On Fri, 10 Jan 2025 01:01:28 GMT, Weijun Wang <wei...@openjdk.org> wrote:
>> Traditionally, an asymmetric key has a key size. The size is displayed by >> `keytool` and `jarsigner`, both in informational output and weak-key >> warnings. However, for the recently added ML-DSA algorithm, key size is not >> defined. >> >> Thus when an ML-DSA key is created, `keytool` shows >> >> Generating -1 bit ML-DSA-65 key pair... >> >> When the entry is being displayed by `keytool -list -v`, it shows >> >> Subject Public Key Algorithm: -1-bit ML-DSA-65 key >> >> If the algorithm is disabled, `keytool -list` shows >> >> <x> uses a -1-bit ML-DSA-65 key which is considered a security risk... >> >> Furthermore, if a JAR file is signed by ML-DSA, `jarsigner -verify` also >> shows >> >> Signature algorithm: ML-DSA-65, unknown size >> >> or when the algorithm is disabled, it shows >> >> Signature algorithm: ML-DSA-65, -1-bit key (disabled) >> The ML-DSA-65 signing key has a keysize of -1 which is considered a security >> risk. >> >> >> With this code change, a key can either has a key size, or characterized by >> a `NamedParameterSpec`, and the display chooses one of them. >> >> One special case is EC keys, which have both a keysize and a >> `NamedParameterSpec`. Both are displayed. > > Weijun Wang has updated the pull request incrementally with one additional > commit since the last revision: > > one missing year change src/java.base/share/classes/sun/security/tools/keytool/Main.java line 2068: > 2066: * Returns the full display name of the given key object. Could be > 2067: * - "X25519", if its getParams() is NamedParameterSpec > 2068: * - "EC (secp256r1)", if it's an EC key Is this true if it is a 3rd-party provider? `NamedCurve` is an internal class. ------------- PR Review Comment: https://git.openjdk.org/jdk/pull/22735#discussion_r1910461431
