[Arabic text] Add whitespace preserve/word wrap performance test.
[WebKit-https.git] / PerformanceTests / ChangeLog
1 2017-03-29  Zalan Bujtas <zalan@apple.com>
2
3         [Arabic text] Add whitespace preserve/word wrap performance test.
4         https://bugs.webkit.org/show_bug.cgi?id=170243
5
6         Reviewed by Simon Fraser.
7
8         * Layout/ArabicLineLayout-pre-and-wrap.html: Added.
9
10 2017-03-08  Yusuke Suzuki  <utatane.tea@gmail.com>
11
12         [JSC] Add helpers to run Six Speed to run-jsc-benchmarks
13         https://bugs.webkit.org/show_bug.cgi?id=169351
14
15         Reviewed by Saam Barati.
16
17         * SixSpeed/wrappers/wrapper.js: Added.
18         (assertEqual):
19         (test):
20         (jscRun):
21
22 2017-03-08  Jon Davis  <jond@apple.com>
23
24         Fixed ARES-6 animations for other browsers
25
26         Rubber stamped by Saam Barati.
27
28         * ARES-6/styles.css:
29         (.swoop):
30         (.logo):
31
32 2017-03-07  Filip Pizlo  <fpizlo@apple.com>
33
34         Revise the ARES-6 explainer text
35         https://bugs.webkit.org/show_bug.cgi?id=169287
36
37         Reviewed by Saam Barati.
38         
39         Just did some copy-editing.
40
41         * ARES-6/about.html:
42
43 2017-03-06  Jon Davis  <jond@apple.com>
44
45         Refined the ARES-6 experience
46         https://bugs.webkit.org/show_bug.cgi?id=169205
47
48         Reviewed by Saam Barati.
49
50         * ARES-6/about.html: Copied from PerformanceTests/ARES-6/index.html.
51         * ARES-6/driver.js:
52         (Driver.prototype.start):
53         (Driver.prototype._updateIterations):
54         (Driver):
55         * ARES-6/index.html:
56         * ARES-6/stats.js:
57         (Stats.prototype.toString.span.span):
58         * ARES-6/styles.css:
59         (body):
60         (.swoop):
61         (p):
62         (.logo):
63         (.start):
64         (.test):
65         (.test:nth-child(odd)):
66         (.overall):
67         (.about h2):
68         (.button):
69         (.button:hover):
70         (.button.return:before):
71         (@keyframes fade-in):
72         (@keyframes scale-in):
73         (@keyframes flicker-in):
74         (@keyframes wipe-in):
75         (.flip): Deleted.
76         (.testing, .about): Deleted.
77         (.testing): Deleted.
78         (.about): Deleted.
79         (#about:target): Deleted.
80
81 2017-03-06  Jeremy Jones  <jeremyj@apple.com>
82
83         StandardCanPlayThrough.html does not work on iOS.
84         https://bugs.webkit.org/show_bug.cgi?id=169207
85
86         Reviewed by Jer Noble.
87
88         On iOS only, add muted=true and autoplay=allowed. The loading policy on iOS requires these attributes to load enough
89         data to trigger canPlayThrough.
90
91         * Media/StandardCanPlayThrough.html:
92
93 2017-02-13  Zalan Bujtas <zalan@apple.com>
94
95         Simple line layout: Add performance test with text-rendering: optimizeSpeed
96         https://bugs.webkit.org/show_bug.cgi?id=168248
97
98         Reviewed by Simon Fraser.
99
100         * Layout/simple-line-layout-with-varying-content-and-optimized-speed.html: Added.
101
102 2017-02-10  Zalan Bujtas  <zalan@apple.com>
103
104         Mail hangs when removing multiple rows from large table.
105         https://bugs.webkit.org/show_bug.cgi?id=168103
106         <rdar://problem/30090186>
107
108         Reviewed by Ryosuke Niwa.
109
110         * DOM/large-table-edit.html: Added.
111
112 2017-02-05  Filip Pizlo  <fpizlo@apple.com>
113
114         Beef up the ARES-6 explainer text
115         https://bugs.webkit.org/show_bug.cgi?id=167867
116
117         Reviewed by Saam Barati.
118         
119         Added appropriate links.
120
121         Added text that explains the subscores.
122
123         * ARES-6/index.html:
124
125 2017-02-06  Jer Noble  <jer.noble@apple.com>
126
127         Playback stalls when a SourceBuffer append causes frame eviction
128         https://bugs.webkit.org/show_bug.cgi?id=167834
129
130         Reviewed by Eric Carlson.
131
132         Add an in-page performance test measuring the amount of time required
133         to append a large amount of media data to a SourceBuffer, and then to
134         completely remove that data 30s at a time.
135
136         Add a microbenchmark for MediaTime which measures the amount of time
137         required to create a 1M entry std::map and traverse the map 1M times.
138
139         * Media/MSERemoveCodedFrames.html: Added.
140         * Media/media-source-loader.js:
141         (MediaSourceLoader.prototype.get duration):
142         * MediaTime/Configurations/Base.xcconfig: Added.
143         * MediaTime/Configurations/DebugRelease.xcconfig: Added.
144         * MediaTime/Makefile: Added.
145         * MediaTime/MediaTime.xcodeproj/project.pbxproj: Added.
146         * MediaTime/main.cpp: Added.
147         (performTest):
148         (test):
149         (main):
150         * Skipped:
151
152 2017-02-06  Saam Barati  <sbarati@apple.com>
153
154         Make ARES-6 work from the CLI again
155         https://bugs.webkit.org/show_bug.cgi?id=167895
156
157         Reviewed by Michael Saboff.
158
159         * ARES-6/driver.js:
160         (Driver.prototype.readyTrigger):
161         (Driver.prototype.disableTrigger):
162         * ARES-6/stats.js:
163         (Stats.prototype.toString.span.span):
164         (Stats.prototype.toString.return.span):
165         (Stats.prototype.toString):
166
167 2017-02-06  Ryan Haddad  <ryanhaddad@apple.com>
168
169         Skip ARES-6 on performance bots
170         https://bugs.webkit.org/show_bug.cgi?id=167863
171
172         Unreviewed test gardening.
173
174         ES6SampleBench was renamed to ARES-6, so this should be updated in PerformanceTests/Skipped as well.
175
176         * Skipped:
177
178 2017-02-05  Filip Pizlo  <fpizlo@apple.com>
179
180         Change ES6SampleBench into ARES-6
181         https://bugs.webkit.org/show_bug.cgi?id=167863
182
183         Rubber stamped by Saam Barati.
184         
185         This imports changes that Jonathan Davis (jond@apple.com) made to ES6SampleBench to make it look
186         super awesome. Its now called ARES-6.
187
188         * ARES-6: Copied from PerformanceTests/ES6SampleBench.
189         * ARES-6/ARES-6.svg: Added.
190         * ARES-6/driver.js:
191         (Driver):
192         (Driver.prototype.readyTrigger):
193         (Driver.prototype.disableTrigger):
194         (Driver.prototype.start):
195         (Driver.prototype._updateIterations):
196         * ARES-6/glue.js:
197         * ARES-6/index.html:
198         * ARES-6/results.js:
199         (Results.prototype.reportRunning):
200         (Results.prototype.reportDone):
201         * ARES-6/stats.js:
202         (Stats.prototype.toString.return.span):
203         (Stats.prototype.toString.span.span):
204         (Stats.prototype.toString): Deleted.
205         * ARES-6/style.css: Removed.
206         * ARES-6/styles.css: Added.
207         (html):
208         (body):
209         (header,):
210         (p a):
211         (p a:hover):
212         (.flip):
213         (header):
214         (h2,):
215         (p):
216         (.about h2):
217         (.logo):
218         (.start):
219         (.start.ready):
220         (.start.ready:hover):
221         (.start:after):
222         (.start.ready .testrun):
223         (.tests):
224         (.test):
225         (.overall h2,):
226         (.test .indicator):
227         (.test .indicator.running):
228         (@keyframes test-running):
229         (100%):
230         (.score):
231         (.score label):
232         (.score .value):
233         (.score .units):
234         (.score .margin):
235         (.overall):
236         (.overall .score):
237         (.overall .margin):
238         (.testing, .about):
239         (.testing):
240         (.about):
241         (#about:target):
242         (@media only screen and (max-width: 784px)):
243         * ARES-6/swoop.svg: Added.
244         * ES6SampleBench: Removed.
245         * ES6SampleBench/Air: Removed.
246         * ES6SampleBench/Air/README.md: Removed.
247         * ES6SampleBench/Air/airjs-tests.yaml: Removed.
248         * ES6SampleBench/Air/all.js: Removed.
249         * ES6SampleBench/Air/allocate_stack.js: Removed.
250         * ES6SampleBench/Air/arg.js: Removed.
251         * ES6SampleBench/Air/basic_block.js: Removed.
252         * ES6SampleBench/Air/benchmark.js: Removed.
253         * ES6SampleBench/Air/code.js: Removed.
254         * ES6SampleBench/Air/custom.js: Removed.
255         * ES6SampleBench/Air/frequented_block.js: Removed.
256         * ES6SampleBench/Air/insertion_set.js: Removed.
257         * ES6SampleBench/Air/inst.js: Removed.
258         * ES6SampleBench/Air/liveness.js: Removed.
259         * ES6SampleBench/Air/make_dist.sh: Removed.
260         * ES6SampleBench/Air/opcode.js: Removed.
261         * ES6SampleBench/Air/payload-airjs-ACLj8C.js: Removed.
262         * ES6SampleBench/Air/payload-gbemu-executeIteration.js: Removed.
263         * ES6SampleBench/Air/payload-imaging-gaussian-blur-gaussianBlur.js: Removed.
264         * ES6SampleBench/Air/payload-typescript-scanIdentifier.js: Removed.
265         * ES6SampleBench/Air/reg.js: Removed.
266         * ES6SampleBench/Air/stack_slot.js: Removed.
267         * ES6SampleBench/Air/stress-test.js: Removed.
268         * ES6SampleBench/Air/strip-hash.rb: Removed.
269         * ES6SampleBench/Air/symbols.js: Removed.
270         * ES6SampleBench/Air/test.html: Removed.
271         * ES6SampleBench/Air/test.js: Removed.
272         * ES6SampleBench/Air/tmp.js: Removed.
273         * ES6SampleBench/Air/tmp_base.js: Removed.
274         * ES6SampleBench/Air/util.js: Removed.
275         * ES6SampleBench/Basic: Removed.
276         * ES6SampleBench/Basic/ast.js: Removed.
277         * ES6SampleBench/Basic/basic-tests.yaml: Removed.
278         * ES6SampleBench/Basic/basic.js: Removed.
279         * ES6SampleBench/Basic/benchmark.js: Removed.
280         * ES6SampleBench/Basic/caseless_map.js: Removed.
281         * ES6SampleBench/Basic/lexer.js: Removed.
282         * ES6SampleBench/Basic/number.js: Removed.
283         * ES6SampleBench/Basic/parser.js: Removed.
284         * ES6SampleBench/Basic/random.js: Removed.
285         * ES6SampleBench/Basic/state.js: Removed.
286         * ES6SampleBench/Basic/stress-test.js: Removed.
287         * ES6SampleBench/Basic/test.html: Removed.
288         * ES6SampleBench/Basic/test.js: Removed.
289         * ES6SampleBench/Basic/util.js: Removed.
290         * ES6SampleBench/air_benchmark.js: Removed.
291         * ES6SampleBench/basic_benchmark.js: Removed.
292         * ES6SampleBench/cli.js: Removed.
293         * ES6SampleBench/driver.js: Removed.
294         * ES6SampleBench/glue.js: Removed.
295         * ES6SampleBench/index.html: Removed.
296         * ES6SampleBench/results.js: Removed.
297         * ES6SampleBench/stats.js: Removed.
298         * ES6SampleBench/style.css: Removed.
299
300 2017-02-05  Zalan Bujtas  <zalan@apple.com>
301
302         Simple line layout: Add performance test with varying string and font-size to minimize width-cache hit.
303         https://bugs.webkit.org/show_bug.cgi?id=167859
304         <rdar://problem/30368070>
305
306         Reviewed by Antti Koivisto.
307
308         This is in preparation for webkit.org/b/167843.
309
310         * Layout/simple-line-layout-with-varying-content.html: Added.
311
312 2017-01-16  Filip Pizlo  <fpizlo@apple.com>
313
314         Add framerate baselines to the splay-detail test
315         https://bugs.webkit.org/show_bug.cgi?id=167659
316
317         Reviewed by Geoffrey Garen.
318
319         * JetStream/Octane2/splay-detail.html:
320
321 2017-01-31  Carlos Alberto Lopez Perez  <clopez@igalia.com>
322
323         [EFL][GTK] Skip some IndexedDB tests that time out on the Perf bots.
324
325         Unreviewed performance test gardening.
326
327         * Skipped:
328
329 2017-01-30  Zalan Bujtas  <zalan@apple.com>
330
331         Simple line layout: Small tweaks to improve performance.
332         https://bugs.webkit.org/show_bug.cgi?id=167611
333         <rdar://problem/30274294>
334
335         Reviewed by Simon Fraser.
336
337         * Layout/simple-line-layout-non-repeating-text.html: Added.
338
339 2017-01-13  Said Abou-Hallawa  <sabouhallawa@apple.com>
340
341         26 MotionMark performance tests failing
342         https://bugs.webkit.org/show_bug.cgi?id=166854
343
344         Reviewed by Ryosuke Niwa.
345         
346         Skip running MotionMark on WebKit perf bots.
347
348         * Skipped:
349
350 2017-01-06  Said Abou-Hallawa  <sabouhallawa@apple.com>
351
352         Rename the directory and the files of MotionMark from Animometer to MotionMark
353         https://bugs.webkit.org/show_bug.cgi?id=166659
354
355         Reviewed by Jon Lee.
356         
357         Rename the directory and the files of the benchmark to its new name.
358
359         * MotionMark/developer.html: Renamed from PerformanceTests/Animometer/developer.html.
360         * MotionMark/index.html: Renamed from PerformanceTests/Animometer/index.html.
361         * MotionMark/resources/debug-runner/d3.min.js: Renamed from PerformanceTests/Animometer/resources/debug-runner/d3.min.js.
362         * MotionMark/resources/debug-runner/graph.js: Renamed from PerformanceTests/Animometer/resources/debug-runner/graph.js.
363         * MotionMark/resources/debug-runner/motionmark.css: Renamed from PerformanceTests/Animometer/resources/debug-runner/animometer.css.
364         * MotionMark/resources/debug-runner/motionmark.js: Renamed from PerformanceTests/Animometer/resources/debug-runner/animometer.js.
365         * MotionMark/resources/debug-runner/tests.js: Renamed from PerformanceTests/Animometer/resources/debug-runner/tests.js.
366         * MotionMark/resources/extensions.js: Renamed from PerformanceTests/Animometer/resources/extensions.js.
367         * MotionMark/resources/runner/benchmark-runner.js: Renamed from PerformanceTests/Animometer/resources/runner/benchmark-runner.js.
368         * MotionMark/resources/runner/crystal.svg: Renamed from PerformanceTests/Animometer/resources/runner/crystal.svg.
369         * MotionMark/resources/runner/lines.svg: Renamed from PerformanceTests/Animometer/resources/runner/lines.svg.
370         * MotionMark/resources/runner/logo.svg: Renamed from PerformanceTests/Animometer/resources/runner/logo.svg.
371         * MotionMark/resources/runner/motionmark.css: Renamed from PerformanceTests/Animometer/resources/runner/animometer.css.
372         * MotionMark/resources/runner/motionmark.js: Renamed from PerformanceTests/Animometer/resources/runner/animometer.js.
373         * MotionMark/resources/runner/tests.js: Renamed from PerformanceTests/Animometer/resources/runner/tests.js.
374         * MotionMark/resources/statistics.js: Renamed from PerformanceTests/Animometer/resources/statistics.js.
375         * MotionMark/resources/strings.js: Renamed from PerformanceTests/Animometer/resources/strings.js.
376         * MotionMark/tests/3d/resources/webgl.js: Renamed from PerformanceTests/Animometer/tests/3d/resources/webgl.js.
377         * MotionMark/tests/3d/webgl.html: Renamed from PerformanceTests/Animometer/tests/3d/webgl.html.
378         * MotionMark/tests/bouncing-particles/bouncing-canvas-images.html: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/bouncing-canvas-images.html.
379         * MotionMark/tests/bouncing-particles/bouncing-canvas-shapes.html: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/bouncing-canvas-shapes.html.
380         * MotionMark/tests/bouncing-particles/bouncing-css-images.html: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/bouncing-css-images.html.
381         * MotionMark/tests/bouncing-particles/bouncing-css-shapes.html: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/bouncing-css-shapes.html.
382         * MotionMark/tests/bouncing-particles/bouncing-svg-images.html: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/bouncing-svg-images.html.
383         * MotionMark/tests/bouncing-particles/bouncing-svg-shapes.html: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/bouncing-svg-shapes.html.
384         * MotionMark/tests/bouncing-particles/bouncing-tagged-images.html: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/bouncing-tagged-images.html.
385         * MotionMark/tests/bouncing-particles/resources/bouncing-canvas-images.js: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/bouncing-canvas-images.js.
386         * MotionMark/tests/bouncing-particles/resources/bouncing-canvas-particles.js: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/bouncing-canvas-particles.js.
387         * MotionMark/tests/bouncing-particles/resources/bouncing-canvas-shapes.js: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/bouncing-canvas-shapes.js.
388         * MotionMark/tests/bouncing-particles/resources/bouncing-css-images.js: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/bouncing-css-images.js.
389         * MotionMark/tests/bouncing-particles/resources/bouncing-css-shapes.js: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/bouncing-css-shapes.js.
390         * MotionMark/tests/bouncing-particles/resources/bouncing-svg-images.js: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/bouncing-svg-images.js.
391         * MotionMark/tests/bouncing-particles/resources/bouncing-svg-particles.js: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/bouncing-svg-particles.js.
392         * MotionMark/tests/bouncing-particles/resources/bouncing-svg-shapes.js: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/bouncing-svg-shapes.js.
393         * MotionMark/tests/bouncing-particles/resources/bouncing-tagged-images.js: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/bouncing-tagged-images.js.
394         * MotionMark/tests/bouncing-particles/resources/image1.jpg: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/image1.jpg.
395         * MotionMark/tests/bouncing-particles/resources/image2.jpg: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/image2.jpg.
396         * MotionMark/tests/bouncing-particles/resources/image3.jpg: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/image3.jpg.
397         * MotionMark/tests/bouncing-particles/resources/image4.jpg: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/image4.jpg.
398         * MotionMark/tests/bouncing-particles/resources/image5.jpg: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/image5.jpg.
399         * MotionMark/tests/dom/compositing-transforms.html: Renamed from PerformanceTests/Animometer/tests/dom/compositing-transforms.html.
400         * MotionMark/tests/dom/focus.html: Renamed from PerformanceTests/Animometer/tests/dom/focus.html.
401         * MotionMark/tests/dom/leaves.html: Renamed from PerformanceTests/Animometer/tests/dom/leaves.html.
402         * MotionMark/tests/dom/particles.html: Renamed from PerformanceTests/Animometer/tests/dom/particles.html.
403         * MotionMark/tests/dom/resources/compositing-transforms.js: Renamed from PerformanceTests/Animometer/tests/dom/resources/compositing-transforms.js.
404         * MotionMark/tests/dom/resources/dom-particles.js: Renamed from PerformanceTests/Animometer/tests/dom/resources/dom-particles.js.
405         * MotionMark/tests/dom/resources/focus.js: Renamed from PerformanceTests/Animometer/tests/dom/resources/focus.js.
406         * MotionMark/tests/dom/resources/leaves.js: Renamed from PerformanceTests/Animometer/tests/dom/resources/leaves.js.
407         * MotionMark/tests/master/canvas-stage.html: Renamed from PerformanceTests/Animometer/tests/master/canvas-stage.html.
408         * MotionMark/tests/master/focus.html: Renamed from PerformanceTests/Animometer/tests/master/focus.html.
409         * MotionMark/tests/master/image-data.html: Renamed from PerformanceTests/Animometer/tests/master/image-data.html.
410         * MotionMark/tests/master/leaves.html: Renamed from PerformanceTests/Animometer/tests/master/leaves.html.
411         * MotionMark/tests/master/multiply.html: Renamed from PerformanceTests/Animometer/tests/master/multiply.html.
412         * MotionMark/tests/master/resources/canvas-stage.js: Renamed from PerformanceTests/Animometer/tests/master/resources/canvas-stage.js.
413         * MotionMark/tests/master/resources/canvas-tests.js: Renamed from PerformanceTests/Animometer/tests/master/resources/canvas-tests.js.
414         * MotionMark/tests/master/resources/compass.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/compass.svg.
415         * MotionMark/tests/master/resources/compass100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/compass100.png.
416         * MotionMark/tests/master/resources/console.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/console.svg.
417         * MotionMark/tests/master/resources/console100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/console100.png.
418         * MotionMark/tests/master/resources/contribute.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/contribute.svg.
419         * MotionMark/tests/master/resources/contribute100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/contribute100.png.
420         * MotionMark/tests/master/resources/debugger.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/debugger.svg.
421         * MotionMark/tests/master/resources/debugger100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/debugger100.png.
422         * MotionMark/tests/master/resources/focus.js: Renamed from PerformanceTests/Animometer/tests/master/resources/focus.js.
423         * MotionMark/tests/master/resources/image-data.js: Renamed from PerformanceTests/Animometer/tests/master/resources/image-data.js.
424         * MotionMark/tests/master/resources/inspector.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/inspector.svg.
425         * MotionMark/tests/master/resources/inspector100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/inspector100.png.
426         * MotionMark/tests/master/resources/layout.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/layout.svg.
427         * MotionMark/tests/master/resources/layout100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/layout100.png.
428         * MotionMark/tests/master/resources/leaves.js: Renamed from PerformanceTests/Animometer/tests/master/resources/leaves.js.
429         * MotionMark/tests/master/resources/multiply.js: Renamed from PerformanceTests/Animometer/tests/master/resources/multiply.js.
430         * MotionMark/tests/master/resources/particles.js: Renamed from PerformanceTests/Animometer/tests/master/resources/particles.js.
431         * MotionMark/tests/master/resources/performance.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/performance.svg.
432         * MotionMark/tests/master/resources/performance100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/performance100.png.
433         * MotionMark/tests/master/resources/script.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/script.svg.
434         * MotionMark/tests/master/resources/script100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/script100.png.
435         * MotionMark/tests/master/resources/shortcuts.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/shortcuts.svg.
436         * MotionMark/tests/master/resources/shortcuts100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/shortcuts100.png.
437         * MotionMark/tests/master/resources/standards.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/standards.svg.
438         * MotionMark/tests/master/resources/standards100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/standards100.png.
439         * MotionMark/tests/master/resources/storage.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/storage.svg.
440         * MotionMark/tests/master/resources/storage100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/storage100.png.
