WebKit-https.git
7 years agoRebaseline ports after bug 109994
commit-queue@webkit.org [Sat, 23 Feb 2013 06:16:14 +0000 (06:16 +0000)]
Rebaseline ports after bug 109994
https://bugs.webkit.org/show_bug.cgi?id=110551

Unreviewed rebaseline of test expectations.

Patch by Christian Biesinger <cbiesinger@chromium.org> on 2013-02-22

* css2.1/20110323/replaced-elements-001-expected.txt: Renamed from LayoutTests/platform/mac/css2.1/20110323/replaced-elements-001-expected.txt.
* fast/forms/button-generated-content-expected.txt:
* fast/forms/select-baseline-expected.txt: Renamed from LayoutTests/platform/chromium-mac/fast/forms/select-baseline-expected.txt.
* platform/chromium-mac-lion/css2.1/20110323/replaced-elements-001-expected.png:
* platform/chromium-mac-lion/fast/forms/button-generated-content-expected.png:
* platform/chromium-mac-lion/fast/forms/select-baseline-expected.png:
* platform/chromium-mac-lion/svg/custom/foreign-object-skew-expected.png:
* platform/chromium-mac-snowleopard/css2.1/20110323/replaced-elements-001-expected.png:
* platform/chromium-mac-snowleopard/fast/forms/button-generated-content-expected.png:
* platform/chromium-mac-snowleopard/fast/forms/select-baseline-expected.png:
* platform/chromium-mac-snowleopard/svg/custom/foreign-object-skew-expected.png:
* platform/chromium-mac/css2.1/20110323/replaced-elements-001-expected.png:
* platform/chromium-mac/fast/forms/button-generated-content-expected.png:
* platform/chromium-mac/fast/forms/select-baseline-expected.png:
* platform/chromium-mac/svg/custom/foreign-object-skew-expected.png:
* platform/chromium-win/css2.1/20110323/replaced-elements-001-expected.png:
* platform/chromium-win/css2.1/20110323/replaced-elements-001-expected.txt:
* platform/chromium-win/fast/forms/button-generated-content-expected.png:
* platform/chromium-win/fast/forms/button-generated-content-expected.txt:
* platform/chromium-win/fast/forms/select-baseline-expected.png:
* platform/chromium-win/fast/forms/select-baseline-expected.txt:
* platform/chromium-win/svg/custom/foreign-object-skew-expected.png:
* platform/chromium-win/svg/custom/foreign-object-skew-expected.txt:
* platform/chromium-win/tables/mozilla/bugs/bug92647-2-expected.png:
* platform/chromium-win/tables/mozilla/bugs/bug92647-2-expected.txt:
* platform/chromium/TestExpectations:
* platform/chromium/css2.1/20110323/replaced-elements-001-expected.txt: Removed.
* platform/efl/TestExpectations:
* platform/efl/css2.1/20110323/replaced-elements-001-expected.txt:
* platform/efl/fast/forms/button-generated-content-expected.txt:
* platform/efl/fast/forms/select-baseline-expected.txt:
* platform/efl/svg/custom/foreign-object-skew-expected.txt: Renamed from LayoutTests/platform/mac/svg/custom/foreign-object-skew-expected.txt.
* platform/efl/tables/mozilla/bugs/bug92647-2-expected.txt:
* platform/gtk/TestExpectations:
* platform/gtk/css2.1/20110323/replaced-elements-001-expected.txt:
* platform/gtk/fast/forms/button-generated-content-expected.txt:
* platform/gtk/fast/forms/select-baseline-expected.txt:
* platform/gtk/svg/custom/foreign-object-skew-expected.txt:
* platform/gtk/tables/mozilla/bugs/bug92647-2-expected.txt:
* platform/mac/TestExpectations:
* platform/mac/fast/forms/select-baseline-expected.txt: Removed.
* platform/mac/tables/mozilla/bugs/bug92647-2-expected.txt: Removed.
* platform/qt/TestExpectations:
* platform/qt/css2.1/20110323/replaced-elements-001-expected.txt:
* platform/qt/fast/forms/button-generated-content-expected.txt:
* platform/qt/fast/forms/select-baseline-expected.txt:
* platform/qt/svg/custom/foreign-object-skew-expected.txt:
* platform/qt/tables/mozilla/bugs/bug92647-2-expected.txt:
* platform/win-future/fast/forms/button-generated-content-expected.txt: Copied from LayoutTests/fast/forms/button-generated-content-expected.txt.
* svg/custom/foreign-object-skew-expected.txt: Renamed from LayoutTests/platform/chromium/svg/custom/foreign-object-skew-expected.txt.
* tables/mozilla/bugs/bug92647-2-expected.txt: Renamed from LayoutTests/platform/chromium/tables/mozilla/bugs/bug92647-2-expected.txt.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143836 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[GTK] Expose all dependencies to the gyp build
mrobinson@webkit.org [Sat, 23 Feb 2013 05:40:08 +0000 (05:40 +0000)]
[GTK] Expose all dependencies to the gyp build
https://bugs.webkit.org/show_bug.cgi?id=110498

Reviewed by Dirk Pranke.

* gyp/Configuration.gypi.in: Add the missing configuration cflags here.
* gyp/Dependencies.gyp: Add targets for all missing dependencies.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143835 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoAttr: Simplify modification callbacks.
akling@apple.com [Sat, 23 Feb 2013 04:47:20 +0000 (04:47 +0000)]
Attr: Simplify modification callbacks.
<http://webkit.org/b/110598>

Reviewed by Antti Koivisto.

Instead of Attr being a friend of Element and grabbing at various internal storage and callbacks,
let Attr simply call Element::setAttribute() when its value is changed, and do all that business
from Element.

* dom/Element.h:
* dom/Element.cpp:
(WebCore::Element::setAttributeInternal):

    Call Attr::recreateTextChildAfterAttributeValueChanged() after modifying an attribute value instead
    of doing the modification through the Attr node.

* dom/Attr.h:
* dom/Attr.cpp:
(WebCore::Attr::Attr):
(WebCore::Attr::childrenChanged):

    Added an m_inChildrenChanged flag so we can prevent infinite callback recursion in childrenChanged().

(WebCore::Attr::setValue):

    Call Element::setAttribute() instead of modifying the attribute storage directly.

(WebCore::Attr::recreateTextChildAfterAttributeValueChanged):

    Factored out the "remove all children / create a new text child with attribute value" logic into
    a separate function.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143834 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUpload results to perf.webkit.org in addition to the one specified by --test-results...
rniwa@webkit.org [Sat, 23 Feb 2013 04:28:47 +0000 (04:28 +0000)]
Upload results to perf.webkit.org in addition to the one specified by --test-results-server
https://bugs.webkit.org/show_bug.cgi?id=108577

Reviewed by Dirk Pranke.

Upload results to perf.webkit.org using new JSON format as well as the host specified by
--test-results-server. The new format is needed to provide extra information perf.webkit.org
need such as the subversion commit time and test URLs. This is a temporarily measure until
we complete the transition and the old JSON format and the code to upload results to
webkit-perf.appspot.com can be deleted.

This patch adds scm.timestamp_of_latest_commit to obtain the timestamp of the latest commit present
in a svn checkout or a git clone. This information is embedded in JSON submitted to perf.webkit.org
so that the app can sort performance test results based on the timestamp of the last commit.

It also changes the repository names returned by port objects to be properly capitalized
human readable names such as WebKit instead of lowercased names such as webkit since these names
are displayed on perf.webkit.org for humans. Several users of this feature has been updated
to explicitly lowercase the names.

* Scripts/webkitpy/common/checkout/scm/git.py:
(Git.timestamp_of_latest_commit): Added. Obtains the timestamp of the last commit. Unfortunately,
git's timestamp granularity is seconds so we're losing some information compared to using a regular
subversion client. To make matters worse, git doesn't have any option to show ISO-format timestamp in
UTC so we're going to manually fiddle with timezone.

* Scripts/webkitpy/common/checkout/scm/scm.py:
(SCM.timestamp_of_latest_commit): Added.

* Scripts/webkitpy/common/checkout/scm/scm_mock.py:
(MockSCM.timestamp_of_latest_commit): Added.

* Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
(test_timestamp_of_latest_commit): Added a test for Git.timestamp_of_latest_commit.

* Scripts/webkitpy/common/checkout/scm/svn.py:
(SVN.timestamp_of_latest_commit): Added. With svn, all we need to do is to use --xml option and parse
the timestamp which is always in UTC.

* Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
(JSONResultsGeneratorBase._insert_generic_metadata): Lowercase the name. Note that the name
'chromium' needs to be substituted by 'chrome' for historical reasons.

* Scripts/webkitpy/layout_tests/port/base.py:
(Port.repository_paths): Return WebKit instead of webkit as noted above.

* Scripts/webkitpy/layout_tests/port/chromium.py:
(ChromiumPort.repository_paths): Return Chromium instead of chromium as noted above.

* Scripts/webkitpy/performance_tests/perftestsrunner.py:
(PerfTestsRunner.__init__): Store the current time in UTC as well as in local time.
(PerfTestsRunner._collect_tests):

(PerfTestsRunner._generate_and_show_results): Retrieve both regular output and one for perf.webkit.org,
and upload them appropriately.

(PerfTestsRunner._generate_results_dict): Store WebKit and Chromium revisions at which tests were ran
in revisions_for_perf_webkit and construct an output for perf.webkit.org.

(PerfTestsRunner._datetime_in_ES5_compatible_iso_format): Added.

(PerfTestsRunner._merge_slave_config_json): Merge slave configuration files into both regular output
and one for perf.webkit.org. Here, we prefix each key with "builder" for perf.webkit.org.
e.g. "processor" would be renamed to "builderProcessor".

(PerfTestsRunner._generate_output_files):

(PerfTestsRunner._upload_json): Added a remote path as an argument since we upload JSONs to /api/report
on perf.webkit.org whereas we upload it to /api/test/report on webkit-perf.appspot.com. Also added the code
to parse response as JSON when possible since perf.webkit.org returns a JSON response as opposed to
webkit-perf.appspot.com which spits out a plaintext response.

* Scripts/webkitpy/performance_tests/perftestsrunner_integrationtest.py:
(MainTest._test_run_with_json_output.mock_upload_json): Tolerate perf.webkit.org/api/report for now.
(MainTest._test_run_with_json_output): Store a UTC time as perftestrunner would do.
(MainTest.test_run_with_upload_json_should_generate_perf_webkit_json): Added.

* Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
(MainTest.test_upload_json): Moved from itegrationtest.py since it really is a unit test. Also added test
cases to parse JSON responses.
(MainTest.test_upload_json.MockFileUploader): Refactored.
(MainTest.test_upload_json.MockFileUploader.reset): Added.
(MainTest.test_upload_json.MockFileUploader.__init__):
(MainTest.test_upload_json.MockFileUploader.upload_single_text_file):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143833 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[EFL] Use EINA_LOG instead of fprintf
gyuyoung.kim@samsung.com [Sat, 23 Feb 2013 04:14:00 +0000 (04:14 +0000)]
[EFL] Use EINA_LOG instead of fprintf
https://bugs.webkit.org/show_bug.cgi?id=110313

Reviewed by Laszlo Gombos.

Replace fprintf with EINA_LOG.

Source/WebCore:

* platform/efl/NetworkInfoProviderEfl.cpp:
(WebCore::NetworkInfoProviderEfl::startUpdating):
* platform/graphics/efl/CairoUtilitiesEfl.cpp:
(WebCore::evasObjectFromCairoImageSurface):

Source/WebKit/efl:

* ewk/ewk_frame.cpp:
(_ewk_frame_debug):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143832 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoThreaded HTML parser fails resources/plain-text-unsafe.dat
tonyg@chromium.org [Sat, 23 Feb 2013 03:28:41 +0000 (03:28 +0000)]
Threaded HTML parser fails resources/plain-text-unsafe.dat
https://bugs.webkit.org/show_bug.cgi?id=110538

Reviewed by Eric Seidel.

With this patch, the background HTML parser passes all tests in plain-text-unsafe.dat.

No new tests because covered by existing tests.

* html/parser/BackgroundHTMLParser.cpp:
(WebCore::tokenExitsSVG):
(WebCore):
(WebCore::tokenExitsMath):
(WebCore::BackgroundHTMLParser::simulateTreeBuilder):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143830 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoAdd temporary work around for 32-bit plugins not working in production builds as...
weinig@apple.com [Sat, 23 Feb 2013 03:18:36 +0000 (03:18 +0000)]
Add temporary work around for 32-bit plugins not working in production builds as XPCServices
https://bugs.webkit.org/show_bug.cgi?id=110680
<rdar://problem/13236883>

Reviewed by Anders Carlsson.

* UIProcess/Plugins/mac/PluginProcessProxyMac.mm:
(WebKit::shouldUseXPC):
Temporarily don't use XPC for plug-ins for Safari while we figure out why the
32-bit XPCService is being built universal.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143829 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoCorrect InspectorInstrumentation for background HTML parser
tonyg@chromium.org [Sat, 23 Feb 2013 03:14:36 +0000 (03:14 +0000)]
Correct InspectorInstrumentation for background HTML parser
https://bugs.webkit.org/show_bug.cgi?id=110678

Reviewed by Adam Barth.

resumeParsingAfterScriptExecution potentially calls processParsedChunkFromBackgroundParser multiple times
without yielding. So it is not correct to count each of those as a separate ParseHTML event.

This patch moves the instrumentation outside of the loop.

No new tests because no new functionality.

* html/parser/HTMLDocumentParser.cpp:
(WebCore::HTMLDocumentParser::didReceiveParsedChunkFromBackgroundParser):
(WebCore::HTMLDocumentParser::processParsedChunkFromBackgroundParser):
(WebCore::HTMLDocumentParser::resumeParsingAfterScriptExecution):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143828 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[chromium] Remove call to detach WebHelperPluginImpl's frame since it is never attached
commit-queue@webkit.org [Sat, 23 Feb 2013 02:54:40 +0000 (02:54 +0000)]
[chromium] Remove call to detach WebHelperPluginImpl's frame since it is never attached
https://bugs.webkit.org/show_bug.cgi?id=110668

Patch by David Dorwin <ddorwin@chromium.org> on 2013-02-22
Reviewed by Adam Barth.

This fixes http://crbug.com/172764

* src/WebHelperPluginImpl.cpp:
(WebKit::WebHelperPluginImpl::destoryPage):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143827 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoFactor MediaSource methods out of MediaPlayer & MediaPlayerPrivate and into a new...
acolwell@chromium.org [Sat, 23 Feb 2013 02:48:57 +0000 (02:48 +0000)]
Factor MediaSource methods out of MediaPlayer & MediaPlayerPrivate and into a new MediaSourcePrivate interface.
https://bugs.webkit.org/show_bug.cgi?id=109857

Reviewed by Adam Barth.

Source/WebCore:

This is a simple refactor that moves MediaSource related methods out of MediaPlayer & MediaPlayerPrivate
so that the MediaSource implementation can be updated without further polluting these interfaces.
MediaSourcePrivate is the new WebCore interface that ports must implement to support the MediaSource API.

No new tests. No user visible behavior has changed.

* Modules/mediasource/MediaSource.cpp:
(WebCore::MediaSource::MediaSource):
(WebCore::MediaSource::duration):
(WebCore::MediaSource::setDuration):
(WebCore::MediaSource::addSourceBuffer):
(WebCore::MediaSource::removeSourceBuffer):
(WebCore::MediaSource::setReadyState):
(WebCore::MediaSource::endOfStream):
(WebCore::MediaSource::buffered):
(WebCore::MediaSource::append):
(WebCore::MediaSource::abort):
(WebCore::MediaSource::setTimestampOffset):
(WebCore::MediaSource::setPrivateAndOpen):
(WebCore):
(WebCore::MediaSource::hasPendingActivity):
(WebCore::MediaSource::stop):
* Modules/mediasource/MediaSource.h:
(MediaSource):
* html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::HTMLMediaElement):
(WebCore::HTMLMediaElement::loadResource):
(WebCore::HTMLMediaElement::createMediaPlayer):
(WebCore::HTMLMediaElement::reportMemoryUsage):
* html/HTMLMediaElement.h:
(HTMLMediaElement):
* platform/graphics/MediaPlayer.cpp:
(NullMediaPlayerPrivate):
(WebCore::NullMediaPlayerPrivate::load):
(WebCore::MediaPlayer::load):
(WebCore):
(WebCore::MediaPlayer::loadWithNextMediaEngine):
* platform/graphics/MediaPlayer.h:
(WebCore):
(MediaPlayerClient):
(MediaPlayer):
* platform/graphics/MediaPlayerPrivate.h:
(MediaPlayerPrivateInterface):
* platform/graphics/MediaSourcePrivate.h: Added.
(WebCore):
(MediaSourcePrivate): Contains the MediaSource methods extracted from MediaPlayer.
(WebCore::MediaSourcePrivate::MediaSourcePrivate):
(WebCore::MediaSourcePrivate::~MediaSourcePrivate):

Source/WebKit/chromium:

WebMediaSource and WebMediaSource client were created to replace the MediaSource methods in WebMediaPlayer and
WebMediaPlayerClient. This allows the MediaSource implementation to be updated without further polluting the
media player interfaces.

* WebKit.gyp:
* public/WebMediaPlayer.h:
(WebMediaPlayer):
(WebKit::WebMediaPlayer::load):
* public/WebMediaPlayerClient.h:
(WebKit):
* public/WebMediaSource.h: Added.
(WebKit):
(WebMediaSource):
(WebKit::WebMediaSource::~WebMediaSource):
* public/WebMediaSourceClient.h: Added.
(WebKit):
(WebMediaSourceClient):
(WebKit::WebMediaSourceClient::~WebMediaSourceClient):
* src/AssertMatchingEnums.cpp:
* src/WebMediaPlayerClientImpl.cpp:
(WebKit):
(WebMediaSourceClientImpl): Temporary WebMediaSourceClient implementation to keep things working until Chromium
                            changes land.
(WebKit::WebMediaSourceClientImpl::WebMediaSourceClientImpl):
(WebKit::WebMediaSourceClientImpl::~WebMediaSourceClientImpl):
(WebKit::WebMediaSourceClientImpl::addId):
(WebKit::WebMediaSourceClientImpl::removeId):
(WebKit::WebMediaSourceClientImpl::buffered):
(WebKit::WebMediaSourceClientImpl::append):
(WebKit::WebMediaSourceClientImpl::abort):
(WebKit::WebMediaSourceClientImpl::duration):
(WebKit::WebMediaSourceClientImpl::setDuration):
(WebKit::WebMediaSourceClientImpl::endOfStream):
(WebKit::WebMediaSourceClientImpl::setTimestampOffset):
(WebKit::WebMediaPlayerClientImpl::sourceOpened):
(WebKit::WebMediaPlayerClientImpl::sourceURL):
(WebKit::WebMediaPlayerClientImpl::load):
(WebKit::WebMediaPlayerClientImpl::loadRequested):
(WebKit::WebMediaPlayerClientImpl::loadInternal):
* src/WebMediaPlayerClientImpl.h:
(WebMediaPlayerClientImpl):
* src/WebMediaSourceImpl.cpp: Added.
(WebKit):
(MediaSourcePrivateImpl):
(WebKit::MediaSourcePrivateImpl::~MediaSourcePrivateImpl):
(WebKit::MediaSourcePrivateImpl::MediaSourcePrivateImpl):
(WebKit::MediaSourcePrivateImpl::addId):
(WebKit::MediaSourcePrivateImpl::removeId):
(WebKit::MediaSourcePrivateImpl::buffered):
(WebKit::MediaSourcePrivateImpl::append):
(WebKit::MediaSourcePrivateImpl::abort):
(WebKit::MediaSourcePrivateImpl::duration):
(WebKit::MediaSourcePrivateImpl::setDuration):
(WebKit::MediaSourcePrivateImpl::endOfStream):
(WebKit::MediaSourcePrivateImpl::setTimestampOffset):
(WebKit::WebMediaSourceImpl::WebMediaSourceImpl):
(WebKit::WebMediaSourceImpl::~WebMediaSourceImpl):
(WebKit::WebMediaSourceImpl::open):
* src/WebMediaSourceImpl.h: Added.
(WebKit):
(WebMediaSourceImpl):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143826 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoRenderLayer::scrollTo() should call FrameLoaderClient::didChangeScrollOffset()
bdakin@apple.com [Sat, 23 Feb 2013 02:13:46 +0000 (02:13 +0000)]
RenderLayer::scrollTo() should call FrameLoaderClient::didChangeScrollOffset()
https://bugs.webkit.org/show_bug.cgi?id=110673
-and corresponding-
<rdar://problem/13258596>

Reviewed by Sam Weinig.

FrameLoaderClient::didChangeScrollOffset() doesn't get called for web pages that
have overflow on the body. We can easily address this by calling it at the
end of RenderLayer::scrollTo().

* rendering/RenderLayer.cpp:
(WebCore::RenderLayer::scrollTo):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143825 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed. Update bot config for OpenSource bots to add two new Win7 Debug testers...
roger_fong@apple.com [Sat, 23 Feb 2013 02:09:56 +0000 (02:09 +0000)]
Unreviewed. Update bot config for OpenSource bots to add two new Win7 Debug testers and get rid of WinXP Debug testers.

* BuildSlaveSupport/build.webkit.org-config/config.json:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143824 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoMore rebaselines after http://trac.webkit.org/changeset/143804.
rniwa@webkit.org [Sat, 23 Feb 2013 02:00:05 +0000 (02:00 +0000)]
More rebaselines after trac.webkit.org/changeset/143804.

* html5lib/run-test1-expected.txt:
* html5lib/run-test18-expected.txt:
* html5lib/run-test7-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143823 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoDisable MediaSource on Apple Windows port
acolwell@chromium.org [Sat, 23 Feb 2013 01:59:01 +0000 (01:59 +0000)]
Disable MediaSource on Apple Windows port
https://bugs.webkit.org/show_bug.cgi?id=110494

Reviewed by Tim Horton.

Source/WebKit:

* WebKit.vcxproj/FeatureDefines.props:

WebKitLibraries:

* win/tools/vsprops/FeatureDefines.vsprops:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143822 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoWebKit can erroneously strip font-size CSS property from font element with size attribute
rniwa@webkit.org [Sat, 23 Feb 2013 01:54:30 +0000 (01:54 +0000)]
WebKit can erroneously strip font-size CSS property from font element with size attribute
https://bugs.webkit.org/show_bug.cgi?id=110657

Reviewed by Justin Garcia.

Source/WebCore:

The bug was caused by ReplaceSelectionCommand::removeRedundantStylesAndKeepStyleSpanInline
erroneously removing style attributes even on an element such as font that implicitly adds
editing style.

Fixed the bug by removing these elements or attributes when they conflict with the inline
style of the element. This is always safe because implicit style of an element is always
overridden by that of inline style.

Note that when the font element becomes "empty" (i.e. doesn't have any attributes), then
we also want to delete this font element as it doesn't contribute anything to the style.

Test: editing/pasteboard/insert-u-with-text-decoration-none.html
      editing/pasteboard/insert-font-with-size-and-css.html

* editing/ApplyStyleCommand.cpp:
(WebCore::isEmptyFontTag): Added ShouldStyleAttributeBeEmpty as an argument. This will
allow removeRedundantStylesAndKeepStyleSpanInline to ignore style attribute when we know
the attribute can be removed.

* editing/ApplyStyleCommand.h: Expose isEmptyFontTag and ShouldStyleAttributeBeEmpty.

* editing/ReplaceSelectionCommand.cpp:
(WebCore::ReplaceSelectionCommand::InsertedNodes::didReplaceNode): Added.
(WebCore::ReplaceSelectionCommand::removeRedundantStylesAndKeepStyleSpanInline): See above.
* editing/ReplaceSelectionCommand.h:
(InsertedNodes): Added a declaration of didReplaceNode.

LayoutTests:

Added two regression tests to ensure WebKit doesn't erroneously remove inline styles of an element
when its implicit styles differ from the inline styles. Also rebaselined some tests as their results
have progressed.

* editing/pasteboard/insert-u-with-text-decoration-none-expected.txt: Added.
* editing/pasteboard/insert-u-with-text-decoration-none.html: Added.
* editing/pasteboard/insert-font-with-size-and-css-expected.txt: Added.
* editing/pasteboard/insert-font-with-size-and-css.html: Added.
* editing/pasteboard/4744008-expected.txt:
* editing/pasteboard/paste-text-with-style-5-expected.txt:
* editing/pasteboard/paste-text-with-style-5.html:
* editing/pasteboard/style-from-rules-expected.txt:
* editing/pasteboard/style-from-rules.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143821 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoRemove unused make variable from DerivedSources.make
commit-queue@webkit.org [Sat, 23 Feb 2013 01:32:31 +0000 (01:32 +0000)]
Remove unused make variable from DerivedSources.make
https://bugs.webkit.org/show_bug.cgi?id=110306

Patch by Laszlo Gombos <l.gombos@samsung.com> on 2013-02-22
Reviewed by Alexey Proskuryakov.

r107026 removed the last piece of logic that was guarded with
ENABLE_DASHBOARD_SUPPORT, so now it can be removed from
DerivedSources.make.

Also combined the way ENABLE_ORIENTATION_EVENTS is set to 0
to make it more readable.

No new tests, no new functionality.

* DerivedSources.make:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143820 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoREGRESSION(r130089): Scrollbar thumb no longer re-rendered on hover
dglazkov@chromium.org [Sat, 23 Feb 2013 01:15:14 +0000 (01:15 +0000)]
REGRESSION(r130089): Scrollbar thumb no longer re-rendered on hover
https://bugs.webkit.org/show_bug.cgi?id=109230

Source/WebCore:

In r130089, I accidentally removed the extra check that made scrollbar
parts styles sad. I put it back and made them happy.

Reviewed by Eric Seidel.

Test: fast/css/pseudo-element-selector-scrollbar-hover.html

* css/SelectorChecker.cpp:
(WebCore::SelectorChecker::match): Added the check back.

LayoutTests:

Reviewed by Eric Seidel.

* fast/css/pseudo-element-selector-scrollbar-hover-expected.html: Added.
* fast/css/pseudo-element-selector-scrollbar-hover.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143819 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoDFG::SpeculativeJIT::speculateNumber() should just use SpeculateDoubleOperand instead...
fpizlo@apple.com [Sat, 23 Feb 2013 01:10:31 +0000 (01:10 +0000)]
DFG::SpeculativeJIT::speculateNumber() should just use SpeculateDoubleOperand instead of doing its own thing
https://bugs.webkit.org/show_bug.cgi?id=110659

Reviewed by Oliver Hunt and Mark Hahnenberg.

This simplifies the code, and also has the effect that if speculateNumber() is called
prior to someone actually using the number in a double context, then the number will
already be up-converted to double and ready to go.

Previously if this ever came up, the subsequent use would have to again branch to see
if the value is tagged as int or tagged as double.

On the other hand, if you ever did speculateNumber() and then used the value as a
JSValue, this will be a slow down now.

I suspect that the former (speculateNumber() and then use as number) is more likely
than the latter (speculateNumber() and then use as JSValue).

* dfg/DFGSpeculativeJIT.cpp:
(JSC::DFG::SpeculativeJIT::speculateNumber):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143818 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoDFG FixupPhase should have one common hook for knowing if a node is ever being specul...
fpizlo@apple.com [Sat, 23 Feb 2013 01:03:10 +0000 (01:03 +0000)]
DFG FixupPhase should have one common hook for knowing if a node is ever being speculated a certain way
https://bugs.webkit.org/show_bug.cgi?id=110650

Reviewed by Mark Hahnenberg.

