[WebKit-https.git] / Source / WebCore / ChangeLog

2012-10-24  Seokju Kwon  <seokju.kwon@samsung.com>

        [SOUP] Provide logging for SocketStreamHandleSoup
        https://bugs.webkit.org/show_bug.cgi?id=100215

        Reviewed by Alexey Proskuryakov.

        Add logging to SocketStreamHandleSoup to facilitate debugging.

        * platform/network/soup/SocketStreamHandleSoup.cpp:
        (WebCore::SocketStreamHandle::SocketStreamHandle):
        (WebCore::SocketStreamHandle::~SocketStreamHandle):
        (WebCore::SocketStreamHandle::platformSend):
        (WebCore::SocketStreamHandle::platformClose):

2012-10-24  Nico Weber  <thakis@chromium.org>

        Pass on exif orientation from ImageSource when using the open-source image decoders
        https://bugs.webkit.org/show_bug.cgi?id=100164

        Reviewed by Eric Seidel.

        The orientation is honored always for image documents, and optionally
        for <img> elements if shouldRespectImageOrientation is set (off by
        default). However, the feature needs port-specific modifications to
        GraphicsContext (without this, webcore will use the rotated bounds but
        draw the image as if it hadn't be rotated, resulting in squished
        pixels), and most ports don't implement these yet -- so put
        turning this on for image documents behind a port-specific #ifdef.

        No observable behavior change. Once it's hooked up, it will be tested
        by fast/images/exif-orientation.html.

        * platform/graphics/ImageSource.cpp:
        (WebCore::ImageSource::size):
        (WebCore::ImageSource::frameSizeAtIndex):
        (WebCore::ImageSource::orientationAtIndex):
        * platform/graphics/chromium/DeferredImageDecoder.cpp:
        (WebCore::DeferredImageDecoder::orientation):
        (WebCore):
        * platform/graphics/chromium/DeferredImageDecoder.h:
        (DeferredImageDecoder):
        * platform/image-decoders/jpeg/JPEGImageDecoder.cpp:
        (WebCore::readImageOrientation):
        * rendering/RenderObject.cpp:
        (WebCore::RenderObject::shouldRespectImageOrientation):

2012-10-24  Beth Dakin  <bdakin@apple.com>

        https://bugs.webkit.org/show_bug.cgi?id=100169
        We should make TileCache tiles the size of the tile coverage rect 
        when we can't do fast scrolling
        -and-
        <rdar://problem/12505021>

        Reviewed by Simon Fraser.

        Some websites that don't do fast scrolling still scroll slower than 
        they do with tiled drawing disabled. 
        https://bugs.webkit.org/show_bug.cgi?id=99768 addressed some of this 
        performance problem, but there is still more ground to make up. This 
        patch addresses the remaining issue by making tiles the size of the 
        window when we can't do fast scrolling. 

        The constructor and create function no longer take a size parameter. 
        That's all fully controlled within TileCache now. m_tileSize is no 
        longer const.
        * platform/graphics/ca/mac/TileCache.h:

        Store the current default size as constants so that we can access it 
        in both the constructor and adjustTileSizeForCoverageRect().
        * platform/graphics/ca/mac/TileCache.mm:
        (WebCore::TileCache::TileCache):

        This new function will set m_tileSize to the size of the tile 
        coverage rect if the tile coverage is limited to the visible area. 
        Otherwise, the tiles are set to be the default size.
        (WebCore::TileCache::adjustTileSizeForCoverageRect):
        
        Call adjustTileSizeForCoverageRect().
        (WebCore::TileCache::revalidateTiles):

        No need to send in a size anymore.
        * platform/graphics/ca/mac/WebTileCacheLayer.h:
        (WebCore):

2012-10-24  David Barton  <dbarton@mathscribe.com>

        MathML tests trigger font cache assertions in debug bots
        https://bugs.webkit.org/show_bug.cgi?id=100268

        Reviewed by Eric Seidel.

        Add a FontCachePurgePreventer to fix this.

        Tested by existing tests.

        * rendering/mathml/RenderMathMLOperator.cpp:
        (WebCore::RenderMathMLOperator::updateFromElement):

2012-10-24  Tim Horton  <timothy_horton@apple.com>

        Update main frame scroll position immediately for programmatic scrolls
        https://bugs.webkit.org/show_bug.cgi?id=98074

        Reviewed by Simon Fraser.

        Immediately update the main frame scroll position after a programmatic scroll,
        so that performing a scrollBy or scrollTo will be instantly reflected in
        all of the scroll offset accessors.

        No new tests; this will be tested by many scrolling tests once WebKitTestRunner
        can use threaded scrolling.

        * page/scrolling/ScrollingCoordinator.cpp:
        (WebCore::ScrollingCoordinator::requestScrollPositionUpdate):

2012-10-24  Mihai Parparita  <mihaip@chromium.org>

        [Chromium] Allow pushState and related history APIs to be disabled per context
        https://bugs.webkit.org/show_bug.cgi?id=99780

        Reviewed by Adam Barth.

        Chrome Apps do not support the history API (or navigation in general).
        Since pushState is generally feature detected, it's cleanest to disable
        it via a V8 per-context feature, so that applications can have the
        appropriate fallback behavior (other history APIs are re-mapped to throw
        exceptions, since history.back() and the link are not feature detected).

        * dom/ContextFeatures.cpp:
        (WebCore::ContextFeatures::pushStateEnabled):
        (WebCore):
        * dom/ContextFeatures.h:
        * dom/Document.cpp:
        (WebCore::Document::enqueuePopstateEvent):
        * page/History.idl:

2012-10-24  Tommy Widenflycht  <tommyw@google.com>

        MediaStream API: Make sure all events are dispatched asynchronously
        https://bugs.webkit.org/show_bug.cgi?id=100286

        Reviewed by Adam Barth.

        This is necessary to safeguard against if the UA uses synchronous UA->WebKit calls,
        and the web application calls methods on the RTCPeerConnection in the event callbacks.

        Test: fast/mediastream/RTCPeerConnection-events.html
        Also tested by the chromium webrtc fuzz tests.

        * Modules/mediastream/RTCPeerConnection.cpp:
        (WebCore::RTCPeerConnection::RTCPeerConnection):
        (WebCore::RTCPeerConnection::negotiationNeeded):
        (WebCore::RTCPeerConnection::didGenerateIceCandidate):
        (WebCore::RTCPeerConnection::didAddRemoteStream):
        (WebCore::RTCPeerConnection::didRemoveRemoteStream):
        (WebCore::RTCPeerConnection::didAddRemoteDataChannel):
        (WebCore::RTCPeerConnection::changeReadyState):
        (WebCore::RTCPeerConnection::changeIceState):
        (WebCore):
        (WebCore::RTCPeerConnection::scheduleDispatchEvent):
        (WebCore::RTCPeerConnection::scheduledEventTimerFired):
        * Modules/mediastream/RTCPeerConnection.h:
        (RTCPeerConnection):

2012-10-24  Mark Pilgrim  <pilgrim@chromium.org>

        [Chromium] Remove screen-related functions from PlatformSupport
        https://bugs.webkit.org/show_bug.cgi?id=97474

        Reviewed by Adam Barth.

        Screen-related functions like screenHorizontalDPI that
        used to be on PlatformSupport are now accessed through a new
        PlatformPageClient attached to Widget in WebCore-land, which is
        implemented by ChromeClientImpl in WebKit-land, which proxies
        calls to WebWidgetClient, which is actually implemented in
        Chromium-land.

        * WebCore.gypi:
        * platform/Widget.h:
        * platform/chromium/PageClientChromium.h: Copied from Source/WebCore/platform/chromium/PlatformWidget.h.
        (PageClientChromium):
        * platform/chromium/PlatformScreenChromium.cpp:
        (WebCore::toPlatformPageClient):
        (WebCore):
        (WebCore::screenHorizontalDPI):
        (WebCore::screenVerticalDPI):
        (WebCore::screenDepth):
        (WebCore::screenDepthPerComponent):
        (WebCore::screenIsMonochrome):
        (WebCore::screenRect):
        (WebCore::screenAvailableRect):
        * platform/chromium/PlatformSupport.h:
        (PlatformSupport):

2012-10-24  Adam Barth  <abarth@webkit.org>

        [V8] WorkerContextExecutionProxy doesn't need to track events
        https://bugs.webkit.org/show_bug.cgi?id=100295

        Reviewed by Eric Seidel.

        This code was added when this code was originally upstreamed as part of
        the Chromium port. There doesn't appear to be any reason why
        WorkerContextExecutionProxy needs to keep a Vector of raw event
        pointers. Those points are likely to become dangling, making the rest
        of what this code does very sketchy.

        * bindings/v8/V8WorkerContextEventListener.cpp:
        (WebCore::V8WorkerContextEventListener::callListenerFunction):
        * bindings/v8/WorkerContextExecutionProxy.cpp:
        (WebCore::WorkerContextExecutionProxy::dispose):
        * bindings/v8/WorkerContextExecutionProxy.h:
        (WebCore::WorkerContextExecutionState::WorkerContextExecutionState):
        (WorkerContextExecutionProxy):

2012-10-24  Max Vujovic  <mvujovic@adobe.com>

        [CSS Shaders] The mesh should be specified using <column, row> order
        https://bugs.webkit.org/show_bug.cgi?id=96285

        Reviewed by Dean Jackson.

        Change StyleResolver and CSSComputedStyleDeclaration to handle the mesh parameters in
        column, row order.

        Test: css3/filters/custom/custom-filter-mesh-column-row-order.html

        * css/CSSComputedStyleDeclaration.cpp:
        (WebCore::CSSComputedStyleDeclaration::valueForFilter):
            Append the number of columns before the number of rows in the meshParameters
            CSSValueList.
        * css/StyleResolver.cpp:
        (WebCore::StyleResolver::createCustomFilterOperation):
            Store the second mesh parameter in meshRows instead of meshColumns.

2012-10-24  Jae Hyun Park  <jae.park@company100.net>

        loaderRunLoop() should use synchronization instead of while loop
        https://bugs.webkit.org/show_bug.cgi?id=55402

        Reviewed by Alexey Proskuryakov.

        Originally, loaderRunLoop() sleeps until a thread has started and set
        the loaderRunLoopObject static variable. This patch uses
        ThreadCondition to synchronize instead of a while loop.

        No new tests (No behavior change).

        * platform/network/cf/LoaderRunLoopCF.cpp:
        (WebCore::runLoaderThread):
        (WebCore::loaderRunLoop):
        * platform/network/cf/LoaderRunLoopCF.h:
        (WebCore):

2012-10-24  Sailesh Agrawal  <sail@chromium.org>

        Incorrect keycodes for numpad /, -, +, .
        https://bugs.webkit.org/show_bug.cgi?id=99188

        Reviewed by Tony Chang.

        In r57951 we switched to mapping keys by character code.
        This regressed the numpad keys which no longer match the Windows virtual key codes.
        This CL fixes this by never mapping numpad keys by character code.

        Test: platform/mac/fast/events/numpad-keycode-mapping.html

        * platform/mac/PlatformEventFactoryMac.mm:
        (WebCore::windowsKeyCodeForKeyEvent):

2012-10-24  Simon Fraser  <simon.fraser@apple.com>

        Null-check the RenderView in ocument::windowScreenDidChange to fix a crash when saving PDFs
        https://bugs.webkit.org/show_bug.cgi?id=100141
        <rdar://problem/12559147>

        Reviewed by Tim Horton.

        For PDF documents the RenderView is null, so null-check it to avoid a crash
        when saving PDFs.

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

2012-10-24  Terry Anderson  <tdanderson@chromium.org>

        Handle two-finger tap gestures in the same way as long-press gestures
        https://bugs.webkit.org/show_bug.cgi?id=99947

        Reviewed by Adam Barth.

        Currently a long-press gesture is used to dispatch a context menu (for platforms 
        defining CONTEXT_MENUS) or to select text (for Android). Additionally, for platforms 
        defining TOUCH_ADJUSTMENT, gesture target fuzzing is performed on the location and 
        touch area of the long-press gesture.

        This CL will cause two-finger tap gestures to be handled in the same way as long-press 
        gestures. The location and touch area of a two-finger tap gesture will correspond to 
        the location and touch area of the first finger down; the location/area of the second 
        finger will be ignored.

        Test: touchadjustment/touch-links-two-finger-tap.html

        * page/EventHandler.cpp:
        (WebCore::EventHandler::handleGestureLongPress):
        (WebCore):
        (WebCore::EventHandler::handleGestureForTextSelectionOrContextMenu):
        (WebCore::EventHandler::handleGestureTwoFingerTap):
        (WebCore::EventHandler::adjustGesturePosition):
        * page/EventHandler.h:
        (EventHandler):

2012-10-24  Chris Fleizach  <cfleizach@apple.com>

        AX:When aria-label is used, the text under an element is still appearing as the AXTitle
        https://bugs.webkit.org/show_bug.cgi?id=98167

        Reviewed by Beth Dakin.

        According to WAI-ARIA text computation, the presence of aria-label and alternative text
        should override the visible text within an element.

        Test: platform/mac/accessibility/aria-label-overrides-visible-text.html

        * accessibility/mac/WebAccessibilityObjectWrapper.mm:
        (-[WebAccessibilityObjectWrapper accessibilityTitle]):

2012-10-24  Sheriff Bot  <webkit.review.bot@gmail.com>

        Unreviewed, rolling out r132303 and r132312.
        http://trac.webkit.org/changeset/132303
        http://trac.webkit.org/changeset/132312
        https://bugs.webkit.org/show_bug.cgi?id=100287

        Triggering crashes on many popular websites (Requested by
        leviw|gardening on #webkit).

        * css/CSSGrammar.y.in:
        * css/CSSParser.cpp:
        (WebCore::CSSParser::detectAtToken):
        * css/CSSParser.h:
        * css/CSSPropertySourceData.h:
        * css/RuleSet.cpp:
        (WebCore::RuleData::RuleData):
        (WebCore::RuleSet::addRule):
        (WebCore::RuleSet::addRegionRule):
        (WebCore::RuleSet::addRulesFromSheet):
        (WebCore::RuleSet::addStyleRule):
        * css/RuleSet.h:
        (RuleData):
        (RuleSet):
        * css/StyleResolver.cpp:
        (WebCore::makeRuleSet):
        (WebCore::StyleResolver::appendAuthorStyleSheets):
        (WebCore::StyleResolver::matchScopedAuthorRules):
        (WebCore::StyleResolver::locateSharedStyle):
        * css/StyleResolver.h:
        (StyleResolver):
        * css/StyleRule.cpp:
        (WebCore::StyleRuleBase::reportMemoryUsage):
        (WebCore::StyleRuleBase::destroy):
        (WebCore::StyleRuleBase::copy):
        (WebCore::StyleRuleBase::createCSSOMWrapper):
        * css/StyleRule.h:
        * css/StyleScopeResolver.cpp:
        (WebCore::StyleScopeResolver::reportMemoryUsage):
        * css/StyleScopeResolver.h:
        (WebCore):
        (StyleScopeResolver):
        * css/StyleSheetContents.cpp:
        (WebCore::childRulesHaveFailedOrCanceledSubresources):

2012-10-24  Joshua Bell  <jsbell@chromium.org>

        IndexedDB: Cursor property value identities should be preserved
        https://bugs.webkit.org/show_bug.cgi?id=100051

        Reviewed by Tony Chang.

        Some W3C test submissions point out that subsequent accesses to cursor properties should
        yield the same value until the cursor advances. We handled this with custom binding code for
        IDBCursor.value but not IDBCursor.key or IDBCursor.primaryKey. The custom value code is
        being removed in webkit.org/b/100034 in favor of returning ScriptValue and the same fix can
        be applied to key/primaryKey.

        Test: storage/indexeddb/cursor-properties.html

        * Modules/indexeddb/IDBCursor.cpp: Compute/store/serve up ScriptValues instead of IDBKeys
        (WebCore::IDBCursor::key):
        (WebCore::IDBCursor::primaryKey):
        (WebCore::IDBCursor::setValueReady):
        * Modules/indexeddb/IDBCursor.h:
        (IDBCursor):
        (WebCore::IDBCursor::idbPrimaryKey): Expose this for callers that need access to the IDBKey
        * Modules/indexeddb/IDBCursor.idl:
        * Modules/indexeddb/IDBObjectStore.cpp: ... like this one.
        (WebCore):
        * Modules/indexeddb/IDBRequest.cpp:
        (WebCore::IDBRequest::dispatchEvent): Pass along script context, for the conversion.
        * bindings/v8/IDBBindingUtilities.cpp:
        (WebCore::idbKeyToScriptValue): New method for explicit conversion.
        (WebCore):
        * bindings/v8/IDBBindingUtilities.h:
        (WebCore):

2012-10-24  Ami Fischman  <fischman@chromium.org>

        call to setNeedsLayout during RenderVideo::paintReplaced
        https://bugs.webkit.org/show_bug.cgi?id=100265

        Reviewed by Eric Carlson.

        Removed unnecessary call and added new defensive guards to catch erroneous setNeedsLayout() calls
        during paints earlier (so the offending calls are in the emitted stacktrace).

        No new tests - new defensive checks are triggered by existing tests.

        * page/FrameView.cpp:
        (WebCore::FrameView::paintContents): forbid setNeedsLayout() during painting
        * rendering/RenderObject.cpp:
        (WebCore):
        (WebCore::RenderObject::SetLayoutNeededForbiddenScope::SetLayoutNeededForbiddenScope):
        (WebCore::RenderObject::SetLayoutNeededForbiddenScope::~SetLayoutNeededForbiddenScope):
        * rendering/RenderObject.h:
        (RenderObject):
        (SetLayoutNeededForbiddenScope): added helper class for forbidding setNeedsLayout() in a scope.
        * rendering/RenderVideo.cpp:
        (WebCore::RenderVideo::paintReplaced): drop the offending & unnecessary call to updatePlayer().

2012-10-24  Adam Barth  <abarth@webkit.org>

        [V8] ActiveDOMObjectEpilogueVisitor is unnecessary and can be deleted
        https://bugs.webkit.org/show_bug.cgi?id=100208

        Reviewed by Eric Seidel.

        Rather than clearing and re-establishing the weak callback for
        ActiveDOMObjects during every GC, this patch puts all the
        ActiveDOMObjects with pending activity into an object group with a live
        object, causing them not to be garbage collected.

        In addition to simplifying this code, this patch makes the patch in
        https://bugs.webkit.org/show_bug.cgi?id=100175 much easier because
        V8GCController no longer needs to know how to configure the weak
        callbacks for these objects.

        * bindings/v8/V8GCController.cpp:
        (WebCore::ActiveDOMObjectPrologueVisitor::ActiveDOMObjectPrologueVisitor):
        (ActiveDOMObjectPrologueVisitor):
        (WebCore::ActiveDOMObjectPrologueVisitor::visitDOMWrapper):
        (WebCore::V8GCController::majorGCPrologue):
        (WebCore::V8GCController::majorGCEpilogue):
        * bindings/v8/V8PerIsolateData.cpp:
        (WebCore::V8PerIsolateData::V8PerIsolateData):
        * bindings/v8/V8PerIsolateData.h:
        (WebCore::V8PerIsolateData::liveRoot):
        (V8PerIsolateData):

2012-10-24  Brady Eidson  <beidson@apple.com>

        Add a strategy for loader customization.
        https://bugs.webkit.org/show_bug.cgi?id=100278

        Reviewed by Alexey Proskuryakov.

