Unreviewed, rolling out r137491.
[WebKit-https.git] / Source / WebCore / ChangeLog
index b988765..44fe3a0 100644 (file)
@@ -1,3 +1,749 @@
+2012-12-12  Sheriff Bot  <webkit.review.bot@gmail.com>
+
+        Unreviewed, rolling out r137491.
+        http://trac.webkit.org/changeset/137491
+        https://bugs.webkit.org/show_bug.cgi?id=104828
+
+        broke the builtd (Requested by rfong on #webkit).
+
+        * DerivedSources.cpp:
+        * WebCore.vcproj/WebCore.vcproj:
+        * WebCore.vcproj/WebCoreCommon.vsprops:
+        * WebCore.vcproj/copyForwardingHeaders.cmd:
+        * bindings/generic/RuntimeEnabledFeatures.cpp:
+        (WebCore):
+        * html/HTMLElementsAllInOne.cpp:
+        * platform/NotImplemented.cpp:
+        (WebCore):
+        (WebCore::notImplementedLoggingChannel):
+        * rendering/RenderingAllInOne.cpp:
+
+2012-12-12  John J. Barton  <johnjbarton@chromium.org>
+
+        Web Inspector: remove protocol-externs.js temp file after compile-front-end.py
+        https://bugs.webkit.org/show_bug.cgi?id=98799
+
+        Reviewed by Vsevolod Vlasov.
+
+        Add os.system rm for protocol-externs.js to the end of the compile script
+
+        * inspector/compile-front-end.py:
+
+2012-12-12  Robert Hogan  <robert@webkit.org>
+
+        White space between inline blocks should be affected by word-spacing property
+        https://bugs.webkit.org/show_bug.cgi?id=69072
+
+        Reviewed by Levi Weintraub.
+
+        Allow word-spacing to get added to the spaces between inline blocks in placeBoxesInInlineDirection by setting
+        needsWordSpacing to true each time an inline block is encountered. This ensures that the next space encountered
+        in a text run will be treated as a space between two words and will get word-spacing added to it if it is present
+        in the style.
+
+        Tests: fast/css/word-spacing-between-blocks.html
+               fast/css/word-spacing-between-inlines.html
+
+        * rendering/InlineFlowBox.cpp:
+        (WebCore::InlineFlowBox::placeBoxRangeInInlineDirection):
+
+2012-12-12  Xianzhu Wang  <wangxianzhu@chromium.org>
+
+        Pre-painting should not paint out-of-view fixed position elements
+        https://bugs.webkit.org/show_bug.cgi?id=104724
+
+        Reviewed by James Robinson.
+
+        Test: ManualTests/compositing/fixed-position-out-of-view-scroll-prepaint.html
+
+        * rendering/RenderLayer.cpp:
+        (WebCore::RenderLayer::paintLayer): Don't paint out-of-view fixed position elements.
+
+2012-12-11  Roger Fong  <roger_fong@apple.com>
+
+        Enable VIDEO_TRACK on Windows.
+        https://bugs.webkit.org/show_bug.cgi?id=104706.
+
+        Reviewed by Timothy Horton.
+
+        Enable VIDEO_TRACK on Windows and update project property files.
+
+        * DerivedSources.cpp:
+        * WebCore.vcproj/WebCore.vcproj:
+        * WebCore.vcproj/WebCoreCommon.vsprops:
+        * WebCore.vcproj/copyForwardingHeaders.cmd:
+        * bindings/generic/RuntimeEnabledFeatures.cpp:
+        (WebCore):
+        * html/HTMLElementsAllInOne.cpp:
+        * rendering/RenderingAllInOne.cpp:
+
+2012-12-12  James Simonsen  <simonjam@chromium.org>
+
+        [Resource Timing] Failed resources shouldn't be recorded in the buffer
+        https://bugs.webkit.org/show_bug.cgi?id=104762
+
+        Reviewed by Tony Gentilcore.
+
+        Test: http/tests/w3c/webperf/submission/resource-timing/html/test_resource_ignore_failures.html
+
+        * loader/cache/CachedResourceLoader.cpp:
+        (WebCore::CachedResourceLoader::loadDone):
+
+2012-12-12  Martin Robinson  <mrobinson@igalia.com>
+
+        [GTK] Add authentication support to DRT and fix exposed issues in the libsoup backend
+        https://bugs.webkit.org/show_bug.cgi?id=104285
+
+        Reviewed by Gustavo Noronha Silva.
+
+        Fix issues in the libsoup networking backend that were causing test
+        failures.
+
+        No new tests. This patch unskips authentication tests.