Changes almost all calls to edge.setUseKind(kind) to be
setUseKindAndUnboxIfProfitable<kind>(edge). This will allow us to use the latter
as a hook for deciding which locals to unbox (webkit.org/b/110433).

* dfg/DFGFixupPhase.cpp:
(JSC::DFG::FixupPhase::fixupNode):
(FixupPhase):
(JSC::DFG::FixupPhase::setUseKindAndUnboxIfProfitable):
(JSC::DFG::FixupPhase::fixIntEdge):
(JSC::DFG::FixupPhase::fixDoubleEdge):
(JSC::DFG::FixupPhase::attemptToMakeIntegerAdd):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143817 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[WK2] The root and page overlay layers should be the size of the view, not the content
timothy_horton@apple.com [Sat, 23 Feb 2013 00:57:10 +0000 (00:57 +0000)]
[WK2] The root and page overlay layers should be the size of the view, not the content
https://bugs.webkit.org/show_bug.cgi?id=110660
<rdar://problem/13096456>

Reviewed by Simon Fraser.

Use the view size for the size of the root and page overlay layers.
For one, we never update these layers' sizes when the content size changes, causing size mismatches.
In addition, there's no reason an overlay shouldn't be able to draw outside of the bounds of the content.

* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.mm:
(WebKit::TiledCoreAnimationDrawingArea::updateGeometry):
(WebKit::TiledCoreAnimationDrawingArea::createPageOverlayLayer):

* WebCore.exp.in:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143816 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agopluginLoadStrategy should take a WKDictionaryRef for extensibility
andersca@apple.com [Sat, 23 Feb 2013 00:43:38 +0000 (00:43 +0000)]
pluginLoadStrategy should take a WKDictionaryRef for extensibility
https://bugs.webkit.org/show_bug.cgi?id=110656
<rdar://problem/13265303>

Reviewed by Jessie Berlin.

Source/WebCore:

Export a symbol needed by WebKit2.

* WebCore.exp.in:

Source/WebKit2:

Change pluginLoadStrategy to take a WKDictionaryRef instead of a number of parameters;
this lets us add more plug-in information if needed. Also add the page URL and rename documentURL to frameURL.

* UIProcess/API/C/WKPage.cpp:
(WKPageGetPluginInformationBundleIdentifierKey):
(WKPageGetPluginInformationBundleVersionKey):
(WKPageGetPluginInformationDisplayNameKey):
(WKPageGetPluginInformationFrameURLKey):
(WKPageGetPluginInformationMIMETypeKey):
(WKPageGetPluginInformationPageURLKey):
* UIProcess/API/C/WKPage.h:
* UIProcess/WebLoaderClient.cpp:
(WebKit):
(WebKit::WebLoaderClient::pluginLoadPolicy):
* UIProcess/WebLoaderClient.h:
(WebLoaderClient):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::getPluginPath):
(WebKit::WebPageProxy::pluginInformationBundleIdentifierKey):
(WebKit):
(WebKit::WebPageProxy::pluginInformationBundleVersionKey):
(WebKit::WebPageProxy::pluginInformationDisplayNameKey):
(WebKit::WebPageProxy::pluginInformationFrameURLKey):
(WebKit::WebPageProxy::pluginInformationMIMETypeKey):
(WebKit::WebPageProxy::pluginInformationPageURLKey):
* UIProcess/WebPageProxy.h:
(WebPageProxy):
* UIProcess/WebPageProxy.messages.in:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::createPlugin):
(WebKit::WebPage::canPluginHandleResponse):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143815 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed gardening, updating test expecations after r140374.
leviw@chromium.org [Sat, 23 Feb 2013 00:35:54 +0000 (00:35 +0000)]
Unreviewed gardening, updating test expecations after r140374.

* platform/chromium-mac-snowleopard/platform/chromium/fast/events/touch: Removed.
* platform/chromium-mac/platform/chromium/fast/events/touch/compositor-touch-hit-rects-expected.txt:
* platform/chromium-win-xp/platform/chromium/fast/events/touch: Removed.
* platform/chromium-win/platform/chromium/fast/events/touch/compositor-touch-hit-rects-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143814 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed, rolling out r143794.
dpranke@chromium.org [Sat, 23 Feb 2013 00:21:59 +0000 (00:21 +0000)]
Unreviewed, rolling out r143794.
http://trac.webkit.org/changeset/143794
https://bugs.webkit.org/show_bug.cgi?id=110542

debug linux builds broken?

Tools:

* DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:

LayoutTests:

* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143813 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoAdd descriptive names for different addMidpoint use cases
leviw@chromium.org [Sat, 23 Feb 2013 00:00:27 +0000 (00:00 +0000)]
Add descriptive names for different addMidpoint use cases
https://bugs.webkit.org/show_bug.cgi?id=110644

Reviewed by Ryosuke Niwa.

Midpoints are used to delineate ranges where we don't add line boxes for contents (collapsed spaces),
and to explicitly split a RenderText into multiple text runs so that text paragraph seperators get
their own line boxes. This patch encapsulates the different cases where midpoints are added to
lineMidpointState into 4 helper functions to make it clearer what's going on in each case.

No new tests. No change in functionality.

* rendering/RenderBlockLineLayout.cpp:
(WebCore::deprecatedAddMidpoint): Original function simply adds a midpoint to the lineMidpointState.
Renaming to deprecated to discourage callers.
(WebCore::startIgnoringSpaces): Adds a midpoint to start collapsing subsequent spaces. Asserts that
we have an even number of midpoints.
(WebCore::stopIgnoringSpaces): Adds the corresponding midpoint from startIgnoringSpaces and asserts
that it's an odd number.
(WebCore::ensureLineBoxInsideIgnoredSpaces): When ignoring spaces and we come across a RenderInline
that needs a line box, this function adds a pair of midpoints which ensures we'll later add a line
box for it.
(WebCore::ensureCharacterGetsLineBox): Adds a pair of midpoints in a text renderer to mark that
the current character needs its own line box. This is used by svg for absolutely positioned
characters, or for text paragraph seperators.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143812 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoRenderArena masking has low entropy
jschuh@chromium.org [Fri, 22 Feb 2013 23:37:20 +0000 (23:37 +0000)]
RenderArena masking has low entropy
https://bugs.webkit.org/show_bug.cgi?id=110394

Reviewed by Oliver Hunt.

No new tests. This is a hardening measure.

* rendering/RenderArena.cpp:
(WebCore::RenderArena::RenderArena):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143811 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoIndexedDB: Remove old SerializedScriptValue-based get() callbacks
alecflett@chromium.org [Fri, 22 Feb 2013 23:28:06 +0000 (23:28 +0000)]
IndexedDB: Remove old SerializedScriptValue-based get() callbacks
https://bugs.webkit.org/show_bug.cgi?id=110626

Reviewed by James Robinson.

Removal now that refactoring is complete.

* public/WebIDBCallbacks.h:
(WebKit):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143810 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoFix the overlay play button position on android
commit-queue@webkit.org [Fri, 22 Feb 2013 23:21:16 +0000 (23:21 +0000)]
Fix the overlay play button position on android
https://bugs.webkit.org/show_bug.cgi?id=110638

Patch by Min Qin <qinmin@chromium.org> on 2013-02-22
Reviewed by Ojan Vafai.

Because we are using flex box, specifying the height is not necessary.
Setting height to 100% will make the enclosure go outside the control panel.
There are no media layout test expecations for android, will do that in another patch.

* css/mediaControlsChromiumAndroid.css:
(video::-webkit-media-controls-overlay-enclosure):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143809 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed. Make WinTools.make stop copying vsprops into nested vsprops folders.
roger_fong@apple.com [Fri, 22 Feb 2013 23:20:22 +0000 (23:20 +0000)]
Unreviewed. Make WinTools.make stop copying vsprops into nested vsprops folders.

* win/tools/WinTools.make:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143808 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Chromium] Test expectations for newish tests
schenney@chromium.org [Fri, 22 Feb 2013 23:15:55 +0000 (23:15 +0000)]
[Chromium] Test expectations for newish tests

Unreviewed test expectations addition.

* platform/chromium-linux-x86/fast/text/decorations-with-text-combine-expected.txt: Added.
* platform/chromium-linux-x86/fast/text/orientation-sideways-expected.txt: Added.
* platform/chromium-linux/fast/text/decorations-with-text-combine-expected.png: Added.
* platform/chromium-linux/fast/text/decorations-with-text-combine-expected.txt: Added.
* platform/chromium-linux/fast/text/orientation-sideways-expected.png: Added.
* platform/chromium-linux/fast/text/orientation-sideways-expected.txt: Added.
* platform/chromium-mac-lion/fast/text/decorations-with-text-combine-expected.png: Added.
* platform/chromium-mac-lion/fast/text/decorations-with-text-combine-expected.txt: Added.
* platform/chromium-mac-lion/fast/text/orientation-sideways-expected.png: Added.
* platform/chromium-mac-lion/fast/text/orientation-sideways-expected.txt: Added.
* platform/chromium-mac-snowleopard/fast/text/decorations-with-text-combine-expected.txt: Added.
* platform/chromium-mac-snowleopard/fast/text/orientation-sideways-expected.txt: Added.
* platform/chromium-mac/fast/text/decorations-with-text-combine-expected.png: Added.
* platform/chromium-mac/fast/text/decorations-with-text-combine-expected.txt: Added.
* platform/chromium-mac/fast/text/orientation-sideways-expected.png: Added.
* platform/chromium-mac/fast/text/orientation-sideways-expected.txt: Added.
* platform/chromium-win-xp/fast/text/decorations-with-text-combine-expected.png: Added.
* platform/chromium-win-xp/fast/text/decorations-with-text-combine-expected.txt: Added.
* platform/chromium-win-xp/fast/text/orientation-sideways-expected.png: Added.
* platform/chromium-win-xp/fast/text/orientation-sideways-expected.txt: Added.
* platform/chromium-win/fast/text/decorations-with-text-combine-expected.png: Added.
* platform/chromium-win/fast/text/decorations-with-text-combine-expected.txt: Added.
* platform/chromium-win/fast/text/orientation-sideways-expected.png: Added.
* platform/chromium-win/fast/text/orientation-sideways-expected.txt: Added.
* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143807 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed build fix. Added missing #if PLATFORM(CHROMIUM).
haraken@chromium.org [Fri, 22 Feb 2013 23:04:36 +0000 (23:04 +0000)]
Unreviewed build fix. Added missing #if PLATFORM(CHROMIUM).

* dom/Document.cpp:
(WebCore::Document::recalcStyle):
* page/FrameView.cpp:
(WebCore::FrameView::layout):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143806 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoDocument::styleRecalc() and FrameView::layout() should be traced
haraken@chromium.org [Fri, 22 Feb 2013 23:00:57 +0000 (23:00 +0000)]
Document::styleRecalc() and FrameView::layout() should be traced
https://bugs.webkit.org/show_bug.cgi?id=110646

Reviewed by Ojan Vafai.

Document::styleRecalc() and FrameView::layout() are key factors
for web page rendering performance. We should add TRACE_EVENT()
macros to these functions.

* dom/Document.cpp:
(WebCore::Document::recalcStyle):
* page/FrameView.cpp:
(WebCore::FrameView::layout):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143805 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoTeach the html5lib tests how to test the threaded HTML parser
eric@webkit.org [Fri, 22 Feb 2013 22:57:14 +0000 (22:57 +0000)]
Teach the html5lib tests how to test the threaded HTML parser
https://bugs.webkit.org/show_bug.cgi?id=110643

Reviewed by Adam Barth.

Source/WebCore:

This adds a way to force data: url parsing to go through
the threaded parser, to enable us to test the threaded parser
using the html5lib test harness.

* html/parser/HTMLParserOptions.cpp:
(WebCore::HTMLParserOptions::HTMLParserOptions):
* page/Settings.in:

LayoutTests:

* TestExpectations:
* html5lib/resources/runner.js:
(window.onload):
* html5lib/run-template-expected.txt:
* html5lib/run-test10-expected.txt:
* html5lib/run-test11-expected.txt:
* html5lib/run-test12-expected.txt:
* html5lib/run-test14-expected.txt:
* html5lib/run-test15-expected.txt:
* html5lib/run-test16-expected.txt:
* html5lib/run-test17-expected.txt:
* html5lib/run-test19-expected.txt:
* html5lib/run-test2-expected.txt:
* html5lib/run-test20-expected.txt:
* html5lib/run-test21-expected.txt:
* html5lib/run-test22-expected.txt:
* html5lib/run-test23-expected.txt:
* html5lib/run-test24-expected.txt:
* html5lib/run-test25-expected.txt:
* html5lib/run-test26-expected.txt:
* html5lib/run-test3-expected.txt:
* html5lib/run-test4-expected.txt:
* html5lib/run-test5-expected.txt:
* html5lib/run-test6-expected.txt:
* html5lib/run-test8-expected.txt:
* html5lib/run-test9-expected.txt:
* html5lib/runner-expected.txt:
* platform/chromium/html5lib/run-test1-expected.txt:
* platform/chromium/html5lib/run-test18-expected.txt:
* platform/chromium/html5lib/run-test7-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143804 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[chromium] Register newly-created layers for animation
commit-queue@webkit.org [Fri, 22 Feb 2013 22:42:58 +0000 (22:42 +0000)]
[chromium] Register newly-created layers for animation
https://bugs.webkit.org/show_bug.cgi?id=106594

Patch by Ali Juma <ajuma@chromium.org> on 2013-02-22
Reviewed by James Robinson.

