Rename the directory and the files of MotionMark from Animometer to MotionMark
[WebKit-https.git] / PerformanceTests / ChangeLog
1 2017-01-06  Said Abou-Hallawa  <sabouhallawa@apple.com>
2
3         Rename the directory and the files of MotionMark from Animometer to MotionMark
4         https://bugs.webkit.org/show_bug.cgi?id=166659
5
6         Reviewed by Jon Lee.
7         
8         Rename the directory and the files of the benchmark to its new name.
9
10         * MotionMark/developer.html: Renamed from PerformanceTests/Animometer/developer.html.
11         * MotionMark/index.html: Renamed from PerformanceTests/Animometer/index.html.
12         * MotionMark/resources/debug-runner/d3.min.js: Renamed from PerformanceTests/Animometer/resources/debug-runner/d3.min.js.
13         * MotionMark/resources/debug-runner/graph.js: Renamed from PerformanceTests/Animometer/resources/debug-runner/graph.js.
14         * MotionMark/resources/debug-runner/motionmark.css: Renamed from PerformanceTests/Animometer/resources/debug-runner/animometer.css.
15         * MotionMark/resources/debug-runner/motionmark.js: Renamed from PerformanceTests/Animometer/resources/debug-runner/animometer.js.
16         * MotionMark/resources/debug-runner/tests.js: Renamed from PerformanceTests/Animometer/resources/debug-runner/tests.js.
17         * MotionMark/resources/extensions.js: Renamed from PerformanceTests/Animometer/resources/extensions.js.
18         * MotionMark/resources/runner/benchmark-runner.js: Renamed from PerformanceTests/Animometer/resources/runner/benchmark-runner.js.
19         * MotionMark/resources/runner/crystal.svg: Renamed from PerformanceTests/Animometer/resources/runner/crystal.svg.
20         * MotionMark/resources/runner/lines.svg: Renamed from PerformanceTests/Animometer/resources/runner/lines.svg.
21         * MotionMark/resources/runner/logo.svg: Renamed from PerformanceTests/Animometer/resources/runner/logo.svg.
22         * MotionMark/resources/runner/motionmark.css: Renamed from PerformanceTests/Animometer/resources/runner/animometer.css.
23         * MotionMark/resources/runner/motionmark.js: Renamed from PerformanceTests/Animometer/resources/runner/animometer.js.
24         * MotionMark/resources/runner/tests.js: Renamed from PerformanceTests/Animometer/resources/runner/tests.js.
25         * MotionMark/resources/statistics.js: Renamed from PerformanceTests/Animometer/resources/statistics.js.
26         * MotionMark/resources/strings.js: Renamed from PerformanceTests/Animometer/resources/strings.js.
27         * MotionMark/tests/3d/resources/webgl.js: Renamed from PerformanceTests/Animometer/tests/3d/resources/webgl.js.
28         * MotionMark/tests/3d/webgl.html: Renamed from PerformanceTests/Animometer/tests/3d/webgl.html.
29         * MotionMark/tests/bouncing-particles/bouncing-canvas-images.html: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/bouncing-canvas-images.html.
30         * MotionMark/tests/bouncing-particles/bouncing-canvas-shapes.html: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/bouncing-canvas-shapes.html.
31         * MotionMark/tests/bouncing-particles/bouncing-css-images.html: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/bouncing-css-images.html.
32         * MotionMark/tests/bouncing-particles/bouncing-css-shapes.html: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/bouncing-css-shapes.html.
33         * MotionMark/tests/bouncing-particles/bouncing-svg-images.html: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/bouncing-svg-images.html.
34         * MotionMark/tests/bouncing-particles/bouncing-svg-shapes.html: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/bouncing-svg-shapes.html.
35         * MotionMark/tests/bouncing-particles/bouncing-tagged-images.html: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/bouncing-tagged-images.html.
36         * MotionMark/tests/bouncing-particles/resources/bouncing-canvas-images.js: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/bouncing-canvas-images.js.
37         * MotionMark/tests/bouncing-particles/resources/bouncing-canvas-particles.js: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/bouncing-canvas-particles.js.
38         * MotionMark/tests/bouncing-particles/resources/bouncing-canvas-shapes.js: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/bouncing-canvas-shapes.js.
39         * MotionMark/tests/bouncing-particles/resources/bouncing-css-images.js: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/bouncing-css-images.js.
40         * MotionMark/tests/bouncing-particles/resources/bouncing-css-shapes.js: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/bouncing-css-shapes.js.
41         * MotionMark/tests/bouncing-particles/resources/bouncing-svg-images.js: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/bouncing-svg-images.js.
42         * MotionMark/tests/bouncing-particles/resources/bouncing-svg-particles.js: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/bouncing-svg-particles.js.
43         * MotionMark/tests/bouncing-particles/resources/bouncing-svg-shapes.js: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/bouncing-svg-shapes.js.
44         * MotionMark/tests/bouncing-particles/resources/bouncing-tagged-images.js: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/bouncing-tagged-images.js.
45         * MotionMark/tests/bouncing-particles/resources/image1.jpg: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/image1.jpg.
46         * MotionMark/tests/bouncing-particles/resources/image2.jpg: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/image2.jpg.
47         * MotionMark/tests/bouncing-particles/resources/image3.jpg: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/image3.jpg.
48         * MotionMark/tests/bouncing-particles/resources/image4.jpg: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/image4.jpg.
49         * MotionMark/tests/bouncing-particles/resources/image5.jpg: Renamed from PerformanceTests/Animometer/tests/bouncing-particles/resources/image5.jpg.
50         * MotionMark/tests/dom/compositing-transforms.html: Renamed from PerformanceTests/Animometer/tests/dom/compositing-transforms.html.
51         * MotionMark/tests/dom/focus.html: Renamed from PerformanceTests/Animometer/tests/dom/focus.html.
52         * MotionMark/tests/dom/leaves.html: Renamed from PerformanceTests/Animometer/tests/dom/leaves.html.
53         * MotionMark/tests/dom/particles.html: Renamed from PerformanceTests/Animometer/tests/dom/particles.html.
54         * MotionMark/tests/dom/resources/compositing-transforms.js: Renamed from PerformanceTests/Animometer/tests/dom/resources/compositing-transforms.js.
55         * MotionMark/tests/dom/resources/dom-particles.js: Renamed from PerformanceTests/Animometer/tests/dom/resources/dom-particles.js.
56         * MotionMark/tests/dom/resources/focus.js: Renamed from PerformanceTests/Animometer/tests/dom/resources/focus.js.
57         * MotionMark/tests/dom/resources/leaves.js: Renamed from PerformanceTests/Animometer/tests/dom/resources/leaves.js.
58         * MotionMark/tests/master/canvas-stage.html: Renamed from PerformanceTests/Animometer/tests/master/canvas-stage.html.
59         * MotionMark/tests/master/focus.html: Renamed from PerformanceTests/Animometer/tests/master/focus.html.
60         * MotionMark/tests/master/image-data.html: Renamed from PerformanceTests/Animometer/tests/master/image-data.html.
61         * MotionMark/tests/master/leaves.html: Renamed from PerformanceTests/Animometer/tests/master/leaves.html.
62         * MotionMark/tests/master/multiply.html: Renamed from PerformanceTests/Animometer/tests/master/multiply.html.
63         * MotionMark/tests/master/resources/canvas-stage.js: Renamed from PerformanceTests/Animometer/tests/master/resources/canvas-stage.js.
64         * MotionMark/tests/master/resources/canvas-tests.js: Renamed from PerformanceTests/Animometer/tests/master/resources/canvas-tests.js.
65         * MotionMark/tests/master/resources/compass.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/compass.svg.
66         * MotionMark/tests/master/resources/compass100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/compass100.png.
67         * MotionMark/tests/master/resources/console.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/console.svg.
68         * MotionMark/tests/master/resources/console100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/console100.png.
69         * MotionMark/tests/master/resources/contribute.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/contribute.svg.
70         * MotionMark/tests/master/resources/contribute100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/contribute100.png.
71         * MotionMark/tests/master/resources/debugger.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/debugger.svg.
72         * MotionMark/tests/master/resources/debugger100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/debugger100.png.
73         * MotionMark/tests/master/resources/focus.js: Renamed from PerformanceTests/Animometer/tests/master/resources/focus.js.
74         * MotionMark/tests/master/resources/image-data.js: Renamed from PerformanceTests/Animometer/tests/master/resources/image-data.js.
75         * MotionMark/tests/master/resources/inspector.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/inspector.svg.
76         * MotionMark/tests/master/resources/inspector100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/inspector100.png.
77         * MotionMark/tests/master/resources/layout.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/layout.svg.
78         * MotionMark/tests/master/resources/layout100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/layout100.png.
79         * MotionMark/tests/master/resources/leaves.js: Renamed from PerformanceTests/Animometer/tests/master/resources/leaves.js.
80         * MotionMark/tests/master/resources/multiply.js: Renamed from PerformanceTests/Animometer/tests/master/resources/multiply.js.
81         * MotionMark/tests/master/resources/particles.js: Renamed from PerformanceTests/Animometer/tests/master/resources/particles.js.
82         * MotionMark/tests/master/resources/performance.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/performance.svg.
83         * MotionMark/tests/master/resources/performance100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/performance100.png.
84         * MotionMark/tests/master/resources/script.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/script.svg.
85         * MotionMark/tests/master/resources/script100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/script100.png.
86         * MotionMark/tests/master/resources/shortcuts.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/shortcuts.svg.
87         * MotionMark/tests/master/resources/shortcuts100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/shortcuts100.png.
88         * MotionMark/tests/master/resources/standards.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/standards.svg.
89         * MotionMark/tests/master/resources/standards100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/standards100.png.
90         * MotionMark/tests/master/resources/storage.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/storage.svg.
91         * MotionMark/tests/master/resources/storage100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/storage100.png.
92         * MotionMark/tests/master/resources/styles.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/styles.svg.
93         * MotionMark/tests/master/resources/styles100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/styles100.png.
94         * MotionMark/tests/master/resources/svg-particles.js: Renamed from PerformanceTests/Animometer/tests/master/resources/svg-particles.js.
95         * MotionMark/tests/master/resources/text.js: Renamed from PerformanceTests/Animometer/tests/master/resources/text.js.
96         * MotionMark/tests/master/resources/timeline.svg: Renamed from PerformanceTests/Animometer/tests/master/resources/timeline.svg.
97         * MotionMark/tests/master/resources/timeline100.png: Renamed from PerformanceTests/Animometer/tests/master/resources/timeline100.png.
98         * MotionMark/tests/master/svg-particles.html: Renamed from PerformanceTests/Animometer/tests/master/svg-particles.html.
99         * MotionMark/tests/master/text.html: Renamed from PerformanceTests/Animometer/tests/master/text.html.
100         * MotionMark/tests/resources/main.js: Renamed from PerformanceTests/Animometer/tests/resources/main.js.
101         * MotionMark/tests/resources/math.js: Renamed from PerformanceTests/Animometer/tests/resources/math.js.
102         * MotionMark/tests/resources/stage.css: Renamed from PerformanceTests/Animometer/tests/resources/stage.css.
103         * MotionMark/tests/resources/star.svg: Renamed from PerformanceTests/Animometer/tests/resources/star.svg.
104         * MotionMark/tests/resources/yin-yang.png: Renamed from PerformanceTests/Animometer/tests/resources/yin-yang.png.
105         * MotionMark/tests/resources/yin-yang.svg: Renamed from PerformanceTests/Animometer/tests/resources/yin-yang.svg.
106         * MotionMark/tests/simple/resources/simple-canvas-paths.js: Renamed from PerformanceTests/Animometer/tests/simple/resources/simple-canvas-paths.js.
107         * MotionMark/tests/simple/resources/simple-canvas.js: Renamed from PerformanceTests/Animometer/tests/simple/resources/simple-canvas.js.
108         * MotionMark/tests/simple/resources/tiled-canvas-image.js: Renamed from PerformanceTests/Animometer/tests/simple/resources/tiled-canvas-image.js.
109         * MotionMark/tests/simple/simple-canvas-paths.html: Renamed from PerformanceTests/Animometer/tests/simple/simple-canvas-paths.html.
110         * MotionMark/tests/simple/tiled-canvas-image.html: Renamed from PerformanceTests/Animometer/tests/simple/tiled-canvas-image.html.
111         * MotionMark/tests/template/resources/template-canvas.js: Renamed from PerformanceTests/Animometer/tests/template/resources/template-canvas.js.
112         * MotionMark/tests/template/resources/template-css.js: Renamed from PerformanceTests/Animometer/tests/template/resources/template-css.js.
113         * MotionMark/tests/template/resources/template-svg.js: Renamed from PerformanceTests/Animometer/tests/template/resources/template-svg.js.
114         * MotionMark/tests/template/template-canvas.html: Renamed from PerformanceTests/Animometer/tests/template/template-canvas.html.
115         * MotionMark/tests/template/template-css.html: Renamed from PerformanceTests/Animometer/tests/template/template-css.html.
116         * MotionMark/tests/template/template-svg.html: Renamed from PerformanceTests/Animometer/tests/template/template-svg.html.
117
118 2016-12-13  Brady Eidson  <beidson@apple.com>
119
120         Add a PerfTest targeting IDBObjectStore.get().
121         https://bugs.webkit.org/show_bug.cgi?id=165816
122
123         Reviewed by Alex Christensen.
124
125         * IndexedDB/objectstore-get.html: Added.
126
127 2016-12-13  Brady Eidson  <beidson@apple.com>
128
129         Add a PerfTest targeting IDBIndex.get().
130         https://bugs.webkit.org/show_bug.cgi?id=165803
131
132         Reviewed by Alex Christensen.
133
134         * IndexedDB/index-get.html: Added.
135
136 2016-12-12  Brady Eidson  <beidson@apple.com>
137
138         More IndexedDB perf tests.
139         https://bugs.webkit.org/show_bug.cgi?id=165634
140
141         Reviewed by Sam Weinig.
142
143         * IndexedDB/index-multientry.html:
144         * IndexedDB/large-number-of-inserts-responsiveness.html: Added.
145         * IndexedDB/large-number-of-inserts.html: Added.
146         * IndexedDB/objectstore-cursor.html: Added. Runtime cut in 1/4th compared to r209672.
147
148         * resources/runner.js:
149
150 2016-12-12  Commit Queue  <commit-queue@webkit.org>
151
152         Unreviewed, rolling out r209672.
153         https://bugs.webkit.org/show_bug.cgi?id=165766
154
155         IndexedDB/objectstore-cursor.html perf test always times out
156         (Requested by ap on #webkit).
157
158         Reverted changeset:
159
160         "More IndexedDB perf tests."
161         https://bugs.webkit.org/show_bug.cgi?id=165634
162         http://trac.webkit.org/changeset/209672
163
164 2016-12-10  Brady Eidson  <beidson@apple.com>
165
166         More IndexedDB perf tests.
167         https://bugs.webkit.org/show_bug.cgi?id=165634
168
169         Reviewed by Sam Weinig.
170
171         * IndexedDB/index-multientry.html:
172         * IndexedDB/large-number-of-inserts-responsiveness.html: Added.
173         * IndexedDB/large-number-of-inserts.html: Added.
174         * IndexedDB/objectstore-cursor.html: Added.
175         
176         * resources/runner.js: Add "track responsiveness" functionality to PerfTestRunner.
177
178 2016-12-06  Filip Pizlo  <fpizlo@apple.com>
179
180         Concurrent GC should be stable enough to land enabled
181         https://bugs.webkit.org/show_bug.cgi?id=164990
182
183         Reviewed by Geoffrey Garen.
184         
185         Made CDjs more configurable and refined the "large.js" configuration. I was using that one and
186         the new "long.js" configuration to tune concurrent eden GCs.
187         
188         Added a new way of running Splay in browser, which using chartjs to plot the execution times of
189         2000 iterations. This includes the minified chartjs.
190
191         * JetStream/Octane2/splay-detail.html: Added.
192         * JetStream/cdjs/benchmark.js:
193         (benchmarkImpl):
194         (benchmark):
195         * JetStream/cdjs/long.js: Added.
196
197 2016-12-07  Brady Eidson  <beidson@apple.com>
198
199         Add IDB perf tests stressing key size.
200         https://bugs.webkit.org/show_bug.cgi?id=165567
201
202         Reviewed by Alex Christensen.
203
204         * IndexedDB/large-array-keys.html: Added.
205         * IndexedDB/large-binary-keys.html: Added.
206         * IndexedDB/large-string-keys.html: Added.
207
208 2016-12-05  Brady Eidson  <beidson@apple.com>
209
210         Add an IndexedDB perf test to PerformanceTests.
211         https://bugs.webkit.org/show_bug.cgi?id=165430
212
213         Reviewed by Alex Christensen.
214
215         * IndexedDB/index-multientry.html: Added.
216
217 2016-12-02  Filip Pizlo  <fpizlo@apple.com>
218
219         ES6SampleBench should report an average for Steady State so that all of the numbers are comparable
220         https://bugs.webkit.org/show_bug.cgi?id=165325
221
222         Reviewed by Saam Barati.
223
224         This makes all of the numbers that ES6SampleBench reports comparable to each other: they all speak of the time
225         it took to run an iteration of something.
226
227         * ES6SampleBench/results.js:
228         (Results.prototype.reportResult):
229
230 2016-11-18  Jer Noble  <jer.noble@apple.com>
231
232         Add some performance tests for various aspects of media loading and playback.
233         https://bugs.webkit.org/show_bug.cgi?id=164977
234
235         Reviewed by Darin Adler.
236
237         Add new tests for a few aspects of media playback; namely: how quickly  media elements fire
238         the "canplaythrough" event when all data is immediately available, how quickly HLS playback
239         switches resolutions, and at what playback rate MSE-backed video can play without dropping
240         frames. Skip these tests by default, as they're not runnable on all ports, and they require
241         a webserver (run-webkit-httpd) to complete on macOS.
242
243         * Media/HLSCanPlayThrough.html: Added.
244         * Media/HLSGearChange.html: Added.
245         * Media/MSECanPlayThrough.html: Added.
246         * Media/MSEPlaybackRate.html: Added.
247         * Media/StandardCanPlayThrough.html: Added.
248         * Media/hls/1080p/iframe_index.m3u8: Added.
249         * Media/hls/1080p/prog_index.m3u8: Added.
250         * Media/hls/1080p/test.ts: Added.
251         * Media/hls/480p/iframe_index.m3u8: Added.
252         * Media/hls/480p/prog_index.m3u8: Added.
253         * Media/hls/480p/test.ts: Added.
254         * Media/hls/720p/iframe_index.m3u8: Added.
255         * Media/hls/720p/prog_index.m3u8: Added.
256         * Media/hls/720p/test.ts: Added.
257         * Media/hls/720p/test.ts.back: Added.
258         * Media/hls/index.m3u8: Added.
259         * Media/media-source-loader.js: Added.
260         (MediaSourceLoader):
261         (MediaSourceLoader.prototype.loadManifest):
262         (MediaSourceLoader.prototype.loadManifestSucceeded):
263         (MediaSourceLoader.prototype.loadManifestFailed):
264         (MediaSourceLoader.prototype.loadMediaData):
265         (MediaSourceLoader.prototype.loadMediaDataSucceeded):
266         (MediaSourceLoader.prototype.loadMediaDataFailed):
267         (MediaSourceLoader.prototype.get type):
268         (MediaSourceLoader.prototype.get duration):
269         (MediaSourceLoader.prototype.get initSegment):
270         (MediaSourceLoader.prototype.get mediaSegmentsLength):
271         (MediaSourceLoader.prototype.mediaSegments):
272         (MediaSourceLoader.prototype.get everyMediaSegment):
273         * Media/test-fragmented-video.json: Added.
274         * Media/test-fragmented-video.mp4: Added.
275         * Media/test.mp4: Added.
276         * Skipped:
277
278 2016-11-18  Filip Pizlo  <fpizlo@apple.com>
279
280         Concurrent GC should be able to run splay in debug mode and earley/raytrace in release mode with no perf regression
281         https://bugs.webkit.org/show_bug.cgi?id=164282
282
283         Reviewed by Geoffrey Garen and Oliver Hunt.
284         
285         CDjs is a fun benchmark for stressing concurrent GCs, but to really give the GC a good
286         workout you need to increase the amount of work that the test does. This adds a second
287         configuration of the benchmark that has more aircraft. It uses much more memory and causes us
288         to do more GCs and those GCs take longer.
289
290         * JetStream/cdjs/benchmark.js:
291         (benchmarkImpl):
292         (benchmark):
293         * JetStream/cdjs/large.js: Added.
294
295 2016-11-14  Filip Pizlo  <fpizlo@apple.com>
296
297         Unreviewed, revert unintended change.
298
299         * ES6SampleBench/Air/benchmark.js:
300
301 2016-10-03  Saam Barati  <sbarati@apple.com>
302
303         MapHash should speculate on the type of its child node
304         https://bugs.webkit.org/show_bug.cgi?id=161922
305
306         Reviewed by Filip Pizlo.
307
308         I gave the main generator a name so it's easier to see what
309         it is when using the sampling profiler.
310
311         * ES6SampleBench/Basic/ast.js:
312         (Basic.Program):
313
314 2016-09-19  Sergio Villar Senin  <svillar@igalia.com>
315
316         [css-grid] Remove the x2 computation of row sizes with indefinite heights
317         https://bugs.webkit.org/show_bug.cgi?id=162150
318
319         Reviewed by Darin Adler.
320
321         Added a new test case which checks the layout performance of grids inside other grids, i.e,
322         grids acting both as grid container and grid item.
323
324         * Layout/nested-grid.html: Added.
325
326 2016-09-09  Simon Fraser  <simon.fraser@apple.com>
327
328         Perf test Animation/css-accelerated-animation.html failing
329         https://bugs.webkit.org/show_bug.cgi?id=161795
330
331         Skip Animation/css-accelerated-animation.html since this only works as a MobileSafari
332         perf test.
333
334         * Skipped:
335
336 2016-09-08  Simon Fraser  <simon.fraser@apple.com>
337
338         Add a content-animation test with accelerated CSS animations
339         https://bugs.webkit.org/show_bug.cgi?id=161776
340
341         Reviewed by Dean Jackson.
342
343         Add a test that measures the frame rate of accelerated CSS animations.
344
345         This is like css-animation.html (which animates 'left' and 'top') but animates
346         transforms instead. In order to get animation in X and Y we make two nested elements
347         and set transformX() on one, and transformY() on the other.
348
349         * Animation/css-accelerated-animation.html: Added.
350
351 2016-08-23  Saam Barati  <sbarati@apple.com>
352
353         It should be easy to run ES6SampleBench from the jsc shell
354         https://bugs.webkit.org/show_bug.cgi?id=161085
355
356         Reviewed by Yusuke Suzuki.
357
358         This patch makes ES6 sample bench runnable through the `jsc` shell.
359         To do that, you need to be in the PerformanceTests/ES6SampleBench
360         directory and run `jsc cli.js`. To make this work, the benchmark
361         is now aware of being run in two modes: via the browser, and via
362         the shell. Each entry point will set a variable `isInBrowser`,
363         and the benchmark will do different things based on if that
364         variable is true or false.
