GitLab

        Reviewed by Darin Adler.

        Calling focus() on an area element not in a document should not cause a NULL ptr crash
        https://bugs.webkit.org/show_bug.cgi?id=54877

        * fast/dom/HTMLAreaElement: Added.
        * fast/dom/HTMLAreaElement/area-islink-focus-null-ptr-crash-expected.txt: Added.
        * fast/dom/HTMLAreaElement/area-islink-focus-null-ptr-crash.html: Added.
2011-03-10  Berend-Jan Wever  <skylined@chromium.org>

        Reviewed by Darin Adler.

        Calling focus() on an area element not in a document should not cause a NULL ptr crash
        https://bugs.webkit.org/show_bug.cgi?id=54877

        Test: fast/dom/HTMLAreaElement/area-islink-focus-null-ptr-crash.html

        * dom/Element.cpp:
        (WebCore::Element::focus): Check element is in the document before allowing focus
        * html/HTMLAreaElement.cpp:
        (WebCore::HTMLAreaElement::imageElement): Check element has a parent before checking if its parent is a map

git-svn-id: svn://svn.chromium.org/blink/trunk@80779 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Windows rebaselines for r80755.

        * platform/win/editing/selection/caret-ltr-right-expected.txt:
        * platform/win/editing/selection/caret-rtl-2-left-expected.txt:
        * platform/win/editing/selection/caret-rtl-right-expected.txt:
        * platform/win/fast/forms/file-input-disabled-expected.txt:


git-svn-id: svn://svn.chromium.org/blink/trunk@80778 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Tony Chang.

        [Chromium] Fix default single selection select's popup background on chromium/linux
        https://bugs.webkit.org/show_bug.cgi?id=56023

        * fast/html/select-dropdown-consistent-background-color-expected.txt: Added.
        * fast/html/select-dropdown-consistent-background-color.html: Added.
2011-03-10  Xiyuan Xia  <xiyuan@chromium.org>

        Reviewed by Tony Chang.

        [Chromium] Fix default single selection select's popup background on chromium/linux
        https://bugs.webkit.org/show_bug.cgi?id=56023

        Test: fast/html/select-dropdown-consistent-background-color.html

        * css/themeChromiumLinux.css:

git-svn-id: svn://svn.chromium.org/blink/trunk@80777 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Alexey Proskuryakov.

        Testing the improved Accept header as per bug 27267.
        https://bugs.webkit.org/show_bug.cgi?id=27267

        * http/tests/misc/xhtml-expected.txt: Tests against improved Accept header.
2011-03-10  Kris Jordan  <krisjordan@gmail.com>

        Reviewed by Alexey Proskuryakov.

        Improve default Accept header to give preference to HTML over XML.
        https://bugs.webkit.org/show_bug.cgi?id=27267

        * loader/FrameLoader.cpp:Changed default accept header to match
        FireFox' as per bug 27267.

git-svn-id: svn://svn.chromium.org/blink/trunk@80776 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Xan Lopez.

        [GTK] [Webkit2] There are no scrollbars visible in the MiniBrowser
        https://bugs.webkit.org/show_bug.cgi?id=56125

        No new tests. This functionality will be tested once we have TestRunner
        implementation for WebKit2 with pixel dumping support.

        * GNUmakefile.am: Add a new define with tracks whether or not we are using
        our specialized GTK+ ScrollView.
        * platform/ScrollView.cpp: Only use the GTK+-specific ScrollView if we are compiling
        WebKit1.
        (WebCore::ScrollView::wheelEvent):
        * platform/gtk/ScrollViewGtk.cpp: Ditto.

git-svn-id: svn://svn.chromium.org/blink/trunk@80775 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Andreas Kling.

        [Qt] QtDeclarative Webview element has a fixed white background
        https://bugs.webkit.org/show_bug.cgi?id=40918

        Implement a way to change the background color of the WebView QML element.
        This feature is activated for QtWebKit 1.1 version of the plugin.

        * declarative/plugin.cpp:
        (WebKitQmlPlugin::registerTypes):
        * declarative/qdeclarativewebview.cpp:
        (QDeclarativeWebView::backgroundColor):
        (QDeclarativeWebView::setBackgroundColor):
        * declarative/qdeclarativewebview_p.h:
        * tests/qdeclarativewebview/resources/webviewbackgroundcolor.qml: Added.
        * tests/qdeclarativewebview/tst_qdeclarativewebview.cpp:
        (tst_QDeclarativeWebView::backgroundColor):
        * tests/qdeclarativewebview/tst_qdeclarativewebview.qrc:

git-svn-id: svn://svn.chromium.org/blink/trunk@80774 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Dimitri Glazkov.

        Assertion fails by validating a form twice very quickly
        https://bugs.webkit.org/show_bug.cgi?id=56069

        * fast/forms/interactive-validation-assertion-by-validate-twice-expected.txt: Added.
        * fast/forms/interactive-validation-assertion-by-validate-twice.html: Added.
        * fast/forms/interactive-validation-attach-assertion.html: Remove unnecessary offsetLeft trick.
        * platform/gtk/Skipped:
        * platform/mac/Skipped:
        * platform/qt/Skipped:
        * platform/win/Skipped:
2011-03-10  Kent Tamura  <tkent@chromium.org>

        Reviewed by Dimitri Glazkov.

