REGRESSION (r129211-r129218): http/tests/loading/redirect-with-no-location-crash...
[WebKit-https.git] / Tools / ChangeLog
1 2012-09-24  Zan Dobersek  <zandobersek@gmail.com>
2
3         REGRESSION (r129211-r129218): http/tests/loading/redirect-with-no-location-crash.html failing on GTK Linux 64-bit Release WK2 (Tests)
4         https://bugs.webkit.org/show_bug.cgi?id=97404
5
6         Reviewed by Philippe Normand.
7
8         Print out the 'didReceiveTitle' information when dumping frame load callbacks.
9
10         * DumpRenderTree/gtk/DumpRenderTree.cpp:
11         (webViewTitleChanged):
12
13 2012-09-24  Simon Hausmann  <simon.hausmann@digia.com>
14
15         [Qt] Fix build with latest Qt 5
16
17         Reviewed by Tor Arne Vestbø.
18
19         The variables for syncqt handling are now set in api.pri, but let's add a comment here that we can
20         remove them from default_pre.prf once we're past Qt 5 beta2.
21
22         * qmake/mkspecs/features/default_pre.prf:
23
24 2012-09-22  Dominic Mazzoni  <dmazzoni@google.com>
25
26         AX: Layout tests would be easier to write if AccessibilityController could find an element by id
27         https://bugs.webkit.org/show_bug.cgi?id=97301
28
29         Reviewed by Chris Fleizach.
30
31         Adds a new accessibleElementById method to AccessibilityController, to make it
32         easier to write cross-platform layout tests.
33
34         Implemented for Chromium and for Mac (DRT and WKTR).
35
36         * DumpRenderTree/AccessibilityController.cpp:
37         (getAccessibleElementByIdCallback):
38         (AccessibilityController::getJSClass):
39         * DumpRenderTree/AccessibilityController.h:
40         (AccessibilityController):
41         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
42         * DumpRenderTree/chromium/TestRunner/AccessibilityControllerChromium.cpp:
43         (AccessibilityController::AccessibilityController):
44         (AccessibilityController::findAccessibleElementByIdRecursive):
45         (AccessibilityController::getAccessibleElementById):
46         (AccessibilityController::accessibleElementByIdGetterCallback):
47         * DumpRenderTree/chromium/TestRunner/AccessibilityControllerChromium.h:
48         (AccessibilityController):
49         * DumpRenderTree/gtk/AccessibilityControllerGtk.cpp:
50         (AccessibilityController::accessibleElementById):
51         * DumpRenderTree/mac/AccessibilityCommonMac.h: added
52         * DumpRenderTree/mac/AccessibilityCommonMac.mm: added
53         (+[NSString stringWithJSStringRef:]):
54         (-[NSString createJSStringRef]):
55         * DumpRenderTree/mac/AccessibilityControllerMac.mm:
56         (findAccessibleObjectById):
57         (AccessibilityController::accessibleElementById):
58         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
59         * DumpRenderTree/win/AccessibilityControllerWin.cpp:
60         (AccessibilityController::accessibleElementById):
61         * WebKitTestRunner/InjectedBundle/AccessibilityController.cpp:
62         (WTR::AccessibilityController::accessibleElementById):
63         * WebKitTestRunner/InjectedBundle/AccessibilityController.h:
64         (AccessibilityController):
65         * WebKitTestRunner/InjectedBundle/Bindings/AccessibilityController.idl:
66         * WebKitTestRunner/InjectedBundle/mac/AccessibilityCommonMac.h: added
67         * WebKitTestRunner/InjectedBundle/mac/AccessibilityCommonMac.mm: added
68         (+[NSString stringWithJSStringRef:]):
69         (-[NSString createJSStringRef]):
70         * WebKitTestRunner/InjectedBundle/mac/AccessibilityControllerMac.mm:
71         (WTR::findAccessibleObjectById):
72         (WTR):
73         (WTR::AccessibilityController::accessibleElementById):
74         * WebKitTestRunner/InjectedBundle/mac/AccessibilityUIElementMac.mm:
75         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
76
77 2012-09-22  Zan Dobersek  <zandobersek@gmail.com>
78
79         Unexpected reftest passes are only reported when pixel testing is enabled
80         https://bugs.webkit.org/show_bug.cgi?id=97242
81
82         Reviewed by Dirk Pranke.
83
84         LayoutTestRunner should know whether the finished test it's handling is
85         a reftest. This is necessary when updating result summary for reftests
86         when pixel testing is not enabled - in these circumstances an unexpectedly
87         passing reftest was not reported as such due to pixel testing then being
88         determined by searching for reftest failures in failures list, which were
89         not found in an empty failure list of a passing test. Now the TestResult
90         objects have a member variable indicating the test is a reftest whenever
91         that's the case.
92
93         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
94         (LayoutTestRunner._update_summary_with_result):
95         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner_unittest.py:
96         (LayoutTestRunnerTests.test_update_summary_with_result):
97         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
98         (SingleTestRunner.run):
99         (SingleTestRunner._run_reftest):
100         * Scripts/webkitpy/layout_tests/models/test_results.py:
101         (TestResult.__init__):
102
103 2012-09-21  Sam Weinig  <sam@webkit.org>
104
105         WebProcess XPC services need have their environment set without disrupting all other XPC services
106         https://bugs.webkit.org/show_bug.cgi?id=95161
107
108         Reviewed by Anders Carlsson.
109
110         Remove setting the __XPC_* environment variables.
111
112         * Scripts/webkitdirs.pm:
113         (setUpGuardMallocIfNeeded):
114         (runMacWebKitApp):
115         (execMacWebKitAppForDebugging):
116
117 2012-09-21  Dirk Pranke  <dpranke@chromium.org>
118
119         Fix typo in additional-platform-directory patch just landed
120         https://bugs.webkit.org/show_bug.cgi?id=97380
121         
122         Unreviewed, build fix.
123
124         * Scripts/webkitpy/layout_tests/port/base.py:
125         (Port.relative_test_filename):
126         (Port.relative_perf_test_filename):
127
128 2012-09-21  Dirk Pranke  <dpranke@chromium.org>
129
130         nrwt: don't require additional-platform-directory to be an abspath or live under LayoutTests
131         https://bugs.webkit.org/show_bug.cgi?id=97380
132
133         Reviewed by Ojan Vafai.
134
135         There doesn't seem to be a good reason for this restriction and
136         it's useful to be able to point to directories outside the
137         checkout for results (e.g., for local failures due to a 10.7.4
138         install ;).
139
140         * Scripts/webkitpy/layout_tests/port/base.py:
141         (Port.relative_test_filename):
142         (Port.relative_perf_test_filename):
143         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
144         (ChromiumAndroidDriver._command_from_driver_input):
145         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
146         (_set_up_derived_options):
147         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
148         (MainTest.test_additional_platform_directory):
149
150 2012-09-21  Dirk Pranke  <dpranke@chromium.org>
151
152         webkitpy: drop support for old TestExpectations syntax
153         https://bugs.webkit.org/show_bug.cgi?id=97364
154
155         Reviewed by Ryosuke Niwa.
156
157         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
158         (TestExpectationParser._collect_matching_tests):
159         (TestExpectationParser):
160         (TestExpectationParser._tokenize_line):
161         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
162         (SkippedTests.test_skipped_entry_dont_exist):
163         (ExpectationSyntaxTests.assert_tokenize_exp):
164         (ExpectationSyntaxTests.test_bare_name):
165         (ExpectationSyntaxTests.test_bare_name_and_bugs):
166         (ExpectationSyntaxTests.test_comments):
167         (ExpectationSyntaxTests.test_config_modifiers):
168         (ExpectationSyntaxTests.test_unknown_config):
169         (ExpectationSyntaxTests.test_unknown_expectation):
170         (ExpectationSyntaxTests.test_skip):
171         (ExpectationSyntaxTests.test_slow):
172         (ExpectationSyntaxTests.test_wontfix):
173         (ExpectationSyntaxTests.test_blank_line):
174         (ExpectationSyntaxTests.test_warnings):
175         (RebaseliningTest.test_no_get_rebaselining_failures):
176
177 2012-09-21  Simon Fraser  <simon.fraser@apple.com>
178
179         Improve WTR unresponsiveness output a little
180         https://bugs.webkit.org/show_bug.cgi?id=97370
181
182         Reviewed by Timothy Horton.
183
184         Distinguish between conditions that already set the errorMessage,
185         and unresponsiveness due to slow about:blank loads when WTR
186         reports unresponsiveness.
187
188         * WebKitTestRunner/TestInvocation.cpp:
189         (WTR::TestInvocation::invoke):
190
191 2012-09-21  Chris Rogers  <crogers@google.com>
192
193         Add Web Audio support for deprecated/legacy APIs
194         https://bugs.webkit.org/show_bug.cgi?id=97050
195
196         Reviewed by Eric Carlson.
197
198         * Scripts/webkitperl/FeatureList.pm:
199
200 2012-09-21  Dirk Pranke  <dpranke@chromium.org>
201
202         webkitpy: update remaining tests to use the new expectation syntax
203         https://bugs.webkit.org/show_bug.cgi?id=97362
204
205         Reviewed by Ojan Vafai.
206
207         This patch updates all the unit tests that were still using the
208         old TestExpectations syntax to use the new syntax *except* for
209         the tests that were specifically testing that we parsed the old
210         syntax correctly.
211
212         Also, a block of tests for the new syntax were duplicated, so
213         I've deleted the duplicate.
214
215         Note that the old syntax is still supported so this change should
216         produce no visible changes.
217
218         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
219         (MiscTests.test_multiple_results):
220         (MiscTests.test_category_expectations):
221         (MiscTests.test_error_on_different_platform):
222         (MiscTests.test_error_on_different_build_type):
223         (MiscTests.test_overrides):
224         (MiscTests.test_overrides__directory):
225         (MiscTests.test_overrides__duplicate):
226         (MiscTests.test_more_specific_override_resets_skip):
227         (SkippedTests.check):
228         (SkippedTests.test_duplicate_skipped_test_fails_lint):
229         (SkippedTests.test_skipped_file_overrides_expectations):
230         (SkippedTests.test_skipped_dir_overrides_expectations):
231         (SkippedTests.test_skipped_file_overrides_overrides):
232         (SkippedTests.test_skipped_dir_overrides_overrides):
233         (ExpectationSyntaxTests.disabled_test_missing_expectation):
234         (ExpectationSyntaxTests.disabled_test_missing_colon):
235         (ExpectationSyntaxTests.disabled_test_too_many_colons):
236         (ExpectationSyntaxTests.disabled_test_too_many_equals_signs):
237         (ExpectationSyntaxTests):
238         (ExpectationSyntaxTests.test_unrecognized_expectation):
239         (ExpectationSyntaxTests.test_macro):
240         (SemanticTests.test_bug_format):
241         (SemanticTests.test_bad_bugid):
242         (SemanticTests.test_missing_bugid):
243         (SemanticTests.test_slow_and_timeout):
244         (SemanticTests.test_rebaseline):
245         (test_missing_file):
246         (test_ambiguous):
247         (test_more_modifiers):
248         (test_order_in_file):
249         (test_macro_overrides):
250         (OldExpectationParserTests):
251         (OldExpectationParserTests._tokenize):
252         (OldExpectationParserTests.test_tokenize_extra_colon):
253         (OldExpectationParserTests.test_tokenize_missing_equal):
254         (OldExpectationParserTests.test_tokenize_extra_equal):
255         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
256         (MainTest.test_skip_failing_tests):
257         (MainTest.test_additional_expectations):
258         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
259         (TestExpectationsTestCase.test_valid_expectations):
260         (TestExpectationsTestCase.test_invalid_expectations):
261         (TestExpectationsTestCase.test_tab):
262
263 2012-09-21  Benjamin Poulain  <bpoulain@apple.com>
264
265         [WK2] Add basic testing support for Geolocation
266         https://bugs.webkit.org/show_bug.cgi?id=97278
267
268         Reviewed by Kenneth Rohde Christiansen.
269
270         Update the Geolocation testing to use the proper API in the UIProcess.
271
272         * WebKitTestRunner/CMakeLists.txt:
273         * WebKitTestRunner/GNUmakefile.am:
274
275         * WebKitTestRunner/GeolocationProviderMock.cpp: Added.
276         (WTR::startUpdatingCallback):
277         (WTR::stopUpdatingCallback):
278         (WTR::GeolocationProviderMock::GeolocationProvierMock):
279         (WTR::GeolocationProviderMock::setMockGeolocationPosition):
280         (WTR::GeolocationProviderMock::startUpdating):
281         (WTR::GeolocationProviderMock::stopUpdating):
282         (GeolocationProviderMock):
283         The GeolocationProvider store the location update and deliver them as needed.
284
285         WebCore GeolocationController do not support asynchronous update on start/stop. This is not
286         a problem in this case because all the messages between the WebProcess and the UIProcess are
287         asynchronous. Because of this, unlike GeolocationClientMock, we do not use a timer for event
288         delivery.
289
290         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
291         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
292         (WTR::InjectedBundle::setGeolocationPermission):
293         (WTR::InjectedBundle::setMockGeolocationPosition):
294         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
295         (InjectedBundle):
296         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
297         (WTR::TestRunner::setGeolocationPermission):
298         (WTR::TestRunner::setMockGeolocationPosition):
299         From the InjectedBundle, we now pass the information to the UIProcess so that
300         GeolocationProvider and the TestController can respond appropriately.
301
302         * WebKitTestRunner/InjectedBundle/TestRunner.h:
303         (TestRunner):
304         * WebKitTestRunner/Target.pri:
305         * WebKitTestRunner/TestController.cpp:
306         (WTR::TestController::TestController):
307         (WTR::decidePolicyForGeolocationPermissionRequest):
308         (WTR::TestController::createOtherPage):
309         (WTR::TestController::initialize):
310         (WTR::TestController::setMockGeolocationPosition):
311         * WebKitTestRunner/TestController.h:
312         (TestController):
313         (WTR::TestController::setGeolocationPermission):
314         (WTR::TestController::isGeolocationPermissionAllowed):
315         * WebKitTestRunner/TestInvocation.cpp:
316         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle):
317         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
318         * WebKitTestRunner/win/WebKitTestRunner.vcproj:
319
320 2012-09-21  Joanmarie Diggs  <jdiggs@igalia.com>
321
322         REGRESSION (r127882): accessibility/spinbutton-value.html failing on GTK
323         https://bugs.webkit.org/show_bug.cgi?id=96196
324
325         Reviewed by Martin Robinson.
326
327         The "regression" is that a new test was added but the support was missing
328         in the Gtk port for spin buttons.
329
330         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
331         (AccessibilityUIElement::valueDescription): Updated the FIXME comment to
332         indicate that this cannot be implemented until it is implemented in ATK.
333         URL of the newly-filed ATK bug included for reference.
334
335 2012-09-21  Dirk Pranke  <dpranke@chromium.org>
336
337         Fix test_skip_and_wontfix failure
338         https://bugs.webkit.org/show_bug.cgi?id=97225
339
340         Unreviewed, build fix.
341
342         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
343         (SemanticTests.test_skip_and_wontfix):
344
345 2012-09-21  Carlos Garcia Campos  <cgarcia@igalia.com>
346
347         [GTK] Implement ViewState methods in PageClientImpl in WebKit2
348         https://bugs.webkit.org/show_bug.cgi?id=97202
349
350         Reviewed by Martin Robinson.
351
352         * WebKitTestRunner/gtk/PlatformWebViewGtk.cpp:
353         (WTR::PlatformWebView::focus): Focus the view.
354
355 2012-09-21  Sudarsana Nagineni  <sudarsana.nagineni@intel.com>
356
357         [WTR] Memory leaks in InjectedBundleController::initialize()
358         https://bugs.webkit.org/show_bug.cgi?id=97329
359
360         Reviewed by Alexey Proskuryakov.
361
362         Fix leaks in InjectedBundleController::initialize() by adopting
363         strings created with WKStringCreateWithUTF8CString().
364
365         * TestWebKitAPI/InjectedBundleController.cpp:
366         (TestWebKitAPI::InjectedBundleController::initialize):
367
368 2012-09-21  Marcelo Lira  <marcelo.lira@openbossa.org>
369
370         [Qt][WK2] Removed duplicated code from EventSenderProxy::keyDown
371         https://bugs.webkit.org/show_bug.cgi?id=97235
372
373         Reviewed by Luiz Agostini.
374
375         * WebKitTestRunner/qt/EventSenderProxyQt.cpp:
376         (WTR::EventSenderProxy::keyDown):
377
378 2012-09-21  Gergely Kis  <gergely@homejinni.com>
379
380         Add MIPS build slave to build.webkit.org
381         https://bugs.webkit.org/show_bug.cgi?id=96713
382
383         Reviewed by Csaba Osztrogonác.
384
385         Added a build slave for MIPS, and enabled a builder for 
386         Qt Linux MIPS32R2 little-endian release build.
387
388         * BuildSlaveSupport/build.webkit.org-config/config.json:
389
390 2012-09-21  Peter Beverloo  <peter@chromium.org>
391
392         [Chromium] Switch back to a fixed fifo path for Android
393         https://bugs.webkit.org/show_bug.cgi?id=97230
394
395         Reviewed by Tony Chang.
396
397         Because not all external storage cards will be formated using a file
398         system that supports named pipes, Chromium has been changed to creating
399         the pipes in a temporary folder on the internal storage. Adapt this in
400         WebKit so we can continue to run layout tests.
401
402         This also includes an *unreviewed* fix for a breakage in the webkitpy
403         tests I made in r129221. Two lines and related to this code, so I decided
404         to include it in this change.
405
406         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
407         (ChromiumAndroidDriver.__init__):
408         (ChromiumAndroidDriver._setup_test):
409         (ChromiumAndroidDriver._update_version):
410
411 2012-09-20  Peter Beverloo  <peter@chromium.org>
412
413         Leverage Chromium's code to set up FIFOs for Chromium Android layout tests
414         https://bugs.webkit.org/show_bug.cgi?id=97227
415
416         Reviewed by Tony Chang.
417
418         We switched Chromium to using FIFOs in order to achieve better consistency,
419         which was done by Marcus in r157541. Remove all custom WebKit code in
420         favor of Chromium's implementation.
421
422         Remove more FIFO code in the test runner itself, including the code in
423         DumpRenderTree that invoked it. We can now switch to Chromium's brand
424         new FIFO-creating code, which is being set-up for all test targets build
425         for Android, including DumpRenderTree, TestWebKitAPI and webkit_unit_tests.
426
427         This also changes the ChromiumAndroidDriver._remove_all_pipes method to
428         delete the files individually. "rm" would fail if one of the earlier files
429         does not exist, and the "-f" argument doesn't seem to be reliable.
430
431         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
432         * DumpRenderTree/chromium/TestShellAndroid.cpp:
433         (platformInit):
434         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
435         (ChromiumAndroidDriver.__init__):
436         (ChromiumAndroidDriver._setup_test):
437         (ChromiumAndroidDriver._get_external_storage):
438         (ChromiumAndroidDriver._drt_cmd_line):
439         (ChromiumAndroidDriver._remove_all_pipes):
440         (ChromiumAndroidDriver.stop):
441         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
442
443 2012-09-21  Christophe Dumez  <christophe.dumez@intel.com>
444
445         [WK2][WKTR] EventSender needs to implement scheduleAsynchronousClick
446         https://bugs.webkit.org/show_bug.cgi?id=97326
447
448         Reviewed by Kenneth Rohde Christiansen.
449
450         Implement scheduleAsynchronousClick() in WebKitTestRunner's
451         EventSender by sending a "MouseDown" and a "MouseUp" message
452         asynchronously to the WebProcess.