365
366         * ES6SampleBench/Air/benchmark.js:
367         * ES6SampleBench/Air/stress-test.js:
368         * ES6SampleBench/Basic/benchmark.js:
369         (runBenchmark):
370         * ES6SampleBench/Basic/stress-test.js:
371         * ES6SampleBench/air_benchmark.js:
372         * ES6SampleBench/basic_benchmark.js:
373         * ES6SampleBench/cli.js: Added.
374         (return.doRun):
375         (makeBenchmarkRunner):
376         * ES6SampleBench/driver.js:
377         (Driver):
378         (Driver.prototype._recomputeSummary):
379         (Driver.prototype._iterate.window.setTimeout):
380         (Driver.prototype._iterate):
381         (Driver.prototype._updateIterations):
382         * ES6SampleBench/glue.js:
383         (reportResult):
384         * ES6SampleBench/index.html:
385         * ES6SampleBench/results.js:
386         (Results):
387         (Results.prototype.reportRunning):
388         (Results.prototype.reportDone):
389         (Results.prototype.reportError):
390         * ES6SampleBench/stats.js:
391         (Stats):
392         (Stats.prototype.toString):
393         (Stats.prototype._update):
394
395 2016-08-01  Said Abou-Hallawa  <sabouhallawa@apple.com>
396
397         Add an option to run the MotionMark tests with classic tile size (512x512)
398         https://bugs.webkit.org/show_bug.cgi?id=160371
399
400         Reviewed by Darin Adler.
401
402         Setting the body to large size (3000x3000) and overflow to scroll forces
403         the classic tile size (512x512). This option does not affect the canvas
404         tests because the canvas is displayed on one tile regardless of its size.
405
406         * Animometer/developer.html:
407         * Animometer/resources/debug-runner/animometer.css:
408         (body.showing-test-container.tiles-big):
409         (body.showing-test-container.tiles-classic):
410         * Animometer/resources/debug-runner/animometer.js:
411         * Animometer/resources/runner/animometer.js:
412         (window.benchmarkController.startBenchmark):
413
414 2016-06-29  Filip Pizlo  <fpizlo@apple.com>
415
416         Generators violate bytecode liveness validation
417         https://bugs.webkit.org/show_bug.cgi?id=159279
418
419         Reviewed by Yusuke Suzuki.
420
421         Add Basic to our test harness.
422
423         Also made some cosmetic changes to the benchmark harness.
424
425         * ES6SampleBench/Basic/basic-tests.yaml: Added.
426         * ES6SampleBench/Basic/stress-test.js: Added.
427         (preciseTime):
428         * ES6SampleBench/driver.js:
429         (Driver):
430         (Driver.prototype.start):
431         (Driver.prototype.reportError):
432         * ES6SampleBench/glue.js:
433         * ES6SampleBench/index.html:
434
435 2016-06-28  Filip Pizlo  <fpizlo@apple.com>
436
437         ES6SampleBench should have a harness
438         https://bugs.webkit.org/show_bug.cgi?id=159246
439
440         Reviewed by Saam Barati.
441         
442         This adds a simple web harness for ES6SampleBench. It runs Air and Basic 10 times for 200
443         iterations each and reports three metrics:
444         
445         First iteration: the time it takes for the first iteration to run. This is the first
446         iteration after the benchmark is loaded into the iframe, so it's representative of what
447         would happen if one of these workloads only ran for a short time.
448         
449         Worst 2%: of the last 199 iterations, the average of the worst 2% iterations. If code like
450         any of these workloads was used in an important event handler, you'd want that code to run
451         well in the worst case in addition to having great throughput.
452         
453         Steady state: the total of the last 199 iterations. This is representative of what would
454         happen if you ran code like this for a long time.
455         
456         The total score is the geomean of the firstIteration/worstCase/steadyState numbers of the
457         two benchmarks.
458         
459         The harness does statistics using Student's T-distribution confidence intervals.
460
461         * ES6SampleBench/Basic/benchmark.js:
462         (Benchmark):
463         * ES6SampleBench/air_benchmark.js: Added.
464         * ES6SampleBench/basic_benchmark.js: Added.
465         * ES6SampleBench/driver.js: Added.
466         (Driver):
467         (Driver.prototype.addBenchmark):
468         (Driver.prototype.start):
469         (Driver.prototype.reportResult):
470         (Driver.prototype.reportError):
471         (Driver.prototype._recomputeSummary.Geomean):
472         (Driver.prototype._recomputeSummary.Geomean.prototype.add):
473         (Driver.prototype._recomputeSummary.Geomean.prototype.get result):
474         (Driver.prototype._recomputeSummary):
475         (Driver.prototype._iterate):
476         (Driver.prototype._updateIterations):
477         * ES6SampleBench/glue.js: Added.
478         * ES6SampleBench/index.html: Added.
479         * ES6SampleBench/results.js: Added.
480         (Results):
481         (Results.prototype.get benchmark):
482         (Results.prototype.reset):
483         (Results.prototype.reportRunning):
484         (Results.prototype.reportDone):
485         (Results.prototype.reportResult.averageAbovePercentile):
486         (Results.prototype.reportResult):
487         (Results.prototype.reportError):
488         * ES6SampleBench/stats.js: Added.
489         (Stats):
490         (Stats.prototype.reset):
491         (Stats.prototype.add):
492         (Stats.prototype.get numIterations):
493         (Stats.prototype.valueForIteration):
494         (Stats.get result.tDist):
495         (Stats.prototype.get result):
496         (Stats.prototype.toString):
497         (Stats.prototype._update):
498         * ES6SampleBench/style.css: Added.
499         (body):
500         (body, th, tr):
501         (h1, h2, h3, h4):
502         (h1):
503         (h2):
504         (h3):
505         (hr):
506         (address):
507         (img):
508         (.underline):
509         (ol.loweralpha):
510         (ol.upperalpha):
511         (ol.lowerroman):
512         (ol.upperroman):
513         (ol.arabic):
514         (.banner-link:link, .banner-link:visited):
515         (:link, :visited):
516         (h4 :link, h4 :visited, h5 :link, h5 :visited, h6 :link, h6 :visited):
517         (.anchor:link, .anchor:visited):
518         (* > .anchor:link, * > .anchor:visited):
519         (span:hover .anchor):
520         (a.forbidden, span.forbidden):
521         (a.missing:hover):
522         (a.closed:link, a.closed:visited, span.closed):
523         (pre):
524         (div.code):
525         (div.code pre):
526         (dt):
527         (dd):
528         (dd:last-child):
529         (.site-logo):
530         (.site-logo .tagline):
531         (table):
532         (#contents):
533         (p):
534         (p:last-child):
535         (th):
536         (td):
537
538 2016-06-28  Jon Lee  <jonlee@apple.com>
539
540         Update focus test
541         https://bugs.webkit.org/show_bug.cgi?id=159242
542         rdar://problem/27070007
543
544         Reviewed by Dean Jackson.
545         Provisionally reviewed by Said Abou-Hallawa.
546
547         Move previous test to dom suite, and update the test for better reporting of frame rate, although
548         it uses a different rendering path.
549
550         * Animometer/resources/debug-runner/tests.js: Add to dom suite.
551         * Animometer/tests/dom/focus.html: Copied from PerformanceTests/Animometer/tests/master/focus.html.
552         * Animometer/tests/dom/resources/focus.js: Copied from PerformanceTests/Animometer/tests/master/resources/focus.js.
553         * Animometer/tests/master/focus.html: Remove center element.
554         * Animometer/tests/master/resources/focus.js: Use narrower size range with smaller particles. Remove the
555         container elements. Inline getBlurValue and getOpacityValue since they are only called once.
556
557 2016-06-28  Filip Pizlo  <fpizlo@apple.com>
558
559         Move Air.js and Basic into ES6SampleBench
560
561         Rubber stamped by Geoffrey Garen.
562         
563         We want to group our ES6 benchmarks together, and eventually, we want to give them a
564         common harness.
565
566         * Air.js: Removed.
567         * Air.js/README.md: Removed.
568         * Air.js/airjs-tests.yaml: Removed.
569         * Air.js/all.js: Removed.
570         * Air.js/allocate_stack.js: Removed.
571         * Air.js/arg.js: Removed.
572         * Air.js/basic_block.js: Removed.
573         * Air.js/benchmark.js: Removed.
574         * Air.js/code.js: Removed.
575         * Air.js/custom.js: Removed.
576         * Air.js/frequented_block.js: Removed.
577         * Air.js/insertion_set.js: Removed.
578         * Air.js/inst.js: Removed.
579         * Air.js/liveness.js: Removed.
580         * Air.js/make_dist.sh: Removed.
581         * Air.js/opcode.js: Removed.
582         * Air.js/payload-airjs-ACLj8C.js: Removed.
583         * Air.js/payload-gbemu-executeIteration.js: Removed.
584         * Air.js/payload-imaging-gaussian-blur-gaussianBlur.js: Removed.
585         * Air.js/payload-typescript-scanIdentifier.js: Removed.
586         * Air.js/reg.js: Removed.
587         * Air.js/stack_slot.js: Removed.
588         * Air.js/stress-test.js: Removed.
589         * Air.js/strip-hash.rb: Removed.
590         * Air.js/symbols.js: Removed.
591         * Air.js/test.html: Removed.
592         * Air.js/test.js: Removed.
593         * Air.js/tmp.js: Removed.
594         * Air.js/tmp_base.js: Removed.
595         * Air.js/util.js: Removed.
596         * Basic: Removed.
597         * Basic/ast.js: Removed.
598         * Basic/basic.js: Removed.
599         * Basic/benchmark.js: Removed.
600         * Basic/caseless_map.js: Removed.
601         * Basic/lexer.js: Removed.
602         * Basic/number.js: Removed.
603         * Basic/parser.js: Removed.
604         * Basic/random.js: Removed.
605         * Basic/state.js: Removed.
606         * Basic/test.html: Removed.
607         * Basic/test.js: Removed.
608         * Basic/util.js: Removed.
609         * ES6SampleBench: Added.
610         * ES6SampleBench/Air: Copied from PerformanceTests/Air.js.
611         * ES6SampleBench/Basic: Copied from PerformanceTests/Basic.
612         * Skipped:
613
614 2016-06-24  Filip Pizlo  <fpizlo@apple.com>
615
616         Add a ES6 generator benchmark
617         https://bugs.webkit.org/show_bug.cgi?id=159101
618
619         Rubber stamped by Keith Miller.
620         
621         This adds a Basic interpreter loosely based on ECMA-55:
622         
623         http://www.ecma-international.org/publications/files/ECMA-ST-WITHDRAWN/ECMA-55,%201st%20Edition,%20January%201978.pdf
624         
625         It includes a lexer that is a generator, a parser that uses regular expressions, and an AST
626         walk interpreter where the walking functions for statements are generators that call each
627         other with yield*. This enables the interpreter to look like an AST walk even though it can
628         yield at INPUT and PRINT statements.
629         
630         This also uses for-of, classes, Map, and WeakMap. It also uses deprecated-but-awesome RegExp
631         features like RegExp.lastMatch and RegExp.rightContext. I did it that way because this is
632         how I've always written lexers in dynamic languages; see offlineasm's lex() method in
633         parser.rb for example.
634         
635         The benchmark runs a handful of simple Basic programs. The longest-running one computes
636         prime numbers.
637         
638         Includes a command-line and web harness. On my machine it runs in 2-3 seconds.
639
640         * Basic: Added.
641         * Basic/ast.js: Added.
642         (Basic.NumberApply):
643         (Basic.Variable):
644         (Basic.Const):
645         (Basic.NumberPow):
646         (Basic.NumberMul):
647         (Basic.NumberDiv):
648         (Basic.NumberNeg):
649         (Basic.NumberAdd):
650         (Basic.NumberSub):
651         (Basic.StringVar):
652         (Basic.Equals):
653         (Basic.NotEquals):
654         (Basic.LessThan):
655         (Basic.GreaterThan):
656         (Basic.LessEqual):
657         (Basic.GreaterEqual):
658         (Basic.GoTo):
659         (Basic.GoSub):
660         (Basic.Def):
661         (Basic.Let):
662         (Basic.If):
663         (Basic.Return):
664         (Basic.Stop):
665         (Basic.On):
666         (sideState.shouldStop):
667         (Basic.For):
668         (Basic.Next):
669         (Basic.Print):
670         (Basic.Input):
671         (Basic.Read):
672         (Basic.Restore):
673         (Basic.Dim):
674         (Basic.Randomize):
675         (Basic.End):
676         (Basic.Program):
677         * Basic/basic.js: Added.
678         (prepare):
679         (simulate):
680         * Basic/benchmark.js: Added.
681         (Benchmark):
682         (Benchmark.prototype.runIteration.expect):
683         (Benchmark.prototype.runIteration):
684         (runBenchmark):
685         * Basic/caseless_map.js: Added.
686         (CaselessMap):
687         * Basic/lexer.js: Added.
688         (lex.consumeWhitespace):
689         (lex.consume):
690         (lex):
691         * Basic/lexer_test.js: Added.
692         * Basic/number.js: Added.
693         (NumberValue):
694         (NumberValue.prototype.get value):
695         (NumberValue.prototype.apply):
696         (NumberValue.prototype.leftApply):
697         (NumberValue.prototype.assign):
698         (NumberArray.):
699         (NumberArray):
700         (NumberArray.prototype.apply):
701         (NumberArray.prototype.leftApply):
702         (NumberFunction):
703         (NumberFunction.prototype.apply):
704         (NumberFunction.prototype.leftApply):
705         (NativeFunction):
706         (NativeFunction.prototype.apply):
707         (NativeFunction.prototype.leftApply):
708         * Basic/parser.js: Added.
709         (parse):
710         (parse.pushToken):
711         (parse.peekToken):
712         (parse.consumeKind):
713         (parse.consumeToken):
714         (parse.parseVariable):
715         (parse.parseNumericExpression.parsePrimary):
716         (parse.parseNumericExpression.parseFactor):
717         (parse.parseNumericExpression.parseTerm):
718         (parse.parseNumericExpression):
719         (parse.parseConstant):
720         (parse.parseStringExpression):
721         (parse.isStringExpression):
722         (parse.parseRelationalExpression):
723         (parse.parseNonNegativeInteger):
724         (parse.parseGoToStatement):
725         (parse.parseGoSubStatement):
726         (parse.parseStatement):
727         (parse.parseStatements):
728         * Basic/random.js: Added.
729         (createRNG):
730         (createRNGWithFixedSeed):
731         (createRNGWithRandomSeed):
732         * Basic/state.js: Added.
733         (State):
734         (State.prototype.getValue):
735         (State.prototype.getSideState):
736         (State.prototype.abort):
737         (State.prototype.validate):
738         * Basic/test.html: Added.
739         * Basic/test.js: Added.
740         * Basic/util.js: Added.
741         (this.performance.performance.now.currentTime):
742         (else.this.preciseTime.currentTime):
743         (else.currentTime):
744         * Skipped: Make sure that we don't run Basic yet.
745
746 2016-06-21  Jon Lee  <jonlee@apple.com>
747
748         Update canvas size when benchmark begins
749         https://bugs.webkit.org/show_bug.cgi?id=159010
750
751         Reviewed by Dean Jackson.
752
753         Previously we would evaluate the media queries and assign the canvas size to the test
754         on body.onload. Instead, do it once the user starts the benchmark.
755
756         * Animometer/resources/debug-runner/animometer.js: Move benchmarkController.determineCanvasSize()
757         to benchmarkController._startBenchmark.
758         * Animometer/resources/runner/animometer.js: Ditto.
759
760 2016-06-21  Jon Lee  <jonlee@apple.com>
761
762         Add new timestamp option
763         https://bugs.webkit.org/show_bug.cgi?id=159006
764
765         Reviewed by Dean Jackson.
766
767         Add a new option to take timestamps from the rAF callback.
768
769         * Animometer/developer.html: Added option.
770         * Animometer/tests/resources/main.js:
771         (Benchmark): Fall back to using the rAF timestamp if performance.now()
772         is not available.
773         (Benchmark._animateLoop): Update how timestamp is set. Prefer to use
774         local var instead of accessing private var.
775
776 2016-06-21  Jon Lee  <jonlee@apple.com>
777
778         Improvements to Animometer benchmark
779         https://bugs.webkit.org/show_bug.cgi?id=157738
780
781         Reviewed by Dean Jackson.
782         Provisionally reviewed by Said Abou-Hallawa.
783
784         Update tests.
785
786         * Animometer/tests/master/text.html: Ensure only three text sizes for
787         the three canvases.
788         * Animometer/tests/master/focus.html: Reduce the text size to fit with smaller
789         particle sizes.
790         * Animometer/tests/master/resources/focus.js: Remove the quadratic distribution for
791         particle sizes, and make it linear. Reduce the size variance. Shuffle the math to
792         reduce some calculations per frame. Fix the placement of the particles which might
793         otherwise be culled.
794         * Animometer/tests/master/resources/image-data.js: Reduce the particle size
795         to encourage larger scores.
796
797 2016-06-21  Jon Lee  <jonlee@apple.com>
798
799         Improvements to Animometer benchmark
800         https://bugs.webkit.org/show_bug.cgi?id=157738
801
802         Reviewed by Dean Jackson.
803         Provisionally reviewed by Said Abou-Hallawa.
804
805         Include confidence interval for the final score, and store the canvas
806         size in the serialization so that it is accurately shown in results.
807
808         * Animometer/developer.html: Add a "confidence" div.
809         * Animometer/index.html: Ditto. Convert "mean" to "confidence".
810         * Animometer/resources/debug-runner/animometer.js: Look at options, and
811         if the configuration is included, update the body class based on it
812         (similar to what we do when we first load the page). That way, if you
813         drag-and-drop previous results in, that configuration is reflected in
814         the dashboard. Show the full confidence interval.
815
816         * Animometer/resources/debug-runner/animometer.css:
817         * Animometer/resources/debug-runner/animometer.js: Style update.
818         * Animometer/resources/runner/animometer.css:
819
820         * Animometer/resources/runner/animometer.js:
821         (_processData): Propagate the confidence interval values out and calculate
822         the lower and upper bounds. For now, shortcut the aggregate calculation,
823         since we only go through one iteration.
824         (this._processData.calculateScore): Propagate the confidence interval out
825         to be next to the score. Depending on the controller these values are
826         calculated differently.
827         (this._processData._getResultsProperty): Convenience function.
828         (this._processData.get score): Refactor.
829         (this._processData.get scoreLowerBound): Get the aggregate lower bound.
830         (this._processData.get scoreUpperBound): Get the aggregate upper bound.
831         (window.sectionsManager.setSectionScore):
832         (window.benchmarkController._startBenchmark): When the benchmark starts, note
833         the canvas size and add it to options. That way it will be automatically be
834         serialized.
835         (window.benchmarkController.showResults): Include the maximum deviation
836         percentage.
837         * Animometer/resources/runner/lines.svg: Make the background line up with the
838         skew.
839         * Animometer/resources/runner/tests.js:
840         (Headers.details.text): Refactor.
841         * Animometer/resources/statistics.js:
842         (Statistics.largestDeviationPercentage): Convenience function to calculate
843         the deviation percentage on either end and return the largest deviation.
844         * Animometer/resources/strings.js:
845
846         Allow specifying a regression profile to use instead of taking the one
847         with the lowest error.
848
849         Address an issue in the focus test when the regression calculated ends
850         up overestimating the change point, causing a cascade of tougher
851         ramps. The reason behind this is that at max complexity of an initial
852         ramp, the frame length is very high, and it influences the second
853         segment of the piecewise regression strongly, causing it to be very
854         steep. As a result, the first segment, expected to be flat, ends up
855         covering a higher range of complexity. That makes the change point
856         much higher than it should be. To avoid this, we will add a sanity
857         check on the maximum value of the ramp. If the regression's projected
858         value at the maximum complexity of the current ramp is very slow (less
859         than 20 fps), 1) reduce the maximum complexity by 20%, and 2) do not
860         include the regression's change point in the change point estimator.
861         That estimator is used as the midpoint of the next ramp, and including
862         the change point from a poor regression can bake in the error. The
863         controller already knows how to adjust for ramps that are too easy for
864         the system.
865
866         * Animometer/resources/runner/animometer.js:
867         (this._processData.findRegression): Takes a preferred profile and gives that to
868         Regression.
869         (this._processData.calculateScore): With the ramp controller, take the profile
870         of the ramp that was used the most when calculating the ramp's regression. That
871         profile is what is used for the test's score.
872         * Animometer/resources/statistics.js:
873         (Regression.Utilities.createClass): Update to take an options object which can
874         specify a profile to calculate with. Otherwise it will continue to use both and
875         select the one with the lower error.
876         (_calculateRegression): Fix an issue where we claim 0 error if the regression
877         calculation fails due to divide-by-zero. Instead reject that regression calculation
878         by giving it Number.MAX_VALUE.
879         * Animometer/resources/strings.js: New strings for marking a regression as flat
880         or slope.
881         * Animometer/tests/resources/main.js:
882         (RampController): Rename the thresholds for clarity. Add a threshold that, if
883         exceeded, will lower the maximum complexity of the next ramp.
884         (tune): Relax the cdf check to consider whether the interval definitely falls in
885         the desired frame length threshold.
886         (processSamples): Include the profile in the ramp.
887
888         Update ramp controller test. Increase the length of the test to 30 seconds, and extend
889         the interval to 120 ms during sampling. Improve the estimation of the ramp parameters.
890
891         * Animometer/developer.html: Change default to 30 seconds, and don't show the progress bar
892         by default.
893         * Animometer/resources/runner/animometer.js: Change default to 30 seconds.
894         * Animometer/tests/resources/main.js: A number of improvements to the ramp controller, in
895         the order in which they appear in the patch:
896
897         - With a longer test length use longer ramps with longer intervals to get more data at each
898         complexity. Keep the 100 ms interval length during the ramp up phase since we don't need to
899         spend more time there to find the right order of magnitude, but increase it during the
900         ramps to 120 ms.
901         - The ramp linearly interpolates the complexity to render based on its timestamp, but it would
902         never sample the minimum complexity. Instead of lerping max to min complexity from time
903         0 to t where t is the ramp length, instead lerp from 0 to (t - intervalSampleLength) so that
904         we can have at least one interval sample at the min complexity for that ramp.
905         - Some regression calculations only come out with one line segment rather than the two
906         we expect. This could be due to a noisy ramp or the ramp's range is too narrow. If that's the
907         case, influence the minimum complexity of the next ramp towards the lowest bound of 1, so that
908         we ensure that at least part of the ramp is covering a complexity range that the system can
909         handle at full 60.
910         - Remove an assignment to interpolatedFrameLength since that is never subsequently used.
911
912         Update the format used to serialize the results for analysis.
913
914         Each data point used to have named properties for fields like complexity and frame rate.
915         In addition the serialized numbers had rounding errors that took up many characters.
916         Update the format by introducing a new data container called SampleData, which contains a
917         field map. The map maps a string to an array index. Each data point is an array, so, to
918         get a stat, use the field map to get the array index into the data point. This allows future
919         versions to track other data, and reduces the size of the output string by two-thirds.