        Assertion fails by validating a form twice very quickly
        https://bugs.webkit.org/show_bug.cgi?id=56069

        If the interactive validation is invoked when a form control is in
        needsLayout() state, an assertion in isFocusable() fails. To avoid it,
        Add a call to updateLayoutIgnorePendingStylesheets() before isFocusable().

        Test: fast/forms/interactive-validation-assertion-by-validate-twice.html

        * html/HTMLFormElement.cpp:
        (WebCore::HTMLFormElement::validateInteractively):
          Calls updateLayoutIgnorePendingStylesheets().

git-svn-id: svn://svn.chromium.org/blink/trunk@80773 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Unskipping some crashers on GTK+. These no longer crash locally and we'd
        like to figure out if they still crash on the bots. Move one flaky test
        to the appropriate section.

        * platform/gtk/Skipped: Unskip tests passing locally.

git-svn-id: svn://svn.chromium.org/blink/trunk@80772 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Antonio Gomes.

        [EFL] Limit touch events when only page required touch events.
        https://bugs.webkit.org/show_bug.cgi?id=54857

        Block touch events when ChromeClient::needTouchEvents got false.

        * WebCoreSupport/ChromeClientEfl.cpp:
        (WebCore::ChromeClientEfl::needTouchEvents):
        * WebCoreSupport/ChromeClientEfl.h:
        * ewk/ewk_frame.cpp:
        (ewk_frame_feed_touch_event):
        * ewk/ewk_private.h:
        * ewk/ewk_view.cpp:
        (ewk_view_need_touch_events_set):
        (ewk_view_need_touch_events_get):

git-svn-id: svn://svn.chromium.org/blink/trunk@80771 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        
Reviewed by Beth Dakin.

Static distance computation is wrong when an ancestor is also positioned.  We need to go up the container()
chain (after initially starting with the parent()) when doing this computation, so that we properly skip
intermediate boxes between two positioned blocks.

Added fast/block/positioning/static-distance-with-positioned-ancestor.html.

Source/WebCore: 

* rendering/RenderBox.cpp:
(WebCore::computeInlineStaticDistance):
(WebCore::computeBlockStaticDistance):

LayoutTests: 

* fast/block/positioning/static-distance-with-positioned-ancestor.html: Added.
* platform/mac/fast/block/positioning/static-distance-with-positioned-ancestor-expected.checksum: Added.
* platform/mac/fast/block/positioning/static-distance-with-positioned-ancestor-expected.png: Added.
* platform/mac/fast/block/positioning/static-distance-with-positioned-ancestor-expected.txt: Added.



git-svn-id: svn://svn.chromium.org/blink/trunk@80770 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Laszlo Gombos.

        [Qt] MemoryCache deadDecodedDataDeletionInterval is not exposed for client's usage
        https://bugs.webkit.org/show_bug.cgi?id=55945

        Added handling of dynamic/runtime property _q_deadDecodedDataDeletionInterval to
        set interval used to trigger when decoded data in dead list of object cache will
        be purged from object cache.

        * Api/qwebpage.cpp:
        (QWebPagePrivate::dynamicPropertyChangeEvent):

git-svn-id: svn://svn.chromium.org/blink/trunk@80769 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Antonio Gomes.

        Simplify how QWebFrame::requestedUrl() is obtained
        https://bugs.webkit.org/show_bug.cgi?id=55842

        When a load starts, store the requested URL until we know that it'll be
        available for us in the document loader -- after load finished.

        The existing auto tests cover the three different code paths in
        requestedUrl() and the new code passes the autotests. In each of those
        cases, we looked for the information in a different place, but in all
        of them, dispatchDidStartProvisionalLoad was called.

        This simplification will be useful to fix bug 32723. The way requestedUrl()
        is implementent, we can't use it as a fallback for url() when the setUrl()
        was called with an invalid URL.

        * Api/qwebframe.cpp:
        (QWebFrame::requestedUrl):
        * WebCoreSupport/FrameLoaderClientQt.cpp:
        (WebCore::FrameLoaderClientQt::dispatchDidStartProvisionalLoad):
        (WebCore::FrameLoaderClientQt::dispatchDidFinishLoad):
        * WebCoreSupport/FrameLoaderClientQt.h:
        (WebCore::FrameLoaderClientQt::lastRequestedUrl):

git-svn-id: svn://svn.chromium.org/blink/trunk@80768 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Kenneth Russell.

        [chromium] Make tiled compositor data structure more efficient.
        https://bugs.webkit.org/show_bug.cgi?id=54133

        Previously, the compositor had a sparse 2D array of tiles for the
        whole page, most of which were null.  The tiles were implicitly
        located based on their position in the array. This was inefficient
        when the page grew (e.g. infinite scrolling) and caused some bugs
        (e.g. width * height > MAX_INT).  This change modifies tiles to have
        explicit positions so they can be stored in a hash map.

        Tests: LayoutTests/compositing/

        * platform/graphics/chromium/LayerTilerChromium.cpp:
        (WebCore::LayerTilerChromium::reset):
        (WebCore::LayerTilerChromium::createTile):
        (WebCore::LayerTilerChromium::invalidateTiles):
        (WebCore::LayerTilerChromium::contentRectToTileIndices):
        (WebCore::LayerTilerChromium::tileAt):
        (WebCore::LayerTilerChromium::tileContentRect):
        (WebCore::LayerTilerChromium::tileLayerRect):
        (WebCore::LayerTilerChromium::invalidateRect):
        (WebCore::LayerTilerChromium::invalidateEntireLayer):
        (WebCore::LayerTilerChromium::update):
        (WebCore::LayerTilerChromium::updateFromPixels):
        (WebCore::LayerTilerChromium::draw):
        (WebCore::LayerTilerChromium::growLayerToContain):
        * platform/graphics/chromium/LayerTilerChromium.h:
        (WebCore::LayerTilerChromium::Tile::Tile):
        (WebCore::LayerTilerChromium::Tile::i):
        (WebCore::LayerTilerChromium::Tile::j):
        (WebCore::LayerTilerChromium::Tile::moveTo):
        (WebCore::LayerTilerChromium::TileMapKeyTraits::emptyValue):
        (WebCore::LayerTilerChromium::TileMapKeyTraits::constructDeletedValue):
        (WebCore::LayerTilerChromium::TileMapKeyTraits::isDeletedValue):

git-svn-id: svn://svn.chromium.org/blink/trunk@80767 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Unreviewed build fix for r80752; Negated the boolean logic.

        * editing/VisiblePosition.cpp:
        (WebCore::VisiblePosition::next):
        (WebCore::VisiblePosition::previous):


git-svn-id: svn://svn.chromium.org/blink/trunk@80766 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Reviewed by Mark Rowe.

* WebKit2.xcodeproj/project.pbxproj:
Make sure to base the PluginProcess target on the correct xcconfig file
in all configurations.



git-svn-id: svn://svn.chromium.org/blink/trunk@80765 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Reviewed by Adam Roben.

* platform/graphics/win/SimpleFontDataCGWin.cpp:
(WebCore::SimpleFontData::platformInit):



git-svn-id: svn://svn.chromium.org/blink/trunk@80764 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Don't paint more than 60 times per second even when scrolling

Reviewed by Sam Weinig.

* WebView/WebClipView.mm:
(-[WebClipView _immediateScrollToPoint:]):



git-svn-id: svn://svn.chromium.org/blink/trunk@80763 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Unreviewed, roll chromium version to r77259 to pick up
        webfontinfo api change.

        * DEPS:

git-svn-id: svn://svn.chromium.org/blink/trunk@80762 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        
Reviewed by Simon Fraser.

Fix for repaint issues when the root element is a table and needs to paint a background that
covers the entire canvas.  Remove paintRootBoxDecorations and replace with paintRootBoxFillLayers,
so that it can be called only for fill layer painting.  Make tables and normal boxes both
call this new method.  Fix the dirty check at the start of RenderTable paint to not happen if
the table is the root of the document (this same check exists in RenderBlock painting already).

Source/WebCore: 

* rendering/RenderBox.cpp:
(WebCore::RenderBox::paintRootBoxFillLayers):
(WebCore::RenderBox::paintBoxDecorations):
(WebCore::RenderBox::paintBoxDecorationsWithSize):
* rendering/RenderBox.h:
* rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::calculateCompositedBounds):
* rendering/RenderTable.cpp:
(WebCore::RenderTable::paint):
(WebCore::RenderTable::paintBoxDecorations):

LayoutTests: 

* fast/backgrounds/positioned-root.html: Added.
* fast/backgrounds/table-root.html: Added.
* platform/mac/fast/backgrounds/positioned-root-expected.checksum: Added.
* platform/mac/fast/backgrounds/positioned-root-expected.png: Added.
* platform/mac/fast/backgrounds/positioned-root-expected.txt: Added.
* platform/mac/fast/backgrounds/table-root-expected.checksum: Added.
* platform/mac/fast/backgrounds/table-root-expected.png: Added.
* platform/mac/fast/backgrounds/table-root-expected.txt: Added.



git-svn-id: svn://svn.chromium.org/blink/trunk@80761 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Update chromium Snow Leopard expectations.

        * platform/chromium/test_expectations.txt:

git-svn-id: svn://svn.chromium.org/blink/trunk@80760 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Unreviewed, rolling out r80749.
        http://trac.webkit.org/changeset/80749
        https://bugs.webkit.org/show_bug.cgi?id=56143

        "Caused at least 3 media tests to be flaky" (Requested by
        tonyg-cr on #webkit).

        * html/parser/HTMLDocumentParser.cpp:
        (WebCore::HTMLDocumentParser::canTakeNextToken):
        (WebCore::HTMLDocumentParser::pumpTokenizer):
        * html/parser/HTMLParserScheduler.cpp:
        * html/parser/HTMLParserScheduler.h:
        (WebCore::HTMLParserScheduler::checkForYieldBeforeToken):
        * page/FrameView.h:

git-svn-id: svn://svn.chromium.org/blink/trunk@80759 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Unreviewed, fixed Efl compilation after r80744.

        * Source/cmake/OptionsEfl.cmake:

git-svn-id: svn://svn.chromium.org/blink/trunk@80758 bbb929c8-8fbe-4397-9dbb-9b2b20218538

https://bugs.webkit.org/show_bug.cgi?id=55077

Reviewed by David Hyatt.

* platform/ScrollView.cpp:
(WebCore::ScrollView::overhangAmount):
(WebCore::ScrollView::calculateOverhangAreasForPainting):
Take the scroll origin into account when calculating overhang.

* platform/ScrollView.h:
* rendering/RenderLayer.h:
* platform/ScrollableArea.h:
(WebCore::ScrollableArea::scrollOrigin):
Move identical scroll origin member from ScrollView and RenderLayer
to shared base ScrollableArea. This is also needed so that the animator
can access it.

* platform/mac/ScrollAnimatorMac.mm:
(WebCore::ScrollAnimatorMac::pinnedInDirection):
(WebCore::ScrollAnimatorMac::smoothScrollWithEvent):
(WebCore::ScrollAnimatorMac::snapRubberBandTimerFired):
Account for a scroll origin when doing calculating scroll offsets.



git-svn-id: svn://svn.chromium.org/blink/trunk@80757 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Simon Fraser.

        Plumb through settings for accelerated drawing for canvas
        https://bugs.webkit.org/show_bug.cgi?id=56039

        * WebView/WebPreferenceKeysPrivate.h:
        * WebView/WebPreferences.mm:
        * WebView/WebPreferencesPrivate.h:
2011-03-09  Matthew Delaney  <mdelaney@apple.com>

        Reviewed by Simon Fraser.

        Plumb through settings for accelerated drawing for canvas
        https://bugs.webkit.org/show_bug.cgi?id=56039

        No new tests. Doesn't affect behavior, just adding a switch to toggle canvas backends.

        * WebCore.exp.in:
        * html/HTMLCanvasElement.cpp:
        * html/canvas/CanvasRenderingContext2D.cpp:
        * page/Settings.cpp:
        * page/Settings.h:
2011-03-09  Matthew Delaney  <mdelaney@apple.com>

        Reviewed by Simon Fraser.

        Plumb through settings for accelerated drawing for canvas
        https://bugs.webkit.org/show_bug.cgi?id=56039

        * Shared/WebPreferencesStore.h:
        * UIProcess/API/C/WKPreferences.cpp:
        * UIProcess/API/C/WKPreferencesPrivate.h:
        * WebProcess/WebPage/WebPage.cpp:


git-svn-id: svn://svn.chromium.org/blink/trunk@80756 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        
Reviewed by Simon Fraser.

CSS2.1 test suite failures because the ex unit is broken with the Ahem font.
        
Remove the code that tries to also include the maxX of the glyph bounds for the
'x' glyph, since it just causes the x-height to be way too large in cases where the 'x' glyph extends
below the baseline.

Remove the Apple Symbol hack for ex units, since the person who added that was confused by another
issue, namely that CGFontGetXHeight wasn't being properly multiplied by the pointSize.  That's why
the value was too small.  Patched the code to just multiply by pointSize and took out the hack.
        
Fix causes many tests in the css2.1 directory to progress, so no new tests required.  Many other
layout tests change because the xHeight for Lucida Grande gets smaller by a little bit, and radio
buttons use ex horizontal margins by default.

Source/WebCore: 

* platform/graphics/mac/SimpleFontDataMac.mm:
(WebCore::SimpleFontData::platformInit):

LayoutTests: 

* platform/mac/accessibility/radio-button-group-members-expected.txt:
* platform/mac/css1/box_properties/acid_test-expected.txt:
* platform/mac/css1/font_properties/font_size-expected.txt:
* platform/mac/css1/text_properties/line_height-expected.txt:
* platform/mac/css1/text_properties/text_indent-expected.txt:
* platform/mac/css1/units/length_units-expected.txt:
* platform/mac/css2.1/t040302-c61-ex-len-00-b-a-expected.txt:
* platform/mac/css2.1/t040302-c61-rel-len-00-b-ag-expected.txt:
* platform/mac/css2.1/t09-c5526c-display-00-e-expected.txt:
* platform/mac/css2.1/t1008-c44-ln-box-02-d-ag-expected.txt:
* platform/mac/css2.1/t100801-c544-valgn-01-d-ag-expected.txt:
* platform/mac/css2.1/t100801-c544-valgn-03-d-agi-expected.txt:
* platform/mac/css2.1/t100801-c548-ln-ht-02-b-ag-expected.txt:
* platform/mac/css2.1/t1507-c526-font-sz-02-b-a-expected.txt:
* platform/mac/css2.1/t1601-c547-indent-00-b-a-expected.txt:
* platform/mac/css2.1/t1604-c541-word-sp-00-b-a-expected.txt:
* platform/mac/css2.1/t1604-c542-letter-sp-00-b-a-expected.txt:
* platform/mac/editing/selection/caret-ltr-2-expected.txt:
* platform/mac/editing/selection/caret-ltr-2-left-expected.txt:
* platform/mac/editing/selection/caret-ltr-expected.txt:
* platform/mac/editing/selection/caret-ltr-right-expected.txt:
* platform/mac/editing/selection/caret-rtl-2-expected.txt:
* platform/mac/editing/selection/caret-rtl-2-left-expected.txt:
* platform/mac/editing/selection/caret-rtl-expected.txt:
* platform/mac/editing/selection/caret-rtl-right-expected.txt:
* platform/mac/fast/block/basic/011-expected.txt:
* platform/mac/fast/block/margin-collapse/103-expected.txt:
* platform/mac/fast/css/ex-after-font-variant-expected.txt:
* platform/mac/fast/css/non-standard-checkbox-size-expected.txt:
* platform/mac/fast/forms/001-expected.txt:
* platform/mac/fast/forms/basic-inputs-expected.txt:
* platform/mac/fast/forms/box-shadow-override-expected.txt:
* platform/mac/fast/forms/checkbox-radio-onchange-expected.txt:
* platform/mac/fast/forms/file-input-disabled-expected.txt:
* platform/mac/fast/forms/form-element-geometry-expected.txt:
* platform/mac/fast/forms/formmove-expected.txt:
* platform/mac/fast/forms/formmove2-expected.txt:
* platform/mac/fast/forms/indeterminate-expected.txt:
* platform/mac/fast/forms/input-appearance-height-expected.txt:
* platform/mac/fast/forms/input-value-expected.txt:
* platform/mac/fast/forms/minWidthPercent-expected.txt:
* platform/mac/fast/forms/radio-attr-order-expected.txt:
* platform/mac/fast/forms/radio-nested-labels-expected.txt:
* platform/mac/fast/forms/radio_checked-expected.txt:
* platform/mac/fast/forms/radio_checked_dynamic-expected.txt:
* platform/mac/fast/inline/positionedLifetime-expected.txt:
* platform/mac/fast/lists/008-expected.txt:
* platform/mac/fast/lists/008-vertical-expected.txt:
* platform/mac/fast/media/mq-relative-constraints-08-expected.txt:
* platform/mac/fast/overflow/overflow-rtl-expected.txt:
* platform/mac/fast/overflow/overflow-rtl-vertical-expected.txt:
* platform/mac/fast/parser/bad-xml-slash-expected.txt:
* platform/mac/fast/replaced/replaced-breaking-expected.txt:
* platform/mac/fast/replaced/width100percent-checkbox-expected.txt:
* platform/mac/fast/replaced/width100percent-radio-expected.txt:
* platform/mac/fast/text/textIteratorNilRenderer-expected.txt:
* platform/mac/fast/text/whitespace/normal-after-nowrap-breaking-expected.txt:
* platform/mac/fast/text/whitespace/pre-wrap-spaces-after-newline-expected.txt:
* platform/mac/http/tests/navigation/javascriptlink-frames-expected.txt:
* platform/mac/svg/custom/alignment-baseline-modes-expected.txt:
* platform/mac/svg/custom/dominant-baseline-modes-expected.txt:
* platform/mac/tables/mozilla/bugs/bug1318-expected.txt:
* platform/mac/tables/mozilla/bugs/bug4527-expected.txt:
* platform/mac/tables/mozilla_expected_failures/core/captions1-expected.txt:
* platform/mac/tables/mozilla_expected_failures/core/captions2-expected.txt:
* svg/custom/invalid-text-content-expected.png:
* svg/custom/text-zoom-expected.png:
* svg/custom/use-clipped-transform-expected.png:



git-svn-id: svn://svn.chromium.org/blink/trunk@80755 bbb929c8-8fbe-4397-9dbb-9b2b20218538

<http://webkit.org/b/56080> tracks the failure.

* platform/win-xp/Skipped: Added jquery/effects.html. (Ideally we'd only skip this in Debug
builds regardless of whether it's XP or Vista or 7, but doing that requires fixing
<http://webkit.org/b/55254>.)

git-svn-id: svn://svn.chromium.org/blink/trunk@80754 bbb929c8-8fbe-4397-9dbb-9b2b20218538

https://bugs.webkit.org/show_bug.cgi?id=56022

Reviewed by Oliver Hunt and Brian Weinstein.

Add a parameter to WKContextClearResourceCaches to specify whether to clear all the caches
or just the memory caches.

* Shared/ResourceCachesToClear.h: Added.
* UIProcess/API/C/WKAPICast.h:
(WebKit::toResourceCachesToClear):

* UIProcess/API/C/WKContext.cpp:
(WKContextClearResourceCaches):
Add the new parameter.
* UIProcess/API/C/WKContext.h:
* UIProcess/WebContext.cpp:
(WebKit::WebContext::clearResourceCaches):
Ditto.
* UIProcess/WebContext.h:

* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::clearResourceCaches):
Pass the new parameter through to the platform-specific functions.
* WebProcess/WebProcess.h:
Add the new parameter and give it a default of AllResourceCaches.
* WebProcess/WebProcess.messages.in:
Ditto.
* WebProcess/gtk/WebProcessGtk.cpp:
(WebKit::WebProcess::platformClearResourceCaches):
Ditto.
* WebProcess/qt/WebProcessQt.cpp:
(WebKit::WebProcess::platformClearResourceCaches):
Ditto.
* WebProcess/win/WebProcessWin.cpp:
(WebKit::WebProcess::platformClearResourceCaches):
When told to clear only the in-memory caches, return early so as not to clear the CFNetwork
disk cache.
* WebProcess/mac/WebProcessMac.mm:
(WebKit::WebProcess::platformClearResourceCaches):
Ditto.

* WebKit2.xcodeproj/project.pbxproj:
Add ResourceCachesToClear.h
* win/WebKit2.vcproj:
Ditto.



git-svn-id: svn://svn.chromium.org/blink/trunk@80753 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Tony Chang.

        VisiblePosition's next and previous should take an enum instead of a boolean
        https://bugs.webkit.org/show_bug.cgi?id=56135

        SelectionController::isAll now takes EditingBoundaryCrossingRule instead of
        StayInEditableContent.

        * WebView/WebView.mm:
        (-[WebView _selectionIsAll]):
2011-03-10  Ryosuke Niwa  <rniwa@webkit.org>

        Reviewed by Tony Chang.

