Re: [PATCH 06/10] acpi: PVH guests need _E02 method

[Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>]

On Mon, Nov 07, 2016 at 11:08:26AM -0500, Boris Ostrovsky wrote:
> On 11/07/2016 10:45 AM, Konrad Rzeszutek Wilk wrote:
> > On Sun, Nov 06, 2016 at 04:42:39PM -0500, Boris Ostrovsky wrote:
> >> This is the method that will get invoked on an SCI.
> >>
> >> Signed-off-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
> >
> > .. thought I wonder - do you want to also make the writes to
> > _GPE.DPT1 and _GPE.DPT2 be optional? Looking at the mk_dsdt.c
> > those are defined at 0xb044 and are done in:
> >  /**** PCI0 start ****/ 
> >
> > aka:
> >
> >   stmt("OperationRegion", "DG1, SystemIO, 0xb044, 0x04");                     
> > 415                                                                                 
> > 416     push_block("Field", "DG1, ByteAcc, NoLock, Preserve");                      
> > 417     indent(); printf("DPT1, 8, DPT2, 8\n");                                     
> > 418     pop_block();                                     
> >
> > Which means if you run 'acpixtract' and 'acpixecute' to play with
> > the DSDTs you will get nasty warnings about the DPT1 and DPT2 not being
> > defined.
> >
> > I am actually not sure what the kernel will do with undefind DPT1 and DPT2?
> > I would have thought the iasl compiler would throw an nasty fit?
> 
> Are these ACPI debug ports?

Yes.
> 
> I don't see any warnings from acpixtract, unless it's a particular AML
> that would cause it.

Could you send the .dsl files over ?
> 
> -boris
> 

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel