WebKit GIT trees - WebKit-https.git/log

Web Inspector: Move sidebar-specific styles to a separate file.
https://bugs.webkit.org/show_bug.cgi?id=111173

Patch by Vladislav Kaznacheev <kaznacheev@chromium.org> on 2013-03-06
Reviewed by Pavel Feldman.

* WebCore.gypi:
* inspector/front-end/BreakpointsSidebarPane.js:
(WebInspector.JavaScriptBreakpointsSidebarPane):
(WebInspector.EventListenerBreakpointsSidebarPane):
* inspector/front-end/NativeBreakpointsSidebarPane.js:
(WebInspector.NativeBreakpointsSidebarPane):
* inspector/front-end/SidebarPane.js:
(WebInspector.SidebarPaneTitle):
(WebInspector.SidebarPaneStack):
(WebInspector.SidebarTabbedPane):
* inspector/front-end/WebKit.qrc:
* inspector/front-end/breakpointsList.css: Added.
(.sidebar-pane > .body .breakpoint-condition):
(#breakpoint-condition-input):
(ol.breakpoint-list):
(.breakpoints-list-deactivated):
(.breakpoint-list li):
(.breakpoint-list li:hover):
(.breakpoint-list .checkbox-elem):
(.breakpoint-list .source-text):
(.sidebar-pane .breakpoint-hit):
(li.breakpoint-hit .breakpoint-hit-marker):
(.event-listener-breakpoints .event-category):
(.event-listener-breakpoints.properties-tree .children li):
(.event-listener-breakpoints .checkbox-elem):
* inspector/front-end/elementsPanel.css:
(.panel.elements .sidebar-pane-toolbar > select):
(.panel.elements .sidebar-pane-toolbar > select:hover):
(.panel.elements .sidebar-pane-toolbar > select:active):
(.panel.elements .sidebar-pane-toolbar > select.select-settings):
(.panel.elements .sidebar-pane-toolbar > select.select-filter):
(.panel.elements .sidebar-pane-toolbar > select > option, .panel.elements .sidebar-pane-toolbar > select > hr):
* inspector/front-end/inspector.css:
(.event-bar .event-properties):
* inspector/front-end/scriptsPanel.css:
(.workers-list > li):
(a.worker-item):
(a.worker-item:hover):
* inspector/front-end/sidebarPane.css: Added.
(.sidebar-pane):
(.sidebar-pane > .body):
(.sidebar-pane > .body .info):
(.sidebar-pane > .body .placard + .info):
(.sidebar-pane.visible > .body):
(.sidebar-pane .section .properties):
(.sidebar-tabbed-pane .tabbed-pane-header):
(.sidebar-pane-stack > .sidebar-pane.visible:nth-last-of-type(1)):
(.sidebar-pane-title):
(.sidebar-pane-title + .sidebar-pane-title, .pane:not(.visible) + .sidebar-pane-title, .sidebar-pane-title:first-of-type):
(.sidebar-pane-title:active):
(.sidebar-pane-title::before):
(.sidebar-pane-title.expanded::before):
(.sidebar-pane-toolbar):
(.sidebar-pane-toolbar > *):
(.sidebar-pane-toolbar > .pane-title-button):
(.sidebar-pane-toolbar > .pane-title-button:hover):
(.sidebar-pane-toolbar > .pane-title-button:active, .sidebar-pane-toolbar > .pane-title-button.toggled):
(.sidebar-pane-toolbar > .pane-title-button.add):
(.sidebar-pane-toolbar > .pane-title-button.element-state):
(.sidebar-pane-toolbar > .pane-title-button.refresh):
(.sidebar-pane-subtitle):
(body.platform-windows .sidebar-pane-subtitle):
(.sidebar-pane-subtitle input, .section .header input[type=checkbox]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144927 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: test that "importScript" is used only in module heads.
https://bugs.webkit.org/show_bug.cgi?id=111548

Patch by Andrey Lushnikov <lushnikov@chromium.org> on 2013-03-06
Reviewed by Pavel Feldman.

Test that output file after all inlines does not have any 'importScript(s)' statements.

* scripts/inline_js_imports.py:
(main):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144926 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: Combine style-related panes into a single tab when Elements panel is split horizontally.
https://bugs.webkit.org/show_bug.cgi?id=111550
Patch by Vladislav Kaznacheev <kaznacheev@chromium.org> on 2013-03-06
Reviewed by Pavel Feldman.

Grouping the Elements sidebar panes into two tabbed panes proved to be
confusing and inconvenient. Put all the Elements sidebar panes into a single tabbed pane.
Combined Styles, Computed Style and Metrics into a single pane.

* inspector/front-end/ElementsPanel.js:
(WebInspector.ElementsPanel.prototype._splitVertically):
(WebInspector.ElementsPanel.prototype.addExtensionSidebarPane):
* inspector/front-end/elementsPanel.css:
(.sidebar-pane.composite):
(.sidebar-pane.composite > .body):
(.sidebar-pane.composite .metrics):
(.sidebar-pane.composite .sidebar-pane-toolbar):
(.sidebar-pane.composite .sidebar-pane-toolbar > .sidebar-pane-subtitle):
(.sidebar-pane.composite .styles-section.read-only):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144925 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: support stable remote object ids
https://bugs.webkit.org/show_bug.cgi?id=109908

Reviewed by Pavel Feldman.

Source/WebCore:

Introduced a map from javascript object to its id. It guarantees
that object is assined id only once (unless it is explicitely
released). The map is per execution context and is stored in
InjectedScriptManager.

Test: inspector-protocol/persistent-id.html

* bindings/js/JSInjectedScriptHostCustom.cpp:
(WebCore::JSInjectedScriptHost::objectId):
(WebCore):
(WebCore::JSInjectedScriptHost::releaseObjectId):
* bindings/js/ScriptObject.h:
(WTF):
(WTF::ScriptObjectHash::hash):
(WTF::ScriptObjectHash::equal):
(ScriptObjectHash):
* bindings/v8/ScriptObject.h:
(WTF):
(WTF::ScriptObjectHash::hash):
(WTF::ScriptObjectHash::equal):
(ScriptObjectHash):
* bindings/v8/custom/V8InjectedScriptHostCustom.cpp:
(WebCore::V8InjectedScriptHost::objectIdMethodCustom):
(WebCore):
(WebCore::V8InjectedScriptHost::releaseObjectIdMethodCustom):
* inspector/InjectedScriptHost.cpp:
(WebCore::InjectedScriptHost::create):
(WebCore::InjectedScriptHost::InjectedScriptHost):
(WebCore::InjectedScriptHost::disconnect):
(WebCore::InjectedScriptHost::objectId):
(WebCore):
(WebCore::InjectedScriptHost::releaseObjectId):
* inspector/InjectedScriptHost.h:
(WebCore):
(InjectedScriptHost):
* inspector/InjectedScriptHost.idl:
* inspector/InjectedScriptManager.cpp:
(InjectedScriptManager::ObjectIdMap):
(WebCore::InjectedScriptManager::ObjectIdMap::ObjectIdMap):
(WebCore::InjectedScriptManager::ObjectIdMap::objectId):
(WebCore::InjectedScriptManager::ObjectIdMap::releaseObjectId):
(WebCore):
(WebCore::InjectedScriptManager::InjectedScriptManager):
(WebCore::InjectedScriptManager::discardInjectedScripts):
(WebCore::InjectedScriptManager::discardInjectedScriptsFor):
(WebCore::InjectedScriptManager::objectId):
(WebCore::InjectedScriptManager::releaseObjectId):
* inspector/InjectedScriptManager.h:
(InjectedScriptManager):
* inspector/InjectedScriptSource.js:
(.):

LayoutTests:

Test that same object will have same id if requested several times.

* inspector-protocol/persistent-id-expected.txt: Added.
* inspector-protocol/persistent-id.html: Added.
* inspector/console/command-line-api-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144924 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[BlackBerry] Use CachedResource::resourceBuffer() instead of CachedResource::data()
https://bugs.webkit.org/show_bug.cgi?id=111555

Patch by Carlos Garcia Campos <cgarcia@igalia.com> on 2013-03-06
Reviewed by Rob Buis.

CachedResource::data() was renamed to
CachedResource::resourceBuffer() in r130947.

* Api/WebPage.cpp:
(BlackBerry::WebKit::WebPagePrivate::webContext):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144923 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[BlackBerry] Remove some unneeded header includes
https://bugs.webkit.org/show_bug.cgi?id=111545

Patch by Carlos Garcia Campos <cgarcia@igalia.com> on 2013-03-06
Reviewed by Rob Buis.

* WebCoreSupport/CredentialTransformData.cpp:
* WebCoreSupport/FrameLoaderClientBlackBerry.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144922 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[BlackBerry] Use hasTagName instead of hasLocalName in CredentialTransformData
https://bugs.webkit.org/show_bug.cgi?id=111544

Patch by Carlos Garcia Campos <cgarcia@igalia.com> on 2013-03-06
Reviewed by Rob Buis.

hasLocalName() expects an AtomicString, but we are passing a
HTMLName which is a QualifiedName.

* WebCoreSupport/CredentialTransformData.cpp:
(WebCore::CredentialTransformData::findPasswordFormFields): Use
hasTagName() instead of hasLocalName().

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144921 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed EFL gardening.
https://bugs.webkit.org/show_bug.cgi?id=111571

Gardening; added accessibility failure tests.

Patch by Krzysztof Czech <k.czech@samsung.com> on 2013-03-06

* platform/efl-wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144920 268f45cc-cd09-0410-ab3c-d52691b4dbfc

REGRESSION(r144617): Wrong usage of overwritten variable in toDOMWindow (V8Binding.cpp)
https://bugs.webkit.org/show_bug.cgi?id=111565

Patch by Marja Hölttä <marja@chromium.org> on 2013-03-06
Reviewed by Jochen Eisinger.

In toDOMWindow, the variable "global" was first nuked, and then used again. This
change was introduced in r144617.

No new tests (no functional changes).

* bindings/v8/V8Binding.cpp:
(WebCore::toDOMWindow):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144919 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt] Illegal narrowing in tst_qwebhistory
https://bugs.webkit.org/show_bug.cgi?id=111562

Reviewed by Jocelyn Turcotte.

The bytearray in tst_QWebHistory::restoreIncompatibleVersion1 has integer
values higher than 127 assigned into an array of signed chars. This causes
compiler errors in C++11 since it assigns constants (>127) that does not
fit the destination (-128 to 127)

* tests/qwebhistory/tst_qwebhistory.cpp:
(tst_QWebHistory::restoreIncompatibleVersion1):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144918 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Typo in inline function in ByteOrder.h
https://bugs.webkit.org/show_bug.cgi?id=111473

Patch by Paweł Forysiuk <tuxator@o2.pl> on 2013-03-06
Reviewed by Benjamin Poulain.

* wtf/ByteOrder.h:
(ntohs):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144917 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: use regular eval instead of injectScript for front-end extension API
https://bugs.webkit.org/show_bug.cgi?id=111529

Reviewed by Pavel Feldman.

Removed InjectedScriptManager::injectScript and replaced all of it calls
except the actual script injection with ScriptController::executeScript.

