From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1030476AbVJGQQF (ORCPT ); Fri, 7 Oct 2005 12:16:05 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1030478AbVJGQQF (ORCPT ); Fri, 7 Oct 2005 12:16:05 -0400 Received: from ms-smtp-02.nyroc.rr.com ([24.24.2.56]:49826 "EHLO ms-smtp-02.nyroc.rr.com") by vger.kernel.org with ESMTP id S1030476AbVJGQQD (ORCPT ); Fri, 7 Oct 2005 12:16:03 -0400 Date: Fri, 7 Oct 2005 12:15:58 -0400 (EDT) From: Steven Rostedt X-X-Sender: rostedt@localhost.localdomain To: linux-kernel@vger.kernel.org Subject: Dell firmware default config options? In-Reply-To: <43469A9A.2070104@beezmo.com> Message-ID: References: <43469A9A.2070104@beezmo.com> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org I'm just curious why the Dell firmware configuration options are default to "m" instead of "n". Since it only matters if you have a Dell System. So for the huge number of systems that are not Dell Systems, they are probably wasting CPU cycles compiling these as modules, and taking up space in loads of /lib/modules directories throughout the world ;-) DCDBAS explicitly states default of "m", and DELL_RBU has no default which just makes it automatically on. Is there any reason that these shouldn't be turned off by default? Thanks, -- Steve