119897 – Object properties added using dot syntax (o.f = ...) from code that isn't in eval should be less likely to cause an object to become a dictionary

Filip Pizlo

Reported 2013-08-16 10:18:36 PDT

Patch forthcoming.

Filip Pizlo

Comment 1 2013-08-16 10:20:46 PDT

Created attachment 208930 [details] the patch

WebKit Commit Bot

Comment 2 2013-08-16 10:23:55 PDT

Attachment 208930 [details] did not pass style-queue: Failed to run "['Tools/Scripts/check-webkit-style', '--diff-files', u'LayoutTests/ChangeLog', u'LayoutTests/fast/js/regress/lots-of-fields-expected.txt', u'LayoutTests/fast/js/regress/lots-of-fields.html', u'LayoutTests/fast/js/regress/script-tests/lots-of-fields.js', u'Source/JavaScriptCore/ChangeLog', u'Source/JavaScriptCore/bytecode/CodeBlock.h', u'Source/JavaScriptCore/dfg/DFGOperations.cpp', u'Source/JavaScriptCore/jit/JITStubs.cpp', u'Source/JavaScriptCore/llint/LLIntSlowPaths.cpp', u'Source/JavaScriptCore/runtime/JSObject.h', u'Source/JavaScriptCore/runtime/PutPropertySlot.h', u'Source/JavaScriptCore/runtime/Structure.cpp', u'Source/JavaScriptCore/runtime/Structure.h']" exit_code: 1 Source/JavaScriptCore/bytecode/CodeBlock.h:58: Alphabetical sorting problem. [build/include_order] [4] Total errors found: 1 in 13 files If any of these errors are false positives, please file a bug against check-webkit-style.

