[INTL] Add PerformanceTests for Intl objects
[WebKit-https.git] / PerformanceTests / ChangeLog
1 2017-05-04  Andy VanWagoner  <thetalecrafter@gmail.com>
2
3         [INTL] Add PerformanceTests for Intl objects
4         https://bugs.webkit.org/show_bug.cgi?id=171695
5
6         Reviewed by Benjamin Poulain.
7
8         * Intl/collator-compare-all-options.html: Added.
9         * Intl/collator-compare-default.html: Added.
10         * Intl/collator-create-all-options.html: Added.
11         * Intl/collator-create-default.html: Added.
12         * Intl/datetimeformat-create-all-options.html: Added.
13         * Intl/datetimeformat-create-default.html: Added.
14         * Intl/datetimeformat-format-all-options.html: Added.
15         * Intl/datetimeformat-format-default.html: Added.
16         * Intl/numberformat-create-all-options.html: Added.
17         * Intl/numberformat-create-default.html: Added.
18         * Intl/numberformat-format-all-options.html: Added.
19         * Intl/numberformat-format-default.html: Added.
20
21 2017-04-27  Michael Saboff  <msaboff@apple.com>
22
23         Add back tests inadvertently deleted in r205032.
24
25         Rubber stamped by Geoffrey Garen.
26
27         * MallocBench/run-malloc-benchmarks:
28
29 2017-04-25  Myles C. Maxfield  <mmaxfield@apple.com>
30
31         Add performance test for FontCache::systemFallbackForCharacters()
32         https://bugs.webkit.org/show_bug.cgi?id=170842
33
34         Reviewed by Tim Horton.
35
36         * Layout/word-joiner.html: Added.
37
38 2017-04-24  Saam Barati  <sbarati@apple.com>
39
40         Add ML to ARES6
41         https://bugs.webkit.org/show_bug.cgi?id=171206
42
43         Rubber stamped by Filip Pizlo.
44
45         This patch adds a new test to ARES6 called ML. ML is an implementation of
46         a feedforward neural network: https://github.com/mljs/feedforward-neural-networks.
47         It makes heavy use of classes, and does non-trivial matrix math using the
48         ml-matrix library: https://github.com/mljs/matrix
49
50         * ARES-6/about.html:
51         * ARES-6/cli.js:
52         * ARES-6/glue.js:
53         * ARES-6/index.html:
54         * ARES-6/ml: Added.
55         * ARES-6/ml/benchmark.js: Added.
56         * ARES-6/ml/index.js: Added.
57         * ARES-6/ml_benchmark.js: Added.
58
59 2017-04-21  Zalan Bujtas  <zalan@apple.com>
60
61         Simple line layout: Add performance test to measure mid-word line breaking with long text.
62         https://bugs.webkit.org/show_bug.cgi?id=171136
63
64         Reviewed by Ryosuke Niwa.
65
66         * Layout/simple-line-layout-word-break-long-text.html: Added.
67
68 2017-04-17  Ryan Haddad  <ryanhaddad@apple.com>
69
70         Unreviewed, rolling out r215366.
71
72         This test is failing on performance bots.
73
74         Reverted changeset:
75
76         "Add performance test for asking the platform for a font for
77         U+2060 WORD JOINER"
78         https://bugs.webkit.org/show_bug.cgi?id=170842
79         http://trac.webkit.org/changeset/215366
80
81 2017-04-14  Myles C. Maxfield  <mmaxfield@apple.com>
82
83         Add performance test for asking the platform for a font for U+2060 WORD JOINER
84         https://bugs.webkit.org/show_bug.cgi?id=170842
85
86         Reviewed by Tim Horton.
87
88         * Layout/word-joiner.html: Added.
89
90 2017-04-03  Saam Barati  <sbarati@apple.com>
91
92         Add a new test to ARES6
93         https://bugs.webkit.org/show_bug.cgi?id=170077
94
95         Rubber stamped by Filip Pizlo.
96
97         This patch adds a new test to ARES6. The test is called Babylon. Babylon is
98         the JavaScript parser that is part of the Babel transpiler. The test runs
99         the parser on 4 different JS scripts.
100
101         * ARES-6/Babylon: Added.
102         * ARES-6/Babylon/AUTHORS: Added.
103         * ARES-6/Babylon/air-blob.js: Added.
104         (Reg):
105         (Reg.fromReg):
106         (Reg.prototype.get index):
107         (Reg.prototype.get type):
108         (Reg.prototype.get name):
109         (Reg.prototype.get isCalleeSave):
110         (Reg.prototype.get isReg):
111         (Reg.prototype.hash):
112         (Reg.prototype.toString):
113         (Reg.extract):
114         (Reg.forEachFast):
115         (Reg.forEach):
116         (newGPR):
117         (Reg.gprs.Reg.fprs.Reg.calleeSaveGPRs.Reg.calleeSaveFPRs.Reg.calleeSaves):
118         * ARES-6/Babylon/babylon-blob.js: Added.
119         (export.default.Parser):
120         (export.default.Parser.prototype.isReservedWord):
121         (export.default.Parser.prototype.hasPlugin):
122         (export.default.Parser.prototype.extend):
123         (export.default.Parser.prototype.loadPlugins):
124         (export.default.Parser.prototype.parse):
125         * ARES-6/Babylon/basic-blob.js: Added.
126         (Basic.NumberApply):
127         (Basic.Variable):
128         (Basic.Const):
129         (Basic.NumberPow):
130         (Basic.NumberMul):
131         (Basic.NumberDiv):
132         (Basic.NumberNeg):
133         (Basic.NumberAdd):
134         (Basic.NumberSub):
135         (Basic.StringVar):
136         (Basic.Equals):
137         (Basic.NotEquals):
138         (Basic.LessThan):
139         (Basic.GreaterThan):
140         (Basic.LessEqual):
141         (Basic.GreaterEqual):
142         (Basic.GoTo):
143         (Basic.GoSub):
144         (Basic.Def):
145         (Basic.Let):
146         (Basic.If):
147         (Basic.Return):
148         (Basic.Stop):
149         (Basic.On):
150         (sideState.shouldStop):
151         (Basic.For):
152         (Basic.Next):
153         (Basic.Print):
154         (Basic.Input):
155         (Basic.Read):
156         (Basic.Restore):
157         (Basic.Dim):
158         (Basic.Randomize):
159         (Basic.End):
160         (Basic.Program):
161         * ARES-6/Babylon/benchmark.js: Added.
162         (this.performance.performance.now.currentTime):
163         (else.this.preciseTime.currentTime):
164         (else.currentTime):
165         (BabylonBenchmark.appendSource):
166         (BabylonBenchmark):
167         (BabylonBenchmark.prototype.runIteration.parse):
168         (BabylonBenchmark.prototype.runIteration.parseExpression):
169         (BabylonBenchmark.prototype.runIteration):
170         (runBenchmark):
171         * ARES-6/Babylon/index.js: Added.
172         (isIdentifierStart):
173         (isIdentifierChar):
174         (isNewLine):
175         (Position):
176         (SourceLocation):
177         (getLineInfo):
178         (TokenType):
179         (KeywordTokenType):
180         (BinopTokenType):
181         (TokContext):
182         (tt.parenR.updateContext.tt.braceR.updateContext):
183         (tt.name.updateContext):
184         (tt.braceL.updateContext):
185         (tt.dollarBraceL.updateContext):
186         (tt.parenL.updateContext):
187         (tt.incDec.updateContext):
188         (tt._function.updateContext):
189         (tt.backQuote.updateContext):
190         (State.prototype.init):
191         (State.prototype.curPosition):
192         (State.prototype.clone):
193         (State):
194         (Token):
195         (codePointToString):
196         (Tokenizer):
197         (Tokenizer.prototype.next):
198         (Tokenizer.prototype.eat):
199         (Tokenizer.prototype.match):
200         (Tokenizer.prototype.isKeyword):
201         (Tokenizer.prototype.lookahead):
202         (Tokenizer.prototype.setStrict):
203         (Tokenizer.prototype.curContext):
204         (Tokenizer.prototype.nextToken):
205         (Tokenizer.prototype.readToken):
206         (Tokenizer.prototype.fullCharCodeAtPos):
207         (Tokenizer.prototype.pushComment):
208         (Tokenizer.prototype.skipBlockComment):
209         (Tokenizer.prototype.skipLineComment):
210         (Tokenizer.prototype.skipSpace):
211         (Tokenizer.prototype.finishToken):
212         (Tokenizer.prototype.readToken_dot):
213         (Tokenizer.prototype.readToken_slash):
214         (Tokenizer.prototype.readToken_mult_modulo):
215         (Tokenizer.prototype.readToken_pipe_amp):
216         (Tokenizer.prototype.readToken_caret):
217         (Tokenizer.prototype.readToken_plus_min):
218         (Tokenizer.prototype.readToken_lt_gt):
219         (Tokenizer.prototype.readToken_eq_excl):
220         (Tokenizer.prototype.getTokenFromCode):
221         (Tokenizer.prototype.finishOp):
222         (Tokenizer.prototype.readRegexp):
223         (Tokenizer.prototype.readInt):
224         (Tokenizer.prototype.readRadixNumber):
225         (Tokenizer.prototype.readNumber):
226         (Tokenizer.prototype.readCodePoint):
227         (Tokenizer.prototype.readString):
228         (Tokenizer.prototype.readTmplToken):
229         (Tokenizer.prototype.readEscapedChar):
230         (Tokenizer.prototype.readHexChar):
231         (Tokenizer.prototype.readWord1):
232         (Tokenizer.prototype.readWord):
233         (Tokenizer.prototype.braceIsBlock):
234         (Tokenizer.prototype.updateContext):
235         (getOptions):
236         (prototype.isReservedWord):
237         (prototype.hasPlugin):
238         (prototype.extend):
239         (prototype.loadPlugins):
240         (prototype.parse):
241         (last):
242         (pp.addComment):
243         (pp.processComment):
244         (pp.checkPropClash):
245         (pp.getExpression):
246         (pp.parseExpression):
247         (pp.parseMaybeAssign):
248         (pos.set i):
249         * ARES-6/Babylon/inspector-blob.js: Added.
250         (WebInspector.ScriptSyntaxTree):
251         (WebInspector.ScriptSyntaxTree.prototype.get parsedSuccessfully):
252         (WebInspector.ScriptSyntaxTree.prototype.forEachNode):
253         (WebInspector.ScriptSyntaxTree.prototype.filter):
254         (WebInspector.ScriptSyntaxTree.prototype.containersOfOffset):
255         (WebInspector.ScriptSyntaxTree.prototype.filterByRange.filterForNodesInRange):
256         (WebInspector.ScriptSyntaxTree.prototype.filterByRange):
257         (WebInspector.ScriptSyntaxTree.prototype.containsNonEmptyReturnStatement.removeFunctionsFilter):
258         (WebInspector.ScriptSyntaxTree.prototype.containsNonEmptyReturnStatement):
259         (WebInspector.ScriptSyntaxTree.functionReturnDivot):
260         (WebInspector.ScriptSyntaxTree.prototype.updateTypes):
261         (WebInspector.ScriptSyntaxTree.prototype._gatherIdentifiersInDeclaration.gatherIdentifiers):
262         (WebInspector.ScriptSyntaxTree.prototype._gatherIdentifiersInDeclaration):
263         (WebInspector.ScriptSyntaxTree.prototype._defaultParserState):
264         (WebInspector.ScriptSyntaxTree.prototype._recurse):
265         (WebInspector.ScriptSyntaxTree.prototype._recurseArray):
266         (WebInspector.ScriptSyntaxTree.prototype._createInternalSyntaxTree):
267         * ARES-6/about.html:
268         * ARES-6/babylon_benchmark.js: Added.
269         * ARES-6/cli.js:
270         * ARES-6/glue.js:
271         * ARES-6/index.html:
272         * ARES-6/results.js:
273         (Results.prototype.reportResult):
274         (Results.prototype.reportResult.averageAbovePercentile): Deleted.
275         * ARES-6/styles.css:
276         (.test):
277         (@media only screen and (max-width: 784px)):
278
279 2017-03-29  Zalan Bujtas <zalan@apple.com>
280
281         [Arabic text] Add whitespace preserve/word wrap performance test.
282         https://bugs.webkit.org/show_bug.cgi?id=170243
283
284         Reviewed by Simon Fraser.
285
286         * Layout/ArabicLineLayout-pre-and-wrap.html: Added.
287
288 2017-03-08  Yusuke Suzuki  <utatane.tea@gmail.com>
289
290         [JSC] Add helpers to run Six Speed to run-jsc-benchmarks
291         https://bugs.webkit.org/show_bug.cgi?id=169351
292
293         Reviewed by Saam Barati.
294
295         * SixSpeed/wrappers/wrapper.js: Added.
296         (assertEqual):
297         (test):
298         (jscRun):
299
300 2017-03-08  Jon Davis  <jond@apple.com>
301
302         Fixed ARES-6 animations for other browsers
303
304         Rubber stamped by Saam Barati.
305
306         * ARES-6/styles.css:
307         (.swoop):
308         (.logo):
309
310 2017-03-07  Filip Pizlo  <fpizlo@apple.com>
311
312         Revise the ARES-6 explainer text
313         https://bugs.webkit.org/show_bug.cgi?id=169287
314
315         Reviewed by Saam Barati.
316         
317         Just did some copy-editing.
318
319         * ARES-6/about.html:
320
321 2017-03-06  Jon Davis  <jond@apple.com>
322
323         Refined the ARES-6 experience
324         https://bugs.webkit.org/show_bug.cgi?id=169205
325
326         Reviewed by Saam Barati.
327
328         * ARES-6/about.html: Copied from PerformanceTests/ARES-6/index.html.
329         * ARES-6/driver.js:
330         (Driver.prototype.start):
331         (Driver.prototype._updateIterations):
332         (Driver):
333         * ARES-6/index.html:
334         * ARES-6/stats.js:
335         (Stats.prototype.toString.span.span):
336         * ARES-6/styles.css:
337         (body):
338         (.swoop):
339         (p):
340         (.logo):
341         (.start):
342         (.test):
343         (.test:nth-child(odd)):
344         (.overall):
345         (.about h2):
346         (.button):
347         (.button:hover):
348         (.button.return:before):
349         (@keyframes fade-in):
350         (@keyframes scale-in):
351         (@keyframes flicker-in):
352         (@keyframes wipe-in):
353         (.flip): Deleted.
354         (.testing, .about): Deleted.
355         (.testing): Deleted.
356         (.about): Deleted.
357         (#about:target): Deleted.
358
359 2017-03-06  Jeremy Jones  <jeremyj@apple.com>
360
361         StandardCanPlayThrough.html does not work on iOS.
362         https://bugs.webkit.org/show_bug.cgi?id=169207
363
364         Reviewed by Jer Noble.
365
366         On iOS only, add muted=true and autoplay=allowed. The loading policy on iOS requires these attributes to load enough
367         data to trigger canPlayThrough.
368
369         * Media/StandardCanPlayThrough.html:
370
371 2017-02-13  Zalan Bujtas <zalan@apple.com>
372
373         Simple line layout: Add performance test with text-rendering: optimizeSpeed
374         https://bugs.webkit.org/show_bug.cgi?id=168248
375
376         Reviewed by Simon Fraser.
377
378         * Layout/simple-line-layout-with-varying-content-and-optimized-speed.html: Added.
379
380 2017-02-10  Zalan Bujtas  <zalan@apple.com>
381
382         Mail hangs when removing multiple rows from large table.
383         https://bugs.webkit.org/show_bug.cgi?id=168103
384         <rdar://problem/30090186>
385
386         Reviewed by Ryosuke Niwa.
387
388         * DOM/large-table-edit.html: Added.
389
390 2017-02-05  Filip Pizlo  <fpizlo@apple.com>
391
392         Beef up the ARES-6 explainer text
393         https://bugs.webkit.org/show_bug.cgi?id=167867
394
395         Reviewed by Saam Barati.
396         
397         Added appropriate links.
398
399         Added text that explains the subscores.
400
401         * ARES-6/index.html:
402
403 2017-02-06  Jer Noble  <jer.noble@apple.com>
404
405         Playback stalls when a SourceBuffer append causes frame eviction
406         https://bugs.webkit.org/show_bug.cgi?id=167834
407
408         Reviewed by Eric Carlson.
409
410         Add an in-page performance test measuring the amount of time required
411         to append a large amount of media data to a SourceBuffer, and then to
412         completely remove that data 30s at a time.
413
414         Add a microbenchmark for MediaTime which measures the amount of time
415         required to create a 1M entry std::map and traverse the map 1M times.
416
417         * Media/MSERemoveCodedFrames.html: Added.
418         * Media/media-source-loader.js:
419         (MediaSourceLoader.prototype.get duration):
420         * MediaTime/Configurations/Base.xcconfig: Added.
421         * MediaTime/Configurations/DebugRelease.xcconfig: Added.
422         * MediaTime/Makefile: Added.
423         * MediaTime/MediaTime.xcodeproj/project.pbxproj: Added.
424         * MediaTime/main.cpp: Added.
425         (performTest):
426         (test):
427         (main):
428         * Skipped:
429
430 2017-02-06  Saam Barati  <sbarati@apple.com>
431
432         Make ARES-6 work from the CLI again
433         https://bugs.webkit.org/show_bug.cgi?id=167895
434
435         Reviewed by Michael Saboff.
436
437         * ARES-6/driver.js:
438         (Driver.prototype.readyTrigger):
439         (Driver.prototype.disableTrigger):
440         * ARES-6/stats.js:
441         (Stats.prototype.toString.span.span):
442         (Stats.prototype.toString.return.span):
443         (Stats.prototype.toString):
444
445 2017-02-06  Ryan Haddad  <ryanhaddad@apple.com>
446
447         Skip ARES-6 on performance bots
448         https://bugs.webkit.org/show_bug.cgi?id=167863
449
450         Unreviewed test gardening.
451
452         ES6SampleBench was renamed to ARES-6, so this should be updated in PerformanceTests/Skipped as well.
453
454         * Skipped:
455
456 2017-02-05  Filip Pizlo  <fpizlo@apple.com>
457
458         Change ES6SampleBench into ARES-6
459         https://bugs.webkit.org/show_bug.cgi?id=167863
460
461         Rubber stamped by Saam Barati.
462         
463         This imports changes that Jonathan Davis (jond@apple.com) made to ES6SampleBench to make it look
464         super awesome. Its now called ARES-6.
465
466         * ARES-6: Copied from PerformanceTests/ES6SampleBench.
467         * ARES-6/ARES-6.svg: Added.
468         * ARES-6/driver.js:
469         (Driver):
470         (Driver.prototype.readyTrigger):
471         (Driver.prototype.disableTrigger):
472         (Driver.prototype.start):
473         (Driver.prototype._updateIterations):
474         * ARES-6/glue.js:
475         * ARES-6/index.html:
476         * ARES-6/results.js:
477         (Results.prototype.reportRunning):
478         (Results.prototype.reportDone):
479         * ARES-6/stats.js:
480         (Stats.prototype.toString.return.span):
481         (Stats.prototype.toString.span.span):
482         (Stats.prototype.toString): Deleted.
483         * ARES-6/style.css: Removed.
484         * ARES-6/styles.css: Added.
485         (html):
486         (body):
487         (header,):
488         (p a):
489         (p a:hover):
490         (.flip):
491         (header):
492         (h2,):
493         (p):
494         (.about h2):
495         (.logo):
496         (.start):
497         (.start.ready):
498         (.start.ready:hover):
499         (.start:after):
500         (.start.ready .testrun):
501         (.tests):
502         (.test):
503         (.overall h2,):
504         (.test .indicator):
505         (.test .indicator.running):
506         (@keyframes test-running):
507         (100%):
508         (.score):
509         (.score label):
510         (.score .value):
511         (.score .units):
512         (.score .margin):
513         (.overall):
514         (.overall .score):
515         (.overall .margin):
516         (.testing, .about):
517         (.testing):
518         (.about):
519         (#about:target):
520         (@media only screen and (max-width: 784px)):
521         * ARES-6/swoop.svg: Added.
522         * ES6SampleBench: Removed.
523         * ES6SampleBench/Air: Removed.
524         * ES6SampleBench/Air/README.md: Removed.
525         * ES6SampleBench/Air/airjs-tests.yaml: Removed.
526         * ES6SampleBench/Air/all.js: Removed.
527         * ES6SampleBench/Air/allocate_stack.js: Removed.
528         * ES6SampleBench/Air/arg.js: Removed.
529         * ES6SampleBench/Air/basic_block.js: Removed.
530         * ES6SampleBench/Air/benchmark.js: Removed.
531         * ES6SampleBench/Air/code.js: Removed.
532         * ES6SampleBench/Air/custom.js: Removed.
533         * ES6SampleBench/Air/frequented_block.js: Removed.
534         * ES6SampleBench/Air/insertion_set.js: Removed.
535         * ES6SampleBench/Air/inst.js: Removed.
536         * ES6SampleBench/Air/liveness.js: Removed.
537         * ES6SampleBench/Air/make_dist.sh: Removed.
538         * ES6SampleBench/Air/opcode.js: Removed.
539         * ES6SampleBench/Air/payload-airjs-ACLj8C.js: Removed.
540         * ES6SampleBench/Air/payload-gbemu-executeIteration.js: Removed.
541         * ES6SampleBench/Air/payload-imaging-gaussian-blur-gaussianBlur.js: Removed.
542         * ES6SampleBench/Air/payload-typescript-scanIdentifier.js: Removed.
543         * ES6SampleBench/Air/reg.js: Removed.
544         * ES6SampleBench/Air/stack_slot.js: Removed.
545         * ES6SampleBench/Air/stress-test.js: Removed.
546         * ES6SampleBench/Air/strip-hash.rb: Removed.
547         * ES6SampleBench/Air/symbols.js: Removed.
548         * ES6SampleBench/Air/test.html: Removed.
549         * ES6SampleBench/Air/test.js: Removed.
550         * ES6SampleBench/Air/tmp.js: Removed.
551         * ES6SampleBench/Air/tmp_base.js: Removed.
552         * ES6SampleBench/Air/util.js: Removed.
553         * ES6SampleBench/Basic: Removed.
554         * ES6SampleBench/Basic/ast.js: Removed.
555         * ES6SampleBench/Basic/basic-tests.yaml: Removed.
556         * ES6SampleBench/Basic/basic.js: Removed.
557         * ES6SampleBench/Basic/benchmark.js: Removed.
558         * ES6SampleBench/Basic/caseless_map.js: Removed.
559         * ES6SampleBench/Basic/lexer.js: Removed.
560         * ES6SampleBench/Basic/number.js: Removed.
561         * ES6SampleBench/Basic/parser.js: Removed.
562         * ES6SampleBench/Basic/random.js: Removed.
563         * ES6SampleBench/Basic/state.js: Removed.
564         * ES6SampleBench/Basic/stress-test.js: Removed.
