From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753351Ab2DWPiE (ORCPT ); Mon, 23 Apr 2012 11:38:04 -0400 Received: from acsinet15.oracle.com ([141.146.126.227]:48535 "EHLO acsinet15.oracle.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752746Ab2DWPiB (ORCPT ); Mon, 23 Apr 2012 11:38:01 -0400 Date: Mon, 23 Apr 2012 11:32:22 -0400 From: Konrad Rzeszutek Wilk To: "Luck, Tony" Cc: Borislav Petkov , "Liu, Jinsong" , "x86@kernel.org" , "linux-edac@vger.kernel.org" , "xen-devel@lists.xensource.com" , "linux-kernel@vger.kernel.org" , "H. Peter Anvin" , Ingo Molnar , Thomas Gleixner Subject: Re: [Xen-devel] [PATCH 1/3] Add mcelog support for xen platform Message-ID: <20120423153222.GE24481@phenom.dumpdata.com> References: <20120421041454.GA29704@phenom.dumpdata.com> <20120421104502.GB17005@aftab.osrc.amd.com> <3908561D78D1C84285E8C5FCA982C28F170EF841@ORSMSX104.amr.corp.intel.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <3908561D78D1C84285E8C5FCA982C28F170EF841@ORSMSX104.amr.corp.intel.com> User-Agent: Mutt/1.5.21 (2010-09-15) X-Source-IP: ucsinet22.oracle.com [156.151.31.94] Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Mon, Apr 23, 2012 at 03:27:32PM +0000, Luck, Tony wrote: > > Because, if you'd hooked into it, just imagine one fine day, when we > > remove mcelog support, what screaming the xen people will be doing when > > mcelog doesn't work anymore. > > Agreed. Even before we get to deleting mcelog, "struct mce" can change (new > fields could be added) ... and you don't want to have your hypervisor to > have to know which version of Linux it is talking to. I am having a hard time seeing how this is different from 'struct mce' changing and the hardware remaining the same? Can't the MCE check drivers deal with that - I mean that is their purpose - to extract some "blob" of data from the hardware, massage it in the software structure, and send its way. If you add some more fields in the software structure - either the hardware can provide it or not. If not, it will have to emulate it. This would be the same case with this driver. This driver is _not_ doing a strict bit-by-bit copy of 'struct mcinfo' to 'struct mce'. It is plucking the right bits and setting them in 'struct mce'.