RFR: Add task termination and enhanced task queue state tracking

[Zhengyu Gu]

On 07/12/2018 10:05 AM, Aleksey Shipilev wrote:
> On 07/12/2018 03:39 PM, Zhengyu Gu wrote:
>> Webrev: http://cr.openjdk.java.net/~zgu/shenandoah/termination_tracking/webrev.00/
> 
> The patch looks good.
> 
> How often do threads request termination? Are we sure this tracking does not introduce its own
> latencies, e.g. due timer calls and contention on counters themselves? Maybe we can guard it with
> the diagnostic ShenandoahTerminationTrace flag in ShenandoahTerminationTimings to be able to dissect?

We enter termination twice per GC cycle. My early instrumentation showed 
that on average about 2.3 workers failed to terminate once entering 
termination phase (of course with ShenandoahOWST), with upcoming 
optimization [1]  (my prototype [2]), the average downs to about 0.2.

Let me rerun the numbers.

Thanks,

-Zhengyu


[1] https://bugs.openjdk.java.net/browse/JDK-8205921
[2] http://cr.openjdk.java.net/~zgu/8205921/webrev.03/
> 
> Thanks,
> -Aleksey
> 
>