jhbuild: Use the multiprocessing module to get the number of CPUs.
[WebKit-https.git] / Tools / ChangeLog
1 2012-03-19  Raphael Kubo da Costa  <rakuco@FreeBSD.org>
2
3         jhbuild: Use the multiprocessing module to get the number of CPUs.
4         https://bugs.webkit.org/show_bug.cgi?id=81585
5
6         Reviewed by Martin Robinson.
7
8         Now that support for Python 2.5 has been dropped, we can safely
9         use multiprocessing.cpu_count() to obtain the current number of
10         CPUs in the system instead of running a webkitperl script for
11         that.
12
13         * efl/common.py:
14         (top_level_path):
15         * efl/jhbuildrc:
16         * gtk/common.py:
17         (build_path):
18         * gtk/jhbuildrc:
19
20 2012-03-19  Ojan Vafai  <ojan@chromium.org>
21
22         We should only add a live event listener once, not once per TestSelector.
23         Adding one per TestSelector would cause accordion('option', 'active') to
24         incorrectly be "false" if you examine multiple sets of tests.
25
26         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
27
28 2012-03-19  Dirk Pranke  <dpranke@chromium.org>
29
30         Fix typo introduced in r111293.
31
32         Unreviewed, build fix.
33
34         * Scripts/webkitpy/layout_tests/port/webkit.py:
35         (WebKitDriver._check_for_driver_crash):
36
37 2012-03-19  Dirk Pranke  <dpranke@chromium.org>
38
39         webkitpy: get ServerProcess out of the reportcrash business
40         https://bugs.webkit.org/show_bug.cgi?id=81600
41
42         Reviewed by Adam Barth.
43
44         ServerProcess was half-aware that ReportCrash might run
45         sometimes, and that the process ServerProcess was talking to
46         might have its own crashing subprocesses; neither of these
47         things really worked right and it made the logic convoluted, so
48         this change makes handling crashes completely separate from the
49         server_process code, so that it can focus on just I/O to the
50         subprocess.
51
52         There should be no functional changes resulting from this patch.
53
54         * Scripts/webkitpy/layout_tests/port/server_process.py:
55         (ServerProcess._reset):
56         (ServerProcess._handle_possible_interrupt):
57         (ServerProcess.write):
58         (ServerProcess.read_stdout):
59         (ServerProcess.has_crashed):
60         (ServerProcess._read):
61         (ServerProcess.stop):
62         * Scripts/webkitpy/layout_tests/port/server_process_unittest.py:
63         (TrivialMockPort.check_for_leaks):
64         (TestServerProcess.test_broken_pipe):
65         * Scripts/webkitpy/layout_tests/port/webkit.py:
66         (WebKitPort._read_image_diff):
67         (WebKitDriver.has_crashed):
68         (WebKitDriver._check_for_driver_crash):
69         (WebKitDriver.run_test):
70         (WebKitDriver._read_block):
71         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
72         (MockServerProcess.__init__):
73         (MockServerProcess):
74         (MockServerProcess.has_crashed):
75
76 2012-03-19  Dirk Pranke  <dpranke@chromium.org>
77
78         webkitpy: crashlog parsing is broken
79         https://bugs.webkit.org/show_bug.cgi?id=81586
80
81         Reviewed by Adam Barth.
82
83         Seeking on a file opened through codecs() appears to not work
84         correctly (at least on SL).
85         
86         The code was more complicated than it needed to be, so I have
87         revamped it to just read a full crashlog at once and then look
88         for matches. 
89         
90         I have also added the ability to optionally propagate
91         errors back (which can be helpful to debug races when ReportCrash is
92         still running and you need to see that you couldn't open some
93         files, rather than just ignoring them).
94
95         * Scripts/webkitpy/common/system/crashlogs.py:
96         (CrashLogs.find_newest_log):
97         (CrashLogs._find_newest_log_darwin):
98         * Scripts/webkitpy/common/system/crashlogs_unittest.py:
99         (CrashLogsTest.test_find_log_darwin):
100
101 2012-03-19  Dirk Pranke  <dpranke@chromium.org>
102
103         webkitpy: clean up some port._filesystem references to not reference private members
104         https://bugs.webkit.org/show_bug.cgi?id=81595
105
106         Reviewed by Eric Seidel.
107
108         There should be no functional changes in this patch, just some
109         refactoring. Also, fixes a minor nit in write_image_diff_files(()
110
111         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
112         (SingleTestRunner.__init__):
113         (SingleTestRunner._run_compare_test):
114         (SingleTestRunner._run_rebaseline):
115         (SingleTestRunner._save_baseline_data):
116         (SingleTestRunner._handle_error):
117         (SingleTestRunner._run_reftest):
118         * Scripts/webkitpy/layout_tests/controllers/test_result_writer.py:
119         (write_test_result):
120         (TestResultWriter.__init__):
121         (TestResultWriter._make_output_directory):
122         (TestResultWriter.output_filename):
123         (TestResultWriter._output_testname):
124         (TestResultWriter.write_output_files):
125         (TestResultWriter.write_stderr):
126         (TestResultWriter.write_crash_report):
127         (TestResultWriter.create_text_diff_and_write_result):
128         (TestResultWriter.write_image_diff_files):
129         (copy_file):
130         * Scripts/webkitpy/layout_tests/controllers/test_result_writer_unittest.py:
131         (TestResultWriterTest.test_reftest_diff_image):
132
133 2012-03-19  Dirk Pranke  <dpranke@chromium.org>
134
135         webkitpy: fix nit - incorrect reference to DRT in test_expectations.py
136         https://bugs.webkit.org/show_bug.cgi?id=81598
137
138         Reviewed by Eric Seidel.
139
140         "DumpRenderTree" shouldn't be hard-coded here, since some other
141         process may actually be crashing (WebKitTestRunner, WebProcess, etc.)
142         
143         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
144         (TestExpectations):
145
146 2012-03-19  Dirk Pranke  <dpranke@chromium.org>
147
148         webkitpy: allow instance data in TestFailure objects
149         https://bugs.webkit.org/show_bug.cgi?id=81580
150
151         Reviewed by Tony Chang.
152
153         Current TestFailures are basically enums, but it would be nice,
154         especially for crashes, if they contained the name and pid of
155         the process that crashed.
156
157         * Scripts/webkitpy/layout_tests/models/test_failures.py:
158         (TestFailure.message):
159         (FailureTimeout.message):
160         (FailureCrash):
161         (FailureCrash.__init__):
162         (FailureCrash.message):
163         (FailureMissingResult.message):
164         (FailureTextMismatch.message):
165         (FailureMissingImageHash.message):
166         (FailureMissingImage.message):
167         (FailureImageHashMismatch.message):
168         (FailureImageHashIncorrect.message):
169         (FailureReftestMismatch.message):
170         (FailureReftestMismatchDidNotOccur.message):
171         (FailureReftestNoImagesGenerated.message):
172         (FailureMissingAudio.message):
173         (FailureAudioMismatch.message):
174         * Scripts/webkitpy/layout_tests/models/test_failures_unittest.py:
175         (TestFailuresTest.test_equals):
176         (TestFailuresTest):
177         (TestFailuresTest.test_crashes):
178
179 2012-03-19  Brady Eidson  <beidson@apple.com>
180
181         <rdar://problem/10917120> and https://bugs.webkit.org/show_bug.cgi?id=81066
182         Crash in 3rd party WebKit apps under XHR/Cache code
183
184         Reviewed by Antti Koivisto.
185
186         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
187         * TestWebKitAPI/Tests/mac/MemoryCachePruneWithinResourceLoadDelegate.html: Added.
188         * TestWebKitAPI/Tests/mac/MemoryCachePruneWithinResourceLoadDelegate.mm: Added.
189         (-[MemoryCachePruneTestResourceLoadDelegate webView:identifierForInitialRequest:fromDataSource:]):
190         (-[MemoryCachePruneTestResourceLoadDelegate webView:resource:willSendRequest:redirectResponse:fromDataSource:]):
191         (-[MemoryCachePruneTestResourceLoadDelegate webView:resource:didFinishLoadingFromDataSource:]):
192         (-[MemoryCachePruneTestResourceLoadDelegate webView:resource:didFailLoadingWithError:fromDataSource:]):
193         (TestWebKitAPI):
194         (TestWebKitAPI::TEST):
195
196 2012-03-19  Brady Eidson  <beidson@apple.com>
197
198         <rdar://problem/10848575> and https://bugs.webkit.org/show_bug.cgi?id=81516
199         REGRESSION (r107435) URLs copied from WebKit apps aren't in the right pasteboard format
200
201         Reviewed by Enrica Casucci.
202
203         Bring up a context menu on a link, copy the link, then verify the pasteboard contents
204         are in the correct format.
205
206         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
207         * TestWebKitAPI/Tests/mac/ContextMenuCanCopyURL.html: Added.
208         * TestWebKitAPI/Tests/mac/ContextMenuCanCopyURL.mm: Added.
209         (-[ContextMenuCanCopyURLDelegate webView:didFinishLoadForFrame:]):
210         (TestWebKitAPI::contextMenuCopyLink):
211         (TestWebKitAPI::TEST):
212
213 2012-03-19  Ojan Vafai  <ojan@chromium.org>
214
215         Make the selected item in the accordion a link to the flakiness dashboard
216         https://bugs.webkit.org/show_bug.cgi?id=81571
217
218         Reviewed by Dimitri Glazkov.
219
220         Also, make it so that you can select the text of the link, instead of the current behavior
221         of acting as if you had clicked on the accordion header.
222
223         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
224         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/results.css:
225
226 2012-03-19  Dirk Pranke  <dpranke@chromium.org>
227
228         NRWT runs some tests that are skipped with -i command line option
229         https://bugs.webkit.org/show_bug.cgi?id=81535
230
231         Reviewed by Ojan Vafai.
232
233         This change modifies the interaction of Skipped files and
234         test_expectations files so that entries in Skipped files (and
235         the -i command line) override *everything* in the expectations
236         file.
237
238         Specifically, a directory in a Skipped file will cause all of
239         the tests in the dir to be skipped even if individual tests in
240         the dir are listed in the test_expectations.txt.
241
242         Skipped files also override anything in an overrides files.
243
244         This seems to make more intuitive sense, since if you list
245         something in the Skips file (and even more specify it on the
246         command line) you probably want it to be universally applied.
247
248         Theoretically we could add more precedence levels and have full
249         paths in an expectations file override dirs in a Skipped file
250         (but not the command line), but I don't know yet that that level
251         of complexity is justified.
252         
253         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
254         (TestExpectations.__init__):
255         (TestExpectations._add_skipped_tests):
256         (TestExpectations._add_test):
257         (TestExpectations._add_expectations):
258         (TestExpectations._add_expectation_line):
259         (TestExpectations._already_seen_better_match):
260         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
261         (test_more_specific_override_resets_skip):
262         (SkippedTests):
263         (SkippedTests.get_exp):
264         (SkippedTests.assert_exp):
265         (SkippedTests.test_skipped_tests_work):
266         (SkippedTests.test_duplicate_skipped_test_fails_lint):
267         (SkippedTests.test_skipped_file_overrides_expectations):
268         (SkippedTests.test_skipped_dir_overrides_expectations):
269         (SkippedTests.test_skipped_file_overrides_overrides):
270         (SkippedTests.test_skipped_dir_overrides_overrides):
271
272 2012-03-19  Eric Seidel  <eric@webkit.org>
273
274         Fix WTF header include discipline in Chromium WebKit
275         https://bugs.webkit.org/show_bug.cgi?id=81281
276
277         Reviewed by James Robinson.
278
279         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
280         * Scripts/update-webkit-chromium: Made it possible to run this from a directory other than the root.
281
282 2012-03-19  Andy Chen  <andchen@rim.com>
283
284         [BlackBerry] Add new layout test for text match marker
285         https://bugs.webkit.org/show_bug.cgi?id=81531
286
287         Make the LayoutTestControllerBlackBerry::findString to test the
288         BlackBerry port implementation.
289
290         Reviewed by Rob Buis.
291
292         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
293         (LayoutTestController::findString):
294
295 2012-03-19  Ojan Vafai  <ojan@chromium.org>
296
297         Don't show WONTFIX tests in garden-o-matic
298         https://bugs.webkit.org/show_bug.cgi?id=81554
299
300         Reviewed by Dimitri Glazkov.
301
302         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results.js:
303         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results_unittests.js:
304
305 2012-03-19  Chris Fleizach  <cfleizach@apple.com>
306
307         platform/mac/accessibility/selected-rows-table.html failing on Lion Intel Debug WebKit2 testers
308         https://bugs.webkit.org/show_bug.cgi?id=81530
309
310         Reviewed by Jessie Berlin.
311
312         Add rowAtIndex to WebKitTestRunner.
313
314         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.cpp:
315         (WTR::AccessibilityUIElement::rowAtIndex):
316         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.h:
317         (AccessibilityUIElement):
318         * WebKitTestRunner/InjectedBundle/Bindings/AccessibilityUIElement.idl:
319         * WebKitTestRunner/InjectedBundle/mac/AccessibilityUIElementMac.mm:
320         (WTR):
321         (WTR::AccessibilityUIElement::rowAtIndex):
322
323 2012-03-19  Ojan Vafai  <ojan@chromium.org>
324
325         Change garden-o-matic All Failures tab to Expected Failures
326         https://bugs.webkit.org/show_bug.cgi?id=81540
327
328         Reviewed by Dimitri Glazkov.
329
330         Also, fix some broken unittests that had just gotten stale.
331
332         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
333         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers/rebaseline.js:
334         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
335         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model.js:
336         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results.js:
337         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results_unittests.js:
338         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
339         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures_unittests.js:
340         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
341         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:
342
343 2012-03-19  Hao Zheng  <zhenghao@chromium.org>
344
345         [chromium] DRT crashes at shutdown.
346         https://bugs.webkit.org/show_bug.cgi?id=81504
347
348         Reviewed by James Robinson.
349
350         WebCompositor must be destroyed after WebView is properly destroyed,
351         or some compositor objects like CCLayerTreeHost may not be destroyed.
352         However, it is really hard to make the destruction order correct
353         explicitly in ~TestShell. So move it out to DumpRenderTree.
354
355         * DumpRenderTree/chromium/DumpRenderTree.cpp:
356         (main):
357         * DumpRenderTree/chromium/TestShell.cpp:
358         (TestShell::~TestShell):
359
360 2012-03-19  Dirk Pranke  <dpranke@chromium.org>
361
362         add dpranke to webkitpy watchlist
363         https://bugs.webkit.org/show_bug.cgi?id=81433
364
365         Reviewed by Benjamin Poulain.
366
367         * Scripts/webkitpy/common/config/watchlist:
368
369 2012-03-19  Dirk Pranke  <dpranke@chromium.org>
370
371         test-webkitpy is failing on SL/Python 2.6.1
372         https://bugs.webkit.org/show_bug.cgi?id=81533
373
374         Reviewed by Ojan Vafai.
375
376         It appears that itertools.combinations() is buggy in 2.6.1; this
377         backports the implementation from 2.7.
378
379         * Scripts/webkitpy/layout_tests/models/test_configuration.py:
380         (TestConfigurationConverter.collapse_macros):
381         (TestConfigurationConverter):
382         (TestConfigurationConverter.combinations):
383
384 2012-03-19  David Michael Barr  <davidbarr@google.com>
385
386         Optimize the cleanup sequence in the start-queue.sh script
387         https://github.com/abarth/webkit/pull/1
388
389         Reviewed by Adam Barth.
390
391         Updating the working copy is an expensive operation
392         for a repository the size of WebKit.
393         Attempt to reset only to the target revision.
394
395         * EWSTools/start-queue.sh:
396
397 2012-03-19  Raphael Kubo da Costa  <kubo@profusion.mobi>
398
399         Adjust architecture type of the EFL build bot.
400         https://bugs.webkit.org/show_bug.cgi?id=81452
401
402         Reviewed by Tony Chang.
403
404         * BuildSlaveSupport/build.webkit.org-config/config.json: The
405         buildslave is a 64-bit machine, therefore describe it as being
406         x86_64, not i386.
407
408 2012-03-19  Alexander Færøy  <alexander.faeroy@nokia.com>
409
410         [Qt] printLoadedUrls prints "Object" instead of the actual URL.
411         https://bugs.webkit.org/show_bug.cgi?id=81514
412
413         Reviewed by Simon Hausmann.
414
415         * MiniBrowser/qt/qml/BrowserWindow.qml:
416
417 2012-03-19  Raphael Kubo da Costa  <rakuco@FreeBSD.org>
418
419         [EFL] jhbuild: Set CMAKE_PREFIX_PATH in jhbuildrc.
420         https://bugs.webkit.org/show_bug.cgi?id=81491
421
422         Reviewed by Philippe Normand.
423
424         For the find_path(), find_library() etc CMake calls to find the
425         dependencies built with jhbuild, CMAKE_PREFIX_PATH must be properly
426         set, otherwise CMake cannot know the dependencies exist and will only
427         look in the standard directories.
428
429         * efl/jhbuildrc: Set the CMAKE_PREFIX_PATH environment variable with
430         the jhbuild Root directory.
431
432 2012-03-19  Dominik Röttsches  <dominik.rottsches@linux.intel.com>
433
434         [EFL] Add XDG env vars to jhbuildrc
435         https://bugs.webkit.org/show_bug.cgi?id=81510
436
437         Needed to ensure correct glib behaviour in jhbuild environment.
438
439         Reviewed by Philippe Normand.
440
441         * efl/jhbuildrc:
442
443 2012-03-19  Allan Sandfeld Jensen  <allan.jensen@nokia.com>
444
445         Select best target for tap gesture.
446         https://bugs.webkit.org/show_bug.cgi?id=78801
447
448         Reviewed by Kenneth Rohde Christiansen.
449
450         Add TOUCH_ADJUSTMENT to enabled features.
451
452         * qmake/mkspecs/features/features.prf:
453
454 2012-03-19  Robert Kroeger  <rjkroege@chromium.org>
455
456         [chromium] synthesize wheel events for fling on main thread
457         https://bugs.webkit.org/show_bug.cgi?id=81458
458
459         Added the ability to dispatch GestureFlingStart and GestureFlingCancel from EventSender.
460
461         Reviewed by James Robinson.
462
463         * DumpRenderTree/chromium/EventSender.cpp:
464         (EventSender::EventSender):
465         (EventSender::gestureFlingCancel):
466         (EventSender::gestureFlingStart):
467         * DumpRenderTree/chromium/EventSender.h:
468         (EventSender):
469
470 2012-03-19  Alexander Færøy  <alexander.faeroy@nokia.com>
471
472         [Qt] Add command line option to set the user agent in the MiniBrowser
473         https://bugs.webkit.org/show_bug.cgi?id=81508
474
475         Reviewed by Simon Hausmann.
476
477         * MiniBrowser/qt/BrowserWindow.cpp:
478         (BrowserWindow::BrowserWindow):
479         (BrowserWindow::webViewExperimental):
480         * MiniBrowser/qt/BrowserWindow.h:
481         (BrowserWindow):
482         * MiniBrowser/qt/MiniBrowserApplication.cpp:
483         (printHelp):
484         (MiniBrowserApplication::handleUserOptions):
485         * MiniBrowser/qt/MiniBrowserApplication.h:
486         (WindowOptions::WindowOptions):
487         (WindowOptions::setUserAgent):
488         (WindowOptions::userAgent):
489         (WindowOptions):
490
491 2012-03-19  Alexander Færøy  <alexander.faeroy@nokia.com>
492
493         [Qt] Remove dead user agent code from the MiniBrowser.
