Add unit test for default builder for the dashboards.
[WebKit-https.git] / Tools / ChangeLog
1 2013-01-03  Julie Parent  <jparent@chromium.org>
2
3         Add unit test for default builder for the dashboards.
4         https://bugs.webkit.org/show_bug.cgi?id=105972
5
6         Reviewed by Dirk Pranke.
7
8         * TestResultServer/static-dashboards/loader_unittests.js:
9
10 2013-01-03  Tony Chang  <tony@chromium.org>
11
12         Generate internal.settings from Settings.in
13         https://bugs.webkit.org/show_bug.cgi?id=104740
14
15         Reviewed by Adam Barth.
16
17         * GNUmakefile.am: Compile new generated files on GTK+.
18         * qmake/mkspecs/features/default_post.prf: Qt: Make it so an action
19         with both extra_sources and add_output_to_sources=false will still
20         compile the files in extra_sources. This makes it possible to compile
21         InternalSettingsGenerated.cpp and not be confused by
22         InternalSettingsGenerated.idl.
23
24 2013-01-03  Zan Dobersek  <zandobersek@gmail.com>
25
26         [style] WebIDL-reflecting uppercase enums reported as style violation
27         https://bugs.webkit.org/show_bug.cgi?id=105832
28
29         Reviewed by Tony Chang.
30
31         Recognize the enum as WebIDL-reflecting one if there's a comment present that
32         indicates this. In enums spanning over multiple lines, the comment is expected
33         on the first line. In enums placed in one line, the comment is expected at the
34         end of that line.
35
36         WebIDL-reflecting enum values are required to be written in uppercase and as such
37         violate the common rule of these values being written in InterCaps. Annotating
38         enums this way makes it possible to avoid false style violation reports.
39
40         The comment is of form '// Web(?:Kit)?IDL enum'.
41
42         * Scripts/webkitpy/style/checkers/cpp.py:
43         (_EnumState.__init__):
44         (_EnumState.process_clean_line):
45         (check_enum_casing):
46         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
47         (NoNonVirtualDestructorsTest.test_enum_casing):
48
49 2013-01-03  Tim 'mithro' Ansell  <mithro@mithis.com>
50
51         Adding favicon.ico to the garden-o-matic.
52         https://bugs.webkit.org/show_bug.cgi?id=105982
53
54         Reviewed by Adam Barth.
55
56         * Scripts/webkitpy/tool/servers/gardeningserver.py:
57         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/favicon.ico:
58
59 2013-01-03  Jonathan Liu  <net147@gmail.com>
60
61         [Qt] Fix library linking order
62         https://bugs.webkit.org/show_bug.cgi?id=105818
63
64         Some Qt libraries are linked before the private static libraries
65         that depend on them which results in QtWebKit failing to link.
66
67         Copy Qt module dependencies from QT to QT_PRIVATE when creating
68         module to ensure the Qt libraries are linked after the private
69         static libraries that depend on them.
70
71         Reviewed by Tor Arne Vestbø.
72
73         * qmake/mkspecs/features/default_post.prf:
74
75 2013-01-02  Byungwoo Lee  <bw80.lee@samsung.com>
76
77         [EFL][WK2] Fix tooltip bugs in MiniBrowser.
78         https://bugs.webkit.org/show_bug.cgi?id=102209
79
80         Reviewed by Gyuyoung Kim.
81
82         Set tooltip on the elm_window instead of ewk_view, and add delay to
83         show the tooltip window.
84
85         Setting tooltip on the ewk_view makes elementary error message because
86         ewk_view is not an elementary widget. To remove the error message,
87         tooltip should be set on the elm_window.
88
89         Most browsers show the tooltip after a mouse pointer stays on an
90         element object that has tooltip text.
91         But elm_tooltip widget itself doesn't have delay or handle mouse
92         events to show/hide.
93
94         So additional codes are added to show/hide tooltip widget along with
95         mouse events.
96           1) When a mouse pointer stays for a second on an element which has
97              tooltip text, then tooltip widget is displayed.
98           2) When a mouse pointer is moved or out of ewk_view, hide the tooltip
99              widget.
100
101         * MiniBrowser/efl/main.c:
102         (_Tooltip_Information):
103         (_Browser_Window):
104         (on_tooltip_show):
105         (window_tooltip_hide):
106         (window_tooltip_update):
107         (on_mouse_in):
108         (on_mouse_move):
109         (on_mouse_out):
110         (window_free):
111         (on_tooltip_text_set):
112         (on_tooltip_text_unset):
113         (window_create):
114
115 2013-01-02  Tony Chang  <tony@chromium.org>
116
117         Unreviewed, rolling out r138661.
118         http://trac.webkit.org/changeset/138661
119         https://bugs.webkit.org/show_bug.cgi?id=104740
120
121         Compile problems on EFL
122
123         * GNUmakefile.am:
124         * qmake/mkspecs/features/default_post.prf:
125
126 2013-01-02  Tony Chang  <tony@chromium.org>
127
128         Generate internal.settings from Settings.in
129         https://bugs.webkit.org/show_bug.cgi?id=104740
130
131         Reviewed by Adam Barth.
132
133         * GNUmakefile.am: Compile new generated files on GTK+.
134         * qmake/mkspecs/features/default_post.prf: Qt: Make it so an action
135         with both extra_sources and add_output_to_sources=false will still
136         compile the files in extra_sources. This makes it possible to compile
137         InternalSettingsGenerated.cpp and not be confused by
138         InternalSettingsGenerated.idl.
139
140 2013-01-02  Adam Barth  <abarth@webkit.org>
141
142         Remove old-run-webkit-tests support from webkitpy
143         https://bugs.webkit.org/show_bug.cgi?id=105619
144
145         Reviewed by Eric Seidel.
146
147         This patch removes support for webkitpy parsing the
148         old-run-webkit-tests results format. There doesn't seem to be any
149         reason to keep this code around anymore now that most bots have
150         switched to new-run-webkit-tests.
151
152         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
153         (Builder.fetch_layout_test_results):
154         * Scripts/webkitpy/common/net/layouttestresults.py:
155         (path_for_layout_test):
156         (LayoutTestResults.results_from_string):
157         (LayoutTestResults):
158         * Scripts/webkitpy/common/net/layouttestresults_unittest.py:
159         (LayoutTestResultsTest.test_results_from_string):
160         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
161         (test_flaky_test_failure):
162         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py:
163         (LayoutTestResultsReaderTest.test_missing_layout_test_results):
164         * Scripts/webkitpy/tool/bot/patchanalysistask.py:
165         (PatchAnalysisTask._test_patch):
166         * Scripts/webkitpy/tool/commands/queuestest.py:
167         (QueuesTest.assert_queue_outputs):
168
169 2013-01-02  Heikki Paajanen  <heikki.paajanen@palm.com>
170
171         [Qt][WK2] Add simple UI to find text from page
172         https://bugs.webkit.org/show_bug.cgi?id=96481
173
174         Reviewed by Jocelyn Turcotte.
175
176         Find bar that slides between navigation bar and WebView.
177
178         * MiniBrowser/qt/BrowserWindow.cpp:
179         (BrowserWindow::toggleFind):
180         * MiniBrowser/qt/BrowserWindow.h:
181         (BrowserWindow):
182         * MiniBrowser/qt/MiniBrowser.qrc:
183         * MiniBrowser/qt/MiniBrowserApplication.cpp:
184         (MiniBrowserApplication::notify):
185         * MiniBrowser/qt/icons/find.png: Added.
186         * MiniBrowser/qt/qml/BrowserWindow.qml:
187
188 2013-01-02  Christophe Dumez  <christophe.dumez@intel.com>
189
190         Regression(r138603): Caused 2 webkitpy tests to fail
191         https://bugs.webkit.org/show_bug.cgi?id=105913
192
193         Reviewed by Kentaro Hara.
194
195         Fix 2 webkitpy tests that started to fail after
196         r138603.
197
198         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
199         (IRCCommandTest._sheriff_test_data_url): Make sure the
200         returned test file URL is valid no matter what the
201         current working directory is.
202         * Scripts/webkitpy/tool/bot/ircbot_unittest.py:
203         (IRCBotTest.test_help): Add "sheriffs" command to the
204         list of available commands.
205
206 2013-01-01  Christophe Dumez  <christophe.dumez@intel.com>
207
208         [WK2] plugins/clicking-missing-plugin-fires-delegate.html is failing
209         https://bugs.webkit.org/show_bug.cgi?id=105897
210
211         Reviewed by Alexey Proskuryakov.
212
213         Dump in WebKitTestRunner the message that is expected by layout tests
214         when the missing plugin button is clicked.
215
216         * WebKitTestRunner/TestController.cpp:
217         (WTR::TestController::createWebViewWithOptions):
218         (WTR::TestController::unavailablePluginButtonClicked):
219         (WTR):
220         * WebKitTestRunner/TestController.h:
221         (TestController):
222
223 2013-01-01  KwangYong Choi  <ky0.choi@samsung.com>
224
225         [EFL] Enable MHTML feature
226         https://bugs.webkit.org/show_bug.cgi?id=105815
227
228         Reviewed by Laszlo Gombos.
229
230         Enable MHTML feature for EFL port by default.
231
232         * Scripts/webkitperl/FeatureList.pm:
233
234 2013-01-01  Alan Cutter  <alancutter@chromium.org>
235
236         sheriff-bot should know who the gardeners/sheriffs are
237         https://bugs.webkit.org/show_bug.cgi?id=105698
238
239         Reviewed by Eric Seidel.
240
241         Added a "sheriffs" command to sheriffbot to fetch the current Chromium Webkit sheriffs.
242
243         * Scripts/webkitpy/tool/bot/irc_command.py:
244         (Current):
245         (Current._retrieve_webkit_sheriffs):
246         (Current.execute):
247         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
248         (IRCCommandTest.test_current):
249         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_0.js: Added.
250         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_1.js: Added.
251         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_2.js: Added.
252         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_malformed.js: Added.
253         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_zero.js: Added.
254
255 2012-12-31  Carlos Garcia Campos  <cgarcia@igalia.com>
256
257         [GTK] Move ImageDiff to the Tools directory root
258         https://bugs.webkit.org/show_bug.cgi?id=105421
259
260         Reviewed by Kenneth Rohde Christiansen.
261
262         * DumpRenderTree/gtk/GNUmakefile.ImageDiff.am: Removed.
263         * GNUmakefile.am:
264         * ImageDiff/gtk/ImageDiff.cpp: Renamed from Tools/DumpRenderTree/gtk/ImageDiff.cpp.
265
266 2012-12-30  Mike West  <mkwst@chromium.org>
267
268         Add myself to watchlists for CSP and console logs.
269         https://bugs.webkit.org/show_bug.cgi?id=105874
270
271         Reviewed by Dirk Pranke.
272
273         Because I really, really, really need more email.
274
275         * Scripts/webkitpy/common/config/committers.py:
276             Add mkwst+watchlist@chromium.org as a contributor.
277         * Scripts/webkitpy/common/config/watchlist:
278             Create watchlist categories for 1) the various mechanisms used to
279             generate console messages, and 2) CSP usage and files.
280
281 2012-12-30  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
282
283         [EFL] Enable TEMPLATE_ELEMENT feature
284         https://bugs.webkit.org/show_bug.cgi?id=105865
285         
286         Reviewed by Laszlo Gombos.
287
288         * Scripts/webkitperl/FeatureList.pm: Enable template element feature for EFL port.
289
290 2012-12-30  Zan Dobersek  <zandobersek@gmail.com>
291
292         Unreviewed GTK gardening of the unit tests.
293
294         Skip the TestInspectorServer unit test that's timing out completely.
295
296         * Scripts/run-gtk-tests:
297         (TestRunner):
298
299 2012-12-29  Ilya Tikhonovsky  <loislo@chromium.org>
300
301         Web Inspector: Native Memory Instrumentation plugin: do not generate 'not instrumented' warning for instrumented mutable members.
302         https://bugs.webkit.org/show_bug.cgi?id=105855
303
304         Reviewed by Vsevolod Vlasov.
305
306         Extract MemberExpr from ImplicitCastExpr. It happens when we instrument a mutable member because
307         addMember expects const T& and the mutable ref to member implicitly converts into const ref to member.
308
309         * clang/ReportMemoryUsagePlugin/ReportMemoryUsage.cpp:
310         (ReportMemoryUsageVisitor):
311         (clang::AddMemberCallVisitor::VisitCallExpr):
312         (clang::ReportMemoryUsageVisitor::VisitCXXMethodDecl):
313         (clang::ReportMemoryUsageVisitor::checkMembersCoverage):
314
315 2012-12-29  Ilya Tikhonovsky  <loislo@chromium.org>
316
317         Web Inspector: Native Memory Instrumentation plugin: move function bodies out of class declarations.
318         https://bugs.webkit.org/show_bug.cgi?id=105852
319
320         Reviewed by Alexander Pavlov.
321
322         * clang/ReportMemoryUsagePlugin/ReportMemoryUsage.cpp:
323         (clang):
324         (AddMemberCallVisitor):
325         (ReportMemoryUsageVisitor):
326         (clang::ReportMemoryUsageAction::ParseArgs):
327         (clang::AddMemberCallVisitor::VisitCallExpr):
328         (clang::ReportMemoryUsageVisitor::VisitCXXMethodDecl):
329         (clang::ReportMemoryUsageVisitor::emitWarning):
330         (clang::ReportMemoryUsageVisitor::findInstrumentationMethod):
331         (clang::ReportMemoryUsageVisitor::needsToBeInstrumented):
332         (clang::ReportMemoryUsageVisitor::CheckMembersCoverage):
333
334 2012-12-29  Zan Dobersek  <zandobersek@gmail.com>
335
336         [webkitpy] Omit webkitpy/thirdparty/BeautifulSoup.py from code coverage checking
337         https://bugs.webkit.org/show_bug.cgi?id=105835
338
339         Reviewed by Dirk Pranke.
340
341         The BeautifulSoup module is a third-party import and as such we don't have much control
342         over its testing. Because of that it should be omitted from code coverage checking.
343
344         * Scripts/webkitpy/test/main.py:
345         (Tester._run_tests):
346
347 2012-12-28  Martin Robinson  <mrobinson@igalia.com>
348
349         [GTK][WK2] Add support for IME Composition
350         https://bugs.webkit.org/show_bug.cgi?id=65093
351
352         Reviewed by Carlos Garcia Campos.
353
354         Update unit tests for GtkInputMethodFilter to note whether or not an event
355         was a fake event synthesized for composition purposes. This is now used
356         for WebKit2.
357
358         * TestWebKitAPI/Tests/gtk/InputMethodFilter.cpp:
359         (TestWebKitAPI::TestInputMethodFilter::sendSimpleKeyEvent):
360         (TestWebKitAPI::TestInputMethodFilter::sendKeyEventWithCompositionResults):
361         (TestWebKitAPI::TEST):
362         * gtk/generate-gtkdoc:
363         (get_webkit2_options): Skip new files in the WebKit2 source directory for documentation
364         generation.
365
366 2012-12-27  Ilya Tikhonovsky  <loislo@chromium.org>
367
368         Web Inspector: Native Memory Instrumentation. Update clang plugin according to the current state of memory instrumentation code.
369         https://bugs.webkit.org/show_bug.cgi?id=105800
370
371         Reviewed by Yury Semikhatsky.
372
373         * clang/ReportMemoryUsagePlugin/ReportMemoryUsage.cpp:
374         (clang::ReportMemoryUsageConsumer::ReportMemoryUsageConsumer):
375         (clang):
376
377 2012-12-27  Zan Dobersek  <zandobersek@gmail.com>
378
379         Create a GTK build system watchlist and add myself to it
380         https://bugs.webkit.org/show_bug.cgi?id=105769
381
382         Reviewed by Dirk Pranke.
383
384         Add a GTK build system watchlist, watching the configure.ac, GNUmakefile.am files
385         and the input GNUmakefile.features.am.in file. The GNUmakefile.list.am files are
386         omitted on purpose due to relatively high change rate they are exposed to.
387
388         The new watchlist would be of use to me as the GTK build system recently went
389         through a series of changes that included a changed policy on adding configuration
390         flags to configure.ac and subsequent additions to GNUmakefile.am files. A PSA on the
391         mailing list about this change will be published later, but I'd also like to be
392         notified of any pending changes to these files just so I can check the changes are
393         in compliance with the new policy.
394
395         * Scripts/webkitpy/common/config/watchlist:
396
397 2012-12-26  Ilya Tikhonovsky  <loislo@chromium.org>
398
399         Unreviewed. Replace find("Tools") with rfind("Tools") because base dir of WebKit may have Tools word.
400         As example ..../DevTools/src/third_party/WebKit
401
402         * Scripts/webkitpy/common/webkit_finder.py:
403         (WebKitFinder.webkit_base):
404
405 2012-12-25  Christophe Dumez  <christophe.dumez@intel.com>
406
407         [EFL][WK2] Refactor snapshot taking code
408         https://bugs.webkit.org/show_bug.cgi?id=105687
409
410         Reviewed by Kenneth Rohde Christiansen.
411
412         Call WKViewCreateSnapshot() instead of WKViewGetSnapshot()
413         since the function was renamed to follow naming style.
414
415         * WebKitTestRunner/efl/PlatformWebViewEfl.cpp:
416         (WTR::PlatformWebView::windowSnapshotImage):
417
418 2012-12-24  Ilya Tikhonovsky  <loislo@chromium.org>
419
420         Web Inspector: Native Memory Instrumentation: propagate member type as edge type to the serialized heap graph.
421         https://bugs.webkit.org/show_bug.cgi?id=105725
422
423         Reviewed by Yury Semikhatsky.
424
425         MemberType value names were adjusted according to Style Guide.
426         Existing tests were extended with link type validation.
427
428         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
429
430 2012-12-24  Xiaobo Wang  <xbwang@torchmobile.com.cn>
431
432         [BlackBerry] DRT - Crashed when running pixels tests
433         https://bugs.webkit.org/show_bug.cgi?id=105711
434
435         RIM PR 270003
436
437         Reviewed by George Staikos.
438
439         Calling window->post() here will result in buffer swapping which is
440         only allowed in UI thread now, and I found we don't need to call it
441         for GL renderer.
442
443         * DumpRenderTree/blackberry/PixelDumpSupportBlackBerry.cpp:
444         (createBitmapContextFromWebView):
445
446 2012-12-22  Dan Winship  <danw@gnome.org>
447
448         Add a patch to fix libxml2 2.9.0 regression
449         https://bugs.webkit.org/show_bug.cgi?id=104494
450
451         Reviewed by Martin Robinson.
452
453         * gtk/jhbuild.modules:
454         * gtk/patches/libxml2-2.9.0-dtd.patch: patch from upstream
455
456 2012-12-23  Allan Sandfeld Jensen  <allan.jensen@digia.com>
457
458         Unreviewed update of status to reviewer.
459
460         * Scripts/webkitpy/common/config/committers.py:
461
462 2012-12-23  Philippe Normand  <pnormand@igalia.com>
463
464         [GTK][jhbuild] Bump GStreamer to 1.0.4
465         https://bugs.webkit.org/show_bug.cgi?id=105640
466
467         Reviewed by Martin Robinson.
468
469         * gtk/jhbuild.modules: Update GStreamer modules to 1.0.4.
