PerformanceTests/ChangeLog

   1 2012-07-24  Ryosuke Niwa  <rniwa@webkit.org>
   2
   3         Add some Japanese and Swedish websites to replay tests
   4         https://bugs.webkit.org/show_bug.cgi?id=92076
   5
   6         Reviewed by Andreas Kling.
   7
   8         Added new replay tests as follows.
   9
  10         * Replay/Japanese: Added.
  11         * Replay/Japanese/2ch.net-newsplus.replay: Added.
  12         * Replay/Japanese/ja.wikipedia.org.replay: Added.
  13         * Replay/Japanese/www.livedoor.com.replay: Added.
  14         * Replay/Japanese/www.rakuten.co.jp.replay: Added.
  15         * Replay/Japanese/www.yahoo.co.jp.replay: Added.
  16         * Replay/Swedish: Added.
  17         * Replay/Swedish/www.flashback.se.replay: Added.
  18         * Replay/Swedish/www.tradera.com.replay: Added.
  19         * Replay/www.techcrunch.com.replay: Added.
  20
  21 2012-07-18  MORITA Hajime  <morrita@google.com>
  22
  23         Add window resize benchmark
  24         https://bugs.webkit.org/show_bug.cgi?id=91070
  25
  26         Reviewed by Ryosuke Niwa.
  27
  28         Added window-resize.html which exercise html5.html. This test
  29         resizes the window and forces re-layout repeatedly.
  30
  31         * Interactive/resources/window-resize.js: Added.
  32         * Interactive/window-resize.html: Added.
  33         * Parser/resources/html5.html: Modified to inject driver scripts.
  34
  35 2012-07-11  Philip Rogers  <pdr@google.com>
  36
  37         Add a performance test for hit testing in SVG
  38         https://bugs.webkit.org/show_bug.cgi?id=90811
  39
  40         Reviewed by Ryosuke Niwa.
  41
  42         The shape rendering code is changing a lot and I'd like a performance test to
  43         make sure we don't regress, and to track our improvements.
  44
  45         Performance results in my Linux desktop:
  46             RESULT SVG: SvgHitTesting= 105.0 ms
  47             median= 103.5 ms, stdev= 4.60434577329 ms, min= 101.0 ms, max= 116.0 ms
  48
  49             RESULT SVG: SvgHitTesting= 97.8 ms
  50             median= 96.5 ms, stdev= 2.67581763205 ms, min= 95.0 ms, max= 102.0 ms
  51
  52             RESULT SVG: SvgHitTesting= 104.3 ms
  53             median= 104.0 ms, stdev= 1.41774468788 ms, min= 102.0 ms, max= 107.0 ms
  54
  55             RESULT SVG: SvgHitTesting= 103.6 ms
  56             median= 103.5 ms, stdev= 1.2 ms, min= 102.0 ms, max= 106.0 ms
  57
  58         * SVG/SvgHitTesting.html: Added.
  59
  60 2012-07-03  Jan Keromnes  <janx@linux.com>
  61
  62         Web Inspector: WebInspector.TextViewer should be renamed WebInspector.TextEditor
  63         https://bugs.webkit.org/show_bug.cgi?id=89939
  64
  65         Reviewed by Vsevolod Vlasov.
  66
  67         * inspector/first-open-scripts.html.broken:
  68
  69 2012-06-22  Philip Rogers  <pdr@google.com>
  70
  71         Make SvgCubics performance test more consistent
  72         https://bugs.webkit.org/show_bug.cgi?id=89778
  73
  74         Reviewed by Ryosuke Niwa.
  75
  76         Previously this test ran twice for each iteration, but this led to inconsistent repaints.
  77         This change updates the test to only run once per iteration, and adds even more cubics
  78         so that the test runs consistently in the 80ms range on my Linux desktop.
  79
  80         Performance results on my desktop:
  81             RESULT SVG: SvgCubics= 80.5 ms
  82             median= 82.0 ms, stdev= 3.04138126515 ms, min= 75.0 ms, max= 85.0 ms
  83
  84             RESULT SVG: SvgCubics= 80.1 ms
  85             median= 81.0 ms, stdev= 4.27668095607 ms, min= 74.0 ms, max= 88.0 ms
  86
  87             RESULT SVG: SvgCubics= 81.45 ms
  88             median= 83.0 ms, stdev= 4.01839520207 ms, min= 75.0 ms, max= 89.0 ms
  89
  90             RESULT SVG: SvgCubics= 78.9 ms
  91             median= 80.0 ms, stdev= 4.75289385533 ms, min= 72.0 ms, max= 87.0 ms
  92
  93         * SVG/SvgCubics.html:
  94
  95 2012-06-22  Ilya Tikhonovsky  <loislo@chromium.org>
  96
  97         Web Inspector: partially instrument DOM Tree native memory.
  98         https://bugs.webkit.org/show_bug.cgi?id=89568
  99
 100         This patch adds MemoryInstrumentation class that counts all visited
 101         objects and calls reportMemoryUsage.
 102
 103         Reviewed by Yury Semikhatsky.
 104
 105         * inspector/native-memory-snapshot.html:
 106
 107 2012-06-21  Kentaro Hara  <haraken@chromium.org>
 108
 109         Add a perf-test for innerHTML setter for a large DOM tree
 110         https://bugs.webkit.org/show_bug.cgi?id=89723
 111
 112         Reviewed by Ryosuke Niwa.
 113
 114         We want a benchmark for innerHTML setter for the following reason:
 115
 116         - innerHTML setter is widely used in the real world.
 117         - I am planning to optimize innerHTML setter in the near future.
 118         - I want to use the innerHTML setter benchmark for the patch of bug 88834.
 119
 120         Performance results in my Linux desktop:
 121
 122         RESULT Parser: innerHTML-setter= 289.782649995 runs/s
 123         median= 290.046269741 runs/s, stdev= 1.06575112224 runs/s, min= 286.831812256 runs/s, max= 291.005291005 runs/s
 124
 125         RESULT Parser: innerHTML-setter= 289.020706132 runs/s
 126         median= 289.093298292 runs/s, stdev= 0.985203313093 runs/s, min= 286.831812256 runs/s, max= 290.620871863 runs/s
 127
 128         RESULT Parser: innerHTML-setter= 288.912051701 runs/s
 129         median= 291.005291005 runs/s, stdev= 3.65241325588 runs/s, min= 283.505154639 runs/s, max= 292.553191489 runs/s
 130
 131         RESULT Parser: innerHTML-setter= 288.644186666 runs/s
 132         median= 288.713910761 runs/s, stdev= 1.31889053717 runs/s, min= 286.085825748 runs/s, max= 290.620871863 runs/s
 133
 134         RESULT Parser: innerHTML-setter= 288.698714577 runs/s
 135         median= 288.713910761 runs/s, stdev= 1.03938198202 runs/s, min= 286.458333333 runs/s, max= 290.237467018 runs/s
 136
 137         * Parser/innerHTML-setter.html: Added.
 138
 139 2012-06-20  Philip Rogers  <pdr@google.com>
 140
 141         Add a performance test for paths in SVG
 142         https://bugs.webkit.org/show_bug.cgi?id=89547
 143
 144         Reviewed by Ryosuke Niwa.
 145
 146         This change adds the first performance test for SVG paths.
 147         In the test we modify complex cubic paths in several ways, testing:
 148         transformations, clipping, d attribute changes, stroke properties,
 149         text on a path, and opacity.
 150
 151         Sample test results on my Linux desktop:
 152             RESULT SVG: SvgCubics= 68.85 ms
 153             median= 69.5 ms, stdev= 4.70398767005 ms, min= 56.0 ms, max= 75.0 ms
 154
 155             RESULT SVG: SvgCubics= 68.4 ms
 156             median= 69.0 ms, stdev= 3.51283361405 ms, min= 59.0 ms, max= 74.0 ms
 157
 158             RESULT SVG: SvgCubics= 66.95 ms
 159             median= 67.0 ms, stdev= 3.4420197559 ms, min= 59.0 ms, max= 74.0 ms
 160
 161             RESULT SVG: SvgCubics= 70.2 ms
 162             median= 71.5 ms, stdev= 3.23419232576 ms, min= 63.0 ms, max= 74.0 ms
 163
 164         * SVG: Added.
 165         * SVG/SvgCubics.html: Added.
 166
 167 2012-06-20  Tom Zakrajsek  <tomz@codeaurora.org>
 168
 169         Import themaninblue.com/experiment/AnimationBenchmark/ as performance tests
 170         https://bugs.webkit.org/show_bug.cgi?id=78789
 171
 172         Reviewed by Ryosuke Niwa.
 173
 174         Added PerfTestRunner fixture around it.
 175
 176         * Animation/balls.html: Added.
 177
 178 2012-06-18  Ilya Tikhonovsky  <loislo@chromium.org>
 179
 180         Web Inspector: native memory snapshot performance and coverage test implementation.
 181         https://bugs.webkit.org/show_bug.cgi?id=89363
 182
 183         We have to track the performance of MemoryAgent.getProcessMemoryDistribution.
 184         Also this test will work as a burn down chart for Unknown memory metric.
 185
 186         Reviewed by Pavel Feldman.
 187
 188         * inspector/native-memory-snapshot.html: Added.
 189         * inspector/performance-test.js:
 190         (initialize_TimeTracker.InspectorTest.runPerformanceTest.Timer):
 191         (initialize_TimeTracker.InspectorTest.runPerformanceTest.Timer.prototype.reportSize):
 192         (initialize_TimeTracker.InspectorTest.runPerformanceTest.Timer.prototype._dump):
 193         (initialize_TimeTracker.InspectorTest.runPerformanceTest):
 194
 195 2012-06-05  Alexei Filippov  <alexeif@chromium.org>
 196
 197         Web Inspector: serialize edge counts instead of indexes in heap snapshot
 198         https://bugs.webkit.org/show_bug.cgi?id=88324
 199
 200         The serialized node structure currently holds an index
 201         of its first containment edge in the edges array.
 202         The index can be quite big (up to 7 digits for large snapshots).
 203         The patch changes the serialization format to pass
 204         node containment edge count instead. For most nodes the count
 205         is just a single digit number.
 206         This reduces serialized snapshot size and therefore its transfer time.
 207
 208         Reviewed by Yury Semikhatsky.
 209
 210         * inspector/heap-snapshot-performance-test.js:
 211
 212 2012-06-04  Alexei Filippov  <alexeif@chromium.org>
 213
 214         Web Inspector: speed up _calculateRetainedSizes function
 215         https://bugs.webkit.org/show_bug.cgi?id=87863
 216
 217         Reviewed by Yury Semikhatsky.
 218
 219         * inspector/heap-snapshot-advanced.html: Added.
 220         * inspector/heap-snapshot-performance-test.js: Copied from PerformanceTests/inspector/heap-snapshot.html.
 221         (test.performanceTest.step0):
 222         (test.performanceTest.step1):
 223         (test.performanceTest.step2):
 224         (test.performanceTest.cleanup):
 225         (test.performanceTest.done):
 226         (test.performanceTest):
 227         (test):
 228         * inspector/heap-snapshot.html:
 229
 230 2012-06-03  Kentaro Hara  <haraken@chromium.org>
 231
 232         [perf-test] Add a benchmark of querySelector() for an element that appears in the depths in the document
 233         https://bugs.webkit.org/show_bug.cgi?id=88202
 234
 235         Reviewed by Ryosuke Niwa.
 236
 237         query-selector-first.html tests querySelector() for an element
 238         that appears at the head of the document. query-selector-last.html
 239         tests querySelector() for an element that appears at the tail of
 240         the document. In addition, we want a benchmark of querySelector()
 241         for an element that appears in the depths in the document.
 242
 243         Test results in my Linux desktop:
 244
 245         RESULT Parser: query-selector-deep= 460.776980611 runs/s
 246         median= 460.947503201 runs/s, stdev= 1.72274941942 runs/s, min= 455.696202532 runs/s, max= 463.32046332 runs/s
 247
 248         RESULT Parser: query-selector-deep= 458.925050915 runs/s
 249         median= 459.183673469 runs/s, stdev= 1.47930124139 runs/s, min= 454.545454545 runs/s, max= 460.947503201 runs/s
 250
 251         RESULT Parser: query-selector-deep= 461.866981491 runs/s
 252         median= 461.538461538 runs/s, stdev= 1.14613123359 runs/s, min= 459.183673469 runs/s, max= 463.917525773 runs/s
 253
 254         RESULT Parser: query-selector-deep= 453.922010788 runs/s
 255         median= 454.258855898 runs/s, stdev= 1.77279123317 runs/s, min= 447.204968944 runs/s, max= 455.696202532 runs/s
 256
 257         RESULT Parser: query-selector-deep= 454.666321221 runs/s
 258         median= 454.545454545 runs/s, stdev= 1.65775667417 runs/s, min= 451.127819549 runs/s, max= 456.852791878 runs/s
 259
 260         * Parser/query-selector-deep.html: Added.
 261
 262 2012-06-03  Kentaro Hara  <haraken@chromium.org>
 263
 264         [perf-test] Make query-selector-last.html more realistic
 265         https://bugs.webkit.org/show_bug.cgi?id=88203
 266
 267         Reviewed by Ryosuke Niwa.
 268
 269         query-selector-last.html tests querySelector() for an element
 270         that appears after 1000 div elements, all of which have the same
 271         id and class. To test the performance of querySelector() under
 272         a lot of ids and classes, we should make the ids and classes of
 273         the 1000 elements different.
 274
 275         This patch degrades the performance of query-selector-last.html
 276         by 3.6%.
 277
 278         [Before]
 279         RESULT Parser: query-selector-last= 476.382274152 runs/s
 280         median= 476.821192053 runs/s, stdev= 1.14578875814 runs/s, min= 474.308300395 runs/s, max= 478.087649402 runs/s
 281
 282         RESULT Parser: query-selector-last= 478.423061861 runs/s
 283         median= 478.723404255 runs/s, stdev= 1.54876867255 runs/s, min= 473.684210526 runs/s, max= 481.347773767 runs/s
 284
 285         RESULT Parser: query-selector-last= 479.849287174 runs/s
 286         median= 480.0 runs/s, stdev= 1.69849752498 runs/s, min= 475.561426684 runs/s, max= 481.927710843 runs/s
 287
 288         RESULT Parser: query-selector-last= 476.765438846 runs/s
 289         median= 477.453580902 runs/s, stdev= 1.8724488664 runs/s, min= 471.821756225 runs/s, max= 478.723404255 runs/s
 290
 291         RESULT Parser: query-selector-last= 481.715340644 runs/s
 292         median= 481.637742305 runs/s, stdev= 1.81446361145 runs/s, min= 475.561426684 runs/s, max= 483.675937122 runs/s
 293
 294         [After]
 295         RESULT Parser: query-selector-last= 458.866623582 runs/s
 296         median= 459.183673469 runs/s, stdev= 1.48881464737 runs/s, min= 453.972257251 runs/s, max= 460.947503201 runs/s
 297
 298         RESULT Parser: query-selector-last= 460.363532897 runs/s
 299         median= 460.358056266 runs/s, stdev= 1.58512056995 runs/s, min= 456.273764259 runs/s, max= 462.130937099 runs/s
 300
 301         RESULT Parser: query-selector-last= 457.727448451 runs/s
 302         median= 458.015267176 runs/s, stdev= 1.23681915516 runs/s, min= 453.972257251 runs/s, max= 459.183673469 runs/s
 303
 304         RESULT Parser: query-selector-last= 465.57336853 runs/s
 305         median= 465.717981889 runs/s, stdev= 1.6616925374 runs/s, min= 460.947503201 runs/s, max= 467.532467532 runs/s
 306
 307         RESULT Parser: query-selector-last= 459.836252019 runs/s
 308         median= 459.770114943 runs/s, stdev= 1.84485630505 runs/s, min= 455.696202532 runs/s, max= 462.724935733 runs/s
 309
 310         * Parser/query-selector-last.html:
 311
 312 2012-06-01  Ryosuke Niwa  <rniwa@webkit.org>
 313
 314         Add public page loading performance tests using web-page-replay
 315         https://bugs.webkit.org/show_bug.cgi?id=84008
 316
 317         Reviewed by Dirk Pranke.
 318
 319         Add replay tests for google.com and youtube.com as examples.
 320
 321         * Replay: Added.
 322         * Replay/www.google.com.replay: Added.
 323         * Replay/www.youtube.com.replay: Added.
