Return-Path: <linux-kernel-owner+w=401wt.eu-S932423AbXAVT6V@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
	id S932423AbXAVT6V (ORCPT <rfc822;w@1wt.eu>);
	Mon, 22 Jan 2007 14:58:21 -0500
Received: (majordomo@vger.kernel.org) by vger.kernel.org id S932433AbXAVT6U
	(ORCPT <rfc822;linux-kernel-outgoing>);
	Mon, 22 Jan 2007 14:58:20 -0500
Received: from smtp.osdl.org ([65.172.181.24]:45786 "EHLO smtp.osdl.org"
	rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
	id S932423AbXAVT6S (ORCPT <rfc822;linux-kernel@vger.kernel.org>);
	Mon, 22 Jan 2007 14:58:18 -0500
Date: Mon, 22 Jan 2007 11:57:03 -0800
From: Andrew Morton <akpm@osdl.org>
To: Justin Piszcz <jpiszcz@lucidpixels.com>
Cc: linux-kernel@vger.kernel.org, linux-raid@vger.kernel.org, xfs@oss.sgi.com
Subject: Re: 2.6.20-rc5: cp 18gb 18gb.2 = OOM killer, reproducible just like
 2.16.19.2
Message-Id: <20070122115703.97ed54f3.akpm@osdl.org>
In-Reply-To: <Pine.LNX.4.64.0701211424170.2552@p34.internal.lan>
References: <Pine.LNX.4.64.0701211424170.2552@p34.internal.lan>
X-Mailer: Sylpheed version 2.2.7 (GTK+ 2.8.17; x86_64-unknown-linux-gnu)
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
X-Mailing-List: linux-kernel@vger.kernel.org
Content-Length: 3591
Lines: 90

> On Sun, 21 Jan 2007 14:27:34 -0500 (EST) Justin Piszcz <jpiszcz@lucidpixels.com> wrote:
> Why does copying an 18GB on a 74GB raptor raid1 cause the kernel to invoke 
> the OOM killer and kill all of my processes?

What's that?   Software raid or hardware raid?  If the latter, which driver?

> Doing this on a single disk 2.6.19.2 is OK, no issues.  However, this 
> happens every time!
> 
> Anything to try?  Any other output needed?  Can someone shed some light on 
> this situation?
> 
> Thanks.
> 
> 
> The last lines of vmstat 1 (right before it kill -9'd my shell/ssh)
> 
> procs -----------memory---------- ---swap-- -----io---- -system-- 
> ----cpu----
>  r  b   swpd   free   buff  cache   si   so    bi    bo   in   cs us sy id 
> wa
>  0  7    764  50348     12 1269988    0    0 53632   172 1902 4600  1  8 
> 29 62
>  0  7    764  49420     12 1260004    0    0 53632 34368 1871 6357  2 11 
> 48 40

The wordwrapping is painful :(

> 
> The last lines of dmesg:
> [ 5947.199985] lowmem_reserve[]: 0 0 0
> [ 5947.199992] DMA: 0*4kB 1*8kB 1*16kB 0*32kB 1*64kB 1*128kB 1*256kB 
> 0*512kB 1*1024kB 1*2048kB 0*4096kB = 3544kB
> [ 5947.200010] Normal: 1*4kB 0*8kB 1*16kB 1*32kB 0*64kB 1*128kB 0*256kB 
> 1*512kB 0*1024kB 1*2048kB 0*4096kB = 2740kB
> [ 5947.200035] HighMem: 98*4kB 35*8kB 9*16kB 69*32kB 4*64kB 1*128kB 
> 1*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB = 3664kB
> [ 5947.200052] Swap cache: add 789, delete 189, find 16/17, race 0+0
> [ 5947.200055] Free swap  = 2197628kB
> [ 5947.200058] Total swap = 2200760kB
> [ 5947.200060] Free swap:       2197628kB
> [ 5947.205664] 517888 pages of RAM 
> [ 5947.205671] 288512 pages of HIGHMEM
> [ 5947.205673] 5666 reserved pages 
> [ 5947.205675] 257163 pages shared
> [ 5947.205678] 600 pages swap cached 
> [ 5947.205680] 88876 pages dirty
> [ 5947.205682] 115111 pages writeback
> [ 5947.205684] 5608 pages mapped
> [ 5947.205686] 49367 pages slab
> [ 5947.205688] 541 pages pagetables
> [ 5947.205795] Out of memory: kill process 1853 (named) score 9937 or a 
> child
> [ 5947.205801] Killed process 1853 (named)
> [ 5947.206616] bash invoked oom-killer: gfp_mask=0x84d0, order=0, 
> oomkilladj=0
> [ 5947.206621]  [<c013e33b>] out_of_memory+0x17b/0x1b0
> [ 5947.206631]  [<c013fcac>] __alloc_pages+0x29c/0x2f0
> [ 5947.206636]  [<c01479ad>] __pte_alloc+0x1d/0x90
> [ 5947.206643]  [<c0148bf7>] copy_page_range+0x357/0x380
> [ 5947.206649]  [<c0119d75>] copy_process+0x765/0xfc0
> [ 5947.206655]  [<c012c3f9>] alloc_pid+0x1b9/0x280
> [ 5947.206662]  [<c011a839>] do_fork+0x79/0x1e0
> [ 5947.206674]  [<c015f91f>] do_pipe+0x5f/0xc0
> [ 5947.206680]  [<c0101176>] sys_clone+0x36/0x40
> [ 5947.206686]  [<c0103138>] syscall_call+0x7/0xb
> [ 5947.206691]  [<c0420033>] __sched_text_start+0x853/0x950
> [ 5947.206698]  ======================= 

Important information from the oom-killing event is missing.  Please send
it all.

>From your earlier reports we have several hundred MB of ZONE_NORMAL memory
which has gone awol.

Please include /proc/meminfo from after the oom-killing.

Please work out what is using all that slab memory, via /proc/slabinfo.

After the oom-killing, please see if you can free up the ZONE_NORMAL memory
via a few `echo 3 > /proc/sys/vm/drop_caches' commands.  See if you can
work out what happened to the missing couple-of-hundred MB from
ZONE_NORMAL.


-
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/