c260c368afbd0e6892577dd6d4c035bd5170f6e8
[WebKit-https.git] / Tools / ChangeLog
1 2012-09-21  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
2
3         WebKitTestRunner needs to print frame load delegate information
4         https://bugs.webkit.org/show_bug.cgi?id=42705
5
6         Reviewed by Kenneth Rohde Christiansen.
7
8         Added missing dumping from WTR::InjectedBundlePage::didFailLoadWithErrorForFrame.
9
10         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
11         (WTR::InjectedBundlePage::didFailLoadWithErrorForFrame):
12
13 2012-09-21  Christophe Dumez  <christophe.dumez@intel.com>
14
15         [EFL] EventSender should mimic CTRL+o emacs shortcut
16         https://bugs.webkit.org/show_bug.cgi?id=97224
17
18         Reviewed by Kenneth Rohde Christiansen.
19
20         Add support for mimicking CTRL+o emacs shortcut in EFL's
21         EventSender in DumpRenderTree and WebKitTestRunner.
22
23         * DumpRenderTree/efl/EventSender.cpp:
24         (sendKeyDown):
25         * WebKitTestRunner/efl/EventSenderProxyEfl.cpp:
26         (WTR::EventSenderProxy::keyDown):
27
28 2012-09-21  Simon Hausmann  <simon.hausmann@digia.com>
29
30         [Qt] Re-fix clean builds
31
32         Reviewed by Tor Arne Vestbø.
33
34         Re-introduce the sanitization for LIBS when creating a module to use
35         LIBS_PRIVATE and otherwise do _not_ use LIBS_PRIVATE. We decided to
36         continue to use QT, LIBS and PKGCONFIG instead of their _PRIVATE variants
37         throughout the code base, so just using LIBS_PRIVATE in linkAgainstLibrary()
38         causes build issues when depending system libraries end up in LIBS before
39         LIBS_PRIVATE.
40
41         * qmake/mkspecs/features/default_post.prf:
42         * qmake/mkspecs/features/functions.prf:
43
44 2012-09-20  Simon Fraser  <simon.fraser@apple.com>
45
46         Comment out a failing webkitpy unit test until Dirk can fix it.
47
48         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
49         (SemanticTests.test_skip_and_wontfix):
50
51 2012-09-20  Dirk Pranke  <dpranke@chromium.org>
52
53         make Skip, WontFix be the only expectations on a line
54         https://bugs.webkit.org/show_bug.cgi?id=97225
55
56         Reviewed by Ojan Vafai.
57
58         It is now incorrect in the new syntax to have a line like:
59
60             foo.html [ WontFix Crash ]
61
62         This will generate a lint warning and be treated as an invalid
63         line. Fixing this caused a whole bunch of unit tests to need updating
64         to no longer be marked as WontFix :). Also, this patch adjusts
65         the warnings so that missing Bug() identifiers will cause lint
66         warnings but will *not* cause the line to be treated as invalid.
67         Fixing these issues also revealed that test_hung_thread was no
68         longer testing the right logic, so I adjusted the timeouts in
69         test.py to make that test work again.
70
71         * Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
72         (Worker._run_test_in_another_thread):
73         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
74         (ResultSummaryTest.test_summarized_results_wontfix):
75         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
76         (TestExpectationParser._parse_modifiers):
77         (TestExpectationParser._tokenize_line_using_new_format):
78         (TestExpectationLine.is_invalid):
79         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
80         (BasicTests.test_basic):
81         (test_get_test_set):
82         (test_parse_warning):
83         (test_pixel_tests_flag):
84         (SemanticTests.test_missing_bugid):
85         (SemanticTests):
86         (SemanticTests.test_skip_and_wontfix):
87         * Scripts/webkitpy/layout_tests/port/test.py:
88         (TestDriver.run_test):
89         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
90         (MainTest.test_hung_thread):
91         * Scripts/webkitpy/tool/commands/queries_unittest.py:
92         (PrintExpectationsTest.test_basic):
93         (PrintExpectationsTest.test_multiple):
94         (PrintExpectationsTest.test_full):
95         (PrintExpectationsTest.test_exclude):
96         (PrintExpectationsTest.test_csv):
97
98 2012-09-20  Lucas Forschler  <lforschler@apple.com>
99
100         Unreviewed.  Start running tests on the mac-ews.
101         
102         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
103         (MacEWS):
104
105 2012-09-20  Tony Chang  <tony@chromium.org>
106
107         [chromium] Enable more clang warnings
108         https://bugs.webkit.org/show_bug.cgi?id=97257
109
110         Reviewed by James Robinson.
111
112         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp: Add -Wunused-parameter. Don't add -Wglobal-constructors because
113         there are lots of violations in these files.
114
115 2012-09-19  Ryosuke Niwa  <rniwa@webkit.org>
116
117         run-perf-tests should record individual value instead of statistics
118         https://bugs.webkit.org/show_bug.cgi?id=97155
119
120         Reviewed by Hajime Morita.
121
122         Parse the list of individual value reported by tests and include them as "values".
123         We strip "values" from the output JSON when uploading it to the perf-o-matic
124         since it doesn't know how to parse "values" or ignore it.
125
126         * Scripts/webkitpy/performance_tests/perftest.py:
127         (PerfTest):
128         (PerfTest.parse_output): Parse and report "values".
129         (PageLoadingPerfTest.run): Report indivisual page loading time in "values".
130         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
131         (MainTest.test_parse_output):
132         (MainTest.test_parse_output_with_failing_line):
133         (TestPageLoadingPerfTest.test_run):
134         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
135         (PerfTestsRunner._generate_and_show_results): Strip "values" from each result
136         until we update perf-o-matic.
137         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
138         (test_run_memory_test):
139         (test_run_with_json_output):
140         (test_run_with_description):
141         (test_run_with_slave_config_json):
142         (test_run_with_multiple_repositories):
143
144 2012-09-20  Simon Hausmann  <simon.hausmann@digia.com>
145
146         [Qt] Fix production builds
147
148         Unreviewed trivial fix: Follow up r129119 and avoid clobbering LIBS through the use of LIBS_PRIVATE.
149
150         Otherwise the libQtWebKit.prl file for example contains -lWebKit1 -lWebCore, etc.
151
152         * qmake/mkspecs/features/functions.prf:
153
154 2012-09-20  Dirk Pranke  <dpranke@chromium.org>
155
156         REGRESSION: layout test results doesn't show diffs
157         https://bugs.webkit.org/show_bug.cgi?id=97182
158
159         Reviewed by Ojan Vafai.
160
161         Go back to storing TEXT, AUDIO, and IMAGE+TEXT in results.json
162         so that results.html (and hopefully garden-o-matic) can
163         determine which things actually failed. However, we keep mapping
164         these results to Failure so that we still only have a single
165         expectation type for them.
166
167         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
168         (JSONLayoutResultsGenerator):
169         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
170         (TestExpectationParser):
171         (TestExpectations):
172         (TestExpectations.result_was_expected):
173         * Scripts/webkitpy/layout_tests/models/test_failures.py:
174         (determine_result_type):
175         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
176         (MainTest.test_missing_and_unexpected_results):
177         (MainTest.test_retrying_and_flaky_tests):
178
179 2012-09-20  Tommy Widenflycht  <tommyw@google.com>
180
181         MediaStream API: Extend UserMediaRequest with a ownerDocument method
182         https://bugs.webkit.org/show_bug.cgi?id=97095
183
184         Reviewed by Adam Barth.
185
186         Extending WebUserMediaClientMock to check that the owning document is valid,
187         and that the document has a frame.
188
189         * DumpRenderTree/chromium/WebUserMediaClientMock.cpp:
190         (WebKit::WebUserMediaClientMock::requestUserMedia):
191
192 2012-09-20  Stephen Chenney  <schenney@chromium.org>
193
194         [Chromium] DRT does not support --dump-all-pixels flag
195         https://bugs.webkit.org/show_bug.cgi?id=95098
196
197         Reviewed by Dirk Pranke.
198
199         Add support for the --pixel-tests and shorthand -p option in Chromium DumpRenderTree. Use
200         of this flag causes pixel results to be created for all tests, regardless of
201         individual test options. If an individual test provides a pixel hash it will be used,
202         otherwise the hash will be empty. This replaces a previously defined but unused option
203         --dump-all-pixels, and is useful primarily when debugging DRT instances.
204
205         * DumpRenderTree/chromium/DumpRenderTree.cpp:
206         (runTest): Add a parameter and code to force pixel results for the test.
207         (main): Add parameter handling for --pixels-test and -p, and remove --dump-all-pixels.
208
209 2012-09-20  Carlos Garcia Campos  <cgarcia@igalia.com>
210
211         [GTK] run-api-tests should not buffer test stdout
212         https://bugs.webkit.org/show_bug.cgi?id=88474
213
214         Reviewed by Philippe Normand.
215
216         * Scripts/run-gtk-tests:
217         (TestRunner._run_test_command): Use os.forkpty() instead of
218         subprocess.Popen() so that gtest sends the output with colors to
219         stdout. Use common.parse_output_lines() to parse the output and
220         write it to stdout while it's read.
221         (TestRunner._run_test_command.parse_line): Parse the line to get
222         the test pid and write the line to stdout.
223         (TestRunner._run_test_command.waitpid): Helper function to call
224         waitpid handling EINTR.
225         (TestRunner._run_test_command.return_code_from_exit_status):
226         Helper function to convert exit status of test commands to a
227         return code.
228         * gtk/common.py:
229         (parse_output_lines): Helper function that uses select to read
230         the given file descriptor and call the given callback for every
231         line read.
232
233 2012-09-20  Simon Hausmann  <simon.hausmann@digia.com>
234
235         [Qt] Fix initial build
236
237         Reviewed by Tor Arne Vestbø.
238
239         When building QtWebKit the first time there is no qt_webkit.pri module pri file, and therefore
240         $$QT.webkit.name isn't set and so creating_module isn't set. That has all sorts of implications
241         causing incorrect linking for Makefile.api, etc.
242
243         Fix the determination by simply checking if MODULE is set, which only happens in api.pri.
244
245         * qmake/mkspecs/features/webkit_modules.prf:
246
247 2012-09-20  Sheriff Bot  <webkit.review.bot@gmail.com>
248
249         Unreviewed, rolling out r129091.
250         http://trac.webkit.org/changeset/129091
251         https://bugs.webkit.org/show_bug.cgi?id=97205
252
253         It broke perf tests everywhere (Requested by Ossy on #webkit).
254
255         * Scripts/webkitpy/performance_tests/perftest.py:
256         (PerfTest):
257         (PerfTest.parse_output):
258         (PageLoadingPerfTest.run):
259         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
260         (MainTest.test_parse_output):
261         (MainTest.test_parse_output_with_failing_line):
262         (TestPageLoadingPerfTest.test_run):
263         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
264         (PerfTestsRunner._generate_and_show_results):
265         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
266         (test_run_memory_test):
267         (test_run_with_json_output):
268         (test_run_with_description):
269         (test_run_with_slave_config_json):
270         (test_run_with_multiple_repositories):
271
272 2012-09-20  Simon Hausmann  <simon.hausmann@digia.com>
273
274         [Qt] QtWebKit module header includes private dependencies
275
276         Reviewed by Tor Arne Vestbø.
277
278         * qmake/mkspecs/features/default_post.prf: Don't try to sanitize LIBS, because we can make sure
279         that LIBS_PRIVATE is set from the beginning. Moved the creating_module and PKGCONFIG/QT(_PRIVATE)
280         sanitization into webkit_modules.prf. creating_module determination requires TARGET to be set, so
281         we can't do it in default_pre.
282         * qmake/mkspecs/features/webkit_modules.prf:
283
284 2012-09-20  Simon Hausmann  <simon.hausmann@digia.com>
285
286         [Qt] Fix build with MingW
287
288         Reviewed by Tor Arne Vestbø.
289
290         Don't try to compile WebCore, etc. with debug symbols for production
291         builds, it's just too big.
292
293         * qmake/mkspecs/features/production_build.prf:
294
295 2012-09-20  Michał Pakuła vel Rutka  <m.pakula@samsung.com>
296
297         [EFL][DRT]DumpRenderTree needs to reset focus state when test starts.
298         https://bugs.webkit.org/show_bug.cgi?id=97087
299
300         Reviewed by Gyuyoung Kim.
301
302         Add focusing a main frame on settings reset.
303         After editing/undo/undo-iframe-location-change was executed a frame
304         was left in unfocused state. This caused flakiness in two tests results.
305
306         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
307         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
308
309 2012-09-20  Allan Sandfeld Jensen  <allan.jensen@digia.com>
310
311         [Qt] Add eventSender.gestureTap
312         https://bugs.webkit.org/show_bug.cgi?id=66173
313
314         Reviewed by Kenneth Rohde Christiansen.
315
316         Implement support for sending raw QGesture events.
317
318         * DumpRenderTree/qt/EventSenderQt.cpp:
319         (EventSender::EventSender):
320         (EventSender::gestureTap):
321         * DumpRenderTree/qt/EventSenderQt.h:
322         (EventSender):
323
324 2012-09-19  Sheriff Bot  <webkit.review.bot@gmail.com>
325
326         Unreviewed, rolling out r129007.
327         http://trac.webkit.org/changeset/129007
328         https://bugs.webkit.org/show_bug.cgi?id=97172
329
330         It broke the build on the Qt bots (Requested by Ossy on
331         #webkit).
332
333         * qmake/config.tests/gccdepends/empty.cpp:
334         * qmake/config.tests/gccdepends/gccdepends.pro:
335         * qmake/mkspecs/features/default_pre.prf:
336
337 2012-09-19  Ryosuke Niwa  <rniwa@webkit.org>
338
339         run-perf-tests should record indivisual value instead of statistics
340         https://bugs.webkit.org/show_bug.cgi?id=97155
341
342         Reviewed by Hajime Morita.
343
344         Parse the list of indivisual value reported by tests and include them as "values".
345         We strip "values" from the output JSON when uploading it to the perf-o-matic
346         since it doesn't know how to parse "values" or ignore it.
347
348         * Scripts/webkitpy/performance_tests/perftest.py:
349         (PerfTest):
350         (PerfTest.parse_output): Parse and report "values".
351         (PageLoadingPerfTest.run): Report indivisual page loading time in "values".
352         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
353         (MainTest.test_parse_output):
354         (MainTest.test_parse_output_with_failing_line):
355         (TestPageLoadingPerfTest.test_run):
356         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
357         (PerfTestsRunner._generate_and_show_results): Strip "values" from each result
358         until we update perf-o-matic.
359         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
360         (test_run_memory_test):
361         (test_run_with_json_output):
362         (test_run_with_description):
363         (test_run_with_slave_config_json):
364         (test_run_with_multiple_repositories):
365
366 2012-09-19  Dirk Pranke  <dpranke@chromium.org>
367
368         Fix regex groups for bug matching in flakiness dashboard.
369         https://bugs.webkit.org/show_bug.cgi?id=97152
370
371         Unreviewed, build fix.
372
373         * TestResultServer/static-dashboards/flakiness_dashboard.js:
374         (htmlForBugs):
375
376 2012-09-19  Dirk Pranke  <dpranke@chromium.org>
377
378         nrwt: print unexpected results using new TestExpectations syntax
379         https://bugs.webkit.org/show_bug.cgi?id=97159
380
381         Unreviewed, build fix.
382
383         Change new-run-webkit-tests to print out failures using the new
384         syntax when there are unexpected results, e.g.:
385
386         52 tests ran as expected, 19 didn't:
387
388
389         Regressions: Unexpected failures : (2)
390           failures/flaky/text.html [ Failure ]
391           failures/unexpected/text-image-checksum.html [ Failure ]
392
393         and so forth
394
395         * Scripts/webkitpy/layout_tests/views/printing.py:
396         (Printer._print_unexpected_results):
397
398 2012-09-19  Dirk Pranke  <dpranke@chromium.org>
399
400         update flakiness dashboard after cutover to new test expectations syntax
401         https://bugs.webkit.org/show_bug.cgi?id=97152
402
403         Unreviewed, build fix.
404
405         Handle (??) the new Bug notations as well. Hopefully we don't
406         still need the old ones.
407
408         * TestResultServer/static-dashboards/flakiness_dashboard.js:
409         (filterBugs):
410         (htmlForBugs):
411         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
412
413 2012-09-19  Dirk Pranke  <dpranke@chromium.org>
414
415         update flakiness dashboard after cutover to new test expectations syntax
416         https://bugs.webkit.org/show_bug.cgi?id=97152
417
418         Reviewed by Ryosuke Niwa.
419
420         This change clones the TestExpectation parsing state machine
421         from python into javascript.
422
423         * TestResultServer/static-dashboards/flakiness_dashboard.js:
424         (parsedExpectations.lines.forEach):
425         (parsedExpectations):
426         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
427
428 2012-09-19  Dana Jansens  <danakj@chromium.org>
429
430         Add backer@chromium.org as contributor
431         https://bugs.webkit.org/show_bug.cgi?id=97150
432
433         * Scripts/webkitpy/common/config/committers.py:
434
435 2012-09-19  Dirk Pranke  <dpranke@chromium.org>
436
437         fix MISSING after TestExpectations conversion
438         https://bugs.webkit.org/show_bug.cgi?id=97148
439
440         Unreviewed, expectations change / build fix.
441
442         * Scripts/convert-test-expectations:
443         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
444         (TestExpectationParser):
445
446 2012-09-19  Dirk Pranke  <dpranke@chromium.org>
447
448         nrwt: convert chromium TestExpectations to the new syntax
449         https://bugs.webkit.org/show_bug.cgi?id=97139
450
451         Reviewed by Ryosuke Niwa.
452     
453         Add in a file temporarily that does the conversion of
454         test expectations formats
455
456         * Tools/Scripts/convert-test-expectations: Added.
457
458 2012-09-19  Ryosuke Niwa  <rniwa@webkit.org>
459
460         REGRESSION: run-perf-tests no longer reports the total test time
461         https://bugs.webkit.org/show_bug.cgi?id=97138
462
463         Reviewed by Tony Chang.
464
465         Report the finished time as a info-level log as opposed to a debug level log.
466
467         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
468         (PerfTestsRunner._run_single_test):
469         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
470         (MainTest.normalizeFinishedTime): Added. It replaces all finished times by 0.1 seconds.
471         (test_run_test_pause_before_testing):
472         (test_run_test_set_for_parser_tests):
473         (test_run_memory_test):
474         (_test_run_with_json_output):
475
476 2012-09-19  Dirk Pranke  <dpranke@chromium.org>
477
478         Support new TestExpectations format alongside old one
479         https://bugs.webkit.org/show_bug.cgi?id=96588
480
481         Reviewed by Ojan Vafai.
482
483         This patch adds support for actually reading in lines formatted
484         in the new style, and when we re-serialize/write out the file,
485         all lines will be written in the new style. Note that reading in
486         the old style is still supported, and no updates are being made
487         to the actual TestExpectations files as part of this change.
488
489         This change updates most but not all of the unit tests to use
490         the new syntax. I will update the rest when (or before where
491         possible) I drop support for the old format.
492
493         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
494         (TestExpectationParser):
495         (TestExpectationParser._tokenize_line):
496         (TestExpectationParser._tokenize_line_using_new_format):
497         (TestExpectationLine.to_string):
498         (TestExpectationLine._format_line):
499         (TestExpectations):
500         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
501         (test_parse_warning):
502         (SkippedTests.test_skipped_entry_dont_exist):
503         (NewExpectationSyntaxTests.test_warnings):
504         (RemoveConfigurationsTest.test_remove):
505         (test_remove_line):
506         (RebaseliningTest.test_remove):
507         (RebaseliningTest.test_no_get_rebaselining_failures):
508         (TestExpectationParserTests.test_tokenize_blank):
509         (TestExpectationParserTests.test_tokenize_extra_colon):
510         (TestExpectationParserTests.test_tokenize_missing_equal):
511         (TestExpectationParserTests.test_tokenize_extra_equal):
512         (TestExpectationSerializationTests.test_unparsed_to_string):
513         (TestExpectationSerializationTests.test_unparsed_list_to_string):
514         (TestExpectationSerializationTests.test_parsed_to_string):
515         (TestExpectationSerializationTests.test_format_line):
516         (TestExpectationSerializationTests.test_string_roundtrip):
517         (TestExpectationSerializationTests.test_list_roundtrip):
518         (TestExpectationSerializationTests.test_reconstitute_only_these):
519         (TestExpectationSerializationTests.test_string_whitespace_stripping):
520         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
521         (LintTest.test_lint_test_files__errors):
522         * Scripts/webkitpy/tool/commands/queries_unittest.py:
523         (PrintExpectationsTest.test_basic):
524         (PrintExpectationsTest.test_multiple):
525         (PrintExpectationsTest.test_full):
526         (PrintExpectationsTest.test_exclude):
527         * Scripts/webkitpy/tool/commands/rebaseline.py:
528         (RebaselineExpectations.execute):
529         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
530         (TestRebaseline.test_rebaseline_updates_expectations_file_noop):
531         (test_rebaseline_updates_expectations_file):
532         (test_rebaseline_does_not_include_overrides):
533         (test_rebaseline_expectations_noop):
534         (test_overrides_are_included_correctly):
535
536 2012-09-19  Dirk Pranke  <dpranke@chromium.org>
537
538         implement first part of support for the new TestExpectations syntax
539         https://bugs.webkit.org/show_bug.cgi?id=96569
540
541         Reviewed by Ryosuke Niwa.
542
543         This patch implements support for parsing a line of the new
544         format for the TestExpectations file and converting it back into
545         the old format for compatibility. This routine is not yet used
546         by anything.
547
548         The new format is documented at:
549           http://trac.webkit.org/wiki/TestExpectations
550
551         but, in short:
552
553           [bugs] [ "[" modifiers "]" ] test_name [ "[" expectations "]" ]
554
555         - Comments are indicated with "#" instead of "//"
556         - If no expectations are specified we default to Skip for
557           compatibility with the Skipped files (these two changes make
558           Skipped files a subset of TestExpectations files)
559
560         - All of the tokens are now CamelCase instead of ALLCAPS.
561         - FAIL -> Failure
562         - IMAGE -> ImageOnlyFailure
563         - WONTFIX -> WontFix
564         - modifiers refer to just the platforms and configurations
565           (release/debug) that the line applies to.
566         - WontFix, Rebaseline, Slow, and Skip move to the right-hand side as
567           expectations
568         - expectations will typically be written out in lexicographic order
569         - We use webkit.org/b/12345, crbug.com/12345, and Bug(dpranke)
570           instead of BUGWK12345, BUGCR12345, and BUGDPRANKE.