441         * MotionMark/tests/master/resources/styles.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/styles.svg.
442         * MotionMark/tests/master/resources/styles100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/styles100.png.
443         * MotionMark/tests/master/resources/svg-particles.js: Renamed from PerformanceTests/Animometer/tests/master/resources/svg-particles.js.
444         * MotionMark/tests/master/resources/text.js: Renamed from PerformanceTests/Animometer/tests/master/resources/text.js.
445         * MotionMark/tests/master/resources/timeline.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/timeline.svg.
446         * MotionMark/tests/master/resources/timeline100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/timeline100.png.
447         * MotionMark/tests/master/svg-particles.html: Renamed from PerformanceTests/Animometer/tests/master/svg-particles.html.
448         * MotionMark/tests/master/text.html: Renamed from PerformanceTests/Animometer/tests/master/text.html.
449         * MotionMark/tests/resources/main.js: Renamed from PerformanceTests/Animometer/tests/resources/main.js.
450         * MotionMark/tests/resources/math.js: Renamed from PerformanceTests/Animometer/tests/resources/math.js.
451         * MotionMark/tests/resources/stage.css: Renamed from PerformanceTests/Animometer/tests/resources/stage.css.
452         * MotionMark/tests/resources/star.svg: Renamed from PerformanceTests/Animometer/tests/resources/star.svg.
453         * MotionMark/tests/resources/yin-yang.png: Renamed from PerformanceTests/Animometer/tests/resources/yin-yang.png.
454         * MotionMark/tests/resources/yin-yang.svg: Renamed from PerformanceTests/Animometer/tests/resources/yin-yang.svg.
455         * MotionMark/tests/simple/resources/simple-canvas-paths.js: Renamed from PerformanceTests/Animometer/tests/simple/resources/simple-canvas-paths.js.
456         * MotionMark/tests/simple/resources/simple-canvas.js: Renamed from PerformanceTests/Animometer/tests/simple/resources/simple-canvas.js.
457         * MotionMark/tests/simple/resources/tiled-canvas-image.js: Renamed from PerformanceTests/Animometer/tests/simple/resources/tiled-canvas-image.js.
458         * MotionMark/tests/simple/simple-canvas-paths.html: Renamed from PerformanceTests/Animometer/tests/simple/simple-canvas-paths.html.
459         * MotionMark/tests/simple/tiled-canvas-image.html: Renamed from PerformanceTests/Animometer/tests/simple/tiled-canvas-image.html.
460         * MotionMark/tests/template/resources/template-canvas.js: Renamed from PerformanceTests/Animometer/tests/template/resources/template-canvas.js.
461         * MotionMark/tests/template/resources/template-css.js: Renamed from PerformanceTests/Animometer/tests/template/resources/template-css.js.
462         * MotionMark/tests/template/resources/template-svg.js: Renamed from PerformanceTests/Animometer/tests/template/resources/template-svg.js.
463         * MotionMark/tests/template/template-canvas.html: Renamed from PerformanceTests/Animometer/tests/template/template-canvas.html.
464         * MotionMark/tests/template/template-css.html: Renamed from PerformanceTests/Animometer/tests/template/template-css.html.
465         * MotionMark/tests/template/template-svg.html: Renamed from PerformanceTests/Animometer/tests/template/template-svg.html.
466
467 2016-12-13  Brady Eidson  <beidson@apple.com>
468
469         Add a PerfTest targeting IDBObjectStore.get().
470         https://bugs.webkit.org/show_bug.cgi?id=165816
471
472         Reviewed by Alex Christensen.
473
474         * IndexedDB/objectstore-get.html: Added.
475
476 2016-12-13  Brady Eidson  <beidson@apple.com>
477
478         Add a PerfTest targeting IDBIndex.get().
479         https://bugs.webkit.org/show_bug.cgi?id=165803
480
481         Reviewed by Alex Christensen.
482
483         * IndexedDB/index-get.html: Added.
484
485 2016-12-12  Brady Eidson  <beidson@apple.com>
486
487         More IndexedDB perf tests.
488         https://bugs.webkit.org/show_bug.cgi?id=165634
489
490         Reviewed by Sam Weinig.
491
492         * IndexedDB/index-multientry.html:
493         * IndexedDB/large-number-of-inserts-responsiveness.html: Added.
494         * IndexedDB/large-number-of-inserts.html: Added.
495         * IndexedDB/objectstore-cursor.html: Added. Runtime cut in 1/4th compared to r209672.
496
497         * resources/runner.js:
498
499 2016-12-12  Commit Queue  <commit-queue@webkit.org>
500
501         Unreviewed, rolling out r209672.
502         https://bugs.webkit.org/show_bug.cgi?id=165766
503
504         IndexedDB/objectstore-cursor.html perf test always times out
505         (Requested by ap on #webkit).
506
507         Reverted changeset:
508
509         "More IndexedDB perf tests."
510         https://bugs.webkit.org/show_bug.cgi?id=165634
511         http://trac.webkit.org/changeset/209672
512
513 2016-12-10  Brady Eidson  <beidson@apple.com>
514
515         More IndexedDB perf tests.
516         https://bugs.webkit.org/show_bug.cgi?id=165634
517
518         Reviewed by Sam Weinig.
519
520         * IndexedDB/index-multientry.html:
521         * IndexedDB/large-number-of-inserts-responsiveness.html: Added.
522         * IndexedDB/large-number-of-inserts.html: Added.
523         * IndexedDB/objectstore-cursor.html: Added.
524         
525         * resources/runner.js: Add "track responsiveness" functionality to PerfTestRunner.
526
527 2016-12-06  Filip Pizlo  <fpizlo@apple.com>
528
529         Concurrent GC should be stable enough to land enabled
530         https://bugs.webkit.org/show_bug.cgi?id=164990
531
532         Reviewed by Geoffrey Garen.
533         
534         Made CDjs more configurable and refined the "large.js" configuration. I was using that one and
535         the new "long.js" configuration to tune concurrent eden GCs.
536         
537         Added a new way of running Splay in browser, which using chartjs to plot the execution times of
538         2000 iterations. This includes the minified chartjs.
539
540         * JetStream/Octane2/splay-detail.html: Added.
541         * JetStream/cdjs/benchmark.js:
542         (benchmarkImpl):
543         (benchmark):
544         * JetStream/cdjs/long.js: Added.
545
546 2016-12-07  Brady Eidson  <beidson@apple.com>
547
548         Add IDB perf tests stressing key size.
549         https://bugs.webkit.org/show_bug.cgi?id=165567
550
551         Reviewed by Alex Christensen.
552
553         * IndexedDB/large-array-keys.html: Added.
554         * IndexedDB/large-binary-keys.html: Added.
555         * IndexedDB/large-string-keys.html: Added.
556
557 2016-12-05  Brady Eidson  <beidson@apple.com>
558
559         Add an IndexedDB perf test to PerformanceTests.
560         https://bugs.webkit.org/show_bug.cgi?id=165430
561
562         Reviewed by Alex Christensen.
563
564         * IndexedDB/index-multientry.html: Added.
565
566 2016-12-02  Filip Pizlo  <fpizlo@apple.com>
567
568         ES6SampleBench should report an average for Steady State so that all of the numbers are comparable
569         https://bugs.webkit.org/show_bug.cgi?id=165325
570
571         Reviewed by Saam Barati.
572
573         This makes all of the numbers that ES6SampleBench reports comparable to each other: they all speak of the time
574         it took to run an iteration of something.
575
576         * ES6SampleBench/results.js:
577         (Results.prototype.reportResult):
578
579 2016-11-18  Jer Noble  <jer.noble@apple.com>
580
581         Add some performance tests for various aspects of media loading and playback.
582         https://bugs.webkit.org/show_bug.cgi?id=164977
583
584         Reviewed by Darin Adler.
585
586         Add new tests for a few aspects of media playback; namely: how quickly  media elements fire
587         the "canplaythrough" event when all data is immediately available, how quickly HLS playback
588         switches resolutions, and at what playback rate MSE-backed video can play without dropping
589         frames. Skip these tests by default, as they're not runnable on all ports, and they require
590         a webserver (run-webkit-httpd) to complete on macOS.
591
592         * Media/HLSCanPlayThrough.html: Added.
593         * Media/HLSGearChange.html: Added.
594         * Media/MSECanPlayThrough.html: Added.
595         * Media/MSEPlaybackRate.html: Added.
596         * Media/StandardCanPlayThrough.html: Added.
597         * Media/hls/1080p/iframe_index.m3u8: Added.
598         * Media/hls/1080p/prog_index.m3u8: Added.
599         * Media/hls/1080p/test.ts: Added.
600         * Media/hls/480p/iframe_index.m3u8: Added.
601         * Media/hls/480p/prog_index.m3u8: Added.
602         * Media/hls/480p/test.ts: Added.
603         * Media/hls/720p/iframe_index.m3u8: Added.
604         * Media/hls/720p/prog_index.m3u8: Added.
605         * Media/hls/720p/test.ts: Added.
606         * Media/hls/720p/test.ts.back: Added.
607         * Media/hls/index.m3u8: Added.
608         * Media/media-source-loader.js: Added.
609         (MediaSourceLoader):
610         (MediaSourceLoader.prototype.loadManifest):
611         (MediaSourceLoader.prototype.loadManifestSucceeded):
612         (MediaSourceLoader.prototype.loadManifestFailed):
613         (MediaSourceLoader.prototype.loadMediaData):
614         (MediaSourceLoader.prototype.loadMediaDataSucceeded):
615         (MediaSourceLoader.prototype.loadMediaDataFailed):
616         (MediaSourceLoader.prototype.get type):
617         (MediaSourceLoader.prototype.get duration):
618         (MediaSourceLoader.prototype.get initSegment):
619         (MediaSourceLoader.prototype.get mediaSegmentsLength):
620         (MediaSourceLoader.prototype.mediaSegments):
621         (MediaSourceLoader.prototype.get everyMediaSegment):
622         * Media/test-fragmented-video.json: Added.
623         * Media/test-fragmented-video.mp4: Added.
624         * Media/test.mp4: Added.
625         * Skipped:
626
627 2016-11-18  Filip Pizlo  <fpizlo@apple.com>
628
629         Concurrent GC should be able to run splay in debug mode and earley/raytrace in release mode with no perf regression
630         https://bugs.webkit.org/show_bug.cgi?id=164282
631
632         Reviewed by Geoffrey Garen and Oliver Hunt.
633         
634         CDjs is a fun benchmark for stressing concurrent GCs, but to really give the GC a good
635         workout you need to increase the amount of work that the test does. This adds a second
636         configuration of the benchmark that has more aircraft. It uses much more memory and causes us
637         to do more GCs and those GCs take longer.
638
639         * JetStream/cdjs/benchmark.js:
640         (benchmarkImpl):
641         (benchmark):
642         * JetStream/cdjs/large.js: Added.
643
644 2016-11-14  Filip Pizlo  <fpizlo@apple.com>
645
646         Unreviewed, revert unintended change.
647
648         * ES6SampleBench/Air/benchmark.js:
649
650 2016-10-03  Saam Barati  <sbarati@apple.com>
651
652         MapHash should speculate on the type of its child node
653         https://bugs.webkit.org/show_bug.cgi?id=161922
654
655         Reviewed by Filip Pizlo.
656
657         I gave the main generator a name so it's easier to see what
658         it is when using the sampling profiler.
659
660         * ES6SampleBench/Basic/ast.js:
661         (Basic.Program):
662
663 2016-09-19  Sergio Villar Senin  <svillar@igalia.com>
664
665         [css-grid] Remove the x2 computation of row sizes with indefinite heights
666         https://bugs.webkit.org/show_bug.cgi?id=162150
667
668         Reviewed by Darin Adler.
669
670         Added a new test case which checks the layout performance of grids inside other grids, i.e,
671         grids acting both as grid container and grid item.
672
673         * Layout/nested-grid.html: Added.
674
675 2016-09-09  Simon Fraser  <simon.fraser@apple.com>
676
677         Perf test Animation/css-accelerated-animation.html failing
678         https://bugs.webkit.org/show_bug.cgi?id=161795
679
680         Skip Animation/css-accelerated-animation.html since this only works as a MobileSafari
681         perf test.
682
683         * Skipped:
684
685 2016-09-08  Simon Fraser  <simon.fraser@apple.com>
686
687         Add a content-animation test with accelerated CSS animations
688         https://bugs.webkit.org/show_bug.cgi?id=161776
689
690         Reviewed by Dean Jackson.
691
692         Add a test that measures the frame rate of accelerated CSS animations.
693
694         This is like css-animation.html (which animates 'left' and 'top') but animates
695         transforms instead. In order to get animation in X and Y we make two nested elements
696         and set transformX() on one, and transformY() on the other.
697
698         * Animation/css-accelerated-animation.html: Added.
699
700 2016-08-23  Saam Barati  <sbarati@apple.com>
701
702         It should be easy to run ES6SampleBench from the jsc shell
703         https://bugs.webkit.org/show_bug.cgi?id=161085
704
705         Reviewed by Yusuke Suzuki.
706
707         This patch makes ES6 sample bench runnable through the `jsc` shell.
708         To do that, you need to be in the PerformanceTests/ES6SampleBench
709         directory and run `jsc cli.js`. To make this work, the benchmark
710         is now aware of being run in two modes: via the browser, and via
711         the shell. Each entry point will set a variable `isInBrowser`,
712         and the benchmark will do different things based on if that
713         variable is true or false.
714
715         * ES6SampleBench/Air/benchmark.js:
716         * ES6SampleBench/Air/stress-test.js:
717         * ES6SampleBench/Basic/benchmark.js:
718         (runBenchmark):
719         * ES6SampleBench/Basic/stress-test.js:
720         * ES6SampleBench/air_benchmark.js:
721         * ES6SampleBench/basic_benchmark.js:
722         * ES6SampleBench/cli.js: Added.
723         (return.doRun):
724         (makeBenchmarkRunner):
725         * ES6SampleBench/driver.js:
726         (Driver):
727         (Driver.prototype._recomputeSummary):
728         (Driver.prototype._iterate.window.setTimeout):
729         (Driver.prototype._iterate):
730         (Driver.prototype._updateIterations):
731         * ES6SampleBench/glue.js:
732         (reportResult):
733         * ES6SampleBench/index.html:
734         * ES6SampleBench/results.js:
735         (Results):
736         (Results.prototype.reportRunning):
737         (Results.prototype.reportDone):
738         (Results.prototype.reportError):
739         * ES6SampleBench/stats.js:
740         (Stats):
741         (Stats.prototype.toString):
742         (Stats.prototype._update):
743
744 2016-08-01  Said Abou-Hallawa  <sabouhallawa@apple.com>
745
746         Add an option to run the MotionMark tests with classic tile size (512x512)
747         https://bugs.webkit.org/show_bug.cgi?id=160371
748
749         Reviewed by Darin Adler.
750
751         Setting the body to large size (3000x3000) and overflow to scroll forces
752         the classic tile size (512x512). This option does not affect the canvas
753         tests because the canvas is displayed on one tile regardless of its size.
754
755         * Animometer/developer.html:
756         * Animometer/resources/debug-runner/animometer.css:
757         (body.showing-test-container.tiles-big):
758         (body.showing-test-container.tiles-classic):
759         * Animometer/resources/debug-runner/animometer.js:
760         * Animometer/resources/runner/animometer.js:
761         (window.benchmarkController.startBenchmark):
762
763 2016-06-29  Filip Pizlo  <fpizlo@apple.com>
764
765         Generators violate bytecode liveness validation
766         https://bugs.webkit.org/show_bug.cgi?id=159279
767
768         Reviewed by Yusuke Suzuki.
769
770         Add Basic to our test harness.
771
772         Also made some cosmetic changes to the benchmark harness.
773
774         * ES6SampleBench/Basic/basic-tests.yaml: Added.
775         * ES6SampleBench/Basic/stress-test.js: Added.
776         (preciseTime):
777         * ES6SampleBench/driver.js:
778         (Driver):
779         (Driver.prototype.start):
780         (Driver.prototype.reportError):
781         * ES6SampleBench/glue.js:
782         * ES6SampleBench/index.html:
783
784 2016-06-28  Filip Pizlo  <fpizlo@apple.com>
785
786         ES6SampleBench should have a harness
787         https://bugs.webkit.org/show_bug.cgi?id=159246
788
789         Reviewed by Saam Barati.
790         
791         This adds a simple web harness for ES6SampleBench. It runs Air and Basic 10 times for 200
792         iterations each and reports three metrics:
793         
794         First iteration: the time it takes for the first iteration to run. This is the first
795         iteration after the benchmark is loaded into the iframe, so it's representative of what
796         would happen if one of these workloads only ran for a short time.
797         
798         Worst 2%: of the last 199 iterations, the average of the worst 2% iterations. If code like
799         any of these workloads was used in an important event handler, you'd want that code to run
800         well in the worst case in addition to having great throughput.
801         
802         Steady state: the total of the last 199 iterations. This is representative of what would
803         happen if you ran code like this for a long time.
804         
805         The total score is the geomean of the firstIteration/worstCase/steadyState numbers of the
806         two benchmarks.
807         
808         The harness does statistics using Student's T-distribution confidence intervals.
809
810         * ES6SampleBench/Basic/benchmark.js:
811         (Benchmark):
812         * ES6SampleBench/air_benchmark.js: Added.
813         * ES6SampleBench/basic_benchmark.js: Added.
814         * ES6SampleBench/driver.js: Added.
815         (Driver):
816         (Driver.prototype.addBenchmark):
817         (Driver.prototype.start):
818         (Driver.prototype.reportResult):
819         (Driver.prototype.reportError):
820         (Driver.prototype._recomputeSummary.Geomean):
821         (Driver.prototype._recomputeSummary.Geomean.prototype.add):
822         (Driver.prototype._recomputeSummary.Geomean.prototype.get result):
823         (Driver.prototype._recomputeSummary):
824         (Driver.prototype._iterate):
825         (Driver.prototype._updateIterations):
826         * ES6SampleBench/glue.js: Added.
827         * ES6SampleBench/index.html: Added.
828         * ES6SampleBench/results.js: Added.
829         (Results):
830         (Results.prototype.get benchmark):
831         (Results.prototype.reset):
832         (Results.prototype.reportRunning):
833         (Results.prototype.reportDone):
834         (Results.prototype.reportResult.averageAbovePercentile):
835         (Results.prototype.reportResult):
836         (Results.prototype.reportError):
837         * ES6SampleBench/stats.js: Added.
838         (Stats):
839         (Stats.prototype.reset):
840         (Stats.prototype.add):
841         (Stats.prototype.get numIterations):
842         (Stats.prototype.valueForIteration):
843         (Stats.get result.tDist):
844         (Stats.prototype.get result):
845         (Stats.prototype.toString):
846         (Stats.prototype._update):
847         * ES6SampleBench/style.css: Added.
848         (body):
849         (body, th, tr):
850         (h1, h2, h3, h4):
851         (h1):
852         (h2):
853         (h3):
854         (hr):
855         (address):
856         (img):
857         (.underline):
858         (ol.loweralpha):
859         (ol.upperalpha):
860         (ol.lowerroman):
861         (ol.upperroman):
862         (ol.arabic):
863         (.banner-link:link, .banner-link:visited):
864         (:link, :visited):
865         (h4 :link, h4 :visited, h5 :link, h5 :visited, h6 :link, h6 :visited):
866         (.anchor:link, .anchor:visited):
867         (* > .anchor:link, * > .anchor:visited):
868         (span:hover .anchor):
869         (a.forbidden, span.forbidden):
870         (a.missing:hover):
871         (a.closed:link, a.closed:visited, span.closed):
872         (pre):
873         (div.code):
874         (div.code pre):
875         (dt):
876         (dd):
877         (dd:last-child):
878         (.site-logo):
879         (.site-logo .tagline):
880         (table):
881         (#contents):
882         (p):
883         (p:last-child):
884         (th):
885         (td):
886
887 2016-06-28  Jon Lee  <jonlee@apple.com>
888
889         Update focus test
890         https://bugs.webkit.org/show_bug.cgi?id=159242
891         rdar://problem/27070007
892
893         Reviewed by Dean Jackson.
894         Provisionally reviewed by Said Abou-Hallawa.
895
896         Move previous test to dom suite, and update the test for better reporting of frame rate, although
897         it uses a different rendering path.
898
899         * Animometer/resources/debug-runner/tests.js: Add to dom suite.
900         * Animometer/tests/dom/focus.html: Copied from PerformanceTests/Animometer/tests/master/focus.html.
901         * Animometer/tests/dom/resources/focus.js: Copied from PerformanceTests/Animometer/tests/master/resources/focus.js.
902         * Animometer/tests/master/focus.html: Remove center element.
903         * Animometer/tests/master/resources/focus.js: Use narrower size range with smaller particles. Remove the
904         container elements. Inline getBlurValue and getOpacityValue since they are only called once.
905
906 2016-06-28  Filip Pizlo  <fpizlo@apple.com>
907
908         Move Air.js and Basic into ES6SampleBench
909
910         Rubber stamped by Geoffrey Garen.
911         
912         We want to group our ES6 benchmarks together, and eventually, we want to give them a
913         common harness.
914
915         * Air.js: Removed.
916         * Air.js/README.md: Removed.
917         * Air.js/airjs-tests.yaml: Removed.
918         * Air.js/all.js: Removed.
919         * Air.js/allocate_stack.js: Removed.
920         * Air.js/arg.js: Removed.
921         * Air.js/basic_block.js: Removed.
922         * Air.js/benchmark.js: Removed.
923         * Air.js/code.js: Removed.
924         * Air.js/custom.js: Removed.
925         * Air.js/frequented_block.js: Removed.
926         * Air.js/insertion_set.js: Removed.
927         * Air.js/inst.js: Removed.
928         * Air.js/liveness.js: Removed.
929         * Air.js/make_dist.sh: Removed.
930         * Air.js/opcode.js: Removed.
931         * Air.js/payload-airjs-ACLj8C.js: Removed.
932         * Air.js/payload-gbemu-executeIteration.js: Removed.
933         * Air.js/payload-imaging-gaussian-blur-gaussianBlur.js: Removed.
934         * Air.js/payload-typescript-scanIdentifier.js: Removed.
935         * Air.js/reg.js: Removed.
936         * Air.js/stack_slot.js: Removed.
937         * Air.js/stress-test.js: Removed.
938         * Air.js/strip-hash.rb: Removed.
939         * Air.js/symbols.js: Removed.
940         * Air.js/test.html: Removed.
941         * Air.js/test.js: Removed.
942         * Air.js/tmp.js: Removed.
943         * Air.js/tmp_base.js: Removed.
944         * Air.js/util.js: Removed.
945         * Basic: Removed.
946         * Basic/ast.js: Removed.
947         * Basic/basic.js: Removed.
948         * Basic/benchmark.js: Removed.
949         * Basic/caseless_map.js: Removed.
950         * Basic/lexer.js: Removed.