        VisiblePosition's next and previous should take an enum instead of a boolean
        https://bugs.webkit.org/show_bug.cgi?id=56135

        Changed the argument type of VisiblePosition::next and VisiblePosition::previous
        from bool to EditingBoundaryCrossingRule. Also got rid of StayInEditableContent enum
        in VisibleSelection and replaced it by EditingBoundaryCrossingRule because the enum
        was only used in VisiblePosition::isAll and subsequently in SelectionController::isAll
        to call VisiblePosition::next and VisiblePosition::previous.

        * WebCore.exp.in:
        * dom/Position.cpp:
        (WebCore::Position::trailingWhitespacePosition):
        * editing/ApplyBlockElementCommand.cpp:
        (WebCore::ApplyBlockElementCommand::doApply):
        * editing/CompositeEditCommand.cpp:
        (WebCore::CompositeEditCommand::breakOutOfEmptyMailBlockquotedParagraph):
        * editing/InsertListCommand.cpp:
        (WebCore::InsertListCommand::doApply):
        (WebCore::InsertListCommand::listifyParagraph):
        * editing/ReplaceSelectionCommand.cpp:
        (WebCore::ReplaceSelectionCommand::shouldMergeStart):
        (WebCore::ReplaceSelectionCommand::shouldMergeEnd):
        (WebCore::ReplaceSelectionCommand::doApply):
        * editing/SelectionController.cpp:
        (WebCore::SelectionController::modifyExtendingRight):
        (WebCore::SelectionController::modifyExtendingForward):
        (WebCore::SelectionController::modifyMovingForward):
        (WebCore::SelectionController::modifyExtendingLeft):
        (WebCore::SelectionController::modifyExtendingBackward):
        (WebCore::SelectionController::modifyMovingBackward):
        * editing/SelectionController.h:
        (WebCore::SelectionController::isAll):
        * editing/TypingCommand.cpp:
        (WebCore::TypingCommand::deleteKeyPressed):
        (WebCore::TypingCommand::forwardDeleteKeyPressed):
        * editing/VisiblePosition.cpp:
        (WebCore::VisiblePosition::next):
        (WebCore::VisiblePosition::previous):
        * editing/VisiblePosition.h:
        * editing/VisibleSelection.cpp:
        (WebCore::VisibleSelection::isAll):
        (WebCore::VisibleSelection::setStartAndEndFromBaseAndExtentRespectingGranularity):
        * editing/VisibleSelection.h:
        * editing/htmlediting.cpp:
        (WebCore::selectionForParagraphIteration):
        * editing/visible_units.cpp:
        (WebCore::startOfNextParagraph):


git-svn-id: svn://svn.chromium.org/blink/trunk@80752 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Gavin Barraclough.

        Fix allocation of native function with a cached thunk
        https://bugs.webkit.org/show_bug.cgi?id=56127

        Fix this race condition found while fixing zombies.

        * collector/handles/HandleHeap.cpp:
        (JSC::HandleHeap::clearWeakPointers):
        * runtime/Heap.cpp:
        (JSC::Heap::reset):
        * runtime/JSFunction.cpp:
        (JSC::JSFunction::JSFunction):
        (JSC::JSFunction::markChildren):
        * runtime/JSValue.h:
        (JSC::JSValue::decode):
        * runtime/JSZombie.cpp:
        (JSC::JSZombie::leakedZombieStructure):
        * runtime/JSZombie.h:
        (JSC::JSZombie::createStructure):
        * runtime/MarkedBlock.cpp:

git-svn-id: svn://svn.chromium.org/blink/trunk@80751 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Tony Gentilcore.

        Remove CRASH() calls added to track down bug 53045
        https://bugs.webkit.org/show_bug.cgi?id=56137

        Remove CRASH() calls added by r76575 and re-label ones added by r80155
        and r80269 as being associated with bug 56124 (which may still be
        happening).

        * css/CSSImageValue.cpp:
        (WebCore::CSSImageValue::cachedImage):
        * css/CSSSelector.h:
        * css/CSSSelectorList.cpp:
        (WebCore::CSSSelectorList::deleteSelectors):
        * loader/cache/CachedResource.cpp:
        (WebCore::CachedResource::CachedResource):
        (WebCore::CachedResource::~CachedResource):
        * loader/cache/CachedResource.h:

git-svn-id: svn://svn.chromium.org/blink/trunk@80750 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Eric Seidel.

        Let the parser yield for layout before running scripts
        https://bugs.webkit.org/show_bug.cgi?id=54355

        Prior to this patch, the parser would yield to perform a layout/paint before running a
        script only if the script or a stylesheet blocking the script is not loaded yet. Since we
        don't preload scan into the body while parsing the head, typically we'll block on a script
        early in the body that causes us to yield to do the first paint within a reasonable time.

        However, I'm planning to change the PreloadScanner to scan into the body from the head.
        That significantly improves overall load time, but would hurt first paint time because
        fewer scripts would be blocked during parsing and thus wouldn't yield.

        This change causes us to yield before running scripts if we haven't painted yet (regardless
        of whether or not the script is loaded). In addition to allowing the above mentioned
        PreloadScanner change to be implemented without regressing first paint time, this also
        improves first paint time by itself.