565         * ES6SampleBench/Basic/test.html: Removed.
566         * ES6SampleBench/Basic/test.js: Removed.
567         * ES6SampleBench/Basic/util.js: Removed.
568         * ES6SampleBench/air_benchmark.js: Removed.
569         * ES6SampleBench/basic_benchmark.js: Removed.
570         * ES6SampleBench/cli.js: Removed.
571         * ES6SampleBench/driver.js: Removed.
572         * ES6SampleBench/glue.js: Removed.
573         * ES6SampleBench/index.html: Removed.
574         * ES6SampleBench/results.js: Removed.
575         * ES6SampleBench/stats.js: Removed.
576         * ES6SampleBench/style.css: Removed.
577
578 2017-02-05  Zalan Bujtas  <zalan@apple.com>
579
580         Simple line layout: Add performance test with varying string and font-size to minimize width-cache hit.
581         https://bugs.webkit.org/show_bug.cgi?id=167859
582         <rdar://problem/30368070>
583
584         Reviewed by Antti Koivisto.
585
586         This is in preparation for webkit.org/b/167843.
587
588         * Layout/simple-line-layout-with-varying-content.html: Added.
589
590 2017-01-16  Filip Pizlo  <fpizlo@apple.com>
591
592         Add framerate baselines to the splay-detail test
593         https://bugs.webkit.org/show_bug.cgi?id=167659
594
595         Reviewed by Geoffrey Garen.
596
597         * JetStream/Octane2/splay-detail.html:
598
599 2017-01-31  Carlos Alberto Lopez Perez  <clopez@igalia.com>
600
601         [EFL][GTK] Skip some IndexedDB tests that time out on the Perf bots.
602
603         Unreviewed performance test gardening.
604
605         * Skipped:
606
607 2017-01-30  Zalan Bujtas  <zalan@apple.com>
608
609         Simple line layout: Small tweaks to improve performance.
610         https://bugs.webkit.org/show_bug.cgi?id=167611
611         <rdar://problem/30274294>
612
613         Reviewed by Simon Fraser.
614
615         * Layout/simple-line-layout-non-repeating-text.html: Added.
616
617 2017-01-13  Said Abou-Hallawa  <sabouhallawa@apple.com>
618
619         26 MotionMark performance tests failing
620         https://bugs.webkit.org/show_bug.cgi?id=166854
621
622         Reviewed by Ryosuke Niwa.
623         
624         Skip running MotionMark on WebKit perf bots.
625
626         * Skipped:
627
628 2017-01-06  Said Abou-Hallawa  <sabouhallawa@apple.com>
629
630         Rename the directory and the files of MotionMark from Animometer to MotionMark
631         https://bugs.webkit.org/show_bug.cgi?id=166659
632
633         Reviewed by Jon Lee.
634         
635         Rename the directory and the files of the benchmark to its new name.
636
637         * MotionMark/developer.html: Renamed from PerformanceTests/Animometer/developer.html.
638         * MotionMark/index.html: Renamed from PerformanceTests/Animometer/index.html.
639         * MotionMark/resources/debug-runner/d3.min.js: Renamed from PerformanceTests/Animometer/resources/debug-runner/d3.min.js.
640         * MotionMark/resources/debug-runner/graph.js: Renamed from PerformanceTests/Animometer/resources/debug-runner/graph.js.
641         * MotionMark/resources/debug-runner/motionmark.css: Renamed from PerformanceTests/Animometer/resources/debug-runner/animometer.css.
642         * MotionMark/resources/debug-runner/motionmark.js: Renamed from PerformanceTests/Animometer/resources/debug-runner/animometer.js.
643         * MotionMark/resources/debug-runner/tests.js: Renamed from PerformanceTests/Animometer/resources/debug-runner/tests.js.
644         * MotionMark/resources/extensions.js: Renamed from PerformanceTests/Animometer/resources/extensions.js.
645         * MotionMark/resources/runner/benchmark-runner.js: Renamed from PerformanceTests/Animometer/resources/runner/benchmark-runner.js.
646         * MotionMark/resources/runner/crystal.svg: Renamed from PerformanceTests/Animometer/resources/runner/crystal.svg.
647         * MotionMark/resources/runner/lines.svg: Renamed from PerformanceTests/Animometer/resources/runner/lines.svg.
648         * MotionMark/resources/runner/logo.svg: Renamed from PerformanceTests/Animometer/resources/runner/logo.svg.
649         * MotionMark/resources/runner/motionmark.css: Renamed from PerformanceTests/Animometer/resources/runner/animometer.css.
650         * MotionMark/resources/runner/motionmark.js: Renamed from PerformanceTests/Animometer/resources/runner/animometer.js.
651         * MotionMark/resources/runner/tests.js: Renamed from PerformanceTests/Animometer/resources/runner/tests.js.
652         * MotionMark/resources/statistics.js: Renamed from PerformanceTests/Animometer/resources/statistics.js.
653         * MotionMark/resources/strings.js: Renamed from PerformanceTests/Animometer/resources/strings.js.
654         * MotionMark/tests/3d/resources/webgl.js: Renamed from PerformanceTests/Animometer/tests/3d/resources/webgl.js.
655         * MotionMark/tests/3d/webgl.html: Renamed from PerformanceTests/Animometer/tests/3d/webgl.html.
656         * MotionMark/tests/bouncing-particles/bouncing-canvas-images.html: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/bouncing-canvas-images.html.
657         * MotionMark/tests/bouncing-particles/bouncing-canvas-shapes.html: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/bouncing-canvas-shapes.html.
658         * MotionMark/tests/bouncing-particles/bouncing-css-images.html: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/bouncing-css-images.html.
659         * MotionMark/tests/bouncing-particles/bouncing-css-shapes.html: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/bouncing-css-shapes.html.
660         * MotionMark/tests/bouncing-particles/bouncing-svg-images.html: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/bouncing-svg-images.html.
661         * MotionMark/tests/bouncing-particles/bouncing-svg-shapes.html: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/bouncing-svg-shapes.html.
662         * MotionMark/tests/bouncing-particles/bouncing-tagged-images.html: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/bouncing-tagged-images.html.
663         * MotionMark/tests/bouncing-particles/resources/bouncing-canvas-images.js: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/bouncing-canvas-images.js.
664         * MotionMark/tests/bouncing-particles/resources/bouncing-canvas-particles.js: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/bouncing-canvas-particles.js.
665         * MotionMark/tests/bouncing-particles/resources/bouncing-canvas-shapes.js: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/bouncing-canvas-shapes.js.
666         * MotionMark/tests/bouncing-particles/resources/bouncing-css-images.js: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/bouncing-css-images.js.
667         * MotionMark/tests/bouncing-particles/resources/bouncing-css-shapes.js: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/bouncing-css-shapes.js.
668         * MotionMark/tests/bouncing-particles/resources/bouncing-svg-images.js: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/bouncing-svg-images.js.
669         * MotionMark/tests/bouncing-particles/resources/bouncing-svg-particles.js: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/bouncing-svg-particles.js.
670         * MotionMark/tests/bouncing-particles/resources/bouncing-svg-shapes.js: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/bouncing-svg-shapes.js.
671         * MotionMark/tests/bouncing-particles/resources/bouncing-tagged-images.js: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/bouncing-tagged-images.js.
672         * MotionMark/tests/bouncing-particles/resources/image1.jpg: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/image1.jpg.
673         * MotionMark/tests/bouncing-particles/resources/image2.jpg: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/image2.jpg.
674         * MotionMark/tests/bouncing-particles/resources/image3.jpg: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/image3.jpg.
675         * MotionMark/tests/bouncing-particles/resources/image4.jpg: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/image4.jpg.
676         * MotionMark/tests/bouncing-particles/resources/image5.jpg: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/image5.jpg.
677         * MotionMark/tests/dom/compositing-transforms.html: Renamed from PerformanceTests/Animometer/tests/dom/compositing-transforms.html.
678         * MotionMark/tests/dom/focus.html: Renamed from PerformanceTests/Animometer/tests/dom/focus.html.
679         * MotionMark/tests/dom/leaves.html: Renamed from PerformanceTests/Animometer/tests/dom/leaves.html.
680         * MotionMark/tests/dom/particles.html: Renamed from PerformanceTests/Animometer/tests/dom/particles.html.
681         * MotionMark/tests/dom/resources/compositing-transforms.js: Renamed from PerformanceTests/Animometer/tests/dom/resources/compositing-transforms.js.
682         * MotionMark/tests/dom/resources/dom-particles.js: Renamed from PerformanceTests/Animometer/tests/dom/resources/dom-particles.js.
683         * MotionMark/tests/dom/resources/focus.js: Renamed from PerformanceTests/Animometer/tests/dom/resources/focus.js.
684         * MotionMark/tests/dom/resources/leaves.js: Renamed from PerformanceTests/Animometer/tests/dom/resources/leaves.js.
685         * MotionMark/tests/master/canvas-stage.html: Renamed from PerformanceTests/Animometer/tests/master/canvas-stage.html.
686         * MotionMark/tests/master/focus.html: Renamed from PerformanceTests/Animometer/tests/master/focus.html.
687         * MotionMark/tests/master/image-data.html: Renamed from PerformanceTests/Animometer/tests/master/image-data.html.
688         * MotionMark/tests/master/leaves.html: Renamed from PerformanceTests/Animometer/tests/master/leaves.html.
689         * MotionMark/tests/master/multiply.html: Renamed from PerformanceTests/Animometer/tests/master/multiply.html.
690         * MotionMark/tests/master/resources/canvas-stage.js: Renamed from PerformanceTests/Animometer/tests/master/resources/canvas-stage.js.
691         * MotionMark/tests/master/resources/canvas-tests.js: Renamed from PerformanceTests/Animometer/tests/master/resources/canvas-tests.js.
692         * MotionMark/tests/master/resources/compass.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/compass.svg.
693         * MotionMark/tests/master/resources/compass100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/compass100.png.
694         * MotionMark/tests/master/resources/console.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/console.svg.
695         * MotionMark/tests/master/resources/console100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/console100.png.
696         * MotionMark/tests/master/resources/contribute.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/contribute.svg.
697         * MotionMark/tests/master/resources/contribute100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/contribute100.png.
698         * MotionMark/tests/master/resources/debugger.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/debugger.svg.
699         * MotionMark/tests/master/resources/debugger100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/debugger100.png.
700         * MotionMark/tests/master/resources/focus.js: Renamed from PerformanceTests/Animometer/tests/master/resources/focus.js.
701         * MotionMark/tests/master/resources/image-data.js: Renamed from PerformanceTests/Animometer/tests/master/resources/image-data.js.
702         * MotionMark/tests/master/resources/inspector.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/inspector.svg.
703         * MotionMark/tests/master/resources/inspector100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/inspector100.png.
704         * MotionMark/tests/master/resources/layout.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/layout.svg.
705         * MotionMark/tests/master/resources/layout100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/layout100.png.
706         * MotionMark/tests/master/resources/leaves.js: Renamed from PerformanceTests/Animometer/tests/master/resources/leaves.js.
707         * MotionMark/tests/master/resources/multiply.js: Renamed from PerformanceTests/Animometer/tests/master/resources/multiply.js.
708         * MotionMark/tests/master/resources/particles.js: Renamed from PerformanceTests/Animometer/tests/master/resources/particles.js.
709         * MotionMark/tests/master/resources/performance.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/performance.svg.
710         * MotionMark/tests/master/resources/performance100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/performance100.png.
711         * MotionMark/tests/master/resources/script.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/script.svg.
712         * MotionMark/tests/master/resources/script100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/script100.png.
713         * MotionMark/tests/master/resources/shortcuts.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/shortcuts.svg.
714         * MotionMark/tests/master/resources/shortcuts100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/shortcuts100.png.
715         * MotionMark/tests/master/resources/standards.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/standards.svg.
716         * MotionMark/tests/master/resources/standards100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/standards100.png.
717         * MotionMark/tests/master/resources/storage.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/storage.svg.
718         * MotionMark/tests/master/resources/storage100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/storage100.png.
719         * MotionMark/tests/master/resources/styles.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/styles.svg.
720         * MotionMark/tests/master/resources/styles100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/styles100.png.
721         * MotionMark/tests/master/resources/svg-particles.js: Renamed from PerformanceTests/Animometer/tests/master/resources/svg-particles.js.
722         * MotionMark/tests/master/resources/text.js: Renamed from PerformanceTests/Animometer/tests/master/resources/text.js.
723         * MotionMark/tests/master/resources/timeline.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/timeline.svg.
724         * MotionMark/tests/master/resources/timeline100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/timeline100.png.
725         * MotionMark/tests/master/svg-particles.html: Renamed from PerformanceTests/Animometer/tests/master/svg-particles.html.
726         * MotionMark/tests/master/text.html: Renamed from PerformanceTests/Animometer/tests/master/text.html.
727         * MotionMark/tests/resources/main.js: Renamed from PerformanceTests/Animometer/tests/resources/main.js.
728         * MotionMark/tests/resources/math.js: Renamed from PerformanceTests/Animometer/tests/resources/math.js.
729         * MotionMark/tests/resources/stage.css: Renamed from PerformanceTests/Animometer/tests/resources/stage.css.
730         * MotionMark/tests/resources/star.svg: Renamed from PerformanceTests/Animometer/tests/resources/star.svg.
731         * MotionMark/tests/resources/yin-yang.png: Renamed from PerformanceTests/Animometer/tests/resources/yin-yang.png.
732         * MotionMark/tests/resources/yin-yang.svg: Renamed from PerformanceTests/Animometer/tests/resources/yin-yang.svg.
733         * MotionMark/tests/simple/resources/simple-canvas-paths.js: Renamed from PerformanceTests/Animometer/tests/simple/resources/simple-canvas-paths.js.
734         * MotionMark/tests/simple/resources/simple-canvas.js: Renamed from PerformanceTests/Animometer/tests/simple/resources/simple-canvas.js.
735         * MotionMark/tests/simple/resources/tiled-canvas-image.js: Renamed from PerformanceTests/Animometer/tests/simple/resources/tiled-canvas-image.js.
736         * MotionMark/tests/simple/simple-canvas-paths.html: Renamed from PerformanceTests/Animometer/tests/simple/simple-canvas-paths.html.
737         * MotionMark/tests/simple/tiled-canvas-image.html: Renamed from PerformanceTests/Animometer/tests/simple/tiled-canvas-image.html.
738         * MotionMark/tests/template/resources/template-canvas.js: Renamed from PerformanceTests/Animometer/tests/template/resources/template-canvas.js.
739         * MotionMark/tests/template/resources/template-css.js: Renamed from PerformanceTests/Animometer/tests/template/resources/template-css.js.
740         * MotionMark/tests/template/resources/template-svg.js: Renamed from PerformanceTests/Animometer/tests/template/resources/template-svg.js.
741         * MotionMark/tests/template/template-canvas.html: Renamed from PerformanceTests/Animometer/tests/template/template-canvas.html.
742         * MotionMark/tests/template/template-css.html: Renamed from PerformanceTests/Animometer/tests/template/template-css.html.
743         * MotionMark/tests/template/template-svg.html: Renamed from PerformanceTests/Animometer/tests/template/template-svg.html.
744
745 2016-12-13  Brady Eidson  <beidson@apple.com>
746
747         Add a PerfTest targeting IDBObjectStore.get().
748         https://bugs.webkit.org/show_bug.cgi?id=165816
749
750         Reviewed by Alex Christensen.
751
752         * IndexedDB/objectstore-get.html: Added.
753
754 2016-12-13  Brady Eidson  <beidson@apple.com>
755
756         Add a PerfTest targeting IDBIndex.get().
757         https://bugs.webkit.org/show_bug.cgi?id=165803
758
759         Reviewed by Alex Christensen.
760
761         * IndexedDB/index-get.html: Added.
762
763 2016-12-12  Brady Eidson  <beidson@apple.com>
764
765         More IndexedDB perf tests.
766         https://bugs.webkit.org/show_bug.cgi?id=165634
767
768         Reviewed by Sam Weinig.
769
770         * IndexedDB/index-multientry.html:
771         * IndexedDB/large-number-of-inserts-responsiveness.html: Added.
772         * IndexedDB/large-number-of-inserts.html: Added.
773         * IndexedDB/objectstore-cursor.html: Added. Runtime cut in 1/4th compared to r209672.
774
775         * resources/runner.js:
776
777 2016-12-12  Commit Queue  <commit-queue@webkit.org>
778
779         Unreviewed, rolling out r209672.
780         https://bugs.webkit.org/show_bug.cgi?id=165766
781
782         IndexedDB/objectstore-cursor.html perf test always times out
783         (Requested by ap on #webkit).
784
785         Reverted changeset:
786
787         "More IndexedDB perf tests."
788         https://bugs.webkit.org/show_bug.cgi?id=165634
789         http://trac.webkit.org/changeset/209672
790
791 2016-12-10  Brady Eidson  <beidson@apple.com>
792
793         More IndexedDB perf tests.
794         https://bugs.webkit.org/show_bug.cgi?id=165634
795
796         Reviewed by Sam Weinig.
797
798         * IndexedDB/index-multientry.html:
799         * IndexedDB/large-number-of-inserts-responsiveness.html: Added.
800         * IndexedDB/large-number-of-inserts.html: Added.
801         * IndexedDB/objectstore-cursor.html: Added.
802         
803         * resources/runner.js: Add "track responsiveness" functionality to PerfTestRunner.
804
805 2016-12-06  Filip Pizlo  <fpizlo@apple.com>
806
807         Concurrent GC should be stable enough to land enabled
808         https://bugs.webkit.org/show_bug.cgi?id=164990
809
810         Reviewed by Geoffrey Garen.
811         
812         Made CDjs more configurable and refined the "large.js" configuration. I was using that one and
813         the new "long.js" configuration to tune concurrent eden GCs.
814         
815         Added a new way of running Splay in browser, which using chartjs to plot the execution times of
816         2000 iterations. This includes the minified chartjs.
817
818         * JetStream/Octane2/splay-detail.html: Added.
819         * JetStream/cdjs/benchmark.js:
820         (benchmarkImpl):
821         (benchmark):
822         * JetStream/cdjs/long.js: Added.
823
824 2016-12-07  Brady Eidson  <beidson@apple.com>
825
826         Add IDB perf tests stressing key size.
827         https://bugs.webkit.org/show_bug.cgi?id=165567
828
829         Reviewed by Alex Christensen.
830
831         * IndexedDB/large-array-keys.html: Added.
832         * IndexedDB/large-binary-keys.html: Added.
833         * IndexedDB/large-string-keys.html: Added.
834
835 2016-12-05  Brady Eidson  <beidson@apple.com>
836
837         Add an IndexedDB perf test to PerformanceTests.
838         https://bugs.webkit.org/show_bug.cgi?id=165430
839
840         Reviewed by Alex Christensen.
841
842         * IndexedDB/index-multientry.html: Added.
843
844 2016-12-02  Filip Pizlo  <fpizlo@apple.com>
845
846         ES6SampleBench should report an average for Steady State so that all of the numbers are comparable
847         https://bugs.webkit.org/show_bug.cgi?id=165325
848
849         Reviewed by Saam Barati.
850
851         This makes all of the numbers that ES6SampleBench reports comparable to each other: they all speak of the time
852         it took to run an iteration of something.
853
854         * ES6SampleBench/results.js:
855         (Results.prototype.reportResult):
856
857 2016-11-18  Jer Noble  <jer.noble@apple.com>
858
859         Add some performance tests for various aspects of media loading and playback.
860         https://bugs.webkit.org/show_bug.cgi?id=164977
861
862         Reviewed by Darin Adler.
863
864         Add new tests for a few aspects of media playback; namely: how quickly  media elements fire
865         the "canplaythrough" event when all data is immediately available, how quickly HLS playback
866         switches resolutions, and at what playback rate MSE-backed video can play without dropping
867         frames. Skip these tests by default, as they're not runnable on all ports, and they require
868         a webserver (run-webkit-httpd) to complete on macOS.
869
870         * Media/HLSCanPlayThrough.html: Added.
871         * Media/HLSGearChange.html: Added.
872         * Media/MSECanPlayThrough.html: Added.
873         * Media/MSEPlaybackRate.html: Added.
874         * Media/StandardCanPlayThrough.html: Added.
875         * Media/hls/1080p/iframe_index.m3u8: Added.
876         * Media/hls/1080p/prog_index.m3u8: Added.
877         * Media/hls/1080p/test.ts: Added.
878         * Media/hls/480p/iframe_index.m3u8: Added.
879         * Media/hls/480p/prog_index.m3u8: Added.
880         * Media/hls/480p/test.ts: Added.
881         * Media/hls/720p/iframe_index.m3u8: Added.
882         * Media/hls/720p/prog_index.m3u8: Added.
883         * Media/hls/720p/test.ts: Added.
884         * Media/hls/720p/test.ts.back: Added.
885         * Media/hls/index.m3u8: Added.
886         * Media/media-source-loader.js: Added.
887         (MediaSourceLoader):
888         (MediaSourceLoader.prototype.loadManifest):
889         (MediaSourceLoader.prototype.loadManifestSucceeded):
890         (MediaSourceLoader.prototype.loadManifestFailed):
891         (MediaSourceLoader.prototype.loadMediaData):
892         (MediaSourceLoader.prototype.loadMediaDataSucceeded):
893         (MediaSourceLoader.prototype.loadMediaDataFailed):
894         (MediaSourceLoader.prototype.get type):
895         (MediaSourceLoader.prototype.get duration):
896         (MediaSourceLoader.prototype.get initSegment):
897         (MediaSourceLoader.prototype.get mediaSegmentsLength):
898         (MediaSourceLoader.prototype.mediaSegments):
899         (MediaSourceLoader.prototype.get everyMediaSegment):
900         * Media/test-fragmented-video.json: Added.
901         * Media/test-fragmented-video.mp4: Added.
902         * Media/test.mp4: Added.
903         * Skipped:
904
905 2016-11-18  Filip Pizlo  <fpizlo@apple.com>
906
907         Concurrent GC should be able to run splay in debug mode and earley/raytrace in release mode with no perf regression
908         https://bugs.webkit.org/show_bug.cgi?id=164282
909
910         Reviewed by Geoffrey Garen and Oliver Hunt.
911         
912         CDjs is a fun benchmark for stressing concurrent GCs, but to really give the GC a good
913         workout you need to increase the amount of work that the test does. This adds a second
914         configuration of the benchmark that has more aircraft. It uses much more memory and causes us
915         to do more GCs and those GCs take longer.
916
917         * JetStream/cdjs/benchmark.js:
918         (benchmarkImpl):
919         (benchmark):
920         * JetStream/cdjs/large.js: Added.
921
922 2016-11-14  Filip Pizlo  <fpizlo@apple.com>
923
924         Unreviewed, revert unintended change.
925
926         * ES6SampleBench/Air/benchmark.js:
927
928 2016-10-03  Saam Barati  <sbarati@apple.com>
929
930         MapHash should speculate on the type of its child node
931         https://bugs.webkit.org/show_bug.cgi?id=161922
932
933         Reviewed by Filip Pizlo.
934
935         I gave the main generator a name so it's easier to see what
936         it is when using the sampling profiler.