951         * Basic/number.js: Removed.
952         * Basic/parser.js: Removed.
953         * Basic/random.js: Removed.
954         * Basic/state.js: Removed.
955         * Basic/test.html: Removed.
956         * Basic/test.js: Removed.
957         * Basic/util.js: Removed.
958         * ES6SampleBench: Added.
959         * ES6SampleBench/Air: Copied from PerformanceTests/Air.js.
960         * ES6SampleBench/Basic: Copied from PerformanceTests/Basic.
961         * Skipped:
962
963 2016-06-24  Filip Pizlo  <fpizlo@apple.com>
964
965         Add a ES6 generator benchmark
966         https://bugs.webkit.org/show_bug.cgi?id=159101
967
968         Rubber stamped by Keith Miller.
969         
970         This adds a Basic interpreter loosely based on ECMA-55:
971         
972         http://www.ecma-international.org/publications/files/ECMA-ST-WITHDRAWN/ECMA-55,%201st%20Edition,%20January%201978.pdf
973         
974         It includes a lexer that is a generator, a parser that uses regular expressions, and an AST
975         walk interpreter where the walking functions for statements are generators that call each
976         other with yield*. This enables the interpreter to look like an AST walk even though it can
977         yield at INPUT and PRINT statements.
978         
979         This also uses for-of, classes, Map, and WeakMap. It also uses deprecated-but-awesome RegExp
980         features like RegExp.lastMatch and RegExp.rightContext. I did it that way because this is
981         how I've always written lexers in dynamic languages; see offlineasm's lex() method in
982         parser.rb for example.
983         
984         The benchmark runs a handful of simple Basic programs. The longest-running one computes
985         prime numbers.
986         
987         Includes a command-line and web harness. On my machine it runs in 2-3 seconds.
988
989         * Basic: Added.
990         * Basic/ast.js: Added.
991         (Basic.NumberApply):
992         (Basic.Variable):
993         (Basic.Const):
994         (Basic.NumberPow):
995         (Basic.NumberMul):
996         (Basic.NumberDiv):
997         (Basic.NumberNeg):
998         (Basic.NumberAdd):
999         (Basic.NumberSub):
1000         (Basic.StringVar):
1001         (Basic.Equals):
1002         (Basic.NotEquals):
1003         (Basic.LessThan):
1004         (Basic.GreaterThan):
1005         (Basic.LessEqual):
1006         (Basic.GreaterEqual):
1007         (Basic.GoTo):
1008         (Basic.GoSub):
1009         (Basic.Def):
1010         (Basic.Let):
1011         (Basic.If):
1012         (Basic.Return):
1013         (Basic.Stop):
1014         (Basic.On):
1015         (sideState.shouldStop):
1016         (Basic.For):
1017         (Basic.Next):
1018         (Basic.Print):
1019         (Basic.Input):
1020         (Basic.Read):
1021         (Basic.Restore):
1022         (Basic.Dim):
1023         (Basic.Randomize):
1024         (Basic.End):
1025         (Basic.Program):
1026         * Basic/basic.js: Added.
1027         (prepare):
1028         (simulate):
1029         * Basic/benchmark.js: Added.
1030         (Benchmark):
1031         (Benchmark.prototype.runIteration.expect):
1032         (Benchmark.prototype.runIteration):
1033         (runBenchmark):
1034         * Basic/caseless_map.js: Added.
1035         (CaselessMap):
1036         * Basic/lexer.js: Added.
1037         (lex.consumeWhitespace):
1038         (lex.consume):
1039         (lex):
1040         * Basic/lexer_test.js: Added.
1041         * Basic/number.js: Added.
1042         (NumberValue):
1043         (NumberValue.prototype.get value):
1044         (NumberValue.prototype.apply):
1045         (NumberValue.prototype.leftApply):
1046         (NumberValue.prototype.assign):
1047         (NumberArray.):
1048         (NumberArray):
1049         (NumberArray.prototype.apply):
1050         (NumberArray.prototype.leftApply):
1051         (NumberFunction):
1052         (NumberFunction.prototype.apply):
1053         (NumberFunction.prototype.leftApply):
1054         (NativeFunction):
1055         (NativeFunction.prototype.apply):
1056         (NativeFunction.prototype.leftApply):
1057         * Basic/parser.js: Added.
1058         (parse):
1059         (parse.pushToken):
1060         (parse.peekToken):
1061         (parse.consumeKind):
1062         (parse.consumeToken):
1063         (parse.parseVariable):
1064         (parse.parseNumericExpression.parsePrimary):
1065         (parse.parseNumericExpression.parseFactor):
1066         (parse.parseNumericExpression.parseTerm):
1067         (parse.parseNumericExpression):
1068         (parse.parseConstant):
1069         (parse.parseStringExpression):
1070         (parse.isStringExpression):
1071         (parse.parseRelationalExpression):
1072         (parse.parseNonNegativeInteger):
1073         (parse.parseGoToStatement):
1074         (parse.parseGoSubStatement):
1075         (parse.parseStatement):
1076         (parse.parseStatements):
1077         * Basic/random.js: Added.
1078         (createRNG):
1079         (createRNGWithFixedSeed):
1080         (createRNGWithRandomSeed):
1081         * Basic/state.js: Added.
1082         (State):
1083         (State.prototype.getValue):
1084         (State.prototype.getSideState):
1085         (State.prototype.abort):
1086         (State.prototype.validate):
1087         * Basic/test.html: Added.
1088         * Basic/test.js: Added.
1089         * Basic/util.js: Added.
1090         (this.performance.performance.now.currentTime):
1091         (else.this.preciseTime.currentTime):
1092         (else.currentTime):
1093         * Skipped: Make sure that we don't run Basic yet.
1094
1095 2016-06-21  Jon Lee  <jonlee@apple.com>
1096
1097         Update canvas size when benchmark begins
1098         https://bugs.webkit.org/show_bug.cgi?id=159010
1099
1100         Reviewed by Dean Jackson.
1101
1102         Previously we would evaluate the media queries and assign the canvas size to the test
1103         on body.onload. Instead, do it once the user starts the benchmark.
1104
1105         * Animometer/resources/debug-runner/animometer.js: Move benchmarkController.determineCanvasSize()
1106         to benchmarkController._startBenchmark.
1107         * Animometer/resources/runner/animometer.js: Ditto.
1108
1109 2016-06-21  Jon Lee  <jonlee@apple.com>
1110
1111         Add new timestamp option
1112         https://bugs.webkit.org/show_bug.cgi?id=159006
1113
1114         Reviewed by Dean Jackson.
1115
1116         Add a new option to take timestamps from the rAF callback.
1117
1118         * Animometer/developer.html: Added option.
1119         * Animometer/tests/resources/main.js:
1120         (Benchmark): Fall back to using the rAF timestamp if performance.now()
1121         is not available.
1122         (Benchmark._animateLoop): Update how timestamp is set. Prefer to use
1123         local var instead of accessing private var.
1124
1125 2016-06-21  Jon Lee  <jonlee@apple.com>
1126
1127         Improvements to Animometer benchmark
1128         https://bugs.webkit.org/show_bug.cgi?id=157738
1129
1130         Reviewed by Dean Jackson.
1131         Provisionally reviewed by Said Abou-Hallawa.
1132
1133         Update tests.
1134
1135         * Animometer/tests/master/text.html: Ensure only three text sizes for
1136         the three canvases.
1137         * Animometer/tests/master/focus.html: Reduce the text size to fit with smaller
1138         particle sizes.
1139         * Animometer/tests/master/resources/focus.js: Remove the quadratic distribution for
1140         particle sizes, and make it linear. Reduce the size variance. Shuffle the math to
1141         reduce some calculations per frame. Fix the placement of the particles which might
1142         otherwise be culled.
1143         * Animometer/tests/master/resources/image-data.js: Reduce the particle size
1144         to encourage larger scores.
1145
1146 2016-06-21  Jon Lee  <jonlee@apple.com>
1147
1148         Improvements to Animometer benchmark
1149         https://bugs.webkit.org/show_bug.cgi?id=157738
1150
1151         Reviewed by Dean Jackson.
1152         Provisionally reviewed by Said Abou-Hallawa.
1153
1154         Include confidence interval for the final score, and store the canvas
1155         size in the serialization so that it is accurately shown in results.
1156
1157         * Animometer/developer.html: Add a "confidence" div.
1158         * Animometer/index.html: Ditto. Convert "mean" to "confidence".
1159         * Animometer/resources/debug-runner/animometer.js: Look at options, and
1160         if the configuration is included, update the body class based on it
1161         (similar to what we do when we first load the page). That way, if you
1162         drag-and-drop previous results in, that configuration is reflected in
1163         the dashboard. Show the full confidence interval.
1164
1165         * Animometer/resources/debug-runner/animometer.css:
1166         * Animometer/resources/debug-runner/animometer.js: Style update.
1167         * Animometer/resources/runner/animometer.css:
1168
1169         * Animometer/resources/runner/animometer.js:
1170         (_processData): Propagate the confidence interval values out and calculate
1171         the lower and upper bounds. For now, shortcut the aggregate calculation,
1172         since we only go through one iteration.
1173         (this._processData.calculateScore): Propagate the confidence interval out
1174         to be next to the score. Depending on the controller these values are
1175         calculated differently.
1176         (this._processData._getResultsProperty): Convenience function.
1177         (this._processData.get score): Refactor.
1178         (this._processData.get scoreLowerBound): Get the aggregate lower bound.
1179         (this._processData.get scoreUpperBound): Get the aggregate upper bound.
1180         (window.sectionsManager.setSectionScore):
1181         (window.benchmarkController._startBenchmark): When the benchmark starts, note
1182         the canvas size and add it to options. That way it will be automatically be
1183         serialized.
1184         (window.benchmarkController.showResults): Include the maximum deviation
1185         percentage.
1186         * Animometer/resources/runner/lines.svg: Make the background line up with the
1187         skew.
1188         * Animometer/resources/runner/tests.js:
1189         (Headers.details.text): Refactor.
1190         * Animometer/resources/statistics.js:
1191         (Statistics.largestDeviationPercentage): Convenience function to calculate
1192         the deviation percentage on either end and return the largest deviation.
1193         * Animometer/resources/strings.js:
1194
1195         Allow specifying a regression profile to use instead of taking the one
1196         with the lowest error.
1197
1198         Address an issue in the focus test when the regression calculated ends
1199         up overestimating the change point, causing a cascade of tougher
1200         ramps. The reason behind this is that at max complexity of an initial
1201         ramp, the frame length is very high, and it influences the second
1202         segment of the piecewise regression strongly, causing it to be very
1203         steep. As a result, the first segment, expected to be flat, ends up
1204         covering a higher range of complexity. That makes the change point
1205         much higher than it should be. To avoid this, we will add a sanity
1206         check on the maximum value of the ramp. If the regression's projected
1207         value at the maximum complexity of the current ramp is very slow (less
1208         than 20 fps), 1) reduce the maximum complexity by 20%, and 2) do not
1209         include the regression's change point in the change point estimator.
1210         That estimator is used as the midpoint of the next ramp, and including
1211         the change point from a poor regression can bake in the error. The
1212         controller already knows how to adjust for ramps that are too easy for
1213         the system.
1214
1215         * Animometer/resources/runner/animometer.js:
1216         (this._processData.findRegression): Takes a preferred profile and gives that to
1217         Regression.
1218         (this._processData.calculateScore): With the ramp controller, take the profile
1219         of the ramp that was used the most when calculating the ramp's regression. That
1220         profile is what is used for the test's score.
1221         * Animometer/resources/statistics.js:
1222         (Regression.Utilities.createClass): Update to take an options object which can
1223         specify a profile to calculate with. Otherwise it will continue to use both and
1224         select the one with the lower error.
1225         (_calculateRegression): Fix an issue where we claim 0 error if the regression
1226         calculation fails due to divide-by-zero. Instead reject that regression calculation
1227         by giving it Number.MAX_VALUE.
1228         * Animometer/resources/strings.js: New strings for marking a regression as flat
1229         or slope.
1230         * Animometer/tests/resources/main.js:
1231         (RampController): Rename the thresholds for clarity. Add a threshold that, if
1232         exceeded, will lower the maximum complexity of the next ramp.
1233         (tune): Relax the cdf check to consider whether the interval definitely falls in
1234         the desired frame length threshold.
1235         (processSamples): Include the profile in the ramp.
1236
1237         Update ramp controller test. Increase the length of the test to 30 seconds, and extend
1238         the interval to 120 ms during sampling. Improve the estimation of the ramp parameters.
1239
1240         * Animometer/developer.html: Change default to 30 seconds, and don't show the progress bar
1241         by default.
1242         * Animometer/resources/runner/animometer.js: Change default to 30 seconds.
1243         * Animometer/tests/resources/main.js: A number of improvements to the ramp controller, in
1244         the order in which they appear in the patch:
1245
1246         - With a longer test length use longer ramps with longer intervals to get more data at each
1247         complexity. Keep the 100 ms interval length during the ramp up phase since we don't need to
1248         spend more time there to find the right order of magnitude, but increase it during the
1249         ramps to 120 ms.
1250         - The ramp linearly interpolates the complexity to render based on its timestamp, but it would
1251         never sample the minimum complexity. Instead of lerping max to min complexity from time
1252         0 to t where t is the ramp length, instead lerp from 0 to (t - intervalSampleLength) so that
1253         we can have at least one interval sample at the min complexity for that ramp.
1254         - Some regression calculations only come out with one line segment rather than the two
1255         we expect. This could be due to a noisy ramp or the ramp's range is too narrow. If that's the
1256         case, influence the minimum complexity of the next ramp towards the lowest bound of 1, so that
1257         we ensure that at least part of the ramp is covering a complexity range that the system can
1258         handle at full 60.
1259         - Remove an assignment to interpolatedFrameLength since that is never subsequently used.
1260
1261         Update the format used to serialize the results for analysis.
1262
1263         Each data point used to have named properties for fields like complexity and frame rate.
1264         In addition the serialized numbers had rounding errors that took up many characters.
1265         Update the format by introducing a new data container called SampleData, which contains a
1266         field map. The map maps a string to an array index. Each data point is an array, so, to
1267         get a stat, use the field map to get the array index into the data point. This allows future
1268         versions to track other data, and reduces the size of the output string by two-thirds.
1269
1270         * Animometer/resources/extensions.js:
1271         (Utilities.toFixedNumber): Add convenience function that truncates the number to a fixed
1272         precision, and converts it back to a number.
1273         (SampleData): New class that contains sample data and a field map that maps properties to
1274         an array index.
1275         (get length): Number of data points.
1276         (addField): Add a field to the field map.
1277         (push): Add a data point.
1278         (sort): Sort the data.
1279         (slice): Return new SampleData object with sliced data.
1280         (forEach): Iterate over the data with provided function.
1281         (createDatum):
1282         (getFieldInDatum): Returns the data point associated with the field name by looking it up
1283         in the field map in the datum provided, which can be the datum object itself (an array) or
1284         an index into the data member variable.
1285         (setFieldInDatum): Sets the data point associated with the field name.
1286         (at): Returns the data point at the provided index.
1287         (toArray): Serializes the data where the field map serves as property names for each point.
1288
1289         * Animometer/resources/debug-runner/graph.js:
1290         (updateGraphData): Remove unused _testData. Convert the data to the old array format for the
1291         graph to use, since the old format was much easier to work with when displaying the graphs.
1292         (onGraphTypeChanged): For some controllers, no alternative score or mean is provided.
1293         * Animometer/resources/runner/animometer.css:
1294         * Animometer/resources/runner/animometer.js: Refactor to use SampleData. Update JSON output
1295         to only go to 3 digits of precision for purposes of reducing the data size.
1296         * Animometer/resources/strings.js: Add new strings to put into the field maps.
1297         * Animometer/tests/resources/main.js: Refactor to use SampleData.
1298
1299         * Animometer/developer.html:
1300         * Animometer/index.html: Restructure results table for both pages. Add charset attribute to
1301         tests.js include.
1302         * Animometer/resources/debug-runner/animometer.css: Clear out styles from release runner.
1303         * Animometer/resources/debug-runner/graph.js:
1304         (onGraphTypeChanged): Update score and mean if bootstrap results are available from the
1305         controller, since not all controllers do bootstrapping.
1306         * Animometer/resources/debug-runner/tests.js: Update header text.
1307         * Animometer/resources/runner/animometer.css: Include confidence interval in results.
1308         * Animometer/resources/runner/animometer.js:
1309         (ResultsTable._addHeader): Header contents can be HTML, so use innerHTML instead.
1310         (ResultsTable._addBody): Add tbody element.
1311         (ResultsTable._addTest): Allow a data cell to invoke a JS function to get its contents.
1312         (window.benchmarkController.showResults): Add table that includes tests' confidence intervals.
1313         * Animometer/resources/runner/tests.js:
1314         (Headers.details.text): Add new details table that includes bootstrap confidence interval.
1315         The interval can be asymmetric, but for simplicity, report the maximum deviation percentage
1316         on either side of the bootstrap median.
1317         * Animometer/resources/statistics.js:
1318         (bootstrap): Include the confidence percentage in the return object.
1319
1320         Report canvas size in results.
1321
1322         * Animometer/developer.html: Add markup to indicate whether a small, medium, or large
1323         canvas was used.
1324         * Animometer/index.html: Ditto.
1325         * Animometer/resources/debug-runner/animometer.js: Call determineCanvasSize().
1326         * Animometer/resources/runner/animometer.css: Update styles to set the canvas based on the
1327         body class size.
1328         * Animometer/resources/runner/animometer.js:
1329         (window.benchmarkController.initialize): Update styles to set the canvas based on the
1330         body class size.
1331         (window.benchmarkController.determineCanvasSize): Run various media queries and set the body
1332         class based on the size of the device.
1333
1334         * Animometer/developer.html: Refactor to include the main CSS file, and redo
1335         the layout so that it doesn't rely on flexbox.
1336         * Animometer/resources/debug-runner/animometer.css:
1337         * Animometer/resources/debug-runner/animometer.js:
1338         (updateDisplay): Since various parts of the script alter the body class, we can't
1339         replace the className directly. Instead, remove all display-based values and then add
1340         the one that was selected.
1341         * Animometer/resources/debug-runner/graph.js:
1342         (updateGraphData): To set the size of the graph, use window.innerHeight.
1343         * Animometer/resources/runner/animometer.js:
1344         (window.sectionsManager.showSection): Since various parts of the script alter the body
1345         class, we can't replace the className directly. Remove all of the section classes
1346         individually and then add the one desired.
1347         * Animometer/tests/resources/stage.css: Remove -apple-system as a font to use in the
1348         stage.
1349
1350 2016-06-12  Filip Pizlo  <fpizlo@apple.com>
1351
1352         Fix round-down goof in Air.js's ShuffleCustom.forEachArg
1353         https://bugs.webkit.org/show_bug.cgi?id=158674
1354
1355         Reviewed by Michael Saboff.
1356         
1357         x / 3 * 3 is not how you round down to multiples of 3 in JavaScript. You need to do
1358         Math.floor(x / 3) * 3 instead.
1359         
1360         This is a benign change, because having extra arguments to a Shuffle is not something we
1361         actually take advantage of yet. But I think it's best to match the original C++ code's
1362         intent.
1363
1364         * Air.js/custom.js:
1365         (const.ShuffleCustom.forEachArg):
1366
1367 2016-06-11  Filip Pizlo  <fpizlo@apple.com>
1368
1369         Unreviewed, skip Air.js. It got unskipped when I renamed it from JSAir.
1370
1371         * Skipped:
1372
1373 2016-06-10  Filip Pizlo  <fpizlo@apple.com>
1374
1375         Make it easy to package Air.js
1376         https://bugs.webkit.org/show_bug.cgi?id=158652
1377
1378         Reviewed by Benjamin Poulain.
1379         
1380         If you want to give Air.js to someone, you can now do:
1381         
1382         cd PerformanceTests/Air.js
1383         ./make_dist.sh
1384         
1385         This will create Air.js.tar.gz and a directory called Air.js. If you want to send someone
1386         the benchmark, just send either of those things and tell them to look at index.html.
1387         
1388         You may have to edit make_dist.sh for wherever you have Markdown.pl. You can get that from
1389         https://daringfireball.net/projects/markdown/
1390
1391         * Air.js/make_dist.sh: Added.
1392         * Air.js/test.html:
1393
1394 2016-06-10  Filip Pizlo  <fpizlo@apple.com>
1395
1396         Unreviewed, make the link to test.html into an actual link.
1397
1398         * Air.js/README.md:
1399
1400 2016-06-10  Filip Pizlo  <fpizlo@apple.com>
1401
1402         Fix some text in Air.js/README.md
1403         https://bugs.webkit.org/show_bug.cgi?id=158650
1404
1405         Reviewed by Benjamin Poulain.
1406         
1407         I read the text again and found bugs:
1408         
1409         - We never actually say how to run the benchmark. This change adds a blurb about how to run
1410           it.
1411
1412         - We both say that allocateStack is responsible for the bulk of the running time and that
1413           we haven't measured where the bulk of the time is spent. This changes the text to say that
1414           it was a goal to make allocateStack be the hottest part of the benchmark, but that we did
1415           not measure this.
1416
1417         * Air.js/README.md:
1418
1419 2016-06-10  Filip Pizlo  <fpizlo@apple.com>
1420
1421         Air.js should have some documentation
1422         https://bugs.webkit.org/show_bug.cgi?id=158648
1423
1424         Reviewed by Keith Miller.
1425         
1426         I want to be able to point people at a document if they want to know more about this
1427         benchmark.
1428
1429         * Air.js/README.md: Added.
1430
1431 2016-06-10  Filip Pizlo  <fpizlo@apple.com>
1432
1433         Rename JSAir to Air.js.
1434
1435         Rubber stamped by Mark Lam.
1436
1437         * Air.js: Copied from PerformanceTests/JSAir.
1438         * Air.js/airjs-tests.yaml: Copied from PerformanceTests/JSAir/jsair-tests.yaml.
1439         * Air.js/benchmark.js: Copied from PerformanceTests/JSAir/benchmark.js.
1440         (Benchmark):
1441         * Air.js/jsair-tests.yaml: Removed.
1442         * Air.js/payload-airjs-ACLj8C.js: Copied from PerformanceTests/JSAir/payload-jsair-ACLj8C.js.
1443         * Air.js/payload-jsair-ACLj8C.js: Removed.
1444         * Air.js/stress-test.js: Copied from PerformanceTests/JSAir/stress-test.js.
1445         * Air.js/test.html: Copied from PerformanceTests/JSAir/test.html.
1446         * Air.js/test.js: Copied from PerformanceTests/JSAir/test.js.
1447         * JSAir: Removed.
1448         * JSAir/all.js: Removed.
1449         * JSAir/allocate_stack.js: Removed.
1450         * JSAir/arg.js: Removed.
1451         * JSAir/basic_block.js: Removed.
1452         * JSAir/benchmark.js: Removed.
