From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755066AbZBIMec (ORCPT ); Mon, 9 Feb 2009 07:34:32 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1753619AbZBIMeY (ORCPT ); Mon, 9 Feb 2009 07:34:24 -0500 Received: from ftp.linux-mips.org ([213.58.128.207]:38869 "EHLO ftp.linux-mips.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753520AbZBIMeX (ORCPT ); Mon, 9 Feb 2009 07:34:23 -0500 Date: Mon, 9 Feb 2009 12:34:21 +0000 (GMT) From: "Maciej W. Rozycki" To: Ingo Molnar cc: Yinghai Lu , Thomas Gleixner , "H. Peter Anvin" , Andrew Morton , "linux-kernel@vger.kernel.org" Subject: Re: [PATCH] x86: check_timer cleanup In-Reply-To: <20090209114520.GE18757@elte.hu> Message-ID: References: <20090206220716.GA15319@elte.hu> <498D55A4.7000608@kernel.org> <20090207165901.GA32116@elte.hu> <498E6A3C.6020004@kernel.org> <20090209081542.GF24420@elte.hu> <20090209114520.GE18757@elte.hu> User-Agent: Alpine 1.10 (LFD 962 2008-03-14) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Mon, 9 Feb 2009, Ingo Molnar wrote: > > What problems are you referring to, Ingo? Perhaps they should be looked > > into specifically? > > I got a spurious boot failure on a 64-bi testbox: > > [ 0.280000] ..TIMER: vector=0x30 apic1=0 pin1=2 apic2=-1 pin2=-1 > [ 0.296000] ..MP-BIOS bug: 8254 timer not connected to IO-APIC > [ 0.296000] ...trying to set up timer (IRQ0) through the 8259A ... > [ 0.296000] ..... (found apic 0 pin 2) ... > [ 0.308000] ....... failed. > [ 0.308000] ...trying to set up timer as Virtual Wire IRQ... > [ 0.324000] ..... failed. > [ 0.324000] ...trying to set up timer as ExtINT IRQ... > [ 0.340000] ..... failed :(. > [ 0.340000] Kernel panic - not syncing: IO-APIC + timer doesn't work! Boot with apic=debug and send a report. Then try booting with the 'noapic' option. > > Not reproducible. Full bootlog below. If so, it could have been anything. Even a casual hardware failure. The timer retriggers by itself producing both edges automatically, so even if an edge is missed because of the mask, another one follows shortly. I don't think the change proposed by Yinghai would make any difference here. Pity you can't use an in-circuit debugger with the x86... :( Maciej