RFR: 8358594: Misleading keyLength value captured in JFR event for ML-KEM key [v2]

[Weijun Wang]

On Wed, 4 Jun 2025 16:08:31 GMT, Artur Barashev <abarashev at openjdk.org> wrote:

>> Weijun Wang has updated the pull request incrementally with one additional commit since the last revision:
>> 
>>   addressing Mark's comments
>
> src/java.base/share/classes/sun/security/util/KeyUtil.java line 62:
> 
>> 60:      * each standardized parameter set. For example, ML-KEM-768 is assigned to
>> 61:      * category 3, and ML-DSA-87 to category 5.
>> 62:      *
> 
> Should we consider returning whatever number is an the end of PQC algorithms as a key size? That would make things consistent and it would allow us to use existing `keySize` algorithm constraints for PQC algorithms. Key sizes for RSA and EC algorithms already differ significantly for the same security level: 3072-bit RSA corresponds to 256-bit EC. So we can return `768` for ML-KEM-768 or `87` for ML-DSA-87.

For ML-DSA-87, 87 isn’t a key size in any sense. Using it as a key size would be misleading. For algorithm constraints, we can use the parameter set name directly.

-------------

PR Review Comment: https://git.openjdk.org/jdk/pull/25642#discussion_r2127040718