Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754986AbZGMJCY (ORCPT ); Mon, 13 Jul 2009 05:02:24 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1755021AbZGMJCV (ORCPT ); Mon, 13 Jul 2009 05:02:21 -0400 Received: from one.firstfloor.org ([213.235.205.2]:60342 "EHLO one.firstfloor.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755050AbZGMJCR (ORCPT ); Mon, 13 Jul 2009 05:02:17 -0400 To: Arjan van de Ven Cc: Greg KH , Rusty Russell , Ingo Molnar , Siarhei Liakh , linux-kernel@vger.kernel.org, linux-security-module@vger.kernel.org, James Morris , Andrew Morton , Andi Kleen , Thomas Gleixner , "H. Peter Anvin" , linux-cris-kernel@axis.com Subject: Re: [PATCH v5] RO/NX protection for loadable kernel modules From: Andi Kleen References: <817ecb6f0907081610p6d60341cudbee42685eac1347@mail.gmail.com> <200907121410.39874.rusty@rustcorp.com.au> <20090712004524.2f0c4e57@infradead.org> <200907121928.28887.rusty@rustcorp.com.au> <20090712083227.39939849@infradead.org> <20090712173329.GB12054@kroah.com> <20090712145804.2f1fce98@infradead.org> Date: Mon, 13 Jul 2009 11:02:14 +0200 In-Reply-To: <20090712145804.2f1fce98@infradead.org> (Arjan van de Ven's message of "Sun, 12 Jul 2009 14:58:04 -0700") Message-ID: <87k52d9crt.fsf@basil.nowhere.org> User-Agent: Gnus/5.1008 (Gnus v5.10.8) Emacs/22.3 (gnu/linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1498 Lines: 36 Arjan van de Ven writes: > I've seen some of these case, where the distro kernel has something as > a module, but the other parts of the distro the unconditionally load > that module always. That makes no sense. One good reason for this is that if something goes wrong with the module you can still remove/blacklist the module. This can be very useful in distro deployment, where telling users "please set flag xyz" is much easier than asking them to get a special kernel build. It also helps debugging when you're trying to narrow down where a problem is. But you can't do that with built-in drivers. One way to avoid this would be to have a standard way to turn off drivers/subsystems that are built in on the command line. Right now that's difficult because the linked kernel doesn't even know the driver names anymore. Perhaps we should keep the module names/metadata even in static kernel? (and make CONFIG_MODULE on the subsystem level disappear?). IMHO that would be a great cleanup anyways, avoiding one special case in the driver build testing. It would be also nice if you could cat some file in sys and it gave you the module descriptions for example. -Andi -- ak@linux.intel.com -- Speaking for myself only. -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/