From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:49711) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1Z1z2N-0004nV-5R for qemu-devel@nongnu.org; Mon, 08 Jun 2015 11:32:38 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1Z1z2I-0004Ez-E1 for qemu-devel@nongnu.org; Mon, 08 Jun 2015 11:32:31 -0400 Received: from mx1.redhat.com ([209.132.183.28]:45243) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1Z1z2I-0004Eo-6O for qemu-devel@nongnu.org; Mon, 08 Jun 2015 11:32:26 -0400 Received: from int-mx14.intmail.prod.int.phx2.redhat.com (int-mx14.intmail.prod.int.phx2.redhat.com [10.5.11.27]) by mx1.redhat.com (Postfix) with ESMTPS id D636D19F275 for ; Mon, 8 Jun 2015 15:32:25 +0000 (UTC) Date: Mon, 8 Jun 2015 17:32:23 +0200 From: "Michael S. Tsirkin" Message-ID: <20150608173142-mutt-send-email-mst@redhat.com> References: <556F2D51.2000203@redhat.com> <55759BB8.3010604@redhat.com> <55759C1B.7000905@redhat.com> <55759E0A.3090102@redhat.com> <55759EC2.5090802@redhat.com> <20150608135854.GC19157@redhat.com> <20150608165951-mutt-send-email-mst@redhat.com> <5575B20B.9000502@redhat.com> <20150608171921-mutt-send-email-mst@redhat.com> <5575B497.9030808@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <5575B497.9030808@redhat.com> Subject: Re: [Qemu-devel] [PATCH V15 4/5] i386: add a Virtual Machine Generation ID device List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Gal Hammer Cc: Paolo Bonzini , qemu-devel@nongnu.org, imammedo@redhat.com On Mon, Jun 08, 2015 at 06:28:23PM +0300, Gal Hammer wrote: > On 08/06/2015 18:22, Michael S. Tsirkin wrote: > >On Mon, Jun 08, 2015 at 05:17:31PM +0200, Paolo Bonzini wrote: > >> > >> > >>On 08/06/2015 17:01, Michael S. Tsirkin wrote: > >>>Are there applications that would actually use this? For microsoft this > >>>seems to be mostly driven by ActiveDirectory needs. > >>>It seems quite possible that applications solve the problem > >>>differently on Linux. > >> > >>I'm not aware of a different solution to the problem on Linux. > >> > >>Paolo > > > >Adding it by default for everyone still seems too aggressive. > >We had a ton of pain with pvpanic exactly because of a > >similar "can't hurt" approach. > > > >I think a better approach would be to merge the code in qemu, > >merge a linux driver. Once that's available, and some apps use it, > >enabling it by default will seem more reasonable. > > > > You should not enable it by default. It should be left for the management > system to decide. QEmu by itself can't follow rules/spec on when to change > the UUID. > > Gal. I agree. And IMHO at this early stage, neither should libvirt without an explicit request. -- MST