Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1760738AbYFDSqk (ORCPT ); Wed, 4 Jun 2008 14:46:40 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1754818AbYFDSqc (ORCPT ); Wed, 4 Jun 2008 14:46:32 -0400 Received: from mx1.redhat.com ([66.187.233.31]:51816 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753286AbYFDSqb (ORCPT ); Wed, 4 Jun 2008 14:46:31 -0400 Subject: Re: [PATCH 0/7] Boot IRQ quirks and rerouting From: Jon Masters To: Stefan Assmann Cc: "Eric W. Biederman" , Olaf Dabrunz , Thomas Gleixner , Ingo Molnar , "H. Peter Anvin" , linux-kernel@vger.kernel.org In-Reply-To: <4846651F.4070802@suse.de> References: <12124107071847-git-send-email-od@suse.de> <4846651F.4070802@suse.de> Content-Type: text/plain Date: Wed, 04 Jun 2008 14:44:00 -0400 Message-Id: <1212605040.8410.8.camel@londonpacket.bos.redhat.com> Mime-Version: 1.0 X-Mailer: Evolution 2.12.3 (2.12.3-4.fc8) Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1422 Lines: 32 On Wed, 2008-06-04 at 11:49 +0200, Stefan Assmann wrote: > On the chips (ICHx, ...) we saw, the interrupt lines on the PIC also go > to the first IO-APIC. So the boot interrupts go to both the PIC and the > first IO-APIC. Yup, it's system dependent too, but it's a mess in many cases. I've seen numerous systems falling over - typically, the classical case will be some IO controller when under heavy load will stop doing interrupts. Another gotcha is that a lot of the time, these legacy interrupt (boot interrupts) happen to be shared between e.g. a disk controller and a USB host controller. On a server system running -RT, it's quite common that there won't be much going on with the other device, so the problem goes unnoticed for a long time...but it still bites you in the end. > When running in APIC mode all PIC IRQs are disabled, except for the > timer maybe. Boot interrupts still arrive on the first IO-APIC and end > up as being counted as spurious interrupts. Yup. I think the good thing is that both SuSE and Red Hat have identified the same problems, posited the same root cause, and contemplated similar workarounds...so I think we understand it. Jon. -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/