From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261754AbVCKWp5 (ORCPT ); Fri, 11 Mar 2005 17:45:57 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261773AbVCKWo0 (ORCPT ); Fri, 11 Mar 2005 17:44:26 -0500 Received: from ozlabs.org ([203.10.76.45]:63139 "EHLO ozlabs.org") by vger.kernel.org with ESMTP id S261726AbVCKWme (ORCPT ); Fri, 11 Mar 2005 17:42:34 -0500 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <16946.7941.404582.764637@cargo.ozlabs.ibm.com> Date: Sat, 12 Mar 2005 09:43:17 +1100 From: Paul Mackerras To: Bjorn Helgaas Cc: Jesse Barnes , Benjamin Herrenschmidt , werner@sgi.com, Linus Torvalds , davej@redhat.com, Linux Kernel list Subject: Re: AGP bogosities In-Reply-To: <1110563965.4822.22.camel@eeyore> References: <16944.62310.967444.786526@cargo.ozlabs.ibm.com> <200503102002.47645.jbarnes@engr.sgi.com> <1110515459.32556.346.camel@gaston> <200503110839.15995.jbarnes@engr.sgi.com> <1110563965.4822.22.camel@eeyore> X-Mailer: VM 7.19 under Emacs 21.3.1 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Bjorn Helgaas writes: > HP ia64 and parisc boxes are similar. The host bridges do not appear > as PCI devices. We discover them via ACPI on ia64 and PDC on parisc. On PPC/PPC64 machines, the host bridges generally do not appear as PCI devices either. *However*, the AGP spec requires a set of registers in PCI config space for controlling the target (host) side of the AGP bus. In other words you are required to have a PCI device to represent the host side of the AGP bus, with a capability structure in its config space which contains the standard AGP registers. The lspci that was posted showed no such device, which was why Ben was querying it. Maybe your systems aren't fully AGP-compliant, in which case much of the generic code won't be usable on your systems. Paul.