+
+        * platform/network/soup/ResourceHandleSoup.cpp:
+        (WebCore::isAuthenticationFailureStatusCode): Added this helper to catch authentication status codes
+        and use it everywhere.
+        (WebCore::applyAuthenticationToRequest): Use an early return instead of an if-statement
+        and instead of always using manually specified credentials, apply those found in the session
+        storage as well. Instead of applying the credentials to firstRequest always, accept the request
+        as a parameter, so this method can be properly used with redirects.
+        (WebCore::restartedCallback): Instead of creating the KURL directly from soup_uri_to_string
+        use soupURIToKURL, which preserves the password from the SoupURI. soup_uri_to_string doesn't
+        include the password. When calling applyAuthenticationToRequest update the new request. Before
+        applyAuthenticationToRequest always updated the firstRequest. Do not call willSendRequest for
+        instances when this method is called for authentication handling. willSendRequest should only
+        be called for server-side redirects.
+        (WebCore::createSoupRequestAndMessageForHandle): Use soup_requester_request_uri instead of
+        soup_requester_request. Request::soupURI properly creates a SoupURI with non-null empty strings
+        when either the username and password are empty.  A null username or password does not trigger
+        the URI-embedded credential override in SoupAuthenticationManager within libsoup. This is important
+        because sometimes either the username or password might empty for a request, but the other
+        part of the credential should still override the per-session credential storage.
+        (WebCore::ResourceHandle::start): We don't need to clear the username and password from the
+        internal data structure any longer. These need to be cleared in the CFNetwork backend because
+        it prevents the same failed per-request credentials from being used over and over again. In our
+        case we don't use these members during didReceiveAuthenticationChallenge, so this isn't an issue.
+        (WebCore::ResourceHandle::didReceiveAuthenticationChallenge): Don't look at the per-request
+        m_user/m_pass pair here any longer. We use these when initially creating a request, so we don't
+        want to re-use them if they fail.
+        * platform/network/soup/ResourceRequest.h:
+        (ResourceRequest): Rename urlStringForSoup to soupURI.
+        * platform/network/soup/ResourceRequestSoup.cpp:
+        (WebCore::ResourceRequest::soupURI): Instead of returning a string for the soup-bound URL
+        return a SoupURI that properly embeds empty usernames and passwords.
+
+2012-12-12  Alexey Proskuryakov  <ap@apple.com>
+
+        Make LOG() work in WebProcess and NetworkProcess
+        https://bugs.webkit.org/show_bug.cgi?id=104718
+
+        Reviewed by Tim Horton.
+
+        Split initialization function into a new InitializeLogging.h file, because it's
+        needed in WebKit, but cannot be included along with WebKit version of this file.
+        We'd have a confilct between LOG_CHANNEL_PREFIX definitions if we tried.
+
+        * WebCore.xcodeproj/project.pbxproj: Make Logging.h a project header, and
+        InitializeLogging.h a private one.
+
+        * GNUmakefile.list.am:
+        * Target.pri:
+        * platform/InitializeLogging.h: Added.
+        * platform/Logging.h:
+
+        * platform/blackberry/LoggingBlackBerry.cpp:
+        * platform/efl/LoggingEfl.cpp:
+        * platform/gtk/LoggingGtk.cpp:
+        * platform/mac/LoggingMac.mm:
+        * platform/qt/LoggingQt.cpp:
+        * platform/win/LoggingWin.cpp:
+        * platform/wx/LoggingWx.cpp:
+        Include InitializeLogging.h where this function is implemented.
+
+        * platform/NotImplemented.h:
+        * platform/NotImplemented.cpp: Added.
+        Changed to not include WebCore's Logging.h in the header, because NotImplemented.h
+        is used in WebKits, which have incompatibe logging headers.
+
+2012-12-12  Chris Fleizach  <cfleizach@apple.com>
+
+        AX: aria-busy should be exposed everywhere, not just on live regions
+        https://bugs.webkit.org/show_bug.cgi?id=104814
+
+        Reviewed by Anders Carlsson.
+
+        The aria-busy attribute should be exposed on all elements. 
+        This isn't entirely clear in the ARIA spec, but that will be fixed with
+        ARIA1.1 (ISSUE-538).
+
+        Existing tests that dump attributes cover this new behavior.
+
+        * accessibility/mac/WebAccessibilityObjectWrapper.mm:
+        (-[WebAccessibilityObjectWrapper additionalAccessibilityAttributeNames]):
+
+2012-12-12  Gavin Peters  <gavinp@chromium.org>
+
+        [chromium] Add destructor to WebPrerender
+        https://bugs.webkit.org/show_bug.cgi?id=104820
+
+        Reviewed by James Robinson.
+
+        When making WebPrerender default constructible and assignable, the destructor should have been moved out of WEBKIT_IMPLEMENTATION.
+
+        Existing WebKit unit tests should test this, as do chromium browser tests.
+
+        * platform/chromium/support/WebPrerender.cpp:
+        (WebKit::WebPrerender::toPrerender):
+        (WebKit::WebPrerender::reset):
+
+2012-12-12  Allan Sandfeld Jensen  <allan.jensen@digia.com>
+
+        [Qt] Animation fails on large layers
+        https://bugs.webkit.org/show_bug.cgi?id=104538
+
+        Reviewed by Noam Rosenthal.
+
+        The backing tiles had no upper limit defined for the non-GL backend, which could cause them
+        to allocate pixmaps widier or higher than what the underlying graphics systems can handle.
+
+        On top of that GraphicsLayerTextureMapper::prepareBackingStore() would allocate an intermediate
+        pixmap the size of the dirty rect, which would at least on the first paint be the size of the
+        entire layer, again causing allocation of pixmaps with dimensions outside of platform bounds.
+
+        This patch introduces a limit to the size of image buffer tiles, and adds an alternative path
+        for painting where the GraphicsLayer paints directly to the tile instead of over an intermediate
+        pixmap. This alternative path can also be useful later to minimize the amount of pixel copying
+        happening in full repaints.
+
+        * platform/graphics/texmap/GraphicsLayerTextureMapper.cpp:
+        (WebCore::GraphicsLayerTextureMapper::prepareBackingStore):
+        * platform/graphics/texmap/GraphicsLayerTextureMapper.h:
+        (GraphicsLayerTextureMapper):
+        * platform/graphics/texmap/TextureMapper.cpp:
+        (WebCore::BitmapTexture::updateContents):
+        * platform/graphics/texmap/TextureMapper.h:
+        (BitmapTexture):
+        (TextureMapper):
+        * platform/graphics/texmap/TextureMapperBackingStore.cpp:
+        (WebCore::TextureMapperTile::updateContents):
+        (WebCore::TextureMapperTiledBackingStore::updateContents):
+        * platform/graphics/texmap/TextureMapperBackingStore.h:
+        (TextureMapperTile):
+        (TextureMapperTiledBackingStore):
+        * platform/graphics/texmap/TextureMapperImageBuffer.cpp:
+        (WebCore::BitmapTextureImageBuffer::updateContents):
+        (WebCore::TextureMapperImageBuffer::maxTextureSize):
+        * platform/graphics/texmap/TextureMapperImageBuffer.h:
+        (BitmapTextureImageBuffer):
+
+2012-12-12  Aaron Colwell  <acolwell@chromium.org>
+
+        Remove MediaSource 'ended' to 'open' transition when seeking.
+        https://bugs.webkit.org/show_bug.cgi?id=104583
+
+        Reviewed by Eric Carlson.
+
+        Remove the 'ended' to 'open' transition on a seek.
+
+        No new tests. The following tests were updated to reflect the new behavior.
+          http/tests/media/media-source/video-media-source-seek.html
+          http/tests/media/media-source/video-media-source-state-changes.html
+
+        * html/HTMLMediaElement.cpp:
+        (WebCore::HTMLMediaElement::seek):
+
+2012-12-12  Julien Chaffraix  <jchaffraix@webkit.org>
+
+        [CSS Grid Layout] Implement CSS parsing and handling for min-content and max-content
+        https://bugs.webkit.org/show_bug.cgi?id=104747
+
+        Reviewed by Alexis Menard.
+
+        This change adds -webkit-min-content and -webkit-max-content to our <track-breadth> parsing.
+        The rendering is not plugged in and will probably depend on bug 103311.
+
+        Tests: fast/css-grid-layout/grid-columns-rows-get-set-multiple.html:
+               fast/css-grid-layout/grid-columns-rows-get-set.html:
+
+        * css/CSSParser.cpp:
+        (WebCore::CSSParser::parseGridBreadth):
+        * css/StyleResolver.cpp:
+        (WebCore::createGridTrackBreadth):
+        Straight-forward extension of the new code. We were already serializing back properly
+        so no need to touch CSSComputedStyleDeclaration.
+
+2012-12-12  Ryosuke Niwa  <rniwa@webkit.org>
+
+        Copying text with ruby inserts new lines around rt elements
+        https://bugs.webkit.org/show_bug.cgi?id=104765
+
+        Reviewed by Dan Bernstein.
+
+        Don't emit new lines before and after rt.
