From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753289AbaEFR6R (ORCPT ); Tue, 6 May 2014 13:58:17 -0400 Received: from vserver.eikelenboom.it ([84.200.39.61]:32916 "EHLO smtp.eikelenboom.it" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750892AbaEFR6Q (ORCPT ); Tue, 6 May 2014 13:58:16 -0400 Date: Tue, 6 May 2014 19:58:11 +0200 From: Sander Eikelenboom Organization: Eikelenboom IT services X-Priority: 3 (Normal) Message-ID: <1029422291.20140506195811@eikelenboom.it> To: David Vrabel CC: Pekka Enberg , Andrew Morton , linux-kernel , , Konrad Rzeszutek Wilk Subject: Re: 3.15-rc3: boot of a xen PV guest fails when using a guest kernel with CONFIG_DEBUG_SLAB=y In-Reply-To: <5368B088.8040206@citrix.com> References: <169388657.20140502230132@eikelenboom.it> <5368B088.8040206@citrix.com> 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 Tuesday, May 6, 2014, 11:51:04 AM, you wrote: > On 02/05/14 22:01, Sander Eikelenboom wrote: >> Hi, >> >> On a machine running Xen-Unstable, i just tried running a 3.15-rc3 kernel with >> CONFIG_DEBUG_SLAB=y >> >> [ 0.748297] Kernel panic - not syncing: Out of memory and no killable processes... >> [ 0.748297] >> [ 0.748305] CPU: 0 PID: 1 Comm: swapper/0 Not tainted 3.15.0-rc3-20140502-nbdebug12-slabdb+ #1 >> [ 0.748310] 00000000002040d0 ffff88000ee9fa48 ffffffff81b8da66 ffff88000ee9c010 >> [ 0.748317] ffffffff81f5a8a0 ffff88000ee9fac8 ffffffff81b89bfb 0000000000000000 >> [ 0.748325] ffffffff00000008 ffff88000ee9fad8 ffff88000ee9fa78 ffffffff811952fa >> [ 0.748332] Call Trace: >> [ 0.748336] [] dump_stack+0x46/0x58 >> [ 0.748340] [] panic+0xd2/0x207 >> [ 0.748345] [] ? out_of_memory+0x39a/0x5a0 >> [ 0.748350] [] out_of_memory+0x59e/0x5a0 > Your guest is out of memory. Give it some more? The guest has 512MB assigned to it and it worked fine with: - SLAB without DEBUG_SLAB - SLUB with and without DEBUG_SLUB It ooms with: - SLAB with DEBUG_SLAB With 1024MB assigned it seems to boot fine in all cases, so it seems DEBUG_SLAB is quite memory consuming when it fails in the first second of boot on 512MB. -- Sander > David