coreboot (rom stitching) not reproducible #1633
Milestone
Comments
|
Only insight of the moment is that we hardcode flags passed down to coreboot when rom is built from modules/coreboot: |
|
Diffoscope doesn't help :( |
|
cbfstool above fails. Needs investigation :/ |
|
But points to romstage being different which is still a start |
|
@tlaurion TL;DR - CI seems to lack the If you pick apart the coreboot stages, that appears to be the only difference (with #1630). E.g.: There is also still a difference in Busybox affecting the initrd also, have not looked into that. One option could be to get CircleCI to sync the tags if that's possible. Or we could stop setting a version based on tags - Heads isn't tagging any releases anyway, but I do tag downstream PureBoot releases and might still want the tag in the version downstream. There could be a middle ground where we only use the tag if it's pointing to this commit (never tag+commit height), but that doesn't seem to be something git-describe will readily do, we'd have to do it ourselves. |
1 task
|
Let's revisit once #1630 is merged. |
|
Deleting the v5.0.1 tag from our local repos and the tlaurion fork fixed these issues - @tlaurion I think we can close this |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
All coreboot forks currently used do not stich the final rom reproducibily.
One example:
Originally posted by @tlaurion in #1630 (comment)
The text was updated successfully, but these errors were encountered: