c03d662c9c481134667652a6b36a19ef8946d412
[WebKit-https.git] / Tools / ChangeLog
1 2013-01-01  Alan Cutter  <alancutter@chromium.org>
2
3         sheriff-bot should know who the gardeners/sheriffs are
4         https://bugs.webkit.org/show_bug.cgi?id=105698
5
6         Reviewed by Eric Seidel.
7
8         Added a "sheriffs" command to sheriffbot to fetch the current Chromium Webkit sheriffs.
9
10         * Scripts/webkitpy/tool/bot/irc_command.py:
11         (Current):
12         (Current._retrieve_webkit_sheriffs):
13         (Current.execute):
14         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
15         (IRCCommandTest.test_current):
16         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_0.js: Added.
17         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_1.js: Added.
18         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_2.js: Added.
19         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_malformed.js: Added.
20         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_zero.js: Added.
21
22 2012-12-31  Carlos Garcia Campos  <cgarcia@igalia.com>
23
24         [GTK] Move ImageDiff to the Tools directory root
25         https://bugs.webkit.org/show_bug.cgi?id=105421
26
27         Reviewed by Kenneth Rohde Christiansen.
28
29         * DumpRenderTree/gtk/GNUmakefile.ImageDiff.am: Removed.
30         * GNUmakefile.am:
31         * ImageDiff/gtk/ImageDiff.cpp: Renamed from Tools/DumpRenderTree/gtk/ImageDiff.cpp.
32
33 2012-12-30  Mike West  <mkwst@chromium.org>
34
35         Add myself to watchlists for CSP and console logs.
36         https://bugs.webkit.org/show_bug.cgi?id=105874
37
38         Reviewed by Dirk Pranke.
39
40         Because I really, really, really need more email.
41
42         * Scripts/webkitpy/common/config/committers.py:
43             Add mkwst+watchlist@chromium.org as a contributor.
44         * Scripts/webkitpy/common/config/watchlist:
45             Create watchlist categories for 1) the various mechanisms used to
46             generate console messages, and 2) CSP usage and files.
47
48 2012-12-30  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
49
50         [EFL] Enable TEMPLATE_ELEMENT feature
51         https://bugs.webkit.org/show_bug.cgi?id=105865
52         
53         Reviewed by Laszlo Gombos.
54
55         * Scripts/webkitperl/FeatureList.pm: Enable template element feature for EFL port.
56
57 2012-12-30  Zan Dobersek  <zandobersek@gmail.com>
58
59         Unreviewed GTK gardening of the unit tests.
60
61         Skip the TestInspectorServer unit test that's timing out completely.
62
63         * Scripts/run-gtk-tests:
64         (TestRunner):
65
66 2012-12-29  Ilya Tikhonovsky  <loislo@chromium.org>
67
68         Web Inspector: Native Memory Instrumentation plugin: do not generate 'not instrumented' warning for instrumented mutable members.
69         https://bugs.webkit.org/show_bug.cgi?id=105855
70
71         Reviewed by Vsevolod Vlasov.
72
73         Extract MemberExpr from ImplicitCastExpr. It happens when we instrument a mutable member because
74         addMember expects const T& and the mutable ref to member implicitly converts into const ref to member.
75
76         * clang/ReportMemoryUsagePlugin/ReportMemoryUsage.cpp:
77         (ReportMemoryUsageVisitor):
78         (clang::AddMemberCallVisitor::VisitCallExpr):
79         (clang::ReportMemoryUsageVisitor::VisitCXXMethodDecl):
80         (clang::ReportMemoryUsageVisitor::checkMembersCoverage):
81
82 2012-12-29  Ilya Tikhonovsky  <loislo@chromium.org>
83
84         Web Inspector: Native Memory Instrumentation plugin: move function bodies out of class declarations.
85         https://bugs.webkit.org/show_bug.cgi?id=105852
86
87         Reviewed by Alexander Pavlov.
88
89         * clang/ReportMemoryUsagePlugin/ReportMemoryUsage.cpp:
90         (clang):
91         (AddMemberCallVisitor):
92         (ReportMemoryUsageVisitor):
93         (clang::ReportMemoryUsageAction::ParseArgs):
94         (clang::AddMemberCallVisitor::VisitCallExpr):
95         (clang::ReportMemoryUsageVisitor::VisitCXXMethodDecl):
96         (clang::ReportMemoryUsageVisitor::emitWarning):
97         (clang::ReportMemoryUsageVisitor::findInstrumentationMethod):
98         (clang::ReportMemoryUsageVisitor::needsToBeInstrumented):
99         (clang::ReportMemoryUsageVisitor::CheckMembersCoverage):
100
101 2012-12-29  Zan Dobersek  <zandobersek@gmail.com>
102
103         [webkitpy] Omit webkitpy/thirdparty/BeautifulSoup.py from code coverage checking
104         https://bugs.webkit.org/show_bug.cgi?id=105835
105
106         Reviewed by Dirk Pranke.
107
108         The BeautifulSoup module is a third-party import and as such we don't have much control
109         over its testing. Because of that it should be omitted from code coverage checking.
110
111         * Scripts/webkitpy/test/main.py:
112         (Tester._run_tests):
113
114 2012-12-28  Martin Robinson  <mrobinson@igalia.com>
115
116         [GTK][WK2] Add support for IME Composition
117         https://bugs.webkit.org/show_bug.cgi?id=65093
118
119         Reviewed by Carlos Garcia Campos.
120
121         Update unit tests for GtkInputMethodFilter to note whether or not an event
122         was a fake event synthesized for composition purposes. This is now used
123         for WebKit2.
124
125         * TestWebKitAPI/Tests/gtk/InputMethodFilter.cpp:
126         (TestWebKitAPI::TestInputMethodFilter::sendSimpleKeyEvent):
127         (TestWebKitAPI::TestInputMethodFilter::sendKeyEventWithCompositionResults):
128         (TestWebKitAPI::TEST):
129         * gtk/generate-gtkdoc:
130         (get_webkit2_options): Skip new files in the WebKit2 source directory for documentation
131         generation.
132
133 2012-12-27  Ilya Tikhonovsky  <loislo@chromium.org>
134
135         Web Inspector: Native Memory Instrumentation. Update clang plugin according to the current state of memory instrumentation code.
136         https://bugs.webkit.org/show_bug.cgi?id=105800
137
138         Reviewed by Yury Semikhatsky.
139
140         * clang/ReportMemoryUsagePlugin/ReportMemoryUsage.cpp:
141         (clang::ReportMemoryUsageConsumer::ReportMemoryUsageConsumer):
142         (clang):
143
144 2012-12-27  Zan Dobersek  <zandobersek@gmail.com>
145
146         Create a GTK build system watchlist and add myself to it
147         https://bugs.webkit.org/show_bug.cgi?id=105769
148
149         Reviewed by Dirk Pranke.
150
151         Add a GTK build system watchlist, watching the configure.ac, GNUmakefile.am files
152         and the input GNUmakefile.features.am.in file. The GNUmakefile.list.am files are
153         omitted on purpose due to relatively high change rate they are exposed to.
154
155         The new watchlist would be of use to me as the GTK build system recently went
156         through a series of changes that included a changed policy on adding configuration
157         flags to configure.ac and subsequent additions to GNUmakefile.am files. A PSA on the
158         mailing list about this change will be published later, but I'd also like to be
159         notified of any pending changes to these files just so I can check the changes are
160         in compliance with the new policy.
161
162         * Scripts/webkitpy/common/config/watchlist:
163
164 2012-12-26  Ilya Tikhonovsky  <loislo@chromium.org>
165
166         Unreviewed. Replace find("Tools") with rfind("Tools") because base dir of WebKit may have Tools word.
167         As example ..../DevTools/src/third_party/WebKit
168
169         * Scripts/webkitpy/common/webkit_finder.py:
170         (WebKitFinder.webkit_base):
171
172 2012-12-25  Christophe Dumez  <christophe.dumez@intel.com>
173
174         [EFL][WK2] Refactor snapshot taking code
175         https://bugs.webkit.org/show_bug.cgi?id=105687
176
177         Reviewed by Kenneth Rohde Christiansen.
178
179         Call WKViewCreateSnapshot() instead of WKViewGetSnapshot()
180         since the function was renamed to follow naming style.
181
182         * WebKitTestRunner/efl/PlatformWebViewEfl.cpp:
183         (WTR::PlatformWebView::windowSnapshotImage):
184
185 2012-12-24  Ilya Tikhonovsky  <loislo@chromium.org>
186
187         Web Inspector: Native Memory Instrumentation: propagate member type as edge type to the serialized heap graph.
188         https://bugs.webkit.org/show_bug.cgi?id=105725
189
190         Reviewed by Yury Semikhatsky.
191
192         MemberType value names were adjusted according to Style Guide.
193         Existing tests were extended with link type validation.
194
195         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
196
197 2012-12-24  Xiaobo Wang  <xbwang@torchmobile.com.cn>
198
199         [BlackBerry] DRT - Crashed when running pixels tests
200         https://bugs.webkit.org/show_bug.cgi?id=105711
201
202         RIM PR 270003
203
204         Reviewed by George Staikos.
205
206         Calling window->post() here will result in buffer swapping which is
207         only allowed in UI thread now, and I found we don't need to call it
208         for GL renderer.
209
210         * DumpRenderTree/blackberry/PixelDumpSupportBlackBerry.cpp:
211         (createBitmapContextFromWebView):
212
213 2012-12-22  Dan Winship  <danw@gnome.org>
214
215         Add a patch to fix libxml2 2.9.0 regression
216         https://bugs.webkit.org/show_bug.cgi?id=104494
217
218         Reviewed by Martin Robinson.
219
220         * gtk/jhbuild.modules:
221         * gtk/patches/libxml2-2.9.0-dtd.patch: patch from upstream
222
223 2012-12-23  Allan Sandfeld Jensen  <allan.jensen@digia.com>
224
225         Unreviewed update of status to reviewer.
226
227         * Scripts/webkitpy/common/config/committers.py:
228
229 2012-12-23  Philippe Normand  <pnormand@igalia.com>
230
231         [GTK][jhbuild] Bump GStreamer to 1.0.4
232         https://bugs.webkit.org/show_bug.cgi?id=105640
233
234         Reviewed by Martin Robinson.
235
236         * gtk/jhbuild.modules: Update GStreamer modules to 1.0.4.
237
238 2012-12-21  Nico Weber  <thakis@chromium.org>
239
240         chromium: webkit-build-directory doesn't handle ninja output dirs
241         https://bugs.webkit.org/show_bug.cgi?id=105671
242
243         Reviewed by Daniel Bates.
244
245         Teach webkit-build-directory to detect the ninja build directory.
246
247         * Scripts/webkitdirs.pm:
248         (determineBaseProductDir):
249         (determineIsChromiumNinja):
250
251 2012-12-21  Ryosuke Niwa  <rniwa@webkit.org>
252
253         [Mountain Lion] platform/mac/editing/spelling/removing-underline-after-accepting-autocorrection-using-punctuation.html failing
254         https://bugs.webkit.org/show_bug.cgi?id=95477
255
256         Reviewed by Sam Weinig.
257
258         Co-authored by Jia Pu.
259
260         Specify a fixed autocorrection dictionary in resetDefaultsToConsistentValues() so that
261         we can have consistent autocorrection behavior when running tests.
262
263         * DumpRenderTree/mac/DumpRenderTree.mm:
264         (resetDefaultsToConsistentValues):
265         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
266         (WTR::InjectedBundle::platformInitialize):
267
268 2012-12-21  Zan Dobersek  <zandobersek@gmail.com>
269
270         [GTK] Add WebKit1-specific TestExpectations file
271         https://bugs.webkit.org/show_bug.cgi?id=97562
272
273         Reviewed by Dirk Pranke.
274
275         Implement the expectations_files method inside the GtkPort class that
276         returns the list of the TestExpectations files the current test configuration
277         requires. The method is implemented in a similar way as on the EFL port.
278         A unit test is also added.
279
280         * Scripts/webkitpy/layout_tests/port/gtk.py:
281         (GtkPort._search_paths):
282         (GtkPort):
283         (GtkPort.expectations_files):
284         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
285         (GtkPortTest.test_expectations_files):
286
287 2012-12-21  Sudarsana Nagineni  <sudarsana.nagineni@intel.com>
288
289         [EFL] Enable feature option for the CSS sticky position
290         https://bugs.webkit.org/show_bug.cgi?id=105630
291
292         Reviewed by Kenneth Rohde Christiansen.
293
294         Enable CSS sticky position feature by default for the
295         EFL port. This feature option was added in r138252.
296
297         No new tests. Existing layout tests pass.
298
299         * Scripts/webkitperl/FeatureList.pm:
300
301 2012-12-21  Jussi Kukkonen  <jussi.kukkonen@intel.com>
302
303         [WK2] REGRESSION(r138232): It made 4 tests timeout
304         https://bugs.webkit.org/show_bug.cgi?id=105623
305
306         Reviewed by Kenneth Rohde Christiansen.
307
308         r138232 changed functionality so that empty text output
309         no longer leads to a dump() call, which leads to a timeout.
310         Instead do a text dump if there is text output or if there
311         is no audio output.
312
313         * WebKitTestRunner/TestInvocation.cpp:
314         (WTR::TestInvocation::dumpResults):
315
316 2012-12-21  Zeno Albisser  <zeno@webkit.org>
317
318         [Qt][Mac] run-launcher cannot find Qt5WebKitWidgets.
319         https://bugs.webkit.org/show_bug.cgi?id=105309
320
321         On Windows the library name is Qt5WebKitWidgets as well.
322         So r138136 should not have adjusted the library name for Windows.
323
324         Reviewed by Simon Hausmann.
325
326         * Scripts/webkitdirs.pm:
327         (builtDylibPathForName):
328
329 2012-12-21  Zeno Albisser  <zeno@webkit.org>
330
331         [Qt] Add new TestExpectations for qt-5.0-mac-wk2.
332         https://bugs.webkit.org/show_bug.cgi?id=105424
333
334         This change will allow us to run a very specific
335         subset of the available LayoutTests on a mac buildbot.
336         The new baseline search path for mac will look as follows:
337         qt-5.0-mac-wk2 -> qt-5.0-wk2 -> qt-5.0 -> qt-mac -> qt
338
339         Reviewed by Dirk Pranke.
340
341         * Scripts/webkitpy/layout_tests/port/qt.py:
342         (QtPort._search_paths):
343         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
344         (QtPortTest):
345
346 2012-12-20  Ryuan Choi  <ryuan.choi@samsung.com>
347
348         [EFL] Build break with latest EFL libraries.
349         https://bugs.webkit.org/show_bug.cgi?id=104827
350
351         Reviewed by Laszlo Gombos.
352
353         The eo EFL package is introduced and evas and ecore use it since 1.8.
354         While introducing Eo, EFL changed several structures of Evas and Ecore
355         from own specific class to Eo.
356
357         So, this patch adds FindEo and changes declaration of Evas, Evas_Object,
358         Ecore_Timer to build with latest EFL libraries.
359
360         * DumpRenderTree/efl/CMakeLists.txt: Includes eo EFL library.
361         * EWebLauncher/CMakeLists.txt: Ditto.
362         * MiniBrowser/efl/CMakeLists.txt: Ditto.
363         * ImageDiff/PlatformEfl.cmake: Ditto.
364         * TestWebKitAPI/PlatformEfl.cmake: Ditto.
365         * TestWebKitAPI/PlatformWebView.h: Modified declaration of Evas_Object.
366         * WebKitTestRunner/InjectedBundle/TestRunner.h: Modified declaration of Ecore_Timer.
367         * WebKitTestRunner/PlatformEfl.cmake: Includes eo EFL library.
368         * WebKitTestRunner/PlatformWebView.h: Modified declaration of Evas_Object.
369
370 2012-12-20  Dirk Pranke  <dpranke@chromium.org>
371
372         TestExpectation fallback is broken
373         https://bugs.webkit.org/show_bug.cgi?id=105583
374
375         Reviewed by Ryosuke Niwa.
376
377         When I added support for the -wk2 in port names, it looks
378         like I broke the way we computed the list of TestExpectations
379         files so that the implementation-version directory was left out :(.
380         Fixed and cleaned up the tests.
381
382         * Scripts/webkitpy/layout_tests/port/base.py:
383         (Port.expectations_files):
384         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
385         (TestWebKitPort.__init__):
386         (PortTestCase.test_skipped_directories_for_symbols):
387         (test_skipped_directories_for_features):
388         (test_skipped_directories_for_features_no_matching_tests_in_test_list):
389         (test_skipped_tests_for_unsupported_features_empty_test_list):
390         (test_skipped_layout_tests):
391         (test_expectations_files):
392
393 2012-12-20  Anders Carlsson  <andersca@apple.com>
394
395         Don't include WebKit2 headers when building TestWebKitAPI on windows
396         https://bugs.webkit.org/show_bug.cgi?id=105582
397
398         Reviewed by Tim Horton.
399
400         * TestWebKitAPI/config.h:
401         * TestWebKitAPI/win/main.cpp:
402
403 2012-12-20  Ryosuke Niwa  <rniwa@webkit.org>
404
405         REGRESSION(r133820?): SimplifyMarkupTest API test asserts
406         https://bugs.webkit.org/show_bug.cgi?id=105370
407
408         Reviewed by Simon Fraser.
409
410         Re-enable the test.
411
412         * TestWebKitAPI/Tests/mac/SimplifyMarkup.mm:
413         (TestWebKitAPI::TEST):
414
415 2012-12-20  Nico Weber  <thakis@chromium.org>
416
417         chromium nrwt: Pick the newest binary found in DEFAULT_BUILD_DIRECTORIES, not the first
418         https://bugs.webkit.org/show_bug.cgi?id=105498
419
420         Reviewed by Dirk Pranke.
421
422         Use the newest binary available rather than an than always picking one
423         build directory over another based on iteration order.
424
425         * Scripts/webkitpy/layout_tests/port/chromium.py:
426         (ChromiumPort._static_build_path):
427         Check for timestamps.
428         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
429         (ChromiumMacPortTest.test_build_path_timestamps):
430         Test that out / xcodebuild selection happens based on timestamps
431         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
432         (ChromiumWinPortTest.test_build_path_timestamps):
433         Test that out / build selection happens based on timestamps
434
435 2012-12-19  Simon Fraser  <simon.fraser@apple.com>
436
437         Lots of sticky tests failing in WK2
438         https://bugs.webkit.org/show_bug.cgi?id=105464
439
440         Reviewed by Beth Dakin.
441
442         WebKitTestRunner had a race between snapshotting in the UI process,
443         and resettting after the test in the web process. InjectedBundle::done()
444         was a bad place to call page()->resetAfterTest(), because of this race;
445         it could reset the scroll position before the UI snapshot had been obtained.
446         
447         Fix by moving the call to page()->resetAfterTest() into didReceiveMessage(),
448         for the "Reset" message which will come in before the next test.
449
450         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
451         (WTR::InjectedBundle::didReceiveMessage):
452         (WTR::InjectedBundle::done):
453
454 2012-12-19  Filip Pizlo  <fpizlo@apple.com>
455
456         DFG speculation checks that take JumpList should consolidate OSRExits
457         https://bugs.webkit.org/show_bug.cgi?id=105401
458
459         Reviewed by Oliver Hunt.
460
461         Update the profiler to understand that an OSR exit may have multiple
462         candidate exit sites.
463
464         * Scripts/display-profiler-output:
465
466 2012-12-20  Eric Seidel  <eric@webkit.org>
467
468         EWSes do not need to dump build/test logs when things pass
469         https://bugs.webkit.org/show_bug.cgi?id=105402
470
471         Reviewed by Dirk Pranke.
472
473         When we were first writing the commit-queue and EWS system
474         and running them locally, we thought it was important to see the
475         build/test output as it happened.  Now these queues have been running
476         for years, and we don't use this feature anymore.  So this patch
477         makes us stop tee-ing the output directly to stdout, and instead
478         only print any output from commands when they fail.
479
480         I also added some logging when we execute each command
481         since we are no longer printing the command output itself.
482         This should make it much easier to see errors with the EWS
483         system itself, now that all the passing-build output is removed.
484
485         This required updating a zillion unittests, since we have a bunch of
486         "integration" style testing of the queue-bots.  Many of the unittests for the bots were
487         already approximating this behavior by using a "logging" Executive.
488         I disabled logging in those cases to avoid redundant logging in those tests.
489
490         * Scripts/webkitpy/common/system/executive.py:
491         (Executive.command_for_printing):
492         (Executive.run_command):
493         * Scripts/webkitpy/common/system/executive_mock.py:
494         (MockExecutive.command_for_printing):
495         (MockExecutive.run_command):
496         * Scripts/webkitpy/common/system/executive_unittest.py:
497         (ExecutiveTest.test_auto_stringify_args):
498         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
499         (EarlyWarningSytemTest._default_expected_logs):
500         (_test_ews):
501         (_test_ewses):
502         * Scripts/webkitpy/tool/commands/queues.py:
503         (AbstractQueue.run_webkit_patch):
504         * Scripts/webkitpy/tool/commands/queues_unittest.py:
505         (AbstractQueueTest._assert_run_webkit_patch):
506         (CommitQueueTest.test_commit_queue):
507         (test_rollout):
508         (test_rollout_lands):
509         (test_manual_reject_during_processing):
510         (StyleQueueTest.test_style_queue_with_style_exception):
511         (test_style_queue_with_watch_list_exception):
512
513 2012-12-20  Dominic Mazzoni  <dmazzoni@google.com>
514
515         AX: support clickPoint in DRT for chromium
516         https://bugs.webkit.org/show_bug.cgi?id=97359
517
518         Reviewed by Chris Fleizach.
519
520         Implement clickPointX, clickPointY, and elementAtPoint for Chromium, so that
521         several more tests can pass.
522
523         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
524         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
525         (WebTestRunner::AccessibilityUIElement::clickPointXGetterCallback):
526         (WebTestRunner):
527         (WebTestRunner::AccessibilityUIElement::clickPointYGetterCallback):
528         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
529         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
530         (AccessibilityUIElement):
531
532 2012-12-20  Dan Bernstein  <mitz@apple.com>
533
534         prepare-ChangeLog adds "Additional information of the change such as approach, rationale" where it doesn't make sense
535         https://bugs.webkit.org/show_bug.cgi?id=99268
536
537         Reviewed by Alexey Proskuryakov.
538
539         * Scripts/prepare-ChangeLog:
540         (generateNewChangeLogs): Removed the "Additional information of the change" line.
