Re: [PATCH 0/2] PCI/AER: Consistently use _OSC to determine who owns AER

[Sinan Kaya]

On 11/19/2018 6:49 PM, Alex_Gagniuc@xxxxxxxxxxxx wrote:
> On 11/19/2018 02:33 PM, Sinan Kaya wrote:
> > However; table assumes governance about for which entities firmware first
> > should be enabled. There is no cross reference to _OSC or permission
> > negotiation like _OST.
>
> Well, from an OSPM perspective, is FFS something that can be enabled or
> disabled? FFS seems to be static to OSPM, which would change the sort of
> assumptions we can reasonably make here.

IMO, it can be enabled/disabled in BIOS. I have seen this implementation before.
If the trigger is the presence of a statically compiled ACPI HEST table (as the
current code does); presence of FFS would be static from OSPM perspective.
BIOS could patch this table or hide it during boot.

If FFS were to be negotiated via _OSC as indirectly implied in this series, then
same BIOS could patch the ACPI table to return different values for the _OSC
return.

> > > > As I said in my previous email, the right place to talk about this is UEFI
> > > > forum.
> > >
> > > The way I would present the problem to he spec writers is that, although
> > > the spec appears to be consistent, we've seen firmware vendors that made
> > > the wrong assumptions about HEST/_OSC. Instead of describing AER
> > > ownership with _OSC, they attempted to do it with HEST. So we should add
> > > an implementation note, or clarification about this.
> >
> > I agree.
>
> Cool. While the UEFI Secret Society debates, can we figure out if/how
> [patch 1/2] breaks those systems, or is it only [patch 2/2] of this
> series that we suspect?

I went back and looked at both patches. Both of them are removing references to
HEST table. I think both patches are impacted by this discussion.

> Alex