Unreviewed, rolling out r111993.
[WebKit-https.git] / Tools / ChangeLog
1 2012-03-26  Sheriff Bot  <webkit.review.bot@gmail.com>
2
3         Unreviewed, rolling out r111993.
4         http://trac.webkit.org/changeset/111993
5         https://bugs.webkit.org/show_bug.cgi?id=82184
6
7         It broke layout and API tests and made WTR crash (Requested by
8         Ossy on #webkit).
9
10         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
11         (WTR::PlatformWebView::PlatformWebView):
12
13 2012-03-25  Kevin Ollivier  <kevino@theolliviers.com>
14
15         [wx] Unreviewed build fix. Move WTF to its own static lib build.
16
17         * waf/build/settings.py:
18         (common_configure):
19
20 2012-03-24  Chris Fleizach  <cfleizach@apple.com>
21
22         AX: Support solution to handle invalid ax text marker
23         https://bugs.webkit.org/show_bug.cgi?id=82023
24
25         Reviewed by Oliver Hunt.
26
27         * DumpRenderTree/AccessibilityUIElement.cpp:
28         (indexForTextMarkerCallback):
29         (isTextMarkerValidCallback):
30         (textMarkerForIndexCallback):
31         (AccessibilityUIElement::indexForTextMarker):
32         (AccessibilityUIElement::isTextMarkerValid):
33         (AccessibilityUIElement::textMarkerForIndex):
34         (AccessibilityUIElement::getJSClass):
35         * DumpRenderTree/AccessibilityUIElement.h:
36         (AccessibilityUIElement):
37         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
38         (AccessibilityUIElement::indexForTextMarker):
39         (AccessibilityUIElement::textMarkerForIndex):
40         (AccessibilityUIElement::isTextMarkerValid):
41         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.cpp:
42         (WTR::AccessibilityUIElement::indexForTextMarker):
43         (WTR::AccessibilityUIElement::isTextMarkerValid):
44         (WTR::AccessibilityUIElement::textMarkerForIndex):
45         (WTR):
46         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.h:
47         (AccessibilityUIElement):
48         * WebKitTestRunner/InjectedBundle/Bindings/AccessibilityUIElement.idl:
49         * WebKitTestRunner/InjectedBundle/mac/AccessibilityUIElementMac.mm:
50         (WTR):
51         (WTR::AccessibilityUIElement::indexForTextMarker):
52         (WTR::AccessibilityUIElement::isTextMarkerValid):
53         (WTR::AccessibilityUIElement::textMarkerForIndex):
54
55 2012-03-24  Sheriff Bot  <webkit.review.bot@gmail.com>
56
57         Unreviewed, rolling out r112014.
58         http://trac.webkit.org/changeset/112014
59         https://bugs.webkit.org/show_bug.cgi?id=82138
60
61         Broke SnowLeopard, Chromium and GTK+ bots (Requested by kov on
62         #webkit).
63
64         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
65         (ManagerTest.test_http_locking):
66         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
67         (AbstractWorker.run):
68         * Scripts/webkitpy/layout_tests/controllers/worker.py:
69         (Worker.__init__):
70         (Worker.safe_init):
71         (Worker.run):
72         (Worker.cleanup):
73         * Scripts/webkitpy/layout_tests/controllers/worker_unittest.py:
74         (WorkerTest.test_default_platform_in_worker):
75         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
76         (run):
77         (main):
78         * Scripts/webkitpy/layout_tests/views/metered_stream.py:
79         (MeteredStream):
80         (MeteredStream.is):
81         (MeteredStream.__init__):
82         (MeteredStream.write):
83         (MeteredStream.update):
84         (MeteredStream._overwrite):
85         (MeteredStream._reset):
86         * Scripts/webkitpy/layout_tests/views/metered_stream_unittest.py:
87         (TestMeteredStream):
88         (TestMeteredStream.test_regular):
89         * Scripts/webkitpy/layout_tests/views/printing.py:
90         (_configure_logging):
91         (_restore_logging):
92         (Printer.__init__):
93         (Printer.cleanup):
94         (Printer.print_progress):
95         (Printer.print_update):
96         (Printer._write):
97         (Printer):
98         (Printer._update):
99         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
100         (TestUtilityFunctions.assertEmpty):
101         (TestUtilityFunctions):
102         (TestUtilityFunctions.assertNotEmpty):
103         (TestUtilityFunctions.assertWritten):
104         (TestUtilityFunctions.test_configure_logging):
105         (Testprinter.get_printer):
106         (Testprinter.test_print_progress):
107         (Testprinter.test_print_progress.mock_time):
108         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
109         (PerfTestsRunner.__init__):
110         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
111         (test_run_test_pause_before_testing):
112
113 2012-03-24  Kevin Ollivier  <kevino@theolliviers.com>
114
115         [wx] Unreviewed. Fix the build after WTF move.
116
117         * waf/build/build_utils.py:
118         (get_excludes):
119         (get_excludes_in_dirs):
120         * waf/build/settings.py:
121         (get_port_excludes):
122         (common_configure):
123
124 2012-03-24  Dirk Pranke  <dpranke@chromium.org>
125
126         clean up NRWT logging/metering, be less verbose
127         https://bugs.webkit.org/show_bug.cgi?id=73846
128
129         Reviewed by Ojan Vafai.
130
131         This patch totally rewrites the metered_stream module and
132         changes the way it and the printing module now interact with the
133         logging subsystem.
134
135         The printing module becomes a dumb layout_tests-specific wrapper around
136         the metered stream. All of the intelligence of how and when to
137         actually write (or erase) lines and how things interact with
138         logging moves into the metered stream.
139
140         As of now, anything written to the metered stream *bypasses the
141         logging system*. If a logger is passed to the meter constructor,
142         the metered stream will also attach a log handler so that all
143         messages fed to the logger will also be written to the metered
144         stream. This detangles the logic that used to exist where
145         sometimes the meter would write to the logger and sometimes the
146         logger would write to the meter.
147
148         All of this makes the logic generic and reusable for
149         test-webkitpy (coming in another patch). It also helps clarify
150         what the mysterious "configure_logging" parameter to the printer
151         constructor used to do (that parameter itself is gone now, and
152         by default the printer and metered_stream will not ever touch
153         the logging system; this makes a bunch of tests less fragile).
154
155         All of the above should produce no user-visible changes.
156
157         However, to also fix the bug in question, we change the message
158         format in --verbose mode to %H:%M:%S.%frac %pid %message,
159         eliminating the date, file:lineno, and level parameters. I am
160         making this change at the same time because it would be
161         difficult to write things consistently between the logs and the
162         metered stream if we were still logging the file:lineno.
163
164         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
165           Remove the configure_logging parameter
166         * Scripts/webkitpy/layout_tests/controllers/worker.py:
167           Use metered stream to configure the logging (we don't need a printer)
168           and be careful about when and how we manipulate the log level
169           and handlers on the root logger.
170         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
171           Pass in the root logger so that log messages are formatted
172           properly, and set the log level outside of any routines called
173           by unit tests so as to not interfere.
174         * Scripts/webkitpy/layout_tests/views/manager_worker_broker.py:
175           remove a duplicate log message.
176         * Scripts/webkitpy/layout_tests/views/metered_stream.py:
177           complete rewrite :).
178         * Scripts/webkitpy/layout_tests/views/metered_stream_unittest.py:
179           complete rewrite :).
180         * Scripts/webkitpy/layout_tests/views/printing.py:
181           move the logging configuration into metered_stream, and also
182           move the "when should I log progress messages" into metered stream.
183         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
184           remove tests that are no longer necessary.
185         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
186           remove the configure_logging parameter.
187         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
188           update the test; since the printer no longer writes to the log,
189           we need to check what is written to the regular_output instead.
190           (test_run_test_pause_before_testing):
191
192 2012-03-24  Gustavo Noronha Silva  <gns@gnome.org>
193
194         Add myself and Martin Robinson to:
195           * GStreamerGraphics
196           * GtkWebKit2PublicAPI
197           * SoupNetwork
198         and to the newly created WebKitGTKTranslations.
199
200         Rubber-stamped by Martin Robinson.
201
202         * Scripts/webkitpy/common/config/watchlist:
203
204 2012-03-24  Jesus Sanchez-Palencia  <jesus.palencia@openbossa.org>
205
206         [Qt][WK2] Implement PageClient::isViewWindowActive()
207         https://bugs.webkit.org/show_bug.cgi?id=81143
208
209         Reviewed by Kenneth Rohde Christiansen.
210
211         Fix the WrapperWindow from our PlatformWebView
212         so it correctly creates the platform related
213         stuff for QWindow, which is not created unless
214         QWindow::setVisible() or QWindow::show() are called.
215
216         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
217         (WTR::PlatformWebView::PlatformWebView):
218
219 2012-03-23  Dirk Pranke  <dpranke@chromium.org>
220
221         nrwt: don't wait for safari to exit before exiting after showing the results file
222         https://bugs.webkit.org/show_bug.cgi?id=81845
223
224         Reviewed by Ryosuke Niwa.
225
226         Prior to this change, if you ran new-run-webkit-tests and
227         displayed the HTML results file at the end, we would block
228         waiting for the user to quit the browser. There doesn't seem to
229         be a need for that, and the Chromium ports don't do this.
230
231         Also, update the mac tests to capture the output and be quiet :).
232
233         * Scripts/webkitpy/common/system/executive_mock.py:
234           Make popen() testable, implement should_log for it.
235         * Scripts/webkitpy/layout_tests/port/mac.py:
236         (MacPort.show_results_html_file):
237         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
238           Silence some of the tests via outputcapture.
239
240 2012-03-23  Gustavo Noronha Silva  <gns@gnome.org>
241
242         [GTK] libgcrypt and p11-kit should not be in jhbuild modules
243         https://bugs.webkit.org/show_bug.cgi?id=82073
244
245         Reviewed by Martin Robinson.
246
247         * Scripts/webkitdirs.pm:
248         (jhbuildConfigurationChanged): new function that isolates the
249         jhbuild configuration change checking from the autogen logic.
250         (mustReRunAutogen): code for the new function lived here.
251         (buildAutotoolsProject): now calls jhbuildConfigurationChanged
252         itself, and cleans up the jhbuild root if configuration has
253         changed; it also forces autogen to be run in that case, to
254         make sure the new libraries are used.
255         * gtk/jhbuild.modules: removed libgcrypt and p11-kit.
256         * jhbuild/jhbuild-wrapper:
257         (ensure_jhbuild): this might be run inside a jhbuild environment,
258         in which case aclocal fails when trying to use the now deleted
259         aclocal directory in the jhbuild prefix, so work around that.
260
261 2012-03-23  Ojan Vafai  <ojan@chromium.org>
262
263         garden-o-matic scrolls to the top of the page when you click an accordion item
264         https://bugs.webkit.org/show_bug.cgi?id=82088
265
266         Reviewed by Adam Barth.
267
268         Check the scroll position and make sure that the clicked accordion item is visible.
269
270         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
271
272 2012-03-23  Dirk Pranke  <dpranke@chromium.org>
273
274         LayoutTestHelper could get torn down earlier (mostly to reset color profile)
275         https://bugs.webkit.org/show_bug.cgi?id=80567
276
277         Reviewed by Ryosuke Niwa.
278
279         Move clean_up_run to occur right after we finish testing and
280         before we start processing the results. Also makes set_up_run()
281         and clean_up_run() private since there's no real need for
282         callers to be aware of them.
283
284         * Scripts/webkitpy/layout_tests/controllers/manager.py:
285         (Manager._set_up_run):
286         (Manager.run):
287         (Manager._clean_up_run):
288         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
289         (ManagerTest.test_http_locking):
290         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
291         (run):
292
293 2012-03-23  Dean Jackson  <dino@apple.com>
294
295         Disable CSS_SHADERS in Apple builds
296         https://bugs.webkit.org/show_bug.cgi?id=81996
297
298         Reviewed by Simon Fraser.
299
300         Default enableCSSShaders to be 0 on all builds that use
301         this script. Use the flag to turn it back on.
302
303         * Scripts/build-webkit:
304
305 2012-03-23  Tony Chang  <tony@chromium.org>
306
307         [chromium] rename newwtf target back to wtf
308         https://bugs.webkit.org/show_bug.cgi?id=82064
309
310         Reviewed by Adam Barth.
311
312         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
313
314 2012-03-23  Jessie Berlin  <jberlin@apple.com>
315
316         Assertion failure in PlatformPasteboardMac constructor causing multiple tests to "crash" on
317         the Lion Intel WebKit2 Debug bots.
318         https://bugs.webkit.org/show_bug.cgi?id=81012.
319
320         Rubber-stamped by Dan Bernstein.
321
322         Add a step to kill the pasteboard server while we investigate what is causing these
323         assertion failures (to get the bots green).
324
325         * BuildSlaveSupport/kill-old-processes:
326
327 2012-03-22  Martin Robinson  <mrobinson@igalia.com>
328
329         [check-webkit-style] Alphabetical sorting errors in headers are reported for the line after the first out of order header
330         https://bugs.webkit.org/show_bug.cgi?id=81986
331
332         Reviewed by David Levin.
333
334         Track whether or not errors are filtered in error handlers. If an
335         alphabetical sorting error is reported for a header and filtered, try
336         reporting it for the other potential source of the error. This ensures
337         that the style bot can still find new sorting errors in both potential
338         situations:
339
340             #include <foo.h> <-- 1. edited line
341             #include <baz.h> <-- 2. edited line
342
343         * Scripts/webkitpy/style/checkers/cpp.py:
344         (check_include_line): Potentially report the error for both lines,
345         if the first error is filtered.
346         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Implement a new style of
347         test that allows only reporting errors for certain lines. This verifies that
348         when one line of an ordering error is filtered, the error is reported on the
349         other line.
350         (ErrorCollector.__init__):
351         (ErrorCollector.__call__):
352         (CppStyleTestBase.perform_lint):
353         (CppStyleTestBase.perform_language_rules_check):
354         (CppStyleTestBase.assert_language_rules_check):
355         (OrderOfIncludesTest.test_check_alphabetical_include_order_errors_reported_for_both_lines):
356         Added a new test that verifies that when one line of the two lines of an ordering
357         error occur, the error is reported on the other line.
358         * Scripts/webkitpy/style/checkers/jsonchecker_unittest.py:
359         (MockErrorHandler.__call__): Report True because the error is handled.
360         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
361         (ErrorCollector.__call__): Ditto.
362         * Scripts/webkitpy/style/checkers/watchlist_unittest.py:
363         (MockErrorHandler.__call__): Ditto.
364         * Scripts/webkitpy/style/checkers/xcodeproj_unittest.py:
365         (TestErrorHandler.__call__): Ditto.
366         * Scripts/webkitpy/style/checkers/xml_unittest.py:
367         (MockErrorHandler.__call__): Ditto.
368         * Scripts/webkitpy/style/error_handlers.py:
369         (DefaultStyleErrorHandler.__call__): Report True if the error is handled
370         and False if it is filtered.
371
372 2012-03-23  Patrick Gansterer  <paroga@webkit.org>
373
374         Build fix for WinCE after r111778.
375
376         * WinCELauncher/CMakeLists.txt:
377
378 2012-03-23  Peter Beverloo  <peter@chromium.org>
379
380         Unreviewed buildfix for the Chromium Android builder.
381
382         Disable the ImageDiff target for now. Android needs to build this for
383         host rather than for target (as other configurations do), but since the
384         WTF move the dependency chain got significantly more complicated.
385
386         I'll work out a proper fix in the following bug, but since several
387         things are broken due to this, this buildfix should be landed first.
388         https://bugs.webkit.org/show_bug.cgi?id=82039
389
390         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
391
392 2012-03-23  Grzegorz Czajkowski  <g.czajkowski@samsung.com>
393
394         [EFL][DRT] Implement LayoutTestController's methods related with editing.
395         https://bugs.webkit.org/show_bug.cgi?id=80136
396
397         Reviewed by Ryosuke Niwa.
398
399         Implements EFL's LayoutTestController methods e.i. setSmartInsertDeleteEnabled,
400         setSelectTrailingWhitespaceEnabled.
401
402         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
403         (LayoutTestController::setSmartInsertDeleteEnabled):
404         (LayoutTestController::setSelectTrailingWhitespaceEnabled):
405
406 2012-03-22  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
407
408         Convert hasGrammarMarker to use Internals interface
409         https://bugs.webkit.org/show_bug.cgi?id=82004
410
411         Reviewed by Ryosuke Niwa.
412
413         Remove hasGrammarMarker functions, because it is able to work in the
414         cross-port way through the Internals interface.
415
416         * DumpRenderTree/LayoutTestController.cpp:
417         (LayoutTestController::staticFunctions):
418         * DumpRenderTree/LayoutTestController.h:
419         (LayoutTestController):
420         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
421         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
422         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
423         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
424         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
425         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
426
427 2012-03-22  Adam Barth  <abarth@webkit.org>
428
429         sync-master-with-upstream creates non-linear git history
430         https://bugs.webkit.org/show_bug.cgi?id=82006
431
432         Reviewed by Julien Chaffraix.
433
434         Previously this script would create a merge commit at the HEAD of
435         origin/master, causing your fork of WebKit to diverge from upstream.
436         The new version of this script pushes directly from upstream to origin
437         (without round-tripping through any local branches), avoiding this
438         problem.
439
440         * Scripts/sync-master-with-upstream:
441
442 2012-03-22  Gustavo Noronha Silva  <gns@gnome.org>
443
444         Unreviewed build fix. Add knowledge of the WTF project to
445         the autotools build system scripts so that build-jsc works
446         correctly for GTK+.
447
448         * Scripts/webkitdirs.pm:
449         (buildAutotoolsProject): handle the WTF project, build the
450         libWTF.la target.
451         (buildGtkProject): add WTF to the list of supported projects.
452
453 2012-03-22  Eric Seidel  <eric@webkit.org>
454
455         Make svn-apply re-write JavaScriptCore/wtf to WTF/wtf
456         https://bugs.webkit.org/show_bug.cgi?id=81984
457
458         Reviewed by Adam Barth.
459
460         * Scripts/VCSUtils.pm:
461         (adjustPathForRecentRenamings):
462
463 2012-03-22  Raphael Kubo da Costa  <rakuco@FreeBSD.org>
464
465         [CMake] Fix build-jsc after r111778 and r111797.
466         https://bugs.webkit.org/show_bug.cgi?id=82001
467
468         Reviewed by Filip Pizlo.
469
470         Contrary to the comment in build-jsc, buildCMakeProjectOrExit()
471         does return if no error occurred, and build-jsc failed with
472         "building not defined for this platform".
473
474         * Scripts/build-jsc: Always exit after caling
475         buildCMakeProjectOrExit().
476         * Scripts/webkitdirs.pm:
477         (buildCMakeProjectOrExit): Return 0 if no error occurred.
478
479 2012-03-22  Raphael Kubo da Costa  <rakuco@FreeBSD.org>
480
481         [CMake] Unreviewed build fix after r111778.
482
483         * DumpRenderTree/efl/CMakeLists.txt: Replace
484         ${JAVASCRIPTCORE_DIR}/wtf with ${WTF_DIR}/wtf in the include
485         paths.
486         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp: Use the right
487         wtf include.
488         * DumpRenderTree/efl/WorkQueueItemEfl.cpp: Ditto.
489         * WinCELauncher/CMakeLists.txt: Replace ${JAVASCRIPTCORE_DIR}/wtf
490         with ${WTF_DIR}/wtf in the include paths.
491
492 2012-03-22  Filip Pizlo  <fpizlo@apple.com>
493
494         Tools/Scripts/build-jsc doesn't work after WTF move
495         https://bugs.webkit.org/show_bug.cgi?id=81999
496
497         Reviewed by Eric Seidel.
498         
499         For ports that require WTF and JSC to be built as two projects and where
500         the machinery to build both does not already exist, teach the tool to do
501         so for us.
502
503         * Scripts/build-jsc:
504         (buildMyProject):
505
506 2012-03-22  Ojan Vafai  <ojan@chromium.org>
507
508         Fix accidental missing line in r111793.
509
510         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base.js:
511         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base_unittests.js:
512         Remove base.values. It was unused and duplicates the builtin Object.values.
513         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
514         Accidentally deleted this line before committing r111793.
515
516 2012-03-22  Ojan Vafai  <ojan@chromium.org>
517
518         Move garden-o-matic progress feedback from non-modal dialogs to a status console anchored to the bottom of the page
519         https://bugs.webkit.org/show_bug.cgi?id=81983
520
521         Reviewed by Adam Barth.
522
523         The dialogs get in the way if you are doing multiple rebaselines. Also,
524         a number of people didn't realize that you could do multiple rebaselines
525         in parallel. A non-modal dialog is just confusing.
526
527         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base.js:
528         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
529         If we're only rebaselining 1 test, show the test name in the initial message. Otherwise,
530         show the number of tests being rebaselined.
531
532         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
533         Turn MessageBox into StatusArea. StatusArea is a singleton and messages in the StatusArea are grouped
534         by ID. That way, the results of each UI action (e.g. clicking rebaseline) are grouped together
535         into a single area.
536
537         Once addFinalMessage has been called for all IDs, we show the close button. Clicking close
538         also serves to clear all teh content in the StatusArea. 
539
540         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:
541         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/results.css:
542
543 2012-03-22  Tony Chang  <tony@chromium.org>
544
545         Unreviewed, fix chromium build after wtf move.
546
547         Only use newwtf, remove references to wtf.
548
549         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
550
551 2012-03-22  Csaba Osztrogonác  <ossy@webkit.org>
552
553         Actually move WTF files to their new home
554         https://bugs.webkit.org/show_bug.cgi?id=81844
555
556         [Qt] Unreviewed buildfix after r111778.
557
558         * DumpRenderTree/qt/DumpRenderTree.pro:
559
560 2012-03-22  Dan Bernstein  <mitz@apple.com>
561
562         Font fallback in WebKitTestRunner depends on user settings
563         https://bugs.webkit.org/show_bug.cgi?id=81978
564
565         Reviewed by Jessie Berlin.
566
567         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
568         (WTR::InjectedBundle::platformInitialize): Added an AppleLanguages key
569         with a value consisting of the array ("en") to the override defaults
570         dictionary. This is the same value used in DumpRenderTree, and it
571         ensures that font fallback follows a consistent path that is not
572         affected by the language preferences of the user running the test.
573
574 2012-03-22  Andy Estes  <aestes@apple.com>
575
576         check-for-inappropriate-objc-class-names should ignore the class defined by libarclite.
577         https://bugs.webkit.org/show_bug.cgi?id=81971
578
579         Reviewed by Dan Bernstein.
580
581         Xcode will automatically link ObjC binaries against libarclite in some
582         cases, which defines a class called __ARCLite__.
583         check-for-inappropriate-objc-class-names shouldn't complain about this
584         class name.
585
586         * Scripts/check-for-inappropriate-objc-class-names:
587
588 2012-03-22  Dirk Pranke  <dpranke@chromium.org>
589
590         new-run-webkit-tests blames the wrong test for crashing
591         https://bugs.webkit.org/show_bug.cgi?id=81951
592
593         Reviewed by Adam Barth.
594
595         NRWT wasn't extracting the WebProcess pid correct; should've
596         used re.search instead of re.match :(. Add more tests!
597
598         * Scripts/webkitpy/layout_tests/port/mac.py:
599         (MacPort._get_crash_log):
600         * Scripts/webkitpy/layout_tests/port/webkit.py:
601         (WebKitDriver.has_crashed):
602         (WebKitDriver._check_for_driver_crash):
603         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
604         (WebKitDriverTest.test_no_timeout):
605         (WebKitDriverTest):
606         (WebKitDriverTest.test_check_for_driver_crash):
607         (WebKitDriverTest.test_check_for_driver_crash.FakeServerProcess):
608         (WebKitDriverTest.test_check_for_driver_crash.FakeServerProcess.__init__):
609         (WebKitDriverTest.test_check_for_driver_crash.FakeServerProcess.pid):
610         (WebKitDriverTest.test_check_for_driver_crash.FakeServerProcess.name):
611         (WebKitDriverTest.test_check_for_driver_crash.FakeServerProcess.has_crashed):
612         (WebKitDriverTest.test_check_for_driver_crash.assert_crash):
613
614 2012-03-22  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
615
616         [EFL] [DRT] Implement LayoutTestController::markerTextForListItem()
617         https://bugs.webkit.org/show_bug.cgi?id=81512
618
619         Adding missing implementation markerTextForListItem to EFL's
620         LayoutTestController so that we can unskip related tests from
621         the skip list.
622
623         Reviewed by Antonio Gomes.
624
625         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
626         (LayoutTestController::markerTextForListItem): Implemented.
627
628 2012-03-22  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
629
630         [EFL] Enable view mode media feature layout tests
631         https://bugs.webkit.org/show_bug.cgi?id=81205
632
633         Adding missing implementation setViewModeMediaFeature to EFL's 
634         LayoutTestController so that we can unskip view mode 
635         media feature layout tests.
636
637         Reviewed by Antonio Gomes.
638
639         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
640         (LayoutTestController::setViewModeMediaFeature): Implemented.
641
642 2012-03-22  Dirk Pranke  <dpranke@chromium.org>
643
644         nrwt: crash while stopping layout test helper on apple mac lion
645         https://bugs.webkit.org/show_bug.cgi?id=81949
646
647         Reviewed by Eric Seidel.
648
649         stop_helper() didn't work right if the helper process wasn't
650         still running or didn't respond properly to being shut down. Fix
651         this and add some tests. Also add code to MockExecutive() to be
652         able to mock executive.popen() properly.
653
654         * Scripts/webkitpy/common/system/executive_mock.py:
655         (MockProcess.__init__):
656         (MockProcess.wait):
657         (MockExecutive.__init__):
658         (MockExecutive.popen):
659         * Scripts/webkitpy/common/system/systemhost_mock.py:
660         (MockSystemHost.__init__):
661         * Scripts/webkitpy/layout_tests/port/mac.py:
662         (MacPort.start_helper):
663         (MacPort.stop_helper):
664         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
665         (test_get_crash_log):
666         (test_helper_starts):
667         (test_helper_fails_to_start):
668         (test_helper_fails_to_stop):
669         (test_helper_fails_to_stop.bad_waiter):
670
671 2012-03-22  Gustavo Noronha Silva  <gns@gnome.org>
672
673         [GTK] Add make to the jhbuild moduleset
674         https://bugs.webkit.org/show_bug.cgi?id=81769
675
676         Reviewed by Martin Robinson.
677
678         Mostly based on a patch by Carlos Garcia Campos <cgarcia@igalia.com>.
679         Add GNU make with custom patches to fix several issues:
680
681         - make-3.82-arg-list-length.patch: Patch by Ralf Wildenhues to
682           fix the argument list limit of GNU make when running make
683           distcheck.
684         - make-3.82-parallel-build.patch: Patch by Kamil Mierzejewski
685           and Jeremy Devenport to fix parallel build issues present in
686           GNU make 3.82
687
688         * gtk/jhbuild.modules: Added make with patches.
689         * gtk/patches/make-3.82-arg-list-length.patch: Added.
690         * gtk/patches/make-3.82-parallel-build.patch: Added.
691
692 2012-03-22  Kevin Ollivier  <kevino@theolliviers.com>
693
694         [wx] Unreviewed. Adding Source/WTF to the build and updating
695         waf port list.
696
697         * DumpRenderTree/wscript:
698         * waf/build/settings.py:
699
700 2012-03-22  Dirk Pranke  <dpranke@chromium.org>
701
702         Re-land fix for r81603 yet again (!) with latest build fix.
703         https://bugs.webkit.org/show_bug.cgi?id=81603
704
705         Unreviewed, build fix.
706
707         The latest problem was a bad sprintf() in
708         BasePort._get_crash_log(). I've reworked that routine and added
709         unit tests for coverage.
710
711         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
712         (SingleTestRunner._handle_error):
713         * Scripts/webkitpy/layout_tests/controllers/test_result_writer.py:
714         (write_test_result):
715         (TestResultWriter.write_crash_log):
716         * Scripts/webkitpy/layout_tests/port/base.py:
717         (Port.diff_text):
718         (Port._get_crash_log):
719         * Scripts/webkitpy/layout_tests/port/chromium.py:
720         (ChromiumDriver.run_test):
721         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
722         (ChromiumDriverTest.test_crash_log):
723         * Scripts/webkitpy/layout_tests/port/driver.py:
724         (DriverOutput.__init__):
725         * Scripts/webkitpy/layout_tests/port/mac.py:
726         (MacPort.is_lion):
727         (MacPort._get_crash_log):
728         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
729         (test_default_child_processes):
730         (test_get_crash_log):
731         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
732         (MockDRTPortTest.test_get_crash_log):
733         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
734         (PortTestCase.test_get_crash_log):
735         * Scripts/webkitpy/layout_tests/port/test.py:
736         (TestDriver.run_test):
737         * Scripts/webkitpy/layout_tests/port/webkit.py:
738         (WebKitDriver.__init__):
739         (WebKitDriver._start):
740         (WebKitDriver.has_crashed):
741         (WebKitDriver._check_for_driver_crash):
742         (WebKitDriver.run_test):
743
744 2012-03-22  Alexander Færøy  <alexander.faeroy@nokia.com>
745
746         [Qt] Enable support for handling database quota's in the MiniBrowser
747         https://bugs.webkit.org/show_bug.cgi?id=81928
748
749         Reviewed by Simon Hausmann.
750
751         * MiniBrowser/qt/qml/BrowserWindow.qml:
752
753 2012-03-22  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
754
755         [EFL] Implement LayoutTestController::setEditingBehavior
756         https://bugs.webkit.org/show_bug.cgi?id=81124
757
758         Adding missing implementation setEditingBehavior to EFL's
759         LayoutTestController. This implementation allows us to
760         unskip some tests from the skip list.
761
762         Reviewed by Martin Robinson.
763
764         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
765         (LayoutTestController::setEditingBehavior): Implemented.
766
767 2012-03-22  Carlos Garcia Campos  <cgarcia@bb-webkit-rel-64.local.igalia.com>
768
769         [GTK] Use the angle-bracket form to include wtf headers
770         https://bugs.webkit.org/show_bug.cgi?id=81884
771
772         Reviewed by Eric Seidel.
773
774         Use #include <wtf/foo> instead of #include "foo".
775
776         * DumpRenderTree/gtk/AccessibilityCallbacks.cpp:
777         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
778         * DumpRenderTree/gtk/DumpRenderTree.cpp:
779         * DumpRenderTree/gtk/EditingCallbacks.cpp:
780         * DumpRenderTree/gtk/WorkQueueItemGtk.cpp:
781         * WebKitTestRunner/gtk/EventSenderProxyGtk.cpp:
782
783 2012-03-22  Alexander Færøy  <alexander.faeroy@nokia.com>
784
785         Unreviewed, add Michael Brüning and Allan Jensen to committers.py.
786
787         * Scripts/webkitpy/common/config/committers.py:
788
789 2012-03-22  Allan Sandfeld Jensen  <allan.jensen@nokia.com>
790
791         [Qt] Enable FAST_MOBILE_SCROLLING when scrolling is delegated.
792         https://bugs.webkit.org/show_bug.cgi?id=81889
793
794         Reviewed by Kenneth Rohde Christiansen.
795
796         Enable FAST_MOBILE_SCROLLING feature.
797
798         * qmake/mkspecs/features/features.prf:
799
800 2012-03-22  Raphael Kubo da Costa  <rakuco@FreeBSD.org>
801
802         [EFL] DRT: Do not pass a relative path to DOWNLOADED_FONTS_DIR.
803         https://bugs.webkit.org/show_bug.cgi?id=81900
804
805         Reviewed by Gustavo Noronha Silva.
806
807         r111598 added the DOWNLOADED_FONTS_DIR define, but it used a relative
808         path which assumed a certain directory structure for the build
809         directory which may noy be true.
810
811         Instead, hardcode the path to the downloaded fonts starting from the
812         top-level source directory. While this solution is also far from
813         perfect, it makes it more consistent and reliable, and easier to
814         tackle at once with the rest of bug 81475.
815
816         * DumpRenderTree/efl/CMakeLists.txt:
817
818 2012-03-22  Carlos Garcia Campos  <cgarcia@igalia.com>
819
820         [GTK] Allow to run any jhbuild command with jhbuild-wrapper script
821         https://bugs.webkit.org/show_bug.cgi?id=81888
822
823         Reviewed by Philippe Normand.
824
825         * Scripts/update-webkit-libs-jhbuild:
826         (runJhbuild): Helper function to run jhbuild with a given command.
827         Instead of running jhbuild directly, use the jhbuild-wrapper, so
828         that we don't need to run it first to check whether jhbuild is
829         installed or not.
830         * efl/run-with-jhbuild: Call jhbuild-wrapper with run command.
831         * gtk/run-with-jhbuild: Ditto.
832         * jhbuild/jhbuild-wrapper: Allow to execute any jhbuild command,
833         not only run.
834
835 2012-03-22  Philippe Normand  <pnormand@igalia.com>
836
837         Unreviewed, test-webkitpy fix after r111661.
838
839         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
840         (MainTest.test_crash_log):
841         (MainTest.test_web_process_crash_log):
842
843 2012-03-22  Philippe Normand  <pnormand@igalia.com>
844
845         Unreviewed, test-webkitpy fix after r111642.
846
847         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
848         (ChromiumDriverTest.setUp):
849         (ChromiumDriverTest.test_two_drivers):
850
851 2012-03-22  Csaba Osztrogonác  <ossy@webkit.org>
852
853         webkitpy: clean up actually getting crash logs from DRT/WTR crashes
854         https://bugs.webkit.org/show_bug.cgi?id=81603
855
856         Unreviewed rolling out r111609 and part of r111615,
857         because it broke NRWT on Qt-WK2 platform.
858
859         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
860         (SingleTestRunner._handle_error):
861         * Scripts/webkitpy/layout_tests/controllers/test_result_writer.py:
862         (write_test_result):
863         (TestResultWriter.write_crash_report):
864         * Scripts/webkitpy/layout_tests/port/base.py:
865         (Port.is_crash_reporter):
866         (Port._driver_class):
867         * Scripts/webkitpy/layout_tests/port/chromium.py:
868         (ChromiumDriver.run_test):
869         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
870         (ChromiumDriverTest.test_crashed_process_name):
871         * Scripts/webkitpy/layout_tests/port/driver.py:
872         (DriverOutput.__init__):
873         * Scripts/webkitpy/layout_tests/port/mac.py:
874         (MacPort):
875         (MacPort.is_crash_reporter):
876         (MacPort.release_http_lock):
877         * Scripts/webkitpy/layout_tests/port/test.py:
878         (TestDriver.run_test):
879         * Scripts/webkitpy/layout_tests/port/webkit.py:
880         (WebKitDriver.__init__):
881         (WebKitDriver._start):
882         (WebKitDriver.has_crashed):
883         (WebKitDriver._check_for_driver_crash):
884         (WebKitDriver.run_test):
885
886 2012-03-22  Csaba Osztrogonác  <ossy@webkit.org>
887
888         [Qt] Add full platforms to Qt buildslaves
889         https://bugs.webkit.org/show_bug.cgi?id=81877
890
891         Reviewed by Ryosuke Niwa.
892
893         * BuildSlaveSupport/build.webkit.org-config/config.json:
894
895 2012-03-21  Johnny Ding  <jnd@chromium.org>
896
897         Customize layout test timeout value for different ports.
898         https://bugs.webkit.org/show_bug.cgi?id=79859
899
900         Reviewed by Ojan Vafai.
901
902         * Scripts/webkitpy/layout_tests/controllers/manager.py:
903         (Manager):
904         * Scripts/webkitpy/layout_tests/port/base.py:
905         (Port.default_test_timeout_ms):
906         * Scripts/webkitpy/layout_tests/port/chromium.py:
907         (ChromiumDriver.stop):
908         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
909         (ChromiumAndroidPort.default_test_timeout_ms):
910         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
911         (_set_up_derived_options):
912
913 2012-03-21  Mark Rowe  <mrowe@apple.com>
914
915         Fix the build.
916
917         Ensure that our test tools don't try and use RTTI unnecessarily.
918
919         * DumpRenderTree/mac/Configurations/Base.xcconfig:
920         * TestWebKitAPI/Configurations/Base.xcconfig:
921         * WebKitTestRunner/Configurations/Base.xcconfig:
922
923 2012-03-21  Adam Barth  <abarth@webkit.org>
924
925         WebKit should have some scripts to smooth interactions with GitHub
926         https://bugs.webkit.org/show_bug.cgi?id=81842
927
928         Reviewed by Eric Seidel.
929
930         These scripts support the workflow in
931         https://trac.webkit.org/wiki/UsingGitHub by automating the
932         configuration and syncing operations needed for a GitHub fork of WebKit
933         to track the main WebKit repository.
934
935         * Scripts/configure-github-as-upstream: Added.
936         * Scripts/sync-master-with-upstream: Added.
937
938 2012-03-21  Dirk Pranke  <dpranke@chromium.org>
939
940         WTR - log the pid of a crashing WebProcess
941         https://bugs.webkit.org/show_bug.cgi?id=81575
942
943         Reviewed by Alexey Proskuryakov.
944
945         This allows run-webkit-tests to find the right crash log for a
946         given crashed WebProcess in WebKit2.
947
948         * WebKitTestRunner/TestController.cpp:
949         (WTR::exceededDatabaseQuota):
950         (WTR::TestController::createOtherPage):
951         (WTR::TestController::initialize):
952         (WTR::TestController::resetStateToConsistentValues):
953         (WTR::TestController::runTest):
954         (WTR::TestController::processDidCrash):
955
956 2012-03-21  Enrica Casucci  <enrica@apple.com>
957
958         WebKitURLWithTitles pasteboard format should support URLs containing Emoji characters.
959         https://bugs.webkit.org/show_bug.cgi?id=81835
960         <rdar://problem/11082749>
961
962         Reviewed by Brady Eidson.
963
964         * TestWebKitAPI/Tests/mac/ContextMenuCanCopyURL.html:
965         * TestWebKitAPI/Tests/mac/ContextMenuCanCopyURL.mm:
966         (TestWebKitAPI::contextMenuCopyLink):
967
968 2012-03-21  Alexandru Chiculita  <achicu@adobe.com>
969
970         [CSS Shaders] Make CSS Shaders compile on Chromium
971         https://bugs.webkit.org/show_bug.cgi?id=81435
972
973         Reviewed by Stephen White.
974         
975         I've added WebKitCSSCustomFilterEnabled support for the layoutTestController.overridePreference function.
976
977         * DumpRenderTree/chromium/LayoutTestController.cpp:
978         (LayoutTestController::overridePreference):
979         * DumpRenderTree/chromium/WebPreferences.cpp:
980         (WebPreferences::reset):
981         (WebPreferences::applyTo):
982         * DumpRenderTree/chromium/WebPreferences.h:
983         (WebPreferences):
984
985 2012-03-21  Dirk Pranke  <dpranke@chromium.org>
986
987         fix regression on gtk port introduced by r111609
988         https://bugs.webkit.org/show_bug.cgi?id=81839
989
990         Reviewed by Philippe Normand.
991
992         r111609 should've initialized a couple of missing fields in
993         WebKitDriver, and also should've checked for method overrides
994         in the other ports :(.
995
996         * Scripts/webkitpy/layout_tests/port/gtk.py:
997         (GtkDriver._start):
998         * Scripts/webkitpy/layout_tests/port/webkit.py:
999         (WebKitDriver.__init__):
1000         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1001         (MainTest.test_crash_log):
1002         (MainTest.test_web_process_crash_log):
1003
1004 2012-03-21  Dirk Pranke  <dpranke@chromium.org>
1005
1006         webkitpy: clean up actually getting crash logs from DRT/WTR crashes
1007         https://bugs.webkit.org/show_bug.cgi?id=81603
1008
1009         Reviewed by Adam Barth.
1010
1011         Add a new _get_crash_log() overridable method on the Port
1012         object to customize how we fetch crash logs per port.
1013         Mac will now slow-spin with a timeout waiting for the crash log, rather
1014         than waiting for the ReportCrash process to exit (it appears
1015         that the mac will manage multiple crashes with a single
1016         ReportCrash process, the process waits around longer than
1017         necessary, presumably to avoid thrashing if processes are
1018         repeatedly crashing).
1019
1020         Also, add the DriverOutput should contain the crash log and other info,
1021         which is created in a port-specific manner but can then be
1022         treated generically. Previously single_test_runner would get
1023         told that something crashed and attempt to do something to get
1024         the crash log, but it didn't have the information it needed to
1025         od the right thing; better to make the driver hand back the
1026         right info.
1027
1028         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
1029         (SingleTestRunner._handle_error):
1030         * Scripts/webkitpy/layout_tests/controllers/test_result_writer.py:
1031         (write_test_result):
1032         (TestResultWriter.write_crash_log):
1033         * Scripts/webkitpy/layout_tests/port/base.py:
1034         (Port.diff_text):
1035         (Port._get_crash_log):
1036         * Scripts/webkitpy/layout_tests/port/chromium.py:
1037         (ChromiumDriver.run_test):
1038         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
1039         (ChromiumDriverTest.test_crash_log):
1040         * Scripts/webkitpy/layout_tests/port/driver.py:
1041         (DriverOutput.__init__):
1042         * Scripts/webkitpy/layout_tests/port/mac.py:
1043         (MacPort.is_lion):
1044         (MacPort._get_crash_log):
1045         * Scripts/webkitpy/layout_tests/port/test.py:
1046         (TestDriver.run_test):
1047         * Scripts/webkitpy/layout_tests/port/webkit.py:
1048         (WebKitDriver.__init__):
1049         (WebKitDriver._start):
1050         (WebKitDriver.has_crashed):
1051         (WebKitDriver._check_for_driver_crash):
1052         (WebKitDriver.run_test):
1053
1054 2012-03-21  Dominik Röttsches  <dominik.rottsches@linux.intel.com>
1055
1056         [EFL] Use jhbuild downloaded fonts instead of hardcoded system font paths
1057         https://bugs.webkit.org/show_bug.cgi?id=81507
1058
1059         Now that we're only loading from one directory, no need to
1060         iterate over a list of directories anymore.
1061
1062         Reviewed by Martin Robinson.
1063
1064         * DumpRenderTree/efl/CMakeLists.txt:
1065         * DumpRenderTree/efl/FontManagement.cpp:
1066         (addFontDirectory):
1067         (addFontsToEnvironment):
1068
1069 2012-03-21  Dirk Pranke  <dpranke@chromium.org>
1070
1071         webkitpy: get ServerProcess out of the reportcrash business
1072         https://bugs.webkit.org/show_bug.cgi?id=81600
1073
1074         Unreviewed, build fix.
1075
1076         Re-land r111307 and r111293 with another fix for a crash in NRWT;
1077         we need to check if the driver has crashed before attempting to
1078         read from it.
1079
1080
1081         * Scripts/webkitpy/layout_tests/port/server_process.py:
1082         (ServerProcess._reset):
1083         (ServerProcess._handle_possible_interrupt):
1084         (ServerProcess.write):
1085         (ServerProcess.read_stdout):
1086         (ServerProcess.has_crashed):
1087         (ServerProcess._read):
1088         (ServerProcess.stop):
1089         * Scripts/webkitpy/layout_tests/port/server_process_unittest.py:
1090         (TrivialMockPort.check_for_leaks):
1091         (TestServerProcess.test_broken_pipe):
1092         * Scripts/webkitpy/layout_tests/port/webkit.py:
1093         (WebKitPort._read_image_diff):
1094         (WebKitDriver.has_crashed):
1095         (WebKitDriver._check_for_driver_crash):
1096         (WebKitDriver.run_test):
1097         (WebKitDriver._read_block):
1098         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
1099         (MockServerProcess.__init__):
1100         (MockServerProcess):
1101         (MockServerProcess.has_crashed):
1102
1103 2012-03-21  Zeno Albisser  <zeno@webkit.org>
1104
1105         [Qt][Mac] ranlib segfaults when creating symbol tables for libWebCore.a.
1106         https://bugs.webkit.org/show_bug.cgi?id=81750
1107
1108         Building WebCore on mac currently creates an archive that is bigger than
1109         4GB. But ranlib fails to create a symbol table for such a big archive,
1110         even on 64bit machines.
1111         Therefore we start using SVGAllInOne.cpp when building debug
1112         on mac. This reduces the size of the archive by about 300MB.
1113
1114         Reviewed by Tor Arne Vestbø.
1115
1116         * qmake/mkspecs/features/mac/default_pre.prf:
1117
1118 2012-03-21  Szilard Ledan  <szledan@inf.u-szeged.hu>
1119
1120         Updated the address list of committers.py.
1121
1122         Reviewed by Csaba Osztrogonác.
1123
1124         * Scripts/webkitpy/common/config/committers.py:
1125
1126 2012-03-21  W. James MacLean  <wjmaclean@chromium.org>
1127
1128         Updating my nick in committers file.
1129
1130         Unreviewed.
1131
1132         * Scripts/webkitpy/common/config/committers.py:
1133
1134 2012-03-21  Peter Beverloo  <peter@chromium.org>
1135
1136         [Chromium] Remove Android build-fix when the proper fix rolled into WebKit
1137         https://bugs.webkit.org/show_bug.cgi?id=80861
1138
1139         Reviewed by Tony Chang.
1140
1141         Remove most Android-specific logic from webkitdirs.pm in favor of
1142         handling this during project generation time.
1143
1144         * Scripts/webkitdirs.pm:
1145         (buildChromiumMakefile):
1146
1147 2012-03-21  Carlos Garcia Campos  <cgarcia@igalia.com>
1148
1149         [GTK] Allow running javascript from location bar in MiniBrowser
1150         https://bugs.webkit.org/show_bug.cgi?id=81331
1151
1152         Reviewed by Martin Robinson.
1153
1154         * MiniBrowser/gtk/BrowserWindow.c:
1155         (activateUriEntryCallback): Use browser_window_load_uri().
1156         (browser_window_load_uri): Check whether uri starts with
1157         javascript: and run the given script, or just call
1158         webkit_web_view_load_uri() to load the given uri.
1159         * MiniBrowser/gtk/BrowserWindow.h:
1160         * MiniBrowser/gtk/main.c:
1161         (createBrowserWindow): Use browser_window_load_uri().
1162
1163 2012-03-21  Carlos Garcia Campos  <cgarcia@igalia.com>
1164
1165         [GTK] Add webkit_web_view_run_javascript() to WebKit2 GTK+
1166         https://bugs.webkit.org/show_bug.cgi?id=75543
1167
1168         Reviewed by Martin Robinson.
1169
1170         * MiniBrowser/gtk/GNUmakefile.am: Add javascriptcore_cppflags to
1171         MiniBrowser CPP flags.
1172
1173 2012-03-20  Eric Seidel  <eric@webkit.org>
1174
1175         Move wtf/Platform.h from JavaScriptCore to Source/WTF/wtf
1176         https://bugs.webkit.org/show_bug.cgi?id=80911
1177
1178         Reviewed by Adam Barth.
1179
1180         Update include paths for Source/WTF.
1181
1182         * DumpRenderTree/efl/CMakeLists.txt:
1183         * DumpRenderTree/qt/DumpRenderTree.pro:
1184         * QtTestBrowser/QtTestBrowser.pro:
1185         * Scripts/build-webkit: Build Source/WTF on AppleWin as well.
1186
1187 2012-03-20  Adam Barth  <abarth@webkit.org>
1188
1189         svn-apply should move inflight patches from WebCore/webaudio to WebCore/Modules/webaudio
1190         https://bugs.webkit.org/show_bug.cgi?id=81567
1191
1192         Reviewed by Eric Seidel.
1193
1194         This temporary measure should help us avoid disrupting the folks
1195         working on WebAudio when we move the webaudio directory to Modules.
1196
1197         * Scripts/VCSUtils.pm:
1198         (adjustPathForRecentRenamings):
1199         (parseGitDiffHeader):
1200         (parseSvnDiffHeader):
1201
1202 2012-03-20  Dirk Pranke  <dpranke@chromium.org>
1203
1204         Unreviewed, rolling out r111453.
1205         http://trac.webkit.org/changeset/111453
1206         https://bugs.webkit.org/show_bug.cgi?id=81600
1207
1208         still broken
1209
1210         * Scripts/webkitpy/layout_tests/port/server_process.py:
1211         (ServerProcess._reset):
1212         (ServerProcess.handle_interrupt):
1213         (ServerProcess.write):
1214         (ServerProcess._check_for_crash):
1215         (ServerProcess._handle_timeout):
1216         (ServerProcess._check_for_abort):
1217         (ServerProcess._read):
1218         (ServerProcess.stop):
1219         (ServerProcess):
1220         (ServerProcess.set_crashed):
1221         * Scripts/webkitpy/layout_tests/port/server_process_unittest.py:
1222         (TrivialMockPort.is_crash_reporter):
1223         (TestServerProcess.test_broken_pipe):
1224         * Scripts/webkitpy/layout_tests/port/webkit.py:
1225         (WebKitPort._read_image_diff):
1226         (WebKitDriver.has_crashed):
1227         (WebKitDriver._check_for_driver_crash):
1228         (WebKitDriver):
1229         (WebKitDriver._detected_crash):
1230         (WebKitDriver._subprocess_crashed):
1231         (WebKitDriver._crashed_process_name):
1232         (WebKitDriver.run_test):
1233         (WebKitDriver._read_block):
1234         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
1235         (MockServerProcess.__init__):
1236
1237 2012-03-20  Dirk Pranke  <dpranke@chromium.org>
1238
1239         webkitpy: get ServerProcess out of the reportcrash business
1240         https://bugs.webkit.org/show_bug.cgi?id=81600
1241
1242         Unreviewed, build fix.
1243
1244         Re-land r111307 and r111293 with fix for crash in NRWT; I
1245         was failing to set the crashing process name.
1246
1247         * Scripts/webkitpy/layout_tests/port/server_process.py:
1248         (ServerProcess._reset):
1249         (ServerProcess._handle_possible_interrupt):
1250         (ServerProcess.write):
1251         (ServerProcess.read_stdout):
1252         (ServerProcess.has_crashed):
1253         (ServerProcess._read):
1254         (ServerProcess.stop):
1255         * Scripts/webkitpy/layout_tests/port/server_process_unittest.py:
1256         (TrivialMockPort.check_for_leaks):
1257         (TestServerProcess.test_broken_pipe):
1258         * Scripts/webkitpy/layout_tests/port/webkit.py:
1259         (WebKitPort._read_image_diff):
1260         (WebKitDriver.has_crashed):
1261         (WebKitDriver._check_for_driver_crash):
1262         (WebKitDriver.run_test):
1263         (WebKitDriver._read_block):
1264         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
1265         (MockServerProcess.__init__):
1266         (MockServerProcess):
1267         (MockServerProcess.has_crashed):
1268
1269 2012-03-20  Tim Horton  <timothy_horton@apple.com>
1270
1271         [mac] Restore color space switching code to run-webkit-tests
1272         https://bugs.webkit.org/show_bug.cgi?id=80571
1273         <rdar://problem/11008529>
1274
1275         Reviewed by Simon Fraser.
1276
1277         http://trac.webkit.org/changeset/111429 broke tools build on Snow Leopard.
1278         
1279         Use the old CM* API on Snow Leopard, since CGDisplayCreateUUIDFromDisplayID
1280         didn't exist until Lion.
1281
1282         * DumpRenderTree/mac/LayoutTestHelper.m:
1283         (installLayoutTestColorProfile):
1284         (restoreUserColorProfile):
1285
1286 2012-03-20  Adele Peterson  <adele@apple.com>
1287
1288         Update the last test to use EXPECT_WK_STREQ.
1289
1290         Reviewed by Dan Bernstein.
1291
1292         * TestWebKitAPI/Tests/mac/AttributedString.mm: (TestWebKitAPI::TEST):
1293
1294 2012-03-20  Adele Peterson  <adele@apple.com>
1295
1296         "Attempt to insert nil value " exception when calling attributed string APIs on content with a custom font
1297         https://bugs.webkit.org/show_bug.cgi?id=81630
1298         <rdar://problem/10650660>
1299
1300         Reviewed by Alexey Proskuryakov.
1301
1302         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1303         * TestWebKitAPI/Tests/mac/Ahem.ttf: Added.
1304         * TestWebKitAPI/Tests/mac/AttributedString.mm: Added.
1305         (-[AttributedStringTest webView:didFinishLoadForFrame:]):
1306         (TestWebKitAPI):
1307         (TestWebKitAPI::didFinishLoadForFrame):
1308         (TestWebKitAPI::TEST):
1309         * TestWebKitAPI/Tests/mac/attributedStringCustomFont.html: Added.
1310
1311 2012-03-20  Eric Seidel  <eric@webkit.org>
1312
1313         Add link on queues.webkit.org to kov's fancy queue-depth graph
1314         https://bugs.webkit.org/show_bug.cgi?id=81690
1315
1316         Reviewed by Adam Barth.
1317
1318         * QueueStatusServer/templates/recentstatus.html:
1319
1320 2012-03-20  Tim Horton  <timothy_horton@apple.com>
1321
1322         [mac] Restore color space switching code to run-webkit-tests
1323         https://bugs.webkit.org/show_bug.cgi?id=80571
1324         <rdar://problem/11008529>
1325
1326         Reviewed by Simon Fraser.
1327
1328         Make use of the layout test helper hooks to restore the Mac system-wide
1329         color profile switching code. This makes WkTR and DRT results match, and
1330         causes the main display's color profile to not affect results on Lion.
1331
1332         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
1333         * DumpRenderTree/mac/LayoutTestHelper.m: Added.
1334         (installLayoutTestColorProfile):
1335         (restoreUserColorProfile):
1336         (simpleSignalHandler):
1337         (main):
1338         * Scripts/webkitpy/layout_tests/port/mac.py:
1339         (MacPort.release_http_lock):
1340         (MacPort):
1341         (MacPort._path_to_helper):
1342         (MacPort.start_helper):
1343         (MacPort.stop_helper):
1344         * WebKitTestRunner/mac/PlatformWebViewMac.mm:
1345         (WTR::PlatformWebView::PlatformWebView):
1346         (WTR::PlatformWebView::windowSnapshotImage):
1347
1348 2012-02-24  Darin Fisher  <darin@chromium.org>
1349
1350         Add more Chromium WebKit API reviewers
1351         https://bugs.webkit.org/show_bug.cgi?id=81576
1352
1353         Reviewed by Adam Barth and David Levin.
1354
1355         Simplify watchlist to lump all WebKit API into one.
1356
1357         * Scripts/webkitpy/common/config/watchlist:
1358
1359 2012-03-20  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
1360
1361         Convert hasSpellingMarker to use Internals interface.
1362         https://bugs.webkit.org/show_bug.cgi?id=81300
1363
1364         Reviewed by Ryosuke Niwa.
1365
1366         Remove hasSpellingMarker functions, because it is able to work in the
1367         cross-port way through the Internals interface.
1368
1369         * DumpRenderTree/LayoutTestController.cpp:
1370         (LayoutTestController::staticFunctions):
1371         * DumpRenderTree/LayoutTestController.h:
1372         (LayoutTestController):
1373         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
1374         * DumpRenderTree/chromium/LayoutTestController.cpp:
1375         (LayoutTestController::LayoutTestController):
1376         * DumpRenderTree/chromium/LayoutTestController.h:
1377         (LayoutTestController):
1378         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1379         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
1380         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
1381         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
1382         * DumpRenderTree/qt/LayoutTestControllerQt.h:
1383         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
1384         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
1385
1386 2012-03-20  Csaba Osztrogonác  <ossy@webkit.org>
1387
1388         Add new Qt-WK2 performance tester bot to build.webkit.org
1389         https://bugs.webkit.org/show_bug.cgi?id=80842
1390
1391         Reviewed by Ryosuke Niwa.
1392
1393         * BuildSlaveSupport/build.webkit.org-config/config.json:
1394         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
1395         (RunAndUploadPerfTestsWebKit2):
1396         (RunAndUploadPerfTestsWebKit2.start):
1397         (BuildAndPerfTestWebKit2Factory):
1398         (BuildAndPerfTestWebKit2Factory.__init__):
1399         (DownloadAndPerfTestWebKit2Factory):
1400         (DownloadAndPerfTestWebKit2Factory.__init__):
1401
1402 2012-03-20  Peter Beverloo  <peter@chromium.org>
1403
1404         [Chromium] Update to Android NDK r7b
1405         https://bugs.webkit.org/show_bug.cgi?id=81005
1406
1407         Reviewed by Adam Barth.
1408
1409         Update to version r7b of the Android NDK, release notes of which are
1410         available on the following page:
1411         http://developer.android.com/sdk/ndk/index.html
1412
1413         * Scripts/update-webkit-chromium:
1414
1415 2012-03-20  Csaba Osztrogonác  <ossy@webkit.org>
1416
1417         x86 GCC bug triggers a compilation error in generated copy constructor of CollapsedBorderValue
1418         https://bugs.webkit.org/show_bug.cgi?id=81502
1419
1420         Unreviewed.
1421
1422         * qmake/mkspecs/features/unix/default_post.prf: Reenable Werror after r111171.
1423
1424 2012-03-19  Jessie Berlin  <jberlin@apple.com>
1425
1426         Unreviewed; roll out http://trac.webkit.org/changeset/111307 and http://trac.webkit.org/changeset/111293
1427         because they broke running the layout tests on mac.
1428
1429         * Scripts/webkitpy/layout_tests/port/webkit.py:
1430         (WebKitDriver._check_for_driver_crash):
1431
1432 2012-03-16  Martin Robinson  <mrobinson@igalia.com>
1433
1434         [GTK] Allow running run-gtk-tests during 'make distcheck'
1435         https://bugs.webkit.org/show_bug.cgi?id=81415
1436
1437         Reviewed by Philippe Normand.
1438
1439         Modify the build and run-gtk-tests so that it can run during 'make distcheck'
1440
1441         * GNUmakefile.am: Distribute run-gtk-tests during 'make dist' and make it the
1442         main command run during 'make check.'
