[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
Wed Sep 28 00:02:25 PDT 2011
https://bugs.webkit.org/show_bug.cgi?id=68593
--- Comment #16 from Filip Pizlo <fpizlo at apple.com> 2011-09-28 00:02:25 PST ---
Updated numbers after merging with https://bugs.webkit.org/show_bug.cgi?id=68580
Conclusion: it's netural.
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 15 samples per benchmark/VM, with 5 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 7.6787+-0.3039 7.4757+-0.2208 might be 1.0271x faster
3d-morph 7.3829+-0.1321 ? 7.4831+-0.1374 ? might be 1.0136x slower
3d-raytrace 8.2394+-0.2402 ? 8.4595+-0.3011 ? might be 1.0267x slower
access-binary-trees 2.0714+-0.0659 ? 2.1538+-0.0820 ? might be 1.0398x slower
access-fannkuch 6.4854+-0.1156 6.4094+-0.1069 might be 1.0119x faster
access-nbody 3.6724+-0.0718 3.6551+-0.0791
access-nsieve 2.6425+-0.0768 2.6057+-0.0687 might be 1.0141x faster
bitops-3bit-bits-in-byte 1.6955+-0.0423 ? 1.7364+-0.0360 ? might be 1.0241x slower
bitops-bits-in-byte 2.7285+-0.0482 ? 2.7671+-0.0680 ? might be 1.0141x slower
bitops-bitwise-and 3.4507+-0.0774 3.3981+-0.0622 might be 1.0155x faster
bitops-nsieve-bits 5.4244+-0.1194 ? 5.4386+-0.1020 ?
controlflow-recursive 2.0497+-0.0537 ? 2.0617+-0.0333 ?
crypto-aes 6.6114+-0.1880 ? 6.6872+-0.1737 ? might be 1.0115x slower
crypto-md5 2.7668+-0.0747 ? 2.8010+-0.0668 ? might be 1.0124x slower
crypto-sha1 2.4589+-0.0671 ? 2.4827+-0.0555 ?
date-format-tofte 10.4541+-0.4384 10.0867+-0.2541 might be 1.0364x faster
date-format-xparb 9.6846+-0.2358 9.6279+-0.2417
math-cordic 6.3572+-0.1142 ? 6.5219+-0.1184 ? might be 1.0259x slower
math-partial-sums 7.5899+-0.1042 7.5543+-0.1343
math-spectral-norm 2.8221+-0.0534 ? 2.8413+-0.0906 ?
regexp-dna 10.9369+-0.2158 10.8010+-0.2099 might be 1.0126x faster
string-base64 6.0534+-0.1700 5.9570+-0.1248 might be 1.0162x faster
string-fasta 6.9345+-0.1592 ? 7.1482+-0.2413 ? might be 1.0308x slower
string-tagcloud 11.8804+-0.2679 11.8612+-0.2276
string-unpack-code 21.3751+-0.4629 ? 21.3774+-0.3379 ?
string-validate-input 6.2531+-0.1936 ? 6.3781+-0.1630 ? might be 1.0200x slower
<arithmetic> 6.3731+-0.0272 ? 6.3758+-0.0352 ?
<geometric> 5.2286+-0.0153 ? 5.2477+-0.0216 ?
<harmonic> 4.2922+-0.0271 ? 4.3243+-0.0223 ?
TipOfTree LiveRangeSplit
V8:
crypto 70.8935+-0.3890 ? 71.2029+-0.2827 ?
deltablue 229.9200+-0.9956 ^ 227.5736+-0.7179 ^ definitely 1.0103x faster
earley-boyer 89.3518+-0.2992 ! 90.3557+-0.1945 ! definitely 1.0112x slower
raytrace 62.5615+-0.4116 ? 62.8549+-0.3808 ?
regexp 104.0308+-0.2646 103.7363+-0.3711
richards 198.6203+-0.7546 ? 198.9464+-0.4766 ?
splay 90.9935+-0.4642 ? 92.3987+-1.6417 ? might be 1.0154x slower
<arithmetic> 120.9102+-0.2689 ? 121.0098+-0.2652 ?
<geometric> 107.9908+-0.2128 ? 108.3575+-0.2639 ?
<harmonic> 98.2147+-0.2025 ! 98.7076+-0.2400 ! definitely 1.0050x slower
TipOfTree LiveRangeSplit
Kraken:
ai-astar 490.2591+-3.2305 ? 490.4662+-2.2753 ?
audio-beat-detection 191.1469+-0.9371 189.4504+-1.0021
audio-dft 278.5623+-2.1923 ? 278.9677+-1.9053 ?
audio-fft 127.7110+-0.9664 ? 127.7116+-0.8813 ?
audio-oscillator 256.5281+-1.6728 ? 267.7998+-12.2863 ? might be 1.0439x slower
imaging-darkroom 419.4694+-1.8256 418.3243+-0.7219
imaging-desaturate 222.7996+-0.6588 ? 223.5182+-0.7306 ?
imaging-gaussian-blur 578.0480+-1.4034 ! 582.3310+-2.1179 ! definitely 1.0074x slower
json-parse-financial 47.6611+-0.1951 ? 47.8559+-0.2437 ?
json-stringify-tinderbox 69.0243+-0.3084 ^ 68.2484+-0.1993 ^ definitely 1.0114x faster
stanford-crypto-aes 130.2373+-1.3560 128.9448+-1.1310 might be 1.0100x faster
stanford-crypto-ccm 102.6255+-0.4921 ^ 99.9854+-0.7414 ^ definitely 1.0264x faster
stanford-crypto-pbkdf2 194.7292+-1.5753 193.4078+-0.6359
stanford-crypto-sha256-iterative 85.3745+-0.4286 84.7869+-0.2958
<arithmetic> 228.1555+-0.4259 ? 228.6999+-0.9140 ?
<geometric> 178.6563+-0.2998 178.4558+-0.5838
<harmonic> 138.6943+-0.2527 138.1404+-0.3229
TipOfTree LiveRangeSplit
All benchmarks:
<arithmetic> 89.4946+-0.1462 ? 89.6731+-0.3029 ?
<geometric> 23.4997+-0.0463 ? 23.5510+-0.0669 ?
<harmonic> 7.5445+-0.0465 ? 7.5992+-0.0383 ?
--
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