b67312fbe7f9b35425714075162110ffd5b4795c
[WebKit-https.git] / Tools / ChangeLog
1 2013-01-03  Tony Chang  <tony@chromium.org>
2
3         Generate internal.settings from Settings.in
4         https://bugs.webkit.org/show_bug.cgi?id=104740
5
6         Reviewed by Adam Barth.
7
8         * GNUmakefile.am: Compile new generated files on GTK+.
9         * qmake/mkspecs/features/default_post.prf: Qt: Make it so an action
10         with both extra_sources and add_output_to_sources=false will still
11         compile the files in extra_sources. This makes it possible to compile
12         InternalSettingsGenerated.cpp and not be confused by
13         InternalSettingsGenerated.idl.
14
15 2013-01-03  Zan Dobersek  <zandobersek@gmail.com>
16
17         [style] WebIDL-reflecting uppercase enums reported as style violation
18         https://bugs.webkit.org/show_bug.cgi?id=105832
19
20         Reviewed by Tony Chang.
21
22         Recognize the enum as WebIDL-reflecting one if there's a comment present that
23         indicates this. In enums spanning over multiple lines, the comment is expected
24         on the first line. In enums placed in one line, the comment is expected at the
25         end of that line.
26
27         WebIDL-reflecting enum values are required to be written in uppercase and as such
28         violate the common rule of these values being written in InterCaps. Annotating
29         enums this way makes it possible to avoid false style violation reports.
30
31         The comment is of form '// Web(?:Kit)?IDL enum'.
32
33         * Scripts/webkitpy/style/checkers/cpp.py:
34         (_EnumState.__init__):
35         (_EnumState.process_clean_line):
36         (check_enum_casing):
37         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
38         (NoNonVirtualDestructorsTest.test_enum_casing):
39
40 2013-01-03  Tim 'mithro' Ansell  <mithro@mithis.com>
41
42         Adding favicon.ico to the garden-o-matic.
43         https://bugs.webkit.org/show_bug.cgi?id=105982
44
45         Reviewed by Adam Barth.
46
47         * Scripts/webkitpy/tool/servers/gardeningserver.py:
48         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/favicon.ico:
49
50 2013-01-03  Jonathan Liu  <net147@gmail.com>
51
52         [Qt] Fix library linking order
53         https://bugs.webkit.org/show_bug.cgi?id=105818
54
55         Some Qt libraries are linked before the private static libraries
56         that depend on them which results in QtWebKit failing to link.
57
58         Copy Qt module dependencies from QT to QT_PRIVATE when creating
59         module to ensure the Qt libraries are linked after the private
60         static libraries that depend on them.
61
62         Reviewed by Tor Arne Vestbø.
63
64         * qmake/mkspecs/features/default_post.prf:
65
66 2013-01-02  Byungwoo Lee  <bw80.lee@samsung.com>
67
68         [EFL][WK2] Fix tooltip bugs in MiniBrowser.
69         https://bugs.webkit.org/show_bug.cgi?id=102209
70
71         Reviewed by Gyuyoung Kim.
72
73         Set tooltip on the elm_window instead of ewk_view, and add delay to
74         show the tooltip window.
75
76         Setting tooltip on the ewk_view makes elementary error message because
77         ewk_view is not an elementary widget. To remove the error message,
78         tooltip should be set on the elm_window.
79
80         Most browsers show the tooltip after a mouse pointer stays on an
81         element object that has tooltip text.
82         But elm_tooltip widget itself doesn't have delay or handle mouse
83         events to show/hide.
84
85         So additional codes are added to show/hide tooltip widget along with
86         mouse events.
87           1) When a mouse pointer stays for a second on an element which has
88              tooltip text, then tooltip widget is displayed.
89           2) When a mouse pointer is moved or out of ewk_view, hide the tooltip
90              widget.
91
92         * MiniBrowser/efl/main.c:
93         (_Tooltip_Information):
94         (_Browser_Window):
95         (on_tooltip_show):
96         (window_tooltip_hide):
97         (window_tooltip_update):
98         (on_mouse_in):
99         (on_mouse_move):
100         (on_mouse_out):
101         (window_free):
102         (on_tooltip_text_set):
103         (on_tooltip_text_unset):
104         (window_create):
105
106 2013-01-02  Tony Chang  <tony@chromium.org>
107
108         Unreviewed, rolling out r138661.
109         http://trac.webkit.org/changeset/138661
110         https://bugs.webkit.org/show_bug.cgi?id=104740
111
112         Compile problems on EFL
113
114         * GNUmakefile.am:
115         * qmake/mkspecs/features/default_post.prf:
116
117 2013-01-02  Tony Chang  <tony@chromium.org>
118
119         Generate internal.settings from Settings.in
120         https://bugs.webkit.org/show_bug.cgi?id=104740
121
122         Reviewed by Adam Barth.
123
124         * GNUmakefile.am: Compile new generated files on GTK+.
125         * qmake/mkspecs/features/default_post.prf: Qt: Make it so an action
126         with both extra_sources and add_output_to_sources=false will still
127         compile the files in extra_sources. This makes it possible to compile
128         InternalSettingsGenerated.cpp and not be confused by
129         InternalSettingsGenerated.idl.
130
131 2013-01-02  Adam Barth  <abarth@webkit.org>
132
133         Remove old-run-webkit-tests support from webkitpy
134         https://bugs.webkit.org/show_bug.cgi?id=105619
135
136         Reviewed by Eric Seidel.
137
138         This patch removes support for webkitpy parsing the
139         old-run-webkit-tests results format. There doesn't seem to be any
140         reason to keep this code around anymore now that most bots have
141         switched to new-run-webkit-tests.
142
143         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
144         (Builder.fetch_layout_test_results):
145         * Scripts/webkitpy/common/net/layouttestresults.py:
146         (path_for_layout_test):
147         (LayoutTestResults.results_from_string):
148         (LayoutTestResults):
149         * Scripts/webkitpy/common/net/layouttestresults_unittest.py:
150         (LayoutTestResultsTest.test_results_from_string):
151         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
152         (test_flaky_test_failure):
153         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py:
154         (LayoutTestResultsReaderTest.test_missing_layout_test_results):
155         * Scripts/webkitpy/tool/bot/patchanalysistask.py:
156         (PatchAnalysisTask._test_patch):
157         * Scripts/webkitpy/tool/commands/queuestest.py:
158         (QueuesTest.assert_queue_outputs):
159
160 2013-01-02  Heikki Paajanen  <heikki.paajanen@palm.com>
161
162         [Qt][WK2] Add simple UI to find text from page
163         https://bugs.webkit.org/show_bug.cgi?id=96481
164
165         Reviewed by Jocelyn Turcotte.
166
167         Find bar that slides between navigation bar and WebView.
168
169         * MiniBrowser/qt/BrowserWindow.cpp:
170         (BrowserWindow::toggleFind):
171         * MiniBrowser/qt/BrowserWindow.h:
172         (BrowserWindow):
173         * MiniBrowser/qt/MiniBrowser.qrc:
174         * MiniBrowser/qt/MiniBrowserApplication.cpp:
175         (MiniBrowserApplication::notify):
176         * MiniBrowser/qt/icons/find.png: Added.
177         * MiniBrowser/qt/qml/BrowserWindow.qml:
178
179 2013-01-02  Christophe Dumez  <christophe.dumez@intel.com>
180
181         Regression(r138603): Caused 2 webkitpy tests to fail
182         https://bugs.webkit.org/show_bug.cgi?id=105913
183
184         Reviewed by Kentaro Hara.
185
186         Fix 2 webkitpy tests that started to fail after
187         r138603.
188
189         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
190         (IRCCommandTest._sheriff_test_data_url): Make sure the
191         returned test file URL is valid no matter what the
192         current working directory is.
193         * Scripts/webkitpy/tool/bot/ircbot_unittest.py:
194         (IRCBotTest.test_help): Add "sheriffs" command to the
195         list of available commands.
196
197 2013-01-01  Christophe Dumez  <christophe.dumez@intel.com>
198
199         [WK2] plugins/clicking-missing-plugin-fires-delegate.html is failing
200         https://bugs.webkit.org/show_bug.cgi?id=105897
201
202         Reviewed by Alexey Proskuryakov.
203
204         Dump in WebKitTestRunner the message that is expected by layout tests
205         when the missing plugin button is clicked.
206
207         * WebKitTestRunner/TestController.cpp:
208         (WTR::TestController::createWebViewWithOptions):
209         (WTR::TestController::unavailablePluginButtonClicked):
210         (WTR):
211         * WebKitTestRunner/TestController.h:
212         (TestController):
213
214 2013-01-01  KwangYong Choi  <ky0.choi@samsung.com>
215
216         [EFL] Enable MHTML feature
217         https://bugs.webkit.org/show_bug.cgi?id=105815
218
219         Reviewed by Laszlo Gombos.
220
221         Enable MHTML feature for EFL port by default.
222
223         * Scripts/webkitperl/FeatureList.pm:
224
225 2013-01-01  Alan Cutter  <alancutter@chromium.org>
226
227         sheriff-bot should know who the gardeners/sheriffs are
228         https://bugs.webkit.org/show_bug.cgi?id=105698
229
230         Reviewed by Eric Seidel.
231
232         Added a "sheriffs" command to sheriffbot to fetch the current Chromium Webkit sheriffs.
233
234         * Scripts/webkitpy/tool/bot/irc_command.py:
235         (Current):
236         (Current._retrieve_webkit_sheriffs):
237         (Current.execute):
238         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
239         (IRCCommandTest.test_current):
240         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_0.js: Added.
241         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_1.js: Added.
242         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_2.js: Added.
243         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_malformed.js: Added.
244         * Scripts/webkitpy/tool/bot/testdata/webkit_sheriff_zero.js: Added.
245
246 2012-12-31  Carlos Garcia Campos  <cgarcia@igalia.com>
247
248         [GTK] Move ImageDiff to the Tools directory root
249         https://bugs.webkit.org/show_bug.cgi?id=105421
250
251         Reviewed by Kenneth Rohde Christiansen.
252
253         * DumpRenderTree/gtk/GNUmakefile.ImageDiff.am: Removed.
254         * GNUmakefile.am:
255         * ImageDiff/gtk/ImageDiff.cpp: Renamed from Tools/DumpRenderTree/gtk/ImageDiff.cpp.
256
257 2012-12-30  Mike West  <mkwst@chromium.org>
258
259         Add myself to watchlists for CSP and console logs.
260         https://bugs.webkit.org/show_bug.cgi?id=105874
261
262         Reviewed by Dirk Pranke.
263
264         Because I really, really, really need more email.
265
266         * Scripts/webkitpy/common/config/committers.py:
267             Add mkwst+watchlist@chromium.org as a contributor.
268         * Scripts/webkitpy/common/config/watchlist:
269             Create watchlist categories for 1) the various mechanisms used to
270             generate console messages, and 2) CSP usage and files.
271
272 2012-12-30  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
273
274         [EFL] Enable TEMPLATE_ELEMENT feature
275         https://bugs.webkit.org/show_bug.cgi?id=105865
276         
277         Reviewed by Laszlo Gombos.
278
279         * Scripts/webkitperl/FeatureList.pm: Enable template element feature for EFL port.
280
281 2012-12-30  Zan Dobersek  <zandobersek@gmail.com>
282
283         Unreviewed GTK gardening of the unit tests.
284
285         Skip the TestInspectorServer unit test that's timing out completely.
286
287         * Scripts/run-gtk-tests:
288         (TestRunner):
289
290 2012-12-29  Ilya Tikhonovsky  <loislo@chromium.org>
291
292         Web Inspector: Native Memory Instrumentation plugin: do not generate 'not instrumented' warning for instrumented mutable members.
293         https://bugs.webkit.org/show_bug.cgi?id=105855
294
295         Reviewed by Vsevolod Vlasov.
296
297         Extract MemberExpr from ImplicitCastExpr. It happens when we instrument a mutable member because
298         addMember expects const T& and the mutable ref to member implicitly converts into const ref to member.
299
300         * clang/ReportMemoryUsagePlugin/ReportMemoryUsage.cpp:
301         (ReportMemoryUsageVisitor):
302         (clang::AddMemberCallVisitor::VisitCallExpr):
303         (clang::ReportMemoryUsageVisitor::VisitCXXMethodDecl):
304         (clang::ReportMemoryUsageVisitor::checkMembersCoverage):
305
306 2012-12-29  Ilya Tikhonovsky  <loislo@chromium.org>
307
308         Web Inspector: Native Memory Instrumentation plugin: move function bodies out of class declarations.
309         https://bugs.webkit.org/show_bug.cgi?id=105852
310
311         Reviewed by Alexander Pavlov.
312
313         * clang/ReportMemoryUsagePlugin/ReportMemoryUsage.cpp:
314         (clang):
315         (AddMemberCallVisitor):
316         (ReportMemoryUsageVisitor):
317         (clang::ReportMemoryUsageAction::ParseArgs):
318         (clang::AddMemberCallVisitor::VisitCallExpr):
319         (clang::ReportMemoryUsageVisitor::VisitCXXMethodDecl):
320         (clang::ReportMemoryUsageVisitor::emitWarning):
321         (clang::ReportMemoryUsageVisitor::findInstrumentationMethod):
322         (clang::ReportMemoryUsageVisitor::needsToBeInstrumented):
323         (clang::ReportMemoryUsageVisitor::CheckMembersCoverage):
324
325 2012-12-29  Zan Dobersek  <zandobersek@gmail.com>
326
327         [webkitpy] Omit webkitpy/thirdparty/BeautifulSoup.py from code coverage checking
328         https://bugs.webkit.org/show_bug.cgi?id=105835
329
330         Reviewed by Dirk Pranke.
331
332         The BeautifulSoup module is a third-party import and as such we don't have much control
333         over its testing. Because of that it should be omitted from code coverage checking.
334
335         * Scripts/webkitpy/test/main.py:
336         (Tester._run_tests):
337
338 2012-12-28  Martin Robinson  <mrobinson@igalia.com>
339
340         [GTK][WK2] Add support for IME Composition
341         https://bugs.webkit.org/show_bug.cgi?id=65093
342
343         Reviewed by Carlos Garcia Campos.
344
345         Update unit tests for GtkInputMethodFilter to note whether or not an event
346         was a fake event synthesized for composition purposes. This is now used
347         for WebKit2.
348
349         * TestWebKitAPI/Tests/gtk/InputMethodFilter.cpp:
350         (TestWebKitAPI::TestInputMethodFilter::sendSimpleKeyEvent):
351         (TestWebKitAPI::TestInputMethodFilter::sendKeyEventWithCompositionResults):
352         (TestWebKitAPI::TEST):
353         * gtk/generate-gtkdoc:
354         (get_webkit2_options): Skip new files in the WebKit2 source directory for documentation
355         generation.
356
357 2012-12-27  Ilya Tikhonovsky  <loislo@chromium.org>
358
359         Web Inspector: Native Memory Instrumentation. Update clang plugin according to the current state of memory instrumentation code.
360         https://bugs.webkit.org/show_bug.cgi?id=105800
361
362         Reviewed by Yury Semikhatsky.
363
364         * clang/ReportMemoryUsagePlugin/ReportMemoryUsage.cpp:
365         (clang::ReportMemoryUsageConsumer::ReportMemoryUsageConsumer):
366         (clang):
367
368 2012-12-27  Zan Dobersek  <zandobersek@gmail.com>
369
370         Create a GTK build system watchlist and add myself to it
371         https://bugs.webkit.org/show_bug.cgi?id=105769
372
373         Reviewed by Dirk Pranke.
374
375         Add a GTK build system watchlist, watching the configure.ac, GNUmakefile.am files
376         and the input GNUmakefile.features.am.in file. The GNUmakefile.list.am files are
377         omitted on purpose due to relatively high change rate they are exposed to.
378
379         The new watchlist would be of use to me as the GTK build system recently went
380         through a series of changes that included a changed policy on adding configuration
381         flags to configure.ac and subsequent additions to GNUmakefile.am files. A PSA on the
382         mailing list about this change will be published later, but I'd also like to be
383         notified of any pending changes to these files just so I can check the changes are
384         in compliance with the new policy.
385
386         * Scripts/webkitpy/common/config/watchlist:
387
388 2012-12-26  Ilya Tikhonovsky  <loislo@chromium.org>
389
390         Unreviewed. Replace find("Tools") with rfind("Tools") because base dir of WebKit may have Tools word.
391         As example ..../DevTools/src/third_party/WebKit
392
393         * Scripts/webkitpy/common/webkit_finder.py:
394         (WebKitFinder.webkit_base):
395
396 2012-12-25  Christophe Dumez  <christophe.dumez@intel.com>
397
398         [EFL][WK2] Refactor snapshot taking code
399         https://bugs.webkit.org/show_bug.cgi?id=105687
400
401         Reviewed by Kenneth Rohde Christiansen.
402
403         Call WKViewCreateSnapshot() instead of WKViewGetSnapshot()
404         since the function was renamed to follow naming style.
405
406         * WebKitTestRunner/efl/PlatformWebViewEfl.cpp:
407         (WTR::PlatformWebView::windowSnapshotImage):
408
409 2012-12-24  Ilya Tikhonovsky  <loislo@chromium.org>
410
411         Web Inspector: Native Memory Instrumentation: propagate member type as edge type to the serialized heap graph.
412         https://bugs.webkit.org/show_bug.cgi?id=105725
413
414         Reviewed by Yury Semikhatsky.
415
416         MemberType value names were adjusted according to Style Guide.
417         Existing tests were extended with link type validation.
418
419         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
420
421 2012-12-24  Xiaobo Wang  <xbwang@torchmobile.com.cn>
422
423         [BlackBerry] DRT - Crashed when running pixels tests
424         https://bugs.webkit.org/show_bug.cgi?id=105711
425
426         RIM PR 270003
427
428         Reviewed by George Staikos.
429
430         Calling window->post() here will result in buffer swapping which is
431         only allowed in UI thread now, and I found we don't need to call it
432         for GL renderer.
433
434         * DumpRenderTree/blackberry/PixelDumpSupportBlackBerry.cpp:
435         (createBitmapContextFromWebView):
436
437 2012-12-22  Dan Winship  <danw@gnome.org>
438
439         Add a patch to fix libxml2 2.9.0 regression
440         https://bugs.webkit.org/show_bug.cgi?id=104494
441
442         Reviewed by Martin Robinson.
443
444         * gtk/jhbuild.modules:
445         * gtk/patches/libxml2-2.9.0-dtd.patch: patch from upstream
446
447 2012-12-23  Allan Sandfeld Jensen  <allan.jensen@digia.com>
448
449         Unreviewed update of status to reviewer.
450
451         * Scripts/webkitpy/common/config/committers.py:
452
453 2012-12-23  Philippe Normand  <pnormand@igalia.com>
454
455         [GTK][jhbuild] Bump GStreamer to 1.0.4
456         https://bugs.webkit.org/show_bug.cgi?id=105640
457
458         Reviewed by Martin Robinson.
459
460         * gtk/jhbuild.modules: Update GStreamer modules to 1.0.4.
461
462 2012-12-21  Nico Weber  <thakis@chromium.org>
463
464         chromium: webkit-build-directory doesn't handle ninja output dirs
465         https://bugs.webkit.org/show_bug.cgi?id=105671
466
467         Reviewed by Daniel Bates.
468
469         Teach webkit-build-directory to detect the ninja build directory.
470
471         * Scripts/webkitdirs.pm:
472         (determineBaseProductDir):
473         (determineIsChromiumNinja):
474
475 2012-12-21  Ryosuke Niwa  <rniwa@webkit.org>
476
477         [Mountain Lion] platform/mac/editing/spelling/removing-underline-after-accepting-autocorrection-using-punctuation.html failing
478         https://bugs.webkit.org/show_bug.cgi?id=95477
479
480         Reviewed by Sam Weinig.
481
482         Co-authored by Jia Pu.
483
484         Specify a fixed autocorrection dictionary in resetDefaultsToConsistentValues() so that
485         we can have consistent autocorrection behavior when running tests.
486
487         * DumpRenderTree/mac/DumpRenderTree.mm:
488         (resetDefaultsToConsistentValues):
489         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
490         (WTR::InjectedBundle::platformInitialize):
491
492 2012-12-21  Zan Dobersek  <zandobersek@gmail.com>
493
494         [GTK] Add WebKit1-specific TestExpectations file
495         https://bugs.webkit.org/show_bug.cgi?id=97562
496
497         Reviewed by Dirk Pranke.
498
499         Implement the expectations_files method inside the GtkPort class that
500         returns the list of the TestExpectations files the current test configuration
501         requires. The method is implemented in a similar way as on the EFL port.
502         A unit test is also added.
503
504         * Scripts/webkitpy/layout_tests/port/gtk.py:
505         (GtkPort._search_paths):
506         (GtkPort):
507         (GtkPort.expectations_files):
508         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
509         (GtkPortTest.test_expectations_files):
510
511 2012-12-21  Sudarsana Nagineni  <sudarsana.nagineni@intel.com>
512
513         [EFL] Enable feature option for the CSS sticky position
514         https://bugs.webkit.org/show_bug.cgi?id=105630
515
516         Reviewed by Kenneth Rohde Christiansen.
517
518         Enable CSS sticky position feature by default for the
519         EFL port. This feature option was added in r138252.
520
521         No new tests. Existing layout tests pass.
522
523         * Scripts/webkitperl/FeatureList.pm:
524
525 2012-12-21  Jussi Kukkonen  <jussi.kukkonen@intel.com>
526
527         [WK2] REGRESSION(r138232): It made 4 tests timeout
528         https://bugs.webkit.org/show_bug.cgi?id=105623
529
530         Reviewed by Kenneth Rohde Christiansen.
531
532         r138232 changed functionality so that empty text output
533         no longer leads to a dump() call, which leads to a timeout.
534         Instead do a text dump if there is text output or if there
535         is no audio output.
536
537         * WebKitTestRunner/TestInvocation.cpp:
538         (WTR::TestInvocation::dumpResults):
539
540 2012-12-21  Zeno Albisser  <zeno@webkit.org>
541
542         [Qt][Mac] run-launcher cannot find Qt5WebKitWidgets.
543         https://bugs.webkit.org/show_bug.cgi?id=105309
544
545         On Windows the library name is Qt5WebKitWidgets as well.
546         So r138136 should not have adjusted the library name for Windows.
547
548         Reviewed by Simon Hausmann.
549
550         * Scripts/webkitdirs.pm:
551         (builtDylibPathForName):
552
553 2012-12-21  Zeno Albisser  <zeno@webkit.org>
554
555         [Qt] Add new TestExpectations for qt-5.0-mac-wk2.
556         https://bugs.webkit.org/show_bug.cgi?id=105424
557
558         This change will allow us to run a very specific
559         subset of the available LayoutTests on a mac buildbot.
