[Webkit-unassigned] [Bug 68320] DFG JIT does not speculate aggressively enough on GetById

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=68320





--- Comment #4 from Filip Pizlo <fpizlo at apple.com>  2011-09-18 13:04:01 PST ---
It's a 4% speed-up on V8.



Benchmark report for V8.

VMs tested:
"TipOfTree" at /Volumes/Data/pizlo/quinary/OpenSource/WebKitBuild/Release/jsc
"GetByOffset" at /Volumes/Data/pizlo/OpenSource/WebKitBuild/Release/jsc

Collected 12 samples per benchmark/VM, with 4 VM invocations per benchmark. Used 1
benchmark iteration per VM invocation for warm-up. Used the jsc-specific preciseTime()
function to get microsecond-level timing. Reporting benchmark execution times with 95%
confidence intervals in milliseconds.

                     TipOfTree              GetByOffset                                   

crypto            83.8914+-1.1391         82.4519+-1.3381         might be 1.0175x faster
deltablue        243.3592+-1.9683    ^   228.2100+-2.6421       ^ definitely 1.0664x faster
earley-boyer      96.1566+-0.2312    ^    92.6819+-0.4199       ^ definitely 1.0375x faster
raytrace          69.7617+-0.6717    ^    63.8253+-0.8824       ^ definitely 1.0930x faster
regexp           106.6008+-0.6734        105.8089+-0.3464       
richards         217.7843+-0.6560    ^   203.0919+-1.3343       ^ definitely 1.0723x faster
splay             98.1221+-0.4274    ?    99.5908+-1.1138       ? might be 1.0150x slower

<arithmetic>     130.8109+-0.2346    ^   125.0944+-0.5873       ^ definitely 1.0457x faster
<geometric>      117.6414+-0.1960    ^   113.1895+-0.5083       ^ definitely 1.0393x faster
<harmonic>       107.7976+-0.2612    ^   103.8596+-0.5277       ^ definitely 1.0379x faster

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