+
+        Test: editing/pasteboard/copy-paste-ruby-text.html
+
+        * editing/TextIterator.cpp:
+        (WebCore::shouldEmitNewlinesBeforeAndAfterNode):
+
+2012-12-12  Krzysztof Czech  <k.czech@samsung.com>
+
+        [EFL] Possibility to turn off accessibility feature for WebKit-EFL.
+        https://bugs.webkit.org/show_bug.cgi?id=103036
+
+        Reviewed by Laszlo Gombos.
+
+        Guard dependencies for accessibility (ATK), so that they can be turned off.
+
+        No new tests as there is no new functionality.
+
+        * PlatformEfl.cmake:
+        * accessibility/AccessibilityObject.cpp:
+        (WebCore::AccessibilityObject::AccessibilityObject):
+        * accessibility/AccessibilityObject.h:
+        (AccessibilityObject):
+
+2012-12-12  Justin Novosad  <junov@google.com>
+
+        Use render box background over border draw strategy in cases with background-image
+        https://bugs.webkit.org/show_bug.cgi?id=103409
+
+        Reviewed by Simon Fraser.
+
+        The BackgroundOverBorderBleedAvoidance mode was not being used in
+        cases where the background is an opaque image. It was also not
+        being used in cases with multiple layers. The multiple layer
+        case is now safe with respect to color bleeding if the top layer
+        is opaque, thank to the recent addition of layer occlusion culling
+        in r135629.
+
+        No new tests. Testing already covered by the following tests:
+        fast/backgrounds/background-opaque-images-over-color.html
+        fast/backgrounds/gradient-background-leakage-2.html
+
+        * rendering/RenderBox.cpp:
+        (WebCore::RenderBox::determineBackgroundBleedAvoidance):
+        (WebCore::RenderBox::backgroundHasOpaqueTopLayer):
+        * rendering/RenderBox.h:
+        (RenderBox):
+
+2012-12-12  Ilya Tikhonovsky  <loislo@chromium.org>
+
+        Web Inspector: instrument static parts of RenderBlock and RenderBox
+        https://bugs.webkit.org/show_bug.cgi?id=104660
+
+        Reviewed by Yury Semikhatsky.
+
+        Plain vanilla instrumentation.
+
+        * dom/WebCoreMemoryInstrumentation.cpp:
+        (WebCore):
+        (WebCore::WebCoreMemoryInstrumentation::reportMemoryUsage):
+        * dom/WebCoreMemoryInstrumentation.h:
+        (WebCoreMemoryTypes):
+        (WebCoreMemoryInstrumentation):
+        (WebCore):
+        * inspector/InspectorMemoryAgent.cpp:
+        (WebCore::InspectorMemoryAgent::getProcessMemoryDistribution):
+        * rendering/RenderBlock.cpp:
+        (WebCore::RenderBlock::reportStaticMembersMemoryUsage):
+        (WebCore):
+        * rendering/RenderBlock.h:
+        (RenderBlock):
+        * rendering/RenderBox.cpp:
+        (WebCore::RenderBox::reportStaticMembersMemoryUsage):
+        (WebCore):
+        * rendering/RenderBox.h:
+        (RenderBox):
+
+2012-12-12  Ilya Tikhonovsky  <loislo@chromium.org>
+
+        Web Inspector: Native Memory Instrumentation: remove fake root MemoryObjectInfo.
+        https://bugs.webkit.org/show_bug.cgi?id=104796
+
+        Reviewed by Yury Semikhatsky.
+
+        It was not a good idea to introduce a fake root MemoryObjectInfo.
+        It makes a problem when we visit an object without its own MemoryObjectType.
+
+        Example: RenderBox has a global pointer to a hash map.
+        HashMap doesn't have its own object type because it is a generic container.
+        It will inherit object type from the fake root memory object info.
+        The same could happen for another container in another class with other MemoryObjectType.
+
+        This fact forces me to create custom process method for root objects
+        because they need to have their own MemoryObjectInfo with customisable memory object type.
+
+        Drive by fix: InstrumentedPointer* was replaced with Wrapper* because actually it is using
+        for instrumented and not instrumented object classes.
+
+        * inspector/MemoryInstrumentationImpl.cpp:
+        (WebCore::MemoryInstrumentationImpl::processDeferredObjects):
+        (WebCore::MemoryInstrumentationImpl::deferObject):
+        (WebCore::MemoryInstrumentationImpl::reportMemoryUsage):
+        * inspector/MemoryInstrumentationImpl.h:
+        (MemoryInstrumentationImpl):
+
+2012-12-12  Kondapally Kalyan  <kalyan.kondapally@intel.com>
+
+        [EFL [WebGL] [Wk2] Resizing the canvas breaks WebGL.