        I tested Alexa's top 45 websites using Web Page Replay to control the content and simulate
        bandwidth. This patch improved average first paint time by 1% over an unlimited connection,
        6% over a 1Mbps connection and 11% over a 5Mbps connection. There was no statistically
        signifcant change in page load time.

        Within the pages tested, 33 had no statistically significant change in time to first paint,
        12 improved, and none regressed. Of the improved, some of the standouts from the 1Mbps set
        are: 20% on youtube, 37% on wiki, 27% on ebay, 13% on cnn, 16% on espn, 74% on sohu.

        * html/parser/HTMLDocumentParser.cpp:
        (WebCore::HTMLDocumentParser::canTakeNextToken): This is the new yield point.
        (WebCore::HTMLDocumentParser::pumpTokenizer): Remove ASSERT that we are not paused. isPaused
        means that we are waiting for a script. Bug 54574 changed pumpTokenizer() so that it does
        the right thing whether we are just before a token or waiting for a script. Now that we may
        yield before a token or before a script, this may be called while paused.
        * html/parser/HTMLParserScheduler.cpp:
        (WebCore::HTMLParserScheduler::checkForYieldBeforeScript): Added.
        * page/FrameView.h:
        (WebCore::FrameView::hasEverPainted): Added.

git-svn-id: svn://svn.chromium.org/blink/trunk@80749 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Fixes <http://webkit.org/b/55581> Test results don't get uploaded if test-result-archive
fails to delete the results directory

Reviewed by Mark Rowe.

* BuildSlaveSupport/test-result-archive:
(archiveTestResults): Ignore errno 2 when deleting the results directory, since it can be
thrown when a process is holding a file open.

git-svn-id: svn://svn.chromium.org/blink/trunk@80748 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Fixed the help message for WebGL support on GTK+ to reflect the fact that
        it's turned off by default and is experimental.

        * configure.ac: Fix WebGL help message.

git-svn-id: svn://svn.chromium.org/blink/trunk@80747 bbb929c8-8fbe-4397-9dbb-9b2b20218538

https://bugs.webkit.org/show_bug.cgi?id=33008

Reviewed by Andreas Kling.

Source/JavaScriptCore:

Defining WTF_USE_PTHREAD_BASED_QT=1 for platforms where QThread uses pthread internally.
Symbian is excluded because pthread_kill does not work on it. Mac is excluded because
it has its own ways to do JSC threading.

Defining WTF_USE_PTHREADS inside MachineStackMarker.cpp if USE(PTHREAD_BASED_QT) is true.

* runtime/MachineStackMarker.cpp:
* wtf/Platform.h:

LayoutTests:

Removing several tests from the Qt skip list. Those tests were flaky because of race conditions
related to web workers.

* platform/qt/Skipped:

git-svn-id: svn://svn.chromium.org/blink/trunk@80746 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Adam Roben.

        WKViewRegisterEditCommandCallback should use WKViewUndoType
        https://bugs.webkit.org/show_bug.cgi?id=56120

        * UIProcess/API/C/win/WKView.h: Change undoOrRedo parameter in WKViewRegisterEditCommandCallback to be WKViewUndoType.


git-svn-id: svn://svn.chromium.org/blink/trunk@80745 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Martin Robinson.

        Some Gtk code uses defined(USE_FREETYPE) instead of just USE(FREETYPE)
        https://bugs.webkit.org/show_bug.cgi?id=55996

        Use the macre USE instead of defined fro WTF_USE_FREETYPE and
        WTF_USE_PANGO.

        * GNUmakefile.am:
        * platform/graphics/cairo/OwnPtrCairo.cpp:
        * platform/graphics/cairo/OwnPtrCairo.h:
        * platform/graphics/cairo/RefPtrCairo.cpp:
        * platform/graphics/cairo/RefPtrCairo.h:
        * platform/graphics/gtk/FontGtk.cpp:
        (WebCore::setPangoAttributes):
        (WebCore::Font::drawComplexText):
        (WebCore::Font::floatWidthForComplexText):
        (WebCore::Font::offsetForPositionForComplexText):
        (WebCore::Font::selectionRectForComplexText):

git-svn-id: svn://svn.chromium.org/blink/trunk@80744 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Reviewed by Oliver Hunt.

There are three issues causing test failures in sputnik.

(1) lastIndex should be converted at the point it is used, not the point it is set (this is visible if valueOf is overridden).
(2) The 'length' property of the test/exec functions should be 1.
(3) If no input is specified, the input to test()/exec() is "undefined" (i.e. ToString(undefined)) - not RegExp.input.

Source/JavaScriptCore: 

* runtime/RegExpObject.cpp:
(JSC::RegExpObject::markChildren):
    - Added to mark lastIndex
(JSC::regExpObjectLastIndex):
(JSC::setRegExpObjectLastIndex):
    - lastIndex is now stored as a JSValue.
(JSC::RegExpObject::match):
    - Use accessor methods to get/set lastIndex, add fast case for isUInt32 (don't convert to double).
* runtime/RegExpObject.h:
(JSC::RegExpObject::setLastIndex):
(JSC::RegExpObject::setLastIndex):
    - Set lastIndex, either from a size_t or a JSValue.
(JSC::RegExpObject::getLastIndex):
    - Get lastIndex.
(JSC::RegExpObject::RegExpObjectData::RegExpObjectData):
    - Initialize as a JSValue.
* runtime/RegExpPrototype.cpp:
(JSC::RegExpPrototype::RegExpPrototype):
    - Add test/exec properties with length 1.
* runtime/StringPrototype.cpp:
(JSC::stringProtoFuncMatch):
(JSC::stringProtoFuncSearch):
    - Do not read RegExp.input if none is provided.
* tests/mozilla/js1_2/regexp/RegExp_input.js:
* tests/mozilla/js1_2/regexp/RegExp_input_as_array.js:
    - Update these tests (they relied on non-ES5 behaviour).

