Na stateful macro

[ashman-p]

The Stateful Hash-Based Signature feature, when enabled, does signature verification by default. In this mode it uses the OQS_SIG API and only populates the 'verify' vector function. In the experimental full mode (verify, key and signature generation) use of OQS_SIG_STFL is needed.
Also, this PR includes code refactored to use macros to generate repetitive functions for HBS variants.

• Does this PR change the input/output behaviour of a cryptographic algorithm (i.e., does it change known answer test values)? (If so, a version bump will be required from x.y.z to x.(y+1).0.)
• Does this PR change the list of algorithms available -- either adding, removing, or renaming? Does this PR otherwise change an API? (If so, PRs in fully supported downstream projects dependent on these, i.e., oqs-provider and OQS-OpenSSH will also need to be ready for review and merge by the time this is merged.)

[baentsch]

Excellent -- Thanks very much! This reduces the LOC count markedly and makes this code much more maintainable. Only nit: Do I understand it right that only a single alg gets tested? No longer too much of an issue as all of them now share the same code base, though.

[ashman-p]

Excellent -- Thanks very much! This reduces the LOC count markedly and makes this code much more maintainable. Only nit: Do I understand it right that only a single alg gets tested? No longer too much of an issue as all of them now share the same code base, though.

Thanks @baentsch.
I may have missed it but, what do you mean "single algorithm gets tested"?
My expectation is that all the enabled HBS variants would be tested.
tests/KATs/sig_stfl/kats.json

[baentsch]

I may have missed it but, what do you mean "single algorithm gets tested"?
My expectation is that all the enabled HBS variants would be tested.
tests/KATs/sig_stfl/kats.json

Oops -- I may have "jumped to conclusions" (in case you know the film Office Space :): I only looked at the file diffs and saw a single test listed -- but on second view, it's not impacted, so indeed, the full alg suite may be subject to testing. I tried to validate that, but I'm hard pressed to find CI runs not skipping the algs: Could you please help me by pointing to them? I'd be particularly interested to see tests doing the verification only (using OQS_SIG) and others doing the full monty. Finally, one more suggestion to make the main "sig-stateful" code base more easily readable: Comments re-stating the macro name on closing "#endif" instructions may be helpful (at least to me) understanding which code should be running under which config.

[ashman-p]

The example that makes the most sense is the KATs used for verification tests.
e.g. python3 tests/test_kat.py

[baentsch]

The example that makes the most sense is the KATs used for verification tests. e.g. python3 tests/test_kat.py

Thanks for this pointer to the test code. Open question remaining: Which CI runs do not skip these tests? Which CI runs trigger only verification, which ones keygen+sign+verify?