Filip Pizlo

Comment 3 2013-08-16 11:20:45 PDT

Complete perf results: Benchmark report for SunSpider, V8Spider, Octane, Kraken, JSBench, JSRegress, and DSP on oldmac (MacPro4,1). VMs tested: "TipOfTree" at /Volumes/Data/pizlo/OpenSource/WebKitBuild/Release/DumpRenderTree (r154157) "LotsOfFields" at /Volumes/Data/fromMiniMe/primary/OpenSource/WebKitBuild/Release/DumpRenderTree (r154157) 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. TipOfTree LotsOfFields SunSpider: 3d-cube 7.6306+-0.2576 ? 7.6367+-0.2239 ? 3d-morph 8.2989+-0.0697 ? 8.3592+-0.0676 ? 3d-raytrace 9.3500+-0.2787 9.2582+-0.2009 access-binary-trees 2.7433+-0.3155 2.7349+-0.3050 access-fannkuch 7.6926+-0.0492 7.6443+-0.0625 access-nbody 4.0687+-0.0629 4.0657+-0.0288 access-nsieve 4.9293+-0.1528 4.8493+-0.0626 might be 1.0165x faster bitops-3bit-bits-in-byte 1.8514+-0.0347 ^ 1.7912+-0.0200 ^ definitely 1.0336x faster bitops-bits-in-byte 6.7498+-0.1207 6.6927+-0.1093 bitops-bitwise-and 2.7585+-0.0520 ? 2.7636+-0.0315 ? bitops-nsieve-bits 4.7127+-0.0472 4.6995+-0.0383 controlflow-recursive 2.9592+-0.0232 2.9437+-0.0232 crypto-aes 5.8872+-0.3061 ? 5.9346+-0.2938 ? crypto-md5 3.2100+-0.0560 3.2083+-0.0556 crypto-sha1 2.9097+-0.0322 ? 2.9694+-0.0510 ? might be 1.0205x slower date-format-tofte 13.3077+-1.0438 ? 13.5622+-1.0428 ? might be 1.0191x slower date-format-xparb 9.2468+-0.6859 ? 9.5744+-0.6891 ? might be 1.0354x slower math-cordic 3.9726+-0.0627 3.9061+-0.0384 might be 1.0170x faster math-partial-sums 10.5248+-0.1104 ? 10.5832+-0.0929 ? math-spectral-norm 2.8657+-0.0278 ? 2.8701+-0.0536 ? regexp-dna 12.9183+-0.5484 12.8464+-0.5488 string-base64 6.4359+-0.5202 6.3592+-0.5208 might be 1.0121x faster string-fasta 12.2263+-0.1732 12.1213+-0.1494 string-tagcloud 14.9299+-0.2430 14.7736+-0.2053 might be 1.0106x faster string-unpack-code 31.3526+-0.3015 ? 31.4395+-0.3153 ? string-validate-input 8.9686+-0.1714 ? 9.0955+-0.2287 ? might be 1.0141x slower <arithmetic> * 7.7885+-0.1381 ? 7.7955+-0.1291 ? might be 1.0009x slower <geometric> 6.2000+-0.1025 6.1948+-0.0940 might be 1.0008x faster <harmonic> 5.0484+-0.0801 5.0312+-0.0714 might be 1.0034x faster TipOfTree LotsOfFields V8Spider: crypto 80.9288+-0.3811 80.8731+-0.3830 deltablue 102.6328+-1.1821 ? 102.6652+-0.6980 ? earley-boyer 71.2676+-0.5727 ? 71.9512+-0.5575 ? raytrace 45.6368+-5.1497 45.5152+-5.1325 regexp 102.8871+-0.4945 ? 103.1333+-1.2980 ? richards 114.7300+-2.4139 113.5802+-0.7325 might be 1.0101x faster splay 59.5613+-3.4098 ? 60.0249+-3.7810 ? <arithmetic> 82.5206+-0.4185 ? 82.5347+-0.6726 ? might be 1.0002x slower <geometric> * 78.6837+-0.8190 ? 78.7581+-0.9845 ? might be 1.0009x slower <harmonic> 74.5392+-1.3711 ? 74.6390+-1.4602 ? might be 1.0013x slower TipOfTree LotsOfFields Octane and V8v7: encrypt 0.47401+-0.00043 0.47333+-0.00032 decrypt 8.54159+-0.02755 ? 