        It's empty for now and does nothing, just like the goggles.

        * WebCore.gypi:
        * WebCore.vcproj/WebCore.vcproj:
        * WebCore.xcodeproj/project.pbxproj:
        * platform/PlatformStrategies.h:
        (WebCore::PlatformStrategies::loaderStrategy):
        (PlatformStrategies):
        (WebCore::PlatformStrategies::PlatformStrategies):

2012-10-24  Tony Chang  <tony@chromium.org>

        Setting width of a flexitem causes the adjacent flex item to be displayed poorly.
        https://bugs.webkit.org/show_bug.cgi?id=99925

        Reviewed by Ojan Vafai.

        Make sure that we always repaint when moving a child. This is similar to what RenderDeprecatedFlexibleBox does.

        Test: css3/flexbox/repaint-during-resize-no-flex.html

        * rendering/RenderFlexibleBox.cpp:
        (WebCore::RenderFlexibleBox::setFlowAwareLocationForChild): Move logic for repaining into the helper method
        for setting the location of a child.
        (WebCore::RenderFlexibleBox::layoutColumnReverse): Remove code for repaint since it's now in setFlowAwareLocationForChild.
        (WebCore::RenderFlexibleBox::adjustAlignmentForChild): Remove code for repaint since it's now in setFlowAwareLocationForChild.

2012-10-24  Simon Fraser  <simon.fraser@apple.com>

        Fix CALayer hiearchy when combining tiling with preserve-3d
        https://bugs.webkit.org/show_bug.cgi?id=100205

        Reviewed by Dean Jackson.

        When an element has "transform-style: preserve-3d", its GraphicsLayerCA has a
        m_structuralLayer which is a CATransformLayer. The primary CALayer which contains rendered
        content becomes a sublayer of the CATransformLayer. If the element has backface-visibility:hidden,
        it is the primary layer that is set to be single-sided.
        
        In r131940 we started to use TileCaches in place of CATiledLayer. TileCaches work via
        "customSublayers" returned from the PlatformCALayer, where the custom sublayer is 
        the tile cache container layer. However, the custom sublayers were being added as
        children of the structural (CATransformLayer) layer, not of the primary (CALayer) layer,
        thus they were not affected by the doubleSided property.
        
        This change cleans up the confusing code in GraphicsLayerCA::updateSublayerList()
        by maintaining two vectors of PlatformCALayers, one for sublayers of the structural
        layer, and one for sublayers of the primary layer. It adds custom sublayers to
        the latter list, so now the tile cache container layer becomes a sublayer of
        the primary layer, so is affected by that layer's doubleSided property.

        Test: compositing/tiling/backface-preserve-3d-tiled.html

        * platform/graphics/ca/GraphicsLayerCA.cpp:
        (WebCore::GraphicsLayerCA::updateSublayerList):

2012-10-23  Zhenyao Mo  <zmo@google.com>

        Update mozilla's CheckedInt.h to the latest version
        https://bugs.webkit.org/show_bug.cgi?id=100177

        Reviewed by Kenneth Russell.

        * html/canvas/CheckedInt.h: Sync with mozilla's copy with minumum modifications.
        (detail):
        (IsSupportedPass2):
        (IsSupported):
        (UnsignedType):
        (IsSigned):
        (TwiceBiggerType):
        (PositionOfSignBit):
        (MinValue):
        (MaxValue):
        (WebCore::detail::HasSignBit):
        (WebCore::detail::BinaryComplement):
        (WebCore::detail::IsInRange):
        (WebCore::detail::IsAddValid):
        (WebCore::detail::IsSubValid):
        (WebCore::detail::IsMulValid):
        (WebCore::detail::IsDivValid):
        (WebCore::detail::OppositeIfSignedImpl::run):
        (WebCore::detail::OppositeIfSigned):
        (WebCore):
        (CheckedInt):
        (WebCore::CheckedInt::CheckedInt):
        (WebCore::CheckedInt::value):
        (WebCore::CheckedInt::isValid):
        (WebCore::CheckedInt::operator -):
        (WebCore::CheckedInt::operator ==):
        (WebCore::CheckedInt::operator++):
        (WebCore::CheckedInt::operator--):
        (CastToCheckedIntImpl):
        (WebCore::detail::CastToCheckedIntImpl::run):
        (WebCore::castToCheckedInt):
        (WebCore::operator ==):
        * html/canvas/DataView.cpp: change valid() to isValid().
        (WebCore::DataView::create):
        * html/canvas/WebGLBuffer.cpp: Ditto.
        (WebCore::WebGLBuffer::associateBufferDataImpl):
        (WebCore::WebGLBuffer::associateBufferSubDataImpl):
        * html/canvas/WebGLRenderingContext.cpp: Ditto.
        (WebCore):
        (WebCore::WebGLRenderingContext::drawArrays):
        * platform/graphics/GraphicsContext3D.cpp: Ditto.
        (WebCore::GraphicsContext3D::computeImageSizeInBytes):

2012-10-24  Noam Rosenthal  <noam.rosenthal@nokia.com>

        [Qt-on-Mac] GraphicsSurfaces should not create a global IOSurface handle
        https://bugs.webkit.org/show_bug.cgi?id=89885

        Reviewed by Kenneth Rohde Christiansen.

        Use mach_port instead of global tokens for IOSurfaces.
        Global IOSurfaces are accessible from other processes using their handle, while mach_ports
        can only be shared directly via IPC.

        Tested by existing WebGL tests.

        * platform/graphics/surfaces/GraphicsSurfaceToken.h:
        (GraphicsSurfaceToken):
        * platform/graphics/surfaces/mac/GraphicsSurfaceMac.cpp:
        (WebCore::GraphicsSurfacePrivate::GraphicsSurfacePrivate):
        (WebCore::GraphicsSurfacePrivate::~GraphicsSurfacePrivate):

2012-10-24  Rick Byers  <rbyers@chromium.org>

        image-set doesn't round-trip properly with cssText
        https://bugs.webkit.org/show_bug.cgi?id=99725

        Reviewed by Beth Dakin.

        Fix serialization of -webkit-image-set rules to use the same format as
        is used for parsing.

        Test: fast/css/image-set-setting.html

        * css/CSSImageSetValue.cpp:
        (WebCore::CSSImageSetValue::customCssText):
        * css/CSSValueList.h:
        (WebCore::CSSValueList::item): Add const overload

2012-10-24  Jonathan Feldstein  <jfeldstein@rim.com>

        BlackBerry fix for webgl-depth-texture.html
        https://bugs.webkit.org/show_bug.cgi?id=100258

        Reviewed by Yong Li.

        Removed a platform specific define that is unnecessary.

        * platform/graphics/GraphicsContext3D.cpp:
        (WebCore::GraphicsContext3D::computeFormatAndTypeParameters):

2012-10-24  Nico Weber  <thakis@chromium.org>

        Honor image orientation in GraphicsContextSkia
        https://bugs.webkit.org/show_bug.cgi?id=100179

        Reviewed by Stephen White.

        Also fix a bug in ImageOrientation.h: Some of the values were
        switched. They now match the description in the exif spec at
        http://www.exif.org/Exif2-2.PDF page 18, and the notes at
        http://www.sno.phy.queensu.ca/~phil/exiftool/TagNames/EXIF.html and
        the picture at http://www.80sidea.com/archives/2316.

        They showed up correctly in Safari because the orientation transform
        was done after flipping Y. Let the flipping happen later. I verified
        that flipping transformation order and fixing the exif enum cancelled
        each other out, so this has no effect on the apple port.

        Not hooked up yet, so no observable difference.

        * platform/graphics/BitmapImage.h:
        * platform/graphics/ImageOrientation.h:
        (WebCore::ImageOrientation::usesWidthAsHeight):
        (WebCore::ImageOrientation::fromEXIFValue):
        * platform/graphics/cg/GraphicsContextCG.cpp:
        (WebCore::GraphicsContext::drawNativeImage):
        * platform/graphics/skia/ImageSkia.cpp:
        (WebCore::paintSkBitmap):
        (WebCore::BitmapImage::draw):
        (WebCore):

2012-10-24  Alexey Proskuryakov  <ap@apple.com>

        Fixed Windows path for SharedWorkerStrategy.h - it's in workers/, not platform/.

        * WebCore.vcproj/WebCore.vcproj:

2012-10-24  Chris Fleizach  <cfleizach@apple.com>

        AX: WebKit exposes abstract ARIA role range as AXSlider; should be generic AXGroup
        https://bugs.webkit.org/show_bug.cgi?id=100204

        Reviewed by Beth Dakin.

        Remove the "range" role as a valid ARIA role.

        Test: platform/mac/accessibility/aria-slider-value.html

        * accessibility/AccessibilityObject.cpp:
        (WebCore::createARIARoleMap):

2012-10-24  Florin Malita  <fmalita@chromium.org>

        [Chromium] SVG repaint issues
        https://bugs.webkit.org/show_bug.cgi?id=99874

        Reviewed by Levi Weintraub.

        RenderSVGRoot::paintReplaced() converts the container offsets to a relative transform,
        but in doing so it ends up accumulating subpixel rounding deltas twice: first for 
        adjustedPaintOffset and second in localToParentTransform(). If coordinates are on
        half-pixel boundaries, the 2x rounding delta yields a full pixel drift and we end up
        painting at the wrong location.

        This can be avoided by using localToBorderBoxTransform() directly, which (unlike
        localToParentTransform()) doesn't perform rounding.

        No new tests: existing pixel results cover this change after rebaseline.

        * rendering/svg/RenderSVGRoot.cpp:
        (WebCore::RenderSVGRoot::paintReplaced):

2012-10-24  Levi Weintraub  <leviw@chromium.org>

        Fixing the Chromium build after typo in r132367.

        * WebCore.gypi:

2012-10-24  Nico Weber  <thakis@chromium.org>

        [chromium] Respect image orientation in image dragging code
        https://bugs.webkit.org/show_bug.cgi?id=100200

        Reviewed by Tony Chang.

        Based on the corresponding code in DragImageMac.mm.  The image
        rotation code isn't hooked up yet, so no observable effect for now.

        * platform/chromium/DragImageChromiumSkia.cpp:
        (WebCore::createDragImageFromImage):

2012-10-24  Michael Saboff  <msaboff@apple.com>

        Code cleanup after r132165
        https://bugs.webkit.org/show_bug.cgi?id=100135

        Reviewed by Geoffrey Garen.

        Fixed up some unneccesary and inefficient constructs in MarkupTokenBase.h.

        Code clean up without functional changes, therefore no new tests.

        * xml/parser/MarkupTokenBase.h:
        (WebCore::MarkupTokenBase::beginEndTag): Changed argument to be const Vector<LChar, 32>&.
        (WebCore::MarkupTokenBase::appendToCharacter): Changed argument to be const Vector<LChar, 32>&.
        (WebCore::MarkupTokenBase::appendToName): Elimintated inline.
        (WebCore::MarkupTokenBase::name): Elimintated inline.
        (WebCore::MarkupTokenBase::nameString): Elimintated inline.

2012-10-24  Toni Barzic  <tbarzic@chromium.org>

        Crash when trying to write exception message to null console
        https://bugs.webkit.org/show_bug.cgi?id=99658

        Reviewed by Adam Barth.

        DOMWindow::console may return NULL, so we should do a NULL check before adding message to it.
        This may happen e.g. if a worker throws an exception just as the document is being replaced in the view.
        The exception task could be processes after current window in the frame changes, and console in the document window is nulled.

        Test: fast/workers/worker-exception-during-navigation.html

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

2012-10-24  Cosmin Truta  <ctruta@rim.com>

        Incorrect conditional use of LogFTP
        https://bugs.webkit.org/show_bug.cgi?id=100260

        Reviewed by Alexey Proskuryakov.

        Use LogFTP if !LOG_DISABLED, to allow toggling ASSERTIONS_DISABLED_DEFAULT
        without breaking the build.

        * html/FTPDirectoryDocument.cpp:
        (WebCore::FTPDirectoryDocument::FTPDirectoryDocument):

2012-10-23  Alexey Proskuryakov  <ap@apple.com>

        Add a strategy for shared workers
        https://bugs.webkit.org/show_bug.cgi?id=100165

        Reviewed by Brady Eidson.

        Also a little bit of alphabetization.

        * WebCore.gypi:
        * WebCore.vcproj/WebCore.vcproj:
        * WebCore.xcodeproj/project.pbxproj:
        * platform/PlatformStrategies.h:
        * workers/SharedWorkerStrategy.h: Added.

2012-10-24  David Barton  <dbarton@mathscribe.com>

        [MathML] Timeouts on linux after r132264
        https://bugs.webkit.org/show_bug.cgi?id=100202

        Reviewed by Eric Seidel.

        When building a stretched operator, such as a large parenthesis or bracket, we need to
        check that the extension glyph's height is > 0, to avoid an infinite loop. The 0 height
        can occur if the glyph is missing on the host system.

        Tested by existing tests.

        * rendering/mathml/RenderMathMLOperator.cpp:
        (WebCore::RenderMathMLOperator::updateFromElement):

2012-10-24  Parth Patel  <parpatel@rim.com>

        [BlackBerry] Extending existing ThreadUnsafe singletons in webkit to
        Generic ThreadUnsafe Singleton
        https://bugs.webkit.org/show_bug.cgi?id=100145

        Reviewed by Rob Buis.

        Extending the singletons to generic singleton and changing
        getInstance() in IconDatabaseClientBlackBerry to instance()
        to match the generic singleton template.

        Reviewed Internally by Yong Li.

        Tests are not modified and added as there was no behavioural change.

        * platform/blackberry/SharedTimerBlackBerry.cpp:
        (SharedTimerBlackBerry):
        (WebCore):

2012-10-24  Ilya Tikhonovsky  <loislo@chromium.org>

        Web Inspector: NMI add instrumentation for ResourceBuffer. It gives us 10mb on gmail
        https://bugs.webkit.org/show_bug.cgi?id=100254

        Reviewed by Yury Semikhatsky.

        * loader/ResourceBuffer.cpp:
        (WebCore::ResourceBuffer::reportMemoryUsage):
        (WebCore):
        * loader/ResourceBuffer.h:
        (ResourceBuffer):

2012-10-24  Eric Carlson  <eric.carlson@apple.com>

        Allow ports to override text track rendering style
        https://bugs.webkit.org/show_bug.cgi?id=97800
        <rdar://problem/12044964>

        Unreviewed Build fix.
        
        CGFloat is already a float in a 32-bit build so narrowPrecisionToFloat() is unnecessary

        * page/CaptionUserPreferencesMac.mm:
        (WebCore::CaptionUserPreferencesMac::captionFontSizeScale):

2012-10-24  Eric Carlson  <eric.carlson@apple.com>

        Allow ports to override text track rendering style
        https://bugs.webkit.org/show_bug.cgi?id=97800
        <rdar://problem/12044964>

        Reviewed by Maciej Stachowiak.

        * WebCore.exp.in: Export new WebkitSystemInterface functions.
        * WebCore.xcodeproj/project.pbxproj: Add CaptionUserPreferences.h, CaptionUserPreferencesMac.mm,
            and CaptionUserPreferencesMac.h.

        * css/mediaControls.css: Rearrange the caption CSS so it is possible to style the cue window,
            background, and text independently.

        * html/HTMLMediaElement.cpp:
        (WebCore::HTMLMediaElement::HTMLMediaElement): Initialize m_disableCaptions with theme->userPrefersCaptions().
        (WebCore::HTMLMediaElement::attach): Register for caption preferences change callbacks.
        (WebCore::HTMLMediaElement::detach): Unregister for caption preferences change callbacks.
        (WebCore::HTMLMediaElement::userPrefersCaptions): Return theme->userPrefersCaptions().
        (WebCore::HTMLMediaElement::userIsInterestedInThisTrackKind): Consider userPrefersCaptions().
        (WebCore::HTMLMediaElement::setClosedCaptionsVisible): Move the code that marks all tracks as
            un-configured to markCaptionAndSubtitleTracksAsUnconfigured so it can be reused.
        (WebCore::HTMLMediaElement::captionPreferencesChanged): New, force a reevaluation of all text tracks.
        (WebCore::HTMLMediaElement::markCaptionAndSubtitleTracksAsUnconfigured): New, code moved from 
            setClosedCaptionsVisible
        * html/HTMLMediaElement.h: Inherit from CaptionPreferencesChangedListener.

        * html/shadow/MediaControlElements.cpp:
        (WebCore::MediaControlTextTrackContainerElement::updateSizes): Get rid of unnecessary member
            variable. Get caption font scale from theme instead of hard coding.
        * html/shadow/MediaControlElements.h:

        * html/track/TextTrack.cpp:
        (WebCore::TextTrack::TextTrack): Change attributes from String to AtomicString.
        (WebCore::TextTrack::isValidKindKeyword): Ditto.
        (WebCore::TextTrack::setKind): Ditto.
        (WebCore::TextTrack::setMode): Ditto.
        (WebCore::TextTrack::mode): Ditto.
        * html/track/TextTrack.h:
        (WebCore::TextTrack::create): Ditto.
        (WebCore::TextTrack::kind): Ditto.
        (WebCore::TextTrack::label): Ditto.
        (WebCore::TextTrack::setLabel): Ditto.
        (WebCore::TextTrack::language): Ditto.
        (WebCore::TextTrack::setLanguage): Ditto.

        * html/track/TextTrackCue.cpp:
        (WebCore::TextTrackCueBox::TextTrackCueBox): Set the shadow pseudo id.
        (WebCore::TextTrackCueBox::textTrackCueBoxShadowPseudoId): New, class method to return the 
            shadow pseudo id so it can be used elsewhere.
        (WebCore::TextTrackCueBox::shadowPseudoId): Call textTrackCueBoxShadowPseudoId.
        (WebCore::TextTrackCue::pastNodesShadowPseudoId): New, class method to return the 
            shadow pseudo id so it can be used elsewhere.
        (WebCore::TextTrackCue::futureNodesShadowPseudoId): Ditto.
        (WebCore::TextTrackCue::updateDisplayTree):
        * html/track/TextTrackCue.h:

        * page/CaptionUserPreferences.h: Added.
        * page/CaptionUserPreferencesMac.h: Added.
        * page/CaptionUserPreferencesMac.mm: Added.
        (WebCore::userCaptionPreferencesChangedNotificationCallback):
        (WebCore::CaptionUserPreferencesMac::CaptionUserPreferencesMac):
        (WebCore::CaptionUserPreferencesMac::~CaptionUserPreferencesMac):
        (WebCore::CaptionUserPreferencesMac::userHasCaptionPreferences): New, passthrough to WKSI function.
        (WebCore::CaptionUserPreferencesMac::userPrefersCaptions): Ditto.
        (WebCore::CaptionUserPreferencesMac::captionsWindowColor): Return Color with user's caption window color preference.
        (WebCore::CaptionUserPreferencesMac::captionsBackgroundColor): Return Color with user's caption 
            background color preference.
        (WebCore::CaptionUserPreferencesMac::captionsTextColor): Return Color with user's caption text color preference.
        (WebCore::CaptionUserPreferencesMac::captionsEdgeColorForTextColor): Return Color for text edge effect.
        (WebCore::CaptionUserPreferencesMac::cssPropertyWithTextEdgeColor): Return String with CSS to set a text-shadow
            or webkit-text-stroke property.
        (WebCore::CaptionUserPreferencesMac::cssColorProperty): Return a String with css to set a property 
            with a color value.
        (WebCore::CaptionUserPreferencesMac::captionsTextEdgeStyle): Return a String with css to style caption 
            text with the user's preferred text edge stye.
        (WebCore::CaptionUserPreferencesMac::captionsDefaultFont): Return a String with css to style caption
            text with the user's preferred font.
        (WebCore::CaptionUserPreferencesMac::captionsStyleSheetOverride): Return a String with css to style captions
            with the user's preferred style.
        (WebCore::CaptionUserPreferencesMac::captionFontSizeScale): Return the user's preferred caption font scale.
        (WebCore::CaptionUserPreferencesMac::captionPreferencesChanged): Notify listeners of caption preference change.
        (WebCore::CaptionUserPreferencesMac::registerForCaptionPreferencesChangedCallbacks): Add a caption preferences 
            changes listener.
        (WebCore::CaptionUserPreferencesMac::unregisterForCaptionPreferencesChangedCallbacks): Remove a caption preferences
            changes listener.
        (WebCore::CaptionUserPreferencesMac::updateCaptionStyleSheetOveride): New, if theme has a captions style sheet override,
            inject it into the current page group, otherwise remove injected sheet.

