oscerd commented on PR #29092: URL: https://github.com/apache/camel/pull/29092#issuecomment-6819913109
Pushed `CAMEL-26545: Address review feedback - an order that already leads with a hybrid is left alone, the unavailable log drops to DEBUG, and the preferred list is derived from the hybrids` (d40e0e4a941e) on top of f4a31f0f7f6e: 1. **Leave the JVM order alone when its first group is already a hybrid:** `resolvePqcNamedGroups` returns `null` in that case, so a CNSA 2.0 style `SecP384r1MLKEM1024,SecP256r1MLKEM768,secp384r1` keeps the P-384 hybrid first, and the default order of a JDK 25 is left as it is. The caller tells that case from "no hybrid at all" through a new `hasPqcNamedGroup`, so only the latter logs. Tests: `testResolvePqcNamedGroupsLeavesAnOrderThatAlreadyLeadsWithAHybridAlone` and `testHasPqcNamedGroup`; the JVM-dependent tests expect the untouched first group when it is a hybrid. 2. **Thread-safety test:** left as it is; as you say it cannot fail without the lock on a JDK whose default already leads with a hybrid, and making the race observable would need a hook into the publication window that is not worth adding for the test. 3. **Java 21 baseline:** the "no post-quantum named group" message is logged at DEBUG now, once per JVM, and reads as a fact rather than an error. 4. **No backport:** none planned, as agreed; this answers gnodet-bot's label note too. On gnodet-bot's thread: `PQC_PREFERRED_NAMED_GROUPS` is now derived from `PQC_NAMED_GROUPS` plus the classical fallbacks, so a hybrid added to the first list takes part in the detection and the ordering at once. `SSLContextParametersTest`: 31 tests, 0 failures, on JDK 21 (no hybrid available locally). Review re-requested. _Claude Code on behalf of oscerd_ -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
