Timeline
Aug 30, 2012:
- 11:31 PM Changeset in webkit [127234] by
-
- 6 edits in trunk
[Forms] Blur from field should reset typeahead buffer in multiple fields time input UI
https://bugs.webkit.org/show_bug.cgi?id=95525
Reviewed by Kent Tamura.
Source/WebCore:
This patch changes to reset typeahead timer used in DateTimeNumericFieldElement
on blur from field. It is intuitive that not merging keyboard inputs
with interleaving focus motion, e.g. sequence type "1", type Tab, type
Shift+Tab, type "2" should set field to "2" instead of "12".
This patch affects ports which enable both ENABLE_INPUT_TYPE_TIME and
ENABLE_INPUT_TYPE_TIME_MULTIPLE_FIELDS.
No new tests. This patch adds a test case into fast/forms/time-multiple-fields/time-multiple-fields-keyboard-events.html
- html/shadow/DateTimeNumericFieldElement.cpp:
(WebCore::DateTimeNumericFieldElement::didBlur): Added to reset typeahead timer, m_lastDigitCharTime.
- html/shadow/DateTimeNumericFieldElement.h:
(DateTimeNumericFieldElement): Added a declaration of didBlur().
LayoutTests:
This patch adds a new test case for checking blur event resets typeahead
buffer in multiple fields time input UI.
This patch affects ports which enable both ENABLE_INPUT_TYPE_TIME and
ENABLE_INPUT_TYPE_TIME_MULTIPLE_FIELDS.
- fast/forms/time-multiple-fields/time-multiple-fields-keyboard-events-expected.txt: Changed to include an expectation of new test case.
- fast/forms/time-multiple-fields/time-multiple-fields-keyboard-events.html: Added a new test case of blur event resets typeahead buffer.
- 10:54 PM Changeset in webkit [127233] by
-
- 2 edits in trunk/Source/WTF
AtomicString(ASCIILiteral) should not compile
https://bugs.webkit.org/show_bug.cgi?id=95413
Patch by Benjamin Poulain <bpoulain@apple.com> on 2012-08-30
Reviewed by Adam Barth.
- wtf/text/AtomicString.h:
(AtomicString): Declare the constructor from ASCIILiteral private to ensure it is
not used by accident.
- 9:53 PM Changeset in webkit [127232] by
-
- 1 copy in tags/Safari-537.8
New tag.
- 9:52 PM Changeset in webkit [127231] by
-
- 1 delete in tags/Safari-537.8
Delete tag.
- 9:51 PM Changeset in webkit [127230] by
-
- 3 edits in trunk/Source/WebCore
Unreviewed, followup build fix against r127228.
- dom/Notation.cpp:
(WebCore::Notation::cloneNode):
- dom/Notation.h:
(Notation):
- 9:51 PM Changeset in webkit [127229] by
-
- 1 copy in tags/Safari-537.8
New tag.
- 9:27 PM Changeset in webkit [127228] by
-
- 26 edits2 deletes in trunk
Unreviewed, rolling out r126127.
http://trac.webkit.org/changeset/126127
This breaks gobject bindings.
Source/WebCore:
- dom/Attr.cpp:
(WebCore::Attr::cloneNode):
- dom/Attr.h:
- dom/CDATASection.cpp:
(WebCore::CDATASection::cloneNode):
- dom/CDATASection.h:
(CDATASection):
- dom/Comment.cpp:
(WebCore::Comment::cloneNode):
- dom/Comment.h:
(Comment):
- dom/Document.cpp:
(WebCore::Document::cloneNode):
- dom/Document.h:
(Document):
- dom/DocumentFragment.cpp:
(WebCore::DocumentFragment::cloneNode):
- dom/DocumentFragment.h:
(DocumentFragment):
- dom/DocumentType.cpp:
(WebCore::DocumentType::cloneNode):
- dom/DocumentType.h:
(DocumentType):
- dom/Element.cpp:
(WebCore::Element::cloneNode):
- dom/Element.h:
(Element):
- dom/EntityReference.cpp:
(WebCore::EntityReference::cloneNode):
- dom/EntityReference.h:
(EntityReference):
- dom/Node.h:
(Node):
- dom/Node.idl:
- dom/ProcessingInstruction.cpp:
(WebCore::ProcessingInstruction::cloneNode):
- dom/ProcessingInstruction.h:
(ProcessingInstruction):
- dom/ShadowRoot.cpp:
(WebCore::ShadowRoot::cloneNode):
- dom/ShadowRoot.h:
(ShadowRoot):
- dom/Text.cpp:
(WebCore::Text::cloneNode):
- dom/Text.h:
(Text):
LayoutTests:
- fast/dom/shadow/shadowroot-clonenode-expected.txt: Removed.
- fast/dom/shadow/shadowroot-clonenode.html: Removed.
- 9:00 PM Changeset in webkit [127227] by
-
- 3 edits in trunk/Source/WebCore
Loading a worker script should not be O(n2)
https://bugs.webkit.org/show_bug.cgi?id=95518
Reviewed by Benjamin Poulain.
Previously, we would malloc a new buffer and memcpy the entire worker
script every time we got another packet of data from the network. This
patch uses StringBuilder to accumulate the buffer more efficiently.
- workers/WorkerScriptLoader.cpp:
(WebCore::WorkerScriptLoader::WorkerScriptLoader):
(WebCore::WorkerScriptLoader::didReceiveData):
(WebCore::WorkerScriptLoader::didFinishLoading):
(WebCore):
(WebCore::WorkerScriptLoader::script):
- workers/WorkerScriptLoader.h:
(WorkerScriptLoader):
- 8:48 PM Changeset in webkit [127226] by
-
- 10 edits2 adds in trunk
[Forms] Shift+Tab should focus the last field of multiple fields time input UI
https://bugs.webkit.org/show_bug.cgi?id=95168
Reviewed by Kent Tamura.
Source/WebCore:
This patch changes focus handling in DateTimeEditElement class for
multiple fields time input UI.
