RFR: 8289943: Simplify some object allocation merges [v10]

[Cesar Soares Lucas]

On Fri, 30 Sep 2022 17:31:11 GMT, Vladimir Kozlov <kvn at openjdk.org> wrote:

>> And I will submit performance runs.
>
>> And I will submit performance runs.
> 
> Performance results are mixed bag. Saw some improvements on x64 in Renaissance but they gone after rerun.
> I see some significant regression in Dacapo on Linux Aarch64 (Ampere CPU) but don't see it at all on MacOS M1.

Hi @vnkozlov  & @merykitty - thank you for the reviews. I'm working to address your comments. I had to take some time off of work due to personal issues but I believe I can address your comments during the weekend or earlier next week.

@vnkozlov - Thank you so much for running the perf. tests! Can you please give me more details about the experiment? What metric did you analyze, which version of Dacapo, and was it any specific benchmark that regressed or all of them? Which JVM parameters?

Regards,
Cesar

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

PR: https://git.openjdk.org/jdk/pull/9073