On Thu, 27 Aug 2026 00:20:58 GMT, Chris Plummer <[email protected]> wrote:
> > Thanks, that makes sense and agreed the full version isn’t worth the > > effort... The cheap version I have just emits the payload oops as > > references from the chunk itself, so it’s the truthful heap graph even if > > not the frame local presentation, follow references and revptrs find the > > holder instead of reporting nothing > > I'm not sure what you are referring to. By "payload oops" do you means the > references in the StackChunk's hidden array? Where is the code emitting the > references? I thought you said the StackChunk instance is represented in the > heap dump using just its visible fields. Sorry for the confusion, conflated two things there. This pr doesn’t emit those references anywhere, The dump record is just the visible fields as I said. The payload oops walk is a separate change sitting on my local branch for the old SA stack chunk bug, and it hooks the reference iteration that follow references and revptrs use, not the dump writer. None of that is in this pr ------------- PR Comment: https://git.openjdk.org/jdk/pull/32525#issuecomment-5432718630
