From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S266450AbUHBK0f (ORCPT ); Mon, 2 Aug 2004 06:26:35 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S266476AbUHBK0G (ORCPT ); Mon, 2 Aug 2004 06:26:06 -0400 Received: from mx1.redhat.com ([66.187.233.31]:33504 "EHLO mx1.redhat.com") by vger.kernel.org with ESMTP id S266487AbUHBKV4 (ORCPT ); Mon, 2 Aug 2004 06:21:56 -0400 Subject: Re: [brett@wrl.org: Stumped about where to post an Oops!] From: "Stephen C. Tweedie" To: Andrew Morton Cc: Marcelo Tosatti , Brett Charbeneau , Stephen Tweedie , linux-kernel In-Reply-To: <20040726133422.66c42143.akpm@osdl.org> References: <20040724235219.GB2614@dmt.cyclades> <20040726133422.66c42143.akpm@osdl.org> Content-Type: text/plain Content-Transfer-Encoding: 7bit Organization: Message-Id: <1091442091.2125.11.camel@sisko.scot.redhat.com> Mime-Version: 1.0 X-Mailer: Ximian Evolution 1.2.2 (1.2.2-5) Date: 02 Aug 2004 11:21:31 +0100 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Hi, On Mon, 2004-07-26 at 21:34, Andrew Morton wrote: > The 0x00040004 access address almost looks like some pointer-poisoning > thing. It's the result of doing > 0: 39 72 04 cmp %esi,0x4(%edx) <===== when %edx is 0x00040000. That looks to me like a single bit-flip in memory, and get_hash_table is one of those functions that shows up in traces all the time when you've got bad memory (it's one of the kernel's hottest functions for doing list-walks over dynamically-allocated data structures.) This footprint would definitely have me reaching for memtest86 as the next step. It certainly has all the hallmarks of a classic bit-flip bad memory trace. --Stephen