        * page/PageGroup.cpp:
        (WebCore::PageGroup::captionPreferences):
        (WebCore::PageGroup::registerForCaptionPreferencesChangedCallbacks): New, passthrough to platform specific function
            of the same name.
        (WebCore::PageGroup::unregisterForCaptionPreferencesChangedCallbacks): Ditto.
        (WebCore::PageGroup::userPrefersCaptions): Ditto.
        (WebCore::PageGroup::userHasCaptionPreferences): Ditto.
        (WebCore::PageGroup::captionFontSizeScale): Ditto.
        * page/PageGroup.h:

        * platform/mac/WebCoreSystemInterface.h: Updated.
        * platform/mac/WebCoreSystemInterface.mm: Ditto.

2012-10-24  Vsevolod Vlasov  <vsevik@chromium.org>

        Web Inspector: Introduce workspace provider as a content providing backend for project.
        https://bugs.webkit.org/show_bug.cgi?id=100244

        Reviewed by Pavel Feldman.

        Introduced WorkspaceProvider interface as a content providing backend for project.
        Added NetworkWorkspaceProvider as a network based (default) implementation.

        * WebCore.gypi:
        * WebCore.vcproj/WebCore.vcproj:
        * inspector/compile-front-end.py:
        * inspector/front-end/CompilerScriptMapping.js:
        (WebInspector.CompilerScriptMapping):
        * inspector/front-end/DebuggerScriptMapping.js:
        (WebInspector.DebuggerScriptMapping):
        * inspector/front-end/NetworkUISourceCodeProvider.js:
        (WebInspector.NetworkUISourceCodeProvider):
        (WebInspector.NetworkUISourceCodeProvider.prototype._parsedScriptSource):
        (WebInspector.NetworkUISourceCodeProvider.prototype._resourceAdded):
        (WebInspector.NetworkUISourceCodeProvider.prototype._addFile):
        * inspector/front-end/NetworkWorkspaceProvider.js: Added.
        (WebInspector.NetworkWorkspaceProvider):
        (WebInspector.NetworkWorkspaceProvider.prototype.requestFileContent):
        (WebInspector.NetworkWorkspaceProvider.prototype.setFileContent):
        (WebInspector.NetworkWorkspaceProvider.prototype.searchInFileContent):
        (WebInspector.NetworkWorkspaceProvider.prototype.addFile):
        (WebInspector.NetworkWorkspaceProvider.prototype.removeFile):
        (WebInspector.NetworkWorkspaceProvider.prototype.reset):
        * inspector/front-end/SASSSourceMapping.js:
        (WebInspector.SASSSourceMapping):
        (_bindUISourceCode):
        * inspector/front-end/ScriptSnippetModel.js:
        (WebInspector.ScriptSnippetModel):
        (WebInspector.ScriptSnippetModel.prototype._addScriptSnippet):
        * inspector/front-end/WebKit.qrc:
        * inspector/front-end/Workspace.js:
        (WebInspector.FileDescriptor):
        (WebInspector.WorkspaceProvider):
        (WebInspector.WorkspaceProvider.prototype.requestFileContent):
        (WebInspector.WorkspaceProvider.prototype.searchInFileContent):
        (WebInspector.WorkspaceProvider.prototype.addEventListener):
        (WebInspector.WorkspaceProvider.prototype.removeEventListener):
        (WebInspector.Project):
        (WebInspector.Project.prototype.reset):
        (WebInspector.Project.prototype._fileAdded):
        (WebInspector.Project.prototype._fileRemoved):
        (WebInspector.Project.prototype.requestFileContent):
        (WebInspector.Project.prototype.searchInFileContent):
        (WebInspector.Workspace):
        (WebInspector.Workspace.prototype.addProject):
        (WebInspector.Workspace.prototype.project):
        * inspector/front-end/inspector.html:
        * inspector/front-end/inspector.js:

2012-10-24  Nikita Vasilyev  <me@elv1s.ru>

        Web Inspector: Styles pane: Don't select whole value when I select just a part
        https://bugs.webkit.org/show_bug.cgi?id=100242

        Reviewed by Alexander Pavlov.

        * inspector/front-end/StylesSidebarPane.js:
        (WebInspector.StylePropertiesSection.prototype._handleEmptySpaceClick):
        (WebInspector.StylePropertyTreeElement.prototype):

2012-10-24  Vsevolod Vlasov  <vsevik@chromium.org>

        Web Inspector: UiSourceCode should rely on the workspace as a content provider.
        https://bugs.webkit.org/show_bug.cgi?id=100216

        Reviewed by Pavel Feldman.

        Workspace is now passed to UISourceCode to be used as a content provider.
        Content providers are now stored in the workspace/project.
        Next step would be to move content providers to workspace providers.

        * inspector/front-end/BreakpointManager.js:
        (WebInspector.BreakpointManager.breakpointStorageId):
        * inspector/front-end/UISourceCode.js:
        (WebInspector.UISourceCode):
        (WebInspector.UISourceCode.prototype.contentType):
        (WebInspector.UISourceCode.prototype.requestContent):
        (WebInspector.UISourceCode.prototype.requestOriginalContent):
        (WebInspector.UISourceCode.prototype._commitContent):
        (WebInspector.UISourceCode.prototype.searchInContent):
        * inspector/front-end/Workspace.js:
        (WebInspector.WorkspaceController.prototype._mainFrameNavigated):
        (WebInspector.Project):
        (WebInspector.Project.prototype.addUISourceCode):
        (WebInspector.Project.prototype.removeUISourceCode):
        (WebInspector.Project.prototype.uiSourceCodes):
        (WebInspector.Project.prototype.requestFileContent):
        (WebInspector.Project.prototype.searchInFileContent):
        (WebInspector.Workspace):
        (WebInspector.Workspace.prototype.addTemporaryUISourceCode):
        (WebInspector.Workspace.prototype.removeTemporaryUISourceCode):
        (WebInspector.Workspace.prototype.reset):

2012-10-24  Zeno Albisser  <zeno@webkit.org>

        Implement GraphicsSurface for Windows.
        https://bugs.webkit.org/show_bug.cgi?id=98147

        Reviewed by Kenneth Rohde Christiansen.

        * Target.pri:
            Include GraphicsSurfaceWin.cpp in SOURCES on Windows.
        * platform/graphics/surfaces/GraphicsSurface.h:
            Add typedef for PlatformGraphicsSurface on Windows.
        * platform/graphics/surfaces/GraphicsSurfaceToken.h:
            Add typedef for BufferHandle on Windows.
        (GraphicsSurfaceToken):
        * platform/graphics/surfaces/win/GraphicsSurfaceWin.cpp: Added.
            The GraphicsSurface implementation on Windows relies on the
            availability of ANGLE and the EGL_ANGLE_query_surface_pointer extension.
            For Qt this requirements are implicitly satisfied, when Qt is built
            on Windows and QT_CONFIG contains OpenGLES2.
            The GraphicsSurface then renders a given texture onto an offscreen
            pixel buffer surface, queries the surface pointer using the
            EGL_ANGLE_query_surface_pointer extension, and transmits the received
            surface pointer (share handle) over IPC.
            On the UIProcess side, the surface pointer can then be resolved
            using eglCreatePbufferFromClientBuffer.
        (WebCore):
        (WebCore::loadShader):
            Initialize the shaders needed for drawing onto the GraphicsSurface.
        (GraphicsSurfacePrivate):
        (WebCore::GraphicsSurfacePrivate::GraphicsSurfacePrivate):
            In case of the instance being on the WebProcess side,
            create an EGLContext that shares the texture objects with the provided
            share context. Also create two pixel buffer surfaces, one as front- and
            one as backbuffer.
            Query the surface pointers for the pixel buffer surfaces and initialize
            the GraphicsSurfaceToken that can be passed over IPC later.
        (WebCore::GraphicsSurfacePrivate::~GraphicsSurfacePrivate):
            Release all aquired resources and destroy the pixel buffer surfaces.
            Also close the EGL-Display connection.
        (WebCore::GraphicsSurfacePrivate::copyFromTexture):
            Make the belonging context current on the back buffer surface
            and use drawTexture() to draw the provided texture onto the GraphicsSurface.
        (WebCore::GraphicsSurfacePrivate::makeCurrent):
            Save the previously current context, then make the context belonging
            to the GraphicsSurface current.
        (WebCore::GraphicsSurfacePrivate::doneCurrent):
            Restore the context that was current before calling makeCurrent().
        (WebCore::GraphicsSurfacePrivate::swapBuffers):
            Swap front and back buffer surfaces and handles.
        (WebCore::GraphicsSurfacePrivate::token):
        (WebCore::GraphicsSurfacePrivate::frontBufferTextureID):
            This function is meant to be called from the UIProcess side.
            If no front buffer surface has been created before for the current
            front buffer handle, one will be created.
            Then eglBindTexImage will be used to actually bind the current
            front buffer surface to a texture as a source for drawing.
        (WebCore::GraphicsSurfacePrivate::initialFrontBufferShareHandle):
        (WebCore::GraphicsSurfacePrivate::frontBufferShareHandle):
        (WebCore::GraphicsSurfacePrivate::backBufferShareHandle):
        (WebCore::GraphicsSurfacePrivate::releaseFrontBufferTexture):
            Free the resources related to the front buffer surface.
            On the UIProcess side we never actually bind the back buffer.
        (WebCore::GraphicsSurfacePrivate::initializeShaderProgram):
            Initialize and link the shader programs necessary for drawing
            onto the GraphicsSurface.
        (WebCore::GraphicsSurfacePrivate::createSurfaceFromShareHandle):
            Creates a single pixel buffer surface from a share Handle.
            This function will be called on the UIProcess side,
            for the front buffer handle, whenever the buffers have been swapped.
        (WebCore::GraphicsSurfacePrivate::drawTexture):
            The WebProcess uses this function to draw a given
            texture onto the GraphicsSurface's back buffer.
        (WebCore::GraphicsSurface::platformExport):
        (WebCore::GraphicsSurface::platformGetTextureID):
        (WebCore::GraphicsSurface::platformCopyToGLTexture):
        (WebCore::GraphicsSurface::platformCopyFromTexture):
        (WebCore::GraphicsSurface::platformPaintToTextureMapper):
            Uses TextureMapperGL::drawTexture() to draw the front buffer texture
            to the TextureMapper on the UIProcess side.
        (WebCore::GraphicsSurface::platformFrontBuffer):
        (WebCore::GraphicsSurface::platformSwapBuffers):
        (WebCore::GraphicsSurface::platformCreate):
        (WebCore::GraphicsSurface::platformImport):
        (WebCore::GraphicsSurface::platformLock):
        (WebCore::GraphicsSurface::platformUnlock):
        (WebCore::GraphicsSurface::platformDestroy):

2012-09-27  Yury Semikhatsky  <yurys@chromium.org>

        Web Inspector: provide memory instrumentation for ListHashSet
        https://bugs.webkit.org/show_bug.cgi?id=97786

        Reviewed by Vsevolod Vlasov.

        Switched existing usages of addListHashSet to the generic instrumentation mechanism
        as it should work just just fine now that there is a memory instrumentation of
        ListHashSet.

        * dom/DocumentEventQueue.cpp:
        * dom/DocumentStyleSheetCollection.cpp:
        (WebCore::DocumentStyleSheetCollection::reportMemoryUsage):
        * loader/cache/CachedResourceLoader.cpp:
        (WebCore::CachedResourceLoader::reportMemoryUsage):

2012-10-24  Charles Wei  <charles.wei@torchmobile.com.cn>

        [BlackBerry] Credentials not re-used for a redirected request to the same domain
        https://bugs.webkit.org/show_bug.cgi?id=100193

        Reviewed by George Staikos.

        We should store the credentials in the redirection response handler, if the request is challenged.
        Because the redirect response suggests the authentication succeeds. 

        Test: http://browsertest01.rim.net/authbasic

        * platform/network/blackberry/NetworkJob.cpp:
        (WebCore::NetworkJob::handleRedirect):

2012-10-24  Alexander Pavlov  <apavlov@chromium.org>

        Web Inspector: Implement CSS reload upon related SASS resource saving
        https://bugs.webkit.org/show_bug.cgi?id=98024

        Reviewed by Vsevolod Vlasov.

        SASS-generated debug info in CSS is parsed to find out which SASS files contributed to this stylesheet.
        Upon SASS file save in the Sources panel, all affected external CSS stylesheets are reloaded to update
        the page styles (presuming that SASS is running in the "watch" mode during the development cycle).

        * English.lproj/localizedStrings.js:
        * inspector/front-end/SASSSourceMapping.js:
        (WebInspector.SASSSourceMapping):
        (WebInspector.SASSSourceMapping.prototype._fileSaveFinished.callback):
        (WebInspector.SASSSourceMapping.prototype._reloadCSS):
        (_bindUISourceCode):
        (_addCSSURLforSASSURL):
        * inspector/front-end/Settings.js:
        * inspector/front-end/SettingsScreen.js:
        (WebInspector.GenericSettingsTab):

2012-10-23  Yury Semikhatsky  <yurys@chromium.org>

        Memory instrumentation: don't count agent-specific front-ends separately
        https://bugs.webkit.org/show_bug.cgi?id=100087

        Reviewed by Alexander Pavlov.

        Pointers to domain-specific inspector front-end interfaces are reported as
        weak pointers instead of members to avoid double-counting.

        Test by comparing set of reported instrumented objects with the set of objects
        allocated by tcmalloc.

        * inspector/InspectorDOMStorageAgent.cpp:
        (WebCore::InspectorDOMStorageAgent::reportMemoryUsage):
        * inspector/InspectorDOMStorageResource.cpp:
        (WebCore::InspectorDOMStorageResource::reportMemoryUsage):
        * inspector/InspectorProfilerAgent.cpp:
        (WebCore::InspectorProfilerAgent::reportMemoryUsage):

2012-10-23  Christophe Dumez  <christophe.dumez@intel.com>

        Regression(r132303) Broke debug build when SHADOW_DOM is enabled but STYLE_SCOPED is disabled
        https://bugs.webkit.org/show_bug.cgi?id=100203

        Unreviewed Build fix.

        Fix compilation error in StyleResolver.h when SHADOW_DOM flag is
        set and STYLE_SCOPED is not.

        No new tests, no behavior change.

        * css/StyleResolver.h:
        (WebCore::StyleResolver::ensureScopeResolver):

2012-10-23  Andreas Kling  <kling@webkit.org>

        Remove specialized hash traits for GlyphPages.
        <http://webkit.org/b/100185>

        Reviewed by Dan Bernstein.

        Now that the default minimum table size has been lowered for all tables, there's no need
        for this specialization anymore.

        * platform/graphics/FontFallbackList.h:
        (FontFallbackList):

2012-10-23  Kent Tamura  <tkent@chromium.org>

        Multiple fields input UI: Don't use CSS properties for physical direction
        https://bugs.webkit.org/show_bug.cgi?id=100195

        Reviewed by Kentaro Hara.

        CSS declarations such as "padding: 0 0 0 1px;" "margin-left: 0.2em;" are
        not direction-neutral, and makes unexpected appearance in RTL. We should
        use -webkit-padding-start or -webkit-margin-start.

        No new tests. Covered by date-multiple-fields-appearance-l10n.html and
        month-multiple-fields-appearance-l10n.html

        * css/html.css:
        (input[type="date"]): Use padding:0 and -webkit-padding-start:1px;
        (input[type="datetime"]): Ditto.
        (input[type="datetime-local"]): Ditto.
        (input[type="month"]): Ditto.
        (input[type="time"]): Ditto.
        (input[type="week"]): Ditto.
        (input[type="week"]::-webkit-inner-spin-button):
        Use -webkit-margin-start. Also use an integral pixel size instead of
        avoid a fractional relative size to avoid subpixel layout rounding.

2012-10-23  Dan Bernstein  <mitz@apple.com>

        The font cache evicts inactive font data too aggressively when not under memory pressure
        https://bugs.webkit.org/show_bug.cgi?id=100194

        Reviewed by Sam Weinig.

        Increasing the number of inactive font data objects the cache may hold increases the hit
        rate, leading to improved page load performance. When memory pressure is detected,
        MemoryPressureHandler evicts all inactive objects, regardless of the limits we are
        increasing here.

        * platform/graphics/FontCache.cpp:
        (WebCore): Increased the maximum number of inactive font data objects in the cache from 50
        to 225, and the number of objects to evict once the limit is met from 20 to 25.

2012-10-23  Takashi Sakamoto  <tasak@google.com>

        [Shadow DOM] Needs @host rule for ShadowDOM styling
        https://bugs.webkit.org/show_bug.cgi?id=88606

        Reviewed by Hajime Morita.