+        https://bugs.webkit.org/show_bug.cgi?id=104535.
+
+        Reviewed by Kenneth Rohde Christiansen.
+
+        GraphicsContext3DEfl does not re-create the surface on canvas resize, this
+        conflicted with the logic in coordinated graphics Layer.
+        CoordinatedGraphicsLayer checks for GraphicsSurfaceToken to decide if the surface
+        in UI-Process should be recreated or not. With this
+        patch we also check for size of the platform Layer.
+
+        * platform/graphics/efl/GraphicsContext3DEfl.cpp:
+        (WebCore::GraphicsContext3D::createGraphicsSurfaces):
+        * platform/graphics/efl/GraphicsContext3DPrivate.cpp:
+        (GraphicsContext3DPrivate::didResizeCanvas):
+        (GraphicsContext3DPrivate::platformLayerSize):
+        * platform/graphics/efl/GraphicsContext3DPrivate.h:
+        (GraphicsContext3DPrivate):
+
+2012-12-12  Antoine Quint  <graouts@apple.com>
+
+        Web Inspector: Crash in InspectorDOMAgent::pushNodePathToFrontend when inspecting document with CSS generated content
+        https://bugs.webkit.org/show_bug.cgi?id=104780
+
+        Reviewed by Pavel Feldman.
+
+        The pushNodePathForRenderLayerToFrontend() method added to InspectorDOMAgent naively assumed
+        that all RenderLayers have a Node associated to them. However, in the case of CSS generated
+        content, that will not be the case and we would call into pushNodePathToFrontend() with a 0
+        parameter that would hit the first ASSERT in this method. We're now checking that we indeed
+        have an associated node for the provided RenderLayer and return 0 in case we don't, without
+        calling into pushNodePathToFrontend().
+
+        Test: inspector-protocol/layer-tree-generated-content.html
+
+        * inspector/InspectorDOMAgent.cpp:
+        (WebCore::InspectorDOMAgent::pushNodePathForRenderLayerToFrontend):
+
+2012-12-12  Alexander Pavlov  <apavlov@chromium.org>
+
+        Web Inspector: [Crash] Clear cached stylesheet rules in InspectorStyleSheet::deleteRule()
+        https://bugs.webkit.org/show_bug.cgi?id=104806
+
+        Reviewed by Vsevolod Vlasov.
+
+        Cached rules (m_flatRules) should get cleared whenever a rule is deleted.
+
+        Test: inspector/styles/undo-add-rule-crash.html
+
+        * inspector/InspectorStyleSheet.cpp:
+        (WebCore::InspectorStyleSheet::deleteRule):
+
+2012-12-12  Stephen White  <senorblanco@chromium.org>
+
+        Unreviewed, rolling out r136975.
+        http://trac.webkit.org/changeset/136975
+        https://bugs.webkit.org/show_bug.cgi?id=90405
+
+        Correctness issues.
+
+        * dom/Element.cpp:
+        * dom/Element.h:
+        (Element):
+        * dom/ElementRareData.h:
+        (ElementRareData):
+        (WebCore::ElementRareData::ElementRareData):
+        * platform/graphics/texmap/TextureMapperImageBuffer.cpp:
+        (WebCore::BitmapTextureImageBuffer::applyFilters):
+        * rendering/FilterEffectRenderer.cpp:
+        (WebCore::createCustomFilterEffect):
+        (WebCore::FilterEffectRenderer::buildReferenceFilter):
+        (WebCore::FilterEffectRenderer::build):
+        * rendering/FilterEffectRenderer.h:
+        (FilterEffectRenderer):
+        * rendering/RenderLayer.cpp:
+        (WebCore::RenderLayer::updateOrRemoveFilterEffectRenderer):
+        * rendering/RenderObject.cpp:
+        (WebCore::RenderObject::willBeDestroyed):
+        * rendering/svg/RenderSVGResourceContainer.cpp:
+        (WebCore::RenderSVGResourceContainer::registerResource):
+        * rendering/svg/SVGResources.cpp:
+        (WebCore::SVGResources::buildCachedResources):
+        * rendering/svg/SVGResourcesCache.cpp:
+        (WebCore::SVGResourcesCache::addResourcesFromRenderObject):
+        (WebCore::SVGResourcesCache::clientStyleChanged):
+        (WebCore::SVGResourcesCache::resourceDestroyed):
+        * svg/SVGDocumentExtensions.cpp:
+        (WebCore::SVGDocumentExtensions::addPendingResource):
+        (WebCore::SVGDocumentExtensions::isElementPendingResources):
+        (WebCore::SVGDocumentExtensions::isElementPendingResource):
+        (WebCore::SVGDocumentExtensions::removeElementFromPendingResources):
+        (WebCore::SVGDocumentExtensions::removeElementFromPendingResourcesForRemoval):
+        * svg/SVGDocumentExtensions.h:
+        (WebCore):
+        (SVGDocumentExtensions):
+        * svg/SVGElement.cpp:
+        (WebCore::SVGElement::hasPendingResources):
+        (WebCore):
+        (WebCore::SVGElement::setHasPendingResources):
+        (WebCore::SVGElement::clearHasPendingResourcesIfPossible):
+        * svg/SVGElement.h:
+        (SVGElement):
+        (WebCore::SVGElement::buildPendingResource):
+        * svg/SVGElementRareData.h:
+        (WebCore::SVGElementRareData::SVGElementRareData):
+        (WebCore::SVGElementRareData::hasPendingResources):
+        (WebCore::SVGElementRareData::setHasPendingResources):
+        (SVGElementRareData):
+        * svg/SVGStyledElement.cpp:
+        (WebCore::SVGStyledElement::buildPendingResourcesIfNeeded):
+
+2012-12-12  Stephen White  <senorblanco@chromium.org>
+
+        Unreviewed, rolling out r137022.
