Re: Regression from efi: call get_event_log before ExitBootServices

On Tue, Mar 13, 2018 at 8:59 AM Ard Biesheuvel <[email protected]>
wrote:

> On 13 March 2018 at 07:47, Hans de Goede <[email protected]> wrote:
> > Hi,
> >
> >
> > On 12-03-18 20:55, Thiebaud Weksteen wrote:
> >>
> ...
> >>
> >> Hans, you said you configured the tablet to use the 32-bit version of
grub
> >> instead
> >> of 64. Why's that?
> >
> >
> > Because this tablet, like (almost?) all Bay Trail hardware has a 32 bit
> > UEFI,
> > even though the processor is 64 bit capable (AFAIK 64 bit Windows
drivers
> > were
> > not ready in time so all Bay Trail devices shipped with a 32 bit
Windows).
> >
> > So this is running a 32 bit grub which boots a 64 bit kernel.
> >
> >> Jeremy, could you confirm if you are building the kernel in 64bit
mode? Is
> >> your Android install working? (that is, what happens if you boot
> >> Boot0000)?
> >
> >
> > AFAIK the kernel on Jeremy's tablet (which I initially installed) is 64
bit.
> >
> > Could the problem perhaps be that the new code for the TPM event-log is
> > missing some handling to deal with running on a 32 bit firmware? I know
the
> > rest of the kernel has special code to deal with this.
> >

Yes, that was my guess as well.

> That is a very good point, and I missed that this is a 64-bit kernel
> running on 32-bit UEFI.

> The TPM code does use efi_call_proto() directly, and now I am thinking
> it is perhaps the allocate_pages() call that simply only initializes
> the low 32-bits of log_tbl.

That make sense. Would you know what happens to the arguments of the
function in this case? (I'm thinking &log_location ?)
Would it be safer to skip the code completely on EFI_MIXED systems?

> Jeremy, could you please try initializing tcg2_protocol and log_tbl to
> NULL at the start of the function?