        Implemented @host-@rules according to the shadow dom spec:
        http://dvcs.w3.org/hg/webcomponents/raw-file/tip/spec/shadow/index.html#host-at-rule
        The design doc is:
        https://docs.google.com/document/d/1P2yorchF8lci2sccr-mVSRf2dtvjKeiuyQWzCB2bEA8/edit

        Test: fast/dom/shadow/athost-atrules.html

        * css/CSSGrammar.y.in:
        Added rules for parsing @host @-rules.
        * css/CSSParser.cpp:
        (WebCore::CSSParser::detectAtToken):
        Added a new token "@host".
        (WebCore::CSSParser::createHostRule):
        Added a new method to create an @host @-rule, which is invoked from
        (WebCore):
        * css/CSSParser.h:
        Added a declaration of the above new method: createHostRule.
        * css/CSSPropertySourceData.h:
        Added HOST_RULE to enum Type.
        * css/RuleSet.cpp:
        (WebCore::RuleData::RuleData):
        Modified multiple bool arguments into one argument. Now it
        uses combinations of values from enum AddRuleFlags.
        (WebCore::RuleSet::addRule):
        (WebCore::RuleSet::addRegionRule):
        (WebCore::RuleSet::addStyleRule):
        Updated according to the RuleData's change.
        Modified to Invoke increaseSpecificity if the given rule is @host
        @-rule.
        (WebCore::RuleSet::addRulesFromSheet):
        Invoked addHostRule if the given rule is @host @-rule.
        * css/RuleSet.h:
        (RuleData):
        (WebCore::RuleData::increaseSpecificity):
        Added a new method to increase selector's specificity. This method is
        used to make @host @-rules' specificity larger than normal author
        rules' specificity.
        (RuleSet):
        * css/StyleResolver.cpp:
        (WebCore::makeRuleSet):
        (WebCore::StyleResolver::addHostRule):
        A wrapper method to invoke StyleScopeResolver::addHostRule. The method
        is used to make only StyleResolver know an implementation detail about
        class StyleScopeResolver.
        (WebCore::StyleResolver::appendAuthorStylesheets):
        Updated according to the RuleData's change.
        (WebCore::StyleResolver::styleSharingCandidateMatchesHostRules):
        A new method to find matched host rules when an element is given.
        This method invokes
        StyleScopeResolver::styleSharingCandidateMatchesHostRules to find
        matched host rules.
        (WebCore):
        (WebCore::StyleResolver::matchHostRules):
        A new method to find matched host rules when an element is given.
        This method invokes StyleScopeResolver::matchHostRules to find
        matched host rules.
        (WebCore::StyleResolver::matchScopedAuthorRules):
        Modified to invoke matchHostRules.
        (WebCore::StyleResolver::locateSharedStyle):
        Disable sibling style cache if the given element is a shadow host and
        any @host @-rules are applied to the element.
        * css/StyleResolver.h:
        (WebCore::StyleResolver::ensureScopeResolver):
        If no scopeResolver is created, create and return the instance.
        If created, just return the instance.
        (StyleResolver):
        * css/StyleRule.cpp:
        (WebCore::StyleRuleBase::reportMemoryUsage):
        (WebCore::StyleRuleBase::destroy):
        (WebCore::StyleRuleBase::copy):
        (WebCore::StyleRuleBase::createCSSOMWrapper):
        (WebCore::StyleRuleHost::StyleRuleHost):
        Implemented class StyleRuleHost. The class is almost the same as
        StyleRuleBlock except type.
        (WebCore):
        * css/StyleRule.h:
        (WebCore::StyleRuleBase::isHostRule):
        (StyleRuleHost):
        (WebCore::StyleRuleHost::create):
        (WebCore::StyleRuleHost::copy):
        (WebCore):
        * css/StyleScopeResolver.cpp:
        (WebCore::StyleScopeResolver::ensureAtHostRuleSetFor):
        A new method to create a new RuleSet for the given shadow root.
        (WebCore):
        (WebCore::StyleScopeResolver::atHostRuleSetFor):
        A new private inline method to obtain @host @-rules declared in
        the given shadow root.
        (WebCore::StyleScopeResolver::addHostRule):
        Added a new method to register @host @-rules with shadow roots.
        (WebCore::StyleScopeResolver::styleSharingCandidateMatchesHostRules):
        A new method to find whether any @host @-rules are applied to
        the given host element.
        (WebCore::StyleScopeResolver::matchHostRules):
        A new method to find matched rules for the given host element.
        (WebCore::StyleScopeResolver::reportMemoryUsage):
        * css/StyleScopeResolver.h:
        (WebCore):
        (StyleScopeResolver):
        * css/StyleSheetContents.cpp:
        (WebCore::childRulesHaveFailedOrCanceledSubresources):

2012-10-23  Andreas Kling  <kling@webkit.org>

        REGRESSION(r130643): ASSERTION FAILED: result.iterator != end() below PluginDatabase::add
        <http://webkit.org/b/100065>

        Reviewed by Anders Carlsson.

        Restore the pre-r130643 hash table size for PluginDatabase to prevent these easily reproducible
        collisions. This will need a proper fix at some point.

        * plugins/PluginDatabase.h:
        * plugins/PluginPackage.h:
        (PluginPackageHashTraits):

2012-10-23  Simon Fraser  <simon.fraser@apple.com>

        Allow tests to run with a tile cache, and optionally dump the tile cache via layerTreeAsText()
        https://bugs.webkit.org/show_bug.cgi?id=100159

        Reviewed by Tim Horton.

        Add a LAYER_TREE_INCLUDES_PAGE_TILE_CACHE flag for layerTreeAsText(), and plumb it
        through to GraphicsLayerCA, which dumps the TiledBacking tile coverage rect.

        Test: platform/mac/tiled-drawing/use-tiled-drawing.html

        * page/Frame.h:
        * platform/graphics/GraphicsLayer.cpp:
        (showGraphicsLayerTree):
        * platform/graphics/GraphicsLayer.h:
        (WebCore::GraphicsLayer::tiledBacking):
        * platform/graphics/TiledBacking.h:
        * platform/graphics/ca/GraphicsLayerCA.cpp:
        (WebCore::GraphicsLayerCA::tiledBacking):
        (WebCore::GraphicsLayerCA::dumpAdditionalProperties):
        * platform/graphics/ca/GraphicsLayerCA.h:
        (GraphicsLayerCA):
        * platform/graphics/ca/mac/TileCache.h:
        (WebCore::TileCache::tileCoverageRect):
        (TileCache):
        * platform/graphics/ca/mac/TileCache.mm:
        (WebCore::TileCache::computeTileCoverageRect):
        (WebCore::TileCache::revalidateTiles):
        * rendering/RenderLayerCompositor.cpp:
        (WebCore::RenderLayerCompositor::layerTreeAsText):
        * testing/Internals.cpp:
        (WebCore::Internals::layerTreeAsText):
        * testing/Internals.h:
        * testing/Internals.idl:

2012-10-23  Simon Fraser  <simon.fraser@apple.com>

        Flush pending GraphicsLayer changes when reattaching compositing layers if necessary
        https://bugs.webkit.org/show_bug.cgi?id=100187
        <rdar://problem/12546770>

        Reviewed by Dan Bernstein.

        Fix a regression from r131940. That revision changed RenderLayerCompositor::flushPendingLayerChanges()
        to bail if the root layer is unattached, which indicates that we're in the page cache, or in a
        background tab. However, that dropped the layer flush on the floor, so that any subsequent changes
        to GraphicsLayerCAs would just pile up and never get flushed. This was most evident on pages
        that require frequent flushing, such as those running animated GIFs.
        
        Fix by setting a flag in flushPendingLayerChanges() if we're unattached. Consult the flag
        when re-attaching the root layer, and if it's set, flush the GraphicsLayers.
        
        Not testable because we can't test detaching and re-adding web views in DRT/WTR.

        * rendering/RenderLayerCompositor.cpp:
        (WebCore::RenderLayerCompositor::RenderLayerCompositor): Initialize m_shouldFlushOnReattach to false.
        (WebCore::RenderLayerCompositor::flushPendingLayerChanges): If the root layer attachment
        is RootLayerUnattached, set the m_shouldFlushOnReattach flag before returning.
        (WebCore::RenderLayerCompositor::attachRootLayer): If m_shouldFlushOnReattach, flush the layers,
        saying that we're the flush root.
        * rendering/RenderLayerCompositor.h:
        (RenderLayerCompositor): Add m_shouldFlushOnReattach flag.

2012-10-23  Sheriff Bot  <webkit.review.bot@gmail.com>

        Unreviewed, rolling out r132276.
        http://trac.webkit.org/changeset/132276
        https://bugs.webkit.org/show_bug.cgi?id=100189

        It broke the Qt-WK2 build intentionally (Requested by
        Ossy_night on #webkit).

        * page/Frame.h:
        * platform/graphics/GraphicsLayer.cpp:
        (showGraphicsLayerTree):
        * platform/graphics/GraphicsLayer.h:
        (WebCore::GraphicsLayer::tiledBacking):
        * platform/graphics/TiledBacking.h:
        * platform/graphics/ca/GraphicsLayerCA.cpp:
        (WebCore::GraphicsLayerCA::tiledBacking):
        (WebCore::GraphicsLayerCA::dumpAdditionalProperties):
        * platform/graphics/ca/GraphicsLayerCA.h:
        (GraphicsLayerCA):
        * platform/graphics/ca/mac/TileCache.h:
        * platform/graphics/ca/mac/TileCache.mm:
        (WebCore::TileCache::tileCoverageRect):
        (WebCore::TileCache::revalidateTiles):
        * rendering/RenderLayerCompositor.cpp:
        (WebCore::RenderLayerCompositor::layerTreeAsText):
        * testing/Internals.cpp:
        (WebCore::Internals::layerTreeAsText):
        * testing/Internals.h:
        * testing/Internals.idl:

2012-10-23  Philip Rogers  <pdr@google.com>

        Add extra check for data() in PageSerializer.
        https://bugs.webkit.org/show_bug.cgi?id=99102

        Reviewed by Eric Seidel.

        The image returned from imageForRenderer() does not contain the raw SVG data
        so this patch adds a check for image->image()->data() before writing SVG
        in PageSerializer::addImageToResources.

        Covered by existing test WebPageNewSerializeTest.SVGImageDontCrash.

        * page/PageSerializer.cpp:
        (WebCore::PageSerializer::addImageToResources):

2012-10-23  Kent Tamura  <tkent@chromium.org>

        Move appendAsLDMLLiteral in LocaleWin.cpp to a common place
        https://bugs.webkit.org/show_bug.cgi?id=100129

        Reviewed by Kentaro Hara.

        We're going to use appendAsLDMLLiteral in other code, and it is
        related to DateTimeFormat class.  So we move it to DateTimeFormat
        class as quoteAndAppendLiteral.

        No new tests because of no behavior change.

        * platform/text/DateTimeFormat.cpp:
        (WebCore::DateTimeFormat::quoteAndAppendLiteral):
        Moved from LocaleWin.cpp
        * platform/text/DateTimeFormat.h:
        Declare StringBuilder by wtf/Forward.h. It also declares String.
        (DateTimeFormat): Declare quoteAndAppendLiteral.
        * platform/text/LocaleWin.cpp:
        (WebCore): Move appendLDMLLiteral to DateTimeFormat.
        (WebCore::convertWindowsDateFormatToLDML):
        Follow the moving.

2012-10-23  Kent Tamura  <tkent@chromium.org>

        REGRESSION(r131421): Text baseline is not aligned in some locales
        https://bugs.webkit.org/show_bug.cgi?id=100088

        Reviewed by Hajime Morita.

        Before this patch, the element with ::-webkit-datetime-edit had
        -webkit-align-items:center. It was incorrect at all because it
        ignores baselines of each of inner fields. We need to change it to
        -webkit-align-items:baseline, or stop using -webkit-flex.

        However -webkit-align-items:baseline doesn't work because a spin
        button element in the element doesn't have the baseline. If we
        specified -webkit-align-items:baseline, the spin button would
        shift up.

        So, we change the element structure for multiple fields input
        elements. Before this patch, the structure was:

        input
          |
           - ::-webkit-date-and-time-container (flex cotainer)
             |
             |- ::-webkit-datetime-edit (has flexibility in date-and-time-container, also this is a flex container)
             |  |- ::-webkit-datetime-edit-foo
             |  |      :
             |  |
             |  |- ::-webkit-date-time-edit-gap  (has flexibility in datetime-edit)
             |   - ::-webkit-inne-spin-button
              - ::-webkit-calendar-picker-indicator

        After the patch, the structure will be:

        input
          |
           - ::-webkit-date-and-time-container (flexible box)
             |
             |- ::-webkit-datetime-edit (has flexibility in date-and-time-container)
             |  |- ::-webkit-datetime-edit-foo
             |  |      :
             |  |
             |
             |- ::-webkit-inne-spin-button
              - ::-webkit-calendar-picker-indicator

        Because the spin button owner is date-and-time-container, we don't
        need to make the datetime-edit element a flex container.

        Also, we had rounding error by subpixel layout. To avoid it, we
        stop using fractional paddings.

        In order to implement this, we need to change the SpinButtonOwner
        interface provider from DateTimeEditElement to
        BaseMultipleFieldsDateAndTimeInputType.

        No new tests. Covered by month-multiple-fields-appearance-l10n.html.

        * css/html.css:
        (input::-webkit-datetime-edit): Stop making this a flex
        container. Inner fields in this use the single baseline.
        (input::-webkit-datetime-edit-ampm-field):
        Don't use fractional paddings to avoid rounding erros of subpixel
        layout.  Use integral margin instead to avoid overwrap of focus
        ring and background-color of the field.
        (input::-webkit-datetime-edit-day-field): Ditto.
        (input::-webkit-datetime-edit-hour-field): Ditto.
        (input::-webkit-datetime-edit-millisecond-field): Ditto.
        (input::-webkit-datetime-edit-minute-field): Ditto.
        (input::-webkit-datetime-edit-month-field): Ditto.
        (input::-webkit-datetime-edit-second-field): Ditto.
        (input::-webkit-datetime-edit-week-field): Ditto.
        (input::-webkit-datetime-edit-year-field): Ditto.
        (input::-webkit-date-and-time-container): Ditto.

        * html/BaseMultipleFieldsDateAndTimeInputType.h:
        (BaseMultipleFieldsDateAndTimeInputType): Overrides
        SpinButtonOwner functions, and add m_spinButtonElement.
        * html/BaseMultipleFieldsDateAndTimeInputType.cpp:
        (WebCore::BaseMultipleFieldsDateAndTimeInputType::focusAndSelectSpinButtonOwner):
        Just delegate to DateTimeEditElement::focusIfNoFocus.
        (WebCore::BaseMultipleFieldsDateAndTimeInputType::shouldSpinButtonRespondToMouseEvents):
        Moved from DateTimeEditElement::shouldSpinButtonRespondToMouseEvents.
        (WebCore::BaseMultipleFieldsDateAndTimeInputType::shouldSpinButtonRespondToWheelEvents):
        Moved from DateTimeEditElement::shouldSpinButtonRespondToWheelEvents.
        (WebCore::BaseMultipleFieldsDateAndTimeInputType::spinButtonStepDown):
        Delegate to DateTimeEditElement::stepDown.
        (WebCore::BaseMultipleFieldsDateAndTimeInputType::spinButtonStepUp):
        Delegate to DateTimeEditElement::stepUp.
        (WebCore::BaseMultipleFieldsDateAndTimeInputType::BaseMultipleFieldsDateAndTimeInputType):
        Initialize m_spinButtonElement.
        (WebCore::BaseMultipleFieldsDateAndTimeInputType::~BaseMultipleFieldsDateAndTimeInputType):
        Reset SpinButtonOwner for m_spinButtonElement like the old code of ~DateTimeEditElement.
        (WebCore::BaseMultipleFieldsDateAndTimeInputType::createShadowSubtree):
        Put a SpinButtonElement between the DateTimeEditElement and the PickerIndicatorElement.
        (WebCore::BaseMultipleFieldsDateAndTimeInputType::forwardEvent):
        Take care of SpinButtonElement.
        (WebCore::BaseMultipleFieldsDateAndTimeInputType::disabledAttributeChanged):
        Release capture of SpinButtonElement. This corresponds to the old
        code of DateTimeEditElement::updateUIState.
        (WebCore::BaseMultipleFieldsDateAndTimeInputType::readonlyAttributeChanged): Ditto.

        * html/TextFieldInputType.h:
        Make SpinButtonElement::SpinButtonOwner protected to allow overriding.

        * html/shadow/DateTimeEditElement.h:
        (DateTimeEditElement): Remove SpniButtonOwner implementation,
        m_spinButton, and rename some functions.
        * html/shadow/DateTimeEditElement.cpp:
        (WebCore::DateTimeEditElement::DateTimeEditElement):
        Remove m_spinButton.
        (WebCore::DateTimeEditElement::~DateTimeEditElement): Ditto.
        (WebCore::DateTimeEditElement::focusIfNoFocus):
        Renamed from focusAndSelectSpinButtonOwner. This is called
        BaseMultipleFieldsDateAndTimeInputType::focusAndSelectSpinButtonOwner.
        (WebCore::DateTimeEditElement::layout):
        Remove creation code for -webkit-datetime-edit-gap and SpinButtonElement.
        (WebCore::DateTimeEditElement::defaultEventHandler):
        Move the code for m_spinButton to
        BaseMultipleFieldsDateAndTimeInputType::forwardEvent.
        (WebCore::DateTimeEditElement::hasFocusedField): A helper for
        BaseMultipleFieldsDateAndTimeInputType::shouldSpinButtonRespondToWheelEvents.
        (WebCore::DateTimeEditElement::stepDown): A helper for
        BaseMultipleFieldsDateAndTimeInputType::spinButtonStepDown.
        (WebCore::DateTimeEditElement::stepUp): A helper for
        BaseMultipleFieldsDateAndTimeInputType::spinButtonStepUp.
        (WebCore::DateTimeEditElement::updateUIState): Move the code to
        BaseMultipleFieldsDateAndTimeInputType::disabledAttributeChanged
        and readonlyAttributeChanged.

2012-10-23  Andreas Kling  <kling@webkit.org>

        Shrink immutable ElementAttributeData and StylePropertySet by one pointer each.
        <http://webkit.org/b/100123>

        Reviewed by Anders Carlsson.

        Remove one pointer of unintentional padding in the immutable versions of these objects.
        583kB progression on Membuster3.

        * css/StylePropertySet.cpp:
        (WebCore::immutableStylePropertySetSize):
        * dom/ElementAttributeData.cpp:
        (WebCore::immutableElementAttributeDataSize):

2012-10-23  Nate Chapin  <japhet@chromium.org>

        Crash in WebCore::SubresourceLoader::willSendRequest.
        https://bugs.webkit.org/show_bug.cgi?id=100147

        Reviewed by Abhishek Arya.

        No new tests. There is a test case that should cover this, but it doesn't
        work correctly on many platforms due to its use of testRunner.addURLToRedirect().
        See http/tests/loading/cross-origin-XHR-willLoadRequest.html.
        Tested manually on http://www.nick.co.uk/shows/spongebob

        * loader/SubresourceLoader.cpp:
        (WebCore::SubresourceLoader::willSendRequest):

2012-10-23  Martin Robinson  <mrobinson@igalia.com>

        [GTK][Soup] Implement the default authentication dialog via WebCoreSupport
        https://bugs.webkit.org/show_bug.cgi?id=99351

        Reviewed by Carlos Garcia Campos.

        Instead of connecting to the SoupSession::authenticate signal in the API layer
        via a SoupSessionFeature, route the message through the typical WebCore authentication
        mechanism. This is a step on the path to having full-blown support for authentication
        in the API allowing for custom authentication dialogs and behavior in clients.

        No new tests. This should not change behavior.

