From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ingo Molnar Subject: Re: 2.6.25 crash: EIP: [] xfrm_output_resume+0x64/0x100 ss:esp 0068:c03a1e5c Date: Fri, 9 May 2008 13:11:30 +0200 Message-ID: <20080509111130.GA7968@elte.hu> References: <20080509102509.GA11710@gondor.apana.org.au> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Herbert Xu , linux-kernel@vger.kernel.org, netdev@vger.kernel.org To: Marco Berizzi Return-path: Received: from mx2.mail.elte.hu ([157.181.151.9]:53165 "EHLO mx2.mail.elte.hu" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1761259AbYEILLq (ORCPT ); Fri, 9 May 2008 07:11:46 -0400 Content-Disposition: inline In-Reply-To: Sender: netdev-owner@vger.kernel.org List-ID: * Marco Berizzi wrote: > Herbert Xu wrote: > > > On Fri, May 09, 2008 at 11:50:00AM +0200, Marco Berizzi wrote: > > > > > > I have rebooted the two boxes with slub_debug. > > > This is the output taken with network console. > > > Herbert does it help you? > > > > > > BUG: unable to handle kernel paging request at 6b6b6b9f > > > > Unfortunately this just confirms that your skb has been freed > > prematurely because 6b is the poison value. > > > > However, it doesn't point us at the offender. > > :-(( you could try x86.git and enable CONFIG_KMEMCHECK, which catches all sorts of memory corruption bugs at its root: http://people.redhat.com/mingo/x86.git/README it's for 32-bit currently, and depends on the following CONFIG details in your .config: depends on X86_32 depends on !X86_USE_3DNOW depends on !CC_OPTIMIZE_FOR_SIZE depends on !DEBUG_PAGEALLOC && SLUB note that CONFIG_DEBUG_PAGEALLOC=y will catch a few types of corruption too. Note that KMEMCHECK and DEBUG_PAGEALLOC are exclusive. Ingo