Current implementation controls focus in fields by DateTimeEditElement
and a focus field is tracked m_focusFieldIndex member variable. In
current implementation, it is hard to set focus to the last field when
it gets focus by Shift+Tab, because we don't know how DateTimeEditElement
gets focus.
In this new implementation, DateTimeEditElement doesn't control focus
rather FocusController class controls focus:
- We set focus to DateTimeFieldElement instead of HTMLInputElement which is shadow host of DateTimeEditElement for controlling focus by FocusController.
- Focus/blur events in DateTimeFieldElement are routed to HTMLInputElement by shadow DOM mechanism.
- TimeInputType class overrides HTMLINputElement::blur() and focus() functions for delegating blur/focus behavior to DateTimeEditElement by calling DateTimeEditElement::blurByOwner()/focusByOwner().
This patch affects only ports which enable both ENABLE_INPUT_TYPE_TIME
and ENABLE_INPUT_TYPE_TIME_MULTIPLE_FIELDS.
Tests: fast/forms/time-multiple-fields/time-multiple-fields-blur-and-focus-events.html
This patch also adds a new test case for Shift+Tab into fast/forms/time-multiple-fields/time-multiple-fields-keyboard-events.html.
- html/TimeInputType.cpp:
(WebCore::TimeInputType::DateTimeEditControlOwnerImpl::didFieldBlur): Added for the input element not to be matched "focus" pseudo class for removing focus ring.
(WebCore::TimeInputType::DateTimeEditControlOwnerImpl::didFieldFocus): Added for the input element to be matched "focus" pseudo class for adding focus ring.
(WebCore::TimeInputType::DateTimeEditControlOwnerImpl::editControlMouseFocus): Removed. DateTimeEditElement no longer set focus to HTMLInputElement.
(WebCore::TimeInputType::DateTimeEditControlOwnerImpl::focusAndSelectEditControlOwner): Removed. In this patch, settting focus to field is handled by DateTimeEditElement. We don't need to call HTMLInputElement::select() which has no effect for "time" input type.
(WebCore::TimeInputType::DateTimeEditControlOwnerImpl::isEditControlOwnerFocused): Revmoed. FocusController manages focus navigation in fields rather than DateTimeEditElement.
(WebCore::TimeInputType::blur): Added for delegating blur() action to DateTimeEditElement.
(WebCore::TimeInputType::focus): Added for delegating focus() action to DateTimeEditElement.
(WebCore::TimeInputType::handleDOMActivateEvent): Removed. HTMLInputElement for "time" input type no longer get focus.
(WebCore::TimeInputType::hasCustomFocusLogic): Added for asking FocusController to walk into shadow DOM tree on input type "time".
(WebCore::TimeInputType::isKeyboardFocusable): Changed for "input" element not to have focus.
(WebCore::TimeInputType::isMouseFocusable): Changed for "input" element not to have focus.
- html/TimeInputType.h:
(DateTimeEditControlOwnerImpl): Changed for new definition of DateTimeEditControlOwner.
(TimeInputType): Changed for new definition of DateTimeEditElement.
- html/shadow/DateTimeEditElement.cpp:
(WebCore::DateTimeEditElement::DateTimeEditElement): Removed dropped member variable m_focusFieldIndex.
(WebCore::DateTimeEditElement::blurByOwner): Added for blur() action asekd by owner.
(WebCore::DateTimeEditElement::didBlurFromField): Added for calling DateTimeEditControlOwner::didBlurFromControl().
(WebCore::DateTimeEditElement::didFocusOnField): Added for calling DateTimeEditControlOwner::didFocusOnControl().
(WebCore::DateTimeEditElement::fieldIndexOf): Added for mapping DateTimeFieldElement to field index.
(WebCore::DateTimeEditElement::focusByOwner): Added for focus() action asekd by owner.
(WebCore::DateTimeEditElement::focusFieldAt): Removed.
(WebCore::DateTimeEditElement::focusedFieldIndex): Added as helper function.
(WebCore::DateTimeEditElement::focusedField): Added.
(WebCore::DateTimeEditElement::focusOnNextField): Changed to pass a field to focusOnNext().
(WebCore::DateTimeEditElement::focusOnPreviousField): ditto.
(WebCore::DateTimeEditElement::focusAndSelectSpinButtonOwner): Changed for setting focus to the first field.
(WebCore::DateTimeEditElement::handleKeyboardEvent): Removed. Moved handling of Left/Right keys to DateTimeFieldElement. Removed Tab-key handling.
(WebCore::DateTimeEditElement::handleMouseEvent): Removed. DateTimeEdit doesn't handle mouse click since this patch.
(WebCore::DateTimeEditElement::layout): Changed to use focusFieldIndex() and DateTimeFieldElement::focus().
(WebCore::DateTimeEditElement::nextFieldIndex): Removed. focusOnNextField does samething and sets focus.
(WebCore::DateTimeEditElement::previousFieldIndex): Removed. focusOnPreviousField does samething and sets focus.
(WebCore::DateTimeEditElement::resetLayout): Changed to remove m_focusFieldIndex variable reference.
(WebCore::DateTimeEditElement::defaultEventHandler): Changed to remove focus navigation related events, blue/focus/mouse, handling.
(WebCore::DateTimeEditElement::shouldSpinButtonRespondToWheelEvents): Changed to check focus field instead of asking owner.
(WebCore::DateTimeEditElement::spinButtonStepDown): Changed to use focusField() instead of m_focusFieldIndex.
(WebCore::DateTimeEditElement::spinButtonStepUp): ditto.
(WebCore::DateTimeEditElement::updateUIState): ditto.
- html/shadow/DateTimeEditElement.h:
(EditControlOwner): Added didBlurFromField() and didFocusOnField() function declarations. Removed focusAndSelectEditControlOwner, it was used for set focus to HTMLInputElement, and isEditControlOwnerFocused.
(DateTimeEditElement):
- html/shadow/DateTimeFieldElement.cpp:
(WebCore::DateTimeFieldElement::defaultEventHandler): Changed to handle blur and focus events.