        * GNUmakefile.list.am: Add the new implementation file for the AuthenticationChallenge.
        * loader/ResourceLoader.cpp:
        (WebCore::ResourceLoader::didReceiveAuthenticationChallenge): Include GTK+ here as well.
        * platform/gtk/GtkAuthenticationDialog.cpp: Instead of carrying a raw pointer to a SoupAuth
        carry a GRefPtr which will allow the dialog to be used even after the AuthenticationChallenge
        is destroyed.
        * platform/gtk/GtkAuthenticationDialog.h: ditto.
        * platform/network/ResourceHandle.h: Add a didReceiveAuthenticationChallenge method.
        * platform/network/ResourceHandleInternal.h:
        (ResourceHandleInternal): Remove the getter for the initiating page ID.
        * platform/network/soup/AuthenticationChallenge.h: Added an implementation of AuthenticationChallenge
        that takes a bit of data from the handle including SoupMessage, SoupAuth, and SoupSession.
        * platform/network/soup/AuthenticationChallengeSoup.cpp: Added. Ditto.
        * platform/network/soup/ResourceError.h:
        (ResourceError): Added a factory for authentication errors.
        * platform/network/soup/ResourceErrorSoup.cpp: Ditto.
        * platform/network/soup/ResourceHandleSoup.cpp:
        (WebCore::setSoupRequestInitiaingPageID): Just get the page directly from the NetworkingContext.
        (WebCore::createSoupMessageForHandleAndRequest): Always set the "handle" data on the message.
        It's always used now.
        (WebCore::ResourceHandle::start): Get the initiating page ID directly from the NetworkingContext.
        (WebCore::ResourceHandle::didReceiveAuthenticationChallenge): Added this method which routes
        the authentication challenge to the client.
        (WebCore::ResourceHandle::receivedRequestToContinueWithoutCredential): Added a stub for this method.
        (WebCore::ResourceHandle::loadResourceSynchronously): Don't pass the session to the loader.
        (WebCore::authenicateCallback): Added.
        (WebCore::ResourceHandle::defaultSession): Connect the authenticate callback.

2012-10-23  Adam Barth  <abarth@webkit.org>

        [V8] Simplify GCEpilogueVisitor along the same lines as GCPrologueVisitor
        https://bugs.webkit.org/show_bug.cgi?id=100157

        Reviewed by Eric Seidel.

        This patch removes the traits template parameter from GCEpilogueVisitor
        and makes it mirror GCPrologueVisitor again.

        * bindings/v8/V8GCController.cpp:
        (WebCore):
        (ActiveDOMObjectEpilogueVisitor):
        (WebCore::ActiveDOMObjectEpilogueVisitor::ActiveDOMObjectEpilogueVisitor):
        (WebCore::ActiveDOMObjectEpilogueVisitor::visitDOMWrapper):
        (WebCore::V8GCController::majorGCEpilogue):

2012-10-23  Varun Jain  <varunjain@chromium.org>

        Context menu generated from touch gestures on textareas has
        context of the cursor position instead of the position where the event occurs.
        https://bugs.webkit.org/show_bug.cgi?id=99520

        Reviewed by Kenneth Rohde Christiansen.

        Send a synthetic mouse down event for context menu-summoning-gesture events so
        that textareas can correctly set cursors before receiving the context menu event.

        Test: fast/events/touch/gesture/right-click-gestures-set-cursor-at-correct-position.html

        * page/EventHandler.cpp:
        (WebCore::EventHandler::handleGestureEvent):
        (WebCore::EventHandler::handleGestureTwoFingerTap):
        (WebCore):
        (WebCore::EventHandler::sendContextMenuEventForGesture):
        * page/EventHandler.h:
        (EventHandler):

2012-10-23  Andy Estes  <aestes@apple.com>

        [WebKit2 API] Add properties to get textRects from a WKDOMRange or WKDOMNode
        https://bugs.webkit.org/show_bug.cgi?id=100162

        Reviewed by Sam Weinig.

        * WebCore.exp.in: Exported symbols needed by WebKit2.
        * bindings/objc/DOM.mm:
        (-[DOMNode textRects]): Moved some logic into Node::textRects()
        (everything but the call to updateLayoutIgnorePendingStylesheets()).
        * dom/Node.cpp:
        (WebCore::Node::textRects):
        * dom/Node.h:

2012-10-23  Adam Barth  <abarth@webkit.org>

        [V8] Remove unused function from DOM wrapper visitor
        https://bugs.webkit.org/show_bug.cgi?id=100163

        Reviewed by Eric Seidel.

        No one overrides these virtual functions. They were added in
        http://trac.webkit.org/changeset/73491, but we don't use them in the GC
        controller anymore.

        * bindings/v8/V8DOMMap.h:
        (WebCore::WeakReferenceMap::visit):

2012-10-23  Aaron Colwell  <acolwell@chromium.org>

        Clear m_mediaSource reference when the MediaSource is closed
        https://bugs.webkit.org/show_bug.cgi?id=100047

        Reviewed by Eric Carlson.

        The m_mediaSource object shouldn't be accessed after the object
        transitions to closed so this change simply clears the reference
        when that transition happens.

        No new tests because the change isn't visible to JavaScript.

        * html/HTMLMediaElement.cpp:
        (WebCore::HTMLMediaElement::loadResource): Removed a closed transition that is actually dead code.
        (WebCore::HTMLMediaElement::setSourceState): Clear the m_mediaSource reference on closed transition.

2012-10-23  Simon Fraser  <simon.fraser@apple.com>

        Allow tests to run with a tile cache, and optionally dump the tile cache via layerTreeAsText()
        https://bugs.webkit.org/show_bug.cgi?id=100159

        Reviewed by Tim Horton.

        Add a LAYER_TREE_INCLUDES_PAGE_TILE_CACHE flag for layerTreeAsText(), and plumb it
        through to GraphicsLayerCA, which dumps the TiledBacking tile coverage rect.

        Test: platform/mac/tiled-drawing/use-tiled-drawing.html

        * page/Frame.h:
        * platform/graphics/GraphicsLayer.cpp:
        (showGraphicsLayerTree):
        * platform/graphics/GraphicsLayer.h:
        (WebCore::GraphicsLayer::tiledBacking):
        * platform/graphics/TiledBacking.h:
        * platform/graphics/ca/GraphicsLayerCA.cpp:
        (WebCore::GraphicsLayerCA::tiledBacking):
        (WebCore::GraphicsLayerCA::dumpAdditionalProperties):
        * platform/graphics/ca/GraphicsLayerCA.h:
        (GraphicsLayerCA):
        * platform/graphics/ca/mac/TileCache.h:
        (WebCore::TileCache::tileCoverageRect):
        (TileCache):
        * platform/graphics/ca/mac/TileCache.mm:
        (WebCore::TileCache::computeTileCoverageRect):
        (WebCore::TileCache::revalidateTiles):
        * rendering/RenderLayerCompositor.cpp:
        (WebCore::RenderLayerCompositor::layerTreeAsText):
        * testing/Internals.cpp:
        (WebCore::Internals::layerTreeAsText):
        * testing/Internals.h:
        * testing/Internals.idl:

2012-10-23  No'am Rosenthal  <noam.rosenthal@nokia.com>

        Coordinated Graphics: Enable threaded/IPC animations
        https://bugs.webkit.org/show_bug.cgi?id=93146

        Reviewed by Kenneth Rohde Christiansen.

        Add enablers to TextureMapper and GraphicsLayerAnimation so that they could be used
        across processes with IPC. Added some public accessors to GraphicsLayerAnimation,
        and allowed a GraphicsLayerTextureMapper to receive a full list of animations.

        Covered by existing animation and compositing tests.

        * platform/graphics/GraphicsLayerAnimation.cpp:
        (WebCore::GraphicsLayerAnimation::GraphicsLayerAnimation):
            Initialize the animation with the startTime instead of the offset.

        (WebCore::GraphicsLayerAnimations::getActiveAnimations):
        * platform/graphics/GraphicsLayerAnimation.h:
        (WebCore::GraphicsLayerAnimation::setState):
            Allow setting the pause time as well.

        (WebCore::GraphicsLayerAnimation::boxSize):
        (WebCore::GraphicsLayerAnimation::startTime):
        (WebCore::GraphicsLayerAnimation::pauseTime):
        (WebCore::GraphicsLayerAnimation::animation):
        (WebCore::GraphicsLayerAnimation::keyframes):
        (WebCore::GraphicsLayerAnimation::listsMatch):
        (WebCore::GraphicsLayerAnimations::size):
        (WebCore::GraphicsLayerAnimations::animations):
            Add public accessors to GraphicsLayerAnimation properties.

        (GraphicsLayerAnimation):
        (GraphicsLayerAnimations):
        * platform/graphics/texmap/GraphicsLayerTextureMapper.cpp:
        (WebCore::GraphicsLayerTextureMapper::addAnimation):
        (WebCore::GraphicsLayerTextureMapper::setAnimations):
            Allow replacing the entire list of animations.

        (WebCore):
        * platform/graphics/texmap/GraphicsLayerTextureMapper.h:
        (GraphicsLayerTextureMapper):
        * platform/graphics/texmap/TextureMapperLayer.cpp:
        (WebCore::TextureMapperLayer::applyAnimationsRecursively):
        (WebCore):
        * platform/graphics/texmap/TextureMapperLayer.h:
        (TextureMapperLayer):

2012-10-23  Chris Rogers  <crogers@google.com>

        Change setTargetValueAtTime() to setTargetAtTime()
        https://bugs.webkit.org/show_bug.cgi?id=100153

        Reviewed by Adam Barth.

        The AudioParam method name has changed due to Web Audio API spec review.
        Keep legacy support for the old name.

        Test: webaudio/audioparam-setTargetAtTime.html

        * Modules/webaudio/AudioParam.h:
        (WebCore::AudioParam::setTargetAtTime):
        * Modules/webaudio/AudioParam.idl:
        * Modules/webaudio/AudioParamTimeline.cpp:
        (WebCore::AudioParamTimeline::setTargetAtTime):
        (WebCore::AudioParamTimeline::valuesForTimeRangeImpl):
        * Modules/webaudio/AudioParamTimeline.h:
        (AudioParamTimeline):

2012-10-23  Benjamin Poulain  <benjamin@webkit.org>

        [Mac] Remove extraneous conversion to String->NSString
        https://bugs.webkit.org/show_bug.cgi?id=100044

        Reviewed by Darin Adler.

        Improve some unfortunate use of String->NSString.

        * accessibility/mac/AXObjectCacheMac.mm:
        (WebCore::AXObjectCache::postPlatformNotification): The variable macNotification is ultimately
        needed as NSString, and all its value are or can be NSString.
        Convert the last 3 char* values to NSString literal and use NSString* all the way.

        * loader/mac/LoaderNSURLExtras.h:
        * loader/mac/LoaderNSURLExtras.mm:
        (suggestedFilenameWithMIMEType):
        We converted the String to NSString to pass to suggestedFilenameWithMIMEType. There is no need
        for the string to be a NSSTring there, we can leave it in its original type.

        The case checking for a nil MIMEType was dead code because the implicit conversion never returns nil.

        * platform/mac/DragImageMac.mm:
        (WebCore::createDragImageForLink):
        Use String's nsStringNilIfEmpty().

        * platform/mac/HTMLConverter.mm:
        (-[WebHTMLConverter _addAttachmentForElement:URL:needsParagraph:usePlaceholder:]):
        Fix the coding style. Do the conversion to NSString only when needed.

2012-10-23  Roger Fong  <roger_fong@apple.com>

        [Win] Popup menus positioning needs to take multiple monitors into account.
        https://bugs.webkit.org/show_bug.cgi?id=100158

        Reviewed by Timothy Horton.

        Add screen.x() to repositioning check to account for position of current monitor.

        * platform/win/PopupMenuWin.cpp:
        (WebCore::PopupMenuWin::calculatePositionAndSize):

2012-10-23  Huang Dongsung  <luxtella@company100.net>

        [CSS Shaders] Set FilterOperations on GraphicsLayer after the program of CSS Shaders is loaded.
        https://bugs.webkit.org/show_bug.cgi?id=99908

        Reviewed by Dean Jackson.

        CSS Shaders can not render anything until the program is loaded. If there is
        partial loaded shaders program, whole FilterOperations chain can not render
        anything. It occurs a flash. So We have to wait until the program is loaded, to
        prevent a flash as RenderLayerBacking::updateImageContents() waits until an
        image is fully loaded.

        No new tests, CSS Shaders on Accelerated Compositing are not activated yet.

        * rendering/RenderLayerBacking.cpp:
        (WebCore::RenderLayerBacking::updateFilters):

2012-10-23  Elliott Sprehn  <esprehn@chromium.org>

        Generated should not be supported for things with a shadow
        https://bugs.webkit.org/show_bug.cgi?id=98836

        Reviewed by Dimitri Glazkov.

        As far as CSS is concerned inputs and things with shadow content inside
        shouldn't support pseudo elements like :before, :after or :first-letter.
        Neither Gecko or Presto supports it, and we only accidentally supported
        it.

        Until the spec tells us what to do we should disable support. This is
        also neccesary because the new generated content implementation doesn't
        support shadows.

        Test: fast/forms/pseudo-elements.html

        * rendering/RenderBlock.cpp:
        (WebCore::RenderBlock::updateFirstLetter):
        * rendering/RenderListBox.h: Added missing canHaveGeneratedChildren() that returns false.
        * rendering/RenderObjectChildList.cpp:
        (WebCore::RenderObjectChildList::updateBeforeAfterContent):

2012-10-23  Huang Dongsung  <luxtella@company100.net>

        [CSS Shaders] Implement overlay, color-dodge, color-burn, hard-light, soft-light blend modes.
        https://bugs.webkit.org/show_bug.cgi?id=98504

        Reviewed by Dean Jackson.

        Add expressions for the aforementioned blend modes. The expressions are lifted
        directly from the CSS Compositing and Blending spec [1]. WebKit adds these
        blending expressions to the author's shader.

        [1]: https://dvcs.w3.org/hg/FXTF/rawfile/tip/compositing/index.html#blendingnormal

        Test: css3/filters/custom/custom-filter-blend-modes.html

        * platform/graphics/filters/CustomFilterValidatedProgram.cpp:
        (WebCore::CustomFilterValidatedProgram::rewriteMixFragmentShader):
        (WebCore::CustomFilterValidatedProgram::blendFunctionString):

2012-10-23  Huang Dongsung  <luxtella@company100.net>

        [CSS Shaders] Implement all composite operators except destination and lighter.
        https://bugs.webkit.org/show_bug.cgi?id=97859

        Reviewed by Dean Jackson.

        Add expressions for all composite operators except destination and
        lighter. The expressions are lifted directly from the CSS Compositing
        and Blending spec [1]. WebKit adds these compositing expressions to the
        author's shader.

        [1]: https://dvcs.w3.org/hg/FXTF/rawfile/tip/compositing/index.html#advancedcompositing

        Test: css3/filters/custom/custom-filter-composite-operators.html

        * platform/graphics/filters/CustomFilterCompiledProgram.cpp:
        (WebCore::CustomFilterCompiledProgram::initializeParameterLocations):
            Remove ASSERTION checking if glGetUniformLocation returns negative,
            because ASSERTION can fail with clear and copy composite operations.
            Clear and copy composite operations do not need "css_u_texture"
            representing the DOM element texture. If the driver compiler is
            smart, "css_u_texture" is not regarded as an active uniform, so
            glGetuniformLocation returns -1. glGetAttribLocation ditto.
        * platform/graphics/filters/CustomFilterValidatedProgram.cpp:
        (WebCore::CustomFilterValidatedProgram::compiledProgram):
            Move the above ASSERTION in CustomFilterCompiledProgram to here.
            ASSERTION checks if m_samplerLocation is not -1 when the author
            shader needs an input texture.
        (WebCore::CustomFilterValidatedProgram::needsInputTexture):
        (WebCore):
        (WebCore::CustomFilterValidatedProgram::compositeFunctionString):
        * platform/graphics/filters/CustomFilterValidatedProgram.h:
        (CustomFilterValidatedProgram):

2012-10-23  Max Vujovic  <mvujovic@adobe.com>

        [CSS Shaders] Changing the blend mode in CSS doesn't update the custom filter rendering
        https://bugs.webkit.org/show_bug.cgi?id=99887

        Reviewed by Dirk Schulze.

        Before this patch, WebKit would not recompute an element's style when just its custom filter
        blend mode changed.

        For example, suppose an element initially has the style:
        -webkit-filter: custom(none mix(url(shader.fs) multiply source-atop));

        Then, we change the blend mode from "multiply" to "normal":
        -webkit-filter: custom(none mix(url(shader.fs) normal source-atop));

        WebKit now detects this change and recomputes the style.

        CustomFilterProgram now has an equals operator that considers the program type and the mix
        settings. The mix settings contain the blend mode.

        Test: css3/filters/custom/custom-filter-change-blend-mode.html

        * platform/graphics/filters/CustomFilterProgram.cpp:
        (WebCore::CustomFilterProgram::operator==):
            Previously, the equals operator was pure virtual in CustomFilterProgram. Now,
            CustomFilterProgram implements it and compares the program type and mix settings, if
            appropriate.
        (WebCore):
        * platform/graphics/filters/CustomFilterProgram.h:
        * platform/graphics/filters/CustomFilterProgramInfo.cpp:
        (WebCore::CustomFilterProgramInfo::hash):
            Refactor a repeated condition into a boolean to improve readability.
        (WebCore::CustomFilterProgramInfo::operator==):
            Refactor repeated conditions to improve readability. Reorder the conditions so the less
            expensive ones occur first.
        * rendering/style/StyleCustomFilterProgram.h:
        (WebCore::StyleCustomFilterProgram::operator==):
            Call the base class's equals operator.

2012-10-23  David Barton  <dbarton@mathscribe.com>

        [MathML] Symbol font uses greek letters for roman ones on linux and Windows
        https://bugs.webkit.org/show_bug.cgi?id=99921

        Reviewed by Eric Seidel.

        For parts of stretched parentheses and brackets, mathml.css currently uses the Symbol font,
        which on linux and Windows has greek letters at roman code points. To fix this, we switch
        to the STIXSizeOneSym font when available. However, this can cause vertical gaps between
        glyph parts, because of hard-coded constants in RenderMathMLOperator.cpp. We eliminate all
        these constants. We also shift glyphs upward by 1px and truncate them to avoid the gaps.

        Tested by existing tests, and one test added to LayoutTests/mathml/presentation/mo.xhtml.

        * css/mathml.css:
        (math, mfenced > *):
            - Set font-family for mn, mi, etc.
        (mo, mfenced):
            - Set font-family for mo and anonymous operators, including extension glyphs for
              stretched operators like parentheses and brackets.
        (math):
        * rendering/mathml/RenderMathMLOperator.cpp:
        (WebCore):
            - Change 0x23d0 to 0x23aa for STIX & Cambria Math fonts.
        (WebCore::RenderMathMLOperator::glyphHeightForCharacter):
        (WebCore::RenderMathMLOperator::updateFromElement):
        (WebCore::RenderMathMLOperator::createStackableStyle):
            - Eliminate unused lineHeight and topRelative parameters, and the hard-coded font size
              of 14.
        (WebCore::RenderMathMLOperator::createGlyph):
            - Eliminate unused lineHeight and topRelative parameters.
        * rendering/mathml/RenderMathMLOperator.h:
        (RenderMathMLOperator):

2012-10-23  Benjamin Poulain  <benjamin@webkit.org>

        WTFURL: Implement KURL::setPort()
        https://bugs.webkit.org/show_bug.cgi?id=99898

        Reviewed by Adam Barth.