571
572         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
573         (TestExpectationParser):
574         (TestExpectationParser._tokenize_line_using_new_format):
575         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
576         (NewExpectationSyntaxTests):
577         (NewExpectationSyntaxTests.assert_exp):
578         (NewExpectationSyntaxTests.test_bare_name):
579         (NewExpectationSyntaxTests.test_bare_name_and_bugs):
580         (NewExpectationSyntaxTests.test_comments):
581         (NewExpectationSyntaxTests.test_config_modifiers):
582         (NewExpectationSyntaxTests.test_unknown_config):
583         (NewExpectationSyntaxTests.test_unknown_expectation):
584         (NewExpectationSyntaxTests.test_skip):
585         (NewExpectationSyntaxTests.test_slow):
586         (NewExpectationSyntaxTests.test_wontfix):
587         (NewExpectationSyntaxTests.test_blank_line):
588         (NewExpectationSyntaxTests.test_warnings):
589
590 2012-09-19  Dirk Pranke  <dpranke@chromium.org>
591
592         nrwt: replace TEXT, AUDIO, and IMAGE+TEXT with FAIL
593         https://bugs.webkit.org/show_bug.cgi?id=96845
594
595         Reviewed by Ojan Vafai.
596
597         In preparation for the new TestExpectations syntax, we replace
598         all TEXT, IMAGE+TEXT, and AUDIO failures with FAIL. This will
599         make switching to the new syntax lossless (i.e., we lose
600         information now, not then).
601
602         Note that we can still parse in results.json files that have the
603         old data for backwards compatibility.
604         
605
606         * Scripts/webkitpy/common/net/resultsjsonparser.py:
607         (JSONTestResult._failure_types_from_actual_result):
608         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
609         (ResultsJSONParserTest):
610         (test_basic):
611         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
612         (ResultSummaryTest.test_summarized_results_wontfix):
613         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
614         (JSONLayoutResultsGenerator):
615         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
616         (TestExpectations):
617         (TestExpectations.remove_pixel_failures):
618         (TestExpectations.has_pixel_failures):
619         (TestExpectations.suffixes_for_expectations):
620         (TestExpectations.get_rebaselining_failures):
621         (TestExpectations.remove_configuration_from_test):
622         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
623         (assert_bad_expectations):
624         (BasicTests):
625         (BasicTests.test_basic):
626         (MiscTests):
627         (MiscTests.test_multiple_results):
628         (MiscTests.test_result_was_expected):
629         (MiscTests.test_remove_pixel_failures):
630         (MiscTests.test_suffixes_for_expectations):
631         (test_get_expectations_string):
632         (test_parse_warning):
633         (test_error_on_different_platform):
634         (test_error_on_different_build_type):
635         (test_overrides):
636         (test_overrides__directory):
637         (test_overrides__duplicate):
638         (test_pixel_tests_flag):
639         (test_more_specific_override_resets_skip):
640         (SkippedTests.test_skipped_file_overrides_expectations):
641         (SkippedTests.test_skipped_dir_overrides_expectations):
642         (SkippedTests.test_skipped_file_overrides_overrides):
643         (SkippedTests.test_skipped_dir_overrides_overrides):
644         (ExpectationSyntaxTests.test_missing_colon):
645         (ExpectationSyntaxTests.test_too_many_equals_signs):
646         (SemanticTests.test_bug_format):
647         (SemanticTests.test_bad_bugid):
648         (SemanticTests.test_missing_bugid):
649         (SemanticTests.test_rebaseline):
650         (test_missing_file):
651         (test_more_modifiers):
652         (test_order_in_file):
653         (test_macro_overrides):
654         (RemoveConfigurationsTest.test_remove):
655         (test_remove_line):
656         (RebaseliningTest.test_remove):
657         (TestExpectationSerializationTests.test_serialize_parsed_expectations):
658         * Scripts/webkitpy/layout_tests/models/test_failures.py:
659         (determine_result_type):
660         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
661         (test_test_expectations):
662         * Scripts/webkitpy/layout_tests/port/test.py:
663         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
664         (MainTest.test_missing_and_unexpected_results):
665         (MainTest.test_retrying_and_flaky_tests):
666         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
667         (TestExpectationsTestCase.test_valid_expectations):
668         * Scripts/webkitpy/tool/commands/queries_unittest.py:
669         (PrintExpectationsTest.test_basic):
670         (PrintExpectationsTest.test_multiple):
671         (PrintExpectationsTest.test_full):
672         (PrintExpectationsTest.test_exclude):
673         (PrintExpectationsTest.test_csv):
674         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
675         (test_rebaseline_does_not_include_overrides):
676         (test_overrides_are_included_correctly):
677
678 2012-09-19  Dominic Mazzoni  <dmazzoni@google.com>
679
680         AX: A few control types are returning the wrong answer for isReadOnly
681         https://bugs.webkit.org/show_bug.cgi?id=96735
682
683         Reviewed by Chris Fleizach.
684
685         Exposing isReadOnly in an AccessibilityObject to DumpRenderTree.
686
687         * DumpRenderTree/chromium/TestRunner/AccessibilityUIElementChromium.cpp:
688         (AccessibilityUIElement::AccessibilityUIElement):
689         (AccessibilityUIElement::isReadOnlyGetterCallback):
690         * DumpRenderTree/chromium/TestRunner/AccessibilityUIElementChromium.h:
691         (AccessibilityUIElement):
692
693 2012-09-19  Sudarsana Nagineni  <sudarsana.nagineni@intel.com>
694
695         [WTR] Memory leaks in TestRunner::deliverWebIntent()
696         https://bugs.webkit.org/show_bug.cgi?id=97111
697
698         Reviewed by Kenneth Rohde Christiansen.
699
700         Fix memory leaks in deliverWebIntent() by adopting strings
701         created with WKStringCreateWithUTF8CString().
702
703         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
704         (WTR::TestRunner::deliverWebIntent):
705
706 2012-09-19  Simon Hausmann  <simon.hausmann@digia.com>
707
708         [Qt] Fix incremental builds with all-in-one-files and gccdepends
709
710         Reviewed by Tor Arne Vestbø.
711
712         Pass -MP to gcc when we use the gcc depends feature, to ensure that implicit rules
713         are not only created for header files but also for .cpp files. AllInOne.cpp files
714         include other .cpp files, and when those are removed we need those dummy rules to
715         avoid a "No rule to make Foo.cpp required by AllInOne.o" error.
716
717         * qmake/config.tests/gccdepends/empty.cpp:
718         (main):
719         * qmake/config.tests/gccdepends/gccdepends.pro:
720         * qmake/mkspecs/features/default_pre.prf:
721
722 2012-09-19  Rick Byers  <rbyers@chromium.org>
723
724         Do touch adjustment on GestureTapDown
725         https://bugs.webkit.org/show_bug.cgi?id=96677
726
727         Reviewed by Antonio Gomes.
728
729         Allow radius to be set for GestureTapDown events.
730         * DumpRenderTree/chromium/TestRunner/EventSender.cpp:
731         (EventSender::gestureEvent):
732
733 2012-09-19  Christophe Dumez  <christophe.dumez@intel.com>
734
735         [EFL] EFL's DRT does not support overriding 'WebKitCSSRegionsEnabled' preference
736         https://bugs.webkit.org/show_bug.cgi?id=97100
737
738         Reviewed by Gyuyoung Kim.
739
740         EFL's DRT now supports overriding the 'WebKitCSSRegionsEnabled'
741         preference, in order to match WebKitTestRunner functionality.
742
743         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
744         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
745         * DumpRenderTree/efl/TestRunnerEfl.cpp:
746         (TestRunner::overridePreference):
747
748 2012-09-18  Kangil Han  <kangil.han@samsung.com>
749
750         [WK2][WTR] CodeGeneratorTestRunner could keep original copyright.
751         https://bugs.webkit.org/show_bug.cgi?id=96181
752
753         Reviewed by Daniel Bates.
754
755         This patch enabled derived files, in DerivedSources/InjectedBundle, to keep original copyright.
756
757         * WebKitTestRunner/InjectedBundle/Bindings/CodeGeneratorTestRunner.pm:
758         (new):
759         (_parseLicenseBlock):
760         (_parseLicenseBlockFromFile):
761         (_defaultLicenseBlock):
762         (_licenseBlock):
763         (_generateHeaderFile):
764         (_generateImplementationFile):
765
766 2012-09-18  Byungwoo Lee  <bw80.lee@samsung.com>
767
768         Title string should be changed when document.title is set to ''.
769         https://bugs.webkit.org/show_bug.cgi?id=96793
770
771         Reviewed by Kenneth Rohde Christiansen.
772
773         Change dump format of dumpTitleChanges more understandable.
774         Uses single quotation marks for the title string.
775
776         * DumpRenderTree/blackberry/DumpRenderTree.cpp:
777         (BlackBerry::WebKit::DumpRenderTree::didReceiveTitleForFrame):
778         * DumpRenderTree/chromium/WebViewHost.cpp:
779         (WebViewHost::didReceiveTitle):
780         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
781         (DumpRenderTreeChrome::onFrameTitleChanged):
782         * DumpRenderTree/gtk/DumpRenderTree.cpp:
783         (webViewTitleChanged):
784         * DumpRenderTree/mac/FrameLoadDelegate.mm:
785         (-[FrameLoadDelegate webView:didReceiveTitle:forFrame:]):
786         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
787         (WebCore::DumpRenderTree::titleChanged):
788         * DumpRenderTree/win/FrameLoadDelegate.cpp:
789         (FrameLoadDelegate::didReceiveTitle):
790         * DumpRenderTree/wx/DumpRenderTreeWx.cpp:
791         (LayoutWebViewEventHandler::OnReceivedTitleEvent):
792         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
793         (WTR::InjectedBundlePage::didReceiveTitleForFrame):
794
795 2012-09-18  Szilard Ledan  <szledan@inf.u-szeged.hu>
796
797         EWS shouldn't sleep if there are new patches in its queue
798         https://bugs.webkit.org/show_bug.cgi?id=83038
799
800         Reviewed by Eric Seidel.
801
802         EWS tries to process a security patch. Of course it can't, because the EWS isn't
803         the member of the security group. But the problem is that after it can't process
804         the attachment, it says that queue is empty (but it isn't!) and it sleeps 2 minutes
805         and push the security patch to the end of the queue.
806         Now it stays in the loop until it finds a patch or the queue gets empty.
807
808         * Scripts/webkitpy/tool/commands/queues.py:
809         (AbstractPatchQueue._next_patch):
810         * Scripts/webkitpy/tool/commands/queues_unittest.py:
811         (AbstractPatchQueueTest.test_next_patch):
812
813 2012-09-18  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
814
815         [WK2][WTR] InjectedBundle::booleanForKey() should handle literals effectively
816         https://bugs.webkit.org/show_bug.cgi?id=97014
817
818         Reviewed by Kenneth Rohde Christiansen.
819
820         According to http://trac.webkit.org/wiki/EfficientStrings WTF::StringBuilder::appendLiteral() shall
821         be used for literals rather than WTF::StringBuilder::append().
822
823         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
824         (WTR::InjectedBundle::booleanForKey):
825
826 2012-09-18  Simon Hausmann  <simon.hausmann@digia.com>
827
828         Update my e-mail address.
829
830         * Scripts/webkitpy/common/config/committers.py:
831
832 2012-09-18  Andras Becsi  <andras.becsi@digia.com>
833
834         Update my e-mail address.
835
836         Unreviewed.
837
838         * Scripts/webkitpy/common/config/committers.py:
839
840 2012-09-18  Andras Becsi  <andras.becsi@digia.com>
841
842         [Qt] qt_webkit.pri should not be listed in Tools.pro
843
844         Reviewed and rubber-stamped by Simon Hausmann and Tor Arne Vestbø.
845
846         Since r128751 the module pri file is auto-generated
847         but it was still listed in OTHER_FILES.
848
849         * Tools.pro: Remove unneeded line.
850
851 2012-09-18  Simon Hausmann  <simon.hausmann@digia.com>
852
853         [Qt] Fix build with some versions of the gold linker
854
855         Reviewed by Tor Arne Vestbø.
856
857         Don't unconditionally pass --no-keep-memory to the linker, some versions might not support it.
858         Instead run a compile/link test first to see if it works.
859
860         * qmake/config.tests/gnuld/gnuld.pro: Added.
861         * qmake/config.tests/gnuld/main.cpp: Added.
862         (main):
863         * qmake/mkspecs/features/unix/default_post.prf:
864
865 2012-09-18  Simon Hausmann  <simon.hausmann@digia.com>
866
867         [Qt] Remove forced use of gold.
868
869         Reviewed by Tor Arne Vestbø.
870
871         The choice of what linker to use with WebKit should be taken by Qt's build system and ideally the same for all
872         modules of Qt. Then in turn it is usually up to the administrator of the machine. Recent Debian based systems
873         often offer the automatic use of gold through a symlink and a dpkg-diversion when installing the gold package.
874
875         * qmake/mkspecs/features/unix/default_post.prf:
876
877 2012-09-18  Simon Hausmann  <simon.hausmann@digia.com>
878
879         [Qt] Prospective Qt/Windows cross-compiling fix
880
881         Reviewed by Tor Arne Vestbø.
882
883         The win32 scope is not set when cross-compiling from Linux to Windows.
884
885         * qmake/mkspecs/features/functions.prf:
886
887 2012-09-18  Seokju Kwon  <seokju.kwon@samsung.com>
888
889         [EFL] Remove background view on EWebLauncher and MiniBrowser
890         https://bugs.webkit.org/show_bug.cgi?id=96905
891
892         Reviewed by Gyuyoung Kim.
893
894         The size of webview was changed after adding url bar.
895         And the background view is not necessary anymore, since it was used for debugging back in the day.
896
897         * EWebLauncher/main.c:
898         (_ELauncher):
899         (on_ecore_evas_resize):
900         (browserCreate):
901         * MiniBrowser/efl/main.c:
902         (_MiniBrowser):
903         (on_ecore_evas_resize):
904         (browserCreate):
905
906 2012-09-18  Sergio Villar Senin  <svillar@igalia.com>
907
908         [GTK] run-webkit-tests unable to find TestExpectations for WK2
909         https://bugs.webkit.org/show_bug.cgi?id=96998
910
911         Reviewed by Philippe Normand.
912
913         We should look for TestExpectations files in all the locations where
914         we currently look for Skipped files. This will allow
915         run-webkit-tests to look for TestExpectations files in
916         platform/gtk-wk2 and platform/wk2 if the "-2" flag is used.
917
918         * Scripts/webkitpy/layout_tests/port/gtk.py:
919         (GtkPort.expectations_files):
920
921 2012-09-18  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
922
923         [Qt] Fix build without the QtQuick module
924
925         Reviewed by Simon Hausmann.
926
927         * qmake/mkspecs/features/features.prf:
928
929 2012-09-18  Stephanie Lewis  <slewis@apple.com>
930
931         Build fix after http://trac.webkit.org/projects/webkit/changeset/128852.
932
933         Unreviewed.
934
935         * DumpRenderTree/mac/DumpRenderTree.mm:
936         (poseAsClass):
937
938 2012-09-18  Allan Sandfeld Jensen  <allan.jensen@nokia.com>
939
940         Unreviewed update of email addresses for Berlin QtWebKit office.
941
942         * Scripts/webkitpy/common/config/committers.py:
943
944 2012-09-17  Zan Dobersek  <zandobersek@gmail.com>
945
946         [GTK] fast/loader/display-image-unset-can-block-image-and-can-reload-in-place.html failing after r128645
947         https://bugs.webkit.org/show_bug.cgi?id=96899
948
949         Reviewed by Martin Robinson.
950
951         When overriding the 'WebKitDisplayImageKey' preference, use the
952         'auto-load-images' property of WebKitWebSettings as the property which
953         should be updated with the corresponding preference value.
954
955         * DumpRenderTree/gtk/DumpRenderTree.cpp:
956         (resetDefaultsToConsistentValues):
957         * DumpRenderTree/gtk/TestRunnerGtk.cpp:
958         (TestRunner::overridePreference):
959
960 2012-09-17  Pratik Solanki  <psolanki@apple.com>
961
962         DumpRenderTree and WebKitTestRunner should compile with -Wundef on Mac
963         https://bugs.webkit.org/show_bug.cgi?id=96973
964
965         Reviewed by Dan Bernstein.
966
967         * DumpRenderTree/mac/Configurations/Base.xcconfig:
968         * WebKitTestRunner/Configurations/Base.xcconfig:
969         * WebKitTestRunner/PlatformWebView.h: Use #ifdef __OBJC__ and not #if.
970
971 2012-09-17  Dirk Pranke  <dpranke@chromium.org>
972
973         nrwt: remove "unexpected EOF" warnings
974         https://bugs.webkit.org/show_bug.cgi?id=96970
975
976         Reviewed by Ojan Vafai.
977
978         After debugging this a bit, it looks like there aren't any cases
979         that I can reproduce where a read() of zero indicates something
980         actually wrong; either it is a prelude to a crash, or a false
981         negative. So, I'm removing these warnings and adding a comment.
982
983         * Scripts/webkitpy/layout_tests/port/server_process.py:
984         (ServerProcess._wait_for_data_and_update_buffers_using_select):
985
986 2012-09-17  Dirk Pranke  <dpranke@chromium.org>
987
988         [chromium] ASAN bot is crashing at the end of the run
989         https://bugs.webkit.org/show_bug.cgi?id=96967
990
991         Reviewed by Abhishek Arya.
992
993         The ASAN bot is crashing attempting to decode some output into
994         UTF-8; there's no reason to do this, so let's not do this and
995         see if something else is going on as well.
996
997         * Scripts/webkitpy/layout_tests/port/chromium.py:
998         (ChromiumPort._get_crash_log):
999
1000 2012-09-17  Dirk Pranke  <dpranke@chromium.org>
1001
1002         nrwt: --results-directory isn't getting printed properly
1003         https://bugs.webkit.org/show_bug.cgi?id=96965
1004
1005         Reviewed by Ojan Vafai.
1006
1007         options.results_directory isn't actually initialized with the
1008         default values until after we call print_config(), so this
1009         changes things to print the value directly.
1010
1011         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1012         (run):
1013         * Scripts/webkitpy/layout_tests/views/printing.py:
1014         (Printer.print_config):
1015         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
1016         (Testprinter.test_print_config):
1017
1018 2012-09-17  Rob Buis  <rbuis@rim.com>
1019
1020         [BlackBerry] Enable VIDEO_TRACK
1021         https://bugs.webkit.org/show_bug.cgi?id=96949
1022
1023         Reviewed by Antonio Gomes.
1024
1025         * Scripts/webkitperl/FeatureList.pm:
1026
1027 2012-09-17  Rick Byers  <rbyers@chromium.org>
1028
1029         Add handling of new GestureTapCancel in DRT
1030
1031         https://bugs.webkit.org/show_bug.cgi?id=96183
1032
1033         Reviewed by Antonio Gomes.
1034
1035         * DumpRenderTree/chromium/TestWebPlugin.cpp:
1036         (TestWebPlugin::handleInputEvent):
1037         * DumpRenderTree/chromium/EventSender.cpp:
1038         (EventSender::gestureTapCancel):
1039
1040 2012-09-17  Philip Rogers  <pdr@google.com>
1041
1042         Teach style checker about preprocessor directive indentation rules
1043         https://bugs.webkit.org/show_bug.cgi?id=96874
1044
1045         Reviewed by Adam Barth.
1046
1047         Preprocessor directives (#ifdef, #include, #define, etc.) should not be indented.
1048         This is not explicit in our style guide but is generally followed in our code.
1049         Searching for violations in our codebase shows these are rarely indented:
1050             #include - indented in 6 files
1051             #ifdef - indented in 0 files
1052             #ifndef - indented in 1 file
1053             #define - indented in 11 files
1054             #if - indented in 7 files
1055
1056         * Scripts/webkitpy/style/checkers/cpp.py:
1057         (check_directive_indentation):
1058
1059             This is the simple test where we look for spaces followed by a #.
1060
1061         (check_style):
1062         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
1063
1064             A few tests needed to be modified because they had unintentionally indented
1065             preprocessor directives.
1066
1067         (CppStyleTest.test_build_class.Foo):
1068         (CppStyleTest.test_build_class):
1069         (CppStyleTest.test_build_class.DERIVE_FROM_GOO):
1070         (WebKitStyleTest.test_line_breaking):
1071         (WebKitStyleTest.test_directive_indentation):
1072
1073 2012-09-17  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1074
1075         [Qt] Auto-generate the module pri file for QtWebKit
1076
1077         Reviewed by Simon Hausmann.
1078
1079         * qmake/qt_webkit.pri: Removed.
1080
1081 2012-09-14  Dirk Pranke  <dpranke@chromium.org>
1082
1083         nrwt: --additional-platform-dir is broken on chromium ports
1084         https://bugs.webkit.org/show_bug.cgi?id=96840
1085
1086         Reviewed by Ojan Vafai.
1087
1088         Looks like I broke this when cleaning things up as part of
1089         adding ML support. Fixing and adding a test.
1090
1091         * Scripts/webkitpy/layout_tests/port/chromium.py:
1092         (ChromiumPort.default_baseline_search_path):
1093         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
1094         (ChromiumAndroidPort.default_baseline_search_path):
1095         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
1096         (ChromiumLinuxPort.default_baseline_search_path):
1097         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
1098         (ChromiumMacPort.__init__):
1099         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
1100         (ChromiumWinPort.setup_environ_for_server):
1101         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
1102         (test_check_build):
1103         (test_additional_platform_directory):
1104
1105 2012-09-13  Stephanie Lewis  <slewis@apple.com>
1106
1107         mac-future ignores previous platform TestExpectations.
1108         https://bugs.webkit.org/show_bug.cgi?id=96718.
1109
1110         Reviewed by Dirk Pranke.
1111
1112         TestExpectations match an expectation's configuraton against a list of configurations that are 
1113         valid for that port.  That list does not contain mac-future so none of the expectations can be
1114         applied.
1115
1116         * Scripts/webkitpy/layout_tests/port/apple.py:
1117         (ApplePort._generate_all_test_configurations):
1118
1119 2012-09-14  Adam Barth  <abarth@webkit.org>
1120
1121         Remove webkitPostMessage
1122         https://bugs.webkit.org/show_bug.cgi?id=96577
1123
1124         Reviewed by Ojan Vafai.
1125
1126         Add ENABLE_LEGACY_VENDOR_PREFIXES flag.
1127
1128         * Scripts/webkitperl/FeatureList.pm:
1129         * qmake/mkspecs/features/features.pri:
1130
1131 2012-09-14  Alexey Proskuryakov  <ap@apple.com>
1132
1133         [WK2] webarchive/loading/javascript-url-iframe-crash.html fails
1134         https://bugs.webkit.org/show_bug.cgi?id=96817
1135
1136         Reviewed by Kenneth Rohde Christiansen.
1137
1138         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1139         (WTR::InjectedBundlePage::didFailProvisionalLoadWithErrorForFrame): This callback
1140         got overlooked when adding logging elsewhere.
1141
1142 2012-09-14  Peter Beverloo  <peter@chromium.org>
1143
1144         [Chromium] Build fix for DumpRenderTree following r128628
1145         https://bugs.webkit.org/show_bug.cgi?id=96808
1146
1147         Reviewed by Tony Chang.
1148
1149         When building DumpRenderTree from the Chromium project, a compile error
1150         shows up because including the header ForwardIOStreamsAndroid.h cannot
1151         be located. TestWebKitAPI does this correctly.