(WebCore::DateTimeFieldElement::defaultKeyboardEventHandler): Changed to handle Left/Right keys.
(WebCore::DateTimeFieldElement::didBlur): Added to call FieldOwner::didBlurFromField() to remove focus ring from owner element.
(WebCore::DateTimeFieldElement::didFocus): Added to call FieldOwner::didFocusOnField() to add focus ring to owner element.
(WebCore::DateTimeFieldElement::focusOnNextField): Added for moving focus to next field for DateTimeNumbericFieldElment
(WebCore::DateTimeFieldElement::isFocusable): Added to make DateTimeFieldElement focusable.
(WebCore::DateTimeFieldElement::supportsFocus): Added to make DateTimeFieldElement focusable.
- html/shadow/DateTimeFieldElement.h:
(FieldOwner): Added declarations for focusable handling.
(DateTimeFieldElement): ditto.
LayoutTests:
This patch adds focus and blur event handling test and a new test case
for Shift+Tab for multiple fields time input UI.
- time-multiple-fields-blur-and-focus-events-expected.txt: Added.
- time-multiple-fields-blur-and-focus-events.html: Added for blur and focus events handling.
- fast/forms/time-multiple-fields/time-multiple-fields-keyboard-events.html: Changed to add a test cases for Shift+Tab.
- 8:30 PM Changeset in webkit [127225] by
-
- 12 edits1 delete in trunk/Source
[chromium] Revert WebCompositorSupport to raw ptrs, make dtor protected
https://bugs.webkit.org/show_bug.cgi?id=95520
Reviewed by Darin Fisher.
Source/Platform:
WebPassOwnPtr<T> isn't quite usable from the chromium side - it needs some more work and isn't worth blocking
WebCompositorSupport for. Also, the d'tor for WebCompositorSupport needs to be protected, not private, so it can
be implemented.
- Platform.gypi:
- chromium/public/WebCompositorSupport.h:
(WebKit):
(WebKit::WebCompositorSupport::createLayerTreeView):
(WebKit::WebCompositorSupport::createLayer):
(WebKit::WebCompositorSupport::createContentLayer):
(WebKit::WebCompositorSupport::createExternalTextureLayer):
(WebKit::WebCompositorSupport::createIOSurfaceLayer):
(WebKit::WebCompositorSupport::createImageLayer):
(WebKit::WebCompositorSupport::createSolidColorLayer):
(WebKit::WebCompositorSupport::createVideoLayer):
(WebKit::WebCompositorSupport::createScrollbarLayer):
(WebKit::WebCompositorSupport::createAnimation):
(WebKit::WebCompositorSupport::createFloatAnimationCurve):
(WebKit::WebCompositorSupport::createTransformAnimationCurve):
(WebCompositorSupport):
- chromium/public/WebPassOwnPtr.h: Removed.
Source/WebCore:
Adopt the return value of WebCompositorSupport explicitly.
- platform/graphics/chromium/AnimationTranslationUtil.cpp:
(WebCore::createWebAnimation):
- platform/graphics/chromium/Canvas2DLayerBridge.cpp:
(WebCore::Canvas2DLayerBridge::Canvas2DLayerBridge):
- platform/graphics/chromium/DrawingBufferChromium.cpp:
(WebCore::DrawingBufferPrivate::DrawingBufferPrivate):
- platform/graphics/chromium/GraphicsLayerChromium.cpp:
(WebCore::GraphicsLayerChromium::GraphicsLayerChromium):
(WebCore::GraphicsLayerChromium::setContentsToImage):
(WebCore::GraphicsLayerChromium::updateLayerPreserves3D):
Source/WebKit/chromium:
Adopt the return value of WebCompositorSupport explicitly.
- src/LinkHighlight.cpp:
(WebKit::LinkHighlight::LinkHighlight):
(WebKit::LinkHighlight::startHighlightAnimation):
- src/WebMediaPlayerClientImpl.cpp:
(WebKit::WebMediaPlayerClientImpl::readyStateChanged):
- src/WebPluginContainerImpl.cpp:
(WebKit::WebPluginContainerImpl::setBackingTextureId):
(WebKit::WebPluginContainerImpl::setBackingIOSurfaceId):
- 6:25 PM Changeset in webkit [127224] by
-
- 16 edits in trunk/Source/WebCore
Replace more instances of += with StringBuilder
https://bugs.webkit.org/show_bug.cgi?id=95502
Reviewed by Darin Adler.
This patch removes many uses of WTF::String::operator+= in WebCore.
Many of these uses are inefficient because they cause us to allocate
and memcpy strings more times than necessary. In most cases, I've
replaced these inefficient patterns with StringBuilder.
This patch makes progress towards removing WTF::String::operator+= from
the project.
We can make cssText() more efficient by passing a single StringBuilder
instance along to the recursive calls, but I've left that for a later
patch.
- css/CSSBorderImageSliceValue.cpp:
(WebCore::CSSBorderImageSliceValue::customCssText):
- css/CSSFontFaceSrcValue.cpp:
(WebCore::CSSFontFaceSrcValue::customCssText):
- css/CSSFunctionValue.cpp:
(WebCore::CSSFunctionValue::customCssText):
- css/CSSGradientValue.cpp:
(WebCore::CSSLinearGradientValue::customCssText):
(WebCore::CSSRadialGradientValue::customCssText):
- css/CSSParser.cpp:
(WebCore::CSSParser::createKeyframe):
- css/CSSPrimitiveValue.cpp:
(WebCore::CSSPrimitiveValue::customCssText):
- css/CSSReflectValue.cpp:
(WebCore::CSSReflectValue::customCssText):
- css/CSSTimingFunctionValue.cpp:
(WebCore::CSSCubicBezierTimingFunctionValue::customCssText):
(WebCore::CSSStepsTimingFunctionValue::customCssText):
- css/StylePropertySet.cpp:
(WebCore::StylePropertySet::get4Values):
(WebCore::StylePropertySet::getLayeredShorthandValue):
(WebCore::StylePropertySet::getShorthandValue):
- fileapi/BlobURL.cpp:
(WebCore::BlobURL::createBlobURL):
- inspector/InspectorCSSAgent.cpp:
(WebCore::InspectorCSSAgent::SetPropertyTextAction::redo):
- inspector/InspectorClient.cpp:
(WebCore::InspectorClient::doDispatchMessageOnFrontendPage):
- inspector/InspectorConsoleAgent.cpp:
(WebCore::InspectorConsoleAgent::didFailLoading):
- inspector/InspectorFileSystemAgent.cpp:
(WebCore):
- inspector/InspectorPageAgent.cpp:
(WebCore::InspectorPageAgent::cachedResourceContent):
(WebCore::InspectorPageAgent::getCookies):
- inspector/InspectorStyleSheet.cpp:
(WebCore::InspectorStyleSheet::addRule):
- 6:19 PM Changeset in webkit [127223] by
-
- 2 edits in trunk/LayoutTests
Unreviewed gardening. fast/events/domactivate-sets-underlying-click-event-as-handled.html fails on Windows.
https://bugs.webkit.org/show_bug.cgi?id=87469
It looks like any test that involves popup windows of some sort confuses DRT.
It doesn't know to continue the test when focus moves away from the main window.
This happens with select menu tests as well when focus is put on the select options popup.
- platform/win/Skipped:
- 6:15 PM Changeset in webkit [127222] by
-
- 4 edits in trunk
ASSERTION FAILURE in JSC::JSGlobalData::float32ArrayDescriptor when running fast/js/dfg-float64array.html
https://bugs.webkit.org/show_bug.cgi?id=95398
Reviewed by Mark Hahnenberg.
Source/JavaScriptCore:
Trying to get the build failure to be a bit more informative.
- runtime/JSGlobalData.h:
(JSGlobalData):
LayoutTests:
Temporarily unskipping tests to figure out what is going.
- platform/mac-wk2/Skipped:
- 6:14 PM Changeset in webkit [127221] by
-
- 3 edits3 copies in branches/chromium/1229
Merge 126839 - REGRESSION(r109480): Form state for iframe content is not restored
https://bugs.webkit.org/show_bug.cgi?id=90870
Reviewed by Jochen Eisinger.
Source/WebCore:
Since r109480, we have restored form state only for documents
loaded by FrameLoader::loadItem(). However we should restore form
state for documents in sub-frames of documents loaded by
FrameLoader::loadItem().
Test: fast/loader/form-state-restore-with-frames.html
- history/HistoryItem.cpp:
(WebCore::HistoryItem::isAncestorOf):
Added. A function to search descendants for the specified
HistoryItem. This is used by isAssociatedToRequestedHistoryItem().
- history/HistoryItem.h:
(HistoryItem): Declare isAncestorOf().
- loader/HistoryController.cpp:
(WebCore::HistoryController::saveDocumentState):
Don't save form state for detached document.
This is needed because saveDocumentState() is called twice; before
document detach and after document detach. We need to avoid the
latter call because formElementsState() for a detached document
produces an empty state.
(WebCore::isAssociatedToRequestedHistoryItem):
Added. This function checks the current HistoryItem is associated
to the HistoryItem specified to FrameLoader::loadItem().
(WebCore::HistoryController::restoreDocumentState):
Uses isAssociatedToRequestedHistoryItem().
LayoutTests:
- fast/loader/form-state-restore-with-frames.html: Added.
- fast/loader/form-state-restore-with-frames-expected.txt: Added.
- fast/loader/resources/form-state-restore-with-frames-1.html: Added.
TBR=tkent@chromium.org
Review URL: https://chromiumcodereview.appspot.com/10918017
- 6:13 PM Changeset in webkit [127220] by
-
- 7 edits10 adds in trunk
Make it possible to use CSS Variables inside Calc expressions.
https://bugs.webkit.org/show_bug.cgi?id=95284
Reviewed by Tony Chang.
Allows calc expressions to contain unevaluated variables, which are then resolved in StyleResolver.cpp when building the RenderStyle tree.
Tests:
fast/css/variables/calc.html
- css/CSSCalculationValue.cpp:
(WebCore::unitCategory):
(WebCore):
(WebCore::CSSCalcValue::customSerializeResolvingVariables):
Generates a CSS expression with variables resolved into their corresponding values.
(WebCore::CSSCalcValue::hasVariableReference):
Returns true if the calculation's expression tree refers to a variable (that needs to be resolved).
(CSSCalcPrimitiveValue):
(WebCore::CSSCalcPrimitiveValue::serializeResolvingVariables):
Resolves the variable using the underlying CSSPrimitiveValue's serializeResolvingVariables function.
(WebCore::CSSCalcPrimitiveValue::hasVariableReference):
(WebCore::CSSCalcPrimitiveValue::toCalcValue):
(WebCore::CSSCalcPrimitiveValue::doubleValue):
(WebCore::CSSCalcPrimitiveValue::computeLengthPx):
(WebCore::CSSCalcBinaryOperation::create):
(CSSCalcBinaryOperation):
(WebCore::CSSCalcBinaryOperation::serializeResolvingVariables):
Builds a CSS expression for contained subtrees.
(WebCore::CSSCalcBinaryOperation::hasVariableReference):
Returns true if either subtree contains a variable.
- css/CSSCalculationValue.h:
(CSSCalcExpressionNode):
(CSSCalcValue):
- css/CSSGrammar.y:
- css/CSSParser.cpp:
(WebCore::CSSParser::validCalculationUnit):
- css/CSSPrimitiveValue.cpp:
(WebCore::CSSPrimitiveValue::primitiveType):
(WebCore::CSSPrimitiveValue::customSerializeResolvingVariables):
- css/StyleResolver.cpp:
(WebCore::StyleResolver::collectMatchingRulesForList):
- 5:49 PM FeatureFlags edited by
- Add COMPUTED_GOTO_OPCODES (diff)
- 5:43 PM Changeset in webkit [127219] by
-
- 2 edits in trunk/Source/WebKit2
~JSNPObject should call invalidate() if it hasn't been called already
https://bugs.webkit.org/show_bug.cgi?id=95497
Reviewed by Geoffrey Garen.
