The status of ACPI S4 support in OVMF

The status of ACPI S4 support in OVMF

[Gary Lin]

Hi,

Years ago, the bug(*) for PEI variable driver was filed against OVMF
and it used to be one reason to block S4 support in OVMF. The bug is
fixed now, and there are several MemTypeInfo fixes merged into OVMF.

I'm curious about the current status of S4 support in OVMF. Is it still
considered as unsupported due to some other bugs?

Thanks,

Gary Lin

(*) https://bugzilla.tianocore.org/show_bug.cgi?id=386

Re: [edk2-devel] The status of ACPI S4 support in OVMF

[Laszlo Ersek]

On 04/19/21 11:54, Gary Lin via groups.io wrote:
> Hi,
>
> Years ago, the bug(*) for PEI variable driver was filed against OVMF
> and it used to be one reason to block S4 support in OVMF. The bug is
> fixed now, and there are several MemTypeInfo fixes merged into OVMF.
>
> I'm curious about the current status of S4 support in OVMF. Is it
> still considered as unsupported due to some other bugs?
>
> Thanks,
>
> Gary Lin
>
> (*) https://bugzilla.tianocore.org/show_bug.cgi?id=386

I recall the last related discussion from February:

* [edk2-devel] [PATCH 0/2] Improve hibernation safety

  https://edk2.groups.io/g/devel/message/71790
  https://listman.redhat.com/archives/edk2-devel-archive/2021-February/msg00842.html
  http://mid.mail-archive.com/20210218200953.20943-1-graf@amazon.com

No patches needed to be merged (Alex was missing some commits that had
been upstreamed by then).

My understanding is that S4 should work at this point, although I've
never really tested it myself.

Thanks
Laszlo

Re: [edk2-devel] The status of ACPI S4 support in OVMF

[Gary Lin]

On Mon, Apr 19, 2021 at 02:32:25PM +0200, Laszlo Ersek wrote:
> On 04/19/21 11:54, Gary Lin via groups.io wrote:
> > Hi,
> >
> > Years ago, the bug(*) for PEI variable driver was filed against OVMF
> > and it used to be one reason to block S4 support in OVMF. The bug is
> > fixed now, and there are several MemTypeInfo fixes merged into OVMF.
> >
> > I'm curious about the current status of S4 support in OVMF. Is it
> > still considered as unsupported due to some other bugs?
> >
> > Thanks,
> >
> > Gary Lin
> >
> > (*) https://bugzilla.tianocore.org/show_bug.cgi?id=386
> 
> I recall the last related discussion from February:
> 
> * [edk2-devel] [PATCH 0/2] Improve hibernation safety
> 
>   https://edk2.groups.io/g/devel/message/71790
>   https://listman.redhat.com/archives/edk2-devel-archive/2021-February/msg00842.html
>   http://mid.mail-archive.com/20210218200953.20943-1-graf@amazon.com
> 
> No patches needed to be merged (Alex was missing some commits that had
> been upstreamed by then).
> 
> My understanding is that S4 should work at this point, although I've
> never really tested it myself.
> 
Thanks for the information. I was reviewing the libvirt descriptor for
ovmf and wondering if it's ready for the "acpi-s4" tag. Maybe it's a
good to give it a try.

Thanks,

Gary Lin