560         The new baseline search path for mac will look as follows:
561         qt-5.0-mac-wk2 -> qt-5.0-wk2 -> qt-5.0 -> qt-mac -> qt
562
563         Reviewed by Dirk Pranke.
564
565         * Scripts/webkitpy/layout_tests/port/qt.py:
566         (QtPort._search_paths):
567         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
568         (QtPortTest):
569
570 2012-12-20  Ryuan Choi  <ryuan.choi@samsung.com>
571
572         [EFL] Build break with latest EFL libraries.
573         https://bugs.webkit.org/show_bug.cgi?id=104827
574
575         Reviewed by Laszlo Gombos.
576
577         The eo EFL package is introduced and evas and ecore use it since 1.8.
578         While introducing Eo, EFL changed several structures of Evas and Ecore
579         from own specific class to Eo.
580
581         So, this patch adds FindEo and changes declaration of Evas, Evas_Object,
582         Ecore_Timer to build with latest EFL libraries.
583
584         * DumpRenderTree/efl/CMakeLists.txt: Includes eo EFL library.
585         * EWebLauncher/CMakeLists.txt: Ditto.
586         * MiniBrowser/efl/CMakeLists.txt: Ditto.
587         * ImageDiff/PlatformEfl.cmake: Ditto.
588         * TestWebKitAPI/PlatformEfl.cmake: Ditto.
589         * TestWebKitAPI/PlatformWebView.h: Modified declaration of Evas_Object.
590         * WebKitTestRunner/InjectedBundle/TestRunner.h: Modified declaration of Ecore_Timer.
591         * WebKitTestRunner/PlatformEfl.cmake: Includes eo EFL library.
592         * WebKitTestRunner/PlatformWebView.h: Modified declaration of Evas_Object.
593
594 2012-12-20  Dirk Pranke  <dpranke@chromium.org>
595
596         TestExpectation fallback is broken
597         https://bugs.webkit.org/show_bug.cgi?id=105583
598
599         Reviewed by Ryosuke Niwa.
600
601         When I added support for the -wk2 in port names, it looks
602         like I broke the way we computed the list of TestExpectations
603         files so that the implementation-version directory was left out :(.
604         Fixed and cleaned up the tests.
605
606         * Scripts/webkitpy/layout_tests/port/base.py:
607         (Port.expectations_files):
608         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
609         (TestWebKitPort.__init__):
610         (PortTestCase.test_skipped_directories_for_symbols):
611         (test_skipped_directories_for_features):
612         (test_skipped_directories_for_features_no_matching_tests_in_test_list):
613         (test_skipped_tests_for_unsupported_features_empty_test_list):
614         (test_skipped_layout_tests):
615         (test_expectations_files):
616
617 2012-12-20  Anders Carlsson  <andersca@apple.com>
618
619         Don't include WebKit2 headers when building TestWebKitAPI on windows
620         https://bugs.webkit.org/show_bug.cgi?id=105582
621
622         Reviewed by Tim Horton.
623
624         * TestWebKitAPI/config.h:
625         * TestWebKitAPI/win/main.cpp:
626
627 2012-12-20  Ryosuke Niwa  <rniwa@webkit.org>
628
629         REGRESSION(r133820?): SimplifyMarkupTest API test asserts
630         https://bugs.webkit.org/show_bug.cgi?id=105370
631
632         Reviewed by Simon Fraser.
633
634         Re-enable the test.
635
636         * TestWebKitAPI/Tests/mac/SimplifyMarkup.mm:
637         (TestWebKitAPI::TEST):
638
639 2012-12-20  Nico Weber  <thakis@chromium.org>
640
641         chromium nrwt: Pick the newest binary found in DEFAULT_BUILD_DIRECTORIES, not the first
642         https://bugs.webkit.org/show_bug.cgi?id=105498
643
644         Reviewed by Dirk Pranke.
645
646         Use the newest binary available rather than an than always picking one
647         build directory over another based on iteration order.
648
649         * Scripts/webkitpy/layout_tests/port/chromium.py:
650         (ChromiumPort._static_build_path):
651         Check for timestamps.
652         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
653         (ChromiumMacPortTest.test_build_path_timestamps):
654         Test that out / xcodebuild selection happens based on timestamps
655         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
656         (ChromiumWinPortTest.test_build_path_timestamps):
657         Test that out / build selection happens based on timestamps
658
659 2012-12-19  Simon Fraser  <simon.fraser@apple.com>
660
661         Lots of sticky tests failing in WK2
662         https://bugs.webkit.org/show_bug.cgi?id=105464
663
664         Reviewed by Beth Dakin.
665
666         WebKitTestRunner had a race between snapshotting in the UI process,
667         and resettting after the test in the web process. InjectedBundle::done()
668         was a bad place to call page()->resetAfterTest(), because of this race;
669         it could reset the scroll position before the UI snapshot had been obtained.
670         
671         Fix by moving the call to page()->resetAfterTest() into didReceiveMessage(),
672         for the "Reset" message which will come in before the next test.
673
674         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
675         (WTR::InjectedBundle::didReceiveMessage):
676         (WTR::InjectedBundle::done):
677
678 2012-12-19  Filip Pizlo  <fpizlo@apple.com>
679
680         DFG speculation checks that take JumpList should consolidate OSRExits
681         https://bugs.webkit.org/show_bug.cgi?id=105401
682
683         Reviewed by Oliver Hunt.
684
685         Update the profiler to understand that an OSR exit may have multiple
686         candidate exit sites.
687
688         * Scripts/display-profiler-output:
689
690 2012-12-20  Eric Seidel  <eric@webkit.org>
691
692         EWSes do not need to dump build/test logs when things pass
693         https://bugs.webkit.org/show_bug.cgi?id=105402
694
695         Reviewed by Dirk Pranke.
696
697         When we were first writing the commit-queue and EWS system
698         and running them locally, we thought it was important to see the
699         build/test output as it happened.  Now these queues have been running
700         for years, and we don't use this feature anymore.  So this patch
701         makes us stop tee-ing the output directly to stdout, and instead
702         only print any output from commands when they fail.
703
704         I also added some logging when we execute each command
705         since we are no longer printing the command output itself.
706         This should make it much easier to see errors with the EWS
707         system itself, now that all the passing-build output is removed.
708
709         This required updating a zillion unittests, since we have a bunch of
710         "integration" style testing of the queue-bots.  Many of the unittests for the bots were
711         already approximating this behavior by using a "logging" Executive.
712         I disabled logging in those cases to avoid redundant logging in those tests.
713
714         * Scripts/webkitpy/common/system/executive.py:
715         (Executive.command_for_printing):
716         (Executive.run_command):
717         * Scripts/webkitpy/common/system/executive_mock.py:
718         (MockExecutive.command_for_printing):
719         (MockExecutive.run_command):
720         * Scripts/webkitpy/common/system/executive_unittest.py:
721         (ExecutiveTest.test_auto_stringify_args):
722         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
723         (EarlyWarningSytemTest._default_expected_logs):
724         (_test_ews):
725         (_test_ewses):
726         * Scripts/webkitpy/tool/commands/queues.py:
727         (AbstractQueue.run_webkit_patch):
728         * Scripts/webkitpy/tool/commands/queues_unittest.py:
729         (AbstractQueueTest._assert_run_webkit_patch):
730         (CommitQueueTest.test_commit_queue):
731         (test_rollout):
732         (test_rollout_lands):
733         (test_manual_reject_during_processing):
734         (StyleQueueTest.test_style_queue_with_style_exception):
735         (test_style_queue_with_watch_list_exception):
736
737 2012-12-20  Dominic Mazzoni  <dmazzoni@google.com>
738
739         AX: support clickPoint in DRT for chromium
740         https://bugs.webkit.org/show_bug.cgi?id=97359
741
742         Reviewed by Chris Fleizach.
743
744         Implement clickPointX, clickPointY, and elementAtPoint for Chromium, so that
745         several more tests can pass.
746
747         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
748         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
749         (WebTestRunner::AccessibilityUIElement::clickPointXGetterCallback):
750         (WebTestRunner):
751         (WebTestRunner::AccessibilityUIElement::clickPointYGetterCallback):
752         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
753         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
754         (AccessibilityUIElement):
755
756 2012-12-20  Dan Bernstein  <mitz@apple.com>
757
758         prepare-ChangeLog adds "Additional information of the change such as approach, rationale" where it doesn't make sense
759         https://bugs.webkit.org/show_bug.cgi?id=99268
760
761         Reviewed by Alexey Proskuryakov.
762
763         * Scripts/prepare-ChangeLog:
764         (generateNewChangeLogs): Removed the "Additional information of the change" line.
765
766 2012-12-20  Carlos Garcia Campos  <cgarcia@igalia.com>
767
768         Unreviewed. Skip unit test /webkit/atk/getTextInParagraphAndBodyModerate
769
770         * Scripts/run-gtk-tests:
771         (TestRunner):
772
773 2012-12-20  Thiago Marcos P. Santos  <thiago.santos@intel.com>
774
775         [Qt] Move ImageDiff to the Tools directory root
776         https://bugs.webkit.org/show_bug.cgi?id=105420
777
778         Reviewed by Tor Arne Vestbø.
779
780         Moved ImageDiff from outside DumpRenderTree because it is also a
781         dependency when running pixel tests for WebKit2.
782
783         * ImageDiff/ImageDiff.pro: Renamed from Tools/DumpRenderTree/qt/ImageDiff.pro.
784         * ImageDiff/qt/ImageDiff.cpp: Renamed from Tools/DumpRenderTree/qt/ImageDiff.cpp.
785         * Tools.pro:
786
787 2012-12-20  Zan Dobersek  <zandobersek@gmail.com>
788
789         [GTK] Remove the --enable-unstable-features configuration option
790         https://bugs.webkit.org/show_bug.cgi?id=105327
791
792         Reviewed by Martin Robinson.
793
794         Add a missing feature option for the CSS sticky position feature.
795
796         * Scripts/webkitperl/FeatureList.pm:
797
798 2012-12-20  Carlos Garcia Campos  <cgarcia@igalia.com>
799
800         Unreviewed. Skip unit test WebKit2.ForceRepaint that times out
801
802         * Scripts/run-gtk-tests:
803         (TestRunner):
804
805 2012-12-20  Dominik Röttsches  <dominik.rottsches@intel.com>
806
807         [EFL] MiniBrowser does not play Infinite Gangnam Style
808         https://bugs.webkit.org/show_bug.cgi?id=103531
809
810         Reviewed by Kenneth Rohde Christiansen.
811
812         For EFL we would like this feature to be built by default.
813
814         * Scripts/webkitperl/FeatureList.pm:
815
816 2012-12-20  Csaba Osztrogonác  <ossy@webkit.org>
817
818         [Qt] EWS bots should call qmake always
819         https://bugs.webkit.org/show_bug.cgi?id=90049
820
821         Reviewed by Simon Hausmann.
822
823         * Scripts/webkitdirs.pm:
824         (buildQMakeProjects):
825
826 2012-12-20  Jochen Eisinger  <jochen@chromium.org>
827
828         [chromium] fix stray printf in TestRunner's WebTestProxy
829         https://bugs.webkit.org/show_bug.cgi?id=105515
830
831         Reviewed by Ryosuke Niwa.
832
833         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
834         (WebTestRunner::WebTestProxyBase::shouldInsertNode):
835
836 2012-12-20  Dominik Röttsches  <dominik.rottsches@intel.com>
837
838         [WK2] WebAudio WKTR support
839         https://bugs.webkit.org/show_bug.cgi?id=95084
840
841         Reviewed by Kenneth Rohde Christiansen.
842
843         Implementing audio data dump support for WebKitTestRunner,
844         by adding the API to the IDL, enabling IPC for transferring
845         the audio data as bytebuffer and dumping it according
846         to run-webkit-tests script's expectations.
847         This enables audio testing on WebKit2 ports.
848         Based on initial wip-patch from Philippe Normand - thanks.
849
850         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl: Adding setAudioData API.
851         * WebKitTestRunner/InjectedBundle/TestRunner.cpp: Assign audio buffer data to a local WKDataRef after extracting it from WK2 using WKBundleDataFromUint8Array bundle API.
852         (WTR::TestRunner::setAudioData):
853         (WTR):
854         * WebKitTestRunner/InjectedBundle/TestRunner.h: Add a whatToDump state for dumping Audio.
855         (TestRunner):
856         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp: Add audio to the message that's transferred to the WKTR process.
857         (WTR::InjectedBundle::done):
858         * WebKitTestRunner/InjectedBundle/InjectedBundle.h: Allow setting and audio result handle to be transferred.
859         (WTR::InjectedBundle::setAudioResult):
860         (InjectedBundle):
861         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
862         (WTR::InjectedBundlePage::dump): Dummy handling the Audio case in the switch, build fix for Mac.
863         * WebKitTestRunner/TestInvocation.cpp: Audio dump support, dump audio if we haven't received a text dump from the web process.
864         (WTR::TestInvocation::dumpResults):
865         (WTR):
866         (WTR::TestInvocation::dumpAudio):
867         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle): Storing received audio data before dumping.
868         * WebKitTestRunner/TestInvocation.h:
869         (TestInvocation):
870
871 2012-12-20  Alberto Garcia  <agarcia@igalia.com>
872
873         [GTK] Count failed and timed out tests separately
874         https://bugs.webkit.org/show_bug.cgi?id=105448
875
876         Reviewed by Carlos Garcia Campos.
877
878         Display the number of tests that fail, time out and are skipped
879         separately in the buildbot page.
880
881         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
882         (RunGtkAPITests.commandComplete):
883         (RunGtkAPITests.evaluateCommand):
884         (RunGtkAPITests.getText2):
885         * Scripts/run-gtk-tests:
886         (TestRunner.run_tests):
887
888 2012-12-19  Ryosuke Niwa  <rniwa@webkit.org>
889
890         PerfTest.parse_output does too much
891         https://bugs.webkit.org/show_bug.cgi?id=105391
892
893         Reviewed by Tony Chang.
894
895         Moved the code to filter lines into PerfTest._filter_output, which has been renamed from _filter_stderr.
896         Also moved the code to output test results into PerfTest._run_with_driver.
897
898         * Scripts/webkitpy/performance_tests/perftest.py:
899         (PerfTest.__init__): Added _description.
900         (PerfTest.description): Added.
901         (PerfTest._run_with_driver): Moved the code to output test description and test results.
902         (PerfTest._filter_output): Moved and renamed from PerfTest._filter_stderr.
903         (PerfTest.parse_output): Removed the code to output test results. Return test results and description
904         respectively. Also removed the code that allowed some tests to omit values since all tests report each
905         iteration now since r136492.
906         (PerfTest.output_statistics): Removed the code to print test description, now done in _run_with_driver.
907         (ChromiumStylePerfTest._run_with_driver): Added. Chromium style tests are sufficiently different from
908         regular PerfTest that it doesn't make much sense to share _run_with_driver. But really, we should just
909         get rid of this type of test altogether in favor of regular performance tests that uses runner.js.
910         (ChromiumStylePerfTest.parse_and_log_output): Renamed from parse_output.
911         (PageLoadingPerfTest._run_with_driver): Removed the explicit '' for the test description.
912         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
913         (MainTest.test_parse_output): Removed the expected logs since parse_output no longer prints out results.
914         Also added a call to _filter_output since parse_output doesn't filter the output text anymore.
915         (MainTest.test_parse_output_with_failing_line): Added a call to _filter_output. Also added ',' after 'Time:'
916         so that it's not string-concatenated with the next line.
917         (MainTest.test_parse_output_with_description): Added; a test for PerfTest.description().
918         (MainTest.test_parse_output_with_subtests): Removed the expected logs, and added a call to _filter_output.
919
920 2012-12-19  Alexis Menard  <alexis@webkit.org>
921
922         Implement CSS parsing for CSS transitions unprefixed.
923         https://bugs.webkit.org/show_bug.cgi?id=104804
924
925         Reviewed by Dean Jackson.
926
927         Add a new flag ENABLE_CSS_TRANSFORMS_ANIMATIONS_TRANSITIONS_UNPREFIXED
928         to cover the work of unprefixing Transforms, Animations and 
929         Transitions. It will let the possibility of each ports to turn it off 
930         in their release branches until we're confident that these CSS 
931         properties are ready to be unprefixed.
932
933         * Scripts/webkitperl/FeatureList.pm:
934         * qmake/mkspecs/features/features.pri:
935
936 2012-12-19  Julien Chaffraix  <jchaffraix@webkit.org>
937
938         2 clampTo unit tests are failing
939         https://bugs.webkit.org/show_bug.cgi?id=105253
940
941         Reviewed by Emil A Eklund.
942
943         Removed clampToUnsigned as it was actually testing exactly what
944         clampToUnsignedLong was doing. Also made clampToIntFloat more robust
945         to precision loss (fix suggested by Emil).
946
947         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
948
949 2012-12-19  Thiago Marcos P. Santos  <thiago.santos@intel.com>
950
951         [EFL] Move ImageDiff to the Tools directory root
952         https://bugs.webkit.org/show_bug.cgi?id=105305
953
954         Reviewed by Laszlo Gombos.
955
956         We previously moved this directory from the DRT directory to
957         efl/ImageDiff. The major reason for doing so was because ImageDiff does
958         not depend on DumpRenderTree but for legacy reasons was living inside the
959         DRT folder, even though it is needed when running pixel tests for WK2.
960
961         Now we are moving to Tool directory root as other ports agreed to do the same.
962
963         * CMakeLists.txt:
964         * ImageDiff/CMakeLists.txt: Added.
965         * ImageDiff/PlatformEfl.cmake: Added.
966         * ImageDiff/efl/ImageDiff.cpp: Renamed from Tools/efl/ImageDiff/ImageDiff.cpp.
967         * efl/ImageDiff/CMakeLists.txt: Removed.
968
969 2012-12-19  Zeno Albisser  <zeno@webkit.org>
970
971         [Qt][Mac] run-launcher cannot find Qt5WebKitWidgets.
972         https://bugs.webkit.org/show_bug.cgi?id=105309
973
974         The version number is only used as a part
975         of the library name on Linux.
976
977         Reviewed by Kenneth Rohde Christiansen.
978
979         * Scripts/webkitdirs.pm:
980         (builtDylibPathForName):
981
982 2012-12-19  Alberto Garcia  <agarcia@igalia.com>
983
984         [GTK] Tests that time out are not considered as failures.
985         https://bugs.webkit.org/show_bug.cgi?id=105318
986
987         Reviewed by Martin Robinson.
988
989         Fix parsing of the GTK API tests log files.
990
991         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
992         (RunGtkAPITests.commandComplete):
993
994 2012-12-18  Ilya Tikhonovsky  <loislo@chromium.org>
995
996         Unreviewed, rolling out r138061.
997         https://bugs.webkit.org/show_bug.cgi?id=105396
998
999         Original patch:
1000         http://trac.webkit.org/changeset/138061
1001         https://bugs.webkit.org/show_bug.cgi?id=97359
1002
1003         It is crashing on Debug bots
1004
1005         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
1006         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
1007         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
1008         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
1009         (AccessibilityUIElement):
1010
1011 2012-12-18  Alice Liu  <alice.liu@apple.com>
1012
1013         Test for https://bugs.webkit.org/show_bug.cgi?id=105106
1014         Add SPI to WebKit1 WebFrame for hit testing
1015
1016         Reviewed by Dan Bernstein.
1017
1018         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added file Tests/mac/ElementAtPointInWebFrame.mm
1019         * TestWebKitAPI/Tests/mac/ElementAtPointInWebFrame.mm: Added.
1020         (TestWebKitAPI::TEST): Loads html with two divs positioned in the 2nd and 4th quadrants of the webview.
1021         Then hit-tests at three points, expecting to hit the two divs and body element. 
1022
1023 2012-12-18  Alice Liu  <alice.liu@apple.com>
1024
1025         Test for https://bugs.webkit.org/show_bug.cgi?id=105262
1026         Add SPI to WebKit1 WebFrame for node conversion to JSValueRef
1027
1028         Reviewed by Anders Carlsson.
1029
1030         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added file Tests/mac/JSWrapperForNodeInWebFrame.mm.mm
1031         * TestWebKitAPI/Tests/mac/JSWrapperForNodeInWebFrame.mm.mm: Added.
1032         (TestWebKitAPI::TEST): Tests for the correct JS wrapper for a DOMNode, provided a WebScriptWorld.
1033         Loads html with a single div element. In an isolated world, creates a property on that node.
1034         Also, in the standard world, creates a different property on that node.  Then tests for 4 things:
1035             - Existence of the isolated property in the isolated world.
1036             - Existence of the standard property in the standard world.
1037             - Non-existence of the isolated property in the standard world.
1038             - Non-existence of the standard property in the isolated world.
1039
1040 2012-12-18  Simon Fraser  <simon.fraser@apple.com>
1041
1042         Disable the SimplifyMarkup API test, wich asserts in debug builds.
1043         Tracked by https://bugs.webkit.org/show_bug.cgi?id=105370.
1044         
1045         * TestWebKitAPI/Tests/mac/SimplifyMarkup.mm:
1046         (TestWebKitAPI::TEST):
1047
1048 2012-12-18  Joseph Pecoraro  <pecoraro@apple.com>
1049
1050         [Mac] Add Build Phase to Check Headers for Inappropriate Macros (Platform.h macros)
1051         https://bugs.webkit.org/show_bug.cgi?id=104279
1052
1053         Reviewed by David Kilzer.
1054
1055         Add a build phase script that checks a framework's header files for
1056         Platform.h and Compiler.h macros. Also add some tests for the script.
1057
1058         * Scripts/check-for-inappropriate-macros-in-external-headers: Added.
1059         Script to be used in build phases to check for inappropriate macros in headers.
1060
1061         * Scripts/test-webkitruby: Added.
1062         Script that can be used to run all webkitruby tests.
1063
1064         * Scripts/test-webkit-scripts:
1065         Also run the ruby tests.
1066
1067         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/fake-data-failing-expected.txt: Added.
1068         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/pass-expected.txt: Added.
1069         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/resources/Fake.framework/Headers/Fail.h: Added.
1070         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/resources/Fake.framework/Headers/Pass.h: Added.
1071         * Scripts/webkitruby/check-for-inappropriate-macros-in-external-headers-tests/run-test.rb: Added.
1072         Tests for the new script.
1073
1074 2012-12-18  Dominic Mazzoni  <dmazzoni@google.com>
1075
1076         AX: support clickPoint in DRT for chromium
1077         https://bugs.webkit.org/show_bug.cgi?id=97359
1078
1079         Reviewed by Chris Fleizach.
1080
1081         Implement clickPointX, clickPointY, and elementAtPoint for Chromium, so that
1082         several more tests can pass.
