GitLab

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66852 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Unreviewed. Build fix.

        Update test expectations for Leopard.

        * platform/mac-leopard/svg/custom/use-invalid-pattern-expected.txt: Added.


git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66851 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Reviewed by Sam Weinig.

        SerializedScriptValue needs to use a flat storage mechanism
        https://bugs.webkit.org/show_bug.cgi?id=45244

        Export JSArray::put

        * JavaScriptCore.exp:
2010-09-05  Oliver Hunt  <oliver@apple.com>

        Reviewed by Sam Weinig.

        SerializedScriptValue needs to use a flat storage mechanism
        https://bugs.webkit.org/show_bug.cgi?id=45244

        Add a few more test cases to cover the new branches in the
        rewritten serialization logic.

        * fast/dom/Window/window-postmessage-clone-expected.txt:
        * fast/dom/Window/window-postmessage-clone.html:
2010-09-05  Oliver Hunt  <oliver@apple.com>

        Reviewed by Sam Weinig.

        SerializedScriptValue needs to use a flat storage mechanism
        https://bugs.webkit.org/show_bug.cgi?id=45244

        Rewrite the old tree to tree serialization logic to use
        flat storage.  Unfortunately this basically required a
        complete rewrite.

        * bindings/js/SerializedScriptValue.cpp:
        (WebCore::CloneBase::CloneBase):
        (WebCore::CloneBase::shouldTerminate):
        (WebCore::CloneBase::ticksUntilNextCheck):
        (WebCore::CloneBase::didTimeOut):
        (WebCore::CloneBase::throwStackOverflow):
        (WebCore::CloneBase::throwInterruptedException):
        (WebCore::CloneBase::fail):
        (WebCore::CloneSerializer::serialize):
        (WebCore::CloneSerializer::CloneSerializer):
        (WebCore::CloneSerializer::isArray):
        (WebCore::CloneSerializer::startObject):
        (WebCore::CloneSerializer::startArray):
        (WebCore::CloneSerializer::endObject):
        (WebCore::CloneSerializer::getSparseIndex):
        (WebCore::CloneSerializer::getProperty):
        (WebCore::CloneSerializer::dumpImmediate):
        (WebCore::CloneSerializer::dumpString):
        (WebCore::CloneSerializer::dumpIfTerminal):
        (WebCore::CloneSerializer::write):
        (WebCore::CloneSerializer::writeLittleEndian):
        (WebCore::CloneSerializer::writeStringIndex):
        (WebCore::CloneDeserializer::deserializeString):
        (WebCore::CloneDeserializer::deserialize):
        (WebCore::CloneDeserializer::CloneDeserializer):
        (WebCore::CloneDeserializer::throwValidationError):
        (WebCore::CloneDeserializer::isValid):
        (WebCore::CloneDeserializer::readLittleEndian):
        (WebCore::CloneDeserializer::read):
        (WebCore::CloneDeserializer::readStringIndex):
        (WebCore::CloneDeserializer::readString):
        (WebCore::CloneDeserializer::readStringData):
        (WebCore::CloneDeserializer::readTag):
        (WebCore::CloneDeserializer::putProperty):
        (WebCore::CloneDeserializer::readFile):
        (WebCore::CloneDeserializer::readTerminal):
        (WebCore::SerializedScriptValue::~SerializedScriptValue):
        (WebCore::SerializedScriptValue::SerializedScriptValue):
        (WebCore::SerializedScriptValue::create):
        (WebCore::SerializedScriptValue::toString):
        (WebCore::SerializedScriptValue::deserialize):
        (WebCore::SerializedScriptValue::nullValue):
        * bindings/js/SerializedScriptValue.h:
        * dom/MessagePortChannel.cpp:
        (WebCore::MessagePortChannel::EventData::EventData):
        * workers/WorkerMessagingProxy.cpp:
        (WebCore::MessageWorkerContextTask::MessageWorkerContextTask):
        (WebCore::MessageWorkerTask::MessageWorkerTask):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66850 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Reviewed by Eric Seidel.

        [reviewtool] Add a field for overall comments
        https://bugs.webkit.org/show_bug.cgi?id=45273

        This patch does a couple logically separate things that could be
        separated into smaller patches:

        1) This patch adds an "overall comments" field where you can enter
           overall comments about the patch.  These comments appear at the top
           of the bugzilla posting.  Currently, these aren't redisplayed when
           viewing the patch, but I plan to add that in a future patch.

        2) This patch renames some of the CSS classes to more consistently
           follow the camelCase style that PrettyPatch uses.

        3) This patch moves the "prepare comments" button to the left of the
           toolbar and renames is to "publish comments".  This makes more sense
           when you scroll to the bottom of the page and enter in some overall
           comments.

        4) When you attempt to add a comment to a line that already has a
           "frozen" comment, we now unfreeze the comment instead of doing
           nothing.  The old behavior was kind of frustrating if you didn't
           know that you could unfreeze a comment by clicking on it.