* inspector/InjectedScriptManager.cpp:
(WebCore::InjectedScriptManager::injectedScriptFor):
* inspector/InjectedScriptManager.h:
(InjectedScriptManager):
* inspector/InspectorAgent.cpp:
(WebCore::InspectorAgent::didClearWindowObjectInWorld):
* inspector/InspectorPageAgent.cpp:
(WebCore::InspectorPageAgent::didClearWindowObjectInWorld):
* inspector/front-end/DOMAgent.js:
(WebInspector.DOMAgent.prototype._emulateTouchEventsChanged.get if):
* inspector/front-end/ExtensionAPI.js:
(buildExtensionAPIInjectedScript):
* inspector/front-end/ExtensionServer.js:
(WebInspector.ExtensionServer.prototype._onReload):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144916 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt][MiniBrowser] Touch mocking broken for chained getures
https://bugs.webkit.org/show_bug.cgi?id=111430

Reviewed by Jocelyn Turcotte.

Make it possible again to use touch mocking to test gesture chains
like pan->pinch->pan->pinch without lifting the primary "finger".
For this to work we need to track the pressed mouse buttons and
only synthesize touch release events when the ctrl key is released
if no mouse button is pressed. Else the pressed mouse button(s)
should result in a stationary touch point in a touch update event.
Since it has been fixed in Qt 5.0.1 use QGuiApplication::keyboardModifiers()
to check for the state of the ctrl key modifier instead of getting
it from the mouse events.

* MiniBrowser/qt/MiniBrowserApplication.cpp:
(MiniBrowserApplication::MiniBrowserApplication):
(MiniBrowserApplication::notify):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144915 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt][WK2] Fix the Mac build after r144787
https://bugs.webkit.org/show_bug.cgi?id=111569

Reviewed by Csaba Osztrogonác.

Use uint_64 explicitly instead of size_t.

* Shared/CoordinatedGraphics/CoordinatedGraphicsArgumentCoders.cpp:
(CoreIPC::::encode):
(CoreIPC::::decode):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144914 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt] Do not disable C++0x
https://bugs.webkit.org/show_bug.cgi?id=111559

Reviewed by Kenneth Rohde Christiansen.

* WebCore.pri:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144913 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remove the (dead) code for handling shorthands in StyleResolver / StyleBuilder
https://bugs.webkit.org/show_bug.cgi?id=111505

Reviewed by Alexis Menard.

The CSS parsing code expands shorthands (with the exception of 'font') to their longhands.
This means that all the code in StyleResolver / StyleBuilder to handle shorthands is basically
dead and shouldn't be there. The reason for it to be is that people didn't know about shorthands
expansion and thus copied existing code.

This change addresses both issues by removing the code and adding ASSERTs that should prevent people
from adding more dead code.

Refactoring, covered by existing tests.

* css/StylePropertyShorthand.h: Added isExpandedShorthand.
* css/StylePropertyShorthand.cpp:
(WebCore::isExpandedShorthand):
Added this function that ignores 'font' as it is the only shorthand that is no consistently expanded
to account for system fonts.

* css/StyleBuilder.cpp:
(WebCore::StyleBuilder::StyleBuilder):
Removed the shorthands.

* css/StyleBuilder.h:
(WebCore::StyleBuilder::setPropertyHandler):
Added ASSERTs that people don't add shorthand handling.

* css/StyleResolver.cpp:
(WebCore::isValidVisitedLinkProperty):
Removed all the shorthands (shouldn't be reachable unless the ASSERT in applyProperty
triggers).

(WebCore::StyleResolver::applyProperty):
Removed all shorthand handling code, moved the shorthands together and added some ASSERT
to ensure people don't abuse the code and re-add some of it back by mistake.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144912 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Rename WebCore/editing/visible_units.* to VisibleUnits.*
https://bugs.webkit.org/show_bug.cgi?id=111426

Reviewed by Alexey Proskuryakov.

Source/WebCore:

No new tests. Just a refactoring.

* CMakeLists.txt:
* GNUmakefile.list.am:
* Target.pri:
* WebCore.gypi:
* WebCore.vcproj/WebCore.vcproj:
* WebCore.vcxproj/WebCore.vcxproj:
* WebCore.vcxproj/WebCore.vcxproj.filters:
* WebCore.xcodeproj/project.pbxproj:
* accessibility/AccessibilityNodeObject.cpp:
* accessibility/AccessibilityObject.cpp:
* accessibility/AccessibilityRenderObject.cpp:
* accessibility/atk/WebKitAccessibleWrapperAtk.cpp:
* accessibility/mac/WebAccessibilityObjectWrapperBase.mm:
* accessibility/mac/WebAccessibilityObjectWrapperMac.mm:
* dom/Position.cpp:
* dom/Range.cpp:
* editing/AlternativeTextController.cpp:
* editing/ApplyBlockElementCommand.cpp:
* editing/ApplyStyleCommand.cpp:
* editing/CompositeEditCommand.cpp:
* editing/DeleteSelectionCommand.cpp:
* editing/EditingAllInOne.cpp:
* editing/EditingStyle.cpp:
* editing/Editor.cpp:
* editing/FormatBlockCommand.cpp:
* editing/FrameSelection.cpp:
* editing/IndentOutdentCommand.cpp:
* editing/InsertLineBreakCommand.cpp:
* editing/InsertListCommand.cpp:
* editing/InsertParagraphSeparatorCommand.cpp:
* editing/InsertTextCommand.cpp:
* editing/ReplaceSelectionCommand.cpp:
* editing/SurroundingText.cpp:
* editing/TextCheckingHelper.cpp:
* editing/TextIterator.cpp:
* editing/TypingCommand.cpp:
* editing/VisiblePosition.cpp:
* editing/VisibleSelection.cpp:
* editing/VisibleUnits.cpp: Copied from Source/WebCore/editing/visible_units.cpp.
* editing/VisibleUnits.h: Copied from Source/WebCore/editing/visible_units.h.
* editing/htmlediting.cpp:
* editing/markup.cpp:
* editing/visible_units.cpp: Removed.
* editing/visible_units.h: Removed.
* page/Frame.cpp:
* page/mac/FrameMac.mm:
* platform/win/EditorWin.cpp:

Source/WebKit/mac:

* WebView/WebFrame.mm:

Source/WebKit2:

* WebProcess/WebPage/mac/WebPageMac.mm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144911 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unused Structure property tables waste 14MB on Membuster.
<http://webkit.org/b/110854>
<rdar://problem/13292104>

Reviewed by Geoffrey Garen.

Turn PropertyTable into a GC object and have Structure drop unpinned tables when marking.
14 MB progression on Membuster3.

This time it should stick; I've been through all the tests with COLLECT_ON_EVERY_ALLOCATION.
The issue with the last version was that Structure::m_offset could be used uninitialized
when re-materializing a previously GC'd property table, causing some sanity checks to fail.

* CMakeLists.txt:
* GNUmakefile.list.am:
* JavaScriptCore.gypi:
* JavaScriptCore.vcxproj/JavaScriptCore.vcxproj:
* JavaScriptCore.xcodeproj/project.pbxproj:
* Target.pri:

    Added PropertyTable.cpp.

* runtime/PropertyTable.cpp: Added.
(JSC::PropertyTable::create):
(JSC::PropertyTable::clone):
(JSC::PropertyTable::PropertyTable):
(JSC::PropertyTable::destroy):
(JSC::PropertyTable::~PropertyTable):
(JSC::PropertyTable::visitChildren):

    Moved marking of property table values here from Structure::visitChildren().

* runtime/WriteBarrier.h:
(JSC::WriteBarrierBase::get):

    Move m_cell to a local before using it multiple times. This avoids a multiple-access race when
    Structure::checkOffsetConsistency() is used in assertions on the main thread while a marking thread
    zaps the property table.

* runtime/Structure.h:
(JSC::Structure::materializePropertyMapIfNecessary):
(JSC::Structure::materializePropertyMapIfNecessaryForPinning):
* runtime/StructureInlines.h:
(JSC::Structure::propertyTable):

    Added a getter for the Structure's PropertyTable that ASSERTs GC currently isn't active.
    Because GC can zap an unpinned property table at any time, it's not entirely safe to access it.
    Renamed the variable itself to m_propertyTableUnsafe to force call sites into explaining themselves.

(JSC::Structure::putWillGrowOutOfLineStorage):
(JSC::Structure::checkOffsetConsistency):

    Moved these out of Structure.h to break header dependency cycle between Structure/PropertyTable.

* runtime/Structure.cpp:
(JSC::Structure::visitChildren):

    Null out m_propertyTable if the table is unpinned. This'll cause the table to get GC'd.

(JSC::Structure::takePropertyTableOrCloneIfPinned):

    Added for setting up the property table in a new transition, this code is now shared between
    addPropertyTransition() and nonPropertyTransition().

* runtime/JSGlobalData.h:
* runtime/JSGlobalData.cpp:
(JSC::JSGlobalData::JSGlobalData):

    Add a global propertyTableStructure.

* runtime/PropertyMapHashTable.h:
(PropertyTable):
(JSC::PropertyTable::createStructure):
(JSC::PropertyTable::copy):

    Make PropertyTable a GC object.

* runtime/Structure.cpp:
(JSC::Structure::dumpStatistics):
(JSC::Structure::materializePropertyMap):
(JSC::Structure::despecifyDictionaryFunction):
(JSC::Structure::addPropertyTransition):
(JSC::Structure::changePrototypeTransition):
(JSC::Structure::despecifyFunctionTransition):
(JSC::Structure::attributeChangeTransition):
(JSC::Structure::toDictionaryTransition):
(JSC::Structure::sealTransition):
(JSC::Structure::freezeTransition):
(JSC::Structure::preventExtensionsTransition):
(JSC::Structure::nonPropertyTransition):
(JSC::Structure::isSealed):
(JSC::Structure::isFrozen):
(JSC::Structure::flattenDictionaryStructure):
(JSC::Structure::pin):
(JSC::Structure::copyPropertyTable):
(JSC::Structure::copyPropertyTableForPinning):
(JSC::Structure::get):
(JSC::Structure::despecifyFunction):
(JSC::Structure::despecifyAllFunctions):
(JSC::Structure::putSpecificValue):
(JSC::Structure::remove):
(JSC::Structure::createPropertyMap):
(JSC::Structure::getPropertyNamesFromStructure):
(JSC::Structure::checkConsistency):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144910 268f45cc-cd09-0410-ab3c-d52691b4dbfc

DataReference::vector() should be const.
https://bugs.webkit.org/show_bug.cgi?id=109928

Reviewed by Anders Carlsson.

* Platform/CoreIPC/DataReference.h:
(CoreIPC::DataReference::vector):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144909 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Do not add CSSPropertyBorderImage shorthand part of the property list when parsing CSS border property
https://bugs.webkit.org/show_bug.cgi?id=111481

Reviewed by Julien Chaffraix.

Source/WebCore:

After http://trac.webkit.org/changeset/135848 and more generally we do not
add shorthand directly into the property list used afterwards to build
StylePropertySet but we rather expand the shorthand to add the
longhands. CSSPropertyBorderImage is a shorthand and it should follow
that rule, so when we successfully parsed CSS border property and we
should reset border-image then we should reset the longhands too.

Test: LayoutTests/inspector/styles/styles-new-API-expected.txt which
reflects the fact that we add longhands instead of the shorthand. All
the other tests should not regress.