1443         * Scripts/run-gtk-tests: Refactored run-gtk-tests so that jhbuild is no longer
1444         a requirement to run it. If jhbuild isn't around it will simply not start the
1445         SPI daemons and skip the WebKit2 accessibility test. Remove test runs from the
1446         main loop and simply use the main loop in a synchronous method to start SPI
1447         daemons. Did a bunch of general cleanup around these tasks.
1448
1449 2012-03-19  Raphael Kubo da Costa  <rakuco@FreeBSD.org>
1450
1451         jhbuild: Use the multiprocessing module to get the number of CPUs.
1452         https://bugs.webkit.org/show_bug.cgi?id=81585
1453
1454         Reviewed by Martin Robinson.
1455
1456         Now that support for Python 2.5 has been dropped, we can safely
1457         use multiprocessing.cpu_count() to obtain the current number of
1458         CPUs in the system instead of running a webkitperl script for
1459         that.
1460
1461         * efl/common.py:
1462         (top_level_path):
1463         * efl/jhbuildrc:
1464         * gtk/common.py:
1465         (build_path):
1466         * gtk/jhbuildrc:
1467
1468 2012-03-19  Ojan Vafai  <ojan@chromium.org>
1469
1470         We should only add a live event listener once, not once per TestSelector.
1471         Adding one per TestSelector would cause accordion('option', 'active') to
1472         incorrectly be "false" if you examine multiple sets of tests.
1473
1474         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
1475
1476 2012-03-19  Dirk Pranke  <dpranke@chromium.org>
1477
1478         Fix typo introduced in r111293.
1479
1480         Unreviewed, build fix.
1481
1482         * Scripts/webkitpy/layout_tests/port/webkit.py:
1483         (WebKitDriver._check_for_driver_crash):
1484
1485 2012-03-19  Dirk Pranke  <dpranke@chromium.org>
1486
1487         webkitpy: get ServerProcess out of the reportcrash business
1488         https://bugs.webkit.org/show_bug.cgi?id=81600
1489
1490         Reviewed by Adam Barth.
1491
1492         ServerProcess was half-aware that ReportCrash might run
1493         sometimes, and that the process ServerProcess was talking to
1494         might have its own crashing subprocesses; neither of these
1495         things really worked right and it made the logic convoluted, so
1496         this change makes handling crashes completely separate from the
1497         server_process code, so that it can focus on just I/O to the
1498         subprocess.
1499
1500         There should be no functional changes resulting from this patch.
1501
1502         * Scripts/webkitpy/layout_tests/port/server_process.py:
1503         (ServerProcess._reset):
1504         (ServerProcess._handle_possible_interrupt):
1505         (ServerProcess.write):
1506         (ServerProcess.read_stdout):
1507         (ServerProcess.has_crashed):
1508         (ServerProcess._read):
1509         (ServerProcess.stop):
1510         * Scripts/webkitpy/layout_tests/port/server_process_unittest.py:
1511         (TrivialMockPort.check_for_leaks):
1512         (TestServerProcess.test_broken_pipe):
1513         * Scripts/webkitpy/layout_tests/port/webkit.py:
1514         (WebKitPort._read_image_diff):
1515         (WebKitDriver.has_crashed):
1516         (WebKitDriver._check_for_driver_crash):
1517         (WebKitDriver.run_test):
1518         (WebKitDriver._read_block):
1519         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
1520         (MockServerProcess.__init__):
1521         (MockServerProcess):
1522         (MockServerProcess.has_crashed):
1523
1524 2012-03-19  Dirk Pranke  <dpranke@chromium.org>
1525
1526         webkitpy: crashlog parsing is broken
1527         https://bugs.webkit.org/show_bug.cgi?id=81586
1528
1529         Reviewed by Adam Barth.
1530
1531         Seeking on a file opened through codecs() appears to not work
1532         correctly (at least on SL).
1533         
1534         The code was more complicated than it needed to be, so I have
1535         revamped it to just read a full crashlog at once and then look
1536         for matches. 
1537         
1538         I have also added the ability to optionally propagate
1539         errors back (which can be helpful to debug races when ReportCrash is
1540         still running and you need to see that you couldn't open some
1541         files, rather than just ignoring them).
1542
1543         * Scripts/webkitpy/common/system/crashlogs.py:
1544         (CrashLogs.find_newest_log):
1545         (CrashLogs._find_newest_log_darwin):
1546         * Scripts/webkitpy/common/system/crashlogs_unittest.py:
1547         (CrashLogsTest.test_find_log_darwin):
1548
1549 2012-03-19  Dirk Pranke  <dpranke@chromium.org>
1550
1551         webkitpy: clean up some port._filesystem references to not reference private members
1552         https://bugs.webkit.org/show_bug.cgi?id=81595
1553
1554         Reviewed by Eric Seidel.
1555
1556         There should be no functional changes in this patch, just some
1557         refactoring. Also, fixes a minor nit in write_image_diff_files(()
1558
1559         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
1560         (SingleTestRunner.__init__):
1561         (SingleTestRunner._run_compare_test):
1562         (SingleTestRunner._run_rebaseline):
1563         (SingleTestRunner._save_baseline_data):
1564         (SingleTestRunner._handle_error):
1565         (SingleTestRunner._run_reftest):
1566         * Scripts/webkitpy/layout_tests/controllers/test_result_writer.py:
1567         (write_test_result):
1568         (TestResultWriter.__init__):
1569         (TestResultWriter._make_output_directory):
1570         (TestResultWriter.output_filename):
1571         (TestResultWriter._output_testname):
1572         (TestResultWriter.write_output_files):
1573         (TestResultWriter.write_stderr):
1574         (TestResultWriter.write_crash_report):
1575         (TestResultWriter.create_text_diff_and_write_result):
1576         (TestResultWriter.write_image_diff_files):
1577         (copy_file):
1578         * Scripts/webkitpy/layout_tests/controllers/test_result_writer_unittest.py:
1579         (TestResultWriterTest.test_reftest_diff_image):
1580
1581 2012-03-19  Dirk Pranke  <dpranke@chromium.org>
1582
1583         webkitpy: fix nit - incorrect reference to DRT in test_expectations.py
1584         https://bugs.webkit.org/show_bug.cgi?id=81598
1585
1586         Reviewed by Eric Seidel.
1587
1588         "DumpRenderTree" shouldn't be hard-coded here, since some other
1589         process may actually be crashing (WebKitTestRunner, WebProcess, etc.)
1590         
1591         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
1592         (TestExpectations):
1593
1594 2012-03-19  Dirk Pranke  <dpranke@chromium.org>
1595
1596         webkitpy: allow instance data in TestFailure objects
1597         https://bugs.webkit.org/show_bug.cgi?id=81580
1598
1599         Reviewed by Tony Chang.
1600
1601         Current TestFailures are basically enums, but it would be nice,
1602         especially for crashes, if they contained the name and pid of
1603         the process that crashed.
1604
1605         * Scripts/webkitpy/layout_tests/models/test_failures.py:
1606         (TestFailure.message):
1607         (FailureTimeout.message):
1608         (FailureCrash):
1609         (FailureCrash.__init__):
1610         (FailureCrash.message):
1611         (FailureMissingResult.message):
1612         (FailureTextMismatch.message):
1613         (FailureMissingImageHash.message):
1614         (FailureMissingImage.message):
1615         (FailureImageHashMismatch.message):
1616         (FailureImageHashIncorrect.message):
1617         (FailureReftestMismatch.message):
1618         (FailureReftestMismatchDidNotOccur.message):
1619         (FailureReftestNoImagesGenerated.message):
1620         (FailureMissingAudio.message):
1621         (FailureAudioMismatch.message):
1622         * Scripts/webkitpy/layout_tests/models/test_failures_unittest.py:
1623         (TestFailuresTest.test_equals):
1624         (TestFailuresTest):
1625         (TestFailuresTest.test_crashes):
1626
1627 2012-03-19  Brady Eidson  <beidson@apple.com>
1628
1629         <rdar://problem/10917120> and https://bugs.webkit.org/show_bug.cgi?id=81066
1630         Crash in 3rd party WebKit apps under XHR/Cache code
1631
1632         Reviewed by Antti Koivisto.
1633
1634         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1635         * TestWebKitAPI/Tests/mac/MemoryCachePruneWithinResourceLoadDelegate.html: Added.
1636         * TestWebKitAPI/Tests/mac/MemoryCachePruneWithinResourceLoadDelegate.mm: Added.
1637         (-[MemoryCachePruneTestResourceLoadDelegate webView:identifierForInitialRequest:fromDataSource:]):
1638         (-[MemoryCachePruneTestResourceLoadDelegate webView:resource:willSendRequest:redirectResponse:fromDataSource:]):
1639         (-[MemoryCachePruneTestResourceLoadDelegate webView:resource:didFinishLoadingFromDataSource:]):
1640         (-[MemoryCachePruneTestResourceLoadDelegate webView:resource:didFailLoadingWithError:fromDataSource:]):
1641         (TestWebKitAPI):
1642         (TestWebKitAPI::TEST):
1643
1644 2012-03-19  Brady Eidson  <beidson@apple.com>
1645
1646         <rdar://problem/10848575> and https://bugs.webkit.org/show_bug.cgi?id=81516
1647         REGRESSION (r107435) URLs copied from WebKit apps aren't in the right pasteboard format
1648
1649         Reviewed by Enrica Casucci.
1650
1651         Bring up a context menu on a link, copy the link, then verify the pasteboard contents
1652         are in the correct format.
1653
1654         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1655         * TestWebKitAPI/Tests/mac/ContextMenuCanCopyURL.html: Added.
1656         * TestWebKitAPI/Tests/mac/ContextMenuCanCopyURL.mm: Added.
1657         (-[ContextMenuCanCopyURLDelegate webView:didFinishLoadForFrame:]):
1658         (TestWebKitAPI::contextMenuCopyLink):
1659         (TestWebKitAPI::TEST):
1660
1661 2012-03-19  Ojan Vafai  <ojan@chromium.org>
1662
1663         Make the selected item in the accordion a link to the flakiness dashboard
1664         https://bugs.webkit.org/show_bug.cgi?id=81571
1665
1666         Reviewed by Dimitri Glazkov.
1667
1668         Also, make it so that you can select the text of the link, instead of the current behavior
1669         of acting as if you had clicked on the accordion header.
1670
1671         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
1672         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/results.css:
1673
1674 2012-03-19  Dirk Pranke  <dpranke@chromium.org>
1675
1676         NRWT runs some tests that are skipped with -i command line option
1677         https://bugs.webkit.org/show_bug.cgi?id=81535
1678
1679         Reviewed by Ojan Vafai.
1680
1681         This change modifies the interaction of Skipped files and
1682         test_expectations files so that entries in Skipped files (and
1683         the -i command line) override *everything* in the expectations
1684         file.
1685
1686         Specifically, a directory in a Skipped file will cause all of
1687         the tests in the dir to be skipped even if individual tests in
1688         the dir are listed in the test_expectations.txt.
1689
1690         Skipped files also override anything in an overrides files.
1691
1692         This seems to make more intuitive sense, since if you list
1693         something in the Skips file (and even more specify it on the
1694         command line) you probably want it to be universally applied.
1695
1696         Theoretically we could add more precedence levels and have full
1697         paths in an expectations file override dirs in a Skipped file
1698         (but not the command line), but I don't know yet that that level
1699         of complexity is justified.
1700         
1701         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
1702         (TestExpectations.__init__):
1703         (TestExpectations._add_skipped_tests):
1704         (TestExpectations._add_test):
1705         (TestExpectations._add_expectations):
1706         (TestExpectations._add_expectation_line):
1707         (TestExpectations._already_seen_better_match):
1708         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
1709         (test_more_specific_override_resets_skip):
1710         (SkippedTests):
1711         (SkippedTests.get_exp):
1712         (SkippedTests.assert_exp):
1713         (SkippedTests.test_skipped_tests_work):
1714         (SkippedTests.test_duplicate_skipped_test_fails_lint):
1715         (SkippedTests.test_skipped_file_overrides_expectations):
1716         (SkippedTests.test_skipped_dir_overrides_expectations):
1717         (SkippedTests.test_skipped_file_overrides_overrides):
1718         (SkippedTests.test_skipped_dir_overrides_overrides):
1719
1720 2012-03-19  Eric Seidel  <eric@webkit.org>
1721
1722         Fix WTF header include discipline in Chromium WebKit
1723         https://bugs.webkit.org/show_bug.cgi?id=81281
1724
1725         Reviewed by James Robinson.
1726
1727         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
1728         * Scripts/update-webkit-chromium: Made it possible to run this from a directory other than the root.
1729
1730 2012-03-19  Andy Chen  <andchen@rim.com>
1731
1732         [BlackBerry] Add new layout test for text match marker
1733         https://bugs.webkit.org/show_bug.cgi?id=81531
1734
1735         Make the LayoutTestControllerBlackBerry::findString to test the
1736         BlackBerry port implementation.
1737
1738         Reviewed by Rob Buis.
1739
1740         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
1741         (LayoutTestController::findString):
1742
1743 2012-03-19  Ojan Vafai  <ojan@chromium.org>
1744
1745         Don't show WONTFIX tests in garden-o-matic
1746         https://bugs.webkit.org/show_bug.cgi?id=81554
1747
1748         Reviewed by Dimitri Glazkov.
1749
1750         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results.js:
1751         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results_unittests.js:
1752
1753 2012-03-19  Chris Fleizach  <cfleizach@apple.com>
1754
1755         platform/mac/accessibility/selected-rows-table.html failing on Lion Intel Debug WebKit2 testers
1756         https://bugs.webkit.org/show_bug.cgi?id=81530
1757
1758         Reviewed by Jessie Berlin.
1759
1760         Add rowAtIndex to WebKitTestRunner.
1761
1762         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.cpp:
1763         (WTR::AccessibilityUIElement::rowAtIndex):
1764         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.h:
1765         (AccessibilityUIElement):
1766         * WebKitTestRunner/InjectedBundle/Bindings/AccessibilityUIElement.idl:
1767         * WebKitTestRunner/InjectedBundle/mac/AccessibilityUIElementMac.mm:
1768         (WTR):
1769         (WTR::AccessibilityUIElement::rowAtIndex):
1770
1771 2012-03-19  Ojan Vafai  <ojan@chromium.org>
1772
1773         Change garden-o-matic All Failures tab to Expected Failures
1774         https://bugs.webkit.org/show_bug.cgi?id=81540
1775
1776         Reviewed by Dimitri Glazkov.
1777
1778         Also, fix some broken unittests that had just gotten stale.
1779
1780         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
1781         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers/rebaseline.js:
1782         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
1783         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model.js:
1784         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results.js:
1785         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results_unittests.js:
1786         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
1787         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures_unittests.js:
1788         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
1789         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:
1790
1791 2012-03-19  Hao Zheng  <zhenghao@chromium.org>
1792
1793         [chromium] DRT crashes at shutdown.
1794         https://bugs.webkit.org/show_bug.cgi?id=81504
1795
1796         Reviewed by James Robinson.
1797
1798         WebCompositor must be destroyed after WebView is properly destroyed,
1799         or some compositor objects like CCLayerTreeHost may not be destroyed.
1800         However, it is really hard to make the destruction order correct
1801         explicitly in ~TestShell. So move it out to DumpRenderTree.
1802
1803         * DumpRenderTree/chromium/DumpRenderTree.cpp:
1804         (main):
1805         * DumpRenderTree/chromium/TestShell.cpp:
1806         (TestShell::~TestShell):
1807
1808 2012-03-19  Dirk Pranke  <dpranke@chromium.org>
1809
1810         add dpranke to webkitpy watchlist
1811         https://bugs.webkit.org/show_bug.cgi?id=81433
1812
1813         Reviewed by Benjamin Poulain.
1814
1815         * Scripts/webkitpy/common/config/watchlist:
1816
1817 2012-03-19  Dirk Pranke  <dpranke@chromium.org>
1818
1819         test-webkitpy is failing on SL/Python 2.6.1
1820         https://bugs.webkit.org/show_bug.cgi?id=81533
1821
1822         Reviewed by Ojan Vafai.
1823
1824         It appears that itertools.combinations() is buggy in 2.6.1; this
1825         backports the implementation from 2.7.
1826
1827         * Scripts/webkitpy/layout_tests/models/test_configuration.py:
1828         (TestConfigurationConverter.collapse_macros):
1829         (TestConfigurationConverter):
1830         (TestConfigurationConverter.combinations):
1831
1832 2012-03-19  David Michael Barr  <davidbarr@google.com>
1833
1834         Optimize the cleanup sequence in the start-queue.sh script
1835         https://github.com/abarth/webkit/pull/1
1836
1837         Reviewed by Adam Barth.
1838
1839         Updating the working copy is an expensive operation
1840         for a repository the size of WebKit.
1841         Attempt to reset only to the target revision.
1842
1843         * EWSTools/start-queue.sh:
1844
1845 2012-03-19  Raphael Kubo da Costa  <kubo@profusion.mobi>
1846
1847         Adjust architecture type of the EFL build bot.
1848         https://bugs.webkit.org/show_bug.cgi?id=81452
1849
1850         Reviewed by Tony Chang.
1851
1852         * BuildSlaveSupport/build.webkit.org-config/config.json: The
1853         buildslave is a 64-bit machine, therefore describe it as being
1854         x86_64, not i386.
1855
1856 2012-03-19  Alexander Færøy  <alexander.faeroy@nokia.com>
1857
1858         [Qt] printLoadedUrls prints "Object" instead of the actual URL.
1859         https://bugs.webkit.org/show_bug.cgi?id=81514
1860
1861         Reviewed by Simon Hausmann.
1862
1863         * MiniBrowser/qt/qml/BrowserWindow.qml:
1864
1865 2012-03-19  Raphael Kubo da Costa  <rakuco@FreeBSD.org>
1866
1867         [EFL] jhbuild: Set CMAKE_PREFIX_PATH in jhbuildrc.
1868         https://bugs.webkit.org/show_bug.cgi?id=81491
1869
1870         Reviewed by Philippe Normand.
1871
1872         For the find_path(), find_library() etc CMake calls to find the
1873         dependencies built with jhbuild, CMAKE_PREFIX_PATH must be properly
1874         set, otherwise CMake cannot know the dependencies exist and will only
1875         look in the standard directories.
1876
1877         * efl/jhbuildrc: Set the CMAKE_PREFIX_PATH environment variable with
1878         the jhbuild Root directory.