453
454         * WebKitTestRunner/InjectedBundle/Bindings/EventSendingController.idl:
455         * WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
456         (WTR::createMouseMessageBody):
457         (WTR):
458         (WTR::EventSendingController::mouseDown):
459         (WTR::EventSendingController::mouseUp):
460         (WTR::EventSendingController::scheduleAsynchronousClick):
461         * WebKitTestRunner/InjectedBundle/EventSendingController.h:
462         (EventSendingController):
463         * WebKitTestRunner/TestController.cpp:
464         (WTR::TestController::didReceiveMessageFromInjectedBundle):
465
466 2012-09-21  Simon Hausmann  <simon.hausmann@digia.com>
467
468         [Qt] Error out early if we don't have ICU available
469
470         Reviewed by Tor Arne Vestbø.
471
472         * qmake/mkspecs/features/configure.prf:
473
474 2012-09-21  Simon Hausmann  <simon.hausmann@digia.com>
475
476         [Qt] Bail out when trying to build WebKit with Qt != 5
477
478         Reviewed by Tor Arne Vestbø.
479
480         Moved check for Qt version out of default_pre into top-level WebKit.pro,
481         because we never reach default_pre.prf due to the lack of .qmake.conf support
482         in older versions of Qt/QMake.
483
484         * qmake/mkspecs/features/default_pre.prf:
485
486 2012-09-21  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
487
488         WebKitTestRunner needs to print frame load delegate information
489         https://bugs.webkit.org/show_bug.cgi?id=42705
490
491         Reviewed by Kenneth Rohde Christiansen.
492
493         Added missing dumping from WTR::InjectedBundlePage::didFailLoadWithErrorForFrame.
494
495         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
496         (WTR::InjectedBundlePage::didFailLoadWithErrorForFrame):
497
498 2012-09-21  Christophe Dumez  <christophe.dumez@intel.com>
499
500         [EFL] EventSender should mimic CTRL+o emacs shortcut
501         https://bugs.webkit.org/show_bug.cgi?id=97224
502
503         Reviewed by Kenneth Rohde Christiansen.
504
505         Add support for mimicking CTRL+o emacs shortcut in EFL's
506         EventSender in DumpRenderTree and WebKitTestRunner.
507
508         * DumpRenderTree/efl/EventSender.cpp:
509         (sendKeyDown):
510         * WebKitTestRunner/efl/EventSenderProxyEfl.cpp:
511         (WTR::EventSenderProxy::keyDown):
512
513 2012-09-21  Simon Hausmann  <simon.hausmann@digia.com>
514
515         [Qt] Re-fix clean builds
516
517         Reviewed by Tor Arne Vestbø.
518
519         Re-introduce the sanitization for LIBS when creating a module to use
520         LIBS_PRIVATE and otherwise do _not_ use LIBS_PRIVATE. We decided to
521         continue to use QT, LIBS and PKGCONFIG instead of their _PRIVATE variants
522         throughout the code base, so just using LIBS_PRIVATE in linkAgainstLibrary()
523         causes build issues when depending system libraries end up in LIBS before
524         LIBS_PRIVATE.
525
526         * qmake/mkspecs/features/default_post.prf:
527         * qmake/mkspecs/features/functions.prf:
528
529 2012-09-20  Simon Fraser  <simon.fraser@apple.com>
530
531         Comment out a failing webkitpy unit test until Dirk can fix it.
532
533         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
534         (SemanticTests.test_skip_and_wontfix):
535
536 2012-09-20  Dirk Pranke  <dpranke@chromium.org>
537
538         make Skip, WontFix be the only expectations on a line
539         https://bugs.webkit.org/show_bug.cgi?id=97225
540
541         Reviewed by Ojan Vafai.
542
543         It is now incorrect in the new syntax to have a line like:
544
545             foo.html [ WontFix Crash ]
546
547         This will generate a lint warning and be treated as an invalid
548         line. Fixing this caused a whole bunch of unit tests to need updating
549         to no longer be marked as WontFix :). Also, this patch adjusts
550         the warnings so that missing Bug() identifiers will cause lint
551         warnings but will *not* cause the line to be treated as invalid.
552         Fixing these issues also revealed that test_hung_thread was no
553         longer testing the right logic, so I adjusted the timeouts in
554         test.py to make that test work again.
555
556         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
557         (Worker._run_test_in_another_thread):
558         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
559         (ResultSummaryTest.test_summarized_results_wontfix):
560         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
561         (TestExpectationParser._parse_modifiers):
562         (TestExpectationParser._tokenize_line_using_new_format):
563         (TestExpectationLine.is_invalid):
564         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
565         (BasicTests.test_basic):
566         (test_get_test_set):
567         (test_parse_warning):
568         (test_pixel_tests_flag):
569         (SemanticTests.test_missing_bugid):
570         (SemanticTests):
571         (SemanticTests.test_skip_and_wontfix):
572         * Scripts/webkitpy/layout_tests/port/test.py:
573         (TestDriver.run_test):
574         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
575         (MainTest.test_hung_thread):
576         * Scripts/webkitpy/tool/commands/queries_unittest.py:
577         (PrintExpectationsTest.test_basic):
578         (PrintExpectationsTest.test_multiple):
579         (PrintExpectationsTest.test_full):
580         (PrintExpectationsTest.test_exclude):
581         (PrintExpectationsTest.test_csv):
582
583 2012-09-20  Lucas Forschler  <lforschler@apple.com>
584
585         Unreviewed.  Start running tests on the mac-ews.
586         
587         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
588         (MacEWS):
589
590 2012-09-20  Tony Chang  <tony@chromium.org>
591
592         [chromium] Enable more clang warnings
593         https://bugs.webkit.org/show_bug.cgi?id=97257
594
595         Reviewed by James Robinson.
596
597         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp: Add -Wunused-parameter. Don't add -Wglobal-constructors because
598         there are lots of violations in these files.
599
600 2012-09-19  Ryosuke Niwa  <rniwa@webkit.org>
601
602         run-perf-tests should record individual value instead of statistics
603         https://bugs.webkit.org/show_bug.cgi?id=97155
604
605         Reviewed by Hajime Morita.
606
607         Parse the list of individual value reported by tests and include them as "values".
608         We strip "values" from the output JSON when uploading it to the perf-o-matic
609         since it doesn't know how to parse "values" or ignore it.
610
611         * Scripts/webkitpy/performance_tests/perftest.py:
612         (PerfTest):
613         (PerfTest.parse_output): Parse and report "values".
614         (PageLoadingPerfTest.run): Report indivisual page loading time in "values".
615         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
616         (MainTest.test_parse_output):
617         (MainTest.test_parse_output_with_failing_line):
618         (TestPageLoadingPerfTest.test_run):
619         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
620         (PerfTestsRunner._generate_and_show_results): Strip "values" from each result
621         until we update perf-o-matic.
622         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
623         (test_run_memory_test):
624         (test_run_with_json_output):
625         (test_run_with_description):
626         (test_run_with_slave_config_json):
627         (test_run_with_multiple_repositories):
628
629 2012-09-20  Simon Hausmann  <simon.hausmann@digia.com>
630
631         [Qt] Fix production builds
632
633         Unreviewed trivial fix: Follow up r129119 and avoid clobbering LIBS through the use of LIBS_PRIVATE.
634
635         Otherwise the libQtWebKit.prl file for example contains -lWebKit1 -lWebCore, etc.
636
637         * qmake/mkspecs/features/functions.prf:
638
639 2012-09-20  Dirk Pranke  <dpranke@chromium.org>
640
641         REGRESSION: layout test results doesn't show diffs
642         https://bugs.webkit.org/show_bug.cgi?id=97182
643
644         Reviewed by Ojan Vafai.
645
646         Go back to storing TEXT, AUDIO, and IMAGE+TEXT in results.json
647         so that results.html (and hopefully garden-o-matic) can
648         determine which things actually failed. However, we keep mapping
649         these results to Failure so that we still only have a single
650         expectation type for them.
651
652         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
653         (JSONLayoutResultsGenerator):
654         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
655         (TestExpectationParser):
656         (TestExpectations):
657         (TestExpectations.result_was_expected):
658         * Scripts/webkitpy/layout_tests/models/test_failures.py:
659         (determine_result_type):
660         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
661         (MainTest.test_missing_and_unexpected_results):
662         (MainTest.test_retrying_and_flaky_tests):
663
664 2012-09-20  Tommy Widenflycht  <tommyw@google.com>
665
666         MediaStream API: Extend UserMediaRequest with a ownerDocument method
667         https://bugs.webkit.org/show_bug.cgi?id=97095
668
669         Reviewed by Adam Barth.
670
671         Extending WebUserMediaClientMock to check that the owning document is valid,
672         and that the document has a frame.
673
674         * DumpRenderTree/chromium/WebUserMediaClientMock.cpp:
675         (WebKit::WebUserMediaClientMock::requestUserMedia):
676
677 2012-09-20  Stephen Chenney  <schenney@chromium.org>
678
679         [Chromium] DRT does not support --dump-all-pixels flag
680         https://bugs.webkit.org/show_bug.cgi?id=95098
681
682         Reviewed by Dirk Pranke.
683
684         Add support for the --pixel-tests and shorthand -p option in Chromium DumpRenderTree. Use
685         of this flag causes pixel results to be created for all tests, regardless of
686         individual test options. If an individual test provides a pixel hash it will be used,
687         otherwise the hash will be empty. This replaces a previously defined but unused option
688         --dump-all-pixels, and is useful primarily when debugging DRT instances.
689
690         * DumpRenderTree/chromium/DumpRenderTree.cpp:
691         (runTest): Add a parameter and code to force pixel results for the test.
692         (main): Add parameter handling for --pixels-test and -p, and remove --dump-all-pixels.
693
694 2012-09-20  Carlos Garcia Campos  <cgarcia@igalia.com>
695
696         [GTK] run-api-tests should not buffer test stdout
697         https://bugs.webkit.org/show_bug.cgi?id=88474
698
699         Reviewed by Philippe Normand.
700
701         * Scripts/run-gtk-tests:
702         (TestRunner._run_test_command): Use os.forkpty() instead of
703         subprocess.Popen() so that gtest sends the output with colors to
704         stdout. Use common.parse_output_lines() to parse the output and
705         write it to stdout while it's read.
706         (TestRunner._run_test_command.parse_line): Parse the line to get
707         the test pid and write the line to stdout.
708         (TestRunner._run_test_command.waitpid): Helper function to call
709         waitpid handling EINTR.
710         (TestRunner._run_test_command.return_code_from_exit_status):
711         Helper function to convert exit status of test commands to a
712         return code.
713         * gtk/common.py:
714         (parse_output_lines): Helper function that uses select to read
715         the given file descriptor and call the given callback for every
716         line read.
717
718 2012-09-20  Simon Hausmann  <simon.hausmann@digia.com>
719
720         [Qt] Fix initial build
721
722         Reviewed by Tor Arne Vestbø.
723
724         When building QtWebKit the first time there is no qt_webkit.pri module pri file, and therefore
725         $$QT.webkit.name isn't set and so creating_module isn't set. That has all sorts of implications
726         causing incorrect linking for Makefile.api, etc.
727
728         Fix the determination by simply checking if MODULE is set, which only happens in api.pri.
729
730         * qmake/mkspecs/features/webkit_modules.prf:
731
732 2012-09-20  Sheriff Bot  <webkit.review.bot@gmail.com>
733
734         Unreviewed, rolling out r129091.
735         http://trac.webkit.org/changeset/129091
736         https://bugs.webkit.org/show_bug.cgi?id=97205
737
738         It broke perf tests everywhere (Requested by Ossy on #webkit).
739
740         * Scripts/webkitpy/performance_tests/perftest.py:
741         (PerfTest):
742         (PerfTest.parse_output):
743         (PageLoadingPerfTest.run):
744         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
745         (MainTest.test_parse_output):
746         (MainTest.test_parse_output_with_failing_line):
747         (TestPageLoadingPerfTest.test_run):
748         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
749         (PerfTestsRunner._generate_and_show_results):
750         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
751         (test_run_memory_test):
752         (test_run_with_json_output):
753         (test_run_with_description):
754         (test_run_with_slave_config_json):
755         (test_run_with_multiple_repositories):
756
757 2012-09-20  Simon Hausmann  <simon.hausmann@digia.com>
758
759         [Qt] QtWebKit module header includes private dependencies
760
761         Reviewed by Tor Arne Vestbø.
762
763         * qmake/mkspecs/features/default_post.prf: Don't try to sanitize LIBS, because we can make sure
764         that LIBS_PRIVATE is set from the beginning. Moved the creating_module and PKGCONFIG/QT(_PRIVATE)
765         sanitization into webkit_modules.prf. creating_module determination requires TARGET to be set, so
766         we can't do it in default_pre.
767         * qmake/mkspecs/features/webkit_modules.prf:
768
769 2012-09-20  Simon Hausmann  <simon.hausmann@digia.com>
770
771         [Qt] Fix build with MingW
772
773         Reviewed by Tor Arne Vestbø.
774
775         Don't try to compile WebCore, etc. with debug symbols for production
776         builds, it's just too big.
777
778         * qmake/mkspecs/features/production_build.prf:
779
780 2012-09-20  Michał Pakuła vel Rutka  <m.pakula@samsung.com>
781
782         [EFL][DRT]DumpRenderTree needs to reset focus state when test starts.
783         https://bugs.webkit.org/show_bug.cgi?id=97087
784
785         Reviewed by Gyuyoung Kim.
786
787         Add focusing a main frame on settings reset.
788         After editing/undo/undo-iframe-location-change was executed a frame
789         was left in unfocused state. This caused flakiness in two tests results.
790
791         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
792         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
793
794 2012-09-20  Allan Sandfeld Jensen  <allan.jensen@digia.com>
795
796         [Qt] Add eventSender.gestureTap
797         https://bugs.webkit.org/show_bug.cgi?id=66173
798
799         Reviewed by Kenneth Rohde Christiansen.
800
801         Implement support for sending raw QGesture events.
802
803         * DumpRenderTree/qt/EventSenderQt.cpp:
804         (EventSender::EventSender):
805         (EventSender::gestureTap):
806         * DumpRenderTree/qt/EventSenderQt.h:
807         (EventSender):
808
809 2012-09-19  Sheriff Bot  <webkit.review.bot@gmail.com>
810
811         Unreviewed, rolling out r129007.
812         http://trac.webkit.org/changeset/129007
813         https://bugs.webkit.org/show_bug.cgi?id=97172
814
815         It broke the build on the Qt bots (Requested by Ossy on
816         #webkit).
817
818         * qmake/config.tests/gccdepends/empty.cpp:
819         * qmake/config.tests/gccdepends/gccdepends.pro:
820         * qmake/mkspecs/features/default_pre.prf:
821
822 2012-09-19  Ryosuke Niwa  <rniwa@webkit.org>
823
824         run-perf-tests should record indivisual value instead of statistics
825         https://bugs.webkit.org/show_bug.cgi?id=97155
826
827         Reviewed by Hajime Morita.
828
829         Parse the list of indivisual value reported by tests and include them as "values".
830         We strip "values" from the output JSON when uploading it to the perf-o-matic
831         since it doesn't know how to parse "values" or ignore it.
832
833         * Scripts/webkitpy/performance_tests/perftest.py:
834         (PerfTest):
835         (PerfTest.parse_output): Parse and report "values".
836         (PageLoadingPerfTest.run): Report indivisual page loading time in "values".
837         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
838         (MainTest.test_parse_output):
839         (MainTest.test_parse_output_with_failing_line):
840         (TestPageLoadingPerfTest.test_run):
841         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
842         (PerfTestsRunner._generate_and_show_results): Strip "values" from each result
843         until we update perf-o-matic.
844         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
845         (test_run_memory_test):
846         (test_run_with_json_output):
847         (test_run_with_description):
848         (test_run_with_slave_config_json):
849         (test_run_with_multiple_repositories):
850
851 2012-09-19  Dirk Pranke  <dpranke@chromium.org>
852
853         Fix regex groups for bug matching in flakiness dashboard.
854         https://bugs.webkit.org/show_bug.cgi?id=97152
855
856         Unreviewed, build fix.
857
858         * TestResultServer/static-dashboards/flakiness_dashboard.js:
859         (htmlForBugs):
860
861 2012-09-19  Dirk Pranke  <dpranke@chromium.org>
862
863         nrwt: print unexpected results using new TestExpectations syntax
864         https://bugs.webkit.org/show_bug.cgi?id=97159
865
866         Unreviewed, build fix.
867
868         Change new-run-webkit-tests to print out failures using the new
869         syntax when there are unexpected results, e.g.:
870
871         52 tests ran as expected, 19 didn't:
872
873
874         Regressions: Unexpected failures : (2)
875           failures/flaky/text.html [ Failure ]
876           failures/unexpected/text-image-checksum.html [ Failure ]
877
878         and so forth
879
880         * Scripts/webkitpy/layout_tests/views/printing.py:
881         (Printer._print_unexpected_results):
882
883 2012-09-19  Dirk Pranke  <dpranke@chromium.org>
884
885         update flakiness dashboard after cutover to new test expectations syntax
886         https://bugs.webkit.org/show_bug.cgi?id=97152
887
888         Unreviewed, build fix.
889
890         Handle (??) the new Bug notations as well. Hopefully we don't
891         still need the old ones.
892
893         * TestResultServer/static-dashboards/flakiness_dashboard.js:
894         (filterBugs):
895         (htmlForBugs):
896         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
897
898 2012-09-19  Dirk Pranke  <dpranke@chromium.org>
899
900         update flakiness dashboard after cutover to new test expectations syntax
901         https://bugs.webkit.org/show_bug.cgi?id=97152
902
903         Reviewed by Ryosuke Niwa.
904
905         This change clones the TestExpectation parsing state machine
906         from python into javascript.
907
908         * TestResultServer/static-dashboards/flakiness_dashboard.js:
909         (parsedExpectations.lines.forEach):
910         (parsedExpectations):
911         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
912
913 2012-09-19  Dana Jansens  <danakj@chromium.org>
914
915         Add backer@chromium.org as contributor
916         https://bugs.webkit.org/show_bug.cgi?id=97150
917
918         * Scripts/webkitpy/common/config/committers.py:
919
920 2012-09-19  Dirk Pranke  <dpranke@chromium.org>
921
922         fix MISSING after TestExpectations conversion
923         https://bugs.webkit.org/show_bug.cgi?id=97148
924
925         Unreviewed, expectations change / build fix.
926
927         * Scripts/convert-test-expectations:
928         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
929         (TestExpectationParser):
930
931 2012-09-19  Dirk Pranke  <dpranke@chromium.org>
932
933         nrwt: convert chromium TestExpectations to the new syntax
934         https://bugs.webkit.org/show_bug.cgi?id=97139
935
936         Reviewed by Ryosuke Niwa.
937     
938         Add in a file temporarily that does the conversion of
939         test expectations formats
940
941         * Tools/Scripts/convert-test-expectations: Added.
942
943 2012-09-19  Ryosuke Niwa  <rniwa@webkit.org>
944
945         REGRESSION: run-perf-tests no longer reports the total test time
946         https://bugs.webkit.org/show_bug.cgi?id=97138
947
948         Reviewed by Tony Chang.
949
950         Report the finished time as a info-level log as opposed to a debug level log.
951
952         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
953         (PerfTestsRunner._run_single_test):
954         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
955         (MainTest.normalizeFinishedTime): Added. It replaces all finished times by 0.1 seconds.
