From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S965316AbXAKIGW (ORCPT ); Thu, 11 Jan 2007 03:06:22 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S965323AbXAKIGW (ORCPT ); Thu, 11 Jan 2007 03:06:22 -0500 Received: from il.qumranet.com ([62.219.232.206]:57963 "EHLO il.qumranet.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S965316AbXAKIGV (ORCPT ); Thu, 11 Jan 2007 03:06:21 -0500 Message-ID: <45A5EFFC.2080609@qumranet.com> Date: Thu, 11 Jan 2007 10:06:20 +0200 From: Avi Kivity User-Agent: Thunderbird 1.5.0.9 (X11/20061219) MIME-Version: 1.0 To: Roland Dreier CC: kvm-devel , linux-kernel@vger.kernel.org Subject: Re: [kvm-devel] guest crash on 2.6.20-rc4 References: <45A40898.4040307@qumranet.com> In-Reply-To: Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Roland Dreier wrote: > > if (is_writeble_pte(*shadow_ent)) > > - return 0; > > + return 1; > > With this patch, it looks like my guest is surviving the load that > triggered the oops before. So I think this fixes the issue I saw as well. > I assume you'll send this in for 2.6.20? > The patch actually replaces one bug (guest pagefaults on writable dirty ptes, under certain conditions) with another, rarer one (spinning on a user-mode pagefault on writable dirty kernel ptes). I'll do it right and re-test, then send for .20 along with a few friends. -- error compiling committee.c: too many arguments to function From mboxrd@z Thu Jan 1 00:00:00 1970 From: Avi Kivity Subject: Re: guest crash on 2.6.20-rc4 Date: Thu, 11 Jan 2007 10:06:20 +0200 Message-ID: <45A5EFFC.2080609@qumranet.com> References: <45A40898.4040307@qumranet.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Cc: kvm-devel , linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org Return-path: To: Roland Dreier In-Reply-To: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: kvm-devel-bounces-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org Errors-To: kvm-devel-bounces-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org List-Id: kvm.vger.kernel.org Roland Dreier wrote: > > if (is_writeble_pte(*shadow_ent)) > > - return 0; > > + return 1; > > With this patch, it looks like my guest is surviving the load that > triggered the oops before. So I think this fixes the issue I saw as well. > I assume you'll send this in for 2.6.20? > The patch actually replaces one bug (guest pagefaults on writable dirty ptes, under certain conditions) with another, rarer one (spinning on a user-mode pagefault on writable dirty kernel ptes). I'll do it right and re-test, then send for .20 along with a few friends. -- error compiling committee.c: too many arguments to function ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV