[jdk17u-dev] RFR: 8266936: Add a finalization JFR event

[Erik Gahlin]

On Tue, 26 Apr 2022 11:45:40 GMT, Jaroslav Bachorik <jbachorik at openjdk.org> wrote:

> I would argue that such crashing event is a haphazard even in a non-update release.

Over time the stability of a release increases if new features are not added. If you are going to adapt a non-update release, you can do extensive testing. This may not be an option if you need to upgrade immediately due to a vulnerability.

If a customer ask why the JVM crashes after an update, it's possible to explain it was a bug that unfortunately came in with another bug fix. I can't explain why a Finalization event was needed for an update release. 

> 
> I agree that if the changes tend to be too complex to be able to have a clean argument about their correctness we should refrain from such backports. But this would be no different from any other backport, AFAIK.

I think there is a difference between bug fixes and features, not just the complexity.

Bug fixes are meant to increase stability. Sometimes you need to backport features, but the bar should be high and there should be a good reason, for example to support new hardware or OS versions. Improve security. Things that are essential to be able to use the existing JDK release. I don't see this JFR event fall into that category. It's also a rather large change.

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

PR: https://git.openjdk.java.net/jdk17u-dev/pull/351