956         (test_run_test_pause_before_testing):
957         (test_run_test_set_for_parser_tests):
958         (test_run_memory_test):
959         (_test_run_with_json_output):
960
961 2012-09-19  Dirk Pranke  <dpranke@chromium.org>
962
963         Support new TestExpectations format alongside old one
964         https://bugs.webkit.org/show_bug.cgi?id=96588
965
966         Reviewed by Ojan Vafai.
967
968         This patch adds support for actually reading in lines formatted
969         in the new style, and when we re-serialize/write out the file,
970         all lines will be written in the new style. Note that reading in
971         the old style is still supported, and no updates are being made
972         to the actual TestExpectations files as part of this change.
973
974         This change updates most but not all of the unit tests to use
975         the new syntax. I will update the rest when (or before where
976         possible) I drop support for the old format.
977
978         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
979         (TestExpectationParser):
980         (TestExpectationParser._tokenize_line):
981         (TestExpectationParser._tokenize_line_using_new_format):
982         (TestExpectationLine.to_string):
983         (TestExpectationLine._format_line):
984         (TestExpectations):
985         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
986         (test_parse_warning):
987         (SkippedTests.test_skipped_entry_dont_exist):
988         (NewExpectationSyntaxTests.test_warnings):
989         (RemoveConfigurationsTest.test_remove):
990         (test_remove_line):
991         (RebaseliningTest.test_remove):
992         (RebaseliningTest.test_no_get_rebaselining_failures):
993         (TestExpectationParserTests.test_tokenize_blank):
994         (TestExpectationParserTests.test_tokenize_extra_colon):
995         (TestExpectationParserTests.test_tokenize_missing_equal):
996         (TestExpectationParserTests.test_tokenize_extra_equal):
997         (TestExpectationSerializationTests.test_unparsed_to_string):
998         (TestExpectationSerializationTests.test_unparsed_list_to_string):
999         (TestExpectationSerializationTests.test_parsed_to_string):
1000         (TestExpectationSerializationTests.test_format_line):
1001         (TestExpectationSerializationTests.test_string_roundtrip):
1002         (TestExpectationSerializationTests.test_list_roundtrip):
1003         (TestExpectationSerializationTests.test_reconstitute_only_these):
1004         (TestExpectationSerializationTests.test_string_whitespace_stripping):
1005         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1006         (LintTest.test_lint_test_files__errors):
1007         * Scripts/webkitpy/tool/commands/queries_unittest.py:
1008         (PrintExpectationsTest.test_basic):
1009         (PrintExpectationsTest.test_multiple):
1010         (PrintExpectationsTest.test_full):
1011         (PrintExpectationsTest.test_exclude):
1012         * Scripts/webkitpy/tool/commands/rebaseline.py:
1013         (RebaselineExpectations.execute):
1014         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
1015         (TestRebaseline.test_rebaseline_updates_expectations_file_noop):
1016         (test_rebaseline_updates_expectations_file):
1017         (test_rebaseline_does_not_include_overrides):
1018         (test_rebaseline_expectations_noop):
1019         (test_overrides_are_included_correctly):
1020
1021 2012-09-19  Dirk Pranke  <dpranke@chromium.org>
1022
1023         implement first part of support for the new TestExpectations syntax
1024         https://bugs.webkit.org/show_bug.cgi?id=96569
1025
1026         Reviewed by Ryosuke Niwa.
1027
1028         This patch implements support for parsing a line of the new
1029         format for the TestExpectations file and converting it back into
1030         the old format for compatibility. This routine is not yet used
1031         by anything.
1032
1033         The new format is documented at:
1034           http://trac.webkit.org/wiki/TestExpectations
1035
1036         but, in short:
1037
1038           [bugs] [ "[" modifiers "]" ] test_name [ "[" expectations "]" ]
1039
1040         - Comments are indicated with "#" instead of "//"
1041         - If no expectations are specified we default to Skip for
1042           compatibility with the Skipped files (these two changes make
1043           Skipped files a subset of TestExpectations files)
1044
1045         - All of the tokens are now CamelCase instead of ALLCAPS.
1046         - FAIL -> Failure
1047         - IMAGE -> ImageOnlyFailure
1048         - WONTFIX -> WontFix
1049         - modifiers refer to just the platforms and configurations
1050           (release/debug) that the line applies to.
1051         - WontFix, Rebaseline, Slow, and Skip move to the right-hand side as
1052           expectations
1053         - expectations will typically be written out in lexicographic order
1054         - We use webkit.org/b/12345, crbug.com/12345, and Bug(dpranke)
1055           instead of BUGWK12345, BUGCR12345, and BUGDPRANKE.
1056
1057         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
1058         (TestExpectationParser):
1059         (TestExpectationParser._tokenize_line_using_new_format):
1060         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
1061         (NewExpectationSyntaxTests):
1062         (NewExpectationSyntaxTests.assert_exp):
1063         (NewExpectationSyntaxTests.test_bare_name):
1064         (NewExpectationSyntaxTests.test_bare_name_and_bugs):
1065         (NewExpectationSyntaxTests.test_comments):
1066         (NewExpectationSyntaxTests.test_config_modifiers):
1067         (NewExpectationSyntaxTests.test_unknown_config):
1068         (NewExpectationSyntaxTests.test_unknown_expectation):
1069         (NewExpectationSyntaxTests.test_skip):
1070         (NewExpectationSyntaxTests.test_slow):
1071         (NewExpectationSyntaxTests.test_wontfix):
1072         (NewExpectationSyntaxTests.test_blank_line):
1073         (NewExpectationSyntaxTests.test_warnings):
1074
1075 2012-09-19  Dirk Pranke  <dpranke@chromium.org>
1076
1077         nrwt: replace TEXT, AUDIO, and IMAGE+TEXT with FAIL
1078         https://bugs.webkit.org/show_bug.cgi?id=96845
1079
1080         Reviewed by Ojan Vafai.
1081
1082         In preparation for the new TestExpectations syntax, we replace
1083         all TEXT, IMAGE+TEXT, and AUDIO failures with FAIL. This will
1084         make switching to the new syntax lossless (i.e., we lose
1085         information now, not then).
1086
1087         Note that we can still parse in results.json files that have the
1088         old data for backwards compatibility.
1089         
1090
1091         * Scripts/webkitpy/common/net/resultsjsonparser.py:
1092         (JSONTestResult._failure_types_from_actual_result):
1093         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
1094         (ResultsJSONParserTest):
1095         (test_basic):
1096         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
1097         (ResultSummaryTest.test_summarized_results_wontfix):
1098         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
1099         (JSONLayoutResultsGenerator):
1100         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
1101         (TestExpectations):
1102         (TestExpectations.remove_pixel_failures):
1103         (TestExpectations.has_pixel_failures):
1104         (TestExpectations.suffixes_for_expectations):
1105         (TestExpectations.get_rebaselining_failures):
1106         (TestExpectations.remove_configuration_from_test):
1107         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
1108         (assert_bad_expectations):
1109         (BasicTests):
1110         (BasicTests.test_basic):
1111         (MiscTests):
1112         (MiscTests.test_multiple_results):
1113         (MiscTests.test_result_was_expected):
1114         (MiscTests.test_remove_pixel_failures):
1115         (MiscTests.test_suffixes_for_expectations):
1116         (test_get_expectations_string):
1117         (test_parse_warning):
1118         (test_error_on_different_platform):
1119         (test_error_on_different_build_type):
1120         (test_overrides):
1121         (test_overrides__directory):
1122         (test_overrides__duplicate):
1123         (test_pixel_tests_flag):
1124         (test_more_specific_override_resets_skip):
1125         (SkippedTests.test_skipped_file_overrides_expectations):
1126         (SkippedTests.test_skipped_dir_overrides_expectations):
1127         (SkippedTests.test_skipped_file_overrides_overrides):
1128         (SkippedTests.test_skipped_dir_overrides_overrides):
1129         (ExpectationSyntaxTests.test_missing_colon):
1130         (ExpectationSyntaxTests.test_too_many_equals_signs):
1131         (SemanticTests.test_bug_format):
1132         (SemanticTests.test_bad_bugid):
1133         (SemanticTests.test_missing_bugid):
1134         (SemanticTests.test_rebaseline):
1135         (test_missing_file):
1136         (test_more_modifiers):
1137         (test_order_in_file):
1138         (test_macro_overrides):
1139         (RemoveConfigurationsTest.test_remove):
1140         (test_remove_line):
1141         (RebaseliningTest.test_remove):
1142         (TestExpectationSerializationTests.test_serialize_parsed_expectations):
1143         * Scripts/webkitpy/layout_tests/models/test_failures.py:
1144         (determine_result_type):
1145         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
1146         (test_test_expectations):
1147         * Scripts/webkitpy/layout_tests/port/test.py:
1148         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1149         (MainTest.test_missing_and_unexpected_results):
1150         (MainTest.test_retrying_and_flaky_tests):
1151         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
1152         (TestExpectationsTestCase.test_valid_expectations):
1153         * Scripts/webkitpy/tool/commands/queries_unittest.py:
1154         (PrintExpectationsTest.test_basic):
1155         (PrintExpectationsTest.test_multiple):
1156         (PrintExpectationsTest.test_full):
1157         (PrintExpectationsTest.test_exclude):
1158         (PrintExpectationsTest.test_csv):
1159         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
1160         (test_rebaseline_does_not_include_overrides):
1161         (test_overrides_are_included_correctly):
1162
1163 2012-09-19  Dominic Mazzoni  <dmazzoni@google.com>
1164
1165         AX: A few control types are returning the wrong answer for isReadOnly
1166         https://bugs.webkit.org/show_bug.cgi?id=96735
1167
1168         Reviewed by Chris Fleizach.
1169
1170         Exposing isReadOnly in an AccessibilityObject to DumpRenderTree.
1171
1172         * DumpRenderTree/chromium/TestRunner/AccessibilityUIElementChromium.cpp:
1173         (AccessibilityUIElement::AccessibilityUIElement):
1174         (AccessibilityUIElement::isReadOnlyGetterCallback):
1175         * DumpRenderTree/chromium/TestRunner/AccessibilityUIElementChromium.h:
1176         (AccessibilityUIElement):
1177
1178 2012-09-19  Sudarsana Nagineni  <sudarsana.nagineni@intel.com>
1179
1180         [WTR] Memory leaks in TestRunner::deliverWebIntent()
1181         https://bugs.webkit.org/show_bug.cgi?id=97111
1182
1183         Reviewed by Kenneth Rohde Christiansen.
1184
1185         Fix memory leaks in deliverWebIntent() by adopting strings
1186         created with WKStringCreateWithUTF8CString().
1187
1188         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
1189         (WTR::TestRunner::deliverWebIntent):
1190
1191 2012-09-19  Simon Hausmann  <simon.hausmann@digia.com>
1192
1193         [Qt] Fix incremental builds with all-in-one-files and gccdepends
1194
1195         Reviewed by Tor Arne Vestbø.
1196
1197         Pass -MP to gcc when we use the gcc depends feature, to ensure that implicit rules
1198         are not only created for header files but also for .cpp files. AllInOne.cpp files
1199         include other .cpp files, and when those are removed we need those dummy rules to
1200         avoid a "No rule to make Foo.cpp required by AllInOne.o" error.
1201
1202         * qmake/config.tests/gccdepends/empty.cpp:
1203         (main):
1204         * qmake/config.tests/gccdepends/gccdepends.pro:
1205         * qmake/mkspecs/features/default_pre.prf:
1206
1207 2012-09-19  Rick Byers  <rbyers@chromium.org>
1208
1209         Do touch adjustment on GestureTapDown
1210         https://bugs.webkit.org/show_bug.cgi?id=96677
1211
1212         Reviewed by Antonio Gomes.
1213
1214         Allow radius to be set for GestureTapDown events.
1215         * DumpRenderTree/chromium/TestRunner/EventSender.cpp:
1216         (EventSender::gestureEvent):
1217
1218 2012-09-19  Christophe Dumez  <christophe.dumez@intel.com>
1219
1220         [EFL] EFL's DRT does not support overriding 'WebKitCSSRegionsEnabled' preference
1221         https://bugs.webkit.org/show_bug.cgi?id=97100
1222
1223         Reviewed by Gyuyoung Kim.
1224
1225         EFL's DRT now supports overriding the 'WebKitCSSRegionsEnabled'
1226         preference, in order to match WebKitTestRunner functionality.
1227
1228         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1229         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
1230         * DumpRenderTree/efl/TestRunnerEfl.cpp:
1231         (TestRunner::overridePreference):
1232
1233 2012-09-18  Kangil Han  <kangil.han@samsung.com>
1234
1235         [WK2][WTR] CodeGeneratorTestRunner could keep original copyright.
1236         https://bugs.webkit.org/show_bug.cgi?id=96181
1237
1238         Reviewed by Daniel Bates.
1239
1240         This patch enabled derived files, in DerivedSources/InjectedBundle, to keep original copyright.
1241
1242         * WebKitTestRunner/InjectedBundle/Bindings/CodeGeneratorTestRunner.pm:
1243         (new):
1244         (_parseLicenseBlock):
1245         (_parseLicenseBlockFromFile):
1246         (_defaultLicenseBlock):
1247         (_licenseBlock):
1248         (_generateHeaderFile):
1249         (_generateImplementationFile):
1250
1251 2012-09-18  Byungwoo Lee  <bw80.lee@samsung.com>
1252
1253         Title string should be changed when document.title is set to ''.
1254         https://bugs.webkit.org/show_bug.cgi?id=96793
1255
1256         Reviewed by Kenneth Rohde Christiansen.
1257
1258         Change dump format of dumpTitleChanges more understandable.
1259         Uses single quotation marks for the title string.
1260
1261         * DumpRenderTree/blackberry/DumpRenderTree.cpp:
1262         (BlackBerry::WebKit::DumpRenderTree::didReceiveTitleForFrame):
1263         * DumpRenderTree/chromium/WebViewHost.cpp:
1264         (WebViewHost::didReceiveTitle):
1265         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1266         (DumpRenderTreeChrome::onFrameTitleChanged):
1267         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1268         (webViewTitleChanged):
1269         * DumpRenderTree/mac/FrameLoadDelegate.mm:
1270         (-[FrameLoadDelegate webView:didReceiveTitle:forFrame:]):
1271         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1272         (WebCore::DumpRenderTree::titleChanged):
1273         * DumpRenderTree/win/FrameLoadDelegate.cpp:
1274         (FrameLoadDelegate::didReceiveTitle):
1275         * DumpRenderTree/wx/DumpRenderTreeWx.cpp:
1276         (LayoutWebViewEventHandler::OnReceivedTitleEvent):
1277         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1278         (WTR::InjectedBundlePage::didReceiveTitleForFrame):
1279
1280 2012-09-18  Szilard Ledan  <szledan@inf.u-szeged.hu>
1281
1282         EWS shouldn't sleep if there are new patches in its queue
1283         https://bugs.webkit.org/show_bug.cgi?id=83038
1284
1285         Reviewed by Eric Seidel.
1286
1287         EWS tries to process a security patch. Of course it can't, because the EWS isn't
1288         the member of the security group. But the problem is that after it can't process
1289         the attachment, it says that queue is empty (but it isn't!) and it sleeps 2 minutes
1290         and push the security patch to the end of the queue.
1291         Now it stays in the loop until it finds a patch or the queue gets empty.
1292
1293         * Scripts/webkitpy/tool/commands/queues.py:
1294         (AbstractPatchQueue._next_patch):
1295         * Scripts/webkitpy/tool/commands/queues_unittest.py:
1296         (AbstractPatchQueueTest.test_next_patch):
1297
1298 2012-09-18  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
1299
1300         [WK2][WTR] InjectedBundle::booleanForKey() should handle literals effectively
1301         https://bugs.webkit.org/show_bug.cgi?id=97014
1302
1303         Reviewed by Kenneth Rohde Christiansen.
1304
1305         According to http://trac.webkit.org/wiki/EfficientStrings WTF::StringBuilder::appendLiteral() shall
1306         be used for literals rather than WTF::StringBuilder::append().
1307
1308         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
1309         (WTR::InjectedBundle::booleanForKey):
1310
1311 2012-09-18  Simon Hausmann  <simon.hausmann@digia.com>
1312
1313         Update my e-mail address.
1314
1315         * Scripts/webkitpy/common/config/committers.py:
1316
1317 2012-09-18  Andras Becsi  <andras.becsi@digia.com>
1318
1319         Update my e-mail address.
1320
1321         Unreviewed.
1322
1323         * Scripts/webkitpy/common/config/committers.py:
1324
1325 2012-09-18  Andras Becsi  <andras.becsi@digia.com>
1326
1327         [Qt] qt_webkit.pri should not be listed in Tools.pro
1328
1329         Reviewed and rubber-stamped by Simon Hausmann and Tor Arne Vestbø.
1330
1331         Since r128751 the module pri file is auto-generated
1332         but it was still listed in OTHER_FILES.
1333
1334         * Tools.pro: Remove unneeded line.
1335
1336 2012-09-18  Simon Hausmann  <simon.hausmann@digia.com>
1337
1338         [Qt] Fix build with some versions of the gold linker
1339
1340         Reviewed by Tor Arne Vestbø.
1341
1342         Don't unconditionally pass --no-keep-memory to the linker, some versions might not support it.
1343         Instead run a compile/link test first to see if it works.
1344
1345         * qmake/config.tests/gnuld/gnuld.pro: Added.
1346         * qmake/config.tests/gnuld/main.cpp: Added.
1347         (main):
1348         * qmake/mkspecs/features/unix/default_post.prf:
1349
1350 2012-09-18  Simon Hausmann  <simon.hausmann@digia.com>
1351
1352         [Qt] Remove forced use of gold.
1353
1354         Reviewed by Tor Arne Vestbø.
1355
1356         The choice of what linker to use with WebKit should be taken by Qt's build system and ideally the same for all
1357         modules of Qt. Then in turn it is usually up to the administrator of the machine. Recent Debian based systems
1358         often offer the automatic use of gold through a symlink and a dpkg-diversion when installing the gold package.
1359
1360         * qmake/mkspecs/features/unix/default_post.prf:
1361
1362 2012-09-18  Simon Hausmann  <simon.hausmann@digia.com>
1363
1364         [Qt] Prospective Qt/Windows cross-compiling fix
1365
1366         Reviewed by Tor Arne Vestbø.
1367
1368         The win32 scope is not set when cross-compiling from Linux to Windows.
1369
1370         * qmake/mkspecs/features/functions.prf:
1371
1372 2012-09-18  Seokju Kwon  <seokju.kwon@samsung.com>
1373
1374         [EFL] Remove background view on EWebLauncher and MiniBrowser
1375         https://bugs.webkit.org/show_bug.cgi?id=96905
1376
1377         Reviewed by Gyuyoung Kim.
1378
1379         The size of webview was changed after adding url bar.
1380         And the background view is not necessary anymore, since it was used for debugging back in the day.
1381
1382         * EWebLauncher/main.c:
1383         (_ELauncher):
1384         (on_ecore_evas_resize):
1385         (browserCreate):
1386         * MiniBrowser/efl/main.c:
1387         (_MiniBrowser):
1388         (on_ecore_evas_resize):
1389         (browserCreate):
1390
1391 2012-09-18  Sergio Villar Senin  <svillar@igalia.com>
1392
1393         [GTK] run-webkit-tests unable to find TestExpectations for WK2
1394         https://bugs.webkit.org/show_bug.cgi?id=96998
1395
1396         Reviewed by Philippe Normand.
1397
1398         We should look for TestExpectations files in all the locations where
1399         we currently look for Skipped files. This will allow
1400         run-webkit-tests to look for TestExpectations files in
1401         platform/gtk-wk2 and platform/wk2 if the "-2" flag is used.
