On Thu, 14 Apr 2022 15:36:20 GMT, Roger Riggs <rri...@openjdk.org> wrote:
> Can you summarize the performance results and implications. Including the > whole jmh run isn't necessary. @RogerRiggs In short, making such a cache can help performance, and the larger int skip length people call `skip()`, the larger difference. And adding SoftReference mechanism have a little acceptable performance cost. ------------- PR: https://git.openjdk.java.net/jdk/pull/5872