From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:57293) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1fdH76-00036C-5Q for qemu-devel@nongnu.org; Wed, 11 Jul 2018 11:33:08 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1fdH70-0005GM-Ah for qemu-devel@nongnu.org; Wed, 11 Jul 2018 11:33:08 -0400 Received: from mx1.redhat.com ([209.132.183.28]:48875) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1fdH70-0005FR-4D for qemu-devel@nongnu.org; Wed, 11 Jul 2018 11:33:02 -0400 Date: Wed, 11 Jul 2018 12:32:54 -0300 From: Eduardo Habkost Message-ID: <20180711153254.GR7451@localhost.localdomain> References: <20180709203731.19865-1-ehabkost@redhat.com> <20180710090731.GC1612@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Disposition: inline In-Reply-To: <20180710090731.GC1612@redhat.com> Content-Transfer-Encoding: quoted-printable Subject: Re: [Qemu-devel] [PATCH for-3.0] pc: Use "3.0+" constant as default SMBIOS version List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Daniel =?iso-8859-1?Q?P=2E_Berrang=E9?= Cc: qemu-devel@nongnu.org, "Michael S. Tsirkin" , Wei Huang , Peter Maydell , Andrew Jones , Igor Mammedov On Tue, Jul 10, 2018 at 10:07:31AM +0100, Daniel P. Berrang=E9 wrote: > On Mon, Jul 09, 2018 at 05:37:31PM -0300, Eduardo Habkost wrote: > > Every time we create new PC machine-types in QEMU, the defaults > > for SMBIOS fields change unnecessarily because the version field > > defaults to MachineClass::name. > >=20 > > This can cause unexpected side-effects, like triggering license > > reactivation on guest software, or changing the VM memory layout > > because of BIOS table size changes. >=20 > Does that really matter though ? By its very nature the 'Version' > field in SMBIOS is expected to change if you alter something about > the hardware. If guests OS don't want to be exposed to changes in > SMBIOS they would be using a fixed machine type, not the variable > "pc" type that continually changes. >=20 > We could put padding in the string if we want to avoid BIOS table > layout changes. >=20 > Having version change though feels like it is working as intended > for the semantics of these Version: fields in BIOS. Michael, do you have additional info on the original motivation for suggesting this change and why do you consider it a bug? (I don't have any concrete examples to justify the change) --=20 Eduardo