1402
1403         * Scripts/webkitpy/layout_tests/port/gtk.py:
1404         (GtkPort.expectations_files):
1405
1406 2012-09-18  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1407
1408         [Qt] Fix build without the QtQuick module
1409
1410         Reviewed by Simon Hausmann.
1411
1412         * qmake/mkspecs/features/features.prf:
1413
1414 2012-09-18  Stephanie Lewis  <slewis@apple.com>
1415
1416         Build fix after http://trac.webkit.org/projects/webkit/changeset/128852.
1417
1418         Unreviewed.
1419
1420         * DumpRenderTree/mac/DumpRenderTree.mm:
1421         (poseAsClass):
1422
1423 2012-09-18  Allan Sandfeld Jensen  <allan.jensen@nokia.com>
1424
1425         Unreviewed update of email addresses for Berlin QtWebKit office.
1426
1427         * Scripts/webkitpy/common/config/committers.py:
1428
1429 2012-09-17  Zan Dobersek  <zandobersek@gmail.com>
1430
1431         [GTK] fast/loader/display-image-unset-can-block-image-and-can-reload-in-place.html failing after r128645
1432         https://bugs.webkit.org/show_bug.cgi?id=96899
1433
1434         Reviewed by Martin Robinson.
1435
1436         When overriding the 'WebKitDisplayImageKey' preference, use the
1437         'auto-load-images' property of WebKitWebSettings as the property which
1438         should be updated with the corresponding preference value.
1439
1440         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1441         (resetDefaultsToConsistentValues):
1442         * DumpRenderTree/gtk/TestRunnerGtk.cpp:
1443         (TestRunner::overridePreference):
1444
1445 2012-09-17  Pratik Solanki  <psolanki@apple.com>
1446
1447         DumpRenderTree and WebKitTestRunner should compile with -Wundef on Mac
1448         https://bugs.webkit.org/show_bug.cgi?id=96973
1449
1450         Reviewed by Dan Bernstein.
1451
1452         * DumpRenderTree/mac/Configurations/Base.xcconfig:
1453         * WebKitTestRunner/Configurations/Base.xcconfig:
1454         * WebKitTestRunner/PlatformWebView.h: Use #ifdef __OBJC__ and not #if.
1455
1456 2012-09-17  Dirk Pranke  <dpranke@chromium.org>
1457
1458         nrwt: remove "unexpected EOF" warnings
1459         https://bugs.webkit.org/show_bug.cgi?id=96970
1460
1461         Reviewed by Ojan Vafai.
1462
1463         After debugging this a bit, it looks like there aren't any cases
1464         that I can reproduce where a read() of zero indicates something
1465         actually wrong; either it is a prelude to a crash, or a false
1466         negative. So, I'm removing these warnings and adding a comment.
1467
1468         * Scripts/webkitpy/layout_tests/port/server_process.py:
1469         (ServerProcess._wait_for_data_and_update_buffers_using_select):
1470
1471 2012-09-17  Dirk Pranke  <dpranke@chromium.org>
1472
1473         [chromium] ASAN bot is crashing at the end of the run
1474         https://bugs.webkit.org/show_bug.cgi?id=96967
1475
1476         Reviewed by Abhishek Arya.
1477
1478         The ASAN bot is crashing attempting to decode some output into
1479         UTF-8; there's no reason to do this, so let's not do this and
1480         see if something else is going on as well.
1481
1482         * Scripts/webkitpy/layout_tests/port/chromium.py:
1483         (ChromiumPort._get_crash_log):
1484
1485 2012-09-17  Dirk Pranke  <dpranke@chromium.org>
1486
1487         nrwt: --results-directory isn't getting printed properly
1488         https://bugs.webkit.org/show_bug.cgi?id=96965
1489
1490         Reviewed by Ojan Vafai.
1491
1492         options.results_directory isn't actually initialized with the
1493         default values until after we call print_config(), so this
1494         changes things to print the value directly.
1495
1496         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1497         (run):
1498         * Scripts/webkitpy/layout_tests/views/printing.py:
1499         (Printer.print_config):
1500         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
1501         (Testprinter.test_print_config):
1502
1503 2012-09-17  Rob Buis  <rbuis@rim.com>
1504
1505         [BlackBerry] Enable VIDEO_TRACK
1506         https://bugs.webkit.org/show_bug.cgi?id=96949
1507
1508         Reviewed by Antonio Gomes.
1509
1510         * Scripts/webkitperl/FeatureList.pm:
1511
1512 2012-09-17  Rick Byers  <rbyers@chromium.org>
1513
1514         Add handling of new GestureTapCancel in DRT
1515
1516         https://bugs.webkit.org/show_bug.cgi?id=96183
1517
1518         Reviewed by Antonio Gomes.
1519
1520         * DumpRenderTree/chromium/TestWebPlugin.cpp:
1521         (TestWebPlugin::handleInputEvent):
1522         * DumpRenderTree/chromium/EventSender.cpp:
1523         (EventSender::gestureTapCancel):
1524
1525 2012-09-17  Philip Rogers  <pdr@google.com>
1526
1527         Teach style checker about preprocessor directive indentation rules
1528         https://bugs.webkit.org/show_bug.cgi?id=96874
1529
1530         Reviewed by Adam Barth.
1531
1532         Preprocessor directives (#ifdef, #include, #define, etc.) should not be indented.
1533         This is not explicit in our style guide but is generally followed in our code.
1534         Searching for violations in our codebase shows these are rarely indented:
1535             #include - indented in 6 files
1536             #ifdef - indented in 0 files
1537             #ifndef - indented in 1 file
1538             #define - indented in 11 files
1539             #if - indented in 7 files
1540
1541         * Scripts/webkitpy/style/checkers/cpp.py:
1542         (check_directive_indentation):
1543
1544             This is the simple test where we look for spaces followed by a #.
1545
1546         (check_style):
1547         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
1548
1549             A few tests needed to be modified because they had unintentionally indented
1550             preprocessor directives.
1551
1552         (CppStyleTest.test_build_class.Foo):
1553         (CppStyleTest.test_build_class):
1554         (CppStyleTest.test_build_class.DERIVE_FROM_GOO):
1555         (WebKitStyleTest.test_line_breaking):
1556         (WebKitStyleTest.test_directive_indentation):
1557
1558 2012-09-17  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1559
1560         [Qt] Auto-generate the module pri file for QtWebKit
1561
1562         Reviewed by Simon Hausmann.
1563
1564         * qmake/qt_webkit.pri: Removed.
1565
1566 2012-09-14  Dirk Pranke  <dpranke@chromium.org>
1567
1568         nrwt: --additional-platform-dir is broken on chromium ports
1569         https://bugs.webkit.org/show_bug.cgi?id=96840
1570
1571         Reviewed by Ojan Vafai.
1572
1573         Looks like I broke this when cleaning things up as part of
1574         adding ML support. Fixing and adding a test.
1575
1576         * Scripts/webkitpy/layout_tests/port/chromium.py:
1577         (ChromiumPort.default_baseline_search_path):
1578         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
1579         (ChromiumAndroidPort.default_baseline_search_path):
1580         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
1581         (ChromiumLinuxPort.default_baseline_search_path):
1582         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
1583         (ChromiumMacPort.__init__):
1584         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
1585         (ChromiumWinPort.setup_environ_for_server):
1586         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
1587         (test_check_build):
1588         (test_additional_platform_directory):
1589
1590 2012-09-13  Stephanie Lewis  <slewis@apple.com>
1591
1592         mac-future ignores previous platform TestExpectations.
1593         https://bugs.webkit.org/show_bug.cgi?id=96718.
1594
1595         Reviewed by Dirk Pranke.
1596
1597         TestExpectations match an expectation's configuraton against a list of configurations that are 
1598         valid for that port.  That list does not contain mac-future so none of the expectations can be
1599         applied.
1600
1601         * Scripts/webkitpy/layout_tests/port/apple.py:
1602         (ApplePort._generate_all_test_configurations):
1603
1604 2012-09-14  Adam Barth  <abarth@webkit.org>
1605
1606         Remove webkitPostMessage
1607         https://bugs.webkit.org/show_bug.cgi?id=96577
1608
1609         Reviewed by Ojan Vafai.
1610
1611         Add ENABLE_LEGACY_VENDOR_PREFIXES flag.
1612
1613         * Scripts/webkitperl/FeatureList.pm:
1614         * qmake/mkspecs/features/features.pri:
1615
1616 2012-09-14  Alexey Proskuryakov  <ap@apple.com>
1617
1618         [WK2] webarchive/loading/javascript-url-iframe-crash.html fails
1619         https://bugs.webkit.org/show_bug.cgi?id=96817
1620
1621         Reviewed by Kenneth Rohde Christiansen.
1622
1623         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1624         (WTR::InjectedBundlePage::didFailProvisionalLoadWithErrorForFrame): This callback
1625         got overlooked when adding logging elsewhere.
1626
1627 2012-09-14  Peter Beverloo  <peter@chromium.org>
1628
1629         [Chromium] Build fix for DumpRenderTree following r128628
1630         https://bugs.webkit.org/show_bug.cgi?id=96808
1631
1632         Reviewed by Tony Chang.
1633
1634         When building DumpRenderTree from the Chromium project, a compile error
1635         shows up because including the header ForwardIOStreamsAndroid.h cannot
1636         be located. TestWebKitAPI does this correctly.
1637
1638         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
1639
1640 2012-09-14  Dana Jansens  <danakj@chromium.org>
1641
1642         Change cc-bugs@google.com watchlist to cc-bugs@chromium.org
1643         https://bugs.webkit.org/show_bug.cgi?id=96805
1644
1645         Reviewed by Dirk Pranke.
1646
1647         * Scripts/webkitpy/common/config/committers.py:
1648         * Scripts/webkitpy/common/config/watchlist:
1649
1650 2012-09-14  Peter Beverloo  <peter@chromium.org>
1651
1652         [Chromium] Support the --{in,out,err}-fifo arguments on TestWebKitAPI and webkit_unit_tests
1653         https://bugs.webkit.org/show_bug.cgi?id=96687
1654
1655         Reviewed by Tony Chang.
1656
1657         Android's DumpRenderTree currently supports these arguments, implemented
1658         as part of TestShellAndroid:
1659         http://trac.webkit.org/browser/trunk/Tools/DumpRenderTree/chromium/TestShellAndroid.cpp?rev=128496
1660
1661         They're used by the layout test runner to get the STDOUT and STDERR while
1662         a layout test run is in process, which is a safer alternative to parsing
1663         all the logcat output manually. The implementation can be seen here:
1664         http://trac.webkit.org/browser/trunk/Tools/Scripts/webkitpy/layout_tests/port/chromium_android.py?rev=128496#L590
1665
1666         This patch generalizes parsing of and applying the effects of these arguments
1667         so that they can be used for TestWebKitAPI and webkit_unit_tests as well.
1668         After this patch, this will make it possible to pull out the output-reading
1669         code from Android's layout test port and generalize it so it can be re-used
1670         in the new test-runner for the other two test suites.
1671
1672         This has no effect when compiling and running these tests as part of Chromium
1673         code, which has a much more advanced test-runner that does parse complete log
1674         output, but also directly depends on code licensed under Apache 2.
1675
1676         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
1677         * DumpRenderTree/chromium/TestShellAndroid.cpp:
1678         (platformInit):
1679         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
1680
1681 2012-09-14  Zoltan Horvath  <zoltan@webkit.org>
1682
1683         check-webkit-style should not warn in case of NONCOPYABLE and FAST_ALLOCATED macros
1684         https://bugs.webkit.org/show_bug.cgi?id=96656
1685
1686         Reviewed by Adam Barth.
1687
1688         We should allow to use WTF_MAKE_NONCOPYABLE(ClassName) and WTF_MAKE_FAST_ALLOCATED
1689         macros in 1 line.
1690
1691         * Scripts/webkitpy/style/checkers/cpp.py:
1692         (check_style.definitions):
1693         (check_style): Add rule.
1694         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
1695         (WebKitStyleTest.test_line_breaking): Add unittest.
1696
1697 2012-09-14  Christophe Dumez  <christophe.dumez@intel.com>
1698
1699         WebKitTestRunner needs layoutTestController.dumpDatabaseCallbacks
1700         https://bugs.webkit.org/show_bug.cgi?id=57570
1701
1702         Reviewed by Kenneth Rohde Christiansen.
1703
1704         Implement testRunner.dumpDatabaseCallbacks in WebKitTestRunner
1705         and dump the information expected by the test cases.
1706
1707         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
1708         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
1709         (WTR::InjectedBundle::beginTesting): Explicitly set the default
1710         database quota.
1711         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1712         (WTR::securityOriginToStr): New convenience function to convert
1713         a security origin to the string expected in test results.
1714         (WTR):
1715         (WTR::InjectedBundlePage::InjectedBundlePage):
1716         (WTR::InjectedBundlePage::didExceedDatabaseQuota):
1717         (WTR::InjectedBundlePage::didReachApplicationCacheOriginQuota): Minor
1718         refactoring to share code with didExceedDatabaseQuota.
1719         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
1720         (InjectedBundlePage):
1721         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
1722         (WTR::TestRunner::TestRunner):
1723         * WebKitTestRunner/InjectedBundle/TestRunner.h:
1724         (WTR::TestRunner::dumpDatabaseCallbacks):
1725         (WTR::TestRunner::shouldDumpDatabaseCallbacks):
1726         (TestRunner):
1727
1728 2012-09-14  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1729
1730         [Qt] Set force_static_libs_as_shared early enough to be caught by configure
1731
1732         Rubber-stamped by Simon Hausmann.
1733
1734         * qmake/mkspecs/features/default_post.prf:
1735
1736 2012-09-14  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1737
1738         [Qt] Remove old cruft from the qmake build system
1739
1740         Rubber-stamped by Simon Hausmann.
1741
1742         * DumpRenderTree/qt/DumpRenderTree.pro:
1743         * qmake/mkspecs/features/default_post.prf:
1744
1745 2012-09-14  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1746
1747         [Qt] Enable force_static_libs_as_shared for development
1748
1749         Not enabled for buildbots or production-builds (part of Qt5)
1750
1751         Reviewed by Simon Hausmann.
1752
1753         * qmake/mkspecs/features/default_post.prf:
1754
1755 2012-09-14  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1756
1757         [Qt] Make force_static_libs_as_shared work on Mac OS
1758
1759         We had to move a few LIBS += around that were in the wrong place,
1760         and not caught when everything was just linked into the final
1761         QtWebKit library.
1762
1763         Reviewed by Simon Hausmann.
1764
1765         * qmake/mkspecs/features/default_post.prf:
1766         * qmake/mkspecs/features/force_static_libs_as_shared.prf:
1767
1768 2012-09-14  Simon Hausmann  <simon.hausmann@nokia.com>
1769
1770         [Qt] Fix build of WTR on Windows
1771
1772         Reviewed by Tor Arne Vestbø.
1773
1774         WTR/qt/main.cpp includes TestController.h, which lives in just WTR/
1775         With qmake's unix makefile generator as well as with the MingW generator,
1776         the directory where the pro file lives (WTR/) is automatically added to
1777         the include search path, but not so with MSVC. Instead of relying on qmake's
1778         implicit feature here, add WTR/ explicitly to the include search path.
1779
1780         * WebKitTestRunner/Target.pri:
1781
1782 2012-09-14  Christophe Dumez  <christophe.dumez@intel.com>
1783
1784         WebKitTestRunner needs layoutTestController.setStopProvisionalFrameLoads
1785         https://bugs.webkit.org/show_bug.cgi?id=42691
1786
1787         Reviewed by Kenneth Rohde Christiansen.
1788
1789         Add implementation for testrunner.setStopProvisionalFrameLoads()
1790         in WebKitTestRunner.
1791
1792         Original patch by Alexey Proskuryakov.
1793
1794         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
1795         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1796         (WTR::InjectedBundlePage::didStartProvisionalLoadForFrame):
1797         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
1798         (WTR::TestRunner::TestRunner):
1799         * WebKitTestRunner/InjectedBundle/TestRunner.h:
1800         (TestRunner):
1801         (WTR::TestRunner::setStopProvisionalFrameLoads):
1802         (WTR::TestRunner::shouldStopProvisionalFrameLoads):
1803
1804 2012-09-14  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1805
1806         [Qt] Fix handling of debug/release/debug_and_release/build_all
1807
1808         We now pick up the defaults from Qt, in default_pre (CONFIG already
1809         contains the appropriate values for debug and release). We then let
1810         the command line arguments to qmake, or the project files themselves,
1811         override the configuration, and finally we sanitize the configuration
1812         in default_post.
1813
1814         Reviewed by Simon Hausmann.
1815
1816         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
1817         * Scripts/webkitdirs.pm:
1818         (buildQMakeProjects):
1819         * qmake/mkspecs/features/default_post.prf:
1820         * qmake/mkspecs/features/default_pre.prf:
1821         * qmake/mkspecs/features/mac/default_pre.prf:
1822
1823 2012-09-14  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
1824
1825         WebKitTestRunner needs layoutTestController.setTabKeyCyclesThroughElements
1826         https://bugs.webkit.org/show_bug.cgi?id=42687
1827
1828         Reviewed by Kenneth Rohde Christiansen.
1829
1830         Implement testRunner.setTabKeyCyclesThroughElements for
1831         WebKitTestRunner.
1832
1833         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
1834         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
1835         (WTR::InjectedBundle::beginTesting):
1836         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
1837         (WTR::TestRunner::setTabKeyCyclesThroughElements):
1838         (WTR):
1839         * WebKitTestRunner/InjectedBundle/TestRunner.h:
1840         (TestRunner):
1841
1842 2012-09-14  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
1843
1844         EventSendingController::keyDown does not support non-array modifier arguments
1845         https://bugs.webkit.org/show_bug.cgi?id=96727
1846
1847         Reviewed by Kenneth Rohde Christiansen.
1848
1849         Add support for handling a string as modifier argument to
1850         EventSendingController::keyDown.
1851
1852         * WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
1853         (WTR::parseModifierArray):
1854
1855 2012-09-14  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
1856
1857         WebKitTestRunner needs layoutTestController.setCustomPolicyDelegate
1858         https://bugs.webkit.org/show_bug.cgi?id=42546
1859
1860         Reviewed by Kenneth Rohde Christiansen.
1861
1862         Exported TestRunner::setCustomPolicyDelegate() method.
1863         Allowed Policy Delegate dumping even if we're not in 'waitUntilDone' mode.
1864
1865         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
1866         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1867         (WTR::InjectedBundlePage::decidePolicyForNavigationAction):
1868
1869 2012-09-14  Zan Dobersek  <zandobersek@gmail.com>
1870
1871         [GTK] Clear application cache between tests in DumpRenderTree
1872         https://bugs.webkit.org/show_bug.cgi?id=96543
1873
1874         Reviewed by Philippe Normand.
1875
1876         Call the new DumpRenderTreeSupportGtk method to clear application cache
1877         after each test in DumpRenderTree.