1083
1084         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
1085         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
1086         (WebTestRunner::AccessibilityUIElement::clickPointXGetterCallback):
1087         (WebTestRunner):
1088         (WebTestRunner::AccessibilityUIElement::clickPointYGetterCallback):
1089         (WebTestRunner::AccessibilityUIElement::elementAtPointCallback):
1090         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
1091         (AccessibilityUIElement):
1092
1093 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
1094
1095         Build fix after r138045.
1096
1097         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1098         (PerfTestsRunner._collect_tests): Use relative_path. We skip only when this path
1099         is not explicitly specified. The condition was backwards but never matched so we
1100         always ran the test regardless of whether it's listed in the skipped list or not.
1101         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
1102         (test_collect_tests_with_ignored_skipped_list): Renamed so that it actually runs.
1103
1104 2012-12-18  Emil A Eklund  <eae@chromium.org>
1105
1106         TestWebKitAPI/Tests/WebCore/LayoutUnit fails on mac
1107         https://bugs.webkit.org/show_bug.cgi?id=105332
1108         
1109         Reviewed by Levi Weintraub.
1110         
1111         Create new test suite for LayoutUnit tests as the KURL tests are
1112         of the incompatible TEST_F type.
1113
1114         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp:
1115         (TestWebKitAPI):
1116         (TestWebKitAPI::TEST_F):
1117
1118 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
1119
1120         Running a skipped test with run-perf-tests could alert the user
1121         https://bugs.webkit.org/show_bug.cgi?id=105219
1122
1123         Reviewed by Eric Seidel.
1124
1125         Force run skipped performance tests explicitly specified in the command line.
1126         Also warn the user when the specified paths don't exist.
1127
1128         * Scripts/webkitpy/layout_tests/port/base.py:
1129         (Port.relative_test_filename):
1130         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1131         (PerfTestsRunner._parse_args):
1132         (PerfTestsRunner._collect_tests):
1133         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
1134         (test_collect_tests_with_skipped_list_and_files):
1135
1136 2012-12-18  Yaron Friedman  <yfriedman@chromium.org>
1137
1138         Chrome Android java output is too verbose
1139         https://bugs.webkit.org/show_bug.cgi?id=105247
1140
1141         Reviewed by Tony Chang.
1142
1143         Pass -quiet to ant to suppress extraneous output.
1144
1145         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
1146         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
1147
1148 2012-12-18  Ryosuke Niwa  <rniwa@webkit.org>
1149
1150         Consider removing --pause-before-testing option
1151         https://bugs.webkit.org/show_bug.cgi?id=105248
1152
1153         Reviewed by Eric Seidel.
1154
1155         Remove --pause-before-testing option.
1156
1157         * Tools/Scripts/webkitpy/layout_tests/port/driver.py:
1158         (DriverProxy.start): Removed.
1159         * Scripts/webkitpy/performance_tests/perftest.py:
1160         (PerfTest.__init__):
1161         (PerfTest.test_path): Renamed from path_or_url.
1162         (PerfTest.run):
1163         (PerfTest): Removed "driver" argument.
1164         (PerfTest._run_with_driver):
1165         (PerfTest.run_single):
1166         (ChromiumStylePerfTest.__init__):
1167         (PageLoadingPerfTest.__init__):
1168         (PageLoadingPerfTest.run_single):
1169         (PageLoadingPerfTest._run_with_driver): Renamed from run so that PerfTest.run can create a driver for us.
1170         (ReplayPerfTest.__init__):
1171         (ReplayPerfTest.prepare):
1172         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
1173         (TestPageLoadingPerfTest.test_run):
1174         (TestPageLoadingPerfTest.test_run_with_memory_output):
1175         (TestPageLoadingPerfTest.test_run_with_bad_output):
1176         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1177         (PerfTestsRunner._parse_args):
1178         (PerfTestsRunner._run_tests_set): Removed the code to create a driver and pause before running the test.
1179         (PerfTestsRunner._run_single_test): Removed "driver" argument.
1180         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
1181         (run_test):
1182         (test_run_test_pause_before_testing): Removed.
1183
1184 2012-12-18  Alberto Garcia  <agarcia@igalia.com>
1185
1186         [GTK] Tests that time out are not considered as failures.
1187         https://bugs.webkit.org/show_bug.cgi?id=105318
1188
1189         Reviewed by Martin Robinson.
1190
1191         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
1192         (RunGtkAPITests.commandComplete):
1193         Consider log lines that end with TIMEOUT as errors.
1194
1195         * Scripts/run-gtk-tests:
1196         (TestRunner.run_tests):
1197         Consider timeouts as failures when returning the number of failed
1198         tests.
1199
1200 2012-12-18  Pravin D  <pravin.d@samsung.com>
1201
1202         Unreviewed, adding secondary email.
1203
1204         * Scripts/webkitpy/common/config/committers.py:
1205
1206 2012-12-18  Pravin D  <pravin.d@samsung.com>
1207
1208         Unreviewed, promoting myself to committer's list.
1209
1210         * Scripts/webkitpy/common/config/committers.py:
1211
1212 2012-12-18  Thiago Marcos P. Santos  <thiago.santos@intel.com>
1213
1214         [EFL] ImageDiff does not build on WebKit2 only builds
1215         https://bugs.webkit.org/show_bug.cgi?id=105288
1216
1217         Reviewed by Kenneth Rohde Christiansen.
1218
1219         Moved ImageDiff to the efl/ folder inside Tools/ (so we don't pollute
1220         the folder with things which are EFL-specific.
1221
1222         ImageDiff will be built by default now regardless of WebKit API flavor
1223         you choose.
1224
1225         * CMakeLists.txt:
1226         * DumpRenderTree/efl/CMakeLists.txt:
1227         * efl/ImageDiff/CMakeLists.txt: Added.
1228         * efl/ImageDiff/ImageDiff.cpp: Renamed from Tools/DumpRenderTree/efl/ImageDiff.cpp.
1229
1230 2012-12-18  Kenneth Rohde Christiansen  <kenneth@webkit.org>
1231
1232         [EFL][WK2] When creating new windows, use the opener's context
1233         https://bugs.webkit.org/show_bug.cgi?id=105283
1234
1235         Reviewed by Laszlo Gombos.
1236
1237         Use the context of the opener if exists.
1238
1239         * MiniBrowser/efl/main.c:
1240         (on_key_down):
1241         (on_window_create):
1242         (window_create):
1243         (elm_main):
1244
1245 2012-12-18  Thiago Marcos P. Santos  <thiago.santos@intel.com>
1246
1247         [EFL][WK2] Gardening of TestWebKitAPI tests
1248         https://bugs.webkit.org/show_bug.cgi?id=105268
1249
1250         Reviewed by Kenneth Rohde Christiansen.
1251
1252         Some tests were marked as failing but are passing and the newly added
1253         LayoutUnit should have it's own binary because it conflicts with the
1254         fixture used by KURL and is failing on the bots because of that.
1255
1256         * TestWebKitAPI/CMakeLists.txt:
1257         * TestWebKitAPI/PlatformEfl.cmake:
1258
1259 2012-12-18  Jochen Eisinger  <jochen@chromium.org>
1260
1261         [chromium] move webintent related methods to TestRunner library
1262         https://bugs.webkit.org/show_bug.cgi?id=105164
1263
1264         Reviewed by Darin Fisher.
1265
1266         Also deleted the unused deliveredIntentFailure and
1267         deliveredIntentResult methods.
1268
1269         * DumpRenderTree/chromium/DRTTestRunner.cpp:
1270         (DRTTestRunner::DRTTestRunner):
1271         * DumpRenderTree/chromium/DRTTestRunner.h:
1272         (DRTTestRunner):
1273         * DumpRenderTree/chromium/TestRunner/public/WebTestDelegate.h:
1274         (WebKit):
1275         (WebTestRunner::WebTestDelegate::setCurrentWebIntentRequest):
1276         (WebTestRunner::WebTestDelegate::currentWebIntentRequest):
1277         * DumpRenderTree/chromium/TestRunner/public/WebTestProxy.h:
1278         (WebKit):
1279         (WebTestProxyBase):
1280         (WebTestRunner::WebTestProxy::registerIntentService):
1281         (WebTestRunner::WebTestProxy::dispatchIntent):
1282         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
1283         (WebTestRunner::TestRunner::TestRunner):
1284         (WebTestRunner::TestRunner::sendWebIntentResponse):
1285         (WebTestRunner):
1286         (WebTestRunner::TestRunner::deliverWebIntent):
1287         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
1288         (TestRunner):
1289         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
1290         (WebTestRunner::WebTestProxyBase::postAccessibilityNotification):
1291         (WebTestRunner::WebTestProxyBase::registerIntentService):
1292         (WebTestRunner):
1293         (WebTestRunner::WebTestProxyBase::dispatchIntent):
1294         * DumpRenderTree/chromium/WebViewHost.cpp:
1295         (WebViewHost::setCurrentWebIntentRequest):
1296         (WebViewHost::currentWebIntentRequest):
1297         * DumpRenderTree/chromium/WebViewHost.h:
1298         (WebKit):
1299         (WebViewHost):
1300
1301 2012-12-18  Jochen Eisinger  <jochen@chromium.org>
1302
1303         [chromium] move editing callback dumping to the WebTestProxy
1304         https://bugs.webkit.org/show_bug.cgi?id=105013
1305
1306         Reviewed by Darin Fisher.
1307
1308         To decide whether or not to dump the editing callbacks, the WebTestProxy
1309         needs to query the TestRunner. For this purpose, I introduce a
1310         WebTestRunner class that just exposes the shouldDumpEditingCallbacks
1311         method. In contrast to e.g. WebEventSender, the WebTestRunner is just
1312         an interface, and it is not owned by the WebTestInterfaces, but the
1313         embedder of the TestRunner library must provide it. This will change
1314         once all TestRunner methods are in the library.
1315
1316         * DumpRenderTree/DumpRenderTree.gypi:
1317         * DumpRenderTree/chromium/DRTTestRunner.cpp:
1318         (DRTTestRunner::DRTTestRunner):
1319         (DRTTestRunner::reset):
1320         * DumpRenderTree/chromium/DRTTestRunner.h:
1321         (DRTTestRunner):
1322         (DRTTestRunner::taskList):
1323         (DRTTestRunner::WorkQueue::taskList):
1324         * DumpRenderTree/chromium/TestRunner/public/WebTestInterfaces.h:
1325         (WebTestRunner):
1326         (WebTestInterfaces):
1327         * DumpRenderTree/chromium/TestRunner/public/WebTestProxy.h:
1328         (WebKit):
1329         (WebTestRunner):
1330         (WebTestProxyBase):
1331         (WebTestRunner::WebTestProxy::shouldBeginEditing):
1332         (WebTestRunner::WebTestProxy::shouldEndEditing):
1333         (WebTestRunner::WebTestProxy::shouldInsertNode):
1334         (WebTestRunner::WebTestProxy::shouldInsertText):
1335         (WebTestRunner::WebTestProxy::shouldChangeSelectedRange):
1336         (WebTestRunner::WebTestProxy::shouldDeleteRange):
1337         (WebTestRunner::WebTestProxy::shouldApplyStyle):
1338         (WebTestRunner::WebTestProxy::didBeginEditing):
1339         (WebTestRunner::WebTestProxy::didChangeSelection):
1340         (WebTestRunner::WebTestProxy::didChangeContents):
1341         (WebTestRunner::WebTestProxy::didEndEditing):
1342         * DumpRenderTree/chromium/TestRunner/public/WebTestRunner.h: Copied from Tools/DumpRenderTree/chromium/TestRunner/public/WebTestInterfaces.h.
1343         (WebTestRunner):
1344         (WebTestRunner::WebTestRunner::shouldDumpEditingCallbacks):
1345         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
1346         (WebTestRunner::TestRunner::TestRunner):
1347         (WebTestRunner::TestRunner::~TestRunner):
1348         (WebTestRunner):
1349         (WebTestRunner::TestRunner::reset):
1350         (WebTestRunner::TestRunner::shouldDumpEditingCallbacks):
1351         (WebTestRunner::TestRunner::dumpEditingCallbacks):
1352         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
1353         (TestRunner):
1354         * DumpRenderTree/chromium/TestRunner/src/WebTestInterfaces.cpp:
1355         (WebTestRunner::WebTestInterfaces::Internal::testRunner):
1356         (WebTestRunner::WebTestInterfaces::Internal::setTestRunner):
1357         (WebTestInterfaces::Internal):
1358         (WebTestRunner::WebTestInterfaces::Internal::Internal):
1359         (WebTestRunner::WebTestInterfaces::testRunner):
1360         (WebTestRunner):
1361         (WebTestRunner::WebTestInterfaces::setTestRunner):
1362         * DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp:
1363         (WebTestRunner::WebTestProxyBase::postAccessibilityNotification):
1364         (WebTestRunner):
1365         (WebTestRunner::WebTestProxyBase::shouldBeginEditing):
1366         (WebTestRunner::WebTestProxyBase::shouldEndEditing):
1367         (WebTestRunner::WebTestProxyBase::shouldInsertNode):
1368         (WebTestRunner::WebTestProxyBase::shouldInsertText):
1369         (WebTestRunner::WebTestProxyBase::shouldChangeSelectedRange):
1370         (WebTestRunner::WebTestProxyBase::shouldDeleteRange):
1371         (WebTestRunner::WebTestProxyBase::shouldApplyStyle):
1372         (WebTestRunner::WebTestProxyBase::didBeginEditing):
1373         (WebTestRunner::WebTestProxyBase::didChangeSelection):
1374         (WebTestRunner::WebTestProxyBase::didChangeContents):
1375         (WebTestRunner::WebTestProxyBase::didEndEditing):
1376         * DumpRenderTree/chromium/TestShell.cpp:
1377         (TestShell::initialize):
1378         (TestShell::createNewWindow):
1379         * DumpRenderTree/chromium/WebViewHost.cpp:
1380         (WebViewHost::shouldBeginEditing):
1381         (WebViewHost::shouldEndEditing):
1382         (WebViewHost::shouldInsertNode):
1383         (WebViewHost::shouldInsertText):
1384         (WebViewHost::shouldChangeSelectedRange):
1385         (WebViewHost::shouldDeleteRange):
1386         (WebViewHost::shouldApplyStyle):
1387         * DumpRenderTree/chromium/WebViewHost.h:
1388         (WebViewHost):
1389
1390 2012-12-17  Jochen Eisinger  <jochen@chromium.org>
1391
1392         [chromium] Fix methods in TestRunner that directly modify the settings instead of going through WebPreferences
1393         https://bugs.webkit.org/show_bug.cgi?id=105154
1394
1395         Reviewed by Tony Chang.
1396
1397         If the settings are modified directly, the embedder can't keep track of
1398         them and might accidentially override the changed settings.
1399
1400         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
1401         (WebPreferences):
1402         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
1403         (WebTestRunner::TestRunner::TestRunner):
1404         (WebTestRunner::TestRunner::setAsynchronousSpellCheckingEnabled):
1405         (WebTestRunner):
1406         (WebTestRunner::TestRunner::setMinimumTimerInterval):
1407         (WebTestRunner::TestRunner::setTouchDragDropEnabled):
1408         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
1409         (TestRunner):
1410         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
1411         (WebTestRunner::WebPreferences::reset):
1412         (WebTestRunner::WebPreferences::applyTo):
1413
1414 2012-12-17  Julien Chaffraix  <jchaffraix@webkit.org>
1415
1416         Unreviewed unit test skipping.
1417
1418         Filed https://bugs.webkit.org/show_bug.cgi?id=105253 to re-enable them.
1419
1420         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
1421         (TestWebKitAPI):
1422         (TestWebKitAPI::TEST):
1423
1424 2012-12-17  Dean Jackson  <dino@apple.com>
1425
1426         Unreviewed. Add Silvia Pfeiffer to contributor list.
1427
1428         * Scripts/webkitpy/common/config/committers.py:
1429
1430 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
1431
1432         build.webkit.org - add more google windows bots
1433         https://bugs.webkit.org/show_bug.cgi?id=105225
1434
1435         Reviewed by Eric Seidel.
1436
1437         Adds two more windows bots (google-windows-3, a builder, and
1438         google-windows-4, a tester) for the Chromium configs.
1439
1440         * BuildSlaveSupport/build.webkit.org-config/config.json:
1441
1442 2012-12-17  Eric Seidel  <eric@webkit.org>
1443
1444         Cleanup --profiler=perf code a little and allow up to 1000 report files (instead of 100)
1445         https://bugs.webkit.org/show_bug.cgi?id=105217
1446
1447         Reviewed by Dirk Pranke.
1448
1449         Previously --profile would fail strangely if find_unused_filename returned None
1450         (which would happen as soon as we had 100 report files).
1451         Now we'll search through 1000 filenames before giving up (and assert if we failed to find one).
1452         I also made some minor cleanups to the output of perf, and tried to make the
1453         commands more self-documenting by using long-names for args.
1454
1455         * Scripts/webkitpy/common/system/profiler.py:
1456         (SingleFileOutputProfiler.__init__):
1457         (Perf.attach_to_pid):
1458         (Perf.profile_after_exit):
1459
1460 2012-12-17  Sheriff Bot  <webkit.review.bot@gmail.com>
1461
1462         Unreviewed, rolling out r137198.
1463         http://trac.webkit.org/changeset/137198
1464         https://bugs.webkit.org/show_bug.cgi?id=105212
1465
1466         This patch is causing API behavior compatibility problems
1467         (Requested by zdobersek on #webkit).
1468
1469         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1470         (willSendRequestCallback):
1471
1472 2012-12-17  Elliott Sprehn  <esprehn@chromium.org>
1473
1474         Unreviewed. Adding myself to committers list.
1475
1476         * Scripts/webkitpy/common/config/committers.py:
1477
1478 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
1479
1480         nrwt: shuffle code around for cleanup in run_webkit_tests.py
1481         https://bugs.webkit.org/show_bug.cgi?id=105078
1482
1483         Reviewed by Ojan Vafai.
1484
1485         This patch reorders functions so that run_webkit_tests can be
1486         understood in a top-down matter better, and to make things slightly
1487         more sensible when I merge manager.py into it.
1488
1489         Also, this adds tests for the actual main() routine, which had several
1490         bugs get through testing a week or two ago.
1491
1492         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1493         (main):
1494         (parse_args):
1495         (_set_up_derived_options):
1496         (run):
1497         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1498         (RunTest):
1499         (RunTest.test_no_http_tests):
1500         (PortTest.disabled_test_mac_lion):
1501         (MainTest):
1502         (MainTest.test_exception_handling):
1503         (MainTest.test_exception_handling.interrupting_run):
1504         (MainTest.test_exception_handling.successful_run):
1505         (MainTest.test_exception_handling.successful_run.FakeRunDetails):
1506         (MainTest.test_exception_handling.exception_raising_run):
1507
1508 2012-12-17  Dirk Pranke  <dpranke@chromium.org>
1509
1510         webkitpy: move --lint-test-files code into its own module
1511         https://bugs.webkit.org/show_bug.cgi?id=105077
1512
1513         Reviewed by Eric Seidel.
1514
1515         It really didn't need to be in run_webkit_tests.py anymore.
1516         This patch adds a new lint-test-expectations scripts to Tools/Scripts;
1517         rwt --lint-test-files is still supported for backwards compatibility
1518         for now.
1519
1520         * Scripts/lint-test-expectations: Added.
1521         * Scripts/webkitpy/layout_tests/lint_test_expectations.py: Added.
1522         (lint):
1523         (main):
1524         * Scripts/webkitpy/layout_tests/lint_test_expectations_unittest.py: Added.
1525         (FakePort):
1526         (FakePort.__init__):
1527         (FakePort.test_configuration):
1528         (FakePort.expectations_dict):
1529         (FakePort.skipped_layout_tests):
1530         (FakePort.all_test_configurations):
1531         (FakePort.configuration_specifier_macros):
1532         (FakePort.get_option):
1533         (FakeFactory):
1534         (FakeFactory.__init__):
1535         (FakeFactory.get):
1536         (FakeFactory.all_port_names):
1537         (LintTest):
1538         (LintTest.test_all_configurations):
1539         (LintTest.test_lint_test_files):
1540         (LintTest.test_lint_test_files__errors):
1541         (MainTest):
1542         (MainTest.test_success):
1543         (MainTest.test_success.interrupting_lint):
1544         (MainTest.test_success.successful_lint):
1545         (MainTest.test_success.exception_raising_lint):
1546         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1547         (main):
1548         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1549         (StreamTestingMixin.assertNotEmpty):
1550
1551 2012-12-17  Adam Klein  <adamk@chromium.org>
1552
1553         build-webkit: rename --template-tag to --template-element to match ENABLE #define name
1554         https://bugs.webkit.org/show_bug.cgi?id=105072
1555
1556         Reviewed by Laszlo Gombos.
1557
1558         * Scripts/webkitperl/FeatureList.pm:
1559
1560 2012-12-17  Julien Chaffraix  <jchaffraix@webkit.org>
1561
1562         Add some unit testing for WTF::clampTo* functions
1563         https://bugs.webkit.org/show_bug.cgi?id=105060
1564
1565         Reviewed by Emil A Eklund.
1566
1567         The new tests are targeted at clamping to int / unsigned and float as those are
1568         the one explicitely exposed through more dedicated functions (like clampToInteger,
1569         clampToPositiveInteger and clampToFloat).
1570
1571         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
1572
1573 2012-12-17  Eric Seidel  <eric@webkit.org>
1574
1575         rpt --profile --chromium-android throws exception
1576         https://bugs.webkit.org/show_bug.cgi?id=105061
1577
1578         Reviewed by Daniel Bates.
1579
1580         Somehow popen(args) got turned into popen(*args) at some point,
1581         which caused my new stringify code to fail when args was passed by name.
1582         Fixed and tested.
1583
1584         * Scripts/webkitpy/common/system/executive.py:
1585         (Executive._stringify_args):
1586         (Executive):
1587         (Executive.popen):
1588         * Scripts/webkitpy/common/system/executive_unittest.py:
1589         (ExecutiveTest.test_popen_args):
1590
1591 2012-12-17  Mark Pilgrim  <pilgrim@chromium.org>
1592
1593         Use Platform::current() instead of webKitPlatformSupport() in DumpRenderTree
1594         https://bugs.webkit.org/show_bug.cgi?id=105054
1595
1596         Reviewed by Jochen Eisinger.
1597
1598         Part of a refactoring series; see tracking bug 82948.
1599
1600         * DumpRenderTree/chromium/NotificationPresenter.cpp:
1601         (NotificationPresenter::show):
1602         * DumpRenderTree/chromium/Task.cpp:
1603         (postTask):
1604         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp:
1605
1606 2012-12-17  Emil A Eklund  <eae@chromium.org>
1607
1608         Clamp values in LayoutUnit::operator/ when SATURATED_LAYOUT_ARITHMETIC is enabled
1609         https://bugs.webkit.org/show_bug.cgi?id=104955
1610
1611         Reviewed by Julien Chaffraix.
1612
1613         Add tests for LayoutUnit.
1614
1615         * TestWebKitAPI/CMakeLists.txt:
1616         * TestWebKitAPI/ForwardingHeaders: Added.