494
495         Reviewed by Simon Hausmann.
496
497         * MiniBrowser/qt/BrowserWindow.cpp:
498         * MiniBrowser/qt/BrowserWindow.h:
499         (BrowserWindow):
500         * MiniBrowser/qt/MiniBrowser.qrc:
501         * MiniBrowser/qt/useragentlist.txt: Removed.
502
503 2012-03-19  Raphael Kubo da Costa  <rakuco@FreeBSD.org>
504
505         jhbuild: Call jhbuild's autogen.sh via bash.
506         https://bugs.webkit.org/show_bug.cgi?id=81485
507
508         Reviewed by Philippe Normand.
509
510         jhbuild's autogen.sh uses "/bin/bash" in the shebang line, however
511         this is usually true only on Linux systems, and even then bash for
512         some reason might not be available in that location. It is safer to
513         call bash directly and pass autogen.sh to it instead.
514
515         * jhbuild/jhbuild-wrapper:
516         (install_jhbuild):
517
518 2012-03-19  Philippe Normand  <pnormand@igalia.com>
519
520         [GTK] Prefix run-launcher with run-with-jhbuild
521         https://bugs.webkit.org/show_bug.cgi?id=81503
522
523         Reviewed by Hajime Morita.
524
525         * Scripts/run-launcher:
526
527 2012-03-19  Dominik Röttsches  <dominik.rottsches@linux.intel.com>
528
529         [EFL] Prefix run-launcher with run-with-jhbuild
530         https://bugs.webkit.org/show_bug.cgi?id=80026
531
532         Reviewed by Hajime Morita.
533
534         * Scripts/run-launcher:
535
536 2012-03-18  Csaba Osztrogonác  <ossy@webkit.org>
537
538         [Qt] Unreviewed. Disable -Werror on x86 until proper fix.
539         https://bugs.webkit.org/show_bug.cgi?id=81498
540
541         * qmake/mkspecs/features/unix/default_post.prf:
542
543 2012-03-17  Raphael Kubo da Costa  <rakuco@FreeBSD.org>
544
545         Unreviewed; switch to using my FreeBSD.org email address, the
546         other one is going to disappear soon.
547
548         * Scripts/webkitpy/common/config/committers.py:
549         * Scripts/webkitpy/common/config/watchlist:
550
551 2012-03-17  Raphael Kubo da Costa  <kubo@profusion.mobi>
552
553         Unreviewed; fix webkitpy unit test failures after r111116.
554
555         It turns out that cmd_line in the patch in bug 80025 had the right
556         signature when it was submitted, but r109242 added a few more
557         required parameters, and since the method (and EflDriver) did not
558         exist at the time we ended up committing a broken method.
559
560         * Scripts/webkitpy/layout_tests/port/efl.py:
561         (EflDriver.cmd_line): Add the `pixel_tests' and `per_test_args'
562         arguments.
563
564 2012-03-17  Charles Wei  <charles.wei@torchmobile.com.cn>
565
566         [BlackBerry] Enable Web Timing for performance profiling and improvement.
567         https://bugs.webkit.org/show_bug.cgi?id=81085
568
569         Reviewed by Rob Buis.
570
571         * Scripts/build-webkit:
572
573 2012-03-17  Dominik Röttsches  <dominik.rottsches@linux.intel.com>
574
575         Add a dependencies installation step for EFL to buildbot master.cfg
576         https://bugs.webkit.org/show_bug.cgi?id=81337
577
578         Reviewed by Gustavo Noronha Silva.
579
580         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
581         (InstallEflDependencies):
582         (Factory.__init__):
583
584 2012-03-17  Dominik Röttsches  <dominik.rottsches@linux.intel.com>
585
586         [EFL] Prefixing run-webkit-tests with run-with-jhbuild
587         https://bugs.webkit.org/show_bug.cgi?id=80025
588
589         Reviewed by Gustavo Noronha Silva.
590
591         * Scripts/webkitpy/layout_tests/port/efl.py:
592         (EflDriver):
593         (EflDriver.cmd_line):
594         (EflPort._driver_class):
595
596 2012-03-17  Dominik Röttsches  <dominik.rottsches@linux.intel.com>
597
598         [EFL] Add and use run-with-jhbuild and update-webkitefl-libs scripts for EFL
599         https://bugs.webkit.org/show_bug.cgi?id=79904
600
601         build-webkit --efl will use jhbuild for fetching and building dependencies.
602
603         Reviewed by Gustavo Noronha Silva.
604
605         * Scripts/update-webkitefl-libs: Added.
606         * Scripts/webkitdirs.pm:
607         (jhbuildWrapperPrefixIfNeeded):
608         (generateBuildSystemFromCMakeProject):
609         (buildCMakeGeneratedProject):
610         * efl/common.py: Added.
611         (script_path):
612         (top_level_path):
613         (number_of_cpus):
614         * efl/jhbuild.modules: Added.
615         * efl/jhbuildrc: Added.
616         * efl/run-with-jhbuild: Added.
617
618 2012-03-16  Stephanie Lewis  <slewis@apple.com>
619
620         https://bugs.webkit.org/show_bug.cgi?id=81065
621         <rdar://problem/10944309> CrashTracer: [USER] 111 crashes in WebProcess at WebKitTestRunnerInjectedBundle: WTR::InjectedBundle::done + 142
622         Fix crash preventing WebKitTestRunner from working in Release.  
623         WebKitTestRunner used ostringstream to collect output from the test.  ostringstream is incompatible with our
624         FastMalloc overrides.  When copying strings out ostringstream a new string is allocated with FastMalloc because
625         the allocation is inlined.  When that string is freed is uses the system malloc because that call is not inlined.
626
627         I removed ostringstream in favor of StringBuilder.  
628
629         Properly set NDEBUG on Release builds so we don't crash when when we free the strings StringBuilder created.
630
631         Reviewed by Geoff Garen.
632
633         Add DEBUG_DEFINES to set NDEBUG on release builds.
634         * WebKitTestRunner/Configurations/Base.xcconfig:
635         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
636
637         Remove ostringstream member in favor of StringBuilder.
638         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
639         (WTR::InjectedBundle::initialize):
640         (WTR::InjectedBundle::beginTesting):
641         (WTR::InjectedBundle::done):
642         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
643         (WTF):
644         (WTR::InjectedBundle::stringBuilder):
645         (InjectedBundle):
646
647         Append output to StringBuilder.
648         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
649         (WTR::hasPrefix):
650         (WTR::dumpPath):
651         (WTR::rangeToStr):
652         (WTR::styleDecToStr):
653         (WTR::frameToStr):
654         (WTR::dumpFrameScrollPosition):
655         (WTR::dumpFrameText):
656         (WTR::dumpDescendantFramesText):
657         (WTR::InjectedBundlePage::dump):
658         (WTR::InjectedBundlePage::didReceiveTitleForFrame):
659         (WTR::InjectedBundlePage::didFinishDocumentLoadForFrame):
660         (WTR::InjectedBundlePage::willSendRequestForFrame):
661         (WTR::InjectedBundlePage::decidePolicyForNavigationAction):
662         (WTR::lastFileURLPathComponent):
663         (WTR::InjectedBundlePage::willAddMessageToConsole):
664         (WTR::InjectedBundlePage::willSetStatusbarText):
665         (WTR::InjectedBundlePage::willRunJavaScriptAlert):
666         (WTR::InjectedBundlePage::willRunJavaScriptConfirm):
667         (WTR::InjectedBundlePage::willRunJavaScriptPrompt):
668         (WTR::InjectedBundlePage::shouldBeginEditing):
669         (WTR::InjectedBundlePage::shouldEndEditing):
670         (WTR::InjectedBundlePage::shouldInsertNode):
671         (WTR::InjectedBundlePage::shouldInsertText):
672         (WTR::InjectedBundlePage::shouldDeleteRange):
673         (WTR::InjectedBundlePage::shouldChangeSelectedRange):
674         (WTR::InjectedBundlePage::shouldApplyStyle):
675         (WTR::InjectedBundlePage::didBeginEditing):
676         (WTR::InjectedBundlePage::didEndEditing):
677         (WTR::InjectedBundlePage::didChange):
678         (WTR::InjectedBundlePage::didChangeSelection):
679         (WTR::InjectedBundlePage::supportsFullScreen):
680         (WTR::InjectedBundlePage::enterFullScreenForElement):
681         (WTR::InjectedBundlePage::exitFullScreenForElement):
682         (WTR::InjectedBundlePage::beganEnterFullScreen):
683         (WTR::InjectedBundlePage::beganExitFullScreen):
684         (WTR::dumpBackForwardListItem):
685         (WTR::InjectedBundlePage::dumpBackForwardList):
686         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
687         (WTR::LayoutTestController::waitToDumpWatchdogTimerFired):
688         (WTR::LayoutTestController::dumpConfigurationForViewport):
689         
690         Convert between WTF::String and WKStringRef.
691         * WebKitTestRunner/StringFunctions.h:
692         (WTR::toWTFString):
693         (WTR):
694         * WebKitTestRunner/TestInvocation.cpp:
695         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle):
696
697 2012-03-16  Eric Seidel  <eric@webkit.org>
698
699         Add a land-from-url command
700         https://bugs.webkit.org/show_bug.cgi?id=81411
701
702         Reviewed by Adam Barth.
703
704         Very basic so far.  The goal is to make it landing smarter
705         so that users don't have to bother to parse out the ids out of the urls.
706
707         * Scripts/webkitpy/common/config/urls.py:
708         (parse_bug_id):
709         (parse_attachment_id):
710         * Scripts/webkitpy/common/config/urls_unittest.py:
711         (URLsTest.test_parse_bug_id):
712         (URLsTest):
713         (URLsTest.test_parse_attachment_id):
714         * Scripts/webkitpy/tool/commands/download.py:
715         (ProcessURLsMixin):
716         (ProcessURLsMixin._fetch_list_of_patches_to_process):
717         (LandFromURL):
718         * Scripts/webkitpy/tool/commands/download_unittest.py:
719         (test_land_from_bug):
720         (test_land_from_url):
721
722 2012-03-16  Dave Tharp  <dtharp@codeaurora.org>
723
724         build-webkit launcher instructions use wrong relative path for Qt
725         https://bugs.webkit.org/show_bug.cgi?id=81357
726
727         Reviewed by Eric Seidel.
728
729         Ensure we are at the proper directory before generating run-launcher
730         instructions.
731
732         * Scripts/build-webkit:
733
734 2012-03-16  Brady Eidson  <beidson@apple.com>
735
736         <rdar://problem/11027997> and https://bugs.webkit.org/show_bug.cgi?id=81412
737         REGRESSION (r107435) Copy a link and paste to Mail: Nothing is pasted
738
739         Reviewed by Geoff Garen.
740
741         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
742         * TestWebKitAPI/Tests/mac/WebViewCanPasteURL.mm: Added.
743
744 2012-03-16  Ojan Vafai  <ojan@chromium.org>
745
746         Mark wontfix tests in full_results.json so we can ignore them in garden-o-matic
747         https://bugs.webkit.org/show_bug.cgi?id=81407
748
749         Reviewed by Adam Barth.
750
751         * Scripts/webkitpy/layout_tests/controllers/manager.py:
752         (summarize_results):
753         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
754         (ResultSummaryTest.summarized_results):
755         (ResultSummaryTest.test_no_svn_revision):
756         (ResultSummaryTest.test_svn_revision):
757         (ResultSummaryTest):
758         (ResultSummaryTest.test_summarized_results_wontfix):
759
760 2012-03-16  Ojan Vafai  <ojan@chromium.org>
761
762         Have webkit-patch rebaseline-test update test_expectations.txt
763         https://bugs.webkit.org/show_bug.cgi?id=81054
764
765         Recommit http://trac.webkit.org/changeset/110783 now that
766         http://trac.webkit.org/changeset/111034 has landed.
767
768         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
769         (TestExpectations.remove_configuration_from_test):
770         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
771         (RemoveConfigurationsTest):
772         (RemoveConfigurationsTest.test_remove):
773         (test_remove_line):
774         * Scripts/webkitpy/tool/commands/rebaseline.py:
775         (RebaselineTest._update_expectations_file):
776         (RebaselineTest._rebaseline_test_and_update_expectations):
777         (RebaselineTest):
778         (RebaselineTest.execute):
779         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
780         (TestRebaseline.test_rebaseline_updates_expectations_file_noop):
781         (test_rebaseline_updates_expectations_file):
782
783 2012-03-15  Ojan Vafai  <ojan@chromium.org>
784
785         Specifier collapsing when writing test expectations lines gets a number of cases wrong
786         https://bugs.webkit.org/show_bug.cgi?id=81309
787
788         Reviewed by Dimitri Glazkov.
789
790         I've run this over all the lines in the current Chromium test_expectations.txt file,
791         so I'm relatively confident we now cover all the cases.
792
793         * Scripts/webkitpy/layout_tests/models/test_configuration.py:
794         (TestConfigurationConverter.__init__):
795         (TestConfigurationConverter.collapse_macros):
796         (TestConfigurationConverter.collapse_macros.collapse_individual_specifier_set):
797         (TestConfigurationConverter):
798         (TestConfigurationConverter.intersect_combination):
799         (TestConfigurationConverter.symmetric_difference):
800         (TestConfigurationConverter.to_specifiers_list):
801         (TestConfigurationConverter.to_specifiers_list.try_collapsing):
802         (TestConfigurationConverter.to_specifiers_list.try_abbreviating):
803         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py:
804         (make_mock_all_test_configurations_set):
805         (TestConfigurationConverterTest.test_symmetric_difference):
806         (TestConfigurationConverterTest.test_to_config_set):
807         (TestConfigurationConverterTest.test_macro_expansion):
808         (TestConfigurationConverterTest.test_to_specifier_lists):
809         (TestConfigurationConverterTest.test_converter_macro_collapsing):
810         * Scripts/webkitpy/layout_tests/port/chromium.py:
811         (ChromiumPort):
812         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
813         (ChromiumAndroidPort.__init__):
814         The android port uses "arm" as it's architecture, which is technically correct,
815         but considerably complicates making collapsing work. We probably should kill
816         the concept of architecture entirely. The benefits are not worth the code
817         complexity.
818
819         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
820         (ChromiumPortTest.test_all_test_configurations):
821         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
822         (TestRebaseline.test_rebaseline_updates_expectations_file_noop):
823         (test_rebaseline_updates_expectations_file):
824
825 2012-03-16  Dinu Jacob  <dinu.jacob@nokia.com>
826
827         [Qt][Wk2] Assertion Failure and crash on file upload
828         https://bugs.webkit.org/show_bug.cgi?id=80854
829
830         Reviewed by Simon Hausmann.
831
832         Added filePicker to WebView using experimental API.
833
834         * MiniBrowser/qt/MiniBrowser.qrc:
835         * MiniBrowser/qt/icons/folder.png: Added.
836         * MiniBrowser/qt/icons/titlebar.png: Added.
837         * MiniBrowser/qt/icons/up.png: Added.
838         * MiniBrowser/qt/qml/BrowserWindow.qml:
839         * MiniBrowser/qt/qml/FilePicker.qml: Added.
840
841 2012-03-16  Mahesh Kulkarni  <mahesh.kulkarni@nokia.com>
842
843         Updating email for committer.py script.
844
845         * Scripts/webkitpy/common/config/committers.py:
846
847 2012-03-16  Robert Kroeger  <rjkroege@chromium.org>
848
849         [chromium] DRT: Add support for sending gesture scrolls without a begin
850         https://bugs.webkit.org/show_bug.cgi?id=81292
851
852         Reviewed by Kent Tamura.
853
854         * DumpRenderTree/chromium/EventSender.cpp:
855         (EventSender::EventSender):
856         (EventSender::gestureScrollFirstPoint):
857         * DumpRenderTree/chromium/EventSender.h:
858         (EventSender):
859
860 2012-03-02  Carlos Garcia Campos  <cgarcia@igalia.com> and Sergio Villar Senin  <svillar@igalia.com>
861
862         [GTK] WebKitIconDatabase doesn't keep icons cached
863         https://bugs.webkit.org/show_bug.cgi?id=56200
864
865         Reviewed by Martin Robinson.
866
867         Add support for favicons in GtkLauncher.
868
869         * GtkLauncher/main.c:
870         (activateUriEntryCb):
871         (iconLoadedCb):
872         (createBrowser):
873
874 2012-03-16  Adam Barth  <abarth@webkit.org>
875
876         Remove sheriff-bot's last-green-revision command
877         https://bugs.webkit.org/show_bug.cgi?id=81314
878
879         Reviewed by Eric Seidel.
880
881         This command is a cruel joke.  WebKit doesn't really ever have a
882         last-green-revision.
883
884         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
885         (BuildBot._find_green_revision):
886         * Scripts/webkitpy/common/net/buildbot/buildbot_mock.py:
887         (MockBuildBot.builder_statuses):
888         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
889         (test_find_green_revision):
890         * Scripts/webkitpy/tool/bot/irc_command.py:
891         (IRCCommand.execute):
892         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
893         (SheriffIRCBotTest.test_help):
894         * Scripts/webkitpy/tool/commands/queries.py:
895         (PatchesToReview.execute):
896
897 2012-03-16  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
898
899         [Qt] Make gccdepends test aware of broken icecream deps generation
900
901         Icecream seems to build the files without passing -o to gcc with
902         the original destination path, which means the target name in the
903         dependency file is wrong.
904
905         Also, echo something to the final file instead of just touching it,
906         so it's not deleted on every build.
907
908         Reviewed by Simon Hausmann.
909
910         * qmake/config.tests/gccdepends/gccdepends.pro:
911
912 2012-03-16  Kihong Kwon  <kihong.kwon@samsung.com>
913
914         Support for Battery Status API
915         https://bugs.webkit.org/show_bug.cgi?id=62698
916
917         Add include directory and file.
918         http://www.w3.org/TR/battery-status/
919
920         Reviewed by Adam Barth.
921
922         * Scripts/build-webkit:
923
924 2012-03-16  Kishore Bolisetty  <kbolisetty@innominds.com>
925
926         check-webkit-style fails on LayoutTests/platform/efl/test_expectations.txt
927         https://bugs.webkit.org/show_bug.cgi?id=81007
928
929         Reviewed by Adam Barth.
930
931         A dedidcated build slave is available for efl on build.webkit.org. 
932         This change add's an entry for the elf port in list of available port builders.
933         This is required because test_expectations.py when triggered, looks for a valid port object for the specified port. 
934
935         * Scripts/webkitpy/layout_tests/port/builders.py:
936
937 2012-03-16  Adam Barth  <abarth@webkit.org>
938
939         [webkitpy] StatusServer.next_work_item has an incorrectly named local variable
940         https://bugs.webkit.org/show_bug.cgi?id=81302
941
942         Reviewed by Eric Seidel.
943
944         This is a trivial patch, but the variable name turned up in a grep of
945         webkitpy and looks like copy pasta.
