nrwt: implement the actual cascade of TestExpectations
[WebKit-https.git] / Tools / ChangeLog
index 21431dd..d5557c2 100644 (file)
+2012-06-13  Dirk Pranke  <dpranke@chromium.org>
+
+        nrwt: implement the actual cascade of TestExpectations
+        https://bugs.webkit.org/show_bug.cgi?id=88947
+
+        Reviewed by Ojan Vafai.
+
+        This change implements the actual cascade by removing the
+        concept of 'overrides' from the TestExpectations object and
+        parsing each file separately.
+
+        There is an actual semantic change in this patch, in that
+        setting an expectation on a directory in one file will override
+        the expectations on any individual tests set in prior files. The
+        test_overrides__directory() unit test verifies this.
+
+        Otherwise, this patch mostly consists of deleting code :).
+
+        * Scripts/webkitpy/layout_tests/models/test_expectations.py:
+        (TestExpectationsModel.__init__):
+        (TestExpectationsModel.add_expectation_line):
+        (TestExpectationsModel._add_test):
+        (TestExpectationsModel._already_seen_better_match):
+        (TestExpectations.__init__):
+        (TestExpectations._add_expectations):
+        * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
+        (test_overrides__directory):
+
+2012-06-13  Dirk Pranke  <dpranke@chromium.org>
+
+        webkitpy: update callers to use port.expectation_dict() instead of test_expectations() and test_expectations_overrides()
+        https://bugs.webkit.org/show_bug.cgi?id=88946
+
+        Reviewed by Ojan Vafai.
+
+        In preparation for fully supporting cascading expectations files
+        and removing the 'overrides' concept.
+
+        There should be no functional changes in this patch.
+
+        * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
+        (ManagerTest.test_update_summary_with_result):
+        (ResultSummaryTest.get_result_summary):
+        * Scripts/webkitpy/layout_tests/models/test_expectations.py:
+        (TestExpectations.__init__):
+        * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
+        (parse_exp):
+        (SkippedTests.check):
+        (RemoveConfigurationsTest.test_remove):
+        (test_remove_line):
+        * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
+        (LintTest.test_all_configurations.FakePort.__init__):
+        (LintTest.test_all_configurations.FakePort.expectations_dict):
+        (LintTest.test_lint_test_files__errors):
+        * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
+        (test_overrides_are_included_correctly):
+
+2012-06-13  Dirk Pranke  <dpranke@chromium.org>
+
+        webkitpy: rework the TestExpectations style checker in preparation for the cascade
+        https://bugs.webkit.org/show_bug.cgi?id=88945
+
+        Reviewed by Ojan Vafai.
+
+        This patch changes the style checker to call the
+        TestExpectations parser directly and be oblivious as to what the
+        port's actual expectations are.
+
+        * Scripts/webkitpy/style/checkers/test_expectations.py:
+        (TestExpectationsChecker.__init__):
+        (TestExpectationsChecker.check_test_expectations):
+
+2012-06-13  Dirk Pranke  <dpranke@chromium.org>
+
+        nrwt: restructure the port classes to handle multiple expectations files
+        https://bugs.webkit.org/show_bug.cgi?id=88944
+
+        Reviewed by Ojan Vafai.
+
+        This patch restructures the Port classes so that each port
+        returns a list of TestExpectations files to be used (in a new, public,
+        expectations_files() method), and the base implementation rolls them
+        up into the ordered expectations_dict(), which becomes a public
+        method.
+
+        port.path_to_test_expectations_file(), port.test_expectations(),
+        and port.test_expectations_overrides() are then reimplemented as
+        wrappers around port.expectations_dict().
+        port.test_expectations() and port.test_expectations_overrides()
+        will eventually be removed when the cascade is fully completed,
+        and port.path_to_test_expectations_file() will be removed when
+        the rebaselining tool supports the cascade
+        (expectations_files() may also become private at that point).
+
+        There should be no functional changes in this patch, but a bunch
+        of custom logic gets deleted!
+
+        * Scripts/webkitpy/layout_tests/port/base.py:
+        (Port.expectations_dict):
+        (Port):
+        (Port.expectations_files):
+        (Port.test_expectations):
+        (Port.test_expectations_overrides):
+        * Scripts/webkitpy/layout_tests/port/base_unittest.py:
+        (PortTest.test_additional_expectations):
+        (PortTest.test_uses_test_expectations_file):
+        * Scripts/webkitpy/layout_tests/port/chromium.py:
+        (ChromiumPort.all_baseline_variants):
+        (ChromiumPort.expectations_files):
+        * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
+        (ChromiumPortTest.test_overrides_and_builder_names):
+        * Scripts/webkitpy/layout_tests/port/google_chrome.py:
+        (_expectations_files):
+        (GoogleChromeLinux32Port.expectations_files):
+        (GoogleChromeLinux64Port.expectations_files):
+        (GoogleChromeMacPort.expectations_files):
+        (GoogleChromeWinPort.expectations_files):
+        * Scripts/webkitpy/layout_tests/port/google_chrome_unittest.py:
+        (TestGoogleChromePort._verify_expectations_overrides):
+        * Scripts/webkitpy/layout_tests/port/port_testcase.py:
+        (PortTestCase.test_expectations_ordering):
+        * Scripts/webkitpy/layout_tests/port/webkit.py:
+        (WebKitPort._skipped_file_search_paths):
+        * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
+        (TestRebaseline.test_rebaseline_updates_expectations_file_noop):
+        (test_rebaseline_updates_expectations_file):
+        (test_rebaseline_does_not_include_overrides):
+        (test_rebaseline_expectations):
+        (test_overrides_are_included_correctly):
+
+2012-06-13  Brent Fulgham  <bfulgham@webkit.org>
+
+        [WinCairo] Update URL of WinCairoRequirements.zip download.
+        https://bugs.webkit.org/show_bug.cgi?id=89009
+
+        Reviewed by Martin Robinson.
+
+        * Tools/Scripts/update-webkit-dependency: Put fallback code
+        to download a "last update" time from servers that don't properly
+        report Last-Modified in their HTTP header.
+        * Tools/Scripts/update-webkit-wincairo-libs: Change download link
+        for the WinCairo build requirements bundle.
+
+2012-06-13  Dirk Pranke  <dpranke@chromium.org>
+
+        nrwt: preliminary cleanup before supporting cascading expectations files
+        https://bugs.webkit.org/show_bug.cgi?id=88942
+
+        Reviewed by Ojan Vafai.
+
+        This change just prepares the TestExpectations parser to get filenames
+        along with the expectations, and improves the warning messages so that
+        they contain the filenames along with the line numbers.
+
+        There should be no functional changes in this patch.
+
+        * Scripts/webkitpy/layout_tests/controllers/test_expectations_editor_unittest.py:
+        (TestExpectationEditorTests.make_parsed_expectation_lines):
+        * Scripts/webkitpy/layout_tests/models/test_expectations.py:
+        (TestExpectationParser.parse):
+        (TestExpectationParser.expectation_for_skipped_test):
+        (TestExpectationParser._tokenize):
+        (TestExpectationParser._tokenize_list):
+        (TestExpectationLine):
+        (TestExpectationsModel.__init__):
+        (TestExpectationsModel._already_seen_better_match):
+        (TestExpectations.__init__):
+        (TestExpectations._shorten_filename):
+        (TestExpectations._report_warnings):
+        (TestExpectations._add_skipped_tests):
+        * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
+        (TestExpectationParserTests._tokenize):
+        (TestExpectationParserTests.test_tokenize_blank):
+        (TestExpectationParserTests.test_tokenize_missing_colon):
+        (TestExpectationParserTests.test_tokenize_extra_colon):
+        (TestExpectationParserTests.test_tokenize_empty_comment):
+        (TestExpectationParserTests.test_tokenize_comment):
+        (TestExpectationParserTests.test_tokenize_missing_equal):
+        (TestExpectationParserTests.test_tokenize_extra_equal):
+        (TestExpectationParserTests.test_tokenize_valid):
+        (TestExpectationParserTests.test_tokenize_valid_with_comment):
+        (TestExpectationParserTests.test_tokenize_valid_with_multiple_modifiers):
+        (TestExpectationParserTests.test_parse_empty_string):
+        (TestExpectationSerializerTests._tokenize):
+        (TestExpectationSerializerTests.assert_round_trip):
+        (TestExpectationSerializerTests.assert_list_round_trip):
+        * Scripts/webkitpy/tool/servers/gardeningserver.py:
+        (GardeningExpectationsUpdater.update_expectations):
+
+2012-06-13  Zan Dobersek  <zandobersek@gmail.com>
+
+        [Gtk] Enable link prefetch support in the developer builds
+        https://bugs.webkit.org/show_bug.cgi?id=89011
+
+        Reviewed by Martin Robinson.