1879
1880 2012-03-19  Dominik Röttsches  <dominik.rottsches@linux.intel.com>
1881
1882         [EFL] Add XDG env vars to jhbuildrc
1883         https://bugs.webkit.org/show_bug.cgi?id=81510
1884
1885         Needed to ensure correct glib behaviour in jhbuild environment.
1886
1887         Reviewed by Philippe Normand.
1888
1889         * efl/jhbuildrc:
1890
1891 2012-03-19  Allan Sandfeld Jensen  <allan.jensen@nokia.com>
1892
1893         Select best target for tap gesture.
1894         https://bugs.webkit.org/show_bug.cgi?id=78801
1895
1896         Reviewed by Kenneth Rohde Christiansen.
1897
1898         Add TOUCH_ADJUSTMENT to enabled features.
1899
1900         * qmake/mkspecs/features/features.prf:
1901
1902 2012-03-19  Robert Kroeger  <rjkroege@chromium.org>
1903
1904         [chromium] synthesize wheel events for fling on main thread
1905         https://bugs.webkit.org/show_bug.cgi?id=81458
1906
1907         Added the ability to dispatch GestureFlingStart and GestureFlingCancel from EventSender.
1908
1909         Reviewed by James Robinson.
1910
1911         * DumpRenderTree/chromium/EventSender.cpp:
1912         (EventSender::EventSender):
1913         (EventSender::gestureFlingCancel):
1914         (EventSender::gestureFlingStart):
1915         * DumpRenderTree/chromium/EventSender.h:
1916         (EventSender):
1917
1918 2012-03-19  Alexander Færøy  <alexander.faeroy@nokia.com>
1919
1920         [Qt] Add command line option to set the user agent in the MiniBrowser
1921         https://bugs.webkit.org/show_bug.cgi?id=81508
1922
1923         Reviewed by Simon Hausmann.
1924
1925         * MiniBrowser/qt/BrowserWindow.cpp:
1926         (BrowserWindow::BrowserWindow):
1927         (BrowserWindow::webViewExperimental):
1928         * MiniBrowser/qt/BrowserWindow.h:
1929         (BrowserWindow):
1930         * MiniBrowser/qt/MiniBrowserApplication.cpp:
1931         (printHelp):
1932         (MiniBrowserApplication::handleUserOptions):
1933         * MiniBrowser/qt/MiniBrowserApplication.h:
1934         (WindowOptions::WindowOptions):
1935         (WindowOptions::setUserAgent):
1936         (WindowOptions::userAgent):
1937         (WindowOptions):
1938
1939 2012-03-19  Alexander Færøy  <alexander.faeroy@nokia.com>
1940
1941         [Qt] Remove dead user agent code from the MiniBrowser.
1942
1943         Reviewed by Simon Hausmann.
1944
1945         * MiniBrowser/qt/BrowserWindow.cpp:
1946         * MiniBrowser/qt/BrowserWindow.h:
1947         (BrowserWindow):
1948         * MiniBrowser/qt/MiniBrowser.qrc:
1949         * MiniBrowser/qt/useragentlist.txt: Removed.
1950
1951 2012-03-19  Raphael Kubo da Costa  <rakuco@FreeBSD.org>
1952
1953         jhbuild: Call jhbuild's autogen.sh via bash.
1954         https://bugs.webkit.org/show_bug.cgi?id=81485
1955
1956         Reviewed by Philippe Normand.
1957
1958         jhbuild's autogen.sh uses "/bin/bash" in the shebang line, however
1959         this is usually true only on Linux systems, and even then bash for
1960         some reason might not be available in that location. It is safer to
1961         call bash directly and pass autogen.sh to it instead.
1962
1963         * jhbuild/jhbuild-wrapper:
1964         (install_jhbuild):
1965
1966 2012-03-19  Philippe Normand  <pnormand@igalia.com>
1967
1968         [GTK] Prefix run-launcher with run-with-jhbuild
1969         https://bugs.webkit.org/show_bug.cgi?id=81503
1970
1971         Reviewed by Hajime Morita.
1972
1973         * Scripts/run-launcher:
1974
1975 2012-03-19  Dominik Röttsches  <dominik.rottsches@linux.intel.com>
1976
1977         [EFL] Prefix run-launcher with run-with-jhbuild
1978         https://bugs.webkit.org/show_bug.cgi?id=80026
1979
1980         Reviewed by Hajime Morita.
1981
1982         * Scripts/run-launcher:
1983
1984 2012-03-18  Csaba Osztrogonác  <ossy@webkit.org>
1985
1986         [Qt] Unreviewed. Disable -Werror on x86 until proper fix.
1987         https://bugs.webkit.org/show_bug.cgi?id=81498
1988
1989         * qmake/mkspecs/features/unix/default_post.prf:
1990
1991 2012-03-17  Raphael Kubo da Costa  <rakuco@FreeBSD.org>
1992
1993         Unreviewed; switch to using my FreeBSD.org email address, the
1994         other one is going to disappear soon.
1995
1996         * Scripts/webkitpy/common/config/committers.py:
1997         * Scripts/webkitpy/common/config/watchlist:
1998
1999 2012-03-17  Raphael Kubo da Costa  <kubo@profusion.mobi>
2000
2001         Unreviewed; fix webkitpy unit test failures after r111116.
2002
2003         It turns out that cmd_line in the patch in bug 80025 had the right
2004         signature when it was submitted, but r109242 added a few more
2005         required parameters, and since the method (and EflDriver) did not
2006         exist at the time we ended up committing a broken method.
2007
2008         * Scripts/webkitpy/layout_tests/port/efl.py:
2009         (EflDriver.cmd_line): Add the `pixel_tests' and `per_test_args'
2010         arguments.
2011
2012 2012-03-17  Charles Wei  <charles.wei@torchmobile.com.cn>
2013
2014         [BlackBerry] Enable Web Timing for performance profiling and improvement.
2015         https://bugs.webkit.org/show_bug.cgi?id=81085
2016
2017         Reviewed by Rob Buis.
2018
2019         * Scripts/build-webkit:
2020
2021 2012-03-17  Dominik Röttsches  <dominik.rottsches@linux.intel.com>
2022
2023         Add a dependencies installation step for EFL to buildbot master.cfg
2024         https://bugs.webkit.org/show_bug.cgi?id=81337
2025
2026         Reviewed by Gustavo Noronha Silva.
2027
2028         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2029         (InstallEflDependencies):
2030         (Factory.__init__):
2031
2032 2012-03-17  Dominik Röttsches  <dominik.rottsches@linux.intel.com>
2033
2034         [EFL] Prefixing run-webkit-tests with run-with-jhbuild
2035         https://bugs.webkit.org/show_bug.cgi?id=80025
2036
2037         Reviewed by Gustavo Noronha Silva.
2038
2039         * Scripts/webkitpy/layout_tests/port/efl.py:
2040         (EflDriver):
2041         (EflDriver.cmd_line):
2042         (EflPort._driver_class):
2043
2044 2012-03-17  Dominik Röttsches  <dominik.rottsches@linux.intel.com>
2045
2046         [EFL] Add and use run-with-jhbuild and update-webkitefl-libs scripts for EFL
2047         https://bugs.webkit.org/show_bug.cgi?id=79904
2048
2049         build-webkit --efl will use jhbuild for fetching and building dependencies.
2050
2051         Reviewed by Gustavo Noronha Silva.
2052
2053         * Scripts/update-webkitefl-libs: Added.
2054         * Scripts/webkitdirs.pm:
2055         (jhbuildWrapperPrefixIfNeeded):
2056         (generateBuildSystemFromCMakeProject):
2057         (buildCMakeGeneratedProject):
2058         * efl/common.py: Added.
2059         (script_path):
2060         (top_level_path):
2061         (number_of_cpus):
2062         * efl/jhbuild.modules: Added.
2063         * efl/jhbuildrc: Added.
2064         * efl/run-with-jhbuild: Added.
2065
2066 2012-03-16  Stephanie Lewis  <slewis@apple.com>
2067
2068         https://bugs.webkit.org/show_bug.cgi?id=81065
2069         <rdar://problem/10944309> CrashTracer: [USER] 111 crashes in WebProcess at WebKitTestRunnerInjectedBundle: WTR::InjectedBundle::done + 142
2070         Fix crash preventing WebKitTestRunner from working in Release.  
2071         WebKitTestRunner used ostringstream to collect output from the test.  ostringstream is incompatible with our
2072         FastMalloc overrides.  When copying strings out ostringstream a new string is allocated with FastMalloc because
2073         the allocation is inlined.  When that string is freed is uses the system malloc because that call is not inlined.
2074
2075         I removed ostringstream in favor of StringBuilder.  
2076
2077         Properly set NDEBUG on Release builds so we don't crash when when we free the strings StringBuilder created.
2078
2079         Reviewed by Geoff Garen.
2080
2081         Add DEBUG_DEFINES to set NDEBUG on release builds.
2082         * WebKitTestRunner/Configurations/Base.xcconfig:
2083         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
2084
2085         Remove ostringstream member in favor of StringBuilder.
2086         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
2087         (WTR::InjectedBundle::initialize):
2088         (WTR::InjectedBundle::beginTesting):
2089         (WTR::InjectedBundle::done):
2090         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
2091         (WTF):
2092         (WTR::InjectedBundle::stringBuilder):
2093         (InjectedBundle):
2094
2095         Append output to StringBuilder.
2096         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
2097         (WTR::hasPrefix):
2098         (WTR::dumpPath):
2099         (WTR::rangeToStr):
2100         (WTR::styleDecToStr):
2101         (WTR::frameToStr):
2102         (WTR::dumpFrameScrollPosition):
2103         (WTR::dumpFrameText):
2104         (WTR::dumpDescendantFramesText):
2105         (WTR::InjectedBundlePage::dump):
2106         (WTR::InjectedBundlePage::didReceiveTitleForFrame):
2107         (WTR::InjectedBundlePage::didFinishDocumentLoadForFrame):
2108         (WTR::InjectedBundlePage::willSendRequestForFrame):
2109         (WTR::InjectedBundlePage::decidePolicyForNavigationAction):
2110         (WTR::lastFileURLPathComponent):
2111         (WTR::InjectedBundlePage::willAddMessageToConsole):
2112         (WTR::InjectedBundlePage::willSetStatusbarText):
2113         (WTR::InjectedBundlePage::willRunJavaScriptAlert):
2114         (WTR::InjectedBundlePage::willRunJavaScriptConfirm):
2115         (WTR::InjectedBundlePage::willRunJavaScriptPrompt):
2116         (WTR::InjectedBundlePage::shouldBeginEditing):
2117         (WTR::InjectedBundlePage::shouldEndEditing):
2118         (WTR::InjectedBundlePage::shouldInsertNode):
2119         (WTR::InjectedBundlePage::shouldInsertText):
2120         (WTR::InjectedBundlePage::shouldDeleteRange):
2121         (WTR::InjectedBundlePage::shouldChangeSelectedRange):
2122         (WTR::InjectedBundlePage::shouldApplyStyle):
2123         (WTR::InjectedBundlePage::didBeginEditing):
2124         (WTR::InjectedBundlePage::didEndEditing):
2125         (WTR::InjectedBundlePage::didChange):
2126         (WTR::InjectedBundlePage::didChangeSelection):
2127         (WTR::InjectedBundlePage::supportsFullScreen):
2128         (WTR::InjectedBundlePage::enterFullScreenForElement):
2129         (WTR::InjectedBundlePage::exitFullScreenForElement):
2130         (WTR::InjectedBundlePage::beganEnterFullScreen):
2131         (WTR::InjectedBundlePage::beganExitFullScreen):
2132         (WTR::dumpBackForwardListItem):
2133         (WTR::InjectedBundlePage::dumpBackForwardList):
2134         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
2135         (WTR::LayoutTestController::waitToDumpWatchdogTimerFired):
2136         (WTR::LayoutTestController::dumpConfigurationForViewport):
2137         
2138         Convert between WTF::String and WKStringRef.
2139         * WebKitTestRunner/StringFunctions.h:
2140         (WTR::toWTFString):
2141         (WTR):
2142         * WebKitTestRunner/TestInvocation.cpp:
2143         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle):
2144
2145 2012-03-16  Eric Seidel  <eric@webkit.org>
2146
2147         Add a land-from-url command
2148         https://bugs.webkit.org/show_bug.cgi?id=81411
2149
2150         Reviewed by Adam Barth.
2151
2152         Very basic so far.  The goal is to make it landing smarter
2153         so that users don't have to bother to parse out the ids out of the urls.
2154
2155         * Scripts/webkitpy/common/config/urls.py:
2156         (parse_bug_id):
2157         (parse_attachment_id):
2158         * Scripts/webkitpy/common/config/urls_unittest.py:
2159         (URLsTest.test_parse_bug_id):
2160         (URLsTest):
2161         (URLsTest.test_parse_attachment_id):
2162         * Scripts/webkitpy/tool/commands/download.py:
2163         (ProcessURLsMixin):
2164         (ProcessURLsMixin._fetch_list_of_patches_to_process):
2165         (LandFromURL):
2166         * Scripts/webkitpy/tool/commands/download_unittest.py:
2167         (test_land_from_bug):
2168         (test_land_from_url):
2169
2170 2012-03-16  Dave Tharp  <dtharp@codeaurora.org>
2171
2172         build-webkit launcher instructions use wrong relative path for Qt
2173         https://bugs.webkit.org/show_bug.cgi?id=81357
2174
2175         Reviewed by Eric Seidel.
2176
2177         Ensure we are at the proper directory before generating run-launcher
2178         instructions.
2179
2180         * Scripts/build-webkit:
2181
2182 2012-03-16  Brady Eidson  <beidson@apple.com>
2183
2184         <rdar://problem/11027997> and https://bugs.webkit.org/show_bug.cgi?id=81412
2185         REGRESSION (r107435) Copy a link and paste to Mail: Nothing is pasted
2186
2187         Reviewed by Geoff Garen.
2188
2189         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
2190         * TestWebKitAPI/Tests/mac/WebViewCanPasteURL.mm: Added.
2191
2192 2012-03-16  Ojan Vafai  <ojan@chromium.org>
2193
2194         Mark wontfix tests in full_results.json so we can ignore them in garden-o-matic
2195         https://bugs.webkit.org/show_bug.cgi?id=81407
2196
2197         Reviewed by Adam Barth.
2198
2199         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2200         (summarize_results):
2201         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
2202         (ResultSummaryTest.summarized_results):
2203         (ResultSummaryTest.test_no_svn_revision):
2204         (ResultSummaryTest.test_svn_revision):
2205         (ResultSummaryTest):
2206         (ResultSummaryTest.test_summarized_results_wontfix):
2207
2208 2012-03-16  Ojan Vafai  <ojan@chromium.org>
2209
2210         Have webkit-patch rebaseline-test update test_expectations.txt
2211         https://bugs.webkit.org/show_bug.cgi?id=81054
2212
2213         Recommit http://trac.webkit.org/changeset/110783 now that
2214         http://trac.webkit.org/changeset/111034 has landed.
2215
2216         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
2217         (TestExpectations.remove_configuration_from_test):
2218         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
2219         (RemoveConfigurationsTest):
2220         (RemoveConfigurationsTest.test_remove):
2221         (test_remove_line):
2222         * Scripts/webkitpy/tool/commands/rebaseline.py:
2223         (RebaselineTest._update_expectations_file):
2224         (RebaselineTest._rebaseline_test_and_update_expectations):
2225         (RebaselineTest):
2226         (RebaselineTest.execute):
2227         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
2228         (TestRebaseline.test_rebaseline_updates_expectations_file_noop):
2229         (test_rebaseline_updates_expectations_file):
2230
2231 2012-03-15  Ojan Vafai  <ojan@chromium.org>
2232
2233         Specifier collapsing when writing test expectations lines gets a number of cases wrong
2234         https://bugs.webkit.org/show_bug.cgi?id=81309
2235
2236         Reviewed by Dimitri Glazkov.
2237
2238         I've run this over all the lines in the current Chromium test_expectations.txt file,
2239         so I'm relatively confident we now cover all the cases.
2240
2241         * Scripts/webkitpy/layout_tests/models/test_configuration.py:
2242         (TestConfigurationConverter.__init__):
2243         (TestConfigurationConverter.collapse_macros):
2244         (TestConfigurationConverter.collapse_macros.collapse_individual_specifier_set):
2245         (TestConfigurationConverter):
2246         (TestConfigurationConverter.intersect_combination):
2247         (TestConfigurationConverter.symmetric_difference):
2248         (TestConfigurationConverter.to_specifiers_list):
2249         (TestConfigurationConverter.to_specifiers_list.try_collapsing):
2250         (TestConfigurationConverter.to_specifiers_list.try_abbreviating):
2251         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py:
2252         (make_mock_all_test_configurations_set):
2253         (TestConfigurationConverterTest.test_symmetric_difference):
2254         (TestConfigurationConverterTest.test_to_config_set):
2255         (TestConfigurationConverterTest.test_macro_expansion):
2256         (TestConfigurationConverterTest.test_to_specifier_lists):
2257         (TestConfigurationConverterTest.test_converter_macro_collapsing):
2258         * Scripts/webkitpy/layout_tests/port/chromium.py:
2259         (ChromiumPort):
2260         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2261         (ChromiumAndroidPort.__init__):
2262         The android port uses "arm" as it's architecture, which is technically correct,
2263         but considerably complicates making collapsing work. We probably should kill
2264         the concept of architecture entirely. The benefits are not worth the code
2265         complexity.
2266
2267         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
2268         (ChromiumPortTest.test_all_test_configurations):
2269         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
2270         (TestRebaseline.test_rebaseline_updates_expectations_file_noop):
2271         (test_rebaseline_updates_expectations_file):
2272
2273 2012-03-16  Dinu Jacob  <dinu.jacob@nokia.com>
2274
2275         [Qt][Wk2] Assertion Failure and crash on file upload
2276         https://bugs.webkit.org/show_bug.cgi?id=80854
2277
2278         Reviewed by Simon Hausmann.
2279
2280         Added filePicker to WebView using experimental API.
2281
2282         * MiniBrowser/qt/MiniBrowser.qrc:
2283         * MiniBrowser/qt/icons/folder.png: Added.
2284         * MiniBrowser/qt/icons/titlebar.png: Added.
2285         * MiniBrowser/qt/icons/up.png: Added.
2286         * MiniBrowser/qt/qml/BrowserWindow.qml:
2287         * MiniBrowser/qt/qml/FilePicker.qml: Added.
2288
2289 2012-03-16  Mahesh Kulkarni  <mahesh.kulkarni@nokia.com>
2290
2291         Updating email for committer.py script.
2292
2293         * Scripts/webkitpy/common/config/committers.py:
2294
2295 2012-03-16  Robert Kroeger  <rjkroege@chromium.org>
2296
2297         [chromium] DRT: Add support for sending gesture scrolls without a begin
2298         https://bugs.webkit.org/show_bug.cgi?id=81292
2299
2300         Reviewed by Kent Tamura.
2301
2302         * DumpRenderTree/chromium/EventSender.cpp:
2303         (EventSender::EventSender):
2304         (EventSender::gestureScrollFirstPoint):
2305         * DumpRenderTree/chromium/EventSender.h:
2306         (EventSender):
2307
2308 2012-03-02  Carlos Garcia Campos  <cgarcia@igalia.com> and Sergio Villar Senin  <svillar@igalia.com>
2309
2310         [GTK] WebKitIconDatabase doesn't keep icons cached
2311         https://bugs.webkit.org/show_bug.cgi?id=56200
2312
2313         Reviewed by Martin Robinson.
2314
2315         Add support for favicons in GtkLauncher.
2316
2317         * GtkLauncher/main.c:
2318         (activateUriEntryCb):
2319         (iconLoadedCb):
2320         (createBrowser):
2321
2322 2012-03-16  Adam Barth  <abarth@webkit.org>
2323
2324         Remove sheriff-bot's last-green-revision command
2325         https://bugs.webkit.org/show_bug.cgi?id=81314
2326
2327         Reviewed by Eric Seidel.
2328
2329         This command is a cruel joke.  WebKit doesn't really ever have a
2330         last-green-revision.
