Attention is currently required from: Matt DeVillier.
Nico Huber has posted comments on this change. ( https://review.coreboot.org/c/coreboot/+/80598?usp=email )
Change subject: device/pnp_device: Skip init on disabled functions ......................................................................
Patch Set 1:
(2 comments)
Commit Message:
https://review.coreboot.org/c/coreboot/+/80598/comment/129d43e4_747212e9 : PS1, Line 15: in the log / don't cause any errors.
[DEBUG] PNP: 002e.1 60 <- [0x00000000000003f8 - 0x00000000000003ff] size 0x00000008 gran 0x03 io […]
So these 'decode base 0' prints are from `soc/intel/common/block/lpc`. And I think they are right. That code shouldn't be called for disabled devices. But that's an issue local to the Intel code which tries to re-implement resource things. Never really understood that code (it was ported from APL just because it was there, IIRC), but I'm sure a simple ``` if (!dev->enabled) return; ``` in pch_lpc_loop_resources() should be safe.
File src/device/pnp_device.c:
https://review.coreboot.org/c/coreboot/+/80598/comment/e80b0061_1f3a0656 : PS1, Line 403: dev->ops = ops;
where do we generate SSDT entries for disabled devices?
Dunno. That was just an example from the top of my head. It's just that we virtually always have the ops assigned and it might take some thought to figure out if it's right not to have them.
If it fixes something beside the Intel code (see other comment), I would prefer to move the new if/continue below the ops assignment.