+
+        Enable the link prefetch support for Gtk by default in developer
+        builds.
+
+        * Scripts/webkitperl/FeatureList.pm:
+
+2012-06-13  Csaba Osztrogonác  <ossy@webkit.org>
+
+        Update buildbot master in autoinstaller to match build.webkit.org
+        https://bugs.webkit.org/show_bug.cgi?id=88992
+
+        Reviewed by Adam Barth.
+
+        * Scripts/webkitpy/thirdparty/__init__.py:
+        (AutoinstallImportHook._install_buildbot):
+
+2012-06-13  Sheriff Bot  <webkit.review.bot@gmail.com>
+
+        Unreviewed, rolling out r120209.
+        http://trac.webkit.org/changeset/120209
+        https://bugs.webkit.org/show_bug.cgi?id=89007
+
+        Broke the WebKit2 mac build. (Requested by andersca on
+        #webkit).
+
+        * MiniBrowser/mac/BrowserWindowController.m:
+        (-[BrowserWindowController awakeFromNib]):
+        * WebKitTestRunner/TestController.cpp:
+        (WTR::TestController::initialize):
+
+2012-06-13  Zoltan Horvath  <zoltan@webkit.org>
+
+        check-webkit-style needs to respect Qt API coding style
+        https://bugs.webkit.org/show_bug.cgi?id=88995
+
+        Reviewed by Csaba Osztrogonác.
+
+        Add -whitespace/declaration exception to Qt API files.
+
+        * Scripts/webkitpy/style/checker.py:
+
+2012-06-13  Christophe Dumez  <christophe.dumez@intel.com>
+
+        [WK2] Add implementation for dispatchIntent in WebFrameLoaderClient
+        https://bugs.webkit.org/show_bug.cgi?id=88340
+
+        Reviewed by Kenneth Rohde Christiansen.
+
+        Update initialization of WKPageLoaderClient.
+
+        * MiniBrowser/mac/BrowserWindowController.m:
+        (-[BrowserWindowController awakeFromNib]):
+        * WebKitTestRunner/TestController.cpp:
+        (WTR::TestController::initialize):
+
+2012-06-13  Allan Sandfeld Jensen  <allan.jensen@nokia.com>
+
+        [Qt] Handle multiple held buttons in touch mocking.
+        https://bugs.webkit.org/show_bug.cgi?id=88865
+
+        Reviewed by Kenneth Rohde Christiansen.
+
+        Add handling for multiple held mouse-buttons in the appropiate places.
+        To do this last-pos and start-pos are now also read from the last recorded
+        touch-point instead from object-global variables.
+
+        * MiniBrowser/qt/MiniBrowserApplication.cpp:
+        (MiniBrowserApplication::notify):
+        (MiniBrowserApplication::updateTouchPoint):
+        * MiniBrowser/qt/MiniBrowserApplication.h:
+        (MiniBrowserApplication):
+
+2012-06-13  Simon Hausmann  <simon.hausmann@nokia.com>
+
+        [Qt] Make it possible to disable -Werror in production builds
+
+        Reviewed by Tor Arne Vestbø.
+
+        Don't do -Werror if build-webkit is called with --qmakearg=CONFIG+=production_build,
+        something qt5.git's qtwebkit.pri will soon do.
+
+        * qmake/mkspecs/features/unix/default_post.prf:
+
+2012-06-13  Dominik Röttsches  <dominik.rottsches@intel.com>
+
+        [EFL] Store and compare md5sum of jhbuild files to trigger cleaning Dependencies
+        https://bugs.webkit.org/show_bug.cgi?id=85790
+
+        Reviewed by Gustavo Noronha Silva.
+
+        Add md5 checks for jhbuild configuration for EFL, reusing the gtk jhbuild related
+        code in webkitdirs.pm.
+
+        * Scripts/webkitdirs.pm:
+        (jhbuildConfigurationChanged):
+        (saveJhbuildMd5):
+        (cleanJhbuild):
+        (buildAutotoolsProject):
+        (generateBuildSystemFromCMakeProject):
+
+2012-06-13  Qi Zhang  <qi.2.zhang@nokia.com>
+
+        [Qt] Add configure test for zlib and set WTF_USE_ZLIB if found
+        https://bugs.webkit.org/show_bug.cgi?id=88763
+
+        Add configure test for zlib and set WTF_USE_ZLIB if found to enable websocket
+        extension.
+
+        Reviewed by Simon Hausmann.
+
+        * qmake/config.tests/libzlib: Added.
+        * qmake/config.tests/libzlib/libzlib.cpp: Added.
+        (main):
+        * qmake/config.tests/libzlib/libzlib.pro: Added.
+        * qmake/mkspecs/features/features.prf:
+        * qmake/sync.profile:
+
+2012-06-13  Robin Cao  <robin.cao@torchmobile.com.cn>
+
+        [BlackBerry] Enable MEDIA_STREAM by default
+        https://bugs.webkit.org/show_bug.cgi?id=88849
+
+        Reviewed by Antonio Gomes.
+
+        * Scripts/webkitperl/FeatureList.pm:
+
+2012-06-13  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
+
+        Prevent stderr output from messing up detection of build path
+
+        https://bugs.webkit.org/show_bug.cgi?id=88075
+
+        Warnings such as 'perl: warning: Setting locale failed' will otherwise
+        end up as the build path when calling out to webkit-build-directory.
+
+        Reviewed by Ojan Vafai.
+
+        * Scripts/webkitpy/layout_tests/port/config.py:
+        (Config.build_directory):
+
+2012-06-13  Ryosuke Niwa  <rniwa@webkit.org>
+
+        Update the builder name for Apple Lion builders.
+
+        * Scripts/webkitpy/layout_tests/port/builders.py:
+
+2012-06-13  Joel Dillon  <joel.dillon@codethink.co.uk>
+
+        [Qt][Win]QtTestBrowser somehow picks up DumpRenderTree's main.cpp, causing a failure to link
+        https://bugs.webkit.org/show_bug.cgi?id=88874
+
+        Reviewed by Simon Hausmann.
+
+        Rename main.cpp in QtTestBrowser because on Windows
+        it conflicts with main.cpp from DumpRenderTree.
+
+        * QtTestBrowser/QtTestBrowser.pro:
+        * QtTestBrowser/qttestbrowser.cpp: Renamed from Tools/QtTestBrowser/main.cpp.
+        (launcherMain):
+        (LauncherApplication):
+        (LauncherApplication::urls):
+        (LauncherApplication::isRobotized):
+        (LauncherApplication::robotTimeout):
+        (LauncherApplication::robotExtraTime):
+        (LauncherApplication::applyDefaultSettings):
+        (LauncherApplication::LauncherApplication):
+        (requiresGraphicsView):
+        (LauncherApplication::handleUserOptions):
+        (main):
+
+2012-06-12  Lucas Forschler  <lforschler@apple.com>
+
+        Wrap bot types inside ().
+        https://bugs.webkit.org/show_bug.cgi?id=88893
+        
+        Reviewed by Stephanie Lewis.
+
+        * BuildSlaveSupport/build.webkit.org-config/config.json:
+
+2012-06-12  Christophe Dumez  <christophe.dumez@intel.com>
+
+        [soup] Prevent setting or editing httpOnly cookies from JavaScript
+        https://bugs.webkit.org/show_bug.cgi?id=88760
+
+        Reviewed by Gustavo Noronha Silva.
+
+        Update libsoup to v2.39.2, glib to v2.33.2 and glib-networking
+        to v2.33.2 for both GTK and EFL ports.
+
+        * efl/jhbuild.modules:
+        * gtk/jhbuild.modules:
+
+2012-06-12  Christophe Dumez  <christophe.dumez@intel.com>
+
+        [EFL] Enable SHADOW_DOM flag
+        https://bugs.webkit.org/show_bug.cgi?id=87732
+
+        Reviewed by Kentaro Hara.
+
+        * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
+        (DumpRenderTreeChrome::resetDefaultsToConsistentValues): Enable
+        SHADOW_DOM by default in DRT.
+        * DumpRenderTree/efl/DumpRenderTreeView.cpp:
+        (onFocusCanCycle): Enable focus cycling in EFL DRT to behave like
+        Chromium and Mac ports.
+        (drtViewAdd):
+        * Scripts/webkitperl/FeatureList.pm: Enable SHADOW_DOM flag by default
+        at compile time for EFL port.
+
+2012-06-12  Ashod Nakashian  <ashodnakashian@yahoo.com>
+
+        WinLauncher should show loading errors
+        https://bugs.webkit.org/show_bug.cgi?id=80760
+
+        Reviewed by Brent Fulgham.
