From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932600AbWB1Xe3 (ORCPT ); Tue, 28 Feb 2006 18:34:29 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S932609AbWB1Xe3 (ORCPT ); Tue, 28 Feb 2006 18:34:29 -0500 Received: from smtp.osdl.org ([65.172.181.4]:39083 "EHLO smtp.osdl.org") by vger.kernel.org with ESMTP id S932600AbWB1Xe2 (ORCPT ); Tue, 28 Feb 2006 18:34:28 -0500 Date: Tue, 28 Feb 2006 15:32:56 -0800 From: Andrew Morton To: Martin Michlmayr Cc: pavel@suse.cz, rmk+lkml@arm.linux.org.uk, linux-kernel@vger.kernel.org Subject: Re: [PATCH] Convert serial_core oopses to BUG_ON Message-Id: <20060228153256.64f4781d.akpm@osdl.org> In-Reply-To: <20060228220128.GA4254@unjust.cyrius.com> References: <20060226100518.GA31256@flint.arm.linux.org.uk> <20060226021414.6a3db942.akpm@osdl.org> <20060227141315.GD2429@ucw.cz> <20060228101713.6fd44027.akpm@osdl.org> <20060228220128.GA4254@unjust.cyrius.com> X-Mailer: Sylpheed version 1.0.4 (GTK+ 1.2.10; i386-redhat-linux-gnu) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Martin Michlmayr wrote: > > * Andrew Morton [2006-02-28 10:17]: > > > It will oops in hard-to-guess, place, anyway. > > Will it? Where? Unfixably? > > http://www.linux-mips.org/archives/linux-mips/2006-02/msg00241.html is > one example we just had on MIPS. On SGI IP22, using the serial > console, you'd get the following on shutdown: > > The system is going down for reboot NOW! > INIT: Sending processes the TERM signal > INIT: Sending proces > > and then nothing at all. I'd never have suspected the serial driver, > had not users reported that the machine shutdowns properly when using > the framebuffer. > > For the record, I don't mind whether it's BUG_ON or WARN_ON, but I > just wanted to give this as an example of an "oops in hard-to-guess, > place". >>From my reading of the above thread, putting the proposed workaround into serial core will indeed allow people's machines to keep running while reminding us about the driver bugs.