Source/Platform:

* chromium/public/WebLayerTreeView.h:
(WebLayerTreeView):
(WebKit::WebLayerTreeView::registerForAnimations):

Source/WebKit/chromium:

* WebKit.gyp:
* src/ChromeClientImpl.cpp:
(WebKit::ChromeClientImpl::ChromeClientImpl):
(WebKit::ChromeClientImpl::graphicsLayerFactory):
* src/ChromeClientImpl.h:
(ChromeClientImpl):
* src/GraphicsLayerFactoryChromium.cpp: Added.
(WebKit):
(WebKit::GraphicsLayerFactoryChromium::GraphicsLayerFactoryChromium):
(WebKit::GraphicsLayerFactoryChromium::~GraphicsLayerFactoryChromium):
(WebKit::GraphicsLayerFactoryChromium::createGraphicsLayer):
* src/GraphicsLayerFactoryChromium.h: Added.
(WebKit):
(GraphicsLayerFactoryChromium):
* src/WebViewImpl.cpp:
(WebKit::WebViewImpl::WebViewImpl):
(WebKit::WebViewImpl::graphicsLayerFactory):
(WebKit):
(WebKit::WebViewImpl::registerForAnimations):
* src/WebViewImpl.h:
(WebViewImpl):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143803 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agofast/parser/parser-yield-timing.html doesn't integrate with js-test-pre correctly
abarth@webkit.org [Fri, 22 Feb 2013 22:39:26 +0000 (22:39 +0000)]
fast/parser/parser-yield-timing.html doesn't integrate with js-test-pre correctly
https://bugs.webkit.org/show_bug.cgi?id=110645

Reviewed by Eric Seidel.

* fast/parser/parser-yield-timing.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143802 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoIncrease the max preferred width of tables to 1000000
ojan@chromium.org [Fri, 22 Feb 2013 22:38:18 +0000 (22:38 +0000)]
Increase the max preferred width of tables to 1000000
https://bugs.webkit.org/show_bug.cgi?id=110545

Reviewed by Emil A Eklund.

Source/WebCore:

Test: fast/table/large-shrink-wrapped-width.html

The old limit of 15000 was picked arbitrarily to avoid overflow.
No need to pick such a small number.

* rendering/AutoTableLayout.cpp:
(WebCore::AutoTableLayout::computeInstrinsicLogicalWidths):
As best I can tell, this code is just wrong. Removing this makes the
new test pass and doesn't cause any test failure. While we have many
tests that hit this codepath, this code would only
show a difference in the cases where the available container width is
greater then 15000, and I don't think we have any tests that hit that case
other than this new one.
* rendering/TableLayout.h:

LayoutTests:

* fast/table/large-shrink-wrapped-width-expected.txt: Added.
* fast/table/large-shrink-wrapped-width.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143801 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoREGRESSION(r143654): some fast/js test crashes on 32 bit build
fpizlo@apple.com [Fri, 22 Feb 2013 22:38:14 +0000 (22:38 +0000)]
REGRESSION(r143654): some fast/js test crashes on 32 bit build
https://bugs.webkit.org/show_bug.cgi?id=110590

Reviewed by Mark Hahnenberg.

In compileValueToInt32, the refactoring in r143654 undid one of the fixes from
r143314 due to a merge goof.

In speculateNumber, we were simply forgetting to indicate that we need a
ManualOperandSpeculation on a JSValueOperand. ManualOperandSpeculation should
be passed whenever you will be performing the type checks yourself rather than
using the operand class to do it for you.

* dfg/DFGSpeculativeJIT.cpp:
(JSC::DFG::SpeculativeJIT::compileValueToInt32):
(JSC::DFG::SpeculativeJIT::speculateNumber):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143800 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[chromium] Update test expectations for XP after 143554
vollick@chromium.org [Fri, 22 Feb 2013 22:30:32 +0000 (22:30 +0000)]
[chromium] Update test expectations for XP after 143554
https://bugs.webkit.org/show_bug.cgi?id=110636

Unreviewed gardening.

* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143799 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoThreaded HTML parser should pass fast/parser/iframe-sets-parent-to-javascript-url...
abarth@webkit.org [Fri, 22 Feb 2013 22:17:49 +0000 (22:17 +0000)]
Threaded HTML parser should pass fast/parser/iframe-sets-parent-to-javascript-url.html
https://bugs.webkit.org/show_bug.cgi?id=110637

Reviewed by Eric Seidel.

With the main thread parser, we always parse the first chunk of content
returned as the result of evaluating a JavaScript URL synchronously. In
particular, if the first chunk has an inline script, we'll execute it
synchronously.

Previous to this patch, the threaded parser would always parse this
content asynchronously. It's conceivable that there could be some
content relying on the synchronous behavior, so this patch introduces
the notion of "pinning" a parser to the main thread and uses that
concept to force the result of JavaScript URLs to be parsed on the main
thread (which is probably desirable anyway because they're likely to be
quite short).

This patch fixes fast/parser/iframe-sets-parent-to-javascript-url.html
and fast/dom/javascript-url-crash-function.html with the threaded
parser with --enable-threaded-html-parser.

* dom/DocumentParser.h:
(WebCore::DocumentParser::pinToMainThread):
(DocumentParser):
* html/parser/HTMLDocumentParser.cpp:
(WebCore::HTMLDocumentParser::HTMLDocumentParser):
(WebCore):
(WebCore::HTMLDocumentParser::pinToMainThread):
* html/parser/HTMLDocumentParser.h:
(HTMLDocumentParser):
(WebCore::HTMLDocumentParser::shouldUseThreading):
* loader/DocumentWriter.cpp:
(WebCore::DocumentWriter::replaceDocument):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143797 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[BlackBerry] Reread cookies when retrying a request with new auth credentials
commit-queue@webkit.org [Fri, 22 Feb 2013 22:16:43 +0000 (22:16 +0000)]
[BlackBerry] Reread cookies when retrying a request with new auth credentials
https://bugs.webkit.org/show_bug.cgi?id=110628

Patch by Joe Mason <jmason@rim.com> on 2013-02-22
Reviewed by George Staikos.

Internal PR: 298805
Internally Reviewed By: Leo Yang

The "isRedirect" parameter of initializePlatformRequest is only used to decide whether to
reread cookies. Since we must reread cookies when resending a request with new auth
credentials as well, rename this to "rereadCookies".

The only caller which sets this parameter rather than using the default value is
NetworkManager::startJob, which also has other processing on redirect. So add a new
"rereadCookies" parameter to startJob and ensure it is always set to true on a redirect, but
can also be set to true on a non-redirect. And when startJob is called due to new auth
credentials, set it to true.

* platform/network/blackberry/NetworkJob.cpp:
(WebCore::NetworkJob::startNewJobWithRequest): Add rereadCookies parameter, pass it to
startJob.
(WebCore::NetworkJob::handleRedirect): Set new rereadCookies param of startNewJob to true.
(WebCore::NetworkJob::notifyChallengeResult): Set new rereadCookies param of startNewJob to
true.
* platform/network/blackberry/NetworkJob.h:
(NetworkJob): Add rereadCookies parameter to startNewJobWithRequest.
* platform/network/blackberry/NetworkManager.cpp:
(WebCore::NetworkManager::startJob): Add rereadCookies parameter, pass it instead of
redirectCount to initializePlatformRequest.
* platform/network/blackberry/NetworkManager.h:
(NetworkManager): Add rereadCookies parameter to startJob.
* platform/network/blackberry/ResourceRequest.h:
(ResourceRequest): Rename isRedirect parameter to rereadCookies.
* platform/network/blackberry/ResourceRequestBlackBerry.cpp:
(WebCore::ResourceRequest::initializePlatformRequest): Rename isRedirect to rereadCookies.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143796 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUse CFNotificationCenter instead of NSNotificationCenter for SharedTimerIOS
benjamin@webkit.org [Fri, 22 Feb 2013 22:15:21 +0000 (22:15 +0000)]
Use CFNotificationCenter instead of NSNotificationCenter for SharedTimerIOS
https://bugs.webkit.org/show_bug.cgi?id=110544

Patch by Benjamin Poulain <bpoulain@apple.com> on 2013-02-22
Reviewed by Daniel Bates.

Previously, we were instantiating the Obj-C object WebCoreResumeNotifierIOS
with the only purpose of forwarding one notification to a C function.

This patch updates the code to use CFNotificationCenter to dispatch the notification
without the intermediary object.

* platform/ios/SharedTimerIOS.mm:
(WebCore::applicationDidBecomeActive):
(WebCore::setSharedTimerFireInterval):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143795 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[chromium] enable the fixed version of freetype on linux
dpranke@chromium.org [Fri, 22 Feb 2013 22:12:47 +0000 (22:12 +0000)]
[chromium] enable the fixed version of freetype on linux
https://bugs.webkit.org/show_bug.cgi?id=110542

Reviewed by Tony Chang.

Tools:

Switch to building with a fixed version of FreeType for
just DumpRenderTree. This allows us to share one version
of pixel baselines for both Lucid and Precise, and run
the layout tests in a stock Precise install (instead of a Lucid
chroot). Going forward, this approach isolates us from os-level
changes to the version of FreeType. Note that Chromium itself
still uses the system FreeType and is unaffected by this change.

This setting can still be turned off (at compile/gyp-time) by
manually setting use_custom_freetype=0.

* DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:

LayoutTests:

Add entries for two tests that pass on Lucid but fail on Precise
even with the Freetype change. This isn't worth adding in
Lucid and Precise keywords for just these two tests.

* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143794 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Chromium] Expose shadowRoot to the WebKit API
fsamuel@chromium.org [Fri, 22 Feb 2013 22:10:38 +0000 (22:10 +0000)]
[Chromium] Expose shadowRoot to the WebKit API
https://bugs.webkit.org/show_bug.cgi?id=110522

Reviewed by Dimitri Glazkov.

* public/WebElement.h:
(WebElement):
* src/WebElement.cpp:
(WebKit::WebElement::shadowRoot):
(WebKit):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143793 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUint8ClampedArray constructor is wrong in WorkerContext.idl
kbr@google.com [Fri, 22 Feb 2013 22:10:29 +0000 (22:10 +0000)]
Uint8ClampedArray constructor is wrong in WorkerContext.idl
https://bugs.webkit.org/show_bug.cgi?id=110634

Reviewed by Adam Barth.

Source/WebCore:

Tests: fast/canvas/webgl/typed-arrays-in-workers.html
       webgl/conformance/typedarrays/typed-arrays-in-workers.html

* workers/WorkerContext.idl:

LayoutTests:

Added copy of new WebGL conformance test both in old location
(fast/canvas/webgl) and new location (webgl -- not yet being run
on any ports) to avoid losing test coverage after cutover.

* fast/canvas/webgl/array-unit-tests.html:
* fast/canvas/webgl/resources/typed-array-test-cases.js: Added.
* fast/canvas/webgl/resources/typed-array-worker.js: Added.
(constructTypedArray):
(constructDataView):
(onmessage):
* fast/canvas/webgl/typed-arrays-in-workers-expected.txt: Added.
* fast/canvas/webgl/typed-arrays-in-workers.html: Added.
* webgl/conformance/typedarrays/typed-arrays-in-workers-expected.txt: Added.
* webgl/conformance/typedarrays/typed-arrays-in-workers.html: Added.
* webgl/resources/webgl_test_files/conformance/typedarrays/00_test_list.txt:
* webgl/resources/webgl_test_files/conformance/typedarrays/array-unit-tests.html:
* webgl/resources/webgl_test_files/conformance/typedarrays/resources/typed-array-test-cases.js: Added.
* webgl/resources/webgl_test_files/conformance/typedarrays/resources/typed-array-worker.js: Added.
(constructTypedArray):
(constructDataView):
(onmessage):
* webgl/resources/webgl_test_files/conformance/typedarrays/typed-arrays-in-workers.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143792 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoAdd some checks to DatabaseTracker::getMaxSizeForDatabase() to ensure
mark.lam@apple.com [Fri, 22 Feb 2013 22:07:44 +0000 (22:07 +0000)]
Add some checks to DatabaseTracker::getMaxSizeForDatabase() to ensure
that it returns a sane value.
https://bugs.webkit.org/show_bug.cgi?id=110557.

Reviewed by Geoffrey Garen.

No layout test, but there is a quota-test.html attached to bugzilla.
The test is a webpage that can be loaded into multiple tabs to
consuming storage space. Once the test webpages are loaded, you will
need to monitor the database directory and its files to confirm that
growth is bounded. Also try reloading the test in the tabs. At no
time should any database file ever exceed the quota.

* Modules/webdatabase/DatabaseTracker.cpp:
(WebCore::DatabaseTracker::getMaxSizeForDatabase):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143791 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoSource/WebKit2: Move pluginLoadPolicy to the page loader client
andersca@apple.com [Fri, 22 Feb 2013 22:06:51 +0000 (22:06 +0000)]
Source/WebKit2: Move pluginLoadPolicy to the page loader client
https://bugs.webkit.org/show_bug.cgi?id=110635
<rdar://problem/13265303>

Reviewed by Sam Weinig.

Move the pluginLoadPolicy callback to the page loader client where it belongs.

* UIProcess/API/C/WKPage.h:
* UIProcess/WebLoaderClient.cpp:
(WebKit::toWKPluginLoadPolicy):
(WebKit):
(WebKit::toPluginModuleLoadPolicy):
(WebKit::WebLoaderClient::pluginLoadPolicy):
* UIProcess/WebLoaderClient.h:
(WebLoaderClient):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::getPluginPath):
* UIProcess/WebUIClient.cpp:
(WebKit):
* UIProcess/WebUIClient.h:
(WebUIClient):
* UIProcess/mac/WebInspectorProxyMac.mm:
(WebKit::WebInspectorProxy::platformCreateInspectorPage):

Tools: Move pluginLoadPolicy to the page loader client
https://bugs.webkit.org/show_bug.cgi?id=110635

Reviewed by Sam Weinig.

* MiniBrowser/mac/WK2BrowserWindowController.m:
(-[WK2BrowserWindowController awakeFromNib]):
* WebKitTestRunner/TestController.cpp:
(WTR::TestController::createOtherPage):
(WTR::TestController::createWebViewWithOptions):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143790 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoPreloads should be cleared when JavaScript cancels loading prematurely.
simonjam@chromium.org [Fri, 22 Feb 2013 22:00:09 +0000 (22:00 +0000)]
Preloads should be cleared when JavaScript cancels loading prematurely.
https://bugs.webkit.org/show_bug.cgi?id=110388

Reviewed by Antti Koivisto.

Source/WebCore:

If a page preloads a resource, but JavaScript aborts parsing for any reason (for example, a
redirect or document.write()), the resource is forever marked a preload and won't be reloaded
even if it's explicitly marked no-cache and the page is reloaded.

Test: http/tests/cache/preload-cleared-after-parrsing-canceled-by-js.html

* dom/Document.cpp:
(WebCore::Document::implicitClose):
(WebCore::Document::finishedParsing):

LayoutTests:

* http/tests/cache/preload-cleared-after-parrsing-canceled-by-js-expected.txt: Added.
* http/tests/cache/preload-cleared-after-parrsing-canceled-by-js.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143789 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoExpect editing/spelling/spelling-changed-text.html to fail on Qt and GTK
commit-queue@webkit.org [Fri, 22 Feb 2013 21:50:47 +0000 (21:50 +0000)]
Expect editing/spelling/spelling-changed-text.html to fail on Qt and GTK
https://bugs.webkit.org/show_bug.cgi?id=110447

Patch by Rouslan Solomakhin <rouslan@chromium.org> on 2013-02-22
Reviewed by Ryosuke Niwa.

Qt and GTK layout tests do not mark the word "wellcome" as a misspelling in
editing/spelling/spelling-changed-text.html layout test. This patch disables
editing/spelling/spelling-changed-text.html on Qt and GTK platforms.

* platform/gtk/TestExpectations: Mark editing/spelling/spelling-changed-text.html as [ Failure ].
* platform/qt/TestExpectations: Mark editing/spelling/spelling-changed-text.html as [ Failure ].

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143788 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoExpose a list of all reasons that qualify a RenderLayer to be composited
graouts@apple.com [Fri, 22 Feb 2013 21:45:55 +0000 (21:45 +0000)]
Expose a list of all reasons that qualify a RenderLayer to be composited
https://bugs.webkit.org/show_bug.cgi?id=110505

Expose the various reasons that can qualify a RenderLayer to be composited
via a new bitmask returned by the reasonForCompositing() method on
RenderLayerCompositor. This method already existed and was used for logging
purposes, but the previous functionality is now accessed via the
logReasonForCompositing() method.

This will allow the the InspectorLayerTreeAgent to provide this information
to the front-end on a per-layer basis as requested by the front-end.

Reviewed by Simon Fraser.

No new tests.

* rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::logLayerInfo):
(WebCore::RenderLayerCompositor::reasonsForCompositing):
(WebCore):
(WebCore::RenderLayerCompositor::logReasonsForCompositing):
* rendering/RenderLayerCompositor.h:
(RenderLayerCompositor):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143787 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[BlackBerry] Eliminate invalid clipping for content editable selections
mifenton@rim.com [Fri, 22 Feb 2013 21:41:53 +0000 (21:41 +0000)]
[BlackBerry] Eliminate invalid clipping for content editable selections
https://bugs.webkit.org/show_bug.cgi?id=110630

Reviewed by Rob Buis.

PR 284629.

Remove selection clipping for content editable
fields as the bounds aren't valid and cause
clipping based on starting node rather than the
union of nodes.

Reviewed Internally by Nima Ghanavatian.

* WebKitSupport/InputHandler.cpp:
(BlackBerry::WebKit::InputHandler::updateFormState):
* WebKitSupport/SelectionHandler.cpp:
(BlackBerry::WebKit::SelectionHandler::clippingRectForVisibleContent):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143786 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoScrollingStateTree::rootStateNode() should be null-checked since it can be null
bdakin@apple.com [Fri, 22 Feb 2013 21:38:08 +0000 (21:38 +0000)]
ScrollingStateTree::rootStateNode() should be null-checked since it can be null
https://bugs.webkit.org/show_bug.cgi?id=110631

Reviewed by Simon Fraser.

This is a speculative fix for <rdar://problem/13233090> and
<rdar://problem/12519348>. It is totally possible for rootStateNode() to be null,
so we should null-check it.

* page/scrolling/mac/ScrollingCoordinatorMac.mm:
(WebCore::ScrollingCoordinatorMac::setShouldUpdateScrollLayerPositionOnMainThread):
(WebCore::ScrollingCoordinatorMac::syncChildPositions):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143785 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoCheck for null-device when calling createCompatibleDevice
reed@google.com [Fri, 22 Feb 2013 21:30:51 +0000 (21:30 +0000)]
Check for null-device when calling createCompatibleDevice
https://bugs.webkit.org/show_bug.cgi?id=107981

Reviewed by James Robinson.

No new tests. Current tests provided coverage. Site in the wild triggered the failure
(cairo failed to allocate a huge surface).

* platform/graphics/skia/ImageBufferSkia.cpp:
(WebCore::ImageBuffer::ImageBuffer):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143784 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[BlackBerry]Adjust fatfinger detection rect size
commit-queue@webkit.org [Fri, 22 Feb 2013 21:24:22 +0000 (21:24 +0000)]
[BlackBerry]Adjust fatfinger detection rect size
https://bugs.webkit.org/show_bug.cgi?id=108678.

Patch by Tiancheng Jiang <tijiang@rim.com> on 2013-02-22
Reviewed by Antonio Gomes.

Changing HitTestRequest::IgnoreClipping does not solve the problem that
FatFinger rect detect the element out of the viewport. We have to clip
the fatfinger rect according to the current viewport size to avoid picking
the element out of the viewport.

* WebKitSupport/FatFingers.cpp:
(BlackBerry::WebKit::FatFingers::fingerRectForPoint):
(BlackBerry::WebKit::FatFingers::findBestPoint):
(BlackBerry::WebKit::FatFingers::getAdjustedPaddings):
(BlackBerry::WebKit::FatFingers::getNodesFromRect):
* WebKitSupport/FatFingers.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143783 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoAdd script to generate LayoutTests from WebGL Conformance Tests
kbr@google.com [Fri, 22 Feb 2013 21:20:57 +0000 (21:20 +0000)]
Add script to generate LayoutTests from WebGL Conformance Tests
https://bugs.webkit.org/show_bug.cgi?id=110525

Patch by Gregg Tavares <gman@chromium.org> on 2013-02-22
Reviewed by Kenneth Russell.

Adds the script generate-webgl-tests.py which given a path
to the WebGL Conformance 'sdk/tests' folder generates and
or updates LayoutTests for WebGL.  Example

git clone git://github.com/KhronosGroup/WebGL.git
generate-webgl-tests.py -w WebGL/sdk/tests -e

* webgl/generate-webgl-tests.py: Added.
(ReadFile):
(WriteFile):
(CopyTree):
(FileReader):
(GreaterThanOrEqualToVersion):
(GetTestList):
(main):
* webgl/resources/webgl-expectation-template.txt: Added.
* webgl/resources/webgl-wrapper-template.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143782 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoAudioDestination::create() needs extra device identification information for live...
crogers@google.com [Fri, 22 Feb 2013 20:49:16 +0000 (20:49 +0000)]
AudioDestination::create() needs extra device identification information for live/local input streams
https://bugs.webkit.org/show_bug.cgi?id=109494

Reviewed by James Robinson.

Source/Platform:

* chromium/public/WebMediaStreamSource.h:
(WebMediaStreamSource):

Source/WebCore:

AudioDestination::create() supports live/local audio input.  But, since there may be multiple
audio input devices available, an identifier for the requested input device needs to be
passed in.  The embedder may then use this information so that the proper audio hardware is
accessed.

* Modules/webaudio/AudioContext.cpp:
(WebCore::AudioContext::createMediaStreamSource):
* Modules/webaudio/AudioDestinationNode.h:
(AudioDestinationNode):
* Modules/webaudio/DefaultAudioDestinationNode.cpp:
(WebCore::DefaultAudioDestinationNode::initialize):
(WebCore::DefaultAudioDestinationNode::createDestination):
(WebCore::DefaultAudioDestinationNode::enableInput):
* Modules/webaudio/DefaultAudioDestinationNode.h:
(DefaultAudioDestinationNode):
* Modules/webaudio/OfflineAudioDestinationNode.h:
* platform/audio/AudioDestination.h:
(AudioDestination):
* platform/audio/chromium/AudioDestinationChromium.cpp:
(WebCore::AudioDestination::create):
(WebCore::AudioDestinationChromium::AudioDestinationChromium):
* platform/audio/chromium/AudioDestinationChromium.h:
(AudioDestinationChromium):
* platform/audio/gstreamer/AudioDestinationGStreamer.cpp:
(WebCore::AudioDestination::create):
* platform/audio/mac/AudioDestinationMac.cpp:
(WebCore::AudioDestination::create):
* platform/chromium/support/WebMediaStreamSource.cpp:
(WebKit::WebMediaStreamSource::deviceId):
(WebKit):
(WebKit::WebMediaStreamSource::setDeviceId):
* platform/mediastream/MediaStreamSource.h:
(WebCore::MediaStreamSource::deviceId):
(WebCore::MediaStreamSource::setDeviceId):
(MediaStreamSource):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143781 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed GTK gardening.
zandobersek@gmail.com [Fri, 22 Feb 2013 20:44:17 +0000 (20:44 +0000)]
Unreviewed GTK gardening.

Removed expectations for tests that now pass. Adjusted an expectation for a flaky crasher.
Marked a few more sputnik tests as slow.

* platform/gtk-wk1/TestExpectations:
* platform/gtk/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143780 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoMac WebKit2 rebaselines.
rniwa@webkit.org [Fri, 22 Feb 2013 20:41:00 +0000 (20:41 +0000)]
Mac WebKit2 rebaselines.

Partially revert http://trac.webkit.org/changeset/143709 since the expected result for
cross-origin-local-storage-wk1-expected.txt is different on Mac WebKit on the contrary to
the change log description in the changeset.

Also skip editing/spelling/spelling-changed-text.html on Mac since the test requires
the support for asynchronous spellchecking.

* platform/mac-wk2/http/tests/security/cross-origin-local-storage-wk1-expected.txt: Copied
    from platform/mac/http/tests/security/cross-origin-local-storage-wk1-expected.txt.
* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143779 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoAdd a Mac baseline for the test added in r142955.
rniwa@webkit.org [Fri, 22 Feb 2013 20:30:58 +0000 (20:30 +0000)]
Add a Mac baseline for the test added in r142955.

* platform/mac/TestExpectations:
* platform/mac/svg/filters/filter-hidden-content-expected.png: Added.
* platform/mac/svg/filters/filter-hidden-content-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143773 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed GTK gardening.
zandobersek@gmail.com [Fri, 22 Feb 2013 20:28:37 +0000 (20:28 +0000)]
Unreviewed GTK gardening.

* platform/gtk/TestExpectations: Marking a bunch of (mostly Sputnik) tests as slow.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143772 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoBinding tests rebaseline after r143737.
rniwa@webkit.org [Fri, 22 Feb 2013 20:22:33 +0000 (20:22 +0000)]
Binding tests rebaseline after r143737.

* bindings/scripts/test/V8/V8Float64Array.cpp:
* bindings/scripts/test/V8/V8TestActiveDOMObject.cpp:
* bindings/scripts/test/V8/V8TestCustomNamedGetter.cpp:
* bindings/scripts/test/V8/V8TestEventConstructor.cpp:
* bindings/scripts/test/V8/V8TestEventTarget.cpp:
* bindings/scripts/test/V8/V8TestException.cpp:
* bindings/scripts/test/V8/V8TestInterface.cpp:
* bindings/scripts/test/V8/V8TestMediaQueryListListener.cpp:
* bindings/scripts/test/V8/V8TestNamedConstructor.cpp:
* bindings/scripts/test/V8/V8TestNode.cpp:
* bindings/scripts/test/V8/V8TestObj.cpp:
* bindings/scripts/test/V8/V8TestOverloadedConstructors.cpp:
* bindings/scripts/test/V8/V8TestSerializedScriptValueInterface.cpp:
* bindings/scripts/test/V8/V8TestTypedefs.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143771 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoINPUT_MULTIPLE_FIELDS_UI: Unable to enter "24" to hour field
commit-queue@webkit.org [Fri, 22 Feb 2013 20:18:37 +0000 (20:18 +0000)]
INPUT_MULTIPLE_FIELDS_UI: Unable to enter "24" to hour field
https://bugs.webkit.org/show_bug.cgi?id=110431

Patch by Kunihiko Sakamoto <ksakamoto@chromium.org> on 2013-02-22
Reviewed by Kent Tamura.

Source/WebCore:

Fixed a bug that hour-field does not accept "24" as a valid input when hour format is 1-24.
To test <input> against different time formats, added "pattern" attribute to DateTimeEditElement.

Test: fast/forms/time-multiple-fields/time-multiple-fields-stepup-stepdown-from-renderer.html

* html/BaseMultipleFieldsDateAndTimeInputType.cpp:
(WebCore::BaseMultipleFieldsDateAndTimeInputType::updateInnerTextValue): Overwrites layoutParameters.dateTimeFormat by pattern attribute of DateTimeEditElement (if exists).
* html/shadow/DateTimeFieldElements.cpp:
(WebCore::DateTimeHourFieldElement::populateDateTimeFieldsState): Fixed a bug where AM/PM value was passed to setHour().
(WebCore::DateTimeHourFieldElement::setValueAsInteger): Clamps the given value to 0-24 when the range of the field is 1-12 or 1-24. Renamed the parameter as it may not be an hour23 value (0-23).

LayoutTests:

Added test cases for hour formats 0-11 and 1-24.