1152
1153         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
1154
1155 2012-09-14  Dana Jansens  <danakj@chromium.org>
1156
1157         Change cc-bugs@google.com watchlist to cc-bugs@chromium.org
1158         https://bugs.webkit.org/show_bug.cgi?id=96805
1159
1160         Reviewed by Dirk Pranke.
1161
1162         * Scripts/webkitpy/common/config/committers.py:
1163         * Scripts/webkitpy/common/config/watchlist:
1164
1165 2012-09-14  Peter Beverloo  <peter@chromium.org>
1166
1167         [Chromium] Support the --{in,out,err}-fifo arguments on TestWebKitAPI and webkit_unit_tests
1168         https://bugs.webkit.org/show_bug.cgi?id=96687
1169
1170         Reviewed by Tony Chang.
1171
1172         Android's DumpRenderTree currently supports these arguments, implemented
1173         as part of TestShellAndroid:
1174         http://trac.webkit.org/browser/trunk/Tools/DumpRenderTree/chromium/TestShellAndroid.cpp?rev=128496
1175
1176         They're used by the layout test runner to get the STDOUT and STDERR while
1177         a layout test run is in process, which is a safer alternative to parsing
1178         all the logcat output manually. The implementation can be seen here:
1179         http://trac.webkit.org/browser/trunk/Tools/Scripts/webkitpy/layout_tests/port/chromium_android.py?rev=128496#L590
1180
1181         This patch generalizes parsing of and applying the effects of these arguments
1182         so that they can be used for TestWebKitAPI and webkit_unit_tests as well.
1183         After this patch, this will make it possible to pull out the output-reading
1184         code from Android's layout test port and generalize it so it can be re-used
1185         in the new test-runner for the other two test suites.
1186
1187         This has no effect when compiling and running these tests as part of Chromium
1188         code, which has a much more advanced test-runner that does parse complete log
1189         output, but also directly depends on code licensed under Apache 2.
1190
1191         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
1192         * DumpRenderTree/chromium/TestShellAndroid.cpp:
1193         (platformInit):
1194         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
1195
1196 2012-09-14  Zoltan Horvath  <zoltan@webkit.org>
1197
1198         check-webkit-style should not warn in case of NONCOPYABLE and FAST_ALLOCATED macros
1199         https://bugs.webkit.org/show_bug.cgi?id=96656
1200
1201         Reviewed by Adam Barth.
1202
1203         We should allow to use WTF_MAKE_NONCOPYABLE(ClassName) and WTF_MAKE_FAST_ALLOCATED
1204         macros in 1 line.
1205
1206         * Scripts/webkitpy/style/checkers/cpp.py:
1207         (check_style.definitions):
1208         (check_style): Add rule.
1209         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
1210         (WebKitStyleTest.test_line_breaking): Add unittest.
1211
1212 2012-09-14  Christophe Dumez  <christophe.dumez@intel.com>
1213
1214         WebKitTestRunner needs layoutTestController.dumpDatabaseCallbacks
1215         https://bugs.webkit.org/show_bug.cgi?id=57570
1216
1217         Reviewed by Kenneth Rohde Christiansen.
1218
1219         Implement testRunner.dumpDatabaseCallbacks in WebKitTestRunner
1220         and dump the information expected by the test cases.
1221
1222         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
1223         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
1224         (WTR::InjectedBundle::beginTesting): Explicitly set the default
1225         database quota.
1226         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1227         (WTR::securityOriginToStr): New convenience function to convert
1228         a security origin to the string expected in test results.
1229         (WTR):
1230         (WTR::InjectedBundlePage::InjectedBundlePage):
1231         (WTR::InjectedBundlePage::didExceedDatabaseQuota):
1232         (WTR::InjectedBundlePage::didReachApplicationCacheOriginQuota): Minor
1233         refactoring to share code with didExceedDatabaseQuota.
1234         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
1235         (InjectedBundlePage):
1236         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
1237         (WTR::TestRunner::TestRunner):
1238         * WebKitTestRunner/InjectedBundle/TestRunner.h:
1239         (WTR::TestRunner::dumpDatabaseCallbacks):
1240         (WTR::TestRunner::shouldDumpDatabaseCallbacks):
1241         (TestRunner):
1242
1243 2012-09-14  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1244
1245         [Qt] Set force_static_libs_as_shared early enough to be caught by configure
1246
1247         Rubber-stamped by Simon Hausmann.
1248
1249         * qmake/mkspecs/features/default_post.prf:
1250
1251 2012-09-14  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1252
1253         [Qt] Remove old cruft from the qmake build system
1254
1255         Rubber-stamped by Simon Hausmann.
1256
1257         * DumpRenderTree/qt/DumpRenderTree.pro:
1258         * qmake/mkspecs/features/default_post.prf:
1259
1260 2012-09-14  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1261
1262         [Qt] Enable force_static_libs_as_shared for development
1263
1264         Not enabled for buildbots or production-builds (part of Qt5)
1265
1266         Reviewed by Simon Hausmann.
1267
1268         * qmake/mkspecs/features/default_post.prf:
1269
1270 2012-09-14  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1271
1272         [Qt] Make force_static_libs_as_shared work on Mac OS
1273
1274         We had to move a few LIBS += around that were in the wrong place,
1275         and not caught when everything was just linked into the final
1276         QtWebKit library.
1277
1278         Reviewed by Simon Hausmann.
1279
1280         * qmake/mkspecs/features/default_post.prf:
1281         * qmake/mkspecs/features/force_static_libs_as_shared.prf:
1282
1283 2012-09-14  Simon Hausmann  <simon.hausmann@nokia.com>
1284
1285         [Qt] Fix build of WTR on Windows
1286
1287         Reviewed by Tor Arne Vestbø.
1288
1289         WTR/qt/main.cpp includes TestController.h, which lives in just WTR/
1290         With qmake's unix makefile generator as well as with the MingW generator,
1291         the directory where the pro file lives (WTR/) is automatically added to
1292         the include search path, but not so with MSVC. Instead of relying on qmake's
1293         implicit feature here, add WTR/ explicitly to the include search path.
1294
1295         * WebKitTestRunner/Target.pri:
1296
1297 2012-09-14  Christophe Dumez  <christophe.dumez@intel.com>
1298
1299         WebKitTestRunner needs layoutTestController.setStopProvisionalFrameLoads
1300         https://bugs.webkit.org/show_bug.cgi?id=42691
1301
1302         Reviewed by Kenneth Rohde Christiansen.
1303
1304         Add implementation for testrunner.setStopProvisionalFrameLoads()
1305         in WebKitTestRunner.
1306
1307         Original patch by Alexey Proskuryakov.
1308
1309         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
1310         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1311         (WTR::InjectedBundlePage::didStartProvisionalLoadForFrame):
1312         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
1313         (WTR::TestRunner::TestRunner):
1314         * WebKitTestRunner/InjectedBundle/TestRunner.h:
1315         (TestRunner):
1316         (WTR::TestRunner::setStopProvisionalFrameLoads):
1317         (WTR::TestRunner::shouldStopProvisionalFrameLoads):
1318
1319 2012-09-14  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1320
1321         [Qt] Fix handling of debug/release/debug_and_release/build_all
1322
1323         We now pick up the defaults from Qt, in default_pre (CONFIG already
1324         contains the appropriate values for debug and release). We then let
1325         the command line arguments to qmake, or the project files themselves,
1326         override the configuration, and finally we sanitize the configuration
1327         in default_post.
1328
1329         Reviewed by Simon Hausmann.
1330
1331         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
1332         * Scripts/webkitdirs.pm:
1333         (buildQMakeProjects):
1334         * qmake/mkspecs/features/default_post.prf:
1335         * qmake/mkspecs/features/default_pre.prf:
1336         * qmake/mkspecs/features/mac/default_pre.prf:
1337
1338 2012-09-14  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
1339
1340         WebKitTestRunner needs layoutTestController.setTabKeyCyclesThroughElements
1341         https://bugs.webkit.org/show_bug.cgi?id=42687
1342
1343         Reviewed by Kenneth Rohde Christiansen.
1344
1345         Implement testRunner.setTabKeyCyclesThroughElements for
1346         WebKitTestRunner.
1347
1348         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
1349         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
1350         (WTR::InjectedBundle::beginTesting):
1351         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
1352         (WTR::TestRunner::setTabKeyCyclesThroughElements):
1353         (WTR):
1354         * WebKitTestRunner/InjectedBundle/TestRunner.h:
1355         (TestRunner):
1356
1357 2012-09-14  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
1358
1359         EventSendingController::keyDown does not support non-array modifier arguments
1360         https://bugs.webkit.org/show_bug.cgi?id=96727
1361
1362         Reviewed by Kenneth Rohde Christiansen.
1363
1364         Add support for handling a string as modifier argument to
1365         EventSendingController::keyDown.
1366
1367         * WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
1368         (WTR::parseModifierArray):
1369
1370 2012-09-14  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
1371
1372         WebKitTestRunner needs layoutTestController.setCustomPolicyDelegate
1373         https://bugs.webkit.org/show_bug.cgi?id=42546
1374
1375         Reviewed by Kenneth Rohde Christiansen.
1376
1377         Exported TestRunner::setCustomPolicyDelegate() method.
1378         Allowed Policy Delegate dumping even if we're not in 'waitUntilDone' mode.
1379
1380         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
1381         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1382         (WTR::InjectedBundlePage::decidePolicyForNavigationAction):
1383
1384 2012-09-14  Zan Dobersek  <zandobersek@gmail.com>
1385
1386         [GTK] Clear application cache between tests in DumpRenderTree
1387         https://bugs.webkit.org/show_bug.cgi?id=96543
1388
1389         Reviewed by Philippe Normand.
1390
1391         Call the new DumpRenderTreeSupportGtk method to clear application cache
1392         after each test in DumpRenderTree.
1393
1394         Set the XDG_CACHE_HOME environment variable when running DumpRenderTree
1395         as well. It's now being set in XvfbDriver._start so that each driver is assigned
1396         a cache directory that's constructed out of the driver's name and its worker
1397         number. These directories are located in the layout tests results directory.
1398         At least on the builders this ensures a driver-specific empty cache directory
1399         with each test run.
1400
1401         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1402         (runTest):
1403         * Scripts/webkitpy/layout_tests/port/gtk.py:
1404         (GtkPort.setup_environ_for_server):
1405         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
1406         (XvfbDriver._start):
1407
1408 2012-09-14  Simon Hausmann  <simon.hausmann@nokia.com>
1409
1410         [Qt] Fix linking on Windows
1411
1412         Reviewed by Tor Arne Vestbø.
1413
1414         When compiling code _after_ QtWebKit.dll, we need to make sure that the QWEBKIT_EXPORT
1415         macro turns into an import macro, which means QT_MAKEDLL should only be set _up until_
1416         the QtWebKit.dll creation in the build process and be off afterwards. This is done in
1417         win32/default_post.prf using a !contains(QT, webkit): DEFINES += QT_MAKEDLL.
1418
1419         r128450 sanitized the values of QT and PKGCONFIG to only contain the dependencies we
1420         want for QtWebKit.dll, but unfortunately that code was applied to any binary built also
1421         after the module. Therefore QT did never contain webkit anymore and we always passed
1422         QT_MAKEDLL to DEFINES, resulting in QWEBKIT_EXPORT to _not_ turn into an import macro.
1423
1424         This patch introduces a creating_module configuration that is used to
1425         scope the qt/pkgconfig sanitization.
1426
1427         * qmake/mkspecs/features/default_post.prf:
1428
1429 2012-09-14  Christophe Dumez  <christophe.dumez@intel.com>
1430
1431         [WK2][WKTR] TestRunner is not dumping some frame load callbacks information
1432         https://bugs.webkit.org/show_bug.cgi?id=96740
1433
1434         Reviewed by Kenneth Rohde Christiansen.
1435
1436         Dump information expected by the test cases about the following
1437         frame load callbacks:
1438         "didDisplayInsecureContentForFrame"
1439         "didRunInsecureContentForFrame"
1440         "didDetectXSSForFrame"
1441
1442         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1443         (WTR::InjectedBundlePage::willPerformClientRedirectForFrame): Print a URL
1444         as expected by test results by calling pathSuitableForTestResult() instead
1445         of printing raw URL.
1446         (WTR::InjectedBundlePage::didDisplayInsecureContentForFrame):
1447         (WTR::InjectedBundlePage::didRunInsecureContentForFrame):
1448         (WTR::InjectedBundlePage::didDetectXSSForFrame):
1449
1450 2012-09-14  Peter Beverloo  <peter@chromium.org>
1451
1452         The runtime/unsigned category should be valid in the cpp style-checker.
1453         https://bugs.webkit.org/show_bug.cgi?id=96748
1454
1455         Reviewed by Csaba Osztrogonác.
1456
1457         This is causing the webkitpy tests to fail because it's not listed in the
1458         category array. Furthermore, change an "int" to "long" because it's
1459         throwing a double error in one of the earlier asserts.
1460
1461         * Scripts/webkitpy/style/checkers/cpp.py:
1462         (CppChecker):
1463         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
1464         (WebKitStyleTest.test_names):
1465
1466 2012-09-13  Kenneth Rohde Christiansen  <kenneth@webkit.org>
1467
1468         Evas_Object* is a ref'ed structure, so tread it as such
1469         https://bugs.webkit.org/show_bug.cgi?id=96659
1470
1471         Reviewed by Adam Barth.
1472
1473         Replace OwnPtr<Evas_Object> with RefPtr.
1474
1475         * DumpRenderTree/efl/ImageDiff.cpp:
1476         (calculateDifference):
1477         (printImageDifferences):
1478         (readImageFromStdin):
1479         (main):
1480
1481 2012-09-14  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
1482
1483         Unreviewed, rolling out r128507.
1484         http://trac.webkit.org/changeset/128507
1485         https://bugs.webkit.org/show_bug.cgi?id=96659
1486
1487         Revert. r128507 makes too many crashes in EFL layout test bots.
1488
1489         * DumpRenderTree/efl/ImageDiff.cpp:
1490         (calculateDifference):
1491         (printImageDifferences):
1492         (readImageFromStdin):
1493         (main):
1494
1495 2012-09-13  Kevin Funk  <kevin.funk@kdab.com>
1496
1497         Make compile with both OS(WINCE) and PLATFORM(QT) support
1498         https://bugs.webkit.org/show_bug.cgi?id=95536
1499
1500         Reviewed by Simon Hausmann.
1501
1502         Fix wince support in qmake files
1503
1504         * Tools.pro:
1505         * qmake/mkspecs/features/configure.prf:
1506         * qmake/mkspecs/features/default_post.prf:
1507         * qmake/mkspecs/features/features.prf:
1508         * qmake/mkspecs/features/functions.prf:
1509
1510 2012-09-13  KwangYong Choi  <ky0.choi@samsung.com>
1511
1512         [WK2] Add color picker API support for WebKit2
1513         https://bugs.webkit.org/show_bug.cgi?id=95058
1514
1515         Reviewed by Gyuyoung Kim.
1516
1517         Initialization code for color picker callbacks in UI client.
1518
1519         * MiniBrowser/mac/BrowserWindowController.m:
1520         (-[BrowserWindowController awakeFromNib]):
1521         * MiniBrowser/win/BrowserView.cpp:
1522         (BrowserView::create):
1523         * WebKitTestRunner/TestController.cpp:
1524         (WTR::TestController::createOtherPage):
1525         (WTR::TestController::initialize):
1526
1527 2012-09-13  Yuta Kitamura  <yutak@chromium.org>
1528
1529         Upgrade yutak to reviewer
1530         https://bugs.webkit.org/show_bug.cgi?id=96721
1531
1532         Reviewed by Kent Tamura.
1533
1534         * Scripts/webkitpy/common/config/committers.py:
1535
1536 2012-09-13  Adenilson Cavalcanti  <cavalcantii@gmail.com>
1537
1538         Make the style checker report usage of "unsigned int"
1539         https://bugs.webkit.org/show_bug.cgi?id=96693
1540
1541         Reviewed by Dirk Pranke.
1542
1543         Add a rule to check for uses of unsigned int, as the updated coding style recomends.
1544
1545         * Scripts/webkitpy/style/checkers/cpp.py:
1546         (check_language):
1547
1548 2012-09-13  Philip Rogers  <pdr@google.com>
1549
1550         Add gender-neutral form of webkit-patch land-cowboy
1551         https://bugs.webkit.org/show_bug.cgi?id=96709
1552
1553         Reviewed by Adam Barth.
1554
1555         Add webkit-patch land-cowhand, the confident and courageous form of webkit-patch for everyone.
1556
1557         * Scripts/webkitpy/tool/commands/download.py:
1558         (LandCowhand):
1559
1560 2012-09-13  Seokju Kwon  <seokju.kwon@samsung.com>
1561
1562         [EFL] Remove some parameters in browserCreate()
1563         https://bugs.webkit.org/show_bug.cgi?id=96499
1564
1565         Reviewed by Kenneth Rohde Christiansen.
1566
1567         There are too many parameters in browserCreate(). Lots of it seems like settings from command line argument.
1568         And I have moved some parameters into User_Arguments.
1569
1570         * EWebLauncher/main.c:
1571         (_User_Arguments):
1572         (_ELauncher):
1573         (on_key_down):
1574         (browserCreate):
1575         (findThemePath):
1576         (parseUserArguments):
1577         (main):
1578
1579 2012-09-13  Mark Lam  <mark.lam@apple.com>
1580
1581         Unreviewed. Adding myself to the committers list.
1582
1583         * Scripts/webkitpy/common/config/committers.py:
1584
1585 2012-09-13  Ryosuke Niwa  <rniwa@webkit.org>
1586
1587         run-perf-tests output cryptic error when the config file is missing.
1588         https://bugs.webkit.org/show_bug.cgi?id=96453
1589
1590         Reviewed by Tony Chang.
1591
1592         Add a special error message when a configuration file is missing.
1593
1594         Also update the help message of --source-json-path to signify the fact it specifies
1595         the configuration file on performance tests bots.
1596
1597         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1598         (PerfTestsRunner._parse_args):
1599         (PerfTestsRunner._generate_and_show_results):
1600         (PerfTestsRunner._merge_slave_config_json):
1601         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
1602         (_test_run_with_json_output): Make upload to fail by default and assert the result
1603         in the function so that we can return logs instead.
1604         (_test_run_with_json_output.mock_upload_json):
1605         (test_run_with_json_output):
1606         (test_run_with_description):
1607         (test_run_respects_no_results):
1608         (test_run_with_slave_config_json):
1609         (test_run_with_bad_slave_config_json):
1610         (test_run_with_multiple_repositories):
1611
1612 2012-09-13  Kenneth Rohde Christiansen  <kenneth@webkit.org>
1613
1614         Evas_Object* is a ref'ed structure, so tread it as such
1615         https://bugs.webkit.org/show_bug.cgi?id=96659
1616
1617         Reviewed by Adam Barth.
1618
1619         Replace OwnPtr<Evas_Object> with RefPtr.
1620
1621         * DumpRenderTree/efl/ImageDiff.cpp:
1622         (calculateDifference):
1623         (printImageDifferences):
1624         (readImageFromStdin):
1625         (main):
1626
1627 2012-09-13  Csaba Osztrogonác  <ossy@webkit.org>
1628
1629         One more unreviewed trivial fix after r128399.
1630
1631         * BuildSlaveSupport/build.webkit.org-config/config.json: Fix URLs of the SVN mirror server.
1632
1633 2012-09-13  Csaba Osztrogonác  <ossy@webkit.org>
1634
1635         Unreviewed trivial fix after r128399.
1636
1637         * BuildSlaveSupport/wait-for-SVN-server.py: Renamed from Tools/BuildSlaveSupport/build.webkit.org-config/wait-for-SVN-server.py.
1638         (getLatestSVNRevision):
1639         (waitForSVNRevision):
1640
1641 2012-09-13  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
1642
1643         [WK2][WTR] WebKitTestRunner needs testRunner.callShouldCloseOnWebView
1644         https://bugs.webkit.org/show_bug.cgi?id=96366
1645
1646         Reviewed by Anders Carlsson.
1647
1648         Implement testRunner.callShouldCloseOnWebView for WebKitTestRunner
1649         by calling shouldClose() on the FrameLoader.
1650
1651         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
1652         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
1653         (WTR::TestRunner::callShouldCloseOnWebView):
1654         (WTR):
1655         * WebKitTestRunner/InjectedBundle/TestRunner.h:
1656         (TestRunner):
1657
1658 2012-09-13  Sheriff Bot  <webkit.review.bot@gmail.com>
1659
1660         Unreviewed, rolling out r128453.
1661         http://trac.webkit.org/changeset/128453
1662         https://bugs.webkit.org/show_bug.cgi?id=96681
1663
1664         Having tests use the same appcache directory leads to timeouts
1665         (Requested by zdobersek on #webkit).
1666
1667         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1668         (runTest):
1669         * Scripts/webkitpy/layout_tests/port/gtk.py:
1670         (GtkPort.setup_environ_for_server):
1671
1672 2012-09-13  James Robinson  <jamesr@chromium.org>
1673
1674         [chromium] Temporarily remove Android Builder (dbg) from garden-o-matic view
1675         https://bugs.webkit.org/show_bug.cgi?id=96678
1676
1677         Reviewed by Dirk Pranke.
1678
1679         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
1680         (.):
1681
1682 2012-09-13  Rob Buis  <rbuis@rim.com>
1683
1684         Unreviewed changes to watchlist.
1685
1686         - Add myself to BlackBerry bugs
1687
1688         * Scripts/webkitpy/common/config/watchlist:
1689
1690 2012-09-13  Stephen Chenney  <schenney@chromium.org>
1691
1692         Add an SVG entry to the watchlist
1693
1694         Unreviewed config change.
1695
1696         * Scripts/webkitpy/common/config/watchlist:
1697         - Created a new SVG watchlist label covering WebCore/svg and WebCore/rendering/svg
1698         - Added myself and pdr as watchers
1699
1700 2012-09-13  Simon Hausmann  <simon.hausmann@nokia.com>
1701
1702         [Qt] Make WebKit2 work on Windows with Qt 5
1703         https://bugs.webkit.org/show_bug.cgi?id=76776
1704
1705         Reviewed by Tor Arne Vestbø.
1706
1707         WebKit2 should compile with Qt on Windows now, so enable it. This patch in fact corrects
1708         the check to disable WebKit when we don't have USE_3D_GRAPHICS available, because we need
1709         that for the GL TextureMapper (it's only built when that feature is set).
1710
1711         * qmake/mkspecs/features/configure.prf:
1712
1713 2012-09-13  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1714
1715         [Qt] Ensure that QT is finalized before loading qt_module.prf
1716
1717         As qt_module.prf does dependency tracking based on the content of the
1718         QT variable. The intermediate WebKit modules modify the variable in
1719         their .pri files, so we have to ensure we've loaded all the modules
1720         before loading qt_module.
