[Webkit-unassigned] [Bug 54707] Introduce lookup-table based approach for applying CSS properties.

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=54707





--- Comment #35 from Eric Seidel <eric at webkit.org>  2011-03-07 19:00:11 PST ---
(In reply to comment #34)
> (In reply to comment #32)
> > (In reply to comment #31)
> > > Let me know if I can make this clearer, but the bottom line is that we now spend slightly less time in applyProperty and its children, for a net speed-up. I expect this will improve further as more properties are converted over.
> > 
> > OK.  Although I believe you, numbers which encompass an entire layout would be more convincing (to me) than sample counts inside those functions.  It's possible (although unlikely) that reducing the samples inside those functions might still have an increase in total time spent for a style-recalc or layout.
> 
> Uh, embarrassingly it looks like I posted numbers which are the time inside applyProperty, but not including its children. I blame shark's intuitive UI. Let me run it again and get back to you.
> 
> Or I could just attach some profiles to the bug and everyone can look at whatever interests them.

The profiles wouldn't actually help answer my query, as they only show the distribution of time spent in functions, not the total time spent running some specific benchmark.

See the PerformanceTest/ directory for examples of benchmarks.  We have many others which are not checked in due to licensing restrictions (ask jamesr or dglaskov for how to run the PLT for example)

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