From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S964931AbWFSWMR (ORCPT ); Mon, 19 Jun 2006 18:12:17 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S964940AbWFSWMQ (ORCPT ); Mon, 19 Jun 2006 18:12:16 -0400 Received: from iabervon.org ([66.92.72.58]:22537 "EHLO iabervon.org") by vger.kernel.org with ESMTP id S964931AbWFSWMP (ORCPT ); Mon, 19 Jun 2006 18:12:15 -0400 Date: Mon, 19 Jun 2006 18:14:28 -0400 (EDT) From: Daniel Barkalow To: "Alexander E. Patrakov" cc: linux-kernel@vger.kernel.org, Joshua Hudson , "H. Peter Anvin" Subject: Re: [GIT PATCH] Remove devfs from 2.6.17 In-Reply-To: <44964EDC.3030104@ums.usu.ru> Message-ID: References: <20060618221343.GA20277@kroah.com> <20060618230041.GG4744@bouh.residence.ens-lyon.fr> <4495F5C3.1030203@zytor.com> <4495FF59.2010100@zytor.com> <8e6f94720606182255u400964c2v1ea16221ffc5c94d@mail.gmail.com> <44964EDC.3030104@ums.usu.ru> 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 On Mon, 19 Jun 2006, Alexander E. Patrakov wrote: > Will Dyson wrote: > > Providing the information about what devices a virtual driver will > > register when loaded seems like a good idea. > > Why? This information is currently useless. What you want is that something > knows that you want this driver to be loaded. The point is that you *don't* want those modules to be loaded. What you want is for the kernel to know that those modules are available, and therefore mention that drivers could be found for those devices, and therefore udev would create the device nodes for them, even though the kernel doesn't contain a module that drives them yet. If something actually opens the device node, the module will be loaded, but until then it isn't using up resources. -Daniel *This .sig left intentionally blank*