1617         * TestWebKitAPI/ForwardingHeaders/WebCore: Added.
1618         * TestWebKitAPI/ForwardingHeaders/WebCore/LayoutUnit.h: Added.
1619         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
1620         * TestWebKitAPI/TestWebKitAPI.gypi:
1621         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1622         * TestWebKitAPI/Tests/WebCore/LayoutUnit.cpp: Added.
1623         (TestWebKitAPI):
1624         (TestWebKitAPI::TEST):
1625         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
1626
1627 2012-12-17  Jochen Eisinger  <jochen@chromium.org>
1628
1629         [chromium] removed unused preferences from testRunner.overridePreference
1630         https://bugs.webkit.org/show_bug.cgi?id=105169
1631
1632         Reviewed by Tony Chang.
1633
1634         Most notably this removes the code related to font family maps which is
1635         now handled via window.internals.settings
1636
1637         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
1638         (WebPreferences):
1639         * DumpRenderTree/chromium/TestRunner/src/TestRunner.cpp:
1640         (WebTestRunner::TestRunner::overridePreference):
1641         * DumpRenderTree/chromium/TestRunner/src/TestRunner.h:
1642         (TestRunner):
1643         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
1644         (WebTestRunner::WebPreferences::applyTo):
1645
1646 2012-12-17  James Robinson  <jamesr@chromium.org>
1647
1648         [chromium] Move creation of WebLayerTreeView to WebWidgetClient
1649         https://bugs.webkit.org/show_bug.cgi?id=105071
1650
1651         Reviewed by Adrienne Walker.
1652
1653         Update WebViewHost for the new WebWidgetClient interface.
1654
1655         * DumpRenderTree/chromium/WebViewHost.cpp:
1656         (WebViewHost::initializeLayerTreeView):
1657         (WebViewHost::layerTreeView):
1658         (WebViewHost::~WebViewHost):
1659         * DumpRenderTree/chromium/WebViewHost.h:
1660         (WebViewHost):
1661
1662 2012-12-17  Kenneth Rohde Christiansen  <kenneth@webkit.org>
1663
1664         [EFL][WK2] window_create doesn't receive the url
1665         https://bugs.webkit.org/show_bug.cgi?id=105184
1666
1667         Reviewed by Alexis Menard.
1668
1669         * MiniBrowser/efl/main.c: Use the url supplied to window_create
1670         (on_window_create):
1671
1672 2012-12-17  Ilya Tikhonovsky  <loislo@chromium.org>
1673
1674         Web Inspector: Native Memory Instrumentation: MemoryInstrumentation doesn't detect reportMemoryUsage method defined in a base class.
1675         https://bugs.webkit.org/show_bug.cgi?id=105026
1676
1677         Reviewed by Yury Semikhatsky.
1678
1679         Old SFINAE test was replaced with new one based on this article:
1680         http://stackoverflow.com/questions/1966362/sfinae-to-check-for-inherited-member-functions
1681
1682         New test which covers this problem was added.
1683
1684         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
1685
1686 2012-12-17  Anthony Scian  <ascian@rim.com>
1687
1688         [Blackberry] Static code analysis warning fixes
1689         https://bugs.webkit.org/show_bug.cgi?id=105129
1690
1691         Reviewed by Rob Buis.
1692
1693         Added null checks for fopen calls, no recovery attempted; access fault prevention only.
1694
1695         * DumpRenderTree/blackberry/DumpRenderTree.cpp:
1696         (BlackBerry::WebKit::createFile):
1697         (BlackBerry::WebKit::DumpRenderTree::runTest):
1698
1699 2012-12-17  John Mellor  <johnme@chromium.org>
1700
1701         Add Anton Vayvod and Tim Volodine as contributors in committers.py
1702         https://bugs.webkit.org/show_bug.cgi?id=105025
1703
1704         Unreviewed. Adds two contributors to committers.py.
1705         They are initially contributing to the Text Autosizing feature (master bug: 84186).
1706
1707         * Scripts/webkitpy/common/config/committers.py:
1708
1709 2012-12-17  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
1710
1711         Unreviewed, add myself to committers.py.
1712
1713         * Scripts/webkitpy/common/config/committers.py:
1714
1715 2012-12-17  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
1716
1717         [Qt] remove some unnecessary CONFIG additions
1718
1719         Reviewed by Simon Hausmann.
1720
1721         qt is already added by spec_pre.prf, warn_on and depend_includepath by
1722         default_pre.prf.
1723
1724         * qmake/mkspecs/features/default_post.prf:
1725
1726 2012-12-17  Ilya Tikhonovsky  <loislo@chromium.org>
1727
1728         Unreviewed, rolling out r137892.
1729         http://trac.webkit.org/changeset/137892
1730         https://bugs.webkit.org/show_bug.cgi?id=105026
1731
1732         it broke compilation on windows
1733
1734         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
1735
1736 2012-12-14  Ilya Tikhonovsky  <loislo@chromium.org>
1737
1738         Web Inspector: Native Memory Instrumentation: MemoryInstrumentation doesn't detect reportMemoryUsage method defined in a base class.
1739         https://bugs.webkit.org/show_bug.cgi?id=105026
1740
1741         Reviewed by Yury Semikhatsky.
1742
1743         Old SFINAE test was replaced with new one based on this article:
1744         http://stackoverflow.com/questions/1966362/sfinae-to-check-for-inherited-member-functions
1745
1746         New test which covers this problem was added.
1747
1748         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
1749
1750 2012-12-17  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
1751
1752         [Qt] rely on automatic output directory setup
1753
1754         Reviewed by Simon Hausmann.
1755
1756         Now being a proper qt module (by virtue of having load(qt_build_config)
1757         in .qmake.conf), webkit gets the path setup goodies for free.
1758
1759         This also fixes Makefile.api.Debug/Release trying to generate qrc_WebKit.cpp
1760         at the same time, instead of the debug and release version ending up in different
1761         directories as they should.
1762
1763         * qmake/mkspecs/features/default_pre.prf:
1764
1765 2012-12-17  Charles Wei  <charles.wei@torchmobile.com.cn>
1766
1767         [BlackBerry] Need to re-enable video track
1768         https://bugs.webkit.org/show_bug.cgi?id=105155
1769
1770         Reviewed by George Staikos.
1771
1772         Just enable Video Track for BlackBerry.
1773
1774         * Scripts/webkitperl/FeatureList.pm:
1775
1776 2012-12-17  János Badics  <jbadics@inf.u-szeged.hu>
1777
1778         Unreviewed. Added myself to committers.py
1779
1780         * Scripts/webkitpy/common/config/committers.py:
1781
1782 2012-12-17  Jonathan Liu  <net147@gmail.com>
1783
1784         Fix linking of libraries with -Wl,-whole-archive
1785         https://bugs.webkit.org/show_bug.cgi?id=104436
1786
1787         Reviewed by Simon Hausmann.
1788
1789         If there is more than one library in LIBS linked with
1790         -Wl,-whole-archive -l... -Wl,-no-whole-archive, qmake tries to
1791         merge the duplicate flags in LIBS which causes linking to fail.
1792
1793         Add no_smart_library_merge to CONFIG to prevent the duplicate
1794         flags from being removed when using -Wl,-whole-archive.
1795
1796         * qmake/mkspecs/features/functions.prf:
1797
1798 2012-12-16  Arvid Nilsson  <anilsson@rim.com>
1799
1800         [BlackBerry] Adapt to new BlackBerry::Platform::TouchPoint API
1801         https://bugs.webkit.org/show_bug.cgi?id=105143
1802         RIM PR 171941
1803
1804         Reviewed by Rob Buis.
1805         Internally reviewed by George Staikos.
1806
1807         Adapt to new method names and encapsulation of TouchPoint data members.
1808
1809         * DumpRenderTree/blackberry/EventSender.cpp:
1810         (addTouchPointCallback):
1811         (updateTouchPointCallback):
1812         (touchEndCallback):
1813         (releaseTouchPointCallback):
1814         (sendTouchEvent):
1815
1816 2012-12-16  Maciej Stachowiak  <mjs@apple.com>
1817
1818         Add a script to compute the mean and 95% confidence interval (using two-sided t-test) of a sample set
1819         https://bugs.webkit.org/show_bug.cgi?id=105148
1820
1821         Reviewed by Filip Pizlo.
1822
1823         Usage examples:
1824         
1825         $ sampstat --help
1826         Usage: sampstat [options] [FILES]
1827           Compute the mean and 95% confidence interval of a sample set.
1828           Standard input or files must contain two or more decimal numbers, one per line.
1829         
1830         Options:
1831           -h, --help            show this help message and exit
1832           -u UNIT, --unit=UNIT  assume values are in units of UNIT
1833           -v, --verbose         print all values (with units)
1834         
1835         $ sampstat -u MB memresults.txt
1836         2356.90 MB +/- 101.34 MB (4.3%)
1837         
1838         $ sampstat -v -u MB memresults.txt
1839               2318.21 MB
1840               2399.56 MB
1841               2352.93 MB
1842         ----------------
1843         Mean: 2356.90 MB +/- 101.34 MB (4.3%)
1844         
1845
1846         * Scripts/sampstat: Added.
1847         (sum): Helper function to compute the sum of a list.
1848         (arithmeticMean): Compute the meam of a list.
1849         (standardDeviation): Compute the sample standard deviation (unbiased estimator).
1850         (standardError): Compute the sample standard error.
1851         (tDist): Compute t(.025, n-1), the t-value for atwo-sided 95% confidence interval. 
1852         (twoSidedConfidenceInterval): Compute the two-sided confidence interval range about the mean.
1853
1854 2012-12-15  Nima Ghanavatian  <nghanavatian@rim.com>
1855
1856         [BlackBerry] Pass key modifiers with touch and mouse events
1857         https://bugs.webkit.org/show_bug.cgi?id=105108
1858
1859         Reviewed by Rob Buis.
1860
1861         Update DRT to new api.
1862
1863         Internally reviewed by Gen Mak.
1864
1865         * DumpRenderTree/blackberry/EventSender.cpp:
1866         (mouseDownCallback):
1867         (mouseUpCallback):
1868         (mouseMoveToCallback):
1869
1870 2012-12-15  Zan Dobersek  <zandobersek@gmail.com>
1871
1872         Stop the XvfbDriver before proceeding with its starting setup
1873         https://bugs.webkit.org/show_bug.cgi?id=105006
1874
1875         Reviewed by Philippe Normand.
1876
1877         Just as with the base Driver implementation, stop the XvfbDriver
1878         before proceeding with setting up lock files and a new Xvfb instance.
1879         This forces any driver or Xvfb instances that are currently under
1880         the driver's control to be closed/killed and cleaned up after.
1881
1882         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
1883         (XvfbDriver._start):
1884
1885 2012-12-14  Dirk Pranke  <dpranke@chromium.org>
1886
1887         webkitpy: make port_name a required parameter to the non-test ports
1888         https://bugs.webkit.org/show_bug.cgi?id=105057
1889
1890         Reviewed by Eric Seidel.
1891
1892         This is just minor code cleanup along the path to making Port
1893         constructors simpler and pushing more logic into the port factory.
1894
1895         * Scripts/webkitpy/layout_tests/port/base.py:
1896         (Port.__init__):
1897         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
1898         (PortTest.make_port):
1899         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
1900         (TestWebKitPort.__init__):
1901         * Scripts/webkitpy/layout_tests/port/test.py:
1902         (TestPort):
1903         (TestPort.determine_full_port_name):
1904         (TestPort.__init__):
1905         * Scripts/webkitpy/layout_tests/port/xvfbdriver_unittest.py:
1906         (XvfbDriverTest.make_driver):
1907         (XvfbDriverTest.test_stop):
1908
1909 2012-12-14  Dirk Pranke  <dpranke@chromium.org>
1910
1911         garden-o-matic doesn't know about reftests
1912         https://bugs.webkit.org/show_bug.cgi?id=101976
1913
1914         Reviewed by Eric Seidel.
1915
1916         Attempt to make garden-o-matic properly aware of reftests again.
1917
1918         This is re-landing r137407 with a change to
1919         rebaselineWithStatusUpdates() to actually pass the filtered
1920         failureInofLost to checkout.rebaseline().
1921
1922         Unfortunately, the testing coverage for this whole module is almost
1923         non-existent, and it's not obvious that there's a good way to
1924         stub out checkout.rebaseline() to test that the right thing
1925         happens. I verified the change manually.
1926
1927         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
1928         (.):
1929         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
1930         (.):
1931         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
1932
1933 2012-12-14  Ryosuke Niwa  <rniwa@webkit.org>
1934
1935         "Running 1 DumpRenderTree over X shards" is not a helpful output
1936         https://bugs.webkit.org/show_bug.cgi?id=104858
1937
1938         Reviewed by Dirk Pranke.
1939
1940         The word "shard" is not typically used in test frameworks. This word is confusing for many WebKit
1941         contributors who typically don't have much experience on distributed computations and databases.
1942         So only print this message out in the debug mode.
1943
1944         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1945         * Scripts/webkitpy/layout_tests/views/printing.py:
1946         (Printer.print_workers_and_shards):
1947
1948 2012-12-13  Eric Seidel  <eric@webkit.org>
1949
1950         Add --profiler=PROFILER option to run-perf-tests to allow specifying which profiler to use on platforms with many
1951         https://bugs.webkit.org/show_bug.cgi?id=104891
1952
1953         Reviewed by Ryosuke Niwa.
1954
1955         I also implemented a very simple "Sample" Profiler using
1956         Mac OS X's /usr/bin/sample command line tool.
1957
1958         The real reason for this abstraction is to make it easy
1959         to support both perf and pprof on linux which seem to
1960         be about equally popular among those I ask in the Chrome team.
1961
1962         * Scripts/webkitpy/common/system/profiler.py:
1963         (ProfilerFactory.create_profiler):
1964         (ProfilerFactory):
1965         (ProfilerFactory.available_profilers_by_name):
1966         (ProfilerFactory.default_profiler_name):
1967         (Sample):
1968         (Sample.__init__):
1969         (Sample.attach_to_pid):
1970         (Sample.profile_after_exit):
1971         (IProfiler.attach_to_pid):
1972         * Scripts/webkitpy/layout_tests/port/driver.py:
1973         (Driver.__init__):
1974         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1975         (PerfTestsRunner._parse_args):
1976
1977 2012-12-13  Eric Seidel  <eric@webkit.org>
1978
1979         Unreviewed, fixing typo in python unittest.
1980
1981         I changed the default profiler on linux from pprof to perf in
1982         bug 104971.  I failed to update the unittest results at that time.
1983
1984         * Scripts/webkitpy/common/system/profiler_unittest.py:
1985         (ProfilerFactoryTest.test_basic):
1986
1987 2012-12-13  Ian Vollick  <vollick@chromium.org>
1988
1989         [chromium] Add a virtual test suite for enabling opt-in to composited scrolling
1990         https://bugs.webkit.org/show_bug.cgi?id=104911
1991
1992         Reviewed by Dirk Pranke.
1993
1994         Adds two virtual tests suites to run the tests in
1995         compositing/overflow/ and scrollbars/ with the flag
1996         --enable-accelerated-overflow-scroll, which I've also plumbed through
1997         in the usual way.
1998
1999         * DumpRenderTree/chromium/DumpRenderTree.cpp:
2000         (main):
2001         * DumpRenderTree/chromium/TestRunner/public/WebPreferences.h:
2002         (WebPreferences):
2003         * DumpRenderTree/chromium/TestRunner/src/WebPreferences.cpp:
2004         (WebTestRunner::WebPreferences::reset):
2005         (WebTestRunner::WebPreferences::applyTo):
2006         * DumpRenderTree/chromium/TestShell.cpp:
2007         (TestShell::TestShell):
2008         (TestShell::resetWebSettings):
2009         * DumpRenderTree/chromium/TestShell.h:
2010         (TestShell::setAcceleratedCompositingForOverflowScrollEnabled):
2011         (TestShell):
2012         * Scripts/webkitpy/layout_tests/port/chromium.py:
2013         (ChromiumPort.virtual_test_suites):
2014
2015 2012-12-13  Eric Seidel  <eric@webkit.org>
2016
2017         Use 'perf' to profile on linux instead of google-pprof by default
2018         https://bugs.webkit.org/show_bug.cgi?id=104971
2019
2020         Reviewed by Dirk Pranke.
2021
2022         This makes Chromium Linux match Chromium Android and use the perf
2023         tool by default.  Once https://bugs.webkit.org/show_bug.cgi?id=104891
2024         lands it will be possible to use pprof again on Linux.
2025
2026         This is slightly more advanced than the AndroidPerf profiler as
2027         instead of using a timeout on "perf record" I instead watch
2028         for the termination of the target process and then control-C
2029         the 'perf record' process.  This required me to add two new
2030         methods to Executive, one to have a limited-time wait() and
2031         the second to be able to send a control-C.  I chose to add
2032         these to Executive to make them easier to mock/fix-for-win32
2033         at a later time if needed.
2034
2035         * Scripts/webkitpy/common/system/executive.py:
2036         (Executive.wait_limited):
2037         (Executive.interrupt):
2038         * Scripts/webkitpy/common/system/profiler.py:
2039         (ProfilerFactory.create_profiler):
2040         (Perf):
2041         (Perf.__init__):
2042         (Perf._perf_path):
2043         (Perf.attach_to_pid):
2044         (Perf._first_ten_lines_of_profile):
2045         (Perf.profile_after_exit):
2046
2047 2012-12-13  Julie Parent  <jparent@chromium.org>
2048
2049         Dashboard cleanup: remove usage of global g_builders.
2050         https://bugs.webkit.org/show_bug.cgi?id=104941
2051
2052         Reviewed by Dirk Pranke.
2053
2054         The dashboards use a lot of global state, which makes hacking on them
2055         complicated. This change removes the use of one such global: g_builders.
2056         In most cases, we can just use currentBuilderGroup().builders instead,
2057         which is now currentBuilders().
2058         Surprisingly, the most changes were required to the unit tests, since
2059         they were even bigger offenders of bad hygiene, relying on global state
2060         set by other tests, randomly clobbering global variables in ways the
2061         real code doesn't, etc.
2062
2063         * TestResultServer/static-dashboards/builders.js:
2064         (BuilderGroup.prototype.setup):
2065         * TestResultServer/static-dashboards/dashboard_base.js:
2066         (.switch.return):
2067         (htmlForTestTypeSwitcher):
2068         * TestResultServer/static-dashboards/flakiness_dashboard.js:
2069         (generatePage):
2070         (getAllTestsTrie):
2071         (processTestRunsForAllBuilders):
2072         (showPopupForBuild):
2073         (generatePageForExpectationsUpdate):
2074         (loadExpectationsLayoutTests):
2075         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2076         (resetGlobals):
2077         (stubResultsByBuilder):
2078         (test):
2079         * TestResultServer/static-dashboards/loader.js:
2080         (.):
2081         * TestResultServer/static-dashboards/loader_unittests.js:
2082
2083 2012-12-13  Eric Seidel  <eric@webkit.org>
2084
2085         Callers should not have to stringify args before calling Executive run_command/popen
2086         https://bugs.webkit.org/show_bug.cgi?id=104975
2087
2088         Reviewed by Dirk Pranke.
2089
2090         One could argue that we should match the python call syntax here,
2091         but I think it's a more friendly API if we automagically handle
2092         stringification of args in run_command, etc.
2093         This removes map(unicode, args) from several callsites.
2094
2095         When I first tried to land this change, I didn't realize that
2096         Executive._command_for_printing depended on this behavior
2097         having been applied to args in run_command.  The fix is to
2098         call _stringify_args in both run_command and popen.
2099         This is slightly redundant, but given how short args have to be
2100         (due to shell limits), I don't think the double-encode check
2101         matters in practice.
2102
2103         This is slightly complicated by the fact that apache_http_server.py
2104         is the one caller in our codebase which uses shell=True.
2105         shell=True is a well-documented trail-of-tears:
2106         http://stackoverflow.com/questions/3172470/actual-meaning-of-shell-true-in-subprocess
2107         but to support this legacy (windows-only) code (which I can't easily test)
2108         I've added an if-hack to avoid stringifying the the popen(shell=True) case.
2109
2110         * Scripts/webkitpy/common/system/executive.py:
2111         (Executive.run_command):
2112         (Executive._stringify_args):
2113         (Executive.popen):
2114         * Scripts/webkitpy/common/system/executive_unittest.py:
2115         (ExecutiveTest.test_auto_stringify_args):
2116         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2117         (attach_to_pid):
2118
2119 2012-12-13  Adrienne Walker  <enne@chromium.org>
2120
2121         Unreviewed, rolling out r137645, r137646, and r137667.
2122         http://trac.webkit.org/changeset/137645
2123         http://trac.webkit.org/changeset/137646
2124         http://trac.webkit.org/changeset/137667
2125         https://bugs.webkit.org/show_bug.cgi?id=104911
2126
2127         Breaks some overflow layout tests
2128
2129         * Scripts/webkitpy/layout_tests/port/chromium.py:
2130         (ChromiumPort.virtual_test_suites):
2131
2132 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
2133
2134         Unreviewed, rolling out r137661.
2135         http://trac.webkit.org/changeset/137661
2136         https://bugs.webkit.org/show_bug.cgi?id=104891
2137
2138         broke unit tests, run-webkit-tests
2139
2140         * Scripts/webkitpy/common/system/executive.py:
2141         (Executive._run_command_with_teed_output):
2142         (Executive.run_command):
2143         (Executive.popen):
2144         * Scripts/webkitpy/common/system/profiler.py:
2145         (ProfilerFactory.create_profiler):
2146         (GooglePProf.profile_after_exit):
2147         (IProfiler.attach_to_pid):
2148         * Scripts/webkitpy/layout_tests/port/driver.py:
2149         (Driver.__init__):
2150         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2151         (PerfTestsRunner._parse_args):
2152
2153 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
2154
2155         nrwt: move rundetails, summarize_results into test_run_results
2156         https://bugs.webkit.org/show_bug.cgi?id=104963
2157
2158         Reviewed by Eric Seidel.
2159
2160         More cleanup; this the plain-old-data functions out of manager.py
2161         and into test_run_results next to the other aggregate data structures.
