From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Tue, 21 Aug 2001 22:49:20 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Tue, 21 Aug 2001 22:49:10 -0400 Received: from femail46.sdc1.sfba.home.com ([24.254.60.40]:49545 "EHLO femail46.sdc1.sfba.home.com") by vger.kernel.org with ESMTP id ; Tue, 21 Aug 2001 22:48:53 -0400 Message-ID: <3B831CDF.4CC930A7@didntduck.org> Date: Tue, 21 Aug 2001 22:45:51 -0400 From: Brian Gerst X-Mailer: Mozilla 4.76 [en] (X11; U; Linux 2.4.9-pre4 i686) X-Accept-Language: en MIME-Version: 1.0 To: Paul , Alan Cox CC: linux-kernel@vger.kernel.org Subject: Re: [OOPS] repeatable 2.4.8-ac7, 2.4.7-ac6 just run xdos In-Reply-To: <20010819004703.A226@squish.home.loc> 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 Paul wrote: > > * Kernel oops, and locks up when I run xdos (dosemu) > > * Occuring in both 2.4.7-ac6 and 2.4.8-ac7. Run 'xdos' in X, and > machine locks hard, only output to console is oops. (no sysrq) > Tried once with strace, but no oops. (didnt wait long, though) > Some oops before window is placed, some a little while after. > (mouse movement?) Repeatable. > > * Kernels are virgin linus patched with ac. AMD-K6(tm) 3D > processor > > * If anyone wants any more info or for me to do anything, just > ask. > > Paul > set@pobox.com > > (2.4.7-ac6 -- two captured, identitcal, first shown) > > ksymoops 2.4.1 on i586 2.4.7-ac6. Options used > -V (default) > -k /proc/ksyms (default) > -l /proc/modules (default) > -o /lib/modules/2.4.7-ac6/ (default) > -m /boot/System.map-2.4.7-ac6 (specified) > > CPU: 0 > EIP: 0010:[] > Using defaults from ksymoops -t elf32-i386 -a i386 > EFLAGS: 00010002 > eax: 00001000 ebx: c4562368 ecx: 00000000 edx: 00000001 > esi: c4562368 edi: c4a954d4 ebp: 00000001 esp: c6887d88 > ds: 008 es: 0000 ss: 0018 ^^^^ Here is your problem. %es is set to the null segment. I had my suspicions about the segment reload optimisation in the -ac kernels, and this proves it. Try backing out the changes to arch/i386/kernel/entry.S and include/asm-i386/hw_irq.h and see if that fixes the problem. -- Brian Gerst