2009-02-26  Darin Fisher  <darin@chromium.org>

        Reviewed by Eric Seidel.

        Minor cleanup of ChromiumBridge:
        Removing unused matchesMIMEType method and s/Javascript/JavaScript/

        * platform/chromium/ChromiumBridge.h:
        * platform/chromium/MimeTypeRegistryChromium.cpp:
        (WebCore::MIMETypeRegistry::isSupportedImageMIMEType):
        (WebCore::MIMETypeRegistry::isSupportedJavaScriptMIMEType):
        (WebCore::MIMETypeRegistry::isSupportedNonImageMIMEType):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41268 268f45cc-cd09-0410-ab3c-d52691b4dbfc

WebCore:

2009-02-26  David Hyatt  <hyatt@apple.com>

        Reviewed by Sam Weinig

        Fix text-bottom vertical alignment.  It was incorrectly aligning the bottom of the descent instead of including
        the line-height below the descent.

        Added fast/inline/vertical-align-text-bottom.html

        * rendering/RenderBoxModelObject.cpp:
        (WebCore::RenderBoxModelObject::verticalPosition):

LayoutTests:

2009-02-26  David Hyatt  <hyatt@apple.com>

        Reviewed by Sam Weinig

        Fix text-bottom vertical alignment.  It was incorrectly aligning the bottom of the descent instead of including
        the line-height below the descent.

        * fast/inline/vertical-align-text-bottom.html: Added.
        * platform/mac/fast/inline/vertical-align-text-bottom-expected.checksum: Added.
        * platform/mac/fast/inline/vertical-align-text-bottom-expected.png: Added.
        * platform/mac/fast/inline/vertical-align-text-bottom-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41267 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-26  Dirk Schulze  <krit@webkit.org>

        Reviewed by Eric Seidel.

        Make SVG Masking platform independet with the use of ImageBuffer::getImageData(),
        ImageBuffer::putImageData() and GraphicsContext::clipToImageBuffer(). Every platform has
        just to implement this three methods to support SVG Masking now.

        Make SVG Masking platform aware
        https://bugs.webkit.org/show_bug.cgi?id=19243

        * GNUmakefile.am:
        * WebCore.pro:
        * WebCore.vcproj/WebCore.vcproj:
        * WebCore.xcodeproj/project.pbxproj:
        * svg/graphics/SVGResourceMasker.cpp:
        (WebCore::SVGResourceMasker::applyMask):
        * svg/graphics/cairo/SVGResourceMaskerCairo.cpp: Removed.
        * svg/graphics/cg/SVGResourceMaskerCg.cpp: Removed.
        * svg/graphics/cg/SVGResourceMaskerCg.mm: Removed.
        * svg/graphics/qt/SVGResourceMaskerQt.cpp: Removed.
        * svg/graphics/skia/SVGResourceMaskerSkia.cpp: Removed.

        LayoutTests:
        Corrected some tests after the new SVG Masking implementation for Mac.

        Make SVG Masking platform aware
        https://bugs.webkit.org/show_bug.cgi?id=19243

        * platform/mac/svg/W3C-SVG-1.1/masking-mask-01-b-expected.checksum:
        * platform/mac/svg/W3C-SVG-1.1/masking-mask-01-b-expected.png:
        * platform/mac/svg/custom/grayscale-gradient-mask-expected.checksum:
        * platform/mac/svg/custom/grayscale-gradient-mask-expected.png:
        * svg/custom/mask-excessive-malloc.svg:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41266 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-26  Yong Li  <yong.li@torchmobile.com>

        Reviewed by Darin Adler, Antti Koivisto and Alexey Proskuryakov.

        Test: http/tests/misc/slow-preload-cancel.html

        https://bugs.webkit.org/show_bug.cgi?id=24133
        Clear all pending preloads in the DocLoader object when we decide to
        cancel its all requests.

        * loader/DocLoader.cpp:
        (WebCore::DocLoader::clearPendingPreloads):
        * loader/DocLoader.h:
        * loader/loader.cpp:
        (WebCore::Loader::cancelRequests):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41265 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-25  Anders Carlsson  <andersca@apple.com>

        Reviewed by Kevin Decker.

        Fix <rdar://problem/6623697>.

        * Plugins/Hosted/NetscapePluginInstanceProxy.h:
        * Plugins/Hosted/NetscapePluginInstanceProxy.mm:
        (WebKit::NetscapePluginInstanceProxy::print):
        Ask the plug-in host to print, create a CGImage of the returned bytes and draw
        the image into the passed in context.

        * Plugins/Hosted/WebHostedNetscapePluginView.mm:
        (-[WebHostedNetscapePluginView drawRect:]):
        When printing, call NetscapePluginInstanceProxy::print.

        * Plugins/Hosted/WebKitPluginHost.defs:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41264 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-26  Dimitri Glazkov  <dglazkov@chromium.org>

        Reviewed by Eric Seidel.

        https://bugs.webkit.org/show_bug.cgi?id=24182
        Add NodeFilter, NodeIterator, and TreeWalker custom V8 bindings.

        * bindings/v8/custom/V8NodeFilterCustom.cpp: Added.
        (WebCore::CALLBACK_FUNC_DECL):
        * bindings/v8/custom/V8NodeIteratorCustom.cpp: Added.
        (WebCore::toV8):
        (WebCore::CALLBACK_FUNC_DECL):
        * bindings/v8/custom/V8TreeWalkerCustom.cpp: Added.
        (WebCore::toV8):
        (WebCore::CALLBACK_FUNC_DECL):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41263 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Darin Adler.

        https://bugs.webkit.org/show_bug.cgi?id=23500
        KURL::parse() incorrectly compares its result to original string

        * platform/KURL.cpp: (WebCore::KURL::parse): Take string length into account.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41262 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-26  Ojan Vafai  <ojan@chromium.org>

        Reviewed by Kevin McCullough.

        Manual test for inspector node highlighting.

        * manual-tests/inspector/highlight-nodes.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41261 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-16  Anantanarayanan Iyengar  <ananta@chromium.org>

        Reviewed by Darin Fisher.

        https://bugs.webkit.org/show_bug.cgi?id=23973
        ScrollView::scrollContents can be invoked during view shutdown.  In
        this scenario the FrameView::hostWindow method can return NULL, which
        indicates that the frame/page is being destroyed. This causes a crash
        when we try to dereference a NULL hostWindow pointer.  Fix is to add a
        NULL check for this.

        * platform/ScrollView.cpp:
        (WebCore::ScrollView::scrollContents):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41260 268f45cc-cd09-0410-ab3c-d52691b4dbfc

WebCore:

2009-02-26  Rahul Kuchhal  <kuchhal@chromium.org>

        Reviewed by Dave Hyatt.

        https://bugs.webkit.org/show_bug.cgi?id=24003
        Fix a crash caused by unsafe type conversion.

        Test: fast/block/positioning/absolute-in-inline-rtl-4.html

        * rendering/RenderBox.cpp:
        (WebCore::RenderBox::calcAbsoluteHorizontalValues):

LayoutTests:

2009-02-26  Rahul Kuchhal  <kuchhal@chromium.org>

        Reviewed by Dave Hyatt.

        https://bugs.webkit.org/show_bug.cgi?id=24003
        Fix a crash caused by unsafe type conversion.

        * fast/block/positioning/absolute-in-inline-rtl-4-expected.txt: Added.
        * fast/block/positioning/absolute-in-inline-rtl-4.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41259 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-26  Ojan Vafai  <ojan@chromium.org>

        Reviewed by Adam Roben.

        https://bugs.webkit.org/show_bug.cgi?id=24202
        Have drawNodeHighlight clip instead of clearing. This makes it work for
        both Safari and Chromium since Chromium draws the inspector highlighting
        in the same buffer as the page.

        * inspector/InspectorController.cpp:
        (WebCore::quadToPath):
        (WebCore::drawOutlinedQuad):
        (WebCore::drawOutlinedQuadWithClip):
        (WebCore::drawHighlightForBox):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41258 268f45cc-cd09-0410-ab3c-d52691b4dbfc

WebCore:

2009-02-26  David Hyatt  <hyatt@apple.com>

        Reviewed by Dan Bernstein & Darin Adler

        Make sure the border/padding are properly omitted at the start of an inline that is a continuation.

        Added fast/inline/inline-continuation-borders.html

        * rendering/InlineFlowBox.cpp:
        (WebCore::InlineFlowBox::determineSpacingForFlowBoxes):

LayoutTests:

2009-02-26  David Hyatt  <hyatt@apple.com>

        Reviewed by Dan Bernstein & Darin Adler

        Make sure the border/padding are properly omitted at the start of an inline that is a continuation.

