RFR: 8319897: Move StackWatermark handling out of LockStack::contains [v2]

[Daniel D. Daugherty]

On Mon, 13 Nov 2023 08:56:09 GMT, Stefan Karlsson <stefank at openjdk.org> wrote:

>> There are StackWatermark handling inside LockStack::contains. This is an indication that either:
>> 1) some thread has failed to start it's own StackWatermark processing, or
>> 2) that some thread is looking into the oops of another thread without starting the StackWatermark handling. The latter is usually more common.
>> 
>> I propose that we hoist the StackWatermark processing out of the LockStack into the code that actually performs pokes around in other thread's data. This is usually where we put the StackWatermark processing calls.
>> 
>> Note: that this doesn't fix a bug, it merely moves the processing code nearer the source of the problematic calls. I've found that jmm_GetThreadInfo looks straight into running thread's lock stacks without while those threads are running. That is a race that we should get rid of and I've created [JDK-8319899](https://bugs.openjdk.org/browse/JDK-8319899) for that issue.
>
> Stefan Karlsson has updated the pull request incrementally with one additional commit since the last revision:
> 
>   Stream line the assert

Thumbs up. Thanks for adding the assert in `LockStack::contains()`
since that should catch any accidental new uses. Also thanks for
documenting the testing that has been done.

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

Marked as reviewed by dcubed (Reviewer).

PR Review: https://git.openjdk.org/jdk/pull/16609#pullrequestreview-1735075281