* fast/forms/time-multiple-fields/time-multiple-fields-stepup-stepdown-from-renderer-expected.txt:
* fast/forms/time-multiple-fields/time-multiple-fields-stepup-stepdown-from-renderer.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143770 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed. Changed gman's primary address for auto-complete in Bugzilla.
kbr@google.com [Fri, 22 Feb 2013 20:08:22 +0000 (20:08 +0000)]
Unreviewed. Changed gman's primary address for auto-complete in Bugzilla.

* Scripts/webkitpy/common/config/committers.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143769 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoNot reviewed.
ggaren@apple.com [Fri, 22 Feb 2013 20:05:21 +0000 (20:05 +0000)]
Not reviewed.

Fix the 32-bit build by using the right data type in more places.

* runtime/CodeCache.h:
(CodeCacheMap):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143768 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[BlackBerry] Accept key events even when composing region is active.
mifenton@rim.com [Fri, 22 Feb 2013 19:59:12 +0000 (19:59 +0000)]
[BlackBerry] Accept key events even when composing region is active.
https://bugs.webkit.org/show_bug.cgi?id=110617

Reviewed by Rob Buis.

PR 293598.

Don't reject key events that arrive when composing region is active,
this can prevent navigation events from firing, instead end
the composition and process the key.

Reviewed Internally by Nima Ghanavatian.

* WebKitSupport/InputHandler.cpp:
(BlackBerry::WebKit::InputHandler::updateFormState):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143767 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[CSS Regions] Region boxes should respect -shape-inside CSS property
zoltan@webkit.org [Fri, 22 Feb 2013 19:53:35 +0000 (19:53 +0000)]
[CSS Regions] Region boxes should respect -shape-inside CSS property
https://bugs.webkit.org/show_bug.cgi?id=107880

Reviewed by David Hyatt.

The shape-inside CSS property modifies the shape of the inner inline flow content from rectangular content box
to an arbitrary geometry. The detailed documentation is located in the CSS-Exclusions specification:
http://dev.w3.org/csswg/css3-exclusions/#shape-inside-property. This change adds support for the shape-inside property
on CSS Regions.

Source/WebCore:

Test: fast/regions/shape-inside-on-regions.html

* rendering/ExclusionShapeInfo.cpp:
(WebCore):
(WebCore::::logicalTopOffset): Shape dimension should be relative to the current region.
* rendering/ExclusionShapeInfo.h:
(ExclusionShapeInfo):
(WebCore::::logicalTopOffset): Move the implementation to ExclusionShapeInfo.cpp.
* rendering/RenderBlockLineLayout.cpp:
(WebCore::logicalHeightForLine): Move the function prior to layoutExclusionShapeInsideInfo, since we need to use it there.
(WebCore):
(WebCore::layoutExclusionShapeInsideInfo): Add support for cases when shape-inside is a region-block property.
(WebCore::RenderBlock::layoutRunsAndFloatsInRange): Update the exclusion shape information for every line, since the content
could flow through several region containers which can have different shape-inside properties.

LayoutTests:

* fast/regions/shape-inside-on-regions-expected.html: Added.
* fast/regions/shape-inside-on-regions.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143766 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoNot reviewed.
ggaren@apple.com [Fri, 22 Feb 2013 19:51:31 +0000 (19:51 +0000)]
Not reviewed.

Fix the 32-bit build by using the right data type.

* runtime/CodeCache.h:
(JSC::CodeCacheMap::find):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143765 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoTwo video contollers appear after exiting the fullscreen video when using menubar...
jer.noble@apple.com [Fri, 22 Feb 2013 19:48:02 +0000 (19:48 +0000)]
Two video contollers appear after exiting the fullscreen video when using menubar to exit.
https://bugs.webkit.org/show_bug.cgi?id=110506

Reviewed by Eric Carlson.

When we are notified that the fullscreen window is about to exit fullscreen mode unrequested,
call requestExitFullScreen(), as that will result in Document::webkitExitFullScreen() correctly
tearing down the fullscreen state.

* UIProcess/mac/WKFullScreenWindowController.mm:
(-[WKFullScreenWindowController _startExitFullScreenAnimationWithDuration:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143764 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoMac OS build broken by new ExpressionParser.cpp
kbr@google.com [Fri, 22 Feb 2013 19:44:44 +0000 (19:44 +0000)]
Mac OS build broken by new ExpressionParser.cpp
https://bugs.webkit.org/show_bug.cgi?id=110629

Unreviewed build fix. Regenerated ExpressionParser with Bison 2.3.

* src/compiler/preprocessor/ExpressionParser.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143763 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoAdd computeInstrinsicLogicalWidths functions to TableLayout subclasses
ojan@chromium.org [Fri, 22 Feb 2013 19:40:43 +0000 (19:40 +0000)]
Add computeInstrinsicLogicalWidths functions to TableLayout subclasses
https://bugs.webkit.org/show_bug.cgi?id=110520

Reviewed by Tony Chang.

This is just a refactor in preparation for adding RenderTable::computeIntinsicLogicalWidths.
No change in behavior.

Split the computePreferredLogicalWidths functions into a computeIntinsicLogicalWidths
and a applyPreferredLogicalWidthQuirks. Also, move the addition of bordersPaddingAndSpacing
into RenderTable since both TableLayout subclasses need this.

This has the added readability benefit of better isolating the quirks that each table layout
mode requires.

* rendering/AutoTableLayout.cpp:
(WebCore::AutoTableLayout::computeIntrinsicLogicalWidths):
(WebCore::AutoTableLayout::applyPreferredLogicalWidthQuirks):
* rendering/AutoTableLayout.h:
(AutoTableLayout):
* rendering/FixedTableLayout.cpp:
(WebCore::FixedTableLayout::computeIntrinsicLogicalWidths):
(WebCore::FixedTableLayout::applyPreferredLogicalWidthQuirks):
* rendering/FixedTableLayout.h:
(FixedTableLayout):
* rendering/RenderTable.cpp:
(WebCore::RenderTable::computePreferredLogicalWidths):
* rendering/TableLayout.h:
(TableLayout):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143762 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoCode cache size should adapt to workload
ggaren@apple.com [Fri, 22 Feb 2013 19:16:03 +0000 (19:16 +0000)]
Code cache size should adapt to workload
https://bugs.webkit.org/show_bug.cgi?id=110560

Reviewed by Antti Koivisto.

(*) 5% PLT arithmetic mean speedup
(*) 10% PLT geometric mean speedup
(*) 3.4X microbenchmark speedup
(*) Reduces initial cache capacity by 16X

* runtime/CodeCache.cpp:
(JSC::CodeCache::CodeCache): Updated for interface change.

* runtime/CodeCache.h:
(JSC::SourceCodeValue::SourceCodeValue):
(SourceCodeValue): Turned the cache value into a struct so it can track its age.

(CodeCacheMap):
(JSC::CodeCacheMap::CodeCacheMap):
(JSC::CodeCacheMap::find):
(JSC::CodeCacheMap::set):
(JSC::CodeCacheMap::clear):
(JSC::CodeCacheMap::pruneIfNeeded):
(CodeCache): Grow and shrink in response to usage.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143759 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[New Multicolumn] Correctly track whether or not a layer is paginated.
hyatt@apple.com [Fri, 22 Feb 2013 19:09:51 +0000 (19:09 +0000)]
[New Multicolumn] Correctly track whether or not a layer is paginated.
https://bugs.webkit.org/show_bug.cgi?id=110625.

Taking the first step towards implementing the correct spec-compliant painting
model for the new multi-column layout. Add code to make sure the paginated
bits in RenderLayer get propagated correctly for the new layout.

The main difference in bit propagation between the old multicolumn layout and
the new multicolumn layout is that the old layout only tracked pagination "roots",
since it has a broken layer painting model. It incorrectly treats a pagination
"root" inside a multi-column block as establishing a stacking context in each
column. This is incorrect.

The new multicolumn code is going to paint each paginated layer individually,
and this means the bits can't just be about roots. Instead they have to be set
on every layer that may be split across columns, without regard for whether parent
layers are also split or not. This is actually a simpler bit propagation model
than the old code.

Reviewed by Beth Dakin.

* rendering/RenderLayer.cpp:
(WebCore::RenderLayer::updateLayerPositions):
Add code to set the m_isPaginated bit for in-flow RenderFlowThreads. This includes
only RenderMultiColumnFlowThread for now, but eventually it will also include
RenderPageFlowThread.

(WebCore::RenderLayer::useRegionBasedColumns):
Add a helper method to check if region-based columns are turned on, since we're
not attempting to alter the old column pagination model.

(WebCore::RenderLayer::updatePagination):
Added the code to propagate bits under the new model. Whether or not you're paginated
just depends on whether or not your enclosing layer along the containing block chain
is paginated.

(WebCore::RenderLayer::paintList):
(WebCore::RenderLayer::hitTestList):
Add code that avoids the old multi-column pagination model when the new column model
is enabled. For now there is no code in the new model that does anything with the bits,
but we're avoiding calling the old code, since it is broken for the new columns.

* rendering/RenderLayer.h:
(RenderLayer):
Added the useRegionBasedColumns method to check the Setting.

* rendering/RenderMultiColumnFlowThread.h:
* rendering/RenderObject.h:
(WebCore::RenderObject::isInFlowRenderFlowThread):
(RenderObject):
Implement isInFlowRenderFlowThread, a virtual function for asking if a renderer is
an in-flow RenderFlowThread, which means that it is going to need paginated layers
because the flow thread has to split across "pages" that do not establish stacking
contexts.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143757 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agotext-overflow:ellipsis is not applied when the block contains nested blocks
commit-queue@webkit.org [Fri, 22 Feb 2013 18:58:19 +0000 (18:58 +0000)]
text-overflow:ellipsis is not applied when the block contains nested blocks
https://bugs.webkit.org/show_bug.cgi?id=101879

Patch by Bem Jones-Bey <bjonesbe@adobe.com> on 2013-02-22
Reviewed by Eric Seidel.

Source/WebCore:

Since text-overflow is not an inherited property, the anonymous blocks
don't know they should ellipse. This causes the anonymous blocks
created in the nested block situation to check their parent to find
out if they should ellipse.

Test: fast/css/text-overflow-ellipsis-anonymous-blocks.html

* rendering/RenderBlockLineLayout.cpp:
(WebCore::RenderBlock::layoutInlineChildren):

LayoutTests:

Use the test from the specification that is supplied as part of the
bug.

* fast/css/text-overflow-ellipsis-anonymous-blocks-expected.html: Added.
* fast/css/text-overflow-ellipsis-anonymous-blocks.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143754 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoREGRESSION(r143664, r143681): http/tests/security/feed-urls-from-remote.html fails
eric@webkit.org [Fri, 22 Feb 2013 18:50:39 +0000 (18:50 +0000)]
REGRESSION(r143664, r143681): http/tests/security/feed-urls-from-remote.html fails
https://bugs.webkit.org/show_bug.cgi?id=110554

Reviewed by Adam Barth.

Will fix http/tests/security/feed-urls-from-remote.html for Mac WK1.

* loader/DocumentLoader.cpp:
(WebCore::DocumentLoader::isLoading):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143753 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoBuild fix after r143637.
mrowe@apple.com [Fri, 22 Feb 2013 18:21:42 +0000 (18:21 +0000)]
Build fix after r143637.

* WebCore.exp.in: Unconditionally export ScriptController::javaScriptContext.
* bindings/js/ScriptController.h: Unconditionally define javaScriptContext for Mac builds.
* bindings/js/ScriptControllerMac.mm:
(WebCore::ScriptController::javaScriptContext): Move the #if inside the function, and return 0 when false.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143750 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[TextAutosizing] Refactoring to eliminate boolean parameter.
commit-queue@webkit.org [Fri, 22 Feb 2013 18:20:19 +0000 (18:20 +0000)]
[TextAutosizing] Refactoring to eliminate boolean parameter.
https://bugs.webkit.org/show_bug.cgi?id=110490

Patch by Anton Vayvod <avayvod@chromium.org> on 2013-02-22
Reviewed by Julien Chaffraix.

A follow-up to the recent change that introduced a boolean parameter to
processClusterInternal method of TextAutosizer. Boolean parameters are discouraged by the
WebKit style guide. See http://trac.webkit.org/changeset/142866

Refactoring so no new tests.

* rendering/TextAutosizer.cpp:
(WebCore::TextAutosizer::clusterMultiplier):

        Calculates the font size multiplier for the specified cluster.

(WebCore::TextAutosizer::processClusterInternal):

        Accepts the font size multiplier instead of |shouldBeAutosized|.

(WebCore::TextAutosizer::processCluster):
(WebCore::TextAutosizer::processCompositeCluster):

        Both methods above now calculate the multiplier and then pass it to
        processClusterInternal.

* rendering/TextAutosizer.h:

        Updated method declarations.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143749 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[BlackBerry] Use KURL::protocolIsInHTTPFamily instead of KURL::protocolInHTTPFamily
commit-queue@webkit.org [Fri, 22 Feb 2013 18:13:31 +0000 (18:13 +0000)]
[BlackBerry] Use KURL::protocolIsInHTTPFamily instead of KURL::protocolInHTTPFamily
https://bugs.webkit.org/show_bug.cgi?id=110581

Patch by Carlos Garcia Campos <cargarcia@rim.com> on 2013-02-22
Reviewed by Rob Buis.

KURL::protocolInHTTPFamily() was removed in r109670.

* platform/network/blackberry/ResourceRequestBlackBerry.cpp:
(WebCore::ResourceRequest::clearHTTPContentLength):
(WebCore::ResourceRequest::clearHTTPContentType):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143748 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed build-fix after r143744.
allan.jensen@digia.com [Fri, 22 Feb 2013 18:07:36 +0000 (18:07 +0000)]
Unreviewed build-fix after r143744.

* dom/Document.cpp:
(WebCore::Document::updateHoverActiveState):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143747 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoWeb Inspector: [Canvas] throttle replayTraceLog messages to backend
aandrey@chromium.org [Fri, 22 Feb 2013 17:56:37 +0000 (17:56 +0000)]
Web Inspector: [Canvas] throttle replayTraceLog messages to backend
https://bugs.webkit.org/show_bug.cgi?id=110591

Reviewed by Pavel Feldman.