        * fast/inline/inline-continuation-borders.html: Added.
        * platform/mac/fast/inline/inline-continuation-borders-expected.checksum: Added.
        * platform/mac/fast/inline/inline-continuation-borders-expected.png: Added.
        * platform/mac/fast/inline/inline-continuation-borders-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41257 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Darin Adler.

        https://bugs.webkit.org/show_bug.cgi?id=24024
        REGRESSION (r39845): Assertion failure in -[WebHistoryItem dictionaryRepresentation] when
        archiving a submission to about:blank

        I don't know how to make an automated test for this bug.

        * History/WebHistory.mm: (-[WebHistory _visitedURL:withTitle:method:wasFailure:]):
        * WebHistory.cpp: (WebHistory::visitedURL):
        Account for the fact that HTTP method may be non-empty for non-HTTP requests.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41256 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-26  Simon Fraser  <simon.fraser@apple.com>

        Build fix, no review.

        Try to fix the wx build after r41218.

        * WebCoreSources.bkl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41255 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-25  Pamela Greene  <pam@chromium.org>

        Reviewed by Darin Fisher.

        Add test that custom DOM properties survive garbage collection.
        https://bugs.webkit.org/show_bug.cgi?id=24142

        * fast/dom/Window/customized-property-survives-gc-expected.txt: Added.
        * fast/dom/Window/customized-property-survives-gc.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41254 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Darin Adler.

        https://bugs.webkit.org/show_bug.cgi?id=19527
        ASSERTION FAILED: containerA && containerB

        Test: fast/dom/Range/bug-19527.html

        * dom/Range.cpp:
        (WebCore::Range::isPointInRange):
        (WebCore::Range::comparePoint):
        Bring the behavior in line with current Firefox, making it impossible for these methods
        to pass a null container to compareBoundaryPoints().

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41253 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-26  Carol Szabo  <carol.szabo@nokia.com>

        Reviewed by Darin Adler.

        https://bugs.webkit.org/show_bug.cgi?id=24099
        ARM Compiler Warnings in pcre_exec.cpp

        * pcre/pcre_exec.cpp:
        (match):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41252 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-26  Xan Lopez  <xan@gnome.org>

        Reviewed by Holger Freyther.

        https://bugs.webkit.org/show_bug.cgi?id=16947
        [GTK] Missing HTTP Auth challenge

        Take marshallers to be built from a manually maintained list
        instead of grepping the sources.

        It's much faster, especially so now that we want to add
        marshallers from WebCore too. A system to only take into account
        the modified files when generating the marshallers from sources
        could be hacked, but I think it's overkill considering how rarely
        a new marshaller is added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41251 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fix the ChangeLog s/frame/content.

I should have read more about the bug history and should have
done the substitution before landing.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41250 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-26  Xan Lopez  <xan@gnome.org>

        Reviewed by Alexey Proskuryakov.

        https://bugs.webkit.org/show_bug.cgi?id=24193
        [GTK] Checkbuttons not activated with space

        Do not swallow key events with GtkIMContext for non-editable
        frames (FIXME: I don't think frame is the right name here).

        * WebCoreSupport/EditorClientGtk.cpp:
        (WebKit::EditorClient::handleInputMethodKeydown):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41249 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-26  Xan Lopez  <xan@gnome.org>

        Reviewed by Alexey Proskuryakov.

        https://bugs.webkit.org/show_bug.cgi?id=24103
        [GTK] Use correct return value for WebView button-release handler

        We are returning whatever the core code tells us it did, but this
        does not play well with the GTK+ model. GTK+ in general expects it
        will see a button-release if it saw a button-press and no
        motion/leave/etc events in between. EventHandler.cpp will, in some
        cases, not handle press but handle release, confusing the parent
        container of the WebView.

        As a workaround return always FALSE for button-release (this is
        the same than the Windows port does).

        * webkit/webkitwebview.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41248 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-26  Xan Lopez  <xan@gnome.org>

        Reviewed by Holger Freyther.

        https://bugs.webkit.org/show_bug.cgi?id=17585
        [gtk] get|set encoding api

        Add functions to get and set a custom encoding an a view.

        * webkit/webkitprivate.h:
        * webkit/webkitwebview.cpp:
        (webkit_web_view_set_custom_encoding):
        (webkit_web_view_get_custom_encoding):
        * webkit/webkitwebview.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41247 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Make font-variant:small-caps work properly (bug #20358)

Set the computed size of the small caps font data rather than the specified size
so that the small-caps text is rendered smaller

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41246 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-26  Benjamin Meyer  <benjamin.meyer@torchmobile.com>

        Reviewed by George Staikos.

        https://bugs.webkit.org/show_bug.cgi?id=24062
        QNetworkCookieJar expects the url and not the policy url.  Sending the
        policy url will cause QNetworkCookieJar to behave incorrectly.  One
        example would be a cookie that does not have a path or domain.
        QNetworkCookieJar will use the url it is given to fill in default values.
        This allows setting cookies on the url of the main frame from an iFrame
        when the cookie should be set on the url of the iFrame.

        Originally noticed on http://writer.zoho.com/jsp/home.jsp?serviceurl=/index.do

        * platform/qt/CookieJarQt.cpp:
        (WebCore::setCookies):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41245 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-26  Charles Wei  <charles.wei@torchmobile.com.cn>

        Reviewed by George Staikos.

        make WebKit/Qt compile with SVG disabled

        * WebCore.pro:
        * platform/graphics/qt/FontPlatformDataQt.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41244 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-25  Cameron Zwarich  <cwzwarich@uwaterloo.ca>

        Reviewed by Gavin Barraclough.

        Bug 24086: Regression (r40993): WebKit crashes after logging in to lists.zenbe
        <https://bugs.webkit.org/show_bug.cgi?id=24086>
        <rdar://problem/6625111>

        The numeric sort optimization in r40993 generated bytecode for a function
        without generating JIT code. This breaks an assumption in some parts of
        the JIT's function calling logic that the presence of a CodeBlock implies
        the existence of JIT code.

        In order to fix this, we simply generate JIT code whenever we check whether
        a function is a numeric sort function. This only incurs an additional cost
        in the case when the function is a numeric sort function, in which case it
        is not expensive to generate JIT code for it.

        JavaScriptCore:

        * runtime/ArrayPrototype.cpp:
        (JSC::isNumericCompareFunction):

        LayoutTests:

        * fast/js/resources/sort-no-jit-code-crash.js: Added.
        * fast/js/sort-no-jit-code-crash-expected.txt: Added.
        * fast/js/sort-no-jit-code-crash.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41243 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-25  Gustavo Noronha Silva  <gns@gnome.org>

        Unreviewed build fix. Add WebCore/workers to the list of paths
        searched by the bindings generator perl script.

        * GNUmakefile.am:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41242 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-25  Sam Weinig  <sam@webkit.org>

        Reviewed by Geoffrey Garen.

        Add canvas to the list of RenderObjects that can mark a page as
        visually not empty.

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

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41241 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-25  Eric Carlson  <eric.carlson@apple.com>

        Reviewed by Dave Hyatt.

        * rendering/RenderBox.cpp:
        (WebCore::RenderBox::nodeAtPoint): Remove assertion fired when child has layer
        as RenderMedia with controls always has a layer.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41240 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-25  Simon Fraser  <simon.fraser@apple.com>

        Reviewed by Eric Seidel

        Fix spew about a WebView being deallocated while key value observers are
        still registered with it by making sure that we always stop observing
        _isUsingAcceleratedCompositing in -[DumpRenderTreeWindow close].

        * DumpRenderTree/mac/DumpRenderTree.mm:
        (dumpRenderTree):
        * DumpRenderTree/mac/DumpRenderTreeWindow.mm:
        (-[DumpRenderTreeWindow close]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41239 268f45cc-cd09-0410-ab3c-d52691b4dbfc

WebCore:

        Reviewed by Simon Fraser.

        - fix https://bugs.webkit.org/show_bug.cgi?id=24130
          <rdar://problem/6618196> Paint very slow when horizontally resizing

        Test: fast/gradients/background-clipped.html

        * rendering/RenderBoxModelObject.cpp:
        (WebCore::RenderBoxModelObject::paintFillLayerExtended): Intersect the
        destination rectangle passed to drawTiledImage() with the dirty
        rectangle. This makes it more likely for the destination rect to be
        contained in a single tile rect, which results in a faster code path
        being taken down the road (just drawing a single tile instead of
        tiling).

LayoutTests:

        Reviewed by Simon Fraser.

