[XL] Change in coreboot[main]: soc/intel/ptl: Add GPIOs for Panther Lake SOC

[Subrata Banik (Code Review)]

Attention is currently required from: Anil Kumar K, Bora Guvendik, Cliff Huang, Elyes Haouas, Jamie Ryu, Jérémy Compostella, Kapil Porwal, Paul Menzel, Pranava Y N, Ravishankar Sarawadi, Saurabh Mishra, Wonkyu Kim.

Subrata Banik has posted comments on this change by Cliff Huang. ( https://review.coreboot.org/c/coreboot/+/83789?usp=email )

Change subject: soc/intel/ptl: Add GPIOs for Panther Lake SOC
......................................................................

Patch Set 67:

(2 comments)

File src/soc/intel/pantherlake/acpi/gpio.asl:

https://review.coreboot.org/c/coreboot/+/83789/comment/8b614b41_6378aa05?usp... :
PS49, Line 436: Name (JTAG, Package (0x02)
> I have asked our kernel pinctrl owner and we cannot skip PADs within a group, so if any of PADs needs to be exposed, the entire group needs to be added. I didn't get the reason, though. To be conservative and consistent, I follow what we did for the past SOCs and LNL (first with this new schema) for now as it requires more validation and finding out the corner cases if we drop any of those non GPP_[group]_[num] group. Will it be okay if we add this in the TODO list for further revisiting?

sure, please add that as TODO. I'm more interested to understand this whole design as we didn't get involved into LNL design

File src/soc/intel/pantherlake/include/soc/gpio_soc_defs.h:

https://review.coreboot.org/c/coreboot/+/83789/comment/b029f0fb_0398ec19?usp... :
PS49, Line 31: 1
> Subrata,
> Currently, we use _UID to describe their community number. 
> from PTL, we have:
> INTC10BC:00/uid:0
> INTC10BC:01/uid:1
> INTC10BC:02/uid:3
> INTC10BC:03/uid:4
> INTC10BC:04/uid:5
> I also checked BIOS for other four SOCs, including LNL, that use this new GPIO schema, and they use _UID for the community number. LNL CB also does the same.
> Name (_UID, 0)          // numeric value, GPIO Community
> Name (_UID, 1)          // numeric value, GPIO Community
> Name (_UID, 3)          // numeric value, GPIO Community
> Name (_UID, 4)          // numeric value, GPIO Community
> Name (_UID, 5)          // numeric value, GPIO Community
>
> As their instance already with :0[n] for index number, should we keep this consistent among these SOCs that uses this new schema? Also, I was told that there is no value/benefit/security implications for hiding their actual community number. At the same time, it is not harmful to use the sequential number for their _UID as you suggest.

although I don't think UID is actually used for anything here, because HID and other details are so unique among each GPIO COMM. But lets not make things complicated. marking this section resolved