1453         * JSAir/code.js: Removed.
1454         * JSAir/custom.js: Removed.
1455         * JSAir/frequented_block.js: Removed.
1456         * JSAir/insertion_set.js: Removed.
1457         * JSAir/inst.js: Removed.
1458         * JSAir/jsair-tests.yaml: Removed.
1459         * JSAir/liveness.js: Removed.
1460         * JSAir/opcode.js: Removed.
1461         * JSAir/payload-gbemu-executeIteration.js: Removed.
1462         * JSAir/payload-imaging-gaussian-blur-gaussianBlur.js: Removed.
1463         * JSAir/payload-jsair-ACLj8C.js: Removed.
1464         * JSAir/payload-typescript-scanIdentifier.js: Removed.
1465         * JSAir/reg.js: Removed.
1466         * JSAir/stack_slot.js: Removed.
1467         * JSAir/stress-test.js: Removed.
1468         * JSAir/strip-hash.rb: Removed.
1469         * JSAir/symbols.js: Removed.
1470         * JSAir/test.html: Removed.
1471         * JSAir/test.js: Removed.
1472         * JSAir/tmp.js: Removed.
1473         * JSAir/tmp_base.js: Removed.
1474         * JSAir/util.js: Removed.
1475
1476 2016-06-10  Filip Pizlo  <fpizlo@apple.com>
1477
1478         JSC Stress Test failing: jsair-tests.yaml/test.js.ftl-eager-no-cjit
1479         https://bugs.webkit.org/show_bug.cgi?id=158571
1480
1481         Reviewed by Keith Miller.
1482         
1483         Introduce a different harness for run-jsc-stress-tests, which runs for a minimum of 10
1484         iterations and then tries to do any number of "bonus" iterations until it's been running for
1485         two seconds. Since this is the sort of test that isn't really meant to stress anything in
1486         particular, I think it's OK if it is time-limited in this way. The worst case is that some
1487         of its failures will be flaky, but I think that they would have been flaky anyway given the
1488         complexity of the test.
1489
1490         * JSAir/benchmark.js:
1491         (Benchmark):
1492         (Benchmark.prototype.runIteration):
1493         (benchmark): Deleted.
1494         * JSAir/jsair-tests.yaml:
1495         * JSAir/stress-test.js: Added.
1496         (preciseTime):
1497         * JSAir/test.html:
1498         * JSAir/test.js:
1499
1500 2016-06-09  Filip Pizlo  <fpizlo@apple.com>
1501
1502         Unreviewed, teach the perf bots not to run JSAir.
1503
1504         * Skipped:
1505
1506 2016-06-08  Filip Pizlo  <fpizlo@apple.com>
1507
1508         Use more ES6 features in JSAir
1509         https://bugs.webkit.org/show_bug.cgi?id=158497
1510
1511         Reviewed by Keith Miller.
1512         
1513         This improves JSAir with the following ES6 features suggested by JoePeck:
1514         
1515         - String interpolation.
1516         - Destructuring inside PatchCustom.
1517         - Default arguments.
1518         
1519         All of these things are on hot paths.
1520         
1521         Note that I didn't use string interpolation everywhere that I could, only in those places
1522         where it made the code more readable. In Ruby, I used the style that if the interpolation
1523         expression has any non-trivial stuff (like a ternary operator, a chain of calls, or embedded
1524         strings) then it's better to use regular strcat. I think that's what I carried over to here.
1525         
1526         Note that the previous change (Add result validation to JSAir) also made the Proxy code not
1527         dead, though it's not necessarily on the hot path. The Proxy isn't called into frequently
1528         but it's used from a function that is otherwise hot, so if calling into the Proxy prevents
1529         that function from being optimized then it will hurt so good.
1530         
1531         I also reenabled tail calls in a few places.
1532         
1533         This change doesn't seem to change the performance of the benchmark for us. That's expected
1534         since these ES6 features are cheap. Note that this claim doesn't include Proxy, which was
1535         added in a separate change and that change did make the benchmark overall more expensive.
1536
1537         * JSAir/allocate_stack.js:
1538         (allocateStack):
1539         * JSAir/arg.js:
1540         (Arg.createBitImm64):
1541         (Arg.createAddr):
1542         (Arg.createStack):
1543         (Arg.logScale):
1544         (Arg.createIndex):
1545         * JSAir/basic_block.js:
1546         (BasicBlock.get headerString):
1547         (BasicBlock.prototype.get if):
1548         (BasicBlock):
1549         * JSAir/benchmark.js:
1550         (benchmark):
1551         * JSAir/code.js:
1552         (Code):
1553         (Code.prototype.addBlock):
1554         (Code.prototype.addStackSlot):
1555         (Code.prototype.newTmp):
1556         (Code.prototype.get size):
1557         (Code.prototype.get blocks):
1558         (Code.prototype.get stackSlots):
1559         (Code.prototype.tmps):
1560         (Code.prototype.get callArgAreaSize):
1561         (Code.prototype.toString):
1562         * JSAir/custom.js:
1563         (const.PatchCustom.forEachArg):
1564         * JSAir/inst.js:
1565         (Inst):
1566         * JSAir/reg.js:
1567         (Reg.prototype.toString):
1568         * JSAir/util.js:
1569         (symbolName):
1570         (lowerSymbolName):
1571
1572 2016-06-07  Filip Pizlo  <fpizlo@apple.com>
1573
1574         Add result validation to JSAir
1575         https://bugs.webkit.org/show_bug.cgi?id=158493
1576
1577         Reviewed by Saam Barati.
1578         
1579         This adds the ability to hash a Code in a way that matches the C++ code's hashing of Code.
1580         This allows us to check if the Code that JSAir sees is the code that C++ saw. We use this to
1581         check the Code before and after allocateStack, and compare against hashes we got from C++.
1582         
1583         Doing this uncovered bugs. roundUpToMultipleOf wasn't doing anything. allocateStack was not
1584         allocating things correctly because I was concatting a Set to an Array, which doesn't really
1585         work. Now these bugs are fixed.
1586         
1587         The checking step adds to the running time so I reduced the number of iterations. The
1588         benchmark spends a decent amount of its time computing Code hashes; I think it's around 1/3
1589         total. This is probably OK. It's better to verify the results even if the running time is
1590         not all in the "core" of the algorithm.
1591         
1592         Also add a run-jsc-stress-tests yaml file to allow this to run as a test.
1593
1594         * JSAir/allocate_stack.js:
1595         * JSAir/arg.js:
1596         (Arg.createImm):
1597         (Arg.createBigImm):
1598         (Arg.createBitImm):
1599         (Arg.createBitImm64):
1600         (Arg.createWidth):
1601         (Arg.createSpecial):
1602         (Arg.prototype.get kind):
1603         (Arg.prototype.get isTmp):
1604         (Arg.prototype.get isImm):
1605         (Arg.prototype.get isSomeImm):
1606         (Arg.prototype.get isSomeBigImm):
1607         (Arg.prototype.get isCondition):
1608         (Arg.prototype.get isWidth):
1609         (Arg.prototype.get isSpecial):
1610         (Arg.prototype.get isAlive):
1611         (Arg.prototype.get tmp):
1612         (Arg.prototype.get value):
1613         (Arg.prototype.get lowValue):
1614         (Arg.prototype.get highValue):
1615         (Arg.prototype.get base):
1616         (Arg.prototype.get isGP):
1617         (Arg.prototype.get isFP):
1618         (Arg.prototype.isValidForm):
1619         (Arg.prototype.get isInvertible):
1620         (Arg.kindCode):
1621         (Arg.prototype.hash):
1622         (Arg.prototype.toString):
1623         (Arg):
1624         * JSAir/basic_block.js:
1625         (BasicBlock.get successorBlocks):
1626         * JSAir/benchmark.js:
1627         (benchmark):
1628         * JSAir/code.js:
1629         (Code.prototype.setFrameSize):
1630         (Code.prototype.hash):
1631         (Code.prototype.toString):
1632         (Code):
1633         * JSAir/inst.js:
1634         (Inst.prototype.get hasNonArgEffects):
1635         (Inst.prototype.hash):
1636         (Inst.prototype.toString):
1637         (Inst):
1638         * JSAir/jsair-tests.yaml: Added.
1639         * JSAir/opcode.js:
1640         (Inst_forEachArg):
1641         (Inst_hasNonArgEffects):
1642         (opcodeCode):
1643         * JSAir/payload-gbemu-executeIteration.js:
1644         (createPayloadGbemuExecuteIteration):
1645         * JSAir/payload-imaging-gaussian-blur-gaussianBlur.js:
1646         (createPayloadImagingGaussianBlurGaussianBlur):
1647         * JSAir/payload-jsair-ACLj8C.js:
1648         (createPayloadJSAirACLj8C):
1649         * JSAir/payload-typescript-scanIdentifier.js:
1650         (createPayloadTypescriptScanIdentifier):
1651         * JSAir/reg.js:
1652         (Reg.prototype.get isReg):
1653         (Reg.prototype.hash):
1654         (Reg.prototype.toString):
1655         * JSAir/stack_slot.js:
1656         (StackSlot.prototype.setOffsetFromFP):
1657         (StackSlot.prototype.hash):
1658         (StackSlot.prototype.toString):
1659         * JSAir/symbols.js:
1660         (relCondCode):
1661         (resCondCode):
1662         (doubleCondCode):
1663         * JSAir/test.html:
1664         * JSAir/tmp.js:
1665         (Tmp.prototype.get isReg):
1666         (Tmp.prototype.hash):
1667         (Tmp.prototype.toString):
1668         * JSAir/util.js:
1669         (roundUpToMultipleOf):
1670         (symbolName):
1671
1672 2016-06-07  Filip Pizlo  <fpizlo@apple.com>
1673
1674         Implement Air::allocateStack() in ES6 to see how much of a bad idea that is
1675         https://bugs.webkit.org/show_bug.cgi?id=158318
1676
1677         Reviewed by Saam Barati.
1678         
1679         This adds a new benchmark for us to play with called JSAir. It's a complete ES6
1680         implementation of Air's allocateStack() phase along with all of Air needed to run that
1681         phase. This includes things like stack slots, registers, temporaries, basic blocks,
1682         instructions, and all of the code for iterating over, inspecting, and modifying those
1683         things.
1684         
1685         To make this work, JSC can now dump Air just before allocateStack() in the form of JS code
1686         that creates a Code object that matches exactly what C++ Air saw. This benchmark comprises
1687         four Air IRs:
1688         
1689         - Octane/gbemu's largest function, executeIteration.
1690         - Kraken/imaging-gaussian-blur's largest function in OSR entry mode, gaussuanBlur.
1691         - Octane/typescript's largest function that is actually hot, scanIdentifier.
1692         - JSAir's largest hot function, which is anonymous, so we call it by its hash (ACLj8C).
1693         
1694         This runs in about 2 seconds on my machine in JSC trunk. It includes both a commandline
1695         harness and a web harness. JSAir is almost exactly 100x slower in ES6 in WebKit than the
1696         C++ Air::allocateStack() phase on which it is based.
1697         
1698         JSAir uses the following ES6 features:
1699         
1700         - Symbol.
1701         - for-of.
1702         - arrow functions.
1703         - Map/Set.
1704         - let/const.
1705         - classes.
1706         
1707         All of these things are used in anger and should end up on the hot path. There is also code
1708         that uses Proxies, but it ends up being dead.
1709         
1710         We can improve this even more:
1711         I still need to add result validation: https://bugs.webkit.org/show_bug.cgi?id=158493
1712         I want to make it use more ES6 features: https://bugs.webkit.org/show_bug.cgi?id=158497
1713         
1714         * JSAir: Added.
1715         * JSAir/all.js: Added.
1716         * JSAir/allocate_stack.js: Added.
1717         (allocateStack.attemptAssignment):
1718         (allocateStack.assign):
1719         (allocateStack.interfere):
1720         (allocateStack.):
1721         (allocateStack):
1722         * JSAir/arg.js: Added.
1723         (Arg):
1724         (Arg.isAnyUse):
1725         (Arg.isColdUse):
1726         (Arg.isWarmUse):
1727         (Arg.cooled):
1728         (Arg.isEarlyUse):
1729         (Arg.isLateUse):
1730         (Arg.isAnyDef):
1731         (Arg.isEarlyDef):
1732         (Arg.isLateDef):
1733         (Arg.isZDef):
1734         (Arg.typeForB3Type):
1735         (Arg.widthForB3Type):
1736         (Arg.conservativeWidth):
1737         (Arg.minimumWidth):
1738         (Arg.bytes):
1739         (Arg.widthForBytes):
1740         (Arg.createTmp):
1741         (Arg.fromReg):
1742         (Arg.createImm):
1743         (Arg.createBigImm):
1744         (Arg.createBitImm):
1745         (Arg.createBitImm64):
1746         (Arg.createAddr):
1747         (Arg.createStack):
1748         (Arg.createCallArg):
1749         (Arg.createStackAddr):
1750         (Arg.isValidScale):
1751         (Arg.logScale):
1752         (Arg.createIndex):
1753         (Arg.createRelCond):
1754         (Arg.createResCond):
1755         (Arg.createDoubleCond):
1756         (Arg.createWidth):
1757         (Arg.prototype.get kind):
1758         (Arg.prototype.get isTmp):
1759         (Arg.prototype.get isImm):
1760         (Arg.prototype.get isBigImm):
1761         (Arg.prototype.get isBitImm):
1762         (Arg.prototype.get isBitImm64):
1763         (Arg.prototype.get isSomeImm):
1764         (Arg.prototype.get isAddr):
1765         (Arg.prototype.get isStack):
1766         (Arg.prototype.get isCallArg):
1767         (Arg.prototype.get isIndex):
1768         (Arg.prototype.get isMemory):
1769         (Arg.prototype.get isStackMemory):
1770         (Arg.prototype.get isRelCond):
1771         (Arg.prototype.get isResCond):
1772         (Arg.prototype.get isDoubleCond):
1773         (Arg.prototype.get isCondition):
1774         (Arg.prototype.get isWidth):
1775         (Arg.prototype.get isAlive):
1776         (Arg.prototype.get tmp):
1777         (Arg.prototype.get value):
1778         (Arg.prototype.get base):
1779         (Arg.prototype.get hasOffset):
1780         (Arg.prototype.get offset):
1781         (Arg.prototype.get stackSlot):
1782         (Arg.prototype.get index):
1783         (Arg.prototype.get scale):
1784         (Arg.prototype.get logScale):
1785         (Arg.prototype.get width):
1786         (Arg.prototype.get isGPTmp):
1787         (Arg.prototype.get isFPTmp):
1788         (Arg.prototype.get isGP):
1789         (Arg.prototype.get isFP):
1790         (Arg.prototype.get hasType):
1791         (Arg.prototype.get type):
1792         (Arg.prototype.isType):
1793         (Arg.prototype.isCompatibleType):
1794         (Arg.prototype.get isGPR):
1795         (Arg.prototype.get gpr):
1796         (Arg.prototype.get isFPR):
1797         (Arg.prototype.get fpr):
1798         (Arg.prototype.get isReg):
1799         (Arg.prototype.get reg):
1800         (Arg.isValidImmForm):
1801         (Arg.isValidBitImmForm):
1802         (Arg.isValidBitImm64Form):
1803         (Arg.isValidAddrForm):
1804         (Arg.isValidIndexForm):
1805         (Arg.prototype.isValidForm):
1806         (Arg.prototype.forEachTmpFast):
1807         (Arg.prototype.usesTmp):
1808         (Arg.prototype.forEachTmp):
1809         (Arg.prototype.is):
1810         (Arg.prototype.as):
1811         (Arg.prototype.forEachFast):
1812         (Arg.prototype.forEach):
1813         (Arg.extract):
1814         (Arg.forEachFast):
1815         (Arg.forEach):
1816         (Arg.prototype.get condition):
1817         (Arg.prototype.get isInvertible):
1818         (Arg.prototype.toString):
1819         * JSAir/basic_block.js: Added.
1820         (BasicBlock):
1821         (BasicBlock.prototype.get index):
1822         (BasicBlock.prototype.get size):
1823         (BasicBlock.prototype.Symbol.iterator):
1824         (BasicBlock.prototype.at):
1825         (BasicBlock.get last):
1826         (BasicBlock.get insts):
1827         (BasicBlock.get numSuccessors):
1828         (BasicBlock.get successors):
1829         (BasicBlock.get successorBlocks.):
1830         (BasicBlock.get successorBlocks):
1831         (BasicBlock.set get numPredecessors):
1832         (BasicBlock.get predecessors):
1833         (BasicBlock.get frequency):
1834         (BasicBlock.get headerString):
1835         (BasicBlock.get footerString):
1836         * JSAir/benchmark.js: Added.
1837         (benchmark):
1838         * JSAir/code.js: Added.
1839         (Code):
1840         (Code.prototype.addBlock):
1841         (Code.prototype.addStackSlot):
1842         (Code.prototype.newTmp):
1843         (Code.prototype.get size):
1844         (Code.prototype.at):
1845         (Code.prototype.Symbol.iterator):
1846         (Code.prototype.get blocks):
1847         (Code.prototype.get stackSlots):
1848         (Code.prototype.tmps):
1849         (Code.prototype.get callArgAreaSize):
1850         (Code.prototype.requestCallArgAreaSize):
1851         (Code.prototype.get frameSize):
1852         (Code.prototype.setFrameSize):
1853         (Code.prototype.toString):
1854         * JSAir/custom.js: Added.
1855         (const.ShuffleCustom.forEachArg):
1856         (const.ShuffleCustom.hasNonArgNonControlEffects):
1857         (const.PatchCustom.forEachArg):
1858         (const.PatchCustom.hasNonArgNonControlEffects):
1859         (const.CCallCustom.forEachArg):
1860         (const.CCallCustom.hasNonArgNonControlEffects):
1861         (const.ColdCCallCustom.forEachArg):
1862         (const.ColdCCallCustom.hasNonArgNonControlEffects):
1863         * JSAir/frequented_block.js: Added.
1864         (FrequentedBlock):
1865         (FrequentedBlock.prototype.toString):
1866         * JSAir/insertion_set.js: Added.
1867         (Insertion):
1868         (Insertion.prototype.get index):
1869         (Insertion.prototype.get element):
1870         (Insertion.prototype.lessThan):
1871         (InsertionSet):
1872         (InsertionSet.prototype.appendInsertion):
1873         (InsertionSet.prototype.append):
1874         (InsertionSet.prototype.execute):
1875         * JSAir/inst.js: Added.
1876         (Inst):
1877         (Inst.prototype.append):
1878         (Inst.prototype.clear):
1879         (Inst.prototype.get opcode):
1880         (Inst.prototype.get args):
1881         (Inst.prototype.visitArg):
1882         (Inst.prototype.forEachTmpFast):
1883         (Inst.prototype.forEachArg):
1884         (Inst.prototype.forEachTmp):
1885         (Inst.prototype.forEach):
1886         (Inst.forEachDef):
1887         (Inst.forEachDefWithExtraClobberedRegs):
1888         (Inst.prototype.get hasNonArgEffects):
1889         (Inst.prototype.toString):
1890         * JSAir/liveness.js: Added.
1891         (Liveness):
1892         (Liveness.prototype.get thing):
1893         (Liveness.prototype.get code):
1894         (Liveness.prototype.get liveAtHead):
1895         (Liveness.prototype.get liveAtTail):
1896         (Liveness.prototype.localCalc.LocalCalc):
1897         (Liveness.prototype.localCalc.LocalCalc.prototype.get liveSet):
1898         (Liveness.prototype.localCalc.LocalCalc.prototype.execute):
1899         (Liveness.prototype.localCalc):
1900         * JSAir/opcode.js: Added.
1901         (Inst_forEachArg):
1902         (Inst_hasNonArgEffects):
1903         * JSAir/payload-gbemu-executeIteration.js: Added.
1904         (createPayloadGbemuExecuteIteration):
1905         * JSAir/payload-imaging-gaussian-blur-gaussianBlur.js: Added.
1906         (createPayloadImagingGaussianBlurGaussianBlur):
1907         * JSAir/payload-jsair-ACLj8C.js: Added.
1908         (createPayloadJSAirACLj8C):
1909         * JSAir/payload-typescript-scanIdentifier.js: Added.
1910         (createPayloadTypescriptScanIdentifier):
1911         * JSAir/reg.js: Added.
1912         (Reg):
1913         (Reg.fromReg):
1914         (Reg.prototype.get index):
1915         (Reg.prototype.get type):
1916         (Reg.prototype.get name):
1917         (Reg.prototype.get isCalleeSave):
1918         (Reg.prototype.get isReg):
1919         (Reg.prototype.toString):
1920         (Reg.extract):
1921         (Reg.forEachFast):
1922         (Reg.forEach):
1923         (newGPR):
1924         (Reg.gprs.Reg.fprs.Reg.calleeSaveGPRs.Reg.calleeSaveFPRs.Reg.calleeSaves):
1925         * JSAir/stack_slot.js: Added.
1926         (StackSlot):
1927         (StackSlot.prototype.get byteSize):
1928         (StackSlot.prototype.get kind):
1929         (StackSlot.prototype.get isLocked):
1930         (StackSlot.prototype.get isSpill):
1931         (StackSlot.prototype.get index):
1932         (StackSlot.prototype.ensureSize):
1933         (StackSlot.prototype.get alignment):
1934         (StackSlot.prototype.get offsetFromFP):
1935         (StackSlot.prototype.setOffsetFromFP):
1936         (StackSlot.prototype.toString):
1937         (StackSlot.extract):
1938         (StackSlot.forEachFast):
1939         (StackSlot.forEach):
1940         * JSAir/symbols.js: Added.
1941         * JSAir/test.html: Added.
1942         * JSAir/test.js: Added.
1943         * JSAir/tmp.js: Added.
1944         (Tmp):
1945         (Tmp.fromReg):
1946         (Tmp.prototype.get index):
1947         (Tmp.prototype.get type):
1948         (Tmp.prototype.get isReg):
1949         (Tmp.prototype.toString):
1950         (Tmp.extract):
1951         (Tmp.forEachFast):
1952         (Tmp.forEach):
1953         * JSAir/tmp_base.js: Added.
1954         (TmpBase.prototype.get isGP):
1955         (TmpBase.prototype.get isFP):
1956         (TmpBase.prototype.get isGPR):
1957         (TmpBase.prototype.get isFPR):
1958         (TmpBase.prototype.get reg):
1959         (TmpBase.prototype.get gpr):
1960         (TmpBase.prototype.get fpr):
1961         (TmpBase):
1962         * JSAir/util.js: Added.
