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