Comment 20 for bug 2004618

Revision history for this message
dann frazier (dannf) wrote :

Thanks for the deeper investigation Mustafa!

fwiw, I believe I came up with the descriptor ordering of secure+enrolled > secure > non-secure just by cargo culting from another package - probably Fedora's. I'm certainly open to rethinking that ordering for 23.10, but I'd prefer to not change the behaviors of existing stable releases. That feels like digging ourselves into a deeper hole instead of just climbing out of this one.

If I understand your findings correctly, it seems like we could restore the original behavior simply by making the first pattern match the 4M images as well, i.e.:

- r".*OVMF_CODE\.fd", # RHEL
+ r".*OVMF_CODE(_4M)?\.fd", # RHEL

That seems like the smallest/safest change we could make for focal to restore existing behavior. If we do that for just focal, it would seem to restore the status quo, where focal behaves differently than >= jammy. IMO, that'd be enough to resolve this issue - what do you think?

Let's also continue conversation about correcting the behavior for future release - and upstream - ideally in a new LP bug.