Laszlo Ersek <ler...@redhat.com> wrote on 07/30/2014 12:18:02 PM: > From: Laszlo Ersek <ler...@redhat.com> > To: "Michael S. Tsirkin" <m...@redhat.com>, Stefan Berger/Watson/IBM@IBMUS > Cc: qemu-devel@nongnu.org, Stefan Berger <stef...@linux.vnet.ibm.com> > Date: 07/30/2014 12:18 PM > Subject: Re: [PATCH v2] Add ACPI tables for TPM > > On 07/30/14 18:10, Michael S. Tsirkin wrote: > > On Wed, Jul 30, 2014 at 12:03:46PM -0400, Stefan Berger wrote: > >> Laszlo Ersek <ler...@redhat.com> wrote on 07/30/2014 11:58:52 AM: > > >>> In the short term, probably skip TCPA, or advise users in documentation > >>> not to enable the TPM device when running OVMF. > > > > Hmm but doesn't OVMF rely on all tables being packed without holes > > in memory? > > I remember it did, and if it still does, this breaks unless TCPA is > > the last ... > > Ah, do you mean that the table size field in the ACPI table header of > TCPA would *not* include the TCG event log that is tacked-on? Yes, that > would certainly trip up the parser. It advances by looking at the table > size fields in the headers.
Is this what finally requires another command for the linker/loader to allocate a chunk of memory where the size of the to-be-allocated memory is found at the location of the pointer ? At least SeaBIOS does not seem to need it. Whether TCPA is the last table or not, doesn't seem to matter. Stefan