From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S965005AbVITM7Z (ORCPT ); Tue, 20 Sep 2005 08:59:25 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S965007AbVITM7Z (ORCPT ); Tue, 20 Sep 2005 08:59:25 -0400 Received: from jurassic.park.msu.ru ([195.208.223.243]:3211 "EHLO jurassic.park.msu.ru") by vger.kernel.org with ESMTP id S965005AbVITM7Z (ORCPT ); Tue, 20 Sep 2005 08:59:25 -0400 Date: Tue, 20 Sep 2005 16:59:02 +0400 From: Ivan Kokshaysky To: Andreas Koch Cc: linux-kernel@vger.kernel.org, linux-pci@atrey.karlin.mff.cuni.cz Subject: Re: Continuing PCI and Yenta troubles in 2.6.13.1 and 2.6.14-rc1 Message-ID: <20050920165902.A27065@jurassic.park.msu.ru> References: <200509030138.11905.koch@esa.informatik.tu-darmstadt.de> <20050913063053.GA24158@suse.de> <20050913154529.C15709@jurassic.park.msu.ru> <200509201202.24318.koch@esa.informatik.tu-darmstadt.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: <200509201202.24318.koch@esa.informatik.tu-darmstadt.de>; from koch@esa.informatik.tu-darmstadt.de on Tue, Sep 20, 2005 at 12:02:23PM +0200 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Tue, Sep 20, 2005 at 12:02:23PM +0200, Andreas Koch wrote: > Thus, I believe there is still something rotten in the PCI code since the > 2.6.12-rc6 days. Thanks a lot for the report. You are right, and I believe the change that broke things is introduction of pci_fixup_parent_subordinate_busnr()... The patch here does two things: - hunk #1 should fix the problems you've seen when you boot without additional "pci" kernel options; - hunk #2 supposedly fixes boot with "pci=assign-busses" option which otherwise hangs Acer TM81xx machines as reported. Please try this with and without "pci=assign-busses". If it boots, I'd like to see 'lspci -vvx' for both cases. Ivan. --- 2.6.13.2/drivers/pci/probe.c Mon Aug 8 12:20:28 2005 +++ linux/drivers/pci/probe.c Tue Sep 20 16:17:23 2005 @@ -400,6 +400,12 @@ static void pci_enable_crs(struct pci_de static void __devinit pci_fixup_parent_subordinate_busnr(struct pci_bus *child, int max) { struct pci_bus *parent = child->parent; + + /* Attempts to fix that up are really dangerous unless + we're going to re-assign all bus numbers. */ + if (!pcibios_assign_all_busses()) + return; + while (parent->parent && parent->subordinate < max) { parent->subordinate = max; pci_write_config_byte(parent->self, PCI_SUBORDINATE_BUS, max); @@ -476,8 +482,18 @@ int __devinit pci_scan_bridge(struct pci * We need to assign a number to this bus which we always * do in the second pass. */ - if (!pass) + if (!pass) { + if (pcibios_assign_all_busses()) + /* Temporarily disable forwarding of the + configuration cycles on all bridges in + this bus segment to avoid possible + conflicts in the second pass between two + bridges programmed with overlapping + bus ranges. */ + pci_write_config_dword(dev, PCI_PRIMARY_BUS, + buses & ~0xffffff); return max; + } /* Clear errors */ pci_write_config_word(dev, PCI_STATUS, 0xffff);