[Webkit-unassigned] [Bug 69235] DFG should speculate more aggressively on obvious cases on polymorphic get_by_id
bugzilla-daemon at webkit.org
bugzilla-daemon at webkit.org
Sun Oct 2 21:30:01 PDT 2011
https://bugs.webkit.org/show_bug.cgi?id=69235
--- Comment #1 from Filip Pizlo <fpizlo at apple.com> 2011-10-02 21:30:01 PST ---
Created an attachment (id=109435)
--> (https://bugs.webkit.org/attachment.cgi?id=109435&action=review)
work in progress
Benchmark report for SunSpider, V8, and Kraken.
VMs tested:
"TipOfTree" at /Volumes/Data/pizlo/quinary/OpenSource/WebKitBuild/Release/jsc
"PolyGetById" at /Volumes/Data/pizlo/senary/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 PolyGetById
SunSpider:
3d-cube 7.4802+-0.2051 ? 7.5662+-0.2457 ? might be 1.0115x slower
3d-morph 7.4047+-0.1666 7.4019+-0.1256
3d-raytrace 8.2055+-0.2450 ^ 7.6090+-0.2031 ^ definitely 1.0784x faster
access-binary-trees 1.9520+-0.1161 ^ 1.7650+-0.0517 ^ definitely 1.1060x faster
access-fannkuch 6.3952+-0.1257 6.3839+-0.1022
access-nbody 3.6135+-0.1259 3.5215+-0.0844 might be 1.0261x faster
access-nsieve 2.6273+-0.0694 ? 2.6825+-0.0772 ? might be 1.0210x slower
bitops-3bit-bits-in-byte 1.7307+-0.0258 ? 1.7378+-0.0204 ?
bitops-bits-in-byte 2.7378+-0.0734 2.7376+-0.0666
bitops-bitwise-and 3.1737+-0.0341 ? 3.2256+-0.0921 ? might be 1.0163x slower
bitops-nsieve-bits 5.5207+-0.1615 5.5002+-0.1703
controlflow-recursive 2.0332+-0.0440 2.0238+-0.0508
crypto-aes 6.8061+-0.2144 6.7950+-0.2456
crypto-md5 2.8211+-0.0715 2.7946+-0.1060
crypto-sha1 2.5510+-0.0921 2.5062+-0.0428 might be 1.0179x faster
date-format-tofte 10.0702+-0.2434 9.9412+-0.3018 might be 1.0130x faster
date-format-xparb 9.3993+-0.3214 9.3923+-0.3008
math-cordic 6.3349+-0.1338 ? 6.4220+-0.2151 ? might be 1.0138x slower
math-partial-sums 7.6382+-0.1566 ? 7.7272+-0.1600 ? might be 1.0117x slower
math-spectral-norm 2.8143+-0.0813 ? 2.8189+-0.0596 ?
regexp-dna 10.7468+-0.0974 ? 10.7851+-0.1443 ?
string-base64 5.6549+-0.1856 ? 5.7496+-0.2816 ? might be 1.0168x slower
string-fasta 6.6589+-0.1586 ? 6.8564+-0.1918 ? might be 1.0297x slower
string-tagcloud 11.8219+-0.2949 11.6058+-0.2792 might be 1.0186x faster
string-unpack-code 21.5388+-0.5473 ? 21.7617+-0.7630 ? might be 1.0103x slower
string-validate-input 6.2991+-0.1684 ? 6.4473+-0.2245 ? might be 1.0235x slower
<arithmetic> * 6.3088+-0.0321 6.2984+-0.0335
<geometric> 5.1750+-0.0327 5.1528+-0.0279
<harmonic> 4.2541+-0.0454 4.2174+-0.0399
TipOfTree PolyGetById
V8:
crypto 72.0420+-0.2361 ! 72.6393+-0.2663 ! definitely 1.0083x slower
deltablue 226.0377+-2.2138 ^ 222.4414+-1.3674 ^ definitely 1.0162x faster
earley-boyer 87.5112+-0.3216 ? 88.2521+-0.6321 ?
raytrace 61.6821+-0.5257 ? 62.4898+-0.4579 ? might be 1.0131x slower
regexp 103.2826+-0.5215 103.0298+-0.6185
richards 185.8201+-0.7279 ? 187.5975+-1.3956 ?
splay 90.7084+-0.3449 90.4355+-0.5553
<arithmetic> 118.1549+-0.4235 118.1265+-0.3624
<geometric> * 106.2625+-0.2895 ? 106.5311+-0.2887 ?
<harmonic> 97.1732+-0.2613 ? 97.6353+-0.2647 ?
TipOfTree PolyGetById
Kraken:
ai-astar 487.9140+-1.2525 486.3753+-1.1255
audio-beat-detection 194.2983+-2.8350 191.1177+-1.6340 might be 1.0166x faster
audio-dft 279.9974+-3.1451 ^ 270.8231+-1.9973 ^ definitely 1.0339x faster
audio-fft 128.0543+-0.8223 ? 128.5125+-1.0330 ?
audio-oscillator 257.3560+-1.9724 ^ 246.3795+-2.0638 ^ definitely 1.0446x faster
imaging-darkroom 419.3905+-0.9765 ? 420.3736+-1.5820 ?
imaging-desaturate 223.6224+-0.9294 ? 224.8275+-0.5870 ?
imaging-gaussian-blur 579.8734+-1.3646 ? 580.5274+-1.4157 ?
json-parse-financial 49.2241+-0.2084 ! 50.7065+-0.8707 ! definitely 1.0301x slower
json-stringify-tinderbox 67.5603+-0.4213 ! 68.4149+-0.4177 ! definitely 1.0126x slower
stanford-crypto-aes 129.8032+-1.5508 ? 131.3744+-1.3683 ? might be 1.0121x slower
stanford-crypto-ccm 100.6605+-0.3288 ? 101.6856+-0.7074 ? might be 1.0102x slower
stanford-crypto-pbkdf2 194.6971+-1.2720 194.3023+-0.5494
stanford-crypto-sha256-iterative 76.9877+-0.3741 76.6157+-0.2952
<arithmetic> * 227.8171+-0.6186 ^ 226.5740+-0.4488 ^ definitely 1.0055x faster
<geometric> 177.5602+-0.4815 177.2219+-0.4616
<harmonic> 137.3444+-0.3335 ? 138.1071+-0.5591 ?
TipOfTree PolyGetById
All benchmarks:
<arithmetic> 88.9479+-0.2062 88.5677+-0.1772
<geometric> 23.2671+-0.0911 23.2075+-0.0818
<harmonic> 7.4772+-0.0778 7.4155+-0.0686
TipOfTree PolyGetById
Geomean of preferred means:
<scaled-result> 53.4525+-0.1159 53.3705+-0.1081
--
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