mikemccand opened a new issue, #15926: URL: https://github.com/apache/lucene/issues/15926
### Description Java 25 added a new CPUTime profiler which uses Linux kernel APIs to sample threads every N CPU instructions instead of every X msec wall-clock time. This eliminates silly ghosts like threads listening forever on a socket. It [sounds really cool](https://mostlynerdless.de/blog/2025/06/11/java-25s-new-cpu-time-profiler-1/). Wow, it uses [JEP 518](https://openjdk.org/jeps/518) -- look how hard it is to eliminate/reduce safepoint bias! I switched nightly benchy to CPU time profiler, but it produces vacuous reports, clearly missing all the meat. See https://github.com/mikemccand/luceneutil/issues/562 for specifics. I think it's because `ProfileResults` tool doesn't know about the new `jdk.CPUTimeSample` event -- the fix should be simple, in the [`isInteresting` method](https://github.com/apache/lucene/blob/e33724c24389d0390767334f051a48884fd6a82b/build-tools/build-infra/src/main/java/org/apache/lucene/gradle/plugins/java/ProfileResults.java#L107-L109). I'll try to make a PR soon. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
