| Commit-Queue | +1 |
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
I forget what was the WPT conclusion. Is there anything here that isn't tested by WPTs? (What's the story with the WPTs?)
case kNoParamsKeyTag: {Why isn't `kKemXwingKeyTag` just `kNoParamsKeyTag`? What did we end up doing for ML-KEM and ML-DSA? I don't see them here either, so I assume we sent them down some generic path?
tag = kKemXwingKeyTag;Ditto from other file.
Although I'm very confused because the `default` case has a `DCHECK` that it seems ML-KEM and ML-DSA would fail. Are ML-KEM and ML-DSA support incomplete? How does this code work?
// Generate a KemXwing key pair.I've never heard the name KemXwing before. Not just "X-Wing"?
Edit: Oh, I see elsewhere you use the string "KEM-XWING". Setting aside that not being a standard name for the algorithm, if the algorithm is to be called KEM-XWING, it seems you should use that in the comment.
// Check that the keys have the same RawSeed representation.Nit: This is also odd to see in prose. "raw seed"?
// Check that the keys have the same RawPublic representation."raw public"?
kKemXwingKeyTag = 9,Ditto. But also why allocate this at all instead of just using the no-params tag?
kKemXwingTag = 26,Ditto re what the name of this algorithm is.
{"KEM-XWING", 9, kWebCryptoAlgorithmIdKemXwing},...oh, is this where KEM-XWING came from? Do you have a link to the spec this is implementing? draft-connolly-cfrg-xwing-kem calls it "X-Wing" while draft-irtf-cfrg-concrete-hybrid-kems calls it "MLKEM768-X25519".
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
I forget what was the WPT conclusion. Is there anything here that isn't tested by WPTs? (What's the story with the WPTs?)
we don't have WPT's right now; I figured I'd add them and remove these tests (or remove these tests if someone else adds them) once the WPTs are added, which I believe we said should happen after X-wing makes it into the modern crypto spec.
Why isn't `kKemXwingKeyTag` just `kNoParamsKeyTag`? What did we end up doing for ML-KEM and ML-DSA? I don't see them here either, so I assume we sent them down some generic path?
resolving as a part of crbug.com/512509718
Ditto from other file.
Although I'm very confused because the `default` case has a `DCHECK` that it seems ML-KEM and ML-DSA would fail. Are ML-KEM and ML-DSA support incomplete? How does this code work?
resolving in crbug.com/512509718
I've never heard the name KemXwing before. Not just "X-Wing"?
Edit: Oh, I see elsewhere you use the string "KEM-XWING". Setting aside that not being a standard name for the algorithm, if the algorithm is to be called KEM-XWING, it seems you should use that in the comment.
changed name to `MLKEM768-X25519` as per discussion outside of CL
// Check that the keys have the same RawSeed representation.Nit: This is also odd to see in prose. "raw seed"?
tests removed (in favor of WPT serialization tests, see crbug.com/512509718)
// Check that the keys have the same RawPublic representation.Hubert Chao"raw public"?
test removed (in favor of WPT serialization tests, see crbug.com/512509718)
Ditto. But also why allocate this at all instead of just using the no-params tag?
resolving as a part of crbug.com/512509718
Ditto re what the name of this algorithm is.
Acknowledged
...oh, is this where KEM-XWING came from? Do you have a link to the spec this is implementing? draft-connolly-cfrg-xwing-kem calls it "X-Wing" while draft-irtf-cfrg-concrete-hybrid-kems calls it "MLKEM768-X25519".
changing name to `MLKEM768-X25519` as per outside discussion.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |