From owner-freebsd-current@FreeBSD.ORG Fri Oct 15 10:48:59 2004 Return-Path: Delivered-To: freebsd-current@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 91A0616A4CE for ; Fri, 15 Oct 2004 10:48:59 +0000 (GMT) Received: from mail07.syd.optusnet.com.au (mail07.syd.optusnet.com.au [211.29.132.188]) by mx1.FreeBSD.org (Postfix) with ESMTP id F0DB243D1F for ; Fri, 15 Oct 2004 10:48:58 +0000 (GMT) (envelope-from PeterJeremy@optushome.com.au) Received: from cirb503493.alcatel.com.au (c211-30-75-229.belrs2.nsw.optusnet.com.au [211.30.75.229]) i9FAmu01009963 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO); Fri, 15 Oct 2004 20:48:57 +1000 Received: from cirb503493.alcatel.com.au (localhost.alcatel.com.au [127.0.0.1])i9FAmuxP046121; Fri, 15 Oct 2004 20:48:56 +1000 (EST) (envelope-from pjeremy@cirb503493.alcatel.com.au) Received: (from pjeremy@localhost)i9FAmuEV046120; Fri, 15 Oct 2004 20:48:56 +1000 (EST) (envelope-from pjeremy) Date: Fri, 15 Oct 2004 20:48:56 +1000 From: Peter Jeremy To: Guido van Rooij Message-ID: <20041015104856.GB45863@cirb503493.alcatel.com.au> References: <20041013214911.GD986@green.homeunix.org> <20041015073118.GA19660@gvr.gvr.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20041015073118.GA19660@gvr.gvr.org> User-Agent: Mutt/1.4.2i cc: freebsd-current@freebsd.org Subject: Re: 5.3-BETA7 install cd: kernel trap 12 with interrupts disabled X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 15 Oct 2004 10:48:59 -0000 On Fri, 2004-Oct-15 09:31:18 +0200, Guido van Rooij wrote: >I am beginning to wonder if we should have a boot option that enables >a thorough memtest from within the kernel...(e.g. boot -m). What do you define as "thorough"? A thorough memory test requires intimate knowledge of the physical memory cell layout (to ensure that pattern tests make sense) as well as the ability to control temperature, the supply voltage, thresholds and timings (to detect marginal conditions). "make buildworld" is probably the best memory test we're likely to find. The speed is on a par with Memtest86 as well. Add/vary '-k' for additional coverage. If you really want a memory test that can run at boot-time, the best approach would seem to be to build a version of sysutils/memtest or sysutils/memtest86 that can be loaded by the bootloader. FWIW, I've have an Alpha/AXP CPU that (apparently) had a pattern- sensitive bug in its cache. It passed all SRM memory tests but consistently died during rc processing. -- Peter Jeremy