 324
 325 2012-05-30  Kentaro Hara  <haraken@chromium.org>
 326
 327         [perf-test] Add a benchmark for querySelector()
 328         https://bugs.webkit.org/show_bug.cgi?id=87742
 329
 330         Reviewed by Ryosuke Niwa.
 331
 332         Now I am optimizing querySelector() (bug 87625).
 333         This patch adds a performance test for querySelector().
 334         query-selector-first.html queries an element that appears
 335         at the head of the document. On the other hand,
 336         query-selector-last.html queries an element that appears
 337         at the tail of the document.
 338
 339         Test results in my desktop:
 340
 341         [query-selector-first.html]
 342         RESULT Parser: query-selector-first= 252.339803014 runs/s
 343         median= 253.646265594 runs/s, stdev= 2.32610873209 runs/s, min= 248.138957816 runs/s, max= 254.777070064 runs/s
 344
 345         RESULT Parser: query-selector-first= 252.688468897 runs/s
 346         median= 252.684876579 runs/s, stdev= 3.60008778708 runs/s, min= 246.002460025 runs/s, max= 258.397932817 runs/s
 347
 348         RESULT Parser: query-selector-first= 253.466019656 runs/s
 349         median= 254.129606099 runs/s, stdev= 1.78002675664 runs/s, min= 247.831474597 runs/s, max= 254.777070064 runs/s
 350
 351         RESULT Parser: query-selector-first= 256.245078189 runs/s
 352         median= 257.06940874 runs/s, stdev= 2.84722072201 runs/s, min= 249.687890137 runs/s, max= 259.74025974 runs/s
 353
 354         RESULT Parser: query-selector-first= 252.203100497 runs/s
 355         median= 252.047989175 runs/s, stdev= 1.75233446036 runs/s, min= 248.138957816 runs/s, max= 255.427841635 runs/s
 356
 357         [query-selector-last.html]
 358         RESULT Parser: query-selector-last= 356.009616076 runs/s
 359         median= 355.781448539 runs/s, stdev= 0.841519086864 runs/s, min= 354.430379747 runs/s, max= 358.056265985 runs/s
 360
 361         RESULT Parser: query-selector-last= 360.735271001 runs/s
 362         median= 361.057532424 runs/s, stdev= 1.12253219867 runs/s, min= 358.514724712 runs/s, max= 362.225097025 runs/s
 363
 364         RESULT Parser: query-selector-last= 359.598592463 runs/s
 365         median= 359.435173299 runs/s, stdev= 0.816538258728 runs/s, min= 358.056265985 runs/s, max= 360.824742268 runs/s
 366
 367         RESULT Parser: query-selector-last= 357.260651715 runs/s
 368         median= 357.598978289 runs/s, stdev= 1.17393159885 runs/s, min= 353.982300885 runs/s, max= 358.974358974 runs/s
 369
 370         RESULT Parser: query-selector-last= 351.696240713 runs/s
 371         median= 351.980025916 runs/s, stdev= 1.14200564038 runs/s, min= 348.692403487 runs/s, max= 353.535353535 runs/s
 372
 373         * Parser/query-selector-first.html: Added.
 374         * Parser/query-selector-last.html: Added.
 375
 376 2012-05-25  Ilya Tikhonovsky  <loislo@chromium.org>
 377
 378         Web Inspector: HeapSnapshot: introduce performance counter for HeapSnapshotConstructorsDataGrid._aggregatesReceived method.
 379         https://bugs.webkit.org/show_bug.cgi?id=87393
 380
 381         Reviewed by Yury Semikhatsky.
 382
 383         * inspector/heap-snapshot.html:
 384
 385 2012-05-25  Ilya Tikhonovsky  <loislo@chromium.org>
 386
 387         Web Inspector: HeapProfiler's perf test: each N-th run pushes the run results N-th times.
 388         https://bugs.webkit.org/show_bug.cgi?id=87476
 389
 390         The time tracking wrappers were attaching on each run.
 391
 392         Reviewed by Yury Semikhatsky.
 393
 394         * inspector/heap-snapshot.html:
 395
 396 2012-05-22  Ilya Tikhonovsky  <loislo@chromium.org>
 397
 398         Web Inspector: HeapProfiler: upstream retainedSize calculation.
 399         https://bugs.webkit.org/show_bug.cgi?id=87107
 400
 401         This patch upstreams V8::HeapSnapshotGenerator::CalculateRetainedSizes function to front-end.
 402         After that we will be able to drop retainedSize field from the snapshot serialized data.
 403         See meta-bug https://bugs.webkit.org/show_bug.cgi?id=87089
 404
 405         Reviewed by Yury Semikhatsky.
 406
 407         * inspector/heap-snapshot.html: add counter for the _calculateRetainedSize method.
 408
 409 2012-05-21  Kent Tamura  <tkent@chromium.org>
 410
 411         Add performance tests for <textarea>
 412         https://bugs.webkit.org/show_bug.cgi?id=86994
 413
 414         Reviewed by Ryosuke Niwa.
 415
 416         * DOM/textarea-dom.html: Added.
 417         * DOM/textarea-edit.html: Added.
 418         * Parser/textarea-parsing.html: Added.
 419
 420 2012-05-18  Ilya Tikhonovsky  <loislo@chromium.org>
 421
 422         Web Inspector: upstream build dominators tree procedure from v8.
 423         https://bugs.webkit.org/show_bug.cgi?id=86640
 424
 425         The idea is to reduce transfer size and move all the post-processing steps to the front-end.
 426         The JS implementation is ~1.5 times slower.
 427
 428         Reviewed by Yury Semikhatsky.
 429
 430         Covered by existing tests and performance tests.
 431
 432         * inspector/heap-snapshot.html:
 433
 434 2012-05-18  Kentaro Hara  <haraken@chromium.org>
 435
 436         [perf-test] Remove Bindings/dom-attributes.html
 437         https://bugs.webkit.org/show_bug.cgi?id=86820
 438
 439         Reviewed by Ryosuke Niwa.
 440
 441         Bindings/dom-attributes.html is just a collection of random
 442         DOM attributes and methods. Now we've added more reasonable
 443         micro benchmarks by classifying DOM binding call paths, we
 444         can remove Bindings/dom-attributes.html.
 445
 446         * Bindings/dom-attributes.html: Removed.
 447
 448 2012-05-18  Kentaro Hara  <haraken@chromium.org>
 449
 450         [perf-test] Add a binding benchmark for getElementById() for an undefined id
 451         https://bugs.webkit.org/show_bug.cgi?id=86812
 452
 453         Reviewed by Ryosuke Niwa.
 454
 455         We should remove Bindings/dom-attributes.html and instead add more
 456         reasonable micro benchmarks by classifying DOM binding call paths.
 457
 458         This patch adds a benchmark for document.getElementById() for an undefined id.
 459         This benchmark covers 'getElementById (not in document)' in Dromaeo/dom-query.html,
 460         and other Node-query methods that return a null.'
 461
 462         (Note: This benchmark is slower than document.getElementById() for a defined id.
 463         We will fix the performance issue soon.)
 464
 465         Test results in my Linux desktop:
 466
 467         RESULT Bindings: undefined-get-element-by-id= 271.945751345 runs/s
 468         median= 273.122389532 runs/s, stdev= 6.78219237299 runs/s, min= 257.4002574 runs/s, max= 281.690140845 runs/s
 469
 470         RESULT Bindings: undefined-get-element-by-id= 269.105872408 runs/s
 471         median= 268.948655257 runs/s, stdev= 5.73707323755 runs/s, min= 257.4002574 runs/s, max= 278.128950695 runs/s
 472
 473         RESULT Bindings: undefined-get-element-by-id= 269.987513372 runs/s
 474         median= 272.952853598 runs/s, stdev= 5.30252336078 runs/s, min= 255.754475703 runs/s, max= 273.97260274 runs/s
 475
 476         RESULT Bindings: undefined-get-element-by-id= 270.169625901 runs/s
 477         median= 271.604938272 runs/s, stdev= 5.29502391906 runs/s, min= 258.064516129 runs/s, max= 278.481012658 runs/s
 478
 479         RESULT Bindings: undefined-get-element-by-id= 269.054631422 runs/s
 480         median= 267.966292802 runs/s, stdev= 6.27197240579 runs/s, min= 258.064516129 runs/s, max= 278.481012658 runs/s
 481
 482         * Bindings/undefined-get-element-by-id.html: Added.
 483
 484 2012-05-18  Kentaro Hara  <haraken@chromium.org>
 485
 486         [perf-test] Add a binding benchmark for getElementsByTagName()
 487         https://bugs.webkit.org/show_bug.cgi?id=86810
 488
 489         Reviewed by Ryosuke Niwa.
 490
 491         We should remove Bindings/dom-attributes.html and instead add more
 492         reasonable micro benchmarks by classifying DOM binding call paths.
 493
 494         This patch adds a benchmark for document.getElementsByTagName().
 495         This benchmark covers 'getElementsByTagName (not in document)',
 496         'getElementsByTagName', 'getElementsByName (not in document)'
 497         and 'getElementsByName' in Dromaeo/dom-query.html, and other DOM
 498         methods that return a NodeList.
 499
 500         Test results in my Linux desktop:
 501
 502         RESULT Bindings: get-elements-by-tag-name= 364.542328875 runs/s
 503         median= 366.972477064 runs/s, stdev= 5.25387955333 runs/s, min= 351.317440402 runs/s, max= 368.421052632 runs/s
 504
 505         RESULT Bindings: get-elements-by-tag-name= 364.341569369 runs/s
 506         median= 366.492146597 runs/s, stdev= 4.867374197 runs/s, min= 352.644836272 runs/s, max= 367.936925099 runs/s
 507
 508         RESULT Bindings: get-elements-by-tag-name= 363.924242967 runs/s
 509         median= 366.013071895 runs/s, stdev= 5.16768212289 runs/s, min= 351.317440402 runs/s, max= 367.936925099 runs/s
 510
 511         RESULT Bindings: get-elements-by-tag-name= 365.332594438 runs/s
 512         median= 367.454068241 runs/s, stdev= 6.17046918572 runs/s, min= 350.0 runs/s, max= 370.37037037 runs/s
 513
 514         RESULT Bindings: get-elements-by-tag-name= 363.999514324 runs/s
 515         median= 366.492146597 runs/s, stdev= 5.31235674905 runs/s, min= 350.877192982 runs/s, max= 368.421052632 runs/s
 516
 517         * Bindings/get-elements-by-tag-name.html: Added.
 518
 519 2012-05-18  Kentaro Hara  <haraken@chromium.org>
 520
 521         [perf-test] Add 7 micro benchmarks for DOM bindings
 522         https://bugs.webkit.org/show_bug.cgi?id=86816
 523
 524         Reviewed by Ryosuke Niwa.
 525
 526         We should remove Bindings/dom-attributes.html and instead add more
 527         reasonable micro benchmarks by classifying DOM binding call paths.
 528         This patch adds seven micro benchmarks for DOM bindings.
 529
 530         id-getter.html covers 'element.property' in Dromaeo/dom-attr.html,
 531         'innerHTML' in Dromaeo/dom-modify.html, and other DOM attributes that
 532         return a String.
 533
 534         id-setter.html covers 'element.property = value' in Dromaeo/dom-attr.html,
 535         and other DOM attributes that sets a String.
 536
 537         undefined-id-getter.html covers undefined DOM attributes that return an empty String.
 538
 539         append-child.html covers 'appendChild' and 'removeChild' in Dromaeo/dom-modify.html.
 540
 541         insert-before.html covers 'insertBefore' in Dromaeo/dom-modify.html and replaceChild().
 542
 543         create-element.html covers 'createElement', 'createTextNode' and 'cloneNode'
 544         in Dromaeo/dom-modify.html, and other DOM methods that return a new Node object.
 545
 546         node-list-access.html covers 'childNodes' in Dromaeo/dom-traverse.html,
 547         and other DOM attributes that access NodeList.
 548
 549         Test results in my Linux desktop:
 550
 551         [id-getter]
 552         RESULT Bindings: id-getter= 550.517821097 runs/s
 553         median= 550.688360451 runs/s, stdev= 0.970723739106 runs/s, min= 548.628428928 runs/s, max= 552.070263488 runs/s
 554
 555         RESULT Bindings: id-getter= 550.655839847 runs/s
 556         median= 550.688360451 runs/s, stdev= 1.03312821884 runs/s, min= 549.313358302 runs/s, max= 552.070263488 runs/s
 557
 558         RESULT Bindings: id-getter= 550.277753355 runs/s
 559         median= 550.344180225 runs/s, stdev= 1.19960683464 runs/s, min= 547.945205479 runs/s, max= 552.070263488 runs/s
 560
 561         RESULT Bindings: id-getter= 550.072271009 runs/s
 562         median= 550.344180225 runs/s, stdev= 1.38925152131 runs/s, min= 547.263681592 runs/s, max= 552.070263488 runs/s
 563
 564         RESULT Bindings: id-getter= 548.802050235 runs/s
 565         median= 548.628428928 runs/s, stdev= 1.18244513683 runs/s, min= 545.905707196 runs/s, max= 550.688360451 runs/s
 566
 567         [id-setter]
 568         RESULT Bindings: id-setter= 211.370591663 runs/s
 569         median= 210.803689065 runs/s, stdev= 1.4555795889 runs/s, min= 209.973753281 runs/s, max= 214.285714286 runs/s
 570
 571         RESULT Bindings: id-setter= 215.439081352 runs/s
 572         median= 214.797136038 runs/s, stdev= 1.50920388121 runs/s, min= 213.049267643 runs/s, max= 218.446601942 runs/s
 573
 574         RESULT Bindings: id-setter= 216.721698466 runs/s
 575         median= 215.956886352 runs/s, stdev= 1.54090749644 runs/s, min= 214.285714286 runs/s, max= 219.78021978 runs/s
 576
 577         RESULT Bindings: id-setter= 216.343019087 runs/s
 578         median= 215.827338129 runs/s, stdev= 1.46619894553 runs/s, min= 215.053763441 runs/s, max= 218.97810219 runs/s
 579
 580         RESULT Bindings: id-setter= 216.799288017 runs/s
 581         median= 215.956886352 runs/s, stdev= 1.49100776978 runs/s, min= 215.311004785 runs/s, max= 219.78021978 runs/s
 582
 583         [undefined-id-getter]
 584         RESULT Bindings: undefined-id-getter= 494.104255241 runs/s
 585         median= 494.132368875 runs/s, stdev= 1.13092821109 runs/s, min= 490.797546012 runs/s, max= 495.662949195 runs/s
 586
 587         RESULT Bindings: undefined-id-getter= 493.921802916 runs/s
 588         median= 493.827160494 runs/s, stdev= 1.25097443119 runs/s, min= 490.196078431 runs/s, max= 495.662949195 runs/s
 589
 590         RESULT Bindings: undefined-id-getter= 491.976485667 runs/s
 591         median= 492.004920049 runs/s, stdev= 0.947090576896 runs/s, min= 490.196078431 runs/s, max= 493.827160494 runs/s
 592
 593         RESULT Bindings: undefined-id-getter= 491.987710185 runs/s
 594         median= 494.132368875 runs/s, stdev= 8.66602543327 runs/s, min= 454.545454545 runs/s, max= 495.662949195 runs/s
 595
 596         RESULT Bindings: undefined-id-getter= 494.195515847 runs/s
 597         median= 494.437577256 runs/s, stdev= 1.06135681702 runs/s, min= 490.797546012 runs/s, max= 495.662949195 runs/s
 598
 599         [append-child]
 600         RESULT Bindings: append-child= 260.628237026 runs/s
 601         median= 260.926399352 runs/s, stdev= 2.55231060598 runs/s, min= 256.739409499 runs/s, max= 265.604249668 runs/s
 602
 603         RESULT Bindings: append-child= 280.796193436 runs/s
 604         median= 280.791435348 runs/s, stdev= 2.53961321867 runs/s, min= 277.427490542 runs/s, max= 286.085825748 runs/s
 605
 606         RESULT Bindings: append-child= 279.856209714 runs/s
 607         median= 280.25477707 runs/s, stdev= 2.85447647301 runs/s, min= 275.344180225 runs/s, max= 284.237726098 runs/s
 608
 609         RESULT Bindings: append-child= 280.834502714 runs/s
 610         median= 280.25477707 runs/s, stdev= 2.67738046429 runs/s, min= 277.427490542 runs/s, max= 287.206266319 runs/s
 611
 612         RESULT Bindings: append-child= 274.723872106 runs/s
 613         median= 273.291925466 runs/s, stdev= 3.03671746706 runs/s, min= 272.614622057 runs/s, max= 282.413350449 runs/s
 614
 615         [insert-before]
 616         RESULT Bindings: insert-before= 180.986913006 runs/s
 617         median= 180.878552972 runs/s, stdev= 0.758038687299 runs/s, min= 180.18018018 runs/s, max= 183.968462549 runs/s
 618
 619         RESULT Bindings: insert-before= 186.339916144 runs/s
 620         median= 186.170212766 runs/s, stdev= 0.869022097127 runs/s, min= 185.676392573 runs/s, max= 189.798339265 runs/s
 621
 622         RESULT Bindings: insert-before= 189.45344838 runs/s
 623         median= 189.349112426 runs/s, stdev= 0.811854866567 runs/s, min= 187.79342723 runs/s, max= 192.539109507 runs/s
 624
 625         RESULT Bindings: insert-before= 189.363295991 runs/s
 626         median= 189.349112426 runs/s, stdev= 0.75387713097 runs/s, min= 188.014101058 runs/s, max= 192.076830732 runs/s
 627
 628         RESULT Bindings: insert-before= 189.252041642 runs/s
 629         median= 189.349112426 runs/s, stdev= 0.840447225363 runs/s, min= 187.573270809 runs/s, max= 192.076830732 runs/s
 630
 631         [create-element]
 632         RESULT Bindings: create-element= 280.675830077 runs/s
 633         median= 280.701824914 runs/s, stdev= 0.788606343054 runs/s, min= 279.441117764 runs/s, max= 282.542885974 runs/s
 634
 635         RESULT Bindings: create-element= 276.427631436 runs/s
 636         median= 277.227722772 runs/s, stdev= 3.39874667549 runs/s, min= 262.417994377 runs/s, max= 279.162512463 runs/s
 637
 638         RESULT Bindings: create-element= 277.156894411 runs/s
 639         median= 277.227722772 runs/s, stdev= 1.13255119226 runs/s, min= 274.914089347 runs/s, max= 278.884462151 runs/s
 640
 641         RESULT Bindings: create-element= 278.066825356 runs/s
 642         median= 278.191822254 runs/s, stdev= 0.686497782792 runs/s, min= 276.953511375 runs/s, max= 279.162512463 runs/s
 643
 644         RESULT Bindings: create-element= 276.431538109 runs/s
 645         median= 276.406712734 runs/s, stdev= 0.759593662298 runs/s, min= 275.229357798 runs/s, max= 278.053624628 runs/s
 646
 647         [node-list-access]
 648         RESULT Bindings: node-list-access= 222.747504073 runs/s
 649         median= 222.772277228 runs/s, stdev= 0.789027527555 runs/s, min= 221.674876847 runs/s, max= 224.159402242 runs/s
 650
 651         RESULT Bindings: node-list-access= 220.722448287 runs/s
 652         median= 221.266217576 runs/s, stdev= 1.65463497875 runs/s, min= 217.391304348 runs/s, max= 222.222222222 runs/s
 653
 654         RESULT Bindings: node-list-access= 222.195870443 runs/s
 655         median= 222.222222222 runs/s, stdev= 0.489709262091 runs/s, min= 221.13022113 runs/s, max= 223.048327138 runs/s
 656
 657         RESULT Bindings: node-list-access= 221.431121677 runs/s
 658         median= 221.538545435 runs/s, stdev= 0.608292116087 runs/s, min= 220.048899756 runs/s, max= 222.772277228 runs/s
 659
 660         RESULT Bindings: node-list-access= 219.001367135 runs/s
 661         median= 221.538545435 runs/s, stdev= 4.65773933007 runs/s, min= 209.973753281 runs/s, max= 222.496909765 runs/s
 662
 663         * Bindings/append-child.html: Added.
