RenderIFrame should display its name correctly in DRT output.
[WebKit-https.git] / Source / WebCore / ChangeLog
index 1a07633..9d7766b 100644 (file)
+2013-11-09  Andreas Kling  <akling@apple.com>
+
+        RenderIFrame should display its name correctly in DRT output.
+        <https://webkit.org/b/124117>
+
+        Nuke an age-old FIXME about making RenderIFrame show its true name
+        in DRT output. No more "RenderPartObject {IFRAME}"!
+
+        Reviewed by Anders Carlsson.
+
+2013-11-09  Andreas Kling  <akling@apple.com>
+
+        SVGTextMetricsBuilder::walkTree() should take a RenderElement.
+        <https://webkit.org/b/124105>
+
+        Make walkTree() take a RenderElement so we can use the non-virtual
+        firstChild() internally. All call sites had pointers to compatible
+        objects already.
+
+        Reviewed by Anders Carlsson.
+
+2013-11-09  Andreas Kling  <akling@apple.com>
+
+        RenderMathMLFenced should pass around operators in tighter types.
+        <https://webkit.org/b/124115>
+
+        Store operator renderers in RenderMathMLOperator pointers instead
+        of passing them around as RenderObject.
+
+        Reviewed by Martin Robinson.
+
+2013-11-09  Andreas Kling  <akling@apple.com>
+
+        Use RENDER_OBJECT_TYPE_CASTS for more types.
+        <https://webkit.org/b/124112>
+
+        Generate toRenderFoo() type casting helpers for these classes:
+
+            - RenderCombineText
+            - RenderDetailsMarker
+            - RenderListMarker
+            - RenderVideo
+            - RenderView
+
+        Reviewed by Anders Carlsson.
+
+2013-11-09  Andreas Kling  <akling@apple.com>
+
+        Move MathML type checking virtuals to RenderObject.
+        <https://webkit.org/b/124111>
+
+        Previously, checking the type of a MathML renderer would require
+        that you first check if it's a RenderMathMLBlock, and then casting
+        to that type to access the check you really wanted.
+
+        This change moves all the isRenderMathMLFoo() virtual functions
+        to RenderObject. I also made sure all the overloads were private
+        and marked them OVERRIDE/FINAL as appropriate.
+
+        Finally I replaced all the hand-written casting functions with
+        autogenerated ones using RENDER_OBJECT_TYPE_CASTS.
+
+        Reviewed by Anders Carlsson.
+
+2013-11-09  Martin Robinson  <mrobinson@igalia.com>
+
+        [MathML] Poor spacing around delimiters in MathML Torture Test 14
+        https://bugs.webkit.org/show_bug.cgi?id=122837
+
+        Reviewed by Brent Fulgham.
+
+        Instead of stretching the vertical bar with the stretchable version, just repeat
+        the normal vertical bar. This follows what Gecko does when rendering tall vertical
+        bars and also works around an issue with STIX fonts leading to poor spacing in
+        formulas.
+
+        * rendering/mathml/RenderMathMLOperator.cpp: Stretch the vertical bar with the
+        normal variant.
+
+2013-11-09  Anders Carlsson  <andersca@apple.com>
+
+        Encode form data using the KeyedEncoder
+        https://bugs.webkit.org/show_bug.cgi?id=124107
+
+        Reviewed by Sam Weinig.
+
+        * platform/KeyedCoding.h:
+        (WebCore::KeyedEncoder::encodeEnum):
+        * platform/network/FormData.cpp:
+        (WebCore::encodeElement):
+        (WebCore::FormData::encode):
+        * platform/network/FormData.h:
+
+2013-11-09  Sam Weinig  <sam@webkit.org>
+
+        Modernize CanvasObserverProxy
+        https://bugs.webkit.org/show_bug.cgi?id=124106
+
+        Reviewed by Anders Carlsson.
+
+        * css/CSSCanvasValue.h:
+
+2013-11-09  Patrick Gansterer  <paroga@webkit.org>
+
+        Move RunLoop from WebCore to WTF
+        https://bugs.webkit.org/show_bug.cgi?id=116606
+
+        Reviewed by Anders Carlsson.
+
+        * CMakeLists.txt:
+        * GNUmakefile.list.am:
+        * PlatformBlackBerry.cmake:
+        * PlatformEfl.cmake:
+        * PlatformGTK.cmake:
+        * PlatformWin.cmake:
+        * WebCore.exp.in:
+        * WebCore.vcxproj/WebCore.vcxproj:
+        * WebCore.xcodeproj/project.pbxproj:
+        * platform/graphics/texmap/coordinated/CoordinatedGraphicsLayer.h:
+        * platform/graphics/texmap/coordinated/CoordinatedGraphicsScene.h:
+
+2013-11-09  Andreas Kling  <akling@apple.com>
+
+        Tighten typing in SVGResourcesCycleSolver a bit.
+        <https://webkit.org/b/124104>
+
+        Make the SVGResourcesCycleSolver constructor take a RenderElement&
+        and a SVGResources&.
+
+        While I was in the neighborhood, also converted one loop to use a
+        renderer iterator instead of walking siblings manually.
+
+        Finally used "auto" to clean up some overly wordy loops.
+
+        Reviewed by Anders Carlsson.
+
+2013-11-09  Andreas Kling  <akling@apple.com>
+
+        Beat SVGRenderSupport with the RenderElement stick.
+        <https://webkit.org/b/124102>
+
+        Tighten up all the SVGRenderSupport helper functions by making them
+        take const RenderElements references instead of raw RenderObject
+        pointers as much as possible.
+
+        This tunes up a bunch of branchy style() calls.
+
+        The patch looks big but it's mostly mechanical. I just changed the
+        signatures and then worked backwards until everything built again.
+
+        Reviewed by Antti Koivisto.
+
+2013-11-09  Andreas Kling  <akling@apple.com>
+
+        SVGTextLayoutAttributes always has a RenderSVGInlineText.
+        <https://webkit.org/b/124101>
+
+        No SVGTextLayoutAttributes object is without a RenderSVGInlineText
+        "context" so make context() return a reference.
+
+        Reviewed by Antti Koivisto.
+
+2013-11-09  Andreas Kling  <akling@apple.com>
+
+        Move BindingSecurity stuff under JSDOMBinding umbrella.
+        <https://webkit.org/b/124099>
+
+        We are hitting shouldAllowAccessToDOMWindow() pretty hard on the
+        demo here: <http://www.jasondavies.com/maps/rotate/>
+
+        Putting it together with the rest of the JSDOMBinding code takes
+        CPU time spent in there from 8.7% to 6.5%. The abstraction was
+        only used to support alternate JS engines in the past.
+
+        Reviewed by Antti Koivisto.
+
+2013-11-08  Brady Eidson  <beidson@apple.com>
+
+        Merge IDBDatabaseBackendInterface and IDBDatabaseBackendImpl
+        https://bugs.webkit.org/show_bug.cgi?id=124088
+
+        Reviewed by Tim Horton.
++
+        * CMakeLists.txt:
+        * GNUmakefile.list.am:
+        * WebCore.vcxproj/WebCore.vcxproj:
+        * WebCore.xcodeproj/project.pbxproj:
+
+        * Modules/indexeddb/IDBDatabaseBackendInterface.h: Removed.
+
+        * Modules/indexeddb/IDBDatabaseBackend.cpp: Renamed from Source/WebCore/Modules/indexeddb/IDBDatabaseBackendImpl.cpp.
+        * Modules/indexeddb/IDBDatabaseBackend.h: Renamed from Source/WebCore/Modules/indexeddb/IDBDatabaseBackendImpl.h.
+
+        * Modules/indexeddb/IDBBackingStoreInterface.h:
+        * Modules/indexeddb/IDBCallbacks.h:
+        * Modules/indexeddb/IDBCursor.cpp:
+        * Modules/indexeddb/IDBCursorBackend.cpp:
+        * Modules/indexeddb/IDBCursorBackend.h:
+        * Modules/indexeddb/IDBDatabase.cpp:
+        * Modules/indexeddb/IDBDatabase.h:
+        * Modules/indexeddb/IDBFactoryBackendInterface.h:
+        * Modules/indexeddb/IDBIndex.cpp:
+        * Modules/indexeddb/IDBIndex.h:
+        * Modules/indexeddb/IDBIndexWriter.h:
+        * Modules/indexeddb/IDBObjectStore.cpp:
+        * Modules/indexeddb/IDBObjectStore.h:
+        * Modules/indexeddb/IDBOpenDBRequest.cpp:
+        * Modules/indexeddb/IDBOpenDBRequest.h:
+        * Modules/indexeddb/IDBPendingDeleteCall.h:
+        * Modules/indexeddb/IDBRequest.cpp:
+        * Modules/indexeddb/IDBRequest.h:
+        * Modules/indexeddb/IDBTransaction.cpp:
+        * Modules/indexeddb/IDBTransaction.h:
+        * Modules/indexeddb/IDBTransactionBackend.cpp:
+        * Modules/indexeddb/IDBTransactionBackend.h:
+        * Modules/indexeddb/IDBTransactionBackendOperations.cpp:
+        * Modules/indexeddb/IDBTransactionBackendOperations.h:
+        * Modules/indexeddb/leveldb/IDBFactoryBackendLevelDB.cpp:
+        * Modules/indexeddb/leveldb/IDBFactoryBackendLevelDB.h:
+        * Modules/indexeddb/leveldb/IDBLevelDBCoding.cpp:
+        * Modules/indexeddb/leveldb/IDBLevelDBCoding.h:
+
+2013-11-08  Andreas Kling  <akling@apple.com>
+
+        Reindent JSDOMBinding.h (finally.)
