745b3b40ce25fe6370397d740c5c783d5a2a05b8
[WebKit-https.git] / Tools / ChangeLog
1 2013-01-08  Zan Dobersek  <zandobersek@gmail.com>
2
3         [EFL][GTK] Make the PulseAudioSanitizer an object on the EflPort, GtkPort
4         https://bugs.webkit.org/show_bug.cgi?id=106354
5
6         Reviewed by Eric Seidel.
7
8         Put the PulseAudioSanitizer object on the EflPort and GtkPort interfaces
9         instead of those two inheriting from it. Also add a mock object of the
10         sanitizer that's used in unit tests.
11
12         * Scripts/webkitpy/layout_tests/port/efl.py:
13         (EflPort):
14         (EflPort.__init__):
15         (EflPort.setup_test_run):
16         (EflPort.clean_up_test_run):
17         * Scripts/webkitpy/layout_tests/port/efl_unittest.py: Also correct the
18         importing order.
19         (EflPortTest.make_port): Put a mock PulseAudioSanitizer on the instance.
20         * Scripts/webkitpy/layout_tests/port/gtk.py:
21         (GtkPort):
22         (GtkPort.__init__):
23         (GtkPort.setup_test_run):
24         (GtkPort.clean_up_test_run):
25         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py: Also correct the
26         importing order.
27         (GtkPortTest.make_port): Put a mock PulseAudioSanitizer on the instance.
28         * Scripts/webkitpy/layout_tests/port/pulseaudio_sanitizer.py:
29         (PulseAudioSanitizer.unload_pulseaudio_module): Stylize the method as public.
30         (PulseAudioSanitizer.restore_pulseaudio_module): Ditto.
31         * Scripts/webkitpy/layout_tests/port/pulseaudio_sanitizer_mock.py: Added.
32         (PulseAudioSanitizerMock): A simple mock interface for PulseAudioSanitizer.
33         (PulseAudioSanitizerMock.unload_pulseaudio_module):
34         (PulseAudioSanitizerMock.restore_pulseaudio_module):
35
36 2013-01-08  Jochen Eisinger  <jochen@chromium.org>
37
38         [chromium] move dumpFrameLoadCallbacks and friends to the TestRunner library
39         https://bugs.webkit.org/show_bug.cgi?id=106324
40
41         Reviewed by Adam Barth.
42
43         Since the frame load callbacks are part of the WebFrameClient, the
44         WebTestProxy is now intercepting both WebViewClient and WebFrameClient
45         methods.
46
47         * DumpRenderTree/chromium/DRTTestRunner.cpp:
48         (DRTTestRunner::DRTTestRunner):
49         (DRTTestRunner::reset):
50         * DumpRenderTree/chromium/DRTTestRunner.h:
51         (DRTTestRunner):
52         * DumpRenderTree/chromium/TestRunner/public/WebTestInterfaces.h:
53         (WebTestInterfaces):
54         * DumpRenderTree/chromium/TestRunner/public/WebTestProxy.h:
55         (WebKit):
56         (WebTestProxyBase):
57         (WebTestRunner):
58         (WebTestRunner::WebTestProxy::WebTestProxy):
59         (WebTestProxy):
60         (WebTestRunner::WebTestProxy::didInvalidateRect):
61         (WebTestRunner::WebTestProxy::didScrollRect):
62         (WebTestRunner::WebTestProxy::scheduleComposite):
63         (WebTestRunner::WebTestProxy::scheduleAnimation):
64         (WebTestRunner::WebTestProxy::setWindowRect):
65         (WebTestRunner::WebTestProxy::show):
66         (WebTestRunner::WebTestProxy::didAutoResize):
67         (WebTestRunner::WebTestProxy::postAccessibilityNotification):
68         (WebTestRunner::WebTestProxy::startDragging):
69         (WebTestRunner::WebTestProxy::shouldBeginEditing):
70         (WebTestRunner::WebTestProxy::shouldEndEditing):
71         (WebTestRunner::WebTestProxy::shouldInsertNode):
72         (WebTestRunner::WebTestProxy::shouldInsertText):
73         (WebTestRunner::WebTestProxy::shouldChangeSelectedRange):
74         (WebTestRunner::WebTestProxy::shouldDeleteRange):
75         (WebTestRunner::WebTestProxy::shouldApplyStyle):
76         (WebTestRunner::WebTestProxy::didBeginEditing):
77         (WebTestRunner::WebTestProxy::didChangeSelection):
78         (WebTestRunner::WebTestProxy::didChangeContents):
79         (WebTestRunner::WebTestProxy::didEndEditing):
80         (WebTestRunner::WebTestProxy::registerIntentService):
81         (WebTestRunner::WebTestProxy::dispatchIntent):
82         (WebTestRunner::WebTestProxy::willPerformClientRedirect):
83         (WebTestRunner::WebTestProxy::didCancelClientRedirect):
84         (WebTestRunner::WebTestProxy::didStartProvisionalLoad):
85         (WebTestRunner::WebTestProxy::didReceiveServerRedirectForProvisionalLoad):
86         (WebTestRunner::WebTestProxy::didFailProvisionalLoad):
87         (WebTestRunner::WebTestProxy::didCommitProvisionalLoad):
88         (WebTestRunner::WebTestProxy::didReceiveTitle):
89         (WebTestRunner::WebTestProxy::didFinishDocumentLoad):
90         (WebTestRunner::WebTestProxy::didHandleOnloadEvents):
91         (WebTestRunner::WebTestProxy::didFailLoad):
92         (WebTestRunner::WebTestProxy::didFinishLoad):
93         (WebTestRunner::WebTestProxy::didChangeLocationWithinPage):
94         (WebTestRunner::WebTestProxy::didDisplayInsecureContent):
95         (WebTestRunner::WebTestProxy::didRunInsecureContent):
96         (WebTestRunner::WebTestProxy::didDetectXSS):
97         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h:
98         (WebTestRunner::WebTestRunner::setTestIsRunning):
99         (WebTestRunner::WebTestRunner::shouldDumpFrameLoadCallbacks):
100         (WebTestRunner::WebTestRunner::setShouldDumpFrameLoadCallbacks):
101         (WebTestRunner::WebTestRunner::shouldDumpUserGestureInFrameLoadCallbacks):
102         (WebTestRunner::WebTestRunner::stopProvisionalFrameLoads):
103         (WebTestRunner::WebTestRunner::shouldDumpTitleChanges):
104         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
105         (WebTestRunner::TestRunner::TestRunner):
106         (WebTestRunner::TestRunner::reset):
107         (WebTestRunner::TestRunner::setTestIsRunning):
108         (WebTestRunner):
109         (WebTestRunner::TestRunner::shouldDumpFrameLoadCallbacks):
110         (WebTestRunner::TestRunner::setShouldDumpFrameLoadCallbacks):
111         (WebTestRunner::TestRunner::shouldDumpUserGestureInFrameLoadCallbacks):
112         (WebTestRunner::TestRunner::stopProvisionalFrameLoads):
113         (WebTestRunner::TestRunner::shouldDumpTitleChanges):
114         (WebTestRunner::TestRunner::dumpFrameLoadCallbacks):
115         (WebTestRunner::TestRunner::dumpUserGestureInFrameLoadCallbacks):
116         (WebTestRunner::TestRunner::setStopProvisionalFrameLoads):
117         (WebTestRunner::TestRunner::dumpTitleChanges):
118         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
119         (TestRunner):
120         * DumpRenderTree/chromium/TestRunner/src/WebTestInterfaces.cpp:
121         (WebTestInterfaces::Internal):
122         (WebTestRunner::WebTestInterfaces::Internal::webView):
123         (WebTestRunner::WebTestInterfaces::Internal::Internal):
124         (WebTestRunner::WebTestInterfaces::Internal::setWebView):
125         (WebTestRunner):
126         (WebTestRunner::WebTestInterfaces::Internal::setTestIsRunning):
127         (WebTestRunner::WebTestInterfaces::setWebView):
128         (WebTestRunner::WebTestInterfaces::setTestIsRunning):
129         (WebTestRunner::WebTestInterfaces::webView):
130         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
131         (WebTestRunner::WebTestProxyBase::willPerformClientRedirect):
132         (WebTestRunner):
133         (WebTestRunner::WebTestProxyBase::didCancelClientRedirect):
134         (WebTestRunner::WebTestProxyBase::didStartProvisionalLoad):
135         (WebTestRunner::WebTestProxyBase::didReceiveServerRedirectForProvisionalLoad):
136         (WebTestRunner::WebTestProxyBase::didFailProvisionalLoad):
137         (WebTestRunner::WebTestProxyBase::didCommitProvisionalLoad):
138         (WebTestRunner::WebTestProxyBase::didReceiveTitle):
139         (WebTestRunner::WebTestProxyBase::didFinishDocumentLoad):
140         (WebTestRunner::WebTestProxyBase::didHandleOnloadEvents):
141         (WebTestRunner::WebTestProxyBase::didFailLoad):
142         (WebTestRunner::WebTestProxyBase::didFinishLoad):
143         (WebTestRunner::WebTestProxyBase::didChangeLocationWithinPage):
144         (WebTestRunner::WebTestProxyBase::didDisplayInsecureContent):
145         (WebTestRunner::WebTestProxyBase::didRunInsecureContent):
146         (WebTestRunner::WebTestProxyBase::didDetectXSS):
147         * DumpRenderTree/chromium/TestShell.cpp:
148         (TestShell::runFileTest):
149         (TestShell::testFinished):
150         * DumpRenderTree/chromium/TestShell.h:
151         (TestShell):
152         * DumpRenderTree/chromium/WebViewHost.cpp:
153         (WebViewHost::didStartProvisionalLoad):
154         (WebViewHost::didReceiveServerRedirectForProvisionalLoad):
155         (WebViewHost::didFailProvisionalLoad):
156         (WebViewHost::didCommitProvisionalLoad):
157         (WebViewHost::didReceiveTitle):
158         (WebViewHost::didFailLoad):
159         (WebViewHost::didFinishLoad):
160         * DumpRenderTree/chromium/WebViewHost.h:
161         (WebViewHost):
162
163 2013-01-08  Zan Dobersek  <zandobersek@gmail.com>
164
165         [GTK] Lower the default timeout value for WebKitTestRunner
166         https://bugs.webkit.org/show_bug.cgi?id=106353
167
168         Reviewed by Dirk Pranke.
169
170         The default timeout value for WebKitTestRunner should be lowered.
171         The current value (80 seconds) is irrational and just leads to producing
172         text failures for layout tests that genuinely time out. This just leads
173         to more deviation from the GTK WK1 port in terms of test expectations.
174
175         The value used is the same as when running the DumpRenderTree, that's
176         6 seconds in release builds and 12 seconds in debug builds.
177
178         * Scripts/webkitpy/layout_tests/port/gtk.py:
179         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
180         (GtkPortTest.test_default_timeout_ms):
181
182 2013-01-08  Zan Dobersek  <zandobersek@gmail.com>
183
184         test-webkitpy shows failure output on my linux box
185         https://bugs.webkit.org/show_bug.cgi?id=101261
186
187         Reviewed by Eric Seidel.
188
189         Mock out the PulseAudioSanitizer methods when creating testing instances
190         of EflPort and GtkPort interfaces in unit tests. This stops throwing up
191         PulseAudio output when the sanitizer was trying to unload modules which
192         were already unloaded by another unit test that was being run in parallel.
193
194         * Scripts/webkitpy/layout_tests/port/efl_unittest.py:
195         (EflPortTest):
196         (EflPortTest.make_port):
197         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
198         (GtkPortTest):
199         (GtkPortTest.make_port):
200
201 2013-01-08  Julie Parent  <jparent@chromium.org>
202
203         Dashboard Cleanup: Remove g_hasDoneInitialPageGeneration
204         https://bugs.webkit.org/show_bug.cgi?id=106254
205
206         Remove the global variable g_hasDoneInitialPageGeneration.  It was used
207         only to figure out if we should call addBuilderLoadErrors during
208         handleLocationChange, which was called when the loader completed and on
209         any hash change, but we only ever wanted to call addBuilderLoadErrors
210         on the initial loader load.  Moved the call of addBuilderLoadErrors to
211         resourceLoadingComplete rather than handleLocationChange, so now we
212         only call it when we want to, and the check can go away completely.
213
214         Reviewed by Dirk Pranke.
215
216         * TestResultServer/static-dashboards/dashboard_base.js:
217         (addBuilderLoadErrors):
218         (resourceLoadingComplete):
219         (handleLocationChange):
220         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
221
222 2013-01-08  Dominic Mazzoni  <dmazzoni@google.com>
223
224         Add aboxhall@chromium.org to accessibility watchlist.
225
226         Unreviewed.
227
228         * Scripts/webkitpy/common/config/watchlist:
229
230 2013-01-08  Dan Winship  <danw@gnome.org>
231
232         Add Apache 2.4 support for Fedora 18
233         https://bugs.webkit.org/show_bug.cgi?id=104478
234
235         Reviewed by Martin Robinson.
236
237         Apache 2.4's httpd.conf is slightly incompatible with 2.2's, so we
238         need a separate config.
239
240         * Scripts/webkitpy/layout_tests/port/base.py:
241         (Port._apache_version):
242         (Port._apache_config_file_name_for_platform):
243         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
244         * Scripts/run-webkit-httpd:
245         * Scripts/webkitperl/httpd.pm:
246         (getApacheVersion):
247         (getDefaultConfigForTestDirectory):
248         (getHTTPDConfigPathForTestDirectory):
249
250 2013-01-08  Zan Dobersek  <zandobersek@gmail.com>
251
252         Make the override-feature-defines script compatible with Python v3
253         https://bugs.webkit.org/show_bug.cgi?id=105742
254
255         Reviewed by Martin Robinson.
256
257         Instead of printing output through a statement, use the print function and format
258         the output using the str.format method. This makes the script runnable with Python v3.
259
260         * gtk/override-feature-defines:
261         (adjust_feature_defines_makefile):
262
263 2013-01-08  Heikki Paajanen  <heikki.paajanen@palm.com>
264
265         [Qt] Fix libxml2 config test
266         https://bugs.webkit.org/show_bug.cgi?id=106317
267
268         Reviewed by Simon Hausmann.
269
270         Don't assume standard include/libxml2, libxml-2.0.pc should know
271         where it's headers are.
272
273         * qmake/config.tests/libxml2/libxml2.cpp: Fix include
274
275 2013-01-08  Zan Dobersek  <zandobersek@gmail.com>
276
277         [GTK] Make Tools/gtk/generate-gtkdoc compatible with Python 3
278         https://bugs.webkit.org/show_bug.cgi?id=106195
279
280         Reviewed by Philippe Normand.
281
282         Perform changes in the generate-gtkdoc script and common and gtkdoc modules
283         to make the Python code compatible with Python 3. This includes modifying
284         print statements, exception handling, dictionary iteration and
285         byte-sequence-to-string conversion.
286
287         * gtk/common.py:
288         (get_build_path):
289         (pkg_config_file_variable):
290         (gtk_version_of_pkg_config_file):
291         * gtk/generate-gtkdoc:
292         (print_missing_api):
293         * gtk/gtkdoc.py:
294         (GTKDoc.__init__):
295         (GTKDoc._run_command):
296
297 2013-01-02  Steve Block  <steveblock@chromium.org>
298
299         Add chromium.org email address for Steve Block.
300
301         Unreviewed.
302
303         * Scripts/webkitpy/common/config/committers.py:
304
305 2013-01-07  Filip Pizlo  <fpizlo@apple.com>
306
307         Rationalize closure call heuristics and profiling
308         https://bugs.webkit.org/show_bug.cgi?id=106270
309
310         Reviewed by Oliver Hunt.
311         
312         Add ability to use display-profiler-output via a pipe, and add the ability to dump
313         all generated code ('display *' or 'd *').
314
315         * Scripts/display-profiler-output:
316
317 2013-01-07  Ryosuke Niwa  <rniwa@webkit.org>
318
319         Sorted more xcodeproj files.
320
321         * MiniBrowser/MiniBrowser.xcodeproj/project.pbxproj:
322         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
323         * WebKitLauncher/WebKitLauncher.xcodeproj/project.pbxproj:
324
325 2013-01-07  Ryosuke Niwa  <rniwa@webkit.org>
326
327         Sorted xcodeproj files.
328
329         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
330         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
331
332 2013-01-07  Dan Bernstein  <mitz@apple.com>
333
334         [mac] WebKit1 clients can’t tell when a frame has been removed from the hierarchy
335         https://bugs.webkit.org/show_bug.cgi?id=106261
336
337         Reviewed by Simon Fraser.
338
339         Added a test for -[WebFrameLoadDelegate webView:didRemoveFrameFromHierarchy:].
340
341         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
342         * TestWebKitAPI/Tests/mac/WebViewDidRemoveFrameFromHierarchy.mm: Added.
343         (-[DidRemoveFrameFromHierarchyFrameLoadDelegate webView:didFinishLoadForFrame:]):
344         (-[DidRemoveFrameFromHierarchyFrameLoadDelegate webView:didRemoveFrameFromHierarchy:]):
345         (TestWebKitAPI):
346         (TestWebKitAPI::TEST):
347
348 2013-01-07  Ojan Vafai  <ojan@chromium.org>
349
350         Add the Apple MountainLion bots to garden-o-matic.
351         Eventually, we really should move away from hardcoding
352         the list of bots.
353
354         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
355
356 2013-01-07  Jochen Eisinger  <jochen@chromium.org>
357
358         [chromium] move web audio related methods to testrunner library
359         https://bugs.webkit.org/show_bug.cgi?id=106211
360
361         Reviewed by Adam Barth.
362
363         * DumpRenderTree/chromium/DRTTestRunner.cpp:
364         (DRTTestRunner::DRTTestRunner):
365         (DRTTestRunner::reset):
366         * DumpRenderTree/chromium/DRTTestRunner.h:
367         (DRTTestRunner):
368         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h:
369         (WebKit):
370         (WebTestRunner::WebTestRunner::shouldDumpAsAudio):
371         (WebTestRunner::WebTestRunner::audioData):
372         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
373         (WebTestRunner::TestRunner::TestRunner):
374         (WebTestRunner::TestRunner::reset):
375         (WebTestRunner::TestRunner::shouldDumpAsAudio):
376         (WebTestRunner):
377         (WebTestRunner::TestRunner::audioData):
378         (WebTestRunner::TestRunner::setAudioData):
379         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
380         (TestRunner):
381         * DumpRenderTree/chromium/TestShell.cpp:
382         (TestShell::dump):
383
384 2013-01-07  Mike West  <mkwst@chromium.org>
385
386         Make the IFRAME_SEAMLESS flag runtime-enabled.
387         https://bugs.webkit.org/show_bug.cgi?id=106213
388
389         Reviewed by Ojan Vafai.
390
391         * DumpRenderTree/chromium/TestShell.cpp:
392         (TestShell::TestShell):
393         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
394         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
395         * DumpRenderTree/gtk/DumpRenderTree.cpp:
396         (resetDefaultsToConsistentValues):
397         * DumpRenderTree/mac/DumpRenderTree.mm:
398         (resetDefaultsToConsistentValues):
399         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
400         (WebCore::WebPage::resetSettings):
401         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
402         (WTR::InjectedBundle::beginTesting):
403             Ensure that seamless IFrames are enabled for layout tests in each of
404             the ports that support seamless IFrames.