        - test for https://bugs.webkit.org/show_bug.cgi?id=24130
          <rdar://problem/6618196> Paint very slow when horizontally resizing

        * fast/gradients/background-clipped.html: Added.
        * platform/mac/fast/gradients/background-clipped-expected.checksum: Added.
        * platform/mac/fast/gradients/background-clipped-expected.png: Added.
        * platform/mac/fast/gradients/background-clipped-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41238 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-25  Ojan Vafai  <ojan@chromium.org>

        Reviewed by Eric Seidel.

        Fix test regressions from positionForCoordinates patch.
        https://bugs.webkit.org/show_bug.cgi?id=24148

        * rendering/RenderBlock.cpp:
        (WebCore::positionForPointWithInlineChildren):
        Remove ASSERT that placeholder text codepath is hitting.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41237 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Bug 24143: Crash occurs at WebCore::AccessibilityTable::isTableExposableThroughAccessibility() when applying a link in GMail
https://bugs.webkit.org/show_bug.cgi?id=24143

When an AX object is marked dirty, do not create AX elements while going up the parent chain.
Do not allow AXRenderObjects to remove their own IDs from the cache, all the cache to do that work
Make sure the AXObjectWrapper's have an object before calling them
In AXObjectCache, change get -> getOrCreate. Use get() to only retrieve an element if it exists

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41236 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-25  Beth Dakin  <bdakin@apple.com>

        Reviewed by Geoff Garen.

        Re-working of <rdar://problem/6487249> repro crash in
        WebCore::CSSParser::parseFillImage copying entire contents of this
        page
        -and corresponding: https://bugs.webkit.org/show_bug.cgi?id=24172

        Fixes a bunch of layout test failures I caused!

        * css/CSSParser.cpp:
        (WebCore::CSSParser::parseValue):
        (WebCore::CSSParser::parseContent):
        (WebCore::CSSParser::parseFillImage):
        (WebCore::CSSParser::parseFontFaceSrc):
        (WebCore::CSSParser::parseBorderImage):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41235 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-25  Dimitri Glazkov  <dglazkov@chromium.org>

        Reviewed by Eric Seidel.

        https://bugs.webkit.org/show_bug.cgi?id=24174
        Add more V8 custom bindings.

        * bindings/v8/custom/V8CSSStyleDeclarationCustom.cpp: Added.
        (WebCore::hasCSSPropertyNamePrefix):
        (WebCore::cssPropertyName):
        (WebCore::NAMED_PROPERTY_GETTER):
        (WebCore::NAMED_PROPERTY_SETTER):
        * bindings/v8/custom/V8DOMStringListCustom.cpp: Added.
        (WebCore::INDEXED_PROPERTY_GETTER):
        (WebCore::CALLBACK_FUNC_DECL):
        * bindings/v8/custom/V8EventCustom.cpp: Added.
        (WebCore::ACCESSOR_SETTER):
        (WebCore::ACCESSOR_GETTER):
        * bindings/v8/custom/V8HTMLDocumentCustom.cpp: Added.
        (WebCore::NAMED_PROPERTY_DELETER):
        (WebCore::NAMED_PROPERTY_SETTER):
        (WebCore::NAMED_PROPERTY_GETTER):
        * bindings/v8/custom/V8HTMLFrameSetElementCustom.cpp: Added.
        (WebCore::NAMED_PROPERTY_GETTER):
        * bindings/v8/custom/V8NamedNodeMapCustom.cpp: Added.
        (WebCore::INDEXED_PROPERTY_GETTER):
        (WebCore::NAMED_PROPERTY_GETTER):
        * bindings/v8/custom/V8NodeListCustom.cpp: Added.
        (WebCore::NAMED_PROPERTY_GETTER):
        * bindings/v8/custom/V8StyleSheetListCustom.cpp: Added.
        (WebCore::NAMED_PROPERTY_GETTER):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41234 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-25  Scott Violet  <sky@google.com>

        Reviewed by Dave Hyatt.

        https://bugs.webkit.org/show_bug.cgi?id=24171
        Provides a RenderTheme method for getting the scroll bar size and
        changes RenderListBox to use it. RenderTheme returns a size of regular,
        and Mac's override to return small. Changes ScrollbarThemeChromium to
        use the scrollbarsize of the scrollbar when getting the size
        instead of passing in no-args.

        * platform/chromium/ScrollbarThemeChromium.cpp:
        (WebCore::ScrollbarThemeChromium::trackRect):
        (WebCore::ScrollbarThemeChromium::buttonSize):
        * rendering/RenderListBox.cpp:
        (WebCore::RenderListBox::createScrollbar):
        * rendering/RenderTheme.h:
        (WebCore::RenderTheme::scrollbarControlSizeForPart):
        * rendering/RenderThemeChromiumMac.h:
        (WebCore::RenderThemeChromiumMac::scrollbarControlSizeForPart):
        * rendering/RenderThemeMac.h:
        (WebCore::RenderThemeMac::scrollbarControlSizeForPart):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41233 268f45cc-cd09-0410-ab3c-d52691b4dbfc

JavaScriptCore:

2009-02-25  Geoffrey Garen  <ggaren@apple.com>

        Reviewed by Maciej Stachowiak.

        Fixed <rdar://problem/6611174> REGRESSION (r36701): Unable to select
        messages on hotmail (24052)

        The bug was that for-in enumeration used a cached prototype chain without
        validating that it was up-to-date.

        This led me to refactor prototype chain caching so it was easier to work
        with and harder to get wrong.

        After a bit of inlining, this patch is performance-neutral on SunSpider
        and the v8 benchmarks.

        * interpreter/Interpreter.cpp:
        (JSC::Interpreter::tryCachePutByID):
        (JSC::Interpreter::tryCacheGetByID):
        * jit/JITStubs.cpp:
        (JSC::JITStubs::tryCachePutByID):
        (JSC::JITStubs::tryCacheGetByID):
        (JSC::JITStubs::cti_op_get_by_id_proto_list): Use the new refactored goodness. See
        lines beginning with "-" and smile.

        * runtime/JSGlobalObject.h:
        (JSC::Structure::prototypeForLookup): A shout out to const.

        * runtime/JSPropertyNameIterator.h:
        (JSC::JSPropertyNameIterator::next): We can use a pointer comparison to
        see if our cached structure chain is equal to the object's structure chain,
        since in the case of a cache hit, we share references to the same structure
        chain.

        * runtime/Operations.h:
        (JSC::countPrototypeChainEntriesAndCheckForProxies): Use the new refactored
        goodness.

        * runtime/PropertyNameArray.h:
        (JSC::PropertyNameArray::PropertyNameArray):
        (JSC::PropertyNameArray::setShouldCache):
        (JSC::PropertyNameArray::shouldCache): Renamed "cacheable" to "shouldCache"
        to communicate that the client is specifying a recommendation, not a
        capability.

        * runtime/Structure.cpp:
        (JSC::Structure::Structure): No need to initialize a RefPtr.
        (JSC::Structure::getEnumerablePropertyNames): Moved some code into helper
        functions.

        (JSC::Structure::prototypeChain): New centralized accessor for a prototype
        chain. Revalidates on every access, since the objects in the prototype
        chain may have mutated.

        (JSC::Structure::isValid): Helper function for revalidating a cached
        prototype chain.

        (JSC::Structure::getEnumerableNamesFromPropertyTable):
        (JSC::Structure::getEnumerableNamesFromClassInfoTable): Factored out of
        getEnumerablePropertyNames.

        * runtime/Structure.h:

        * runtime/StructureChain.cpp:
        (JSC::StructureChain::StructureChain):
        * runtime/StructureChain.h:
        (JSC::StructureChain::create): No need for structureChainsAreEqual, since
        we use pointer equality now. Refactored StructureChain to make a little
        more sense and eliminate special cases for null prototypes.

LayoutTests:

2009-02-24  Geoffrey Garen  <ggaren@apple.com>

        Reviewed by Maciej Stachowiak.

        Added a test for <rdar://problem/6611174> REGRESSION (r36701): Unable to
        select messages on hotmail (24052)

        * fast/js/for-in-cached-expected.txt: Added.
        * fast/js/for-in-cached.html: Added.
        * fast/js/resources/for-in-cached.js: Added.
        (forIn):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41232 268f45cc-cd09-0410-ab3c-d52691b4dbfc

WebCore:

2009-02-25  Beth Dakin  <bdakin@apple.com>

        Reviewed by Darin Adler.

        Fix for <rdar://problem/6487249> repro crash in
        WebCore::CSSParser::parseFillImage copying entire contents of this
        page
        -and corresponding: https://bugs.webkit.org/show_bug.cgi?id=24172

        * css/CSSParser.cpp:
        (WebCore::CSSParser::parseValue): Null-check m_styleSheet

LayoutTests:

2009-02-25  Beth Dakin  <bdakin@apple.com>

        Reviewed by Darin Adler.

