JDK-8259593 : Test plan performance success criteria.
  • Type: Sub-task
  • Component: client-libs
  • Sub-Component: 2d
  • Affected Version: internal
  • Priority: P4
  • Status: Closed
  • Resolution: Delivered
  • Submitted: 2021-01-11
  • Updated: 2021-01-11
  • Resolved: 2021-01-11
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.
Other
internalResolved
Description
The tools and tests used to measure performance are described in the test plan (https://bugs.openjdk.java.net/browse/JDK-8251396)

Since the point of this JEP is to replace a deprecated pipeline we need to be flexible on performance
so the following outlines goals but we have to reserve the right to waive specific performance regressions

Success criteria for performance

- No open P1 or P2 performance bugs

- we expect most performance measures to at least be similar to OpenGL (within 10% on like-for-like h/w)

-  we are not blocked by performance regressions that are marginal (5-10%) even if in some widely used case.

-  we aren't blocked by performance regressions no matter how large, if it is deemed
 an unimportant/marginal/uncommon case.
Comments
Results There are no open P1 or P2 perfomance bugs Relying on J2DBench the majority of tests are as fast or faster. We have a notable regression with XOR but this is a corner case and is due to Metal not providing the necessary support
11-01-2021