Finalization is no longer eager, just like destruction, so the original intent behind
this ASSERT in ~JSNPObject is no longer relevant. Therefore, we can just call invalidate()
ourselves in ~JSNPObject.
- WebProcess/Plugins/Netscape/JSNPObject.cpp:
(WebKit::JSNPObject::~JSNPObject):
- 5:39 PM Changeset in webkit [127218] by
-
- 2 edits in trunk/LayoutTests
Unreviewed Gardening. http/security/referrer tests fail. Unable to load resources over https.
https://bugs.webkit.org/show_bug.cgi?id=73913
- platform/win/Skipped:
- 5:34 PM Changeset in webkit [127217] by
-
- 27 edits1 copy13 adds2 deletes in trunk
[CSS Shaders] Implement normal blend mode and source-atop compositing mode
https://bugs.webkit.org/show_bug.cgi?id=93869
Patch by Max Vujovic <mvujovic@adobe.com> on 2012-08-30
Reviewed by Dean Jackson.
Source/WebCore:
Instead of allowing direct texture access in an author's shader via u_texture, CSS
Shaders blends special symbols in the author's shader (css_MixColor and
css_ColorMatrix) with the DOM element texture.
The author specifies the blend mode and composite operator via the CSS mix
function like this:
-webkit-filter: custom(none mix(shader.fs normal source-atop));
This patch implements the normal blend mode and the source-atop composite
operator. The other blend modes and composite operators will come in later
patches.
This patch introduces a new class, CustomFilterValidatedProgram, which validates
the shader using ANGLE. If the shader uses blending and compositing,
CustomFilterValidatedProgram uses ANGLE's SH_CSS_SHADERS_SPEC flag. This allows
the author's shader to compile successfully with special symbols like
"css_MixColor". ANGLE also reserves the "css_" prefix. If the shader doesn't use
blending and compositing, CustomFilterValidatedProgram validates the shader using
ANGLE's SH_WEBGL_SPEC flag.
After validation, CustomFilterValidatedProgram adds blending, compositing, and
texture access shader code to the author's original shaders. The definitions for
css_MixColor and css_ColorMatrix are added before the author's fragment shader
code so that the author code can access them. The blending, compositing, and
texture access code is added after the author code and is thus inaccessible to the
author code. Since ANGLE reserves the "css_" prefix during the validation phase,
no collisions are possible between the author's code and the code that WebKit adds.
The CustomFilterGlobalContext now caches CustomFilterValidatedProgram instead
of CustomFilterCompiledProgram. CustomFilterValidatedProgram owns a
CustomFilterCompiledProgram. This way, we can use the platform-independent
CustomFilterValidatedProgram to validate and rewrite the shaders, regardless of
the platform representation of the program (e.g. CustomFilterCompiledProgram).
Tests: css3/filters/custom/custom-filter-color-matrix.html
css3/filters/custom/custom-filter-composite-source-atop.html
- GNUmakefile.list.am:
- Target.pri:
- WebCore.gyp/WebCore.gyp:
- WebCore.gypi:
- WebCore.vcproj/WebCore.vcproj:
- WebCore.xcodeproj/project.pbxproj:
- platform/graphics/ANGLEWebKitBridge.cpp:
(WebCore::ANGLEWebKitBridge::ANGLEWebKitBridge):
Add a shader spec parameter, since sometimes we want to validate the shader
against the CSS Shaders spec and other times we want to validate the shader
against the WebGL spec. Note that the CSS Shaders spec is treated as a subset
of the WebGL spec in ANGLE.
(WebCore::ANGLEWebKitBridge::validateShaderSource):
- platform/graphics/ANGLEWebKitBridge.h:
(ANGLEWebKitBridge):
- platform/graphics/filters/CustomFilterCompiledProgram.cpp:
(WebCore::CustomFilterCompiledProgram::CustomFilterCompiledProgram):
(WebCore::CustomFilterCompiledProgram::compileShader):
(WebCore::CustomFilterCompiledProgram::initializeParameterLocations):
(WebCore::CustomFilterCompiledProgram::~CustomFilterCompiledProgram):
- platform/graphics/filters/CustomFilterCompiledProgram.h:
(WebCore):
- platform/graphics/filters/CustomFilterGlobalContext.cpp:
(WebCore::CustomFilterGlobalContext::~CustomFilterGlobalContext):
(WebCore::CustomFilterGlobalContext::webglShaderValidator):
(WebCore):
(WebCore::CustomFilterGlobalContext::mixShaderValidator):
(WebCore::CustomFilterGlobalContext::createShaderValidator):
(WebCore::CustomFilterGlobalContext::getValidatedProgram):
(WebCore::CustomFilterGlobalContext::removeValidatedProgram):
- platform/graphics/filters/CustomFilterGlobalContext.h:
(WebCore):
(CustomFilterGlobalContext):
- platform/graphics/filters/CustomFilterProgramInfo.h:
(WebCore::CustomFilterProgramInfo::mixSettings):
- platform/graphics/filters/CustomFilterValidatedProgram.cpp: Added.
(WebCore):
(WebCore::CustomFilterValidatedProgram::defaultVertexShaderString):
(WebCore::CustomFilterValidatedProgram::defaultFragmentShaderString):
(WebCore::CustomFilterValidatedProgram::CustomFilterValidatedProgram):
(WebCore::CustomFilterValidatedProgram::compiledProgram):
(WebCore::CustomFilterValidatedProgram::rewriteMixVertexShader):
(WebCore::CustomFilterValidatedProgram::rewriteMixFragmentShader):
(WebCore::CustomFilterValidatedProgram::blendFunctionString):
(WebCore::CustomFilterValidatedProgram::compositeFunctionString):
(WebCore::CustomFilterValidatedProgram::~CustomFilterValidatedProgram):
- platform/graphics/filters/CustomFilterValidatedProgram.h: Added.