+        http://trac.webkit.org/changeset/137022
+
+        Precursor to rollout of 136975.
+
+        * rendering/svg/SVGResources.cpp:
+        (WebCore::SVGResources::buildCachedResources):
+
+2012-12-12  Stephen White  <senorblanco@chromium.org>
+
+        Unreviewed, rolling out r137027.
+        http://trac.webkit.org/changeset/137027
+
+        Precursor to rollout of 136975.
+
+        * dom/ElementRareData.h:
+        (ElementRareData):
+
+2012-12-12  Alexei Filippov  <alph@chromium.org>
+
+        Web Inspector: add memory used by memory allocator itself to NMI
+        https://bugs.webkit.org/show_bug.cgi?id=93372
+
+        Make native memory instrumentation report memory retained
+        for the memory allocator internal purposes.
+
+        Reviewed by Yury Semikhatsky.
+
+        * inspector/front-end/NativeMemorySnapshotView.js:
+        (WebInspector.MemoryBlockViewProperties._initialize):
+        * platform/chromium/MemoryUsageSupportChromium.cpp:
+        (WebCore::MemoryUsageSupport::memoryUsageByComponents):
+
+2012-12-12  Sheriff Bot  <webkit.review.bot@gmail.com>
+
+        Unreviewed, rolling out r137438.
+        http://trac.webkit.org/changeset/137438
+        https://bugs.webkit.org/show_bug.cgi?id=104798
+
+        assertion reached when removing tile (Requested by tmpsantos
+        on #webkit).
+
+        * page/Frame.cpp:
+        (WebCore::Frame::setTiledBackingStoreEnabled):
+        * platform/graphics/TiledBackingStore.cpp:
+        (WebCore::TiledBackingStore::TiledBackingStore):
+        (WebCore::TiledBackingStore::setTileCreationDelay):
+        (WebCore::TiledBackingStore::coverWithTilesIfNeeded):
+        (WebCore::TiledBackingStore::updateTileBuffers):
+        (WebCore::TiledBackingStore::createTiles):
+        (WebCore::TiledBackingStore::isTileBufferUpdatesSuspended):
+        (WebCore::TiledBackingStore::startTileBufferUpdateTimer):
+        (WebCore::TiledBackingStore::tileBufferUpdateTimerFired):
+        (WebCore::TiledBackingStore::startBackingStoreUpdateTimer):
+        (WebCore::TiledBackingStore::backingStoreUpdateTimerFired):
+        * platform/graphics/TiledBackingStore.h:
+        (TiledBackingStore):
+        (WebCore::TiledBackingStore::tileCreationDelay):
+        * platform/graphics/TiledBackingStoreClient.h:
+        (WebCore::TiledBackingStoreClient::tiledBackingStoreUpdatesAllowed):
+
+2012-12-12  Csaba Osztrogonác  <ossy@webkit.org>
+
+        [Qt] Unreviewed typo fix after r137446.
+
+        * WebCore.pri:
+
+2012-12-12  Zan Dobersek  <zandobersek@gmail.com>
+
+        [GTK] Remove the last of unnecessary configuration options in configure.ac
+        https://bugs.webkit.org/show_bug.cgi?id=104793
+
+        Reviewed by Martin Robinson.
+
+        Remove the feature define overriding that's based on the automake variables which are
+        not set anymore due to the corresponding configuration options being removed.