946
947         * Scripts/webkitpy/common/net/statusserver.py:
948         (StatusServer.next_work_item):
949
950 2012-03-16  Andrew Lo  <anlo@rim.com>
951
952         [BlackBerry] Enable PAGE_VISIBILITY_API
953         https://bugs.webkit.org/show_bug.cgi?id=81257
954
955         Enable Page Visibility API for BlackBerry.
956
957         Reviewed by Rob Buis.
958
959         * Scripts/build-webkit:
960
961 2012-03-16  Zan Dobersek  <zandobersek@gmail.com>
962
963         [Gtk] Add computedStyleIncludingVisitedInfo to DumpRenderTreeSupport
964         https://bugs.webkit.org/show_bug.cgi?id=81232
965
966         Reviewed by Philippe Normand.
967
968         Use the newly-added helper method in DumpRenderTreeSupportGtk.
969
970         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
971         (LayoutTestController::computedStyleIncludingVisitedInfo):
972
973 2012-03-15  Eric Seidel  <eric@webkit.org>
974
975         Move parse_bug_id into config/urls in preparation for re-use elsewhere.
976         https://bugs.webkit.org/show_bug.cgi?id=81313
977
978         Reviewed by Adam Barth.
979
980         * Scripts/webkitpy/common/checkout/changelog.py:
981         (parse_bug_id_from_changelog):
982         * Scripts/webkitpy/common/config/urls.py:
983         (parse_bug_id):
984         * Scripts/webkitpy/common/config/urls_unittest.py: Copied from Tools/Scripts/webkitpy/common/config/urls.py.
985         (URLsTest):
986         (URLsTest.test_parse_bug_id):
987         * Scripts/webkitpy/common/net/bugzilla/bugzilla_unittest.py:
988         (test_parse_bug_id):
989         * Scripts/webkitpy/tool/bot/irc_command.py:
990         (_post_error_and_check_for_bug_url):
991         * Scripts/webkitpy/tool/bot/sheriff.py:
992         (Sheriff.post_rollout_patch):
993         (Sheriff.post_chromium_deps_roll):
994
995 2012-03-15  Adam Barth  <abarth@webkit.org>
996
997         The commit-queue should fast-track patches that have already passed the testing EWS bots
998         https://bugs.webkit.org/show_bug.cgi?id=81305
999
1000         Reviewed by Eric Seidel.
1001
1002         If a patch has already passed the testing EWS bots (currently only the
1003         chromium-ews bot), we should be able to land the patch without running
1004         the tests again.  This patch is somewhat risky in that the tests might
1005         have been broken between when the EWS bots ran and when the
1006         commit-queue is processing the patch.  If that turns out to be a
1007         problem in practice, we'll tighten up this fast-track in some way,
1008         likely by making sure that the pass status is recent.
1009
1010         * Scripts/webkitpy/tool/bot/commitqueuetask.py:
1011         (CommitQueueTaskDelegate.did_pass_testing_ews):
1012         (CommitQueueTask._did_pass_tests_recently):
1013         (CommitQueueTask.run):
1014         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
1015         (MockCommitQueue.did_pass_testing_ews):
1016         (test_fast_success_case):
1017         * Scripts/webkitpy/tool/commands/queues.py:
1018         (CommitQueue.did_pass_testing_ews):
1019         * Scripts/webkitpy/tool/commands/queues_unittest.py:
1020         (test_did_pass_testing_ews):
1021
1022 2012-03-15  Kentaro Hara  <haraken@chromium.org>
1023
1024         [PerformanceTests] run-perf-tests should output correct units
1025         https://bugs.webkit.org/show_bug.cgi?id=78303
1026
1027         Reviewed by Ryosuke Niwa.
1028
1029         Previously run-perf-tests just supported "ms" for units.
1030         Consequently, Dromaeo perf tests had been reporting runs/s as "ms".
1031         This patch fixes run-perf-tests so that they can support custom units.
1032         The default unit is "ms".
1033
1034         The test result looks like this:
1035
1036         $ ./Tools/Scripts/run-perf-tests Dromaeo/dom-attr.html
1037         RESULT Dromaeo: dom-attr= 6465.9525483 runs/s
1038         median= 0.0 runs/s, stdev= 24.2983433436 runs/s, min= 6435.87649402 runs/s, max= 6515.63693392 runs/s
1039
1040         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1041         (PerfTestsRunner._process_parser_test_result):
1042
1043 2012-03-15  Brent Fulgham  <bfulgham@webkit.org>
1044
1045         [WinCairo] Unreviewed build change after wtf path changes.
1046
1047         * DumpRenderTree/win/ImageDiffWinCairoCommon.vsprops: Updated include
1048         directories to look at the private header directory to find 'wtf'.
1049
1050 2012-03-15  Kent Tamura  <tkent@chromium.org>
1051
1052         REGRESSION(r110595): webkit-patch suggest-reviewers doesn't work
1053         https://bugs.webkit.org/show_bug.cgi?id=81293
1054
1055         Reviewed by Adam Barth.
1056
1057         r110595 introduced new format of the author field of ChangeLog entry:
1058         "Adam Barth  <abarth@webkit.org> && Benjamin Poulain  <bpoulain@apple.com>"
1059         Updte chanelog.py to support it.
1060
1061         * Scripts/webkitpy/common/checkout/changelog.py:
1062         (ChangeLogEntry._split_contributor_names):
1063         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
1064         (test_parse_authors):
1065
1066 2012-03-15  Lynn Neir  <lynn.neir@skype.net>
1067
1068         Implement LayoutTestController::setValueForUser on Windows https://bugs.webkit.org/show_bug.cgi?id=55834
1069
1070         Implemented setValueForUser in both DumpRenderTree and WebKitTestRunner.
1071
1072         Reviewed by Jessie Berlin.
1073
1074         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
1075         (LayoutTestController::setValueForUser):
1076         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
1077         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
1078         (WTR::LayoutTestController::setValueForUser):
1079         (WTR):
1080         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
1081         (LayoutTestController):
1082
1083 2012-03-15  Adam Barth  <abarth@webkit.org>
1084
1085         Update the start-queue.sh script to match what we're using on the bots
1086         https://bugs.webkit.org/show_bug.cgi?id=81277
1087
1088         Reviewed by Eric Seidel.
1089
1090         Over time we've added more magic git commands to clean up the local repo after
1091         it has been borked in various ways. The idea behind this sequence is to clear
1092         out any changes in the working copy and then to rebuild the master branch from
1093         origin/master.
1094
1095         This patch also removes the watchdog timer that we've stopped using on the
1096         bots.
1097
1098         This patch was reviewed on <https://github.com/eseidel/webkit/pull/1>.
1099         Note: Eric and I are experimenting with GitHub to see if it might be
1100         useful for WebKit.  If you have thoughts on this topic, please feel
1101         free to grab Eric or me on #webkit to chat.  I assume that he and I are
1102         the only ones who care about this file so using this alternative
1103         process for this patch shouldn't cause any trouble.
1104
1105         * EWSTools/start-queue.sh:
1106
1107 2012-03-15  Adam Barth  <abarth@webkit.org>
1108
1109         The commit-queue can hang when the test_expectations.txt style check fails during commit
1110         https://bugs.webkit.org/show_bug.cgi?id=81251
1111
1112         Reviewed by Eric Seidel.
1113
1114         We forgot to check the --non-interactive flag, so we end up prompting
1115         the "user" who doesn't exist on the bot.  We might consider teaching
1116         the User class about non-interactive if we have this sort of bug again.
1117
1118         * Scripts/webkitpy/tool/steps/commit.py:
1119         (Commit.options):
1120         (Commit._check_test_expectations):
1121         (Commit.run):
1122         * Scripts/webkitpy/tool/steps/commit_unittest.py:
1123         (CommitTest.test_check_test_expectations):
1124
1125 2012-03-15  Jessie Berlin  <jberlin@apple.com>
1126
1127         fast/dom/Window/window-properties.html and fast/dom/prototype-inheritance-2.html failing on
1128         Mac WK2 because WebAudio is not enabled
1129         https://bugs.webkit.org/show_bug.cgi?id=81259
1130
1131         Reviewed by Dan Bernstein.
1132
1133         Enable WebAudio so that it appears in the window object in the tests.
1134         The webaudio tests are all still skipped in the Mac Skipped list.
1135
1136         * WebKitTestRunner/TestController.cpp:
1137         (WTR::TestController::resetStateToConsistentValues):
1138
1139 2012-03-15  David Levin  <levin@chromium.org>
1140
1141         Various autoresize fixes.
1142         https://bugs.webkit.org/show_bug.cgi?id=81199
1143
1144         Reviewed by Darin Fisher.
1145
1146         * DumpRenderTree/chromium/LayoutTestController.cpp:
1147         (LayoutTestController::reset): Stop using the deprecated api that was removed.
1148
1149 2012-03-15  Carlos Garcia Campos  <cgarcia@igalia.com>
1150
1151         [GTK] Actually use --release and --debug command line options in run-gtk-test
1152         https://bugs.webkit.org/show_bug.cgi?id=81234
1153
1154         Reviewed by Philippe Normand.
1155
1156         Actually handle --release and --debug command line options to
1157         create the build directory, falling back to current hack to detect
1158         whether it's a Release or Debug build when both options are missed.
1159
1160         * Scripts/run-gtk-tests:
1161         (TestRunner.__init__): Receive options too.
1162         (TestRunner._get_top_level_directory): Helper function to get the
1163         top level directory.
1164         (TestRunner._get_build_directory): Helper function to get the
1165         build directory depending on --release/--debug command line
1166         options or gessing it if both options are missing.
1167         (TestRunner._ensure_accessibility_daemon_is_running): Use
1168         self-options now.
1169         (TestRunner.run): Ditto.
1170         (TestRunner.run.run_tests): Pass options to the constructor
1171         instead of run().
1172
1173 2012-03-15  Carlos Garcia Campos  <cgarcia@igalia.com>
1174
1175         [GTK] Run unit tests with --verbose in the bots
1176         https://bugs.webkit.org/show_bug.cgi?id=81230
1177
1178         Reviewed by Martin Robinson.
1179
1180         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
1181         (RunGtkAPITests): Add --verbose to run-gtk-test command.
1182
1183 2012-03-15  Kevin Ollivier  <kevino@theolliviers.com>
1184
1185         Move wx port to using export macros
1186         https://bugs.webkit.org/show_bug.cgi?id=77279
1187
1188         Reviewed by Hajime Morita.
1189
1190         * waf/build/settings.py:
1191         (common_configure):
1192
1193 2012-03-15  Philippe Normand  <pnormand@igalia.com>
1194
1195         Unreviewed, GTK build fix attempt after r110845.
1196
1197         * Scripts/run-gtk-tests: Added --release and --debug options.
1198
1199 2012-03-15  Dominik Röttsches  <dominik.rottsches@linux.intel.com>
1200
1201         Parametrize run-with-jhbuild and update-webkitgtk-libs with platform --gtk/--efl
1202         https://bugs.webkit.org/show_bug.cgi?id=79673
1203
1204         Preparatory work for bringing jhbuild enabled build to EFL port.
1205
1206         Reviewed by Philippe Normand.
1207
1208         * Scripts/update-webkit-libs-jhbuild: Copied from Tools/Scripts/update-webkitgtk-libs.
1209         * Scripts/update-webkitgtk-libs:
1210         * gtk/run-with-jhbuild:
1211         * jhbuild/jhbuild-wrapper: Copied from Tools/gtk/run-with-jhbuild.
1212         (top_level_path):
1213         (jhbuild_installed):
1214         (jhbuild_cloned):
1215         (jhbuild_at_expected_revision):
1216         (update_jhbuild):
1217         (clone_jhbuild):
1218         (install_jhbuild):
1219         (update_webkit_libs_jhbuild):
1220         (determine_platform):
1221         (ensure_jhbuild):
1222
1223 2012-03-15  Carlos Garcia Campos  <cgarcia@igalia.com>
1224
1225         [GTK] Add command line options to run-gtk-tests script
1226         https://bugs.webkit.org/show_bug.cgi?id=81218
1227
1228         Reviewed by Philippe Normand.
1229
1230         Following command line options have been added:
1231
1232         -v, --verbose: to run gtester with --verbose. That would be used
1233         by the bots to make it easier to know which test case it failing.
1234
1235         --display: to use a different display for Xvfb
1236
1237         It's also possible to pass one or more tests as arguments.
1238
1239         * Scripts/run-gtk-tests:
1240         (TestRunner.__init__): Receive command line arguments.
1241         (TestRunner._get_tests): Get all the tests or the ones passed by
1242         command line arguments.
1243         (TestRunner._ensure_accessibility_daemon_is_running): Use display
1244         command line argument.
1245         (TestRunner.run): Pas command line options.
1246         (TestRunner.run.run_tests): Pass command line arguments.
1247
1248 2012-03-15  Carlos Garcia Campos  <cgarcia@igalia.com>
1249
1250         [GTK] WebKit2APITests/TestPrinting timeout in the bots
1251         https://bugs.webkit.org/show_bug.cgi?id=81211
1252
1253         Reviewed by Philippe Normand.
1254
1255         Add gtk+ patch to fix timeout in the bots. A similar patch has
1256         already landed in updatream gtk+, see
1257         http://git.gnome.org/browse/gtk+/commit/?id=5ebfad5d091f4c9d9c6e60364fad4c60842f027f
1258
1259         * gtk/jhbuild.modules:
1260         * gtk/patches/gtk+-3.0.12-gtkprinter.patch: Added.
1261
1262 2012-03-14  Gustavo Noronha Silva  <gns@gnome.org>
1263
1264         sheriffbot should also be addressable with a comma in addition to colon
1265         https://bugs.webkit.org/show_bug.cgi?id=81101
1266
1267         Reviewed by Adam Barth.
1268
1269         * Scripts/webkitpy/common/net/irc/ircbot.py:
1270         (IRCBot.on_pubmsg):
1271
1272 2012-03-15  Alexander Færøy  <alexander.faeroy@nokia.com>
1273
1274         [Qt] MiniBrowser scroll indicators should not scroll out of view
1275         https://bugs.webkit.org/show_bug.cgi?id=81131
1276
1277         Reviewed by Kenneth Rohde Christiansen.
1278
1279         * MiniBrowser/qt/qml/ScrollIndicator.qml:
1280
1281 2012-03-14  Chris Fleizach  <cfleizach@apple.com>
1282
1283         AX: ARIA tables need to support selected rows attribute/select rows changed notification
1284         https://bugs.webkit.org/show_bug.cgi?id=79821
1285
1286         Reviewed by Beth Dakin.
1287
1288         * DumpRenderTree/AccessibilityUIElement.cpp:
1289         (rowAtIndexCallback):
1290         (AccessibilityUIElement::getJSClass):
1291         * DumpRenderTree/AccessibilityUIElement.h:
1292         (AccessibilityUIElement):
1293         * DumpRenderTree/blackberry/AccessibilityUIElementBlackBerry.cpp:
1294         (AccessibilityUIElement::rowAtIndex):
1295         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
1296         (AccessibilityUIElement::rowAtIndex):
1297         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
1298         (AccessibilityUIElement::rowAtIndex):
1299         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
1300         (AccessibilityUIElement::rowAtIndex):
1301
1302 2012-03-14  Tim Horton  <timothy_horton@apple.com>
1303
1304         NRWT should respect --additional-platform-directory's Skipped list
1305         https://bugs.webkit.org/show_bug.cgi?id=81172
1306
1307         Reviewed by Timothy Hatcher.
1308
1309         Respect the Skipped file inside all paths included via the --additional-platform-directory option. Test results
1310         from said path are already used, but the Skipped file is ignored.
1311
1312         * Scripts/webkitpy/layout_tests/port/webkit.py:
1313         (WebKitPort._skipped_file_search_paths):
1314         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
1315         (test_skipped_file_search_paths):
1316
1317 2012-03-14  Ojan Vafai  <ojan@chromium.org>
1318
1319         Unreviewed, rolling out r110783.
1320         http://trac.webkit.org/changeset/110783
1321         https://bugs.webkit.org/show_bug.cgi?id=81054
1322
1323         Caused existing test expectation lines to expand to one for
1324         each test configuration.
1325
1326         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
1327         (TestExpectations.has_warnings):
1328         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
1329         (test_macro_overrides):
1330         * Scripts/webkitpy/tool/commands/rebaseline.py:
1331         (RebaselineTest._save_baseline):
1332         (RebaselineTest._rebaseline_test):
1333         (RebaselineTest.execute):
1334         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
1335         (TestRebaseline.test_tests_to_update):
1336
1337 2012-03-13  Jon Lee  <jonlee@apple.com>
1338
1339         Separate NOTIFICATIONS and LEGACY_NOTIFICATIONS
1340         https://bugs.webkit.org/show_bug.cgi?id=80922
1341         <rdar://problem/11035082>
1342
1343         Reviewed by Jian Li.
1344
1345         You can include either NOTIFICATIONS or LEGACY_NOTIFICATIONS and have a complete API.
1346         LEGACY_NOTIFICATIONS should cover all of the previous functionality, and NOTIFICATIONS will cover the
1347         new API. Therefore, APIs that are common between the two will have:
1348         #if ENABLE(NOTIFICATIONS) || ENABLE(LEGACY_NOTIFICATIONS)
1349
1350         This patch initially sets both to be exactly the same. As other bugs with patches begin to migrate to
1351         the new API, the defines will begin to split. This allows ports to decide which set of APIs to include.
1352
1353         Update everything to be #if ENABLE(NOTIFICATIONS) || ENABLE(LEGACY_NOTIFICATIONS)
1354         * DumpRenderTree/chromium/LayoutTestController.cpp:
1355         (LayoutTestController::grantDesktopNotificationPermission):
1356         (LayoutTestController::simulateDesktopNotificationClick):
1357         * DumpRenderTree/chromium/NotificationPresenter.cpp:
1358         * DumpRenderTree/chromium/TestShell.cpp:
1359         (TestShell::TestShell):
1360         (TestShell::resetTestController):
1361         * DumpRenderTree/win/DRTDesktopNotificationPresenter.cpp:
1362         (DRTDesktopNotificationPresenter::checkNotificationPermission):
1363
1364 2012-03-13  Ojan Vafai  <ojan@chromium.org>
1365
1366         Have webkit-patch rebaseline-test update test_expectations.txt
1367         https://bugs.webkit.org/show_bug.cgi?id=81054
1368
1369         Reviewed by Dirk Pranke.
1370
1371         This will allow us to use the All Failures tab in garden-o-matic
1372         to rebaseline tests without making test_expectations.txt stale.
1373
1374         It's conservative and only removes lines if the test is
1375         not flaky and fails in a way the rebaseline would fix (e.g.
1376         CRASH/TIMEOUT will not be touched).
1377
1378         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
1379         (TestExpectations.remove_configuration_from_test):
1380         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
1381         (RemoveConfigurationsTest):
1382         (RemoveConfigurationsTest.test_remove):
1383         (test_remove_line):
1384         * Scripts/webkitpy/tool/commands/rebaseline.py:
1385         (RebaselineTest._is_supported_port):
1386         (RebaselineTest):
1387         (RebaselineTest._update_expectations_file):
1388         (RebaselineTest._rebaseline_test_and_update_expectations):
1389         (RebaselineTest.execute):
1390         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
1391         (TestRebaseline.test_rebaseline_updates_expectations_file):
1392
1393 2012-03-14  Jessie Berlin  <jberlin@apple.com>
1394
1395         Crash in WKTR under addChromeInputField due to using 0 as a key in a HashMap
1396         https://bugs.webkit.org/show_bug.cgi?id=81167
1397
1398         Reviewed by Sam Weinig.
1399
1400         Do not use 0 as a key in a HashMap.
1401
1402         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
1403         cacheLayoutTestControllerCallback uses the enum as keys for the callback map.
1404
1405 2012-03-14  Ojan Vafai  <ojan@chromium.org>
1406
1407         Add the chromium lion bot to garden-o-matic and remove
1408         the platform_move_to in builders.py since the Chromium Lion
1409         port is now a normal port.