541
542 2012-12-20  Carlos Garcia Campos  <cgarcia@igalia.com>
543
544         Unreviewed. Skip unit test /webkit/atk/getTextInParagraphAndBodyModerate
545
546         * Scripts/run-gtk-tests:
547         (TestRunner):
548
549 2012-12-20  Thiago Marcos P. Santos  <thiago.santos@intel.com>
550
551         [Qt] Move ImageDiff to the Tools directory root
552         https://bugs.webkit.org/show_bug.cgi?id=105420
553
554         Reviewed by Tor Arne Vestbø.
555
556         Moved ImageDiff from outside DumpRenderTree because it is also a
557         dependency when running pixel tests for WebKit2.
558
559         * ImageDiff/ImageDiff.pro: Renamed from Tools/DumpRenderTree/qt/ImageDiff.pro.
560         * ImageDiff/qt/ImageDiff.cpp: Renamed from Tools/DumpRenderTree/qt/ImageDiff.cpp.
561         * Tools.pro:
562
563 2012-12-20  Zan Dobersek  <zandobersek@gmail.com>
564
565         [GTK] Remove the --enable-unstable-features configuration option
566         https://bugs.webkit.org/show_bug.cgi?id=105327
567
568         Reviewed by Martin Robinson.
569
570         Add a missing feature option for the CSS sticky position feature.
571
572         * Scripts/webkitperl/FeatureList.pm:
573
574 2012-12-20  Carlos Garcia Campos  <cgarcia@igalia.com>
575
576         Unreviewed. Skip unit test WebKit2.ForceRepaint that times out
577
578         * Scripts/run-gtk-tests:
579         (TestRunner):
580
581 2012-12-20  Dominik Röttsches  <dominik.rottsches@intel.com>
582
583         [EFL] MiniBrowser does not play Infinite Gangnam Style
584         https://bugs.webkit.org/show_bug.cgi?id=103531
585
586         Reviewed by Kenneth Rohde Christiansen.
587
588         For EFL we would like this feature to be built by default.
589
590         * Scripts/webkitperl/FeatureList.pm:
591
592 2012-12-20  Csaba Osztrogonác  <ossy@webkit.org>
593
594         [Qt] EWS bots should call qmake always
595         https://bugs.webkit.org/show_bug.cgi?id=90049
596
597         Reviewed by Simon Hausmann.
598
599         * Scripts/webkitdirs.pm:
600         (buildQMakeProjects):
601
602 2012-12-20  Jochen Eisinger  <jochen@chromium.org>
603
604         [chromium] fix stray printf in TestRunner's WebTestProxy
605         https://bugs.webkit.org/show_bug.cgi?id=105515
606
607         Reviewed by Ryosuke Niwa.
608
609         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
610         (WebTestRunner::WebTestProxyBase::shouldInsertNode):
611
612 2012-12-20  Dominik Röttsches  <dominik.rottsches@intel.com>
613
614         [WK2] WebAudio WKTR support
615         https://bugs.webkit.org/show_bug.cgi?id=95084
616
617         Reviewed by Kenneth Rohde Christiansen.
618
619         Implementing audio data dump support for WebKitTestRunner,
620         by adding the API to the IDL, enabling IPC for transferring
621         the audio data as bytebuffer and dumping it according
622         to run-webkit-tests script's expectations.
623         This enables audio testing on WebKit2 ports.
624         Based on initial wip-patch from Philippe Normand - thanks.
625
626         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl: Adding setAudioData API.
627         * WebKitTestRunner/InjectedBundle/TestRunner.cpp: Assign audio buffer data to a local WKDataRef after extracting it from WK2 using WKBundleDataFromUint8Array bundle API.
628         (WTR::TestRunner::setAudioData):
629         (WTR):
630         * WebKitTestRunner/InjectedBundle/TestRunner.h: Add a whatToDump state for dumping Audio.
631         (TestRunner):
632         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp: Add audio to the message that's transferred to the WKTR process.
633         (WTR::InjectedBundle::done):
634         * WebKitTestRunner/InjectedBundle/InjectedBundle.h: Allow setting and audio result handle to be transferred.
635         (WTR::InjectedBundle::setAudioResult):
636         (InjectedBundle):
637         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
638         (WTR::InjectedBundlePage::dump): Dummy handling the Audio case in the switch, build fix for Mac.
639         * WebKitTestRunner/TestInvocation.cpp: Audio dump support, dump audio if we haven't received a text dump from the web process.
640         (WTR::TestInvocation::dumpResults):
641         (WTR):
642         (WTR::TestInvocation::dumpAudio):
643         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle): Storing received audio data before dumping.
644         * WebKitTestRunner/TestInvocation.h:
645         (TestInvocation):
646
647 2012-12-20  Alberto Garcia  <agarcia@igalia.com>
648
649         [GTK] Count failed and timed out tests separately
650         https://bugs.webkit.org/show_bug.cgi?id=105448
651
652         Reviewed by Carlos Garcia Campos.
653
654         Display the number of tests that fail, time out and are skipped
655         separately in the buildbot page.
656
657         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
658         (RunGtkAPITests.commandComplete):
659         (RunGtkAPITests.evaluateCommand):
660         (RunGtkAPITests.getText2):
661         * Scripts/run-gtk-tests:
662         (TestRunner.run_tests):
663
664 2012-12-19  Ryosuke Niwa  <rniwa@webkit.org>
665
666         PerfTest.parse_output does too much
667         https://bugs.webkit.org/show_bug.cgi?id=105391
668
669         Reviewed by Tony Chang.
670
671         Moved the code to filter lines into PerfTest._filter_output, which has been renamed from _filter_stderr.
672         Also moved the code to output test results into PerfTest._run_with_driver.
673
674         * Scripts/webkitpy/performance_tests/perftest.py:
675         (PerfTest.__init__): Added _description.
676         (PerfTest.description): Added.
677         (PerfTest._run_with_driver): Moved the code to output test description and test results.
678         (PerfTest._filter_output): Moved and renamed from PerfTest._filter_stderr.
679         (PerfTest.parse_output): Removed the code to output test results. Return test results and description
680         respectively. Also removed the code that allowed some tests to omit values since all tests report each
681         iteration now since r136492.
682         (PerfTest.output_statistics): Removed the code to print test description, now done in _run_with_driver.
683         (ChromiumStylePerfTest._run_with_driver): Added. Chromium style tests are sufficiently different from
684         regular PerfTest that it doesn't make much sense to share _run_with_driver. But really, we should just
685         get rid of this type of test altogether in favor of regular performance tests that uses runner.js.
686         (ChromiumStylePerfTest.parse_and_log_output): Renamed from parse_output.
687         (PageLoadingPerfTest._run_with_driver): Removed the explicit '' for the test description.
688         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
689         (MainTest.test_parse_output): Removed the expected logs since parse_output no longer prints out results.
690         Also added a call to _filter_output since parse_output doesn't filter the output text anymore.
691         (MainTest.test_parse_output_with_failing_line): Added a call to _filter_output. Also added ',' after 'Time:'
692         so that it's not string-concatenated with the next line.
693         (MainTest.test_parse_output_with_description): Added; a test for PerfTest.description().
694         (MainTest.test_parse_output_with_subtests): Removed the expected logs, and added a call to _filter_output.
695
696 2012-12-19  Alexis Menard  <alexis@webkit.org>
697
698         Implement CSS parsing for CSS transitions unprefixed.
699         https://bugs.webkit.org/show_bug.cgi?id=104804
700
701         Reviewed by Dean Jackson.
702
703         Add a new flag ENABLE_CSS_TRANSFORMS_ANIMATIONS_TRANSITIONS_UNPREFIXED
704         to cover the work of unprefixing Transforms, Animations and 
705         Transitions. It will let the possibility of each ports to turn it off 
706         in their release branches until we're confident that these CSS 
707         properties are ready to be unprefixed.
708
709         * Scripts/webkitperl/FeatureList.pm:
710         * qmake/mkspecs/features/features.pri:
711
712 2012-12-19  Julien Chaffraix  <jchaffraix@webkit.org>
713
714         2 clampTo unit tests are failing
715         https://bugs.webkit.org/show_bug.cgi?id=105253
716
717         Reviewed by Emil A Eklund.
718
719         Removed clampToUnsigned as it was actually testing exactly what
720         clampToUnsignedLong was doing. Also made clampToIntFloat more robust
721         to precision loss (fix suggested by Emil).
722
723         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
724
725 2012-12-19  Thiago Marcos P. Santos  <thiago.santos@intel.com>
726
727         [EFL] Move ImageDiff to the Tools directory root
728         https://bugs.webkit.org/show_bug.cgi?id=105305
729
730         Reviewed by Laszlo Gombos.
731
732         We previously moved this directory from the DRT directory to
733         efl/ImageDiff. The major reason for doing so was because ImageDiff does
734         not depend on DumpRenderTree but for legacy reasons was living inside the
735         DRT folder, even though it is needed when running pixel tests for WK2.
736
737         Now we are moving to Tool directory root as other ports agreed to do the same.
738
739         * CMakeLists.txt:
740         * ImageDiff/CMakeLists.txt: Added.
741         * ImageDiff/PlatformEfl.cmake: Added.
742         * ImageDiff/efl/ImageDiff.cpp: Renamed from Tools/efl/ImageDiff/ImageDiff.cpp.
743         * efl/ImageDiff/CMakeLists.txt: Removed.
744
745 2012-12-19  Zeno Albisser  <zeno@webkit.org>
746
747         [Qt][Mac] run-launcher cannot find Qt5WebKitWidgets.
748         https://bugs.webkit.org/show_bug.cgi?id=105309
749
750         The version number is only used as a part
751         of the library name on Linux.
752
753         Reviewed by Kenneth Rohde Christiansen.
754
755         * Scripts/webkitdirs.pm:
756         (builtDylibPathForName):
757
758 2012-12-19  Alberto Garcia  <agarcia@igalia.com>
759
760         [GTK] Tests that time out are not considered as failures.
761         https://bugs.webkit.org/show_bug.cgi?id=105318
762
763         Reviewed by Martin Robinson.
764
765         Fix parsing of the GTK API tests log files.
766
767         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
768         (RunGtkAPITests.commandComplete):
769
770 2012-12-18  Ilya Tikhonovsky  <loislo@chromium.org>
771
772         Unreviewed, rolling out r138061.
773         https://bugs.webkit.org/show_bug.cgi?id=105396
774
775         Original patch:
776         http://trac.webkit.org/changeset/138061
777         https://bugs.webkit.org/show_bug.cgi?id=97359
778
779         It is crashing on Debug bots
780
781         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
782         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
783         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
784         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
785         (AccessibilityUIElement):
786
787 2012-12-18  Alice Liu  <alice.liu@apple.com>
788
789         Test for https://bugs.webkit.org/show_bug.cgi?id=105106
790         Add SPI to WebKit1 WebFrame for hit testing
791
792         Reviewed by Dan Bernstein.
793
794         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added file Tests/mac/ElementAtPointInWebFrame.mm
795         * TestWebKitAPI/Tests/mac/ElementAtPointInWebFrame.mm: Added.
796         (TestWebKitAPI::TEST): Loads html with two divs positioned in the 2nd and 4th quadrants of the webview.
797         Then hit-tests at three points, expecting to hit the two divs and body element. 
798
799 2012-12-18  Alice Liu  <alice.liu@apple.com>
800
801         Test for https://bugs.webkit.org/show_bug.cgi?id=105262
802         Add SPI to WebKit1 WebFrame for node conversion to JSValueRef
803
804         Reviewed by Anders Carlsson.
805
806         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added file Tests/mac/JSWrapperForNodeInWebFrame.mm.mm
807         * TestWebKitAPI/Tests/mac/JSWrapperForNodeInWebFrame.mm.mm: Added.
808         (TestWebKitAPI::TEST): Tests for the correct JS wrapper for a DOMNode, provided a WebScriptWorld.
809         Loads html with a single div element. In an isolated world, creates a property on that node.
810         Also, in the standard world, creates a different property on that node.  Then tests for 4 things:
811             - Existence of the isolated property in the isolated world.
812             - Existence of the standard property in the standard world.
813             - Non-existence of the isolated property in the standard world.
814             - Non-existence of the standard property in the isolated world.
815
816 2012-12-18  Simon Fraser  <simon.fraser@apple.com>
817
818         Disable the SimplifyMarkup API test, wich asserts in debug builds.
819         Tracked by https://bugs.webkit.org/show_bug.cgi?id=105370.
820         
821         * TestWebKitAPI/Tests/mac/SimplifyMarkup.mm:
822         (TestWebKitAPI::TEST):
823
824 2012-12-18  Joseph Pecoraro  <pecoraro@apple.com>
825
826         [Mac] Add Build Phase to Check Headers for Inappropriate Macros (Platform.h macros)
827         https://bugs.webkit.org/show_bug.cgi?id=104279
828
829         Reviewed by David Kilzer.
830
831         Add a build phase script that checks a framework's header files for
832         Platform.h and Compiler.h macros. Also add some tests for the script.
833
834         * Scripts/check-for-inappropriate-macros-in-external-headers: Added.
835         Script to be used in build phases to check for inappropriate macros in headers.
836
837         * Scripts/test-webkitruby: Added.
838         Script that can be used to run all webkitruby tests.
839
840         * Scripts/test-webkit-scripts:
841         Also run the ruby tests.
842
843         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/fake-data-failing-expected.txt: Added.
844         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/pass-expected.txt: Added.
845         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/resources/Fake.framework/Headers/Fail.h: Added.
846         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/resources/Fake.framework/Headers/Pass.h: Added.
847         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/run-test.rb: Added.
848         Tests for the new script.
849
850 2012-12-18  Dominic Mazzoni  <dmazzoni@google.com>
851
852         AX: support clickPoint in DRT for chromium
853         https://bugs.webkit.org/show_bug.cgi?id=97359
854
855         Reviewed by Chris Fleizach.
856
857         Implement clickPointX, clickPointY, and elementAtPoint for Chromium, so that
858         several more tests can pass.
859
860         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
861         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
862         (WebTestRunner::AccessibilityUIElement::clickPointXGetterCallback):
863         (WebTestRunner):
864         (WebTestRunner::AccessibilityUIElement::clickPointYGetterCallback):
865         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
866         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
867         (AccessibilityUIElement):
868
869 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
870
871         Build fix after r138045.
872
873         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
874         (PerfTestsRunner._collect_tests): Use relative_path. We skip only when this path
875         is not explicitly specified. The condition was backwards but never matched so we
876         always ran the test regardless of whether it's listed in the skipped list or not.
877         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
878         (test_collect_tests_with_ignored_skipped_list): Renamed so that it actually runs.
879
880 2012-12-18  Emil A Eklund  <eae@chromium.org>
881
882         TestWebKitAPI/Tests/WebCore/LayoutUnit fails on mac
883         https://bugs.webkit.org/show_bug.cgi?id=105332
884         
885         Reviewed by Levi Weintraub.
886         
887         Create new test suite for LayoutUnit tests as the KURL tests are
888         of the incompatible TEST_F type.
889
890         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp:
891         (TestWebKitAPI):
892         (TestWebKitAPI::TEST_F):
893
894 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
895
896         Running a skipped test with run-perf-tests could alert the user
897         https://bugs.webkit.org/show_bug.cgi?id=105219
898
899         Reviewed by Eric Seidel.
900
901         Force run skipped performance tests explicitly specified in the command line.
902         Also warn the user when the specified paths don't exist.
903
904         * Scripts/webkitpy/layout_tests/port/base.py:
905         (Port.relative_test_filename):
906         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
907         (PerfTestsRunner._parse_args):
908         (PerfTestsRunner._collect_tests):
909         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
910         (test_collect_tests_with_skipped_list_and_files):
911
912 2012-12-18  Yaron Friedman  <yfriedman@chromium.org>
913
914         Chrome Android java output is too verbose
915         https://bugs.webkit.org/show_bug.cgi?id=105247
916
917         Reviewed by Tony Chang.
918
919         Pass -quiet to ant to suppress extraneous output.
920
921         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
922         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
923
924 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
925
926         Consider removing --pause-before-testing option
927         https://bugs.webkit.org/show_bug.cgi?id=105248
928
929         Reviewed by Eric Seidel.
930
931         Remove --pause-before-testing option.
932
933         * Tools/Scripts/webkitpy/layout_tests/port/driver.py:
934         (DriverProxy.start): Removed.
935         * Scripts/webkitpy/performance_tests/perftest.py:
936         (PerfTest.__init__):
937         (PerfTest.test_path): Renamed from path_or_url.
938         (PerfTest.run):
939         (PerfTest): Removed "driver" argument.
940         (PerfTest._run_with_driver):
941         (PerfTest.run_single):
942         (ChromiumStylePerfTest.__init__):
943         (PageLoadingPerfTest.__init__):
944         (PageLoadingPerfTest.run_single):
945         (PageLoadingPerfTest._run_with_driver): Renamed from run so that PerfTest.run can create a driver for us.
946         (ReplayPerfTest.__init__):
947         (ReplayPerfTest.prepare):
948         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
949         (TestPageLoadingPerfTest.test_run):
950         (TestPageLoadingPerfTest.test_run_with_memory_output):
951         (TestPageLoadingPerfTest.test_run_with_bad_output):
952         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
953         (PerfTestsRunner._parse_args):
954         (PerfTestsRunner._run_tests_set): Removed the code to create a driver and pause before running the test.
955         (PerfTestsRunner._run_single_test): Removed "driver" argument.
956         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
957         (run_test):
958         (test_run_test_pause_before_testing): Removed.
959
960 2012-12-18  Alberto Garcia  <agarcia@igalia.com>
961
962         [GTK] Tests that time out are not considered as failures.
963         https://bugs.webkit.org/show_bug.cgi?id=105318
964
965         Reviewed by Martin Robinson.
966
967         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
968         (RunGtkAPITests.commandComplete):
969         Consider log lines that end with TIMEOUT as errors.
970
971         * Scripts/run-gtk-tests:
972         (TestRunner.run_tests):
973         Consider timeouts as failures when returning the number of failed
974         tests.
975
976 2012-12-18  Pravin D  <pravin.d@samsung.com>
977
978         Unreviewed, adding secondary email.
979
980         * Scripts/webkitpy/common/config/committers.py:
981
982 2012-12-18  Pravin D  <pravin.d@samsung.com>
983
984         Unreviewed, promoting myself to committer's list.
985
986         * Scripts/webkitpy/common/config/committers.py:
987
988 2012-12-18  Thiago Marcos P. Santos  <thiago.santos@intel.com>
989
990         [EFL] ImageDiff does not build on WebKit2 only builds
991         https://bugs.webkit.org/show_bug.cgi?id=105288
992
993         Reviewed by Kenneth Rohde Christiansen.
994
995         Moved ImageDiff to the efl/ folder inside Tools/ (so we don't pollute
996         the folder with things which are EFL-specific.
997
998         ImageDiff will be built by default now regardless of WebKit API flavor
999         you choose.
1000
1001         * CMakeLists.txt:
1002         * DumpRenderTree/efl/CMakeLists.txt:
1003         * efl/ImageDiff/CMakeLists.txt: Added.
1004         * efl/ImageDiff/ImageDiff.cpp: Renamed from Tools/DumpRenderTree/efl/ImageDiff.cpp.
1005
1006 2012-12-18  Kenneth Rohde Christiansen  <kenneth@webkit.org>
1007
1008         [EFL][WK2] When creating new windows, use the opener's context
1009         https://bugs.webkit.org/show_bug.cgi?id=105283
1010
1011         Reviewed by Laszlo Gombos.
1012
1013         Use the context of the opener if exists.
1014
1015         * MiniBrowser/efl/main.c:
1016         (on_key_down):
1017         (on_window_create):
1018         (window_create):
1019         (elm_main):
1020
1021 2012-12-18  Thiago Marcos P. Santos  <thiago.santos@intel.com>
1022
1023         [EFL][WK2] Gardening of TestWebKitAPI tests
1024         https://bugs.webkit.org/show_bug.cgi?id=105268
1025
1026         Reviewed by Kenneth Rohde Christiansen.
1027
1028         Some tests were marked as failing but are passing and the newly added
1029         LayoutUnit should have it's own binary because it conflicts with the
1030         fixture used by KURL and is failing on the bots because of that.
1031
1032         * TestWebKitAPI/CMakeLists.txt:
1033         * TestWebKitAPI/PlatformEfl.cmake:
1034
1035 2012-12-18  Jochen Eisinger  <jochen@chromium.org>
1036
1037         [chromium] move webintent related methods to TestRunner library
1038         https://bugs.webkit.org/show_bug.cgi?id=105164
1039
1040         Reviewed by Darin Fisher.
1041
1042         Also deleted the unused deliveredIntentFailure and
1043         deliveredIntentResult methods.
1044
1045         * DumpRenderTree/chromium/DRTTestRunner.cpp:
1046         (DRTTestRunner::DRTTestRunner):
1047         * DumpRenderTree/chromium/DRTTestRunner.h:
1048         (DRTTestRunner):
1049         * DumpRenderTree/chromium/TestRunner/public/WebTestDelegate.h:
1050         (WebKit):
1051         (WebTestRunner::WebTestDelegate::setCurrentWebIntentRequest):
1052         (WebTestRunner::WebTestDelegate::currentWebIntentRequest):
1053         * DumpRenderTree/chromium/TestRunner/public/WebTestProxy.h:
1054         (WebKit):
1055         (WebTestProxyBase):
1056         (WebTestRunner::WebTestProxy::registerIntentService):
1057         (WebTestRunner::WebTestProxy::dispatchIntent):
1058         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
1059         (WebTestRunner::TestRunner::TestRunner):
1060         (WebTestRunner::TestRunner::sendWebIntentResponse):
1061         (WebTestRunner):
1062         (WebTestRunner::TestRunner::deliverWebIntent):
1063         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
1064         (TestRunner):
1065         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
1066         (WebTestRunner::WebTestProxyBase::postAccessibilityNotification):
1067         (WebTestRunner::WebTestProxyBase::registerIntentService):
1068         (WebTestRunner):
1069         (WebTestRunner::WebTestProxyBase::dispatchIntent):
1070         * DumpRenderTree/chromium/WebViewHost.cpp:
1071         (WebViewHost::setCurrentWebIntentRequest):
1072         (WebViewHost::currentWebIntentRequest):
1073         * DumpRenderTree/chromium/WebViewHost.h:
1074         (WebKit):
1075         (WebViewHost):
1076
1077 2012-12-18  Jochen Eisinger  <jochen@chromium.org>
1078
1079         [chromium] move editing callback dumping to the WebTestProxy
1080         https://bugs.webkit.org/show_bug.cgi?id=105013
1081
1082         Reviewed by Darin Fisher.
1083
1084         To decide whether or not to dump the editing callbacks, the WebTestProxy
1085         needs to query the TestRunner. For this purpose, I introduce a
1086         WebTestRunner class that just exposes the shouldDumpEditingCallbacks
1087         method. In contrast to e.g. WebEventSender, the WebTestRunner is just
1088         an interface, and it is not owned by the WebTestInterfaces, but the
1089         embedder of the TestRunner library must provide it. This will change
1090         once all TestRunner methods are in the library.