        * platform/KURLWTFURL.cpp:
        (WebCore::KURL::setPort):

2012-10-23  Nico Weber  <thakis@chromium.org>

        In the open-source jpeg decoder, read image orientation from the exif data
        https://bugs.webkit.org/show_bug.cgi?id=100144

        Reviewed by Eric Seidel.

        This will be used to implement the shouldRespectImageOrientation
        setting (see bug 19688). Currently this isn't hooked up anywhere, so
        it has no observable effect for now.

        Once it's hooked up, it will be tested by
        fast/images/exif-orientation.html

        * platform/graphics/BitmapImage.h:
        * platform/image-decoders/ImageDecoder.h:
        (WebCore::ImageDecoder::orientation):
        (ImageDecoder):
        * platform/image-decoders/jpeg/JPEGImageDecoder.cpp:
        (WebCore::readUint16):
        (WebCore):
        (WebCore::readUint32):
        (WebCore::checkExifHeader):
        (WebCore::readImageOrientation):
        (WebCore::JPEGImageReader::JPEGImageReader):
        (WebCore::JPEGImageReader::decode):
        * platform/image-decoders/jpeg/JPEGImageDecoder.h:
        (WebCore::JPEGImageDecoder::setOrientation):

2012-10-23  Chris Rogers  <crogers@google.com>

        Fix thread safety issue in AudioParamTimeline
        https://bugs.webkit.org/show_bug.cgi?id=100154

        Reviewed by Kenneth Russell.

        Add appropriate locking in AudioParamTimeline::valueForContextTime()

        * Modules/webaudio/AudioParamTimeline.cpp:
        (WebCore::AudioParamTimeline::valueForContextTime):

2012-10-23  Alec Flett  <alecflett@chromium.org>

        IndexedDB: refactor backend to use IDB*Metadata
        https://bugs.webkit.org/show_bug.cgi?id=100055

        Reviewed by Tony Chang.

        This further encapsulates the static data in
        IDBObjectStoreBackendImpl and IDBIndexBackendImp into their
        respective IDB*Metadata structs, in preparation for
        https://bugs.webkit.org/show_bug.cgi?id=99774.

        No new tests as this is purely a refactor.

        * Modules/indexeddb/IDBBackingStore.h:
        (IDBBackingStore):
        * Modules/indexeddb/IDBDatabaseBackendImpl.cpp:
        (WebCore::IDBDatabaseBackendImpl::createObjectStore):
        (WebCore::IDBDatabaseBackendImpl::loadObjectStores):
        * Modules/indexeddb/IDBIndexBackendImpl.cpp:
        (WebCore::IDBIndexBackendImpl::IDBIndexBackendImpl):
        (WebCore::IDBIndexBackendImpl::metadata):
        * Modules/indexeddb/IDBIndexBackendImpl.h:
        (WebCore::IDBIndexBackendImpl::create):
        (WebCore::IDBIndexBackendImpl::id):
        (WebCore::IDBIndexBackendImpl::setId):
        (WebCore::IDBIndexBackendImpl::hasValidId):
        (WebCore::IDBIndexBackendImpl::name):
        (WebCore::IDBIndexBackendImpl::keyPath):
        (WebCore::IDBIndexBackendImpl::unique):
        (WebCore::IDBIndexBackendImpl::multiEntry):
        (IDBIndexBackendImpl):
        * Modules/indexeddb/IDBLevelDBBackingStore.cpp:
        (WebCore::IDBLevelDBBackingStore::getObjectStores):
        (WebCore::IDBLevelDBBackingStore::getIndexes):
        * Modules/indexeddb/IDBLevelDBBackingStore.h:
        (IDBLevelDBBackingStore):
        * Modules/indexeddb/IDBMetadata.h:
        (WebCore::IDBObjectStoreMetadata::IDBObjectStoreMetadata):
        * Modules/indexeddb/IDBObjectStoreBackendImpl.cpp:
        (WebCore::IDBObjectStoreBackendImpl::IDBObjectStoreBackendImpl):
        (WebCore::IDBObjectStoreBackendImpl::metadata):
        (WebCore::IDBObjectStoreBackendImpl::setIndexKeys):
        (WebCore::IDBObjectStoreBackendImpl::putInternal):
        (WebCore::IDBObjectStoreBackendImpl::createIndex):
        (WebCore::IDBObjectStoreBackendImpl::loadIndexes):
        * Modules/indexeddb/IDBObjectStoreBackendImpl.h:
        (WebCore::IDBObjectStoreBackendImpl::create):
        (WebCore::IDBObjectStoreBackendImpl::id):
        (WebCore::IDBObjectStoreBackendImpl::setId):
        (WebCore::IDBObjectStoreBackendImpl::name):
        (WebCore::IDBObjectStoreBackendImpl::keyPath):
        (WebCore::IDBObjectStoreBackendImpl::autoIncrement):
        (IDBObjectStoreBackendImpl):

2012-10-19  Roger Fong  <roger_fong@apple.com>

        [WebGL] conformance/textures/texture-size.html is failing on Apple Mountain Lion
        https://bugs.webkit.org/show_bug.cgi?id=94041

        Reviewed by Dean Jackson.

        When binding a texture to GL_TEXTURE_2D when GL_ACTIVE_TEXTURE is 0, we set m_boundTexture0 to the texture unit. 
        However when we delete the texture, we need to be setting m_boundTexture0 to 0.
        Otherwise when we draw to the screen we bind m_boundTexture0 in the prepareTexture() method and since the associated texture 
        has already been deleted we end up in an error state.

        Tested using Khronos WebGL conformance suite:
        conformance/textures/texture-size.html

        * platform/graphics/opengl/GraphicsContext3DOpenGLCommon.cpp:
        (WebCore::GraphicsContext3D::deleteTexture):

2012-10-22  Dirk Schulze  <krit@webkit.org>

        BasicShapePolygon::path takes width instead of height for boundary calculation
        https://bugs.webkit.org/show_bug.cgi?id=99919

        Reviewed by Darin Adler.

        The 'y' parameters of polygon were calculated by the with of the bounding box of the object.
        This caused problems on percentage values for point positions. 
        Changed it to the height of the bounding box.

        Test: css3/masking/clip-path-polygon-percentage.html

        * rendering/style/BasicShapes.cpp:
        (WebCore::BasicShapePolygon::path):

2012-10-23  Dominik Röttsches  <dominik.rottsches@intel.com>

        Add timeout support to XMLHttpRequest
        https://bugs.webkit.org/show_bug.cgi?id=74802

        Reviewed by Nate Chapin.

        An implementation of XHR2 timeouts by using ResourceRequest's setTimeoutInterval.
        This made several changes necessary in CachedResource and SubresourceLoader in order
        to distinguish and forward the timeout case from there.

        The case of late updates to the timeout property, changing the timeout value after send()
        is not supported yet and handled separately in bug 98156.

        XHR2 timeout tests were initially written by Mozilla's Alex Vincent's. He granted
        permission to reuse them under PD/BSD license
        in https://bugzilla.mozilla.org/show_bug.cgi?id=525816#c86 - big thanks!
        I adapted them for W3C testharness.js and split them into groups with shorter test running time
        so that they can be used as WebKit layout tests.

        Tests: http/tests/xmlhttprequest/timeout/xmlhttprequest-timeout-aborted.html
               http/tests/xmlhttprequest/timeout/xmlhttprequest-timeout-abortedonmain.html
               http/tests/xmlhttprequest/timeout/xmlhttprequest-timeout-overridesexpires.html
               http/tests/xmlhttprequest/timeout/xmlhttprequest-timeout-simple.html
               http/tests/xmlhttprequest/timeout/xmlhttprequest-timeout-synconmain.html
               http/tests/xmlhttprequest/timeout/xmlhttprequest-timeout-twice.html
               http/tests/xmlhttprequest/timeout/xmlhttprequest-timeout-worker-aborted.html
               http/tests/xmlhttprequest/timeout/xmlhttprequest-timeout-worker-overridesexpires.html
               http/tests/xmlhttprequest/timeout/xmlhttprequest-timeout-worker-simple.html
               http/tests/xmlhttprequest/timeout/xmlhttprequest-timeout-worker-synconworker.html
               http/tests/xmlhttprequest/timeout/xmlhttprequest-timeout-worker-twice.html

        * loader/DocumentThreadableLoader.cpp:
        (WebCore::DocumentThreadableLoader::notifyFinished): Forward information about timeout case.
        * loader/SubresourceLoader.cpp:
        (WebCore::SubresourceLoader::didFail): Distinguish timeout case when informing client.
        * loader/cache/CachedResource.h: Distinguishing timeout case for errors.
        (WebCore::CachedResource::errorOccurred):
        (WebCore::CachedResource::loadFailedOrCanceled):
        (WebCore::CachedResource::timedOut):
        * xml/XMLHttpRequest.cpp:
        (WebCore::XMLHttpRequest::XMLHttpRequest): Initializing m_timeout value to zero.
        (WebCore::XMLHttpRequest::setTimeout): Setter function, possibly raising JS exception.
        (WebCore):
        (WebCore::XMLHttpRequest::open): Open call may raise exception for synchronous requests when timeout value is set.
        (WebCore::XMLHttpRequest::createRequest): Assigning timeout value to ResourceRequest.
        (WebCore::XMLHttpRequest::didFail): Handling timeout case separately.
        (WebCore::XMLHttpRequest::didTimeout): Timeout case state transisition and event firing as spec'ed.
        * xml/XMLHttpRequest.h: New event listener, member and callback for handling timeout.
        (WebCore::XMLHttpRequest::timeout):
        (XMLHttpRequest):
        * xml/XMLHttpRequest.idl: New event listener and property added.
        * xml/XMLHttpRequestException.cpp:
        * xml/XMLHttpRequestException.h: Added an exception value for the timeout case.


2012-10-23  Christophe Dumez  <christophe.dumez@intel.com>

        Possible assertion hit in WebCore::HTMLSelectElement::updateListBoxSelection()
        https://bugs.webkit.org/show_bug.cgi?id=99967

        Reviewed by Tony Chang.

        Fix assertion hit in WebCore::HTMLSelectElement::updateListBoxSelection() when
        pressing the left mouse button outside a multiselect and then moving the cursor
        over the multiselect element while holding the button down.

        The issue is that the HTMLSelectElement handler for the mouse move event does
        not check if there is a selection before trying to extend the selection.

        Test: fast/dom/HTMLSelectElement/select-selectedIndex-noAnchorIndex-crash.html

        * html/HTMLSelectElement.cpp:
        (WebCore::HTMLSelectElement::listBoxDefaultEventHandler):

2012-10-23  Adam Barth  <abarth@webkit.org>

        [V8] ScriptWrappable should hold the wrapper handle directly (Dromaeo/dom-modify and dom-traverse get ~2.5% faster)
        https://bugs.webkit.org/show_bug.cgi?id=97974

        Reviewed by Eric Seidel.

        Previously, we stored a pointer to a handle to a wrapper in Node. That
        is an extra layer of indirection that slows down finding the wrapper
        for the node. A handle is just a pointer, so we might as we just store
        the handle in the Node directly. That speeds up dom-modify and
        dom-traverse by about 2.5%.

        This change also lets us get rid of the ChunkedTable we were using to
        store all the wrappers because they're now stored in the Nodes
        directly.

        * bindings/scripts/CodeGeneratorV8.pm:
        (GenerateHeader):
        * bindings/v8/IntrusiveDOMWrapperMap.h:
        (WebCore::IntrusiveDOMWrapperMap::IntrusiveDOMWrapperMap):
        (WebCore::IntrusiveDOMWrapperMap::get):
        (WebCore::IntrusiveDOMWrapperMap::set):
        (WebCore::IntrusiveDOMWrapperMap::contains):
        (WebCore::IntrusiveDOMWrapperMap::visit):
        (WebCore::IntrusiveDOMWrapperMap::removeIfPresent):
        (WebCore::IntrusiveDOMWrapperMap::clear):
        * bindings/v8/ScriptWrappable.h:
        (WebCore::ScriptWrappable::ScriptWrappable):
        (WebCore::ScriptWrappable::wrapper):
        (WebCore::ScriptWrappable::setWrapper):
        (WebCore::ScriptWrappable::disposeWrapper):
        (WebCore::ScriptWrappable::reportMemoryUsage):
        (ScriptWrappable):
        * bindings/v8/V8DOMWrapper.h:
        (WebCore::V8DOMWrapper::getCachedWrapper):

2012-10-23  Kentaro Hara  <haraken@chromium.org>

        [V8] Replace SetGlobalGCPrologueCallback() with AddGCPrologueCallback()
        https://bugs.webkit.org/show_bug.cgi?id=100140

        Reviewed by Adam Barth.

        SetGlobalGCPrologueCallback() and SetGlobalGCEpilogueCallback()
        are deprecated (See http://code.google.com/codesearch#OAMlx_jo-ck/src/v8/include/v8.h&exact_package=chromium&q=v8.h&type=cs&l=3149)
        Instead we should use AddGCPrologueCallback()
        and AddGCEpilogueCallback().

        No tests. No change in behavior.

        * bindings/v8/V8DOMWindowShell.cpp:
        (WebCore::initializeV8IfNeeded):
        * bindings/v8/V8GCController.cpp:
        (WebCore::V8GCController::gcPrologue):
        (WebCore):
        (WebCore::V8GCController::minorGCPrologue):
        (WebCore::V8GCController::majorGCPrologue):
        (WebCore::V8GCController::gcEpilogue):
        (WebCore::V8GCController::minorGCEpilogue):
        (WebCore::V8GCController::majorGCEpilogue):
        * bindings/v8/V8GCController.h:
        (V8GCController):
        * bindings/v8/WorkerContextExecutionProxy.cpp:
        (WebCore::WorkerContextExecutionProxy::initIsolate):

2012-10-23  Adam Barth  <abarth@webkit.org>

        [V8] Enumerate Nodes via the V8 heap rather than via a list in WebCore
        https://bugs.webkit.org/show_bug.cgi?id=100033

        Reviewed by Eric Seidel.

        This patch changes how we enumerate nodes during garbage collection.
        After this patch, we use V8's list of open handles to enumerate node
        wrappers rather than using a separate list that we maintain in WebCore
        for this purpose. A future patch will remove the list in WebCore for a
        DOM performance gain.

        * bindings/js/ScriptProfiler.h:
        (WebCore):
        (WebCore::ScriptProfiler::visitNodeWrappers):
        * bindings/scripts/CodeGeneratorV8.pm:
        (GenerateToV8Converters):
        * bindings/scripts/test/V8/V8TestNode.cpp:
        (WebCore::V8TestNode::wrapSlow):
        * bindings/v8/IntrusiveDOMWrapperMap.h:
        (WebCore::IntrusiveDOMWrapperMap::set):
        * bindings/v8/ScriptProfiler.cpp:
        (WebCore::ScriptProfiler::visitNodeWrappers):
        * bindings/v8/ScriptProfiler.h:
        (WebCore):
        (ScriptProfiler):
        * bindings/v8/V8DOMMap.cpp:
        (WebCore::NodeWrapperVisitor::~NodeWrapperVisitor):
        (WebCore):
        (WebCore::visitAllDOMNodes):
        * bindings/v8/V8DOMMap.h:
        (WebCore):
        (NodeWrapperVisitor):
        * bindings/v8/V8DOMWrapper.cpp:
        (WebCore::V8DOMWrapper::setJSWrapperForDOMNode):
        (WebCore::V8DOMWrapper::setJSWrapperForActiveDOMNode):
        * bindings/v8/V8GCController.cpp:
        (WebCore::NodeVisitor::visitNodeWrapper):
        (WebCore::V8GCController::gcPrologue):
        (WebCore::V8GCController::gcEpilogue):
        * inspector/BindingVisitors.h:
        (WebCore::WrappedNodeVisitor::~WrappedNodeVisitor):
        * inspector/InspectorMemoryAgent.cpp:
        (WebCore):

2012-10-22  Andrey Kosyakov  <caseq@chromium.org>

        Web Inspector: paint rectangles are incorrectly shown in case subframes are present
        https://bugs.webkit.org/show_bug.cgi?id=99849

        Reviewed by Pavel Feldman.

        Move GraphicsContext and paint rectangle from willPaint() to didPaint(), so we don't have
        to store them as a state of InspectorPageAgent

        * inspector/InspectorInstrumentation.cpp:
        (WebCore):
        (WebCore::InspectorInstrumentation::willPaintImpl):
        (WebCore::InspectorInstrumentation::didPaintImpl):
        * inspector/InspectorInstrumentation.h:
        (InspectorInstrumentation):
        (WebCore::InspectorInstrumentation::willPaint):
        (WebCore::InspectorInstrumentation::didPaint):
        * inspector/InspectorPageAgent.cpp:
        (WebCore::InspectorPageAgent::didPaint):
        * inspector/InspectorPageAgent.h:
        * inspector/InspectorTimelineAgent.cpp:
        (WebCore::InspectorTimelineAgent::willPaint):
        (WebCore::InspectorTimelineAgent::didPaint):
        * inspector/InspectorTimelineAgent.h:
        (InspectorTimelineAgent):
        * inspector/TimelineRecordFactory.cpp:
        * inspector/TimelineRecordFactory.h:
        (TimelineRecordFactory):
        * page/FrameView.cpp:
        (WebCore::FrameView::paintContents):
        * rendering/RenderLayerBacking.cpp:
        (WebCore::RenderLayerBacking::paintContents):

2012-10-23  Pavel Feldman  <pfeldman@chromium.org>

        Web Inspector: array grouping does not work for arrays with exactly 10000 elements.
        https://bugs.webkit.org/show_bug.cgi?id=100131

        Reviewed by Vsevolod Vlasov.

        Using ceil() - 1 instead of floor() in bucket size calculation.

        * inspector/front-end/ObjectPropertiesSection.js:

2012-10-23  Shinya Kawanaka  <shinyak@chromium.org>

        The order of resolving distribution in tree composition is wrong.
        https://bugs.webkit.org/show_bug.cgi?id=99552

        Reviewed by Dimitri Glazkov.

        According to the current ShadowDOM spec, we have to resolve <content> first, then resolve <shadow>.
        However, the order of resolution is now the mixed tree order of <content> and <shadow>.

        Test: fast/dom/shadow/content-reprojection-order.html

        * html/shadow/ContentDistributor.cpp:
        (WebCore::ContentDistributor::distribute): We should resolve <content> before <shadow>.
        Only the first active shadow insertion point can select the rest of contents.
        * html/shadow/HTMLContentElement.h:
        (HTMLContentElement):
        * html/shadow/HTMLShadowElement.cpp:
        * html/shadow/HTMLShadowElement.h:
        (WebCore::isHTMLShadowElement):
        (WebCore):
        (WebCore::toHTMLShadowElement):
        * html/shadow/InsertionPoint.h:
        (InsertionPoint): We don't need doesSelectFromHostChildren() anymore.

2012-10-23  Vsevolod Vlasov  <vsevik@chromium.org>

        Web Inspector: Move UISourceCode creation out of mappings to workspace.
        https://bugs.webkit.org/show_bug.cgi?id=100092

        Reviewed by Pavel Feldman.

        Moved uiSourceCode constructor calls out of mappings to workspace.