        Test for <rdar://problem/6487249> repro crash in
        WebCore::CSSParser::parseFillImage copying entire contents of this
        page
        -and corresponding: https://bugs.webkit.org/show_bug.cgi?id=24172

        * editing/pasteboard/copy-crash-with-extraneous-attribute-expected.txt: Added.
        * editing/pasteboard/copy-crash-with-extraneous-attribute.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41231 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-25  Adam Treat  <adam.treat@torchmobile.com>

        Fix the Qt build.

        * WebCore.pro:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41230 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-25  Simon Fraser  <simon.fraser@apple.com>

        Build fix with ACCELERATED_COMPOSITING turned on.

        I missed a spot in my last commit in renaming to
        _stoppedAcceleratedCompositingForFrame:

        * WebView/WebHTMLView.mm:
        (-[WebHTMLView close]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41229 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-25  Adam Treat  <adam.treat@torchmobile.com>

        Reviewed by Alexey Proskuryakov.

        Do not queue the calls to 'DumpRenderTree::dump()' as this can result
        in more than one call as a test that calls 'notifyDone()' can then be
        subsequently fully loaded and initiate a second dump.  Also make sure
        to stop any existing page load that is happening before running the next
        test.  Combined this serves to produce 217 more passing tests for the
        Qt port or roughly 5% at this point.

        * DumpRenderTree/qt/DumpRenderTree.cpp:
        (WebCore::DumpRenderTree::DumpRenderTree):
        (WebCore::DumpRenderTree::open):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41228 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-25  Jan Michael Alonzo  <jmalonzo@webkit.org>

        Gtk build fix. Not reviewed.

        Add files to the build per r41218.
        Rename WebkitPoint.h to WebKitPoint.h in DOMWindow.cpp

        * GNUmakefile.am:
        * page/DOMWindow.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41227 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-25  Zan Dobersek  <zandobersek@gmail.com>

        Reviewed by Alexey Proskuryakov.

        https://bugs.webkit.org/show_bug.cgi?id=24043
        When faced with URLs with unsupported protocol on Gtk port,
        report the error through an idle function and return true, so
        a proper resource handle is created.

        * platform/network/soup/ResourceHandleSoup.cpp:
        (WebCore::reportUnknownProtocolError):
        (WebCore::ResourceHandle::start):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41226 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Windows build fix.
        Use struct to forward declare ResourceRequest.

        * history/HistoryItem.h:
        * inspector/InspectorController.h:
        * loader/DocumentThreadableLoader.h:
        * loader/FrameLoaderClient.h:
        * loader/MainResourceLoader.h:
        * loader/SubresourceLoader.h:
        * loader/SubresourceLoaderClient.h:
        * loader/ThreadableLoader.h:
        * loader/WorkerThreadableLoader.h:
        * platform/CrossThreadCopier.h:
        * platform/network/ResourceHandle.h:
        * platform/network/ResourceHandleClient.h:
        * platform/network/ResourceRequestBase.h:
        * platform/network/cf/ResourceRequestCFNet.h:
        * xml/XMLHttpRequest.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41225 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-25  Simon Fraser  <simon.fraser@apple.com>

        Reviewed by Dan Bernstein

        https://bugs.webkit.org/show_bug.cgi?id=23854

        Make an observable property, _isUsingAcceleratedCompositing, on
        WebView that DumpRenderTree can use to specialize behavior.

        This is implemented via a count of Frames that are using
        accelerated compositing.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41224 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-25  Andreia Gaita  <shana@jitted.com>

        Reviewed by Alp Toker.

        Update Win def files with recent JavaScriptCore API additions.

        * WebKit.vcproj/WebKit.def:
        * WebKit.vcproj/WebKit_Cairo.def:
        * WebKit.vcproj/WebKit_debug.def:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41223 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Partial Windows build fix.

        * DerivedSources.cpp:
        * WebCore.vcproj/WebCore.vcproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41222 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-25  Dirk Schulze  <krit@webkit.org>

        Reviewed by Oliver Hunt.

        Ported arcTo to Qt. Qt has no native support for arcTo. This changes
        calculate the behavior of arcTo and draws it with lineTo and arc.

        [QT] implement Canvas arcTo
        https://bugs.webkit.org/show_bug.cgi?id=23873

        * platform/graphics/qt/PathQt.cpp:
        (WebCore::Path::addArcTo):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41221 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-25  Simon Fraser  <simon.fraser@apple.com>

        Reviewed by Dan Bernstein

        Remove idl files from Resources, and sort.

        * WebCore.xcodeproj/project.pbxproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41220 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Use timeBeginPeriod to enable timing resolution greater than 16ms in command line jsc for Windows.
        Allows more accurate reporting of benchmark times via command line jsc.exe. Doesn't affect WebKit's use of JavaScriptCore.

        Reviewed by Adam Roben.

        * jsc.cpp:
        (main):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41219 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Simon Fraser.

        https://bugs.webkit.org/show_bug.cgi?id=23943

        Added webkitConvertPointFromNodeToPage and webkitConvertPointFromPageToNode on
        the window object. Also added WebKitPoint object, which is passed in and out
        of these functions.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41218 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-25  Jian Li  <jianli@chromium.org>

        Reviewed by Alexey Proskuryakov.

        Remove "#if ENABLE(WORKERS)" wrap from CrossThreadCopier files.
        https://bugs.webkit.org/show_bug.cgi?id=24145

        * platform/CrossThreadCopier.cpp:
        * platform/CrossThreadCopier.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41217 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-25  David Levin  <levin@chromium.org>

        Reviewed by Alexey Proskuryakov.

        Bug 23688: ThreadableLoader needs a sync implementation for Workers.
        <https://bugs.webkit.org/show_bug.cgi?id=23688>

        No observable change in behavior, so no test.

        * loader/ThreadableLoader.cpp:
        (WebCore::ThreadableLoader::loadResourceSynchronously):
        * loader/ThreadableLoaderClientWrapper.h:
        (WebCore::ThreadableLoaderClientWrapper::clearClient):
        (WebCore::ThreadableLoaderClientWrapper::done):
        (WebCore::ThreadableLoaderClientWrapper::didFinishLoading):
        (WebCore::ThreadableLoaderClientWrapper::didFail):
        (WebCore::ThreadableLoaderClientWrapper::didFailRedirectCheck):
        (WebCore::ThreadableLoaderClientWrapper::ThreadableLoaderClientWrapper):
        Expose whether the loader is done (based on what callbacks were done).

        * loader/WorkerThreadableLoader.cpp:
        (WebCore::WorkerThreadableLoader::WorkerThreadableLoader):
        (WebCore::WorkerThreadableLoader::loadResourceSynchronously):
        Each loader is given its own mode so that only its callbacks get through the run loop.

        The xhr spec says that the readystatechange events are synchronous, so in the case of a
        nested sync xhr no readystatechange events should be fired for the outer xhr.

        (WebCore::WorkerThreadableLoader::MainThreadBridge::MainThreadBridge):

        * loader/WorkerThreadableLoader.h:
        (WebCore::WorkerThreadableLoader::create):
        (WebCore::WorkerThreadableLoader::done):

        * workers/WorkerRunLoop.cpp:
        (WebCore::WorkerRunLoop::WorkerRunLoop):
        * workers/WorkerRunLoop.h:
        (WebCore::WorkerRunLoop::createUniqueId):
        Simple method to create a uniqueId on demand with respect to the run loop.

        * workers/WorkerThread.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41216 268f45cc-cd09-0410-ab3c-d52691b4dbfc

WebCore:

2009-02-25  David Levin  <levin@chromium.org>

        Reviewed by Alexey Proskuryakov.

        Bug 24089: ThreadableLoader::loadResourceSynchronously should do callbacks like the async code.
        <https://bugs.webkit.org/show_bug.cgi?id=24089>

        Make threadable loader callbacks to happen during the sync load call.

        Changes the behavior of sync xhr for insecure redirects in two ways:
        + Sends an error event instead of an abort event (which is the same as async xhr's behavior).
        + Throws a network exception which is what other browsers do and what the spec
        says to do (http://www.w3.org/TR/XMLHttpRequest/).

        * loader/DocumentThreadableLoader.cpp:
        (WebCore::DocumentThreadableLoader::loadResourceSynchronously):
        * loader/DocumentThreadableLoader.h:
        * loader/ThreadableLoader.cpp:
        (WebCore::ThreadableLoader::loadResourceSynchronously):
        * loader/ThreadableLoader.h:
        * xml/XMLHttpRequest.cpp:
        (WebCore::XMLHttpRequest::XMLHttpRequest):
        (WebCore::XMLHttpRequest::loadRequestSynchronously):
        (WebCore::XMLHttpRequest::loadRequestAsynchronously):
        (WebCore::XMLHttpRequest::didFail):
        (WebCore::XMLHttpRequest::didFailRedirectCheck):
        * xml/XMLHttpRequest.h:

LayoutTests:

2009-02-25  David Levin  <levin@chromium.org>

        Reviewed by Alexey Proskuryakov.

