RFR: 8343704: Bad GC parallelism with processing Cleaner queues

[Brent Christian]

On Tue, 12 Nov 2024 16:00:39 GMT, Aleksey Shipilev <shade at openjdk.org> wrote:

> See the bug for more discussion and reproducer. This PR replaces the ad-hoc linked list with the `ArrayList` wrapper that manages synchronization, search and replacements efficiently. Arrays are easy targets for GC. There are possible improvements here, most glaring is parallelism that is currently knee-capped by global synchronization. The synchronization scheme follows what we have in original code, and I think it is safer to continue with it right now.
> 
> I'll put performance data in a separate comment.
> 
> Additional testing:
>  - [x] Original reproducer improves drastically
>  - [x] New microbenchmark shows no regression on "churning" tests, which covers insertion/removal perf
>  - [x] New microbenchmark shows improvement on Full GC times (crude, but repeatable), serves as a proxy for reproducer
>  - [x] `java/lang/ref` tests in release 
>  - [ ] `all` tests in fastdebug

Changes requested by bchristi (Reviewer).

src/java.base/share/classes/jdk/internal/ref/CleanerImpl.java line 300:

> 298:                 return true;
> 299:             }
> 300:         }

It looks like the new `PhantomCleanableList.remove()` puts 100% trust in the correctness of the PhantomCleanable's `index`.

What would you think of adding some sort of check/assert that `list.get(phc.index) == phc` ?

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

PR Review: https://git.openjdk.org/jdk/pull/22043#pullrequestreview-2431221723
PR Review Comment: https://git.openjdk.org/jdk/pull/22043#discussion_r1839073366