        * PrettyPatch/PrettyPatch.rb:
            - Update CSS.
        * code-review.js:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66849 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Reviewed by Gustavo Noronha Silva.

        [GTK] ScrollbarThemeGtk should be enabled for interior frame scrollbars
        https://bugs.webkit.org/show_bug.cgi?id=45046

        Make interior frame scrollbars "fully-fake," which means they do not rely
        at all on GTK+ for painting or behavior, but instead are typical WebCore
        scrollbars drawn according to the GTK+ theme. Move ScrollbarGtk to
        MainFrameScrollbarGtk and remove all logic in that file dealing with interior
        frame scrollbars.

        No new tests, as this is already covered by pixel tests for scrollbars,
        which were added in r66605.

        * GNUmakefile.am: Update sources list.
        * platform/ScrollView.h: Change the adjustment members to be smart pointers.
        * platform/Scrollbar.cpp: Enable THUMB_POSITION_AFFECTS_BUTTONS for GTK+ as well.
        * platform/gtk/MainFrameScrollbarGtk.cpp: Added.
        (MainFrameScrollbarGtk::create): Create a MainFrameScrollbarGtk instead of ScrollbarGtk.
        (MainFrameScrollbarGtk::MainFrameScrollbarGtk):
        (MainFrameScrollbarGtk::~MainFrameScrollbarGtk):
        (MainFrameScrollbarGtk::attachAdjustment):
        (MainFrameScrollbarGtk::detachAdjustment):
        (MainFrameScrollbarGtk::updateThumbPosition):
        (MainFrameScrollbarGtk::updateThumbProportion):
        (MainFrameScrollbarGtk::gtkValueChanged):
        (MainFrameScrollbarGtk::paint):
        * platform/gtk/MainFrameScrollbarGtk.h: Added.
        * platform/gtk/ScrollViewGtk.cpp:
        (WebCore::ScrollView::platformInit): Remove unnecessary initialization
        because of smart pointer change.
        (WebCore::ScrollView::createScrollbar): Instead of creating ScrollbarGtk for
        interior frame scrollbars, create a normal WebCore scrollbar.
        (WebCore::ScrollView::setGtkAdjustments): Added an extra assert which ensures
        that this method is never called with a non-null adjustment on an interior frame
        scrollbar.
        (WebCore::ScrollView::platformAddChild): Remove logic for interior frame scrollbars.
        (WebCore::ScrollView::platformRemoveChild): Remove logic for interior frame scrollbars.
        (WebCore::ScrollView::visibleContentRect): Use a more accurate guard for detecting
        transitionary states when accessing parent widgets. Explcitly guard against interior
        frame ScrollView's trying to determine size based on parent widgets.
        (WebCore::ScrollView::setScrollbarModes): This method was out of sync with the one
        it copy-and-pasted from. Update it and change the logic to do the right thing for
        interior frame scrollbars.
        * platform/gtk/ScrollbarGtk.cpp: Removed.
        * platform/gtk/ScrollbarGtk.h: Removed.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66848 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Reviewed by Nikolas Zimmermann.

        Make SVG PendingResources use RefPtr
        https://bugs.webkit.org/show_bug.cgi?id=43587

        Convert SVGDocumentExtensions::m_pendingResources to use a RefPtr for 
        pending elements instead of a raw pointer so that pending elements can't
        be freed prematurely.

        Test: svg/custom/use-invalid-pattern.svg

        * rendering/RenderSVGResourceContainer.cpp:
        (WebCore::RenderSVGResourceContainer::registerResource):
        * svg/SVGDocumentExtensions.cpp:
        (WebCore::SVGDocumentExtensions::addPendingResource):
        (WebCore::SVGDocumentExtensions::removePendingResource):
        * svg/SVGDocumentExtensions.h:
        * svg/SVGElement.cpp:
        (WebCore::SVGElement::insertedIntoDocument):
2010-09-06  Justin Schuh  <jschuh@chromium.org>

        Reviewed by Nikolas Zimmermann.

        Test that an invalid pending resource pattern doesn't crash
        https://bugs.webkit.org/show_bug.cgi?id=43587

        * platform/mac-snowleopard/svg: Added.
        * platform/mac-snowleopard/svg/custom: Added.
        * platform/mac-snowleopard/svg/custom/use-invalid-pattern-expected.txt: Added.
        * svg/custom/use-invalid-pattern.svg: Added.


git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66847 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Reviewed by Oliver Hunt.

        Support JSVALUE32_64 on MIPS
        https://bugs.webkit.org/show_bug.cgi?id=43999

        Add missing functions to support JSVALUE32_64 on MIPS.
        Remove JSVALUE32 as the default for MIPS.