        Bug 24089: ThreadableLoader::loadResourceSynchronously should do callbacks like the async code.
        <https://bugs.webkit.org/show_bug.cgi?id=24089>

        Modified the test to account for behavior change and output more information to better detect
        future changes in behavior.

        * http/tests/xmlhttprequest/xmlhttprequest-unsafe-redirect-expected.txt:
        * http/tests/xmlhttprequest/xmlhttprequest-unsafe-redirect.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41215 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by David Hyatt.

        https://bugs.webkit.org/show_bug.cgi?id=23368

        Added logic to correctly hit test accelerated layers.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41214 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Darin Adler.

        https://bugs.webkit.org/show_bug.cgi?id=24067
        REGRESSION: Crash in WebCore::Document::initSecurityContext

        The crash started to happen when we removed a check for frame->document() being null.
        However, the original document shouldn't be null here, because it is needed to alias
        security origins. So, this patch fixes the crash by correcting security origin behavior.

        Test: http/tests/security/aboutBlank/window-open-self-about-blank.html
        This tests for not crashing, and for inheriting the domain from the document being replaced.
        Preserving the aliasing cannot be tested for automatically, because we'd need a non-trivial
        domain to remove a prefix from.

        * loader/FrameLoader.cpp: (WebCore::FrameLoader::begin): Create a new document before
        clearing the frame, so that Document::initSecurityContext() could access the old one.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41213 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-25  Jay Campan  <jcampan@google.com>

        Reviewed by Darin Fisher.

        https://bugs.webkit.org/show_bug.cgi?id=24066

        Items in drop-downs were not painted correctly.  Makes sure the
        PopupListBox invalidates in the coordinates of the window as this is
        FramelessScrollView::invalidateRect paints to.

        * platform/chromium/PopupMenuChromium.cpp:
        (WebCore::PopupListBox::invalidateRow):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41212 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-25  Adam Treat  <adam.treat@torchmobile.com>

        Reviewed by Zack Rusin.

        https://bugs.webkit.org/show_bug.cgi?id=24158
        Implement the queue*() methods of the layoutTestController and begin
        implementing the dump of the back/forward list.  This results in 2% more
        tests passing as well as 23 currently skipped tests now passing.

        * DumpRenderTree/qt/DumpRenderTree.cpp:
        (WebCore::DumpRenderTree::dumpBackForwardList):
        (WebCore::DumpRenderTree::dump):
        * DumpRenderTree/qt/DumpRenderTree.h:
        * DumpRenderTree/qt/DumpRenderTree.pro:
        * DumpRenderTree/qt/jsobjects.cpp:
        (findFrameNamed):
        (LoadItem::invoke):
        (ReloadItem::invoke):
        (ScriptItem::invoke):
        (BackForwardItem::invoke):
        (LayoutTestController::reset):
        (LayoutTestController::processWork):
        (LayoutTestController::maybeDump):
        (LayoutTestController::queueBackNavigation):
        (LayoutTestController::queueForwardNavigation):
        (LayoutTestController::queueLoad):
        (LayoutTestController::queueReload):
        (LayoutTestController::queueScript):
        * DumpRenderTree/qt/jsobjects.h:
        (LayoutTestController::shouldDumpBackForwardList):
        (LayoutTestController::dumpBackForwardList):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41211 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-25  Kavindra Palaraja  <kavindra.palaraja@nokia.com>

        Reviewed by Simon Hausmann.

        Fix the documentation for the linkHovered signal.

        * Api/qwebpage.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41210 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-24  Simon Fraser  <simon.fraser@apple.com>

        Reviewed by Anders Carlsson

        https://bugs.webkit.org/show_bug.cgi?id=15081

        Make display:none work for applet, emebed and object elements
        by calling rendererIsNeeded() on superclasses.

        Tests: fast/replaced/applet-display-none.html
               fast/replaced/embed-display-none.html
               fast/replaced/object-display-none.html

        * html/HTMLAppletElement.cpp:
        (WebCore::HTMLAppletElement::rendererIsNeeded):
        * html/HTMLElement.cpp:
        (WebCore::HTMLElement::rendererIsNeeded):
        * html/HTMLEmbedElement.cpp:
        (WebCore::HTMLEmbedElement::rendererIsNeeded):
        * html/HTMLObjectElement.cpp:
        (WebCore::HTMLObjectElement::rendererIsNeeded):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41209 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-24  Simon Fraser  <simon.fraser@apple.com>

        Reviewed by Dave Hyatt

        https://bugs.webkit.org/show_bug.cgi?id=24137

        Fix localToAbsolute() and absoluteToLocal() to map points through 3d transforms,
        taking perspective and transform-style: preserve-3d into account.

        In order to support transform-style: preserve-3d, which keeps elements in a
        3d space, we have to carry along an accumulated matrix in TransformState.
        We also need to apply the perspective from the parent, if any, with the
        appropriate origin.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41208 268f45cc-cd09-0410-ab3c-d52691b4dbfc

WebCore:

2009-02-24  Sam Weinig  <sam@webkit.org>

        Reviewed by David Hyatt.

        Fix for https://bugs.webkit.org/show_bug.cgi?id=23990
        Regression (r40837): JavaScript image popup doesn't work

        Make the getClientRects and getBoundingClientRect methods return rects
        relative to the viewport.

        Test: fast/dom/getBoundingClientRect-getClientRects-relative-to-viewport.html

        * dom/Element.cpp:
        (WebCore::Element::getClientRects):
        (WebCore::Element::getBoundingClientRect):

LayoutTests:

2009-02-24  Sam Weinig  <sam@webkit.org>

        Reviewed by David Hyatt.

        Test for https://bugs.webkit.org/show_bug.cgi?id=23990
        Regression (r40837): JavaScript image popup doesn't work

        * fast/dom/getBoundingClientRect-getClientRects-relative-to-viewport-expected.txt: Added.
        * fast/dom/getBoundingClientRect-getClientRects-relative-to-viewport.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41207 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-24  Ojan Vafai  <ojan@chromium.org>

        Reviewed by Eric Seidel.

        SVG pages don't have a body or an html element!
        Don't return early if there is no body.

        * editing/VisiblePosition.cpp:
        (WebCore::VisiblePosition::canonicalPosition):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41206 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Update some layout test results from my previous checkin.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41205 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Update some layout test results from my previous checkin.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41204 268f45cc-cd09-0410-ab3c-d52691b4dbfc

WebCore:

2009-02-23  David Hyatt  <hyatt@apple.com>

        Reviewed by Eric Seidel

        https://bugs.webkit.org/show_bug.cgi?id=23740, painting order wrong for normal flow elements with overflow: hidden

        This patch reworks the painting of overflow.  There is now the concept of a "self-painting layer."  All
        layers are considered to be self-painting except for overflow layers that are normal flow (and that don't have
        reflections or masks).

        If an overflow layer is not self-painting, then it ends up painted by its parent just like any other normal flow object.
        The only difference is that the clip has to be pushed and popped when painting the object's children.

        The lightweight clipping scheme used for controls has been extended to cover overflow now in this simplified case.  With
        the code consolidated into reusable push/pop functions, all of the renderers that use overflow have been patched to
        use the new functions.

        Hit testing has also been patched to check the overflow clip rect first before recurring into children.

        Scrollbar paint has been moved into RenderBlock for now, since none of the table objects support scrollbars
        yet, and scrollbar hit testing was already there anyway.  Now the two code paths are more symmetrical.

        Masks are now treated like normal flow layers (just like reflections).

        A couple of test cases have been added to fast/overflow to test the stacking order.

