2010 13:21:36 +0900 From: KAMEZAWA Hiroyuki To: Andrew Hendry Cc: Richard Yao , linux-kernel@vger.kernel.org Subject: Re: Kernel does strange things when compilations push memory usage above physical memory and the compilations are being done in a tmpfs, despite having ample swap Message-Id: <20100621132136.b8734516.kamezawa.hiroyu@jp.fujitsu.com> In-Reply-To: References: <20100621120256.8d2a2fca.kamezawa.hiroyu@jp.fujitsu.com> Organization: FUJITSU Co. LTD. X-Mailer: Sylpheed 3.0.2 (GTK+ 2.10.14; i686-pc-mingw32) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 2044 Lines: 74 On Mon, 21 Jun 2010 13:53:02 +1000 Andrew Hendry wrote: > Kame, > > Could the tempfs revert be the same and fix the ramdisk issue I have seen? > http://marc.info/?l=linux-kernel&m=127569877714937&w=2 > I can re-test this evening. > I think no. (and IIUC, the patch is included in 2.6.35-rc1) Anyway, ramdisk isn't swappable, is it ? And, this doesn't happen in 2.6.33 or 2.6.34 ? >From your log. == Jun 5 04:51:58 jaunty kernel: [25913.310588] firefox-bin invoked oom-killer: gfp_mask=0xd0, order=1, oom_adj=0 == Then, order=1 (2-pages contiguous page allocation) failure. == Jun 5 05:12:29 jaunty kernel: [27142.500150] Node 0 DMA: 1*4kB 0*8kB 0*16kB 2*32kB 2*64kB 0*128kB 1*256kB 0*512kB 1*1024kB 1*2048kB 3*4096kB = 15812kB Jun 5 05:12:29 jaunty kernel: [27142.500157] Node 0 DMA32: 49660*4kB 14*8kB 4*16kB 6*32kB 1*64kB 2*128kB 0*256kB 1*512kB 1*1024kB 0*2048kB 0*4096kB = 200864kB Jun 5 05:12:29 jaunty kernel: [27142.500165] Node 0 Normal: 63050*4kB 5*8kB 5*16kB 9*32kB 1*64kB 2*128kB 1*256kB 1*512kB 0*1024kB 1*2048kB 0*4096kB = 255744kB == It seems fragmented. Hmm.. == MemTotal: 8187304 kB MemFree: 231976 kB Active: 14056 kB Inactive: 72592 kB Active(anon): 12900 kB Inactive(anon): 53940 kB Active(file): 1156 kB Inactive(file): 18652 kB Unevictable: 0 kB Mlocked: 0 kB HugePages_Total: 0 HugePages_Free: 0 HugePages_Rsvd: 0 HugePages_Surp: 0 AND 5GBytes for ramdisk. (Right ?) == There are too small memory used for user program or file caches. In this kind of case, I tend to doubt memory leak (in kernel) But I feel ramdisk is very large, too... How about updating to -rc3 or turning on kmemcheck and see what happens ? Thanks, -Kame -- 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/