RFR: 7903450: JMH: Improve -prof perfnorm accuracy with robust estimations

[Aleksey Shipilev]

If we measure the mostly-idle workload, like the sleep-wait:
  while (target > System.nanoTime()) Thread.sleep(1000);

...then the `-prof perfnorm` measurement would mostly contain the iteration prologs/epilogs, which take orders of magnitude more cycles than the sleep itself. This can be alleviated by using a more robust estimator in perfnorm. The estimator can filter out the "outliers" caused by infrastructural overhead.

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

Commit messages:
 - Work

Changes: https://git.openjdk.org/jmh/pull/114/files
 Webrev: https://webrevs.openjdk.org/?repo=jmh&pr=114&range=00
  Issue: https://bugs.openjdk.org/browse/CODETOOLS-7903450
  Stats: 181 lines in 1 file changed: 48 ins; 63 del; 70 mod
  Patch: https://git.openjdk.org/jmh/pull/114.diff
  Fetch: git fetch https://git.openjdk.org/jmh.git pull/114/head:pull/114

PR: https://git.openjdk.org/jmh/pull/114