        * assembler/MIPSAssembler.h:
        (JSC::MIPSAssembler::divd):
        (JSC::MIPSAssembler::mthc1):
        (JSC::MIPSAssembler::cvtwd):
        * assembler/MacroAssemblerMIPS.h:
        (JSC::MacroAssemblerMIPS::neg32):
        (JSC::MacroAssemblerMIPS::branchOr32):
        (JSC::MacroAssemblerMIPS::set8):
        (JSC::MacroAssemblerMIPS::loadDouble):
        (JSC::MacroAssemblerMIPS::divDouble):
        (JSC::MacroAssemblerMIPS::convertInt32ToDouble):
        (JSC::MacroAssemblerMIPS::branchDouble):
        (JSC::MacroAssemblerMIPS::branchConvertDoubleToInt32):
        (JSC::MacroAssemblerMIPS::zeroDouble):
        * jit/JIT.h:
        * jit/JITOpcodes32_64.cpp:
        (JSC::JIT::privateCompileCTINativeCall):
        * jit/JITPropertyAccess32_64.cpp:
        (JSC::JIT::privateCompilePutByIdTransition):
        * jit/JITStubs.cpp:
        (JSC::JITThunks::JITThunks):
        * jit/JITStubs.h:
        * wtf/Platform.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66846 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Reviewed by Gustavo Noronha Silva.

        [GTK] ScrollbarThemeGtk should support secondary steppers
        https://bugs.webkit.org/show_bug.cgi?id=44791

        Add support to ScrollbarThemeGtk for drawing alternate steppers. Adjust
        the algorithms for calculating forward and back button rects, if they
        are active in the theme. Expose this information via GtkScrollbarMetrics.

        * platform/gtk/ScrollbarThemeGtk.cpp:
        (WebCore::ScrollbarThemeGtk::updateThemeProperties): Access the secondary stepper properties
        of GtkScrollbarMetrics when updating the style cache.
        (WebCore::ScrollbarThemeGtk::backButtonRect): Account for alternate steppers.
        (WebCore::ScrollbarThemeGtk::forwardButtonRect): Ditto.
        (WebCore::ScrollbarThemeGtk::trackRect): Ditto.
        (WebCore::ScrollbarThemeGtk::paintButton): Ditto.
        * platform/gtk/ScrollbarThemeGtk.h:
        * platform/gtk/gtk2drawing.c: Expose whether or not the style uses alternate steppers
        via GtkScrollbarMetrics.
        * platform/gtk/gtkdrawing.h: Added fields to GtkScrollbarMetrics.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66845 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Revert to original patch in bugzilla. Leave bug open for
discussion on potential removal of double utf8 conversion.

https://bugs.webkit.org/show_bug.cgi?id=45240

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66844 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Reviewed by Andreas Kling.

        [Qt] utf8 encoding of console() messages

        Unskip:
        http/tests/security/xssAuditor/embed-tag-null-char.html
        http/tests/security/xssAuditor/object-embed-tag-null-char.html

        Both tests failed because Qt's implementation of String::format()
        is casting a utf8 result to String, which assumes latin1 in
        its constructor. So instead of casting a QString to a String, use
        StringImpl::create() instead. Unfortunately, this involves a lot
        of extra casts but the end result is correct.

        https://bugs.webkit.org/show_bug.cgi?id=45240

        * wtf/text/WTFString.cpp:
        (WTF::String::format):
2010-09-06  Robert Hogan  <robert@webkit.org>

        Reviewed by Andreas Kling.

        [Qt] utf8 encoding of console() messages

        Unskip:
        http/tests/security/xssAuditor/embed-tag-null-char.html
        http/tests/security/xssAuditor/object-embed-tag-null-char.html

        Both tests failed because Qt's implementation of String::format()
        is casting a utf8 result to String, which assumes latin1 in
        its constructor. So instead of casting a QString to a String, use
        StringImpl::create() instead. Unfortunately, this involves a lot
        of extra casts but the end result is correct.

        https://bugs.webkit.org/show_bug.cgi?id=45240

        * platform/qt/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66843 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Unreviewed. Build fix.

        Update test expectations for Chromium Mac.

        * platform/chromium-mac/svg/custom/use-invalid-style-expected.txt: Added.


git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66842 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Reviewed by Adam Barth.

        Implement HTML5 definition of document.readyState
        https://bugs.webkit.org/show_bug.cgi?id=45119

        * fast/dom/Document/readystate-expected.txt: Added.
        * fast/dom/Document/readystate.html: Added. Reads readyState inline script, external script, deferred script, DOMContentLoaded, onload, and dynamic post-onload script.
2010-09-06  Tony Gentilcore  <tonyg@chromium.org>

        Reviewed by Adam Barth.

        Implement HTML5 definition of document.readyState
        https://bugs.webkit.org/show_bug.cgi?id=45119

        The legacy behavior was "loading" -> "loaded" -> "complete". The new
        HTML5 behavior is "loading" -> "interactive" -> "complete". There is
        some potential for this to cause compat problems if for instance a
        page expects readyState to be "loaded" during the DOMContentLoaded event.

