JDK-8100627 : TreeItem expand/collapse performance in hardware pipeline is almost 20 fps worse than in j2d pipeline
  • Type: Enhancement
  • Component: javafx
  • Sub-Component: graphics
  • Affected Version: fx2.0
  • Priority: P3
  • Status: Resolved
  • Resolution: Cannot Reproduce
  • Submitted: 2011-08-05
  • Updated: 2015-06-16
  • Resolved: 2012-11-02
The Version table provides details related to the release that this issue/RFE will be addressed.

Unresolved : Release in which this issue/RFE will be addressed.
Resolved: Release in which this issue/RFE has been resolved.
Fixed : Release in which this issue/RFE has been fixed. The release containing this fix may be available for download as an Early Access Release or a General Availability Release.

To download the current JDK release, click here.
JDK 8
8Resolved
Related Reports
Relates :  
JDK-8097926 - Pisces Renderer shows huge performance win when coded in C
Description
TreeItem expand/collapse performance in hardware pipeline is worse than in j2d pipeline.

Here are results of running "TreeViewExpandPerfTest -size 100 -type wide" with fx2.0-b39 on windows xp midrange machine.
 hw: 34.92 fps
 j2d: 51.56 fps

I also see similar on my Dell E6410 laptop.

It would be good to investigate and improve.
Comments
I'm marking this as "cannot reproduce" rather than "fixed" because there was no specific change we made to address this exact problem. But, since the combined work on performance in a number of areas has brought us up to (near) parity for this benchmark, the original "performance issue" is not currently reproducible. If we have performance regressions on this benchmark, they can be filed as new issues that we can fix as they arise...
02-11-2012
The latest results we have for both j2d and hw pipelines are for build 8.0-b55. WinXP-mid-range results were: TreeView-Expand-items100-depth: j2d: 166 hw: 168 1% (1.72) TreeView-Expand-items100-wide: j2d: 401 hw: 368 -8% (-33.11) Win7-High-end results were: TreeView-Expand-items100-depth: j2d: 209 hw: 209 0% (0.19) TreeView-Expand-items100-wide: j2d: 468 hw: 477 2% (9.73) So, looks like the issue is fixed.
01-11-2012
ok to defer since current diff is not big
23-08-2011
OK to defer.
23-08-2011
Note that with build #2692 of presidio-graphics-scrum I see a much smaller difference on my laptop (Intel Core I5 with Intel HD graphics): D3D :57.0 fps J2D : 61.6 fps If I turn full-speed on, I see a somewhat greater difference (d3d:84.0, j2d:97.3). This is very likely reflective of the performance difference in rendering general shapes, e.g., as reflected in GUIMark2.Vector.
17-08-2011
This is a performance optimization which can be looked at in Lombard.
17-08-2011
this is quite strange, I just downloaded latest graphics-scrum build 2610 and run attached TreeViewExpandPerfTest. I am getting the following results on fx-win-hi-2 machine: d3d: up to 37 fps j2d: up tp 55 fps
08-08-2011
This is likely a graphics team bug, and likely related to RT-15344.
07-08-2011
I just tried running the test app and it seems that all FPS values are being printed out as 0.00. I'm not sure why but until this is fixed I can't a great deal further.
07-08-2011
potential performance bottleneck which perhaps is easy to improve.
05-08-2011