From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ben Hutchings Subject: Re: 8139cp vs 8139too, request_module ? Date: Wed, 10 Jun 2009 18:00:06 +0100 Message-ID: <1244653206.2848.15.camel@achroite> References: <1244649729.14601.1.camel@plop> Mime-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 7bit Cc: netdev@vger.kernel.org To: Pascal Terjan Return-path: Received: from smarthost03.mail.zen.net.uk ([212.23.3.142]:51651 "EHLO smarthost03.mail.zen.net.uk" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754843AbZFJRAJ (ORCPT ); Wed, 10 Jun 2009 13:00:09 -0400 In-Reply-To: <1244649729.14601.1.camel@plop> Sender: netdev-owner@vger.kernel.org List-ID: On Wed, 2009-06-10 at 18:02 +0200, Pascal Terjan wrote: > Hello, > 8139cp and 8139too both handle the same id and then test revision. > > If revision is wrong they tell to load the other module and return > ENODEV. > > Why not doing a request_module instead of printing a message ? The kernel requests a module to handle each PCI device it doesn't already have a driver for. modprobe will then load all modules that match that PCI device and vendor id. So 8139cp and 8139too are both loaded for PCI devices that might be handled by either. There is no need for them to try to load each other, and I suspect there would be a risk of deadlock if they did. Ben. -- Ben Hutchings, Senior Software Engineer, Solarflare Communications Not speaking for my employer; that's the marketing department's job. They asked us to note that Solarflare product names are trademarked.