405
406 2013-01-07  Jochen Eisinger  <jochen@chromium.org>
407
408         [chromium] move dumpAsText and friends to the TestRunner library
409         https://bugs.webkit.org/show_bug.cgi?id=106222
410
411         Reviewed by Adam Barth.
412
413         This moves dumpAsText, dumpChildFramesAsText, and dumpChildFrameScrollPositions to the TestRunner library.
414
415         * DumpRenderTree/chromium/DRTTestRunner.cpp:
416         (DRTTestRunner::DRTTestRunner):
417         (DRTTestRunner::reset):
418         * DumpRenderTree/chromium/DRTTestRunner.h:
419         (DRTTestRunner):
420         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h:
421         (WebTestRunner::WebTestRunner::shouldDumpAsText):
422         (WebTestRunner::WebTestRunner::setShouldDumpAsText):
423         (WebTestRunner::WebTestRunner::shouldGeneratePixelResults):
424         (WebTestRunner::WebTestRunner::setShouldGeneratePixelResults):
425         (WebTestRunner::WebTestRunner::shouldDumpChildFrameScrollPositions):
426         (WebTestRunner::WebTestRunner::shouldDumpChildFramesAsText):
427         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
428         (WebTestRunner::TestRunner::TestRunner):
429         (WebTestRunner::TestRunner::reset):
430         (WebTestRunner::TestRunner::shouldDumpAsText):
431         (WebTestRunner):
432         (WebTestRunner::TestRunner::setShouldDumpAsText):
433         (WebTestRunner::TestRunner::shouldGeneratePixelResults):
434         (WebTestRunner::TestRunner::setShouldGeneratePixelResults):
435         (WebTestRunner::TestRunner::shouldDumpChildFrameScrollPositions):
436         (WebTestRunner::TestRunner::shouldDumpChildFramesAsText):
437         (WebTestRunner::TestRunner::dumpAsText):
438         (WebTestRunner::TestRunner::dumpChildFrameScrollPositions):
439         (WebTestRunner::TestRunner::dumpChildFramesAsText):
440         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
441         (TestRunner):
442
443 2013-01-07  Ryosuke Niwa  <rniwa@webkit.org>
444
445         Build fix attempt after r138810.
446
447         Apparently python 2.6 is really bad at floating point computation.
448
449         * Scripts/webkitpy/performance_tests/perftestsrunner_integrationtest.py:
450         (MainTest.test_run_memory_test):
451
452 2013-01-07  Ryosuke Niwa  <rniwa@webkit.org>
453
454         Extract a class to represent a test result to simplify PerfTest.parse_output and PageLoadPerfTest._run_with_driver
455         https://bugs.webkit.org/show_bug.cgi?id=106172
456
457         Reviewed by Tony Chang.
458
459         Extracted PerfTestMetric, which represents a test result and knows how to compute statistics and convert itself
460         to a dictionary. A metric is a type of measurement each test makes such as "Time", "Malloc", and "JSHeap".
461         A single test (e.g. html5-full-render.html) may have multiple metrics, each of which is reported as a separate
462         test on perf-o-matic and perf. test's results page.
463
464         Also, use python's compute_statistics instead of reading statistics off of DriverOutput to simplify the code.
465
466         * Scripts/webkitpy/performance_tests/perftest.py:
467         (PerfTestMetric): Added.
468         (PerfTestMetric.__init__): Added.
469         (PerfTestMetric.metric): Added.
470         (PerfTestMetric.has_values): Added.
471         (PerfTestMetric.legacy_chromium_bot_compatible_test_name): Added. Following the legacy convention, "Malloc" and
472         "JSHeap" will use "test_name:Malloc" whereas "Time" uses "test_name" itself. We should eventually move these
473         conversions to be done on the client side (perf-o-matic and results page).
474         (PerfTestMetric.append): Added.
475         (PerfTestMetric.to_dict): Added. Computes the statistics and generates the dictionary object that represents
476         a result for this metric.
477         (PerfTestMetric.metric_to_unit): Added. Each metric maps to exactly one unit (it's not one-to-one).
478         (PerfTestMetric.compute_statistics): Moved from PerfTest.
479
480         (PerfTest._create_driver): Extracted from PerfTest.run.
481         (PerfTest.run): Given an array of PerfTestMetric, generate the results dictionary. Also print out results
482         as needed.
483         (PerfTest._run_with_driver):
484         (PerfTest): Removed _result_classes and _result_class_regex. "result class" is the old name for "metric".
485         (PerfTest.parse_output): Dramatically simplified. We now just return an array of PerfTestMetric's. Also ignore
486         all keys except "values". We rely on PerfTestMetric to compute the statistics for us.
487
488         (ChromiumStylePerfTest.run): Added and merged _run_with_driver. Since chromium style perf tests are so different
489         from other perf tests, there isn't much point in sharing the code here.
490
491         (PageLoadingPerfTest._run_with_driver): Simplified. Just create an array of PerfTestMetric's without computing
492         statistics for the measured values.
493
494         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
495         (TestPerfTestMetric): Added.
496         (TestPerfTestMetric.test_init_set_missing_unit): Added.
497         (TestPerfTestMetric.test_legacy_chromium_bot_compatible_test_name): Added.
498         (TestPerfTestMetric.test_has_values): Added.
499         (TestPerfTestMetric.test_append): Added.
500         (TestPerfTestMetric.test_compute_statistics.compute_statistics):
501         (TestPerfTestMetric.test_compute_statistics): Added a trivial test case.
502
503         (TestPerfTest._assert_results_are_correct): Updated as needed.
504         (TestPageLoadingPerfTest.test_run): Ditto. Notice that we no longer log anything in _run_with_driver. It's done
505         in PerfTest.run now after generating the results dictionary.
506         (TestPageLoadingPerfTest.test_run_with_memory_output): Ditto.
507
508 2013-01-07  Julien Chaffraix  <jchaffraix@webkit.org>
509
510         Support size_t multiplication and division operators on LayoutUnit
511         https://bugs.webkit.org/show_bug.cgi?id=83848
512
513         Reviewed by Emil A Eklund.
514
515         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp:
516         Added some simple tests for the new operators using size_t.
517
518 2013-01-07  Csaba Osztrogonác  <ossy@webkit.org>
519
520         [Qt] run-webkit-tests cleanup for parallel NRWT
521         https://bugs.webkit.org/show_bug.cgi?id=106220
522
523         Reviewed by Dirk Pranke.
524
525         * Scripts/run-webkit-tests:
526         (useNewRunWebKitTests):
527         * Scripts/webkitpy/layout_tests/port/qt.py:
528         (QtPort):
529         (QtPort.default_child_processes):
530
531 2013-01-07  Philippe Normand  <pnormand@igalia.com>
532
533         Create a GStreamerAudio watchlist and add myself to it.
534
535         Unreviewed.
536
537         * Scripts/webkitpy/common/config/watchlist:
538
539 2013-01-05  Ryosuke Niwa  <rniwa@webkit.org>
540
541         Add svn:executable to a bunch of scripts in BuildSlaveSupport.
542         Also removed the empty mac directory.
543
544         * BuildSlaveSupport/built-product-archive: Added property svn:executable.
545         * BuildSlaveSupport/download-built-product: Added property svn:executable.
546         * BuildSlaveSupport/mac: Removed.
547         * BuildSlaveSupport/test-result-archive: Added property svn:executable.
548
549 2013-01-04  Ryosuke Niwa  <rniwa@webkit.org>
550
551         Add a script to delete stale zero-byte build files
552         https://bugs.webkit.org/show_bug.cgi?id=106131
553
554         Reviewed by Tony Chang.
555
556         Delete zero-byte build files on Mac port for now. It doesn't do anything on other ports.
557
558         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
559         (DeleteStaleBuildFiles): Added.
560         (Factory.__init__): Add DeleteStaleBuildFiles step.
561         * BuildSlaveSupport/delete-stale-build-files: Added.
562         (main): Delete zero-byte files except ones intentionally generated by our build process.
563         (webkitBuildDirectory): Copied from build-product-archive.
564
565 2013-01-04  Michael Pruett  <michael@68k.org>
566
567         [GTK] Fix build error on GTK due to r138836
568         https://bugs.webkit.org/show_bug.cgi?id=106154
569
570         Reviewed by Kentaro Hara.
571
572         * GNUmakefile.am:
573
574 2013-01-04  Dan Bernstein  <mitz@apple.com>
575
576         [mac] WebKit1 has no equivalent of WKBundleNodeHandleCopyHTMLTableCellElementCellAbove
577         https://bugs.webkit.org/show_bug.cgi?id=106121
578
579         Reviewed by Adele Peterson.
580
581         Added test for -[DOMTHMLTableCellElement _cellAbove].
582
583         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
584         * TestWebKitAPI/Tests/mac/DOMHTMLTableCellCellAbove.mm: Added.
585         (-[HTMLTableCellElementCellAboveTest webView:didFinishLoadForFrame:]):
586         (TestWebKitAPI):
587         (TestWebKitAPI::expectCellAboveCell):
588         (TestWebKitAPI::TEST):
589         * TestWebKitAPI/Tests/mac/DOMHTMLTableCellElementCellAbove.html: Added.
590
591 2013-01-04  Eric Seidel  <eric@webkit.org>
592
593         FlakyTestReporter should be re-enabled and taught how to post patches
594         https://bugs.webkit.org/show_bug.cgi?id=103839
595
596         Reviewed by Adam Barth.
597
598         This moves LayoutTestResultsReader off of the DeprecatedPort object.
599         It also adds a real Port object to the EWS and CommitQueue objects
600         using a new PatchProcessingQueue superclass.
601         PatchProcessingQueue doesn't seem to be the cleanest insertion point
602         but works for this purpose.
603         I also moved _upload_results_archive_for_patch into PatchProcessingQueue
604         to resolve a FIXME.
605
606         This required some hacks/shiming between the non-specific "old"
607         port names which the EWS system is currently using, and the new
608         port objects.  This shimming is done in PatchProcessingQueue._new_port_name_from_old.
609
610         Hopefully this will fix the flaky test reporter and results uploading.
611         Certainly it's another small step towards getting rid of the old ports.py
612         infrastructure and making the NRWT Port class common webkitpy infrastructure.
613
614         Now that DeprecatedPort is no longer used for the results directory
615         I removed all the explicit results-dir code which webkit-patch was using
616         in --non-interactive mode to tell NRWT to use a results-dir it understood. :)
617
618         * Scripts/webkitpy/common/config/ports.py:
619         (DeprecatedPort.run_perl_unittests_command):
620         * Scripts/webkitpy/common/config/ports_mock.py:
621         (MockPort.name):
622         * Scripts/webkitpy/common/system/outputcapture.py:
623         (OutputCapture.assert_outputs):
624         * Scripts/webkitpy/layout_tests/port/apple.py:
625         (ApplePort.determine_full_port_name):
626         * Scripts/webkitpy/tool/bot/layouttestresultsreader.py:
627         (LayoutTestResultsReader.__init__):
628         (LayoutTestResultsReader):
629         (LayoutTestResultsReader._read_file_contents):
630         (LayoutTestResultsReader._create_layout_test_results):
631         (LayoutTestResultsReader._create_unit_test_results):
632         (LayoutTestResultsReader.results):
633         (LayoutTestResultsReader.archive):
634         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py:
635         (LayoutTestResultsReaderTest.test_missing_layout_test_results):
636         (LayoutTestResultsReaderTest.test_create_unit_test_results):
637         (test_missing_unit_test_results_path):
638         (test_layout_test_results):
639         (test_archive_last_layout_test_results):
640         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
641         (AbstractEarlyWarningSystem):
642         (AbstractEarlyWarningSystem.__init__):
643         (AbstractEarlyWarningSystem.begin_work_queue):
644         (AbstractEarlyWarningSystem.run_command):
645         * Scripts/webkitpy/tool/commands/queues.py:
646         (AbstractPatchQueue.work_item_log_path):
647         (PatchProcessingQueue):
648         (PatchProcessingQueue._new_port_name_from_old):
649         (PatchProcessingQueue.begin_work_queue):
650         (PatchProcessingQueue._upload_results_archive_for_patch):
651         (CommitQueue):
652         (CommitQueue.begin_work_queue):
653         (CommitQueue.run_command):
654         (AbstractReviewQueue):
655         (AbstractReviewQueue.__init__):
656         (AbstractReviewQueue.begin_work_queue):
657         * Scripts/webkitpy/tool/commands/queues_unittest.py:
658         (AbstractPatchQueueTest.test_next_patch):
659         (PatchProcessingQueueTest):
660         (PatchProcessingQueueTest.test_upload_results_archive_for_patch):
661         * Scripts/webkitpy/tool/steps/runtests.py:
662         (RunTests.run):
663         * Scripts/webkitpy/tool/steps/runtests_unittest.py:
664
665 2013-01-04  Dana Jansens  <danakj@chromium.org>
666
667         [chromium] Remove lost-compositor-context layout tests
668         https://bugs.webkit.org/show_bug.cgi?id=106089
669
670         Reviewed by James Robinson.
671
672         Remove the loseCompositorContext() test function.
673
674         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
675         (WebTestRunner::TestRunner::TestRunner):
676         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
677         (TestRunner):
678
679 2013-01-04  Adam Klein  <adamk@chromium.org>
680
681         Remove ENABLE_MUTATION_OBSERVERS #define
682         https://bugs.webkit.org/show_bug.cgi?id=105459
683
684         Reviewed by Ryosuke Niwa.
685
686         * Scripts/webkitperl/FeatureList.pm:
687         * qmake/mkspecs/features/features.pri:
688
689 2013-01-04  Ryosuke Niwa  <rniwa@webkit.org>
690
691         Statistics used in perftest_unittest.py and perftest_integrationtest.py are bogus
692         https://bugs.webkit.org/show_bug.cgi?id=106078
693
694         Reviewed by Tony Chang.
695
696         This patch updates values for individual iterations in dummy performance tests used in
697         perftest_unittest.py and perftest_integrationtest.py so that they match up with
698         statistics such as stdev and median reported in those tests. This change is required
699         to resolve the bug 97510.
700
701         This patch also refactors integration tests by extracting *TestData classes that
702         centralizes the definitions of each dummy performance test's text (DRT output),
703         runner output, and parsed results (JSON) so that we may share more data between
704         python tests.
705
706         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
707         (MainTest._assert_results_are_correct): Extracted from test_parse_output.
708         Use self.assertAlmostEqual when comparing the standard deviation.
709         (MainTest.test_parse_output):
710         (test_parse_output_with_failing_line):
711         (test_parse_output_with_description): Use _assert_results_are_correct.
712         (test_parse_output_with_subtests): Ditto.
713
714         * Scripts/webkitpy/performance_tests/perftestsrunner_integrationtest.py:
715         (InspectorPassTestData): Extracted from tests below.
716         (EventTargetWrapperTestData): Ditto.
717         (SomeParserTestData): Ditto.
718         (MemoryTestData): Ditto.
719         (TestDriver.run_test): Moved test definitions out of the method.
720         (MainTest._normalize_output): Truncate standard deviations at 5th decimal point.
721         (MainTest._load_output_json): Extracted from create_runner.
722         (MainTest.create_runner):
723         (MainTest.test_run_test_set_for_parser_tests):
724         (MainTest.test_run_memory_test):
725         (MainTest._test_run_with_json_output):
726         (MainTest.test_run_with_json_output):
727         (MainTest.test_run_with_description):
728         (MainTest.test_run_generates_json_by_default):
729         (MainTest.test_run_merges_output_by_default):
730         (MainTest.test_run_respects_reset_results):
731         (MainTest.test_run_generates_and_show_results_page):
732         (MainTest.test_run_with_slave_config_json):
733         (MainTest.test_run_with_multiple_repositories):
734
735 2013-01-04  Martin Robinson  <mrobinson@igalia.com>
736
737         [GTK] [WebKit2] Clicking on a word that is selected in an editable area deletes it
738         https://bugs.webkit.org/show_bug.cgi?id=106045
739
740         Reviewed by Carlos Garcia Campos.
741
742         Update test results to confirm that compositions without preedits do not trigger
743         a composition confirmation.
744
745         * TestWebKitAPI/Tests/gtk/InputMethodFilter.cpp:
746         (TestWebKitAPI::verifyCanceledComposition):
747
748 2013-01-04  Sailesh Agrawal  <sail@chromium.org>
749
750         [Mac] [WK2] platform/mac/fast/events/numpad-keycode-mapping.html fails
751         https://bugs.webkit.org/show_bug.cgi?id=105958
752
753         Reviewed by Alexey Proskuryakov.
754
755         Ported code from DumpRenderTree to WebKitTestRunner to map all numpad keys.
756
757         * WebKitTestRunner/mac/EventSenderProxy.mm:
758         (KeyMappingEntry):
759         (WTR):
760         (WTR::EventSenderProxy::keyDown): Added all numpad keys.
761
762 2013-01-03  Chris Fleizach  <cfleizach@apple.com>
763
764         REGRESSION: [Mac] Intermittent crash in WTR::AccessibilityUIElement::isEqual
765         https://bugs.webkit.org/show_bug.cgi?id=106073
766
767         Reviewed by Ryosuke Niwa.
768
769         Check that the element being compared to is not nil. 
770         This crash is happening because one of the accessibility tests does an isEqual comparison within a callback function.
771         It looks like occasionally, one of the elements passed in to the isEqual is nil because the notification handler is on the accessibility
772         controller, and the notification is being delivered after the document objects are cleared. This shouldn't happen but I've seen it 
773         sometimes before on parallel WebKitTestRunner runs.
774
775         * DumpRenderTree/AccessibilityUIElement.cpp:
776         (AccessibilityUIElement::isEqual):
777         * WebKitTestRunner/InjectedBundle/mac/AccessibilityUIElementMac.mm:
778         (WTR::AccessibilityUIElement::isEqual):
779
780 2013-01-03  Tim 'mithro' Ansell  <mithro@mithis.com>
781
782         Downloads can be flaky, so try downloading multiple times.
783         https://bugs.webkit.org/show_bug.cgi?id=105976
784
785         Reviewed by Eric Seidel.
786
787         * Scripts/webkitpy/common/system/autoinstall.py:
788         (AutoInstaller._download_to_stream):
789
790 2013-01-03  Eric Seidel  <eric@webkit.org>
791
792         Rename Tool.port to deprecated_port in preparation for removing more callers
793         https://bugs.webkit.org/show_bug.cgi?id=106076
794
795         Reviewed by Ryosuke Niwa.
796
797         This resolves a long-standing FIXME to rename this method. :)
798
799         * Scripts/webkitpy/tool/bot/botinfo.py:
800         (BotInfo.summary_text):
801         * Scripts/webkitpy/tool/bot/irc_command.py:
802         (Rollout._update_working_copy):
803         * Scripts/webkitpy/tool/bot/layouttestresultsreader.py:
804         (LayoutTestResultsReader._create_layout_test_results):
805         (LayoutTestResultsReader._create_unit_test_results):
806         (LayoutTestResultsReader._results_directory):
807         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py:
808         (test_missing_unit_test_results_path):
809         * Scripts/webkitpy/tool/main.py:
810         (WebKitPatch.deprecated_port):
811         * Scripts/webkitpy/tool/mocktool.py:
812         (MockTool.deprecated_port):
813         * Scripts/webkitpy/tool/steps/build.py:
814         (Build.build):
815         * Scripts/webkitpy/tool/steps/checkstyle.py:
816         (CheckStyle.run):
817         * Scripts/webkitpy/tool/steps/commit.py:
818         (Commit._check_test_expectations):
819         * Scripts/webkitpy/tool/steps/preparechangelog.py:
820         (PrepareChangeLog.run):
821         * Scripts/webkitpy/tool/steps/preparechangelogfordepsroll.py:
822         (PrepareChangeLogForDEPSRoll.run):
823         * Scripts/webkitpy/tool/steps/preparechangelogforrevert.py:
824         (PrepareChangeLogForRevert.run):
825         * Scripts/webkitpy/tool/steps/runtests.py:
826         (RunTests.run):
827         * Scripts/webkitpy/tool/steps/steps_unittest.py:
828         (StepsTest.test_runtests_args):
829         * Scripts/webkitpy/tool/steps/update.py:
830         (Update._update_command):
831
832 2013-01-03  Eric Seidel  <eric@webkit.org>
833
834         Add myself to the webkitpy watchlist.
835
836         Unreviewed.
837
838         * Scripts/webkitpy/common/config/watchlist:
839
840 2013-01-03  Ryosuke Niwa  <rniwa@webkit.org>
841
842         Cleanup perftest* tests and add a test for computing statistics
843         https://bugs.webkit.org/show_bug.cgi?id=105685
844
845         Reviewed by Eric Seidel.
846
847         Add a test for PerfTest.compute_statistics (moved and renamed from PageLoadingPerfTest.calculate_statistics) and
848         extracted perftestsrunner_itegrationtests.py from perftestsrunner_unittests.py.
