Do we need to do more robust validation of programs and links?

The current code validates that that program and link IDs have been saved in the state CRDs and that bpfman knows about them.

The previous code would also check that all the data in the program+link matched what is intended. A better option would be to make load and attach operations idempotent (see: https://github.com/bpfman/bpfman-operator/issues/395).

It would currently be a bug if the program or link ID didn't match what we expected, so that shouldn't happen. However, if we don't do #395 it wouldn't hurt to make the check more robust just in case. If we do #395, bpfman would essentially do this for us so it wouldn't be needed.

Start by reading the load and attach operations that validate program and link IDs against state CRDs, then review the related idempotency proposal in issue #395. Determine whether #395 will make the extra validation unnecessary; done should be a decided approach with checks covering the intended program and link data if validation remains required.

Written by the indexing model from the issue text.