        * rendering/InlineFlowBox.cpp:
        (WebCore::InlineFlowBox::nodeAtPoint):
        (WebCore::InlineFlowBox::paint):
        * rendering/RenderBlock.cpp:
        (WebCore::RenderBlock::repaintOverhangingFloats):
        (WebCore::RenderBlock::paint):
        (WebCore::RenderBlock::paintChildren):
        (WebCore::RenderBlock::paintObject):
        (WebCore::RenderBlock::paintFloats):
        (WebCore::RenderBlock::insertFloatingObject):
        (WebCore::RenderBlock::floatRect):
        (WebCore::RenderBlock::lowestPosition):
        (WebCore::RenderBlock::rightmostPosition):
        (WebCore::RenderBlock::leftmostPosition):
        (WebCore::RenderBlock::addOverhangingFloats):
        (WebCore::RenderBlock::nodeAtPoint):
        (WebCore::RenderBlock::hitTestContents):
        * rendering/RenderBox.cpp:
        (WebCore::RenderBox::nodeAtPoint):
        (WebCore::RenderBox::pushContentsClip):
        (WebCore::RenderBox::popContentsClip):
        * rendering/RenderBox.h:
        (WebCore::RenderBox::paintObject):
        * rendering/RenderBoxModelObject.cpp:
        (WebCore::RenderBoxModelObject::hasSelfPaintingLayer):
        * rendering/RenderBoxModelObject.h:
        * rendering/RenderLayer.cpp:
        (WebCore::RenderLayer::paintLayer):
        (WebCore::RenderLayer::hitTestLayer):
        (WebCore::RenderLayer::shouldBeNormalFlowOnly):
        (WebCore::RenderLayer::isSelfPaintingLayer):
        * rendering/RenderLayer.h:
        * rendering/RenderTable.cpp:
        (WebCore::RenderTable::paint):
        (WebCore::RenderTable::paintObject):
        (WebCore::RenderTable::nodeAtPoint):
        * rendering/RenderTable.h:
        * rendering/RenderTableCell.cpp:
        (WebCore::RenderTableCell::paint):
        * rendering/RenderTableRow.cpp:
        (WebCore::RenderTableRow::nodeAtPoint):
        (WebCore::RenderTableRow::paint):
        * rendering/RenderTableRow.h:
        * rendering/RenderTableSection.cpp:
        (WebCore::RenderTableSection::paint):
        (WebCore::RenderTableSection::paintObject):
        (WebCore::RenderTableSection::nodeAtPoint):
        * rendering/RenderTableSection.h:
        * rendering/RenderTreeAsText.cpp:
        (WebCore::write):
        (WebCore::writeLayers):

LayoutTests:

2009-02-24  David Hyatt  <hyatt@apple.com>

        Reviewed by Eric Seidel

        https://bugs.webkit.org/show_bug.cgi?id=23740, painting order wrong for normal flow elements with overflow: hidden

        This patch reworks the painting of overflow.  There is now the concept of a "self-painting layer."  All
        layers are considered to be self-painting except for overflow layers that are normal flow (and that don't have
        reflections or masks).

        If an overflow layer is not self-painting, then it ends up painted by its parent just like any other normal flow object.
        The only difference is that the clip has to be pushed and popped when painting the object's children.

        The lightweight clipping scheme used for controls has been extended to cover overflow now in this simplified case.  With
        the code consolidated into reusable push/pop functions, all of the renderers that use overflow have been patched to
        use the new functions.

        Hit testing has also been patched to check the overflow clip rect first before recurring into children.

        Scrollbar paint has been moved into RenderBlock for now, since none of the table objects support scrollbars
        yet, and scrollbar hit testing was already there anyway.  Now the two code paths are more symmetrical.

        Masks are now treated like normal flow layers (just like reflections).

        A couple of test cases have been added to fast/overflow to test the stacking order.

        * fast/overflow/overflow-float-stacking.html: Added.
        * fast/overflow/overflow-stacking.html: Added.
        * platform/mac/fast/overflow/overflow-float-stacking-expected.checksum: Added.
        * platform/mac/fast/overflow/overflow-float-stacking-expected.png: Added.
        * platform/mac/fast/overflow/overflow-float-stacking-expected.txt: Added.
        * platform/mac/fast/overflow/overflow-stacking-expected.checksum: Added.
        * platform/mac/fast/overflow/overflow-stacking-expected.png: Added.
        * platform/mac/fast/overflow/overflow-stacking-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41203 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-24  David Levin  <levin@chromium.org>

        Reviewed by NOBODY (build fix).

        Attempted build fix for wx-mac.

        * WebCoreSources.bkl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41202 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-24  Simon Fraser  <simon.fraser@apple.com>

        Reviewed by Eric Seidel

        https://bugs.webkit.org/show_bug.cgi?id=24135

        Update the expected value after testcase change.

        * fast/events/offsetX-offsetY-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41201 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-24  David Levin  <levin@chromium.org>

        Reviewed by NOBODY (build fix).

        Attempted build fixes for qt-linux and wx-mac.

        * WebCore.pro:
        * webcore-base.bkl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41200 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-24  Ojan Vafai  <ojan@chromium.org>

        Reviewed by Sam Weinig

        Fix null pointer error. If the node is the Document, then ownerDocument()
        returns null, document() does not.

        * editing/VisiblePosition.cpp:
        (WebCore::VisiblePosition::canonicalPosition):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41199 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-24  Jian Li  <jianli@chromium.org>

        Reviewed by NOBODY (attempted build fixes).

        Fix build break for Windows and Linux.

        * GNUmakefile.am:
        * WebCore.vcproj/WebCore.vcproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41198 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-24  Dimitri Glazkov  <dglazkov@chromium.org>

        Reviewed by Eric Seidel.

        https://bugs.webkit.org/show_bug.cgi?id=24131
        Fix-up COM/RefCounted dichotomy in Chromium port.

        * page/chromium/AccessibilityObjectWrapper.h:
        (WebCore::AccessibilityObjectWrapper::AccessibilityObjectWrapper): Added
        explicit setting of recount.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41197 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-24  Dimitri Glazkov  <dglazkov@chromium.org>

        Reviewed by Eric Seidel.

        https://bugs.webkit.org/show_bug.cgi?id=24141
        Add clarity to V8 interceptor helper function.

        * bindings/v8/custom/V8HTMLPlugInElementCustom.cpp:
        (WebCore::NAMED_PROPERTY_GETTER): Renamed to notHandledByInterceptor.
        (WebCore::NAMED_PROPERTY_SETTER): Ditto.
        (WebCore::INDEXED_PROPERTY_GETTER): Ditto.
        (WebCore::INDEXED_PROPERTY_SETTER): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41196 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-24  Simon Fraser  <simon.fraser@apple.com>

        Reviewed by Eric Seidel

        Minor numeric cleanup: convert float literals to doubles.

        * platform/graphics/transforms/RotateTransformOperation.cpp:
        (WebCore::RotateTransformOperation::blend):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41195 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-24  Mark Mentovai  <mark@chromium.org>

        Reviewed by Eric Seidel.

        https://bugs.webkit.org/show_bug.cgi?id=24139
        Add missing include.

        * bindings/v8/ScriptInstance.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41194 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-24  Jian Li  <jianli@chromium.org>

        Reviewed by Alexey Proskuryakov.

        Move worker related files from dom directory to worker directory under WebCore.
        https://bugs.webkit.org/show_bug.cgi?id=24123

        * DerivedSources.make:
        * GNUmakefile.am:
        * WebCore.pro:
        * WebCore.scons:
        * WebCore.vcproj/MigrateIDLAndScripts:
        * WebCore.vcproj/WebCore.vcproj:
        * WebCore.xcodeproj/project.pbxproj:
        * platform/CrossThreadCopier.cpp: Renamed from WebCore/dom/CrossThreadCopier.cpp.
        * platform/CrossThreadCopier.h: Renamed from WebCore/dom/CrossThreadCopier.h.
        * workers/GenericWorkerTask.h: Renamed from WebCore/dom/GenericWorkerTask.h.
        * workers/Worker.cpp: Renamed from WebCore/dom/Worker.cpp.
        * workers/Worker.h: Renamed from WebCore/dom/Worker.h.
        * workers/Worker.idl: Renamed from WebCore/dom/Worker.idl.
        * workers/WorkerContext.cpp: Renamed from WebCore/dom/WorkerContext.cpp.
        * workers/WorkerContext.h: Renamed from WebCore/dom/WorkerContext.h.
        * workers/WorkerContext.idl: Renamed from WebCore/dom/WorkerContext.idl.
        * workers/WorkerContextProxy.h: Renamed from WebCore/dom/WorkerContextProxy.h.
        * workers/WorkerLocation.cpp: Renamed from WebCore/dom/WorkerLocation.cpp.
        * workers/WorkerLocation.h: Renamed from WebCore/dom/WorkerLocation.h.
        * workers/WorkerLocation.idl: Renamed from WebCore/dom/WorkerLocation.idl.
        * workers/WorkerMessagingProxy.cpp: Renamed from WebCore/dom/WorkerMessagingProxy.cpp.
        * workers/WorkerMessagingProxy.h: Renamed from WebCore/dom/WorkerMessagingProxy.h.
        * workers/WorkerObjectProxy.h: Renamed from WebCore/dom/WorkerObjectProxy.h.
        * workers/WorkerRunLoop.cpp: Renamed from WebCore/dom/WorkerRunLoop.cpp.
        * workers/WorkerRunLoop.h: Renamed from WebCore/dom/WorkerRunLoop.h.
        * workers/WorkerThread.cpp: Renamed from WebCore/dom/WorkerThread.cpp.
        * workers/WorkerThread.h: Renamed from WebCore/dom/WorkerThread.h.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41193 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-24  Simon Fraser  <simon.fraser@apple.com>

        Fix build when ACCELERATED_COMPOSITING is turned on
        (overflowList() -> normalFlowList()).