+
+        No new tests - no new functionality.
+
+        * GNUmakefile.am:
+
+2012-12-12  Simon Hausmann  <simon.hausmann@digia.com>
+
+        [Qt] Fix build on Mac
+
+        Reviewed by Csaba Osztrogonác.
+
+        On Mac OS X we have sqlite as a system library available.
+
+        * WebCore.pri:
+
+2012-12-12  Simon Hausmann  <simon.hausmann@digia.com>
+
+        [Qt] Fix linkage against SQLite
+        https://bugs.webkit.org/show_bug.cgi?id=104781
+
+        Reviewed by Csaba Osztrogonác.
+
+        Detect SQLite either through pkg-config, from the SQLITE3SRCDIR
+        environment variable or from qtbase via a qt5.git build. Bail out early
+        with an error message if neither option applies.
+
+        This removes the usage of $$QT.core.sources and the dependency on Qt's
+        system-sqlite configure setting. We should always favour a system
+        library over a copy in Qt.
+
+        * Target.pri:
+        * WebCore.pri:
+
+2012-12-12  Allan Sandfeld Jensen  <allan.jensen@digia.com>
+
+        [TexMap] Uninitialized m_context pointer in TextureMapper
+        https://bugs.webkit.org/show_bug.cgi?id=104785
+
+        Reviewed by Kenneth Rohde Christiansen.
+
+        * platform/graphics/texmap/TextureMapper.cpp:
+        (WebCore::TextureMapper::TextureMapper):
+        * platform/graphics/texmap/TextureMapperGL.cpp:
+        (WebCore::TextureMapperGL::TextureMapperGL):
+
+2012-12-12  Tommy Widenflycht  <tommyw@google.com>
+
+        MediaStream API: Change the data channel descriptor pattern to a handler pattern
+        https://bugs.webkit.org/show_bug.cgi?id=104543
+
+        Reviewed by Kent Tamura.
+
+        In short the code is refactured to use the Handler/Client pattern (see RTCPeerConnectionHandler)
+        instead of the Descriptor pattern.
+        This will fix lifetime/memory/code structure issues in the chromium port.
+
+        Existing tests cover this patch.
+
+        * CMakeLists.txt:
+        * GNUmakefile.list.am:
+        * Modules/mediastream/RTCDataChannel.cpp:
+        (WebCore::RTCDataChannel::create):
+        (WebCore::RTCDataChannel::RTCDataChannel):
+        (WebCore::RTCDataChannel::~RTCDataChannel):
+        (WebCore::RTCDataChannel::label):
+        (WebCore::RTCDataChannel::reliable):
+        (WebCore::RTCDataChannel::readyState):
+        (WebCore::RTCDataChannel::bufferedAmount):
+        (WebCore::RTCDataChannel::send):
+        (WebCore::RTCDataChannel::close):
+        (WebCore::RTCDataChannel::didChangeReadyState):
+        (WebCore::RTCDataChannel::didReceiveStringData):
+        (WebCore::RTCDataChannel::didReceiveRawData):
+        (WebCore::RTCDataChannel::didDetectError):
+        (WebCore::RTCDataChannel::stop):
+        * Modules/mediastream/RTCDataChannel.h:
+        (WebCore):
+        (RTCDataChannel):
+        * Modules/mediastream/RTCPeerConnection.cpp:
+        (WebCore::RTCPeerConnection::didAddRemoteDataChannel):
+        * Modules/mediastream/RTCPeerConnection.h:
+        (RTCPeerConnection):
+        * WebCore.gypi:
+        * platform/chromium/support/WebRTCDataChannel.cpp: Removed.
+        * platform/mediastream/RTCDataChannelDescriptor.h: Removed.
+        * platform/mediastream/RTCDataChannelHandler.h: Copied from Source/WebCore/platform/mediastream/RTCDataChannelDescriptor.cpp.
+        (WebCore):
+        (RTCDataChannelHandler):
+        (WebCore::RTCDataChannelHandler::~RTCDataChannelHandler):
+        * platform/mediastream/RTCDataChannelHandlerClient.h: Renamed from Source/WebCore/platform/mediastream/RTCDataChannelDescriptor.cpp.
+        (WebCore):
+        (RTCDataChannelHandlerClient):
+        (WebCore::RTCDataChannelHandlerClient::~RTCDataChannelHandlerClient):
+        * platform/mediastream/RTCPeerConnectionHandler.cpp:
+        (WebCore):
+        (WebCore::RTCPeerConnectionHandler::create):
+        * platform/mediastream/RTCPeerConnectionHandler.h:
+        (WebCore):
+        (RTCPeerConnectionHandler):
+        * platform/mediastream/RTCPeerConnectionHandlerClient.h:
+        (WebCore):
+        (RTCPeerConnectionHandlerClient):
+        * platform/mediastream/chromium/RTCDataChannelHandlerChromium.cpp: Added.