1963         (isRepresentableAsInt32):
1964         (addIndexed):
1965         (roundUpToMultipleOf):
1966         (symbolName):
1967         (mergeIntoSet):
1968         (nonEmptyRangesOverlap):
1969         (rangesOverlap):
1970         (removeAllMatching):
1971         (swap):
1972         (bubble):
1973         (bubbleSort):
1974         (currentTime):
1975         (else.currentTime):
1976
1977 2016-05-26  Yusuke Suzuki  <utatane.tea@gmail.com>
1978
1979         [JSC] Allow JSBench to use precise time
1980         https://bugs.webkit.org/show_bug.cgi?id=158050
1981
1982         Reviewed by Geoffrey Garen.
1983
1984         * JSBench/amazon-chrome-win/urem.js:
1985         (else.window.performance.window.performance.now.currentTimeInMS):
1986         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1987         (else.else.currentTimeInMS):
1988         (onload.cb):
1989         (onload):
1990         * JSBench/amazon-chrome/urem.js:
1991         (else.window.performance.window.performance.now.currentTimeInMS):
1992         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1993         (else.else.currentTimeInMS):
1994         (onload.cb):
1995         (onload):
1996         * JSBench/amazon-firefox-win/urm.js:
1997         (else.window.performance.window.performance.now.currentTimeInMS):
1998         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1999         (else.else.currentTimeInMS):
2000         (onload.cb):
2001         (onload):
2002         * JSBench/amazon-firefox/urm.js:
2003         (else.window.performance.window.performance.now.currentTimeInMS):
2004         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2005         (else.else.currentTimeInMS):
2006         (onload.cb):
2007         (onload):
2008         * JSBench/amazon-safari/urem.js:
2009         (else.window.performance.window.performance.now.currentTimeInMS):
2010         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2011         (else.else.currentTimeInMS):
2012         (onload.cb):
2013         (onload):
2014         * JSBench/facebook-chrome-win/urem.js:
2015         (else.window.performance.window.performance.now.currentTimeInMS):
2016         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2017         (else.else.currentTimeInMS):
2018         (onload.cb):
2019         (onload):
2020         * JSBench/facebook-chrome/urem.js:
2021         (else.window.performance.window.performance.now.currentTimeInMS):
2022         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2023         (else.else.currentTimeInMS):
2024         (onload.cb):
2025         (onload):
2026         * JSBench/facebook-firefox-win/urem.js:
2027         (else.window.performance.window.performance.now.currentTimeInMS):
2028         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2029         (else.else.currentTimeInMS):
2030         (onload.cb):
2031         (onload):
2032         * JSBench/facebook-firefox/urem.js:
2033         (else.window.performance.window.performance.now.currentTimeInMS):
2034         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2035         (else.else.currentTimeInMS):
2036         (onload.cb):
2037         (onload):
2038         * JSBench/facebook-safari/urem.js:
2039         (else.window.performance.window.performance.now.currentTimeInMS):
2040         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2041         (else.else.currentTimeInMS):
2042         (onload.cb):
2043         (onload):
2044         * JSBench/google-chrome-win/urem.js:
2045         (else.window.performance.window.performance.now.currentTimeInMS):
2046         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2047         (else.else.currentTimeInMS):
2048         (onload.cb):
2049         (onload):
2050         * JSBench/google-chrome/urem.js:
2051         (else.window.performance.window.performance.now.currentTimeInMS):
2052         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2053         (else.else.currentTimeInMS):
2054         (onload.cb):
2055         (onload):
2056         * JSBench/google-firefox-win/urem.js:
2057         (else.window.performance.window.performance.now.currentTimeInMS):
2058         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2059         (else.else.currentTimeInMS):
2060         (onload.cb):
2061         (onload):
2062         * JSBench/google-firefox/uem.js:
2063         (else.window.performance.window.performance.now.currentTimeInMS):
2064         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2065         (else.else.currentTimeInMS):
2066         (onload.cb):
2067         (onload):
2068         * JSBench/google-safari/urem.js:
2069         (else.window.performance.window.performance.now.currentTimeInMS):
2070         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2071         (else.else.currentTimeInMS):
2072         (onload.cb):
2073         (onload):
2074         * JSBench/harness.js:
2075         (runBenchmark.window.currentTimeInMS):
2076         (runBenchmark.else.window.currentTimeInMS):
2077         * JSBench/twitter-chrome-win/rem.js:
2078         (else.window.performance.window.performance.now.currentTimeInMS):
2079         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2080         (else.else.currentTimeInMS):
2081         (onload.cb):
2082         (onload):
2083         * JSBench/twitter-chrome/urem.js:
2084         (else.window.performance.window.performance.now.currentTimeInMS):
2085         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2086         (else.else.currentTimeInMS):
2087         (onload.cb):
2088         (onload):
2089         * JSBench/twitter-firefox-win/urem.js:
2090         (else.window.performance.window.performance.now.currentTimeInMS):
2091         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2092         (else.else.currentTimeInMS):
2093         (onload.cb):
2094         (onload):
2095         * JSBench/twitter-firefox/urem.js:
2096         (else.window.performance.window.performance.now.currentTimeInMS):
2097         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2098         (else.else.currentTimeInMS):
2099         (onload.cb):
2100         (onload):
2101         * JSBench/twitter-safari/urem.js:
2102         (else.window.performance.window.performance.now.currentTimeInMS):
2103         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2104         (else.else.currentTimeInMS):
2105         (onload.cb):
2106         (onload):
2107         * JSBench/yahoo-chrome-win/urem.js:
2108         (else.window.performance.window.performance.now.currentTimeInMS):
2109         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2110         (else.else.currentTimeInMS):
2111         (onload.cb):
2112         (onload):
2113         * JSBench/yahoo-chrome/urem.js:
2114         (else.window.performance.window.performance.now.currentTimeInMS):
2115         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2116         (else.else.currentTimeInMS):
2117         (onload.cb):
2118         (onload):
2119         * JSBench/yahoo-firefox-win/urem.js:
2120         (else.window.performance.window.performance.now.currentTimeInMS):
2121         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2122         (else.else.currentTimeInMS):
2123         (onload.cb):
2124         (onload):
2125         * JSBench/yahoo-firefox/urem.js:
2126         (else.window.performance.window.performance.now.currentTimeInMS):
2127         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2128         (else.else.currentTimeInMS):
2129         (onload.cb):
2130         (onload):
2131         * JSBench/yahoo-safari/urem.js:
2132         (else.window.performance.window.performance.now.currentTimeInMS):
2133         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
2134         (else.else.currentTimeInMS):
2135         (onload.cb):
2136         (onload):
2137
2138 2016-05-25  Geoffrey Garen  <ggaren@apple.com>
2139
2140         replaceable own properties seem to ignore replacement after property caching
2141         https://bugs.webkit.org/show_bug.cgi?id=158091
2142
2143         Reviewed by Darin Adler.
2144
2145         * MallocBench/MallocBench.xcodeproj/project.pbxproj:
2146         * MallocBench/MallocBench/Benchmark.cpp:
2147         * MallocBench/MallocBench/Interpreter.cpp:
2148         (Interpreter::doMallocOp):
2149         * MallocBench/MallocBench/Interpreter.h:
2150         * MallocBench/MallocBench/fastMallocLog.63316.ops: Added.
2151         * MallocBench/MallocBench/jetstream.cpp: Added.
2152         (benchmark_jetstream):
2153         * MallocBench/MallocBench/jetstream.h: Added.
2154
2155 2016-05-25  Keith Miller  <keith_miller@apple.com>
2156
2157         Unreviewed, add JSBench to the skipped list for now since it doesn't
2158         work currently.
2159
2160         * Skipped:
2161
2162 2016-05-17  Keith Miller  <keith_miller@apple.com>
2163         We should have JSBench in PerformanceTests
2164         https://bugs.webkit.org/show_bug.cgi?id=157952
2165
2166         Rubber-stamped by Saam Barati.
2167
2168         There are some slight changes to the layout of the test directory
2169         to make it work nicely with run-jsc-benchmarks. Before JSBench
2170         had each of the browser specific sub-tests in a sub-directory.
2171         These have been flattened e.g. amazon/safari/ has become
2172         amazon-safari/.
2173
2174         * JSBench/amazon-chrome-win/urem.html: Added.
2175         * JSBench/amazon-chrome-win/urem.js: Added.
2176         * JSBench/amazon-chrome/urem.html: Added.
2177         * JSBench/amazon-chrome/urem.js: Added.
2178         * JSBench/amazon-firefox-win/urm.html: Added.
2179         * JSBench/amazon-firefox-win/urm.js: Added.
2180         * JSBench/amazon-firefox/urm.html: Added.
2181         * JSBench/amazon-firefox/urm.js: Added.
2182         * JSBench/amazon-safari/urem.html: Added.
2183         * JSBench/amazon-safari/urem.js: Added.
2184         * JSBench/browsercheck.js: Added.
2185         * JSBench/facebook-chrome-win/urem.html: Added.
2186         * JSBench/facebook-chrome-win/urem.js: Added.
2187         * JSBench/facebook-chrome/urem.html: Added.
2188         * JSBench/facebook-chrome/urem.js: Added.
2189         * JSBench/facebook-firefox-win/urem.html: Added.
2190         * JSBench/facebook-firefox-win/urem.js: Added.
2191         * JSBench/facebook-firefox/urem.html: Added.
2192         * JSBench/facebook-firefox/urem.js: Added.
2193         * JSBench/facebook-safari/urem.html: Added.
2194         * JSBench/facebook-safari/urem.js: Added.
2195         * JSBench/google-chrome-win/urem.html: Added.
2196         * JSBench/google-chrome-win/urem.js: Added.
2197         * JSBench/google-chrome/urem.html: Added.
2198         * JSBench/google-chrome/urem.js: Added.
2199         * JSBench/google-firefox-win/urem.html: Added.
2200         * JSBench/google-firefox-win/urem.js: Added.
2201         * JSBench/google-firefox/uem.html: Added.
2202         * JSBench/google-firefox/uem.js: Added.
2203         * JSBench/google-safari/urem.html: Added.
2204         * JSBench/google-safari/urem.js: Added.
2205         * JSBench/harness.html: Added.
2206         * JSBench/harness.js: Added.
2207         * JSBench/harness.py: Added.
2208         * JSBench/index.html: Added.
2209         * JSBench/reload.html: Added.
2210         * JSBench/twitter-chrome-win/rem.html: Added.
2211         * JSBench/twitter-chrome-win/rem.js: Added.
2212         * JSBench/twitter-chrome/urem.html: Added.
2213         * JSBench/twitter-chrome/urem.js: Added.
2214         * JSBench/twitter-firefox-win/urem.html: Added.
2215         * JSBench/twitter-firefox-win/urem.js: Added.
2216         * JSBench/twitter-firefox/urem.html: Added.
2217         * JSBench/twitter-firefox/urem.js: Added.
2218         * JSBench/twitter-safari/urem.html: Added.
2219         * JSBench/twitter-safari/urem.js: Added.
2220         * JSBench/yahoo-chrome-win/urem.html: Added.
2221         * JSBench/yahoo-chrome-win/urem.js: Added.
2222         * JSBench/yahoo-chrome/urem.html: Added.
2223         * JSBench/yahoo-chrome/urem.js: Added.
2224         * JSBench/yahoo-firefox-win/urem.html: Added.
2225         * JSBench/yahoo-firefox-win/urem.js: Added.
2226         * JSBench/yahoo-firefox/urem.html: Added.
2227         * JSBench/yahoo-firefox/urem.js: Added.
2228         * JSBench/yahoo-safari/urem.html: Added.
2229         * JSBench/yahoo-safari/urem.js: Added.
2230
2231 2016-05-20  Ryosuke Niwa  <rniwa@webkit.org>
2232
2233         run-benchmark's results should contain Animometer's debug output
2234         https://bugs.webkit.org/show_bug.cgi?id=157941
2235
2236         Reviewed by Stephanie Lewis.
2237
2238         Made developer.html support the JSON generated by run-benchmark which stores everything under debugOutput.
2239
2240         * Animometer/resources/debug-runner/animometer.js:
2241         (Utilities.initialize): Unwrap debugOutput in the case run-benchmark's result JSON is used.
2242
2243 2016-05-18  Timothy Hatcher  <timothy@apple.com>
2244
2245         Make Animometer work in all browsers
2246         https://bugs.webkit.org/show_bug.cgi?id=157855
2247         rdar://problem/26338521
2248
2249         Reviewed by Jon Lee.
2250
2251         * Animometer/resources/runner/logo.svg: Use a symbol instead
2252         of putting the id on the root svg. This works in all browsers.
2253
2254 2016-05-07  Timothy Hatcher  <timothy@apple.com>
2255
2256         Update Animometer design
2257         https://bugs.webkit.org/show_bug.cgi?id=157449
2258
2259         Reviewed by Darin Adler.
2260
2261         * Animometer/developer.html: Stop including runner/animometer.css.
2262         * Animometer/index.html: Udpated elements to fit new styles.
2263
2264         * Animometer/resources/debug-runner/animometer.css:
2265         Copy the old runner/animometer.css to here so the debug runner is unaffected.
2266
2267         * Animometer/resources/runner/animometer.css:
2268         New styles for the design. Drop flex box for broad browser support.
2269
2270         * Animometer/resources/runner/animometer.js:
2271         (ResultsTable.clear): Use textContent.
2272         (window.sectionsManager.showSection): Add class to the body.
2273         (window.sectionsManager.setSectionScore): Use textContent, not innerHTML.
2274
2275         * Animometer/resources/runner/crystal.svg: Added.
2276         * Animometer/resources/runner/lines.svg: Added.
2277         * Animometer/resources/runner/logo.svg: Added.
2278
2279         * Animometer/resources/runner/tests.js: Shorten test name and title case them.
2280
2281 2016-05-06  Manuel Rego Casasnovas  <rego@igalia.com>
2282
2283         [css-grid] Unprefix CSS Grid Layout properties
2284         https://bugs.webkit.org/show_bug.cgi?id=157137
2285
2286         Reviewed by Simon Fraser.
2287
2288         Remove "-webkit" prefix from all the grid layout properties,
2289         including the display value.
2290         Update the source code to remove the prefix where it was used too.
2291
2292         * Layout/auto-grid-lots-of-data.html:
2293         * Layout/fixed-grid-lots-of-data.html:
2294         * Layout/fixed-grid-lots-of-stretched-data.html:
2295
2296 2016-04-25  Geoffrey Garen  <ggaren@apple.com>
2297
2298         bmalloc: Misc improvements to MallocBench
2299         https://bugs.webkit.org/show_bug.cgi?id=157004
2300
2301         Reviewed by Darin Adler.
2302
2303         * MallocBench/run-malloc-benchmarks: Added --memory and --memory_warning
2304         modes for focused memory testing.
2305
2306         * MallocBench/MallocBench/Benchmark.cpp:
2307         (Benchmark::printReport): Clarified output.
2308
2309         (Benchmark::currentMemoryBytes): Added compressed memory because top
2310         does the same. (It always happens to zero in the benchmarks we run. But
2311         this is good for sanity.)
2312
2313         * MallocBench/MallocBench/CommandLine.cpp: Moved up to 8 runs to reduce
2314         variance.
2315
2316         * MallocBench/MallocBench/alloc_free.cpp:
2317         (benchmark_alloc_free): Cycle a single allocation in order to stress
2318         the effect of merging on calls to madvise.
2319
2320         * MallocBench/MallocBench/big.cpp:
2321         (benchmark_big): Graduated to 8kB-128kB because medium tests up to 8 and
2322         our large allocator doesn't kick in until 64kB.
2323
2324         * MallocBench/MallocBench/medium.cpp:
2325         (benchmark_medium): Test all the way down to 1kB because our large
2326         allocator used to service 1kB allocations and 1kB is an interesting
2327         middle size where memory is unusually large but allocation throughput
2328         still matters.
2329
2330         * MallocBench/MallocBench/stress.cpp:
2331         (benchmark_stress): Reduced the churn count to match stress_aligned
2332         because this test was taking too long to complete.
2333
2334         * MallocBench/MallocBench/stress_aligned.cpp:
2335         (benchmark_stress_aligned): Our new large allocator can handle even
2336         more absurdly large values.
2337
2338 2016-04-25  Simon Fraser  <simon.fraser@apple.com>
2339
2340         Add a content animation test that uses SVG animation.
2341         https://bugs.webkit.org/show_bug.cgi?id=156827
2342
2343         * Animation/svg-animation.html: Added.
2344         * Skipped:
2345
2346 2016-04-22  Simon Fraser  <simon.fraser@apple.com>
2347
2348         Skip two content animation tests which are only meant for iOS testing.
2349
2350         * Animation/css-animation.html: Added.
2351         * Animation/raf-animation.html: Added.
2352
2353         * Skipped:
2354
2355 2016-04-20  Simon Fraser  <simon.fraser@apple.com>
2356
2357         Add content animation tests to benchmark_runner, and allow the runner to collect device data as part of the results
2358         https://bugs.webkit.org/show_bug.cgi?id=156827
2359
2360         Add two files missing from the previous commit.
2361
2362         * Animation/css-animation.html: Added.
2363         * Animation/raf-animation.html: Added.
2364
2365 2016-04-20  Simon Fraser  <simon.fraser@apple.com>
2366
2367         Add content animation tests to benchmark_runner, and allow the runner to collect device data as part of the results
2368         https://bugs.webkit.org/show_bug.cgi?id=156827
2369
2370         Reviewed by Stephanie Lewis.
2371
2372         Two new files for the content-animation suite.
2373
2374         After patching for use in the benchmark, these tests are a little different to other benchmarks.
2375         The test content itself does not collect performance data; the tests spit out a boilerplate JSON
2376         result with a placeholder for framerate, which is collected in native code.
2377
2378         When run as a benchmark, test completion involves a couple of bounces between the test and MobileSafari.
2379         Test completion is initiated by a setTimeout() in the test, at which point location.hash is set to "#done".
2380         MobileSafari detects that, and asynchronously dumps performance data. Once this is complete, MobileSafari
2381         changes the URL has to "#submit", which triggers the XHR to the test relay that indicates the test is
2382         complete.
2383
2384         * Animation/css-animation.html: Added.
2385         * Animation/raf-animation.html: Added.
2386
2387 2016-04-15  Jon Lee  <jonlee@apple.com>
2388
2389         Animometer test could report a NaN
2390         https://bugs.webkit.org/show_bug.cgi?id=156646
2391
2392         Reviewed by Darin Adler.
2393         Provisionally reviewed by Said Abou-Hallawa.
2394
2395         * Animometer/tests/resources/main.js:
2396         (didFinishInterval): The ramp controller has a first phase where it ramps up the complexity and
2397         reacts based on how well the system handles the load. The assumption was that it would handle at least
2398         1 particle easily. That is not always the case. As a result, an interpolation calculation could
2399         end up setting an upper bound of NaN.
2400
2401         This occurs because we never get out of the first tier, so this._lastTierComplexity is undefined.
2402         Now that we guarantee a minimum complexity of 1, modify the conditional to check for this._lastTierComplexity
2403         before interpolating the upper bound of the first ramp. In the case where the system struggles with
2404         1 particle, set it to currentComplexity.
2405
2406 2016-04-09  Jon Lee  <jonlee@apple.com>
2407
2408         Update Animometer to accommodate different screens
2409         https://bugs.webkit.org/show_bug.cgi?id=156449
2410
2411         Reviewed by Darin Adler.
2412         Provisionally reviewed by Said Abou-Hallawa.
2413
2414         * Animometer/index.html: Wrap button in a container to add padding at the bottom.
2415         * Animometer/resources/debug-runner/animometer.css:
2416         (@media screen and (min-device-width: 1800px)): Deleted.
2417         (@media screen and (min-width: 1800px)): Cannot use min-device-width since it may match incorrectly.
2418         (screen and (max-device-height: 414px) and (orientation: landscape)): Some devices swap device width
2419         and height with orientation change.
2420         * Animometer/resources/runner/animometer.css: Similar.
2421         (screen and (min-device-width: 1024px) and (orientation: landscape)):
2422         (screen and (max-device-height: 414px) and (orientation: landscape)):
2423         (.frame-container): On smaller iPhones, adding 1px prevents the navigation bars from appearing.
2424         (@media screen and (min-device-width: 768px) and (max-device-width: 1024px)): Deleted.
2425         (@media (min-device-height: 768px) and (max-device-height: 1024px)): Target iPad Airs and similar.
2426         (@media screen and (min-device-width: 1024px) and (max-device-width: 1366px)): Deleted.
2427         (@media screen and (max-device-width: 1024px) and (min-device-height: 1366px)): Target iPad Pro.
2428         (#results footer): Add padding below the button for testing again.
2429         * Animometer/tests/master/multiply.html: Remove the center text.
2430         * Animometer/tests/master/resources/text.js: Update the test so that in every frame the text moves.
2431         * Animometer/tests/master/text.html: Update the text sizing depending on the size of the device.
2432
2433 2016-04-08  Jon Lee  <jonlee@apple.com>
2434
2435         Have Animometer benchmark always start with complexity of 1
2436         https://bugs.webkit.org/show_bug.cgi?id=156432
2437
2438         Reviewed by Ryosuke Niwa.
2439
2440         * Animometer/tests/resources/main.js: Update the default Controller and RampController to
2441         set its minimum complexities to 1 instead of 0.
2442
2443 2016-04-08  Jon Lee  <jonlee@apple.com>
2444
2445         Fix SVG benchmark test
2446         https://bugs.webkit.org/show_bug.cgi?id=156410
2447
2448         Reviewed by Dean Jackson.
2449
2450         * Animometer/resources/extensions.js: Update Point.zero to be a static Point.
2451         * Animometer/tests/simple/resources/tiled-canvas-image.js:
2452         (Stage.call._setupTiles): Refactor.
2453         * Animometer/tests/master/resources/particles.js:
2454         (Particle.prototype.reset): Use Point.center.
2455         (complexity): We are not using a gradient background anymore, so remove the +1.
2456         * Animometer/tests/master/resources/svg-particles.js: Update to use SVG transform
2457         instead of CSS transform.
2458
2459 2016-04-06  Jon Lee  <jonlee@apple.com>
2460
2461         Update master benchmark with SVG test
2462         https://bugs.webkit.org/show_bug.cgi?id=156273
2463
2464         Reviewed by Dean Jackson.
2465         Provisionally reviewed by Said Abou-Hallawa.
2466
2467         Switch masks tests for SVG path test.
2468
2469         * Animometer/resources/debug-runner/tests.js: Move mask test here.
2470         * Animometer/resources/runner/tests.js: Add SVG test here.
2471         * Animometer/tests/dom/particles.html: Renamed from PerformanceTests/Animometer/tests/master/particles.html.
2472         * Animometer/tests/dom/resources/dom-particles.js: Renamed from PerformanceTests/Animometer/tests/master/resources/dom-particles.js.
2473
2474         * Animometer/tests/master/resources/particles.js: Add minPosition for bounds checking. Prevents particle from being
2475         partially obscured.
2476         * Animometer/tests/master/resources/svg-particles.js: Added.
2477         (Particle): The particle is either a path object or a rect using a path as a clip. The
2478         same path is used either way. For each particle create a linear gradient with a random
2479         rotation.