 664         * Bindings/create-element.html: Added.
 665         * Bindings/id-getter.html: Added.
 666         * Bindings/id-setter.html: Added.
 667         * Bindings/insert-before.html: Added.
 668         * Bindings/node-list-access.html: Added.
 669         * Bindings/undefined-id-getter.html: Added.
 670
 671 2012-05-18  Kentaro Hara  <haraken@chromium.org>
 672
 673         [perf-test] Add a binding benchmark for div.setAttribute()
 674         https://bugs.webkit.org/show_bug.cgi?id=86815
 675
 676         Reviewed by Ryosuke Niwa.
 677
 678         We should remove Bindings/dom-attributes.html and instead add more
 679         reasonable micro benchmarks by classifying DOM binding call paths.
 680
 681         This patch adds a benchmark for div.setAttribute(). This benchmark
 682         covers 'setAttribute' in Dromaeo/dom-attr.html and other DOM methods
 683         that return an undefined.
 684
 685         Test results in my Linux desktop:
 686
 687         RESULT Bindings: set-attribute= 793.745073252 runs/s
 688         median= 791.589665765 runs/s, stdev= 6.00467159709 runs/s, min= 785.27607362 runs/s, max= 808.080808081 runs/s
 689
 690         RESULT Bindings: set-attribute= 790.858444164 runs/s
 691         median= 787.918921436 runs/s, stdev= 7.86264953655 runs/s, min= 780.234070221 runs/s, max= 806.04534005 runs/s
 692
 693         RESULT Bindings: set-attribute= 787.277571417 runs/s
 694         median= 784.31372549 runs/s, stdev= 7.45363346657 runs/s, min= 778.588807786 runs/s, max= 804.020100503 runs/s
 695
 696         RESULT Bindings: set-attribute= 782.401393791 runs/s
 697         median= 779.062978802 runs/s, stdev= 6.48250081621 runs/s, min= 777.642770352 runs/s, max= 796.019900498 runs/s
 698
 699         RESULT Bindings: set-attribute= 791.792516608 runs/s
 700         median= 791.556728232 runs/s, stdev= 6.49182980397 runs/s, min= 779.220779221 runs/s, max= 808.080808081 runs/s
 701
 702         * Bindings/set-attribute.html: Added.
 703
 704 2012-05-17  Kentaro Hara  <haraken@chromium.org>
 705
 706         [perf-test] Add a binding benchmark for div.getAttribute()
 707         https://bugs.webkit.org/show_bug.cgi?id=86814
 708
 709         Reviewed by Ryosuke Niwa.
 710
 711         We should remove Bindings/dom-attributes.html and instead add more
 712         reasonable micro benchmarks by classifying DOM binding call paths.
 713
 714         This patch adds a benchmark for div.getAttribute(). This benchmark
 715         covers 'getAttribute' in Dromaeo/dom-attr.html, and other DOM methods
 716         that return a String.
 717
 718         Test results in my Linux desktop:
 719
 720         RESULT Bindings: get-attribute= 250.088158538 runs/s
 721         median= 250.469728765 runs/s, stdev= 4.66551901315 runs/s, min= 240.673886883 runs/s, max= 256.739409499 runs/s
 722
 723         RESULT Bindings: get-attribute= 250.209764698 runs/s
 724         median= 250.626959986 runs/s, stdev= 4.12030002477 runs/s, min= 241.254523522 runs/s, max= 255.102040816 runs/s
 725
 726         RESULT Bindings: get-attribute= 248.907532887 runs/s
 727         median= 250.312891114 runs/s, stdev= 3.23173125874 runs/s, min= 240.673886883 runs/s, max= 251.572327044 runs/s
 728
 729         RESULT Bindings: get-attribute= 250.243680844 runs/s
 730         median= 250.783797637 runs/s, stdev= 4.19327170024 runs/s, min= 240.963855422 runs/s, max= 255.102040816 runs/s
 731
 732         RESULT Bindings: get-attribute= 245.329046608 runs/s
 733         median= 244.948040416 runs/s, stdev= 2.24314248534 runs/s, min= 240.384615385 runs/s, max= 248.756218905 runs/s
 734
 735         * Bindings/get-attribute.html: Added.
 736
 737 2012-05-17  Kentaro Hara  <haraken@chromium.org>
 738
 739         [perf-test] Add a binding benchmark for getElementById()
 740         https://bugs.webkit.org/show_bug.cgi?id=86808
 741
 742         Reviewed by Ryosuke Niwa.
 743
 744         Currently Bindings/dom-attributes.html is just a collection
 745         of random DOM attributes and methods. We should remove it and
 746         instead add more reasonable micro benchmarks by classifying
 747         DOM binding call paths.
 748
 749         This patch adds a benchmark for document.getElementById(),
 750         which covers other DOM methods that return a Node Object.
 751
 752         The reason why we want to add a benchmark for document.getElementById()
 753         in spite of the fact that Dromaeo/dom-query.html is already testing
 754         document.getElementById(), is that Dromaeo/dom-query.html outputs one
 755         result for several benchmarks (i.e. getElementById().nodeType,
 756         getElementsByTagName()[0].nodeType etc), and thus we cannot observe
 757         regression of getElementById() itself. (Catching regression of
 758         getElementById() would be critical.)
 759
 760         Test results in my Linux desktop:
 761
 762         RESULT Bindings: get-element-by-id= 300.933707795 runs/s
 763         median= 302.267002519 runs/s, stdev= 3.04954648504 runs/s, min= 295.930949445 runs/s, max= 305.34351145 runs/s
 764
 765         RESULT Bindings: get-element-by-id= 302.030324483 runs/s
 766         median= 302.267481977 runs/s, stdev= 1.78518068367 runs/s, min= 298.507462687 runs/s, max= 304.568527919 runs/s
 767
 768         RESULT Bindings: get-element-by-id= 301.993483538 runs/s
 769         median= 302.648171501 runs/s, stdev= 1.87604126626 runs/s, min= 296.662546354 runs/s, max= 304.568527919 runs/s
 770
 771         RESULT Bindings: get-element-by-id= 301.936607982 runs/s
 772         median= 302.45758701 runs/s, stdev= 1.88991087072 runs/s, min= 298.879202989 runs/s, max= 304.955527319 runs/s
 773
 774         RESULT Bindings: get-element-by-id= 301.952234563 runs/s
 775         median= 302.648171501 runs/s, stdev= 1.59929752614 runs/s, min= 299.625468165 runs/s, max= 304.568527919 runs/s
 776
 777         * Bindings/get-element-by-id.html: Added.
 778
 779 2012-05-17  Kentaro Hara  <haraken@chromium.org>
 780
 781         [perf-test] Add a benchmark for body.scrollTop
 782         https://bugs.webkit.org/show_bug.cgi?id=86806
 783
 784         Reviewed by Ryosuke Niwa.
 785
 786         Currently Bindings/dom-attributes.html is just a collection of random
 787         DOM attributes and methods. We should remove it and instead add more
 788         reasonable micro benchmarks by classifying DOM binding call paths.
 789
 790         This patch adds a benchmark for body.scrollTop, which covers other
 791         DOM attributes that return small integers.
 792
 793         (Note: Performance optimization for the benchmark will be coming soon.)
 794
 795         Test results in my Linux desktop:
 796
 797         RESULT Bindings: scroll-top= 209.410654239 runs/s
 798         median= 211.64021164 runs/s, stdev= 4.09958741195 runs/s, min= 201.005025126 runs/s, max= 212.483399734 runs/s
 799
 800         RESULT Bindings: scroll-top= 195.651111157 runs/s
 801         median= 196.319018405 runs/s, stdev= 1.77011627406 runs/s, min= 191.846522782 runs/s, max= 197.28729963 runs/s
 802         RESULT Bindings: scroll-top= 200.004907612 runs/s
 803         median= 201.511335013 runs/s, stdev= 3.27455403827 runs/s, min= 193.236714976 runs/s, max= 202.53164557 runs/s
 804
 805         RESULT Bindings: scroll-top= 209.582737512 runs/s
 806         median= 210.803689065 runs/s, stdev= 3.16769655464 runs/s, min= 203.045685279 runs/s, max= 211.920529801 runs/s
 807
 808         RESULT Bindings: scroll-top= 199.924607687 runs/s
 809         median= 201.13144338 runs/s, stdev= 3.11080503619 runs/s, min= 193.003618818 runs/s, max= 202.02020202 runs/s
 810
 811         * Bindings/scroll-top.html: Added.
 812
 813 2012-05-17  Kentaro Hara  <haraken@chromium.org>
 814
 815         [perf-test] Add a micro benchmark for div.firstChild getter for an undefined firstChild
 816         https://bugs.webkit.org/show_bug.cgi?id=86725
 817
 818         Reviewed by Ryosuke Niwa.
 819
 820         We are going to remove Bindings/dom-attributes.html and instead add more
 821         reasonable micro benchmarks by classifying DOM binding call paths.
 822
 823         In this bug, we add a micro benchmark for div.firstChild getter for an
 824         undefined firstChild. This benchmark covers other undefined DOM attributes
 825         that return a null.
 826
 827         (Note: In V8, div.firstChild for an undefined firstChild is currently slower
 828         than div.firstChild for a defined firstChild.)
 829
 830         Test results in my Linux desktop:
 831
 832         RESULT Bindings: undefined-first-child= 541.408311042 runs/s
 833         median= 541.205412054 runs/s, stdev= 1.29841219187 runs/s, min= 537.897310513 runs/s, max= 543.209876543 runs/s
 834
 835         RESULT Bindings: undefined-first-child= 541.00750695 runs/s
 836         median= 540.540540541 runs/s, stdev= 0.967521230965 runs/s, min= 539.215686275 runs/s, max= 543.209876543 runs/s
 837
 838         RESULT Bindings: undefined-first-child= 523.630714897 runs/s
 839         median= 523.560209424 runs/s, stdev= 1.01570919645 runs/s, min= 522.193211488 runs/s, max= 525.624178712 runs/s
 840
 841         RESULT Bindings: undefined-first-child= 553.077663588 runs/s
 842         median= 553.459119497 runs/s, stdev= 0.805697397863 runs/s, min= 551.378446115 runs/s, max= 554.156171285 runs/s
 843
 844         RESULT Bindings: undefined-first-child= 541.5401432 runs/s
 845         median= 541.538666618 runs/s, stdev= 0.954301364767 runs/s, min= 539.877300613 runs/s, max= 543.209876543 runs/s
 846
 847         * Bindings/undefined-first-child.html: Added.
 848
 849 2012-05-17  Kentaro Hara  <haraken@chromium.org>
 850
 851         [Performance test] Add a micro benchmark for div.firstChild getter
 852         https://bugs.webkit.org/show_bug.cgi?id=86582
 853
 854         Reviewed by Ryosuke Niwa.
 855
 856         We are going to remove Bindings/dom-attributes.html and instead add more
 857         reasonable micro benchmarks by classifying DOM binding call paths.
 858
 859         In this bug, we add a micro benchmark for div.firstChild getter.
 860         This benchmark covers 'firstChild', 'lastChild', 'nextSibling' and
 861         'previousSibling' in Dromaeo/dom-traverse.html, and other DOM attributes
 862         that return a Node object.
 863
 864         Multiple test results in my Linux desktop:
 865
 866         RESULT Bindings: first-child= 798.157160346 runs/s
 867         median= 798.004987531 runs/s, stdev= 1.52006063407 runs/s, min= 796.019900498 runs/s, max= 801.001251564 runs/s
 868         RESULT Bindings: first-child= 797.603608554 runs/s
 869         median= 797.872340426 runs/s, stdev= 2.2522621261 runs/s, min= 791.556728232 runs/s, max= 801.001251564 runs/s
 870         RESULT Bindings: first-child= 798.656295468 runs/s
 871         median= 798.004987531 runs/s, stdev= 1.79367478063 runs/s, min= 797.01120797 runs/s, max= 803.011292346 runs/s
 872         RESULT Bindings: first-child= 797.812784267 runs/s
 873         median= 798.004987531 runs/s, stdev= 2.31766523191 runs/s, min= 791.100123609 runs/s, max= 802.005012531 runs/s
 874         RESULT Bindings: first-child= 797.963311597 runs/s
 875         median= 798.004987531 runs/s, stdev= 2.53014907337 runs/s, min= 789.14919852 runs/s, max= 801.001251564 runs/s
 876
 877         * Bindings/first-child.html: Added.
 878
 879 2012-05-16  Kentaro Hara  <haraken@chromium.org>
 880
 881         [Performance test] Support "description" for PerfTestRunner.run and PerfTestRunner.runPerSecond
 882         https://bugs.webkit.org/show_bug.cgi?id=86696
 883
 884         Reviewed by Ryosuke Niwa.
 885
 886         We want to add some description for each PerfTestRunner.run and
 887         PerfTestRunner.runPerSecond, so that people can know what each
 888         run is testing.