1410
1411         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
1412         * Scripts/webkitpy/layout_tests/port/builders.py:
1413
1414 2012-03-14  Dirk Pranke  <dpranke@chromium.org>
1415
1416         Fix regression/crash introduced in r110724 - bad type passed from skipped_tests() by default.
1417
1418         Unreviewed, build fix.
1419
1420         * Scripts/webkitpy/layout_tests/port/base.py:
1421         (Port.skipped_tests):
1422
1423 2012-03-14  Dirk Pranke  <dpranke@chromium.org>
1424
1425         Please add a way to manually skip some tests in NRWT
1426         https://bugs.webkit.org/show_bug.cgi?id=81019
1427
1428         Reviewed by Ojan Vafai.
1429
1430         This patch implements the -i / --ignore-tests flag from ORWT
1431         and refactors the test_expectations.py code slightly to handle
1432         it.
1433
1434         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1435         (Manager.parse_expectations):
1436         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
1437         (TestExpectations.__init__):
1438         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
1439         (test_add_skipped_tests):
1440         (test_add_skipped_tests_duplicate):
1441         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1442         (parse_args):
1443         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1444         (MainTest.test_ignore_tests):
1445         (MainTest.test_ignore_tests.assert_ignored):
1446
1447 2012-03-14  Dirk Pranke  <dpranke@chromium.org>
1448
1449         nrwt: run with no args on lion, is putting baselines in platform/mac-lion, not platform/mac
1450         https://bugs.webkit.org/show_bug.cgi?id=81028
1451
1452         Reviewed by Ojan Vafai.
1453
1454         The mac port had incorrect lookups for 'mac-lion' and
1455         'mac-future' in some places. This patch fixes that and adds a
1456         bunch of tests.
1457
1458         * Scripts/webkitpy/layout_tests/port/mac.py:
1459         (MacPort.baseline_path):
1460         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
1461         (MacTest.assert_skipped_file_search_paths):
1462         (MacTest.test_skipped_file_search_paths):
1463         (_assert_search_path):
1464         (test_baseline_search_path):
1465         * Scripts/webkitpy/layout_tests/port/webkit.py:
1466         (WebKitPort._skipped_file_search_paths):
1467
1468 2012-03-13  Jer Noble  <jer.noble@apple.com>
1469
1470         Lion Intel Debug WebKit2 Tests crashing under [WKFullScreenWindowController _startEnterFullScreenAnimationWithDuration:]
1471         https://bugs.webkit.org/show_bug.cgi?id=81056
1472
1473         Reviewed by Jessie Berlin.
1474
1475         Intercept beganEnterFullScreen and beganExitFullScreen and turn them into no-ops (plus logging).
1476
1477         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1478         (WTR::InjectedBundlePage::InjectedBundlePage):
1479         (WTR::InjectedBundlePage::beganEnterFullScreen):
1480         (WTR::InjectedBundlePage::beganExitFullScreen):
1481
1482 2012-03-14  Alexander Færøy  <alexander.faeroy@nokia.com>
1483
1484         The width and height of the scroll indicators in the MiniBrowser does
1485         not consider the offset margins.
1486         https://bugs.webkit.org/show_bug.cgi?id=81112
1487
1488         Reviewed by Kenneth Rohde Christiansen.
1489
1490         This patch fixes the issue by withdrawing two times the anchor
1491         margins.
1492
1493         * MiniBrowser/qt/qml/ScrollIndicator.qml:
1494
1495 2012-03-14  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
1496
1497         [EFL] Implement LayoutTestController::clearBackForwardList
1498         https://bugs.webkit.org/show_bug.cgi?id=80142
1499
1500         Adding missing implementation clearBackForwardList to EFL's
1501         LayoutTestController. This implementation allows us to unskip 
1502         tests connected with history.
1503
1504         Reviewed by Antonio Gomes.
1505
1506         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1507         (LayoutTestController::clearBackForwardList): Implemented.
1508
1509 2012-03-13  Thiago Marcos P. Santos  <thiago.santos@intel.com>
1510
1511         [EFL][DRT] Expose window.internals object
1512         https://bugs.webkit.org/show_bug.cgi?id=79900
1513
1514         Exposes window.internals for EFL's DRT. Makes possible to
1515         remove some tests from the skip list.
1516
1517         Reviewed by Antonio Gomes.
1518
1519         * DumpRenderTree/efl/CMakeLists.txt:
1520         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1521         (DumpRenderTreeChrome::onWindowObjectCleared):
1522
1523 2012-03-13  Adam Barth  <abarth@webkit.org> && Benjamin Poulain  <bpoulain@apple.com>
1524
1525         Always enable ENABLE(CLIENT_BASED_GEOLOCATION)
1526         https://bugs.webkit.org/show_bug.cgi?id=78853
1527
1528         Reviewed by Adam Barth.
1529
1530         * Scripts/build-webkit:
1531         * qmake/mkspecs/features/features.prf:
1532
1533 2012-03-12  Antonio Gomes  <agomes@rim.com>
1534
1535         Convert nodesFromRect tests to use Internals interface
1536         https://bugs.webkit.org/show_bug.cgi?id=80886
1537
1538         Reviewed by Ryosuke Niwa.
1539
1540         This clean up allows to remove port specific bits of each
1541         DRT implementation that supports testing Document::nodesFromRect,
1542         in favor of making use of the 'Internals' interface to test it.
1543
1544         This makes its code conveniently more expansible and cleaner.
1545
1546         Also removed stubs from Win and EFL DRTs.
1547
1548         * DumpRenderTree/LayoutTestController.cpp:
1549         (LayoutTestController::staticFunctions):
1550         * DumpRenderTree/LayoutTestController.h:
1551         (LayoutTestController):
1552         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1553         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
1554         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
1555         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
1556         * DumpRenderTree/qt/LayoutTestControllerQt.h:
1557         (LayoutTestController):
1558         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
1559
1560 2012-03-13  Ojan Vafai  <ojan@chromium.org>
1561
1562         Rebaselining for a new port doesn't work right with multiple fallback ports
1563         https://bugs.webkit.org/show_bug.cgi?id=80932
1564
1565         Reviewed by Adam Barth.
1566
1567         Since chromium-leopard also falls back to mac-leopard, we need to
1568         copy the existing result for both chromium-leopard and chromium-snowleopard
1569         before doing lion rebaselines.
1570
1571         * Scripts/webkitpy/layout_tests/port/builders.py:
1572         (fallback_port_names_for_new_port):
1573         * Scripts/webkitpy/tool/commands/rebaseline.py:
1574         (RebaselineTest):
1575         (RebaselineTest._copy_existing_baseline):
1576         (RebaselineTest._rebaseline_test):
1577         (RebaselineTest.execute):
1578         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
1579         * Scripts/webkitpy/tool/servers/gardeningserver.py:
1580         (GardeningHTTPRequestHandler.rebaseline):
1581         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
1582         (GardeningServerTest.test_rebaseline_new_port):
1583
1584 2012-03-13  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1585
1586         [Qt] Tweak minibrowser UI
1587
1588         Add some margins to the scroll indicators, and set a dark background
1589         color so you see the page dimensions.
1590
1591         Reviewed by Antonio Gomes.
1592
1593         * MiniBrowser/qt/qml/BrowserWindow.qml:
1594         * MiniBrowser/qt/qml/ScrollIndicator.qml:
1595
1596 2012-03-13  ChangSeok Oh  <shivamidow@gmail.com>
1597
1598         [EFL] [DRT] Implement scheduleAsynchronousKeyDown.
1599         https://bugs.webkit.org/show_bug.cgi?id=79601
1600
1601         Reviewed by Gustavo Noronha Silva.
1602
1603         Extracted shareable code from keyDownCallback and then added createKeyEventInfo.
1604         createKeyEventInfo creates a new structure KeyEventInfo that has keyName and modifiers
1605         and is taken over to sendKeyDown.
1606         Added sendKeyDown and sendAsynchronousKeyDown to evaluate keyDown event directly
1607         and asynchronously. Eventually, sendAsynchronousKeyDown calls sendKeyDown.
1608
1609         * DumpRenderTree/efl/EventSender.cpp:
1610         (KeyEventInfo::KeyEventInfo):
1611         (KeyEventInfo):
1612         (createKeyEventInfo):
1613         (sendKeyDown):
1614         (keyDownCallback):
1615         (sendAsynchronousKeyDown):
1616         (scheduleAsynchronousKeyDownCallback):
1617
1618 2012-03-12  Ojan Vafai  <ojan@chromium.org>
1619
1620         webkit-patch optimize-baselines sometimes creates a mac-future result
1621         https://bugs.webkit.org/show_bug.cgi?id=80717
1622
1623         Reviewed by Adam Barth.
1624
1625         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
1626         (_baseline_search_hypergraph):
1627         (BaselineOptimizer._find_optimal_result_placement):
1628         (BaselineOptimizer._filter_virtual_ports):
1629         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
1630         (BaselineOptimizerTest.test_no_add_mac_future):
1631
1632 2012-03-12  Jessie Berlin  <jberlin@apple.com>
1633
1634         More Windows build fixing.
1635
1636         * DumpRenderTree/win/WorkQueueItemWin.cpp:
1637
1638 2012-03-12  Jessie Berlin  <jberlin@apple.com>
1639
1640         More Windows build fixing.
1641
1642         * DumpRenderTree/win/AccessibilityControllerWin.cpp:
1643         * DumpRenderTree/win/EditingDelegate.cpp:
1644         * DumpRenderTree/win/EventSender.cpp:
1645         * DumpRenderTree/win/FrameLoadDelegate.cpp:
1646         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
1647         * DumpRenderTree/win/UIDelegate.cpp:
1648
1649 2012-03-12  Dirk Pranke  <dpranke@chromium.org>
1650
1651         remove the 'gpu' group from the flakiness dashboard
1652         https://bugs.webkit.org/show_bug.cgi?id=80353
1653
1654         Reviewed by Adam Barth.
1655
1656         Since those bots are obsolete, we don't need this any more.
1657
1658         * TestResultServer/static-dashboards/builders.js:
1659         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
1660         (testPlatformAndBuildType):
1661         (testHtmlForTestTypeSwitcherGroup):
1662
1663 2012-03-12  Dirk Pranke  <dpranke@chromium.org>
1664
1665         remove graphics_type from webkitpy.layout_tests.models.test_configuration
1666         https://bugs.webkit.org/show_bug.cgi?id=80643
1667
1668         Reviewed by Ojan Vafai.
1669
1670         Now that the chromium-gpu ports are gone, there's no need for
1671         the graphics_type CPU/GPU distinction. Most of this change is
1672         just updating tests accordingly.
1673
1674         * Scripts/webkitpy/layout_tests/controllers/test_expectations_editor_unittest.py:
1675         (TestExpectationEditorTests):
1676         * Scripts/webkitpy/layout_tests/models/test_configuration.py:
1677         (TestConfiguration.__init__):
1678         (TestConfiguration.category_order):
1679         (TestConfiguration.__str__):
1680         (TestConfiguration.__repr__):
1681         (TestConfiguration.__hash__):
1682         (TestConfigurationConverter.to_specifiers_list):
1683         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py:
1684         (make_mock_all_test_configurations_set):
1685         (TestConfigurationTest.test_items):
1686         (TestConfigurationTest.test_keys):
1687         (TestConfigurationTest.test_str):
1688         (TestConfigurationTest.test_repr):
1689         (TestConfigurationTest.test_hash):
1690         (TestConfigurationTest.test_hash.query_unknown_key):
1691         (TestConfigurationTest.test_eq):
1692         (TestConfigurationTest.test_values):
1693         (SpecifierSorterTest.test_specifier_priority):
1694         (SpecifierSorterTest.test_sort_specifiers):
1695         (TestConfigurationConverterTest.test_to_config_set):
1696         (TestConfigurationConverterTest.test_macro_expansion):
1697         (TestConfigurationConverterTest.test_to_specifier_lists):
1698         (TestConfigurationConverterTest.test_converter_macro_collapsing):
1699         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
1700         (test_error_on_different_build_type):
1701         (TestExpectationSerializerTests.test_parsed_to_string):
1702         (TestExpectationSerializerTests.test_reconstitute_only_these):
1703         * Scripts/webkitpy/layout_tests/port/apple.py:
1704         (ApplePort._generate_all_test_configurations):
1705         * Scripts/webkitpy/layout_tests/port/base.py:
1706         (Port.__init__):
1707         (Port.name):
1708         (Port.version):
1709         (Port.test_configuration):
1710         * Scripts/webkitpy/layout_tests/port/builders.py:
1711         * Scripts/webkitpy/layout_tests/port/chromium.py:
1712         (ChromiumPort):
1713         (ChromiumPort._generate_all_test_configurations):
1714         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
1715         (ChromiumLinuxPort.__init__):
1716         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
1717         (ChromiumLinuxPortTest.test_check_illegal_port_names):
1718         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
1719         (ChromiumMacPortTest.test_baseline_path):
1720         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
1721         (ChromiumPortTest.test_all_test_configurations):
1722         * Scripts/webkitpy/layout_tests/port/efl.py:
1723         (EflPort._generate_all_test_configurations):
1724         * Scripts/webkitpy/layout_tests/port/gtk.py:
1725         (GtkPort._generate_all_test_configurations):
1726         * Scripts/webkitpy/layout_tests/port/qt.py:
1727         (QtPort._generate_all_test_configurations):
1728         * Scripts/webkitpy/layout_tests/port/test.py:
1729         (TestPort.all_test_configurations):
1730         (TestPort._all_build_types):
1731         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
1732         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
1733         (BuildCoverageExtrapolatorTest.test_extrapolate):
1734         (GardeningExpectationsUpdaterTest.test_unknown_failure_type):
1735         (GardeningExpectationsUpdaterTest.test_add_new_expectation):
1736         (GardeningExpectationsUpdaterTest.test_replace_old_expectation):
1737         (GardeningExpectationsUpdaterTest.test_pass_expectation):
1738         (GardeningExpectationsUpdaterTest.test_supplement_old_expectation):
1739         (GardeningExpectationsUpdaterTest.test_spurious_updates):
1740
1741 2012-03-12  Dirk Pranke  <dpranke@chromium.org>
1742
1743         _runtime_feature_list isn't supported on Apple Mac DRT
1744         https://bugs.webkit.org/show_bug.cgi?id=80906
1745
1746         Reviewed by Eric Seidel.
1747
1748         It looks like this feature is actually only supported on the Win
1749         DRT port, so this patch reworks the code to make that clearer
1750         and reduce unnecessary overrides.
1751
1752         * Scripts/webkitpy/layout_tests/port/efl.py:
1753         (EflPort._path_to_webcore_library):
1754         * Scripts/webkitpy/layout_tests/port/gtk.py:
1755         (GtkPort._path_to_webcore_library):
1756         * Scripts/webkitpy/layout_tests/port/qt.py:
1757         (QtPort._skipped_file_search_paths):
1758         * Scripts/webkitpy/layout_tests/port/webkit.py:
1759         (WebKitPort._runtime_feature_list):
1760         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
1761         (TestWebKitPort.__init__):
1762         (TestWebKitPort.all_test_configurations):
1763         (test_skipped_directories_for_features):
1764         * Scripts/webkitpy/layout_tests/port/win.py:
1765         (WinPort):
1766         (WinPort._runtime_feature_list):
1767         * Scripts/webkitpy/layout_tests/port/win_unittest.py:
1768         (WinPortTest.test_operating_system):
1769         (WinPortTest):
1770         (WinPortTest.test_runtime_feature_list):
1771
1772 2012-03-12  Peter Beverloo  <peter@chromium.org>
1773
1774         [Chromium] Temporary build-fix for the Android bot
1775         https://bugs.webkit.org/show_bug.cgi?id=80862
1776
1777         Reviewed by Adam Barth.
1778
1779         Manually set the environmental variables required to compile source
1780         files through the Android NDK instead of relying on the envsetup.sh
1781         script. A proper fix will be landed to Chromium ASAP, and following
1782         a Chromium roll into WebKit this code will be removed.
1783
1784         * Scripts/webkitdirs.pm:
1785         (buildChromiumMakefile):
1786
1787 2012-03-12  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1788
1789         [Qt] Use right construct to detect config_test_gccdepends
1790
1791         Reviewed by Andreas Kling.
1792
1793         * qmake/mkspecs/features/default_pre.prf:
1794
1795 2012-03-12  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1796
1797         [Qt] Expose location of WebKit's bin and imports dir through module file
1798
1799         Reviewed by Andreas Kling.
1800
1801         * qmake/qt_webkit.pri:
1802
1803 2012-03-12  Andras Becsi  <andras.becsi@nokia.com>
1804
1805         [Qt][WK2] Add support for rudimentary scroll indicators in MiniBrowser
1806         https://bugs.webkit.org/show_bug.cgi?id=80832
1807
1808         Reviewed by Tor Arne Vestbø.
1809
1810         Add a basic scroll indicator component to MiniBrowser so that it
1811         behaves similar to the ScrollDecorator QML component.
1812
1813         * MiniBrowser/qt/MiniBrowser.qrc:
1814         * MiniBrowser/qt/qml/BrowserWindow.qml:
1815         * MiniBrowser/qt/qml/ScrollIndicator.qml: Added.
1816
1817 2012-03-12  C Anthony Risinger  <anthony@xtfx.me>
1818
1819         [GTK] r110296 included an extra $ in @$(AM_V_GEN)
1820         https://bugs.webkit.org/show_bug.cgi?id=80809
1821
1822         Reviewed by Philippe Normand.
1823
1824         * GNUmakefile.am: Add `$` to @(AM_V_GEN), else /bin/sh syntax error
1825
1826 2012-03-12  Ming Xie  <mxie@rim.com>
1827
1828         [BlackBerry] build-webkit doesn't recognize "--cmakearg" for the BlackBerry port
1829         https://bugs.webkit.org/show_bug.cgi?id=80812
1830
1831         Reviewed by Rob Buis.
1832
1833         * Scripts/build-webkit: Add $cmakeArgs to buildCMakeProjectOrExit()
1834         * Scripts/webkitdirs.pm:
1835         (blackberryCMakeArguments): Remove "PUBLIC_BUILD=0" define
1836
1837
1838 2012-03-12  Simon Hausmann  <simon.hausmann@nokia.com>
1839
1840         [Qt] Fix build with newer Qt 5.
1841
1842         Reviewed by Tor Arne Vestbø.
1843
1844         Newer Qt 5 renames QDeclarativeFoo to QQmlFoo and provides backwards compatible
1845         renames, which however do trigger a #warning. Until we _require_ a newer Qt 5
1846         we should disable -Werror.
1847
1848         Also load(moc) for the availability of QMAKE_MOC in various derived sources. The
1849         implicit load was removed recently.
1850
1851         * qmake/mkspecs/features/default_pre.prf:
1852         * qmake/mkspecs/features/unix/default_post.prf:
1853
1854 2012-03-11  Maciej Stachowiak  <mjs@apple.com>
1855
1856         Add --instruments option to sunspider to profile with Instruments
1857         https://bugs.webkit.org/show_bug.cgi?id=80783
1858
1859         Reviewed by Filip Pizlo.
1860
1861         * Scripts/run-sunspider:
1862
1863 2012-03-11  Viatcheslav Ostapenko  <ostapenko.viatcheslav@nokia.com>
1864
1865         [Qt] [WK2] Support threaded renderer in WK2
1866         https://bugs.webkit.org/show_bug.cgi?id=76661
1867
1868         Remove disabling of threaded rendering.
1869
1870         Reviewed by Noam Rosenthal.
1871
1872         * MiniBrowser/qt/main.cpp:
1873         (main):
1874
1875 2012-03-09  Zan Dobersek  <zandobersek@gmail.com>
1876
1877         [WK2] run-perf-tests should be able to run with WTR
1878         https://bugs.webkit.org/show_bug.cgi?id=80042
1879
1880         Reviewed by Martin Robinson.
1881
1882         Add an option to PerfTestRunner to use WebKitTestRunner instead of
1883         DumpRenderTree and add the '--no-timeout' flag to WebKitTestRunner
1884         to avoid timing out when running performance tests.