8.54717+-0.02089 ? deltablue x2 0.60204+-0.00653 0.59603+-0.00432 might be 1.0101x faster earley 0.90955+-0.00773 0.89925+-0.00551 might be 1.0115x faster boyer 12.63949+-0.04299 ? 12.67616+-0.05574 ? raytrace x2 4.45775+-0.05991 4.43165+-0.06299 regexp x2 33.27820+-0.16605 33.18729+-0.35190 richards x2 0.31100+-0.00493 ? 0.31480+-0.00435 ? might be 1.0122x slower splay x2 0.64825+-0.00743 0.63607+-0.00620 might be 1.0191x faster navier-stokes x2 10.94183+-0.00860 10.94123+-0.01137 closure 0.38217+-0.01832 0.38161+-0.01826 jquery 4.74511+-0.55382 4.71803+-0.55098 gbemu x2 133.04056+-2.35823 ^ 91.91793+-3.10193 ^ definitely 1.4474x faster mandreel x2 195.02732+-0.78930 194.72279+-0.96283 pdfjs x2 99.14485+-0.39014 ? 99.70104+-0.27412 ? box2d x2 34.86907+-0.22845 34.71528+-0.16714 V8v7: <arithmetic> 7.69017+-0.02097 7.67563+-0.04739 might be 1.0019x faster <geometric> * 2.46044+-0.00783 2.45122+-0.00805 might be 1.0038x faster <harmonic> 0.94455+-0.00556 0.94279+-0.00514 might be 1.0019x faster Octane including V8v7: <arithmetic> 40.47437+-0.14514 ^ 37.30861+-0.26548 ^ definitely 1.0849x faster <geometric> * 7.27821+-0.04587 ^ 7.05540+-0.06084 ^ definitely 1.0316x faster <harmonic> 1.30745+-0.01361 1.30462+-0.01384 might be 1.0022x faster TipOfTree LotsOfFields Kraken: ai-astar 492.136+-0.871 ? 492.397+-1.022 ? audio-beat-detection 251.478+-2.523 247.994+-1.891 might be 1.0140x faster audio-dft 312.536+-1.341 ? 314.603+-1.894 ? audio-fft 145.891+-0.301 ? 146.473+-0.538 ? audio-oscillator 248.979+-0.918 248.634+-0.565 imaging-darkroom 294.609+-1.484 ? 295.022+-1.294 ? imaging-desaturate 158.948+-0.219 158.926+-0.263 imaging-gaussian-blur 403.228+-0.272 ? 403.299+-0.398 ? json-parse-financial 79.455+-0.410 ! 81.786+-0.347 ! definitely 1.0293x slower json-stringify-tinderbox 104.001+-0.315 ? 104.249+-0.450 ? stanford-crypto-aes 108.451+-0.900 107.336+-0.803 might be 1.0104x faster stanford-crypto-ccm 125.075+-2.146 ? 125.792+-1.695 ? stanford-crypto-pbkdf2 271.564+-1.119 ^ 268.164+-1.503 ^ definitely 1.0127x faster stanford-crypto-sha256-iterative 127.720+-0.482 127.097+-0.324 <arithmetic> * 223.148+-0.371 222.984+-0.375 might be 1.0007x faster <geometric> 193.488+-0.365 ? 193.579+-0.376 ? might be 1.0005x slower <harmonic> 168.305+-0.370 ? 168.794+-0.359 ? might be 1.0029x slower TipOfTree LotsOfFields JSBench: amazon 10.0833+-0.1834 ? 10.2500+-0.2874 ? might be 1.0165x slower facebook 38.9167+-0.5038 ? 39.0833+-0.5038 ? google 68.0000+-1.1170 67.1667+-1.2659 might be 1.0124x faster twitter 10.1667+-0.2473 ? 10.2500+-0.2874 ? yahoo 3.6667+-0.3128 ? 3.9167+-0.1834 ? might be 1.0682x slower <arithmetic> * 26.1667+-0.3105 26.1333+-0.2122 might be 1.0013x faster <geometric> 15.8078+-0.3429 ? 16.0808+-0.1768 ? might be 1.0173x slower <harmonic> 9.7466+-0.4748 ? 