470
471 2012-12-21  Nico Weber  <thakis@chromium.org>
472
473         chromium: webkit-build-directory doesn't handle ninja output dirs
474         https://bugs.webkit.org/show_bug.cgi?id=105671
475
476         Reviewed by Daniel Bates.
477
478         Teach webkit-build-directory to detect the ninja build directory.
479
480         * Scripts/webkitdirs.pm:
481         (determineBaseProductDir):
482         (determineIsChromiumNinja):
483
484 2012-12-21  Ryosuke Niwa  <rniwa@webkit.org>
485
486         [Mountain Lion] platform/mac/editing/spelling/removing-underline-after-accepting-autocorrection-using-punctuation.html failing
487         https://bugs.webkit.org/show_bug.cgi?id=95477
488
489         Reviewed by Sam Weinig.
490
491         Co-authored by Jia Pu.
492
493         Specify a fixed autocorrection dictionary in resetDefaultsToConsistentValues() so that
494         we can have consistent autocorrection behavior when running tests.
495
496         * DumpRenderTree/mac/DumpRenderTree.mm:
497         (resetDefaultsToConsistentValues):
498         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
499         (WTR::InjectedBundle::platformInitialize):
500
501 2012-12-21  Zan Dobersek  <zandobersek@gmail.com>
502
503         [GTK] Add WebKit1-specific TestExpectations file
504         https://bugs.webkit.org/show_bug.cgi?id=97562
505
506         Reviewed by Dirk Pranke.
507
508         Implement the expectations_files method inside the GtkPort class that
509         returns the list of the TestExpectations files the current test configuration
510         requires. The method is implemented in a similar way as on the EFL port.
511         A unit test is also added.
512
513         * Scripts/webkitpy/layout_tests/port/gtk.py:
514         (GtkPort._search_paths):
515         (GtkPort):
516         (GtkPort.expectations_files):
517         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
518         (GtkPortTest.test_expectations_files):
519
520 2012-12-21  Sudarsana Nagineni  <sudarsana.nagineni@intel.com>
521
522         [EFL] Enable feature option for the CSS sticky position
523         https://bugs.webkit.org/show_bug.cgi?id=105630
524
525         Reviewed by Kenneth Rohde Christiansen.
526
527         Enable CSS sticky position feature by default for the
528         EFL port. This feature option was added in r138252.
529
530         No new tests. Existing layout tests pass.
531
532         * Scripts/webkitperl/FeatureList.pm:
533
534 2012-12-21  Jussi Kukkonen  <jussi.kukkonen@intel.com>
535
536         [WK2] REGRESSION(r138232): It made 4 tests timeout
537         https://bugs.webkit.org/show_bug.cgi?id=105623
538
539         Reviewed by Kenneth Rohde Christiansen.
540
541         r138232 changed functionality so that empty text output
542         no longer leads to a dump() call, which leads to a timeout.
543         Instead do a text dump if there is text output or if there
544         is no audio output.
545
546         * WebKitTestRunner/TestInvocation.cpp:
547         (WTR::TestInvocation::dumpResults):
548
549 2012-12-21  Zeno Albisser  <zeno@webkit.org>
550
551         [Qt][Mac] run-launcher cannot find Qt5WebKitWidgets.
552         https://bugs.webkit.org/show_bug.cgi?id=105309
553
554         On Windows the library name is Qt5WebKitWidgets as well.
555         So r138136 should not have adjusted the library name for Windows.
556
557         Reviewed by Simon Hausmann.
558
559         * Scripts/webkitdirs.pm:
560         (builtDylibPathForName):
561
562 2012-12-21  Zeno Albisser  <zeno@webkit.org>
563
564         [Qt] Add new TestExpectations for qt-5.0-mac-wk2.
565         https://bugs.webkit.org/show_bug.cgi?id=105424
566
567         This change will allow us to run a very specific
568         subset of the available LayoutTests on a mac buildbot.
569         The new baseline search path for mac will look as follows:
570         qt-5.0-mac-wk2 -> qt-5.0-wk2 -> qt-5.0 -> qt-mac -> qt
571
572         Reviewed by Dirk Pranke.
573
574         * Scripts/webkitpy/layout_tests/port/qt.py:
575         (QtPort._search_paths):
576         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
577         (QtPortTest):
578
579 2012-12-20  Ryuan Choi  <ryuan.choi@samsung.com>
580
581         [EFL] Build break with latest EFL libraries.
582         https://bugs.webkit.org/show_bug.cgi?id=104827
583
584         Reviewed by Laszlo Gombos.
585
586         The eo EFL package is introduced and evas and ecore use it since 1.8.
587         While introducing Eo, EFL changed several structures of Evas and Ecore
588         from own specific class to Eo.
589
590         So, this patch adds FindEo and changes declaration of Evas, Evas_Object,
591         Ecore_Timer to build with latest EFL libraries.
592
593         * DumpRenderTree/efl/CMakeLists.txt: Includes eo EFL library.
594         * EWebLauncher/CMakeLists.txt: Ditto.
595         * MiniBrowser/efl/CMakeLists.txt: Ditto.
596         * ImageDiff/PlatformEfl.cmake: Ditto.
597         * TestWebKitAPI/PlatformEfl.cmake: Ditto.
598         * TestWebKitAPI/PlatformWebView.h: Modified declaration of Evas_Object.
599         * WebKitTestRunner/InjectedBundle/TestRunner.h: Modified declaration of Ecore_Timer.
600         * WebKitTestRunner/PlatformEfl.cmake: Includes eo EFL library.
601         * WebKitTestRunner/PlatformWebView.h: Modified declaration of Evas_Object.
602
603 2012-12-20  Dirk Pranke  <dpranke@chromium.org>
604
605         TestExpectation fallback is broken
606         https://bugs.webkit.org/show_bug.cgi?id=105583
607
608         Reviewed by Ryosuke Niwa.
609
610         When I added support for the -wk2 in port names, it looks
611         like I broke the way we computed the list of TestExpectations
612         files so that the implementation-version directory was left out :(.
613         Fixed and cleaned up the tests.
614
615         * Scripts/webkitpy/layout_tests/port/base.py:
616         (Port.expectations_files):
617         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
618         (TestWebKitPort.__init__):
619         (PortTestCase.test_skipped_directories_for_symbols):
620         (test_skipped_directories_for_features):
621         (test_skipped_directories_for_features_no_matching_tests_in_test_list):
622         (test_skipped_tests_for_unsupported_features_empty_test_list):
623         (test_skipped_layout_tests):
624         (test_expectations_files):
625
626 2012-12-20  Anders Carlsson  <andersca@apple.com>
627
628         Don't include WebKit2 headers when building TestWebKitAPI on windows
629         https://bugs.webkit.org/show_bug.cgi?id=105582
630
631         Reviewed by Tim Horton.
632
633         * TestWebKitAPI/config.h:
634         * TestWebKitAPI/win/main.cpp:
635
636 2012-12-20  Ryosuke Niwa  <rniwa@webkit.org>
637
638         REGRESSION(r133820?): SimplifyMarkupTest API test asserts
639         https://bugs.webkit.org/show_bug.cgi?id=105370
640
641         Reviewed by Simon Fraser.
642
643         Re-enable the test.
644
645         * TestWebKitAPI/Tests/mac/SimplifyMarkup.mm:
646         (TestWebKitAPI::TEST):
647
648 2012-12-20  Nico Weber  <thakis@chromium.org>
649
650         chromium nrwt: Pick the newest binary found in DEFAULT_BUILD_DIRECTORIES, not the first
651         https://bugs.webkit.org/show_bug.cgi?id=105498
652
653         Reviewed by Dirk Pranke.
654
655         Use the newest binary available rather than an than always picking one
656         build directory over another based on iteration order.
657
658         * Scripts/webkitpy/layout_tests/port/chromium.py:
659         (ChromiumPort._static_build_path):
660         Check for timestamps.
661         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
662         (ChromiumMacPortTest.test_build_path_timestamps):
663         Test that out / xcodebuild selection happens based on timestamps
664         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
665         (ChromiumWinPortTest.test_build_path_timestamps):
666         Test that out / build selection happens based on timestamps
667
668 2012-12-19  Simon Fraser  <simon.fraser@apple.com>
669
670         Lots of sticky tests failing in WK2
671         https://bugs.webkit.org/show_bug.cgi?id=105464
672
673         Reviewed by Beth Dakin.
674
675         WebKitTestRunner had a race between snapshotting in the UI process,
676         and resettting after the test in the web process. InjectedBundle::done()
677         was a bad place to call page()->resetAfterTest(), because of this race;
678         it could reset the scroll position before the UI snapshot had been obtained.
679         
680         Fix by moving the call to page()->resetAfterTest() into didReceiveMessage(),
681         for the "Reset" message which will come in before the next test.
682
683         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
684         (WTR::InjectedBundle::didReceiveMessage):
685         (WTR::InjectedBundle::done):
686
687 2012-12-19  Filip Pizlo  <fpizlo@apple.com>
688
689         DFG speculation checks that take JumpList should consolidate OSRExits
690         https://bugs.webkit.org/show_bug.cgi?id=105401
691
692         Reviewed by Oliver Hunt.
693
694         Update the profiler to understand that an OSR exit may have multiple
695         candidate exit sites.
696
697         * Scripts/display-profiler-output:
698
699 2012-12-20  Eric Seidel  <eric@webkit.org>
700
701         EWSes do not need to dump build/test logs when things pass
702         https://bugs.webkit.org/show_bug.cgi?id=105402
703
704         Reviewed by Dirk Pranke.
705
706         When we were first writing the commit-queue and EWS system
707         and running them locally, we thought it was important to see the
708         build/test output as it happened.  Now these queues have been running
709         for years, and we don't use this feature anymore.  So this patch
710         makes us stop tee-ing the output directly to stdout, and instead
711         only print any output from commands when they fail.
712
713         I also added some logging when we execute each command
714         since we are no longer printing the command output itself.
715         This should make it much easier to see errors with the EWS
716         system itself, now that all the passing-build output is removed.
717
718         This required updating a zillion unittests, since we have a bunch of
719         "integration" style testing of the queue-bots.  Many of the unittests for the bots were
720         already approximating this behavior by using a "logging" Executive.
721         I disabled logging in those cases to avoid redundant logging in those tests.
722
723         * Scripts/webkitpy/common/system/executive.py:
724         (Executive.command_for_printing):
725         (Executive.run_command):
726         * Scripts/webkitpy/common/system/executive_mock.py:
727         (MockExecutive.command_for_printing):
728         (MockExecutive.run_command):
729         * Scripts/webkitpy/common/system/executive_unittest.py:
730         (ExecutiveTest.test_auto_stringify_args):
731         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
732         (EarlyWarningSytemTest._default_expected_logs):
733         (_test_ews):
734         (_test_ewses):
735         * Scripts/webkitpy/tool/commands/queues.py:
736         (AbstractQueue.run_webkit_patch):
737         * Scripts/webkitpy/tool/commands/queues_unittest.py:
738         (AbstractQueueTest._assert_run_webkit_patch):
739         (CommitQueueTest.test_commit_queue):
740         (test_rollout):
741         (test_rollout_lands):
742         (test_manual_reject_during_processing):
743         (StyleQueueTest.test_style_queue_with_style_exception):
744         (test_style_queue_with_watch_list_exception):
745
746 2012-12-20  Dominic Mazzoni  <dmazzoni@google.com>
747
748         AX: support clickPoint in DRT for chromium
749         https://bugs.webkit.org/show_bug.cgi?id=97359
750
751         Reviewed by Chris Fleizach.
752
753         Implement clickPointX, clickPointY, and elementAtPoint for Chromium, so that
754         several more tests can pass.
755
756         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
757         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
758         (WebTestRunner::AccessibilityUIElement::clickPointXGetterCallback):
759         (WebTestRunner):
760         (WebTestRunner::AccessibilityUIElement::clickPointYGetterCallback):
761         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
762         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
763         (AccessibilityUIElement):
764
765 2012-12-20  Dan Bernstein  <mitz@apple.com>
766
767         prepare-ChangeLog adds "Additional information of the change such as approach, rationale" where it doesn't make sense
768         https://bugs.webkit.org/show_bug.cgi?id=99268
769
770         Reviewed by Alexey Proskuryakov.
771
772         * Scripts/prepare-ChangeLog:
773         (generateNewChangeLogs): Removed the "Additional information of the change" line.
774
775 2012-12-20  Carlos Garcia Campos  <cgarcia@igalia.com>
776
777         Unreviewed. Skip unit test /webkit/atk/getTextInParagraphAndBodyModerate
778
779         * Scripts/run-gtk-tests:
780         (TestRunner):
781
782 2012-12-20  Thiago Marcos P. Santos  <thiago.santos@intel.com>
783
784         [Qt] Move ImageDiff to the Tools directory root
785         https://bugs.webkit.org/show_bug.cgi?id=105420
786
787         Reviewed by Tor Arne Vestbø.
788
789         Moved ImageDiff from outside DumpRenderTree because it is also a
790         dependency when running pixel tests for WebKit2.
791
792         * ImageDiff/ImageDiff.pro: Renamed from Tools/DumpRenderTree/qt/ImageDiff.pro.
793         * ImageDiff/qt/ImageDiff.cpp: Renamed from Tools/DumpRenderTree/qt/ImageDiff.cpp.
794         * Tools.pro:
795
796 2012-12-20  Zan Dobersek  <zandobersek@gmail.com>
797
798         [GTK] Remove the --enable-unstable-features configuration option
799         https://bugs.webkit.org/show_bug.cgi?id=105327
800
801         Reviewed by Martin Robinson.
802
803         Add a missing feature option for the CSS sticky position feature.
804
805         * Scripts/webkitperl/FeatureList.pm:
806
807 2012-12-20  Carlos Garcia Campos  <cgarcia@igalia.com>
808
809         Unreviewed. Skip unit test WebKit2.ForceRepaint that times out
810
811         * Scripts/run-gtk-tests:
812         (TestRunner):
813
814 2012-12-20  Dominik Röttsches  <dominik.rottsches@intel.com>
815
816         [EFL] MiniBrowser does not play Infinite Gangnam Style
817         https://bugs.webkit.org/show_bug.cgi?id=103531
818
819         Reviewed by Kenneth Rohde Christiansen.
820
821         For EFL we would like this feature to be built by default.
822
823         * Scripts/webkitperl/FeatureList.pm:
824
825 2012-12-20  Csaba Osztrogonác  <ossy@webkit.org>
826
827         [Qt] EWS bots should call qmake always
828         https://bugs.webkit.org/show_bug.cgi?id=90049
829
830         Reviewed by Simon Hausmann.
831
832         * Scripts/webkitdirs.pm:
833         (buildQMakeProjects):
834
835 2012-12-20  Jochen Eisinger  <jochen@chromium.org>
836
837         [chromium] fix stray printf in TestRunner's WebTestProxy
838         https://bugs.webkit.org/show_bug.cgi?id=105515
839
840         Reviewed by Ryosuke Niwa.
841
842         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
843         (WebTestRunner::WebTestProxyBase::shouldInsertNode):
844
845 2012-12-20  Dominik Röttsches  <dominik.rottsches@intel.com>
846
847         [WK2] WebAudio WKTR support
848         https://bugs.webkit.org/show_bug.cgi?id=95084
849
850         Reviewed by Kenneth Rohde Christiansen.
851
852         Implementing audio data dump support for WebKitTestRunner,
853         by adding the API to the IDL, enabling IPC for transferring
854         the audio data as bytebuffer and dumping it according
855         to run-webkit-tests script's expectations.
856         This enables audio testing on WebKit2 ports.
857         Based on initial wip-patch from Philippe Normand - thanks.
858
859         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl: Adding setAudioData API.
860         * WebKitTestRunner/InjectedBundle/TestRunner.cpp: Assign audio buffer data to a local WKDataRef after extracting it from WK2 using WKBundleDataFromUint8Array bundle API.
861         (WTR::TestRunner::setAudioData):
862         (WTR):
863         * WebKitTestRunner/InjectedBundle/TestRunner.h: Add a whatToDump state for dumping Audio.
864         (TestRunner):
865         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp: Add audio to the message that's transferred to the WKTR process.
866         (WTR::InjectedBundle::done):
867         * WebKitTestRunner/InjectedBundle/InjectedBundle.h: Allow setting and audio result handle to be transferred.
868         (WTR::InjectedBundle::setAudioResult):
869         (InjectedBundle):
870         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
871         (WTR::InjectedBundlePage::dump): Dummy handling the Audio case in the switch, build fix for Mac.
872         * WebKitTestRunner/TestInvocation.cpp: Audio dump support, dump audio if we haven't received a text dump from the web process.
873         (WTR::TestInvocation::dumpResults):
874         (WTR):
875         (WTR::TestInvocation::dumpAudio):
876         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle): Storing received audio data before dumping.
877         * WebKitTestRunner/TestInvocation.h:
878         (TestInvocation):
879
880 2012-12-20  Alberto Garcia  <agarcia@igalia.com>
881
882         [GTK] Count failed and timed out tests separately
883         https://bugs.webkit.org/show_bug.cgi?id=105448
884
885         Reviewed by Carlos Garcia Campos.
886
887         Display the number of tests that fail, time out and are skipped
888         separately in the buildbot page.
889
890         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
891         (RunGtkAPITests.commandComplete):
892         (RunGtkAPITests.evaluateCommand):
893         (RunGtkAPITests.getText2):
894         * Scripts/run-gtk-tests:
895         (TestRunner.run_tests):
896
897 2012-12-19  Ryosuke Niwa  <rniwa@webkit.org>
898
899         PerfTest.parse_output does too much
900         https://bugs.webkit.org/show_bug.cgi?id=105391
901
902         Reviewed by Tony Chang.
903
904         Moved the code to filter lines into PerfTest._filter_output, which has been renamed from _filter_stderr.
905         Also moved the code to output test results into PerfTest._run_with_driver.
906
907         * Scripts/webkitpy/performance_tests/perftest.py:
908         (PerfTest.__init__): Added _description.
909         (PerfTest.description): Added.
910         (PerfTest._run_with_driver): Moved the code to output test description and test results.
911         (PerfTest._filter_output): Moved and renamed from PerfTest._filter_stderr.
912         (PerfTest.parse_output): Removed the code to output test results. Return test results and description
913         respectively. Also removed the code that allowed some tests to omit values since all tests report each
914         iteration now since r136492.
915         (PerfTest.output_statistics): Removed the code to print test description, now done in _run_with_driver.
916         (ChromiumStylePerfTest._run_with_driver): Added. Chromium style tests are sufficiently different from
917         regular PerfTest that it doesn't make much sense to share _run_with_driver. But really, we should just
918         get rid of this type of test altogether in favor of regular performance tests that uses runner.js.
