ron minnich has posted comments on this change. ( https://review.coreboot.org/c/coreboot/+/36221 )
Change subject: Add configurable ramstage support for minimal PCI scanning ......................................................................
Patch Set 8:
so far 4 companies have indicated an interest in this CL System76 9elements intel google
no anonymity there.
It's a trivial CL. We did it this way because of the objections to just removing the ramstage in the last 12 months of discussion; we were asked to look at a configurable ramstage. This is the first step on the road. It shows the variable to enable it and the variable to use it.
This has nothing to do with ENV_PAYLOAD_LOADER in my view. It is concerned only with turning off coreboot enumeration of PCI space, in part because coreboot is still 32-bit firmware on x86 and that's causing trouble for some companies; in part because it's not needed anyway for things like Linux: Linux has a much better idea of how it needs configuration space laid out than coreboot does.
We're hoping on subsequent CLs to make further changes in how this is achieved, but the goal was to present a very minimal change that people could evaluate for their needs to get some experience with it. The need is real, however, and the need has yet to be accommodated. This simple change accommodates it.
Nico, I understand your point, but that looks to me like a far larger effort. I'd like to get some evaluation if this is taking us at least the right direction (from a user point of view) before we start making too many far-reaching changes.
as for hidden vs. mandatory: I can find no use of the hidden keyword in any Kconfig; am I missing something? If it is in fact never used, should it be in the parser at all?