 889
 890         Output example:
 891
 892         $ ./Tools/Scripts/run-perf-tests Bindings/first-child.html
 893         Running Bindings/first-child.html (1 of 1)
 894         DESCRIPTION: Benchmark for DOM attributes that return a Node object.
 895         RESULT Bindings: first-child= 788.359076534 runs/s
 896         median= 797.508097751 runs/s, stdev= 19.0972905207 runs/s, min= 746.666666667 runs/s, max= 801.001251564 runs/s
 897
 898         * resources/runner.js:
 899         (PerfTestRunner.logStatistics):
 900         (PerfTestRunner.printStatistics):
 901         (PerfTestRunner.runPerSecond):
 902
 903 2012-05-16  Yury Semikhatsky  <yurys@chromium.org>
 904
 905         Unreviewed. Fix heap profiler performance test after r117234.
 906
 907         * inspector/heap-snapshot.html:
 908
 909 2012-05-14  Ryosuke Niwa  <rniwa@webkit.org>
 910
 911         Convert CSS and Parser tests to runs/s tests
 912         https://bugs.webkit.org/show_bug.cgi?id=86399
 913
 914         Reviewed by Ojan Vafai.
 915
 916         Use runPerSecond instead of run to measure results in function calls per second (run/s) instead of time (ms).
 917         This allows us to continue measuring performance reliably even when WebKit's performance improves and results
 918         become too small for Date.now()'s precision (around 15 ms).
 919
 920         It should also reduce perf. bots cycle time for slower tests because we limit the number of function calls by time.
 921
 922         * CSS/CSSPropertySetterGetter.html:
 923         * CSS/CSSPropertyUpdateValue.html:
 924         * Parser/css-parser-yui.html:
 925         * Parser/simple-url.html:
 926         * Parser/tiny-innerHTML.html:
 927         * Parser/url-parser.html:
 928         * Parser/xml-parser.html:
 929
 930 2012-05-14  Dan Bernstein  <mitz@apple.com>
 931
 932         Add a line layout performance test
 933         https://bugs.webkit.org/show_bug.cgi?id=86216
 934
 935         Reviewed by Ryosuke Niwa.
 936
 937         * Layout/line-layout.html: Added.
 938
 939 2012-05-14  Tony Chang  <tony@chromium.org>
 940
 941         Don't append log lines while perf tests are running.
 942         https://bugs.webkit.org/show_bug.cgi?id=86028
 943
 944         Reviewed by Ryosuke Niwa.
 945
 946         In some tests, the extra DOM nodes can cause the test timing to change.
 947         Avoid adding DOM nodes until the test is over. When running in a browser,
 948         we append nodes as we run so the user can have some feedback.
 949
 950         * resources/runner.js:
 951         (PerfTestRunner.log): Store the log lines in an array until the test is finished.
 952
 953 2012-05-14  Ryosuke Niwa  <rniwa@webkit.org>
 954
 955         Build fix. Don't adjust calls per iteration in the second run.
 956
 957         * resources/runner.js:
 958         (PerfTestRunner._perSecondRunner):
 959
 960 2012-05-13  Ryosuke Niwa  <rniwa@webkit.org>
 961
 962         performance tests should be able to measure runs/sec rather than time
 963         https://bugs.webkit.org/show_bug.cgi?id=86021
 964
 965         Reviewed by Ojan Vafai.
 966
 967         Add PerfTestRunner.runPerSecond. It uses _runLoop but replaces _runner by _perSecondRunner
 968         to compute runs/s of runFunction.
 969
 970         When _perSecondRunner is called for the first time, i.e. _completedRuns is 0 (notice this is -1
 971         in regular run/_runner), it slowly increases the number of function calls to runFunction between
 972         time measurements in order to discount the time used by new Date() calls themselves until the
 973         total time spent reaches 100 milliseconds.
 974
 975         By default, runPerSecond runs the test for at least 750 milliseconds in each run, and executes
 976         21 runs, yielding the total run time of roughly 18 seconds. This is significantly faster than
 977         most of existing performance tests. Also see http://ejohn.org/blog/accuracy-of-javascript-time/.
 978
 979         Finally, refactored the existing methods of PerfTestRunner to allow "runs/s" unit and share code.
 980
 981         * Layout/flexbox-column-nowrap.html:
 982         * Layout/flexbox-column-wrap.html:
 983         * Layout/flexbox-row-nowrap.html:
 984         * Layout/flexbox-row-wrap.html:
 985         * resources/runner.js:
 986         (PerfTestRunner.computeStatistics): Takes unit.
 987         (PerfTestRunner.logStatistics): Ditto.
 988         (PerfTestRunner._runLoop):
 989         (PerfTestRunner._runner):
 990         (PerfTestRunner.runPerSecond): Added.
 991         (PerfTestRunner._perSecondRunner): Added. Called by _runLoop.
 992         (PerfTestRunner._perSecondRunnerIterator): Added.
 993
 994 2012-05-09  Tony Chang  <tony@chromium.org>
 995
 996         add some basic perf-o-matic tests for flexbox
 997         https://bugs.webkit.org/show_bug.cgi?id=86017
 998
 999         Reviewed by Ryosuke Niwa.
1000
1001         Each of these tests take < 3s on my machine.
1002
1003         * Layout/flexbox-column-nowrap.html: Added.
1004         * Layout/flexbox-column-wrap.html: Added.
1005         * Layout/flexbox-row-nowrap.html: Added.
1006         * Layout/flexbox-row-wrap.html: Added.
1007
1008 2012-05-05  Ilya Tikhonovsky  <loislo@chromium.org>
1009
1010         Web Inspector: unreviewed followup fix for the performance test after r116218.
1011
1012         * inspector/heap-snapshot.html:
1013
1014 2012-05-02  Tom Zakrajsek  <tomz@codeaurora.org>
1015
1016         Need tests for PerfTestRunner.computeStatistics
1017         https://bugs.webkit.org/show_bug.cgi?id=85410
1018
1019         Reviewed by Ryosuke Niwa.
1020
1021         * resources/runner.js:
1022
1023 2012-05-02  Sheriff Bot  <webkit.review.bot@gmail.com>
1024
1025         Unreviewed, rolling out r115905.
1026         http://trac.webkit.org/changeset/115905
1027         https://bugs.webkit.org/show_bug.cgi?id=85435
1028
1029         New computeStatistics test is failing on chromium ports
1030         (Requested by tomz on #webkit).
1031
1032         * resources/runner.js:
1033
1034 2012-05-02  Tom Zakrajsek  <tomz@codeaurora.org>
1035
1036         Need tests for PerfTestRunner.computeStatistics
1037         https://bugs.webkit.org/show_bug.cgi?id=85410
1038
1039         Reviewed by Ryosuke Niwa.
1040
1041         * resources/runner.js:
1042
1043 2012-04-29  Tom Zakrajsek  <tomz@codeaurora.org>
1044
1045         PerfTestRunner.computeStatistics incorrectly calculates min, max and median
1046         https://bugs.webkit.org/show_bug.cgi?id=85111
1047
1048         Reviewed by Ryosuke Niwa.
1049
1050         The sort of the data input was being done alphabetically.
1051         So I provided a numeric compare function.
1052
1053         * resources/runner.js:
1054
1055 2012-04-17  Ilya Tikhonovsky  <loislo@chromium.org>
1056
1057         Web Inspector: the fix for test after massive rename in r114271.
1058         https://bugs.webkit.org/show_bug.cgi?id=84147
1059
1060         Drive by fix. The test was renamed to heap-snapshot.html
1061
1062         Reviewed by Yury Semikhatsky.
1063
1064         * inspector/heap-snapshot.html: Renamed from PerformanceTests/inspector/detailed-heapshots-smoke-test.html.
1065
1066 2012-04-16  Yury Semikhatsky  <yurys@chromium.org>
1067
1068         Web Inspector: rename heap profiler files and classes DetailedHeapshot* ->HeapSnapshot*
1069         https://bugs.webkit.org/show_bug.cgi?id=84038
1070
1071         A bunch of renames "detailed heapshot" -> "heap snapshot".
1072
1073         Reviewed by Pavel Feldman.
1074
1075         * inspector/detailed-heapshots-smoke-test.html:
1076
1077 2012-04-13  Ryosuke Niwa  <rniwa@webkit.org>
1078
1079         Fix skipped list.
1080
1081         * Skipped:
1082
1083 2012-04-12  Ryosuke Niwa  <rniwa@webkit.org>
1084
1085         Enable SVG page loading performance tests
1086         https://bugs.webkit.org/show_bug.cgi?id=83861
1087
1088         Reviewed by Adam Barth.
1089
1090         Unskip PageLoad directory and skip tests that load too fast (less than 100ms) and are unstable,
1091         and the ones that take too long to load (more than 20s).
1092
1093         * Skipped:
1094
1095 2012-04-04  Ilya Tikhonovsky  <loislo@chromium.org>
1096
1097         Web Inspector: Unreviewed. temporary disable broken inspector tests.
1098
1099         * inspector/first-open-scripts.html.broken: Renamed from PerformanceTests/inspector/first-open-scripts.html.
1100         * inspector/network-append-30-requests.html.broken: Renamed from PerformanceTests/inspector/network-append-30-requests.html.
1101         * inspector/show-panel.html.broken: Renamed from PerformanceTests/inspector/show-panel.html.
1102
1103 2012-04-04  Ilya Tikhonovsky  <loislo@chromium.org>
1104
1105         Web Inspector: linearise aggregate's retaining size calculation.
1106         https://bugs.webkit.org/show_bug.cgi?id=83125
1107
1108         This version is twice as fast as the original and it is non-recursive.
1109
1110         Reviewed by Yury Semikhatsky.
1111
1112         * inspector/detailed-heapshots-smoke-test.html:
1113
1114 2012-03-30  David Barr  <davidbarr@chromium.org>
1115
1116         Split up top-level .gitignore and .gitattributes
1117         https://bugs.webkit.org/show_bug.cgi?id=82687
1118
1119         Reviewed by Tor Arne Vestbø.
1120
1121         * .gitattributes: Added.
1122
1123 2012-03-28  Yury Semikhatsky  <yurys@chromium.org>
1124
1125         Web Inspector: switch heap profiler front-end to separate storage of nodes and edges
1126         https://bugs.webkit.org/show_bug.cgi?id=82453
1127
1128         Updated heap profiler performance test after heap profiler front-end
1129         changes.
1130
1131         Reviewed by Pavel Feldman.
1132
1133         * inspector/detailed-heapshots-smoke-test.html:
1134
1135 2012-03-27  Alexis Menard  <alexis.menard@openbossa.org>
1136
1137         Add a perf test which updates the value of an already declared CSS property.
1138         https://bugs.webkit.org/show_bug.cgi?id=82321
1139
1140         Reviewed by Andreas Kling.
1141
1142         This benchmark update CSS existing properties of the style from JS.
1143
1144         * CSS/CSSPropertyUpdateValue.html: Added.
1145
1146 2012-03-15  Kentaro Hara  <haraken@chromium.org>
1147
1148         [PerformanceTests] run-perf-tests should output correct units
1149         https://bugs.webkit.org/show_bug.cgi?id=78303
1150
1151         Reviewed by Ryosuke Niwa.
1152
1153         Previously run-perf-tests just supported "ms" for units.
1154         Consequently, Dromaeo perf tests had been reporting runs/s as "ms".
1155         This patch fixes run-perf-tests so that they can support custom units.
1156         The default unit is "ms".
1157
1158         The test result looks like this:
1159
1160         $ ./Tools/Scripts/run-perf-tests Dromaeo/dom-attr.html
1161         RESULT Dromaeo: dom-attr= 6465.9525483 runs/s
1162         median= 0.0 runs/s, stdev= 24.2983433436 runs/s, min= 6435.87649402 runs/s, max= 6515.63693392 runs/s
1163
1164         * Dromaeo/resources/dromaeorunner.js:
1165         (.):
1166         * resources/runner.js:
1167         (PerfTestRunner.computeStatistics):
1168         (PerfTestRunner.printStatistics):
1169
1170 2012-03-15  Alexis Menard  <alexis.menard@openbossa.org>
1171
1172         Fix the test failing with run-perf-tests.
1173
1174         Rubber stamped by Ryosuke Niwa.
1175
1176         * CSS/CSSPropertySetterGetter.html:
1177
1178 2012-03-14  Alexis Menard  <alexis.menard@openbossa.org>
1179
1180         Add a new perf test to benchmark setting and reading CSS properties from JS.
1181         https://bugs.webkit.org/show_bug.cgi?id=81155
1182
1183         Reviewed by Ryosuke Niwa.
1184
1185         This benchmark read and set CSS properties from JS. It contains a list
1186         of common CSS properties.
1187
1188         * CSS/CSSPropertySetterGetter.html: Added.
1189
1190 2012-03-14  Alexandru Chiculita  <achicu@adobe.com>
1191
1192         Layout/floats.html should be runnable by run-perf-tests
1193         https://bugs.webkit.org/show_bug.cgi?id=77051
1194
1195         Reviewed by Ryosuke Niwa.
1196
1197         I've split Layout/floats.html into multiple tests for each button in the original manual test.
1198         Some tests take longer to run and I've changed the number of iterations, so that each tests finishes under 30s per run.
1199
1200         * DOM/resources/dom-perf.js:
1201         * Layout/floats.html: Removed.
1202         * Layout/floats_100_100.html: Added.
1203         * Layout/floats_100_100_nested.html: Added.
1204         * Layout/floats_20_100.html: Added.
1205         * Layout/floats_20_100_nested.html: Added.
1206         * Layout/floats_2_100.html: Added.
1207         * Layout/floats_2_100_nested.html: Added.
1208         * Layout/floats_50_100.html: Added.
1209         * Layout/floats_50_100_nested.html: Added.
1210         * Layout/resources/floats.css: Added.
1211         (.container):
1212         (.float):
1213         (.big):
1214         (.float-end):
1215         * Layout/resources/floats.js: Added.
1216         (.):
1217         * Skipped:
1218         * resources/runner.js:
1219         (PerfTestRunner.resetRandomSeed):
1220         (PerfTestRunner.random):
1221         (Math.random):
1222             Moved the Math.random to runner.js to be used by all tests.
1223             Added resetRandomSeed to bring the randomizer back to initial seed.
1224             It is useful to get the same results at every run and minimize the
1225             differences between runs.
1226
1227 2012-03-14  Sheriff Bot  <webkit.review.bot@gmail.com>
1228
1229         Unreviewed, rolling out r110559.
1230         http://trac.webkit.org/changeset/110559
1231         https://bugs.webkit.org/show_bug.cgi?id=81142
1232
1233         Broke Dromaeo tests (Requested by rniwa on #webkit).
1234
1235         * Dromaeo/resources/dromaeorunner.js:
1236         (.):
1237
1238 2012-03-13  Kentaro Hara  <haraken@chromium.org>
1239
1240         Dromaeo perf-tests results are wrong
1241         https://bugs.webkit.org/show_bug.cgi?id=80974
1242
1243         Reviewed by Hajime Morita.
1244
1245         Dromaeo perf-tests had reported the sum of runs/seconds as "ms".
1246         This patch fixes them to report the execution time per run.
1247
1248         At present, they report "0.0ms" for "stdev". This is because
1249         dromaeorunner.js can just know stdev of "runs/seconds"
1250         and cannot calculate stdev of "ms" from that. We can fix it later.
1251
1252         * Dromaeo/resources/dromaeorunner.js:
1253         (.):
1254
1255 2012-03-12  Ilya Tikhonovsky  <loislo@chromium.org>
1256
1257         Web Inspector: chromium: increase heap snapshot performance coverage.
1258         https://bugs.webkit.org/show_bug.cgi?id=80829
1259
1260         Reviewed by Yury Semikhatsky.
1261
1262         * inspector/detailed-heapshots-smoke-test.html:
1263         * inspector/performance-test.js:
1264         (initialize_TimeTracker.InspectorTest.measureFunction):
1265
1266 2012-03-06  Kentaro Hara  <haraken@chromium.org>
1267
1268         [Perf tests] Reduce the number of loops in dom-attributes.html
1269         https://bugs.webkit.org/show_bug.cgi?id=80077
1270
1271         Reviewed by Ryosuke Niwa.
1272
1273         dom-attributes.html takes 100~ seconds in the perf bots. This patch reduces
1274         the time by reducing the number of loops.
1275
1276         This is just a temporary fix. After more detailed investigation, I will reduce
1277         the number of tests in dom-attributes.html, and revert back the number of loops
1278         to guarantee the reliability of the perf test results.
