From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755304Ab0EGOsB (ORCPT ); Fri, 7 May 2010 10:48:01 -0400 Received: from nef2.ens.fr ([129.199.96.40]:2974 "EHLO nef2.ens.fr" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755125Ab0EGOr7 convert rfc822-to-8bit (ORCPT ); Fri, 7 May 2010 10:47:59 -0400 X-Greylist: delayed 3040 seconds by postgrey-1.27 at vger.kernel.org; Fri, 07 May 2010 10:47:59 EDT From: =?utf-8?q?=C3=89ric_Brunet?= Organization: UPMC LPS-ENS To: linux-kernel@vger.kernel.org Subject: Some problem with nouveau Date: Fri, 7 May 2010 15:57:15 +0200 User-Agent: KMail/1.13.2 (Linux/2.6.32.11-99.fc12.x86_64; KDE/4.4.2; x86_64; ; ) Cc: eric.brunet@lps.ens.fr MIME-Version: 1.0 Content-Type: Text/Plain; charset="utf-8" Content-Transfer-Encoding: 8BIT Message-Id: <201005071557.15728.Eric.Brunet@lps.ens.fr> X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.1.4 (nef2.ens.fr [129.199.96.32]); Fri, 07 May 2010 15:57:18 +0200 (CEST) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hello, I have HP Z400 desktop computer running in x86_64 mode with a fedora 12 installed on it. The computer has an nvidia (NV96) card, for which I am using the nouveau driver. Everything (meaning 2D acceleration) was working fine up to kernel-2.6.31.12-174.2.22, and when fedora updated to 2.6.32.xxx (now kernel-2.6.32.11-99) it stopped working. I also tried a vanilla 2.6.33.3 and it doesn't work either (I wouldn't dream bothering LKML without trying a vanilla kernel...) Here are the symptoms: when X is launched, the computer reboots instantaneously to the BIOS greeting screen, leaving the computer in such a state that the BIOS is enable to work (it never reaches grub, but keeps on rebooting after a few seconds). Shutting down the computer is necessary to recover. I have no useful log as the buffer have not been flushed: /var/log/Xorg.0.log is found empty on the next reboot. I filled a bug with fedora with more info (dmesg, lspci) on https://bugzilla.redhat.com/show_bug.cgi?id=578108 but I had zero response. So to sum up: 2.6.31.12 + fedora patches is working 2.6.32.xx + fedora patches is crashing horribly 2.6.33.xx vanilla is crashing in the same way. Does that count as a regression ? I am a little bit at a loss as how to provide debugging information on that one. Thanks, Éric Brunet