10.1628+-0.2731 ? might be 1.0427x slower TipOfTree LotsOfFields JSRegress: adapt-to-double-divide 22.5474+-0.0508 22.5089+-0.0316 aliased-arguments-getbyval 0.9550+-0.0307 ? 0.9894+-0.0201 ? might be 1.0360x slower allocate-big-object 4.8139+-1.4028 4.4599+-1.4003 might be 1.0794x faster arity-mismatch-inlining 0.8823+-0.0284 0.8820+-0.0199 array-access-polymorphic-structure 8.9643+-1.9976 ? 8.9765+-2.0080 ? array-nonarray-polymorhpic-access 56.3120+-0.1053 ! 56.5248+-0.0705 ! definitely 1.0038x slower array-with-double-add 5.4305+-0.0502 5.4266+-0.0516 array-with-double-increment 4.2583+-0.1396 ? 4.2765+-0.1258 ? array-with-double-mul-add 7.3773+-0.2817 ? 7.3995+-0.2552 ? array-with-double-sum 7.8757+-0.0227 ? 7.8776+-0.0371 ? array-with-int32-add-sub 9.3844+-0.0812 9.3250+-0.0346 array-with-int32-or-double-sum 7.9654+-0.0188 7.9542+-0.0225 ArrayBuffer-Int8Array-alloc-huge-long-lived 220.0331+-2.0901 215.9948+-3.3348 might be 1.0187x faster ArrayBuffer-Int8Array-alloc-large-long-lived 133.3690+-2.5465 ? 133.8109+-1.2729 ? ArrayBuffer-Int8Array-alloc-long-lived-buffer 84.2347+-10.3846 82.7975+-10.9949 might be 1.0174x faster ArrayBuffer-Int8Array-alloc-long-lived 42.5424+-3.8013 42.4116+-3.9430 ArrayBuffer-Int8Array-alloc 38.2171+-4.0754 37.8311+-3.9915 might be 1.0102x faster big-int-mul 5.0521+-0.4907 4.7614+-0.0348 might be 1.0610x faster boolean-test 4.3534+-0.0403 4.3448+-0.0229 branch-fold 4.7262+-0.0366 ? 4.7654+-0.0595 ? cast-int-to-double 14.1235+-0.0566 14.1019+-0.0520 cell-argument 14.8322+-0.3058 14.7910+-0.2806 cfg-simplify 3.8329+-0.0296 3.7887+-0.0254 might be 1.0117x faster cmpeq-obj-to-obj-other 12.5485+-0.1840 ? 12.5865+-0.1602 ? constant-test 8.6841+-0.1461 8.6183+-0.0909 delay-tear-off-arguments-strictmode 3.2889+-0.0652 3.2414+-0.0586 might be 1.0147x faster direct-arguments-getbyval 0.7927+-0.0155 ? 0.8066+-0.0189 ? might be 1.0176x slower double-pollution-getbyval 10.8172+-0.0484 10.7951+-0.0403 double-pollution-putbyoffset 6.6318+-0.6037 6.6069+-0.5972 empty-string-plus-int 13.8492+-0.2098 ? 13.9976+-0.2536 ? might be 1.0107x slower external-arguments-getbyval 2.3743+-0.1374 ? 2.4098+-0.1539 ? might be 1.0149x slower external-arguments-putbyval 3.9155+-0.2764 ? 4.0131+-0.2389 ? might be 1.0249x slower Float32Array-matrix-mult 7.3228+-0.2498 7.2189+-0.2020 might be 1.0144x faster fold-double-to-int 25.3121+-0.1062 ? 25.4200+-0.2714 ? function-dot-apply 3.0168+-0.0262 3.0117+-0.0226 function-test 5.0213+-0.1159 4.8767+-0.1084 might be 1.0297x faster get-by-id-chain-from-try-block 6.6927+-0.1163 ? 6.8459+-0.1081 ? might be 1.0229x slower get-by-id-proto-or-self 22.9548+-0.3922 ? 23.1817+-0.4059 ? get-by-id-self-or-proto 23.2727+-0.4556 23.1579+-0.3773 HashMap-put-get-iterate-keys 44.3111+-0.6032 ? 