2480         (SVGParticlesStage): Look in #shapes to see how many different kinds of paths are available.
2481         This makes the test more generic in case other shapes need to be tested.
2482         * Animometer/tests/master/svg-particles.html: Added. Have two defs, one that houses each
2483         particle's gradient, and one that holds the shape templates.
2484
2485 2016-03-27  Hunseop Jeong  <hs85.jeong@samsung.com>
2486
2487         [EFL] REGRESSION(r188793): It made 200 layout tests and Bindings/event-target-wrapper.html performance test fail
2488         https://bugs.webkit.org/show_bug.cgi?id=148470 
2489
2490         Reviewed by Darin Adler.
2491
2492         * Skipped: Unskip the Bindings/event-target-wrapper test.
2493
2494 2016-03-23  Jon Lee  <jonlee@apple.com>
2495
2496         Update focus and leaves tests
2497         https://bugs.webkit.org/show_bug.cgi?id=155825
2498
2499         Reviewed by Simon Fraser.
2500
2501         Particles would improperly disappear, before they were fully off stage. Update the calculation.
2502         * Animometer/tests/dom/resources/leaves.js:
2503         * Animometer/tests/master/resources/leaves.js:
2504
2505         * Animometer/tests/master/focus.html: Update the center element styling.
2506         * Animometer/tests/master/resources/focus.js:
2507         (FocusElement.Utilities.createClass): Give enough space in the container for each particle for the blur
2508         to spread. It should be about 3x more space than the blur radius on each side.
2509         (Stage.call.initialize): Ditto for the center element.
2510
2511 2016-03-23  Jon Lee  <jonlee@apple.com>
2512
2513         Update support for other platforms
2514         https://bugs.webkit.org/show_bug.cgi?id=155824
2515
2516         Reviewed by Simon Fraser.
2517
2518         * Animometer/resources/debug-runner/animometer.css: Remove extraneous media query. All of it should
2519         be handled in the release stylesheet.
2520         * Animometer/resources/debug-runner/animometer.js: Add support for browsers that return an HTMLCollection
2521         when retrieving radio elements with a name, instead of the selected value.
2522         * Animometer/resources/runner/animometer.css: Adjust the queries for devices of varying widths.
2523         (.frame-container): Move overflow:hidden to #stage.
2524         * Animometer/tests/master/resources/stage.css: Consolidate into tests/resources/stage.css.
2525         * Animometer/tests/resources/stage.css:
2526         (body): Remove overflow:hidden.
2527         (#stage): Set overflow:hidden here.
2528         (#center-text): Moved from master/resources/stage.css.
2529
2530         Update stylesheet hrefs.
2531         * Animometer/tests/master/canvas-stage.html:
2532         * Animometer/tests/master/focus.html:
2533         * Animometer/tests/master/leaves.html:
2534         * Animometer/tests/master/multiply.html:
2535         * Animometer/tests/master/particles.html:
2536         * Animometer/tests/master/text.html: Adjust text size to fit in the device screen.
2537
2538 2016-03-23  Jon Lee  <jonlee@apple.com>
2539
2540         Make the benchmark require device to be in landscape orientation
2541         https://bugs.webkit.org/show_bug.cgi?id=155822
2542         rdar://problem/25258650
2543
2544         Reviewed by Ryosuke Niwa.
2545
2546         Check that the device is in landscape orientation. If not, disable the button to start it,
2547         and show a warning message.
2548
2549         * Animometer/developer.html: Add a message about how the browser should be set up. Include a
2550         warning message that appears if the orientation is incorrect on mobile devices.
2551         * Animometer/index.html: Ditto.
2552
2553         * Animometer/resources/debug-runner/animometer.css: Migrate the .hidden rule into the release
2554         stylesheet. Update the style to accommodate the new UI.
2555         * Animometer/resources/runner/animometer.css:
2556
2557         * Animometer/resources/runner/animometer.js:
2558         (window.benchmarkController.initialize): Add the orientation listener is needed.
2559         (window.benchmarkController.addOrientationListenerIfNecessary): Only mobile devices need this. Check
2560         to see for support of window.orientation.
2561         (window.benchmarkController._orientationChanged): Toggle the warning based on whether we match the
2562         landscape query. Set a state variable, which is needed for the debug harness. Call updateStartButtonState.
2563         (window.benchmarkController.updateStartButtonState):
2564         * Animometer/resources/debug-runner/animometer.js: Change _updateStartButtonState to return
2565         a boolean about whether at least one test is selected. That will be used in
2566         benchmarkController.updateStartButtonState(). Move the load event listener to the release version.
2567         (window.benchmarkController.updateStartButtonState): Override the release version, and also check
2568         that at least one test is selected.
2569
2570 2016-03-22  Geoffrey Garen  <ggaren@apple.com>
2571
2572         MallocBench: consolidate regression testing for aligned allocation
2573         https://bugs.webkit.org/show_bug.cgi?id=155762
2574
2575         Reviewed by Michael Saboff.
2576
2577         * MallocBench/MallocBench.xcodeproj/project.pbxproj:
2578         * MallocBench/MallocBench/Benchmark.cpp:
2579         * MallocBench/MallocBench/memalign.cpp: Removed.
2580         * MallocBench/MallocBench/memalign.h: Removed. The stress_aligned test
2581         covers this and much more.
2582
2583         * MallocBench/MallocBench/stress_aligned.cpp:
2584         (benchmark_stress_aligned): Include specific tests for extreme sizes
2585         and alignments.
2586
2587 2016-03-21  Jon Lee  <jonlee@apple.com>
2588
2589         Update benchmark tests
2590         https://bugs.webkit.org/show_bug.cgi?id=155723
2591
2592         Reviewed by Darin Adler.
2593         Provisionally reviewed by Said Abou-Hallawa.
2594
2595         Add quadratic and bezier segments to the canvas path test.
2596
2597         * Animometer/resources/runner/tests.js: Some of the query strings are unnecessary.
2598         Rename the test.
2599         * Animometer/tests/master/resources/canvas-stage.js:
2600         (tune): Update to be able to take an array of possible constructors. Choose one randomly.
2601         * Animometer/tests/master/resources/canvas-tests.js:
2602         (CanvasLinePoint.Utilities.createClass): Move the point selection out to a separate
2603         function called randomPoint() for reuse. Move X_LOOPS, Y_LOOPS, and offsets into the
2604         class definition.
2605         (randomPoint): Scale the grid down a little bit so that the lines along the edge of the
2606         canvas are not cut off when the stroke size is thick.
2607         (CanvasQuadraticSegment): Added.
2608         (CanvasBezierSegment): Added.
2609         (SimpleCanvasStage): Pass in an array of the different segment types. Since line segments
2610         are short compared to the curved ones, make it twice as likely to render a line segment.
2611         (SimpleCanvasStage.animate): Update the drawing code so that we render all line segments.
2612
2613         Add a helper method that selects a random element from an array.
2614
2615         * Animometer/tests/resources/main.js:
2616         (Stage.randomElementInArray): Select a random element from the provided array.
2617
2618         * Animometer/tests/bouncing-particles/resources/bouncing-tagged-images.js: Refactor.
2619         * Animometer/tests/dom/resources/leaves.js: Ditto.
2620         * Animometer/tests/master/resources/dom-particles.js: Ditto.
2621         * Animometer/tests/master/resources/image-data.js: Ditto.
2622         * Animometer/tests/master/resources/leaves.js: Ditto.
2623         * Animometer/tests/simple/resources/simple-canvas-paths.js: Ditto.
2624
2625         Add canvas tests that includes all stroke and fill paths. This makes it possible to avoid having to
2626         include the full simple canvas suite for perf testing.
2627
2628         * Animometer/resources/debug-runner/tests.js: Add new tests. Move the canvas test into the Canvas
2629         suite.
2630         * Animometer/tests/simple/resources/simple-canvas-paths.js: Add a CanvasStroke and CanvasFill particle
2631         that random selects an object to render.
2632
2633         * Animometer/resources/debug-runner/tests.js: Move 3D suite before basic canvas suite.
2634         When updating the perf bot script, we will include the suites up to this one, but exclude
2635         the basic canvas suite.
2636
2637         Move compositing transforms test to HTML suite and remove the empty Miscellaneous suite.
2638
2639         * Animometer/resources/debug-runner/tests.js:
2640         * Animometer/tests/dom/compositing-transforms.html: Renamed from PerformanceTests/Animometer/tests/misc/compositing-transforms.html.
2641         * Animometer/tests/dom/resources/compositing-transforms.js: Renamed from PerformanceTests/Animometer/tests/misc/resources/compositing-transforms.js.
2642
2643         Clean up miscellaneous test suite. Add a canvas ellipse test, and remove the other
2644         canvas tests.
2645
2646         * Animometer/resources/debug-runner/tests.js: Add ellipse tests to the simple suite.
2647         * Animometer/tests/misc/canvas-electrons.html: Removed.
2648         * Animometer/tests/misc/canvas-stars.html: Removed.
2649         * Animometer/tests/misc/resources/canvas-electrons.js: Removed.
2650         * Animometer/tests/misc/resources/canvas-stars.js: Removed.
2651         * Animometer/tests/simple/resources/simple-canvas-paths.js: Add ellipse primitives.
2652
2653         Merge text tests together into one.
2654
2655         * Animometer/resources/runner/tests.js: Remove international.html.
2656         * Animometer/tests/master/international.html: Removed.
2657         * Animometer/tests/master/resources/text.js:
2658         (animate): Update styling. Manually calculate gradients for the shadow particles.
2659         Reduce the step size for y direction to avoid cutting text off at the margins. Increase
2660         step for x since there will be a little more room.
2661         * Animometer/tests/master/text.html: Add more translations and lay it out in a table.
2662
2663 2016-03-20  Jon Lee  <jonlee@apple.com>
2664
2665         Add a link to show debug data
2666         https://bugs.webkit.org/show_bug.cgi?id=155724
2667
2668         Reviewed by Simon Fraser.
2669
2670         * Animometer/developer.html: Attach the onclick handler to the score.
2671         * Animometer/index.html: Ditto.
2672
2673 2016-03-18  Dean Jackson  <dino@apple.com>
2674
2675         Add a basic WebGL test to Animometer
2676         https://bugs.webkit.org/show_bug.cgi?id=155475
2677         <rdar://problem/25156860>
2678
2679         Reviewed by Simon Fraser and Jon Lee.
2680
2681         Add a pretty simple WebGL test to Animometer
2682         which tries to draw a number of animated triangles.
2683         The result still needs a bit of tuning.
2684
2685         * Animometer/resources/debug-runner/tests.js: Add a new
2686         Suite for "3D" tests.
2687         * Animometer/tests/3d/resources/webgl.js: Added.
2688         * Animometer/tests/3d/webgl.html: Added.
2689
2690 2016-03-18  Jon Lee  <jonlee@apple.com>
2691
2692         Add support for statically linking to a specific test
2693         https://bugs.webkit.org/show_bug.cgi?id=155631
2694
2695         Rubber-stamped by Darin Adler.
2696
2697         * Animometer/developer.html: Update to call restartBenchmark instead so that it works for
2698         both the preset and the manually set test sessions.
2699         * Animometer/resources/debug-runner/animometer.css:
2700         (.tree .link): Style the "link" UI.
2701         * Animometer/resources/debug-runner/animometer.js: Move updateDisplay from suitesManager
2702         to optionsManager since this is a visual option and has nothing to do with the suites
2703         settings. Update the construction of the test UI by including a "link" after each test.
2704         Clicking on that link brings up a JS prompt with a URL and a query string with the current
2705         parameters of the controller, and selected test. Pasting this URL into the location bar will
2706         automatically start running the selected test.
2707         (suitesManager.suitesFromQueryString): Iterate through the Suites and tests and find the one
2708         that matches the provided parameters. Returns an object similar to the form in
2709         suitesManager.updateLocalStorageFromUI.
2710         (benchmarkController.initialize): After settings up the events and options, try parsing the
2711         query string and running the benchmark immediately. Otherwise, fall back to the form.
2712         (benchmarkController.startBenchmark): Store the options and suites into member variables for
2713         reuse in restartBenchmark.
2714         (benchmarkController.startBenchmarkImmediatelyIfEncoded): Convert the query string to an object.
2715         If that's successful, find the suite and test referenced in the query string. Start the benchmark if
2716         the search for the test succeeded.
2717         * Animometer/resources/extensions.js:
2718         (Utilities.stripNonASCIICharacters): Helper method to convert the name of the suite and test
2719         into a query-string-friendly version.
2720         (Utilities.convertObjectToQueryString): Helper method to convert an object to query string
2721         format.
2722         (Utilities.convertQueryStringToObject): Helper method to convert query string into an object
2723         with properties and values.
2724
2725 2016-03-16  Jon Lee  <jonlee@apple.com>
2726
2727         Add a new benchmark test
2728         https://bugs.webkit.org/show_bug.cgi?id=155570
2729
2730         Reviewed by Simon Fraser.
2731
2732         New Leaves test includes various image sizes and opacity.
2733
2734         * Animometer/resources/debug-runner/tests.js: Add it to the HTML test suite.
2735         * Animometer/tests/dom/leaves.html: Added.
2736         * Animometer/tests/dom/resources/leaves.js: Added. Override the
2737         (Particle.call.reset): Uses a range of sizes, and opacity.
2738         (Particle.call.animate): Opacity goes up then down. When it hits 0, reset the particle.
2739         (Particle.call.move): Set transform and opacity.
2740         * Animometer/tests/master/resources/leaves.js: Get rid of the closure so that it
2741         can be used in this test. Update the relative path so that it works in both the master
2742         and dom test suite.
2743
2744 2016-03-15  Simon Fraser  <simon.fraser@apple.com>
2745
2746         Add developer Animometer test that bounces P3-tagged images
2747         https://bugs.webkit.org/show_bug.cgi?id=155511
2748
2749         Reviewed by Tim Horton.
2750         
2751         Add a test for rendering performance of tagged images. The 5 images are tagged
2752         with the Display P3 colorspace.
2753
2754         * Animometer/resources/debug-runner/tests.js:
2755         * Animometer/tests/bouncing-particles/bouncing-tagged-images.html: Added.
2756         * Animometer/tests/bouncing-particles/resources/bouncing-tagged-images.js: Added.
2757         * Animometer/tests/bouncing-particles/resources/image1.jpg: Added.
2758         * Animometer/tests/bouncing-particles/resources/image2.jpg: Added.
2759         * Animometer/tests/bouncing-particles/resources/image3.jpg: Added.
2760         * Animometer/tests/bouncing-particles/resources/image4.jpg: Added.
2761         * Animometer/tests/bouncing-particles/resources/image5.jpg: Added.
2762
2763 2016-03-09  Jon Lee  <jonlee@apple.com>
2764
2765         Enhance existing Animometer tests
2766         https://bugs.webkit.org/show_bug.cgi?id=155261
2767
2768         Reviewed by Simon Fraser.
2769
2770         * Animometer/tests/master/resources/canvas-tests.js: Add a gradient to
2771         the filled circles. To expose more of the gradient, add another circle.
2772         (SimpleCanvasStage.call.animate): For each frame, create a gradient with
2773         undulating stop points and colors. Fill the circles twice; once with
2774         the solid color, and once with the gradient.
2775
2776         * Animometer/tests/master/resources/dom-particles.js: Refactor the
2777         emission variables into a separate stage for this test. Add a colorOffset
2778         variable to make the colors of each particle slightly different, since
2779         the ramp controller can add large numbers of particles all at once, which
2780         would otherwise get all the same color.
2781         * Animometer/tests/master/resources/particles.js:
2782         (initialize): Remove the code specific to the SVG mask test.
2783
2784         * Animometer/tests/master/resources/image-data.js:
2785         (initialize): Remove unused local variable.
2786         * Animometer/tests/master/resources/multiply.js:
2787         (initialize): Make the test harder by adding more total particles.
2788
2789 2016-03-09  Jon Lee  <jonlee@apple.com>
2790
2791         Add text tests
2792         https://bugs.webkit.org/show_bug.cgi?id=155257
2793
2794         Reviewed by Simon Fraser.
2795
2796         * Animometer/resources/extensions.js:
2797         (UnitBezier.Utilities.createClass): Add a class that computes Bezier points
2798         assuming that two of the control points are at (0,0) and (1,1). Taken from
2799         WebCore/platform/graphics/UnitBezier.h
2800         * Animometer/resources/runner/tests.js: Add a test for English text and one
2801         for international text.
2802         * Animometer/tests/master/international.html: Added.
2803         * Animometer/tests/master/resources/text.js: Added. The test assumes there is
2804         #template div which it will copy. The copies are placed behind the template, and
2805         are set with different colors each frame. They are moved around with CSS transform.
2806         * Animometer/tests/master/text.html: Added.
2807
2808         Remove the other text tests, since these ones cover the same techniques.
2809         * Animometer/tests/text/layering-text.html: Removed.
2810         * Animometer/tests/text/resources/layering-text.js: Removed.
2811         * Animometer/tests/text/resources/text-boxes.js: Removed.
2812         * Animometer/tests/text/text-boxes.html: Removed.
2813         * Animometer/resources/debug-runner/tests.js:
2814
2815 2016-03-09  Jon Lee  <jonlee@apple.com>
2816
2817         Add a new image test
2818         https://bugs.webkit.org/show_bug.cgi?id=155232
2819
2820         Reviewed by Dean Jackson.
2821         Provisionally reviewed by Said Abou-Hallawa.
2822
2823         The image test renders PNGs and moves them with translate and rotate
2824         transforms. Each particle has a lifetime, and when the lifetime ends or
2825         the particle goes offscreen, it resets itself somewhere on the stage.
2826
2827         * Animometer/resources/debug-runner/tests.js: Remove the CSS bouncing PNG
2828         images test, because this one tests the same technique.
2829         * Animometer/resources/runner/tests.js: Added here as "Leaves".
2830         * Animometer/tests/master/leaves.html: Added.
2831         * Animometer/tests/master/resources/particles.js: Refactor out the parts
2832         specific to the DOM particles test. Consequently make velocity a public
2833         member.
2834         (initialize): The options parameter is never used, so remove it.
2835         (animate): Remove unused local variable.
2836         * Animometer/tests/master/resources/leaves.js: Maintains a focal point that
2837         moves back and forth across the canvas. That point affects the velocity
2838         of the particles. When the focal point is on the leftmost side, it is
2839         between [-6, -2], and set in reset(). When the focal point is on the rightmost
2840         side, it will be from [2, 6].
2841         * Animometer/tests/master/resources/dom-particles.js: Move JS specific to
2842         this test here from particles.js.
2843
2844         New images.
2845         * Animometer/tests/master/resources/compass100.png: Added.
2846         * Animometer/tests/master/resources/console100.png: Added.
2847         * Animometer/tests/master/resources/contribute100.png: Added.
2848         * Animometer/tests/master/resources/debugger100.png: Added.
2849         * Animometer/tests/master/resources/inspector100.png: Added.
2850         * Animometer/tests/master/resources/layout100.png: Added.
2851         * Animometer/tests/master/resources/performance100.png: Added.
2852         * Animometer/tests/master/resources/script100.png: Added.
2853         * Animometer/tests/master/resources/shortcuts100.png: Added.
2854         * Animometer/tests/master/resources/standards100.png: Added.
2855         * Animometer/tests/master/resources/storage100.png: Added.
2856         * Animometer/tests/master/resources/styles100.png: Added.
2857         * Animometer/tests/master/resources/timeline100.png: Added.
2858
2859 2016-03-03  Jon Lee  <jonlee@apple.com>
2860
2861         Add ability to retrieve raw data from release harness
2862         https://bugs.webkit.org/show_bug.cgi?id=155026
2863
2864         Reviewed by Simon Fraser.
2865
2866         * Animometer/developer.html: Remove the special UI in the debug
2867         harness.
2868         * Animometer/resources/runner/animometer.css: Add styles for the
2869         overlay.
2870         * Animometer/resources/runner/animometer.js: Let 'j' show the JSON
2871         results, but only if the overlay doesn't exist. Add 'esc' key to dismiss
2872         the overlay.
2873         (window.benchmarkController.selectResults): Cycle the cases around
2874         so that the first 's' press selects both the benchmark score and
2875         the individual test scores.
2876         * Animometer/resources/debug-runner/animometer.css: Remove
2877         unneeded rules.
2878         * Animometer/resources/debug-runner/animometer.js: Make the same
2879         call to handleKeyPress.
2880
2881 2016-03-03  Jon Lee  <jonlee@apple.com>
2882
2883         Make sure multiply test particles have at least some opacity
2884         https://bugs.webkit.org/show_bug.cgi?id=155027
2885
2886         Reviewed by Simon Fraser.
2887
2888         * Animometer/tests/master/resources/multiply.js:
2889         (initialize): Have the elements spin a little faster to make it
2890         more obvious when the system is being stressed.
2891         (_addTile):
2892         (animate): Make sure each element being animated has at least 1%
2893         opacity.
2894
2895 2016-03-03  Jon Lee  <jonlee@apple.com>
2896
2897         Update image test
2898         https://bugs.webkit.org/show_bug.cgi?id=154962
2899
2900         Rubber-stamped by Darin Adler.
2901
2902         * Animometer/tests/master/image-data.html: Make each canvas have a compositing layer.
2903         * Animometer/tests/master/resources/image-data.js:
2904         (initialize): Update to load multiple images serially.
2905         (_loadImage):
2906         (tune): Use display instead of visibility. When showing an element anew, move it somewhere else.
2907         (_createTestElement): Set up the element, refactor out placement of element to _refreshElement.
2908         (_refreshElement): Place element in a tile grid.
2909         (animate):
2910         (_getRandomNeighboringPixelIndex): Use a more conservative distribution to make the effect last a little longer.
2911         * Animometer/tests/master/resources/compass.svg: Added.
2912         * Animometer/tests/master/resources/console.svg: Added.
2913         * Animometer/tests/master/resources/contribute.svg: Added.
2914         * Animometer/tests/master/resources/debugger.svg: Added.
2915         * Animometer/tests/master/resources/inspector.svg: Added.
2916         * Animometer/tests/master/resources/layout.svg: Added.
2917         * Animometer/tests/master/resources/performance.svg: Added.
2918         * Animometer/tests/master/resources/script.svg: Added.
2919         * Animometer/tests/master/resources/shortcuts.svg: Added.
2920         * Animometer/tests/master/resources/standards.svg: Added.
2921         * Animometer/tests/master/resources/storage.svg: Added.
2922         * Animometer/tests/master/resources/styles.svg: Added.
2923         * Animometer/tests/master/resources/timeline.svg: Added.
2924
2925 2016-03-02  Jon Lee  <jonlee@apple.com>
2926
2927         Add some new controllers, and refine tests
2928         https://bugs.webkit.org/show_bug.cgi?id=154914
2929
2930         Reviewed by Simon Fraser.
2931
2932         Improve tests.
2933
2934         * Animometer/tests/master/focus.html: Move each particle into a layer. Put the particle
2935         in a container element. Expand the container element by the max blur radius, and clip
2936         overflow. This way, when the blur is applied, it avoids causing layer resizes.