(WebCore):
(CustomFilterValidatedProgram):
(WebCore::CustomFilterValidatedProgram::create):
(WebCore::CustomFilterValidatedProgram::programInfo):
(WebCore::CustomFilterValidatedProgram::isInitialized):
(WebCore::CustomFilterValidatedProgram::detachFromGlobalContext):
- platform/graphics/filters/FECustomFilter.cpp:
(WebCore::FECustomFilter::FECustomFilter):
Accept a CustomFilterValidatedProgram instead of CustomFilterProgram.
(WebCore::FECustomFilter::create):
(WebCore::FECustomFilter::initializeContext):
(WebCore::FECustomFilter::bindVertexAttribute):
(WebCore::FECustomFilter::bindProgramAndBuffers):
- platform/graphics/filters/FECustomFilter.h:
(WebCore):
(FECustomFilter):
- rendering/FilterEffectRenderer.cpp:
(WebCore):
(WebCore::createCustomFilterEffect):
(WebCore::FilterEffectRenderer::build):
Only create an FECustomFilter if the program validates.
- rendering/FilterEffectRenderer.h:
(WebCore):
(FilterEffectRenderer):
LayoutTests:
We've added two new special built-in symbols in CSS Shaders, css_MixColor and
css_ColorMatrix.
This change adds custom-filter-composite-source-atop.html, a new test that uses css_MixColor
in a shader with the normal blend mode and the source-atop compositing operator.
This change also adds custom-filter-color-matrix.html, a new test that uses css_ColorMatrix
in the shader.
Authors can read and write to these special built-ins in their shader code. WebKit will
premultiply the DOM element texture with css_ColorMatrix. Additionally, WebKit will blend
and composite css_MixColor with the DOM element texture, according to the blend mode and
compositing operator that the author specifies in CSS.
For example, the following line of CSS tells WebKit how to blend and composite the
css_MixColor from shader.fs:
-webkit-filter: custom(none mix(shader.fs normal source-atop));
- css3/filters/custom/custom-filter-color-matrix-expected.png: Added.
- css3/filters/custom/custom-filter-color-matrix-expected.txt: Added.
- css3/filters/custom/custom-filter-color-matrix.html: Added.
- css3/filters/custom/custom-filter-composite-source-atop-expected.png: Added.
- css3/filters/custom/custom-filter-composite-source-atop-expected.txt: Added.
- css3/filters/custom/custom-filter-composite-source-atop.html: Added.
- css3/filters/custom/custom-filter-shader-cache.html:
- css3/filters/custom/effect-color-check.html:
Use pass-tex-coord.vs now. Since v_texCoord in not automatically passed in the default
shaders anymore, we have to pass it in our own vertex shader. After we implement more
blending and compositing modes, we will be able to convert most of the tests to use the
CSS mix function. Then, we won't be sampling the DOM element texture directly, so we
won't need a tex coord, so we won't need this shader anymore.
- css3/filters/custom/effect-custom.html:
The default CSS value for <fragmentShader> needs to be specified explicitly until we
change it from its current default. This will be fixed in:
https://bugs.webkit.org/show_bug.cgi?id=94020
- css3/filters/custom/effect-custom-transform-parameters.html:
Ditto.
- css3/filters/custom/filter-fallback-to-software.html:
Ditto.
- css3/filters/custom/invalid-custom-filter-shader-expected.html:
These tests will not attempt to apply a filter because the programs do not validate.
This means we don't need to trigger FilterEffectRenderer with grayscale(0) in
the reference file anymore.
- css3/filters/resources/composite.fs: Added.
- css3/filters/resources/empty-shader.fs: Added.
- css3/filters/resources/grayscale-color-matrix.fs: Added.
- css3/filters/resources/pass-tex-coord.vs: Added.
- platform/chromium/css3/filters/custom/custom-filter-color-matrix-expected.png: Added.
- platform/chromium/css3/filters/custom/custom-filter-composite-source-atop-expected.png: Added.
- platform/chromium/TestExpectations:
- platform/mac-snowleopard/css3/filters/custom/effect-custom-expected.png: Removed.
We should use the generic expectation, because this expectation does not show the
shaders applied.
- platform/mac-snowleopard/css3/filters/custom/effect-custom-parameters-expected.png: Removed.
Ditto.
- 5:21 PM Changeset in webkit [127216] by
-
- 2 edits in trunk/LayoutTests
Remove the failing expectation from some tests that are no longer failing on Lion.
- platform/mac-lion/TestExpectations:
- 5:04 PM Changeset in webkit [127215] by
-
- 2 edits in trunk/Source/WebKit2
Allow dynamic mach-lookup extensions in the WebProcess
<rdar://problem/12207996>
Reviewed by Gavin Barraclough.
Like we allow file read/write dynamic sandbox extensions, we should allow
mach-lookup extensions.
- WebProcess/com.apple.WebProcess.sb.in:
- 5:00 PM Changeset in webkit [127214] by
-
- 4 edits in trunk/Source/JavaScriptCore
Try to fix the Qt build: add some #includes that, for some reason, only the Qt linker requires.
- runtime/BooleanObject.cpp:
- runtime/ErrorInstance.cpp:
- runtime/NameInstance.cpp:
- 4:49 PM Changeset in webkit [127213] by
-
- 2 edits in trunk/Source/JavaScriptCore
Fix the Qt build: Removed a now-dead variable.
- interpreter/Interpreter.cpp:
(JSC::Interpreter::execute):
- 4:41 PM Changeset in webkit [127212] by
-
- 4 edits in trunk/Source
Ambiguous operator[] after r127191 on some compiler
https://bugs.webkit.org/show_bug.cgi?id=95509
Patch by Benjamin Poulain <bpoulain@apple.com> on 2012-08-30
Reviewed by Simon Fraser.
Source/JavaScriptCore:
On some compilers, the operator[] conflicts with the Obj-C++ operators. This attempts to solve
the issue.
- runtime/JSString.h:
(JSC::jsSingleCharacterSubstring):
(JSC::jsString):
(JSC::jsSubstring8):
(JSC::jsSubstring):
(JSC::jsOwnedString):
Source/WTF:
- wtf/text/WTFString.h:
(WTF::String::characterAt): At this as a synonym to operator[] to attempt a build fix.