2331
2332         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
2333         (BuildBot._find_green_revision):
2334         * Scripts/webkitpy/common/net/buildbot/buildbot_mock.py:
2335         (MockBuildBot.builder_statuses):
2336         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2337         (test_find_green_revision):
2338         * Scripts/webkitpy/tool/bot/irc_command.py:
2339         (IRCCommand.execute):
2340         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
2341         (SheriffIRCBotTest.test_help):
2342         * Scripts/webkitpy/tool/commands/queries.py:
2343         (PatchesToReview.execute):
2344
2345 2012-03-16  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2346
2347         [Qt] Make gccdepends test aware of broken icecream deps generation
2348
2349         Icecream seems to build the files without passing -o to gcc with
2350         the original destination path, which means the target name in the
2351         dependency file is wrong.
2352
2353         Also, echo something to the final file instead of just touching it,
2354         so it's not deleted on every build.
2355
2356         Reviewed by Simon Hausmann.
2357
2358         * qmake/config.tests/gccdepends/gccdepends.pro:
2359
2360 2012-03-16  Kihong Kwon  <kihong.kwon@samsung.com>
2361
2362         Support for Battery Status API
2363         https://bugs.webkit.org/show_bug.cgi?id=62698
2364
2365         Add include directory and file.
2366         http://www.w3.org/TR/battery-status/
2367
2368         Reviewed by Adam Barth.
2369
2370         * Scripts/build-webkit:
2371
2372 2012-03-16  Kishore Bolisetty  <kbolisetty@innominds.com>
2373
2374         check-webkit-style fails on LayoutTests/platform/efl/test_expectations.txt
2375         https://bugs.webkit.org/show_bug.cgi?id=81007
2376
2377         Reviewed by Adam Barth.
2378
2379         A dedidcated build slave is available for efl on build.webkit.org. 
2380         This change add's an entry for the elf port in list of available port builders.
2381         This is required because test_expectations.py when triggered, looks for a valid port object for the specified port. 
2382
2383         * Scripts/webkitpy/layout_tests/port/builders.py:
2384
2385 2012-03-16  Adam Barth  <abarth@webkit.org>
2386
2387         [webkitpy] StatusServer.next_work_item has an incorrectly named local variable
2388         https://bugs.webkit.org/show_bug.cgi?id=81302
2389
2390         Reviewed by Eric Seidel.
2391
2392         This is a trivial patch, but the variable name turned up in a grep of
2393         webkitpy and looks like copy pasta.
2394
2395         * Scripts/webkitpy/common/net/statusserver.py:
2396         (StatusServer.next_work_item):
2397
2398 2012-03-16  Andrew Lo  <anlo@rim.com>
2399
2400         [BlackBerry] Enable PAGE_VISIBILITY_API
2401         https://bugs.webkit.org/show_bug.cgi?id=81257
2402
2403         Enable Page Visibility API for BlackBerry.
2404
2405         Reviewed by Rob Buis.
2406
2407         * Scripts/build-webkit:
2408
2409 2012-03-16  Zan Dobersek  <zandobersek@gmail.com>
2410
2411         [Gtk] Add computedStyleIncludingVisitedInfo to DumpRenderTreeSupport
2412         https://bugs.webkit.org/show_bug.cgi?id=81232
2413
2414         Reviewed by Philippe Normand.
2415
2416         Use the newly-added helper method in DumpRenderTreeSupportGtk.
2417
2418         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
2419         (LayoutTestController::computedStyleIncludingVisitedInfo):
2420
2421 2012-03-15  Eric Seidel  <eric@webkit.org>
2422
2423         Move parse_bug_id into config/urls in preparation for re-use elsewhere.
2424         https://bugs.webkit.org/show_bug.cgi?id=81313
2425
2426         Reviewed by Adam Barth.
2427
2428         * Scripts/webkitpy/common/checkout/changelog.py:
2429         (parse_bug_id_from_changelog):
2430         * Scripts/webkitpy/common/config/urls.py:
2431         (parse_bug_id):
2432         * Scripts/webkitpy/common/config/urls_unittest.py: Copied from Tools/Scripts/webkitpy/common/config/urls.py.
2433         (URLsTest):
2434         (URLsTest.test_parse_bug_id):
2435         * Scripts/webkitpy/common/net/bugzilla/bugzilla_unittest.py:
2436         (test_parse_bug_id):
2437         * Scripts/webkitpy/tool/bot/irc_command.py:
2438         (_post_error_and_check_for_bug_url):
2439         * Scripts/webkitpy/tool/bot/sheriff.py:
2440         (Sheriff.post_rollout_patch):
2441         (Sheriff.post_chromium_deps_roll):
2442
2443 2012-03-15  Adam Barth  <abarth@webkit.org>
2444
2445         The commit-queue should fast-track patches that have already passed the testing EWS bots
2446         https://bugs.webkit.org/show_bug.cgi?id=81305
2447
2448         Reviewed by Eric Seidel.
2449
2450         If a patch has already passed the testing EWS bots (currently only the
2451         chromium-ews bot), we should be able to land the patch without running
2452         the tests again.  This patch is somewhat risky in that the tests might
2453         have been broken between when the EWS bots ran and when the
2454         commit-queue is processing the patch.  If that turns out to be a
2455         problem in practice, we'll tighten up this fast-track in some way,
2456         likely by making sure that the pass status is recent.
2457
2458         * Scripts/webkitpy/tool/bot/commitqueuetask.py:
2459         (CommitQueueTaskDelegate.did_pass_testing_ews):
2460         (CommitQueueTask._did_pass_tests_recently):
2461         (CommitQueueTask.run):
2462         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
2463         (MockCommitQueue.did_pass_testing_ews):
2464         (test_fast_success_case):
2465         * Scripts/webkitpy/tool/commands/queues.py:
2466         (CommitQueue.did_pass_testing_ews):
2467         * Scripts/webkitpy/tool/commands/queues_unittest.py:
2468         (test_did_pass_testing_ews):
2469
2470 2012-03-15  Kentaro Hara  <haraken@chromium.org>
2471
2472         [PerformanceTests] run-perf-tests should output correct units
2473         https://bugs.webkit.org/show_bug.cgi?id=78303
2474
2475         Reviewed by Ryosuke Niwa.
2476
2477         Previously run-perf-tests just supported "ms" for units.
2478         Consequently, Dromaeo perf tests had been reporting runs/s as "ms".
2479         This patch fixes run-perf-tests so that they can support custom units.
2480         The default unit is "ms".
2481
2482         The test result looks like this:
2483
2484         $ ./Tools/Scripts/run-perf-tests Dromaeo/dom-attr.html
2485         RESULT Dromaeo: dom-attr= 6465.9525483 runs/s
2486         median= 0.0 runs/s, stdev= 24.2983433436 runs/s, min= 6435.87649402 runs/s, max= 6515.63693392 runs/s
2487
2488         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2489         (PerfTestsRunner._process_parser_test_result):
2490
2491 2012-03-15  Brent Fulgham  <bfulgham@webkit.org>
2492
2493         [WinCairo] Unreviewed build change after wtf path changes.
2494
2495         * DumpRenderTree/win/ImageDiffWinCairoCommon.vsprops: Updated include
2496         directories to look at the private header directory to find 'wtf'.
2497
2498 2012-03-15  Kent Tamura  <tkent@chromium.org>
2499
2500         REGRESSION(r110595): webkit-patch suggest-reviewers doesn't work
2501         https://bugs.webkit.org/show_bug.cgi?id=81293
2502
2503         Reviewed by Adam Barth.
2504
2505         r110595 introduced new format of the author field of ChangeLog entry:
2506         "Adam Barth  <abarth@webkit.org> && Benjamin Poulain  <bpoulain@apple.com>"
2507         Updte chanelog.py to support it.
2508
2509         * Scripts/webkitpy/common/checkout/changelog.py:
2510         (ChangeLogEntry._split_contributor_names):
2511         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
2512         (test_parse_authors):
2513
2514 2012-03-15  Lynn Neir  <lynn.neir@skype.net>
2515
2516         Implement LayoutTestController::setValueForUser on Windows https://bugs.webkit.org/show_bug.cgi?id=55834
2517
2518         Implemented setValueForUser in both DumpRenderTree and WebKitTestRunner.
2519
2520         Reviewed by Jessie Berlin.
2521
2522         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
2523         (LayoutTestController::setValueForUser):
2524         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
2525         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
2526         (WTR::LayoutTestController::setValueForUser):
2527         (WTR):
2528         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
2529         (LayoutTestController):
2530
2531 2012-03-15  Adam Barth  <abarth@webkit.org>
2532
2533         Update the start-queue.sh script to match what we're using on the bots
2534         https://bugs.webkit.org/show_bug.cgi?id=81277
2535
2536         Reviewed by Eric Seidel.
2537
2538         Over time we've added more magic git commands to clean up the local repo after
2539         it has been borked in various ways. The idea behind this sequence is to clear
2540         out any changes in the working copy and then to rebuild the master branch from
2541         origin/master.
2542
2543         This patch also removes the watchdog timer that we've stopped using on the
2544         bots.
2545
2546         This patch was reviewed on <https://github.com/eseidel/webkit/pull/1>.
2547         Note: Eric and I are experimenting with GitHub to see if it might be
2548         useful for WebKit.  If you have thoughts on this topic, please feel
2549         free to grab Eric or me on #webkit to chat.  I assume that he and I are
2550         the only ones who care about this file so using this alternative
2551         process for this patch shouldn't cause any trouble.
2552
2553         * EWSTools/start-queue.sh:
2554
2555 2012-03-15  Adam Barth  <abarth@webkit.org>
2556
2557         The commit-queue can hang when the test_expectations.txt style check fails during commit
2558         https://bugs.webkit.org/show_bug.cgi?id=81251
2559
2560         Reviewed by Eric Seidel.
2561
2562         We forgot to check the --non-interactive flag, so we end up prompting
2563         the "user" who doesn't exist on the bot.  We might consider teaching
2564         the User class about non-interactive if we have this sort of bug again.
2565
2566         * Scripts/webkitpy/tool/steps/commit.py:
2567         (Commit.options):
2568         (Commit._check_test_expectations):
2569         (Commit.run):
2570         * Scripts/webkitpy/tool/steps/commit_unittest.py:
2571         (CommitTest.test_check_test_expectations):
2572
2573 2012-03-15  Jessie Berlin  <jberlin@apple.com>
2574
2575         fast/dom/Window/window-properties.html and fast/dom/prototype-inheritance-2.html failing on
2576         Mac WK2 because WebAudio is not enabled
2577         https://bugs.webkit.org/show_bug.cgi?id=81259
2578
2579         Reviewed by Dan Bernstein.
2580
2581         Enable WebAudio so that it appears in the window object in the tests.
2582         The webaudio tests are all still skipped in the Mac Skipped list.
2583
2584         * WebKitTestRunner/TestController.cpp:
2585         (WTR::TestController::resetStateToConsistentValues):
2586
2587 2012-03-15  David Levin  <levin@chromium.org>
2588
2589         Various autoresize fixes.
2590         https://bugs.webkit.org/show_bug.cgi?id=81199
2591
2592         Reviewed by Darin Fisher.
2593
2594         * DumpRenderTree/chromium/LayoutTestController.cpp:
2595         (LayoutTestController::reset): Stop using the deprecated api that was removed.
2596
2597 2012-03-15  Carlos Garcia Campos  <cgarcia@igalia.com>
2598
2599         [GTK] Actually use --release and --debug command line options in run-gtk-test
2600         https://bugs.webkit.org/show_bug.cgi?id=81234
2601
2602         Reviewed by Philippe Normand.
2603
2604         Actually handle --release and --debug command line options to
2605         create the build directory, falling back to current hack to detect
2606         whether it's a Release or Debug build when both options are missed.
2607
2608         * Scripts/run-gtk-tests:
2609         (TestRunner.__init__): Receive options too.
2610         (TestRunner._get_top_level_directory): Helper function to get the
2611         top level directory.
2612         (TestRunner._get_build_directory): Helper function to get the
2613         build directory depending on --release/--debug command line
2614         options or gessing it if both options are missing.
2615         (TestRunner._ensure_accessibility_daemon_is_running): Use
2616         self-options now.
2617         (TestRunner.run): Ditto.
2618         (TestRunner.run.run_tests): Pass options to the constructor
2619         instead of run().
2620
2621 2012-03-15  Carlos Garcia Campos  <cgarcia@igalia.com>
2622
2623         [GTK] Run unit tests with --verbose in the bots
2624         https://bugs.webkit.org/show_bug.cgi?id=81230
2625
2626         Reviewed by Martin Robinson.
2627
2628         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2629         (RunGtkAPITests): Add --verbose to run-gtk-test command.
2630
2631 2012-03-15  Kevin Ollivier  <kevino@theolliviers.com>
2632
2633         Move wx port to using export macros
2634         https://bugs.webkit.org/show_bug.cgi?id=77279
2635
2636         Reviewed by Hajime Morita.
2637
2638         * waf/build/settings.py:
2639         (common_configure):
2640
2641 2012-03-15  Philippe Normand  <pnormand@igalia.com>
2642
2643         Unreviewed, GTK build fix attempt after r110845.
2644
2645         * Scripts/run-gtk-tests: Added --release and --debug options.
2646
2647 2012-03-15  Dominik Röttsches  <dominik.rottsches@linux.intel.com>
2648
2649         Parametrize run-with-jhbuild and update-webkitgtk-libs with platform --gtk/--efl
2650         https://bugs.webkit.org/show_bug.cgi?id=79673
2651
2652         Preparatory work for bringing jhbuild enabled build to EFL port.
2653
2654         Reviewed by Philippe Normand.
2655
2656         * Scripts/update-webkit-libs-jhbuild: Copied from Tools/Scripts/update-webkitgtk-libs.
2657         * Scripts/update-webkitgtk-libs:
2658         * gtk/run-with-jhbuild:
2659         * jhbuild/jhbuild-wrapper: Copied from Tools/gtk/run-with-jhbuild.
2660         (top_level_path):
2661         (jhbuild_installed):
2662         (jhbuild_cloned):
2663         (jhbuild_at_expected_revision):
2664         (update_jhbuild):
2665         (clone_jhbuild):
2666         (install_jhbuild):
2667         (update_webkit_libs_jhbuild):
2668         (determine_platform):
2669         (ensure_jhbuild):
2670
2671 2012-03-15  Carlos Garcia Campos  <cgarcia@igalia.com>
2672
2673         [GTK] Add command line options to run-gtk-tests script
2674         https://bugs.webkit.org/show_bug.cgi?id=81218
2675
2676         Reviewed by Philippe Normand.
2677
2678         Following command line options have been added:
2679
2680         -v, --verbose: to run gtester with --verbose. That would be used
2681         by the bots to make it easier to know which test case it failing.
2682
2683         --display: to use a different display for Xvfb
2684
2685         It's also possible to pass one or more tests as arguments.
2686
2687         * Scripts/run-gtk-tests:
2688         (TestRunner.__init__): Receive command line arguments.
2689         (TestRunner._get_tests): Get all the tests or the ones passed by
2690         command line arguments.
2691         (TestRunner._ensure_accessibility_daemon_is_running): Use display
2692         command line argument.
2693         (TestRunner.run): Pas command line options.
2694         (TestRunner.run.run_tests): Pass command line arguments.
2695
2696 2012-03-15  Carlos Garcia Campos  <cgarcia@igalia.com>
2697
2698         [GTK] WebKit2APITests/TestPrinting timeout in the bots
2699         https://bugs.webkit.org/show_bug.cgi?id=81211
2700
2701         Reviewed by Philippe Normand.
2702
2703         Add gtk+ patch to fix timeout in the bots. A similar patch has
2704         already landed in updatream gtk+, see
2705         http://git.gnome.org/browse/gtk+/commit/?id=5ebfad5d091f4c9d9c6e60364fad4c60842f027f
2706
2707         * gtk/jhbuild.modules:
2708         * gtk/patches/gtk+-3.0.12-gtkprinter.patch: Added.
2709
2710 2012-03-14  Gustavo Noronha Silva  <gns@gnome.org>
2711
2712         sheriffbot should also be addressable with a comma in addition to colon
2713         https://bugs.webkit.org/show_bug.cgi?id=81101
2714
2715         Reviewed by Adam Barth.
2716
2717         * Scripts/webkitpy/common/net/irc/ircbot.py:
2718         (IRCBot.on_pubmsg):
2719
2720 2012-03-15  Alexander Færøy  <alexander.faeroy@nokia.com>
2721
2722         [Qt] MiniBrowser scroll indicators should not scroll out of view
2723         https://bugs.webkit.org/show_bug.cgi?id=81131
2724
2725         Reviewed by Kenneth Rohde Christiansen.
2726
2727         * MiniBrowser/qt/qml/ScrollIndicator.qml:
2728
2729 2012-03-14  Chris Fleizach  <cfleizach@apple.com>
2730
2731         AX: ARIA tables need to support selected rows attribute/select rows changed notification
2732         https://bugs.webkit.org/show_bug.cgi?id=79821
2733
2734         Reviewed by Beth Dakin.
2735
2736         * DumpRenderTree/AccessibilityUIElement.cpp:
2737         (rowAtIndexCallback):
2738         (AccessibilityUIElement::getJSClass):
2739         * DumpRenderTree/AccessibilityUIElement.h:
2740         (AccessibilityUIElement):
2741         * DumpRenderTree/blackberry/AccessibilityUIElementBlackBerry.cpp:
2742         (AccessibilityUIElement::rowAtIndex):
2743         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
2744         (AccessibilityUIElement::rowAtIndex):
2745         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
2746         (AccessibilityUIElement::rowAtIndex):
2747         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
2748         (AccessibilityUIElement::rowAtIndex):
2749
2750 2012-03-14  Tim Horton  <timothy_horton@apple.com>
2751
2752         NRWT should respect --additional-platform-directory's Skipped list
2753         https://bugs.webkit.org/show_bug.cgi?id=81172
2754
2755         Reviewed by Timothy Hatcher.
2756
2757         Respect the Skipped file inside all paths included via the --additional-platform-directory option. Test results
2758         from said path are already used, but the Skipped file is ignored.
2759
2760         * Scripts/webkitpy/layout_tests/port/webkit.py:
2761         (WebKitPort._skipped_file_search_paths):
2762         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
2763         (test_skipped_file_search_paths):
2764
2765 2012-03-14  Ojan Vafai  <ojan@chromium.org>
2766
2767         Unreviewed, rolling out r110783.
2768         http://trac.webkit.org/changeset/110783
2769         https://bugs.webkit.org/show_bug.cgi?id=81054
2770
2771         Caused existing test expectation lines to expand to one for
2772         each test configuration.
2773
2774         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
2775         (TestExpectations.has_warnings):
2776         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
2777         (test_macro_overrides):
2778         * Scripts/webkitpy/tool/commands/rebaseline.py:
2779         (RebaselineTest._save_baseline):
2780         (RebaselineTest._rebaseline_test):
2781         (RebaselineTest.execute):
2782         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
2783         (TestRebaseline.test_tests_to_update):
2784
2785 2012-03-13  Jon Lee  <jonlee@apple.com>
2786
2787         Separate NOTIFICATIONS and LEGACY_NOTIFICATIONS
2788         https://bugs.webkit.org/show_bug.cgi?id=80922
2789         <rdar://problem/11035082>
2790
2791         Reviewed by Jian Li.
2792
2793         You can include either NOTIFICATIONS or LEGACY_NOTIFICATIONS and have a complete API.
2794         LEGACY_NOTIFICATIONS should cover all of the previous functionality, and NOTIFICATIONS will cover the
2795         new API. Therefore, APIs that are common between the two will have:
2796         #if ENABLE(NOTIFICATIONS) || ENABLE(LEGACY_NOTIFICATIONS)
2797
2798         This patch initially sets both to be exactly the same. As other bugs with patches begin to migrate to
2799         the new API, the defines will begin to split. This allows ports to decide which set of APIs to include.
