[jdk11u-dev] RFR: 8242312: use reproducible random in hotspot gc tests

[Martin Doerr]

On Fri, 8 Apr 2022 07:42:38 GMT, Goetz Lindenmaier <goetz at openjdk.org> wrote:

> I backport this for parity with 11.0.16-oracle.
> 
> Resolved context, "8272783: Epsilon: Refactor tests to improve performance" already backported
>  test/hotspot/jtreg/gc/epsilon/TestByteArrays.java
>  test/hotspot/jtreg/gc/epsilon/TestElasticTLAB.java
>  test/hotspot/jtreg/gc/epsilon/TestElasticTLABDecay.java
>  test/hotspot/jtreg/gc/epsilon/TestRefArrays.java
> 
> I had to "implement" the change for this test. The test was removed with CMS
> before the backported change was pushed.
>  test/hotspot/jtreg/gc/stress/gcold/TestGCOldWithCMS.java
> 
> Copyright
>  test/hotspot/jtreg/gc/g1/TestShrinkAuxiliaryData10.java
> 
> Context
>  test/hotspot/jtreg/gc/g1/humongousObjects/TestNoAllocationsInHRegions.java
>  test/hotspot/jtreg/gc/logging/TestUnifiedLoggingSwitchStress.java
> 
> "8256806: Shenandoah: optimize shenandoah/jni/TestPinnedGarbage.java test" was already 
> backported. I had to redo the change, differences wrt. random to head are gone now.
>  test/hotspot/jtreg/gc/shenandoah/jni/TestPinnedGarbage.java
> 
> Test at different location, patched manually
>  test/hotspot/jtreg/gc/stress/CriticalNativeStress.java

Thanks for the detailed description. LGTM. This backport should make other backports a bit easier.
Note: We don't build Shenandoah in 11u by default, so it may be missed by our testing.

-------------

Marked as reviewed by mdoerr (Reviewer).

PR: https://git.openjdk.java.net/jdk11u-dev/pull/1021