937
938         * ES6SampleBench/Basic/ast.js:
939         (Basic.Program):
940
941 2016-09-19  Sergio Villar Senin  <svillar@igalia.com>
942
943         [css-grid] Remove the x2 computation of row sizes with indefinite heights
944         https://bugs.webkit.org/show_bug.cgi?id=162150
945
946         Reviewed by Darin Adler.
947
948         Added a new test case which checks the layout performance of grids inside other grids, i.e,
949         grids acting both as grid container and grid item.
950
951         * Layout/nested-grid.html: Added.
952
953 2016-09-09  Simon Fraser  <simon.fraser@apple.com>
954
955         Perf test Animation/css-accelerated-animation.html failing
956         https://bugs.webkit.org/show_bug.cgi?id=161795
957
958         Skip Animation/css-accelerated-animation.html since this only works as a MobileSafari
959         perf test.
960
961         * Skipped:
962
963 2016-09-08  Simon Fraser  <simon.fraser@apple.com>
964
965         Add a content-animation test with accelerated CSS animations
966         https://bugs.webkit.org/show_bug.cgi?id=161776
967
968         Reviewed by Dean Jackson.
969
970         Add a test that measures the frame rate of accelerated CSS animations.
971
972         This is like css-animation.html (which animates 'left' and 'top') but animates
973         transforms instead. In order to get animation in X and Y we make two nested elements
974         and set transformX() on one, and transformY() on the other.
975
976         * Animation/css-accelerated-animation.html: Added.
977
978 2016-08-23  Saam Barati  <sbarati@apple.com>
979
980         It should be easy to run ES6SampleBench from the jsc shell
981         https://bugs.webkit.org/show_bug.cgi?id=161085
982
983         Reviewed by Yusuke Suzuki.
984
985         This patch makes ES6 sample bench runnable through the `jsc` shell.
986         To do that, you need to be in the PerformanceTests/ES6SampleBench
987         directory and run `jsc cli.js`. To make this work, the benchmark
988         is now aware of being run in two modes: via the browser, and via
989         the shell. Each entry point will set a variable `isInBrowser`,
990         and the benchmark will do different things based on if that
991         variable is true or false.
992
993         * ES6SampleBench/Air/benchmark.js:
994         * ES6SampleBench/Air/stress-test.js:
995         * ES6SampleBench/Basic/benchmark.js:
996         (runBenchmark):
997         * ES6SampleBench/Basic/stress-test.js:
998         * ES6SampleBench/air_benchmark.js:
999         * ES6SampleBench/basic_benchmark.js:
1000         * ES6SampleBench/cli.js: Added.
1001         (return.doRun):
1002         (makeBenchmarkRunner):
1003         * ES6SampleBench/driver.js:
1004         (Driver):
1005         (Driver.prototype._recomputeSummary):
1006         (Driver.prototype._iterate.window.setTimeout):
1007         (Driver.prototype._iterate):
1008         (Driver.prototype._updateIterations):
1009         * ES6SampleBench/glue.js:
1010         (reportResult):
1011         * ES6SampleBench/index.html:
1012         * ES6SampleBench/results.js:
1013         (Results):
1014         (Results.prototype.reportRunning):
1015         (Results.prototype.reportDone):
1016         (Results.prototype.reportError):
1017         * ES6SampleBench/stats.js:
1018         (Stats):
1019         (Stats.prototype.toString):
1020         (Stats.prototype._update):
1021
1022 2016-08-01  Said Abou-Hallawa  <sabouhallawa@apple.com>
1023
1024         Add an option to run the MotionMark tests with classic tile size (512x512)
1025         https://bugs.webkit.org/show_bug.cgi?id=160371
1026
1027         Reviewed by Darin Adler.
1028
1029         Setting the body to large size (3000x3000) and overflow to scroll forces
1030         the classic tile size (512x512). This option does not affect the canvas
1031         tests because the canvas is displayed on one tile regardless of its size.
1032
1033         * Animometer/developer.html:
1034         * Animometer/resources/debug-runner/animometer.css:
1035         (body.showing-test-container.tiles-big):
1036         (body.showing-test-container.tiles-classic):
1037         * Animometer/resources/debug-runner/animometer.js:
1038         * Animometer/resources/runner/animometer.js:
1039         (window.benchmarkController.startBenchmark):
1040
1041 2016-06-29  Filip Pizlo  <fpizlo@apple.com>
1042
1043         Generators violate bytecode liveness validation
1044         https://bugs.webkit.org/show_bug.cgi?id=159279
1045
1046         Reviewed by Yusuke Suzuki.
1047
1048         Add Basic to our test harness.
1049
1050         Also made some cosmetic changes to the benchmark harness.
1051
1052         * ES6SampleBench/Basic/basic-tests.yaml: Added.
1053         * ES6SampleBench/Basic/stress-test.js: Added.
1054         (preciseTime):
1055         * ES6SampleBench/driver.js:
1056         (Driver):
1057         (Driver.prototype.start):
1058         (Driver.prototype.reportError):
1059         * ES6SampleBench/glue.js:
1060         * ES6SampleBench/index.html:
1061
1062 2016-06-28  Filip Pizlo  <fpizlo@apple.com>
1063
1064         ES6SampleBench should have a harness
1065         https://bugs.webkit.org/show_bug.cgi?id=159246
1066
1067         Reviewed by Saam Barati.
1068         
1069         This adds a simple web harness for ES6SampleBench. It runs Air and Basic 10 times for 200
1070         iterations each and reports three metrics:
1071         
1072         First iteration: the time it takes for the first iteration to run. This is the first
1073         iteration after the benchmark is loaded into the iframe, so it's representative of what
1074         would happen if one of these workloads only ran for a short time.
1075         
1076         Worst 2%: of the last 199 iterations, the average of the worst 2% iterations. If code like
1077         any of these workloads was used in an important event handler, you'd want that code to run
1078         well in the worst case in addition to having great throughput.
1079         
1080         Steady state: the total of the last 199 iterations. This is representative of what would
1081         happen if you ran code like this for a long time.
1082         
1083         The total score is the geomean of the firstIteration/worstCase/steadyState numbers of the
1084         two benchmarks.
1085         
1086         The harness does statistics using Student's T-distribution confidence intervals.
1087
1088         * ES6SampleBench/Basic/benchmark.js:
1089         (Benchmark):
1090         * ES6SampleBench/air_benchmark.js: Added.
1091         * ES6SampleBench/basic_benchmark.js: Added.
1092         * ES6SampleBench/driver.js: Added.
1093         (Driver):
1094         (Driver.prototype.addBenchmark):
1095         (Driver.prototype.start):
1096         (Driver.prototype.reportResult):
1097         (Driver.prototype.reportError):
1098         (Driver.prototype._recomputeSummary.Geomean):
1099         (Driver.prototype._recomputeSummary.Geomean.prototype.add):
1100         (Driver.prototype._recomputeSummary.Geomean.prototype.get result):
1101         (Driver.prototype._recomputeSummary):
1102         (Driver.prototype._iterate):
1103         (Driver.prototype._updateIterations):
1104         * ES6SampleBench/glue.js: Added.
1105         * ES6SampleBench/index.html: Added.
1106         * ES6SampleBench/results.js: Added.
1107         (Results):
1108         (Results.prototype.get benchmark):
1109         (Results.prototype.reset):
1110         (Results.prototype.reportRunning):
1111         (Results.prototype.reportDone):
1112         (Results.prototype.reportResult.averageAbovePercentile):
1113         (Results.prototype.reportResult):
1114         (Results.prototype.reportError):
1115         * ES6SampleBench/stats.js: Added.
1116         (Stats):
1117         (Stats.prototype.reset):
1118         (Stats.prototype.add):
1119         (Stats.prototype.get numIterations):
1120         (Stats.prototype.valueForIteration):
1121         (Stats.get result.tDist):
1122         (Stats.prototype.get result):
1123         (Stats.prototype.toString):
1124         (Stats.prototype._update):
1125         * ES6SampleBench/style.css: Added.
1126         (body):
1127         (body, th, tr):
1128         (h1, h2, h3, h4):
1129         (h1):
1130         (h2):
1131         (h3):
1132         (hr):
1133         (address):
1134         (img):
1135         (.underline):
1136         (ol.loweralpha):
1137         (ol.upperalpha):
1138         (ol.lowerroman):
1139         (ol.upperroman):
1140         (ol.arabic):
1141         (.banner-link:link, .banner-link:visited):
1142         (:link, :visited):
1143         (h4 :link, h4 :visited, h5 :link, h5 :visited, h6 :link, h6 :visited):
1144         (.anchor:link, .anchor:visited):
1145         (* > .anchor:link, * > .anchor:visited):
1146         (span:hover .anchor):
1147         (a.forbidden, span.forbidden):
1148         (a.missing:hover):
1149         (a.closed:link, a.closed:visited, span.closed):
1150         (pre):
1151         (div.code):
1152         (div.code pre):
1153         (dt):
1154         (dd):
1155         (dd:last-child):
1156         (.site-logo):
1157         (.site-logo .tagline):
1158         (table):
1159         (#contents):
1160         (p):
1161         (p:last-child):
1162         (th):
1163         (td):
1164
1165 2016-06-28  Jon Lee  <jonlee@apple.com>
1166
1167         Update focus test
1168         https://bugs.webkit.org/show_bug.cgi?id=159242
1169         rdar://problem/27070007
1170
1171         Reviewed by Dean Jackson.
1172         Provisionally reviewed by Said Abou-Hallawa.
1173
1174         Move previous test to dom suite, and update the test for better reporting of frame rate, although
1175         it uses a different rendering path.
1176
1177         * Animometer/resources/debug-runner/tests.js: Add to dom suite.
1178         * Animometer/tests/dom/focus.html: Copied from PerformanceTests/Animometer/tests/master/focus.html.
1179         * Animometer/tests/dom/resources/focus.js: Copied from PerformanceTests/Animometer/tests/master/resources/focus.js.
1180         * Animometer/tests/master/focus.html: Remove center element.
1181         * Animometer/tests/master/resources/focus.js: Use narrower size range with smaller particles. Remove the
1182         container elements. Inline getBlurValue and getOpacityValue since they are only called once.
1183
1184 2016-06-28  Filip Pizlo  <fpizlo@apple.com>
1185
1186         Move Air.js and Basic into ES6SampleBench
1187
1188         Rubber stamped by Geoffrey Garen.
1189         
1190         We want to group our ES6 benchmarks together, and eventually, we want to give them a
1191         common harness.
1192
1193         * Air.js: Removed.
1194         * Air.js/README.md: Removed.
1195         * Air.js/airjs-tests.yaml: Removed.
1196         * Air.js/all.js: Removed.
1197         * Air.js/allocate_stack.js: Removed.
1198         * Air.js/arg.js: Removed.
1199         * Air.js/basic_block.js: Removed.
1200         * Air.js/benchmark.js: Removed.
1201         * Air.js/code.js: Removed.
1202         * Air.js/custom.js: Removed.
1203         * Air.js/frequented_block.js: Removed.
1204         * Air.js/insertion_set.js: Removed.
1205         * Air.js/inst.js: Removed.
1206         * Air.js/liveness.js: Removed.
1207         * Air.js/make_dist.sh: Removed.
1208         * Air.js/opcode.js: Removed.
1209         * Air.js/payload-airjs-ACLj8C.js: Removed.
1210         * Air.js/payload-gbemu-executeIteration.js: Removed.
1211         * Air.js/payload-imaging-gaussian-blur-gaussianBlur.js: Removed.
1212         * Air.js/payload-typescript-scanIdentifier.js: Removed.
1213         * Air.js/reg.js: Removed.
1214         * Air.js/stack_slot.js: Removed.
1215         * Air.js/stress-test.js: Removed.
1216         * Air.js/strip-hash.rb: Removed.
1217         * Air.js/symbols.js: Removed.
1218         * Air.js/test.html: Removed.
1219         * Air.js/test.js: Removed.
1220         * Air.js/tmp.js: Removed.
1221         * Air.js/tmp_base.js: Removed.
1222         * Air.js/util.js: Removed.
1223         * Basic: Removed.
1224         * Basic/ast.js: Removed.
1225         * Basic/basic.js: Removed.
1226         * Basic/benchmark.js: Removed.
1227         * Basic/caseless_map.js: Removed.
1228         * Basic/lexer.js: Removed.
1229         * Basic/number.js: Removed.
1230         * Basic/parser.js: Removed.
1231         * Basic/random.js: Removed.
1232         * Basic/state.js: Removed.
1233         * Basic/test.html: Removed.
1234         * Basic/test.js: Removed.
1235         * Basic/util.js: Removed.
1236         * ES6SampleBench: Added.
1237         * ES6SampleBench/Air: Copied from PerformanceTests/Air.js.
1238         * ES6SampleBench/Basic: Copied from PerformanceTests/Basic.
1239         * Skipped:
1240
1241 2016-06-24  Filip Pizlo  <fpizlo@apple.com>
1242
1243         Add a ES6 generator benchmark
1244         https://bugs.webkit.org/show_bug.cgi?id=159101
1245
1246         Rubber stamped by Keith Miller.
1247         
1248         This adds a Basic interpreter loosely based on ECMA-55:
1249         
1250         http://www.ecma-international.org/publications/files/ECMA-ST-WITHDRAWN/ECMA-55,%201st%20Edition,%20January%201978.pdf
1251         
1252         It includes a lexer that is a generator, a parser that uses regular expressions, and an AST
1253         walk interpreter where the walking functions for statements are generators that call each
1254         other with yield*. This enables the interpreter to look like an AST walk even though it can
1255         yield at INPUT and PRINT statements.
1256         
1257         This also uses for-of, classes, Map, and WeakMap. It also uses deprecated-but-awesome RegExp
1258         features like RegExp.lastMatch and RegExp.rightContext. I did it that way because this is
1259         how I've always written lexers in dynamic languages; see offlineasm's lex() method in
1260         parser.rb for example.
1261         
1262         The benchmark runs a handful of simple Basic programs. The longest-running one computes
1263         prime numbers.
1264         
1265         Includes a command-line and web harness. On my machine it runs in 2-3 seconds.
1266
1267         * Basic: Added.
1268         * Basic/ast.js: Added.
1269         (Basic.NumberApply):
1270         (Basic.Variable):
1271         (Basic.Const):
1272         (Basic.NumberPow):
1273         (Basic.NumberMul):
1274         (Basic.NumberDiv):
1275         (Basic.NumberNeg):
1276         (Basic.NumberAdd):
1277         (Basic.NumberSub):
1278         (Basic.StringVar):
1279         (Basic.Equals):
1280         (Basic.NotEquals):
1281         (Basic.LessThan):
1282         (Basic.GreaterThan):
1283         (Basic.LessEqual):
1284         (Basic.GreaterEqual):
1285         (Basic.GoTo):
1286         (Basic.GoSub):
1287         (Basic.Def):
1288         (Basic.Let):
1289         (Basic.If):
1290         (Basic.Return):
1291         (Basic.Stop):
1292         (Basic.On):
1293         (sideState.shouldStop):
1294         (Basic.For):
1295         (Basic.Next):
1296         (Basic.Print):
1297         (Basic.Input):
1298         (Basic.Read):
1299         (Basic.Restore):
1300         (Basic.Dim):
1301         (Basic.Randomize):
1302         (Basic.End):
1303         (Basic.Program):
1304         * Basic/basic.js: Added.
1305         (prepare):
1306         (simulate):
1307         * Basic/benchmark.js: Added.
1308         (Benchmark):
1309         (Benchmark.prototype.runIteration.expect):
1310         (Benchmark.prototype.runIteration):
1311         (runBenchmark):
1312         * Basic/caseless_map.js: Added.
1313         (CaselessMap):
1314         * Basic/lexer.js: Added.
1315         (lex.consumeWhitespace):
1316         (lex.consume):
1317         (lex):
1318         * Basic/lexer_test.js: Added.
1319         * Basic/number.js: Added.
1320         (NumberValue):
1321         (NumberValue.prototype.get value):
1322         (NumberValue.prototype.apply):
1323         (NumberValue.prototype.leftApply):
1324         (NumberValue.prototype.assign):
1325         (NumberArray.):
1326         (NumberArray):
1327         (NumberArray.prototype.apply):
1328         (NumberArray.prototype.leftApply):
1329         (NumberFunction):
1330         (NumberFunction.prototype.apply):
1331         (NumberFunction.prototype.leftApply):
1332         (NativeFunction):
1333         (NativeFunction.prototype.apply):
1334         (NativeFunction.prototype.leftApply):
1335         * Basic/parser.js: Added.
1336         (parse):
1337         (parse.pushToken):
1338         (parse.peekToken):
1339         (parse.consumeKind):
1340         (parse.consumeToken):
1341         (parse.parseVariable):
1342         (parse.parseNumericExpression.parsePrimary):
1343         (parse.parseNumericExpression.parseFactor):
1344         (parse.parseNumericExpression.parseTerm):
1345         (parse.parseNumericExpression):
1346         (parse.parseConstant):
1347         (parse.parseStringExpression):
1348         (parse.isStringExpression):
1349         (parse.parseRelationalExpression):
1350         (parse.parseNonNegativeInteger):
1351         (parse.parseGoToStatement):
1352         (parse.parseGoSubStatement):
1353         (parse.parseStatement):
1354         (parse.parseStatements):
1355         * Basic/random.js: Added.
1356         (createRNG):
1357         (createRNGWithFixedSeed):
1358         (createRNGWithRandomSeed):
1359         * Basic/state.js: Added.
1360         (State):
1361         (State.prototype.getValue):
1362         (State.prototype.getSideState):
1363         (State.prototype.abort):
1364         (State.prototype.validate):
1365         * Basic/test.html: Added.
1366         * Basic/test.js: Added.
1367         * Basic/util.js: Added.
1368         (this.performance.performance.now.currentTime):
1369         (else.this.preciseTime.currentTime):
1370         (else.currentTime):
1371         * Skipped: Make sure that we don't run Basic yet.
1372
1373 2016-06-21  Jon Lee  <jonlee@apple.com>
1374
1375         Update canvas size when benchmark begins
1376         https://bugs.webkit.org/show_bug.cgi?id=159010
1377
1378         Reviewed by Dean Jackson.
1379
1380         Previously we would evaluate the media queries and assign the canvas size to the test
1381         on body.onload. Instead, do it once the user starts the benchmark.
1382
1383         * Animometer/resources/debug-runner/animometer.js: Move benchmarkController.determineCanvasSize()
1384         to benchmarkController._startBenchmark.
1385         * Animometer/resources/runner/animometer.js: Ditto.
1386
1387 2016-06-21  Jon Lee  <jonlee@apple.com>
1388
1389         Add new timestamp option
1390         https://bugs.webkit.org/show_bug.cgi?id=159006
1391
1392         Reviewed by Dean Jackson.
1393
1394         Add a new option to take timestamps from the rAF callback.
1395
1396         * Animometer/developer.html: Added option.
1397         * Animometer/tests/resources/main.js:
1398         (Benchmark): Fall back to using the rAF timestamp if performance.now()
1399         is not available.
1400         (Benchmark._animateLoop): Update how timestamp is set. Prefer to use
1401         local var instead of accessing private var.
1402
1403 2016-06-21  Jon Lee  <jonlee@apple.com>
1404
1405         Improvements to Animometer benchmark
1406         https://bugs.webkit.org/show_bug.cgi?id=157738
1407
1408         Reviewed by Dean Jackson.
1409         Provisionally reviewed by Said Abou-Hallawa.
1410
1411         Update tests.
1412
1413         * Animometer/tests/master/text.html: Ensure only three text sizes for
1414         the three canvases.
1415         * Animometer/tests/master/focus.html: Reduce the text size to fit with smaller
1416         particle sizes.
1417         * Animometer/tests/master/resources/focus.js: Remove the quadratic distribution for
1418         particle sizes, and make it linear. Reduce the size variance. Shuffle the math to
1419         reduce some calculations per frame. Fix the placement of the particles which might
1420         otherwise be culled.
1421         * Animometer/tests/master/resources/image-data.js: Reduce the particle size
1422         to encourage larger scores.
1423
1424 2016-06-21  Jon Lee  <jonlee@apple.com>
1425
1426         Improvements to Animometer benchmark
1427         https://bugs.webkit.org/show_bug.cgi?id=157738
1428
1429         Reviewed by Dean Jackson.
1430         Provisionally reviewed by Said Abou-Hallawa.
1431
1432         Include confidence interval for the final score, and store the canvas
1433         size in the serialization so that it is accurately shown in results.
1434
1435         * Animometer/developer.html: Add a "confidence" div.
1436         * Animometer/index.html: Ditto. Convert "mean" to "confidence".
1437         * Animometer/resources/debug-runner/animometer.js: Look at options, and
1438         if the configuration is included, update the body class based on it
1439         (similar to what we do when we first load the page). That way, if you
1440         drag-and-drop previous results in, that configuration is reflected in
1441         the dashboard. Show the full confidence interval.
1442
1443         * Animometer/resources/debug-runner/animometer.css:
1444         * Animometer/resources/debug-runner/animometer.js: Style update.
1445         * Animometer/resources/runner/animometer.css:
1446
1447         * Animometer/resources/runner/animometer.js:
1448         (_processData): Propagate the confidence interval values out and calculate
1449         the lower and upper bounds. For now, shortcut the aggregate calculation,
1450         since we only go through one iteration.
1451         (this._processData.calculateScore): Propagate the confidence interval out
1452         to be next to the score. Depending on the controller these values are
1453         calculated differently.
1454         (this._processData._getResultsProperty): Convenience function.
1455         (this._processData.get score): Refactor.
1456         (this._processData.get scoreLowerBound): Get the aggregate lower bound.
1457         (this._processData.get scoreUpperBound): Get the aggregate upper bound.
1458         (window.sectionsManager.setSectionScore):
1459         (window.benchmarkController._startBenchmark): When the benchmark starts, note
1460         the canvas size and add it to options. That way it will be automatically be
1461         serialized.
1462         (window.benchmarkController.showResults): Include the maximum deviation
1463         percentage.
1464         * Animometer/resources/runner/lines.svg: Make the background line up with the
1465         skew.
1466         * Animometer/resources/runner/tests.js:
1467         (Headers.details.text): Refactor.
1468         * Animometer/resources/statistics.js:
1469         (Statistics.largestDeviationPercentage): Convenience function to calculate
1470         the deviation percentage on either end and return the largest deviation.
1471         * Animometer/resources/strings.js:
1472
1473         Allow specifying a regression profile to use instead of taking the one
1474         with the lowest error.
