From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:51940) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1UpHrT-0003F6-Uo for qemu-devel@nongnu.org; Wed, 19 Jun 2013 08:51:49 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1UpHrP-0008Fp-4S for qemu-devel@nongnu.org; Wed, 19 Jun 2013 08:51:43 -0400 Received: from mx1.redhat.com ([209.132.183.28]:3065) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1UpHrO-0008Dt-M5 for qemu-devel@nongnu.org; Wed, 19 Jun 2013 08:51:38 -0400 Message-ID: <51C1A94C.2090704@redhat.com> Date: Wed, 19 Jun 2013 14:51:24 +0200 From: Paolo Bonzini MIME-Version: 1.0 References: <1371117054-5694-1-git-send-email-paul.durrant@citrix.com> <1371145442.6955.64.camel@zakaz.uk.xensource.com> <51BA0943.7050705@redhat.com> <9AAE0902D5BC7E449B7C8E4E778ABCD00394C5@LONPEX01CL01.citrite.net> <9AAE0902D5BC7E449B7C8E4E778ABCD0039637@LONPEX01CL01.citrite.net> <51BB1FB4.1030006@redhat.com> <9AAE0902D5BC7E449B7C8E4E778ABCD0039B3F@LONPEX01CL01.citrite.net> <51BB2F73.70902@redhat.com> <9AAE0902D5BC7E449B7C8E4E778ABCD0039C9E@LONPEX01CL01.citrite.net> <51C0B119.8070704@redhat.com> <51C0D352.7070701@suse.de> <1371630543.22783.66.camel@zakaz.uk.xensource.com> <51C16EB5.9060804@redhat.com> <1371632169.22783.81.camel@zakaz.uk.xensource.com> In-Reply-To: <1371632169.22783.81.camel@zakaz.uk.xensource.com> Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Subject: Re: [Qemu-devel] [Xen-devel] [PATCH] Remove hardcoded xen-platform device initialization List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Ian Campbell Cc: "qemu-devel@nongnu.org" , Paul Durrant , "xen-devel@lists.xen.org" , =?UTF-8?B?QW5kcmVhcyBGw6RyYmVy?= , Stefano Stabellini Il 19/06/2013 10:56, Ian Campbell ha scritto: > The case with xl is more like your example 5: Ok, that's what I remembered. :) > > Example 5: you use "virsh create" to start a VM based on an XML file, > > rather than "virsh define"+"virsh start" as in examples 1-2. You lose > > any guarantee that hardware does not change. Not frowned upon as much > > as example 4, since the VM is supposed to be transient. > > For something like xapi we'd likely want to support some sort of model > similar to libvirt, so whatever we do at the libxl layer needs to > consider both approaches. I think xapi can itself query qemu if this is the case (in this respect, libvirt is somewhat in the middle between xl and xapi, depending on whether you use transient or persistent domains). Or perhaps you can add an xl command to query qemu and provide capabilities. xl can default to some "blessed" old-enough machine if it is available or just the newest "pc" if the blessed machine is unavailable. And of course it can be overridden by the guest config. Paolo