My goal was pretty simple: Kill the UEFI HOBs, and the FSP UPD, and put something better in their place. coreboot tables could easily replace HOBs, save that intel will never accept that; but I don't see coreboot tables replacing UPD.
[one might argue that what Intel will accept matters a lot less than it did 5 years ago, and I would agree. So maybe we can worry less about what Intel will accept, but still ... :-) ] I like self describing data as it avoids that mess that we are in with UPD today, where you can end up with problems if the compilers you use for FSP and (e.g.) coreboot don't agree totally on how to lay out data structures. UPD are also a major pain for non-C firmware, such as oreboot. So I'd like a data format that is not defined by a compiler or language. But maybe I'm the only person who wants that :-) On Tue, Apr 12, 2022 at 11:45 AM Peter Stuge <[email protected]> wrote: > > ron minnich wrote: > > peter, you are right about CBOR, and that says to me it does not > > really meet the original goal of self-describing data. > > Hm, whose goal is that? > > Anyway, using some data structure serialized in CBOR requires > defining the structure somewhere. Using coreboot tables requires > definitions too, they are currently defined in coreboot, > standardizing coreboot tables would probably see them move to a > repo of their own. > > > > But coreboot tables, at least in my understanding, is also not > > self-describing. > > I don't know? What do you mean by self-describing actually? > > > > Do you have some thoughts on a good format that is self-describing? > > So what's the expectation there; what does a self-describing format > enable or need to enable? And what's the complexity tradeoff involved? > > As Arthur pointed out, coreboot tables have the quite significant > advantage of being very very simple to read and write. > > > I think this is still interesting to pursue: > > > > > So if the idea is to create a payload handoff format that can be > > > > shared and used by multiple different firmware packages, do you have > > > > a better option? > > > > > > I'd ask what other boot firmware is missing from coreboot tables for > > > them to be universally acceptable. > > Martin wrote that the goal is to create a handoff format that can be > shared and I'm asking what coreboot tables are missing to serve others, > because I think we have a really good (simple) technical solution there. > > > //Peter > _______________________________________________ > coreboot mailing list -- [email protected] > To unsubscribe send an email to [email protected] _______________________________________________ coreboot mailing list -- [email protected] To unsubscribe send an email to [email protected]