2162
2163         Also this cleans up a bunch of test code and reduces duplication.
2164
2165         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2166         (Manager.run):
2167         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
2168         (ManagerTest.test_look_for_new_crash_logs):
2169         * Scripts/webkitpy/layout_tests/models/test_run_results.py:
2170         (TestRunResults.__init__):
2171         (TestRunResults.add):
2172         (RunDetails):
2173         (RunDetails.__init__):
2174         (_interpret_test_failures):
2175         (summarize_results):
2176         * Scripts/webkitpy/layout_tests/models/test_run_results_unittest.py: Added.
2177         (get_result):
2178         (run_results):
2179         (summarized_results):
2180         (InterpretTestFailuresTest):
2181         (InterpretTestFailuresTest.setUp):
2182         (InterpretTestFailuresTest.test_interpret_test_failures):
2183         (SummarizedResultsTest):
2184         (SummarizedResultsTest.setUp):
2185         (SummarizedResultsTest.test_no_svn_revision):
2186         (SummarizedResultsTest.test_svn_revision):
2187         (SummarizedResultsTest.test_summarized_results_wontfix):
2188         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
2189         (BuildBotPrinterTests.get_printer):
2190         (BuildBotPrinterTests.test_print_unexpected_results):
2191         (BuildBotPrinterTests.test_print_results):
2192
2193 2012-12-13  Eric Seidel  <eric@webkit.org>
2194
2195         Make --profile output easier to understand for pprof on linux
2196         https://bugs.webkit.org/show_bug.cgi?id=104957
2197
2198         Reviewed by Dirk Pranke.
2199
2200         pprof is kinda mysterious, so we print some information
2201         on how to use it, and how to learn how to read the sample dump.
2202         This is similar to what we've done for perf output for Chromium Android.
2203
2204         * Scripts/webkitpy/common/system/profiler.py:
2205         (GooglePProf._pprof_path):
2206         (GooglePProf.profile_after_exit):
2207
2208 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
2209
2210         Revert r137407 - it didn't actually work right.
2211         https://bugs.webkit.org/show_bug.cgi?id=101976
2212
2213         Unreviewed, build fix.
2214
2215         Turns out that rebaselining was completely broken with this change :(
2216         We're passing in a list of test names rather than the list of
2217         failureInfo objects.
2218
2219         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
2220         (.):
2221         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
2222         (.):
2223         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
2224
2225 2012-12-13  Antoine Labour  <piman@chromium.org>
2226
2227         [chromium] Small fixes for WebPluginContainer::setWebLayer
2228         https://bugs.webkit.org/show_bug.cgi?id=104953
2229
2230         Reviewed by James Robinson.
2231
2232         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp: Clean up
2233         dangling pointers, force correct invalidation of the layer.
2234
2235 2012-12-13  Eric Seidel  <eric@webkit.org>
2236
2237         Add --profiler=PROFILER option to run-perf-tests to allow specifying which profiler to use on platforms with many
2238         https://bugs.webkit.org/show_bug.cgi?id=104891
2239
2240         Reviewed by Dirk Pranke.
2241
2242         I also implemented a very simple "Sample" Profiler using
2243         Mac OS X's /usr/bin/sample command line tool.
2244
2245         I also moved the map(unicode, args) call into Executive.popen
2246         so that all callers don't have to do it themselves.
2247
2248         The real reason for this abstraction is to make it easy
2249         to support both perf and pprof on linux which seem to
2250         be about equally popular among those I ask in the Chrome team.
2251
2252         * Scripts/webkitpy/common/system/executive.py:
2253         (Executive.popen):
2254         * Scripts/webkitpy/common/system/profiler.py:
2255         (ProfilerFactory.create_profiler):
2256         (ProfilerFactory):
2257         (ProfilerFactory.available_profilers_by_name):
2258         (ProfilerFactory.default_profiler_name):
2259         (Sample):
2260         (Sample.__init__):
2261         (Sample.attach_to_pid):
2262         (Sample.profile_after_exit):
2263         (IProfiler.attach_to_pid):
2264         * Scripts/webkitpy/layout_tests/port/driver.py:
2265         (Driver.__init__):
2266         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2267         (PerfTestsRunner._parse_args):
2268
2269 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
2270
2271         support -wk2 port names properly in webkitpy.layout_tests.port
2272         https://bugs.webkit.org/show_bug.cgi?id=104761
2273
2274         Reviewed by Eric Seidel.
2275
2276         Our support for WK2 port names was inconsistent at best; this patch
2277         modifies the code so that <port_name>-wk2 is supported for all ports that
2278         have WK2 implementations and is equivalent to specifying <port_name>
2279         and -2/--webkit-test-runner. In addition, this modifies
2280         builders.all_port_names() to include the wk2 variants.
2281
2282         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
2283         * Scripts/webkitpy/layout_tests/port/apple.py:
2284         (ApplePort.determine_full_port_name):
2285         * Scripts/webkitpy/layout_tests/port/base.py:
2286         (Port.determine_full_port_name):
2287         (Port.__init__):
2288         * Scripts/webkitpy/layout_tests/port/builders.py:
2289         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
2290         (DriverTest.make_port):
2291         * Scripts/webkitpy/layout_tests/port/mac.py:
2292         (MacPort.default_baseline_search_path):
2293         * Scripts/webkitpy/layout_tests/port/win.py:
2294         (WinPort.default_baseline_search_path):
2295         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
2296         (TestRebaselineExpectations.test_rebaseline_expectations):
2297
2298 2012-12-13  Ian Vollick  <vollick@chromium.org>
2299
2300         [chromium] Add a virtual test suite for enabling opt-in to composited scrolling
2301         https://bugs.webkit.org/show_bug.cgi?id=104911
2302
2303         Reviewed by Adrienne Walker.
2304
2305         Adds two virtual tests suites to run the tests in
2306         compositing/overflow/ and scrollbars/ with the flag
2307         --enable-accelerated-compositing-for-overflow-scroll
2308
2309         * Scripts/webkitpy/layout_tests/port/chromium.py:
2310         (ChromiumPort.virtual_test_suites):
2311
2312 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
2313
2314         delete extraneous self.fail in mastercfg_unittest.py
2315         https://bugs.webkit.org/show_bug.cgi?id=104940
2316
2317         Reviewed by Eric Seidel.
2318
2319         This was accidentally introduced in r137404.
2320
2321         * BuildSlaveSupport/build.webkit.org-config/mastercfg_unittest.py:
2322
2323 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
2324
2325         Fix a mis-named variable after r137621.
2326         https://bugs.webkit.org/show_bug.cgi?id=104871
2327
2328         Unreviewed, build fix.
2329
2330         * Scripts/webkitpy/layout_tests/views/buildbot_results.py:
2331         (BuildBotPrinter.print_results):
2332         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
2333
2334 2012-12-13  Dirk Pranke  <dpranke@chromium.org>
2335
2336         webkitpy: rename ResultSummary
2337         https://bugs.webkit.org/show_bug.cgi?id=104871
2338
2339         Reviewed by Eric Seidel.
2340
2341         This patch changes the names of ResultSummary and related modules,
2342         classes, and variables as a part of the larger clean up in bug 103824.
2343
2344         This change is basically a bunch of search and replaces:
2345         - result_summary.py -> test_run_results.py
2346         - class ResultSummary -> class TestRunResults
2347         - "result_summary" or "rs" as a generic variable name -> "run_results",
2348           when it doesn't matter (or we don't know) if this refers to the
2349           initial pass over the tests or the retry pass.
2350         - "result_summary" as a variable containing the results of the initial,
2351           complete test run -> "initial_results"
2352         - "retry_summary" as a variable containing the results of retried
2353           results -> "retry_results"
2354         - ResultSummary.results (the property) -> TestRunResults.results_by_name
2355         - ResultSummary.failures (the property) -> TestRunResults.failures_by_name
2356         - RunDetails.result_summary -> RunDetails.initial_results
2357         - RunDetails.retry_summary -> RunDetails.retry_results
2358
2359         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
2360         (LayoutTestRunner.__init__):
2361         (LayoutTestRunner.run_tests):
2362         (LayoutTestRunner._mark_interrupted_tests_as_skipped):
2363         (LayoutTestRunner._interrupt_if_at_failure_limits):
2364         (LayoutTestRunner._interrupt_if_at_failure_limits.interrupt_if_at_failure_limit):
2365         (LayoutTestRunner._update_summary_with_result):
2366         (LayoutTestRunner._handle_finished_test):
2367         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner_unittest.py:
2368         (FakePrinter.print_expected):
2369         (LayoutTestRunnerTests.test_interrupt_if_at_failure_limits):
2370         (LayoutTestRunnerTests.test_update_summary_with_result):
2371         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2372         (RunDetails.__init__):
2373         (summarize_results):
2374         (Manager.run):
2375         (Manager._look_for_new_crash_logs):
2376         (Manager._tests_to_retry):
2377         (Manager._upload_json_files):
2378         (Manager._stats_trie):
2379         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
2380         (ManagerTest.test_look_for_new_crash_logs):
2381         (SummarizeResultsTest):
2382         (SummarizeResultsTest.get_run_results):
2383         (SummarizeResultsTest.summarized_results):
2384         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
2385         (JSONLayoutResultsGenerator.__init__):
2386         (JSONLayoutResultsGenerator._get_modifier_char):
2387         (JSONLayoutResultsGenerator._insert_failure_summaries):
2388         (JSONLayoutResultsGenerator._get_failure_summary_entry):
2389         * Scripts/webkitpy/layout_tests/models/test_run_results.py: Renamed from Tools/Scripts/webkitpy/layout_tests/models/result_summary.py.
2390         (TestRunResults):
2391         (TestRunResults.__init__):
2392         (TestRunResults.add):
2393         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2394         (get_test_results):
2395         (MainTest.test_basic):
2396         * Scripts/webkitpy/layout_tests/views/buildbot_results.py:
2397         (BuildBotPrinter.print_results):
2398         (BuildBotPrinter.print_run_results):
2399         (BuildBotPrinter._print_run_results_entry):
2400         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
2401         (BuildBotPrinterTests.get_run_results):
2402         (BuildBotPrinterTests.test_print_unexpected_results.get_unexpected_results):
2403         (BuildBotPrinterTests):
2404         * Scripts/webkitpy/layout_tests/views/printing.py:
2405         (Printer.print_expected):
2406         (Printer._print_expected_results_of_type):
2407         (Printer.print_results):
2408         (Printer._print_timing_statistics):
2409         (Printer._print_worker_statistics):
2410         (Printer._print_aggregate_test_statistics):
2411         (Printer._print_individual_test_times):
2412         (Printer._print_directory_timings):
2413         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
2414
2415 2012-12-13  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2416
2417         [EFL] The WebKit2 bots are building and executing WebKit1 tests
2418         https://bugs.webkit.org/show_bug.cgi?id=104690
2419
2420         Reviewed by Csaba Osztrogonác.
2421
2422         Pass --no-webkit1 option to the EFL WK2 bots when building. This
2423         will disable WK1 API and thus, the WK1 API tests.
2424
2425         * BuildSlaveSupport/build.webkit.org-config/config.json:
2426         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2427         (CompileWebKit2Only):
2428         (BuildAndTestFactory):
2429         (BuildAndTestFactory.__init__):
2430         (BuildAndTestWebKit2OnlyFactory):
2431         Now bots using BuildAndTestWebKit2OnlyFactory (for instance only EFL) will
2432         get --no-webkit1 as a parameter on the build-webkit script.
2433         * Scripts/build-webkit:
2434         Added support for -no-webkit1 for the Qt and EFL port.
2435
2436 2012-12-13  Eric Seidel  <eric@webkit.org>
2437
2438         build-webkit --chromium-android forces Make even when update-webkit used --ninja
2439         https://bugs.webkit.org/show_bug.cgi?id=104839
2440
2441         Reviewed by Tony Chang.
2442
2443         This was recommended by Peter as part of:
2444         https://bugs.webkit.org/show_bug.cgi?id=104434#c4
2445         Fixing this makes build-webkit --chromium-android work for the common (ninja) case.
2446
2447         * Scripts/webkitdirs.pm:
2448         (buildChromium):
2449
2450 2012-12-13  Julie Parent  <jparent@chromium.org>
2451
2452         Loader_unittests should reset state between runs.
2453         https://bugs.webkit.org/show_bug.cgi?id=104862
2454
2455         Reviewed by Dirk Pranke.
2456
2457         The loader_unittests were not resetting state between runs, leaving
2458         a lot of data behind in global objects that the dashboard and tests
2459         rely on.  A handful were resetting just a few objects.  Use the common
2460         resetGlobals() method instead.
2461
2462         * TestResultServer/static-dashboards/loader_unittests.js:
2463
2464 2012-12-13  Andras Becsi  <andras.becsi@digia.com>
2465
2466         [Qt][WK2] Fix painting on Mac with retina display
2467         https://bugs.webkit.org/show_bug.cgi?id=104574
2468
2469         Reviewed by Kenneth Rohde Christiansen.
2470
2471         Remove setting the devicePixelRatio experimental property
2472         since the value is now automatically picked up from Qt.
2473
2474         * MiniBrowser/qt/qml/BrowserWindow.qml:
2475
2476 2012-12-13  Jussi Kukkonen  <jussi.kukkonen@intel.com>
2477
2478         [EFL][GTK] Don't call deprecated g_type_init when glib => 2.35
2479         https://bugs.webkit.org/show_bug.cgi?id=103209
2480
2481         Reviewed by Kenneth Rohde Christiansen.
2482
2483         g_type_init() is deprecated from 2.35.0 onwards. Don't call it
2484         in that case to avoid warnings and/or build failure.
2485
2486         * DumpRenderTree/gtk/ImageDiff.cpp:
2487         (main):
2488
2489 2012-12-13  Claudio Saavedra  <csaavedra@igalia.com>
2490
2491         [Tools] Add exception for gtk_style_context_get()
2492         https://bugs.webkit.org/show_bug.cgi?id=104896
2493
2494         Reviewed by Martin Robinson.
2495
2496         Add new exception for gtk_style_context_get().
2497
2498         * Scripts/webkitpy/style/checkers/cpp.py:
2499         (check_for_null): Add exception for gtk_style_context_get().
2500         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
2501         (WebKitStyleTest.test_null_false_zero): Add a unit
2502         test for the new exception.
2503
2504 2012-12-13  Joaquim Rocha  <jrocha@igalia.com>
2505
2506         REGRESSION (r137432): The /webkit2/WebKitCookieManager/accept-policy unit test is failing
2507         https://bugs.webkit.org/show_bug.cgi?id=104790
2508
2509         Reviewed by Carlos Garcia Campos.
2510
2511         Unskip the GTK accept policy test.
2512
2513         * Scripts/run-gtk-tests:
2514         (TestRunner):
2515
2516 2012-12-13  Xingnan Wang  <xingnan.wang@intel.com>
2517
2518         Unreviewed. Adding myself to committers list.
2519
2520         * Scripts/webkitpy/common/config/committers.py:
2521
2522 2012-12-12  Roger Fong  <roger_fong@apple.com>
2523
2524         Unreviewed. Fix typo in NRWT.
2525
2526         * Scripts/webkitpy/layout_tests/port/base.py:
2527         (Port.is_reference_html_file):
2528
2529 2012-12-12  Roger Fong  <roger_fong@apple.com>
2530
2531         Fix detection of ref test files in old and new run webkit tests.
2532         https://bugs.webkit.org/show_bug.cgi?id=104850.
2533
2534         Reviewed by Timothy Horton.
2535
2536         * Scripts/old-run-webkit-tests:
2537         (isUsedInReftest): 
2538         Add notref and ref prefix and suffix cases to isUsedinReftest.
2539         * Scripts/webkitpy/layout_tests/port/base.py:
2540         (Port):
2541         (Port.is_reference_html_file): 
2542         Typo in NRWT.
2543
2544 2012-12-12  Dirk Pranke  <dpranke@chromium.org>
2545
2546         rework ChromiumAndroidDriver.start() et. al to avoid funky layering
2547         https://bugs.webkit.org/show_bug.cgi?id=104856
2548
2549         Reviewed by Eric Seidel.
2550
2551         r137523 / bug 10357 introduced some back-and-forth hopping during
2552         driver.start() in the chromium android port. This patch cleans it
2553         up a bit.
2554
2555         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2556         (ChromiumAndroidDriver.__init__):
2557         (ChromiumAndroidDriver._start_once):
2558         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
2559         * Scripts/webkitpy/layout_tests/port/driver.py:
2560         (Driver.start):
2561         (Driver._start):
2562
2563 2012-12-12  Eric Seidel  <eric@webkit.org>
2564
2565         Add support for run-perf-tests --chromium-android --profile
2566         https://bugs.webkit.org/show_bug.cgi?id=103587
2567
2568         Reviewed by Dirk Pranke.
2569
2570         This makes it possible to profile PerformanceTests on an Android device
2571         using run-perf-tests --chromium-android --profile.
2572         CPU profiling on Android is sadly non-trivial, but this goes a long
2573         ways towards making it more accessible to non-Android developers.
2574
2575         Although this script requires your device to be flashed with a
2576         userdebug build (an annoyingly non-trivial endeavor for external contributors)
2577         it endeavors to detect and explain the common errors encountered
2578         while trying to setup a device for perf-based profiling.
2579
2580         * Scripts/webkitpy/common/system/profiler.py:
2581         (GooglePProf.profile_after_exit): Discovered that pprof will just hang if passed a missing file.
2582         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2583         (AndroidPerf):
2584         (AndroidPerf.__init__):
2585         (AndroidPerf.check_configuration):
2586         (AndroidPerf.print_setup_instructions):
2587         (_file_exists_on_device):
2588         (_run_adb_command):
2589         (attach_to_pid):
2590         (_perf_version_string):
2591         (_find_perfhost_binary):
2592         (_perfhost_path):
2593         (_first_ten_lines_of_profile):
2594         (profile_after_exit):
2595         (ChromiumAndroidDriver.__init__):
2596         (ChromiumAndroidDriver.__init__.methods):
2597         (ChromiumAndroidDriver._find_or_create_symfs):
2598         (ChromiumAndroidDriver._start_once):
2599         (ChromiumAndroidDriver):
2600         (ChromiumAndroidDriver._run_post_start_tasks):
2601         (ChromiumAndroidDriver._pid_from_android_ps_output):
2602         (ChromiumAndroidDriver._pid_on_target):
2603         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
2604         (ChromiumAndroidDriverTest.test_pid_from_android_ps_output):
2605         (AndroidPerfTest):
2606         (AndroidPerfTest.test_perf_output_regexp):
2607         * Scripts/webkitpy/layout_tests/port/driver.py:
2608         (Driver._setup_environ_for_driver):
2609         (Driver):
2610         (Driver._start):
2611         (Driver._run_post_start_tasks):
2612         (Driver._pid_on_target):
2613
2614 2012-12-12  Dirk Pranke  <dpranke@chromium.org>
2615
2616         nrwt: remove deprecated ORWT compatibility command line flags
2617         https://bugs.webkit.org/show_bug.cgi?id=104301
2618
2619         Reviewed by Eric Seidel.
2620
2621         The command line flags for --noshow-results, --[no-]launch-safari,
2622         and --use-remote-links-to-tests were for compatibility with
2623         old-run-webkit-tests, but the compatibility is not needed any more
2624         since NRWT is basically default and the scripts that passed the
2625         old flags have been updated.
2626
2627         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2628         (_set_up_derived_options):
2629         (parse_args):
2630
2631 2012-12-12  John Griggs  <jgriggs@rim.com>
2632
2633         2012-12-12  John Griggs  <jgriggs@rim.com>
2634
2635         [BlackBerry] Update Media Controls
2636         https://bugs.webkit.org/show_bug.cgi?id=104443
2637         https://przilla.ott.qnx.com/bugzilla/show_bug.cgi?id=204748
2638
2639         Update Media Controls for BlackBerry platform to allow audio, embedded video and fullscreen video controls to be positioned and styled differently.
2640
2641         Reviewed by Rob Buis.
2642
2643         BlackBerry-specific update to Media Controls - No new tests required.
2644
2645         * Scripts/webkitperl/FeatureList.pm:
2646
2647 2012-12-12  Justin Novosad  <junov@google.com>
2648
2649         Adding myself to SkiaGraphics watchlist
2650
2651         Unreviewed
2652
2653         * Scripts/webkitpy/common/config/watchlist:
2654
2655 2012-12-12  Martin Robinson  <mrobinson@igalia.com>
2656
2657         [GTK] Add authentication support to DRT and fix exposed issues in the libsoup backend
2658         https://bugs.webkit.org/show_bug.cgi?id=104285
2659
2660         Reviewed by Gustavo Noronha Silva.
2661
2662         Add support to DumpRenderTree for running authentication tests. Since the DRT
2663         expects an authentication callback, we add one to DRTSupport to avoid #ifdefs
2664         in platform-independent code for GTK+.
2665
2666         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2667         (resetDefaultsToConsistentValues): Reset the authentication password and username.
2668         (authenticationCallback): Added.
2669         (createWebView): Attach the authentiation callback during startup.
2670         * DumpRenderTree/gtk/TestRunnerGtk.cpp:
2671         (soupURIToKURL): soup_uri_to_string does not preserve passwords embedded
2672         in the URL so we add a somewhat messy method of re-adding them when they exist.
2673         It would be nice to use soupURIToKURL here, but it seems we cannot use KURL without
2674         pulling in lots of WebCore code so we use string search and replace.
2675         (TestRunner::queueLoad): Use the new helper.
2676
2677 2012-12-12  Dirk Pranke  <dpranke@chromium.org>
2678
2679         garden-o-matic should prefer efl/ over efl-wk1/ and efl-wk2/ when rebaselining
2680         https://bugs.webkit.org/show_bug.cgi?id=102389
2681
2682         Reviewed by Tony Chang.
2683
2684         The code wasn't aware of both the efl-wk1 and efl-wk2 variants,
2685         because all_port_names() by itself doesn't tell you about the
2686         variants. Work around this for now, and filed bug 104761 to
2687         address the underlying problem.
2688
2689         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
2690         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
2691         (BaselineOptimizerTest.test_efl):
2692
2693 2012-12-12  Krzysztof Czech  <k.czech@samsung.com>
2694
2695         [EFL] Possibility to turn off accessibility feature for WebKit-EFL.
