From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andi Kleen Subject: Re: [PATCH] Add MCE support to KVM Date: Mon, 13 Apr 2009 10:26:03 +0200 Message-ID: <20090413082603.GS14687@one.firstfloor.org> References: <1239155601.6384.3.camel@yhuang-dev.sh.intel.com> <49DE195D.1020303@redhat.com> <1239332455.6384.108.camel@yhuang-dev.sh.intel.com> <49E08762.1010206@redhat.com> <20090411121911.GR14687@one.firstfloor.org> <49E08C2F.4010003@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Andi Kleen , Huang Ying , "kvm@vger.kernel.org" , "linux-kernel@vger.kernel.org" To: Avi Kivity Return-path: Received: from one.firstfloor.org ([213.235.205.2]:45456 "EHLO one.firstfloor.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752775AbZDMIX1 (ORCPT ); Mon, 13 Apr 2009 04:23:27 -0400 Content-Disposition: inline In-Reply-To: <49E08C2F.4010003@redhat.com> Sender: kvm-owner@vger.kernel.org List-ID: On Sat, Apr 11, 2009 at 03:25:19PM +0300, Avi Kivity wrote: > Andi Kleen wrote: > >>Right, but we can allocate the maximum number, no? it's a fairly small > >>amount of memory. > >> > > > >There are 255 banks worst case which need upto 4 (in theory 5) MSRs each > > > > > > That's 8KB, which would normally not be allocated. But I think we can > live with it. > > Do we actually need to support 255 banks? Probably not. Current CPUs all have a single digit number. e.g. 64 would be plenty for now. > Or can we support a smaller > number transparently? Not transparently. In theory software can hardcode that for cpu model X bank Y is available and does Z. Software doesn't have too, Intel provides a bank independent way to report errors too, but we don't know if everyone uses it. Linux doesn't rely on bank numbers on Intel at least. -Andi -- ak@linux.intel.com -- Speaking for myself only.