From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1758752AbXGXPUb (ORCPT ); Tue, 24 Jul 2007 11:20:31 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751821AbXGXPUX (ORCPT ); Tue, 24 Jul 2007 11:20:23 -0400 Received: from tomts16.bellnexxia.net ([209.226.175.4]:47208 "EHLO tomts16-srv.bellnexxia.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750903AbXGXPUW (ORCPT ); Tue, 24 Jul 2007 11:20:22 -0400 Date: Tue, 24 Jul 2007 11:20:18 -0400 From: Mathieu Desnoyers To: Jeremy Fitzhardinge Cc: Rusty Russell , Andi Kleen , jbeulich@novell.com, zach@vmware.com, patches@x86-64.org, linux-kernel@vger.kernel.org Subject: Re: [PATCH] [2/11] x86: Fix alternatives and kprobes to remap write-protected kernel text Message-ID: <20070724152018.GA31729@Krystal> References: <20070720532.949522000@suse.de> <20070720153246.6597614E79@wotan.suse.de> <1185257169.1803.213.camel@localhost.localdomain> <20070724135750.GD26634@Krystal> <46A61821.3070703@goop.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Content-Disposition: inline In-Reply-To: <46A61821.3070703@goop.org> X-Editor: vi X-Info: http://krystal.dyndns.org:8080 X-Operating-System: Linux/2.6.21.3-grsec (i686) X-Uptime: 11:19:42 up 7 days, 9:53, 3 users, load average: 0.56, 0.39, 0.38 User-Agent: Mutt/1.5.13 (2006-08-11) Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org * Jeremy Fitzhardinge (jeremy@goop.org) wrote: > Mathieu Desnoyers wrote: > > Hi Rusty, > > > > Good catch. I also still wonder why alternative (except alternative smp, > > which is ok) and paravirt patching functions are not marked __init. I really > > hope they are never used when NMI, MCE are enabled or when threads may > > have been preempted in the site being patched, or it could result in an > > illegal instruction. > > > > We need to patch modules at load time. Presumably an NMI/MCE handler > will not depend on code in a being-loaded module. > > J > Oh yes, I see. That's ok since the code of a being-loaded module is not executed yet. Thanks, Mathieu -- Mathieu Desnoyers Computer Engineering Ph.D. Student, Ecole Polytechnique de Montreal OpenPGP key fingerprint: 8CD5 52C3 8E3C 4140 715F BA06 3F25 A8FE 3BAE 9A68