RFR: 8334247: [PPC64] Consider trap based nmethod entry barriers [v5]

[Sorna Sarathi N]

On Mon, 29 Sep 2025 10:48:41 GMT, Richard Reingruber <rrich at openjdk.org> wrote:

> > On the performance side, the benchmark results consistently demonstrate a substantial improvement in performance compared to the baseline. Importantly, I did not see any regressions in functionality.
> 
> Hi @Sorna-Sarathi, would you mind sharing the benchmark results you are referring to? Thanks, Richard.

Hi, @reinrich 
I checked with Dacapo benchmark in aix-ppc64. There is a significant decrease in benchmark runtime(sec) after the patch
<img width="1400" height="800" alt="gc_runtime_heatmap-final" src="https://github.com/user-attachments/assets/9f5b3f1d-8785-4d6e-9149-f8f8bb8292c9" />
 
 Thanks, Sorna Sarathi N.

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

PR Comment: https://git.openjdk.org/jdk/pull/24135#issuecomment-3346355070