[Webkit-unassigned] [Bug 68593] DFG JIT should infer which uses of a variable are not aliased
bugzilla-daemon at webkit.org
bugzilla-daemon at webkit.org
Tue Sep 27 22:28:56 PDT 2011
https://bugs.webkit.org/show_bug.cgi?id=68593
--- Comment #13 from Filip Pizlo <fpizlo at apple.com> 2011-09-27 22:28:56 PST ---
Fixed some stuff. Now it's totally neutral.
Benchmark report for SunSpider, V8, and Kraken.
VMs tested:
"TipOfTree" at /Volumes/Data/pizlo/quinary/OpenSource/WebKitBuild/Release/jsc
"LiveRangeSplit" 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 LiveRangeSplit
SunSpider:
3d-cube 7.6922+-0.1988 ? 7.7205+-0.2574 ?
3d-morph 7.5029+-0.1592 7.4454+-0.2336
3d-raytrace 8.2500+-0.2146 ? 8.4582+-0.2255 ? might be 1.0252x slower
access-binary-trees 2.1528+-0.1100 2.1172+-0.1243 might be 1.0168x faster
access-fannkuch 6.2283+-0.0899 ? 6.4395+-0.1506 ? might be 1.0339x slower
access-nbody 3.6747+-0.0747 ? 3.6900+-0.1150 ?
access-nsieve 2.5620+-0.0497 ? 2.6403+-0.0766 ? might be 1.0306x slower
bitops-3bit-bits-in-byte 1.7323+-0.0467 1.7322+-0.0522
bitops-bits-in-byte 2.7092+-0.0460 ? 2.7689+-0.0710 ? might be 1.0221x slower
bitops-bitwise-and 3.5607+-0.1322 3.4690+-0.0751 might be 1.0264x faster
bitops-nsieve-bits 5.5097+-0.1361 5.4230+-0.1047 might be 1.0160x faster
controlflow-recursive 2.0410+-0.0514 ? 2.0848+-0.0459 ? might be 1.0215x slower
crypto-aes 6.3633+-0.1242 ? 6.4963+-0.2182 ? might be 1.0209x slower
crypto-md5 2.8494+-0.1204 2.7613+-0.0634 might be 1.0319x faster
crypto-sha1 2.5110+-0.0591 2.5104+-0.0746
date-format-tofte 10.0627+-0.3360 ? 10.2101+-0.3020 ? might be 1.0147x slower
date-format-xparb 9.5168+-0.1626 ? 9.6750+-0.2667 ? might be 1.0166x slower
math-cordic 6.3692+-0.0948 6.3406+-0.1316
math-partial-sums 8.0563+-0.1988 7.8165+-0.2004 might be 1.0307x faster
math-spectral-norm 2.9357+-0.1066 2.8295+-0.0678 might be 1.0375x faster
regexp-dna 10.8642+-0.2494 ? 11.0709+-0.2275 ? might be 1.0190x slower
string-base64 5.9855+-0.3214 5.7881+-0.1345 might be 1.0341x faster
string-fasta 7.1136+-0.1528 ? 7.2783+-0.2611 ? might be 1.0231x slower
string-tagcloud 11.7383+-0.2994 ? 12.2077+-0.3409 ? might be 1.0400x slower
string-unpack-code 21.5002+-0.7005 ? 21.5974+-0.3771 ?
string-validate-input 6.5260+-0.2503 6.3355+-0.1992 might be 1.0301x faster
<arithmetic> 6.3849+-0.0276 ? 6.4195+-0.0282 ?
<geometric> 5.2580+-0.0177 ? 5.2671+-0.0246 ?
<harmonic> 4.3333+-0.0238 4.3310+-0.0326
TipOfTree LiveRangeSplit
V8:
crypto 71.7616+-1.0628 71.1205+-0.3344
deltablue 228.1658+-0.7650 ? 231.5297+-2.8410 ? might be 1.0147x slower
earley-boyer 89.8243+-0.5228 ? 89.8258+-0.5621 ?
raytrace 62.2547+-0.7894 ? 63.6586+-0.9188 ? might be 1.0226x slower
regexp 106.3893+-1.0650 105.9907+-1.6466
richards 198.1538+-0.8433 ! 200.9998+-1.5931 ! definitely 1.0144x slower
splay 96.5465+-0.8717 ^ 95.0097+-0.4065 ^ definitely 1.0162x faster
<arithmetic> 121.8709+-0.3295 ? 122.5907+-0.6098 ?
<geometric> 109.2908+-0.4012 ? 109.6402+-0.4618 ?
<harmonic> 99.5289+-0.4996 ? 99.7573+-0.4226 ?
TipOfTree LiveRangeSplit
Kraken:
ai-astar 499.2630+-5.3220 495.1316+-4.8329
audio-beat-detection 207.5893+-2.3626 205.4705+-1.9525 might be 1.0103x faster
audio-dft 425.4155+-5.1167 ? 439.0691+-16.9925 ? might be 1.0321x slower
audio-fft 139.5907+-0.8042 ? 141.1493+-1.4043 ? might be 1.0112x slower
audio-oscillator 257.4750+-3.5507 ? 257.8873+-2.2155 ?
imaging-darkroom 440.2766+-4.1131 ^ 426.5286+-6.2953 ^ definitely 1.0322x faster
imaging-desaturate 224.8910+-1.3828 ? 228.9449+-7.2459 ? might be 1.0180x slower
imaging-gaussian-blur 586.4749+-4.4782 586.4513+-4.4580
json-parse-financial 48.5181+-0.6663 ? 48.7977+-0.7105 ?
json-stringify-tinderbox 68.7911+-0.6904 ^ 67.5631+-0.5023 ^ definitely 1.0182x faster
stanford-crypto-aes 134.0193+-1.6182 ? 135.1391+-0.7355 ?
stanford-crypto-ccm 105.1607+-0.7178 ^ 103.1710+-0.7521 ^ definitely 1.0193x faster
stanford-crypto-pbkdf2 202.8855+-0.9739 ^ 200.2139+-0.8190 ^ definitely 1.0133x faster
stanford-crypto-sha256-iterative 84.8164+-0.9626 ? 85.4529+-1.3834 ?
<arithmetic> 244.6548+-0.6791 244.3550+-1.4587
<geometric> 189.0306+-0.6238 188.7703+-0.9134
<harmonic> 143.5915+-0.7115 143.3510+-0.7654
TipOfTree LiveRangeSplit
All benchmarks:
<arithmetic> 94.5590+-0.2062 ? 94.5960+-0.3798 ?
<geometric> 24.0152+-0.0550 ? 24.0395+-0.0798 ?
<harmonic> 7.6202+-0.0407 7.6162+-0.0558
--
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