919         (ChromiumStylePerfTest.parse_and_log_output): Renamed from parse_output.
920         (PageLoadingPerfTest._run_with_driver): Removed the explicit '' for the test description.
921         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
922         (MainTest.test_parse_output): Removed the expected logs since parse_output no longer prints out results.
923         Also added a call to _filter_output since parse_output doesn't filter the output text anymore.
924         (MainTest.test_parse_output_with_failing_line): Added a call to _filter_output. Also added ',' after 'Time:'
925         so that it's not string-concatenated with the next line.
926         (MainTest.test_parse_output_with_description): Added; a test for PerfTest.description().
927         (MainTest.test_parse_output_with_subtests): Removed the expected logs, and added a call to _filter_output.
928
929 2012-12-19  Alexis Menard  <alexis@webkit.org>
930
931         Implement CSS parsing for CSS transitions unprefixed.
932         https://bugs.webkit.org/show_bug.cgi?id=104804
933
934         Reviewed by Dean Jackson.
935
936         Add a new flag ENABLE_CSS_TRANSFORMS_ANIMATIONS_TRANSITIONS_UNPREFIXED
937         to cover the work of unprefixing Transforms, Animations and 
938         Transitions. It will let the possibility of each ports to turn it off 
939         in their release branches until we're confident that these CSS 
940         properties are ready to be unprefixed.
941
942         * Scripts/webkitperl/FeatureList.pm:
943         * qmake/mkspecs/features/features.pri:
944
945 2012-12-19  Julien Chaffraix  <jchaffraix@webkit.org>
946
947         2 clampTo unit tests are failing
948         https://bugs.webkit.org/show_bug.cgi?id=105253
949
950         Reviewed by Emil A Eklund.
951
952         Removed clampToUnsigned as it was actually testing exactly what
953         clampToUnsignedLong was doing. Also made clampToIntFloat more robust
954         to precision loss (fix suggested by Emil).
955
956         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
957
958 2012-12-19  Thiago Marcos P. Santos  <thiago.santos@intel.com>
959
960         [EFL] Move ImageDiff to the Tools directory root
961         https://bugs.webkit.org/show_bug.cgi?id=105305
962
963         Reviewed by Laszlo Gombos.
964
965         We previously moved this directory from the DRT directory to
966         efl/ImageDiff. The major reason for doing so was because ImageDiff does
967         not depend on DumpRenderTree but for legacy reasons was living inside the
968         DRT folder, even though it is needed when running pixel tests for WK2.
969
970         Now we are moving to Tool directory root as other ports agreed to do the same.
971
972         * CMakeLists.txt:
973         * ImageDiff/CMakeLists.txt: Added.
974         * ImageDiff/PlatformEfl.cmake: Added.
975         * ImageDiff/efl/ImageDiff.cpp: Renamed from Tools/efl/ImageDiff/ImageDiff.cpp.
976         * efl/ImageDiff/CMakeLists.txt: Removed.
977
978 2012-12-19  Zeno Albisser  <zeno@webkit.org>
979
980         [Qt][Mac] run-launcher cannot find Qt5WebKitWidgets.
981         https://bugs.webkit.org/show_bug.cgi?id=105309
982
983         The version number is only used as a part
984         of the library name on Linux.
985
986         Reviewed by Kenneth Rohde Christiansen.
987
988         * Scripts/webkitdirs.pm:
989         (builtDylibPathForName):
990
991 2012-12-19  Alberto Garcia  <agarcia@igalia.com>
992
993         [GTK] Tests that time out are not considered as failures.
994         https://bugs.webkit.org/show_bug.cgi?id=105318
995
996         Reviewed by Martin Robinson.
997
998         Fix parsing of the GTK API tests log files.
999
1000         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
1001         (RunGtkAPITests.commandComplete):
1002
1003 2012-12-18  Ilya Tikhonovsky  <loislo@chromium.org>
1004
1005         Unreviewed, rolling out r138061.
1006         https://bugs.webkit.org/show_bug.cgi?id=105396
1007
1008         Original patch:
1009         http://trac.webkit.org/changeset/138061
1010         https://bugs.webkit.org/show_bug.cgi?id=97359
1011
1012         It is crashing on Debug bots
1013
1014         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
1015         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
1016         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
1017         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
1018         (AccessibilityUIElement):
1019
1020 2012-12-18  Alice Liu  <alice.liu@apple.com>
1021
1022         Test for https://bugs.webkit.org/show_bug.cgi?id=105106
1023         Add SPI to WebKit1 WebFrame for hit testing
1024
1025         Reviewed by Dan Bernstein.
1026
1027         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added file Tests/mac/ElementAtPointInWebFrame.mm
1028         * TestWebKitAPI/Tests/mac/ElementAtPointInWebFrame.mm: Added.
1029         (TestWebKitAPI::TEST): Loads html with two divs positioned in the 2nd and 4th quadrants of the webview.
1030         Then hit-tests at three points, expecting to hit the two divs and body element. 
1031
1032 2012-12-18  Alice Liu  <alice.liu@apple.com>
1033
1034         Test for https://bugs.webkit.org/show_bug.cgi?id=105262
1035         Add SPI to WebKit1 WebFrame for node conversion to JSValueRef
1036
1037         Reviewed by Anders Carlsson.
1038
1039         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added file Tests/mac/JSWrapperForNodeInWebFrame.mm.mm
1040         * TestWebKitAPI/Tests/mac/JSWrapperForNodeInWebFrame.mm.mm: Added.
1041         (TestWebKitAPI::TEST): Tests for the correct JS wrapper for a DOMNode, provided a WebScriptWorld.
1042         Loads html with a single div element. In an isolated world, creates a property on that node.
1043         Also, in the standard world, creates a different property on that node.  Then tests for 4 things:
1044             - Existence of the isolated property in the isolated world.
1045             - Existence of the standard property in the standard world.
1046             - Non-existence of the isolated property in the standard world.
1047             - Non-existence of the standard property in the isolated world.
1048
1049 2012-12-18  Simon Fraser  <simon.fraser@apple.com>
1050
1051         Disable the SimplifyMarkup API test, wich asserts in debug builds.
1052         Tracked by https://bugs.webkit.org/show_bug.cgi?id=105370.
1053         
1054         * TestWebKitAPI/Tests/mac/SimplifyMarkup.mm:
1055         (TestWebKitAPI::TEST):
1056
1057 2012-12-18  Joseph Pecoraro  <pecoraro@apple.com>
1058
1059         [Mac] Add Build Phase to Check Headers for Inappropriate Macros (Platform.h macros)
1060         https://bugs.webkit.org/show_bug.cgi?id=104279
1061
1062         Reviewed by David Kilzer.
1063
1064         Add a build phase script that checks a framework's header files for
1065         Platform.h and Compiler.h macros. Also add some tests for the script.
1066
1067         * Scripts/check-for-inappropriate-macros-in-external-headers: Added.
1068         Script to be used in build phases to check for inappropriate macros in headers.
1069
1070         * Scripts/test-webkitruby: Added.
1071         Script that can be used to run all webkitruby tests.
1072
1073         * Scripts/test-webkit-scripts:
1074         Also run the ruby tests.
1075
1076         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/fake-data-failing-expected.txt: Added.
1077         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/pass-expected.txt: Added.
1078         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/resources/Fake.framework/Headers/Fail.h: Added.
1079         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/resources/Fake.framework/Headers/Pass.h: Added.
1080         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/run-test.rb: Added.
1081         Tests for the new script.
1082
1083 2012-12-18  Dominic Mazzoni  <dmazzoni@google.com>
1084
1085         AX: support clickPoint in DRT for chromium
1086         https://bugs.webkit.org/show_bug.cgi?id=97359
1087
1088         Reviewed by Chris Fleizach.
1089
1090         Implement clickPointX, clickPointY, and elementAtPoint for Chromium, so that
1091         several more tests can pass.
1092
1093         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
1094         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
1095         (WebTestRunner::AccessibilityUIElement::clickPointXGetterCallback):
1096         (WebTestRunner):
1097         (WebTestRunner::AccessibilityUIElement::clickPointYGetterCallback):
1098         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
1099         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
1100         (AccessibilityUIElement):
1101
1102 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
1103
1104         Build fix after r138045.
1105
1106         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1107         (PerfTestsRunner._collect_tests): Use relative_path. We skip only when this path
1108         is not explicitly specified. The condition was backwards but never matched so we
1109         always ran the test regardless of whether it's listed in the skipped list or not.
1110         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
1111         (test_collect_tests_with_ignored_skipped_list): Renamed so that it actually runs.
1112
1113 2012-12-18  Emil A Eklund  <eae@chromium.org>
1114
1115         TestWebKitAPI/Tests/WebCore/LayoutUnit fails on mac
1116         https://bugs.webkit.org/show_bug.cgi?id=105332
1117         
1118         Reviewed by Levi Weintraub.
1119         
1120         Create new test suite for LayoutUnit tests as the KURL tests are
1121         of the incompatible TEST_F type.
1122
1123         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp:
1124         (TestWebKitAPI):
1125         (TestWebKitAPI::TEST_F):
1126
1127 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
1128
1129         Running a skipped test with run-perf-tests could alert the user
1130         https://bugs.webkit.org/show_bug.cgi?id=105219
1131
1132         Reviewed by Eric Seidel.
1133
1134         Force run skipped performance tests explicitly specified in the command line.
1135         Also warn the user when the specified paths don't exist.
1136
1137         * Scripts/webkitpy/layout_tests/port/base.py:
1138         (Port.relative_test_filename):
1139         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1140         (PerfTestsRunner._parse_args):
1141         (PerfTestsRunner._collect_tests):
1142         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
1143         (test_collect_tests_with_skipped_list_and_files):
1144
1145 2012-12-18  Yaron Friedman  <yfriedman@chromium.org>
1146
1147         Chrome Android java output is too verbose
1148         https://bugs.webkit.org/show_bug.cgi?id=105247
1149
1150         Reviewed by Tony Chang.
1151
1152         Pass -quiet to ant to suppress extraneous output.
1153
1154         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
1155         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
1156
1157 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
1158
1159         Consider removing --pause-before-testing option
1160         https://bugs.webkit.org/show_bug.cgi?id=105248
1161
1162         Reviewed by Eric Seidel.
1163
1164         Remove --pause-before-testing option.
1165
1166         * Tools/Scripts/webkitpy/layout_tests/port/driver.py:
1167         (DriverProxy.start): Removed.
1168         * Scripts/webkitpy/performance_tests/perftest.py:
1169         (PerfTest.__init__):
1170         (PerfTest.test_path): Renamed from path_or_url.
1171         (PerfTest.run):
1172         (PerfTest): Removed "driver" argument.
1173         (PerfTest._run_with_driver):
1174         (PerfTest.run_single):
1175         (ChromiumStylePerfTest.__init__):
1176         (PageLoadingPerfTest.__init__):
1177         (PageLoadingPerfTest.run_single):
1178         (PageLoadingPerfTest._run_with_driver): Renamed from run so that PerfTest.run can create a driver for us.
1179         (ReplayPerfTest.__init__):
1180         (ReplayPerfTest.prepare):
1181         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
1182         (TestPageLoadingPerfTest.test_run):
1183         (TestPageLoadingPerfTest.test_run_with_memory_output):
1184         (TestPageLoadingPerfTest.test_run_with_bad_output):
1185         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1186         (PerfTestsRunner._parse_args):
1187         (PerfTestsRunner._run_tests_set): Removed the code to create a driver and pause before running the test.
1188         (PerfTestsRunner._run_single_test): Removed "driver" argument.
1189         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
1190         (run_test):
1191         (test_run_test_pause_before_testing): Removed.
1192
1193 2012-12-18  Alberto Garcia  <agarcia@igalia.com>
1194
1195         [GTK] Tests that time out are not considered as failures.
1196         https://bugs.webkit.org/show_bug.cgi?id=105318
1197
1198         Reviewed by Martin Robinson.
1199
1200         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
1201         (RunGtkAPITests.commandComplete):
1202         Consider log lines that end with TIMEOUT as errors.
1203
1204         * Scripts/run-gtk-tests:
1205         (TestRunner.run_tests):
1206         Consider timeouts as failures when returning the number of failed
1207         tests.
1208
1209 2012-12-18  Pravin D  <pravin.d@samsung.com>
1210
1211         Unreviewed, adding secondary email.
1212
1213         * Scripts/webkitpy/common/config/committers.py:
1214
1215 2012-12-18  Pravin D  <pravin.d@samsung.com>
1216
1217         Unreviewed, promoting myself to committer's list.
1218
1219         * Scripts/webkitpy/common/config/committers.py:
1220
1221 2012-12-18  Thiago Marcos P. Santos  <thiago.santos@intel.com>
1222
1223         [EFL] ImageDiff does not build on WebKit2 only builds
1224         https://bugs.webkit.org/show_bug.cgi?id=105288
1225
1226         Reviewed by Kenneth Rohde Christiansen.
1227
1228         Moved ImageDiff to the efl/ folder inside Tools/ (so we don't pollute
1229         the folder with things which are EFL-specific.
1230
1231         ImageDiff will be built by default now regardless of WebKit API flavor
1232         you choose.
1233
1234         * CMakeLists.txt:
1235         * DumpRenderTree/efl/CMakeLists.txt:
1236         * efl/ImageDiff/CMakeLists.txt: Added.
1237         * efl/ImageDiff/ImageDiff.cpp: Renamed from Tools/DumpRenderTree/efl/ImageDiff.cpp.
1238
1239 2012-12-18  Kenneth Rohde Christiansen  <kenneth@webkit.org>
1240
1241         [EFL][WK2] When creating new windows, use the opener's context
1242         https://bugs.webkit.org/show_bug.cgi?id=105283
1243
1244         Reviewed by Laszlo Gombos.
1245
1246         Use the context of the opener if exists.
1247
1248         * MiniBrowser/efl/main.c:
1249         (on_key_down):
1250         (on_window_create):
1251         (window_create):
1252         (elm_main):
1253
1254 2012-12-18  Thiago Marcos P. Santos  <thiago.santos@intel.com>
1255
1256         [EFL][WK2] Gardening of TestWebKitAPI tests
1257         https://bugs.webkit.org/show_bug.cgi?id=105268
1258
1259         Reviewed by Kenneth Rohde Christiansen.
1260
1261         Some tests were marked as failing but are passing and the newly added
1262         LayoutUnit should have it's own binary because it conflicts with the
1263         fixture used by KURL and is failing on the bots because of that.
1264
1265         * TestWebKitAPI/CMakeLists.txt:
1266         * TestWebKitAPI/PlatformEfl.cmake:
1267
1268 2012-12-18  Jochen Eisinger  <jochen@chromium.org>
1269
1270         [chromium] move webintent related methods to TestRunner library
1271         https://bugs.webkit.org/show_bug.cgi?id=105164
1272
1273         Reviewed by Darin Fisher.
1274
1275         Also deleted the unused deliveredIntentFailure and
1276         deliveredIntentResult methods.
1277
1278         * DumpRenderTree/chromium/DRTTestRunner.cpp:
1279         (DRTTestRunner::DRTTestRunner):
1280         * DumpRenderTree/chromium/DRTTestRunner.h:
1281         (DRTTestRunner):
1282         * DumpRenderTree/chromium/TestRunner/public/WebTestDelegate.h:
1283         (WebKit):
1284         (WebTestRunner::WebTestDelegate::setCurrentWebIntentRequest):
1285         (WebTestRunner::WebTestDelegate::currentWebIntentRequest):
1286         * DumpRenderTree/chromium/TestRunner/public/WebTestProxy.h:
1287         (WebKit):
1288         (WebTestProxyBase):
1289         (WebTestRunner::WebTestProxy::registerIntentService):
1290         (WebTestRunner::WebTestProxy::dispatchIntent):
1291         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
1292         (WebTestRunner::TestRunner::TestRunner):
1293         (WebTestRunner::TestRunner::sendWebIntentResponse):
1294         (WebTestRunner):
1295         (WebTestRunner::TestRunner::deliverWebIntent):
1296         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
1297         (TestRunner):
1298         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
1299         (WebTestRunner::WebTestProxyBase::postAccessibilityNotification):
1300         (WebTestRunner::WebTestProxyBase::registerIntentService):
1301         (WebTestRunner):
1302         (WebTestRunner::WebTestProxyBase::dispatchIntent):
1303         * DumpRenderTree/chromium/WebViewHost.cpp:
1304         (WebViewHost::setCurrentWebIntentRequest):
1305         (WebViewHost::currentWebIntentRequest):
1306         * DumpRenderTree/chromium/WebViewHost.h:
1307         (WebKit):
1308         (WebViewHost):
1309
1310 2012-12-18  Jochen Eisinger  <jochen@chromium.org>
1311
1312         [chromium] move editing callback dumping to the WebTestProxy
1313         https://bugs.webkit.org/show_bug.cgi?id=105013
1314
1315         Reviewed by Darin Fisher.
1316
1317         To decide whether or not to dump the editing callbacks, the WebTestProxy
1318         needs to query the TestRunner. For this purpose, I introduce a
1319         WebTestRunner class that just exposes the shouldDumpEditingCallbacks
1320         method. In contrast to e.g. WebEventSender, the WebTestRunner is just
1321         an interface, and it is not owned by the WebTestInterfaces, but the
1322         embedder of the TestRunner library must provide it. This will change
1323         once all TestRunner methods are in the library.
1324
1325         * DumpRenderTree/DumpRenderTree.gypi:
1326         * DumpRenderTree/chromium/DRTTestRunner.cpp:
1327         (DRTTestRunner::DRTTestRunner):
1328         (DRTTestRunner::reset):
1329         * DumpRenderTree/chromium/DRTTestRunner.h:
1330         (DRTTestRunner):
1331         (DRTTestRunner::taskList):
1332         (DRTTestRunner::WorkQueue::taskList):
1333         * DumpRenderTree/chromium/TestRunner/public/WebTestInterfaces.h:
1334         (WebTestRunner):
1335         (WebTestInterfaces):
1336         * DumpRenderTree/chromium/TestRunner/public/WebTestProxy.h:
1337         (WebKit):
1338         (WebTestRunner):
1339         (WebTestProxyBase):
1340         (WebTestRunner::WebTestProxy::shouldBeginEditing):
1341         (WebTestRunner::WebTestProxy::shouldEndEditing):
1342         (WebTestRunner::WebTestProxy::shouldInsertNode):
1343         (WebTestRunner::WebTestProxy::shouldInsertText):
1344         (WebTestRunner::WebTestProxy::shouldChangeSelectedRange):
1345         (WebTestRunner::WebTestProxy::shouldDeleteRange):
1346         (WebTestRunner::WebTestProxy::shouldApplyStyle):
1347         (WebTestRunner::WebTestProxy::didBeginEditing):
1348         (WebTestRunner::WebTestProxy::didChangeSelection):
1349         (WebTestRunner::WebTestProxy::didChangeContents):
1350         (WebTestRunner::WebTestProxy::didEndEditing):
1351         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h: Copied from Tools/DumpRenderTree/chromium/TestRunner/public/WebTestInterfaces.h.