1878
1879         Set the XDG_CACHE_HOME environment variable when running DumpRenderTree
1880         as well. It's now being set in XvfbDriver._start so that each driver is assigned
1881         a cache directory that's constructed out of the driver's name and its worker
1882         number. These directories are located in the layout tests results directory.
1883         At least on the builders this ensures a driver-specific empty cache directory
1884         with each test run.
1885
1886         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1887         (runTest):
1888         * Scripts/webkitpy/layout_tests/port/gtk.py:
1889         (GtkPort.setup_environ_for_server):
1890         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
1891         (XvfbDriver._start):
1892
1893 2012-09-14  Simon Hausmann  <simon.hausmann@nokia.com>
1894
1895         [Qt] Fix linking on Windows
1896
1897         Reviewed by Tor Arne Vestbø.
1898
1899         When compiling code _after_ QtWebKit.dll, we need to make sure that the QWEBKIT_EXPORT
1900         macro turns into an import macro, which means QT_MAKEDLL should only be set _up until_
1901         the QtWebKit.dll creation in the build process and be off afterwards. This is done in
1902         win32/default_post.prf using a !contains(QT, webkit): DEFINES += QT_MAKEDLL.
1903
1904         r128450 sanitized the values of QT and PKGCONFIG to only contain the dependencies we
1905         want for QtWebKit.dll, but unfortunately that code was applied to any binary built also
1906         after the module. Therefore QT did never contain webkit anymore and we always passed
1907         QT_MAKEDLL to DEFINES, resulting in QWEBKIT_EXPORT to _not_ turn into an import macro.
1908
1909         This patch introduces a creating_module configuration that is used to
1910         scope the qt/pkgconfig sanitization.
1911
1912         * qmake/mkspecs/features/default_post.prf:
1913
1914 2012-09-14  Christophe Dumez  <christophe.dumez@intel.com>
1915
1916         [WK2][WKTR] TestRunner is not dumping some frame load callbacks information
1917         https://bugs.webkit.org/show_bug.cgi?id=96740
1918
1919         Reviewed by Kenneth Rohde Christiansen.
1920
1921         Dump information expected by the test cases about the following
1922         frame load callbacks:
1923         "didDisplayInsecureContentForFrame"
1924         "didRunInsecureContentForFrame"
1925         "didDetectXSSForFrame"
1926
1927         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1928         (WTR::InjectedBundlePage::willPerformClientRedirectForFrame): Print a URL
1929         as expected by test results by calling pathSuitableForTestResult() instead
1930         of printing raw URL.
1931         (WTR::InjectedBundlePage::didDisplayInsecureContentForFrame):
1932         (WTR::InjectedBundlePage::didRunInsecureContentForFrame):
1933         (WTR::InjectedBundlePage::didDetectXSSForFrame):
1934
1935 2012-09-14  Peter Beverloo  <peter@chromium.org>
1936
1937         The runtime/unsigned category should be valid in the cpp style-checker.
1938         https://bugs.webkit.org/show_bug.cgi?id=96748
1939
1940         Reviewed by Csaba Osztrogonác.
1941
1942         This is causing the webkitpy tests to fail because it's not listed in the
1943         category array. Furthermore, change an "int" to "long" because it's
1944         throwing a double error in one of the earlier asserts.
1945
1946         * Scripts/webkitpy/style/checkers/cpp.py:
1947         (CppChecker):
1948         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
1949         (WebKitStyleTest.test_names):
1950
1951 2012-09-13  Kenneth Rohde Christiansen  <kenneth@webkit.org>
1952
1953         Evas_Object* is a ref'ed structure, so tread it as such
1954         https://bugs.webkit.org/show_bug.cgi?id=96659
1955
1956         Reviewed by Adam Barth.
1957
1958         Replace OwnPtr<Evas_Object> with RefPtr.
1959
1960         * DumpRenderTree/efl/ImageDiff.cpp:
1961         (calculateDifference):
1962         (printImageDifferences):
1963         (readImageFromStdin):
1964         (main):
1965
1966 2012-09-14  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
1967
1968         Unreviewed, rolling out r128507.
1969         http://trac.webkit.org/changeset/128507
1970         https://bugs.webkit.org/show_bug.cgi?id=96659
1971
1972         Revert. r128507 makes too many crashes in EFL layout test bots.
1973
1974         * DumpRenderTree/efl/ImageDiff.cpp:
1975         (calculateDifference):
1976         (printImageDifferences):
1977         (readImageFromStdin):
1978         (main):
1979
1980 2012-09-13  Kevin Funk  <kevin.funk@kdab.com>
1981
1982         Make compile with both OS(WINCE) and PLATFORM(QT) support
1983         https://bugs.webkit.org/show_bug.cgi?id=95536
1984
1985         Reviewed by Simon Hausmann.
1986
1987         Fix wince support in qmake files
1988
1989         * Tools.pro:
1990         * qmake/mkspecs/features/configure.prf:
1991         * qmake/mkspecs/features/default_post.prf:
1992         * qmake/mkspecs/features/features.prf:
1993         * qmake/mkspecs/features/functions.prf:
1994
1995 2012-09-13  KwangYong Choi  <ky0.choi@samsung.com>
1996
1997         [WK2] Add color picker API support for WebKit2
1998         https://bugs.webkit.org/show_bug.cgi?id=95058
1999
2000         Reviewed by Gyuyoung Kim.
2001
2002         Initialization code for color picker callbacks in UI client.
2003
2004         * MiniBrowser/mac/BrowserWindowController.m:
2005         (-[BrowserWindowController awakeFromNib]):
2006         * MiniBrowser/win/BrowserView.cpp:
2007         (BrowserView::create):
2008         * WebKitTestRunner/TestController.cpp:
2009         (WTR::TestController::createOtherPage):
2010         (WTR::TestController::initialize):
2011
2012 2012-09-13  Yuta Kitamura  <yutak@chromium.org>
2013
2014         Upgrade yutak to reviewer
2015         https://bugs.webkit.org/show_bug.cgi?id=96721
2016
2017         Reviewed by Kent Tamura.
2018
2019         * Scripts/webkitpy/common/config/committers.py:
2020
2021 2012-09-13  Adenilson Cavalcanti  <cavalcantii@gmail.com>
2022
2023         Make the style checker report usage of "unsigned int"
2024         https://bugs.webkit.org/show_bug.cgi?id=96693
2025
2026         Reviewed by Dirk Pranke.
2027
2028         Add a rule to check for uses of unsigned int, as the updated coding style recomends.
2029
2030         * Scripts/webkitpy/style/checkers/cpp.py:
2031         (check_language):
2032
2033 2012-09-13  Philip Rogers  <pdr@google.com>
2034
2035         Add gender-neutral form of webkit-patch land-cowboy
2036         https://bugs.webkit.org/show_bug.cgi?id=96709
2037
2038         Reviewed by Adam Barth.
2039
2040         Add webkit-patch land-cowhand, the confident and courageous form of webkit-patch for everyone.
2041
2042         * Scripts/webkitpy/tool/commands/download.py:
2043         (LandCowhand):
2044
2045 2012-09-13  Seokju Kwon  <seokju.kwon@samsung.com>
2046
2047         [EFL] Remove some parameters in browserCreate()
2048         https://bugs.webkit.org/show_bug.cgi?id=96499
2049
2050         Reviewed by Kenneth Rohde Christiansen.
2051
2052         There are too many parameters in browserCreate(). Lots of it seems like settings from command line argument.
2053         And I have moved some parameters into User_Arguments.
2054
2055         * EWebLauncher/main.c:
2056         (_User_Arguments):
2057         (_ELauncher):
2058         (on_key_down):
2059         (browserCreate):
2060         (findThemePath):
2061         (parseUserArguments):
2062         (main):
2063
2064 2012-09-13  Mark Lam  <mark.lam@apple.com>
2065
2066         Unreviewed. Adding myself to the committers list.
2067
2068         * Scripts/webkitpy/common/config/committers.py:
2069
2070 2012-09-13  Ryosuke Niwa  <rniwa@webkit.org>
2071
2072         run-perf-tests output cryptic error when the config file is missing.
2073         https://bugs.webkit.org/show_bug.cgi?id=96453
2074
2075         Reviewed by Tony Chang.
2076
2077         Add a special error message when a configuration file is missing.
2078
2079         Also update the help message of --source-json-path to signify the fact it specifies
2080         the configuration file on performance tests bots.
2081
2082         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2083         (PerfTestsRunner._parse_args):
2084         (PerfTestsRunner._generate_and_show_results):
2085         (PerfTestsRunner._merge_slave_config_json):
2086         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
2087         (_test_run_with_json_output): Make upload to fail by default and assert the result
2088         in the function so that we can return logs instead.
2089         (_test_run_with_json_output.mock_upload_json):
2090         (test_run_with_json_output):
2091         (test_run_with_description):
2092         (test_run_respects_no_results):
2093         (test_run_with_slave_config_json):
2094         (test_run_with_bad_slave_config_json):
2095         (test_run_with_multiple_repositories):
2096
2097 2012-09-13  Kenneth Rohde Christiansen  <kenneth@webkit.org>
2098
2099         Evas_Object* is a ref'ed structure, so tread it as such
2100         https://bugs.webkit.org/show_bug.cgi?id=96659
2101
2102         Reviewed by Adam Barth.
2103
2104         Replace OwnPtr<Evas_Object> with RefPtr.
2105
2106         * DumpRenderTree/efl/ImageDiff.cpp:
2107         (calculateDifference):
2108         (printImageDifferences):
2109         (readImageFromStdin):
2110         (main):
2111
2112 2012-09-13  Csaba Osztrogonác  <ossy@webkit.org>
2113
2114         One more unreviewed trivial fix after r128399.
2115
2116         * BuildSlaveSupport/build.webkit.org-config/config.json: Fix URLs of the SVN mirror server.
2117
2118 2012-09-13  Csaba Osztrogonác  <ossy@webkit.org>
2119
2120         Unreviewed trivial fix after r128399.
2121
2122         * BuildSlaveSupport/wait-for-SVN-server.py: Renamed from Tools/BuildSlaveSupport/build.webkit.org-config/wait-for-SVN-server.py.
2123         (getLatestSVNRevision):
2124         (waitForSVNRevision):
2125
2126 2012-09-13  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
2127
2128         [WK2][WTR] WebKitTestRunner needs testRunner.callShouldCloseOnWebView
2129         https://bugs.webkit.org/show_bug.cgi?id=96366
2130
2131         Reviewed by Anders Carlsson.
2132
2133         Implement testRunner.callShouldCloseOnWebView for WebKitTestRunner
2134         by calling shouldClose() on the FrameLoader.
2135
2136         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
2137         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
2138         (WTR::TestRunner::callShouldCloseOnWebView):
2139         (WTR):
2140         * WebKitTestRunner/InjectedBundle/TestRunner.h:
2141         (TestRunner):
2142
2143 2012-09-13  Sheriff Bot  <webkit.review.bot@gmail.com>
2144
2145         Unreviewed, rolling out r128453.
2146         http://trac.webkit.org/changeset/128453
2147         https://bugs.webkit.org/show_bug.cgi?id=96681
2148
2149         Having tests use the same appcache directory leads to timeouts
2150         (Requested by zdobersek on #webkit).
2151
2152         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2153         (runTest):
2154         * Scripts/webkitpy/layout_tests/port/gtk.py:
2155         (GtkPort.setup_environ_for_server):
2156
2157 2012-09-13  James Robinson  <jamesr@chromium.org>
2158
2159         [chromium] Temporarily remove Android Builder (dbg) from garden-o-matic view
2160         https://bugs.webkit.org/show_bug.cgi?id=96678
2161
2162         Reviewed by Dirk Pranke.
2163
2164         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
2165         (.):
2166
2167 2012-09-13  Rob Buis  <rbuis@rim.com>
2168
2169         Unreviewed changes to watchlist.
2170
2171         - Add myself to BlackBerry bugs
2172
2173         * Scripts/webkitpy/common/config/watchlist:
2174
2175 2012-09-13  Stephen Chenney  <schenney@chromium.org>
2176
2177         Add an SVG entry to the watchlist
2178
2179         Unreviewed config change.
2180
2181         * Scripts/webkitpy/common/config/watchlist:
2182         - Created a new SVG watchlist label covering WebCore/svg and WebCore/rendering/svg
2183         - Added myself and pdr as watchers
2184
2185 2012-09-13  Simon Hausmann  <simon.hausmann@nokia.com>
2186
2187         [Qt] Make WebKit2 work on Windows with Qt 5
2188         https://bugs.webkit.org/show_bug.cgi?id=76776
2189
2190         Reviewed by Tor Arne Vestbø.
2191
2192         WebKit2 should compile with Qt on Windows now, so enable it. This patch in fact corrects
2193         the check to disable WebKit when we don't have USE_3D_GRAPHICS available, because we need
2194         that for the GL TextureMapper (it's only built when that feature is set).
2195
2196         * qmake/mkspecs/features/configure.prf:
2197
2198 2012-09-13  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2199
2200         [Qt] Ensure that QT is finalized before loading qt_module.prf
2201
2202         As qt_module.prf does dependency tracking based on the content of the
2203         QT variable. The intermediate WebKit modules modify the variable in
2204         their .pri files, so we have to ensure we've loaded all the modules
2205         before loading qt_module.
2206
2207         Reviewed by Tor Arne Vestbø.
2208         Patch by Oswald Buddenhagen <oswald.buddenhagen@nokia.com> on 2012-09-13
2209
2210         * qmake/mkspecs/features/default_post.prf:
2211         * qmake/mkspecs/features/webkit_modules.prf: Added.
2212
2213 2012-09-13  Simon Hausmann  <simon.hausmann@nokia.com>
2214
2215         [Qt] Fix build of Qt WTR/MiniBrowser with MSVC and nmake
2216         https://bugs.webkit.org/show_bug.cgi?id=96633
2217
2218         Reviewed by Csaba Osztrogonác.
2219
2220         WTR and MiniBrowser both have DRT in their include path (for the shared font stuff). Batch
2221         builds with MSVC and nmake result in nmake picking up main.cpp from the wrong directory.
2222         It's the usual problem and the usual solution is to rename one of the conflicting files.
2223         This patch chooses to rename DRT instead of WTR and MBR.
2224
2225         Fixed include order at the same time to match style guide.
2226
2227         * DumpRenderTree/qt/DumpRenderTree.pro:
2228         * DumpRenderTree/qt/DumpRenderTreeMain.cpp: Renamed from Tools/DumpRenderTree/qt/main.cpp.
2229         (messageHandler):
2230         (isOption):
2231         (takeOptionValue):
2232         (printUsage):
2233         (main):
2234
2235 2012-09-13  Kenneth Rohde Christiansen  <kenneth@webkit.org>
2236
2237         The Script run-launcher should support -2 in combination with --efl
2238         https://bugs.webkit.org/show_bug.cgi?id=96639
2239
2240         Reviewed by Simon Hausmann.
2241
2242         Support running the MiniBrowser then supplied -2 as argument.
2243
2244         * Scripts/run-launcher:
2245
2246 2012-09-13  Raphael Kubo da Costa  <raphael.kubo.da.costa@intel.com>
2247
2248         Unreviewed, add my Intel email address to the list.
2249
2250         * Scripts/webkitpy/common/config/committers.py:
2251
2252 2012-09-13  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2253
2254         [Qt] Don't overwrite LIBS_PRIVATE when sanitizing LIBS
2255
2256         Reviewed by Ossy.
2257
2258         * qmake/mkspecs/features/default_post.prf:
2259
2260 2012-09-13  Kenneth Rohde Christiansen  <kenneth@webkit.org>
2261
2262         [EFL][DRT] Implement LayoutTestController::layerTreeAsText
2263         https://bugs.webkit.org/show_bug.cgi?id=82294
2264
2265         Reviewed by Gyuyoung Kim.
2266
2267         Implement layerTreeAsText using the DumpRenderTreeSupportEfl method.
2268
2269         * DumpRenderTree/efl/TestRunnerEfl.cpp:
2270         (TestRunner::layerTreeAsText):
2271
2272 2012-09-13  Zan Dobersek  <zandobersek@gmail.com>
2273
2274         [GTK] Clear application cache between tests in DumpRenderTree
2275         https://bugs.webkit.org/show_bug.cgi?id=96543
2276
2277         Reviewed by Philippe Normand.
2278
2279         Call the new DumpRenderTreeSupportGtk method to clear application cache
2280         after each test.
2281
2282         Set the XDG_CACHE_HOME environment variable when running DumpRenderTree
2283         as well. Currently it points to a subdirectory in the layout tests results
2284         directory. At least on the builders this ensures an empty cache directory
2285         with each test run.
2286
2287         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2288         (runTest):
2289         * Scripts/webkitpy/layout_tests/port/gtk.py:
2290         (GtkPort.setup_environ_for_server):
2291
2292 2012-09-13  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2293
2294         [Qt] Make sure LIBS, PKGCONFIG, and QT, do not affect QtWebKit's prl file
2295
2296         Unless it's actually a dependency of the public QtWebKit API.
2297
2298         We could replace every occurance of LIBS, PKGCONFIG, and QT, in the pri
2299         files with their _PRIVATE equivivalent, but that's likely to break when
2300         someone adds a new QT/LIBS/PKGCONFIG += foo line somewhere. Instead we
2301         clean up the variables in default_post.
2302
2303         This means that the CONFIG -= explicitlib in linkAgainstLibrary() is no
2304         longer needed, as it was a workaround for the intermediate libraries ending
2305         up in the prl file. And, since CONFIG -= staticlib was there to support
2306         the explicitlib option, we can remove that too (as well as the exception
2307         for gprof, since it would be empty). If gprof needs tweaks to always link
2308         statically they should go in gprof.prf.
2309
2310         Reviewed by Simon Hausmann.
2311
2312         * qmake/mkspecs/features/default_post.prf:
2313         * qmake/mkspecs/features/functions.prf:
2314
2315 2012-09-13  Mario Sanchez Prada  <msanchez@igalia.com>
2316
2317         [GTK] "Infinite" loop in AccessibilityUIElementGtk.cpp
2318         https://bugs.webkit.org/show_bug.cgi?id=96632
2319
2320         Reviewed by Carlos Garcia Campos.
2321
2322         Fix this by using atk_object_get_n_accessible_children instead of
2323         calling getChildren() from childCount.
2324
2325         * WebKitTestRunner/InjectedBundle/gtk/AccessibilityUIElementGtk.cpp:
2326         (WTR::AccessibilityUIElement::childrenCount): Avoid the infinite
2327         loop by using atk_object_get_n_accessible_children().
2328
2329 2012-09-13  Simon Hausmann  <simon.hausmann@nokia.com>
2330
2331         [Qt][Win] Enable USE(3D_GRAPHICS)
2332         https://bugs.webkit.org/show_bug.cgi?id=90879
2333
2334         Reviewed by Kenneth Rohde Christiansen.
2335
2336         Enable it on Windows as well as on QNX (bug fixed).
2337
2338         * qmake/mkspecs/features/features.prf:
2339
2340 2012-09-13  Simon Hausmann  <simon.hausmann@nokia.com>
2341
2342         [Qt] The ANGLE build on Windows breaks with GnuWin32's flex because it is too old
2343         https://bugs.webkit.org/show_bug.cgi?id=96359
2344
2345         Reviewed by Tor Arne Vestbø.
2346
2347         GnuWin32's flex is 2.5.4 and that is too old for ANGLE's preprocessor
2348         lexer. The only viable alternative I could find on Windows that provides
2349         binaries is the winflexbison distribution (http://sourceforge.net/projects/winflexbison/),
2350         so let's use that one instead.