1091
1092         * DumpRenderTree/DumpRenderTree.gypi:
1093         * DumpRenderTree/chromium/DRTTestRunner.cpp:
1094         (DRTTestRunner::DRTTestRunner):
1095         (DRTTestRunner::reset):
1096         * DumpRenderTree/chromium/DRTTestRunner.h:
1097         (DRTTestRunner):
1098         (DRTTestRunner::taskList):
1099         (DRTTestRunner::WorkQueue::taskList):
1100         * DumpRenderTree/chromium/TestRunner/public/WebTestInterfaces.h:
1101         (WebTestRunner):
1102         (WebTestInterfaces):
1103         * DumpRenderTree/chromium/TestRunner/public/WebTestProxy.h:
1104         (WebKit):
1105         (WebTestRunner):
1106         (WebTestProxyBase):
1107         (WebTestRunner::WebTestProxy::shouldBeginEditing):
1108         (WebTestRunner::WebTestProxy::shouldEndEditing):
1109         (WebTestRunner::WebTestProxy::shouldInsertNode):
1110         (WebTestRunner::WebTestProxy::shouldInsertText):
1111         (WebTestRunner::WebTestProxy::shouldChangeSelectedRange):
1112         (WebTestRunner::WebTestProxy::shouldDeleteRange):
1113         (WebTestRunner::WebTestProxy::shouldApplyStyle):
1114         (WebTestRunner::WebTestProxy::didBeginEditing):
1115         (WebTestRunner::WebTestProxy::didChangeSelection):
1116         (WebTestRunner::WebTestProxy::didChangeContents):
1117         (WebTestRunner::WebTestProxy::didEndEditing):
1118         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h: Copied from Tools/DumpRenderTree/chromium/TestRunner/public/WebTestInterfaces.h.
1119         (WebTestRunner):
1120         (WebTestRunner::WebTestRunner::shouldDumpEditingCallbacks):
1121         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
1122         (WebTestRunner::TestRunner::TestRunner):
1123         (WebTestRunner::TestRunner::~TestRunner):
1124         (WebTestRunner):
1125         (WebTestRunner::TestRunner::reset):
1126         (WebTestRunner::TestRunner::shouldDumpEditingCallbacks):
1127         (WebTestRunner::TestRunner::dumpEditingCallbacks):
1128         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
1129         (TestRunner):
1130         * DumpRenderTree/chromium/TestRunner/src/WebTestInterfaces.cpp:
1131         (WebTestRunner::WebTestInterfaces::Internal::testRunner):
1132         (WebTestRunner::WebTestInterfaces::Internal::setTestRunner):
1133         (WebTestInterfaces::Internal):
1134         (WebTestRunner::WebTestInterfaces::Internal::Internal):
1135         (WebTestRunner::WebTestInterfaces::testRunner):
1136         (WebTestRunner):
1137         (WebTestRunner::WebTestInterfaces::setTestRunner):
1138         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
1139         (WebTestRunner::WebTestProxyBase::postAccessibilityNotification):
1140         (WebTestRunner):
1141         (WebTestRunner::WebTestProxyBase::shouldBeginEditing):
1142         (WebTestRunner::WebTestProxyBase::shouldEndEditing):
1143         (WebTestRunner::WebTestProxyBase::shouldInsertNode):
1144         (WebTestRunner::WebTestProxyBase::shouldInsertText):
1145         (WebTestRunner::WebTestProxyBase::shouldChangeSelectedRange):
1146         (WebTestRunner::WebTestProxyBase::shouldDeleteRange):
1147         (WebTestRunner::WebTestProxyBase::shouldApplyStyle):
1148         (WebTestRunner::WebTestProxyBase::didBeginEditing):
1149         (WebTestRunner::WebTestProxyBase::didChangeSelection):
1150         (WebTestRunner::WebTestProxyBase::didChangeContents):
1151         (WebTestRunner::WebTestProxyBase::didEndEditing):
1152         * DumpRenderTree/chromium/TestShell.cpp:
1153         (TestShell::initialize):
1154         (TestShell::createNewWindow):
1155         * DumpRenderTree/chromium/WebViewHost.cpp:
1156         (WebViewHost::shouldBeginEditing):
1157         (WebViewHost::shouldEndEditing):
1158         (WebViewHost::shouldInsertNode):
1159         (WebViewHost::shouldInsertText):
1160         (WebViewHost::shouldChangeSelectedRange):
1161         (WebViewHost::shouldDeleteRange):
1162         (WebViewHost::shouldApplyStyle):
1163         * DumpRenderTree/chromium/WebViewHost.h:
1164         (WebViewHost):
1165
1166 2012-12-17  Jochen Eisinger  <jochen@chromium.org>
1167
1168         [chromium] Fix methods in TestRunner that directly modify the settings instead of going through WebPreferences
1169         https://bugs.webkit.org/show_bug.cgi?id=105154
1170
1171         Reviewed by Tony Chang.
1172
1173         If the settings are modified directly, the embedder can't keep track of
1174         them and might accidentially override the changed settings.
1175
1176         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
1177         (WebPreferences):
1178         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
1179         (WebTestRunner::TestRunner::TestRunner):
1180         (WebTestRunner::TestRunner::setAsynchronousSpellCheckingEnabled):
1181         (WebTestRunner):
1182         (WebTestRunner::TestRunner::setMinimumTimerInterval):
1183         (WebTestRunner::TestRunner::setTouchDragDropEnabled):
1184         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
1185         (TestRunner):
1186         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
1187         (WebTestRunner::WebPreferences::reset):
1188         (WebTestRunner::WebPreferences::applyTo):
1189
1190 2012-12-17  Julien Chaffraix  <jchaffraix@webkit.org>
1191
1192         Unreviewed unit test skipping.
1193
1194         Filed https://bugs.webkit.org/show_bug.cgi?id=105253 to re-enable them.
1195
1196         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
1197         (TestWebKitAPI):
1198         (TestWebKitAPI::TEST):
1199
1200 2012-12-17  Dean Jackson  <dino@apple.com>
1201
1202         Unreviewed. Add Silvia Pfeiffer to contributor list.
1203
1204         * Scripts/webkitpy/common/config/committers.py:
1205
1206 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
1207
1208         build.webkit.org - add more google windows bots
1209         https://bugs.webkit.org/show_bug.cgi?id=105225
1210
1211         Reviewed by Eric Seidel.
1212
1213         Adds two more windows bots (google-windows-3, a builder, and
1214         google-windows-4, a tester) for the Chromium configs.
1215
1216         * BuildSlaveSupport/build.webkit.org-config/config.json:
1217
1218 2012-12-17  Eric Seidel  <eric@webkit.org>
1219
1220         Cleanup --profiler=perf code a little and allow up to 1000 report files (instead of 100)
1221         https://bugs.webkit.org/show_bug.cgi?id=105217
1222
1223         Reviewed by Dirk Pranke.
1224
1225         Previously --profile would fail strangely if find_unused_filename returned None
1226         (which would happen as soon as we had 100 report files).
1227         Now we'll search through 1000 filenames before giving up (and assert if we failed to find one).
1228         I also made some minor cleanups to the output of perf, and tried to make the
1229         commands more self-documenting by using long-names for args.
1230
1231         * Scripts/webkitpy/common/system/profiler.py:
1232         (SingleFileOutputProfiler.__init__):
1233         (Perf.attach_to_pid):
1234         (Perf.profile_after_exit):
1235
1236 2012-12-17  Sheriff Bot  <webkit.review.bot@gmail.com>
1237
1238         Unreviewed, rolling out r137198.
1239         http://trac.webkit.org/changeset/137198
1240         https://bugs.webkit.org/show_bug.cgi?id=105212
1241
1242         This patch is causing API behavior compatibility problems
1243         (Requested by zdobersek on #webkit).
1244
1245         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1246         (willSendRequestCallback):
1247
1248 2012-12-17  Elliott Sprehn  <esprehn@chromium.org>
1249
1250         Unreviewed. Adding myself to committers list.
1251
1252         * Scripts/webkitpy/common/config/committers.py:
1253
1254 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
1255
1256         nrwt: shuffle code around for cleanup in run_webkit_tests.py
1257         https://bugs.webkit.org/show_bug.cgi?id=105078
1258
1259         Reviewed by Ojan Vafai.
1260
1261         This patch reorders functions so that run_webkit_tests can be
1262         understood in a top-down matter better, and to make things slightly
1263         more sensible when I merge manager.py into it.
1264
1265         Also, this adds tests for the actual main() routine, which had several
1266         bugs get through testing a week or two ago.
1267
1268         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1269         (main):
1270         (parse_args):
1271         (_set_up_derived_options):
1272         (run):
1273         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1274         (RunTest):
1275         (RunTest.test_no_http_tests):
1276         (PortTest.disabled_test_mac_lion):
1277         (MainTest):
1278         (MainTest.test_exception_handling):
1279         (MainTest.test_exception_handling.interrupting_run):
1280         (MainTest.test_exception_handling.successful_run):
1281         (MainTest.test_exception_handling.successful_run.FakeRunDetails):
1282         (MainTest.test_exception_handling.exception_raising_run):
1283
1284 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
1285
1286         webkitpy: move --lint-test-files code into its own module
1287         https://bugs.webkit.org/show_bug.cgi?id=105077
1288
1289         Reviewed by Eric Seidel.
1290
1291         It really didn't need to be in run_webkit_tests.py anymore.
1292         This patch adds a new lint-test-expectations scripts to Tools/Scripts;
1293         rwt --lint-test-files is still supported for backwards compatibility
1294         for now.
1295
1296         * Scripts/lint-test-expectations: Added.
1297         * Scripts/webkitpy/layout_tests/lint_test_expectations.py: Added.
1298         (lint):
1299         (main):
1300         * Scripts/webkitpy/layout_tests/lint_test_expectations_unittest.py: Added.
1301         (FakePort):
1302         (FakePort.__init__):
1303         (FakePort.test_configuration):
1304         (FakePort.expectations_dict):
1305         (FakePort.skipped_layout_tests):
1306         (FakePort.all_test_configurations):
1307         (FakePort.configuration_specifier_macros):
1308         (FakePort.get_option):
1309         (FakeFactory):
1310         (FakeFactory.__init__):
1311         (FakeFactory.get):
1312         (FakeFactory.all_port_names):
1313         (LintTest):
1314         (LintTest.test_all_configurations):
1315         (LintTest.test_lint_test_files):
1316         (LintTest.test_lint_test_files__errors):
1317         (MainTest):
1318         (MainTest.test_success):
1319         (MainTest.test_success.interrupting_lint):
1320         (MainTest.test_success.successful_lint):
1321         (MainTest.test_success.exception_raising_lint):
1322         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1323         (main):
1324         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1325         (StreamTestingMixin.assertNotEmpty):
1326
1327 2012-12-17  Adam Klein  <adamk@chromium.org>
1328
1329         build-webkit: rename --template-tag to --template-element to match ENABLE #define name
1330         https://bugs.webkit.org/show_bug.cgi?id=105072
1331
1332         Reviewed by Laszlo Gombos.
1333
1334         * Scripts/webkitperl/FeatureList.pm:
1335
1336 2012-12-17  Julien Chaffraix  <jchaffraix@webkit.org>
1337
1338         Add some unit testing for WTF::clampTo* functions
1339         https://bugs.webkit.org/show_bug.cgi?id=105060
1340
1341         Reviewed by Emil A Eklund.
1342
1343         The new tests are targeted at clamping to int / unsigned and float as those are
1344         the one explicitely exposed through more dedicated functions (like clampToInteger,
1345         clampToPositiveInteger and clampToFloat).
1346
1347         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
1348
1349 2012-12-17  Eric Seidel  <eric@webkit.org>
1350
1351         rpt --profile --chromium-android throws exception
1352         https://bugs.webkit.org/show_bug.cgi?id=105061
1353
1354         Reviewed by Daniel Bates.
1355
1356         Somehow popen(args) got turned into popen(*args) at some point,
1357         which caused my new stringify code to fail when args was passed by name.
1358         Fixed and tested.
1359
1360         * Scripts/webkitpy/common/system/executive.py:
1361         (Executive._stringify_args):
1362         (Executive):
1363         (Executive.popen):
1364         * Scripts/webkitpy/common/system/executive_unittest.py:
1365         (ExecutiveTest.test_popen_args):
1366
1367 2012-12-17  Mark Pilgrim  <pilgrim@chromium.org>
1368
1369         Use Platform::current() instead of webKitPlatformSupport() in DumpRenderTree
1370         https://bugs.webkit.org/show_bug.cgi?id=105054
1371
1372         Reviewed by Jochen Eisinger.
1373
1374         Part of a refactoring series; see tracking bug 82948.
1375
1376         * DumpRenderTree/chromium/NotificationPresenter.cpp:
1377         (NotificationPresenter::show):
1378         * DumpRenderTree/chromium/Task.cpp:
1379         (postTask):
1380         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp:
1381
1382 2012-12-17  Emil A Eklund  <eae@chromium.org>
1383
1384         Clamp values in LayoutUnit::operator/ when SATURATED_LAYOUT_ARITHMETIC is enabled
1385         https://bugs.webkit.org/show_bug.cgi?id=104955
1386
1387         Reviewed by Julien Chaffraix.
1388
1389         Add tests for LayoutUnit.
1390
1391         * TestWebKitAPI/CMakeLists.txt:
1392         * TestWebKitAPI/ForwardingHeaders: Added.
1393         * TestWebKitAPI/ForwardingHeaders/WebCore: Added.
1394         * TestWebKitAPI/ForwardingHeaders/WebCore/LayoutUnit.h: Added.
1395         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
1396         * TestWebKitAPI/TestWebKitAPI.gypi:
1397         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1398         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp: Added.
1399         (TestWebKitAPI):
1400         (TestWebKitAPI::TEST):
1401         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
1402
1403 2012-12-17  Jochen Eisinger  <jochen@chromium.org>
1404
1405         [chromium] removed unused preferences from testRunner.overridePreference
1406         https://bugs.webkit.org/show_bug.cgi?id=105169
1407
1408         Reviewed by Tony Chang.
1409
1410         Most notably this removes the code related to font family maps which is
1411         now handled via window.internals.settings
1412
1413         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
1414         (WebPreferences):
1415         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
1416         (WebTestRunner::TestRunner::overridePreference):
1417         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
1418         (TestRunner):
1419         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
1420         (WebTestRunner::WebPreferences::applyTo):
1421
1422 2012-12-17  James Robinson  <jamesr@chromium.org>
1423
1424         [chromium] Move creation of WebLayerTreeView to WebWidgetClient
1425         https://bugs.webkit.org/show_bug.cgi?id=105071
1426
1427         Reviewed by Adrienne Walker.
1428
1429         Update WebViewHost for the new WebWidgetClient interface.
1430
1431         * DumpRenderTree/chromium/WebViewHost.cpp:
1432         (WebViewHost::initializeLayerTreeView):
1433         (WebViewHost::layerTreeView):
1434         (WebViewHost::~WebViewHost):
1435         * DumpRenderTree/chromium/WebViewHost.h:
1436         (WebViewHost):
1437
1438 2012-12-17  Kenneth Rohde Christiansen  <kenneth@webkit.org>
1439
1440         [EFL][WK2] window_create doesn't receive the url
1441         https://bugs.webkit.org/show_bug.cgi?id=105184
1442
1443         Reviewed by Alexis Menard.
1444
1445         * MiniBrowser/efl/main.c: Use the url supplied to window_create
1446         (on_window_create):
1447
1448 2012-12-17  Ilya Tikhonovsky  <loislo@chromium.org>
1449
1450         Web Inspector: Native Memory Instrumentation: MemoryInstrumentation doesn't detect reportMemoryUsage method defined in a base class.
1451         https://bugs.webkit.org/show_bug.cgi?id=105026
1452
1453         Reviewed by Yury Semikhatsky.
1454
1455         Old SFINAE test was replaced with new one based on this article:
1456         http://stackoverflow.com/questions/1966362/sfinae-to-check-for-inherited-member-functions
1457
1458         New test which covers this problem was added.
1459
1460         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
1461
1462 2012-12-17  Anthony Scian  <ascian@rim.com>
1463
1464         [Blackberry] Static code analysis warning fixes
1465         https://bugs.webkit.org/show_bug.cgi?id=105129
1466
1467         Reviewed by Rob Buis.
1468
1469         Added null checks for fopen calls, no recovery attempted; access fault prevention only.
1470
1471         * DumpRenderTree/blackberry/DumpRenderTree.cpp:
1472         (BlackBerry::WebKit::createFile):
1473         (BlackBerry::WebKit::DumpRenderTree::runTest):
1474
1475 2012-12-17  John Mellor  <johnme@chromium.org>
1476
1477         Add Anton Vayvod and Tim Volodine as contributors in committers.py
1478         https://bugs.webkit.org/show_bug.cgi?id=105025
1479
1480         Unreviewed. Adds two contributors to committers.py.
1481         They are initially contributing to the Text Autosizing feature (master bug: 84186).
1482
1483         * Scripts/webkitpy/common/config/committers.py:
1484
1485 2012-12-17  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
1486
1487         Unreviewed, add myself to committers.py.
1488
1489         * Scripts/webkitpy/common/config/committers.py:
1490
1491 2012-12-17  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
1492
1493         [Qt] remove some unnecessary CONFIG additions
1494
1495         Reviewed by Simon Hausmann.
1496
1497         qt is already added by spec_pre.prf, warn_on and depend_includepath by
1498         default_pre.prf.
1499
1500         * qmake/mkspecs/features/default_post.prf:
1501
1502 2012-12-17  Ilya Tikhonovsky  <loislo@chromium.org>
1503
1504         Unreviewed, rolling out r137892.
1505         http://trac.webkit.org/changeset/137892
1506         https://bugs.webkit.org/show_bug.cgi?id=105026
1507
1508         it broke compilation on windows
1509
1510         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
1511
1512 2012-12-14  Ilya Tikhonovsky  <loislo@chromium.org>
1513
1514         Web Inspector: Native Memory Instrumentation: MemoryInstrumentation doesn't detect reportMemoryUsage method defined in a base class.
1515         https://bugs.webkit.org/show_bug.cgi?id=105026
1516
1517         Reviewed by Yury Semikhatsky.
1518
1519         Old SFINAE test was replaced with new one based on this article:
1520         http://stackoverflow.com/questions/1966362/sfinae-to-check-for-inherited-member-functions
1521
1522         New test which covers this problem was added.
1523
1524         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
1525
1526 2012-12-17  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
1527
1528         [Qt] rely on automatic output directory setup
1529
1530         Reviewed by Simon Hausmann.
1531
1532         Now being a proper qt module (by virtue of having load(qt_build_config)
1533         in .qmake.conf), webkit gets the path setup goodies for free.
1534
1535         This also fixes Makefile.api.Debug/Release trying to generate qrc_WebKit.cpp
1536         at the same time, instead of the debug and release version ending up in different
1537         directories as they should.
1538
1539         * qmake/mkspecs/features/default_pre.prf:
1540
1541 2012-12-17  Charles Wei  <charles.wei@torchmobile.com.cn>
1542
1543         [BlackBerry] Need to re-enable video track
1544         https://bugs.webkit.org/show_bug.cgi?id=105155
1545
1546         Reviewed by George Staikos.
1547
1548         Just enable Video Track for BlackBerry.
1549
1550         * Scripts/webkitperl/FeatureList.pm:
1551
1552 2012-12-17  János Badics  <jbadics@inf.u-szeged.hu>
1553
1554         Unreviewed. Added myself to committers.py
1555
1556         * Scripts/webkitpy/common/config/committers.py:
1557
1558 2012-12-17  Jonathan Liu  <net147@gmail.com>
1559
1560         Fix linking of libraries with -Wl,-whole-archive
1561         https://bugs.webkit.org/show_bug.cgi?id=104436
1562
1563         Reviewed by Simon Hausmann.
1564
1565         If there is more than one library in LIBS linked with
1566         -Wl,-whole-archive -l... -Wl,-no-whole-archive, qmake tries to
1567         merge the duplicate flags in LIBS which causes linking to fail.
1568
1569         Add no_smart_library_merge to CONFIG to prevent the duplicate
1570         flags from being removed when using -Wl,-whole-archive.
1571
1572         * qmake/mkspecs/features/functions.prf:
1573
1574 2012-12-16  Arvid Nilsson  <anilsson@rim.com>
1575
1576         [BlackBerry] Adapt to new BlackBerry::Platform::TouchPoint API
1577         https://bugs.webkit.org/show_bug.cgi?id=105143
1578         RIM PR 171941
1579
1580         Reviewed by Rob Buis.
1581         Internally reviewed by George Staikos.
1582
1583         Adapt to new method names and encapsulation of TouchPoint data members.
1584
1585         * DumpRenderTree/blackberry/EventSender.cpp:
1586         (addTouchPointCallback):
1587         (updateTouchPointCallback):
1588         (touchEndCallback):
1589         (releaseTouchPointCallback):
1590         (sendTouchEvent):
1591
1592 2012-12-16  Maciej Stachowiak  <mjs@apple.com>
1593
1594         Add a script to compute the mean and 95% confidence interval (using two-sided t-test) of a sample set
1595         https://bugs.webkit.org/show_bug.cgi?id=105148
1596
1597         Reviewed by Filip Pizlo.
1598
1599         Usage examples:
1600         
1601         $ sampstat --help
1602         Usage: sampstat [options] [FILES]
1603           Compute the mean and 95% confidence interval of a sample set.
1604           Standard input or files must contain two or more decimal numbers, one per line.
1605         
1606         Options:
1607           -h, --help            show this help message and exit
1608           -u UNIT, --unit=UNIT  assume values are in units of UNIT
1609           -v, --verbose         print all values (with units)
1610         
1611         $ sampstat -u MB memresults.txt
1612         2356.90 MB +/- 101.34 MB (4.3%)
1613         
1614         $ sampstat -v -u MB memresults.txt
1615               2318.21 MB
1616               2399.56 MB
1617               2352.93 MB
1618         ----------------
1619         Mean: 2356.90 MB +/- 101.34 MB (4.3%)
1620         
1621
1622         * Scripts/sampstat: Added.
1623         (sum): Helper function to compute the sum of a list.
1624         (arithmeticMean): Compute the meam of a list.
1625         (standardDeviation): Compute the sample standard deviation (unbiased estimator).
