Nico Huber has posted comments on this change. ( https://review.coreboot.org/c/coreboot/+/36493 )
Change subject: [WIP] Documentation: PCI resource allocations ......................................................................
Patch Set 1:
(2 comments)
https://review.coreboot.org/c/coreboot/+/36493/1/Documentation/device/pci_re... File Documentation/device/pci_resource.md:
https://review.coreboot.org/c/coreboot/+/36493/1/Documentation/device/pci_re... PS1, Line 13: 1. Within PCI drivers, *.init* is the only callback where resource assignment may be required. And `.final`?
https://review.coreboot.org/c/coreboot/+/36493/1/Documentation/device/pci_re... PS1, Line 25: Maybe
1. Platforms with blobbed chipset initialisation that hook into boot states and expect resources assigned are borked.
I guess the best thing we can do with chipset devices on FSP platforms is to allocate resources and unallocate like described below. Alternatively, we could go trough intensive vendor-code reviews or testing to figure out what happens when we violate FSP.