sntrup761: replace PQClean code with public domain OpenSSH code

[bbbrumley]

PQClean does not support sntrup761 anymore, since it did not proceed in the NIST standardization process. The liboqs code for sntrup761 needs updating, so it makes sense to ditch the PQClean code and go with "something else."

sntrup761 remains relevant for many reasons, IMO the most significant being its (standardized) use in SSH key exchange. (OK, maybe it's a draft rn, but it's deployed, keys are out there, and they're only growing, the details are historical footnotes.)

In this case, OpenSSH maintains a public domain single-file implementation. (Big props upstream! Thank you for your contribution.)

So this PR removes the PQClean code for sntrup761 and shims in that OpenSSH code.

Fixes #2324.

This is not a new feature.

• [N] 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.)
• [N] 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 will also need to be ready for review and merge by the time this is merged. Also, make sure to update the list of algorithms in the continuous benchmarking files: .github/workflows/kem-bench.yml and sig-bench.yml)

I think this change is transparent to linking applications. But idk liboqs that well, just my best guess.

No AI.

[bbbrumley]

I don't really know what that DCO check means, I added the string to the commit message, idk if that was correct 🤷

[coveralls]

coverage: 82.799% (-0.2%) from 82.96%
when pulling eb5ccc2 on bbbrumley:bbb_sntrup
into c58e93e on open-quantum-safe:main.

[bbbrumley]

0764558 should solve some of the CI failures. Will take a look at what's still failing after the CI runs.

[bbbrumley]

... and @dstebila thank you for the push

[bbbrumley]

Thanks for your patience while I stumble through your CI!

[baentsch]

Thanks for your patience while I stumble through your CI!

Thanks for your willingness to do that. Just triggered a full run.

[baentsch]

One quick initial comment on the PR: Could you please also amend the ntruprime files in the folder https://github.com/open-quantum-safe/liboqs/tree/main/docs/algorithms/kem as part of the PR? Those files had been auto-generated while importing from PQClean -- but that obviously now no longer works.... If at all possible, please retain the format but replace everything that no longer applies (optimizations, license info, etc.). Thanks in advance!

[bbbrumley]

Could you please also amend the ntruprime files in the folder

@baentsch absolutely, I saw those but wasn't sure exactly what the procedure is -- does the md get generated from the yml or are these independent files? Just so I know what to edit.

[dstebila]

Could you please also amend the ntruprime files in the folder

@baentsch absolutely, I saw those but wasn't sure exactly what the procedure is -- does the md get generated from the yml or are these independent files? Just so I know what to edit.

Yes, the Markdown gets generated from the YAML file by the copy_from_upstream script.

[dstebila]

Thanks for your patience while I stumble through your CI!

Let us know if we can help with that. Although my own expertise with the Windows builds (which are the ones currently failing) is low.

[bbbrumley]

Let us know if we can help with that

will do, making progress. (I think I'm almost there, also no direct windows experience since early 2000s ...)

[bbbrumley]

Yes, the Markdown gets generated from the YAML file by the copy_from_upstream script

Ah I see, I think I got it now, changes from running that script like the CI does are in 89eeb44

Also rebased on current main

[bbbrumley]

If you need a GH handle for sntrup761 going forward, I added myself in 6edba19