920
921         * Animometer/resources/extensions.js:
922         (Utilities.toFixedNumber): Add convenience function that truncates the number to a fixed
923         precision, and converts it back to a number.
924         (SampleData): New class that contains sample data and a field map that maps properties to
925         an array index.
926         (get length): Number of data points.
927         (addField): Add a field to the field map.
928         (push): Add a data point.
929         (sort): Sort the data.
930         (slice): Return new SampleData object with sliced data.
931         (forEach): Iterate over the data with provided function.
932         (createDatum):
933         (getFieldInDatum): Returns the data point associated with the field name by looking it up
934         in the field map in the datum provided, which can be the datum object itself (an array) or
935         an index into the data member variable.
936         (setFieldInDatum): Sets the data point associated with the field name.
937         (at): Returns the data point at the provided index.
938         (toArray): Serializes the data where the field map serves as property names for each point.
939
940         * Animometer/resources/debug-runner/graph.js:
941         (updateGraphData): Remove unused _testData. Convert the data to the old array format for the
942         graph to use, since the old format was much easier to work with when displaying the graphs.
943         (onGraphTypeChanged): For some controllers, no alternative score or mean is provided.
944         * Animometer/resources/runner/animometer.css:
945         * Animometer/resources/runner/animometer.js: Refactor to use SampleData. Update JSON output
946         to only go to 3 digits of precision for purposes of reducing the data size.
947         * Animometer/resources/strings.js: Add new strings to put into the field maps.
948         * Animometer/tests/resources/main.js: Refactor to use SampleData.
949
950         * Animometer/developer.html:
951         * Animometer/index.html: Restructure results table for both pages. Add charset attribute to
952         tests.js include.
953         * Animometer/resources/debug-runner/animometer.css: Clear out styles from release runner.
954         * Animometer/resources/debug-runner/graph.js:
955         (onGraphTypeChanged): Update score and mean if bootstrap results are available from the
956         controller, since not all controllers do bootstrapping.
957         * Animometer/resources/debug-runner/tests.js: Update header text.
958         * Animometer/resources/runner/animometer.css: Include confidence interval in results.
959         * Animometer/resources/runner/animometer.js:
960         (ResultsTable._addHeader): Header contents can be HTML, so use innerHTML instead.
961         (ResultsTable._addBody): Add tbody element.
962         (ResultsTable._addTest): Allow a data cell to invoke a JS function to get its contents.
963         (window.benchmarkController.showResults): Add table that includes tests' confidence intervals.
964         * Animometer/resources/runner/tests.js:
965         (Headers.details.text): Add new details table that includes bootstrap confidence interval.
966         The interval can be asymmetric, but for simplicity, report the maximum deviation percentage
967         on either side of the bootstrap median.
968         * Animometer/resources/statistics.js:
969         (bootstrap): Include the confidence percentage in the return object.
970
971         Report canvas size in results.
972
973         * Animometer/developer.html: Add markup to indicate whether a small, medium, or large
974         canvas was used.
975         * Animometer/index.html: Ditto.
976         * Animometer/resources/debug-runner/animometer.js: Call determineCanvasSize().
977         * Animometer/resources/runner/animometer.css: Update styles to set the canvas based on the
978         body class size.
979         * Animometer/resources/runner/animometer.js:
980         (window.benchmarkController.initialize): Update styles to set the canvas based on the
981         body class size.
982         (window.benchmarkController.determineCanvasSize): Run various media queries and set the body
983         class based on the size of the device.
984
985         * Animometer/developer.html: Refactor to include the main CSS file, and redo
986         the layout so that it doesn't rely on flexbox.
987         * Animometer/resources/debug-runner/animometer.css:
988         * Animometer/resources/debug-runner/animometer.js:
989         (updateDisplay): Since various parts of the script alter the body class, we can't
990         replace the className directly. Instead, remove all display-based values and then add
991         the one that was selected.
992         * Animometer/resources/debug-runner/graph.js:
993         (updateGraphData): To set the size of the graph, use window.innerHeight.
994         * Animometer/resources/runner/animometer.js:
995         (window.sectionsManager.showSection): Since various parts of the script alter the body
996         class, we can't replace the className directly. Remove all of the section classes
997         individually and then add the one desired.
998         * Animometer/tests/resources/stage.css: Remove -apple-system as a font to use in the
999         stage.
1000
1001 2016-06-12  Filip Pizlo  <fpizlo@apple.com>
1002
1003         Fix round-down goof in Air.js's ShuffleCustom.forEachArg
1004         https://bugs.webkit.org/show_bug.cgi?id=158674
1005
1006         Reviewed by Michael Saboff.
1007         
1008         x / 3 * 3 is not how you round down to multiples of 3 in JavaScript. You need to do
1009         Math.floor(x / 3) * 3 instead.
1010         
1011         This is a benign change, because having extra arguments to a Shuffle is not something we
1012         actually take advantage of yet. But I think it's best to match the original C++ code's
1013         intent.
1014
1015         * Air.js/custom.js:
1016         (const.ShuffleCustom.forEachArg):
1017
1018 2016-06-11  Filip Pizlo  <fpizlo@apple.com>
1019
1020         Unreviewed, skip Air.js. It got unskipped when I renamed it from JSAir.
1021
1022         * Skipped:
1023
1024 2016-06-10  Filip Pizlo  <fpizlo@apple.com>
1025
1026         Make it easy to package Air.js
1027         https://bugs.webkit.org/show_bug.cgi?id=158652
1028
1029         Reviewed by Benjamin Poulain.
1030         
1031         If you want to give Air.js to someone, you can now do:
1032         
1033         cd PerformanceTests/Air.js
1034         ./make_dist.sh
1035         
1036         This will create Air.js.tar.gz and a directory called Air.js. If you want to send someone
1037         the benchmark, just send either of those things and tell them to look at index.html.
1038         
1039         You may have to edit make_dist.sh for wherever you have Markdown.pl. You can get that from
1040         https://daringfireball.net/projects/markdown/
1041
1042         * Air.js/make_dist.sh: Added.
1043         * Air.js/test.html:
1044
1045 2016-06-10  Filip Pizlo  <fpizlo@apple.com>
1046
1047         Unreviewed, make the link to test.html into an actual link.
1048
1049         * Air.js/README.md:
1050
1051 2016-06-10  Filip Pizlo  <fpizlo@apple.com>
1052
1053         Fix some text in Air.js/README.md
1054         https://bugs.webkit.org/show_bug.cgi?id=158650
1055
1056         Reviewed by Benjamin Poulain.
1057         
1058         I read the text again and found bugs:
1059         
1060         - We never actually say how to run the benchmark. This change adds a blurb about how to run
1061           it.
1062
1063         - We both say that allocateStack is responsible for the bulk of the running time and that
1064           we haven't measured where the bulk of the time is spent. This changes the text to say that
1065           it was a goal to make allocateStack be the hottest part of the benchmark, but that we did
1066           not measure this.
1067
1068         * Air.js/README.md:
1069
1070 2016-06-10  Filip Pizlo  <fpizlo@apple.com>
1071
1072         Air.js should have some documentation
1073         https://bugs.webkit.org/show_bug.cgi?id=158648
1074
1075         Reviewed by Keith Miller.
1076         
1077         I want to be able to point people at a document if they want to know more about this
1078         benchmark.
1079
1080         * Air.js/README.md: Added.
1081
1082 2016-06-10  Filip Pizlo  <fpizlo@apple.com>
1083
1084         Rename JSAir to Air.js.
1085
1086         Rubber stamped by Mark Lam.
1087
1088         * Air.js: Copied from PerformanceTests/JSAir.
1089         * Air.js/airjs-tests.yaml: Copied from PerformanceTests/JSAir/jsair-tests.yaml.
1090         * Air.js/benchmark.js: Copied from PerformanceTests/JSAir/benchmark.js.
1091         (Benchmark):
1092         * Air.js/jsair-tests.yaml: Removed.
1093         * Air.js/payload-airjs-ACLj8C.js: Copied from PerformanceTests/JSAir/payload-jsair-ACLj8C.js.
1094         * Air.js/payload-jsair-ACLj8C.js: Removed.
1095         * Air.js/stress-test.js: Copied from PerformanceTests/JSAir/stress-test.js.
1096         * Air.js/test.html: Copied from PerformanceTests/JSAir/test.html.
1097         * Air.js/test.js: Copied from PerformanceTests/JSAir/test.js.
1098         * JSAir: Removed.
1099         * JSAir/all.js: Removed.
1100         * JSAir/allocate_stack.js: Removed.
1101         * JSAir/arg.js: Removed.
1102         * JSAir/basic_block.js: Removed.
1103         * JSAir/benchmark.js: Removed.
1104         * JSAir/code.js: Removed.
1105         * JSAir/custom.js: Removed.
1106         * JSAir/frequented_block.js: Removed.
1107         * JSAir/insertion_set.js: Removed.
1108         * JSAir/inst.js: Removed.
1109         * JSAir/jsair-tests.yaml: Removed.
1110         * JSAir/liveness.js: Removed.
1111         * JSAir/opcode.js: Removed.
1112         * JSAir/payload-gbemu-executeIteration.js: Removed.
1113         * JSAir/payload-imaging-gaussian-blur-gaussianBlur.js: Removed.
1114         * JSAir/payload-jsair-ACLj8C.js: Removed.
1115         * JSAir/payload-typescript-scanIdentifier.js: Removed.
1116         * JSAir/reg.js: Removed.
1117         * JSAir/stack_slot.js: Removed.
1118         * JSAir/stress-test.js: Removed.
1119         * JSAir/strip-hash.rb: Removed.
1120         * JSAir/symbols.js: Removed.
1121         * JSAir/test.html: Removed.
1122         * JSAir/test.js: Removed.
1123         * JSAir/tmp.js: Removed.
1124         * JSAir/tmp_base.js: Removed.
1125         * JSAir/util.js: Removed.
1126
1127 2016-06-10  Filip Pizlo  <fpizlo@apple.com>
1128
1129         JSC Stress Test failing: jsair-tests.yaml/test.js.ftl-eager-no-cjit
1130         https://bugs.webkit.org/show_bug.cgi?id=158571
1131
1132         Reviewed by Keith Miller.
1133         
1134         Introduce a different harness for run-jsc-stress-tests, which runs for a minimum of 10
1135         iterations and then tries to do any number of "bonus" iterations until it's been running for
1136         two seconds. Since this is the sort of test that isn't really meant to stress anything in
1137         particular, I think it's OK if it is time-limited in this way. The worst case is that some
1138         of its failures will be flaky, but I think that they would have been flaky anyway given the
1139         complexity of the test.
1140
1141         * JSAir/benchmark.js:
1142         (Benchmark):
1143         (Benchmark.prototype.runIteration):
1144         (benchmark): Deleted.
1145         * JSAir/jsair-tests.yaml:
1146         * JSAir/stress-test.js: Added.
1147         (preciseTime):
1148         * JSAir/test.html:
1149         * JSAir/test.js:
1150
1151 2016-06-09  Filip Pizlo  <fpizlo@apple.com>
1152
1153         Unreviewed, teach the perf bots not to run JSAir.
1154
1155         * Skipped:
1156
1157 2016-06-08  Filip Pizlo  <fpizlo@apple.com>
1158
1159         Use more ES6 features in JSAir
1160         https://bugs.webkit.org/show_bug.cgi?id=158497
1161
1162         Reviewed by Keith Miller.
1163         
1164         This improves JSAir with the following ES6 features suggested by JoePeck:
1165         
1166         - String interpolation.
1167         - Destructuring inside PatchCustom.
1168         - Default arguments.
1169         
1170         All of these things are on hot paths.
1171         
1172         Note that I didn't use string interpolation everywhere that I could, only in those places
1173         where it made the code more readable. In Ruby, I used the style that if the interpolation
1174         expression has any non-trivial stuff (like a ternary operator, a chain of calls, or embedded
1175         strings) then it's better to use regular strcat. I think that's what I carried over to here.
1176         
1177         Note that the previous change (Add result validation to JSAir) also made the Proxy code not
1178         dead, though it's not necessarily on the hot path. The Proxy isn't called into frequently
1179         but it's used from a function that is otherwise hot, so if calling into the Proxy prevents
1180         that function from being optimized then it will hurt so good.
1181         
1182         I also reenabled tail calls in a few places.
1183         
1184         This change doesn't seem to change the performance of the benchmark for us. That's expected
1185         since these ES6 features are cheap. Note that this claim doesn't include Proxy, which was
1186         added in a separate change and that change did make the benchmark overall more expensive.
1187
1188         * JSAir/allocate_stack.js:
1189         (allocateStack):
1190         * JSAir/arg.js:
1191         (Arg.createBitImm64):
1192         (Arg.createAddr):
1193         (Arg.createStack):
1194         (Arg.logScale):
1195         (Arg.createIndex):
1196         * JSAir/basic_block.js:
1197         (BasicBlock.get headerString):
1198         (BasicBlock.prototype.get if):
1199         (BasicBlock):
1200         * JSAir/benchmark.js:
1201         (benchmark):
1202         * JSAir/code.js:
1203         (Code):
1204         (Code.prototype.addBlock):
1205         (Code.prototype.addStackSlot):
1206         (Code.prototype.newTmp):
1207         (Code.prototype.get size):
1208         (Code.prototype.get blocks):
1209         (Code.prototype.get stackSlots):
1210         (Code.prototype.tmps):
1211         (Code.prototype.get callArgAreaSize):
1212         (Code.prototype.toString):
1213         * JSAir/custom.js:
1214         (const.PatchCustom.forEachArg):
1215         * JSAir/inst.js:
1216         (Inst):
1217         * JSAir/reg.js:
1218         (Reg.prototype.toString):
1219         * JSAir/util.js:
1220         (symbolName):
1221         (lowerSymbolName):
1222
1223 2016-06-07  Filip Pizlo  <fpizlo@apple.com>
1224
1225         Add result validation to JSAir
1226         https://bugs.webkit.org/show_bug.cgi?id=158493
1227
1228         Reviewed by Saam Barati.
1229         
1230         This adds the ability to hash a Code in a way that matches the C++ code's hashing of Code.
1231         This allows us to check if the Code that JSAir sees is the code that C++ saw. We use this to
1232         check the Code before and after allocateStack, and compare against hashes we got from C++.
1233         
1234         Doing this uncovered bugs. roundUpToMultipleOf wasn't doing anything. allocateStack was not
1235         allocating things correctly because I was concatting a Set to an Array, which doesn't really
1236         work. Now these bugs are fixed.
1237         
1238         The checking step adds to the running time so I reduced the number of iterations. The
1239         benchmark spends a decent amount of its time computing Code hashes; I think it's around 1/3
1240         total. This is probably OK. It's better to verify the results even if the running time is
1241         not all in the "core" of the algorithm.
1242         
1243         Also add a run-jsc-stress-tests yaml file to allow this to run as a test.
1244
1245         * JSAir/allocate_stack.js:
1246         * JSAir/arg.js:
1247         (Arg.createImm):
1248         (Arg.createBigImm):
1249         (Arg.createBitImm):
1250         (Arg.createBitImm64):
1251         (Arg.createWidth):
1252         (Arg.createSpecial):
1253         (Arg.prototype.get kind):
1254         (Arg.prototype.get isTmp):
1255         (Arg.prototype.get isImm):
1256         (Arg.prototype.get isSomeImm):
1257         (Arg.prototype.get isSomeBigImm):
1258         (Arg.prototype.get isCondition):
1259         (Arg.prototype.get isWidth):
1260         (Arg.prototype.get isSpecial):
1261         (Arg.prototype.get isAlive):
1262         (Arg.prototype.get tmp):
1263         (Arg.prototype.get value):
1264         (Arg.prototype.get lowValue):
1265         (Arg.prototype.get highValue):
1266         (Arg.prototype.get base):
1267         (Arg.prototype.get isGP):
1268         (Arg.prototype.get isFP):
1269         (Arg.prototype.isValidForm):
1270         (Arg.prototype.get isInvertible):
1271         (Arg.kindCode):
1272         (Arg.prototype.hash):
1273         (Arg.prototype.toString):
1274         (Arg):
1275         * JSAir/basic_block.js:
1276         (BasicBlock.get successorBlocks):
1277         * JSAir/benchmark.js:
1278         (benchmark):
1279         * JSAir/code.js:
1280         (Code.prototype.setFrameSize):
1281         (Code.prototype.hash):
1282         (Code.prototype.toString):
1283         (Code):
1284         * JSAir/inst.js:
1285         (Inst.prototype.get hasNonArgEffects):
1286         (Inst.prototype.hash):
1287         (Inst.prototype.toString):
1288         (Inst):
1289         * JSAir/jsair-tests.yaml: Added.
1290         * JSAir/opcode.js:
1291         (Inst_forEachArg):
1292         (Inst_hasNonArgEffects):
1293         (opcodeCode):
1294         * JSAir/payload-gbemu-executeIteration.js:
1295         (createPayloadGbemuExecuteIteration):
1296         * JSAir/payload-imaging-gaussian-blur-gaussianBlur.js:
1297         (createPayloadImagingGaussianBlurGaussianBlur):
1298         * JSAir/payload-jsair-ACLj8C.js:
1299         (createPayloadJSAirACLj8C):
1300         * JSAir/payload-typescript-scanIdentifier.js:
1301         (createPayloadTypescriptScanIdentifier):
1302         * JSAir/reg.js:
1303         (Reg.prototype.get isReg):
1304         (Reg.prototype.hash):
1305         (Reg.prototype.toString):
1306         * JSAir/stack_slot.js:
1307         (StackSlot.prototype.setOffsetFromFP):
1308         (StackSlot.prototype.hash):
1309         (StackSlot.prototype.toString):
1310         * JSAir/symbols.js:
1311         (relCondCode):
1312         (resCondCode):
1313         (doubleCondCode):
1314         * JSAir/test.html:
1315         * JSAir/tmp.js:
1316         (Tmp.prototype.get isReg):
1317         (Tmp.prototype.hash):
1318         (Tmp.prototype.toString):
1319         * JSAir/util.js:
1320         (roundUpToMultipleOf):
1321         (symbolName):
1322
1323 2016-06-07  Filip Pizlo  <fpizlo@apple.com>
1324
1325         Implement Air::allocateStack() in ES6 to see how much of a bad idea that is
1326         https://bugs.webkit.org/show_bug.cgi?id=158318
1327
1328         Reviewed by Saam Barati.
1329         
1330         This adds a new benchmark for us to play with called JSAir. It's a complete ES6
1331         implementation of Air's allocateStack() phase along with all of Air needed to run that
1332         phase. This includes things like stack slots, registers, temporaries, basic blocks,
1333         instructions, and all of the code for iterating over, inspecting, and modifying those
1334         things.
1335         
1336         To make this work, JSC can now dump Air just before allocateStack() in the form of JS code
1337         that creates a Code object that matches exactly what C++ Air saw. This benchmark comprises
1338         four Air IRs:
1339         
1340         - Octane/gbemu's largest function, executeIteration.
1341         - Kraken/imaging-gaussian-blur's largest function in OSR entry mode, gaussuanBlur.
1342         - Octane/typescript's largest function that is actually hot, scanIdentifier.
1343         - JSAir's largest hot function, which is anonymous, so we call it by its hash (ACLj8C).
1344         
1345         This runs in about 2 seconds on my machine in JSC trunk. It includes both a commandline
1346         harness and a web harness. JSAir is almost exactly 100x slower in ES6 in WebKit than the
1347         C++ Air::allocateStack() phase on which it is based.
1348         
1349         JSAir uses the following ES6 features:
1350         
1351         - Symbol.
1352         - for-of.
1353         - arrow functions.
1354         - Map/Set.
1355         - let/const.
1356         - classes.
1357         
1358         All of these things are used in anger and should end up on the hot path. There is also code
1359         that uses Proxies, but it ends up being dead.
1360         
1361         We can improve this even more:
1362         I still need to add result validation: https://bugs.webkit.org/show_bug.cgi?id=158493
1363         I want to make it use more ES6 features: https://bugs.webkit.org/show_bug.cgi?id=158497
1364         
1365         * JSAir: Added.
1366         * JSAir/all.js: Added.
1367         * JSAir/allocate_stack.js: Added.
1368         (allocateStack.attemptAssignment):
1369         (allocateStack.assign):
1370         (allocateStack.interfere):
1371         (allocateStack.):
1372         (allocateStack):
1373         * JSAir/arg.js: Added.
1374         (Arg):
1375         (Arg.isAnyUse):
1376         (Arg.isColdUse):
1377         (Arg.isWarmUse):
1378         (Arg.cooled):
1379         (Arg.isEarlyUse):
1380         (Arg.isLateUse):
1381         (Arg.isAnyDef):
1382         (Arg.isEarlyDef):
1383         (Arg.isLateDef):
1384         (Arg.isZDef):
1385         (Arg.typeForB3Type):
1386         (Arg.widthForB3Type):
1387         (Arg.conservativeWidth):
1388         (Arg.minimumWidth):
1389         (Arg.bytes):
1390         (Arg.widthForBytes):
1391         (Arg.createTmp):
1392         (Arg.fromReg):
1393         (Arg.createImm):
1394         (Arg.createBigImm):
1395         (Arg.createBitImm):
1396         (Arg.createBitImm64):
1397         (Arg.createAddr):
1398         (Arg.createStack):
1399         (Arg.createCallArg):
1400         (Arg.createStackAddr):
1401         (Arg.isValidScale):
1402         (Arg.logScale):
1403         (Arg.createIndex):
1404         (Arg.createRelCond):
1405         (Arg.createResCond):
1406         (Arg.createDoubleCond):
1407         (Arg.createWidth):
1408         (Arg.prototype.get kind):
1409         (Arg.prototype.get isTmp):
1410         (Arg.prototype.get isImm):
1411         (Arg.prototype.get isBigImm):
1412         (Arg.prototype.get isBitImm):
1413         (Arg.prototype.get isBitImm64):
1414         (Arg.prototype.get isSomeImm):
1415         (Arg.prototype.get isAddr):
1416         (Arg.prototype.get isStack):
1417         (Arg.prototype.get isCallArg):
1418         (Arg.prototype.get isIndex):
1419         (Arg.prototype.get isMemory):
1420         (Arg.prototype.get isStackMemory):
1421         (Arg.prototype.get isRelCond):
1422         (Arg.prototype.get isResCond):
1423         (Arg.prototype.get isDoubleCond):
1424         (Arg.prototype.get isCondition):
1425         (Arg.prototype.get isWidth):
1426         (Arg.prototype.get isAlive):
1427         (Arg.prototype.get tmp):
1428         (Arg.prototype.get value):
1429         (Arg.prototype.get base):
1430         (Arg.prototype.get hasOffset):
1431         (Arg.prototype.get offset):
1432         (Arg.prototype.get stackSlot):
1433         (Arg.prototype.get index):
1434         (Arg.prototype.get scale):
1435         (Arg.prototype.get logScale):
1436         (Arg.prototype.get width):
1437         (Arg.prototype.get isGPTmp):
1438         (Arg.prototype.get isFPTmp):
1439         (Arg.prototype.get isGP):
1440         (Arg.prototype.get isFP):
1441         (Arg.prototype.get hasType):
1442         (Arg.prototype.get type):
1443         (Arg.prototype.isType):
1444         (Arg.prototype.isCompatibleType):
1445         (Arg.prototype.get isGPR):
1446         (Arg.prototype.get gpr):
1447         (Arg.prototype.get isFPR):
1448         (Arg.prototype.get fpr):
1449         (Arg.prototype.get isReg):
1450         (Arg.prototype.get reg):
1451         (Arg.isValidImmForm):
1452         (Arg.isValidBitImmForm):
1453         (Arg.isValidBitImm64Form):
1454         (Arg.isValidAddrForm):
1455         (Arg.isValidIndexForm):
1456         (Arg.prototype.isValidForm):
1457         (Arg.prototype.forEachTmpFast):
1458         (Arg.prototype.usesTmp):
1459         (Arg.prototype.forEachTmp):
1460         (Arg.prototype.is):
1461         (Arg.prototype.as):
1462         (Arg.prototype.forEachFast):
1463         (Arg.prototype.forEach):
1464         (Arg.extract):
1465         (Arg.forEachFast):
1466         (Arg.forEach):
1467         (Arg.prototype.get condition):
1468         (Arg.prototype.get isInvertible):
1469         (Arg.prototype.toString):
1470         * JSAir/basic_block.js: Added.