1721
1722         Reviewed by Tor Arne Vestbø.
1723         Patch by Oswald Buddenhagen <oswald.buddenhagen@nokia.com> on 2012-09-13
1724
1725         * qmake/mkspecs/features/default_post.prf:
1726         * qmake/mkspecs/features/webkit_modules.prf: Added.
1727
1728 2012-09-13  Simon Hausmann  <simon.hausmann@nokia.com>
1729
1730         [Qt] Fix build of Qt WTR/MiniBrowser with MSVC and nmake
1731         https://bugs.webkit.org/show_bug.cgi?id=96633
1732
1733         Reviewed by Csaba Osztrogonác.
1734
1735         WTR and MiniBrowser both have DRT in their include path (for the shared font stuff). Batch
1736         builds with MSVC and nmake result in nmake picking up main.cpp from the wrong directory.
1737         It's the usual problem and the usual solution is to rename one of the conflicting files.
1738         This patch chooses to rename DRT instead of WTR and MBR.
1739
1740         Fixed include order at the same time to match style guide.
1741
1742         * DumpRenderTree/qt/DumpRenderTree.pro:
1743         * DumpRenderTree/qt/DumpRenderTreeMain.cpp: Renamed from Tools/DumpRenderTree/qt/main.cpp.
1744         (messageHandler):
1745         (isOption):
1746         (takeOptionValue):
1747         (printUsage):
1748         (main):
1749
1750 2012-09-13  Kenneth Rohde Christiansen  <kenneth@webkit.org>
1751
1752         The Script run-launcher should support -2 in combination with --efl
1753         https://bugs.webkit.org/show_bug.cgi?id=96639
1754
1755         Reviewed by Simon Hausmann.
1756
1757         Support running the MiniBrowser then supplied -2 as argument.
1758
1759         * Scripts/run-launcher:
1760
1761 2012-09-13  Raphael Kubo da Costa  <raphael.kubo.da.costa@intel.com>
1762
1763         Unreviewed, add my Intel email address to the list.
1764
1765         * Scripts/webkitpy/common/config/committers.py:
1766
1767 2012-09-13  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1768
1769         [Qt] Don't overwrite LIBS_PRIVATE when sanitizing LIBS
1770
1771         Reviewed by Ossy.
1772
1773         * qmake/mkspecs/features/default_post.prf:
1774
1775 2012-09-13  Kenneth Rohde Christiansen  <kenneth@webkit.org>
1776
1777         [EFL][DRT] Implement LayoutTestController::layerTreeAsText
1778         https://bugs.webkit.org/show_bug.cgi?id=82294
1779
1780         Reviewed by Gyuyoung Kim.
1781
1782         Implement layerTreeAsText using the DumpRenderTreeSupportEfl method.
1783
1784         * DumpRenderTree/efl/TestRunnerEfl.cpp:
1785         (TestRunner::layerTreeAsText):
1786
1787 2012-09-13  Zan Dobersek  <zandobersek@gmail.com>
1788
1789         [GTK] Clear application cache between tests in DumpRenderTree
1790         https://bugs.webkit.org/show_bug.cgi?id=96543
1791
1792         Reviewed by Philippe Normand.
1793
1794         Call the new DumpRenderTreeSupportGtk method to clear application cache
1795         after each test.
1796
1797         Set the XDG_CACHE_HOME environment variable when running DumpRenderTree
1798         as well. Currently it points to a subdirectory in the layout tests results
1799         directory. At least on the builders this ensures an empty cache directory
1800         with each test run.
1801
1802         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1803         (runTest):
1804         * Scripts/webkitpy/layout_tests/port/gtk.py:
1805         (GtkPort.setup_environ_for_server):
1806
1807 2012-09-13  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1808
1809         [Qt] Make sure LIBS, PKGCONFIG, and QT, do not affect QtWebKit's prl file
1810
1811         Unless it's actually a dependency of the public QtWebKit API.
1812
1813         We could replace every occurance of LIBS, PKGCONFIG, and QT, in the pri
1814         files with their _PRIVATE equivivalent, but that's likely to break when
1815         someone adds a new QT/LIBS/PKGCONFIG += foo line somewhere. Instead we
1816         clean up the variables in default_post.
1817
1818         This means that the CONFIG -= explicitlib in linkAgainstLibrary() is no
1819         longer needed, as it was a workaround for the intermediate libraries ending
1820         up in the prl file. And, since CONFIG -= staticlib was there to support
1821         the explicitlib option, we can remove that too (as well as the exception
1822         for gprof, since it would be empty). If gprof needs tweaks to always link
1823         statically they should go in gprof.prf.
1824
1825         Reviewed by Simon Hausmann.
1826
1827         * qmake/mkspecs/features/default_post.prf:
1828         * qmake/mkspecs/features/functions.prf:
1829
1830 2012-09-13  Mario Sanchez Prada  <msanchez@igalia.com>
1831
1832         [GTK] "Infinite" loop in AccessibilityUIElementGtk.cpp
1833         https://bugs.webkit.org/show_bug.cgi?id=96632
1834
1835         Reviewed by Carlos Garcia Campos.
1836
1837         Fix this by using atk_object_get_n_accessible_children instead of
1838         calling getChildren() from childCount.
1839
1840         * WebKitTestRunner/InjectedBundle/gtk/AccessibilityUIElementGtk.cpp:
1841         (WTR::AccessibilityUIElement::childrenCount): Avoid the infinite
1842         loop by using atk_object_get_n_accessible_children().
1843
1844 2012-09-13  Simon Hausmann  <simon.hausmann@nokia.com>
1845
1846         [Qt][Win] Enable USE(3D_GRAPHICS)
1847         https://bugs.webkit.org/show_bug.cgi?id=90879
1848
1849         Reviewed by Kenneth Rohde Christiansen.
1850
1851         Enable it on Windows as well as on QNX (bug fixed).
1852
1853         * qmake/mkspecs/features/features.prf:
1854
1855 2012-09-13  Simon Hausmann  <simon.hausmann@nokia.com>
1856
1857         [Qt] The ANGLE build on Windows breaks with GnuWin32's flex because it is too old
1858         https://bugs.webkit.org/show_bug.cgi?id=96359
1859
1860         Reviewed by Tor Arne Vestbø.
1861
1862         GnuWin32's flex is 2.5.4 and that is too old for ANGLE's preprocessor
1863         lexer. The only viable alternative I could find on Windows that provides
1864         binaries is the winflexbison distribution (http://sourceforge.net/projects/winflexbison/),
1865         so let's use that one instead.
1866
1867         * Scripts/webkitdirs.pm:
1868         (checkRequiredSystemConfig):
1869         * qmake/mkspecs/features/default_pre.prf:
1870
1871 2012-09-13  Simon Hausmann  <simon.hausmann@nokia.com>
1872
1873         [Qt] Add configuration for production builds
1874         https://bugs.webkit.org/show_bug.cgi?id=96607
1875
1876         Reviewed by Kenneth Rohde Christiansen.
1877
1878         Add configuration tweaks for production builds, to ease memory pressure with the GNU toolchain
1879         and enable all-in-one-files optimizations.
1880
1881         * qmake/mkspecs/features/production_build.prf: Added.
1882
1883 2012-09-13  Simon Hausmann  <simon.hausmann@nokia.com>
1884
1885         [Qt] GStreamer detection does not work when cross-compiling
1886         https://bugs.webkit.org/show_bug.cgi?id=96608
1887
1888         Reviewed by Kenneth Rohde Christiansen.
1889
1890         When cross-compiling it is not safe to just use system(pkg-config),
1891         because there is a high probability that it will find packages in the
1892         host system instead of in the target one. Instead use qmake's new
1893         packagesExist() function that takes care of this delicate aspect of
1894         cross compilation by setting PKG_CONFIG_SYSROOT_DIR and
1895         PKG_CONFIG_LIBDIR accordingly.
1896
1897         * qmake/mkspecs/features/features.prf:
1898
1899 2012-09-13  Christophe Dumez  <christophe.dumez@intel.com>
1900
1901         [WK2][WTR] More efficient string handling in InjectedBundlePage
1902         https://bugs.webkit.org/show_bug.cgi?id=96609
1903
1904         Reviewed by Kenneth Rohde Christiansen.
1905
1906         Update StringBuilder usage in InjectedBundlePage to follow
1907         the latest recommandations for efficient string handling:
1908         http://trac.webkit.org/wiki/EfficientStrings
1909
1910         This involves using appendLiteral() or appendNumber()
1911         instead of append() whenever possible.
1912
1913         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1914         (WTR::dumpPath):
1915         (WTR::rangeToStr):
1916         (WTR::styleDecToStr):
1917         (WTR::frameToStr):
1918         (WTR::dumpResourceURL):
1919         (WTR):
1920         (WTR::dumpFrameDescriptionSuitableForTestResult):
1921         (WTR::dumpRequestDescriptionSuitableForTestResult):
1922         (WTR::dumpResponseDescriptionSuitableForTestResult):
1923         (WTR::dumpErrorDescriptionSuitableForTestResult):
1924         (WTR::InjectedBundlePage::didReceiveIntentForFrame):
1925         (WTR::InjectedBundlePage::registerIntentServiceForFrame):
1926         (WTR::InjectedBundlePage::didStartProvisionalLoadForFrame):
1927         (WTR::InjectedBundlePage::didReceiveServerRedirectForProvisionalLoadForFrame):
1928         (WTR::InjectedBundlePage::didCommitLoadForFrame):
1929         (WTR::InjectedBundlePage::didFinishProgress):
1930         (WTR::dumpFrameScrollPosition):
1931         (WTR::dumpFrameText):
1932         (WTR::dumpDescendantFramesText):
1933         (WTR::InjectedBundlePage::didFinishLoadForFrame):
1934         (WTR::InjectedBundlePage::didReceiveTitleForFrame):
1935         (WTR::InjectedBundlePage::didCancelClientRedirectForFrame):
1936         (WTR::InjectedBundlePage::willPerformClientRedirectForFrame):
1937         (WTR::InjectedBundlePage::didFinishDocumentLoadForFrame):
1938         (WTR::InjectedBundlePage::didHandleOnloadEventsForFrame):
1939         (WTR::InjectedBundlePage::willSendRequestForFrame):
1940         (WTR::InjectedBundlePage::didReceiveResponseForResource):
1941         (WTR::InjectedBundlePage::didFinishLoadForResource):
1942         (WTR::InjectedBundlePage::didFailLoadForResource):
1943         (WTR::InjectedBundlePage::shouldCacheResponse):
1944         (WTR::InjectedBundlePage::decidePolicyForNavigationAction):
1945         (WTR::InjectedBundlePage::willAddMessageToConsole):
1946         (WTR::InjectedBundlePage::willSetStatusbarText):
1947         (WTR::InjectedBundlePage::willRunJavaScriptAlert):
1948         (WTR::InjectedBundlePage::willRunJavaScriptConfirm):
1949         (WTR::InjectedBundlePage::willRunJavaScriptPrompt):
1950         (WTR::InjectedBundlePage::didReachApplicationCacheOriginQuota):
1951         (WTR::InjectedBundlePage::shouldBeginEditing):
1952         (WTR::InjectedBundlePage::shouldEndEditing):
1953         (WTR::InjectedBundlePage::shouldInsertNode):
1954         (WTR::InjectedBundlePage::shouldInsertText):
1955         (WTR::InjectedBundlePage::shouldDeleteRange):
1956         (WTR::InjectedBundlePage::shouldChangeSelectedRange):
1957         (WTR::InjectedBundlePage::shouldApplyStyle):
1958         (WTR::InjectedBundlePage::didBeginEditing):
1959         (WTR::InjectedBundlePage::didEndEditing):
1960         (WTR::InjectedBundlePage::didChange):
1961         (WTR::InjectedBundlePage::didChangeSelection):
1962         (WTR::InjectedBundlePage::supportsFullScreen):
1963         (WTR::InjectedBundlePage::enterFullScreenForElement):
1964         (WTR::InjectedBundlePage::exitFullScreenForElement):
1965         (WTR::InjectedBundlePage::beganEnterFullScreen):
1966         (WTR::InjectedBundlePage::beganExitFullScreen):
1967         (WTR::InjectedBundlePage::closeFullScreen):
1968         (WTR::dumpBackForwardListItem):
1969         (WTR::InjectedBundlePage::dumpBackForwardList):
1970
1971 2012-09-13  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
1972
1973         [Qt][WK2] Memory leaks in Qt's TestRunner
1974         https://bugs.webkit.org/show_bug.cgi?id=96603
1975
1976         Reviewed by Csaba Osztrogonác.
1977
1978         Fix memory leaks in Qt's TestRunner code by adopting WKString
1979         created with WKStringCreateWithUTF8CString().
1980
1981         * WebKitTestRunner/qt/TestControllerQt.cpp:
1982         (WTR::TestController::initializeInjectedBundlePath):
1983         (WTR::TestController::initializeTestPluginDirectory):
1984
1985 2012-09-13  Zan Dobersek  <zandobersek@gmail.com>
1986
1987         [GTK][NRWT] Lower timeout value to 6 seconds (12 for debug configuration)
1988         https://bugs.webkit.org/show_bug.cgi?id=96506
1989
1990         Reviewed by Dirk Pranke.
1991
1992         Lower the timeout value to 6 seconds for release configuration and 12 seconds
1993         for debug configuration. These are the same values the Chromium port uses.
1994
1995         Currently the value is lowered only for the DumpRenderTree, WebKitTestRunner (and
1996         the layout tests under WebKit2 generally) requires more attention and gardening
1997         than what the current effort produces.
1998
1999         The unit test is updated as well to test these changes.
2000
2001         * Scripts/webkitpy/layout_tests/port/gtk.py:
2002         (GtkPort.default_timeout_ms):
2003         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
2004         (GtkPortTest.test_default_timeout_ms):
2005         (GtkPortTest.assertLinesEqual):
2006
2007 2012-09-12  Csaba Osztrogonác  <ossy@webkit.org>
2008
2009         Add SVN mirror handling feature to build.webkit.org
2010         https://bugs.webkit.org/show_bug.cgi?id=85887
2011
2012         Reviewed by Dirk Pranke.
2013
2014         * BuildSlaveSupport/build.webkit.org-config/config.json: Add SVN mirrors for Qt buildslaves hosted in Szeged.
2015         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2016         (ConfigureBuild.__init__): Add SVNMirror property to be able to watch which slaves use which mirror on build.webkit.org.
2017         (ConfigureBuild.start):
2018         (CheckOutSource.__init__): Set baseURL to the SVNMirror or the default http://svn.webkit.org/repository/webkit/
2019         (WaitForSVNServer): Add new buildstep for waiting the SVNMirror to be in sync with http://svn.webkit.org/repository/webkit/
2020         (Factory.__init__):
2021         (BuildFactory.__init__):
2022         (TestFactory.__init__):
2023         (BuildAndTestFactory.__init__):
2024         (BuildAndPerfTestFactory.__init__):
2025         (BuildAndPerfTestWebKit2Factory.__init__):
2026         (DownloadAndPerfTestFactory.__init__):
2027         (DownloadAndPerfTestWebKit2Factory.__init__):
2028         (loadBuilderConfig): Use kwargs instead of args to be able to add the optional SVNMirror factory argument.
2029         * BuildSlaveSupport/build.webkit.org-config/mastercfg_unittest.py:
2030         (SVNMirrorTest):
2031         (SVNMirrorTest.setUp): Load config.json once at startup.
2032         (SVNMirrorTest.get_SVNMirrorFromConfig): Get the SVNMirror from config.json for a given buildslave.
2033         (SVNMirrorTest.test_CheckOutSource): Compare CheckOutSource.baseURL with SVNMirror in config.json for all builders
2034         * BuildSlaveSupport/build.webkit.org-config/wait-for-SVN-server.py: Added.
2035         (getLatestSVNRevision): Get the latest SVN revison from the given server.
2036         (waitForSVNRevision): Wait until the given SVN revision is committed to the given server. It doesn't wait if the
2037                               SVN revision is empty (force build triggered without revision) or the server is unavailable.
2038
2039 2012-09-12  Kangil Han  <kangil.han@samsung.com>
2040
2041         [WK2][WTR] Set waitUntilDone watchdog timer value equal to WK1.
2042         https://bugs.webkit.org/show_bug.cgi?id=81606
2043
2044         Reviewed by Tim Horton.
2045
2046         6 seconds is harsh for some jquery test cases.
2047         Therefore, adjust the value to match DumpRenderTree.
2048
2049         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
2050         (WTR):
2051
2052 2012-09-12  KyungTae Kim  <ktf.kim@samsung.com>
2053
2054         [EFL] Support download attribute feature
2055         https://bugs.webkit.org/show_bug.cgi?id=96462
2056
2057         Reviewed by Gyuyoung Kim.
2058
2059         Enable DOWNLOAD_ATTRIBUTE feature for EFL port,
2060         and add "download,request" callback to the DumpRenderTree for EFL.
2061
2062         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
2063         (DumpRenderTreeChrome::createView):
2064         (DumpRenderTreeChrome::onDownloadRequest):
2065         * DumpRenderTree/efl/DumpRenderTreeChrome.h:
2066         (DumpRenderTreeChrome):
2067         * Scripts/webkitperl/FeatureList.pm:
2068
2069 2012-09-12  Stephanie Lewis  <slewis@apple.com>
2070
2071         Reduce parallism on the wk2 testers.
2072         https://bugs.webkit.org/show_bug.cgi?id=95906
2073
2074         Reviewed by Dirk Pranke.
2075
2076         The wk2 testers on Mountain Lion are getting stuck and timing out.  The problem appears to be due to
2077         resource contention.  Reducing the number of processes alleviates the issue.  
2078
2079         Starting by reducing 25%.
2080
2081         * Scripts/webkitpy/layout_tests/port/mac.py:
2082         (MacPort.default_child_processes):
2083
2084 2012-09-12  Brady Eidson  <beidson@apple.com>
2085
2086         Assert in NetscapePlugin::destroy() with async plugin init
2087         <rdar://problem/12277595> and https://bugs.webkit.org/show_bug.cgi?id=96576
2088
2089         Reviewed by Anders Carlsson.
2090
2091         Expose NPN_Invoke to plug-in tests:
2092         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
2093         (PluginTest::NPN_Invoke):
2094         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
2095         (PluginTest):
2096
2097         Add a test that uses NPN_Invoke on the window object from inside NPP_New to remove the plug-in element:
2098         * DumpRenderTree/TestNetscapePlugIn/Tests/InvokeDestroysPluginWithinNPP_New.cpp: Added.
2099         (InvokeDestroysPluginWithinNPP_New):
2100         (InvokeDestroysPluginWithinNPP_New::InvokeDestroysPluginWithinNPP_New):
2101         (InvokeDestroysPluginWithinNPP_New::NPP_New):
2102         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
2103
2104 2012-09-12  Dirk Pranke  <dpranke@chromium.org>
2105
2106         refactor TestExpectations tokenization slightly in preparation for the new syntax
2107         https://bugs.webkit.org/show_bug.cgi?id=96564
2108
2109         Reviewed by Ojan Vafai.
2110
2111         This change changes how we will tokenize/lex the
2112         TestExpectations files, in preparation for supporting both the
2113         old syntax and the new one. This change by itself should be a
2114         no-op.
2115
2116         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
2117         (TestExpectationParser._tokenize_line):
2118         (TestExpectationParser):
2119         (TestExpectationParser._tokenize_line_using_old_format):
2120         (TestExpectationParser._tokenize_line_using_new_format):
2121
2122 2012-09-12  Dirk Pranke  <dpranke@chromium.org>
2123
2124         remove unused testexpectations editing code from webkitpy, garden-o-matic
2125         https://bugs.webkit.org/show_bug.cgi?id=96562
2126
2127         Reviewed by Adam Barth.
2128
2129         This code was prototyped but never fully realized. Removing
2130         until it reanimates :).
2131
2132         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout.js:
2133         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout_unittests.js:
2134         * Scripts/webkitpy/layout_tests/controllers/test_expectations_editor.py: Removed.
2135         * Scripts/webkitpy/layout_tests/controllers/test_expectations_editor_unittest.py: Removed.
2136         * Scripts/webkitpy/tool/servers/gardeningserver.py:
2137         (BuildCoverageExtrapolator.extrapolate_test_configurations):
2138         (GardeningHTTPRequestHandler._run_webkit_patch):
2139         (GardeningHTTPRequestHandler.ping):
2140         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
2141         (BuildCoverageExtrapolatorTest.test_extrapolate):
2142         (GardeningServerTest.test_rebaseline_new_port):
2143
2144 2012-09-12  Lucas Forschler  <lforschler@apple.com>
2145
2146         Remove unused bot, apple-xserve-1.
2147         https://bugs.webkit.org/show_bug.cgi?id=96570
2148         
2149         Reviewed by Stephanie Lewis.
2150
2151         * BuildSlaveSupport/build.webkit.org-config/config.json:
2152
2153 2012-09-12  Xan Lopez  <xlopez@igalia.com>
2154
2155         [GTK] We attempt to rebase documentation even if it's not present
2156         https://bugs.webkit.org/show_bug.cgi?id=96553
2157
2158         Reviewed by Martin Robinson.
2159
2160         Do not make the documentation rebase step fatal. This allows make
2161         install to succeed when there's no documentation generated.
2162
2163         * gtk/generate-gtkdoc:
2164
2165 2012-09-12  Kenneth Rohde Christiansen  <kenneth@webkit.org>
2166
2167         Respect WEBKITOUTPUTDIR when running EFL tests
2168         https://bugs.webkit.org/show_bug.cgi?id=96528
2169
2170         Reviewed by Dirk Pranke.
2171
2172         Expose user set WEBKITOUTPUTDIR to the web process.
2173
2174         * Scripts/webkitpy/layout_tests/port/driver.py:
2175         (Driver._start): Add WEBKITOUTPUTDIR to the environment
2176         of the web process and its potential jhbuild wrapper.
2177
2178 2012-09-12  Dirk Pranke  <dpranke@chromium.org>
2179
2180         webkitdirs: fix uname version handling for cygwin
2181         https://bugs.webkit.org/show_bug.cgi?id=96436
2182
2183         Reviewed by Jon Honeycutt.
2184
2185         Newer versions of cygwin embed an additional version string
2186         inside parentheses, so you get "1.7.16(0.249/5/3)" instead of "1.7.16".
2187         Update the code to handle that.
2188
2189         * Scripts/webkitdirs.pm:
2190         (setupAppleWinEnv):
2191
2192 2012-09-12  Christophe Dumez  <christophe.dumez@intel.com>
2193
2194         Regression(r128338): Broke Windows build
2195         https://bugs.webkit.org/show_bug.cgi?id=96537
2196
2197         Unreviewed build fix.
2198
2199         Fix Apple-Win build by allocating array dynamically
2200         since its size is not constant.
2201
2202         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
2203         (WTR::stringArrayToJS):
2204
2205 2012-09-12  Alex Sakhartchouk  <alexst@chromium.org>
2206
2207         [chromium] Add a virtual test suite for fast/hidpi
2208         https://bugs.webkit.org/show_bug.cgi?id=90192
2209
2210         Reviewed by Dirk Pranke.
2211
2212         Add a virtual test suite to make sure the pixel tests in fast/hidpi give the same result
2213         on the hardware accelerated path as the software path.
