I think we still need to have a difference between hacky vendor stuff
and normal coreboot code. For example, the Eltan mboot stuff is
something we didn't really want to have in coreboot in that form, and
so they kinda put it in vendorcode as a compromise. We should make
sure it remains clear that that code isn't "proper" coreboot code and
didn't go through the same level of review.
It might have started that way, but I don't think that's an accurate portrayal of eltan's work at this point:
The eltan code uses vboot for the cryptographic primitives these days and as far as I can see, extends it for measured boot - which vboot itself doesn't do, ever.
Also, regarding your point on gerrit (collecting arguments in this thread) that we don't have duplicate things, look no further than graphics init:
- src/device/oprom/realmode
- src/device/oprom/yabel
- src/drivers/intel/gma
- FSP can do the graphics init, too (and report it in cbtables)
(I didn't count the native ARM graphics init routines because we don't ship alternatives for those)
Patrick
Google Germany GmbH, ABC-Str. 19, 20354 Hamburg
Registergericht und -nummer: Hamburg, HRB 86891, Sitz der Gesellschaft: Hamburg
Geschäftsführer: Paul Manicle, Halimah DeLaine Prado