Patch Set 1:
Does IASL or FWTS complain about that?

Complain about the original or the proposed patch? The original would only exhibit an issue (or at least questionable behavior) for a device that lacks properties, and so far I haven't come across an existing example in coreboot (they all have properties or properties+children).

I noticed the problem when trying to port the various ipu_mainboard.asl and ipu_endpoints.asl files in coreboot to acpigen (e.g., under mainboard/google/poppy/variants/baseboard/include/baseboard/acpi/ and nearby directories). The ipu_mainboard.asl specifies a _DSD that only has children--when I trying porting this to acpigen and diff'ed the result I noticed the extra empty properties UUID/package. With this change these is no difference after the port. These asl files create ACPI devices with tables that have 1) only properties, 2) only children, and 3) both properties and children, so my port to acpigen happens to exercise the 3 main use cases of this function (the 4th case is neither properties nor children, but I wasn't sure whether to also change the handling for that or just expect that a caller shouldn't call the method if there's no properties/children to write).

View Change

To view, visit change 31161. To unsubscribe, or for help writing mail filters, visit settings.