        * inspector/front-end/CompilerScriptMapping.js:
        * inspector/front-end/NetworkUISourceCodeProvider.js:
        (WebInspector.NetworkUISourceCodeProvider.prototype._parsedScriptSource):
        (WebInspector.NetworkUISourceCodeProvider.prototype._addUISourceCode):
        * inspector/front-end/ResourceScriptMapping.js:
        (WebInspector.ResourceScriptMapping.prototype._deleteOriginalUISourceCodeForScripts):
        (WebInspector.ResourceScriptMapping.prototype._deleteTemporaryUISourceCodeForScripts):
        (WebInspector.ResourceScriptMapping.prototype._bindUISourceCodeToScripts):
        (WebInspector.ResourceScriptMapping.prototype._createUISourceCode):
        (WebInspector.ResourceScriptMapping.prototype._getOrCreateTemporaryUISourceCode):
        (WebInspector.ResourceScriptMapping.prototype._getOrCreateOriginalUISourceCode):
        * inspector/front-end/SASSSourceMapping.js:
        (_bindUISourceCode):
        * inspector/front-end/ScriptSnippetModel.js:
        (WebInspector.ScriptSnippetModel.prototype._addScriptSnippet):
        (WebInspector.ScriptSnippetModel.prototype._createUISourceCodeForScript):
        * inspector/front-end/Workspace.js:
        (WebInspector.Project.prototype.addUISourceCode):
        (WebInspector.Project.prototype.removeUISourceCode):
        (WebInspector.Workspace.prototype.addTemporaryUISourceCode):
        (WebInspector.Workspace.prototype.removeTemporaryUISourceCode):

2012-10-23  Emil A Eklund  <eae@chromium.org>

        Remove unnecessary m_layoutDelta[XY]Saturated initialization
        https://bugs.webkit.org/show_bug.cgi?id=100018

        Reviewed by Julien Chaffraix.
        
        Remove unnecessary initialization from LayoutState constructor
        added in r132105.

        No new tests, no change in functionality.

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

2012-10-23  Zeno Albisser  <zeno@webkit.org>

        [Texmap] Fix drawTextureRectangleARB after r131485.
        https://bugs.webkit.org/show_bug.cgi?id=100133

        Consistently rename u_textureSize to u_samplerSize.

        Reviewed by Noam Rosenthal.

        * platform/graphics/texmap/TextureMapperGL.cpp:
        (WebCore::TextureMapperGL::drawTextureRectangleARB):
        * platform/graphics/texmap/TextureMapperShaderManager.cpp:
        (WebCore::getShaderSpec):
        * platform/graphics/texmap/TextureMapperShaderManager.h:
        (TextureMapperShaderProgram):

2012-10-23  Mike West  <mkwst@chromium.org>

        Web Inspector: 'data:' URLs should be properly trimmed for readability.
        https://bugs.webkit.org/show_bug.cgi?id=100083

        Reviewed by Pavel Feldman.

        We recently landed a patch to trim the middle out of long URLs in
        console messages in order to improve readability. That patch didn't
        effect 'data:' URLs, as they didn't match the regex in the linkifier.
        This patch ensures that 'data:' URLs are properly trimmed down to
        size.

        This problem came to light while resolving a different, smaller issue:
        'image/jpg' wasn't whitelisted as an image MIME type. That trivial fix
        is included in this patch.

        Test: http/tests/inspector/network/image-as-text-loading-data-url.html

        * inspector/front-end/NetworkManager.js:
        (WebInspector.NetworkManager):
            Adds 'image/jpg' as a valid image type.
        * inspector/front-end/ResourceUtils.js:
        (WebInspector.linkifyStringAsFragmentWithCustomLinkifier):
            Supports 'data:' URLs in the linkifier's regex.

2012-10-23  Mike West  <mkwst@chromium.org>

        Web Inspector: Floated anchor element sometimes overlaps following content.
        https://bugs.webkit.org/show_bug.cgi?id=100105

        Reviewed by Pavel Feldman.

        This patch ensures that each console message clears the
        potentially-overlapping floated anchor element.

        * inspector/front-end/inspector.css:
        (#console-prompt):
        (.console-message, .console-user-command):
            Clear the float, and flip the border from the bottom of the
            message to the top. Do the same for the prompt.
        (.console-message:first-child):
            Ensure that the first message doesn't have a top border.

2012-10-23  Alexander Pavlov  <apavlov@chromium.org>

        Web Inspector: Crash when adding a keyframes rule in the Styles pane
        https://bugs.webkit.org/show_bug.cgi?id=99826

        Reviewed by Pavel Feldman.

        The client-supplied selector text is first parsed to make sure it results in a valid style rule selector.

        Test: inspector/styles/add-new-rule-invalid-selector.html

        * inspector/InspectorStyleSheet.cpp:
        (WebCore::createCSSParser):
        (WebCore):
        (WebCore::InspectorStyle::setPropertyText):
        (WebCore::checkStyleRuleSelector):
        (WebCore::InspectorStyleSheet::addRule):
        (WebCore::InspectorStyleSheet::ensureSourceData):
        (WebCore::InspectorStyleSheetForInlineStyle::getStyleAttributeRanges):

2012-10-15  Kenneth Rohde Christiansen  <kenneth@webkit.org>

        Add support for resolution media query
        https://bugs.webkit.org/show_bug.cgi?id=99077

        Reviewed by Antti Koivisto.

        Add support for 'resolution' media query feature.

        Background info:
        http://www.w3.org/blog/CSS/2012/06/14/unprefix-webkit-device-pixel-ratio

        Related spec links:
        http://www.w3.org/TR/css3-mediaqueries/#resolution (recommendation)
        http://www.w3.org/TR/css3-values/#resolution (candidate recommentation)

        Add infrastructure to make it testable.

        Test: fast/media/mq-resolution.html

        * css/CSSParser.cpp:
        (WebCore::CSSParser::validUnit):
        (WebCore::CSSParser::createPrimitiveNumericValue):
        (WebCore::CSSParser::parseValidPrimitive):
        (WebCore::CSSParser::detectNumberToken):
        * css/CSSParser.h:
        * css/CSSPrimitiveValue.cpp:
        (WebCore::isValidCSSUnitTypeForDoubleConversion):
        (WebCore::unitCategory):
        (WebCore::CSSPrimitiveValue::canonicalUnitTypeForCategory):
        (WebCore::CSSPrimitiveValue::customCssText):
        (WebCore::CSSPrimitiveValue::cloneForCSSOM):
        * css/CSSPrimitiveValue.h:

            Enable dpi, dpcm and dppx units when RESOLUTION_MEDIA_QUERY
            is enabled.

        * WebCore.exp.in:

            Export the WebCore::Settings setting.

        * css/CSSPrimitiveValue.h:
        (WebCore::CSSPrimitiveValue::isDotsPerInch):
        (WebCore::CSSPrimitiveValue::isDotsPerPixel):
        (WebCore::CSSPrimitiveValue::isDotsPerCentimeter):
        (CSSPrimitiveValue):

            Add function for checking the recently added density types.

        * css/MediaFeatureNames.h:
        (MediaFeatureNames):

            Add support for resolution, min-resolution and max-resolution.

        * css/MediaQueryEvaluator.cpp:
        (WebCore::compareResolution): Add methods for comparing resolutions.
        (WebCore):
        (WebCore::resolutionMediaFeatureEval):
        (WebCore::min_resolutionMediaFeatureEval):
        (WebCore::max_resolutionMediaFeatureEval):

            Implement the resolution method evaluation.

        * css/MediaQueryExp.cpp:
        (WebCore::featureWithValidPositiveDensity):
        (WebCore):
        (WebCore::featureWithoutValue):
        (WebCore::MediaQueryExp::MediaQueryExp):

            Hook up resolution with the right pre-checks.

       * page/Screen.cpp:
        (WebCore::Screen::horizontalDPI):
        (WebCore::Screen::verticalDPI):

            Check whether an override exists, and if so, uses it.
            If not calculate the value given the device scale factor.

        * page/Settings.cpp:
        (WebCore::Settings::setResolutionOverride):
        (WebCore):
        * page/Settings.h:
        (Settings):
        (WebCore::Settings::resolutionOverride):

            Add a resolution override to settings.

        * testing/InternalSettings.cpp:
        (WebCore::InternalSettings::Backup::Backup):
        (WebCore::InternalSettings::Backup::restoreTo):
        (WebCore::InternalSettings::setResolutionOverride):
        (WebCore):
        * testing/InternalSettings.h:
        (Backup):
        (InternalSettings):
        * testing/InternalSettings.idl:

            Add a new setResolutionOverride method to internals.settings.

2012-10-23  Filip Spacek  <fspacek@rim.com>

        [BlackBerry] Improve the use of stencil buffer during compositing
        https://bugs.webkit.org/show_bug.cgi?id=100020

        We always want to scissor so remove the define.
        Only turn stenciling on if needed.

        Reviewed by Rob Buis.

        Reviewed internally by Arvid Nilsson.

        * platform/graphics/blackberry/LayerRenderer.cpp:
        (WebCore::LayerRenderer::setViewport):
        (WebCore::LayerRenderer::compositeLayers):
        (WebCore::LayerRenderer::drawLayersOnSurfaces):
        (WebCore::LayerRenderer::compositeLayersRecursive):
        (WebCore::LayerRenderer::updateScissorIfNeeded):

2012-10-23  'Pavel Feldman'  <pfeldman@chromium.org>

        Not reviewed: kick out non-chromium files from WebCore.gypi.

        * WebCore.gypi:

2012-10-23  Alexander Shalamov  <alexander.shalamov@intel.com>

        [EFL][WK2] ecore_x should be initialised in WebProcess to avoid re-initialization by PlatformScreenEfl utilities and systemBeep() function
        https://bugs.webkit.org/show_bug.cgi?id=100110

        Reviewed by Kenneth Rohde Christiansen.

        Removed initialization of ecore_x, since it is initialized when process starts.

        test: fast/media/*

        * platform/efl/PlatformScreenEfl.cpp:
        (WebCore::screenDepth):
        (WebCore::screenRect):
        * platform/efl/SoundEfl.cpp:
        (WebCore::systemBeep):

2012-10-23  Adam Klein  <adamk@chromium.org>

        Always parse pasted fragments as HTML even on XHTML pages
        https://bugs.webkit.org/show_bug.cgi?id=99880

        Reviewed by Ojan Vafai.

        When pasting HTML into a page, using the XML parser is unlikely
        to work correctly, as the contents of the clipboard are unlikely
        to be properly-formed XHTML. Thus, for the pasting case, it's always
        better to use HTML parsing, which will properly parse either HTML
        (which is what's usually in the clipboard) or XHTML (which is
        sometimes there as well).

        The Mac port previously worked around this problem by falling back to plain text
        when parsing failed, but switching to HTML seems like a clear improvement.

        This also fixes a crash in Chromium (see http://webkit.org/b/99607
        and http://crbug.com/136218); it erroneously assumed that createFragmentFromMarkup()
        would never return null. This patch makes that true.

        * editing/markup.cpp:
        (WebCore::createFragmentFromMarkup): Don't delegate to createContextualFragment:
        we already know our context element is safe (i.e., it's <body>),
        and we want to force HTML (not XML) parsing.

2012-10-23  Allan Sandfeld Jensen  <allan.jensen@digia.com>

        [Qt] REGRESSION (r130851): fast/text/word-space-with-kerning.html fails
        https://bugs.webkit.org/show_bug.cgi?id=98876

        Reviewed by Simon Hausmann.

        Do not add word-spacing for leading space. This matches what simple path font-width does.

        Tested by existing tests.

        * platform/graphics/qt/FontQt.cpp:
        (WebCore::Font::floatWidthForComplexText):

2012-10-23  Alexander Pavlov  <apavlov@chromium.org>

        Web Inspector: Incorrect resolution of relative URLs containing a scheme in query parameters
        https://bugs.webkit.org/show_bug.cgi?id=100084

        Reviewed by Vsevolod Vlasov.

        Use the RFC 3986 grammar for the URL scheme.

        * inspector/front-end/ParsedURL.js:
        (WebInspector.ParsedURL):

2012-10-23  Eugene Klyuchnikov  <eustas.bug@gmail.com>

        Web Inspector: Elaborate source panel sidebar context menus.
        https://bugs.webkit.org/show_bug.cgi?id=99980

        Reviewed by Vsevolod Vlasov.

        - Watches: add "Add watch expression" item to items and empty element
        - Watches: hide "Remove watch expression" from editing prompt context menu
        - Watches: add titles to header buttons
        - XHR Breakpoints: add "Add Breakpoint" item to items and empty element
        - XHR Breakpoints: add "Remove all breakpoints" item to items (when >1)
        - XHR Breakpoints: add title to header button
        - JS Breakpoints: hide "Remove/(De)Activate breakpoints" when only 1 item present
        - JS Breakpoints: add "(De)Activate breakpoints" to empty element context menu

        * English.lproj/localizedStrings.js: Added corresponding strings.
        * inspector/front-end/BreakpointsSidebarPane.js: Adjusted context menu.
        (WebInspector.XHRBreakpointsSidebarPane): Ditto.
        * inspector/front-end/ObjectPropertiesSection.js: Added "isEditing()"
        * inspector/front-end/WatchExpressionsSidebarPane.js: Adjusted context menu.

2012-10-23  Kent Tamura  <tkent@chromium.org>

        Update binding test results for r132194
        https://bugs.webkit.org/show_bug.cgi?id=100097

        * bindings/scripts/test/V8/V8TestObj.cpp:
        (WebCore::V8TestObj::installPerContextProperties):

2012-10-23  Alexander Pavlov  <apavlov@chromium.org>

        Web Inspector: Sass can only resolve same folder paths
        https://bugs.webkit.org/show_bug.cgi?id=99259

        Reviewed by Vsevolod Vlasov.

        The actual reason is that the rule source location linkifier tries to linkify a resource (*.scss), which does not exist,
        and falls back to just stripping the main page URL prefix from the rule location URL. This change introduces LiveLocations
        for CSSRule locations and makes sure they are linkified using uiSourceCode's parsedURL.displayName.

        * inspector/front-end/CSSStyleModel.js:
        (WebInspector.CSSStyleModel): Introduced LiveLocation management for CSSRules.
        (WebInspector.CSSStyleModel.prototype.setSourceMapping):
        (WebInspector.CSSStyleModel.prototype._updateLocations):
        (WebInspector.CSSStyleModel.prototype.createLiveLocation):
        (WebInspector.CSSStyleModel.prototype.updateLocations):
        (WebInspector.CSSStyleModel.LiveLocation): A LiveLocation for the CSS domain.
        (WebInspector.CSSStyleModel.LiveLocation.prototype.uiLocation):
        (WebInspector.CSSStyleModel.LiveLocation.prototype.dispose):
        * inspector/front-end/Linkifier.js:
        (WebInspector.Linkifier.prototype.linkifyCSSRuleLocation): CSSRule LiveLocation-based link builder.
        (WebInspector.Linkifier.DefaultFormatter.prototype.formatLiveAnchor): Add a title for anchors.
        (WebInspector.Linkifier.DefaultCSSFormatter): Formatter for CSS location links.
        (WebInspector.Linkifier.DefaultCSSFormatter.prototype.formatLiveAnchor):
        * inspector/front-end/ResourceUtils.js:
        (WebInspector.displayNameForURL): Use parsedURL.displayName if uiSourceCode is present for the specified URL.
        * inspector/front-end/SASSSourceMapping.js:
        * inspector/front-end/StylesSidebarPane.js:
        (WebInspector.StylesSidebarPane):
        (WebInspector.StylesSidebarPane.prototype._innerRebuildUpdate):
        * inspector/front-end/inspector.html:

2012-10-23  Sheriff Bot  <webkit.review.bot@gmail.com>

        Unreviewed, rolling out r132149.
        http://trac.webkit.org/changeset/132149
        https://bugs.webkit.org/show_bug.cgi?id=100098

        Breaks inspector profiler tests in debug mode. (Requested by
        pfeldman1 on #webkit).

        * rendering/AutoTableLayout.cpp:
        (WebCore::AutoTableLayout::recalcColumn):
        * rendering/FixedTableLayout.cpp:
        (WebCore::FixedTableLayout::calcWidthArray):
        * rendering/RenderTable.cpp:
        (WebCore::RenderTable::layout):
        * rendering/RenderTableCol.cpp:
        (WebCore::RenderTableCol::styleDidChange):
        (WebCore::RenderTableCol::updateFromElement):
        (WebCore::RenderTableCol::computePreferredLogicalWidths):
        * rendering/RenderTableCol.h:
        (RenderTableCol):

2012-10-23  Simon Hausmann  <simon.hausmann@digia.com>

        Unreviewed trivial Qt build fix: Fix build without USE_3D_GRAPHICS

        Move the #if USE(GRAPHICS_SURFACE) up to protect the inclusion of
        GraphicsContext3D.h to be done only if we use the surface.

        * platform/graphics/surfaces/GraphicsSurface.h:

2012-10-23  Sheriff Bot  <webkit.review.bot@gmail.com>

        Unreviewed, rolling out r132033.
        http://trac.webkit.org/changeset/132033
        https://bugs.webkit.org/show_bug.cgi?id=100097

        Broke calendar picker (Requested by tkent on #webkit).

        * bindings/scripts/CodeGeneratorV8.pm:
        (GenerateImplementation):

2012-10-23  Andras Becsi  <andras.becsi@digia.com>

        Remove devicePixelRatio from ViewportAttributes
        https://bugs.webkit.org/show_bug.cgi?id=99845

        Reviewed by Adam Barth.

        Since r121555 the devicePixelRatio is not calculated any more
        and the scale factor is stored in Page::m_deviceScaleFactor,
        thus it can be removed from ViewportAttributes to reduce
        redundancy and unnecessary client code.
        Use a new parameter in viewport calculation functions using
        the visible viewport size (instead of passing the adjusted
        viewport size) so that after this change clients do not end
        up using the unadjusted viewport size for calculations.

        No behavioural change, no new tests needed.

        * WebCore.exp.in:
        * dom/ViewportArguments.cpp:
        (WebCore::computeViewportAttributes):
        (WebCore::computeMinimumScaleFactorForContentContained):
        Add the devicePixelRatio as a parameter.
        (WebCore::restrictMinimumScaleFactorToViewportSize): Ditto.
        * dom/ViewportArguments.h:
        (ViewportAttributes):
        (WebCore):
        * testing/InternalSettings.cpp:
        (WebCore::InternalSettings::configurationForViewport):

2012-10-23  Kent Tamura  <tkent@chromium.org>

        Support full month names in DateTimeEditElement, and use them in input[type=month] by default
        https://bugs.webkit.org/show_bug.cgi?id=100060

        Reviewed by Kentaro Hara.

        According to https://plus.google.com/104770450049736549185/posts/4zsoeHoa7SM
        no one wants to show abbreviated month names for input[type=month].
        This change add support for full month names in DateTimeEditELement, and
        LocaleICU and LocaleMac retun month formats with full month names. Note
        that LocaleWin::monthFormat returns formats for full month names.

        No new tests. Covered by fast/forms/month-multiple-fields/month-multiple-fields-appearance-*.html

        * html/shadow/DateTimeEditElement.cpp:
        (WebCore::DateTimeEditBuilder::visitField):
        If count is 4, use Localizer::monthLabels or standAloneMonthLabels.
        * platform/text/LocaleICU.cpp:
        (WebCore::LocaleICU::monthFormat):
        Returns a format for full month names.
        * platform/text/mac/LocaleMac.mm:
        (WebCore::LocaleMac::monthFormat): Ditto.