1475
1476         Address an issue in the focus test when the regression calculated ends
1477         up overestimating the change point, causing a cascade of tougher
1478         ramps. The reason behind this is that at max complexity of an initial
1479         ramp, the frame length is very high, and it influences the second
1480         segment of the piecewise regression strongly, causing it to be very
1481         steep. As a result, the first segment, expected to be flat, ends up
1482         covering a higher range of complexity. That makes the change point
1483         much higher than it should be. To avoid this, we will add a sanity
1484         check on the maximum value of the ramp. If the regression's projected
1485         value at the maximum complexity of the current ramp is very slow (less
1486         than 20 fps), 1) reduce the maximum complexity by 20%, and 2) do not
1487         include the regression's change point in the change point estimator.
1488         That estimator is used as the midpoint of the next ramp, and including
1489         the change point from a poor regression can bake in the error. The
1490         controller already knows how to adjust for ramps that are too easy for
1491         the system.
1492
1493         * Animometer/resources/runner/animometer.js:
1494         (this._processData.findRegression): Takes a preferred profile and gives that to
1495         Regression.
1496         (this._processData.calculateScore): With the ramp controller, take the profile
1497         of the ramp that was used the most when calculating the ramp's regression. That
1498         profile is what is used for the test's score.
1499         * Animometer/resources/statistics.js:
1500         (Regression.Utilities.createClass): Update to take an options object which can
1501         specify a profile to calculate with. Otherwise it will continue to use both and
1502         select the one with the lower error.
1503         (_calculateRegression): Fix an issue where we claim 0 error if the regression
1504         calculation fails due to divide-by-zero. Instead reject that regression calculation
1505         by giving it Number.MAX_VALUE.
1506         * Animometer/resources/strings.js: New strings for marking a regression as flat
1507         or slope.
1508         * Animometer/tests/resources/main.js:
1509         (RampController): Rename the thresholds for clarity. Add a threshold that, if
1510         exceeded, will lower the maximum complexity of the next ramp.
1511         (tune): Relax the cdf check to consider whether the interval definitely falls in
1512         the desired frame length threshold.
1513         (processSamples): Include the profile in the ramp.
1514
1515         Update ramp controller test. Increase the length of the test to 30 seconds, and extend
1516         the interval to 120 ms during sampling. Improve the estimation of the ramp parameters.
1517
1518         * Animometer/developer.html: Change default to 30 seconds, and don't show the progress bar
1519         by default.
1520         * Animometer/resources/runner/animometer.js: Change default to 30 seconds.
1521         * Animometer/tests/resources/main.js: A number of improvements to the ramp controller, in
1522         the order in which they appear in the patch:
1523
1524         - With a longer test length use longer ramps with longer intervals to get more data at each
1525         complexity. Keep the 100 ms interval length during the ramp up phase since we don't need to
1526         spend more time there to find the right order of magnitude, but increase it during the
1527         ramps to 120 ms.
1528         - The ramp linearly interpolates the complexity to render based on its timestamp, but it would
1529         never sample the minimum complexity. Instead of lerping max to min complexity from time
1530         0 to t where t is the ramp length, instead lerp from 0 to (t - intervalSampleLength) so that
1531         we can have at least one interval sample at the min complexity for that ramp.
1532         - Some regression calculations only come out with one line segment rather than the two
1533         we expect. This could be due to a noisy ramp or the ramp's range is too narrow. If that's the
1534         case, influence the minimum complexity of the next ramp towards the lowest bound of 1, so that
1535         we ensure that at least part of the ramp is covering a complexity range that the system can
1536         handle at full 60.
1537         - Remove an assignment to interpolatedFrameLength since that is never subsequently used.
1538
1539         Update the format used to serialize the results for analysis.
1540
1541         Each data point used to have named properties for fields like complexity and frame rate.
1542         In addition the serialized numbers had rounding errors that took up many characters.
1543         Update the format by introducing a new data container called SampleData, which contains a
1544         field map. The map maps a string to an array index. Each data point is an array, so, to
1545         get a stat, use the field map to get the array index into the data point. This allows future
1546         versions to track other data, and reduces the size of the output string by two-thirds.
1547
1548         * Animometer/resources/extensions.js:
1549         (Utilities.toFixedNumber): Add convenience function that truncates the number to a fixed
1550         precision, and converts it back to a number.
1551         (SampleData): New class that contains sample data and a field map that maps properties to
1552         an array index.
1553         (get length): Number of data points.
1554         (addField): Add a field to the field map.
1555         (push): Add a data point.
1556         (sort): Sort the data.
1557         (slice): Return new SampleData object with sliced data.
1558         (forEach): Iterate over the data with provided function.
1559         (createDatum):
1560         (getFieldInDatum): Returns the data point associated with the field name by looking it up
1561         in the field map in the datum provided, which can be the datum object itself (an array) or
1562         an index into the data member variable.
1563         (setFieldInDatum): Sets the data point associated with the field name.
1564         (at): Returns the data point at the provided index.
1565         (toArray): Serializes the data where the field map serves as property names for each point.
1566
1567         * Animometer/resources/debug-runner/graph.js:
1568         (updateGraphData): Remove unused _testData. Convert the data to the old array format for the
1569         graph to use, since the old format was much easier to work with when displaying the graphs.
1570         (onGraphTypeChanged): For some controllers, no alternative score or mean is provided.
1571         * Animometer/resources/runner/animometer.css:
1572         * Animometer/resources/runner/animometer.js: Refactor to use SampleData. Update JSON output
1573         to only go to 3 digits of precision for purposes of reducing the data size.
1574         * Animometer/resources/strings.js: Add new strings to put into the field maps.
1575         * Animometer/tests/resources/main.js: Refactor to use SampleData.
1576
1577         * Animometer/developer.html:
1578         * Animometer/index.html: Restructure results table for both pages. Add charset attribute to
1579         tests.js include.
1580         * Animometer/resources/debug-runner/animometer.css: Clear out styles from release runner.
1581         * Animometer/resources/debug-runner/graph.js:
1582         (onGraphTypeChanged): Update score and mean if bootstrap results are available from the
1583         controller, since not all controllers do bootstrapping.
1584         * Animometer/resources/debug-runner/tests.js: Update header text.
1585         * Animometer/resources/runner/animometer.css: Include confidence interval in results.
1586         * Animometer/resources/runner/animometer.js:
1587         (ResultsTable._addHeader): Header contents can be HTML, so use innerHTML instead.
1588         (ResultsTable._addBody): Add tbody element.
1589         (ResultsTable._addTest): Allow a data cell to invoke a JS function to get its contents.
1590         (window.benchmarkController.showResults): Add table that includes tests' confidence intervals.
1591         * Animometer/resources/runner/tests.js:
1592         (Headers.details.text): Add new details table that includes bootstrap confidence interval.
1593         The interval can be asymmetric, but for simplicity, report the maximum deviation percentage
1594         on either side of the bootstrap median.
1595         * Animometer/resources/statistics.js:
1596         (bootstrap): Include the confidence percentage in the return object.
1597
1598         Report canvas size in results.
1599
1600         * Animometer/developer.html: Add markup to indicate whether a small, medium, or large
1601         canvas was used.
1602         * Animometer/index.html: Ditto.
1603         * Animometer/resources/debug-runner/animometer.js: Call determineCanvasSize().
1604         * Animometer/resources/runner/animometer.css: Update styles to set the canvas based on the
1605         body class size.
1606         * Animometer/resources/runner/animometer.js:
1607         (window.benchmarkController.initialize): Update styles to set the canvas based on the
1608         body class size.
1609         (window.benchmarkController.determineCanvasSize): Run various media queries and set the body
1610         class based on the size of the device.
1611
1612         * Animometer/developer.html: Refactor to include the main CSS file, and redo
1613         the layout so that it doesn't rely on flexbox.
1614         * Animometer/resources/debug-runner/animometer.css:
1615         * Animometer/resources/debug-runner/animometer.js:
1616         (updateDisplay): Since various parts of the script alter the body class, we can't
1617         replace the className directly. Instead, remove all display-based values and then add
1618         the one that was selected.
1619         * Animometer/resources/debug-runner/graph.js:
1620         (updateGraphData): To set the size of the graph, use window.innerHeight.
1621         * Animometer/resources/runner/animometer.js:
1622         (window.sectionsManager.showSection): Since various parts of the script alter the body
1623         class, we can't replace the className directly. Remove all of the section classes
1624         individually and then add the one desired.
1625         * Animometer/tests/resources/stage.css: Remove -apple-system as a font to use in the
1626         stage.
1627
1628 2016-06-12  Filip Pizlo  <fpizlo@apple.com>
1629
1630         Fix round-down goof in Air.js's ShuffleCustom.forEachArg
1631         https://bugs.webkit.org/show_bug.cgi?id=158674
1632
1633         Reviewed by Michael Saboff.
1634         
1635         x / 3 * 3 is not how you round down to multiples of 3 in JavaScript. You need to do
1636         Math.floor(x / 3) * 3 instead.
1637         
1638         This is a benign change, because having extra arguments to a Shuffle is not something we
1639         actually take advantage of yet. But I think it's best to match the original C++ code's
1640         intent.
1641
1642         * Air.js/custom.js:
1643         (const.ShuffleCustom.forEachArg):
1644
1645 2016-06-11  Filip Pizlo  <fpizlo@apple.com>
1646
1647         Unreviewed, skip Air.js. It got unskipped when I renamed it from JSAir.
1648
1649         * Skipped:
1650
1651 2016-06-10  Filip Pizlo  <fpizlo@apple.com>
1652
1653         Make it easy to package Air.js
1654         https://bugs.webkit.org/show_bug.cgi?id=158652
1655
1656         Reviewed by Benjamin Poulain.
1657         
1658         If you want to give Air.js to someone, you can now do:
1659         
1660         cd PerformanceTests/Air.js
1661         ./make_dist.sh
1662         
1663         This will create Air.js.tar.gz and a directory called Air.js. If you want to send someone
1664         the benchmark, just send either of those things and tell them to look at index.html.
1665         
1666         You may have to edit make_dist.sh for wherever you have Markdown.pl. You can get that from
1667         https://daringfireball.net/projects/markdown/
1668
1669         * Air.js/make_dist.sh: Added.
1670         * Air.js/test.html:
1671
1672 2016-06-10  Filip Pizlo  <fpizlo@apple.com>
1673
1674         Unreviewed, make the link to test.html into an actual link.
1675
1676         * Air.js/README.md:
1677
1678 2016-06-10  Filip Pizlo  <fpizlo@apple.com>
1679
1680         Fix some text in Air.js/README.md
1681         https://bugs.webkit.org/show_bug.cgi?id=158650
1682
1683         Reviewed by Benjamin Poulain.
1684         
1685         I read the text again and found bugs:
1686         
1687         - We never actually say how to run the benchmark. This change adds a blurb about how to run
1688           it.
1689
1690         - We both say that allocateStack is responsible for the bulk of the running time and that
1691           we haven't measured where the bulk of the time is spent. This changes the text to say that
1692           it was a goal to make allocateStack be the hottest part of the benchmark, but that we did
1693           not measure this.
1694
1695         * Air.js/README.md:
1696
1697 2016-06-10  Filip Pizlo  <fpizlo@apple.com>
1698
1699         Air.js should have some documentation
1700         https://bugs.webkit.org/show_bug.cgi?id=158648
1701
1702         Reviewed by Keith Miller.
1703         
1704         I want to be able to point people at a document if they want to know more about this
1705         benchmark.
1706
1707         * Air.js/README.md: Added.
1708
1709 2016-06-10  Filip Pizlo  <fpizlo@apple.com>
1710
1711         Rename JSAir to Air.js.
1712
1713         Rubber stamped by Mark Lam.
1714
1715         * Air.js: Copied from PerformanceTests/JSAir.
1716         * Air.js/airjs-tests.yaml: Copied from PerformanceTests/JSAir/jsair-tests.yaml.
1717         * Air.js/benchmark.js: Copied from PerformanceTests/JSAir/benchmark.js.
1718         (Benchmark):
1719         * Air.js/jsair-tests.yaml: Removed.
1720         * Air.js/payload-airjs-ACLj8C.js: Copied from PerformanceTests/JSAir/payload-jsair-ACLj8C.js.
1721         * Air.js/payload-jsair-ACLj8C.js: Removed.
1722         * Air.js/stress-test.js: Copied from PerformanceTests/JSAir/stress-test.js.
1723         * Air.js/test.html: Copied from PerformanceTests/JSAir/test.html.
1724         * Air.js/test.js: Copied from PerformanceTests/JSAir/test.js.
1725         * JSAir: Removed.
1726         * JSAir/all.js: Removed.
1727         * JSAir/allocate_stack.js: Removed.
1728         * JSAir/arg.js: Removed.
1729         * JSAir/basic_block.js: Removed.
1730         * JSAir/benchmark.js: Removed.
1731         * JSAir/code.js: Removed.
1732         * JSAir/custom.js: Removed.
1733         * JSAir/frequented_block.js: Removed.
1734         * JSAir/insertion_set.js: Removed.
1735         * JSAir/inst.js: Removed.
1736         * JSAir/jsair-tests.yaml: Removed.
1737         * JSAir/liveness.js: Removed.
1738         * JSAir/opcode.js: Removed.
1739         * JSAir/payload-gbemu-executeIteration.js: Removed.
1740         * JSAir/payload-imaging-gaussian-blur-gaussianBlur.js: Removed.
1741         * JSAir/payload-jsair-ACLj8C.js: Removed.
1742         * JSAir/payload-typescript-scanIdentifier.js: Removed.
1743         * JSAir/reg.js: Removed.
1744         * JSAir/stack_slot.js: Removed.
1745         * JSAir/stress-test.js: Removed.
1746         * JSAir/strip-hash.rb: Removed.
1747         * JSAir/symbols.js: Removed.
1748         * JSAir/test.html: Removed.
1749         * JSAir/test.js: Removed.
1750         * JSAir/tmp.js: Removed.
1751         * JSAir/tmp_base.js: Removed.
1752         * JSAir/util.js: Removed.
1753
1754 2016-06-10  Filip Pizlo  <fpizlo@apple.com>
1755
1756         JSC Stress Test failing: jsair-tests.yaml/test.js.ftl-eager-no-cjit
1757         https://bugs.webkit.org/show_bug.cgi?id=158571
1758
1759         Reviewed by Keith Miller.
1760         
1761         Introduce a different harness for run-jsc-stress-tests, which runs for a minimum of 10
1762         iterations and then tries to do any number of "bonus" iterations until it's been running for
1763         two seconds. Since this is the sort of test that isn't really meant to stress anything in
1764         particular, I think it's OK if it is time-limited in this way. The worst case is that some
1765         of its failures will be flaky, but I think that they would have been flaky anyway given the
1766         complexity of the test.
1767
1768         * JSAir/benchmark.js:
1769         (Benchmark):
1770         (Benchmark.prototype.runIteration):
1771         (benchmark): Deleted.
1772         * JSAir/jsair-tests.yaml:
1773         * JSAir/stress-test.js: Added.
1774         (preciseTime):
1775         * JSAir/test.html:
1776         * JSAir/test.js:
1777
1778 2016-06-09  Filip Pizlo  <fpizlo@apple.com>
1779
1780         Unreviewed, teach the perf bots not to run JSAir.
1781
1782         * Skipped:
1783
1784 2016-06-08  Filip Pizlo  <fpizlo@apple.com>
1785
1786         Use more ES6 features in JSAir
1787         https://bugs.webkit.org/show_bug.cgi?id=158497
1788
1789         Reviewed by Keith Miller.
1790         
1791         This improves JSAir with the following ES6 features suggested by JoePeck:
1792         
1793         - String interpolation.
1794         - Destructuring inside PatchCustom.
1795         - Default arguments.
1796         
1797         All of these things are on hot paths.
1798         
1799         Note that I didn't use string interpolation everywhere that I could, only in those places
1800         where it made the code more readable. In Ruby, I used the style that if the interpolation
1801         expression has any non-trivial stuff (like a ternary operator, a chain of calls, or embedded
1802         strings) then it's better to use regular strcat. I think that's what I carried over to here.
1803         
1804         Note that the previous change (Add result validation to JSAir) also made the Proxy code not
1805         dead, though it's not necessarily on the hot path. The Proxy isn't called into frequently
1806         but it's used from a function that is otherwise hot, so if calling into the Proxy prevents
1807         that function from being optimized then it will hurt so good.
1808         
1809         I also reenabled tail calls in a few places.
1810         
1811         This change doesn't seem to change the performance of the benchmark for us. That's expected
1812         since these ES6 features are cheap. Note that this claim doesn't include Proxy, which was
1813         added in a separate change and that change did make the benchmark overall more expensive.
1814
1815         * JSAir/allocate_stack.js:
1816         (allocateStack):
1817         * JSAir/arg.js:
1818         (Arg.createBitImm64):
1819         (Arg.createAddr):
1820         (Arg.createStack):
1821         (Arg.logScale):
1822         (Arg.createIndex):
1823         * JSAir/basic_block.js:
1824         (BasicBlock.get headerString):
1825         (BasicBlock.prototype.get if):
1826         (BasicBlock):
1827         * JSAir/benchmark.js:
1828         (benchmark):
1829         * JSAir/code.js:
1830         (Code):
1831         (Code.prototype.addBlock):
1832         (Code.prototype.addStackSlot):
1833         (Code.prototype.newTmp):
1834         (Code.prototype.get size):
1835         (Code.prototype.get blocks):
1836         (Code.prototype.get stackSlots):
1837         (Code.prototype.tmps):
1838         (Code.prototype.get callArgAreaSize):
1839         (Code.prototype.toString):
1840         * JSAir/custom.js:
1841         (const.PatchCustom.forEachArg):
1842         * JSAir/inst.js:
1843         (Inst):
1844         * JSAir/reg.js:
1845         (Reg.prototype.toString):
1846         * JSAir/util.js:
1847         (symbolName):
1848         (lowerSymbolName):
1849
1850 2016-06-07  Filip Pizlo  <fpizlo@apple.com>
1851
1852         Add result validation to JSAir
1853         https://bugs.webkit.org/show_bug.cgi?id=158493
1854
1855         Reviewed by Saam Barati.
1856         
1857         This adds the ability to hash a Code in a way that matches the C++ code's hashing of Code.
1858         This allows us to check if the Code that JSAir sees is the code that C++ saw. We use this to
1859         check the Code before and after allocateStack, and compare against hashes we got from C++.
1860         
1861         Doing this uncovered bugs. roundUpToMultipleOf wasn't doing anything. allocateStack was not
1862         allocating things correctly because I was concatting a Set to an Array, which doesn't really
1863         work. Now these bugs are fixed.
1864         
1865         The checking step adds to the running time so I reduced the number of iterations. The
1866         benchmark spends a decent amount of its time computing Code hashes; I think it's around 1/3
1867         total. This is probably OK. It's better to verify the results even if the running time is
1868         not all in the "core" of the algorithm.
1869         
1870         Also add a run-jsc-stress-tests yaml file to allow this to run as a test.
1871
1872         * JSAir/allocate_stack.js:
1873         * JSAir/arg.js:
1874         (Arg.createImm):
1875         (Arg.createBigImm):
1876         (Arg.createBitImm):
1877         (Arg.createBitImm64):
1878         (Arg.createWidth):
1879         (Arg.createSpecial):
1880         (Arg.prototype.get kind):
1881         (Arg.prototype.get isTmp):
1882         (Arg.prototype.get isImm):
1883         (Arg.prototype.get isSomeImm):
1884         (Arg.prototype.get isSomeBigImm):
1885         (Arg.prototype.get isCondition):
1886         (Arg.prototype.get isWidth):
1887         (Arg.prototype.get isSpecial):
1888         (Arg.prototype.get isAlive):
1889         (Arg.prototype.get tmp):
1890         (Arg.prototype.get value):
1891         (Arg.prototype.get lowValue):
1892         (Arg.prototype.get highValue):
1893         (Arg.prototype.get base):
1894         (Arg.prototype.get isGP):
1895         (Arg.prototype.get isFP):
1896         (Arg.prototype.isValidForm):
1897         (Arg.prototype.get isInvertible):
1898         (Arg.kindCode):
1899         (Arg.prototype.hash):
1900         (Arg.prototype.toString):
1901         (Arg):
1902         * JSAir/basic_block.js:
1903         (BasicBlock.get successorBlocks):
1904         * JSAir/benchmark.js:
1905         (benchmark):
1906         * JSAir/code.js:
1907         (Code.prototype.setFrameSize):
1908         (Code.prototype.hash):
1909         (Code.prototype.toString):
1910         (Code):
1911         * JSAir/inst.js:
1912         (Inst.prototype.get hasNonArgEffects):
1913         (Inst.prototype.hash):
1914         (Inst.prototype.toString):
1915         (Inst):
1916         * JSAir/jsair-tests.yaml: Added.
1917         * JSAir/opcode.js:
1918         (Inst_forEachArg):
1919         (Inst_hasNonArgEffects):
1920         (opcodeCode):
1921         * JSAir/payload-gbemu-executeIteration.js:
1922         (createPayloadGbemuExecuteIteration):
1923         * JSAir/payload-imaging-gaussian-blur-gaussianBlur.js:
1924         (createPayloadImagingGaussianBlurGaussianBlur):
1925         * JSAir/payload-jsair-ACLj8C.js:
1926         (createPayloadJSAirACLj8C):
1927         * JSAir/payload-typescript-scanIdentifier.js:
1928         (createPayloadTypescriptScanIdentifier):
1929         * JSAir/reg.js:
1930         (Reg.prototype.get isReg):
1931         (Reg.prototype.hash):
1932         (Reg.prototype.toString):
1933         * JSAir/stack_slot.js:
1934         (StackSlot.prototype.setOffsetFromFP):
1935         (StackSlot.prototype.hash):
1936         (StackSlot.prototype.toString):
1937         * JSAir/symbols.js:
1938         (relCondCode):
1939         (resCondCode):
1940         (doubleCondCode):
1941         * JSAir/test.html:
1942         * JSAir/tmp.js:
1943         (Tmp.prototype.get isReg):
1944         (Tmp.prototype.hash):
1945         (Tmp.prototype.toString):
1946         * JSAir/util.js:
1947         (roundUpToMultipleOf):
1948         (symbolName):
1949
1950 2016-06-07  Filip Pizlo  <fpizlo@apple.com>
1951
1952         Implement Air::allocateStack() in ES6 to see how much of a bad idea that is
1953         https://bugs.webkit.org/show_bug.cgi?id=158318
1954
1955         Reviewed by Saam Barati.
1956         
1957         This adds a new benchmark for us to play with called JSAir. It's a complete ES6
1958         implementation of Air's allocateStack() phase along with all of Air needed to run that
1959         phase. This includes things like stack slots, registers, temporaries, basic blocks,
1960         instructions, and all of the code for iterating over, inspecting, and modifying those
1961         things.
1962         
1963         To make this work, JSC can now dump Air just before allocateStack() in the form of JS code
1964         that creates a Code object that matches exactly what C++ Air saw. This benchmark comprises
1965         four Air IRs:
1966         
1967         - Octane/gbemu's largest function, executeIteration.
