AOT MXBean

[macarte]

passes tests

[swesonga]

@macarte, which is the upstream JBS issue that got this work into tip?

[macarte]

@swesonga , here's the upstream issue: https://bugs.openjdk.org/browse/JDK-8369736

Note that the upstream implementation only provides one method; as the Leyden group wants to slowly introduce other functionality (currently the priority was only to provide the ability to stop training)

Where as this implementation which we introduced in our jdk25 release as more methods

[swesonga]

@swesonga , here's the upstream issue: https://bugs.openjdk.org/browse/JDK-8369736

Note that the upstream implementation only provides one method; as the Leyden group wants to slowly introduce other functionality (currently the priority was only to provide the ability to stop training)

Where as this implementation which we introduced in our jdk25 release as more methods

Which JBS issue corresponds to this work in Leyden/premain?

[macarte]

@swesonga , here's the upstream issue: https://bugs.openjdk.org/browse/JDK-8369736
Note that the upstream implementation only provides one method; as the Leyden group wants to slowly introduce other functionality (currently the priority was only to provide the ability to stop training)
Where as this implementation which we introduced in our jdk25 release as more methods

Which JBS issue corresponds to this work in Leyden/premain?

https://bugs.openjdk.org/browse/JDK-8356572
openjdk/leyden#52

[swesonga]

Should this test have the if (isLeydenWorkflow()) checks in the Leyden PR? Why are these tests different?

[macarte]

@swesonga , here's the upstream issue: https://bugs.openjdk.org/browse/JDK-8369736
Note that the upstream implementation only provides one method; as the Leyden group wants to slowly introduce other functionality (currently the priority was only to provide the ability to stop training)
Where as this implementation which we introduced in our jdk25 release as more methods

Which JBS issue corresponds to this work in Leyden/premain?

There was no JBS issue, work going into leyden/premain doesn't always have an associated JBS issue; here's the PR: openjdk/leyden#52

[macarte]

@swesonga , here's the upstream issue: https://bugs.openjdk.org/browse/JDK-8369736
Note that the upstream implementation only provides one method; as the Leyden group wants to slowly introduce other functionality (currently the priority was only to provide the ability to stop training)
Where as this implementation which we introduced in our jdk25 release as more methods

Which JBS issue corresponds to this work in Leyden/premain?

Sorry I thought I'd already replied to this; I didn't recall there being a JBS issue, but it turns out there was: https://bugs.openjdk.org/browse/JDK-8356572

here's the PR: openjdk/leyden#52

Of note, the not all the functionality in premain was moved to tip; but we kept that functionality in our MS build in order to prove out its value as we have a lot less overhead. Upstream would require documentation CSRs, and there's a process to removing bean functionality that the Leyden team wanted to avoid while we proved out its usefulness (which hopefully we get to do by supporting Jaz)