From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from [140.186.70.92] (port=39674 helo=eggs.gnu.org) by lists.gnu.org with esmtp (Exim 4.43) id 1P3E68-0006Wg-2s for qemu-devel@nongnu.org; Tue, 05 Oct 2010 16:26:53 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1P3E67-0005dX-1T for qemu-devel@nongnu.org; Tue, 05 Oct 2010 16:26:51 -0400 Received: from mx1.redhat.com ([209.132.183.28]:46458) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1P3E66-0005dD-QR for qemu-devel@nongnu.org; Tue, 05 Oct 2010 16:26:50 -0400 Date: Tue, 5 Oct 2010 17:24:53 -0300 From: Marcelo Tosatti Subject: Re: [Qemu-devel] [patch uq/master 0/8] port qemu-kvm's MCE support Message-ID: <20101005202453.GA6319@amt.cnet> References: <20101004185447.891324545@redhat.com> <20101005185813.GC11396@sequoia.sous-sol.org> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Disposition: inline In-Reply-To: <20101005185813.GC11396@sequoia.sous-sol.org> Content-Transfer-Encoding: quoted-printable List-Id: qemu-devel.nongnu.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Chris Wright Cc: Andreas =?iso-8859-1?Q?F=E4rber?= , QEMU Developers , kvm@vger.kernel.org On Tue, Oct 05, 2010 at 11:58:13AM -0700, Chris Wright wrote: > * Andreas F=E4rber (andreas.faerber@web.de) wrote: > > Am 04.10.2010 um 20:54 schrieb Marcelo Tosatti: > >=20 > > I assume something went wrong with your cover letter here. It > > would've been nice to see MCE spelled out or summarized for those of > > us that don't speak x86. Sorry about that. Will improve on next submission. > It would help. The acronym is Machine Check Exception. The patchset > should allow (on newer Intel x86 hw with a newer linux kernel) a class = of > memory errors delivered to the host OS as MCEs to be propagated to the > guest OS. Without the patchset, the qemu process assoicated with the > memory where the error took place would be killed. With the patchset, > qemu can propagate the error into the guest and allow the guest to kill > only the process within the guest that is assocated with the memory err= or. > -- Thanks Chris.