RFR: 8290025: Remove the Sweeper

[Boris Ulasevich]

On Thu, 4 Aug 2022 20:20:53 GMT, Erik Österlund <eosterlund at openjdk.org> wrote:

> It still might be that my new algorithm is more aggressive, but the data does not conclude that

I tweaked JMH code to run Warmup iterations for all benchmarks first (a single vm run mode), and than run the final Iterations for the benchmarks (renaissance suite in my case). The idea is to check if a new sweeper algorithm flush the hot methods before the final iteration and get a performance penalty. I played with a benchmark set, combining them into different groups, changing the number of Warmup iterations and VM options. With this approach I do not find evidence that the new algorithm is more aggressive than the previous one.

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

PR: https://git.openjdk.org/jdk/pull/9741