1471         (BasicBlock):
1472         (BasicBlock.prototype.get index):
1473         (BasicBlock.prototype.get size):
1474         (BasicBlock.prototype.Symbol.iterator):
1475         (BasicBlock.prototype.at):
1476         (BasicBlock.get last):
1477         (BasicBlock.get insts):
1478         (BasicBlock.get numSuccessors):
1479         (BasicBlock.get successors):
1480         (BasicBlock.get successorBlocks.):
1481         (BasicBlock.get successorBlocks):
1482         (BasicBlock.set get numPredecessors):
1483         (BasicBlock.get predecessors):
1484         (BasicBlock.get frequency):
1485         (BasicBlock.get headerString):
1486         (BasicBlock.get footerString):
1487         * JSAir/benchmark.js: Added.
1488         (benchmark):
1489         * JSAir/code.js: Added.
1490         (Code):
1491         (Code.prototype.addBlock):
1492         (Code.prototype.addStackSlot):
1493         (Code.prototype.newTmp):
1494         (Code.prototype.get size):
1495         (Code.prototype.at):
1496         (Code.prototype.Symbol.iterator):
1497         (Code.prototype.get blocks):
1498         (Code.prototype.get stackSlots):
1499         (Code.prototype.tmps):
1500         (Code.prototype.get callArgAreaSize):
1501         (Code.prototype.requestCallArgAreaSize):
1502         (Code.prototype.get frameSize):
1503         (Code.prototype.setFrameSize):
1504         (Code.prototype.toString):
1505         * JSAir/custom.js: Added.
1506         (const.ShuffleCustom.forEachArg):
1507         (const.ShuffleCustom.hasNonArgNonControlEffects):
1508         (const.PatchCustom.forEachArg):
1509         (const.PatchCustom.hasNonArgNonControlEffects):
1510         (const.CCallCustom.forEachArg):
1511         (const.CCallCustom.hasNonArgNonControlEffects):
1512         (const.ColdCCallCustom.forEachArg):
1513         (const.ColdCCallCustom.hasNonArgNonControlEffects):
1514         * JSAir/frequented_block.js: Added.
1515         (FrequentedBlock):
1516         (FrequentedBlock.prototype.toString):
1517         * JSAir/insertion_set.js: Added.
1518         (Insertion):
1519         (Insertion.prototype.get index):
1520         (Insertion.prototype.get element):
1521         (Insertion.prototype.lessThan):
1522         (InsertionSet):
1523         (InsertionSet.prototype.appendInsertion):
1524         (InsertionSet.prototype.append):
1525         (InsertionSet.prototype.execute):
1526         * JSAir/inst.js: Added.
1527         (Inst):
1528         (Inst.prototype.append):
1529         (Inst.prototype.clear):
1530         (Inst.prototype.get opcode):
1531         (Inst.prototype.get args):
1532         (Inst.prototype.visitArg):
1533         (Inst.prototype.forEachTmpFast):
1534         (Inst.prototype.forEachArg):
1535         (Inst.prototype.forEachTmp):
1536         (Inst.prototype.forEach):
1537         (Inst.forEachDef):
1538         (Inst.forEachDefWithExtraClobberedRegs):
1539         (Inst.prototype.get hasNonArgEffects):
1540         (Inst.prototype.toString):
1541         * JSAir/liveness.js: Added.
1542         (Liveness):
1543         (Liveness.prototype.get thing):
1544         (Liveness.prototype.get code):
1545         (Liveness.prototype.get liveAtHead):
1546         (Liveness.prototype.get liveAtTail):
1547         (Liveness.prototype.localCalc.LocalCalc):
1548         (Liveness.prototype.localCalc.LocalCalc.prototype.get liveSet):
1549         (Liveness.prototype.localCalc.LocalCalc.prototype.execute):
1550         (Liveness.prototype.localCalc):
1551         * JSAir/opcode.js: Added.
1552         (Inst_forEachArg):
1553         (Inst_hasNonArgEffects):
1554         * JSAir/payload-gbemu-executeIteration.js: Added.
1555         (createPayloadGbemuExecuteIteration):
1556         * JSAir/payload-imaging-gaussian-blur-gaussianBlur.js: Added.
1557         (createPayloadImagingGaussianBlurGaussianBlur):
1558         * JSAir/payload-jsair-ACLj8C.js: Added.
1559         (createPayloadJSAirACLj8C):
1560         * JSAir/payload-typescript-scanIdentifier.js: Added.
1561         (createPayloadTypescriptScanIdentifier):
1562         * JSAir/reg.js: Added.
1563         (Reg):
1564         (Reg.fromReg):
1565         (Reg.prototype.get index):
1566         (Reg.prototype.get type):
1567         (Reg.prototype.get name):
1568         (Reg.prototype.get isCalleeSave):
1569         (Reg.prototype.get isReg):
1570         (Reg.prototype.toString):
1571         (Reg.extract):
1572         (Reg.forEachFast):
1573         (Reg.forEach):
1574         (newGPR):
1575         (Reg.gprs.Reg.fprs.Reg.calleeSaveGPRs.Reg.calleeSaveFPRs.Reg.calleeSaves):
1576         * JSAir/stack_slot.js: Added.
1577         (StackSlot):
1578         (StackSlot.prototype.get byteSize):
1579         (StackSlot.prototype.get kind):
1580         (StackSlot.prototype.get isLocked):
1581         (StackSlot.prototype.get isSpill):
1582         (StackSlot.prototype.get index):
1583         (StackSlot.prototype.ensureSize):
1584         (StackSlot.prototype.get alignment):
1585         (StackSlot.prototype.get offsetFromFP):
1586         (StackSlot.prototype.setOffsetFromFP):
1587         (StackSlot.prototype.toString):
1588         (StackSlot.extract):
1589         (StackSlot.forEachFast):
1590         (StackSlot.forEach):
1591         * JSAir/symbols.js: Added.
1592         * JSAir/test.html: Added.
1593         * JSAir/test.js: Added.
1594         * JSAir/tmp.js: Added.
1595         (Tmp):
1596         (Tmp.fromReg):
1597         (Tmp.prototype.get index):
1598         (Tmp.prototype.get type):
1599         (Tmp.prototype.get isReg):
1600         (Tmp.prototype.toString):
1601         (Tmp.extract):
1602         (Tmp.forEachFast):
1603         (Tmp.forEach):
1604         * JSAir/tmp_base.js: Added.
1605         (TmpBase.prototype.get isGP):
1606         (TmpBase.prototype.get isFP):
1607         (TmpBase.prototype.get isGPR):
1608         (TmpBase.prototype.get isFPR):
1609         (TmpBase.prototype.get reg):
1610         (TmpBase.prototype.get gpr):
1611         (TmpBase.prototype.get fpr):
1612         (TmpBase):
1613         * JSAir/util.js: Added.
1614         (isRepresentableAsInt32):
1615         (addIndexed):
1616         (roundUpToMultipleOf):
1617         (symbolName):
1618         (mergeIntoSet):
1619         (nonEmptyRangesOverlap):
1620         (rangesOverlap):
1621         (removeAllMatching):
1622         (swap):
1623         (bubble):
1624         (bubbleSort):
1625         (currentTime):
1626         (else.currentTime):
1627
1628 2016-05-26  Yusuke Suzuki  <utatane.tea@gmail.com>
1629
1630         [JSC] Allow JSBench to use precise time
1631         https://bugs.webkit.org/show_bug.cgi?id=158050
1632
1633         Reviewed by Geoffrey Garen.
1634
1635         * JSBench/amazon-chrome-win/urem.js:
1636         (else.window.performance.window.performance.now.currentTimeInMS):
1637         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1638         (else.else.currentTimeInMS):
1639         (onload.cb):
1640         (onload):
1641         * JSBench/amazon-chrome/urem.js:
1642         (else.window.performance.window.performance.now.currentTimeInMS):
1643         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1644         (else.else.currentTimeInMS):
1645         (onload.cb):
1646         (onload):
1647         * JSBench/amazon-firefox-win/urm.js:
1648         (else.window.performance.window.performance.now.currentTimeInMS):
1649         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1650         (else.else.currentTimeInMS):
1651         (onload.cb):
1652         (onload):
1653         * JSBench/amazon-firefox/urm.js:
1654         (else.window.performance.window.performance.now.currentTimeInMS):
1655         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1656         (else.else.currentTimeInMS):
1657         (onload.cb):
1658         (onload):
1659         * JSBench/amazon-safari/urem.js:
1660         (else.window.performance.window.performance.now.currentTimeInMS):
1661         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1662         (else.else.currentTimeInMS):
1663         (onload.cb):
1664         (onload):
1665         * JSBench/facebook-chrome-win/urem.js:
1666         (else.window.performance.window.performance.now.currentTimeInMS):
1667         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1668         (else.else.currentTimeInMS):
1669         (onload.cb):
1670         (onload):
1671         * JSBench/facebook-chrome/urem.js:
1672         (else.window.performance.window.performance.now.currentTimeInMS):
1673         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1674         (else.else.currentTimeInMS):
1675         (onload.cb):
1676         (onload):
1677         * JSBench/facebook-firefox-win/urem.js:
1678         (else.window.performance.window.performance.now.currentTimeInMS):
1679         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1680         (else.else.currentTimeInMS):
1681         (onload.cb):
1682         (onload):
1683         * JSBench/facebook-firefox/urem.js:
1684         (else.window.performance.window.performance.now.currentTimeInMS):
1685         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1686         (else.else.currentTimeInMS):
1687         (onload.cb):
1688         (onload):
1689         * JSBench/facebook-safari/urem.js:
1690         (else.window.performance.window.performance.now.currentTimeInMS):
1691         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1692         (else.else.currentTimeInMS):
1693         (onload.cb):
1694         (onload):
1695         * JSBench/google-chrome-win/urem.js:
1696         (else.window.performance.window.performance.now.currentTimeInMS):
1697         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1698         (else.else.currentTimeInMS):
1699         (onload.cb):
1700         (onload):
1701         * JSBench/google-chrome/urem.js:
1702         (else.window.performance.window.performance.now.currentTimeInMS):
1703         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1704         (else.else.currentTimeInMS):
1705         (onload.cb):
1706         (onload):
1707         * JSBench/google-firefox-win/urem.js:
1708         (else.window.performance.window.performance.now.currentTimeInMS):
1709         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1710         (else.else.currentTimeInMS):
1711         (onload.cb):
1712         (onload):
1713         * JSBench/google-firefox/uem.js:
1714         (else.window.performance.window.performance.now.currentTimeInMS):
1715         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1716         (else.else.currentTimeInMS):
1717         (onload.cb):
1718         (onload):
1719         * JSBench/google-safari/urem.js:
1720         (else.window.performance.window.performance.now.currentTimeInMS):
1721         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1722         (else.else.currentTimeInMS):
1723         (onload.cb):
1724         (onload):
1725         * JSBench/harness.js:
1726         (runBenchmark.window.currentTimeInMS):
1727         (runBenchmark.else.window.currentTimeInMS):
1728         * JSBench/twitter-chrome-win/rem.js:
1729         (else.window.performance.window.performance.now.currentTimeInMS):
1730         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1731         (else.else.currentTimeInMS):
1732         (onload.cb):
1733         (onload):
1734         * JSBench/twitter-chrome/urem.js:
1735         (else.window.performance.window.performance.now.currentTimeInMS):
1736         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1737         (else.else.currentTimeInMS):
1738         (onload.cb):
1739         (onload):
1740         * JSBench/twitter-firefox-win/urem.js:
1741         (else.window.performance.window.performance.now.currentTimeInMS):
1742         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1743         (else.else.currentTimeInMS):
1744         (onload.cb):
1745         (onload):
1746         * JSBench/twitter-firefox/urem.js:
1747         (else.window.performance.window.performance.now.currentTimeInMS):
1748         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1749         (else.else.currentTimeInMS):
1750         (onload.cb):
1751         (onload):
1752         * JSBench/twitter-safari/urem.js:
1753         (else.window.performance.window.performance.now.currentTimeInMS):
1754         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1755         (else.else.currentTimeInMS):
1756         (onload.cb):
1757         (onload):
1758         * JSBench/yahoo-chrome-win/urem.js:
1759         (else.window.performance.window.performance.now.currentTimeInMS):
1760         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1761         (else.else.currentTimeInMS):
1762         (onload.cb):
1763         (onload):
1764         * JSBench/yahoo-chrome/urem.js:
1765         (else.window.performance.window.performance.now.currentTimeInMS):
1766         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1767         (else.else.currentTimeInMS):
1768         (onload.cb):
1769         (onload):
1770         * JSBench/yahoo-firefox-win/urem.js:
1771         (else.window.performance.window.performance.now.currentTimeInMS):
1772         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1773         (else.else.currentTimeInMS):
1774         (onload.cb):
1775         (onload):
1776         * JSBench/yahoo-firefox/urem.js:
1777         (else.window.performance.window.performance.now.currentTimeInMS):
1778         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1779         (else.else.currentTimeInMS):
1780         (onload.cb):
1781         (onload):
1782         * JSBench/yahoo-safari/urem.js:
1783         (else.window.performance.window.performance.now.currentTimeInMS):
1784         (else.else.typeof.preciseTime.string_appeared_here.currentTimeInMS):
1785         (else.else.currentTimeInMS):
1786         (onload.cb):
1787         (onload):
1788
1789 2016-05-25  Geoffrey Garen  <ggaren@apple.com>
1790
1791         replaceable own properties seem to ignore replacement after property caching
1792         https://bugs.webkit.org/show_bug.cgi?id=158091
1793
1794         Reviewed by Darin Adler.
1795
1796         * MallocBench/MallocBench.xcodeproj/project.pbxproj:
1797         * MallocBench/MallocBench/Benchmark.cpp:
1798         * MallocBench/MallocBench/Interpreter.cpp:
1799         (Interpreter::doMallocOp):
1800         * MallocBench/MallocBench/Interpreter.h:
1801         * MallocBench/MallocBench/fastMallocLog.63316.ops: Added.
1802         * MallocBench/MallocBench/jetstream.cpp: Added.
1803         (benchmark_jetstream):
1804         * MallocBench/MallocBench/jetstream.h: Added.
1805
1806 2016-05-25  Keith Miller  <keith_miller@apple.com>
1807
1808         Unreviewed, add JSBench to the skipped list for now since it doesn't
1809         work currently.
1810
1811         * Skipped:
1812
1813 2016-05-17  Keith Miller  <keith_miller@apple.com>
1814         We should have JSBench in PerformanceTests
1815         https://bugs.webkit.org/show_bug.cgi?id=157952
1816
1817         Rubber-stamped by Saam Barati.
1818
1819         There are some slight changes to the layout of the test directory
1820         to make it work nicely with run-jsc-benchmarks. Before JSBench
1821         had each of the browser specific sub-tests in a sub-directory.
1822         These have been flattened e.g. amazon/safari/ has become
1823         amazon-safari/.
1824
1825         * JSBench/amazon-chrome-win/urem.html: Added.
1826         * JSBench/amazon-chrome-win/urem.js: Added.
1827         * JSBench/amazon-chrome/urem.html: Added.
1828         * JSBench/amazon-chrome/urem.js: Added.
1829         * JSBench/amazon-firefox-win/urm.html: Added.
1830         * JSBench/amazon-firefox-win/urm.js: Added.
1831         * JSBench/amazon-firefox/urm.html: Added.
1832         * JSBench/amazon-firefox/urm.js: Added.
1833         * JSBench/amazon-safari/urem.html: Added.
1834         * JSBench/amazon-safari/urem.js: Added.
1835         * JSBench/browsercheck.js: Added.
1836         * JSBench/facebook-chrome-win/urem.html: Added.
1837         * JSBench/facebook-chrome-win/urem.js: Added.
1838         * JSBench/facebook-chrome/urem.html: Added.
1839         * JSBench/facebook-chrome/urem.js: Added.
1840         * JSBench/facebook-firefox-win/urem.html: Added.
1841         * JSBench/facebook-firefox-win/urem.js: Added.
1842         * JSBench/facebook-firefox/urem.html: Added.
1843         * JSBench/facebook-firefox/urem.js: Added.
1844         * JSBench/facebook-safari/urem.html: Added.
1845         * JSBench/facebook-safari/urem.js: Added.
1846         * JSBench/google-chrome-win/urem.html: Added.
1847         * JSBench/google-chrome-win/urem.js: Added.
1848         * JSBench/google-chrome/urem.html: Added.
1849         * JSBench/google-chrome/urem.js: Added.
1850         * JSBench/google-firefox-win/urem.html: Added.
1851         * JSBench/google-firefox-win/urem.js: Added.
1852         * JSBench/google-firefox/uem.html: Added.
1853         * JSBench/google-firefox/uem.js: Added.
1854         * JSBench/google-safari/urem.html: Added.
1855         * JSBench/google-safari/urem.js: Added.
1856         * JSBench/harness.html: Added.
1857         * JSBench/harness.js: Added.
1858         * JSBench/harness.py: Added.
1859         * JSBench/index.html: Added.
1860         * JSBench/reload.html: Added.
1861         * JSBench/twitter-chrome-win/rem.html: Added.
1862         * JSBench/twitter-chrome-win/rem.js: Added.
1863         * JSBench/twitter-chrome/urem.html: Added.
1864         * JSBench/twitter-chrome/urem.js: Added.
1865         * JSBench/twitter-firefox-win/urem.html: Added.
1866         * JSBench/twitter-firefox-win/urem.js: Added.
1867         * JSBench/twitter-firefox/urem.html: Added.
1868         * JSBench/twitter-firefox/urem.js: Added.
1869         * JSBench/twitter-safari/urem.html: Added.
1870         * JSBench/twitter-safari/urem.js: Added.
1871         * JSBench/yahoo-chrome-win/urem.html: Added.
1872         * JSBench/yahoo-chrome-win/urem.js: Added.
1873         * JSBench/yahoo-chrome/urem.html: Added.
1874         * JSBench/yahoo-chrome/urem.js: Added.
1875         * JSBench/yahoo-firefox-win/urem.html: Added.
1876         * JSBench/yahoo-firefox-win/urem.js: Added.
1877         * JSBench/yahoo-firefox/urem.html: Added.
1878         * JSBench/yahoo-firefox/urem.js: Added.
1879         * JSBench/yahoo-safari/urem.html: Added.
1880         * JSBench/yahoo-safari/urem.js: Added.
1881
1882 2016-05-20  Ryosuke Niwa  <rniwa@webkit.org>
1883
1884         run-benchmark's results should contain Animometer's debug output
1885         https://bugs.webkit.org/show_bug.cgi?id=157941
1886
1887         Reviewed by Stephanie Lewis.
1888
1889         Made developer.html support the JSON generated by run-benchmark which stores everything under debugOutput.
1890
1891         * Animometer/resources/debug-runner/animometer.js:
1892         (Utilities.initialize): Unwrap debugOutput in the case run-benchmark's result JSON is used.
1893
1894 2016-05-18  Timothy Hatcher  <timothy@apple.com>
1895
1896         Make Animometer work in all browsers
1897         https://bugs.webkit.org/show_bug.cgi?id=157855
1898         rdar://problem/26338521
1899
1900         Reviewed by Jon Lee.
1901
1902         * Animometer/resources/runner/logo.svg: Use a symbol instead
1903         of putting the id on the root svg. This works in all browsers.
1904
1905 2016-05-07  Timothy Hatcher  <timothy@apple.com>
1906
1907         Update Animometer design
1908         https://bugs.webkit.org/show_bug.cgi?id=157449
1909
1910         Reviewed by Darin Adler.
1911
1912         * Animometer/developer.html: Stop including runner/animometer.css.
1913         * Animometer/index.html: Udpated elements to fit new styles.
1914
1915         * Animometer/resources/debug-runner/animometer.css:
1916         Copy the old runner/animometer.css to here so the debug runner is unaffected.
1917
1918         * Animometer/resources/runner/animometer.css:
1919         New styles for the design. Drop flex box for broad browser support.
1920
1921         * Animometer/resources/runner/animometer.js:
1922         (ResultsTable.clear): Use textContent.
1923         (window.sectionsManager.showSection): Add class to the body.
1924         (window.sectionsManager.setSectionScore): Use textContent, not innerHTML.
1925
1926         * Animometer/resources/runner/crystal.svg: Added.
1927         * Animometer/resources/runner/lines.svg: Added.
1928         * Animometer/resources/runner/logo.svg: Added.
1929
1930         * Animometer/resources/runner/tests.js: Shorten test name and title case them.
1931
1932 2016-05-06  Manuel Rego Casasnovas  <rego@igalia.com>
1933
1934         [css-grid] Unprefix CSS Grid Layout properties
1935         https://bugs.webkit.org/show_bug.cgi?id=157137
1936
1937         Reviewed by Simon Fraser.
1938
1939         Remove "-webkit" prefix from all the grid layout properties,
1940         including the display value.
1941         Update the source code to remove the prefix where it was used too.
1942
1943         * Layout/auto-grid-lots-of-data.html:
1944         * Layout/fixed-grid-lots-of-data.html:
1945         * Layout/fixed-grid-lots-of-stretched-data.html:
1946
1947 2016-04-25  Geoffrey Garen  <ggaren@apple.com>
1948
1949         bmalloc: Misc improvements to MallocBench
1950         https://bugs.webkit.org/show_bug.cgi?id=157004
1951
1952         Reviewed by Darin Adler.
1953
1954         * MallocBench/run-malloc-benchmarks: Added --memory and --memory_warning
1955         modes for focused memory testing.
1956
1957         * MallocBench/MallocBench/Benchmark.cpp:
1958         (Benchmark::printReport): Clarified output.
1959
1960         (Benchmark::currentMemoryBytes): Added compressed memory because top
1961         does the same. (It always happens to zero in the benchmarks we run. But
1962         this is good for sanity.)
1963
1964         * MallocBench/MallocBench/CommandLine.cpp: Moved up to 8 runs to reduce
1965         variance.
1966
1967         * MallocBench/MallocBench/alloc_free.cpp:
1968         (benchmark_alloc_free): Cycle a single allocation in order to stress
1969         the effect of merging on calls to madvise.