The replayTraceLog message may be quite expensive. So, while such a request is processed by the
backend, user may well generate a dozen of new ones simply by walking through the DataGrid.
We should actually send to the backend only the last event from the queue.

* inspector/front-end/CanvasProfileView.js:
(WebInspector.CanvasProfileView.prototype._replayTraceLog.didReplayTraceLog):
(WebInspector.CanvasProfileView.prototype._replayTraceLog):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143745 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoMultiple Layout Test crashes (ASSERT) on chromium linux debug after r143727
allan.jensen@digia.com [Fri, 22 Feb 2013 17:53:07 +0000 (17:53 +0000)]
Multiple Layout Test crashes (ASSERT) on chromium linux debug after r143727
https://bugs.webkit.org/show_bug.cgi?id=110609

Reviewed by Antonio Gomes.

Revert r128869. InnerNode may now again be from a child document,
and should be handled and not asserted against.

* dom/Document.cpp:
(WebCore::Document::updateHoverActiveState):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143744 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoElement: Make updateName/updateId/updateLabel private.
akling@apple.com [Fri, 22 Feb 2013 17:47:51 +0000 (17:47 +0000)]
Element: Make updateName/updateId/updateLabel private.
<http://webkit.org/b/110603>

Reviewed by Antti Koivisto.

These methods are only called from inside Element so make them private and
move them to the .cpp file.

* dom/Element.cpp:
(WebCore::Element::updateName):
(WebCore::Element::updateId):
* dom/Element.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143743 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoRoll ANGLE to r1833
zmo@google.com [Fri, 22 Feb 2013 17:42:21 +0000 (17:42 +0000)]
Roll ANGLE to r1833
https://bugs.webkit.org/show_bug.cgi?id=110513

Reviewed by Kenneth Russell.

Source/ThirdParty/ANGLE:

Upstream ANGLE r1833 includes webkit side local fixes
1) r139758 build fix for -Wshorten-64-to-32
2) r139665 build fix for uninitialized variables

After this roll, WebKit side ANGLE is the same as upstream ANGLE r1833, except:
1) include/GLSLANG/ShaderLang.h, where "KHR" is removed from include path.

With this roll, we have EXT_draw_buffers support.

* ANGLE.xcodeproj/project.pbxproj:
* Target.pri:
* include/GLSLANG/ShaderLang.h:
* src/common/debug.cpp:
* src/common/system.h: Added.
* src/common/version.h:
* src/compiler/ArrayBoundsClamper.cpp: Removed.
* src/compiler/ArrayBoundsClamper.h: Removed.
* src/compiler/Compiler.cpp:
(TCompiler::TCompiler):
(TCompiler::Init):
(TCompiler::compile):
(TCompiler::getArrayBoundsClamper):
(TCompiler::getArrayIndexClampingStrategy):
(TCompiler::getBuiltInFunctionEmulator):
* src/compiler/DetectDiscontinuity.cpp:
(sh::DetectLoopDiscontinuity::traverse):
(sh::DetectLoopDiscontinuity::visitLoop):
(sh):
(sh::DetectLoopDiscontinuity::visitBranch):
* src/compiler/DetectDiscontinuity.h:
(DetectLoopDiscontinuity):
* src/compiler/Initialize.cpp:
(InitExtensionBehavior):
* src/compiler/Intermediate.cpp:
(TIntermConstantUnion::fold):
(TIntermTraverser::hash):
* src/compiler/MapLongVariableNames.cpp:
(LongNameMap::Size):
(MapLongVariableNames::mapGlobalLongName):
* src/compiler/MapLongVariableNames.h:
(LongNameMap):
* src/compiler/OutputESSL.cpp:
(TOutputESSL::TOutputESSL):
* src/compiler/OutputESSL.h:
(TOutputESSL):
* src/compiler/OutputGLSL.cpp:
(TOutputGLSL::TOutputGLSL):
* src/compiler/OutputGLSL.h:
(TOutputGLSL):
* src/compiler/OutputGLSLBase.cpp:
(TOutputGLSLBase::TOutputGLSLBase):
(TOutputGLSLBase::writeConstantUnion):
(TOutputGLSLBase::visitBinary):
* src/compiler/OutputGLSLBase.h:
(TOutputGLSLBase):
* src/compiler/OutputHLSL.cpp:
(sh::OutputHLSL::output):
(sh::OutputHLSL::visitLoop):
(sh::OutputHLSL::writeConstantUnion):
* src/compiler/ParseHelper.cpp:
(TParseContext::constructorErrorCheck):
(PaParseStrings):
* src/compiler/ParseHelper.h:
* src/compiler/ShHandle.h:
(TCompiler):
* src/compiler/ShaderLang.cpp:
(checkActiveUniformAndAttribMaxLengths):
(checkMappedNameMaxLength):
(getVariableInfo):
(ShInitBuiltInResources):
(ShCompile):
(ShGetInfo):
(ShGetActiveAttrib):
(ShGetActiveUniform):
(ShGetNameHashingEntry):
* src/compiler/SymbolTable.h:
(TFunction::getParamCount):
(TFunction::getParam):
(TSymbolTable::setDefaultPrecision):
* src/compiler/TranslatorESSL.cpp:
(TranslatorESSL::translate):
* src/compiler/TranslatorGLSL.cpp:
(TranslatorGLSL::translate):
* src/compiler/ValidateLimitations.cpp:
(ValidateLimitations::validateFunctionCall):
* src/compiler/VariablePacker.cpp:
(VariablePacker::GetNumRows):
* src/compiler/glslang.h:
* src/compiler/glslang.l:
* src/compiler/glslang.y:
* src/compiler/glslang_lex.cpp:
(string_input):
(glslang_scan):
* src/compiler/glslang_tab.cpp:
* src/compiler/preprocessor/ExpressionParser.cpp:
* src/compiler/preprocessor/ExpressionParser.y:
* src/compiler/preprocessor/Input.cpp:
(pp::Input::Input):
(pp::Input::read):
* src/compiler/preprocessor/Input.h:
(Input):
(pp::Input::count):
(pp::Input::string):
(pp::Input::length):
(Location):
* src/compiler/preprocessor/Preprocessor.cpp:
(pp::Preprocessor::init):
* src/compiler/preprocessor/Preprocessor.h:
(Preprocessor):
* src/compiler/preprocessor/Tokenizer.cpp:
(pp::Tokenizer::init):
* src/compiler/preprocessor/Tokenizer.h:
(Tokenizer):
* src/compiler/preprocessor/Tokenizer.l:
* src/third_party: Added.
* src/third_party/compiler: Added.
* src/third_party/compiler/ArrayBoundsClamper.cpp: Added.
(ArrayBoundsClamper::ArrayBoundsClamper):
(ArrayBoundsClamper::SetClampingStrategy):
(ArrayBoundsClamper::MarkIndirectArrayBoundsForClamping):
(ArrayBoundsClamper::OutputClampingFunctionDefinition):
* src/third_party/compiler/ArrayBoundsClamper.h: Added.
(ArrayBoundsClamper):
(ArrayBoundsClamper::Cleanup):
(ArrayBoundsClamper::GetArrayBoundsClampDefinitionNeeded):
(ArrayBoundsClamper::SetArrayBoundsClampDefinitionNeeded):
* src/third_party/compiler/LICENSE: Added.
* src/third_party/compiler/README.angle: Added.

Source/WebCore:

* CMakeLists.txt:
* GNUmakefile.list.am:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143742 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed, rolling out r143734.
commit-queue@webkit.org [Fri, 22 Feb 2013 17:28:33 +0000 (17:28 +0000)]
Unreviewed, rolling out r143734.
http://trac.webkit.org/changeset/143734
https://bugs.webkit.org/show_bug.cgi?id=110615

Causes linux build failures. (Requested by vollick on
#webkit).

Patch by Sheriff Bot <webkit.review.bot@gmail.com> on 2013-02-22

* public/WebKit.h:
(WebKit):
* src/WebKit.cpp:
(WebKit::initialize):
(WebKit::initializeWithoutV8):
(WebKit::webKitPlatformSupport):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143741 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[BlackBerry] Rename first/second to key/value in HashMap iterators
commit-queue@webkit.org [Fri, 22 Feb 2013 17:26:04 +0000 (17:26 +0000)]
[BlackBerry] Rename first/second to key/value in HashMap iterators
https://bugs.webkit.org/show_bug.cgi?id=110577

Patch by Carlos Garcia Campos <cargarcia@rim.com> on 2013-02-22
Reviewed by Rob Buis.

Iterators were renamed to key/value in r130612.

Source/WebCore:

* platform/blackberry/AuthenticationChallengeManager.cpp:
(WebCore::AuthenticationChallengeManager::pageVisibilityChanged):
* platform/network/blackberry/NetworkJob.cpp:
(WebCore::NetworkJob::handleNotifyMultipartHeaderReceived):

Source/WebKit/blackberry:

* WebKitSupport/NotificationManager.cpp:
(BlackBerry::WebKit::NotificationManager::show):
(BlackBerry::WebKit::NotificationManager::clearNotifications):
(BlackBerry::WebKit::NotificationManager::removeNotificationFromContextMap):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143740 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[WTR] Do not dump information about empty subframes.
rakuco@webkit.org [Fri, 22 Feb 2013 17:23:54 +0000 (17:23 +0000)]
[WTR] Do not dump information about empty subframes.
https://bugs.webkit.org/show_bug.cgi?id=110585

Reviewed by Simon Fraser.

Tools:

Follow DumpRenderTree more closely by skipping empty frames
(frames which have no document). We are not really interested in
showing anything about them.

* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::dumpDescendantFramesText): Do not print subframe names when they
have no associated Document.

LayoutTests:

Remove some platform-specific expectations now that the results match
the base expectations.

* platform/efl/TestExpectations: Unskip tests that are now passing.
* platform/mac-wk2/http/tests/security/cross-origin-xsl-BLOCKED-expected.txt: Removed.
* platform/mac-wk2/http/tests/security/cross-origin-xsl-redirect-BLOCKED-expected.txt: Removed.
* platform/qt-5.0-wk2/http/tests/security/cross-origin-xsl-BLOCKED-expected.txt: Removed.
* platform/qt-5.0-wk2/http/tests/security/cross-origin-xsl-redirect-BLOCKED-expected.txt: Removed.
* platform/win-wk2/http/tests/security/cross-origin-xsl-BLOCKED-expected.txt: Removed.
* platform/win-wk2/http/tests/security/cross-origin-xsl-redirect-BLOCKED-expected.txt: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143739 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed, rolling out r143678 and r143702.
commit-queue@webkit.org [Fri, 22 Feb 2013 16:55:23 +0000 (16:55 +0000)]
Unreviewed, rolling out r143678 and r143702.
http://trac.webkit.org/changeset/143678
http://trac.webkit.org/changeset/143702
https://bugs.webkit.org/show_bug.cgi?id=110613

Breaks fast/loader/stateobjects/state-url-sets-links-
visited.html in Chromium (Requested by abarth on #webkit).

Patch by Sheriff Bot <webkit.review.bot@gmail.com> on 2013-02-22

* fast/js/resources/js-test-pre.js:
* fast/loader/stateobjects/state-url-sets-links-visited-expected.txt:
* fast/loader/stateobjects/state-url-sets-links-visited.html:
* platform/chromium-win/fast/loader/stateobjects/state-url-sets-links-visited-expected.txt: Copied from LayoutTests/fast/loader/stateobjects/state-url-sets-links-visited-expected.txt.
* platform/chromium/TestExpectations:
* platform/win/fast/loader/stateobjects/state-url-sets-links-visited-expected.txt: Copied from LayoutTests/fast/loader/stateobjects/state-url-sets-links-visited-expected.txt.
* platform/wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143738 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[V8] Add comments to checkTypeOrDieTrying()
haraken@chromium.org [Fri, 22 Feb 2013 16:54:03 +0000 (16:54 +0000)]
[V8] Add comments to checkTypeOrDieTrying()
https://bugs.webkit.org/show_bug.cgi?id=110553

Reviewed by Adam Barth.

It is mysterious how the check is helpful. We need comments.

No tests. Just added comments.

* bindings/scripts/CodeGeneratorV8.pm:
(GenerateImplementation):
* bindings/v8/ScriptWrappable.h:
(ScriptWrappable):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143737 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[BlackBerry] Remove ProtocolHandler methods from ChromeClientBlackBerry
commit-queue@webkit.org [Fri, 22 Feb 2013 16:51:50 +0000 (16:51 +0000)]
[BlackBerry] Remove ProtocolHandler methods from ChromeClientBlackBerry
https://bugs.webkit.org/show_bug.cgi?id=110567

Patch by Carlos Garcia Campos <cargarcia@rim.com> on 2013-02-22
Reviewed by Rob Buis.

They were removed from ChromeClient in r122810 and are now
implemented in NavigatorContentUtilsClientBlackBerry since r131056.

* WebCoreSupport/ChromeClientBlackBerry.cpp:
* WebCoreSupport/ChromeClientBlackBerry.h:
(ChromeClientBlackBerry):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143736 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Chromium] Add support for emulating legacy Android WebView 'setInitialScale' method
mnaganov@chromium.org [Fri, 22 Feb 2013 16:46:55 +0000 (16:46 +0000)]
[Chromium] Add support for emulating legacy Android WebView 'setInitialScale' method
https://bugs.webkit.org/show_bug.cgi?id=109946

Adding a WebView method for permanently setting initial page scale.
This override has higher priority than any calculated page scale
and viewport meta tag value.

Also, this patch eliminates a dubious "fixed layout enabled, viewport disabled" mode
previously used by Android WebView.

Reviewed by Adam Barth.