+
+        * WinLauncher/WinLauncher.cpp: Added load-fail handler to show messagebox with error.
+        (WinLauncherWebHost::didFailProvisionalLoadWithError):
+        * WinLauncher/WinLauncher.h:
+        (WinLauncherWebHost::didFailProvisionalLoadWithError):
+        (WinLauncherWebHost):
+
+2012-06-12  Ryuan Choi  <ryuan.choi@samsung.com>
+
+        [EFL][WK2] Add MiniBrowserEfl.
+        https://bugs.webkit.org/show_bug.cgi?id=61850
+
+        Reviewed by Chang Shu.
+
+        Implement MiniBrowser for EFL.
+
+        * CMakeLists.txt:
+        * MiniBrowser/efl/CMakeLists.txt: Added.
+        * MiniBrowser/efl/main.c: Added.
+        (_MiniBrowser):
+        (main_signal_exit):
+        (on_ecore_evas_resize):
+        (browserCreate):
+        (main):
+
+2012-06-12  Helder Correia  <helder.correia@nokia.com>
+
+        Unreviewed. Update my email address in committers.py.
+
+        * Scripts/webkitpy/common/config/committers.py:
+
+2012-06-12  Dominik Röttsches  <dominik.rottsches@intel.com>
+
+        [EFL] Resolve CMake warnings on overlapping search paths for EFL jhbuild-enabled build
+        https://bugs.webkit.org/show_bug.cgi?id=84707
+
+        Fixing CMAKE prefix path for 64bit systems to avoid CMAKE warnings on those.
+        use_lib64 variable is preconfigured to True by jhbuild.
+
+        Reviewed by Dirk Pranke.
+
+        * efl/jhbuildrc:
+
+2012-06-12  Dana Jansens  <danakj@chromium.org>
+
+        Add zlieber@chromium.org to contributors list
+        https://bugs.webkit.org/show_bug.cgi?id=88904
+
+        Unreviewed.
+
+        * Scripts/webkitpy/common/config/committers.py:
+
+2012-06-12  Lucas Forschler  <lforschler@apple.com>
+
+        Rename Apple bots to a standard convention.
+        https://bugs.webkit.org/show_bug.cgi?id=88893
+        
+        Note: This has the side effect of grouping all of the Apple bots together in the waterfall.
+
+        Reviewed by Simon Fraser.
+
+        * BuildSlaveSupport/build.webkit.org-config/config.json:
+
+2012-06-12  Mike West  <mkwst@chromium.org>
+
+        Make document.documentURI readonly from JavaScript
+        https://bugs.webkit.org/show_bug.cgi?id=65187
+
+        Reviewed by Alexey Proskuryakov.
+
+        * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
+        * TestWebKitAPI/Tests/mac/SetDocumentURI.html: Added.
+        * TestWebKitAPI/Tests/mac/SetDocumentURI.mm: Added.
+        (-[SetDocumentURITest webView:didFinishLoadForFrame:]):
+        (TestWebKitAPI):
+        (TestWebKitAPI::TEST):
+
+2012-06-12  Christophe Dumez  <christophe.dumez@intel.com>
+
+        [EFL] enable LEGACY_WEBKIT_BLOB_BUILDER flag
+        https://bugs.webkit.org/show_bug.cgi?id=88715
+
+        Reviewed by Noam Rosenthal.
+
+        Enable LEGACY_WEBKIT_BLOB_BUILDER flag by default on EFL port.
+
+        * Scripts/webkitperl/FeatureList.pm:
+
+2012-06-12  Csaba Osztrogonác  <ossy@webkit.org>
+
+        [Qt] Fix -Werror for the buildbots
+        https://bugs.webkit.org/show_bug.cgi?id=88223
+
+        Rubber-stamped by Tor Arne Vestbø.
+
+        Revert a part of r119269 to enable Werror again for non qt-developer-builds. (for bots for example)
+
+        * qmake/mkspecs/features/unix/default_post.prf:
+
+2012-06-12  Balazs Kelemen  <kbalazs@webkit.org>
+
+        [REGRESSION][WTR] Build fail with werror after r120054
+        https://bugs.webkit.org/show_bug.cgi?id=88862
+
+        Reviewed by Csaba Osztrogonác.
+
+        Some static helpers become unused after counterValueForElementById
+        has moved to Internals. Let's remove them.
+
+        * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
+        (WTR):
+
+2012-06-12  Csaba Osztrogonác  <ossy@webkit.org>
+
+        [Qt] Slave losts cause build break on bots
+        https://bugs.webkit.org/show_bug.cgi?id=38980
+
+        Minor fix after r59261.
+
+        Reviewed by Tor Arne Vestbø.
+
+        * Scripts/build-webkit:
+        (unlinkZeroFiles): Don't remove 0 byte sized directories, because they are always 0 byte sized on Windows.
+
+2012-06-11  Kaustubh Atrawalkar  <kaustubh@motorola.com>
+
+        [DRT] LTC:: counterValueForElementById() could be moved to Internals.
+        https://bugs.webkit.org/show_bug.cgi?id=84406
+
+        Reviewed by Hajime Morita.
+
+        Move the counterValueForElementById from LayoutTestCotroller to Internals and
+        remove the old platform specific implementations as it exclusively tests WebCore functionality.
+
+        * DumpRenderTree/LayoutTestController.cpp:
+        (LayoutTestController::staticFunctions):
+        * DumpRenderTree/LayoutTestController.h:
+        (LayoutTestController):
+        * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
+        * DumpRenderTree/chromium/LayoutTestController.cpp:
+        (LayoutTestController::LayoutTestController):
+        (LayoutTestController::setPOSIXLocale):
+        * DumpRenderTree/chromium/LayoutTestController.h:
+        (LayoutTestController):
+        * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
+        * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
+        * DumpRenderTree/mac/LayoutTestControllerMac.mm:
+        * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
+        * DumpRenderTree/qt/LayoutTestControllerQt.h:
+        (LayoutTestController):
+        * DumpRenderTree/win/LayoutTestControllerWin.cpp:
+        * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
+        * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
+        * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
+        (WTR):
+        * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
+        (LayoutTestController):
+
+2012-06-11  Hans Wennborg  <hans@chromium.org>
+
+        Speech JavaScript API: Make SpeechRecognitionError an Event
+        https://bugs.webkit.org/show_bug.cgi?id=88784
+
+        Reviewed by Adam Barth.
+
+        Make it possible to have MockWebSpeechRecognizer fire error events.
+
+        * DumpRenderTree/chromium/LayoutTestController.cpp:
+        (LayoutTestController::LayoutTestController):
+        (LayoutTestController::setMockSpeechRecognitionError):
+        * DumpRenderTree/chromium/LayoutTestController.h:
+        (LayoutTestController):
+        * DumpRenderTree/chromium/MockWebSpeechRecognizer.cpp:
+        (WebKit):
+        (ErrorTask):
+        (WebKit::ErrorTask::ErrorTask):
+        (MockWebSpeechRecognizer::addMockResult):
+        (MockWebSpeechRecognizer::setError):
+        * DumpRenderTree/chromium/MockWebSpeechRecognizer.h:
+        (MockWebSpeechRecognizer):
+
+2012-06-11  Ojan Vafai  <ojan@chromium.org>
+
+        Don't show the ASAN builders in garden-o-matic since they don't get block WebKit rolls.
+        https://bugs.webkit.org/show_bug.cgi?id=88836
+
+        Reviewed by Dirk Pranke.
+
+        * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders.js:
+        * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders_unittests.js:
+
+2012-06-11  Christophe Dumez  <christophe.dumez@intel.com>
+
+        [EFL] Remove CMakeCache.txt before building
+        https://bugs.webkit.org/show_bug.cgi?id=88717
+
+        Reviewed by Antonio Gomes.
+
+        Remove CMake cache file before every build to avoid using outdated
+        build flags when new features get enabled. This will avoid unexpected
+        failures on the bots due to outdated build flags.
+
+        * Scripts/webkitdirs.pm:
+        (generateBuildSystemFromCMakeProject):
+
+2012-06-11  Vincent Scheib  <scheib@chromium.org>
+
+        Add new Pointer Lock spec attribute webkitPointerLockElement.
+        https://bugs.webkit.org/show_bug.cgi?id=88799
+
+        Reviewed by Dimitri Glazkov.
+
+        Part of a series of refactoring changes to update pointer lock API to
+        the fullscreen locking style. https://bugs.webkit.org/show_bug.cgi?id=84402
+
+        New attribute webkitPointerLockElement added. Follow up patches
+        will remove the previous isLocked attribute. Tests updated to use
+        the new attribute.