1970
1971         * MallocBench/MallocBench/big.cpp:
1972         (benchmark_big): Graduated to 8kB-128kB because medium tests up to 8 and
1973         our large allocator doesn't kick in until 64kB.
1974
1975         * MallocBench/MallocBench/medium.cpp:
1976         (benchmark_medium): Test all the way down to 1kB because our large
1977         allocator used to service 1kB allocations and 1kB is an interesting
1978         middle size where memory is unusually large but allocation throughput
1979         still matters.
1980
1981         * MallocBench/MallocBench/stress.cpp:
1982         (benchmark_stress): Reduced the churn count to match stress_aligned
1983         because this test was taking too long to complete.
1984
1985         * MallocBench/MallocBench/stress_aligned.cpp:
1986         (benchmark_stress_aligned): Our new large allocator can handle even
1987         more absurdly large values.
1988
1989 2016-04-25  Simon Fraser  <simon.fraser@apple.com>
1990
1991         Add a content animation test that uses SVG animation.
1992         https://bugs.webkit.org/show_bug.cgi?id=156827
1993
1994         * Animation/svg-animation.html: Added.
1995         * Skipped:
1996
1997 2016-04-22  Simon Fraser  <simon.fraser@apple.com>
1998
1999         Skip two content animation tests which are only meant for iOS testing.
2000
2001         * Animation/css-animation.html: Added.
2002         * Animation/raf-animation.html: Added.
2003
2004         * Skipped:
2005
2006 2016-04-20  Simon Fraser  <simon.fraser@apple.com>
2007
2008         Add content animation tests to benchmark_runner, and allow the runner to collect device data as part of the results
2009         https://bugs.webkit.org/show_bug.cgi?id=156827
2010
2011         Add two files missing from the previous commit.
2012
2013         * Animation/css-animation.html: Added.
2014         * Animation/raf-animation.html: Added.
2015
2016 2016-04-20  Simon Fraser  <simon.fraser@apple.com>
2017
2018         Add content animation tests to benchmark_runner, and allow the runner to collect device data as part of the results
2019         https://bugs.webkit.org/show_bug.cgi?id=156827
2020
2021         Reviewed by Stephanie Lewis.
2022
2023         Two new files for the content-animation suite.
2024
2025         After patching for use in the benchmark, these tests are a little different to other benchmarks.
2026         The test content itself does not collect performance data; the tests spit out a boilerplate JSON
2027         result with a placeholder for framerate, which is collected in native code.
2028
2029         When run as a benchmark, test completion involves a couple of bounces between the test and MobileSafari.
2030         Test completion is initiated by a setTimeout() in the test, at which point location.hash is set to "#done".
2031         MobileSafari detects that, and asynchronously dumps performance data. Once this is complete, MobileSafari
2032         changes the URL has to "#submit", which triggers the XHR to the test relay that indicates the test is
2033         complete.
2034
2035         * Animation/css-animation.html: Added.
2036         * Animation/raf-animation.html: Added.
2037
2038 2016-04-15  Jon Lee  <jonlee@apple.com>
2039
2040         Animometer test could report a NaN
2041         https://bugs.webkit.org/show_bug.cgi?id=156646
2042
2043         Reviewed by Darin Adler.
2044         Provisionally reviewed by Said Abou-Hallawa.
2045
2046         * Animometer/tests/resources/main.js:
2047         (didFinishInterval): The ramp controller has a first phase where it ramps up the complexity and
2048         reacts based on how well the system handles the load. The assumption was that it would handle at least
2049         1 particle easily. That is not always the case. As a result, an interpolation calculation could
2050         end up setting an upper bound of NaN.
2051
2052         This occurs because we never get out of the first tier, so this._lastTierComplexity is undefined.
2053         Now that we guarantee a minimum complexity of 1, modify the conditional to check for this._lastTierComplexity
2054         before interpolating the upper bound of the first ramp. In the case where the system struggles with
2055         1 particle, set it to currentComplexity.
2056
2057 2016-04-09  Jon Lee  <jonlee@apple.com>
2058
2059         Update Animometer to accommodate different screens
2060         https://bugs.webkit.org/show_bug.cgi?id=156449
2061
2062         Reviewed by Darin Adler.
2063         Provisionally reviewed by Said Abou-Hallawa.
2064
2065         * Animometer/index.html: Wrap button in a container to add padding at the bottom.
2066         * Animometer/resources/debug-runner/animometer.css:
2067         (@media screen and (min-device-width: 1800px)): Deleted.
2068         (@media screen and (min-width: 1800px)): Cannot use min-device-width since it may match incorrectly.
2069         (screen and (max-device-height: 414px) and (orientation: landscape)): Some devices swap device width
2070         and height with orientation change.
2071         * Animometer/resources/runner/animometer.css: Similar.
2072         (screen and (min-device-width: 1024px) and (orientation: landscape)):
2073         (screen and (max-device-height: 414px) and (orientation: landscape)):
2074         (.frame-container): On smaller iPhones, adding 1px prevents the navigation bars from appearing.
2075         (@media screen and (min-device-width: 768px) and (max-device-width: 1024px)): Deleted.
2076         (@media (min-device-height: 768px) and (max-device-height: 1024px)): Target iPad Airs and similar.
2077         (@media screen and (min-device-width: 1024px) and (max-device-width: 1366px)): Deleted.
2078         (@media screen and (max-device-width: 1024px) and (min-device-height: 1366px)): Target iPad Pro.
2079         (#results footer): Add padding below the button for testing again.
2080         * Animometer/tests/master/multiply.html: Remove the center text.
2081         * Animometer/tests/master/resources/text.js: Update the test so that in every frame the text moves.
2082         * Animometer/tests/master/text.html: Update the text sizing depending on the size of the device.
2083
2084 2016-04-08  Jon Lee  <jonlee@apple.com>
2085
2086         Have Animometer benchmark always start with complexity of 1
2087         https://bugs.webkit.org/show_bug.cgi?id=156432
2088
2089         Reviewed by Ryosuke Niwa.
2090
2091         * Animometer/tests/resources/main.js: Update the default Controller and RampController to
2092         set its minimum complexities to 1 instead of 0.
2093
2094 2016-04-08  Jon Lee  <jonlee@apple.com>
2095
2096         Fix SVG benchmark test
2097         https://bugs.webkit.org/show_bug.cgi?id=156410
2098
2099         Reviewed by Dean Jackson.
2100
2101         * Animometer/resources/extensions.js: Update Point.zero to be a static Point.
2102         * Animometer/tests/simple/resources/tiled-canvas-image.js:
2103         (Stage.call._setupTiles): Refactor.
2104         * Animometer/tests/master/resources/particles.js:
2105         (Particle.prototype.reset): Use Point.center.
2106         (complexity): We are not using a gradient background anymore, so remove the +1.
2107         * Animometer/tests/master/resources/svg-particles.js: Update to use SVG transform
2108         instead of CSS transform.
2109
2110 2016-04-06  Jon Lee  <jonlee@apple.com>
2111
2112         Update master benchmark with SVG test
2113         https://bugs.webkit.org/show_bug.cgi?id=156273
2114
2115         Reviewed by Dean Jackson.
2116         Provisionally reviewed by Said Abou-Hallawa.
2117
2118         Switch masks tests for SVG path test.
2119
2120         * Animometer/resources/debug-runner/tests.js: Move mask test here.
2121         * Animometer/resources/runner/tests.js: Add SVG test here.
2122         * Animometer/tests/dom/particles.html: Renamed from PerformanceTests/Animometer/tests/master/particles.html.
2123         * Animometer/tests/dom/resources/dom-particles.js: Renamed from PerformanceTests/Animometer/tests/master/resources/dom-particles.js.
2124
2125         * Animometer/tests/master/resources/particles.js: Add minPosition for bounds checking. Prevents particle from being
2126         partially obscured.
2127         * Animometer/tests/master/resources/svg-particles.js: Added.
2128         (Particle): The particle is either a path object or a rect using a path as a clip. The
2129         same path is used either way. For each particle create a linear gradient with a random
2130         rotation.
2131         (SVGParticlesStage): Look in #shapes to see how many different kinds of paths are available.
2132         This makes the test more generic in case other shapes need to be tested.
2133         * Animometer/tests/master/svg-particles.html: Added. Have two defs, one that houses each
2134         particle's gradient, and one that holds the shape templates.
2135
2136 2016-03-27  Hunseop Jeong  <hs85.jeong@samsung.com>
2137
2138         [EFL] REGRESSION(r188793): It made 200 layout tests and Bindings/event-target-wrapper.html performance test fail
2139         https://bugs.webkit.org/show_bug.cgi?id=148470 
2140
2141         Reviewed by Darin Adler.
2142
2143         * Skipped: Unskip the Bindings/event-target-wrapper test.
2144
2145 2016-03-23  Jon Lee  <jonlee@apple.com>
2146
2147         Update focus and leaves tests
2148         https://bugs.webkit.org/show_bug.cgi?id=155825
2149
2150         Reviewed by Simon Fraser.
2151
2152         Particles would improperly disappear, before they were fully off stage. Update the calculation.
2153         * Animometer/tests/dom/resources/leaves.js:
2154         * Animometer/tests/master/resources/leaves.js:
2155
2156         * Animometer/tests/master/focus.html: Update the center element styling.
2157         * Animometer/tests/master/resources/focus.js:
2158         (FocusElement.Utilities.createClass): Give enough space in the container for each particle for the blur
2159         to spread. It should be about 3x more space than the blur radius on each side.
2160         (Stage.call.initialize): Ditto for the center element.
2161
2162 2016-03-23  Jon Lee  <jonlee@apple.com>
2163
2164         Update support for other platforms
2165         https://bugs.webkit.org/show_bug.cgi?id=155824
2166
2167         Reviewed by Simon Fraser.
2168
2169         * Animometer/resources/debug-runner/animometer.css: Remove extraneous media query. All of it should
2170         be handled in the release stylesheet.
2171         * Animometer/resources/debug-runner/animometer.js: Add support for browsers that return an HTMLCollection
2172         when retrieving radio elements with a name, instead of the selected value.
2173         * Animometer/resources/runner/animometer.css: Adjust the queries for devices of varying widths.
2174         (.frame-container): Move overflow:hidden to #stage.
2175         * Animometer/tests/master/resources/stage.css: Consolidate into tests/resources/stage.css.
2176         * Animometer/tests/resources/stage.css:
2177         (body): Remove overflow:hidden.
2178         (#stage): Set overflow:hidden here.
2179         (#center-text): Moved from master/resources/stage.css.
2180
2181         Update stylesheet hrefs.
2182         * Animometer/tests/master/canvas-stage.html:
2183         * Animometer/tests/master/focus.html:
2184         * Animometer/tests/master/leaves.html:
2185         * Animometer/tests/master/multiply.html:
2186         * Animometer/tests/master/particles.html:
2187         * Animometer/tests/master/text.html: Adjust text size to fit in the device screen.
2188
2189 2016-03-23  Jon Lee  <jonlee@apple.com>
2190
2191         Make the benchmark require device to be in landscape orientation
2192         https://bugs.webkit.org/show_bug.cgi?id=155822
2193         rdar://problem/25258650
2194
2195         Reviewed by Ryosuke Niwa.
2196
2197         Check that the device is in landscape orientation. If not, disable the button to start it,
2198         and show a warning message.
2199
2200         * Animometer/developer.html: Add a message about how the browser should be set up. Include a
2201         warning message that appears if the orientation is incorrect on mobile devices.
2202         * Animometer/index.html: Ditto.
2203
2204         * Animometer/resources/debug-runner/animometer.css: Migrate the .hidden rule into the release
2205         stylesheet. Update the style to accommodate the new UI.
2206         * Animometer/resources/runner/animometer.css:
2207
2208         * Animometer/resources/runner/animometer.js:
2209         (window.benchmarkController.initialize): Add the orientation listener is needed.
2210         (window.benchmarkController.addOrientationListenerIfNecessary): Only mobile devices need this. Check
2211         to see for support of window.orientation.
2212         (window.benchmarkController._orientationChanged): Toggle the warning based on whether we match the
2213         landscape query. Set a state variable, which is needed for the debug harness. Call updateStartButtonState.
2214         (window.benchmarkController.updateStartButtonState):
2215         * Animometer/resources/debug-runner/animometer.js: Change _updateStartButtonState to return
2216         a boolean about whether at least one test is selected. That will be used in
2217         benchmarkController.updateStartButtonState(). Move the load event listener to the release version.
2218         (window.benchmarkController.updateStartButtonState): Override the release version, and also check
2219         that at least one test is selected.
2220
2221 2016-03-22  Geoffrey Garen  <ggaren@apple.com>
2222
2223         MallocBench: consolidate regression testing for aligned allocation
2224         https://bugs.webkit.org/show_bug.cgi?id=155762
2225
2226         Reviewed by Michael Saboff.
2227
2228         * MallocBench/MallocBench.xcodeproj/project.pbxproj:
2229         * MallocBench/MallocBench/Benchmark.cpp:
2230         * MallocBench/MallocBench/memalign.cpp: Removed.
2231         * MallocBench/MallocBench/memalign.h: Removed. The stress_aligned test
2232         covers this and much more.
2233
2234         * MallocBench/MallocBench/stress_aligned.cpp:
2235         (benchmark_stress_aligned): Include specific tests for extreme sizes
2236         and alignments.
2237
2238 2016-03-21  Jon Lee  <jonlee@apple.com>
2239
2240         Update benchmark tests
2241         https://bugs.webkit.org/show_bug.cgi?id=155723
2242
2243         Reviewed by Darin Adler.
2244         Provisionally reviewed by Said Abou-Hallawa.
2245
2246         Add quadratic and bezier segments to the canvas path test.
2247
2248         * Animometer/resources/runner/tests.js: Some of the query strings are unnecessary.
2249         Rename the test.
2250         * Animometer/tests/master/resources/canvas-stage.js:
2251         (tune): Update to be able to take an array of possible constructors. Choose one randomly.
2252         * Animometer/tests/master/resources/canvas-tests.js:
2253         (CanvasLinePoint.Utilities.createClass): Move the point selection out to a separate
2254         function called randomPoint() for reuse. Move X_LOOPS, Y_LOOPS, and offsets into the
2255         class definition.
2256         (randomPoint): Scale the grid down a little bit so that the lines along the edge of the
2257         canvas are not cut off when the stroke size is thick.
2258         (CanvasQuadraticSegment): Added.
2259         (CanvasBezierSegment): Added.
2260         (SimpleCanvasStage): Pass in an array of the different segment types. Since line segments
2261         are short compared to the curved ones, make it twice as likely to render a line segment.
2262         (SimpleCanvasStage.animate): Update the drawing code so that we render all line segments.
2263
2264         Add a helper method that selects a random element from an array.
2265
2266         * Animometer/tests/resources/main.js:
2267         (Stage.randomElementInArray): Select a random element from the provided array.
2268
2269         * Animometer/tests/bouncing-particles/resources/bouncing-tagged-images.js: Refactor.
2270         * Animometer/tests/dom/resources/leaves.js: Ditto.
2271         * Animometer/tests/master/resources/dom-particles.js: Ditto.
2272         * Animometer/tests/master/resources/image-data.js: Ditto.
2273         * Animometer/tests/master/resources/leaves.js: Ditto.
2274         * Animometer/tests/simple/resources/simple-canvas-paths.js: Ditto.
2275
2276         Add canvas tests that includes all stroke and fill paths. This makes it possible to avoid having to
2277         include the full simple canvas suite for perf testing.
2278
2279         * Animometer/resources/debug-runner/tests.js: Add new tests. Move the canvas test into the Canvas
2280         suite.
2281         * Animometer/tests/simple/resources/simple-canvas-paths.js: Add a CanvasStroke and CanvasFill particle
2282         that random selects an object to render.
2283
2284         * Animometer/resources/debug-runner/tests.js: Move 3D suite before basic canvas suite.
2285         When updating the perf bot script, we will include the suites up to this one, but exclude
2286         the basic canvas suite.
2287
2288         Move compositing transforms test to HTML suite and remove the empty Miscellaneous suite.
2289
2290         * Animometer/resources/debug-runner/tests.js:
2291         * Animometer/tests/dom/compositing-transforms.html: Renamed from PerformanceTests/Animometer/tests/misc/compositing-transforms.html.
2292         * Animometer/tests/dom/resources/compositing-transforms.js: Renamed from PerformanceTests/Animometer/tests/misc/resources/compositing-transforms.js.
2293
2294         Clean up miscellaneous test suite. Add a canvas ellipse test, and remove the other
2295         canvas tests.
2296
2297         * Animometer/resources/debug-runner/tests.js: Add ellipse tests to the simple suite.
2298         * Animometer/tests/misc/canvas-electrons.html: Removed.
2299         * Animometer/tests/misc/canvas-stars.html: Removed.
2300         * Animometer/tests/misc/resources/canvas-electrons.js: Removed.
2301         * Animometer/tests/misc/resources/canvas-stars.js: Removed.
2302         * Animometer/tests/simple/resources/simple-canvas-paths.js: Add ellipse primitives.
2303
2304         Merge text tests together into one.
2305
2306         * Animometer/resources/runner/tests.js: Remove international.html.
2307         * Animometer/tests/master/international.html: Removed.
2308         * Animometer/tests/master/resources/text.js:
2309         (animate): Update styling. Manually calculate gradients for the shadow particles.
2310         Reduce the step size for y direction to avoid cutting text off at the margins. Increase
2311         step for x since there will be a little more room.
2312         * Animometer/tests/master/text.html: Add more translations and lay it out in a table.
2313
2314 2016-03-20  Jon Lee  <jonlee@apple.com>
2315
2316         Add a link to show debug data
2317         https://bugs.webkit.org/show_bug.cgi?id=155724
2318
2319         Reviewed by Simon Fraser.
2320
2321         * Animometer/developer.html: Attach the onclick handler to the score.
2322         * Animometer/index.html: Ditto.
2323
2324 2016-03-18  Dean Jackson  <dino@apple.com>
2325
2326         Add a basic WebGL test to Animometer
2327         https://bugs.webkit.org/show_bug.cgi?id=155475
2328         <rdar://problem/25156860>
2329
2330         Reviewed by Simon Fraser and Jon Lee.
2331
2332         Add a pretty simple WebGL test to Animometer
2333         which tries to draw a number of animated triangles.
2334         The result still needs a bit of tuning.
2335
2336         * Animometer/resources/debug-runner/tests.js: Add a new
2337         Suite for "3D" tests.
2338         * Animometer/tests/3d/resources/webgl.js: Added.
2339         * Animometer/tests/3d/webgl.html: Added.
2340
2341 2016-03-18  Jon Lee  <jonlee@apple.com>
2342
2343         Add support for statically linking to a specific test
2344         https://bugs.webkit.org/show_bug.cgi?id=155631
2345
2346         Rubber-stamped by Darin Adler.
2347
2348         * Animometer/developer.html: Update to call restartBenchmark instead so that it works for
2349         both the preset and the manually set test sessions.
2350         * Animometer/resources/debug-runner/animometer.css:
2351         (.tree .link): Style the "link" UI.
2352         * Animometer/resources/debug-runner/animometer.js: Move updateDisplay from suitesManager
2353         to optionsManager since this is a visual option and has nothing to do with the suites
2354         settings. Update the construction of the test UI by including a "link" after each test.
2355         Clicking on that link brings up a JS prompt with a URL and a query string with the current
2356         parameters of the controller, and selected test. Pasting this URL into the location bar will
2357         automatically start running the selected test.
2358         (suitesManager.suitesFromQueryString): Iterate through the Suites and tests and find the one
2359         that matches the provided parameters. Returns an object similar to the form in
2360         suitesManager.updateLocalStorageFromUI.
2361         (benchmarkController.initialize): After settings up the events and options, try parsing the
2362         query string and running the benchmark immediately. Otherwise, fall back to the form.
2363         (benchmarkController.startBenchmark): Store the options and suites into member variables for
2364         reuse in restartBenchmark.
2365         (benchmarkController.startBenchmarkImmediatelyIfEncoded): Convert the query string to an object.
2366         If that's successful, find the suite and test referenced in the query string. Start the benchmark if
2367         the search for the test succeeded.
2368         * Animometer/resources/extensions.js:
2369         (Utilities.stripNonASCIICharacters): Helper method to convert the name of the suite and test
2370         into a query-string-friendly version.
2371         (Utilities.convertObjectToQueryString): Helper method to convert an object to query string
2372         format.
2373         (Utilities.convertQueryStringToObject): Helper method to convert query string into an object
2374         with properties and values.
2375
2376 2016-03-16  Jon Lee  <jonlee@apple.com>
2377
2378         Add a new benchmark test
2379         https://bugs.webkit.org/show_bug.cgi?id=155570
2380
2381         Reviewed by Simon Fraser.
2382
2383         New Leaves test includes various image sizes and opacity.
2384
2385         * Animometer/resources/debug-runner/tests.js: Add it to the HTML test suite.
2386         * Animometer/tests/dom/leaves.html: Added.
2387         * Animometer/tests/dom/resources/leaves.js: Added. Override the
2388         (Particle.call.reset): Uses a range of sizes, and opacity.
2389         (Particle.call.animate): Opacity goes up then down. When it hits 0, reset the particle.
2390         (Particle.call.move): Set transform and opacity.
2391         * Animometer/tests/master/resources/leaves.js: Get rid of the closure so that it
2392         can be used in this test. Update the relative path so that it works in both the master
2393         and dom test suite.
2394
2395 2016-03-15  Simon Fraser  <simon.fraser@apple.com>
2396
2397         Add developer Animometer test that bounces P3-tagged images
2398         https://bugs.webkit.org/show_bug.cgi?id=155511
2399
2400         Reviewed by Tim Horton.
2401         
2402         Add a test for rendering performance of tagged images. The 5 images are tagged
2403         with the Display P3 colorspace.
2404
2405         * Animometer/resources/debug-runner/tests.js:
2406         * Animometer/tests/bouncing-particles/bouncing-tagged-images.html: Added.
2407         * Animometer/tests/bouncing-particles/resources/bouncing-tagged-images.js: Added.
2408         * Animometer/tests/bouncing-particles/resources/image1.jpg: Added.
2409         * Animometer/tests/bouncing-particles/resources/image2.jpg: Added.
2410         * Animometer/tests/bouncing-particles/resources/image3.jpg: Added.
2411         * Animometer/tests/bouncing-particles/resources/image4.jpg: Added.
2412         * Animometer/tests/bouncing-particles/resources/image5.jpg: Added.
2413
2414 2016-03-09  Jon Lee  <jonlee@apple.com>
2415
2416         Enhance existing Animometer tests
2417         https://bugs.webkit.org/show_bug.cgi?id=155261
2418
2419         Reviewed by Simon Fraser.
2420
2421         * Animometer/tests/master/resources/canvas-tests.js: Add a gradient to
2422         the filled circles. To expose more of the gradient, add another circle.
2423         (SimpleCanvasStage.call.animate): For each frame, create a gradient with
2424         undulating stop points and colors. Fill the circles twice; once with
2425         the solid color, and once with the gradient.