45.2180+-0.7726 ? might be 1.0205x slower HashMap-put-get-iterate 54.0832+-0.4741 ? 54.6537+-0.3497 ? might be 1.0105x slower HashMap-string-put-get-iterate 64.2464+-0.8872 ? 64.5848+-0.9610 ? imul-double-only 17.4536+-0.0177 ? 17.4631+-0.0213 ? imul-int-only 14.7905+-0.2016 14.6604+-0.0298 imul-mixed 21.9772+-0.0143 ? 21.9872+-0.0207 ? in-four-cases 26.2086+-0.1381 26.1202+-0.0659 in-one-case-false 12.2063+-0.0290 ? 12.2193+-0.0237 ? in-one-case-true 12.1915+-0.0273 ? 12.1936+-0.0159 ? in-two-cases 12.6249+-0.0439 12.6049+-0.0322 indexed-properties-in-objects 4.1663+-0.0217 ? 4.1682+-0.0303 ? inline-arguments-access 1.4529+-0.0294 1.4293+-0.0245 might be 1.0165x faster inline-arguments-local-escape 25.7837+-0.0865 ? 26.8111+-1.6743 ? might be 1.0398x slower inline-get-scoped-var 6.7868+-0.0928 ? 6.8317+-0.0877 ? inlined-put-by-id-transition 14.8572+-0.2129 14.8348+-0.2482 int-or-other-abs-then-get-by-val 9.0660+-0.0176 ? 9.0671+-0.0311 ? int-or-other-abs-zero-then-get-by-val 36.9854+-0.4685 ^ 36.3317+-0.1171 ^ definitely 1.0180x faster int-or-other-add-then-get-by-val 10.1873+-0.0278 ? 10.3339+-0.1347 ? might be 1.0144x slower int-or-other-add 10.6921+-0.1550 10.6638+-0.1581 int-or-other-div-then-get-by-val 6.2118+-0.0221 ? 6.2160+-0.0423 ? int-or-other-max-then-get-by-val 7.0557+-0.0597 ? 7.1113+-0.0789 ? int-or-other-min-then-get-by-val 7.1219+-0.0247 7.1199+-0.0262 int-or-other-mod-then-get-by-val 5.9189+-0.1115 5.8033+-0.0137 might be 1.0199x faster int-or-other-mul-then-get-by-val 6.4896+-0.0692 6.4284+-0.0375 int-or-other-neg-then-get-by-val 7.9098+-0.0497 ? 7.9857+-0.0483 ? int-or-other-neg-zero-then-get-by-val 37.0525+-0.5909 36.2443+-0.5725 might be 1.0223x faster int-or-other-sub-then-get-by-val 10.1715+-0.0349 ? 10.2105+-0.0408 ? int-or-other-sub 8.1607+-0.0550 ? 8.2421+-0.1013 ? int-overflow-local 11.8056+-0.0179 11.7970+-0.0729 Int16Array-bubble-sort 48.2996+-0.2397 48.2615+-0.3273 Int16Array-load-int-mul 1.9894+-0.0248 ? 2.0176+-0.0203 ? might be 1.0142x slower Int32Array-alloc-huge-long-lived 716.7318+-16.1847 716.4501+-15.8666 Int32Array-alloc-huge 818.8742+-7.4920 ? 822.0723+-6.0544 ? Int32Array-alloc-large-long-lived 1023.7960+-4.0836 ? 1027.9814+-6.3803 ? Int32Array-alloc-large 45.3640+-0.8524 44.5212+-0.7116 might be 1.0189x faster Int32Array-alloc-long-lived 190.7977+-3.4005 186.2670+-1.9737 might be 1.0243x faster Int32Array-alloc 17.6830+-3.0530 ? 17.7047+-3.0345 ? Int32Array-Int8Array-view-alloc 25.8518+-2.0160 25.2155+-1.7917 might be 1.0252x faster Int8Array-load 4.8298+-0.0231 ? 4.8368+-0.0336 ? integer-divide 15.3758+-0.0741 ? 15.3889+-0.0352 ? integer-modulo 1.9150+-0.0252 ? 1.9263+-0.0355 ? make-indexed-storage 4.9761+-0.6638 ? 5.1524+-0.6284 ? might be 1.0354x slower make-rope-cse 16.7504+-2.9356 16.6062+-3.0075 marsaglia 521.9096+-1.3320 521.2599+-0.6239 method-on-number 25.0728+-0.5448 24.8846+-0.4745 negative-zero-divide 0.3355+-0.0150 0.3352+-0.0127 negative-zero-modulo 0.3196+-0.0120 ? 