* css/CSSParser.cpp:
(WebCore::CSSParser::parseValue):

LayoutTests:

As now border-image is expanded when we sucessfully parsed a CSS border,
this test needs to be updated to reflect that the longhands are now
part of the declarated style.

* inspector/styles/styles-new-API-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144908 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Accelerated overflow scrolling for Coordinated Graphics.
https://bugs.webkit.org/show_bug.cgi?id=110323

Source/WebCore:

Patch by Luiz Agostini <luiz.agostini@nokia.com> on 2013-03-06
Reviewed by Noam Rosenthal.

A scroll offset is applied to individual layers on UI process and the
web process is notified assynchronously.
The new API WKCoordinatedScene may be used to scroll layers in the UI process.

Tests will be added on bug #111140.

Scrolling coordinator is used to provide a ScrollableArea to CoordinatedGraphicsLayer.

* page/scrolling/coordinatedgraphics/ScrollingCoordinatorCoordinatedGraphics.cpp:
(WebCore::ScrollingCoordinatorCoordinatedGraphics::scrollableAreaScrollLayerDidChange):
(WebCore):
(WebCore::ScrollingCoordinatorCoordinatedGraphics::willDestroyScrollableArea):
* page/scrolling/coordinatedgraphics/ScrollingCoordinatorCoordinatedGraphics.h:
(ScrollingCoordinatorCoordinatedGraphics):

GraphicsLayerTextureMapper is responsible for receiving the 'did commit' messages and to pass
them to its TextureMapperLayer on the appropriated time.

* platform/graphics/texmap/GraphicsLayerTextureMapper.cpp:
(WebCore::GraphicsLayerTextureMapper::GraphicsLayerTextureMapper):
(WebCore::GraphicsLayerTextureMapper::didCommitScrollOffset):
(WebCore):
(WebCore::GraphicsLayerTextureMapper::setIsScrollable):
(WebCore::GraphicsLayerTextureMapper::commitLayerChanges):
* platform/graphics/texmap/GraphicsLayerTextureMapper.h:
(WebCore::GraphicsLayerTextureMapper::setScrollClient):
(WebCore::GraphicsLayerTextureMapper::setID):
(GraphicsLayerTextureMapper):
(WebCore::GraphicsLayerTextureMapper::isScrollable):

TextureMapperLayer is responsible for applying the requested scroll offset and for
sending the commit messages to web process. It is also able to hittest layers and identify
the ones that are scrollable.

* platform/graphics/texmap/TextureMapperLayer.cpp:
(WebCore::TextureMapperLayer::hitTest):
(WebCore):
(WebCore::TextureMapperLayer::scrollableLayerHitTestCondition):
(WebCore::TextureMapperLayer::findScrollableContentsLayerAt):
(WebCore::TextureMapperLayer::mapScrollOffset):
(WebCore::TextureMapperLayer::commitScrollOffset):
(WebCore::TextureMapperLayer::scrollBy):
(WebCore::TextureMapperLayer::didCommitScrollOffset):
* platform/graphics/texmap/TextureMapperLayer.h:
(TextureMapperLayer):
(ScrollingClient):
(WebCore::TextureMapperLayer::TextureMapperLayer):
(WebCore::TextureMapperLayer::setID):
(WebCore::TextureMapperLayer::id):
(WebCore::TextureMapperLayer::setScrollClient):
(WebCore::TextureMapperLayer::setIsScrollable):
(WebCore::TextureMapperLayer::isScrollable):
(WebCore::TextureMapperLayer::adjustedPosition):

CoordinatedGraphicsLayer receives the commit messages and updates its ScrollableArea.

* platform/graphics/texmap/coordinated/CoordinatedGraphicsLayer.cpp:
(WebCore::CoordinatedGraphicsLayer::CoordinatedGraphicsLayer):
(WebCore::CoordinatedGraphicsLayer::setScrollableArea):
(WebCore):
(WebCore::CoordinatedGraphicsLayer::commitScrollOffset):
(WebCore::CoordinatedGraphicsLayer::resetLayerState):
* platform/graphics/texmap/coordinated/CoordinatedGraphicsLayer.h:
(WebCore):
(CoordinatedGraphicsLayer):
(WebCore::CoordinatedGraphicsLayer::isScrollable):

CoordinatedGraphicsScene receives the commit messages from TextureMapperLayer and
dispatches them on the main thread. It also has a new method that is used to find
the topmost scrollable layer at a given point.

* platform/graphics/texmap/coordinated/CoordinatedGraphicsScene.cpp:
(WebCore::CoordinatedGraphicsScene::setLayerState):
(WebCore::CoordinatedGraphicsScene::createLayer):
(WebCore::CoordinatedGraphicsScene::dispatchCommitScrollOffset):
(WebCore):
(WebCore::CoordinatedGraphicsScene::commitScrollOffset):
(WebCore::CoordinatedGraphicsScene::findScrollableContentsLayerAt):
* platform/graphics/texmap/coordinated/CoordinatedGraphicsScene.h:
(WebCore):
(CoordinatedGraphicsSceneClient):
(CoordinatedGraphicsScene):

New fields have been added to CoordinatedGraphicsLayerState.

* platform/graphics/texmap/coordinated/CoordinatedGraphicsState.h:
(CoordinatedGraphicsLayerState):

Source/WebKit2:

Patch by Luiz Agostini <luiz.agostini@nokia.com> on 2013-03-06
Reviewed by Noam Rosenthal.
Signed off for WebKit2 by Simon Fraser.

A scroll offset is applied to individual layers on UI process and the
web process is notified assynchronously.
The new API WKCoordinatedScene may be used to scroll layers in the UI process.

* CMakeLists.txt:

Encoding and decoding the new members of CoordinatedGraphicsLayerState.

* Shared/CoordinatedGraphics/CoordinatedGraphicsArgumentCoders.cpp:
(CoreIPC::::encode):
(CoreIPC::::decode):

Adding acceleratedCompositingForOverflowScrollEnabled to WebPreferencesStore.

* Shared/WebPreferencesStore.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updatePreferences):

New API WKCoordinatedScene is used to scroll layers on UI side.

* UIProcess/API/CoordinatedGraphics/WKCoordinatedScene.cpp:
* UIProcess/API/CoordinatedGraphics/WKCoordinatedScene.h:
* UIProcess/CoordinatedGraphics/WKCoordinatedSceneAPICast.h:

A new message is used to send scroll commit messages to web process.

* UIProcess/CoordinatedGraphics/CoordinatedLayerTreeHostProxy.cpp:
(WebKit::CoordinatedLayerTreeHostProxy::commitScrollOffset):
(WebKit):
* UIProcess/CoordinatedGraphics/CoordinatedLayerTreeHostProxy.h:
(CoordinatedLayerTreeHostProxy):
* WebProcess/WebPage/CoordinatedGraphics/CoordinatedLayerTreeHost.cpp:
(WebKit::CoordinatedLayerTreeHost::~CoordinatedLayerTreeHost):
(WebKit::CoordinatedLayerTreeHost::detachLayer):
(WebKit::CoordinatedLayerTreeHost::createGraphicsLayer):
(WebKit::CoordinatedLayerTreeHost::setVisibleContentsRect):
(WebKit::CoordinatedLayerTreeHost::purgeBackingStores):
(WebKit::CoordinatedLayerTreeHost::commitScrollOffset):
(WebKit):
* WebProcess/WebPage/CoordinatedGraphics/CoordinatedLayerTreeHost.h:
(CoordinatedLayerTreeHost):
* WebProcess/WebPage/CoordinatedGraphics/CoordinatedLayerTreeHost.messages.in:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144907 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[EFL] Unreviewed gardening

Added new baselines after r144837 and marked as failures
some pasteboard tests that we don't support yet.

* platform/efl/TestExpectations:
* platform/efl/fast/invalid/017-expected.txt:
* platform/efl/fast/invalid/018-expected.txt:
* platform/efl/tables/mozilla/bugs/bug113235-2-expected.txt:
* platform/efl/tables/mozilla/bugs/bug9024-expected.txt: Added.
* platform/efl/tables/mozilla_expected_failures/bugs/bug3517-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144906 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed EFL gardening.

Unskip authentication test that is now passing on EFL port.

* platform/efl/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144905 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed GTK gardening.

* platform/gtk/TestExpectations: Adding a failure expectation for a test added in r144877.
* platform/gtk/fast/invalid/017-expected.txt: Rebaselining after r144837.
* platform/gtk/fast/invalid/018-expected.txt: Ditto.
* platform/gtk/tables/mozilla/bugs/bug113235-2-expected.txt: Added. Ditto.
* platform/gtk/tables/mozilla/bugs/bug9024-expected.txt: Added. Ditto.
* platform/gtk/tables/mozilla_expected_failures/bugs/bug3517-expected.txt: Added. Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144904 268f45cc-cd09-0410-ab3c-d52691b4dbfc

REGGRESSION (r142112): It made compositing tests fail on EFL, Qt.
https://bugs.webkit.org/show_bug.cgi?id=109291

Reviewed by Kenneth Rohde Christiansen.

Source/WebCore:

Report we support fixed layers.

* page/scrolling/coordinatedgraphics/ScrollingCoordinatorCoordinatedGraphics.h:
(WebCore::ScrollingCoordinatorCoordinatedGraphics::supportsFixedPositionLayers):
(ScrollingCoordinatorCoordinatedGraphics):

LayoutTests:

Unskip now passing test, and move baselines specific to having no scrolling-coordinator to WK1.

* platform/qt-5.0-wk1/scrollingcoordinator/non-fast-scrollable-region-scaled-iframe-expected.png: Renamed from LayoutTests/platform/qt/scrollingcoordinator/non-fast-scrollable-region-scaled-iframe-expected.png.
* platform/qt-5.0-wk1/scrollingcoordinator/non-fast-scrollable-region-scaled-iframe-expected.txt: Renamed from LayoutTests/platform/qt/scrollingcoordinator/non-fast-scrollable-region-scaled-iframe-expected.txt.
* platform/qt-5.0-wk1/scrollingcoordinator/non-fast-scrollable-region-transformed-iframe-expected.png: Renamed from LayoutTests/platform/qt/scrollingcoordinator/non-fast-scrollable-region-transformed-iframe-expected.png.
* platform/qt-5.0-wk1/scrollingcoordinator/non-fast-scrollable-region-transformed-iframe-expected.txt: Renamed from LayoutTests/platform/qt/scrollingcoordinator/non-fast-scrollable-region-transformed-iframe-expected.txt.
* platform/qt-5.0-wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144901 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[EFL][WK2] Convert from device view size to UI view size only in EwkView.
https://bugs.webkit.org/show_bug.cgi?id=110197

Patch by Huang Dongsung <luxtella@company100.net> on 2013-03-06
Reviewed by Kenneth Rohde Christiansen.

Signed off for WebKit2 by Benjamin Poulain.

Device scale factor is a property of the physical display, so only EwkView
should deal with converting device view size to UI (a.k.a Density Independent Pixel)
view size. It increases readability because we can regard a view size in
other classes (e.g. WebView, PageViewportController, WebPage) except for
EwkView as UI size.