2426
2427         * Animometer/tests/master/resources/dom-particles.js: Refactor the
2428         emission variables into a separate stage for this test. Add a colorOffset
2429         variable to make the colors of each particle slightly different, since
2430         the ramp controller can add large numbers of particles all at once, which
2431         would otherwise get all the same color.
2432         * Animometer/tests/master/resources/particles.js:
2433         (initialize): Remove the code specific to the SVG mask test.
2434
2435         * Animometer/tests/master/resources/image-data.js:
2436         (initialize): Remove unused local variable.
2437         * Animometer/tests/master/resources/multiply.js:
2438         (initialize): Make the test harder by adding more total particles.
2439
2440 2016-03-09  Jon Lee  <jonlee@apple.com>
2441
2442         Add text tests
2443         https://bugs.webkit.org/show_bug.cgi?id=155257
2444
2445         Reviewed by Simon Fraser.
2446
2447         * Animometer/resources/extensions.js:
2448         (UnitBezier.Utilities.createClass): Add a class that computes Bezier points
2449         assuming that two of the control points are at (0,0) and (1,1). Taken from
2450         WebCore/platform/graphics/UnitBezier.h
2451         * Animometer/resources/runner/tests.js: Add a test for English text and one
2452         for international text.
2453         * Animometer/tests/master/international.html: Added.
2454         * Animometer/tests/master/resources/text.js: Added. The test assumes there is
2455         #template div which it will copy. The copies are placed behind the template, and
2456         are set with different colors each frame. They are moved around with CSS transform.
2457         * Animometer/tests/master/text.html: Added.
2458
2459         Remove the other text tests, since these ones cover the same techniques.
2460         * Animometer/tests/text/layering-text.html: Removed.
2461         * Animometer/tests/text/resources/layering-text.js: Removed.
2462         * Animometer/tests/text/resources/text-boxes.js: Removed.
2463         * Animometer/tests/text/text-boxes.html: Removed.
2464         * Animometer/resources/debug-runner/tests.js:
2465
2466 2016-03-09  Jon Lee  <jonlee@apple.com>
2467
2468         Add a new image test
2469         https://bugs.webkit.org/show_bug.cgi?id=155232
2470
2471         Reviewed by Dean Jackson.
2472         Provisionally reviewed by Said Abou-Hallawa.
2473
2474         The image test renders PNGs and moves them with translate and rotate
2475         transforms. Each particle has a lifetime, and when the lifetime ends or
2476         the particle goes offscreen, it resets itself somewhere on the stage.
2477
2478         * Animometer/resources/debug-runner/tests.js: Remove the CSS bouncing PNG
2479         images test, because this one tests the same technique.
2480         * Animometer/resources/runner/tests.js: Added here as "Leaves".
2481         * Animometer/tests/master/leaves.html: Added.
2482         * Animometer/tests/master/resources/particles.js: Refactor out the parts
2483         specific to the DOM particles test. Consequently make velocity a public
2484         member.
2485         (initialize): The options parameter is never used, so remove it.
2486         (animate): Remove unused local variable.
2487         * Animometer/tests/master/resources/leaves.js: Maintains a focal point that
2488         moves back and forth across the canvas. That point affects the velocity
2489         of the particles. When the focal point is on the leftmost side, it is
2490         between [-6, -2], and set in reset(). When the focal point is on the rightmost
2491         side, it will be from [2, 6].
2492         * Animometer/tests/master/resources/dom-particles.js: Move JS specific to
2493         this test here from particles.js.
2494
2495         New images.
2496         * Animometer/tests/master/resources/compass100.png: Added.
2497         * Animometer/tests/master/resources/console100.png: Added.
2498         * Animometer/tests/master/resources/contribute100.png: Added.
2499         * Animometer/tests/master/resources/debugger100.png: Added.
2500         * Animometer/tests/master/resources/inspector100.png: Added.
2501         * Animometer/tests/master/resources/layout100.png: Added.
2502         * Animometer/tests/master/resources/performance100.png: Added.
2503         * Animometer/tests/master/resources/script100.png: Added.
2504         * Animometer/tests/master/resources/shortcuts100.png: Added.
2505         * Animometer/tests/master/resources/standards100.png: Added.
2506         * Animometer/tests/master/resources/storage100.png: Added.
2507         * Animometer/tests/master/resources/styles100.png: Added.
2508         * Animometer/tests/master/resources/timeline100.png: Added.
2509
2510 2016-03-03  Jon Lee  <jonlee@apple.com>
2511
2512         Add ability to retrieve raw data from release harness
2513         https://bugs.webkit.org/show_bug.cgi?id=155026
2514
2515         Reviewed by Simon Fraser.
2516
2517         * Animometer/developer.html: Remove the special UI in the debug
2518         harness.
2519         * Animometer/resources/runner/animometer.css: Add styles for the
2520         overlay.
2521         * Animometer/resources/runner/animometer.js: Let 'j' show the JSON
2522         results, but only if the overlay doesn't exist. Add 'esc' key to dismiss
2523         the overlay.
2524         (window.benchmarkController.selectResults): Cycle the cases around
2525         so that the first 's' press selects both the benchmark score and
2526         the individual test scores.
2527         * Animometer/resources/debug-runner/animometer.css: Remove
2528         unneeded rules.
2529         * Animometer/resources/debug-runner/animometer.js: Make the same
2530         call to handleKeyPress.
2531
2532 2016-03-03  Jon Lee  <jonlee@apple.com>
2533
2534         Make sure multiply test particles have at least some opacity
2535         https://bugs.webkit.org/show_bug.cgi?id=155027
2536
2537         Reviewed by Simon Fraser.
2538
2539         * Animometer/tests/master/resources/multiply.js:
2540         (initialize): Have the elements spin a little faster to make it
2541         more obvious when the system is being stressed.
2542         (_addTile):
2543         (animate): Make sure each element being animated has at least 1%
2544         opacity.
2545
2546 2016-03-03  Jon Lee  <jonlee@apple.com>
2547
2548         Update image test
2549         https://bugs.webkit.org/show_bug.cgi?id=154962
2550
2551         Rubber-stamped by Darin Adler.
2552
2553         * Animometer/tests/master/image-data.html: Make each canvas have a compositing layer.
2554         * Animometer/tests/master/resources/image-data.js:
2555         (initialize): Update to load multiple images serially.
2556         (_loadImage):
2557         (tune): Use display instead of visibility. When showing an element anew, move it somewhere else.
2558         (_createTestElement): Set up the element, refactor out placement of element to _refreshElement.
2559         (_refreshElement): Place element in a tile grid.
2560         (animate):
2561         (_getRandomNeighboringPixelIndex): Use a more conservative distribution to make the effect last a little longer.
2562         * Animometer/tests/master/resources/compass.svg: Added.
2563         * Animometer/tests/master/resources/console.svg: Added.
2564         * Animometer/tests/master/resources/contribute.svg: Added.
2565         * Animometer/tests/master/resources/debugger.svg: Added.
2566         * Animometer/tests/master/resources/inspector.svg: Added.
2567         * Animometer/tests/master/resources/layout.svg: Added.
2568         * Animometer/tests/master/resources/performance.svg: Added.
2569         * Animometer/tests/master/resources/script.svg: Added.
2570         * Animometer/tests/master/resources/shortcuts.svg: Added.
2571         * Animometer/tests/master/resources/standards.svg: Added.
2572         * Animometer/tests/master/resources/storage.svg: Added.
2573         * Animometer/tests/master/resources/styles.svg: Added.
2574         * Animometer/tests/master/resources/timeline.svg: Added.
2575
2576 2016-03-02  Jon Lee  <jonlee@apple.com>
2577
2578         Add some new controllers, and refine tests
2579         https://bugs.webkit.org/show_bug.cgi?id=154914
2580
2581         Reviewed by Simon Fraser.
2582
2583         Improve tests.
2584
2585         * Animometer/tests/master/focus.html: Move each particle into a layer. Put the particle
2586         in a container element. Expand the container element by the max blur radius, and clip
2587         overflow. This way, when the blur is applied, it avoids causing layer resizes.
2588         * Animometer/tests/master/resources/focus.js:
2589         (FocusElement.Utilities.createClass): Create a container element, and put the particle
2590         inside.
2591         (hide): Set display:none.
2592         (show): Set display:block.
2593         (animate): Apply filters to the container element.
2594         (FocusStage.call.initialize): Instead of inserting and removing elements from the DOM,
2595         keep them in the stage, but set the display style instead. Use this._offsetIndex to
2596         keep track of which elements are displayed.
2597         (FocusStage.call.animate): Move some calculations around to avoid doing unneeded math.
2598         (FocusStage.call.getBlurValue): Make sure elements are always blurred.
2599         (FocusStage.call.getOpacityValue): Make sure elements have some opacity.
2600         * Animometer/tests/master/resources/multiply.js: Have the particles on the edge of the
2601         stage appear less black when the complexity gets large enough.
2602         (tune): Have this._distanceFactor calculate the factor. Avoid calculating square root each
2603         frame.
2604
2605 2016-03-02  Jon Lee  <jonlee@apple.com>
2606
2607         Add some new controllers, and refine tests
2608         https://bugs.webkit.org/show_bug.cgi?id=154914
2609
2610         Reviewed by Simon Fraser.
2611
2612         Add a controller that centers around 30 fps instead of 60 fps.
2613
2614         * Animometer/developer.html: Add a new option.
2615         * Animometer/resources/debug-runner/animometer.js:
2616         * Animometer/resources/runner/animometer.js:
2617         (this._processData.findRegression): When calculating the complexity-frameLength regression,
2618         check the controller, and use a 30 fps baseline if needed.
2619         * Animometer/resources/statistics.js:
2620         (Regression.Utilities.createClass): Update to allow clients to specify the baseline
2621         frame length.
2622         * Animometer/tests/resources/main.js:
2623         (tune): Override some of the constants in RampController. Move those constants out for
2624         easier reading.
2625
2626         Add a fixed controller, with no step.
2627
2628         * Animometer/developer.html: Add a controller that takes no step.
2629         * Animometer/resources/debug-runner/animometer.js:
2630         * Animometer/tests/resources/main.js:
2631         (Rotater.Utilities.createClass):
2632
2633         Switch to ramp controller as default.
2634
2635         * Animometer/developer.html: Increase the test length to 20 seconds.
2636         * Animometer/resources/debug-runner/animometer.js: Rename the "adjustment" field to "controller"
2637         since that is a more accurate description.
2638         * Animometer/resources/debug-runner/graph.js:
2639         * Animometer/resources/runner/animometer.js: Update preferences for release suite.
2640
2641 2016-02-26  Jon Lee  <jonlee@apple.com>
2642
2643         Address Dean's comments in 154673.
2644
2645         * Animometer/developer.html:
2646         * Animometer/resources/debug-runner/animometer.js:
2647         * Animometer/resources/debug-runner/graph.js:
2648
2649 2016-02-25  Jon Lee  <jonlee@apple.com>
2650
2651         Update animation benchmark and tests
2652         https://bugs.webkit.org/show_bug.cgi?id=154673
2653
2654         Reviewed by Dean Jackson.
2655
2656         Update test visuals.
2657
2658         * Animometer/tests/master/resources/canvas-tests.js:
2659         (CanvasLineSegment.Utilities.createClass): Line segments near the edge of the stage get clipped.
2660         Update circle position and radius to minimize impact.
2661         * Animometer/tests/master/resources/particles.js: Get rid of rotating gradient background, and
2662         have 3 locations for emitting particles.
2663         * Animometer/tests/master/resources/multiply.js: Update distance metric so that fringe tiles get
2664         more color.
2665         (initialize): Fix some of the math for laying out the tiles.
2666         (animate): When a tile isn't used, set visibility: hidden.
2667
2668 2016-02-24  Jon Lee  <jonlee@apple.com>
2669
2670         Update animation benchmark and tests
2671         https://bugs.webkit.org/show_bug.cgi?id=154673
2672
2673         Reviewed by Dean Jackson.
2674
2675         Update the ramp controller.
2676
2677         The controller refines the complexity interval to test across.
2678
2679         * Animometer/resources/statistics.js: Add functions that estimate cumulative distribution function.
2680         (Regression): For the flat regression, force the first segment to be at 60 fps.
2681         (valueAt): Add convenience function to return interpolated value based on the regression used.
2682         (_calculateRegression): Include the number of points included for both segments, and the piecewise
2683         errors.
2684         * Animometer/tests/resources/math.js: Make the Kalman estimator subclass Experiment, and allow it
2685         to be reset.
2686
2687         * Animometer/tests/resources/main.js: Initialize the tier such that it starts at 10^0 = 1.
2688         Increase the number of ramps. Maintain three FPS thresholds-- the frame rate of interest, a limit
2689         on the lowest FPS we care to go for later interpolation, and a minimum FPS threshold we want to
2690         aim for each ramp. Also keep three estimators: a running average of the change point, a minimum
2691         boundary for each ramp, and an estimator for all the frames within an interval. The first two
2692         are used to determine the parameters of the next ramp, and the latter allows us to refine the
2693         parameters.
2694         (update): During the tier phase, it is possible that the highest complexity possible for a test
2695         won't stress the system enough to trigger stopping the tier phase and transitioning to the ramps.
2696         If the complexity doesn't change when going to the next tier, we've maxed the test out, and move
2697         on. When the tier phase completed, turn off Controller.frameLengthEstimator, which estimates the
2698         FPS at each tier.
2699         (tune): At each interval, look at the confidence distribution of being on the 60 FPS side or the
2700         slow side. If the slowest FPS we achieve at the ramp's maximum complexity is not at least
2701         _fpsRampSlowThreshold, then increase the maximum complexity. If we ever achieve 60 FPS, increase
2702         the ramp's minimum complexity to that level. If, at an even lower complexity, a glitch causes the
2703         FPS to drop, we reset the minimum complexity.
2704
2705         Have the bootstrap calculation occur between tests. Clean up harness.
2706
2707         * Animometer/resources/debug-runner/animometer.js: Run bootstrap after a test has
2708         completed to avoid doing all of it at the end before showing the results. Clean up
2709         parameters being passed around.
2710         * Animometer/resources/debug-runner/tests.js:
2711         (text):
2712         * Animometer/resources/runner/animometer.js:
2713         (this._processData.calculateScore): Save the results to the same object holding the data.
2714         (this._processData._processData): In the case where a file is dragged, calculate the score
2715         serially. Grab the results object and move it to the results variable and remove it from
2716         the data object. This avoids serializing the results into the JSON.
2717         (this._processData.findRegression): Include the samples used for bootstrapping. Reduce the
2718         resample size to shorten the wait.
2719         * Animometer/resources/runner/benchmark-runner.js:
2720         * Animometer/resources/statistics.js:
2721         (bootstrap): Update how bootstrapData is sorted. In some regression results the mix of
2722         floats and integers causes an alphabetical sort to occur.
2723         * Animometer/resources/strings.js:
2724
2725         Add meta charset so that encodings between harness and test match.
2726
2727         * Animometer/tests/bouncing-particles/bouncing-canvas-images.html:
2728         * Animometer/tests/bouncing-particles/bouncing-canvas-shapes.html:
2729         * Animometer/tests/bouncing-particles/bouncing-css-images.html:
2730         * Animometer/tests/bouncing-particles/bouncing-css-shapes.html:
2731         * Animometer/tests/bouncing-particles/bouncing-svg-images.html:
2732         * Animometer/tests/bouncing-particles/bouncing-svg-shapes.html:
2733         * Animometer/tests/master/canvas-stage.html:
2734         * Animometer/tests/master/focus.html:
2735         * Animometer/tests/master/image-data.html:
2736         * Animometer/tests/master/multiply.html:
2737         * Animometer/tests/master/particles.html:
2738         * Animometer/tests/misc/canvas-electrons.html:
2739         * Animometer/tests/misc/canvas-stars.html:
2740         * Animometer/tests/misc/compositing-transforms.html:
2741         * Animometer/tests/simple/simple-canvas-paths.html:
2742         * Animometer/tests/simple/tiled-canvas-image.html:
2743         * Animometer/tests/template/template-canvas.html:
2744         * Animometer/tests/template/template-css.html:
2745         * Animometer/tests/template/template-svg.html:
2746         * Animometer/tests/text/layering-text.html:
2747         * Animometer/tests/text/text-boxes.html:
2748
2749         Update test harness reporting.
2750
2751         * Animometer/developer.html: Add missing meta charset.
2752         * Animometer/index.html: Remove unnecessary utf-8 declaration.
2753         * Animometer/resources/debug-runner/animometer.css: Add convenience classes for
2754         formatting the results table.
2755         * Animometer/resources/debug-runner/animometer.js: Adjust which stats are shown.
2756         * Animometer/resources/debug-runner/tests.js: Display bootstrapping statistics.
2757         * Animometer/resources/strings.js: Move strings not used by the release harness.
2758
2759         Switch to a pseudo-random number generator.
2760
2761         * Animometer/resources/statistics.js: Add a Pseudo class, with a simple
2762         pseudo-random number generator.
2763         (_calculateRegression): Reset the generator before running bootstrap.
2764         (bootstrap): Deleted.
2765
2766         Replace Math.random with Pseudo.random.
2767         * Animometer/tests/master/resources/canvas-tests.js:
2768         * Animometer/tests/master/resources/focus.js:
2769         * Animometer/tests/master/resources/particles.js:
2770         * Animometer/tests/resources/main.js:
2771
2772         Use bootstrapping to get confidence interval in the breakpoint.
2773
2774         For the ramp controller, calculate the piecewise regression, and then use
2775         bootstrapping in order to find the 95% confidence interval. Use the raw data.
2776
2777         * Animometer/developer.html: Default to the complexity graph. Add a legend
2778         checkbox to toggle visibility of the bootstrap score and histogram.
2779         * Animometer/resources/debug-runner/animometer.css: Make some more space to show
2780         the old raw and average scores in the legend. Add new styles for the data.
2781         * Animometer/resources/debug-runner/graph.js:
2782         (_addRegressionLine): Allow passing an array for the variance bar tied to the
2783         regression line. Now |stdev| is |range|.
2784         (createComplexityGraph): Add bootstrap median, and overlay a histogram of
2785         the bootstrap samples. Switch raw samples from circles to X's.
2786         (onComplexityGraphOptionsChanged): Allow toggling of the bootstrap data.
2787         (onGraphTypeChanged): Move the regressions for the raw and average samples to the
2788         legend. In the subtitle use the bootstrap median, and include the 95% confidence
2789         interval.
2790         * Animometer/resources/runner/animometer.js:
2791         (this._processData.findRegression): Factor out the code that determines which
2792         samples to include when calculating the piecewise regression. For series that have
2793         many samples, or a wider range of recorded complexities, throw away the 2.5%
2794         lowest and highest samples before calculating the regression. Keep all samples
2795         if the number of samples to regress is small or the range of complexities is
2796         narrow.
2797         (this._processData._calculateScore): Factor out regression calculation to
2798         findRegression(). Bootstrap the change point of the regression. The score is the
2799         median.
2800         * Animometer/resources/statistics.js:
2801         (_calculateRegression): Correct an issue in the calculation of the regression, where
2802         the denominator can be 0.
2803         (bootstrap): Template for bootstrapping. Create a bootstrap sample array, Create
2804         re-samples by random selection with replacement. Return the 95% confidence samples,
2805         the bootstrap median, mean, and the data itself.
2806         * Animometer/resources/strings.js: Add bootstrap.
2807         * Animometer/tests/resources/main.js:
2808         (processSamples): Don't prematurely cut the sample data.
2809
2810         Fix graph drawing.
2811
2812         * Animometer/resources/debug-runner/animometer.js: Add spacing in the JSON output.
2813         Multiple tests output a lot of JSON and can hang when selecting JSON with no whitespace.
2814         * Animometer/resources/debug-runner/animometer.css:
2815         (#complexity-graph .series.raw circle): Update the color.
2816         * Animometer/resources/debug-runner/graph.js: Use the FPS axis instead of the
2817         complexity axis, which can vary in domain. For determining the complexity domain,
2818         only use samples after samplingTimeOffset.
2819
2820         Allow dropping results JSON.
2821
2822         * Animometer/developer.html: Add a button.
2823         * Animometer/resources/debug-runner/animometer.css:
2824         * Animometer/resources/debug-runner/animometer.js: Read the data and go straight
2825         to the dashboard. With JSON output, write out only the options and the raw data.
2826
2827         Teach the harness to evaluate the samples and determine the test score.
2828
2829         This will allow us to update how the score is calculated separately from the samples recorded.
2830         This also prepares the harness to be able to accept JSON of prior runs.
2831
2832         * Animometer/resources/strings.js: Clean up and remove unneeded strings and reduce some of the
2833         hierarchy.
2834         * Animometer/resources/debug-runner/tests.js: Update to use the new strings.
2835
2836         * Animometer/tests/resources/main.js: Allow all controllers to show a complexity-FPS graph.
2837         (_processComplexitySamples): Factor out some of the sample processing done in the ramp
2838         controller for the benefit of the other controllers. |complexitySamples| contains a list of
2839         samples. Sort the samples by complexity. Optionally remove the top x% of samples.
2840         Group them, and calculate distribution of samples within the same complexity, and add those as
2841         new entries into |complexityAverageSamples|.
2842         (Controller.processSamples): Move the code responsible for determining the complexity and FPS
2843         scores out to ResultsDashboard. The structure of the data returned by the controller is:
2844
2845         {
2846             controller: [time-regression, time-regression, ...], // optional, data specific to controller
2847             marks: [...],
2848             samples: {                    // all of the sample data
2849                 controller: [...],
2850                 complexity: [...],        // processed from controller samples
2851                 complexityAverage: [...], // processed from complexity samples
2852             }
2853         }
2854
2855         (AdaptiveController.processSamples): Adding the target frame length is no longer necessary; we
2856         now pass the test options to the graph.
2857         (Regression): Move to statistics.js.
2858         * Animometer/resources/statistics.js: Move Regression to here. Add a check if the sampling range
2859         only contains one sample, since we cannot calculate a regression from one sample point.
2860
2861         Teach the test harness to evaluate the data.
2862         * Animometer/resources/runner/animometer.js:
2863         (ResultsDashboard): Store the options used to run the test and the computed results/score separately
2864         from the data. The results are stored as:
2865
2866         {
2867             score: /* geomean of iteration score */,
2868             iterationsResults: [
2869                 {
2870                     score: /* geomean of tests */,
2871                     testsResults: {
2872                         suiteName: {
2873                             testName: {
2874                                 controller: {
2875                                     average:
2876                                     concern:
2877                                     stdev:
2878                                     percent:
2879                                 },
2880                                 frameLength: { ... },
2881                                 complexity: {
2882                                     complexity:
2883                                     stdev:
2884                                     segment1:
2885                                     segment2:
2886                                 },
2887                                 complexityAverage: { ... }
2888                             },
2889                             testName: { ... },
2890                         },
2891                         ... next suite ...
2892                     }
2893                 },
2894                 { ...next iteration... }
2895             ]
2896         }
2897
2898         * Animometer/resources/debug-runner/animometer.js: Pass options around instead of relying
2899         on what was selected in the form. This will later allow for dropping previous results, and
2900         using those runs' options when calculating scores.