+
+        Rubber-stamped by Sam Weinig.
+
+2013-11-08  Sam Weinig  <sam@webkit.org>
+
+        Change collectStyleForPresentationAttribute and related functions to take a MutableStylePropertySet by reference
+        https://bugs.webkit.org/show_bug.cgi?id=124096
+
+        Reviewed by Andreas Kling.
+
+        Pass MutableStylePropertySet by reference. It is never null.
+
+2013-11-08  Anders Carlsson  <andersca@apple.com>
+
+        Implement encoding of arrays of objects
+        https://bugs.webkit.org/show_bug.cgi?id=124091
+
+        Reviewed by Beth Dakin.
+
+        * history/HistoryItem.cpp:
+        (WebCore::HistoryItem::encodeBackForwardTreeNode):
+        Encode the rest of the members.
+
+        * platform/KeyedCoding.h:
+        (WebCore::KeyedEncoder::encodeObjects):
+        Call beginArray, then beginArrayElement/endArrayElement for every element and lastly endArray.
+
+2013-11-08  Tim Horton  <timothy_horton@apple.com>
+
+        Remote Layer Tree: RemoteLayerBackingStore partial repaint is broken for the tile cache
+        https://bugs.webkit.org/show_bug.cgi?id=123944
+
+        Reviewed by Simon Fraser.
+
+        Ensure that the tile cache retrieves repaint rects from the tile layer,
+        not from the tiled backing layer.
+
+        * platform/graphics/ca/PlatformCALayer.h:
+        * platform/graphics/ca/mac/PlatformCALayerMac.h:
+        * platform/graphics/ca/mac/PlatformCALayerMac.mm:
+        (PlatformCALayerMac::enumerateRectsBeingDrawn):
+        Allow the PlatformCALayer to decide how to enumerate rects to paint.
+
+        * WebCore.exp.in:
+        * platform/graphics/mac/WebLayer.h:
+        Add RepaintRectList, remove some unnecessary WebCore::s.
+
+        * platform/graphics/mac/WebLayer.mm:
+        (WebCore::collectRectsToPaint):
+        Factor collectRectsToPaint out; it just grabs the rects from
+        the layer and makes the decision whether to repaint the bounds
+        of the dirty region or just the subregions.
+
+        Move calls to collectRectsToPaint() to callers of drawLayerContents(),
+        so that TileController can collect rects from the appropriate source
+        (the Tile layer) and other layers just continue grabbing them from
+        their relevant layers.
+
+        Make sure that the list that comes from collectRectsToPaint() always
+        has at least one rect in it (appending the clip bounds if we don't
+        want to repaint subregions) so we can simplify logic in drawLayerContents.
+
+        (WebCore::drawLayerContents):
+        Remove code to support CompositingCoordinatesBottomUp, as it's only
+        used on Windows, so this Mac-specific code doesn't need to support it.
+
+        Simplify logic given that dirtyRects will always be non-empty.
+
+        (-[WebLayer drawInContext:]):
+        (-[WebSimpleLayer setNeedsDisplayInRect:]):
+        * platform/graphics/ca/mac/TileController.mm:
+        (WebCore::TileController::platformCALayerPaintContents):
+        Adopt collectRectsToPaint.
+
+2013-11-08  Anders Carlsson  <andersca@apple.com>
+
+        Implement more KeyedEncoder functionality
+        https://bugs.webkit.org/show_bug.cgi?id=124089
+
+        Reviewed by Beth Dakin.
+
+        * bindings/js/SerializedScriptValue.h:
+        * history/HistoryItem.cpp:
+        (WebCore::HistoryItem::encodeBackForwardTreeNode):
+        * platform/KeyedCoding.h:
+        (WebCore::KeyedEncoder::encodeConditionalObject):
+
+2013-11-08  Eric Carlson  <eric.carlson@apple.com>
+
+        getCueAsHTML() on an empty cue should return a document fragment
+        https://bugs.webkit.org/show_bug.cgi?id=124084
+
+        Reviewed by Darin Adler.
+
+        Test: media/track/track-cue-empty-cue-text.html
+
+        * html/track/WebVTTParser.cpp:
+        (WebCore::WebVTTParser::createDocumentFragmentFromCueText): Don't return early when the
+            passed an empty string.
+
+2013-11-08  Anders Carlsson  <andersca@apple.com>
+
+        KeyedEncoder should be able to encoder objects
+        https://bugs.webkit.org/show_bug.cgi?id=124085
+
+        Reviewed by Sam Weinig.
+
+        * history/HistoryItem.cpp:
+        (WebCore::HistoryItem::encodeBackForwardTree):
+        Encode the root object.
+
+        (WebCore::HistoryItem::encodeBackForwardTreeNode):
+        Encode the target.
+
+        * history/HistoryItem.h:
+        Add new members.
+
+        * platform/KeyedCoding.h:
+        (WebCore::KeyedEncoder::encodeObject):
+        Call beginObject, call the functor and then call endObject().
+
+2013-11-08  Sam Weinig  <sam@webkit.org>
+
+        Teach CanvasObserver about references
+        https://bugs.webkit.org/show_bug.cgi?id=124082
+
+        Reviewed by Anders Carlsson.
+
+        * css/CSSCanvasValue.cpp:
+        * css/CSSCanvasValue.h:
+        * html/HTMLCanvasElement.cpp:
+        * html/HTMLCanvasElement.h:
+
+2013-11-08  Anders Carlsson  <andersca@apple.com>
+
+        Begin stubbing out a KeyedEncoder class in WebCore
+        https://bugs.webkit.org/show_bug.cgi?id=124079
+
+        Reviewed by Sam Weinig.
+
+        * WebCore.exp.in:
+        Add symbol needed by WebCore.
+
+        * WebCore.xcodeproj/project.pbxproj:
+        Add new header file.
+
+        * history/HistoryItem.cpp:
+        (WebCore::HistoryItem::encodeBackForwardTree):
+        * history/HistoryItem.h:
+        Add an encodeBackForwardTree overload that takes a KeyedEncoder object. Encode the version.
+
+        * platform/KeyedCoding.h:
+        Add a KeyedEncoder class that just has a single encodeUInt32 member function for now.
+
+2013-11-08  Brady Eidson  <beidson@apple.com>
+
+        Merge IDBTransactionBackendInterface and IDBTransactionBackendImpl
+        https://bugs.webkit.org/show_bug.cgi?id=124077
+
+        Reviewed by Alexey Proskuryakov.
+
+        The abstraction is no longer needed.
+
+        * CMakeLists.txt:
+        * GNUmakefile.list.am:
+        * WebCore.xcodeproj/project.pbxproj:
+
+        * Modules/indexeddb/IDBTransactionBackendInterface.h: Removed.
+
+        * Modules/indexeddb/IDBTransactionBackend.cpp: Renamed from Source/WebCore/Modules/indexeddb/IDBTransactionBackendImpl.cpp.
+        * Modules/indexeddb/IDBTransactionBackend.h: Renamed from Source/WebCore/Modules/indexeddb/IDBTransactionBackendImpl.h.
+
+        * Modules/indexeddb/IDBBackingStoreInterface.h:
+        * Modules/indexeddb/IDBCursorBackend.cpp:
+        * Modules/indexeddb/IDBCursorBackend.h:
+        * Modules/indexeddb/IDBDatabaseBackendImpl.cpp:
+        * Modules/indexeddb/IDBDatabaseBackendImpl.h:
+        * Modules/indexeddb/IDBFactoryBackendInterface.h:
+        * Modules/indexeddb/IDBTransactionBackendOperations.h:
+        * Modules/indexeddb/IDBTransactionCoordinator.cpp:
+        * Modules/indexeddb/IDBTransactionCoordinator.h:
+        * Modules/indexeddb/leveldb/IDBBackingStoreLevelDB.cpp:
+        * Modules/indexeddb/leveldb/IDBBackingStoreLevelDB.h:
+        * Modules/indexeddb/leveldb/IDBFactoryBackendLevelDB.cpp:
+        * Modules/indexeddb/leveldb/IDBFactoryBackendLevelDB.h:
+
+2013-11-08  Simon Fraser  <simon.fraser@apple.com>
+
+        REGRESSION (r155660): Some Etherpad pages not scrollable with overlay scrollbars
+        https://bugs.webkit.org/show_bug.cgi?id=124075
+
+        Reviewed by Beth Dakin.