+
+        WebViewHost logic required modification to correctly repond to
+        a lock, unlock, pointerLockElement call series. Specifically,
+        unlocking must be queued after a lock command is issued always
+        as the lock state may not be set yet with a lock request in flight.
+
+        * DumpRenderTree/chromium/WebViewHost.cpp:
+        (WebViewHost::requestPointerUnlock): Always post didLosePointerLock task.
+        (WebViewHost::didLosePointerLock): Signal didLosePointerLock only if pointer was locked.
+
+2012-06-11  Alexis Menard  <alexis.menard@openbossa.org>
+
+        [CSS3 Backgrounds and Borders] Protect box-decoration-break behind a feature flag.
+        https://bugs.webkit.org/show_bug.cgi?id=88804
+
+        Reviewed by Tony Chang.
+
+        Protect box-decoration-break behind a feature flag enabled by default.
+
+        * Scripts/webkitperl/FeatureList.pm:
+        * qmake/mkspecs/features/features.pri:
+
+2012-06-11  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
+
+        [CMAKE][EFL] Remove duplicated executable output path
+        https://bugs.webkit.org/show_bug.cgi?id=88765
+
+        Reviewed by Daniel Bates.
+
+        CMake files for EFL port have redefined executable output path. However, EFL port doesn't
+        need to define again because it is already defined in top-level CMake file.
+
+        * DumpRenderTree/efl/CMakeLists.txt:
+        * EWebLauncher/CMakeLists.txt:
+
+2012-06-11  Ojan Vafai  <ojan@chromium.org>
+
+        Don't consider show warnings in the garden-o-matic UI for non-layout test failures
+        https://bugs.webkit.org/show_bug.cgi?id=88816
+
+        Reviewed by Dirk Pranke.
+
+        * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base.js:
+        Expose a way to clear AsynchronousCaches.
+        * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders.js:
+        * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders_unittests.js:
+        Clear the buildInfo cache at the beginning of each test in order to avoid having one test affect another.
+
+2012-06-11  Dirk Pranke  <dpranke@chromium.org>
+
+        nrwt exits early too frequently
+        https://bugs.webkit.org/show_bug.cgi?id=73843
+
+        Reviewed by Ryosuke Niwa.
+
+        Change the defaults for new-run-webkit-tests for both
+        --exit-after-n-failures and --exit-after-n-crashes-or-timeouts
+        to zero (i.e., we will never exit early by default).
+
+        If people want their buildbots to exit early it is their
+        responsibility to update their buildbot configs appropriately.
+
+        Note that we have tests that ensure that these flags work for
+        non-zero numbers, but we don't have tests that either test the
+        default or ensure that zero means "run all of the tests". I
+        think that's okay but can add tests if others disagree.
+
+        Also note that it looks like no changes on the build.webkit.org
+        master.cfg are necessary; all of the bots seem to be using the
+        RunWebKitTests wrapper which is already setting the flags. There
+        is a NewRunWebKitTests wrapper, but that looks to be unused; I
+        will delete that in a separate patch.
+
+        * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
+        (parse_args):
+
+2012-06-11  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
+
+        [EFL] [DRT] Reset CacheModel before running each test
+        https://bugs.webkit.org/show_bug.cgi?id=87534
+
+        Reviewed by Gustavo Noronha Silva.
+
+        Reset cache model to default before running each test to fix some
+        flaky tests on EFL build bots. Also fix case value in setCacheModel.
+
+        * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
+        (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
+        * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
+        (LayoutTestController::setCacheModel):
+
+2012-06-11  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
+
+        [GTK] Memory leaks in DRT EventSender.cpp
+        https://bugs.webkit.org/show_bug.cgi?id=88774
+
+        Reviewed by Martin Robinson.
+
+        Fix a few possible memory leaks in EventSender code.
+
+        * DumpRenderTree/gtk/EventSender.cpp:
+        (contextClickCallback): Free the event returned by gdk_event_new().
+        (mouseDownCallback): Ditto.
+        (mouseUpCallback): Ditto.
+
+2012-06-11  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
+
+        [EFL] REGRESSION (r119788): tests rely on pathToLocalResource are failing after r119788
+        https://bugs.webkit.org/show_bug.cgi?id=88661
+
+        Reviewed by Dirk Pranke.
+
+        Pass string length explicitly when creating String object from
+        non-null-terminated UChar* returned by characters().
+
+        * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
+        (LayoutTestController::pathToLocalResource):
+
+2012-06-11  David Dorwin  <ddorwin@chromium.org>
+
+        [chromium] Provide access to the WebPlugin created by the helper plugin widget
+        https://bugs.webkit.org/show_bug.cgi?id=88028
+
+        Reviewed by Adam Barth.
+
+        Added isPlaceholder() to WebPlugin.
+
+        * DumpRenderTree/chromium/TestWebPlugin.h:
+        (TestWebPlugin::isPlaceholder):
+
+2012-06-11  Tony Chang  <tony@chromium.org>
+
+        rebaseline from garden-o-matic leaves N processes each time it is run
+        https://bugs.webkit.org/show_bug.cgi?id=88586
+
+        Reviewed by Dirk Pranke.
+
+        This appears to only be a problem on python 2.7. Maybe a bug causing
+        pools to not be garbage collected?
+
+        * Scripts/webkitpy/common/system/executive.py:
+        (Executive.run_in_parallel): close() and join() the process pool.
+        * Scripts/webkitpy/common/system/executive_unittest.py:
+        (ExecutiveTest.test_run_in_parallel):
+
+2012-06-11  Zoltan Horvath  <zoltan@webkit.org>
+
+        [Qt] Add config tests for WEBP imagedecoder library, modify HAVE(decoderlibrary) to USE(...)
+        https://bugs.webkit.org/show_bug.cgi?id=87841
+
+        Reviewed by Simon Hausmann.
+
+        * qmake/config.tests/libwebp/libwebp.cpp: Added.
+        (main):
+        * qmake/config.tests/libwebp/libwebp.pro: Added.
+        * qmake/sync.profile:
+
+2012-06-10  Ryosuke Niwa  <rniwa@webkit.org>
+
+        Expose layoutTestController as testRunner
+        https://bugs.webkit.org/show_bug.cgi?id=88210
+
+        Reviewed by Darin Adler.
+
+        * DumpRenderTree/LayoutTestController.cpp:
+        (LayoutTestController::makeWindowObject):
+        * DumpRenderTree/chromium/TestShell.cpp:
+        (TestShell::bindJSObjectsToWindow):
+        * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
+        (WebCore::DumpRenderTree::initJSObjects):
+        * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
+        (WTR::LayoutTestController::makeWindowObject):
+
+2012-06-09  Dominic Cooney  <dominicc@chromium.org>
+
+        [Chromium] Remove JavaScriptCore dependencies from gyp
+        https://bugs.webkit.org/show_bug.cgi?id=88510
+
+        Reviewed by Adam Barth.
+
+        Chromium doesn't support JSC any more and there doesn't seem to be
+        a strong interest in using GYP as the common build system in other
+        ports.
+
+        * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
+        * Scripts/generate-project-files: Removed.
+
+2012-06-09  Sukolsak Sakshuwong  <sukolsak@google.com>
+
+        Add UNDO_MANAGER flag
+        https://bugs.webkit.org/show_bug.cgi?id=87908
+
+        Reviewed by Tony Chang.
+
+        * Scripts/webkitperl/FeatureList.pm:
+        * qmake/mkspecs/features/features.pri:
+
+2012-06-08  Ryosuke Niwa  <rniwa@webkit.org>
+
+        webkit-patch land-safely should set cq? if the patch author is not in committers.py
+        https://bugs.webkit.org/show_bug.cgi?id=88689
+
+        Reviewed by Dirk Pranke.
+
+        When posting a patch on Bugzilla, trun cq+ into cq? if the Bugzilla login is not listed
+        in committers.py or the contributor cannot commit. If anything, the contributor can still
+        set cq+ on thier patches manually so I don't think this will be an issue for people who
+        don't list their Bugzille email on committers.py.
+
+        * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
+        (Bugzilla._commit_queue_flag):
+        * Scripts/webkitpy/common/net/bugzilla/bugzilla_unittest.py:
+        (test_commit_queue_flag):
+        (test_commit_queue_flag.assert_commit_queue_flag):
+
+2012-06-08  Dirk Pranke  <dpranke@chromium.org>
+
+        remove obsolete NewRunWebKitTests buildbot master factory classes
+        https://bugs.webkit.org/show_bug.cgi?id=88693
+
+        Reviewed by Ryosuke Niwa.
