Can we address UEF-CMA issues?

[danvangeest]

See whole thread: https://mailarchive.ietf.org/arch/msg/spasm/yM8kS1kCoizWCMjS8pdcV3IFaDg/ and https://mailarchive.ietf.org/arch/msg/spasm/uAwmfFDCP2gAggqYSu4ykRCk4mU/

Possible solutions:

• (pure) ML-DSA for SignedArtifacts, HashML-DSA for non-signed artifacts.
• Use context string, "CMW-with-SignedAttributes" and "CMS-without-SignedAttributes" (see also To use context string or not to use context string? #2 ).
• ML-DSA MUST be used with SignedAttributes.

[adam-r-ncsc]

As discussed in #6 (comment), if we don't address the issue directly in the draft (as is looking likely), it could be mentioned in the Security Considerations alongside any interim mitigations prior to the issue being addressed for CMS more generally.