2800
2801         Update everything to be #if ENABLE(NOTIFICATIONS) || ENABLE(LEGACY_NOTIFICATIONS)
2802         * DumpRenderTree/chromium/LayoutTestController.cpp:
2803         (LayoutTestController::grantDesktopNotificationPermission):
2804         (LayoutTestController::simulateDesktopNotificationClick):
2805         * DumpRenderTree/chromium/NotificationPresenter.cpp:
2806         * DumpRenderTree/chromium/TestShell.cpp:
2807         (TestShell::TestShell):
2808         (TestShell::resetTestController):
2809         * DumpRenderTree/win/DRTDesktopNotificationPresenter.cpp:
2810         (DRTDesktopNotificationPresenter::checkNotificationPermission):
2811
2812 2012-03-13  Ojan Vafai  <ojan@chromium.org>
2813
2814         Have webkit-patch rebaseline-test update test_expectations.txt
2815         https://bugs.webkit.org/show_bug.cgi?id=81054
2816
2817         Reviewed by Dirk Pranke.
2818
2819         This will allow us to use the All Failures tab in garden-o-matic
2820         to rebaseline tests without making test_expectations.txt stale.
2821
2822         It's conservative and only removes lines if the test is
2823         not flaky and fails in a way the rebaseline would fix (e.g.
2824         CRASH/TIMEOUT will not be touched).
2825
2826         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
2827         (TestExpectations.remove_configuration_from_test):
2828         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
2829         (RemoveConfigurationsTest):
2830         (RemoveConfigurationsTest.test_remove):
2831         (test_remove_line):
2832         * Scripts/webkitpy/tool/commands/rebaseline.py:
2833         (RebaselineTest._is_supported_port):
2834         (RebaselineTest):
2835         (RebaselineTest._update_expectations_file):
2836         (RebaselineTest._rebaseline_test_and_update_expectations):
2837         (RebaselineTest.execute):
2838         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
2839         (TestRebaseline.test_rebaseline_updates_expectations_file):
2840
2841 2012-03-14  Jessie Berlin  <jberlin@apple.com>
2842
2843         Crash in WKTR under addChromeInputField due to using 0 as a key in a HashMap
2844         https://bugs.webkit.org/show_bug.cgi?id=81167
2845
2846         Reviewed by Sam Weinig.
2847
2848         Do not use 0 as a key in a HashMap.
2849
2850         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
2851         cacheLayoutTestControllerCallback uses the enum as keys for the callback map.
2852
2853 2012-03-14  Ojan Vafai  <ojan@chromium.org>
2854
2855         Add the chromium lion bot to garden-o-matic and remove
2856         the platform_move_to in builders.py since the Chromium Lion
2857         port is now a normal port.
2858
2859         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
2860         * Scripts/webkitpy/layout_tests/port/builders.py:
2861
2862 2012-03-14  Dirk Pranke  <dpranke@chromium.org>
2863
2864         Fix regression/crash introduced in r110724 - bad type passed from skipped_tests() by default.
2865
2866         Unreviewed, build fix.
2867
2868         * Scripts/webkitpy/layout_tests/port/base.py:
2869         (Port.skipped_tests):
2870
2871 2012-03-14  Dirk Pranke  <dpranke@chromium.org>
2872
2873         Please add a way to manually skip some tests in NRWT
2874         https://bugs.webkit.org/show_bug.cgi?id=81019
2875
2876         Reviewed by Ojan Vafai.
2877
2878         This patch implements the -i / --ignore-tests flag from ORWT
2879         and refactors the test_expectations.py code slightly to handle
2880         it.
2881
2882         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2883         (Manager.parse_expectations):
2884         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
2885         (TestExpectations.__init__):
2886         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
2887         (test_add_skipped_tests):
2888         (test_add_skipped_tests_duplicate):
2889         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2890         (parse_args):
2891         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2892         (MainTest.test_ignore_tests):
2893         (MainTest.test_ignore_tests.assert_ignored):
2894
2895 2012-03-14  Dirk Pranke  <dpranke@chromium.org>
2896
2897         nrwt: run with no args on lion, is putting baselines in platform/mac-lion, not platform/mac
2898         https://bugs.webkit.org/show_bug.cgi?id=81028
2899
2900         Reviewed by Ojan Vafai.
2901
2902         The mac port had incorrect lookups for 'mac-lion' and
2903         'mac-future' in some places. This patch fixes that and adds a
2904         bunch of tests.
2905
2906         * Scripts/webkitpy/layout_tests/port/mac.py:
2907         (MacPort.baseline_path):
2908         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
2909         (MacTest.assert_skipped_file_search_paths):
2910         (MacTest.test_skipped_file_search_paths):
2911         (_assert_search_path):
2912         (test_baseline_search_path):
2913         * Scripts/webkitpy/layout_tests/port/webkit.py:
2914         (WebKitPort._skipped_file_search_paths):
2915
2916 2012-03-13  Jer Noble  <jer.noble@apple.com>
2917
2918         Lion Intel Debug WebKit2 Tests crashing under [WKFullScreenWindowController _startEnterFullScreenAnimationWithDuration:]
2919         https://bugs.webkit.org/show_bug.cgi?id=81056
2920
2921         Reviewed by Jessie Berlin.
2922
2923         Intercept beganEnterFullScreen and beganExitFullScreen and turn them into no-ops (plus logging).
2924
2925         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
2926         (WTR::InjectedBundlePage::InjectedBundlePage):
2927         (WTR::InjectedBundlePage::beganEnterFullScreen):
2928         (WTR::InjectedBundlePage::beganExitFullScreen):
2929
2930 2012-03-14  Alexander Færøy  <alexander.faeroy@nokia.com>
2931
2932         The width and height of the scroll indicators in the MiniBrowser does
2933         not consider the offset margins.
2934         https://bugs.webkit.org/show_bug.cgi?id=81112
2935
2936         Reviewed by Kenneth Rohde Christiansen.
2937
2938         This patch fixes the issue by withdrawing two times the anchor
2939         margins.
2940
2941         * MiniBrowser/qt/qml/ScrollIndicator.qml:
2942
2943 2012-03-14  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
2944
2945         [EFL] Implement LayoutTestController::clearBackForwardList
2946         https://bugs.webkit.org/show_bug.cgi?id=80142
2947
2948         Adding missing implementation clearBackForwardList to EFL's
2949         LayoutTestController. This implementation allows us to unskip 
2950         tests connected with history.
2951
2952         Reviewed by Antonio Gomes.
2953
2954         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
2955         (LayoutTestController::clearBackForwardList): Implemented.
2956
2957 2012-03-13  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2958
2959         [EFL][DRT] Expose window.internals object
2960         https://bugs.webkit.org/show_bug.cgi?id=79900
2961
2962         Exposes window.internals for EFL's DRT. Makes possible to
2963         remove some tests from the skip list.
2964
2965         Reviewed by Antonio Gomes.
2966
2967         * DumpRenderTree/efl/CMakeLists.txt:
2968         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
2969         (DumpRenderTreeChrome::onWindowObjectCleared):
2970
2971 2012-03-13  Adam Barth  <abarth@webkit.org> && Benjamin Poulain  <bpoulain@apple.com>
2972
2973         Always enable ENABLE(CLIENT_BASED_GEOLOCATION)
2974         https://bugs.webkit.org/show_bug.cgi?id=78853
2975
2976         Reviewed by Adam Barth.
2977
2978         * Scripts/build-webkit:
2979         * qmake/mkspecs/features/features.prf:
2980
2981 2012-03-12  Antonio Gomes  <agomes@rim.com>
2982
2983         Convert nodesFromRect tests to use Internals interface
2984         https://bugs.webkit.org/show_bug.cgi?id=80886
2985
2986         Reviewed by Ryosuke Niwa.
2987
2988         This clean up allows to remove port specific bits of each
2989         DRT implementation that supports testing Document::nodesFromRect,
2990         in favor of making use of the 'Internals' interface to test it.
2991
2992         This makes its code conveniently more expansible and cleaner.
2993
2994         Also removed stubs from Win and EFL DRTs.
2995
2996         * DumpRenderTree/LayoutTestController.cpp:
2997         (LayoutTestController::staticFunctions):
2998         * DumpRenderTree/LayoutTestController.h:
2999         (LayoutTestController):
3000         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
3001         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
3002         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
3003         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
3004         * DumpRenderTree/qt/LayoutTestControllerQt.h:
3005         (LayoutTestController):
3006         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
3007
3008 2012-03-13  Ojan Vafai  <ojan@chromium.org>
3009
3010         Rebaselining for a new port doesn't work right with multiple fallback ports
3011         https://bugs.webkit.org/show_bug.cgi?id=80932
3012
3013         Reviewed by Adam Barth.
3014
3015         Since chromium-leopard also falls back to mac-leopard, we need to
3016         copy the existing result for both chromium-leopard and chromium-snowleopard
3017         before doing lion rebaselines.
3018
3019         * Scripts/webkitpy/layout_tests/port/builders.py:
3020         (fallback_port_names_for_new_port):
3021         * Scripts/webkitpy/tool/commands/rebaseline.py:
3022         (RebaselineTest):
3023         (RebaselineTest._copy_existing_baseline):
3024         (RebaselineTest._rebaseline_test):
3025         (RebaselineTest.execute):
3026         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
3027         * Scripts/webkitpy/tool/servers/gardeningserver.py:
3028         (GardeningHTTPRequestHandler.rebaseline):
3029         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
3030         (GardeningServerTest.test_rebaseline_new_port):
3031
3032 2012-03-13  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3033
3034         [Qt] Tweak minibrowser UI
3035
3036         Add some margins to the scroll indicators, and set a dark background
3037         color so you see the page dimensions.
3038
3039         Reviewed by Antonio Gomes.
3040
3041         * MiniBrowser/qt/qml/BrowserWindow.qml:
3042         * MiniBrowser/qt/qml/ScrollIndicator.qml:
3043
3044 2012-03-13  ChangSeok Oh  <shivamidow@gmail.com>
3045
3046         [EFL] [DRT] Implement scheduleAsynchronousKeyDown.
3047         https://bugs.webkit.org/show_bug.cgi?id=79601
3048
3049         Reviewed by Gustavo Noronha Silva.
3050
3051         Extracted shareable code from keyDownCallback and then added createKeyEventInfo.
3052         createKeyEventInfo creates a new structure KeyEventInfo that has keyName and modifiers
3053         and is taken over to sendKeyDown.
3054         Added sendKeyDown and sendAsynchronousKeyDown to evaluate keyDown event directly
3055         and asynchronously. Eventually, sendAsynchronousKeyDown calls sendKeyDown.
3056
3057         * DumpRenderTree/efl/EventSender.cpp:
3058         (KeyEventInfo::KeyEventInfo):
3059         (KeyEventInfo):
3060         (createKeyEventInfo):
3061         (sendKeyDown):
3062         (keyDownCallback):
3063         (sendAsynchronousKeyDown):
3064         (scheduleAsynchronousKeyDownCallback):
3065
3066 2012-03-12  Ojan Vafai  <ojan@chromium.org>
3067
3068         webkit-patch optimize-baselines sometimes creates a mac-future result
3069         https://bugs.webkit.org/show_bug.cgi?id=80717
3070
3071         Reviewed by Adam Barth.
3072
3073         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
3074         (_baseline_search_hypergraph):
3075         (BaselineOptimizer._find_optimal_result_placement):
3076         (BaselineOptimizer._filter_virtual_ports):
3077         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
3078         (BaselineOptimizerTest.test_no_add_mac_future):
3079
3080 2012-03-12  Jessie Berlin  <jberlin@apple.com>
3081
3082         More Windows build fixing.
3083
3084         * DumpRenderTree/win/WorkQueueItemWin.cpp:
3085
3086 2012-03-12  Jessie Berlin  <jberlin@apple.com>
3087
3088         More Windows build fixing.
3089
3090         * DumpRenderTree/win/AccessibilityControllerWin.cpp:
3091         * DumpRenderTree/win/EditingDelegate.cpp:
3092         * DumpRenderTree/win/EventSender.cpp:
3093         * DumpRenderTree/win/FrameLoadDelegate.cpp:
3094         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
3095         * DumpRenderTree/win/UIDelegate.cpp:
3096
3097 2012-03-12  Dirk Pranke  <dpranke@chromium.org>
3098
3099         remove the 'gpu' group from the flakiness dashboard
3100         https://bugs.webkit.org/show_bug.cgi?id=80353
3101
3102         Reviewed by Adam Barth.
3103
3104         Since those bots are obsolete, we don't need this any more.
3105
3106         * TestResultServer/static-dashboards/builders.js:
3107         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
3108         (testPlatformAndBuildType):
3109         (testHtmlForTestTypeSwitcherGroup):
3110
3111 2012-03-12  Dirk Pranke  <dpranke@chromium.org>
3112
3113         remove graphics_type from webkitpy.layout_tests.models.test_configuration
3114         https://bugs.webkit.org/show_bug.cgi?id=80643
3115
3116         Reviewed by Ojan Vafai.
3117
3118         Now that the chromium-gpu ports are gone, there's no need for
3119         the graphics_type CPU/GPU distinction. Most of this change is
3120         just updating tests accordingly.
3121
3122         * Scripts/webkitpy/layout_tests/controllers/test_expectations_editor_unittest.py:
3123         (TestExpectationEditorTests):
3124         * Scripts/webkitpy/layout_tests/models/test_configuration.py:
3125         (TestConfiguration.__init__):
3126         (TestConfiguration.category_order):
3127         (TestConfiguration.__str__):
3128         (TestConfiguration.__repr__):
3129         (TestConfiguration.__hash__):
3130         (TestConfigurationConverter.to_specifiers_list):
3131         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py:
3132         (make_mock_all_test_configurations_set):
3133         (TestConfigurationTest.test_items):
3134         (TestConfigurationTest.test_keys):
3135         (TestConfigurationTest.test_str):
3136         (TestConfigurationTest.test_repr):
3137         (TestConfigurationTest.test_hash):
3138         (TestConfigurationTest.test_hash.query_unknown_key):
3139         (TestConfigurationTest.test_eq):
3140         (TestConfigurationTest.test_values):
3141         (SpecifierSorterTest.test_specifier_priority):
3142         (SpecifierSorterTest.test_sort_specifiers):
3143         (TestConfigurationConverterTest.test_to_config_set):
3144         (TestConfigurationConverterTest.test_macro_expansion):
3145         (TestConfigurationConverterTest.test_to_specifier_lists):
3146         (TestConfigurationConverterTest.test_converter_macro_collapsing):
3147         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
3148         (test_error_on_different_build_type):
3149         (TestExpectationSerializerTests.test_parsed_to_string):
3150         (TestExpectationSerializerTests.test_reconstitute_only_these):
3151         * Scripts/webkitpy/layout_tests/port/apple.py:
3152         (ApplePort._generate_all_test_configurations):
3153         * Scripts/webkitpy/layout_tests/port/base.py:
3154         (Port.__init__):
3155         (Port.name):
3156         (Port.version):
3157         (Port.test_configuration):
3158         * Scripts/webkitpy/layout_tests/port/builders.py:
3159         * Scripts/webkitpy/layout_tests/port/chromium.py:
3160         (ChromiumPort):
3161         (ChromiumPort._generate_all_test_configurations):
3162         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
3163         (ChromiumLinuxPort.__init__):
3164         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
3165         (ChromiumLinuxPortTest.test_check_illegal_port_names):
3166         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
3167         (ChromiumMacPortTest.test_baseline_path):
3168         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
3169         (ChromiumPortTest.test_all_test_configurations):
3170         * Scripts/webkitpy/layout_tests/port/efl.py:
3171         (EflPort._generate_all_test_configurations):
3172         * Scripts/webkitpy/layout_tests/port/gtk.py:
3173         (GtkPort._generate_all_test_configurations):
3174         * Scripts/webkitpy/layout_tests/port/qt.py:
3175         (QtPort._generate_all_test_configurations):
3176         * Scripts/webkitpy/layout_tests/port/test.py:
3177         (TestPort.all_test_configurations):
3178         (TestPort._all_build_types):
3179         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
3180         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
3181         (BuildCoverageExtrapolatorTest.test_extrapolate):
3182         (GardeningExpectationsUpdaterTest.test_unknown_failure_type):
3183         (GardeningExpectationsUpdaterTest.test_add_new_expectation):
3184         (GardeningExpectationsUpdaterTest.test_replace_old_expectation):
3185         (GardeningExpectationsUpdaterTest.test_pass_expectation):
3186         (GardeningExpectationsUpdaterTest.test_supplement_old_expectation):
3187         (GardeningExpectationsUpdaterTest.test_spurious_updates):
3188
3189 2012-03-12  Dirk Pranke  <dpranke@chromium.org>
3190
3191         _runtime_feature_list isn't supported on Apple Mac DRT
3192         https://bugs.webkit.org/show_bug.cgi?id=80906
3193
3194         Reviewed by Eric Seidel.
3195
3196         It looks like this feature is actually only supported on the Win
3197         DRT port, so this patch reworks the code to make that clearer
3198         and reduce unnecessary overrides.
3199
3200         * Scripts/webkitpy/layout_tests/port/efl.py:
3201         (EflPort._path_to_webcore_library):
3202         * Scripts/webkitpy/layout_tests/port/gtk.py:
3203         (GtkPort._path_to_webcore_library):
3204         * Scripts/webkitpy/layout_tests/port/qt.py:
3205         (QtPort._skipped_file_search_paths):
3206         * Scripts/webkitpy/layout_tests/port/webkit.py:
3207         (WebKitPort._runtime_feature_list):
3208         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
3209         (TestWebKitPort.__init__):
3210         (TestWebKitPort.all_test_configurations):
3211         (test_skipped_directories_for_features):
3212         * Scripts/webkitpy/layout_tests/port/win.py:
3213         (WinPort):
3214         (WinPort._runtime_feature_list):
3215         * Scripts/webkitpy/layout_tests/port/win_unittest.py:
3216         (WinPortTest.test_operating_system):
3217         (WinPortTest):
3218         (WinPortTest.test_runtime_feature_list):
3219
3220 2012-03-12  Peter Beverloo  <peter@chromium.org>
3221
3222         [Chromium] Temporary build-fix for the Android bot
3223         https://bugs.webkit.org/show_bug.cgi?id=80862
3224
3225         Reviewed by Adam Barth.
3226
3227         Manually set the environmental variables required to compile source
3228         files through the Android NDK instead of relying on the envsetup.sh
3229         script. A proper fix will be landed to Chromium ASAP, and following
3230         a Chromium roll into WebKit this code will be removed.
3231
3232         * Scripts/webkitdirs.pm:
3233         (buildChromiumMakefile):
3234
3235 2012-03-12  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3236
3237         [Qt] Use right construct to detect config_test_gccdepends
3238
3239         Reviewed by Andreas Kling.
3240
3241         * qmake/mkspecs/features/default_pre.prf:
3242
3243 2012-03-12  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3244
3245         [Qt] Expose location of WebKit's bin and imports dir through module file
3246
3247         Reviewed by Andreas Kling.
3248
3249         * qmake/qt_webkit.pri:
3250
3251 2012-03-12  Andras Becsi  <andras.becsi@nokia.com>
3252
3253         [Qt][WK2] Add support for rudimentary scroll indicators in MiniBrowser
3254         https://bugs.webkit.org/show_bug.cgi?id=80832
3255
3256         Reviewed by Tor Arne Vestbø.
3257
3258         Add a basic scroll indicator component to MiniBrowser so that it
3259         behaves similar to the ScrollDecorator QML component.