2696         https://bugs.webkit.org/show_bug.cgi?id=103036
2697
2698         Reviewed by Laszlo Gombos.
2699
2700         Guard dependencies for accessibility (ATK), so that they can be turned off.
2701
2702         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.h:
2703         (AccessibilityUIElement):
2704         * WebKitTestRunner/PlatformEfl.cmake:
2705
2706 2012-12-12  Simon Fraser  <simon.fraser@apple.com>
2707
2708         Add some entries in the watchlist, and register me for them.
2709
2710         * Scripts/webkitpy/common/config/watchlist:
2711
2712 2012-12-12  Ilya Tikhonovsky  <loislo@chromium.org>
2713
2714         Web Inspector: Native Memory Instrumentation: remove fake root MemoryObjectInfo.
2715         https://bugs.webkit.org/show_bug.cgi?id=104796
2716
2717         Reviewed by Yury Semikhatsky.
2718
2719         It was not a good idea to introduce a fake root MemoryObjectInfo.
2720         It makes a problem when we visit an object without its own MemoryObjectType.
2721
2722         Example: RenderBox has a global pointer to a hash map.
2723         HashMap doesn't have its own object type because it is a generic container.
2724         It will inherit object type from the fake root memory object info.
2725         The same could happen for another container in another class with other MemoryObjectType.
2726
2727         This fact forces me to create custom process method for root objects
2728         because they need to have their own MemoryObjectInfo with customisable memory object type.
2729
2730         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
2731
2732         Drive by fix: InstrumentedPointer* was replaced with Wrapper* because actually it is using
2733         for instrumented and not instrumented object classes.
2734
2735 2012-12-12  Zan Dobersek  <zandobersek@gmail.com>
2736
2737         [GTK] Remove the last of unnecessary configuration options in configure.ac
2738         https://bugs.webkit.org/show_bug.cgi?id=104793
2739
2740         Reviewed by Martin Robinson.
2741
2742         Removing the filters and indexed database features from the list of features
2743         being configurable through the configure script. This is not the case anymore
2744         as the corresponding configuration options are being removed.
2745
2746         * Scripts/webkitdirs.pm:
2747         (buildAutotoolsProject):
2748
2749 2012-12-12  Mario Sanchez Prada  <mario@webkit.org>
2750
2751         [GTK][WK2] New API to detect display/execution of insecure content
2752         https://bugs.webkit.org/show_bug.cgi?id=104578
2753
2754         Reviewed by Martin Robinson.
2755
2756         Add coding style exception for 'readability/enum_casing' in GTK.
2757
2758         The GTK+ API use upper case, underscore separated, words in
2759         certain types of enums (e.g. signals, properties).
2760
2761         * Scripts/webkitpy/style/checker.py: Added exception, both for
2762         WebKitGTK+ and WebKit2GTK+.
2763
2764 2012-12-12  Jochen Eisinger  <jochen@chromium.org>
2765
2766         [chromium] Fix build on chromium-win
2767
2768         Unreviewed build-fix.
2769
2770         MSVS indeed doesn't have snprintf, adding wtf/StringExtras.h include
2771         to work around that.
2772
2773         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp:
2774
2775 2012-12-12  Simon Hausmann  <simon.hausmann@digia.com>
2776
2777         [Qt] Fix build on Mac
2778
2779         Reviewed by Csaba Osztrogonác.
2780
2781         On Mac OS X we have sqlite as a system library available.
2782
2783         * qmake/mkspecs/features/features.prf:
2784
2785 2012-12-12  Zan Dobersek  <zandobersek@gmail.com>
2786
2787         Unreviewed GTK gardening.
2788
2789         Skipping another unit test that regressed in r137432.
2790
2791         * Scripts/run-gtk-tests:
2792         (TestRunner):
2793
2794 2012-12-12  Simon Hausmann  <simon.hausmann@digia.com>
2795
2796         [Qt] Fix linkage against SQLite
2797         https://bugs.webkit.org/show_bug.cgi?id=104781
2798
2799         Reviewed by Csaba Osztrogonác.
2800
2801         Detect SQLite either through pkg-config, from the SQLITE3SRCDIR
2802         environment variable or from qtbase via a qt5.git build. Bail out early
2803         with an error message if neither option applies.
2804
2805         This removes the usage of $$QT.core.sources and the dependency on Qt's
2806         system-sqlite configure setting. We should always favour a system
2807         library over a copy in Qt.
2808
2809         * qmake/mkspecs/features/features.prf:
2810
2811 2012-12-12  Tommy Widenflycht  <tommyw@google.com>
2812
2813         MediaStream API: Change the data channel descriptor pattern to a handler pattern
2814         https://bugs.webkit.org/show_bug.cgi?id=104543
2815
2816         Reviewed by Kent Tamura.
2817
2818         Adding mocks to be able to test this refactoring.
2819
2820         * DumpRenderTree/DumpRenderTree.gypi:
2821         * DumpRenderTree/chromium/MockWebRTCDataChannelHandler.cpp: Added.
2822         (DataChannelReadyStateTask):
2823         (DataChannelReadyStateTask::DataChannelReadyStateTask):
2824         (MockWebRTCDataChannelHandler::MockWebRTCDataChannelHandler):
2825         (MockWebRTCDataChannelHandler::setClient):
2826         (MockWebRTCDataChannelHandler::bufferedAmount):
2827         (MockWebRTCDataChannelHandler::sendStringData):
2828         (MockWebRTCDataChannelHandler::sendRawData):
2829         (MockWebRTCDataChannelHandler::close):
2830         * DumpRenderTree/chromium/MockWebRTCDataChannelHandler.h: Added.
2831         (MockWebRTCDataChannelHandler):
2832         (MockWebRTCDataChannelHandler::taskList):
2833         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
2834         (RTCPeerConnectionReadyStateTask::RTCPeerConnectionReadyStateTask):
2835         (RemoteDataChannelTask::RemoteDataChannelTask):
2836         (MockWebRTCPeerConnectionHandler::createDataChannel):
2837         (MockWebRTCPeerConnectionHandler::stop):
2838         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h:
2839         (MockWebRTCPeerConnectionHandler):
2840
2841 2012-12-12  Jochen Eisinger  <jochen@chromium.org>
2842
2843         [chromium] WebTestPlugin must not depend on WTF::String
2844         https://bugs.webkit.org/show_bug.cgi?id=104783
2845
2846         Reviewed by Kent Tamura.
2847
2848         In a component build, we can't access the WTF that is part of the
2849         WebKit component. Accessing WTF instead will pull in a separate copy
2850         into the TestRunner library. Instead of WTF::String::number use
2851         snprintf() to print numbers.
2852
2853         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp:
2854
2855 2012-12-12  Jocelyn Turcotte  <jocelyn.turcotte@digia.com>
2856
2857         [Qt] Do not automatically enable force_static_libs_as_shared when using debug_and_release
2858         https://bugs.webkit.org/show_bug.cgi?id=104562
2859
2860         Reviewed by Tor Arne Vestbø.
2861
2862         force_static_libs_as_shared creates issues on Mac if QtWebKit is built as a framework.
2863         QtWebProcess, for example, will link to the framework's release library and will rely
2864         on DYLD_IMAGE_SUFFIX to switch to the debug build. Since QtWebProcess will also links
2865         directly to the debug build of WebCore, and that QtWebKit release will link in the
2866         release build of WebCore, both the debug and release version will be loaded.
2867
2868         To fix this properly we should make sure that applications always link to the release
2869         build of dependent libraries by bypassing resolveFinalLibraryName.
2870         Disable it for now as force_static_libs_as_shared is mainly for development while
2871         debug_and_release is mainly for production builds.
2872
2873         * qmake/mkspecs/features/default_post.prf:
2874
2875 2012-12-12  Zan Dobersek  <zandobersek@gmail.com>
2876
2877         Unreviewed GTK gardening of the unit tests.
2878
2879         Skip two unit tests that regressed during the hackfest.
2880
2881         * Scripts/run-gtk-tests:
2882         (TestRunner):
2883
2884 2012-12-11  Ryosuke Niwa  <rniwa@webkit.org>
2885
2886         Add Koji Ishii to the list of contributors.
2887
2888         * Scripts/webkitpy/common/config/committers.py:
2889
2890 2012-12-11  Dominic Mazzoni  <dmazzoni@google.com>
2891
2892         AX: Make isActionSupported cross-platform.
2893         https://bugs.webkit.org/show_bug.cgi?id=104681
2894
2895         Reviewed by Chris Fleizach.
2896
2897         Replace isActionSupported, which takes a platform-specific
2898         argument, with three cross-platform methods.
2899
2900         * DumpRenderTree/AccessibilityUIElement.cpp:
2901         (isPressActionSupportedCallback):
2902         (isIncrementActionSupportedCallback):
2903         (isDecrementActionSupportedCallback):
2904         (AccessibilityUIElement::getJSClass):
2905         * DumpRenderTree/AccessibilityUIElement.h:
2906         (AccessibilityUIElement):
2907         * DumpRenderTree/blackberry/AccessibilityUIElementBlackBerry.cpp:
2908         (AccessibilityUIElement::isPressActionSupported):
2909         (AccessibilityUIElement::isIncrementActionSupported):
2910         (AccessibilityUIElement::isDecrementActionSupported):
2911         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.cpp:
2912         (WebTestRunner::AccessibilityUIElement::AccessibilityUIElement):
2913         (WebTestRunner::AccessibilityUIElement::isPressActionSupportedCallback):
2914         (WebTestRunner):
2915         (WebTestRunner::AccessibilityUIElement::isIncrementActionSupportedCallback):
2916         (WebTestRunner::AccessibilityUIElement::isDecrementActionSupportedCallback):
2917         (WebTestRunner::AccessibilityUIElement::incrementCallback):
2918         (WebTestRunner::AccessibilityUIElement::decrementCallback):
2919         (WebTestRunner::AccessibilityUIElement::pressCallback):
2920         * DumpRenderTree/chromium/TestRunner/src/AccessibilityUIElementChromium.h:
2921         (AccessibilityUIElement):
2922         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
2923         (AccessibilityUIElement::isPressActionSupported):
2924         (AccessibilityUIElement::isIncrementActionSupported):
2925         (AccessibilityUIElement::isDecrementActionSupported):
2926         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
2927         (AccessibilityUIElement::isPressActionSupported):
2928         (AccessibilityUIElement::isIncrementActionSupported):
2929         (AccessibilityUIElement::isDecrementActionSupported):
2930         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
2931         (AccessibilityUIElement::isPressActionSupported):
2932         (AccessibilityUIElement::isIncrementActionSupported):
2933         (AccessibilityUIElement::isDecrementActionSupported):
2934         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.cpp:
2935         (WTR::AccessibilityUIElement::isPressActionSupported):
2936         (WTR::AccessibilityUIElement::isIncrementActionSupported):
2937         (WTR::AccessibilityUIElement::isDecrementActionSupported):
2938         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.h:
2939         (AccessibilityUIElement):
2940         * WebKitTestRunner/InjectedBundle/Bindings/AccessibilityUIElement.idl:
2941         * WebKitTestRunner/InjectedBundle/mac/AccessibilityUIElementMac.mm:
2942         (WTR::AccessibilityUIElement::isPressActionSupported):
2943         (WTR):
2944         (WTR::AccessibilityUIElement::isIncrementActionSupported):
2945         (WTR::AccessibilityUIElement::isDecrementActionSupported):
2946
2947 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2948
2949         nrwt: remove result_summary.total_tests_by_expectation
2950         https://bugs.webkit.org/show_bug.cgi?id=104756
2951
2952         Reviewed by Eric Seidel.
2953
2954         This field wasn't being used by anything.
2955
2956         * Scripts/webkitpy/layout_tests/models/result_summary.py:
2957         (ResultSummary.__init__):
2958
2959 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2960
2961         garden-o-matic doesn't know about reftests
2962         https://bugs.webkit.org/show_bug.cgi?id=101976
2963
2964         Reviewed by Ojan Vafai.
2965
2966         This is a first attempt at fixing this, sort of a "stop-the-bleeding"
2967         fix. We will now look at the list of tests we're asked to rebaseline,
2968         and log which ones are reftests into the status area, and rebaseline
2969         the rest.
2970
2971         Longer-term, we should display feedback in the results area based on
2972         which tests we're looking at. If the test list contains no reftests,
2973         we should behave as today; if the list is all reftest, we should
2974         probably disable the button, and if the list is a mix, we should
2975         probably display an alert next to the button or something. It would
2976         also be nice to annotate which tests were reftests, maybe with a
2977         "(reftest)" next to the test name or something.
2978
2979         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
2980         (.):
2981         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
2982         (.):
2983         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
2984
2985 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
2986
2987         REGRESSION(r133380): new tests without expected file reported as failing tests on the bots
2988         https://bugs.webkit.org/show_bug.cgi?id=101469
2989
2990         Reviewed by Ryosuke Niwa.
2991
2992         Fix parsing of NRWT's output to identify missing results separately
2993         from other regressions.
2994
2995         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2996         (RunWebKitTests._parseNewRunWebKitTestsOutput):
2997         * BuildSlaveSupport/build.webkit.org-config/mastercfg_unittest.py:
2998         (MasterCfgTest.test_nrwt_leaks_parsing):
2999         (test_nrwt_missing_results):
3000
3001 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
3002
3003         buildbot: change run-webkit-tests --no-launch-safari to --no-show-results
3004         https://bugs.webkit.org/show_bug.cgi?id=104299
3005
3006         Reviewed by Ryosuke Niwa.
3007
3008         Since --no-show-results is the more generically-named flag that NRWT uses.
3009
3010         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
3011         (RunWebKitTests):
3012
3013 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
3014
3015         remove --use-remote-links-to-tests from buildbot configs (except apple win)
3016         https://bugs.webkit.org/show_bug.cgi?id=104291
3017
3018         Reviewed by Csaba Osztrogonác.
3019
3020         This flag is ignored by NRWT and was just implemented for compatibility.
3021         ORWT still uses it, though.
3022
3023         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
3024         (RunWebKitTests):
3025         (RunWebKitTests.start):
3026
3027 2012-12-11  Anders Carlsson  <andersca@apple.com>
3028
3029         ASSERT running platform/mac/plugins/convert-point.html on the WK2 bots
3030         https://bugs.webkit.org/show_bug.cgi?id=104735
3031
3032         Reviewed by Jon Honeycutt.
3033
3034         Do the testing from within NPP_New instead of NPN_SetWindow. Plug-ins aren't supposed to do any real work inside NPP_SetWindow,
3035         and doing so will more easily trigger the assertion failure. Ideally we should get rid of the initial sync painting pass, but this
3036         gets the test passing again!
3037
3038         * DumpRenderTree/TestNetscapePlugIn/Tests/mac/ConvertPoint.cpp:
3039         (ConvertPoint::NPP_New):
3040
3041 2012-12-11  Julie Parent  <jparent@chromium.org>
3042
3043         Dashboard unit tests should not make xhrs.
3044         https://bugs.webkit.org/show_bug.cgi?id=104598
3045
3046         Reviewed by Dirk Pranke.
3047
3048         The unit test were requesting 25 files that they did not use.
3049         Stub out loader.request to avoid making the xhrs.
3050         Result: order of magnitude speed improvement in test run time (from ~1600ms to 160ms on my machine).
3051         
3052         * TestResultServer/static-dashboards/run-embedded-unittests.html:
3053         * TestResultServer/static-dashboards/run-unittests.html:
3054
3055 2012-12-11  Filip Pizlo  <fpizlo@apple.com>
3056
3057         Profiler should show bytecode dumps as they would have been visible to the JITs, including the profiling data that the JITs would see
3058         https://bugs.webkit.org/show_bug.cgi?id=104647
3059
3060         Reviewed by Oliver Hunt.
3061
3062         Added a "profiling" (or "p") command to show the profiling data that the JITs saw
3063         for each JIT compilation of a code block.
3064         
3065         Also added instruction counts in the "full" display and made the "full" display the
3066         default thing you see.
3067
3068         * Scripts/display-profiler-output:
3069
3070 2012-12-11  Eric Seidel  <eric@webkit.org>
3071
3072         Unreviewed, rolling out r137371.
3073         http://trac.webkit.org/changeset/137371
3074         https://bugs.webkit.org/show_bug.cgi?id=104434
3075
3076         Various scripts are not ready for out/ as the build directory,
3077         this can't work as written.
3078
3079         * Scripts/update-webkit:
3080         * Scripts/webkitdirs.pm:
3081         (buildChromium):
3082
3083 2012-12-11  Zan Dobersek  <zandobersek@gmail.com>
3084
3085         Unreviewed GTK build functionality fix.
3086
3087         Turns out the CSP feature was not enabled, that's why all the tests were failing.
3088
3089         * Scripts/webkitperl/FeatureList.pm:
3090
3091 2012-12-11  Eric Seidel  <eric@webkit.org>
3092
3093         Make Ninja the default build system for build-webkit --chromium
3094         https://bugs.webkit.org/show_bug.cgi?id=104434
3095
3096         Reviewed by Daniel Bates.
3097
3098         As far as I can tell, most Chromium developers are now using Ninja
3099         as their default build system instead of the native Xcode/Visual Studio/Make.
3100         This change makes Ninja the default for developers as well as all Chromium webkit bots.
3101
3102         Developers can pass --no-ninja to get the platform-native buildsystem instead of Ninja.
3103         It's possible we may wish to make this a per-platform default, and have
3104         Ninja only be the default for Mac/Linux.
3105
3106         There was a thread on webkit-dev discussing this change:
3107         http://lists.webkit.org/pipermail/webkit-dev/2012-December/023113.html
3108
3109         * Scripts/update-webkit:
3110
3111 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
3112
3113         nrwt: record the order and run times of all of the tests in a single file
3114         https://bugs.webkit.org/show_bug.cgi?id=104184
3115
3116         Reviewed by Ryosuke Niwa.
3117
3118         Replace the N tests_run*.txt files that we output into
3119         layout-test-results with a single packed JSON trie that contains
3120         the test ordering and run statistics. This can eventually be uploaded
3121         to the TestResultsServer to replace the times.json file we currently
3122         upload as well, and will let us replay the results of a previous
3123         run fairly deterministically.
3124
3125         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
3126         (Worker.__init__):
3127         (Worker.start):
3128         (Worker._run_test):
3129         (Worker.stop):
3130         (Worker._clean_up_after_test):
3131         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3132         (Manager._upload_json_files):
3133         (Manager._copy_results_html_file):
3134         (Manager):
3135         (Manager._stats_trie):
3136         (Manager._stats_trie._worker_number):
3137         * Scripts/webkitpy/layout_tests/models/test_results.py:
3138         (TestResult.__init__):
3139         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3140         (MainTest.test_missing_and_unexpected_results):
3141         (MainTest.test_retrying_and_flaky_tests):
3142         (EndToEndTest.test_end_to_end):
3143         (EndToEndTest.test_reftest_with_two_notrefs):
3144         (RebaselineTest.test_reset_results):
3145         (RebaselineTest.test_missing_results):
3146         (RebaselineTest.test_new_baseline):
3147
3148 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
3149
3150         check-webkit-style complaining about TestExpectations lines a patch didn't modify
3151         https://bugs.webkit.org/show_bug.cgi?id=104687
3152
3153         Reviewed by Ojan Vafai.
3154
3155         Way back in bug 60466 / r86165, we changed the style checker
3156         for TestExpectations to ensure that the entire file linted, not
3157         just the lines modified by a patch, in order to ensure that a
3158         new line didn't cause errors elsewhere in the file (e.g., by
3159         introducing duplicate lines). At the time, errors in the
3160         expectations files were fatal and we didn't have a separate step
3161         on the bots that would lint the files to ensure things were still
3162         okay, so this was probably the right thing to do, but it could
3163         cause some patches to fail the style check even though the patches
3164         themselves were fine.
3165
3166         However, now errors aren't always fatal, and we have a separate build
3167         step to lint the files. So, we should probably not punish the
3168         patches and the style checker should filter out "unrelated" errors.
3169
3170         In theory we should be comparing the output pre- and post-patch
3171         to filter out pre-existing errors, but I don't think the style
3172         checker has that ability.
3173
3174         * Scripts/webkitpy/style/checkers/test_expectations.py:
3175         (TestExpectationsChecker.__init__):
3176         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
3177         (TestExpectationsTestCase.assert_lines_lint):
3178
3179 2012-12-11  Dirk Pranke  <dpranke@chromium.org>
3180
3181         nrwt isn't rejecting unrecognized expectations
3182         https://bugs.webkit.org/show_bug.cgi?id=104701
3183
3184         Reviewed by Ojan Vafai.
3185
3186         NRWT is silently ignoring unrecognized expectations like "Text"
3187         and "Audio" (largely because I haven't fully removed support for
3188         the old syntax and keywords and restructured the code to do the
3189         proper checking). Temporarily change the tokenizer to reject
3190         unrecognized expectations until I can make those changes.
3191
3192         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
3193         (TestExpectationParser._tokenize_line):
3194         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
3195         (ExpectationSyntaxTests.test_unknown_expectation):
3196         (TestExpectationSerializationTests.disabled_test_string_whitespace_stripping):
3197
3198 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
3199
3200         Fix a typo in my previous commit that is causing test failures.
3201
3202         Reviewed by Gustavo Noronha Silva.
3203
3204         There was an extra bit of statement appended after the end of a conditional. Somehow
3205         this compiled and caused test failures.
3206
3207         * DumpRenderTree/gtk/EventSender.cpp:
3208         (continuousMouseScrollByCallback):
3209
3210 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
3211
3212         [GTK] Spurious stderr output from the event sender
3213         https://bugs.webkit.org/show_bug.cgi?id=104678
3214
3215         Reviewed by Gustavo Noronha Silva.
3216
3217         Eliminate some spurious stderr output from the EventSender by using an early return
3218         instead of an assertion that prints a warning to the console.
3219
3220         * DumpRenderTree/gtk/EventSender.cpp:
3221         (continuousMouseScrollByCallback): We don't support continuous scroll events that
3222         move by page, but we should just fail silently instead of printing an error.
3223
3224 2012-12-11  Joone Hur  <joone.hur@intel.com>
3225
3226         [GTK] Add GraphicsLayerActor
3227         https://bugs.webkit.org/show_bug.cgi?id=91940
3228
3229         Reviewed by Gustavo Noronha Silva.
3230
3231         Add GraphicsLayerActor to the list of classes that contain GObjects
3232         in the style checker script so that it does not complain about GObject
3233         coding style.
3234
3235         * Scripts/webkitpy/style/checker.py:
3236
3237 2012-12-11  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
3238
3239         [Qt] remove hack which avoids unused .obj directories all over the place
3240
3241         Reviewed by Simon Hausmann.
3242
3243         The feature is subject to upstreaming in Qt itself in
3244         https://codereview.qt-project.org/#change,42011
3245
3246         * qmake/mkspecs/features/default_post.prf:
3247
3248 2012-12-11  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
3249
3250         [Qt] move SQLITE3SRCDIR assignment to only use site
3251
3252         Reviewed by Simon Hausmann.
3253
3254         Target.pri includes WebCore.pri the almost first thing, so that is
3255         covered as well.