        Test: fast/dom/Document/readystate.html

        * dom/Document.cpp:
        (WebCore::Document::Document): Initial value is Complete because according to http://www.whatwg.org/specs/web-apps/current-work/#dom-document-readystate,
        when a Document is created the initial value is "complete" unless it has a parser associated with it, in which case it is "loading".
        So the ctor starts it Complete, and when the parser is created it is flipped to Loading.
        (WebCore::Document::readyState):
        (WebCore::Document::setReadyState):
        (WebCore::Document::implicitOpen):
        (WebCore::Document::finishedParsing): Ensure that XML and HTML parser have transition to Stopping state.
        * dom/Document.h:
        * dom/DocumentParser.cpp:
        (WebCore::DocumentParser::prepareToStopParsing): Previously this was being called when parsing had stopped.
        It is better to ensure it is only called while parsing.
        * dom/XMLDocumentParser.cpp:
        (WebCore::XMLDocumentParser::end): Transition to stopping before calling document finishedParsiong().
        * html/parser/HTMLDocumentParser.cpp:
        (WebCore::HTMLDocumentParser::prepareToStopParsing): Set state to interactive before running deferred scripts.
        This method is also called when parsing fragments, so we need to ensure it isn't done in that case.
        (WebCore::HTMLDocumentParser::attemptToRunDeferredScriptsAndEnd): Added. Break out this part s that notifyFinished doesn't go through
        the additional steps of pumping tokenizer, setting the state, etc.
        (WebCore::HTMLDocumentParser::notifyFinished): Now that prepareToStopParsing is split up, we must protect. It also makes sense to add a couple of ASSERTs.
        * loader/FrameLoader.cpp:
        (WebCore::FrameLoader::stopLoading): It looks like an aborted load should never transition to "complete" according the HTML5. I've left the legacy behavior for now though.
        (WebCore::FrameLoader::checkCompleted): The FrameLoader now sets the state on the Document instead of the Document polling the FrameLoader.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66841 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Reviewed by Adam Barth.

        [v8] Inline hot methods for V8 to WebCore and back conversions
        https://bugs.webkit.org/show_bug.cgi?id=45270

        Inline fast paths of hot functions performing conversions from V8 wrappers
        to WebCore native objects and back.
        That slightly increases the size of binary (within 0.1% for both Ubuntu
        and Windows, but those builds are slightly different from official ones),
        but gives performance boost (3--5% on Windows, up to 8% on Ubuntu).

        * bindings/scripts/CodeGeneratorV8.pm:
        * bindings/v8/V8DOMWindowShell.cpp:
        (WebCore::V8DOMWindowShell::initContextIfNeeded):
        * bindings/v8/V8DOMWindowShell.h:
        * bindings/v8/V8DOMWrapper.cpp:
        (WebCore::V8DOMWrapper::getWrapperSlow):
        * bindings/v8/V8DOMWrapper.h:
        (WebCore::V8DOMWrapper::getWrapper):
        * bindings/v8/custom/V8NodeCustom.cpp:
        (WebCore::toV8Slow):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66840 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Unreviewed build fix.

        [EFL] REGRESSION(66794) Need to fix build break.
        https://bugs.webkit.org/show_bug.cgi?id=45241

        Add dummy class(FrameNetworkingContextEfl) like GTK+ port and implement
        as EFL style.

        * WebCoreSupport/FrameLoaderClientEfl.cpp:
        (WebCore::FrameLoaderClientEfl::createNetworkingContext):
        * WebCoreSupport/FrameNetworkingContextEfl.h: Added.
        (WebCore::FrameNetworkingContextEfl::create):
        (WebCore::FrameNetworkingContextEfl::coreFrame):
        (WebCore::FrameNetworkingContextEfl::FrameNetworkingContextEfl):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66839 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        [reviewtool] Tweak the ok button to cancel the comment if the comment
        is empty.  Previously we would get into a bad state where a line had a
        comment but there was no longer any way to access it.

        * code-review.js:


git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66838 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Reviewed by Jeremy Orlow.

        Add new interface and empty impl. as precursor to bug 45112
        https://bugs.webkit.org/show_bug.cgi?id=45257

        * public/WebGeolocationServiceBridge.h:
        (WebKit::WebGeolocationServiceBridge::~WebGeolocationServiceBridge):
        * src/WebGeolocationServiceBridgeImpl.cpp:
        (WebKit::WebGeolocationServiceBridgeImpl::onWebGeolocationServiceDestroyed):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66837 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Reviewed by Dimitri Glazkov.

        [Crash] <animateMotion> element directly inside <symbol> element causes crash when referenced by <use>
        https://bugs.webkit.org/show_bug.cgi?id=44750

        Added test case to demonstrate crash bug.

        * svg/dom/symbol-embeddedAnimation-expected.txt: Added.
        * svg/dom/symbol-embeddedAnimation.svg: Added.
2010-09-06  Shane Stephens  <shanestephens@google.com>

        Reviewed by Dimitri Glazkov.

        [Crash] <animateMotion> element directly inside <symbol> element causes crash when referenced by <use>
        https://bugs.webkit.org/show_bug.cgi?id=44750

        Fixes crash by checking for null transforms and skipping update step
        when appropriate.