1968         - Kraken/imaging-gaussian-blur's largest function in OSR entry mode, gaussuanBlur.
1969         - Octane/typescript's largest function that is actually hot, scanIdentifier.
1970         - JSAir's largest hot function, which is anonymous, so we call it by its hash (ACLj8C).
1971         
1972         This runs in about 2 seconds on my machine in JSC trunk. It includes both a commandline
1973         harness and a web harness. JSAir is almost exactly 100x slower in ES6 in WebKit than the
1974         C++ Air::allocateStack() phase on which it is based.
1975         
1976         JSAir uses the following ES6 features:
1977         
1978         - Symbol.
1979         - for-of.
1980         - arrow functions.
1981         - Map/Set.
1982         - let/const.
1983         - classes.
1984         
1985         All of these things are used in anger and should end up on the hot path. There is also code
1986         that uses Proxies, but it ends up being dead.
1987         
1988         We can improve this even more:
1989         I still need to add result validation: https://bugs.webkit.org/show_bug.cgi?id=158493
1990         I want to make it use more ES6 features: https://bugs.webkit.org/show_bug.cgi?id=158497
1991         
1992         * JSAir: Added.
1993         * JSAir/all.js: Added.
1994         * JSAir/allocate_stack.js: Added.
1995         (allocateStack.attemptAssignment):
1996         (allocateStack.assign):
1997         (allocateStack.interfere):
1998         (allocateStack.):
1999         (allocateStack):
2000         * JSAir/arg.js: Added.
2001         (Arg):
2002         (Arg.isAnyUse):
2003         (Arg.isColdUse):
2004         (Arg.isWarmUse):
2005         (Arg.cooled):
2006         (Arg.isEarlyUse):
2007         (Arg.isLateUse):
2008         (Arg.isAnyDef):
2009         (Arg.isEarlyDef):
2010         (Arg.isLateDef):
2011         (Arg.isZDef):
2012         (Arg.typeForB3Type):
2013         (Arg.widthForB3Type):
2014         (Arg.conservativeWidth):
2015         (Arg.minimumWidth):
2016         (Arg.bytes):
2017         (Arg.widthForBytes):
2018         (Arg.createTmp):
2019         (Arg.fromReg):
2020         (Arg.createImm):
2021         (Arg.createBigImm):
2022         (Arg.createBitImm):
2023         (Arg.createBitImm64):
2024         (Arg.createAddr):
2025         (Arg.createStack):
2026         (Arg.createCallArg):
2027         (Arg.createStackAddr):
2028         (Arg.isValidScale):
2029         (Arg.logScale):
2030         (Arg.createIndex):
2031         (Arg.createRelCond):
2032         (Arg.createResCond):
2033         (Arg.createDoubleCond):
2034         (Arg.createWidth):
2035         (Arg.prototype.get kind):
2036         (Arg.prototype.get isTmp):
2037         (Arg.prototype.get isImm):
2038         (Arg.prototype.get isBigImm):
2039         (Arg.prototype.get isBitImm):
2040         (Arg.prototype.get isBitImm64):
2041         (Arg.prototype.get isSomeImm):
2042         (Arg.prototype.get isAddr):
2043         (Arg.prototype.get isStack):
2044         (Arg.prototype.get isCallArg):
2045         (Arg.prototype.get isIndex):
2046         (Arg.prototype.get isMemory):
2047         (Arg.prototype.get isStackMemory):
2048         (Arg.prototype.get isRelCond):
2049         (Arg.prototype.get isResCond):
2050         (Arg.prototype.get isDoubleCond):
2051         (Arg.prototype.get isCondition):
2052         (Arg.prototype.get isWidth):
2053         (Arg.prototype.get isAlive):
2054         (Arg.prototype.get tmp):
2055         (Arg.prototype.get value):
2056         (Arg.prototype.get base):
2057         (Arg.prototype.get hasOffset):
2058         (Arg.prototype.get offset):
2059         (Arg.prototype.get stackSlot):
2060         (Arg.prototype.get index):
2061         (Arg.prototype.get scale):
2062         (Arg.prototype.get logScale):
2063         (Arg.prototype.get width):
2064         (Arg.prototype.get isGPTmp):
2065         (Arg.prototype.get isFPTmp):
2066         (Arg.prototype.get isGP):
2067         (Arg.prototype.get isFP):
2068         (Arg.prototype.get hasType):
2069         (Arg.prototype.get type):
2070         (Arg.prototype.isType):
2071         (Arg.prototype.isCompatibleType):
2072         (Arg.prototype.get isGPR):
2073         (Arg.prototype.get gpr):
2074         (Arg.prototype.get isFPR):
2075         (Arg.prototype.get fpr):
2076         (Arg.prototype.get isReg):
2077         (Arg.prototype.get reg):
2078         (Arg.isValidImmForm):
2079         (Arg.isValidBitImmForm):
2080         (Arg.isValidBitImm64Form):
2081         (Arg.isValidAddrForm):
2082         (Arg.isValidIndexForm):
2083         (Arg.prototype.isValidForm):
2084         (Arg.prototype.forEachTmpFast):
2085         (Arg.prototype.usesTmp):
2086         (Arg.prototype.forEachTmp):
2087         (Arg.prototype.is):
2088         (Arg.prototype.as):
2089         (Arg.prototype.forEachFast):
2090         (Arg.prototype.forEach):
2091         (Arg.extract):
2092         (Arg.forEachFast):
2093         (Arg.forEach):
2094         (Arg.prototype.get condition):
2095         (Arg.prototype.get isInvertible):
2096         (Arg.prototype.toString):
2097         * JSAir/basic_block.js: Added.
2098         (BasicBlock):
2099         (BasicBlock.prototype.get index):
2100         (BasicBlock.prototype.get size):
2101         (BasicBlock.prototype.Symbol.iterator):
2102         (BasicBlock.prototype.at):
2103         (BasicBlock.get last):
2104         (BasicBlock.get insts):
2105         (BasicBlock.get numSuccessors):
2106         (BasicBlock.get successors):
2107         (BasicBlock.get successorBlocks.):
2108         (BasicBlock.get successorBlocks):
2109         (BasicBlock.set get numPredecessors):
2110         (BasicBlock.get predecessors):
2111         (BasicBlock.get frequency):
2112         (BasicBlock.get headerString):
2113         (BasicBlock.get footerString):
2114         * JSAir/benchmark.js: Added.
2115         (benchmark):
2116         * JSAir/code.js: Added.
2117         (Code):
2118         (Code.prototype.addBlock):
2119         (Code.prototype.addStackSlot):
2120         (Code.prototype.newTmp):
2121         (Code.prototype.get size):
2122         (Code.prototype.at):
2123         (Code.prototype.Symbol.iterator):
2124         (Code.prototype.get blocks):
2125         (Code.prototype.get stackSlots):
2126         (Code.prototype.tmps):
2127         (Code.prototype.get callArgAreaSize):
2128         (Code.prototype.requestCallArgAreaSize):
2129         (Code.prototype.get frameSize):
2130         (Code.prototype.setFrameSize):
2131         (Code.prototype.toString):
2132         * JSAir/custom.js: Added.
2133         (const.ShuffleCustom.forEachArg):
2134         (const.ShuffleCustom.hasNonArgNonControlEffects):
2135         (const.PatchCustom.forEachArg):
2136         (const.PatchCustom.hasNonArgNonControlEffects):
2137         (const.CCallCustom.forEachArg):
2138         (const.CCallCustom.hasNonArgNonControlEffects):
2139         (const.ColdCCallCustom.forEachArg):
2140         (const.ColdCCallCustom.hasNonArgNonControlEffects):
2141         * JSAir/frequented_block.js: Added.
2142         (FrequentedBlock):
2143         (FrequentedBlock.prototype.toString):
2144         * JSAir/insertion_set.js: Added.
2145         (Insertion):
2146         (Insertion.prototype.get index):
2147         (Insertion.prototype.get element):
2148         (Insertion.prototype.lessThan):
2149         (InsertionSet):
2150         (InsertionSet.prototype.appendInsertion):
2151         (InsertionSet.prototype.append):
2152         (InsertionSet.prototype.execute):
2153         * JSAir/inst.js: Added.
2154         (Inst):
2155         (Inst.prototype.append):
2156         (Inst.prototype.clear):
2157         (Inst.prototype.get opcode):
2158         (Inst.prototype.get args):
2159         (Inst.prototype.visitArg):
2160         (Inst.prototype.forEachTmpFast):
2161         (Inst.prototype.forEachArg):
2162         (Inst.prototype.forEachTmp):
2163         (Inst.prototype.forEach):
2164         (Inst.forEachDef):
2165         (Inst.forEachDefWithExtraClobberedRegs):
2166         (Inst.prototype.get hasNonArgEffects):
2167         (Inst.prototype.toString):
2168         * JSAir/liveness.js: Added.
2169         (Liveness):
2170         (Liveness.prototype.get thing):
2171         (Liveness.prototype.get code):
2172         (Liveness.prototype.get liveAtHead):
2173         (Liveness.prototype.get liveAtTail):
2174         (Liveness.prototype.localCalc.LocalCalc):
2175         (Liveness.prototype.localCalc.LocalCalc.prototype.get liveSet):
2176         (Liveness.prototype.localCalc.LocalCalc.prototype.execute):
2177         (Liveness.prototype.localCalc):
2178         * JSAir/opcode.js: Added.
2179         (Inst_forEachArg):
2180         (Inst_hasNonArgEffects):
2181         * JSAir/payload-gbemu-executeIteration.js: Added.
2182         (createPayloadGbemuExecuteIteration):
2183         * JSAir/payload-imaging-gaussian-blur-gaussianBlur.js: Added.
2184         (createPayloadImagingGaussianBlurGaussianBlur):
2185         * JSAir/payload-jsair-ACLj8C.js: Added.
2186         (createPayloadJSAirACLj8C):
2187         * JSAir/payload-typescript-scanIdentifier.js: Added.
2188         (createPayloadTypescriptScanIdentifier):
2189         * JSAir/reg.js: Added.
2190         (Reg):
2191         (Reg.fromReg):
2192         (Reg.prototype.get index):
2193         (Reg.prototype.get type):
2194         (Reg.prototype.get name):
2195         (Reg.prototype.get isCalleeSave):
2196         (Reg.prototype.get isReg):
2197         (Reg.prototype.toString):
2198         (Reg.extract):
2199         (Reg.forEachFast):
2200         (Reg.forEach):
2201         (newGPR):
2202         (Reg.gprs.Reg.fprs.Reg.calleeSaveGPRs.Reg.calleeSaveFPRs.Reg.calleeSaves):
2203         * JSAir/stack_slot.js: Added.
2204         (StackSlot):
2205         (StackSlot.prototype.get byteSize):
2206         (StackSlot.prototype.get kind):
2207         (StackSlot.prototype.get isLocked):
2208         (StackSlot.prototype.get isSpill):
2209         (StackSlot.prototype.get index):
2210         (StackSlot.prototype.ensureSize):
2211         (StackSlot.prototype.get alignment):
2212         (StackSlot.prototype.get offsetFromFP):
2213         (StackSlot.prototype.setOffsetFromFP):
2214         (StackSlot.prototype.toString):
2215         (StackSlot.extract):
2216         (StackSlot.forEachFast):
2217         (StackSlot.forEach):
2218         * JSAir/symbols.js: Added.
2219         * JSAir/test.html: Added.
2220         * JSAir/test.js: Added.
2221         * JSAir/tmp.js: Added.
2222         (Tmp):
2223         (Tmp.fromReg):
2224         (Tmp.prototype.get index):
2225         (Tmp.prototype.get type):
2226         (Tmp.prototype.get isReg):
2227         (Tmp.prototype.toString):
2228         (Tmp.extract):
2229         (Tmp.forEachFast):
2230         (Tmp.forEach):
2231         * JSAir/tmp_base.js: Added.
2232         (TmpBase.prototype.get isGP):
2233         (TmpBase.prototype.get isFP):
2234         (TmpBase.prototype.get isGPR):
2235         (TmpBase.prototype.get isFPR):
2236         (TmpBase.prototype.get reg):
2237         (TmpBase.prototype.get gpr):
2238         (TmpBase.prototype.get fpr):
2239         (TmpBase):
2240         * JSAir/util.js: Added.
2241         (isRepresentableAsInt32):
2242         (addIndexed):
2243         (roundUpToMultipleOf):
2244         (symbolName):
2245         (mergeIntoSet):
2246         (nonEmptyRangesOverlap):
2247         (rangesOverlap):
2248         (removeAllMatching):
2249         (swap):
2250         (bubble):
2251         (bubbleSort):
2252         (currentTime):
2253         (else.currentTime):
2254
2255 2016-05-26  Yusuke Suzuki  <utatane.tea@gmail.com>
2256
2257         [JSC] Allow JSBench to use precise time
2258         https://bugs.webkit.org/show_bug.cgi?id=158050
2259
2260         Reviewed by Geoffrey Garen.
2261
2262         * JSBench/amazon-chrome-win/urem.js:
2263         (else.window.performance.window.performance.now.currentTimeInMS):
2264         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2265         (else.else.currentTimeInMS):
2266         (onload.cb):
2267         (onload):
2268         * JSBench/amazon-chrome/urem.js:
2269         (else.window.performance.window.performance.now.currentTimeInMS):
2270         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2271         (else.else.currentTimeInMS):
2272         (onload.cb):
2273         (onload):
2274         * JSBench/amazon-firefox-win/urm.js:
2275         (else.window.performance.window.performance.now.currentTimeInMS):
2276         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2277         (else.else.currentTimeInMS):
2278         (onload.cb):
2279         (onload):
2280         * JSBench/amazon-firefox/urm.js:
2281         (else.window.performance.window.performance.now.currentTimeInMS):
2282         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2283         (else.else.currentTimeInMS):
2284         (onload.cb):
2285         (onload):
2286         * JSBench/amazon-safari/urem.js:
2287         (else.window.performance.window.performance.now.currentTimeInMS):
2288         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2289         (else.else.currentTimeInMS):
2290         (onload.cb):
2291         (onload):
2292         * JSBench/facebook-chrome-win/urem.js:
2293         (else.window.performance.window.performance.now.currentTimeInMS):
2294         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2295         (else.else.currentTimeInMS):
2296         (onload.cb):
2297         (onload):
2298         * JSBench/facebook-chrome/urem.js:
2299         (else.window.performance.window.performance.now.currentTimeInMS):
2300         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2301         (else.else.currentTimeInMS):
2302         (onload.cb):
2303         (onload):
2304         * JSBench/facebook-firefox-win/urem.js:
2305         (else.window.performance.window.performance.now.currentTimeInMS):
2306         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2307         (else.else.currentTimeInMS):
2308         (onload.cb):
2309         (onload):
2310         * JSBench/facebook-firefox/urem.js:
2311         (else.window.performance.window.performance.now.currentTimeInMS):
2312         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2313         (else.else.currentTimeInMS):
2314         (onload.cb):
2315         (onload):
2316         * JSBench/facebook-safari/urem.js:
2317         (else.window.performance.window.performance.now.currentTimeInMS):
2318         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2319         (else.else.currentTimeInMS):
2320         (onload.cb):
2321         (onload):
2322         * JSBench/google-chrome-win/urem.js:
2323         (else.window.performance.window.performance.now.currentTimeInMS):
2324         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2325         (else.else.currentTimeInMS):
2326         (onload.cb):
2327         (onload):
2328         * JSBench/google-chrome/urem.js:
2329         (else.window.performance.window.performance.now.currentTimeInMS):
2330         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2331         (else.else.currentTimeInMS):
2332         (onload.cb):
2333         (onload):
2334         * JSBench/google-firefox-win/urem.js:
2335         (else.window.performance.window.performance.now.currentTimeInMS):
2336         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2337         (else.else.currentTimeInMS):
2338         (onload.cb):
2339         (onload):
2340         * JSBench/google-firefox/uem.js:
2341         (else.window.performance.window.performance.now.currentTimeInMS):
2342         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2343         (else.else.currentTimeInMS):
2344         (onload.cb):
2345         (onload):
2346         * JSBench/google-safari/urem.js:
2347         (else.window.performance.window.performance.now.currentTimeInMS):
2348         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2349         (else.else.currentTimeInMS):
2350         (onload.cb):
2351         (onload):
2352         * JSBench/harness.js:
2353         (runBenchmark.window.currentTimeInMS):
2354         (runBenchmark.else.window.currentTimeInMS):
2355         * JSBench/twitter-chrome-win/rem.js:
2356         (else.window.performance.window.performance.now.currentTimeInMS):
2357         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2358         (else.else.currentTimeInMS):
2359         (onload.cb):
2360         (onload):
2361         * JSBench/twitter-chrome/urem.js:
2362         (else.window.performance.window.performance.now.currentTimeInMS):
2363         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2364         (else.else.currentTimeInMS):
2365         (onload.cb):
2366         (onload):
2367         * JSBench/twitter-firefox-win/urem.js:
2368         (else.window.performance.window.performance.now.currentTimeInMS):
2369         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2370         (else.else.currentTimeInMS):
2371         (onload.cb):
2372         (onload):
2373         * JSBench/twitter-firefox/urem.js:
2374         (else.window.performance.window.performance.now.currentTimeInMS):
2375         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2376         (else.else.currentTimeInMS):
2377         (onload.cb):
2378         (onload):
2379         * JSBench/twitter-safari/urem.js:
2380         (else.window.performance.window.performance.now.currentTimeInMS):
2381         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2382         (else.else.currentTimeInMS):
2383         (onload.cb):
2384         (onload):
2385         * JSBench/yahoo-chrome-win/urem.js:
2386         (else.window.performance.window.performance.now.currentTimeInMS):
2387         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2388         (else.else.currentTimeInMS):
2389         (onload.cb):
2390         (onload):
2391         * JSBench/yahoo-chrome/urem.js:
2392         (else.window.performance.window.performance.now.currentTimeInMS):
2393         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2394         (else.else.currentTimeInMS):
2395         (onload.cb):
2396         (onload):
2397         * JSBench/yahoo-firefox-win/urem.js:
2398         (else.window.performance.window.performance.now.currentTimeInMS):
2399         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2400         (else.else.currentTimeInMS):
2401         (onload.cb):
2402         (onload):
2403         * JSBench/yahoo-firefox/urem.js:
2404         (else.window.performance.window.performance.now.currentTimeInMS):
2405         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2406         (else.else.currentTimeInMS):
2407         (onload.cb):
2408         (onload):
2409         * JSBench/yahoo-safari/urem.js:
2410         (else.window.performance.window.performance.now.currentTimeInMS):
2411         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2412         (else.else.currentTimeInMS):
2413         (onload.cb):
2414         (onload):
2415
2416 2016-05-25  Geoffrey Garen  <ggaren@apple.com>
2417
2418         replaceable own properties seem to ignore replacement after property caching
2419         https://bugs.webkit.org/show_bug.cgi?id=158091
2420
2421         Reviewed by Darin Adler.
2422
2423         * MallocBench/MallocBench.xcodeproj/project.pbxproj:
2424         * MallocBench/MallocBench/Benchmark.cpp:
2425         * MallocBench/MallocBench/Interpreter.cpp:
2426         (Interpreter::doMallocOp):
2427         * MallocBench/MallocBench/Interpreter.h:
2428         * MallocBench/MallocBench/fastMallocLog.63316.ops: Added.
2429         * MallocBench/MallocBench/jetstream.cpp: Added.
2430         (benchmark_jetstream):
2431         * MallocBench/MallocBench/jetstream.h: Added.
2432
2433 2016-05-25  Keith Miller  <keith_miller@apple.com>
2434
2435         Unreviewed, add JSBench to the skipped list for now since it doesn't
2436         work currently.
2437
2438         * Skipped:
2439
2440 2016-05-17  Keith Miller  <keith_miller@apple.com>
2441         We should have JSBench in PerformanceTests
2442         https://bugs.webkit.org/show_bug.cgi?id=157952
2443
2444         Rubber-stamped by Saam Barati.
2445
2446         There are some slight changes to the layout of the test directory
2447         to make it work nicely with run-jsc-benchmarks. Before JSBench
2448         had each of the browser specific sub-tests in a sub-directory.
2449         These have been flattened e.g. amazon/safari/ has become
2450         amazon-safari/.
2451
2452         * JSBench/amazon-chrome-win/urem.html: Added.
2453         * JSBench/amazon-chrome-win/urem.js: Added.
2454         * JSBench/amazon-chrome/urem.html: Added.
2455         * JSBench/amazon-chrome/urem.js: Added.
2456         * JSBench/amazon-firefox-win/urm.html: Added.
2457         * JSBench/amazon-firefox-win/urm.js: Added.
2458         * JSBench/amazon-firefox/urm.html: Added.
2459         * JSBench/amazon-firefox/urm.js: Added.
2460         * JSBench/amazon-safari/urem.html: Added.
2461         * JSBench/amazon-safari/urem.js: Added.
2462         * JSBench/browsercheck.js: Added.
2463         * JSBench/facebook-chrome-win/urem.html: Added.
2464         * JSBench/facebook-chrome-win/urem.js: Added.
2465         * JSBench/facebook-chrome/urem.html: Added.
2466         * JSBench/facebook-chrome/urem.js: Added.
2467         * JSBench/facebook-firefox-win/urem.html: Added.
2468         * JSBench/facebook-firefox-win/urem.js: Added.
2469         * JSBench/facebook-firefox/urem.html: Added.
2470         * JSBench/facebook-firefox/urem.js: Added.
2471         * JSBench/facebook-safari/urem.html: Added.
2472         * JSBench/facebook-safari/urem.js: Added.
2473         * JSBench/google-chrome-win/urem.html: Added.
2474         * JSBench/google-chrome-win/urem.js: Added.
2475         * JSBench/google-chrome/urem.html: Added.
2476         * JSBench/google-chrome/urem.js: Added.
2477         * JSBench/google-firefox-win/urem.html: Added.
2478         * JSBench/google-firefox-win/urem.js: Added.
2479         * JSBench/google-firefox/uem.html: Added.
2480         * JSBench/google-firefox/uem.js: Added.
2481         * JSBench/google-safari/urem.html: Added.
2482         * JSBench/google-safari/urem.js: Added.
2483         * JSBench/harness.html: Added.
2484         * JSBench/harness.js: Added.
2485         * JSBench/harness.py: Added.
2486         * JSBench/index.html: Added.
2487         * JSBench/reload.html: Added.
2488         * JSBench/twitter-chrome-win/rem.html: Added.
2489         * JSBench/twitter-chrome-win/rem.js: Added.
2490         * JSBench/twitter-chrome/urem.html: Added.
2491         * JSBench/twitter-chrome/urem.js: Added.
2492         * JSBench/twitter-firefox-win/urem.html: Added.
2493         * JSBench/twitter-firefox-win/urem.js: Added.
2494         * JSBench/twitter-firefox/urem.html: Added.
