Re: ForkJoin refresh
Doug Lea via Concurrency-interest <[email protected]>
| Newsgroups | gmane.comp.java.jsr.166-concurrency |
|---|---|
| Message-ID | <[email protected]> |
On 1/17/20 4:54 PM, Dr Heinz M. Kabutz wrote: > Thanks Doug - how different is this to the Java 14-EA version? I One thing you might notice is the lack of performance oddities when tasks unblock but GC hasn't yet unbiased locks. Also better behaved when a bunch of exceptions are thrown by tasks. Also less flailing with relatively slow streams of incoming tasks. Plus the other improvements I mentioned > noticed the ManagedBlocker implementation today whilst showing the > Condition.await() implementation to a class I was teaching and even > tweeted about it: https://twitter.com/heinzkabutz/status/1218204292626243585 > Right. I had forgotten that because few people use non-LTS JDKs, most haven't seen that blocking on BlockingQueues etc, along with most other j.u.c locking/sync work better especially in FJ programs. -Doug _______________________________________________ Concurrency-interest mailing list [email protected] http://cs.oswego.edu/mailman/listinfo/concurrency-interest