        Test: svg/dom/symbol-embeddedAnimation.svg

        * svg/SVGAnimateMotionElement.cpp:
        (WebCore::SVGAnimateMotionElement::applyResultsToTarget):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66836 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Reviewed by Antonio Gomes.

        Web Inspector: Tests crash on Qt bots revealed by r66720
        https://bugs.webkit.org/show_bug.cgi?id=45256

        * WebCoreSupport/InspectorClientQt.cpp:
        (WebCore::InspectorClientQt::InspectorClientQt): m_frontendClient must be initialized by constructor


git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66835 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Reviewed by Xan Lopez.

        [GTK] Small code cleanup in DumpRenderTreeGtk.cpp
        https://bugs.webkit.org/show_bug.cgi?id=45213

        * DumpRenderTree/gtk/DumpRenderTree.cpp:
        (initializeFonts): Made this function do nothing for non-X11 platforms, so we don't
        have to surround the invocation with #ifdefs.
        (useLongRunningServerMode): Added.
        (runTestingServerLoop): Added.
        (initializeGlobalsFromCommandLineOptions): Added.
        (runTest): Removed ifdefs.
        (main): Use new helper functions.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66834 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Unskip a couple of plugin tests which are now passing.

        * platform/gtk/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66833 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Unreviewed build fix.

        * inspector/report-API-errors.html:




git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66832 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Rubber-stamped by Tor Arne Vestbø.

        WebCore already generates a focus-{out,in} events on its setFocus
        method, there's no need to duplicate it here. Previously this was
        needed because we were not catching all possible names of focus
        events in PluginView::handleEvent and were missing the ones
        WebCore does, but this has been fixed in r66827.

        * plugins/qt/PluginViewQt.cpp:
        (WebCore::PluginView::setFocus):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66831 268f45cc-cd09-0410-ab3c-d52691b4dbfc

http://trac.webkit.org/changeset/66823
https://bugs.webkit.org/show_bug.cgi?id=45266

Patch by Sheriff Bot <webkit.review.bot@gmail.com> on 2010-09-06
Windows build fails with unknown reason (Requested by zherczeg
on #webkit).

WebCore: 

* Android.mk:
* CMakeLists.txt:
* GNUmakefile.am:
* WebCore.gypi:
* WebCore.pro:
* WebCore.vcproj/WebCore.vcproj:
* WebCore.xcodeproj/project.pbxproj:
* rendering/RenderObject.h:
* rendering/RenderSVGResourceFilterPrimitive.cpp: Removed.
* rendering/RenderSVGResourceFilterPrimitive.h: Removed.
* rendering/SVGRenderTreeAsText.cpp:
(WebCore::writeSVGContainer):
* svg/SVGFEDiffuseLightingElement.cpp:
(WebCore::SVGFEDiffuseLightingElement::svgAttributeChanged):
* svg/SVGFELightElement.cpp:
(WebCore::SVGFELightElement::svgAttributeChanged):
(WebCore::SVGFELightElement::childrenChanged):
* svg/SVGFEOffsetElement.cpp:
(WebCore::SVGFEOffsetElement::svgAttributeChanged):
* svg/SVGFilterElement.h:
(WebCore::SVGFilterElement::invalidateFilter):
* svg/SVGFilterPrimitiveStandardAttributes.cpp:
(WebCore::SVGFilterPrimitiveStandardAttributes::svgAttributeChanged):
(WebCore::SVGFilterPrimitiveStandardAttributes::childrenChanged):
* svg/SVGFilterPrimitiveStandardAttributes.h:
(WebCore::SVGFilterPrimitiveStandardAttributes::rendererIsNeeded):

LayoutTests: 

* platform/mac/svg/dynamic-updates/SVGFEDiffuseLightingElement-lighting-color-css-prop-expected.checksum:
* platform/mac/svg/dynamic-updates/SVGFEDiffuseLightingElement-lighting-color-css-prop-expected.png:



git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66830 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Reviewed by Xan Lopez.

        [GTK] EventSender should support modifier keys with mouseDown and mouseUp events
        https://bugs.webkit.org/show_bug.cgi?id=45235

        * platform/gtk/Skipped: Unskip a test which is now passing.
2010-09-06  Martin Robinson  <mrobinson@igalia.com>

        Reviewed by Xan Lopez.

        [GTK] EventSender should support modifier keys with mouseDown and mouseUp events
        https://bugs.webkit.org/show_bug.cgi?id=45235

        Add support for interpreting the modifier key arguments to the mouseDown and mouseUp
        methods of the EventSender.

        * DumpRenderTree/gtk/EventSender.cpp:
        (prepareMouseButtonEvent): Allow passing in a modifier bitmask, which will be OR'd
        with the current modifiers.
        (contextClickCallback): Always send no modifiers when preparing the mouse event.
        (gdkModifersFromJSValue): Added, converts a JSValue array into a GDK modifier bitmask.
        (mouseDownCallback): Send in the requested modifiers to prepareMouseButtonEvent.
        (mouseUpCallback): Ditto.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66829 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Reviewed by Antonio Gomes.

