It would be largely unrelated to the project's actual source control,
but I could manufacture one, given a day or two, if you want. It
might even be from HEAD and tested if I can find the time. I've been
using my personal account for interaction with the mailing list until
there's an SOP for contributing back from work.
This patch could probably use a little outside review, honestly.
Systemd ordering cycles have mostly non-deterministic failure
behavior, as part of the spec. If someone out there is performing TPM
operations in a systemd-based initramfs (maybe Dracut) and could check
they don't end up with any cycles in their graph it would be nice.
This has only been tested with a manual init system and no ABRMD in
the initramfs in my case.