1279
1280         * Bindings/dom-attributes.html:
1281
1282 2012-03-06  Ilya Tikhonovsky  <loislo@chromium.org>
1283
1284         Web Inspector: Heap Snapshot test: split overal time metric into separate metrics.
1285         https://bugs.webkit.org/show_bug.cgi?id=80409
1286
1287         Reviewed by Yury Semikhatsky.
1288
1289         * inspector/detailed-heapshots-smoke-test.html:
1290
1291 2012-03-05  Ilya Tikhonovsky  <loislo@chromium.org>
1292
1293         Web Inspector: [chromium] introduce HeapSnapshot performance test.
1294         https://bugs.webkit.org/show_bug.cgi?id=80280
1295
1296         Reviewed by Pavel Feldman.
1297
1298         * inspector/detailed-heapshots-smoke-test.html: Added.
1299
1300 2012-03-01  Ryosuke Niwa  <rniwa@webkit.org>
1301
1302         DOM tests take too long to run
1303         https://bugs.webkit.org/show_bug.cgi?id=80082
1304
1305         Reviewed by Eric Seidel.
1306
1307         Reduce benchmark.timeToRun from 500ms to 100ms. Doesn't seem to affect test results much
1308         (except DOM/DOMTable.html) but significantly reduces the runtime of tests.
1309
1310         * DOM/resources/dom-perf.js:
1311         (BenchmarkSuite.prototype.RunSingle):
1312
1313 2012-02-26  Kentaro Hara  <haraken@chromium.org>
1314
1315         [Performance Tests] [Chromium] Bindings/dom-attributes.html is timing out
1316         https://bugs.webkit.org/show_bug.cgi?id=79593
1317
1318         Reviewed by Adam Barth.
1319
1320         This patch halves the execution time of Bindings/dom-attributes.html
1321         in order to avoid timeout in Chromium. The reason for the timeout is that
1322         V8 DOM bindings are too much slower than JSC bindings.
1323         I am a bit afraid that this patch will reduce the accuracy of the
1324         perf test results, but it would make sense to reduce the execution time
1325         until we fix the performance issue in V8 DOM bindings.
1326
1327         * Bindings/dom-attributes.html:
1328
1329 2012-02-24  Ryosuke Niwa  <rniwa@webkit.org>
1330
1331         Rename dom*_*attributes.html to dom*-*attributes.html to follow the WebKit convention.
1332
1333         * Bindings/dom-attributes.html: Copied from PerformanceTests/Bindings/dom_attributes.html.
1334         * Bindings/dom_attributes.html: Removed.
1335
1336 2012-02-24  Kentaro Hara  <haraken@chromium.org>
1337
1338         Unreviewed, rolling out r108477.
1339         http://trac.webkit.org/changeset/108477
1340         https://bugs.webkit.org/show_bug.cgi?id=79209
1341
1342         The perf tests committed in r108748 covers the perf tests
1343         committed in r108477. So I'll roll out r108477.
1344
1345         * Bindings/dom_attributes_id_getter.html: Removed.
1346         * Bindings/dom_attributes_id_setter.html: Removed.
1347         * Bindings/resources/dom_attributes_common.js: Removed.
1348
1349 2012-02-24  Kentaro Hara  <haraken@chromium.org>
1350
1351         Add performance tests for DOM attribute getters and setters
1352         https://bugs.webkit.org/show_bug.cgi?id=79208
1353
1354         Reviewed by Ryosuke Niwa.
1355
1356         This patch adds performance micro-benchmarks for DOM attribute
1357         getters and setters. We selected DOM attributes from HTMLElement,
1358         Element and Node which will most impact on the DOM performance
1359         in the real Web and which have different call-paths from another
1360         DOM attribute.
1361
1362         The perf test result in my local Chromium/Linux environment is as follows:
1363         median= 14967.0 ms, stdev= 85.2385476178 ms, min= 14861.0 ms, max= 15155.0 ms
1364
1365         If we run the test with DumpRenderTree, we can get the following test results:
1366         Info: [HTMLElement.id] avg=65.76ms, median=65.00ms, stdev=2.14ms, min=63, max=70
1367         Info: [HTMLElement.title] avg=64.05ms, median=63.00ms, stdev=1.50ms, min=63, max=67
1368         Info: [HTMLElement.lang] avg=64.19ms, median=63.00ms, stdev=1.82ms, min=62, max=67
1369         Info: [HTMLElement.dir] avg=64.10ms, median=63.00ms, stdev=1.72ms, min=62, max=68
1370         Info: [HTMLElement.className] avg=63.33ms, median=63.00ms, stdev=0.89ms, min=63, max=67
1371         Info: [HTMLElement.classList] avg=1946.00ms, median=1945.00ms, stdev=38.22ms, min=1921, max=2110
1372         Info: [HTMLElement.tabIndex] avg=83.52ms, median=80.00ms, stdev=6.00ms, min=79, max=94
1373         Info: [HTMLElement.draggable] avg=78.14ms, median=81.00ms, stdev=3.51ms, min=74, max=82
1374         Info: [HTMLElement.webkitdropzone] avg=64.33ms, median=63.00ms, stdev=1.98ms, min=63, max=68
1375         Info: [HTMLElement.hidden] avg=65.57ms, median=64.00ms, stdev=3.27ms, min=63, max=71
1376         Info: [HTMLElement.accessKey] avg=64.43ms, median=63.00ms, stdev=2.11ms, min=62, max=68
1377         Info: [HTMLElement.children] avg=130.76ms, median=128.00ms, stdev=9.75ms, min=109, max=143
1378         Info: [HTMLElement.isContentEditable] avg=74.29ms, median=75.00ms, stdev=2.00ms, min=71, max=77
1379         Info: [HTMLElement.spellcheck] avg=79.52ms, median=80.00ms, stdev=2.52ms, min=76, max=82
1380         Info: [HTMLElement.innerHTML] avg=279.05ms, median=279.00ms, stdev=1.96ms, min=276, max=283
1381         Info: [HTMLElement.innerText] avg=204.14ms, median=203.00ms, stdev=2.95ms, min=200, max=210
1382         Info: [HTMLElement.outerHTML] avg=1544.10ms, median=1546.00ms, stdev=15.56ms, min=1514, max=1568
1383         Info: [HTMLElement.outerText] avg=216.05ms, median=207.00ms, stdev=18.19ms, min=201, max=247
1384         Info: [HTMLElement.contentEditable] avg=334.24ms, median=323.00ms, stdev=36.06ms, min=318, max=446
1385         Info: [HTMLElement.id="foo"] avg=53.19ms, median=53.00ms, stdev=1.76ms, min=52, max=58
1386         Info: [HTMLElement.title="foo"] avg=63.14ms, median=63.00ms, stdev=0.35ms, min=63, max=64
1387         Info: [HTMLElement.lang="foo"] avg=60.86ms, median=61.00ms, stdev=0.56ms, min=60, max=62
1388         Info: [HTMLElement.dir="foo"] avg=53.57ms, median=54.00ms, stdev=0.58ms, min=53, max=55
1389         Info: [HTMLElement.className="foo"] avg=64.52ms, median=64.00ms, stdev=0.91ms, min=63, max=66
1390         Info: [HTMLElement.tabIndex="foo"] avg=367.86ms, median=369.00ms, stdev=8.06ms, min=352, max=380
1391         Info: [HTMLElement.draggable="foo"] avg=79.43ms, median=79.00ms, stdev=0.58ms, min=79, max=81
1392         Info: [HTMLElement.webkitdropzone="foo"] avg=62.14ms, median=62.00ms, stdev=0.35ms, min=62, max=63
1393         Info: [HTMLElement.hidden="foo"] avg=61.67ms, median=62.00ms, stdev=0.71ms, min=61, max=64
1394         Info: [HTMLElement.accessKey="foo"] avg=61.86ms, median=62.00ms, stdev=0.56ms, min=61, max=63
1395         Info: [HTMLElement.spellcheck="foo"] avg=79.57ms, median=80.00ms, stdev=0.58ms, min=78, max=80
1396         Info: [HTMLElement.innerHTML="foo"] avg=1176.67ms, median=1169.00ms, stdev=26.60ms, min=1146, max=1246
1397         Info: [HTMLElement.innerText="foo"] avg=51.62ms, median=52.00ms, stdev=0.72ms, min=50, max=53
1398         Info: [Element.tagName] avg=79.52ms, median=80.00ms, stdev=1.05ms, min=78, max=81
1399         Info: [Element.style] avg=126.19ms, median=126.00ms, stdev=10.40ms, min=104, max=161
1400         Info: [Element.offsetLeft] avg=98.95ms, median=104.00ms, stdev=4.01ms, min=100, max=99
1401         Info: [Element.clientLeft] avg=98.33ms, median=102.00ms, stdev=2.59ms, min=100, max=99
1402         Info: [Element.scrollLeft] avg=103.29ms, median=96.00ms, stdev=15.63ms, min=100, max=97
1403         Info: [Element.dataset] avg=2448.62ms, median=2441.00ms, stdev=20.43ms, min=2426, max=2503
1404         Info: [Element.firstElementChild] avg=58.71ms, median=58.00ms, stdev=2.37ms, min=57, max=65
1405         Info: [Element.lastElementChild] avg=56.33ms, median=55.00ms, stdev=1.83ms, min=54, max=59
1406         Info: [Element.previousElementSibling] avg=52.95ms, median=52.00ms, stdev=1.70ms, min=52, max=57
1407         Info: [Element.nextElementSibling] avg=58.48ms, median=58.00ms, stdev=1.79ms, min=57, max=63
1408         Info: [Element.childElementCount] avg=64.57ms, median=64.00ms, stdev=1.81ms, min=63, max=71
1409         Info: [Element.onload] avg=98.38ms, median=98.00ms, stdev=0.49ms, min=98, max=99
1410         Info: [Element.scrollLeft="foo"] avg=95.24ms, median=95.00ms, stdev=0.92ms, min=94, max=98
1411         Info: [Element.onload="foo"] avg=67.14ms, median=67.00ms, stdev=0.56ms, min=66, max=68
1412         Info: [Node.nodeName] avg=76.29ms, median=76.00ms, stdev=0.63ms, min=75, max=77
1413         Info: [Node.nodeType] avg=53.05ms, median=52.00ms, stdev=2.61ms, min=52, max=61
1414         Info: [Node.parentNode] avg=53.52ms, median=53.00ms, stdev=1.22ms, min=52, max=56
1415         Info: [Node.childNodes] avg=118.33ms, median=123.00ms, stdev=8.85ms, min=107, max=129
1416         Info: [Node.firstChild] avg=53.62ms, median=53.00ms, stdev=1.13ms, min=51, max=55
1417         Info: [Node.lastChild] avg=53.14ms, median=53.00ms, stdev=1.21ms, min=52, max=55
1418         Info: [Node.previousSibling] avg=51.86ms, median=51.00ms, stdev=1.55ms, min=50, max=54
1419         Info: [Node.nextSibling] avg=52.86ms, median=54.00ms, stdev=1.81ms, min=50, max=55
1420         Info: [Node.attributes] avg=2027.71ms, median=2026.00ms, stdev=10.48ms, min=2010, max=2051
1421         Info: [Node.ownerDocument] avg=152.14ms, median=152.00ms, stdev=1.73ms, min=149, max=155
1422         Info: [Node.namespaceURI] avg=60.86ms, median=59.00ms, stdev=3.85ms, min=56, max=66
1423         Info: [Node.localName] avg=60.00ms, median=59.00ms, stdev=2.71ms, min=57, max=66
1424         Info: [Node.baseURI] avg=671.10ms, median=668.00ms, stdev=10.66ms, min=659, max=706
1425         Info: [Node.parentElement] avg=55.05ms, median=56.00ms, stdev=2.15ms, min=52, max=58
1426         Info: [Node.nodeValue] avg=61.24ms, median=60.00ms, stdev=2.56ms, min=59, max=68
1427         Info: [Node.prefix] avg=60.90ms, median=58.00ms, stdev=4.26ms, min=57, max=68
1428         Info: [Node.textContent] avg=149.71ms, median=149.00ms, stdev=1.12ms, min=148, max=152
1429         Info: [Node.nodeValue="foo"] avg=23.62ms, median=23.00ms, stdev=0.72ms, min=23, max=25
1430         Info: [Node.prefix="foo"] avg=49.57ms, median=49.00ms, stdev=0.73ms, min=49, max=51
1431         Info: [Node.textContent="foo"] avg=145.81ms, median=140.00ms, stdev=17.43ms, min=138, max=199
1432
1433         * Bindings/dom_attributes.html: Added.
1434         * resources/runner.js:
1435         (PerfTestRunner.info):
1436
1437 2012-02-22  Kentaro Hara  <haraken@chromium.org>
1438
1439         Add performance tests for the Element.id getter and setter
1440         https://bugs.webkit.org/show_bug.cgi?id=79209
1441
1442         Reviewed by Adam Barth.
1443
1444         This patch adds performance tests for the Element.id getter and setter.
1445
1446         The results of 10 runs in my local Chromium/Linux are as follows:
1447
1448         dom_attributes_id_getter.html:
1449         median= 92.0 ms, stdev= 1.65151445649 ms, min= 89.0 ms, max= 96.0 ms
1450         median= 87.5 ms, stdev= 1.84390889146 ms, min= 85.0 ms, max= 92.0 ms
1451         median= 94.0 ms, stdev= 1.46543508898 ms, min= 90.0 ms, max= 95.0 ms
1452         median= 86.0 ms, stdev= 1.66958078571 ms, min= 84.0 ms, max= 91.0 ms
1453         median= 94.0 ms, stdev= 2.25610283454 ms, min= 103.0 ms, max= 98.0 ms
1454         median= 86.0 ms, stdev= 1.49666295471 ms, min= 84.0 ms, max= 91.0 ms
1455         median= 86.0 ms, stdev= 1.5321553446 ms, min= 84.0 ms, max= 90.0 ms
1456         median= 86.0 ms, stdev= 2.20227155455 ms, min= 84.0 ms, max= 94.0 ms
1457         median= 96.0 ms, stdev= 1.74570902501 ms, min= 101.0 ms, max= 99.0 ms
1458         median= 86.0 ms, stdev= 1.63935963108 ms, min= 84.0 ms, max= 90.0 ms
1459         median= 94.0 ms, stdev= 1.49248115566 ms, min= 90.0 ms, max= 95.0 ms
1460         median= 88.0 ms, stdev= 5.20192272146 ms, min= 84.0 ms, max= 98.0 ms
1461         median= 86.0 ms, stdev= 1.41067359797 ms, min= 84.0 ms, max= 89.0 ms
1462         median= 86.0 ms, stdev= 1.48660687473 ms, min= 84.0 ms, max= 89.0 ms
1463         median= 94.0 ms, stdev= 1.58113883008 ms, min= 93.0 ms, max= 97.0 ms
1464         median= 86.0 ms, stdev= 1.95959179423 ms, min= 84.0 ms, max= 93.0 ms
1465         median= 86.0 ms, stdev= 1.8547236991 ms, min= 84.0 ms, max= 92.0 ms
1466         median= 86.0 ms, stdev= 1.39283882772 ms, min= 84.0 ms, max= 89.0 ms
1467         median= 88.0 ms, stdev= 1.48660687473 ms, min= 85.0 ms, max= 90.0 ms
1468         median= 86.5 ms, stdev= 2.3 ms, min= 84.0 ms, max= 93.0 ms
1469
1470         dom_attributes_id_setter.html:
1471         median= 123.0 ms, stdev= 2.06397674406 ms, min= 119.0 ms, max= 128.0 ms
1472         median= 119.0 ms, stdev= 1.92028643697 ms, min= 117.0 ms, max= 123.0 ms
1473         median= 121.0 ms, stdev= 1.8781639971 ms, min= 118.0 ms, max= 123.0 ms
1474         median= 123.0 ms, stdev= 2.21585198062 ms, min= 119.0 ms, max= 130.0 ms
1475         median= 124.5 ms, stdev= 2.36378933071 ms, min= 119.0 ms, max= 129.0 ms
1476         median= 122.0 ms, stdev= 2.4474476501 ms, min= 118.0 ms, max= 128.0 ms
1477         median= 123.0 ms, stdev= 1.76351920885 ms, min= 118.0 ms, max= 125.0 ms
1478         median= 119.0 ms, stdev= 1.79652442232 ms, min= 117.0 ms, max= 125.0 ms
1479         median= 121.0 ms, stdev= 1.49916643506 ms, min= 119.0 ms, max= 123.0 ms
1480         median= 119.5 ms, stdev= 1.85135085816 ms, min= 117.0 ms, max= 125.0 ms
1481         median= 124.0 ms, stdev= 1.43090880213 ms, min= 120.0 ms, max= 124.0 ms
1482         median= 122.0 ms, stdev= 1.45859521458 ms, min= 118.0 ms, max= 123.0 ms
1483         median= 123.0 ms, stdev= 1.50332963784 ms, min= 120.0 ms, max= 124.0 ms
1484         median= 123.0 ms, stdev= 1.3638181697 ms, min= 120.0 ms, max= 125.0 ms
1485         median= 119.0 ms, stdev= 1.43527000944 ms, min= 117.0 ms, max= 121.0 ms
1486         median= 122.5 ms, stdev= 1.71682847134 ms, min= 118.0 ms, max= 124.0 ms
1487         median= 121.0 ms, stdev= 1.44481832768 ms, min= 118.0 ms, max= 123.0 ms
1488         median= 120.0 ms, stdev= 1.40978721799 ms, min= 117.0 ms, max= 122.0 ms
1489         median= 122.0 ms, stdev= 1.68522995464 ms, min= 119.0 ms, max= 125.0 ms
1490         median= 122.0 ms, stdev= 1.55804364509 ms, min= 118.0 ms, max= 123.0 ms
1491
1492
1493         * Bindings/dom_attributes_id_getter.html: Added.