1352         (WebTestRunner):
1353         (WebTestRunner::WebTestRunner::shouldDumpEditingCallbacks):
1354         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
1355         (WebTestRunner::TestRunner::TestRunner):
1356         (WebTestRunner::TestRunner::~TestRunner):
1357         (WebTestRunner):
1358         (WebTestRunner::TestRunner::reset):
1359         (WebTestRunner::TestRunner::shouldDumpEditingCallbacks):
1360         (WebTestRunner::TestRunner::dumpEditingCallbacks):
1361         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
1362         (TestRunner):
1363         * DumpRenderTree/chromium/TestRunner/src/WebTestInterfaces.cpp:
1364         (WebTestRunner::WebTestInterfaces::Internal::testRunner):
1365         (WebTestRunner::WebTestInterfaces::Internal::setTestRunner):
1366         (WebTestInterfaces::Internal):
1367         (WebTestRunner::WebTestInterfaces::Internal::Internal):
1368         (WebTestRunner::WebTestInterfaces::testRunner):
1369         (WebTestRunner):
1370         (WebTestRunner::WebTestInterfaces::setTestRunner):
1371         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
1372         (WebTestRunner::WebTestProxyBase::postAccessibilityNotification):
1373         (WebTestRunner):
1374         (WebTestRunner::WebTestProxyBase::shouldBeginEditing):
1375         (WebTestRunner::WebTestProxyBase::shouldEndEditing):
1376         (WebTestRunner::WebTestProxyBase::shouldInsertNode):
1377         (WebTestRunner::WebTestProxyBase::shouldInsertText):
1378         (WebTestRunner::WebTestProxyBase::shouldChangeSelectedRange):
1379         (WebTestRunner::WebTestProxyBase::shouldDeleteRange):
1380         (WebTestRunner::WebTestProxyBase::shouldApplyStyle):
1381         (WebTestRunner::WebTestProxyBase::didBeginEditing):
1382         (WebTestRunner::WebTestProxyBase::didChangeSelection):
1383         (WebTestRunner::WebTestProxyBase::didChangeContents):
1384         (WebTestRunner::WebTestProxyBase::didEndEditing):
1385         * DumpRenderTree/chromium/TestShell.cpp:
1386         (TestShell::initialize):
1387         (TestShell::createNewWindow):
1388         * DumpRenderTree/chromium/WebViewHost.cpp:
1389         (WebViewHost::shouldBeginEditing):
1390         (WebViewHost::shouldEndEditing):
1391         (WebViewHost::shouldInsertNode):
1392         (WebViewHost::shouldInsertText):
1393         (WebViewHost::shouldChangeSelectedRange):
1394         (WebViewHost::shouldDeleteRange):
1395         (WebViewHost::shouldApplyStyle):
1396         * DumpRenderTree/chromium/WebViewHost.h:
1397         (WebViewHost):
1398
1399 2012-12-17  Jochen Eisinger  <jochen@chromium.org>
1400
1401         [chromium] Fix methods in TestRunner that directly modify the settings instead of going through WebPreferences
1402         https://bugs.webkit.org/show_bug.cgi?id=105154
1403
1404         Reviewed by Tony Chang.
1405
1406         If the settings are modified directly, the embedder can't keep track of
1407         them and might accidentially override the changed settings.
1408
1409         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
1410         (WebPreferences):
1411         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
1412         (WebTestRunner::TestRunner::TestRunner):
1413         (WebTestRunner::TestRunner::setAsynchronousSpellCheckingEnabled):
1414         (WebTestRunner):
1415         (WebTestRunner::TestRunner::setMinimumTimerInterval):
1416         (WebTestRunner::TestRunner::setTouchDragDropEnabled):
1417         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
1418         (TestRunner):
1419         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
1420         (WebTestRunner::WebPreferences::reset):
1421         (WebTestRunner::WebPreferences::applyTo):
1422
1423 2012-12-17  Julien Chaffraix  <jchaffraix@webkit.org>
1424
1425         Unreviewed unit test skipping.
1426
1427         Filed https://bugs.webkit.org/show_bug.cgi?id=105253 to re-enable them.
1428
1429         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
1430         (TestWebKitAPI):
1431         (TestWebKitAPI::TEST):
1432
1433 2012-12-17  Dean Jackson  <dino@apple.com>
1434
1435         Unreviewed. Add Silvia Pfeiffer to contributor list.
1436
1437         * Scripts/webkitpy/common/config/committers.py:
1438
1439 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
1440
1441         build.webkit.org - add more google windows bots
1442         https://bugs.webkit.org/show_bug.cgi?id=105225
1443
1444         Reviewed by Eric Seidel.
1445
1446         Adds two more windows bots (google-windows-3, a builder, and
1447         google-windows-4, a tester) for the Chromium configs.
1448
1449         * BuildSlaveSupport/build.webkit.org-config/config.json:
1450
1451 2012-12-17  Eric Seidel  <eric@webkit.org>
1452
1453         Cleanup --profiler=perf code a little and allow up to 1000 report files (instead of 100)
1454         https://bugs.webkit.org/show_bug.cgi?id=105217
1455
1456         Reviewed by Dirk Pranke.
1457
1458         Previously --profile would fail strangely if find_unused_filename returned None
1459         (which would happen as soon as we had 100 report files).
1460         Now we'll search through 1000 filenames before giving up (and assert if we failed to find one).
1461         I also made some minor cleanups to the output of perf, and tried to make the
1462         commands more self-documenting by using long-names for args.
1463
1464         * Scripts/webkitpy/common/system/profiler.py:
1465         (SingleFileOutputProfiler.__init__):
1466         (Perf.attach_to_pid):
1467         (Perf.profile_after_exit):
1468
1469 2012-12-17  Sheriff Bot  <webkit.review.bot@gmail.com>
1470
1471         Unreviewed, rolling out r137198.
1472         http://trac.webkit.org/changeset/137198
1473         https://bugs.webkit.org/show_bug.cgi?id=105212
1474
1475         This patch is causing API behavior compatibility problems
1476         (Requested by zdobersek on #webkit).
1477
1478         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1479         (willSendRequestCallback):
1480
1481 2012-12-17  Elliott Sprehn  <esprehn@chromium.org>
1482
1483         Unreviewed. Adding myself to committers list.
1484
1485         * Scripts/webkitpy/common/config/committers.py:
1486
1487 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
1488
1489         nrwt: shuffle code around for cleanup in run_webkit_tests.py
1490         https://bugs.webkit.org/show_bug.cgi?id=105078
1491
1492         Reviewed by Ojan Vafai.
1493
1494         This patch reorders functions so that run_webkit_tests can be
1495         understood in a top-down matter better, and to make things slightly
1496         more sensible when I merge manager.py into it.
1497
1498         Also, this adds tests for the actual main() routine, which had several
1499         bugs get through testing a week or two ago.
1500
1501         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1502         (main):
1503         (parse_args):
1504         (_set_up_derived_options):
1505         (run):
1506         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1507         (RunTest):
1508         (RunTest.test_no_http_tests):
1509         (PortTest.disabled_test_mac_lion):
1510         (MainTest):
1511         (MainTest.test_exception_handling):
1512         (MainTest.test_exception_handling.interrupting_run):
1513         (MainTest.test_exception_handling.successful_run):
1514         (MainTest.test_exception_handling.successful_run.FakeRunDetails):
1515         (MainTest.test_exception_handling.exception_raising_run):
1516
1517 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
1518
1519         webkitpy: move --lint-test-files code into its own module
1520         https://bugs.webkit.org/show_bug.cgi?id=105077
1521
1522         Reviewed by Eric Seidel.
1523
1524         It really didn't need to be in run_webkit_tests.py anymore.
1525         This patch adds a new lint-test-expectations scripts to Tools/Scripts;
1526         rwt --lint-test-files is still supported for backwards compatibility
1527         for now.
1528
1529         * Scripts/lint-test-expectations: Added.
1530         * Scripts/webkitpy/layout_tests/lint_test_expectations.py: Added.
1531         (lint):
1532         (main):
1533         * Scripts/webkitpy/layout_tests/lint_test_expectations_unittest.py: Added.
1534         (FakePort):
1535         (FakePort.__init__):
1536         (FakePort.test_configuration):
1537         (FakePort.expectations_dict):
1538         (FakePort.skipped_layout_tests):
1539         (FakePort.all_test_configurations):
1540         (FakePort.configuration_specifier_macros):
1541         (FakePort.get_option):
1542         (FakeFactory):
1543         (FakeFactory.__init__):
1544         (FakeFactory.get):
1545         (FakeFactory.all_port_names):
1546         (LintTest):
1547         (LintTest.test_all_configurations):
1548         (LintTest.test_lint_test_files):
1549         (LintTest.test_lint_test_files__errors):
1550         (MainTest):
1551         (MainTest.test_success):
1552         (MainTest.test_success.interrupting_lint):
1553         (MainTest.test_success.successful_lint):
1554         (MainTest.test_success.exception_raising_lint):
1555         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1556         (main):
1557         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1558         (StreamTestingMixin.assertNotEmpty):
1559
1560 2012-12-17  Adam Klein  <adamk@chromium.org>
1561
1562         build-webkit: rename --template-tag to --template-element to match ENABLE #define name
1563         https://bugs.webkit.org/show_bug.cgi?id=105072
1564
1565         Reviewed by Laszlo Gombos.
1566
1567         * Scripts/webkitperl/FeatureList.pm:
1568
1569 2012-12-17  Julien Chaffraix  <jchaffraix@webkit.org>
1570
1571         Add some unit testing for WTF::clampTo* functions
1572         https://bugs.webkit.org/show_bug.cgi?id=105060
1573
1574         Reviewed by Emil A Eklund.
1575
1576         The new tests are targeted at clamping to int / unsigned and float as those are
1577         the one explicitely exposed through more dedicated functions (like clampToInteger,
1578         clampToPositiveInteger and clampToFloat).
1579
1580         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
1581
1582 2012-12-17  Eric Seidel  <eric@webkit.org>
1583
1584         rpt --profile --chromium-android throws exception
1585         https://bugs.webkit.org/show_bug.cgi?id=105061
1586
1587         Reviewed by Daniel Bates.
1588
1589         Somehow popen(args) got turned into popen(*args) at some point,
1590         which caused my new stringify code to fail when args was passed by name.
1591         Fixed and tested.
1592
1593         * Scripts/webkitpy/common/system/executive.py:
1594         (Executive._stringify_args):
1595         (Executive):
1596         (Executive.popen):
1597         * Scripts/webkitpy/common/system/executive_unittest.py:
1598         (ExecutiveTest.test_popen_args):
1599
1600 2012-12-17  Mark Pilgrim  <pilgrim@chromium.org>
1601
1602         Use Platform::current() instead of webKitPlatformSupport() in DumpRenderTree
1603         https://bugs.webkit.org/show_bug.cgi?id=105054
1604
1605         Reviewed by Jochen Eisinger.
1606
1607         Part of a refactoring series; see tracking bug 82948.
1608
1609         * DumpRenderTree/chromium/NotificationPresenter.cpp:
1610         (NotificationPresenter::show):
1611         * DumpRenderTree/chromium/Task.cpp:
1612         (postTask):
1613         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp:
1614
1615 2012-12-17  Emil A Eklund  <eae@chromium.org>
1616
1617         Clamp values in LayoutUnit::operator/ when SATURATED_LAYOUT_ARITHMETIC is enabled
1618         https://bugs.webkit.org/show_bug.cgi?id=104955
1619
1620         Reviewed by Julien Chaffraix.
1621
1622         Add tests for LayoutUnit.
1623
1624         * TestWebKitAPI/CMakeLists.txt:
1625         * TestWebKitAPI/ForwardingHeaders: Added.
1626         * TestWebKitAPI/ForwardingHeaders/WebCore: Added.
1627         * TestWebKitAPI/ForwardingHeaders/WebCore/LayoutUnit.h: Added.
1628         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
1629         * TestWebKitAPI/TestWebKitAPI.gypi:
1630         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1631         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp: Added.
1632         (TestWebKitAPI):
1633         (TestWebKitAPI::TEST):
1634         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
1635
1636 2012-12-17  Jochen Eisinger  <jochen@chromium.org>
1637
1638         [chromium] removed unused preferences from testRunner.overridePreference
1639         https://bugs.webkit.org/show_bug.cgi?id=105169
1640
1641         Reviewed by Tony Chang.
1642
1643         Most notably this removes the code related to font family maps which is
1644         now handled via window.internals.settings
1645
1646         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
1647         (WebPreferences):
1648         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
1649         (WebTestRunner::TestRunner::overridePreference):
1650         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
1651         (TestRunner):
1652         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
1653         (WebTestRunner::WebPreferences::applyTo):
1654
1655 2012-12-17  James Robinson  <jamesr@chromium.org>
1656
1657         [chromium] Move creation of WebLayerTreeView to WebWidgetClient
1658         https://bugs.webkit.org/show_bug.cgi?id=105071
1659
1660         Reviewed by Adrienne Walker.
1661
1662         Update WebViewHost for the new WebWidgetClient interface.
1663
1664         * DumpRenderTree/chromium/WebViewHost.cpp:
1665         (WebViewHost::initializeLayerTreeView):
1666         (WebViewHost::layerTreeView):
1667         (WebViewHost::~WebViewHost):
1668         * DumpRenderTree/chromium/WebViewHost.h:
1669         (WebViewHost):
1670
1671 2012-12-17  Kenneth Rohde Christiansen  <kenneth@webkit.org>
1672
1673         [EFL][WK2] window_create doesn't receive the url
1674         https://bugs.webkit.org/show_bug.cgi?id=105184
1675
1676         Reviewed by Alexis Menard.
1677
1678         * MiniBrowser/efl/main.c: Use the url supplied to window_create
1679         (on_window_create):
1680
1681 2012-12-17  Ilya Tikhonovsky  <loislo@chromium.org>
1682
1683         Web Inspector: Native Memory Instrumentation: MemoryInstrumentation doesn't detect reportMemoryUsage method defined in a base class.
1684         https://bugs.webkit.org/show_bug.cgi?id=105026
1685
1686         Reviewed by Yury Semikhatsky.
1687
1688         Old SFINAE test was replaced with new one based on this article:
1689         http://stackoverflow.com/questions/1966362/sfinae-to-check-for-inherited-member-functions
1690
1691         New test which covers this problem was added.
1692
1693         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
1694
1695 2012-12-17  Anthony Scian  <ascian@rim.com>
1696
1697         [Blackberry] Static code analysis warning fixes
1698         https://bugs.webkit.org/show_bug.cgi?id=105129
1699
1700         Reviewed by Rob Buis.
1701
1702         Added null checks for fopen calls, no recovery attempted; access fault prevention only.
1703
1704         * DumpRenderTree/blackberry/DumpRenderTree.cpp:
1705         (BlackBerry::WebKit::createFile):
1706         (BlackBerry::WebKit::DumpRenderTree::runTest):
1707
1708 2012-12-17  John Mellor  <johnme@chromium.org>
1709
1710         Add Anton Vayvod and Tim Volodine as contributors in committers.py
1711         https://bugs.webkit.org/show_bug.cgi?id=105025
1712
1713         Unreviewed. Adds two contributors to committers.py.
1714         They are initially contributing to the Text Autosizing feature (master bug: 84186).
1715
1716         * Scripts/webkitpy/common/config/committers.py:
1717
1718 2012-12-17  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
1719
1720         Unreviewed, add myself to committers.py.
1721
1722         * Scripts/webkitpy/common/config/committers.py:
1723
1724 2012-12-17  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
1725
1726         [Qt] remove some unnecessary CONFIG additions
1727
1728         Reviewed by Simon Hausmann.
1729
1730         qt is already added by spec_pre.prf, warn_on and depend_includepath by
1731         default_pre.prf.
1732
1733         * qmake/mkspecs/features/default_post.prf:
1734
1735 2012-12-17  Ilya Tikhonovsky  <loislo@chromium.org>
1736
1737         Unreviewed, rolling out r137892.
1738         http://trac.webkit.org/changeset/137892
1739         https://bugs.webkit.org/show_bug.cgi?id=105026
1740
1741         it broke compilation on windows
1742
1743         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
1744
1745 2012-12-14  Ilya Tikhonovsky  <loislo@chromium.org>
1746
1747         Web Inspector: Native Memory Instrumentation: MemoryInstrumentation doesn't detect reportMemoryUsage method defined in a base class.
1748         https://bugs.webkit.org/show_bug.cgi?id=105026
1749
1750         Reviewed by Yury Semikhatsky.
1751
1752         Old SFINAE test was replaced with new one based on this article:
1753         http://stackoverflow.com/questions/1966362/sfinae-to-check-for-inherited-member-functions
1754
1755         New test which covers this problem was added.
1756
1757         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
1758
1759 2012-12-17  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
1760
1761         [Qt] rely on automatic output directory setup
1762
1763         Reviewed by Simon Hausmann.
1764
1765         Now being a proper qt module (by virtue of having load(qt_build_config)
1766         in .qmake.conf), webkit gets the path setup goodies for free.
1767
1768         This also fixes Makefile.api.Debug/Release trying to generate qrc_WebKit.cpp
1769         at the same time, instead of the debug and release version ending up in different
1770         directories as they should.
1771
1772         * qmake/mkspecs/features/default_pre.prf:
1773
1774 2012-12-17  Charles Wei  <charles.wei@torchmobile.com.cn>
1775
1776         [BlackBerry] Need to re-enable video track
1777         https://bugs.webkit.org/show_bug.cgi?id=105155
1778
1779         Reviewed by George Staikos.
1780
1781         Just enable Video Track for BlackBerry.
1782
1783         * Scripts/webkitperl/FeatureList.pm:
1784
1785 2012-12-17  János Badics  <jbadics@inf.u-szeged.hu>
1786
1787         Unreviewed. Added myself to committers.py
1788
1789         * Scripts/webkitpy/common/config/committers.py:
1790
1791 2012-12-17  Jonathan Liu  <net147@gmail.com>
1792
1793         Fix linking of libraries with -Wl,-whole-archive
1794         https://bugs.webkit.org/show_bug.cgi?id=104436
1795
1796         Reviewed by Simon Hausmann.
1797
1798         If there is more than one library in LIBS linked with
1799         -Wl,-whole-archive -l... -Wl,-no-whole-archive, qmake tries to
1800         merge the duplicate flags in LIBS which causes linking to fail.
1801
1802         Add no_smart_library_merge to CONFIG to prevent the duplicate
1803         flags from being removed when using -Wl,-whole-archive.