0.3251+-0.0146 ? might be 1.0173x slower negative-zero-negate 0.3159+-0.0223 0.3144+-0.0165 nested-function-parsing-random 415.4240+-12.2845 414.7116+-12.3628 nested-function-parsing 57.4897+-3.9657 57.1888+-4.0080 new-array-buffer-dead 3.9233+-0.1033 ? 3.9774+-0.1005 ? might be 1.0138x slower new-array-buffer-push 14.9355+-2.3663 14.8791+-2.4229 new-array-dead 28.4313+-0.0805 28.3617+-0.0265 new-array-push 12.4365+-2.0323 ? 12.4803+-1.9863 ? number-test 4.2544+-0.0296 ? 4.2782+-0.0273 ? object-closure-call 9.5872+-1.4358 9.5515+-1.5489 object-test 4.8037+-0.0674 4.7733+-0.0967 poly-stricteq 90.8852+-0.4190 ? 91.2134+-0.5133 ? polymorphic-structure 21.0515+-0.0983 ? 21.2184+-0.2963 ? polyvariant-monomorphic-get-by-id 12.5030+-0.0106 12.4771+-0.0310 put-by-val-large-index-blank-indexing-type 15.7844+-3.0905 ? 16.0252+-3.1243 ? might be 1.0153x slower rare-osr-exit-on-local 20.0755+-0.0764 ? 20.0890+-0.0489 ? register-pressure-from-osr 31.6079+-0.1742 31.4887+-0.0800 simple-activation-demo 34.6946+-0.2468 34.6633+-0.2257 slow-array-profile-convergence 4.6886+-0.2443 ? 4.7483+-0.2643 ? might be 1.0127x slower slow-convergence 3.6016+-0.0570 3.5927+-0.0322 sparse-conditional 1.3130+-0.0188 ? 1.3410+-0.0387 ? might be 1.0214x slower splice-to-remove 47.7261+-0.2052 47.6646+-0.1707 string-concat-object 5.0712+-1.3919 ? 5.0734+-1.3913 ? string-concat-pair-object 4.9803+-1.4035 4.9752+-1.4026 string-concat-pair-simple 19.5451+-0.7055 ? 19.6131+-0.7365 ? string-concat-simple 19.5682+-0.5615 19.4766+-0.6678 string-cons-repeat 14.9962+-1.0241 14.9879+-1.0036 string-cons-tower 15.6142+-1.0371 ? 15.6652+-0.9777 ? string-equality 43.7321+-0.2138 ? 43.7960+-0.3039 ? string-get-by-val-big-char 17.4107+-0.1908 17.2645+-0.2062 string-get-by-val-out-of-bounds-insane 5.2083+-0.1991 5.1245+-0.1086 might be 1.0164x faster string-get-by-val-out-of-bounds 4.6259+-0.0185 ? 4.6643+-0.0342 ? string-get-by-val 4.1438+-0.0491 ^ 4.0154+-0.0480 ^ definitely 1.0320x faster string-hash 2.6154+-0.0286 2.5983+-0.0280 string-long-ident-equality 38.6480+-0.5328 38.4075+-0.4446 string-repeat-arith 48.1144+-0.3490 ? 48.6016+-0.3459 ? might be 1.0101x slower string-sub 94.0298+-0.6855 ? 94.3448+-1.1631 ? string-test 4.2173+-0.0203 ! 4.3249+-0.0729 ! definitely 1.0255x slower string-var-equality 72.2075+-0.4365 72.0858+-0.4525 structure-hoist-over-transitions 4.0423+-0.5908 ? 4.3110+-0.6219 ? might be 1.0665x slower switch-char-constant 3.3729+-0.0585 3.3463+-0.0191 switch-char 7.9776+-0.0266 ? 7.9892+-0.0304 ? switch-constant 13.0220+-2.6930 12.8386+-2.5725 might be 1.0143x faster switch-string-basic-big-var 27.1417+-1.3070 ? 