2495         * JSBench/twitter-firefox/urem.js: Added.
2496         * JSBench/twitter-safari/urem.html: Added.
2497         * JSBench/twitter-safari/urem.js: Added.
2498         * JSBench/yahoo-chrome-win/urem.html: Added.
2499         * JSBench/yahoo-chrome-win/urem.js: Added.
2500         * JSBench/yahoo-chrome/urem.html: Added.
2501         * JSBench/yahoo-chrome/urem.js: Added.
2502         * JSBench/yahoo-firefox-win/urem.html: Added.
2503         * JSBench/yahoo-firefox-win/urem.js: Added.
2504         * JSBench/yahoo-firefox/urem.html: Added.
2505         * JSBench/yahoo-firefox/urem.js: Added.
2506         * JSBench/yahoo-safari/urem.html: Added.
2507         * JSBench/yahoo-safari/urem.js: Added.
2508
2509 2016-05-20  Ryosuke Niwa  <rniwa@webkit.org>
2510
2511         run-benchmark's results should contain Animometer's debug output
2512         https://bugs.webkit.org/show_bug.cgi?id=157941
2513
2514         Reviewed by Stephanie Lewis.
2515
2516         Made developer.html support the JSON generated by run-benchmark which stores everything under debugOutput.
2517
2518         * Animometer/resources/debug-runner/animometer.js:
2519         (Utilities.initialize): Unwrap debugOutput in the case run-benchmark's result JSON is used.
2520
2521 2016-05-18  Timothy Hatcher  <timothy@apple.com>
2522
2523         Make Animometer work in all browsers
2524         https://bugs.webkit.org/show_bug.cgi?id=157855
2525         rdar://problem/26338521
2526
2527         Reviewed by Jon Lee.
2528
2529         * Animometer/resources/runner/logo.svg: Use a symbol instead
2530         of putting the id on the root svg. This works in all browsers.
2531
2532 2016-05-07  Timothy Hatcher  <timothy@apple.com>
2533
2534         Update Animometer design
2535         https://bugs.webkit.org/show_bug.cgi?id=157449
2536
2537         Reviewed by Darin Adler.
2538
2539         * Animometer/developer.html: Stop including runner/animometer.css.
2540         * Animometer/index.html: Udpated elements to fit new styles.
2541
2542         * Animometer/resources/debug-runner/animometer.css:
2543         Copy the old runner/animometer.css to here so the debug runner is unaffected.
2544
2545         * Animometer/resources/runner/animometer.css:
2546         New styles for the design. Drop flex box for broad browser support.
2547
2548         * Animometer/resources/runner/animometer.js:
2549         (ResultsTable.clear): Use textContent.
2550         (window.sectionsManager.showSection): Add class to the body.
2551         (window.sectionsManager.setSectionScore): Use textContent, not innerHTML.
2552
2553         * Animometer/resources/runner/crystal.svg: Added.
2554         * Animometer/resources/runner/lines.svg: Added.
2555         * Animometer/resources/runner/logo.svg: Added.
2556
2557         * Animometer/resources/runner/tests.js: Shorten test name and title case them.
2558
2559 2016-05-06  Manuel Rego Casasnovas  <rego@igalia.com>
2560
2561         [css-grid] Unprefix CSS Grid Layout properties
2562         https://bugs.webkit.org/show_bug.cgi?id=157137
2563
2564         Reviewed by Simon Fraser.
2565
2566         Remove "-webkit" prefix from all the grid layout properties,
2567         including the display value.
2568         Update the source code to remove the prefix where it was used too.
2569
2570         * Layout/auto-grid-lots-of-data.html:
2571         * Layout/fixed-grid-lots-of-data.html:
2572         * Layout/fixed-grid-lots-of-stretched-data.html:
2573
2574 2016-04-25  Geoffrey Garen  <ggaren@apple.com>
2575
2576         bmalloc: Misc improvements to MallocBench
2577         https://bugs.webkit.org/show_bug.cgi?id=157004
2578
2579         Reviewed by Darin Adler.
2580
2581         * MallocBench/run-malloc-benchmarks: Added --memory and --memory_warning
2582         modes for focused memory testing.
2583
2584         * MallocBench/MallocBench/Benchmark.cpp:
2585         (Benchmark::printReport): Clarified output.
2586
2587         (Benchmark::currentMemoryBytes): Added compressed memory because top
2588         does the same. (It always happens to zero in the benchmarks we run. But
2589         this is good for sanity.)
2590
2591         * MallocBench/MallocBench/CommandLine.cpp: Moved up to 8 runs to reduce
2592         variance.
2593
2594         * MallocBench/MallocBench/alloc_free.cpp:
2595         (benchmark_alloc_free): Cycle a single allocation in order to stress
2596         the effect of merging on calls to madvise.
2597
2598         * MallocBench/MallocBench/big.cpp:
2599         (benchmark_big): Graduated to 8kB-128kB because medium tests up to 8 and
2600         our large allocator doesn't kick in until 64kB.
2601
2602         * MallocBench/MallocBench/medium.cpp:
2603         (benchmark_medium): Test all the way down to 1kB because our large
2604         allocator used to service 1kB allocations and 1kB is an interesting
2605         middle size where memory is unusually large but allocation throughput
2606         still matters.
2607
2608         * MallocBench/MallocBench/stress.cpp:
2609         (benchmark_stress): Reduced the churn count to match stress_aligned
2610         because this test was taking too long to complete.
2611
2612         * MallocBench/MallocBench/stress_aligned.cpp:
2613         (benchmark_stress_aligned): Our new large allocator can handle even
2614         more absurdly large values.
2615
2616 2016-04-25  Simon Fraser  <simon.fraser@apple.com>
2617
2618         Add a content animation test that uses SVG animation.
2619         https://bugs.webkit.org/show_bug.cgi?id=156827
2620
2621         * Animation/svg-animation.html: Added.
2622         * Skipped:
2623
2624 2016-04-22  Simon Fraser  <simon.fraser@apple.com>
2625
2626         Skip two content animation tests which are only meant for iOS testing.
2627
2628         * Animation/css-animation.html: Added.
2629         * Animation/raf-animation.html: Added.
2630
2631         * Skipped:
2632
2633 2016-04-20  Simon Fraser  <simon.fraser@apple.com>
2634
2635         Add content animation tests to benchmark_runner, and allow the runner to collect device data as part of the results
2636         https://bugs.webkit.org/show_bug.cgi?id=156827
2637
2638         Add two files missing from the previous commit.
2639
2640         * Animation/css-animation.html: Added.
2641         * Animation/raf-animation.html: Added.
2642
2643 2016-04-20  Simon Fraser  <simon.fraser@apple.com>
2644
2645         Add content animation tests to benchmark_runner, and allow the runner to collect device data as part of the results
2646         https://bugs.webkit.org/show_bug.cgi?id=156827
2647
2648         Reviewed by Stephanie Lewis.
2649
2650         Two new files for the content-animation suite.
2651
2652         After patching for use in the benchmark, these tests are a little different to other benchmarks.
2653         The test content itself does not collect performance data; the tests spit out a boilerplate JSON
2654         result with a placeholder for framerate, which is collected in native code.
2655
2656         When run as a benchmark, test completion involves a couple of bounces between the test and MobileSafari.
2657         Test completion is initiated by a setTimeout() in the test, at which point location.hash is set to "#done".
2658         MobileSafari detects that, and asynchronously dumps performance data. Once this is complete, MobileSafari
2659         changes the URL has to "#submit", which triggers the XHR to the test relay that indicates the test is
2660         complete.
2661
2662         * Animation/css-animation.html: Added.
2663         * Animation/raf-animation.html: Added.
2664
2665 2016-04-15  Jon Lee  <jonlee@apple.com>
2666
2667         Animometer test could report a NaN
2668         https://bugs.webkit.org/show_bug.cgi?id=156646
2669
2670         Reviewed by Darin Adler.
2671         Provisionally reviewed by Said Abou-Hallawa.
2672
2673         * Animometer/tests/resources/main.js:
2674         (didFinishInterval): The ramp controller has a first phase where it ramps up the complexity and
2675         reacts based on how well the system handles the load. The assumption was that it would handle at least
2676         1 particle easily. That is not always the case. As a result, an interpolation calculation could
2677         end up setting an upper bound of NaN.
2678
2679         This occurs because we never get out of the first tier, so this._lastTierComplexity is undefined.
2680         Now that we guarantee a minimum complexity of 1, modify the conditional to check for this._lastTierComplexity
2681         before interpolating the upper bound of the first ramp. In the case where the system struggles with
2682         1 particle, set it to currentComplexity.
2683
2684 2016-04-09  Jon Lee  <jonlee@apple.com>
2685
2686         Update Animometer to accommodate different screens
2687         https://bugs.webkit.org/show_bug.cgi?id=156449
2688
2689         Reviewed by Darin Adler.
2690         Provisionally reviewed by Said Abou-Hallawa.
2691
2692         * Animometer/index.html: Wrap button in a container to add padding at the bottom.
2693         * Animometer/resources/debug-runner/animometer.css:
2694         (@media screen and (min-device-width: 1800px)): Deleted.
2695         (@media screen and (min-width: 1800px)): Cannot use min-device-width since it may match incorrectly.
2696         (screen and (max-device-height: 414px) and (orientation: landscape)): Some devices swap device width
2697         and height with orientation change.
2698         * Animometer/resources/runner/animometer.css: Similar.
2699         (screen and (min-device-width: 1024px) and (orientation: landscape)):
2700         (screen and (max-device-height: 414px) and (orientation: landscape)):
2701         (.frame-container): On smaller iPhones, adding 1px prevents the navigation bars from appearing.
2702         (@media screen and (min-device-width: 768px) and (max-device-width: 1024px)): Deleted.
2703         (@media (min-device-height: 768px) and (max-device-height: 1024px)): Target iPad Airs and similar.
2704         (@media screen and (min-device-width: 1024px) and (max-device-width: 1366px)): Deleted.
2705         (@media screen and (max-device-width: 1024px) and (min-device-height: 1366px)): Target iPad Pro.
2706         (#results footer): Add padding below the button for testing again.
2707         * Animometer/tests/master/multiply.html: Remove the center text.
2708         * Animometer/tests/master/resources/text.js: Update the test so that in every frame the text moves.
2709         * Animometer/tests/master/text.html: Update the text sizing depending on the size of the device.
2710
2711 2016-04-08  Jon Lee  <jonlee@apple.com>
2712
2713         Have Animometer benchmark always start with complexity of 1
2714         https://bugs.webkit.org/show_bug.cgi?id=156432
2715
2716         Reviewed by Ryosuke Niwa.
2717
2718         * Animometer/tests/resources/main.js: Update the default Controller and RampController to
2719         set its minimum complexities to 1 instead of 0.
2720
2721 2016-04-08  Jon Lee  <jonlee@apple.com>
2722
2723         Fix SVG benchmark test
2724         https://bugs.webkit.org/show_bug.cgi?id=156410
2725
2726         Reviewed by Dean Jackson.
2727
2728         * Animometer/resources/extensions.js: Update Point.zero to be a static Point.
2729         * Animometer/tests/simple/resources/tiled-canvas-image.js:
2730         (Stage.call._setupTiles): Refactor.
2731         * Animometer/tests/master/resources/particles.js:
2732         (Particle.prototype.reset): Use Point.center.
2733         (complexity): We are not using a gradient background anymore, so remove the +1.
2734         * Animometer/tests/master/resources/svg-particles.js: Update to use SVG transform
2735         instead of CSS transform.
2736
2737 2016-04-06  Jon Lee  <jonlee@apple.com>
2738
2739         Update master benchmark with SVG test
2740         https://bugs.webkit.org/show_bug.cgi?id=156273
2741
2742         Reviewed by Dean Jackson.
2743         Provisionally reviewed by Said Abou-Hallawa.
2744
2745         Switch masks tests for SVG path test.
2746
2747         * Animometer/resources/debug-runner/tests.js: Move mask test here.
2748         * Animometer/resources/runner/tests.js: Add SVG test here.
2749         * Animometer/tests/dom/particles.html: Renamed from PerformanceTests/Animometer/tests/master/particles.html.
2750         * Animometer/tests/dom/resources/dom-particles.js: Renamed from PerformanceTests/Animometer/tests/master/resources/dom-particles.js.
2751
2752         * Animometer/tests/master/resources/particles.js: Add minPosition for bounds checking. Prevents particle from being
2753         partially obscured.
2754         * Animometer/tests/master/resources/svg-particles.js: Added.
2755         (Particle): The particle is either a path object or a rect using a path as a clip. The
2756         same path is used either way. For each particle create a linear gradient with a random
2757         rotation.
2758         (SVGParticlesStage): Look in #shapes to see how many different kinds of paths are available.
2759         This makes the test more generic in case other shapes need to be tested.
2760         * Animometer/tests/master/svg-particles.html: Added. Have two defs, one that houses each
2761         particle's gradient, and one that holds the shape templates.
2762
2763 2016-03-27  Hunseop Jeong  <hs85.jeong@samsung.com>
2764
2765         [EFL] REGRESSION(r188793): It made 200 layout tests and Bindings/event-target-wrapper.html performance test fail
2766         https://bugs.webkit.org/show_bug.cgi?id=148470 
2767
2768         Reviewed by Darin Adler.
2769
2770         * Skipped: Unskip the Bindings/event-target-wrapper test.
2771
2772 2016-03-23  Jon Lee  <jonlee@apple.com>
2773
2774         Update focus and leaves tests
2775         https://bugs.webkit.org/show_bug.cgi?id=155825
2776
2777         Reviewed by Simon Fraser.
2778
2779         Particles would improperly disappear, before they were fully off stage. Update the calculation.
2780         * Animometer/tests/dom/resources/leaves.js:
2781         * Animometer/tests/master/resources/leaves.js:
2782
2783         * Animometer/tests/master/focus.html: Update the center element styling.
2784         * Animometer/tests/master/resources/focus.js:
2785         (FocusElement.Utilities.createClass): Give enough space in the container for each particle for the blur
2786         to spread. It should be about 3x more space than the blur radius on each side.
2787         (Stage.call.initialize): Ditto for the center element.
2788
2789 2016-03-23  Jon Lee  <jonlee@apple.com>
2790
2791         Update support for other platforms
2792         https://bugs.webkit.org/show_bug.cgi?id=155824
2793
2794         Reviewed by Simon Fraser.
2795
2796         * Animometer/resources/debug-runner/animometer.css: Remove extraneous media query. All of it should
2797         be handled in the release stylesheet.
2798         * Animometer/resources/debug-runner/animometer.js: Add support for browsers that return an HTMLCollection
2799         when retrieving radio elements with a name, instead of the selected value.
2800         * Animometer/resources/runner/animometer.css: Adjust the queries for devices of varying widths.
2801         (.frame-container): Move overflow:hidden to #stage.
2802         * Animometer/tests/master/resources/stage.css: Consolidate into tests/resources/stage.css.
2803         * Animometer/tests/resources/stage.css:
2804         (body): Remove overflow:hidden.
2805         (#stage): Set overflow:hidden here.
2806         (#center-text): Moved from master/resources/stage.css.
2807
2808         Update stylesheet hrefs.
2809         * Animometer/tests/master/canvas-stage.html:
2810         * Animometer/tests/master/focus.html:
2811         * Animometer/tests/master/leaves.html:
2812         * Animometer/tests/master/multiply.html:
2813         * Animometer/tests/master/particles.html:
2814         * Animometer/tests/master/text.html: Adjust text size to fit in the device screen.
2815
2816 2016-03-23  Jon Lee  <jonlee@apple.com>
2817
2818         Make the benchmark require device to be in landscape orientation
2819         https://bugs.webkit.org/show_bug.cgi?id=155822
2820         rdar://problem/25258650
2821
2822         Reviewed by Ryosuke Niwa.
2823
2824         Check that the device is in landscape orientation. If not, disable the button to start it,
2825         and show a warning message.
2826
2827         * Animometer/developer.html: Add a message about how the browser should be set up. Include a
2828         warning message that appears if the orientation is incorrect on mobile devices.
2829         * Animometer/index.html: Ditto.
2830
2831         * Animometer/resources/debug-runner/animometer.css: Migrate the .hidden rule into the release
2832         stylesheet. Update the style to accommodate the new UI.
2833         * Animometer/resources/runner/animometer.css:
2834
2835         * Animometer/resources/runner/animometer.js:
2836         (window.benchmarkController.initialize): Add the orientation listener is needed.
2837         (window.benchmarkController.addOrientationListenerIfNecessary): Only mobile devices need this. Check
2838         to see for support of window.orientation.
2839         (window.benchmarkController._orientationChanged): Toggle the warning based on whether we match the
2840         landscape query. Set a state variable, which is needed for the debug harness. Call updateStartButtonState.
2841         (window.benchmarkController.updateStartButtonState):
2842         * Animometer/resources/debug-runner/animometer.js: Change _updateStartButtonState to return
2843         a boolean about whether at least one test is selected. That will be used in
2844         benchmarkController.updateStartButtonState(). Move the load event listener to the release version.
2845         (window.benchmarkController.updateStartButtonState): Override the release version, and also check
2846         that at least one test is selected.
2847
2848 2016-03-22  Geoffrey Garen  <ggaren@apple.com>
2849
2850         MallocBench: consolidate regression testing for aligned allocation
2851         https://bugs.webkit.org/show_bug.cgi?id=155762
2852
2853         Reviewed by Michael Saboff.
2854
2855         * MallocBench/MallocBench.xcodeproj/project.pbxproj:
2856         * MallocBench/MallocBench/Benchmark.cpp:
2857         * MallocBench/MallocBench/memalign.cpp: Removed.
2858         * MallocBench/MallocBench/memalign.h: Removed. The stress_aligned test
2859         covers this and much more.
2860
2861         * MallocBench/MallocBench/stress_aligned.cpp:
2862         (benchmark_stress_aligned): Include specific tests for extreme sizes
2863         and alignments.
2864
2865 2016-03-21  Jon Lee  <jonlee@apple.com>
2866
2867         Update benchmark tests
2868         https://bugs.webkit.org/show_bug.cgi?id=155723
2869
2870         Reviewed by Darin Adler.
2871         Provisionally reviewed by Said Abou-Hallawa.
2872
2873         Add quadratic and bezier segments to the canvas path test.
2874
2875         * Animometer/resources/runner/tests.js: Some of the query strings are unnecessary.
2876         Rename the test.
2877         * Animometer/tests/master/resources/canvas-stage.js:
2878         (tune): Update to be able to take an array of possible constructors. Choose one randomly.
2879         * Animometer/tests/master/resources/canvas-tests.js:
2880         (CanvasLinePoint.Utilities.createClass): Move the point selection out to a separate
2881         function called randomPoint() for reuse. Move X_LOOPS, Y_LOOPS, and offsets into the
2882         class definition.
2883         (randomPoint): Scale the grid down a little bit so that the lines along the edge of the
2884         canvas are not cut off when the stroke size is thick.
2885         (CanvasQuadraticSegment): Added.
2886         (CanvasBezierSegment): Added.
2887         (SimpleCanvasStage): Pass in an array of the different segment types. Since line segments
2888         are short compared to the curved ones, make it twice as likely to render a line segment.
2889         (SimpleCanvasStage.animate): Update the drawing code so that we render all line segments.
2890
2891         Add a helper method that selects a random element from an array.
2892
2893         * Animometer/tests/resources/main.js:
2894         (Stage.randomElementInArray): Select a random element from the provided array.
2895
2896         * Animometer/tests/bouncing-particles/resources/bouncing-tagged-images.js: Refactor.
2897         * Animometer/tests/dom/resources/leaves.js: Ditto.
2898         * Animometer/tests/master/resources/dom-particles.js: Ditto.
2899         * Animometer/tests/master/resources/image-data.js: Ditto.
2900         * Animometer/tests/master/resources/leaves.js: Ditto.
2901         * Animometer/tests/simple/resources/simple-canvas-paths.js: Ditto.
2902
2903         Add canvas tests that includes all stroke and fill paths. This makes it possible to avoid having to
2904         include the full simple canvas suite for perf testing.
2905
2906         * Animometer/resources/debug-runner/tests.js: Add new tests. Move the canvas test into the Canvas
2907         suite.
2908         * Animometer/tests/simple/resources/simple-canvas-paths.js: Add a CanvasStroke and CanvasFill particle
2909         that random selects an object to render.
2910
2911         * Animometer/resources/debug-runner/tests.js: Move 3D suite before basic canvas suite.
2912         When updating the perf bot script, we will include the suites up to this one, but exclude
2913         the basic canvas suite.
2914
2915         Move compositing transforms test to HTML suite and remove the empty Miscellaneous suite.
2916
2917         * Animometer/resources/debug-runner/tests.js:
2918         * Animometer/tests/dom/compositing-transforms.html: Renamed from PerformanceTests/Animometer/tests/misc/compositing-transforms.html.
2919         * Animometer/tests/dom/resources/compositing-transforms.js: Renamed from PerformanceTests/Animometer/tests/misc/resources/compositing-transforms.js.
2920
2921         Clean up miscellaneous test suite. Add a canvas ellipse test, and remove the other
2922         canvas tests.
2923
2924         * Animometer/resources/debug-runner/tests.js: Add ellipse tests to the simple suite.
2925         * Animometer/tests/misc/canvas-electrons.html: Removed.
2926         * Animometer/tests/misc/canvas-stars.html: Removed.
2927         * Animometer/tests/misc/resources/canvas-electrons.js: Removed.
2928         * Animometer/tests/misc/resources/canvas-stars.js: Removed.
2929         * Animometer/tests/simple/resources/simple-canvas-paths.js: Add ellipse primitives.
2930
2931         Merge text tests together into one.
2932
2933         * Animometer/resources/runner/tests.js: Remove international.html.
2934         * Animometer/tests/master/international.html: Removed.
2935         * Animometer/tests/master/resources/text.js:
2936         (animate): Update styling. Manually calculate gradients for the shadow particles.
2937         Reduce the step size for y direction to avoid cutting text off at the margins. Increase
2938         step for x since there will be a little more room.
2939         * Animometer/tests/master/text.html: Add more translations and lay it out in a table.
2940
2941 2016-03-20  Jon Lee  <jonlee@apple.com>
2942
2943         Add a link to show debug data
2944         https://bugs.webkit.org/show_bug.cgi?id=155724
2945
2946         Reviewed by Simon Fraser.
2947
2948         * Animometer/developer.html: Attach the onclick handler to the score.
2949         * Animometer/index.html: Ditto.
2950
2951 2016-03-18  Dean Jackson  <dino@apple.com>
2952
2953         Add a basic WebGL test to Animometer
2954         https://bugs.webkit.org/show_bug.cgi?id=155475
2955         <rdar://problem/25156860>
2956
2957         Reviewed by Simon Fraser and Jon Lee.
2958
2959         Add a pretty simple WebGL test to Animometer
2960         which tries to draw a number of animated triangles.