* UIProcess/API/efl/EwkView.cpp:
(EwkView::setDeviceScaleFactor):
(EwkView::setDeviceSize):
(EwkView::size):
(EwkView::deviceSize):
(EwkView::scheduleUpdateDisplay):
(EwkView::createGLSurface):
(EwkView::handleEvasObjectCalculate):
* UIProcess/API/efl/EwkView.h:
(EwkView):
* UIProcess/efl/WebView.cpp:
(WebKit::WebView::paintToCurrentGLContext):
(WebKit::WebView::updateViewportSize):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144899 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[EFL][WK2] Use Vector::reserveInitialCapacity() in WebEventFactory::createWebTouchEvent()
https://bugs.webkit.org/show_bug.cgi?id=111530

Reviewed by Benjamin Poulain.

In WebEventFactory::createWebTouchEvent(), when constructing the vector of touch points,
we already know in advance the total number of touch points. Therefore, we should
leverage WTF::Vector::reserveInitialCapacity() and WTF::Vector::uncheckedAppend() to save
some capacity checks.

* Shared/efl/WebEventFactory.cpp:
(WebKit::WebEventFactory::createWebTouchEvent):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144898 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed gardening.

* platform/chromium/TestExpectations: Delete http/tests/misc/delete-frame-during-readystatechange-with-gc-after-video-removal.html

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144897 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, rolling out r144859.
http://trac.webkit.org/changeset/144859
https://bugs.webkit.org/show_bug.cgi?id=111519

Introduced a test that calls non-exist function on non-
Chromium ports and crash on Chromium (Requested by rniwa on
#webkit).

Patch by Sheriff Bot <webkit.review.bot@gmail.com> on 2013-03-06

Source/WebCore:

* html/HTMLAudioElement.h:
(WebCore::HTMLAudioElement::hasPendingActivity):
(HTMLAudioElement):
* html/HTMLAudioElement.idl:
* html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::hasPendingActivity):
* html/HTMLMediaElement.idl:

LayoutTests:

* http/tests/misc/delete-frame-during-readystatechange-with-gc-after-video-removal-expected.txt: Removed.
* http/tests/misc/delete-frame-during-readystatechange-with-gc-after-video-removal.html: Removed.
* http/tests/misc/resources/delete-frame-during-readystatechange-frame-with-gc-after-video-removal.html: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144896 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: Could not open Profiles panel.
https://bugs.webkit.org/show_bug.cgi?id=111535

Reviewed by Alexander Pavlov.

* inspector/front-end/CPUProfileView.js:
* inspector/front-end/ProfilesPanel.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144895 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Crash during middle mouse click when page is removed
https://bugs.webkit.org/show_bug.cgi?id=111489

Reviewed by Hajime Morrita.

Source/WebCore:

Check for a null page that can happen if we remove the document from the DOM.
The other methods in EventHandler also check for a null page.

Test: editing/pasteboard/selection-paste-crash.html

* page/EventHandler.cpp:
(WebCore::EventHandler::handlePasteGlobalSelection):

LayoutTests:

* editing/pasteboard/resources/selection-paste-crash.html: Added.
* editing/pasteboard/selection-paste-crash-expected.txt: Added.
* editing/pasteboard/selection-paste-crash.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144894 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[WK2][EFL] Enable asynchronous spell checking by default
https://bugs.webkit.org/show_bug.cgi?id=111295

Reviewed by Benjamin Poulain.

Asynchronous spell checking feature is disabled by default for wk2 platforms.
It means all requests of spell checking are sent synchronously.

WebKit-EFL implements requestCheckingOfString method which allows
to check spelling asynchronously.

* UIProcess/API/efl/EwkView.cpp:
(EwkView::EwkView):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144892 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Replace shouldLazyAttach bool with an enum for readability
https://bugs.webkit.org/show_bug.cgi?id=111503

Reviewed by Dimitri Glazkov.

No behavioral change.

* bindings/cpp/WebDOMNodeCustom.cpp:
(WebDOMNode::insertBefore):
(WebDOMNode::replaceChild):
(WebDOMNode::appendChild):
* bindings/js/JSNodeCustom.cpp:
(WebCore::JSNode::insertBefore):
(WebCore::JSNode::replaceChild):
(WebCore::JSNode::appendChild):
* bindings/v8/custom/V8NodeCustom.cpp:
(WebCore):
(WebCore::V8Node::insertBeforeMethodCustom):
(WebCore::V8Node::replaceChildMethodCustom):
(WebCore::V8Node::appendChildMethodCustom):
* dom/ContainerNode.cpp:
(WebCore):
(WebCore::ContainerNode::insertBefore):
(WebCore::ContainerNode::replaceChild):
(WebCore::ContainerNode::appendChild):
(WebCore::updateTreeAfterInsertion):
* dom/ContainerNode.h:
(ContainerNode):
* dom/Node.cpp:
(WebCore::Node::insertBefore):
(WebCore::Node::replaceChild):
(WebCore::Node::appendChild):
* dom/Node.h:
(Node):
* editing/AppendNodeCommand.cpp:
(WebCore::AppendNodeCommand::doApply):
* editing/InsertNodeBeforeCommand.cpp:
(WebCore::InsertNodeBeforeCommand::doApply):
* html/shadow/MediaControlsApple.cpp:
(WebCore::MediaControlsApple::createControls):
* html/shadow/MediaControlsBlackBerry.cpp:
(WebCore::MediaControlsBlackBerry::createControls):
* html/shadow/MediaControlsChromium.cpp:
(WebCore::MediaControlsChromium::initializeControls):
* html/shadow/MediaControlsChromiumAndroid.cpp:
(WebCore::MediaControlsChromiumAndroid::createControls):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144891 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Custom Elements: CustomElementConstructor::m_name doesn't have any good use.
https://bugs.webkit.org/show_bug.cgi?id=111523

Reviewed by Kentaro Hara.

This change removes CustomElementConstructor::m_name and renames
m_tagName to m_name. Now generated function name is same as interface name.

Originally, m_name was added to specify the generated constructor
name. An early version of custom element allows to give the name,
but it no longer says anything about that.

The rename is to align the semantics of type extension mechanism
of Custom Elements, where the name of a custom element isn't
necessarily a tag name.

No new tests. Just removing old code.

* bindings/scripts/CodeGeneratorV8.pm:
(GenerateHeader):
* dom/CustomElementConstructor.cpp:
(WebCore::CustomElementConstructor::create):
(WebCore::CustomElementConstructor::CustomElementConstructor):
(WebCore::CustomElementConstructor::createElement):
* dom/CustomElementConstructor.h:
(CustomElementConstructor):
(WebCore::CustomElementConstructor::name):
* dom/CustomElementRegistry.cpp:
(WebCore::CustomElementRegistry::registerElement):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144889 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: Flame Chart. Provide node context info in a popover when mouse hovers over a flame chart item.
https://bugs.webkit.org/show_bug.cgi?id=111442

Reviewed by Yury Semikhatsky.

FlameChart is a canvas so it is not possible to attach the popover to the items on it.
It would be a bit tricky to support an offset due to complex behaviour of the popover.
The simplest solution is to create a fake anchor element and move it to the right position.

* inspector/front-end/FlameChart.js:
(WebInspector.FlameChart):
(WebInspector.FlameChart.prototype._getPopoverAnchor):
(WebInspector.FlameChart.prototype._showPopover):
(WebInspector.FlameChart.prototype._onMouseMove):
(WebInspector.FlameChart.prototype.findNodeCallback):
(WebInspector.FlameChart.prototype._coordinatesToNode):
(WebInspector.FlameChart.prototype.onResize):
* inspector/front-end/flameChart.css:
(.flame-chart .item-anchor):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144888 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[EFL] Ensure right format is used in SnapshotImageGL.
https://bugs.webkit.org/show_bug.cgi?id=111333

Patch by Kondapally Kalyan <kalyan.kondapally@intel.com> on 2013-03-05
Reviewed by Benjamin Poulain.

GL_BGRA format is not standard for glReadPixels with GLES.
This patch ensures that GL_RGBA is used in SnapshotImageGL
while using GLES.

* UIProcess/API/efl/SnapshotImageGL.cpp:
(getImageSurfaceFromFrameBuffer):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144887 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Get rid of the invert argument to SpeculativeJIT::jumpSlowForUnwantedArrayMode
https://bugs.webkit.org/show_bug.cgi?id=105624

Reviewed by Oliver Hunt.

All callers pass invert = false, which is the default value of the argument. So, get
rid of the argument and fold away all code that checks it.

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

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144886 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed buildfix after r144787. Add missing USE(GRAPHICS_SURFACE) guards.

* platform/graphics/texmap/coordinated/CoordinatedGraphicsScene.cpp:
(WebCore::CoordinatedGraphicsScene::setLayerState):
* platform/graphics/texmap/coordinated/CoordinatedGraphicsState.h:
(WebCore::CoordinatedGraphicsLayerState::CoordinatedGraphicsLayerState):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144885 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Autoreleased cached pages slow down the PLT by 2%
https://bugs.webkit.org/show_bug.cgi?id=111522

Reviewed by Filip Pizlo.

Let's stop doing that.

2% PLT speedup.

* history/PageCache.cpp:
(WebCore::PageCache::remove):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144884 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Make toggle-unlink.html test Mac specific behaviour
https://bugs.webkit.org/show_bug.cgi?id=111321

Reviewed by Ryosuke Niwa.

A new test for Windows behavior will have to be added once bug
110487 is landed.

* editing/execCommand/script-tests/toggle-unlink-mac.js: Renamed
from
LayoutTests/editing/execCommand/script-tests/toggle-unlink.js.
Also set editing behaviour to Mac.
* editing/execCommand/toggle-unlink-mac-expected.txt: Renamed from
LayoutTests/editing/execCommand/toggle-unlink-expected.txt.
* editing/execCommand/toggle-unlink-mac.html: Renamed from
LayoutTests/editing/execCommand/toggle-unlink.html.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144883 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed chromium test expectations update.

Patch by Vsevolod Vlasov <vsevik@chromium.org> on 2013-03-05

* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144882 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed chromium test expectations update.

* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144881 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Add a flaky crash test expectation to editing/selection/selection-invalid-offset.html
on Mac per bug 111521.

* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144880 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Add test coverage for DCE and type checks.

Rubber stamped by Geoffrey Garen.

* fast/js/dfg-cfg-simplify-eliminate-set-local-type-check-then-branch-not-null-and-decrement-expected.txt: Added.
* fast/js/dfg-cfg-simplify-eliminate-set-local-type-check-then-branch-not-null-and-decrement.html: Added.
* fast/js/dfg-cfg-simplify-eliminate-set-local-type-check-then-branch-not-null-expected.txt: Added.
* fast/js/dfg-cfg-simplify-eliminate-set-local-type-check-then-branch-not-null.html: Added.
* fast/js/jsc-test-list:
* fast/js/script-tests/dfg-cfg-simplify-eliminate-set-local-type-check-then-branch-not-null-and-decrement.js: Added.
(foo):
* fast/js/script-tests/dfg-cfg-simplify-eliminate-set-local-type-check-then-branch-not-null.js: Added.
(foo):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144879 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed gardening.