        * rendering/RenderLayerBacking.cpp:
        (WebCore::RenderLayerBacking::hasNonCompositingContent):
        (WebCore::RenderLayerBacking::paintIntoLayer):
        * rendering/RenderLayerCompositor.cpp:
        (WebCore::RenderLayerCompositor::calculateCompositedBounds):
        (WebCore::RenderLayerCompositor::computeCompositingRequirements):
        (WebCore::RenderLayerCompositor::rebuildCompositingLayerTree):
        (WebCore::RenderLayerCompositor::recursiveRepaintLayerRect):
        (WebCore::RenderLayerCompositor::layerHas3DContent):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41192 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-05  Ojan Vafai  <ojan@chromium.org> and Eric Seidel <eric@webkit.org>

        Reviewed by Dave Hyatt.

        Make cursor positions match IE6/IE7/FF3 when clicking in margins/padding
        around divs inside editable regions.
        https://bugs.webkit.org/show_bug.cgi?id=23605

        Fix clicks outside editable regions from focusing the editable region.
        https://bugs.webkit.org/show_bug.cgi?id=23607

        Removed editing/selection/contenteditable-click-outside.html as it's
        not as useful as our new tests.

        Clean up RenderBlock::positionForCoordinates to remove dead code,
        duplicate code, and generally make it more readable.

        Tests: editing/selection/click-in-margins-inside-editable-div.html
               editing/selection/click-in-padding-with-multiple-line-boxes.html
               editing/selection/click-outside-editable-div.html

        * editing/VisiblePosition.cpp:
        (WebCore::VisiblePosition::canonicalPosition):
        * rendering/RenderBlock.cpp:
        (WebCore::positionForPointRespectingEditingBoundaries):
        (WebCore::positionForPointWithInlineChildren):
        (WebCore::RenderBlock::positionForCoordinates):
        (WebCore::RenderBlock::updateFirstLetter):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41191 268f45cc-cd09-0410-ab3c-d52691b4dbfc

WebCore:

2009-02-24  Sam Weinig  <sam@webkit.org>

        Reviewed by Geoffrey Garen.

        Related to <rdar://problem/6590295>
        Allow disabling javascript: urls.

        * WebCore.base.exp:
        * html/HTMLAnchorElement.cpp:
        (WebCore::HTMLAnchorElement::parseMappedAttribute):
        * loader/FrameLoader.cpp:
        (WebCore::FrameLoader::executeIfJavaScriptURL):
        * page/Page.cpp:
        (WebCore::Page::Page):
        (WebCore::Page::setJavaScriptURLsAreAllowed):
        (WebCore::Page::javaScriptURLsAreAllowed):
        * page/Page.h:

WebKit/mac:

2009-02-24  Sam Weinig  <sam@webkit.org>

        Reviewed by Geoffrey Garen.

        Related to <rdar://problem/6590295>
        Allow disabling javascript: urls.

        * WebView/WebView.mm:
        (-[WebView _setJavaScriptURLsAreAllowed:]):
        * WebView/WebViewPrivate.h:

WebKit/win:

2009-02-24  Sam Weinig  <sam@webkit.org>

        Reviewed by Geoffrey Garen.

        Related to <rdar://problem/6590295>
        Allow disabling javascript: urls.

        * Interfaces/IWebViewPrivate.idl:
        * WebView.cpp:
        (WebView::setJavaScriptURLsAreAllowed):
        * WebView.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41190 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-24  Simon Fraser  <simon.fraser@apple.com>

        Reviewed by Darin Adler

        https://bugs.webkit.org/show_bug.cgi?id=24135

        Round the FloatPoint returned by absoluteToLocal(), rather than
        truncating it.

        * dom/MouseRelatedEvent.cpp:
        (WebCore::MouseRelatedEvent::receivedTarget):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41189 268f45cc-cd09-0410-ab3c-d52691b4dbfc

WebCore:

2009-02-24  Beth Dakin  <bdakin@apple.com>

        Reviewed by Sam Weinig.

        Fix for https://bugs.webkit.org/show_bug.cgi?id=24004 REGRESSION:
        Ordered list item marker misaligned when line height is not 1.2
        -and corresponding <rdar://problem/6602506>

        This is a regression from http://trac.webkit.org/changeset/40880
        where createInlineBox was taken off of RenderObject and moved to
        RenderBox. The problem was that the RenderBox version still needs
        to be virtual because there is still an existing implementation in
        RenderListMarker.

        * rendering/RenderBox.h:
        * rendering/RenderListMarker.cpp:
        (WebCore::RenderListMarker::createInlineBox):
        * rendering/RenderListMarker.h:

LayoutTests:

2009-02-24  Beth Dakin  <bdakin@apple.com>

        Reviewed by Sam Weinig.

        Test for https://bugs.webkit.org/show_bug.cgi?id=24004 REGRESSION:
        Ordered list item marker misaligned when line height is not 1.2
        -and corresponding <rdar://problem/6602506>

        * fast/lists/list-marker-with-line-height.html: Added.
        * platform/mac/fast/lists/list-marker-with-line-height-expected.checksum: Added.
        * platform/mac/fast/lists/list-marker-with-line-height-expected.png: Added.
        * platform/mac/fast/lists/list-marker-with-line-height-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41188 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Windows build fix

        * WebKit.vcproj/WebKit.def:
        * WebKit.vcproj/WebKit_debug.def:
        Export WTF::ThreadCondition::timedWait.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41187 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-24  Dimitri Glazkov  <dglazkov@chromium.org>

        Reviewed by Eric Seidel.

        https://bugs.webkit.org/show_bug.cgi?id=24128
        Upstream more V8 custom bindings: constructors, XSLTProcessor.

        * bindings/v8/custom/V8DOMParserConstructor.cpp: Added.
        (WebCore::CALLBACK_FUNC_DECL):
        * bindings/v8/custom/V8MessageChannelConstructor.cpp: Added.
        (WebCore::CALLBACK_FUNC_DECL):
        * bindings/v8/custom/V8WebKitCSSMatrixConstructor.cpp: Added.
        (WebCore::CALLBACK_FUNC_DECL):
        * bindings/v8/custom/V8XMLSerializerConstructor.cpp: Added.
        (WebCore::CALLBACK_FUNC_DECL):
        * bindings/v8/custom/V8XPathEvaluatorConstructor.cpp: Added.
        (WebCore::CALLBACK_FUNC_DECL):
        * bindings/v8/custom/V8XSLTProcessorCustom.cpp: Added.
        (WebCore::CALLBACK_FUNC_DECL):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41186 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-24  Adam Treat  <adam.treat@torchmobile.com>

        Reviewed by Simon Fraser.

        Adding these tests back in and removing from the Skipped list as
        these tests are all now passing on my machine as of latest revision.

        * platform/qt/Skipped:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41185 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-24  David Levin  <levin@chromium.org>

        Reviewed by Alexey Proskuryakov.

        Bug 24090: WorkerThreadableLoader needs to be able to post tasks for a mode.
        <https://bugs.webkit.org/show_bug.cgi?id=24090>

        No observable change in behavior, so no test.

        * dom/WorkerMessagingProxy.cpp:
        (WebCore::WorkerMessagingProxy::postMessageToWorkerContext):
        (WebCore::WorkerMessagingProxy::postTaskToWorkerContext):
        Fixed the bug that m_unconfirmedMessageCount was getting incremented for non-message tasks.

        (WebCore::WorkerMessagingProxy::postTaskForModeToWorkerContext):

        * dom/WorkerMessagingProxy.h:
        * loader/ThreadableLoader.cpp:
        (WebCore::ThreadableLoader::create):
        * loader/WorkerThreadableLoader.cpp:
        (WebCore::WorkerThreadableLoader::WorkerThreadableLoader):
        (WebCore::WorkerThreadableLoader::MainThreadBridge::MainThreadBridge):
        (WebCore::WorkerThreadableLoader::MainThreadBridge::didSendData):
        (WebCore::WorkerThreadableLoader::MainThreadBridge::didReceiveResponse):
        (WebCore::WorkerThreadableLoader::MainThreadBridge::didReceiveData):
        (WebCore::WorkerThreadableLoader::MainThreadBridge::didFinishLoading):
        (WebCore::WorkerThreadableLoader::MainThreadBridge::didFail):
        (WebCore::WorkerThreadableLoader::MainThreadBridge::didFailRedirectCheck):
        (WebCore::WorkerThreadableLoader::MainThreadBridge::didReceiveAuthenticationCancellation):
        Changed these methods to post task using a mode.