849
850         Also fixed a bug in compute_statistics that the mean ('avg') value can have a large rounding errors in some cases.
851
852         * Scripts/webkitpy/performance_tests/perftest.py:
853         (PerfTest.compute_statistics): Moved from PageLoadingPerfTest to prepare for the bug 97510. Also compute the mean
854         directly from sorted_values instead of using the one from Knuth's online algorithm. This approach gives more
855         accurate result for the mean.
856         (PageLoadingPerfTest.run_single):
857         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
858         (MainTest.test_compute_statistics):
859         (MainTest.test_compute_statistics.compute_statistics): Added.
860         (TestPageLoadingPerfTest.test_run): floatify values.
861         (TestPageLoadingPerfTest.test_run_with_memory_output): Ditto. Also got rid of ".0" from mean values now that Python
862         correctly recognizes them as integers.
863
864         * Scripts/webkitpy/performance_tests/perftestsrunner_integrationtest.py: Copied from
865         Tools/Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py.
866         (TestDriver): Moved out of MainTest.
867         (MainTest): Got rid of assertWritten and all unit tests.
868         (MainTest._normalize_output): Renamed from normalizeFinishedTime to match the PEP8 naming convention.
869         (MainTest.test_run_test_set_kills_drt_per_run.TestDriverWithStopCount):
870         (MainTest.test_run_test_set_for_parser_tests):
871         (MainTest.test_run_memory_test):
872         (MainTest._test_run_with_json_output):
873         (MainTest.test_run_generates_json_by_default):
874         (MainTest.test_run_merges_output_by_default):
875         (MainTest.test_run_respects_reset_results):
876         (MainTest.test_run_generates_and_show_results_page): Use runner.load_output_json() instead of manually loading and
877         parsing output JSON files. Just verify that the output path is correct instead.
878         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
879         (MainTest): Removed all integration tests.
880         (MainTest.create_runner): Simplified.
881
882 2013-01-03  Julie Parent  <jparent@chromium.org>
883
884         Dashboard cleanup: remove usage of global g_defaultBuilderName
885         https://bugs.webkit.org/show_bug.cgi?id=106043
886
887         Use now public defaultBuilder() from the BuilderGroup, rather
888         than global variable.
889
890         Also eliminates the setup() and initBuilders() functions since we
891         no longer need to track the global.
892
893         Reviewed by Adam Barth.
894
895         * TestResultServer/static-dashboards/builders.js:
896         (BuilderGroup.prototype.defaultBuilder):
897         (BuilderGroup.prototype.master):
898         * TestResultServer/static-dashboards/dashboard_base.js:
899         (parseParameters):
900         * TestResultServer/static-dashboards/flakiness_dashboard.js:
901         (updateDefaultBuilderState):
902         * TestResultServer/static-dashboards/loader.js:
903         (.):
904         * TestResultServer/static-dashboards/loader_unittests.js:
905
906 2013-01-03  Ryosuke Niwa  <rniwa@webkit.org>
907
908         Disable an intermittently failing unit test for now.
909         The failure is tracked by the bug 106051.
910
911         * TestWebKitAPI/Tests/mac/MemoryCachePruneWithinResourceLoadDelegate.mm:
912         (TestWebKitAPI::TEST):
913
914 2013-01-03  Dan Bernstein  <mitz@apple.com>
915
916         No way to obtain a DOMNode given a JS wrapper for a Node
917         https://bugs.webkit.org/show_bug.cgi?id=106033
918
919         Added a test for +[DOMNode _nodeFromJSWrapper:].
920
921         Reviewed by Sam Weinig.
922
923         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
924         * TestWebKitAPI/Tests/mac/DOMNodeFromJSObject.mm: Added.
925         (TestWebKitAPI):
926         (TestWebKitAPI::TEST):
927
928 2013-01-03  Emil A Eklund  <eae@chromium.org>
929
930         Fix overflow in LayoutUnit::ceil and floor for SATURATED_LAYOUT_ARITHMETIC
931         https://bugs.webkit.org/show_bug.cgi?id=105961
932
933         Reviewed by Levi Weintraub.
934         
935         Add tests for LayoutUnit::ceil and floor.
936
937         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp:
938         (TestWebKitAPI::TEST):
939         (TestWebKitAPI):
940
941 2013-01-03  Julie Parent  <jparent@chromium.org>
942
943         Add unit test for default builder for the dashboards.
944         https://bugs.webkit.org/show_bug.cgi?id=105972
945
946         Reviewed by Dirk Pranke.
947
948         * TestResultServer/static-dashboards/loader_unittests.js:
949
950 2013-01-03  Tony Chang  <tony@chromium.org>
951
952         Generate internal.settings from Settings.in
953         https://bugs.webkit.org/show_bug.cgi?id=104740
954
955         Reviewed by Adam Barth.
956
957         * GNUmakefile.am: Compile new generated files on GTK+.
958         * qmake/mkspecs/features/default_post.prf: Qt: Make it so an action
959         with both extra_sources and add_output_to_sources=false will still
960         compile the files in extra_sources. This makes it possible to compile
961         InternalSettingsGenerated.cpp and not be confused by
962         InternalSettingsGenerated.idl.
963
964 2013-01-03  Zan Dobersek  <zandobersek@gmail.com>
965
966         [style] WebIDL-reflecting uppercase enums reported as style violation
967         https://bugs.webkit.org/show_bug.cgi?id=105832
968
969         Reviewed by Tony Chang.
970
971         Recognize the enum as WebIDL-reflecting one if there's a comment present that
972         indicates this. In enums spanning over multiple lines, the comment is expected
973         on the first line. In enums placed in one line, the comment is expected at the
974         end of that line.
975
976         WebIDL-reflecting enum values are required to be written in uppercase and as such
977         violate the common rule of these values being written in InterCaps. Annotating
978         enums this way makes it possible to avoid false style violation reports.
979
980         The comment is of form '// Web(?:Kit)?IDL enum'.
981
982         * Scripts/webkitpy/style/checkers/cpp.py:
983         (_EnumState.__init__):
984         (_EnumState.process_clean_line):
985         (check_enum_casing):
986         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
987         (NoNonVirtualDestructorsTest.test_enum_casing):
988
989 2013-01-03  Tim 'mithro' Ansell  <mithro@mithis.com>
990
991         Adding favicon.ico to the garden-o-matic.
992         https://bugs.webkit.org/show_bug.cgi?id=105982
993
994         Reviewed by Adam Barth.
995
996         * Scripts/webkitpy/tool/servers/gardeningserver.py:
997         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/favicon.ico:
998
999 2013-01-03  Jonathan Liu  <net147@gmail.com>
1000
1001         [Qt] Fix library linking order
1002         https://bugs.webkit.org/show_bug.cgi?id=105818
1003
1004         Some Qt libraries are linked before the private static libraries
1005         that depend on them which results in QtWebKit failing to link.
1006
1007         Copy Qt module dependencies from QT to QT_PRIVATE when creating
1008         module to ensure the Qt libraries are linked after the private
1009         static libraries that depend on them.
1010
1011         Reviewed by Tor Arne Vestbø.
1012
1013         * qmake/mkspecs/features/default_post.prf:
1014
1015 2013-01-02  Byungwoo Lee  <bw80.lee@samsung.com>
1016
1017         [EFL][WK2] Fix tooltip bugs in MiniBrowser.
1018         https://bugs.webkit.org/show_bug.cgi?id=102209
1019
1020         Reviewed by Gyuyoung Kim.
1021
1022         Set tooltip on the elm_window instead of ewk_view, and add delay to
1023         show the tooltip window.
1024
1025         Setting tooltip on the ewk_view makes elementary error message because
1026         ewk_view is not an elementary widget. To remove the error message,
1027         tooltip should be set on the elm_window.
1028
1029         Most browsers show the tooltip after a mouse pointer stays on an
1030         element object that has tooltip text.
1031         But elm_tooltip widget itself doesn't have delay or handle mouse
1032         events to show/hide.
1033
1034         So additional codes are added to show/hide tooltip widget along with
1035         mouse events.
1036           1) When a mouse pointer stays for a second on an element which has
1037              tooltip text, then tooltip widget is displayed.
1038           2) When a mouse pointer is moved or out of ewk_view, hide the tooltip
1039              widget.
1040
1041         * MiniBrowser/efl/main.c:
1042         (_Tooltip_Information):
1043         (_Browser_Window):
1044         (on_tooltip_show):
1045         (window_tooltip_hide):
1046         (window_tooltip_update):
1047         (on_mouse_in):
1048         (on_mouse_move):
1049         (on_mouse_out):
1050         (window_free):
1051         (on_tooltip_text_set):
1052         (on_tooltip_text_unset):
1053         (window_create):
1054
1055 2013-01-02  Tony Chang  <tony@chromium.org>
1056
1057         Unreviewed, rolling out r138661.
1058         http://trac.webkit.org/changeset/138661
1059         https://bugs.webkit.org/show_bug.cgi?id=104740
1060
1061         Compile problems on EFL
1062
1063         * GNUmakefile.am:
1064         * qmake/mkspecs/features/default_post.prf:
1065
1066 2013-01-02  Tony Chang  <tony@chromium.org>
1067
1068         Generate internal.settings from Settings.in
1069         https://bugs.webkit.org/show_bug.cgi?id=104740
1070
1071         Reviewed by Adam Barth.
1072
1073         * GNUmakefile.am: Compile new generated files on GTK+.
1074         * qmake/mkspecs/features/default_post.prf: Qt: Make it so an action
1075         with both extra_sources and add_output_to_sources=false will still
1076         compile the files in extra_sources. This makes it possible to compile
1077         InternalSettingsGenerated.cpp and not be confused by
1078         InternalSettingsGenerated.idl.
1079
1080 2013-01-02  Adam Barth  <abarth@webkit.org>
1081
1082         Remove old-run-webkit-tests support from webkitpy
1083         https://bugs.webkit.org/show_bug.cgi?id=105619
1084
1085         Reviewed by Eric Seidel.
1086
1087         This patch removes support for webkitpy parsing the
1088         old-run-webkit-tests results format. There doesn't seem to be any
1089         reason to keep this code around anymore now that most bots have
1090         switched to new-run-webkit-tests.
1091
1092         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
1093         (Builder.fetch_layout_test_results):
1094         * Scripts/webkitpy/common/net/layouttestresults.py:
1095         (path_for_layout_test):
1096         (LayoutTestResults.results_from_string):
1097         (LayoutTestResults):
1098         * Scripts/webkitpy/common/net/layouttestresults_unittest.py:
1099         (LayoutTestResultsTest.test_results_from_string):
1100         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
1101         (test_flaky_test_failure):
1102         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py:
1103         (LayoutTestResultsReaderTest.test_missing_layout_test_results):
1104         * Scripts/webkitpy/tool/bot/patchanalysistask.py:
1105         (PatchAnalysisTask._test_patch):
1106         * Scripts/webkitpy/tool/commands/queuestest.py:
1107         (QueuesTest.assert_queue_outputs):
1108
1109 2013-01-02  Heikki Paajanen  <heikki.paajanen@palm.com>
1110
1111         [Qt][WK2] Add simple UI to find text from page
1112         https://bugs.webkit.org/show_bug.cgi?id=96481
1113
1114         Reviewed by Jocelyn Turcotte.
1115
1116         Find bar that slides between navigation bar and WebView.
1117
1118         * MiniBrowser/qt/BrowserWindow.cpp:
1119         (BrowserWindow::toggleFind):
1120         * MiniBrowser/qt/BrowserWindow.h:
1121         (BrowserWindow):
1122         * MiniBrowser/qt/MiniBrowser.qrc:
1123         * MiniBrowser/qt/MiniBrowserApplication.cpp:
1124         (MiniBrowserApplication::notify):
1125         * MiniBrowser/qt/icons/find.png: Added.
1126         * MiniBrowser/qt/qml/BrowserWindow.qml:
1127
1128 2013-01-02  Christophe Dumez  <christophe.dumez@intel.com>
1129
1130         Regression(r138603): Caused 2 webkitpy tests to fail
1131         https://bugs.webkit.org/show_bug.cgi?id=105913
1132
1133         Reviewed by Kentaro Hara.
1134
1135         Fix 2 webkitpy tests that started to fail after
1136         r138603.
1137
1138         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
1139         (IRCCommandTest._sheriff_test_data_url): Make sure the
1140         returned test file URL is valid no matter what the
1141         current working directory is.
1142         * Scripts/webkitpy/tool/bot/ircbot_unittest.py:
1143         (IRCBotTest.test_help): Add "sheriffs" command to the
1144         list of available commands.
1145
1146 2013-01-01  Christophe Dumez  <christophe.dumez@intel.com>
1147
1148         [WK2] plugins/clicking-missing-plugin-fires-delegate.html is failing
1149         https://bugs.webkit.org/show_bug.cgi?id=105897
1150
1151         Reviewed by Alexey Proskuryakov.
1152
1153         Dump in WebKitTestRunner the message that is expected by layout tests
1154         when the missing plugin button is clicked.
1155
1156         * WebKitTestRunner/TestController.cpp:
1157         (WTR::TestController::createWebViewWithOptions):
1158         (WTR::TestController::unavailablePluginButtonClicked):
1159         (WTR):
1160         * WebKitTestRunner/TestController.h:
1161         (TestController):
1162
1163 2013-01-01  KwangYong Choi  <ky0.choi@samsung.com>
1164
1165         [EFL] Enable MHTML feature
1166         https://bugs.webkit.org/show_bug.cgi?id=105815
1167
1168         Reviewed by Laszlo Gombos.
1169
1170         Enable MHTML feature for EFL port by default.
1171
1172         * Scripts/webkitperl/FeatureList.pm:
1173
1174 2013-01-01  Alan Cutter  <alancutter@chromium.org>
1175
1176         sheriff-bot should know who the gardeners/sheriffs are
1177         https://bugs.webkit.org/show_bug.cgi?id=105698
1178
1179         Reviewed by Eric Seidel.
1180
1181         Added a "sheriffs" command to sheriffbot to fetch the current Chromium Webkit sheriffs.
1182
1183         * Scripts/webkitpy/tool/bot/irc_command.py:
1184         (Current):
1185         (Current._retrieve_webkit_sheriffs):
1186         (Current.execute):
1187         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
1188         (IRCCommandTest.test_current):
1189         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_0.js: Added.
1190         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_1.js: Added.
1191         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_2.js: Added.
1192         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_malformed.js: Added.
1193         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_zero.js: Added.
1194
1195 2012-12-31  Carlos Garcia Campos  <cgarcia@igalia.com>
1196
1197         [GTK] Move ImageDiff to the Tools directory root
1198         https://bugs.webkit.org/show_bug.cgi?id=105421
1199
1200         Reviewed by Kenneth Rohde Christiansen.
1201
1202         * DumpRenderTree/gtk/GNUmakefile.ImageDiff.am: Removed.
1203         * GNUmakefile.am:
1204         * ImageDiff/gtk/ImageDiff.cpp: Renamed from Tools/DumpRenderTree/gtk/ImageDiff.cpp.
1205
1206 2012-12-30  Mike West  <mkwst@chromium.org>
1207
1208         Add myself to watchlists for CSP and console logs.
1209         https://bugs.webkit.org/show_bug.cgi?id=105874
1210
1211         Reviewed by Dirk Pranke.
1212
1213         Because I really, really, really need more email.
1214
1215         * Scripts/webkitpy/common/config/committers.py:
1216             Add mkwst+watchlist@chromium.org as a contributor.
1217         * Scripts/webkitpy/common/config/watchlist:
1218             Create watchlist categories for 1) the various mechanisms used to
1219             generate console messages, and 2) CSP usage and files.
1220
1221 2012-12-30  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
1222
1223         [EFL] Enable TEMPLATE_ELEMENT feature
1224         https://bugs.webkit.org/show_bug.cgi?id=105865
1225         
1226         Reviewed by Laszlo Gombos.
1227
1228         * Scripts/webkitperl/FeatureList.pm: Enable template element feature for EFL port.
1229
1230 2012-12-30  Zan Dobersek  <zandobersek@gmail.com>
1231
1232         Unreviewed GTK gardening of the unit tests.
1233
1234         Skip the TestInspectorServer unit test that's timing out completely.
1235
1236         * Scripts/run-gtk-tests:
1237         (TestRunner):
1238
1239 2012-12-29  Ilya Tikhonovsky  <loislo@chromium.org>
1240
1241         Web Inspector: Native Memory Instrumentation plugin: do not generate 'not instrumented' warning for instrumented mutable members.
1242         https://bugs.webkit.org/show_bug.cgi?id=105855
1243
1244         Reviewed by Vsevolod Vlasov.
1245
1246         Extract MemberExpr from ImplicitCastExpr. It happens when we instrument a mutable member because
1247         addMember expects const T& and the mutable ref to member implicitly converts into const ref to member.
1248
1249         * clang/ReportMemoryUsagePlugin/ReportMemoryUsage.cpp:
1250         (ReportMemoryUsageVisitor):
1251         (clang::AddMemberCallVisitor::VisitCallExpr):
1252         (clang::ReportMemoryUsageVisitor::VisitCXXMethodDecl):
1253         (clang::ReportMemoryUsageVisitor::checkMembersCoverage):
1254
1255 2012-12-29  Ilya Tikhonovsky  <loislo@chromium.org>
1256
1257         Web Inspector: Native Memory Instrumentation plugin: move function bodies out of class declarations.
1258         https://bugs.webkit.org/show_bug.cgi?id=105852
1259
1260         Reviewed by Alexander Pavlov.
1261
1262         * clang/ReportMemoryUsagePlugin/ReportMemoryUsage.cpp:
1263         (clang):
1264         (AddMemberCallVisitor):
1265         (ReportMemoryUsageVisitor):
1266         (clang::ReportMemoryUsageAction::ParseArgs):
1267         (clang::AddMemberCallVisitor::VisitCallExpr):
1268         (clang::ReportMemoryUsageVisitor::VisitCXXMethodDecl):
1269         (clang::ReportMemoryUsageVisitor::emitWarning):
1270         (clang::ReportMemoryUsageVisitor::findInstrumentationMethod):
1271         (clang::ReportMemoryUsageVisitor::needsToBeInstrumented):
1272         (clang::ReportMemoryUsageVisitor::CheckMembersCoverage):
1273
1274 2012-12-29  Zan Dobersek  <zandobersek@gmail.com>
1275
1276         [webkitpy] Omit webkitpy/thirdparty/BeautifulSoup.py from code coverage checking
1277         https://bugs.webkit.org/show_bug.cgi?id=105835
1278
1279         Reviewed by Dirk Pranke.
1280
1281         The BeautifulSoup module is a third-party import and as such we don't have much control
1282         over its testing. Because of that it should be omitted from code coverage checking.
1283
1284         * Scripts/webkitpy/test/main.py:
1285         (Tester._run_tests):
1286
1287 2012-12-28  Martin Robinson  <mrobinson@igalia.com>
1288
1289         [GTK][WK2] Add support for IME Composition
1290         https://bugs.webkit.org/show_bug.cgi?id=65093
1291
1292         Reviewed by Carlos Garcia Campos.
1293
1294         Update unit tests for GtkInputMethodFilter to note whether or not an event
1295         was a fake event synthesized for composition purposes. This is now used
1296         for WebKit2.