2901         (ResultsTable._addGraphButton): Simplify button action by using attached test data.
2902         * Animometer/resources/debug-runner/graph.js: Refactor to use the data.
2903
2904         Consolidate JS files, and move statistics out to a separate JS.
2905
2906         Preparation for having the Controller only handle recording and storage of the samples,
2907         and leave the evaluation of the test score out to the harness. Move Experiment to
2908         a new statistics.js, where Regression will also eventually go. Get rid of algorithm.js
2909         and move it to utilities.js since the Heap is used only for Experiments.
2910
2911         * Animometer/tests/resources/algorithm.js: Removed. Heap is in utilities.js.
2912         * Animometer/tests/resources/sampler.js: Removed. Experiment is in statistics.js,
2913         Sampler in main.js.
2914         * Animometer/tests/resources/main.js: Move Sampler here.
2915         * Animometer/resources/statistics.js: Added. Move Statistics and Experiment here.
2916         * Animometer/resources/extensions.js: Move Heap here. Attach static method to create
2917         a max or min heap to Heap, instead of a new Algorithm object.
2918
2919         Update JS files.
2920         * Animometer/developer.html:
2921         * Animometer/index.html:
2922         * Animometer/tests/bouncing-particles/bouncing-canvas-images.html:
2923         * Animometer/tests/bouncing-particles/bouncing-canvas-shapes.html:
2924         * Animometer/tests/bouncing-particles/bouncing-css-images.html:
2925         * Animometer/tests/bouncing-particles/bouncing-css-shapes.html:
2926         * Animometer/tests/bouncing-particles/bouncing-svg-images.html:
2927         * Animometer/tests/bouncing-particles/bouncing-svg-shapes.html:
2928         * Animometer/tests/master/canvas-stage.html:
2929         * Animometer/tests/master/focus.html:
2930         * Animometer/tests/master/image-data.html:
2931         * Animometer/tests/master/multiply.html:
2932         * Animometer/tests/master/particles.html:
2933         * Animometer/tests/misc/canvas-electrons.html:
2934         * Animometer/tests/misc/canvas-stars.html:
2935         * Animometer/tests/misc/compositing-transforms.html:
2936         * Animometer/tests/simple/simple-canvas-paths.html:
2937         * Animometer/tests/simple/tiled-canvas-image.html:
2938         * Animometer/tests/template/template-canvas.html:
2939         * Animometer/tests/template/template-css.html:
2940         * Animometer/tests/template/template-svg.html:
2941         * Animometer/tests/text/layering-text.html:
2942         * Animometer/tests/text/text-boxes.html:
2943
2944         Fix the cursor in the graph analysis when the min
2945         complexity is not 0.
2946
2947         * Animometer/resources/debug-runner/graph.js:
2948         (_addRegression):
2949         (createComplexityGraph):
2950
2951 2016-02-23  Geoffrey Garen  <ggaren@apple.com>
2952
2953         Fix some issues in MallocBench
2954         https://bugs.webkit.org/show_bug.cgi?id=154600
2955
2956         Reviewed by Oliver Hunt.
2957
2958         * MallocBench/MallocBench.xcodeproj/project.pbxproj: Make nimlang.ops
2959         a part of the build so that it copies to the right place for execution.
2960
2961         * MallocBench/MallocBench/Interpreter.cpp:
2962         (Interpreter::Interpreter): Don't try to open .ops files for writing
2963         because we're only going to read and we might need extra permissions to
2964         write.
2965
2966         (Interpreter::~Interpreter): Give more context when opening a file fails
2967         to help with debugging.
2968
2969         * MallocBench/MallocBench/stress.cpp:
2970         (benchmark_stress): Reduce iterations to complete in less than 10 seconds.
2971
2972 2016-02-23  Michael Saboff  <msaboff@apple.com>
2973
2974         Unreviewed change to revert extraneous changes made part of change set 196955.
2975
2976         * MallocBench/MallocBench/Interpreter.cpp:
2977         (Interpreter::doMallocOp):
2978         (Interpreter::Thread::switchTo):
2979         (writeData): Deleted.
2980
2981 2016-02-19  Michael Saboff  <msaboff@apple.com>
2982
2983         MallocBench: Added recording for nimlang website, new recording details and added new options
2984         https://bugs.webkit.org/show_bug.cgi?id=154485
2985
2986         Reviewed by Geoff Garen.
2987
2988         Added new capabilities to MallocBench.  These include:
2989             Added a recording of http://nim-lang.org/docs/lib.html.
2990             Added thread id to the recording and the ability to playback switching threads in MallocBench
2991             Added aligned allocations to recordings and the ability to playback
2992             Added --use-thread-id option to honor recorded thread ids
2993             Added --detailed-report to output remaining allocations by size after playback
2994             Added --no-warmup to not run the warm up iteration
2995
2996         Changed the way that options are passed down to the benchmarks.  Instead of passing individual
2997         boolean or numeric option values, just pass a reference the CommandLine itself.  Each benchmark
2998         can access the options that are appropriate.  The Benchmark class also uses the options for
2999         is parallel, run counts and warm up.
3000
3001         Added thread id and aligned malloc to the Op by noticing that structure padding and Opcode allowed
3002         for another 32 bits of data.  Breaking that unused 32 bits into a 16 bit thread id value and a
3003         16 bit log base 2 of the alignment for aligned malloc allowed for existing recordings to playback
3004         without any incompatibilities.
3005
3006         Threaded operation is simulated by creating threads as needed.  As long as the next Op's thread id
3007         is the same as the last, operation continues as normal.  When the next Op has a different thread id,
3008         we switch to that thread using the shared Op stream to continue playing back.  There is a mutex to
3009         assure that only one thread is really running at a time and a condition variable used to wait
3010         that the current thread id matches each block thread's thread id.  This doesn't simulate true
3011         concurrent threading, but is instead plays back Ops recorded for multiple thread faithfully.
3012
3013         * MallocBench/MallocBench.xcodeproj/project.pbxproj:
3014         * MallocBench/MallocBench/Benchmark.cpp:
3015         (deallocateHeap):
3016         (Benchmark::Benchmark):
3017         (Benchmark::runOnce):
3018         (Benchmark::run):
3019         * MallocBench/MallocBench/Benchmark.h:
3020         (Benchmark::isValid):
3021         * MallocBench/MallocBench/CommandLine.cpp:
3022         (CommandLine::printUsage):
3023         * MallocBench/MallocBench/CommandLine.h:
3024         (CommandLine::isValid):
3025         (CommandLine::benchmarkName):
3026         (CommandLine::isParallel):
3027         (CommandLine::useThreadID):
3028         (CommandLine::detailedReport):
3029         (CommandLine::warmUp):
3030         (CommandLine::heapSize):
3031         (CommandLine::runs):
3032         * MallocBench/MallocBench/Interpreter.cpp:
3033         (Interpreter::Interpreter):
3034         (Interpreter::run):
3035         (Interpreter::readOps):
3036         (Interpreter::doOnSameThread):
3037         (Interpreter::switchToThread):
3038         (Interpreter::detailedReport):
3039         (compute2toPower):
3040         (writeData):
3041         (Interpreter::doMallocOp):
3042         (Interpreter::Thread::Thread):
3043         (Interpreter::Thread::stop):
3044         (Interpreter::Thread::~Thread):
3045         (Interpreter::Thread::runThread):
3046         (Interpreter::Thread::waitToRun):
3047         (Interpreter::Thread::switchTo):
3048         * MallocBench/MallocBench/Interpreter.h:
3049         (Interpreter::Thread::isMainThread):
3050         * MallocBench/MallocBench/alloc_free.cpp: Added.
3051         (benchmark_alloc_free):
3052         * MallocBench/MallocBench/alloc_free.h: Added.
3053         * MallocBench/MallocBench/balloon.cpp:
3054         (benchmark_balloon):
3055         * MallocBench/MallocBench/balloon.h:
3056         * MallocBench/MallocBench/big.cpp:
3057         (benchmark_big):
3058         * MallocBench/MallocBench/big.h:
3059         * MallocBench/MallocBench/churn.cpp:
3060         (benchmark_churn):
3061         * MallocBench/MallocBench/churn.h:
3062         * MallocBench/MallocBench/facebook.cpp:
3063         (benchmark_facebook):
3064         * MallocBench/MallocBench/facebook.h:
3065         * MallocBench/MallocBench/flickr.cpp:
3066         (benchmark_flickr):
3067         (benchmark_flickr_memory_warning):
3068         * MallocBench/MallocBench/flickr.h:
3069         * MallocBench/MallocBench/fragment.cpp:
3070         (validate):
3071         (benchmark_fragment):
3072         (benchmark_fragment_iterate):
3073         * MallocBench/MallocBench/fragment.h:
3074         * MallocBench/MallocBench/list.cpp:
3075         (benchmark_list_allocate):
3076         (benchmark_list_traverse):
3077         * MallocBench/MallocBench/list.h:
3078         * MallocBench/MallocBench/main.cpp:
3079         (main):
3080         * MallocBench/MallocBench/medium.cpp:
3081         (benchmark_medium):
3082         * MallocBench/MallocBench/medium.h:
3083         * MallocBench/MallocBench/memalign.cpp:
3084         (test):
3085         (benchmark_memalign):
3086         * MallocBench/MallocBench/memalign.h:
3087         * MallocBench/MallocBench/message.cpp:
3088         (benchmark_message_one):
3089         (benchmark_message_many):
3090         * MallocBench/MallocBench/message.h:
3091         * MallocBench/MallocBench/nimlang.cpp: Added.
3092         (benchmark_nimlang):
3093         * MallocBench/MallocBench/nimlang.h: Added.
3094         * MallocBench/MallocBench/nimlang.ops: Added.
3095         * MallocBench/MallocBench/realloc.cpp:
3096         (benchmark_realloc):
3097         * MallocBench/MallocBench/realloc.h:
3098         * MallocBench/MallocBench/reddit.cpp:
3099         (benchmark_reddit):
3100         (benchmark_reddit_memory_warning):
3101         * MallocBench/MallocBench/reddit.h:
3102         * MallocBench/MallocBench/stress.cpp:
3103         (deallocate):
3104         (benchmark_stress):
3105         * MallocBench/MallocBench/stress.h:
3106         * MallocBench/MallocBench/stress_aligned.cpp:
3107         (benchmark_stress_aligned):
3108         * MallocBench/MallocBench/stress_aligned.h:
3109         * MallocBench/MallocBench/theverge.cpp:
3110         (benchmark_theverge):
3111         (benchmark_theverge_memory_warning):
3112         * MallocBench/MallocBench/theverge.h:
3113         * MallocBench/MallocBench/tree.cpp:
3114         (benchmark_tree_allocate):
3115         (benchmark_tree_traverse):
3116         (benchmark_tree_churn):
3117         * MallocBench/MallocBench/tree.h:
3118         * MallocBench/run-malloc-benchmarks:
3119
3120 2016-02-11  Jon Lee  <jonlee@apple.com>
3121
3122         Fix a missing refactoring.
3123
3124         * Animometer/tests/master/resources/multiply.js:
3125         (animate): Move to Utilities.lerp.
3126         (_lerp): Deleted.
3127
3128 2016-02-10  Jon Lee  <jonlee@apple.com>
3129
3130         Add new benchmark tests.
3131         https://bugs.webkit.org/show_bug.cgi?id=154063
3132
3133         Provisionally reviewed by Said Abou-Hallawa.
3134
3135         Add tests for get/put image data, filters, opacity, and css transforms.
3136
3137         * Animometer/resources/runner/benchmark-runner.js:
3138         (_runBenchmarkAndRecordResults): Update the body background color to match that of
3139         the stage.
3140         (this._runNextIteration): Clear the background color style for the results page.
3141         * Animometer/resources/runner/tests.js:
3142         * Animometer/tests/master/focus.html: Added.
3143         * Animometer/tests/master/image-data.html: Added.
3144         * Animometer/tests/master/multiply.html: Added.
3145         * Animometer/tests/master/resources/focus.js: Added.
3146         * Animometer/tests/master/resources/image-data.js: Added.
3147         * Animometer/tests/master/resources/multiply.js: Added.
3148         * Animometer/tests/master/resources/stage.css: Move common styles out.
3149         * Animometer/tests/resources/main.js: Update Stage.randomBool to use Math.random.
3150         Add Stage.randomSign for randomly setting a direction. Add the notion of the
3151         current timestamp of the test to Benchmark, since some animations cycle through
3152         colors and rely on an incremental counter like the time.
3153
3154 2016-02-09  Said Abou-Hallawa  <sabouhallawa@apple.com>
3155
3156         Add internal benchmark tests for CSS mix-blend-modes and filters
3157         https://bugs.webkit.org/show_bug.cgi?id=154058
3158
3159         Provisionally reviewed by Jon Lee.
3160
3161         * Animometer/resources/debug-runner/tests.js: Include the new tests in the
3162         "HTML suite" of the debug runner.
3163
3164         * Animometer/resources/extensions.js:
3165         (Utilities.browserPrefix):
3166         (Utilities.setElementPrefixedProperty): Utility functions to allow setting
3167         prefixed style properties.
3168
3169         * Animometer/tests/bouncing-particles/resources/bouncing-css-shapes.js:
3170         Set the mix-blend-mode and the filter to some random values if the options
3171         of the test requested that.
3172
3173         * Animometer/tests/bouncing-particles/resources/bouncing-particles.js:
3174         (parseShapeParameters): Parse the url options "blend" and "filter" and set
3175         the corresponding flags.
3176
3177         * Animometer/tests/resources/main.js:
3178         (randomStyleMixBlendMode):
3179         (randomStyleFilter): Return random mix-blend-mode and filter.
3180
3181 2016-02-08  Jon Lee  <jonlee@apple.com>
3182
3183         Add a ramp controller
3184         https://bugs.webkit.org/show_bug.cgi?id=154028
3185
3186         Provisionally reviewed by Said Abou-Hallawa.
3187
3188         Enhance the graph to include a complexity-fps graph, in addition
3189         to the time graph.
3190
3191         * Animometer/developer.html: Add a ramp option.
3192         * Animometer/resources/debug-runner/animometer.css: Update the style.
3193         * Animometer/resources/strings.js: Flatten the Strings.text constants.
3194         * Animometer/resources/debug-runner/animometer.js:
3195         (ResultsTable.call._addGraphButton): Refactor.
3196         (ResultsTable.call._addTest): Add regression data.
3197         (benchmarkController): Add a form that allows the user to switch between the two forms,
3198         Add a form that allows the user to toggle different data. Hide certain header columns
3199         depending on the selected controller.
3200         * Animometer/resources/debug-runner/graph.js: Add the complexity regressions.
3201         * Animometer/resources/debug-runner/tests.js: Add headers for the ramp results.
3202         * Animometer/resources/runner/animometer.js:
3203         (ResultsTable): If a header is disabled don't include them in _flattenedHeaders.
3204         * Animometer/tests/resources/main.js:
3205         (Controller): Allow options to specify the capacity for sample arrays.
3206         (Regression): A piecewise regression that tries to fit a slope and a flat profile.
3207         (_calculateRegression): Options can fix the slope and bias when calculating the minimal
3208         error. Sweep across the samples in time (which could be backward depending on the controller)
3209         and calculate the intersection point.
3210         (RampController): This controller assumes that the target frame rate is below
3211         58 FPS. It runs in two stages. The first stage quickly determines the order of
3212         magnitude of objects needed to stress the system by the setting the complexity
3213         to increasingly difficult tiers. Perform a series of ramps descending from a
3214         high-water mark of complexity. The complexity needed to reach the target frame
3215         length is done by performing a piecewise regression on each ramp, and track a
3216         running average of these values. For the next ramp, make that running average
3217         the center of the ramp. With a minimum complexity of 0, the high-water mark is
3218         twice that average. The score is based on the highest complexity that can
3219         reach 60 fps.
3220
3221 2016-02-08  Jon Lee  <jonlee@apple.com>
3222
3223         Address Said's comments on the benchmark, and do some clean up.
3224
3225         * Animometer/developer.html:
3226         * Animometer/resources/debug-runner/animometer.css: Add styles for averages.
3227         * Animometer/resources/debug-runner/animometer.js: Use the right
3228         Strings constants.
3229         * Animometer/resources/debug-runner/graph.js:
3230         (_addRegressionLine): Add missing code to draw the line and standard
3231         deviation highlight.
3232         (onGraphTypeChanged): Remove unneeded variables
3233         (onTimeGraphOptionsChanged):
3234         * Animometer/resources/runner/benchmark-runner.js:
3235         (_runBenchmarkAndRecordResults): Rename samplers to suiteResults and
3236         _suitesSamplers to _suitesResults.
3237         * Animometer/tests/resources/main.js:
3238         (results): Call processSamples().
3239         (update): Change sampling timestamp comparison.
3240         (_animateLoop): Move shouldStop call to before the update.
3241         * Animometer/tests/resources/sampler.js:
3242         (process): Rename to processSamples().
3243
3244 2016-02-07  Jon Lee  <jonlee@apple.com>
3245
3246         Teach Controller to measure intervals, and turn off the frame length estimator.
3247
3248         * Animometer/tests/resources/main.js: Default interval length is 100 ms.
3249         (start): Set the first interval.
3250         (_measureAndResetInterval): Reports the average frame length of the interval that just
3251         completed, and sets up the next interval.
3252         (update): If there is no length, then just use the estimator per frame, otherwise the
3253         estimator measures per interval. Add a didFinishInterval for subclasses to process
3254         prior to recording the sample. Update tune() to include whether an interval had
3255         finished.
3256         (StepController): Step controllers don't measure on an interval basis.
3257
3258 2016-02-07  Jon Lee  <jonlee@apple.com>
3259
3260         Minor refactoring. Rename Controller._estimator to Controller._frameLengthEstimator
3261         and switch the parameters for start(), update(), and tune(), so that the timestamp
3262         is first and stage is second.
3263
3264         * Animometer/tests/resources/main.js:
3265
3266 2016-02-07  Jon Lee  <jonlee@apple.com>
3267
3268         Move ResultsTable functionality not needed for release tests out.
3269         Move reporting of score and mean to selection of the time-based graph.
3270
3271         * Animometer/developer.html: Rename graph-options to time-graph-options.
3272         * Animometer/resources/debug-runner/animometer.js:
3273         (DeveloperResultsTable): Moved from runner/animometer.js. Switch from mean
3274         values to "average" objects which can hold stdev. Move graph button and
3275         calculation of noisy measurements here. Sophisticated header processing
3276         is not needed in release suite.
3277         (populateTable): Use DeveloperResultsTable.
3278         * Animometer/resources/debug-runner/graph.js: Pull time graph creation to
3279         its own function, and add a new onGraphTypeChanged handler in preparation
3280         of a complexity graph to be added later.
3281         * Animometer/resources/runner/animometer.js:
3282         (ResultsTable): Simplify to just handle test names and scores.
3283
3284 2016-02-07  Jon Lee  <jonlee@apple.com>
3285
3286         Tests: reuse objects already made.
3287
3288         Avoid thrash of object creation and removal by maintaining an index that
3289         moves along the array as the adjust values change. If the tune value
3290         requires more objects than the maximum size of the object array, then create
3291         new objects. This means that the object array size never decreases.
3292
3293         * Animometer/tests/master/resources/canvas-stage.js: Maintain a separate
3294         offsetIndex. For these tests, we want to avoid drawing the oldest objects,
3295         so the scene will draw the object at offsetIndex to the end of the array.
3296         (tune): Reverse the logic since "removal" of objects is much simpler and
3297         involves simply changing the offsetIndex.
3298         (animate): Update the for loop to draw from offsetIndex to the end.
3299         (complexity): Update the definition.
3300         * Animometer/tests/master/resources/canvas-tests.js: Maintain a separate
3301         offsetIndex. For these tests, we want to avoid drawing the newest objects,
3302         so the scene will draw the object at index 0 to the object at offsetIndex.
3303         (SimpleCanvasStage.animate): Fly-by removal of local stage variable,
3304         which is unneeded. Update the for loop to draw from offsetIndex to the end.
3305         * Animometer/tests/simple/resources/simple-canvas-paths.js:
3306         (SimpleCanvasStage.animate): Update the for loop to draw from 0 to
3307         offsetIndex.
3308         * Animometer/tests/simple/resources/simple-canvas.js:
3309         (tune): Update logic. Here, offsetIndex represents the boundary of the last
3310         index to render.
3311         (animate): Update the for loop to draw from 0 to offsetIndex.
3312         (complexity): Update the definition.
3313
3314 2016-02-07  Jon Lee  <jonlee@apple.com>
3315
3316         Tests: refactor and update styles.
3317
3318         * Animometer/tests/resources/main.js: Add helper methods that return
3319         a color that hue rotates based on the date, and a counter value that
3320         increases based on the date. Fix randomInt() to not bias against the min and
3321         max values.
3322
3323         * Animometer/tests/master/resources/canvas-tests.js: Use new helper methods.
3324         * Animometer/tests/master/resources/dom-particles.js: Ditto.
3325         * Animometer/tests/master/resources/particles.js: Ditto.
3326         * Animometer/tests/simple/resources/simple-canvas-paths.js: Refactor to
3327         use a rotating color instead of a random color. The fast switching of color
3328         is too vivid to watch.
3329
3330         * Animometer/tests/bouncing-particles/resources/bouncing-svg-shapes.js:
3331         (BouncingSvgParticlesStage.call.createGradient): Fix the gradient so
3332         that the last stop is located at the end.
3333
3334 2016-02-07  Jon Lee  <jonlee@apple.com>
3335
3336         Refactor tune() to not return the complexity of the scene.
3337
3338         We have stage.complexity() now, so returning the complexity through tune
3339         is unnecessary.
3340
3341         * Animometer/tests/bouncing-particles/resources/bouncing-particles.js:
3342         * Animometer/tests/master/resources/canvas-stage.js:
3343         * Animometer/tests/master/resources/particles.js:
3344         * Animometer/tests/misc/resources/canvas-electrons.js:
3345         * Animometer/tests/misc/resources/canvas-stars.js:
3346         * Animometer/tests/resources/main.js:
3347         * Animometer/tests/simple/resources/simple-canvas.js:
3348         * Animometer/tests/simple/resources/tiled-canvas-image.js:
3349         * Animometer/tests/template/resources/template-canvas.js:
3350         * Animometer/tests/template/resources/template-css.js:
3351         * Animometer/tests/template/resources/template-svg.js:
3352         * Animometer/tests/text/resources/layering-text.js:
3353
3354 2016-02-07  Jon Lee  <jonlee@apple.com>
3355
3356         Make the fixed controller a step controller instead. Halfway through the test
3357         it will bump up the complexity 4-fold. Calculate the step timestamp using options
3358         instead of a separate parameter to the Controller constructor.
3359
3360         * Animometer/developer.html: Change value to "step"
3361         * Animometer/resources/debug-runner/animometer.js:
3362         (window.suitesManager.updateEditsElementsState): Show number inputs when set to "step".
3363         * Animometer/tests/resources/main.js:
3364         (update): Provide a hook for subclasses to tune.
3365         (StepController): Maintain a flag determining whether we've stepped, and the time
3366         we should step.
3367         (Benchmark): Use the new StepController.
