[Gtk] Add a configuration option for disabling unstable features in releases
[WebKit-https.git] / Tools / ChangeLog
1 2012-07-06  Zan Dobersek  <zandobersek@gmail.com>
2
3         [Gtk] Add a configuration option for disabling unstable features in releases
4         https://bugs.webkit.org/show_bug.cgi?id=87995
5
6         Reviewed by Martin Robinson.
7
8         Pass the --enable-unstable-features flag when building the Gtk port through build-webkit.
9
10         * Scripts/webkitdirs.pm:
11         (buildAutotoolsProject):
12
13 2012-07-06  Jessie Berlin  <jberlin@apple.com>
14
15         WKContext should ask for its initialization data when it needs it so the client doesn't have
16         to keep it up to date.
17         https://bugs.webkit.org/show_bug.cgi?id=90627
18
19         Reviewed by Anders Carlsson.
20
21         Add tests and update other WKContextInjectedBundleClients.
22
23         * MiniBrowser/mac/AppDelegate.m:
24         (-[BrowserAppDelegate init]):
25         Updated for the change to WKContextInjectedBundleClient.
26         * WebKitTestRunner/TestController.cpp:
27         (WTR::TestController::initialize):
28         Ditto.
29
30         * TestWebKitAPI/PlatformUtilities.cpp:
31         (TestWebKitAPI::Util::createInitializationDictionaryForInjectedBundleTest):
32         Moved the logic to create the initialization dictionary here ...
33         (TestWebKitAPI::Util::createContextForInjectedBundleTest):
34         ... from here so that it can be used without automatically using
35         WKContextSetInitializationUserDataForInjectedBundle.
36         * TestWebKitAPI/PlatformUtilities.h:
37
38         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
39         Add the new test files.
40         * TestWebKitAPI/GNUmakefile.am:
41         Ditto.
42
43         * TestWebKitAPI/Tests/WebKit2/GetInjectedBundleInitializationUserDataCallback.cpp: Added.
44         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
45         Check that the message received from the injected bundle matches the user data it was
46         initialized with.
47         (TestWebKitAPI::getInjectedBundleInitializationUserData):
48         Return the user data that the injected bundle should be initialized with.
49         (TestWebKitAPI::TEST):
50         Set up WKContextInjectedBundleClient and load a page.
51         * TestWebKitAPI/Tests/WebKit2/GetInjectedBundleInitializationUserDataCallback_Bundle.cpp: Added.
52         (TestWebKitAPI::GetInjectedBundleInitializationUserDataCallbackTest::initialize):
53         Send the initialization user data back up to the UI Process.
54
55         * TestWebKitAPI/Tests/WebKit2/InjectedBundleInitializationUserDataCallbackWins.cpp: Added.
56         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
57         Check that the message received from the injected bundle matches the user data it was
58         initialized with in the callback.
59         (TestWebKitAPI::getInjectedBundleInitializationUserData):
60         Return the user data that the injected bundle should be initialized with.
61         (TestWebKitAPI::TEST):
62         Set up the context and use WKContextSetInitializationUserDataForInjectedBundle to set the
63         initialization user data (which should be overridden by the user data returned in
64         getInjectedBundleInitializationUserData).
65         * TestWebKitAPI/Tests/WebKit2/InjectedBundleInitializationUserDataCallbackWins_Bundle.cpp: Added.
66         (TestWebKitAPI::InjectedBundleInitializationUserDataCallbackWinsTest::initialize):
67         Send the initialization user data back up to the UI Process.
68
69 2012-07-06  Dongwoo Im  <dw.im@samsung.com>
70
71         [EFL][GTK] jhbuild : Disable pixman demos build depending on GTK+
72         https://bugs.webkit.org/show_bug.cgi?id=90593
73
74         Reviewed by Philippe Normand.
75
76         * efl/jhbuild.modules: Give '--enable-gtk=no' option to pixman.
77         * gtk/jhbuild.modules: Give '--enable-gtk=no' option to pixman.
78
79 2012-07-06  Luciano Wolf <luciano.wolf@openbossa.org>
80
81         [Qt] DumpRenderTree does not use 'monospace' font when directed
82         https://bugs.webkit.org/show_bug.cgi?id=85203
83
84         The test fonts used for Qt tests were moved to the Liberation font family.
85
86         Due to this change we are skipping tons of tests. They will be gradually
87         unskipped in batches, ASAP.
88
89         Also, the now unneeded special font configuration for Qt5 WK1 was removed.
90
91         Reviewed by Tor Arne Vestbø.
92
93         * DumpRenderTree/qt/QtInitializeTestFonts.cpp:
94         (WebKit::initializeTestFonts):
95
96 2012-07-06  Csaba Osztrogonác  <ossy@webkit.org>
97
98         [Qt] Buildfix for newer Qt5
99         https://bugs.webkit.org/show_bug.cgi?id=90519
100
101         Reviewed by Tor Arne Vestbø.
102
103         * MiniBrowser/qt/MiniBrowser.pro: We also need to depend on gui-private in the MiniBrowser to get access to these headers.
104         * MiniBrowser/qt/MiniBrowserApplication.h: Include qpa/qwindowsysteminterface.h instead of deprecated qwindowsysteminterface_qpa.h
105         * WebKitTestRunner/qt/PlatformWebViewQt.cpp: Include qpa/qwindowsysteminterface.h instead of deprecated qwindowsysteminterface_qpa.h
106
107 2012-07-06  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
108
109         [Qt] Make use of Qt5 qmake's changed makefile recursion behavior
110
111         When not using the -r option, qmake now interleaves qmake and make calls,
112         so we don't need custom logic for this. There's also an option to supress
113         the effect of the -r option, which we use to make WebKit.pro the only
114         project file parsed in a recursive qmake-run.
115
116         https://bugs.webkit.org/show_bug.cgi?id=90461
117
118         Reviewed by Tor Arne Vestbø.
119
120         * qmake/mkspecs/features/default_post.prf:
121         * qmake/mkspecs/features/functions.prf:
122
123 2012-07-06  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
124
125         [Qt] Remove custom qmake logic for module creation
126
127         Qmake now has the necessary hooks to cleanly override the build locations.
128
129         https://bugs.webkit.org/show_bug.cgi?id=90461
130
131         Reviewed by Tor Arne Vestbø.
132
133         * qmake/mkspecs/features/default_pre.prf:
134
135 2012-07-06  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
136
137         [Qt] Add top-level .qmake.conf
138
139         With Qt5, this makes setting $QMAKEPATH externally unnecessary.
140
141         The magic in the perl scripts to set QMAKEPATH is still there, as it doesn't
142         hurt, and is still required for Qt4.
143
144         https://bugs.webkit.org/show_bug.cgi?id=90461
145
146         Reviewed by Tor Arne Vestbø.
147
148         * qmake/mkspecs/features/default_pre.prf:
149
150 2012-07-06  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
151
152         [Qt] Unify qtFeatureDefaults code paths
153
154         There's no reason to run qmake on features.prf directly anymore, as we
155         can selectivly run configure tests in the project file now.
156
157         https://bugs.webkit.org/show_bug.cgi?id=90461
158
159         Reviewed by Tor Arne Vestbø.
160
161         * Scripts/webkitdirs.pm:
162         (qtFeatureDefaults):
163         * qmake/configure.pri:
164         * qmake/mkspecs/features/features.prf:
165
166 2012-07-06  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
167
168         [Qt] Don't let qt_webkit.pri proclaim its own location
169
170         This won't work any more with recent Qt5 versions, as the forwarding
171         pri is created by qt_module_config, which needs MODULE_PRI to be set
172         up already.
173
174         We also need to load build_config, not qt_module.
175
176         https://bugs.webkit.org/show_bug.cgi?id=90461
177
178         Reviewed by Tor Arne Vestbø.
179
180         * qmake/qt_webkit.pri:
181
182 2012-07-06  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
183
184         [Qt] Remove redundant CONFIG+=module
185
186         qt_module_config takes care of that.
187
188         In api.pri we are actually testing the flag ourselves, so now we need to
189         test a related flag qt_module_config sets instead.
190
191         Reviewed by Tor Arne Vestbø.
192
193         * qmake/mkspecs/features/default_post.prf:
194
195 2012-07-06  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
196
197         [Qt] Switch to new-style Qt 5 configure tests
198
199         Use explicit project file action instead of syncqt magic.
200
201         https://bugs.webkit.org/show_bug.cgi?id=90461
202
203         Reviewed by Tor Arne Vestbø.
204
205         * Scripts/webkitdirs.pm:
206         (buildQMakeProjects):
207         * qmake/configure.pri: Copied from Tools/qmake/configure.pro.
208         * qmake/configure.pro:
209         * qmake/mkspecs/features/default_pre.prf:
210         * qmake/mkspecs/features/features.prf:
211         * qmake/sync.profile: Removed.
212
213 2012-07-05  Jason Liu  <jason.liu@torchmobile.com.cn>
214
215         Add Jason Liu to committers.py.
216         https://bugs.webkit.org/show_bug.cgi?id=90654
217
218         Unreviewed.
219
220         * Scripts/webkitpy/common/config/committers.py:
221
222 2012-07-05  Benjamin Poulain  <bpoulain@apple.com>
223
224         Double release of resources if the load is canceled in a callback of ResourceLoader::didFinishLoading
225         https://bugs.webkit.org/show_bug.cgi?id=90431
226
227         Reviewed by Anders Carlsson.
228
229         Add a Mac API test.
230
231         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
232         * TestWebKitAPI/Tests/mac/CancelLoadFromResourceLoadDelegate.html: Added.
233         * TestWebKitAPI/Tests/mac/CancelLoadFromResourceLoadDelegate.mm: Added.
234         (-[CancelLoadFromResourceLoadDelegate webView:resource:didFinishLoadingFromDataSource:]):
235         (-[CancelLoadFromResourceLoadDelegateFrameLoadDelegate webView:didFinishLoadForFrame:]):
236         (TestWebKitAPI):
237         (TestWebKitAPI::TEST):
238
239 2012-07-05  Dave Tharp  <dtharp@codeaurora.org>
240
241         Adding myself as committer to committers.py
242
243         Unreviewed.
244
245         * Scripts/webkitpy/common/config/committers.py:
246
247 2012-07-05  Alexey Proskuryakov  <ap@apple.com>
248
249         [Mac][WK2] Enable HTTPS tests
250         https://bugs.webkit.org/show_bug.cgi?id=90600
251
252         Reviewed by Brady Eidson.
253
254         * DumpRenderTree/mac/DumpRenderTree.mm: (dumpRenderTree): Removed a misleading comment.
255         It explained why we were doing this here instead of a frame load delegate, but that's
256         just history, not an explanation of why we need this code.
257
258         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm: (WTR::InjectedBundle::platformInitialize):
259         Ignore certificate errors for localhost and 127.0.0.1. Since this is only needed
260         for tests, it seems appropriate to have platform specific solutions in every WTR,
261         and not add a WebKit2 API.
262
263 2012-07-05  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
264
265         [Qt] Remove rpath-hacks in qt_webkit.pri
266
267         Having it in qt_webkit.pri meant that all client apps would inherit the rpath,
268         even if they were not using QtWebKit.
269
270         This is a regression for people who build client apps against a non-installed
271         QtWebKit, as the client app will be missing a rpath to the standalone WebKit
272         build dir. This can be solved by LD_LIBRARY_PATH or custom QMAKE_RPATHDIR
273         logic, until the real issue is fixed in Qt.
274
275         Reviewed by Csaba Osztrogonác.
276
277         * qmake/mkspecs/features/rpath.prf:
278         * qmake/qt_webkit.pri:
279
280 2012-07-05  Dongwoo Im  <dw.im@samsung.com>
281
282         [EFL] Enable the CUSTOM_SCHEME_HANDLER feature as default.
283         https://bugs.webkit.org/show_bug.cgi?id=88608
284
285         Reviewed by Hajime Morita.
286
287         * Scripts/webkitperl/FeatureList.pm: Enable the CUSTOM_SCHEME_HANDLER feature as default.
288
289 2012-07-05  Ryuan Choi  <ryuan.choi@samsung.com>
290
291         [Wk2][EFL] EFL needs a WebKitTestRunner
292         https://bugs.webkit.org/show_bug.cgi?id=87659
293
294         Reviewed by Kenneth Rohde Christiansen.
295
296         Add an implementation of WebKitTestRunner for Efl port.
297
298         * CMakeLists.txt: Includes WebKitTestRunner build script.
299         * Scripts/build-webkittestrunner: Added knowledge of WebKitTestRunner/Efl.
300         * Scripts/webkitpy/layout_tests/port/efl.py: Ditto.
301         (EflPort.setup_environ_for_server):
302         * WebKitTestRunner/CMakeLists.txt: Added.
303         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
304         * WebKitTestRunner/InjectedBundle/efl/ActivateFontsEfl.cpp: Added.
305         * WebKitTestRunner/InjectedBundle/efl/InjectedBundleEfl.cpp: Added.
306         * WebKitTestRunner/InjectedBundle/efl/LayoutTestControllerEfl.cpp: Added.
307         * WebKitTestRunner/PlatformEfl.cmake: Added.
308         * WebKitTestRunner/PlatformWebView.h:
309         * WebKitTestRunner/TestController.cpp:
310         * WebKitTestRunner/config.h: Added knowledge of WebKitTestRunner/Efl.
311         * WebKitTestRunner/efl/PlatformWebViewEfl.cpp: Added.
312         * WebKitTestRunner/efl/TestControllerEfl.cpp: Added.
313         * WebKitTestRunner/efl/main.cpp: Added.
314         (main):
315
316 2012-07-05  Christophe Dumez  <christophe.dumez@intel.com>
317
318         [WK2][EFL] Ewk_View needs API to load HTML data
319         https://bugs.webkit.org/show_bug.cgi?id=90540
320
321         Reviewed by Kenneth Rohde Christiansen.
322
323         Update EFL MiniBrowser to catch the "load,error" signal
324         on the view and display an error page.
325
326         * MiniBrowser/efl/main.c:
327         (on_error):
328         (browserCreate):
329
330 2012-07-05  Sergio Villar Senin  <svillar@igalia.com>
331
332         Hidden dirs are not copied when creating the built product archive
333         https://bugs.webkit.org/show_bug.cgi?id=90559
334
335         Reviewed by Carlos Garcia Campos.
336
337         Hidden dirs are not copied inside the built product archive
338         because the recursive copy command was ignoring them. We need those
339         hidden directories for the GTK WebKit2 testing bot to work fine.
340
341         * BuildSlaveSupport/built-product-archive:
342         (archiveBuiltProduct): replaced "*" by "." to include hidden
343         dirs/files.
344
345 2012-07-04  Yoshifumi Inoue  <yosin@chromium.org>
346
347         Unreviewed Chromium gardening, Disable always failed python tests on Windows.
348         https://bugs.webkit.org/show_bug.cgi?id=90587
349
350         PortTest.qt_linux and mac_lion are always failed on Chromium Windows at
351         creating child process.
352
353         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
354         (PortTest.disabled_test_qt_linux):
355         (PortTest.disabled_test_mac_lion):
356
357 2012-07-04  Yoshifumi Inoue  <yosin@chromium.org>
358
359         [Tools] webkit_unittest.py got assertion
360         https://bugs.webkit.org/show_bug.cgi?id=90579
361
362         Reviewed by Hajime Morita.
363
364         This patch updates expectation of assertion in test_skipped_directories_for_symbols().
365         r121874 added new (exclude) directory (inspector/styles/variables), however it didn't
366         update one of two assertions.
367
368         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
369
370 2012-07-04  Balazs Ankes  <bank@inf.u-szeged.hu>
371
372         webkit-patch should add reviewer if "Reviewed by NOBODY ..." is missing
373         https://bugs.webkit.org/show_bug.cgi?id=67935
374
375         Reviewed by Ryosuke Niwa.
376
377         * Scripts/webkitpy/common/checkout/changelog.py:
378         (ChangeLog.set_reviewer):
379         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
380         (test_set_reviewer):
381         (test_set_short_description_and_bug_url):
382
383 2012-07-04  Sheriff Bot  <webkit.review.bot@gmail.com>
384
385         Unreviewed, rolling out r121862.
386         http://trac.webkit.org/changeset/121862
387         https://bugs.webkit.org/show_bug.cgi?id=90569
388
389         It broke built-product-archive for GTK, Qt and EFL (Requested
390         by svillar on #webkit).
391
392         * BuildSlaveSupport/built-product-archive:
393         (archiveBuiltProduct):
394
395 2012-07-04  Sergio Villar Senin  <svillar@igalia.com>
396
397         Hidden dirs are not copied when creating the built product archive
398         https://bugs.webkit.org/show_bug.cgi?id=90559
399
400         Reviewed by Carlos Garcia Campos.
401
402         Hidden dirs are not copied inside the built product archive because
403         the recursive copy command was ignoring them. We need those hidden
404         directories for the GTK WebKit2 testing bot to work fine.
405
406         * BuildSlaveSupport/built-product-archive:
407         (archiveBuiltProduct): removed the trailing "*" from the copy
408         command.
409
410 2012-07-04  John Mellor  <johnme@chromium.org>
411
412         Text Autosizing: Add compile flag and runtime setting
413         https://bugs.webkit.org/show_bug.cgi?id=87394
414
415         This patch renames Font Boosting to Text Autosizing.
416
417         Reviewed by Adam Barth.
418
419         * Scripts/webkitperl/FeatureList.pm:
420         * qmake/mkspecs/features/features.pri:
421
422 2012-07-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
423
424         [Qt] Replace deprecated IN_PWD with PWD
425
426         Reviewed by Csaba Osztrogonác.
427
428         * qmake/mkspecs/features/default_pre.prf:
429
430 2012-07-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
431
432         [Qt] Don't try to build WTR twice
433
434         Reviewed by Csaba Osztrogonác.
435
436         * Tools.pro:
437
438 2012-07-04  Sheriff Bot  <webkit.review.bot@gmail.com>
439
440         Unreviewed, rolling out r121821.
441         http://trac.webkit.org/changeset/121821
442         https://bugs.webkit.org/show_bug.cgi?id=90551
443
444         This patch did not receive a high-quality review and has a
445         number of errors (Requested by abarth on #webkit).
446
447         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
448         (Build.results_url):
449         * Scripts/webkitpy/common/net/buildbot/chromiumbuildbot.py:
450         (ChromiumBuilder.accumulated_results_url):
451         * Scripts/webkitpy/layout_tests/port/builders.py:
452         * Scripts/webkitpy/tool/commands/rebaseline.py:
453         (AbstractParallelRebaselineCommand.__init__):
454         (Rebaseline):
455         (Rebaseline._builder_to_pull_from):
456         (Rebaseline._tests_to_update):
457         (Rebaseline.execute):
458         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
459         (test_rebaseline.mock_builder_to_pull_from):
460         (test_rebaseline):
461
462 2012-07-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
463
464         [Qt] Get rid of un-needed QT += declarative for Qt 5
465
466         The declarative module has been renamed to quick1 in Qt 5, and the
467         engine-only module for Qt 5 is named 'qml'. For many of the instances
468         we could just remove 'declarative', since the project file was only
469         used for Qt5/WebKit2 builds. In the other cases the module was wrapped
470         in a haveQt(4) scope.