2351
2352         * Scripts/webkitdirs.pm:
2353         (checkRequiredSystemConfig):
2354         * qmake/mkspecs/features/default_pre.prf:
2355
2356 2012-09-13  Simon Hausmann  <simon.hausmann@nokia.com>
2357
2358         [Qt] Add configuration for production builds
2359         https://bugs.webkit.org/show_bug.cgi?id=96607
2360
2361         Reviewed by Kenneth Rohde Christiansen.
2362
2363         Add configuration tweaks for production builds, to ease memory pressure with the GNU toolchain
2364         and enable all-in-one-files optimizations.
2365
2366         * qmake/mkspecs/features/production_build.prf: Added.
2367
2368 2012-09-13  Simon Hausmann  <simon.hausmann@nokia.com>
2369
2370         [Qt] GStreamer detection does not work when cross-compiling
2371         https://bugs.webkit.org/show_bug.cgi?id=96608
2372
2373         Reviewed by Kenneth Rohde Christiansen.
2374
2375         When cross-compiling it is not safe to just use system(pkg-config),
2376         because there is a high probability that it will find packages in the
2377         host system instead of in the target one. Instead use qmake's new
2378         packagesExist() function that takes care of this delicate aspect of
2379         cross compilation by setting PKG_CONFIG_SYSROOT_DIR and
2380         PKG_CONFIG_LIBDIR accordingly.
2381
2382         * qmake/mkspecs/features/features.prf:
2383
2384 2012-09-13  Christophe Dumez  <christophe.dumez@intel.com>
2385
2386         [WK2][WTR] More efficient string handling in InjectedBundlePage
2387         https://bugs.webkit.org/show_bug.cgi?id=96609
2388
2389         Reviewed by Kenneth Rohde Christiansen.
2390
2391         Update StringBuilder usage in InjectedBundlePage to follow
2392         the latest recommandations for efficient string handling:
2393         http://trac.webkit.org/wiki/EfficientStrings
2394
2395         This involves using appendLiteral() or appendNumber()
2396         instead of append() whenever possible.
2397
2398         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
2399         (WTR::dumpPath):
2400         (WTR::rangeToStr):
2401         (WTR::styleDecToStr):
2402         (WTR::frameToStr):
2403         (WTR::dumpResourceURL):
2404         (WTR):
2405         (WTR::dumpFrameDescriptionSuitableForTestResult):
2406         (WTR::dumpRequestDescriptionSuitableForTestResult):
2407         (WTR::dumpResponseDescriptionSuitableForTestResult):
2408         (WTR::dumpErrorDescriptionSuitableForTestResult):
2409         (WTR::InjectedBundlePage::didReceiveIntentForFrame):
2410         (WTR::InjectedBundlePage::registerIntentServiceForFrame):
2411         (WTR::InjectedBundlePage::didStartProvisionalLoadForFrame):
2412         (WTR::InjectedBundlePage::didReceiveServerRedirectForProvisionalLoadForFrame):
2413         (WTR::InjectedBundlePage::didCommitLoadForFrame):
2414         (WTR::InjectedBundlePage::didFinishProgress):
2415         (WTR::dumpFrameScrollPosition):
2416         (WTR::dumpFrameText):
2417         (WTR::dumpDescendantFramesText):
2418         (WTR::InjectedBundlePage::didFinishLoadForFrame):
2419         (WTR::InjectedBundlePage::didReceiveTitleForFrame):
2420         (WTR::InjectedBundlePage::didCancelClientRedirectForFrame):
2421         (WTR::InjectedBundlePage::willPerformClientRedirectForFrame):
2422         (WTR::InjectedBundlePage::didFinishDocumentLoadForFrame):
2423         (WTR::InjectedBundlePage::didHandleOnloadEventsForFrame):
2424         (WTR::InjectedBundlePage::willSendRequestForFrame):
2425         (WTR::InjectedBundlePage::didReceiveResponseForResource):
2426         (WTR::InjectedBundlePage::didFinishLoadForResource):
2427         (WTR::InjectedBundlePage::didFailLoadForResource):
2428         (WTR::InjectedBundlePage::shouldCacheResponse):
2429         (WTR::InjectedBundlePage::decidePolicyForNavigationAction):
2430         (WTR::InjectedBundlePage::willAddMessageToConsole):
2431         (WTR::InjectedBundlePage::willSetStatusbarText):
2432         (WTR::InjectedBundlePage::willRunJavaScriptAlert):
2433         (WTR::InjectedBundlePage::willRunJavaScriptConfirm):
2434         (WTR::InjectedBundlePage::willRunJavaScriptPrompt):
2435         (WTR::InjectedBundlePage::didReachApplicationCacheOriginQuota):
2436         (WTR::InjectedBundlePage::shouldBeginEditing):
2437         (WTR::InjectedBundlePage::shouldEndEditing):
2438         (WTR::InjectedBundlePage::shouldInsertNode):
2439         (WTR::InjectedBundlePage::shouldInsertText):
2440         (WTR::InjectedBundlePage::shouldDeleteRange):
2441         (WTR::InjectedBundlePage::shouldChangeSelectedRange):
2442         (WTR::InjectedBundlePage::shouldApplyStyle):
2443         (WTR::InjectedBundlePage::didBeginEditing):
2444         (WTR::InjectedBundlePage::didEndEditing):
2445         (WTR::InjectedBundlePage::didChange):
2446         (WTR::InjectedBundlePage::didChangeSelection):
2447         (WTR::InjectedBundlePage::supportsFullScreen):
2448         (WTR::InjectedBundlePage::enterFullScreenForElement):
2449         (WTR::InjectedBundlePage::exitFullScreenForElement):
2450         (WTR::InjectedBundlePage::beganEnterFullScreen):
2451         (WTR::InjectedBundlePage::beganExitFullScreen):
2452         (WTR::InjectedBundlePage::closeFullScreen):
2453         (WTR::dumpBackForwardListItem):
2454         (WTR::InjectedBundlePage::dumpBackForwardList):
2455
2456 2012-09-13  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
2457
2458         [Qt][WK2] Memory leaks in Qt's TestRunner
2459         https://bugs.webkit.org/show_bug.cgi?id=96603
2460
2461         Reviewed by Csaba Osztrogonác.
2462
2463         Fix memory leaks in Qt's TestRunner code by adopting WKString
2464         created with WKStringCreateWithUTF8CString().
2465
2466         * WebKitTestRunner/qt/TestControllerQt.cpp:
2467         (WTR::TestController::initializeInjectedBundlePath):
2468         (WTR::TestController::initializeTestPluginDirectory):
2469
2470 2012-09-13  Zan Dobersek  <zandobersek@gmail.com>
2471
2472         [GTK][NRWT] Lower timeout value to 6 seconds (12 for debug configuration)
2473         https://bugs.webkit.org/show_bug.cgi?id=96506
2474
2475         Reviewed by Dirk Pranke.
2476
2477         Lower the timeout value to 6 seconds for release configuration and 12 seconds
2478         for debug configuration. These are the same values the Chromium port uses.
2479
2480         Currently the value is lowered only for the DumpRenderTree, WebKitTestRunner (and
2481         the layout tests under WebKit2 generally) requires more attention and gardening
2482         than what the current effort produces.
2483
2484         The unit test is updated as well to test these changes.
2485
2486         * Scripts/webkitpy/layout_tests/port/gtk.py:
2487         (GtkPort.default_timeout_ms):
2488         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
2489         (GtkPortTest.test_default_timeout_ms):
2490         (GtkPortTest.assertLinesEqual):
2491
2492 2012-09-12  Csaba Osztrogonác  <ossy@webkit.org>
2493
2494         Add SVN mirror handling feature to build.webkit.org
2495         https://bugs.webkit.org/show_bug.cgi?id=85887
2496
2497         Reviewed by Dirk Pranke.
2498
2499         * BuildSlaveSupport/build.webkit.org-config/config.json: Add SVN mirrors for Qt buildslaves hosted in Szeged.
2500         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2501         (ConfigureBuild.__init__): Add SVNMirror property to be able to watch which slaves use which mirror on build.webkit.org.
2502         (ConfigureBuild.start):
2503         (CheckOutSource.__init__): Set baseURL to the SVNMirror or the default http://svn.webkit.org/repository/webkit/
2504         (WaitForSVNServer): Add new buildstep for waiting the SVNMirror to be in sync with http://svn.webkit.org/repository/webkit/
2505         (Factory.__init__):
2506         (BuildFactory.__init__):
2507         (TestFactory.__init__):
2508         (BuildAndTestFactory.__init__):
2509         (BuildAndPerfTestFactory.__init__):
2510         (BuildAndPerfTestWebKit2Factory.__init__):
2511         (DownloadAndPerfTestFactory.__init__):
2512         (DownloadAndPerfTestWebKit2Factory.__init__):
2513         (loadBuilderConfig): Use kwargs instead of args to be able to add the optional SVNMirror factory argument.
2514         * BuildSlaveSupport/build.webkit.org-config/mastercfg_unittest.py:
2515         (SVNMirrorTest):
2516         (SVNMirrorTest.setUp): Load config.json once at startup.
2517         (SVNMirrorTest.get_SVNMirrorFromConfig): Get the SVNMirror from config.json for a given buildslave.
2518         (SVNMirrorTest.test_CheckOutSource): Compare CheckOutSource.baseURL with SVNMirror in config.json for all builders
2519         * BuildSlaveSupport/build.webkit.org-config/wait-for-SVN-server.py: Added.
2520         (getLatestSVNRevision): Get the latest SVN revison from the given server.
2521         (waitForSVNRevision): Wait until the given SVN revision is committed to the given server. It doesn't wait if the
2522                               SVN revision is empty (force build triggered without revision) or the server is unavailable.
2523
2524 2012-09-12  Kangil Han  <kangil.han@samsung.com>
2525
2526         [WK2][WTR] Set waitUntilDone watchdog timer value equal to WK1.
2527         https://bugs.webkit.org/show_bug.cgi?id=81606
2528
2529         Reviewed by Tim Horton.
2530
2531         6 seconds is harsh for some jquery test cases.
2532         Therefore, adjust the value to match DumpRenderTree.
2533
2534         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
2535         (WTR):
2536
2537 2012-09-12  KyungTae Kim  <ktf.kim@samsung.com>
2538
2539         [EFL] Support download attribute feature
2540         https://bugs.webkit.org/show_bug.cgi?id=96462
2541
2542         Reviewed by Gyuyoung Kim.
2543
2544         Enable DOWNLOAD_ATTRIBUTE feature for EFL port,
2545         and add "download,request" callback to the DumpRenderTree for EFL.
2546
2547         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
2548         (DumpRenderTreeChrome::createView):
2549         (DumpRenderTreeChrome::onDownloadRequest):
2550         * DumpRenderTree/efl/DumpRenderTreeChrome.h:
2551         (DumpRenderTreeChrome):
2552         * Scripts/webkitperl/FeatureList.pm:
2553
2554 2012-09-12  Stephanie Lewis  <slewis@apple.com>
2555
2556         Reduce parallism on the wk2 testers.
2557         https://bugs.webkit.org/show_bug.cgi?id=95906
2558
2559         Reviewed by Dirk Pranke.
2560
2561         The wk2 testers on Mountain Lion are getting stuck and timing out.  The problem appears to be due to
2562         resource contention.  Reducing the number of processes alleviates the issue.  
2563
2564         Starting by reducing 25%.
2565
2566         * Scripts/webkitpy/layout_tests/port/mac.py:
2567         (MacPort.default_child_processes):
2568
2569 2012-09-12  Brady Eidson  <beidson@apple.com>
2570
2571         Assert in NetscapePlugin::destroy() with async plugin init
2572         <rdar://problem/12277595> and https://bugs.webkit.org/show_bug.cgi?id=96576
2573
2574         Reviewed by Anders Carlsson.
2575
2576         Expose NPN_Invoke to plug-in tests:
2577         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
2578         (PluginTest::NPN_Invoke):
2579         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
2580         (PluginTest):
2581
2582         Add a test that uses NPN_Invoke on the window object from inside NPP_New to remove the plug-in element:
2583         * DumpRenderTree/TestNetscapePlugIn/Tests/InvokeDestroysPluginWithinNPP_New.cpp: Added.
2584         (InvokeDestroysPluginWithinNPP_New):
2585         (InvokeDestroysPluginWithinNPP_New::InvokeDestroysPluginWithinNPP_New):
2586         (InvokeDestroysPluginWithinNPP_New::NPP_New):
2587         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
2588
2589 2012-09-12  Dirk Pranke  <dpranke@chromium.org>
2590
2591         refactor TestExpectations tokenization slightly in preparation for the new syntax
2592         https://bugs.webkit.org/show_bug.cgi?id=96564
2593
2594         Reviewed by Ojan Vafai.
2595
2596         This change changes how we will tokenize/lex the
2597         TestExpectations files, in preparation for supporting both the
2598         old syntax and the new one. This change by itself should be a
2599         no-op.
2600
2601         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
2602         (TestExpectationParser._tokenize_line):
2603         (TestExpectationParser):
2604         (TestExpectationParser._tokenize_line_using_old_format):
2605         (TestExpectationParser._tokenize_line_using_new_format):
2606
2607 2012-09-12  Dirk Pranke  <dpranke@chromium.org>
2608
2609         remove unused testexpectations editing code from webkitpy, garden-o-matic
2610         https://bugs.webkit.org/show_bug.cgi?id=96562
2611
2612         Reviewed by Adam Barth.
2613
2614         This code was prototyped but never fully realized. Removing
2615         until it reanimates :).
2616
2617         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout.js:
2618         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout_unittests.js:
2619         * Scripts/webkitpy/layout_tests/controllers/test_expectations_editor.py: Removed.
2620         * Scripts/webkitpy/layout_tests/controllers/test_expectations_editor_unittest.py: Removed.
2621         * Scripts/webkitpy/tool/servers/gardeningserver.py:
2622         (BuildCoverageExtrapolator.extrapolate_test_configurations):
2623         (GardeningHTTPRequestHandler._run_webkit_patch):
2624         (GardeningHTTPRequestHandler.ping):
2625         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
2626         (BuildCoverageExtrapolatorTest.test_extrapolate):
2627         (GardeningServerTest.test_rebaseline_new_port):
2628
2629 2012-09-12  Lucas Forschler  <lforschler@apple.com>
2630
2631         Remove unused bot, apple-xserve-1.
2632         https://bugs.webkit.org/show_bug.cgi?id=96570
2633         
2634         Reviewed by Stephanie Lewis.
2635
2636         * BuildSlaveSupport/build.webkit.org-config/config.json:
2637
2638 2012-09-12  Xan Lopez  <xlopez@igalia.com>
2639
2640         [GTK] We attempt to rebase documentation even if it's not present
2641         https://bugs.webkit.org/show_bug.cgi?id=96553
2642
2643         Reviewed by Martin Robinson.
2644
2645         Do not make the documentation rebase step fatal. This allows make
2646         install to succeed when there's no documentation generated.
2647
2648         * gtk/generate-gtkdoc:
2649
2650 2012-09-12  Kenneth Rohde Christiansen  <kenneth@webkit.org>
2651
2652         Respect WEBKITOUTPUTDIR when running EFL tests
2653         https://bugs.webkit.org/show_bug.cgi?id=96528
2654
2655         Reviewed by Dirk Pranke.
2656
2657         Expose user set WEBKITOUTPUTDIR to the web process.
2658
2659         * Scripts/webkitpy/layout_tests/port/driver.py:
2660         (Driver._start): Add WEBKITOUTPUTDIR to the environment
2661         of the web process and its potential jhbuild wrapper.
2662
2663 2012-09-12  Dirk Pranke  <dpranke@chromium.org>
2664
2665         webkitdirs: fix uname version handling for cygwin
2666         https://bugs.webkit.org/show_bug.cgi?id=96436
2667
2668         Reviewed by Jon Honeycutt.
2669
2670         Newer versions of cygwin embed an additional version string
2671         inside parentheses, so you get "1.7.16(0.249/5/3)" instead of "1.7.16".
2672         Update the code to handle that.
2673
2674         * Scripts/webkitdirs.pm:
2675         (setupAppleWinEnv):
2676
2677 2012-09-12  Christophe Dumez  <christophe.dumez@intel.com>
2678
2679         Regression(r128338): Broke Windows build
2680         https://bugs.webkit.org/show_bug.cgi?id=96537
2681
2682         Unreviewed build fix.
2683
2684         Fix Apple-Win build by allocating array dynamically
2685         since its size is not constant.
2686
2687         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
2688         (WTR::stringArrayToJS):
2689
2690 2012-09-12  Alex Sakhartchouk  <alexst@chromium.org>
2691
2692         [chromium] Add a virtual test suite for fast/hidpi
2693         https://bugs.webkit.org/show_bug.cgi?id=90192
2694
2695         Reviewed by Dirk Pranke.
2696
2697         Add a virtual test suite to make sure the pixel tests in fast/hidpi give the same result
2698         on the hardware accelerated path as the software path.
2699
2700         * DumpRenderTree/chromium/WebPreferences.cpp:
2701         (WebPreferences::applyTo):
2702         * Scripts/webkitpy/layout_tests/port/chromium.py:
2703         (ChromiumPort.virtual_test_suites):
2704
2705 2012-09-12  Adam Klein  <adamk@chromium.org>
2706
2707         [chromium] Add content_browsertests to the flakiness dashboard
2708         https://bugs.webkit.org/show_bug.cgi?id=96535
2709
2710         Reviewed by Ojan Vafai.
2711
2712         * TestResultServer/static-dashboards/dashboard_base.js:
2713
2714 2012-09-12  Christophe Dumez  <christophe.dumez@intel.com>
2715
2716         [WK2][WKTR] TestRunner needs to implement originsWithApplicationCache
2717         https://bugs.webkit.org/show_bug.cgi?id=96496
2718
2719         Reviewed by Kenneth Rohde Christiansen.
2720
2721         Add implementation for originsWithApplicationCache to
2722         WebKitTestRunner.
2723
2724         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
2725         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
2726         (WTR::stringArrayToJS):
2727         (WTR):
2728         (WTR::TestRunner::originsWithApplicationCache):
2729         * WebKitTestRunner/InjectedBundle/TestRunner.h:
2730         (TestRunner):
2731
2732 2012-09-12  Simon Hausmann  <simon.hausmann@nokia.com>
2733
2734         Fix Qt/Windows build with Python3
2735         https://bugs.webkit.org/show_bug.cgi?id=96473
2736
2737         Reviewed by Csaba Osztrogonác.
2738
2739         In Python 3 print is a real function, so we must use parentheses around
2740         the function parameters. This is backwards compatible with Python 2.
2741
2742         * Scripts/generate-win32-export-forwards:
2743
2744 2012-09-12  Simon Hausmann  <simon.hausmann@nokia.com>
2745
2746         Unreviewed trivial build fix for Qt/Windows after bug #96358.
2747
2748         As pointed out by Kevin, we should inject the "set PATH=..." statement
2749         to add the GnuWin32 directory only if that directory actually exists.