1297
1298         * TestWebKitAPI/Tests/gtk/InputMethodFilter.cpp:
1299         (TestWebKitAPI::TestInputMethodFilter::sendSimpleKeyEvent):
1300         (TestWebKitAPI::TestInputMethodFilter::sendKeyEventWithCompositionResults):
1301         (TestWebKitAPI::TEST):
1302         * gtk/generate-gtkdoc:
1303         (get_webkit2_options): Skip new files in the WebKit2 source directory for documentation
1304         generation.
1305
1306 2012-12-27  Ilya Tikhonovsky  <loislo@chromium.org>
1307
1308         Web Inspector: Native Memory Instrumentation. Update clang plugin according to the current state of memory instrumentation code.
1309         https://bugs.webkit.org/show_bug.cgi?id=105800
1310
1311         Reviewed by Yury Semikhatsky.
1312
1313         * clang/ReportMemoryUsagePlugin/ReportMemoryUsage.cpp:
1314         (clang::ReportMemoryUsageConsumer::ReportMemoryUsageConsumer):
1315         (clang):
1316
1317 2012-12-27  Zan Dobersek  <zandobersek@gmail.com>
1318
1319         Create a GTK build system watchlist and add myself to it
1320         https://bugs.webkit.org/show_bug.cgi?id=105769
1321
1322         Reviewed by Dirk Pranke.
1323
1324         Add a GTK build system watchlist, watching the configure.ac, GNUmakefile.am files
1325         and the input GNUmakefile.features.am.in file. The GNUmakefile.list.am files are
1326         omitted on purpose due to relatively high change rate they are exposed to.
1327
1328         The new watchlist would be of use to me as the GTK build system recently went
1329         through a series of changes that included a changed policy on adding configuration
1330         flags to configure.ac and subsequent additions to GNUmakefile.am files. A PSA on the
1331         mailing list about this change will be published later, but I'd also like to be
1332         notified of any pending changes to these files just so I can check the changes are
1333         in compliance with the new policy.
1334
1335         * Scripts/webkitpy/common/config/watchlist:
1336
1337 2012-12-26  Ilya Tikhonovsky  <loislo@chromium.org>
1338
1339         Unreviewed. Replace find("Tools") with rfind("Tools") because base dir of WebKit may have Tools word.
1340         As example ..../DevTools/src/third_party/WebKit
1341
1342         * Scripts/webkitpy/common/webkit_finder.py:
1343         (WebKitFinder.webkit_base):
1344
1345 2012-12-25  Christophe Dumez  <christophe.dumez@intel.com>
1346
1347         [EFL][WK2] Refactor snapshot taking code
1348         https://bugs.webkit.org/show_bug.cgi?id=105687
1349
1350         Reviewed by Kenneth Rohde Christiansen.
1351
1352         Call WKViewCreateSnapshot() instead of WKViewGetSnapshot()
1353         since the function was renamed to follow naming style.
1354
1355         * WebKitTestRunner/efl/PlatformWebViewEfl.cpp:
1356         (WTR::PlatformWebView::windowSnapshotImage):
1357
1358 2012-12-24  Ilya Tikhonovsky  <loislo@chromium.org>
1359
1360         Web Inspector: Native Memory Instrumentation: propagate member type as edge type to the serialized heap graph.
1361         https://bugs.webkit.org/show_bug.cgi?id=105725
1362
1363         Reviewed by Yury Semikhatsky.
1364
1365         MemberType value names were adjusted according to Style Guide.
1366         Existing tests were extended with link type validation.
1367
1368         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
1369
1370 2012-12-24  Xiaobo Wang  <xbwang@torchmobile.com.cn>
1371
1372         [BlackBerry] DRT - Crashed when running pixels tests
1373         https://bugs.webkit.org/show_bug.cgi?id=105711
1374
1375         RIM PR 270003
1376
1377         Reviewed by George Staikos.
1378
1379         Calling window->post() here will result in buffer swapping which is
1380         only allowed in UI thread now, and I found we don't need to call it
1381         for GL renderer.
1382
1383         * DumpRenderTree/blackberry/PixelDumpSupportBlackBerry.cpp:
1384         (createBitmapContextFromWebView):
1385
1386 2012-12-22  Dan Winship  <danw@gnome.org>
1387
1388         Add a patch to fix libxml2 2.9.0 regression
1389         https://bugs.webkit.org/show_bug.cgi?id=104494
1390
1391         Reviewed by Martin Robinson.
1392
1393         * gtk/jhbuild.modules:
1394         * gtk/patches/libxml2-2.9.0-dtd.patch: patch from upstream
1395
1396 2012-12-23  Allan Sandfeld Jensen  <allan.jensen@digia.com>
1397
1398         Unreviewed update of status to reviewer.
1399
1400         * Scripts/webkitpy/common/config/committers.py:
1401
1402 2012-12-23  Philippe Normand  <pnormand@igalia.com>
1403
1404         [GTK][jhbuild] Bump GStreamer to 1.0.4
1405         https://bugs.webkit.org/show_bug.cgi?id=105640
1406
1407         Reviewed by Martin Robinson.
1408
1409         * gtk/jhbuild.modules: Update GStreamer modules to 1.0.4.
1410
1411 2012-12-21  Nico Weber  <thakis@chromium.org>
1412
1413         chromium: webkit-build-directory doesn't handle ninja output dirs
1414         https://bugs.webkit.org/show_bug.cgi?id=105671
1415
1416         Reviewed by Daniel Bates.
1417
1418         Teach webkit-build-directory to detect the ninja build directory.
1419
1420         * Scripts/webkitdirs.pm:
1421         (determineBaseProductDir):
1422         (determineIsChromiumNinja):
1423
1424 2012-12-21  Ryosuke Niwa  <rniwa@webkit.org>
1425
1426         [Mountain Lion] platform/mac/editing/spelling/removing-underline-after-accepting-autocorrection-using-punctuation.html failing
1427         https://bugs.webkit.org/show_bug.cgi?id=95477
1428
1429         Reviewed by Sam Weinig.
1430
1431         Co-authored by Jia Pu.
1432
1433         Specify a fixed autocorrection dictionary in resetDefaultsToConsistentValues() so that
1434         we can have consistent autocorrection behavior when running tests.
1435
1436         * DumpRenderTree/mac/DumpRenderTree.mm:
1437         (resetDefaultsToConsistentValues):
1438         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
1439         (WTR::InjectedBundle::platformInitialize):
1440
1441 2012-12-21  Zan Dobersek  <zandobersek@gmail.com>
1442
1443         [GTK] Add WebKit1-specific TestExpectations file
1444         https://bugs.webkit.org/show_bug.cgi?id=97562
1445
1446         Reviewed by Dirk Pranke.
1447
1448         Implement the expectations_files method inside the GtkPort class that
1449         returns the list of the TestExpectations files the current test configuration
1450         requires. The method is implemented in a similar way as on the EFL port.
1451         A unit test is also added.
1452
1453         * Scripts/webkitpy/layout_tests/port/gtk.py:
1454         (GtkPort._search_paths):
1455         (GtkPort):
1456         (GtkPort.expectations_files):
1457         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
1458         (GtkPortTest.test_expectations_files):
1459
1460 2012-12-21  Sudarsana Nagineni  <sudarsana.nagineni@intel.com>
1461
1462         [EFL] Enable feature option for the CSS sticky position
1463         https://bugs.webkit.org/show_bug.cgi?id=105630
1464
1465         Reviewed by Kenneth Rohde Christiansen.
1466
1467         Enable CSS sticky position feature by default for the
1468         EFL port. This feature option was added in r138252.
1469
1470         No new tests. Existing layout tests pass.
1471
1472         * Scripts/webkitperl/FeatureList.pm:
1473
1474 2012-12-21  Jussi Kukkonen  <jussi.kukkonen@intel.com>
1475
1476         [WK2] REGRESSION(r138232): It made 4 tests timeout
1477         https://bugs.webkit.org/show_bug.cgi?id=105623
1478
1479         Reviewed by Kenneth Rohde Christiansen.
1480
1481         r138232 changed functionality so that empty text output
1482         no longer leads to a dump() call, which leads to a timeout.
1483         Instead do a text dump if there is text output or if there
1484         is no audio output.
1485
1486         * WebKitTestRunner/TestInvocation.cpp:
1487         (WTR::TestInvocation::dumpResults):
1488
1489 2012-12-21  Zeno Albisser  <zeno@webkit.org>
1490
1491         [Qt][Mac] run-launcher cannot find Qt5WebKitWidgets.
1492         https://bugs.webkit.org/show_bug.cgi?id=105309
1493
1494         On Windows the library name is Qt5WebKitWidgets as well.
1495         So r138136 should not have adjusted the library name for Windows.
1496
1497         Reviewed by Simon Hausmann.
1498
1499         * Scripts/webkitdirs.pm:
1500         (builtDylibPathForName):
1501
1502 2012-12-21  Zeno Albisser  <zeno@webkit.org>
1503
1504         [Qt] Add new TestExpectations for qt-5.0-mac-wk2.
1505         https://bugs.webkit.org/show_bug.cgi?id=105424
1506
1507         This change will allow us to run a very specific
1508         subset of the available LayoutTests on a mac buildbot.
1509         The new baseline search path for mac will look as follows:
1510         qt-5.0-mac-wk2 -> qt-5.0-wk2 -> qt-5.0 -> qt-mac -> qt
1511
1512         Reviewed by Dirk Pranke.
1513
1514         * Scripts/webkitpy/layout_tests/port/qt.py:
1515         (QtPort._search_paths):
1516         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
1517         (QtPortTest):
1518
1519 2012-12-20  Ryuan Choi  <ryuan.choi@samsung.com>
1520
1521         [EFL] Build break with latest EFL libraries.
1522         https://bugs.webkit.org/show_bug.cgi?id=104827
1523
1524         Reviewed by Laszlo Gombos.
1525
1526         The eo EFL package is introduced and evas and ecore use it since 1.8.
1527         While introducing Eo, EFL changed several structures of Evas and Ecore
1528         from own specific class to Eo.
1529
1530         So, this patch adds FindEo and changes declaration of Evas, Evas_Object,
1531         Ecore_Timer to build with latest EFL libraries.
1532
1533         * DumpRenderTree/efl/CMakeLists.txt: Includes eo EFL library.
1534         * EWebLauncher/CMakeLists.txt: Ditto.
1535         * MiniBrowser/efl/CMakeLists.txt: Ditto.
1536         * ImageDiff/PlatformEfl.cmake: Ditto.
1537         * TestWebKitAPI/PlatformEfl.cmake: Ditto.
1538         * TestWebKitAPI/PlatformWebView.h: Modified declaration of Evas_Object.
1539         * WebKitTestRunner/InjectedBundle/TestRunner.h: Modified declaration of Ecore_Timer.
1540         * WebKitTestRunner/PlatformEfl.cmake: Includes eo EFL library.
1541         * WebKitTestRunner/PlatformWebView.h: Modified declaration of Evas_Object.
1542
1543 2012-12-20  Dirk Pranke  <dpranke@chromium.org>
1544
1545         TestExpectation fallback is broken
1546         https://bugs.webkit.org/show_bug.cgi?id=105583
1547
1548         Reviewed by Ryosuke Niwa.
1549
1550         When I added support for the -wk2 in port names, it looks
1551         like I broke the way we computed the list of TestExpectations
1552         files so that the implementation-version directory was left out :(.
1553         Fixed and cleaned up the tests.
1554
1555         * Scripts/webkitpy/layout_tests/port/base.py:
1556         (Port.expectations_files):
1557         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
1558         (TestWebKitPort.__init__):
1559         (PortTestCase.test_skipped_directories_for_symbols):
1560         (test_skipped_directories_for_features):
1561         (test_skipped_directories_for_features_no_matching_tests_in_test_list):
1562         (test_skipped_tests_for_unsupported_features_empty_test_list):
1563         (test_skipped_layout_tests):
1564         (test_expectations_files):
1565
1566 2012-12-20  Anders Carlsson  <andersca@apple.com>
1567
1568         Don't include WebKit2 headers when building TestWebKitAPI on windows
1569         https://bugs.webkit.org/show_bug.cgi?id=105582
1570
1571         Reviewed by Tim Horton.
1572
1573         * TestWebKitAPI/config.h:
1574         * TestWebKitAPI/win/main.cpp:
1575
1576 2012-12-20  Ryosuke Niwa  <rniwa@webkit.org>
1577
1578         REGRESSION(r133820?): SimplifyMarkupTest API test asserts
1579         https://bugs.webkit.org/show_bug.cgi?id=105370
1580
1581         Reviewed by Simon Fraser.
1582
1583         Re-enable the test.
1584
1585         * TestWebKitAPI/Tests/mac/SimplifyMarkup.mm:
1586         (TestWebKitAPI::TEST):
1587
1588 2012-12-20  Nico Weber  <thakis@chromium.org>
1589
1590         chromium nrwt: Pick the newest binary found in DEFAULT_BUILD_DIRECTORIES, not the first
1591         https://bugs.webkit.org/show_bug.cgi?id=105498
1592
1593         Reviewed by Dirk Pranke.
1594
1595         Use the newest binary available rather than an than always picking one
1596         build directory over another based on iteration order.
1597
1598         * Scripts/webkitpy/layout_tests/port/chromium.py:
1599         (ChromiumPort._static_build_path):
1600         Check for timestamps.
1601         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
1602         (ChromiumMacPortTest.test_build_path_timestamps):
1603         Test that out / xcodebuild selection happens based on timestamps
1604         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
1605         (ChromiumWinPortTest.test_build_path_timestamps):
1606         Test that out / build selection happens based on timestamps
1607
1608 2012-12-19  Simon Fraser  <simon.fraser@apple.com>
1609
1610         Lots of sticky tests failing in WK2
1611         https://bugs.webkit.org/show_bug.cgi?id=105464
1612
1613         Reviewed by Beth Dakin.
1614
1615         WebKitTestRunner had a race between snapshotting in the UI process,
1616         and resettting after the test in the web process. InjectedBundle::done()
1617         was a bad place to call page()->resetAfterTest(), because of this race;
1618         it could reset the scroll position before the UI snapshot had been obtained.
1619         
1620         Fix by moving the call to page()->resetAfterTest() into didReceiveMessage(),
1621         for the "Reset" message which will come in before the next test.
1622
1623         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
1624         (WTR::InjectedBundle::didReceiveMessage):
1625         (WTR::InjectedBundle::done):
1626
1627 2012-12-19  Filip Pizlo  <fpizlo@apple.com>
1628
1629         DFG speculation checks that take JumpList should consolidate OSRExits
1630         https://bugs.webkit.org/show_bug.cgi?id=105401
1631
1632         Reviewed by Oliver Hunt.
1633
1634         Update the profiler to understand that an OSR exit may have multiple
1635         candidate exit sites.
1636
1637         * Scripts/display-profiler-output:
1638
1639 2012-12-20  Eric Seidel  <eric@webkit.org>
1640
1641         EWSes do not need to dump build/test logs when things pass
1642         https://bugs.webkit.org/show_bug.cgi?id=105402
1643
1644         Reviewed by Dirk Pranke.
1645
1646         When we were first writing the commit-queue and EWS system
1647         and running them locally, we thought it was important to see the
1648         build/test output as it happened.  Now these queues have been running
1649         for years, and we don't use this feature anymore.  So this patch
1650         makes us stop tee-ing the output directly to stdout, and instead
1651         only print any output from commands when they fail.
1652
1653         I also added some logging when we execute each command
1654         since we are no longer printing the command output itself.
1655         This should make it much easier to see errors with the EWS
1656         system itself, now that all the passing-build output is removed.
1657
1658         This required updating a zillion unittests, since we have a bunch of
1659         "integration" style testing of the queue-bots.  Many of the unittests for the bots were
1660         already approximating this behavior by using a "logging" Executive.
1661         I disabled logging in those cases to avoid redundant logging in those tests.
1662
1663         * Scripts/webkitpy/common/system/executive.py:
1664         (Executive.command_for_printing):
1665         (Executive.run_command):
1666         * Scripts/webkitpy/common/system/executive_mock.py:
1667         (MockExecutive.command_for_printing):
1668         (MockExecutive.run_command):
1669         * Scripts/webkitpy/common/system/executive_unittest.py:
1670         (ExecutiveTest.test_auto_stringify_args):
1671         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
1672         (EarlyWarningSytemTest._default_expected_logs):
1673         (_test_ews):
1674         (_test_ewses):
1675         * Scripts/webkitpy/tool/commands/queues.py:
1676         (AbstractQueue.run_webkit_patch):
1677         * Scripts/webkitpy/tool/commands/queues_unittest.py:
1678         (AbstractQueueTest._assert_run_webkit_patch):
1679         (CommitQueueTest.test_commit_queue):
1680         (test_rollout):
1681         (test_rollout_lands):
1682         (test_manual_reject_during_processing):
1683         (StyleQueueTest.test_style_queue_with_style_exception):
1684         (test_style_queue_with_watch_list_exception):
1685
1686 2012-12-20  Dominic Mazzoni  <dmazzoni@google.com>
1687
1688         AX: support clickPoint in DRT for chromium
1689         https://bugs.webkit.org/show_bug.cgi?id=97359
1690
1691         Reviewed by Chris Fleizach.
1692
1693         Implement clickPointX, clickPointY, and elementAtPoint for Chromium, so that
1694         several more tests can pass.
1695
1696         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
1697         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
1698         (WebTestRunner::AccessibilityUIElement::clickPointXGetterCallback):
1699         (WebTestRunner):
1700         (WebTestRunner::AccessibilityUIElement::clickPointYGetterCallback):
1701         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
1702         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
1703         (AccessibilityUIElement):
1704
1705 2012-12-20  Dan Bernstein  <mitz@apple.com>
1706
1707         prepare-ChangeLog adds "Additional information of the change such as approach, rationale" where it doesn't make sense
1708         https://bugs.webkit.org/show_bug.cgi?id=99268
1709
1710         Reviewed by Alexey Proskuryakov.
1711
1712         * Scripts/prepare-ChangeLog:
1713         (generateNewChangeLogs): Removed the "Additional information of the change" line.
1714
1715 2012-12-20  Carlos Garcia Campos  <cgarcia@igalia.com>
1716
1717         Unreviewed. Skip unit test /webkit/atk/getTextInParagraphAndBodyModerate
1718
1719         * Scripts/run-gtk-tests:
1720         (TestRunner):
1721
1722 2012-12-20  Thiago Marcos P. Santos  <thiago.santos@intel.com>
1723
1724         [Qt] Move ImageDiff to the Tools directory root
1725         https://bugs.webkit.org/show_bug.cgi?id=105420
1726
1727         Reviewed by Tor Arne Vestbø.
1728
1729         Moved ImageDiff from outside DumpRenderTree because it is also a
1730         dependency when running pixel tests for WebKit2.
1731
1732         * ImageDiff/ImageDiff.pro: Renamed from Tools/DumpRenderTree/qt/ImageDiff.pro.
1733         * ImageDiff/qt/ImageDiff.cpp: Renamed from Tools/DumpRenderTree/qt/ImageDiff.cpp.