+
+        I believe these aren't used any more ...
+
+        * BuildSlaveSupport/build.webkit.org-config/master.cfg:
+        (RunWebKitTests.getText2):
+        (BuildAndTestLeaksFactory):
+
+2012-06-08  Xianzhu Wang  <wangxianzhu@chromium.org>
+
+        [Chromium-Android] Build DumpRenderTree with Android SDK
+        https://bugs.webkit.org/show_bug.cgi?id=88598
+
+        Change WebKit side of DRT related code according to some recent changes at chromium side.
+
+        Reviewed by Adam Barth.
+
+        * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
+        * Scripts/webkitpy/layout_tests/port/chromium_android.py:
+        (ChromiumAndroidPort.__init__):
+        (ChromiumAndroidPort._path_to_driver):
+
+2012-06-08  Wei James  <james.wei@intel.com>
+
+        Add app_abi option for DumpRenderTree_apk
+        https://bugs.webkit.org/show_bug.cgi?id=88626
+
+        Reviewed by Adam Barth.
+
+        * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
+
+2012-06-08  Carlos Garcia Campos  <cgarcia@igalia.com>
+
+        [GTK] Add API to get the library version to WebKit2 GTK+
+        https://bugs.webkit.org/show_bug.cgi?id=88426
+
+        Reviewed by Martin Robinson.
+
+        * gtk/generate-gtkdoc:
+        (get_webkit2_options): Add built sources to the list of files, but
+        ignore enum types and marshallers.
+
+2012-06-07  Vincent Scheib  <scheib@chromium.org>
+
+        Layout Test pointer-lock/pointerlockchange-pointerlockerror-events.html is failing
+        https://bugs.webkit.org/show_bug.cgi?id=88604
+
+        Reviewed by Kent Tamura.
+
+        DumpRenderTree fixed to not call didLosePointerLock if pointer is not currently locked.
+        Fixes debug only assert crash in pointer-lock/pointerlockchange-pointerlockerror-events.html test.
+
+        * DumpRenderTree/chromium/WebViewHost.cpp:
+        (WebViewHost::requestPointerUnlock):
+
+2012-06-08  Hans Wennborg  <hans@chromium.org>
+
+        Speech JavaScript API: mock WebSpeechRecognizer for DumpRenderTree
+        https://bugs.webkit.org/show_bug.cgi?id=87976
+
+        Reviewed by Kent Tamura.
+
+        Provide a mock implementation of WebSpeechRecognizer for
+        DumpRenderTree. This will allow better testing of the API via layout
+        tests.
+
+        * DumpRenderTree/DumpRenderTree.gypi:
+        * DumpRenderTree/chromium/LayoutTestController.cpp:
+        (LayoutTestController::LayoutTestController):
+        (LayoutTestController::addMockSpeechRecognitionResult):
+        * DumpRenderTree/chromium/LayoutTestController.h:
+        (LayoutTestController):
+        * DumpRenderTree/chromium/MockWebSpeechRecognizer.cpp: Added.
+        (MockWebSpeechRecognizer::create):
+        (MockWebSpeechRecognizer::start):
+        (MockWebSpeechRecognizer::stop):
+        (MockWebSpeechRecognizer::abort):
+        (MockWebSpeechRecognizer::MockWebSpeechRecognizer):
+        (MockWebSpeechRecognizer::~MockWebSpeechRecognizer):
+        (MockWebSpeechRecognizer::ResultTask::runIfValid):
+        * DumpRenderTree/chromium/MockWebSpeechRecognizer.h: Added.
+        (WebKit):
+        (MockWebSpeechRecognizer):
+        (MockWebSpeechRecognizer::addMockResult):
+        (MockWebSpeechRecognizer::taskList):
+        (ClientCallTask):
+        (MockWebSpeechRecognizer::ClientCallTask::ClientCallTask):
+        (ResultTask):
+        (MockWebSpeechRecognizer::ResultTask::ResultTask):
+        (NoMatchTask):
+        (MockWebSpeechRecognizer::NoMatchTask::NoMatchTask):
+        * DumpRenderTree/chromium/WebViewHost.cpp:
+        (WebViewHost::speechRecognizer):
+        * DumpRenderTree/chromium/WebViewHost.h:
+        (WebViewHost):
+        (WebViewHost::mockSpeechRecognizer):
+
+2012-06-07  Kinuko Yasuda  <kinuko@chromium.org>
+
+        check-webkit-style needs to be taught about <public/Foo.h>
+        https://bugs.webkit.org/show_bug.cgi?id=88524
+
+        Reviewed by David Levin.
+
+        * Scripts/webkitpy/style/checkers/cpp.py:
+        (_classify_include):
+        * Scripts/webkitpy/style/checkers/cpp_unittest.py:
+        (OrderOfIncludesTest.test_public_primary_header):
+        (OrderOfIncludesTest.test_classify_include):
+
+2012-06-07  Joshua Lock  <joshua.lock@intel.com>
+
+        [EFL][DRT] Normalize file:///tmp/LayoutTests in LayoutTestController::pathToLocalResource()
+        https://bugs.webkit.org/show_bug.cgi?id=67255
+
+        Reviewed by Dirk Pranke.
+
+        Implement pathToLocalResource to redirect file:///tmp URL's to
+        DUMPRENDERTREE_TEMP and file:///tmp/LayoutTests to point to the
+        LayoutTests sub-directory of the WebKit source tree.
+
+        * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
+        (LayoutTestController::pathToLocalResource):
+
+2012-06-07  Tony Payne  <tpayne@chromium.org>
+
+        [chromium] Allow LayoutTests under new git workflow
+        https://bugs.webkit.org/show_bug.cgi?id=88575
+
+        On Windows, chromium indirects through git.bat. Adds the same hack that
+        is currently used for discovering svn.bat to support the new git
+        workflow.
+
+        Reviewed by Dirk Pranke.
+
+        * Scripts/webkitpy/common/checkout/scm/git.py:
+        (Git): Replaces all hard-coded references to 'git' with the class
+        variable executable_name.
+        (Git._check_git_architecture):
+        (Git.in_working_directory):
+        (Git.find_checkout_root):
+        (Git.read_git_config):
+        (Git.discard_local_commits):
+        (Git.local_commits):
+        (Git.working_directory_is_clean):
+        (Git.clean_working_directory):
+        (Git.status_command):
+        (Git.add_list):
+        (Git.delete_list):
+        (Git.exists):
+        (Git._current_branch):
+        (Git.changed_files):
+        (Git._changes_files_for_commit):
+        (Git.revisions_changing_file):
+        (Git.conflicted_files):
+        (Git.svn_revision):
+        (Git.create_patch):
+        (Git._run_git_svn_find_rev):
+        (Git.contents_at_revision):
+        (Git.diff_for_file):
+        (Git.show_head):
+        (Git.committer_email_for_revision):
+        (Git.apply_reverse_diff):
+        (Git.revert_files):
+        (Git.commit_with_message):
+        (Git._commit_on_branch):
+        (Git.svn_commit_log):
+        (Git.last_svn_commit_log):
+        (Git.svn_blame):
+        (Git._branch_ref_exists):
+        (Git.delete_branch):
+        (Git.remote_merge_base):
+        (Git.commit_locally_with_message):
+        (Git.push_local_commits_to_server):
+        (Git.commit_ids_from_commitish_arguments):
+        (Git.commit_message_for_local_commit):
+        (Git.files_changed_summary_for_commit):
+        * Scripts/webkitpy/common/host.py:
+        (Host._engage_awesome_windows_hacks): Adds an awesome 'git.bat' hack
+        that duplicates the logic in the existing 'svn.bat' hack.
+
+2012-06-07  Ojan Vafai  <ojan@chromium.org>
+
+        Style build architectures like build versions for non-layout test failures
+        https://bugs.webkit.org/show_bug.cgi?id=88590
+
+        Reviewed by Tony Chang.
+
+        Since we show the list of failing tests after the architecture, we want the
+        buildtype and architecture to be grouped together visually.
+
+        * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/failures.css:
+        (.effects a.failing-builder>span.architecture):
+
+2012-06-07  Xianzhu Wang  <wangxianzhu@chromium.org>
+
+        [Chromium] Remove header dependency of ImageDiff to WTF
+        https://bugs.webkit.org/show_bug.cgi?id=88422
+
+        Reviewed by Ryosuke Niwa.
+
+        * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
+        * DumpRenderTree/chromium/ImageDiff.cpp:
+        (untestedCompareImages):
+
+2012-06-07  Dongwoo Im  <dw.im@samsung.com>
+
+        [EFL] [DRT] DRT enables WebAudio when WebAudio feature is enabled.
