RFR: 8256265: G1: Improve parallelism in regions that failed evacuation [v2]

[Thomas Schatzl]

On Tue, 13 Sep 2022 11:17:07 GMT, Stefan Johansson <sjohanss at openjdk.org> wrote:

>> I think that's it. We expect few failed regions (1-2) with pinned regions, so many threads working on the same region at the same time. If you want we can drop the cache, I did not (re-)measure.
>
> I would prefer to drop it if not needed. From the comment it sounds like the cache was created to avoid updating the global value for each "garbage area", but a middle road would be to update the global value after each chunk. How much work is it to re-measure?

I intend to work and measure in this area again soon'ish so I'll have plenty of opportunity to see if makes a difference. For this change I'll remove the cache.

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

PR: https://git.openjdk.org/jdk/pull/9980