1626         (standardError): Compute the sample standard error.
1627         (tDist): Compute t(.025, n-1), the t-value for atwo-sided 95% confidence interval. 
1628         (twoSidedConfidenceInterval): Compute the two-sided confidence interval range about the mean.
1629
1630 2012-12-15  Nima Ghanavatian  <nghanavatian@rim.com>
1631
1632         [BlackBerry] Pass key modifiers with touch and mouse events
1633         https://bugs.webkit.org/show_bug.cgi?id=105108
1634
1635         Reviewed by Rob Buis.
1636
1637         Update DRT to new api.
1638
1639         Internally reviewed by Gen Mak.
1640
1641         * DumpRenderTree/blackberry/EventSender.cpp:
1642         (mouseDownCallback):
1643         (mouseUpCallback):
1644         (mouseMoveToCallback):
1645
1646 2012-12-15  Zan Dobersek  <zandobersek@gmail.com>
1647
1648         Stop the XvfbDriver before proceeding with its starting setup
1649         https://bugs.webkit.org/show_bug.cgi?id=105006
1650
1651         Reviewed by Philippe Normand.
1652
1653         Just as with the base Driver implementation, stop the XvfbDriver
1654         before proceeding with setting up lock files and a new Xvfb instance.
1655         This forces any driver or Xvfb instances that are currently under
1656         the driver's control to be closed/killed and cleaned up after.
1657
1658         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
1659         (XvfbDriver._start):
1660
1661 2012-12-14  Dirk Pranke  <dpranke@chromium.org>
1662
1663         webkitpy: make port_name a required parameter to the non-test ports
1664         https://bugs.webkit.org/show_bug.cgi?id=105057
1665
1666         Reviewed by Eric Seidel.
1667
1668         This is just minor code cleanup along the path to making Port
1669         constructors simpler and pushing more logic into the port factory.
1670
1671         * Scripts/webkitpy/layout_tests/port/base.py:
1672         (Port.__init__):
1673         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
1674         (PortTest.make_port):
1675         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
1676         (TestWebKitPort.__init__):
1677         * Scripts/webkitpy/layout_tests/port/test.py:
1678         (TestPort):
1679         (TestPort.determine_full_port_name):
1680         (TestPort.__init__):
1681         * Scripts/webkitpy/layout_tests/port/xvfbdriver_unittest.py:
1682         (XvfbDriverTest.make_driver):
1683         (XvfbDriverTest.test_stop):
1684
1685 2012-12-14  Dirk Pranke  <dpranke@chromium.org>
1686
1687         garden-o-matic doesn't know about reftests
1688         https://bugs.webkit.org/show_bug.cgi?id=101976
1689
1690         Reviewed by Eric Seidel.
1691
1692         Attempt to make garden-o-matic properly aware of reftests again.
1693
1694         This is re-landing r137407 with a change to
1695         rebaselineWithStatusUpdates() to actually pass the filtered
1696         failureInofLost to checkout.rebaseline().
1697
1698         Unfortunately, the testing coverage for this whole module is almost
1699         non-existent, and it's not obvious that there's a good way to
1700         stub out checkout.rebaseline() to test that the right thing
1701         happens. I verified the change manually.
1702
1703         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
1704         (.):
1705         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
1706         (.):
1707         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
1708
1709 2012-12-14  Ryosuke Niwa  <rniwa@webkit.org>
1710
1711         "Running 1 DumpRenderTree over X shards" is not a helpful output
1712         https://bugs.webkit.org/show_bug.cgi?id=104858
1713
1714         Reviewed by Dirk Pranke.
1715
1716         The word "shard" is not typically used in test frameworks. This word is confusing for many WebKit
1717         contributors who typically don't have much experience on distributed computations and databases.
1718         So only print this message out in the debug mode.
1719
1720         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1721         * Scripts/webkitpy/layout_tests/views/printing.py:
1722         (Printer.print_workers_and_shards):
1723
1724 2012-12-13  Eric Seidel  <eric@webkit.org>
1725
1726         Add --profiler=PROFILER option to run-perf-tests to allow specifying which profiler to use on platforms with many
1727         https://bugs.webkit.org/show_bug.cgi?id=104891
1728
1729         Reviewed by Ryosuke Niwa.
1730
1731         I also implemented a very simple "Sample" Profiler using
1732         Mac OS X's /usr/bin/sample command line tool.
1733
1734         The real reason for this abstraction is to make it easy
1735         to support both perf and pprof on linux which seem to
1736         be about equally popular among those I ask in the Chrome team.
1737
1738         * Scripts/webkitpy/common/system/profiler.py:
1739         (ProfilerFactory.create_profiler):
1740         (ProfilerFactory):
1741         (ProfilerFactory.available_profilers_by_name):
1742         (ProfilerFactory.default_profiler_name):
1743         (Sample):
1744         (Sample.__init__):
1745         (Sample.attach_to_pid):
1746         (Sample.profile_after_exit):
1747         (IProfiler.attach_to_pid):
1748         * Scripts/webkitpy/layout_tests/port/driver.py:
1749         (Driver.__init__):
1750         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1751         (PerfTestsRunner._parse_args):
1752
1753 2012-12-13  Eric Seidel  <eric@webkit.org>
1754
1755         Unreviewed, fixing typo in python unittest.
1756
1757         I changed the default profiler on linux from pprof to perf in
1758         bug 104971.  I failed to update the unittest results at that time.
1759
1760         * Scripts/webkitpy/common/system/profiler_unittest.py:
1761         (ProfilerFactoryTest.test_basic):
1762
1763 2012-12-13  Ian Vollick  <vollick@chromium.org>
1764
1765         [chromium] Add a virtual test suite for enabling opt-in to composited scrolling
1766         https://bugs.webkit.org/show_bug.cgi?id=104911
1767
1768         Reviewed by Dirk Pranke.
1769
1770         Adds two virtual tests suites to run the tests in
1771         compositing/overflow/ and scrollbars/ with the flag
1772         --enable-accelerated-overflow-scroll, which I've also plumbed through
1773         in the usual way.
1774
1775         * DumpRenderTree/chromium/DumpRenderTree.cpp:
1776         (main):
1777         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
1778         (WebPreferences):
1779         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
1780         (WebTestRunner::WebPreferences::reset):
1781         (WebTestRunner::WebPreferences::applyTo):
1782         * DumpRenderTree/chromium/TestShell.cpp:
1783         (TestShell::TestShell):
1784         (TestShell::resetWebSettings):
1785         * DumpRenderTree/chromium/TestShell.h:
1786         (TestShell::setAcceleratedCompositingForOverflowScrollEnabled):
1787         (TestShell):
1788         * Scripts/webkitpy/layout_tests/port/chromium.py:
1789         (ChromiumPort.virtual_test_suites):
1790
1791 2012-12-13  Eric Seidel  <eric@webkit.org>
1792
1793         Use 'perf' to profile on linux instead of google-pprof by default
1794         https://bugs.webkit.org/show_bug.cgi?id=104971
1795
1796         Reviewed by Dirk Pranke.
1797
1798         This makes Chromium Linux match Chromium Android and use the perf
1799         tool by default.  Once https://bugs.webkit.org/show_bug.cgi?id=104891
1800         lands it will be possible to use pprof again on Linux.
1801
1802         This is slightly more advanced than the AndroidPerf profiler as
1803         instead of using a timeout on "perf record" I instead watch
1804         for the termination of the target process and then control-C
1805         the 'perf record' process.  This required me to add two new
1806         methods to Executive, one to have a limited-time wait() and
1807         the second to be able to send a control-C.  I chose to add
1808         these to Executive to make them easier to mock/fix-for-win32
1809         at a later time if needed.
1810
1811         * Scripts/webkitpy/common/system/executive.py:
1812         (Executive.wait_limited):
1813         (Executive.interrupt):
1814         * Scripts/webkitpy/common/system/profiler.py:
1815         (ProfilerFactory.create_profiler):
1816         (Perf):
1817         (Perf.__init__):
1818         (Perf._perf_path):
1819         (Perf.attach_to_pid):
1820         (Perf._first_ten_lines_of_profile):
1821         (Perf.profile_after_exit):
1822
1823 2012-12-13  Julie Parent  <jparent@chromium.org>
1824
1825         Dashboard cleanup: remove usage of global g_builders.
1826         https://bugs.webkit.org/show_bug.cgi?id=104941
1827
1828         Reviewed by Dirk Pranke.
1829
1830         The dashboards use a lot of global state, which makes hacking on them
1831         complicated. This change removes the use of one such global: g_builders.
1832         In most cases, we can just use currentBuilderGroup().builders instead,
1833         which is now currentBuilders().
1834         Surprisingly, the most changes were required to the unit tests, since
1835         they were even bigger offenders of bad hygiene, relying on global state
1836         set by other tests, randomly clobbering global variables in ways the
1837         real code doesn't, etc.
1838
1839         * TestResultServer/static-dashboards/builders.js:
1840         (BuilderGroup.prototype.setup):
1841         * TestResultServer/static-dashboards/dashboard_base.js:
1842         (.switch.return):
1843         (htmlForTestTypeSwitcher):
1844         * TestResultServer/static-dashboards/flakiness_dashboard.js:
1845         (generatePage):
1846         (getAllTestsTrie):
1847         (processTestRunsForAllBuilders):
1848         (showPopupForBuild):
1849         (generatePageForExpectationsUpdate):
1850         (loadExpectationsLayoutTests):
1851         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
1852         (resetGlobals):
1853         (stubResultsByBuilder):
1854         (test):
1855         * TestResultServer/static-dashboards/loader.js:
1856         (.):
1857         * TestResultServer/static-dashboards/loader_unittests.js:
1858
1859 2012-12-13  Eric Seidel  <eric@webkit.org>
1860
1861         Callers should not have to stringify args before calling Executive run_command/popen
1862         https://bugs.webkit.org/show_bug.cgi?id=104975
1863
1864         Reviewed by Dirk Pranke.
1865
1866         One could argue that we should match the python call syntax here,
1867         but I think it's a more friendly API if we automagically handle
1868         stringification of args in run_command, etc.
1869         This removes map(unicode, args) from several callsites.
1870
1871         When I first tried to land this change, I didn't realize that
1872         Executive._command_for_printing depended on this behavior
1873         having been applied to args in run_command.  The fix is to
1874         call _stringify_args in both run_command and popen.
1875         This is slightly redundant, but given how short args have to be
1876         (due to shell limits), I don't think the double-encode check
1877         matters in practice.
1878
1879         This is slightly complicated by the fact that apache_http_server.py
1880         is the one caller in our codebase which uses shell=True.
1881         shell=True is a well-documented trail-of-tears:
1882         http://stackoverflow.com/questions/3172470/actual-meaning-of-shell-true-in-subprocess
1883         but to support this legacy (windows-only) code (which I can't easily test)
1884         I've added an if-hack to avoid stringifying the the popen(shell=True) case.
1885
1886         * Scripts/webkitpy/common/system/executive.py:
1887         (Executive.run_command):
1888         (Executive._stringify_args):
1889         (Executive.popen):
1890         * Scripts/webkitpy/common/system/executive_unittest.py:
1891         (ExecutiveTest.test_auto_stringify_args):
1892         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
1893         (attach_to_pid):
1894
1895 2012-12-13  Adrienne Walker  <enne@chromium.org>
1896
1897         Unreviewed, rolling out r137645, r137646, and r137667.
1898         http://trac.webkit.org/changeset/137645
1899         http://trac.webkit.org/changeset/137646
1900         http://trac.webkit.org/changeset/137667
1901         https://bugs.webkit.org/show_bug.cgi?id=104911
1902
1903         Breaks some overflow layout tests
1904
1905         * Scripts/webkitpy/layout_tests/port/chromium.py:
1906         (ChromiumPort.virtual_test_suites):
1907
1908 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
1909
1910         Unreviewed, rolling out r137661.
1911         http://trac.webkit.org/changeset/137661
1912         https://bugs.webkit.org/show_bug.cgi?id=104891
1913
1914         broke unit tests, run-webkit-tests
1915
1916         * Scripts/webkitpy/common/system/executive.py:
1917         (Executive._run_command_with_teed_output):
1918         (Executive.run_command):
1919         (Executive.popen):
1920         * Scripts/webkitpy/common/system/profiler.py:
1921         (ProfilerFactory.create_profiler):
1922         (GooglePProf.profile_after_exit):
1923         (IProfiler.attach_to_pid):
1924         * Scripts/webkitpy/layout_tests/port/driver.py:
1925         (Driver.__init__):
1926         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1927         (PerfTestsRunner._parse_args):
1928
1929 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
1930
1931         nrwt: move rundetails, summarize_results into test_run_results
1932         https://bugs.webkit.org/show_bug.cgi?id=104963
1933
1934         Reviewed by Eric Seidel.
1935
1936         More cleanup; this the plain-old-data functions out of manager.py
1937         and into test_run_results next to the other aggregate data structures.
1938
1939         Also this cleans up a bunch of test code and reduces duplication.
1940
1941         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1942         (Manager.run):
1943         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
1944         (ManagerTest.test_look_for_new_crash_logs):
1945         * Scripts/webkitpy/layout_tests/models/test_run_results.py:
1946         (TestRunResults.__init__):
1947         (TestRunResults.add):
1948         (RunDetails):
1949         (RunDetails.__init__):
1950         (_interpret_test_failures):
1951         (summarize_results):
1952         * Scripts/webkitpy/layout_tests/models/test_run_results_unittest.py: Added.
1953         (get_result):
1954         (run_results):
1955         (summarized_results):
1956         (InterpretTestFailuresTest):
1957         (InterpretTestFailuresTest.setUp):
1958         (InterpretTestFailuresTest.test_interpret_test_failures):
1959         (SummarizedResultsTest):
1960         (SummarizedResultsTest.setUp):
1961         (SummarizedResultsTest.test_no_svn_revision):
1962         (SummarizedResultsTest.test_svn_revision):
1963         (SummarizedResultsTest.test_summarized_results_wontfix):
1964         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
1965         (BuildBotPrinterTests.get_printer):
1966         (BuildBotPrinterTests.test_print_unexpected_results):
1967         (BuildBotPrinterTests.test_print_results):
1968
1969 2012-12-13  Eric Seidel  <eric@webkit.org>
1970
1971         Make --profile output easier to understand for pprof on linux
1972         https://bugs.webkit.org/show_bug.cgi?id=104957
1973
1974         Reviewed by Dirk Pranke.
1975
1976         pprof is kinda mysterious, so we print some information
1977         on how to use it, and how to learn how to read the sample dump.
1978         This is similar to what we've done for perf output for Chromium Android.
1979
1980         * Scripts/webkitpy/common/system/profiler.py:
1981         (GooglePProf._pprof_path):
1982         (GooglePProf.profile_after_exit):
1983
1984 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
1985
1986         Revert r137407 - it didn't actually work right.
1987         https://bugs.webkit.org/show_bug.cgi?id=101976
1988
1989         Unreviewed, build fix.
1990
1991         Turns out that rebaselining was completely broken with this change :(
1992         We're passing in a list of test names rather than the list of
1993         failureInfo objects.
1994
1995         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
1996         (.):
1997         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
1998         (.):
1999         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
2000
2001 2012-12-13  Antoine Labour  <piman@chromium.org>
2002
2003         [chromium] Small fixes for WebPluginContainer::setWebLayer
2004         https://bugs.webkit.org/show_bug.cgi?id=104953
2005
2006         Reviewed by James Robinson.
2007
2008         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp: Clean up
2009         dangling pointers, force correct invalidation of the layer.
2010
2011 2012-12-13  Eric Seidel  <eric@webkit.org>
2012
2013         Add --profiler=PROFILER option to run-perf-tests to allow specifying which profiler to use on platforms with many
2014         https://bugs.webkit.org/show_bug.cgi?id=104891
2015
2016         Reviewed by Dirk Pranke.
2017
2018         I also implemented a very simple "Sample" Profiler using
2019         Mac OS X's /usr/bin/sample command line tool.
2020
2021         I also moved the map(unicode, args) call into Executive.popen
2022         so that all callers don't have to do it themselves.
2023
2024         The real reason for this abstraction is to make it easy
2025         to support both perf and pprof on linux which seem to
2026         be about equally popular among those I ask in the Chrome team.
2027
2028         * Scripts/webkitpy/common/system/executive.py:
2029         (Executive.popen):
2030         * Scripts/webkitpy/common/system/profiler.py:
2031         (ProfilerFactory.create_profiler):
2032         (ProfilerFactory):
2033         (ProfilerFactory.available_profilers_by_name):
2034         (ProfilerFactory.default_profiler_name):
2035         (Sample):
2036         (Sample.__init__):
2037         (Sample.attach_to_pid):
2038         (Sample.profile_after_exit):
2039         (IProfiler.attach_to_pid):
2040         * Scripts/webkitpy/layout_tests/port/driver.py:
2041         (Driver.__init__):
2042         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2043         (PerfTestsRunner._parse_args):
2044
2045 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
2046
2047         support -wk2 port names properly in webkitpy.layout_tests.port
2048         https://bugs.webkit.org/show_bug.cgi?id=104761
2049
2050         Reviewed by Eric Seidel.
2051
2052         Our support for WK2 port names was inconsistent at best; this patch
2053         modifies the code so that <port_name>-wk2 is supported for all ports that
2054         have WK2 implementations and is equivalent to specifying <port_name>
2055         and -2/--webkit-test-runner. In addition, this modifies
2056         builders.all_port_names() to include the wk2 variants.
2057
2058         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
2059         * Scripts/webkitpy/layout_tests/port/apple.py:
2060         (ApplePort.determine_full_port_name):
2061         * Scripts/webkitpy/layout_tests/port/base.py:
2062         (Port.determine_full_port_name):
2063         (Port.__init__):
2064         * Scripts/webkitpy/layout_tests/port/builders.py:
2065         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
2066         (DriverTest.make_port):
2067         * Scripts/webkitpy/layout_tests/port/mac.py:
2068         (MacPort.default_baseline_search_path):
2069         * Scripts/webkitpy/layout_tests/port/win.py:
2070         (WinPort.default_baseline_search_path):
2071         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
2072         (TestRebaselineExpectations.test_rebaseline_expectations):
2073
2074 2012-12-13  Ian Vollick  <vollick@chromium.org>
2075
2076         [chromium] Add a virtual test suite for enabling opt-in to composited scrolling
2077         https://bugs.webkit.org/show_bug.cgi?id=104911
2078
2079         Reviewed by Adrienne Walker.
2080
2081         Adds two virtual tests suites to run the tests in
2082         compositing/overflow/ and scrollbars/ with the flag
2083         --enable-accelerated-compositing-for-overflow-scroll
2084
2085         * Scripts/webkitpy/layout_tests/port/chromium.py:
2086         (ChromiumPort.virtual_test_suites):
2087
2088 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
2089
2090         delete extraneous self.fail in mastercfg_unittest.py
2091         https://bugs.webkit.org/show_bug.cgi?id=104940
2092
2093         Reviewed by Eric Seidel.
2094
2095         This was accidentally introduced in r137404.
2096
2097         * BuildSlaveSupport/build.webkit.org-config/mastercfg_unittest.py:
2098
2099 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
2100
2101         Fix a mis-named variable after r137621.
2102         https://bugs.webkit.org/show_bug.cgi?id=104871
2103
2104         Unreviewed, build fix.
2105
2106         * Scripts/webkitpy/layout_tests/views/buildbot_results.py:
2107         (BuildBotPrinter.print_results):
2108         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
2109
2110 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
2111
2112         webkitpy: rename ResultSummary
2113         https://bugs.webkit.org/show_bug.cgi?id=104871
2114
2115         Reviewed by Eric Seidel.
2116
2117         This patch changes the names of ResultSummary and related modules,
2118         classes, and variables as a part of the larger clean up in bug 103824.
2119
2120         This change is basically a bunch of search and replaces:
2121         - result_summary.py -> test_run_results.py
2122         - class ResultSummary -> class TestRunResults
2123         - "result_summary" or "rs" as a generic variable name -> "run_results",
2124           when it doesn't matter (or we don't know) if this refers to the
2125           initial pass over the tests or the retry pass.
2126         - "result_summary" as a variable containing the results of the initial,
2127           complete test run -> "initial_results"
2128         - "retry_summary" as a variable containing the results of retried
2129           results -> "retry_results"
2130         - ResultSummary.results (the property) -> TestRunResults.results_by_name
2131         - ResultSummary.failures (the property) -> TestRunResults.failures_by_name
2132         - RunDetails.result_summary -> RunDetails.initial_results
2133         - RunDetails.retry_summary -> RunDetails.retry_results
2134
2135         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
2136         (LayoutTestRunner.__init__):
2137         (LayoutTestRunner.run_tests):
2138         (LayoutTestRunner._mark_interrupted_tests_as_skipped):
2139         (LayoutTestRunner._interrupt_if_at_failure_limits):
2140         (LayoutTestRunner._interrupt_if_at_failure_limits.interrupt_if_at_failure_limit):
2141         (LayoutTestRunner._update_summary_with_result):
2142         (LayoutTestRunner._handle_finished_test):
2143         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner_unittest.py:
2144         (FakePrinter.print_expected):
2145         (LayoutTestRunnerTests.test_interrupt_if_at_failure_limits):
2146         (LayoutTestRunnerTests.test_update_summary_with_result):
2147         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2148         (RunDetails.__init__):
2149         (summarize_results):
2150         (Manager.run):
2151         (Manager._look_for_new_crash_logs):
2152         (Manager._tests_to_retry):
2153         (Manager._upload_json_files):
2154         (Manager._stats_trie):
2155         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
2156         (ManagerTest.test_look_for_new_crash_logs):
2157         (SummarizeResultsTest):
2158         (SummarizeResultsTest.get_run_results):
2159         (SummarizeResultsTest.summarized_results):
2160         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
2161         (JSONLayoutResultsGenerator.__init__):
2162         (JSONLayoutResultsGenerator._get_modifier_char):
2163         (JSONLayoutResultsGenerator._insert_failure_summaries):
2164         (JSONLayoutResultsGenerator._get_failure_summary_entry):
2165         * Scripts/webkitpy/layout_tests/models/test_run_results.py: Renamed from Tools/Scripts/webkitpy/layout_tests/models/result_summary.py.