471
472         Reviewed by Csaba Osztrogonác.
473
474         * MiniBrowser/qt/MiniBrowser.pro:
475         * WebKitTestRunner/InjectedBundle/Target.pri:
476         * WebKitTestRunner/Target.pri:
477
478 2012-07-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
479
480         [Qt] Remove a few un-needed load(features) after r121777
481
482         The features are computed by configure.pro and cached in .qmake.cache.
483
484         Reviewed by Jocelyn Turcotte..
485
486         * MiniBrowser/qt/raw/Target.pri:
487
488 2012-07-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
489
490         [Qt] Use haveQtModule() in project files instead of manual checks
491
492         Reviewed by Jocelyn Turcotte.
493
494         * qmake/mkspecs/features/default_pre.prf:
495         * qmake/mkspecs/features/features.prf:
496
497 2012-07-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
498
499         [Qt] Fix haveQtModule() check
500
501         Without {} brackets we were pulling out the value of the module.name variable.
502
503         Reviewed by Jocelyn Turcotte.
504
505         * qmake/mkspecs/features/functions.prf:
506
507 2012-07-04  Sergio Villar Senin  <svillar@igalia.com>
508
509         [WK2] [GTK] WK2 testing bot unable to find the InjectedBundle library
510         https://bugs.webkit.org/show_bug.cgi?id=90541
511
512         Reviewed by Carlos Garcia Campos.
513
514         We must add the contents of WebKitBuild/*/Libraries/ directory to
515         the file to be uploaded to b.w.o to successfully run WebKit2 tests
516         in the WebKit2 testing bot.
517
518         * BuildSlaveSupport/built-product-archive:
519         (archiveBuiltProduct): appended Libraries to the list of
520         neededDirectories.
521
522 2012-07-04  Sergio Villar Senin  <svillar@igalia.com>
523
524         [GTK] WebKit2 testing bot unable to find the build directory
525         https://bugs.webkit.org/show_bug.cgi?id=90523
526
527         Reviewed by Carlos Garcia Campos.
528
529         The presence of GNUmakefile was used to determine whether a given
530         directory was a valid build directory or not. That's not correct for
531         testing bots as we export only the minimum required stuff to run the
532         tests.
533
534         * gtk/common.py:
535         (get_build_path.is_valid_build_directory): added an extra check.:
536
537 2012-07-04  Joe Thomas  <joethomas@motorola.com>
538
539         Adding myself to committers list.
540
541         Unreviewed.
542
543         * Scripts/webkitpy/common/config/committers.py:
544
545 2012-07-03  Christophe Dumez  <christophe.dumez@intel.com>
546
547         [WK2][EFL] Ewk_View should report the load progress
548         https://bugs.webkit.org/show_bug.cgi?id=90457
549
550         Reviewed by Kenneth Rohde Christiansen.
551
552         Update EFL MiniBrowser so that it listens for the
553         "load,progress" on the Ewk_View and updates its
554         window title accordingly.
555
556         * MiniBrowser/efl/main.c:
557         (title_set):
558         (on_title_changed):
559         (on_progress):
560         (browserCreate):
561
562 2012-07-03  Dirk Pranke  <dpranke@chromium.org>
563
564         disable failing mock-chromium-mac python unit tests
565
566         Unreviewed, build fix.
567
568         Disable these tests until I can figure out why they're failing
569         on the bots.
570
571         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
572         (PortTest.disabled_test_chromium_mac_lion):
573         (PortTest.disabled_test_chromium_mac_lion_in_test_shell_mode):
574
575 2012-07-03  Ian Vollick  <vollick@chromium.org>
576
577         Add vollick to the list of committers
578         https://bugs.webkit.org/show_bug.cgi?id=90512
579
580         Reviewed by Adrienne Walker.
581
582         * Scripts/webkitpy/common/config/committers.py:
583
584 2012-07-03  Ojan Vafai  <ojan@chromium.org>
585
586         Improve webkit-patch rebaseline to work for more cases
587         https://bugs.webkit.org/show_bug.cgi?id=90504
588
589         Reviewed by Dirk Pranke.
590
591         -Makes it work for the build.chromium.org bots.
592         -Lets you rebaseline all builders instead of just one.
593         -Lets you pass in the builder or test to rebaseline.
594
595         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
596         (Build.results_url):
597         Make this work for build.chromium.org builders as well.
598
599         * Scripts/webkitpy/common/net/buildbot/chromiumbuildbot.py:
600         (ChromiumBuilder):
601         (ChromiumBuilder.latest_cached_build):
602         ChromiumBuilder doesn't support large parts of the Builder interface.
603         This provides the bare minimum for this patch to work. Eventually,
604         we should create a single interface that can be supported for both
605         sets of buildbots.
606
607         * Scripts/webkitpy/layout_tests/port/builders.py:
608         (builder_path_from_name):
609         (all_builder_names):
610         memoizing here is incorrect because the test override _exact_matches,
611         so these can return different values. In either case, I'm pretty sure these
612         are not remotely hot codepaths.
613
614         * Scripts/webkitpy/tool/commands/rebaseline.py:
615         (AbstractParallelRebaselineCommand.__init__):
616         (Rebaseline):
617         (Rebaseline.__init__):
618         (Rebaseline._builders_to_pull_from):
619         (Rebaseline._builder_with_name):
620         (Rebaseline._tests_to_update):
621         (Rebaseline.execute):
622         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
623         (test_rebaseline.mock_builders_to_pull_from):
624         (test_rebaseline):
625         (test_rebaseline_command_line_flags):
626         (test_rebaseline_multiple_builders):
627         (test_rebaseline_multiple_builders.mock_builders_to_pull_from):
628         (test_rebaseline_multiple_builders.mock_tests_to_update):
629
630 2012-07-03  Christophe Dumez  <christophe.dumez@intel.com>
631
632         [EFL] Enable CSS variables support at compile time
633         https://bugs.webkit.org/show_bug.cgi?id=90448
634
635         Reviewed by Kenneth Rohde Christiansen.
636
637         Turn on CSS_VARIABLES flag by default on EFL port.
638
639         * Scripts/webkitperl/FeatureList.pm:
640
641 2012-07-03  Dirk Pranke  <dpranke@chromium.org>
642
643         nrwt: clean up exception handling and make sure we log some more failures
644         https://bugs.webkit.org/show_bug.cgi?id=90503
645
646         Reviewed by Ojan Vafai.
647
648         There were several places where exceptions weren't getting
649         logged, most notably if you passed a bad value to --platform.
650         This change tests that and cleans things up a bit; more cleanup
651         will be possible when we rework the manager_worker_broker code.
652
653         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
654         (_BrokerConnection.raise_exception):
655         (_InlineWorkerConnection.raise_exception):
656         * Scripts/webkitpy/layout_tests/controllers/worker.py:
657         (Worker.run):
658         (Worker.kill_driver):
659         * Scripts/webkitpy/layout_tests/port/factory.py:
660         (PortFactory.get):
661         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
662         (run):
663         (main):
664         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
665         (MainTest.test_unsupported_platfrom):
666
667 2012-07-03  Dirk Pranke  <dpranke@chromium.org>
668
669         nrwt: fix mock port
670         https://bugs.webkit.org/show_bug.cgi?id=90500
671
672         Reviewed by Ojan Vafai.
673
674         The MockDRT code was never updated when we switched the chromium
675         ports to using "drt mode" by default. This change updates that
676         code, fixes a typo in the chromium port that went undetected
677         (default_test_timeout_ms -> default_timeout_ms), and adds tests
678         that actually exercise some of the mock ports. These tests are
679         useful in that they will exercise the port-specific code in an
680         end-to-end-manner, but they are a bit slow for some reason (>1s
681         each) that I need to look into.
682
683         * Scripts/webkitpy/layout_tests/port/chromium.py:
684         (ChromiumDriver.stop):
685         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
686         (ChromiumAndroidPort.default_timeout_ms):
687         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
688         (TestChromiumAndroidPort.test_default_timeout_ms):
689         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
690         (main):
691         (parse_options):
692         (MockTestShell):
693         (MockTestShell.output_for_test):
694         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
695         (MockDRTTest.assertTest):
696         (MockTestShellTest):
697         (MockTestShellTest.make_drt):
698         (MockTestShellTest.test_test_shell_parse_options):
699         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
700         (PortTest):
701         (PortTest.assert_mock_port_works):
702         (PortTest.test_chromium_mac_lion):
703         (PortTest.test_chromium_mac_lion_in_test_shell_mode):
704         (PortTest.test_qt_linux):
705         (PortTest.test_mac_lion):
706
707 2012-07-03  Dirk Pranke  <dpranke@chromium.org>
708
709         nrwt: make the worker class stand alone with a cleaner interface
710         https://bugs.webkit.org/show_bug.cgi?id=90409
711
712         Reviewed by Ojan Vafai.
713
714         Currently the Worker class derives from AbstractWorker, which is
715         kind of crufty and awkward; it would be better if we did not
716         rely on shared state.
717
718         This change changes that so that Worker derives from object, and
719         exposes the following interface:
720           __init__() - called in the manager process
721           safe_init() - called in the worker process to initialize
722             unpicklable state
723           handle() - a single routine to handle all messages
724           cleanup() - called so the worker can clean up
725
726         Also, all of the "administrative" messages that are handled by
727         the worker (notification of start/stop/etc.) move into
728         manager_worker_broker - this reduces worker.py to just handling
729         the mechanics of actually running each test.
730
731         For the moment, we do this by creating Yet Another wrapper/proxy
732         class in manager_worker_broker, but this will get simpler
733         shortly when the rest of m_w_b is cleaned up.
734
735         With this change worker is now in its new form but there will be
736         a follow-on change that cleans up some names and other minor
737         things.
738
739         This change is again mostly just moving things around and should
740         be covered by the (updated) existing tests.
741
742         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
743         (get):
744         (AbstractWorker.__init__):
745         (AbstractWorker.run):
746         (AbstractWorker):
747         (AbstractWorker.handle_stop):
748         (AbstractWorker.handle_test_list):
749         (AbstractWorker.yield_to_broker):
750         (AbstractWorker.post_message):
751         (_WorkerConnection.__init__):
752         (_Process.run):
753         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py:
754         (_TestWorker):
755         (_TestWorker.__init__):
756         (_TestWorker.name):
757         (_TestWorker.cleanup):
758         (_TestWorker.handle):
759         (_TestWorker.safe_init):
760         (_TestWorker.stop):
761         (_TestsMixin.handle_finished_test):
762         (_TestsMixin.setUp):
763         (_TestsMixin.test_cancel):
764         (_TestsMixin.test_done):
765         * Scripts/webkitpy/layout_tests/controllers/worker.py:
766         (Worker):
767         (Worker.__init__):
768         (Worker.safe_init):
769         (Worker.handle):
770
771 2012-07-03  Dirk Pranke  <dpranke@chromium.org>
772
773         nrwt: moving child process logging code into manager_worker_broker
774         https://bugs.webkit.org/show_bug.cgi?id=90408
775
776         Reviewed by Ojan Vafai.
777
778         Users of manager_worker_broker should not have to be aware of
779         whether they're in the same process or different processes and
780         configure logging themselves; mwb should hide this complexity.
781         We can't quite do this completely/correctly yet, since the
782         manager expects to get a list of messages to log, but this
783         change fixes the worker side of it, at least.
784
785         This is just moving code around, there is no new functionality
786         and this should be covered by existing tests.
787
788         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
789         (AbstractWorker.__init__):
790         (_WorkerConnection.__init__):
791         (_WorkerConnection.post_message):
792         (_WorkerConnection):
793         (_WorkerConnection.set_up_logging):
794         (_WorkerConnection.clean_up_logging):
795         (_InlineWorkerConnection.run):
796         (_MultiProcessWorkerConnection.run):
797         (_WorkerLogHandler):
798         (_WorkerLogHandler.__init__):
799         (_WorkerLogHandler.emit):
800         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py:
801         (_TestWorker.run):
802         (_TestsMixin.handle_done):
803         * Scripts/webkitpy/layout_tests/controllers/worker.py:
804         (Worker.__init__):
805         (Worker.run):
806         (Worker._run_test):
807         (Worker.cleanup):
808         (Worker.run_single_test):
809
810 2012-07-03  Tony Chang  <tony@chromium.org>
811
812         [chromium] Don't archive build files generated by VS2010
813         https://bugs.webkit.org/show_bug.cgi?id=90493
814
815         Reviewed by Ojan Vafai.
816
817         In particular, the pch (pre compiled header) files are gigantic, making
818         release.zip almost a 1gb in size.
819
820         * BuildSlaveSupport/built-product-archive:
821         (archiveBuiltProduct):
822
823 2012-07-03  Tony Chang  <tony@chromium.org>
824
825         [chromium] Unreviewed, update .gitignore to handle VS2010 files.
826
827         * .gitignore:
828         * DumpRenderTree/DumpRenderTree.gyp/.gitignore:
829         * TestWebKitAPI/TestWebKitAPI.gyp/.gitignore:
830
831 2012-07-03  Ojan Vafai  <ojan@chromium.org>
832
833         Have webkit-patch rebaseline use rebaseline-test-internal
834         https://bugs.webkit.org/show_bug.cgi?id=90491
835
836         Reviewed by Dirk Pranke.
837
838         This lets it handle new files, reduces duplicate code and lays the
839         groundwork for making rebaseline have a richer interface.
840
841         * Scripts/webkitpy/common/net/buildbot/buildbot_mock.py:
842         (MockBuild):
843         (MockBuild.__init__):
844         (MockBuilder.build):
845         * Scripts/webkitpy/tool/commands/rebaseline.py:
846         (AbstractParallelRebaselineCommand._files_to_add):
847         (Rebaseline):
848         (Rebaseline._builder_to_pull_from):
849         (Rebaseline._tests_to_update):
850         (Rebaseline.execute):
851         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
852         (test_overrides_are_included_correctly):
853         (test_rebaseline):
854         (test_rebaseline.mock_builder_to_pull_from):
855         (test_rebaseline.mock_tests_to_update):
856
857 2012-07-03  Ojan Vafai  <ojan@chromium.org>
858
859         Rename rebaseline-test to rebaseline-test-internal
860         https://bugs.webkit.org/show_bug.cgi?id=90485
861
862         Reviewed by Adam Barth.
863
864         It's now only used by other rebaseline commands. It's still useful to leave it
865         as it's own command to aid in debugging when something goes wrong. In a followup
866         patch, I'll make webkit-patch rebaseline cover any use-cases that rebaseline-test
867         might have covered.
868
869         We no longer need the --print-scm-changes option since the only caller always passes
870         that option in.
871
872         Also, make all the arguments command-line flags instead. Simplifies the code
873         a bit in my opinion.
874
875         * Scripts/webkitpy/tool/commands/rebaseline.py:
876         (RebaselineTest):
877         (RebaselineTest.__init__):
878         (RebaselineTest.execute):
879         (AbstractParallelRebaselineCommand._rebaseline_commands):
880         (RebaselineExpectations.execute):
881         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
882         (test_rebaseline_all):
883
884 2012-07-03  Balazs Kelemen  <kbalazs@webkit.org>
885
886         [Qt][WTR] Get rid of using DumpRenderTreeSupportQt
887         https://bugs.webkit.org/show_bug.cgi?id=90262
888
889         Unreviewed build fix.
890
891         Reverting r121550 manually because it caused build
892         break on ARM and Mac. It turned out that the Qt build
893         system is too broken for this change now, first we have
894         to address two issues:
895             * symbol hiding was forbidden accidentally on x86 Linux
896             * Qt lacks a separate library for test support
897
898         * Tools.pro:
899         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
900         (WTR::InjectedBundlePage::resetAfterTest):
901         (WTR::InjectedBundlePage::didClearWindowForFrame):
902         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
903
904 2012-07-03  Tony Chang  <tony@chromium.org>
905
906         [chromium] Handle trailing backslash on %VSINSTALLDIR%
907         https://bugs.webkit.org/show_bug.cgi?id=90410
908
909         Reviewed by Ojan Vafai.
910
911         * Scripts/webkitdirs.pm:
912         (buildChromiumVisualStudioProject): Handle VSINSTALLDIR ending in a backslash, the backslash escapes the closing double quote.
913
914 2012-07-02  Ojan Vafai  <ojan@chromium.org>
915
916         Make the skia_test_expectations.txt file optional.
917         https://bugs.webkit.org/show_bug.cgi?id=90400
918
919         Reviewed by Dirk Pranke.
920
921         It used to be optional. This regressed at some point. It's important that it be
922         optional so that webkit-patch commands work in a pure-webkit checkout for chromium bots.
923         Specifically, this was breaking webkit-patch rebaseline-test when it would go to update
924         TestExpectations.
925
926         * Scripts/webkitpy/layout_tests/port/chromium.py:
927         (ChromiumPort.expectations_files):
928         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
929         (ChromiumDriverTest.test_expectations_dict):
930
931 2012-07-03  Raphael Kubo da Costa  <rakuco@webkit.org>
932
933         [jhbuild][EFL] Bump libffi dependency.
934         https://bugs.webkit.org/show_bug.cgi?id=85373
935
936         Reviewed by Gustavo Noronha Silva.
937
938         Update libffi to 3.0.11, which fixes the build on some platforms
939         where 3.0.10 had problems.
940
941         * efl/jhbuild.modules:
942
943 2012-07-03  George Staikos  <staikos@webkit.org>
944
945         [BlackBerry] Enable microdata support for BlackBerry.
946         https://bugs.webkit.org/show_bug.cgi?id=90429
947
948         Reviewed by Rob Buis.
949
950         * Scripts/webkitperl/FeatureList.pm:
951
952 2012-07-03  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
953
954         [Qt] Make use of .qmake.cache for caching features
955
956         Instead of loading() features from the files that need them (and re-running
957         a bunch of checks), we now run feature detection as part of configure.pro,
958         and have build-webkit write the computed feature-defines and CONFIG to
959         .qmake.cache, which is then loaded by qmake _before_ even defaults_pre
960         when building WebKit.pro.
961
962         At some point we'll be able to selectivly prevent running of config tests
963         in configure.pro, which means we don't need a separate code-path for
964         the build-webkit --help case.
965
966         We should also move the code in build-webkit that now uses .webkit.config
967         to detect clean builds, to use .qmake.cache, since we now store the same
968         thing there.