1494         * Bindings/dom_attributes_id_setter.html: Added.
1495         * Bindings/resources/dom_attributes_common.js: Added.
1496
1497 2012-02-16  Alexis Menard  <alexis.menard@openbossa.org>
1498
1499         Add a perf test for the CSS parser.
1500         https://bugs.webkit.org/show_bug.cgi?id=78114
1501
1502         Reviewed by Ryosuke Niwa.
1503
1504         Add a perf test for the CSS parser. It load a big CSS chunk
1505         which contains some snippets of the YUI library from Yahoo!.
1506
1507         * Parser/css-parser-yui.html: Added.
1508
1509 2012-02-08  Hajime Morrita  <morrita@chromium.org>
1510
1511         [PerformanceTests] sunspider-crypto-md5.html and sunspider-crypto-sha1.html are timing out on cr-mac perf bot
1512         https://bugs.webkit.org/show_bug.cgi?id=78174
1513
1514         Reviewed by Ryosuke Niwa.
1515
1516         Pulled back removed files.
1517
1518         * Dromaeo/resources/dromaeo/web/tests/sunspider-crypto-md5.html: Added.
1519         * Dromaeo/resources/dromaeo/web/tests/sunspider-crypto-sha1.html: Added.
1520
1521 2012-02-02  Hajime Morrita  <morrita@chromium.org>
1522
1523         [PerformanceTests] Each Dromaeo test needs its landing html.
1524         https://bugs.webkit.org/show_bug.cgi?id=77504
1525
1526         Reviewed by Ryosuke Niwa.
1527
1528         - Added landing pages for each test which are listed in MANIFEST.json
1529         - Removed some tests which depends on the library whose license is incompatible to WebKit.
1530         - Added local copy of corresponding library. And replace the references to the remote sources
1531           to the local copies.
1532         - Removed old library files under Dromaeo/resources/dromaeo/web/lib/.
1533         - Skipped Sunsupier and v8 test. Such JS centric test should have its own test instead of
1534           running as a part of Dromaeo.
1535
1536         * Dromaeo/cssquery-dojo.html: Added.
1537         * Dromaeo/cssquery-jquery.html: Added.
1538         * Dromaeo/cssquery-prototype.html: Added.
1539         * Dromaeo/dom-attr.html: Added.
1540         * Dromaeo/dom-modify.html: Added.
1541         * Dromaeo/dom-traverse.html: Added.
1542         * Dromaeo/dromaeo-3d-cube.html: Added.
1543         * Dromaeo/dromaeo-core-eval.html: Added.
1544         * Dromaeo/dromaeo-object-array.html: Added.
1545         * Dromaeo/dromaeo-object-regexp.html: Added.
1546         * Dromaeo/dromaeo-object-string.html: Added.
1547         * Dromaeo/dromaeo-string-base64.html: Added.
1548         * Dromaeo/jslib-attr-jquery.html: Added.
1549         * Dromaeo/jslib-attr-prototype.html: Added.
1550         * Dromaeo/jslib-event-jquery.html: Added.
1551         * Dromaeo/jslib-event-prototype.html: Added.
1552         * Dromaeo/jslib-modify-jquery.html: Added.
1553         * Dromaeo/jslib-modify-prototype.html: Added.
1554         * Dromaeo/jslib-style-jquery.html: Added.
1555         * Dromaeo/jslib-style-prototype.html: Added.
1556         * Dromaeo/jslib-traverse-jquery.html: Added.
1557         * Dromaeo/jslib-traverse-prototype.html: Added.
1558         * Dromaeo/resources/dromaeo/web/lib/dojo.js: Removed.
1559         * Dromaeo/resources/dromaeo/web/lib/dojo-1.6.1.js: Added.
1560         * Dromaeo/resources/dromaeo/web/lib/jquery.js: Removed.
1561         * Dromaeo/resources/dromaeo/web/lib/jquery-1.6.4.js: Added.
1562         * Dromaeo/resources/dromaeo/web/lib/mootools.js: Removed.
1563         * Dromaeo/resources/dromaeo/web/lib/prototype.js: Removed.
1564         * Dromaeo/resources/dromaeo/web/lib/prototype-1.7.js: Added.
1565         * Dromaeo/resources/dromaeo/web/tests/cssquery-dojo.html:
1566         * Dromaeo/resources/dromaeo/web/tests/cssquery-jquery.html:
1567         * Dromaeo/resources/dromaeo/web/tests/cssquery-mootools.html: Removed.
1568         * Dromaeo/resources/dromaeo/web/tests/cssquery-prototype.html:
1569         * Dromaeo/resources/dromaeo/web/tests/jslib-attr-jquery.html:
1570         * Dromaeo/resources/dromaeo/web/tests/jslib-attr-prototype.html:
1571         * Dromaeo/resources/dromaeo/web/tests/jslib-event-jquery.html:
1572         * Dromaeo/resources/dromaeo/web/tests/jslib-event-prototype.html:
1573         * Dromaeo/resources/dromaeo/web/tests/jslib-modify-jquery.html:
1574         * Dromaeo/resources/dromaeo/web/tests/jslib-modify-prototype.html:
1575         * Dromaeo/resources/dromaeo/web/tests/jslib-style-jquery.html:
1576         * Dromaeo/resources/dromaeo/web/tests/jslib-style-prototype.html:
1577         * Dromaeo/resources/dromaeo/web/tests/jslib-traverse-jquery.html:
1578         * Dromaeo/resources/dromaeo/web/tests/jslib-traverse-prototype.html:
1579         * Dromaeo/sunspider-3d-morph.html: Added.
1580         * Dromaeo/sunspider-3d-raytrace.html: Added.
1581         * Dromaeo/sunspider-access-binary-trees.html: Added.
1582         * Dromaeo/sunspider-access-fannkuch.html: Added.
1583         * Dromaeo/sunspider-access-nbody.html: Added.
1584         * Dromaeo/sunspider-access-nsieve.html: Added.
1585         * Dromaeo/sunspider-bitops-3bit-bits-in-byte.html: Added.
1586         * Dromaeo/sunspider-bitops-bits-in-byte.html: Added.
1587         * Dromaeo/sunspider-bitops-bitwise-and.html: Added.
1588         * Dromaeo/sunspider-bitops-nsieve-bits.html: Added.
1589         * Dromaeo/sunspider-controlflow-recursive.html: Added.
1590         * Dromaeo/sunspider-crypto-aes.html: Added.
1591         * Dromaeo/sunspider-crypto-md5.html: Added.
1592         * Dromaeo/sunspider-crypto-sha1.html: Added.
1593         * Dromaeo/sunspider-date-format-tofte.html: Added.
1594         * Dromaeo/sunspider-date-format-xparb.html: Added.
1595         * Dromaeo/sunspider-math-cordic.html: Added.
1596         * Dromaeo/sunspider-math-partial-sums.html: Added.
1597         * Dromaeo/sunspider-math-spectral-norm.html: Added.
1598         * Dromaeo/sunspider-regexp-dna.html: Added.
1599         * Dromaeo/sunspider-string-fasta.html: Added.
1600         * Dromaeo/sunspider-string-tagcloud.html: Added.
1601         * Dromaeo/sunspider-string-unpack-code.html: Added.
1602         * Dromaeo/sunspider-string-validate-input.html: Added.
1603         * Dromaeo/v8-crypto.html: Added.
1604         * Dromaeo/v8-deltablue.html: Added.
1605         * Dromaeo/v8-earley-boyer.html: Added.
1606         * Dromaeo/v8-raytrace.html: Added.
1607         * Dromaeo/v8-richards.html: Added.
1608         * Skipped:
1609
1610 2012-02-02  Sheriff Bot  <webkit.review.bot@gmail.com>
1611
1612         Unreviewed, rolling out r106543.
1613         http://trac.webkit.org/changeset/106543
1614         https://bugs.webkit.org/show_bug.cgi?id=77671
1615
1616         Made tests less stable (Requested by rniwa on #webkit).
1617
1618         * resources/init.html: Removed.
1619
1620 2012-02-02  Ryosuke Niwa  <rniwa@webkit.org>
1621
1622         [PerformanceTests] tests have dependencies
1623         https://bugs.webkit.org/show_bug.cgi?id=77506
1624
1625         Reviewed by Hajime Morita.
1626
1627         * resources/init.html: Added.
1628
1629 2012-02-01  Ryosuke Niwa  <rniwa@webkit.org>
1630
1631         Fix after r106386. The empty anchor element was generating an erroneous whitespace.
1632
1633         * Bindings/event-target-wrapper.html:
1634
1635 2012-02-01  Ryosuke Niwa  <rniwa@webkit.org>
1636
1637         Another fix after r106379.
1638
1639         * Parser/html5-full-render.html:
1640
1641 2012-01-31  Hajime Morrita  <morrita@chromium.org>
1642
1643         Import Dromaeo to WebKit repository.
1644         https://bugs.webkit.org/show_bug.cgi?id=77328
1645
1646         Reviewed by Ryosuke Niwa.
1647
1648         Imported a modified version of Dromaeo which is able to
1649         run on the DRT.
1650
1651         The imported version originally came from:
1652         https://github.com/omo/dromaeo/commit/c942a1452109cdd5f759b3c4aee8d68d490d44a2
1653
1654         * Dromaeo/resources/dromaeo/web/application.css: Added.
1655         * Dromaeo/resources/dromaeo/web/favicon.ico: Added.
1656         * Dromaeo/resources/dromaeo/web/favicon.png: Added.
1657         * Dromaeo/resources/dromaeo/web/htmlrunner.js: Added.
1658         * Dromaeo/resources/dromaeo/web/ie.css: Added.
1659         * Dromaeo/resources/dromaeo/web/images/bg.png: Added.
1660         * Dromaeo/resources/dromaeo/web/images/clouds.png: Added.
1661         * Dromaeo/resources/dromaeo/web/images/clouds2.png: Added.
1662         * Dromaeo/resources/dromaeo/web/images/comets.png: Added.
1663         * Dromaeo/resources/dromaeo/web/images/dino1.png: Added.
1664         * Dromaeo/resources/dromaeo/web/images/dino2.png: Added.
1665         * Dromaeo/resources/dromaeo/web/images/dino3.png: Added.
1666         * Dromaeo/resources/dromaeo/web/images/dino4.png: Added.
1667         * Dromaeo/resources/dromaeo/web/images/dino5.png: Added.
1668         * Dromaeo/resources/dromaeo/web/images/dino6.png: Added.
1669         * Dromaeo/resources/dromaeo/web/images/dino7.png: Added.
1670         * Dromaeo/resources/dromaeo/web/images/dino8.png: Added.
1671         * Dromaeo/resources/dromaeo/web/images/left.png: Added.
1672         * Dromaeo/resources/dromaeo/web/images/logo.png: Added.
1673         * Dromaeo/resources/dromaeo/web/images/logo2.png: Added.
1674         * Dromaeo/resources/dromaeo/web/images/logo3.png: Added.
1675         * Dromaeo/resources/dromaeo/web/images/right.png: Added.
1676         * Dromaeo/resources/dromaeo/web/images/top.png: Added.
1677         * Dromaeo/resources/dromaeo/web/images/water.png: Added.
1678         * Dromaeo/resources/dromaeo/web/index.html: Added.
1679         * Dromaeo/resources/dromaeo/web/jquery.js: Added.
1680         * Dromaeo/resources/dromaeo/web/lib/dojo.js: Added.
1681         * Dromaeo/resources/dromaeo/web/lib/jquery.js: Added.
1682         * Dromaeo/resources/dromaeo/web/lib/mootools.js: Added.
1683         * Dromaeo/resources/dromaeo/web/lib/prototype.js: Added.
1684         * Dromaeo/resources/dromaeo/web/pngfix.js: Added.
1685         * Dromaeo/resources/dromaeo/web/reset.css: Added.
1686         * Dromaeo/resources/dromaeo/web/test-head.html: Added.
1687         * Dromaeo/resources/dromaeo/web/test-head.js: Added.
1688         * Dromaeo/resources/dromaeo/web/test-tail.html: Added.
1689         * Dromaeo/resources/dromaeo/web/test-tail.js: Added.
1690         * Dromaeo/resources/dromaeo/web/tests/MANIFEST.json: Added.
1691         * Dromaeo/resources/dromaeo/web/tests/cssquery-dojo.html: Added.
1692         * Dromaeo/resources/dromaeo/web/tests/cssquery-jquery.html: Added.
1693         * Dromaeo/resources/dromaeo/web/tests/cssquery-mootools.html: Added.
1694         * Dromaeo/resources/dromaeo/web/tests/cssquery-prototype.html: Added.
1695         * Dromaeo/resources/dromaeo/web/tests/cssquery-yui.html: Added.
1696         * Dromaeo/resources/dromaeo/web/tests/dom-attr.html: Added.
1697         * Dromaeo/resources/dromaeo/web/tests/dom-modify.html: Added.
1698         * Dromaeo/resources/dromaeo/web/tests/dom-query.html: Added.
1699         * Dromaeo/resources/dromaeo/web/tests/dom-traverse.html: Added.
1700         * Dromaeo/resources/dromaeo/web/tests/dromaeo-3d-cube.html: Added.
1701         * Dromaeo/resources/dromaeo/web/tests/dromaeo-core-eval.html: Added.
1702         * Dromaeo/resources/dromaeo/web/tests/dromaeo-object-array.html: Added.
1703         * Dromaeo/resources/dromaeo/web/tests/dromaeo-object-regexp.html: Added.
1704         * Dromaeo/resources/dromaeo/web/tests/dromaeo-object-string.html: Added.
1705         * Dromaeo/resources/dromaeo/web/tests/dromaeo-string-base64.html: Added.
1706         * Dromaeo/resources/dromaeo/web/tests/jslib-attr-jquery.html: Added.
1707         * Dromaeo/resources/dromaeo/web/tests/jslib-attr-prototype.html: Added.
1708         * Dromaeo/resources/dromaeo/web/tests/jslib-event-jquery.html: Added.
1709         * Dromaeo/resources/dromaeo/web/tests/jslib-event-prototype.html: Added.
1710         * Dromaeo/resources/dromaeo/web/tests/jslib-modify-jquery.html: Added.
1711         * Dromaeo/resources/dromaeo/web/tests/jslib-modify-prototype.html: Added.
1712         * Dromaeo/resources/dromaeo/web/tests/jslib-style-jquery.html: Added.
1713         * Dromaeo/resources/dromaeo/web/tests/jslib-style-prototype.html: Added.
1714         * Dromaeo/resources/dromaeo/web/tests/jslib-traverse-jquery.html: Added.
1715         * Dromaeo/resources/dromaeo/web/tests/jslib-traverse-prototype.html: Added.
1716         * Dromaeo/resources/dromaeo/web/tests/sunspider-3d-morph.html: Added.
1717         * Dromaeo/resources/dromaeo/web/tests/sunspider-3d-raytrace.html: Added.
1718         * Dromaeo/resources/dromaeo/web/tests/sunspider-access-binary-trees.html: Added.
1719         * Dromaeo/resources/dromaeo/web/tests/sunspider-access-fannkuch.html: Added.
1720         * Dromaeo/resources/dromaeo/web/tests/sunspider-access-nbody.html: Added.
1721         * Dromaeo/resources/dromaeo/web/tests/sunspider-access-nsieve.html: Added.
1722         * Dromaeo/resources/dromaeo/web/tests/sunspider-bitops-3bit-bits-in-byte.html: Added.
1723         * Dromaeo/resources/dromaeo/web/tests/sunspider-bitops-bits-in-byte.html: Added.
1724         * Dromaeo/resources/dromaeo/web/tests/sunspider-bitops-bitwise-and.html: Added.
