From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S268294AbUHKWoN (ORCPT ); Wed, 11 Aug 2004 18:44:13 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S268293AbUHKWnM (ORCPT ); Wed, 11 Aug 2004 18:43:12 -0400 Received: from atlrel7.hp.com ([156.153.255.213]:9449 "EHLO atlrel7.hp.com") by vger.kernel.org with ESMTP id S268292AbUHKWnC (ORCPT ); Wed, 11 Aug 2004 18:43:02 -0400 From: Bjorn Helgaas To: Ralf Gerbig Subject: Re: rc4-mm1 pci-routing Date: Wed, 11 Aug 2004 16:42:59 -0600 User-Agent: KMail/1.6.2 Cc: linux-kernel@vger.kernel.org MIME-Version: 1.0 Content-Disposition: inline Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Message-Id: <200408111642.59938.bjorn.helgaas@hp.com> Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Hi Ralf, Thanks very much for your report. It looks like this device is the problem: 0000:00:06.0 Multimedia audio controller: nVidia Corporation nForce2 AC97 Audio Controler (MCP) (rev a1) and it should get IRQ 21: ACPI: PCI Interrupt Link [APCJ] enabled at IRQ 21 ACPI: PCI interrupt 0000:00:06.0[A] -> GSI 21 (level, high) -> IRQ 21 The driver for this should be intel8x0.o, and it looks like you are loading it as a module. Could you build it in statically to the kernel and collect the complete dmesg from a boot with "pci=routeirq" (one from the boot that hangs would be nice as well, but that is a pain to collect unless you're using a serial console)? The usual problem is that a driver looks at pci_dev->irq before calling pci_enable_device(), but intel8x0.c seems to be doing the right thing in this regard. Bjorn