From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S265341AbTGCT7m (ORCPT ); Thu, 3 Jul 2003 15:59:42 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S265346AbTGCT7m (ORCPT ); Thu, 3 Jul 2003 15:59:42 -0400 Received: from dbl.q-ag.de ([80.146.160.66]:737 "EHLO dbl.q-ag.de") by vger.kernel.org with ESMTP id S265341AbTGCT7h (ORCPT ); Thu, 3 Jul 2003 15:59:37 -0400 Message-ID: <3F048E77.8080402@colorfullife.com> Date: Thu, 03 Jul 2003 22:13:43 +0200 From: Manfred Spraul User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.3) Gecko/20030313 X-Accept-Language: en-us, en MIME-Version: 1.0 To: Andy Pfiffer CC: linux-kernel@vger.kernel.org, yoshfuji@linux-ipv6.org, Andrew Morton Subject: Re: Linux 2.5.74: BUG at mm/slab.c:1537 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org The problem is caused by changeset 1.1310.102.3, 2003/07/01 02:01:51+10:00, yoshfuji@linux-ipv6.org: http://linus.bkbits.net:8080/linux-2.5/diffs/net/ipv4/raw.c@3f005eebc5YsuvTFXhDo-QDhPEgh5Q?nav=index.html (Subject: [NET] fixed /proc/net/raw{,6} seq_file support) raw_iter_state is just an integer. Without that patch, the integer is stored directly in the seq->private pointer (note the & in the define of raw_seq_private, around line 690 of net/ipv4/raw.c). The patch converts part of the code to an pointer to an integer, but other parts still consider seq->private as an integer. The oops is actually a BUG statement in kmalloc: it complains (if CONFIG_DEBUG_SLAB is enabled) about the invalid pointer. -- Manfred