2961         The result still needs a bit of tuning.
2962
2963         * Animometer/resources/debug-runner/tests.js: Add a new
2964         Suite for "3D" tests.
2965         * Animometer/tests/3d/resources/webgl.js: Added.
2966         * Animometer/tests/3d/webgl.html: Added.
2967
2968 2016-03-18  Jon Lee  <jonlee@apple.com>
2969
2970         Add support for statically linking to a specific test
2971         https://bugs.webkit.org/show_bug.cgi?id=155631
2972
2973         Rubber-stamped by Darin Adler.
2974
2975         * Animometer/developer.html: Update to call restartBenchmark instead so that it works for
2976         both the preset and the manually set test sessions.
2977         * Animometer/resources/debug-runner/animometer.css:
2978         (.tree .link): Style the "link" UI.
2979         * Animometer/resources/debug-runner/animometer.js: Move updateDisplay from suitesManager
2980         to optionsManager since this is a visual option and has nothing to do with the suites
2981         settings. Update the construction of the test UI by including a "link" after each test.
2982         Clicking on that link brings up a JS prompt with a URL and a query string with the current
2983         parameters of the controller, and selected test. Pasting this URL into the location bar will
2984         automatically start running the selected test.
2985         (suitesManager.suitesFromQueryString): Iterate through the Suites and tests and find the one
2986         that matches the provided parameters. Returns an object similar to the form in
2987         suitesManager.updateLocalStorageFromUI.
2988         (benchmarkController.initialize): After settings up the events and options, try parsing the
2989         query string and running the benchmark immediately. Otherwise, fall back to the form.
2990         (benchmarkController.startBenchmark): Store the options and suites into member variables for
2991         reuse in restartBenchmark.
2992         (benchmarkController.startBenchmarkImmediatelyIfEncoded): Convert the query string to an object.
2993         If that's successful, find the suite and test referenced in the query string. Start the benchmark if
2994         the search for the test succeeded.
2995         * Animometer/resources/extensions.js:
2996         (Utilities.stripNonASCIICharacters): Helper method to convert the name of the suite and test
2997         into a query-string-friendly version.
2998         (Utilities.convertObjectToQueryString): Helper method to convert an object to query string
2999         format.
3000         (Utilities.convertQueryStringToObject): Helper method to convert query string into an object
3001         with properties and values.
3002
3003 2016-03-16  Jon Lee  <jonlee@apple.com>
3004
3005         Add a new benchmark test
3006         https://bugs.webkit.org/show_bug.cgi?id=155570
3007
3008         Reviewed by Simon Fraser.
3009
3010         New Leaves test includes various image sizes and opacity.
3011
3012         * Animometer/resources/debug-runner/tests.js: Add it to the HTML test suite.
3013         * Animometer/tests/dom/leaves.html: Added.
3014         * Animometer/tests/dom/resources/leaves.js: Added. Override the
3015         (Particle.call.reset): Uses a range of sizes, and opacity.
3016         (Particle.call.animate): Opacity goes up then down. When it hits 0, reset the particle.
3017         (Particle.call.move): Set transform and opacity.
3018         * Animometer/tests/master/resources/leaves.js: Get rid of the closure so that it
3019         can be used in this test. Update the relative path so that it works in both the master
3020         and dom test suite.
3021
3022 2016-03-15  Simon Fraser  <simon.fraser@apple.com>
3023
3024         Add developer Animometer test that bounces P3-tagged images
3025         https://bugs.webkit.org/show_bug.cgi?id=155511
3026
3027         Reviewed by Tim Horton.
3028         
3029         Add a test for rendering performance of tagged images. The 5 images are tagged
3030         with the Display P3 colorspace.
3031
3032         * Animometer/resources/debug-runner/tests.js:
3033         * Animometer/tests/bouncing-particles/bouncing-tagged-images.html: Added.
3034         * Animometer/tests/bouncing-particles/resources/bouncing-tagged-images.js: Added.
3035         * Animometer/tests/bouncing-particles/resources/image1.jpg: Added.
3036         * Animometer/tests/bouncing-particles/resources/image2.jpg: Added.
3037         * Animometer/tests/bouncing-particles/resources/image3.jpg: Added.
3038         * Animometer/tests/bouncing-particles/resources/image4.jpg: Added.
3039         * Animometer/tests/bouncing-particles/resources/image5.jpg: Added.
3040
3041 2016-03-09  Jon Lee  <jonlee@apple.com>
3042
3043         Enhance existing Animometer tests
3044         https://bugs.webkit.org/show_bug.cgi?id=155261
3045
3046         Reviewed by Simon Fraser.
3047
3048         * Animometer/tests/master/resources/canvas-tests.js: Add a gradient to
3049         the filled circles. To expose more of the gradient, add another circle.
3050         (SimpleCanvasStage.call.animate): For each frame, create a gradient with
3051         undulating stop points and colors. Fill the circles twice; once with
3052         the solid color, and once with the gradient.
3053
3054         * Animometer/tests/master/resources/dom-particles.js: Refactor the
3055         emission variables into a separate stage for this test. Add a colorOffset
3056         variable to make the colors of each particle slightly different, since
3057         the ramp controller can add large numbers of particles all at once, which
3058         would otherwise get all the same color.
3059         * Animometer/tests/master/resources/particles.js:
3060         (initialize): Remove the code specific to the SVG mask test.
3061
3062         * Animometer/tests/master/resources/image-data.js:
3063         (initialize): Remove unused local variable.
3064         * Animometer/tests/master/resources/multiply.js:
3065         (initialize): Make the test harder by adding more total particles.
3066
3067 2016-03-09  Jon Lee  <jonlee@apple.com>
3068
3069         Add text tests
3070         https://bugs.webkit.org/show_bug.cgi?id=155257
3071
3072         Reviewed by Simon Fraser.
3073
3074         * Animometer/resources/extensions.js:
3075         (UnitBezier.Utilities.createClass): Add a class that computes Bezier points
3076         assuming that two of the control points are at (0,0) and (1,1). Taken from
3077         WebCore/platform/graphics/UnitBezier.h
3078         * Animometer/resources/runner/tests.js: Add a test for English text and one
3079         for international text.
3080         * Animometer/tests/master/international.html: Added.
3081         * Animometer/tests/master/resources/text.js: Added. The test assumes there is
3082         #template div which it will copy. The copies are placed behind the template, and
3083         are set with different colors each frame. They are moved around with CSS transform.
3084         * Animometer/tests/master/text.html: Added.
3085
3086         Remove the other text tests, since these ones cover the same techniques.
3087         * Animometer/tests/text/layering-text.html: Removed.
3088         * Animometer/tests/text/resources/layering-text.js: Removed.
3089         * Animometer/tests/text/resources/text-boxes.js: Removed.
3090         * Animometer/tests/text/text-boxes.html: Removed.
3091         * Animometer/resources/debug-runner/tests.js:
3092
3093 2016-03-09  Jon Lee  <jonlee@apple.com>
3094
3095         Add a new image test
3096         https://bugs.webkit.org/show_bug.cgi?id=155232
3097
3098         Reviewed by Dean Jackson.
3099         Provisionally reviewed by Said Abou-Hallawa.
3100
3101         The image test renders PNGs and moves them with translate and rotate
3102         transforms. Each particle has a lifetime, and when the lifetime ends or
3103         the particle goes offscreen, it resets itself somewhere on the stage.
3104
3105         * Animometer/resources/debug-runner/tests.js: Remove the CSS bouncing PNG
3106         images test, because this one tests the same technique.
3107         * Animometer/resources/runner/tests.js: Added here as "Leaves".
3108         * Animometer/tests/master/leaves.html: Added.
3109         * Animometer/tests/master/resources/particles.js: Refactor out the parts
3110         specific to the DOM particles test. Consequently make velocity a public
3111         member.
3112         (initialize): The options parameter is never used, so remove it.
3113         (animate): Remove unused local variable.
3114         * Animometer/tests/master/resources/leaves.js: Maintains a focal point that
3115         moves back and forth across the canvas. That point affects the velocity
3116         of the particles. When the focal point is on the leftmost side, it is
3117         between [-6, -2], and set in reset(). When the focal point is on the rightmost
3118         side, it will be from [2, 6].
3119         * Animometer/tests/master/resources/dom-particles.js: Move JS specific to
3120         this test here from particles.js.
3121
3122         New images.
3123         * Animometer/tests/master/resources/compass100.png: Added.
3124         * Animometer/tests/master/resources/console100.png: Added.
3125         * Animometer/tests/master/resources/contribute100.png: Added.
3126         * Animometer/tests/master/resources/debugger100.png: Added.
3127         * Animometer/tests/master/resources/inspector100.png: Added.
3128         * Animometer/tests/master/resources/layout100.png: Added.
3129         * Animometer/tests/master/resources/performance100.png: Added.
3130         * Animometer/tests/master/resources/script100.png: Added.
3131         * Animometer/tests/master/resources/shortcuts100.png: Added.
3132         * Animometer/tests/master/resources/standards100.png: Added.
3133         * Animometer/tests/master/resources/storage100.png: Added.
3134         * Animometer/tests/master/resources/styles100.png: Added.
3135         * Animometer/tests/master/resources/timeline100.png: Added.
3136
3137 2016-03-03  Jon Lee  <jonlee@apple.com>
3138
3139         Add ability to retrieve raw data from release harness
3140         https://bugs.webkit.org/show_bug.cgi?id=155026
3141
3142         Reviewed by Simon Fraser.
3143
3144         * Animometer/developer.html: Remove the special UI in the debug
3145         harness.
3146         * Animometer/resources/runner/animometer.css: Add styles for the
3147         overlay.
3148         * Animometer/resources/runner/animometer.js: Let 'j' show the JSON
3149         results, but only if the overlay doesn't exist. Add 'esc' key to dismiss
3150         the overlay.
3151         (window.benchmarkController.selectResults): Cycle the cases around
3152         so that the first 's' press selects both the benchmark score and
3153         the individual test scores.
3154         * Animometer/resources/debug-runner/animometer.css: Remove
3155         unneeded rules.
3156         * Animometer/resources/debug-runner/animometer.js: Make the same
3157         call to handleKeyPress.
3158
3159 2016-03-03  Jon Lee  <jonlee@apple.com>
3160
3161         Make sure multiply test particles have at least some opacity
3162         https://bugs.webkit.org/show_bug.cgi?id=155027
3163
3164         Reviewed by Simon Fraser.
3165
3166         * Animometer/tests/master/resources/multiply.js:
3167         (initialize): Have the elements spin a little faster to make it
3168         more obvious when the system is being stressed.
3169         (_addTile):
3170         (animate): Make sure each element being animated has at least 1%
3171         opacity.
3172
3173 2016-03-03  Jon Lee  <jonlee@apple.com>
3174
3175         Update image test
3176         https://bugs.webkit.org/show_bug.cgi?id=154962
3177
3178         Rubber-stamped by Darin Adler.
3179
3180         * Animometer/tests/master/image-data.html: Make each canvas have a compositing layer.
3181         * Animometer/tests/master/resources/image-data.js:
3182         (initialize): Update to load multiple images serially.
3183         (_loadImage):
3184         (tune): Use display instead of visibility. When showing an element anew, move it somewhere else.
3185         (_createTestElement): Set up the element, refactor out placement of element to _refreshElement.
3186         (_refreshElement): Place element in a tile grid.
3187         (animate):
3188         (_getRandomNeighboringPixelIndex): Use a more conservative distribution to make the effect last a little longer.
3189         * Animometer/tests/master/resources/compass.svg: Added.
3190         * Animometer/tests/master/resources/console.svg: Added.
3191         * Animometer/tests/master/resources/contribute.svg: Added.
3192         * Animometer/tests/master/resources/debugger.svg: Added.
3193         * Animometer/tests/master/resources/inspector.svg: Added.
3194         * Animometer/tests/master/resources/layout.svg: Added.
3195         * Animometer/tests/master/resources/performance.svg: Added.
3196         * Animometer/tests/master/resources/script.svg: Added.
3197         * Animometer/tests/master/resources/shortcuts.svg: Added.
3198         * Animometer/tests/master/resources/standards.svg: Added.
3199         * Animometer/tests/master/resources/storage.svg: Added.
3200         * Animometer/tests/master/resources/styles.svg: Added.
3201         * Animometer/tests/master/resources/timeline.svg: Added.
3202
3203 2016-03-02  Jon Lee  <jonlee@apple.com>
3204
3205         Add some new controllers, and refine tests
3206         https://bugs.webkit.org/show_bug.cgi?id=154914
3207
3208         Reviewed by Simon Fraser.
3209
3210         Improve tests.
3211
3212         * Animometer/tests/master/focus.html: Move each particle into a layer. Put the particle
3213         in a container element. Expand the container element by the max blur radius, and clip
3214         overflow. This way, when the blur is applied, it avoids causing layer resizes.
3215         * Animometer/tests/master/resources/focus.js:
3216         (FocusElement.Utilities.createClass): Create a container element, and put the particle
3217         inside.
3218         (hide): Set display:none.
3219         (show): Set display:block.
3220         (animate): Apply filters to the container element.
3221         (FocusStage.call.initialize): Instead of inserting and removing elements from the DOM,
3222         keep them in the stage, but set the display style instead. Use this._offsetIndex to
3223         keep track of which elements are displayed.
3224         (FocusStage.call.animate): Move some calculations around to avoid doing unneeded math.
3225         (FocusStage.call.getBlurValue): Make sure elements are always blurred.
3226         (FocusStage.call.getOpacityValue): Make sure elements have some opacity.
3227         * Animometer/tests/master/resources/multiply.js: Have the particles on the edge of the
3228         stage appear less black when the complexity gets large enough.
3229         (tune): Have this._distanceFactor calculate the factor. Avoid calculating square root each
3230         frame.
3231
3232 2016-03-02  Jon Lee  <jonlee@apple.com>
3233
3234         Add some new controllers, and refine tests
3235         https://bugs.webkit.org/show_bug.cgi?id=154914
3236
3237         Reviewed by Simon Fraser.
3238
3239         Add a controller that centers around 30 fps instead of 60 fps.
3240
3241         * Animometer/developer.html: Add a new option.
3242         * Animometer/resources/debug-runner/animometer.js:
3243         * Animometer/resources/runner/animometer.js:
3244         (this._processData.findRegression): When calculating the complexity-frameLength regression,
3245         check the controller, and use a 30 fps baseline if needed.
3246         * Animometer/resources/statistics.js:
3247         (Regression.Utilities.createClass): Update to allow clients to specify the baseline
3248         frame length.
3249         * Animometer/tests/resources/main.js:
3250         (tune): Override some of the constants in RampController. Move those constants out for
3251         easier reading.
3252
3253         Add a fixed controller, with no step.
3254
3255         * Animometer/developer.html: Add a controller that takes no step.
3256         * Animometer/resources/debug-runner/animometer.js:
3257         * Animometer/tests/resources/main.js:
3258         (Rotater.Utilities.createClass):
3259
3260         Switch to ramp controller as default.
3261
3262         * Animometer/developer.html: Increase the test length to 20 seconds.
3263         * Animometer/resources/debug-runner/animometer.js: Rename the "adjustment" field to "controller"
3264         since that is a more accurate description.
3265         * Animometer/resources/debug-runner/graph.js:
3266         * Animometer/resources/runner/animometer.js: Update preferences for release suite.
3267
3268 2016-02-26  Jon Lee  <jonlee@apple.com>
3269
3270         Address Dean's comments in 154673.
3271
3272         * Animometer/developer.html:
3273         * Animometer/resources/debug-runner/animometer.js:
3274         * Animometer/resources/debug-runner/graph.js:
3275
3276 2016-02-25  Jon Lee  <jonlee@apple.com>
3277
3278         Update animation benchmark and tests
3279         https://bugs.webkit.org/show_bug.cgi?id=154673
3280
3281         Reviewed by Dean Jackson.
3282
3283         Update test visuals.
3284
3285         * Animometer/tests/master/resources/canvas-tests.js:
3286         (CanvasLineSegment.Utilities.createClass): Line segments near the edge of the stage get clipped.
3287         Update circle position and radius to minimize impact.
3288         * Animometer/tests/master/resources/particles.js: Get rid of rotating gradient background, and
3289         have 3 locations for emitting particles.
3290         * Animometer/tests/master/resources/multiply.js: Update distance metric so that fringe tiles get
3291         more color.
3292         (initialize): Fix some of the math for laying out the tiles.
3293         (animate): When a tile isn't used, set visibility: hidden.
3294
3295 2016-02-24  Jon Lee  <jonlee@apple.com>
3296
3297         Update animation benchmark and tests
3298         https://bugs.webkit.org/show_bug.cgi?id=154673
3299
3300         Reviewed by Dean Jackson.
3301
3302         Update the ramp controller.
3303
3304         The controller refines the complexity interval to test across.
3305
3306         * Animometer/resources/statistics.js: Add functions that estimate cumulative distribution function.
3307         (Regression): For the flat regression, force the first segment to be at 60 fps.
3308         (valueAt): Add convenience function to return interpolated value based on the regression used.
3309         (_calculateRegression): Include the number of points included for both segments, and the piecewise
3310         errors.
3311         * Animometer/tests/resources/math.js: Make the Kalman estimator subclass Experiment, and allow it
3312         to be reset.
3313
3314         * Animometer/tests/resources/main.js: Initialize the tier such that it starts at 10^0 = 1.
3315         Increase the number of ramps. Maintain three FPS thresholds-- the frame rate of interest, a limit
3316         on the lowest FPS we care to go for later interpolation, and a minimum FPS threshold we want to
3317         aim for each ramp. Also keep three estimators: a running average of the change point, a minimum
3318         boundary for each ramp, and an estimator for all the frames within an interval. The first two
3319         are used to determine the parameters of the next ramp, and the latter allows us to refine the
3320         parameters.
3321         (update): During the tier phase, it is possible that the highest complexity possible for a test
3322         won't stress the system enough to trigger stopping the tier phase and transitioning to the ramps.
3323         If the complexity doesn't change when going to the next tier, we've maxed the test out, and move
3324         on. When the tier phase completed, turn off Controller.frameLengthEstimator, which estimates the
3325         FPS at each tier.
3326         (tune): At each interval, look at the confidence distribution of being on the 60 FPS side or the
3327         slow side. If the slowest FPS we achieve at the ramp's maximum complexity is not at least
3328         _fpsRampSlowThreshold, then increase the maximum complexity. If we ever achieve 60 FPS, increase
3329         the ramp's minimum complexity to that level. If, at an even lower complexity, a glitch causes the
3330         FPS to drop, we reset the minimum complexity.
3331
3332         Have the bootstrap calculation occur between tests. Clean up harness.
3333
3334         * Animometer/resources/debug-runner/animometer.js: Run bootstrap after a test has
3335         completed to avoid doing all of it at the end before showing the results. Clean up
3336         parameters being passed around.
3337         * Animometer/resources/debug-runner/tests.js:
3338         (text):
3339         * Animometer/resources/runner/animometer.js:
3340         (this._processData.calculateScore): Save the results to the same object holding the data.
3341         (this._processData._processData): In the case where a file is dragged, calculate the score
3342         serially. Grab the results object and move it to the results variable and remove it from
3343         the data object. This avoids serializing the results into the JSON.
3344         (this._processData.findRegression): Include the samples used for bootstrapping. Reduce the
3345         resample size to shorten the wait.
3346         * Animometer/resources/runner/benchmark-runner.js:
3347         * Animometer/resources/statistics.js:
3348         (bootstrap): Update how bootstrapData is sorted. In some regression results the mix of
3349         floats and integers causes an alphabetical sort to occur.
3350         * Animometer/resources/strings.js:
3351
3352         Add meta charset so that encodings between harness and test match.
3353
3354         * Animometer/tests/bouncing-particles/bouncing-canvas-images.html:
3355         * Animometer/tests/bouncing-particles/bouncing-canvas-shapes.html:
3356         * Animometer/tests/bouncing-particles/bouncing-css-images.html:
3357         * Animometer/tests/bouncing-particles/bouncing-css-shapes.html:
3358         * Animometer/tests/bouncing-particles/bouncing-svg-images.html:
3359         * Animometer/tests/bouncing-particles/bouncing-svg-shapes.html:
3360         * Animometer/tests/master/canvas-stage.html:
3361         * Animometer/tests/master/focus.html:
3362         * Animometer/tests/master/image-data.html:
3363         * Animometer/tests/master/multiply.html:
3364         * Animometer/tests/master/particles.html:
3365         * Animometer/tests/misc/canvas-electrons.html:
3366         * Animometer/tests/misc/canvas-stars.html:
3367         * Animometer/tests/misc/compositing-transforms.html:
3368         * Animometer/tests/simple/simple-canvas-paths.html:
3369         * Animometer/tests/simple/tiled-canvas-image.html:
3370         * Animometer/tests/template/template-canvas.html:
3371         * Animometer/tests/template/template-css.html:
3372         * Animometer/tests/template/template-svg.html:
3373         * Animometer/tests/text/layering-text.html:
3374         * Animometer/tests/text/text-boxes.html:
3375
3376         Update test harness reporting.
3377
3378         * Animometer/developer.html: Add missing meta charset.
3379         * Animometer/index.html: Remove unnecessary utf-8 declaration.
3380         * Animometer/resources/debug-runner/animometer.css: Add convenience classes for
3381         formatting the results table.
3382         * Animometer/resources/debug-runner/animometer.js: Adjust which stats are shown.
3383         * Animometer/resources/debug-runner/tests.js: Display bootstrapping statistics.
3384         * Animometer/resources/strings.js: Move strings not used by the release harness.
3385
3386         Switch to a pseudo-random number generator.
3387
3388         * Animometer/resources/statistics.js: Add a Pseudo class, with a simple
3389         pseudo-random number generator.
3390         (_calculateRegression): Reset the generator before running bootstrap.
3391         (bootstrap): Deleted.
3392
3393         Replace Math.random with Pseudo.random.
3394         * Animometer/tests/master/resources/canvas-tests.js:
3395         * Animometer/tests/master/resources/focus.js:
3396         * Animometer/tests/master/resources/particles.js:
3397         * Animometer/tests/resources/main.js:
3398
3399         Use bootstrapping to get confidence interval in the breakpoint.
3400
3401         For the ramp controller, calculate the piecewise regression, and then use
3402         bootstrapping in order to find the 95% confidence interval. Use the raw data.
3403
3404         * Animometer/developer.html: Default to the complexity graph. Add a legend
3405         checkbox to toggle visibility of the bootstrap score and histogram.
3406         * Animometer/resources/debug-runner/animometer.css: Make some more space to show
3407         the old raw and average scores in the legend. Add new styles for the data.
3408         * Animometer/resources/debug-runner/graph.js:
3409         (_addRegressionLine): Allow passing an array for the variance bar tied to the
3410         regression line. Now |stdev| is |range|.
3411         (createComplexityGraph): Add bootstrap median, and overlay a histogram of
3412         the bootstrap samples. Switch raw samples from circles to X's.