* platform/chromium/TestExpectations: http/tests/misc/delete-frame-during-readystatechange-with-gc-after-video-removal.html

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144878 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Shadow] Implements event re-targeting for Touch Events.
https://bugs.webkit.org/show_bug.cgi?id=107800

Reviewed by Dimitri Glazkov.

Source/WebCore:

So far, Touch Events have been intentionally disabled in shadow
tree since it allows users to access internal nodes in shadow tree
by accessing Touch Event's information.

This patch re-targets all individual touch point's target in
TouchList (touches, targetTouches and changedTouches) owned by a Touch Event
before the Touch Event is dispatched so that we won't break an
upper boundary of shadow DOM.

Touch Events re-targeting is similar to that of Mouse Events
re-targeting. The difference is that a Touch Event can have more
than one related target because of multiple touches. We must
re-target all individual touch points which can be referred from
the Touch Event.

Now, Touch Events can be fired in shadow tree. I've removed the
existing limitation.

Test: fast/dom/shadow/touch-event-retargeting.html

* dom/EventContext.cpp:
(WebCore::EventContext::isTouchEventContext):
(WebCore):
(WebCore::TouchEventContext::TouchEventContext):
(WebCore::TouchEventContext::~TouchEventContext):
(WebCore::TouchEventContext::handleLocalEvents):
(WebCore::TouchEventContext::isTouchEventContext):
(WebCore::TouchEventContext::checkReachability): Used for assertion of reachability for each Touch point.
* dom/EventContext.h:
(WebCore):
(EventContext):
(TouchEventContext): Introduced to hold necessary information for re-targeting Touch event.
(WebCore::TouchEventContext::touches):
(WebCore::TouchEventContext::targetTouches):
(WebCore::TouchEventContext::changedTouches):
(WebCore::EventContext::isReachable):
* dom/EventRetargeter.cpp:
(WebCore::EventRetargeter::calculateEventPath):
(WebCore::EventRetargeter::adjustForTouchEvent):
(WebCore):
(WebCore::EventRetargeter::adjustTouchList):
(WebCore::EventRetargeter::adjustForRelatedTarget):
(WebCore::EventRetargeter::calculateAdjustedNodes): Updated so that this can be used for Touch point re-targeting.
(WebCore::EventRetargeter::buildRelatedNodeMap):
* dom/EventRetargeter.h:
(WebCore):
(EventRetargeter):
* dom/Node.cpp:
(WebCore::Node::dispatchEvent):
(WebCore):
(WebCore::Node::dispatchTouchEvent):
* dom/Node.h:
(WebCore):
(Node):
* dom/Touch.cpp:
(WebCore::Touch::Touch):
(WebCore):
(WebCore::Touch::cloneWithNewTarget):
* dom/Touch.h:
(Touch):
* dom/TouchEvent.cpp:
(WebCore::TouchEventDispatchMediator::create):
(WebCore):
(WebCore::TouchEventDispatchMediator::TouchEventDispatchMediator):
(WebCore::TouchEventDispatchMediator::event):
(WebCore::TouchEventDispatchMediator::dispatchEvent):
* dom/TouchEvent.h:
(TouchEvent):
(WebCore::TouchEvent::setTouches):
(WebCore::TouchEvent::setTargetTouches):
(WebCore::TouchEvent::setChangedTouches):
(TouchEventDispatchMediator):
(WebCore):
(WebCore::toTouchEvent):
* dom/TouchList.cpp:
(WebCore::TouchList::item):
(WebCore):
* dom/TouchList.h:
(TouchList):
* page/EventHandler.cpp:
(WebCore::EventHandler::hitTestResultInFrame): Removed the limitation. Adjusting is no longer done here.
(WebCore::EventHandler::handleTouchEvent):

LayoutTests:

* fast/dom/shadow/resources/event-dispatching.js:
(recordEvent):
(dumpTouchList):
(sortDispatchedEvent):
(showSandboxTree):
* fast/dom/shadow/shadow-dom-event-dispatching-distributed-text-node-expected.txt:
* fast/dom/shadow/shadow-dom-event-dispatching-distributed-text-node.html:
* fast/dom/shadow/shadow-dom-event-dispatching-text-node-in-shadow-root-expected.txt:
* fast/dom/shadow/touch-event-retargeting-expected.txt: Added.
* fast/dom/shadow/touch-event-retargeting.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144877 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Add test coverage for DCE and type checks.

Rubber stampted by Geoffrey Garen.

* fast/js/dfg-cfg-simplify-eliminate-set-local-type-check-then-typeof-expected.txt: Added.
* fast/js/dfg-cfg-simplify-eliminate-set-local-type-check-then-typeof.html: Added.
* fast/js/jsc-test-list:
* fast/js/script-tests/dfg-cfg-simplify-eliminate-set-local-type-check-then-typeof.js: Added.
(foo):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144876 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed gardening. AppleWin port.

* platform/win/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144875 268f45cc-cd09-0410-ab3c-d52691b4dbfc

DFG may infer an edge to be a double-using edge and then later experience multiple-personality disorder with respect to that particular decision
https://bugs.webkit.org/show_bug.cgi?id=109393

Rubber stamped by Geoffrey Garen.

* fast/js/dfg-double-addition-simplify-to-int-expected.txt: Added.
* fast/js/dfg-double-addition-simplify-to-int.html: Added.
* fast/js/jsc-test-list:
* fast/js/script-tests/dfg-double-addition-simplify-to-int.js: Added.
(foo):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144874 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed gardening. AppleWin port.

* platform/win/cssom: Added.
* platform/win/cssom/cssvalue-comparison-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144873 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed gardening. AppleWin port.

Patch by Roger Fong <roger_fong@apple.com> on 2013-03-05

* platform/win/TestExpectations:
* platform/win/fast/forms/number: Added.
* platform/win/fast/forms/number/number-l10n-input-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144872 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed gardening. AppleWin port.

* platform/win/TestExpectations:
* platform/win/fast/forms/number: Added.
* platform/win/fast/forms/number/number-l10n-input-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144871 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[chromium] don't use custom freetype on linux asan for now
https://bugs.webkit.org/show_bug.cgi?id=111516

Unreviewed, build fix.

Turns out the custom lib is getting picked up by other binaries
on our bot, which is bad.

* DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144870 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed gardening.

* platform/chromium/TestExpectations: plugins/plugin-clip-subframe.html

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144869 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed gardening.

* platform/chromium/TestExpectations: fast/regions/region-style-in-columns.html

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144868 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[chromium] making DumpRenderTree_resources a separate bundle breaks xcode
https://bugs.webkit.org/show_bug.cgi?id=111509

Unreviewed, build fix.

Actually comment out the flag :(.

* DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144867 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[chromium] making DumpRenderTree_resources a separate bundle breaks xcode
https://bugs.webkit.org/show_bug.cgi?id=111509

Unreviewed, build fix.

r144863 appears to have broken the xcode build; I'm temporarily
trying this instead, and if this doesn't work, I'll revert the
other changes.

* DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144866 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Custom Elements][V8] Custom Element doesn't need its own WrapperTypeInfo
https://bugs.webkit.org/show_bug.cgi?id=111411

Reviewed by Kentaro Hara.

No new tests. No observable change yet.

This is a preparation for Bug 110436, where WebKit is going to
allow any HTML element to be a superclass of custom elements.

This change eliminates WrapperTypeInfo for custom element. Instead
of that, the wrapper of a custom element picks WrapperTypeInfo
from most descendant built-in HTML element. For example, think about
following hierarchy:

- ... <- HTMLElement <- HTMLDivElememnt <- ExistingCustomElement <- NewCustomElement

In this case, both ExistingCustomElement and NewCustomElement has
HTMLDivElememnt's type info. This is because HTMLDivElememnt is
the most descendant (or "the nearest") built-in HTML
element. (HTMLElement isn't the most descendant.
ExistingCustomElement isn't a built-in.)

Using such WrapperTypeInfos, created custom element wrappers can
get appropriate set of method and properties regardless of its
super-elements.

This change tells the prototype object of each built-in object
which WrapperTypeInfo the native backing wants:
HTMLDivElememnt.prototype knows V8HTMLDivElement::info for
example. The pointer to the type info is stored into an internal
field. On wrapper construction, createWrapper() walks up the
prototype chain and pick the first one.

* bindings/v8/V8DOMConfiguration.cpp:
(WebCore::V8DOMConfiguration::configureTemplate):
* bindings/v8/V8HTMLCustomElement.cpp:
(WebCore::findWrapperTypeOf):
(WebCore::V8HTMLCustomElement::createWrapper):
* bindings/v8/V8HTMLCustomElement.h:
* bindings/v8/V8PerContextData.cpp:
(WebCore::V8PerContextData::constructorForTypeSlowCase):
* bindings/v8/WrapperTypeInfo.h:
(WebCore):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144865 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, fix an incorrect comment. The comment was a holdover from a work-in-progress version of this code.

* dfg/DFGDCEPhase.cpp:
(JSC::DFG::DCEPhase::run):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144864 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[chromium] build a DumpRenderTree.pak even on windows
https://bugs.webkit.org/show_bug.cgi?id=111504

Reviewed by Tony Chang.

The Chromium Aura build uses the default theme to draw
controls, and that theme requires the ui ResourceBundle
to be initialized. On every other platform, we build
a DumpRenderTree.pak that works for this, but on Windows we
still compile the resources directly into DRT. As a first step,
let's build the pak file. I will then need to land a patch
in Chromium to actually load it. Then I can go back and move
the resources over to just use the .pak file and no longer need
the .rc's.

* DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144863 268f45cc-cd09-0410-ab3c-d52691b4dbfc

DFG DCE might eliminate checks unsoundly
https://bugs.webkit.org/show_bug.cgi?id=109389

Source/JavaScriptCore:

Reviewed by Oliver Hunt.

This gets rid of all eager reference counting, and does all dead code elimination
in one phase - the DCEPhase. This phase also sets up the node reference counts,
which are then used not just for DCE but also register allocation and stack slot
allocation.

Doing this required a number of surgical changes in places that previously relied
on always having liveness information. For example, the structure check hoisting
phase must now consult whether a VariableAccessData is profitable for unboxing to
make sure that it doesn't try to do hoisting on set SetLocals. The arguments
simplification phase employs its own light-weight liveness analysis. Both phases
previously just used reference counts.

The largest change is that now, dead nodes get turned into Phantoms. Those
Phantoms will retain those child edges that are not proven. This ensures that any
type checks performed by a dead node remain even after the node is killed. On the
other hand, this Phantom conversion means that we need special handling for
SetLocal. I decided to make the four forms of SetLocal explicit:

MovHint(@a, rK): Just indicates that node @a contains the value that would have
     now been placed into virtual register rK. Does not actually cause @a to be
     stored into rK. This would have previously been a dead SetLocal with @a
     being live. MovHints are always dead.

ZombieHint(rK): Indicates that at this point, register rK will contain a dead
     value and OSR should put Undefined into it. This would have previously been
     a dead SetLocal with @a being dead also. ZombieHints are always dead.

MovHintAndCheck(@a, rK): Identical to MovHint except @a is also type checked,
     according to whatever UseKind the edge to @a has. The type check is always a
     forward exit. MovHintAndChecks are always live, since they are
     NodeMustGenerate. Previously this would have been a dead SetLocal with a
     live @a, and the check would have disappeared. This is one of the bugs that
     this patch solves.