1734         * Tools.pro:
1735
1736 2012-12-20  Zan Dobersek  <zandobersek@gmail.com>
1737
1738         [GTK] Remove the --enable-unstable-features configuration option
1739         https://bugs.webkit.org/show_bug.cgi?id=105327
1740
1741         Reviewed by Martin Robinson.
1742
1743         Add a missing feature option for the CSS sticky position feature.
1744
1745         * Scripts/webkitperl/FeatureList.pm:
1746
1747 2012-12-20  Carlos Garcia Campos  <cgarcia@igalia.com>
1748
1749         Unreviewed. Skip unit test WebKit2.ForceRepaint that times out
1750
1751         * Scripts/run-gtk-tests:
1752         (TestRunner):
1753
1754 2012-12-20  Dominik Röttsches  <dominik.rottsches@intel.com>
1755
1756         [EFL] MiniBrowser does not play Infinite Gangnam Style
1757         https://bugs.webkit.org/show_bug.cgi?id=103531
1758
1759         Reviewed by Kenneth Rohde Christiansen.
1760
1761         For EFL we would like this feature to be built by default.
1762
1763         * Scripts/webkitperl/FeatureList.pm:
1764
1765 2012-12-20  Csaba Osztrogonác  <ossy@webkit.org>
1766
1767         [Qt] EWS bots should call qmake always
1768         https://bugs.webkit.org/show_bug.cgi?id=90049
1769
1770         Reviewed by Simon Hausmann.
1771
1772         * Scripts/webkitdirs.pm:
1773         (buildQMakeProjects):
1774
1775 2012-12-20  Jochen Eisinger  <jochen@chromium.org>
1776
1777         [chromium] fix stray printf in TestRunner's WebTestProxy
1778         https://bugs.webkit.org/show_bug.cgi?id=105515
1779
1780         Reviewed by Ryosuke Niwa.
1781
1782         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
1783         (WebTestRunner::WebTestProxyBase::shouldInsertNode):
1784
1785 2012-12-20  Dominik Röttsches  <dominik.rottsches@intel.com>
1786
1787         [WK2] WebAudio WKTR support
1788         https://bugs.webkit.org/show_bug.cgi?id=95084
1789
1790         Reviewed by Kenneth Rohde Christiansen.
1791
1792         Implementing audio data dump support for WebKitTestRunner,
1793         by adding the API to the IDL, enabling IPC for transferring
1794         the audio data as bytebuffer and dumping it according
1795         to run-webkit-tests script's expectations.
1796         This enables audio testing on WebKit2 ports.
1797         Based on initial wip-patch from Philippe Normand - thanks.
1798
1799         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl: Adding setAudioData API.
1800         * WebKitTestRunner/InjectedBundle/TestRunner.cpp: Assign audio buffer data to a local WKDataRef after extracting it from WK2 using WKBundleDataFromUint8Array bundle API.
1801         (WTR::TestRunner::setAudioData):
1802         (WTR):
1803         * WebKitTestRunner/InjectedBundle/TestRunner.h: Add a whatToDump state for dumping Audio.
1804         (TestRunner):
1805         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp: Add audio to the message that's transferred to the WKTR process.
1806         (WTR::InjectedBundle::done):
1807         * WebKitTestRunner/InjectedBundle/InjectedBundle.h: Allow setting and audio result handle to be transferred.
1808         (WTR::InjectedBundle::setAudioResult):
1809         (InjectedBundle):
1810         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1811         (WTR::InjectedBundlePage::dump): Dummy handling the Audio case in the switch, build fix for Mac.
1812         * WebKitTestRunner/TestInvocation.cpp: Audio dump support, dump audio if we haven't received a text dump from the web process.
1813         (WTR::TestInvocation::dumpResults):
1814         (WTR):
1815         (WTR::TestInvocation::dumpAudio):
1816         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle): Storing received audio data before dumping.
1817         * WebKitTestRunner/TestInvocation.h:
1818         (TestInvocation):
1819
1820 2012-12-20  Alberto Garcia  <agarcia@igalia.com>
1821
1822         [GTK] Count failed and timed out tests separately
1823         https://bugs.webkit.org/show_bug.cgi?id=105448
1824
1825         Reviewed by Carlos Garcia Campos.
1826
1827         Display the number of tests that fail, time out and are skipped
1828         separately in the buildbot page.
1829
1830         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
1831         (RunGtkAPITests.commandComplete):
1832         (RunGtkAPITests.evaluateCommand):
1833         (RunGtkAPITests.getText2):
1834         * Scripts/run-gtk-tests:
1835         (TestRunner.run_tests):
1836
1837 2012-12-19  Ryosuke Niwa  <rniwa@webkit.org>
1838
1839         PerfTest.parse_output does too much
1840         https://bugs.webkit.org/show_bug.cgi?id=105391
1841
1842         Reviewed by Tony Chang.
1843
1844         Moved the code to filter lines into PerfTest._filter_output, which has been renamed from _filter_stderr.
1845         Also moved the code to output test results into PerfTest._run_with_driver.
1846
1847         * Scripts/webkitpy/performance_tests/perftest.py:
1848         (PerfTest.__init__): Added _description.
1849         (PerfTest.description): Added.
1850         (PerfTest._run_with_driver): Moved the code to output test description and test results.
1851         (PerfTest._filter_output): Moved and renamed from PerfTest._filter_stderr.
1852         (PerfTest.parse_output): Removed the code to output test results. Return test results and description
1853         respectively. Also removed the code that allowed some tests to omit values since all tests report each
1854         iteration now since r136492.
1855         (PerfTest.output_statistics): Removed the code to print test description, now done in _run_with_driver.
1856         (ChromiumStylePerfTest._run_with_driver): Added. Chromium style tests are sufficiently different from
1857         regular PerfTest that it doesn't make much sense to share _run_with_driver. But really, we should just
1858         get rid of this type of test altogether in favor of regular performance tests that uses runner.js.
1859         (ChromiumStylePerfTest.parse_and_log_output): Renamed from parse_output.
1860         (PageLoadingPerfTest._run_with_driver): Removed the explicit '' for the test description.
1861         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
1862         (MainTest.test_parse_output): Removed the expected logs since parse_output no longer prints out results.
1863         Also added a call to _filter_output since parse_output doesn't filter the output text anymore.
1864         (MainTest.test_parse_output_with_failing_line): Added a call to _filter_output. Also added ',' after 'Time:'
1865         so that it's not string-concatenated with the next line.
1866         (MainTest.test_parse_output_with_description): Added; a test for PerfTest.description().
1867         (MainTest.test_parse_output_with_subtests): Removed the expected logs, and added a call to _filter_output.
1868
1869 2012-12-19  Alexis Menard  <alexis@webkit.org>
1870
1871         Implement CSS parsing for CSS transitions unprefixed.
1872         https://bugs.webkit.org/show_bug.cgi?id=104804
1873
1874         Reviewed by Dean Jackson.
1875
1876         Add a new flag ENABLE_CSS_TRANSFORMS_ANIMATIONS_TRANSITIONS_UNPREFIXED
1877         to cover the work of unprefixing Transforms, Animations and 
1878         Transitions. It will let the possibility of each ports to turn it off 
1879         in their release branches until we're confident that these CSS 
1880         properties are ready to be unprefixed.
1881
1882         * Scripts/webkitperl/FeatureList.pm:
1883         * qmake/mkspecs/features/features.pri:
1884
1885 2012-12-19  Julien Chaffraix  <jchaffraix@webkit.org>
1886
1887         2 clampTo unit tests are failing
1888         https://bugs.webkit.org/show_bug.cgi?id=105253
1889
1890         Reviewed by Emil A Eklund.
1891
1892         Removed clampToUnsigned as it was actually testing exactly what
1893         clampToUnsignedLong was doing. Also made clampToIntFloat more robust
1894         to precision loss (fix suggested by Emil).
1895
1896         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
1897
1898 2012-12-19  Thiago Marcos P. Santos  <thiago.santos@intel.com>
1899
1900         [EFL] Move ImageDiff to the Tools directory root
1901         https://bugs.webkit.org/show_bug.cgi?id=105305
1902
1903         Reviewed by Laszlo Gombos.
1904
1905         We previously moved this directory from the DRT directory to
1906         efl/ImageDiff. The major reason for doing so was because ImageDiff does
1907         not depend on DumpRenderTree but for legacy reasons was living inside the
1908         DRT folder, even though it is needed when running pixel tests for WK2.
1909
1910         Now we are moving to Tool directory root as other ports agreed to do the same.
1911
1912         * CMakeLists.txt:
1913         * ImageDiff/CMakeLists.txt: Added.
1914         * ImageDiff/PlatformEfl.cmake: Added.
1915         * ImageDiff/efl/ImageDiff.cpp: Renamed from Tools/efl/ImageDiff/ImageDiff.cpp.
1916         * efl/ImageDiff/CMakeLists.txt: Removed.
1917
1918 2012-12-19  Zeno Albisser  <zeno@webkit.org>
1919
1920         [Qt][Mac] run-launcher cannot find Qt5WebKitWidgets.
1921         https://bugs.webkit.org/show_bug.cgi?id=105309
1922
1923         The version number is only used as a part
1924         of the library name on Linux.
1925
1926         Reviewed by Kenneth Rohde Christiansen.
1927
1928         * Scripts/webkitdirs.pm:
1929         (builtDylibPathForName):
1930
1931 2012-12-19  Alberto Garcia  <agarcia@igalia.com>
1932
1933         [GTK] Tests that time out are not considered as failures.
1934         https://bugs.webkit.org/show_bug.cgi?id=105318
1935
1936         Reviewed by Martin Robinson.
1937
1938         Fix parsing of the GTK API tests log files.
1939
1940         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
1941         (RunGtkAPITests.commandComplete):
1942
1943 2012-12-18  Ilya Tikhonovsky  <loislo@chromium.org>
1944
1945         Unreviewed, rolling out r138061.
1946         https://bugs.webkit.org/show_bug.cgi?id=105396
1947
1948         Original patch:
1949         http://trac.webkit.org/changeset/138061
1950         https://bugs.webkit.org/show_bug.cgi?id=97359
1951
1952         It is crashing on Debug bots
1953
1954         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
1955         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
1956         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
1957         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
1958         (AccessibilityUIElement):
1959
1960 2012-12-18  Alice Liu  <alice.liu@apple.com>
1961
1962         Test for https://bugs.webkit.org/show_bug.cgi?id=105106
1963         Add SPI to WebKit1 WebFrame for hit testing
1964
1965         Reviewed by Dan Bernstein.
1966
1967         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added file Tests/mac/ElementAtPointInWebFrame.mm
1968         * TestWebKitAPI/Tests/mac/ElementAtPointInWebFrame.mm: Added.
1969         (TestWebKitAPI::TEST): Loads html with two divs positioned in the 2nd and 4th quadrants of the webview.
1970         Then hit-tests at three points, expecting to hit the two divs and body element. 
1971
1972 2012-12-18  Alice Liu  <alice.liu@apple.com>
1973
1974         Test for https://bugs.webkit.org/show_bug.cgi?id=105262
1975         Add SPI to WebKit1 WebFrame for node conversion to JSValueRef
1976
1977         Reviewed by Anders Carlsson.
1978
1979         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added file Tests/mac/JSWrapperForNodeInWebFrame.mm.mm
1980         * TestWebKitAPI/Tests/mac/JSWrapperForNodeInWebFrame.mm.mm: Added.
1981         (TestWebKitAPI::TEST): Tests for the correct JS wrapper for a DOMNode, provided a WebScriptWorld.
1982         Loads html with a single div element. In an isolated world, creates a property on that node.
1983         Also, in the standard world, creates a different property on that node.  Then tests for 4 things:
1984             - Existence of the isolated property in the isolated world.
1985             - Existence of the standard property in the standard world.
1986             - Non-existence of the isolated property in the standard world.
1987             - Non-existence of the standard property in the isolated world.
1988
1989 2012-12-18  Simon Fraser  <simon.fraser@apple.com>
1990
1991         Disable the SimplifyMarkup API test, wich asserts in debug builds.
1992         Tracked by https://bugs.webkit.org/show_bug.cgi?id=105370.
1993         
1994         * TestWebKitAPI/Tests/mac/SimplifyMarkup.mm:
1995         (TestWebKitAPI::TEST):
1996
1997 2012-12-18  Joseph Pecoraro  <pecoraro@apple.com>
1998
1999         [Mac] Add Build Phase to Check Headers for Inappropriate Macros (Platform.h macros)
2000         https://bugs.webkit.org/show_bug.cgi?id=104279
2001
2002         Reviewed by David Kilzer.
2003
2004         Add a build phase script that checks a framework's header files for
2005         Platform.h and Compiler.h macros. Also add some tests for the script.
2006
2007         * Scripts/check-for-inappropriate-macros-in-external-headers: Added.
2008         Script to be used in build phases to check for inappropriate macros in headers.
2009
2010         * Scripts/test-webkitruby: Added.
2011         Script that can be used to run all webkitruby tests.
2012
2013         * Scripts/test-webkit-scripts:
2014         Also run the ruby tests.
2015
2016         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/fake-data-failing-expected.txt: Added.
2017         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/pass-expected.txt: Added.
2018         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/resources/Fake.framework/Headers/Fail.h: Added.
2019         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/resources/Fake.framework/Headers/Pass.h: Added.
2020         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/run-test.rb: Added.
2021         Tests for the new script.
2022
2023 2012-12-18  Dominic Mazzoni  <dmazzoni@google.com>
2024
2025         AX: support clickPoint in DRT for chromium
2026         https://bugs.webkit.org/show_bug.cgi?id=97359
2027
2028         Reviewed by Chris Fleizach.
2029
2030         Implement clickPointX, clickPointY, and elementAtPoint for Chromium, so that
2031         several more tests can pass.
2032
2033         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
2034         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
2035         (WebTestRunner::AccessibilityUIElement::clickPointXGetterCallback):
2036         (WebTestRunner):
2037         (WebTestRunner::AccessibilityUIElement::clickPointYGetterCallback):
2038         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
2039         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
2040         (AccessibilityUIElement):
2041
2042 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
2043
2044         Build fix after r138045.
2045
2046         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2047         (PerfTestsRunner._collect_tests): Use relative_path. We skip only when this path
2048         is not explicitly specified. The condition was backwards but never matched so we
2049         always ran the test regardless of whether it's listed in the skipped list or not.
2050         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
2051         (test_collect_tests_with_ignored_skipped_list): Renamed so that it actually runs.
2052
2053 2012-12-18  Emil A Eklund  <eae@chromium.org>
2054
2055         TestWebKitAPI/Tests/WebCore/LayoutUnit fails on mac
2056         https://bugs.webkit.org/show_bug.cgi?id=105332
2057         
2058         Reviewed by Levi Weintraub.
2059         
2060         Create new test suite for LayoutUnit tests as the KURL tests are
2061         of the incompatible TEST_F type.
2062
2063         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp:
2064         (TestWebKitAPI):
2065         (TestWebKitAPI::TEST_F):
2066
2067 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
2068
2069         Running a skipped test with run-perf-tests could alert the user
2070         https://bugs.webkit.org/show_bug.cgi?id=105219
2071
2072         Reviewed by Eric Seidel.
2073
2074         Force run skipped performance tests explicitly specified in the command line.
2075         Also warn the user when the specified paths don't exist.
2076
2077         * Scripts/webkitpy/layout_tests/port/base.py:
2078         (Port.relative_test_filename):
2079         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2080         (PerfTestsRunner._parse_args):
2081         (PerfTestsRunner._collect_tests):
2082         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
2083         (test_collect_tests_with_skipped_list_and_files):
2084
2085 2012-12-18  Yaron Friedman  <yfriedman@chromium.org>
2086
2087         Chrome Android java output is too verbose
2088         https://bugs.webkit.org/show_bug.cgi?id=105247
2089
2090         Reviewed by Tony Chang.
2091
2092         Pass -quiet to ant to suppress extraneous output.
2093
2094         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
2095         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
2096
2097 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
2098
2099         Consider removing --pause-before-testing option
2100         https://bugs.webkit.org/show_bug.cgi?id=105248
2101
2102         Reviewed by Eric Seidel.
2103
2104         Remove --pause-before-testing option.
2105
2106         * Tools/Scripts/webkitpy/layout_tests/port/driver.py:
2107         (DriverProxy.start): Removed.
2108         * Scripts/webkitpy/performance_tests/perftest.py:
2109         (PerfTest.__init__):
2110         (PerfTest.test_path): Renamed from path_or_url.
2111         (PerfTest.run):
2112         (PerfTest): Removed "driver" argument.
2113         (PerfTest._run_with_driver):
2114         (PerfTest.run_single):
2115         (ChromiumStylePerfTest.__init__):
2116         (PageLoadingPerfTest.__init__):
2117         (PageLoadingPerfTest.run_single):
2118         (PageLoadingPerfTest._run_with_driver): Renamed from run so that PerfTest.run can create a driver for us.
2119         (ReplayPerfTest.__init__):
2120         (ReplayPerfTest.prepare):
2121         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
2122         (TestPageLoadingPerfTest.test_run):
2123         (TestPageLoadingPerfTest.test_run_with_memory_output):
2124         (TestPageLoadingPerfTest.test_run_with_bad_output):
2125         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2126         (PerfTestsRunner._parse_args):
2127         (PerfTestsRunner._run_tests_set): Removed the code to create a driver and pause before running the test.
2128         (PerfTestsRunner._run_single_test): Removed "driver" argument.
2129         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
2130         (run_test):
2131         (test_run_test_pause_before_testing): Removed.
2132
2133 2012-12-18  Alberto Garcia  <agarcia@igalia.com>
2134
2135         [GTK] Tests that time out are not considered as failures.
2136         https://bugs.webkit.org/show_bug.cgi?id=105318
2137
2138         Reviewed by Martin Robinson.
2139
2140         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2141         (RunGtkAPITests.commandComplete):
2142         Consider log lines that end with TIMEOUT as errors.
2143
2144         * Scripts/run-gtk-tests:
2145         (TestRunner.run_tests):
2146         Consider timeouts as failures when returning the number of failed
2147         tests.
2148
2149 2012-12-18  Pravin D  <pravin.d@samsung.com>
2150
2151         Unreviewed, adding secondary email.
2152
2153         * Scripts/webkitpy/common/config/committers.py:
2154
2155 2012-12-18  Pravin D  <pravin.d@samsung.com>
2156
2157         Unreviewed, promoting myself to committer's list.
2158
2159         * Scripts/webkitpy/common/config/committers.py:
2160
2161 2012-12-18  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2162
2163         [EFL] ImageDiff does not build on WebKit2 only builds
2164         https://bugs.webkit.org/show_bug.cgi?id=105288
2165
2166         Reviewed by Kenneth Rohde Christiansen.
2167
2168         Moved ImageDiff to the efl/ folder inside Tools/ (so we don't pollute
2169         the folder with things which are EFL-specific.
2170
2171         ImageDiff will be built by default now regardless of WebKit API flavor
2172         you choose.
2173
2174         * CMakeLists.txt:
2175         * DumpRenderTree/efl/CMakeLists.txt:
2176         * efl/ImageDiff/CMakeLists.txt: Added.
2177         * efl/ImageDiff/ImageDiff.cpp: Renamed from Tools/DumpRenderTree/efl/ImageDiff.cpp.
2178
2179 2012-12-18  Kenneth Rohde Christiansen  <kenneth@webkit.org>
2180
2181         [EFL][WK2] When creating new windows, use the opener's context
2182         https://bugs.webkit.org/show_bug.cgi?id=105283
2183
2184         Reviewed by Laszlo Gombos.
2185
2186         Use the context of the opener if exists.
2187
2188         * MiniBrowser/efl/main.c:
2189         (on_key_down):
2190         (on_window_create):
2191         (window_create):
2192         (elm_main):
2193
2194 2012-12-18  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2195
2196         [EFL][WK2] Gardening of TestWebKitAPI tests
2197         https://bugs.webkit.org/show_bug.cgi?id=105268
2198
2199         Reviewed by Kenneth Rohde Christiansen.
2200
2201         Some tests were marked as failing but are passing and the newly added
2202         LayoutUnit should have it's own binary because it conflicts with the
2203         fixture used by KURL and is failing on the bots because of that.
2204
2205         * TestWebKitAPI/CMakeLists.txt:
2206         * TestWebKitAPI/PlatformEfl.cmake:
2207
2208 2012-12-18  Jochen Eisinger  <jochen@chromium.org>
2209
2210         [chromium] move webintent related methods to TestRunner library
2211         https://bugs.webkit.org/show_bug.cgi?id=105164
2212
2213         Reviewed by Darin Fisher.
2214
2215         Also deleted the unused deliveredIntentFailure and
2216         deliveredIntentResult methods.
