[Webkit-unassigned] [Bug 240605] New: Increase maximum complexity of multiply Suite in MotionMark

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=240605

            Bug ID: 240605
           Summary: Increase maximum complexity of multiply Suite in
                    MotionMark
           Product: WebKit
           Version: WebKit Nightly Build
          Hardware: Unspecified
                OS: Unspecified
            Status: NEW
          Severity: Normal
          Priority: P2
         Component: New Bugs
          Assignee: webkit-unassigned at lists.webkit.org
          Reporter: sky at chromium.org

Created attachment 459557

  --> https://bugs.webkit.org/attachment.cgi?id=459557&action=review

Chrome running the multiply suite

On laptops the multiply suite has a maximum complexity of ~3100. STP 145 is pretty close to this as is Chrome. On recent MacBook Pros with ProMotion, Chrome receives an extremely low score (typically less than 10) even though it can generally handle the complexity (maintains a frame rate of 60hz). Chrome receives a low score because of two factors:

1. Chrome processes requestAnimationFrame at 120Hz.
2. The maximum complexity is not high enough.

These two result in Chrome getting a low score on multiply because the line fitting used to determine the score generally does not cross the 60hz line. In this case the scoring typically uses the first value seen, which is a complexity of around 1. I've attached a screenshot illustrating this. I'm happy to attach any other data that would be helpful (perhaps the JSON results).

All the other suites in MotionMark run long enough such that the line fitting uses a reasonable value.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20220518/72741408/attachment-0001.htm>