Using a bunch of Semaphores here seems a bit surprising. One more CountDownLatch or CyclicBarrier *might* have been better. But I guess it's not of concern.
For the diagnostic purposes I would add a newline between t4.join(); t3.join(); since they throw exceptions. Otherwise looks good. Thanks for doing this. -Pavel > On 13 Mar 2019, at 14:41, Pavel Rappo <pavel.ra...@oracle.com> wrote: > > I'm reading the test. Still trying to figure out how it works in detail. Give > me > some more time to get comfortable with it and I will reply to the list. At the > very least I would (suggestion) change the @summary. After all if this group > is > being destroyed, and this is the only call, then 0 would be a correct value, > wouldn't it? > > -Pavel