+        
+        In r155660 I removed some scrollbar-related layouts when scrollbars
+        are in overlay mode.
+        
+        However, ScrollView::updateScrollbars() has a case where we still need
+        to do multiple pases, related to its "Never ever try to both gain/lose a
+        scrollbar in the same pass" comment. When we avoid making a new scrollbar
+        because the other was removed, we need to do another pass to bring the
+        correct scrollbar back.
+
+        Can't test overlay scrollbars in tests.
+
+        * platform/ScrollView.cpp:
+        (WebCore::ScrollView::updateScrollbars):
+
+2013-11-08  Hans Muller  <hmuller@adobe.com>
+
+        [CSS Shapes] Image valued shape-outside that extends vertically into the margin-box is top-clipped
+        https://bugs.webkit.org/show_bug.cgi?id=123769
+
+        Reviewed by Dirk Schulze.
+
+        Remove the assumption that Y coordinates are >= 0 from the RasterShapeIntervals class
+        and correct its computeShapeMarginIntervals() method. The computeShapeMarginIntervals()
+        method now generates intervals with Y coordinates that begin at the image shape's
+        bounds.y - shape-margin, which may be less than 0.
+
+        The RasterShapeIntervals::intervalsAt() method now offsets its Y coordinate parameter
+        by the shape-margin. A non-const overload of the method was added to centralize all
+        access to m_intervalLists.
+
+        Test: fast/shapes/shape-outside-floats/shape-outside-floats-image-margin-004.html
+              fast/shapes/shape-outside-floats/shape-outside-floats-image-margin-005.html
+
+        * rendering/shapes/RasterShape.cpp:
+        (WebCore::MarginIntervalGenerator::intervalAt): Don't clip X coordinates to 0 since they can extend into the margin-box.
+        (WebCore::RasterShapeIntervals::appendInterval): Use the non-const intervalsAt() method.
+        (WebCore::RasterShapeIntervals::uniteMarginInterval): Ditto.
+        (WebCore::RasterShapeIntervals::computeShapeMarginIntervals): See above.
+        * rendering/shapes/RasterShape.h:
+        (WebCore::RasterShapeIntervals::RasterShapeIntervals): Added a field for the margin.
+        (WebCore::RasterShapeIntervals::intervalsAt): Offset y coordinates by the margin; added a non-const overload.
+
+2013-11-08  Piotr Grad  <p.grad@samsung.com>
+
+        Ended event should work also when playback rate is negative
+        https://bugs.webkit.org/show_bug.cgi?id=123879
+
+        Reviewed by Eric Carlson.
+
+        According to W3C specification playback is ended also when playback rate is
+        negative and position is the earliest possible position.
+
+        Test: media/video-ended-event-negative-playback.html
+
+        * html/HTMLMediaElement.cpp:
+        (WebCore::HTMLMediaElement::mediaPlayerTimeChanged):
+
+2013-11-08  Thiago de Barros Lacerda  <thiago.lacerda@openbossa.org>
+
+        Checking for TypeError in RTCPeerConnection object creation
+        https://bugs.webkit.org/show_bug.cgi?id=124049
+
+        Reviewed by Eric Carlson.
+
+        If invalid parameters are passed on RTCPeerConnection creation we must throw a TypeError exception.
+        According to the spec it requires a Dictionary argument, the RTCConfiguration, which is mandatory.
+
+        Please notice that this patch does not make every tests run as expected,
+        RTCPeerConnectionHandlerMock needs to be update to deal with contraints.
+
+        Existing tests were updated.
+
+        * GNUmakefile.list.am:
+        * Modules/mediastream/RTCPeerConnection.idl:
+        * UseJSC.cmake:
+        * WebCore.vcxproj/WebCore.vcxproj:
+        * WebCore.vcxproj/WebCore.vcxproj.filters:
+        * WebCore.xcodeproj/project.pbxproj:
+        * bindings/js/JSRTCPeerConnectionCustom.cpp: Added.
+        (WebCore::JSRTCPeerConnectionConstructor::constructJSRTCPeerConnection):
+
+2013-11-08  Bem Jones-Bey  <bjonesbe@adobe.com>
+
+        Use references instead of pointers to RenderBlockFlow in FloatingObjects and ComputeFloatOffsetAdapter
+        https://bugs.webkit.org/show_bug.cgi?id=124074
+
+        Reviewed by Sam Weinig.
+
+        Just a straightforward conversion from const pointers to const references.
+
+        Also, remove unneeded argument from FloatingObjects constructor.
+
+        No new tests, no behavior change.
+
+        * rendering/FloatingObjects.cpp:
+        (WebCore::ComputeFloatOffsetAdapter::ComputeFloatOffsetAdapter):
+        (WebCore::FloatingObjects::findNextFloatLogicalBottomBelow):
+        (WebCore::FloatingObjects::findNextFloatLogicalBottomBelowForBlock):
+        (WebCore::FloatingObjects::FloatingObjects):
+        (WebCore::FloatingObjects::clearLineBoxTreePointers):
+        (WebCore::FloatingObjects::computePlacedFloatsTree):
+        (WebCore::shapeInfoForFloat):
+        (WebCore::::updateOffsetIfNeeded):
+        (WebCore::::collectIfNeeded):
+        (WebCore::::heightRemaining):
+        * rendering/FloatingObjects.h:
+        * rendering/RenderBlockFlow.cpp:
+        (WebCore::RenderBlockFlow::createFloatingObjects):
+
+2013-11-08  Sam Weinig  <sam@webkit.org>
+
+        Teach SubframeLoader a bit about references
+        https://bugs.webkit.org/show_bug.cgi?id=124076
+
+        Reviewed by Anders Carlsson.
+
+        * html/HTMLAppletElement.cpp:
+        * html/HTMLFrameElementBase.cpp:
+        * html/HTMLMediaElement.cpp:
+        * loader/SubframeLoader.cpp:
+        * loader/SubframeLoader.h:
+
+2013-11-08  Brady Eidson  <beidson@apple.com>
+
+        Blind Windows build-fix attempt after r158959
+
+        * WebCore.vcxproj/WebCore.vcxproj:
+
+2013-11-08  Brady Eidson  <beidson@apple.com>
+
+        Merge IDBCursorBackendInterface and IDBCursorBackendImpl
+        https://bugs.webkit.org/show_bug.cgi?id=124068
+
+        Reviewed by Anders Carlsson.
+
+        * CMakeLists.txt:
+        * GNUmakefile.list.am:
+        * WebCore.xcodeproj/project.pbxproj:
+
+        * Modules/indexeddb/IDBCursorBackendInterface.h: Removed.
+
+        * Modules/indexeddb/IDBCursorBackend.cpp: Renamed from Source/WebCore/Modules/indexeddb/IDBCursorBackendImpl.cpp.
+        * Modules/indexeddb/IDBCursorBackend.h: Renamed from Source/WebCore/Modules/indexeddb/IDBCursorBackendImpl.h.
+
+        * Modules/indexeddb/IDBCallbacks.h:
+        * Modules/indexeddb/IDBCursor.cpp:
+        * Modules/indexeddb/IDBCursor.h:
+        * Modules/indexeddb/IDBCursorWithValue.cpp:
+        * Modules/indexeddb/IDBCursorWithValue.h:
+        * Modules/indexeddb/IDBDatabaseBackendImpl.cpp:
+        * Modules/indexeddb/IDBFactoryBackendInterface.h:
+        * Modules/indexeddb/IDBRequest.cpp:
+        * Modules/indexeddb/IDBRequest.h:
+        * Modules/indexeddb/IDBTransactionBackendImpl.cpp:
+        * Modules/indexeddb/IDBTransactionBackendImpl.h:
+        * Modules/indexeddb/IDBTransactionBackendInterface.h:
+        * Modules/indexeddb/IDBTransactionBackendOperations.cpp:
+        * Modules/indexeddb/leveldb/IDBFactoryBackendLevelDB.cpp:
+        * Modules/indexeddb/leveldb/IDBFactoryBackendLevelDB.h:
+
+2013-11-08  Sam Weinig  <sam@webkit.org>
+
+        Modernize FrameLoader a bit
+        https://bugs.webkit.org/show_bug.cgi?id=124073
+
+        Reviewed by Anders Carlsson.
+
+        * loader/FrameLoader.cpp:
+        * loader/FrameLoader.h:
+        Use std::unique_ptrs rather than OwnPtrs.
+
+        * loader/MixedContentChecker.cpp:
+        * loader/MixedContentChecker.h:
+        Switch to hold a Frame& rather than Frame*.
+
+2013-11-08  Zan Dobersek  <zdobersek@igalia.com>
+
+        Remove code guarded with ENABLE(STREAM)
+        https://bugs.webkit.org/show_bug.cgi?id=123667
+
+        Reviewed by Anders Carlsson.
+
+        Remove ENABLE(STREAM)-guarded code. This was added in the effort to provide Stream API support, but
+        no port enables the feature and the work on this feature has wound down after the Chromium port forked,
+        leaving this code unmaintained.
+
+        * fileapi/FileReaderLoader.cpp:
+        * fileapi/FileReaderLoader.h:
+
+2013-11-08  Bem Jones-Bey  <bjonesbe@adobe.com>
+
+        RenderBlockFlow::nextFloatLogicalBottomBelow should not use ShapeOutsideFloatOffsetMode
+        https://bugs.webkit.org/show_bug.cgi?id=123931
+
+        Reviewed by Sam Weinig.
+
+        Rewrite nextFloatLogicalBottomBelow to use the placed floats tree for
+        the search and to not need ShapeOutsideFloatOffsetMode anymore. This
+        moves almost all of the logic into FloatingObjects, making a small
+        reduction in the amount that RenderBlockFlow needs to know about the
+        implementation of FloatingObjects.
+
+        In addition, change ComputeFloatOffsetAdapter to take in LayoutUnits
+        and roundToInt itself so that all of it's callers can be simplified.
+
+        No new tests, no new behavior.
+
+        * rendering/FloatingObjects.cpp:
+        (WebCore::rangesIntersect):
+        (WebCore::ComputeFloatOffsetAdapter::ComputeFloatOffsetAdapter):
+        (WebCore::FindNextFloatLogicalBottomAdapter::FindNextFloatLogicalBottomAdapter):
+        (WebCore::FindNextFloatLogicalBottomAdapter::lowValue):
+        (WebCore::FindNextFloatLogicalBottomAdapter::highValue):
+        (WebCore::FindNextFloatLogicalBottomAdapter::nextLogicalBottom):
+        (WebCore::FindNextFloatLogicalBottomAdapter::nextShapeLogicalBottom):
+        (WebCore::FindNextFloatLogicalBottomAdapter::collectIfNeeded):
+        (WebCore::FloatingObjects::findNextFloatLogicalBottomBelow):
+        (WebCore::FloatingObjects::findNextFloatLogicalBottomBelowForBlock):
+        (WebCore::FloatingObjects::logicalLeftOffsetForPositioningFloat):
+        (WebCore::FloatingObjects::logicalRightOffsetForPositioningFloat):
+        (WebCore::FloatingObjects::logicalLeftOffset):
+        (WebCore::FloatingObjects::logicalRightOffset):
+        * rendering/FloatingObjects.h:
+        * rendering/LineWidth.cpp:
+        (WebCore::LineWidth::fitBelowFloats):
+        * rendering/RenderBlockFlow.cpp:
+        (WebCore::RenderBlockFlow::nextFloatLogicalBottomBelow):
+        (WebCore::RenderBlockFlow::nextFloatLogicalBottomBelowForBlock):
+        (WebCore::RenderBlockFlow::getClearDelta):
+        * rendering/RenderBlockFlow.h:
+
+2013-11-08  Alexey Proskuryakov  <ap@apple.com>
+
+        https://bugs.webkit.org/show_bug.cgi?id=124064
+        Some WebCrypto files are not in correct directories
+
+        Rubber-stamped by Anders Carlsson.
+
+        * WebCore.xcodeproj/project.pbxproj:
+        * crypto/CryptoAlgorithmAesCbcParams.h: Removed.
+        * crypto/CryptoAlgorithmAesKeyGenParams.h: Removed.
+        * crypto/CryptoKeyAES.cpp: Removed.
+        * crypto/CryptoKeyAES.h: Removed.
+        * crypto/CryptoKeyMac.cpp: Removed.
+        * crypto/keys/CryptoKeyAES.cpp: Copied from Source/WebCore/crypto/CryptoKeyAES.cpp.
+        * crypto/keys/CryptoKeyAES.h: Copied from Source/WebCore/crypto/CryptoKeyAES.h.
+        * crypto/mac/CryptoKeyMac.cpp: Copied from Source/WebCore/crypto/CryptoKeyMac.cpp.
+        * crypto/parameters/CryptoAlgorithmAesCbcParams.h: Copied from Source/WebCore/crypto/CryptoAlgorithmAesCbcParams.h.
+        * crypto/parameters/CryptoAlgorithmAesKeyGenParams.h: Copied from Source/WebCore/crypto/CryptoAlgorithmAesKeyGenParams.h.
+
+2013-11-08  Alexey Proskuryakov  <ap@apple.com>
+
+        Implement JWK key import for HMAC and AES-CBC
+        https://bugs.webkit.org/show_bug.cgi?id=124059
+
+        Reviewed by Anders Carlsson.
+
+        Tests: crypto/subtle/aes-cbc-import-jwk.html
+               crypto/subtle/hmac-import-jwk.html
+               crypto/subtle/import-jwk.html
+
+        WebCrypto supports multiple key formats - raw, pkcs8, spki, jwk. The design is that
+        we'll transform these into parsed KeyData subclasses before passing to algorithms.
+
+        CryptoKeySerialization is a base class for handling all these formats.
+
+        * WebCore.xcodeproj/project.pbxproj: Added new files. Removed CryptoKeyFormat.h.
+        
+        * bindings/js/JSCryptoKeySerializationJWK.h: Added.
+        * bindings/js/JSCryptoKeySerializationJWK.cpp: Added.
+        (WebCore::getStringFromJSON): A helper. Note that we can rely on the object being
+        a nice freshly parsed JSON, no getters or anything.
+        (WebCore::getBooleanFromJSON): Ditto.
+        (WebCore::JSCryptoKeySerializationJWK::JSCryptoKeySerializationJWK):
+        (WebCore::JSCryptoKeySerializationJWK::~JSCryptoKeySerializationJWK):
+        (WebCore::createHMACParameters): A simple helper.
+        (WebCore::JSCryptoKeySerializationJWK::reconcileAlgorithm): WebCrypto API is weird,
+        you can have algorithm parameters both inside a JWK key and passed as importKey()
+        arguments. They need to agree, whatever that means for specific serialization's
+        algorithm options (not all necessarily have a 1-1 matching to WebCrypto).
+        (WebCore::JSCryptoKeySerializationJWK::reconcileUsages): Take an intersection of usages.
+        (WebCore::JSCryptoKeySerializationJWK::reconcileExtractable): Only extractable if
+        both JWK and the caller agree.
+        (WebCore::JSCryptoKeySerializationJWK::keySizeIsValid): Verify validity of JWK key.
+        (WebCore::JSCryptoKeySerializationJWK::keyData): Return an appropriate KeyData
+        subclass.
+
+        * bindings/js/JSSubtleCryptoCustom.cpp:
+        (WebCore::ENUM_CLASS): Moved CryptoKeyFormat here.
+        (WebCore::cryptoKeyFormatFromJSValue): Added a human readable string to an exception.
+        (WebCore::JSSubtleCrypto::importKey): Support multiple key formats, not just raw.
+
+        * crypto/CryptoAlgorithm.cpp: (WebCore::CryptoAlgorithm::importKey):
+        * crypto/CryptoAlgorithm.h:
+        Updated signature for importKey to one that makes more sense. Decoding formats all
+        the way from a binary blob is not something that CryptoAlgorithm subclasses should
+        do, we now pass a KeyData subclass instead.
+        Removed exportKey/wrapKey/unwrapKey altogether, because I don't yet know what the
+        signatures will be.
+
+        * crypto/CryptoKeyData.h: Added.
+        (WebCore::CryptoKeyData::ENUM_CLASS):
+        (WebCore::CryptoKeyData::CryptoKeyData):
+        (WebCore::CryptoKeyData::~CryptoKeyData):
+        (WebCore::CryptoKeyData::format):
+        A base class for passing key material to algorithms. Currently, only one type is
+        supported, that being OctetSequence for secret keys. Keys for RSA and EC are more
+        complicated, and secret/public ones are different too.
+
+        * crypto/CryptoKeyFormat.h: Removed. There are too many key format classes
+        confusingly floating around, and this was only needed in one file for parsing.
+
+        * crypto/CryptoKeySerialization.h: Added.
+        Base class for handling raw/pkcs8/spki/jwk keys.
+
+        * crypto/algorithms/CryptoAlgorithmAES_CBC.cpp:
+        (WebCore::CryptoAlgorithmAES_CBC::importKey):
+        * crypto/algorithms/CryptoAlgorithmAES_CBC.h:
+        Updated to use CryptoKeyData.
