[Webkit-unassigned] [Bug 70107] Custom written CSS lexer

Mon Dec 12 14:49:58 PST 2011

https://bugs.webkit.org/show_bug.cgi?id=70107

Andreas Kling <kling at webkit.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |kling at webkit.org

--- Comment #15 from Andreas Kling <kling at webkit.org>  2011-12-12 14:49:58 PST ---
(In reply to comment #0)
> In the current state it is far from ready, I am just curious about EWS.
> 
> On methanol it is more than twice (~2.18x) as fast as the original.
> 
> The result of 3x2 runs in CPU cycles
> Original:  1354030040 1377989672 1345539012
> Optimized:  622771368  614838444  635476680

Are these callgrind CPU cycle counts, or measured using a hardware CPU profiler? What's the improvement in wall-time? I'm curious because I've seen some rather unbelievable improvement numbers in association with the "methanol" benchmark more than once before and I'm not convinced it can be trusted. (Examples: bug 66851, bug 64262, bug 63909)

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