2166         (TestRunResults):
2167         (TestRunResults.__init__):
2168         (TestRunResults.add):
2169         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2170         (get_test_results):
2171         (MainTest.test_basic):
2172         * Scripts/webkitpy/layout_tests/views/buildbot_results.py:
2173         (BuildBotPrinter.print_results):
2174         (BuildBotPrinter.print_run_results):
2175         (BuildBotPrinter._print_run_results_entry):
2176         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
2177         (BuildBotPrinterTests.get_run_results):
2178         (BuildBotPrinterTests.test_print_unexpected_results.get_unexpected_results):
2179         (BuildBotPrinterTests):
2180         * Scripts/webkitpy/layout_tests/views/printing.py:
2181         (Printer.print_expected):
2182         (Printer._print_expected_results_of_type):
2183         (Printer.print_results):
2184         (Printer._print_timing_statistics):
2185         (Printer._print_worker_statistics):
2186         (Printer._print_aggregate_test_statistics):
2187         (Printer._print_individual_test_times):
2188         (Printer._print_directory_timings):
2189         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
2190
2191 2012-12-13  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2192
2193         [EFL] The WebKit2 bots are building and executing WebKit1 tests
2194         https://bugs.webkit.org/show_bug.cgi?id=104690
2195
2196         Reviewed by Csaba Osztrogonác.
2197
2198         Pass --no-webkit1 option to the EFL WK2 bots when building. This
2199         will disable WK1 API and thus, the WK1 API tests.
2200
2201         * BuildSlaveSupport/build.webkit.org-config/config.json:
2202         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2203         (CompileWebKit2Only):
2204         (BuildAndTestFactory):
2205         (BuildAndTestFactory.__init__):
2206         (BuildAndTestWebKit2OnlyFactory):
2207         Now bots using BuildAndTestWebKit2OnlyFactory (for instance only EFL) will
2208         get --no-webkit1 as a parameter on the build-webkit script.
2209         * Scripts/build-webkit:
2210         Added support for -no-webkit1 for the Qt and EFL port.
2211
2212 2012-12-13  Eric Seidel  <eric@webkit.org>
2213
2214         build-webkit --chromium-android forces Make even when update-webkit used --ninja
2215         https://bugs.webkit.org/show_bug.cgi?id=104839
2216
2217         Reviewed by Tony Chang.
2218
2219         This was recommended by Peter as part of:
2220         https://bugs.webkit.org/show_bug.cgi?id=104434#c4
2221         Fixing this makes build-webkit --chromium-android work for the common (ninja) case.
2222
2223         * Scripts/webkitdirs.pm:
2224         (buildChromium):
2225
2226 2012-12-13  Julie Parent  <jparent@chromium.org>
2227
2228         Loader_unittests should reset state between runs.
2229         https://bugs.webkit.org/show_bug.cgi?id=104862
2230
2231         Reviewed by Dirk Pranke.
2232
2233         The loader_unittests were not resetting state between runs, leaving
2234         a lot of data behind in global objects that the dashboard and tests
2235         rely on.  A handful were resetting just a few objects.  Use the common
2236         resetGlobals() method instead.
2237
2238         * TestResultServer/static-dashboards/loader_unittests.js:
2239
2240 2012-12-13  Andras Becsi  <andras.becsi@digia.com>
2241
2242         [Qt][WK2] Fix painting on Mac with retina display
2243         https://bugs.webkit.org/show_bug.cgi?id=104574
2244
2245         Reviewed by Kenneth Rohde Christiansen.
2246
2247         Remove setting the devicePixelRatio experimental property
2248         since the value is now automatically picked up from Qt.
2249
2250         * MiniBrowser/qt/qml/BrowserWindow.qml:
2251
2252 2012-12-13  Jussi Kukkonen  <jussi.kukkonen@intel.com>
2253
2254         [EFL][GTK] Don't call deprecated g_type_init when glib => 2.35
2255         https://bugs.webkit.org/show_bug.cgi?id=103209
2256
2257         Reviewed by Kenneth Rohde Christiansen.
2258
2259         g_type_init() is deprecated from 2.35.0 onwards. Don't call it
2260         in that case to avoid warnings and/or build failure.
2261
2262         * DumpRenderTree/gtk/ImageDiff.cpp:
2263         (main):
2264
2265 2012-12-13  Claudio Saavedra  <csaavedra@igalia.com>
2266
2267         [Tools] Add exception for gtk_style_context_get()
2268         https://bugs.webkit.org/show_bug.cgi?id=104896
2269
2270         Reviewed by Martin Robinson.
2271
2272         Add new exception for gtk_style_context_get().
2273
2274         * Scripts/webkitpy/style/checkers/cpp.py:
2275         (check_for_null): Add exception for gtk_style_context_get().
2276         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
2277         (WebKitStyleTest.test_null_false_zero): Add a unit
2278         test for the new exception.
2279
2280 2012-12-13  Joaquim Rocha  <jrocha@igalia.com>
2281
2282         REGRESSION (r137432): The /webkit2/WebKitCookieManager/accept-policy unit test is failing
2283         https://bugs.webkit.org/show_bug.cgi?id=104790
2284
2285         Reviewed by Carlos Garcia Campos.
2286
2287         Unskip the GTK accept policy test.
2288
2289         * Scripts/run-gtk-tests:
2290         (TestRunner):
2291
2292 2012-12-13  Xingnan Wang  <xingnan.wang@intel.com>
2293
2294         Unreviewed. Adding myself to committers list.
2295
2296         * Scripts/webkitpy/common/config/committers.py:
2297
2298 2012-12-12  Roger Fong  <roger_fong@apple.com>
2299
2300         Unreviewed. Fix typo in NRWT.
2301
2302         * Scripts/webkitpy/layout_tests/port/base.py:
2303         (Port.is_reference_html_file):
2304
2305 2012-12-12  Roger Fong  <roger_fong@apple.com>
2306
2307         Fix detection of ref test files in old and new run webkit tests.
2308         https://bugs.webkit.org/show_bug.cgi?id=104850.
2309
2310         Reviewed by Timothy Horton.
2311
2312         * Scripts/old-run-webkit-tests:
2313         (isUsedInReftest): 
2314         Add notref and ref prefix and suffix cases to isUsedinReftest.
2315         * Scripts/webkitpy/layout_tests/port/base.py:
2316         (Port):
2317         (Port.is_reference_html_file): 
2318         Typo in NRWT.
2319
2320 2012-12-12  Dirk Pranke  <dpranke@chromium.org>
2321
2322         rework ChromiumAndroidDriver.start() et. al to avoid funky layering
2323         https://bugs.webkit.org/show_bug.cgi?id=104856
2324
2325         Reviewed by Eric Seidel.
2326
2327         r137523 / bug 10357 introduced some back-and-forth hopping during
2328         driver.start() in the chromium android port. This patch cleans it
2329         up a bit.
2330
2331         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2332         (ChromiumAndroidDriver.__init__):
2333         (ChromiumAndroidDriver._start_once):
2334         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
2335         * Scripts/webkitpy/layout_tests/port/driver.py:
2336         (Driver.start):
2337         (Driver._start):
2338
2339 2012-12-12  Eric Seidel  <eric@webkit.org>
2340
2341         Add support for run-perf-tests --chromium-android --profile
2342         https://bugs.webkit.org/show_bug.cgi?id=103587
2343
2344         Reviewed by Dirk Pranke.
2345
2346         This makes it possible to profile PerformanceTests on an Android device
2347         using run-perf-tests --chromium-android --profile.
2348         CPU profiling on Android is sadly non-trivial, but this goes a long
2349         ways towards making it more accessible to non-Android developers.
2350
2351         Although this script requires your device to be flashed with a
2352         userdebug build (an annoyingly non-trivial endeavor for external contributors)
2353         it endeavors to detect and explain the common errors encountered
2354         while trying to setup a device for perf-based profiling.
2355
2356         * Scripts/webkitpy/common/system/profiler.py:
2357         (GooglePProf.profile_after_exit): Discovered that pprof will just hang if passed a missing file.
2358         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2359         (AndroidPerf):
2360         (AndroidPerf.__init__):
2361         (AndroidPerf.check_configuration):
2362         (AndroidPerf.print_setup_instructions):
2363         (_file_exists_on_device):
2364         (_run_adb_command):
2365         (attach_to_pid):
2366         (_perf_version_string):
2367         (_find_perfhost_binary):
2368         (_perfhost_path):
2369         (_first_ten_lines_of_profile):
2370         (profile_after_exit):
2371         (ChromiumAndroidDriver.__init__):
2372         (ChromiumAndroidDriver.__init__.methods):
2373         (ChromiumAndroidDriver._find_or_create_symfs):
2374         (ChromiumAndroidDriver._start_once):
2375         (ChromiumAndroidDriver):
2376         (ChromiumAndroidDriver._run_post_start_tasks):
2377         (ChromiumAndroidDriver._pid_from_android_ps_output):
2378         (ChromiumAndroidDriver._pid_on_target):
2379         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
2380         (ChromiumAndroidDriverTest.test_pid_from_android_ps_output):
2381         (AndroidPerfTest):
2382         (AndroidPerfTest.test_perf_output_regexp):
2383         * Scripts/webkitpy/layout_tests/port/driver.py:
2384         (Driver._setup_environ_for_driver):
2385         (Driver):
2386         (Driver._start):
2387         (Driver._run_post_start_tasks):
2388         (Driver._pid_on_target):
2389
2390 2012-12-12  Dirk Pranke  <dpranke@chromium.org>
2391
2392         nrwt: remove deprecated ORWT compatibility command line flags
2393         https://bugs.webkit.org/show_bug.cgi?id=104301
2394
2395         Reviewed by Eric Seidel.
2396
2397         The command line flags for --noshow-results, --[no-]launch-safari,
2398         and --use-remote-links-to-tests were for compatibility with
2399         old-run-webkit-tests, but the compatibility is not needed any more
2400         since NRWT is basically default and the scripts that passed the
2401         old flags have been updated.
2402
2403         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2404         (_set_up_derived_options):
2405         (parse_args):
2406
2407 2012-12-12  John Griggs  <jgriggs@rim.com>
2408
2409         2012-12-12  John Griggs  <jgriggs@rim.com>
2410
2411         [BlackBerry] Update Media Controls
2412         https://bugs.webkit.org/show_bug.cgi?id=104443
2413         https://przilla.ott.qnx.com/bugzilla/show_bug.cgi?id=204748
2414
2415         Update Media Controls for BlackBerry platform to allow audio, embedded video and fullscreen video controls to be positioned and styled differently.
2416
2417         Reviewed by Rob Buis.
2418
2419         BlackBerry-specific update to Media Controls - No new tests required.
2420
2421         * Scripts/webkitperl/FeatureList.pm:
2422
2423 2012-12-12  Justin Novosad  <junov@google.com>
2424
2425         Adding myself to SkiaGraphics watchlist
2426
2427         Unreviewed
2428
2429         * Scripts/webkitpy/common/config/watchlist:
2430
2431 2012-12-12  Martin Robinson  <mrobinson@igalia.com>
2432
2433         [GTK] Add authentication support to DRT and fix exposed issues in the libsoup backend
2434         https://bugs.webkit.org/show_bug.cgi?id=104285
2435
2436         Reviewed by Gustavo Noronha Silva.
2437
2438         Add support to DumpRenderTree for running authentication tests. Since the DRT
2439         expects an authentication callback, we add one to DRTSupport to avoid #ifdefs
2440         in platform-independent code for GTK+.
2441
2442         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2443         (resetDefaultsToConsistentValues): Reset the authentication password and username.
2444         (authenticationCallback): Added.
2445         (createWebView): Attach the authentiation callback during startup.
2446         * DumpRenderTree/gtk/TestRunnerGtk.cpp:
2447         (soupURIToKURL): soup_uri_to_string does not preserve passwords embedded
2448         in the URL so we add a somewhat messy method of re-adding them when they exist.
2449         It would be nice to use soupURIToKURL here, but it seems we cannot use KURL without
2450         pulling in lots of WebCore code so we use string search and replace.
2451         (TestRunner::queueLoad): Use the new helper.
2452
2453 2012-12-12  Dirk Pranke  <dpranke@chromium.org>
2454
2455         garden-o-matic should prefer efl/ over efl-wk1/ and efl-wk2/ when rebaselining
2456         https://bugs.webkit.org/show_bug.cgi?id=102389
2457
2458         Reviewed by Tony Chang.
2459
2460         The code wasn't aware of both the efl-wk1 and efl-wk2 variants,
2461         because all_port_names() by itself doesn't tell you about the
2462         variants. Work around this for now, and filed bug 104761 to
2463         address the underlying problem.
2464
2465         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
2466         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
2467         (BaselineOptimizerTest.test_efl):
2468
2469 2012-12-12  Krzysztof Czech  <k.czech@samsung.com>
2470
2471         [EFL] Possibility to turn off accessibility feature for WebKit-EFL.
2472         https://bugs.webkit.org/show_bug.cgi?id=103036
2473
2474         Reviewed by Laszlo Gombos.
2475
2476         Guard dependencies for accessibility (ATK), so that they can be turned off.
2477
2478         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.h:
2479         (AccessibilityUIElement):
2480         * WebKitTestRunner/PlatformEfl.cmake:
2481
2482 2012-12-12  Simon Fraser  <simon.fraser@apple.com>
2483
2484         Add some entries in the watchlist, and register me for them.
2485
2486         * Scripts/webkitpy/common/config/watchlist:
2487
2488 2012-12-12  Ilya Tikhonovsky  <loislo@chromium.org>
2489
2490         Web Inspector: Native Memory Instrumentation: remove fake root MemoryObjectInfo.
2491         https://bugs.webkit.org/show_bug.cgi?id=104796
2492
2493         Reviewed by Yury Semikhatsky.
2494
2495         It was not a good idea to introduce a fake root MemoryObjectInfo.
2496         It makes a problem when we visit an object without its own MemoryObjectType.
2497
2498         Example: RenderBox has a global pointer to a hash map.
2499         HashMap doesn't have its own object type because it is a generic container.
2500         It will inherit object type from the fake root memory object info.
2501         The same could happen for another container in another class with other MemoryObjectType.
2502
2503         This fact forces me to create custom process method for root objects
2504         because they need to have their own MemoryObjectInfo with customisable memory object type.
2505
2506         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
2507
2508         Drive by fix: InstrumentedPointer* was replaced with Wrapper* because actually it is using
2509         for instrumented and not instrumented object classes.
2510
2511 2012-12-12  Zan Dobersek  <zandobersek@gmail.com>
2512
2513         [GTK] Remove the last of unnecessary configuration options in configure.ac
2514         https://bugs.webkit.org/show_bug.cgi?id=104793
2515
2516         Reviewed by Martin Robinson.
2517
2518         Removing the filters and indexed database features from the list of features
2519         being configurable through the configure script. This is not the case anymore
2520         as the corresponding configuration options are being removed.
2521
2522         * Scripts/webkitdirs.pm:
2523         (buildAutotoolsProject):
2524
2525 2012-12-12  Mario Sanchez Prada  <mario@webkit.org>
2526
2527         [GTK][WK2] New API to detect display/execution of insecure content
2528         https://bugs.webkit.org/show_bug.cgi?id=104578
2529
2530         Reviewed by Martin Robinson.
2531
2532         Add coding style exception for 'readability/enum_casing' in GTK.
2533
2534         The GTK+ API use upper case, underscore separated, words in
2535         certain types of enums (e.g. signals, properties).
2536
2537         * Scripts/webkitpy/style/checker.py: Added exception, both for
2538         WebKitGTK+ and WebKit2GTK+.
2539
2540 2012-12-12  Jochen Eisinger  <jochen@chromium.org>
2541
2542         [chromium] Fix build on chromium-win
2543
2544         Unreviewed build-fix.
2545
2546         MSVS indeed doesn't have snprintf, adding wtf/StringExtras.h include
2547         to work around that.
2548
2549         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp:
2550
2551 2012-12-12  Simon Hausmann  <simon.hausmann@digia.com>
2552
2553         [Qt] Fix build on Mac
2554
2555         Reviewed by Csaba Osztrogonác.
2556
2557         On Mac OS X we have sqlite as a system library available.
2558
2559         * qmake/mkspecs/features/features.prf:
2560
2561 2012-12-12  Zan Dobersek  <zandobersek@gmail.com>
2562
2563         Unreviewed GTK gardening.
2564
2565         Skipping another unit test that regressed in r137432.
2566
2567         * Scripts/run-gtk-tests:
2568         (TestRunner):
2569
2570 2012-12-12  Simon Hausmann  <simon.hausmann@digia.com>
2571
2572         [Qt] Fix linkage against SQLite
2573         https://bugs.webkit.org/show_bug.cgi?id=104781
2574
2575         Reviewed by Csaba Osztrogonác.
2576
2577         Detect SQLite either through pkg-config, from the SQLITE3SRCDIR
2578         environment variable or from qtbase via a qt5.git build. Bail out early
2579         with an error message if neither option applies.
2580
2581         This removes the usage of $$QT.core.sources and the dependency on Qt's
2582         system-sqlite configure setting. We should always favour a system
2583         library over a copy in Qt.
2584
2585         * qmake/mkspecs/features/features.prf:
2586
2587 2012-12-12  Tommy Widenflycht  <tommyw@google.com>
2588
2589         MediaStream API: Change the data channel descriptor pattern to a handler pattern
2590         https://bugs.webkit.org/show_bug.cgi?id=104543
2591
2592         Reviewed by Kent Tamura.
2593
2594         Adding mocks to be able to test this refactoring.
2595
2596         * DumpRenderTree/DumpRenderTree.gypi:
2597         * DumpRenderTree/chromium/MockWebRTCDataChannelHandler.cpp: Added.
2598         (DataChannelReadyStateTask):
2599         (DataChannelReadyStateTask::DataChannelReadyStateTask):
2600         (MockWebRTCDataChannelHandler::MockWebRTCDataChannelHandler):
2601         (MockWebRTCDataChannelHandler::setClient):
2602         (MockWebRTCDataChannelHandler::bufferedAmount):
2603         (MockWebRTCDataChannelHandler::sendStringData):
2604         (MockWebRTCDataChannelHandler::sendRawData):
2605         (MockWebRTCDataChannelHandler::close):
2606         * DumpRenderTree/chromium/MockWebRTCDataChannelHandler.h: Added.
2607         (MockWebRTCDataChannelHandler):
2608         (MockWebRTCDataChannelHandler::taskList):
2609         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
2610         (RTCPeerConnectionReadyStateTask::RTCPeerConnectionReadyStateTask):
2611         (RemoteDataChannelTask::RemoteDataChannelTask):
2612         (MockWebRTCPeerConnectionHandler::createDataChannel):
2613         (MockWebRTCPeerConnectionHandler::stop):
2614         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h:
2615         (MockWebRTCPeerConnectionHandler):
2616
2617 2012-12-12  Jochen Eisinger  <jochen@chromium.org>
2618
2619         [chromium] WebTestPlugin must not depend on WTF::String
2620         https://bugs.webkit.org/show_bug.cgi?id=104783
2621
2622         Reviewed by Kent Tamura.
2623
2624         In a component build, we can't access the WTF that is part of the
2625         WebKit component. Accessing WTF instead will pull in a separate copy
2626         into the TestRunner library. Instead of WTF::String::number use
2627         snprintf() to print numbers.
2628
2629         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp:
2630
2631 2012-12-12  Jocelyn Turcotte  <jocelyn.turcotte@digia.com>
2632
2633         [Qt] Do not automatically enable force_static_libs_as_shared when using debug_and_release
2634         https://bugs.webkit.org/show_bug.cgi?id=104562
2635
2636         Reviewed by Tor Arne Vestbø.
2637
2638         force_static_libs_as_shared creates issues on Mac if QtWebKit is built as a framework.
2639         QtWebProcess, for example, will link to the framework's release library and will rely
2640         on DYLD_IMAGE_SUFFIX to switch to the debug build. Since QtWebProcess will also links
2641         directly to the debug build of WebCore, and that QtWebKit release will link in the
2642         release build of WebCore, both the debug and release version will be loaded.
2643
2644         To fix this properly we should make sure that applications always link to the release
2645         build of dependent libraries by bypassing resolveFinalLibraryName.
2646         Disable it for now as force_static_libs_as_shared is mainly for development while
2647         debug_and_release is mainly for production builds.
2648
2649         * qmake/mkspecs/features/default_post.prf:
2650
2651 2012-12-12  Zan Dobersek  <zandobersek@gmail.com>
2652
2653         Unreviewed GTK gardening of the unit tests.
2654
2655         Skip two unit tests that regressed during the hackfest.
2656
2657         * Scripts/run-gtk-tests:
2658         (TestRunner):
2659
2660 2012-12-11  Ryosuke Niwa  <rniwa@webkit.org>
2661
2662         Add Koji Ishii to the list of contributors.
2663
2664         * Scripts/webkitpy/common/config/committers.py:
2665
2666 2012-12-11  Dominic Mazzoni  <dmazzoni@google.com>
2667
2668         AX: Make isActionSupported cross-platform.
2669         https://bugs.webkit.org/show_bug.cgi?id=104681
2670
2671         Reviewed by Chris Fleizach.
2672
2673         Replace isActionSupported, which takes a platform-specific
2674         argument, with three cross-platform methods.
2675
2676         * DumpRenderTree/AccessibilityUIElement.cpp:
2677         (isPressActionSupportedCallback):
2678         (isIncrementActionSupportedCallback):
2679         (isDecrementActionSupportedCallback):
2680         (AccessibilityUIElement::getJSClass):
2681         * DumpRenderTree/AccessibilityUIElement.h:
2682         (AccessibilityUIElement):
2683         * DumpRenderTree/blackberry/AccessibilityUIElementBlackBerry.cpp:
2684         (AccessibilityUIElement::isPressActionSupported):
2685         (AccessibilityUIElement::isIncrementActionSupported):
2686         (AccessibilityUIElement::isDecrementActionSupported):
2687         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
2688         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
2689         (WebTestRunner::AccessibilityUIElement::isPressActionSupportedCallback):
2690         (WebTestRunner):
2691         (WebTestRunner::AccessibilityUIElement::isIncrementActionSupportedCallback):
2692         (WebTestRunner::AccessibilityUIElement::isDecrementActionSupportedCallback):
2693         (WebTestRunner::AccessibilityUIElement::incrementCallback):
2694         (WebTestRunner::AccessibilityUIElement::decrementCallback):
2695         (WebTestRunner::AccessibilityUIElement::pressCallback):
2696         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
2697         (AccessibilityUIElement):
2698         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
2699         (AccessibilityUIElement::isPressActionSupported):
2700         (AccessibilityUIElement::isIncrementActionSupported):
2701         (AccessibilityUIElement::isDecrementActionSupported):
2702         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
2703         (AccessibilityUIElement::isPressActionSupported):
2704         (AccessibilityUIElement::isIncrementActionSupported):
2705         (AccessibilityUIElement::isDecrementActionSupported):
2706         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
2707         (AccessibilityUIElement::isPressActionSupported):
2708         (AccessibilityUIElement::isIncrementActionSupported):
2709         (AccessibilityUIElement::isDecrementActionSupported):
2710         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.cpp:
2711         (WTR::AccessibilityUIElement::isPressActionSupported):
2712         (WTR::AccessibilityUIElement::isIncrementActionSupported):
2713         (WTR::AccessibilityUIElement::isDecrementActionSupported):
2714         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.h:
2715         (AccessibilityUIElement):
2716         * WebKitTestRunner/InjectedBundle/Bindings/AccessibilityUIElement.idl:
2717         * WebKitTestRunner/InjectedBundle/mac/AccessibilityUIElementMac.mm:
2718         (WTR::AccessibilityUIElement::isPressActionSupported):
2719         (WTR):
2720         (WTR::AccessibilityUIElement::isIncrementActionSupported):
2721         (WTR::AccessibilityUIElement::isDecrementActionSupported):
2722
2723 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2724
2725         nrwt: remove result_summary.total_tests_by_expectation
2726         https://bugs.webkit.org/show_bug.cgi?id=104756
2727
2728         Reviewed by Eric Seidel.
2729
2730         This field wasn't being used by anything.
2731
2732         * Scripts/webkitpy/layout_tests/models/result_summary.py:
2733         (ResultSummary.__init__):
2734
2735 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2736
2737         garden-o-matic doesn't know about reftests
2738         https://bugs.webkit.org/show_bug.cgi?id=101976
2739
2740         Reviewed by Ojan Vafai.
2741
2742         This is a first attempt at fixing this, sort of a "stop-the-bleeding"
2743         fix. We will now look at the list of tests we're asked to rebaseline,
2744         and log which ones are reftests into the status area, and rebaseline
2745         the rest.
