On Tue, 15 Sep 2026 22:06:28 GMT, Weijun Wang <[email protected]> wrote:
>> src/java.base/share/classes/sun/security/tools/keytool/Main.java line 1935:
>>
>>> 1933: keysize =
>>> SecurityProviderConstants.DEF_DSA_KEY_SIZE;
>>> 1934: } else if ("EdDSA".equalsIgnoreCase(keyAlgName)) {
>>> 1935: keysize =
>>> SecurityProviderConstants.DEF_ED_KEY_SIZE;
>>
>> Can we also remove the constants not used here like `DEF_ED_KEY_SIZE` now?
>> For example, it looks like we can initialize with the Ed25519
>> `NamedParameterSpec`
>> [here](https://github.com/openjdk/jdk/blob/d9397cebdb4945dbf10f9d9751fee72781dd89cf/src/java.base/share/classes/sun/security/ec/ed/EdDSAKeyPairGenerator.java#L54)
>> instead of the key size. That is the only other place it is used.
>
> Yes.
>
> `DEF_ED_KEY_SIZE` is only used in `EdDSAKeyPairGenerator` which could have
> been a `NamedParameterSpec`. `DEF_XEC_KEY_SIZE` is not used at all because
> `XDHKeyPairGenerator` has its own default (a `NamedParameterSpec`) inside.
> `DEF_EC_KEY_SIZE` is used by both SunEC and SunPKCS11 and is translated to a
> name using `ECUtil.getECParameterSpec` at a very early stage.
>
> I suggest we migrate all `DEF_EC_KEY_SIZE`, `DEF_ED_KEY_SIZE` and
> `DEF_XEC_KEY_SIZE` to names and reference them from our own implementations.
> When we add PKCS11 implementations on EdDSA and XDH we can use them.
>
> What do you think? We can create another issue on this change.
I'm fine with doing it in a separate issue.
-------------
PR Review Comment: https://git.openjdk.org/jdk/pull/32616#discussion_r4027316558