969
970         Original patch by Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
971
972         Reviewed by Tor Arne Vestbø.
973
974         * DumpRenderTree/qt/DumpRenderTree.pro:
975         * QtTestBrowser/QtTestBrowser.pro:
976         * Scripts/webkitdirs.pm:
977         (qtFeatureDefaults):
978         (buildQMakeProjects):
979         * Tools.pro:
980         * WebKitTestRunner/InjectedBundle/DerivedSources.pri:
981         * WebKitTestRunner/InjectedBundle/Target.pri:
982         * WebKitTestRunner/Target.pri:
983         * qmake/.qmake.conf: Added.
984         * qmake/configure.pro:
985         * qmake/mkspecs/features/default_post.prf:
986         * qmake/mkspecs/features/default_pre.prf:
987         * qmake/mkspecs/features/features.prf:
988
989 2012-07-03  Nico Weber  <thakis@chromium.org>
990
991         -Wunused-private-field cleanup, linux edition
992         https://bugs.webkit.org/show_bug.cgi?id=90463
993
994         Reviewed by Stephen White.
995
996         * DumpRenderTree/chromium/DRTDevToolsAgent.h:
997         * DumpRenderTree/chromium/GamepadController.h:
998         (GamepadController):
999         * DumpRenderTree/chromium/NotificationPresenter.h:
1000         (NotificationPresenter::NotificationPresenter):
1001         (NotificationPresenter):
1002
1003 2012-07-03  Sheriff Bot  <webkit.review.bot@gmail.com>
1004
1005         Unreviewed, rolling out r121766.
1006         http://trac.webkit.org/changeset/121766
1007         https://bugs.webkit.org/show_bug.cgi?id=90465
1008
1009         It caused flakey build errors on the bots (Requested by Ossy
1010         on #webkit).
1011
1012         * DumpRenderTree/qt/DumpRenderTree.pro:
1013         * QtTestBrowser/QtTestBrowser.pro:
1014         * Scripts/webkitdirs.pm:
1015         (qtFeatureDefaults):
1016         (buildQMakeProjects):
1017         * Tools.pro:
1018         * WebKitTestRunner/InjectedBundle/DerivedSources.pri:
1019         * WebKitTestRunner/InjectedBundle/Target.pri:
1020         * WebKitTestRunner/Target.pri:
1021         * qmake/.qmake.conf: Removed.
1022         * qmake/configure.pro:
1023         * qmake/mkspecs/features/default_post.prf:
1024         * qmake/mkspecs/features/default_pre.prf:
1025         * qmake/mkspecs/features/features.prf:
1026
1027 2012-07-03  George Staikos  <staikos@webkit.org>
1028
1029         [BlackBerry] Enable Custom Scheme Handlers for BlackBerry.
1030         https://bugs.webkit.org/show_bug.cgi?id=90422
1031
1032         Reviewed by Rob Buis.
1033
1034         * Scripts/webkitperl/FeatureList.pm:
1035
1036 2012-07-03  George Staikos  <staikos@webkit.org>
1037
1038         [BlackBerry] Enable RegisterProtocolHandler for BlackBerry.
1039         https://bugs.webkit.org/show_bug.cgi?id=90422
1040
1041         Reviewed by Rob Buis.
1042
1043         * Scripts/webkitperl/FeatureList.pm:
1044
1045 2012-07-03  Thiago Marcos P. Santos  <thiago.santos@intel.com>
1046
1047         [EFL] Add run API tests as a step on the build bots
1048         https://bugs.webkit.org/show_bug.cgi?id=87252
1049
1050         Reviewed by Daniel Bates.
1051
1052         Run EFL API test suite on the build bots.
1053
1054         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
1055         (RunEflAPITests):
1056         (TestFactory.__init__):
1057         (BuildAndTestFactory.__init__):
1058         * Scripts/run-efl-tests: Added.
1059
1060 2012-07-03  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1061
1062         [Qt] Make use of .qmake.cache for caching features
1063
1064         Instead of loading() features from the files that need them (and re-running
1065         a bunch of checks), we now run feature detection as part of configure.pro,
1066         and have build-webkit write the computed feature-defines and CONFIG to
1067         .qmake.cache, which is then loaded by qmake _before_ even defaults_pre
1068         when building WebKit.pro.
1069
1070         At some point we'll be able to selectivly prevent running of config tests
1071         in configure.pro, which means we don't need a separate code-path for
1072         the build-webkit --help case.
1073
1074         We should also move the code in build-webkit that now uses .webkit.config
1075         to detect clean builds, to use .qmake.cache, since we now store the same
1076         thing there.
1077
1078         Original patch by Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
1079
1080         Reviewed by Tor Arne Vestbø.
1081
1082         * DumpRenderTree/qt/DumpRenderTree.pro:
1083         * QtTestBrowser/QtTestBrowser.pro:
1084         * Scripts/webkitdirs.pm:
1085         (qtFeatureDefaults):
1086         (buildQMakeProjects):
1087         * Tools.pro:
1088         * WebKitTestRunner/InjectedBundle/DerivedSources.pri:
1089         * WebKitTestRunner/InjectedBundle/Target.pri:
1090         * WebKitTestRunner/Target.pri:
1091         * qmake/.qmake.conf: Added.
1092         * qmake/configure.pro:
1093         * qmake/mkspecs/features/default_post.prf:
1094         * qmake/mkspecs/features/default_pre.prf:
1095         * qmake/mkspecs/features/features.prf:
1096
1097 2012-07-03  Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>  Joel Dillon <joel.dillon@codethink.co.uk>
1098
1099         [Qt][Win] Fix broken QtWebKit5.lib linking
1100         https://bugs.webkit.org/show_bug.cgi?id=88321
1101
1102         Reviewed by Kenneth Rohde Christiansen.
1103
1104         On windows the import/export definition is part of the symbol's signature.
1105         Define STATICALLY_LINKED_WITH_$$library for each dependend module
1106         that is being linked statically to make sure that they can be linked together
1107         later on.
1108
1109         Also do not compile Assertions.cpp in DumpRenderTree anymore since all the
1110         used symbols are exported and it would cause a duplicate symbols error.
1111
1112         * DumpRenderTree/qt/DumpRenderTree.pro:
1113         * qmake/mkspecs/features/default_post.prf:
1114
1115 2012-07-03  Szilard Ledan  <szledan@inf.u-szeged.hu>
1116
1117         master.cfg cleanup, pass BuildStep instances instead of BuildStep subclasses
1118         https://bugs.webkit.org/show_bug.cgi?id=89001
1119
1120         Reviewed by Csaba Osztrogonác.
1121
1122         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
1123         (UploadBuiltProduct.__init__):
1124         (UploadTestResults.__init__):
1125         (ExtractTestResults.__init__):
1126         (Factory.__init__):
1127         (BuildFactory.__init__):
1128         (TestFactory.__init__):
1129         (BuildAndTestFactory.__init__):
1130         (BuildAndPerfTestFactory.__init__):
1131         (BuildAndPerfTestWebKit2Factory.__init__):
1132         (DownloadAndPerfTestFactory.__init__):
1133
1134 2012-07-03  Ádám Kallai  <kadam@inf.u-szeged.hu>
1135
1136         Update unit test after r121724.
1137         https://bugs.webkit.org/show_bug.cgi?id=90444
1138
1139         Reviewed by Csaba Osztrogonác.
1140
1141         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
1142         (GardeningServerTest.test_rebaseline_new_port):
1143
1144 2012-07-03  Sheriff Bot  <webkit.review.bot@gmail.com>
1145
1146         Unreviewed, rolling out r121749.
1147         http://trac.webkit.org/changeset/121749
1148         https://bugs.webkit.org/show_bug.cgi?id=90441
1149
1150         Tests crash on the paralel NRWT Qt bot (Requested by Ossy on
1151         #webkit).
1152
1153         * Scripts/webkitpy/layout_tests/port/qt.py:
1154         (QtPort._build_driver):
1155         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
1156         (XvfbDriver._start):
1157         (XvfbDriver._start.x_filter):
1158         (XvfbDriver.stop):
1159
1160 2012-07-03  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
1161
1162         [NRWT] XvfbDriver should choose the next free display
1163         https://bugs.webkit.org/show_bug.cgi?id=88414
1164
1165         Reviewed by Dirk Pranke.
1166
1167         * Scripts/webkitpy/layout_tests/port/qt.py:
1168         (QtPort._driver_class):
1169         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
1170         (XvfbDriver.__init__):
1171         (XvfbDriver._start):
1172         (XvfbDriver._start.next_free_id):
1173         (XvfbDriver.stop):
1174
1175 2012-07-03  Christophe Dumez  <christophe.dumez@intel.com>
1176
1177         [EFL] Enable MICRODATA support
1178         https://bugs.webkit.org/show_bug.cgi?id=90377
1179
1180         Reviewed by Kenneth Rohde Christiansen.
1181
1182         Turn on MICRODATA flag for EFL port by default.
1183
1184         * Scripts/webkitperl/FeatureList.pm:
1185
1186 2012-07-02  George Staikos  <staikos@webkit.org>
1187
1188         [BlackBerry] Enable scoped style for BlackBerry.
1189         https://bugs.webkit.org/show_bug.cgi?id=90418
1190
1191         Reviewed by Rob Buis.
1192
1193         * Scripts/webkitperl/FeatureList.pm:
1194
1195 2012-07-02  Xiaobo Wang  <xbwang@torchmobile.com.cn>
1196
1197         [BlackBerry] Use PUBLIC_BUILD to enable/disable DRT
1198         https://bugs.webkit.org/show_bug.cgi?id=90271
1199
1200         Reviewed by George Staikos.
1201
1202         RIM PR #154707
1203
1204         Currently DRT code will be compiled only if ENABLE_DRT is set, and it's not
1205         defined by default.
1206         We should enable DRT by default unless PUBLIC_BUILD is set. In this way we don't
1207         need to rebuild webkit before running DRT.
1208
1209         * Scripts/webkitdirs.pm:
1210         (blackberryCMakeArguments):
1211
1212 2012-07-02  Ojan Vafai  <ojan@chromium.org>
1213
1214         webkit-patch rebaseline-expectations should share code with rebaseline-all
1215         https://bugs.webkit.org/show_bug.cgi?id=90413
1216
1217         Reviewed by Dirk Pranke.
1218
1219         Make them share code. In addition to reducing code duplication this makes
1220         rebaseline-expectations considerably faster by rebaselining in parallel.
1221
1222         * Scripts/webkitpy/tool/commands/rebaseline.py:
1223         (AbstractParallelRebaselineCommand):
1224         (AbstractParallelRebaselineCommand._run_webkit_patch):
1225         (AbstractParallelRebaselineCommand._rebaseline):
1226         (RebaselineJson):
1227         (RebaselineJson.execute):
1228         (RebaselineExpectations):
1229         (RebaselineExpectations._update_expectations_file):
1230         (RebaselineExpectations._tests_to_rebaseline):
1231         (RebaselineExpectations._add_tests_to_rebaseline_for_port):
1232         (RebaselineExpectations.execute):
1233         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
1234         (test_rebaseline_all):
1235         (test_rebaseline_expectations.run_in_parallel):
1236         (test_rebaseline_expectations):
1237
1238 2012-07-02  Xiaobo Wang  <xbwang@torchmobile.com.cn>
1239
1240         [BlackBerry] Update DumpRenderTree to have it work interactively in parallel
1241         https://bugs.webkit.org/show_bug.cgi?id=88326
1242
1243         Reviewed by Rob Buis.
1244
1245         1. Test name is sent to torch-launcher via PPS message(from host machine).
1246            So we get test list by monitoring and receiving PPS message instead of
1247            reading file index.drt.
1248         2. Torch-launcher create a <test file>.done file when it finished a test.
1249         3. We don't need to search for Ref-tests in DumpRenderTree.cpp any more. NRWT
1250            will get them for us.
1251
1252         * DumpRenderTree/blackberry/DumpRenderTree.cpp:
1253         (BlackBerry::WebKit::DumpRenderTree::DumpRenderTree):
1254         (BlackBerry::WebKit::DumpRenderTree::doneDrt):
1255         (BlackBerry::WebKit::DumpRenderTree::runRemainingTests):
1256         (BlackBerry::WebKit::DumpRenderTree::ensurePPS):
1257         (WebKit):
1258         (BlackBerry::WebKit::DumpRenderTree::handlePPSData):
1259         (BlackBerry::WebKit::DumpRenderTree::waitForTest):
1260         (BlackBerry::WebKit::DumpRenderTree::runTests):
1261         (BlackBerry::WebKit::DumpRenderTree::dump):
1262         * DumpRenderTree/blackberry/DumpRenderTreeBlackBerry.h:
1263         (DumpRenderTree):
1264
1265 2012-07-02  Adam Barth  <abarth@webkit.org>
1266
1267         Remove flashplugin-installer from the EWS image because it causes some tests to crash
1268         https://bugs.webkit.org/show_bug.cgi?id=90403
1269
1270         Reviewed by Tony Chang.
1271
1272         * EWSTools/cold-boot.sh:
1273
1274 2012-07-02  Leandro Gracia Gil  <leandrogracia@chromium.org>
1275
1276         [Chromium] Implement a Layout Test for editing/SurroundingText
1277         https://bugs.webkit.org/show_bug.cgi?id=82461
1278
1279         Reviewed by Ryosuke Niwa.
1280
1281         Add a new method to the layout test controller in order to retrieve the
1282         text surrounding a provided element.
1283
1284         * DumpRenderTree/chromium/LayoutTestController.cpp:
1285         (LayoutTestController::LayoutTestController):
1286         (LayoutTestController::textSurroundingElement):
1287         * DumpRenderTree/chromium/LayoutTestController.h:
1288         (LayoutTestController):
1289
1290 2012-07-02  Ojan Vafai  <ojan@chromium.org>
1291
1292         Delete unused rebaseline method in gardeningserver.py
1293         https://bugs.webkit.org/show_bug.cgi?id=90396
1294
1295         Reviewed by Eric Seidel.
1296
1297         As best I can tell, the only usage was in this unittest.
1298
1299         * Scripts/webkitpy/tool/servers/gardeningserver.py:
1300         (GardeningHTTPRequestHandler.updateexpectations):
1301         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
1302         (GardeningServerTest.test_rebaseline_new_port):
1303
1304 2012-07-02  Csaba Osztrogonác  <ossy@webkit.org>
1305
1306         REGRESSION(r121497): It switched off and broke many unittests
1307         https://bugs.webkit.org/show_bug.cgi?id=90371
1308
1309         Reviewed by Adam Barth.
1310
1311         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
1312         (ChromiumMacPortTest): Use snowleopard as os_version instead of leopard, because it isn't supported anymore.
1313         * Scripts/webkitpy/layout_tests/port/chromium_port_testcase.py:
1314         (ChromiumPortTestCase.test_all_test_configurations): Remove leopard testcases, because it isn't supported anymore.
1315         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
1316         (PortTestCase): Inherit class PortTestCase from unittest.TestCase instead of object.
1317
1318 2012-07-02  Ojan Vafai  <ojan@chromium.org>
1319
1320         Move rebaseline-all command from the gardening-server down into webkit-patch
1321         https://bugs.webkit.org/show_bug.cgi?id=90395
1322
1323         Reviewed by Adam Barth.
1324
1325         This is just moving code. It it in preparation for making rebaseline-expectations
1326         use the same code in order to get the parallelism benefits and reduces the amount
1327         of code we have for doing rebaselines.
1328
1329         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
1330         (CheckoutTest.test_apply_patch):
1331         Updated due to the change to executive_mock.
1332         * Scripts/webkitpy/common/system/executive_mock.py:
1333         (MockExecutive.run_command):
1334         Update to print out the input passed to stdin.
1335         * Scripts/webkitpy/tool/commands/download_unittest.py:
1336         Updated due to executive_mock change.
1337         * Scripts/webkitpy/tool/commands/rebaseline.py:
1338         (RebaselineAll):
1339         (RebaselineAll._run_webkit_patch):
1340         (RebaselineAll._builders_to_fetch_from):
1341         (RebaselineAll._rebaseline_commands):
1342         (RebaselineAll._files_to_add):
1343         (RebaselineAll._optimize_baselines):
1344         (RebaselineAll._rebaseline):
1345         (RebaselineAll.execute):
1346         All this code is just copy-pasted except for mechanical changes
1347         (e.g. self.server.tool --> self._tool) and the reading in of the 
1348         JSON from stdin instead of the post body.
1349         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
1350         (test_rebaseline_all):
1351         Copied the test-case out of gardeningserver_unittest.py.
1352         * Scripts/webkitpy/tool/servers/gardeningserver.py:
1353         (GardeningHTTPRequestHandler):
1354         (GardeningHTTPRequestHandler.rebaseline):
1355         (GardeningHTTPRequestHandler.rebaselineall):
1356         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
1357         (GardeningServerTest.test_rebaselineall):
1358         (GardeningServerTest.test_rebaselineall.run_command):
1359
1360 2012-07-02  Ojan Vafai  <ojan@chromium.org>
1361
1362         Remove Leopard support from the flakiness dashboard
1363         https://bugs.webkit.org/show_bug.cgi?id=90390
1364
1365         Reviewed by Adam Barth.
1366
1367         * TestResultServer/static-dashboards/flakiness_dashboard.js:
1368         (nonChromiumPlatform):
1369         (chromiumPlatform):
1370         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
1371         (test):
1372         * TestResultServer/static-dashboards/run-embedded-unittests.html:
1373         * TestResultServer/static-dashboards/run-unittests.html:
1374
1375 2012-07-02  Ojan Vafai  <ojan@chromium.org>
1376
1377         Fix posting from garden-o-matic. This broke in moving away from jquery's ajax method
1378         in http://trac.webkit.org/changeset/121617.
1379         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/net.js:
1380
1381 2012-07-02  Carlos Garcia Campos  <cgarcia@igalia.com>
1382
1383         [GTK] Read fonts path when running layout tests from alternative fonts dir when main dir doesn't exist
1384         https://bugs.webkit.org/show_bug.cgi?id=89437
1385
1386         Reviewed by Martin Robinson.
1387
1388         If main fonts directory doesn't exist, try with an alternative
1389         fonts directory at build directory.
1390
1391         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1392         (getOutputDir):
1393         (getFontsPath):
1394         (initializeFonts):
1395         * WebKitTestRunner/InjectedBundle/gtk/ActivateFontsGtk.cpp:
1396         (WTR::getOutputDir):
1397         (WTR):
1398         (WTR::getFontsPath):
1399         (WTR::inititializeFontConfigSetting):
1400
1401 2012-07-02  Carlos Garcia Campos  <cgarcia@igalia.com>
1402
1403         [GTK] Don't run the tests with jhbuild wrapper if it's already running under jhbuild
1404         https://bugs.webkit.org/show_bug.cgi?id=89435
1405
1406         Reviewed by Martin Robinson.
1407
1408         * Scripts/new-run-webkit-tests: Don't run the tests with the
1409         jhbuild wrapper if there isn't a Dependencies directory inside the
1410         build directory.
