RFR: JDK-8323695 RenderPerf (2D) enhancements (23.12)

[Laurent Bourgès]

On Tue, 16 Jan 2024 21:46:24 GMT, Phil Race <prr at openjdk.org> wrote:

>> - new executor modes (buffer & volatile)
>> - support parallel rendering on several frames and / or screens
>> - added robot calibration + optionally use rounded duration to min frame latency (120Hz)
>> - give more statistics in verbose (-v)
>> - added new command-line arguments
>> - added version + help information
>
> test/jdk/performance/client/RenderPerfTest/src/renderperf/RenderPerfTest.java line 204:
> 
>> 202:                     enabled ? RenderingHints.VALUE_TEXT_ANTIALIAS_LCD_HRGB
>> 203:                             : RenderingHints.VALUE_TEXT_ANTIALIAS_DEFAULT);
>> 204:         }
> 
> Although it makes no practical difference, I am wondering why in this case, enabled==false sets "DEFAULT" and in the case above it sets "OFF" ?

Will check

> test/jdk/performance/client/RenderPerfTest/src/renderperf/RenderPerfTest.java line 460:
> 
>> 458:             if (id % 100 != 0) return;
>> 459:             if (font == null) {
>> 460:                 font = new Font("LucidaGrande", Font.PLAIN, 32);
> 
> I see this is pre-existing, but we should change this.
> That's a font available only on macOS. You should either look up a font to use on the system at run time,
> or use a logical font.

Of course, I will make constant Font instances (no new Font() in benchmark code!) and use family name (Dialog or SansSerif), what do you prefer?

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

PR Review Comment: https://git.openjdk.org/jdk/pull/17408#discussion_r1457122289
PR Review Comment: https://git.openjdk.org/jdk/pull/17408#discussion_r1457123987