+        https://bugs.webkit.org/show_bug.cgi?id=88491
+
+        Reviewed by Dirk Pranke.
+
+        DRT should enable WebAudio when WebAudio feature is enabled,
+        so that the WebAudio feature could be tested by the Layouttest.
+
+        * DumpRenderTree/efl/LayoutTestControllerEfl.cpp: Enable WebAudio when WebAudio feature is enabled.
+        (LayoutTestController::overridePreference):
+
+2012-06-07  Dirk Pranke  <dpranke@chromium.org>
+
+        webkit-patch rebaseline-expectations should only rebaseline the appropriate suffixes for the failure in question
+        https://bugs.webkit.org/show_bug.cgi?id=88581
+
+        Reviewed by Ojan Vafai.
+
+        webkit-paptch rebaseline-expectations was ignoring what the
+        expected results for a test were and pulling down all the result
+        files from a bot. This could lead to us pulling down old,
+        incorrect .txt or .png files. With this patch we will now only
+        pull down expectations of the appropriate file types (so an
+        IMAGE failure will only pull the png and not the txt file).
+
+        * Scripts/webkitpy/layout_tests/models/test_expectations.py:
+        (has_pixel_failures):
+        (suffixes_for_expectations):
+        * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
+        (FunctionsTest.test_suffixes_for_expectations):
+        * Scripts/webkitpy/tool/commands/rebaseline.py:
+        (AbstractRebaseliningCommand.__init__):
+        (RebaselineExpectations._is_supported_port):
+        (RebaselineExpectations._update_expectations_file):
+        (RebaselineExpectations._tests_to_rebaseline):
+        (RebaselineExpectations._rebaseline_port):
+        * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
+        (test_overrides_are_included_correctly):
+
+2012-06-07  Ojan Vafai  <ojan@chromium.org>
+
+        Fix recent null pointer error regression in the flakiness dashboard
+        https://bugs.webkit.org/show_bug.cgi?id=88573
+
+        Reviewed by Dirk Pranke.
+
+        * TestResultServer/static-dashboards/builders.js:
+        (requestBuilderList):
+        * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
+
+2012-06-07  Konrad Piascik  <kpiascik@rim.com>
+
+        Add Konrad Piascik to committers.py
+        https://bugs.webkit.org/show_bug.cgi?id=88560
+
+        Unreviewed.
+
+        * Scripts/webkitpy/common/config/committers.py:
+
+2012-06-07  Tony Chang  <tony@chromium.org>
+
+        [chromium] Win mock scrollbars appear to overflow by a pixel to the right and bottom
+        https://bugs.webkit.org/show_bug.cgi?id=77368
+
+        Reviewed by Ojan Vafai.
+
+        It doesn't look like we ever get a rect with width or height of 0.
+        I ran css1, css2.1, fast/forms, and tables without any crashes.
+
+        * DumpRenderTree/chromium/WebThemeEngineDRTWin.cpp:
+        (webRectToSkIRect): Correctly convert a WebRect to a SkIRect.
+
+2012-06-07  Dirk Pranke  <dpranke@chromium.org>
+
+        run-webkit-tests --lint-test-files gives an error when run on linux
+        https://bugs.webkit.org/show_bug.cgi?id=88551
+
+        Reviewed by Ojan Vafai.
+
+        We weren't mapping the expectations properly for the
+        google-chrome ports.
+
+        * Scripts/webkitpy/layout_tests/port/base.py:
+        (Port.path_to_test_expectations_file):
+        * Scripts/webkitpy/layout_tests/port/google_chrome_unittest.py:
+        (TestGoogleChromePort.test_path_to_expectations):
+
+2012-06-07  Daniel Erat  <derat@chromium.org>
+
+        [chromium] Add LayoutTestController::setTextSubpixelPositioning() for Linux.
+        https://bugs.webkit.org/show_bug.cgi?id=88263
+
+        Reviewed by Tony Chang.
+
+        * DumpRenderTree/chromium/LayoutTestController.cpp:
+        (LayoutTestController::LayoutTestController):
+        (LayoutTestController::reset):
+        (LayoutTestController::setTextSubpixelPositioning):
+        * DumpRenderTree/chromium/LayoutTestController.h:
+        (LayoutTestController):
+        * DumpRenderTree/chromium/fonts.conf:
+
+2012-06-07  Zan Dobersek  <zandobersek@gmail.com>
+
+        [Gtk] Assertions in gtk_widget_draw when acquiring pixel output
+        https://bugs.webkit.org/show_bug.cgi?id=88060
+
+        Reviewed by Martin Robinson.
+
+        Spin the main loop before acquiring the pixel content. This ensures things
+        like late size allocations are finished before calling gtk_widget_draw,
+        avoiding assertions and subsequent empty pixel results.
+
+        * DumpRenderTree/gtk/PixelDumpSupportGtk.cpp:
+        (createBitmapContextFromWebView):
+
+2012-06-06  Kenneth Rohde Christiansen  <kenneth@webkit.org>
+
+        [Qt] Improve the visual of the viewport info box
+        https://bugs.webkit.org/show_bug.cgi?id=88410
+
+        Reviewed by Tor Arne Vestbø.
+
+        Improve the visual and make it possible to see the content more easy.
+
+        * MiniBrowser/qt/qml/ViewportInfoItem.qml:
+
+2012-06-06  Mariusz Grzegorczyk  <mariusz.g@samsung.com>
+
+        [EFL][WK2] Compile webkit1 and webkit2 by default.
+        https://bugs.webkit.org/show_bug.cgi?id=85041
+
+        Reviewed by Chang Shu.
+
+        Compile webkit1 and webkit2 while running Tools/Scripts/build-webkit --efl.
+        Webkit2 build can be skipped by adding --no-webkit2 option to build-webkit script.
+
+        * Scripts/build-webkit:
+
+2012-06-06  Sheriff Bot  <webkit.review.bot@gmail.com>
+
+        Unreviewed, rolling out r119662.
+        http://trac.webkit.org/changeset/119662
+        https://bugs.webkit.org/show_bug.cgi?id=88487
+
+        Broke builds (Requested by rniwa on #webkit).
+
+        * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
+        * DumpRenderTree/chromium/ImageDiff.cpp:
+        (untestedCompareImages):
+
+2012-06-06  Xianzhu Wang  <wangxianzhu@chromium.org>
+
+        [Chromium] Remove header dependency of ImageDiff to WTF
+        https://bugs.webkit.org/show_bug.cgi?id=88422
+
+        Reviewed by Adam Barth.
+
+        * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
+        * DumpRenderTree/chromium/ImageDiff.cpp:
+        (untestedCompareImages):
+
+2012-06-06  Amy Ousterhout  <aousterh@chromium.org>
+
+        [WebKit] DeviceOrientation cleanup
+        https://bugs.webkit.org/show_bug.cgi?id=88406
+
+        Reviewed by Kent Tamura.
+
+        Updated LayoutTestController to use new public default constructor in WebDeviceOrientation.h.
+
+        * DumpRenderTree/chromium/LayoutTestController.cpp:
+        (LayoutTestController::setMockDeviceOrientation):
+
+2012-05-29  Dirk Pranke  <dpranke@chromium.org>
+
+        webkitpy: add support for an ordered dict of test expectations
+        https://bugs.webkit.org/show_bug.cgi?id=87802
+
+        Reviewed by Ojan Vafai.
+
+        As per https://bugs.webkit.org/show_bug.cgi?id=65834 we want to
+        support a cascaded list of test expectations files. An easy way
+        to think of this is as an ordered dictionary of name -> contents
+        for files, where the name is usually a path on disk (I say
+        usually because we want to support other kinds of expectations
+        like the compile-time skips for unsupported features in
+        webkit.py, and we want to continue to support "in-memory" test
+        expectations that don't require a filesystem).
+
+        Conveniently there is an OrderedDict implementation in Python
+        2.7+ and it is available as a backport, so the first step in
+        adding this support is to use that implementation. Subsequent
+        patches will update the test_expectations.py module (and other
+        callers) to access the dict directly.
+
+        This patch just changes the base internal implementation and
+        provides wrappers for compatibility. The derived ports
+        (WebKitPort, TestPort, etc.) still need to be updated.
+
+        * Scripts/webkitpy/layout_tests/port/base.py:
+        (Port._expectations_dict):
+        (Port.test_expectations):
+        (Port):
+        (Port._expectations_overrides_dict):
+        (Port.test_expectations_overrides):
+        * Scripts/webkitpy/layout_tests/port/port_testcase.py:
+        (PortTestCase.test_expectations_ordering):
+        (PortTestCase):
+        (PortTestCase.test_expectations_overrides_ordering):
+        * Scripts/webkitpy/thirdparty/ordered_dict.py: Added.