1411
1412 2012-06-29  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
1413
1414         [Qt] Simplify detection of non-installed module file
1415
1416         Has the additional advantage that we do not rely on additional information.
1417
1418         Reviewed by Tor Arne Vestbø.
1419
1420         * qmake/qt_webkit.pri:
1421
1422 2012-07-01  Christophe Dumez  <christophe.dumez@intel.com>
1423
1424         [EFL] Add Gamepad support
1425         https://bugs.webkit.org/show_bug.cgi?id=90170
1426
1427         Reviewed by Kenneth Rohde Christiansen.
1428
1429         * Scripts/webkitperl/FeatureList.pm: Turn on GAMEPAD flag
1430         by default for EFL port.
1431         * efl/jhbuild.modules: Bump dependency versions of EFL libs
1432         since the latest Eeze is required for gamepad support.
1433
1434 2012-07-01  Adam Barth  <abarth@webkit.org>
1435
1436         Unreviewed.
1437
1438         Using the default start cylinder for fdisk causes a warning about the
1439         partition not starting on physical sector boundary.  The Ubuntu forums
1440         recommend using a number that's divisible by 8, which is what we do in
1441         this patch.
1442
1443         * EWSTools/cold-boot.sh:
1444
1445 2012-07-01  Adam Barth  <abarth@webkit.org>
1446
1447         Unreviewed.
1448
1449         It turns out we need to use Ubuntu 10.04 to get the right image results
1450         for chromium-linux. This patch updates our scripts to be compatible
1451         with Ubuntu 10.04.
1452
1453         * EWSTools/cold-boot.sh:
1454             - fdisk doesn't have p and 1 as default commands in 10.04.
1455         * EWSTools/start-queue.sh:
1456             - git doesn't understanding the -B argument in 10.04. We've been
1457               using this change locally on the EC2 bots for a while.
1458
1459 2012-07-01  Adam Barth  <abarth@webkit.org>
1460
1461         Unreviewed.
1462
1463         * EWSTools/boot.sh:
1464             - We need to start the screen in detached mode so that we can run
1465               it remotely via ssh.
1466
1467 2012-07-01  Adam Barth  <abarth@webkit.org>
1468
1469         Unreviewed.
1470
1471         * EWSTools/cold-boot.sh:
1472             - Turns out we need to sudo this command in order for it to
1473               actually dimiss the EULA screen. :)
1474
1475 2012-06-29  Adam Barth  <abarth@webkit.org>
1476
1477         Turns out we need zip too.
1478
1479         * EWSTools/cold-boot.sh:
1480
1481 2012-06-29  Adam Barth  <abarth@webkit.org>
1482
1483         Add a cold-boot.sh script for the EWS
1484         https://bugs.webkit.org/show_bug.cgi?id=90330
1485
1486         Unreviewed.
1487
1488         * EWSTools/cold-boot.sh: Added.
1489             - This script can take us from a cold GCE instance to a running EWS
1490               bot in one fell swoop.
1491         * EWSTools/start-queue.sh:
1492             - The if-block at the top of this script was causing trouble. I
1493               removed it from the bots a while ago. Now that we're using SVN to
1494               cold-boot the EWS bots, we need to make this change in the repo.
1495
1496 2012-06-29  Luiz Agostini  <luiz.agostini@nokia.com>
1497
1498         [Qt][WK2] Private non-QtQuick API
1499         https://bugs.webkit.org/show_bug.cgi?id=84532
1500
1501         MiniBrowserRaw is an usage example for the QRawWebView API.
1502         It is only an example and is not fully implemented.
1503
1504         Reviewed by Noam Rosenthal.
1505
1506         * MiniBrowser/qt/raw/DerivedSources.pri: Added.
1507         * MiniBrowser/qt/raw/MiniBrowserRaw.pro: Added.
1508         * MiniBrowser/qt/raw/Target.pri: Added.
1509         * MiniBrowser/qt/raw/View.cpp: Added.
1510         * MiniBrowser/qt/raw/View.h: Added.
1511         * Scripts/webkitpy/style/checker.py:
1512         * Tools.pro:
1513
1514 2012-06-29  Ojan Vafai  <ojan@chromium.org>
1515
1516         Remove leopard bots from garden-o-matic
1517         https://bugs.webkit.org/show_bug.cgi?id=90328
1518
1519         Reviewed by Adam Barth.
1520
1521         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
1522         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures_unittests.js:
1523         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
1524
1525 2012-06-29  Ojan Vafai  <ojan@chromium.org>
1526
1527         garden-o-matic broken: TypeError: 'undefined' is not an object (evaluating 'buildLocations[currentIndex].url')
1528         https://bugs.webkit.org/show_bug.cgi?id=90243
1529
1530         Reviewed by Dirk Pranke.
1531
1532         jQuery was trying to be too smart and parsing the jsonp as json because of it's content-type.
1533         Excise jQuery and just use XHR directly since it's easier to maintain something where we control it
1534         all.
1535
1536         Not really sure how to unittest this. I tested it all manually of course.
1537
1538         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/net.js:
1539         Made net.ajax a drop-in replacement for the features of $.ajax that we were using.
1540         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:
1541         Not related to this patch, but figured I'd update the failing test while I was at it.
1542
1543 2012-06-29  Yaron Friedman  <yfriedman@chromium.org>
1544
1545         Fix layout test runner for Android after https://bugs.webkit.org/show_bug.cgi?id=88134
1546         https://bugs.webkit.org/show_bug.cgi?id=90309
1547
1548         Reviewed by Adam Barth.
1549
1550         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
1551         (ChromiumAndroidPort.start_http_server):
1552
1553 2012-06-29  James Weatherall  <wez@chromium.org>
1554
1555         NPObjectWrapper may not address all window script object lifetime issues
1556         https://bugs.webkit.org/show_bug.cgi?id=85679
1557
1558         TestNetscapePlugin now has a leak-window-scriptable-object test which takes a reference to the window script object, and a second reference to it via the "self" property, and does not release those references. This is used to simulate a leaky plugin in layout tests of the NPAPI scripting interface glue code.
1559
1560         Reviewed by Nate Chapin.
1561
1562         * DumpRenderTree/DumpRenderTree.gypi:
1563         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
1564         (PluginTest::NPN_GetProperty):
1565         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
1566         (PluginTest):
1567         * DumpRenderTree/TestNetscapePlugIn/Tests/LeakWindowScriptableObject.cpp: Added.
1568         (LeakWindowScriptableObject):
1569         (LeakWindowScriptableObject::LeakWindowScriptableObject):
1570         (LeakWindowScriptableObject::NPP_New):
1571
1572 2012-06-29  Dirk Pranke  <dpranke@chromium.org>
1573
1574         webkitpy: add comment about how determine_full_port_name() works for apple ports, fix -wk2 bug
1575         https://bugs.webkit.org/show_bug.cgi?id=90314
1576
1577         Reviewed by Ojan Vafai.
1578
1579         Add comments and fix a bug in how we would handle the mac-wk2
1580         and win-wk2 port names after confusion around in bug 90312 :).
1581
1582         * Scripts/webkitpy/layout_tests/port/apple.py:
1583         (ApplePort.determine_full_port_name):
1584
1585 2012-06-29  Dirk Pranke  <dpranke@chromium.org>
1586
1587         webkitpy: remove support for mac leopard from chromium configurations
1588         https://bugs.webkit.org/show_bug.cgi?id=90313
1589
1590         Reviewed by Tony Chang.
1591
1592         Google has shipped the last version of Chrome that will support
1593         Mac OS 10.5 (Leopard), and we no longer have bots that run this
1594         configuration, so we're removing support for it.
1595
1596         A subsequent change will remove the baselines in platform/chromium-mac-leopard.
1597
1598         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
1599         (BaselineOptimizerTest.test_complex_shadowing):
1600         * Scripts/webkitpy/layout_tests/port/builders.py:
1601         * Scripts/webkitpy/layout_tests/port/chromium.py:
1602         (ChromiumPort):
1603         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
1604         (ChromiumMacPort):
1605         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
1606         (ChromiumMacPortTest.test_versions):
1607         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
1608         (FactoryTest.test_mac):
1609         (FactoryTest.test_chromium_mac):
1610         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
1611         (test_rebaseline_and_copy_test_with_lion_result):
1612
1613 2012-06-29  Ojan Vafai  <ojan@chromium.org>
1614
1615         Fix optimize-baselines to not move baselines from win to win-7sp0
1616         https://bugs.webkit.org/show_bug.cgi?id=90312
1617
1618         Reviewed by Dirk Pranke.
1619
1620         It used to consider win-7sp0 as the common directory for all the Apple
1621         windows ports and incorrectly move results out of win.
1622
1623         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
1624         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
1625         (BaselineOptimizerTest.test_win_does_not_drop_to_win_7sp0):
1626         (BaselineOptimizerTest.test_common_directory_includes_root):
1627
1628 2012-06-28  Ojan Vafai  <ojan@chromium.org>
1629
1630         Fix rebaselining for Qt and Apple ports
1631         https://bugs.webkit.org/show_bug.cgi?id=90204
1632
1633         Reviewed by Dirk Pranke.
1634
1635         -Apporpriately put wk2 results in the -wk2 directories.
1636         -Since Qt and Apple-Win don't have bots that correspond to the
1637         platform/qt and platform/win directories, we need to fudge it
1638         and always put the results in those directories for those ports.
1639
1640         * Scripts/webkitpy/layout_tests/port/builders.py:
1641         (rebaseline_override_dir):
1642         * Scripts/webkitpy/layout_tests/port/factory.py:
1643         (_builder_options):
1644         Identify webkit2 builders by the WK2 in the builder name.
1645         * Scripts/webkitpy/tool/commands/rebaseline.py:
1646         (RebaselineTest._baseline_directory):
1647         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
1648         (TestRebaseline.test_baseline_directory):
1649
1650 2012-06-29  Zoltan Horvath  <zoltan@webkit.org>
1651
1652         Add support for --force parameter to run-performance-tests
1653         https://bugs.webkit.org/show_bug.cgi?id=90279
1654
1655         Reviewed by Dirk Pranke.
1656
1657         It's helpful to be able to run tests from the Skipped list of the performance tests.
1658
1659         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1660         (PerfTestsRunner._parse_args):
1661         (PerfTestsRunner._collect_tests):
1662         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py: Add test.
1663         (test_collect_tests_with_skipped_list):
1664
1665 2012-06-29  Tony Chang  <tony@chromium.org>
1666
1667         [GTK] Enable CSS grid layout LayoutTests on GTK+
1668         https://bugs.webkit.org/show_bug.cgi?id=90226
1669
1670         Reviewed by Martin Robinson.
1671
1672         This feature is disabled via Settings by default, but for testing,
1673         we enable it using layoutTestController.overridePreferences. Add the
1674         necessary plumbing for DRT.
1675
1676         WTR already works because support was added for Apple Mac earlier.
1677
1678         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1679         (resetDefaultsToConsistentValues): Feature is off by default.
1680         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
1681         (LayoutTestController::overridePreference): Add handling of WebKitCSSGridLayoutEnabled.
1682
1683 2012-06-29  Ádám Kallai  <kadam@inf.u-szeged.hu>
1684
1685         [Qt] Added Qt port for garden-o-matic.
1686         https://bugs.webkit.org/show_bug.cgi?id=82719
1687
1688         Reviewed by Adam Barth.
1689
1690         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
1691         (.):
1692
1693 2012-06-29  Balazs Kelemen  <kbalazs@webkit.org>
1694
1695         [Qt][WTR] Get rid of using DumpRenderTreeSupportQt
1696         https://bugs.webkit.org/show_bug.cgi?id=90262
1697
1698         Reviewed by Alexey Proskuryakov.
1699
1700         Now that we decided to not support v8 in WebKit2
1701         we can get rid of using DumpRenderTreeSupportQt
1702         in WebKitTestRunner.
1703
1704         * Tools.pro:
1705         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1706         (WTR::InjectedBundlePage::resetAfterTest):
1707         (WTR::InjectedBundlePage::didClearWindowForFrame):
1708         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
1709
1710 2012-06-29  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1711
1712         Revert r121540, it broke most Qt builds
1713
1714         * qmake/mkspecs/features/default_post.prf:
1715         * qmake/mkspecs/features/functions.prf:
1716
1717 2012-06-29  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
1718
1719         [Qt] Make build-webkit reject uknown configurations, eg. --profile
1720
1721         The qmake-based buildsystem doesn't support it.
1722
1723         Reviewed by Tor Arne Vestbø.
1724
1725         * Scripts/webkitdirs.pm:
1726         (buildQMakeProjects):
1727
1728 2012-06-29  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
1729
1730         [Qt] Don't add Qt module dependencies in features.prf
1731
1732         The required dependencies are already added in WebCore.pri.
1733
1734         Reviewed by Tor Arne Vestbø.
1735
1736         * qmake/mkspecs/features/features.prf:
1737
1738 2012-06-27  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
1739
1740         [Qt] Use LIBS_PRIVATE instead of putting dependencies into LIBS
1741
1742         Reviewed by Tor Arne Vestbø..
1743
1744         * qmake/mkspecs/features/default_post.prf:
1745         * qmake/mkspecs/features/functions.prf:
1746
1747 2012-06-28  Ryosuke Niwa  <rniwa@webkit.org>
1748
1749         DOMHTMLCollection::item may return a wrong element after namedItem is called
1750         https://bugs.webkit.org/show_bug.cgi?id=90240
1751
1752         Reviewed by Antti Koivisto.
1753
1754         Add a WebKit API test since namedItem is not used in the JS/V8 binding code.
1755
1756         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1757         * TestWebKitAPI/Tests/mac/HTMLCollectionNamedItem.html: Copied from Tools/TestWebKitAPI/Tests/mac/HTMLFormCollectionNamedItem.html.
1758         * TestWebKitAPI/Tests/mac/HTMLCollectionNamedItem.mm: Copied from Tools/TestWebKitAPI/Tests/mac/HTMLFormCollectionNamedItem.mm.
1759         (TestWebKitAPI::TEST):
1760
1761 2012-06-28  Dirk Pranke  <dpranke@chromium.org>
1762
1763         nrwt: remove the 'google-chrome' port code
1764         https://bugs.webkit.org/show_bug.cgi?id=88824
1765
1766         Reviewed by Ojan Vafai.
1767
1768         NRWT now supports passing additional baseline directories
1769         and expectations files on the command line, so there's no need
1770         to support the concept of a 'google-chrome' port directly.
1771
1772         * Scripts/webkitpy/layout_tests/port/base.py:
1773         (Port.path_to_test_expectations_file):
1774         * Scripts/webkitpy/layout_tests/port/builders.py:
1775         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
1776         (ChromiumMacPort.__init__):
1777         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
1778         (ChromiumWinPort.__init__):
1779         * Scripts/webkitpy/layout_tests/port/factory.py:
1780         (PortFactory):
1781         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
1782         (FactoryTest.test_win):
1783         * Scripts/webkitpy/layout_tests/port/google_chrome.py: Removed.
1784         * Scripts/webkitpy/layout_tests/port/google_chrome_unittest.py: Removed.
1785
1786 2012-06-28  Dirk Pranke  <dpranke@chromium.org>
1787
1788         nrwt: clean up how arguments are passed to workers
1789         https://bugs.webkit.org/show_bug.cgi?id=90126
1790
1791         Reviewed by Ojan Vafai.
1792
1793         The way arguments are passed to workers has been crufty. It
1794         turns out it can be a lot cleaner via two things:
1795         1) using a factory method instead of instantiating objects
1796         directly in manager_worker_broker removes the need for passing
1797         'worker arguments' to the broker.
1798         2) it turns out that since mock hosts and test ports are purely
1799         in-memory constructions, they can be pickled and passed to child
1800         workers, meaning that the worker no longer needs hacky code to
1801         pass the port in a special case or to guess what to do if we
1802         don't have a host - all of the test-specific logic can move to
1803         the test file, where we can stub out the mock host's
1804         port_factory to return the same already-created port when it
1805         needs to be shared.
1806
1807         This change also moves WorkerException to manager_worker_broker.py
1808         where it belongs, and removes several useless tests that were
1809         just a maintenance burden (and would've needed rewriting when we
1810         change the rest of the broker implementation).
1811
1812         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1813         (Manager._run_tests.worker_factory):
1814         (Manager._run_tests):
1815         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
1816         (get):
1817         (WorkerException):
1818         (AbstractWorker.__init__):
1819         (_ManagerConnection.__init__):
1820         (_ManagerConnection.start_worker):
1821         (_InlineManager.__init__):
1822         (_InlineManager.start_worker):
1823         (_MultiProcessManager._can_pickle_host):
1824         (_MultiProcessManager):
1825         (_MultiProcessManager.start_worker):
1826         (_WorkerConnection.__init__):
1827         (_InlineWorkerConnection.__init__):
1828         (_InlineWorkerConnection.join):
1829         (_InlineWorkerConnection.run):
1830         (_Process.run):
1831         (_MultiProcessWorkerConnection.__init__):
1832         (_MultiProcessWorkerConnection.start):
1833         (_MultiProcessWorkerConnection):
1834         (_MultiProcessWorkerConnection.run):
1835         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py:
1836         (_TestWorker.__init__):
1837         (_TestWorker.run):
1838         (_TestsMixin.test_name):
1839         (_TestsMixin.test_cancel):
1840         (_TestsMixin.test_done):
1841         (_TestsMixin.test_unknown_message):
1842         (InlineBrokerTests.setUp):
1843         (MultiProcessBrokerTests.setUp):
1844         * Scripts/webkitpy/layout_tests/controllers/worker.py:
1845         (Worker.__init__):
1846         (Worker.run):
1847         * Scripts/webkitpy/layout_tests/controllers/worker_unittest.py: Removed.
1848         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1849         (passing_run):
1850         (logging_run):
1851         (run_and_capture):
1852         (MainTest.test_child_processes_2):
1853         (MainTest.test_child_processes_min):
1854         (MainTest.test_exception_raised):
1855         (MainTest.test_keyboard_interrupt):
1856         (MainTest.test_retrying_and_flaky_tests):
1857         (MainTest.test_run_order__inline):
1858
1859 2012-06-28  Dirk Pranke  <dpranke@chromium.org>
1860
1861         nrwt: don't try to catch worker exceptions in run_webkit_tests.__main__
1862         https://bugs.webkit.org/show_bug.cgi?id=90125
1863
1864         Reviewed by Ojan Vafai.
1865
1866         It turns out run_webkit_tests.py wasn't really using
1867         WorkerException and the catch we had for it was pointless. I've
1868         removed the symbol import and moved it to the integration tests
1869         where it is needed. Eventually the definition of the exception
1870         moves to the broker module, and so minimizing the number of
1871         users of it is a good thing.
1872
1873         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1874         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1875         (MainTest.test_exception_raised):
1876
1877 2012-06-28  Dirk Pranke  <dpranke@chromium.org>
1878
1879         nrwt: clean up passing of log messages between processes
1880         https://bugs.webkit.org/show_bug.cgi?id=90123
1881
1882         Reviewed by Ojan Vafai.
1883
1884         It turns out log messages are perfectly picklable by themselves
1885         and contain the process id of the process that generated the
1886         message, so if we just pass the record from the worker to the
1887         manager and call logger.handle() in the manager, everything just
1888         works :).
1889
1890         The change is covered by existing tests.
1891
1892         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1893         (Manager._log_messages):
1894         * Scripts/webkitpy/layout_tests/controllers/worker.py:
1895         (_WorkerLogHandler.emit):
1896         * Scripts/webkitpy/layout_tests/views/metered_stream.py:
1897         (MeteredStream.__init__):
1898         (_LogHandler.emit):
1899
1900 2012-06-28  Dirk Pranke  <dpranke@chromium.org>
1901
1902         add a pylint wrapper for linting python code
1903         https://bugs.webkit.org/show_bug.cgi?id=90232
1904
1905         Reviewed by Adam Barth.
1906
1907         Currently we use 'pep8' to check python code in
1908         check-webkit-style. pep8 is fast but simple; pylint is slower
1909         but has much more robust linting capabilities and will catch
1910         variable typos and all sorts of other things. Eventually we
1911         should switch check-webkit-style to use this, but our code is
1912         far from linting now so it needs to be cleaned up first.
1913
1914         This change adds the infrastructure and a wrapper so we can
1915         start doing that.
1916
1917         * Scripts/lint-webkitpy: Added.
1918         * Scripts/webkitpy/pylintrc: Added.
1919         * Scripts/webkitpy/thirdparty/__init__.py:
1920         (AutoinstallImportHook.find_module):
1921         (AutoinstallImportHook._install_pylint):
1922
1923 2012-06-28  Tony Chang  <tony@chromium.org>
1924
1925         [GTK] Use WEBKITOUTPUTDIR to find fonts in DumpRenderTree
1926         https://bugs.webkit.org/show_bug.cgi?id=90215
1927
1928         Reviewed by Martin Robinson.
1929
1930         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1931         (initializeFonts): Check for WEBKITOUTPUTDIR first.
1932         * Scripts/webkitpy/layout_tests/port/gtk.py:
1933         (GtkPort.setup_environ_for_server): Copy the environment variable to the child process.
1934         * WebKitTestRunner/InjectedBundle/gtk/ActivateFontsGtk.cpp:
1935         (WTR::inititializeFontConfigSetting): Check for WEBKITOUTPUTDIR first.
1936
1937 2012-06-28  Dirk Pranke  <dpranke@chromium.org>
1938
1939         derive ChromiumPort from WebKitPort in NRWT in order to support skipping tests if symbols are missing
1940         https://bugs.webkit.org/show_bug.cgi?id=89706
1941
1942         Reviewed by Ojan Vafai.
1943
1944         Try again to land the change first landed in r121363. This patch
1945         adds a bunch more tests and reworks the handling of
1946         port-specific default values for --pixel-tests and --time-out-ms
1947         to be more consistent (adding a default_pixel_tests() method,
1948         pushing the webkit default_timeout_ms() value up into base.py,
1949         and overriding it properly in the chromium and apple mac ports.
1950
1951         Also change the logic in
1952         run_webkit_tests._setup_derived_options() to not second-guess
1953         what the port wants the timeout value to be for debug builds;
1954         computing this in two different places led to several bugs.
1955
1956         This change also changes the Chromium unittest ports to derive
1957         from ChromiumPortTestCase instead of PortTestCase, so that we
1958         ensure that we're running the same tests on all port variants
1959         more easily. There's more cleanup that can be done here, but
1960         this is good enough for now
1961
1962         * Scripts/webkitpy/layout_tests/port/base.py:
1963         (Port.default_pixel_tests):
1964         (Port):
1965         (Port.default_timeout_ms):
1966         * Scripts/webkitpy/layout_tests/port/chromium.py:
1967         (ChromiumPort):
1968         (ChromiumPort.__init__):
1969         (ChromiumPort.default_pixel_tests):
1970         (ChromiumPort.default_timeout_ms):
1971         (ChromiumPort.driver_name):
1972         (ChromiumPort._driver_class):
1973         (ChromiumPort._missing_symbol_to_skipped_tests):
1974         (ChromiumPort.skipped_layout_tests):
1975         (ChromiumPort.setup_test_run):
1976         (ChromiumPort._path_to_image_diff):
1977         (ChromiumPort._convert_path):
1978         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
1979         (ChromiumAndroidPortTest):
1980         (ChromiumAndroidPortTest.test_expectations_files):
1981         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
1982         (ChromiumLinuxPort._modules_to_search_for_symbols):
1983         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
1984         (ChromiumLinuxPortTest):
1985         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
1986         (ChromiumMacPort._modules_to_search_for_symbols):
1987         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
1988         (ChromiumMacPortTest):
1989         * Scripts/webkitpy/layout_tests/port/chromium_port_testcase.py:
1990         split off from chromium_unittest.
1991         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
1992         (ChromiumWinPort._modules_to_search_for_symbols):
1993         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
1994         (ChromiumWinTest):
1995         * Scripts/webkitpy/layout_tests/port/mac.py:
1996         (MacPort.default_timeout_ms):
1997         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
1998         (MacTest.test_default_timeout_ms):
1999         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
2000         (MockDRTPort.start_http_server):
2001         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
2002         (PortTestCase.test_default_timeout_ms):
2003         (PortTestCase):
2004         (PortTestCase.test_default_pixel_tests):
2005         * Scripts/webkitpy/layout_tests/port/test.py:
2006         (TestPort.default_pixel_tests):
2007         * Scripts/webkitpy/layout_tests/port/webkit.py:
2008         (WebKitPort._modules_to_search_for_symbols):
2009         (WebKitPort):
2010         (WebKitPort._symbols_string):
2011         (WebKitPort._skipped_tests_for_unsupported_features):
2012         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
2013         (TestWebKitPort._symbols_string):
2014         (TestWebKitPort._tests_for_disabled_features):
2015         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2016         (_set_up_derived_options):
2017
2018 2012-06-28  Joshua Bell  <jsbell@chromium.org>
2019
2020         run-bindings-tests should return non-zero exit code on test failure
2021         https://bugs.webkit.org/show_bug.cgi?id=90205
2022
2023         Reviewed by Adam Barth.
2024
2025         * Scripts/run-bindings-tests:
2026         (main):
2027
2028 2012-06-28  Terry Anderson  <tdanderson@chromium.org>
2029
2030         [chromium] Reset the device scale factor to 1 before each test is run
2031         https://bugs.webkit.org/show_bug.cgi?id=90212
2032
2033         Reviewed by Adam Barth.
2034
2035         Some tests change the device scale factor, so this needs to be reset to
2036         1.0 at the start of each test to avoid test flakiness.
2037
2038         * DumpRenderTree/chromium/LayoutTestController.cpp:
2039         (LayoutTestController::reset):
2040
2041 2012-06-28  Ryosuke Niwa  <rniwa@webkit.org>
2042
2043         Cleanup HTMLFormCollection
2044         https://bugs.webkit.org/show_bug.cgi?id=90111
2045
2046         Reviewed by Andreas Kling.
2047
2048         Add a WebKit API test using copy-paste design pattern per kling's request.
2049
2050         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
2051         * TestWebKitAPI/Tests/mac/HTMLFormCollectionNamedItem.html: Added.
2052         * TestWebKitAPI/Tests/mac/HTMLFormCollectionNamedItem.mm: Added.
2053         (-[HTMLFormCollectionNamedItemTest webView:didFinishLoadForFrame:]):
2054         (TestWebKitAPI):
2055         (TestWebKitAPI::TEST):
2056
2057 2012-06-28  Tim Horton  <timothy_horton@apple.com>
2058
2059         [mac] WKTR windows still don't stay off screen sometimes
2060         https://bugs.webkit.org/show_bug.cgi?id=90214
2061         <rdar://problem/11760263>
2062
2063         Reviewed by Simon Fraser.
2064
2065         In some cases, the system itself will consult [WebKitTestRunnerWindow frame], so we should refrain from
2066         overriding it and instead provide a different method to use when we want the web-facing "fake" window origin
2067         (for PlatformWebView::windowFrame()).
2068
2069         * WebKitTestRunner/mac/PlatformWebViewMac.mm:
2070         (-[WebKitTestRunnerWindow frameRespectingFakeOrigin]):
2071         (WTR::PlatformWebView::windowFrame):
2072
2073 2012-06-28  Martin Robinson  <mrobinson@igalia.com>
2074
2075         [GTK] Add unit tests for GtkInputMethodFilter
2076         https://bugs.webkit.org/show_bug.cgi?id=88698
2077
2078         Reviewed by Carlos Garcia Campos.
2079
2080         Add unit tests for GtkInputMethodFilter in the WebCore platform layer.
2081         This change adds the TestGtk test suite which will be used for all non-API
2082         layer GTK unit tests.
2083
2084         * TestWebKitAPI/GNUmakefile.am: Update the build to include the new tests.
2085         * TestWebKitAPI/Tests/gtk/InputMethodFilter.cpp: Added.
2086         (TestWebKitAPI::PlatformWebView::PlatformWebView): Remove the call to gtk_init here
2087         as it's now in main.cpp.
2088         * TestWebKitAPI/gtk/main.cpp: Change the g_type_init call to gtk_init, because now
2089         a majority of all unit tests depend on GTK+ being initialized.
2090
2091 2012-06-28  Christophe Dumez  <christophe.dumez@intel.com>
2092
2093         [EFL] Enable support for HTML5 datalist
2094         https://bugs.webkit.org/show_bug.cgi?id=90157
2095
2096         Reviewed by Martin Robinson.
2097
2098         Turn on DATALIST flag by default on EFL port to
2099         support HTML5 datalist tag.
2100
2101         * Scripts/webkitperl/FeatureList.pm:
2102
2103 2012-06-28  Tony Chang  <tony@chromium.org>
2104
2105         Enable CSS grid layout LayoutTests on platform Mac
2106         https://bugs.webkit.org/show_bug.cgi?id=90113
2107
2108         Reviewed by Ojan Vafai.
2109
2110         * DumpRenderTree/mac/DumpRenderTree.mm:
2111         (resetDefaultsToConsistentValues): Reset the value to NO between tests.
2112
2113 2012-06-28  Mike Fenton  <mifenton@rim.com>
2114
2115         [BlackBerry] Add watchlist options for Blackberry and editing.
2116         https://bugs.webkit.org/show_bug.cgi?id=90193
2117
2118         Unreviewed.
2119
2120         Add BlackBerry and Editing watchlist and monitor them.
2121
2122         * Scripts/webkitpy/common/config/watchlist:
2123
2124 2012-06-28  Tim Horton  <timothy_horton@apple.com>
2125
2126         Unreviewed, upgrade myself to a reviewer!
2127
2128         http://www.webkit.org/blog/2082/tim-horton-is-now-a-webkit-reviewer/
2129
2130         * Scripts/webkitpy/common/config/committers.py:
2131
2132 2012-06-28  Terry Anderson  <tdanderson@chromium.org>
2133
2134         [Chromium] Chromium's LayoutTestController is missing setBackingScaleFactor
2135         https://bugs.webkit.org/show_bug.cgi?id=83635
2136
2137         Reviewed by Adam Barth.
2138
2139         Added InvokeCallbackTask, a new derived class of MethodTask. When
2140         setBackingScaleFactor is called, a call to setDeviceScaleFactor in
2141         WebView is made and then postTask is used to invoke the callback
2142         function specified in testRunner.setBackingScaleFactor. 
2143
2144         * DumpRenderTree/chromium/LayoutTestController.cpp:
2145         (LayoutTestController::LayoutTestController):
2146         (InvokeCallbackTask):
2147         (InvokeCallbackTask::InvokeCallbackTask):
2148         (InvokeCallbackTask::runIfValid):
2149         (LayoutTestController::setBackingScaleFactor):
2150         * DumpRenderTree/chromium/LayoutTestController.h:
2151         (LayoutTestController):
2152
2153 2012-06-28  Ojan Vafai  <ojan@chromium.org>
2154
2155         Make rebaseline-test and rebaseline-expectations work for non-Chromium ports
2156         https://bugs.webkit.org/show_bug.cgi?id=90186
2157
2158         Reviewed by Adam Barth.
2159
2160         This makes rebaselining work for all ports that have a TestExpectations file
2161         in the tree. I didn't test other ports.
2162
2163         This doesn't address 100% of the problem. The rebaseline code puts the expectations
2164         in the most specific directory and relies on optimize-baselines to merge baselines
2165         appropriately. This only works if every platform directory has an equivalent bot
2166         that runs the tests, which is not true for most ports.
2167
2168         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
2169         (Builder._revision_and_build_for_filename):
2170         Some bots have filenames that aren't revision/build number pairs
2171         e.g. they are random junk like aQhxvx. Handle this gracefully.
2172         (Builder._fetch_revision_to_build_map):
2173         (Builder._file_info_list_to_revision_to_build_list):
2174         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2175         (BuilderTest.test_build_and_revision_for_filename):
2176         (BuilderTest.test_file_info_list_to_revision_to_build_list):
2177         * Scripts/webkitpy/layout_tests/port/builders.py:
2178         Update the list of builders. This list needs to be kept up
2179         to date for the rebaseline tool to work.
2180         * Scripts/webkitpy/tool/commands/rebaseline.py:
2181         (RebaselineTest._results_url):
2182         (RebaselineExpectations._run_webkit_patch):
2183         (RebaselineExpectations._rebaseline_port):
2184         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
2185         Qt port uses qmake to determine the right version. Systems without qmake correctly fallback
2186         to a specific version.
2187
2188 2012-06-28  Csaba Osztrogonác  <ossy@webkit.org>
2189
2190         [Qt][NRWT] Fix baseline and skipped file search path.
2191         https://bugs.webkit.org/show_bug.cgi?id=89882
2192
2193         Unreviewed trivial typo fix after r121430.
2194
2195         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
2196         (QtPortTest._assert_search_path):
2197         (QtPortTest._assert_skipped_path):
2198
2199 2012-06-28  János Badics  <jbadics@inf.u-szeged.hu>
2200
2201         [Qt][NRWT] Fix baseline and skipped file search path.
2202         https://bugs.webkit.org/show_bug.cgi?id=89882
2203
2204         Reviewed by Csaba Osztrogonác.
2205
2206         * Scripts/webkitpy/layout_tests/port/qt.py:
2207         (QtPort._search_paths):
2208         (QtPort):
2209         (QtPort.baseline_search_path):
2210         (QtPort._skipped_file_search_paths):
2211         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
2212         (QtPortTest):
2213         (QtPortTest._assert_skipped_path):
2214         (QtPortTest.test_baseline_search_path):
2215         (QtPortTest.test_skipped_file_search_path):
2216
2217 2012-06-28  Alexander Pavlov  <apavlov@chromium.org>
2218
2219         [Qt] inspector/styles/inject-stylesheet.html makes 4 tests flakey (TEXT PASS)
2220         https://bugs.webkit.org/show_bug.cgi?id=90167
2221
2222         Reviewed by Csaba Osztrogonác.
2223
2224         DRT should remove user stylesheets from the page group when resetting before running another test.
2225
2226         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
2227         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
2228
2229 2012-06-28  Csaba Osztrogonác  <ossy@webkit.org>
2230
2231         [Qt][DRT] Reset AcceleratedCompositingEnabled between tests
2232         https://bugs.webkit.org/show_bug.cgi?id=90164
2233
2234         Reviewed by Simon Hausmann.
2235
2236         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
2237         (WebCore::WebPage::resetSettings):
2238
2239 2012-06-28  Csaba Osztrogonác  <ossy@webkit.org>
2240
2241         [Qt] Restore original value of mock scrollbars between tests
2242         https://bugs.webkit.org/show_bug.cgi?id=90155
2243
2244         Reviewed by Simon Hausmann.
2245
2246         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
2247         (WebCore::WebPage::resetSettings):
2248         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
2249         (LayoutTestController::setMockScrollbarsEnabled):
2250         * DumpRenderTree/qt/LayoutTestControllerQt.h:
2251         (LayoutTestController):
2252
2253 2012-06-27  Mark Rowe  <mrowe@apple.com>
2254
2255         <http://webkit.org/b/90096> Building within Xcode sometimes does a full rebuild after building via build-webkit
2256
2257         We need to ensure that build-webkit uses the same setting for SHARED_PRECOMPS_DIR
2258         as what Xcode itself will use when building, otherwise switching between the two
2259         can result in the precompiled headers being rebuilt and thus the entire world
2260         rebuilding.
2261
2262         Reviewed by Dan Bernstein.
2263
2264         * Scripts/webkitdirs.pm:
2265         (determineBaseProductDir):
2266
2267 2012-06-27  Sheriff Bot  <webkit.review.bot@gmail.com>
2268
2269         Unreviewed, rolling out r121363, r121367, r121384, and
2270         r121390.
2271         http://trac.webkit.org/changeset/121363
2272         http://trac.webkit.org/changeset/121367
2273         http://trac.webkit.org/changeset/121384
2274         http://trac.webkit.org/changeset/121390
2275         https://bugs.webkit.org/show_bug.cgi?id=90134
2276
2277         It broke debug NRWT on GTK and on Qt (Requested by Ossy_NIGHT
2278         on #webkit).
2279
2280         * Scripts/webkitpy/layout_tests/port/chromium.py:
2281         (ChromiumPort):
2282         (ChromiumPort.__init__):
2283         (ChromiumPort._check_file_exists):
2284         (ChromiumPort.default_results_directory):
2285         (ChromiumPort._driver_class):
2286         (ChromiumPort._build_path):
2287         (ChromiumPort._path_to_image_diff):
2288         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2289         (ChromiumLinuxPort.baseline_search_path):
2290         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
2291         (ChromiumMacPort.baseline_search_path):
2292         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
2293         (ChromiumPortTest):
2294         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2295         (ChromiumWinPort.baseline_search_path):
2296         * Scripts/webkitpy/layout_tests/port/webkit.py:
2297         (WebKitPort.__init__):
2298         (WebKitPort._webcore_symbols_string):
2299         (WebKitPort._skipped_tests_for_unsupported_features):
2300         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
2301         (TestWebKitPort._webcore_symbols_string):
2302         (WebKitPortUnitTests.test_default_options):
2303
2304 2012-06-27  Dirk Pranke  <dpranke@chromium.org>
2305
2306         Fix typo in r121384 :(
2307
2308         Unreviewed, build fix.
2309
2310         * Scripts/webkitpy/layout_tests/port/webkit.py:
2311         (WebKitPort.default_test_timeout_ms):
2312
2313 2012-06-27  Dirk Pranke  <dpranke@chromium.org>
2314
2315         nrwt: default timeout for debug bots broke in r121363
2316         https://bugs.webkit.org/show_bug.cgi?id=90112
2317
2318         Unreviewed, build fix.
2319
2320         I forgot to account for release and debug having different
2321         default values :(.
2322
2323         * Scripts/webkitpy/layout_tests/port/chromium.py:
2324         (ChromiumPort.default_test_timeout_ms):
2325         * Scripts/webkitpy/layout_tests/port/webkit.py:
2326         (WebKitPort):
2327         (WebKitPort.default_test_timeout_ms):
2328         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
2329         (WebKitPortUnitTests.test_default_options):
2330
2331 2012-06-27  Dirk Pranke  <dpranke@chromium.org>
2332
2333         webkitpy: fix a couple of issues running under cygwin
2334         https://bugs.webkit.org/show_bug.cgi?id=90035
2335
2336         Reviewed by Eric Seidel.
2337
2338         These were causing unit tests to fail on cygwin (apple win bot).
2339
2340         * Scripts/webkitpy/layout_tests/port/chromium.py:
2341         * Scripts/webkitpy/performance_tests/perftest.py:
2342
2343 2012-06-27  Adam Barth  <abarth@webkit.org>
2344
2345         [Chromium] DumpRenderTree on Android should call SkUseTestFontConfigFile once available
2346         https://bugs.webkit.org/show_bug.cgi?id=89801
2347
2348         Reviewed by Nate Chapin.
2349
2350         Let's call SkUseTestFontConfigFile now that it exists.
2351
2352         * DumpRenderTree/chromium/TestShellAndroid.cpp:
2353         (platformInit):
2354
2355 2012-06-27  Dirk Pranke  <dpranke@chromium.org>
2356
2357         Fix typo introduced in r121363.
2358
2359         Unreviewed, build fix.
2360
2361         * Scripts/webkitpy/layout_tests/port/webkit.py:
2362         (WebKitPort._symbols_string):
2363
2364 2012-06-27  Mark Hahnenberg  <mhahnenberg@apple.com>
2365
2366         Moving from committers to reviewers :-)
2367
2368         Rubber stamped by Filip Pizlo.
2369
2370         * Scripts/webkitpy/common/config/committers.py:
2371
2372 2012-06-27  Dirk Pranke  <dpranke@chromium.org>
2373
2374         Derive ChromiumPort from WebKitPort to add support for missing symbols to skip tests
2375         https://bugs.webkit.org/show_bug.cgi?id=89706
2376
2377         Reviewed by Adam Barth.
2378
2379         Based on the original patch by Raymond Toy.
2380
2381         This patch changes ChromiumPort to derive from webkit.WebKitPort
2382         instead of base.Port. This is a long-awaited change and a
2383         precursor to merging base.Port and webkit.WebKitPort, but is
2384         driven by the desire to dynamically detect whether the MP3 and
2385         AAC codecs are compiled into DRT on Chromium, for which we
2386         wanted to re-use the existing logic in WebKit port for determine
2387         what to skip at compile time.
2388
2389         Most of the changes are shuffling things around so that we don't
2390         change any other logic and so we override the necessary methods
2391         in WebKitPort (and try to follow the same method definition
2392         order where possible).
2393
2394         Also, on the Chromium port the mp3 and aac codecs are actually
2395         defined in a separate library, so scanning webcore isn't
2396         sufficient. This patch generalizes the symbol lookup to handle
2397         multiple libraries, and uses different libraries as appropriate
2398         for Chromium.
2399
2400         The only functional/visible changes should be in the value
2401         returned for skipped_layout_tests().
2402
2403         * Scripts/webkitpy/layout_tests/port/chromium.py:
2404         (ChromiumPort):
2405         (ChromiumPort.__init__):
2406         (ChromiumPort.driver_name):
2407         (ChromiumPort._driver_class):
2408         (ChromiumPort._missing_symbol_to_skipped_tests):
2409         (ChromiumPort.skipped_layout_tests):
2410         (ChromiumPort.setup_test_run):
2411         (ChromiumPort._path_to_image_diff):
2412         (ChromiumPort._convert_path):
2413         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
2414         (ChromiumPortTest.test_missing_symbol_to_skipped_tests):
2415         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2416         (ChromiumLinuxPort._modules_to_search_for_symbols):
2417         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
2418         (ChromiumLinuxPort._modules_to_search_for_symbols):
2419         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2420         (ChromiumLinuxPort._modules_to_search_for_symbols):
2421         * Scripts/webkitpy/layout_tests/port/webkit.py:
2422         (WebKitPort):
2423         (WebKitPort.__init__):
2424         (WebKitPort._symbols_string):
2425         (WebKitPort._modules_to_search_for_symbols):
2426
2427 2012-06-27  Csaba Osztrogonác  <ossy@webkit.org>
2428
2429         [Qt] REGRESSION(r121339): It broke the build on the Qt Windows bots
2430         https://bugs.webkit.org/show_bug.cgi?id=90081
2431
2432         Buildfix for Qt 4.8 Windows. Use the former path for Qt 4.8, and the newer one for Qt 5.
2433
2434         Reviewed by Noam Rosenthal.
2435
2436         * qmake/mkspecs/features/features.prf:
2437
2438 2012-06-27  Sergio Villar Senin  <svillar@igalia.com>
2439
2440         [WK2] [GTK] WebKit2 testing bot fails to run tests due to missing files
2441         https://bugs.webkit.org/show_bug.cgi?id=90061
2442
2443         Reviewed by Gustavo Noronha Silva.
2444
2445         Add "-no-install -no-fast-install" to the LDFLAGS in bots. With
2446         this flag libtool tells the linker to set the rpath for the output
2447         file to the full path of the .libs directory, instead of using a
2448         wrapper script to set up the LD_LIBRARY_PATH. This will allow us
2449         to directly reuse builds in the pure testing bots.
2450
2451         * BuildSlaveSupport/gtk/daemontools-buildbot.conf:
2452
2453 2012-06-27  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2454
2455         [Qt] No need to save and restore TEMPLATE in a function
2456
2457         This was a leftover from when the logic was not in its own function scope.
2458
2459         QMAKE_FRAMEWORK_BUNDLE_NAME on the other hand is exported in qtLibraryTarget, which
2460         will surprisingly affect the global scope as well, so we have to save and restore it.
2461
2462         Original patch by Oswald Buddenhagen <oswald.buddenhagen@nokia.com> on 2012-06-27
2463
2464         Reviewed by Tor Arne Vestbø.
2465
2466         * qmake/mkspecs/features/functions.prf:
2467
2468 2012-06-27  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
2469
2470         [Qt] Fix lookup location for sqlite sources
2471
2472         Don't look in the install dir - we are unlikely to find anything there
2473         unless we are doing a developer build.
2474
2475         Reviewed by Tor Arne Vestbø.
2476
2477         * qmake/mkspecs/features/features.prf:
2478
2479 2012-06-27  Zan Dobersek  <zandobersek@gmail.com>
2480
2481         [Gtk] Add support for the Gamepad API
2482         https://bugs.webkit.org/show_bug.cgi?id=87503
2483
2484         Reviewed by Carlos Garcia Campos.
2485
2486         Enable the gamepad support for the GTK port.
2487
2488         * Scripts/webkitperl/FeatureList.pm:
2489
2490 2012-06-27  Ryosuke Niwa  <rniwa@webkit.org>
2491
2492         Fix gcc build after r121302
2493         https://bugs.webkit.org/show_bug.cgi?id=90055
2494
2495         Reviewed by Mark Rowe.
2496
2497         Assume RTTI is always disabled so that gtest builds on XCode 3.2.6.
2498
2499         It appears that gcc doesn't like window.get().* inside a template so replace that by [window.get() *] instead.
2500
2501         * TestWebKitAPI/Configurations/Base.xcconfig:
2502         * TestWebKitAPI/Tests/mac/AcceptsFirstMouse.mm:
2503         (TestWebKitAPI::AcceptsFirstMouse::runTest):
2504
2505 2012-06-27  Csaba Osztrogonác  <ossy@webkit.org>
2506
2507         Add master.cfg unittest to help migration - pass BuildStep instances instead of BuildStep subclasses
2508         https://bugs.webkit.org/show_bug.cgi?id=89564
2509
2510         Reviewed by Tony Chang.
2511
2512         * BuildSlaveSupport/build.webkit.org-config/mastercfg_unittest.py:
2513         (BuildStepsConstructorTest):
2514         (BuildStepsConstructorTest.generateTests):
2515         (BuildStepsConstructorTest.createTest):
2516         (BuildStepsConstructorTest.createTest.doTest):
2517
2518 2012-06-26  Mark Hahnenberg  <mhahnenberg@apple.com>
2519
2520         Add support for preciseTime() to WebKitTestRunner
2521         https://bugs.webkit.org/show_bug.cgi?id=90027
2522
2523         Reviewed by Darin Adler.
2524
2525         It would be nice to be able to use preciseTime() in WebKitTestRunner like we can in DumpRenderTree.
2526
2527         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
2528         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp: 
2529         (WTR::LayoutTestController::preciseTime): 
2530         (WTR):
2531         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
2532         (LayoutTestController):
2533
2534 2012-06-26  Lucas Forschler  <lforschler@apple.com>
2535
2536       Teach the Apple port how to build the test tools in build-webkit
2537       https://bugs.webkit.org/show_bug.cgi?id=89540
2538
2539       Reviewed by Jon Lee & Simon Fraser.
2540
2541       * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2542       (RunWebKitTests.start): pass --no-build since tools should now be in the downloaded archive
2543       (RunUnitTests.start): ditto
2544       * Scripts/build-webkit: add tools to the projects build list
2545
2546 2012-06-26  Ojan Vafai  <ojan@chromium.org>
2547
2548         Only show the platform-appropriate builders for non-layout test failures in garden-o-matic
2549         https://bugs.webkit.org/show_bug.cgi?id=90025
2550
2551         Reviewed by Simon Fraser.
2552
2553         Move the chromium-specific filtering code into config.js and replace it with a method on each platform
2554         config. Also, let the webkit test step name be webkit_tests (build.chromium.org) or layout-test (build.webkit.org).
2555
2556         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders.js:
2557         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders_unittests.js:
2558         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
2559
2560 2012-06-26  Ojan Vafai  <ojan@chromium.org>
2561
2562         Fix failing garden-o-matic unittests
2563         https://bugs.webkit.org/show_bug.cgi?id=90021
2564
2565         Reviewed by Adam Barth.
2566
2567         These had all just fallen out of date.
2568
2569         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout_unittests.js:
2570         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
2571         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:
2572
2573 2012-06-26  Ojan Vafai  <ojan@chromium.org>
2574
2575         Fix platform picker change handler in garden-o-matic
2576         https://bugs.webkit.org/show_bug.cgi?id=90010
2577
2578         Reviewed by Simon Fraser.
2579
2580         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
2581         The old code never worked. This is hard to test because change handlers require
2582         a user-initiated action and the code is changing the window's location, which would
2583         navigate the test page.
2584
2585         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/onebar.css:
2586         Fix the CSS so it doesn't cause the tabstrip to be disconnected from the tabs on Linux
2587         due to the large margin-bottom.
2588
2589 2012-06-26  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2590
2591         webkitpy: Make webkit-patch patches-to-review useful
2592         https://bugs.webkit.org/show_bug.cgi?id=89470
2593
2594         Reviewed by Eric Seidel.
2595
2596         webkit-patch patches-to-review will now output the list of
2597         bugs with patches pending for review that has the user on CC,
2598         excluding patches with cq-, sorted by the age of the patch.
2599
2600         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
2601         (BugzillaQueries.fetch_bugs_from_review_queue):
2602         * Scripts/webkitpy/common/net/bugzilla/bugzilla_mock.py:
2603         (MockBugzillaQueries.fetch_bugs_from_review_queue):
2604         (MockBugzilla.__init__):
2605         (MockBugzilla.authenticate):
2606         * Scripts/webkitpy/tool/commands/queries.py:
2607         (PatchesToReview):
2608         (PatchesToReview.__init__):
2609         (PatchesToReview._print_report):
2610         (PatchesToReview._generate_report):
2611         (PatchesToReview.execute):
2612         * Scripts/webkitpy/tool/commands/queries_unittest.py:
2613         (QueryCommandsTest.test_patches_to_review):
2614
2615 2012-06-26  Dirk Pranke  <dpranke@chromium.org>
2616
2617         nrwt: broken for chromium on vista
2618         https://bugs.webkit.org/show_bug.cgi?id=89988
2619
2620         Reviewed by Tony Chang.
2621
2622         r121194 removed support for 'chromium-win-vista' as a separate
2623         port, but this actually prevented the code from running on vista
2624         at all, which is unduly harsh and broke the websocket tests on
2625         the (non-webkit) chromium bots that are still running on vista.
2626
2627         It's probably good enough to pretend that vista is win7 instead;
2628         some layout tests will still fail but at least things'll run.
2629
2630         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2631         (ChromiumWinPort.determine_full_port_name):
2632         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
2633         (ChromiumWinTest.test_versions):
2634
2635 2012-06-26  Csaba Osztrogonác  <ossy@webkit.org>
2636
2637         master.cfg cleanup: Pass CheckOutSource instance instead of class to BuildStep.addStep
2638         https://bugs.webkit.org/show_bug.cgi?id=89215
2639
2640         We need it because it is deprecated and will be dropped in buildbot 0.8.7
2641
2642         Reviewed by Tony Chang.
2643
2644         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2645         (CheckOutSource.__init__):
2646         (Factory.__init__):
2647
2648 2012-06-26  Simon Hausmann  <simon.hausmann@nokia.com>
2649
2650         [Qt][Win] Symbols are not exported in QtWebKit5.dll
2651         https://bugs.webkit.org/show_bug.cgi?id=88873
2652
2653         Reviewed by Tor Arne Vestbø.
2654
2655         When linking the target dll make sure to re-export the symbols from
2656         the static libraries marked as export, with the help of a little python
2657         script and a qmake extra compiler.
2658
2659         * Scripts/generate-win32-export-forwards: Added.
2660         * qmake/mkspecs/features/win32/default_post.prf:
2661
2662 2012-06-25  Jocelyn Turcotte  <turcotte.j@gmail.com>
2663
2664         Add a note about hostname completion not working well with --cc completion
2665
2666         Reviewed by Simon Hausmann.
2667
2668         Hostname completion tries to resolve anything after an @ sign which is present
2669         in the completed list of contributor emails to CC.
2670
2671         * Scripts/webkit-tools-completion.sh:
2672
2673 2012-06-26  Sheriff Bot  <webkit.review.bot@gmail.com>
2674
2675         Unreviewed, rolling out r121244.
2676         http://trac.webkit.org/changeset/121244
2677         https://bugs.webkit.org/show_bug.cgi?id=89966
2678
2679         skip list path is incorrect now. (Requested by Ossy on
2680         #webkit).
2681
2682         * Scripts/webkitpy/layout_tests/port/qt.py:
2683         (QtPort.baseline_search_path):
2684         (QtPort._skipped_file_search_paths):
2685         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
2686         (QtPortTest.test_baseline_search_path):
2687
2688 2012-06-26  János Badics  <jbadics@inf.u-szeged.hu>
2689
2690         [Qt][NRWT] Fix baseline and skipped file search path.
2691         https://bugs.webkit.org/show_bug.cgi?id=89882
2692
2693         Reviewed by Csaba Osztrogonác.
2694
2695         * Scripts/webkitpy/layout_tests/port/qt.py:
2696         (QtPort.baseline_search_path):
2697         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
2698         (QtPortTest.test_baseline_search_path):
2699
2700 2012-06-26  Tony Chang  <tony@chromium.org>
2701
2702         [Qt] Enable grid layout LayoutTests
2703         https://bugs.webkit.org/show_bug.cgi?id=89909
2704
2705         Reviewed by Csaba Osztrogonác.
2706
2707         These tests pass, we just need to hook up the overridePreference.
2708
2709         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
2710         (WebCore::WebPage::resetSettings): Reset grid layout and regions between tests.
2711         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
2712         (LayoutTestController::overridePreference): Add WebKitCSSGridLayoutEnabled.
2713
2714 2012-06-26  Sheriff Bot  <webkit.review.bot@gmail.com>
2715
2716         Unreviewed, rolling out r121236.
2717         http://trac.webkit.org/changeset/121236
2718         https://bugs.webkit.org/show_bug.cgi?id=89956
2719
2720         It's broke the nrwt on qt and gtk platform (Requested by
2721         kkristof on #webkit).
2722
2723         * Scripts/webkitpy/layout_tests/port/qt.py:
2724         (QtPort._build_driver):
2725         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
2726         (XvfbDriver._start):
2727         (XvfbDriver._start.x_filter):
2728         (XvfbDriver.stop):
2729
2730 2012-06-26  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
2731
2732         [NRWT] XvfbDriver should choose the next free display
2733         https://bugs.webkit.org/show_bug.cgi?id=88414
2734
2735         Reviewed by Dirk Pranke.
2736
2737         * Scripts/webkitpy/layout_tests/port/qt.py:
2738         (QtPort._driver_class):
2739         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
2740         (XvfbDriver.__init__):
2741         (XvfbDriver._start):
2742         (XvfbDriver._start.next_free_id):
2743         (XvfbDriver.stop):
2744
2745 2012-06-25  Simon Hausmann  <simon.hausmann@nokia.com>
2746
2747         [Qt] Make it possible to build WebKit without QtWidgets
2748         https://bugs.webkit.org/show_bug.cgi?id=78109
2749
2750         Reviewed by Tor Arne Vestbø.
2751
2752         * Tools.pro: Disable various tools that depend on WK1.
2753         * qmake/mkspecs/features/default_post.prf: Add WK1 as supported static lib in WEBKIT
2754         * qmake/mkspecs/features/default_pre.prf: Disable WK1 if QtWidgets is not built.
2755         * qmake/mkspecs/modules/webkit1.prf: Added.
2756
2757 2012-06-25  Dirk Pranke  <dpranke@chromium.org>
2758
2759         put vista back for chromium (non-webkit) bots ...
2760         https://bugs.webkit.org/show_bug.cgi?id=89929
2761
2762         Unreviewed, build fix.
2763
2764         Turns out we've only gotten rid of vista on the webkit tests so
2765         far. Put vista back for the other chromium bots ...
2766
2767         * TestResultServer/static-dashboards/flakiness_dashboard.js:
2768         (chromiumPlatform):
2769         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2770         (test):
2771
2772 2012-06-25  Dirk Pranke  <dpranke@chromium.org>
2773
2774         update flakiness dashboard after removing chromium-vista bots
2775         https://bugs.webkit.org/show_bug.cgi?id=89925
2776
2777         Unreviewed, build fix.
2778
2779         More changes missed in r121194 :(.
2780
2781         * TestResultServer/static-dashboards/flakiness_dashboard.js:
2782         (chromiumPlatform):
2783         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2784         (test):
2785
2786 2012-06-25  Dirk Pranke  <dpranke@chromium.org>
2787
2788         remove support for chromium vista from tools
2789         https://bugs.webkit.org/show_bug.cgi?id=89915
2790
2791         Reviewed by Tony Chang.
2792
2793         This change is all deleting code and updating tests to handle
2794         the configuration being gone. All tests now pass except for a
2795         few unittests for the flakiness dashboard which were failing
2796         before this patch.
2797
2798         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders_unittests.js:
2799         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout_unittests.js:
2800         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
2801         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
2802         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
2803         (BaselineOptimizerTest.test_complex_shadowing):
2804         * Scripts/webkitpy/layout_tests/port/builders.py:
2805         * Scripts/webkitpy/layout_tests/port/chromium.py:
2806         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
2807         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2808         (ChromiumWinPort):
2809         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
2810         (ChromiumWinTest.test_versions):
2811         (ChromiumWinTest.test_baseline_path):
2812         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
2813         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
2814         (BuildCoverageExtrapolatorTest.test_extrapolate):
2815
2816 2012-06-25  Scott Graham  <scottmg@chromium.org>
2817
2818         [Chromium] add empty impl of WebThemeEngine::getSize() for DRT
2819         https://bugs.webkit.org/show_bug.cgi?id=89907
2820
2821         Not reviewed. Fix build in Chromium.
2822
2823         * DumpRenderTree/chromium/WebThemeEngineDRTWin.cpp:
2824         (WebThemeEngineDRTWin::getSize):
2825         * DumpRenderTree/chromium/WebThemeEngineDRTWin.h:
2826         (WebThemeEngineDRTWin):
2827
2828 2012-06-25  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
2829
2830         [EFL] Implement Network Information API
2831         https://bugs.webkit.org/show_bug.cgi?id=87067
2832
2833         Reviewed by Kenneth Rohde Christiansen.
2834
2835         * efl/jhbuild.modules: Add eeze library dependency.
2836
2837 2012-06-25  Simon Hausmann  <simon.hausmann@nokia.com>
2838
2839         [Qt] Fix Win32 debug build
2840
2841         Reviewed by Tor Arne Vestbø.
2842
2843         Consolidate use_all_in_one_files into mac/unix/win32 default_pre.prf.
2844
2845         * qmake/mkspecs/features/unix/default_pre.prf:
2846         * qmake/mkspecs/features/win32/default_pre.prf: Added.
2847
2848 2012-06-24  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2849
2850         [Qt] Enable <input type="color"> support by default
2851         https://bugs.webkit.org/show_bug.cgi?id=89653
2852
2853         Reviewed by Kenneth Rohde Christiansen.
2854
2855         This feature is complete for Qt WebKit2. Enabling by default
2856         will help to mature the implementation.
2857
2858         * Scripts/webkitperl/FeatureList.pm:
2859         * qmake/mkspecs/features/features.pri:
2860
2861 2012-06-24  Dan Bernstein  <mitz@apple.com>
2862
2863         Made debug-{minibrowser,safari,test-runner} work with LLDB.
2864
2865         Reviewed by Sam Weinig.
2866
2867         * Scripts/debug-minibrowser: Pass INCLUDE_OPTIONS_FOR_DEBUGGING to
2868         printHelpAndExitForRunAndDebugWebKitAppIfNeeded().
2869         * Scripts/debug-safari: Ditto.
2870         * Scripts/debug-test-runner: Ditto.
2871         * Scripts/webkitdirs.pm:
2872         (debugger): Added. Calls determineDebugger() if needed and returns the chosen debugger.
2873         (determineDebugger): Added. Sets the debugger to "lldb" if the --use-lldb switch is present,
2874         and to "gdb" otherwise.
2875         (printHelpAndExitForRunAndDebugWebKitAppIfNeeded): Changed to print help for the
2876         --target-web-process and --use-lldb switches if passed INCLUDE_OPTIONS_FOR_DEBUGGING.
2877         (execMacWebKitAppForDebugging): Changed to use the chosen debugger.
2878
2879 2012-06-24  Adam Barth  <abarth@webkit.org>
2880
2881         [Chromium] Release media resources after each LayoutTest on Android
2882         https://bugs.webkit.org/show_bug.cgi?id=89720
2883
2884         Reviewed by Eric Carlson.
2885
2886         The LayoutTests can hang when run on Android because we sometimes run
2887         out of media resources. This patch causes us to release our media
2888         resources after each test in order to avoid running out of this
2889         resource. In production, there are other mechanisms that manage this
2890         resource.
2891
2892         * DumpRenderTree/chromium/TestShell.cpp:
2893         (TestShell::resetTestController):
2894
2895 2012-06-23  Zan Dobersek  <zandobersek@gmail.com>
2896
2897         [Gtk] REGRESSION(r120918): Causes flaky DND tests
2898         https://bugs.webkit.org/show_bug.cgi?id=89770
2899
2900         Reviewed by Martin Robinson.
2901
2902         Connect to the run-file-chooser signal of the WebKitWebView and
2903         return TRUE so no file chooser dialog is run. This prevents certain
2904         test failures.
2905
2906         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2907         (webViewRunFileChooser):
2908         (createWebView):
2909
2910 2012-06-23  Dan Bernstein  <mitz@apple.com>
2911
2912         debug-{minibrowser,safari,test-runner} fail when gdb is not installed in /usr/bin.
2913
2914         Reviewed by Sam Weinig.
2915
2916         * Scripts/webkitdirs.pm:
2917         (execMacWebKitAppForDebugging): Use xcrun to locate gdb.
2918
2919 2012-06-23  Carlos Garcia Campos  <cgarcia@igalia.com>
2920
2921         Unreviewed. Skip GTK+ unit test /webkit2/WebKitFindController/hide.
2922
2923         It fails always when running it in Xvfb.
2924
2925         * gtk/run-api-tests:
2926         (TestRunner):
2927
2928 2012-06-23  Wajahat Siddiqui  <mdwajahatali.siddiqui@motorola.com>
2929
2930         [GTK][WK2]Mis-spelt defaultWindowTitle
2931         https://bugs.webkit.org/show_bug.cgi?id=89808
2932
2933         Reviewed by Carlos Garcia Campos.
2934
2935         * MiniBrowser/gtk/BrowserWindow.c:
2936
2937 2012-06-23  Sergio Villar Senin  <svillar@igalia.com>
2938
2939         [GTK] Add a new webkit2 tests slave bot
2940         https://bugs.webkit.org/show_bug.cgi?id=89336
2941
2942         Reviewed by Csaba Osztrogonác.
2943
2944         Slave configuration for a new GTK 64 bit release bot that will run
2945         WebKit2 tests.
2946
2947         Also BuildAndTest now accepts the "triggers" parametter,
2948         this means that apart from building and testing it can trigger
2949         some other build/test sequences in other slaves.
2950
2951         * BuildSlaveSupport/build.webkit.org-config/config.json:
2952         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2953         (BuildAndTestFactory.__init__):
2954
2955 2012-06-23  Kwang Yul Seo  <skyul@company100.net>
2956
2957         Unreviewed. Update my email.
2958
2959         * Scripts/webkitpy/common/config/committers.py:
2960
2961 2012-06-22  Adam Barth  <abarth@webkit.org>
2962
2963         [Chromium] DumpRenderTree on Android needs to configure fonts for testing
2964         https://bugs.webkit.org/show_bug.cgi?id=89721
2965
2966         Reviewed by Nate Chapin.
2967
2968         This patch teaches DumpRenderTree to configure Skia to use the fallback
2969         fonts that the LayoutTests assume. This patch reduces the number result
2970         differences between chromium-android and chromium-linux.
2971
2972         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
2973         * DumpRenderTree/chromium/TestShellAndroid.cpp:
2974         (platformInit):
2975         * DumpRenderTree/chromium/android_fallback_fonts.xml: Added.
2976         * DumpRenderTree/chromium/android_main_fonts.xml: Added.
2977         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2978
2979 2012-06-22  Alexandru Chiculita  <achicu@adobe.com>
2980
2981         [CSS Shaders] Re-enable the CSS Shaders compile time flag on Safari Mac
2982         https://bugs.webkit.org/show_bug.cgi?id=89781
2983
2984         Reviewed by Dean Jackson.
2985
2986         Added CSS Shaders as enabled by default on Safari for Mac.
2987
2988         * Scripts/webkitperl/FeatureList.pm:
2989
2990 2012-06-22  Adam Barth  <abarth@webkit.org>
2991
2992         Add support for test_expectations_android.txt for overriding test expecations on the chromium-android branch
2993         https://bugs.webkit.org/show_bug.cgi?id=89791
2994
2995         Reviewed by Dirk Pranke.
2996
2997         Downstream, the chromium-android port maintains a test expectations
2998         file to keep track of which tests pass or fail. That causes them to
2999         keep a patch in chromium_android.py in their branch.
3000
3001         This patch teaches chromium_android.py to use test_expectations_android.txt.
3002         We don't intend to use this file upstream, but having this code patch
3003         upstream lets us unfork webkitpy.
3004
3005         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
3006         (ChromiumAndroidPort.expectations_files):
3007
3008 2012-06-22  Dirk Pranke  <dpranke@chromium.org>
3009
3010         run_webkit_tests.py failed with AttributeError(''NoneType' object has no attribute 'pid'')
3011         https://bugs.webkit.org/show_bug.cgi?id=89734
3012
3013         Reviewed by Ryosuke Niwa.
3014
3015         Fix a crash in ServerProcess if you called .pid() after it
3016         crashed during a write(). We had a test for this case but the
3017         test wasn't calling pid(), just has_crashed(). Fixed the problem
3018         and the test.
3019
3020         * Scripts/webkitpy/layout_tests/port/server_process.py:
3021         (ServerProcess.__init__):
3022         (ServerProcess.pid):
3023         (ServerProcess._start):
3024         (ServerProcess.stop):
3025         * Scripts/webkitpy/layout_tests/port/server_process_unittest.py:
3026         (FakeServerProcess._start):
3027         (TestServerProcess.test_broken_pipe):
3028
3029 2012-06-22  Peter Beverloo  <peter@chromium.org>
3030
3031         [Chromium] Disable c++0x compatibility warnings in JavaScriptCore.gyp when building for Android
3032         https://bugs.webkit.org/show_bug.cgi?id=88853
3033
3034         Reviewed by Steve Block.
3035
3036         The Android exclusions were necessary to fix a gyp generation error, as
3037         the gcc_version variable wasn't being defined for Android. Remove these
3038         exceptions when Chromium is able to define the gcc_version variable.
3039
3040         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
3041
3042 2012-06-22  Mario Sanchez Prada  <msanchez@igalia.com>
3043
3044         Unreviewed gardening. Unskipping test passing after fixing bug 88419.
3045
3046         * gtk/run-api-tests:
3047         (TestRunner): Unskip WTF.HashMap.
3048
3049 2012-06-22  Zoltan Horvath  <zoltan@webkit.org>
3050
3051         [Qt] Allow DumpRenderTree to dump about:blank
3052         https://bugs.webkit.org/show_bug.cgi?id=89685
3053
3054         Reviewed by Ryosuke Niwa.
3055
3056         We need to allow DumpRenderTree to dump about:blank page, then it will be consistent with other ports,
3057         additionaly this behavior is required for running WTR performance tests.
3058
3059         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
3060         (WebCore::DumpRenderTree::processLine):
3061
3062 2012-06-21  Adam Barth  <abarth@webkit.org>
3063
3064         [Chromium] Reset mediaPlaybackRequiresUserGesture WebSettings after each test
3065         https://bugs.webkit.org/show_bug.cgi?id=89718
3066
3067         Reviewed by Kent Tamura.
3068
3069         We should reset this WebSetting to its default value so that it behaves
3070         consistently across tests.
3071
3072         * DumpRenderTree/chromium/WebPreferences.cpp:
3073         (WebPreferences::reset):
3074         (WebPreferences::applyTo):
3075         * DumpRenderTree/chromium/WebPreferences.h:
3076         (WebPreferences):
3077
3078 2012-06-21  Zoltan Horvath  <zoltan@webkit.org>
3079
3080         DRT/WTR python interface handles about:blank incorrectly
3081         https://bugs.webkit.org/show_bug.cgi?id=89563
3082
3083         Reviewed by Dirk Pranke.
3084
3085         Currently DRT/WTR python interface's _command_from_driver_input function
3086         puts the test directory path before about:blank. Remove this incorrect behavior.
3087
3088         * Scripts/webkitpy/layout_tests/port/webkit.py:
3089         (WebKitDriver._command_from_driver_input):
3090
3091 2012-06-21  Dirk Pranke  <dpranke@chromium.org>
3092
3093         reenable perf tests on win
3094         https://bugs.webkit.org/show_bug.cgi?id=89690
3095
3096         Reviewed by Ryosuke Niwa.
3097
3098         Only the replay tests don't work, and those are disabled by
3099         default. This change also fixes the undefined _log reference
3100         that was causing a crash.
3101
3102         * Scripts/run-perf-tests:
3103
3104 2012-06-21  Sheriff Bot  <webkit.review.bot@gmail.com>
3105
3106         Unreviewed, rolling out r120937.
3107         http://trac.webkit.org/changeset/120937
3108         https://bugs.webkit.org/show_bug.cgi?id=89679
3109
3110         This patch brought buildbot master down (Requested by svillar
3111         on #webkit).
3112
3113         * BuildSlaveSupport/build.webkit.org-config/config.json:
3114
3115 2012-06-21  Martin Robinson  <mrobinson@igalia.com>
3116
3117         [GTK] Combine WebKit API tests into fewer binaries
3118         https://bugs.webkit.org/show_bug.cgi?id=88458
3119
3120         Reviewed by Carlos Garcia Campos.
3121
3122         Instead of creating one binary per-test file, create binaries for each
3123         category of tests. Right now this includes WTF and the WebKit2 C API,
3124         but later tests can be added for the GTK+ platform layer and the WebKit1
3125         and WebKit2 API layers.
3126
3127         * TestWebKitAPI/GNUmakefile.am: Compile only two test binaries, one for the WebKit2
3128           C API and one for WTF. Refresh the source list to ensure that new test are active.
3129         * gtk/run-api-tests: Change the way that tests are skipped by splitting out the
3130         concept of skipping a test and skipping a suite (program) of tests. Test cases are
3131         skipped because of legitimate failures, but entire programs are skipped because of
3132         problems in the harness. As of right now a test program is only skipped if the
3133         accessibility bus cannot be started.
3134         (SkippedTest.__init__): Make the test case a required argument and have one skipped
3135         test case per SkippedTest instance.
3136         (SkippedTest.__str__): Ditto.
3137         (TestRunner): Update the directory list. We only have two gtest binaries now and they are
3138         in one directory. Reformat the test list to make it slightly easier to read.
3139         (TestRunner.__init__): Add the _skipped_test_program member, which handles entire
3140         test programs that are skipped.
3141         (TestRunner._setup_testing_environment): Use the new member.
3142         (TestRunner._test_cases_to_skip): Collect all skipped test cases now instead of just
3143         the first SkippedTest that matches.
3144         (TestRunner._should_run_test_program): Take a look at the new member to make this
3145         decision.
3146         (TestRunner._run_test_command): Use the name test_program instead of test
3147         to disambiguate between test cases and test suites.
3148         (TestRunner._run_test_glib): ditto.
3149         (TestRunner._run_test_google): Ditto.
3150         (TestRunner._run_test): Ditto.
3151         (TestRunner.run_tests): Ditto.
3152
3153 2012-06-21  Yong Li  <yoli@rim.com>
3154
3155         Unreviewed. Moving myself from committer to reviewer.
3156
3157         * Scripts/webkitpy/common/config/committers.py:
3158
3159 2012-06-21  Sergio Villar Senin  <svillar@igalia.com>
3160
3161         [GTK] Add a new webkit2 tests slave bot
3162         https://bugs.webkit.org/show_bug.cgi?id=89336
3163
3164         Reviewed by Philippe Normand.
3165
3166         Slave configuration for a new GTK bot that will run WebKit2 tests.
3167
3168         * BuildSlaveSupport/build.webkit.org-config/config.json:
3169
3170 2012-06-21  Thiago Marcos P. Santos  <thiago.santos@intel.com>
3171
3172         [Qt] Add a custom Color Chooser widget to MiniBrowser
3173         https://bugs.webkit.org/show_bug.cgi?id=87988
3174
3175         Reviewed by Alexis Menard.
3176
3177         Added custom color chooser example for the MiniBrowser.
3178
3179         * MiniBrowser/qt/MiniBrowser.qrc:
3180         * MiniBrowser/qt/qml/BrowserWindow.qml:
3181         * MiniBrowser/qt/qml/ColorChooser.qml: Added.
3182
3183 2012-06-21  Vivek Galatage  <vivekgalatage@gmail.com>
3184
3185         Remove the warning "File not found" in MiniBrowser post-build event
3186         https://bugs.webkit.org/show_bug.cgi?id=89601
3187
3188         Reviewed by Adam Roben.
3189
3190         * MiniBrowser/MiniBrowserPostBuild.cmd:
3191
3192 2012-06-21  Balazs Kelemen  <kbalazs@webkit.org>
3193
3194         [Qt] DRT in standalone mode hangs after the first test
3195         https://bugs.webkit.org/show_bug.cgi?id=89613
3196
3197         Reviewed by Csaba Osztrogonác.
3198
3199         Break an inline signal slot chain that ends up in setting
3200         LayoutTestController::m_hasDumped too early. In LayoutTestController::maybeDump()
3201         we emit done() which starts the chain that finishes in DumpRenderTree::open().
3202         The next line in maybeDump sets m_hasDumped to true. We could simply reorder
3203         these two lines but it seems to be wrong in general to start the next load
3204         from this call chain. The server mode (when DRT is runned by the test harness)
3205         also uses a queued connection to read the next test from stdin.
3206
3207         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
3208         (WebCore::DumpRenderTree::processArgsLine):
3209
3210 2012-06-20  Hans Wennborg  <hans@chromium.org>
3211
3212         Speech JavaScript API: SpeechRecognition should hook up with ActiveDOMObject more
3213         https://bugs.webkit.org/show_bug.cgi?id=89217
3214
3215         Reviewed by Adam Barth.
3216
3217         Add a method for checking whether the mock speech
3218         recognition was aborted.
3219
3220         Also redo the way the mock posts tasks. Instead of posting them all at once,
3221         maintain an internal queue of task objects, and call postTask() for them
3222         once at the time. This means that for example when the page is navigated
3223         away and abort() is called, that call doesn't end up after a bunch
3224         of previously posted events on the event loop.
3225
3226         * DumpRenderTree/chromium/LayoutTestController.cpp:
3227         (LayoutTestController::LayoutTestController):
3228         (LayoutTestController::wasMockSpeechRecognitionAborted):
3229         * DumpRenderTree/chromium/LayoutTestController.h:
3230         (LayoutTestController):
3231         * DumpRenderTree/chromium/MockWebSpeechRecognizer.cpp:
3232         (WebKit::ClientCallTask::ClientCallTask):
3233         (WebKit::ResultTask::ResultTask):
3234         (WebKit::NoMatchTask::NoMatchTask):
3235         (WebKit::ErrorTask::ErrorTask):
3236         (MockWebSpeechRecognizer::start):
3237         (MockWebSpeechRecognizer::abort):
3238         (MockWebSpeechRecognizer::setError):
3239         (MockWebSpeechRecognizer::MockWebSpeechRecognizer):
3240         (MockWebSpeechRecognizer::startTaskQueue):
3241         (MockWebSpeechRecognizer::StepTask::runIfValid):
3242         * DumpRenderTree/chromium/MockWebSpeechRecognizer.h:
3243         (MockWebSpeechRecognizer::hasBeenAborted):
3244         (MockWebSpeechRecognizer):
3245         (MockWebSpeechRecognizer::taskList):
3246         (Task):
3247         (MockWebSpeechRecognizer::Task::Task):
3248         (MockWebSpeechRecognizer::Task::~Task):
3249         (StepTask):
3250         (MockWebSpeechRecognizer::StepTask::StepTask):
3251
3252 2012-06-21  Mario Sanchez Prada  <msanchez@igalia.com>
3253
3254         [GTK] Add support for window.showModalDialog in WebKit2GTK+
3255         https://bugs.webkit.org/show_bug.cgi?id=79500
3256
3257         Reviewed by Carlos Garcia Campos.
3258
3259         Add support for modal dialogs in GTK's MiniBrowser.
3260
3261         * MiniBrowser/gtk/BrowserWindow.c:
3262         (webViewRunAsModal):
3263         (webViewCreate):
3264         (webViewDecidePolicy):
3265         (browser_window_new):
3266         * MiniBrowser/gtk/BrowserWindow.h:
3267         * MiniBrowser/gtk/main.c:
3268         (createBrowserWindow):
3269
3270 2012-06-20  Brent Fulgham  <bfulgham@webkit.org>
3271
3272         [WinCairo] Unreviewed build fix.
3273         The wrapper script that launched new- and old-run-webkit-tests
3274         was discarding the --wincairo argument needed to get the proper
3275         test infrastructure to build under WinCairo.
3276
3277         * Scripts/run-webkit-tests: Emulate the Qt, Wx, Chromium, etc.,
3278         behavior to chain the --wincairo flag through to the new build
3279         and test scripts.
3280
3281 2012-06-20  Mark Rowe  <mrowe@apple.com>
3282
3283         <http://webkit.org/b/89606> Teach run-safari and debug-safari to work with a Safari.app that has entitlements
3284
3285         Reviewed by Dan Bernstein.
3286
3287         * Scripts/webkitdirs.pm:
3288         (executableHasEntitlements):
3289         (safariPathFromSafariBundle):
3290
3291 2012-06-20  Sheriff Bot  <webkit.review.bot@gmail.com>
3292
3293         Unreviewed, rolling out r120821.
3294         http://trac.webkit.org/changeset/120821
3295         https://bugs.webkit.org/show_bug.cgi?id=89605
3296
3297         It made duplicated reviewer entries (Requested by Ossy on
3298         #webkit).
3299
3300         * Scripts/webkitpy/common/checkout/changelog.py:
3301         (ChangeLog.set_reviewer):
3302         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
3303         (test_set_reviewer):
3304         (test_set_short_description_and_bug_url):
3305
3306 2012-06-20  Dirk Pranke  <dpranke@chromium.org>
3307
3308         nrwt outputs empty files for wdiff output if wdiff is not installed
3309         https://bugs.webkit.org/show_bug.cgi?id=88709
3310
3311         Reviewed by Tony Chang.
3312
3313         Reviewed by Tony Chang.
3314
3315         Don't write -wdiff or -pretty.html files if wdiff or prettypatch
3316         aren't available, and clean up the handling for them in the port
3317         code.
3318
3319         * Scripts/webkitpy/run_webkit_tests_integrationtest.py:
3320         (MainTest.test_output_diffs):
3321         * Scripts/webkitpy/layout_tests/controllers/test_result_writer.py:
3322         (TestResultWriter.create_text_diff_and_write_result):
3323         * Scripts/webkitpy/layout_tests/port/base.py:
3324         (Port.check_pretty_patch):
3325         (Port.check_wdiff):
3326         (Port._wdiff_missing_message):
3327         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
3328         (ChromiumLinuxPort.check_build):
3329         (ChromiumLinuxPort._wdiff_missing_message):
3330         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
3331         (ChromiumMacPort.check_build):
3332         (ChromiumMacPort):
3333         (ChromiumMacPort._wdiff_missing_message):
3334         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
3335         (ChromiumMacPortTest):
3336
3337 2012-06-20  Dirk Pranke  <dpranke@chromium.org>
3338
3339         nrwt: fix unit tests for ensuring svn revision is correct
3340         https://bugs.webkit.org/show_bug.cgi?id=89498
3341
3342         Reviewed by Eric Seidel.
3343
3344         Reviewed by Eric Seidel.
3345
3346         Fix the unit tests for testing that we are embedding the SVN
3347         revision in the results json for NRWT properly; this was broken
3348         in r120646 but because the MockHost worked differently than the
3349         real one (by always initializing the MockSCM object) we didn't
3350         notice. Unfortunately, just changing the default breaks all
3351         sorts of unit tests ...
3352
3353         * Scripts/webkitpy/common/host_mock.py:
3354         (MockHost.__init__):
3355         (MockHost._initialize_scm):
3356         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
3357         (ResultSummaryTest.test_no_svn_revision):
3358         (ResultSummaryTest.test_svn_revision):
3359
3360 2012-06-20  Dirk Pranke  <dpranke@chromium.org>
3361
3362         new-run-webkit-tests appends "/Debug" or "/Release" to $WEBKITOUTPUTDIR
3363         https://bugs.webkit.org/show_bug.cgi?id=69360
3364
3365         Reviewed by Eric Seidel.
3366
3367         Reviewed by Eric Seidel.
3368
3369         Propagate the 'port_implementation' part of the platform (i.e.,
3370         gtk,qt,chromium) to webkit-build-directory so that we can pick
3371         up the gtk-specific handling of WEBKITOUTPUTDIR ...
3372
3373         I didn't write any additional tests for this; testing it
3374         properly is an integration test between the python code and the
3375         perl code, which I verified by hand.
3376
3377         * Scripts/webkitpy/layout_tests/port/base.py:
3378         (Port.__init__):
3379         * Scripts/webkitpy/layout_tests/port/config.py:
3380         (Config.__init__):
3381         (Config.build_directory):
3382         * Scripts/webkitpy/layout_tests/port/config_mock.py:
3383         (MockConfig.__init__):
3384         * Scripts/webkitpy/layout_tests/port/config_unittest.py:
3385         (ConfigTest.test_build_directory_passes_port_implementation):
3386
3387 2012-06-20  Dirk Pranke  <dpranke@chromium.org>
3388
3389         Fix import sorting missed in r120846
3390
3391         Reviewed by Tony Chang.
3392
3393         * Scripts/webkitpy/layout_tests/servers/apache_http_server_unittest.py:
3394
3395 2012-06-20  Dirk Pranke  <dpranke@chromium.org>
3396
3397         Fix chromium win http servers after breakage introduced in r120846.
3398
3399         Unreviewed, build fix.
3400
3401         * Scripts/webkitpy/layout_tests/port/base.py:
3402         (Port.to.start_http_server):
3403         * Scripts/webkitpy/layout_tests/servers/http_server.py:
3404         (Lighttpd.__init__):
3405
3406 2012-06-20  Dirk Pranke  <dpranke@chromium.org>
3407
3408         tweak output of webkit-patch print-{baselines,expectations}
3409         https://bugs.webkit.org/show_bug.cgi?id=89588
3410
3411         Reviewed by Tony Chang.
3412
3413         Minor tweaks to the output to make it more readable, including
3414         turning off csv by default for print-baselines with multiple ports.
3415
3416         * Scripts/webkitpy/layout_tests/port/test.py:
3417         (TestPort):
3418         * Scripts/webkitpy/tool/commands/queries.py:
3419         (PrintExpectations.execute):
3420         (PrintBaselines.execute):
3421         * Scripts/webkitpy/tool/commands/queries_unittest.py:
3422         (PrintExpectationsTest.run_test):
3423         (PrintExpectationsTest.test_multiple):
3424         (PrintBaselinesTest.setUp):
3425         (PrintBaselinesTest.test_multiple):
3426
3427 2012-06-18  Dirk Pranke  <dpranke@chromium.org>
3428
3429         new-run-webkit-tests should spin-up enough httpd processes to avoid timeouts
3430         https://bugs.webkit.org/show_bug.cgi?id=88134
3431
3432         Reviewed by Tony Chang.
3433
3434         Change NRWT to spin up 2*min(child_processes, locked_shards)
3435         http servers by default so that we are less likely to get a
3436         bunch of http timeouts at the beginning of a test run.
3437
3438         Note that I had to tweak executive_mock to support mocked stderr
3439         because the apache_http_server code reads stderr when starting a
3440         process to ensure it started okay.
3441
3442         * Scripts/webkitpy/common/system/executive_mock.py:
3443         (MockProcess.__init__):
3444         * Scripts/webkitpy/common/system/outputcapture.py:
3445         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3446         (Manager._run_tests):
3447         (Manager.start_servers_with_lock):
3448         * Scripts/webkitpy/layout_tests/port/base.py:
3449         (Port.to.start_http_server):
3450         * Scripts/webkitpy/layout_tests/port/test.py:
3451         (TestPort.start_http_server):
3452         (TestPort._path_to_apache):
3453         (TestPort):
3454         (TestPort._path_to_apache_config_file):
3455         * Scripts/webkitpy/layout_tests/servers/apache_http_server.py:
3456         (LayoutTestApacheHttpd.__init__):
3457         * Scripts/webkitpy/layout_tests/servers/apache_http_server_unittest.py: Added
3458         (LayoutTestApacheHttpd.__init__):
3459         * Scripts/webkitpy/layout_tests/servers/http_server.py:
3460         (Lighttpd.__init__):
3461         * Scripts/webkitpy/layout_tests/servers/http_server_base.py:
3462         (HttpServerBase.__init__):
3463
3464 2012-06-20  Tom Zakrajsek  <tomz@codeaurora.org>
3465
3466         Unreviewed: Back out accidentally checked in debug print which broke a test
3467
3468         * Scripts/webkitpy/performance_tests/perftest.py:
3469         (PerfTest.parse_output):
3470
3471 2012-06-20  Tom Zakrajsek  <tomz@codeaurora.org>
3472
3473         Import themaninblue.com/experiment/AnimationBenchmark/ as performance tests
3474         https://bugs.webkit.org/show_bug.cgi?id=78789
3475
3476         Reviewed by Ryosuke Niwa.
3477
3478         Updated parser to include fps as a valid unit.
3479
3480         * Scripts/webkitpy/performance_tests/perftest.py:
3481         (PerfTest):
3482         (PerfTest.parse_output):
3483
3484 2012-06-20  Balazs Ankes  <bank@inf.u-szeged.hu>
3485
3486         webkit-patch should add reviewer if "Reviewed by NOBODY ..." is missing
3487         https://bugs.webkit.org/show_bug.cgi?id=67935
3488
3489         Reviewed by Ryosuke Niwa.
3490
3491         * Scripts/webkitpy/common/checkout/changelog.py:
3492         (ChangeLog.set_reviewer):
3493         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
3494         (test_set_reviewer):
3495         (test_set_short_description_and_bug_url):
3496
3497 2012-06-20  Simon Hausmann  <simon.hausmann@nokia.com>
3498
3499         Unreviewed trivial permissions fix.
3500
3501         * qmake/mkspecs/features/features.prf: Don't mark this file as executable.
3502
3503 2012-06-19  Ryuan Choi  <ryuan.choi@samsung.com>
3504
3505         [EFL][Regression] Build break after r120786
3506         https://bugs.webkit.org/show_bug.cgi?id=89549
3507
3508         Unreviewed build fix for Efl port.
3509
3510         * DumpRenderTree/efl/CMakeLists.txt: Added disassembler in includes.
3511
3512 2012-06-19  Dirk Pranke  <dpranke@chromium.org>
3513
3514         NRWT spins up and down the WebSocket server when running a single HTTP test from the command line
3515         https://bugs.webkit.org/show_bug.cgi?id=64489
3516
3517         Reviewed by Tony Chang.
3518
3519         This change changes NRWT so that we don't start the websocket
3520         server if we aren't running websocket tests and don't start the
3521         HTTP server if we aren't running http tests.
3522         
3523         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3524         (Manager._websocket_tests):
3525         (Manager.start_servers_with_lock):
3526         (Manager.stop_servers_with_lock):
3527         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
3528         (ManagerTest.test_servers_started):
3529
3530 2012-06-19  Dirk Pranke  <dpranke@chromium.org>
3531
3532         test-webkitpy: add a -p flag to pass through captured output to enable debugging
3533         https://bugs.webkit.org/show_bug.cgi?id=89158
3534
3535         Reviewed by Eric Seidel.
3536
3537         Normally when outputcapture runs it intercepts stdout/stderr,
3538         which makes running under the debugger difficult and mostly
3539         pointless (since you can't see any output). This change adds a
3540         flag to test-webkitpy (-p) that will cause outputcapture to pass
3541         through the output as well as capture it.
3542
3543         * Scripts/webkitpy/common/system/outputcapture.py:
3544         (OutputCapture.stream_wrapper):
3545         (OutputCapture._capture_output_with_name):
3546         * Scripts/webkitpy/common/system/outputcapture_unittest.py:
3547         (OutputCaptureTest.setUp):
3548         * Scripts/webkitpy/test/main.py:
3549         (Tester._parse_args):
3550         (Tester._run_tests):
3551         (Tester._log_exception):
3552         (_CaptureAndPassThroughStream):
3553         (_CaptureAndPassThroughStream.__init__):
3554         (_CaptureAndPassThroughStream.write):
3555         (_CaptureAndPassThroughStream._message_is_from_pdb):
3556         (_CaptureAndPassThroughStream.flush):
3557         (_CaptureAndPassThroughStream.getvalue):
3558
3559 2012-06-19  Adam Barth  <abarth@webkit.org>
3560
3561         garden-o-matic fails to rebaseline tests with MISSING results
3562         https://bugs.webkit.org/show_bug.cgi?id=89521
3563
3564         Reviewed by Kenneth Russell.
3565
3566         If the bots report "MISSING", we should try to find PNG and TXT results on the server.
3567
3568         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results.js:
3569
3570 2012-06-19  Adam Barth  <abarth@webkit.org>
3571
3572         svn.delete_list fails to delete empty parent directories
3573         https://bugs.webkit.org/show_bug.cgi?id=89520
3574
3575         Reviewed by Dirk Pranke.
3576
3577         These functions were calling scm.delete and scm.add, which already
3578         delete the parent directory. We'd get an exception when we tried to
3579         delete the parent directory again.
3580
3581         * Scripts/webkitpy/common/checkout/scm/svn.py:
3582         (SVN._add_parent_directories):
3583         (SVN._delete_parent_directories):
3584
3585 2012-06-19  Thiago Marcos P. Santos  <thiago.santos@intel.com>
3586
3587         webkitpy: Simplify fetch_bugs_matching_search()
3588         https://bugs.webkit.org/show_bug.cgi?id=89497
3589
3590         Reviewed by Adam Barth.
3591
3592         Removed parameter with default value. It's not being used
3593         and the handling of the parameter was wrong. Saves a check.
3594
3595         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
3596         (BugzillaQueries.fetch_bugs_matching_search):
3597         * Scripts/webkitpy/common/net/bugzilla/bugzilla_mock.py:
3598         (MockBugzillaQueries.fetch_bugs_matching_search):
3599
3600 2012-06-19  Dirk Pranke  <dpranke@chromium.org>
3601
3602         webkitpy: remove unneeded imports in layout_tests/port/chromium.py
3603         https://bugs.webkit.org/show_bug.cgi?id=89522
3604
3605         Reviewed by Eric Seidel.
3606
3607         They were no longer being used. No tests necessary.
3608
3609         * Scripts/webkitpy/layout_tests/port/chromium.py:
3610
3611 2012-06-19  Adam Barth  <abarth@webkit.org>
3612
3613         Missing results aren't shown in garden-o-matic
3614
3615         Reviewed by Kenneth Russell.
3616
3617         Previously we didn't know to look for text and image results when the
3618         bot told us that results were missing.
3619
3620         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results.js:
3621
3622 2012-06-19  Dirk Pranke  <dpranke@chromium.org>
3623
3624         Fix regression introduced in r120646 where the svn revision isn't saved in results.json
3625
3626         Unreviewed, build fix.
3627
3628         It turns out that we store the svn revision two different ways
3629         in the JSON, and I only caught one of them :(. This should fix
3630         the other.
3631
3632         Also, this part of the code is poorly tested. I will investigate
3633         adding at least a basic test for this in a separate fix.
3634
3635         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3636         (summarize_results):
3637
3638 2012-06-19  Dirk Pranke  <dpranke@chromium.org>
3639
3640         REGRESSION (NRWT): Results for new non-text-only tests are always put in the most-specific platform directory
3641         https://bugs.webkit.org/show_bug.cgi?id=78127
3642
3643         Reviewed by Ryosuke Niwa.
3644
3645         Add the --add-platform-exceptions flag from ORWT, and ensure
3646         that --new-baseline is equivalent to --reset-results
3647         --add-platform-exceptions. Also fix the default logic for where to
3648         put PNGs and render tree dumps if --new-test-results is true:
3649         if --add-platform-exceptions is False, baselines should go in
3650         the least-specific platform directory (e.g., platform/mac)
3651         rather than the most-specific (platform/mac-snowleopard).
3652
3653         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
3654         (SingleTestRunner):
3655         (SingleTestRunner.run):
3656         (SingleTestRunner._run_rebaseline):
3657         (SingleTestRunner._add_missing_baselines):
3658         (SingleTestRunner._location_for_new_baseline):
3659         (SingleTestRunner._overwrite_baselines):
3660         (SingleTestRunner._save_baseline_data):
3661         * Scripts/webkitpy/layout_tests/port/base.py:
3662         (Port.baseline_path):
3663         (Port):
3664         (Port.baseline_platform_dir):
3665         (Port.baseline_version_dir):
3666         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3667         (_set_up_derived_options):
3668         (parse_args):
3669         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3670         (RebaselineTest.assertBaselines):
3671         (RebaselineTest.test_reset_results):
3672         (RebaselineTest.test_missing_results):
3673         (RebaselineTest.test_new_baseline):