[Xen-devel] [PATCH v2] tools/firmware: add ACPI device for Windows laptop/slate mode switch

["Jan Beulich" <JBeulich@xxxxxxxx>]

>>> On 24.03.17 at 11:10, <paul.durrant@xxxxxxxxxx> wrote:
> @@ -406,6 +407,16 @@ static int construct_secondary_tables(struct acpi_ctxt 
> *ctxt,
>          printf("S4 disabled\n");
>      }
>  
> +    if ( config->table_flags & ACPI_HAS_SSDT_CONV )
> +    {
> +        ssdt = ctxt->mem_ops.alloc(ctxt, sizeof(ssdt_conv), 16);
> +        if (!ssdt) return -1;

This being an optional thing anyway, would it perhaps be better to
allow guest boot to continue, just logging a message? Otoh, if we
have no memory here anymore, the guest ought to not get very
far anyway.

> +DefinitionBlock ("SSDT_CONV.aml", "SSDT", 2, "Xen", "HVM", 0)
> +{
> +    Device (CONV) {
> +        Method (_HID, 0x0, NotSerialized) {
> +            Return("ID9001")
> +        }
> +        Name (_CID, "PNP0C60")
> +    }
> +}

And that's it? This device doesn't do anything.

The only other concern I have here is that the abbreviation "conv"
used throughout the patch is sort of ambiguous. I think it means
"convertible" here, but without knowing the context it could easily
be "conventional" or "convenience". Would there be anything
wrong with spelling it out wherever name length limitations don't
require it to be just four characters?

Jan


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