RFR: 8320304: Refactor and simplify monitor deflation functions [v3]

[Stefan Karlsson]

On Mon, 20 Nov 2023 20:39:26 GMT, Daniel D. Daugherty <dcubed at openjdk.org> wrote:

>> The call to `log.before_blocked()` on L1725 will stop the timer so it will not
>> be running during the ThreadBlockInVM.
>> The call to `log.after_blocked()` on L1731 will restart the timer.
>> The advantage of having these two calls is that you know how long we spent
>> in ThreadBlockInVM based on the timestamps in the two log mesgs.
>
> Definitely going to have to merge with [JDK-8319137](https://bugs.openjdk.org/browse/JDK-8319137)!!

>The call to log.before_blocked() on L1725 will stop the timer so it will not
be running during the ThreadBlockInVM.
The call to log.after_blocked() on L1731 will restart the timer.
The advantage of having these two calls is that you know how long we spent in ThreadBlockInVM based on the timestamps in the two log mesgs.

It's unclear to me if you want me to change anything. Patricio pointed out that the pre-existing logs are around the "wrong" part of the code. It's not interesting to log the constructor, but rather the destructor. Alternatively, the code that deletes the monitors.

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

PR Review Comment: https://git.openjdk.org/jdk/pull/16706#discussion_r1399758776