2746
2747         Longer-term, we should display feedback in the results area based on
2748         which tests we're looking at. If the test list contains no reftests,
2749         we should behave as today; if the list is all reftest, we should
2750         probably disable the button, and if the list is a mix, we should
2751         probably display an alert next to the button or something. It would
2752         also be nice to annotate which tests were reftests, maybe with a
2753         "(reftest)" next to the test name or something.
2754
2755         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
2756         (.):
2757         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
2758         (.):
2759         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
2760
2761 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2762
2763         REGRESSION(r133380): new tests without expected file reported as failing tests on the bots
2764         https://bugs.webkit.org/show_bug.cgi?id=101469
2765
2766         Reviewed by Ryosuke Niwa.
2767
2768         Fix parsing of NRWT's output to identify missing results separately
2769         from other regressions.
2770
2771         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2772         (RunWebKitTests._parseNewRunWebKitTestsOutput):
2773         * BuildSlaveSupport/build.webkit.org-config/mastercfg_unittest.py:
2774         (MasterCfgTest.test_nrwt_leaks_parsing):
2775         (test_nrwt_missing_results):
2776
2777 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2778
2779         buildbot: change run-webkit-tests --no-launch-safari to --no-show-results
2780         https://bugs.webkit.org/show_bug.cgi?id=104299
2781
2782         Reviewed by Ryosuke Niwa.
2783
2784         Since --no-show-results is the more generically-named flag that NRWT uses.
2785
2786         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2787         (RunWebKitTests):
2788
2789 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2790
2791         remove --use-remote-links-to-tests from buildbot configs (except apple win)
2792         https://bugs.webkit.org/show_bug.cgi?id=104291
2793
2794         Reviewed by Csaba Osztrogonác.
2795
2796         This flag is ignored by NRWT and was just implemented for compatibility.
2797         ORWT still uses it, though.
2798
2799         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2800         (RunWebKitTests):
2801         (RunWebKitTests.start):
2802
2803 2012-12-11  Anders Carlsson  <andersca@apple.com>
2804
2805         ASSERT running platform/mac/plugins/convert-point.html on the WK2 bots
2806         https://bugs.webkit.org/show_bug.cgi?id=104735
2807
2808         Reviewed by Jon Honeycutt.
2809
2810         Do the testing from within NPP_New instead of NPN_SetWindow. Plug-ins aren't supposed to do any real work inside NPP_SetWindow,
2811         and doing so will more easily trigger the assertion failure. Ideally we should get rid of the initial sync painting pass, but this
2812         gets the test passing again!
2813
2814         * DumpRenderTree/TestNetscapePlugIn/Tests/mac/ConvertPoint.cpp:
2815         (ConvertPoint::NPP_New):
2816
2817 2012-12-11  Julie Parent  <jparent@chromium.org>
2818
2819         Dashboard unit tests should not make xhrs.
2820         https://bugs.webkit.org/show_bug.cgi?id=104598
2821
2822         Reviewed by Dirk Pranke.
2823
2824         The unit test were requesting 25 files that they did not use.
2825         Stub out loader.request to avoid making the xhrs.
2826         Result: order of magnitude speed improvement in test run time (from ~1600ms to 160ms on my machine).
2827         
2828         * TestResultServer/static-dashboards/run-embedded-unittests.html:
2829         * TestResultServer/static-dashboards/run-unittests.html:
2830
2831 2012-12-11  Filip Pizlo  <fpizlo@apple.com>
2832
2833         Profiler should show bytecode dumps as they would have been visible to the JITs, including the profiling data that the JITs would see
2834         https://bugs.webkit.org/show_bug.cgi?id=104647
2835
2836         Reviewed by Oliver Hunt.
2837
2838         Added a "profiling" (or "p") command to show the profiling data that the JITs saw
2839         for each JIT compilation of a code block.
2840         
2841         Also added instruction counts in the "full" display and made the "full" display the
2842         default thing you see.
2843
2844         * Scripts/display-profiler-output:
2845
2846 2012-12-11  Eric Seidel  <eric@webkit.org>
2847
2848         Unreviewed, rolling out r137371.
2849         http://trac.webkit.org/changeset/137371
2850         https://bugs.webkit.org/show_bug.cgi?id=104434
2851
2852         Various scripts are not ready for out/ as the build directory,
2853         this can't work as written.
2854
2855         * Scripts/update-webkit:
2856         * Scripts/webkitdirs.pm:
2857         (buildChromium):
2858
2859 2012-12-11  Zan Dobersek  <zandobersek@gmail.com>
2860
2861         Unreviewed GTK build functionality fix.
2862
2863         Turns out the CSP feature was not enabled, that's why all the tests were failing.
2864
2865         * Scripts/webkitperl/FeatureList.pm:
2866
2867 2012-12-11  Eric Seidel  <eric@webkit.org>
2868
2869         Make Ninja the default build system for build-webkit --chromium
2870         https://bugs.webkit.org/show_bug.cgi?id=104434
2871
2872         Reviewed by Daniel Bates.
2873
2874         As far as I can tell, most Chromium developers are now using Ninja
2875         as their default build system instead of the native Xcode/Visual Studio/Make.
2876         This change makes Ninja the default for developers as well as all Chromium webkit bots.
2877
2878         Developers can pass --no-ninja to get the platform-native buildsystem instead of Ninja.
2879         It's possible we may wish to make this a per-platform default, and have
2880         Ninja only be the default for Mac/Linux.
2881
2882         There was a thread on webkit-dev discussing this change:
2883         http://lists.webkit.org/pipermail/webkit-dev/2012-December/023113.html
2884
2885         * Scripts/update-webkit:
2886
2887 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2888
2889         nrwt: record the order and run times of all of the tests in a single file
2890         https://bugs.webkit.org/show_bug.cgi?id=104184
2891
2892         Reviewed by Ryosuke Niwa.
2893
2894         Replace the N tests_run*.txt files that we output into
2895         layout-test-results with a single packed JSON trie that contains
2896         the test ordering and run statistics. This can eventually be uploaded
2897         to the TestResultsServer to replace the times.json file we currently
2898         upload as well, and will let us replay the results of a previous
2899         run fairly deterministically.
2900
2901         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
2902         (Worker.__init__):
2903         (Worker.start):
2904         (Worker._run_test):
2905         (Worker.stop):
2906         (Worker._clean_up_after_test):
2907         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2908         (Manager._upload_json_files):
2909         (Manager._copy_results_html_file):
2910         (Manager):
2911         (Manager._stats_trie):
2912         (Manager._stats_trie._worker_number):
2913         * Scripts/webkitpy/layout_tests/models/test_results.py:
2914         (TestResult.__init__):
2915         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2916         (MainTest.test_missing_and_unexpected_results):
2917         (MainTest.test_retrying_and_flaky_tests):
2918         (EndToEndTest.test_end_to_end):
2919         (EndToEndTest.test_reftest_with_two_notrefs):
2920         (RebaselineTest.test_reset_results):
2921         (RebaselineTest.test_missing_results):
2922         (RebaselineTest.test_new_baseline):
2923
2924 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2925
2926         check-webkit-style complaining about TestExpectations lines a patch didn't modify
2927         https://bugs.webkit.org/show_bug.cgi?id=104687
2928
2929         Reviewed by Ojan Vafai.
2930
2931         Way back in bug 60466 / r86165, we changed the style checker
2932         for TestExpectations to ensure that the entire file linted, not
2933         just the lines modified by a patch, in order to ensure that a
2934         new line didn't cause errors elsewhere in the file (e.g., by
2935         introducing duplicate lines). At the time, errors in the
2936         expectations files were fatal and we didn't have a separate step
2937         on the bots that would lint the files to ensure things were still
2938         okay, so this was probably the right thing to do, but it could
2939         cause some patches to fail the style check even though the patches
2940         themselves were fine.
2941
2942         However, now errors aren't always fatal, and we have a separate build
2943         step to lint the files. So, we should probably not punish the
2944         patches and the style checker should filter out "unrelated" errors.
2945
2946         In theory we should be comparing the output pre- and post-patch
2947         to filter out pre-existing errors, but I don't think the style
2948         checker has that ability.
2949
2950         * Scripts/webkitpy/style/checkers/test_expectations.py:
2951         (TestExpectationsChecker.__init__):
2952         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
2953         (TestExpectationsTestCase.assert_lines_lint):
2954
2955 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2956
2957         nrwt isn't rejecting unrecognized expectations
2958         https://bugs.webkit.org/show_bug.cgi?id=104701
2959
2960         Reviewed by Ojan Vafai.
2961
2962         NRWT is silently ignoring unrecognized expectations like "Text"
2963         and "Audio" (largely because I haven't fully removed support for
2964         the old syntax and keywords and restructured the code to do the
2965         proper checking). Temporarily change the tokenizer to reject
2966         unrecognized expectations until I can make those changes.
2967
2968         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
2969         (TestExpectationParser._tokenize_line):
2970         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
2971         (ExpectationSyntaxTests.test_unknown_expectation):
2972         (TestExpectationSerializationTests.disabled_test_string_whitespace_stripping):
2973
2974 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
2975
2976         Fix a typo in my previous commit that is causing test failures.
2977
2978         Reviewed by Gustavo Noronha Silva.
2979
2980         There was an extra bit of statement appended after the end of a conditional. Somehow
2981         this compiled and caused test failures.
2982
2983         * DumpRenderTree/gtk/EventSender.cpp:
2984         (continuousMouseScrollByCallback):
2985
2986 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
2987
2988         [GTK] Spurious stderr output from the event sender
2989         https://bugs.webkit.org/show_bug.cgi?id=104678
2990
2991         Reviewed by Gustavo Noronha Silva.
2992
2993         Eliminate some spurious stderr output from the EventSender by using an early return
2994         instead of an assertion that prints a warning to the console.
2995
2996         * DumpRenderTree/gtk/EventSender.cpp:
2997         (continuousMouseScrollByCallback): We don't support continuous scroll events that
2998         move by page, but we should just fail silently instead of printing an error.
2999
3000 2012-12-11  Joone Hur  <joone.hur@intel.com>
3001
3002         [GTK] Add GraphicsLayerActor
3003         https://bugs.webkit.org/show_bug.cgi?id=91940
3004
3005         Reviewed by Gustavo Noronha Silva.
3006
3007         Add GraphicsLayerActor to the list of classes that contain GObjects
3008         in the style checker script so that it does not complain about GObject
3009         coding style.
3010
3011         * Scripts/webkitpy/style/checker.py:
3012
3013 2012-12-11  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
3014
3015         [Qt] remove hack which avoids unused .obj directories all over the place
3016
3017         Reviewed by Simon Hausmann.
3018
3019         The feature is subject to upstreaming in Qt itself in
3020         https://codereview.qt-project.org/#change,42011
3021
3022         * qmake/mkspecs/features/default_post.prf:
3023
3024 2012-12-11  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
3025
3026         [Qt] move SQLITE3SRCDIR assignment to only use site
3027
3028         Reviewed by Simon Hausmann.
3029
3030         Target.pri includes WebCore.pri the almost first thing, so that is
3031         covered as well.
3032
3033         * qmake/mkspecs/features/default_pre.prf:
3034
3035 2012-12-11  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
3036
3037         [Qt] delete dead include paths
3038
3039         Reviewed by Simon Hausmann.
3040
3041         followup to https://bugs.webkit.org/show_bug.cgi?id=93446
3042
3043         * DumpRenderTree/qt/DumpRenderTree.pro:
3044
3045 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
3046
3047         [GTK] Implement some missing FrameLoaderClient message in DRT
3048         https://bugs.webkit.org/show_bug.cgi?id=104444
3049
3050         Reviewed by Gustavo Noronha Silva.
3051
3052         Install a FrameLoadEvent callback that prints the necessary output for some
3053         missing FrameLoader events.
3054
3055         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3056         (frameLoadEventCallback): Added.
3057         (createWebView): Install the FrameLoader event callback.
3058
3059 2012-12-11  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
3060
3061         [Qt] change setup of QMAKE_MOC
3062
3063         Reviewed by Simon Hausmann.
3064
3065         don't load(moc) explicitly - turns out that this breaks
3066         debug_and_release builds (the feature does not like being loaded before
3067         default_post has been loaded).
3068         so instead just do a minimal call to find the moc executable.
3069
3070         as it's in fact only two files which need it, so de-centralize this call
3071         for the sake of saving some cpu cycles spent in qmake.
3072
3073         * WebKitTestRunner/InjectedBundle/DerivedSources.pri:
3074         * qmake/mkspecs/features/default_pre.prf:
3075
3076 2012-12-11  Jocelyn Turcotte  <jocelyn.turcotte@digia.com>
3077
3078         [Qt] Fix the linking of debug builds on Mac with force_static_libs_as_shared
3079         https://bugs.webkit.org/show_bug.cgi?id=104559
3080
3081         Reviewed by Tor Arne Vestbø.
3082
3083         "contains(libraries, $$TARGET): return(true)" fails for intermediate libraries since
3084         at this point TARGET may contain the "_debug" suffix.
3085
3086         Fix it by moving the TARGET adjustment later in the file.
3087
3088         * qmake/mkspecs/features/default_post.prf:
3089
3090 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
3091
3092         [Coverity] [GTK] Remove some redundant null checks in DumpRenderTree
3093         https://bugs.webkit.org/show_bug.cgi?id=104570
3094
3095         Reviewed by Daniel Bates.
3096
3097         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3098         (descriptionSuitableForTestResult): Remove a redundant null check.
3099
3100 2012-12-11  Alberto Garcia  <agarcia@igalia.com>
3101
3102         Remove unused GErrors
3103         https://bugs.webkit.org/show_bug.cgi?id=104657
3104
3105         Reviewed by Martin Robinson.
3106
3107         Replace unused GErrors with null pointers.
3108
3109         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3110         (initializeFonts):
3111         * DumpRenderTree/gtk/EditingCallbacks.cpp:
3112         (dumpRange):
3113         * WebKitTestRunner/InjectedBundle/gtk/ActivateFontsGtk.cpp:
3114         (WTR::inititializeFontConfigSetting):
3115
3116 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
3117
3118         [GTK] Bump the jhbuild libsoup version to fix an authentication issue
3119         https://bugs.webkit.org/show_bug.cgi?id=104448
3120
3121         Reviewed by Gustavo Noronha Silva.
3122
3123         Bump the libsoup version to one from the git repository. This requires updating
3124         GLib as well which in turn requires adjusting the CFLAGS we pass when compiling
3125         the scanobj step of the gtkdoc generation.
3126
3127         * gtk/generate-gtkdoc:
3128         * gtk/jhbuild.modules:
3129
3130 2012-12-11  Zan Dobersek  <zandobersek@gmail.com>
3131
3132         Unreviewed functionality fix for the GTK port.
3133
3134         Enable the MicroData feature support for the GTK port in the FeatureList module.
3135         This feature listing now again affects which features are enabled and which disabled
3136         on the development builds of the GTK port.
3137
3138         * Scripts/webkitperl/FeatureList.pm:
3139
3140 2012-12-11  Jinwoo Song  <jinwoo7.song@samsung.com>
3141
3142         [EFL][WK2] MiniBrowser is not built
3143         https://bugs.webkit.org/show_bug.cgi?id=104629
3144
3145         Reviewed by Gyuyoung Kim.
3146
3147         As we are finding an elementary package in Tools/MiniBrowser/efl/CMakeList.txt since r137203,
3148         we need to check if the package is found in the same place not in the Tools/CMakeList.txt.
3149
3150         * CMakeLists.txt:
3151         * MiniBrowser/efl/CMakeLists.txt:
3152
3153 2012-12-11  Xabier Rodriguez Calvar  <calvaris@igalia.com>
3154
3155         [GTK][jhbuild] Switch to GStreamer 1.0 build
3156         https://bugs.webkit.org/show_bug.cgi?id=91727
3157
3158         Reviewed by Philippe Normand.
3159
3160         Switch build-webkit --gtk to GStreamer 1.0 support and build the
3161         necessary GStreamer git modules from JHBuild.
3162
3163         * gtk/jhbuild.modules: Added GStreamer 1.0.3 build support.
3164
3165 2012-12-11  Zan Dobersek  <zandobersek@gmail.com>
3166
3167         [GTK] Feature enabling/disabling should be possible through build-webkit
3168         https://bugs.webkit.org/show_bug.cgi?id=99271
3169
3170         Reviewed by Gustavo Noronha Silva.
3171
3172         Refactor the code in webkitdirs.pm that builds an autotools project to generate
3173         autogen.sh arguments, make arguments and installation prefix from the passed-in
3174         parameters rather than generating all of that in build-webkit. The autogen.sh
3175         arguments now contain only enable/disable flags for options that are actually
3176         configurable in configure.ac, the flag value actually reflecting whether the feature
3177         is enabled or disabled in the feature list.
3178
3179         Other features are overridable through modifying the GNUmakefile.features.am file.
3180         All these features are now stored in the build directory and upon change trigger
3181         a rerun of the autogen.sh script, pretty much like the autogen.sh arguments do.
3182
3183         The override-feature-defines script is called by autogen.sh, before GNUmakefile.in is
3184         generated when calling autoreconf. Its task is to copy the GNUmakefile.features.am.in
3185         into GNUmakefile.features.am and modify it if there's a text file in the build directory
3186         that contains all the feature defines the build-webkit script has written. If the build
3187         is not done through build-webkit (and there's no file in build directory listing all the
3188         feature defines) the script does not advance further from copying.
3189
3190         This approach is taken to overcome the rigidness of the automake system as it's impossible
3191         to effectively generate and use GNUmakefile.features.am or even GNUmakefile.features file
3192         after the autoreconf command execution in autogen.sh.
3193
3194         * Scripts/build-jsc:
3195         (buildMyProject):
3196         * Scripts/build-webkit:
3197         * Scripts/webkitdirs.pm:
3198         (runAutogenForAutotoolsProjectIfNecessary):
3199         (mustReRunAutogen):
3200         (buildAutotoolsProject):
3201         (buildGtkProject):
3202         * Scripts/webkitperl/FeatureList.pm: Update the features that are currently enabled in
3203         Source/WebCore/GNUmakefile.features.am.in but aren't in the feature list.
3204         * Scripts/webkitpy/style/checker.py: Source/WebCore/GNUmakefile.features.am.in is recognized
3205         as a text file and tabulation errors are reported. Skip the file to suppress them.
3206         * gtk/override-feature-defines: Added.
3207         (copy_feature_defines_makefile):
3208         (adjust_feature_defines_makefile):
3209         (adjust_feature_defines_makefile.override_feature_define):
3210         (override_feature_defines):
3211
3212 2012-12-10  Yury Semikhatsky  <yurys@chromium.org>
3213
3214         Memory instrumentation: make sure each edge is reported only once
3215         https://bugs.webkit.org/show_bug.cgi?id=104630
3216
3217         Reviewed by Pavel Feldman.
3218
3219         Test that outgoing edges are reported only once if we come to the same
3220         object twice: first by an address of a base class and then by a real address
3221         of the object.
3222
3223         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
3224
3225 2012-12-10  Dana Jansens  <danakj@chromium.org>
3226
3227         [chromium] Remove WebCompositorOutputSurface implementations from DRT
3228         https://bugs.webkit.org/show_bug.cgi?id=104046
3229
3230         Reviewed by Darin Fisher.
3231
3232         The classes are now provided through WebCompositorSupport.
3233
3234         * DumpRenderTree/DumpRenderTree.gypi:
3235         * DumpRenderTree/chromium/WebViewHost.cpp:
3236         (WebViewHost::createOutputSurface):
3237         * DumpRenderTree/chromium/WebViewHostOutputSurface.cpp: Removed.
3238         * DumpRenderTree/chromium/WebViewHostOutputSurface.h: Removed.
3239         * DumpRenderTree/chromium/WebViewHostSoftwareOutputDevice.cpp: Removed.
3240         * DumpRenderTree/chromium/WebViewHostSoftwareOutputDevice.h: Removed.
3241
3242 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3243
3244         change --no-launch-safari to --no-show-results in scripts
3245         https://bugs.webkit.org/show_bug.cgi?id=104298
3246
3247         Reviewed by Eric Seidel.
3248
3249         Since Safari only exists on the Apple ports, NRWT calls the
3250         command line flag --[no-]show-results instead. This updates
3251         the scripts to use the new flag (but not the buildbot config yet).