27.2371+-1.2736 ? switch-string-basic-big 24.7609+-1.1488 24.7441+-1.2537 switch-string-basic-var 21.7910+-0.4766 ? 21.8415+-0.5946 ? switch-string-basic 22.1232+-1.8982 20.7840+-1.5499 might be 1.0644x faster switch-string-big-length-tower-var 29.5607+-0.0488 ? 29.6101+-0.0803 ? switch-string-length-tower-var 22.6528+-0.0179 ! 22.7990+-0.0735 ! definitely 1.0065x slower switch-string-length-tower 17.6489+-0.0504 17.6469+-0.0237 switch-string-short 17.8351+-0.1967 17.7903+-0.1857 switch 18.9431+-3.2131 ? 19.5547+-3.4109 ? might be 1.0323x slower tear-off-arguments-simple 2.3193+-0.0688 2.2684+-0.0379 might be 1.0224x faster tear-off-arguments 3.4846+-0.0300 3.4634+-0.0385 temporal-structure 21.0315+-0.0278 ? 21.1970+-0.2880 ? to-int32-boolean 21.9562+-0.0500 21.9408+-0.0255 undefined-test 4.4995+-0.0596 4.4600+-0.0301 <arithmetic> 44.9426+-0.2209 44.9099+-0.2248 might be 1.0007x faster <geometric> * 12.5672+-0.1730 12.5600+-0.1794 might be 1.0006x faster <harmonic> 4.7927+-0.0860 ? 4.8080+-0.0769 ? might be 1.0032x slower TipOfTree LotsOfFields DSP: filtrr-posterize-tint 43.9497+-1.3647 ? 44.0499+-1.3026 ? filtrr-tint-contrast-sat-bright 71.9355+-1.0284 ? 72.1006+-1.0201 ? filtrr-tint-sat-adj-contr-mult 85.3233+-3.9803 ? 85.5898+-3.9459 ? filtrr-blur-overlay-sat-contr 217.9329+-5.7883 ? 219.3346+-6.0586 ? filtrr-sat-blur-mult-sharpen-contr 259.1763+-2.0209 ? 259.6100+-2.5434 ? filtrr-sepia-bias 33.9087+-3.4755 33.8869+-3.4395 route9-vp8 x5 1081.6394+-7.2687 ? 1082.2983+-8.3244 ? starfield x5 1185.4524+-4.6188 1183.8739+-4.7232 bellard-jslinux x5 2872.4167+-7.4956 2870.8333+-5.4376 zynaps-quake3 x5 1395.9535+-25.0805 1393.9805+-24.5054 zynaps-mandelbrot x5 1108.1667+-5.5310 ? 1110.5154+-7.2420 ? ammojs-asm-js x5 264.6556+-14.4313 262.6845+-14.9480 ammojs-regular-js x5 254.3963+-11.6450 252.1862+-10.1824 <arithmetic> 1012.8202+-2.5780 1012.1081+-2.3668 might be 1.0007x faster <geometric> * 624.4494+-8.1878 623.4205+-7.8186 might be 1.0017x faster <harmonic> 284.6529+-11.2768 284.3369+-11.1313 might be 1.0011x faster TipOfTree LotsOfFields All benchmarks: <arithmetic> 203.0454+-0.3117 202.5921+-0.2989 might be 1.0022x faster <geometric> 25.0281+-0.2339 24.9436+-0.2233 might be 1.0034x faster <harmonic> 4.7014+-0.0538 ? 4.7065+-0.0523 ? might be 1.0011x slower TipOfTree LotsOfFields Geomean of preferred means: <scaled-result> 41.2783+-0.2071 41.0814+-0.1933 might be 1.0048x faster

Filip Pizlo

Comment 4 2013-08-16 11:42:23 PDT

On my laptop, the gbemu speed-up is more pronounced: gbemu x2 95.13075+-2.83926 ^ 57.76286+-2.12455 ^ definitely 1.6469x faster

Filip Pizlo

Comment 5 2013-08-16 12:14:36 PDT

Landed in http://trac.webkit.org/changeset/154199

Attachments
the patch (22.56 KB, patch) 2013-08-16 10:20 PDT, Filip Pizlo	oliver: review+	Details Formatted Diff Diff
View All Add attachment proposed patch, testcase, etc.