[Webkit-unassigned] [Bug 63531] New: Turn PreloadScanner implementation to a thread based solution

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=63531

           Summary: Turn PreloadScanner implementation to a thread based
                    solution
           Product: WebKit
           Version: 528+ (Nightly build)
          Platform: Unspecified
        OS/Version: Unspecified
            Status: NEW
          Severity: Normal
          Priority: P2
         Component: WebCore Misc.
        AssignedTo: webkit-unassigned at lists.webkit.org
        ReportedBy: zoltan at webkit.org


Created an attachment (id=98909)
 --> (https://bugs.webkit.org/attachment.cgi?id=98909&action=review)
patch for comments

I implemented PreloadScanner to use a worker thread.

My observations...

Some "performance" results:

# Measurement with html-parser performance test (trunk/PerformanceTests/Parser/html-parser.html) runs:

with PreloadScanner: 1774 ms (+/-5.8%)
without PreloadScanner: 23.7 ms (+/-0.55%)
with my ThreadedPreloadScanner: 77.15 ms (+/-1.5%)

So I think html-parser test is not suitable for htmlparser benchmarking.

# Measurement with our Methanol benchmark (23 locally mirrored sites) + 1 site what contains 3 iframes

with PreloadScanner: 19208 ms (+/-2.61%)
without PreloadScanner: 18335 ms (+/-3.74%)
with my ThreadedPreloadScanner: 17423 ms (+/-2.65%)

I had issues with AtomicString and QualifiedName comparison on worker thread, so I turned some variables into simple WTF::String.

I haven't modified the layouttests of the preloadscanner yet since my preloadscanner is working in a bit different way.

I didn't find anything about the simple preloadscanner's performance results... Has anybody measured anything?

I'm waiting your comments about the patch!

Thanks in advance,

Z

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