2214
2215         * DumpRenderTree/chromium/WebPreferences.cpp:
2216         (WebPreferences::applyTo):
2217         * Scripts/webkitpy/layout_tests/port/chromium.py:
2218         (ChromiumPort.virtual_test_suites):
2219
2220 2012-09-12  Adam Klein  <adamk@chromium.org>
2221
2222         [chromium] Add content_browsertests to the flakiness dashboard
2223         https://bugs.webkit.org/show_bug.cgi?id=96535
2224
2225         Reviewed by Ojan Vafai.
2226
2227         * TestResultServer/static-dashboards/dashboard_base.js:
2228
2229 2012-09-12  Christophe Dumez  <christophe.dumez@intel.com>
2230
2231         [WK2][WKTR] TestRunner needs to implement originsWithApplicationCache
2232         https://bugs.webkit.org/show_bug.cgi?id=96496
2233
2234         Reviewed by Kenneth Rohde Christiansen.
2235
2236         Add implementation for originsWithApplicationCache to
2237         WebKitTestRunner.
2238
2239         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
2240         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
2241         (WTR::stringArrayToJS):
2242         (WTR):
2243         (WTR::TestRunner::originsWithApplicationCache):
2244         * WebKitTestRunner/InjectedBundle/TestRunner.h:
2245         (TestRunner):
2246
2247 2012-09-12  Simon Hausmann  <simon.hausmann@nokia.com>
2248
2249         Fix Qt/Windows build with Python3
2250         https://bugs.webkit.org/show_bug.cgi?id=96473
2251
2252         Reviewed by Csaba Osztrogonác.
2253
2254         In Python 3 print is a real function, so we must use parentheses around
2255         the function parameters. This is backwards compatible with Python 2.
2256
2257         * Scripts/generate-win32-export-forwards:
2258
2259 2012-09-12  Simon Hausmann  <simon.hausmann@nokia.com>
2260
2261         Unreviewed trivial build fix for Qt/Windows after bug #96358.
2262
2263         As pointed out by Kevin, we should inject the "set PATH=..." statement
2264         to add the GnuWin32 directory only if that directory actually exists.
2265
2266         * qmake/mkspecs/features/default_post.prf:
2267
2268 2012-09-12  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
2269
2270         [EFL] [WK2] Memory leaks in TestControllerEfl
2271         https://bugs.webkit.org/show_bug.cgi?id=96525
2272
2273         Reviewed by Kenneth Rohde Christiansen.
2274
2275         Fix memory leaks in EFL's TestRunner code by adopting an allocation
2276         of WKString created with WKStringCreateWithUTF8CString().
2277
2278         * WebKitTestRunner/efl/TestControllerEfl.cpp:
2279         (WTR::TestController::initializeInjectedBundlePath):
2280         (WTR::TestController::initializeTestPluginDirectory):
2281
2282 2012-09-12  Christophe Dumez  <christophe.dumez@intel.com>
2283
2284         [WK2][WKTR] TestRunner needs to implement dumpApplicationCacheDelegateCallbacks
2285         https://bugs.webkit.org/show_bug.cgi?id=96374
2286
2287         Reviewed by Kenneth Rohde Christiansen.
2288
2289         Implement support for dumpApplicationCacheDelegateCallbacks
2290         and disallowIncreaseForApplicationCacheQuota in
2291         WebKitTestRunner and properly dump the information
2292         expected by the tests.
2293
2294         If the application cache quota is reached for a given
2295         security origin, WebKitTestRunner will reset the quota
2296         to its default value, unless intructed not to via
2297         disallowIncreaseForApplicationCacheQuota().
2298
2299         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
2300         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
2301         (WTR::InjectedBundlePage::InjectedBundlePage):
2302         (WTR::InjectedBundlePage::didReachApplicationCacheOriginQuota):
2303         (WTR):
2304         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
2305         (InjectedBundlePage):
2306         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
2307         (WTR::TestRunner::TestRunner):
2308         (WTR::TestRunner::disallowIncreaseForApplicationCacheQuota):
2309         (WTR):
2310         * WebKitTestRunner/InjectedBundle/TestRunner.h:
2311         (WTR::TestRunner::dumpApplicationCacheDelegateCallbacks):
2312         (TestRunner):
2313         (WTR::TestRunner::shouldDisallowIncreaseForApplicationCacheQuota):
2314         (WTR::TestRunner::shouldDumpApplicationCacheDelegateCallbacks):
2315
2316 2012-09-12  Jochen Eisinger  <jochen@chromium.org>
2317
2318         [chromium] remove deprecated and unused sets import from chromium_android driver
2319         https://bugs.webkit.org/show_bug.cgi?id=96485
2320
2321         Reviewed by Dirk Pranke.
2322
2323         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2324
2325 2012-09-12  Zan Dobersek  <zandobersek@gmail.com>
2326
2327         Flakiness dashboard doesn't recognize new Chromium Android test builder
2328         https://bugs.webkit.org/show_bug.cgi?id=96523
2329
2330         Reviewed by Ojan Vafai.
2331
2332         Properly return 'ANDROID' as the Chromium platform for Android builders.
2333
2334         * TestResultServer/static-dashboards/flakiness_dashboard.js:
2335         (chromiumPlatform):
2336         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
2337         (test):
2338
2339 2012-09-12  Simon Hausmann  <simon.hausmann@nokia.com>
2340
2341         [Qt] Build on Windows requires bison/flex in PATH
2342         https://bugs.webkit.org/show_bug.cgi?id=96358
2343
2344         Reviewed by Tor Arne Vestbø.
2345
2346         The build requires flex, bison, etc. and they need to be in the PATH when building. On Mac OS X
2347         and Linux that is rarely a problem given how easily available the tools are. On Windows however
2348         a separate installation of various GNU tools is required as the operating system doesn't come with
2349         them. To make the development more convenient, Qt 5 provides a copy of the most essential tools in
2350         the gnuwin32 directory of the qt5.git top-level repository.
2351
2352         This patch tries to detect the presence of those tools and prepends them to the PATH if found.
2353
2354         This is required in preparation for the elimination of qt5/qtwebkit.pri, which currently expands
2355         PATH before calling build-webkit. It it also required for the upcoming introduction of win_flex
2356         as dependency over flex, which can be done with less hassle when qt5's gnuwin32 directory has been
2357         updated with the new tool.
2358
2359         * Scripts/webkitdirs.pm:
2360         (checkRequiredSystemConfig):
2361         * qmake/mkspecs/features/default_post.prf:
2362
2363 2012-09-12  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2364
2365         [Qt] Update build-jsc after r128174
2366
2367         Reviewed by Ossy.
2368
2369         * Scripts/build-jsc:
2370
2371 2012-09-12  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2372
2373         [Qt] Teach addStrictSubdirOrderBetween to handle more than two targets
2374
2375         By hard-coding the names of the targets we defined we ended up just
2376         redefining the previous target when using addStrictSubdirOrderBetween
2377         more than once in a single project file.
2378
2379         We now embed the two base targets into the target names.
2380
2381         Reviewed by Simon Hausmann.
2382
2383         * qmake/mkspecs/features/functions.prf:
2384
2385 2012-09-12  Philippe Liard  <pliard@google.com>
2386
2387         Depend on {base,net} GYP targets rather than {base,net}_java.
2388         https://bugs.webkit.org/show_bug.cgi?id=95690
2389
2390         Reviewed by Adam Barth.
2391
2392         {base,net}_java became 'private' targets on the Chromium side that
2393         clients should not depend on (see
2394         https://chromiumcodereview.appspot.com/10913083/).
2395         This removes a bunch of trailing white spaces additionally in
2396         TestWebkitAPI.gyp (which includes non-trailing white space changes).
2397
2398         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
2399         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
2400
2401 2012-09-12  Simon Hausmann  <simon.hausmann@nokia.com>
2402
2403         [Qt] Fix the build with ENABLE_NETSCAPE_PLUGIN_API=0
2404         https://bugs.webkit.org/show_bug.cgi?id=96494
2405
2406         Reviewed by Tor Arne Vestbø.
2407
2408         WK2's ENABLE_PLUGIN_PROCESS uses the NPAPI functions unconditionally, so disable the
2409         plugin process feature if we don't have NPAPI.
2410
2411         * qmake/mkspecs/features/features.prf:
2412
2413 2012-09-12  Sheriff Bot  <webkit.review.bot@gmail.com>
2414
2415         Unreviewed, rolling out r128280.
2416         http://trac.webkit.org/changeset/128280
2417         https://bugs.webkit.org/show_bug.cgi?id=96498
2418
2419         it broke compilation on windows debug bot (Requested by loislo
2420         on #webkit).
2421
2422         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
2423         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
2424         (WTR::InjectedBundlePage::InjectedBundlePage):
2425         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
2426         (InjectedBundlePage):
2427         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
2428         (WTR::TestRunner::TestRunner):
2429         * WebKitTestRunner/InjectedBundle/TestRunner.h:
2430         (TestRunner):
2431
2432 2012-09-12  Simon Hausmann  <simon.hausmann@nokia.com>
2433
2434         [Qt] Make it possible to build with "make release" and "make debug" on Windows
2435         https://bugs.webkit.org/show_bug.cgi?id=96488
2436
2437         Reviewed by Tor Arne Vestbø.
2438
2439         A "make debug" is passed through recursively and currently it aborts at Makefile.DerivedSources
2440         because there are no such targets. We want the generated sources to be independent from release
2441         or debug build configurations, so it is sufficient to provide fake debug and release targets that
2442         redirect to the same general-purpose target (first) of creating the derived sources.
2443
2444         * qmake/mkspecs/features/default_post.prf:
2445
2446 2012-09-12  Kenneth Rohde Christiansen  <kenneth@webkit.org>
2447
2448         [EFL] Make DumpRenderTree smarter at finding the fonts
2449         http://webkit.org/b/96281
2450
2451         Reviewed by Gyuyoung Kim.
2452
2453         Respect WEBKITOUTPUTDIR and expand the font dir from it.
2454         Use CString consistently.
2455
2456         * DumpRenderTree/efl/FontManagement.cpp:
2457         (buildPath):
2458         (getCoreFontFiles):
2459         (addFontDirectory):
2460         (addFontFiles):
2461         (getCustomBuildDir):
2462         (getPlatformFontsPath):
2463         (addFontsToEnvironment):
2464
2465 2012-09-12  Christophe Dumez  <christophe.dumez@intel.com>
2466
2467         [WK2][WKTR] TestRunner needs to implement dumpApplicationCacheDelegateCallbacks
2468         https://bugs.webkit.org/show_bug.cgi?id=96374
2469
2470         Reviewed by Kenneth Rohde Christiansen.
2471
2472         Implement support for dumpApplicationCacheDelegateCallbacks
2473         and disallowIncreaseForApplicationCacheQuota in
2474         WebKitTestRunner and properly dump the information
2475         expected by the tests.
2476
2477         If the application cache quota is reached for a given
2478         security origin, WebKitTestRunner will reset the quota
2479         to its default value, unless intructed not to via
2480         disallowIncreaseForApplicationCacheQuota().
2481
2482         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
2483         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
2484         (WTR::InjectedBundlePage::InjectedBundlePage):
2485         (WTR::InjectedBundlePage::didReachApplicationCacheOriginQuota):
2486         (WTR):
2487         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
2488         (InjectedBundlePage):
2489         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
2490         (WTR::TestRunner::TestRunner):
2491         (WTR::TestRunner::disallowIncreaseForApplicationCacheQuota):
2492         (WTR):
2493         * WebKitTestRunner/InjectedBundle/TestRunner.h:
2494         (WTR::TestRunner::dumpApplicationCacheDelegateCallbacks):
2495         (TestRunner):
2496         (WTR::TestRunner::shouldDisallowIncreaseForApplicationCacheQuota):
2497         (WTR::TestRunner::shouldDumpApplicationCacheDelegateCallbacks):
2498
2499 2012-09-12  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
2500
2501         [WK2] [WTR] WebKitTestRunner needs TestRunner.workerThreadCount
2502         https://bugs.webkit.org/show_bug.cgi?id=96388
2503
2504         Reviewed by Kenneth Rohde Christiansen.
2505
2506         Exported TestRunner.workerThreadCount as readonly attribute.
2507
2508         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
2509         * WebKitTestRunner/InjectedBundle/TestRunner.cpp: Added workerThreadCount() method implementation.
2510         (WTR::TestRunner::workerThreadCount): Returns count of worker threads.
2511         (WTR):
2512         * WebKitTestRunner/InjectedBundle/TestRunner.h: Added workerThreadCount() method. 
2513         (TestRunner):
2514
2515 2012-09-12  Jochen Eisinger  <jochen@chromium.org>
2516
2517         [chromium] Consume a user gesture when creating a new view.
2518         https://bugs.webkit.org/show_bug.cgi?id=96373
2519
2520         Reviewed by Adam Barth.
2521
2522         * DumpRenderTree/chromium/WebViewHost.cpp:
2523         (WebViewHost::createView):
2524
2525 2012-09-12  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
2526
2527         [WK2][WTR] Some of TestRunner special options are not reset before testing
2528         https://bugs.webkit.org/show_bug.cgi?id=96384
2529
2530         Reviewed by Kenneth Rohde Christiansen.
2531
2532         Now values of the following special options are reset:
2533         void setAcceptsEditing(in boolean value);
2534         void setCloseRemainingWindowsWhenComplete(in boolean value);
2535         void setXSSAuditorEnabled(in boolean value);
2536         void setAllowFileAccessFromFileURLs(in boolean value);
2537         void setPluginsEnabled(in boolean value);
2538         void setPopupBlockingEnabled(in boolean value);
2539
2540         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
2541         (WTR::InjectedBundle::beginTesting):
2542
2543 2012-09-11  Ryuan Choi  <ryuan.choi@samsung.com>
2544
2545         [CMAKE] Supply feature defines to CodeGeneratorTestRunner.
2546         https://bugs.webkit.org/show_bug.cgi?id=96273
2547
2548         Reviewed by Gyuyoung Kim.
2549
2550         Passed FEATURE_DEFINES_WITH_SPACE_SEPARATOR to CodeGeneratorTestRunner
2551         in order to distinguish compile time options such as ENABLE(TOUCH_EVENTS).
2552
2553         * WebKitTestRunner/CMakeLists.txt:
2554
2555 2012-09-11  Dirk Pranke  <dpranke@chromium.org>
2556
2557         [chromium] rename webkit_resources
2558         https://bugs.webkit.org/show_bug.cgi?id=95874
2559
2560         Reviewed by James Robinson.
2561
2562         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
2563
2564 2012-09-11  Jochen Eisinger  <jochen@chromium.org>
2565
2566         [chromium] don't include content shell builders in the garden-o-matic view
2567         https://bugs.webkit.org/show_bug.cgi?id=96418
2568
2569         Reviewed by James Robinson.
2570
2571         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
2572
2573 2012-09-11  Adam Klein  <adamk@chromium.org>
2574
2575         Garden-o-matic should ignore a wider variety of warnings in buildbot json
2576         https://bugs.webkit.org/show_bug.cgi?id=96411
2577
2578         Reviewed by Adam Barth.
2579
2580         Previously only the exact string "warning" was treated as a warning
2581         result. This patch treats any string with "warning" as a substring
2582         as a warning (e.g., "warnings", as seen on the cr-win buildbots).
2583
2584         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders.js:
2585         (.):
2586         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders_unittests.js:
2587
2588 2012-09-11  Marcelo Lira  <marcelo.lira@openbossa.org>
2589
2590         Restore original value of mock scrollbars enabled in InternalSettings
2591         https://bugs.webkit.org/show_bug.cgi?id=87680
2592
2593         Reviewed by Adam Barth.
2594
2595         Added backup support for mock scrollbars state in
2596         InternalSettings::Backup, and removed support for these
2597         from Qt's DumpRenderTree.
2598
2599         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
2600         (WebCore::WebPage::resetSettings):
2601         * DumpRenderTree/qt/TestRunnerQt.cpp:
2602         (TestRunner::setFrameFlatteningEnabled):
2603         * DumpRenderTree/qt/TestRunnerQt.h:
2604         (TestRunner):
2605
2606 2012-09-11  Dan Bernstein  <mitz@apple.com>
2607
2608         Fixed a bug where kill-old-processes failed to kill clang++.
2609
2610         Rubber-stamped by Alexey Proskuryakov.
2611
2612         * BuildSlaveSupport/kill-old-processes: Escape the + signs in "clang++"
2613         where it is passed as a regular expression to killall.
2614
2615 2012-09-11  Philippe Liard  <pliard@google.com>
2616
2617         Make sure that md5sum is not setup on non-Chromium Android platforms.
2618         https://bugs.webkit.org/show_bug.cgi?id=96393
2619
2620         Reviewed by Adam Barth.
2621
2622         Previously md5sum was setup in ChromiumAndroidDriver's constructor
2623         which is also invoked on non-Chromium Android platforms. The following patch
2624         moves this logic to _startup_test() which is guaranteed to be invoked
2625         only on Android.
2626
2627         This also updates the prompt unit test in chromium_android_unittest.py.
2628
2629         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2630         (ChromiumAndroidDriver.__init__):
2631         (ChromiumAndroidDriver._setup_md5sum_and_push_data_if_needed):
2632         (ChromiumAndroidDriver._setup_test):
2633
2634 2012-09-11  Raphael Kubo da Costa  <rakuco@webkit.org>
2635
2636         [EFL] Rewrite the EFL-related Find modules
2637         https://bugs.webkit.org/show_bug.cgi?id=95237
2638
2639         Reviewed by Kenneth Rohde Christiansen.
2640
2641         * DumpRenderTree/efl/CMakeLists.txt: Stop setting the LINK_FLAGS
2642         property and include all the Enlightenment Foundation Libraries
2643         required by the target.
2644         * EWebLauncher/CMakeLists.txt: Ditto.
2645         * MiniBrowser/efl/CMakeLists.txt: Ditto.
2646         * TestWebKitAPI/PlatformEfl.cmake: Add missing include directories
2647         now that they are not added implicitly.
2648         * WebKitTestRunner/CMakeLists.txt: Stop setting the LINK_FLAGS
2649         property.
2650         * WebKitTestRunner/PlatformEfl.cmake: Stop setting the LINK_FLAGS
2651         property and include all the Enlightenment Foundation Libraries
2652         required by the target.
2653
2654 2012-09-11  Christophe Dumez  <christophe.dumez@intel.com>
2655
2656         [WK2][WKTR] TestRunner needs to implement setApplicationCacheOriginQuota
2657         https://bugs.webkit.org/show_bug.cgi?id=96379
2658
2659         Reviewed by Kenneth Rohde Christiansen.
2660
2661         Add implementation for setApplicationCacheOriginQuota
2662         to WebKitTestRunner.
2663
2664         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
2665         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
2666         (WTR::TestRunner::setApplicationCacheOriginQuota):
2667         (WTR):
2668         * WebKitTestRunner/InjectedBundle/TestRunner.h:
2669         (TestRunner):
2670
2671 2012-09-11  Luciano Wolf  <luciano.wolf@openbossa.org>
2672
2673         [Qt] [WK2] editing/inserting/typing-tab-designmode tests are failing
2674         https://bugs.webkit.org/show_bug.cgi?id=91988
2675
2676         Reviewed by Kenneth Rohde Christiansen.
2677
2678         Regular key events generated by Qt for the tab key fill both
2679         keyCode and keyText, so we should do the same for our fake events.
2680
2681         * WebKitTestRunner/qt/EventSenderProxyQt.cpp:
2682         (WTR::EventSenderProxy::keyDown):
2683
2684 2012-09-11  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
2685
2686         [WK2][WTR] WebKitTestRunner needs testRunner.setSpatialNavigationEnabled
2687         https://bugs.webkit.org/show_bug.cgi?id=96269
2688
2689         Reviewed by Kenneth Rohde Christiansen.
2690
2691         Exported missing testRunner.setSpatialNavigationEnabled() method.
2692
2693         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
2694         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
2695         (WTR::InjectedBundle::beginTesting):
2696         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
2697         (WTR::TestRunner::setSpatialNavigationEnabled):
2698         (WTR):
2699         * WebKitTestRunner/InjectedBundle/TestRunner.h:
2700         (TestRunner):
2701
2702 2012-09-11  Philippe Liard  <pliard@chromium.org>
2703
2704         Support LayoutTests on non-rooted devices for Chromium Android.
2705         https://bugs.webkit.org/show_bug.cgi?id=95346
2706
2707         Reviewed by Dirk Pranke.
2708
2709         We were requiring root permissions to sync resource files with the
2710         device by comparing host and device files' timestamp after we
2711         synchronized the device and host clocks.
2712         We now rely on md5sum provided by Chromium to handle data files
2713         synchronization.
2714
2715         * Scripts/webkitpy/layout_tests/port/chromium.py:
2716         (ChromiumPort._check_driver_build_up_to_date):
2717         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2718         (ChromiumAndroidDriver.__init__):
2719         (ChromiumAndroidDriver._setup_test):
2720         (ChromiumAndroidDriver._push_executable):
2721         (ChromiumAndroidDriver._push_test_resources):
2722         (ChromiumAndroidDriver._read_prompt):
2723
2724 2012-09-11  Simon Hausmann  <simon.hausmann@nokia.com>
2725
2726         [Qt] Fix output paths on Windows
2727
2728         Reviewed by Tor Arne Vestbø.
2729
2730         BUILD_ROOT_RELATIVE_TO_OUT_PWD was calculated incorrectly Windows, resulting in an incorrect
2731         DESTDIR transformation at the end of default_post.prf. The calculation should be baased no qmake
2732         paths, which can use drive letters but always use forward slashes. So the use of QMAKE_DIR_SEP is
2733         incorrect here and instead for the calculation we should use forward slashes.
2734
2735         * qmake/mkspecs/features/default_pre.prf:
2736
2737 2012-09-11  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2738
2739         [Qt] Fix passing of defines from the build system
2740
2741         We explicitly have to pass -DENABLE_FOO=0 for every single feature
2742         that's not explicitly enabled in WEBKIT_CONFIG, since Platform.h
2743         will add it's own defaults if the features are not defined.
2744
2745         At some point we might want to run a configure-test to inspect what
2746         Platform.h will do, so we can sync up the build system's view of
2747         the feature-defines with the compile-time situation.
2748
2749         Reviewed by Simon Hausmann.
2750
2751         * qmake/mkspecs/features/functions.prf:
2752
2753 2012-09-11  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2754
2755         [Qt] Fix --qt option to build-webkit
2756
2757         Reviewed by Ossy.
2758
2759         * Scripts/webkitdirs.pm:
2760         (buildQMakeProjects):
2761
2762 2012-09-11  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2763
2764         [Qt] Add a configure step to the Qt build system
2765
2766         This allows building the Qt port using just 'qmake WebKit.pro'. Using
2767         the build-webkit script is still supported, and will add slightly more
2768         logic to the build, such as detecting the need for clean or incremental
2769         builds.