1804
1805         * qmake/mkspecs/features/functions.prf:
1806
1807 2012-12-16  Arvid Nilsson  <anilsson@rim.com>
1808
1809         [BlackBerry] Adapt to new BlackBerry::Platform::TouchPoint API
1810         https://bugs.webkit.org/show_bug.cgi?id=105143
1811         RIM PR 171941
1812
1813         Reviewed by Rob Buis.
1814         Internally reviewed by George Staikos.
1815
1816         Adapt to new method names and encapsulation of TouchPoint data members.
1817
1818         * DumpRenderTree/blackberry/EventSender.cpp:
1819         (addTouchPointCallback):
1820         (updateTouchPointCallback):
1821         (touchEndCallback):
1822         (releaseTouchPointCallback):
1823         (sendTouchEvent):
1824
1825 2012-12-16  Maciej Stachowiak  <mjs@apple.com>
1826
1827         Add a script to compute the mean and 95% confidence interval (using two-sided t-test) of a sample set
1828         https://bugs.webkit.org/show_bug.cgi?id=105148
1829
1830         Reviewed by Filip Pizlo.
1831
1832         Usage examples:
1833         
1834         $ sampstat --help
1835         Usage: sampstat [options] [FILES]
1836           Compute the mean and 95% confidence interval of a sample set.
1837           Standard input or files must contain two or more decimal numbers, one per line.
1838         
1839         Options:
1840           -h, --help            show this help message and exit
1841           -u UNIT, --unit=UNIT  assume values are in units of UNIT
1842           -v, --verbose         print all values (with units)
1843         
1844         $ sampstat -u MB memresults.txt
1845         2356.90 MB +/- 101.34 MB (4.3%)
1846         
1847         $ sampstat -v -u MB memresults.txt
1848               2318.21 MB
1849               2399.56 MB
1850               2352.93 MB
1851         ----------------
1852         Mean: 2356.90 MB +/- 101.34 MB (4.3%)
1853         
1854
1855         * Scripts/sampstat: Added.
1856         (sum): Helper function to compute the sum of a list.
1857         (arithmeticMean): Compute the meam of a list.
1858         (standardDeviation): Compute the sample standard deviation (unbiased estimator).
1859         (standardError): Compute the sample standard error.
1860         (tDist): Compute t(.025, n-1), the t-value for atwo-sided 95% confidence interval. 
1861         (twoSidedConfidenceInterval): Compute the two-sided confidence interval range about the mean.
1862
1863 2012-12-15  Nima Ghanavatian  <nghanavatian@rim.com>
1864
1865         [BlackBerry] Pass key modifiers with touch and mouse events
1866         https://bugs.webkit.org/show_bug.cgi?id=105108
1867
1868         Reviewed by Rob Buis.
1869
1870         Update DRT to new api.
1871
1872         Internally reviewed by Gen Mak.
1873
1874         * DumpRenderTree/blackberry/EventSender.cpp:
1875         (mouseDownCallback):
1876         (mouseUpCallback):
1877         (mouseMoveToCallback):
1878
1879 2012-12-15  Zan Dobersek  <zandobersek@gmail.com>
1880
1881         Stop the XvfbDriver before proceeding with its starting setup
1882         https://bugs.webkit.org/show_bug.cgi?id=105006
1883
1884         Reviewed by Philippe Normand.
1885
1886         Just as with the base Driver implementation, stop the XvfbDriver
1887         before proceeding with setting up lock files and a new Xvfb instance.
1888         This forces any driver or Xvfb instances that are currently under
1889         the driver's control to be closed/killed and cleaned up after.
1890
1891         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
1892         (XvfbDriver._start):
1893
1894 2012-12-14  Dirk Pranke  <dpranke@chromium.org>
1895
1896         webkitpy: make port_name a required parameter to the non-test ports
1897         https://bugs.webkit.org/show_bug.cgi?id=105057
1898
1899         Reviewed by Eric Seidel.
1900
1901         This is just minor code cleanup along the path to making Port
1902         constructors simpler and pushing more logic into the port factory.
1903
1904         * Scripts/webkitpy/layout_tests/port/base.py:
1905         (Port.__init__):
1906         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
1907         (PortTest.make_port):
1908         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
1909         (TestWebKitPort.__init__):
1910         * Scripts/webkitpy/layout_tests/port/test.py:
1911         (TestPort):
1912         (TestPort.determine_full_port_name):
1913         (TestPort.__init__):
1914         * Scripts/webkitpy/layout_tests/port/xvfbdriver_unittest.py:
1915         (XvfbDriverTest.make_driver):
1916         (XvfbDriverTest.test_stop):
1917
1918 2012-12-14  Dirk Pranke  <dpranke@chromium.org>
1919
1920         garden-o-matic doesn't know about reftests
1921         https://bugs.webkit.org/show_bug.cgi?id=101976
1922
1923         Reviewed by Eric Seidel.
1924
1925         Attempt to make garden-o-matic properly aware of reftests again.
1926
1927         This is re-landing r137407 with a change to
1928         rebaselineWithStatusUpdates() to actually pass the filtered
1929         failureInofLost to checkout.rebaseline().
1930
1931         Unfortunately, the testing coverage for this whole module is almost
1932         non-existent, and it's not obvious that there's a good way to
1933         stub out checkout.rebaseline() to test that the right thing
1934         happens. I verified the change manually.
1935
1936         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
1937         (.):
1938         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
1939         (.):
1940         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
1941
1942 2012-12-14  Ryosuke Niwa  <rniwa@webkit.org>
1943
1944         "Running 1 DumpRenderTree over X shards" is not a helpful output
1945         https://bugs.webkit.org/show_bug.cgi?id=104858
1946
1947         Reviewed by Dirk Pranke.
1948
1949         The word "shard" is not typically used in test frameworks. This word is confusing for many WebKit
1950         contributors who typically don't have much experience on distributed computations and databases.
1951         So only print this message out in the debug mode.
1952
1953         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1954         * Scripts/webkitpy/layout_tests/views/printing.py:
1955         (Printer.print_workers_and_shards):
1956
1957 2012-12-13  Eric Seidel  <eric@webkit.org>
1958
1959         Add --profiler=PROFILER option to run-perf-tests to allow specifying which profiler to use on platforms with many
1960         https://bugs.webkit.org/show_bug.cgi?id=104891
1961
1962         Reviewed by Ryosuke Niwa.
1963
1964         I also implemented a very simple "Sample" Profiler using
1965         Mac OS X's /usr/bin/sample command line tool.
1966
1967         The real reason for this abstraction is to make it easy
1968         to support both perf and pprof on linux which seem to
1969         be about equally popular among those I ask in the Chrome team.
1970
1971         * Scripts/webkitpy/common/system/profiler.py:
1972         (ProfilerFactory.create_profiler):
1973         (ProfilerFactory):
1974         (ProfilerFactory.available_profilers_by_name):
1975         (ProfilerFactory.default_profiler_name):
1976         (Sample):
1977         (Sample.__init__):
1978         (Sample.attach_to_pid):
1979         (Sample.profile_after_exit):
1980         (IProfiler.attach_to_pid):
1981         * Scripts/webkitpy/layout_tests/port/driver.py:
1982         (Driver.__init__):
1983         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1984         (PerfTestsRunner._parse_args):
1985
1986 2012-12-13  Eric Seidel  <eric@webkit.org>
1987
1988         Unreviewed, fixing typo in python unittest.
1989
1990         I changed the default profiler on linux from pprof to perf in
1991         bug 104971.  I failed to update the unittest results at that time.
1992
1993         * Scripts/webkitpy/common/system/profiler_unittest.py:
1994         (ProfilerFactoryTest.test_basic):
1995
1996 2012-12-13  Ian Vollick  <vollick@chromium.org>
1997
1998         [chromium] Add a virtual test suite for enabling opt-in to composited scrolling
1999         https://bugs.webkit.org/show_bug.cgi?id=104911
2000
2001         Reviewed by Dirk Pranke.
2002
2003         Adds two virtual tests suites to run the tests in
2004         compositing/overflow/ and scrollbars/ with the flag
2005         --enable-accelerated-overflow-scroll, which I've also plumbed through
2006         in the usual way.
2007
2008         * DumpRenderTree/chromium/DumpRenderTree.cpp:
2009         (main):
2010         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
2011         (WebPreferences):
2012         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
2013         (WebTestRunner::WebPreferences::reset):
2014         (WebTestRunner::WebPreferences::applyTo):
2015         * DumpRenderTree/chromium/TestShell.cpp:
2016         (TestShell::TestShell):
2017         (TestShell::resetWebSettings):
2018         * DumpRenderTree/chromium/TestShell.h:
2019         (TestShell::setAcceleratedCompositingForOverflowScrollEnabled):
2020         (TestShell):
2021         * Scripts/webkitpy/layout_tests/port/chromium.py:
2022         (ChromiumPort.virtual_test_suites):
2023
2024 2012-12-13  Eric Seidel  <eric@webkit.org>
2025
2026         Use 'perf' to profile on linux instead of google-pprof by default
2027         https://bugs.webkit.org/show_bug.cgi?id=104971
2028
2029         Reviewed by Dirk Pranke.
2030
2031         This makes Chromium Linux match Chromium Android and use the perf
2032         tool by default.  Once https://bugs.webkit.org/show_bug.cgi?id=104891
2033         lands it will be possible to use pprof again on Linux.
2034
2035         This is slightly more advanced than the AndroidPerf profiler as
2036         instead of using a timeout on "perf record" I instead watch
2037         for the termination of the target process and then control-C
2038         the 'perf record' process.  This required me to add two new
2039         methods to Executive, one to have a limited-time wait() and
2040         the second to be able to send a control-C.  I chose to add
2041         these to Executive to make them easier to mock/fix-for-win32
2042         at a later time if needed.
2043
2044         * Scripts/webkitpy/common/system/executive.py:
2045         (Executive.wait_limited):
2046         (Executive.interrupt):
2047         * Scripts/webkitpy/common/system/profiler.py:
2048         (ProfilerFactory.create_profiler):
2049         (Perf):
2050         (Perf.__init__):
2051         (Perf._perf_path):
2052         (Perf.attach_to_pid):
2053         (Perf._first_ten_lines_of_profile):
2054         (Perf.profile_after_exit):
2055
2056 2012-12-13  Julie Parent  <jparent@chromium.org>
2057
2058         Dashboard cleanup: remove usage of global g_builders.
2059         https://bugs.webkit.org/show_bug.cgi?id=104941
2060
2061         Reviewed by Dirk Pranke.
2062
2063         The dashboards use a lot of global state, which makes hacking on them
2064         complicated. This change removes the use of one such global: g_builders.
2065         In most cases, we can just use currentBuilderGroup().builders instead,
2066         which is now currentBuilders().
2067         Surprisingly, the most changes were required to the unit tests, since
2068         they were even bigger offenders of bad hygiene, relying on global state
2069         set by other tests, randomly clobbering global variables in ways the
2070         real code doesn't, etc.
2071
2072         * TestResultServer/static-dashboards/builders.js:
2073         (BuilderGroup.prototype.setup):
2074         * TestResultServer/static-dashboards/dashboard_base.js:
2075         (.switch.return):
2076         (htmlForTestTypeSwitcher):
2077         * TestResultServer/static-dashboards/flakiness_dashboard.js:
2078         (generatePage):
2079         (getAllTestsTrie):
2080         (processTestRunsForAllBuilders):
2081         (showPopupForBuild):
2082         (generatePageForExpectationsUpdate):
2083         (loadExpectationsLayoutTests):
2084         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2085         (resetGlobals):
2086         (stubResultsByBuilder):
2087         (test):
2088         * TestResultServer/static-dashboards/loader.js:
2089         (.):
2090         * TestResultServer/static-dashboards/loader_unittests.js:
2091
2092 2012-12-13  Eric Seidel  <eric@webkit.org>
2093
2094         Callers should not have to stringify args before calling Executive run_command/popen
2095         https://bugs.webkit.org/show_bug.cgi?id=104975
2096
2097         Reviewed by Dirk Pranke.
2098
2099         One could argue that we should match the python call syntax here,
2100         but I think it's a more friendly API if we automagically handle
2101         stringification of args in run_command, etc.
2102         This removes map(unicode, args) from several callsites.
2103
2104         When I first tried to land this change, I didn't realize that
2105         Executive._command_for_printing depended on this behavior
2106         having been applied to args in run_command.  The fix is to
2107         call _stringify_args in both run_command and popen.
2108         This is slightly redundant, but given how short args have to be
2109         (due to shell limits), I don't think the double-encode check
2110         matters in practice.
2111
2112         This is slightly complicated by the fact that apache_http_server.py
2113         is the one caller in our codebase which uses shell=True.
2114         shell=True is a well-documented trail-of-tears:
2115         http://stackoverflow.com/questions/3172470/actual-meaning-of-shell-true-in-subprocess
2116         but to support this legacy (windows-only) code (which I can't easily test)
2117         I've added an if-hack to avoid stringifying the the popen(shell=True) case.
2118
2119         * Scripts/webkitpy/common/system/executive.py:
2120         (Executive.run_command):
2121         (Executive._stringify_args):
2122         (Executive.popen):
2123         * Scripts/webkitpy/common/system/executive_unittest.py:
2124         (ExecutiveTest.test_auto_stringify_args):
2125         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2126         (attach_to_pid):
2127
2128 2012-12-13  Adrienne Walker  <enne@chromium.org>
2129
2130         Unreviewed, rolling out r137645, r137646, and r137667.
2131         http://trac.webkit.org/changeset/137645
2132         http://trac.webkit.org/changeset/137646
2133         http://trac.webkit.org/changeset/137667
2134         https://bugs.webkit.org/show_bug.cgi?id=104911
2135
2136         Breaks some overflow layout tests
2137
2138         * Scripts/webkitpy/layout_tests/port/chromium.py:
2139         (ChromiumPort.virtual_test_suites):
2140
2141 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
2142
2143         Unreviewed, rolling out r137661.
2144         http://trac.webkit.org/changeset/137661
2145         https://bugs.webkit.org/show_bug.cgi?id=104891
2146
2147         broke unit tests, run-webkit-tests
2148
2149         * Scripts/webkitpy/common/system/executive.py:
2150         (Executive._run_command_with_teed_output):
2151         (Executive.run_command):
2152         (Executive.popen):
2153         * Scripts/webkitpy/common/system/profiler.py:
2154         (ProfilerFactory.create_profiler):
2155         (GooglePProf.profile_after_exit):
2156         (IProfiler.attach_to_pid):
2157         * Scripts/webkitpy/layout_tests/port/driver.py:
2158         (Driver.__init__):
2159         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2160         (PerfTestsRunner._parse_args):
2161
2162 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
2163
2164         nrwt: move rundetails, summarize_results into test_run_results
2165         https://bugs.webkit.org/show_bug.cgi?id=104963
2166
2167         Reviewed by Eric Seidel.
2168
2169         More cleanup; this the plain-old-data functions out of manager.py
2170         and into test_run_results next to the other aggregate data structures.
2171
2172         Also this cleans up a bunch of test code and reduces duplication.
2173
2174         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2175         (Manager.run):
2176         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
2177         (ManagerTest.test_look_for_new_crash_logs):
2178         * Scripts/webkitpy/layout_tests/models/test_run_results.py:
2179         (TestRunResults.__init__):
2180         (TestRunResults.add):
2181         (RunDetails):
2182         (RunDetails.__init__):
2183         (_interpret_test_failures):
2184         (summarize_results):
2185         * Scripts/webkitpy/layout_tests/models/test_run_results_unittest.py: Added.
2186         (get_result):
2187         (run_results):
2188         (summarized_results):
2189         (InterpretTestFailuresTest):
2190         (InterpretTestFailuresTest.setUp):
2191         (InterpretTestFailuresTest.test_interpret_test_failures):
2192         (SummarizedResultsTest):
2193         (SummarizedResultsTest.setUp):
2194         (SummarizedResultsTest.test_no_svn_revision):
2195         (SummarizedResultsTest.test_svn_revision):
2196         (SummarizedResultsTest.test_summarized_results_wontfix):
2197         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
2198         (BuildBotPrinterTests.get_printer):
2199         (BuildBotPrinterTests.test_print_unexpected_results):
2200         (BuildBotPrinterTests.test_print_results):
2201
2202 2012-12-13  Eric Seidel  <eric@webkit.org>
2203
2204         Make --profile output easier to understand for pprof on linux
2205         https://bugs.webkit.org/show_bug.cgi?id=104957
2206
2207         Reviewed by Dirk Pranke.
2208
2209         pprof is kinda mysterious, so we print some information
2210         on how to use it, and how to learn how to read the sample dump.
2211         This is similar to what we've done for perf output for Chromium Android.
2212
2213         * Scripts/webkitpy/common/system/profiler.py:
2214         (GooglePProf._pprof_path):
2215         (GooglePProf.profile_after_exit):
2216
2217 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
2218
2219         Revert r137407 - it didn't actually work right.
2220         https://bugs.webkit.org/show_bug.cgi?id=101976
2221
2222         Unreviewed, build fix.
2223
2224         Turns out that rebaselining was completely broken with this change :(
2225         We're passing in a list of test names rather than the list of
2226         failureInfo objects.
2227
2228         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
2229         (.):
2230         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
2231         (.):
2232         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
2233
2234 2012-12-13  Antoine Labour  <piman@chromium.org>
2235
2236         [chromium] Small fixes for WebPluginContainer::setWebLayer
2237         https://bugs.webkit.org/show_bug.cgi?id=104953
2238
2239         Reviewed by James Robinson.
2240
2241         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp: Clean up
2242         dangling pointers, force correct invalidation of the layer.
2243
2244 2012-12-13  Eric Seidel  <eric@webkit.org>
2245
2246         Add --profiler=PROFILER option to run-perf-tests to allow specifying which profiler to use on platforms with many
2247         https://bugs.webkit.org/show_bug.cgi?id=104891
2248
2249         Reviewed by Dirk Pranke.
2250
2251         I also implemented a very simple "Sample" Profiler using
2252         Mac OS X's /usr/bin/sample command line tool.
2253
2254         I also moved the map(unicode, args) call into Executive.popen
2255         so that all callers don't have to do it themselves.
2256
2257         The real reason for this abstraction is to make it easy
2258         to support both perf and pprof on linux which seem to
2259         be about equally popular among those I ask in the Chrome team.
2260
2261         * Scripts/webkitpy/common/system/executive.py:
2262         (Executive.popen):
2263         * Scripts/webkitpy/common/system/profiler.py:
2264         (ProfilerFactory.create_profiler):
2265         (ProfilerFactory):
2266         (ProfilerFactory.available_profilers_by_name):
2267         (ProfilerFactory.default_profiler_name):
2268         (Sample):
2269         (Sample.__init__):
2270         (Sample.attach_to_pid):
2271         (Sample.profile_after_exit):
2272         (IProfiler.attach_to_pid):
2273         * Scripts/webkitpy/layout_tests/port/driver.py:
2274         (Driver.__init__):
2275         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2276         (PerfTestsRunner._parse_args):
2277
2278 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
2279
2280         support -wk2 port names properly in webkitpy.layout_tests.port
2281         https://bugs.webkit.org/show_bug.cgi?id=104761
2282
2283         Reviewed by Eric Seidel.
2284
2285         Our support for WK2 port names was inconsistent at best; this patch
2286         modifies the code so that <port_name>-wk2 is supported for all ports that
2287         have WK2 implementations and is equivalent to specifying <port_name>
2288         and -2/--webkit-test-runner. In addition, this modifies
2289         builders.all_port_names() to include the wk2 variants.