1885
1886         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1887         (PerfTestsRunner._parse_args):
1888         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
1889         (WTR::InjectedBundle::InjectedBundle): Initialize m_useWaitToDumpWatchdogTimer
1890         to true.
1891         (WTR::InjectedBundle::didReceiveMessage): Unpack the 'BeginTest' message
1892         as a dictionary containing DumpPixels and UseWaitToDumpWatchdogTimer keys.
1893         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
1894         (WTR::InjectedBundle::useWaitToDumpWatchdogTimer): Make m_useWaitToDumpWatchdogTimer
1895         publicly accessible.
1896         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
1897         (WTR::LayoutTestController::waitUntilDone): Only initialize the watchdog if it
1898         is required.
1899         * WebKitTestRunner/TestController.cpp:
1900         (WTR::TestController::TestController): Initialize m_useWaitToDumpWatchdogTimer
1901         to true.
1902         (WTR::TestController::initialize): Set m_useWaitToDumpWatchdogTimer to false
1903         if '--no-timeout' option is present.
1904         (WTR::TestController::runUntil): Translate TimeoutDuration to a proper timeout
1905         value, or -1.0 if timeout should not occur.
1906         * WebKitTestRunner/TestController.h:
1907         (WTR::TestController::useWaitToDumpWatchdogTimer): Make m_useWaitToDumpWatchdogTimer
1908         publicly accessible.
1909         * WebKitTestRunner/TestInvocation.cpp:
1910         (WTR::TestInvocation::invoke): Pack the 'BeginTest' message as a dictionary,
1911         containing DumpPixels and UseWaitToDumpWatchdogTimer keys and corresponding values.
1912         Waiting for final message should not time out if not necessary.
1913         * WebKitTestRunner/gtk/TestControllerGtk.cpp:
1914         (WTR::TestController::platformRunUntil): Only set up a timeout if necessary.
1915         * WebKitTestRunner/qt/TestControllerQt.cpp:
1916         (WTR::TestController::platformRunUntil): Ditto.
1917         * WebKitTestRunner/mac/TestControllerMac.mm:
1918         (WTR::TestController::platformRunUntil): Add a FIXME, indicating functionality
1919         is missing.
1920         * WebKitTestRunner/win/TestControllerWin.cpp:
1921         (WTR::TestController::platformRunUntil): Ditto.
1922
1923 2012-03-09  Jon Lee  <jonlee@apple.com>
1924
1925         Rename NotificationPresenter to NotificationClient
1926         https://bugs.webkit.org/show_bug.cgi?id=80488
1927         <rdar://problem/10965558>
1928
1929         Reviewed by Kentaro Hara.
1930
1931         Refactor to use renamed WebCore::NotificationClient.
1932         * DumpRenderTree/win/DRTDesktopNotificationPresenter.cpp:
1933         (DRTDesktopNotificationPresenter::checkNotificationPermission):
1934
1935 2012-03-07  James Robinson  <jamesr@chromium.org>
1936
1937         [chromium] Deprecate renderDirectlyToWebView parameter of WebViewClient::createGraphicsContext3D()
1938         https://bugs.webkit.org/show_bug.cgi?id=80523
1939
1940         Reviewed by Darin Fisher.
1941
1942         * DumpRenderTree/chromium/WebViewHost.cpp:
1943         (WebViewHost::createGraphicsContext3D):
1944         * DumpRenderTree/chromium/WebViewHost.h:
1945         (WebViewHost):
1946
1947 2012-03-09  Eric Seidel  <eric@webkit.org>
1948
1949         build-webkit should build Source/WTF on Mac
1950         https://bugs.webkit.org/show_bug.cgi?id=80728
1951
1952         Reviewed by Adam Barth.
1953
1954         We'll add other ports to this list in future commits.
1955
1956         * Scripts/build-webkit:
1957
1958 2012-03-09  Jon Lee  <jonlee@apple.com>
1959
1960         Add support for ENABLE(LEGACY_NOTIFICATIONS)
1961         https://bugs.webkit.org/show_bug.cgi?id=80497
1962
1963         Reviewed by Adam Barth.
1964
1965         Prep for b80472: Update API for Web Notifications
1966         * Scripts/build-webkit:
1967         * qmake/mkspecs/features/features.prf:
1968
1969 2012-03-09  Ashod Nakashian  <ashodnakashian@yahoo.com>
1970
1971         Bash scripts should support LF endings only
1972         https://bugs.webkit.org/show_bug.cgi?id=79509
1973
1974         Reviewed by David Kilzer.
1975
1976         * EWSTools/boot.sh: Added properties svn:executable and svn:eol-style.
1977         * EWSTools/start-queue.sh: Added property svn:eol-style.
1978         * Scripts/webkit-tools-completion.sh: Added properties svn:executable and svn:eol-style.
1979         * WebKitTestRunner/win/build-generated-files.sh: Added properties svn:executable and svn:eol-style.
1980         * iExploder/iexploder-1.7.2/tools/release_src.sh: Added property svn:eol-style.
1981         * iExploder/iexploder-1.7.2/tools/update_html_tags_from_sources.sh: Added property svn:eol-style.
1982
1983 2012-03-09  Ionut Biru  <ibiru@archlinux.org>
1984
1985         [BUILD] use detected python to run generate-gtkdoc
1986         https://bugs.webkit.org/show_bug.cgi?id=80695
1987
1988         Reviewed by Gustavo Noronha Silva.
1989
1990         * GNUmakefile.am: Use $(PYTHON) instead of plain `python`
1991
1992 2012-03-09  Nat Duca  <nduca@chromium.org>
1993
1994         Make check-webkit-style complain about "no new tests" lines in changelogs
1995         https://bugs.webkit.org/show_bug.cgi?id=80582
1996
1997         Reviewed by David Levin.
1998
1999         * Scripts/webkitpy/style/checkers/changelog.py:
2000         (ChangeLogChecker.check_entry):
2001         * Scripts/webkitpy/style/checkers/changelog_unittest.py:
2002         (ChangeLogCheckerTest.test_no_new_tests):
2003
2004 2012-03-09  Sheriff Bot  <webkit.review.bot@gmail.com>
2005
2006         Unreviewed, rolling out r110250.
2007         http://trac.webkit.org/changeset/110250
2008         https://bugs.webkit.org/show_bug.cgi?id=80682
2009
2010         It made NRWT ignore test_expectations.txt on Qt-WK2 and ignore
2011         qt-linux at all. (Requested by Ossy on #webkit).
2012
2013         * Scripts/old-run-webkit-tests:
2014         * Scripts/webkitpy/layout_tests/port/qt.py:
2015         (QtPort):
2016         (QtPort.determine_full_port_name):
2017         (QtPort.baseline_search_path):
2018         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
2019         (QtPortTest):
2020         (QtPortTest._assert_search_path):
2021         (QtPortTest.test_baseline_search_path):
2022
2023 2012-03-08  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2024
2025         [Qt] Use Qt's module system for install rules and depending on QtWebKit
2026
2027         Instead of rolling our own install rules we now use the same approach as
2028         every other Qt module, by loading qt_module.prf and qt_module_config.prf.
2029
2030         This ensures that we follow the same semantics as the rest of Qt on
2031         what sort of config options are enabled by default (create_cmake eg.).
2032         It also allows us to use QT += webkit instead of the workaround we had
2033         with CONFIG += qtwebkit.
2034
2035         We do however force Qt to always treat our build as a non-developer build,
2036         so the libraries will end up in the WebKit lib directory instead of the
2037         qtbase directory (as with a normal developer-build). This allows us to
2038         keep the webkit-build self-contained. If Qt is a developer build we still
2039         copy the module file manually to Qt, so that you don't have to install
2040         WebKit to make it available.
2041
2042         For non-developer builds of Qt, it is still possible to use the built
2043         WebKit libraries without having to install them, by having the variable
2044         QMAKE_EXTRA_MODULE_FORWARDS set in the project's .qmake.cache file,
2045         pointing to $WEBKITOUTUTDIR/$CONFIGURATION/modules.
2046
2047         https://bugs.webkit.org/show_bug.cgi?id=80590
2048
2049         Reviewed by Simon Hausmann.
2050
2051         * DumpRenderTree/qt/DumpRenderTree.pro:
2052         * MiniBrowser/qt/BrowserWindow.cpp:
2053         * MiniBrowser/qt/MiniBrowser.pro:
2054         * MiniBrowser/qt/MiniBrowserApplication.cpp:
2055         * MiniBrowser/qt/UrlLoader.cpp:
2056         * QtTestBrowser/QtTestBrowser.pro:
2057         * Tools.pro:
2058         * WebKitTestRunner/InjectedBundle/Target.pri:
2059         * WebKitTestRunner/Target.pri:
2060         * qmake/mkspecs/features/default_post.prf:
2061         * qmake/mkspecs/features/default_pre.prf:
2062         * qmake/mkspecs/features/functions.prf:
2063         * qmake/mkspecs/features/qtwebkit-private.prf: Removed.
2064         * qmake/mkspecs/features/qtwebkit.prf:
2065         * qmake/mkspecs/features/rpath.prf:
2066         * qmake/mkspecs/modules/qt_webkit.pri: Removed.
2067         * qmake/qt_webkit.pri: Added.
2068
2069 2012-03-08  Igor Oliveira  <igor.o@sisa.samsung.com>
2070
2071         Unreviewed. Update my email.
2072
2073         * Scripts/webkitpy/common/config/committers.py:
2074
2075 2012-03-08  Xianzhu Wang  <wangxianzhu@chromium.org>
2076
2077         Recent changes to chromium.py etc. broke chromium_android.py
2078         https://bugs.webkit.org/show_bug.cgi?id=80623
2079
2080         Reviewed by Dirk Pranke.
2081
2082         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2083         (ChromiumAndroidDriver.__init__):
2084         (ChromiumAndroidDriver._start):
2085
2086 2012-03-08  Xianzhu Wang  <wangxianzhu@chromium.org>
2087
2088         [Chromium] Exception running reftest with --no-pixel-tests
2089         https://bugs.webkit.org/show_bug.cgi?id=80642
2090
2091         Reviewed by Dirk Pranke.
2092
2093         * Scripts/webkitpy/layout_tests/port/chromium.py:
2094         (ChromiumDriver.__init__):
2095         (ChromiumDriver._wrapper_options):
2096
2097 2012-03-08  Rafael Brandao  <rafael.lobo@openbossa.org>
2098
2099         [NRWT] Fix --platform=qt-5.0 --new-baseline combo
2100         https://bugs.webkit.org/show_bug.cgi?id=72489
2101
2102         Reviewed by Dirk Pranke.
2103
2104         NRWT: The default platform name for Qt combined with --webkit-test-runner is now
2105         "qt-5.0-wk2" instead of "qt-linux" and that name is now prefered for tests.
2106         --platform will override the platform name and then we'll use it.
2107
2108         ORWT: We check for qt version and use platform name "qt-4.8" if the version is lower
2109         than 5.0.0, and then we choose between "qt-5.0-wk1" and "qt-5.0-wk2". --platform will
2110         override it in the end.
2111
2112         * Scripts/old-run-webkit-tests:
2113         * Scripts/webkitpy/layout_tests/port/qt.py:
2114         (QtPort):
2115         (QtPort.determine_full_port_name):
2116         (QtPort.baseline_search_path):
2117         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
2118         (QtPortTest):
2119         (QtPortTest._assert_search_path):
2120         (QtPortTest.test_baseline_search_path):
2121
2122 2012-03-08  Beth Dakin  <bdakin@apple.com>
2123
2124         https://bugs.webkit.org/show_bug.cgi?id=80463
2125         RenderImage is using the wrong origin when calling addRelevantRepaintedObject
2126         -and corresponding-
2127         <rdar://problem/10970221>
2128
2129         Reviewed by Dan Bernstein.
2130
2131         New test!
2132         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
2133         * TestWebKitAPI/Tests/WebKit2/NewFirstVisuallyNonEmptyLayoutForImages.cpp: Added.
2134         (TestWebKitAPI):
2135         (TestWebKitAPI::didNewFirstVisuallyNonEmptyLayout):
2136         (TestWebKitAPI::setPageLoaderClient):
2137         (TestWebKitAPI::TEST):
2138         * TestWebKitAPI/Tests/WebKit2/NewFirstVisuallyNonEmptyLayoutForImages_Bundle.cpp: Added.
2139         (TestWebKitAPI):
2140         (NewFirstVisuallyNonEmptyLayoutForImagesTest):
2141         (TestWebKitAPI::NewFirstVisuallyNonEmptyLayoutForImagesTest::NewFirstVisuallyNonEmptyLayoutForImagesTest):
2142         (TestWebKitAPI::NewFirstVisuallyNonEmptyLayoutForImagesTest::didCreatePage):
2143         * TestWebKitAPI/Tests/WebKit2/lots-of-images.html: Added.
2144
2145 2012-03-08  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2146
2147         [Qt] Prospective Windows build fix
2148
2149         Don't assume that windows-builds will always run inside a cmd.exe shell.
2150
2151         * qmake/mkspecs/features/default_pre.prf:
2152
2153 2012-03-08  Dan Bernstein  <mitz@apple.com>
2154
2155         <rdar://problem/10981173> Dashboard regions should not be in device space
2156
2157         Reviewed by John Sullivan.
2158
2159         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
2160         * TestWebKitAPI/Tests/mac/DeviceScaleFactorInDashboardRegions.mm: Added.
2161         (-[DeviceScaleFactorInDashboardRegionsUIDelegate webView:dashboardRegionsChanged:]):
2162         (TestWebKitAPI):
2163         (TestWebKitAPI::TEST):
2164
2165 2012-03-08  Dirk Pranke  <dpranke@chromium.org>
2166
2167         webkit-patch works oddly on local Git branches
2168         https://bugs.webkit.org/show_bug.cgi?id=76958
2169
2170         Reviewed by Ojan Vafai.
2171
2172         This change makes the '-g' flag to webkit-patch behave a
2173         little more consistently with git. The 'HEAD..' syntax
2174         is changed to 'HEAD....' (four dots), so as to not
2175         collide with the underlying git syntax. We also generalize
2176         the syntax to support any refname, so 'foo....' means
2177         diff against foo and include working copy changes.
2178
2179         This change also adds the 'UPSTREAM' shortcut to identify
2180         whatever the current upstream/tracking branch is; this is
2181         useful if you have multiple local branches stacked up.
2182
2183         * Scripts/webkitpy/common/checkout/scm/git.py:
2184         (Git._branch_from_ref):
2185         (Git):
2186         (Git._current_branch):
2187         (Git._upstream_branch):
2188         (Git.merge_base):
2189         (Git._commit_on_branch):
2190         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
2191         (GitSVNTest._second_local_commit):
2192         (GitSVNTest._two_local_commits):
2193         (GitSVNTest.test_commit_with_message_git_commit_range):
2194         (GitSVNTest.test_upstream_branch):
2195         (GitSVNTest.test_create_patch_working_copy_only):
2196         (GitSVNTest.test_changed_files_local_plus_working_copy):
2197         (GitSVNTest.test_changed_files_working_copy_only):
2198         (GitSVNTest.test_changed_files_upstream):
2199         * Scripts/webkitpy/tool/steps/options.py:
2200         (Options):
2201
2202 2012-03-08  John Sullivan  <sullivan@apple.com>
2203
2204         Made extract-localizable-strings ignore NSLocalizedFileSizeDescription
2205
2206         Reviewed by Eric Carlson.
2207
2208         * Scripts/extract-localizable-strings:
2209         Added NSLocalizedFileSizeDescription to allowed list of tokens that start
2210         with NSLocalized. Also improved an error message to include the troublesome token.
2211
2212 2012-03-08  Sheriff Bot  <webkit.review.bot@gmail.com>
2213
2214         Unreviewed, rolling out r110115.
2215         http://trac.webkit.org/changeset/110115
2216         https://bugs.webkit.org/show_bug.cgi?id=80581
2217
2218         broke test-webkitpy (Requested by philn-tp on #webkit).
2219
2220         * Scripts/old-run-webkit-tests:
2221         * Scripts/webkitpy/layout_tests/port/qt.py:
2222         (QtPort):
2223         (QtPort.determine_full_port_name):
2224         (QtPort.baseline_search_path):
2225
2226 2012-03-08  Kentaro Hara  <haraken@chromium.org>
2227
2228         Unreviewed. Added myself to watchlist.
2229
2230         * Scripts/webkitpy/common/config/watchlist:
2231
2232 2012-03-07  Jessie Berlin  <jberlin@apple.com>
2233
2234         Clean Windows build fails after r110033
2235         https://bugs.webkit.org/show_bug.cgi?id=80553
2236
2237         Rubber-stamped by Jon Honeycutt and Eric Seidel.
2238
2239         Add the new JavaScriptCore include directory to some more vsprops files.
2240
2241         * DumpRenderTree/win/ImageDiffCommon.vsprops:
2242         * WebKitTestRunner/Configurations/InjectedBundleCommon.vsprops:
2243         * WebKitTestRunner/Configurations/WebKitTestRunnerCommon.vsprops:
2244
2245 2012-03-07  Rafael Brandao  <rafael.lobo@openbossa.org>
2246
2247         [NRWT] Fix --platform=qt-5.0 --new-baseline combo
2248         https://bugs.webkit.org/show_bug.cgi?id=72489
2249
2250         Reviewed by Dirk Pranke.
2251
2252         NRWT: The default platform name for Qt combined with --webkit-test-runner is now
2253         "qt-5.0-wk2" instead of "qt-linux" and that name is now prefered for tests.
2254         --platform will override the platform name and then we'll use it.
2255
2256         ORWT: We check for qt version and use platform name "qt-4.8" if the version is lower
2257         than 5.0.0, and then we choose between "qt-5.0-wk1" and "qt-5.0-wk2". --platform will
2258         override it in the end.
2259
2260         * Scripts/old-run-webkit-tests:
2261         * Scripts/webkitpy/layout_tests/port/qt.py:
2262         (QtPort):
2263         (QtPort.determine_full_port_name):
2264         (QtPort.baseline_search_path):
2265
2266 2012-03-07  Hugo Parente Lima  <hugo.lima@openbossa.org>
2267
2268         [Qt][WK2] MiniBrowser on desktop does not react to click events
2269         https://bugs.webkit.org/show_bug.cgi?id=80314
2270
2271         Reviewed by Kenneth Rohde Christiansen.
2272
2273         Do not enable touch mocking on desktop mode.
2274
2275         * MiniBrowser/qt/MiniBrowserApplication.cpp:
2276         (MiniBrowserApplication::handleUserOptions):
2277
2278 2012-03-07  Ojan Vafai  <ojan@chromium.org>
2279
2280         Automatically move to the next test after rebaseline in the results view
2281         https://bugs.webkit.org/show_bug.cgi?id=80527
2282
2283         Reviewed by Dimitri Glazkov.
2284
2285         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
2286         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
2287
2288 2012-03-07  Ojan Vafai  <ojan@chromium.org>
2289
2290         garden-o-matic rebaseline is broken for ports that don't have a fallback port
2291         https://bugs.webkit.org/show_bug.cgi?id=80526
2292
2293         Reviewed by Adam Barth.
2294
2295         Passing "None" as the fallback port throws an error in "webkit-patch rebaseline-test"
2296         since there is no "None" port.