2770
2771         Internally, the build system now uses a new variable, WEBKIT_CONFIG, for
2772         all things related to configuring the build, such as use/have/enable flags,
2773         and these are translated to defines in default_post. Project files should
2774         no longer check contains(DEFINES,...) to detect features, but use the new
2775         enable?(), use?(), and have?() functions. The no_webkit1 and no_webkit2
2776         options have been translated into WEBKIT_CONFIG options as well, and can
2777         be checked using build?().
2778
2779         Reviewed by Simon Hausmann.
2780
2781         * DumpRenderTree/qt/DumpRenderTree.pro:
2782         * MiniBrowser/qt/MiniBrowser.pro:
2783         * QtTestBrowser/QtTestBrowser.pro:
2784         * Scripts/build-webkit:
2785         * Scripts/webkitdirs.pm:
2786         (qtFeatureDefaults):
2787         (buildQMakeProjects):
2788         * Tools.pro:
2789         * WebKitTestRunner/InjectedBundle/DerivedSources.pri:
2790         * WebKitTestRunner/InjectedBundle/Target.pri:
2791         * WebKitTestRunner/Target.pri:
2792         * qmake/.qmake.cache: Added.
2793         * qmake/configure.pri: Removed.
2794         * qmake/configure.pro: Removed.
2795         * qmake/dump-features: Added.
2796         * qmake/mkspecs/features/configure.prf: Added.
2797         * qmake/mkspecs/features/default_post.prf:
2798         * qmake/mkspecs/features/default_pre.prf:
2799         * qmake/mkspecs/features/features.prf:
2800         * qmake/mkspecs/features/features.pri:
2801         * qmake/mkspecs/features/functions.prf:
2802         * qmake/mkspecs/features/gprof.prf:
2803         * qmake/mkspecs/features/mac/default_pre.prf:
2804         * qmake/mkspecs/features/win32/default_pre.prf:
2805
2806 2012-09-11  Christophe Dumez  <christophe.dumez@intel.com>
2807
2808         [WK2][WKTR] TestRunner needs to implement clearApplicationCacheForOrigin
2809         https://bugs.webkit.org/show_bug.cgi?id=96372
2810
2811         Reviewed by Kenneth Rohde Christiansen.
2812
2813         Implement clearApplicationCacheForOrigin() in WebKitTestRunner.
2814
2815         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
2816         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
2817         (WTR::TestRunner::clearApplicationCacheForOrigin):
2818         (WTR):
2819         * WebKitTestRunner/InjectedBundle/TestRunner.h:
2820         (TestRunner):
2821
2822 2012-09-11  Peter Beverloo  <peter@chromium.org>
2823
2824         Don't ignore .pak files when making an archive for Android
2825         https://bugs.webkit.org/show_bug.cgi?id=96375
2826
2827         Reviewed by Jochen Eisinger.
2828
2829         The .pak files do need to be pushed to the device, which is currently
2830         preventing the layout test runner from correctly pushing all files.
2831
2832         * BuildSlaveSupport/built-product-archive:
2833         (archiveBuiltProduct):
2834
2835 2012-09-11  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
2836
2837         [WK2][WTR] WebKitTestRunner needs layoutTestController.setMinimumTimerInterval
2838         https://bugs.webkit.org/show_bug.cgi?id=96256
2839
2840         Reviewed by Kenneth Rohde Christiansen.
2841
2842         Exported TestRunner::setMinimumTimerInterval() method.
2843
2844         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
2845         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
2846         (WTR::InjectedBundle::beginTesting):
2847         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
2848         (WTR::TestRunner::setMinimumTimerInterval):
2849         (WTR):
2850         * WebKitTestRunner/InjectedBundle/TestRunner.h:
2851         (TestRunner):
2852
2853 2012-09-11  Tommy Widenflycht  <tommyw@google.com>
2854
2855         MediaStream API: add RTCPeerConnection::onnegotiationneeded
2856         https://bugs.webkit.org/show_bug.cgi?id=96097
2857
2858         Reviewed by Adam Barth.
2859
2860         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
2861         (MockWebRTCPeerConnectionHandler::addStream):
2862         (MockWebRTCPeerConnectionHandler::removeStream):
2863
2864 2012-09-11  Peter Beverloo  <peter@chromium.org>
2865
2866         Recognize the "--chromium-android" argument in run-webkit-tests
2867         https://bugs.webkit.org/show_bug.cgi?id=96369
2868
2869         Reviewed by Jochen Eisinger.
2870
2871         Chromium for the Android platform cannot determine the platform based on
2872         the host itself, so we'll use the --chromium-android argument as the build
2873         master passes to run-webkit-tests. Make sure the layout test runner
2874         can pick the right port when doing so.
2875
2876         * Scripts/run-webkit-tests:
2877         * Scripts/webkitpy/layout_tests/port/factory.py:
2878         (port_options):
2879
2880 2012-09-11  Simon Hausmann  <simon.hausmann@nokia.com>
2881
2882         [QT] Fix build with newer qmake from Qt 5
2883
2884         Rubber-stamped by Tor Arne Vestbø.
2885
2886         Support for "qmake -query QMAKE_MKSPECS" was removed from qmake. Replace
2887         it with the use of QT_HOST_DATA/mkspecs.
2888         Based on patch by Oswald Buddenhagen <oswald.buddenhagen@nokia.com>.
2889
2890         * Scripts/webkitdirs.pm:
2891         (builtDylibPathForName):
2892         (qtMakeCommand):
2893
2894 2012-09-11  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2895
2896         [Qt] Remove Qt 4 syncqt, no longer needed
2897
2898         Reviewed by Simon Hausmann.
2899
2900         * qmake/syncqt-4.8: Removed.
2901
2902 2012-09-11  Andras Becsi  <andras.becsi@nokia.com>
2903
2904         [Qt][WK2] Fix the build with recent Qt5
2905         https://bugs.webkit.org/show_bug.cgi?id=96283
2906
2907         Reviewed by Simon Hausmann.
2908
2909         We should not use deprecated Qt API because doing that results
2910         in build failure since all the deprecated API was disabled
2911         in qtbase (https://codereview.qt-project.org/#change,24890).
2912
2913         * DumpRenderTree/qt/main.cpp:
2914         (main):
2915         * MiniBrowser/qt/MiniBrowserApplication.cpp:
2916         (MiniBrowserApplication::notify):
2917         * MiniBrowser/qt/raw/View.cpp:
2918         (View::exposeEvent):
2919         (main):
2920
2921 2012-09-11  Christophe Dumez  <christophe.dumez@intel.com>
2922
2923         WTR needs an implementation of applicationCacheDiskUsageForOrigin
2924         https://bugs.webkit.org/show_bug.cgi?id=87188
2925
2926         Reviewed by Kenneth Rohde Christiansen.
2927
2928         Add implementation for applicationCacheDiskUsageForOrigin()
2929         in WebKitTestRunner.
2930
2931         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
2932         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
2933         (WTR::TestRunner::applicationCacheDiskUsageForOrigin):
2934         (WTR):
2935         * WebKitTestRunner/InjectedBundle/TestRunner.h:
2936         (TestRunner):
2937
2938 2012-09-10  Joanmarie Diggs  <jdiggs@igalia.com>
2939
2940         Properly expose <legend> elements to ATs
2941         https://bugs.webkit.org/show_bug.cgi?id=84137
2942
2943         Reviewed by Chris Fleizach.
2944
2945         Created a new WebCore Accessibility Role, LegendRole. Used it to map to
2946         the expected platform role, ATK_ROLE_LABEL. Also established the needed
2947         AtkRelation pair, label-for/labelled-by between the legend and fieldset.
2948
2949         * DumpRenderTree/chromium/TestRunner/AccessibilityUIElementChromium.cpp:
2950         (roleToString): added "Legend" string for WebAccessibilityRoleLegend
2951
2952 2012-09-10  Jer Noble  <jer.noble@apple.com>
2953
2954         Unreviewed; rolling out r128081.
2955         http://trac.webkit.org/changeset/128081
2956         https://bugs.webkit.org/show_bug.cgi?id=96134
2957
2958         New test http/tests/media/video-donottrack.html fails to run on Apple test bots.
2959
2960         * DumpRenderTree/TestRunner.cpp:
2961         (TestRunner::TestRunner):
2962         (TestRunner::staticFunctions):
2963         * DumpRenderTree/TestRunner.h:
2964         (TestRunner):
2965         * DumpRenderTree/mac/FrameLoadDelegate.mm:
2966         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
2967         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
2968         (WTR::InjectedBundlePage::InjectedBundlePage):
2969         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
2970         (InjectedBundlePage):
2971         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
2972         (WTR::TestRunner::TestRunner):
2973         * WebKitTestRunner/InjectedBundle/TestRunner.h:
2974         (TestRunner):
2975
2976 2012-09-10  Christophe Dumez  <christophe.dumez@intel.com>
2977
2978         TestExpectationsChecker._determine_port_from_expectations_path() does not support cascaded TestExpectations
2979         https://bugs.webkit.org/show_bug.cgi?id=96205
2980
2981         Reviewed by Tony Chang.
2982
2983         _determine_port_from_expectations_path() was calling port.path_to_test_expectations_file()
2984         internally, which means that it would support only 1 TestExpectations file per port. This
2985         is an issue for ports such as EFL that support cascased TestExpectations (efl-wk2 -> efl
2986         and efl-wk1 -> efl).
2987
2988         This patch makes _determine_port_from_expectations_path() call port.expectations_files()
2989         instead so that all the ports TestExpectations are recognized. The ports are also
2990         constructed twice, with "webkit_test_runner" option set to True and False so that we
2991         retrieve the TestExpectations paths for both WebKit1 and WebKit2.
2992
2993         * Scripts/webkitpy/style/checkers/test_expectations.py:
2994         (TestExpectationsChecker._determine_port_from_expectations_path):
2995         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py: Add corresponding unit test.
2996         (TestExpectationsTestCase.test_determine_port_from_expectations_path):
2997
2998 2012-09-10  Ryuan Choi  <ryuan.choi@samsung.com>
2999
3000         [WTR] Generated source files should include config.h
3001         https://bugs.webkit.org/show_bug.cgi?id=96254
3002
3003         Reviewed by Benjamin Poulain.
3004
3005         Current coding style has encouraged to include config.h at the first position,
3006         But derived sources, generated by CodeGeneratorTestRunner.pm, don't follow
3007         this rule.
3008         Therefore, this patch added it.
3009
3010         * WebKitTestRunner/InjectedBundle/Bindings/CodeGeneratorTestRunner.pm:
3011         (_generateImplementationFile):
3012
3013 2012-09-10  Rick Byers  <rbyers@chromium.org>
3014
3015         Add handling of new GestureTapCancel in DRT
3016         
3017         https://bugs.webkit.org/show_bug.cgi?id=96183
3018
3019         Reviewed by Adam Barth.
3020
3021         * DumpRenderTree/chromium/TestWebPlugin.cpp:
3022         (TestWebPlugin::handleInputEvent):
3023
3024 2012-09-10  Brady Eidson  <beidson@apple.com>
3025
3026         Javascript in foreground tabs should not wait synchronously for plug-ins to load
3027         <rdar://problem/12067415> and https://bugs.webkit.org/show_bug.cgi?id=96167
3028
3029         Reviewed by Geoff Garen.
3030
3031         Enhance the "Slow NPP_New" plug-in to also be able to return properties to javascript.
3032
3033         * DumpRenderTree/TestNetscapePlugIn/Tests/SlowNPPNew.cpp:
3034         (PluginObject):
3035         (SlowNPPNew::PluginObject::PluginObject):
3036         (SlowNPPNew::PluginObject::~PluginObject):
3037         (SlowNPPNew::PluginObject::hasProperty):
3038         (SlowNPPNew::PluginObject::getProperty):
3039         (SlowNPPNew::NPP_GetValue):
3040         (SlowNPPNew):
3041
3042 2012-09-10  Tim Horton  <timothy_horton@apple.com>
3043
3044         WKTR often reports an unresponsive WebProcess on Mac bots
3045         https://bugs.webkit.org/show_bug.cgi?id=95906
3046
3047         Reviewed by Dan Bernstein.
3048
3049         Attempt to work around resource contention issues by increasing the NRWT
3050         timeout when running against WKTR, as well as WKTR's own WebProcess timeout.
3051
3052         * Scripts/webkitpy/layout_tests/port/base.py:
3053         (Port.default_timeout_ms): Default timeout when using WKTR -> 80 seconds.
3054         * WebKitTestRunner/TestController.cpp:
3055         (WTR): LongTimeout (used when waiting for WebProcess to finish) -> 60 seconds.
3056
3057 2012-09-07  Jer Noble  <jer.noble@apple.com>
3058
3059         <audio> and <video> should send Do Not Track when appropriate
3060         https://bugs.webkit.org/show_bug.cgi?id=96134
3061
3062         Reviewed by Eric Carlson.
3063
3064         Add a testRunner JS API for calling setShouldSendDoNotTrackHTTPHeader to DRT and 
3065         WTR. Add FrameLoaderClient client methods for retrieving that value.
3066
3067         * DumpRenderTree/TestRunner.cpp:
3068         (setShouldSendDoNotTrackHTTPHeaderCallback): Pass through to the TestRunner.
3069         (TestRunner::staticFunctions): Add the setShouldSendDoNotTrackHTTPHeader function.
3070         * DumpRenderTree/TestRunner.h:
3071         (TestRunner::setShouldSendDoNotTrackHTTPHeader): Simple setter.
3072         (TestRunner::shouldSendDoNotTrackHTTPHeader): Simple getter.
3073         * DumpRenderTree/mac/FrameLoadDelegate.mm:
3074         (-[FrameLoadDelegate webViewShouldSendDoNotTrackHTTPHeader:]): 
3075             Pass the question on to the TestRunner.
3076         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
3077         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
3078         (WTR::InjectedBundlePage::InjectedBundlePage):
3079             Add the setShouldSendDoNotTrackHTTPHeader client function.
3080         (WTR::InjectedBundlePage::shouldSendDoNotTrackHTTPHeader):
3081             Pass the question on to the TestRunner.
3082         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
3083         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
3084         * WebKitTestRunner/InjectedBundle/TestRunner.h:
3085         (WTR::TestRunner::shouldSendDoNotTrackHTTPHeader): Simple getter.
3086         (WTR::TestRunner::setShouldSendDoNotTrackHTTPHeader): Simple Setter.
3087
3088 2012-09-10  Ojan Vafai  <ojan@chromium.org>
3089
3090         Make EFL bots work on the flakiness dashboard
3091         https://bugs.webkit.org/show_bug.cgi?id=96164
3092
3093         Reviewed by Tony Chang.
3094
3095         Also, update outdated unittests that had started failing.
3096
3097         * TestResultServer/static-dashboards/builders.js:
3098         (isWebkitTestRunner):
3099         * TestResultServer/static-dashboards/flakiness_dashboard.js:
3100         (nonChromiumPlatform):
3101         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
3102         (test):
3103
3104 2012-09-10  Tommy Widenflycht  <tommyw@google.com>
3105
3106         [chromium] MediaStream API: Remove the Descriptor postfix
3107         https://bugs.webkit.org/show_bug.cgi?id=96268
3108
3109         Reviewed by Adam Barth.
3110
3111         After consideration I realized that there is no need whatsoever in using the Descriptor prefix
3112         in the WekKit embedder API. Removed from WebRTCSessionDescription and WebRTCICECandidate to start with.
3113
3114         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
3115         (RTCSessionDescriptionRequestSuccededTask::RTCSessionDescriptionRequestSuccededTask):
3116         (MockWebRTCPeerConnectionHandler::createOffer):
3117         (MockWebRTCPeerConnectionHandler::createAnswer):
3118         (MockWebRTCPeerConnectionHandler::setLocalDescription):
3119         (MockWebRTCPeerConnectionHandler::setRemoteDescription):
3120         (MockWebRTCPeerConnectionHandler::localDescription):
3121         (MockWebRTCPeerConnectionHandler::remoteDescription):
3122         (MockWebRTCPeerConnectionHandler::addICECandidate):
3123         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h:
3124         (MockWebRTCPeerConnectionHandler):
3125
3126 2012-09-10  Arnaud Renevier  <a.renevier@sisa.samsung.com>
3127
3128         [Gtk] failure to build jhbuild dependencies with cups 1.6
3129         https://bugs.webkit.org/show_bug.cgi?id=95991
3130
3131         Reviewed by Gustavo Noronha Silva.
3132
3133         Bump gtk+ and gnome-themes-standard dependencies to 3.4.2 to allow
3134         building with cups >= 1.6 [GNOME #670373].
3135
3136         * gtk/jhbuild.modules:
3137
3138 2012-09-10  Christophe Dumez  <christophe.dumez@intel.com>
3139
3140         [EFL][DRT] Encoding issue in printed console messages
3141         https://bugs.webkit.org/show_bug.cgi?id=96263
3142
3143         Reviewed by Kenneth Rohde Christiansen.
3144
3145         Correctly handle CONSOLE messages that contain non-ascii
3146         characters in EFL's DRT. We need to use String::fromUTF8()
3147         to construct a WTF::String from a raw string.
3148
3149         * DumpRenderTree/efl/DumpRenderTreeView.cpp:
3150         (onConsoleMessage):
3151
3152 2012-09-10  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
3153
3154         [WK2][WTR] User style sheet location should be reset before each test
3155         https://bugs.webkit.org/show_bug.cgi?id=96262
3156
3157         Reviewed by Kenneth Rohde Christiansen.
3158
3159         User style sheet location is reset before each test now.
3160
3161         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
3162         (WTR::InjectedBundle::beginTesting):
3163
3164 2012-09-10  Simon Hausmann  <simon.hausmann@nokia.com>
3165
3166         Fix build of WTR with Qt on Windows
3167         https://bugs.webkit.org/show_bug.cgi?id=96249
3168
3169         Reviewed by Kenneth Rohde Christiansen.
3170
3171         Windows loves to define min and max as macros, which beautifully
3172         clashes in many places where min and max are used as variable or method
3173         names. It can be convinced not to do this when min and max are already
3174         define as macros (expanding to themselves). We have this kind of
3175         workaround in config.h files all over WebKit and as it turns out when
3176         building WTR with Qt on Windows we do end up with such a clash. Therefore
3177         we need the same workaround.
3178
3179         * WebKitTestRunner/config.h:
3180
3181 2012-09-10  Christophe Dumez  <christophe.dumez@intel.com>
3182
3183         [WK2] New fast/forms/number/number-interactive-validation-required.html fails
3184         https://bugs.webkit.org/show_bug.cgi?id=95936
3185
3186         Reviewed by Kenneth Rohde Christiansen.
3187
3188         Enable interactive form validation setting in WebKitTestRunner as
3189         we are currently doing in DumpRenderTree. This is needed for some
3190         tests to pass.
3191
3192         * WebKitTestRunner/TestController.cpp:
3193         (WTR::TestController::resetStateToConsistentValues):
3194
3195 2012-09-09  Zan Dobersek  <zandobersek@gmail.com>
3196
3197         Clean up the code in XvfbDriver
3198         https://bugs.webkit.org/show_bug.cgi?id=96202
3199
3200         Reviewed by Philippe Normand.
3201
3202         Changes:
3203         - Remove unneeded imports,
3204         - assign the _lock_file member variable when the display number is determined,
3205         - no need for parentheses around a single string format argument,
3206         - use FileSystem.exists method when checking if lock file is present as it
3207         makes the intention much more clearer.
3208
3209         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
3210         (XvfbDriver._start):
3211         (XvfbDriver.stop):
3212
3213 2012-09-09  Patrick Gansterer  <paroga@webkit.org>
3214
3215         Make the String initialization on the function side of String::number()
3216         https://bugs.webkit.org/show_bug.cgi?id=95940
3217
3218         Reviewed by Benjamin Poulain.
3219
3220         * TestWebKitAPI/Tests/WTF/IntegerToStringConversion.cpp:
3221         (testBoundaries):
3222         (testNumbers):
3223
3224 2012-09-09  Pierre Rossi  <pierre.rossi@gmail.com>
3225
3226         [WK2] expose element rect for color input type
3227         https://bugs.webkit.org/show_bug.cgi?id=91664
3228
3229         Reviewed by Simon Hausmann.
3230
3231         Update the QML logic in MiniBrowser to take advantage of this.
3232         The picker size is now hardcoded and we try to place it below the
3233         input element if possible.
3234
3235         * MiniBrowser/qt/qml/ColorChooser.qml:
3236
3237 2012-09-08  Christophe Dumez  <christophe.dumez@intel.com>
3238
3239         [WK2] New fast/events/tab-focus-link-in-canvas fails from r126908
3240         https://bugs.webkit.org/show_bug.cgi?id=95329
3241
3242         Reviewed by Kenneth Rohde Christiansen.
3243
3244         Explicitely disable TabToLinks setting in WebKitTestRunner.
3245         Tests override the "WebKitTabToLinksPreferenceKey" preference
3246         to enable it.
3247
3248         * WebKitTestRunner/TestController.cpp:
3249         (WTR::TestController::resetStateToConsistentValues):
3250
3251 2012-09-08  Kenichi Ishibashi  <bashi@chromium.org>
3252
3253         webkit.py gdb pretty printer can't print CString
3254         https://bugs.webkit.org/show_bug.cgi?id=96068
3255
3256         Reviewed by Tony Chang.
3257
3258         Update WTFCStringPrinter to follow r126191.
3259
3260         * gdb/webkit.py:
3261         (WTFCStringPrinter.to_string):
3262
3263 2012-09-07  Benjamin Poulain  <bpoulain@apple.com>
3264
3265         Rename the ustring() accessor to string()
3266         https://bugs.webkit.org/show_bug.cgi?id=95919
3267
3268         Reviewed by Geoffrey Garen.
3269
3270         Update EFL's DRT.
3271
3272         Fix some string conversion that have been made useless
3273         with recent changes.