- 4:36 PM Changeset in webkit [127211] by
-
- 2 edits in trunk/Source/JavaScriptCore
Try to fix the Qt build: Remove the inline keyword at the declaration
site.
The Qt compiler seems to be confused, complaining about these functions
not being defined in a translation unit, even though no generated code
in the unit calls these functions. Maybe removing the keyword at the
declaration site will change its mind.
This shouldn't change the inlining decision at all: the definition is
still inline.
- interpreter/CallFrame.h:
(ExecState):
- 4:33 PM Changeset in webkit [127210] by
-
- 2 edits in trunk/Source/JavaScriptCore
Undo Qt build fix guess, since it breaks other builds.
- runtime/JSArray.h:
- 4:31 PM Changeset in webkit [127209] by
-
- 2 edits in trunk/LayoutTests
CSS 2.1 failure: margin-collapse-clear-012 fails
https://bugs.webkit.org/show_bug.cgi?id=80394
Mark fast/block/margin-collapse/empty-clear-blocks.html as an expected text failure on Mac,
not both an image and text failure, since the bots are only seeing it as a text failure.
- platform/mac/TestExpectations:
- 4:30 PM Changeset in webkit [127208] by
-
- 2 edits in trunk/Source/JavaScriptCore
Try to fix the Qt build: add an #include to JSArray.h, since
it's included by some of the files Qt complains about, and
some of is functions call the functions Qt complains about.
- runtime/JSArray.h:
- 4:23 PM Changeset in webkit [127207] by
-
- 2 edits in trunk/Source/JavaScriptCore
Second step toward fixing the Windows build: Add new symbols.
- JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:
- 4:22 PM Changeset in webkit [127206] by
-
- 3 edits2 adds in trunk
Crash in RenderTable::calcBorderEnd
https://bugs.webkit.org/show_bug.cgi?id=95487
Reviewed by Abhishek Arya.
Source/WebCore:
r126590 opened the window for a race condition in RenderObjectChildList::removeChildNode.
This is caused because willBeRemovedFromTree should be strictly following by the actual removal
and wasn't.
This race condition was caused by clearSelection() being called just after willBeRemovedFromTree,
which forced a section's cells recalc and would re-add the soon-to-be-removed child, causing the
crash.
Test: fast/table/crash-clearSelection-collapsed-borders.html
- rendering/RenderObjectChildList.cpp:
(WebCore::RenderObjectChildList::removeChildNode):
Moved the clearSeletion call before willBeRemovedFromTree. Added a warning about running code
after willBeRemovedFromTree and before removing the child from the tree.
LayoutTests:
- fast/table/crash-clearSelection-collapsed-borders-expected.txt: Added.
- fast/table/crash-clearSelection-collapsed-borders.html: Added.
- 4:07 PM Changeset in webkit [127205] by
-
- 2 edits in trunk/Source/JavaScriptCore
Try to fix the Qt build: add an #include.
- bytecode/GetByIdStatus.cpp:
- 4:03 PM Changeset in webkit [127204] by
-
- 2 edits in trunk/Source/JavaScriptCore
First step toward fixing the Windows build: Remove old symbols.
- JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:
- 3:57 PM Changeset in webkit [127203] by
-
- 2 edits in trunk/Tools
[Chromium-Android] Skip compositing/webgl and platform/chromium/virtual/threaded/compositing/webgl tests
https://bugs.webkit.org/show_bug.cgi?id=95400
Reviewed by Dirk Pranke.
WebGL is not enabled and should be skipped on Android (http://crbug.com/135877).
- Scripts/webkitpy/layout_tests/port/chromium_android.py:
(ChromiumAndroidPort.skipped_layout_tests):
- 3:50 PM Changeset in webkit [127202] by
-
- 120 edits1 add4 deletes in trunk/Source
Use one object instead of two for closures, eliminating ScopeChainNode
https://bugs.webkit.org/show_bug.cgi?id=95501
Reviewed by Filip Pizlo.
This patch removes ScopeChainNode, and moves all the data and related
functions that used to be in ScopeChainNode into JSScope.
Most of this patch is mechanical changes to use a JSScope* where we used
to use a ScopeChainNode*. I've only specifically commented about items
that were non-mechanical.
- runtime/Completion.cpp:
(JSC::evaluate):
- runtime/Completion.h: Don't require an explicit scope chain argument
when evaluating code. Clients never wanted anything other than the
global scope, and other arbitrary scopes probably wouldn't work
correctly, anyway.
- runtime/JSScope.cpp:
- runtime/JSScope.h:
(JSC::JSScope::JSScope): JSScope now requires the data we used to pass to
ScopeChainNode, so it can link itself into the scope chain correctly.
- runtime/JSWithScope.h:
(JSC::JSWithScope::create):
(JSC::JSWithScope::JSWithScope): JSWithScope gets an extra constructor
for specifically supplying your own scope chain. The DOM needs this
interface for setting up the scope chain for certain event handlers.
Other clients always just push the JSWithScope to the head of the current
scope chain.
../WebCore:
Mechanical changes to update for JSC interface changes.
../WebKit/mac:
Mechanical change to update for JSC interface change.
../WebKit/qt:
Mechanical change to update for JSC interface change.
- Api/qwebelement.cpp:
(QWebElement::evaluateJavaScript):
../WebKit2:
Mechanical changes to update for JSC interface change.
- 3:49 PM Changeset in webkit [127201] by
-
- 2 edits in trunk/Tools
[GTK] Tries to run empty string when calling generate-gtkdoc when not using jhbuild
https://bugs.webkit.org/show_bug.cgi?id=95499
Unreviewed. One more fix to avoid trying to run the empty string.
- Scripts/webkitdirs.pm:
(buildAutotoolsProject):
- 3:23 PM Changeset in webkit [127200] by
-
- 2 edits in trunk/LayoutTests
Unreviewed update to test expectations.
fast/css/font-weight-1.html stopped failing for IMAGE+TEXT and
started failing for just IMAGE.