LayoutTests: 

* fast/js/kde/function_length-expected.txt:
* fast/js/kde/script-tests/function_length.js:
    - 'length' of RexExp.test/RexExp.exec is 1
* fast/js/regexp-caching-expected.txt:
* fast/js/regexp-test-null-string-expected.txt:
* fast/js/regexp-test-null-string.html:
    - test() is equivalent to test("undefined")
* sputnik/Conformance/15_Native_Objects/15.10_RegExp/15.10.6/15.10.6.2_RegExp.prototype.exec/S15.10.6.2_A11-expected.txt:
* sputnik/Conformance/15_Native_Objects/15.10_RegExp/15.10.6/15.10.6.2_RegExp.prototype.exec/S15.10.6.2_A1_T16-expected.txt:
* sputnik/Conformance/15_Native_Objects/15.10_RegExp/15.10.6/15.10.6.2_RegExp.prototype.exec/S15.10.6.2_A4_T11-expected.txt:
* sputnik/Conformance/15_Native_Objects/15.10_RegExp/15.10.6/15.10.6.3_RegExp.prototype.test/S15.10.6.3_A11-expected.txt:
* sputnik/Conformance/15_Native_Objects/15.10_RegExp/15.10.6/15.10.6.3_RegExp.prototype.test/S15.10.6.3_A1_T16-expected.txt:
    - These tests now pass!



git-svn-id: svn://svn.chromium.org/blink/trunk@80743 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Oliver Hunt.

        Rolled back in 80277 and 80280 with event handler layout test failures fixed.
        https://bugs.webkit.org/show_bug.cgi?id=55653        

        The failures were caused by a last minute typo: assigning to currentEvent
        instead of m_currentEvent.

        * JavaScriptCore.exp:
        * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:
        * JavaScriptCore.xcodeproj/project.pbxproj:
        * bytecompiler/BytecodeGenerator.cpp:
        * jit/JITOpcodes.cpp:
        * jit/JITOpcodes32_64.cpp:
        * runtime/Arguments.h:
        * runtime/JSActivation.cpp:
        * runtime/JSActivation.h:
        * runtime/JSCell.h:
        * runtime/JSGlobalObject.cpp:
        * runtime/JSGlobalObject.h:
        * runtime/JSObject.cpp:
        * runtime/JSStaticScopeObject.cpp:
        * runtime/JSStaticScopeObject.h:
        * runtime/JSVariableObject.h:
        * runtime/MarkedSpace.cpp:
        * runtime/MarkedSpace.h:
2011-03-10  Geoffrey Garen  <ggaren@apple.com>

        Reviewed by Oliver Hunt.

        Rolled back in 80277 and 80280 with event handler layout test failures fixed.
        https://bugs.webkit.org/show_bug.cgi?id=55653        

        The failures were caused by a last minute typo: assigning to currentEvent
        instead of m_currentEvent.

        * JSRun.cpp:
        * JSRun.h:
2011-03-10  Geoffrey Garen  <ggaren@apple.com>

        Reviewed by Oliver Hunt.

        Rolled back in 80277 and 80280 with event handler layout test failures fixed.
        https://bugs.webkit.org/show_bug.cgi?id=55653

        The failures were caused by a last minute typo: assigning to currentEvent
        instead of m_currentEvent.

        * WebCore.xcodeproj/project.pbxproj:
        * bindings/js/JSDOMGlobalObject.cpp:
        * bindings/js/JSDOMGlobalObject.h:
        * bindings/js/JSDOMWindowBase.cpp:
        * bindings/js/JSDOMWindowBase.h:
        * bindings/js/JSDOMWindowCustom.h:
        * bindings/js/JSWorkerContextBase.cpp:


git-svn-id: svn://svn.chromium.org/blink/trunk@80742 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Doing so can lead to hangs (or at least really slow execution).

Fixes <http://webkit.org/b/56109> run-webkit-tests sometimes times out on Windows XP Debug
(Tests) after fast/text/large-text-composed-char-dos.html fails (due to PrettyPatch
hanging?)

Reviewed by David Kilzer.

* PrettyPatch/PrettyPatch.rb:
(PrettyPatch.MAXIMUM_INTRALINE_DIFF_LINE_LENGTH): Added this new constant.
(PrettyPatch.DiffSection.initialize): Don't bother looking for intra-line diffs in lines
longer than the maximum length.

git-svn-id: svn://svn.chromium.org/blink/trunk@80741 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Reviewed by Dan Bernstein.

Add Snow Leopard ifdefs for the scaling by point size and then the division by unitsPerEm to the
translationsTransform applied to the results from CTFontGetVerticalTranslationsForGlyphs, since
this is done already on Lion and Leopard.

* platform/graphics/mac/FontMac.mm:
(WebCore::showGlyphsWithAdvances):



git-svn-id: svn://svn.chromium.org/blink/trunk@80740 bbb929c8-8fbe-4397-9dbb-9b2b20218538