        * loader/WorkerThreadableLoader.h:
        (WebCore::WorkerThreadableLoader::create):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41184 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-24  Adam Treat  <adam.treat@torchmobile.com>

        Reviewed by Alexey Proskuryakov.

        Don't print out that you are generating new results if you are not in fact
        generating new results and disable generating new results by default for
        the all ports other than the canonical Apple Mac port since this can litter
        the source directory with hundreds of new results since other ports
        are not as up to date.

        * Scripts/run-webkit-tests:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41183 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-24  Adam Treat  <adam.treat@torchmobile.com>

        Reviewed by Cameron Zwarich.

        The Qt port does not support these yet nor does the nm check work with
        QMake based build.

        * Scripts/webkitdirs.pm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41181 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-24  Adam Treat  <adam.treat@torchmobile.com>

        Reviewed by Cameron Zwarich.

        Remove tests from the Qt Skipped list that no longer even exist.

        * platform/qt/Skipped:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41180 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-24  Geoffrey Garen  <ggaren@apple.com>

        Build fix?

        * GNUmakefile.am:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41179 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Darin Adler.

        https://bugs.webkit.org/show_bug.cgi?id=24091
        <rdar://problem/6468660> Start of redirect chain ends up as master entry in Application Cache

        Test: http/tests/appcache/access-via-redirect.php

        * loader/appcache/ApplicationCacheGroup.cpp:
        (WebCore::ApplicationCacheGroup::finishedLoadingMainResource): Use the URL we ended up with,
        not the original one.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41178 268f45cc-cd09-0410-ab3c-d52691b4dbfc

<rdar://problem/6407844> JavaScriptGlue should use -Os rather than -O2

Rubber-stamped by Dan Bernstein.

* Configurations/Base.xcconfig:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41177 268f45cc-cd09-0410-ab3c-d52691b4dbfc

<rdar://problem/6259220> Rename AVAILABLE_AFTER_WEBKIT_VERSION_3_1 (etc.) to match the other macros

Reviewed by Oliver Hunt.

* API/JSBasePrivate.h:
* API/JSContextRef.h:
* API/JSObjectRef.h:
* API/WebKitAvailability.h:

* Carbon/CarbonUtils.h:
* Carbon/HIWebView.h:
* Plugins/WebPlugin.h:
* Plugins/WebPluginViewFactory.h:
* WebView/WebUIDelegate.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41176 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Mark Rowe ("awesome!").

        Make test for <https://bugs.webkit.org/show_bug.cgi?id=15707> more reliable.

        * fast/dom/onload-open.html: Added waitUntilDone().

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41175 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-24  Jian Li  <jianli@chromium.org>

        Reviewed by Alexey Proskuryakov.

        Change to include WorkerObjectProxy.h instead of WorkerMessagingProxy.h in WorkerContext.cpp and WorkerScriptController.cpp.
        https://bugs.webkit.org/show_bug.cgi?id=24112

        * bindings/js/WorkerScriptController.cpp:
        * dom/WorkerContext.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41174 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-24  Peter Ammon  <pammon@apple.com>

        Reviewed by Mark Rowe.

        Fix <rdar://problem/6251410> Services can modify non-editable content in Safari

        * WebView/WebHTMLView.mm:
        (-[WebHTMLView validRequestorForSendType:returnType:]): Return self only if we can handle
        both the send and return type.  We should also handle a nil send or return type by ignoring
        the argument and returning whether we can handle the other type passed in.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41173 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Gtk] add options for 3D transforms and HTML5 channel messaging to the build
https://bugs.webkit.org/show_bug.cgi?id=24072

Reviewed by Alexey Proskuryakov.

Add options for toggling 3D transforms and HTML5 channel messaging
support on or off.

Also fix the web-workers option. It should be web-workers and not workers.

    * configure.ac:

Allow toggling of 3D transforms and HTML5 channel messaging
support for the Autotools (Gtk) build.

Also add '--gtk' in the build-webkit help doc and fix autotools
option for web-workers support.

    * Scripts/build-webkit:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41172 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-24  David Levin  <levin@chromium.org>

        Reviewed by Dave Hyatt.

        Update expected results due to better results introduced by r41090.

        * platform/mac/media/video-layer-crash-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41171 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-23  Antti Koivisto  <antti@apple.com>

        Reviewed by Oliver Hunt.

        <rdar://problem/6613796> Extended text codecs registered on webview creation

        Comparing a text encoding with string "GBK" ended up constructing
        TextEncoding("GBK") which in turn initialized all extended
        text codecs.

        * platform/text/TextCodecICU.cpp:
        (WebCore::TextCodecICU::decode):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41170 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-23  Sam Weinig  <sam@webkit.org>

        Fix Qt build.

        * bridge/qt/qt_runtime.cpp:
        (JSC::Bindings::valueRealType):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41169 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-02-23  Geoffrey Garen  <ggaren@apple.com>

        Reviewed by Sam Weinig.

        Next step in splitting JIT functionality out of the Interpreter class:
        Moved vptr storage from Interpreter to JSGlobalData, so it could be shared
        between Interpreter and JITStubs, and moved the *Trampoline JIT stubs
        into the JITStubs class. Also added a VPtrSet class to encapsulate vptr
        hacks during JSGlobalData initialization.

        SunSpider says 0.4% faster. Meh.

        * JavaScriptCore.exp:
        * JavaScriptCore.xcodeproj/project.pbxproj:
        * interpreter/Interpreter.cpp:
        (JSC::Interpreter::Interpreter):
        (JSC::Interpreter::tryCacheGetByID):
        (JSC::Interpreter::privateExecute):
        * interpreter/Interpreter.h:
        * jit/JIT.cpp:
        (JSC::JIT::privateCompileMainPass):
        (JSC::JIT::privateCompile):
        (JSC::JIT::privateCompileCTIMachineTrampolines):
        * jit/JIT.h:
        (JSC::JIT::compileCTIMachineTrampolines):
        * jit/JITCall.cpp:
        (JSC::JIT::compileOpCall):
        (JSC::JIT::compileOpCallSlowCase):
        * jit/JITPropertyAccess.cpp:
        (JSC::JIT::privateCompilePatchGetArrayLength):
        * jit/JITStubs.cpp:
        (JSC::JITStubs::JITStubs):
        (JSC::JITStubs::tryCacheGetByID):
        (JSC::JITStubs::cti_vm_dontLazyLinkCall):
        (JSC::JITStubs::cti_op_get_by_val):
        (JSC::JITStubs::cti_op_get_by_val_byte_array):
        (JSC::JITStubs::cti_op_put_by_val):
        (JSC::JITStubs::cti_op_put_by_val_array):
        (JSC::JITStubs::cti_op_put_by_val_byte_array):
        (JSC::JITStubs::cti_op_is_string):
        * jit/JITStubs.h:
        (JSC::JITStubs::ctiArrayLengthTrampoline):
        (JSC::JITStubs::ctiStringLengthTrampoline):
        (JSC::JITStubs::ctiVirtualCallPreLink):
        (JSC::JITStubs::ctiVirtualCallLink):
        (JSC::JITStubs::ctiVirtualCall):
        * runtime/ArrayPrototype.cpp:
        (JSC::arrayProtoFuncPop):
        (JSC::arrayProtoFuncPush):
        * runtime/FunctionPrototype.cpp:
        (JSC::functionProtoFuncApply):
        * runtime/JSArray.h:
        (JSC::isJSArray):
        * runtime/JSByteArray.h:
        (JSC::asByteArray):
        (JSC::isJSByteArray):
        * runtime/JSCell.h:
        * runtime/JSFunction.h:
        * runtime/JSGlobalData.cpp:
        (JSC::VPtrSet::VPtrSet):
        (JSC::JSGlobalData::JSGlobalData):
        (JSC::JSGlobalData::create):
        (JSC::JSGlobalData::sharedInstance):
        * runtime/JSGlobalData.h:
        * runtime/JSString.h:
        (JSC::isJSString):
        * runtime/Operations.h:
        (JSC::jsLess):
        (JSC::jsLessEq):
        * wrec/WREC.cpp:
        (JSC::WREC::Generator::compileRegExp):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41168 268f45cc-cd09-0410-ab3c-d52691b4dbfc