Re: [PATCH 1/3] acpi/processor: fix evaluating _PDC method when running as Xen dom0

[Roger Pau Monné]

On Fri, Dec 02, 2022 at 06:06:26PM +0100, Rafael J. Wysocki wrote:
> On Fri, Dec 2, 2022 at 5:37 PM Roger Pau Monné <roger.pau@xxxxxxxxxx> wrote:
> >
> > On Fri, Dec 02, 2022 at 08:17:56AM -0800, Dave Hansen wrote:
> > > On 12/2/22 04:24, Roger Pau Monné wrote:
> > > > On the implementation side, is the proposed approach acceptable?
> > > > Mostly asking because it adds Xen conditionals to otherwise generic
> > > > ACPI code.
> > >
> > > That's a good Rafael question.
> 
> Sorry for joining late, but first off _PDC has been deprecated since
> ACPI 3.0 (2004) and it is not even present in ACPI 6.5 any more.
> 
> It follows from your description that _PDC is still used in the field,
> though, after 18 years of deprecation.  Who uses it, if I may know?

I saw this issue on a Sapphire Rapids SDP from Intel, but I would
think there are other platforms affected.

> > > But, how do other places in the ACPI code handle things like this?
> >
> > Hm, I don't know of other places in the Xen case, the only resource
> > in ACPI AML tables managed by Xen are Processor objects/devices AFAIK.
> > The rest of devices are fully managed by the dom0 guest.
> >
> > I think such special handling is very specific to Xen, but maybe I'm
> > wrong and there are similar existing cases in ACPI code already.
> >
> > We could add some kind of hook (iow: a function pointer in some struct
> > that could be filled on a implementation basis?) but I didn't want
> > overengineering this if adding a conditional was deemed OK.
> 
> What _PDC capabilities specifically do you need to pass to the
> firmware for things to work correctly?

I'm not sure what capabilities do I need to pass explicitly to _PSD,
but the call to _PDC as done by Linux currently changes the reported
_PSD Coordination Field (vs not doing the call).

Thanks, Roger.