RFR: 8302147: Speed up compiler/jvmci/compilerToVM/IterateFramesNative.java [v3]

[Emanuel Peter]

> We have two run statements.
> 
> The `-Xbatch`  version takes `5 sec`, now `1.5`.
> And the `-Xcomp`  took `19` and now takes `0.5`.
> 
> I discussed it with @woess the original author of the test. He suggested this fix.
> The first run statement still works as before, the second one, which sets `-Dcompiler.jvmci.compilerToVM.IterateFramesNative.checkCompiled=true` and has `CHECK_COMPILED == true` now only runs one iteration. This should be sufficient, we see that the whitebox checks successfully that the methods were compiled.
> 
> **Update**
> For `-Xbatch`, I simply reduced the iteration count. I saw that after `255` iterations both methods were already compiled, so I thought `1000` should suffice. I moved the compilation-verification outside the loop - so now we check that we compile in both cases.

Emanuel Peter has updated the pull request incrementally with one additional commit since the last revision:

  alternative that looks a bit better

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

Changes:
  - all: https://git.openjdk.org/jdk/pull/12496/files
  - new: https://git.openjdk.org/jdk/pull/12496/files/72c53dfb..095ee38b

Webrevs:
 - full: https://webrevs.openjdk.org/?repo=jdk&pr=12496&range=02
 - incr: https://webrevs.openjdk.org/?repo=jdk&pr=12496&range=01-02

  Stats: 15 lines in 1 file changed: 6 ins; 8 del; 1 mod
  Patch: https://git.openjdk.org/jdk/pull/12496.diff
  Fetch: git fetch https://git.openjdk.org/jdk pull/12496/head:pull/12496

PR: https://git.openjdk.org/jdk/pull/12496