[webkit-reviews] review requested: [Bug 79899] DFG BasicBlocks should not require that their nodes have contiguous indices in the graph : [Attachment 129627] patch updated
bugzilla-daemon at webkit.org
bugzilla-daemon at webkit.org
Wed Feb 29 21:23:50 PST 2012
Yuqiang Xian <yuqiang.xian at intel.com> has asked for review:
Bug 79899: DFG BasicBlocks should not require that their nodes have contiguous
indices in the graph
https://bugs.webkit.org/show_bug.cgi?id=79899
Attachment 129627: patch updated
https://bugs.webkit.org/attachment.cgi?id=129627&action=review
------- Additional Comments from Yuqiang Xian <yuqiang.xian at intel.com>
Correct CSE.
Performance result here:
Collected 12 samples per benchmark/VM, with 4 VM invocations per benchmark.
Emitted a call to gc() between sample
measurements. 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.
ToT 79899
SunSpider:
3d-cube 8.2464+-0.0825 ?
8.8842+-1.3563 ? might be 1.0773x slower
3d-morph 9.7461+-0.1194
9.6008+-0.0471 might be 1.0151x faster
3d-raytrace 10.3699+-0.0746 ?
10.4985+-0.0897 ? might be 1.0124x slower
access-binary-trees 2.4174+-0.0491 ?
2.4208+-0.0299 ?
access-fannkuch 10.2097+-0.0340 !
10.3671+-0.0738 ! definitely 1.0154x slower
access-nbody 5.4205+-0.0502 ?
5.4971+-0.0408 ? might be 1.0141x slower
access-nsieve 4.0139+-0.0354
3.9855+-0.0549
bitops-3bit-bits-in-byte 1.2879+-0.1218
1.2674+-0.0267 might be 1.0162x faster
bitops-bits-in-byte 4.6899+-0.0529 ?
4.6899+-0.0680 ?
bitops-bitwise-and 4.3005+-0.0526
4.2784+-0.0348
bitops-nsieve-bits 4.2527+-0.0555 ?
4.2611+-0.0592 ?
controlflow-recursive 2.8832+-0.0560
2.8495+-0.0330 might be 1.0118x faster
crypto-aes 9.9004+-0.8040
9.7395+-0.1091 might be 1.0165x faster
crypto-md5 3.7363+-0.7201
3.5368+-0.0762 might be 1.0564x faster
crypto-sha1 2.7769+-0.1318 ?
2.8187+-0.0382 ? might be 1.0151x slower
date-format-tofte 12.7713+-0.1900
12.6657+-0.0813
date-format-xparb 11.9215+-0.0611
11.8989+-0.0899
math-cordic 4.0575+-0.0668
4.0442+-0.0729
math-partial-sums 14.8560+-0.0503 ?
14.8976+-0.0510 ?
math-spectral-norm 2.6806+-0.0414
2.6710+-0.0367
regexp-dna 9.4064+-0.0848
9.3769+-0.0829
string-base64 5.5638+-0.0453 ?
5.6208+-0.0781 ? might be 1.0102x slower
string-fasta 9.4595+-0.0784 ?
9.4883+-0.0626 ?
string-tagcloud 16.2971+-0.1989 ?
16.6938+-1.4490 ? might be 1.0243x slower
string-unpack-code 27.5464+-0.1902
27.2705+-0.1486 might be 1.0101x faster
string-validate-input 8.6276+-0.3451
8.4289+-0.0586 might be 1.0236x faster
<arithmetic> * 7.9784+-0.0601 ?
7.9905+-0.0691 ? might be 1.0015x slower
<geometric> 6.3202+-0.0628 ?
6.3234+-0.0358 ? might be 1.0005x slower
<harmonic> 4.9249+-0.0723
4.9247+-0.0266 might be 1.0000x faster
ToT 79899
V8:
crypto 95.0226+-1.0047
94.3739+-0.3426
deltablue 165.2617+-0.6624 ?
165.7057+-0.6259 ?
earley-boyer 120.6456+-2.7215 ?
122.9765+-2.5056 ? might be 1.0193x slower
raytrace 58.9135+-0.3810 ?
59.0790+-0.4627 ?
regexp 111.1890+-0.4246
110.9795+-0.4306
richards 190.9017+-0.7440 ^
189.6743+-0.2815 ^ definitely 1.0065x faster
splay 76.9456+-0.1664
76.8783+-0.1922
<arithmetic> 116.9828+-0.5396 ?
117.0953+-0.4770 ? might be 1.0010x slower
<geometric> * 108.9471+-0.5280 ?
109.0833+-0.4856 ? might be 1.0013x slower
<harmonic> 101.3365+-0.4907 ?
101.4683+-0.4743 ? might be 1.0013x slower
ToT 79899
Kraken:
ai-astar 785.0024+-0.6913 ?
785.6353+-1.8885 ?
audio-beat-detection 237.7250+-1.4305 ?
238.6727+-0.6064 ?
audio-dft 368.7606+-7.9661
366.7011+-3.3630
audio-fft 152.0918+-1.0270
151.3073+-0.0661
audio-oscillator 343.2595+-1.3591 ?
344.1854+-1.4707 ?
imaging-darkroom 375.6221+-9.2659
374.0356+-8.6721
imaging-desaturate 303.7351+-0.8658
303.5210+-0.5746
imaging-gaussian-blur 506.6966+-1.0673
506.6839+-0.5188
json-parse-financial 86.3672+-0.7082 ?
87.4680+-0.3961 ? might be 1.0127x slower
json-stringify-tinderbox 106.0603+-0.8037
106.0054+-0.6595
stanford-crypto-aes 106.1432+-0.8878
105.0881+-0.3653 might be 1.0100x faster
stanford-crypto-ccm 99.5565+-0.6853 ?
100.0071+-0.7250 ?
stanford-crypto-pbkdf2 235.4754+-0.3006 !
238.4268+-1.2700 ! definitely 1.0125x slower
stanford-crypto-sha256-iterative 102.9337+-0.3225
102.6497+-0.5483
<arithmetic> * 272.1021+-1.1641 ?
272.1705+-0.9473 ? might be 1.0003x slower
<geometric> 216.1712+-0.8053 ?
216.3056+-0.6672 ? might be 1.0006x slower
<harmonic> 174.8804+-0.5656 ?
175.0663+-0.5010 ? might be 1.0011x slower
More information about the webkit-reviews
mailing list