SetLocal(@a, rK): This still does exactly what it does now, if the SetLocal is
     live.

Basically this patch makes it so that dead SetLocals eventually decay to MovHint,
ZombieHint, or MovHintAndCheck depending on the situation. If the child @a is
also dead, then you get a ZombieHint. If the child @a is live but the SetLocal
has a type check and @a's type hasn't been proven to have that type then you get
a MovHintAndCheck. Otherwise you get a MovHint.

This is performance neutral.

* CMakeLists.txt:
* GNUmakefile.list.am:
* JavaScriptCore.xcodeproj/project.pbxproj:
* Target.pri:
* dfg/DFGAbstractState.cpp:
(JSC::DFG::AbstractState::executeEffects):
(JSC::DFG::AbstractState::mergeStateAtTail):
* dfg/DFGArgumentsSimplificationPhase.cpp:
(JSC::DFG::ArgumentsSimplificationPhase::run):
(ArgumentsSimplificationPhase):
(JSC::DFG::ArgumentsSimplificationPhase::removeArgumentsReferencingPhantomChild):
* dfg/DFGBasicBlock.h:
(BasicBlock):
* dfg/DFGBasicBlockInlines.h:
(DFG):
* dfg/DFGByteCodeParser.cpp:
(JSC::DFG::ByteCodeParser::addToGraph):
(JSC::DFG::ByteCodeParser::insertPhiNode):
(JSC::DFG::ByteCodeParser::emitFunctionChecks):
* dfg/DFGCFAPhase.cpp:
(JSC::DFG::CFAPhase::run):
* dfg/DFGCFGSimplificationPhase.cpp:
(JSC::DFG::CFGSimplificationPhase::run):
(JSC::DFG::CFGSimplificationPhase::keepOperandAlive):
* dfg/DFGCPSRethreadingPhase.cpp:
(JSC::DFG::CPSRethreadingPhase::run):
(JSC::DFG::CPSRethreadingPhase::addPhiSilently):
* dfg/DFGCSEPhase.cpp:
(JSC::DFG::CSEPhase::eliminateIrrelevantPhantomChildren):
(JSC::DFG::CSEPhase::setReplacement):
(JSC::DFG::CSEPhase::performNodeCSE):
* dfg/DFGCommon.cpp:
(WTF::printInternal):
(WTF):
* dfg/DFGCommon.h:
(WTF):
* dfg/DFGConstantFoldingPhase.cpp:
(JSC::DFG::ConstantFoldingPhase::foldConstants):
(JSC::DFG::ConstantFoldingPhase::addStructureTransitionCheck):
(JSC::DFG::ConstantFoldingPhase::paintUnreachableCode):
* dfg/DFGDCEPhase.cpp: Added.
(DFG):
(DCEPhase):
(JSC::DFG::DCEPhase::DCEPhase):
(JSC::DFG::DCEPhase::run):
(JSC::DFG::DCEPhase::findTypeCheckRoot):
(JSC::DFG::DCEPhase::countEdge):
(JSC::DFG::DCEPhase::eliminateIrrelevantPhantomChildren):
(JSC::DFG::performDCE):
* dfg/DFGDCEPhase.h: Added.
(DFG):
* dfg/DFGDriver.cpp:
(JSC::DFG::compile):
* dfg/DFGFixupPhase.cpp:
(JSC::DFG::FixupPhase::fixupNode):
(JSC::DFG::FixupPhase::checkArray):
(JSC::DFG::FixupPhase::blessArrayOperation):
(JSC::DFG::FixupPhase::fixIntEdge):
(JSC::DFG::FixupPhase::injectInt32ToDoubleNode):
(JSC::DFG::FixupPhase::truncateConstantToInt32):
* dfg/DFGGraph.cpp:
(JSC::DFG::Graph::Graph):
(JSC::DFG::Graph::dump):
(DFG):
* dfg/DFGGraph.h:
(JSC::DFG::Graph::changeChild):
(JSC::DFG::Graph::changeEdge):
(JSC::DFG::Graph::compareAndSwap):
(JSC::DFG::Graph::clearAndDerefChild):
(JSC::DFG::Graph::performSubstitution):
(JSC::DFG::Graph::performSubstitutionForEdge):
(Graph):
(JSC::DFG::Graph::substitute):
* dfg/DFGInsertionSet.h:
(InsertionSet):
* dfg/DFGNode.h:
(JSC::DFG::Node::Node):
(JSC::DFG::Node::convertToConstant):
(JSC::DFG::Node::convertToGetLocalUnlinked):
(JSC::DFG::Node::containsMovHint):
(Node):
(JSC::DFG::Node::hasVariableAccessData):
(JSC::DFG::Node::willHaveCodeGenOrOSR):
* dfg/DFGNodeType.h:
(DFG):
* dfg/DFGPredictionPropagationPhase.cpp:
(JSC::DFG::PredictionPropagationPhase::propagate):
* dfg/DFGSpeculativeJIT.cpp:
(JSC::DFG::SpeculativeJIT::convertLastOSRExitToForward):
(JSC::DFG::SpeculativeJIT::compileMovHint):
(JSC::DFG::SpeculativeJIT::compileMovHintAndCheck):
(DFG):
(JSC::DFG::SpeculativeJIT::compileInlineStart):
(JSC::DFG::SpeculativeJIT::compile):
* dfg/DFGSpeculativeJIT.h:
(SpeculativeJIT):
* dfg/DFGSpeculativeJIT32_64.cpp:
(JSC::DFG::SpeculativeJIT::compile):
* dfg/DFGSpeculativeJIT64.cpp:
(JSC::DFG::SpeculativeJIT::compile):
* dfg/DFGStructureCheckHoistingPhase.cpp:
(JSC::DFG::StructureCheckHoistingPhase::run):
(JSC::DFG::StructureCheckHoistingPhase::shouldConsiderForHoisting):
(StructureCheckHoistingPhase):
* dfg/DFGValidate.cpp:
(JSC::DFG::Validate::validate):

LayoutTests:

Reviewed by Oliver Hunt.

* fast/js/dfg-arguments-osr-exit-multiple-blocks-before-exit-expected.txt: Added.
* fast/js/dfg-arguments-osr-exit-multiple-blocks-before-exit.html: Added.
* fast/js/dfg-arguments-osr-exit-multiple-blocks-expected.txt: Added.
* fast/js/dfg-arguments-osr-exit-multiple-blocks.html: Added.
* fast/js/jsc-test-list:
* fast/js/script-tests/dfg-arguments-osr-exit-multiple-blocks-before-exit.js: Added.
(baz):
(foo):
(bar):
* fast/js/script-tests/dfg-arguments-osr-exit-multiple-blocks.js: Added.
(baz):
(foo):
(bar):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144862 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed gardening.

* platform/chromium/TestExpectations: inspector/console/command-line-api-inspect.html

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144861 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Each web process truncates the disk cache to zero on launch
https://bugs.webkit.org/show_bug.cgi?id=111467

Reviewed by Darin Adler.

Let's not do that.

* WebProcess/mac/WebProcessMac.mm:
(WebKit::WebProcess::platformInitializeWebProcess):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144860 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Heap-use-after-free in WebCore::HTMLMediaElement::~HTMLMediaElement
https://bugs.webkit.org/show_bug.cgi?id=110623

Reviewed by Eric Seidel.

Source/WebCore:

Test: http/tests/misc/delete-frame-during-readystatechange-with-gc-after-video-removal.html

* html/HTMLAudioElement.h:
(HTMLAudioElement):
* html/HTMLAudioElement.idl:
* html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::hasPendingActivity):
* html/HTMLMediaElement.idl:

LayoutTests:

* http/tests/misc/delete-frame-during-readystatechange-with-gc-after-video-removal-expected.txt: Added.
* http/tests/misc/delete-frame-during-readystatechange-with-gc-after-video-removal.html: Added.
* http/tests/misc/resources/delete-frame-during-readystatechange-frame-with-gc-after-video-removal.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144859 268f45cc-cd09-0410-ab3c-d52691b4dbfc

AX: Support aria-posinset/setsize
https://bugs.webkit.org/show_bug.cgi?id=109725

Reviewed by Tim Horton.

Source/WebCore:

Expose aria-setsize and aria-posinset through accessibility.
Right now only the Mac platform exposes them.

Test: platform/mac/accessibility/aria-setsize-posinset.html

* accessibility/AccessibilityObject.cpp:
(WebCore::AccessibilityObject::supportsARIASetSize):
(WebCore):
(WebCore::AccessibilityObject::supportsARIAPosInset):
(WebCore::AccessibilityObject::ariaSetSize):
(WebCore::AccessibilityObject::ariaPosInset):
* accessibility/AccessibilityObject.h:
(AccessibilityObject):
* accessibility/mac/WebAccessibilityObjectWrapper.mm:
(-[WebAccessibilityObjectWrapper additionalAccessibilityAttributeNames]):
(-[WebAccessibilityObjectWrapper accessibilityAttributeValue:]):
* html/HTMLAttributeNames.in:

LayoutTests:

* platform/mac/accessibility/aria-setsize-posinset-expected.txt: Added.
* platform/mac/accessibility/aria-setsize-posinset.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144858 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remove deprecated process suppression assertion SPI from WebKitSystemInterface
https://bugs.webkit.org/show_bug.cgi?id=111501

Patch by Kiran Muppala <cmuppala@apple.com> on 2013-03-05
Reviewed by Alexey Proskuryakov.

https://bugs.webkit.org/show_bug.cgi?id=111387 replaced all usage of
WKNSProcessInfoProcessAssertionWithTypes() with
-[NSProcessInfo beginSuspensionOfSystemBehaviors:]. Hence, remove
the now deprecated SPI.

* WebKitSystemInterface.h: Remove
WKNSProcessInfoProcessAssertionWithTypes and associated declarations.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144857 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[chromium] Scroll deltas should be floats
https://bugs.webkit.org/show_bug.cgi?id=111465

Reviewed by James Robinson.

This converts the remaining scroll delta fields to floats so that
we no longer lose precision due to DIP conversion. The natural
type for deltas is WebFloatSize, which was missing, so I also created
that type based on the existing WebSize header.

Source/Platform:

* Platform.gypi:
* chromium/public/WebFloatSize.h: Added.
(WebKit):
(WebFloatSize):
(WebKit::WebFloatSize::isEmpty):
(WebKit::WebFloatSize::WebFloatSize):
(WebKit::WebFloatSize::operator=):
(WebKit::WebFloatSize::operator WebCore::FloatSize):
(WebKit::WebFloatSize::operator gfx::SizeF):
(WebKit::WebFloatSize::operator gfx::Vector2dF):
(WebKit::operator==):
(WebKit::operator!=):
* chromium/public/WebGestureCurveTarget.h:
(WebKit::WebGestureCurveTarget::scrollBy):
(WebGestureCurveTarget):
* chromium/public/WebInputHandlerClient.h:
(WebKit::WebInputHandlerClient::scrollByIfPossible):
(WebInputHandlerClient):

Source/WebKit/chromium:

* src/WebCompositorInputHandlerImpl.cpp:
(WebKit::WebCompositorInputHandlerImpl::handleInputEventInternal):
(WebKit::WebCompositorInputHandlerImpl::touchpadFlingScroll):
(WebKit::WebCompositorInputHandlerImpl::scrollBy):
* src/WebCompositorInputHandlerImpl.h:
(WebCompositorInputHandlerImpl):
* src/WebViewImpl.cpp:
(WebKit::WebViewImpl::scrollBy):
* src/WebViewImpl.h:
(WebViewImpl):
* tests/WebCompositorInputHandlerImplTest.cpp:
(MockWebInputHandlerClient):
(WebKit::TEST_F):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144856 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed apple rebaselines for r144837.