+
+        * crypto/algorithms/CryptoAlgorithmHMAC.cpp:
+        (WebCore::CryptoAlgorithmHMAC::importKey):
+        * crypto/algorithms/CryptoAlgorithmHMAC.h:
+        Updated to use CryptoKeyData.
+
+        * crypto/keys/CryptoKeyDataOctetSequence.cpp: Added.
+        (WebCore::CryptoKeyDataOctetSequence::CryptoKeyDataOctetSequence):
+        (WebCore::CryptoKeyDataOctetSequence::~CryptoKeyDataOctetSequence):
+        * crypto/keys/CryptoKeyDataOctetSequence.h: Added.
+        (WebCore::asCryptoKeyDataOctetSequence):
+        * crypto/keys/CryptoKeySerializationRaw.cpp: Added.
+        (WebCore::CryptoKeySerializationRaw::CryptoKeySerializationRaw):
+        (WebCore::CryptoKeySerializationRaw::~CryptoKeySerializationRaw):
+        (WebCore::CryptoKeySerializationRaw::reconcileAlgorithm):
+        (WebCore::CryptoKeySerializationRaw::reconcileUsages):
+        (WebCore::CryptoKeySerializationRaw::reconcileExtractable):
+        (WebCore::CryptoKeySerializationRaw::keyData):
+        * crypto/keys/CryptoKeySerializationRaw.h: Added.
+        Much code to pass around a Vector<char>.
+
+2013-11-08  Mark Lam  <mark.lam@apple.com>
+
+        Move breakpoint (and exception break) functionality into JSC::Debugger.
+        https://bugs.webkit.org/show_bug.cgi?id=121796.
+
+        Reviewed by Geoffrey Garen.
+
+        No new tests.
+
+        - In ScriptDebugServer and JSC::Debugger, SourceID and BreakpointID are
+          now numeric tokens.
+
+        - JSC::Debugger now tracks user defined breakpoints in a JSC::Breakpoint
+          record. Previously, this info is tracked in the ScriptBreakpoint record
+          in ScriptDebugServer. The only element of ScriptBreakpoint that is not
+          being tracked by JSC::Breakpoint is the ScriptBreakpointAction.
+             The ScriptBreakpointAction is still tracked by the ScriptDebugServer
+          in a list keyed on the corresponding BreakpointID.
+             The ScriptBreakpoint record is now only used as a means of passing
+          breakpoint paramaters to the ScriptDebugServer.
+
+        - ScriptDebugServer now no longer accesses the JSC::CallFrame* directly.
+          It always goes through the DebuggerCallFrame.
+
+        * GNUmakefile.list.am:
+        * WebCore.vcxproj/WebCore.vcxproj:
+        * WebCore.vcxproj/WebCore.vcxproj.filters:
+        * WebCore.xcodeproj/project.pbxproj:
+        * bindings/js/BreakpointID.h: Added.
+        * bindings/js/ScriptDebugServer.cpp:
+        (WebCore::ScriptDebugServer::ScriptDebugServer):
+        (WebCore::ScriptDebugServer::setBreakpoint):
+        (WebCore::ScriptDebugServer::removeBreakpoint):
+        (WebCore::ScriptDebugServer::clearBreakpoints):
+        (WebCore::ScriptDebugServer::dispatchDidPause):
+        (WebCore::ScriptDebugServer::dispatchDidContinue):
+        (WebCore::ScriptDebugServer::dispatchDidParseSource):
+        (WebCore::ScriptDebugServer::notifyDoneProcessingDebuggerEvents):
+        (WebCore::ScriptDebugServer::needPauseHandling):
+        (WebCore::ScriptDebugServer::handleBreakpointHit):
+        (WebCore::ScriptDebugServer::handleExceptionInBreakpointCondition):
+        (WebCore::ScriptDebugServer::handlePause):
+        * bindings/js/ScriptDebugServer.h:
+        * bindings/js/SourceID.h: Added.
+        * bindings/js/WorkerScriptDebugServer.cpp:
+        (WebCore::WorkerScriptDebugServer::WorkerScriptDebugServer):
+        * bindings/js/WorkerScriptDebugServer.h:
+        * inspector/InspectorDebuggerAgent.cpp:
+        (WebCore::InspectorDebuggerAgent::InspectorDebuggerAgent):
+        (WebCore::parseLocation):
+        (WebCore::InspectorDebuggerAgent::setBreakpoint):
+        (WebCore::InspectorDebuggerAgent::continueToLocation):
+        (WebCore::InspectorDebuggerAgent::resolveBreakpoint):
+        (WebCore::InspectorDebuggerAgent::searchInContent):
+        (WebCore::InspectorDebuggerAgent::getScriptSource):
+        (WebCore::InspectorDebuggerAgent::didParseSource):
+        (WebCore::InspectorDebuggerAgent::didPause):
+        (WebCore::InspectorDebuggerAgent::clear):
+        * inspector/InspectorDebuggerAgent.h:
+        * inspector/ScriptDebugListener.h:
+
+2013-11-08  László Langó  <lango@inf.u-szeged.hu>
+
+        InspectorConsoleAgent::didFinishXHRLoading ConsoleMessage should include a column number
+        https://bugs.webkit.org/show_bug.cgi?id=114316
+
+        Reviewed by Timothy Hatcher.
+
+        InspectorConsoleAgent::didFinishXHRLoading creates a ConsoleMessage with a line number, 
+        but it should also include a column number. It looks like ultimately the line number comes from
+        JSXMLHttpRequest::send, it should also be possible to get the column number at the time.
+        The column number would be needed by the Web Inspector to jump to the proper place in source code 
+        to show where the XHR originated from.
+
+        * bindings/js/JSXMLHttpRequestCustom.cpp:
+        (WebCore::SendFunctor::SendFunctor):
+        (WebCore::SendFunctor::column):
+        (WebCore::SendFunctor::operator()):
+        (WebCore::JSXMLHttpRequest::send):
+        * inspector/InspectorConsoleAgent.cpp:
+        (WebCore::InspectorConsoleAgent::didFinishXHRLoading):
+        * inspector/InspectorConsoleAgent.h:
+        * inspector/InspectorInstrumentation.cpp:
+        (WebCore::InspectorInstrumentation::didFinishXHRLoadingImpl):
+        * inspector/InspectorInstrumentation.h:
+        (WebCore::InspectorInstrumentation::didFinishXHRLoading):
+        * xml/XMLHttpRequest.cpp:
+        (WebCore::XMLHttpRequest::XMLHttpRequest):
+        (WebCore::XMLHttpRequest::setLastSendLineAndColumnNumber):
+        (WebCore::XMLHttpRequest::didFinishLoading):
+        * xml/XMLHttpRequest.h:
+
+2013-11-08  Simon Fraser  <simon.fraser@apple.com>
+
+        Left sidebar on cubic-bezier.com flickers
+        https://bugs.webkit.org/show_bug.cgi?id=123128
+
+        Reviewed by Dean Jackson.
+        
+        The logic that determined whether position:fixed elements outside the viewport
+        should be composited was incorrect if the fixed element also had a transform.
+        
+        layer.calculateLayerBounds() only takes into account painted transforms (since they
+        affect layer bounds). So we need to compute the bounds relative to the layer
+        itself, then use localToContainerQuad() to map them to document coordinates,
+        but only to the RenderView so that we don't hit the page scale transform.
+
+        Tests: compositing/layer-creation/fixed-position-transformed-into-view.html
+               compositing/layer-creation/fixed-position-transformed-outside-view.html
+
+        * rendering/RenderLayerCompositor.cpp:
+        (WebCore::RenderLayerCompositor::requiresCompositingForPosition):
+
+2013-11-08  Martin Robinson  <mrobinson@igalia.com>
+
+        [MathML] Center of stretched curly bracket not always vertically centered
+        https://bugs.webkit.org/show_bug.cgi?id=123715
+
+        Reviewed by Brent Fulgham.
+
+        * rendering/mathml/RenderMathMLOperator.cpp:
+        (WebCore::RenderMathMLOperator::fillWithExtensionGlyph): Update an assertion and
+        handle the case where two glyph pieces abut.
+        (WebCore::RenderMathMLOperator::paint): Do not offset the center glyph by y().
+
+2013-10-30  Jer Noble  <jer.noble@apple.com>
+
+        [MSE] Bring SourceBuffer.append up to the most recent spec.
+        https://bugs.webkit.org/show_bug.cgi?id=123377
+
+        Reviewed by Eric Carlson.
+
+        Test: media/media-source/media-source-append-failed.html
+
+        Bring the MediaSource append() implementation up to the current spec.