1725         * Dromaeo/resources/dromaeo/web/tests/sunspider-bitops-nsieve-bits.html: Added.
1726         * Dromaeo/resources/dromaeo/web/tests/sunspider-controlflow-recursive.html: Added.
1727         * Dromaeo/resources/dromaeo/web/tests/sunspider-crypto-aes.html: Added.
1728         * Dromaeo/resources/dromaeo/web/tests/sunspider-crypto-md5.html: Added.
1729         * Dromaeo/resources/dromaeo/web/tests/sunspider-crypto-sha1.html: Added.
1730         * Dromaeo/resources/dromaeo/web/tests/sunspider-date-format-tofte.html: Added.
1731         * Dromaeo/resources/dromaeo/web/tests/sunspider-date-format-xparb.html: Added.
1732         * Dromaeo/resources/dromaeo/web/tests/sunspider-math-cordic.html: Added.
1733         * Dromaeo/resources/dromaeo/web/tests/sunspider-math-partial-sums.html: Added.
1734         * Dromaeo/resources/dromaeo/web/tests/sunspider-math-spectral-norm.html: Added.
1735         * Dromaeo/resources/dromaeo/web/tests/sunspider-regexp-dna.html: Added.
1736         * Dromaeo/resources/dromaeo/web/tests/sunspider-string-fasta.html: Added.
1737         * Dromaeo/resources/dromaeo/web/tests/sunspider-string-tagcloud.html: Added.
1738         * Dromaeo/resources/dromaeo/web/tests/sunspider-string-unpack-code.html: Added.
1739         * Dromaeo/resources/dromaeo/web/tests/sunspider-string-validate-input.html: Added.
1740         * Dromaeo/resources/dromaeo/web/tests/v8-crypto.html: Added.
1741         * Dromaeo/resources/dromaeo/web/tests/v8-deltablue.html: Added.
1742         * Dromaeo/resources/dromaeo/web/tests/v8-earley-boyer.html: Added.
1743         * Dromaeo/resources/dromaeo/web/tests/v8-raytrace.html: Added.
1744         * Dromaeo/resources/dromaeo/web/tests/v8-richards.html: Added.
1745         * Dromaeo/resources/dromaeo/web/tests/72px.png: Added.
1746         * Dromaeo/resources/dromaeo/web/web-style.css: Added.
1747         * Dromaeo/resources/dromaeo/web/webrunner.js: Added.
1748         * Skipped: Unskip the dromaeo driver.
1749
1750 2012-01-31  Ryosuke Niwa  <rniwa@webkit.org>
1751
1752         Fix Parser/html5-full-render.html after r106379.
1753
1754         * Parser/html5-full-render.html:
1755
1756 2012-01-31  Ryosuke Niwa  <rniwa@webkit.org>
1757
1758         PerfTestRunner should automatically create pre#log
1759         https://bugs.webkit.org/show_bug.cgi?id=77469
1760
1761         Reviewed by Tony Chang.
1762
1763         Automatically generate <pre id="log"></pre> if there isn't one.
1764
1765         * Bindings/event-target-wrapper.html:
1766         * DOM/Accessors.html:
1767         * DOM/CloneNodes.html:
1768         * DOM/CreateNodes.html:
1769         * DOM/DOMDivWalk.html:
1770         * DOM/DOMTable.html:
1771         * DOM/DOMWalk.html:
1772         * DOM/Events.html:
1773         * DOM/GetElement.html:
1774         * DOM/GridSort.html:
1775         * DOM/Template.html:
1776         * Dromaeo/dom-query.html:
1777         * Dromaeo/resources/dromaeorunner.js:
1778         * Parser/html-parser.html:
1779         * Parser/html5-full-render.html:
1780         * Parser/simple-url.html:
1781         * Parser/tiny-innerHTML.html:
1782         * Parser/url-parser.html:
1783         * Parser/xml-parser.html:
1784         * resources/runner.js:
1785         (PerfTestRunner.log):
1786         (PerfTestRunner.printStatistics):
1787
1788 2012-01-31  Ryosuke Niwa  <rniwa@webkit.org>
1789
1790         runner.js in performance tests should define a class
1791         https://bugs.webkit.org/show_bug.cgi?id=77074
1792
1793         Reviewed by Eric Seidel.
1794
1795         Wrap all functions in runner.js by PerfTestRunner and update tests that runner.js accordingly.
1796         Also replace compute* functions in runner.js by more robust code from dom-perf.js.
1797
1798         * Bindings/event-target-wrapper.html:
1799         * DOM/DOMTable.html:
1800         * DOM/resources/dom-perf.js:
1801         (BenchmarkSuite.prototype.RunSingle):
1802         (runBenchmarkSuite):
1803         * Mutation/append-child-deep.html:
1804         * Mutation/append-child.html:
1805         * Mutation/inner-html.html:
1806         * Mutation/remove-child-deep.html:
1807         * Mutation/remove-child.html:
1808         * Parser/html-parser.html:
1809         * Parser/html5-full-render.html:
1810         * Parser/simple-url.html:
1811         * Parser/tiny-innerHTML.html:
1812         * Parser/url-parser.html:
1813         * Parser/xml-parser.html:
1814         * resources/runner.js:
1815         (PerfTestRunner.log):
1816         (PerfTestRunner.logInfo):
1817         (PerfTestRunner.loadFile):
1818         (PerfTestRunner.computeStatistics):
1819         (PerfTestRunner.logStatistics):
1820         (PerfTestRunner._runLoop.else):
1821         (PerfTestRunner._runLoop):
1822         (PerfTestRunner._runner):
1823
1824 2012-01-31  Hajime Morrita  <morrita@chromium.org>
1825
1826         [PerformanceTests] Add landing html for Dromaeo dom-query test
1827         https://bugs.webkit.org/show_bug.cgi?id=77329
1828
1829         Reviewed by Ryosuke Niwa.
1830
1831         Added a "landing html" which includes actual dromaemo page and
1832         send messages to interact with it. The landing html also dumps the
1833         score in run-perf-tests friendly format.
1834
1835         This test isn't enabled until Dromaemo itself is available for run-perf-tests.
1836         It will happen in a separate change.
1837
1838         * Dromaeo/dom-query.html: Added.
1839         * Dromaeo/resources/dromaeorunner.js: Added.
1840         * Skipped: Skipping this for now.
1841
1842 2012-01-30  Ryosuke Niwa  <rniwa@webkit.org>
1843
1844         Skip inspector tests since they have been timing out.
1845
1846         * Skipped:
1847
1848 2012-01-29  Ryosuke Niwa  <rniwa@webkit.org>
1849
1850         DOM/DOMDivWalk.html result is unreliable
1851         https://bugs.webkit.org/show_bug.cgi?id=77313
1852
1853         Reviewed by Adam Barth.
1854
1855         Trigger garbage collection manually between test runs to avoid GC
1856         to occur in the middle of runs.
1857
1858         * DOM/resources/dom-perf.js:
1859         (BenchmarkSuite.prototype.RunSingle):
1860         * resources/runner.js:
1861         (gc.else.gcRec):
1862         (gc):
1863         (runLoop):
1864
1865 2012-01-26  Ryosuke Niwa  <rniwa@webkit.org>
1866
1867         Import Chromium's dom_perf test
1868         https://bugs.webkit.org/show_bug.cgi?id=77175
1869
1870         Reviewed by Adam Barth.
1871
1872         Import dom_perf.
1873
1874         Note resources/dom/suites.js isn't used by any html file yet but it will be used by Chromium port
1875         once its perf bots start pulling test files from WebKit repository instead of Google's internal repository.
1876
1877         * DOM: Added.
1878         * DOM/Accessors.html: Added.
1879         * DOM/CloneNodes.html: Added.
1880         * DOM/CreateNodes.html: Added.
1881         * DOM/DOMDivWalk.html: Added.
1882         * DOM/DOMTable.html: Added.
1883         * DOM/DOMWalk.html: Added.
1884         * DOM/Events.html: Added.
1885         * DOM/GetElement.html: Added.
1886         * DOM/GridSort.html: Added.
1887         * DOM/Template.html: Added.
1888         * DOM/resources: Added.
1889         * DOM/resources/dom-perf.js: Added.
1890         * DOM/resources/dom-perf: Added.
1891         * DOM/resources/dom-perf/accessors.js: Added.
1892         * DOM/resources/dom-perf/clonenodes.js: Added.
1893         * DOM/resources/dom-perf/createnodes.js: Added.
1894         * DOM/resources/dom-perf/domdivwalk.js: Added.
1895         * DOM/resources/dom-perf/domtable.js: Added.
1896         * DOM/resources/dom-perf/domwalk.js: Added.
1897         * DOM/resources/dom-perf/events.js: Added.
1898         * DOM/resources/dom-perf/getelement.js: Added.
1899         * DOM/resources/dom-perf/gridsort.js: Added.
1900         * DOM/resources/dom-perf/suites.js: Added.
1901         * DOM/resources/dom-perf/template.js: Added.
1902         * resources/runner.js:
1903
1904 2012-01-25  Ryosuke Niwa  <rniwa@webkit.org>
1905
1906         html5-full-render.html fails due to a log
1907         https://bugs.webkit.org/show_bug.cgi?id=77046
1908
1909         Reviewed by Adam Barth.
1910
1911         Replace the call to log() in html5-full-render.html by a call to newly added logInfo(),
1912         which doesn't print anything inside DRT.
1913
1914         * Parser/html5-full-render.html:
1915         * resources/runner.js:
1916         (logInfo):
1917
1918 2012-01-19  Adam Barth  <abarth@webkit.org>
1919
1920         PerformanceTests's runner.js shouldn't be Parser-specific
1921         https://bugs.webkit.org/show_bug.cgi?id=76670
1922
1923         Reviewed by Ryosuke Niwa.
1924
1925         This runner script is used by a bunch of difference performance tests.
1926         It shouldn't be in the Parser directory anymore.
1927
1928         * Bindings/event-target-wrapper.html:
1929         * Mutation/append-child-deep.html:
1930         * Mutation/append-child.html:
1931         * Mutation/inner-html.html:
1932         * Mutation/remove-child-deep.html:
1933         * Mutation/remove-child.html:
1934         * Parser/html-parser.html:
1935         * Parser/html5-full-render.html:
1936         * Parser/resources/runner.js: Removed.
1937         * Parser/simple-url.html:
1938         * Parser/tiny-innerHTML.html:
1939         * Parser/url-parser.html:
1940         * Parser/xml-parser.html:
1941         * resources: Added.
1942         * resources/runner.js: Copied from PerformanceTests/Parser/resources/runner.js.
1943
1944 2012-01-19  Ryosuke Niwa  <rniwa@webkit.org>
1945
1946         Some perf tests time out when ran by run-perf-tests
1947         https://bugs.webkit.org/show_bug.cgi?id=76612
1948
1949         Reviewed by Dirk Pranke and Eric Seidel.
1950
1951         Replace all images in html5.html by geenbox.png to avoid accessing whatwg.org when
1952         running the parser tests. Also call dumpAsText, waitUntilDone, and notifyDone automatically
1953         inside runner.js to avoid having to call them in individual tests.
1954
1955         * Bindings/event-target-wrapper.html: Removed calls to layoutTestController methods since
1956         they are now called by runner.js automatically.
1957         * Parser/resources/greenbox.png: Copied from LayoutTests/fast/css/resources/greenbox.png.
1958         * Parser/resources/html5.html:
1959         * Parser/resources/runner.js:
1960         (runLoop):
1961
1962 2012-01-18  Ryosuke Niwa  <rniwa@webkit.org>
1963
1964         run-perf-tests should support Skipped list
1965         https://bugs.webkit.org/show_bug.cgi?id=76594
1966
1967         Reviewed by Adam Barth.
1968
1969         Add Skipped list to be used by run-perf-tests as it can only runs tests in
1970         Bindings, Parser, and inspector at the moment.
1971
1972         * Skipped: Added.
1973
1974 2012-01-12  Ilya Tikhonovsky  <loislo@chromium.org>
1975
1976         Web Inspector: performance: restore 'log 300 messages into console' test.
1977         https://bugs.webkit.org/show_bug.cgi?id=76170
1978
1979         It was removed in order of transition from layout tests to perf tests.
1980
1981         Reviewed by Yury Semikhatsky.
1982
1983         * inspector/console-300-lines.html: Added.
1984
1985 2012-01-12  Ilya Tikhonovsky  <loislo@chromium.org>
1986
1987         Web Inspector: performance tests: fix memory leak in first-open-resources test.
1988         https://bugs.webkit.org/show_bug.cgi?id=76049
1989
1990         first-open-resources doesn't reset resource panel properly after test run.
1991
1992         Reviewed by Yury Semikhatsky.
1993
1994         * inspector/first-open-resources.html:
1995
1996 2012-01-10  Ilya Tikhonovsky  <loislo@chromium.org>
1997
1998         Unreviewed. WebInspector: Remove unnecessary logging from the test.
1999
2000         * inspector/first-open-resources.html:
2001
2002 2012-01-10  Ilya Tikhonovsky  <loislo@chromium.org>
2003
2004         Unreviewed. Restore test that was removed during migration from LayoutTests folder.
2005
2006         * inspector/first-open-resources.html: Added.
2007
2008 2012-01-09  Ilya Tikhonovsky  <loislo@chromium.org>
2009
2010         Unreviewed test fix after r103683.
2011
2012         * inspector/first-open-scripts.html:
2013         * inspector/show-panel.html:
2014
2015 2012-01-03  Adam Barth  <abarth@webkit.org>
2016
2017         html-parser.html takes too long to run
2018         https://bugs.webkit.org/show_bug.cgi?id=75515
2019
2020         Reviewed by Eric Seidel.
2021
2022         In working on runner.js, we introduced a factor of 10 increase to the
2023         running time of this benchmark.  This patch dials down the number of
2024         iterations to something more user-friendly.
2025
2026         * Parser/html-parser.html:
2027
2028 2011-12-18  Ilya Tikhonovsky  <loislo@chromium.org>
2029
2030         Web Inspector: move heap data stats to separate tab.
2031         https://bugs.webkit.org/show_bug.cgi?id=74704
2032
2033         Reviewed by Pavel Feldman.
2034
2035         * inspector/performance-test.js:
2036         (initialize_TimeTracker.InspectorTest.runPerformanceTest.Timer.prototype._dump):
2037         (initialize_TimeTracker.InspectorTest.runPerformanceTest):
2038
2039 2011-12-01  Ilya Tikhonovsky  <loislo@chromium.org>
2040
2041         Web Inspector: chromium: move and adapt Inspector's performance tests for running with run-inspector-perf-tests.py.
2042         https://bugs.webkit.org/show_bug.cgi?id=72260
2043
2044         Reviewed by Pavel Feldman.
2045
2046         * inspector/first-open-elements.html: Renamed from LayoutTests/inspector/performance/resources/first-open-elements.html.
2047         * inspector/first-open-scripts.html: Renamed from LayoutTests/inspector/performance/resources/first-open-scripts.html.
2048         * inspector/inspector-startup-time.html: Renamed from LayoutTests/inspector/performance/resources/inspector-startup-time.html.
2049         * inspector/network-append-30-requests.html: Renamed from LayoutTests/inspector/performance/resources/network-append-30-requests.html.
2050         * inspector/performance-test.js: Renamed from LayoutTests/inspector/performance/resources/performance-test.js.
2051         (initialize_TimeTracker.InspectorTest.runPerformanceTest.Timer):
2052         (initialize_TimeTracker.InspectorTest.runPerformanceTest.Timer.prototype.start):
2053         (initialize_TimeTracker.InspectorTest.runPerformanceTest.Timer.prototype.finish):
2054         (initialize_TimeTracker.InspectorTest.runPerformanceTest.Timer.prototype._getJSHeapSize):
2055         (initialize_TimeTracker.InspectorTest.runPerformanceTest.Timer.prototype.done):
2056         (initialize_TimeTracker.InspectorTest.runPerformanceTest.Timer.prototype._runTest):
2057         (initialize_TimeTracker.InspectorTest.runPerformanceTest.Timer.prototype._dump):
2058         (initialize_TimeTracker.InspectorTest.runPerformanceTest):
2059         (initialize_TimeTracker.InspectorTest.mark):
2060         * inspector/show-panel.html: Renamed from LayoutTests/inspector/performance/resources/show-panel.html.
2061
2062 2011-11-09  Adam Klein  <adamk@chromium.org>
2063
2064         [MutationObservers] Microbenchmarks for appendChild, removeChild, and innerHTML
2065         https://bugs.webkit.org/show_bug.cgi?id=71939
2066
2067         Reviewed by Ojan Vafai.
2068
2069         These benchmarks time both the DOM mutations themselves and the time taken
2070         to call the MutationCallback. This measurement is achieved by
2071         executing the code entirely within the MutationCallback itself.