2217
2218         * DumpRenderTree/chromium/DRTTestRunner.cpp:
2219         (DRTTestRunner::DRTTestRunner):
2220         * DumpRenderTree/chromium/DRTTestRunner.h:
2221         (DRTTestRunner):
2222         * DumpRenderTree/chromium/TestRunner/public/WebTestDelegate.h:
2223         (WebKit):
2224         (WebTestRunner::WebTestDelegate::setCurrentWebIntentRequest):
2225         (WebTestRunner::WebTestDelegate::currentWebIntentRequest):
2226         * DumpRenderTree/chromium/TestRunner/public/WebTestProxy.h:
2227         (WebKit):
2228         (WebTestProxyBase):
2229         (WebTestRunner::WebTestProxy::registerIntentService):
2230         (WebTestRunner::WebTestProxy::dispatchIntent):
2231         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
2232         (WebTestRunner::TestRunner::TestRunner):
2233         (WebTestRunner::TestRunner::sendWebIntentResponse):
2234         (WebTestRunner):
2235         (WebTestRunner::TestRunner::deliverWebIntent):
2236         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
2237         (TestRunner):
2238         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
2239         (WebTestRunner::WebTestProxyBase::postAccessibilityNotification):
2240         (WebTestRunner::WebTestProxyBase::registerIntentService):
2241         (WebTestRunner):
2242         (WebTestRunner::WebTestProxyBase::dispatchIntent):
2243         * DumpRenderTree/chromium/WebViewHost.cpp:
2244         (WebViewHost::setCurrentWebIntentRequest):
2245         (WebViewHost::currentWebIntentRequest):
2246         * DumpRenderTree/chromium/WebViewHost.h:
2247         (WebKit):
2248         (WebViewHost):
2249
2250 2012-12-18  Jochen Eisinger  <jochen@chromium.org>
2251
2252         [chromium] move editing callback dumping to the WebTestProxy
2253         https://bugs.webkit.org/show_bug.cgi?id=105013
2254
2255         Reviewed by Darin Fisher.
2256
2257         To decide whether or not to dump the editing callbacks, the WebTestProxy
2258         needs to query the TestRunner. For this purpose, I introduce a
2259         WebTestRunner class that just exposes the shouldDumpEditingCallbacks
2260         method. In contrast to e.g. WebEventSender, the WebTestRunner is just
2261         an interface, and it is not owned by the WebTestInterfaces, but the
2262         embedder of the TestRunner library must provide it. This will change
2263         once all TestRunner methods are in the library.
2264
2265         * DumpRenderTree/DumpRenderTree.gypi:
2266         * DumpRenderTree/chromium/DRTTestRunner.cpp:
2267         (DRTTestRunner::DRTTestRunner):
2268         (DRTTestRunner::reset):
2269         * DumpRenderTree/chromium/DRTTestRunner.h:
2270         (DRTTestRunner):
2271         (DRTTestRunner::taskList):
2272         (DRTTestRunner::WorkQueue::taskList):
2273         * DumpRenderTree/chromium/TestRunner/public/WebTestInterfaces.h:
2274         (WebTestRunner):
2275         (WebTestInterfaces):
2276         * DumpRenderTree/chromium/TestRunner/public/WebTestProxy.h:
2277         (WebKit):
2278         (WebTestRunner):
2279         (WebTestProxyBase):
2280         (WebTestRunner::WebTestProxy::shouldBeginEditing):
2281         (WebTestRunner::WebTestProxy::shouldEndEditing):
2282         (WebTestRunner::WebTestProxy::shouldInsertNode):
2283         (WebTestRunner::WebTestProxy::shouldInsertText):
2284         (WebTestRunner::WebTestProxy::shouldChangeSelectedRange):
2285         (WebTestRunner::WebTestProxy::shouldDeleteRange):
2286         (WebTestRunner::WebTestProxy::shouldApplyStyle):
2287         (WebTestRunner::WebTestProxy::didBeginEditing):
2288         (WebTestRunner::WebTestProxy::didChangeSelection):
2289         (WebTestRunner::WebTestProxy::didChangeContents):
2290         (WebTestRunner::WebTestProxy::didEndEditing):
2291         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h: Copied from Tools/DumpRenderTree/chromium/TestRunner/public/WebTestInterfaces.h.
2292         (WebTestRunner):
2293         (WebTestRunner::WebTestRunner::shouldDumpEditingCallbacks):
2294         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
2295         (WebTestRunner::TestRunner::TestRunner):
2296         (WebTestRunner::TestRunner::~TestRunner):
2297         (WebTestRunner):
2298         (WebTestRunner::TestRunner::reset):
2299         (WebTestRunner::TestRunner::shouldDumpEditingCallbacks):
2300         (WebTestRunner::TestRunner::dumpEditingCallbacks):
2301         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
2302         (TestRunner):
2303         * DumpRenderTree/chromium/TestRunner/src/WebTestInterfaces.cpp:
2304         (WebTestRunner::WebTestInterfaces::Internal::testRunner):
2305         (WebTestRunner::WebTestInterfaces::Internal::setTestRunner):
2306         (WebTestInterfaces::Internal):
2307         (WebTestRunner::WebTestInterfaces::Internal::Internal):
2308         (WebTestRunner::WebTestInterfaces::testRunner):
2309         (WebTestRunner):
2310         (WebTestRunner::WebTestInterfaces::setTestRunner):
2311         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
2312         (WebTestRunner::WebTestProxyBase::postAccessibilityNotification):
2313         (WebTestRunner):
2314         (WebTestRunner::WebTestProxyBase::shouldBeginEditing):
2315         (WebTestRunner::WebTestProxyBase::shouldEndEditing):
2316         (WebTestRunner::WebTestProxyBase::shouldInsertNode):
2317         (WebTestRunner::WebTestProxyBase::shouldInsertText):
2318         (WebTestRunner::WebTestProxyBase::shouldChangeSelectedRange):
2319         (WebTestRunner::WebTestProxyBase::shouldDeleteRange):
2320         (WebTestRunner::WebTestProxyBase::shouldApplyStyle):
2321         (WebTestRunner::WebTestProxyBase::didBeginEditing):
2322         (WebTestRunner::WebTestProxyBase::didChangeSelection):
2323         (WebTestRunner::WebTestProxyBase::didChangeContents):
2324         (WebTestRunner::WebTestProxyBase::didEndEditing):
2325         * DumpRenderTree/chromium/TestShell.cpp:
2326         (TestShell::initialize):
2327         (TestShell::createNewWindow):
2328         * DumpRenderTree/chromium/WebViewHost.cpp:
2329         (WebViewHost::shouldBeginEditing):
2330         (WebViewHost::shouldEndEditing):
2331         (WebViewHost::shouldInsertNode):
2332         (WebViewHost::shouldInsertText):
2333         (WebViewHost::shouldChangeSelectedRange):
2334         (WebViewHost::shouldDeleteRange):
2335         (WebViewHost::shouldApplyStyle):
2336         * DumpRenderTree/chromium/WebViewHost.h:
2337         (WebViewHost):
2338
2339 2012-12-17  Jochen Eisinger  <jochen@chromium.org>
2340
2341         [chromium] Fix methods in TestRunner that directly modify the settings instead of going through WebPreferences
2342         https://bugs.webkit.org/show_bug.cgi?id=105154
2343
2344         Reviewed by Tony Chang.
2345
2346         If the settings are modified directly, the embedder can't keep track of
2347         them and might accidentially override the changed settings.
2348
2349         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
2350         (WebPreferences):
2351         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
2352         (WebTestRunner::TestRunner::TestRunner):
2353         (WebTestRunner::TestRunner::setAsynchronousSpellCheckingEnabled):
2354         (WebTestRunner):
2355         (WebTestRunner::TestRunner::setMinimumTimerInterval):
2356         (WebTestRunner::TestRunner::setTouchDragDropEnabled):
2357         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
2358         (TestRunner):
2359         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
2360         (WebTestRunner::WebPreferences::reset):
2361         (WebTestRunner::WebPreferences::applyTo):
2362
2363 2012-12-17  Julien Chaffraix  <jchaffraix@webkit.org>
2364
2365         Unreviewed unit test skipping.
2366
2367         Filed https://bugs.webkit.org/show_bug.cgi?id=105253 to re-enable them.
2368
2369         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
2370         (TestWebKitAPI):
2371         (TestWebKitAPI::TEST):
2372
2373 2012-12-17  Dean Jackson  <dino@apple.com>
2374
2375         Unreviewed. Add Silvia Pfeiffer to contributor list.
2376
2377         * Scripts/webkitpy/common/config/committers.py:
2378
2379 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
2380
2381         build.webkit.org - add more google windows bots
2382         https://bugs.webkit.org/show_bug.cgi?id=105225
2383
2384         Reviewed by Eric Seidel.
2385
2386         Adds two more windows bots (google-windows-3, a builder, and
2387         google-windows-4, a tester) for the Chromium configs.
2388
2389         * BuildSlaveSupport/build.webkit.org-config/config.json:
2390
2391 2012-12-17  Eric Seidel  <eric@webkit.org>
2392
2393         Cleanup --profiler=perf code a little and allow up to 1000 report files (instead of 100)
2394         https://bugs.webkit.org/show_bug.cgi?id=105217
2395
2396         Reviewed by Dirk Pranke.
2397
2398         Previously --profile would fail strangely if find_unused_filename returned None
2399         (which would happen as soon as we had 100 report files).
2400         Now we'll search through 1000 filenames before giving up (and assert if we failed to find one).
2401         I also made some minor cleanups to the output of perf, and tried to make the
2402         commands more self-documenting by using long-names for args.
2403
2404         * Scripts/webkitpy/common/system/profiler.py:
2405         (SingleFileOutputProfiler.__init__):
2406         (Perf.attach_to_pid):
2407         (Perf.profile_after_exit):
2408
2409 2012-12-17  Sheriff Bot  <webkit.review.bot@gmail.com>
2410
2411         Unreviewed, rolling out r137198.
2412         http://trac.webkit.org/changeset/137198
2413         https://bugs.webkit.org/show_bug.cgi?id=105212
2414
2415         This patch is causing API behavior compatibility problems
2416         (Requested by zdobersek on #webkit).
2417
2418         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2419         (willSendRequestCallback):
2420
2421 2012-12-17  Elliott Sprehn  <esprehn@chromium.org>
2422
2423         Unreviewed. Adding myself to committers list.
2424
2425         * Scripts/webkitpy/common/config/committers.py:
2426
2427 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
2428
2429         nrwt: shuffle code around for cleanup in run_webkit_tests.py
2430         https://bugs.webkit.org/show_bug.cgi?id=105078
2431
2432         Reviewed by Ojan Vafai.
2433
2434         This patch reorders functions so that run_webkit_tests can be
2435         understood in a top-down matter better, and to make things slightly
2436         more sensible when I merge manager.py into it.
2437
2438         Also, this adds tests for the actual main() routine, which had several
2439         bugs get through testing a week or two ago.
2440
2441         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2442         (main):
2443         (parse_args):
2444         (_set_up_derived_options):
2445         (run):
2446         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2447         (RunTest):
2448         (RunTest.test_no_http_tests):
2449         (PortTest.disabled_test_mac_lion):
2450         (MainTest):
2451         (MainTest.test_exception_handling):
2452         (MainTest.test_exception_handling.interrupting_run):
2453         (MainTest.test_exception_handling.successful_run):
2454         (MainTest.test_exception_handling.successful_run.FakeRunDetails):
2455         (MainTest.test_exception_handling.exception_raising_run):
2456
2457 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
2458
2459         webkitpy: move --lint-test-files code into its own module
2460         https://bugs.webkit.org/show_bug.cgi?id=105077
2461
2462         Reviewed by Eric Seidel.
2463
2464         It really didn't need to be in run_webkit_tests.py anymore.
2465         This patch adds a new lint-test-expectations scripts to Tools/Scripts;
2466         rwt --lint-test-files is still supported for backwards compatibility
2467         for now.
2468
2469         * Scripts/lint-test-expectations: Added.
2470         * Scripts/webkitpy/layout_tests/lint_test_expectations.py: Added.
2471         (lint):
2472         (main):
2473         * Scripts/webkitpy/layout_tests/lint_test_expectations_unittest.py: Added.
2474         (FakePort):
2475         (FakePort.__init__):
2476         (FakePort.test_configuration):
2477         (FakePort.expectations_dict):
2478         (FakePort.skipped_layout_tests):
2479         (FakePort.all_test_configurations):
2480         (FakePort.configuration_specifier_macros):
2481         (FakePort.get_option):
2482         (FakeFactory):
2483         (FakeFactory.__init__):
2484         (FakeFactory.get):
2485         (FakeFactory.all_port_names):
2486         (LintTest):
2487         (LintTest.test_all_configurations):
2488         (LintTest.test_lint_test_files):
2489         (LintTest.test_lint_test_files__errors):
2490         (MainTest):
2491         (MainTest.test_success):
2492         (MainTest.test_success.interrupting_lint):
2493         (MainTest.test_success.successful_lint):
2494         (MainTest.test_success.exception_raising_lint):
2495         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2496         (main):
2497         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2498         (StreamTestingMixin.assertNotEmpty):
2499
2500 2012-12-17  Adam Klein  <adamk@chromium.org>
2501
2502         build-webkit: rename --template-tag to --template-element to match ENABLE #define name
2503         https://bugs.webkit.org/show_bug.cgi?id=105072
2504
2505         Reviewed by Laszlo Gombos.
2506
2507         * Scripts/webkitperl/FeatureList.pm:
2508
2509 2012-12-17  Julien Chaffraix  <jchaffraix@webkit.org>
2510
2511         Add some unit testing for WTF::clampTo* functions
2512         https://bugs.webkit.org/show_bug.cgi?id=105060
2513
2514         Reviewed by Emil A Eklund.
2515
2516         The new tests are targeted at clamping to int / unsigned and float as those are
2517         the one explicitely exposed through more dedicated functions (like clampToInteger,
2518         clampToPositiveInteger and clampToFloat).
2519
2520         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
2521
2522 2012-12-17  Eric Seidel  <eric@webkit.org>
2523
2524         rpt --profile --chromium-android throws exception
2525         https://bugs.webkit.org/show_bug.cgi?id=105061
2526
2527         Reviewed by Daniel Bates.
2528
2529         Somehow popen(args) got turned into popen(*args) at some point,
2530         which caused my new stringify code to fail when args was passed by name.
2531         Fixed and tested.
2532
2533         * Scripts/webkitpy/common/system/executive.py:
2534         (Executive._stringify_args):
2535         (Executive):
2536         (Executive.popen):
2537         * Scripts/webkitpy/common/system/executive_unittest.py:
2538         (ExecutiveTest.test_popen_args):
2539
2540 2012-12-17  Mark Pilgrim  <pilgrim@chromium.org>
2541
2542         Use Platform::current() instead of webKitPlatformSupport() in DumpRenderTree
2543         https://bugs.webkit.org/show_bug.cgi?id=105054
2544
2545         Reviewed by Jochen Eisinger.
2546
2547         Part of a refactoring series; see tracking bug 82948.
2548
2549         * DumpRenderTree/chromium/NotificationPresenter.cpp:
2550         (NotificationPresenter::show):
2551         * DumpRenderTree/chromium/Task.cpp:
2552         (postTask):
2553         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp:
2554
2555 2012-12-17  Emil A Eklund  <eae@chromium.org>
2556
2557         Clamp values in LayoutUnit::operator/ when SATURATED_LAYOUT_ARITHMETIC is enabled
2558         https://bugs.webkit.org/show_bug.cgi?id=104955
2559
2560         Reviewed by Julien Chaffraix.
2561
2562         Add tests for LayoutUnit.
2563
2564         * TestWebKitAPI/CMakeLists.txt:
2565         * TestWebKitAPI/ForwardingHeaders: Added.
2566         * TestWebKitAPI/ForwardingHeaders/WebCore: Added.
2567         * TestWebKitAPI/ForwardingHeaders/WebCore/LayoutUnit.h: Added.
2568         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
2569         * TestWebKitAPI/TestWebKitAPI.gypi:
2570         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
2571         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp: Added.
2572         (TestWebKitAPI):
2573         (TestWebKitAPI::TEST):
2574         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
2575
2576 2012-12-17  Jochen Eisinger  <jochen@chromium.org>
2577
2578         [chromium] removed unused preferences from testRunner.overridePreference
2579         https://bugs.webkit.org/show_bug.cgi?id=105169
2580
2581         Reviewed by Tony Chang.
2582
2583         Most notably this removes the code related to font family maps which is
2584         now handled via window.internals.settings
2585
2586         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
2587         (WebPreferences):
2588         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
2589         (WebTestRunner::TestRunner::overridePreference):
2590         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
2591         (TestRunner):
2592         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
2593         (WebTestRunner::WebPreferences::applyTo):
2594
2595 2012-12-17  James Robinson  <jamesr@chromium.org>
2596
2597         [chromium] Move creation of WebLayerTreeView to WebWidgetClient
2598         https://bugs.webkit.org/show_bug.cgi?id=105071
2599
2600         Reviewed by Adrienne Walker.
2601
2602         Update WebViewHost for the new WebWidgetClient interface.
2603
2604         * DumpRenderTree/chromium/WebViewHost.cpp:
2605         (WebViewHost::initializeLayerTreeView):
2606         (WebViewHost::layerTreeView):
2607         (WebViewHost::~WebViewHost):
2608         * DumpRenderTree/chromium/WebViewHost.h:
2609         (WebViewHost):
2610
2611 2012-12-17  Kenneth Rohde Christiansen  <kenneth@webkit.org>
2612
2613         [EFL][WK2] window_create doesn't receive the url
2614         https://bugs.webkit.org/show_bug.cgi?id=105184
2615
2616         Reviewed by Alexis Menard.
2617
2618         * MiniBrowser/efl/main.c: Use the url supplied to window_create
2619         (on_window_create):
2620
2621 2012-12-17  Ilya Tikhonovsky  <loislo@chromium.org>
2622
2623         Web Inspector: Native Memory Instrumentation: MemoryInstrumentation doesn't detect reportMemoryUsage method defined in a base class.
2624         https://bugs.webkit.org/show_bug.cgi?id=105026
2625
2626         Reviewed by Yury Semikhatsky.
2627
2628         Old SFINAE test was replaced with new one based on this article:
2629         http://stackoverflow.com/questions/1966362/sfinae-to-check-for-inherited-member-functions
2630
2631         New test which covers this problem was added.
2632
2633         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
2634
2635 2012-12-17  Anthony Scian  <ascian@rim.com>
2636
2637         [Blackberry] Static code analysis warning fixes
2638         https://bugs.webkit.org/show_bug.cgi?id=105129
2639
2640         Reviewed by Rob Buis.
2641
2642         Added null checks for fopen calls, no recovery attempted; access fault prevention only.
2643
2644         * DumpRenderTree/blackberry/DumpRenderTree.cpp:
2645         (BlackBerry::WebKit::createFile):
2646         (BlackBerry::WebKit::DumpRenderTree::runTest):
2647
2648 2012-12-17  John Mellor  <johnme@chromium.org>
2649
2650         Add Anton Vayvod and Tim Volodine as contributors in committers.py
2651         https://bugs.webkit.org/show_bug.cgi?id=105025
2652
2653         Unreviewed. Adds two contributors to committers.py.
2654         They are initially contributing to the Text Autosizing feature (master bug: 84186).
2655
2656         * Scripts/webkitpy/common/config/committers.py:
2657
2658 2012-12-17  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
2659
2660         Unreviewed, add myself to committers.py.
2661
2662         * Scripts/webkitpy/common/config/committers.py:
2663
2664 2012-12-17  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
2665
2666         [Qt] remove some unnecessary CONFIG additions
2667
2668         Reviewed by Simon Hausmann.
2669
2670         qt is already added by spec_pre.prf, warn_on and depend_includepath by
2671         default_pre.prf.
