On Tue, 28 Apr 2026 13:22:26 GMT, Alan Bateman <[email protected]> wrote:
>> That combined with the new `_processed` flag just makes the assert for !NEW >> even more redundant. I think you can just delete that assertion as it was a >> proxy for the new assertion, and you can restore the default status to NEW >> instead of TERMINATED. > > That would be a bit clearer, and would be nicer if it could return nullptr if > !cl.processed or the target threadhas the TERMINATED state. Ok, I restored the default status to NEW and brought `_processed` out of `DEBUG_ONLY`. Also, I kept the original assert since `_processed` is not exactly the same as `cl._thread_status != NEW`. We could have processed the handshake but read a NEW state (of course it should never happen, but that is the point of the assert). ------------- PR Review Comment: https://git.openjdk.org/jdk/pull/30903#discussion_r3156806130