+        (OrderedDict):
+        (OrderedDict.__init__):
+        (OrderedDict.clear):
+        (OrderedDict.__setitem__):
+        (OrderedDict.__delitem__):
+        (OrderedDict.__iter__):
+        (OrderedDict.__reversed__):
+        (OrderedDict.popitem):
+        (OrderedDict.__reduce__):
+        (OrderedDict.__repr__):
+        (OrderedDict.copy):
+        (OrderedDict.fromkeys):
+
+2012-06-06  Mario Sanchez Prada  <msanchez@igalia.com>
+
+        Unreviewed gardening. Unit test fails on GTK 64bit debug bot.
+
+        * gtk/run-api-tests:
+        (TestRunner): Skip TestWebKitAPI/WTF/TestHashMap.
+
+2012-06-06  Dirk Pranke  <dpranke@chromium.org>
+
+        nrwt should look in 'out' for binaries on chromium win to support ninja
+        https://bugs.webkit.org/show_bug.cgi?id=88273
+
+        Reviewed by Tony Chang.
+
+        This patch standardizes the search algorithm the chromium ports
+        use to figure out which driver to run. We will look in a
+        chromium location before a webkit location (e.g., in
+        Source/WebKit/chromium/out before out/) and we will look in the
+        "legacy" directory (xcodebuild) before the directory ninja uses
+        (out).
+
+        Unfortunately due to the way the test code is set up testing the
+        properly requires some duplication of test code. I will fix that
+        in a followup patch.
+
+        * Scripts/webkitpy/layout_tests/port/chromium.py:
+        (ChromiumPort):
+        (ChromiumPort._static_build_path):
+        (ChromiumPort._build_path):
+        * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
+        (ChromiumLinuxPort):
+        (ChromiumLinuxPort._determine_driver_path_statically):
+        * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
+        (ChromiumLinuxPortTest.test_build_path):
+        (ChromiumLinuxPortTest):
+        (ChromiumLinuxPortTest.test_driver_name_option):
+        (ChromiumLinuxPortTest.path_to_image_diff):
+        * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
+        (ChromiumMacPort):
+        * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
+        (ChromiumMacPortTest.test_build_path):
+        (ChromiumMacPortTest):
+        (ChromiumMacPortTest.test_driver_name_option):
+        (ChromiumMacPortTest.path_to_image_diff):
+        * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
+        (ChromiumPortTest.test_overrides_and_builder_names):
+        * Scripts/webkitpy/layout_tests/port/chromium_win.py:
+        (ChromiumWinPort):
+        * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
+        (ChromiumWinTest.test_build_path):
+        (ChromiumWinTest.test_operating_system):
+        (ChromiumWinTest):
+        (ChromiumWinTest.test_driver_name_option):
+        (ChromiumWinPortTest.path_to_image_diff):
+        * Scripts/webkitpy/layout_tests/port/port_testcase.py:
+        (PortTestCase.assert_build_path):
+
+2012-06-06  Jessie Berlin  <jberlin@apple.com>
+
+        Remove very red Windows WebKit2 testers
+        https://bugs.webkit.org/show_bug.cgi?id=88452
+
+        Reviewed by Steve Falkenburg.
+
+        Remove the Windows 7 Release WK2 testers so that the hardware can be repurposed. We are
+        need to focus on getting the WK1 testers green.
+
+        * BuildSlaveSupport/build.webkit.org-config/config.json:
+
+2012-06-06  Dirk Pranke  <dpranke@chromium.org>
+
+        get rebaselining tools to kinda work with the skia overrides file
+        https://bugs.webkit.org/show_bug.cgi?id=88456
+
+        Reviewed by Ryosuke Niwa.
+
+        Update rebaseline-expectations so that we read in the overrides
+        when looking for tests to rebaseline, but don't read in the
+        overrides when we are writing the expectations file back out.
+        This prevents the overrides from getting written into the main
+        file. This is kind of a hack but will have to do until we
+        support multiple expectations files properly.
+
+        * Scripts/webkitpy/tool/commands/rebaseline.py:
+        (RebaselineExpectations._expectations):
+        (RebaselineExpectations._update_expectations_file):
+        (RebaselineExpectations._tests_to_rebaseline):
+        * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
+        (test_overrides_are_included_correctly):
+
+2012-06-06  Dirk Pranke  <dpranke@chromium.org>
+
+        webkitpy: two manager_worker_broker_unittest tests are broken
+        https://bugs.webkit.org/show_bug.cgi?id=88445
+
+        Reviewed by Ojan Vafai.
+
+        These tests were never updated when we removed the
+        --worker-model argument, and we were passing 'inline' instead of
+        1 (for the number of workers) :(. As a result we were creating
+        multiprocessing queues unnecessarily and that was failing on cygwin.
+
+        * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py:
+        (InterfaceTest.test_managerconnection_is_abstract):
+        (InterfaceTest.test_workerconnection_is_abstract):
+
+2012-06-06  Ryosuke Niwa  <rniwa@webkit.org>
+
+        REGRESSION(r117789): "webkit-patch rebaseline-expectations" copies contents in skia_test_expectations.txt into test_expectations.txt
+        https://bugs.webkit.org/show_bug.cgi?id=87406
+
+        Reviewed by Dirk Pranke.
+
+        * Scripts/webkitpy/layout_tests/models/test_expectations.py:
+        (TestExpectations.__init__):
+        * Scripts/webkitpy/tool/commands/rebaseline.py:
+        (RebaselineTest._update_expectations_file):
+        * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
+        (test_rebaseline_does_not_include_overrides):
+
+2012-06-06  Dirk Pranke  <dpranke@chromium.org>
+
+        webkitpy: perf tests unit tests fail on chromium win
+        https://bugs.webkit.org/show_bug.cgi?id=88279
+
+        Reviewed by Ryosuke Niwa.
+
+        I'm seeing weird failures running the unit tests in my win32
+        checkout, and we don't really expect webpagereplay to work on
+        win32 anyway, so I'm skipping the import of that module for now
+        and adding a warning.
+
+        * Scripts/run-perf-tests:
+        * Scripts/webkitpy/performance_tests/perftest.py:
+
+2012-06-06  Yong Li <yoli@rim.com>
+
+        Unreviewed. Update my email address in committers.py.
+
+        * Scripts/webkitpy/common/config/committers.py:
+
+2012-06-06  Robert Kroeger  <rjkroege@chromium.org>
+
+        Add Robert Kroeger  to commiters.py
+        https://bugs.webkit.org/show_bug.cgi?id=88415
+
+        Unreviewed.
+
+        * Scripts/webkitpy/common/config/committers.py:
+
+2012-06-06  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
+
+        [EFL][DRT] http/tests/globalhistory testcases do not pass
+        https://bugs.webkit.org/show_bug.cgi?id=82579
+
+        Reviewed by Csaba Osztrogonác.
+
+        DRT has provided callbacks for ewk_view "global history delegate" signals.
+
+        * DumpRenderTree/efl/DumpRenderTree.cpp:
+        (isGlobalHistoryTest): Aux function to ident global history testcase.
+        (createLayoutTestController):
+        * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
+        (DumpRenderTreeChrome::createView):
+        (DumpRenderTreeChrome::onTitleChanged): Handles also some global history testcases.
+        (DumpRenderTreeChrome::onWebViewNavigatedWithData): New callback function.
+        (DumpRenderTreeChrome::onWebViewServerRedirect): Ditto.
+        (DumpRenderTreeChrome::onWebViewClientRedirect): Ditto.
+        (DumpRenderTreeChrome::onWebViewPopulateVisitedLinks): Ditto.
+        * DumpRenderTree/efl/DumpRenderTreeChrome.h:
+        (DumpRenderTreeChrome):
+
+2012-06-06  János Badics  <jbadics@inf.u-szeged.hu>
+
+        [Qt] run-qtwebkit-tests hangs with -j1
+        run-qtwebkit-tests used to hang when wk2 API tests were run with -j1 and --timeout.
+        This could happen due to an uncancelled timer after results were returned. This timer was
+        started only if --timeout was set.
+        https://bugs.webkit.org/show_bug.cgi?id=85475
+
+        Reviewed by Csaba Osztrogonác.
+
+        * Scripts/run-qtwebkit-tests:
+        (run_test):
+
+2012-06-06  Michael Brüning  <michael.bruning@nokia.com>
+
+        [Qt][WK2] Add preferredMinimumContentsWidth to the viewport info in MiniBrowser.
+        https://bugs.webkit.org/show_bug.cgi?id=88327
+
+        Reviewed by Kenneth Rohde Christiansen.
