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