Subrata Banik has posted comments on this change. ( https://review.coreboot.org/c/coreboot/+/33143 )
Change subject: Rampayload: Able to build coreboot without ramstage ......................................................................
Patch Set 7:
(1 comment)
https://review.coreboot.org/#/c/33143/3/toolchain.inc File toolchain.inc:
https://review.coreboot.org/#/c/33143/3/toolchain.inc@50 PS3, Line 50: ifeq ($(CONFIG_RAMPAYLOAD),y)
Today ARCH_RAMSTAGE_X86 is the way to tell if we like to build ramstage or not. At least thats how coreboot soc is structured now.
That's not true. That's what you're trying to turn it into and I'm trying to argue to please not do that. ;)
Today ARCH_RAMSTAGE_X86 tells what architecture the ramstage is and nothing else. There is no way to not build a ramstage today.
You got the right point, here the new requirement is to build coreboot without RAMSTAGE like verstage/postcar :)
Thats the reason we are trying to remove root dependency from ramstage in coreboot code today.
There is a way to not build two other stages (verstage and postcar), but for both of those the option that controls whether the stage is built is completely orthogonal to the option that controls with architecture it is (and the latter is usually still set even if the former is not... e.g. ARCH_VERSTAGE_X86 is still set even if SEPERATE_VERSTAGE=n).
So you are telling to have ENABLE_STAGE_RAMSTAGE=n and ARCH_RAMSTAGE_X86 can still be yes ? is that what you wish to conclude then i can make code changes accordingly.
I guess Aaron has recommended to have ENABLE_STAGE_<bootblock/romstage/ramsatge> selection from ARCH_ config. And looks like you don't wnat that to be interdependent ?