3252
3253         * Scripts/generate-coverage-data:
3254         * Scripts/webkit-tools-completion.sh:
3255         * Scripts/webkitpy/tool/steps/runtests.py:
3256         (RunTests.run):
3257         * Scripts/webkitpy/tool/steps/runtests_unittest.py:
3258
3259 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3260
3261         nrwt: fix mapping of AUDIO onto FAIL when printing unexpected results
3262         https://bugs.webkit.org/show_bug.cgi?id=104597
3263
3264         Reviewed by Ojan Vafai.
3265
3266         This was broken in the recent printer refactoring.
3267
3268         * Scripts/webkitpy/layout_tests/views/buildbot_results.py:
3269         (BuildBotPrinter.print_unexpected_results.add_result.is_expected):
3270         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
3271         (BuildBotPrinterTests.get_result):
3272         (BuildBotPrinterTests.test_print_unexpected_results):
3273         (BuildBotPrinterTests.test_print_unexpected_results.get_unexpected_results):
3274         (BuildBotPrinterTests):
3275
3276 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3277
3278         old-run-webkit-tests: add --show-results properly
3279         https://bugs.webkit.org/show_bug.cgi?id=104608
3280
3281         Reviewed by Eric Seidel.
3282
3283         Rework r137219 / bug 104601 to use the correct perl idiom.
3284
3285         * Scripts/old-run-webkit-tests:
3286
3287 2012-12-10  Aaron Colwell  <acolwell@chromium.org>
3288
3289         Unreviewed. Adding Aaron Colwell to the contributor list.
3290
3291         * Scripts/webkitpy/common/config/committers.py:
3292
3293 2012-12-10  Anders Carlsson  <andersca@apple.com>
3294
3295         Remove WebKit2 tests on Windows
3296         https://bugs.webkit.org/show_bug.cgi?id=104607
3297
3298         Reviewed by Tim Horton.
3299
3300         Stop building WebKit2 tests and remove associated files.
3301
3302         * TestWebKitAPI/win/InjectedBundleControllerWin.cpp: Removed.
3303         * TestWebKitAPI/win/PlatformUtilitiesWin.cpp: Removed.
3304         * TestWebKitAPI/win/PlatformWebViewWin.cpp: Removed.
3305         * TestWebKitAPI/win/TestWebKitAPI.sln: Removed.
3306         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
3307         * TestWebKitAPI/win/TestWebKitAPIGenerated.vcproj: Removed.
3308         * TestWebKitAPI/win/TestWebKitAPIGeneratedCommon.vsprops: Removed.
3309         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj: Removed.
3310         * TestWebKitAPI/win/WindowMessageObserver.h: Removed.
3311         * TestWebKitAPI/win/copy-resources.cmd: Removed.
3312
3313 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3314
3315         old-run-webkit-tests: support --no-show-results for compatibility w/ NRWT
3316         https://bugs.webkit.org/show_bug.cgi?id=104601
3317
3318         Reviewed by Eric Siedel.
3319
3320         This flag is being added so we can move all of the calling scripts
3321         over to using --no-show-results instead of --no-launch-safari.
3322
3323         * Scripts/old-run-webkit-tests:
3324
3325 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3326
3327         nrwt: use the default logic for skipping platform-specific tests for chromium
3328         https://bugs.webkit.org/show_bug.cgi?id=104403
3329
3330         Reviewed by Stephen White.
3331
3332         I thought I had fixed this a while ago, but the chromium port was
3333         using custom logic to figure out which platform/* tests to run (or
3334         skip) and as a result we were getting some odd behavior (like running
3335         tests in chromium-linux on chromium-mac, but only when there was
3336         an expectation listed for the test).
3337
3338         Ports are supposed to only run the tests found under their
3339         baseline_search_path() by default.
3340
3341         Not writing a test for this; theoretically there shouldn't been a
3342         test checking that the behavior for Chromium was *different* than
3343         the other ports before.
3344
3345         * Scripts/webkitpy/layout_tests/port/chromium.py:
3346         (ChromiumPort._missing_symbol_to_skipped_tests):
3347
3348 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3349
3350         fix regression in test-webkitpy after r137189
3351         https://bugs.webkit.org/show_bug.cgi?id=104589
3352
3353         Reviewed by Eric Seidel.
3354
3355         Need to specify full port names.
3356
3357         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3358         (MainTest.test_build_check):
3359
3360 2012-12-10  Laszlo Gombos  <l.gombos@samsung.com>
3361
3362         [EFL] Change the minimum required EFL version to 1.6
3363         https://bugs.webkit.org/show_bug.cgi?id=104431
3364
3365         Reviewed by Kenneth Rohde Christiansen.
3366
3367         Change the minimum required EFL version to 1.6 from 1.7.
3368         The elementary EFL package is only required to build MiniBrowser,
3369         so I moved the required only to the Minibrowser CMake file.
3370
3371         * MiniBrowser/efl/CMakeLists.txt: In addition removed
3372         ENABLE_GLIB_SUPPORT as it is obsolete and it was always set to be on.
3373
3374 2012-12-10  Dean Jackson  <dino@apple.com>
3375
3376         Unreviewed. Adding Antoine to the contributor list so his name autocompletes in bugzilla.
3377
3378         * Scripts/webkitpy/common/config/committers.py:
3379
3380 2012-12-10  Zan Dobersek  <zandobersek@gmail.com>
3381
3382         REGRESSION (r118735): svg/dom/complex-svgView-specification.html, svg/dom/SVGViewSpec.html, svg/dom/viewspec-parser.html failing on GTK Linux 64-bit Release
3383         https://bugs.webkit.org/show_bug.cgi?id=87734
3384
3385         Reviewed by Martin Robinson.
3386
3387         When willSendRequest should return null in testing, cancel the SoupMessage
3388         instead of setting the request's URI to about:blank.
3389
3390         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3391         (willSendRequestCallback):
3392
3393 2012-12-10  Jochen Eisinger  <jochen@chromium.org>
3394
3395         [chromium] Move the test implementation of WebKit::WebPlugin to the TestRunner library
3396         https://bugs.webkit.org/show_bug.cgi?id=104537
3397
3398         Reviewed by Tony Chang.
3399
3400         In order to hide implementation details from the public TestRunner API,
3401         I moved the plugin to WebTestPluginImpl and added a static create
3402         method to WebTestPlugin. Also, I renamed the plugin from TestWebPlugin
3403         to WebTestPlugin to be more in line with other classes in the
3404         TestRunner API.
3405
3406         * DumpRenderTree/DumpRenderTree.gypi:
3407         * DumpRenderTree/chromium/TestRunner/public/WebTestPlugin.h: Added.
3408         (WebTestRunner):
3409         (WebTestPlugin):
3410         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp: Renamed from Tools/DumpRenderTree/chromium/TestWebPlugin.cpp.
3411         (WebTestRunner::WebTestPlugin::create):
3412         (WebTestRunner):
3413         (WebTestRunner::WebTestPlugin::WebTestPlugin):
3414         (WebTestRunner::WebTestPlugin::~WebTestPlugin):
3415         (WebTestRunner::WebTestPlugin::mimeType):
3416         * DumpRenderTree/chromium/TestWebPlugin.h: Removed.
3417         * DumpRenderTree/chromium/WebViewHost.cpp:
3418         (WebViewHost::createPlugin):
3419
3420 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3421
3422         run_webkit_tests throws exception when DRT is not present
3423         https://bugs.webkit.org/show_bug.cgi?id=104565
3424
3425         Reviewed by Jochen Eisinger.
3426
3427         Don't actually try to print the bot results if there aren't
3428         any because we exited before running any tests.
3429
3430         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3431         (main):
3432         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3433         (MainTest.test_build_check):
3434
3435 2012-12-10  Benjamin Poulain  <benjamin@webkit.org>
3436
3437         Add convenience methods to use ListHashSet for a LRU cache
3438         https://bugs.webkit.org/show_bug.cgi?id=104499
3439
3440         Reviewed by Sam Weinig.
3441
3442         Test the new methods added to ListHashSet.
3443
3444         * TestWebKitAPI/CMakeLists.txt:
3445         * TestWebKitAPI/GNUmakefile.am:
3446         * TestWebKitAPI/TestWebKitAPI.gypi:
3447         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
3448         * TestWebKitAPI/Tests/WTF/ListHashSet.cpp: Added.
3449         (TestWebKitAPI):
3450         (TestWebKitAPI::TEST):
3451
3452 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3453
3454         webkit-patch print-expectations doesn't work right for platforms w/ shared expectations
3455         https://bugs.webkit.org/show_bug.cgi?id=104405
3456
3457         Reviewed by Eric Seidel.
3458
3459         If multiple ports shared a TestExpectations file, webkit-patch
3460         print-expectations would use the expectation for the first port
3461         using that file and print the result for all ports, rather than
3462         re-parse and print each port separately (e.g., we might print
3463         the chromium-linux expectation for a file for both chromium-linux
3464         and chromium-mac). Discovered while testing/debugging the fix
3465         for bug 104403.
3466
3467         * Scripts/webkitpy/layout_tests/port/test.py:
3468         (TestPort._skipped_tests_for_unsupported_features):
3469         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3470         (lint):
3471         * Scripts/webkitpy/tool/commands/queries.py:
3472         (PrintExpectations._model):
3473         * Scripts/webkitpy/tool/commands/queries_unittest.py:
3474         (PrintExpectationsTest.test_platform):
3475
3476 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3477
3478         nrwt: fix total number of tests computed in the result summary
3479         https://bugs.webkit.org/show_bug.cgi?id=104061
3480
3481         Reviewed by Jochen Eisinger.
3482
3483         Good thing I left this open, because I actually broke it again
3484         the same way in a merge. Fix the computation of the one_line_summary
3485         again, and add tests for it (also clean up some redundant tests).
3486
3487         Also, add constants into the test port code so that the knowledge
3488         of how many tests should fail and how is next to the list of tests,
3489         rather than a totally different place.
3490
3491         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
3492         (LayoutTestRunner.run_tests):
3493         * Scripts/webkitpy/layout_tests/port/test.py:
3494         (TestList.__getitem__):
3495         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3496         (get_test_results):
3497         (MainTest.setUp):
3498         (MainTest.test_basic):
3499         (MainTest.test_run_singly_actually_runs_tests):
3500         (MainTest.test_test_list_with_prefix):
3501         (MainTest.test_missing_and_unexpected_results_with_custom_exit_code.CustomExitCodePort):
3502         (MainTest.test_tolerance.ImageDiffTestPort):
3503         (EndToEndTest.parse_full_results):
3504
3505 2012-12-10  Mike West  <mkwst@chromium.org>
3506
3507         Log to console when script is blocked by sandbox attributes.
3508         https://bugs.webkit.org/show_bug.cgi?id=104365
3509
3510         Reviewed by Ojan Vafai.
3511
3512         * Scripts/webkitpy/performance_tests/perftest.py:
3513         (PerfTest):
3514             Ignoring the new console message in perftests, as it broke parsing
3515             for the Parser/html-parser.html microbenchmark.
3516
3517 2012-12-10  Filip Pizlo  <fpizlo@apple.com>
3518
3519         JSC profiling and debug dump code should use inferred names when possible
3520         https://bugs.webkit.org/show_bug.cgi?id=104519
3521
3522         Reviewed by Oliver Hunt.
3523
3524         The format I'm using for referring to a code block is now name#hash. For example,
3525         v8-crypto has something called bnpSquareTo#B5QFbU. The profiler allows you to use
3526         either the hash, the inferred name, or the combined hash and full name when referring
3527         to blocks.
3528
3529         * Scripts/display-profiler-output:
3530
3531 2012-12-09  Filip Pizlo  <fpizlo@apple.com>
3532
3533         Profiler should say things about OSR exits
3534         https://bugs.webkit.org/show_bug.cgi?id=104497
3535
3536         Reviewed by Oliver Hunt.
3537
3538         Adds support for displaying OSR exit information for full summary (just displays the
3539         counts and the number of recompilations), bytecode display (says which bytecodes
3540         exited), and DFG display (annotates disassembly with taken OSR exits and their
3541         counts).
3542
3543         * Scripts/display-profiler-output:
3544
3545 2012-12-10  Alexis Menard  <alexis@webkit.org>
3546
3547         [CSS3 Backgrounds and Borders] Remove CSS3_BACKGROUND feature flag.
3548         https://bugs.webkit.org/show_bug.cgi?id=104539
3549
3550         Reviewed by Antonio Gomes.
3551
3552         As discussed on webkit-dev it is not needed to keep this feature flag 
3553         as support for <position> type is a small feature that is already 
3554         implemented by three other UAs. It was useful while landing this 
3555         feature as partial bits were landed one after one.
3556
3557         * Scripts/webkitperl/FeatureList.pm:
3558         * qmake/mkspecs/features/features.pri:
3559
3560 2012-12-10  Martin Robinson  <mrobinson@igalia.com>
3561
3562         Build fix.
3563
3564         Add a patchfile missing from my previous commit.
3565
3566         * gtk/patches/harfbuzz-icu-detection-fix.patch: Added.
3567
3568 2012-12-10  Martin Robinson  <mrobinson@igalia.com>
3569
3570         REGRESSION(137127): Causes assertion failures on the debug bots
3571         https://bugs.webkit.org/show_bug.cgi?id=104549
3572
3573         Reviewed by Alejandro G. Castro.
3574
3575         After handling the new TextOutput message we should return the method
3576         handling messages received.
3577
3578         * WebKitTestRunner/TestInvocation.cpp:
3579         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle): Fix a missing early return.
3580
3581 2012-12-09  Gustavo Noronha Silva  <gustavo.noronha@collabora.com>
3582
3583         [Soup] utilize multipart/x-mixed-replace support recently added to libsoup
3584         https://bugs.webkit.org/show_bug.cgi?id=94515
3585
3586         Reviewed by Martin Robinson.
3587
3588         This change requires a very recent libsoup, so update our jhbuild
3589         requirements.
3590
3591         * gtk/jhbuild.modules: depend on soup version that adds the multipart/x-mixed-replace
3592         support
3593         * efl/jhbuild.modules: ditto
3594
3595 2012-12-10  Zoltan Nyul  <zoltan.nyul@intel.com>
3596
3597         Implement testRunner.setViewModeMediaFeature() in WebKitTestRunner
3598         https://bugs.webkit.org/show_bug.cgi?id=103886
3599
3600         Reviewed by Kenneth Rohde Christiansen.
3601
3602         Implement testRunner.setViewModeMediaFeature() in WebKitTestRunner.
3603         This API is required by layout tests.
3604
3605         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
3606         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
3607         (WTR::TestRunner::setViewModeMediaFeature):
3608         (WTR):
3609         * WebKitTestRunner/InjectedBundle/TestRunner.h:
3610         (TestRunner):
3611
3612 2012-12-10  Martin Robinson  <mrobinson@igalia.com>
3613
3614         [GTK] Bring Harfbuzz-ng support to Gtk
3615         https://bugs.webkit.org/show_bug.cgi?id=92098
3616
3617         Reviewed by Gustavo Noronha Silva.
3618
3619         Add HarfBuzz dependency.
3620
3621         * GNUmakefile.am: Add FreeType/HarfBuzz CFLAGS to the WebCore Internals library.
3622         * gtk/jhbuild.modules: Add HarfBuzz to the JHbuild and remove Pango, since it
3623         should no longer affect layout test results.
3624
3625 2012-12-10  Simon Hausmann  <simon.hausmann@digia.com>
3626
3627         [Qt] Use QLibraryInfo::LibraryExecutablesPath unconditionally
3628         https://bugs.webkit.org/show_bug.cgi?id=104541
3629
3630         Reviewed by Jocelyn Turcotte.
3631
3632         We now depend on a Qt 5 version that is guaranteed to have this API, so we
3633         can remove the configure checks for it.
3634
3635         * qmake/config.tests/libexecdir/libexecdir.cpp: Removed.
3636         * qmake/config.tests/libexecdir/libexecdir.pro: Removed.
3637         * qmake/mkspecs/features/features.prf:
3638
3639 2012-12-10  Víctor Manuel Jáquez Leal  <vjaquez@igalia.com>
3640
3641         REGRESSION(r135863): GtkLauncher --help no longer displays all help options
3642         https://bugs.webkit.org/show_bug.cgi?id=104382
3643
3644         Reviewed by Philippe Normand.
3645
3646         Remove gst_init() because it steals the GOptionContext.
3647         gst_init_get_option_group() already does the initialization job.
3648
3649         * GtkLauncher/main.c:
3650         (main):
3651
3652 2012-12-10  Zan Dobersek  <zandobersek@gmail.com>
3653
3654         [GTK][WTR] Implement AccessibilityUIElement::titleUIElement
3655         https://bugs.webkit.org/show_bug.cgi?id=102953
3656
3657         Reviewed by Martin Robinson.
3658
3659         Implement AccessibilityUIElement::titleUIElement in the ATK backend
3660         in the same manner it is implemented in DumpRenderTree. Both implementations
3661         are updated to unref the relation set before the final function return
3662         rather than at the end of each loop through the relations held in the said
3663         relation set.
3664
3665         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
3666         (AccessibilityUIElement::titleUIElement):
3667         * WebKitTestRunner/InjectedBundle/atk/AccessibilityUIElementAtk.cpp:
3668         (WTR::AccessibilityUIElement::titleUIElement):
3669
3670 2012-12-10  Martin Robinson  <mrobinson@igalia.com>
3671
3672         Fix the WebKit2 debug build
3673
3674         * WebKitTestRunner/TestInvocation.cpp:
3675         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle): Fix a typo of WKStringGetTypeID().
3676
3677 2012-12-10  Martin Robinson  <mrobinson@igalia.com>
3678
3679         [WTR] Move text output accumulation to the UIProcess
3680         https://bugs.webkit.org/show_bug.cgi?id=104214
3681
3682         Reviewed by Darin Adler.
3683
3684         Instead of accumulating text output in the InjectedBundle and then sending it to the UIProcess 
3685         once a test is finished, immediately send any text output to the UIProcess. This will allow
3686         WebKitTestRunner to output text from the UIProcess as well.
3687
3688         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp: Remove the handling of the text output
3689         StringBuilder. Add a method to send output to the UIProcess.
3690         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
3691         (InjectedBundle):
3692         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp: Change code that appends to the StringBuilder
3693         to use the new outputText method.
3694         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
3695         * WebKitTestRunner/InjectedBundle/TestRunner.cpp: Ditto.
3696         * WebKitTestRunner/InjectedBundle/atk/AccessibilityControllerAtk.cpp: Ditto.
3697         * WebKitTestRunner/TestInvocation.cpp:
3698         (WTR::TestInvocation::TestInvocation): Intialize the StringBuilder.
3699         (WTR::TestInvocation::invoke): Clear the StringBuilder when a new tests starts.
3700         (WTR::TestInvocation::dumpResults): m_textOutput is a StringBuilder now so the
3701         method of printing the text is slightly different.
3702         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle): Handle the new TextOutput
3703         message by appending the results to the StringBuilder.
3704         * WebKitTestRunner/TestInvocation.h:
3705         (TestInvocation): Switch the type of m_textOutput.
3706
3707 2012-12-10  Mihnea Ovidenie  <mihnea@adobe.com>
3708
3709         Unreviewed. Adding Mihai Maerean to the list of contributors, working on CSSRegions.
3710
3711         * Scripts/webkitpy/common/config/committers.py:
3712
3713 2012-12-10  Joanmarie Diggs  <jdiggs@igalia.com>
3714
3715         [GTK] accessibility/language-attribute.html is failing
3716         https://bugs.webkit.org/show_bug.cgi?id=98369
3717
3718         Reviewed by Chris Fleizach.
3719
3720         The test is failing for two reasons: Not exposing the language
3721         for accessible objects and AccessibilityUIElement::language() not
3722         being implemented.
3723
3724         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
3725         (AccessibilityUIElement::language): Implemented.
3726
3727 2012-12-10  Joanmarie Diggs  <jdiggs@igalia.com>
3728
3729         [GTK] accessibility/editable-webarea-context-menu-point.html is failing
3730         https://bugs.webkit.org/show_bug.cgi?id=98364
3731
3732         Reviewed by Chris Fleizach.
3733
3734         The test is failing because the clickPoint values remain unchanged. The
3735         values remain unchanged because ATK lacks a means to keep track of this
3736         information.
3737
3738         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
3739         (AccessibilityUIElement::clickPointX): Added comment that ATK doesn't have clickPoint details
3740         (AccessibilityUIElement::clickPointY): Added comment that ATK doesn't have clickPoint details
3741
3742 2012-12-09  Alejandro Piñeiro  <apinheiro@igalia.com>
3743
3744         [GTK] lack of implemention of AccessibilityUIElementGtk::isSelectable and AccessibilityUIElementGtk::isMultiSelectable
3745         https://bugs.webkit.org/show_bug.cgi?id=104481
3746
3747         Reviewed by Martin Robinson.
3748
3749         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
3750         (AccessibilityUIElement::isSelectable): Implement this by checking
3751         the current state of the AtkObject.
3752         (AccessibilityUIElement::isMultiSelectable): Implement this by
3753         checking the current state of the AtkObject.
3754
3755 2012-12-09  Joanmarie Diggs  <jdiggs@igalia.com>
3756
3757         [GTK] accessibility/placeholder.html is failing
3758         https://bugs.webkit.org/show_bug.cgi?id=98373
3759
3760         Reviewed by Martin Robinson.
3761
3762         The test was failing because the placeholder text was not supported and
3763         AccessibilityUIElement::stringAttributeValue() was not implemented.
3764
3765         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
3766         (coreAttributeToAtkAttribute): New convenience method to convert WebCore attribute
3767         names into AtkObject attribute names
3768         (AccessibilityUIElement::stringAttributeValue): implemented
3769
3770 2012-12-09  Dan Winship  <danw@gnome.org>
3771
3772         [GTK] Bump libxml2 requirement in jhbuild.modules
3773         https://bugs.webkit.org/show_bug.cgi?id=104477
3774
3775         Reviewed by Martin Robinson.
3776
3777         Fedora 18's Apache depends on symbols in libxml2 2.9.0, so the
3778         tests won't run if we only have libxml2 2.8.0 installed in the
3779         jhbuild root.
3780
3781         * gtk/jhbuild.modules:
3782
3783 2012-12-09  Zan Dobersek  <zandobersek@gmail.com>
3784
3785         Remove testRunner.disableImageLoading()
3786         https://bugs.webkit.org/show_bug.cgi?id=98083
3787
3788         Reviewed by Gustavo Noronha Silva.
3789
3790         Remove the disableImageLoading methods and callback.
3791         No tests use the method anymore and all ports except Wx
3792         now support this behavior through overriding the
3793         WebKitDisplayImagesKey preference.
3794
3795         * DumpRenderTree/TestRunner.cpp:
3796         (TestRunner::staticFunctions):
3797         * DumpRenderTree/TestRunner.h:
3798         (TestRunner):
3799         * DumpRenderTree/blackberry/TestRunnerBlackBerry.cpp:
3800         (TestRunner::overridePreference): Add support for overriding the
3801         WebKitDisplayImagesKey preference which is now the preferred way
3802         of disabling image loading.