3274
3275         * DumpRenderTree/efl/EventSender.cpp:
3276         (keyPadNameFromJSValue):
3277         (keyNameFromJSValue):
3278         * DumpRenderTree/efl/TestRunnerEfl.cpp:
3279         (TestRunner::queueLoad):
3280         (TestRunner::addOriginAccessWhitelistEntry):
3281         (TestRunner::removeOriginAccessWhitelistEntry):
3282         (TestRunner::setUserStyleSheetLocation):
3283         (TestRunner::setValueForUser):
3284         (TestRunner::elementDoesAutoCompleteForElementWithId):
3285         (TestRunner::execCommand):
3286         (TestRunner::findString):
3287         (TestRunner::isCommandEnabled):
3288         (TestRunner::clearApplicationCacheForOrigin):
3289         (TestRunner::setDomainRelaxationForbiddenForURLScheme):
3290         (TestRunner::pauseAnimationAtTimeOnElementWithId):
3291         (TestRunner::pauseTransitionAtTimeOnElementWithId):
3292         (toInt):
3293         (TestRunner::overridePreference):
3294         (TestRunner::addUserScript):
3295         (TestRunner::addUserStyleSheet):
3296         (TestRunner::evaluateInWebInspector):
3297         (TestRunner::evaluateScriptInIsolatedWorld):
3298         (TestRunner::setTextDirection):
3299         * DumpRenderTree/efl/WorkQueueItemEfl.cpp:
3300         (LoadItem::invoke):
3301         (LoadHTMLStringItem::invoke):
3302         (ScriptItem::invoke):
3303
3304 2012-09-06  Jeffrey Pfau  <jpfau@apple.com>
3305
3306         Extend third-party storage blocking API to optionally allow blocking all storage
3307         https://bugs.webkit.org/show_bug.cgi?id=95915
3308
3309         Reviewed by Brady Eidson.
3310
3311         Modified API test to use new revised API.
3312
3313         * TestWebKitAPI/Tests/WebKit2/WKPreferences.cpp:
3314         (TestWebKitAPI::TEST):
3315
3316 2012-09-07  Dominic Mazzoni  <dmazzoni@google.com>
3317
3318         AX: WebCore accessibility roles should be cross-platform
3319         https://bugs.webkit.org/show_bug.cgi?id=94870
3320
3321         Reviewed by Chris Fleizach.
3322
3323         Add debug strings to Chromium for new accessibility roles.
3324
3325         * DumpRenderTree/chromium/TestRunner/AccessibilityUIElementChromium.cpp:
3326         (roleToString):
3327
3328 2012-09-07  Dominic Mazzoni  <dmazzoni@google.com>
3329
3330         AX: Chromium needs access to canvasHasFallbackContent
3331         https://bugs.webkit.org/show_bug.cgi?id=96124
3332
3333         Reviewed by Chris Fleizach.
3334
3335         Distinguish between canvas roles with and without fallback
3336         content on Chromium.
3337
3338         * DumpRenderTree/chromium/TestRunner/AccessibilityUIElementChromium.cpp:
3339         (getRole):
3340
3341 2012-09-07  Ojan Vafai  <ojan@chromium.org>
3342
3343         Handle non-existant TestExpectations files gracefully
3344         https://bugs.webkit.org/show_bug.cgi?id=96149
3345
3346         Reviewed by Eric Seidel.
3347
3348         Even when we get rid of Skipped files, we shouldn't have each port
3349         require a TestExpectations file for each platform if it doesn't
3350         need one.
3351
3352         * Scripts/webkitpy/layout_tests/port/base.py:
3353         (Port.expectations_dict):
3354         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
3355         (PortTest.test_nonexistant_expectations):
3356
3357 2012-09-07  Tony Chang  <tony@chromium.org>
3358
3359         Regression: check-webkit-style no longer checks TestExpectations for syntax errors
3360         https://bugs.webkit.org/show_bug.cgi?id=96136
3361
3362         Reviewed by Ojan Vafai.
3363
3364         When we renamed the file to TestExpectations, it was no longer considered a text file so
3365         it was no longer being checked. Since then, test_expectations_overrides() has been removed
3366         from the port object, so I've removed that from the check. I'm not sure what it was
3367         supposed to do.
3368
3369         * Scripts/webkitpy/style/checker.py:
3370         (CheckerDispatcher._file_type): Mark TestExpectations as a text file.
3371         * Scripts/webkitpy/style/checkers/test_expectations.py:
3372         (TestExpectationsChecker.check_test_expectations): Remove overrides param.
3373         (TestExpectationsChecker.check): Remove overrides param.
3374         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
3375         (TestExpectationsTestCase.assert_lines_lint): Remove overrides param.
3376
3377 2012-09-07  Christophe Dumez  <christophe.dumez@intel.com>
3378
3379         [EFL] Use same default minimum logical font size in DRT and WTR
3380         https://bugs.webkit.org/show_bug.cgi?id=96116
3381
3382         Reviewed by Martin Robinson.
3383
3384         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
3385         (DumpRenderTreeChrome::resetDefaultsToConsistentValues): Set minimum
3386         logical font size to 9 to match the value used in WTR.
3387         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
3388         (WTR::InjectedBundle::beginTesting): Explicitly set the minimum
3389         logical font size to 9 (default value in WK2)
3390
3391 2012-09-07  Tommy Widenflycht  <tommyw@google.com>
3392
3393         MediaStream API: add RTCPeerConnection::createAnswer
3394         https://bugs.webkit.org/show_bug.cgi?id=96092
3395
3396         Reviewed by Adam Barth.
3397
3398         Expanding and enhancing MockWebRTCPeerConnectionHandler.
3399
3400         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
3401         (MockWebRTCPeerConnectionHandler::createOffer):
3402         (MockWebRTCPeerConnectionHandler::createAnswer):
3403         (MockWebRTCPeerConnectionHandler::setLocalDescription):
3404         (MockWebRTCPeerConnectionHandler::setRemoteDescription):
3405         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h:
3406         (MockWebRTCPeerConnectionHandler):
3407
3408 2012-09-07  Dean Jackson  <dino@apple.com>
3409
3410         Unreviewed changes to watchlist.
3411
3412         - Add new WebGL and Filters definitions
3413         - Add myself to OpenGL bugs
3414
3415         * Scripts/webkitpy/common/config/watchlist:
3416
3417 2012-09-07  Peter Beverloo  <peter@chromium.org>
3418
3419         Actually trigger the Android Tester
3420         https://bugs.webkit.org/show_bug.cgi?id=96118
3421
3422         Reviewed by Adam Barth.
3423
3424         * BuildSlaveSupport/build.webkit.org-config/config.json:
3425
3426 2012-09-07  Dominic Mazzoni <dmazzoni@google.com>
3427
3428         New time input needs accessibility
3429         https://bugs.webkit.org/show_bug.cgi?id=96032
3430
3431         Reviewed by Chris Fleizach.
3432
3433         Add support for valueDescription for testing.
3434
3435         * DumpRenderTree/chromium/TestRunner/AccessibilityUIElementChromium.cpp:
3436         (getValueDescription):
3437         (AccessibilityUIElement::AccessibilityUIElement):
3438         (AccessibilityUIElement::valueDescriptionGetterCallback):
3439         * DumpRenderTree/chromium/TestRunner/AccessibilityUIElementChromium.h:
3440         (AccessibilityUIElement):
3441
3442 2012-09-07  Peter Beverloo  <peter@chromium.org>
3443
3444         Add the google-android-tests build slave to the master's config
3445         https://bugs.webkit.org/show_bug.cgi?id=96113
3446
3447         Reviewed by Csaba Osztrogonác.
3448
3449         * BuildSlaveSupport/build.webkit.org-config/config.json:
3450
3451 2012-09-07  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
3452
3453         [WK2] [EFL] Set theme for WebKit2 API test runner
3454         https://bugs.webkit.org/show_bug.cgi?id=96100
3455
3456         Reviewed by Kenneth Rohde Christiansen.
3457
3458         Set the theme generated by the build for the WebKit2 API test runner.
3459
3460         * TestWebKitAPI/PlatformEfl.cmake:
3461         * TestWebKitAPI/efl/PlatformWebView.cpp:
3462         (TestWebKitAPI::PlatformWebView::PlatformWebView):
3463
3464 2012-09-07  Csaba Osztrogonác  <ossy@webkit.org>
3465
3466         Unreviewed. Add Gábor Ballabás as contributor.
3467
3468         * Scripts/webkitpy/common/config/committers.py:
3469
3470 2012-09-07  Zoltan Arvai  <zarvai@inf.u-szeged.hu>
3471
3472         [Qt] Fix ROOT_BUILD_DIR on Windows MSVC build.
3473         https://bugs.webkit.org/show_bug.cgi?id=96078
3474
3475         Reviewed by Tor Arne Vestbø.
3476
3477         Correct JSC, DumpRenderTree, ImageDiff, QtTestBrowser binaries location by adding $$toSystemPath to ROOT_BUILD_DIR generation.
3478
3479         * qmake/mkspecs/features/default_pre.prf:
3480
3481 2012-09-07  Christophe Dumez  <christophe.dumez@intel.com>
3482
3483         [EFL] Add TestExpectations file for efl-wk1
3484         https://bugs.webkit.org/show_bug.cgi?id=96019
3485
3486         Reviewed by Gyuyoung Kim.
3487
3488         Update TestExpectations cascading for EFL port
3489         so that we use efl-wk1 -> efl search paths
3490         for WebKit1 EFL.
3491
3492         * Scripts/webkitpy/layout_tests/port/efl.py:
3493         (EflPort._search_paths):
3494
3495 2012-09-07  Zan Dobersek  <zandobersek@gmail.com>
3496
3497         [NRWT] XvfbDriver should stop the Xvfb more aggressively
3498         https://bugs.webkit.org/show_bug.cgi?id=95097
3499
3500         Reviewed by Dirk Pranke.
3501
3502         The Xvfb process is now killed when the XvfbDriver is stopping. Simply
3503         terminating the process does not always work as expected, leading to timeouts
3504         in run-webkit-tests on the buildbot. The lock file Xvfb creates is also removed
3505         as it's not removed by the killed Xvfb process.
3506
3507         Also included is a thorough unittest for XvfbDriver. Proper starting of the driver
3508         is tested in different circumstances - whether pixel tests are enabled or not,
3509         with an arbitrary worker number or with an Xorg process already running on the system.
3510         The stopping of the driver is tested by checking that the Xvfb process is truly killed
3511         and then set to None.
3512
3513         * BuildSlaveSupport/kill-old-processes:
3514         Kill any stale Xvfb processes at the start of the build cycle so no Xvfb
3515         is left running, potentially clogging up a display at a certain position.
3516         * Scripts/webkitpy/common/system/executive_mock.py:
3517         (MockExecutive.__init__):
3518         The _running_pids is now a dictionary with process names as keys and pids
3519         as values.
3520         (MockExecutive.check_running_pid):
3521         Check whether the passed-in pid is in the _running_pids dictionary's values.
3522         (MockExecutive):
3523         (MockExecutive.running_pids):
3524         This method returns the list of all the process pids of which the name passes
3525         through the process_name_filter.
3526         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
3527         Use the Executive object of the port's host to check for running pids and open
3528         new subprocesses or kill them throughout the class.
3529         (XvfbDriver._start):
3530         Use the _server_process_constructor to make testing the XvfbDriver possible. Also,
3531         start the server process after it's created.
3532         (XvfbDriver.stop):
3533         Now kills the Xvfb process instead of terminating it and waiting for it to close.
3534         * Scripts/webkitpy/layout_tests/port/xvfbdriver_unittest.py: Added.
3535         (XvfbDriverTest):
3536         (XvfbDriverTest.make_driver):
3537         (XvfbDriverTest.assertDriverStartSuccessful):
3538         (XvfbDriverTest.test_start_no_pixel_tests):
3539         (XvfbDriverTest.test_start_pixel_tests):
3540         (XvfbDriverTest.test_start_arbitrary_worker_number):
3541         (XvfbDriverTest.test_start_existing_xorg_process):
3542         (XvfbDriverTest.test_stop):
3543         (XvfbDriverTest.test_stop.FakeXvfbProcess):
3544
3545 2012-09-06  Peter Beverloo  <peter@chromium.org>
3546
3547         Introduce the Chromium Android Release (Tests) bot
3548         https://bugs.webkit.org/show_bug.cgi?id=95980
3549
3550         Reviewed by Adam Barth.
3551
3552         This adds the Chromium Android Release (Tests) bot to the WebKit waterfall,
3553         which will be running layout tests on actual Android devices. Right now,
3554         four Galaxy Nexus phones are attached.
3555
3556         * BuildSlaveSupport/build.webkit.org-config/config.json:
3557
3558 2012-09-06  James Robinson  <jamesr@chromium.org>
3559
3560         [chromium] Use WebCompositorSupport functions instead of WebCompositor statics
3561         https://bugs.webkit.org/show_bug.cgi?id=96007
3562
3563         Reviewed by Adrienne Walker.
3564
3565         Fixes initialization / settings code to use WebCompositorSupport.
3566
3567         * DumpRenderTree/chromium/DumpRenderTree.cpp:
3568         (main):
3569         * DumpRenderTree/chromium/TestShell.cpp:
3570         (TestShell::initialize):
3571         (TestShell::setPerTilePaintingEnabled):
3572
3573 2012-09-06  Christophe Dumez  <christophe.dumez@intel.com>
3574
3575         [EFL][WK2] Provide implementation for PlatformWebView::windowFrame()
3576         https://bugs.webkit.org/show_bug.cgi?id=95995
3577
3578         Reviewed by Kenneth Rohde Christiansen.
3579
3580         Provide proper implementation for PlatformWebView::windowFrame()
3581         in EFL's WebKitTestRunner.
3582
3583         * WebKitTestRunner/efl/PlatformWebViewEfl.cpp:
3584         (WTR::PlatformWebView::windowFrame):
3585
3586 2012-09-06  Fady Samuel  <fsamuel@chromium.org>
3587
3588         Fix WTF.Lrint on 32-bit platforms
3589         https://bugs.webkit.org/show_bug.cgi?id=96003
3590
3591         Reviewed by Benjamin Poulain.
3592
3593         Fixed failing WTF.Lrint test on 32-bit platforms.
3594
3595         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
3596         (TestWebKitAPI::TEST):
3597
3598 2012-09-06  Tommy Widenflycht  <tommyw@google.com>
3599
3600         MediaStream API: Add the local and remote description functionality to RTCPeerConnection
3601         https://bugs.webkit.org/show_bug.cgi?id=95839
3602
3603         Reviewed by Adam Barth.
3604
3605         In addition to adding the new methods to MockWebRTCPeerConnectionHandler I cleaned
3606         up the task definitions a bit. There was no need to have them in the header file.
3607
3608         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
3609         (RTCSessionDescriptionRequestSuccededTask):
3610         (RTCSessionDescriptionRequestSuccededTask::RTCSessionDescriptionRequestSuccededTask):
3611         (RTCSessionDescriptionRequestFailedTask):
3612         (RTCSessionDescriptionRequestFailedTask::RTCSessionDescriptionRequestFailedTask):
3613         (RTCVoidRequestTask):
3614         (RTCVoidRequestTask::RTCVoidRequestTask):
3615         (MockWebRTCPeerConnectionHandler::createOffer):
3616         (MockWebRTCPeerConnectionHandler::setLocalDescription):
3617         (MockWebRTCPeerConnectionHandler::setRemoteDescription):
3618         (MockWebRTCPeerConnectionHandler::localDescription):
3619         (MockWebRTCPeerConnectionHandler::remoteDescription):
3620         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h:
3621         (MockWebRTCPeerConnectionHandler):
3622
3623 2012-09-06  Simon Hausmann  <simon.hausmann@nokia.com>
3624
3625         [Qt] Fix build on Mac OS X 10.6
3626         https://bugs.webkit.org/show_bug.cgi?id=95200
3627
3628         Reviewed by Tor Arne Vestbø.
3629
3630         The build on Mac OS X 10.6 is broken when using QuickTime. Enable it only for
3631         10.7 and newer.
3632
3633         * qmake/mkspecs/features/features.prf:
3634
3635 2012-09-06  Christophe Dumez  <christophe.dumez@intel.com>
3636
3637         [WKTR] addSelectionKey should be mapped to "Control" modifiers on other OSes than Mac
3638         https://bugs.webkit.org/show_bug.cgi?id=95973
3639
3640         Reviewed by Kenneth Rohde Christiansen.
3641
3642         Correctly map "addSelectionKey" to "Control" modifier
3643         in WKTR's EventSender on other OSes than Mac OS X,
3644         instead of mapping it to the "Meta" modifier.
3645
3646         * WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
3647         (WTR::parseModifier):
3648
3649 2012-09-06  Balazs Kelemen  <kbalazs@webkit.org>
3650
3651         [nrwt] Harden against empty stdout/stderr after stopping a crashed/timed out driver
3652         https://bugs.webkit.org/show_bug.cgi?id=95710
3653
3654         Reviewed by Dirk Pranke.
3655
3656         The stdout / stderr output can be None so make sure
3657         we won't raise an exception in this case.
3658
3659         * Scripts/webkitpy/layout_tests/port/driver.py:
3660         (Driver.run_test):
3661
3662 2012-09-06  Christophe Dumez  <christophe.dumez@intel.com>
3663
3664         [EFL] Sanitize URLs in EWebLauncher / Minibrowser
3665         https://bugs.webkit.org/show_bug.cgi?id=95849
3666
3667         Reviewed by Kenneth Rohde Christiansen.
3668
3669         Sanitize URLs passed by the user as program
3670         argument or typed in the URL bar in order to
3671         be more tolerant.
3672
3673         * EWebLauncher/CMakeLists.txt:
3674         * EWebLauncher/main.c:
3675         (main):
3676         * EWebLauncher/url_bar.c:
3677         (_url_bar_url_get_with_protocol):
3678         (on_urlbar_key_down):
3679         * EWebLauncher/url_utils.c: Added.
3680         (has_scheme):
3681         (url_from_user_input): Add utility function to construct a
3682         URL from user input.
3683         * EWebLauncher/url_utils.h: Added.
3684         * MiniBrowser/efl/CMakeLists.txt:
3685         * MiniBrowser/efl/main.c:
3686         (main):
3687
3688 2012-09-06  Simon Hausmann  <simon.hausmann@nokia.com>
3689
3690         [Qt] Various small Windows / MSVC build fixes
3691         https://bugs.webkit.org/show_bug.cgi?id=95934
3692
3693         Reviewed by Kenneth Rohde Christiansen.
3694
3695         setFocus takes a boolean as argument: Don't try to implicitly convert
3696         an enum to a boolean, MSVC doesn't like that. The intent here is to
3697         enable focus, so just pass true.
3698
3699         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
3700         (WTR::PlatformWebView::focus):
3701
3702 2012-09-06  Simon Hausmann  <simon.hausmann@nokia.com>
3703
3704         [Qt] Fix build with ANGLE on platforms that provide EGL/GLESv2
3705         https://bugs.webkit.org/show_bug.cgi?id=95965
3706
3707         Reviewed by Tor Arne Vestbø.
3708
3709         Add ANGLE as a static library module to the Qt WebKit build system.
3710
3711         * qmake/mkspecs/features/default_post.prf:
3712         * qmake/mkspecs/modules/angle.prf: Added.
3713
3714 2012-09-06  Andras Becsi  <andras.becsi@nokia.com>
3715
3716         [WK2] Make [Qt]ViewportHandler cross platform
3717         https://bugs.webkit.org/show_bug.cgi?id=91257
3718
3719         Reviewed by Kenneth Rohde Christiansen.
3720
3721         Remove initial scale from viewport info item since it is
3722         set to -1 after applied thus it does not hold any valuable
3723         information.
3724
3725         * MiniBrowser/qt/qml/ViewportInfoItem.qml:
3726
3727 2012-09-05  Tony Chang  <tony@chromium.org>
3728
3729         [chromium] Remove dependency on webkit_user_agent
3730         https://bugs.webkit.org/show_bug.cgi?id=95886
3731
3732         Reviewed by Adam Barth.
3733
3734         The code isn't used by DumpRenderTree.
3735
3736         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
3737
3738 2012-09-05  Rick Byers  <rbyers@chromium.org>
3739
3740         Update chromium DumpRenderTree for new WebGestureEvent fields
3741         https://bugs.webkit.org/show_bug.cgi?id=95573
3742
3743         Reviewed by James Robinson.
3744
3745         Update EventSender to write gesture details into the appropriate
3746         type-specific WebGestureEvent fields.
3747         * DumpRenderTree/chromium/TestRunner/EventSender.cpp:
3748         (EventSender::gestureEvent):
3749         (EventSender::gestureFlingStart):
3750
3751 2012-09-05  Dirk Pranke  <dpranke@chromium.org>
3752
3753         NRWT lies about the number of tests that it's skipping when used with --repeat-each
3754         https://bugs.webkit.org/show_bug.cgi?id=95789
3755
3756         Reviewed by Ojan Vafai.
3757
3758         Fix miscalculation in log message.
3759
3760         * Scripts/webkitpy/layout_tests/views/printing.py:
3761         (Printer.print_found):
3762
3763 2012-09-05  Kenichi Ishibashi  <bashi@chromium.org>
3764
3765         Unreviewed, rolling out r127612, r127660, and r127664.
3766         http://trac.webkit.org/changeset/127612
3767         http://trac.webkit.org/changeset/127660
3768         http://trac.webkit.org/changeset/127664
3769         https://bugs.webkit.org/show_bug.cgi?id=95920
3770
3771         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
3772         (MockWebRTCPeerConnectionHandler::SuccessCallbackTask::SuccessCallbackTask):
3773         (MockWebRTCPeerConnectionHandler::SuccessCallbackTask::runIfValid):
3774         (MockWebRTCPeerConnectionHandler::FailureCallbackTask::FailureCallbackTask):
3775         (MockWebRTCPeerConnectionHandler::FailureCallbackTask::runIfValid):
3776         (MockWebRTCPeerConnectionHandler::createOffer):
3777         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h:
3778         (MockWebRTCPeerConnectionHandler):
3779         (SuccessCallbackTask):
3780         (FailureCallbackTask):
3781
3782 2012-09-05  Allan Sandfeld Jensen  <allan.jensen@nokia.com>
3783
3784         Ignore-metrics includes scroll offsets.
3785         https://bugs.webkit.org/show_bug.cgi?id=95845
3786
3787         Reviewed by Dirk Pranke.
3788
3789         Strip scroll-offsets the same way scroll-height and scroll-width is stripped.
3790
3791         * Scripts/old-run-webkit-tests:
3792         (stripMetrics):
3793         * Scripts/webkitpy/layout_tests/port/driver.py:
3794         (DriverOutput):
3795
3796 2012-09-05  Zan Dobersek  <zandobersek@gmail.com>
3797
3798         Unreviewed test fix.
3799
3800         Including a plugin source file into the build that was added in r127595.
3801
3802         * GNUmakefile.am:
3803
3804 2012-09-05  Tommy Widenflycht  <tommyw@google.com>
3805
3806         MediaStream API: Add the local and remote description functionality to RTCPeerConnection
3807         https://bugs.webkit.org/show_bug.cgi?id=95839
3808
3809         Reviewed by Adam Barth.
3810
3811         In addition to adding the new methods to MockWebRTCPeerConnectionHandler I cleaned
3812         up the task definitions a bit. There was no need to have them in the header file.
3813
3814         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
3815         (RTCSessionDescriptionRequestSuccededTask):
3816         (RTCSessionDescriptionRequestSuccededTask::RTCSessionDescriptionRequestSuccededTask):
3817         (RTCSessionDescriptionRequestFailedTask):
3818         (RTCSessionDescriptionRequestFailedTask::RTCSessionDescriptionRequestFailedTask):
3819         (RTCVoidRequestTask):
3820         (RTCVoidRequestTask::RTCVoidRequestTask):
3821         (MockWebRTCPeerConnectionHandler::createOffer):
3822         (MockWebRTCPeerConnectionHandler::setLocalDescription):
3823         (MockWebRTCPeerConnectionHandler::setRemoteDescription):
3824         (MockWebRTCPeerConnectionHandler::localDescription):
3825         (MockWebRTCPeerConnectionHandler::remoteDescription):
3826         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h:
3827         (MockWebRTCPeerConnectionHandler):
3828
3829 2012-09-05  Kangil Han  <kangil.han@samsung.com>
3830
3831         [EFL] Remove '+=' usage in String
3832         https://bugs.webkit.org/show_bug.cgi?id=95808
3833
3834         Reviewed by Benjamin Poulain.
3835
3836         Replaced use of String operator '+=' with StringBuilder.append*() in order to adopt efficient String concatenation.