        Convert editing/selection/shift-click.html to use setEditingBehavior
        https://bugs.webkit.org/show_bug.cgi?id=45233

        * editing/selection/script-tests/shift-click.js:
        (shiftClick): Take only one expected string.
        (assertSelectionString): Take only one expected string.
        (runShiftClickTest): Added. Runs the test with a particular editing behavior.
        * editing/selection/shift-click-expected.txt: Updated to reflect new results.
        * platform/win/editing/selection/shift-click-expected.txt: Removed.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66828 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-09-06  Xan Lopez  <xlopez@igalia.com>

        Reviewed by Martin Robinson.

        [GTK] Provide Keyboard Events to Windowless plugins
        https://bugs.webkit.org/show_bug.cgi?id=44613

        Fix keyboard event delivery for windowless plugins.

        Based on a patch by Bharathwaaj.

        * plugins/PluginView.cpp:
        (WebCore::PluginView::handleEvent): also take into account
        focused{in,out} events
        * plugins/gtk/PluginViewGtk.cpp:
        (WebCore::PluginView::handleKeyboardEvent): we want the keyval
        here, not the hardware keycode
        (WebCore::PluginView::handleMouseEvent): focus the pluginview on mouse events

LayoutTests:

2010-09-06  Xan Lopez  <xlopez@igalia.com>

        Reviewed by Martin Robinson.

        [GTK] Provide Keyboard Events to Windowless plugins
        https://bugs.webkit.org/show_bug.cgi?id=44613

        Unskip passing test.

        * platform/gtk/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66827 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Reviewed by Yury Semikhatsky.

        WebInspector: it'd be nice to be able to pass undefined argument as a callback to InspectorBackend functions.

        There are some places where callback is passing via some wrapper. In that case sendMessageToBackend
        will be called with additional argument for callback but with undefined value.

        https://bugs.webkit.org/show_bug.cgi?id=45265

        * inspector/CodeGeneratorInspector.pm:




git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66826 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Reviewed by Yury Semikhatsky.

        Web Inspector: fix chromium devtools tests
        https://bugs.webkit.org/show_bug.cgi?id=45258

        * src/js/Tests.js:
        (.TestSuite.prototype._waitForScriptPause):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66825 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Unreviewed. Attempt to fix random test crashes after r66720.

        * WebCoreSupport/InspectorClientQt.cpp:
        (WebCore::InspectorFrontendClientQt::~InspectorFrontendClientQt):
        * WebCoreSupport/InspectorClientQt.h:


git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66824 268f45cc-cd09-0410-ab3c-d52691b4dbfc

https://bugs.webkit.org/show_bug.cgi?id=43954

Reviewed by Dirk Schulze.

WebCore: 

RenderSVGResourceFilterPrimitive renderer is added to
the project, and assigned to each object, which class is
derived from SVGFilterPrimitiveStandardAttributes. The patch
mainly contains build system changes, and it fixes one layout
test in svg/dynamic-updates.

* Android.mk:
* CMakeLists.txt:
* GNUmakefile.am:
* WebCore.gypi:
* WebCore.pro:
* WebCore.vcproj/WebCore.vcproj:
* WebCore.xcodeproj/project.pbxproj:
* rendering/RenderObject.h:
(WebCore::RenderObject::isSVGResourceFilterPrimitive):
* rendering/RenderSVGResourceFilterPrimitive.cpp: Added.
(WebCore::RenderSVGResourceFilterPrimitive::RenderSVGResourceFilterPrimitive):
* rendering/RenderSVGResourceFilterPrimitive.h: Added.
(WebCore::RenderSVGResourceFilterPrimitive::isSVGResourceFilterPrimitive):
* rendering/SVGRenderTreeAsText.cpp:
(WebCore::writeSVGContainer):
* svg/SVGFEDiffuseLightingElement.cpp:
(WebCore::SVGFEDiffuseLightingElement::svgAttributeChanged):
* svg/SVGFELightElement.cpp:
(WebCore::SVGFELightElement::svgAttributeChanged):
(WebCore::SVGFELightElement::childrenChanged):
* svg/SVGFEOffsetElement.cpp:
(WebCore::SVGFEOffsetElement::svgAttributeChanged):
* svg/SVGFilterElement.h:
* svg/SVGFilterPrimitiveStandardAttributes.cpp:
(WebCore::SVGFilterPrimitiveStandardAttributes::svgAttributeChanged):
(WebCore::SVGFilterPrimitiveStandardAttributes::childrenChanged):
(WebCore::SVGFilterPrimitiveStandardAttributes::createRenderer):
* svg/SVGFilterPrimitiveStandardAttributes.h:
(WebCore::SVGFilterPrimitiveStandardAttributes::invalidate):

LayoutTests: 

The layout test below is fixed by this patch.

