Attention is currently required from: Nico Huber, Furquan Shaikh, Tim Wawrzynczak. Shelley Chen has posted comments on this change. ( https://review.coreboot.org/c/coreboot/+/58331 )
Change subject: [RFC] pci_mmio_cfg: Move guard around pci_s_* functions to x86 ......................................................................
Patch Set 2:
(1 comment)
File src/arch/x86/include/arch/pci_ops.h:
https://review.coreboot.org/c/coreboot/+/58331/comment/4080491b_7d5c9d90 PS2, Line 7: MMCONF_SUPPORT
I don't think we need a Kconfig option for mmio
ops in general.
Sorry, I'm a little confused by this statement. Just to clarify, do you imagine:
MMCONF_SUPPORT : enable mmio ops (mostly non-x86 platforms) MMCONF_ECAM_SUPPORT: enable mmio ops using ECAM (mostly x86 platforms) NO_MMCONF_SUPPORT: no mmio ops support (revert to io ops)
The reason that we wanted to rename MMCONF was because it seems like there was a lot of confusion around the meaning of MMCONF: 1. does it mean mmio ops? 2. does it only cover ECAM? I think that I actually started out thinking #1, but realized that I was wrong. That was the reason we wanted to introduce the PCI_MMIO_OPS Kconfig in the first place. Our proposal was:
PCI_MMIO_OPS : enable mmio ops (mostly non-x86 platforms) PCI_ECAM: enable mmio ops using ECAM (mostly x86 platforms) PCI_NO_MMIO_OPS: no mmio ops support (revert to io ops)
I am happy to add all this explanation to the doc (CB:57861) if that is useful.