2012-10-23  Mike West  <mkwst@chromium.org>

        Viewport errors should be slightly friendlier with regard to ';'.
        https://bugs.webkit.org/show_bug.cgi?id=100003

        Reviewed by Adam Barth.

        This patch scans viewport values that cause errors for ';'. If found,
        a quick message is appended to the error, noting that semicolons are
        not valid separators in viewport contents, and that commas would be
        the proper substitute.

        No functional changes, just a better error message.

        * dom/ViewportArguments.cpp:
        (WebCore::reportViewportWarning):

2012-10-23  Timothy Hatcher  <timothy@apple.com>

        Docking/undocking the Web Inspector does not work correctly in Safari.

        The "docked" and "bottom" strings need quoted instead of being passed as identifiers.

        https://bugs.webkit.org/show_bug.cgi?id=100080

        Reviewed by Yury Semikhatsky.

        * inspector/InspectorFrontendClientLocal.cpp:
        (WebCore::InspectorFrontendClientLocal::setAttachedWindow): Put quotes around the %s.

2012-10-23  Sudarsana Nagineni  <sudarsana.nagineni@intel.com>

        [EFL][WK2] Compilation warning in GraphicsContext3DPrivate.cpp when AC is enabled
        https://bugs.webkit.org/show_bug.cgi?id=99723

        Reviewed by Kenneth Rohde Christiansen.

        Fix compilation warning in GraphicsContext3DPrivate.cpp when AC is
        enabled.

        No new tests, no change in behavior.

        * platform/graphics/efl/GraphicsContext3DPrivate.cpp:
        (WebCore::GraphicsContext3DPrivate::createSurface):

2012-10-23  Dan Carney  <dcarney@google.com>

        When blocking localStorage, Firefox throws a security exception on access, and maybe so should we
        https://bugs.webkit.org/show_bug.cgi?id=63257

        Reviewed by Jochen Eisinger.

        Throw security exception when local storage is accessed
        under certain circumstances to match firefox.

        No new tests. Existing tests modified.

        * bindings/js/JSStorageCustom.cpp:
        (WebCore::JSStorage::canGetItemsForName):
        (WebCore::JSStorage::nameGetter):
        (WebCore::JSStorage::deleteProperty):
        (WebCore::JSStorage::getOwnPropertyNames):
        * bindings/v8/custom/V8StorageCustom.cpp:
        (WebCore):
        (WebCore::setDOMException):
        (WebCore::V8Storage::namedPropertyEnumerator):
        (WebCore::storageGetter):
        (WebCore::V8Storage::namedPropertyQuery):
        (WebCore::storageDeleter):
        * inspector/InspectorDOMStorageAgent.cpp:
        (WebCore::InspectorDOMStorageAgent::getDOMStorageEntries):
        (WebCore::InspectorDOMStorageAgent::removeDOMStorageItem):
        * page/DOMWindow.cpp:
        (WebCore::DOMWindow::sessionStorage):
        (WebCore::DOMWindow::localStorage):
        * storage/Storage.cpp:
        * storage/Storage.h:
        (WebCore):
        (WebCore::Storage::length):
        (WebCore::Storage::key):
        (WebCore::Storage::getItem):
        (WebCore::Storage::setItem):
        (WebCore::Storage::removeItem):
        (WebCore::Storage::clear):
        (WebCore::Storage::contains):
        * storage/Storage.idl:
        * storage/StorageArea.h:
        (StorageArea):
        * storage/StorageAreaImpl.cpp:
        (WebCore::StorageAreaImpl::canAccessStorage): Checks whether access to storage is a security violation.
        (WebCore):
        (WebCore::StorageAreaImpl::disabledByPrivateBrowsingInFrame):
        (WebCore::StorageAreaImpl::length):
        (WebCore::StorageAreaImpl::key):
        (WebCore::StorageAreaImpl::getItem):
        (WebCore::StorageAreaImpl::setItem):
        (WebCore::StorageAreaImpl::removeItem):
        (WebCore::StorageAreaImpl::clear):
        (WebCore::StorageAreaImpl::contains):
        * storage/StorageAreaImpl.h:
        (StorageAreaImpl):

2012-10-22  Joshua Bell  <jsbell@chromium.org>

        IndexedDB: Remove custom binding code for IDBCursor.value
        https://bugs.webkit.org/show_bug.cgi?id=100034

        Reviewed by Kentaro Hara.

        Now that we're using ScriptValue instead of SerializedScriptValue we can just expose
        IDBCursor.value as an |any| (IDL) or |ScriptValue| (C++) to maintain the specified
        semantics that the object identity is retained across accesses.

        Test: storage/indexeddb/cursor-value.html

        * Modules/indexeddb/IDBCursor.cpp: Remove "dirty" tracking.
        (WebCore::IDBCursor::IDBCursor):
        (WebCore::IDBCursor::value):
        (WebCore::IDBCursor::setValueReady):
        * Modules/indexeddb/IDBCursor.h: IDBAny -> ScriptValue
        (IDBCursor):
        * Modules/indexeddb/IDBCursorWithValue.idl: IDBAny -> any
        * Modules/indexeddb/IDBObjectStore.cpp: No need to route through IDBAny to get ScriptValue.
        (WebCore):
        * UseV8.cmake: Remove references to IDBCustomBindings.cpp
        * WebCore.gypi: Ditto.
        * WebCore.vcproj/WebCore.vcproj: Ditto.
        * bindings/v8/IDBCustomBindings.cpp: Removed.

2012-10-22  Dan Bernstein  <mitz@apple.com>

        Font’s fast code path is used for partial runs with kerning and ligatures, but shouldn’t be
        https://bugs.webkit.org/show_bug.cgi?id=100068

        Reviewed by Sam Weinig.

        As described in <http://webkit.org/b/100050>, the fast code path doesn’t handle partial runs
        correctly when kerning or ligatures are enabled. Since the partial-run case is uncommon,
        for now just use the complex code path in this case.

        * platform/graphics/Font.cpp:
        (WebCore::Font::drawText): Changed to use the complex path for partial runs if there are any
        typesetting features.
        (WebCore::Font::drawEmphasisMarks): Ditto.
        (WebCore::Font::selectionRectForText): Ditto.
        (WebCore::Font::offsetForPosition): Changed to use the complex path if there are any
        typesetting features.

2012-10-22  Peter Wang  <peter.wang@torchmobile.com.cn>

        [BlackBerry] Missing some cookies in HTTP response header when set several cookies in one "Set-Cookie" header.
        https://bugs.webkit.org/show_bug.cgi?id=99950

        Reviewed by George Staikos.

        In "NetworkJob::handleNotifyHeaderReceived", if there are several "Set-Cookie" headers, 
        we should combine the following ones with the first.

        No new test case.

        * platform/network/blackberry/NetworkJob.cpp:
        (WebCore::NetworkJob::handleNotifyHeaderReceived):

2012-10-22  MORITA Hajime  <morrita@google.com>

        Assertion failed at WebCore::toInsertionPoint / WebCore::ContentDistributor::distribute
        https://bugs.webkit.org/show_bug.cgi?id=100038

        Reviewed by Kent Tamura.

        isHTMLContentElement() assumes that the content element always has
        a tag name "content" but it doesn't when Shadow DOM feature is
        disabled. This fix let the function see the correct tag name.

        Test: fast/dom/shadow/insertion-points-with-shadow-disabled.html

        * html/shadow/HTMLContentElement.cpp:
        (WebCore::HTMLContentElement::contentTagName):
        * html/shadow/HTMLContentElement.h:
        (HTMLContentElement):
        (WebCore::isHTMLContentElement):

2012-10-22  Shinya Kawanaka  <shinyak@chromium.org>

        [Shadow] Fallback content should also be reprojection.
        https://bugs.webkit.org/show_bug.cgi?id=99750

        Reviewed by Dimitri Glazkov.

        Fallback content of InsertionPoint should be reprojected. The existing assumption that
        only the direct child of host element can be distributed to InsertionPoint does not hold anymore.
        So, if the parent of an element is InsertionPoint which should show fallback element, we have to
        check the grand parent of the element instead of the element.

        Tests: fast/dom/shadow/content-reprojection-fallback-reprojection.html
               fast/dom/shadow/content-reprojection-fallback.html

        * css/StyleResolver.cpp:
        (WebCore::shouldResetStyleInheritance): Checks the grandparent of the element if the parent is
        an InsertionPoint which uses fallback content.
        * dom/ComposedShadowTreeWalker.cpp:
        (WebCore::shadowOfParentForDistribution):
        (WebCore):
        (WebCore::resolveReprojection):
        (WebCore::ComposedShadowTreeWalker::traverseParent):
        * html/shadow/InsertionPoint.cpp:
        (WebCore::InsertionPoint::shouldUseFallbackElements): True if the InsertionPoint should use fallback content.
        (WebCore):
        * html/shadow/InsertionPoint.h:
        (WebCore::parentElementForDistribution): Returns the grandparent element if the parent is InsertionPoint which uses
        fallback content. Returns parent element otherwise.
        (WebCore):

2012-10-22  Keishi Hattori  <keishi@webkit.org>

        Label position is wrong in the suggestion picker when all the suggestions have labels
        https://bugs.webkit.org/show_bug.cgi?id=99965

        Reviewed by Kent Tamura.

        Somehow the scrollbar was appearing and so the label element was being wrapped to the next line.

        No new tests. Can't reproduce in layout test.

        * Resources/pagepopups/suggestionPicker.css:
        (.suggestion-list):
        * Resources/pagepopups/suggestionPicker.js:
        (SuggestionPicker.prototype._fixWindowSize): Explicitly show the scroll bar.

2012-10-22  MORITA Hajime  <morrita@google.com>

        [Chromium] Needs to track ShadowRoot usage
        https://bugs.webkit.org/show_bug.cgi?id=99955

        Reviewed by Dimitri Glazkov.

        Added an UMA instrumentation.

        * dom/ShadowRoot.cpp:
        (WebCore::determineUsageType):
        (WebCore):
        (WebCore::ShadowRoot::create):

2012-10-22  Kent Tamura  <tkent@chromium.org>

        Introduce Localizer::standAloneMonthLabels
        https://bugs.webkit.org/show_bug.cgi?id=99963

        Reviewed by Kentaro Hara.

        We realized full month names and full stand-alone month names were
        necessary for input[type=month] UI. We change the compile-flag for
        Localizer::monthLabels from "ENABLE(CALENDAR_PICKER)" to
        "ENABLE(CALENDAR_PICKER) || ENABLE(INPUT_MULTIPLE_FIELDS_UI)," and
        introduce Localizer::standAloneMonthLabels.

        Tests: Add some test cases to Source/WebKit/chromium/LocaleMacTest.cpp
        and LocalizedDateICUTest.cpp.

        * platform/text/Localizer.h:
        (Localizer):
        - Add pure virtual standAloneMonthLabels.
        - Change the condition for monthLabels.

        * platform/text/LocaleNone.cpp:
        (LocaleNone): Declare monthLabels, standAloneMonthLabels, and m_monthLabels.
        (WebCore::LocaleNone::monthLabels):
        Added. It always returns English month names.
        (WebCore::LocaleNone::standAloneMonthLabels):
        Added. Just calls monthLabels.

        * platform/text/LocaleWin.h:
        (LocaleWin):
        Declare standAloneMonthLabels, and change the condition for monthLabels.
        * platform/text/LocaleWin.cpp:
        (WebCore): Change the condition for monthLabels.
        (WebCore::LocaleWin::standAloneMonthLabels):
        Added. Just calls monthLabels.

        * platform/text/mac/LocaleMac.h:
        (LocaleMac):
        - Add standAloneMonthLabels and m_standAloneMonthLabels
        - Change the condition for monthLabels and m_monthLabels.
        * platform/text/mac/LocaleMac.mm:
        (WebCore): Change the condition for monthLabels.
        (WebCore::LocaleMac::standAloneMonthLabels):
        Added. Get the information with NSDateFormatter::standaloneMonthSymbols.

        * platform/text/LocaleICU.h:
        (LocaleICU):
        - Add standAloneMonthLabels and m_standAloneMonthLabels
        - Change the condition for monthLabels and m_monthLabels.
        * platform/text/LocaleICU.cpp:
        (WebCore::LocaleICU::initializeCalendar):
        Remove m_monthLabels initialization in order to avoid dependecy from monthLabels.
        (WebCore):
        (WebCore::createFallbackMonthLabels): Change the compile condition.
        (WebCore::LocaleICU::monthLabels):
        - Change the compile condition.
        - Don't depend on initializeCalendar to make the code for
        ENABLE(INPUT_MULTIPLE_FIELDS_UI) && !ENABLE(CALENDAR_PICKER) minimal.
        (WebCore::LocaleICU::standAloneMonthLabels):
        Added. The code is similar to shortStandAloneMonthLabels.

2012-10-22  Shinya Kawanaka  <shinyak@chromium.org>

        Refactoring around ContainerNode::attachChildren
        https://bugs.webkit.org/show_bug.cgi?id=99968

        Reviewed by Hajime Morita.

        Since ContainerNode::attach() is now equivalent to ContainerNode::attachChildren() + Node::attach(), we should call
        ContainerNode::attach() instead of calling them.

        No new tests, no change in behavior.

        * dom/ContainerNode.h:
        (ContainerNode):
        * dom/Element.cpp:
        (WebCore::Element::attach):
        * dom/ShadowRoot.cpp:
        (WebCore::ShadowRoot::attach):

2012-10-22  Michael Saboff  <msaboff@apple.com>

        HTML Parser should produce 8 bit strings for doctype, comment and tagName tokens
        https://bugs.webkit.org/show_bug.cgi?id=99889

        Reviewed by Geoffrey Garen.

        Added 8 bit check for accumulating all token data in MarkupTokenBase.  Added code to convert
        "name" token data directly to a string (8 or 16 as appropriate).  Changed to accumulate
        m_bufferedEndTagName as LChar's.

        No new tests, covered by existing tests.

        * html/parser/HTMLToken.h:
        (HTMLToken):
        * html/parser/HTMLTokenizer.cpp:
        (WebCore::HTMLTokenizer::nextToken):
        (WebCore::HTMLTokenizer::addToPossibleEndTag):
        (WebCore::HTMLTokenizer::isAppropriateEndTag):
        * html/parser/HTMLTokenizer.h:
        (HTMLTokenizer):
        * html/parser/HTMLTreeBuilder.cpp:
        (WebCore::HTMLTreeBuilder::ExternalCharacterTokenBuffer::takeLeading):
        * xml/parser/MarkupTokenBase.h:
        (WebCore::MarkupTokenBase::beginStartTag):
        (WebCore::MarkupTokenBase::beginEndTag):
        (MarkupTokenBase):
        (WebCore::MarkupTokenBase::beginDOCTYPE):
        (WebCore::MarkupTokenBase::appendToComment):
        (WebCore::MarkupTokenBase::appendToName):
        (WebCore::MarkupTokenBase::nameString):
        (WebCore::AtomicMarkupTokenBase::AtomicMarkupTokenBase):

2012-10-22  Tony Chang  <tony@chromium.org>

        margin-top/bottom has no effect for child nodes of flex items
        https://bugs.webkit.org/show_bug.cgi?id=99923

        Reviewed by Ojan Vafai.

        Flexitems, like table cells, shouldn't collapse margins.

        Test: css3/flexbox/flexitem-no-margin-collapsing.html

        * rendering/RenderBlock.cpp:
        (WebCore::RenderBlock::MarginInfo::MarginInfo): Check to see if the parent is a flexible box.
        We should always have a parent if we make it this far in the check.

2012-10-22  Marja Hölttä  <marja@chromium.org>

        Refactor CachedResourceLoader: add CachedResourceRequest
        https://bugs.webkit.org/show_bug.cgi?id=99736

        Reviewed by Adam Barth.

        For fixing bugs 84883 and 92761,
        CachedResourceLoader::requestResource should take as parameter
        information about who initiated the request. But the parameter
        list was already long. This gathers all the parameters into a
        separate class, CachedResourceRequest. The next step is to add
        information about who initiated the request into
        CachedResourceRequest.

        No new tests because no changes in functionality, just moving code
        around.

        * CMakeLists.txt:
        * GNUmakefile.list.am:
        * Target.pri:
        * WebCore.gypi:
        * WebCore.vcproj/WebCore.vcproj:
        * WebCore.xcodeproj/project.pbxproj:
        * css/CSSFontFaceSrcValue.cpp:
        (WebCore::CSSFontFaceSrcValue::cachedFont):
        * css/CSSImageSetValue.cpp:
        (WebCore::CSSImageSetValue::cachedImageSet):
        * css/CSSImageValue.cpp:
        (WebCore::CSSImageValue::cachedImage):
        * css/StyleRuleImport.cpp:
        (WebCore::StyleRuleImport::requestStyleSheet):
        * css/WebKitCSSSVGDocumentValue.cpp:
        (WebCore::WebKitCSSSVGDocumentValue::load):
        * css/WebKitCSSShaderValue.cpp:
        (WebCore::WebKitCSSShaderValue::cachedShader):
        * dom/ProcessingInstruction.cpp:
        (WebCore::ProcessingInstruction::checkStyleSheet):
        * dom/ScriptElement.cpp:
        (WebCore::ScriptElement::requestScript):
        * html/HTMLLinkElement.cpp:
        (WebCore::HTMLLinkElement::process):
        * loader/DocumentThreadableLoader.cpp:
        (WebCore::DocumentThreadableLoader::loadRequest):
        * loader/ImageLoader.cpp:
        (WebCore::ImageLoader::updateFromElement):
        * loader/LinkLoader.cpp:
        (WebCore::LinkLoader::loadLink):
        * loader/TextTrackLoader.cpp:
        (WebCore::TextTrackLoader::load):
        * loader/cache/CachedResourceLoader.cpp:
        (WebCore::CachedResourceLoader::requestImage):
        (WebCore::CachedResourceLoader::requestFont):
        (WebCore::CachedResourceLoader::requestTextTrack):
        (WebCore::CachedResourceLoader::requestShader):
        (WebCore::CachedResourceLoader::requestCSSStyleSheet):
        (WebCore::CachedResourceLoader::requestUserCSSStyleSheet):
        (WebCore::CachedResourceLoader::requestScript):
        (WebCore::CachedResourceLoader::requestXSLStyleSheet):
        (WebCore::CachedResourceLoader::requestSVGDocument):
        (WebCore::CachedResourceLoader::requestLinkResource):
        (WebCore::CachedResourceLoader::requestRawResource):
        (WebCore::CachedResourceLoader::requestResource):
        (WebCore::CachedResourceLoader::requestPreload):
        (WebCore::CachedResourceLoader::defaultCachedResourceOptions):
        (WebCore):
        * loader/cache/CachedResourceLoader.h:
        (WebCore):
        (CachedResourceLoader):
        * loader/cache/CachedResourceRequest.cpp: Added.
        (WebCore):
        (WebCore::CachedResourceRequest::CachedResourceRequest):
        * loader/cache/CachedResourceRequest.h: Added.