Re: [libvirt] inconsistent handling of "qemu64" CPU model