2672
2673         * qmake/mkspecs/features/default_post.prf:
2674
2675 2012-12-17  Ilya Tikhonovsky  <loislo@chromium.org>
2676
2677         Unreviewed, rolling out r137892.
2678         http://trac.webkit.org/changeset/137892
2679         https://bugs.webkit.org/show_bug.cgi?id=105026
2680
2681         it broke compilation on windows
2682
2683         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
2684
2685 2012-12-14  Ilya Tikhonovsky  <loislo@chromium.org>
2686
2687         Web Inspector: Native Memory Instrumentation: MemoryInstrumentation doesn't detect reportMemoryUsage method defined in a base class.
2688         https://bugs.webkit.org/show_bug.cgi?id=105026
2689
2690         Reviewed by Yury Semikhatsky.
2691
2692         Old SFINAE test was replaced with new one based on this article:
2693         http://stackoverflow.com/questions/1966362/sfinae-to-check-for-inherited-member-functions
2694
2695         New test which covers this problem was added.
2696
2697         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
2698
2699 2012-12-17  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
2700
2701         [Qt] rely on automatic output directory setup
2702
2703         Reviewed by Simon Hausmann.
2704
2705         Now being a proper qt module (by virtue of having load(qt_build_config)
2706         in .qmake.conf), webkit gets the path setup goodies for free.
2707
2708         This also fixes Makefile.api.Debug/Release trying to generate qrc_WebKit.cpp
2709         at the same time, instead of the debug and release version ending up in different
2710         directories as they should.
2711
2712         * qmake/mkspecs/features/default_pre.prf:
2713
2714 2012-12-17  Charles Wei  <charles.wei@torchmobile.com.cn>
2715
2716         [BlackBerry] Need to re-enable video track
2717         https://bugs.webkit.org/show_bug.cgi?id=105155
2718
2719         Reviewed by George Staikos.
2720
2721         Just enable Video Track for BlackBerry.
2722
2723         * Scripts/webkitperl/FeatureList.pm:
2724
2725 2012-12-17  János Badics  <jbadics@inf.u-szeged.hu>
2726
2727         Unreviewed. Added myself to committers.py
2728
2729         * Scripts/webkitpy/common/config/committers.py:
2730
2731 2012-12-17  Jonathan Liu  <net147@gmail.com>
2732
2733         Fix linking of libraries with -Wl,-whole-archive
2734         https://bugs.webkit.org/show_bug.cgi?id=104436
2735
2736         Reviewed by Simon Hausmann.
2737
2738         If there is more than one library in LIBS linked with
2739         -Wl,-whole-archive -l... -Wl,-no-whole-archive, qmake tries to
2740         merge the duplicate flags in LIBS which causes linking to fail.
2741
2742         Add no_smart_library_merge to CONFIG to prevent the duplicate
2743         flags from being removed when using -Wl,-whole-archive.
2744
2745         * qmake/mkspecs/features/functions.prf:
2746
2747 2012-12-16  Arvid Nilsson  <anilsson@rim.com>
2748
2749         [BlackBerry] Adapt to new BlackBerry::Platform::TouchPoint API
2750         https://bugs.webkit.org/show_bug.cgi?id=105143
2751         RIM PR 171941
2752
2753         Reviewed by Rob Buis.
2754         Internally reviewed by George Staikos.
2755
2756         Adapt to new method names and encapsulation of TouchPoint data members.
2757
2758         * DumpRenderTree/blackberry/EventSender.cpp:
2759         (addTouchPointCallback):
2760         (updateTouchPointCallback):
2761         (touchEndCallback):
2762         (releaseTouchPointCallback):
2763         (sendTouchEvent):
2764
2765 2012-12-16  Maciej Stachowiak  <mjs@apple.com>
2766
2767         Add a script to compute the mean and 95% confidence interval (using two-sided t-test) of a sample set
2768         https://bugs.webkit.org/show_bug.cgi?id=105148
2769
2770         Reviewed by Filip Pizlo.
2771
2772         Usage examples:
2773         
2774         $ sampstat --help
2775         Usage: sampstat [options] [FILES]
2776           Compute the mean and 95% confidence interval of a sample set.
2777           Standard input or files must contain two or more decimal numbers, one per line.
2778         
2779         Options:
2780           -h, --help            show this help message and exit
2781           -u UNIT, --unit=UNIT  assume values are in units of UNIT
2782           -v, --verbose         print all values (with units)
2783         
2784         $ sampstat -u MB memresults.txt
2785         2356.90 MB +/- 101.34 MB (4.3%)
2786         
2787         $ sampstat -v -u MB memresults.txt
2788               2318.21 MB
2789               2399.56 MB
2790               2352.93 MB
2791         ----------------
2792         Mean: 2356.90 MB +/- 101.34 MB (4.3%)
2793         
2794
2795         * Scripts/sampstat: Added.
2796         (sum): Helper function to compute the sum of a list.
2797         (arithmeticMean): Compute the meam of a list.
2798         (standardDeviation): Compute the sample standard deviation (unbiased estimator).
2799         (standardError): Compute the sample standard error.
2800         (tDist): Compute t(.025, n-1), the t-value for atwo-sided 95% confidence interval. 
2801         (twoSidedConfidenceInterval): Compute the two-sided confidence interval range about the mean.
2802
2803 2012-12-15  Nima Ghanavatian  <nghanavatian@rim.com>
2804
2805         [BlackBerry] Pass key modifiers with touch and mouse events
2806         https://bugs.webkit.org/show_bug.cgi?id=105108
2807
2808         Reviewed by Rob Buis.
2809
2810         Update DRT to new api.
2811
2812         Internally reviewed by Gen Mak.
2813
2814         * DumpRenderTree/blackberry/EventSender.cpp:
2815         (mouseDownCallback):
2816         (mouseUpCallback):
2817         (mouseMoveToCallback):
2818
2819 2012-12-15  Zan Dobersek  <zandobersek@gmail.com>
2820
2821         Stop the XvfbDriver before proceeding with its starting setup
2822         https://bugs.webkit.org/show_bug.cgi?id=105006
2823
2824         Reviewed by Philippe Normand.
2825
2826         Just as with the base Driver implementation, stop the XvfbDriver
2827         before proceeding with setting up lock files and a new Xvfb instance.
2828         This forces any driver or Xvfb instances that are currently under
2829         the driver's control to be closed/killed and cleaned up after.
2830
2831         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
2832         (XvfbDriver._start):
2833
2834 2012-12-14  Dirk Pranke  <dpranke@chromium.org>
2835
2836         webkitpy: make port_name a required parameter to the non-test ports
2837         https://bugs.webkit.org/show_bug.cgi?id=105057
2838
2839         Reviewed by Eric Seidel.
2840
2841         This is just minor code cleanup along the path to making Port
2842         constructors simpler and pushing more logic into the port factory.
2843
2844         * Scripts/webkitpy/layout_tests/port/base.py:
2845         (Port.__init__):
2846         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
2847         (PortTest.make_port):
2848         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
2849         (TestWebKitPort.__init__):
2850         * Scripts/webkitpy/layout_tests/port/test.py:
2851         (TestPort):
2852         (TestPort.determine_full_port_name):
2853         (TestPort.__init__):
2854         * Scripts/webkitpy/layout_tests/port/xvfbdriver_unittest.py:
2855         (XvfbDriverTest.make_driver):
2856         (XvfbDriverTest.test_stop):
2857
2858 2012-12-14  Dirk Pranke  <dpranke@chromium.org>
2859
2860         garden-o-matic doesn't know about reftests
2861         https://bugs.webkit.org/show_bug.cgi?id=101976
2862
2863         Reviewed by Eric Seidel.
2864
2865         Attempt to make garden-o-matic properly aware of reftests again.
2866
2867         This is re-landing r137407 with a change to
2868         rebaselineWithStatusUpdates() to actually pass the filtered
2869         failureInofLost to checkout.rebaseline().
2870
2871         Unfortunately, the testing coverage for this whole module is almost
2872         non-existent, and it's not obvious that there's a good way to
2873         stub out checkout.rebaseline() to test that the right thing
2874         happens. I verified the change manually.
2875
2876         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
2877         (.):
2878         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
2879         (.):
2880         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
2881
2882 2012-12-14  Ryosuke Niwa  <rniwa@webkit.org>
2883
2884         "Running 1 DumpRenderTree over X shards" is not a helpful output
2885         https://bugs.webkit.org/show_bug.cgi?id=104858
2886
2887         Reviewed by Dirk Pranke.
2888
2889         The word "shard" is not typically used in test frameworks. This word is confusing for many WebKit
2890         contributors who typically don't have much experience on distributed computations and databases.
2891         So only print this message out in the debug mode.
2892
2893         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2894         * Scripts/webkitpy/layout_tests/views/printing.py:
2895         (Printer.print_workers_and_shards):
2896
2897 2012-12-13  Eric Seidel  <eric@webkit.org>
2898
2899         Add --profiler=PROFILER option to run-perf-tests to allow specifying which profiler to use on platforms with many
2900         https://bugs.webkit.org/show_bug.cgi?id=104891
2901
2902         Reviewed by Ryosuke Niwa.
2903
2904         I also implemented a very simple "Sample" Profiler using
2905         Mac OS X's /usr/bin/sample command line tool.
2906
2907         The real reason for this abstraction is to make it easy
2908         to support both perf and pprof on linux which seem to
2909         be about equally popular among those I ask in the Chrome team.
2910
2911         * Scripts/webkitpy/common/system/profiler.py:
2912         (ProfilerFactory.create_profiler):
2913         (ProfilerFactory):
2914         (ProfilerFactory.available_profilers_by_name):
2915         (ProfilerFactory.default_profiler_name):
2916         (Sample):
2917         (Sample.__init__):
2918         (Sample.attach_to_pid):
2919         (Sample.profile_after_exit):
2920         (IProfiler.attach_to_pid):
2921         * Scripts/webkitpy/layout_tests/port/driver.py:
2922         (Driver.__init__):
2923         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2924         (PerfTestsRunner._parse_args):
2925
2926 2012-12-13  Eric Seidel  <eric@webkit.org>
2927
2928         Unreviewed, fixing typo in python unittest.
2929
2930         I changed the default profiler on linux from pprof to perf in
2931         bug 104971.  I failed to update the unittest results at that time.
2932
2933         * Scripts/webkitpy/common/system/profiler_unittest.py:
2934         (ProfilerFactoryTest.test_basic):
2935
2936 2012-12-13  Ian Vollick  <vollick@chromium.org>
2937
2938         [chromium] Add a virtual test suite for enabling opt-in to composited scrolling
2939         https://bugs.webkit.org/show_bug.cgi?id=104911
2940
2941         Reviewed by Dirk Pranke.
2942
2943         Adds two virtual tests suites to run the tests in
2944         compositing/overflow/ and scrollbars/ with the flag
2945         --enable-accelerated-overflow-scroll, which I've also plumbed through
2946         in the usual way.
2947
2948         * DumpRenderTree/chromium/DumpRenderTree.cpp:
2949         (main):
2950         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
2951         (WebPreferences):
2952         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
2953         (WebTestRunner::WebPreferences::reset):
2954         (WebTestRunner::WebPreferences::applyTo):
2955         * DumpRenderTree/chromium/TestShell.cpp:
2956         (TestShell::TestShell):
2957         (TestShell::resetWebSettings):
2958         * DumpRenderTree/chromium/TestShell.h:
2959         (TestShell::setAcceleratedCompositingForOverflowScrollEnabled):
2960         (TestShell):
2961         * Scripts/webkitpy/layout_tests/port/chromium.py:
2962         (ChromiumPort.virtual_test_suites):
2963
2964 2012-12-13  Eric Seidel  <eric@webkit.org>
2965
2966         Use 'perf' to profile on linux instead of google-pprof by default
2967         https://bugs.webkit.org/show_bug.cgi?id=104971
2968
2969         Reviewed by Dirk Pranke.
2970
2971         This makes Chromium Linux match Chromium Android and use the perf
2972         tool by default.  Once https://bugs.webkit.org/show_bug.cgi?id=104891
2973         lands it will be possible to use pprof again on Linux.
2974
2975         This is slightly more advanced than the AndroidPerf profiler as
2976         instead of using a timeout on "perf record" I instead watch
2977         for the termination of the target process and then control-C
2978         the 'perf record' process.  This required me to add two new
2979         methods to Executive, one to have a limited-time wait() and
2980         the second to be able to send a control-C.  I chose to add
2981         these to Executive to make them easier to mock/fix-for-win32
2982         at a later time if needed.
2983
2984         * Scripts/webkitpy/common/system/executive.py:
2985         (Executive.wait_limited):
2986         (Executive.interrupt):
2987         * Scripts/webkitpy/common/system/profiler.py:
2988         (ProfilerFactory.create_profiler):
2989         (Perf):
2990         (Perf.__init__):
2991         (Perf._perf_path):
2992         (Perf.attach_to_pid):
2993         (Perf._first_ten_lines_of_profile):
2994         (Perf.profile_after_exit):
2995
2996 2012-12-13  Julie Parent  <jparent@chromium.org>
2997
2998         Dashboard cleanup: remove usage of global g_builders.
2999         https://bugs.webkit.org/show_bug.cgi?id=104941
3000
3001         Reviewed by Dirk Pranke.
3002
3003         The dashboards use a lot of global state, which makes hacking on them
3004         complicated. This change removes the use of one such global: g_builders.
3005         In most cases, we can just use currentBuilderGroup().builders instead,
3006         which is now currentBuilders().
3007         Surprisingly, the most changes were required to the unit tests, since
3008         they were even bigger offenders of bad hygiene, relying on global state
3009         set by other tests, randomly clobbering global variables in ways the
3010         real code doesn't, etc.
3011
3012         * TestResultServer/static-dashboards/builders.js:
3013         (BuilderGroup.prototype.setup):
3014         * TestResultServer/static-dashboards/dashboard_base.js:
3015         (.switch.return):
3016         (htmlForTestTypeSwitcher):
3017         * TestResultServer/static-dashboards/flakiness_dashboard.js:
3018         (generatePage):
3019         (getAllTestsTrie):
3020         (processTestRunsForAllBuilders):
3021         (showPopupForBuild):
3022         (generatePageForExpectationsUpdate):
3023         (loadExpectationsLayoutTests):
3024         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
3025         (resetGlobals):
3026         (stubResultsByBuilder):
3027         (test):
3028         * TestResultServer/static-dashboards/loader.js:
3029         (.):
3030         * TestResultServer/static-dashboards/loader_unittests.js:
3031
3032 2012-12-13  Eric Seidel  <eric@webkit.org>
3033
3034         Callers should not have to stringify args before calling Executive run_command/popen
3035         https://bugs.webkit.org/show_bug.cgi?id=104975
3036
3037         Reviewed by Dirk Pranke.
3038
3039         One could argue that we should match the python call syntax here,
3040         but I think it's a more friendly API if we automagically handle
3041         stringification of args in run_command, etc.
3042         This removes map(unicode, args) from several callsites.
3043
3044         When I first tried to land this change, I didn't realize that
3045         Executive._command_for_printing depended on this behavior
3046         having been applied to args in run_command.  The fix is to
3047         call _stringify_args in both run_command and popen.
3048         This is slightly redundant, but given how short args have to be
3049         (due to shell limits), I don't think the double-encode check
3050         matters in practice.
3051
3052         This is slightly complicated by the fact that apache_http_server.py
3053         is the one caller in our codebase which uses shell=True.
3054         shell=True is a well-documented trail-of-tears:
3055         http://stackoverflow.com/questions/3172470/actual-meaning-of-shell-true-in-subprocess
3056         but to support this legacy (windows-only) code (which I can't easily test)
3057         I've added an if-hack to avoid stringifying the the popen(shell=True) case.
3058
3059         * Scripts/webkitpy/common/system/executive.py:
3060         (Executive.run_command):
3061         (Executive._stringify_args):
3062         (Executive.popen):
3063         * Scripts/webkitpy/common/system/executive_unittest.py:
3064         (ExecutiveTest.test_auto_stringify_args):
3065         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
3066         (attach_to_pid):
3067
3068 2012-12-13  Adrienne Walker  <enne@chromium.org>
3069
3070         Unreviewed, rolling out r137645, r137646, and r137667.
3071         http://trac.webkit.org/changeset/137645
3072         http://trac.webkit.org/changeset/137646
3073         http://trac.webkit.org/changeset/137667
3074         https://bugs.webkit.org/show_bug.cgi?id=104911
3075
3076         Breaks some overflow layout tests
3077
3078         * Scripts/webkitpy/layout_tests/port/chromium.py:
3079         (ChromiumPort.virtual_test_suites):
3080
3081 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
3082
3083         Unreviewed, rolling out r137661.
3084         http://trac.webkit.org/changeset/137661
3085         https://bugs.webkit.org/show_bug.cgi?id=104891
3086
3087         broke unit tests, run-webkit-tests
3088
3089         * Scripts/webkitpy/common/system/executive.py:
3090         (Executive._run_command_with_teed_output):
3091         (Executive.run_command):
3092         (Executive.popen):
3093         * Scripts/webkitpy/common/system/profiler.py:
3094         (ProfilerFactory.create_profiler):
3095         (GooglePProf.profile_after_exit):
3096         (IProfiler.attach_to_pid):
3097         * Scripts/webkitpy/layout_tests/port/driver.py:
3098         (Driver.__init__):
3099         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
3100         (PerfTestsRunner._parse_args):
3101
3102 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
3103
3104         nrwt: move rundetails, summarize_results into test_run_results
3105         https://bugs.webkit.org/show_bug.cgi?id=104963
3106
3107         Reviewed by Eric Seidel.
3108
3109         More cleanup; this the plain-old-data functions out of manager.py
3110         and into test_run_results next to the other aggregate data structures.
3111
3112         Also this cleans up a bunch of test code and reduces duplication.
3113
3114         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3115         (Manager.run):
3116         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
3117         (ManagerTest.test_look_for_new_crash_logs):
3118         * Scripts/webkitpy/layout_tests/models/test_run_results.py:
3119         (TestRunResults.__init__):
3120         (TestRunResults.add):
3121         (RunDetails):
3122         (RunDetails.__init__):
3123         (_interpret_test_failures):
3124         (summarize_results):
3125         * Scripts/webkitpy/layout_tests/models/test_run_results_unittest.py: Added.
3126         (get_result):
3127         (run_results):
3128         (summarized_results):
3129         (InterpretTestFailuresTest):
3130         (InterpretTestFailuresTest.setUp):
3131         (InterpretTestFailuresTest.test_interpret_test_failures):
3132         (SummarizedResultsTest):
3133         (SummarizedResultsTest.setUp):
3134         (SummarizedResultsTest.test_no_svn_revision):
3135         (SummarizedResultsTest.test_svn_revision):
3136         (SummarizedResultsTest.test_summarized_results_wontfix):
3137         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
3138         (BuildBotPrinterTests.get_printer):
3139         (BuildBotPrinterTests.test_print_unexpected_results):
3140         (BuildBotPrinterTests.test_print_results):
3141
3142 2012-12-13  Eric Seidel  <eric@webkit.org>
3143
3144         Make --profile output easier to understand for pprof on linux
3145         https://bugs.webkit.org/show_bug.cgi?id=104957
3146
3147         Reviewed by Dirk Pranke.
3148
3149         pprof is kinda mysterious, so we print some information
3150         on how to use it, and how to learn how to read the sample dump.
3151         This is similar to what we've done for perf output for Chromium Android.
3152
3153         * Scripts/webkitpy/common/system/profiler.py:
3154         (GooglePProf._pprof_path):
3155         (GooglePProf.profile_after_exit):
3156
3157 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
3158
3159         Revert r137407 - it didn't actually work right.
3160         https://bugs.webkit.org/show_bug.cgi?id=101976
3161
3162         Unreviewed, build fix.
3163
3164         Turns out that rebaselining was completely broken with this change :(
3165         We're passing in a list of test names rather than the list of
3166         failureInfo objects.
