From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261570AbUJXRwt (ORCPT ); Sun, 24 Oct 2004 13:52:49 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261565AbUJXRwt (ORCPT ); Sun, 24 Oct 2004 13:52:49 -0400 Received: from clock-tower.bc.nu ([81.2.110.250]:44423 "EHLO localhost.localdomain") by vger.kernel.org with ESMTP id S261570AbUJXRuB (ORCPT ); Sun, 24 Oct 2004 13:50:01 -0400 Subject: Re: [PATCH] PCI fixes for 2.6.9 From: Alan Cox To: Greg KH Cc: Linux Kernel Mailing List In-Reply-To: <20041022234508.GA28380@kroah.com> References: <10982257353682@kroah.com> <10982257352301@kroah.com> <20041020091045.D1047@flint.arm.linux.org.uk> <20041022234508.GA28380@kroah.com> Content-Type: text/plain Content-Transfer-Encoding: 7bit Message-Id: <1098636436.24301.10.camel@localhost.localdomain> Mime-Version: 1.0 X-Mailer: Ximian Evolution 1.4.6 (1.4.6-2) Date: Sun, 24 Oct 2004 17:47:18 +0100 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Sad, 2004-10-23 at 00:45, Greg KH wrote: > > It's not that the driver is buggy. It's that the driver has far > > more information than the PCI layer could ever have. > > Ugh, I hate broken hardware. I'll revert this in my next round of pci > changes (sometime next week.) It isnt always broken hardware either. You have no idea at the core level how the internals are configured. Thus for example if a device provides both legacy and non-legacy ports disabling it can do horrible things on pci unregister (eg soundcards with legacy joystick mappings) As to video, well we don't always have an owner for video but unloading frame buffers doesn't want to turn the video card off either..