2290
2291         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
2292         * Scripts/webkitpy/layout_tests/port/apple.py:
2293         (ApplePort.determine_full_port_name):
2294         * Scripts/webkitpy/layout_tests/port/base.py:
2295         (Port.determine_full_port_name):
2296         (Port.__init__):
2297         * Scripts/webkitpy/layout_tests/port/builders.py:
2298         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
2299         (DriverTest.make_port):
2300         * Scripts/webkitpy/layout_tests/port/mac.py:
2301         (MacPort.default_baseline_search_path):
2302         * Scripts/webkitpy/layout_tests/port/win.py:
2303         (WinPort.default_baseline_search_path):
2304         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
2305         (TestRebaselineExpectations.test_rebaseline_expectations):
2306
2307 2012-12-13  Ian Vollick  <vollick@chromium.org>
2308
2309         [chromium] Add a virtual test suite for enabling opt-in to composited scrolling
2310         https://bugs.webkit.org/show_bug.cgi?id=104911
2311
2312         Reviewed by Adrienne Walker.
2313
2314         Adds two virtual tests suites to run the tests in
2315         compositing/overflow/ and scrollbars/ with the flag
2316         --enable-accelerated-compositing-for-overflow-scroll
2317
2318         * Scripts/webkitpy/layout_tests/port/chromium.py:
2319         (ChromiumPort.virtual_test_suites):
2320
2321 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
2322
2323         delete extraneous self.fail in mastercfg_unittest.py
2324         https://bugs.webkit.org/show_bug.cgi?id=104940
2325
2326         Reviewed by Eric Seidel.
2327
2328         This was accidentally introduced in r137404.
2329
2330         * BuildSlaveSupport/build.webkit.org-config/mastercfg_unittest.py:
2331
2332 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
2333
2334         Fix a mis-named variable after r137621.
2335         https://bugs.webkit.org/show_bug.cgi?id=104871
2336
2337         Unreviewed, build fix.
2338
2339         * Scripts/webkitpy/layout_tests/views/buildbot_results.py:
2340         (BuildBotPrinter.print_results):
2341         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
2342
2343 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
2344
2345         webkitpy: rename ResultSummary
2346         https://bugs.webkit.org/show_bug.cgi?id=104871
2347
2348         Reviewed by Eric Seidel.
2349
2350         This patch changes the names of ResultSummary and related modules,
2351         classes, and variables as a part of the larger clean up in bug 103824.
2352
2353         This change is basically a bunch of search and replaces:
2354         - result_summary.py -> test_run_results.py
2355         - class ResultSummary -> class TestRunResults
2356         - "result_summary" or "rs" as a generic variable name -> "run_results",
2357           when it doesn't matter (or we don't know) if this refers to the
2358           initial pass over the tests or the retry pass.
2359         - "result_summary" as a variable containing the results of the initial,
2360           complete test run -> "initial_results"
2361         - "retry_summary" as a variable containing the results of retried
2362           results -> "retry_results"
2363         - ResultSummary.results (the property) -> TestRunResults.results_by_name
2364         - ResultSummary.failures (the property) -> TestRunResults.failures_by_name
2365         - RunDetails.result_summary -> RunDetails.initial_results
2366         - RunDetails.retry_summary -> RunDetails.retry_results
2367
2368         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
2369         (LayoutTestRunner.__init__):
2370         (LayoutTestRunner.run_tests):
2371         (LayoutTestRunner._mark_interrupted_tests_as_skipped):
2372         (LayoutTestRunner._interrupt_if_at_failure_limits):
2373         (LayoutTestRunner._interrupt_if_at_failure_limits.interrupt_if_at_failure_limit):
2374         (LayoutTestRunner._update_summary_with_result):
2375         (LayoutTestRunner._handle_finished_test):
2376         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner_unittest.py:
2377         (FakePrinter.print_expected):
2378         (LayoutTestRunnerTests.test_interrupt_if_at_failure_limits):
2379         (LayoutTestRunnerTests.test_update_summary_with_result):
2380         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2381         (RunDetails.__init__):
2382         (summarize_results):
2383         (Manager.run):
2384         (Manager._look_for_new_crash_logs):
2385         (Manager._tests_to_retry):
2386         (Manager._upload_json_files):
2387         (Manager._stats_trie):
2388         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
2389         (ManagerTest.test_look_for_new_crash_logs):
2390         (SummarizeResultsTest):
2391         (SummarizeResultsTest.get_run_results):
2392         (SummarizeResultsTest.summarized_results):
2393         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
2394         (JSONLayoutResultsGenerator.__init__):
2395         (JSONLayoutResultsGenerator._get_modifier_char):
2396         (JSONLayoutResultsGenerator._insert_failure_summaries):
2397         (JSONLayoutResultsGenerator._get_failure_summary_entry):
2398         * Scripts/webkitpy/layout_tests/models/test_run_results.py: Renamed from Tools/Scripts/webkitpy/layout_tests/models/result_summary.py.
2399         (TestRunResults):
2400         (TestRunResults.__init__):
2401         (TestRunResults.add):
2402         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2403         (get_test_results):
2404         (MainTest.test_basic):
2405         * Scripts/webkitpy/layout_tests/views/buildbot_results.py:
2406         (BuildBotPrinter.print_results):
2407         (BuildBotPrinter.print_run_results):
2408         (BuildBotPrinter._print_run_results_entry):
2409         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
2410         (BuildBotPrinterTests.get_run_results):
2411         (BuildBotPrinterTests.test_print_unexpected_results.get_unexpected_results):
2412         (BuildBotPrinterTests):
2413         * Scripts/webkitpy/layout_tests/views/printing.py:
2414         (Printer.print_expected):
2415         (Printer._print_expected_results_of_type):
2416         (Printer.print_results):
2417         (Printer._print_timing_statistics):
2418         (Printer._print_worker_statistics):
2419         (Printer._print_aggregate_test_statistics):
2420         (Printer._print_individual_test_times):
2421         (Printer._print_directory_timings):
2422         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
2423
2424 2012-12-13  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2425
2426         [EFL] The WebKit2 bots are building and executing WebKit1 tests
2427         https://bugs.webkit.org/show_bug.cgi?id=104690
2428
2429         Reviewed by Csaba Osztrogonác.
2430
2431         Pass --no-webkit1 option to the EFL WK2 bots when building. This
2432         will disable WK1 API and thus, the WK1 API tests.
2433
2434         * BuildSlaveSupport/build.webkit.org-config/config.json:
2435         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2436         (CompileWebKit2Only):
2437         (BuildAndTestFactory):
2438         (BuildAndTestFactory.__init__):
2439         (BuildAndTestWebKit2OnlyFactory):
2440         Now bots using BuildAndTestWebKit2OnlyFactory (for instance only EFL) will
2441         get --no-webkit1 as a parameter on the build-webkit script.
2442         * Scripts/build-webkit:
2443         Added support for -no-webkit1 for the Qt and EFL port.
2444
2445 2012-12-13  Eric Seidel  <eric@webkit.org>
2446
2447         build-webkit --chromium-android forces Make even when update-webkit used --ninja
2448         https://bugs.webkit.org/show_bug.cgi?id=104839
2449
2450         Reviewed by Tony Chang.
2451
2452         This was recommended by Peter as part of:
2453         https://bugs.webkit.org/show_bug.cgi?id=104434#c4
2454         Fixing this makes build-webkit --chromium-android work for the common (ninja) case.
2455
2456         * Scripts/webkitdirs.pm:
2457         (buildChromium):
2458
2459 2012-12-13  Julie Parent  <jparent@chromium.org>
2460
2461         Loader_unittests should reset state between runs.
2462         https://bugs.webkit.org/show_bug.cgi?id=104862
2463
2464         Reviewed by Dirk Pranke.
2465
2466         The loader_unittests were not resetting state between runs, leaving
2467         a lot of data behind in global objects that the dashboard and tests
2468         rely on.  A handful were resetting just a few objects.  Use the common
2469         resetGlobals() method instead.
2470
2471         * TestResultServer/static-dashboards/loader_unittests.js:
2472
2473 2012-12-13  Andras Becsi  <andras.becsi@digia.com>
2474
2475         [Qt][WK2] Fix painting on Mac with retina display
2476         https://bugs.webkit.org/show_bug.cgi?id=104574
2477
2478         Reviewed by Kenneth Rohde Christiansen.
2479
2480         Remove setting the devicePixelRatio experimental property
2481         since the value is now automatically picked up from Qt.
2482
2483         * MiniBrowser/qt/qml/BrowserWindow.qml:
2484
2485 2012-12-13  Jussi Kukkonen  <jussi.kukkonen@intel.com>
2486
2487         [EFL][GTK] Don't call deprecated g_type_init when glib => 2.35
2488         https://bugs.webkit.org/show_bug.cgi?id=103209
2489
2490         Reviewed by Kenneth Rohde Christiansen.
2491
2492         g_type_init() is deprecated from 2.35.0 onwards. Don't call it
2493         in that case to avoid warnings and/or build failure.
2494
2495         * DumpRenderTree/gtk/ImageDiff.cpp:
2496         (main):
2497
2498 2012-12-13  Claudio Saavedra  <csaavedra@igalia.com>
2499
2500         [Tools] Add exception for gtk_style_context_get()
2501         https://bugs.webkit.org/show_bug.cgi?id=104896
2502
2503         Reviewed by Martin Robinson.
2504
2505         Add new exception for gtk_style_context_get().
2506
2507         * Scripts/webkitpy/style/checkers/cpp.py:
2508         (check_for_null): Add exception for gtk_style_context_get().
2509         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
2510         (WebKitStyleTest.test_null_false_zero): Add a unit
2511         test for the new exception.
2512
2513 2012-12-13  Joaquim Rocha  <jrocha@igalia.com>
2514
2515         REGRESSION (r137432): The /webkit2/WebKitCookieManager/accept-policy unit test is failing
2516         https://bugs.webkit.org/show_bug.cgi?id=104790
2517
2518         Reviewed by Carlos Garcia Campos.
2519
2520         Unskip the GTK accept policy test.
2521
2522         * Scripts/run-gtk-tests:
2523         (TestRunner):
2524
2525 2012-12-13  Xingnan Wang  <xingnan.wang@intel.com>
2526
2527         Unreviewed. Adding myself to committers list.
2528
2529         * Scripts/webkitpy/common/config/committers.py:
2530
2531 2012-12-12  Roger Fong  <roger_fong@apple.com>
2532
2533         Unreviewed. Fix typo in NRWT.
2534
2535         * Scripts/webkitpy/layout_tests/port/base.py:
2536         (Port.is_reference_html_file):
2537
2538 2012-12-12  Roger Fong  <roger_fong@apple.com>
2539
2540         Fix detection of ref test files in old and new run webkit tests.
2541         https://bugs.webkit.org/show_bug.cgi?id=104850.
2542
2543         Reviewed by Timothy Horton.
2544
2545         * Scripts/old-run-webkit-tests:
2546         (isUsedInReftest): 
2547         Add notref and ref prefix and suffix cases to isUsedinReftest.
2548         * Scripts/webkitpy/layout_tests/port/base.py:
2549         (Port):
2550         (Port.is_reference_html_file): 
2551         Typo in NRWT.
2552
2553 2012-12-12  Dirk Pranke  <dpranke@chromium.org>
2554
2555         rework ChromiumAndroidDriver.start() et. al to avoid funky layering
2556         https://bugs.webkit.org/show_bug.cgi?id=104856
2557
2558         Reviewed by Eric Seidel.
2559
2560         r137523 / bug 10357 introduced some back-and-forth hopping during
2561         driver.start() in the chromium android port. This patch cleans it
2562         up a bit.
2563
2564         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2565         (ChromiumAndroidDriver.__init__):
2566         (ChromiumAndroidDriver._start_once):
2567         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
2568         * Scripts/webkitpy/layout_tests/port/driver.py:
2569         (Driver.start):
2570         (Driver._start):
2571
2572 2012-12-12  Eric Seidel  <eric@webkit.org>
2573
2574         Add support for run-perf-tests --chromium-android --profile
2575         https://bugs.webkit.org/show_bug.cgi?id=103587
2576
2577         Reviewed by Dirk Pranke.
2578
2579         This makes it possible to profile PerformanceTests on an Android device
2580         using run-perf-tests --chromium-android --profile.
2581         CPU profiling on Android is sadly non-trivial, but this goes a long
2582         ways towards making it more accessible to non-Android developers.
2583
2584         Although this script requires your device to be flashed with a
2585         userdebug build (an annoyingly non-trivial endeavor for external contributors)
2586         it endeavors to detect and explain the common errors encountered
2587         while trying to setup a device for perf-based profiling.
2588
2589         * Scripts/webkitpy/common/system/profiler.py:
2590         (GooglePProf.profile_after_exit): Discovered that pprof will just hang if passed a missing file.
2591         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2592         (AndroidPerf):
2593         (AndroidPerf.__init__):
2594         (AndroidPerf.check_configuration):
2595         (AndroidPerf.print_setup_instructions):
2596         (_file_exists_on_device):
2597         (_run_adb_command):
2598         (attach_to_pid):
2599         (_perf_version_string):
2600         (_find_perfhost_binary):
2601         (_perfhost_path):
2602         (_first_ten_lines_of_profile):
2603         (profile_after_exit):
2604         (ChromiumAndroidDriver.__init__):
2605         (ChromiumAndroidDriver.__init__.methods):
2606         (ChromiumAndroidDriver._find_or_create_symfs):
2607         (ChromiumAndroidDriver._start_once):
2608         (ChromiumAndroidDriver):
2609         (ChromiumAndroidDriver._run_post_start_tasks):
2610         (ChromiumAndroidDriver._pid_from_android_ps_output):
2611         (ChromiumAndroidDriver._pid_on_target):
2612         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
2613         (ChromiumAndroidDriverTest.test_pid_from_android_ps_output):
2614         (AndroidPerfTest):
2615         (AndroidPerfTest.test_perf_output_regexp):
2616         * Scripts/webkitpy/layout_tests/port/driver.py:
2617         (Driver._setup_environ_for_driver):
2618         (Driver):
2619         (Driver._start):
2620         (Driver._run_post_start_tasks):
2621         (Driver._pid_on_target):
2622
2623 2012-12-12  Dirk Pranke  <dpranke@chromium.org>
2624
2625         nrwt: remove deprecated ORWT compatibility command line flags
2626         https://bugs.webkit.org/show_bug.cgi?id=104301
2627
2628         Reviewed by Eric Seidel.
2629
2630         The command line flags for --noshow-results, --[no-]launch-safari,
2631         and --use-remote-links-to-tests were for compatibility with
2632         old-run-webkit-tests, but the compatibility is not needed any more
2633         since NRWT is basically default and the scripts that passed the
2634         old flags have been updated.
2635
2636         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2637         (_set_up_derived_options):
2638         (parse_args):
2639
2640 2012-12-12  John Griggs  <jgriggs@rim.com>
2641
2642         2012-12-12  John Griggs  <jgriggs@rim.com>
2643
2644         [BlackBerry] Update Media Controls
2645         https://bugs.webkit.org/show_bug.cgi?id=104443
2646         https://przilla.ott.qnx.com/bugzilla/show_bug.cgi?id=204748
2647
2648         Update Media Controls for BlackBerry platform to allow audio, embedded video and fullscreen video controls to be positioned and styled differently.
2649
2650         Reviewed by Rob Buis.
2651
2652         BlackBerry-specific update to Media Controls - No new tests required.
2653
2654         * Scripts/webkitperl/FeatureList.pm:
2655
2656 2012-12-12  Justin Novosad  <junov@google.com>
2657
2658         Adding myself to SkiaGraphics watchlist
2659
2660         Unreviewed
2661
2662         * Scripts/webkitpy/common/config/watchlist:
2663
2664 2012-12-12  Martin Robinson  <mrobinson@igalia.com>
2665
2666         [GTK] Add authentication support to DRT and fix exposed issues in the libsoup backend
2667         https://bugs.webkit.org/show_bug.cgi?id=104285
2668
2669         Reviewed by Gustavo Noronha Silva.
2670
2671         Add support to DumpRenderTree for running authentication tests. Since the DRT
2672         expects an authentication callback, we add one to DRTSupport to avoid #ifdefs
2673         in platform-independent code for GTK+.
2674
2675         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2676         (resetDefaultsToConsistentValues): Reset the authentication password and username.
2677         (authenticationCallback): Added.
2678         (createWebView): Attach the authentiation callback during startup.
2679         * DumpRenderTree/gtk/TestRunnerGtk.cpp:
2680         (soupURIToKURL): soup_uri_to_string does not preserve passwords embedded
2681         in the URL so we add a somewhat messy method of re-adding them when they exist.
2682         It would be nice to use soupURIToKURL here, but it seems we cannot use KURL without
2683         pulling in lots of WebCore code so we use string search and replace.
2684         (TestRunner::queueLoad): Use the new helper.
2685
2686 2012-12-12  Dirk Pranke  <dpranke@chromium.org>
2687
2688         garden-o-matic should prefer efl/ over efl-wk1/ and efl-wk2/ when rebaselining
2689         https://bugs.webkit.org/show_bug.cgi?id=102389
2690
2691         Reviewed by Tony Chang.
2692
2693         The code wasn't aware of both the efl-wk1 and efl-wk2 variants,
2694         because all_port_names() by itself doesn't tell you about the
2695         variants. Work around this for now, and filed bug 104761 to
2696         address the underlying problem.
2697
2698         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
2699         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
2700         (BaselineOptimizerTest.test_efl):
2701
2702 2012-12-12  Krzysztof Czech  <k.czech@samsung.com>
2703
2704         [EFL] Possibility to turn off accessibility feature for WebKit-EFL.
2705         https://bugs.webkit.org/show_bug.cgi?id=103036
2706
2707         Reviewed by Laszlo Gombos.
2708
2709         Guard dependencies for accessibility (ATK), so that they can be turned off.
2710
2711         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.h:
2712         (AccessibilityUIElement):
2713         * WebKitTestRunner/PlatformEfl.cmake:
2714
2715 2012-12-12  Simon Fraser  <simon.fraser@apple.com>
2716
2717         Add some entries in the watchlist, and register me for them.
2718
2719         * Scripts/webkitpy/common/config/watchlist:
2720
2721 2012-12-12  Ilya Tikhonovsky  <loislo@chromium.org>
2722
2723         Web Inspector: Native Memory Instrumentation: remove fake root MemoryObjectInfo.
2724         https://bugs.webkit.org/show_bug.cgi?id=104796
2725
2726         Reviewed by Yury Semikhatsky.
2727
2728         It was not a good idea to introduce a fake root MemoryObjectInfo.
2729         It makes a problem when we visit an object without its own MemoryObjectType.
2730
2731         Example: RenderBox has a global pointer to a hash map.
2732         HashMap doesn't have its own object type because it is a generic container.
2733         It will inherit object type from the fake root memory object info.
2734         The same could happen for another container in another class with other MemoryObjectType.
2735
2736         This fact forces me to create custom process method for root objects
2737         because they need to have their own MemoryObjectInfo with customisable memory object type.