2750
2751         * qmake/mkspecs/features/default_post.prf:
2752
2753 2012-09-12  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
2754
2755         [EFL] [WK2] Memory leaks in TestControllerEfl
2756         https://bugs.webkit.org/show_bug.cgi?id=96525
2757
2758         Reviewed by Kenneth Rohde Christiansen.
2759
2760         Fix memory leaks in EFL's TestRunner code by adopting an allocation
2761         of WKString created with WKStringCreateWithUTF8CString().
2762
2763         * WebKitTestRunner/efl/TestControllerEfl.cpp:
2764         (WTR::TestController::initializeInjectedBundlePath):
2765         (WTR::TestController::initializeTestPluginDirectory):
2766
2767 2012-09-12  Christophe Dumez  <christophe.dumez@intel.com>
2768
2769         [WK2][WKTR] TestRunner needs to implement dumpApplicationCacheDelegateCallbacks
2770         https://bugs.webkit.org/show_bug.cgi?id=96374
2771
2772         Reviewed by Kenneth Rohde Christiansen.
2773
2774         Implement support for dumpApplicationCacheDelegateCallbacks
2775         and disallowIncreaseForApplicationCacheQuota in
2776         WebKitTestRunner and properly dump the information
2777         expected by the tests.
2778
2779         If the application cache quota is reached for a given
2780         security origin, WebKitTestRunner will reset the quota
2781         to its default value, unless intructed not to via
2782         disallowIncreaseForApplicationCacheQuota().
2783
2784         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
2785         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
2786         (WTR::InjectedBundlePage::InjectedBundlePage):
2787         (WTR::InjectedBundlePage::didReachApplicationCacheOriginQuota):
2788         (WTR):
2789         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
2790         (InjectedBundlePage):
2791         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
2792         (WTR::TestRunner::TestRunner):
2793         (WTR::TestRunner::disallowIncreaseForApplicationCacheQuota):
2794         (WTR):
2795         * WebKitTestRunner/InjectedBundle/TestRunner.h:
2796         (WTR::TestRunner::dumpApplicationCacheDelegateCallbacks):
2797         (TestRunner):
2798         (WTR::TestRunner::shouldDisallowIncreaseForApplicationCacheQuota):
2799         (WTR::TestRunner::shouldDumpApplicationCacheDelegateCallbacks):
2800
2801 2012-09-12  Jochen Eisinger  <jochen@chromium.org>
2802
2803         [chromium] remove deprecated and unused sets import from chromium_android driver
2804         https://bugs.webkit.org/show_bug.cgi?id=96485
2805
2806         Reviewed by Dirk Pranke.
2807
2808         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2809
2810 2012-09-12  Zan Dobersek  <zandobersek@gmail.com>
2811
2812         Flakiness dashboard doesn't recognize new Chromium Android test builder
2813         https://bugs.webkit.org/show_bug.cgi?id=96523
2814
2815         Reviewed by Ojan Vafai.
2816
2817         Properly return 'ANDROID' as the Chromium platform for Android builders.
2818
2819         * TestResultServer/static-dashboards/flakiness_dashboard.js:
2820         (chromiumPlatform):
2821         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2822         (test):
2823
2824 2012-09-12  Simon Hausmann  <simon.hausmann@nokia.com>
2825
2826         [Qt] Build on Windows requires bison/flex in PATH
2827         https://bugs.webkit.org/show_bug.cgi?id=96358
2828
2829         Reviewed by Tor Arne Vestbø.
2830
2831         The build requires flex, bison, etc. and they need to be in the PATH when building. On Mac OS X
2832         and Linux that is rarely a problem given how easily available the tools are. On Windows however
2833         a separate installation of various GNU tools is required as the operating system doesn't come with
2834         them. To make the development more convenient, Qt 5 provides a copy of the most essential tools in
2835         the gnuwin32 directory of the qt5.git top-level repository.
2836
2837         This patch tries to detect the presence of those tools and prepends them to the PATH if found.
2838
2839         This is required in preparation for the elimination of qt5/qtwebkit.pri, which currently expands
2840         PATH before calling build-webkit. It it also required for the upcoming introduction of win_flex
2841         as dependency over flex, which can be done with less hassle when qt5's gnuwin32 directory has been
2842         updated with the new tool.
2843
2844         * Scripts/webkitdirs.pm:
2845         (checkRequiredSystemConfig):
2846         * qmake/mkspecs/features/default_post.prf:
2847
2848 2012-09-12  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2849
2850         [Qt] Update build-jsc after r128174
2851
2852         Reviewed by Ossy.
2853
2854         * Scripts/build-jsc:
2855
2856 2012-09-12  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2857
2858         [Qt] Teach addStrictSubdirOrderBetween to handle more than two targets
2859
2860         By hard-coding the names of the targets we defined we ended up just
2861         redefining the previous target when using addStrictSubdirOrderBetween
2862         more than once in a single project file.
2863
2864         We now embed the two base targets into the target names.
2865
2866         Reviewed by Simon Hausmann.
2867
2868         * qmake/mkspecs/features/functions.prf:
2869
2870 2012-09-12  Philippe Liard  <pliard@google.com>
2871
2872         Depend on {base,net} GYP targets rather than {base,net}_java.
2873         https://bugs.webkit.org/show_bug.cgi?id=95690
2874
2875         Reviewed by Adam Barth.
2876
2877         {base,net}_java became 'private' targets on the Chromium side that
2878         clients should not depend on (see
2879         https://chromiumcodereview.appspot.com/10913083/).
2880         This removes a bunch of trailing white spaces additionally in
2881         TestWebkitAPI.gyp (which includes non-trailing white space changes).
2882
2883         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
2884         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
2885
2886 2012-09-12  Simon Hausmann  <simon.hausmann@nokia.com>
2887
2888         [Qt] Fix the build with ENABLE_NETSCAPE_PLUGIN_API=0
2889         https://bugs.webkit.org/show_bug.cgi?id=96494
2890
2891         Reviewed by Tor Arne Vestbø.
2892
2893         WK2's ENABLE_PLUGIN_PROCESS uses the NPAPI functions unconditionally, so disable the
2894         plugin process feature if we don't have NPAPI.
2895
2896         * qmake/mkspecs/features/features.prf:
2897
2898 2012-09-12  Sheriff Bot  <webkit.review.bot@gmail.com>
2899
2900         Unreviewed, rolling out r128280.
2901         http://trac.webkit.org/changeset/128280
2902         https://bugs.webkit.org/show_bug.cgi?id=96498
2903
2904         it broke compilation on windows debug bot (Requested by loislo
2905         on #webkit).
2906
2907         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
2908         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
2909         (WTR::InjectedBundlePage::InjectedBundlePage):
2910         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
2911         (InjectedBundlePage):
2912         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
2913         (WTR::TestRunner::TestRunner):
2914         * WebKitTestRunner/InjectedBundle/TestRunner.h:
2915         (TestRunner):
2916
2917 2012-09-12  Simon Hausmann  <simon.hausmann@nokia.com>
2918
2919         [Qt] Make it possible to build with "make release" and "make debug" on Windows
2920         https://bugs.webkit.org/show_bug.cgi?id=96488
2921
2922         Reviewed by Tor Arne Vestbø.
2923
2924         A "make debug" is passed through recursively and currently it aborts at Makefile.DerivedSources
2925         because there are no such targets. We want the generated sources to be independent from release
2926         or debug build configurations, so it is sufficient to provide fake debug and release targets that
2927         redirect to the same general-purpose target (first) of creating the derived sources.
2928
2929         * qmake/mkspecs/features/default_post.prf:
2930
2931 2012-09-12  Kenneth Rohde Christiansen  <kenneth@webkit.org>
2932
2933         [EFL] Make DumpRenderTree smarter at finding the fonts
2934         http://webkit.org/b/96281
2935
2936         Reviewed by Gyuyoung Kim.
2937
2938         Respect WEBKITOUTPUTDIR and expand the font dir from it.
2939         Use CString consistently.
2940
2941         * DumpRenderTree/efl/FontManagement.cpp:
2942         (buildPath):
2943         (getCoreFontFiles):
2944         (addFontDirectory):
2945         (addFontFiles):
2946         (getCustomBuildDir):
2947         (getPlatformFontsPath):
2948         (addFontsToEnvironment):
2949
2950 2012-09-12  Christophe Dumez  <christophe.dumez@intel.com>
2951
2952         [WK2][WKTR] TestRunner needs to implement dumpApplicationCacheDelegateCallbacks
2953         https://bugs.webkit.org/show_bug.cgi?id=96374
2954
2955         Reviewed by Kenneth Rohde Christiansen.
2956
2957         Implement support for dumpApplicationCacheDelegateCallbacks
2958         and disallowIncreaseForApplicationCacheQuota in
2959         WebKitTestRunner and properly dump the information
2960         expected by the tests.
2961
2962         If the application cache quota is reached for a given
2963         security origin, WebKitTestRunner will reset the quota
2964         to its default value, unless intructed not to via
2965         disallowIncreaseForApplicationCacheQuota().
2966
2967         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
2968         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
2969         (WTR::InjectedBundlePage::InjectedBundlePage):
2970         (WTR::InjectedBundlePage::didReachApplicationCacheOriginQuota):
2971         (WTR):
2972         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
2973         (InjectedBundlePage):
2974         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
2975         (WTR::TestRunner::TestRunner):
2976         (WTR::TestRunner::disallowIncreaseForApplicationCacheQuota):
2977         (WTR):
2978         * WebKitTestRunner/InjectedBundle/TestRunner.h:
2979         (WTR::TestRunner::dumpApplicationCacheDelegateCallbacks):
2980         (TestRunner):
2981         (WTR::TestRunner::shouldDisallowIncreaseForApplicationCacheQuota):
2982         (WTR::TestRunner::shouldDumpApplicationCacheDelegateCallbacks):
2983
2984 2012-09-12  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
2985
2986         [WK2] [WTR] WebKitTestRunner needs TestRunner.workerThreadCount
2987         https://bugs.webkit.org/show_bug.cgi?id=96388
2988
2989         Reviewed by Kenneth Rohde Christiansen.
2990
2991         Exported TestRunner.workerThreadCount as readonly attribute.
2992
2993         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
2994         * WebKitTestRunner/InjectedBundle/TestRunner.cpp: Added workerThreadCount() method implementation.
2995         (WTR::TestRunner::workerThreadCount): Returns count of worker threads.
2996         (WTR):
2997         * WebKitTestRunner/InjectedBundle/TestRunner.h: Added workerThreadCount() method. 
2998         (TestRunner):
2999
3000 2012-09-12  Jochen Eisinger  <jochen@chromium.org>
3001
3002         [chromium] Consume a user gesture when creating a new view.
3003         https://bugs.webkit.org/show_bug.cgi?id=96373
3004
3005         Reviewed by Adam Barth.
3006
3007         * DumpRenderTree/chromium/WebViewHost.cpp:
3008         (WebViewHost::createView):
3009
3010 2012-09-12  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
3011
3012         [WK2][WTR] Some of TestRunner special options are not reset before testing
3013         https://bugs.webkit.org/show_bug.cgi?id=96384
3014
3015         Reviewed by Kenneth Rohde Christiansen.
3016
3017         Now values of the following special options are reset:
3018         void setAcceptsEditing(in boolean value);
3019         void setCloseRemainingWindowsWhenComplete(in boolean value);
3020         void setXSSAuditorEnabled(in boolean value);
3021         void setAllowFileAccessFromFileURLs(in boolean value);
3022         void setPluginsEnabled(in boolean value);
3023         void setPopupBlockingEnabled(in boolean value);
3024
3025         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
3026         (WTR::InjectedBundle::beginTesting):
3027
3028 2012-09-11  Ryuan Choi  <ryuan.choi@samsung.com>
3029
3030         [CMAKE] Supply feature defines to CodeGeneratorTestRunner.
3031         https://bugs.webkit.org/show_bug.cgi?id=96273
3032
3033         Reviewed by Gyuyoung Kim.
3034
3035         Passed FEATURE_DEFINES_WITH_SPACE_SEPARATOR to CodeGeneratorTestRunner
3036         in order to distinguish compile time options such as ENABLE(TOUCH_EVENTS).
3037
3038         * WebKitTestRunner/CMakeLists.txt:
3039
3040 2012-09-11  Dirk Pranke  <dpranke@chromium.org>
3041
3042         [chromium] rename webkit_resources
3043         https://bugs.webkit.org/show_bug.cgi?id=95874
3044
3045         Reviewed by James Robinson.
3046
3047         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
3048
3049 2012-09-11  Jochen Eisinger  <jochen@chromium.org>
3050
3051         [chromium] don't include content shell builders in the garden-o-matic view
3052         https://bugs.webkit.org/show_bug.cgi?id=96418
3053
3054         Reviewed by James Robinson.
3055
3056         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
3057
3058 2012-09-11  Adam Klein  <adamk@chromium.org>
3059
3060         Garden-o-matic should ignore a wider variety of warnings in buildbot json
3061         https://bugs.webkit.org/show_bug.cgi?id=96411
3062
3063         Reviewed by Adam Barth.
3064
3065         Previously only the exact string "warning" was treated as a warning
3066         result. This patch treats any string with "warning" as a substring
3067         as a warning (e.g., "warnings", as seen on the cr-win buildbots).
3068
3069         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders.js:
3070         (.):
3071         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders_unittests.js:
3072
3073 2012-09-11  Marcelo Lira  <marcelo.lira@openbossa.org>
3074
3075         Restore original value of mock scrollbars enabled in InternalSettings
3076         https://bugs.webkit.org/show_bug.cgi?id=87680
3077
3078         Reviewed by Adam Barth.
3079
3080         Added backup support for mock scrollbars state in
3081         InternalSettings::Backup, and removed support for these
3082         from Qt's DumpRenderTree.
3083
3084         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
3085         (WebCore::WebPage::resetSettings):
3086         * DumpRenderTree/qt/TestRunnerQt.cpp:
3087         (TestRunner::setFrameFlatteningEnabled):
3088         * DumpRenderTree/qt/TestRunnerQt.h:
3089         (TestRunner):
3090
3091 2012-09-11  Dan Bernstein  <mitz@apple.com>
3092
3093         Fixed a bug where kill-old-processes failed to kill clang++.
3094
3095         Rubber-stamped by Alexey Proskuryakov.
3096
3097         * BuildSlaveSupport/kill-old-processes: Escape the + signs in "clang++"
3098         where it is passed as a regular expression to killall.
3099
3100 2012-09-11  Philippe Liard  <pliard@google.com>
3101
3102         Make sure that md5sum is not setup on non-Chromium Android platforms.
3103         https://bugs.webkit.org/show_bug.cgi?id=96393
3104
3105         Reviewed by Adam Barth.
3106
3107         Previously md5sum was setup in ChromiumAndroidDriver's constructor
3108         which is also invoked on non-Chromium Android platforms. The following patch
3109         moves this logic to _startup_test() which is guaranteed to be invoked
3110         only on Android.
3111
3112         This also updates the prompt unit test in chromium_android_unittest.py.
3113
3114         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
3115         (ChromiumAndroidDriver.__init__):
3116         (ChromiumAndroidDriver._setup_md5sum_and_push_data_if_needed):
3117         (ChromiumAndroidDriver._setup_test):
3118
3119 2012-09-11  Raphael Kubo da Costa  <rakuco@webkit.org>
3120
3121         [EFL] Rewrite the EFL-related Find modules
3122         https://bugs.webkit.org/show_bug.cgi?id=95237
3123
3124         Reviewed by Kenneth Rohde Christiansen.
3125
3126         * DumpRenderTree/efl/CMakeLists.txt: Stop setting the LINK_FLAGS
3127         property and include all the Enlightenment Foundation Libraries
3128         required by the target.
3129         * EWebLauncher/CMakeLists.txt: Ditto.
3130         * MiniBrowser/efl/CMakeLists.txt: Ditto.
3131         * TestWebKitAPI/PlatformEfl.cmake: Add missing include directories
3132         now that they are not added implicitly.
3133         * WebKitTestRunner/CMakeLists.txt: Stop setting the LINK_FLAGS
3134         property.
3135         * WebKitTestRunner/PlatformEfl.cmake: Stop setting the LINK_FLAGS
3136         property and include all the Enlightenment Foundation Libraries
3137         required by the target.
3138
3139 2012-09-11  Christophe Dumez  <christophe.dumez@intel.com>
3140
3141         [WK2][WKTR] TestRunner needs to implement setApplicationCacheOriginQuota
3142         https://bugs.webkit.org/show_bug.cgi?id=96379
3143
3144         Reviewed by Kenneth Rohde Christiansen.
3145
3146         Add implementation for setApplicationCacheOriginQuota
3147         to WebKitTestRunner.
3148
3149         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
3150         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
3151         (WTR::TestRunner::setApplicationCacheOriginQuota):
3152         (WTR):
3153         * WebKitTestRunner/InjectedBundle/TestRunner.h:
3154         (TestRunner):
3155
3156 2012-09-11  Luciano Wolf  <luciano.wolf@openbossa.org>
3157
3158         [Qt] [WK2] editing/inserting/typing-tab-designmode tests are failing
3159         https://bugs.webkit.org/show_bug.cgi?id=91988
3160
3161         Reviewed by Kenneth Rohde Christiansen.
3162
3163         Regular key events generated by Qt for the tab key fill both
3164         keyCode and keyText, so we should do the same for our fake events.
3165
3166         * WebKitTestRunner/qt/EventSenderProxyQt.cpp:
3167         (WTR::EventSenderProxy::keyDown):
3168
3169 2012-09-11  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
3170
3171         [WK2][WTR] WebKitTestRunner needs testRunner.setSpatialNavigationEnabled
3172         https://bugs.webkit.org/show_bug.cgi?id=96269
3173
3174         Reviewed by Kenneth Rohde Christiansen.
3175
3176         Exported missing testRunner.setSpatialNavigationEnabled() method.
3177
3178         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
3179         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
3180         (WTR::InjectedBundle::beginTesting):
3181         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
3182         (WTR::TestRunner::setSpatialNavigationEnabled):
3183         (WTR):
3184         * WebKitTestRunner/InjectedBundle/TestRunner.h:
3185         (TestRunner):
3186
3187 2012-09-11  Philippe Liard  <pliard@chromium.org>
3188
3189         Support LayoutTests on non-rooted devices for Chromium Android.
3190         https://bugs.webkit.org/show_bug.cgi?id=95346
3191
3192         Reviewed by Dirk Pranke.
3193
3194         We were requiring root permissions to sync resource files with the
3195         device by comparing host and device files' timestamp after we
3196         synchronized the device and host clocks.
3197         We now rely on md5sum provided by Chromium to handle data files
3198         synchronization.
3199
3200         * Scripts/webkitpy/layout_tests/port/chromium.py:
3201         (ChromiumPort._check_driver_build_up_to_date):
3202         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
3203         (ChromiumAndroidDriver.__init__):
3204         (ChromiumAndroidDriver._setup_test):
3205         (ChromiumAndroidDriver._push_executable):
3206         (ChromiumAndroidDriver._push_test_resources):
3207         (ChromiumAndroidDriver._read_prompt):
3208
3209 2012-09-11  Simon Hausmann  <simon.hausmann@nokia.com>
3210
3211         [Qt] Fix output paths on Windows
3212
3213         Reviewed by Tor Arne Vestbø.
3214
3215         BUILD_ROOT_RELATIVE_TO_OUT_PWD was calculated incorrectly Windows, resulting in an incorrect
3216         DESTDIR transformation at the end of default_post.prf. The calculation should be baased no qmake
3217         paths, which can use drive letters but always use forward slashes. So the use of QMAKE_DIR_SEP is
3218         incorrect here and instead for the calculation we should use forward slashes.
