RFR: 8317453: NMT: Performance benchmarks are needed to measure speed and memory [v4]

[David Holmes]

On Fri, 17 Jan 2025 16:29:00 GMT, Gerard Ziemski <gziemski at openjdk.org> wrote:

>> src/hotspot/share/nmt/memLogRecorder.cpp line 43:
>> 
>>> 41: // then to actually run the benchmark:
>>> 42: //
>>> 43: // ./build/macosx-aarch64-server-release/xcode/build/jdk/bin/java -XX:+UnlockDiagnosticVMOptions -XX:NMTBenchmarkRecordedPID=43100 -XX:NMTBenchmarkRecordedLoops=10
>> 
>> ?? How does this run anything - you haven't passed an initial class to the Java invocation?
>
> We check for the flags when we initialize NMT and run only when NMTBenchmarkRecordedPID is set.

We do have some terminal flags like `-version` and `-Xshare:dump`, but generally the VM does not have magic flags that behave that way. I still don't have a clear picture of the different phases involved with this work. We have a recording phase. Then we have the benchmarking phase that does something (I can't figure out what) with the recording. Then we have the Java tool which does ... what?

In any case I have concerns about how the "benchmarking" is initiated.

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

PR Review Comment: https://git.openjdk.org/jdk/pull/23115#discussion_r1921701378