+
+        * Modules/mediasource/SourceBuffer.cpp:
+        (WebCore::SourceBuffer::appendBufferInternal):
+        (WebCore::SourceBuffer::appendBufferTimerFired):
+        * platform/graphics/SourceBufferPrivate.h:
+        * platform/mock/mediasource/MockSourceBufferPrivate.cpp:
+        (WebCore::MockSourceBufferPrivate::append):
+        (WebCore::MockSourceBufferPrivate::evictCodedFrames):
+        (WebCore::MockSourceBufferPrivate::isFull):
+        * platform/mock/mediasource/MockSourceBufferPrivate.h:
+
+2013-11-07  Jer Noble  <jer.noble@apple.com>
+
+        [Mac] Crash at com.apple.WebCore: WebCore::MediaPlayerPrivateAVFoundationObjC::tracksDidChange + 26
+        https://bugs.webkit.org/show_bug.cgi?id=124031
+
+        Reviewed by Eric Carlson.
+
+        WTF::bind() causes errors when given a bare id pointer as a parameter,
+        when that parameter is casted to a specific NS type pointer (in this
+        case, a NSArray*) in order to pass it as a parameter to the bound
+        function.
+
+        Instead of passing around bare id pointers, wrap them in RetainPtr<>
+        objects before passing them to WTF::bind().
+
+        * platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.h:
+        * platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.mm:
+        (WebCore::MediaPlayerPrivateAVFoundationObjC::seekableTimeRangesDidChange):
+        (WebCore::MediaPlayerPrivateAVFoundationObjC::loadedTimeRangesDidChange):
+        (WebCore::MediaPlayerPrivateAVFoundationObjC::tracksDidChange):
+        (-[WebCoreAVFMovieObserver observeValueForKeyPath:ofObject:change:context:]):
+
+2013-11-08  Antti Koivisto  <antti@apple.com>
+
+        Templated LChar/UChar paths for simple line layout
+        https://bugs.webkit.org/show_bug.cgi?id=124035
+
+        Reviewed by Andreas Kling.
+
+        * rendering/SimpleLineLayout.cpp:
+        (WebCore::SimpleLineLayout::canUseForText):
+        (WebCore::SimpleLineLayout::canUseFor):
+        
+            Use a templated function to check for illegal characters.
+
+        (WebCore::SimpleLineLayout::skipWhitespaces):
+        
+            Make a template function.
+
+        (WebCore::SimpleLineLayout::textWidth):
+        
+            Make a template function plus some argument changes.
+
+        (WebCore::SimpleLineLayout::createTextRuns):
+        
+            Template function for creating runs while operating with either LChars or UChar.
+            Also simplified line breaking and text width measuring logic.
+
+        (WebCore::SimpleLineLayout::create):
+        
+            Pick the template.
+
+        * rendering/break_lines.cpp:
+        * rendering/break_lines.h:
+        
+            Move the implementation to the header (except for the table) so we can use the template
+            versions directly.
+
+        (WebCore::isBreakableSpace):
+        (WebCore::shouldBreakAfter):
+        (WebCore::needsLineBreakIterator):
+        (WebCore::nextBreakablePosition):
+        (WebCore::nextBreakablePositionIgnoringNBSP):
+
+2013-11-08  Mario Sanchez Prada  <mario.prada@samsung.com>
+
+        AX: [ATK] <span> elements exposed through ATK when not needed
+        https://bugs.webkit.org/show_bug.cgi?id=123885
+
+        Reviewed by Chris Fleizach.
+
+        As per SVN r158195, the way it's decided whether <span> elements
+        should be ignored or not has slightly changed, causing that the
+        GTK/EFL ports expose them in cases that they should be ignored,
+        such as for text elements that neither are focusable (e.g. by
+        explicitly setting tabindex) nor have a meaningful accessible name
+        suggesting they should be exposed.
+
+        As a result, the flattening that ATK based ports normally do for
+        this kind of elements (by folding them into their parents) do not
+        work correctly anymore, making two tests to fail:
+
+            platform/gtk/accessibility/spans-paragraphs-and-divs.html
+            platform/gtk/accessibility/spans.html
+
+        This patch encapsulates the part of the logic that affects this in
+        the computeAccessibilityIsIgnored() method, placing it in a
+        new method of AccessibilityObject that we can call from ATK's
+        accessibilityPlatformIncludesObject() to ensure we hide those
+        <span> elements when they don't fulfill those requirements.
+
+        * accessibility/AccessibilityObject.cpp:
+        (WebCore::AccessibilityObject::hasAttributesRequiredForInclusion):
+        New virtual method encapsulating part of the logic from the function
+        that computes whether accessibility should be ignored or not.
+        * accessibility/AccessibilityObject.h:
+
+        * accessibility/AccessibilityNodeObject.cpp:
+        (WebCore::AccessibilityNodeObject::hasAttributesRequiredForInclusion):
+        Override of the new method adding additional checks, as extracted from
+        the original bits in computeAccessibilityIsIgnored().
+        * accessibility/AccessibilityNodeObject.h:
+
+        * accessibility/AccessibilityRenderObject.cpp:
+        (WebCore::AccessibilityRenderObject::computeAccessibilityIsIgnored):
+        Use the newly added function where we had the original code before.
+
+        * accessibility/atk/AccessibilityObjectAtk.cpp:
+        (WebCore::AccessibilityObject::accessibilityPlatformIncludesObject):
+        Make sure <span> elements are ignored if they are not focusable
+        and they don't have a meaningful accessible name.
+
+2013-11-08  Carlos Garcia Campos  <cgarcia@igalia.com>
+
+        [GTK] Add missing symbols to WebKitDOMEventTarget.symbols
+        https://bugs.webkit.org/show_bug.cgi?id=123990
+
+        Reviewed by Philippe Normand.
+
+        Add webkit_dom_event_target_add_event_listener_with_closure and
+        webkit_dom_event_target_remove_event_listener_with_closure to the
+        symbols files.
+
+        * bindings/gobject/WebKitDOMEventTarget.symbols:
+        * bindings/gobject/webkitdom.symbols:
+
+2013-11-07  Carlos Garcia Campos  <cgarcia@igalia.com>
+
+        [GTK] Use deprecation guards around deprecated API in GObject DOM bindings
+        https://bugs.webkit.org/show_bug.cgi?id=123899
+
+        Reviewed by Martin Robinson.
+
+        Do not include deprecated API when compiling with
+        WEBKIT_DISABLE_DEPRECATED option.
+
+        * bindings/scripts/CodeGeneratorGObject.pm:
+        (GenerateFunction):
+        * bindings/scripts/test/GObject/WebKitDOMTestEventTarget.cpp:
+        * bindings/scripts/test/GObject/WebKitDOMTestEventTarget.h:
+
+2013-11-07  Brady Eidson  <beidson@apple.com>
+
+        Enhance SQL journal_mode setting code to be less likely to log an error.
+        <rdar://problem/15418577> and https://bugs.webkit.org/show_bug.cgi?id=124018
+
+        Reviewed by Anders Carlsson.
+
+        Even though the docs says SQLITE_ROW will always be returned, apparently SQLITE_OK is sometimes returned.
+        Change the code to handle that.
+
+        * platform/sql/SQLiteDatabase.cpp:
+        (WebCore::SQLiteDatabase::open): Save the statement result value, and accept SQLITE_OK as a non-error condition.
+
+2013-11-07  Brady Eidson  <beidson@apple.com>
+
+        Update an out-dated ASSERT in IconDatabase code.
+        <rdar://problem/15171118> and https://bugs.webkit.org/show_bug.cgi?id=124030.
+
+        Reviewed by Andreas Kling.
+
+        With the asynchronous interfaces that have been added and the support for WK2 that has been added, 
+        this ASSERT can incorrectly fire if an icon is asked for before database cleanup is allowed.
+
+        * loader/icon/IconDatabase.cpp:
+        (WebCore::IconDatabase::synchronousIconForPageURL): Update an invalid ASSERT.
+
+2013-11-07  Andreas Kling  <akling@apple.com>
+
+        RenderSVGResource helpers should take RenderStyle by const reference.
+        <https://webkit.org/b/124029>
+
+        Take const RenderStyle& instead of RenderStyle* in a few more places
+        so we can get rid of some ampersands and assertions.
+
+        Reviewed by Anders Carlsson.
+
+2013-11-07  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
+
+        [AX] Generate toAccessibilityTableRow|Column|Cell to detect bad type casts
+        https://bugs.webkit.org/show_bug.cgi?id=123984
+
+        Reviewed by Mario Sanchez Prada.
+
+        As a step to let static_cast<> use TYPE_CASTS_BASE, AccessibilityTableRow|Column|Cell use
+        ACCESSIBILITY_OBJECT_TYPE_CASTS which can support more helper functions rather than manual
+        static_cast<>. This change will help to detect bad type casts further.
+
+        No new tests, no behavior changes.
