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