RFR(xs): 8186402: [TESTBUG] "Balance queues" output expected by test

Thomas Schatzl thomas.schatzl at oracle.com
Tue Aug 22 09:20:18 UTC 2017


On Fri, 2017-08-18 at 13:17 -0700, sangheon.kim wrote:
> Hi all,
> Could I have some reviews for this tiny change of adding
> ParallelGCThreads=2 in TestPrintReferences.java?
> The test is assuming always to have 'Balance queues:' log but it is
> only generated when the VM did queue balancing. Queue balancing
> happens when MT process is enabled(with
> +ParallelRefProcBalancingEnabled) or MT discovery is enabled. And MT
> discovery is enabled if we have multiple ParallelGCThreads (> 1).
> Adding explicit thread setting of 'ParallelGCThreads=1' is easiest
> way to avoid this test failure on limited environments. i.e. machines
> which determine to use 1 ParallelGCThreads.
> The test failure log also includes:
> "[Java HotSpot(TM) 64-Bit Server VM warning: If the number of
> processors is expected to increase from one, then you should
> configure the number of parallel GC threads appropriately using
> -XX:ParallelGCThreads=N"
> CR: https://bugs.openjdk.java.net/browse/JDK-8186402
> Webrev: http://cr.openjdk.java.net/~sangheki/8186402/webrev.0/
> Testing: JPRT

  looks good.


More information about the hotspot-gc-dev mailing list