+        (WebCore):
+        (WebCore::RTCDataChannelHandlerChromium::create):
+        (WebCore::RTCDataChannelHandlerChromium::RTCDataChannelHandlerChromium):
+        (WebCore::RTCDataChannelHandlerChromium::~RTCDataChannelHandlerChromium):
+        (WebCore::RTCDataChannelHandlerChromium::setClient):
+        (WebCore::RTCDataChannelHandlerChromium::label):
+        (WebCore::RTCDataChannelHandlerChromium::isReliable):
+        (WebCore::RTCDataChannelHandlerChromium::bufferedAmount):
+        (WebCore::RTCDataChannelHandlerChromium::sendStringData):
+        (WebCore::RTCDataChannelHandlerChromium::sendRawData):
+        (WebCore::RTCDataChannelHandlerChromium::close):
+        (WebCore::RTCDataChannelHandlerChromium::didChangeReadyState):
+        (WebCore::RTCDataChannelHandlerChromium::didReceiveStringData):
+        (WebCore::RTCDataChannelHandlerChromium::didReceiveRawData):
+        (WebCore::RTCDataChannelHandlerChromium::didDetectError):
+        * platform/mediastream/chromium/RTCDataChannelHandlerChromium.h: Added.
+        (WebCore):
+        (RTCDataChannelHandlerChromium):
+        * platform/mediastream/chromium/RTCPeerConnectionHandlerChromium.cpp:
+        (WebCore::RTCPeerConnectionHandlerChromium::createDataChannel):
+        (WebCore::RTCPeerConnectionHandlerChromium::didAddRemoteDataChannel):
+        * platform/mediastream/chromium/RTCPeerConnectionHandlerChromium.h:
+        (WebCore):
+        (RTCPeerConnectionHandlerChromium):
+
+2012-12-12  Huang Dongsung  <luxtella@company100.net>
+
+        Coordinated Graphics: Refactor TiledBackingStore code in CoordinatedGraphicsLayer.
+        https://bugs.webkit.org/show_bug.cgi?id=103959
+
+        Reviewed by Kenneth Rohde Christiansen.
+
+        TiledBackingStore::setCommitTileUpdatesOnIdleEventLoop() is used when class
+        methods cannot be called asynchronously by client. Updates of tiles are
+        committed as soon as all the events in event queue have been processed.
+        After this patch Frame sets m_commitTileUpdatesOnIdleEventLoop to true.
+
+        In addition, remove TiledBackingStoreClient::tiledBackingStoreUpdatesAllowed()
+        which was introduced for Coordinated Graphics.
+
+        Refactoring covered by existing tests.
+
+        * page/Frame.cpp:
+        (WebCore::Frame::setTiledBackingStoreEnabled):
+        * platform/graphics/TiledBackingStore.cpp:
+        (WebCore::TiledBackingStore::TiledBackingStore):
+        (WebCore::TiledBackingStore::setTrajectoryVector):
+          Separate setting a trajectory vector from coverWithTilesIfNeeded().
+        (WebCore::TiledBackingStore::coverWithTilesIfNeeded):
+        (WebCore::TiledBackingStore::updateTileBuffers):
+        (WebCore::TiledBackingStore::createTiles):
+        (WebCore::TiledBackingStore::isTileBufferUpdatesSuspended):
+        (WebCore::TiledBackingStore::startTileBufferUpdateTimer):
+        (WebCore::TiledBackingStore::tileBufferUpdateTimerFired):
+        (WebCore::TiledBackingStore::startBackingStoreUpdateTimer):
+        (WebCore::TiledBackingStore::backingStoreUpdateTimerFired):
+        * platform/graphics/TiledBackingStore.h:
+        (TiledBackingStore):
+        (WebCore::TiledBackingStore::setCommitTileUpdatesOnIdleEventLoop):
+        * platform/graphics/TiledBackingStoreClient.h:
+        (WebCore::TiledBackingStoreClient::tiledBackingStoreHasPendingTileCreation):
+            If TiledBackingStore does not create all tiles, TiledBackingStore
+            notifies a client of needing to create tiles more, when
+            m_commitTileUpdatesOnIdleEventLoop is false.
+
 2012-12-11  Jocelyn Turcotte  <jocelyn.turcotte@digia.com>
 
         [Qt] Fix the inspector not showing up on Windows