2297
2298         * Scripts/webkitpy/tool/servers/gardeningserver.py:
2299         (GardeningHTTPRequestHandler.rebaseline):
2300         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
2301         (GardeningServerTest.test_rebaseline):
2302
2303 2012-03-07  James Robinson  <jamesr@chromium.org>
2304
2305         [chromium] Remove the TestWebPlugin's use of implicit FBO for offscreen contexts
2306         https://bugs.webkit.org/show_bug.cgi?id=80521
2307
2308         Reviewed by Kenneth Russell.
2309
2310         Switches TestWebPlugin from using a context created by WebViewClient::createGraphicsContext3D(..., false) over
2311         to an explicitly offscreen context with explicitly managed texture / FBO, so we can deprecate the
2312         renderDirectlyToHostWindow flag from WebViewClient::createGraphicsContext3D.
2313
2314         Covered by platform/chromium/compositing/plugins/*
2315
2316         * DumpRenderTree/chromium/TestWebPlugin.cpp:
2317         (TestWebPlugin::TestWebPlugin):
2318         (TestWebPlugin::initialize):
2319         (TestWebPlugin::updateGeometry):
2320         (TestWebPlugin::initScene):
2321         (TestWebPlugin::destroyScene):
2322         * DumpRenderTree/chromium/TestWebPlugin.h:
2323         (TestWebPlugin):
2324         * DumpRenderTree/chromium/WebViewHost.cpp:
2325         (WebViewHost::createPlugin):
2326
2327 2012-03-07  Philippe Normand  <pnormand@igalia.com>
2328
2329         [GTK] remove webkitpy dependency in run-gtk-tests
2330         https://bugs.webkit.org/show_bug.cgi?id=80500
2331
2332         Reviewed by Martin Robinson.
2333
2334         Replace Executive calls with bare subprocess calls.
2335
2336         * Scripts/run-gtk-tests:
2337         (TestRunner):
2338         (TestRunner._lookup_atspi2_binary):
2339         (TestRunner.run):
2340         (TestRunner.run.run_for_real):
2341
2342 2012-03-07  Philippe Normand  <pnormand@igalia.com>
2343
2344         Unreviewed, GTK build fix after r110059.
2345
2346         * Scripts/run-gtk-tests:
2347         (TestRunner._check_if_tests_have_timed_out):
2348         (TestRunner.run.run_tests):
2349         (TestRunner):
2350
2351 2012-03-07  Philippe Normand  <pnormand@igalia.com>
2352
2353         [GTK] race condition in run-gtk-tests
2354         https://bugs.webkit.org/show_bug.cgi?id=80495
2355
2356         Reviewed by Martin Robinson.
2357
2358         Refactored the script to wait the a11y dbus service becomes
2359         available before starting the tests.
2360
2361         * Scripts/run-gtk-tests:
2362         (TestRunner._lookup_atspi2_binary):
2363         (TestRunner):
2364         (TestRunner._wait_dbus_service_and_run):
2365         (TestRunner._wait_dbus_service_and_run.on_name_appeared):
2366         (TestRunner._wait_dbus_service_and_run.on_name_vanished):
2367         (TestRunner.run):
2368         (TestRunner.run.bailout):
2369         (TestRunner.run.run_for_real):
2370
2371 2012-03-07  Dinu Jacob  <dinu.jacob@nokia.com>
2372
2373         [Qt] REGRESSION(r109575) Broke tap-to-zoom gesture
2374         https://bugs.webkit.org/show_bug.cgi?id=80425
2375
2376         Reviewed by Simon Hausmann.
2377
2378         Set startScreenPos in mock touch point
2379
2380         * MiniBrowser/qt/MiniBrowserApplication.cpp:
2381         (MiniBrowserApplication::notify):
2382         * MiniBrowser/qt/MiniBrowserApplication.h:
2383         (MiniBrowserApplication):
2384
2385 2012-03-07  Simon Hausmann  <simon.hausmann@nokia.com>
2386
2387         [Qt] Windows build fix.
2388
2389         Reviewed by Tor Arne Vestbø.
2390
2391         Pass the -j parameter only to "make", not "nmake".
2392
2393         * Scripts/webkitdirs.pm:
2394         (buildQMakeProjects):
2395
2396 2012-03-07  Simon Hausmann  <simon.hausmann@nokia.com>
2397
2398         [Qt] Fix compilation without QtQuick1
2399         https://bugs.webkit.org/show_bug.cgi?id=80503
2400
2401         Reviewed by Tor Arne Vestbø.
2402
2403         Introduce HAVE_QQUICK1 as feature and added haveQtModule
2404         as convenience function.
2405
2406         * qmake/mkspecs/features/features.prf:
2407         * qmake/mkspecs/features/functions.prf:
2408
2409 2012-03-07  Zoltan Horvath  <zoltan@webkit.org>
2410
2411         [Qt] Add qmake config tests for JPEG and PNG library
2412         https://bugs.webkit.org/show_bug.cgi?id=80398
2413
2414         Add JPEG and PNG library tests to config.tests and
2415         add error() abort to WebCore.pri if tests don't work.
2416
2417         Reviewed by Tor Arne Vestbø.
2418
2419         * qmake/config.tests/libjpeg/libjpeg.cpp: Added.
2420         (main):
2421         * qmake/config.tests/libjpeg/libjpeg.pro: Added.
2422         * qmake/config.tests/libpng/libpng.cpp: Added.
2423         (main):
2424         * qmake/config.tests/libpng/libpng.pro: Added.
2425         * qmake/sync.profile:
2426
2427 2012-03-07  Byungwoo Lee  <bw80.lee@samsung.com>
2428
2429         [EFL] Build warning: Fix warn_unused_result warnings.
2430         https://bugs.webkit.org/show_bug.cgi?id=79194
2431
2432         Reviewed by Filip Pizlo.
2433
2434         Fixed build warning which is getting generated when not using return
2435         value of a fuction, especially declared with warn_unused_result
2436         attribute.
2437
2438         * DumpRenderTree/efl/ImageDiff.cpp:
2439         (printImage):
2440
2441 2012-03-07  Kangil Han  <kangil.han@samsung.com>
2442
2443         [DRT] Remove PlainTextController implementations.
2444         https://bugs.webkit.org/show_bug.cgi?id=79959
2445
2446         Reviewed by Hajime Morita.
2447
2448         PlainTextController usages in existing tests have been
2449         replaced by internals API by bug 78570.
2450         So this patch will remove PlainTextController implementations
2451         to avoid further usage in new tests.
2452
2453         * DumpRenderTree/DumpRenderTree.gypi:
2454         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
2455         * DumpRenderTree/chromium/PlainTextController.cpp: Removed.
2456         * DumpRenderTree/chromium/PlainTextController.h: Removed.
2457         * DumpRenderTree/chromium/TestShell.cpp:
2458         (TestShell::TestShell):
2459         (TestShell::bindJSObjectsToWindow):
2460         * DumpRenderTree/chromium/TestShell.h:
2461         (TestShell):
2462         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2463         (webViewWindowObjectCleared):
2464         * DumpRenderTree/gtk/PlainTextController.cpp: Removed.
2465         * DumpRenderTree/gtk/PlainTextController.h: Removed.
2466         * DumpRenderTree/mac/FrameLoadDelegate.mm:
2467         (-[FrameLoadDelegate didClearWindowObjectInStandardWorldForFrame:]):
2468         * DumpRenderTree/mac/PlainTextController.h: Removed.
2469         * DumpRenderTree/mac/PlainTextController.mm: Removed.
2470         * DumpRenderTree/qt/DumpRenderTree.pro:
2471         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
2472         (WebCore::DumpRenderTree::DumpRenderTree):
2473         (WebCore::DumpRenderTree::initJSObjects):
2474         * DumpRenderTree/qt/DumpRenderTreeQt.h:
2475         (DumpRenderTree):
2476         * DumpRenderTree/qt/PlainTextControllerQt.cpp: Removed.
2477         * DumpRenderTree/qt/PlainTextControllerQt.h: Removed.
2478         * GNUmakefile.am:
2479
2480 2012-03-06  Grzegorz Czajkowski  <g.czajkowski@samsung.com>
2481
2482         [EFL][DRT] Implements LayoutTestController's setPopupBlockingEnabled.
2483         https://bugs.webkit.org/show_bug.cgi?id=80137
2484
2485         Reviewed by Eric Seidel.
2486
2487         Adds implementation LayoutTestController's setPopupBlockingEnabled.
2488         To pass popup-blocking-click-in-iframe.html EventSender should not
2489         repeat click events in idler's function. EventSender allows to
2490         asynchronously send click event through ecore_idler_add.
2491         This event will be called repeatedly as long as it return true.
2492         Click event should be called only once by returning ECORE_CALLBACK_CANCEL.
2493
2494         * DumpRenderTree/efl/EventSender.cpp:
2495         (sendMouseEvent): Changed return value from bool to void because it always returns true.
2496         (sendClick): Idler is deleted by returning ECORE_CALLBACK_CANCEL.
2497         (mouseDownCallback): Remove unnecessary checking of return value.
2498         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
2499         (LayoutTestController::setPopupBlockingEnabled):
2500
2501 2012-03-06  Dirk Pranke  <dpranke@chromium.org>
2502
2503         rebaseline_chromium_webkit_tests is obsolete and should be removed
2504         https://bugs.webkit.org/show_bug.cgi?id=80347
2505
2506         Reviewed by Tony Chang.
2507
2508         * Scripts/rebaseline-chromium-webkit-tests: Removed.
2509         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py: Removed.
2510         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests_unittest.py: Removed.
2511
2512 2012-03-06  Dirk Pranke  <dpranke@chromium.org>
2513
2514         test_start_cmd (webkitpy.layout_tests.servers.http_server_unittest.TestHttpServer) is broken
2515         https://bugs.webkit.org/show_bug.cgi?id=80468
2516
2517         Unreviewed, build fix.
2518
2519         Fixing a bug revealed by r80443 when we fixed filesystem.copyfile().
2520
2521         * Scripts/webkitpy/layout_tests/servers/http_server.py:
2522         (Lighttpd._prepare_config):
2523
2524 2012-03-06  Dirk Pranke  <dpranke@chromium.org>
2525
2526         remove chromium_gpu ports
2527         https://bugs.webkit.org/show_bug.cgi?id=80345
2528
2529         Reviewed by Ojan Vafai.
2530
2531         This change leaves in the 'graphics_type' distinction in the
2532         test_expectations; removing that will require more updates and a
2533         pass over test_expectations.txt, so I'm separating that out.
2534
2535         * Scripts/webkitpy/layout_tests/port/builders.py:
2536         * Scripts/webkitpy/layout_tests/port/chromium.py:
2537         (ChromiumPort):
2538         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py: Removed.
2539         * Scripts/webkitpy/layout_tests/port/chromium_gpu_unittest.py: Removed.
2540         * Scripts/webkitpy/layout_tests/port/factory.py:
2541         (PortFactory):
2542         (PortFactory.get):
2543         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
2544         (FactoryTest.test_qt):
2545         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
2546         (TestRebaseline.test_rebaseline_expectations):
2547
2548 2012-03-06  Raphael Kubo da Costa  <kubo@profusion.mobi>
2549
2550         undefined reference to JSC::IdentifierTable::~IdentifierTable() on EFL port
2551         https://bugs.webkit.org/show_bug.cgi?id=80282
2552
2553         Reviewed by Antonio Gomes.
2554
2555         Remove transitive library dependencies; they are especially
2556         dangerous when one ends up linking against WTF, JSC and then WTF
2557         again, since some symbols will not be defined.
2558
2559         Passing --no-copy-dt-needed-entries and --as-needed to the linker
2560         (which some recent Linux distros do by default) makes the issue
2561         even more evident.
2562
2563         * DumpRenderTree/efl/CMakeLists.txt: Do not link directly to WTF,
2564         as the other libraries have the needed symbols. Explicitly link
2565         against fontconfig.
2566
2567 2012-03-06  Dean Jackson  <dino@apple.com>
2568
2569         filter-build-webkit needs more filters
2570         https://bugs.webkit.org/show_bug.cgi?id=80440
2571
2572         Reviewed by Joseph Pecoraro.
2573
2574         Added as many missing filters as I could find for Apple
2575         builds. Plus a few minor style updates.
2576
2577         * Scripts/filter-build-webkit:
2578         (printLine):
2579
2580 2012-03-06  Ojan Vafai  <ojan@chromium.org>
2581
2582         Copying old baselines throws an error if the new directory doesn't exist
2583         https://bugs.webkit.org/show_bug.cgi?id=80443
2584
2585         Reviewed by Adam Barth.
2586
2587         Existing tests cover the new code with the fix to filesystem_mock.
2588
2589         * Scripts/webkitpy/common/system/filesystem_mock.py:
2590         (MockFileSystem.copyfile):
2591         Fix the mock to better match python's copyfile semantics.
2592
2593         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
2594         (Rebaseliner._create_html_baseline_files):
2595         * Scripts/webkitpy/tool/commands/rebaseline.py:
2596         (RebaselineTest._copy_existing_baseline):
2597
2598 2012-03-05  Ojan Vafai  <ojan@chromium.org>
2599
2600         Add a mechanism to rebaseline new ports
2601         https://bugs.webkit.org/show_bug.cgi?id=80355
2602
2603         Reviewed by Adam Barth.
2604
2605         For a test that's only failing on a new port, we want
2606         to first copy the existing result into the location for the
2607         port it's replacing in order to not break that port.
2608
2609         For example, bringing up the chromium-lion port, if we just
2610         stick the results in the chromium-mac directory, the snow leopard
2611         will start failing. Instead, we first copy the existing result
2612         to the chromium-mac-snowleopard directory.
2613
2614         * Scripts/webkitpy/layout_tests/port/builders.py:
2615         (builder_path_for_port_name):
2616         (fallback_port_name_for_new_port):
2617         * Scripts/webkitpy/tool/commands/rebaseline.py:
2618         (RebaselineTest):
2619         (RebaselineTest._copy_existing_baseline):
2620         (RebaselineTest._rebaseline_test):
2621         (RebaselineTest.execute):
2622         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
2623         (TestRebaseline.test_rebaseline_test):
2624         (TestRebaseline):
2625         (TestRebaseline.test_rebaseline_and_copy_test):
2626         (test_rebaseline_and_copy_test_no_existing_result):
2627         (test_rebaseline_and_copy_test_with_lion_result):
2628         (test_rebaseline_and_copy_no_overwrite_test):
2629         * Scripts/webkitpy/tool/servers/gardeningserver.py:
2630         (GardeningHTTPRequestHandler.rebaseline):
2631         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
2632         (GardeningServerTest.test_rebaseline):
2633         (GardeningServerTest):
2634         (GardeningServerTest.test_rebaseline_new_port):
2635
2636 2012-03-06  Beth Dakin  <bdakin@apple.com>
2637
2638         https://bugs.webkit.org/show_bug.cgi?id=80351
2639         InlineTextBox is using the wrong origin when calling 
2640         addRelevantRepaintedObject
2641         -and corresponding-
2642         <rdar://problem/10970221>
2643
2644         Reviewed by Sam Weinig.
2645
2646         This test has actually been broken since revision 109273. This change makes 
2647         the test pass again, and it makes it specifically exercise the fix for this 
2648         bug.
2649         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
2650         * TestWebKitAPI/Tests/WebKit2/NewFirstVisuallyNonEmptyLayout.cpp:
2651         (TestWebKitAPI::TEST):
2652
2653         Adjust comments in the other tests to reflect the API changes from 109273.
2654         * TestWebKitAPI/Tests/WebKit2/NewFirstVisuallyNonEmptyLayoutFails.cpp:
2655         (TestWebKitAPI::TEST):
2656         * TestWebKitAPI/Tests/WebKit2/NewFirstVisuallyNonEmptyLayoutFails_Bundle.cpp:
2657         (TestWebKitAPI::NewFirstVisuallyNonEmptyLayoutFailsTest::didCreatePage):
2658         * TestWebKitAPI/Tests/WebKit2/NewFirstVisuallyNonEmptyLayoutFrames_Bundle.cpp:
2659         (TestWebKitAPI::NewFirstVisuallyNonEmptyLayoutFramesTest::didCreatePage):
2660         * TestWebKitAPI/Tests/WebKit2/NewFirstVisuallyNonEmptyLayout_Bundle.cpp:
2661         (TestWebKitAPI::NewFirstVisuallyNonEmptyLayoutTest::didCreatePage):
2662         * TestWebKitAPI/Tests/WebKit2/lots-of-text.html: Added.
2663
2664 2012-03-06  Hans Wennborg  <hans@chromium.org>
2665
2666         Speech JavaScript API: add SpeechRecognitionError
2667         https://bugs.webkit.org/show_bug.cgi?id=80410
2668
2669         Chromium: enable the runtime flag for scripted speech so we can run
2670         the layout tests for it.
2671
2672         Reviewed by Adam Barth.
2673
2674         * DumpRenderTree/chromium/TestShell.cpp:
2675         (TestShell::TestShell):
2676
2677 2012-03-06  Sami Kyostila  <skyostil@chromium.org>
2678
2679         Enable universal newlines in chromium android layout test runner
2680         https://bugs.webkit.org/show_bug.cgi?id=80418
2681
2682         Reviewed by Adam Barth.
2683
2684         The Chromium Android layout test runner uses the 'adb' program to
2685         communicate with the Android device. Adb uses CRLF ('\r\n') to delimit
2686         output lines from the test runner, while the test expectation files are
2687         written with LF ('\n') delimiters. This causes the results to have
2688         spurious whitespace differences, making them more difficult to
2689         interpret.
2690
2691         The fix is to enable universal newlines for the DumpRenderTree pipe,
2692         causing the output to only contain LF newlines like on other platforms.
2693
2694         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2695         (ChromiumAndroidDriver._start):
2696
2697 2012-03-06  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2698
2699         [Qt] Revert r109900, it was completly busted
2700
2701         HAVE_FOO does _not_ have a WTF_ prefix, go figure.
2702
2703         We still leave in the HAVE(FOO) in QtInitializeTestFonts.cpp, instead
2704         of manual define checks.
2705
2706         Reviewed by Ossy.
2707
2708         * DumpRenderTree/qt/DumpRenderTree.pro:
2709         * QtTestBrowser/QtTestBrowser.pro:
2710         * WebKitTestRunner/InjectedBundle/Target.pri:
2711         * qmake/mkspecs/features/features.prf:
2712
2713 2012-03-05  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2714
2715         [Qt] Use config test to determine if the compiler supports -MD
2716
2717         We used to assume that unix systems building against Qt5 would have
2718         a compiler that supported -MD. Although this was usually the case,
2719         we failed to take into account broken tools for distributed builds
2720         (e.g. teambuilder or icecream) which would end up producing .d files
2721         in the root of the build dir instead of along-side the object file.
2722
2723         Now that we use a config-test to determine the support, we'll capture
2724         those cases, unless distributed builds are enabled after the initial
2725         build-webkit run.
2726
2727         Reviewed by Csaba Osztrogonác.
2728
2729         * Tools.pro:
2730         * qmake/config.tests/gccdepends/empty.cpp: Added.
2731         * qmake/config.tests/gccdepends/gccdepends.pro: Added.