2738
2739         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
2740
2741         Drive by fix: InstrumentedPointer* was replaced with Wrapper* because actually it is using
2742         for instrumented and not instrumented object classes.
2743
2744 2012-12-12  Zan Dobersek  <zandobersek@gmail.com>
2745
2746         [GTK] Remove the last of unnecessary configuration options in configure.ac
2747         https://bugs.webkit.org/show_bug.cgi?id=104793
2748
2749         Reviewed by Martin Robinson.
2750
2751         Removing the filters and indexed database features from the list of features
2752         being configurable through the configure script. This is not the case anymore
2753         as the corresponding configuration options are being removed.
2754
2755         * Scripts/webkitdirs.pm:
2756         (buildAutotoolsProject):
2757
2758 2012-12-12  Mario Sanchez Prada  <mario@webkit.org>
2759
2760         [GTK][WK2] New API to detect display/execution of insecure content
2761         https://bugs.webkit.org/show_bug.cgi?id=104578
2762
2763         Reviewed by Martin Robinson.
2764
2765         Add coding style exception for 'readability/enum_casing' in GTK.
2766
2767         The GTK+ API use upper case, underscore separated, words in
2768         certain types of enums (e.g. signals, properties).
2769
2770         * Scripts/webkitpy/style/checker.py: Added exception, both for
2771         WebKitGTK+ and WebKit2GTK+.
2772
2773 2012-12-12  Jochen Eisinger  <jochen@chromium.org>
2774
2775         [chromium] Fix build on chromium-win
2776
2777         Unreviewed build-fix.
2778
2779         MSVS indeed doesn't have snprintf, adding wtf/StringExtras.h include
2780         to work around that.
2781
2782         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp:
2783
2784 2012-12-12  Simon Hausmann  <simon.hausmann@digia.com>
2785
2786         [Qt] Fix build on Mac
2787
2788         Reviewed by Csaba Osztrogonác.
2789
2790         On Mac OS X we have sqlite as a system library available.
2791
2792         * qmake/mkspecs/features/features.prf:
2793
2794 2012-12-12  Zan Dobersek  <zandobersek@gmail.com>
2795
2796         Unreviewed GTK gardening.
2797
2798         Skipping another unit test that regressed in r137432.
2799
2800         * Scripts/run-gtk-tests:
2801         (TestRunner):
2802
2803 2012-12-12  Simon Hausmann  <simon.hausmann@digia.com>
2804
2805         [Qt] Fix linkage against SQLite
2806         https://bugs.webkit.org/show_bug.cgi?id=104781
2807
2808         Reviewed by Csaba Osztrogonác.
2809
2810         Detect SQLite either through pkg-config, from the SQLITE3SRCDIR
2811         environment variable or from qtbase via a qt5.git build. Bail out early
2812         with an error message if neither option applies.
2813
2814         This removes the usage of $$QT.core.sources and the dependency on Qt's
2815         system-sqlite configure setting. We should always favour a system
2816         library over a copy in Qt.
2817
2818         * qmake/mkspecs/features/features.prf:
2819
2820 2012-12-12  Tommy Widenflycht  <tommyw@google.com>
2821
2822         MediaStream API: Change the data channel descriptor pattern to a handler pattern
2823         https://bugs.webkit.org/show_bug.cgi?id=104543
2824
2825         Reviewed by Kent Tamura.
2826
2827         Adding mocks to be able to test this refactoring.
2828
2829         * DumpRenderTree/DumpRenderTree.gypi:
2830         * DumpRenderTree/chromium/MockWebRTCDataChannelHandler.cpp: Added.
2831         (DataChannelReadyStateTask):
2832         (DataChannelReadyStateTask::DataChannelReadyStateTask):
2833         (MockWebRTCDataChannelHandler::MockWebRTCDataChannelHandler):
2834         (MockWebRTCDataChannelHandler::setClient):
2835         (MockWebRTCDataChannelHandler::bufferedAmount):
2836         (MockWebRTCDataChannelHandler::sendStringData):
2837         (MockWebRTCDataChannelHandler::sendRawData):
2838         (MockWebRTCDataChannelHandler::close):
2839         * DumpRenderTree/chromium/MockWebRTCDataChannelHandler.h: Added.
2840         (MockWebRTCDataChannelHandler):
2841         (MockWebRTCDataChannelHandler::taskList):
2842         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
2843         (RTCPeerConnectionReadyStateTask::RTCPeerConnectionReadyStateTask):
2844         (RemoteDataChannelTask::RemoteDataChannelTask):
2845         (MockWebRTCPeerConnectionHandler::createDataChannel):
2846         (MockWebRTCPeerConnectionHandler::stop):
2847         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h:
2848         (MockWebRTCPeerConnectionHandler):
2849
2850 2012-12-12  Jochen Eisinger  <jochen@chromium.org>
2851
2852         [chromium] WebTestPlugin must not depend on WTF::String
2853         https://bugs.webkit.org/show_bug.cgi?id=104783
2854
2855         Reviewed by Kent Tamura.
2856
2857         In a component build, we can't access the WTF that is part of the
2858         WebKit component. Accessing WTF instead will pull in a separate copy
2859         into the TestRunner library. Instead of WTF::String::number use
2860         snprintf() to print numbers.
2861
2862         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp:
2863
2864 2012-12-12  Jocelyn Turcotte  <jocelyn.turcotte@digia.com>
2865
2866         [Qt] Do not automatically enable force_static_libs_as_shared when using debug_and_release
2867         https://bugs.webkit.org/show_bug.cgi?id=104562
2868
2869         Reviewed by Tor Arne Vestbø.
2870
2871         force_static_libs_as_shared creates issues on Mac if QtWebKit is built as a framework.
2872         QtWebProcess, for example, will link to the framework's release library and will rely
2873         on DYLD_IMAGE_SUFFIX to switch to the debug build. Since QtWebProcess will also links
2874         directly to the debug build of WebCore, and that QtWebKit release will link in the
2875         release build of WebCore, both the debug and release version will be loaded.
2876
2877         To fix this properly we should make sure that applications always link to the release
2878         build of dependent libraries by bypassing resolveFinalLibraryName.
2879         Disable it for now as force_static_libs_as_shared is mainly for development while
2880         debug_and_release is mainly for production builds.
2881
2882         * qmake/mkspecs/features/default_post.prf:
2883
2884 2012-12-12  Zan Dobersek  <zandobersek@gmail.com>
2885
2886         Unreviewed GTK gardening of the unit tests.
2887
2888         Skip two unit tests that regressed during the hackfest.
2889
2890         * Scripts/run-gtk-tests:
2891         (TestRunner):
2892
2893 2012-12-11  Ryosuke Niwa  <rniwa@webkit.org>
2894
2895         Add Koji Ishii to the list of contributors.
2896
2897         * Scripts/webkitpy/common/config/committers.py:
2898
2899 2012-12-11  Dominic Mazzoni  <dmazzoni@google.com>
2900
2901         AX: Make isActionSupported cross-platform.
2902         https://bugs.webkit.org/show_bug.cgi?id=104681
2903
2904         Reviewed by Chris Fleizach.
2905
2906         Replace isActionSupported, which takes a platform-specific
2907         argument, with three cross-platform methods.
2908
2909         * DumpRenderTree/AccessibilityUIElement.cpp:
2910         (isPressActionSupportedCallback):
2911         (isIncrementActionSupportedCallback):
2912         (isDecrementActionSupportedCallback):
2913         (AccessibilityUIElement::getJSClass):
2914         * DumpRenderTree/AccessibilityUIElement.h:
2915         (AccessibilityUIElement):
2916         * DumpRenderTree/blackberry/AccessibilityUIElementBlackBerry.cpp:
2917         (AccessibilityUIElement::isPressActionSupported):
2918         (AccessibilityUIElement::isIncrementActionSupported):
2919         (AccessibilityUIElement::isDecrementActionSupported):
2920         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
2921         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
2922         (WebTestRunner::AccessibilityUIElement::isPressActionSupportedCallback):
2923         (WebTestRunner):
2924         (WebTestRunner::AccessibilityUIElement::isIncrementActionSupportedCallback):
2925         (WebTestRunner::AccessibilityUIElement::isDecrementActionSupportedCallback):
2926         (WebTestRunner::AccessibilityUIElement::incrementCallback):
2927         (WebTestRunner::AccessibilityUIElement::decrementCallback):
2928         (WebTestRunner::AccessibilityUIElement::pressCallback):
2929         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
2930         (AccessibilityUIElement):
2931         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
2932         (AccessibilityUIElement::isPressActionSupported):
2933         (AccessibilityUIElement::isIncrementActionSupported):
2934         (AccessibilityUIElement::isDecrementActionSupported):
2935         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
2936         (AccessibilityUIElement::isPressActionSupported):
2937         (AccessibilityUIElement::isIncrementActionSupported):
2938         (AccessibilityUIElement::isDecrementActionSupported):
2939         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
2940         (AccessibilityUIElement::isPressActionSupported):
2941         (AccessibilityUIElement::isIncrementActionSupported):
2942         (AccessibilityUIElement::isDecrementActionSupported):
2943         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.cpp:
2944         (WTR::AccessibilityUIElement::isPressActionSupported):
2945         (WTR::AccessibilityUIElement::isIncrementActionSupported):
2946         (WTR::AccessibilityUIElement::isDecrementActionSupported):
2947         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.h:
2948         (AccessibilityUIElement):
2949         * WebKitTestRunner/InjectedBundle/Bindings/AccessibilityUIElement.idl:
2950         * WebKitTestRunner/InjectedBundle/mac/AccessibilityUIElementMac.mm:
2951         (WTR::AccessibilityUIElement::isPressActionSupported):
2952         (WTR):
2953         (WTR::AccessibilityUIElement::isIncrementActionSupported):
2954         (WTR::AccessibilityUIElement::isDecrementActionSupported):
2955
2956 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2957
2958         nrwt: remove result_summary.total_tests_by_expectation
2959         https://bugs.webkit.org/show_bug.cgi?id=104756
2960
2961         Reviewed by Eric Seidel.
2962
2963         This field wasn't being used by anything.
2964
2965         * Scripts/webkitpy/layout_tests/models/result_summary.py:
2966         (ResultSummary.__init__):
2967
2968 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2969
2970         garden-o-matic doesn't know about reftests
2971         https://bugs.webkit.org/show_bug.cgi?id=101976
2972
2973         Reviewed by Ojan Vafai.
2974
2975         This is a first attempt at fixing this, sort of a "stop-the-bleeding"
2976         fix. We will now look at the list of tests we're asked to rebaseline,
2977         and log which ones are reftests into the status area, and rebaseline
2978         the rest.
2979
2980         Longer-term, we should display feedback in the results area based on
2981         which tests we're looking at. If the test list contains no reftests,
2982         we should behave as today; if the list is all reftest, we should
2983         probably disable the button, and if the list is a mix, we should
2984         probably display an alert next to the button or something. It would
2985         also be nice to annotate which tests were reftests, maybe with a
2986         "(reftest)" next to the test name or something.
2987
2988         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
2989         (.):
2990         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
2991         (.):
2992         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
2993
2994 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2995
2996         REGRESSION(r133380): new tests without expected file reported as failing tests on the bots
2997         https://bugs.webkit.org/show_bug.cgi?id=101469
2998
2999         Reviewed by Ryosuke Niwa.
3000
3001         Fix parsing of NRWT's output to identify missing results separately
3002         from other regressions.
3003
3004         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
3005         (RunWebKitTests._parseNewRunWebKitTestsOutput):
3006         * BuildSlaveSupport/build.webkit.org-config/mastercfg_unittest.py:
3007         (MasterCfgTest.test_nrwt_leaks_parsing):
3008         (test_nrwt_missing_results):
3009
3010 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
3011
3012         buildbot: change run-webkit-tests --no-launch-safari to --no-show-results
3013         https://bugs.webkit.org/show_bug.cgi?id=104299
3014
3015         Reviewed by Ryosuke Niwa.
3016
3017         Since --no-show-results is the more generically-named flag that NRWT uses.
3018
3019         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
3020         (RunWebKitTests):
3021
3022 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
3023
3024         remove --use-remote-links-to-tests from buildbot configs (except apple win)
3025         https://bugs.webkit.org/show_bug.cgi?id=104291
3026
3027         Reviewed by Csaba Osztrogonác.
3028
3029         This flag is ignored by NRWT and was just implemented for compatibility.
3030         ORWT still uses it, though.
3031
3032         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
3033         (RunWebKitTests):
3034         (RunWebKitTests.start):
3035
3036 2012-12-11  Anders Carlsson  <andersca@apple.com>
3037
3038         ASSERT running platform/mac/plugins/convert-point.html on the WK2 bots
3039         https://bugs.webkit.org/show_bug.cgi?id=104735
3040
3041         Reviewed by Jon Honeycutt.
3042
3043         Do the testing from within NPP_New instead of NPN_SetWindow. Plug-ins aren't supposed to do any real work inside NPP_SetWindow,
3044         and doing so will more easily trigger the assertion failure. Ideally we should get rid of the initial sync painting pass, but this
3045         gets the test passing again!
3046
3047         * DumpRenderTree/TestNetscapePlugIn/Tests/mac/ConvertPoint.cpp:
3048         (ConvertPoint::NPP_New):
3049
3050 2012-12-11  Julie Parent  <jparent@chromium.org>
3051
3052         Dashboard unit tests should not make xhrs.
3053         https://bugs.webkit.org/show_bug.cgi?id=104598
3054
3055         Reviewed by Dirk Pranke.
3056
3057         The unit test were requesting 25 files that they did not use.
3058         Stub out loader.request to avoid making the xhrs.
3059         Result: order of magnitude speed improvement in test run time (from ~1600ms to 160ms on my machine).
3060         
3061         * TestResultServer/static-dashboards/run-embedded-unittests.html:
3062         * TestResultServer/static-dashboards/run-unittests.html:
3063
3064 2012-12-11  Filip Pizlo  <fpizlo@apple.com>
3065
3066         Profiler should show bytecode dumps as they would have been visible to the JITs, including the profiling data that the JITs would see
3067         https://bugs.webkit.org/show_bug.cgi?id=104647
3068
3069         Reviewed by Oliver Hunt.
3070
3071         Added a "profiling" (or "p") command to show the profiling data that the JITs saw
3072         for each JIT compilation of a code block.
3073         
3074         Also added instruction counts in the "full" display and made the "full" display the
3075         default thing you see.
3076
3077         * Scripts/display-profiler-output:
3078
3079 2012-12-11  Eric Seidel  <eric@webkit.org>
3080
3081         Unreviewed, rolling out r137371.
3082         http://trac.webkit.org/changeset/137371
3083         https://bugs.webkit.org/show_bug.cgi?id=104434
3084
3085         Various scripts are not ready for out/ as the build directory,
3086         this can't work as written.
3087
3088         * Scripts/update-webkit:
3089         * Scripts/webkitdirs.pm:
3090         (buildChromium):
3091
3092 2012-12-11  Zan Dobersek  <zandobersek@gmail.com>
3093
3094         Unreviewed GTK build functionality fix.
3095
3096         Turns out the CSP feature was not enabled, that's why all the tests were failing.
3097
3098         * Scripts/webkitperl/FeatureList.pm:
3099
3100 2012-12-11  Eric Seidel  <eric@webkit.org>
3101
3102         Make Ninja the default build system for build-webkit --chromium
3103         https://bugs.webkit.org/show_bug.cgi?id=104434
3104
3105         Reviewed by Daniel Bates.
3106
3107         As far as I can tell, most Chromium developers are now using Ninja
3108         as their default build system instead of the native Xcode/Visual Studio/Make.
3109         This change makes Ninja the default for developers as well as all Chromium webkit bots.
3110
3111         Developers can pass --no-ninja to get the platform-native buildsystem instead of Ninja.
3112         It's possible we may wish to make this a per-platform default, and have
3113         Ninja only be the default for Mac/Linux.
3114
3115         There was a thread on webkit-dev discussing this change:
3116         http://lists.webkit.org/pipermail/webkit-dev/2012-December/023113.html
3117
3118         * Scripts/update-webkit:
3119
3120 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
3121
3122         nrwt: record the order and run times of all of the tests in a single file
3123         https://bugs.webkit.org/show_bug.cgi?id=104184
3124
3125         Reviewed by Ryosuke Niwa.
3126
3127         Replace the N tests_run*.txt files that we output into
3128         layout-test-results with a single packed JSON trie that contains
3129         the test ordering and run statistics. This can eventually be uploaded
3130         to the TestResultsServer to replace the times.json file we currently
3131         upload as well, and will let us replay the results of a previous
3132         run fairly deterministically.
3133
3134         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
3135         (Worker.__init__):
3136         (Worker.start):
3137         (Worker._run_test):
3138         (Worker.stop):
3139         (Worker._clean_up_after_test):
3140         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3141         (Manager._upload_json_files):
3142         (Manager._copy_results_html_file):
3143         (Manager):
3144         (Manager._stats_trie):
3145         (Manager._stats_trie._worker_number):
3146         * Scripts/webkitpy/layout_tests/models/test_results.py:
3147         (TestResult.__init__):
3148         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3149         (MainTest.test_missing_and_unexpected_results):
3150         (MainTest.test_retrying_and_flaky_tests):
3151         (EndToEndTest.test_end_to_end):
3152         (EndToEndTest.test_reftest_with_two_notrefs):
3153         (RebaselineTest.test_reset_results):
3154         (RebaselineTest.test_missing_results):
3155         (RebaselineTest.test_new_baseline):
3156
3157 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
3158
3159         check-webkit-style complaining about TestExpectations lines a patch didn't modify
3160         https://bugs.webkit.org/show_bug.cgi?id=104687
3161
3162         Reviewed by Ojan Vafai.
3163
3164         Way back in bug 60466 / r86165, we changed the style checker
3165         for TestExpectations to ensure that the entire file linted, not
3166         just the lines modified by a patch, in order to ensure that a
3167         new line didn't cause errors elsewhere in the file (e.g., by
3168         introducing duplicate lines). At the time, errors in the
3169         expectations files were fatal and we didn't have a separate step
3170         on the bots that would lint the files to ensure things were still
3171         okay, so this was probably the right thing to do, but it could
3172         cause some patches to fail the style check even though the patches
3173         themselves were fine.
3174
3175         However, now errors aren't always fatal, and we have a separate build
3176         step to lint the files. So, we should probably not punish the
3177         patches and the style checker should filter out "unrelated" errors.
3178
3179         In theory we should be comparing the output pre- and post-patch
3180         to filter out pre-existing errors, but I don't think the style
3181         checker has that ability.