3256
3257         * qmake/mkspecs/features/default_pre.prf:
3258
3259 2012-12-11  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
3260
3261         [Qt] delete dead include paths
3262
3263         Reviewed by Simon Hausmann.
3264
3265         followup to https://bugs.webkit.org/show_bug.cgi?id=93446
3266
3267         * DumpRenderTree/qt/DumpRenderTree.pro:
3268
3269 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
3270
3271         [GTK] Implement some missing FrameLoaderClient message in DRT
3272         https://bugs.webkit.org/show_bug.cgi?id=104444
3273
3274         Reviewed by Gustavo Noronha Silva.
3275
3276         Install a FrameLoadEvent callback that prints the necessary output for some
3277         missing FrameLoader events.
3278
3279         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3280         (frameLoadEventCallback): Added.
3281         (createWebView): Install the FrameLoader event callback.
3282
3283 2012-12-11  Oswald Buddenhagen  <oswald.buddenhagen@digia.com>
3284
3285         [Qt] change setup of QMAKE_MOC
3286
3287         Reviewed by Simon Hausmann.
3288
3289         don't load(moc) explicitly - turns out that this breaks
3290         debug_and_release builds (the feature does not like being loaded before
3291         default_post has been loaded).
3292         so instead just do a minimal call to find the moc executable.
3293
3294         as it's in fact only two files which need it, so de-centralize this call
3295         for the sake of saving some cpu cycles spent in qmake.
3296
3297         * WebKitTestRunner/InjectedBundle/DerivedSources.pri:
3298         * qmake/mkspecs/features/default_pre.prf:
3299
3300 2012-12-11  Jocelyn Turcotte  <jocelyn.turcotte@digia.com>
3301
3302         [Qt] Fix the linking of debug builds on Mac with force_static_libs_as_shared
3303         https://bugs.webkit.org/show_bug.cgi?id=104559
3304
3305         Reviewed by Tor Arne Vestbø.
3306
3307         "contains(libraries, $$TARGET): return(true)" fails for intermediate libraries since
3308         at this point TARGET may contain the "_debug" suffix.
3309
3310         Fix it by moving the TARGET adjustment later in the file.
3311
3312         * qmake/mkspecs/features/default_post.prf:
3313
3314 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
3315
3316         [Coverity] [GTK] Remove some redundant null checks in DumpRenderTree
3317         https://bugs.webkit.org/show_bug.cgi?id=104570
3318
3319         Reviewed by Daniel Bates.
3320
3321         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3322         (descriptionSuitableForTestResult): Remove a redundant null check.
3323
3324 2012-12-11  Alberto Garcia  <agarcia@igalia.com>
3325
3326         Remove unused GErrors
3327         https://bugs.webkit.org/show_bug.cgi?id=104657
3328
3329         Reviewed by Martin Robinson.
3330
3331         Replace unused GErrors with null pointers.
3332
3333         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3334         (initializeFonts):
3335         * DumpRenderTree/gtk/EditingCallbacks.cpp:
3336         (dumpRange):
3337         * WebKitTestRunner/InjectedBundle/gtk/ActivateFontsGtk.cpp:
3338         (WTR::inititializeFontConfigSetting):
3339
3340 2012-12-11  Martin Robinson  <mrobinson@igalia.com>
3341
3342         [GTK] Bump the jhbuild libsoup version to fix an authentication issue
3343         https://bugs.webkit.org/show_bug.cgi?id=104448
3344
3345         Reviewed by Gustavo Noronha Silva.
3346
3347         Bump the libsoup version to one from the git repository. This requires updating
3348         GLib as well which in turn requires adjusting the CFLAGS we pass when compiling
3349         the scanobj step of the gtkdoc generation.
3350
3351         * gtk/generate-gtkdoc:
3352         * gtk/jhbuild.modules:
3353
3354 2012-12-11  Zan Dobersek  <zandobersek@gmail.com>
3355
3356         Unreviewed functionality fix for the GTK port.
3357
3358         Enable the MicroData feature support for the GTK port in the FeatureList module.
3359         This feature listing now again affects which features are enabled and which disabled
3360         on the development builds of the GTK port.
3361
3362         * Scripts/webkitperl/FeatureList.pm:
3363
3364 2012-12-11  Jinwoo Song  <jinwoo7.song@samsung.com>
3365
3366         [EFL][WK2] MiniBrowser is not built
3367         https://bugs.webkit.org/show_bug.cgi?id=104629
3368
3369         Reviewed by Gyuyoung Kim.
3370
3371         As we are finding an elementary package in Tools/MiniBrowser/efl/CMakeList.txt since r137203,
3372         we need to check if the package is found in the same place not in the Tools/CMakeList.txt.
3373
3374         * CMakeLists.txt:
3375         * MiniBrowser/efl/CMakeLists.txt:
3376
3377 2012-12-11  Xabier Rodriguez Calvar  <calvaris@igalia.com>
3378
3379         [GTK][jhbuild] Switch to GStreamer 1.0 build
3380         https://bugs.webkit.org/show_bug.cgi?id=91727
3381
3382         Reviewed by Philippe Normand.
3383
3384         Switch build-webkit --gtk to GStreamer 1.0 support and build the
3385         necessary GStreamer git modules from JHBuild.
3386
3387         * gtk/jhbuild.modules: Added GStreamer 1.0.3 build support.
3388
3389 2012-12-11  Zan Dobersek  <zandobersek@gmail.com>
3390
3391         [GTK] Feature enabling/disabling should be possible through build-webkit
3392         https://bugs.webkit.org/show_bug.cgi?id=99271
3393
3394         Reviewed by Gustavo Noronha Silva.
3395
3396         Refactor the code in webkitdirs.pm that builds an autotools project to generate
3397         autogen.sh arguments, make arguments and installation prefix from the passed-in
3398         parameters rather than generating all of that in build-webkit. The autogen.sh
3399         arguments now contain only enable/disable flags for options that are actually
3400         configurable in configure.ac, the flag value actually reflecting whether the feature
3401         is enabled or disabled in the feature list.
3402
3403         Other features are overridable through modifying the GNUmakefile.features.am file.
3404         All these features are now stored in the build directory and upon change trigger
3405         a rerun of the autogen.sh script, pretty much like the autogen.sh arguments do.
3406
3407         The override-feature-defines script is called by autogen.sh, before GNUmakefile.in is
3408         generated when calling autoreconf. Its task is to copy the GNUmakefile.features.am.in
3409         into GNUmakefile.features.am and modify it if there's a text file in the build directory
3410         that contains all the feature defines the build-webkit script has written. If the build
3411         is not done through build-webkit (and there's no file in build directory listing all the
3412         feature defines) the script does not advance further from copying.
3413
3414         This approach is taken to overcome the rigidness of the automake system as it's impossible
3415         to effectively generate and use GNUmakefile.features.am or even GNUmakefile.features file
3416         after the autoreconf command execution in autogen.sh.
3417
3418         * Scripts/build-jsc:
3419         (buildMyProject):
3420         * Scripts/build-webkit:
3421         * Scripts/webkitdirs.pm:
3422         (runAutogenForAutotoolsProjectIfNecessary):
3423         (mustReRunAutogen):
3424         (buildAutotoolsProject):
3425         (buildGtkProject):
3426         * Scripts/webkitperl/FeatureList.pm: Update the features that are currently enabled in
3427         Source/WebCore/GNUmakefile.features.am.in but aren't in the feature list.
3428         * Scripts/webkitpy/style/checker.py: Source/WebCore/GNUmakefile.features.am.in is recognized
3429         as a text file and tabulation errors are reported. Skip the file to suppress them.
3430         * gtk/override-feature-defines: Added.
3431         (copy_feature_defines_makefile):
3432         (adjust_feature_defines_makefile):
3433         (adjust_feature_defines_makefile.override_feature_define):
3434         (override_feature_defines):
3435
3436 2012-12-10  Yury Semikhatsky  <yurys@chromium.org>
3437
3438         Memory instrumentation: make sure each edge is reported only once
3439         https://bugs.webkit.org/show_bug.cgi?id=104630
3440
3441         Reviewed by Pavel Feldman.
3442
3443         Test that outgoing edges are reported only once if we come to the same
3444         object twice: first by an address of a base class and then by a real address
3445         of the object.
3446
3447         * TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:
3448
3449 2012-12-10  Dana Jansens  <danakj@chromium.org>
3450
3451         [chromium] Remove WebCompositorOutputSurface implementations from DRT
3452         https://bugs.webkit.org/show_bug.cgi?id=104046
3453
3454         Reviewed by Darin Fisher.
3455
3456         The classes are now provided through WebCompositorSupport.
3457
3458         * DumpRenderTree/DumpRenderTree.gypi:
3459         * DumpRenderTree/chromium/WebViewHost.cpp:
3460         (WebViewHost::createOutputSurface):
3461         * DumpRenderTree/chromium/WebViewHostOutputSurface.cpp: Removed.
3462         * DumpRenderTree/chromium/WebViewHostOutputSurface.h: Removed.
3463         * DumpRenderTree/chromium/WebViewHostSoftwareOutputDevice.cpp: Removed.
3464         * DumpRenderTree/chromium/WebViewHostSoftwareOutputDevice.h: Removed.
3465
3466 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3467
3468         change --no-launch-safari to --no-show-results in scripts
3469         https://bugs.webkit.org/show_bug.cgi?id=104298
3470
3471         Reviewed by Eric Seidel.
3472
3473         Since Safari only exists on the Apple ports, NRWT calls the
3474         command line flag --[no-]show-results instead. This updates
3475         the scripts to use the new flag (but not the buildbot config yet).
3476
3477         * Scripts/generate-coverage-data:
3478         * Scripts/webkit-tools-completion.sh:
3479         * Scripts/webkitpy/tool/steps/runtests.py:
3480         (RunTests.run):
3481         * Scripts/webkitpy/tool/steps/runtests_unittest.py:
3482
3483 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3484
3485         nrwt: fix mapping of AUDIO onto FAIL when printing unexpected results
3486         https://bugs.webkit.org/show_bug.cgi?id=104597
3487
3488         Reviewed by Ojan Vafai.
3489
3490         This was broken in the recent printer refactoring.
3491
3492         * Scripts/webkitpy/layout_tests/views/buildbot_results.py:
3493         (BuildBotPrinter.print_unexpected_results.add_result.is_expected):
3494         * Scripts/webkitpy/layout_tests/views/buildbot_results_unittest.py:
3495         (BuildBotPrinterTests.get_result):
3496         (BuildBotPrinterTests.test_print_unexpected_results):
3497         (BuildBotPrinterTests.test_print_unexpected_results.get_unexpected_results):
3498         (BuildBotPrinterTests):
3499
3500 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3501
3502         old-run-webkit-tests: add --show-results properly
3503         https://bugs.webkit.org/show_bug.cgi?id=104608
3504
3505         Reviewed by Eric Seidel.
3506
3507         Rework r137219 / bug 104601 to use the correct perl idiom.
3508
3509         * Scripts/old-run-webkit-tests:
3510
3511 2012-12-10  Aaron Colwell  <acolwell@chromium.org>
3512
3513         Unreviewed. Adding Aaron Colwell to the contributor list.
3514
3515         * Scripts/webkitpy/common/config/committers.py:
3516
3517 2012-12-10  Anders Carlsson  <andersca@apple.com>
3518
3519         Remove WebKit2 tests on Windows
3520         https://bugs.webkit.org/show_bug.cgi?id=104607
3521
3522         Reviewed by Tim Horton.
3523
3524         Stop building WebKit2 tests and remove associated files.
3525
3526         * TestWebKitAPI/win/InjectedBundleControllerWin.cpp: Removed.
3527         * TestWebKitAPI/win/PlatformUtilitiesWin.cpp: Removed.
3528         * TestWebKitAPI/win/PlatformWebViewWin.cpp: Removed.
3529         * TestWebKitAPI/win/TestWebKitAPI.sln: Removed.
3530         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
3531         * TestWebKitAPI/win/TestWebKitAPIGenerated.vcproj: Removed.
3532         * TestWebKitAPI/win/TestWebKitAPIGeneratedCommon.vsprops: Removed.
3533         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj: Removed.
3534         * TestWebKitAPI/win/WindowMessageObserver.h: Removed.
3535         * TestWebKitAPI/win/copy-resources.cmd: Removed.
3536
3537 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3538
3539         old-run-webkit-tests: support --no-show-results for compatibility w/ NRWT
3540         https://bugs.webkit.org/show_bug.cgi?id=104601
3541
3542         Reviewed by Eric Siedel.
3543
3544         This flag is being added so we can move all of the calling scripts
3545         over to using --no-show-results instead of --no-launch-safari.
3546
3547         * Scripts/old-run-webkit-tests:
3548
3549 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3550
3551         nrwt: use the default logic for skipping platform-specific tests for chromium
3552         https://bugs.webkit.org/show_bug.cgi?id=104403
3553
3554         Reviewed by Stephen White.
3555
3556         I thought I had fixed this a while ago, but the chromium port was
3557         using custom logic to figure out which platform/* tests to run (or
3558         skip) and as a result we were getting some odd behavior (like running
3559         tests in chromium-linux on chromium-mac, but only when there was
3560         an expectation listed for the test).
3561
3562         Ports are supposed to only run the tests found under their
3563         baseline_search_path() by default.
3564
3565         Not writing a test for this; theoretically there shouldn't been a
3566         test checking that the behavior for Chromium was *different* than
3567         the other ports before.
3568
3569         * Scripts/webkitpy/layout_tests/port/chromium.py:
3570         (ChromiumPort._missing_symbol_to_skipped_tests):
3571
3572 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3573
3574         fix regression in test-webkitpy after r137189
3575         https://bugs.webkit.org/show_bug.cgi?id=104589
3576
3577         Reviewed by Eric Seidel.
3578
3579         Need to specify full port names.
3580
3581         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3582         (MainTest.test_build_check):
3583
3584 2012-12-10  Laszlo Gombos  <l.gombos@samsung.com>
3585
3586         [EFL] Change the minimum required EFL version to 1.6
3587         https://bugs.webkit.org/show_bug.cgi?id=104431
3588
3589         Reviewed by Kenneth Rohde Christiansen.
3590
3591         Change the minimum required EFL version to 1.6 from 1.7.
3592         The elementary EFL package is only required to build MiniBrowser,
3593         so I moved the required only to the Minibrowser CMake file.
3594
3595         * MiniBrowser/efl/CMakeLists.txt: In addition removed
3596         ENABLE_GLIB_SUPPORT as it is obsolete and it was always set to be on.
3597
3598 2012-12-10  Dean Jackson  <dino@apple.com>
3599
3600         Unreviewed. Adding Antoine to the contributor list so his name autocompletes in bugzilla.
3601
3602         * Scripts/webkitpy/common/config/committers.py:
3603
3604 2012-12-10  Zan Dobersek  <zandobersek@gmail.com>
3605
3606         REGRESSION (r118735): svg/dom/complex-svgView-specification.html, svg/dom/SVGViewSpec.html, svg/dom/viewspec-parser.html failing on GTK Linux 64-bit Release
3607         https://bugs.webkit.org/show_bug.cgi?id=87734
3608
3609         Reviewed by Martin Robinson.
3610
3611         When willSendRequest should return null in testing, cancel the SoupMessage
3612         instead of setting the request's URI to about:blank.
3613
3614         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3615         (willSendRequestCallback):
3616
3617 2012-12-10  Jochen Eisinger  <jochen@chromium.org>
3618
3619         [chromium] Move the test implementation of WebKit::WebPlugin to the TestRunner library
3620         https://bugs.webkit.org/show_bug.cgi?id=104537
3621
3622         Reviewed by Tony Chang.
3623
3624         In order to hide implementation details from the public TestRunner API,
3625         I moved the plugin to WebTestPluginImpl and added a static create
3626         method to WebTestPlugin. Also, I renamed the plugin from TestWebPlugin
3627         to WebTestPlugin to be more in line with other classes in the
3628         TestRunner API.
3629
3630         * DumpRenderTree/DumpRenderTree.gypi:
3631         * DumpRenderTree/chromium/TestRunner/public/WebTestPlugin.h: Added.
3632         (WebTestRunner):
3633         (WebTestPlugin):
3634         * DumpRenderTree/chromium/TestRunner/src/WebTestPlugin.cpp: Renamed from Tools/DumpRenderTree/chromium/TestWebPlugin.cpp.
3635         (WebTestRunner::WebTestPlugin::create):
3636         (WebTestRunner):
3637         (WebTestRunner::WebTestPlugin::WebTestPlugin):
3638         (WebTestRunner::WebTestPlugin::~WebTestPlugin):
3639         (WebTestRunner::WebTestPlugin::mimeType):
3640         * DumpRenderTree/chromium/TestWebPlugin.h: Removed.
3641         * DumpRenderTree/chromium/WebViewHost.cpp:
3642         (WebViewHost::createPlugin):
3643
3644 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3645
3646         run_webkit_tests throws exception when DRT is not present
3647         https://bugs.webkit.org/show_bug.cgi?id=104565
3648
3649         Reviewed by Jochen Eisinger.
3650
3651         Don't actually try to print the bot results if there aren't
3652         any because we exited before running any tests.
3653
3654         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3655         (main):
3656         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3657         (MainTest.test_build_check):
3658
3659 2012-12-10  Benjamin Poulain  <benjamin@webkit.org>
3660
3661         Add convenience methods to use ListHashSet for a LRU cache
3662         https://bugs.webkit.org/show_bug.cgi?id=104499
3663
3664         Reviewed by Sam Weinig.
3665
3666         Test the new methods added to ListHashSet.
3667
3668         * TestWebKitAPI/CMakeLists.txt:
3669         * TestWebKitAPI/GNUmakefile.am:
3670         * TestWebKitAPI/TestWebKitAPI.gypi:
3671         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
3672         * TestWebKitAPI/Tests/WTF/ListHashSet.cpp: Added.
3673         (TestWebKitAPI):
3674         (TestWebKitAPI::TEST):
3675
3676 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3677
3678         webkit-patch print-expectations doesn't work right for platforms w/ shared expectations
3679         https://bugs.webkit.org/show_bug.cgi?id=104405
3680
3681         Reviewed by Eric Seidel.
3682
3683         If multiple ports shared a TestExpectations file, webkit-patch
3684         print-expectations would use the expectation for the first port
3685         using that file and print the result for all ports, rather than
3686         re-parse and print each port separately (e.g., we might print
3687         the chromium-linux expectation for a file for both chromium-linux
3688         and chromium-mac). Discovered while testing/debugging the fix
3689         for bug 104403.
3690
3691         * Scripts/webkitpy/layout_tests/port/test.py:
3692         (TestPort._skipped_tests_for_unsupported_features):
3693         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3694         (lint):
3695         * Scripts/webkitpy/tool/commands/queries.py:
3696         (PrintExpectations._model):
3697         * Scripts/webkitpy/tool/commands/queries_unittest.py:
3698         (PrintExpectationsTest.test_platform):
3699
3700 2012-12-10  Dirk Pranke  <dpranke@chromium.org>
3701
3702         nrwt: fix total number of tests computed in the result summary
3703         https://bugs.webkit.org/show_bug.cgi?id=104061
3704
3705         Reviewed by Jochen Eisinger.
3706
3707         Good thing I left this open, because I actually broke it again
3708         the same way in a merge. Fix the computation of the one_line_summary
3709         again, and add tests for it (also clean up some redundant tests).
3710
3711         Also, add constants into the test port code so that the knowledge
3712         of how many tests should fail and how is next to the list of tests,
3713         rather than a totally different place.
3714
3715         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
3716         (LayoutTestRunner.run_tests):
3717         * Scripts/webkitpy/layout_tests/port/test.py:
3718         (TestList.__getitem__):
3719         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3720         (get_test_results):
3721         (MainTest.setUp):
3722         (MainTest.test_basic):
3723         (MainTest.test_run_singly_actually_runs_tests):
3724         (MainTest.test_test_list_with_prefix):
3725         (MainTest.test_missing_and_unexpected_results_with_custom_exit_code.CustomExitCodePort):
3726         (MainTest.test_tolerance.ImageDiffTestPort):
3727         (EndToEndTest.parse_full_results):
3728
3729 2012-12-10  Mike West  <mkwst@chromium.org>
3730
3731         Log to console when script is blocked by sandbox attributes.
3732         https://bugs.webkit.org/show_bug.cgi?id=104365
3733
3734         Reviewed by Ojan Vafai.
3735
3736         * Scripts/webkitpy/performance_tests/perftest.py:
3737         (PerfTest):
3738             Ignoring the new console message in perftests, as it broke parsing
3739             for the Parser/html-parser.html microbenchmark.
3740
3741 2012-12-10  Filip Pizlo  <fpizlo@apple.com>
3742
3743         JSC profiling and debug dump code should use inferred names when possible
3744         https://bugs.webkit.org/show_bug.cgi?id=104519
3745
3746         Reviewed by Oliver Hunt.
3747
3748         The format I'm using for referring to a code block is now name#hash. For example,
3749         v8-crypto has something called bnpSquareTo#B5QFbU. The profiler allows you to use
3750         either the hash, the inferred name, or the combined hash and full name when referring
3751         to blocks.
3752
3753         * Scripts/display-profiler-output:
3754
3755 2012-12-09  Filip Pizlo  <fpizlo@apple.com>
3756
3757         Profiler should say things about OSR exits
3758         https://bugs.webkit.org/show_bug.cgi?id=104497
3759
3760         Reviewed by Oliver Hunt.
3761
3762         Adds support for displaying OSR exit information for full summary (just displays the
3763         counts and the number of recompilations), bytecode display (says which bytecodes
3764         exited), and DFG display (annotates disassembly with taken OSR exits and their
3765         counts).
3766
3767         * Scripts/display-profiler-output:
3768
3769 2012-12-10  Alexis Menard  <alexis@webkit.org>
3770
3771         [CSS3 Backgrounds and Borders] Remove CSS3_BACKGROUND feature flag.
3772         https://bugs.webkit.org/show_bug.cgi?id=104539
3773
3774         Reviewed by Antonio Gomes.
3775
3776         As discussed on webkit-dev it is not needed to keep this feature flag 
3777         as support for <position> type is a small feature that is already 
3778         implemented by three other UAs. It was useful while landing this 
3779         feature as partial bits were landed one after one.
3780
3781         * Scripts/webkitperl/FeatureList.pm:
3782         * qmake/mkspecs/features/features.pri:
3783
3784 2012-12-10  Martin Robinson  <mrobinson@igalia.com>
3785
3786         Build fix.
3787
3788         Add a patchfile missing from my previous commit.
3789
3790         * gtk/patches/harfbuzz-icu-detection-fix.patch: Added.
3791
3792 2012-12-10  Martin Robinson  <mrobinson@igalia.com>
3793
3794         REGRESSION(137127): Causes assertion failures on the debug bots
3795         https://bugs.webkit.org/show_bug.cgi?id=104549
3796
3797         Reviewed by Alejandro G. Castro.
3798
3799         After handling the new TextOutput message we should return the method
3800         handling messages received.