+
+        * accessibility/AccessibilityARIAGrid.cpp:
+        (WebCore::AccessibilityARIAGrid::addTableCellChild):
+        (WebCore::AccessibilityARIAGrid::addChildren):
+        * accessibility/AccessibilityARIAGridCell.cpp:
+        (WebCore::AccessibilityARIAGridCell::rowIndexRange):
+        * accessibility/AccessibilityTable.cpp:
+        (WebCore::AccessibilityTable::addChildren):
+        (WebCore::AccessibilityTable::rowHeaders):
+        (WebCore::AccessibilityTable::columnHeaders):
+        (WebCore::AccessibilityTable::cellForColumnAndRow):
+        * accessibility/AccessibilityTableCell.h:
+        * accessibility/AccessibilityTableColumn.h:
+        * accessibility/AccessibilityTableRow.cpp:
+        (WebCore::AccessibilityTableRow::headerObject):
+        * accessibility/AccessibilityTableRow.h:
+        * accessibility/atk/WebKitAccessibleInterfaceTable.cpp:
+        (cellAtIndex):
+        (webkitAccessibleTableGetColumnHeader):
+        (webkitAccessibleTableGetRowHeader):
+        * accessibility/ios/WebAccessibilityObjectWrapperIOS.mm:
+        (-[WebAccessibilityObjectWrapper tableCellParent]):
+        * accessibility/mac/WebAccessibilityObjectWrapperMac.mm:
+        (-[WebAccessibilityObjectWrapper accessibilityAttributeValue:]):
+
+2013-11-07  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
+
+        REGRESSION (r154375): Image is oriented incorrectly
+        https://bugs.webkit.org/show_bug.cgi?id=123831
+
+        Reviewed by Antonio Gomes.
+
+        r154375 made that shouldRespectImageOrientation() isn't used by drawImage().
+        It causes an image isn't oriented correctly. This patch sets shouldRespectImageOrientation()
+        value by default.
+
+        * rendering/RenderImage.cpp:
+        (WebCore::RenderImage::paintReplaced):
+        (WebCore::RenderImage::paintIntoRect):
+
+2013-11-07  Hans Muller  <hmuller@adobe.com>
+
+        [CSS Shapes] Image shape-outside with vertical gaps is handled incorrectly
+        https://bugs.webkit.org/show_bug.cgi?id=123934
+
+        Reviewed by Andreas Kling.
+
+        RasterShapeIntervals::getExcludedIntervals() was returning an empty
+        list when the line overlapped any vertical gap in the image. This short-circuit
+        had been mistakenly copied from getIncludedIntervals(), where it makes sense.
+
+        Test: fast/shapes/shape-outside-floats/shape-outside-floats-image-vgap.html
+
+        * rendering/shapes/RasterShape.cpp:
+        (WebCore::RasterShapeIntervals::getExcludedIntervals):
+
+2013-11-07  Simon Fraser  <simon.fraser@apple.com>
+
+        Make contents layer borders more visible
+        https://bugs.webkit.org/show_bug.cgi?id=124025
+
+        Reviewed by Tim Horton.
+
+        Layer borders for contents layers are impossible to see when the
+        contentsLayer has the same bounds as its parent; make the contents
+        layer border 4px thick to make it more visible.
+
+        * platform/graphics/ca/GraphicsLayerCA.cpp:
+        (WebCore::GraphicsLayerCA::setupContentsLayer):
+
+2013-11-07  Andreas Kling  <akling@apple.com>
+
+        InlineFlowBox always has a RenderBoxModelObject, take advantage.
+        <https://webkit.org/b/124024>
+
+        Since InlineFlowBox already has the branch-less renderer() returning
+        a RenderBoxModelObject&, avoid using InlineBox::boxModelObject()
+        wherever we have a tightly-typed box. One branch disappears from
+        every call site.
+
+        Deleted boxModelObject() on InlineFlowBox to prevent new code from
+        calling the less efficient function.
+
+        Reviewed by Anders Carlsson.
+
+2013-11-07  Andreas Kling  <akling@apple.com>
+
+        CTTE: Scrolling tree nodes should always have a ScrollingTree&.
+        <https://webkit.org/b/124022>
+
+        Let ScrollingTreeNode and subclasses store the backpointer to the
+        tree as a ScrollingTree& reference.
+
+        Reviewed by Anders Carlsson.
+
+2013-11-07  Simon Fraser  <simon.fraser@apple.com>
+
+        Lots of layers get solid color but transparent contents layers now
+        https://bugs.webkit.org/show_bug.cgi?id=123537
+
+        Reviewed by Tim Horton.
+        
+        We call rendererBackgroundColor() to determine the layer's background color,
+        but on most elements this returns the transparent color (a valid color).
+        This caused us to allocate a contentsLayer, and use the transparent color as its
+        backgroundColor, which was wasteful.
+        
+        Fix by only making a background-color layer if the color is not transparent (zero alpha).
+        
+        Also avoid making a new contents layer on every color change, and make sure that
+        we don't do implicit animations for backgroundColor, and some other properties
+        that were omitted by mistake.
+
+        Layer tree dumps don't dump content layers, so no way to test easily.
+
+        * platform/graphics/ca/GraphicsLayerCA.cpp:
+        (WebCore::GraphicsLayerCA::setContentsToSolidColor):
+        * platform/graphics/ca/mac/PlatformCALayerMac.mm:
+        (nullActionsDictionary):
+
+2013-11-07  Ryosuke Niwa  <rniwa@webkit.org>
+
+        DOMTokenList::add can add duplicated values if arguments had duplicated values
+        https://bugs.webkit.org/show_bug.cgi?id=123962
+
+        Reviewed by Benjamin Poulain.
+
+        Merge https://chromium.googlesource.com/chromium/blink/+/bd3822ad4ae3fc5d8f89f433a7bf04f697334305
+
+        In case we do element.classList.add('a', 'a') we need to ensure that we do not add the same token twice.
+        See http://dom.spec.whatwg.org/#dom-domtokenlist-add
+
+        * html/DOMTokenList.cpp:
+        (WebCore::DOMTokenList::add): Make sure filtered tokens are unique among themselves.
+
+2013-11-07  Eric Carlson  <eric.carlson@apple.com>
+
+        Remove npr.org specific hack in HTMLMediaElement
+        https://bugs.webkit.org/show_bug.cgi?id=123859
+
+        Reviewed by Jer Noble.
+
+        Remove the site specific hack added in r57820, it is no longer necessary.
+
+        * html/HTMLMediaElement.cpp:
+        (WebCore::HTMLMediaElement::HTMLMediaElement): Remove m_dispatchingCanPlayEvent.
+        (HTMLMediaElement::play): Don't special case npr.org.
+        * html/HTMLMediaElement.h:
+
+2013-11-07  Simon Fraser  <simon.fraser@apple.com>
+
+        Attempt to fix the 32-bit build. Virtual thunks seem to have different
+        symbol names between 32- and 64-bit.
+        
+        * WebCore.exp.in:
+
+2013-11-07  Ryosuke Niwa  <rniwa@webkit.org>
+
+        Crash in HTMLMediaElement::contextDestroyed
+        https://bugs.webkit.org/show_bug.cgi?id=123963
+
+        Reviewed by Eric Carlson.
+
+        Merge https://chromium.googlesource.com/chromium/blink/+/177999cdb34b707465670f0feff723922939f278
+
+        * html/HTMLMediaElement.cpp:
+        (WebCore::HTMLMediaElement::~HTMLMediaElement):
+
+2013-11-07  Jer Noble  <jer.noble@apple.com>
+
+        [Mac] Crash at com.apple.WebCore: -[WebCoreAVFMovieObserver observeValueForKeyPath:ofObject:change:context:] + 2084 
+        https://bugs.webkit.org/show_bug.cgi?id=124012
+
+        Reviewed by Eric Carlson.
+
+        The value of the 'duration' key is a NSConcreteValue wrapping a CMTime, not a NSNumber.
+
+        * platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.mm:
+        (-[WebCoreAVFMovieObserver observeValueForKeyPath:ofObject:change:context:]):
+
+2013-11-07  Ryosuke Niwa  <rniwa@webkit.org>
+
+        Crash when submitting form in a document with null encoding
+        https://bugs.webkit.org/show_bug.cgi?id=123975
+
+        Reviewed by Alexey Proskuryakov.
+
+        Merge https://chromium.googlesource.com/chromium/blink/+/bba01a7fff09e3053ada96ababac2a6e4261fe5f
+        
+        The CString object which is passed to normalizeLineEndingsToCRLF() can be
+        a null string. It is created in FormDataList::appendString(), and it
+        produces a null CString if FormDataList::m_encoding is a null encoding.
+
+        Test: fast/forms/form-submit-in-image-document.html
+
+        * platform/text/LineEnding.cpp:
+        (internalNormalizeLineEndingsToCRLF):
+
+2013-11-07  Anders Carlsson  <andersca@apple.com>
+
+        Use std::function for all policy continuation functions
+        https://bugs.webkit.org/show_bug.cgi?id=124011
+
+        Reviewed by Sam Weinig.