* platform/mac/svg/dynamic-updates/SVGFEDiffuseLightingElement-lighting-color-css-prop-expected.checksum:
* platform/mac/svg/dynamic-updates/SVGFEDiffuseLightingElement-lighting-color-css-prop-expected.png:



git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66823 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Reviewed by Dirk Schulze.

        [WML] Use RenderImageResource in WMLImageElement.cpp
        https://bugs.webkit.org/show_bug.cgi?id=44952

        The hasImage() and setCachedImage() were moved to RenderImageResource class.
        So, WML also should use the functions from RenderImageResource.

        * wml/WMLImageElement.cpp:
        (WebCore::WMLImageElement::attach):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66822 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Reviewed by Eric Seidel.

        [reviewtool] Add an "ok" button that collapses review comments
        https://bugs.webkit.org/show_bug.cgi?id=45255

        Many folks (including myself) are too trigger-happy with the "delete"
        button that we end up deleting perfectly good review comments.  This
        patch adds an "ok" button to absorb these clicks (and renames "delete"
        to "cancel").  When you click the ok button, the comment becomes
        non-editable and the box shrinks to fit the text.  You can click the
        text to get back to the editable version.

        * code-review.js:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66821 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Reviewed by Steve Block.

        Add index insertion support to IndexedDB.
        https://bugs.webkit.org/show_bug.cgi?id=44695

        * storage/indexeddb/objectstore-basics-expected.txt:
        * storage/indexeddb/script-tests/objectstore-basics.js:
        (createSuccess):
        (addIndexSuccess):
        (getSuccess):
2010-08-26  Jeremy Orlow  <jorlow@chromium.org>

        Reviewed by Steve Block.

        Add index insertion support to IndexedDB.
        https://bugs.webkit.org/show_bug.cgi?id=44695

        Whenever you insert an item into an ObjectStore, it should use all
        indexes' key paths to insert corresponding entries into each index.
        Also data should be deleted out of the index when it goes away.

        Not much testing yet since there's no way to directly observe indexes.
        More will be in next patch.

        * storage/IDBDatabaseBackendImpl.cpp:
        (WebCore::IDBDatabaseBackendImpl::removeObjectStore):
        * storage/IDBFactoryBackendImpl.cpp:
        (WebCore::createTables):
        * storage/IDBIndex.idl:
        * storage/IDBIndexBackendImpl.cpp:
        (WebCore::whereClause):
        (WebCore::bindWhereClause):
        (WebCore::IDBIndexBackendImpl::addingKeyAllowed):
        * storage/IDBIndexBackendImpl.h:
        (WebCore::IDBIndexBackendImpl::id):
        * storage/IDBKey.cpp:
        (WebCore::IDBKey::whereSyntax):
        (WebCore::IDBKey::bind):
        (WebCore::IDBKey::bindWithNulls):
        * storage/IDBKey.h:
        * storage/IDBObjectStore.idl:
        * storage/IDBObjectStoreBackendImpl.cpp:
        (WebCore::whereClause):
        (WebCore::bindWhereClause):
        (WebCore::IDBObjectStoreBackendImpl::get):
        (WebCore::fetchKeyFromKeyPath):
        (WebCore::putObjectStoreData):
        (WebCore::putIndexData):
        (WebCore::IDBObjectStoreBackendImpl::put):
        (WebCore::IDBObjectStoreBackendImpl::remove):
        (WebCore::IDBObjectStoreBackendImpl::createIndex):
        (WebCore::doDelete):
        (WebCore::IDBObjectStoreBackendImpl::removeIndex):
        (WebCore::IDBObjectStoreBackendImpl::openCursor):
        * storage/IDBObjectStoreBackendImpl.h:
2010-08-26  Jeremy Orlow  <jorlow@chromium.org>

        Reviewed by Steve Block.

        Add index insertion support to IndexedDB.
        https://bugs.webkit.org/show_bug.cgi?id=44695

        Add asserts on the [] operator.

        * public/WebVector.h:
        (WebKit::WebVector::operator[]):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66820 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Reviewed by Adam Barth.

        https://bugs.webkit.org/show_bug.cgi?id=43982
        This js file seems to be a popular script for detecting mobile WebKit but
        it has a flaw that makes it fail on Android and BlackBerry.

        The issue is that it looks for "Mobile/" but Android and BlackBerry
        use "Mobile" with no numbering after it.
        Changed RegExp pattern " Mobile/" to " Mobile\\b".

        * misc/WebKitDetect.js:
        (WebKitDetect.isMobile):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66819 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Reviewed by Adam Barth.

        [v8] bypass caches when query memory usage from post GC and in crash handler.
        https://bugs.webkit.org/show_bug.cgi?id=45036

        Add Chromium-specific API to query actual memory usage which bypasses any caches.

        * platform/chromium/ChromiumBridge.h:
2010-09-06  Anton Muhin  <antonm@chromium.org>

        Reviewed by Adam Barth.

        [v8] bypass caches when query memory usage from post GC and in crash handler.
        https://bugs.webkit.org/show_bug.cgi?id=45036

        Add Chromium-specific API to query actual memory usage which bypasses any caches.