* fast/invalid/017-expected.txt: Added.
* fast/invalid/018-expected.txt: Added.
* platform/chromium-mac/fast/invalid/017-expected.txt: Removed.
* platform/chromium-mac/fast/invalid/018-expected.txt: Removed.
* platform/mac-lion/tables/mozilla/bugs: Added.
* platform/mac-lion/tables/mozilla/bugs/bug113235-2-expected.txt: Added.
* platform/mac-lion/tables/mozilla/bugs/bug9024-expected.txt: Added.
* platform/mac-lion/tables/mozilla_expected_failures: Removed.
* platform/mac-wk2/tables/mozilla/bugs/bug113235-2-expected.txt: Added.
* platform/mac-wk2/tables/mozilla/bugs/bug9024-expected.txt: Added.
* platform/mac/fast/invalid/017-expected.txt: Removed.
* platform/mac/fast/invalid/018-expected.txt: Removed.
* platform/mac/tables/mozilla/bugs/bug113235-2-expected.txt:
* platform/mac/tables/mozilla/bugs/bug9024-expected.txt: Added.
* platform/mac/tables/mozilla_expected_failures/bugs/bug3517-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144855 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed chromium rebaselines for r144837.

* platform/chromium-linux/fast/invalid/017-expected.txt: Added.
* platform/chromium-linux/fast/invalid/018-expected.txt: Added.
* platform/chromium-linux/tables/mozilla/bugs/bug113235-2-expected.txt: Added.
* platform/chromium-linux/tables/mozilla/bugs/bug9024-expected.txt: Added.
* platform/chromium-linux/tables/mozilla_expected_failures/bugs/bug3517-expected.txt: Added.
* platform/chromium-mac-lion/tables/mozilla/bugs/bug9024-expected.txt: Added.
* platform/chromium-mac-snowleopard/tables/mozilla/bugs/bug9024-expected.txt: Added.
* platform/chromium-mac/fast/invalid/017-expected.txt: Added.
* platform/chromium-mac/fast/invalid/018-expected.txt: Added.
* platform/chromium-mac/tables/mozilla/bugs/bug113235-2-expected.txt: Removed.
* platform/chromium-mac/tables/mozilla/bugs/bug9024-expected.txt: Added.
* platform/chromium-win/tables/mozilla_expected_failures/bugs/bug3517-expected.txt: Added.
* platform/chromium/fast/invalid/017-expected.txt: Removed.
* platform/chromium/fast/invalid/018-expected.txt: Removed.
* platform/chromium/tables/mozilla/bugs/bug113235-2-expected.txt: Added.
* platform/chromium/tables/mozilla_expected_failures/bugs/bug3517-expected.txt: Replaced.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144854 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Source/WebCore: [WinCairo] Support for cookies is incomplete
https://bugs.webkit.org/show_bug.cgi?id=110147

Expired and HttpOnly cookies no longer accessible from JavaScript.
Cookies set in JavaScript now have correct domain/path.

Patch by Peter Nelson <peter@peterdn.com> on 2013-03-05
Reviewed by Brent Fulgham.

Test: http/tests/cookies/http-get-cookie-set-in-js.html

* platform/network/curl/CookieJarCurl.cpp:
(WebCore):
(WebCore::addMatchingCurlCookie):
(WebCore::getNetscapeCookieFormat):
(WebCore::setCookiesFromDOM):
(WebCore::cookieRequestHeaderFieldValue):

LayoutTests: [WinCairo] Support for cookies is incomplete
https://bugs.webkit.org/show_bug.cgi?id=110147

Patch by Peter Nelson <peter@peterdn.com> on 2013-03-05
Reviewed by Brent Fulgham.

Re-enabled cookie tests for WinCairo.
Added test to check whether cookie set in HTTP response is accessible in JS.

* http/tests/cookies/http-get-cookie-set-in-js-expected.txt: Added.
* http/tests/cookies/http-get-cookie-set-in-js.html: Added.
* http/tests/cookies/resources/cookies-test-pre.js:
(clearAllCookies): Cookies set in JS now correctly cleared.
* platform/wincairo/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144853 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed. AppleWin gardening.

* platform/win/TestExpectations:
* platform/win/css3/filters/effect-reference-expected.txt:
* platform/win/css3/filters/effect-reference-hw-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144852 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Reset the values if there is a quota error
https://bugs.webkit.org/show_bug.cgi?id=111500

Reviewed by Beth Dakin.

* WebProcess/Storage/StorageAreaProxy.cpp:
(WebKit::StorageAreaProxy::didSetItem):
(WebKit::StorageAreaProxy::resetValues):
(WebKit):
* WebProcess/Storage/StorageAreaProxy.h:
(StorageAreaProxy):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144851 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Keep track of pending value changes in StorageAreaProxy
https://bugs.webkit.org/show_bug.cgi?id=111496

Reviewed by Beth Dakin.

Add a HashCountedSet to keep track of the number of pending value
changes for a given key. If we get incoming storage events from other
processes while we have pending value changes, ignore the events.

* WebProcess/Storage/StorageAreaProxy.cpp:
(WebKit::StorageAreaProxy::setItem):
(WebKit::StorageAreaProxy::didSetItem):
(WebKit::StorageAreaProxy::dispatchStorageEvent):
(WebKit::StorageAreaProxy::shouldApplyChangesForKey):
(WebKit):
* WebProcess/Storage/StorageAreaProxy.h:
(StorageAreaProxy):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144850 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fix mac clang build with long long
https://bugs.webkit.org/show_bug.cgi?id=111495

Unreviewed fix for mac build.

* Modules/indexeddb/IDBBackingStore.cpp:
(WebCore::IDBBackingStore::IDBBackingStore):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144849 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed rebaseline of media/track/track-cue-rendering-*.html
https://bugs.webkit.org/show_bug.cgi?id=111388

* platform/chromium-linux/media/track/track-cue-rendering-horizontal-expected.png:
* platform/chromium-linux/media/track/track-cue-rendering-vertical-expected.png:
* platform/chromium-mac-lion/media/track/track-cue-rendering-horizontal-expected.png:
* platform/chromium-mac-lion/media/track/track-cue-rendering-vertical-expected.png:
* platform/chromium-mac-snowleopard/media/track/track-cue-rendering-horizontal-expected.png:
* platform/chromium-mac-snowleopard/media/track/track-cue-rendering-vertical-expected.png:
* platform/chromium-mac/media/track/track-cue-rendering-horizontal-expected.png:
* platform/chromium-mac/media/track/track-cue-rendering-vertical-expected.png:
* platform/chromium-win-xp/media/track/track-cue-rendering-vertical-expected.png:
* platform/chromium-win/media/track/track-cue-rendering-horizontal-expected.png:
* platform/chromium-win/media/track/track-cue-rendering-vertical-expected.png:
* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144847 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Add V8Document.h include when the V8 code generator creates a toV8(Document*) call.
https://bugs.webkit.org/show_bug.cgi?id=111486

Reviewed by Adam Barth.

No new tests. This doesn't change user visible behavior.

* bindings/scripts/CodeGeneratorV8.pm:
(GenerateNamedConstructor):
* bindings/scripts/test/V8/V8TestNamedConstructor.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144846 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Dashboard cleanup: move string utilities into their own namespace.

https://bugs.webkit.org/show_bug.cgi?id=111485

Reviewed by Ojan Vafai.

One of many pieces to clean up dashboard_base.

Moves string utility functions into separate file (string.js) and into their own namespace rather than cluttering up the global scope.

* TestResultServer/static-dashboards/aggregate_results.html:
* TestResultServer/static-dashboards/dashboard_base.js:
(isTreeMap):
(isFlakinessDashboard):
* TestResultServer/static-dashboards/flakiness_dashboard.html:
* TestResultServer/static-dashboards/flakiness_dashboard.js:
(determineWKPlatform):
(chromiumPlatform):
(platformAndBuildType):
(substringList):
(individualTestsForSubstringList):
(filterBugs):
(populateExpectationsData):
(processMissingTestsWithExpectations):
(processMissingAndExtraExpectations):
(htmlForSingleTestRow):
(.dummyNode.onload):
(expectationsTitle):
* TestResultServer/static-dashboards/run-embedded-unittests.html:
* TestResultServer/static-dashboards/run-unittests.html:
* TestResultServer/static-dashboards/string.js: Added.
* TestResultServer/static-dashboards/timeline_explorer.html:
* TestResultServer/static-dashboards/treemap.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144845 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Objective-C API: JSValue should implement init and return nil in exceptional cases
https://bugs.webkit.org/show_bug.cgi?id=111487

Reviewed by Darin Adler.

* API/JSValue.mm:
(-[JSValue init]): We return nil here because there is no way to get the instance into a coherent state
without a JSContext.
(-[JSValue initWithValue:inContext:]): Similarly, we should also return nil here if either of the arguments is 0.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144843 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[EFL] Build break with latest EFL libraries
https://bugs.webkit.org/show_bug.cgi?id=111028

Reviewed by Dirk Pranke.

.:

In latest EFL trunk, include path of ecore sub modules are changed from
ecore-1 to ecore-XXX-1.
So, this patch adds missing ECORE_XXX_INCLUDE_DIRS.

* Source/cmake/FindEcore.cmake: Added additional path suffixes to find include directories of ecore-XXX
* Source/cmake/FindElementary.cmake: Checked Ecore_Con dependency which Elementary requires.
* Source/cmake/OptionsEfl.cmake: Made Imf and Imf_Evas mandatory.

Source/WebKit:

* PlatformEfl.cmake: Added ECORE_X_INCLUDE_DIRS.

Source/WebKit2:

* PlatformEfl.cmake: Added ECORE_XXX_INCLUDE_DIRS.

Source/WTF:

* wtf/PlatformEfl.cmake: Added ECORE_IMF_INCLUDE_DIRS.

Tools:

* MiniBrowser/efl/CMakeLists.txt: Added ECORE_XXX_INCLUDE_DIRS.
* TestWebKitAPI/PlatformEfl.cmake: Ditto.
* WebKitTestRunner/PlatformEfl.cmake: Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144842 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fix default background of a dragged image.
https://bugs.webkit.org/show_bug.cgi?id=110512

Reviewed by Tony Chang.

.:

* ManualTests/drag-background-with-padding.html: Added.

Source/WebCore:

Transparent background color (particularly visible in the padding
of an element) was not being respected. The solution is to
initially fill the drag image with transparent instead of white,
so that any color background works correctly. This patch does not
affect Apple Mac or Win platforms, given that this part of the
code is #if guarded from those platforms.

Manual test added:
ManualTests/drag-background-with-padding.html

Infrastructure for retrieving the drag image for pixel testing
does not yet exist.

