Attention is currently required from: Angel Pons, Arthur Heymans, Christian Walter, David Milosevic, Felix Singer, Lean Sheng Tan, Martin L Roth, Maximilian Brune, Nico Huber.
Julius Werner has posted comments on this change. ( https://review.coreboot.org/c/coreboot/+/74798?usp=email )
Change subject: arch/arm64: Add EL1/EL2/EL3 support for arm64 ......................................................................
Patch Set 9:
(2 comments)
Commit Message:
https://review.coreboot.org/c/coreboot/+/74798/comment/222b419e_0bb54c66 : PS5, Line 11: one boots into TF-A first and drops into EL2 for coreboot afterwards.
So, if we shut ourselves off to even make it possible to enable major Arm SoC platforms today, and don’t even have the patience to start the journey to build the coreboot ecosystem for arm
I don't want to add much more to this discussion than I already have, but please don't present this as if you were only just now trying to bring coreboot to the Arm world. We have been supporting real, physical Arm hardware for over a decade, including commercial products numbering millions of devices. In fact some of the most FOSS / blobless platforms in all of coreboot are Arm devices.
I appreciate that your particular platform for the vendor you're trying to add is stuck on this problem, and that's a fair concern since we generally want to try to support as many platforms as possible, but it's also not more than that. And as far as I understand the problem with that platform is also less of a "we really need this SoC vendor secret sauce to initialize the platform which they are absolutely adamant they can't open-source or let us reimplement", as it is often with Intel (at least as far as they present it), for you it's more just a matter of "we only have existing closed source code for this right now and we don't have the time/manpower to port our silicon init to coreboot at the moment".
Patchset:
PS9:
Maybe to differentiate this, we can go further than just making it a Kconfig option and actually add a new makefile target. That would make it even more plain that this isn't the full coreboot build.
I don't really see what that would give us other than a big maintenance nightmare to be honest. I think we should decouple the technical implementation details from the philosophical question here. From the technical side, I think the patch as proposed is pretty much the right way to implement this if we want to implement it at all.