2732         * qmake/mkspecs/features/default_pre.prf:
2733         * qmake/mkspecs/features/features.prf:
2734         * qmake/mkspecs/features/unix/default_pre.prf:
2735         * qmake/sync.profile:
2736
2737 2012-03-06  Csaba Osztrogonác  <ossy@webkit.org>
2738
2739         REGRESSION(r109294): It broke the unittest of master.cfg
2740         https://bugs.webkit.org/show_bug.cgi?id=80253
2741
2742         Reviewed by Tor Arne Vestbø.
2743
2744         * BuildSlaveSupport/build.webkit.org-config/mastercfg_unittest.py: Use json instead of simplejson.
2745         (BuildBotConfigLoader._create_mock_passwords_dict):
2746         (BuildBotConfigLoader._mock_open):
2747         (BuildBotConfigLoader._add_dependant_modules_to_sys_modules):
2748         (BuildBotConfigLoader.load_config):
2749
2750 2012-03-06  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
2751
2752         Qt API test buildstep throws exception after buildmaster update
2753         https://bugs.webkit.org/show_bug.cgi?id=80249
2754
2755         This patch based on Peter Gal's work.
2756
2757         Reviewed by Csaba Osztrogonác.
2758
2759         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2760         (RunQtAPITests):
2761         (BuildAndTestFactory.__init__): Enable API tests on the Qt platform again.
2762
2763 2012-03-06  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2764
2765         [Qt] Use correct define names for HAVE(FOO) style optional dependencies
2766
2767         Reviewed by Simon Hausmann.
2768         Reviewed by Ossy.
2769
2770         * DumpRenderTree/qt/DumpRenderTree.pro:
2771         * DumpRenderTree/qt/QtInitializeTestFonts.cpp:
2772         (WebKit::initializeTestFonts):
2773         * QtTestBrowser/QtTestBrowser.pro:
2774         * WebKitTestRunner/InjectedBundle/Target.pri:
2775         * qmake/mkspecs/features/features.prf:
2776
2777 2012-03-06  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2778
2779         [Qt] Don't add recursive qmake_all target if there are no subdirs to recurse
2780
2781         Reviewed by Ossy.
2782
2783         * qmake/mkspecs/features/default_post.prf:
2784
2785 2012-03-06  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2786
2787         [Qt] Make sure clean builds are followed by make qmake && make if needed
2788
2789         On non-GNUmake systems we have to run make qmake so that the generated
2790         sources are generated before we run qmake on the target pro files. This
2791         is achieved by using the 'incremental' target.
2792
2793         Reviewed by Ossy.
2794
2795         * Scripts/webkitdirs.pm:
2796         (buildQMakeProjects):
2797
2798 2012-03-06  Filip Pizlo  <fpizlo@apple.com>
2799
2800         DumpRenderTree should support preciseTime()
2801         https://bugs.webkit.org/show_bug.cgi?id=80397
2802
2803         Reviewed by Gavin Barraclough.
2804
2805         * DumpRenderTree/LayoutTestController.cpp:
2806         (preciseTimeCallback):
2807         (LayoutTestController::staticFunctions):
2808
2809 2012-03-06  Sheriff Bot  <webkit.review.bot@gmail.com>
2810
2811         Unreviewed, rolling out r109887.
2812         http://trac.webkit.org/changeset/109887
2813         https://bugs.webkit.org/show_bug.cgi?id=80402
2814
2815         It broke the build (Requested by Ossy on #webkit).
2816
2817         * DumpRenderTree/qt/DumpRenderTree.pro:
2818         * DumpRenderTree/qt/QtInitializeTestFonts.cpp:
2819         (WebKit::initializeTestFonts):
2820         * QtTestBrowser/QtTestBrowser.pro:
2821         * WebKitTestRunner/InjectedBundle/Target.pri:
2822         * qmake/mkspecs/features/features.prf:
2823
2824 2012-03-06  Sheriff Bot  <webkit.review.bot@gmail.com>
2825
2826         Unreviewed, rolling out r109837.
2827         http://trac.webkit.org/changeset/109837
2828         https://bugs.webkit.org/show_bug.cgi?id=80399
2829
2830         breaks Mac Productions builds, too late to try and fix it
2831         tonight (Requested by eseidel on #webkit).
2832
2833         * Scripts/build-webkit:
2834
2835 2012-03-06  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2836
2837         [Qt] Use correct define names for HAVE(FOO) style optional dependencies
2838
2839         Reviewed by Simon Hausmann.
2840
2841         * DumpRenderTree/qt/DumpRenderTree.pro:
2842         * DumpRenderTree/qt/QtInitializeTestFonts.cpp:
2843         (WebKit::initializeTestFonts):
2844         * QtTestBrowser/QtTestBrowser.pro:
2845         * WebKitTestRunner/InjectedBundle/Target.pri:
2846         * qmake/mkspecs/features/features.prf:
2847
2848 2012-03-06  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
2849
2850         [Qt] Create EWS to build WebKit2
2851         https://bugs.webkit.org/show_bug.cgi?id=80277
2852
2853         Reviewed by Csaba Osztrogonác.
2854
2855         * QueueStatusServer/model/queues.py:
2856         (Queue):
2857         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
2858         (QtWK2EWS):
2859         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
2860         (EarlyWarningSytemTest.test_builder_ewses):
2861
2862 2012-03-05  Sheriff Bot  <webkit.review.bot@gmail.com>
2863
2864         Unreviewed, rolling out r109733.
2865         http://trac.webkit.org/changeset/109733
2866         https://bugs.webkit.org/show_bug.cgi?id=80378
2867
2868         It broke Qt5 build system (Requested by Ossy on #webkit).
2869
2870         * Tools.pro:
2871         * qmake/config.tests/gccdepends/empty.cpp: Removed.
2872         * qmake/config.tests/gccdepends/gccdepends.pro: Removed.
2873         * qmake/mkspecs/features/default_pre.prf:
2874         * qmake/mkspecs/features/features.prf:
2875         * qmake/mkspecs/features/unix/default_pre.prf:
2876         * qmake/sync.profile:
2877
2878 2012-03-05  Eric Seidel  <eric@webkit.org>
2879
2880         Update JavaScriptCore files to use fully-qualified WTF include path
2881         https://bugs.webkit.org/show_bug.cgi?id=79960
2882
2883         Reviewed by Adam Barth.
2884
2885         Build WTF/WTF.xcodeproj by default on Mac.
2886
2887         * Scripts/build-webkit:
2888
2889 2012-03-05  Dirk Pranke  <dpranke@chromium.org>
2890
2891         There should be a way to disable optimizer in webkit-patch rebaseline-expectations
2892         https://bugs.webkit.org/show_bug.cgi?id=69590
2893
2894         Reviewed by Ryosuke Niwa.
2895
2896         Adds a --no-optimize flag to webkit-patch rebaseline-expectations
2897         to skip the 'optimize-expectations' step.
2898
2899         * Scripts/webkitpy/tool/commands/rebaseline.py:
2900         (RebaselineExpectations.__init__):
2901         (RebaselineExpectations.execute):
2902         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
2903
2904 2012-03-05  Kent Tamura  <tkent@chromium.org>
2905
2906         [Chromium] Add new popup type: PagePopup
2907         https://bugs.webkit.org/show_bug.cgi?id=80106
2908
2909         Reviewed by Darin Fisher.
2910
2911         * DumpRenderTree/chromium/WebViewHost.cpp:
2912         (WebViewHost::createPopupMenu):
2913
2914 2012-03-05  Jessie Berlin  <jberlin@apple.com>
2915
2916         Add myself as a reviewer.
2917
2918         * Scripts/webkitpy/common/config/committers.py:
2919         Also, add my IRC nick so it is easier to find and yell at me.
2920
2921 2012-03-05  Robert Kroeger  <rjkroege@chromium.org>
2922
2923         Correctly compute the positions of GestureScroll* events on
2924         Chromium.
2925         https://bugs.webkit.org/show_bug.cgi?id=80075
2926
2927         Reviewed by Adam Barth.
2928
2929         * DumpRenderTree/chromium/EventSender.cpp:
2930         (EventSender::reset):
2931         (EventSender::gestureEvent):
2932         * DumpRenderTree/chromium/EventSender.h:
2933         (EventSender):
2934
2935 2012-03-05  Dirk Pranke  <dpranke@chromium.org>
2936
2937         flakiness dashboard doesn't display baselines for virtual tests correctly
2938         https://bugs.webkit.org/show_bug.cgi?id=80341
2939
2940         Reviewed by Ojan Vafai.
2941
2942         If the test is part of a virtual test suite, we need to look for
2943         baselines in the underlying fallback path for the actual test
2944         being run, as well as baselines for the virtual version.
2945
2946         This patch adds code to handle both branches and also pulls the
2947         mapping of the test suites up into a constant.
2948
2949         * TestResultServer/static-dashboards/flakiness_dashboard.html:
2950         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
2951         (testLookupVirtualTestSuite):
2952         (testBaseTest):
2953
2954 2012-03-05  Sheriff Bot  <webkit.review.bot@gmail.com>
2955
2956         Unreviewed, rolling out r109760.
2957         http://trac.webkit.org/changeset/109760
2958         https://bugs.webkit.org/show_bug.cgi?id=80320
2959
2960         Caused many GTK+ tests to crash (Requested by mrobinson on
2961         #webkit).
2962
2963         * gtk/jhbuild.modules:
2964
2965 2012-03-05  Charlie Reis  <creis@chromium.org>
2966
2967         [chromium] Pass WebNavigationPolicy to WebViewClient::createView
2968         https://bugs.webkit.org/show_bug.cgi?id=80057
2969
2970         Reviewed by Darin Fisher.
2971
2972         * DumpRenderTree/chromium/WebViewHost.cpp:
2973         (WebViewHost::createView):
2974         * DumpRenderTree/chromium/WebViewHost.h:
2975         (WebViewHost):
2976
2977 2012-03-05  Martin Robinson  <mrobinson@igalia.com>
2978
2979         [soup] Crash while loading http://www.jusco.cn
2980         https://bugs.webkit.org/show_bug.cgi?id=68238
2981
2982         Reviewed by Philippe Normand.
2983
2984         * gtk/jhbuild.modules: Bumped the libsoup and glib dependencies
2985         in the jhbuild file.
2986
2987 2012-03-05  Sheriff Bot  <webkit.review.bot@gmail.com>
2988
2989         Unreviewed, rolling out r109748.
2990         http://trac.webkit.org/changeset/109748
2991         https://bugs.webkit.org/show_bug.cgi?id=80296
2992
2993         Made some tests crash, will fix and recommit (Requested by
2994         noamr on #webkit).
2995
2996         * MiniBrowser/qt/main.cpp:
2997         (main):
2998
2999 2012-03-05  Zan Dobersek  <zandobersek@gmail.com>
3000
3001         [GTK] plugins/netscape-plugin-page-cache-works.html fails
3002         https://bugs.webkit.org/show_bug.cgi?id=74409
3003
3004         Reviewed by Martin Robinson.
3005
3006         Add support for overriding the page-cache-for-plugins preference
3007         through DumpRenderTreeSupport.
3008
3009         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
3010         (booleanFromValue): A helper function to convert a string to boolean.
3011         (LayoutTestController::overridePreference):
3012
3013 2012-03-05  Rob Buis  <rbuis@rim.com>
3014
3015         [BlackBerry] Update Tools/DumpRenderTree
3016         https://bugs.webkit.org/show_bug.cgi?id=80287
3017
3018         Reviewed by Antonio Gomes.
3019
3020         Update so we are sure we can build our implementation of DRT.
3021
3022         * DumpRenderTree/DumpRenderTree.h: include DumpRenderTreeBlackBerry.h.
3023         * DumpRenderTree/PixelDumpSupport.cpp: include PixelDumpSupportBlackBerry.h.
3024         * DumpRenderTree/blackberry/AccessibilityControllerBlackBerry.cpp:
3025         (AccessibilityController::addNotificationListener): add missing hook.
3026         (AccessibilityController::removeNotificationListener): add missing hook.
3027         * DumpRenderTree/blackberry/AccessibilityUIElementBlackBerry.cpp:
3028         (AccessibilityUIElement::isSelectedOptionActive): add missing hook.
3029         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
3030         (LayoutTestController::setMockSpeechInputDumpRect): add missing hook.
3031         (LayoutTestController::simulateDesktopNotificationClick): add missing hook.
3032
3033 2012-03-05  No'am Rosenthal  <noam.rosenthal@nokia.com>
3034
3035         [Qt] [WK2] Support threaded renderer in WK2
3036         https://bugs.webkit.org/show_bug.cgi?id=76661
3037
3038         Remove the QML_NO_THREADED_RENDERER environment variable from MiniBrowser.
3039
3040         Reviewed by Kenneth Rohde Christiansen.
3041
3042         * MiniBrowser/qt/main.cpp:
3043         (main):
3044
3045 2012-03-05  Alexander Færøy  <alexander.faeroy@nokia.com>
3046
3047         Unreviewed. Add myself to committer list.
3048
3049         * Scripts/webkitpy/common/config/committers.py:
3050
3051 2012-03-05  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3052
3053         [Qt] Use config test to determine if the compiler supports -MD
3054
3055         We used to assume that unix systems building against Qt5 would have
3056         a compiler that supported -MD. Although this was usually the case,
3057         we failed to take into account broken tools for distributed builds
3058         (e.g. teambuilder or icecream) which would end up producing .d files
3059         in the root of the build dir instead of along-side the object file.
3060
3061         Now that we use a config-test to determine the support, we'll capture
3062         those cases, unless distributed builds are enabled after the initial
3063         build-webkit run.
3064
3065         Reviewed by Csaba Osztrogonác.
3066
3067         * Tools.pro:
3068         * qmake/config.tests/gccdepends/empty.cpp: Added.
3069         * qmake/config.tests/gccdepends/gccdepends.pro: Added.
3070         * qmake/mkspecs/features/default_pre.prf:
3071         * qmake/mkspecs/features/features.prf:
3072         * qmake/mkspecs/features/unix/default_pre.prf:
3073         * qmake/sync.profile:
3074
3075 2012-03-05  Csaba Osztrogonác  <ossy@webkit.org>
3076
3077         Qt API test buildstep throws exception after buildmaster update
3078         https://bugs.webkit.org/show_bug.cgi?id=80249
3079
3080         Reviewed by Zoltan Herczeg.
3081
3082         * BuildSlaveSupport/build.webkit.org-config/master.cfg: Disable API tests on the Qt bot temporarily.
3083         (BuildAndTestFactory.__init__):
3084
3085 2012-03-04  Martin Robinson  <mrobinson@igalia.com>
3086
3087         [GTK] Unit tests are noisy when run under jhbuild
3088         https://bugs.webkit.org/show_bug.cgi?id=80241
3089
3090         Reviewed by Philippe Normand.
3091
3092         Force the 'memory' backend in GSettings so that each unit
3093         test doesn't produce a warning when it's forced to fall back
3094         to the 'memory' backend.
3095
3096         * Scripts/run-gtk-tests:
3097         (TestRunner.run): Force the 'memory' backend.
3098
3099 2012-03-04  Martin Robinson  <mrobinson@igalia.com>
3100
3101         [GTK] "Console: undefined @0: ReferenceError: Can't find variable: WebInspector" during testwebinspector
3102         https://bugs.webkit.org/show_bug.cgi?id=80235
3103
3104         Reviewed by Philippe Normand.
3105
3106         * Scripts/run-gtk-tests:
3107         (TestRunner.run): Use the WEBKIT_INSPECTOR_PATH environment variable so that
3108         inspector resources path determination is shared by all tests.
3109
3110 2012-03-04  Martin Robinson  <mrobinson@igalia.com>
3111
3112         [GTK] REGRESSION: r108278 broke some tests that expect focus
3113         https://bugs.webkit.org/show_bug.cgi?id=79084
3114
3115         Reviewed by Philippe Normand.
3116
3117         Grab focus when dumping test results and also when sending
3118         events. During the course of testing, the WebView may lose
3119         focus and this seems to snap it back when tests demand it.
3120
3121         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3122         (dump): Grab focus.
3123         * DumpRenderTree/gtk/EventSender.cpp:
3124         (dispatchEvent): Ditto.
3125
3126 2012-03-04  Charles Wei  <charles.wei@torchmobile.com.cn>
3127
3128         Unreviewed, Add myself to committer list.
3129         https://bugs.webkit.org/show_bug.cgi?id=80232
3130
3131         * Scripts/webkitpy/common/config/committers.py:
3132
3133 2012-03-04  Anders Carlsson  <andersca@apple.com>
3134
3135         Newer versions of clang fail to compile WebKit.
3136
3137         Fix build.
3138
3139         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
3140         (descriptionOfValue):
3141         (AccessibilityUIElement::rowIndexRange):
3142         (AccessibilityUIElement::columnIndexRange):
3143         (AccessibilityUIElement::selectedTextRange):
3144         * DumpRenderTree/mac/EditingDelegate.mm:
3145         (-[DOMRange dump]):
3146         * DumpRenderTree/mac/PolicyDelegate.mm:
3147         (-[PolicyDelegate webView:unableToImplementPolicyWithError:frame:]):
3148         * DumpRenderTree/mac/ResourceLoadDelegate.mm:
3149         (-[NSError _drt_descriptionSuitableForTestResult]):
3150         * WebKitTestRunner/InjectedBundle/mac/AccessibilityUIElementMac.mm:
3151         (WTR::descriptionOfValue):
3152         (WTR::AccessibilityUIElement::rowIndexRange):
3153         (WTR::AccessibilityUIElement::columnIndexRange):
3154         (WTR::AccessibilityUIElement::selectedTextRange):
3155
3156 2012-03-03  Simon Hausmann  <simon.hausmann@nokia.com>
3157
3158         [Qt] Fix static_libs_as_shared build
3159         https://bugs.webkit.org/show_bug.cgi?id=80214
3160
3161         Reviewed by Tor Arne Vestbø.
3162
3163         * WebKitTestRunner/qt/main.cpp:
3164         (main): Set the correct environment variable for choosing the theme in the
3165         web process.
3166
3167 2012-03-03  Hans Wennborg  <hans@chromium.org>
3168
3169         Implement Speech JavaScript API
3170         https://bugs.webkit.org/show_bug.cgi?id=80019
3171
3172         Reviewed by Adam Barth.
3173
3174         Add ENABLE_SCRIPTED_SPEECH.
3175
3176         * Scripts/build-webkit:
3177         * qmake/mkspecs/features/features.prf:
3178
3179 2012-03-02  Alexey Proskuryakov  <ap@apple.com>
3180
3181         Fix commit queue, remove duplicate entry.
3182
3183         * Scripts/webkitpy/common/config/committers.py:
3184
3185 2012-03-02  Stephen Chenney  <schenney@chromium.org>
3186
3187         Unreviewed, add myself as a committer
3188
3189         * Scripts/webkitpy/common/config/committers.py:
3190
3191 2012-03-02  Ojan Vafai  <ojan@chromium.org>
3192
3193         Add the Chromium Lion canary bot to the flakiness dashboard.
3194         * TestResultServer/static-dashboards/builders.js:
3195
3196 2012-03-02  Dirk Pranke  <dpranke@chromium.org>
3197
3198         links for virtual tests are broken in the flakiness dashboard
3199         https://bugs.webkit.org/show_bug.cgi?id=80180
3200
3201         Reviewed by Ojan Vafai.
3202
3203         * TestResultServer/static-dashboards/flakiness_dashboard.html:
3204         (linkHTMLToOpenWindow):
3205         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
3206         (testLinkHTMLToOpenWindow): added.