* page/Frame.cpp:
(WebCore::Frame::nodeImage):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144841 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Increase number of WinEWS iterations to 20.
Don't reboot after each iteration set.
Make EWS bots just build temporarily so that the queue catches up again.

* EWSTools/start-queue-win.sh:
* Scripts/webkitpy/tool/commands/earlywarningsystem.py:
(WinEWS):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144840 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed gardening.

* platform/chromium/TestExpectations: /inspector/debugger/debugger-script-preprocessor.html

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144839 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Use a named parameter for whether a quota error ocurred or not
https://bugs.webkit.org/show_bug.cgi?id=111490

Reviewed by Beth Dakin.

This is clearer than setItem returning true on error.
Also, actually set the item in the storage map.

* UIProcess/Storage/StorageManager.cpp:
(StorageManager::StorageArea):
(WebKit::StorageManager::StorageArea::setItem):
(WebKit::StorageManager::setItem):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144838 268f45cc-cd09-0410-ab3c-d52691b4dbfc

ASSERTION FAILED: i < size(), UNKNOWN in WebCore::RenderTableSection::layout
https://bugs.webkit.org/show_bug.cgi?id=108838

Source/WebCore:

Reviewed by Levi Weintraub.

Ensure that RenderTable::recalcSections is called when moving a
TableSection by calling setNeedsSectionRecalc in addChild for
TableSections.

Test: fullscreen/full-screen-table-section.html

* rendering/RenderTable.cpp:
(WebCore::RenderTable::addChild):

LayoutTests:

Reviewed by Levi Weintraub.

Add test for calling requestFullscreen on an element with a
table-section display type.

* fullscreen/full-screen-table-section-expected.txt: Added.
* fullscreen/full-screen-table-section.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144837 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed. Skip tests on Windows to make EWS bots happier.

* platform/win/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144835 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed gardening.

* platform/chromium/TestExpectations: fullscreen/full-screen-iframe-allowed-prexied.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144834 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Upgrade ANGLE to r1987
https://bugs.webkit.org/show_bug.cgi?id=111381

Reviewed by Dean Jackson.

Source/ThirdParty/ANGLE:

* include/GLSLANG/ShaderLang.h:
* src/common/version.h:
* src/compiler/Compiler.cpp:
(TCompiler::TCompiler):
(TCompiler::Init):
(TCompiler::compile):
* src/compiler/ParseHelper.h:
(TParseContext):
* src/compiler/ShHandle.h:
(TCompiler):
* src/compiler/ShaderLang.cpp:
(ShInitBuiltInResources):
* src/compiler/UnfoldShortCircuit.cpp:
(sh::UnfoldShortCircuit::visitSelection):
* src/compiler/glslang.l:
* src/compiler/glslang.y:
* src/compiler/glslang_lex.cpp:
(glslang_scan):
* src/compiler/glslang_tab.cpp:
     Updated ANGLE to r1987.

Source/WebCore:

Covered by existing tests. Ran WebGL conformance tests in Safari
to ensure correct operation of patch.

* platform/graphics/blackberry/GraphicsContext3DBlackBerry.cpp:
(WebCore::GraphicsContext3D::GraphicsContext3D):
* platform/graphics/cairo/GraphicsContext3DCairo.cpp:
(WebCore::GraphicsContext3D::GraphicsContext3D):
* platform/graphics/efl/GraphicsContext3DEfl.cpp:
(WebCore::GraphicsContext3D::GraphicsContext3D):
* platform/graphics/mac/GraphicsContext3DMac.mm:
(WebCore::GraphicsContext3D::GraphicsContext3D):
* platform/graphics/qt/GraphicsContext3DQt.cpp:
(WebCore::GraphicsContext3DPrivate::initializeANGLE):
    Changed all ports to query for highp support in fragment
    shaders and conditionally enable it in ANGLE shader translator.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144833 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Use new assertion API for process suppression on Mac
https://bugs.webkit.org/show_bug.cgi?id=111387

Patch by Kiran Muppala <cmuppala@apple.com> on 2013-03-05
Reviewed by Alexey Proskuryakov.

Source/WebKit2:

* Shared/ChildProcess.h:
(WebKit::ChildProcess::processSuppressionEnabled): Rename
m_processVisibleAssertion to m_processSuppressionAssertion.
* Shared/mac/ChildProcessMac.mm:
(WebKit::ChildProcess::setProcessSuppressionEnabled): Ditto. And,
replace calls to WKNSProcessInfoProcessAssertionWithTypes() with call to
-[NSProcessInfo beginSuspensionOfSystemBehaviors:].

WebKitLibraries:

Add WKProcessSuppressionSystemBehaviors constant for use in WebKit2 to
take a process suppression assertion.

* WebKitSystemInterface.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144832 268f45cc-cd09-0410-ab3c-d52691b4dbfc

IndexedDB: Add IDBBackingStoreTest
https://bugs.webkit.org/show_bug.cgi?id=111233

Reviewed by Adam Barth.

Source/WebCore:

These are tweaks to make IDBBackingStore more easily testable.

* Modules/indexeddb/IDBBackingStore.cpp:
(WebCore::IDBBackingStore::IDBBackingStore): Allow use without a backing store.

Source/WebKit/chromium:

Add some really simple IDBBackingStore tests
as a seed for additional tests later.

* WebKit.gypi:
* tests/IDBBackingStoreTest.cpp: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144830 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Move storageType() to StorageAreaProxy
https://bugs.webkit.org/show_bug.cgi?id=111488

Reviewed by Andreas Kling.

* WebProcess/Storage/StorageAreaProxy.cpp:
(WebKit::StorageAreaProxy::StorageAreaProxy):
(WebKit::StorageAreaProxy::storageType):
(WebKit::StorageAreaProxy::disabledByPrivateBrowsingInFrame):
* WebProcess/Storage/StorageAreaProxy.h:
(StorageAreaProxy):
* WebProcess/Storage/StorageNamespaceProxy.cpp:
* WebProcess/Storage/StorageNamespaceProxy.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144829 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remove old Mac test expectations for the bug 106412.

* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144828 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[chromium] Move check for threaded compositing from WebCompositorSupport to Platform
https://bugs.webkit.org/show_bug.cgi?id=111386

Reviewed by Adrienne Walker.

Source/Platform:

This moves the query for if threaded compositing is enabled from WebCompositorSupport
to Platform. This is part of an effort to remove statefulness on WebCompositorSupport
and because it's much easier to override Platform functions for different tests.

* chromium/public/Platform.h:
(WebKit::Platform::isThreadedCompositingEnabled):
(Platform):
* chromium/public/WebCompositorSupport.h:

Source/WebCore:

This moves checks for whether threaded compositing is enabled from WebCompositorSupport
to Platform. In the case of Canvas2DLayerBridge, the check is moved completely out of
the class to make it easier for tests to specify the threaded configuration to use in the test.

* platform/graphics/chromium/Canvas2DLayerBridge.cpp:
(WebCore::Canvas2DLayerBridge::Canvas2DLayerBridge):
* platform/graphics/chromium/Canvas2DLayerBridge.h:
(WebCore::Canvas2DLayerBridge::create):
(Canvas2DLayerBridge):
* platform/graphics/chromium/DrawingBufferChromium.cpp:
(WebCore::DrawingBuffer::DrawingBuffer):
* platform/graphics/skia/ImageBufferSkia.cpp:
(WebCore::createAcceleratedCanvas):

Source/WebKit/chromium:

* src/WebViewImpl.cpp:
(WebKit::WebViewImpl::composite):
(WebKit::WebViewImpl::scheduleAnimation):
(WebKit::WebViewImpl::setIsAcceleratedCompositingActive):
(WebKit::WebViewImpl::scheduleComposite):
* tests/Canvas2DLayerBridgeTest.cpp:
(Canvas2DLayerBridgeTest::fullLifecycleTest):
* tests/Canvas2DLayerManagerTest.cpp:
(FakeCanvas2DLayerBridge::FakeCanvas2DLayerBridge):

Tools:

* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::initialize):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144827 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Split up the storage event dispatch functions further
https://bugs.webkit.org/show_bug.cgi?id=111482

Reviewed by Beth Dakin.

Add new functions for dispatching storage events to a set of frames.

* storage/StorageAreaImpl.cpp:
(WebCore::StorageAreaImpl::dispatchStorageEvent):
* storage/StorageEventDispatcher.cpp:
(WebCore::StorageEventDispatcher::dispatchSessionStorageEvents):
(WebCore::StorageEventDispatcher::dispatchLocalStorageEvents):
(WebCore):
(WebCore::StorageEventDispatcher::dispatchSessionStorageEventsToFrames):
(WebCore::StorageEventDispatcher::dispatchLocalStorageEventsToFrames):
* storage/StorageEventDispatcher.h:
(WebCore):
(StorageEventDispatcher):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144826 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Crash when ImageLoader deletes Element inside SVGImageElement
https://bugs.webkit.org/show_bug.cgi?id=111085

Reviewed by Abhishek Arya.

Source/WebCore:

Elements with ImageLoader objects associated with them may have their
final reference held by the ImageLoader (to allow events to be sent
and handled). Any call on Element that causes the ImageLoader to
dispatch events might then result in the final deref of the Element
itself, thus leaving all the Element's "this" pointers up the stack
pointing to invalid memory.

This change puts the deref of the Element on a timer so that, if the
deref is called via a method on Element, the call stack will unwind
before the deref occurs.

Test: svg/custom/image-with-attr-change-after-delete-crash.html

* loader/ImageLoader.cpp:
(WebCore::ImageLoader::ImageLoader): Initialize a timer
(WebCore::ImageLoader::updatedHasPendingEvent): Put deref of the
element on a oneShotTimer, with appropriate assertions and checks to
ensure we only ref/deref once.
(WebCore::ImageLoader::timerFired): Deref the element when the timer fires.
* loader/ImageLoader.h:
(ImageLoader): Define a timer for controlling deref of the element.

LayoutTests:

* svg/custom/image-with-attr-change-after-delete-crash-expected.txt: Added.
* svg/custom/image-with-attr-change-after-delete-crash.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144825 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: identify reflection layers in LayerTreeAgent
https://bugs.webkit.org/show_bug.cgi?id=111419

Source/WebCore:

We add a new property "isReflection" to the Layer objects returned
to the front-end to indicate whether the layer was created to host
the reflection for an element. When this new property is true, the
"nodeId" property is set to point to the reflected element's node
id, since there is no node per-se directly associated with this layer.
We also change the "bounds" property to return absolute page coordinates
such that the front-end can use this information to highlight that
region in the page since metrics would otherwise be unavailable.

Reviewed by Dean Jackson.

Test: inspector-protocol/layers/layers-reflected-content.html

* inspector/Inspector.json:
* inspector/InspectorLayerTreeAgent.cpp:
(WebCore::InspectorLayerTreeAgent::buildObjectForLayer):

LayoutTests:

New test for reflection layers and an updated output for the
existing layersForNode() test since we changed the coordinates
for the "bounds" property to be absolute instead of local.

Reviewed by Dean Jackson.

* inspector-protocol/layers/layers-for-node-expected.txt:
* inspector-protocol/layers/layers-reflected-content-expected.txt: Added.
* inspector-protocol/layers/layers-reflected-content.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@144824 268f45cc-cd09-0410-ab3c-d52691b4dbfc