3803         * DumpRenderTree/efl/TestRunnerEfl.cpp:
3804         * DumpRenderTree/gtk/TestRunnerGtk.cpp:
3805         * DumpRenderTree/mac/TestRunnerMac.mm:
3806         * DumpRenderTree/qt/TestRunnerQt.cpp:
3807         * DumpRenderTree/qt/TestRunnerQt.h:
3808         (TestRunner):
3809         * DumpRenderTree/win/TestRunnerWin.cpp:
3810         * DumpRenderTree/wx/TestRunnerWx.cpp:
3811
3812 2012-12-09  Joanmarie Diggs  <jdiggs@igalia.com>
3813
3814         [GTK] accessibility/label-for-control-hittest.html is failing
3815         https://bugs.webkit.org/show_bug.cgi?id=98368
3816
3817         Reviewed by Martin Robinson.
3818
3819         The test was failing because AccessibilityUIElement::elementAtPoint was
3820         not implemented.
3821
3822         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
3823         (AccessibilityUIElement::elementAtPoint): implemented
3824
3825 2012-12-09  Jon Lee  <jonlee@apple.com>
3826
3827         [WK2] Move button image to injected bundle
3828         https://bugs.webkit.org/show_bug.cgi?id=104107
3829         <rdar://problem/12813842>
3830
3831         Reviewed by Simon Fraser.
3832
3833         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
3834         (WTR::InjectedBundlePage::InjectedBundlePage): Update the bundle page UI client.
3835
3836 2012-12-08  Halton Huo  <halton.huo@intel.com>
3837
3838         [CMake] Add CMake style checker
3839         https://bugs.webkit.org/show_bug.cgi?id=104240
3840
3841         Reviewed by Eric Seidel.
3842
3843         The CMake style checker is followed by existing style checkers like cpp, python.
3844         Unit test is added as well.
3845
3846         * Scripts/webkitpy/style/checker.py: Use CMakeChecker CMAKE type files(.cmake or CMakeLists.txt)
3847         (FileType):
3848         (CheckerDispatcher._file_type):
3849         (CheckerDispatcher._create_checker):
3850         * Scripts/webkitpy/style/checkers/cmake.py: Added.
3851         (CMakeChecker):
3852         (CMakeChecker.__init__): 
3853         (CMakeChecker.check):
3854         (CMakeChecker._process_line):
3855         (CMakeChecker._check_trailing_whitespace):
3856         (CMakeChecker._check_no_space_cmds):
3857         (CMakeChecker._check_one_space_cmds):
3858         (CMakeChecker._check_non_lowercase_cmd):
3859         (CMakeChecker._check_indent):
3860         * Scripts/webkitpy/style/checkers/cmake_unittest.py: Added.
3861         (CMakeCheckerTest):
3862         (CMakeCheckerTest.test_init):
3863         (CMakeCheckerTest.test_init._mock_handle_style_error):
3864         (CMakeCheckerTest.test_check):
3865         (CMakeCheckerTest.test_check._mock_handle_style_error):
3866
3867 2012-12-08  Kondapally Kalyan  <kalyan.kondapally@intel.com>
3868
3869         [EFL] Enable WebGL by default.
3870         https://bugs.webkit.org/show_bug.cgi?id=102991.
3871
3872         Reviewed by Kenneth Rohde Christiansen.
3873
3874         Enable WebGL by default.
3875
3876         * Scripts/webkitperl/FeatureList.pm:
3877
3878 2012-12-07  Scott Violet  <sky@chromium.org>
3879
3880         [chromium] Remove linux theme related files and switch to default
3881         https://bugs.webkit.org/show_bug.cgi?id=103897
3882
3883         Reviewed by Dimitri Glazkov.
3884
3885         * Scripts/webkitpy/common/config/build_unittest.py:
3886         (ShouldBuildTest): linux->default
3887
3888 2012-12-07  Mike Lattanzio  <mlattanzio@rim.com>
3889
3890         [BlackBerry] Update BlackBerry Port to use new Graphics::createBuffer BufferType enum values.
3891         https://bugs.webkit.org/show_bug.cgi?id=104121
3892
3893         Reviewed by Rob Buis.
3894
3895         Update createBuffer call to reflect new BufferType enum.
3896         Internal PR 254649
3897         Internal Review: Jakob Petsovits
3898
3899         DRT is always backed for maximum compatibility.
3900
3901         * DumpRenderTree/blackberry/PixelDumpSupportBlackBerry.cpp:
3902         (createBitmapContextFromWebView):
3903
3904 2012-12-07  Jochen Eisinger  <jochen@chromium.org>
3905
3906         Add 'CHROMIUM_DEVEL_SANDBOX' to the environment variables to copy for layout test drivers
3907         https://bugs.webkit.org/show_bug.cgi?id=104355
3908
3909         Reviewed by Dirk Pranke.
3910
3911         This enables chromium's content_shell driver to run with the full SUID
3912         sandbox.
3913
3914         * Scripts/webkitpy/layout_tests/port/base.py:
3915         (Port.to.setup_environ_for_server):
3916
3917 2012-12-07  KyungTae Kim  <ktf.kim@samsung.com>
3918
3919         [EFL] Fix tools for WebKit2-only builds
3920         https://bugs.webkit.org/show_bug.cgi?id=103904
3921
3922         Reviewed by Laszlo Gombos.
3923
3924         The libewebkit2.so should be used for library path for EFL WebKit2.
3925
3926         * Scripts/webkitdirs.pm:
3927         (builtDylibPathForName):
3928
3929 2012-12-07  Xiaobo Wang  <xbwang@torchmobile.com.cn>
3930
3931         [BlackBerry] DRT - Update implementation of TestRunner.setCustomPolicyDelegate()
3932         https://bugs.webkit.org/show_bug.cgi?id=104243
3933
3934         RIM PR 259337
3935
3936         Reviewed by Rob Buis.
3937
3938         1. Implement calculation of navigation originating in
3939         didDecidePolicyForNavigationAction().
3940         2. Implement policy delegate for resource response.
3941         3. Enable customPolicyDelegate if waitForPolicyDelegate() is called.
3942         4. Minor optional updates in dump() and dumpFrameLoadCallbacks().
3943
3944         The patch fixed the following tests:
3945         - http/tests/download
3946         - fast/forms/mailto/formenctype-attribute-button-html.html
3947         - fast/forms/mailto/formenctype-attribute-input-html.html
3948         - fast/encoding/mailto-always-utf-8.html
3949
3950         * DumpRenderTree/blackberry/DumpRenderTree.cpp:
3951         (BlackBerry::WebKit::DumpRenderTree::DumpRenderTree):
3952         (BlackBerry::WebKit::DumpRenderTree::resetToConsistentStateBeforeTesting):
3953         (BlackBerry::WebKit::DumpRenderTree::dump):
3954         (BlackBerry::WebKit::DumpRenderTree::didFinishLoadForFrame):
3955         (BlackBerry::WebKit::DumpRenderTree::didDecidePolicyForNavigationAction):
3956         (BlackBerry::WebKit::DumpRenderTree::didDecidePolicyForResponse):
3957         (BlackBerry::WebKit::DumpRenderTree::setCustomPolicyDelegate):
3958         * DumpRenderTree/blackberry/DumpRenderTreeBlackBerry.h:
3959         (DumpRenderTree):
3960         * DumpRenderTree/blackberry/TestRunnerBlackBerry.cpp:
3961         (TestRunner::setCustomPolicyDelegate):
3962         (TestRunner::waitForPolicyDelegate):
3963         (TestRunner::overridePreference):
3964
3965 2012-12-07  Gwang Yoon Hwang  <ryumiel@company100.net>
3966
3967         [Qt] Adjust library name in layout_tests/port/qt.py after Qt5-r40
3968         https://bugs.webkit.org/show_bug.cgi?id=104348
3969
3970         Reviewed by Csaba Osztrogonác.
3971
3972         After Qt-5.0.0-r40, the library was renamed from QtWebKitWidgets to
3973         Qt5WebKitWidgets. run-webkit-tests were still looking for QtWebKitWidgets
3974
3975         * Scripts/webkitpy/layout_tests/port/qt.py:
3976         (QtPort._path_to_webcore_library):
3977
3978 2012-12-07  Csaba Osztrogonác  <ossy@webkit.org>
3979
3980         Unreviewed fix, unbreak run-launcher after library renaming.
3981
3982         * Scripts/webkitdirs.pm:
3983         (builtDylibPathForName):
3984
3985 2012-12-06  Csaba Osztrogonác  <ossy@webkit.org>
3986
3987         Unreviewed fix after r136896, unbreak --debug-rwt to make bots result verbose again.
3988         https://bugs.webkit.org/show_bug.cgi?id=104296
3989
3990         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3991         (main):
3992
3993 2012-12-06  Rick Byers  <rbyers@chromium.org>
3994
3995         CSS cursor property should support webkit-image-set
3996         https://bugs.webkit.org/show_bug.cgi?id=99493
3997
3998         Reviewed by Beth Dakin.
3999
4000         Add ENABLE_MOUSE_CURSOR_SCALE - disabled by default
4001
4002         * Scripts/webkitperl/FeatureList.pm:
4003         * qmake/mkspecs/features/features.pri:
4004
4005 2012-12-06  Vivek Galatage  <vivek.vg@samsung.com>
4006
4007         Unreviewed. Fixing the email address order to work correctly with bugzilla.
4008
4009         * Scripts/webkitpy/common/config/committers.py:
4010
4011 2012-12-06  Jochen Eisinger  <jochen@chromium.org>
4012
4013         [chromium] when forwarding events through the WebPluginContainer, create a UserGestureIndicator if processing a user gesture
4014         https://bugs.webkit.org/show_bug.cgi?id=104306
4015
4016         Reviewed by Dimitri Glazkov.
4017
4018         * DumpRenderTree/chromium/TestWebPlugin.cpp:
4019         (TestWebPlugin::TestWebPlugin):
4020         (TestWebPlugin::handleInputEvent): add support for printing the user gesture status
4021         * DumpRenderTree/chromium/TestWebPlugin.h:
4022         (TestWebPlugin):
4023
4024 2012-12-06  Dirk Pranke  <dpranke@chromium.org>
4025
4026         new-run-webkit-tests --lint-test-files seems to be broken
4027         https://bugs.webkit.org/show_bug.cgi?id=104296
4028
4029         Unreviewed, build fix.
4030
4031         My recent refactoring caused me to not initialize the printer
4032         before calling lint(); this patch restructures the code to make
4033         the lint routine self-contained and do all the work necessary.
4034         This will also make it easier to move this code out into a standalone
4035         file.
4036
4037         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
4038         (lint):
4039         (main):
4040         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
4041         (LintTest.test_all_configurations):
4042         (LintTest.test_lint_test_files):
4043         (LintTest.test_lint_test_files__errors):
4044
4045 2012-12-05  Filip Pizlo  <fpizlo@apple.com>
4046
4047         DFG profiler should be helpful about gem installation
4048         https://bugs.webkit.org/show_bug.cgi?id=104217
4049
4050         Reviewed by Oliver Hunt.
4051         
4052         'json' and 'highline' are not installed by default on the ruby distributions with
4053         which I am familiar, and the default error messages don't make me happy. This makes
4054         display-profiler-output print a helpful message if those gems are not found.
4055
4056         * Scripts/display-profiler-output:
4057
4058 2012-12-05  Filip Pizlo  <fpizlo@apple.com>
4059
4060         Profiler should print a helpful message if you pass the wrong arguments
4061         https://bugs.webkit.org/show_bug.cgi?id=104222
4062
4063         Reviewed by Oliver Hunt.
4064
4065         * Scripts/display-profiler-output:
4066
4067 2012-12-06  Ojan Vafai  <ojan@chromium.org>
4068
4069         Allow for CCing a secondary email address from watchlists
4070         https://bugs.webkit.org/show_bug.cgi?id=104286
4071
4072         Reviewed by Dirk Pranke.
4073
4074         Change my watchlist CC address so I can filter these separately from
4075         when people explicitly CC me and update the watchlist parser to allow that.
4076
4077         * Scripts/webkitpy/common/config/committers.py:
4078         * Scripts/webkitpy/common/config/watchlist:
4079         * Scripts/webkitpy/common/watchlist/watchlistparser.py:
4080         (WatchListParser._validate):
4081         * Scripts/webkitpy/common/watchlist/watchlistparser_unittest.py:
4082         (WatchListParserTest.test_cc_rule_with_invalid_email):
4083         (WatchListParserTest.test_cc_rule_with_secondary_email):
4084
4085 2012-12-06  Max Vujovic  <mvujovic@adobe.com>
4086
4087         Unreviewed. Add myself to committers.py.
4088
4089         * Scripts/webkitpy/common/config/committers.py:
4090
4091 2012-12-06  Yael Aharon  <yael.aharon@intel.com>
4092
4093         [EFL] [WK2] fast/dom/vertical-scrollbar-in-rtl.html makes fast/regions tests flaky
4094         https://bugs.webkit.org/show_bug.cgi?id=104139
4095
4096         Reviewed by Kenneth Rohde Christiansen.
4097
4098         Resize the window back to 800x600 between tests.
4099         This should be a no-op if the test did not resize the window.
4100
4101         * WebKitTestRunner/TestController.cpp:
4102         (WTR::TestController::resetStateToConsistentValues):
4103
4104 2012-12-06  Tommy Widenflycht  <tommyw@google.com>
4105
4106         Speech Recognition API: Change the error code to a string on SpeechRecognitionError
4107         https://bugs.webkit.org/show_bug.cgi?id=104254
4108
4109         Reviewed by Adam Barth.
4110
4111         Changes the mock error function to take a string instead of an int.
4112
4113         * DumpRenderTree/chromium/DRTTestRunner.cpp:
4114         (DRTTestRunner::setMockSpeechRecognitionError):
4115         * DumpRenderTree/chromium/MockWebSpeechRecognizer.cpp:
4116         (WebTestRunner::ErrorTask::ErrorTask):
4117         (MockWebSpeechRecognizer::setError):
4118         * DumpRenderTree/chromium/MockWebSpeechRecognizer.h:
4119         (MockWebSpeechRecognizer):
4120
4121 2012-12-06  Andras Becsi  <andras.becsi@digia.com>
4122
4123         [Qt][Mac] Fix libxslt and libxml2 config tests
4124         https://bugs.webkit.org/show_bug.cgi?id=104164
4125
4126         Reviewed by Simon Hausmann.
4127
4128         We should not use pkg-config on Mac instead use direct include
4129         paths and add needed libraries to the linker options to detect
4130         libxslt and libxml2 provided by the system.
4131         Previously we would always fall back to qtxmlpatterns.
4132
4133         * qmake/config.tests/libxml2/libxml2.pro:
4134         * qmake/config.tests/libxslt/libxslt.pro:
4135
4136 2012-12-06  Csaba Osztrogonác  <ossy@webkit.org>
4137
4138         [Qt] Fix the build if the path contains "+" character
4139         https://bugs.webkit.org/show_bug.cgi?id=104120
4140
4141         Reviewed by Tor Arne Vestbø.
4142
4143         The second parameter of qmake's replace() is regular expression, so
4144         we have to escape special characters in paths passed to replace().
4145
4146         * qmake/mkspecs/features/default_post.prf:
4147         * qmake/mkspecs/features/default_pre.prf:
4148         * qmake/mkspecs/features/functions.prf:
4149
4150 2012-12-06  Sheriff Bot  <webkit.review.bot@gmail.com>
4151
4152         Unreviewed, rolling out r136825.
4153         http://trac.webkit.org/changeset/136825
4154         https://bugs.webkit.org/show_bug.cgi?id=104251
4155
4156         It made 100+ test fail on Qt-WK2 pixel bot, and 1800+ test
4157         flakey on Qt-WK2 non-pixel bot (Requested by Ossy on #webkit).
4158
4159         * WebKitTestRunner/TestController.cpp:
4160         (WTR::TestController::resetStateToConsistentValues):
4161
4162 2012-12-06  Yael Aharon  <yael.aharon@intel.com>
4163
4164         [EFL] [WK2] fast/dom/vertical-scrollbar-in-rtl.html makes fast/regions tests flaky
4165         https://bugs.webkit.org/show_bug.cgi?id=104139
4166
4167         Reviewed by Kenneth Rohde Christiansen.
4168
4169         Resize the window back to 800x600 between tests.
4170         This should be a no-op if the test did not resize the window.
4171
4172         * WebKitTestRunner/TestController.cpp:
4173         (WTR::TestController::resetStateToConsistentValues):
4174
4175 2012-12-06  Sheriff Bot  <webkit.review.bot@gmail.com>
4176
4177         Unreviewed, rolling out r136784 and r136802.
4178         http://trac.webkit.org/changeset/136784
4179         http://trac.webkit.org/changeset/136802
4180         https://bugs.webkit.org/show_bug.cgi?id=104231
4181
4182         breaks chromium canary (Requested by morrita on #webkit).
4183
4184         * Scripts/webkitpy/common/config/build_unittest.py:
4185         (ShouldBuildTest):
4186
4187 2012-12-05  Ryosuke Niwa  <rniwa@webkit.org>
4188
4189         Add myself to watch changes in PerformanceTests/*.
4190
4191         * Scripts/webkitpy/common/config/watchlist:
4192
4193 2012-12-05  Shinya Kawanaka  <shinyak@chromium.org>
4194
4195         [Chromium] DRT on Windows crashes when a testcase has a progress element in indeterminate state
4196         https://bugs.webkit.org/show_bug.cgi?id=102459
4197
4198         Reviewed by Dimitri Glazkov.
4199
4200         Since WebThemeControlDRTWin::markState does not consider indeterminate state, DRT caused a crash.
4201
4202         * DumpRenderTree/chromium/WebThemeControlDRTWin.cpp:
4203         (WebThemeControlDRTWin::markState):
4204
4205 2012-12-05  Filip Pizlo  <fpizlo@apple.com>
4206
4207         DFG profiler should be able to display where a code block was inlined into, and what it inlined
4208         https://bugs.webkit.org/show_bug.cgi?id=104216
4209
4210         Reviewed by Oliver Hunt.
4211
4212         Added an "inlines" command that shows the inline stacks that a code block participated
4213         in. Also cleaned up some of the ways we print compilation units.
4214
4215         * Scripts/display-profiler-output:
4216
4217 2012-12-05  Halton Huo  <halton.huo@intel.com>
4218
4219         [CMake] Unify coding style for CMake files
4220         https://bugs.webkit.org/show_bug.cgi?id=103605
4221
4222         Reviewed by Laszlo Gombos.
4223
4224         Update cmake files(.cmake, CMakeLists.txt) with following style rules:
4225         1. Indentation
4226         1.1 Use spaces, not tabs.
4227         1.2 Four spaces as indent.
4228         2. Spacing
4229         2.1 Place one space between control statements and their parentheses.
4230             For eg, if (), else (), elseif (), endif (), foreach (),
4231             endforeach (), while (), endwhile (), break ().
4232         2.2 Do not place spaces between function and macro statements and
4233             their parentheses. For eg, macro(), endmacro(), function(),
4234             endfunction().
4235         2.3 Do not place spaces between a command or function or macro and its
4236             parentheses, or between a parenthesis and its content. For eg,
4237             message("testing") not message( "testing") or message ("testing" )
4238         2.4 No space at line ending.
4239         3. Lowercase when call commands macros and functions. For eg,
4240            add_executable() not ADD_EXECUTABLE(), set() not SET().
4241
4242         * CMakeLists.txt:
4243         * DumpRenderTree/TestNetscapePlugIn/CMakeLists.txt:
4244         * DumpRenderTree/efl/CMakeLists.txt:
4245         * EWebLauncher/CMakeLists.txt:
4246         * EWebLauncher/ControlTheme/CMakeLists.txt:
4247         * MiniBrowser/efl/CMakeLists.txt:
4248         * TestWebKitAPI/CMakeLists.txt:
4249         * TestWebKitAPI/PlatformEfl.cmake:
4250         * WebKitTestRunner/CMakeLists.txt:
4251         * WebKitTestRunner/PlatformEfl.cmake:
4252         * WinCELauncher/CMakeLists.txt:
4253         * clang/ReportMemoryUsagePlugin/CMakeLists.txt:
4254
4255 2012-12-05  Scott Violet  <sky@chromium.org>
4256
4257         [chromium] Remove linux theme related files and switch to default
4258         https://bugs.webkit.org/show_bug.cgi?id=103897
4259
4260         Reviewed by Tony Chang.
4261
4262         * Scripts/webkitpy/common/config/build_unittest.py:
4263         (ShouldBuildTest): linux->default
4264
4265 2012-12-05  Dirk Pranke  <dpranke@chromium.org>
4266
4267         nrwt: run_webkit_tests_integrationtest.py depends on being able to share a port w/ between master and worker
4268         https://bugs.webkit.org/show_bug.cgi?id=78168
4269
4270         Reviewed by Ojan Vafai.
4271
4272         This patch adds a single ordered list to ResultSummary that contains
4273         the results of every test run in order of result received, and
4274         modifies the TestResults class to contain the list of references used
4275         for each test as well as the pid of the driver; this allows us to stop
4276         having to override the test drivers to record the tests, and makes the
4277         integration tests that test rdering, retries, restarts, etc. black-box
4278         tests.
4279
4280         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
4281         (SingleTestRunner._run_rebaseline):
4282         (SingleTestRunner._compare_output):
4283         (SingleTestRunner._run_reftest):
4284         (SingleTestRunner._compare_output_with_reference):
4285         * Scripts/webkitpy/layout_tests/models/result_summary.py:
4286         (ResultSummary.__init__):
4287         (ResultSummary.add):
4288         * Scripts/webkitpy/layout_tests/models/test_results.py:
4289         (TestResult.__init__):
4290         * Scripts/webkitpy/layout_tests/port/driver.py:
4291         (DriverOutput.__init__):
4292         (Driver.run_test):
4293         * Scripts/webkitpy/layout_tests/port/test.py:
4294         (TestDriver):
4295         (TestDriver.__init__):
4296         (TestDriver.run_test):
4297         (TestDriver.stop):
4298         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
4299         (get_tests_run):
4300         (get_test_batches):
4301         (get_test_results):
4302         (MainTest.test_batch_size):
4303         (MainTest.test_dryrun):
4304         (MainTest.test_natural_order):
4305         (MainTest.test_natural_order_test_specified_multiple_times):
4306         (MainTest.test_random_order):
4307         (MainTest.test_random_order_test_specified_multiple_times):
4308         (MainTest.test_no_order):
4309         (MainTest.test_no_order_test_specified_multiple_times):
4310         (MainTest.test_no_order_with_directory_entries_in_natural_order):
4311         (MainTest.test_repeat_each):
4312         (MainTest.test_ignore_flag):
4313         (MainTest.test_skipped_flag):
4314         (MainTest.test_iterations):
4315   &n