[Webkit-unassigned] [Bug 77509] Enable IPP for Biquad filter

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Tue Feb 7 21:01:38 PST 2012


--- Comment #23 from Raymond Toy <rtoy at chromium.org>  2012-02-07 21:01:38 PST ---
(In reply to comment #22)
> > But I will look into this some more on my system (but without IPP.)
> I cleaned the out dir and recompiled the DumpRenderTree, the tests still failed.

Ok.  I think I will need to get a copy of ipp to test with.

> > > 
> > > I tried to adjust the value of 'timeStep' in biquad-testing.js as the comments, but it did not work. I could not figure out the problem in patch or in the test scripts, could you provide some help on this?  Thanks a lot.
> > 
> > What did you change timeStep to?  And in what way did it not work?
> > 
> As your comments, I tried enlarge the timeStep to the values like 0.2, 0.5 ,1, 10 and I even tried to set it to some small values, like 0.01. All the tryings could not make all the tests pass always while running the tests together. For every single value, the tests are not passed occasionally, and different tests failed by different values.
> Most of the failed tests shows it has infinites or NaNs, some of them shows err values.

If you can, open up the javascript console and look at renderedData and see if the values make sense.  You may want to modify the test and have the location of the infinity or NaN printed out so you can verify renderedData really has those bad values.

Otherwise, I'll need to get ipp going on my linux machine to debug this.

Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.

More information about the webkit-unassigned mailing list