3207
3208 2012-03-02  Philippe Normand  <pnormand@igalia.com>
3209
3210         [GTK] security/block-test.html baselines differ from mac
3211         https://bugs.webkit.org/show_bug.cgi?id=78343
3212
3213         Reviewed by Martin Robinson.
3214
3215         Improved the FrameLoader error heuristics to match mac behavior a
3216         bit more.
3217
3218         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3219         (pathFromSoupURI): Handle ftp urls the same as http urls.
3220         (descriptionSuitableForTestResult): Map WebKitPolicyError to WebKitErrorDomain.
3221
3222 2012-03-02  Andras Becsi  <andras.becsi@nokia.com>
3223
3224         [Qt][WK2] Fix typo in patch committed in r109575
3225
3226         Unreviewed typo fix.
3227
3228         * MiniBrowser/qt/MiniBrowserApplication.cpp:
3229         (MiniBrowserApplication::notify): Cache the position
3230         in floating point coordinates.
3231
3232 2012-03-02  Andras Becsi  <andras.becsi@nokia.com>
3233
3234         [Qt][WK2] Infinite loop on history navigation, when panning
3235         https://bugs.webkit.org/show_bug.cgi?id=79119
3236
3237         Reviewed by Simon Hausmann.
3238
3239         Make the touch mocking more robust by creating touch events
3240         directly and sending them synchronously to the browser window.
3241         If the sent touch event was not accepted we propagate the
3242         original mouse event to its target.
3243         This way we make sure we only send either a touch or a mouse
3244         event and do not end up in an infinite loop between our mocking
3245         code and the Qt5 automatic mouse event synthesizing.
3246
3247         * MiniBrowser/qt/BrowserWindow.cpp:
3248         (BrowserWindow::updateVisualMockTouchPoints):
3249         * MiniBrowser/qt/BrowserWindow.h:
3250         (BrowserWindow):
3251         * MiniBrowser/qt/MiniBrowserApplication.cpp:
3252         (touchRectForPosition):
3253         (MiniBrowserApplication::notify):
3254         (MiniBrowserApplication::sendTouchEvent):
3255         * MiniBrowser/qt/MiniBrowserApplication.h:
3256         (MiniBrowserApplication):
3257
3258 2012-03-02  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3259
3260         [Qt] Use 'all' as default target when debug_and_release is in effect
3261
3262         When debug_and_release is enabled, in combination with build_all,
3263         qmake will generate an aggregate makefile that calls make on both
3264         the debug and the release makefiles, but with the 'all' target.
3265
3266         In this case we can't use 'first' as the default target, as the
3267         extra targets would then not be run, so we append extra targets
3268         to 'all' instead.
3269
3270         Reviewed by Simon Hausmann.
3271
3272         * qmake/mkspecs/features/default_post.prf:
3273
3274 2012-03-02  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3275
3276         [Qt] Make sure activeBuildConfig() returns the right config
3277
3278         Using isActiveConfig fails, as that only checks if the given
3279         argument is in CONFIG, not the order.
3280
3281         Reviewed by Simon Hausmann.
3282
3283         * qmake/mkspecs/features/functions.prf:
3284
3285 2012-03-02  Simon Hausmann  <simon.hausmann@nokia.com>
3286
3287         [Qt] Fix tests run with WTR not using QStyle theme
3288         https://bugs.webkit.org/show_bug.cgi?id=80147
3289
3290         Reviewed by Csaba Osztrogonác.
3291
3292         * WebKitTestRunner/qt/main.cpp:
3293         (main): Set environment variable for web process to use QStyle theme.
3294
3295 2012-03-02  Philippe Normand  <pnormand@igalia.com>
3296
3297         Unreviewed, remove debug print call.
3298
3299         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3300         (descriptionSuitableForTestResult):
3301
3302 2012-03-02  Robert Kroeger  <rjkroege@chromium.org>
3303
3304         Updated EventSender to send complete gesture events.
3305         https://bugs.webkit.org/show_bug.cgi?id=80075
3306
3307         Reviewed by Adam Barth.
3308
3309         * DumpRenderTree/chromium/EventSender.cpp:
3310         (EventSender::gestureEvent): set additional parameters in sent events.
3311
3312 2012-03-01  Kalev Lember  <kalevlember@gmail.com>
3313
3314         WebKitTestRunner: Explicitly include unistd.h for gcc 4.7
3315         https://bugs.webkit.org/show_bug.cgi?id=80033
3316
3317         Reviewed by Gustavo Noronha Silva.
3318
3319         GCC 4.7 no longer includes unistd.h from most of the standard C++
3320         header files, so we need to explicitly include it for getcwd().
3321
3322         * WebKitTestRunner/TestInvocation.cpp: Make sure unistd.h gets included
3323         not only on Mac, but on all Unix platforms.
3324
3325 2012-03-01  Dirk Pranke  <dpranke@chromium.org>
3326
3327         Actually look at the args in driver_input :(.
3328
3329         Unreviewed, build fix.
3330
3331         * Scripts/webkitpy/layout_tests/port/chromium.py:
3332         (ChromiumDriver.run_test):
3333
3334 2012-03-01  Dirk Pranke  <dpranke@chromium.org>
3335
3336         Fix typo in flag used to run gpu virtual suites.
3337
3338         Unreviewed, build fix.
3339
3340         * Scripts/webkitpy/layout_tests/port/chromium.py:
3341         (ChromiumPort.virtual_test_suites):
3342
3343 2012-03-01  Adam Barth  <abarth@webkit.org>
3344
3345         Unreviewed, rolling out r109317.
3346         http://trac.webkit.org/changeset/109317
3347
3348         Let's try watchlists again now that we're using a pure git
3349         working copy
3350
3351         * Scripts/webkitpy/tool/bot/stylequeuetask.py:
3352         (StyleQueueTask.run):
3353         * Scripts/webkitpy/tool/commands/queues_unittest.py:
3354
3355 2012-03-01  Dirk Pranke  <dpranke@chromium.org>
3356
3357         webkitpy: test_run_test_pause_before_testing is failing after r109482
3358         https://bugs.webkit.org/show_bug.cgi?id=80084
3359
3360         Unreviewed, build fix.
3361
3362         Fixes the test string compare to ignore the time spent running
3363         the tests.
3364
3365         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
3366         (test_run_test_pause_before_testing):
3367
3368 2012-03-01  Dirk Pranke  <dpranke@chromium.org>
3369
3370         nrwt: virtual tests aren't being run
3371         https://bugs.webkit.org/show_bug.cgi?id=80083
3372
3373         Reviewed by Eric Seidel.
3374
3375         If we tried to run all of the tests by default (and
3376         didn't specify any paths), the virtual tests weren't
3377         picked up.
3378
3379         * Scripts/webkitpy/layout_tests/port/base.py:
3380         (Port._virtual_tests):
3381         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
3382         (PortTest.test_tests):
3383         * Scripts/webkitpy/layout_tests/port/test.py:
3384         (TestPort.virtual_test_suites):
3385
3386 2012-03-01  Ryosuke Niwa  <rniwa@webkit.org>
3387
3388         run-perf-tests should report the test runtime
3389         https://bugs.webkit.org/show_bug.cgi?id=80078
3390
3391         Reviewed by Hajime Morita.
3392
3393         Record and output the aggreated test run time.
3394
3395         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
3396         (PerfTestsRunner._run_single_test):
3397
3398 2012-03-01  Michael Nordman  <michaeln@google.com>
3399
3400         [chromium] Use webkit_support to create session storage namespaces in DRT.
3401         https://bugs.webkit.org/show_bug.cgi?id=79933
3402
3403         Reviewed by Tony Chang.
3404
3405         * DumpRenderTree/chromium/WebViewHost.cpp:
3406         (WebViewHost::createSessionStorageNamespace):
3407
3408 2012-03-01  Dirk Pranke  <dpranke@chromium.org>
3409
3410         nrwt: test_isfile, test_isdir, and test_exists don't work for virtual tests
3411         https://bugs.webkit.org/show_bug.cgi?id=80048
3412
3413         Reviewed by Adam Barth.
3414
3415         * Scripts/webkitpy/layout_tests/port/base.py:
3416         (Port.test_isfile):
3417         (Port.test_isdir):
3418         (Port):
3419         (Port.test_exists):
3420         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
3421         (PortTest.test_test_exists):
3422         (PortTest):
3423         (PortTest.test_test_isfile):
3424         (PortTest.test_test_isdir):
3425         * Scripts/webkitpy/layout_tests/port/test.py:
3426         (TestPort.virtual_test_suites):
3427         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3428         (MainTest.test_virtual):
3429
3430 2012-03-01  Kalev Lember  <kalevlember@gmail.com>
3431
3432         [GTK] Avoid building any parts of WebKitTestRunner when WK2 is disabled
3433         https://bugs.webkit.org/show_bug.cgi?id=80031
3434
3435         Reviewed by Gustavo Noronha Silva.
3436
3437         Only build the libTestRunnerInjectedBundle.la convenience library when
3438         ENABLE_WEBKIT2 is defined.
3439
3440         * WebKitTestRunner/GNUmakefile.am:
3441
3442 2012-03-01  Philippe Normand  <pnormand@igalia.com>
3443
3444         Unreviewed, second attempt to fix GTK NRWT after r109242.
3445
3446         * Scripts/webkitpy/layout_tests/port/gtk.py:
3447         (GtkDriver._start):
3448
3449 2012-03-01  Philippe Normand  <pnormand@igalia.com>
3450
3451         Unreviewed, GTK NRWT build fix attempt after r109242.
3452
3453         * Scripts/webkitpy/layout_tests/port/gtk.py:
3454         (GtkDriver._start):
3455
3456 2012-03-01  Kentaro Hara  <haraken@chromium.org>
3457
3458         REGRESSION(r109036): run-bindings-tests fails because resolve-supplemental.pl was renamed
3459         https://bugs.webkit.org/show_bug.cgi?id=80008
3460
3461         Reviewed by Adam Barth.
3462
3463         In r109036, I forgot to rename "resolve-supplemental.pl" in run-bindings-tests
3464         to "preprocess-idls.pl". This patch renames it.
3465
3466         Test: Confirm that run-bindings-tests runs.
3467
3468         * Scripts/webkitpy/bindings/main.py:
3469         (BindingsTests.generate_supplemental_dependency):
3470
3471 2012-02-29  Simon Hausmann  <simon.hausmann@nokia.com>
3472
3473         [Qt][WK2] QQuickWebView::event should lookup faster which events QQuickWebPage can handle
3474         https://bugs.webkit.org/show_bug.cgi?id=78047
3475
3476         Reviewed by Kenneth Rohde Christiansen.
3477
3478         Send events for the webview through the QQuickCanvas instead of directly to the item. Direct
3479         sending to the item requires the QObject::event re-implementation to dispatch to the correct
3480         event handling function. However QQuickItem::event doesn't do that. Instead QQuickCanvas calls
3481         the event handlers directly, and with the removal of our own dispatching we need to go through
3482         that code path.
3483
3484         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
3485         (WTR::PlatformWebView::sendEvent):
3486         (WTR::PlatformWebView::postEvent):
3487
3488 2012-02-29  Adam Barth  <abarth@webkit.org>
3489
3490         Disable watchlist.  It's been spamming CCs.
3491
3492         * Scripts/webkitpy/tool/bot/stylequeuetask.py:
3493         (StyleQueueTask.run):
3494         * Scripts/webkitpy/tool/commands/queues_unittest.py:
3495
3496 2012-02-29  Dirk Pranke  <dpranke@chromium.org>
3497
3498         need to clone the gpu baselines and expectations into the new "virtual" dir
3499         https://bugs.webkit.org/show_bug.cgi?id=79956
3500
3501         Reviewed by Adam Barth.
3502
3503         * Scripts/webkitpy/layout_tests/port/chromium.py:
3504         (ChromiumPort.virtual_test_suites):
3505
3506 2012-02-29  Mark Rowe  <mrowe@apple.com>
3507
3508         Switch to using json rather than simplejson now that the master is using Python 2.7.
3509
3510         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
3511         (loadBuilderConfig):
3512
3513 2012-02-29  David Levin  <levin@chromium.org>
3514
3515         [chromium] Add the ability to turn off autoresize.
3516         https://bugs.webkit.org/show_bug.cgi?id=77452
3517
3518         Reviewed by Darin Fisher.
3519
3520         * DumpRenderTree/chromium/LayoutTestController.cpp:
3521         (LayoutTestController::LayoutTestController): Expose a way to turn off auto-resize.
3522         (LayoutTestController::enableAutoResizeMode): Ditto and adjust to new WebKit api.
3523         (LayoutTestController::disableAutoResizeMode):  Ditto.
3524         * DumpRenderTree/chromium/LayoutTestController.h: Ditto.
3525
3526 2012-02-29  Dirk Pranke  <dpranke@chromium.org>
3527
3528         nrwt: make --print trace work with virtual tests
3529         https://bugs.webkit.org/show_bug.cgi?id=79952
3530
3531         Reviewed by Adam Barth.
3532
3533         * Scripts/webkitpy/layout_tests/views/printing.py:
3534         (Printer._print_test_trace):
3535         (Printer._print_baseline):
3536
3537 2012-02-28  Brian Weinstein  <bweinstein@apple.com>
3538
3539         WebKit2: didNewFirstVisuallyNonEmptyLayout should be sent to injected bundle
3540         https://bugs.webkit.org/show_bug.cgi?id=79849
3541
3542         Reviewed by Beth Dakin.
3543
3544         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
3545         (WTR::InjectedBundlePage::InjectedBundlePage): Add an entry to the struct.
3546
3547 2012-02-29  Lucas Forschler  <lforschler@apple.com>
3548
3549         Removed unused Apple buildslaves from the master configuration
3550         https://bugs.webkit.org/show_bug.cgi?id=79925
3551
3552         Reviewed by Stephanie Lewis.
3553
3554         * BuildSlaveSupport/build.webkit.org-config/config.json:
3555
3556 2012-02-29  Dirk Pranke  <dpranke@chromium.org>
3557
3558         nrwt: implement simple 'virtual test suite' support
3559         https://bugs.webkit.org/show_bug.cgi?id=79737
3560
3561         Reviewed by Adam Barth.
3562
3563         Add very basic support for 'virtual test suites' to NRWT:
3564         a virtual suite is a directory that may contain new baselines,
3565         contains a pointer to a "base" directory of tests, and an
3566         optional list of command line arguments to pass to DRT;
3567         it gives us a way to run all of the tests in a given directory
3568         multiple times with multiple (potentially differing) sets of
3569         results and baselines.
3570
3571         This patch implements the support needed for this feature, and
3572         some basic tests, but no actual port will use the feature yet.
3573
3574         This is probably the simplest implementation possible - the
3575         list of virtual suites will be hard-coded into the port's
3576         implementation. One can imagine a more data-driven approach
3577         where the suite instructions are stored in a manifest file
3578         either in LayoutTests (all suites in one file), or per-port,
3579         or per-directory. If this feature ends up being useful we
3580         should probably add something like that so people don't have
3581         to hunt around in the code to add suites.
3582
3583         * Scripts/webkitpy/layout_tests/port/base.py:
3584         (Port.__init__):
3585         (Port.expected_filename):
3586         (Port.tests):
3587         (Port):
3588         (Port._real_tests):
3589         (Port._driver_class):
3590         (Port.virtual_test_suites):
3591         (Port.virtual_suite):
3592         (Port.virtual_suite.VirtualTestSuite):
3593         (Port.virtual_suite.VirtualTestSuite.__init__):
3594         (Port.virtual_suite.VirtualTestSuite.__repr__):
3595         (Port.populated_virtual_test_suites):
3596         (Port._virtual_tests):
3597         (Port.lookup_virtual_test_base):
3598         (Port.lookup_virtual_test_args):
3599         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
3600         (PortTest.test_find_with_skipped_directories):
3601         * Scripts/webkitpy/layout_tests/port/driver.py:
3602         (DriverInput.__init__):
3603         (DriverProxy.run_test):
3604         * Scripts/webkitpy/layout_tests/port/test.py:
3605         (TestPort.virtual_test_suites):
3606         (TestDriver.run_test):
3607         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3608         (MainTest.test_all):
3609         (MainTest.test_virtual):
3610
3611 2012-02-29  Dirk Pranke  <dpranke@chromium.org>
3612
3613         nrwt: support more than two drivers in DriverProxy
3614         https://bugs.webkit.org/show_bug.cgi?id=79736
3615
3616         Reviewed by Adam Barth.
3617
3618         Now that we can support per-test command lines for
3619         Drivers, modify DriverProxy to keep a map of running
3620         drivers for each needed command-line; this will allow
3621         us to transparently maintain a pool of appropriately
3622         configured DRTs without having to constantly start and stop
3623         them.
3624
3625         Note that this potentially raises a garbage collection
3626         problem - the number of running DRTs will grow with the
3627         number of different sets of command line args. For now
3628         this is no worse than the current code - if you're running
3629         with pixel tests, you will only need one DRT per worker,
3630         and if you aren't, you'll need two (one for text-only tests,
3631         and one for reftests).
3632
3633         An alternative would be to only ever have one running driver,
3634         and restart the driver as the command line changes, but this
3635         might (?) slow down execution in the text-only case - we
3636         should benchmark this because it would be simpler and possibly
3637         allow us to eliminate DriverProxy altogether.
3638
3639         * Scripts/webkitpy/layout_tests/port/driver.py:
3640         (DriverProxy.__init__):
3641         (DriverProxy):
3642         (DriverProxy._make_driver):
3643         (DriverProxy.run_test):
3644         (DriverProxy.has_crashed):
3645         (DriverProxy.stop):
3646         (DriverProxy.cmd_line):
3647         (DriverProxy._cmd_line_as_key):
3648
3649 2012-02-29  Adrienne Walker  <enne@google.com>
3650
3651         Unreviewed, add myself as a reviewer
3652
3653         * Scripts/webkitpy/common/config/committers.py:
3654
3655 2012-02-29  Greg Billock  <gbillock@google.com>
3656
3657         Add more tests for web intents
3658         https://bugs.webkit.org/show_bug.cgi?id=79527
3659
3660         Reviewed by Adam Barth.
3661
3662         * DumpRenderTree/chromium/LayoutTestController.cpp:
3663         (LayoutTestController::LayoutTestController):
3664         (LayoutTestController::sendWebIntentResponse):
3665         * DumpRenderTree/chromium/LayoutTestController.h:
3666         (LayoutTestController):
3667         * DumpRenderTree/chromium/WebViewHost.h:
3668         (WebViewHost):
3669         (WebViewHost::currentIntentRequest):
3670
3671 2012-02-29  Dirk Pranke  <dpranke@chromium.org>
3672
3673         nrwt: port/Driver needs to support per-test command line args
3674         https://bugs.webkit.org/show_bug.cgi?id=79733
3675
3676         Reviewed by Adam Barth.
3677
3678         As part of removing the 'gpu' configurations and adding support
3679         for 'virtual test suites', the Driver classes need to support
3680         per-test command lines (since different tests will need to be
3681         run with different command line options).
3682
3683         The per-test args are not yet used, so this change should have
3684         no visible effects and need no additional testing.
3685
3686         * Scripts/webkitpy/layout_tests/port/base.py:
3687         (Port.driver_cmd_line):
3688         * Scripts/webkitpy/layout_tests/port/chromium.py:
3689         (ChromiumDriver.cmd_line):
3690         (ChromiumDriver._start):
3691         (ChromiumDriver.start):
3692         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
3693         (ChromiumDriverTest.test_two_drivers.MockDriver.cmd_line):
3694         (ChromiumDriverTest):
3695   &nbs