3837
3838         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
3839         (descriptionSuitableForTestResult):
3840
3841 2012-09-05  Christophe Dumez  <christophe.dumez@intel.com>
3842
3843         [EFL][WK2] Provide implementation for TestRunner::pathToLocalResource()
3844         https://bugs.webkit.org/show_bug.cgi?id=95842
3845
3846         Reviewed by Kenneth Rohde Christiansen.
3847
3848         Provide proper implementation for TestRunner::pathToLocalResource()
3849         in EFL WKTR, identical to the one for DumpRenderTree.
3850
3851         Map /tmp to ${DUMPRENDERTREE_TEMP} environment variable and
3852         /tmp/LayoutTests to ${LOCAL_RESOURCE_ROOT} so that local resources
3853         are found my WebKitTestRunner.
3854
3855         * WebKitTestRunner/InjectedBundle/efl/TestRunnerEfl.cpp:
3856         (WTR::TestRunner::pathToLocalResource):
3857
3858 2012-09-05  Brady Eidson  <beidson@apple.com>
3859
3860         Frequent crashes in PluginView::scriptObject under runtimeObjectCustomGetOwnPropertySlot
3861         <rdar://problem/12142226> and https://bugs.webkit.org/show_bug.cgi?id=95026
3862
3863         Reviewed by Andy Estes.
3864
3865         Add a plug-in that always fails to initialize:
3866         * DumpRenderTree/TestNetscapePlugIn/Tests/NPPNewFails.cpp: Added.
3867         (NPPNewFails):
3868         (NPPNewFails::NPPNewFails):
3869         (NPPNewFails::NPP_New):
3870
3871         Add it to all the project files:
3872         * DumpRenderTree/DumpRenderTree.gypi:
3873         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
3874         * DumpRenderTree/TestNetscapePlugIn/CMakeLists.txt:
3875         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
3876         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
3877
3878 2012-09-05  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3879
3880         [Qt] Fix makefile race condition between derived sources and target
3881
3882         r121966 removed too much of the old qmake logic, including the rule
3883         that ensured that qmake_all would not trigger dual runs of qmake --
3884         one of the derived sources -qmake-all rule, and one for making the
3885         derived sources.
3886
3887         We now follow a similar pattern as before, by introducing a custom
3888         target that explicitly depends on the derived sources' qmake-all before
3889         doing a manual run of make. That ensures a single qmake run for the derived
3890         sources.
3891
3892         https://bugs.webkit.org/show_bug.cgi?id=93847
3893
3894         Reviewed by Csaba Osztrogonác.
3895
3896         * qmake/mkspecs/features/functions.prf:
3897
3898 2012-09-05  Christophe Dumez  <christophe.dumez@intel.com>
3899
3900         WKTR doesn't implement dumpWillCacheResponse()
3901         https://bugs.webkit.org/show_bug.cgi?id=94607
3902
3903         Reviewed by Kenneth Rohde Christiansen.
3904
3905         Define shouldCacheResponse() callback in resource
3906         load client and provide implementation for
3907         dumpWillCacheResponse() in WebKitTestRunner.
3908
3909         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
3910         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
3911         (WTR::InjectedBundlePage::InjectedBundlePage):
3912         (WTR::InjectedBundlePage::shouldCacheResponse):
3913         (WTR):
3914         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
3915         (InjectedBundlePage):
3916         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
3917         (WTR::TestRunner::TestRunner):
3918         * WebKitTestRunner/InjectedBundle/TestRunner.h:
3919         (WTR::TestRunner::dumpWillCacheResponse):
3920         (WTR::TestRunner::shouldDumpWillCacheResponse):
3921         (TestRunner):
3922
3923 2012-09-05  Kaustubh Atrawalkar  <kaustubh@motorola.com>
3924
3925         [DRT] LTC:: Move printing related APIs from LayoutTestController to Internals
3926         https://bugs.webkit.org/show_bug.cgi?id=92735
3927
3928         Reviewed by Hajime Morita.
3929
3930         Move numberOfPages, pageProperty & pageSizeAndMarginsInPixels in Internals and remove duplicated code from DumprenderTree & WebkitTestRunner.
3931
3932         * DumpRenderTree/TestRunner.cpp:
3933         (TestRunner::staticFunctions):
3934         * DumpRenderTree/TestRunner.h:
3935         (TestRunner):
3936         * DumpRenderTree/blackberry/TestRunnerBlackBerry.cpp:
3937         (TestRunner::addUserStyleSheet):
3938         * DumpRenderTree/chromium/DRTTestRunner.cpp:
3939         (DRTTestRunner::DRTTestRunner):
3940         (DRTTestRunner::hasCustomPageSizeStyle):
3941         * DumpRenderTree/chromium/DRTTestRunner.h:
3942         (DRTTestRunner):
3943         * DumpRenderTree/efl/TestRunnerEfl.cpp:
3944         * DumpRenderTree/gtk/TestRunnerGtk.cpp:
3945         * DumpRenderTree/mac/TestRunnerMac.mm:
3946         * DumpRenderTree/qt/TestRunnerQt.cpp:
3947         (TestRunner::evaluateScriptInIsolatedWorld):
3948         * DumpRenderTree/qt/TestRunnerQt.h:
3949         (TestRunner):
3950         * DumpRenderTree/win/TestRunnerWin.cpp:
3951         * DumpRenderTree/wx/TestRunnerWx.cpp:
3952         (TestRunner::abortModal):
3953         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
3954         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
3955         (WTR):
3956         * WebKitTestRunner/InjectedBundle/TestRunner.h:
3957         (TestRunner):
3958
3959 2012-09-05  Simon Hausmann  <simon.hausmann@nokia.com>
3960
3961         [Qt] Automate the use of AllInOne files in the Qt build
3962         https://bugs.webkit.org/show_bug.cgi?id=95837
3963
3964         Reviewed by Tor Arne Vestbø.
3965
3966         New feature file that implements the all-in-one file automation by
3967         adding the requested all-in-one files to SOURCES and removing the files
3968         it includes from SOURCES.
3969
3970         * qmake/mkspecs/features/use_all_in_one_files.prf: Added.
3971
3972 2012-09-04  Tim Horton  <timothy_horton@apple.com>
3973
3974         Have NRWT log more output when crash logs are missing
3975         https://bugs.webkit.org/show_bug.cgi?id=95814
3976
3977         Reviewed by Dan Bernstein.
3978
3979         Log the stderr and stdout of the test, in order to provide slightly more
3980         context when the WebProcess times out while running tests.
3981
3982         * Scripts/webkitpy/layout_tests/port/driver.py:
3983         (Driver.run_test):
3984
3985 2012-09-04  Xianzhu Wang  <wangxianzhu@chromium.org>
3986
3987         [Chromium-Android] Include device serial number in ChromiumAndroidDriver log messages
3988         https://bugs.webkit.org/show_bug.cgi?id=95770
3989
3990         Reviewed by Adam Barth.
3991
3992         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
3993         (ChromiumAndroidDriver.__init__):
3994         (ChromiumAndroidDriver._log_error): Added to replace original _log.error().
3995         (ChromiumAndroidDriver._log_debug): Added to replace original _log.debug().
3996         (ChromiumAndroidDriver._abort): Added to replace original raise AssertionError().
3997         (ChromiumAndroidDriver._push_executable):
3998         (ChromiumAndroidDriver._push_fonts):
3999         (ChromiumAndroidDriver._push_test_resources):
4000         (ChromiumAndroidDriver._run_adb_command):
4001         (ChromiumAndroidDriver._get_last_stacktrace):
4002         (ChromiumAndroidDriver._get_crash_log):
4003         (ChromiumAndroidDriver._start):
4004         (ChromiumAndroidDriver._start_once):
4005         (ChromiumAndroidDriver._start_once.deadlock_detector):
4006         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
4007         (ChromiumAndroidDriverTest.test_get_crash_log):
4008
4009 2012-09-04  Satish Sampath  <satish@chromium.org>
4010
4011         Run git commands within the checkout directory
4012         https://bugs.webkit.org/show_bug.cgi?id=95742
4013
4014         Reviewed by Eric Seidel.
4015
4016         git commands run from scm/git.py are sometimes run with cwd set to the
4017         checkout_root and other times with whatever directory was the default cwd.
4018         Instead all these should run with the checkout root set as the cwd
4019         (except where we explicitly need a different path)
4020
4021         * Scripts/webkitpy/common/checkout/scm/git.py:
4022         (Git._run_git):
4023         (Git.find_checkout_root):
4024         (Git.discard_local_commits):
4025         (Git.local_commits):
4026         (Git.working_directory_is_clean):
4027         (Git.clean_working_directory):
4028         (Git.add_list):
4029         (Git.delete_list):
4030         (Git.exists):
4031         (Git._current_branch):
4032         (Git._changes_files_for_commit):
4033         (Git.revisions_changing_file):
4034         (Git.svn_revision):
4035         (Git._run_git_svn_find_rev):
4036         (Git.contents_at_revision):
4037         (Git.diff_for_file):
4038         (Git.show_head):
4039         (Git.committer_email_for_revision):
4040         (Git.apply_reverse_diff):
4041         (Git.revert_files):
4042         (Git.commit_with_message):
4043         (Git._commit_on_branch):
4044         (Git.svn_commit_log):
4045         (Git.last_svn_commit_log):
4046         (Git.svn_blame):
4047         (Git._branch_ref_exists):
4048         (Git.delete_branch):
4049         (Git.remote_merge_base):
4050         (Git.commit_locally_with_message):
4051         (Git.push_local_commits_to_server):
4052         (Git.commit_ids_from_commitish_arguments):
4053         (Git.commit_message_for_local_commit):
4054         (Git.files_changed_summary_for_commit):
4055
4056 2012-09-04  Tommy Widenflycht  <tommyw@google.com>
4057
4058         MediaStream API: Add the async createOffer functionality to RTCPeerConnection
4059         https://bugs.webkit.org/show_bug.cgi?id=95734
4060
4061         Reviewed by Adam Barth.
4062
4063         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
4064         (MockWebRTCPeerConnectionHandler::SuccessCallbackTask::SuccessCallbackTask):
4065         (MockWebRTCPeerConnectionHandler::SuccessCallbackTask::runIfValid):
4066         (MockWebRTCPeerConnectionHandler::FailureCallbackTask::FailureCallbackTask):
4067         (MockWebRTCPeerConnectionHandler::FailureCallbackTask::runIfValid):
4068         (MockWebRTCPeerConnectionHandler::createOffer):
4069         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h:
4070         (MockWebRTCPeerConnectionHandler):
4071         (MockWebRTCPeerConnectionHandler::taskList):
4072         (SuccessCallbackTask):
4073         (FailureCallbackTask):
4074
4075 2012-09-04  Xianzhu Wang  <wangxianzhu@chromium.org>
4076
4077         [Chromium-Android] Output time and thread info in crash log
4078         https://bugs.webkit.org/show_bug.cgi?id=95768
4079
4080         Reviewed by Adam Barth.
4081
4082         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
4083         (ChromiumAndroidDriver._get_logcat): Add '-v threadtime' to adb logcat command line to get time and thread info.
4084
4085 2012-09-04  Tommy Widenflycht  <tommyw@google.com>
4086
4087         MediaStream API: Change the MediaStreamTrackList track added/removed signaling
4088         https://bugs.webkit.org/show_bug.cgi?id=95721
4089
4090         Reviewed by Adam Barth.
4091
4092         * DumpRenderTree/chromium/MockWebMediaStreamCenter.cpp:
4093         (MockWebMediaStreamCenter::didAddMediaStreamTrack):
4094         (MockWebMediaStreamCenter::didRemoveMediaStreamTrack):
4095         * DumpRenderTree/chromium/MockWebMediaStreamCenter.h:
4096         (MockWebMediaStreamCenter):
4097
4098 2012-09-04  Tony Chang  <tony@chromium.org>
4099
4100         Elide test names when running new-run-webkit-tests
4101         https://bugs.webkit.org/show_bug.cgi?id=95626
4102
4103         Reviewed by Dirk Pranke.
4104
4105         When the test name exceeds the number of terminal columns, it leaves
4106         text on the screen. Avoid this by eliding the test name when running
4107         tests. This is similar to what ninja does when compiling.
4108
4109         * Scripts/webkitpy/layout_tests/views/metered_stream.py:
4110         (MeteredStream):
4111         (MeteredStream.number_of_columns): Helper method to get number of
4112         terminal columns. Won't work on Win32 python, but meh.
4113         * Scripts/webkitpy/layout_tests/views/printing.py:
4114         (Printer._test_status_line): New method for eliding the text. Doesn't use \u2026
4115         because stderr appears to be ascii only.
4116         (Printer.print_started_test): Route through _test_status_line.
4117         (Printer.print_finished_test): Route through _test_status_line.
4118         (Printer._print_test_trace): Route through _test_status_line.
4119         (Printer._print_baseline): Route through _test_status_line.
4120         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
4121         (test_test_status_line): Test for _test_status_line.
4122
4123 2012-09-04  Martin Robinson  <mrobinson@igalia.com>
4124
4125         [GTK] Print API missing documentation when generating gtkdoc
4126         https://bugs.webkit.org/show_bug.cgi?id=95703
4127
4128         Reviewed by Carlos Garcia Campos.
4129
4130         Print out API missing documentation when generating gtkdoc. This makes it
4131         easier to fix documentation errors.
4132
4133         * gtk/generate-gtkdoc:
4134         (print_missing_api): Added.
4135         (generate_doc): Inline the rebase step, since it now reports an error
4136         when you try to rebase without generating documentation first.
4137         * gtk/gtkdoc.py:
4138         (GTKDoc.rebase_installed_docs): Raise an error when rebasing without
4139         generating documentation first to enable a small cleanup at the caller.
4140         (GTKDoc.api_missing_documentation): Added.
4141
4142 2012-09-04  Vincent Scheib  <scheib@chromium.org>
4143
4144         [Chromium] Allow asyncronous response of pointer lock requests in layout tests.
4145         https://bugs.webkit.org/show_bug.cgi?id=95628
4146
4147         Reviewed by James Robinson.
4148
4149         Future tests need the ability to respond asynchronously that
4150         pointer lock requests have been accepted or rejected. This patch
4151         adds the following:
4152             testRunner.setPointerLockWillRespondAsynchronously();
4153             testRunner.didAcquirePointerLock();
4154             testRunner.didNotAcquirePointerLock();
4155
4156         Tested in existing test:
4157             pointerlockchange-pointerlockerror-events.html
4158
4159         * DumpRenderTree/chromium/DRTTestRunner.cpp:
4160         (DRTTestRunner::DRTTestRunner):
4161         (DRTTestRunner::didAcquirePointerLock):
4162         (DRTTestRunner::didNotAcquirePointerLock):
4163         (DRTTestRunner::setPointerLockWillRespondAsynchronously):
4164         * DumpRenderTree/chromium/DRTTestRunner.h:
4165         (DRTTestRunner):
4166         * DumpRenderTree/chromium/WebViewHost.cpp:
4167         (WebViewHost::requestPointerLock):
4168         (WebViewHost::didAcquirePointerLock):
4169         (WebViewHost::didNotAcquirePointerLock):
4170         * DumpRenderTree/chromium/WebViewHost.h:
4171         (WebViewHost):
4172         (WebViewHost::setPointerLockWillRespondAsynchronously):
4173
4174 2012-09-04  Ojan Vafai  <ojan@chromium.org>
4175
4176         Move content shell bots to their own builder group
4177         https://bugs.webkit.org/show_bug.cgi?id=95765
4178
4179         Reviewed by Adam Barth.
4180
4181         In their current state of instability, they add too much noise. Once they get more stable,
4182         we can add them back to the regular group or just have them replace the other bots.
4183
4184         Also, fix the filter for Apple Windows bots since some of them are now "Apple Win ...".
4185
4186         * TestResultServer/static-dashboards/builders.js:
4187         (isWebkitTestRunner):
4188         (isChromiumContentShellTestRunner):
4189         (isChromiumWebkitTipOfTreeTestRunner):
4190         (loadBuildersList):
4191
4192 2012-09-03  Ryuan Choi  <ryuan.choi@samsung.com>
4193
4194         [EFL] Change the focus when the web view is clicked in MiniBrowser and EWebLauncher
4195         https://bugs.webkit.org/show_bug.cgi?id=95689
4196
4197         Reviewed by Kenneth Rohde Christiansen.
4198
4199         In Minibrowser & EWebLauncher, the web view loses the focus when url bar is clicked.
4200         So, there are bugs that key events are always generated in url bar although
4201         the user clicked the web view.
4202
4203         This patch recovers the focus of the web view when left button is pressed and
4204         adds for MiniBrowser to toggle the focus when middle button is pressed
4205         like EWebLauncher already did.
4206
4207         * EWebLauncher/main.c:
4208         (on_mouse_down):
4209         * MiniBrowser/efl/main.c:
4210         (on_mouse_down):
4211         (browserCreate):
4212
4213 2012-09-03  Tao Bai  <michaelbai@chromium.org>
4214
4215         Added ANDROID_GDBSERVER definition for Ant
4216         https://bugs.webkit.org/show_bug.cgi?id=95265
4217
4218         Reviewed by Adam Barth.
4219
4220         ANDROID_GDBSERVER will replace ANDROID_TOOLCHAIN, because the ANDROID_TOOLCHAIN
4221         is only used for finding the gdbserver.
4222
4223         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
4224         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
4225
4226 2012-09-03  Tommy Widenflycht  <tommyw@google.com>
4227
4228         MediaStream API: Add Ice-related functionality to RTCPeerConnection
4229         https://bugs.webkit.org/show_bug.cgi?id=95565
4230
4231         Reviewed by Adam Barth.
4232
4233         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
4234         (MockWebRTCPeerConnectionHandler::MockWebRTCPeerConnectionHandler):
4235         (MockWebRTCPeerConnectionHandler::updateICE):
4236         (MockWebRTCPeerConnectionHandler::addICECandidate):
4237         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h:
4238         (MockWebRTCPeerConnectionHandler):
4239
4240 2012-09-03  Christophe Dumez  <christophe.dumez@intel.com>
4241
4242         [EFL] Fix resizing issues related to URL bar in Minibrowser / EWebLauncher
4243         https://bugs.webkit.org/show_bug.cgi?id=95670
4244
4245         Reviewed by Kenneth Rohde Christiansen.
4246
4247         Fix layout issues related to the URL bar in Minibrowser
4248         and EWebLauncher when the window is resized.
4249
4250         The URL bar was not resized accordingly and the Web
4251         page was moved behing the URL bar, hiding part of it.
4252
4253         * EWebLauncher/main.c:
4254         (on_ecore_evas_resize):
4255         (closeWindow):
4256         * EWebLauncher/url_bar.c:
4257         (url_bar_width_set):
4258         (url_bar_add):
4259         * EWebLauncher/url_bar.h:
4260         * MiniBrowser/efl/main.c:
4261         (on_ecore_evas_resize):
4262         (main):
4263
4264 2012-09-01  Xianzhu Wang  <wangxianzhu@chromium.org>
4265
4266         [Chromium-Android] Remove hard-coded skipped directories in chromium_android.py
4267         https://bugs.webkit.org/show_bug.cgi?id=95605
4268
4269         Reviewed by Dirk Pranke.
4270
4271         Now skipping rules are put in LayoutTests/platform/chromium-android/TestExpectations.
4272
4273         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
4274         (ChromiumAndroidPort.check_sys_deps):
4275         (ChromiumAndroidPort.expectations_files):
4276         (ChromiumAndroidPort.clean_up_test_run):
4277
4278 2012-09-01  Joanmarie Diggs  <jdiggs@igalia.com>
4279
4280         [Gtk] Incorrect/unexpected characters in the text of certain accessibles
4281         https://bugs.webkit.org/show_bug.cgi?id=95180
4282
4283         Reviewed by Chris Fleizach.
4284
4285         The bug that was fixed stood in the way of fully implementing stringValue().
4286         Testing that the bug is fixed requires stringValue() to be fully implemented
4287         and object replacement characters to be printable.
4288
4289         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
4290         (replaceCharactersForResults): New method which turns object replacement
4291         characters into "<obj>" so that the characters can be properly shown in
4292         Layout Test results. Also turns "\n" into "<\\n>" so that printing the
4293         accessible text of a single object in the accessible tree doesn't mess up
4294         the readibility of the results.
4295         (AccessibilityUIElement::stringValue): Remove the code that immediately
4296         returned upon encountering an object of ATK_ROLE_PANEL and call the new
4297         replaceCharactersForResults() prior to returning the accessible string
4298         value.
4299
4300 2012-09-01  Tommy Widenflycht  <tommyw@google.com>
4301
4302         MediaStream API: Add MediaStream management to RTCPeerConnection
4303         https://bugs.webkit.org/show_bug.cgi?id=95543
4304
4305         Reviewed by Adam Barth.
4306
4307         Extending MockWebRTCPeerConnectionHandler to handle addStream/removeStream.
4308
4309         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
4310         (MockWebRTCPeerConnectionHandler::MockWebRTCPeerConnectionHandler):
4311         (MockWebRTCPeerConnectionHandler::addStream):
4312         (MockWebRTCPeerConnectionHandler::removeStream):
4313         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h:
4314         (MockWebRTCPeerConnectionHandler):
4315
4316 2012-08-31  Tony Chang  <tony@chromium.org>
4317
4318         Remove ENABLE_CSS3_FLEXBOX compile time flag
4319         https://bugs.webkit.org/show_bug.cgi?id=95382
4320
4321         Reviewed by Ojan Vafai.
4322
4323         Everyone is already enabling this by default and the spec has stablized.
4324
4325         * Scripts/webkitperl/FeatureList.pm:
4326         * qmake/mkspecs/features/features.pri:
4327
4328 2012-08-31  Pratik Solanki  <psolanki@apple.com>
4329
4330         objc_msgSend and IMP should be cast appropriately before using
4331         https://bugs.webkit.org/show_bug.cgi?id=95242
4332
4333         Reviewed by Benjamin Poulain.
4334
4335         Fix for older compilers. Pass id as the return type to the template
4336         instead of relying on default type.
4337
4338         * DumpRenderTree/mac/DumpRenderTree.mm:
4339         (drt_NSFontManager_availableFontFamilies):
4340         * WebKitTestRunner/InjectedBundle/mac/ActivateFonts.mm:
4341         (WTR::wtr_NSFontManager_availableFontFamilies):
4342
4343 2012-08-31  Zan Dobersek  <zandobersek@gmail.com>
4344
4345         nrwt: use scm instead of calling svn directly to get the revision in json results generator
4346         https://bugs.webkit.org/show_bug.cgi?id=89616
4347