+
+        * loader/DocumentLoader.cpp:
+        (WebCore::DocumentLoader::willSendRequest):
+        (WebCore::DocumentLoader::responseReceived):
+        * loader/DocumentLoader.h:
+        * loader/FrameLoader.cpp:
+        (WebCore::FrameLoader::loadURL):
+        (WebCore::FrameLoader::load):
+        (WebCore::FrameLoader::loadWithDocumentLoader):
+        (WebCore::FrameLoader::loadPostRequest):
+        * loader/PolicyCallback.cpp:
+        (WebCore::PolicyCallback::clear):
+        (WebCore::PolicyCallback::set):
+        (WebCore::PolicyCallback::call):
+        (WebCore::PolicyCallback::clearRequest):
+        (WebCore::PolicyCallback::cancel):
+        * loader/PolicyCallback.h:
+        * loader/PolicyChecker.cpp:
+        (WebCore::PolicyChecker::checkNavigationPolicy):
+        (WebCore::PolicyChecker::checkNewWindowPolicy):
+        (WebCore::PolicyChecker::checkContentPolicy):
+        * loader/PolicyChecker.h:
+
+2013-11-07  Brady Eidson  <beidson@apple.com>
+
+        Use SQLite journal mode WAL (WriteAheadLogging)
+        https://bugs.webkit.org/show_bug.cgi?id=124009
+
+        Reviewed by Anders Carlsson.
+
+        WriteAheadLogging journalling is better than the traditional rollback model.
+
+        * platform/sql/SQLiteDatabase.cpp:
+        (WebCore::SQLiteDatabase::open): Use a PRAGMA to set journal_mode to WAL.
+
+2013-11-07  Mark Lam  <mark.lam@apple.com>
+
+        Cosmetic: rename xxxId to xxxID for ScriptId, SourceId, and BreakpointId.
+        https://bugs.webkit.org/show_bug.cgi?id=123945.
+
+        Reviewed by Geoffrey Garen.
+
+        No new tests.
+
+        * bindings/js/JSInjectedScriptHostCustom.cpp:
+        (WebCore::JSInjectedScriptHost::functionDetails):
+        * bindings/js/JavaScriptCallFrame.h:
+        (WebCore::JavaScriptCallFrame::sourceID):
+        * bindings/js/ScriptDebugServer.cpp:
+        (WebCore::ScriptDebugServer::ScriptDebugServer):
+        (WebCore::ScriptDebugServer::setBreakpoint):
+        (WebCore::ScriptDebugServer::removeBreakpoint):
+        (WebCore::ScriptDebugServer::hasBreakpoint):
+        (WebCore::ScriptDebugServer::clearBreakpoints):
+        (WebCore::ScriptDebugServer::updateCallFrame):
+        (WebCore::ScriptDebugServer::pauseIfNeeded):
+        * bindings/js/ScriptDebugServer.h:
+        * inspector/InspectorConsoleAgent.cpp:
+        (WebCore::InspectorConsoleAgent::addMessageToConsole):
+        * inspector/InspectorConsoleAgent.h:
+        * inspector/InspectorConsoleInstrumentation.h:
+        (WebCore::InspectorInstrumentation::addMessageToConsole):
+        * inspector/InspectorDOMAgent.cpp:
+        (WebCore::InspectorDOMAgent::buildObjectForEventListener):
+        * inspector/InspectorDebuggerAgent.cpp:
+        (WebCore::InspectorDebuggerAgent::setBreakpointByUrl):
+        (WebCore::parseLocation):
+        (WebCore::InspectorDebuggerAgent::setBreakpoint):
+        (WebCore::InspectorDebuggerAgent::removeBreakpoint):
+        (WebCore::InspectorDebuggerAgent::continueToLocation):
+        (WebCore::InspectorDebuggerAgent::resolveBreakpoint):
+        (WebCore::InspectorDebuggerAgent::searchInContent):
+        (WebCore::InspectorDebuggerAgent::setScriptSource):
+        (WebCore::InspectorDebuggerAgent::getScriptSource):
+        (WebCore::InspectorDebuggerAgent::compileScript):
+        (WebCore::InspectorDebuggerAgent::runScript):
+        (WebCore::InspectorDebuggerAgent::didParseSource):
+        (WebCore::InspectorDebuggerAgent::didPause):
+        (WebCore::InspectorDebuggerAgent::clear):
+        (WebCore::InspectorDebuggerAgent::reset):
+        * inspector/InspectorDebuggerAgent.h:
+        * inspector/InspectorInstrumentation.cpp:
+        (WebCore::InspectorInstrumentation::addMessageToConsoleImpl):
+        * inspector/InspectorInstrumentation.h:
+        * inspector/ScriptDebugListener.h:
+
+2013-11-07  Cidorvan Leite  <cidorvan.leite@openbossa.org>
+
+        Avoid invalid cairo matrix when drawing surfaces too small
+        https://bugs.webkit.org/show_bug.cgi?id=123810
+
+        Drawing surfaces too small makes inverse matrix with values too big,
+        when this happen, cairo context is not valid anymore and it stops to draw anything.
+
+        Reviewed by Martin Robinson.
+
+        Test: fast/canvas/drawImage-with-small-values.html
+
+        * platform/graphics/cairo/PlatformContextCairo.cpp:
+        (WebCore::PlatformContextCairo::drawSurfaceToContext):
+
+2013-11-07  Antti Koivisto  <antti@apple.com>
+
+        Simple line layout crashes with SVG fonts
+        https://bugs.webkit.org/show_bug.cgi?id=124002
+
+        Reviewed by Simon Fraser.
+        
+        Don't use simple line layout for flows using SVG fonts. They crash if kerning is enabled.
+
+        Test: fast/text/svg-font-simple-line-crash.html
+
+        * platform/graphics/Font.h:
+        (WebCore::Font::isSVGFont):
+        
+            Add isSVGFont() so callers don't need to go via primaryFont().
+
+        * rendering/InlineTextBox.cpp:
+        (WebCore::InlineTextBox::constructTextRun):
+        * rendering/RenderBlock.cpp:
+        (WebCore::constructTextRunInternal):
+        * rendering/SimpleLineLayout.cpp:
+        (WebCore::SimpleLineLayout::canUseFor):
+        
+            Disallow SVG fonts.
+
+        * rendering/svg/SVGInlineTextBox.cpp:
+        (WebCore::SVGInlineTextBox::constructTextRun):
+        * rendering/svg/SVGTextMetrics.cpp:
+        (WebCore::SVGTextMetrics::constructTextRun):
+        (WebCore::SVGTextMetrics::SVGTextMetrics):
+        * rendering/svg/SVGTextRunRenderingContext.h:
+        
+            Get rid of the abstract textRunNeedsRenderingContext in favor of just testing isSVGFont().
+
+2013-11-07  Simon Fraser  <simon.fraser@apple.com>
+
+        Allow customization of the contentsScale of TileController tiles
+        https://bugs.webkit.org/show_bug.cgi?id=124004
+
+        Reviewed by Tim Horton.
+
+        On some platorms, zooming out on pages with TiledBacking compositing
+        layers can cause very high memory use, because the TiledBacking retains
+        the original page scale while the zoom is in flight, but can be asked
+        to cover a large area.
+        
+        Make it possible to reduce memory use in this case by allowing RenderLayerCompositor
+        to provide an additional scale factor for newly created tiles. Platforms can
+        then customize this to create low-res tiles when necessary.
+
+        * WebCore.exp.in:
+        * platform/graphics/GraphicsLayerClient.h:
+        (WebCore::GraphicsLayerClient::contentsScaleMultiplierForNewTiles):
+        * platform/graphics/ca/GraphicsLayerCA.cpp:
+        (WebCore::GraphicsLayerCA::platformCALayerContentsScaleMultiplierForNewTiles):
+        * platform/graphics/ca/GraphicsLayerCA.h:
+        * platform/graphics/ca/PlatformCALayerClient.h:
+        (WebCore::PlatformCALayerClient::platformCALayerContentsScaleMultiplierForNewTiles):
+        * platform/graphics/ca/mac/TileController.h:
+        * platform/graphics/ca/mac/TileController.mm:
+        (WebCore::TileController::TileController):
+        (WebCore::TileController::setScale):
+        (WebCore::TileController::createTileLayer):
+        * rendering/RenderLayerBacking.cpp:
+        (WebCore::RenderLayerBacking::contentsScaleMultiplierForNewTiles):
+        * rendering/RenderLayerBacking.h:
+        * rendering/RenderLayerCompositor.cpp:
+        (WebCore::RenderLayerCompositor::contentsScaleMultiplierForNewTiles):
+        * rendering/RenderLayerCompositor.h:
+
 2013-11-07  Jer Noble  <jer.noble@apple.com>
 
         Unreviewed Win build fix after r158855; wrap shapeInfoForFloat() in an ENABLE(CSS_SHAPES) guard.