3182
3183         * Scripts/webkitpy/style/checkers/test_expectations.py:
3184         (TestExpectationsChecker.__init__):
3185         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
3186         (TestExpectationsTestCase.assert_lines_lint):
3187
3188 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
3189
3190         nrwt isn't rejecting unrecognized expectations
3191         https://bugs.webkit.org/show_bug.cgi?id=104701
3192
3193         Reviewed by Ojan Vafai.
3194
3195         NRWT is silently ignoring unrecognized expectations like "Text"
3196         and "Audio" (largely because I haven't fully removed support for
3197         the old syntax and keywords and restructured the code to do the
3198         proper checking). Temporarily change the tokenizer to reject
3199         unrecognized expectations until I can make those changes.
3200
3201         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
3202         (TestExpectationParser._tokenize_line):
3203         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
3204         (ExpectationSyntaxTests.test_unknown_expectation):
3205         (TestExpectationSerializationTests.disabled_test_string_whitespace_stripping):
3206
3207 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
3208
3209         Fix a typo in my previous commit that is causing test failures.
3210
3211         Reviewed by Gustavo Noronha Silva.
3212
3213         There was an extra bit of statement appended after the end of a conditional. Somehow
3214         this compiled and caused test failures.
3215
3216         * DumpRenderTree/gtk/EventSender.cpp:
3217         (continuousMouseScrollByCallback):
3218
3219 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
3220
3221         [GTK] Spurious stderr output from the event sender
3222         https://bugs.webkit.org/show_bug.cgi?id=104678
3223
3224         Reviewed by Gustavo Noronha Silva.
3225
3226         Eliminate some spurious stderr output from the EventSender by using an early return
3227         instead of an assertion that prints a warning to the console.
3228
3229         * DumpRenderTree/gtk/EventSender.cpp:
3230         (continuousMouseScrollByCallback): We don't support continuous scroll events that
3231         move by page, but we should just fail silently instead of printing an error.
3232
3233 2012-12-11  Joone Hur  <joone.hur@intel.com>
3234
3235         [GTK] Add GraphicsLayerActor
3236         https://bugs.webkit.org/show_bug.cgi?id=91940
3237
3238         Reviewed by Gustavo Noronha Silva.
3239
3240         Add GraphicsLayerActor to the list of classes that contain GObjects
3241         in the style checker script so that it does not complain about GObject
3242         coding style.
3243
3244         * Scripts/webkitpy/style/checker.py:
3245
3246 2012-12-11  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
3247
3248         [Qt] remove hack which avoids unused .obj directories all over the place
3249
3250         Reviewed by Simon Hausmann.
3251
3252         The feature is subject to upstreaming in Qt itself in
3253         https://codereview.qt-project.org/#change,42011
3254
3255         * qmake/mkspecs/features/default_post.prf:
3256
3257 2012-12-11  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
3258
3259         [Qt] move SQLITE3SRCDIR assignment to only use site
3260
3261         Reviewed by Simon Hausmann.
3262
3263         Target.pri includes WebCore.pri the almost first thing, so that is
3264         covered as well.
3265
3266         * qmake/mkspecs/features/default_pre.prf:
3267
3268 2012-12-11  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
3269
3270         [Qt] delete dead include paths
3271
3272         Reviewed by Simon Hausmann.
3273
3274         followup to https://bugs.webkit.org/show_bug.cgi?id=93446
3275
3276         * DumpRenderTree/qt/DumpRenderTree.pro:
3277
3278 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
3279
3280         [GTK] Implement some missing FrameLoaderClient message in DRT
3281         https://bugs.webkit.org/show_bug.cgi?id=104444
3282
3283         Reviewed by Gustavo Noronha Silva.
3284
3285         Install a FrameLoadEvent callback that prints the necessary output for some
3286         missing FrameLoader events.
3287
3288         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3289         (frameLoadEventCallback): Added.
3290         (createWebView): Install the FrameLoader event callback.
3291
3292 2012-12-11  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
3293
3294         [Qt] change setup of QMAKE_MOC
3295
3296         Reviewed by Simon Hausmann.
3297
3298         don't load(moc) explicitly - turns out that this breaks
3299         debug_and_release builds (the feature does not like being loaded before
3300         default_post has been loaded).
3301         so instead just do a minimal call to find the moc executable.
3302
3303         as it's in fact only two files which need it, so de-centralize this call
3304         for the sake of saving some cpu cycles spent in qmake.
3305
3306         * WebKitTestRunner/InjectedBundle/DerivedSources.pri:
3307         * qmake/mkspecs/features/default_pre.prf:
3308
3309 2012-12-11  Jocelyn Turcotte  <jocelyn.turcotte@digia.com>
3310
3311         [Qt] Fix the linking of debug builds on Mac with force_static_libs_as_shared
3312         https://bugs.webkit.org/show_bug.cgi?id=104559
3313
3314         Reviewed by Tor Arne Vestbø.
3315
3316         "contains(libraries, $$TARGET): return(true)" fails for intermediate libraries since
3317         at this point TARGET may contain the "_debug" suffix.
3318
3319         Fix it by moving the TARGET adjustment later in the file.
3320
3321         * qmake/mkspecs/features/default_post.prf:
3322
3323 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
3324
3325         [Coverity] [GTK] Remove some redundant null checks in DumpRenderTree
3326         https://bugs.webkit.org/show_bug.cgi?id=104570
3327
3328         Reviewed by Daniel Bates.
3329
3330         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3331         (descriptionSuitableForTestResult): Remove a redundant null check.
3332
3333 2012-12-11  Alberto Garcia  <agarcia@igalia.com>
3334
3335         Remove unused GErrors
3336         https://bugs.webkit.org/show_bug.cgi?id=104657
3337
3338         Reviewed by Martin Robinson.
3339
3340         Replace unused GErrors with null pointers.
3341
3342         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3343         (initializeFonts):
3344         * DumpRenderTree/gtk/EditingCallbacks.cpp:
3345         (dumpRange):
3346         * WebKitTestRunner/InjectedBundle/gtk/ActivateFontsGtk.cpp:
3347         (WTR::inititializeFontConfigSetting):
3348
3349 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
3350
3351         [GTK] Bump the jhbuild libsoup version to fix an authentication issue
3352         https://bugs.webkit.org/show_bug.cgi?id=104448
3353
3354         Reviewed by Gustavo Noronha Silva.
3355
3356         Bump the libsoup version to one from the git repository. This requires updating
3357         GLib as well which in turn requires adjusting the CFLAGS we pass when compiling
3358         the scanobj step of the gtkdoc generation.
3359
3360         * gtk/generate-gtkdoc:
3361         * gtk/jhbuild.modules:
3362
3363 2012-12-11  Zan Dobersek  <zandobersek@gmail.com>
3364
3365         Unreviewed functionality fix for the GTK port.
3366
3367         Enable the MicroData feature support for the GTK port in the FeatureList module.
3368         This feature listing now again affects which features are enabled and which disabled
3369         on the development builds of the GTK port.
3370
3371         * Scripts/webkitperl/FeatureList.pm:
3372
3373 2012-12-11  Jinwoo Song  <jinwoo7.song@samsung.com>
3374
3375         [EFL][WK2] MiniBrowser is not built
3376         https://bugs.webkit.org/show_bug.cgi?id=104629
3377
3378         Reviewed by Gyuyoung Kim.
3379
3380         As we are finding an elementary package in Tools/MiniBrowser/efl/CMakeList.txt since r137203,
3381         we need to check if the package is found in the same place not in the Tools/CMakeList.txt.
3382
3383         * CMakeLists.txt:
3384         * MiniBrowser/efl/CMakeLists.txt:
3385
3386 2012-12-11  Xabier Rodriguez Calvar  <calvaris@igalia.com>
3387
3388         [GTK][jhbuild] Switch to GStreamer 1.0 build
3389         https://bugs.webkit.org/show_bug.cgi?id=91727
3390
3391         Reviewed by Philippe Normand.
3392
3393         Switch build-webkit --gtk to GStreamer 1.0 support and build the
3394         necessary GStreamer git modules from JHBuild.
3395
3396         * gtk/jhbuild.modules: Added GStreamer 1.0.3 build support.
3397
3398 2012-12-11  Zan Dobersek  <zandobersek@gmail.com>
3399
3400         [GTK] Feature enabling/disabling should be possible through build-webkit
3401         https://bugs.webkit.org/show_bug.cgi?id=99271
3402
3403         Reviewed by Gustavo Noronha Silva.
3404
3405         Refactor the code in webkitdirs.pm that builds an autotools project to generate
3406         autogen.sh arguments, make arguments and installation prefix from the passed-in
3407         parameters rather than generating all of that in build-webkit. The autogen.sh
3408         arguments now contain only enable/disable flags for options that are actually
3409         configurable in configure.ac, the flag value actually reflecting whether the feature
3410         is enabled or disabled in the feature list.
3411
3412         Other features are overridable through modifying the GNUmakefile.features.am file.
3413         All these features are now stored in the build directory and upon change trigger
3414         a rerun of the autogen.sh script, pretty much like the autogen.sh arguments do.
3415
3416         The override-feature-defines script is called by autogen.sh, before GNUmakefile.in is
3417         generated when calling autoreconf. Its task is to copy the GNUmakefile.features.am.in
3418         into GNUmakefile.features.am and modify it if there's a text file in the build directory
3419         that contains all the feature defines the build-webkit script has written. If the build
3420         is not done through build-webkit (and there's no file in build directory listing all the
3421         feature defines) the script does not advance further from copying.
3422
3423         This approach is taken to overcome the rigidness of the automake system as it's impossible
3424         to effectively generate and use GNUmakefile.features.am or even GNUmakefile.features file
3425         after the autoreconf command execution in autogen.sh.
3426
3427         * Scripts/build-jsc:
3428         (buildMyProject):
3429         * Scripts/build-webkit:
3430         * Scripts/webkitdirs.pm:
3431         (runAutogenForAutotoolsProjectIfNecessary):
3432         (mustReRunAutogen):
3433         (buildAutotoolsProject):
3434         (buildGtkProject):
3435         * Scripts/webkitperl/FeatureList.pm: Update the features that are currently enabled in
3436         Source/WebCore/GNUmakefile.features.am.in but aren't in the feature list.
3437         * Scripts/webkitpy/style/checker.py: Source/WebCore/GNUmakefile.features.am.in is recognized
3438         as a text file and tabulation errors are reported. Skip the file to suppress them.
3439         * gtk/override-feature-defines: Added.
3440         (copy_feature_defines_makefile):
3441         (adjust_feature_defines_makefile):
3442         (adjust_feature_defines_makefile.override_feature_define):
3443         (override_feature_defines):
3444
3445 2012-12-10  Yury Semikhatsky  <yurys@chromium.org>
3446
3447         Memory instrumentation: make sure each edge is reported only once
3448         https://bugs.webkit.org/show_bug.cgi?id=104630
3449
3450         Reviewed by Pavel Feldman.
3451
3452         Test that outgoing edges are reported only once if we come to the same
3453         object twice: first by an address of a base class and then by a real address
3454         of the object.
3455
3456         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
3457
3458 2012-12-10  Dana Jansens  <danakj@chromium.org>
3459
3460         [chromium] Remove WebCompositorOutputSurface implementations from DRT
3461         https://bugs.webkit.org/show_bug.cgi?id=104046
3462
3463         Reviewed by Darin Fisher.
3464
3465         The classes are now provided through WebCompositorSupport.
3466
3467         * DumpRenderTree/DumpRenderTree.gypi:
3468         * DumpRenderTree/chromium/WebViewHost.cpp:
3469         (WebViewHost::createOutputSurface):
3470         * DumpRenderTree/chromium/WebViewHostOutputSurface.cpp: Removed.
3471         * DumpRenderTree/chromium/WebViewHostOutputSurface.h: Removed.
3472         * DumpRenderTree/chromium/WebViewHostSoftwareOutputDevice.cpp: Removed.
3473         * DumpRenderTree/chromium/WebViewHostSoftwareOutputDevice.h: Removed.
3474
3475 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3476
3477         change --no-launch-safari to --no-show-results in scripts
3478         https://bugs.webkit.org/show_bug.cgi?id=104298
3479
3480         Reviewed by Eric Seidel.
3481
3482         Since Safari only exists on the Apple ports, NRWT calls the
3483         command line flag --[no-]show-results instead. This updates
3484         the scripts to use the new flag (but not the buildbot config yet).
3485
3486         * Scripts/generate-coverage-data:
3487         * Scripts/webkit-tools-completion.sh:
3488         * Scripts/webkitpy/tool/steps/runtests.py:
3489         (RunTests.run):
3490         * Scripts/webkitpy/tool/steps/runtests_unittest.py:
3491
3492 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3493
3494         nrwt: fix mapping of AUDIO onto FAIL when printing unexpected results
3495         https://bugs.webkit.org/show_bug.cgi?id=104597
3496
3497         Reviewed by Ojan Vafai.
3498
3499         This was broken in the recent printer refactoring.
3500
3501         * Scripts/webkitpy/layout_tests/views/buildbot_results.py:
3502         (BuildBotPrinter.print_unexpected_results.add_result.is_expected):
3503         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
3504         (BuildBotPrinterTests.get_result):
3505         (BuildBotPrinterTests.test_print_unexpected_results):
3506         (BuildBotPrinterTests.test_print_unexpected_results.get_unexpected_results):
3507         (BuildBotPrinterTests):
3508
3509 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3510
3511         old-run-webkit-tests: add --show-results properly
3512         https://bugs.webkit.org/show_bug.cgi?id=104608
3513
3514         Reviewed by Eric Seidel.
3515
3516         Rework r137219 / bug 104601 to use the correct perl idiom.
3517
3518         * Scripts/old-run-webkit-tests:
3519
3520 2012-12-10  Aaron Colwell  <acolwell@chromium.org>
3521
3522         Unreviewed. Adding Aaron Colwell to the contributor list.
3523
3524         * Scripts/webkitpy/common/config/committers.py:
3525
3526 2012-12-10  Anders Carlsson  <andersca@apple.com>
3527
3528         Remove WebKit2 tests on Windows
3529         https://bugs.webkit.org/show_bug.cgi?id=104607
3530
3531         Reviewed by Tim Horton.
3532
3533         Stop building WebKit2 tests and remove associated files.
3534
3535         * TestWebKitAPI/win/InjectedBundleControllerWin.cpp: Removed.
3536         * TestWebKitAPI/win/PlatformUtilitiesWin.cpp: Removed.
3537         * TestWebKitAPI/win/PlatformWebViewWin.cpp: Removed.
3538         * TestWebKitAPI/win/TestWebKitAPI.sln: Removed.
3539         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
3540         * TestWebKitAPI/win/TestWebKitAPIGenerated.vcproj: Removed.
3541         * TestWebKitAPI/win/TestWebKitAPIGeneratedCommon.vsprops: Removed.
3542         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj: Removed.
3543         * TestWebKitAPI/win/WindowMessageObserver.h: Removed.
3544         * TestWebKitAPI/win/copy-resources.cmd: Removed.
3545
3546 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3547
3548         old-run-webkit-tests: support --no-show-results for compatibility w/ NRWT
3549         https://bugs.webkit.org/show_bug.cgi?id=104601
3550
3551         Reviewed by Eric Siedel.
3552
3553         This flag is being added so we can move all of the calling scripts
3554         over to using --no-show-results instead of --no-launch-safari.
3555
3556         * Scripts/old-run-webkit-tests:
3557
3558 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3559
3560         nrwt: use the default logic for skipping platform-specific tests for chromium
3561         https://bugs.webkit.org/show_bug.cgi?id=104403
3562
3563         Reviewed by Stephen White.
3564
3565         I thought I had fixed this a while ago, but the chromium port was
3566         using custom logic to figure out which platform/* tests to run (or
3567         skip) and as a result we were getting some odd behavior (like running
3568         tests in chromium-linux on chromium-mac, but only when there was
3569         an expectation listed for the test).
3570
3571         Ports are supposed to only run the tests found under their
3572         baseline_search_path() by default.
3573
3574         Not writing a test for this; theoretically there shouldn't been a
3575         test checking that the behavior for Chromium was *different* than
3576         the other ports before.
3577
3578         * Scripts/webkitpy/layout_tests/port/chromium.py:
3579         (ChromiumPort._missing_symbol_to_skipped_tests):
3580
3581 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3582
3583         fix regression in test-webkitpy after r137189
3584         https://bugs.webkit.org/show_bug.cgi?id=104589
3585
3586         Reviewed by Eric Seidel.
3587
3588         Need to specify full port names.
3589
3590         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3591         (MainTest.test_build_check):
3592
3593 2012-12-10  Laszlo Gombos  <l.gombos@samsung.com>
3594
3595         [EFL] Change the minimum required EFL version to 1.6
3596         https://bugs.webkit.org/show_bug.cgi?id=104431
3597
3598         Reviewed by Kenneth Rohde Christiansen.
3599
3600         Change the minimum required EFL version to 1.6 from 1.7.
3601         The elementary EFL package is only required to build MiniBrowser,
3602         so I moved the required only to the Minibrowser CMake file.
3603
3604         * MiniBrowser/efl/CMakeLists.txt: In addition removed
3605         ENABLE_GLIB_SUPPORT as it is obsolete and it was always set to be on.
3606
3607 2012-12-10  Dean Jackson  <dino@apple.com>
3608
3609         Unreviewed. Adding Antoine to the contributor list so his name autocompletes in bugzilla.
3610
3611         * Scripts/webkitpy/common/config/committers.py:
3612
3613 2012-12-10  Zan Dobersek  <zandobersek@gmail.com>
3614
3615         REGRESSION (r118735): svg/dom/complex-svgView-specification.html, svg/dom/SVGViewSpec.html, svg/dom/viewspec-parser.html failing on GTK Linux 64-bit Release
3616         https://bugs.webkit.org/show_bug.cgi?id=87734
3617
3618         Reviewed by Martin Robinson.
3619
3620         When willSendRequest should return null in testing, cancel the SoupMessage
3621         instead of setting the request's URI to about:blank.
3622
3623         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3624         (willSendRequestCallback):
3625
3626 2012-12-10  Jochen Eisinger  <jochen@chromium.org>
3627
3628         [chromium] Move the test implementation of WebKit::WebPlugin to the TestRunner library
3629         https://bugs.webkit.org/show_bug.cgi?id=104537
3630
3631         Reviewed by Tony Chang.
3632
3633         In order to hide implementation details from the public TestRunner API,
3634         I moved the plugin to WebTestPluginImpl and added a static create
3635         method to WebTestPlugin. Also, I renamed the plugin from TestWebPlugin
3636         to WebTestPlugin to be more in line with other classes in the
3637         TestRunner API.
3638
3639         * DumpRenderTree/DumpRenderTree.gypi:
3640         * DumpRenderTree/chromium/TestRunner/public/WebTestPlugin.h: Added.
3641         (WebTestRunner):
3642         (WebTestPlugin):
3643         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp: Renamed from Tools/DumpRenderTree/chromium/TestWebPlugin.cpp.
3644         (WebTestRunner: