e0f8a62e7aa78f0d2ba6e3b075dd648f31bc8251
[WebKit-https.git] / Tools / ChangeLog
1 2012-07-06  Jessie Berlin  <jberlin@apple.com>
2
3         WKContext should ask for its initialization data when it needs it so the client doesn't have
4         to keep it up to date.
5         https://bugs.webkit.org/show_bug.cgi?id=90627
6
7         Reviewed by Anders Carlsson.
8
9         Add tests and update other WKContextInjectedBundleClients.
10
11         * MiniBrowser/mac/AppDelegate.m:
12         (-[BrowserAppDelegate init]):
13         Updated for the change to WKContextInjectedBundleClient.
14         * WebKitTestRunner/TestController.cpp:
15         (WTR::TestController::initialize):
16         Ditto.
17
18         * TestWebKitAPI/PlatformUtilities.cpp:
19         (TestWebKitAPI::Util::createInitializationDictionaryForInjectedBundleTest):
20         Moved the logic to create the initialization dictionary here ...
21         (TestWebKitAPI::Util::createContextForInjectedBundleTest):
22         ... from here so that it can be used without automatically using
23         WKContextSetInitializationUserDataForInjectedBundle.
24         * TestWebKitAPI/PlatformUtilities.h:
25
26         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
27         Add the new test files.
28         * TestWebKitAPI/GNUmakefile.am:
29         Ditto.
30
31         * TestWebKitAPI/Tests/WebKit2/GetInjectedBundleInitializationUserDataCallback.cpp: Added.
32         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
33         Check that the message received from the injected bundle matches the user data it was
34         initialized with.
35         (TestWebKitAPI::getInjectedBundleInitializationUserData):
36         Return the user data that the injected bundle should be initialized with.
37         (TestWebKitAPI::TEST):
38         Set up WKContextInjectedBundleClient and load a page.
39         * TestWebKitAPI/Tests/WebKit2/GetInjectedBundleInitializationUserDataCallback_Bundle.cpp: Added.
40         (TestWebKitAPI::GetInjectedBundleInitializationUserDataCallbackTest::initialize):
41         Send the initialization user data back up to the UI Process.
42
43         * TestWebKitAPI/Tests/WebKit2/InjectedBundleInitializationUserDataCallbackWins.cpp: Added.
44         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
45         Check that the message received from the injected bundle matches the user data it was
46         initialized with in the callback.
47         (TestWebKitAPI::getInjectedBundleInitializationUserData):
48         Return the user data that the injected bundle should be initialized with.
49         (TestWebKitAPI::TEST):
50         Set up the context and use WKContextSetInitializationUserDataForInjectedBundle to set the
51         initialization user data (which should be overridden by the user data returned in
52         getInjectedBundleInitializationUserData).
53         * TestWebKitAPI/Tests/WebKit2/InjectedBundleInitializationUserDataCallbackWins_Bundle.cpp: Added.
54         (TestWebKitAPI::InjectedBundleInitializationUserDataCallbackWinsTest::initialize):
55         Send the initialization user data back up to the UI Process.
56
57 2012-07-06  Dongwoo Im  <dw.im@samsung.com>
58
59         [EFL][GTK] jhbuild : Disable pixman demos build depending on GTK+
60         https://bugs.webkit.org/show_bug.cgi?id=90593
61
62         Reviewed by Philippe Normand.
63
64         * efl/jhbuild.modules: Give '--enable-gtk=no' option to pixman.
65         * gtk/jhbuild.modules: Give '--enable-gtk=no' option to pixman.
66
67 2012-07-06  Luciano Wolf <luciano.wolf@openbossa.org>
68
69         [Qt] DumpRenderTree does not use 'monospace' font when directed
70         https://bugs.webkit.org/show_bug.cgi?id=85203
71
72         The test fonts used for Qt tests were moved to the Liberation font family.
73
74         Due to this change we are skipping tons of tests. They will be gradually
75         unskipped in batches, ASAP.
76
77         Also, the now unneeded special font configuration for Qt5 WK1 was removed.
78
79         Reviewed by Tor Arne Vestbø.
80
81         * DumpRenderTree/qt/QtInitializeTestFonts.cpp:
82         (WebKit::initializeTestFonts):
83
84 2012-07-06  Csaba Osztrogonác  <ossy@webkit.org>
85
86         [Qt] Buildfix for newer Qt5
87         https://bugs.webkit.org/show_bug.cgi?id=90519
88
89         Reviewed by Tor Arne Vestbø.
90
91         * MiniBrowser/qt/MiniBrowser.pro: We also need to depend on gui-private in the MiniBrowser to get access to these headers.
92         * MiniBrowser/qt/MiniBrowserApplication.h: Include qpa/qwindowsysteminterface.h instead of deprecated qwindowsysteminterface_qpa.h
93         * WebKitTestRunner/qt/PlatformWebViewQt.cpp: Include qpa/qwindowsysteminterface.h instead of deprecated qwindowsysteminterface_qpa.h
94
95 2012-07-06  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
96
97         [Qt] Make use of Qt5 qmake's changed makefile recursion behavior
98
99         When not using the -r option, qmake now interleaves qmake and make calls,
100         so we don't need custom logic for this. There's also an option to supress
101         the effect of the -r option, which we use to make WebKit.pro the only
102         project file parsed in a recursive qmake-run.
103
104         https://bugs.webkit.org/show_bug.cgi?id=90461
105
106         Reviewed by Tor Arne Vestbø.
107
108         * qmake/mkspecs/features/default_post.prf:
109         * qmake/mkspecs/features/functions.prf:
110
111 2012-07-06  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
112
113         [Qt] Remove custom qmake logic for module creation
114
115         Qmake now has the necessary hooks to cleanly override the build locations.
116
117         https://bugs.webkit.org/show_bug.cgi?id=90461
118
119         Reviewed by Tor Arne Vestbø.
120
121         * qmake/mkspecs/features/default_pre.prf:
122
123 2012-07-06  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
124
125         [Qt] Add top-level .qmake.conf
126
127         With Qt5, this makes setting $QMAKEPATH externally unnecessary.
128
129         The magic in the perl scripts to set QMAKEPATH is still there, as it doesn't
130         hurt, and is still required for Qt4.
131
132         https://bugs.webkit.org/show_bug.cgi?id=90461
133
134         Reviewed by Tor Arne Vestbø.
135
136         * qmake/mkspecs/features/default_pre.prf:
137
138 2012-07-06  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
139
140         [Qt] Unify qtFeatureDefaults code paths
141
142         There's no reason to run qmake on features.prf directly anymore, as we
143         can selectivly run configure tests in the project file now.
144
145         https://bugs.webkit.org/show_bug.cgi?id=90461
146
147         Reviewed by Tor Arne Vestbø.
148
149         * Scripts/webkitdirs.pm:
150         (qtFeatureDefaults):
151         * qmake/configure.pri:
152         * qmake/mkspecs/features/features.prf:
153
154 2012-07-06  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
155
156         [Qt] Don't let qt_webkit.pri proclaim its own location
157
158         This won't work any more with recent Qt5 versions, as the forwarding
159         pri is created by qt_module_config, which needs MODULE_PRI to be set
160         up already.
161
162         We also need to load build_config, not qt_module.
163
164         https://bugs.webkit.org/show_bug.cgi?id=90461
165
166         Reviewed by Tor Arne Vestbø.
167
168         * qmake/qt_webkit.pri:
169
170 2012-07-06  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
171
172         [Qt] Remove redundant CONFIG+=module
173
174         qt_module_config takes care of that.
175
176         In api.pri we are actually testing the flag ourselves, so now we need to
177         test a related flag qt_module_config sets instead.
178
179         Reviewed by Tor Arne Vestbø.
180
181         * qmake/mkspecs/features/default_post.prf:
182
183 2012-07-06  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
184
185         [Qt] Switch to new-style Qt 5 configure tests
186
187         Use explicit project file action instead of syncqt magic.
188
189         https://bugs.webkit.org/show_bug.cgi?id=90461
190
191         Reviewed by Tor Arne Vestbø.
192
193         * Scripts/webkitdirs.pm:
194         (buildQMakeProjects):
195         * qmake/configure.pri: Copied from Tools/qmake/configure.pro.
196         * qmake/configure.pro:
197         * qmake/mkspecs/features/default_pre.prf:
198         * qmake/mkspecs/features/features.prf:
199         * qmake/sync.profile: Removed.
200
201 2012-07-05  Jason Liu  <jason.liu@torchmobile.com.cn>
202
203         Add Jason Liu to committers.py.
204         https://bugs.webkit.org/show_bug.cgi?id=90654
205
206         Unreviewed.
207
208         * Scripts/webkitpy/common/config/committers.py:
209
210 2012-07-05  Benjamin Poulain  <bpoulain@apple.com>
211
212         Double release of resources if the load is canceled in a callback of ResourceLoader::didFinishLoading
213         https://bugs.webkit.org/show_bug.cgi?id=90431
214
215         Reviewed by Anders Carlsson.
216
217         Add a Mac API test.
218
219         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
220         * TestWebKitAPI/Tests/mac/CancelLoadFromResourceLoadDelegate.html: Added.
221         * TestWebKitAPI/Tests/mac/CancelLoadFromResourceLoadDelegate.mm: Added.
222         (-[CancelLoadFromResourceLoadDelegate webView:resource:didFinishLoadingFromDataSource:]):
223         (-[CancelLoadFromResourceLoadDelegateFrameLoadDelegate webView:didFinishLoadForFrame:]):
224         (TestWebKitAPI):
225         (TestWebKitAPI::TEST):
226
227 2012-07-05  Dave Tharp  <dtharp@codeaurora.org>
228
229         Adding myself as committer to committers.py
230
231         Unreviewed.
232
233         * Scripts/webkitpy/common/config/committers.py:
234
235 2012-07-05  Alexey Proskuryakov  <ap@apple.com>
236
237         [Mac][WK2] Enable HTTPS tests
238         https://bugs.webkit.org/show_bug.cgi?id=90600
239
240         Reviewed by Brady Eidson.
241
242         * DumpRenderTree/mac/DumpRenderTree.mm: (dumpRenderTree): Removed a misleading comment.
243         It explained why we were doing this here instead of a frame load delegate, but that's
244         just history, not an explanation of why we need this code.
245
246         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm: (WTR::InjectedBundle::platformInitialize):
247         Ignore certificate errors for localhost and 127.0.0.1. Since this is only needed
248         for tests, it seems appropriate to have platform specific solutions in every WTR,
249         and not add a WebKit2 API.
250
251 2012-07-05  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
252
253         [Qt] Remove rpath-hacks in qt_webkit.pri
254
255         Having it in qt_webkit.pri meant that all client apps would inherit the rpath,
256         even if they were not using QtWebKit.
257
258         This is a regression for people who build client apps against a non-installed
259         QtWebKit, as the client app will be missing a rpath to the standalone WebKit
260         build dir. This can be solved by LD_LIBRARY_PATH or custom QMAKE_RPATHDIR
261         logic, until the real issue is fixed in Qt.
262
263         Reviewed by Csaba Osztrogonác.
264
265         * qmake/mkspecs/features/rpath.prf:
266         * qmake/qt_webkit.pri:
267
268 2012-07-05  Dongwoo Im  <dw.im@samsung.com>
269
270         [EFL] Enable the CUSTOM_SCHEME_HANDLER feature as default.
271         https://bugs.webkit.org/show_bug.cgi?id=88608
272
273         Reviewed by Hajime Morita.
274
275         * Scripts/webkitperl/FeatureList.pm: Enable the CUSTOM_SCHEME_HANDLER feature as default.
276
277 2012-07-05  Ryuan Choi  <ryuan.choi@samsung.com>
278
279         [Wk2][EFL] EFL needs a WebKitTestRunner
280         https://bugs.webkit.org/show_bug.cgi?id=87659
281
282         Reviewed by Kenneth Rohde Christiansen.
283
284         Add an implementation of WebKitTestRunner for Efl port.
285
286         * CMakeLists.txt: Includes WebKitTestRunner build script.
287         * Scripts/build-webkittestrunner: Added knowledge of WebKitTestRunner/Efl.
288         * Scripts/webkitpy/layout_tests/port/efl.py: Ditto.
289         (EflPort.setup_environ_for_server):
290         * WebKitTestRunner/CMakeLists.txt: Added.
291         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
292         * WebKitTestRunner/InjectedBundle/efl/ActivateFontsEfl.cpp: Added.
293         * WebKitTestRunner/InjectedBundle/efl/InjectedBundleEfl.cpp: Added.
294         * WebKitTestRunner/InjectedBundle/efl/LayoutTestControllerEfl.cpp: Added.
295         * WebKitTestRunner/PlatformEfl.cmake: Added.
296         * WebKitTestRunner/PlatformWebView.h:
297         * WebKitTestRunner/TestController.cpp:
298         * WebKitTestRunner/config.h: Added knowledge of WebKitTestRunner/Efl.
299         * WebKitTestRunner/efl/PlatformWebViewEfl.cpp: Added.
300         * WebKitTestRunner/efl/TestControllerEfl.cpp: Added.
301         * WebKitTestRunner/efl/main.cpp: Added.
302         (main):
303
304 2012-07-05  Christophe Dumez  <christophe.dumez@intel.com>
305
306         [WK2][EFL] Ewk_View needs API to load HTML data
307         https://bugs.webkit.org/show_bug.cgi?id=90540
308
309         Reviewed by Kenneth Rohde Christiansen.
310
311         Update EFL MiniBrowser to catch the "load,error" signal
312         on the view and display an error page.
313
314         * MiniBrowser/efl/main.c:
315         (on_error):
316         (browserCreate):
317
318 2012-07-05  Sergio Villar Senin  <svillar@igalia.com>
319
320         Hidden dirs are not copied when creating the built product archive
321         https://bugs.webkit.org/show_bug.cgi?id=90559
322
323         Reviewed by Carlos Garcia Campos.
324
325         Hidden dirs are not copied inside the built product archive
326         because the recursive copy command was ignoring them. We need those
327         hidden directories for the GTK WebKit2 testing bot to work fine.
328
329         * BuildSlaveSupport/built-product-archive:
330         (archiveBuiltProduct): replaced "*" by "." to include hidden
331         dirs/files.
332
333 2012-07-04  Yoshifumi Inoue  <yosin@chromium.org>
334
335         Unreviewed Chromium gardening, Disable always failed python tests on Windows.
336         https://bugs.webkit.org/show_bug.cgi?id=90587
337
338         PortTest.qt_linux and mac_lion are always failed on Chromium Windows at
339         creating child process.
340
341         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
342         (PortTest.disabled_test_qt_linux):
343         (PortTest.disabled_test_mac_lion):
344
345 2012-07-04  Yoshifumi Inoue  <yosin@chromium.org>
346
347         [Tools] webkit_unittest.py got assertion
348         https://bugs.webkit.org/show_bug.cgi?id=90579
349
350         Reviewed by Hajime Morita.
351
352         This patch updates expectation of assertion in test_skipped_directories_for_symbols().
353         r121874 added new (exclude) directory (inspector/styles/variables), however it didn't
354         update one of two assertions.
355
356         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
357
358 2012-07-04  Balazs Ankes  <bank@inf.u-szeged.hu>
359
360         webkit-patch should add reviewer if "Reviewed by NOBODY ..." is missing
361         https://bugs.webkit.org/show_bug.cgi?id=67935
362
363         Reviewed by Ryosuke Niwa.
364
365         * Scripts/webkitpy/common/checkout/changelog.py:
366         (ChangeLog.set_reviewer):
367         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
368         (test_set_reviewer):
369         (test_set_short_description_and_bug_url):
370
371 2012-07-04  Sheriff Bot  <webkit.review.bot@gmail.com>
372
373         Unreviewed, rolling out r121862.
374         http://trac.webkit.org/changeset/121862
375         https://bugs.webkit.org/show_bug.cgi?id=90569
376
377         It broke built-product-archive for GTK, Qt and EFL (Requested
378         by svillar on #webkit).
379
380         * BuildSlaveSupport/built-product-archive:
381         (archiveBuiltProduct):
382
383 2012-07-04  Sergio Villar Senin  <svillar@igalia.com>
384
385         Hidden dirs are not copied when creating the built product archive
386         https://bugs.webkit.org/show_bug.cgi?id=90559
387
388         Reviewed by Carlos Garcia Campos.
389
390         Hidden dirs are not copied inside the built product archive because
391         the recursive copy command was ignoring them. We need those hidden
392         directories for the GTK WebKit2 testing bot to work fine.
393
394         * BuildSlaveSupport/built-product-archive:
395         (archiveBuiltProduct): removed the trailing "*" from the copy
396         command.
397
398 2012-07-04  John Mellor  <johnme@chromium.org>
399
400         Text Autosizing: Add compile flag and runtime setting
401         https://bugs.webkit.org/show_bug.cgi?id=87394
402
403         This patch renames Font Boosting to Text Autosizing.
404
405         Reviewed by Adam Barth.
406
407         * Scripts/webkitperl/FeatureList.pm:
408         * qmake/mkspecs/features/features.pri:
409
410 2012-07-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
411
412         [Qt] Replace deprecated IN_PWD with PWD
413
414         Reviewed by Csaba Osztrogonác.
415
416         * qmake/mkspecs/features/default_pre.prf:
417
418 2012-07-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
419
420         [Qt] Don't try to build WTR twice
421
422         Reviewed by Csaba Osztrogonác.
423
424         * Tools.pro:
425
426 2012-07-04  Sheriff Bot  <webkit.review.bot@gmail.com>
427
428         Unreviewed, rolling out r121821.
429         http://trac.webkit.org/changeset/121821
430         https://bugs.webkit.org/show_bug.cgi?id=90551
431
432         This patch did not receive a high-quality review and has a
433         number of errors (Requested by abarth on #webkit).
434
435         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
436         (Build.results_url):
437         * Scripts/webkitpy/common/net/buildbot/chromiumbuildbot.py:
438         (ChromiumBuilder.accumulated_results_url):
439         * Scripts/webkitpy/layout_tests/port/builders.py:
440         * Scripts/webkitpy/tool/commands/rebaseline.py:
441         (AbstractParallelRebaselineCommand.__init__):
442         (Rebaseline):
443         (Rebaseline._builder_to_pull_from):
444         (Rebaseline._tests_to_update):
445         (Rebaseline.execute):
446         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
447         (test_rebaseline.mock_builder_to_pull_from):
448         (test_rebaseline):
449
450 2012-07-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
451
452         [Qt] Get rid of un-needed QT += declarative for Qt 5
453
454         The declarative module has been renamed to quick1 in Qt 5, and the
455         engine-only module for Qt 5 is named 'qml'. For many of the instances
456         we could just remove 'declarative', since the project file was only
457         used for Qt5/WebKit2 builds. In the other cases the module was wrapped
458         in a haveQt(4) scope.
459
460         Reviewed by Csaba Osztrogonác.
461
462         * MiniBrowser/qt/MiniBrowser.pro:
463         * WebKitTestRunner/InjectedBundle/Target.pri:
464         * WebKitTestRunner/Target.pri:
465
466 2012-07-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
467
468         [Qt] Remove a few un-needed load(features) after r121777
469
470         The features are computed by configure.pro and cached in .qmake.cache.
471
472         Reviewed by Jocelyn Turcotte..
473
474         * MiniBrowser/qt/raw/Target.pri:
475
476 2012-07-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
477
478         [Qt] Use haveQtModule() in project files instead of manual checks
479
480         Reviewed by Jocelyn Turcotte.
481
482         * qmake/mkspecs/features/default_pre.prf:
483         * qmake/mkspecs/features/features.prf:
484
485 2012-07-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
486
487         [Qt] Fix haveQtModule() check
488
489         Without {} brackets we were pulling out the value of the module.name variable.
490
491         Reviewed by Jocelyn Turcotte.
492
493         * qmake/mkspecs/features/functions.prf:
494
495 2012-07-04  Sergio Villar Senin  <svillar@igalia.com>
496
497         [WK2] [GTK] WK2 testing bot unable to find the InjectedBundle library
498         https://bugs.webkit.org/show_bug.cgi?id=90541
499
500         Reviewed by Carlos Garcia Campos.
501
502         We must add the contents of WebKitBuild/*/Libraries/ directory to
503         the file to be uploaded to b.w.o to successfully run WebKit2 tests
504         in the WebKit2 testing bot.
505
506         * BuildSlaveSupport/built-product-archive:
507         (archiveBuiltProduct): appended Libraries to the list of
508         neededDirectories.
509
510 2012-07-04  Sergio Villar Senin  <svillar@igalia.com>
511
512         [GTK] WebKit2 testing bot unable to find the build directory
513         https://bugs.webkit.org/show_bug.cgi?id=90523
514
515         Reviewed by Carlos Garcia Campos.
516
517         The presence of GNUmakefile was used to determine whether a given
518         directory was a valid build directory or not. That's not correct for
519         testing bots as we export only the minimum required stuff to run the
520         tests.
521
522         * gtk/common.py:
523         (get_build_path.is_valid_build_directory): added an extra check.:
524
525 2012-07-04  Joe Thomas  <joethomas@motorola.com>
526
527         Adding myself to committers list.
528
529         Unreviewed.
530
531         * Scripts/webkitpy/common/config/committers.py:
532
533 2012-07-03  Christophe Dumez  <christophe.dumez@intel.com>
534
535         [WK2][EFL] Ewk_View should report the load progress
536         https://bugs.webkit.org/show_bug.cgi?id=90457
537
538         Reviewed by Kenneth Rohde Christiansen.
539
540         Update EFL MiniBrowser so that it listens for the
541         "load,progress" on the Ewk_View and updates its
542         window title accordingly.
543
544         * MiniBrowser/efl/main.c:
545         (title_set):
546         (on_title_changed):
547         (on_progress):
548         (browserCreate):
549
550 2012-07-03  Dirk Pranke  <dpranke@chromium.org>
551
552         disable failing mock-chromium-mac python unit tests
553
554         Unreviewed, build fix.
555
556         Disable these tests until I can figure out why they're failing
557         on the bots.
558
559         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
560         (PortTest.disabled_test_chromium_mac_lion):
561         (PortTest.disabled_test_chromium_mac_lion_in_test_shell_mode):
562
563 2012-07-03  Ian Vollick  <vollick@chromium.org>
564
565         Add vollick to the list of committers
566         https://bugs.webkit.org/show_bug.cgi?id=90512
567
568         Reviewed by Adrienne Walker.
569
570         * Scripts/webkitpy/common/config/committers.py:
571
572 2012-07-03  Ojan Vafai  <ojan@chromium.org>
573
574         Improve webkit-patch rebaseline to work for more cases
575         https://bugs.webkit.org/show_bug.cgi?id=90504
576
577         Reviewed by Dirk Pranke.
578
579         -Makes it work for the build.chromium.org bots.
580         -Lets you rebaseline all builders instead of just one.
581         -Lets you pass in the builder or test to rebaseline.
582
583         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
584         (Build.results_url):
585         Make this work for build.chromium.org builders as well.
586
587         * Scripts/webkitpy/common/net/buildbot/chromiumbuildbot.py:
588         (ChromiumBuilder):
589         (ChromiumBuilder.latest_cached_build):
590         ChromiumBuilder doesn't support large parts of the Builder interface.
591         This provides the bare minimum for this patch to work. Eventually,
592         we should create a single interface that can be supported for both
593         sets of buildbots.
594
595         * Scripts/webkitpy/layout_tests/port/builders.py:
596         (builder_path_from_name):
597         (all_builder_names):
598         memoizing here is incorrect because the test override _exact_matches,
599         so these can return different values. In either case, I'm pretty sure these
600         are not remotely hot codepaths.
601
602         * Scripts/webkitpy/tool/commands/rebaseline.py:
603         (AbstractParallelRebaselineCommand.__init__):
604         (Rebaseline):
605         (Rebaseline.__init__):
606         (Rebaseline._builders_to_pull_from):
607         (Rebaseline._builder_with_name):
608         (Rebaseline._tests_to_update):
609         (Rebaseline.execute):
610         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
611         (test_rebaseline.mock_builders_to_pull_from):
612         (test_rebaseline):
613         (test_rebaseline_command_line_flags):
614         (test_rebaseline_multiple_builders):
615         (test_rebaseline_multiple_builders.mock_builders_to_pull_from):
616         (test_rebaseline_multiple_builders.mock_tests_to_update):
617
618 2012-07-03  Christophe Dumez  <christophe.dumez@intel.com>
619
620         [EFL] Enable CSS variables support at compile time
621         https://bugs.webkit.org/show_bug.cgi?id=90448
622
623         Reviewed by Kenneth Rohde Christiansen.
624
625         Turn on CSS_VARIABLES flag by default on EFL port.
626
627         * Scripts/webkitperl/FeatureList.pm:
628
629 2012-07-03  Dirk Pranke  <dpranke@chromium.org>
630
631         nrwt: clean up exception handling and make sure we log some more failures
632         https://bugs.webkit.org/show_bug.cgi?id=90503
633
634         Reviewed by Ojan Vafai.
635
636         There were several places where exceptions weren't getting
637         logged, most notably if you passed a bad value to --platform.
638         This change tests that and cleans things up a bit; more cleanup
639         will be possible when we rework the manager_worker_broker code.
640
641         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
642         (_BrokerConnection.raise_exception):
643         (_InlineWorkerConnection.raise_exception):
644         * Scripts/webkitpy/layout_tests/controllers/worker.py:
645         (Worker.run):
646         (Worker.kill_driver):
647         * Scripts/webkitpy/layout_tests/port/factory.py:
648         (PortFactory.get):
649         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
650         (run):
651         (main):
652         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
653         (MainTest.test_unsupported_platfrom):
654
655 2012-07-03  Dirk Pranke  <dpranke@chromium.org>
656
657         nrwt: fix mock port
658         https://bugs.webkit.org/show_bug.cgi?id=90500
659
660         Reviewed by Ojan Vafai.
661
662         The MockDRT code was never updated when we switched the chromium
663         ports to using "drt mode" by default. This change updates that
664         code, fixes a typo in the chromium port that went undetected
665         (default_test_timeout_ms -> default_timeout_ms), and adds tests
666         that actually exercise some of the mock ports. These tests are
667         useful in that they will exercise the port-specific code in an
668         end-to-end-manner, but they are a bit slow for some reason (>1s
669         each) that I need to look into.
670
671         * Scripts/webkitpy/layout_tests/port/chromium.py:
672         (ChromiumDriver.stop):
673         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
674         (ChromiumAndroidPort.default_timeout_ms):
675         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
676         (TestChromiumAndroidPort.test_default_timeout_ms):
677         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
678         (main):
679         (parse_options):
680         (MockTestShell):
681         (MockTestShell.output_for_test):
682         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
683         (MockDRTTest.assertTest):
684         (MockTestShellTest):
685         (MockTestShellTest.make_drt):
686         (MockTestShellTest.test_test_shell_parse_options):
687         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
688         (PortTest):
689         (PortTest.assert_mock_port_works):
690         (PortTest.test_chromium_mac_lion):
691         (PortTest.test_chromium_mac_lion_in_test_shell_mode):
692         (PortTest.test_qt_linux):
693         (PortTest.test_mac_lion):
694
695 2012-07-03  Dirk Pranke  <dpranke@chromium.org>
696
697         nrwt: make the worker class stand alone with a cleaner interface
698         https://bugs.webkit.org/show_bug.cgi?id=90409
699
700         Reviewed by Ojan Vafai.
701
702         Currently the Worker class derives from AbstractWorker, which is
703         kind of crufty and awkward; it would be better if we did not
704         rely on shared state.
705
706         This change changes that so that Worker derives from object, and
707         exposes the following interface:
708           __init__() - called in the manager process
709           safe_init() - called in the worker process to initialize
710             unpicklable state
711           handle() - a single routine to handle all messages
712           cleanup() - called so the worker can clean up
713
714         Also, all of the "administrative" messages that are handled by
715         the worker (notification of start/stop/etc.) move into
716         manager_worker_broker - this reduces worker.py to just handling
717         the mechanics of actually running each test.
718
719         For the moment, we do this by creating Yet Another wrapper/proxy
720         class in manager_worker_broker, but this will get simpler
721         shortly when the rest of m_w_b is cleaned up.
722
723         With this change worker is now in its new form but there will be
724         a follow-on change that cleans up some names and other minor
725         things.
726
727         This change is again mostly just moving things around and should
728         be covered by the (updated) existing tests.
729
730         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
731         (get):
732         (AbstractWorker.__init__):
733         (AbstractWorker.run):
734         (AbstractWorker):
735         (AbstractWorker.handle_stop):
736         (AbstractWorker.handle_test_list):
737         (AbstractWorker.yield_to_broker):
738         (AbstractWorker.post_message):
739         (_WorkerConnection.__init__):
740         (_Process.run):
741         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py:
742         (_TestWorker):
743         (_TestWorker.__init__):
744         (_TestWorker.name):
745         (_TestWorker.cleanup):
746         (_TestWorker.handle):
747         (_TestWorker.safe_init):
748         (_TestWorker.stop):
749         (_TestsMixin.handle_finished_test):
750         (_TestsMixin.setUp):
751         (_TestsMixin.test_cancel):
752         (_TestsMixin.test_done):
753         * Scripts/webkitpy/layout_tests/controllers/worker.py:
754         (Worker):
755         (Worker.__init__):
756         (Worker.safe_init):
757         (Worker.handle):
758
759 2012-07-03  Dirk Pranke  <dpranke@chromium.org>
760
761         nrwt: moving child process logging code into manager_worker_broker
762         https://bugs.webkit.org/show_bug.cgi?id=90408
763
764         Reviewed by Ojan Vafai.
765
766         Users of manager_worker_broker should not have to be aware of
767         whether they're in the same process or different processes and
768         configure logging themselves; mwb should hide this complexity.
769         We can't quite do this completely/correctly yet, since the
770         manager expects to get a list of messages to log, but this
771         change fixes the worker side of it, at least.
772
773         This is just moving code around, there is no new functionality
774         and this should be covered by existing tests.
775
776         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
777         (AbstractWorker.__init__):
778         (_WorkerConnection.__init__):
779         (_WorkerConnection.post_message):
780         (_WorkerConnection):
781         (_WorkerConnection.set_up_logging):
782         (_WorkerConnection.clean_up_logging):
783         (_InlineWorkerConnection.run):
784         (_MultiProcessWorkerConnection.run):
785         (_WorkerLogHandler):
786         (_WorkerLogHandler.__init__):
787         (_WorkerLogHandler.emit):
788         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py:
789         (_TestWorker.run):
790         (_TestsMixin.handle_done):
791         * Scripts/webkitpy/layout_tests/controllers/worker.py:
792         (Worker.__init__):
793         (Worker.run):
794         (Worker._run_test):
795         (Worker.cleanup):
796         (Worker.run_single_test):
797
798 2012-07-03  Tony Chang  <tony@chromium.org>
799
800         [chromium] Don't archive build files generated by VS2010
801         https://bugs.webkit.org/show_bug.cgi?id=90493
802
803         Reviewed by Ojan Vafai.
804
805         In particular, the pch (pre compiled header) files are gigantic, making
806         release.zip almost a 1gb in size.
807
808         * BuildSlaveSupport/built-product-archive:
809         (archiveBuiltProduct):
810
811 2012-07-03  Tony Chang  <tony@chromium.org>
812
813         [chromium] Unreviewed, update .gitignore to handle VS2010 files.
814
815         * .gitignore:
816         * DumpRenderTree/DumpRenderTree.gyp/.gitignore:
817         * TestWebKitAPI/TestWebKitAPI.gyp/.gitignore:
818
819 2012-07-03  Ojan Vafai  <ojan@chromium.org>
820
821         Have webkit-patch rebaseline use rebaseline-test-internal
822         https://bugs.webkit.org/show_bug.cgi?id=90491
823
824         Reviewed by Dirk Pranke.
825
826         This lets it handle new files, reduces duplicate code and lays the
827         groundwork for making rebaseline have a richer interface.
828
829         * Scripts/webkitpy/common/net/buildbot/buildbot_mock.py:
830         (MockBuild):
831         (MockBuild.__init__):
832         (MockBuilder.build):
833         * Scripts/webkitpy/tool/commands/rebaseline.py:
834         (AbstractParallelRebaselineCommand._files_to_add):
835         (Rebaseline):
836         (Rebaseline._builder_to_pull_from):
837         (Rebaseline._tests_to_update):
838         (Rebaseline.execute):
839         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
840         (test_overrides_are_included_correctly):
841         (test_rebaseline):
842         (test_rebaseline.mock_builder_to_pull_from):
843         (test_rebaseline.mock_tests_to_update):
844
845 2012-07-03  Ojan Vafai  <ojan@chromium.org>
846
847         Rename rebaseline-test to rebaseline-test-internal
848         https://bugs.webkit.org/show_bug.cgi?id=90485
849
850         Reviewed by Adam Barth.
851
852         It's now only used by other rebaseline commands. It's still useful to leave it
853         as it's own command to aid in debugging when something goes wrong. In a followup
854         patch, I'll make webkit-patch rebaseline cover any use-cases that rebaseline-test
855         might have covered.
856
857         We no longer need the --print-scm-changes option since the only caller always passes
858         that option in.
859
860         Also, make all the arguments command-line flags instead. Simplifies the code
861         a bit in my opinion.
862
863         * Scripts/webkitpy/tool/commands/rebaseline.py:
864         (RebaselineTest):
865         (RebaselineTest.__init__):
866         (RebaselineTest.execute):
867         (AbstractParallelRebaselineCommand._rebaseline_commands):
868         (RebaselineExpectations.execute):
869         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
870         (test_rebaseline_all):
871
872 2012-07-03  Balazs Kelemen  <kbalazs@webkit.org>
873
874         [Qt][WTR] Get rid of using DumpRenderTreeSupportQt
875         https://bugs.webkit.org/show_bug.cgi?id=90262
876
877         Unreviewed build fix.
878
879         Reverting r121550 manually because it caused build
880         break on ARM and Mac. It turned out that the Qt build
881         system is too broken for this change now, first we have
882         to address two issues:
883             * symbol hiding was forbidden accidentally on x86 Linux
884             * Qt lacks a separate library for test support
885
886         * Tools.pro:
887         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
888         (WTR::InjectedBundlePage::resetAfterTest):
889         (WTR::InjectedBundlePage::didClearWindowForFrame):
890         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
891
892 2012-07-03  Tony Chang  <tony@chromium.org>
893
894         [chromium] Handle trailing backslash on %VSINSTALLDIR%
895         https://bugs.webkit.org/show_bug.cgi?id=90410
896
897         Reviewed by Ojan Vafai.
898
899         * Scripts/webkitdirs.pm:
900         (buildChromiumVisualStudioProject): Handle VSINSTALLDIR ending in a backslash, the backslash escapes the closing double quote.
901
902 2012-07-02  Ojan Vafai  <ojan@chromium.org>
903
904         Make the skia_test_expectations.txt file optional.
905         https://bugs.webkit.org/show_bug.cgi?id=90400
906
907         Reviewed by Dirk Pranke.
908
909         It used to be optional. This regressed at some point. It's important that it be
910         optional so that webkit-patch commands work in a pure-webkit checkout for chromium bots.
911         Specifically, this was breaking webkit-patch rebaseline-test when it would go to update
912         TestExpectations.
913
914         * Scripts/webkitpy/layout_tests/port/chromium.py:
915         (ChromiumPort.expectations_files):
916         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
917         (ChromiumDriverTest.test_expectations_dict):
918
919 2012-07-03  Raphael Kubo da Costa  <rakuco@webkit.org>
920
921         [jhbuild][EFL] Bump libffi dependency.
922         https://bugs.webkit.org/show_bug.cgi?id=85373
923
924         Reviewed by Gustavo Noronha Silva.
925
926         Update libffi to 3.0.11, which fixes the build on some platforms
927         where 3.0.10 had problems.
928
929         * efl/jhbuild.modules:
930
931 2012-07-03  George Staikos  <staikos@webkit.org>
932
933         [BlackBerry] Enable microdata support for BlackBerry.
934         https://bugs.webkit.org/show_bug.cgi?id=90429
935
936         Reviewed by Rob Buis.
937
938         * Scripts/webkitperl/FeatureList.pm:
939
940 2012-07-03  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
941
942         [Qt] Make use of .qmake.cache for caching features
943
944         Instead of loading() features from the files that need them (and re-running
945         a bunch of checks), we now run feature detection as part of configure.pro,
946         and have build-webkit write the computed feature-defines and CONFIG to
947         .qmake.cache, which is then loaded by qmake _before_ even defaults_pre
948         when building WebKit.pro.
949
950         At some point we'll be able to selectivly prevent running of config tests
951         in configure.pro, which means we don't need a separate code-path for
952         the build-webkit --help case.
953
954         We should also move the code in build-webkit that now uses .webkit.config
955         to detect clean builds, to use .qmake.cache, since we now store the same
956         thing there.
957
958         Original patch by Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
959
960         Reviewed by Tor Arne Vestbø.
961
962         * DumpRenderTree/qt/DumpRenderTree.pro:
963         * QtTestBrowser/QtTestBrowser.pro:
964         * Scripts/webkitdirs.pm:
965         (qtFeatureDefaults):
966         (buildQMakeProjects):
967         * Tools.pro:
968         * WebKitTestRunner/InjectedBundle/DerivedSources.pri:
969         * WebKitTestRunner/InjectedBundle/Target.pri:
970         * WebKitTestRunner/Target.pri:
971         * qmake/.qmake.conf: Added.
972         * qmake/configure.pro:
973         * qmake/mkspecs/features/default_post.prf:
974         * qmake/mkspecs/features/default_pre.prf:
975         * qmake/mkspecs/features/features.prf:
976
977 2012-07-03  Nico Weber  <thakis@chromium.org>
978
979         -Wunused-private-field cleanup, linux edition
980         https://bugs.webkit.org/show_bug.cgi?id=90463
981
982         Reviewed by Stephen White.
983
984         * DumpRenderTree/chromium/DRTDevToolsAgent.h:
985         * DumpRenderTree/chromium/GamepadController.h:
986         (GamepadController):
987         * DumpRenderTree/chromium/NotificationPresenter.h:
988         (NotificationPresenter::NotificationPresenter):
989         (NotificationPresenter):
990
991 2012-07-03  Sheriff Bot  <webkit.review.bot@gmail.com>
992
993         Unreviewed, rolling out r121766.
994         http://trac.webkit.org/changeset/121766
995         https://bugs.webkit.org/show_bug.cgi?id=90465
996
997         It caused flakey build errors on the bots (Requested by Ossy
998         on #webkit).
999
1000         * DumpRenderTree/qt/DumpRenderTree.pro:
1001         * QtTestBrowser/QtTestBrowser.pro:
1002         * Scripts/webkitdirs.pm:
1003         (qtFeatureDefaults):
1004         (buildQMakeProjects):
1005         * Tools.pro:
1006         * WebKitTestRunner/InjectedBundle/DerivedSources.pri:
1007         * WebKitTestRunner/InjectedBundle/Target.pri:
1008         * WebKitTestRunner/Target.pri:
1009         * qmake/.qmake.conf: Removed.
1010         * qmake/configure.pro:
1011         * qmake/mkspecs/features/default_post.prf:
1012         * qmake/mkspecs/features/default_pre.prf:
1013         * qmake/mkspecs/features/features.prf:
1014
1015 2012-07-03  George Staikos  <staikos@webkit.org>
1016
1017         [BlackBerry] Enable Custom Scheme Handlers for BlackBerry.
1018         https://bugs.webkit.org/show_bug.cgi?id=90422
1019
1020         Reviewed by Rob Buis.
1021
1022         * Scripts/webkitperl/FeatureList.pm:
1023
1024 2012-07-03  George Staikos  <staikos@webkit.org>
1025
1026         [BlackBerry] Enable RegisterProtocolHandler for BlackBerry.
1027         https://bugs.webkit.org/show_bug.cgi?id=90422
1028
1029         Reviewed by Rob Buis.
1030
1031         * Scripts/webkitperl/FeatureList.pm:
1032
1033 2012-07-03  Thiago Marcos P. Santos  <thiago.santos@intel.com>
1034
1035         [EFL] Add run API tests as a step on the build bots
1036         https://bugs.webkit.org/show_bug.cgi?id=87252
1037
1038         Reviewed by Daniel Bates.
1039
1040         Run EFL API test suite on the build bots.
1041
1042         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
1043         (RunEflAPITests):
1044         (TestFactory.__init__):
1045         (BuildAndTestFactory.__init__):
1046         * Scripts/run-efl-tests: Added.
1047
1048 2012-07-03  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1049
1050         [Qt] Make use of .qmake.cache for caching features
1051
1052         Instead of loading() features from the files that need them (and re-running
1053         a bunch of checks), we now run feature detection as part of configure.pro,
1054         and have build-webkit write the computed feature-defines and CONFIG to
1055         .qmake.cache, which is then loaded by qmake _before_ even defaults_pre
1056         when building WebKit.pro.
1057
1058         At some point we'll be able to selectivly prevent running of config tests
1059         in configure.pro, which means we don't need a separate code-path for
1060         the build-webkit --help case.
1061
1062         We should also move the code in build-webkit that now uses .webkit.config
1063         to detect clean builds, to use .qmake.cache, since we now store the same
1064         thing there.
1065
1066         Original patch by Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
1067
1068         Reviewed by Tor Arne Vestbø.
1069
1070         * DumpRenderTree/qt/DumpRenderTree.pro:
1071         * QtTestBrowser/QtTestBrowser.pro:
1072         * Scripts/webkitdirs.pm:
1073         (qtFeatureDefaults):
1074         (buildQMakeProjects):
1075         * Tools.pro:
1076         * WebKitTestRunner/InjectedBundle/DerivedSources.pri:
1077         * WebKitTestRunner/InjectedBundle/Target.pri:
1078         * WebKitTestRunner/Target.pri:
1079         * qmake/.qmake.conf: Added.
1080         * qmake/configure.pro:
1081         * qmake/mkspecs/features/default_post.prf:
1082         * qmake/mkspecs/features/default_pre.prf:
1083         * qmake/mkspecs/features/features.prf:
1084
1085 2012-07-03  Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>  Joel Dillon <joel.dillon@codethink.co.uk>
1086
1087         [Qt][Win] Fix broken QtWebKit5.lib linking
1088         https://bugs.webkit.org/show_bug.cgi?id=88321
1089
1090         Reviewed by Kenneth Rohde Christiansen.
1091
1092         On windows the import/export definition is part of the symbol's signature.
1093         Define STATICALLY_LINKED_WITH_$$library for each dependend module
1094         that is being linked statically to make sure that they can be linked together
1095         later on.
1096
1097         Also do not compile Assertions.cpp in DumpRenderTree anymore since all the
1098         used symbols are exported and it would cause a duplicate symbols error.
1099
1100         * DumpRenderTree/qt/DumpRenderTree.pro:
1101         * qmake/mkspecs/features/default_post.prf:
1102
1103 2012-07-03  Szilard Ledan  <szledan@inf.u-szeged.hu>
1104
1105         master.cfg cleanup, pass BuildStep instances instead of BuildStep subclasses
1106         https://bugs.webkit.org/show_bug.cgi?id=89001
1107
1108         Reviewed by Csaba Osztrogonác.
1109
1110         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
1111         (UploadBuiltProduct.__init__):
1112         (UploadTestResults.__init__):
1113         (ExtractTestResults.__init__):
1114         (Factory.__init__):
1115         (BuildFactory.__init__):
1116         (TestFactory.__init__):
1117         (BuildAndTestFactory.__init__):
1118         (BuildAndPerfTestFactory.__init__):
1119         (BuildAndPerfTestWebKit2Factory.__init__):
1120         (DownloadAndPerfTestFactory.__init__):
1121
1122 2012-07-03  Ádám Kallai  <kadam@inf.u-szeged.hu>
1123
1124         Update unit test after r121724.
1125         https://bugs.webkit.org/show_bug.cgi?id=90444
1126
1127         Reviewed by Csaba Osztrogonác.
1128
1129         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
1130         (GardeningServerTest.test_rebaseline_new_port):
1131
1132 2012-07-03  Sheriff Bot  <webkit.review.bot@gmail.com>
1133
1134         Unreviewed, rolling out r121749.
1135         http://trac.webkit.org/changeset/121749
1136         https://bugs.webkit.org/show_bug.cgi?id=90441
1137
1138         Tests crash on the paralel NRWT Qt bot (Requested by Ossy on
1139         #webkit).
1140
1141         * Scripts/webkitpy/layout_tests/port/qt.py:
1142         (QtPort._build_driver):
1143         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
1144         (XvfbDriver._start):
1145         (XvfbDriver._start.x_filter):
1146         (XvfbDriver.stop):
1147
1148 2012-07-03  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
1149
1150         [NRWT] XvfbDriver should choose the next free display
1151         https://bugs.webkit.org/show_bug.cgi?id=88414
1152
1153         Reviewed by Dirk Pranke.
1154
1155         * Scripts/webkitpy/layout_tests/port/qt.py:
1156         (QtPort._driver_class):
1157         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
1158         (XvfbDriver.__init__):
1159         (XvfbDriver._start):
1160         (XvfbDriver._start.next_free_id):
1161         (XvfbDriver.stop):
1162
1163 2012-07-03  Christophe Dumez  <christophe.dumez@intel.com>
1164
1165         [EFL] Enable MICRODATA support
1166         https://bugs.webkit.org/show_bug.cgi?id=90377
1167
1168         Reviewed by Kenneth Rohde Christiansen.
1169
1170         Turn on MICRODATA flag for EFL port by default.
1171
1172         * Scripts/webkitperl/FeatureList.pm:
1173
1174 2012-07-02  George Staikos  <staikos@webkit.org>
1175
1176         [BlackBerry] Enable scoped style for BlackBerry.
1177         https://bugs.webkit.org/show_bug.cgi?id=90418
1178
1179         Reviewed by Rob Buis.
1180
1181         * Scripts/webkitperl/FeatureList.pm:
1182
1183 2012-07-02  Xiaobo Wang  <xbwang@torchmobile.com.cn>
1184
1185         [BlackBerry] Use PUBLIC_BUILD to enable/disable DRT
1186         https://bugs.webkit.org/show_bug.cgi?id=90271
1187
1188         Reviewed by George Staikos.
1189
1190         RIM PR #154707
1191
1192         Currently DRT code will be compiled only if ENABLE_DRT is set, and it's not
1193         defined by default.
1194         We should enable DRT by default unless PUBLIC_BUILD is set. In this way we don't
1195         need to rebuild webkit before running DRT.
1196
1197         * Scripts/webkitdirs.pm:
1198         (blackberryCMakeArguments):
1199
1200 2012-07-02  Ojan Vafai  <ojan@chromium.org>
1201
1202         webkit-patch rebaseline-expectations should share code with rebaseline-all
1203         https://bugs.webkit.org/show_bug.cgi?id=90413
1204
1205         Reviewed by Dirk Pranke.
1206
1207         Make them share code. In addition to reducing code duplication this makes
1208         rebaseline-expectations considerably faster by rebaselining in parallel.
1209
1210         * Scripts/webkitpy/tool/commands/rebaseline.py:
1211         (AbstractParallelRebaselineCommand):
1212         (AbstractParallelRebaselineCommand._run_webkit_patch):
1213         (AbstractParallelRebaselineCommand._rebaseline):
1214         (RebaselineJson):
1215         (RebaselineJson.execute):
1216         (RebaselineExpectations):
1217         (RebaselineExpectations._update_expectations_file):
1218         (RebaselineExpectations._tests_to_rebaseline):
1219         (RebaselineExpectations._add_tests_to_rebaseline_for_port):
1220         (RebaselineExpectations.execute):
1221         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
1222         (test_rebaseline_all):
1223         (test_rebaseline_expectations.run_in_parallel):
1224         (test_rebaseline_expectations):
1225
1226 2012-07-02  Xiaobo Wang  <xbwang@torchmobile.com.cn>
1227
1228         [BlackBerry] Update DumpRenderTree to have it work interactively in parallel
1229         https://bugs.webkit.org/show_bug.cgi?id=88326
1230
1231         Reviewed by Rob Buis.
1232
1233         1. Test name is sent to torch-launcher via PPS message(from host machine).
1234            So we get test list by monitoring and receiving PPS message instead of
1235            reading file index.drt.
1236         2. Torch-launcher create a <test file>.done file when it finished a test.
1237         3. We don't need to search for Ref-tests in DumpRenderTree.cpp any more. NRWT
1238            will get them for us.
1239
1240         * DumpRenderTree/blackberry/DumpRenderTree.cpp:
1241         (BlackBerry::WebKit::DumpRenderTree::DumpRenderTree):
1242         (BlackBerry::WebKit::DumpRenderTree::doneDrt):
1243         (BlackBerry::WebKit::DumpRenderTree::runRemainingTests):
1244         (BlackBerry::WebKit::DumpRenderTree::ensurePPS):
1245         (WebKit):
1246         (BlackBerry::WebKit::DumpRenderTree::handlePPSData):
1247         (BlackBerry::WebKit::DumpRenderTree::waitForTest):
1248         (BlackBerry::WebKit::DumpRenderTree::runTests):
1249         (BlackBerry::WebKit::DumpRenderTree::dump):
1250         * DumpRenderTree/blackberry/DumpRenderTreeBlackBerry.h:
1251         (DumpRenderTree):
1252
1253 2012-07-02  Adam Barth  <abarth@webkit.org>
1254
1255         Remove flashplugin-installer from the EWS image because it causes some tests to crash
1256         https://bugs.webkit.org/show_bug.cgi?id=90403
1257
1258         Reviewed by Tony Chang.
1259
1260         * EWSTools/cold-boot.sh:
1261
1262 2012-07-02  Leandro Gracia Gil  <leandrogracia@chromium.org>
1263
1264         [Chromium] Implement a Layout Test for editing/SurroundingText
1265         https://bugs.webkit.org/show_bug.cgi?id=82461
1266
1267         Reviewed by Ryosuke Niwa.
1268
1269         Add a new method to the layout test controller in order to retrieve the
1270         text surrounding a provided element.
1271
1272         * DumpRenderTree/chromium/LayoutTestController.cpp:
1273         (LayoutTestController::LayoutTestController):
1274         (LayoutTestController::textSurroundingElement):
1275         * DumpRenderTree/chromium/LayoutTestController.h:
1276         (LayoutTestController):
1277
1278 2012-07-02  Ojan Vafai  <ojan@chromium.org>
1279
1280         Delete unused rebaseline method in gardeningserver.py
1281         https://bugs.webkit.org/show_bug.cgi?id=90396
1282
1283         Reviewed by Eric Seidel.
1284
1285         As best I can tell, the only usage was in this unittest.
1286
1287         * Scripts/webkitpy/tool/servers/gardeningserver.py:
1288         (GardeningHTTPRequestHandler.updateexpectations):
1289         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
1290         (GardeningServerTest.test_rebaseline_new_port):
1291
1292 2012-07-02  Csaba Osztrogonác  <ossy@webkit.org>
1293
1294         REGRESSION(r121497): It switched off and broke many unittests
1295         https://bugs.webkit.org/show_bug.cgi?id=90371
1296
1297         Reviewed by Adam Barth.
1298
1299         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
1300         (ChromiumMacPortTest): Use snowleopard as os_version instead of leopard, because it isn't supported anymore.
1301         * Scripts/webkitpy/layout_tests/port/chromium_port_testcase.py:
1302         (ChromiumPortTestCase.test_all_test_configurations): Remove leopard testcases, because it isn't supported anymore.
1303         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
1304         (PortTestCase): Inherit class PortTestCase from unittest.TestCase instead of object.
1305
1306 2012-07-02  Ojan Vafai  <ojan@chromium.org>
1307
1308         Move rebaseline-all command from the gardening-server down into webkit-patch
1309         https://bugs.webkit.org/show_bug.cgi?id=90395
1310
1311         Reviewed by Adam Barth.
1312
1313         This is just moving code. It it in preparation for making rebaseline-expectations
1314         use the same code in order to get the parallelism benefits and reduces the amount
1315         of code we have for doing rebaselines.
1316
1317         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
1318         (CheckoutTest.test_apply_patch):
1319         Updated due to the change to executive_mock.
1320         * Scripts/webkitpy/common/system/executive_mock.py:
1321         (MockExecutive.run_command):
1322         Update to print out the input passed to stdin.
1323         * Scripts/webkitpy/tool/commands/download_unittest.py:
1324         Updated due to executive_mock change.
1325         * Scripts/webkitpy/tool/commands/rebaseline.py:
1326         (RebaselineAll):
1327         (RebaselineAll._run_webkit_patch):
1328         (RebaselineAll._builders_to_fetch_from):
1329         (RebaselineAll._rebaseline_commands):
1330         (RebaselineAll._files_to_add):
1331         (RebaselineAll._optimize_baselines):
1332         (RebaselineAll._rebaseline):
1333         (RebaselineAll.execute):
1334         All this code is just copy-pasted except for mechanical changes
1335         (e.g. self.server.tool --> self._tool) and the reading in of the 
1336         JSON from stdin instead of the post body.
1337         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
1338         (test_rebaseline_all):
1339         Copied the test-case out of gardeningserver_unittest.py.
1340         * Scripts/webkitpy/tool/servers/gardeningserver.py:
1341         (GardeningHTTPRequestHandler):
1342         (GardeningHTTPRequestHandler.rebaseline):
1343         (GardeningHTTPRequestHandler.rebaselineall):
1344         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
1345         (GardeningServerTest.test_rebaselineall):
1346         (GardeningServerTest.test_rebaselineall.run_command):
1347
1348 2012-07-02  Ojan Vafai  <ojan@chromium.org>
1349
1350         Remove Leopard support from the flakiness dashboard
1351         https://bugs.webkit.org/show_bug.cgi?id=90390
1352
1353         Reviewed by Adam Barth.
1354
1355         * TestResultServer/static-dashboards/flakiness_dashboard.js:
1356         (nonChromiumPlatform):
1357         (chromiumPlatform):
1358         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
1359         (test):
1360         * TestResultServer/static-dashboards/run-embedded-unittests.html:
1361         * TestResultServer/static-dashboards/run-unittests.html:
1362
1363 2012-07-02  Ojan Vafai  <ojan@chromium.org>
1364
1365         Fix posting from garden-o-matic. This broke in moving away from jquery's ajax method
1366         in http://trac.webkit.org/changeset/121617.
1367         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/net.js:
1368
1369 2012-07-02  Carlos Garcia Campos  <cgarcia@igalia.com>
1370
1371         [GTK] Read fonts path when running layout tests from alternative fonts dir when main dir doesn't exist
1372         https://bugs.webkit.org/show_bug.cgi?id=89437
1373
1374         Reviewed by Martin Robinson.
1375
1376         If main fonts directory doesn't exist, try with an alternative
1377         fonts directory at build directory.
1378
1379         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1380         (getOutputDir):
1381         (getFontsPath):
1382         (initializeFonts):
1383         * WebKitTestRunner/InjectedBundle/gtk/ActivateFontsGtk.cpp:
1384         (WTR::getOutputDir):
1385         (WTR):
1386         (WTR::getFontsPath):
1387         (WTR::inititializeFontConfigSetting):
1388
1389 2012-07-02  Carlos Garcia Campos  <cgarcia@igalia.com>
1390
1391         [GTK] Don't run the tests with jhbuild wrapper if it's already running under jhbuild
1392         https://bugs.webkit.org/show_bug.cgi?id=89435
1393
1394         Reviewed by Martin Robinson.
1395
1396         * Scripts/new-run-webkit-tests: Don't run the tests with the
1397         jhbuild wrapper if there isn't a Dependencies directory inside the
1398         build directory.
1399
1400 2012-06-29  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
1401
1402         [Qt] Simplify detection of non-installed module file
1403
1404         Has the additional advantage that we do not rely on additional information.
1405
1406         Reviewed by Tor Arne Vestbø.
1407
1408         * qmake/qt_webkit.pri:
1409
1410 2012-07-01  Christophe Dumez  <christophe.dumez@intel.com>
1411
1412         [EFL] Add Gamepad support
1413         https://bugs.webkit.org/show_bug.cgi?id=90170
1414
1415         Reviewed by Kenneth Rohde Christiansen.
1416
1417         * Scripts/webkitperl/FeatureList.pm: Turn on GAMEPAD flag
1418         by default for EFL port.
1419         * efl/jhbuild.modules: Bump dependency versions of EFL libs
1420         since the latest Eeze is required for gamepad support.
1421
1422 2012-07-01  Adam Barth  <abarth@webkit.org>
1423
1424         Unreviewed.
1425
1426         Using the default start cylinder for fdisk causes a warning about the
1427         partition not starting on physical sector boundary.  The Ubuntu forums
1428         recommend using a number that's divisible by 8, which is what we do in
1429         this patch.
1430
1431         * EWSTools/cold-boot.sh:
1432
1433 2012-07-01  Adam Barth  <abarth@webkit.org>
1434
1435         Unreviewed.
1436
1437         It turns out we need to use Ubuntu 10.04 to get the right image results
1438         for chromium-linux. This patch updates our scripts to be compatible
1439         with Ubuntu 10.04.
1440
1441         * EWSTools/cold-boot.sh:
1442             - fdisk doesn't have p and 1 as default commands in 10.04.
1443         * EWSTools/start-queue.sh:
1444             - git doesn't understanding the -B argument in 10.04. We've been
1445               using this change locally on the EC2 bots for a while.
1446
1447 2012-07-01  Adam Barth  <abarth@webkit.org>
1448
1449         Unreviewed.
1450
1451         * EWSTools/boot.sh:
1452             - We need to start the screen in detached mode so that we can run
1453               it remotely via ssh.
1454
1455 2012-07-01  Adam Barth  <abarth@webkit.org>
1456
1457         Unreviewed.
1458
1459         * EWSTools/cold-boot.sh:
1460             - Turns out we need to sudo this command in order for it to
1461               actually dimiss the EULA screen. :)
1462
1463 2012-06-29  Adam Barth  <abarth@webkit.org>
1464
1465         Turns out we need zip too.
1466
1467         * EWSTools/cold-boot.sh:
1468
1469 2012-06-29  Adam Barth  <abarth@webkit.org>
1470
1471         Add a cold-boot.sh script for the EWS
1472         https://bugs.webkit.org/show_bug.cgi?id=90330
1473
1474         Unreviewed.
1475
1476         * EWSTools/cold-boot.sh: Added.
1477             - This script can take us from a cold GCE instance to a running EWS
1478               bot in one fell swoop.
1479         * EWSTools/start-queue.sh:
1480             - The if-block at the top of this script was causing trouble. I
1481               removed it from the bots a while ago. Now that we're using SVN to
1482               cold-boot the EWS bots, we need to make this change in the repo.
1483
1484 2012-06-29  Luiz Agostini  <luiz.agostini@nokia.com>
1485
1486         [Qt][WK2] Private non-QtQuick API
1487         https://bugs.webkit.org/show_bug.cgi?id=84532
1488
1489         MiniBrowserRaw is an usage example for the QRawWebView API.
1490         It is only an example and is not fully implemented.
1491
1492         Reviewed by Noam Rosenthal.
1493
1494         * MiniBrowser/qt/raw/DerivedSources.pri: Added.
1495         * MiniBrowser/qt/raw/MiniBrowserRaw.pro: Added.
1496         * MiniBrowser/qt/raw/Target.pri: Added.
1497         * MiniBrowser/qt/raw/View.cpp: Added.
1498         * MiniBrowser/qt/raw/View.h: Added.
1499         * Scripts/webkitpy/style/checker.py:
1500         * Tools.pro:
1501
1502 2012-06-29  Ojan Vafai  <ojan@chromium.org>
1503
1504         Remove leopard bots from garden-o-matic
1505         https://bugs.webkit.org/show_bug.cgi?id=90328
1506
1507         Reviewed by Adam Barth.
1508
1509         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
1510         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures_unittests.js:
1511         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
1512
1513 2012-06-29  Ojan Vafai  <ojan@chromium.org>
1514
1515         garden-o-matic broken: TypeError: 'undefined' is not an object (evaluating 'buildLocations[currentIndex].url')
1516         https://bugs.webkit.org/show_bug.cgi?id=90243
1517
1518         Reviewed by Dirk Pranke.
1519
1520         jQuery was trying to be too smart and parsing the jsonp as json because of it's content-type.
1521         Excise jQuery and just use XHR directly since it's easier to maintain something where we control it
1522         all.
1523
1524         Not really sure how to unittest this. I tested it all manually of course.
1525
1526         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/net.js:
1527         Made net.ajax a drop-in replacement for the features of $.ajax that we were using.
1528         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:
1529         Not related to this patch, but figured I'd update the failing test while I was at it.
1530
1531 2012-06-29  Yaron Friedman  <yfriedman@chromium.org>
1532
1533         Fix layout test runner for Android after https://bugs.webkit.org/show_bug.cgi?id=88134
1534         https://bugs.webkit.org/show_bug.cgi?id=90309
1535
1536         Reviewed by Adam Barth.
1537
1538         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
1539         (ChromiumAndroidPort.start_http_server):
1540
1541 2012-06-29  James Weatherall  <wez@chromium.org>
1542
1543         NPObjectWrapper may not address all window script object lifetime issues
1544         https://bugs.webkit.org/show_bug.cgi?id=85679
1545
1546         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.
1547
1548         Reviewed by Nate Chapin.
1549
1550         * DumpRenderTree/DumpRenderTree.gypi:
1551         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
1552         (PluginTest::NPN_GetProperty):
1553         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
1554         (PluginTest):
1555         * DumpRenderTree/TestNetscapePlugIn/Tests/LeakWindowScriptableObject.cpp: Added.
1556         (LeakWindowScriptableObject):
1557         (LeakWindowScriptableObject::LeakWindowScriptableObject):
1558         (LeakWindowScriptableObject::NPP_New):
1559
1560 2012-06-29  Dirk Pranke  <dpranke@chromium.org>
1561
1562         webkitpy: add comment about how determine_full_port_name() works for apple ports, fix -wk2 bug
1563         https://bugs.webkit.org/show_bug.cgi?id=90314
1564
1565         Reviewed by Ojan Vafai.
1566
1567         Add comments and fix a bug in how we would handle the mac-wk2
1568         and win-wk2 port names after confusion around in bug 90312 :).
1569
1570         * Scripts/webkitpy/layout_tests/port/apple.py:
1571         (ApplePort.determine_full_port_name):
1572
1573 2012-06-29  Dirk Pranke  <dpranke@chromium.org>
1574
1575         webkitpy: remove support for mac leopard from chromium configurations
1576         https://bugs.webkit.org/show_bug.cgi?id=90313
1577
1578         Reviewed by Tony Chang.
1579
1580         Google has shipped the last version of Chrome that will support
1581         Mac OS 10.5 (Leopard), and we no longer have bots that run this
1582         configuration, so we're removing support for it.
1583
1584         A subsequent change will remove the baselines in platform/chromium-mac-leopard.
1585
1586         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
1587         (BaselineOptimizerTest.test_complex_shadowing):
1588         * Scripts/webkitpy/layout_tests/port/builders.py:
1589         * Scripts/webkitpy/layout_tests/port/chromium.py:
1590         (ChromiumPort):
1591         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
1592         (ChromiumMacPort):
1593         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
1594         (ChromiumMacPortTest.test_versions):
1595         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
1596         (FactoryTest.test_mac):
1597         (FactoryTest.test_chromium_mac):
1598         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
1599         (test_rebaseline_and_copy_test_with_lion_result):
1600
1601 2012-06-29  Ojan Vafai  <ojan@chromium.org>
1602
1603         Fix optimize-baselines to not move baselines from win to win-7sp0
1604         https://bugs.webkit.org/show_bug.cgi?id=90312
1605
1606         Reviewed by Dirk Pranke.
1607
1608         It used to consider win-7sp0 as the common directory for all the Apple
1609         windows ports and incorrectly move results out of win.
1610
1611         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
1612         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
1613         (BaselineOptimizerTest.test_win_does_not_drop_to_win_7sp0):
1614         (BaselineOptimizerTest.test_common_directory_includes_root):
1615
1616 2012-06-28  Ojan Vafai  <ojan@chromium.org>
1617
1618         Fix rebaselining for Qt and Apple ports
1619         https://bugs.webkit.org/show_bug.cgi?id=90204
1620
1621         Reviewed by Dirk Pranke.
1622
1623         -Apporpriately put wk2 results in the -wk2 directories.
1624         -Since Qt and Apple-Win don't have bots that correspond to the
1625         platform/qt and platform/win directories, we need to fudge it
1626         and always put the results in those directories for those ports.
1627
1628         * Scripts/webkitpy/layout_tests/port/builders.py:
1629         (rebaseline_override_dir):
1630         * Scripts/webkitpy/layout_tests/port/factory.py:
1631         (_builder_options):
1632         Identify webkit2 builders by the WK2 in the builder name.
1633         * Scripts/webkitpy/tool/commands/rebaseline.py:
1634         (RebaselineTest._baseline_directory):
1635         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
1636         (TestRebaseline.test_baseline_directory):
1637
1638 2012-06-29  Zoltan Horvath  <zoltan@webkit.org>
1639
1640         Add support for --force parameter to run-performance-tests
1641         https://bugs.webkit.org/show_bug.cgi?id=90279
1642
1643         Reviewed by Dirk Pranke.
1644
1645         It's helpful to be able to run tests from the Skipped list of the performance tests.
1646
1647         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1648         (PerfTestsRunner._parse_args):
1649         (PerfTestsRunner._collect_tests):
1650         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py: Add test.
1651         (test_collect_tests_with_skipped_list):
1652
1653 2012-06-29  Tony Chang  <tony@chromium.org>
1654
1655         [GTK] Enable CSS grid layout LayoutTests on GTK+
1656         https://bugs.webkit.org/show_bug.cgi?id=90226
1657
1658         Reviewed by Martin Robinson.
1659
1660         This feature is disabled via Settings by default, but for testing,
1661         we enable it using layoutTestController.overridePreferences. Add the
1662         necessary plumbing for DRT.
1663
1664         WTR already works because support was added for Apple Mac earlier.
1665
1666         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1667         (resetDefaultsToConsistentValues): Feature is off by default.
1668         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
1669         (LayoutTestController::overridePreference): Add handling of WebKitCSSGridLayoutEnabled.
1670
1671 2012-06-29  Ádám Kallai  <kadam@inf.u-szeged.hu>
1672
1673         [Qt] Added Qt port for garden-o-matic.
1674         https://bugs.webkit.org/show_bug.cgi?id=82719
1675
1676         Reviewed by Adam Barth.
1677
1678         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
1679         (.):
1680
1681 2012-06-29  Balazs Kelemen  <kbalazs@webkit.org>
1682
1683         [Qt][WTR] Get rid of using DumpRenderTreeSupportQt
1684         https://bugs.webkit.org/show_bug.cgi?id=90262
1685
1686         Reviewed by Alexey Proskuryakov.
1687
1688         Now that we decided to not support v8 in WebKit2
1689         we can get rid of using DumpRenderTreeSupportQt
1690         in WebKitTestRunner.
1691
1692         * Tools.pro:
1693         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1694         (WTR::InjectedBundlePage::resetAfterTest):
1695         (WTR::InjectedBundlePage::didClearWindowForFrame):
1696         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
1697
1698 2012-06-29  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1699
1700         Revert r121540, it broke most Qt builds
1701
1702         * qmake/mkspecs/features/default_post.prf:
1703         * qmake/mkspecs/features/functions.prf:
1704
1705 2012-06-29  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
1706
1707         [Qt] Make build-webkit reject uknown configurations, eg. --profile
1708
1709         The qmake-based buildsystem doesn't support it.
1710
1711         Reviewed by Tor Arne Vestbø.
1712
1713         * Scripts/webkitdirs.pm:
1714         (buildQMakeProjects):
1715
1716 2012-06-29  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
1717
1718         [Qt] Don't add Qt module dependencies in features.prf
1719
1720         The required dependencies are already added in WebCore.pri.
1721
1722         Reviewed by Tor Arne Vestbø.
1723
1724         * qmake/mkspecs/features/features.prf:
1725
1726 2012-06-27  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
1727
1728         [Qt] Use LIBS_PRIVATE instead of putting dependencies into LIBS
1729
1730         Reviewed by Tor Arne Vestbø..
1731
1732         * qmake/mkspecs/features/default_post.prf:
1733         * qmake/mkspecs/features/functions.prf:
1734
1735 2012-06-28  Ryosuke Niwa  <rniwa@webkit.org>
1736
1737         DOMHTMLCollection::item may return a wrong element after namedItem is called
1738         https://bugs.webkit.org/show_bug.cgi?id=90240
1739
1740         Reviewed by Antti Koivisto.
1741
1742         Add a WebKit API test since namedItem is not used in the JS/V8 binding code.
1743
1744         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1745         * TestWebKitAPI/Tests/mac/HTMLCollectionNamedItem.html: Copied from Tools/TestWebKitAPI/Tests/mac/HTMLFormCollectionNamedItem.html.
1746         * TestWebKitAPI/Tests/mac/HTMLCollectionNamedItem.mm: Copied from Tools/TestWebKitAPI/Tests/mac/HTMLFormCollectionNamedItem.mm.
1747         (TestWebKitAPI::TEST):
1748
1749 2012-06-28  Dirk Pranke  <dpranke@chromium.org>
1750
1751         nrwt: remove the 'google-chrome' port code
1752         https://bugs.webkit.org/show_bug.cgi?id=88824
1753
1754         Reviewed by Ojan Vafai.
1755
1756         NRWT now supports passing additional baseline directories
1757         and expectations files on the command line, so there's no need
1758         to support the concept of a 'google-chrome' port directly.
1759
1760         * Scripts/webkitpy/layout_tests/port/base.py:
1761         (Port.path_to_test_expectations_file):
1762         * Scripts/webkitpy/layout_tests/port/builders.py:
1763         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
1764         (ChromiumMacPort.__init__):
1765         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
1766         (ChromiumWinPort.__init__):
1767         * Scripts/webkitpy/layout_tests/port/factory.py:
1768         (PortFactory):
1769         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
1770         (FactoryTest.test_win):
1771         * Scripts/webkitpy/layout_tests/port/google_chrome.py: Removed.
1772         * Scripts/webkitpy/layout_tests/port/google_chrome_unittest.py: Removed.
1773
1774 2012-06-28  Dirk Pranke  <dpranke@chromium.org>
1775
1776         nrwt: clean up how arguments are passed to workers
1777         https://bugs.webkit.org/show_bug.cgi?id=90126
1778
1779         Reviewed by Ojan Vafai.
1780
1781         The way arguments are passed to workers has been crufty. It
1782         turns out it can be a lot cleaner via two things:
1783         1) using a factory method instead of instantiating objects
1784         directly in manager_worker_broker removes the need for passing
1785         'worker arguments' to the broker.
1786         2) it turns out that since mock hosts and test ports are purely
1787         in-memory constructions, they can be pickled and passed to child
1788         workers, meaning that the worker no longer needs hacky code to
1789         pass the port in a special case or to guess what to do if we
1790         don't have a host - all of the test-specific logic can move to
1791         the test file, where we can stub out the mock host's
1792         port_factory to return the same already-created port when it
1793         needs to be shared.
1794
1795         This change also moves WorkerException to manager_worker_broker.py
1796         where it belongs, and removes several useless tests that were
1797         just a maintenance burden (and would've needed rewriting when we
1798         change the rest of the broker implementation).
1799
1800         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1801         (Manager._run_tests.worker_factory):
1802         (Manager._run_tests):
1803         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
1804         (get):
1805         (WorkerException):
1806         (AbstractWorker.__init__):
1807         (_ManagerConnection.__init__):
1808         (_ManagerConnection.start_worker):
1809         (_InlineManager.__init__):
1810         (_InlineManager.start_worker):
1811         (_MultiProcessManager._can_pickle_host):
1812         (_MultiProcessManager):
1813         (_MultiProcessManager.start_worker):
1814         (_WorkerConnection.__init__):
1815         (_InlineWorkerConnection.__init__):
1816         (_InlineWorkerConnection.join):
1817         (_InlineWorkerConnection.run):
1818         (_Process.run):
1819         (_MultiProcessWorkerConnection.__init__):
1820         (_MultiProcessWorkerConnection.start):
1821         (_MultiProcessWorkerConnection):
1822         (_MultiProcessWorkerConnection.run):
1823         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py:
1824         (_TestWorker.__init__):
1825         (_TestWorker.run):
1826         (_TestsMixin.test_name):
1827         (_TestsMixin.test_cancel):
1828         (_TestsMixin.test_done):
1829         (_TestsMixin.test_unknown_message):
1830         (InlineBrokerTests.setUp):
1831         (MultiProcessBrokerTests.setUp):
1832         * Scripts/webkitpy/layout_tests/controllers/worker.py:
1833         (Worker.__init__):
1834         (Worker.run):
1835         * Scripts/webkitpy/layout_tests/controllers/worker_unittest.py: Removed.
1836         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1837         (passing_run):
1838         (logging_run):
1839         (run_and_capture):
1840         (MainTest.test_child_processes_2):
1841         (MainTest.test_child_processes_min):
1842         (MainTest.test_exception_raised):
1843         (MainTest.test_keyboard_interrupt):
1844         (MainTest.test_retrying_and_flaky_tests):
1845         (MainTest.test_run_order__inline):
1846
1847 2012-06-28  Dirk Pranke  <dpranke@chromium.org>
1848
1849         nrwt: don't try to catch worker exceptions in run_webkit_tests.__main__
1850         https://bugs.webkit.org/show_bug.cgi?id=90125
1851
1852         Reviewed by Ojan Vafai.
1853
1854         It turns out run_webkit_tests.py wasn't really using
1855         WorkerException and the catch we had for it was pointless. I've
1856         removed the symbol import and moved it to the integration tests
1857         where it is needed. Eventually the definition of the exception
1858         moves to the broker module, and so minimizing the number of
1859         users of it is a good thing.
1860
1861         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1862         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1863         (MainTest.test_exception_raised):
1864
1865 2012-06-28  Dirk Pranke  <dpranke@chromium.org>
1866
1867         nrwt: clean up passing of log messages between processes
1868         https://bugs.webkit.org/show_bug.cgi?id=90123
1869
1870         Reviewed by Ojan Vafai.
1871
1872         It turns out log messages are perfectly picklable by themselves
1873         and contain the process id of the process that generated the
1874         message, so if we just pass the record from the worker to the
1875         manager and call logger.handle() in the manager, everything just
1876         works :).
1877
1878         The change is covered by existing tests.
1879
1880         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1881         (Manager._log_messages):
1882         * Scripts/webkitpy/layout_tests/controllers/worker.py:
1883         (_WorkerLogHandler.emit):
1884         * Scripts/webkitpy/layout_tests/views/metered_stream.py:
1885         (MeteredStream.__init__):
1886         (_LogHandler.emit):
1887
1888 2012-06-28  Dirk Pranke  <dpranke@chromium.org>
1889
1890         add a pylint wrapper for linting python code
1891         https://bugs.webkit.org/show_bug.cgi?id=90232
1892
1893         Reviewed by Adam Barth.
1894
1895         Currently we use 'pep8' to check python code in
1896         check-webkit-style. pep8 is fast but simple; pylint is slower
1897         but has much more robust linting capabilities and will catch
1898         variable typos and all sorts of other things. Eventually we
1899         should switch check-webkit-style to use this, but our code is
1900         far from linting now so it needs to be cleaned up first.
1901
1902         This change adds the infrastructure and a wrapper so we can
1903         start doing that.
1904
1905         * Scripts/lint-webkitpy: Added.
1906         * Scripts/webkitpy/pylintrc: Added.
1907         * Scripts/webkitpy/thirdparty/__init__.py:
1908         (AutoinstallImportHook.find_module):
1909         (AutoinstallImportHook._install_pylint):
1910
1911 2012-06-28  Tony Chang  <tony@chromium.org>
1912
1913         [GTK] Use WEBKITOUTPUTDIR to find fonts in DumpRenderTree
1914         https://bugs.webkit.org/show_bug.cgi?id=90215
1915
1916         Reviewed by Martin Robinson.
1917
1918         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1919         (initializeFonts): Check for WEBKITOUTPUTDIR first.
1920         * Scripts/webkitpy/layout_tests/port/gtk.py:
1921         (GtkPort.setup_environ_for_server): Copy the environment variable to the child process.
1922         * WebKitTestRunner/InjectedBundle/gtk/ActivateFontsGtk.cpp:
1923         (WTR::inititializeFontConfigSetting): Check for WEBKITOUTPUTDIR first.
1924
1925 2012-06-28  Dirk Pranke  <dpranke@chromium.org>
1926
1927         derive ChromiumPort from WebKitPort in NRWT in order to support skipping tests if symbols are missing
1928         https://bugs.webkit.org/show_bug.cgi?id=89706
1929
1930         Reviewed by Ojan Vafai.
1931
1932         Try again to land the change first landed in r121363. This patch
1933         adds a bunch more tests and reworks the handling of
1934         port-specific default values for --pixel-tests and --time-out-ms
1935         to be more consistent (adding a default_pixel_tests() method,
1936         pushing the webkit default_timeout_ms() value up into base.py,
1937         and overriding it properly in the chromium and apple mac ports.
1938
1939         Also change the logic in
1940         run_webkit_tests._setup_derived_options() to not second-guess
1941         what the port wants the timeout value to be for debug builds;
1942         computing this in two different places led to several bugs.
1943
1944         This change also changes the Chromium unittest ports to derive
1945         from ChromiumPortTestCase instead of PortTestCase, so that we
1946         ensure that we're running the same tests on all port variants
1947         more easily. There's more cleanup that can be done here, but
1948         this is good enough for now
1949
1950         * Scripts/webkitpy/layout_tests/port/base.py:
1951         (Port.default_pixel_tests):
1952         (Port):
1953         (Port.default_timeout_ms):
1954         * Scripts/webkitpy/layout_tests/port/chromium.py:
1955         (ChromiumPort):
1956         (ChromiumPort.__init__):
1957         (ChromiumPort.default_pixel_tests):
1958         (ChromiumPort.default_timeout_ms):
1959         (ChromiumPort.driver_name):
1960         (ChromiumPort._driver_class):
1961         (ChromiumPort._missing_symbol_to_skipped_tests):
1962         (ChromiumPort.skipped_layout_tests):
1963         (ChromiumPort.setup_test_run):
1964         (ChromiumPort._path_to_image_diff):
1965         (ChromiumPort._convert_path):
1966         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
1967         (ChromiumAndroidPortTest):
1968         (ChromiumAndroidPortTest.test_expectations_files):
1969         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
1970         (ChromiumLinuxPort._modules_to_search_for_symbols):
1971         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
1972         (ChromiumLinuxPortTest):
1973         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
1974         (ChromiumMacPort._modules_to_search_for_symbols):
1975         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
1976         (ChromiumMacPortTest):
1977         * Scripts/webkitpy/layout_tests/port/chromium_port_testcase.py:
1978         split off from chromium_unittest.
1979         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
1980         (ChromiumWinPort._modules_to_search_for_symbols):
1981         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
1982         (ChromiumWinTest):
1983         * Scripts/webkitpy/layout_tests/port/mac.py:
1984         (MacPort.default_timeout_ms):
1985         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
1986         (MacTest.test_default_timeout_ms):
1987         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
1988         (MockDRTPort.start_http_server):
1989         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
1990         (PortTestCase.test_default_timeout_ms):
1991         (PortTestCase):
1992         (PortTestCase.test_default_pixel_tests):
1993         * Scripts/webkitpy/layout_tests/port/test.py:
1994         (TestPort.default_pixel_tests):
1995         * Scripts/webkitpy/layout_tests/port/webkit.py:
1996         (WebKitPort._modules_to_search_for_symbols):
1997         (WebKitPort):
1998         (WebKitPort._symbols_string):
1999         (WebKitPort._skipped_tests_for_unsupported_features):
2000         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
2001         (TestWebKitPort._symbols_string):
2002         (TestWebKitPort._tests_for_disabled_features):
2003         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2004         (_set_up_derived_options):
2005
2006 2012-06-28  Joshua Bell  <jsbell@chromium.org>
2007
2008         run-bindings-tests should return non-zero exit code on test failure
2009         https://bugs.webkit.org/show_bug.cgi?id=90205
2010
2011         Reviewed by Adam Barth.
2012
2013         * Scripts/run-bindings-tests:
2014         (main):
2015
2016 2012-06-28  Terry Anderson  <tdanderson@chromium.org>
2017
2018         [chromium] Reset the device scale factor to 1 before each test is run
2019         https://bugs.webkit.org/show_bug.cgi?id=90212
2020
2021         Reviewed by Adam Barth.
2022
2023         Some tests change the device scale factor, so this needs to be reset to
2024         1.0 at the start of each test to avoid test flakiness.
2025
2026         * DumpRenderTree/chromium/LayoutTestController.cpp:
2027         (LayoutTestController::reset):
2028
2029 2012-06-28  Ryosuke Niwa  <rniwa@webkit.org>
2030
2031         Cleanup HTMLFormCollection
2032         https://bugs.webkit.org/show_bug.cgi?id=90111
2033
2034         Reviewed by Andreas Kling.
2035
2036         Add a WebKit API test using copy-paste design pattern per kling's request.
2037
2038         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
2039         * TestWebKitAPI/Tests/mac/HTMLFormCollectionNamedItem.html: Added.
2040         * TestWebKitAPI/Tests/mac/HTMLFormCollectionNamedItem.mm: Added.
2041         (-[HTMLFormCollectionNamedItemTest webView:didFinishLoadForFrame:]):
2042         (TestWebKitAPI):
2043         (TestWebKitAPI::TEST):
2044
2045 2012-06-28  Tim Horton  <timothy_horton@apple.com>
2046
2047         [mac] WKTR windows still don't stay off screen sometimes
2048         https://bugs.webkit.org/show_bug.cgi?id=90214
2049         <rdar://problem/11760263>
2050
2051         Reviewed by Simon Fraser.
2052
2053         In some cases, the system itself will consult [WebKitTestRunnerWindow frame], so we should refrain from
2054         overriding it and instead provide a different method to use when we want the web-facing "fake" window origin
2055         (for PlatformWebView::windowFrame()).
2056
2057         * WebKitTestRunner/mac/PlatformWebViewMac.mm:
2058         (-[WebKitTestRunnerWindow frameRespectingFakeOrigin]):
2059         (WTR::PlatformWebView::windowFrame):
2060
2061 2012-06-28  Martin Robinson  <mrobinson@igalia.com>
2062
2063         [GTK] Add unit tests for GtkInputMethodFilter
2064         https://bugs.webkit.org/show_bug.cgi?id=88698
2065
2066         Reviewed by Carlos Garcia Campos.
2067
2068         Add unit tests for GtkInputMethodFilter in the WebCore platform layer.
2069         This change adds the TestGtk test suite which will be used for all non-API
2070         layer GTK unit tests.
2071
2072         * TestWebKitAPI/GNUmakefile.am: Update the build to include the new tests.
2073         * TestWebKitAPI/Tests/gtk/InputMethodFilter.cpp: Added.
2074         (TestWebKitAPI::PlatformWebView::PlatformWebView): Remove the call to gtk_init here
2075         as it's now in main.cpp.
2076         * TestWebKitAPI/gtk/main.cpp: Change the g_type_init call to gtk_init, because now
2077         a majority of all unit tests depend on GTK+ being initialized.
2078
2079 2012-06-28  Christophe Dumez  <christophe.dumez@intel.com>
2080
2081         [EFL] Enable support for HTML5 datalist
2082         https://bugs.webkit.org/show_bug.cgi?id=90157
2083
2084         Reviewed by Martin Robinson.
2085
2086         Turn on DATALIST flag by default on EFL port to
2087         support HTML5 datalist tag.
2088
2089         * Scripts/webkitperl/FeatureList.pm:
2090
2091 2012-06-28  Tony Chang  <tony@chromium.org>
2092
2093         Enable CSS grid layout LayoutTests on platform Mac
2094         https://bugs.webkit.org/show_bug.cgi?id=90113
2095
2096         Reviewed by Ojan Vafai.
2097
2098         * DumpRenderTree/mac/DumpRenderTree.mm:
2099         (resetDefaultsToConsistentValues): Reset the value to NO between tests.
2100
2101 2012-06-28  Mike Fenton  <mifenton@rim.com>
2102
2103         [BlackBerry] Add watchlist options for Blackberry and editing.
2104         https://bugs.webkit.org/show_bug.cgi?id=90193
2105
2106         Unreviewed.
2107
2108         Add BlackBerry and Editing watchlist and monitor them.
2109
2110         * Scripts/webkitpy/common/config/watchlist:
2111
2112 2012-06-28  Tim Horton  <timothy_horton@apple.com>
2113
2114         Unreviewed, upgrade myself to a reviewer!
2115
2116         http://www.webkit.org/blog/2082/tim-horton-is-now-a-webkit-reviewer/
2117
2118         * Scripts/webkitpy/common/config/committers.py:
2119
2120 2012-06-28  Terry Anderson  <tdanderson@chromium.org>
2121
2122         [Chromium] Chromium's LayoutTestController is missing setBackingScaleFactor
2123         https://bugs.webkit.org/show_bug.cgi?id=83635
2124
2125         Reviewed by Adam Barth.
2126
2127         Added InvokeCallbackTask, a new derived class of MethodTask. When
2128         setBackingScaleFactor is called, a call to setDeviceScaleFactor in
2129         WebView is made and then postTask is used to invoke the callback
2130         function specified in testRunner.setBackingScaleFactor. 
2131
2132         * DumpRenderTree/chromium/LayoutTestController.cpp:
2133         (LayoutTestController::LayoutTestController):
2134         (InvokeCallbackTask):
2135         (InvokeCallbackTask::InvokeCallbackTask):
2136         (InvokeCallbackTask::runIfValid):
2137         (LayoutTestController::setBackingScaleFactor):
2138         * DumpRenderTree/chromium/LayoutTestController.h:
2139         (LayoutTestController):
2140
2141 2012-06-28  Ojan Vafai  <ojan@chromium.org>
2142
2143         Make rebaseline-test and rebaseline-expectations work for non-Chromium ports
2144         https://bugs.webkit.org/show_bug.cgi?id=90186
2145
2146         Reviewed by Adam Barth.
2147
2148         This makes rebaselining work for all ports that have a TestExpectations file
2149         in the tree. I didn't test other ports.
2150
2151         This doesn't address 100% of the problem. The rebaseline code puts the expectations
2152         in the most specific directory and relies on optimize-baselines to merge baselines
2153         appropriately. This only works if every platform directory has an equivalent bot
2154         that runs the tests, which is not true for most ports.
2155
2156         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
2157         (Builder._revision_and_build_for_filename):
2158         Some bots have filenames that aren't revision/build number pairs
2159         e.g. they are random junk like aQhxvx. Handle this gracefully.
2160         (Builder._fetch_revision_to_build_map):
2161         (Builder._file_info_list_to_revision_to_build_list):
2162         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2163         (BuilderTest.test_build_and_revision_for_filename):
2164         (BuilderTest.test_file_info_list_to_revision_to_build_list):
2165         * Scripts/webkitpy/layout_tests/port/builders.py:
2166         Update the list of builders. This list needs to be kept up
2167         to date for the rebaseline tool to work.
2168         * Scripts/webkitpy/tool/commands/rebaseline.py:
2169         (RebaselineTest._results_url):
2170         (RebaselineExpectations._run_webkit_patch):
2171         (RebaselineExpectations._rebaseline_port):
2172         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
2173         Qt port uses qmake to determine the right version. Systems without qmake correctly fallback
2174         to a specific version.
2175
2176 2012-06-28  Csaba Osztrogonác  <ossy@webkit.org>
2177
2178         [Qt][NRWT] Fix baseline and skipped file search path.
2179         https://bugs.webkit.org/show_bug.cgi?id=89882
2180
2181         Unreviewed trivial typo fix after r121430.
2182
2183         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
2184         (QtPortTest._assert_search_path):
2185         (QtPortTest._assert_skipped_path):
2186
2187 2012-06-28  János Badics  <jbadics@inf.u-szeged.hu>
2188
2189         [Qt][NRWT] Fix baseline and skipped file search path.
2190         https://bugs.webkit.org/show_bug.cgi?id=89882
2191
2192         Reviewed by Csaba Osztrogonác.
2193
2194         * Scripts/webkitpy/layout_tests/port/qt.py:
2195         (QtPort._search_paths):
2196         (QtPort):
2197         (QtPort.baseline_search_path):
2198         (QtPort._skipped_file_search_paths):
2199         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
2200         (QtPortTest):
2201         (QtPortTest._assert_skipped_path):
2202         (QtPortTest.test_baseline_search_path):
2203         (QtPortTest.test_skipped_file_search_path):
2204
2205 2012-06-28  Alexander Pavlov  <apavlov@chromium.org>
2206
2207         [Qt] inspector/styles/inject-stylesheet.html makes 4 tests flakey (TEXT PASS)
2208         https://bugs.webkit.org/show_bug.cgi?id=90167
2209
2210         Reviewed by Csaba Osztrogonác.
2211
2212         DRT should remove user stylesheets from the page group when resetting before running another test.
2213
2214         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
2215         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
2216
2217 2012-06-28  Csaba Osztrogonác  <ossy@webkit.org>
2218
2219         [Qt][DRT] Reset AcceleratedCompositingEnabled between tests
2220         https://bugs.webkit.org/show_bug.cgi?id=90164
2221
2222         Reviewed by Simon Hausmann.
2223
2224         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
2225         (WebCore::WebPage::resetSettings):
2226
2227 2012-06-28  Csaba Osztrogonác  <ossy@webkit.org>
2228
2229         [Qt] Restore original value of mock scrollbars between tests
2230         https://bugs.webkit.org/show_bug.cgi?id=90155
2231
2232         Reviewed by Simon Hausmann.
2233
2234         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
2235         (WebCore::WebPage::resetSettings):
2236         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
2237         (LayoutTestController::setMockScrollbarsEnabled):
2238         * DumpRenderTree/qt/LayoutTestControllerQt.h:
2239         (LayoutTestController):
2240
2241 2012-06-27  Mark Rowe  <mrowe@apple.com>
2242
2243         <http://webkit.org/b/90096> Building within Xcode sometimes does a full rebuild after building via build-webkit
2244
2245         We need to ensure that build-webkit uses the same setting for SHARED_PRECOMPS_DIR
2246         as what Xcode itself will use when building, otherwise switching between the two
2247         can result in the precompiled headers being rebuilt and thus the entire world
2248         rebuilding.
2249
2250         Reviewed by Dan Bernstein.
2251
2252         * Scripts/webkitdirs.pm:
2253         (determineBaseProductDir):
2254
2255 2012-06-27  Sheriff Bot  <webkit.review.bot@gmail.com>
2256
2257         Unreviewed, rolling out r121363, r121367, r121384, and
2258         r121390.
2259         http://trac.webkit.org/changeset/121363
2260         http://trac.webkit.org/changeset/121367
2261         http://trac.webkit.org/changeset/121384
2262         http://trac.webkit.org/changeset/121390
2263         https://bugs.webkit.org/show_bug.cgi?id=90134
2264
2265         It broke debug NRWT on GTK and on Qt (Requested by Ossy_NIGHT
2266         on #webkit).
2267
2268         * Scripts/webkitpy/layout_tests/port/chromium.py:
2269         (ChromiumPort):
2270         (ChromiumPort.__init__):
2271         (ChromiumPort._check_file_exists):
2272         (ChromiumPort.default_results_directory):
2273         (ChromiumPort._driver_class):
2274         (ChromiumPort._build_path):
2275         (ChromiumPort._path_to_image_diff):
2276         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2277         (ChromiumLinuxPort.baseline_search_path):
2278         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
2279         (ChromiumMacPort.baseline_search_path):
2280         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
2281         (ChromiumPortTest):
2282         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2283         (ChromiumWinPort.baseline_search_path):
2284         * Scripts/webkitpy/layout_tests/port/webkit.py:
2285         (WebKitPort.__init__):
2286         (WebKitPort._webcore_symbols_string):
2287         (WebKitPort._skipped_tests_for_unsupported_features):
2288         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
2289         (TestWebKitPort._webcore_symbols_string):
2290         (WebKitPortUnitTests.test_default_options):
2291
2292 2012-06-27  Dirk Pranke  <dpranke@chromium.org>
2293
2294         Fix typo in r121384 :(
2295
2296         Unreviewed, build fix.
2297
2298         * Scripts/webkitpy/layout_tests/port/webkit.py:
2299         (WebKitPort.default_test_timeout_ms):
2300
2301 2012-06-27  Dirk Pranke  <dpranke@chromium.org>
2302
2303         nrwt: default timeout for debug bots broke in r121363
2304         https://bugs.webkit.org/show_bug.cgi?id=90112
2305
2306         Unreviewed, build fix.
2307
2308         I forgot to account for release and debug having different
2309         default values :(.
2310
2311         * Scripts/webkitpy/layout_tests/port/chromium.py:
2312         (ChromiumPort.default_test_timeout_ms):
2313         * Scripts/webkitpy/layout_tests/port/webkit.py:
2314         (WebKitPort):
2315         (WebKitPort.default_test_timeout_ms):
2316         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
2317         (WebKitPortUnitTests.test_default_options):
2318
2319 2012-06-27  Dirk Pranke  <dpranke@chromium.org>
2320
2321         webkitpy: fix a couple of issues running under cygwin
2322         https://bugs.webkit.org/show_bug.cgi?id=90035
2323
2324         Reviewed by Eric Seidel.
2325
2326         These were causing unit tests to fail on cygwin (apple win bot).
2327
2328         * Scripts/webkitpy/layout_tests/port/chromium.py:
2329         * Scripts/webkitpy/performance_tests/perftest.py:
2330
2331 2012-06-27  Adam Barth  <abarth@webkit.org>
2332
2333         [Chromium] DumpRenderTree on Android should call SkUseTestFontConfigFile once available
2334         https://bugs.webkit.org/show_bug.cgi?id=89801
2335
2336         Reviewed by Nate Chapin.
2337
2338         Let's call SkUseTestFontConfigFile now that it exists.
2339
2340         * DumpRenderTree/chromium/TestShellAndroid.cpp:
2341         (platformInit):
2342
2343 2012-06-27  Dirk Pranke  <dpranke@chromium.org>
2344
2345         Fix typo introduced in r121363.
2346
2347         Unreviewed, build fix.
2348
2349         * Scripts/webkitpy/layout_tests/port/webkit.py:
2350         (WebKitPort._symbols_string):
2351
2352 2012-06-27  Mark Hahnenberg  <mhahnenberg@apple.com>
2353
2354         Moving from committers to reviewers :-)
2355
2356         Rubber stamped by Filip Pizlo.
2357
2358         * Scripts/webkitpy/common/config/committers.py:
2359
2360 2012-06-27  Dirk Pranke  <dpranke@chromium.org>
2361
2362         Derive ChromiumPort from WebKitPort to add support for missing symbols to skip tests
2363         https://bugs.webkit.org/show_bug.cgi?id=89706
2364
2365         Reviewed by Adam Barth.
2366
2367         Based on the original patch by Raymond Toy.
2368
2369         This patch changes ChromiumPort to derive from webkit.WebKitPort
2370         instead of base.Port. This is a long-awaited change and a
2371         precursor to merging base.Port and webkit.WebKitPort, but is
2372         driven by the desire to dynamically detect whether the MP3 and
2373         AAC codecs are compiled into DRT on Chromium, for which we
2374         wanted to re-use the existing logic in WebKit port for determine
2375         what to skip at compile time.
2376
2377         Most of the changes are shuffling things around so that we don't
2378         change any other logic and so we override the necessary methods
2379         in WebKitPort (and try to follow the same method definition
2380         order where possible).
2381
2382         Also, on the Chromium port the mp3 and aac codecs are actually
2383         defined in a separate library, so scanning webcore isn't
2384         sufficient. This patch generalizes the symbol lookup to handle
2385         multiple libraries, and uses different libraries as appropriate
2386         for Chromium.
2387
2388         The only functional/visible changes should be in the value
2389         returned for skipped_layout_tests().
2390
2391         * Scripts/webkitpy/layout_tests/port/chromium.py:
2392         (ChromiumPort):
2393         (ChromiumPort.__init__):
2394         (ChromiumPort.driver_name):
2395         (ChromiumPort._driver_class):
2396         (ChromiumPort._missing_symbol_to_skipped_tests):
2397         (ChromiumPort.skipped_layout_tests):
2398         (ChromiumPort.setup_test_run):
2399         (ChromiumPort._path_to_image_diff):
2400         (ChromiumPort._convert_path):
2401         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
2402         (ChromiumPortTest.test_missing_symbol_to_skipped_tests):
2403         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2404         (ChromiumLinuxPort._modules_to_search_for_symbols):
2405         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
2406         (ChromiumLinuxPort._modules_to_search_for_symbols):
2407         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2408         (ChromiumLinuxPort._modules_to_search_for_symbols):
2409         * Scripts/webkitpy/layout_tests/port/webkit.py:
2410         (WebKitPort):
2411         (WebKitPort.__init__):
2412         (WebKitPort._symbols_string):
2413         (WebKitPort._modules_to_search_for_symbols):
2414
2415 2012-06-27  Csaba Osztrogonác  <ossy@webkit.org>
2416
2417         [Qt] REGRESSION(r121339): It broke the build on the Qt Windows bots
2418         https://bugs.webkit.org/show_bug.cgi?id=90081
2419
2420         Buildfix for Qt 4.8 Windows. Use the former path for Qt 4.8, and the newer one for Qt 5.
2421
2422         Reviewed by Noam Rosenthal.
2423
2424         * qmake/mkspecs/features/features.prf:
2425
2426 2012-06-27  Sergio Villar Senin  <svillar@igalia.com>
2427
2428         [WK2] [GTK] WebKit2 testing bot fails to run tests due to missing files
2429         https://bugs.webkit.org/show_bug.cgi?id=90061
2430
2431         Reviewed by Gustavo Noronha Silva.
2432
2433         Add "-no-install -no-fast-install" to the LDFLAGS in bots. With
2434         this flag libtool tells the linker to set the rpath for the output
2435         file to the full path of the .libs directory, instead of using a
2436         wrapper script to set up the LD_LIBRARY_PATH. This will allow us
2437         to directly reuse builds in the pure testing bots.
2438
2439         * BuildSlaveSupport/gtk/daemontools-buildbot.conf:
2440
2441 2012-06-27  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2442
2443         [Qt] No need to save and restore TEMPLATE in a function
2444
2445         This was a leftover from when the logic was not in its own function scope.
2446
2447         QMAKE_FRAMEWORK_BUNDLE_NAME on the other hand is exported in qtLibraryTarget, which
2448         will surprisingly affect the global scope as well, so we have to save and restore it.
2449
2450         Original patch by Oswald Buddenhagen <oswald.buddenhagen@nokia.com> on 2012-06-27
2451
2452         Reviewed by Tor Arne Vestbø.
2453
2454         * qmake/mkspecs/features/functions.prf:
2455
2456 2012-06-27  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
2457
2458         [Qt] Fix lookup location for sqlite sources
2459
2460         Don't look in the install dir - we are unlikely to find anything there
2461         unless we are doing a developer build.
2462
2463         Reviewed by Tor Arne Vestbø.
2464
2465         * qmake/mkspecs/features/features.prf:
2466
2467 2012-06-27  Zan Dobersek  <zandobersek@gmail.com>
2468
2469         [Gtk] Add support for the Gamepad API
2470         https://bugs.webkit.org/show_bug.cgi?id=87503
2471
2472         Reviewed by Carlos Garcia Campos.
2473
2474         Enable the gamepad support for the GTK port.
2475
2476         * Scripts/webkitperl/FeatureList.pm:
2477
2478 2012-06-27  Ryosuke Niwa  <rniwa@webkit.org>
2479
2480         Fix gcc build after r121302
2481         https://bugs.webkit.org/show_bug.cgi?id=90055
2482
2483         Reviewed by Mark Rowe.
2484
2485         Assume RTTI is always disabled so that gtest builds on XCode 3.2.6.
2486
2487         It appears that gcc doesn't like window.get().* inside a template so replace that by [window.get() *] instead.
2488
2489         * TestWebKitAPI/Configurations/Base.xcconfig:
2490         * TestWebKitAPI/Tests/mac/AcceptsFirstMouse.mm:
2491         (TestWebKitAPI::AcceptsFirstMouse::runTest):
2492
2493 2012-06-27  Csaba Osztrogonác  <ossy@webkit.org>
2494
2495         Add master.cfg unittest to help migration - pass BuildStep instances instead of BuildStep subclasses
2496         https://bugs.webkit.org/show_bug.cgi?id=89564
2497
2498         Reviewed by Tony Chang.
2499
2500         * BuildSlaveSupport/build.webkit.org-config/mastercfg_unittest.py:
2501         (BuildStepsConstructorTest):
2502         (BuildStepsConstructorTest.generateTests):
2503         (BuildStepsConstructorTest.createTest):
2504         (BuildStepsConstructorTest.createTest.doTest):
2505
2506 2012-06-26  Mark Hahnenberg  <mhahnenberg@apple.com>
2507
2508         Add support for preciseTime() to WebKitTestRunner
2509         https://bugs.webkit.org/show_bug.cgi?id=90027
2510
2511         Reviewed by Darin Adler.
2512
2513         It would be nice to be able to use preciseTime() in WebKitTestRunner like we can in DumpRenderTree.
2514
2515         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
2516         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp: 
2517         (WTR::LayoutTestController::preciseTime): 
2518         (WTR):
2519         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
2520         (LayoutTestController):
2521
2522 2012-06-26  Lucas Forschler  <lforschler@apple.com>
2523
2524       Teach the Apple port how to build the test tools in build-webkit
2525       https://bugs.webkit.org/show_bug.cgi?id=89540
2526
2527       Reviewed by Jon Lee & Simon Fraser.
2528
2529       * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2530       (RunWebKitTests.start): pass --no-build since tools should now be in the downloaded archive
2531       (RunUnitTests.start): ditto
2532       * Scripts/build-webkit: add tools to the projects build list
2533
2534 2012-06-26  Ojan Vafai  <ojan@chromium.org>
2535
2536         Only show the platform-appropriate builders for non-layout test failures in garden-o-matic
2537         https://bugs.webkit.org/show_bug.cgi?id=90025
2538
2539         Reviewed by Simon Fraser.
2540
2541         Move the chromium-specific filtering code into config.js and replace it with a method on each platform
2542         config. Also, let the webkit test step name be webkit_tests (build.chromium.org) or layout-test (build.webkit.org).
2543
2544         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders.js:
2545         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders_unittests.js:
2546         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
2547
2548 2012-06-26  Ojan Vafai  <ojan@chromium.org>
2549
2550         Fix failing garden-o-matic unittests
2551         https://bugs.webkit.org/show_bug.cgi?id=90021
2552
2553         Reviewed by Adam Barth.
2554
2555         These had all just fallen out of date.
2556
2557         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout_unittests.js:
2558         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
2559         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:
2560
2561 2012-06-26  Ojan Vafai  <ojan@chromium.org>
2562
2563         Fix platform picker change handler in garden-o-matic
2564         https://bugs.webkit.org/show_bug.cgi?id=90010
2565
2566         Reviewed by Simon Fraser.
2567
2568         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
2569         The old code never worked. This is hard to test because change handlers require
2570         a user-initiated action and the code is changing the window's location, which would
2571         navigate the test page.
2572
2573         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/onebar.css:
2574         Fix the CSS so it doesn't cause the tabstrip to be disconnected from the tabs on Linux
2575         due to the large margin-bottom.
2576
2577 2012-06-26  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2578
2579         webkitpy: Make webkit-patch patches-to-review useful
2580         https://bugs.webkit.org/show_bug.cgi?id=89470
2581
2582         Reviewed by Eric Seidel.
2583
2584         webkit-patch patches-to-review will now output the list of
2585         bugs with patches pending for review that has the user on CC,
2586         excluding patches with cq-, sorted by the age of the patch.
2587
2588         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
2589         (BugzillaQueries.fetch_bugs_from_review_queue):
2590         * Scripts/webkitpy/common/net/bugzilla/bugzilla_mock.py:
2591         (MockBugzillaQueries.fetch_bugs_from_review_queue):
2592         (MockBugzilla.__init__):
2593         (MockBugzilla.authenticate):
2594         * Scripts/webkitpy/tool/commands/queries.py:
2595         (PatchesToReview):
2596         (PatchesToReview.__init__):
2597         (PatchesToReview._print_report):
2598         (PatchesToReview._generate_report):
2599         (PatchesToReview.execute):
2600         * Scripts/webkitpy/tool/commands/queries_unittest.py:
2601         (QueryCommandsTest.test_patches_to_review):
2602
2603 2012-06-26  Dirk Pranke  <dpranke@chromium.org>
2604
2605         nrwt: broken for chromium on vista
2606         https://bugs.webkit.org/show_bug.cgi?id=89988
2607
2608         Reviewed by Tony Chang.
2609
2610         r121194 removed support for 'chromium-win-vista' as a separate
2611         port, but this actually prevented the code from running on vista
2612         at all, which is unduly harsh and broke the websocket tests on
2613         the (non-webkit) chromium bots that are still running on vista.
2614
2615         It's probably good enough to pretend that vista is win7 instead;
2616         some layout tests will still fail but at least things'll run.
2617
2618         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2619         (ChromiumWinPort.determine_full_port_name):
2620         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
2621         (ChromiumWinTest.test_versions):
2622
2623 2012-06-26  Csaba Osztrogonác  <ossy@webkit.org>
2624
2625         master.cfg cleanup: Pass CheckOutSource instance instead of class to BuildStep.addStep
2626         https://bugs.webkit.org/show_bug.cgi?id=89215
2627
2628         We need it because it is deprecated and will be dropped in buildbot 0.8.7
2629
2630         Reviewed by Tony Chang.
2631
2632         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2633         (CheckOutSource.__init__):
2634         (Factory.__init__):
2635
2636 2012-06-26  Simon Hausmann  <simon.hausmann@nokia.com>
2637
2638         [Qt][Win] Symbols are not exported in QtWebKit5.dll
2639         https://bugs.webkit.org/show_bug.cgi?id=88873
2640
2641         Reviewed by Tor Arne Vestbø.
2642
2643         When linking the target dll make sure to re-export the symbols from
2644         the static libraries marked as export, with the help of a little python
2645         script and a qmake extra compiler.
2646
2647         * Scripts/generate-win32-export-forwards: Added.
2648         * qmake/mkspecs/features/win32/default_post.prf:
2649
2650 2012-06-25  Jocelyn Turcotte  <turcotte.j@gmail.com>
2651
2652         Add a note about hostname completion not working well with --cc completion
2653
2654         Reviewed by Simon Hausmann.
2655
2656         Hostname completion tries to resolve anything after an @ sign which is present
2657         in the completed list of contributor emails to CC.
2658
2659         * Scripts/webkit-tools-completion.sh:
2660
2661 2012-06-26  Sheriff Bot  <webkit.review.bot@gmail.com>
2662
2663         Unreviewed, rolling out r121244.
2664         http://trac.webkit.org/changeset/121244
2665         https://bugs.webkit.org/show_bug.cgi?id=89966
2666
2667         skip list path is incorrect now. (Requested by Ossy on
2668         #webkit).
2669
2670         * Scripts/webkitpy/layout_tests/port/qt.py:
2671         (QtPort.baseline_search_path):
2672         (QtPort._skipped_file_search_paths):
2673         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
2674         (QtPortTest.test_baseline_search_path):
2675
2676 2012-06-26  János Badics  <jbadics@inf.u-szeged.hu>
2677
2678         [Qt][NRWT] Fix baseline and skipped file search path.
2679         https://bugs.webkit.org/show_bug.cgi?id=89882
2680
2681         Reviewed by Csaba Osztrogonác.
2682
2683         * Scripts/webkitpy/layout_tests/port/qt.py:
2684         (QtPort.baseline_search_path):
2685         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
2686         (QtPortTest.test_baseline_search_path):
2687
2688 2012-06-26  Tony Chang  <tony@chromium.org>
2689
2690         [Qt] Enable grid layout LayoutTests
2691         https://bugs.webkit.org/show_bug.cgi?id=89909
2692
2693         Reviewed by Csaba Osztrogonác.
2694
2695         These tests pass, we just need to hook up the overridePreference.
2696
2697         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
2698         (WebCore::WebPage::resetSettings): Reset grid layout and regions between tests.
2699         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
2700         (LayoutTestController::overridePreference): Add WebKitCSSGridLayoutEnabled.
2701
2702 2012-06-26  Sheriff Bot  <webkit.review.bot@gmail.com>
2703
2704         Unreviewed, rolling out r121236.
2705         http://trac.webkit.org/changeset/121236
2706         https://bugs.webkit.org/show_bug.cgi?id=89956
2707
2708         It's broke the nrwt on qt and gtk platform (Requested by
2709         kkristof on #webkit).
2710
2711         * Scripts/webkitpy/layout_tests/port/qt.py:
2712         (QtPort._build_driver):
2713         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
2714         (XvfbDriver._start):
2715         (XvfbDriver._start.x_filter):
2716         (XvfbDriver.stop):
2717
2718 2012-06-26  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
2719
2720         [NRWT] XvfbDriver should choose the next free display
2721         https://bugs.webkit.org/show_bug.cgi?id=88414
2722
2723         Reviewed by Dirk Pranke.
2724
2725         * Scripts/webkitpy/layout_tests/port/qt.py:
2726         (QtPort._driver_class):
2727         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
2728         (XvfbDriver.__init__):
2729         (XvfbDriver._start):
2730         (XvfbDriver._start.next_free_id):
2731         (XvfbDriver.stop):
2732
2733 2012-06-25  Simon Hausmann  <simon.hausmann@nokia.com>
2734
2735         [Qt] Make it possible to build WebKit without QtWidgets
2736         https://bugs.webkit.org/show_bug.cgi?id=78109
2737
2738         Reviewed by Tor Arne Vestbø.
2739
2740         * Tools.pro: Disable various tools that depend on WK1.
2741         * qmake/mkspecs/features/default_post.prf: Add WK1 as supported static lib in WEBKIT
2742         * qmake/mkspecs/features/default_pre.prf: Disable WK1 if QtWidgets is not built.
2743         * qmake/mkspecs/modules/webkit1.prf: Added.
2744
2745 2012-06-25  Dirk Pranke  <dpranke@chromium.org>
2746
2747         put vista back for chromium (non-webkit) bots ...
2748         https://bugs.webkit.org/show_bug.cgi?id=89929
2749
2750         Unreviewed, build fix.
2751
2752         Turns out we've only gotten rid of vista on the webkit tests so
2753         far. Put vista back for the other chromium bots ...
2754
2755         * TestResultServer/static-dashboards/flakiness_dashboard.js:
2756         (chromiumPlatform):
2757         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2758         (test):
2759
2760 2012-06-25  Dirk Pranke  <dpranke@chromium.org>
2761
2762         update flakiness dashboard after removing chromium-vista bots
2763         https://bugs.webkit.org/show_bug.cgi?id=89925
2764
2765         Unreviewed, build fix.
2766
2767         More changes missed in r121194 :(.
2768
2769         * TestResultServer/static-dashboards/flakiness_dashboard.js:
2770         (chromiumPlatform):
2771         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2772         (test):
2773
2774 2012-06-25  Dirk Pranke  <dpranke@chromium.org>
2775
2776         remove support for chromium vista from tools
2777         https://bugs.webkit.org/show_bug.cgi?id=89915
2778
2779         Reviewed by Tony Chang.
2780
2781         This change is all deleting code and updating tests to handle
2782         the configuration being gone. All tests now pass except for a
2783         few unittests for the flakiness dashboard which were failing
2784         before this patch.
2785
2786         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders_unittests.js:
2787         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout_unittests.js:
2788         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
2789         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
2790         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
2791         (BaselineOptimizerTest.test_complex_shadowing):
2792         * Scripts/webkitpy/layout_tests/port/builders.py:
2793         * Scripts/webkitpy/layout_tests/port/chromium.py:
2794         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
2795         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2796         (ChromiumWinPort):
2797         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
2798         (ChromiumWinTest.test_versions):
2799         (ChromiumWinTest.test_baseline_path):
2800         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
2801         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
2802         (BuildCoverageExtrapolatorTest.test_extrapolate):
2803
2804 2012-06-25  Scott Graham  <scottmg@chromium.org>
2805
2806         [Chromium] add empty impl of WebThemeEngine::getSize() for DRT
2807         https://bugs.webkit.org/show_bug.cgi?id=89907
2808
2809         Not reviewed. Fix build in Chromium.
2810
2811         * DumpRenderTree/chromium/WebThemeEngineDRTWin.cpp:
2812         (WebThemeEngineDRTWin::getSize):
2813         * DumpRenderTree/chromium/WebThemeEngineDRTWin.h:
2814         (WebThemeEngineDRTWin):
2815
2816 2012-06-25  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
2817
2818         [EFL] Implement Network Information API
2819         https://bugs.webkit.org/show_bug.cgi?id=87067
2820
2821         Reviewed by Kenneth Rohde Christiansen.
2822
2823         * efl/jhbuild.modules: Add eeze library dependency.
2824
2825 2012-06-25  Simon Hausmann  <simon.hausmann@nokia.com>
2826
2827         [Qt] Fix Win32 debug build
2828
2829         Reviewed by Tor Arne Vestbø.
2830
2831         Consolidate use_all_in_one_files into mac/unix/win32 default_pre.prf.
2832
2833         * qmake/mkspecs/features/unix/default_pre.prf:
2834         * qmake/mkspecs/features/win32/default_pre.prf: Added.
2835
2836 2012-06-24  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2837
2838         [Qt] Enable <input type="color"> support by default
2839         https://bugs.webkit.org/show_bug.cgi?id=89653
2840
2841         Reviewed by Kenneth Rohde Christiansen.
2842
2843         This feature is complete for Qt WebKit2. Enabling by default
2844         will help to mature the implementation.
2845
2846         * Scripts/webkitperl/FeatureList.pm:
2847         * qmake/mkspecs/features/features.pri:
2848
2849 2012-06-24  Dan Bernstein  <mitz@apple.com>
2850
2851         Made debug-{minibrowser,safari,test-runner} work with LLDB.
2852
2853         Reviewed by Sam Weinig.
2854
2855         * Scripts/debug-minibrowser: Pass INCLUDE_OPTIONS_FOR_DEBUGGING to
2856         printHelpAndExitForRunAndDebugWebKitAppIfNeeded().
2857         * Scripts/debug-safari: Ditto.
2858         * Scripts/debug-test-runner: Ditto.
2859         * Scripts/webkitdirs.pm:
2860         (debugger): Added. Calls determineDebugger() if needed and returns the chosen debugger.
2861         (determineDebugger): Added. Sets the debugger to "lldb" if the --use-lldb switch is present,
2862         and to "gdb" otherwise.
2863         (printHelpAndExitForRunAndDebugWebKitAppIfNeeded): Changed to print help for the
2864         --target-web-process and --use-lldb switches if passed INCLUDE_OPTIONS_FOR_DEBUGGING.
2865         (execMacWebKitAppForDebugging): Changed to use the chosen debugger.
2866
2867 2012-06-24  Adam Barth  <abarth@webkit.org>
2868
2869         [Chromium] Release media resources after each LayoutTest on Android
2870         https://bugs.webkit.org/show_bug.cgi?id=89720
2871
2872         Reviewed by Eric Carlson.
2873
2874         The LayoutTests can hang when run on Android because we sometimes run
2875         out of media resources. This patch causes us to release our media
2876         resources after each test in order to avoid running out of this
2877         resource. In production, there are other mechanisms that manage this
2878         resource.
2879
2880         * DumpRenderTree/chromium/TestShell.cpp:
2881         (TestShell::resetTestController):
2882
2883 2012-06-23  Zan Dobersek  <zandobersek@gmail.com>
2884
2885         [Gtk] REGRESSION(r120918): Causes flaky DND tests
2886         https://bugs.webkit.org/show_bug.cgi?id=89770
2887
2888         Reviewed by Martin Robinson.
2889
2890         Connect to the run-file-chooser signal of the WebKitWebView and
2891         return TRUE so no file chooser dialog is run. This prevents certain
2892         test failures.
2893
2894         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2895         (webViewRunFileChooser):
2896         (createWebView):
2897
2898 2012-06-23  Dan Bernstein  <mitz@apple.com>
2899
2900         debug-{minibrowser,safari,test-runner} fail when gdb is not installed in /usr/bin.
2901
2902         Reviewed by Sam Weinig.
2903
2904         * Scripts/webkitdirs.pm:
2905         (execMacWebKitAppForDebugging): Use xcrun to locate gdb.
2906
2907 2012-06-23  Carlos Garcia Campos  <cgarcia@igalia.com>
2908
2909         Unreviewed. Skip GTK+ unit test /webkit2/WebKitFindController/hide.
2910
2911         It fails always when running it in Xvfb.
2912
2913         * gtk/run-api-tests:
2914         (TestRunner):
2915
2916 2012-06-23  Wajahat Siddiqui  <mdwajahatali.siddiqui@motorola.com>
2917
2918         [GTK][WK2]Mis-spelt defaultWindowTitle
2919         https://bugs.webkit.org/show_bug.cgi?id=89808
2920
2921         Reviewed by Carlos Garcia Campos.
2922
2923         * MiniBrowser/gtk/BrowserWindow.c:
2924
2925 2012-06-23  Sergio Villar Senin  <svillar@igalia.com>
2926
2927         [GTK] Add a new webkit2 tests slave bot
2928         https://bugs.webkit.org/show_bug.cgi?id=89336
2929
2930         Reviewed by Csaba Osztrogonác.
2931
2932         Slave configuration for a new GTK 64 bit release bot that will run
2933         WebKit2 tests.
2934
2935         Also BuildAndTest now accepts the "triggers" parametter,
2936         this means that apart from building and testing it can trigger
2937         some other build/test sequences in other slaves.
2938
2939         * BuildSlaveSupport/build.webkit.org-config/config.json:
2940         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2941         (BuildAndTestFactory.__init__):
2942
2943 2012-06-23  Kwang Yul Seo  <skyul@company100.net>
2944
2945         Unreviewed. Update my email.
2946
2947         * Scripts/webkitpy/common/config/committers.py:
2948
2949 2012-06-22  Adam Barth  <abarth@webkit.org>
2950
2951         [Chromium] DumpRenderTree on Android needs to configure fonts for testing
2952         https://bugs.webkit.org/show_bug.cgi?id=89721
2953
2954         Reviewed by Nate Chapin.
2955
2956         This patch teaches DumpRenderTree to configure Skia to use the fallback
2957         fonts that the LayoutTests assume. This patch reduces the number result
2958         differences between chromium-android and chromium-linux.
2959
2960         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
2961         * DumpRenderTree/chromium/TestShellAndroid.cpp:
2962         (platformInit):
2963         * DumpRenderTree/chromium/android_fallback_fonts.xml: Added.
2964         * DumpRenderTree/chromium/android_main_fonts.xml: Added.
2965         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2966
2967 2012-06-22  Alexandru Chiculita  <achicu@adobe.com>
2968
2969         [CSS Shaders] Re-enable the CSS Shaders compile time flag on Safari Mac
2970         https://bugs.webkit.org/show_bug.cgi?id=89781
2971
2972         Reviewed by Dean Jackson.
2973
2974         Added CSS Shaders as enabled by default on Safari for Mac.
2975
2976         * Scripts/webkitperl/FeatureList.pm:
2977
2978 2012-06-22  Adam Barth  <abarth@webkit.org>
2979
2980         Add support for test_expectations_android.txt for overriding test expecations on the chromium-android branch
2981         https://bugs.webkit.org/show_bug.cgi?id=89791
2982
2983         Reviewed by Dirk Pranke.
2984
2985         Downstream, the chromium-android port maintains a test expectations
2986         file to keep track of which tests pass or fail. That causes them to
2987         keep a patch in chromium_android.py in their branch.
2988
2989         This patch teaches chromium_android.py to use test_expectations_android.txt.
2990         We don't intend to use this file upstream, but having this code patch
2991         upstream lets us unfork webkitpy.
2992
2993         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2994         (ChromiumAndroidPort.expectations_files):
2995
2996 2012-06-22  Dirk Pranke  <dpranke@chromium.org>
2997
2998         run_webkit_tests.py failed with AttributeError(''NoneType' object has no attribute 'pid'')
2999         https://bugs.webkit.org/show_bug.cgi?id=89734
3000
3001         Reviewed by Ryosuke Niwa.
3002
3003         Fix a crash in ServerProcess if you called .pid() after it
3004         crashed during a write(). We had a test for this case but the
3005         test wasn't calling pid(), just has_crashed(). Fixed the problem
3006         and the test.
3007
3008         * Scripts/webkitpy/layout_tests/port/server_process.py:
3009         (ServerProcess.__init__):
3010         (ServerProcess.pid):
3011         (ServerProcess._start):
3012         (ServerProcess.stop):
3013         * Scripts/webkitpy/layout_tests/port/server_process_unittest.py:
3014         (FakeServerProcess._start):
3015         (TestServerProcess.test_broken_pipe):
3016
3017 2012-06-22  Peter Beverloo  <peter@chromium.org>
3018
3019         [Chromium] Disable c++0x compatibility warnings in JavaScriptCore.gyp when building for Android
3020         https://bugs.webkit.org/show_bug.cgi?id=88853
3021
3022         Reviewed by Steve Block.
3023
3024         The Android exclusions were necessary to fix a gyp generation error, as
3025         the gcc_version variable wasn't being defined for Android. Remove these
3026         exceptions when Chromium is able to define the gcc_version variable.
3027
3028         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
3029
3030 2012-06-22  Mario Sanchez Prada  <msanchez@igalia.com>
3031
3032         Unreviewed gardening. Unskipping test passing after fixing bug 88419.
3033
3034         * gtk/run-api-tests:
3035         (TestRunner): Unskip WTF.HashMap.
3036
3037 2012-06-22  Zoltan Horvath  <zoltan@webkit.org>
3038
3039         [Qt] Allow DumpRenderTree to dump about:blank
3040         https://bugs.webkit.org/show_bug.cgi?id=89685
3041
3042         Reviewed by Ryosuke Niwa.
3043
3044         We need to allow DumpRenderTree to dump about:blank page, then it will be consistent with other ports,
3045         additionaly this behavior is required for running WTR performance tests.
3046
3047         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
3048         (WebCore::DumpRenderTree::processLine):
3049
3050 2012-06-21  Adam Barth  <abarth@webkit.org>
3051
3052         [Chromium] Reset mediaPlaybackRequiresUserGesture WebSettings after each test
3053         https://bugs.webkit.org/show_bug.cgi?id=89718
3054
3055         Reviewed by Kent Tamura.
3056
3057         We should reset this WebSetting to its default value so that it behaves
3058         consistently across tests.
3059
3060         * DumpRenderTree/chromium/WebPreferences.cpp:
3061         (WebPreferences::reset):
3062         (WebPreferences::applyTo):
3063         * DumpRenderTree/chromium/WebPreferences.h:
3064         (WebPreferences):
3065
3066 2012-06-21  Zoltan Horvath  <zoltan@webkit.org>
3067
3068         DRT/WTR python interface handles about:blank incorrectly
3069         https://bugs.webkit.org/show_bug.cgi?id=89563
3070
3071         Reviewed by Dirk Pranke.
3072
3073         Currently DRT/WTR python interface's _command_from_driver_input function
3074         puts the test directory path before about:blank. Remove this incorrect behavior.
3075
3076         * Scripts/webkitpy/layout_tests/port/webkit.py:
3077         (WebKitDriver._command_from_driver_input):
3078
3079 2012-06-21  Dirk Pranke  <dpranke@chromium.org>
3080
3081         reenable perf tests on win
3082         https://bugs.webkit.org/show_bug.cgi?id=89690
3083
3084         Reviewed by Ryosuke Niwa.
3085
3086         Only the replay tests don't work, and those are disabled by
3087         default. This change also fixes the undefined _log reference
3088         that was causing a crash.
3089
3090         * Scripts/run-perf-tests:
3091
3092 2012-06-21  Sheriff Bot  <webkit.review.bot@gmail.com>
3093
3094         Unreviewed, rolling out r120937.
3095         http://trac.webkit.org/changeset/120937
3096         https://bugs.webkit.org/show_bug.cgi?id=89679
3097
3098         This patch brought buildbot master down (Requested by svillar
3099         on #webkit).
3100
3101         * BuildSlaveSupport/build.webkit.org-config/config.json:
3102
3103 2012-06-21  Martin Robinson  <mrobinson@igalia.com>
3104
3105         [GTK] Combine WebKit API tests into fewer binaries
3106         https://bugs.webkit.org/show_bug.cgi?id=88458
3107
3108         Reviewed by Carlos Garcia Campos.
3109
3110         Instead of creating one binary per-test file, create binaries for each
3111         category of tests. Right now this includes WTF and the WebKit2 C API,
3112         but later tests can be added for the GTK+ platform layer and the WebKit1
3113         and WebKit2 API layers.
3114
3115         * TestWebKitAPI/GNUmakefile.am: Compile only two test binaries, one for the WebKit2
3116           C API and one for WTF. Refresh the source list to ensure that new test are active.
3117         * gtk/run-api-tests: Change the way that tests are skipped by splitting out the
3118         concept of skipping a test and skipping a suite (program) of tests. Test cases are
3119         skipped because of legitimate failures, but entire programs are skipped because of
3120         problems in the harness. As of right now a test program is only skipped if the
3121         accessibility bus cannot be started.
3122         (SkippedTest.__init__): Make the test case a required argument and have one skipped
3123         test case per SkippedTest instance.
3124         (SkippedTest.__str__): Ditto.
3125         (TestRunner): Update the directory list. We only have two gtest binaries now and they are
3126         in one directory. Reformat the test list to make it slightly easier to read.
3127         (TestRunner.__init__): Add the _skipped_test_program member, which handles entire
3128         test programs that are skipped.
3129         (TestRunner._setup_testing_environment): Use the new member.
3130         (TestRunner._test_cases_to_skip): Collect all skipped test cases now instead of just
3131         the first SkippedTest that matches.
3132         (TestRunner._should_run_test_program): Take a look at the new member to make this
3133         decision.
3134         (TestRunner._run_test_command): Use the name test_program instead of test
3135         to disambiguate between test cases and test suites.
3136         (TestRunner._run_test_glib): ditto.
3137         (TestRunner._run_test_google): Ditto.
3138         (TestRunner._run_test): Ditto.
3139         (TestRunner.run_tests): Ditto.
3140
3141 2012-06-21  Yong Li  <yoli@rim.com>
3142
3143         Unreviewed. Moving myself from committer to reviewer.
3144
3145         * Scripts/webkitpy/common/config/committers.py:
3146
3147 2012-06-21  Sergio Villar Senin  <svillar@igalia.com>
3148
3149         [GTK] Add a new webkit2 tests slave bot
3150         https://bugs.webkit.org/show_bug.cgi?id=89336
3151
3152         Reviewed by Philippe Normand.
3153
3154         Slave configuration for a new GTK bot that will run WebKit2 tests.
3155
3156         * BuildSlaveSupport/build.webkit.org-config/config.json:
3157
3158 2012-06-21  Thiago Marcos P. Santos  <thiago.santos@intel.com>
3159
3160         [Qt] Add a custom Color Chooser widget to MiniBrowser
3161         https://bugs.webkit.org/show_bug.cgi?id=87988
3162
3163         Reviewed by Alexis Menard.
3164
3165         Added custom color chooser example for the MiniBrowser.
3166
3167         * MiniBrowser/qt/MiniBrowser.qrc:
3168         * MiniBrowser/qt/qml/BrowserWindow.qml:
3169         * MiniBrowser/qt/qml/ColorChooser.qml: Added.
3170
3171 2012-06-21  Vivek Galatage  <vivekgalatage@gmail.com>
3172
3173         Remove the warning "File not found" in MiniBrowser post-build event
3174         https://bugs.webkit.org/show_bug.cgi?id=89601
3175
3176         Reviewed by Adam Roben.
3177
3178         * MiniBrowser/MiniBrowserPostBuild.cmd:
3179
3180 2012-06-21  Balazs Kelemen  <kbalazs@webkit.org>
3181
3182         [Qt] DRT in standalone mode hangs after the first test
3183         https://bugs.webkit.org/show_bug.cgi?id=89613
3184
3185         Reviewed by Csaba Osztrogonác.
3186
3187         Break an inline signal slot chain that ends up in setting
3188         LayoutTestController::m_hasDumped too early. In LayoutTestController::maybeDump()
3189         we emit done() which starts the chain that finishes in DumpRenderTree::open().
3190         The next line in maybeDump sets m_hasDumped to true. We could simply reorder
3191         these two lines but it seems to be wrong in general to start the next load
3192         from this call chain. The server mode (when DRT is runned by the test harness)
3193         also uses a queued connection to read the next test from stdin.
3194
3195         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
3196         (WebCore::DumpRenderTree::processArgsLine):
3197
3198 2012-06-20  Hans Wennborg  <hans@chromium.org>
3199
3200         Speech JavaScript API: SpeechRecognition should hook up with ActiveDOMObject more
3201         https://bugs.webkit.org/show_bug.cgi?id=89217
3202
3203         Reviewed by Adam Barth.
3204
3205         Add a method for checking whether the mock speech
3206         recognition was aborted.
3207
3208         Also redo the way the mock posts tasks. Instead of posting them all at once,
3209         maintain an internal queue of task objects, and call postTask() for them
3210         once at the time. This means that for example when the page is navigated
3211         away and abort() is called, that call doesn't end up after a bunch
3212         of previously posted events on the event loop.
3213
3214         * DumpRenderTree/chromium/LayoutTestController.cpp:
3215         (LayoutTestController::LayoutTestController):
3216         (LayoutTestController::wasMockSpeechRecognitionAborted):
3217         * DumpRenderTree/chromium/LayoutTestController.h:
3218         (LayoutTestController):
3219         * DumpRenderTree/chromium/MockWebSpeechRecognizer.cpp:
3220         (WebKit::ClientCallTask::ClientCallTask):
3221         (WebKit::ResultTask::ResultTask):
3222         (WebKit::NoMatchTask::NoMatchTask):
3223         (WebKit::ErrorTask::ErrorTask):
3224         (MockWebSpeechRecognizer::start):
3225         (MockWebSpeechRecognizer::abort):
3226         (MockWebSpeechRecognizer::setError):
3227         (MockWebSpeechRecognizer::MockWebSpeechRecognizer):
3228         (MockWebSpeechRecognizer::startTaskQueue):
3229         (MockWebSpeechRecognizer::StepTask::runIfValid):
3230         * DumpRenderTree/chromium/MockWebSpeechRecognizer.h:
3231         (MockWebSpeechRecognizer::hasBeenAborted):
3232         (MockWebSpeechRecognizer):
3233         (MockWebSpeechRecognizer::taskList):
3234         (Task):
3235         (MockWebSpeechRecognizer::Task::Task):
3236         (MockWebSpeechRecognizer::Task::~Task):
3237         (StepTask):
3238         (MockWebSpeechRecognizer::StepTask::StepTask):
3239
3240 2012-06-21  Mario Sanchez Prada  <msanchez@igalia.com>
3241
3242         [GTK] Add support for window.showModalDialog in WebKit2GTK+
3243         https://bugs.webkit.org/show_bug.cgi?id=79500
3244
3245         Reviewed by Carlos Garcia Campos.
3246
3247         Add support for modal dialogs in GTK's MiniBrowser.
3248
3249         * MiniBrowser/gtk/BrowserWindow.c:
3250         (webViewRunAsModal):
3251         (webViewCreate):
3252         (webViewDecidePolicy):
3253         (browser_window_new):
3254         * MiniBrowser/gtk/BrowserWindow.h:
3255         * MiniBrowser/gtk/main.c:
3256         (createBrowserWindow):
3257
3258 2012-06-20  Brent Fulgham  <bfulgham@webkit.org>
3259
3260         [WinCairo] Unreviewed build fix.
3261         The wrapper script that launched new- and old-run-webkit-tests
3262         was discarding the --wincairo argument needed to get the proper
3263         test infrastructure to build under WinCairo.
3264
3265         * Scripts/run-webkit-tests: Emulate the Qt, Wx, Chromium, etc.,
3266         behavior to chain the --wincairo flag through to the new build
3267         and test scripts.
3268
3269 2012-06-20  Mark Rowe  <mrowe@apple.com>
3270
3271         <http://webkit.org/b/89606> Teach run-safari and debug-safari to work with a Safari.app that has entitlements
3272
3273         Reviewed by Dan Bernstein.
3274
3275         * Scripts/webkitdirs.pm:
3276         (executableHasEntitlements):
3277         (safariPathFromSafariBundle):
3278
3279 2012-06-20  Sheriff Bot  <webkit.review.bot@gmail.com>
3280
3281         Unreviewed, rolling out r120821.
3282         http://trac.webkit.org/changeset/120821
3283         https://bugs.webkit.org/show_bug.cgi?id=89605
3284
3285         It made duplicated reviewer entries (Requested by Ossy on
3286         #webkit).
3287
3288         * Scripts/webkitpy/common/checkout/changelog.py:
3289         (ChangeLog.set_reviewer):
3290         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
3291         (test_set_reviewer):
3292         (test_set_short_description_and_bug_url):
3293
3294 2012-06-20  Dirk Pranke  <dpranke@chromium.org>
3295
3296         nrwt outputs empty files for wdiff output if wdiff is not installed
3297         https://bugs.webkit.org/show_bug.cgi?id=88709
3298
3299         Reviewed by Tony Chang.
3300
3301         Reviewed by Tony Chang.
3302
3303         Don't write -wdiff or -pretty.html files if wdiff or prettypatch
3304         aren't available, and clean up the handling for them in the port
3305         code.
3306
3307         * Scripts/webkitpy/run_webkit_tests_integrationtest.py:
3308         (MainTest.test_output_diffs):
3309         * Scripts/webkitpy/layout_tests/controllers/test_result_writer.py:
3310         (TestResultWriter.create_text_diff_and_write_result):
3311         * Scripts/webkitpy/layout_tests/port/base.py:
3312         (Port.check_pretty_patch):
3313         (Port.check_wdiff):
3314         (Port._wdiff_missing_message):
3315         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
3316         (ChromiumLinuxPort.check_build):
3317         (ChromiumLinuxPort._wdiff_missing_message):
3318         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
3319         (ChromiumMacPort.check_build):
3320         (ChromiumMacPort):
3321         (ChromiumMacPort._wdiff_missing_message):
3322         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
3323         (ChromiumMacPortTest):
3324
3325 2012-06-20  Dirk Pranke  <dpranke@chromium.org>
3326
3327         nrwt: fix unit tests for ensuring svn revision is correct
3328         https://bugs.webkit.org/show_bug.cgi?id=89498
3329
3330         Reviewed by Eric Seidel.
3331
3332         Reviewed by Eric Seidel.
3333
3334         Fix the unit tests for testing that we are embedding the SVN
3335         revision in the results json for NRWT properly; this was broken
3336         in r120646 but because the MockHost worked differently than the
3337         real one (by always initializing the MockSCM object) we didn't
3338         notice. Unfortunately, just changing the default breaks all
3339         sorts of unit tests ...
3340
3341         * Scripts/webkitpy/common/host_mock.py:
3342         (MockHost.__init__):
3343         (MockHost._initialize_scm):
3344         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
3345         (ResultSummaryTest.test_no_svn_revision):
3346         (ResultSummaryTest.test_svn_revision):
3347
3348 2012-06-20  Dirk Pranke  <dpranke@chromium.org>
3349
3350         new-run-webkit-tests appends "/Debug" or "/Release" to $WEBKITOUTPUTDIR
3351         https://bugs.webkit.org/show_bug.cgi?id=69360
3352
3353         Reviewed by Eric Seidel.
3354
3355         Reviewed by Eric Seidel.
3356
3357         Propagate the 'port_implementation' part of the platform (i.e.,
3358         gtk,qt,chromium) to webkit-build-directory so that we can pick
3359         up the gtk-specific handling of WEBKITOUTPUTDIR ...
3360
3361         I didn't write any additional tests for this; testing it
3362         properly is an integration test between the python code and the
3363         perl code, which I verified by hand.
3364
3365         * Scripts/webkitpy/layout_tests/port/base.py:
3366         (Port.__init__):
3367         * Scripts/webkitpy/layout_tests/port/config.py:
3368         (Config.__init__):
3369         (Config.build_directory):
3370         * Scripts/webkitpy/layout_tests/port/config_mock.py:
3371         (MockConfig.__init__):
3372         * Scripts/webkitpy/layout_tests/port/config_unittest.py:
3373         (ConfigTest.test_build_directory_passes_port_implementation):
3374
3375 2012-06-20  Dirk Pranke  <dpranke@chromium.org>
3376
3377         Fix import sorting missed in r120846
3378
3379         Reviewed by Tony Chang.
3380
3381         * Scripts/webkitpy/layout_tests/servers/apache_http_server_unittest.py:
3382
3383 2012-06-20  Dirk Pranke  <dpranke@chromium.org>
3384
3385         Fix chromium win http servers after breakage introduced in r120846.
3386
3387         Unreviewed, build fix.
3388
3389         * Scripts/webkitpy/layout_tests/port/base.py:
3390         (Port.to.start_http_server):
3391         * Scripts/webkitpy/layout_tests/servers/http_server.py:
3392         (Lighttpd.__init__):
3393
3394 2012-06-20  Dirk Pranke  <dpranke@chromium.org>
3395
3396         tweak output of webkit-patch print-{baselines,expectations}
3397         https://bugs.webkit.org/show_bug.cgi?id=89588
3398
3399         Reviewed by Tony Chang.
3400
3401         Minor tweaks to the output to make it more readable, including
3402         turning off csv by default for print-baselines with multiple ports.
3403
3404         * Scripts/webkitpy/layout_tests/port/test.py:
3405         (TestPort):
3406         * Scripts/webkitpy/tool/commands/queries.py:
3407         (PrintExpectations.execute):
3408         (PrintBaselines.execute):
3409         * Scripts/webkitpy/tool/commands/queries_unittest.py:
3410         (PrintExpectationsTest.run_test):
3411         (PrintExpectationsTest.test_multiple):
3412         (PrintBaselinesTest.setUp):
3413         (PrintBaselinesTest.test_multiple):
3414
3415 2012-06-18  Dirk Pranke  <dpranke@chromium.org>
3416
3417         new-run-webkit-tests should spin-up enough httpd processes to avoid timeouts
3418         https://bugs.webkit.org/show_bug.cgi?id=88134
3419
3420         Reviewed by Tony Chang.
3421
3422         Change NRWT to spin up 2*min(child_processes, locked_shards)
3423         http servers by default so that we are less likely to get a
3424         bunch of http timeouts at the beginning of a test run.
3425
3426         Note that I had to tweak executive_mock to support mocked stderr
3427         because the apache_http_server code reads stderr when starting a
3428         process to ensure it started okay.
3429
3430         * Scripts/webkitpy/common/system/executive_mock.py:
3431         (MockProcess.__init__):
3432         * Scripts/webkitpy/common/system/outputcapture.py:
3433         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3434         (Manager._run_tests):
3435         (Manager.start_servers_with_lock):
3436         * Scripts/webkitpy/layout_tests/port/base.py:
3437         (Port.to.start_http_server):
3438         * Scripts/webkitpy/layout_tests/port/test.py:
3439         (TestPort.start_http_server):
3440         (TestPort._path_to_apache):
3441         (TestPort):
3442         (TestPort._path_to_apache_config_file):
3443         * Scripts/webkitpy/layout_tests/servers/apache_http_server.py:
3444         (LayoutTestApacheHttpd.__init__):
3445         * Scripts/webkitpy/layout_tests/servers/apache_http_server_unittest.py: Added
3446         (LayoutTestApacheHttpd.__init__):
3447         * Scripts/webkitpy/layout_tests/servers/http_server.py:
3448         (Lighttpd.__init__):
3449         * Scripts/webkitpy/layout_tests/servers/http_server_base.py:
3450         (HttpServerBase.__init__):
3451
3452 2012-06-20  Tom Zakrajsek  <tomz@codeaurora.org>
3453
3454         Unreviewed: Back out accidentally checked in debug print which broke a test
3455
3456         * Scripts/webkitpy/performance_tests/perftest.py:
3457         (PerfTest.parse_output):
3458
3459 2012-06-20  Tom Zakrajsek  <tomz@codeaurora.org>
3460
3461         Import themaninblue.com/experiment/AnimationBenchmark/ as performance tests
3462         https://bugs.webkit.org/show_bug.cgi?id=78789
3463
3464         Reviewed by Ryosuke Niwa.
3465
3466         Updated parser to include fps as a valid unit.
3467
3468         * Scripts/webkitpy/performance_tests/perftest.py:
3469         (PerfTest):
3470         (PerfTest.parse_output):
3471
3472 2012-06-20  Balazs Ankes  <bank@inf.u-szeged.hu>
3473
3474         webkit-patch should add reviewer if "Reviewed by NOBODY ..." is missing
3475         https://bugs.webkit.org/show_bug.cgi?id=67935
3476
3477         Reviewed by Ryosuke Niwa.
3478
3479         * Scripts/webkitpy/common/checkout/changelog.py:
3480         (ChangeLog.set_reviewer):
3481         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
3482         (test_set_reviewer):
3483         (test_set_short_description_and_bug_url):
3484
3485 2012-06-20  Simon Hausmann  <simon.hausmann@nokia.com>
3486
3487         Unreviewed trivial permissions fix.
3488
3489         * qmake/mkspecs/features/features.prf: Don't mark this file as executable.
3490
3491 2012-06-19  Ryuan Choi  <ryuan.choi@samsung.com>
3492
3493         [EFL][Regression] Build break after r120786
3494         https://bugs.webkit.org/show_bug.cgi?id=89549
3495
3496         Unreviewed build fix for Efl port.
3497
3498         * DumpRenderTree/efl/CMakeLists.txt: Added disassembler in includes.
3499
3500 2012-06-19  Dirk Pranke  <dpranke@chromium.org>
3501
3502         NRWT spins up and down the WebSocket server when running a single HTTP test from the command line
3503         https://bugs.webkit.org/show_bug.cgi?id=64489
3504
3505         Reviewed by Tony Chang.
3506
3507         This change changes NRWT so that we don't start the websocket
3508         server if we aren't running websocket tests and don't start the
3509         HTTP server if we aren't running http tests.
3510         
3511         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3512         (Manager._websocket_tests):
3513         (Manager.start_servers_with_lock):
3514         (Manager.stop_servers_with_lock):
3515         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
3516         (ManagerTest.test_servers_started):
3517
3518 2012-06-19  Dirk Pranke  <dpranke@chromium.org>
3519
3520         test-webkitpy: add a -p flag to pass through captured output to enable debugging
3521         https://bugs.webkit.org/show_bug.cgi?id=89158
3522
3523         Reviewed by Eric Seidel.
3524
3525         Normally when outputcapture runs it intercepts stdout/stderr,
3526         which makes running under the debugger difficult and mostly
3527         pointless (since you can't see any output). This change adds a
3528         flag to test-webkitpy (-p) that will cause outputcapture to pass
3529         through the output as well as capture it.
3530
3531         * Scripts/webkitpy/common/system/outputcapture.py:
3532         (OutputCapture.stream_wrapper):
3533         (OutputCapture._capture_output_with_name):
3534         * Scripts/webkitpy/common/system/outputcapture_unittest.py:
3535         (OutputCaptureTest.setUp):
3536         * Scripts/webkitpy/test/main.py:
3537         (Tester._parse_args):
3538         (Tester._run_tests):
3539         (Tester._log_exception):
3540         (_CaptureAndPassThroughStream):
3541         (_CaptureAndPassThroughStream.__init__):
3542         (_CaptureAndPassThroughStream.write):
3543         (_CaptureAndPassThroughStream._message_is_from_pdb):
3544         (_CaptureAndPassThroughStream.flush):
3545         (_CaptureAndPassThroughStream.getvalue):
3546
3547 2012-06-19  Adam Barth  <abarth@webkit.org>
3548
3549         garden-o-matic fails to rebaseline tests with MISSING results
3550         https://bugs.webkit.org/show_bug.cgi?id=89521
3551
3552         Reviewed by Kenneth Russell.
3553
3554         If the bots report "MISSING", we should try to find PNG and TXT results on the server.
3555
3556         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results.js:
3557
3558 2012-06-19  Adam Barth  <abarth@webkit.org>
3559
3560         svn.delete_list fails to delete empty parent directories
3561         https://bugs.webkit.org/show_bug.cgi?id=89520
3562
3563         Reviewed by Dirk Pranke.
3564
3565         These functions were calling scm.delete and scm.add, which already
3566         delete the parent directory. We'd get an exception when we tried to
3567         delete the parent directory again.
3568
3569         * Scripts/webkitpy/common/checkout/scm/svn.py:
3570         (SVN._add_parent_directories):
3571         (SVN._delete_parent_directories):
3572
3573 2012-06-19  Thiago Marcos P. Santos  <thiago.santos@intel.com>
3574
3575         webkitpy: Simplify fetch_bugs_matching_search()
3576         https://bugs.webkit.org/show_bug.cgi?id=89497
3577
3578         Reviewed by Adam Barth.
3579
3580         Removed parameter with default value. It's not being used
3581         and the handling of the parameter was wrong. Saves a check.
3582
3583         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
3584         (BugzillaQueries.fetch_bugs_matching_search):
3585         * Scripts/webkitpy/common/net/bugzilla/bugzilla_mock.py:
3586         (MockBugzillaQueries.fetch_bugs_matching_search):
3587
3588 2012-06-19  Dirk Pranke  <dpranke@chromium.org>
3589
3590         webkitpy: remove unneeded imports in layout_tests/port/chromium.py
3591         https://bugs.webkit.org/show_bug.cgi?id=89522
3592
3593         Reviewed by Eric Seidel.
3594
3595         They were no longer being used. No tests necessary.
3596
3597         * Scripts/webkitpy/layout_tests/port/chromium.py:
3598
3599 2012-06-19  Adam Barth  <abarth@webkit.org>
3600
3601         Missing results aren't shown in garden-o-matic
3602
3603         Reviewed by Kenneth Russell.
3604
3605         Previously we didn't know to look for text and image results when the
3606         bot told us that results were missing.
3607
3608         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results.js:
3609
3610 2012-06-19  Dirk Pranke  <dpranke@chromium.org>
3611
3612         Fix regression introduced in r120646 where the svn revision isn't saved in results.json
3613
3614         Unreviewed, build fix.
3615
3616         It turns out that we store the svn revision two different ways
3617         in the JSON, and I only caught one of them :(. This should fix
3618         the other.
3619
3620         Also, this part of the code is poorly tested. I will investigate
3621         adding at least a basic test for this in a separate fix.
3622
3623         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3624         (summarize_results):
3625
3626 2012-06-19  Dirk Pranke  <dpranke@chromium.org>
3627
3628         REGRESSION (NRWT): Results for new non-text-only tests are always put in the most-specific platform directory
3629         https://bugs.webkit.org/show_bug.cgi?id=78127
3630
3631         Reviewed by Ryosuke Niwa.
3632
3633         Add the --add-platform-exceptions flag from ORWT, and ensure
3634         that --new-baseline is equivalent to --reset-results
3635         --add-platform-exceptions. Also fix the default logic for where to
3636         put PNGs and render tree dumps if --new-test-results is true:
3637         if --add-platform-exceptions is False, baselines should go in
3638         the least-specific platform directory (e.g., platform/mac)
3639         rather than the most-specific (platform/mac-snowleopard).
3640
3641         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
3642         (SingleTestRunner):
3643         (SingleTestRunner.run):
3644         (SingleTestRunner._run_rebaseline):
3645         (SingleTestRunner._add_missing_baselines):
3646         (SingleTestRunner._location_for_new_baseline):
3647         (SingleTestRunner._overwrite_baselines):
3648         (SingleTestRunner._save_baseline_data):
3649         * Scripts/webkitpy/layout_tests/port/base.py:
3650         (Port.baseline_path):
3651         (Port):
3652         (Port.baseline_platform_dir):
3653         (Port.baseline_version_dir):
3654         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3655         (_set_up_derived_options):
3656         (parse_args):
3657         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3658         (RebaselineTest.assertBaselines):
3659         (RebaselineTest.test_reset_results):
3660         (RebaselineTest.test_missing_results):
3661         (RebaselineTest.test_new_baseline):
3662
3663 2012-06-19  Dirk Pranke  <dpranke@chromium.org>
3664
3665         NRWT should not take memory used as disk cache into account when deciding how many processes to launch
3666         https://bugs.webkit.org/show_bug.cgi?id=81379
3667
3668         Reviewed by Eric Seidel.
3669
3670         The 'free memory' calculation we were running on the mac seemed
3671         to underestimate how many children we can run in parallel, and
3672         it was complex. This patch replaces that calculation with a
3673         simpler one that reserves 2GB for overhead and assumes 256MB per
3674         DRT/WTR, so if we had 4GB of RAM we can run up to 8 DRTs.
3675
3676         Also, there was a bug where we were truncating the memory
3677         installed on the machine to 4GB by casting to an int instead of
3678         a long; this was probably the source of some of the earlier
3679         problems when using total memory.
3680
3681         This patch also removes the no-longer-needed restrictions on the
3682         number of workers on beefy Lion boxes for both Apple Mac and Chromium
3683         Mac; we should now use all of the cores by default.
3684
3685         The memory calculations have only been implemented on the mac;
3686         having the calculation in base.default_child_processes() was IMO
3687         misleading, and so this patch also moves the computation into
3688         the MacPort. I have not heard of the # of workers being an issue
3689         on any other ports, so this should be fine.
3690
3691         * Scripts/webkitpy/common/system/platforminfo.py:
3692         (PlatformInfo.total_bytes_memory):
3693         * Scripts/webkitpy/common/system/platforminfo_mock.py:
3694         (PlatformInfo.total_bytes_memory):
3695         (PlatformInfo.total_bytes_memory):
3696         (PlatformInfo._win_version_tuple_from_cmd):
3697         * Scripts/webkitpy/common/system/platforminfo_unittest.py:
3698         (TestPlatformInfo.test_total_bytes_memory):
3699         * Scripts/webkitpy/layout_tests/port/base.py:
3700         (Port.default_child_processes):
3701         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
3702         (PortTest.test_default_child_processes):
3703         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
3704         (ChromiumMacPort.operating_system):
3705         * Scripts/webkitpy/layout_tests/port/mac.py:
3706         (MacPort.default_child_processes):
3707         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
3708         (TestMacPort.test_default_child_processes):
3709
3710 2012-06-19  Dirk Pranke  <dpranke@chromium.org>
3711
3712         new-run-webkit-tests reports unexpected pass of pixel tests when pixel testing is disabled
3713         https://bugs.webkit.org/show_bug.cgi?id=85446
3714
3715         Reviewed by Simon Fraser.
3716
3717         Embed whether pixel testing was enabled into the results.json.
3718
3719         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3720         (summarize_results):
3721
3722 2012-06-19  Zoltan Horvath  <zoltan@webkit.org>
3723
3724         [Qt] Modify HTTPS port to 8443 for performance tests
3725         https://bugs.webkit.org/show_bug.cgi?id=89442
3726
3727         Reviewed by Ryosuke Niwa.
3728
3729         We should be consistent with the default HTTPS port and with the changelog of r119188.
3730
3731         * Scripts/webkitpy/performance_tests/perftest.py:
3732         (ReplayServer.__init__):
3733
3734 2012-06-19  Mike West  <mkwst@chromium.org>
3735
3736         Introduce ENABLE_CSP_NEXT configuration flag.
3737         https://bugs.webkit.org/show_bug.cgi?id=89300
3738
3739         Reviewed by Adam Barth.
3740
3741         The 1.0 draft of the Content Security Policy spec is just about to
3742         move to Last Call. We'll hide work on the upcoming 1.1 spec behind
3743         this ENABLE flag, disabled by default.
3744
3745         Spec: https://dvcs.w3.org/hg/content-security-policy/raw-file/tip/csp-specification.dev.html
3746
3747         * Scripts/webkitperl/FeatureList.pm:
3748         * qmake/mkspecs/features/features.pri:
3749
3750 2012-06-19  Christophe Dumez  <christophe.dumez@intel.com>
3751
3752         [EFL] EFL's LayoutTestController does not support setTextDirection
3753         https://bugs.webkit.org/show_bug.cgi?id=87481
3754
3755         Reviewed by Hajime Morita.
3756
3757         Implement setTextDirection in EFL's LayoutTestController and
3758         properly reset its value between test cases to avoid flakiness.
3759
3760         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
3761         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
3762         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
3763         (LayoutTestController::setTextDirection):
3764
3765 2012-06-19  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
3766
3767         [Qt] Temporarily disable xvfb driver in nrwt
3768         https://bugs.webkit.org/show_bug.cgi?id=88414
3769
3770         Reviewed by Csaba Osztrogonác.
3771
3772         * Scripts/webkitpy/layout_tests/port/qt.py:
3773         (QtPort._build_driver):
3774
3775 2012-06-18  Joone Hur  <joone.hur@intel.com>
3776
3777         [EFL] Fontconfig can't be linked properly
3778         https://bugs.webkit.org/show_bug.cgi?id=89418
3779
3780         Unreviewed build fix.
3781
3782         Make Fontconfig be linked correctly.
3783
3784         * MiniBrowser/efl/CMakeLists.txt: add FONTCONFIG_LIBRARIES to MiniBrowser_LIBRARIES.
3785
3786 2012-06-18  Joone Hur  <joone.hur@intel.com>
3787
3788         Unreviewed. Updating email for committers.py script.
3789
3790         * Scripts/webkitpy/common/config/committers.py:
3791
3792 2012-06-18  Simon Fraser  <simon.fraser@apple.com>
3793
3794         Fix a typo that caused TestFailures to have a syntax error.
3795         
3796         Sadly TestFailures is broken for another reason now too:
3797         https://bugs.webkit.org/show_bug.cgi?id=89419
3798
3799         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ViewController.js:
3800         (ViewController.prototype._domForBuildName):
3801
3802 2012-06-18  Simon Fraser  <simon.fraser@apple.com>
3803
3804         Provide bug links for suspicious commits
3805         https://bugs.webkit.org/show_bug.cgi?id=89408
3806         
3807         Reviewed by Adam Barth.
3808
3809         Add links to bugs for the suspicious commits if we have bug data.
3810
3811         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/Bugzilla.js:
3812         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
3813         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/notifications.css:
3814         (ol.notifications>li ul.causes>li>div.description>span>span.bugID::before):
3815         (ol.notifications>li ul.causes>li>div.description>span>span.bugID>a):
3816
3817 2012-06-18  Amy Ousterhout  <aousterh@chromium.org>
3818
3819         [Chromium] DeviceOrientation Cleanup
3820         https://bugs.webkit.org/show_bug.cgi?id=89354
3821
3822         Reviewed by Kent Tamura.
3823
3824         * DumpRenderTree/chromium/WebViewHost.h: added OVERRIDE specifier for virtual function
3825         (WebViewHost):
3826
3827 2012-06-18  James Robinson  <jamesr@chromium.org>
3828
3829         [chromium] Stop passing deprecated 'direct' parameter to webkit_support::CreateGraphicsContext3D
3830         https://bugs.webkit.org/show_bug.cgi?id=89254
3831
3832         Reviewed by Adrienne Walker.
3833
3834         This parameter doesn't mean anything since this codepath is only used for onscreen contexts and is deprecated
3835         upstream.
3836
3837         * DumpRenderTree/chromium/WebViewHost.cpp:
3838         (WebViewHost::createGraphicsContext3D):
3839
3840 2012-06-18  Simon Fraser  <simon.fraser@apple.com>
3841
3842         Minor fix suggested by Adam Barth.
3843
3844         Use $(this._what).empty(); rather than removing children one at a time.
3845         For some reason removeAllChildren() is undefined on this node.
3846         
3847         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
3848
3849 2012-06-18  Simon Fraser  <simon.fraser@apple.com>
3850
3851         Show TOT revision in garden-o-matic, and make the revisions into links
3852         https://bugs.webkit.org/show_bug.cgi?id=89396
3853
3854         Reviewed by Adam Barth.
3855
3856         The "latest revision" notification now reads "Latest revision processed by every bot: NN (trunk is at NN)"
3857         where the revisions are links to trac.
3858         
3859         To avoid setting innerHTML, I made base.createLinkNode() and used it in a bunch of places, which had
3860         the knock-on effect of flipping the target and href attributes for anchors in some test output.
3861
3862         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ViewController.js:
3863         (ViewController.prototype._displayTesters.updateList.list):
3864         (ViewController.prototype._displayTesters.updateList):
3865         (ViewController.prototype._domForRegressionRange.trac.commitDataForRevisionRange):
3866         (ViewController.prototype._domForRegressionRange):
3867         (ViewController.prototype._domForAuxiliaryUIElements):
3868         (ViewController.prototype._domForBuildName):
3869         (ViewController.prototype):
3870         (ViewController.prototype.):
3871         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base.js:
3872         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
3873         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model.js:
3874         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
3875         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
3876
3877 2012-06-18  Csaba Osztrogonác  <ossy@webkit.org>
3878
3879         REGRESSION(r100558): NRWT should work without SVN or GIT
3880         https://bugs.webkit.org/show_bug.cgi?id=76630
3881
3882         Reviewed by Dirk Pranke.
3883
3884         NRWT is not supposed to have a requirement that we have an
3885         actual SVN or Git checkout, and we were unnecessarily
3886         initializing the checkout. This patch removes that line :).
3887
3888         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3889         (main):
3890
3891 2012-06-18  Zan Dobersek  <zandobersek@gmail.com>
3892
3893         [garden-o-matic] Add support for the GTK builders
3894         https://bugs.webkit.org/show_bug.cgi?id=89360
3895
3896         Reviewed by Adam Barth.
3897
3898         Add the 'gtk' platform, essentially a copy of the 'apple' platform since
3899         both get their data from build.webkit.org.
3900
3901         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
3902         (.):
3903
3904 2012-06-18  Nico Weber  <thakis@chromium.org>
3905
3906         [chromium/mac] Unbreak smooth scrolling.
3907         https://bugs.webkit.org/show_bug.cgi?id=89327
3908
3909         Reviewed by Dimitri Glazkov.
3910
3911         Broken by Sam in r115589 / r115591.
3912
3913         * DumpRenderTree/mac/DumpRenderTree.mm:
3914         (resetDefaultsToConsistentValues):
3915         * TestWebKitAPI/mac/InjectedBundleControllerMac.mm:
3916         (TestWebKitAPI::InjectedBundleController::platformInitialize):
3917         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
3918         (WTR::InjectedBundle::platformInitialize):
3919
3920 2012-06-18  Sheriff Bot  <webkit.review.bot@gmail.com>
3921
3922         Unreviewed, rolling out r120541, r120547, and r120548.
3923         http://trac.webkit.org/changeset/120541
3924         http://trac.webkit.org/changeset/120547
3925         http://trac.webkit.org/changeset/120548
3926         https://bugs.webkit.org/show_bug.cgi?id=89383
3927
3928         This approach doesn't work for WebKit2 (Requested by abarth on
3929         #webkit).
3930
3931         * DumpRenderTree/LayoutTestController.cpp:
3932         (setBackingScaleFactorCallback):
3933         (LayoutTestController::staticFunctions):
3934         * DumpRenderTree/LayoutTestController.h:
3935         (LayoutTestController):
3936         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
3937         (LayoutTestController::setBackingScaleFactor):
3938         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
3939         (LayoutTestController::setBackingScaleFactor):
3940         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
3941         (LayoutTestController::setBackingScaleFactor):
3942         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
3943         (LayoutTestController::setBackingScaleFactor):
3944         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
3945         (LayoutTestController::setBackingScaleFactor):
3946         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
3947         (LayoutTestController::setBackingScaleFactor):
3948         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
3949         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
3950         (WTR::InjectedBundle::didReceiveMessage):
3951         (WTR::InjectedBundle::postSetBackingScaleFactor):
3952         (WTR):
3953         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
3954         (InjectedBundle):
3955         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
3956         (WTR::LayoutTestController::setBackingScaleFactor):
3957         (WTR):
3958         (WTR::LayoutTestController::callSetBackingScaleFactorCallback):
3959         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
3960         (LayoutTestController):
3961         * WebKitTestRunner/TestInvocation.cpp:
3962         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle):
3963
3964 2012-06-18  Zan Dobersek  <zandobersek@gmail.com>
3965
3966         REGRESSION(r120546): It made 3 webkitpy tests fail
3967         https://bugs.webkit.org/show_bug.cgi?id=89332
3968
3969         Reviewed by Adam Barth.
3970
3971         Update the BaselineOptimizer unittest after r120546.
3972
3973         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
3974         (BaselineOptimizerTest.test_no_add_mac_future):
3975         (BaselineOptimizerTest.test_mac_future):
3976         (BaselineOptimizerTest.test_complex_shadowing):
3977
3978 2012-06-18  Dirk Pranke  <dpranke@chromium.org>
3979
3980         new-run-webkit-tests output gets confused when logging
3981         https://bugs.webkit.org/show_bug.cgi?id=63793
3982
3983         Reviewed by Adam Barth.
3984
3985         Ship all of the log messages generated in the worker back
3986         to the manager; this allows the messages to be properly
3987         serialized to stderr and for the meter to flush its output correctly.
3988
3989         Note however that this will likely result in messages to be
3990         logged out of order between workers; I'm not sure that there's
3991         an easy way to fix this short of buffering a potentially
3992         unbounded amount of data. It might be better to just log through
3993         stderr unless we're in 'metering' mode, but it's also worth
3994         noting that we already get messages out of order through stderr
3995         on Windows.
3996
3997         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3998         (Manager.handle_done):
3999         (Manager.handle_finished_test):
4000         (Manager._log_messages):
4001         * Scripts/webkitpy/layout_tests/controllers/worker.py:
4002         (Worker.__init__):
4003         (Worker._set_up_logging):
4004         (Worker.run):
4005         (Worker._run_test):
4006         (Worker.cleanup):
4007         (Worker.run_single_test):
4008         (_WorkerLogHandler):
4009         (_WorkerLogHandler.__init__):
4010         (_WorkerLogHandler.emit):
4011         * Scripts/webkitpy/layout_tests/views/metered_stream.py:
4012         (MeteredStream.write):
4013         (MeteredStream.writeln):
4014         * Scripts/webkitpy/layout_tests/views/printing.py:
4015         (Printer.writeln):
4016
4017 2012-06-18  Zan Dobersek  <zandobersek@gmail.com>
4018
4019         [garden-o-matic] Builder names without underscores cause incorrect BuildSelector behavior
4020         https://bugs.webkit.org/show_bug.cgi?id=89362
4021
4022         Reviewed by Simon Fraser.
4023
4024         Replace white spaces, braces and dots in the builder name with underscores. This
4025         way the JQuery tabs will be properly matched with the according container.
4026
4027         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
4028
4029 2012-06-18  Dirk Pranke  <dpranke@chromium.org>
4030
4031         nrwt: metered output doesn't handle ^C cleanly
4032         https://bugs.webkit.org/show_bug.cgi?id=89249
4033
4034         Reviewed by Tony Chang.
4035
4036         If you ctrl-c a running nrwt (w/o --verbose), then all but
4037         two characters of the last update are erased, and then we print
4038         "interrupted, exiting" as an update, which itself then gets
4039         erased. Fix this so that we flush the meter (making the last
4040         update persistent, so that we print the ^C for a visual clue),
4041         print the interrupt message, and move on ...
4042
4043         * Scripts/webkitpy/layout_tests/controllers/manager.py:
4044         (Manager._run_tests):
4045         * Scripts/webkitpy/layout_tests/views/metered_stream.py:
4046         (MeteredStream.flush):
4047         * Scripts/webkitpy/layout_tests/views/printing.py:
4048         (Printer._write):
4049         (Printer):
4050         (Printer.flush):
4051
4052 2012-06-18  Xianzhu Wang  <wangxianzhu@chromium.org>
4053
4054         [Chromium] Change back "linuxish" to "linux" and include WebFontRendering.cpp on Android
4055         https://bugs.webkit.org/show_bug.cgi?id=89228
4056
4057         Reviewed by Tony Chang.
4058
4059         * DumpRenderTree/chromium/LayoutTestController.cpp:
4060         * DumpRenderTree/chromium/TestShellAndroid.cpp:
4061
4062 2012-06-18  Mario Sanchez Prada  <msanchez@igalia.com>
4063
4064         [GTK] Get rid of DumpRenderTreeSupportGtk::{in|de}crementAccessibilityValue
4065         https://bugs.webkit.org/show_bug.cgi?id=89226
4066
4067         Reviewed by Martin Robinson.
4068
4069         Implement increment() and decrement() functions in term of the AtkValue interface,
4070         instead of using DumpRenderTreeSupportGtk helper class.
4071
4072         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
4073         (AccessibilityUIElement::intValue):
4074         (AccessibilityUIElement::minValue):
4075         (AccessibilityUIElement::maxValue):
4076         (alterCurrentValue):
4077         (AccessibilityUIElement::increment):
4078         (AccessibilityUIElement::decrement):
4079
4080 2012-06-17  Simon Fraser  <simon.fraser@apple.com>
4081
4082         garden-o-matic 'Results' panel is broken for the Apple platform
4083         https://bugs.webkit.org/show_bug.cgi?id=89310
4084
4085         Reviewed by Adam Barth.
4086         
4087         For platforms that don't use accumulated build directories on the server,
4088         carry along buildLocation data with buildInfo, so that we know where to
4089         look for the test results files for a given test.
4090         
4091         Renamed historicalResultsSummaryURLs to historicalResultsLocations because
4092         it now returns an array of objects with buildNumber, revision and url info.
4093         
4094         Fixed results.fetchResultsURLs() to use this data to find results.
4095         
4096         Fixed the results display to not scroll when you click on a test to see the results.
4097         Made the h3s look less ugly
4098         
4099         Don't try to show the flakiness dashboard for the Apple platform.
4100
4101         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout.js:
4102         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
4103         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results.js:
4104         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
4105         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
4106         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/results.css:
4107         (.results-view .top-panel):
4108         (.results-view h3):
4109
4110 2012-06-16  Simon Fraser  <simon.fraser@apple.com>
4111
4112         Make garden-o-matic work for the Apple Mac port
4113         https://bugs.webkit.org/show_bug.cgi?id=84642
4114
4115         Reviewed by Adam Barth.
4116         
4117         Make garden-o-matic work for the Apple, webkit.org-hosted builders and testers. This involved
4118         educating the scripts in various ways:
4119         * Wrap up platform-related differences in config.kPlatforms[]
4120         * Add a <select> to switch between platforms, and support a url parameter, ?platform=chromium/apple
4121         * The webkit.org bots don't accumulate test results into a single directory like the chromium.org ones do,
4122           so add config.haveBuilderAccumulatedResults and logic in fetchResultsByBuilder() to find the most
4123           recent build with valid results.
4124         * The webkit.org bots often generate results directories with no layout test data (e.g. when testers
4125           try to test a build that has already been deleted). Make walkHistory() more robust here.
4126         * webkit.org uses differently named test result directories, that include the SVN revision as
4127           well as the build number. That forces us to fetch more build info before we can get the
4128           result directory URL.
4129         * chromium.org serves raw directory listings for a builder's results directories. webkit.org serves
4130           those with Twisted, so rather than scrape directory listings, we use buildbot JSON to find results
4131           dirs.
4132         * Various URLs differ between webkit.org and chromium.org
4133         * Better UI for the failures info, so that some info is visible even when not hovered.
4134
4135         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
4136         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/LayoutTestResultsLoader.js:
4137         (LayoutTestResultsLoader.prototype.set _fetchAndParseNRWTResults):
4138         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders.js:
4139         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
4140         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model.js:
4141         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/net.js:
4142         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results.js:
4143         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results_unittests.js:
4144         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/summary-mock.js:
4145         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
4146         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures.js:
4147         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
4148         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/perf.js:
4149         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:
4150         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/notifications.css:
4151         (ol.notifications>li table.failures):
4152         (ol.notifications>li:hover table.failures):
4153         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/onebar.css:
4154         (#onebar #platform-picker):
4155         * Scripts/webkitpy/layout_tests/port/builders.py:
4156
4157 2012-06-16  Sheriff Bot  <webkit.review.bot@gmail.com>
4158
4159         Unreviewed, rolling out r120536.
4160         http://trac.webkit.org/changeset/120536
4161         https://bugs.webkit.org/show_bug.cgi?id=89296
4162
4163         Does not compile on chromium-linux (Requested by abarth on
4164         #webkit).
4165
4166         * DumpRenderTree/chromium/LayoutTestController.cpp:
4167         (LayoutTestController::reset):
4168         (LayoutTestController::setTextSubpixelPositioning):
4169         * DumpRenderTree/chromium/TestShellAndroid.cpp:
4170
4171 2012-06-16  Adam Barth  <abarth@webkit.org>
4172
4173         layoutTestController.setBackingScaleFactor is redundant with (and less awesome than) internals.settings.setDeviceScaleFactor
4174         https://bugs.webkit.org/show_bug.cgi?id=89274
4175
4176         Reviewed by Levi Weintraub.
4177
4178         Delete (mostly stub) implementations of layoutTestController.setBackingScaleFactor.
4179         Note: The WebKit2 API that setBackingScaleFactor exercises is still
4180         tested by API-level tests.
4181
4182         * DumpRenderTree/LayoutTestController.cpp:
4183         (LayoutTestController::staticFunctions):
4184         * DumpRenderTree/LayoutTestController.h:
4185         (LayoutTestController):
4186         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
4187         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
4188         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
4189         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
4190         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
4191         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
4192         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
4193         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
4194         (WTR::InjectedBundle::didReceiveMessage):
4195         (WTR):
4196         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
4197         (InjectedBundle):
4198         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
4199         (WTR):
4200         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
4201         (LayoutTestController):
4202         * WebKitTestRunner/TestInvocation.cpp:
4203         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle):
4204
4205 2012-06-16  Xianzhu Wang  <wangxianzhu@chromium.org>
4206
4207         [Chromium] Move chromium/public/linuxish/WebFontRendering.h out of linuxish directory
4208         https://bugs.webkit.org/show_bug.cgi?id=89228
4209
4210         Reviewed by Tony Chang.
4211
4212         * DumpRenderTree/chromium/LayoutTestController.cpp:
4213         (LayoutTestController::reset):
4214         (LayoutTestController::setTextSubpixelPositioning):
4215         * DumpRenderTree/chromium/TestShellAndroid.cpp:
4216
4217 2012-06-15  Adam Barth  <abarth@webkit.org>
4218
4219         garden-o-matic's results.js should use RequestTracker
4220         https://bugs.webkit.org/show_bug.cgi?id=89257
4221
4222         Reviewed by Dimitri Glazkov.
4223
4224         We wrote results.js before we recognized the RequestTracker pattern.
4225         This patch replaces the manual implementations of RequestTracker with
4226         the real deal.
4227
4228         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results.js:
4229
4230 2012-06-15  Darin Adler  <darin@apple.com>
4231
4232         * Scripts/webkitpy/bindings: Added property svn:ignore.
4233
4234 2012-06-15  Sheriff Bot  <webkit.review.bot@gmail.com>
4235
4236         Unreviewed, rolling out r120511.
4237         http://trac.webkit.org/changeset/120511
4238         https://bugs.webkit.org/show_bug.cgi?id=89255
4239
4240         Breaks at least Android builder (Requested by wangxianzhu on
4241         #webkit).
4242
4243         * DumpRenderTree/chromium/LayoutTestController.cpp:
4244         (LayoutTestController::reset):
4245         (LayoutTestController::setTextSubpixelPositioning):
4246
4247 2012-06-15  Xianzhu Wang  <wangxianzhu@chromium.org>
4248
4249         [Chromium] Move chromium/public/linuxish/WebFontRendering.h out of linuxish directory
4250         https://bugs.webkit.org/show_bug.cgi?id=89228
4251
4252         Reviewed by Tony Chang.
4253
4254         * DumpRenderTree/chromium/LayoutTestController.cpp:
4255         (LayoutTestController::reset):
4256         (LayoutTestController::setTextSubpixelPositioning):
4257
4258 2012-06-15  Dirk Pranke  <dpranke@chromium.org>
4259
4260         webkitpy: remove DummyOptions and clean up the code in Port.get_option() and Port.set_option_default()
4261         https://bugs.webkit.org/show_bug.cgi?id=89135
4262
4263         Re-land change in r120370 with fix in
4264         PortFactory.get_from_builder_name() that changes BuilderOptions
4265         to an actual optparse.Values object.
4266
4267         * Scripts/webkitpy/layout_tests/port/base.py:
4268         (Port.__init__):
4269         (Port.get_option):
4270         (Port.set_option_default):
4271         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
4272         (ChromiumWinTest.test_setup_environ_for_server_register_cygwin):
4273         * Scripts/webkitpy/layout_tests/port/factory.py:
4274         (_builder_options):
4275         (PortFactory.get_from_builder_name):
4276         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
4277         (FactoryTest.test_get_from_builder_name):
4278         * Scripts/webkitpy/style/checkers/test_expectations.py:
4279         (TestExpectationsChecker._determine_port_from_expectations_path):
4280         * Scripts/webkitpy/tool/mocktool.py:
4281         (MockOptions.ensure_value):
4282
4283 2012-06-15  Bill Budge  <bbudge@chromium.org>
4284
4285         Add bbudge@chromium.org to committers.py
4286
4287         Unreviewed.
4288
4289         * Scripts/webkitpy/common/config/committers.py:
4290
4291 2012-06-15  Csaba Osztrogonác  <ossy@webkit.org>
4292
4293         Unreviewed style fix after r120351.
4294
4295         * BuildSlaveSupport/build.webkit.org-config/public_html/default.css:
4296         (body.interface):
4297
4298 2012-06-15  Csaba Osztrogonác  <ossy@webkit.org>
4299
4300         master.cfg cleanup, remove unnecessary workaround
4301         https://bugs.webkit.org/show_bug.cgi?id=88994
4302
4303         Reviewed by Lucas Forschler.
4304
4305         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
4306         (loadBuilderConfig):
4307
4308 2012-06-15  Dirk Pranke  <dpranke@chromium.org>
4309
4310         webkit-patch rebaseline-expectations should only rebaseline the appropriate suffixes for the failure in question
4311         https://bugs.webkit.org/show_bug.cgi?id=88581
4312
4313         Reviewed by Adam Barth.
4314
4315         Make sure we only optimize the suffixes we rebaselined during
4316         rebaseline-expectations, and not all suffixes for a test.
4317         While optimizing is somewhere between harmless and good, it's also confusing :)
4318
4319         * Scripts/webkitpy/tool/commands/rebaseline.py:
4320         (RebaselineExpectations._rebaseline_port):
4321         (RebaselineExpectations.execute):
4322         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
4323
4324 2012-06-15  Csaba Osztrogonác  <ossy@webkit.org>
4325
4326         Update builder.html template for newer buildmaster
4327         https://bugs.webkit.org/show_bug.cgi?id=89207
4328
4329         Rebasing builder.html template from v0.8.3 to v0.8.6p1.
4330
4331         Reviewed by Zoltan Herczeg.
4332
4333         * BuildSlaveSupport/build.webkit.org-config/templates/builder.html:
4334