[Webkit-unassigned] [Bug 69114] DFG does not speculate aggressively enough on put_by_id
bugzilla-daemon at webkit.org
bugzilla-daemon at webkit.org
Thu Sep 29 19:20:57 PDT 2011
https://bugs.webkit.org/show_bug.cgi?id=69114
--- Comment #3 from Filip Pizlo <fpizlo at apple.com> 2011-09-29 19:20:57 PST ---
Looks like it's only a win on V8.
Benchmark report for SunSpider, V8, and Kraken.
VMs tested:
"TipOfTree" at /Volumes/Data/pizlo/quinary/OpenSource/WebKitBuild/Release/jsc
"PutByOffset" at /Volumes/Data/pizlo/septenary/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 PutByOffset
SunSpider:
3d-cube 7.4733+-0.2370 ? 7.4909+-0.2434 ?
3d-morph 7.3221+-0.1314 ? 7.4902+-0.1818 ? might be 1.0230x slower
3d-raytrace 8.2079+-0.2675 ? 8.2614+-0.2639 ?
access-binary-trees 2.0855+-0.0733 ^ 1.7651+-0.0578 ^ definitely 1.1816x faster
access-fannkuch 6.5079+-0.1798 6.3323+-0.0819 might be 1.0277x faster
access-nbody 3.6268+-0.0917 3.4877+-0.0829 might be 1.0399x faster
access-nsieve 2.6092+-0.0605 ? 2.6369+-0.0700 ? might be 1.0106x slower
bitops-3bit-bits-in-byte 1.7039+-0.0381 ? 1.7147+-0.0353 ?
bitops-bits-in-byte 2.7511+-0.0767 ? 2.7978+-0.0933 ? might be 1.0170x slower
bitops-bitwise-and 3.4511+-0.0848 ^ 3.2750+-0.0814 ^ definitely 1.0538x faster
bitops-nsieve-bits 5.5331+-0.1009 5.4708+-0.0805 might be 1.0114x faster
controlflow-recursive 2.0846+-0.0539 2.0461+-0.0608 might be 1.0188x faster
crypto-aes 6.6652+-0.2079 ? 6.7579+-0.1919 ? might be 1.0139x slower
crypto-md5 2.8164+-0.0746 ? 2.8615+-0.0950 ? might be 1.0160x slower
crypto-sha1 2.5247+-0.0600 2.4973+-0.0644 might be 1.0110x faster
date-format-tofte 10.0118+-0.2674 ? 10.1345+-0.2735 ? might be 1.0123x slower
date-format-xparb 9.2736+-0.3044 ? 9.5957+-0.3815 ? might be 1.0347x slower
math-cordic 6.5799+-0.2671 6.4915+-0.1929 might be 1.0136x faster
math-partial-sums 7.5752+-0.1633 ? 7.6859+-0.1510 ? might be 1.0146x slower
math-spectral-norm 2.8814+-0.0999 2.8715+-0.0738
regexp-dna 10.8292+-0.1427 ? 10.8850+-0.1618 ?
string-base64 6.0760+-0.2190 ? 6.1591+-0.2329 ? might be 1.0137x slower
string-fasta 7.0525+-0.2171 ? 7.1913+-0.3145 ? might be 1.0197x slower
string-tagcloud 11.8762+-0.3947 ? 11.9610+-0.3092 ?
string-unpack-code 21.3559+-0.4601 21.2620+-0.4328
string-validate-input 6.2368+-0.1467 ? 6.3990+-0.2173 ? might be 1.0260x slower
<arithmetic> 6.3504+-0.0324 ? 6.3662+-0.0315 ?
<geometric> 5.2336+-0.0294 5.2091+-0.0221
<harmonic> 4.3158+-0.0355 ^ 4.2486+-0.0303 ^ definitely 1.0158x faster
TipOfTree PutByOffset
V8:
crypto 71.1752+-0.4499 ! 72.1471+-0.3928 ! definitely 1.0137x slower
deltablue 228.0470+-1.1437 ? 228.2657+-2.3416 ?
earley-boyer 89.8495+-0.6176 ^ 88.1906+-0.5496 ^ definitely 1.0188x faster
raytrace 62.3209+-0.3664 ^ 61.1129+-0.4345 ^ definitely 1.0198x faster
regexp 104.0114+-0.7839 103.5840+-0.6265
richards 199.0179+-1.0726 ^ 187.5831+-1.0648 ^ definitely 1.0610x faster
splay 91.0199+-0.7136 ? 91.4576+-0.6711 ?
<arithmetic> 120.7774+-0.3729 ^ 118.9059+-0.3800 ^ definitely 1.0157x faster
<geometric> 107.9834+-0.3348 ^ 106.7217+-0.2557 ^ definitely 1.0118x faster
<harmonic> 98.2570+-0.3084 ^ 97.3992+-0.2199 ^ definitely 1.0088x faster
TipOfTree PutByOffset
Kraken:
ai-astar 493.2977+-4.6697 491.8601+-3.7826
audio-beat-detection 190.4707+-1.7293 ? 190.7607+-1.5873 ?
audio-dft 280.9297+-2.4491 279.2958+-2.6566
audio-fft 128.3063+-1.0409 ? 129.6684+-1.1003 ? might be 1.0106x slower
audio-oscillator 256.1838+-1.9104 255.2704+-2.2988
imaging-darkroom 422.3094+-1.3344 ^ 418.5479+-1.3490 ^ definitely 1.0090x faster
imaging-desaturate 223.5703+-0.8074 ? 224.3784+-1.7922 ?
imaging-gaussian-blur 582.5605+-2.3726 ? 583.2726+-2.1460 ?
json-parse-financial 48.5278+-0.8188 47.7118+-0.3265 might be 1.0171x faster
json-stringify-tinderbox 68.5631+-0.5214 68.4865+-0.3847
stanford-crypto-aes 130.9508+-1.2565 130.0566+-1.3290
stanford-crypto-ccm 101.1984+-0.9027 ? 102.0392+-0.9587 ?
stanford-crypto-pbkdf2 193.0240+-1.6461 ! 196.9612+-1.7595 ! definitely 1.0204x slower
stanford-crypto-sha256-iterative 83.5403+-0.4886 ! 85.1920+-0.5650 ! definitely 1.0198x slower
<arithmetic> 228.8166+-0.5134 ? 228.8215+-0.4248 ?
<geometric> 178.7126+-0.6271 ? 178.9531+-0.3788 ?
<harmonic> 138.6559+-0.7664 ? 138.7200+-0.3518 ?
TipOfTree PutByOffset
All benchmarks:
<arithmetic> 89.6593+-0.1651 89.3907+-0.1544
<geometric> 23.5139+-0.0709 23.4213+-0.0565
<harmonic> 7.5847+-0.0606 ^ 7.4693+-0.0520 ^ definitely 1.0155x faster
--
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