Shenandoah performance problem

[Aleksey Shipilev]

On 10/17/19 12:49 PM, Aleksey Shipilev wrote:
> This workload runs:
>   Parallel GC, -t 1:     0.112 ± 0.006  ms/op
>   Parallel GC, -t max:   0.146 ± 0.002  ms/op
>   Shenandoah GC, -t 1:   0.120 ± 0.002  ms/op
>   Shenandoah GC, -t max: 1.595 ± 0.050  ms/op <--- scalability bottleneck
> 
> And you are right, that's the contention on pinning/unpinning, I am looking what can be done about
> that...

This bunch of fixes makes it much better:
  https://cr.openjdk.java.net/~shade/shenandoah/async-pinning/

Shenandoah GC, -t max:   0.157 ± 0.003  ms/op

Going to test and RFR the pieces of it now.

-- 
Thanks,
-Aleksey