+
+        * MiniBrowser/qt/MiniBrowser.qrc: Added contents_width.png.
+        * MiniBrowser/qt/icons/contents_width.png: Added.
+        * MiniBrowser/qt/qml/BrowserWindow.qml: Added button for toggling preferredMinimumContentsWidth.
+        * MiniBrowser/qt/qml/ViewportInfoItem.qml: Added display of preferredMinimumContentsWidth.
+
+2012-06-05  Jon Lee  <jonlee@apple.com>
+
+        Buildbot fix.
+
+        * BuildSlaveSupport/build.webkit.org-config/master.cfg: We lose the type attribute earlier
+        in the function. Store in a local variable.
+        (loadBuilderConfig):
+
+2012-06-05  Jon Lee  <jonlee@apple.com>
+
+        Buildbot fix.
+
+        * BuildSlaveSupport/build.webkit.org-config/master.cfg:
+        (loadBuilderConfig):
+
+2012-06-04  Dirk Pranke  <dpranke@chromium.org>
+
+        webkitpy: clean up win-specific path handling and tests
+        https://bugs.webkit.org/show_bug.cgi?id=88281
+
+        Reviewed by Adam Barth.
+
+        There was a bunch of inconsistent logic for handling path
+        conversions for win32 and cygwin paths due to us sometimes
+        looking at sys.platform and sometimes using mock hosts. This
+        patch cleans everything up so that we are required to pass
+        PlatformInfo objects to the path module and stop trying to do
+        different things when running on cygwin or win32 hosts (except
+        in the path_unittest module itself).
+
+        This may slightly reduce test coverage for the win32 code paths
+        but will be a lot easier to follow and maintain.
+
+        * Scripts/webkitpy/common/system/path.py:
+        (abspath_to_uri):
+        (_convert_path):
+        * Scripts/webkitpy/common/system/path_unittest.py:
+        (AbspathTest.platforminfo):
+        (AbspathTest.test_abspath_to_uri_cygwin):
+        (AbspathTest.test_abspath_to_uri_unixy):
+        (AbspathTest.test_abspath_to_uri_win):
+        (AbspathTest.test_abspath_to_uri_escaping_unixy):
+        (AbspathTest.test_abspath_to_uri_escaping_cygwin):
+        (AbspathTest.test_stop_cygpath_subprocess):
+        * Scripts/webkitpy/common/system/platforminfo.py:
+        (PlatformInfo.__init__):
+        (PlatformInfo.is_cygwin):
+        * Scripts/webkitpy/common/system/platforminfo_mock.py:
+        (MockPlatformInfo.is_cygwin):
+        * Scripts/webkitpy/layout_tests/port/base.py:
+        (Port.to.show_results_html_file):
+        * Scripts/webkitpy/layout_tests/port/driver.py:
+        (Driver.test_to_uri):
+        (Driver.uri_to_test):
+        * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
+        (DriverTest.test_test_to_uri):
+        (DriverTest.test_uri_to_test):
+        * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
+        (MockDRTPortTest.make_port):
+        (MockDRTTest.input_line):
+        (MockChromiumDRTTest.test_pixeltest__fails):
+        * Scripts/webkitpy/layout_tests/port/test.py:
+        * Scripts/webkitpy/layout_tests/port/win.py:
+        (WinPort.show_results_html_file):
+        * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
+        (MainTest.test_unexpected_failures):
+        (MainTest.test_results_directory_absolute):
+        (MainTest.test_results_directory_default):
+        (MainTest.test_results_directory_relative):
+
+2012-06-05  Jon Lee  <jonlee@apple.com>
+
+        Workaround buildbot bug when merging build requests.
+        https://bugs.webkit.org/show_bug.cgi?id=88158
+
+        Reviewed by Mark Rowe.
+
+        For expediency, we still want testers to pick the latest build to test, but to ensure
+        that the builders are reporting the correct revision number, we should let the building bots
+        run as they normally would. That is, the builder picks the oldest revision, and merges
+        subsequent requests. Because merged requests get appended to the list of changes for a given
+        build, the last change in that list is also the latest change. When we used pickLatestBuild
+        for the builders, this was not the case (it would always report the penultimate change).
+
+        http://trac.buildbot.net/ticket/2309 is tracking the buildbot bug.
+
+        * BuildSlaveSupport/build.webkit.org-config/config.json: Remove the mergeRequests override
+        for all Apple builders.
+        * BuildSlaveSupport/build.webkit.org-config/master.cfg: Have only non-building bots pick
+        the latest build to run.
+        (loadBuilderConfig):
+
+2012-06-05  Jessie Berlin  <jberlin@apple.com>
+
+        [Win] DumpRenderTree should call resetInternalsObject
+        https://bugs.webkit.org/show_bug.cgi?id=88346
+
+        Reviewed by Simon Fraser.
+
+        Invoke it on Windows in the same place it is invoked on Mac.
+
+        * DumpRenderTree/win/DumpRenderTree.cpp:
+        (resetWebViewToConsistentStateBeforeTesting):
+
+2012-06-05  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
+
+        [GTK] Fix two more memory leaks in DRT
+        https://bugs.webkit.org/show_bug.cgi?id=88255
+
+        Reviewed by Martin Robinson.
+
+        Fixed memory leaks in GTK's DRT code.
+
+        * DumpRenderTree/gtk/DumpRenderTree.cpp:
+        (dumpHistoryItem): webkit_web_history_item_get_target() returns a
+        g_strdup()'d string. So, use GOwnPtr to manage memory automatically.
+        * DumpRenderTree/gtk/PixelDumpSupportGtk.cpp:
+        (createBitmapContextFromWebView): Free the cairo surface.
+
+2012-06-05  Allan Sandfeld Jensen  <allan.jensen@nokia.com>
+
+        [Qt] Building when a top directory is named Source
+        https://bugs.webkit.org/show_bug.cgi?id=77727
+
+        Reviewed by Tor Arne Vestbø.
+
+        Change regex to only replace the last instance of OUT_PWD.
+
+        * qmake/mkspecs/features/default_pre.prf:
+
+2012-06-05  Dongwoo Im  <dw.im@samsung.com>
+
+        Add 'isProtocolHandlerRegistered' and 'unregisterProtocolHandler'.
+        https://bugs.webkit.org/show_bug.cgi?id=73176
+
+        Reviewed by Adam Barth.
+
+        Two more APIs are added in Custom Scheme Handler specification.
+        http://dev.w3.org/html5/spec/Overview.html#custom-handlers
+        One is 'isProtocolHandlerRegistered' to query whether the specific URL
+        is registered or not.
+        The other is 'unregisterProtocolHandler' to remove the registered URL.
+
+        * Scripts/webkitperl/FeatureList.pm: Add a macro 'ENABLE_CUSTOM_SCHEME_HANDLER'.
+
+2012-06-05  Mario Sanchez Prada  <msanchez@igalia.com>
+
+        [GTK][WK2] Implement API for Geolocation permission requests in the GTK port
+        https://bugs.webkit.org/show_bug.cgi?id=83879
+
+        Reviewed by Carlos Garcia Campos.
+
+        Make minibrowser connect to the new 'permission requests' signal
+        to allow users handle the Geolocation permission requests.
+
+        * MiniBrowser/gtk/BrowserWindow.c:
+        (geolocationRequestDialogCallback): Callback for the dialog asking
+        the user for allowing or not geolocation permission requests.
+        (webViewDecidePermissionRequest): Launch a dialog to ask the user
+        whether to allow or deny geolocation permission requests.
+        (browserWindowConstructed): Connect to 'permission-request' signal.
+
+2012-06-04  Chase Phillips  <cmp@google.com>
+
+        Make a builder group support+expect multiple loads.
+        https://bugs.webkit.org/show_bug.cgi?id=88260
+
+        Change BuilderGroup to allow expecting multiple
+        loads.  This allows merging results for a given set
+        of builders into one group.
+
+        Add a test that verifies that the group count
+        increments correctly based on successful and failed
+        list loads.
+
+        Reviewed by Ojan Vafai.
+
+        * TestResultServer/static-dashboards/builders.js:
+        (BuilderGroup):
+        (BuilderGroup.prototype.setbuilder):
+        (BuilderGroup.prototype.append):
+        (BuilderGroup.prototype.loaded):
+        (requestBuilderList.xhr.onload):
+        (requestBuilderList.xhr.onerror):
+        (onErrorLoadingBuilderList):
+        (loadBuildersList):
+        * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
+        (test):
+        * TestResultServer/static-dashboards/run-unittests.html:
+
 2012-06-04  Ryosuke Niwa  <rniwa@webkit.org>
 
         Python test fix attempt for Chromium Windows.