2072
2073         * Mutation/append-child-deep.html: Added. Covers appendChild in a 200-node-deep tree with subtree observation.
2074         * Mutation/append-child.html: Added. Covers appendChild in a flat tree.
2075         * Mutation/inner-html.html: Added. Covers innerHTML adding and removing many nodes at a time.
2076         * Mutation/remove-child-deep.html: Added. Covers removeChild in a 200-node-deep tree with subtree observation.
2077         * Mutation/remove-child.html: Added. Covers removeChild in a flat tree.
2078
2079 2011-10-25  Adam Barth  <abarth@webkit.org>
2080
2081         EventTargetFactory.in is not sorted
2082         https://bugs.webkit.org/show_bug.cgi?id=70855
2083
2084         Reviewed by Eric Seidel.
2085
2086         Add a microbenchmark for wrapping event target.
2087
2088         * Bindings: Added.
2089         * Bindings/event-target-wrapper.html: Added.
2090         * Parser/resources/runner.js:
2091         (runLoop):
2092             - Let the user of the harness get a callback when the test is done.
2093
2094 2011-10-04  Eric Seidel  <eric@webkit.org>
2095
2096         Update html5-full-render.html to load the HTML5 spec incrementally, closer to how the browser would
2097         https://bugs.webkit.org/show_bug.cgi?id=69374
2098
2099         Unreviewed.
2100
2101         Per https://bugs.webkit.org/show_bug.cgi?id=69374#c7
2102         updating the chunksize to be a bit smaller now that the total
2103         size of the html5 spec is smaller than when I wrote the test.
2104
2105         This should increase the prevalence of style resolution in the
2106         sample time, which more closely matches instrument samples
2107         taken when loading the full page in Safari.
2108
2109         * Parser/html5-full-render.html:
2110
2111 2011-10-04  Eric Seidel  <eric@webkit.org>
2112
2113         Update html5-full-render.html to load the HTML5 spec incrementally, closer to how the browser would
2114         https://bugs.webkit.org/show_bug.cgi?id=69374
2115
2116         Reviewed by James Robinson.
2117
2118         This should finally be able to provide us with a repeatable metric
2119         for how fast we're currently able to load the HTML5 spec.
2120         There are a variety of interesting functions which show up in this
2121         sample, including of course style resolution.
2122
2123         * Parser/html5-full-render.html:
2124
2125 2011-10-04  Eric Seidel  <eric@webkit.org>
2126
2127         Update our copy of the HTML5 spec used for performance testing to match the latest version
2128         https://bugs.webkit.org/show_bug.cgi?id=69364
2129
2130         Unreviewed.  The patch is too large to post, so landing this unreviewed. :(
2131
2132         The copy of the html5 spec we were using for testing was over a year old
2133         and didn't have any of the subresources with it (we were only using it
2134         as an HTML parsing benchmark).  I'm about to use it for testing
2135         the full load/render time of the benchmark, so I've updated to the latest
2136         version of the spec (now 8mb instead of 5mb) and inlined two of the
2137         CSS files.  (Inlining the CSS was necessary to make my testing consistent
2138         as there seems to be a race with some timer fired after CSS load.)
2139
2140         Increasing from 5mb to 8mb showed a similar 30% slowdown in our parser
2141         benchmarks as expected.
2142
2143         * Parser/resources/html5.html:
2144
2145 2011-10-04  Eric Seidel  <eric@webkit.org>
2146
2147         Add loopsPerRun option to Parser performance test runner
2148         https://bugs.webkit.org/show_bug.cgi?id=69363
2149
2150         Reviewed by Antti Koivisto.
2151
2152         Add loopsPerRun option and abstract out runLoop function (to make the UI more responsive).
2153         No change in behavior in any of the tests.
2154
2155         * Parser/resources/runner.js:
2156         (runLoop):
2157         (run):
2158         (start):
2159
2160 2011-10-03  Eric Seidel  <eric@webkit.org>
2161
2162         Add a microbenchmark for a full-page render of the HTML5 spec
2163         https://bugs.webkit.org/show_bug.cgi?id=69285
2164
2165         Reviewed by Adam Barth.
2166
2167         This was designed to be a test for https://bugs.webkit.org/show_bug.cgi?id=68944
2168         But it seems that by far our dominating cost for the HTML5 benchmark is
2169         time spent laying out lines (which isn't actually that surprising).
2170
2171         I'm adding the performance test for posterity.
2172
2173         * Parser/html-parser.html:
2174          - Removed use of "about:blank" which abarth says is a no-op.  Made comment more accurate.
2175         * Parser/html5-full-render.html: Copied from PerformanceTests/Parser/html-parser.html.
2176
2177 2011-10-03  Eric Seidel  <eric@webkit.org>
2178
2179         PerformanceTests/Parser/html-parser is only testing parsing of the head element
2180         https://bugs.webkit.org/show_bug.cgi?id=69283
2181
2182         Reviewed by Adam Barth.
2183
2184         While investigating https://bugs.webkit.org/show_bug.cgi?id=68944
2185         I found that the html-parser benchmark was only parsing up to the
2186         first script tag per loop!  We've fixed this by adding the sandbox
2187         tag which will deny all external loads and allow the parser to
2188         synchronously continue parsing the entire document to completion
2189         (as we had expected it was doing).
2190
2191         This changes the profile somewhat.  Line number counting is much
2192         hotter, since we're actually accounting for the parse of the entire
2193         document in our sample.
2194
2195         Total sample time only about doubles, from 1800ms to 3800ms on my machine
2196         which is less than I would have expected.
2197
2198         * Parser/html-parser.html:
2199
2200 2011-08-09  Alexandru Chiculita  <achicu@adobe.com>
2201
2202         Fixing 65868 REGRESSION(r92610) caused by 65668 - Optimize floating elements lookup
2203         https://bugs.webkit.org/show_bug.cgi?id=65871
2204
2205         Reviewed by Dave Hyatt.
2206
2207         * Layout/floats.html: Added the nested divs, so that we can test the propagation impact of the floats tree.
2208
2209 2011-08-08  Sheriff Bot  <webkit.review.bot@gmail.com>
2210
2211         Unreviewed, rolling out r92610.
2212         http://trac.webkit.org/changeset/92610
2213         https://bugs.webkit.org/show_bug.cgi?id=65868
2214
2215         Caused assertion failures when running fast/multicol tests
2216         (Requested by andersca on #webkit).
2217
2218         * Layout/floats.html:
2219
2220 2011-08-08  Alexandru Chiculita  <achicu@adobe.com>
2221
2222         Optimize floating elements lookup
2223         https://bugs.webkit.org/show_bug.cgi?id=65668
2224
2225         Reviewed by David Hyatt.
2226
2227         * Layout/floats.html: Added the nested divs, so that we can test the propagation impact of the floats tree.
2228
2229 2011-08-04  Alexandru Chiculita  <achicu@adobe.com>
2230
2231         Add a performance test for floating elements layout algorithm
2232         https://bugs.webkit.org/show_bug.cgi?id=65741
2233
2234         The test creating lots of small floats and randomly changes
2235         the width of one element at a time. It measures how often it can do that.
2236
2237         Reviewed by Adam Barth.
2238
2239         * Layout/floats.html: Added.
2240
2241 2011-05-30  Daniel Bates  <dbates@webkit.org>
2242
2243         Reviewed by Adam Barth.
2244
2245         Rename XSSFilter to XSSAuditor
2246         https://bugs.webkit.org/show_bug.cgi?id=61718
2247
2248         Currently we use the names XSSFilter and XSSAuditor throughout the project.
2249         Instead, we should choose one name for consistency.
2250
2251         No functionality was changed. So, no new tests.
2252
2253         * XSSAuditor: Copied from PerformanceTests/XSSFilter.
2254         * XSSFilter: Removed.
2255         * XSSFilter/large-post-many-events.html: Removed.
2256         * XSSFilter/large-post-many-inline-scripts-and-events.html: Removed.
2257         * XSSFilter/resources: Removed.
2258         * XSSFilter/resources/target-for-large-post-many-inline-scripts-and-events.html: Removed.
2259
2260 2011-03-25  Leo Yang  <leo.yang@torchmobile.com.cn>
2261
2262         Reviewed by Dirk Schulze.
2263
2264         SVG <use> element performance improvement
2265         https://bugs.webkit.org/show_bug.cgi?id=57077
2266
2267         Add a manual test case which is from
2268         http://upload.wikimedia.org/wikipedia/commons/4/4e/Sierpinski_carpet_6.svg
2269         for svg <use> element performance test.
2270
2271         * PageLoad/svg/files/Sierpinski_carpet_6.svg: Added.
2272         * PageLoad/svg/svg.pltsuite:
2273
2274 2011-02-05  Eric Seidel  <eric@webkit.org>
2275
2276         Reviewed by Adam Barth.
2277
2278         Add performance tests for URL parsing
2279         https://bugs.webkit.org/show_bug.cgi?id=53845
2280
2281         The final-url-en URL corpus is from:
2282         http://corpus.leeds.ac.uk/internet.html
2283
2284         There are also other language corpuses which we may want to use in the future.
2285
2286         The usage instructions were as follows:
2287         "URL lists and other resources: you can freely use them in your research
2288         provided that you supply a link to this website: http://corpus.leeds.ac.uk/."
2289         Done.
2290
2291         * Parser/simple-url.html: Added.
2292          On my machine:
2293          Firefox: 255ms
2294          TOT WebKit: 378ms
2295          Chrome: 286ms
2296          Opera: 2830ms (no, that is not a typo)
2297         * Parser/url-parser.html: Added.
2298          Firefox: 381ms
2299          TOT WebKit: 216ms
2300          Chrome: 131ms
2301          Opera: 1383ms (again, not a typo)
2302         * Parser/resources/final-url-en: Added.
2303
2304 2011-02-03  Adam Barth  <abarth@webkit.org>
2305
2306         Reviewed by Daniel Bates.
2307
2308         Add another XSSFilter PerformanceTest
2309         https://bugs.webkit.org/show_bug.cgi?id=53750
2310
2311         This PerformanceTest tests the case that was slow in
2312         https://bugs.webkit.org/show_bug.cgi?id=49845
2313
2314         * XSSFilter/large-post-many-inline-scripts-and-events.html: Added.
2315         * XSSFilter/resources: Added.
2316         * XSSFilter/resources/target-for-large-post-many-inline-scripts-and-events.html: Added.
2317
2318 2011-02-03  Adam Barth  <abarth@webkit.org>
2319
2320         Reviewed by Eric Seidel.
2321
2322         Add PerformanceTest for XSSFilter
2323         https://bugs.webkit.org/show_bug.cgi?id=53741
2324
2325         This performance tests covers the case of a large POST data and many
2326         small event handlers.
2327
2328         * XSSFilter: Added.
2329         * XSSFilter/large-post-many-events.html: Added.
2330
2331 2011-02-02  Adam Barth  <abarth@webkit.org>
2332
2333         Rubber-stamped by Eric Seidel.
2334
2335         Merge PerformanceTests/Parser/ChangeLog and
2336         PerformanceTests/PageLoad/ChangeLog into PerformanceTests/ChangeLog.
2337
2338         It's silly to have ChangeLogs for each of these directories separately.
2339         I've left SunSpider with its own ChangeLog because it's more of an
2340         independent entity.
2341
2342         * ChangeLog: Added.
2343
2344 2011-02-02  Eric Seidel  <eric@webkit.org>
2345
2346         Unreviewed.  Just fixing an exception seen in Firefox.
2347
2348         HTML5 TreeBuilder regressed a Peacekeeper DOM test by 40%
2349         https://bugs.webkit.org/show_bug.cgi?id=48719
2350
2351         Make the benchmarks work in Firefox/Opera.
2352
2353         * resources/runner.js:
2354         (log):
2355
2356 2011-01-27  Eric Seidel  <eric@webkit.org>
2357
2358         Reviewed by Darin Adler.
2359
2360         HTML5 TreeBuilder regressed a Peacekeeper DOM test by 40%
2361         https://bugs.webkit.org/show_bug.cgi?id=48719
2362
2363         It's unclear exactly what the Peacekeeper benchmark is testing,
2364         because I haven't found a way to run it myself.
2365
2366         However, I constructed a benchmark which shows at least one possible slow point.
2367         The HTML5 spec talks about creating a new document for every time we use
2368         the fragment parsing algorithm.  Document() it turns out, it a huge bloated
2369         mess, and the constructor and destructor do a huge amount of work.
2370
2371         * benchmarks/parser/tiny-innerHTML.html: Added.
2372
2373 2011-01-29  Sheriff Bot  <webkit.review.bot@gmail.com>
2374
2375         Unreviewed, rolling out r77050.
2376         http://trac.webkit.org/changeset/77050
2377         https://bugs.webkit.org/show_bug.cgi?id=53371
2378
2379         Caused a crash in Chromium's test_shell_tests (Requested by
2380         rniwa on #webkit).
2381
2382         * resources/performance-test.js: Removed.
2383         * tiny-innerHTML.html: Removed.
2384
2385 2011-01-28  Eric Seidel  <eric@webkit.org>
2386
2387         Reviewed by Darin Adler.
2388
2389         HTML5 TreeBuilder regressed a Peacekeeper DOM test by 40%
2390         https://bugs.webkit.org/show_bug.cgi?id=48719
2391
2392         It's unclear exactly what the Peacekeeper benchmark is testing,
2393         because I haven't found a way to run it myself.
2394
2395         However, I constructed a benchmark which shows at least one possible slow point.
2396         The HTML5 spec talks about creating a new document for every time we use
2397         the fragment parsing algorithm.  Document() it turns out, it a huge bloated
2398         mess, and the constructor and destructor do a huge amount of work.
2399         To avoid constructing (or destructing) documents for each innerHTML call,
2400         this patch adds a shared dummy document used by all innerHTML calls.
2401
2402         * benchmarks/parser/tiny-innerHTML.html: Added.
2403
2404 2010-12-31  Adam Barth  <abarth@webkit.org>
2405
2406         Rubber-stamped by Eric Seidel.
2407
2408         Move HTML and XML parser benchmarks into PerformanceTests/Parser
2409         https://bugs.webkit.org/show_bug.cgi?id=51772
2410
2411         Add a ChangeLog for tracking changes to the Parser PerformanceTest.
2412
2413         * ChangeLog: Added.
2414
2415 2010-12-31  Adam Barth  <abarth@webkit.org>
2416
2417         Rubber-stamped by Eric Seidel.
2418
2419         Move PageLoadTests to PerformanceTests/PageLoad
2420         https://bugs.webkit.org/show_bug.cgi?id=51771
2421
2422         Update URLs to point to the new directory name.
2423
2424         * svg/svg.pltsuite:
2425
2426 2006-12-26  Eric Seidel  <eric@webkit.org>
2427
2428         Reviewed by olliej.
2429
2430         * svg/svg.pltsuite: re-enable word-iso.svg after fixing http://bugs.webkit.org/show_bug.cgi?id=11987
2431
2432 2006-12-26  Eric Seidel  <eric@webkit.org>
2433
2434         Reviewed by bradee-oh.
2435
2436         Add new PageLoadTests directory (this one)
2437         Add LICENSES file to explain where each SVG came from.
2438
2439         * ChangeLog: Added.
2440         * svg/LICENSES: Added.
2441         * svg/files/33041-Samurai.svg: Added.
2442         * svg/files/42470-flower_from_my_garden_v2.svg: Added.
2443         * svg/files/Harvey_Rayner.svg: Added.
2444         * svg/files/az-lizard_benji_park_01.svg: Added.
2445         * svg/files/bamboo_01.svg: Added.
2446         * svg/files/cacuts_01.svg: Added.
2447         * svg/files/cowboy.svg: Added.
2448         * svg/files/crawfish2_ganson.svg: Added.
2449         * svg/files/deb9frac1.svg: Added.
2450         * svg/files/food_leif_lodahl_01.svg: Added.
2451         * svg/files/france.svg: Added.
2452         * svg/files/francobollo_gnome_ezechi_02.svg: Added.
2453         * svg/files/gearflowers.svg: Added.
2454         * svg/files/hereGear4.svg: Added.
2455         * svg/files/mtsthelens.svg: Added.
2456         * svg/files/mtsthelens0.jpg: Added.
2457         * svg/files/world-iso.svg: Added.
2458         * svg/files/worldcup.svg: Added.
2459         * svg/svg.pltsuite: Added.
2460