> From: Hanjun Guo [mailto:hanjun.guo@linaro.org]

> Sent: Sunday, September 22, 2013 11:27 AM

>
> On 2013-9-19 10:37, Moore, Robert wrote:
> > While we are at it, here is the *complete* list of ACPICA interfaces that are
> > meaningless on a hardware-reduced platform. If we are going to dynamically
> > disable some of these interfaces, we will need to disable all of them -- for
> > completeness. So, this is actually not a trivial change.
>
> Yes, you are right.
> Actually, before I send this patch out, I searched all the ACPICA API on
> a hardware-reduced platform as you did, I found some of them will finally call
> acpi_hw_register_read/write(), so some of the ACPICA API do not need to change.
>

I agreed.

My idea is to keep the call path implemented using non-HWREDUCED ACPI silicon not-changed, but changing the code inside of a function implemented with the non-HWREDUCED logic to contain a HWREDUCED logic.
This might be more reasonable.
So IMO, it is not a good idea to reduce the APIs, but need to re-implement the APIs using HWREDUCED logic.
And the non-HWREDUCED code inside of an API should be marked with ACPI_HARDWARE_REDUCED macro rather than the API itself be marked out.
If the APIs (aka., ACPI functionalities) are kept un-modified, the device drivers can be kept un-modified on HWREDUCED platforms.