Convert oqs-kem-info.md code points to hex

[WillChilds-Klein]

Please give a brief explanation of the purpose of this pull request.

Minor documentation update to use hexadecimal notation for KEM code points.

Does this PR resolve any issue? If so, please reference it using automatic-closing keywords like "Fixes #123."

Not that I'm aware of.

---

Notes

The initial motivation for this PR was a small documentation change. That change uncovered an unexpected warning message. It appears that the erroneous warning message motivating this commit is due to some vestigial code special-casing x-curves. The nid_ecx_hyybrid kem field was only used to generate the duplicate ecx nid warning message, so we track duplicates in a local set instead. Additionally, we expand the warning message to all curve NIDs, not just x-curves and treat duplicates as fatal because extra curve NIDs should be unique for each KEM.

Duplicate NID detection is tested below:

$ git diff
diff --git a/oqs-template/generate.yml b/oqs-template/generate.yml
index 4865eb7..fa3ddfe 100644
--- a/oqs-template/generate.yml
+++ b/oqs-template/generate.yml
@@ -114,6 +114,8 @@ kems:
           nid: '0x2F90'
         - hybrid_group: "x25519"
           nid: '0x6399'
+        - hybrid_group: "p521"
+          nid: '0x6399'
         - hybrid_group: "p256"
           nid: '0x639A'
       old:

$ LIBOQS_SRC_DIR=$(pwd)/../liboqs python3 oqs-template/generate.py
ERROR: duplicate hybrid NID for kyber768 : 0x6399 in generate.yml. Curve NIDs may only be specified once per KEM.

$ echo $?
1

I'm not sure what's up with the CI; I don't appear to have permissions to view the build errors in CircleCI. I ran the tests locally on a linux box:

$ cmake --build build && ( cd build; ctest -V )
...
100% tests passed, 0 tests failed out of 5

Total Test time (real) =  29.85 sec

[baentsch]

This is a change to an automatically generated file and as such won't persist. Please change at the master file (oqs-template/generate.yml), re-generate (python3 oqs-template/generate.py) and re-run all tests.

[WillChilds-Klein]

I'm not able to see the CircleCI results. I'll try to re-run the tests to reproduce the CI failures locally.

Also, I've noticed that both before and after this PR's changeset, the generate.py script emits (what appears to be) an erroneous warning. I'll try to update the script to fix this, but I may need some guidance on that warning's original intent.

[baentsch]

Thanks for the update: This is now perfect to retain.

some guidance on that warning's original intent

This is basically a question/reminder to @bhess to re-evaluate the original ECX logic (via an extra-YML file) he originally introduced. IMO it could be removed as #177 completely changed the build logic to permit more than a single X-hybrid for the same PQ alg.

[baentsch]

@WillChilds-Klein Thanks again for the contribution! CI has run successfully so we'll merge this tomorrow. As this is more than a trivial change, please consider amending this PR before merge by adding yourself to the Contributors list in the README.