Attention is currently required from: Martin Roth, Julius Werner, Felix Held. Piotr Król has posted comments on this change. ( https://review.coreboot.org/c/coreboot/+/52576 )
Change subject: Documentation: Add suggestion to document flag days ......................................................................
Patch Set 1:
(1 comment)
File Documentation/getting_started/gerrit_guidelines.md:
https://review.coreboot.org/c/coreboot/+/52576/comment/7f4095a0_803164ef PS1, Line 332: requires a change in all mainboards using it) needs to be documented.
I think this is your real problem tbh, and it's one of your own making. Why isn't your code upstream? Google had a downstream fork of coreboot for years, and we suffered immensely under the constant back and forward porting effort just like you describe here, until we eventually learned our lesson forced everything to be landed upstream first before we would ship it. I don't think that this is any "10k+ employees vs small company" difference, the effort is still the same, and the answer is just that working directly upstream is the only way to prevent all that churn. (Google may have 10k+ employees but they're not all working on coreboot, btw. It's more like 20 or so.)
We can try to upstream things, but some customers requirements are bizarre. To count few: 1. Sign of life - certain string showing as early as possible on platform boot 2. sortbootorder as payload - would payload supporting only one vendor will be merged to coreboot? 3. Custom SeaBIOS features which SeaBIOS upstream won't accept - this is general pattern related to other payloads (memtest, ipxe), including customization and configuration of those payloads 4. Experimental uCode inclusion 5. ACPI difference, mentioned in email thread, required for OS compatibility 6. Runtime configuration 7. CI/CD related stuff etc.
I don't think upstream is the place for any customer requirement.
As you said before there should be fork on certain point to provide stability and working always with upstream and doing "stable" releases on upstream seem to be against that argument.
There are also features which simply pending upstream, there are things which are under review. We trying to be as close to upstream as possible using resources we have.
Difference in budget between small player and huge player is significant. Maybe we should be clear about saying that coreboot is too expensive for small OEM/ODM, they should go to IBVs.
-
Piotr Król (Code Review)