From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1422875AbWKPKEr (ORCPT ); Thu, 16 Nov 2006 05:04:47 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1422910AbWKPKEq (ORCPT ); Thu, 16 Nov 2006 05:04:46 -0500 Received: from iona.labri.fr ([147.210.8.143]:3763 "EHLO iona.labri.fr") by vger.kernel.org with ESMTP id S1422875AbWKPKEp (ORCPT ); Thu, 16 Nov 2006 05:04:45 -0500 Message-ID: <455C37B9.4020000@ens-lyon.org> Date: Thu, 16 Nov 2006 11:04:41 +0100 From: Brice Goglin User-Agent: Icedove 1.5.0.7 (X11/20061013) MIME-Version: 1.0 To: Roland Dreier CC: LKML Subject: [PATCH] ALSA: hda-intel - Disable MSI support by default] X-Enigmail-Version: 0.94.0.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org (forgot to cc: LKML) Roland Dreier wrote: > > A whitelist is an awkward solution, the problem is the number of > > chipsets available with MSI will continue to grow. And the assumption > > is that after Microsoft OS supports MSI, that newer chipsets will work. > > Maybe a whitelist for older systems and then enable everything after a > DMI cutoff date by default? Doesn't work on non-PC stuff though... When I started sending patches about all this, I proposed a whitelist for PCI bridges and a blacklist for PCIe bridges, but Greg was against whitelisting. Now, reading this thread, I am very disturbed with the idea of disabling MSI on all non-Intel bridges. Blacklisting correctly shouldn't be that hard. From my point of view, there are several non-Intel bridges that seem to support MSI very well (especially some HT<->PCIe bridges). And apart from being required for ipath, MSI improves the performance of network drivers a bit, so I'd rather keep it enabled on trusted bridges... Brice