* public/WebView.h:
(WebView):
* src/ChromeClientImpl.cpp:
(WebKit::ChromeClientImpl::dispatchViewportPropertiesDidChange):
* src/WebViewImpl.cpp:
(WebKit::WebViewImpl::WebViewImpl):
(WebKit::WebViewImpl::setInitialPageScaleOverride):
(WebKit):
(WebKit::WebViewImpl::computePageScaleFactorLimits):
* src/WebViewImpl.h:
* tests/WebFrameTest.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143735 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Chromium] WebKit::initialize should take a Platform* now that WebKitPlatformSupport...
pilgrim@chromium.org [Fri, 22 Feb 2013 16:41:20 +0000 (16:41 +0000)]
[Chromium] WebKit::initialize should take a Platform* now that WebKitPlatformSupport is empty
https://bugs.webkit.org/show_bug.cgi?id=110605

Reviewed by Adam Barth.

Part of a larger refactoring series; see tracking bug 82948.

* public/WebKit.h:
(WebKit):
* src/WebKit.cpp:
(WebKit::initialize):
(WebKit::initializeWithoutV8):
(WebKit::webKitPlatformSupport):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143734 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago32-bit build fix after r143706
jberlin@webkit.org [Fri, 22 Feb 2013 16:33:54 +0000 (16:33 +0000)]
32-bit build fix after r143706

* fileapi/FileReaderLoader.cpp:
(WebCore::FileReaderLoader::didReceiveData):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143733 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Qt] Unreviewed gardening. Added platform specific expected files after 143626.
kadam@inf.u-szeged.hu [Fri, 22 Feb 2013 16:24:48 +0000 (16:24 +0000)]
[Qt] Unreviewed gardening. Added platform specific expected files after 143626.

* platform/qt-5.0-wk2/compositing/layer-creation/fixed-position-change-out-of-view-in-view-expected.txt:
* platform/qt-5.0-wk2/compositing/rtl/rtl-iframe-absolute-overflow-expected.txt:
* platform/qt-5.0-wk2/compositing/rtl/rtl-iframe-absolute-overflow-scrolled-expected.txt:
* platform/qt-5.0-wk2/compositing/rtl/rtl-iframe-fixed-overflow-expected.txt:
* platform/qt-5.0-wk2/compositing/rtl/rtl-iframe-fixed-overflow-scrolled-expected.txt:
* platform/qt/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143732 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoWeb Inspector: [Canvas] do no show a single frame node in data grid
aandrey@chromium.org [Fri, 22 Feb 2013 16:17:31 +0000 (16:17 +0000)]
Web Inspector: [Canvas] do no show a single frame node in data grid
https://bugs.webkit.org/show_bug.cgi?id=110473

Reviewed by Pavel Feldman.

In a single canvas frame capture mode do not show the single "Frame #1" node.
Drive-by: Fix a wrong "this" pointer.
* inspector/front-end/CanvasProfileView.js:
(WebInspector.CanvasProfileView.prototype._didReceiveTraceLog):
(WebInspector.CanvasProfileView.prototype.appendDrawCallGroup):
(WebInspector.CanvasProfileView.prototype._flattenSingleFrameNode):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143731 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoMultiple Layout Test crashes (ASSERT) on chromium linux debug after r143727
allan.jensen@digia.com [Fri, 22 Feb 2013 16:12:40 +0000 (16:12 +0000)]
Multiple Layout Test crashes (ASSERT) on chromium linux debug after r143727
https://bugs.webkit.org/show_bug.cgi?id=110609

Reviewed by Antonio Gomes.

Point based hit tests should not append but assign HitTestResults.

* rendering/RenderFrameBase.cpp:
(WebCore::RenderFrameBase::nodeAtPoint):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143730 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoWeb Inspector: show source location after drawer views
commit-queue@webkit.org [Fri, 22 Feb 2013 15:54:12 +0000 (15:54 +0000)]
Web Inspector: show source location after drawer views
https://bugs.webkit.org/show_bug.cgi?id=110156

Patch by Andrey Lushnikov <lushnikov@chromium.org> on 2013-02-22
Reviewed by Pavel Feldman.

- Add statusBarText method
- Add #drawer-view-anchor, which determines layout position of
drawerView
- Fix layout in #panel-status-bar to avoid floating elements.

No new tests.

* inspector/front-end/Panel.js:
(WebInspector.Panel.prototype.wasShown): Add statusBarItems before
drawer-view-anchor and statusBarText after.
(WebInspector.Panel.prototype.willHide): Remove statusBarText from DOM in the
way it's done for statusBarItems
* inspector/front-end/ScriptsPanel.js:
(WebInspector.ScriptsPanel.prototype.statusBarText): Added.
* inspector/front-end/SourceFrame.js:
(WebInspector.SourceFrame.prototype.statusBarText): Added.
(WebInspector.SourceFrame.prototype.statusBarItems):
* inspector/front-end/TimelinePanel.js:
(WebInspector.TimelinePanel.prototype.get statusBarItems):
* inspector/front-end/View.js:
(WebInspector.View.prototype.statusBarText):
* inspector/front-end/inspector.css:
(#drawer-view-anchor):
(.source-frame-cursor-position):
* inspector/front-end/inspector.html:
* inspector/front-end/inspector.js:
(WebInspector.showViewInDrawer):
* inspector/front-end/timelinePanel.css:
(.timeline-records-stats):
(.timeline-records-stats-container):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143729 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed EFL gardening.
sudarsana.nagineni@linux.intel.com [Fri, 22 Feb 2013 15:31:39 +0000 (15:31 +0000)]
Unreviewed EFL gardening.

Skip failing tests on EFL bots.

* platform/efl/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143728 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoAllow child-frame content in hit-tests.
allan.jensen@digia.com [Fri, 22 Feb 2013 14:46:35 +0000 (14:46 +0000)]
Allow child-frame content in hit-tests.
https://bugs.webkit.org/show_bug.cgi?id=95204

Reviewed by Julien Chaffraix.

Source/WebCore:

Refactors how EventHandler::hitTestResultAtPoint handles child-frame content,
it is now handled by the hit test itself controlled by the AllowChildFrameContent
flag in HitTestRequest.

This means that area-based hit-tests can now return elements from all the child frames
they intersect instead of just the one frame containing the center point. The improved
results from area-based hit-tests will among other things also improve touch adjustment
near frame boundaries.

Tests: fast/dom/nodesFromRect/nodesFromRect-child-frame-content.html

* page/EventHandler.cpp:
(WebCore::EventHandler::hitTestResultAtPoint):
    Recursion into child-frames have been moved to RenderFrameBase::nodeAtPoint, so
    now hitTestResultAtPoint just needs to set AllowChildFrameContent.
* page/TouchAdjustment.cpp:
(WebCore::TouchAdjustment::parentShadowHostOrOwner):
    New function to iterate up across frame boundaries.
(WebCore::TouchAdjustment::compileSubtargetList):
    We need to iterate up across frame boundaries to avoid iframes competing with their
    own content for touch adjustment.
* rendering/HitTestRequest.h:
(WebCore::HitTestRequest::allowsChildFrameContent):
(WebCore::HitTestRequest::isChildFrameHitTest):
* rendering/HitTestResult.cpp:
(WebCore::HitTestResult::HitTestResult):
(WebCore::HitTestResult::operator=):
(WebCore::HitTestResult::append):
(WebCore::HitTestResult::dictationAlternatives):
* rendering/HitTestResult.h:
(WebCore::HitTestResult::pointInMainFrame):
(WebCore::HitTestResult::pointInInnerNodeFrame):
(HitTestResult):
    m_hitTestLocation is now in main frame coordinates, which make m_pointInMainFrame
    unnecessary, but requires the introduction of m_pointInInnerFrame, to remember
    the coordinates of inner-node in its own frame.
* rendering/RenderFrameBase.cpp:
(WebCore::RenderFrameBase::nodeAtPoint):
    The recursion into child-frames is now handled here instead of in hitTestResultAtPoint, this
    allows us to recurse into multiple frames, instead of just one.
* rendering/RenderFrameBase.h:
(RenderFrameBase):
* rendering/RenderLayer.cpp:
(WebCore::RenderLayer::hitTest):
    RenderLayer should not lie about being hit if the request is child-frame request.
* testing/Internals.cpp:
(WebCore::Internals::nodesFromRect):
* testing/Internals.h:
(Internals):
* testing/Internals.idl:
    Extended so nodesFromRect with child-frame content can be tested.

LayoutTests:

A new tests for nodesFromRect that returns result from child frames.

* fast/dom/nodesFromRect/nodesFromRect-child-frame-content-expected.txt: Added.
* fast/dom/nodesFromRect/nodesFromRect-child-frame-content.html: Added.
* fast/dom/nodesFromRect/nodesFromRect-continuation-crash.html:
* fast/dom/nodesFromRect/resources/child-frame.html: Added.
* fast/dom/nodesFromRect/resources/nodesFromRect.js:
(check):
(checkShadowContent):
(nodesFromRectAsString):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143727 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoShareableElementData should use zero-length array for storage.
akling@apple.com [Fri, 22 Feb 2013 14:39:42 +0000 (14:39 +0000)]
ShareableElementData should use zero-length array for storage.
<http://webkit.org/b/109959>

Reviewed by Anders Carlsson.

Use a zero-length Attribute array instead of always casting from void* to an array.
It was done this way originally because I didn't know we could sidestep the MSVC
build error with some #pragma hackery and a default constructor for Attribute.

* dom/Attribute.cpp:
(WebCore::Attribute::Attribute):
* dom/DocumentSharedObjectPool.cpp:
(WebCore::DocumentSharedObjectPool::cachedShareableElementDataWithAttributes):
* dom/Element.cpp:
(WebCore::sizeForShareableElementDataWithAttributeCount):
(WebCore::ShareableElementData::ShareableElementData):
(WebCore::ShareableElementData::~ShareableElementData):
(WebCore::UniqueElementData::UniqueElementData):
* dom/Element.h:
(ShareableElementData):
(WebCore::ElementData::attributeItem):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143726 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoWeb Inspector: [Chromium] add a browser test for frames on timeline
caseq@chromium.org [Fri, 22 Feb 2013 13:52:46 +0000 (13:52 +0000)]
Web Inspector: [Chromium] add a browser test for frames on timeline
https://bugs.webkit.org/show_bug.cgi?id=110592

- factor out timeline recording logic from testPageOverlayUpdate for reuse;
- record timeline while running a simple DOM-based animation;
- assure we have frames and Style Recalc/Layout/Paint events in between.

Reviewed by Yury Semikhatsky.

* src/js/Tests.js:
(.TestSuite.prototype.assertHasKey):
(.TestSuite.prototype.testTimelineFrames.step1):
(.TestSuite.prototype.testTimelineFrames.onTimelineRecorded):
(.TestSuite.prototype.testTimelineFrames):
(.TestSuite.prototype.testPageOverlayUpdate.step4):
(.TestSuite.prototype.testPageOverlayUpdate.onTimelineRecorded):
(.TestSuite.prototype.recordTimeline.addRecord):
(.TestSuite.prototype.recordTimeline.innerAddRecord):
(.TestSuite.prototype.recordTimeline.done):
(.TestSuite.prototype.recordTimeline):
(.TestSuite.prototype.stopTimeline):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143725 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoRemove unused declaration: Element::removeCachedHTMLCollection()
akling@apple.com [Fri, 22 Feb 2013 13:50:51 +0000 (13:50 +0000)]
Remove unused declaration: Element::removeCachedHTMLCollection()

Not reviewed by Antti Koivisto.

* dom/Element.h:
(Element):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143724 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoAdd scroll view for new calendar picker
keishi@webkit.org [Fri, 22 Feb 2013 13:44:06 +0000 (13:44 +0000)]
Add scroll view for new calendar picker
https://bugs.webkit.org/show_bug.cgi?id=110137

Reviewed by Kent Tamura.

Adding a scroll view class as part of the new calendar picker (Bug 109439).

No new tests. Code not yet used.

* Resources/pagepopups/calendarPicker.js:
(View):
(View.prototype.offsetRelativeTo): Returns the offset position of this view relative to the given ancestor element.
(View.prototype.attachTo): Attaches view to a node or view.
(View.prototype.bindCallbackMethods): Binds all methods starting with "on" to this.
(ScrollView): A custom scroll view that can contain extremely long content. CSS values have a limit. This can go beyond that.
(ScrollView.prototype.setWidth): Sets the view width.
(ScrollView.prototype.width):
(ScrollView.prototype.setHeight):Sets the view height.
(ScrollView.prototype.height):
(ScrollView.prototype.onScrollAnimatorStep): Callback for scroll animator step.
(ScrollView.prototype.scrollTo): Scrolls to a certain offset.
(ScrollView.prototype.scrollBy): Scrolls by a certain amount.
(ScrollView.prototype.contentOffset): Current content offset.
(ScrollView.prototype.onMouseWheel): Scroll with the mouse wheel.
(ScrollView.prototype.setContentOffset): Sets the content offset.
(ScrollView.prototype.contentPositionForContentOffset): Returns where the content element should be positioned.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143723 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoWeb Inspector: InspectorPageAgent::disable() should not update view metrics regardles...
apavlov@chromium.org [Fri, 22 Feb 2013 13:41:30 +0000 (13:41 +0000)]
Web Inspector: InspectorPageAgent::disable() should not update view metrics regardless of the override state
https://bugs.webkit.org/show_bug.cgi?id=110593

Reviewed by Vsevolod Vlasov.

InspectorPageAgent::disable() now checks if the device metrics are overridden before
telling the client to disable the override.

* inspector/InspectorPageAgent.cpp:
(WebCore::InspectorPageAgent::disable):
(WebCore::InspectorPageAgent::setDeviceMetricsOverride):
(WebCore::InspectorPageAgent::deviceMetricsChanged):
* inspector/InspectorPageAgent.h:
(InspectorPageAgent):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@143721 268f45cc-cd09-0410-ab3c-d52691b4dbfc