Github user JoshRosen commented on the issue: https://github.com/apache/spark/pull/18400 I actually asked @liufengdb to submit this PR after he spotted this problem while debugging a test. I did a bit of digging and spotted where this bug was introduced: in https://github.com/JoshRosen/spark/commit/b9d6e181db0c7293b0b00b3a679716f652df936f, an intermediate commit in an old PR, this was introduced when refactoring one map from a thread-safe LoadingCache (where entires were implicitly created) to an explicit map with explicit registration (which used synchronization). In making this change I incorrectly preserved a bad synchronization pattern which was unnecessarily fine-grained. Therefore this looks like the right change to me so I'm going to retest and merge.
--- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. --- --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org