3801
3802         * WebKitTestRunner/TestInvocation.cpp:
3803         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle): Fix a missing early return.
3804
3805 2012-12-09  Gustavo Noronha Silva  <gustavo.noronha@collabora.com>
3806
3807         [Soup] utilize multipart/x-mixed-replace support recently added to libsoup
3808         https://bugs.webkit.org/show_bug.cgi?id=94515
3809
3810         Reviewed by Martin Robinson.
3811
3812         This change requires a very recent libsoup, so update our jhbuild
3813         requirements.
3814
3815         * gtk/jhbuild.modules: depend on soup version that adds the multipart/x-mixed-replace
3816         support
3817         * efl/jhbuild.modules: ditto
3818
3819 2012-12-10  Zoltan Nyul  <zoltan.nyul@intel.com>
3820
3821         Implement testRunner.setViewModeMediaFeature() in WebKitTestRunner
3822         https://bugs.webkit.org/show_bug.cgi?id=103886
3823
3824         Reviewed by Kenneth Rohde Christiansen.
3825
3826         Implement testRunner.setViewModeMediaFeature() in WebKitTestRunner.
3827         This API is required by layout tests.
3828
3829         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
3830         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
3831         (WTR::TestRunner::setViewModeMediaFeature):
3832         (WTR):
3833         * WebKitTestRunner/InjectedBundle/TestRunner.h:
3834         (TestRunner):
3835
3836 2012-12-10  Martin Robinson  <mrobinson@igalia.com>
3837
3838         [GTK] Bring Harfbuzz-ng support to Gtk
3839         https://bugs.webkit.org/show_bug.cgi?id=92098
3840
3841         Reviewed by Gustavo Noronha Silva.
3842
3843         Add HarfBuzz dependency.
3844
3845         * GNUmakefile.am: Add FreeType/HarfBuzz CFLAGS to the WebCore Internals library.
3846         * gtk/jhbuild.modules: Add HarfBuzz to the JHbuild and remove Pango, since it
3847         should no longer affect layout test results.
3848
3849 2012-12-10  Simon Hausmann  <simon.hausmann@digia.com>
3850
3851         [Qt] Use QLibraryInfo::LibraryExecutablesPath unconditionally
3852         https://bugs.webkit.org/show_bug.cgi?id=104541
3853
3854         Reviewed by Jocelyn Turcotte.
3855
3856         We now depend on a Qt 5 version that is guaranteed to have this API, so we
3857         can remove the configure checks for it.
3858
3859         * qmake/config.tests/libexecdir/libexecdir.cpp: Removed.
3860         * qmake/config.tests/libexecdir/libexecdir.pro: Removed.
3861         * qmake/mkspecs/features/features.prf:
3862
3863 2012-12-10  Víctor Manuel Jáquez Leal  <vjaquez@igalia.com>
3864
3865         REGRESSION(r135863): GtkLauncher --help no longer displays all help options
3866         https://bugs.webkit.org/show_bug.cgi?id=104382
3867
3868         Reviewed by Philippe Normand.
3869
3870         Remove gst_init() because it steals the GOptionContext.
3871         gst_init_get_option_group() already does the initialization job.
3872
3873         * GtkLauncher/main.c:
3874         (main):
3875
3876 2012-12-10  Zan Dobersek  <zandobersek@gmail.com>
3877
3878         [GTK][WTR] Implement AccessibilityUIElement::titleUIElement
3879         https://bugs.webkit.org/show_bug.cgi?id=102953
3880
3881         Reviewed by Martin Robinson.
3882
3883         Implement AccessibilityUIElement::titleUIElement in the ATK backend
3884         in the same manner it is implemented in DumpRenderTree. Both implementations
3885         are updated to unref the relation set before the final function return
3886         rather than at the end of each loop through the relations held in the said
3887         relation set.
3888
3889         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
3890         (AccessibilityUIElement::titleUIElement):
3891         * WebKitTestRunner/InjectedBundle/atk/AccessibilityUIElementAtk.cpp:
3892         (WTR::AccessibilityUIElement::titleUIElement):
3893
3894 2012-12-10  Martin Robinson  <mrobinson@igalia.com>
3895
3896         Fix the WebKit2 debug build
3897
3898         * WebKitTestRunner/TestInvocation.cpp:
3899         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle): Fix a typo of WKStringGetTypeID().
3900
3901 2012-12-10  Martin Robinson  <mrobinson@igalia.com>
3902
3903         [WTR] Move text output accumulation to the UIProcess
3904         https://bugs.webkit.org/show_bug.cgi?id=104214
3905
3906         Reviewed by Darin Adler.
3907
3908         Instead of accumulating text output in the InjectedBundle and then sending it to the UIProcess 
3909         once a test is finished, immediately send any text output to the UIProcess. This will allow
3910         WebKitTestRunner to output text from the UIProcess as well.
3911
3912         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp: Remove the handling of the text output
3913         StringBuilder. Add a method to send output to the UIProcess.
3914         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
3915         (InjectedBundle):
3916         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp: Change code that appends to the StringBuilder
3917         to use the new outputText method.
3918         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
3919         * WebKitTestRunner/InjectedBundle/TestRunner.cpp: Ditto.
3920         * WebKitTestRunner/InjectedBundle/atk/AccessibilityControllerAtk.cpp: Ditto.
3921         * WebKitTestRunner/TestInvocation.cpp:
3922         (WTR::TestInvocation::TestInvocation): Intialize the StringBuilder.
3923         (WTR::TestInvocation::invoke): Clear the StringBuilder when a new tests starts.
3924         (WTR::TestInvocation::dumpResults): m_textOutput is a StringBuilder now so the
3925         method of printing the text is slightly different.
3926         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle): Handle the new TextOutput
3927         message by appending the results to the StringBuilder.
3928         * WebKitTestRunner/TestInvocation.h:
3929         (TestInvocation): Switch the type of m_textOutput.
3930
3931 2012-12-10  Mihnea Ovidenie  <mihnea@adobe.com>
3932
3933         Unreviewed. Adding Mihai Maerean to the list of contributors, working on CSSRegions.
3934
3935         * Scripts/webkitpy/common/config/committers.py:
3936
3937 2012-12-10  Joanmarie Diggs  <jdiggs@igalia.com>
3938
3939         [GTK] accessibility/language-attribute.html is failing
3940         https://bugs.webkit.org/show_bug.cgi?id=98369
3941
3942         Reviewed by Chris Fleizach.
3943
3944         The test is failing for two reasons: Not exposing the language
3945         for accessible objects and AccessibilityUIElement::language() not
3946         being implemented.
3947
3948         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
3949         (AccessibilityUIElement::language): Implemented.
3950
3951 2012-12-10  Joanmarie Diggs  <jdiggs@igalia.com>
3952
3953         [GTK] accessibility/editable-webarea-context-menu-point.html is failing
3954         https://bugs.webkit.org/show_bug.cgi?id=98364
3955
3956         Reviewed by Chris Fleizach.
3957
3958         The test is failing because the clickPoint values remain unchanged. The
3959         values remain unchanged because ATK lacks a means to keep track of this
3960         information.
3961
3962         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
3963         (AccessibilityUIElement::clickPointX): Added comment that ATK doesn't have clickPoint details
3964         (AccessibilityUIElement::clickPointY): Added comment that ATK doesn't have clickPoint details
3965
3966 2012-12-09  Alejandro Piñeiro  <apinheiro@igalia.com>
3967
3968         [GTK] lack of implemention of AccessibilityUIElementGtk::isSelectable and AccessibilityUIElementGtk::isMultiSelectable
3969         https://bugs.webkit.org/show_bug.cgi?id=104481
3970
3971         Reviewed by Martin Robinson.
3972
3973         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
3974         (AccessibilityUIElement::isSelectable): Implement this by checking
3975         the current state of the AtkObject.
3976         (AccessibilityUIElement::isMultiSelectable): Implement this by
3977         checking the current state of the AtkObject.
3978
3979 2012-12-09  Joanmarie Diggs  <jdiggs@igalia.com>
3980
3981         [GTK] accessibility/placeholder.html is failing
3982         https://bugs.webkit.org/show_bug.cgi?id=98373
3983
3984         Reviewed by Martin Robinson.
3985
3986         The test was failing because the placeholder text was not supported and
3987         AccessibilityUIElement::stringAttributeValue() was not implemented.
3988
3989         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
3990         (coreAttributeToAtkAttribute): New convenience method to convert WebCore attribute
3991         names into AtkObject attribute names
3992         (AccessibilityUIElement::stringAttributeValue): implemented
3993
3994 2012-12-09  Dan Winship  <danw@gnome.org>
3995
3996         [GTK] Bump libxml2 requirement in jhbuild.modules
3997         https://bugs.webkit.org/show_bug.cgi?id=104477
3998
3999         Reviewed by Martin Robinson.
4000
4001         Fedora 18's Apache depends on symbols in libxml2 2.9.0, so the
4002         tests won't run if we only have libxml2 2.8.0 installed in the
4003         jhbuild root.
4004
4005         * gtk/jhbuild.modules:
4006
4007 2012-12-09  Zan Dobersek  <zandobersek@gmail.com>
4008
4009         Remove testRunner.disableImageLoading()
4010         https://bugs.webkit.org/show_bug.cgi?id=98083
4011
4012         Reviewed by Gustavo Noronha Silva.
4013
4014         Remove the disableImageLoading methods and callback.
4015         No tests use the method anymore and all ports except Wx
4016         now support this behavior through overriding the
4017         WebKitDisplayImagesKey preference.
4018
4019         * DumpRenderTree/TestRunner.cpp:
4020         (TestRunner::staticFunctions):
4021         * DumpRenderTree/TestRunner.h:
4022         (TestRunner):
4023         * DumpRenderTree/blackberry/TestRunnerBlackBerry.cpp:
4024         (TestRunner::overridePreference): Add support for overriding the
4025         WebKitDisplayImagesKey preference which is now the preferred way
4026         of disabling image loading.
4027         * DumpRenderTree/efl/TestRunnerEfl.cpp:
4028         * DumpRenderTree/gtk/TestRunnerGtk.cpp:
4029         * DumpRenderTree/mac/TestRunnerMac.mm:
4030         * DumpRenderTree/qt/TestRunnerQt.cpp:
4031         * DumpRenderTree/qt/TestRunnerQt.h:
4032         (TestRunner):
4033         * DumpRenderTree/win/TestRunnerWin.cpp:
4034         * DumpRenderTree/wx/TestRunnerWx.cpp:
4035
4036 2012-12-09  Joanmarie Diggs  <jdiggs@igalia.com>
4037
4038         [GTK] accessibility/label-for-control-hittest.html is failing
4039         https://bugs.webkit.org/show_bug.cgi?id=98368
4040
4041         Reviewed by Martin Robinson.
4042
4043         The test was failing because AccessibilityUIElement::elementAtPoint was
4044         not implemented.
4045
4046         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
4047         (AccessibilityUIElement::elementAtPoint): implemented
4048
4049 2012-12-09  Jon Lee  <jonlee@apple.com>
4050
4051         [WK2] Move button image to injected bundle
4052         https://bugs.webkit.org/show_bug.cgi?id=104107
4053         <rdar://problem/12813842>
4054
4055         Reviewed by Simon Fraser.
4056
4057         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
4058         (WTR::InjectedBundlePage::InjectedBundlePage): Update the bundle page UI client.
4059
4060 2012-12-08  Halton Huo  <halton.huo@intel.com>
4061
4062         [CMake] Add CMake style checker
4063         https://bugs.webkit.org/show_bug.cgi?id=104240
4064
4065         Reviewed by Eric Seidel.
4066
4067         The CMake style checker is followed by existing style checkers like cpp, python.
4068         Unit test is added as well.
4069
4070         * Scripts/webkitpy/style/checker.py: Use CMakeChecker CMAKE type files(.cmake or CMakeLists.txt)
4071         (FileType):
4072         (CheckerDispatcher._file_type):
4073         (CheckerDispatcher._create_checker):
4074         * Scripts/webkitpy/style/checkers/cmake.py: Added.
4075         (CMakeChecker):
4076         (CMakeChecker.__init__): 
4077         (CMakeChecker.check):
4078         (CMakeChecker._process_line):
4079         (CMakeChecker._check_trailing_whitespace):
4080         (CMakeChecker._check_no_space_cmds):
4081         (CMakeChecker._check_one_space_cmds):
4082         (CMakeChecker._check_non_lowercase_cmd):
4083         (CMakeChecker._check_indent):
4084         * Scripts/webkitpy/style/checkers/cmake_unittest.py: Added.
4085         (CMakeCheckerTest):
4086         (CMakeCheckerTest.test_init):
4087         (CMakeCheckerTest.test_init._mock_handle_style_error):
4088         (CMakeCheckerTest.test_check):
4089         (CMakeCheckerTest.test_check._mock_handle_style_error):
4090
4091 2012-12-08  Kondapally Kalyan  <kalyan.kondapally@intel.com>
4092
4093         [EFL] Enable WebGL by default.
4094         https://bugs.webkit.org/show_bug.cgi?id=102991.
4095
4096         Reviewed by Kenneth Rohde Christiansen.
4097
4098         Enable WebGL by default.
4099
4100         * Scripts/webkitperl/FeatureList.pm:
4101
4102 2012-12-07  Scott Violet  <sky@chromium.org>
4103
4104         [chromium] Remove linux theme related files and switch to default
4105         https://bugs.webkit.org/show_bug.cgi?id=103897
4106
4107         Reviewed by Dimitri Glazkov.
4108
4109         * Scripts/webkitpy/common/config/build_unittest.py:
4110         (ShouldBuildTest): linux->default
4111
4112 2012-12-07  Mike Lattanzio  <mlattanzio@rim.com>
4113
4114         [BlackBerry] Update BlackBerry Port to use new Graphics::createBuffer BufferType enum values.
4115         https://bugs.webkit.org/show_bug.cgi?id=104121
4116
4117         Reviewed by Rob Buis.
4118
4119         Update createBuffer call to reflect new BufferType enum.
4120         Internal PR 254649
4121         Internal Review: Jakob Petsovits
4122
4123         DRT is always backed for maximum compatibility.
4124
4125         * DumpRenderTree/blackberry/PixelDumpSupportBlackBerry.cpp:
4126         (createBitmapContextFromWebView):
4127
4128 2012-12-07  Jochen Eisinger  <jochen@chromium.org>
4129
4130         Add 'CHROMIUM_DEVEL_SANDBOX' to the environment variables to copy for layout test drivers
4131         https://bugs.webkit.org/show_bug.cgi?id=104355
4132
4133         Reviewed by Dirk Pranke.
4134
4135         This enables chromium's content_shell driver to run with the full SUID
4136         sandbox.
4137
4138         * Scripts/webkitpy/layout_tests/port/base.py:
4139         (Port.to.setup_environ_for_server):
4140
4141 2012-12-07  KyungTae Kim  <ktf.kim@samsung.com>
4142
4143         [EFL] Fix tools for WebKit2-only builds
4144         https://bugs.webkit.org/show_bug.cgi?id=103904
4145
4146         Reviewed by Laszlo Gombos.
4147
4148         The libewebkit2.so should be used for library path for EFL WebKit2.
4149
4150         * Scripts/webkitdirs.pm:
4151         (builtDylibPathForName):
4152
4153 2012-12-07  Xiaobo Wang  <xbwang@torchmobile.com.cn>
4154
4155         [BlackBerry] DRT - Update implementation of TestRunner.setCustomPolicyDelegate()
4156         https://bugs.webkit.org/show_bug.cgi?id=104243
4157
4158         RIM PR 259337
4159
4160         Reviewed by Rob Buis.
4161
4162         1. Implement calculation of navigation originating in
4163         didDecidePolicyForNavigationAction().
4164         2. Implement policy delegate for resource response.
4165         3. Enable customPolicyDelegate if waitForPolicyDelegate() is called.
4166         4. Minor optional updates in dump() and dumpFrameLoadCallbacks().
4167
4168         The patch fixed the following tests:
4169         - http/tests/download
4170         - fast/forms/mailto/formenctype-attribute-button-html.html
4171         - fast/forms/mailto/formenctype-attribute-input-html.html
4172         - fast/encoding/mailto-always-utf-8.html
4173
4174         * DumpRenderTree/blackberry/DumpRenderTree.cpp:
4175         (BlackBerry::WebKit::DumpRenderTree::DumpRenderTree):
4176         (BlackBerry::WebKit::DumpRenderTree::resetToConsistentStateBeforeTesting):
4177         (BlackBerry::WebKit::DumpRenderTree::dump):
4178         (BlackBerry::WebKit::DumpRenderTree::didFinishLoadForFrame):
4179         (BlackBerry::WebKit::DumpRenderTree::didDecidePolicyForNavigationAction):
4180         (BlackBerry::WebKit::DumpRenderTree::didDecidePolicyForResponse):
4181         (BlackBerry::WebKit::DumpRenderTree::setCustomPolicyDelegate):
4182         * DumpRenderTree/blackberry/DumpRenderTreeBlackBerry.h:
4183         (DumpRenderTree):
4184         * DumpRenderTree/blackberry/TestRunnerBlackBerry.cpp:
4185         (TestRunner::setCustomPolicyDelegate):
4186         (TestRunner::waitForPolicyDelegate):
4187         (TestRunner::overridePreference):
4188
4189 2012-12-07  Gwang Yoon Hwang  <ryumiel@company100.net>
4190
4191         [Qt] Adjust library name in layout_tests/port/qt.py after Qt5-r40
4192         https://bugs.webkit.org/show_bug.cgi?id=104348
4193
4194         Reviewed by Csaba Osztrogonác.
4195
4196         After Qt-5.0.0-r40, the library was renamed from QtWebKitWidgets to
4197         Qt5WebKitWidgets. run-webkit-tests were still looking for QtWebKitWidgets
4198
4199         * Scripts/webkitpy/layout_tests/port/qt.py:
4200         (QtPort._path_to_webcore_library):
4201
4202 2012-12-07  Csaba Osztrogonác  <ossy@webkit.org>
4203
4204         Unreviewed fix, unbreak run-launcher after library renaming.
4205
4206         * Scripts/webkitdirs.pm:
4207         (builtDylibPathForName):
4208
4209 2012-12-06  Csaba Osztrogonác  <ossy@webkit.org>
4210
4211         Unreviewed fix after r136896, unbreak --debug-rwt to make bots result verbose again.
4212         https://bugs.webkit.org/show_bug.cgi?id=104296
4213
4214         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
4215         (main):
4216
4217 2012-12-06  Rick Byers  <rbyers@chromium.org>
4218
4219         CSS cursor property should support webkit-image-set
4220         https://bugs.webkit.org/show_bug.cgi?id=99493
4221
4222         Reviewed by Beth Dakin.
4223
4224         Add ENABLE_MOUSE_CURSOR_SCALE - disabled by default
4225
4226         * Scripts/webkitperl/FeatureList.pm:
4227         * qmake/mkspecs/features/features.pri:
4228
4229 2012-12-06  Vivek Galatage  <vivek.vg@samsung.com>
4230
4231         Unreviewed. Fixing the email address order to work correctly with bugzilla.
4232
4233         * Scripts/webkitpy/common/config/committers.py:
4234
4235 2012-12-06  Jochen Eisinger  <jochen@chromium.org>
4236
4237         [chromium] when forwarding events through the WebPluginContainer, create a UserGestureIndicator if processing a user gesture
4238         https://bugs.webkit.org/show_bug.cgi?id=104306
4239
4240         Reviewed by Dimitri Glazkov.
4241
4242         * DumpRenderTree/chromium/TestWebPlugin.cpp:
4243         (TestWebPlugin::TestWebPlugin):
4244         (TestWebPlugin::handleInputEvent): add support for printing the user gesture status
4245         * DumpRenderTree/chromium/TestWebPlugin.h:
4246         (TestWebPlugin):
4247
4248 2012-12-06  Dirk Pranke  <dpranke@chromium.org>
4249
4250         new-run-webkit-tests --lint-test-files seems to be broken
4251         https://bugs.webkit.org/show_bug.cgi?id=104296
4252
4253         Unreviewed, build fix.
4254
4255         My recent refactoring caused me to not initialize the printer
4256         before calling lint(); this patch restructures the code to make
4257         the lint routine self-contained and do all the work necessary.
4258         This will also make it easier to move this code out into a standalone
4259         file.
4260
4261         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
4262         (lint):
4263         (main):
4264         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
4265         (LintTest.test_all_configurations):
4266         (LintTest.test_lint_test_files):
4267         (LintTest.test_lint_test_files__errors):
4268
4269 2012-12-05  Filip Pizlo  <fpizlo@apple.com>
4270
4271         DFG profiler should be helpful about gem installation
4272         https://bugs.webkit.org/show_bug.cgi?id=104217
4273
4274         Reviewed by Oliver Hunt.
4275         
4276         'json' and 'highline' are not installed by default on the ruby distributions with
4277         which I am familiar, and the default error messages don't make me happy. This makes
4278         display-profiler-output print a helpful message if those gems are not found.
4279
4280         * Scripts/display-profiler-output:
4281
4282 2012-12-05  Filip Pizlo  <fpizlo@apple.com>
4283
4284         Profiler should print a helpful message if you pass the wrong arguments
4285         https://bugs.webkit.org/show_bug.cgi?id=104222
4286
4287         Reviewed by Oliver Hunt.
4288
4289         * Scripts/display-profiler-output:
4290
4291 2012-12-06  Ojan Vafai  <ojan@chromium.org>
4292
4293         Allow for CCing a secondary email address from watchlists
4294         https://bugs.webkit.org/show_bug.cgi?id=104286
4295
4296         Reviewed by Dirk Pranke.
4297
4298         Change my watchlist CC address so I can filter these separately from
4299         when people explicitly CC me and update the watchlist parser to allow that.
4300
4301         * Scripts/webkitpy/common/config/committers.py:
4302         * Scripts/webkitpy/common/config/watchlist:
4303         * Scripts/webkitpy/common/watchlist/watchlistparser.py:
4304         (WatchListParser._validate):
4305         * Scripts/webkitpy/common/watchlist/watchlistparser_unittest.py:
4306         (WatchListParserTest.test_cc_rule_with_invalid_email):
4307         (WatchListParserTest.test_cc_rule_with_secondary_email):
4308
4309 2012-12-06  Max Vujovic  <mvujovic@adobe.com>
4310
4311         Unreviewed. Add myself to committers.py.
4312
4313         * Scripts/webkitpy/common/config/committers.py:
4314
4315 2012-12-06  Yael Aharon  <yael.aharon@intel.com>
4316
4317         [EFL] [WK2] fast/dom/vertical-scrollbar-in-rtl.html makes fast/regions tests flaky
4318         https://bugs.w