3260
3261         * MiniBrowser/qt/MiniBrowser.qrc:
3262         * MiniBrowser/qt/qml/BrowserWindow.qml:
3263         * MiniBrowser/qt/qml/ScrollIndicator.qml: Added.
3264
3265 2012-03-12  C Anthony Risinger  <anthony@xtfx.me>
3266
3267         [GTK] r110296 included an extra $ in @$(AM_V_GEN)
3268         https://bugs.webkit.org/show_bug.cgi?id=80809
3269
3270         Reviewed by Philippe Normand.
3271
3272         * GNUmakefile.am: Add `$` to @(AM_V_GEN), else /bin/sh syntax error
3273
3274 2012-03-12  Ming Xie  <mxie@rim.com>
3275
3276         [BlackBerry] build-webkit doesn't recognize "--cmakearg" for the BlackBerry port
3277         https://bugs.webkit.org/show_bug.cgi?id=80812
3278
3279         Reviewed by Rob Buis.
3280
3281         * Scripts/build-webkit: Add $cmakeArgs to buildCMakeProjectOrExit()
3282         * Scripts/webkitdirs.pm:
3283         (blackberryCMakeArguments): Remove "PUBLIC_BUILD=0" define
3284
3285
3286 2012-03-12  Simon Hausmann  <simon.hausmann@nokia.com>
3287
3288         [Qt] Fix build with newer Qt 5.
3289
3290         Reviewed by Tor Arne Vestbø.
3291
3292         Newer Qt 5 renames QDeclarativeFoo to QQmlFoo and provides backwards compatible
3293         renames, which however do trigger a #warning. Until we _require_ a newer Qt 5
3294         we should disable -Werror.
3295
3296         Also load(moc) for the availability of QMAKE_MOC in various derived sources. The
3297         implicit load was removed recently.
3298
3299         * qmake/mkspecs/features/default_pre.prf:
3300         * qmake/mkspecs/features/unix/default_post.prf:
3301
3302 2012-03-11  Maciej Stachowiak  <mjs@apple.com>
3303
3304         Add --instruments option to sunspider to profile with Instruments
3305         https://bugs.webkit.org/show_bug.cgi?id=80783
3306
3307         Reviewed by Filip Pizlo.
3308
3309         * Scripts/run-sunspider:
3310
3311 2012-03-11  Viatcheslav Ostapenko  <ostapenko.viatcheslav@nokia.com>
3312
3313         [Qt] [WK2] Support threaded renderer in WK2
3314         https://bugs.webkit.org/show_bug.cgi?id=76661
3315
3316         Remove disabling of threaded rendering.
3317
3318         Reviewed by Noam Rosenthal.
3319
3320         * MiniBrowser/qt/main.cpp:
3321         (main):
3322
3323 2012-03-09  Zan Dobersek  <zandobersek@gmail.com>
3324
3325         [WK2] run-perf-tests should be able to run with WTR
3326         https://bugs.webkit.org/show_bug.cgi?id=80042
3327
3328         Reviewed by Martin Robinson.
3329
3330         Add an option to PerfTestRunner to use WebKitTestRunner instead of
3331         DumpRenderTree and add the '--no-timeout' flag to WebKitTestRunner
3332         to avoid timing out when running performance tests.
3333
3334         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
3335         (PerfTestsRunner._parse_args):
3336         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
3337         (WTR::InjectedBundle::InjectedBundle): Initialize m_useWaitToDumpWatchdogTimer
3338         to true.
3339         (WTR::InjectedBundle::didReceiveMessage): Unpack the 'BeginTest' message
3340         as a dictionary containing DumpPixels and UseWaitToDumpWatchdogTimer keys.
3341         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
3342         (WTR::InjectedBundle::useWaitToDumpWatchdogTimer): Make m_useWaitToDumpWatchdogTimer
3343         publicly accessible.
3344         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
3345         (WTR::LayoutTestController::waitUntilDone): Only initialize the watchdog if it
3346         is required.
3347         * WebKitTestRunner/TestController.cpp:
3348         (WTR::TestController::TestController): Initialize m_useWaitToDumpWatchdogTimer
3349         to true.
3350         (WTR::TestController::initialize): Set m_useWaitToDumpWatchdogTimer to false
3351         if '--no-timeout' option is present.
3352         (WTR::TestController::runUntil): Translate TimeoutDuration to a proper timeout
3353         value, or -1.0 if timeout should not occur.
3354         * WebKitTestRunner/TestController.h:
3355         (WTR::TestController::useWaitToDumpWatchdogTimer): Make m_useWaitToDumpWatchdogTimer
3356         publicly accessible.
3357         * WebKitTestRunner/TestInvocation.cpp:
3358         (WTR::TestInvocation::invoke): Pack the 'BeginTest' message as a dictionary,
3359         containing DumpPixels and UseWaitToDumpWatchdogTimer keys and corresponding values.
3360         Waiting for final message should not time out if not necessary.
3361         * WebKitTestRunner/gtk/TestControllerGtk.cpp:
3362         (WTR::TestController::platformRunUntil): Only set up a timeout if necessary.
3363         * WebKitTestRunner/qt/TestControllerQt.cpp:
3364         (WTR::TestController::platformRunUntil): Ditto.
3365         * WebKitTestRunner/mac/TestControllerMac.mm:
3366         (WTR::TestController::platformRunUntil): Add a FIXME, indicating functionality
3367         is missing.
3368         * WebKitTestRunner/win/TestControllerWin.cpp:
3369         (WTR::TestController::platformRunUntil): Ditto.
3370
3371 2012-03-09  Jon Lee  <jonlee@apple.com>
3372
3373         Rename NotificationPresenter to NotificationClient
3374         https://bugs.webkit.org/show_bug.cgi?id=80488
3375         <rdar://problem/10965558>
3376
3377         Reviewed by Kentaro Hara.
3378
3379         Refactor to use renamed WebCore::NotificationClient.
3380         * DumpRenderTree/win/DRTDesktopNotificationPresenter.cpp:
3381         (DRTDesktopNotificationPresenter::checkNotificationPermission):
3382
3383 2012-03-07  James Robinson  <jamesr@chromium.org>
3384
3385         [chromium] Deprecate renderDirectlyToWebView parameter of WebViewClient::createGraphicsContext3D()
3386         https://bugs.webkit.org/show_bug.cgi?id=80523
3387
3388         Reviewed by Darin Fisher.
3389
3390         * DumpRenderTree/chromium/WebViewHost.cpp:
3391         (WebViewHost::createGraphicsContext3D):
3392         * DumpRenderTree/chromium/WebViewHost.h:
3393         (WebViewHost):
3394
3395 2012-03-09  Eric Seidel  <eric@webkit.org>
3396
3397         build-webkit should build Source/WTF on Mac
3398         https://bugs.webkit.org/show_bug.cgi?id=80728
3399
3400         Reviewed by Adam Barth.
3401
3402         We'll add other ports to this list in future commits.
3403
3404         * Scripts/build-webkit:
3405
3406 2012-03-09  Jon Lee  <jonlee@apple.com>
3407
3408         Add support for ENABLE(LEGACY_NOTIFICATIONS)
3409         https://bugs.webkit.org/show_bug.cgi?id=80497
3410
3411         Reviewed by Adam Barth.
3412
3413         Prep for b80472: Update API for Web Notifications
3414         * Scripts/build-webkit:
3415         * qmake/mkspecs/features/features.prf:
3416
3417 2012-03-09  Ashod Nakashian  <ashodnakashian@yahoo.com>
3418
3419         Bash scripts should support LF endings only
3420         https://bugs.webkit.org/show_bug.cgi?id=79509
3421
3422         Reviewed by David Kilzer.
3423
3424         * EWSTools/boot.sh: Added properties svn:executable and svn:eol-style.
3425         * EWSTools/start-queue.sh: Added property svn:eol-style.
3426         * Scripts/webkit-tools-completion.sh: Added properties svn:executable and svn:eol-style.
3427         * WebKitTestRunner/win/build-generated-files.sh: Added properties svn:executable and svn:eol-style.
3428         * iExploder/iexploder-1.7.2/tools/release_src.sh: Added property svn:eol-style.
3429         * iExploder/iexploder-1.7.2/tools/update_html_tags_from_sources.sh: Added property svn:eol-style.
3430
3431 2012-03-09  Ionut Biru  <ibiru@archlinux.org>
3432
3433         [BUILD] use detected python to run generate-gtkdoc
3434         https://bugs.webkit.org/show_bug.cgi?id=80695
3435
3436         Reviewed by Gustavo Noronha Silva.
3437
3438         * GNUmakefile.am: Use $(PYTHON) instead of plain `python`
3439
3440 2012-03-09  Nat Duca  <nduca@chromium.org>
3441
3442         Make check-webkit-style complain about "no new tests" lines in changelogs
3443         https://bugs.webkit.org/show_bug.cgi?id=80582
3444
3445         Reviewed by David Levin.
3446
3447         * Scripts/webkitpy/style/checkers/changelog.py:
3448         (ChangeLogChecker.check_entry):
3449         * Scripts/webkitpy/style/checkers/changelog_unittest.py:
3450         (ChangeLogCheckerTest.test_no_new_tests):
3451
3452 2012-03-09  Sheriff Bot  <webkit.review.bot@gmail.com>
3453
3454         Unreviewed, rolling out r110250.
3455         http://trac.webkit.org/changeset/110250
3456         https://bugs.webkit.org/show_bug.cgi?id=80682
3457
3458         It made NRWT ignore test_expectations.txt on Qt-WK2 and ignore
3459         qt-linux at all. (Requested by Ossy on #webkit).
3460
3461         * Scripts/old-run-webkit-tests:
3462         * Scripts/webkitpy/layout_tests/port/qt.py:
3463         (QtPort):
3464         (QtPort.determine_full_port_name):
3465         (QtPort.baseline_search_path):
3466         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
3467         (QtPortTest):
3468         (QtPortTest._assert_search_path):
3469         (QtPortTest.test_baseline_search_path):
3470
3471 2012-03-08  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3472
3473         [Qt] Use Qt's module system for install rules and depending on QtWebKit
3474
3475         Instead of rolling our own install rules we now use the same approach as
3476         every other Qt module, by loading qt_module.prf and qt_module_config.prf.
3477
3478         This ensures that we follow the same semantics as the rest of Qt on
3479         what sort of config options are enabled by default (create_cmake eg.).
3480         It also allows us to use QT += webkit instead of the workaround we had
3481         with CONFIG += qtwebkit.
3482
3483         We do however force Qt to always treat our build as a non-developer build,
3484         so the libraries will end up in the WebKit lib directory instead of the
3485         qtbase directory (as with a normal developer-build). This allows us to
3486         keep the webkit-build self-contained. If Qt is a developer build we still
3487         copy the module file manually to Qt, so that you don't have to install
3488         WebKit to make it available.
3489
3490         For non-developer builds of Qt, it is still possible to use the built
3491         WebKit libraries without having to install them, by having the variable
3492         QMAKE_EXTRA_MODULE_FORWARDS set in the project's .qmake.cache file,
3493         pointing to $WEBKITOUTUTDIR/$CONFIGURATION/modules.
3494
3495         https://bugs.webkit.org/show_bug.cgi?id=80590
3496
3497         Reviewed by Simon Hausmann.
3498
3499         * DumpRenderTree/qt/DumpRenderTree.pro:
3500         * MiniBrowser/qt/BrowserWindow.cpp:
3501         * MiniBrowser/qt/MiniBrowser.pro:
3502         * MiniBrowser/qt/MiniBrowserApplication.cpp:
3503         * MiniBrowser/qt/UrlLoader.cpp:
3504         * QtTestBrowser/QtTestBrowser.pro:
3505         * Tools.pro:
3506         * WebKitTestRunner/InjectedBundle/Target.pri:
3507         * WebKitTestRunner/Target.pri:
3508         * qmake/mkspecs/features/default_post.prf:
3509         * qmake/mkspecs/features/default_pre.prf:
3510         * qmake/mkspecs/features/functions.prf:
3511         * qmake/mkspecs/features/qtwebkit-private.prf: Removed.
3512         * qmake/mkspecs/features/qtwebkit.prf:
3513         * qmake/mkspecs/features/rpath.prf:
3514         * qmake/mkspecs/modules/qt_webkit.pri: Removed.
3515         * qmake/qt_webkit.pri: Added.
3516
3517 2012-03-08  Igor Oliveira  <igor.o@sisa.samsung.com>
3518
3519         Unreviewed. Update my email.
3520
3521         * Scripts/webkitpy/common/config/committers.py:
3522
3523 2012-03-08  Xianzhu Wang  <wangxianzhu@chromium.org>
3524
3525         Recent changes to chromium.py etc. broke chromium_android.py
3526         https://bugs.webkit.org/show_bug.cgi?id=80623
3527
3528         Reviewed by Dirk Pranke.
3529
3530         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
3531         (ChromiumAndroidDriver.__init__):
3532         (ChromiumAndroidDriver._start):
3533
3534 2012-03-08  Xianzhu Wang  <wangxianzhu@chromium.org>
3535
3536         [Chromium] Exception running reftest with --no-pixel-tests
3537         https://bugs.webkit.org/show_bug.cgi?id=80642
3538
3539         Reviewed by Dirk Pranke.
3540
3541         * Scripts/webkitpy/layout_tests/port/chromium.py:
3542         (ChromiumDriver.__init__):
3543         (ChromiumDriver._wrapper_options):
3544
3545 2012-03-08  Rafael Brandao  <rafael.lobo@openbossa.org>
3546
3547         [NRWT] Fix --platform=qt-5.0 --new-baseline combo
3548         https://bugs.webkit.org/show_bug.cgi?id=72489
3549
3550         Reviewed by Dirk Pranke.
3551
3552         NRWT: The default platform name for Qt combined with --webkit-test-runner is now
3553         "qt-5.0-wk2" instead of "qt-linux" and that name is now prefered for tests.
3554         --platform will override the platform name and then we'll use it.
3555
3556         ORWT: We check for qt version and use platform name "qt-4.8" if the version is lower
3557         than 5.0.0, and then we choose between "qt-5.0-wk1" and "qt-5.0-wk2". --platform will
3558         override it in the end.
3559
3560         * Scripts/old-run-webkit-tests:
3561         * Scripts/webkitpy/layout_tests/port/qt.py:
3562         (QtPort):
3563         (QtPort.determine_full_port_name):
3564         (QtPort.baseline_search_path):
3565         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
3566         (QtPortTest):
3567         (QtPortTest._assert_search_path):
3568         (QtPortTest.test_baseline_search_path):
3569
3570 2012-03-08  Beth Dakin  <bdakin@apple.com>
3571
3572         https://bugs.webkit.org/show_bug.cgi?id=80463
3573         RenderImage is using the wrong origin when calling addRelevantRepaintedObject
3574         -and corresponding-
3575         <rdar://problem/10970221>
3576
3577         Reviewed by Dan Bernstein.
3578
3579         New test!
3580         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
3581         * TestWebKitAPI/Tests/WebKit2/NewFirstVisuallyNonEmptyLayoutForImages.cpp: Added.
3582         (TestWebKitAPI):
3583         (TestWebKitAPI::didNewFirstVisuallyNonEmptyLayout):
3584         (TestWebKitAPI::setPageLoaderClient):
3585         (TestWebKitAPI::TEST):
3586         * TestWebKitAPI/Tests/WebKit2/NewFirstVisuallyNonEmptyLayoutForImages_Bundle.cpp: Added.
3587         (TestWebKitAPI):
3588         (NewFirstVisuallyNonEmptyLayoutForImagesTest):
3589         (TestWebKitAPI::NewFirstVisuallyNonEmptyLayoutForImagesTest::NewFirstVisuallyNonEmptyLayoutForImagesTest):
3590         (TestWebKitAPI::NewFirstVisuallyNonEmptyLayoutForImagesTest::didCreatePage):
3591         * TestWebKitAPI/Tests/WebKit2/lots-of-images.html: Added.
3592
3593 2012-03-08  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3594
3595         [Qt] Prospective Windows build fix
3596
3597         Don't assume that windows-builds will always run inside a cmd.exe shell.
3598
3599         * qmake/mkspecs/features/default_pre.prf:
3600
3601 2012-03-08  Dan Bernstein  <mitz@apple.com>
3602
3603         <rdar://problem/10981173> Dashboard regions should not be in device space
3604
3605         Reviewed by John Sullivan.
3606
3607         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
3608         * TestWebKitAPI/Tests/mac/DeviceScaleFactorInDashboardRegions.mm: Added.
3609         (-[DeviceScaleFactorInDashboardRegionsUIDelegate webView:dashboardRegionsChanged:]):
3610         (TestWebKitAPI):
3611         (TestWebKitAPI::TEST):
3612
3613 2012-03-08  Dirk Pranke  <dpranke@chromium.org>
3614
3615         webkit-patch works oddly on local Git branches
3616         https://bugs.webkit.org/show_bug.cgi?id=76958
3617
3618         Reviewed by Ojan Vafai.
3619
3620         This change makes the '-g' flag to webkit-patch behave a
3621         little more consistently with git. The 'HEAD..' syntax
3622         is changed to 'HEAD....' (four dots), so as to not
3623         collide with the underlying git syntax. We also generalize
3624         the syntax to support any refname, so 'foo....' means
3625         diff against foo and include working copy changes.
3626
3627         This change also adds the 'UPSTREAM' shortcut to identify
3628         whatever the current upstream/tracking branch is; this is
3629         useful if you have multiple local branches stacked up.
3630
3631         * Scripts/webkitpy/common/checkout/scm/git.py:
3632         (Git._branch_from_ref):
3633         (Git):
3634         (Git._current_branch):
3635         (Git._upstream_branch):
3636         (Git.merge_base):
3637         (Git._commit_on_branch):
3638         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
3639         (GitSVNTest._second_local_commit):
3640         (GitSVNTest._two_local_commits):
3641         (GitSVNTest.test_commit_with_message_git_commit_range):
3642         (GitSVNTest.test_upstream_branch):
3643         (GitSVNTest.test_create_patch_working_copy_only):
3644         (GitSVNTest.test_changed_files_local_plus_working_copy):
3645         (GitSVNTest.test_changed_files_working_copy_only):
3646         (GitSVNTest.test_changed_files_upstream):
3647         * Scripts/webkitpy/tool/steps/options.py:
3648         (Options):
3649
3650 2012-03-08  John Sullivan  <sullivan@apple.com>
3651
3652         Made extract-localizable-strings ignore NSLocalizedFileSizeDescription
3653
3654         Reviewed by Eric Carlson.
3655
3656         * Scripts/extract-localizable-strings:
3657         Added NSLocalizedFileSizeDescription to allowed list of tokens that start
3658         with NSLocalized. Also improved an error message to include the troublesome token.
3659
3660 2012-03-08  Sheriff Bot  <webkit.review.bot@gmail.com>
3661
3662         Unreviewed, rolling out r110115.
3663         http://trac.webkit.org/changeset/110115
3664         https://bugs.webkit.org/show_bug.cgi?id=80581
3665