From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:35336) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1VEYrA-0005Fh-N3 for qemu-devel@nongnu.org; Wed, 28 Aug 2013 02:03:57 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1VEYr5-0002QQ-Qv for qemu-devel@nongnu.org; Wed, 28 Aug 2013 02:03:52 -0400 Received: from mx1.redhat.com ([209.132.183.28]:16392) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1VEYr5-0002QH-Hh for qemu-devel@nongnu.org; Wed, 28 Aug 2013 02:03:47 -0400 Date: Wed, 28 Aug 2013 09:05:42 +0300 From: "Michael S. Tsirkin" Message-ID: <20130828060542.GA2386@redhat.com> References: <1376659114-6630-1-git-send-email-armbru@redhat.com> <1376659114-6630-8-git-send-email-armbru@redhat.com> <20130827170444.GA29703@redhat.com> <521D34AB.5080700@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <521D34AB.5080700@redhat.com> Subject: Re: [Qemu-devel] [PATCH v2 7/7] smbios: Set system manufacturer, product & version by default List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Eric Blake Cc: aliguori@us.ibm.com, lersek@redhat.com, armbru@redhat.com, ehabkost@redhat.com, qemu-devel@nongnu.org On Tue, Aug 27, 2013 at 05:22:19PM -0600, Eric Blake wrote: > On 08/27/2013 11:04 AM, Michael S. Tsirkin wrote: > > On Fri, Aug 16, 2013 at 03:18:34PM +0200, armbru@redhat.com wrote: > >> From: Markus Armbruster > >> > >> Currently, we get SeaBIOS defaults: manufacturer Bochs, product Bochs, > >> no version. Best SeaBIOS can do, but we can provide better defaults: > >> manufacturer QEMU, product & version taken from QEMUMachine desc and > >> name. > >> > >> Take care to do this only for new machine types, of course. > >> > >> Signed-off-by: Markus Armbruster > >> Reviewed-by: Eric Blake > > > > We can do this of course, but why is this better? > > It seems to expose new information to the guest > > for no reason, we just might come to regret it > > later if guests start implementing hacks keying > > off the version number. > > Some guests (cough: some OEM builds of windows) already DO key off of > SMBIOS information to determine if they are running in a valid > environment. Well they are unlikely to like either QEMU or Bochs. > Furthermore, making this change to qemu makes it easier to > decouple the strings being presented to guests by default; it's always > better to have a situation where changing just qemu works, instead of > having to patch both qemu and BIOS in tandem. > > The choice of whether to present the different information MUST be tied > to machine types (we cannot change SMBIOS data without an explicit > change to a newer machine type, precisely _because_ there are guests > that base their licensing decisions on constancy of BIOS information). > But for a new machine type, presenting qemu as the machine type rather > than being stuck to a particular SeaBIOS build seems nicer. Yes but why change anything at all? We have command line flags to set these fields for whatever you want to boot OEM windows - which is highly unlikely to match any QEMU version. Why isn't this enough? > -- > Eric Blake eblake redhat com +1-919-301-3266 > Libvirt virtualization library http://libvirt.org >