        * public/WebKitClient.h:
        (WebKit::WebKitClient::actualMemoryUsageMB):
        * src/ChromiumBridge.cpp:
        (WebCore::ChromiumBridge::actualMemoryUsageMB):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66818 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Unreviewed, skipping a crashing media test.

        [GStreamer] multiple <video> elements + PulseAudio are crash-prone
        https://bugs.webkit.org/show_bug.cgi?id=44235

        * platform/gtk/Skipped: Skip media/controls-styling.html

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66817 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Unreviewed: upstream Chromium SVG test_expectations

        * platform/chromium/test_expectations.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66816 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Reviewed by Darin Adler.

        Rename SecurityOrigin::canLoad to canDisplay
        https://bugs.webkit.org/show_bug.cgi?id=45214

        canLoad is a pretty opaque name.  This function is really about whether
        you can display the contents of the URL in an iframe, an image, or a
        plugin.

        * WebCore.exp.in:
        * html/HTMLMediaElement.cpp:
        (WebCore::HTMLMediaElement::isSafeToLoadURL):
        * loader/Cache.cpp:
        (WebCore::Cache::requestResource):
        * loader/FrameLoader.cpp:
        (WebCore::FrameLoader::loadFrameRequest):
        * loader/PingLoader.cpp:
        (WebCore::PingLoader::loadImage):
        * loader/SubframeLoader.cpp:
        (WebCore::SubframeLoader::loadMediaPlayerProxyPlugin):
        (WebCore::SubframeLoader::createJavaAppletWidget):
        (WebCore::SubframeLoader::loadSubframe):
        (WebCore::SubframeLoader::loadPlugin):
        * loader/SubresourceLoader.cpp:
        (WebCore::SubresourceLoader::create):
        * page/SecurityOrigin.cpp:
        (WebCore::SecurityOrigin::canDisplay):
        * page/SecurityOrigin.h:
        * plugins/PluginView.cpp:
        (WebCore::PluginView::load):
2010-09-06  Adam Barth  <abarth@webkit.org>

        Reviewed by Darin Adler.

        Rename SecurityOrigin::canLoad to canDisplay
        https://bugs.webkit.org/show_bug.cgi?id=45214

        Propagate name change.

        * Plugins/Hosted/NetscapePluginInstanceProxy.mm:
        (WebKit::NetscapePluginInstanceProxy::loadRequest):
        * Plugins/WebNetscapePluginStream.mm:
        (WebNetscapePluginStream::WebNetscapePluginStream):
        * Plugins/WebNetscapePluginView.mm:
        (-[WebNetscapePluginView loadRequest:inTarget:withNotifyData:sendNotification:]):
        * Plugins/WebPluginContainerCheck.mm:
        (-[WebPluginContainerCheck _isForbiddenFileLoad]):
        * WebView/WebFrame.mm:
        (-[WebFrame _allowsFollowingLink:]):
2010-09-06  Adam Barth  <abarth@webkit.org>

        Reviewed by Darin Adler.

        Rename SecurityOrigin::canLoad to canDisplay
        https://bugs.webkit.org/show_bug.cgi?id=45214

        Propagate name change.

        * WebFrame.cpp:
        (WebFrame::allowsFollowingLink):
2010-09-06  Adam Barth  <abarth@webkit.org>

        Reviewed by Darin Adler.

        Rename SecurityOrigin::canLoad to canDisplay
        https://bugs.webkit.org/show_bug.cgi?id=45214

        Propagate name change.

        * WebProcess/Plugins/PluginView.cpp:
        (WebKit::PluginView::performFrameLoadURLRequest):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66815 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Update expected results to match change in test.  Qt's behavior is
        unchanged.  It's just the redundant expectations in the test that are
        changing.

        * platform/qt/fast/events/continuous-platform-wheelevent-in-scrolling-div-expected.txt:


git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66814 268f45cc-cd09-0410-ab3c-d52691b4dbfc

        Reviewed by Joseph Pecoraro.

        Web Inspector: browser crashes on attempt to evaluate "alert(1)" while staying on a breakpoint
        https://bugs.webkit.org/show_bug.cgi?id=44943

        Test: inspector/debugger-suspend-active-dom-objects.html

        * page/PageGroupLoadDeferrer.cpp:
        (WebCore::PageGroupLoadDeferrer::PageGroupLoadDeferrer):

2010-08-31  Yury Semikhatsky  <yurys@chromium.org>

        Reviewed by Joseph Pecoraro.

        Web Inspector: browser crashes on attempt to evaluate "alert(1)" while staying on a breakpoint
        https://bugs.webkit.org/show_bug.cgi?id=44943

        * inspector/debugger-suspend-active-dom-objects-expected.txt: Added.
        * inspector/debugger-suspend-active-dom-objects.html: Added.


git-svn-id: http://svn.webkit.org/repository/webkit/trunk@66813 268f45cc-cd09-0410-ab3c-d52691b4dbfc