2937         * Animometer/tests/master/resources/focus.js:
2938         (FocusElement.Utilities.createClass): Create a container element, and put the particle
2939         inside.
2940         (hide): Set display:none.
2941         (show): Set display:block.
2942         (animate): Apply filters to the container element.
2943         (FocusStage.call.initialize): Instead of inserting and removing elements from the DOM,
2944         keep them in the stage, but set the display style instead. Use this._offsetIndex to
2945         keep track of which elements are displayed.
2946         (FocusStage.call.animate): Move some calculations around to avoid doing unneeded math.
2947         (FocusStage.call.getBlurValue): Make sure elements are always blurred.
2948         (FocusStage.call.getOpacityValue): Make sure elements have some opacity.
2949         * Animometer/tests/master/resources/multiply.js: Have the particles on the edge of the
2950         stage appear less black when the complexity gets large enough.
2951         (tune): Have this._distanceFactor calculate the factor. Avoid calculating square root each
2952         frame.
2953
2954 2016-03-02  Jon Lee  <jonlee@apple.com>
2955
2956         Add some new controllers, and refine tests
2957         https://bugs.webkit.org/show_bug.cgi?id=154914
2958
2959         Reviewed by Simon Fraser.
2960
2961         Add a controller that centers around 30 fps instead of 60 fps.
2962
2963         * Animometer/developer.html: Add a new option.
2964         * Animometer/resources/debug-runner/animometer.js:
2965         * Animometer/resources/runner/animometer.js:
2966         (this._processData.findRegression): When calculating the complexity-frameLength regression,
2967         check the controller, and use a 30 fps baseline if needed.
2968         * Animometer/resources/statistics.js:
2969         (Regression.Utilities.createClass): Update to allow clients to specify the baseline
2970         frame length.
2971         * Animometer/tests/resources/main.js:
2972         (tune): Override some of the constants in RampController. Move those constants out for
2973         easier reading.
2974
2975         Add a fixed controller, with no step.
2976
2977         * Animometer/developer.html: Add a controller that takes no step.
2978         * Animometer/resources/debug-runner/animometer.js:
2979         * Animometer/tests/resources/main.js:
2980         (Rotater.Utilities.createClass):
2981
2982         Switch to ramp controller as default.
2983
2984         * Animometer/developer.html: Increase the test length to 20 seconds.
2985         * Animometer/resources/debug-runner/animometer.js: Rename the "adjustment" field to "controller"
2986         since that is a more accurate description.
2987         * Animometer/resources/debug-runner/graph.js:
2988         * Animometer/resources/runner/animometer.js: Update preferences for release suite.
2989
2990 2016-02-26  Jon Lee  <jonlee@apple.com>
2991
2992         Address Dean's comments in 154673.
2993
2994         * Animometer/developer.html:
2995         * Animometer/resources/debug-runner/animometer.js:
2996         * Animometer/resources/debug-runner/graph.js:
2997
2998 2016-02-25  Jon Lee  <jonlee@apple.com>
2999
3000         Update animation benchmark and tests
3001         https://bugs.webkit.org/show_bug.cgi?id=154673
3002
3003         Reviewed by Dean Jackson.
3004
3005         Update test visuals.
3006
3007         * Animometer/tests/master/resources/canvas-tests.js:
3008         (CanvasLineSegment.Utilities.createClass): Line segments near the edge of the stage get clipped.
3009         Update circle position and radius to minimize impact.
3010         * Animometer/tests/master/resources/particles.js: Get rid of rotating gradient background, and
3011         have 3 locations for emitting particles.
3012         * Animometer/tests/master/resources/multiply.js: Update distance metric so that fringe tiles get
3013         more color.
3014         (initialize): Fix some of the math for laying out the tiles.
3015         (animate): When a tile isn't used, set visibility: hidden.
3016
3017 2016-02-24  Jon Lee  <jonlee@apple.com>
3018
3019         Update animation benchmark and tests
3020         https://bugs.webkit.org/show_bug.cgi?id=154673
3021
3022         Reviewed by Dean Jackson.
3023
3024         Update the ramp controller.
3025
3026         The controller refines the complexity interval to test across.
3027
3028         * Animometer/resources/statistics.js: Add functions that estimate cumulative distribution function.
3029         (Regression): For the flat regression, force the first segment to be at 60 fps.
3030         (valueAt): Add convenience function to return interpolated value based on the regression used.
3031         (_calculateRegression): Include the number of points included for both segments, and the piecewise
3032         errors.
3033         * Animometer/tests/resources/math.js: Make the Kalman estimator subclass Experiment, and allow it
3034         to be reset.
3035
3036         * Animometer/tests/resources/main.js: Initialize the tier such that it starts at 10^0 = 1.
3037         Increase the number of ramps. Maintain three FPS thresholds-- the frame rate of interest, a limit
3038         on the lowest FPS we care to go for later interpolation, and a minimum FPS threshold we want to
3039         aim for each ramp. Also keep three estimators: a running average of the change point, a minimum
3040         boundary for each ramp, and an estimator for all the frames within an interval. The first two
3041         are used to determine the parameters of the next ramp, and the latter allows us to refine the
3042         parameters.
3043         (update): During the tier phase, it is possible that the highest complexity possible for a test
3044         won't stress the system enough to trigger stopping the tier phase and transitioning to the ramps.
3045         If the complexity doesn't change when going to the next tier, we've maxed the test out, and move
3046         on. When the tier phase completed, turn off Controller.frameLengthEstimator, which estimates the
3047         FPS at each tier.
3048         (tune): At each interval, look at the confidence distribution of being on the 60 FPS side or the
3049         slow side. If the slowest FPS we achieve at the ramp's maximum complexity is not at least
3050         _fpsRampSlowThreshold, then increase the maximum complexity. If we ever achieve 60 FPS, increase
3051         the ramp's minimum complexity to that level. If, at an even lower complexity, a glitch causes the
3052         FPS to drop, we reset the minimum complexity.
3053
3054         Have the bootstrap calculation occur between tests. Clean up harness.
3055
3056         * Animometer/resources/debug-runner/animometer.js: Run bootstrap after a test has
3057         completed to avoid doing all of it at the end before showing the results. Clean up
3058         parameters being passed around.
3059         * Animometer/resources/debug-runner/tests.js:
3060         (text):
3061         * Animometer/resources/runner/animometer.js:
3062         (this._processData.calculateScore): Save the results to the same object holding the data.
3063         (this._processData._processData): In the case where a file is dragged, calculate the score
3064         serially. Grab the results object and move it to the results variable and remove it from
3065         the data object. This avoids serializing the results into the JSON.
3066         (this._processData.findRegression): Include the samples used for bootstrapping. Reduce the
3067         resample size to shorten the wait.
3068         * Animometer/resources/runner/benchmark-runner.js:
3069         * Animometer/resources/statistics.js:
3070         (bootstrap): Update how bootstrapData is sorted. In some regression results the mix of
3071         floats and integers causes an alphabetical sort to occur.
3072         * Animometer/resources/strings.js:
3073
3074         Add meta charset so that encodings between harness and test match.
3075
3076         * Animometer/tests/bouncing-particles/bouncing-canvas-images.html:
3077         * Animometer/tests/bouncing-particles/bouncing-canvas-shapes.html:
3078         * Animometer/tests/bouncing-particles/bouncing-css-images.html:
3079         * Animometer/tests/bouncing-particles/bouncing-css-shapes.html:
3080         * Animometer/tests/bouncing-particles/bouncing-svg-images.html:
3081         * Animometer/tests/bouncing-particles/bouncing-svg-shapes.html:
3082         * Animometer/tests/master/canvas-stage.html:
3083         * Animometer/tests/master/focus.html:
3084         * Animometer/tests/master/image-data.html:
3085         * Animometer/tests/master/multiply.html:
3086         * Animometer/tests/master/particles.html:
3087         * Animometer/tests/misc/canvas-electrons.html:
3088         * Animometer/tests/misc/canvas-stars.html:
3089         * Animometer/tests/misc/compositing-transforms.html:
3090         * Animometer/tests/simple/simple-canvas-paths.html:
3091         * Animometer/tests/simple/tiled-canvas-image.html:
3092         * Animometer/tests/template/template-canvas.html:
3093         * Animometer/tests/template/template-css.html:
3094         * Animometer/tests/template/template-svg.html:
3095         * Animometer/tests/text/layering-text.html:
3096         * Animometer/tests/text/text-boxes.html:
3097
3098         Update test harness reporting.
3099
3100         * Animometer/developer.html: Add missing meta charset.
3101         * Animometer/index.html: Remove unnecessary utf-8 declaration.
3102         * Animometer/resources/debug-runner/animometer.css: Add convenience classes for
3103         formatting the results table.
3104         * Animometer/resources/debug-runner/animometer.js: Adjust which stats are shown.
3105         * Animometer/resources/debug-runner/tests.js: Display bootstrapping statistics.
3106         * Animometer/resources/strings.js: Move strings not used by the release harness.
3107
3108         Switch to a pseudo-random number generator.
3109
3110         * Animometer/resources/statistics.js: Add a Pseudo class, with a simple
3111         pseudo-random number generator.
3112         (_calculateRegression): Reset the generator before running bootstrap.
3113         (bootstrap): Deleted.
3114
3115         Replace Math.random with Pseudo.random.
3116         * Animometer/tests/master/resources/canvas-tests.js:
3117         * Animometer/tests/master/resources/focus.js:
3118         * Animometer/tests/master/resources/particles.js:
3119         * Animometer/tests/resources/main.js:
3120
3121         Use bootstrapping to get confidence interval in the breakpoint.
3122
3123         For the ramp controller, calculate the piecewise regression, and then use
3124         bootstrapping in order to find the 95% confidence interval. Use the raw data.
3125
3126         * Animometer/developer.html: Default to the complexity graph. Add a legend
3127         checkbox to toggle visibility of the bootstrap score and histogram.
3128         * Animometer/resources/debug-runner/animometer.css: Make some more space to show
3129         the old raw and average scores in the legend. Add new styles for the data.
3130         * Animometer/resources/debug-runner/graph.js:
3131         (_addRegressionLine): Allow passing an array for the variance bar tied to the
3132         regression line. Now |stdev| is |range|.
3133         (createComplexityGraph): Add bootstrap median, and overlay a histogram of
3134         the bootstrap samples. Switch raw samples from circles to X's.
3135         (onComplexityGraphOptionsChanged): Allow toggling of the bootstrap data.
3136         (onGraphTypeChanged): Move the regressions for the raw and average samples to the
3137         legend. In the subtitle use the bootstrap median, and include the 95% confidence
3138         interval.
3139         * Animometer/resources/runner/animometer.js:
3140         (this._processData.findRegression): Factor out the code that determines which
3141         samples to include when calculating the piecewise regression. For series that have
3142         many samples, or a wider range of recorded complexities, throw away the 2.5%
3143         lowest and highest samples before calculating the regression. Keep all samples
3144         if the number of samples to regress is small or the range of complexities is
3145         narrow.
3146         (this._processData._calculateScore): Factor out regression calculation to
3147         findRegression(). Bootstrap the change point of the regression. The score is the
3148         median.
3149         * Animometer/resources/statistics.js:
3150         (_calculateRegression): Correct an issue in the calculation of the regression, where
3151         the denominator can be 0.
3152         (bootstrap): Template for bootstrapping. Create a bootstrap sample array, Create
3153         re-samples by random selection with replacement. Return the 95% confidence samples,
3154         the bootstrap median, mean, and the data itself.
3155         * Animometer/resources/strings.js: Add bootstrap.
3156         * Animometer/tests/resources/main.js:
3157         (processSamples): Don't prematurely cut the sample data.
3158
3159         Fix graph drawing.
3160
3161         * Animometer/resources/debug-runner/animometer.js: Add spacing in the JSON output.
3162         Multiple tests output a lot of JSON and can hang when selecting JSON with no whitespace.
3163         * Animometer/resources/debug-runner/animometer.css:
3164         (#complexity-graph .series.raw circle): Update the color.
3165         * Animometer/resources/debug-runner/graph.js: Use the FPS axis instead of the
3166         complexity axis, which can vary in domain. For determining the complexity domain,
3167         only use samples after samplingTimeOffset.
3168
3169         Allow dropping results JSON.
3170
3171         * Animometer/developer.html: Add a button.
3172         * Animometer/resources/debug-runner/animometer.css:
3173         * Animometer/resources/debug-runner/animometer.js: Read the data and go straight
3174         to the dashboard. With JSON output, write out only the options and the raw data.
3175
3176         Teach the harness to evaluate the samples and determine the test score.
3177
3178         This will allow us to update how the score is calculated separately from the samples recorded.
3179         This also prepares the harness to be able to accept JSON of prior runs.
3180
3181         * Animometer/resources/strings.js: Clean up and remove unneeded strings and reduce some of the
3182         hierarchy.
3183         * Animometer/resources/debug-runner/tests.js: Update to use the new strings.
3184
3185         * Animometer/tests/resources/main.js: Allow all controllers to show a complexity-FPS graph.
3186         (_processComplexitySamples): Factor out some of the sample processing done in the ramp
3187         controller for the benefit of the other controllers. |complexitySamples| contains a list of
3188         samples. Sort the samples by complexity. Optionally remove the top x% of samples.
3189         Group them, and calculate distribution of samples within the same complexity, and add those as
3190         new entries into |complexityAverageSamples|.
3191         (Controller.processSamples): Move the code responsible for determining the complexity and FPS
3192         scores out to ResultsDashboard. The structure of the data returned by the controller is:
3193
3194         {
3195             controller: [time-regression, time-regression, ...], // optional, data specific to controller
3196             marks: [...],
3197             samples: {                    // all of the sample data
3198                 controller: [...],
3199                 complexity: [...],        // processed from controller samples
3200                 complexityAverage: [...], // processed from complexity samples
3201             }
3202         }
3203
3204         (AdaptiveController.processSamples): Adding the target frame length is no longer necessary; we
3205         now pass the test options to the graph.
3206         (Regression): Move to statistics.js.
3207         * Animometer/resources/statistics.js: Move Regression to here. Add a check if the sampling range
3208         only contains one sample, since we cannot calculate a regression from one sample point.
3209
3210         Teach the test harness to evaluate the data.
3211         * Animometer/resources/runner/animometer.js:
3212         (ResultsDashboard): Store the options used to run the test and the computed results/score separately
3213         from the data. The results are stored as:
3214
3215         {
3216             score: /* geomean of iteration score */,
3217             iterationsResults: [
3218                 {
3219                     score: /* geomean of tests */,
3220                     testsResults: {
3221                         suiteName: {
3222                             testName: {
3223                                 controller: {
3224                                     average:
3225                                     concern:
3226                                     stdev:
3227                                     percent:
3228                                 },
3229                                 frameLength: { ... },
3230                                 complexity: {
3231                                     complexity:
3232                                     stdev:
3233                                     segment1:
3234                                     segment2:
3235                                 },
3236                                 complexityAverage: { ... }
3237                             },
3238                             testName: { ... },
3239                         },
3240                         ... next suite ...
3241                     }
3242                 },
3243                 { ...next iteration... }
3244             ]
3245         }
3246
3247         * Animometer/resources/debug-runner/animometer.js: Pass options around instead of relying
3248         on what was selected in the form. This will later allow for dropping previous results, and
3249         using those runs' options when calculating scores.
3250         (ResultsTable._addGraphButton): Simplify button action by using attached test data.
3251         * Animometer/resources/debug-runner/graph.js: Refactor to use the data.
3252
3253         Consolidate JS files, and move statistics out to a separate JS.
3254
3255         Preparation for having the Controller only handle recording and storage of the samples,
3256         and leave the evaluation of the test score out to the harness. Move Experiment to
3257         a new statistics.js, where Regression will also eventually go. Get rid of algorithm.js
3258         and move it to utilities.js since the Heap is used only for Experiments.
3259
3260         * Animometer/tests/resources/algorithm.js: Removed. Heap is in utilities.js.
3261         * Animometer/tests/resources/sampler.js: Removed. Experiment is in statistics.js,
3262         Sampler in main.js.
3263         * Animometer/tests/resources/main.js: Move Sampler here.
3264         * Animometer/resources/statistics.js: Added. Move Statistics and Experiment here.
3265         * Animometer/resources/extensions.js: Move Heap here. Attach static method to create
3266         a max or min heap to Heap, instead of a new Algorithm object.
3267
3268         Update JS files.
3269         * Animometer/developer.html:
3270         * Animometer/index.html:
3271         * Animometer/tests/bouncing-particles/bouncing-canvas-images.html:
3272         * Animometer/tests/bouncing-particles/bouncing-canvas-shapes.html:
3273         * Animometer/tests/bouncing-particles/bouncing-css-images.html:
3274         * Animometer/tests/bouncing-particles/bouncing-css-shapes.html:
3275         * Animometer/tests/bouncing-particles/bouncing-svg-images.html:
3276         * Animometer/tests/bouncing-particles/bouncing-svg-shapes.html:
3277         * Animometer/tests/master/canvas-stage.html:
3278         * Animometer/tests/master/focus.html:
3279         * Animometer/tests/master/image-data.html:
3280         * Animometer/tests/master/multiply.html:
3281         * Animometer/tests/master/particles.html:
3282         * Animometer/tests/misc/canvas-electrons.html:
3283         * Animometer/tests/misc/canvas-stars.html:
3284         * Animometer/tests/misc/compositing-transforms.html:
3285         * Animometer/tests/simple/simple-canvas-paths.html:
3286         * Animometer/tests/simple/tiled-canvas-image.html:
3287         * Animometer/tests/template/template-canvas.html:
3288         * Animometer/tests/template/template-css.html:
3289         * Animometer/tests/template/template-svg.html:
3290         * Animometer/tests/text/layering-text.html:
3291         * Animometer/tests/text/text-boxes.html:
3292
3293         Fix the cursor in the graph analysis when the min
3294         complexity is not 0.
3295
3296         * Animometer/resources/debug-runner/graph.js:
3297         (_addRegression):
3298         (createComplexityGraph):
3299
3300 2016-02-23  Geoffrey Garen  <ggaren@apple.com>
3301
3302         Fix some issues in MallocBench
3303         https://bugs.webkit.org/show_bug.cgi?id=154600
3304
3305         Reviewed by Oliver Hunt.
3306
3307         * MallocBench/MallocBench.xcodeproj/project.pbxproj: Make nimlang.ops
3308         a part of the build so that it copies to the right place for execution.
3309
3310         * MallocBench/MallocBench/Interpreter.cpp:
3311         (Interpreter::Interpreter): Don't try to open .ops files for writing
3312         because we're only going to read and we might need extra permissions to
3313         write.
3314
3315         (Interpreter::~Interpreter): Give more context when opening a file fails
3316         to help with debugging.
3317
3318         * MallocBench/MallocBench/stress.cpp:
3319         (benchmark_stress): Reduce iterations to complete in less than 10 seconds.
3320
3321 2016-02-23  Michael Saboff  <msaboff@apple.com>
3322
3323         Unreviewed change to revert extraneous changes made part of change set 196955.
3324
3325         * MallocBench/MallocBench/Interpreter.cpp:
3326         (Interpreter::doMallocOp):
3327         (Interpreter::Thread::switchTo):
3328         (writeData): Deleted.
3329
3330 2016-02-19  Michael Saboff  <msaboff@apple.com>
3331
3332         MallocBench: Added recording for nimlang website, new recording details and added new options
3333         https://bugs.webkit.org/show_bug.cgi?id=154485
3334
3335         Reviewed by Geoff Garen.
3336
3337         Added new capabilities to MallocBench.  These include:
3338             Added a recording of http://nim-lang.org/docs/lib.html.
3339             Added thread id to the recording and the ability to playback switching threads in MallocBench
3340             Added aligned allocations to recordings and the ability to playback
3341             Added --use-thread-id option to honor recorded thread ids
3342             Added --detailed-report to output remaining allocations by size after playback
3343             Added --no-warmup to not run the warm up iteration
3344
3345         Changed the way that options are passed down to the benchmarks.  Instead of passing individual
3346         boolean or numeric option values, just pass a reference the CommandLine itself.  Each benchmark
3347         can access the options that are appropriate.  The Benchmark class also uses the options for
3348         is parallel, run counts and warm up.
3349
3350         Added thread id and aligned malloc to the Op by noticing that structure padding and Opcode allowed
3351         for another 32 bits of data.  Breaking that unused 32 bits into a 16 bit thread id value and a
3352         16 bit log base 2 of the alignment for aligned malloc allowed for existing recordings to playback
3353         without any incompatibilities.
3354
3355         Threaded operation is simulated by creating threads as needed.  As long as the next Op's thread id
3356         is the same as the last, operation continues as normal.  When the next Op has a different thread id,
3357         we switch to that thread using the shared Op stream to continue playing back.  There is a mutex to
3358         assure that only one thread is really running at a time and a condition variable used to wait
3359         that the current thread id matches each block thread's thread id.  This doesn't simulate true
3360         concurrent threading, but is instead plays back Ops recorded for multiple thread faithfully.
3361
3362         * MallocBench/MallocBench.xcodeproj/project.pbxproj:
3363         * MallocBench/MallocBench/Benchmark.cpp:
3364         (deallocateHeap):
3365         (Benchmark::Benchmark):
3366         (Benchmark::runOnce):
3367         (Benchmark::run):
3368         * MallocBench/MallocBench/Benchmark.h:
3369         (Benchmark::isValid):
3370         * MallocBench/MallocBench/CommandLine.cpp:
3371         (CommandLine::printUsage):
3372         * MallocBench/MallocBench/CommandLine.h:
3373         (CommandLine::isValid):
3374         (CommandLine::benchmarkName):
3375         (CommandLine::isParallel):
3376         (CommandLine::useThreadID):
3377         (CommandLine::detailedReport):
3378         (CommandLine::warmUp):
3379         (CommandLine::heapSize):
3380         (CommandLine::runs):
3381         * MallocBench/MallocBench/Interpreter.cpp:
3382         (Interpreter::Interpreter):
3383         (Interpreter::run):
3384         (Interpreter::readOps):
3385         (Interpreter::doOnSameThread):
3386         (Interpreter::switchToThread):
3387         (Interpreter::detailedReport):
3388         (compute2toPower):
3389         (writeData):
3390         (Interpreter::doMallocOp):
3391         (Interpreter::Thread::Thread):
3392         (Interpreter::Thread::stop):
3393         (Interpreter::Thread::~Thread):
3394         (Interpreter::Thread::runThread):
3395         (Interpreter::Thread::waitToRun):
3396         (Interpreter::Thread::switchTo):
3397         * MallocBench/MallocBench/Interpreter.h:
3398         (Interpreter::Thread::isMainThread):
3399         * MallocBench/MallocBench/alloc_free.cpp: Added.
3400         (benchmark_alloc_free):
3401         * MallocBench/MallocBench/alloc_free.h: Added.