3167
3168         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
3169         (.):
3170         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
3171         (.):
3172         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
3173
3174 2012-12-13  Antoine Labour  <piman@chromium.org>
3175
3176         [chromium] Small fixes for WebPluginContainer::setWebLayer
3177         https://bugs.webkit.org/show_bug.cgi?id=104953
3178
3179         Reviewed by James Robinson.
3180
3181         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp: Clean up
3182         dangling pointers, force correct invalidation of the layer.
3183
3184 2012-12-13  Eric Seidel  <eric@webkit.org>
3185
3186         Add --profiler=PROFILER option to run-perf-tests to allow specifying which profiler to use on platforms with many
3187         https://bugs.webkit.org/show_bug.cgi?id=104891
3188
3189         Reviewed by Dirk Pranke.
3190
3191         I also implemented a very simple "Sample" Profiler using
3192         Mac OS X's /usr/bin/sample command line tool.
3193
3194         I also moved the map(unicode, args) call into Executive.popen
3195         so that all callers don't have to do it themselves.
3196
3197         The real reason for this abstraction is to make it easy
3198         to support both perf and pprof on linux which seem to
3199         be about equally popular among those I ask in the Chrome team.
3200
3201         * Scripts/webkitpy/common/system/executive.py:
3202         (Executive.popen):
3203         * Scripts/webkitpy/common/system/profiler.py:
3204         (ProfilerFactory.create_profiler):
3205         (ProfilerFactory):
3206         (ProfilerFactory.available_profilers_by_name):
3207         (ProfilerFactory.default_profiler_name):
3208         (Sample):
3209         (Sample.__init__):
3210         (Sample.attach_to_pid):
3211         (Sample.profile_after_exit):
3212         (IProfiler.attach_to_pid):
3213         * Scripts/webkitpy/layout_tests/port/driver.py:
3214         (Driver.__init__):
3215         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
3216         (PerfTestsRunner._parse_args):
3217
3218 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
3219
3220         support -wk2 port names properly in webkitpy.layout_tests.port
3221         https://bugs.webkit.org/show_bug.cgi?id=104761
3222
3223         Reviewed by Eric Seidel.
3224
3225         Our support for WK2 port names was inconsistent at best; this patch
3226         modifies the code so that <port_name>-wk2 is supported for all ports that
3227         have WK2 implementations and is equivalent to specifying <port_name>
3228         and -2/--webkit-test-runner. In addition, this modifies
3229         builders.all_port_names() to include the wk2 variants.
3230
3231         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
3232         * Scripts/webkitpy/layout_tests/port/apple.py:
3233         (ApplePort.determine_full_port_name):
3234         * Scripts/webkitpy/layout_tests/port/base.py:
3235         (Port.determine_full_port_name):
3236         (Port.__init__):
3237         * Scripts/webkitpy/layout_tests/port/builders.py:
3238         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
3239         (DriverTest.make_port):
3240         * Scripts/webkitpy/layout_tests/port/mac.py:
3241         (MacPort.default_baseline_search_path):
3242         * Scripts/webkitpy/layout_tests/port/win.py:
3243         (WinPort.default_baseline_search_path):
3244         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
3245         (TestRebaselineExpectations.test_rebaseline_expectations):
3246
3247 2012-12-13  Ian Vollick  <vollick@chromium.org>
3248
3249         [chromium] Add a virtual test suite for enabling opt-in to composited scrolling
3250         https://bugs.webkit.org/show_bug.cgi?id=104911
3251
3252         Reviewed by Adrienne Walker.
3253
3254         Adds two virtual tests suites to run the tests in
3255         compositing/overflow/ and scrollbars/ with the flag
3256         --enable-accelerated-compositing-for-overflow-scroll
3257
3258         * Scripts/webkitpy/layout_tests/port/chromium.py:
3259         (ChromiumPort.virtual_test_suites):
3260
3261 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
3262
3263         delete extraneous self.fail in mastercfg_unittest.py
3264         https://bugs.webkit.org/show_bug.cgi?id=104940
3265
3266         Reviewed by Eric Seidel.
3267
3268         This was accidentally introduced in r137404.
3269
3270         * BuildSlaveSupport/build.webkit.org-config/mastercfg_unittest.py:
3271
3272 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
3273
3274         Fix a mis-named variable after r137621.
3275         https://bugs.webkit.org/show_bug.cgi?id=104871
3276
3277         Unreviewed, build fix.
3278
3279         * Scripts/webkitpy/layout_tests/views/buildbot_results.py:
3280         (BuildBotPrinter.print_results):
3281         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
3282
3283 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
3284
3285         webkitpy: rename ResultSummary
3286         https://bugs.webkit.org/show_bug.cgi?id=104871
3287
3288         Reviewed by Eric Seidel.
3289
3290         This patch changes the names of ResultSummary and related modules,
3291         classes, and variables as a part of the larger clean up in bug 103824.
3292
3293         This change is basically a bunch of search and replaces:
3294         - result_summary.py -> test_run_results.py
3295         - class ResultSummary -> class TestRunResults
3296         - "result_summary" or "rs" as a generic variable name -> "run_results",
3297           when it doesn't matter (or we don't know) if this refers to the
3298           initial pass over the tests or the retry pass.
3299         - "result_summary" as a variable containing the results of the initial,
3300           complete test run -> "initial_results"
3301         - "retry_summary" as a variable containing the results of retried
3302           results -> "retry_results"
3303         - ResultSummary.results (the property) -> TestRunResults.results_by_name
3304         - ResultSummary.failures (the property) -> TestRunResults.failures_by_name
3305         - RunDetails.result_summary -> RunDetails.initial_results
3306         - RunDetails.retry_summary -> RunDetails.retry_results
3307
3308         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
3309         (LayoutTestRunner.__init__):
3310         (LayoutTestRunner.run_tests):
3311         (LayoutTestRunner._mark_interrupted_tests_as_skipped):
3312         (LayoutTestRunner._interrupt_if_at_failure_limits):
3313         (LayoutTestRunner._interrupt_if_at_failure_limits.interrupt_if_at_failure_limit):
3314         (LayoutTestRunner._update_summary_with_result):
3315         (LayoutTestRunner._handle_finished_test):
3316         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner_unittest.py:
3317         (FakePrinter.print_expected):
3318         (LayoutTestRunnerTests.test_interrupt_if_at_failure_limits):
3319         (LayoutTestRunnerTests.test_update_summary_with_result):
3320         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3321         (RunDetails.__init__):
3322         (summarize_results):
3323         (Manager.run):
3324         (Manager._look_for_new_crash_logs):
3325         (Manager._tests_to_retry):
3326         (Manager._upload_json_files):
3327         (Manager._stats_trie):
3328         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
3329         (ManagerTest.test_look_for_new_crash_logs):
3330         (SummarizeResultsTest):
3331         (SummarizeResultsTest.get_run_results):
3332         (SummarizeResultsTest.summarized_results):
3333         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
3334         (JSONLayoutResultsGenerator.__init__):
3335         (JSONLayoutResultsGenerator._get_modifier_char):
3336         (JSONLayoutResultsGenerator._insert_failure_summaries):
3337         (JSONLayoutResultsGenerator._get_failure_summary_entry):
3338         * Scripts/webkitpy/layout_tests/models/test_run_results.py: Renamed from Tools/Scripts/webkitpy/layout_tests/models/result_summary.py.
3339         (TestRunResults):
3340         (TestRunResults.__init__):
3341         (TestRunResults.add):
3342         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3343         (get_test_results):
3344         (MainTest.test_basic):
3345         * Scripts/webkitpy/layout_tests/views/buildbot_results.py:
3346         (BuildBotPrinter.print_results):
3347         (BuildBotPrinter.print_run_results):
3348         (BuildBotPrinter._print_run_results_entry):
3349         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
3350         (BuildBotPrinterTests.get_run_results):
3351         (BuildBotPrinterTests.test_print_unexpected_results.get_unexpected_results):
3352         (BuildBotPrinterTests):
3353         * Scripts/webkitpy/layout_tests/views/printing.py:
3354         (Printer.print_expected):
3355         (Printer._print_expected_results_of_type):
3356         (Printer.print_results):
3357         (Printer._print_timing_statistics):
3358         (Printer._print_worker_statistics):
3359         (Printer._print_aggregate_test_statistics):
3360         (Printer._print_individual_test_times):
3361         (Printer._print_directory_timings):
3362         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
3363
3364 2012-12-13  Thiago Marcos P. Santos  <thiago.santos@intel.com>
3365
3366         [EFL] The WebKit2 bots are building and executing WebKit1 tests
3367         https://bugs.webkit.org/show_bug.cgi?id=104690
3368
3369         Reviewed by Csaba Osztrogonác.
3370
3371         Pass --no-webkit1 option to the EFL WK2 bots when building. This
3372         will disable WK1 API and thus, the WK1 API tests.
3373
3374         * BuildSlaveSupport/build.webkit.org-config/config.json:
3375         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
3376         (CompileWebKit2Only):
3377         (BuildAndTestFactory):
3378         (BuildAndTestFactory.__init__):
3379         (BuildAndTestWebKit2OnlyFactory):
3380         Now bots using BuildAndTestWebKit2OnlyFactory (for instance only EFL) will
3381         get --no-webkit1 as a parameter on the build-webkit script.
3382         * Scripts/build-webkit:
3383         Added support for -no-webkit1 for the Qt and EFL port.
3384
3385 2012-12-13  Eric Seidel  <eric@webkit.org>
3386
3387         build-webkit --chromium-android forces Make even when update-webkit used --ninja
3388         https://bugs.webkit.org/show_bug.cgi?id=104839
3389
3390         Reviewed by Tony Chang.
3391
3392         This was recommended by Peter as part of:
3393         https://bugs.webkit.org/show_bug.cgi?id=104434#c4
3394         Fixing this makes build-webkit --chromium-android work for the common (ninja) case.
3395
3396         * Scripts/webkitdirs.pm:
3397         (buildChromium):
3398
3399 2012-12-13  Julie Parent  <jparent@chromium.org>
3400
3401         Loader_unittests should reset state between runs.
3402         https://bugs.webkit.org/show_bug.cgi?id=104862
3403
3404         Reviewed by Dirk Pranke.
3405
3406         The loader_unittests were not resetting state between runs, leaving
3407         a lot of data behind in global objects that the dashboard and tests
3408         rely on.  A handful were resetting just a few objects.  Use the common
3409         resetGlobals() method instead.
3410
3411         * TestResultServer/static-dashboards/loader_unittests.js:
3412
3413 2012-12-13  Andras Becsi  <andras.becsi@digia.com>
3414
3415         [Qt][WK2] Fix painting on Mac with retina display
3416         https://bugs.webkit.org/show_bug.cgi?id=104574
3417
3418         Reviewed by Kenneth Rohde Christiansen.
3419
3420         Remove setting the devicePixelRatio experimental property
3421         since the value is now automatically picked up from Qt.
3422
3423         * MiniBrowser/qt/qml/BrowserWindow.qml:
3424
3425 2012-12-13  Jussi Kukkonen  <jussi.kukkonen@intel.com>
3426
3427         [EFL][GTK] Don't call deprecated g_type_init when glib => 2.35
3428         https://bugs.webkit.org/show_bug.cgi?id=103209
3429
3430         Reviewed by Kenneth Rohde Christiansen.
3431
3432         g_type_init() is deprecated from 2.35.0 onwards. Don't call it
3433         in that case to avoid warnings and/or build failure.
3434
3435         * DumpRenderTree/gtk/ImageDiff.cpp:
3436         (main):
3437
3438 2012-12-13  Claudio Saavedra  <csaavedra@igalia.com>
3439
3440         [Tools] Add exception for gtk_style_context_get()
3441         https://bugs.webkit.org/show_bug.cgi?id=104896
3442
3443         Reviewed by Martin Robinson.
3444
3445         Add new exception for gtk_style_context_get().
3446
3447         * Scripts/webkitpy/style/checkers/cpp.py:
3448         (check_for_null): Add exception for gtk_style_context_get().
3449         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
3450         (WebKitStyleTest.test_null_false_zero): Add a unit
3451         test for the new exception.
3452
3453 2012-12-13  Joaquim Rocha  <jrocha@igalia.com>
3454
3455         REGRESSION (r137432): The /webkit2/WebKitCookieManager/accept-policy unit test is failing
3456         https://bugs.webkit.org/show_bug.cgi?id=104790
3457
3458         Reviewed by Carlos Garcia Campos.
3459
3460         Unskip the GTK accept policy test.
3461
3462         * Scripts/run-gtk-tests:
3463         (TestRunner):
3464
3465 2012-12-13  Xingnan Wang  <xingnan.wang@intel.com>
3466
3467         Unreviewed. Adding myself to committers list.
3468
3469         * Scripts/webkitpy/common/config/committers.py:
3470
3471 2012-12-12  Roger Fong  <roger_fong@apple.com>
3472
3473         Unreviewed. Fix typo in NRWT.
3474
3475         * Scripts/webkitpy/layout_tests/port/base.py:
3476         (Port.is_reference_html_file):
3477
3478 2012-12-12  Roger Fong  <roger_fong@apple.com>
3479
3480         Fix detection of ref test files in old and new run webkit tests.
3481         https://bugs.webkit.org/show_bug.cgi?id=104850.
3482
3483         Reviewed by Timothy Horton.
3484
3485         * Scripts/old-run-webkit-tests:
3486         (isUsedInReftest): 
3487         Add notref and ref prefix and suffix cases to isUsedinReftest.
3488         * Scripts/webkitpy/layout_tests/port/base.py:
3489         (Port):
3490         (Port.is_reference_html_file): 
3491         Typo in NRWT.
3492
3493 2012-12-12  Dirk Pranke  <dpranke@chromium.org>
3494
3495         rework ChromiumAndroidDriver.start() et. al to avoid funky layering
3496         https://bugs.webkit.org/show_bug.cgi?id=104856
3497
3498         Reviewed by Eric Seidel.
3499
3500         r137523 / bug 10357 introduced some back-and-forth hopping during
3501         driver.start() in the chromium android port. This patch cleans it
3502         up a bit.
3503
3504         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
3505         (ChromiumAndroidDriver.__init__):
3506         (ChromiumAndroidDriver._start_once):
3507         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
3508         * Scripts/webkitpy/layout_tests/port/driver.py:
3509         (Driver.start):
3510         (Driver._start):
3511
3512 2012-12-12  Eric Seidel  <eric@webkit.org>
3513
3514         Add support for run-perf-tests --chromium-android --profile
3515         https://bugs.webkit.org/show_bug.cgi?id=103587
3516
3517         Reviewed by Dirk Pranke.
3518
3519         This makes it possible to profile PerformanceTests on an Android device
3520         using run-perf-tests --chromium-android --profile.
3521         CPU profiling on Android is sadly non-trivial, but this goes a long
3522         ways towards making it more accessible to non-Android developers.
3523
3524         Although this script requires your device to be flashed with a
3525         userdebug build (an annoyingly non-trivial endeavor for external contributors)
3526         it endeavors to detect and explain the common errors encountered
3527         while trying to setup a device for perf-based profiling.
3528
3529         * Scripts/webkitpy/common/system/profiler.py:
3530         (GooglePProf.profile_after_exit): Discovered that pprof will just hang if passed a missing file.
3531         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
3532         (AndroidPerf):
3533         (AndroidPerf.__init__):
3534         (AndroidPerf.check_configuration):
3535         (AndroidPerf.print_setup_instructions):
3536         (_file_exists_on_device):
3537         (_run_adb_command):
3538         (attach_to_pid):
3539         (_perf_version_string):
3540         (_find_perfhost_binary):
3541         (_perfhost_path):
3542         (_first_ten_lines_of_profile):
3543         (profile_after_exit):
3544         (ChromiumAndroidDriver.__init__):
3545         (ChromiumAndroidDriver.__init__.methods):
3546         (ChromiumAndroidDriver._find_or_create_symfs):
3547         (ChromiumAndroidDriver._start_once):
3548         (ChromiumAndroidDriver):
3549         (ChromiumAndroidDriver._run_post_start_tasks):
3550         (ChromiumAndroidDriver._pid_from_android_ps_output):
3551         (ChromiumAndroidDriver._pid_on_target):
3552         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
3553         (ChromiumAndroidDriverTest.test_pid_from_android_ps_output):
3554         (AndroidPerfTest):
3555         (AndroidPerfTest.test_perf_output_regexp):
3556         * Scripts/webkitpy/layout_tests/port/driver.py:
3557         (Driver._setup_environ_for_driver):
3558         (Driver):
3559         (Driver._start):
3560         (Driver._run_post_start_tasks):
3561         (Driver._pid_on_target):
3562
3563 2012-12-12  Dirk Pranke  <dpranke@chromium.org>
3564
3565         nrwt: remove deprecated ORWT compatibility command line flags
3566         https://bugs.webkit.org/show_bug.cgi?id=104301
3567
3568         Reviewed by Eric Seidel.
3569
3570         The command line flags for --noshow-results, --[no-]launch-safari,
3571         and --use-remote-links-to-tests were for compatibility with
3572         old-run-webkit-tests, but the compatibility is not needed any more
3573         since NRWT is basically default and the scripts that passed the
3574         old flags have been updated.
3575
3576         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3577         (_set_up_derived_options):
3578         (parse_args):
3579
3580 2012-12-12  John Griggs  <jgriggs@rim.com>
3581
3582         2012-12-12  John Griggs  <jgriggs@rim.com>
3583
3584         [BlackBerry] Update Media Controls
3585         https://bugs.webkit.org/show_bug.cgi?id=104443
3586         https://przilla.ott.qnx.com/bugzilla/show_bug.cgi?id=204748
3587
3588         Update Media Controls for BlackBerry platform to allow audio, embedded video and fullscreen video controls to be positioned and styled differently.
3589
3590         Reviewed by Rob Buis.
3591
3592         BlackBerry-specific update to Media Controls - No new tests required.
3593
3594         * Scripts/webkitperl/FeatureList.pm:
3595
3596 2012-12-12  Justin Novosad  <junov@google.com>
3597
3598         Adding myself to SkiaGraphics watchlist
3599
3600         Unreviewed
3601
3602         * Scripts/webkitpy/common/config/watchlist:
3603
3604 2012-12-12  Martin Robinson  <mrobinson@igalia.com>
3605
3606         [GTK] Add authentication support to DRT and fix exposed issues in the libsoup backend
3607         https://bugs.webkit.org/show_bug.cgi?id=104285
3608
3609         Reviewed by Gustavo Noronha Silva.
3610
3611         Add support to DumpRenderTree for running authentication tests. Since the DRT
3612         expects an authentication callback, we add one to DRTSupport to avoid #ifdefs
3613         in platform-independent code for GTK+.
3614
3615         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3616         (resetDefaultsToConsistentValues): Reset the authentication password and username.
3617         (authenticationCallback): Added.
3618         (createWebView): Attach the authentiation callback during startup.
3619         * DumpRenderTree/gtk/TestRunnerGtk.cpp:
3620         (soupURIToKURL): soup_uri_to_string does not preserve passwords embedded
3621         in the URL so we add a somewhat messy method of re-adding them when they exist.
3622         It would be nice to use soupURIToKURL here, but it seems we cannot use KURL without
3623         pulling in lots of WebCore code so we use string search and replace.
3624         (TestRunner::queueLoad): Use the new helper.
3625
3626 2012-12-12  Dirk Pranke  <dpranke@chromium.org>
3627
3628         garden-o-matic should prefer efl/ over efl-wk1/ and efl-wk2/ when rebaselining
3629         https://bugs.webkit.org/show_bug.cgi?id=102389
3630
3631         Reviewed by Tony Chang.
3632
3633         The code wasn't aware of both the efl-wk1 and efl-wk2 variants,
3634         because all_port_names() by itself doesn't tell you about the
3635         variants. Work around this for now, and filed bug 104761 to
3636