3219
3220         * qmake/mkspecs/features/default_pre.prf:
3221
3222 2012-09-11  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3223
3224         [Qt] Fix passing of defines from the build system
3225
3226         We explicitly have to pass -DENABLE_FOO=0 for every single feature
3227         that's not explicitly enabled in WEBKIT_CONFIG, since Platform.h
3228         will add it's own defaults if the features are not defined.
3229
3230         At some point we might want to run a configure-test to inspect what
3231         Platform.h will do, so we can sync up the build system's view of
3232         the feature-defines with the compile-time situation.
3233
3234         Reviewed by Simon Hausmann.
3235
3236         * qmake/mkspecs/features/functions.prf:
3237
3238 2012-09-11  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3239
3240         [Qt] Fix --qt option to build-webkit
3241
3242         Reviewed by Ossy.
3243
3244         * Scripts/webkitdirs.pm:
3245         (buildQMakeProjects):
3246
3247 2012-09-11  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3248
3249         [Qt] Add a configure step to the Qt build system
3250
3251         This allows building the Qt port using just 'qmake WebKit.pro'. Using
3252         the build-webkit script is still supported, and will add slightly more
3253         logic to the build, such as detecting the need for clean or incremental
3254         builds.
3255
3256         Internally, the build system now uses a new variable, WEBKIT_CONFIG, for
3257         all things related to configuring the build, such as use/have/enable flags,
3258         and these are translated to defines in default_post. Project files should
3259         no longer check contains(DEFINES,...) to detect features, but use the new
3260         enable?(), use?(), and have?() functions. The no_webkit1 and no_webkit2
3261         options have been translated into WEBKIT_CONFIG options as well, and can
3262         be checked using build?().
3263
3264         Reviewed by Simon Hausmann.
3265
3266         * DumpRenderTree/qt/DumpRenderTree.pro:
3267         * MiniBrowser/qt/MiniBrowser.pro:
3268         * QtTestBrowser/QtTestBrowser.pro:
3269         * Scripts/build-webkit:
3270         * Scripts/webkitdirs.pm:
3271         (qtFeatureDefaults):
3272         (buildQMakeProjects):
3273         * Tools.pro:
3274         * WebKitTestRunner/InjectedBundle/DerivedSources.pri:
3275         * WebKitTestRunner/InjectedBundle/Target.pri:
3276         * WebKitTestRunner/Target.pri:
3277         * qmake/.qmake.cache: Added.
3278         * qmake/configure.pri: Removed.
3279         * qmake/configure.pro: Removed.
3280         * qmake/dump-features: Added.
3281         * qmake/mkspecs/features/configure.prf: Added.
3282         * qmake/mkspecs/features/default_post.prf:
3283         * qmake/mkspecs/features/default_pre.prf:
3284         * qmake/mkspecs/features/features.prf:
3285         * qmake/mkspecs/features/features.pri:
3286         * qmake/mkspecs/features/functions.prf:
3287         * qmake/mkspecs/features/gprof.prf:
3288         * qmake/mkspecs/features/mac/default_pre.prf:
3289         * qmake/mkspecs/features/win32/default_pre.prf:
3290
3291 2012-09-11  Christophe Dumez  <christophe.dumez@intel.com>
3292
3293         [WK2][WKTR] TestRunner needs to implement clearApplicationCacheForOrigin
3294         https://bugs.webkit.org/show_bug.cgi?id=96372
3295
3296         Reviewed by Kenneth Rohde Christiansen.
3297
3298         Implement clearApplicationCacheForOrigin() in WebKitTestRunner.
3299
3300         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
3301         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
3302         (WTR::TestRunner::clearApplicationCacheForOrigin):
3303         (WTR):
3304         * WebKitTestRunner/InjectedBundle/TestRunner.h:
3305         (TestRunner):
3306
3307 2012-09-11  Peter Beverloo  <peter@chromium.org>
3308
3309         Don't ignore .pak files when making an archive for Android
3310         https://bugs.webkit.org/show_bug.cgi?id=96375
3311
3312         Reviewed by Jochen Eisinger.
3313
3314         The .pak files do need to be pushed to the device, which is currently
3315         preventing the layout test runner from correctly pushing all files.
3316
3317         * BuildSlaveSupport/built-product-archive:
3318         (archiveBuiltProduct):
3319
3320 2012-09-11  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
3321
3322         [WK2][WTR] WebKitTestRunner needs layoutTestController.setMinimumTimerInterval
3323         https://bugs.webkit.org/show_bug.cgi?id=96256
3324
3325         Reviewed by Kenneth Rohde Christiansen.
3326
3327         Exported TestRunner::setMinimumTimerInterval() method.
3328
3329         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
3330         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
3331         (WTR::InjectedBundle::beginTesting):
3332         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
3333         (WTR::TestRunner::setMinimumTimerInterval):
3334         (WTR):
3335         * WebKitTestRunner/InjectedBundle/TestRunner.h:
3336         (TestRunner):
3337
3338 2012-09-11  Tommy Widenflycht  <tommyw@google.com>
3339
3340         MediaStream API: add RTCPeerConnection::onnegotiationneeded
3341         https://bugs.webkit.org/show_bug.cgi?id=96097
3342
3343         Reviewed by Adam Barth.
3344
3345         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
3346         (MockWebRTCPeerConnectionHandler::addStream):
3347         (MockWebRTCPeerConnectionHandler::removeStream):
3348
3349 2012-09-11  Peter Beverloo  <peter@chromium.org>
3350
3351         Recognize the "--chromium-android" argument in run-webkit-tests
3352         https://bugs.webkit.org/show_bug.cgi?id=96369
3353
3354         Reviewed by Jochen Eisinger.
3355
3356         Chromium for the Android platform cannot determine the platform based on
3357         the host itself, so we'll use the --chromium-android argument as the build
3358         master passes to run-webkit-tests. Make sure the layout test runner
3359         can pick the right port when doing so.
3360
3361         * Scripts/run-webkit-tests:
3362         * Scripts/webkitpy/layout_tests/port/factory.py:
3363         (port_options):
3364
3365 2012-09-11  Simon Hausmann  <simon.hausmann@nokia.com>
3366
3367         [QT] Fix build with newer qmake from Qt 5
3368
3369         Rubber-stamped by Tor Arne Vestbø.
3370
3371         Support for "qmake -query QMAKE_MKSPECS" was removed from qmake. Replace
3372         it with the use of QT_HOST_DATA/mkspecs.
3373         Based on patch by Oswald Buddenhagen <oswald.buddenhagen@nokia.com>.
3374
3375         * Scripts/webkitdirs.pm:
3376         (builtDylibPathForName):
3377         (qtMakeCommand):
3378
3379 2012-09-11  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3380
3381         [Qt] Remove Qt 4 syncqt, no longer needed
3382
3383         Reviewed by Simon Hausmann.
3384
3385         * qmake/syncqt-4.8: Removed.
3386
3387 2012-09-11  Andras Becsi  <andras.becsi@nokia.com>
3388
3389         [Qt][WK2] Fix the build with recent Qt5
3390         https://bugs.webkit.org/show_bug.cgi?id=96283
3391
3392         Reviewed by Simon Hausmann.
3393
3394         We should not use deprecated Qt API because doing that results
3395         in build failure since all the deprecated API was disabled
3396         in qtbase (https://codereview.qt-project.org/#change,24890).
3397
3398         * DumpRenderTree/qt/main.cpp:
3399         (main):
3400         * MiniBrowser/qt/MiniBrowserApplication.cpp:
3401         (MiniBrowserApplication::notify):
3402         * MiniBrowser/qt/raw/View.cpp:
3403         (View::exposeEvent):
3404         (main):
3405
3406 2012-09-11  Christophe Dumez  <christophe.dumez@intel.com>
3407
3408         WTR needs an implementation of applicationCacheDiskUsageForOrigin
3409         https://bugs.webkit.org/show_bug.cgi?id=87188
3410
3411         Reviewed by Kenneth Rohde Christiansen.
3412
3413         Add implementation for applicationCacheDiskUsageForOrigin()
3414         in WebKitTestRunner.
3415
3416         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
3417         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
3418         (WTR::TestRunner::applicationCacheDiskUsageForOrigin):
3419         (WTR):
3420         * WebKitTestRunner/InjectedBundle/TestRunner.h:
3421         (TestRunner):
3422
3423 2012-09-10  Joanmarie Diggs  <jdiggs@igalia.com>
3424
3425         Properly expose <legend> elements to ATs
3426         https://bugs.webkit.org/show_bug.cgi?id=84137
3427
3428         Reviewed by Chris Fleizach.
3429
3430         Created a new WebCore Accessibility Role, LegendRole. Used it to map to
3431         the expected platform role, ATK_ROLE_LABEL. Also established the needed
3432         AtkRelation pair, label-for/labelled-by between the legend and fieldset.
3433
3434         * DumpRenderTree/chromium/TestRunner/AccessibilityUIElementChromium.cpp:
3435         (roleToString): added "Legend" string for WebAccessibilityRoleLegend
3436
3437 2012-09-10  Jer Noble  <jer.noble@apple.com>
3438
3439         Unreviewed; rolling out r128081.
3440         http://trac.webkit.org/changeset/128081
3441         https://bugs.webkit.org/show_bug.cgi?id=96134
3442
3443         New test http/tests/media/video-donottrack.html fails to run on Apple test bots.
3444
3445         * DumpRenderTree/TestRunner.cpp:
3446         (TestRunner::TestRunner):
3447         (TestRunner::staticFunctions):
3448         * DumpRenderTree/TestRunner.h:
3449         (TestRunner):
3450         * DumpRenderTree/mac/FrameLoadDelegate.mm:
3451         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
3452         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
3453         (WTR::InjectedBundlePage::InjectedBundlePage):
3454         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
3455         (InjectedBundlePage):
3456         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
3457         (WTR::TestRunner::TestRunner):
3458         * WebKitTestRunner/InjectedBundle/TestRunner.h:
3459         (TestRunner):
3460
3461 2012-09-10  Christophe Dumez  <christophe.dumez@intel.com>
3462
3463         TestExpectationsChecker._determine_port_from_expectations_path() does not support cascaded TestExpectations
3464         https://bugs.webkit.org/show_bug.cgi?id=96205
3465
3466         Reviewed by Tony Chang.
3467
3468         _determine_port_from_expectations_path() was calling port.path_to_test_expectations_file()
3469         internally, which means that it would support only 1 TestExpectations file per port. This
3470         is an issue for ports such as EFL that support cascased TestExpectations (efl-wk2 -> efl
3471         and efl-wk1 -> efl).
3472
3473         This patch makes _determine_port_from_expectations_path() call port.expectations_files()
3474         instead so that all the ports TestExpectations are recognized. The ports are also
3475         constructed twice, with "webkit_test_runner" option set to True and False so that we
3476         retrieve the TestExpectations paths for both WebKit1 and WebKit2.
3477
3478         * Scripts/webkitpy/style/checkers/test_expectations.py:
3479         (TestExpectationsChecker._determine_port_from_expectations_path):
3480         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py: Add corresponding unit test.
3481         (TestExpectationsTestCase.test_determine_port_from_expectations_path):
3482
3483 2012-09-10  Ryuan Choi  <ryuan.choi@samsung.com>
3484
3485         [WTR] Generated source files should include config.h
3486         https://bugs.webkit.org/show_bug.cgi?id=96254
3487
3488         Reviewed by Benjamin Poulain.
3489
3490         Current coding style has encouraged to include config.h at the first position,
3491         But derived sources, generated by CodeGeneratorTestRunner.pm, don't follow
3492         this rule.
3493         Therefore, this patch added it.
3494
3495         * WebKitTestRunner/InjectedBundle/Bindings/CodeGeneratorTestRunner.pm:
3496         (_generateImplementationFile):
3497
3498 2012-09-10  Rick Byers  <rbyers@chromium.org>
3499
3500         Add handling of new GestureTapCancel in DRT
3501         
3502         https://bugs.webkit.org/show_bug.cgi?id=96183
3503
3504         Reviewed by Adam Barth.
3505
3506         * DumpRenderTree/chromium/TestWebPlugin.cpp:
3507         (TestWebPlugin::handleInputEvent):
3508
3509 2012-09-10  Brady Eidson  <beidson@apple.com>
3510
3511         Javascript in foreground tabs should not wait synchronously for plug-ins to load
3512         <rdar://problem/12067415> and https://bugs.webkit.org/show_bug.cgi?id=96167
3513
3514         Reviewed by Geoff Garen.
3515
3516         Enhance the "Slow NPP_New" plug-in to also be able to return properties to javascript.
3517
3518         * DumpRenderTree/TestNetscapePlugIn/Tests/SlowNPPNew.cpp:
3519         (PluginObject):
3520         (SlowNPPNew::PluginObject::PluginObject):
3521         (SlowNPPNew::PluginObject::~PluginObject):
3522         (SlowNPPNew::PluginObject::hasProperty):
3523         (SlowNPPNew::PluginObject::getProperty):
3524         (SlowNPPNew::NPP_GetValue):
3525         (SlowNPPNew):
3526
3527 2012-09-10  Tim Horton  <timothy_horton@apple.com>
3528
3529         WKTR often reports an unresponsive WebProcess on Mac bots
3530         https://bugs.webkit.org/show_bug.cgi?id=95906
3531
3532         Reviewed by Dan Bernstein.
3533
3534         Attempt to work around resource contention issues by increasing the NRWT
3535         timeout when running against WKTR, as well as WKTR's own WebProcess timeout.
3536
3537         * Scripts/webkitpy/layout_tests/port/base.py:
3538         (Port.default_timeout_ms): Default timeout when using WKTR -> 80 seconds.
3539         * WebKitTestRunner/TestController.cpp:
3540         (WTR): LongTimeout (used when waiting for WebProcess to finish) -> 60 seconds.
3541
3542 2012-09-07  Jer Noble  <jer.noble@apple.com>
3543
3544         <audio> and <video> should send Do Not Track when appropriate
3545         https://bugs.webkit.org/show_bug.cgi?id=96134
3546
3547         Reviewed by Eric Carlson.
3548
3549         Add a testRunner JS API for calling setShouldSendDoNotTrackHTTPHeader to DRT and 
3550         WTR. Add FrameLoaderClient client methods for retrieving that value.
3551
3552         * DumpRenderTree/TestRunner.cpp:
3553         (setShouldSendDoNotTrackHTTPHeaderCallback): Pass through to the TestRunner.
3554         (TestRunner::staticFunctions): Add the setShouldSendDoNotTrackHTTPHeader function.
3555         * DumpRenderTree/TestRunner.h:
3556         (TestRunner::setShouldSendDoNotTrackHTTPHeader): Simple setter.
3557         (TestRunner::shouldSendDoNotTrackHTTPHeader): Simple getter.
3558         * DumpRenderTree/mac/FrameLoadDelegate.mm:
3559         (-[FrameLoadDelegate webViewShouldSendDoNotTrackHTTPHeader:]): 
3560             Pass the question on to the TestRunner.
3561         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
3562         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
3563         (WTR::InjectedBundlePage::InjectedBundlePage):
3564             Add the setShouldSendDoNotTrackHTTPHeader client function.
3565         (WTR::InjectedBundlePage::shouldSendDoNotTrackHTTPHeader):
3566             Pass the question on to the TestRunner.
3567         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
3568         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
3569         * WebKitTestRunner/InjectedBundle/TestRunner.h:
3570         (WTR::TestRunner::shouldSendDoNotTrackHTTPHeader): Simple getter.
3571         (WTR::TestRunner::setShouldSendDoNotTrackHTTPHeader): Simple Setter.
3572
3573 2012-09-10  Ojan Vafai  <ojan@chromium.org>
3574
3575         Make EFL bots work on the flakiness dashboard
3576         https://bugs.webkit.org/show_bug.cgi?id=96164
3577
3578         Reviewed by Tony Chang.
3579
3580         Also, update outdated unittests that had started failing.
3581
3582         * TestResultServer/static-dashboards/builders.js:
3583         (isWebkitTestRunner):
3584         * TestResultServer/static-dashboards/flakiness_dashboard.js:
3585         (nonChromiumPlatform):
3586         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
3587         (test):
3588
3589 2012-09-10  Tommy Widenflycht  <tommyw@google.com>
3590
3591         [chromium] MediaStream API: Remove the Descriptor postfix
3592         https://bugs.webkit.org/show_bug.cgi?id=96268
3593
3594         Reviewed by Adam Barth.
3595
3596         After consideration I realized that there is no need whatsoever in using the Descriptor prefix
3597         in the WekKit embedder API. Removed from WebRTCSessionDescription and WebRTCICECandidate to start with.
3598
3599         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
3600         (RTCSessionDescriptionRequestSuccededTask::RTCSessionDescriptionRequestSuccededTask):
3601         (MockWebRTCPeerConnectionHandler::createOffer):
3602         (MockWebRTCPeerConnectionHandler::createAnswer):
3603         (MockWebRTCPeerConnectionHandler::setLocalDescription):
3604         (MockWebRTCPeerConnectionHandler::setRemoteDescription):
3605         (MockWebRTCPeerConnectionHandler::localDescription):
3606         (MockWebRTCPeerConnectionHandler::remoteDescription):
3607         (MockWebRTCPeerConnectionHandler::addICECandidate):
3608         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h:
3609         (MockWebRTCPeerConnectionHandler):
3610
3611 2012-09-10  Arnaud Renevier  <a.renevier@sisa.samsung.com>
3612
3613         [Gtk] failure to build jhbuild dependencies with cups 1.6
3614         https://bugs.webkit.org/show_bug.cgi?id=95991
3615
3616         Reviewed by Gustavo Noronha Silva.
3617
3618         Bump gtk+ and gnome-themes-standard dependencies to 3.4.2 to allow
3619         building with cups >= 1.6 [GNOME #670373].
3620
3621         * gtk/jhbuild.modules:
3622
3623 2012-09-10  Christophe Dumez  <christophe.dumez@intel.com>
3624
3625         [EFL][DRT] Encoding issue in printed console messages
3626         https://bugs.webkit.org/show_bug.cgi?id=96263
3627
3628         Reviewed by Kenneth Rohde Christiansen.
3629
3630         Correctly handle CONSOLE messages that contain non-ascii
3631         characters in EFL's DRT. We need to use String::fromUTF8()
3632         to construct a WTF::String from a raw string.
3633
3634         * DumpRenderTree/efl/DumpRenderTreeView.cpp:
3635         (onConsoleMessage):
3636
3637 2012-09-10  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
3638
3639         [WK2][WTR] User style sheet location should be reset before each test
3640         https://bugs.webkit.org/show_bug.cgi?id=96262
3641
3642         Reviewed by Kenneth Rohde Christiansen.
3643
3644         User style sheet location is reset before each test now.
3645
3646         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
3647         (WTR::InjectedBundle::beginTesting):
3648
3649 2012-09-10  Simon Hausmann  <simon.hausmann@nokia.com>
3650
3651         Fix build of WTR with Qt on Windows
3652         https://bugs.webkit.org/show_bug.cgi?id=96249
3653
3654         Reviewed by Kenneth Rohde Christiansen.
3655
3656         Windows loves to define min and max as macros, which beautifully
3657         clashes in many places where min and max are used as variable or method
3658         names. It can be convinced not to do this when min and max are already
3659         define as macros (expanding to themselves). We have this kind of
3660         workaround in config.h files all over WebKit