3402         * MallocBench/MallocBench/balloon.cpp:
3403         (benchmark_balloon):
3404         * MallocBench/MallocBench/balloon.h:
3405         * MallocBench/MallocBench/big.cpp:
3406         (benchmark_big):
3407         * MallocBench/MallocBench/big.h:
3408         * MallocBench/MallocBench/churn.cpp:
3409         (benchmark_churn):
3410         * MallocBench/MallocBench/churn.h:
3411         * MallocBench/MallocBench/facebook.cpp:
3412         (benchmark_facebook):
3413         * MallocBench/MallocBench/facebook.h:
3414         * MallocBench/MallocBench/flickr.cpp:
3415         (benchmark_flickr):
3416         (benchmark_flickr_memory_warning):
3417         * MallocBench/MallocBench/flickr.h:
3418         * MallocBench/MallocBench/fragment.cpp:
3419         (validate):
3420         (benchmark_fragment):
3421         (benchmark_fragment_iterate):
3422         * MallocBench/MallocBench/fragment.h:
3423         * MallocBench/MallocBench/list.cpp:
3424         (benchmark_list_allocate):
3425         (benchmark_list_traverse):
3426         * MallocBench/MallocBench/list.h:
3427         * MallocBench/MallocBench/main.cpp:
3428         (main):
3429         * MallocBench/MallocBench/medium.cpp:
3430         (benchmark_medium):
3431         * MallocBench/MallocBench/medium.h:
3432         * MallocBench/MallocBench/memalign.cpp:
3433         (test):
3434         (benchmark_memalign):
3435         * MallocBench/MallocBench/memalign.h:
3436         * MallocBench/MallocBench/message.cpp:
3437         (benchmark_message_one):
3438         (benchmark_message_many):
3439         * MallocBench/MallocBench/message.h:
3440         * MallocBench/MallocBench/nimlang.cpp: Added.
3441         (benchmark_nimlang):
3442         * MallocBench/MallocBench/nimlang.h: Added.
3443         * MallocBench/MallocBench/nimlang.ops: Added.
3444         * MallocBench/MallocBench/realloc.cpp:
3445         (benchmark_realloc):
3446         * MallocBench/MallocBench/realloc.h:
3447         * MallocBench/MallocBench/reddit.cpp:
3448         (benchmark_reddit):
3449         (benchmark_reddit_memory_warning):
3450         * MallocBench/MallocBench/reddit.h:
3451         * MallocBench/MallocBench/stress.cpp:
3452         (deallocate):
3453         (benchmark_stress):
3454         * MallocBench/MallocBench/stress.h:
3455         * MallocBench/MallocBench/stress_aligned.cpp:
3456         (benchmark_stress_aligned):
3457         * MallocBench/MallocBench/stress_aligned.h:
3458         * MallocBench/MallocBench/theverge.cpp:
3459         (benchmark_theverge):
3460         (benchmark_theverge_memory_warning):
3461         * MallocBench/MallocBench/theverge.h:
3462         * MallocBench/MallocBench/tree.cpp:
3463         (benchmark_tree_allocate):
3464         (benchmark_tree_traverse):
3465         (benchmark_tree_churn):
3466         * MallocBench/MallocBench/tree.h:
3467         * MallocBench/run-malloc-benchmarks:
3468
3469 2016-02-11  Jon Lee  <jonlee@apple.com>
3470
3471         Fix a missing refactoring.
3472
3473         * Animometer/tests/master/resources/multiply.js:
3474         (animate): Move to Utilities.lerp.
3475         (_lerp): Deleted.
3476
3477 2016-02-10  Jon Lee  <jonlee@apple.com>
3478
3479         Add new benchmark tests.
3480         https://bugs.webkit.org/show_bug.cgi?id=154063
3481
3482         Provisionally reviewed by Said Abou-Hallawa.
3483
3484         Add tests for get/put image data, filters, opacity, and css transforms.
3485
3486         * Animometer/resources/runner/benchmark-runner.js:
3487         (_runBenchmarkAndRecordResults): Update the body background color to match that of
3488         the stage.
3489         (this._runNextIteration): Clear the background color style for the results page.
3490         * Animometer/resources/runner/tests.js:
3491         * Animometer/tests/master/focus.html: Added.
3492         * Animometer/tests/master/image-data.html: Added.
3493         * Animometer/tests/master/multiply.html: Added.
3494         * Animometer/tests/master/resources/focus.js: Added.
3495         * Animometer/tests/master/resources/image-data.js: Added.
3496         * Animometer/tests/master/resources/multiply.js: Added.
3497         * Animometer/tests/master/resources/stage.css: Move common styles out.
3498         * Animometer/tests/resources/main.js: Update Stage.randomBool to use Math.random.
3499         Add Stage.randomSign for randomly setting a direction. Add the notion of the
3500         current timestamp of the test to Benchmark, since some animations cycle through
3501         colors and rely on an incremental counter like the time.
3502
3503 2016-02-09  Said Abou-Hallawa  <sabouhallawa@apple.com>
3504
3505         Add internal benchmark tests for CSS mix-blend-modes and filters
3506         https://bugs.webkit.org/show_bug.cgi?id=154058
3507
3508         Provisionally reviewed by Jon Lee.
3509
3510         * Animometer/resources/debug-runner/tests.js: Include the new tests in the
3511         "HTML suite" of the debug runner.
3512
3513         * Animometer/resources/extensions.js:
3514         (Utilities.browserPrefix):
3515         (Utilities.setElementPrefixedProperty): Utility functions to allow setting
3516         prefixed style properties.
3517
3518         * Animometer/tests/bouncing-particles/resources/bouncing-css-shapes.js:
3519         Set the mix-blend-mode and the filter to some random values if the options
3520         of the test requested that.
3521
3522         * Animometer/tests/bouncing-particles/resources/bouncing-particles.js:
3523         (parseShapeParameters): Parse the url options "blend" and "filter" and set
3524         the corresponding flags.
3525
3526         * Animometer/tests/resources/main.js:
3527         (randomStyleMixBlendMode):
3528         (randomStyleFilter): Return random mix-blend-mode and filter.
3529
3530 2016-02-08  Jon Lee  <jonlee@apple.com>
3531
3532         Add a ramp controller
3533         https://bugs.webkit.org/show_bug.cgi?id=154028
3534
3535         Provisionally reviewed by Said Abou-Hallawa.
3536
3537         Enhance the graph to include a complexity-fps graph, in addition
3538         to the time graph.
3539
3540         * Animometer/developer.html: Add a ramp option.
3541         * Animometer/resources/debug-runner/animometer.css: Update the style.
3542         * Animometer/resources/strings.js: Flatten the Strings.text constants.
3543         * Animometer/resources/debug-runner/animometer.js:
3544         (ResultsTable.call._addGraphButton): Refactor.
3545         (ResultsTable.call._addTest): Add regression data.
3546         (benchmarkController): Add a form that allows the user to switch between the two forms,
3547         Add a form that allows the user to toggle different data. Hide certain header columns
3548         depending on the selected controller.
3549         * Animometer/resources/debug-runner/graph.js: Add the complexity regressions.
3550         * Animometer/resources/debug-runner/tests.js: Add headers for the ramp results.
3551         * Animometer/resources/runner/animometer.js:
3552         (ResultsTable): If a header is disabled don't include them in _flattenedHeaders.
3553         * Animometer/tests/resources/main.js:
3554         (Controller): Allow options to specify the capacity for sample arrays.
3555         (Regression): A piecewise regression that tries to fit a slope and a flat profile.
3556         (_calculateRegression): Options can fix the slope and bias when calculating the minimal
3557         error. Sweep across the samples in time (which could be backward depending on the controller)
3558         and calculate the intersection point.
3559         (RampController): This controller assumes that the target frame rate is below
3560         58 FPS. It runs in two stages. The first stage quickly determines the order of
3561         magnitude of objects needed to stress the system by the setting the complexity
3562         to increasingly difficult tiers. Perform a series of ramps descending from a
3563         high-water mark of complexity. The complexity needed to reach the target frame
3564         length is done by performing a piecewise regression on each ramp, and track a
3565         running average of these values. For the next ramp, make that running average
3566         the center of the ramp. With a minimum complexity of 0, the high-water mark is
3567         twice that average. The score is based on the highest complexity that can
3568         reach 60 fps.
3569
3570 2016-02-08  Jon Lee  <jonlee@apple.com>
3571
3572         Address Said's comments on the benchmark, and do some clean up.
3573
3574         * Animometer/developer.html:
3575         * Animometer/resources/debug-runner/animometer.css: Add styles for averages.
3576         * Animometer/resources/debug-runner/animometer.js: Use the right
3577         Strings constants.
3578         * Animometer/resources/debug-runner/graph.js:
3579         (_addRegressionLine): Add missing code to draw the line and standard
3580         deviation highlight.
3581         (onGraphTypeChanged): Remove unneeded variables
3582         (onTimeGraphOptionsChanged):
3583         * Animometer/resources/runner/benchmark-runner.js:
3584         (_runBenchmarkAndRecordResults): Rename samplers to suiteResults and
3585         _suitesSamplers to _suitesResults.
3586         * Animometer/tests/resources/main.js:
3587         (results): Call processSamples().
3588         (update): Change sampling timestamp comparison.
3589         (_animateLoop): Move shouldStop call to before the update.
3590         * Animometer/tests/resources/sampler.js:
3591         (process): Rename to processSamples().
3592
3593 2016-02-07  Jon Lee  <jonlee@apple.com>
3594
3595         Teach Controller to measure intervals, and turn off the frame length estimator.
3596
3597         * Animometer/tests/resources/main.js: Default interval length is 100 ms.
3598         (start): Set the first interval.
3599         (_measureAndResetInterval): Reports the average frame length of the interval that just
3600         completed, and sets up the next interval.
3601         (update): If there is no length, then just use the estimator per frame, otherwise the
3602         estimator measures per interval. Add a didFinishInterval for subclasses to process
3603         prior to recording the sample. Update tune() to include whether an interval had
3604         finished.
3605         (StepController): Step controllers don't measure on an interval basis.
3606
3607 2016-02-07  Jon Lee  <jonlee@apple.com>
3608
3609         Minor refactoring. Rename Controller._estimator to Controller._frameLengthEstimator
3610         and switch the parameters for start(), update(), and tune(), so that the timestamp
3611         is first and stage is second.
3612
3613         * Animometer/tests/resources/main.js:
3614
3615 2016-02-07  Jon Lee  <jonlee@apple.com>
3616
3617         Move ResultsTable functionality not needed for release tests out.
3618         Move reporting of score and mean to selection of the time-based graph.
3619
3620         * Animometer/developer.html: Rename graph-options to time-graph-options.
3621         * Animometer/resources/debug-runner/animometer.js:
3622         (DeveloperResultsTable): Moved from runner/animometer.js. Switch from mean
3623         values to "average" objects which can hold stdev. Move graph button and
3624         calculation of noisy measurements here. Sophisticated header processing
3625         is not needed in release suite.
3626         (populateTable): Use DeveloperResultsTable.
3627         * Animometer/resources/debug-runner/graph.js: Pull time graph creation to
3628         its own function, and add a new onGraphTypeChanged handler in preparation
3629         of a complexity graph to be added later.
3630         * Animometer/resources/runner/animometer.js:
3631         (ResultsTable): Simplify to just handle test names and scores.
3632
3633 2016-02-07  Jon Lee  <jonlee@apple.com>
3634
3635         Tests: reuse objects already made.
3636
3637         Avoid thrash of object creation and removal by maintaining an index that
3638         moves along the array as the adjust values change. If the tune value
3639         requires more objects than the maximum size of the object array, then create
3640         new objects. This means that the object array size never decreases.
3641
3642         * Animometer/tests/master/resources/canvas-stage.js: Maintain a separate
3643         offsetIndex. For these tests, we want to avoid drawing the oldest objects,
3644         so the scene will draw the object at offsetIndex to the end of the array.
3645         (tune): Reverse the logic since "removal" of objects is much simpler and
3646         involves simply changing the offsetIndex.
3647         (animate): Update the for loop to draw from offsetIndex to the end.
3648         (complexity): Update the definition.
3649         * Animometer/tests/master/resources/canvas-tests.js: Maintain a separate
3650         offsetIndex. For these tests, we want to avoid drawing the newest objects,
3651         so the scene will draw the object at index 0 to the object at offsetIndex.
3652         (SimpleCanvasStage.animate): Fly-by removal of local stage variable,
3653         which is unneeded. Update the for loop to draw from offsetIndex to the end.
3654         * Animometer/tests/simple/resources/simple-canvas-paths.js:
3655         (SimpleCanvasStage.animate): Update the for loop to draw from 0 to
3656         offsetIndex.
3657         * Animometer/tests/simple/resources/simple-canvas.js:
3658         (tune): Update logic. Here, offsetIndex represents the boundary of the last
3659         index to render.
3660         (animate): Update the for loop to draw from 0 to offsetIndex.
3661         (complexity): Update the definition.
3662
3663 2016-02-07  Jon Lee  <jonlee@apple.com>
3664
3665         Tests: refactor and update styles.
3666
3667         * Animometer/tests/resources/main.js: Add helper methods that return
3668         a color that hue rotates based on the date, and a counter value that
3669         increases based on the date. Fix randomInt() to not bias against the min and
3670         max values.
3671
3672         * Animometer/tests/master/resources/canvas-tests.js: Use new helper methods.
3673         * Animometer/tests/master/resources/dom-particles.js: Ditto.
3674         * Animometer/tests/master/resources/particles.js: Ditto.
3675         * Animometer/tests/simple/resources/simple-canvas-paths.js: Refactor to
3676         use a rotating color instead of a random color. The fast switching of color
3677         is too vivid to watch.
3678
3679         * Animometer/tests/bouncing-particles/resources/bouncing-svg-shapes.js:
3680         (BouncingSvgParticlesStage.call.createGradient): Fix the gradient so
3681         that the last stop is located at the end.
3682
3683 2016-02-07  Jon Lee  <jonlee@apple.com>
3684
3685         Refactor tune() to not return the complexity of the scene.
3686
3687         We have stage.complexity() now, so returning the complexity through tune
3688         is unnecessary.
3689
3690         * Animometer/tests/bouncing-particles/resources/bouncing-particles.js:
3691         * Animometer/tests/master/resources/canvas-stage.js:
3692         * Animometer/tests/master/resources/particles.js:
3693         * Animometer/tests/misc/resources/canvas-electrons.js:
3694         * Animometer/tests/misc/resources/canvas-stars.js:
3695         * Animometer/tests/resources/main.js:
3696         * Animometer/tests/simple/resources/simple-canvas.js:
3697         * Animometer/tests/simple/resources/tiled-canvas-image.js:
3698         * Animometer/tests/template/resources/template-canvas.js:
3699         * Animometer/tests/template/resources/template-css.js:
3700         * Animometer/tests/template/resources/template-svg.js:
3701         * Animometer/tests/text/resources/layering-text.js:
3702
3703 2016-02-07  Jon Lee  <jonlee@apple.com>
3704
3705         Make the fixed controller a step controller instead. Halfway through the test
3706         it will bump up the complexity 4-fold. Calculate the step timestamp using options
3707         instead of a separate parameter to the Controller constructor.
3708
3709         * Animometer/developer.html: Change value to "step"
3710         * Animometer/resources/debug-runner/animometer.js:
3711         (window.suitesManager.updateEditsElementsState): Show number inputs when set to "step".
3712         * Animometer/tests/resources/main.js:
3713         (update): Provide a hook for subclasses to tune.
3714         (StepController): Maintain a flag determining whether we've stepped, and the time
3715         we should step.
3716         (Benchmark): Use the new StepController.
3717
3718 2016-02-07  Jon Lee  <jonlee@apple.com>
3719
3720         Adjust the FPS graph scale.
3721
3722         Instead of making the FPS graph linearly scale, scale it based on the frame length,
3723         but show the data in terms of FPS. Because it is inversely proportional, and most
3724         of the data never gets below 20, concentrate the axis from 20-60 FPS, since otherwise
3725         over half of the available graph space ends up blank.
3726
3727         This means we should convert all of the FPS data to frame length data.
3728
3729         * Animometer/resources/debug-runner/graph.js: Update the domain to be based on
3730         frame length in milliseconds instead of FPS. Update the cursor to consider all of the
3731         values being shown, and then pick the min and max values to represent the length of the
3732         cursor.
3733         * Animometer/resources/runner/animometer.js:
3734         * Animometer/resources/strings.js:
3735         * Animometer/tests/resources/main.js:
3736         (processSamples): Add the ability to only sample a range of the data instead of everything
3737         after an offset index. Update sampler to record the frame lengths instead of the frame
3738         rate.
3739
3740 2016-02-07  Jon Lee  <jonlee@apple.com>
3741
3742         Add option to use different methods for retrieving a timestamp.
3743
3744         * Animometer/developer.html: Add performance.now and Date.now options.
3745         * Animometer/resources/runner/animometer.js: Default to performance.now.
3746         (window.benchmarkController.startBenchmark):
3747         * Animometer/tests/resources/main.js: Tie the desired method to _getTimestamp.
3748         (run): Use _getTimestamp.
3749         (_animateLoop): Ditto.
3750
3751 2016-02-07  Jon Lee  <jonlee@apple.com>
3752
3753         Allow adding any number of markers to the graph. The markers can be labeled
3754         and contain timestamp and sample index data. Make it a part of the controller
3755         rather than keeping it in the sampler.
3756
3757         * Animometer/resources/debug-runner/animometer.css: Add styles for markers
3758         * Animometer/resources/debug-runner/graph.js: Create the markers and add
3759         text labels.
3760         * Animometer/resources/runner/animometer.js: Assume the samplingTimeOffset
3761         is just one of the marks provided.
3762         * Animometer/resources/strings.js: Add Strings.json.marks.
3763         * Animometer/tests/resources/main.js:
3764         (Controller): Keep marks here. They are keyed by the marker name, so no two
3765         markers should have the same name.
3766         (recordFirstSample): Refactor to use mark.
3767         (mark): Allows for arbitrary data if needed later. The timestamp maintained
3768         is relative to the absolute start timestamp.
3769         (containsMark): Checks whether a mark with a specific comment exists.
3770         (processSamples): Removes the _startTimestamp offset from the marks before
3771         setting it in results.
3772         * Animometer/tests/resources/sampler.js: Remove marks.
3773
3774 2016-02-07  Jon Lee  <jonlee@apple.com>
3775
3776         Get rid of options member variable in Benchmark.
3777
3778         Options are only needed when initializing the stage or benchmark, so there's no
3779         need to also keep a reference to it.
3780
3781         * Animometer/tests/resources/main.js: Get rid of options variable in Benchmark.
3782         Pass options to Controllers and Stages.
3783         (Controller.Utilities.createClass):
3784         (Benchmark.Utilities.createClass):
3785         (get options): Deleted.
3786
3787         * Animometer/tests/bouncing-particles/resources/bouncing-canvas-images.js:
3788         * Animometer/tests/bouncing-particles/resources/bouncing-canvas-particles.js:
3789         * Animometer/tests/bouncing-particles/resources/bouncing-canvas-shapes.js:
3790         * Animometer/tests/bouncing-particles/resources/bouncing-css-images.js:
3791         * Animometer/tests/bouncing-particles/resources/bouncing-css-shapes.js:
3792         * Animometer/tests/bouncing-particles/resources/bouncing-particles.js:
3793         * Animometer/tests/bouncing-particles/resources/bouncing-svg-images.js:
3794         * Animometer/tests/bouncing-particles/resources/bouncing-svg-shapes.js:
3795         * Animometer/tests/master/resources/canvas-stage.js:
3796         * Animometer/tests/master/resources/canvas-tests.js:
3797         * Animometer/tests/master/resources/particles.js:
3798         * Animometer/tests/misc/resources/canvas-electrons.js:
3799         * Animometer/tests/misc/resources/canvas-stars.js:
3800         * Animometer/tests/misc/resources/compositing-transforms.js:
3801         * Animometer/tests/simple/resources/simple-canvas-paths.js:
3802         * Animometer/tests/simple/resources/tiled-canvas-image.js:
3803         * Animometer/tests/template/resources/template-canvas.js:
3804         * Animometer/tests/template/resources/template-css.js:
3805         * Animometer/tests/template/resources/template-svg.js:
3806         * Animometer/tests/text/resources/layering-text.js:
3807
3808 2016-02-07  Jon Lee  <jonlee@apple.com>
3809
3810         Update how the benchmark is run
3811         https://bugs.webkit.org/show_bug.cgi?id=153960
3812
3813         Provisionally reviewed by Said Abou-Hallawa.
3814
3815         Introduce the notion of a Controller. It is responsible for recording, updating,
3816         and processing the statistics and complexity of the benchmark. This allows
3817         plugging in different Controllers.
3818
3819         This strips most of the functionality from Animator and BenchmarkState, so fold
3820         what's left into Benchmark. Now, Benchmarks only own a stage and a controller, but
3821         are responsible for driving the animation loop.
3822
3823         Rewrite Animator._shouldRequestAnotherFrame into two different Controllers. One
3824         maintains a fixed complexity, and the other adapts the complexity to meet a
3825         fixed FPS.
3826
3827         Fix the Kalman estimator to be modeled on a scalar variable with no model.
3828
3829         * Animometer/tests/resources/main.js: Remove BenchmarkState and Animator, and
3830         replace it with a Controller. Add a FixedController and refactor the previous controller
3831         to an AdaptiveController.
3832
3833         (Controller): Controllers own the estimator and the sampler. When a new frame is
3834         displayed, the animation loop calls update(). The estimator and sampler record
3835         stats, then tune. Samplers can track multiple series of data. The basic controller
3836         tracks timestamp, complexity, and estimated frame rate.
3837                 The Kalman estimation is based on the frame length rather than the frame
3838         rate. Because FPS is inverse proportional to frame length, in the case where the measured
3839         frame length is very small, the FPS ends up being a wildly large number (in the order of
3840         600-1000 "FPS"), and it pulls the estimator up drastically enough that it takes a while
3841         for it to settle back down. Using frame length reduces the impact of these spikes.
3842                 Converging the estimation takes enough time to avoid initializing it immediately
3843         when the benchmark starts. Instead, the benchmark runs for a brief period of time (100ms)
3844         before running it in earnest. Allow controllers an opportunity to set the complexity
3845         before starting recording.
3846                 When the benchmark is complete, the controller has an opportunity to process
3847         the samples. The default implementation calculates the raw FPS based on the time
3848         difference of the samples, and calculates the complexity score. This is moved from
3849         Benchmark.processSamples.
3850
3851         (Controller): Initialize timestamps. These are at first relative to the start of the
3852         benchmark, but are offset by the absolute start time during start(). By default maintain
3853         3 data series, but subclasses can override.
3854         (start): Calls recordFirstSample() for subclasses to override if needed.
3855         (recordFirstSample): For basic controller, start sampling at the beginning.