3413         (onComplexityGraphOptionsChanged): Allow toggling of the bootstrap data.
3414         (onGraphTypeChanged): Move the regressions for the raw and average samples to the
3415         legend. In the subtitle use the bootstrap median, and include the 95% confidence
3416         interval.
3417         * Animometer/resources/runner/animometer.js:
3418         (this._processData.findRegression): Factor out the code that determines which
3419         samples to include when calculating the piecewise regression. For series that have
3420         many samples, or a wider range of recorded complexities, throw away the 2.5%
3421         lowest and highest samples before calculating the regression. Keep all samples
3422         if the number of samples to regress is small or the range of complexities is
3423         narrow.
3424         (this._processData._calculateScore): Factor out regression calculation to
3425         findRegression(). Bootstrap the change point of the regression. The score is the
3426         median.
3427         * Animometer/resources/statistics.js:
3428         (_calculateRegression): Correct an issue in the calculation of the regression, where
3429         the denominator can be 0.
3430         (bootstrap): Template for bootstrapping. Create a bootstrap sample array, Create
3431         re-samples by random selection with replacement. Return the 95% confidence samples,
3432         the bootstrap median, mean, and the data itself.
3433         * Animometer/resources/strings.js: Add bootstrap.
3434         * Animometer/tests/resources/main.js:
3435         (processSamples): Don't prematurely cut the sample data.
3436
3437         Fix graph drawing.
3438
3439         * Animometer/resources/debug-runner/animometer.js: Add spacing in the JSON output.
3440         Multiple tests output a lot of JSON and can hang when selecting JSON with no whitespace.
3441         * Animometer/resources/debug-runner/animometer.css:
3442         (#complexity-graph .series.raw circle): Update the color.
3443         * Animometer/resources/debug-runner/graph.js: Use the FPS axis instead of the
3444         complexity axis, which can vary in domain. For determining the complexity domain,
3445         only use samples after samplingTimeOffset.
3446
3447         Allow dropping results JSON.
3448
3449         * Animometer/developer.html: Add a button.
3450         * Animometer/resources/debug-runner/animometer.css:
3451         * Animometer/resources/debug-runner/animometer.js: Read the data and go straight
3452         to the dashboard. With JSON output, write out only the options and the raw data.
3453
3454         Teach the harness to evaluate the samples and determine the test score.
3455
3456         This will allow us to update how the score is calculated separately from the samples recorded.
3457         This also prepares the harness to be able to accept JSON of prior runs.
3458
3459         * Animometer/resources/strings.js: Clean up and remove unneeded strings and reduce some of the
3460         hierarchy.
3461         * Animometer/resources/debug-runner/tests.js: Update to use the new strings.
3462
3463         * Animometer/tests/resources/main.js: Allow all controllers to show a complexity-FPS graph.
3464         (_processComplexitySamples): Factor out some of the sample processing done in the ramp
3465         controller for the benefit of the other controllers. |complexitySamples| contains a list of
3466         samples. Sort the samples by complexity. Optionally remove the top x% of samples.
3467         Group them, and calculate distribution of samples within the same complexity, and add those as
3468         new entries into |complexityAverageSamples|.
3469         (Controller.processSamples): Move the code responsible for determining the complexity and FPS
3470         scores out to ResultsDashboard. The structure of the data returned by the controller is:
3471
3472         {
3473             controller: [time-regression, time-regression, ...], // optional, data specific to controller
3474             marks: [...],
3475             samples: {                    // all of the sample data
3476                 controller: [...],
3477                 complexity: [...],        // processed from controller samples
3478                 complexityAverage: [...], // processed from complexity samples
3479             }
3480         }
3481
3482         (AdaptiveController.processSamples): Adding the target frame length is no longer necessary; we
3483         now pass the test options to the graph.
3484         (Regression): Move to statistics.js.
3485         * Animometer/resources/statistics.js: Move Regression to here. Add a check if the sampling range
3486         only contains one sample, since we cannot calculate a regression from one sample point.
3487
3488         Teach the test harness to evaluate the data.
3489         * Animometer/resources/runner/animometer.js:
3490         (ResultsDashboard): Store the options used to run the test and the computed results/score separately
3491         from the data. The results are stored as:
3492
3493         {
3494             score: /* geomean of iteration score */,
3495             iterationsResults: [
3496                 {
3497                     score: /* geomean of tests */,
3498                     testsResults: {
3499                         suiteName: {
3500                             testName: {
3501                                 controller: {
3502                                     average:
3503                                     concern:
3504                                     stdev:
3505                                     percent:
3506                                 },
3507                                 frameLength: { ... },
3508                                 complexity: {
3509                                     complexity:
3510                                     stdev:
3511                                     segment1:
3512                                     segment2:
3513                                 },
3514                                 complexityAverage: { ... }
3515                             },
3516                             testName: { ... },
3517                         },
3518                         ... next suite ...
3519                     }
3520                 },
3521                 { ...next iteration... }
3522             ]
3523         }
3524
3525         * Animometer/resources/debug-runner/animometer.js: Pass options around instead of relying
3526         on what was selected in the form. This will later allow for dropping previous results, and
3527         using those runs' options when calculating scores.
3528         (ResultsTable._addGraphButton): Simplify button action by using attached test data.
3529         * Animometer/resources/debug-runner/graph.js: Refactor to use the data.
3530
3531         Consolidate JS files, and move statistics out to a separate JS.
3532
3533         Preparation for having the Controller only handle recording and storage of the samples,
3534         and leave the evaluation of the test score out to the harness. Move Experiment to
3535         a new statistics.js, where Regression will also eventually go. Get rid of algorithm.js
3536         and move it to utilities.js since the Heap is used only for Experiments.
3537
3538         * Animometer/tests/resources/algorithm.js: Removed. Heap is in utilities.js.
3539         * Animometer/tests/resources/sampler.js: Removed. Experiment is in statistics.js,
3540         Sampler in main.js.
3541         * Animometer/tests/resources/main.js: Move Sampler here.
3542         * Animometer/resources/statistics.js: Added. Move Statistics and Experiment here.
3543         * Animometer/resources/extensions.js: Move Heap here. Attach static method to create
3544         a max or min heap to Heap, instead of a new Algorithm object.
3545
3546         Update JS files.
3547         * Animometer/developer.html:
3548         * Animometer/index.html:
3549         * Animometer/tests/bouncing-particles/bouncing-canvas-images.html:
3550         * Animometer/tests/bouncing-particles/bouncing-canvas-shapes.html:
3551         * Animometer/tests/bouncing-particles/bouncing-css-images.html:
3552         * Animometer/tests/bouncing-particles/bouncing-css-shapes.html:
3553         * Animometer/tests/bouncing-particles/bouncing-svg-images.html:
3554         * Animometer/tests/bouncing-particles/bouncing-svg-shapes.html:
3555         * Animometer/tests/master/canvas-stage.html:
3556         * Animometer/tests/master/focus.html:
3557         * Animometer/tests/master/image-data.html:
3558         * Animometer/tests/master/multiply.html:
3559         * Animometer/tests/master/particles.html:
3560         * Animometer/tests/misc/canvas-electrons.html:
3561         * Animometer/tests/misc/canvas-stars.html:
3562         * Animometer/tests/misc/compositing-transforms.html:
3563         * Animometer/tests/simple/simple-canvas-paths.html:
3564         * Animometer/tests/simple/tiled-canvas-image.html:
3565         * Animometer/tests/template/template-canvas.html:
3566         * Animometer/tests/template/template-css.html:
3567         * Animometer/tests/template/template-svg.html:
3568         * Animometer/tests/text/layering-text.html:
3569         * Animometer/tests/text/text-boxes.html:
3570
3571         Fix the cursor in the graph analysis when the min
3572         complexity is not 0.
3573
3574         * Animometer/resources/debug-runner/graph.js:
3575         (_addRegression):
3576         (createComplexityGraph):
3577
3578 2016-02-23  Geoffrey Garen  <ggaren@apple.com>
3579
3580         Fix some issues in MallocBench
3581         https://bugs.webkit.org/show_bug.cgi?id=154600
3582
3583         Reviewed by Oliver Hunt.
3584
3585         * MallocBench/MallocBench.xcodeproj/project.pbxproj: Make nimlang.ops
3586         a part of the build so that it copies to the right place for execution.
3587
3588         * MallocBench/MallocBench/Interpreter.cpp:
3589         (Interpreter::Interpreter): Don't try to open .ops files for writing
3590         because we're only going to read and we might need extra permissions to
3591         write.
3592
3593         (Interpreter::~Interpreter): Give more context when opening a file fails
3594         to help with debugging.
3595
3596         * MallocBench/MallocBench/stress.cpp:
3597         (benchmark_stress): Reduce iterations to complete in less than 10 seconds.
3598
3599 2016-02-23  Michael Saboff  <msaboff@apple.com>
3600
3601         Unreviewed change to revert extraneous changes made part of change set 196955.
3602
3603         * MallocBench/MallocBench/Interpreter.cpp:
3604         (Interpreter::doMallocOp):
3605         (Interpreter::Thread::switchTo):
3606         (writeData): Deleted.
3607
3608 2016-02-19  Michael Saboff  <msaboff@apple.com>
3609
3610         MallocBench: Added recording for nimlang website, new recording details and added new options
3611         https://bugs.webkit.org/show_bug.cgi?id=154485
3612
3613         Reviewed by Geoff Garen.
3614
3615         Added new capabilities to MallocBench.  These include:
3616             Added a recording of http://nim-lang.org/docs/lib.html.
3617             Added thread id to the recording and the ability to playback switching threads in MallocBench
3618             Added aligned allocations to recordings and the ability to playback
3619             Added --use-thread-id option to honor recorded thread ids
3620             Added --detailed-report to output remaining allocations by size after playback
3621             Added --no-warmup to not run the warm up iteration
3622
3623         Changed the way that options are passed down to the benchmarks.  Instead of passing individual
3624         boolean or numeric option values, just pass a reference the CommandLine itself.  Each benchmark
3625         can access the options that are appropriate.  The Benchmark class also uses the options for
3626         is parallel, run counts and warm up.
3627
3628         Added thread id and aligned malloc to the Op by noticing that structure padding and Opcode allowed
3629         for another 32 bits of data.  Breaking that unused 32 bits into a 16 bit thread id value and a
3630         16 bit log base 2 of the alignment for aligned malloc allowed for existing recordings to playback
3631         without any incompatibilities.
3632
3633         Threaded operation is simulated by creating threads as needed.  As long as the next Op's thread id
3634         is the same as the last, operation continues as normal.  When the next Op has a different thread id,
3635         we switch to that thread using the shared Op stream to continue playing back.  There is a mutex to
3636         assure that only one thread is really running at a time and a condition variable used to wait
3637         that the current thread id matches each block thread's thread id.  This doesn't simulate true
3638         concurrent threading, but is instead plays back Ops recorded for multiple thread faithfully.
3639
3640         * MallocBench/MallocBench.xcodeproj/project.pbxproj:
3641         * MallocBench/MallocBench/Benchmark.cpp:
3642         (deallocateHeap):
3643         (Benchmark::Benchmark):
3644         (Benchmark::runOnce):
3645         (Benchmark::run):
3646         * MallocBench/MallocBench/Benchmark.h:
3647         (Benchmark::isValid):
3648         * MallocBench/MallocBench/CommandLine.cpp:
3649         (CommandLine::printUsage):
3650         * MallocBench/MallocBench/CommandLine.h:
3651         (CommandLine::isValid):
3652         (CommandLine::benchmarkName):
3653         (CommandLine::isParallel):
3654         (CommandLine::useThreadID):
3655         (CommandLine::detailedReport):
3656         (CommandLine::warmUp):
3657         (CommandLine::heapSize):
3658         (CommandLine::runs):
3659         * MallocBench/MallocBench/Interpreter.cpp:
3660         (Interpreter::Interpreter):
3661         (Interpreter::run):
3662         (Interpreter::readOps):
3663         (Interpreter::doOnSameThread):
3664         (Interpreter::switchToThread):
3665         (Interpreter::detailedReport):
3666         (compute2toPower):
3667         (writeData):
3668         (Interpreter::doMallocOp):
3669         (Interpreter::Thread::Thread):
3670         (Interpreter::Thread::stop):
3671         (Interpreter::Thread::~Thread):
3672         (Interpreter::Thread::runThread):
3673         (Interpreter::Thread::waitToRun):
3674         (Interpreter::Thread::switchTo):
3675         * MallocBench/MallocBench/Interpreter.h:
3676         (Interpreter::Thread::isMainThread):
3677         * MallocBench/MallocBench/alloc_free.cpp: Added.
3678         (benchmark_alloc_free):
3679         * MallocBench/MallocBench/alloc_free.h: Added.
3680         * MallocBench/MallocBench/balloon.cpp:
3681         (benchmark_balloon):
3682         * MallocBench/MallocBench/balloon.h:
3683         * MallocBench/MallocBench/big.cpp:
3684         (benchmark_big):
3685         * MallocBench/MallocBench/big.h:
3686         * MallocBench/MallocBench/churn.cpp:
3687         (benchmark_churn):
3688         * MallocBench/MallocBench/churn.h:
3689         * MallocBench/MallocBench/facebook.cpp:
3690         (benchmark_facebook):
3691         * MallocBench/MallocBench/facebook.h:
3692         * MallocBench/MallocBench/flickr.cpp:
3693         (benchmark_flickr):
3694         (benchmark_flickr_memory_warning):
3695         * MallocBench/MallocBench/flickr.h:
3696         * MallocBench/MallocBench/fragment.cpp:
3697         (validate):
3698         (benchmark_fragment):
3699         (benchmark_fragment_iterate):
3700         * MallocBench/MallocBench/fragment.h:
3701         * MallocBench/MallocBench/list.cpp:
3702         (benchmark_list_allocate):
3703         (benchmark_list_traverse):
3704         * MallocBench/MallocBench/list.h:
3705         * MallocBench/MallocBench/main.cpp:
3706         (main):
3707         * MallocBench/MallocBench/medium.cpp:
3708         (benchmark_medium):
3709         * MallocBench/MallocBench/medium.h:
3710         * MallocBench/MallocBench/memalign.cpp:
3711         (test):
3712         (benchmark_memalign):
3713         * MallocBench/MallocBench/memalign.h:
3714         * MallocBench/MallocBench/message.cpp:
3715         (benchmark_message_one):
3716         (benchmark_message_many):
3717         * MallocBench/MallocBench/message.h:
3718         * MallocBench/MallocBench/nimlang.cpp: Added.
3719         (benchmark_nimlang):
3720         * MallocBench/MallocBench/nimlang.h: Added.
3721         * MallocBench/MallocBench/nimlang.ops: Added.
3722         * MallocBench/MallocBench/realloc.cpp:
3723         (benchmark_realloc):
3724         * MallocBench/MallocBench/realloc.h:
3725         * MallocBench/MallocBench/reddit.cpp:
3726         (benchmark_reddit):
3727         (benchmark_reddit_memory_warning):
3728         * MallocBench/MallocBench/reddit.h:
3729         * MallocBench/MallocBench/stress.cpp:
3730         (deallocate):
3731         (benchmark_stress):
3732         * MallocBench/MallocBench/stress.h:
3733         * MallocBench/MallocBench/stress_aligned.cpp:
3734         (benchmark_stress_aligned):
3735         * MallocBench/MallocBench/stress_aligned.h:
3736         * MallocBench/MallocBench/theverge.cpp:
3737         (benchmark_theverge):
3738         (benchmark_theverge_memory_warning):
3739         * MallocBench/MallocBench/theverge.h:
3740         * MallocBench/MallocBench/tree.cpp:
3741         (benchmark_tree_allocate):
3742         (benchmark_tree_traverse):
3743         (benchmark_tree_churn):
3744         * MallocBench/MallocBench/tree.h:
3745         * MallocBench/run-malloc-benchmarks:
3746
3747 2016-02-11  Jon Lee  <jonlee@apple.com>
3748
3749         Fix a missing refactoring.
3750
3751         * Animometer/tests/master/resources/multiply.js:
3752         (animate): Move to Utilities.lerp.
3753         (_lerp): Deleted.
3754
3755 2016-02-10  Jon Lee  <jonlee@apple.com>
3756
3757         Add new benchmark tests.
3758         https://bugs.webkit.org/show_bug.cgi?id=154063
3759
3760         Provisionally reviewed by Said Abou-Hallawa.
3761
3762         Add tests for get/put image data, filters, opacity, and css transforms.
3763
3764         * Animometer/resources/runner/benchmark-runner.js:
3765         (_runBenchmarkAndRecordResults): Update the body background color to match that of
3766         the stage.
3767         (this._runNextIteration): Clear the background color style for the results page.
3768         * Animometer/resources/runner/tests.js:
3769         * Animometer/tests/master/focus.html: Added.
3770         * Animometer/tests/master/image-data.html: Added.
3771         * Animometer/tests/master/multiply.html: Added.
3772         * Animometer/tests/master/resources/focus.js: Added.
3773         * Animometer/tests/master/resources/image-data.js: Added.
3774         * Animometer/tests/master/resources/multiply.js: Added.
3775         * Animometer/tests/master/resources/stage.css: Move common styles out.
3776         * Animometer/tests/resources/main.js: Update Stage.randomBool to use Math.random.
3777         Add Stage.randomSign for randomly setting a direction. Add the notion of the
3778         current timestamp of the test to Benchmark, since some animations cycle through
3779         colors and rely on an incremental counter like the time.
3780
3781 2016-02-09  Said Abou-Hallawa  <sabouhallawa@apple.com>
3782
3783         Add internal benchmark tests for CSS mix-blend-modes and filters
3784         https://bugs.webkit.org/show_bug.cgi?id=154058
3785
3786         Provisionally reviewed by Jon Lee.
3787
3788         * Animometer/resources/debug-runner/tests.js: Include the new tests in the
3789         "HTML suite" of the debug runner.
3790
3791         * Animometer/resources/extensions.js:
3792         (Utilities.browserPrefix):
3793         (Utilities.setElementPrefixedProperty): Utility functions to allow setting
3794         prefixed style properties.
3795
3796         * Animometer/tests/bouncing-particles/resources/bouncing-css-shapes.js:
3797         Set the mix-blend-mode and the filter to some random values if the options
3798         of the test requested that.
3799
3800         * Animometer/tests/bouncing-particles/resources/bouncing-particles.js:
3801         (parseShapeParameters): Parse the url options "blend" and "filter" and set
3802         the corresponding flags.
3803
3804         * Animometer/tests/resources/main.js:
3805         (randomStyleMixBlendMode):
3806         (randomStyleFilter): Return random mix-blend-mode and filter.
3807
3808 2016-02-08  Jon Lee  <jonlee@apple.com>
3809
3810         Add a ramp controller
3811         https://bugs.webkit.org/show_bug.cgi?id=154028
3812
3813         Provisionally reviewed by Said Abou-Hallawa.
3814
3815         Enhance the graph to include a complexity-fps graph, in addition
3816         to the time graph.
3817
3818         * Animometer/developer.html: Add a ramp option.
3819         * Animometer/resources/debug-runner/animometer.css: Update the style.
3820         * Animometer/resources/strings.js: Flatten the Strings.text constants.
3821         * Animometer/resources/debug-runner/animometer.js:
3822         (ResultsTable.call._addGraphButton): Refactor.
3823         (ResultsTable.call._addTest): Add regression data.
3824         (benchmarkController): Add a form that allows the user to switch between the two forms,
3825         Add a form that allows the user to toggle different data. Hide certain header columns
3826         depending on the selected controller.
3827         * Animometer/resources/debug-runner/graph.js: Add the complexity regressions.
3828         * Animometer/resources/debug-runner/tests.js: Add headers for the ramp results.
3829         * Animometer/resources/runner/animometer.js:
3830         (ResultsTable): If a header is disabled don't include them in _flattenedHeaders.
3831         * Animometer/tests/resources/main.js:
3832         (Controller): Allow options to specify the capacity for sample arrays.
3833         (Regression): A piecewise regression that tries to fit a slope and a flat profile.
3834         (_calculateRegression): Options can fix the slope and bias when calculating the minimal
3835         error. Sweep across the samples in time (which could be backward depending on the controller)
3836         and calculate the intersection point.
3837         (RampController): This controller assumes that the target frame rate is below
3838         58 FPS. It runs in two stages. The first stage quickly determines the order of
3839         magnitude of objects needed to stress the system by the setting the complexity
3840         to increasingly difficult tiers. Perform a series of ramps descending from a
3841         high-water mark of complexity. The complexity needed to reach the target frame
3842         length is done by performing a piecewise regression on each ramp, and track a
3843         running average of these values. For the next ramp, make that running average
3844         the center of the ramp. With a minimum complexity of 0, the high-water mark is
3845         twice that average. The score is based on the highest complexity that can
3846         reach 60 fps.
3847
3848 2016-02-08  Jon Lee  <jonlee@apple.com>
3849
3850         Address Said's comments on the benchmark, and do some clean up.
3851
3852         * Animometer/developer.html:
3853         * Animometer/resources/debug-runner/animometer.css: Add styles for averages.
3854         * Animometer/resources/debug-runner/animometer.js: Use the right
3855         Strings constants.
3856         * Animometer/resources/debug-runner/graph.js:
3857         (_addRegressionLine): Add missing code to draw the line and standard
3858         deviation highlight.
3859         (onGraphTypeChanged): Remove unneeded variables
3860         (onTimeGraphOptionsChanged):
3861         * Animometer/resources/runner/benchmark-runner.js:
3862         (_runBenchmarkAndRecordResults): Rename samplers to suiteResults and
3863         _suitesSamplers to _suitesResults.
3864         * Animometer/tests/resources/main.js:
3865         (results): Call processSamples().
3866         (update): Change sampling timestamp comparison.
3867         (_animateLoop): Move shouldStop call to before the update.
3868         * Animometer/tests/resources/sampler.js:
3869         (process): Rename to processSamples().
3870
3871 2016-02-07  Jon Lee  <jonlee@apple.com>
3872
3873         Teach Controller to measure intervals, and turn off the frame length estimator.
3874
3875         * Animometer/tests/resources/main.js: Default interval length is 100 ms.
3876         (start): Set the first interval.
3877         (_measureAndResetInterval): Reports the average frame length of the interval that just
3878         completed, and sets up the next interval.
3879         (update): If there is no length, then just use the estimator per frame, otherwise the
3880         estimator measures per interval. Add a didFinishInterval for subclasses to process
3881         prior to recording the sample. Update tune() to include whether an interval had
3882         finished.
3883         (StepController): Step controllers don't measure on an interval basis.
3884
3885 2016-02-07  Jon Lee  <jonlee@apple.com>
3886
3887         Minor refactoring. Rename Controller._estimator to Controller._frameLengthEstimator
3888         and switch the parameters for start(), update(), and tune(), so that the timestamp
3889         is first and stage is second.