3368
3369 2016-02-07  Jon Lee  <jonlee@apple.com>
3370
3371         Adjust the FPS graph scale.
3372
3373         Instead of making the FPS graph linearly scale, scale it based on the frame length,
3374         but show the data in terms of FPS. Because it is inversely proportional, and most
3375         of the data never gets below 20, concentrate the axis from 20-60 FPS, since otherwise
3376         over half of the available graph space ends up blank.
3377
3378         This means we should convert all of the FPS data to frame length data.
3379
3380         * Animometer/resources/debug-runner/graph.js: Update the domain to be based on
3381         frame length in milliseconds instead of FPS. Update the cursor to consider all of the
3382         values being shown, and then pick the min and max values to represent the length of the
3383         cursor.
3384         * Animometer/resources/runner/animometer.js:
3385         * Animometer/resources/strings.js:
3386         * Animometer/tests/resources/main.js:
3387         (processSamples): Add the ability to only sample a range of the data instead of everything
3388         after an offset index. Update sampler to record the frame lengths instead of the frame
3389         rate.
3390
3391 2016-02-07  Jon Lee  <jonlee@apple.com>
3392
3393         Add option to use different methods for retrieving a timestamp.
3394
3395         * Animometer/developer.html: Add performance.now and Date.now options.
3396         * Animometer/resources/runner/animometer.js: Default to performance.now.
3397         (window.benchmarkController.startBenchmark):
3398         * Animometer/tests/resources/main.js: Tie the desired method to _getTimestamp.
3399         (run): Use _getTimestamp.
3400         (_animateLoop): Ditto.
3401
3402 2016-02-07  Jon Lee  <jonlee@apple.com>
3403
3404         Allow adding any number of markers to the graph. The markers can be labeled
3405         and contain timestamp and sample index data. Make it a part of the controller
3406         rather than keeping it in the sampler.
3407
3408         * Animometer/resources/debug-runner/animometer.css: Add styles for markers
3409         * Animometer/resources/debug-runner/graph.js: Create the markers and add
3410         text labels.
3411         * Animometer/resources/runner/animometer.js: Assume the samplingTimeOffset
3412         is just one of the marks provided.
3413         * Animometer/resources/strings.js: Add Strings.json.marks.
3414         * Animometer/tests/resources/main.js:
3415         (Controller): Keep marks here. They are keyed by the marker name, so no two
3416         markers should have the same name.
3417         (recordFirstSample): Refactor to use mark.
3418         (mark): Allows for arbitrary data if needed later. The timestamp maintained
3419         is relative to the absolute start timestamp.
3420         (containsMark): Checks whether a mark with a specific comment exists.
3421         (processSamples): Removes the _startTimestamp offset from the marks before
3422         setting it in results.
3423         * Animometer/tests/resources/sampler.js: Remove marks.
3424
3425 2016-02-07  Jon Lee  <jonlee@apple.com>
3426
3427         Get rid of options member variable in Benchmark.
3428
3429         Options are only needed when initializing the stage or benchmark, so there's no
3430         need to also keep a reference to it.
3431
3432         * Animometer/tests/resources/main.js: Get rid of options variable in Benchmark.
3433         Pass options to Controllers and Stages.
3434         (Controller.Utilities.createClass):
3435         (Benchmark.Utilities.createClass):
3436         (get options): Deleted.
3437
3438         * Animometer/tests/bouncing-particles/resources/bouncing-canvas-images.js:
3439         * Animometer/tests/bouncing-particles/resources/bouncing-canvas-particles.js:
3440         * Animometer/tests/bouncing-particles/resources/bouncing-canvas-shapes.js:
3441         * Animometer/tests/bouncing-particles/resources/bouncing-css-images.js:
3442         * Animometer/tests/bouncing-particles/resources/bouncing-css-shapes.js:
3443         * Animometer/tests/bouncing-particles/resources/bouncing-particles.js:
3444         * Animometer/tests/bouncing-particles/resources/bouncing-svg-images.js:
3445         * Animometer/tests/bouncing-particles/resources/bouncing-svg-shapes.js:
3446         * Animometer/tests/master/resources/canvas-stage.js:
3447         * Animometer/tests/master/resources/canvas-tests.js:
3448         * Animometer/tests/master/resources/particles.js:
3449         * Animometer/tests/misc/resources/canvas-electrons.js:
3450         * Animometer/tests/misc/resources/canvas-stars.js:
3451         * Animometer/tests/misc/resources/compositing-transforms.js:
3452         * Animometer/tests/simple/resources/simple-canvas-paths.js:
3453         * Animometer/tests/simple/resources/tiled-canvas-image.js:
3454         * Animometer/tests/template/resources/template-canvas.js:
3455         * Animometer/tests/template/resources/template-css.js:
3456         * Animometer/tests/template/resources/template-svg.js:
3457         * Animometer/tests/text/resources/layering-text.js:
3458
3459 2016-02-07  Jon Lee  <jonlee@apple.com>
3460
3461         Update how the benchmark is run
3462         https://bugs.webkit.org/show_bug.cgi?id=153960
3463
3464         Provisionally reviewed by Said Abou-Hallawa.
3465
3466         Introduce the notion of a Controller. It is responsible for recording, updating,
3467         and processing the statistics and complexity of the benchmark. This allows
3468         plugging in different Controllers.
3469
3470         This strips most of the functionality from Animator and BenchmarkState, so fold
3471         what's left into Benchmark. Now, Benchmarks only own a stage and a controller, but
3472         are responsible for driving the animation loop.
3473
3474         Rewrite Animator._shouldRequestAnotherFrame into two different Controllers. One
3475         maintains a fixed complexity, and the other adapts the complexity to meet a
3476         fixed FPS.
3477
3478         Fix the Kalman estimator to be modeled on a scalar variable with no model.
3479
3480         * Animometer/tests/resources/main.js: Remove BenchmarkState and Animator, and
3481         replace it with a Controller. Add a FixedController and refactor the previous controller
3482         to an AdaptiveController.
3483
3484         (Controller): Controllers own the estimator and the sampler. When a new frame is
3485         displayed, the animation loop calls update(). The estimator and sampler record
3486         stats, then tune. Samplers can track multiple series of data. The basic controller
3487         tracks timestamp, complexity, and estimated frame rate.
3488                 The Kalman estimation is based on the frame length rather than the frame
3489         rate. Because FPS is inverse proportional to frame length, in the case where the measured
3490         frame length is very small, the FPS ends up being a wildly large number (in the order of
3491         600-1000 "FPS"), and it pulls the estimator up drastically enough that it takes a while
3492         for it to settle back down. Using frame length reduces the impact of these spikes.
3493                 Converging the estimation takes enough time to avoid initializing it immediately
3494         when the benchmark starts. Instead, the benchmark runs for a brief period of time (100ms)
3495         before running it in earnest. Allow controllers an opportunity to set the complexity
3496         before starting recording.
3497                 When the benchmark is complete, the controller has an opportunity to process
3498         the samples. The default implementation calculates the raw FPS based on the time
3499         difference of the samples, and calculates the complexity score. This is moved from
3500         Benchmark.processSamples.
3501
3502         (Controller): Initialize timestamps. These are at first relative to the start of the
3503         benchmark, but are offset by the absolute start time during start(). By default maintain
3504         3 data series, but subclasses can override.
3505         (start): Calls recordFirstSample() for subclasses to override if needed.
3506         (recordFirstSample): For basic controller, start sampling at the beginning.
3507         (update): Update the frame length estimator and sample.
3508         (shouldStop): Checks that the time is before _endTimestamp.
3509         (results): Returns the processed samples.
3510         (processSamples): Iterate through the sample data and collate them. Include scores.
3511
3512         (FixedComplexityController): Controller that tunes the stage to the desired complexity
3513         prior to starting, and keeps it at that complexity.
3514
3515         (AdaptiveController): Have the estimator estimate the interval frame rate instead of the
3516         raw frame rate.
3517                 The previous version of this controller ignored the frame that came after the
3518         adjustment. The raw FPS show that whatever noise the scene change adds is negligible
3519         compared to the noise of the system overall. Stop ignoring that frame and include all
3520         frames in the measurements.
3521
3522         (Benchmark): Remove dependency on animator, and instantiate a runner based on what is
3523         selected. Most of the loop's functionality is in Controller, so remove here.
3524         (Benchmark.run): Remove start() since it is only called from run(), and fold it in here.
3525         (Benchmark._animateLoop): Fold in from Animator.animateLoop. Let the benchmark run for
3526         a brief period before calling Controller.start().
3527
3528         * Animometer/tests/resources/math.js: Fix the Kalman estimator. The filter estimates
3529         a scalar variable, and makes basic assumptions regarding the model. As a result
3530         none of the linear algebra classes are needed, so remove Matrix, Vector3, and Matrix3.
3531         (SimpleKalmanEstimator): Calculate the gain based on the provided process and
3532         measurement errors.
3533         (KalmanEstimator): Deleted.
3534         (IdentityEstimator): Deleted.
3535         (PIDController): Refactor to use the Utilities.createClass() helper.
3536
3537         The Kalman filter algorithm is explained here http://greg.czerniak.info/guides/kalman1/.
3538         The state, represented by a scalar, is the estimated frame length. There is no user
3539         transition of the state, and the state is the same as the measurement. With this model,
3540         the estimation error converges, so calculate the gain ahead of time.
3541
3542         * Animometer/developer.html: Remove fixed-after-warmup since it is not useful.
3543         Replace the option to toggle the estimator, and make it possible to customize the
3544         estimator's error parameters. Show raw FPS by default, and remove interval FPS,
3545         which will be shown instead of the filtered raw FPS.
3546         * Animometer/resources/debug-runner/animometer.css: Put the header behind the graph.
3547         Remove #intervalFPS rules; move the color to #filteredFPS.
3548         * Animometer/resources/debug-runner/graph.js:
3549         (updateGraphData): Update the hr style to force the layout to be calculated
3550         correctly. Change the tick format to be in terms of seconds, since the timestamps
3551         are in milliseconds. Remove interval data.
3552         * Animometer/resources/runner/animometer.js:
3553         (window.benchmarkController.startBenchmark): Set Kalman parameters.
3554         * Animometer/resources/runner/benchmark-runner.js:
3555         (_runBenchmarkAndRecordResults): When a benchmark completes, expect it to return
3556         the final data, rather than passing a sampler from the controller. This avoids
3557         needing to expose the sampler variable in the benchmark.
3558         * Animometer/tests/resources/sampler.js:
3559         (process): Move the setting of the target frame rate to AdaptiveController.
3560
3561 2016-02-06  Jon Lee  <jonlee@apple.com>
3562
3563         Code clean up: Move Rotater function closer to Stage static methods.
3564         The Rotater is used together with those methods; keep them close.
3565
3566         * Animometer/tests/resources/main.js:
3567
3568 2016-02-06  Jon Lee  <jonlee@apple.com>
3569
3570         Update the JS includes due to ResultsTable move.
3571
3572         * Animometer/developer.html:
3573         * Animometer/index.html:
3574
3575 2016-02-06  Jon Lee  <jonlee@apple.com>
3576
3577         Move createElement and createSVGElement to Utilities.
3578
3579         * Animometer/resources/extensions.js:
3580         (Utilities.createElement): Added.
3581         (Utilities.createSVGElement): Added.
3582         (DocumentExtension.createElement): Deleted.
3583         (DocumentExtension.createSvgElement): Deleted.
3584
3585         * Animometer/resources/debug-runner/animometer.js:
3586         * Animometer/resources/runner/animometer.js:
3587         * Animometer/tests/bouncing-particles/resources/bouncing-svg-images.js:
3588         * Animometer/tests/bouncing-particles/resources/bouncing-svg-particles.js:
3589         * Animometer/tests/bouncing-particles/resources/bouncing-svg-shapes.js:
3590
3591 2016-02-06  Jon Lee  <jonlee@apple.com>
3592
3593         Add a convenience function for creating a class.
3594
3595         The pattern for creating a class is common enough to add as a Utilities
3596         helper function. It also makes it easy to collapse class definitions when
3597         editing.
3598
3599         * Animometer/resources/debug-runner/animometer.js: Move ProgressBar definition,
3600         since it is only used here.
3601         * Animometer/resources/runner/animometer.js: Move ResultsDashboard and
3602         ResultsTable definition, since it is only used here.
3603         * Animometer/resources/extensions.js: Move Utilities definition to the top. Convert
3604         Point, Insets, SimplePromise.
3605         (ProgressBar): Moved to animometer.js.
3606         (ResultsDashboard): Moved to animometer.js.
3607         (ResultsTable): Moved to animometer.js.
3608         * Animometer/resources/runner/benchmark-runner.js: Convert BenchmarkRunnerState,
3609         BenchmarkRunner.
3610         * Animometer/tests/resources/main.js: Convert Rotater, Stage, Animator, Benchmark.
3611         * Animometer/tests/resources/sampler.js: Convert Experiment, Sampler.
3612
3613         Convert test primitives.
3614         * Animometer/tests/master/resources/canvas-tests.js: Convert CanvasLineSegment,
3615         CanvasArc, CanvasLinePoint.
3616         * Animometer/tests/simple/resources/simple-canvas-paths.js: Convert CanvasLineSegment,
3617         CanvasLinePoint, CanvasQuadraticSegment, CanvasQuadraticPoint, CanvasBezierSegment,
3618         CanvasBezierPoint, CanvasArcToSegment, CanvasArcToSegmentFill, CanvasArcSegment,
3619         CanvasArcSegmentFill, CanvasRect, CanvasRectFill.
3620         * Animometer/tests/simple/resources/tiled-canvas-image.js: Convert CanvasImageTile.
3621
3622 2016-02-06  Jon Lee  <jonlee@apple.com>
3623
3624         Minor improvements to debug harness.
3625
3626         * Animometer/developer.html:
3627         * Animometer/resources/debug-runner/animometer.css:
3628         (#suites): Put the complexity text boxes closer to the test names.
3629         (#options):
3630         (#rawFPS circle): Make the interval FPS appear as a separate data series, with a line.
3631         (#intervalFPS path):
3632         (#intervalFPS circle):
3633         * Animometer/resources/debug-runner/animometer.js:
3634         (window.optionsManager.updateLocalStorageFromUI): Convert number inputs from text.
3635         (window.suitesManager._onChangeTestCheckbox): Refactor to take a checkbox.
3636         (window.suitesManager._createTestElement): Enhance such that typing into the complexity
3637         input will automatically select that test for running.
3638         (window.suitesManager.updateLocalStorageFromJSON): Make the harness work for private
3639         browsing.
3640         * Animometer/resources/debug-runner/graph.js: Separate the intervalFPS data, and show
3641         more accuracy in timestamps.
3642
3643 2016-02-06  Jon Lee  <jonlee@apple.com>
3644
3645         Refactor helper methods for getting random values for a stage.
3646
3647         Instead of requiring a Stage instance, just attach it to the Stage object.
3648
3649         * Animometer/tests/bouncing-particles/resources/bouncing-canvas-shapes.js:
3650         * Animometer/tests/bouncing-particles/resources/bouncing-css-shapes.js:
3651         * Animometer/tests/bouncing-particles/resources/bouncing-particles.js:
3652         * Animometer/tests/bouncing-particles/resources/bouncing-svg-shapes.js:
3653         * Animometer/tests/master/resources/canvas-tests.js:
3654         * Animometer/tests/master/resources/particles.js:
3655         * Animometer/tests/misc/resources/canvas-electrons.js:
3656         * Animometer/tests/misc/resources/canvas-stars.js:
3657         * Animometer/tests/misc/resources/compositing-transforms.js:
3658         * Animometer/tests/resources/main.js:
3659         * Animometer/tests/simple/resources/simple-canvas-paths.js:
3660
3661 2016-02-05  Said Abou-Hallawa  <sabouhallawa@apple.com>
3662
3663         Add a new graphics test for CanvasRenderingContext2D functions: getImageData and putImageData
3664         https://bugs.webkit.org/show_bug.cgi?id=151716
3665
3666         Reviewed by Darin Adler.
3667
3668         The purpose of this test is to measure the performance of getImageData
3669         and putImageData functions. This test draws a background on the canvas
3670         and then gets some random tiles from this background and draw them in
3671         destinations different from their original sources.
3672
3673         * Animometer/resources/debug-runner/tests.js: Adding the new test to the canvas simple tests suite.
3674
3675         * Animometer/resources/extensions.js:
3676         (Array.prototype.shuffle): Shuffles the elements of an array.
3677
3678         (Point.zero): Returns a new Point object whose x and y are equal zero.
3679         (Point.prototype.str): Used for debugging the Point object.
3680
3681         * Animometer/tests/simple/resources/tiled-canvas-image.js: Added.
3682         (CanvasImageTile):
3683         (CanvasImageTile.prototype.getImageData):
3684         (CanvasImageTile.prototype.putImageData):
3685         (Stage.call.initialize):
3686         (Stage.call._createTiles):
3687         (Stage.call._nextTilePosition):
3688         (Stage.call.tune):
3689         (Stage.call._drawBackground):
3690         (Stage.call.animate):
3691         (Stage.call.complexity):
3692         (Stage.call):
3693         * Animometer/tests/simple/tiled-canvas-image.html: Added.
3694
3695 2016-01-07  Jon Lee  <jonlee@apple.com>
3696
3697         Fix new test.
3698
3699         * Animometer/resources/runner/tests.js: Wrong URL from an
3700         old patch.
3701         * Animometer/tests/master/particles.html:
3702         * Animometer/tests/master/resources/dom-particles.js:
3703         (Particle.call.reset): Figured out a simpler way to set up
3704         the particles.
3705         (this.move.reset): Deleted.
3706         (this.move._applyAttributes): Deleted.
3707         * Animometer/tests/master/resources/particles.js:
3708         (Particle): Call move() after reset().
3709
3710 2016-01-07  Jon Lee  <jonlee@apple.com>
3711
3712         Update benchmark test suite
3713         https://bugs.webkit.org/show_bug.cgi?id=152679
3714
3715         Reviewed by Simon Fraser.
3716
3717         Add a new test. The test has a rotating background
3718         gradient, and does a better job physically simulating
3719         particles.
3720
3721         * Animometer/resources/extensions.js: Teach Point to take constants as well as other Points.
3722         (Point.prototype.length): Added.
3723         (Point.prototype.normalize): Added.
3724         * Animometer/resources/runner/tests.js: Add the test to the master suite.
3725         * Animometer/tests/master/particles.html: Added.
3726         * Animometer/tests/master/resources/particles.js: Added. Parent class for different kinds of particles.
3727         (Particle):
3728         (Particle.prototype.reset): If the particle starts slowing down in terms of its animation,
3729         reset it.
3730         (Particle.prototype.animate): Bounce off the walls elastically, and include gravity.
3731         (Particle.prototype.move): Subclasses should override.
3732         (ParticlesStage): Stage includes a rotating gradient background.
3733         * Animometer/tests/master/resources/dom-particles.js: Added. Creates a <div> and adds it to
3734         the stage.
3735         * Animometer/tests/resources/star.svg: Added.
3736
3737 2016-01-03  Jon Lee  <jonlee@apple.com>
3738
3739         Update benchmark test suite
3740         https://bugs.webkit.org/show_bug.cgi?id=152679
3741
3742         Reviewed by Simon Fraser.
3743
3744         Move algorithm.js and sampler.js to tests/ and benchmark-runner.js to runner/.
3745
3746         Needed by both harnesses.
3747         * Animometer/resources/runner/benchmark-runner.js: Renamed from PerformanceTests/Animometer/resources/debug-runner/benchmark-runner.js.
3748         * Animometer/developer.html:
3749         * Animometer/index.html:
3750
3751         Needed only by the tests. Move to tests/. Statistics, in sampler.js, is used by ResultsDashboard, so move that
3752         into extensions.js.
3753         * Animometer/resources/extensions.js:
3754         * Animometer/tests/resources/algorithm.js: Renamed from PerformanceTests/Animometer/resources/algorithm.js.
3755         * Animometer/tests/resources/sampler.js: Renamed from PerformanceTests/Animometer/resources/sampler.js.
3756         * Animometer/tests/bouncing-particles/bouncing-canvas-images.html:
3757         * Animometer/tests/bouncing-particles/bouncing-canvas-shapes.html:
3758         * Animometer/tests/bouncing-particles/bouncing-css-images.html:
3759         * Animometer/tests/bouncing-particles/bouncing-css-shapes.html:
3760         * Animometer/tests/bouncing-particles/bouncing-svg-images.html:
3761         * Animometer/tests/bouncing-particles/bouncing-svg-shapes.html:
3762         * Animometer/tests/master/canvas-stage.html:
3763         * Animometer/tests/misc/canvas-electrons.html:
3764         * Animometer/tests/misc/canvas-stars.html:
3765         * Animometer/tests/misc/compositing-transforms.html:
3766         * Animometer/tests/simple/simple-canvas-paths.html:
3767         * Animometer/tests/template/template-canvas.html:
3768         * Animometer/tests/template/template-css.html:
3769         * Animometer/tests/template/template-svg.html:
3770         * Animometer/tests/text/layering-text.html:
3771         * Animometer/tests/text/text-boxes.html:
3772
3773 2016-01-07  Jon Lee  <jonlee@apple.com>
3774
3775         Update benchmark test suite
3776         https://bugs.webkit.org/show_bug.cgi?id=152679
3777
3778         Reviewed by Simon Fraser.
3779
3780         Fix tests for other browsers.
3781
3782         * Animometer/resources/extensions.js:
3783         (Point.elementClientSize): Some browsers return 0 for SVG clientWidth and clientHeight.
3784         Use getBoundingClientRect() instead.
3785         * Animometer/tests/misc/resources/canvas-electrons.js:
3786         (CanvasElectron.prototype._draw): Some browsers don't support ellipse.
3787
3788 2016-01-07  Jon Lee  <jonlee@apple.com>
3789
3790         Add a waitUntilReady() step
3791         https://bugs.webkit.org/show_bug.cgi?id=152862
3792
3793         Reviewed by Simon Fraser.
3794
3795         Add a waitUntilReady() callback that lets the benchmark complete
3796         its setup before running the benchmark.
3797
3798         * Animometer/tests/resources/main.js:
3799         (Benchmark.prototype.run): First call waitUntilReady, which returns
3800         a promise. When the promise resolves, run everything that was in this
3801         function.
3802         (Benchmark.prototype.waitUntilReady): Default implementation returns
3803         a resolved promise.
3804         (Benchmark.prototype.resolveWhenFinished): Deleted.
3805         * Animometer/tests/template/resources/template-canvas.js:
3806         (new.TemplateCanvasStage.waitUntilReady): Example on how to override.
3807
3808 2016-01-03  Jon Lee  <jonlee@apple.com>
3809
3810         Update data reporting and analysis
3811         https://bugs.webkit.org/show_bug.cgi?id=152670
3812
3813         Reviewed by Simon Fraser.
3814
3815         Show new graph data. Provide controls to show different series data. Provide an
3816         interactive cursor that shows the data at a given sample.
3817
3818         * Animometer/developer.html: Add a nav section in #results. Each part of the graph
3819         has a checkbox for visual toggling, as well as companion spans to contain the data.
3820                 The numbers will always be shown even if the SVG isn't.