OpenJDK / amber / amber
changeset 42318:35806bea1e90
8169243: java/util/concurrent/ThreadPoolExecutor/ConfigChanges.java still fails intermittently
Reviewed-by: martin, smarks, psandoz, amlu
| author | dl |
|---|---|
| date | Mon, 28 Nov 2016 23:33:25 -0800 |
| parents | debe8d32a0a4 |
| children | 0193886267c3 |
| files | jdk/test/java/util/concurrent/ThreadPoolExecutor/ConfigChanges.java |
| diffstat | 1 files changed, 7 insertions(+), 1 deletions(-) [+] |
line wrap: on
line diff
--- a/jdk/test/java/util/concurrent/ThreadPoolExecutor/ConfigChanges.java Mon Nov 28 17:32:49 2016 -0800 +++ b/jdk/test/java/util/concurrent/ThreadPoolExecutor/ConfigChanges.java Mon Nov 28 23:33:25 2016 -0800 @@ -247,7 +247,13 @@ tg.activeCount() > 0) Thread.sleep(4); equal(tg.activeCount(), 0); - check(System.nanoTime() - t0 >= tpe.getKeepAliveTime(NANOSECONDS)); + + // The following assertion is almost always true, but may + // exceptionally not be during a transition from core count + // too high to allowCoreThreadTimeOut. Users will never + // notice, and we accept the small loss of testability. + // + // check(System.nanoTime() - t0 >= tpe.getKeepAliveTime(NANOSECONDS)); //report("idle", tpe);