- platform/chromium/TestExpectations:
- 3:21 PM Changeset in webkit [127199] by
-
- 22 edits in trunk/Source
Render unto #ifdef's that which belong to them.
https://bugs.webkit.org/show_bug.cgi?id=95482.
Patch by Mark Lam <mark.lam@apple.com> on 2012-08-30
Reviewed by Filip Pizlo.
Source/JavaScriptCore:
Refining / disambiguating between #ifdefs and adding some. For
example, ENABLE(JIT) is conflated with ENABLE(LLINT) in some places.
Also, we need to add ENABLE(COMPUTED_GOTO_OPCODES) to indicate that we
want interpreted opcodes to use COMPUTED GOTOs apart from ENABLE(LLINT)
and ENABLE(COMPUTED_GOTO_CLASSIC_INTERPRETER). Also cleaned up #ifdefs
in certain places which were previously incorrect.
- bytecode/CodeBlock.cpp:
(JSC):
(JSC::CodeBlock::bytecodeOffset):
- bytecode/CodeBlock.h:
(CodeBlock):
- bytecode/Opcode.h:
(JSC::padOpcodeName):
- config.h:
- dfg/DFGOperations.cpp:
- interpreter/AbstractPC.cpp:
(JSC::AbstractPC::AbstractPC):
- interpreter/CallFrame.h:
(ExecState):
- interpreter/Interpreter.cpp:
(JSC::Interpreter::~Interpreter):
(JSC::Interpreter::initialize):
(JSC::Interpreter::isOpcode):
(JSC::Interpreter::unwindCallFrame):
(JSC::getLineNumberForCallFrame):
(JSC::getCallerInfo):
(JSC::Interpreter::execute):
(JSC::Interpreter::executeCall):
(JSC::Interpreter::executeConstruct):
(JSC::Interpreter::privateExecute):
- interpreter/Interpreter.h:
(JSC::Interpreter::getOpcode):
(JSC::Interpreter::getOpcodeID):
(Interpreter):
- jit/HostCallReturnValue.h:
- jit/JITCode.h:
(JITCode):
- jit/JITExceptions.cpp:
- jit/JITExceptions.h:
- jit/JSInterfaceJIT.h:
- llint/LLIntData.h:
(JSC::LLInt::getOpcode):
- llint/LLIntEntrypoints.cpp:
(JSC::LLInt::getFunctionEntrypoint):
(JSC::LLInt::getEvalEntrypoint):
(JSC::LLInt::getProgramEntrypoint):
- llint/LLIntOffsetsExtractor.cpp:
(JSC::LLIntOffsetsExtractor::dummy):
- llint/LLIntSlowPaths.cpp:
(LLInt):
- runtime/JSGlobalData.cpp:
(JSC):
Source/WTF:
- wtf/Platform.h: Added ENABLE(COMPUTED_GOTO_OPCODES).
- 3:13 PM Changeset in webkit [127198] by
-
- 7 edits in trunk
Unreviewed, rolling out r127171.
http://trac.webkit.org/changeset/127171
https://bugs.webkit.org/show_bug.cgi?id=95505
testRunner does not need dumpWebNotificationCallbacks().
(Requested by jonlee on #webkit).
Patch by Sheriff Bot <webkit.review.bot@gmail.com> on 2012-08-30
Source/WebKit/mac:
- WebKit.exp:
Tools:
- DumpRenderTree/TestRunner.cpp:
(TestRunner::TestRunner):
(TestRunner::staticFunctions):
- DumpRenderTree/TestRunner.h:
(TestRunner):
- DumpRenderTree/mac/MockWebNotificationProvider.mm:
(-[MockWebNotificationProvider webView:didShowNotification:]):
(-[MockWebNotificationProvider webView:didClickNotification:]):
(-[MockWebNotificationProvider webView:didCloseNotifications:]):
- DumpRenderTree/mac/UIDelegate.mm:
(-[UIDelegate webView:decidePolicyForNotificationRequestFromOrigin:listener:]):
- 3:13 PM Changeset in webkit [127197] by
-
- 2 edits in trunk/LayoutTests
[EFL] Add tables/mozilla_expected_failures/ expectations back
https://bugs.webkit.org/show_bug.cgi?id=95483
Unreviewed gardening.
It was removed by mistake on r127152. But now I'm moving the tests
to TestExpectations since we want to get rid of Skipped.
Patch by Thiago Marcos P. Santos <thiago.santos@intel.com> on 2012-08-30
- platform/efl/TestExpectations:
- 3:01 PM Changeset in webkit [127196] by
-
- 3 edits in trunk/LayoutTests
Unreviewed gardening. Update flexitem-stretch-image.html expectations.
- platform/chromium/TestExpectations: Mark the test as flaky (passes on Linux and sporadically on Win).
- platform/efl/TestExpectations: Remove since the test is passing.
- 2:56 PM Changeset in webkit [127195] by
-
- 9 edits2 adds in trunk
[EFL][WK2] Add WebMemorySampler feature.
https://bugs.webkit.org/show_bug.cgi?id=91214
Patch by JungJik Lee <jungjik.lee@samsung.com> on 2012-08-30
Reviewed by Kenneth Rohde Christiansen.
.:
Set WebMemorySampler feature on in EFL port.
- Source/cmake/OptionsEfl.cmake:
- Source/cmake/WebKitFeatures.cmake:
- Source/cmakeconfig.h.cmake:
Source/JavaScriptCore:
WebMemorySampler collects Javascript stack and JIT memory usage in globalMemoryStatistics.
- PlatformEfl.cmake:
Source/WebKit2:
Add WebMemorySampler feature to EFL port. WebMemorySampler records memory usage of
WebProcess and UI Process and also it records application memory status in real time.
Included items on the result are JIT, JS heap, fastmalloc bytes and
application memory info from /proc/process_id/statm.
- PlatformEfl.cmake:
- Shared/linux/WebMemorySamplerLinux.cpp: Added.
(WebKit):
(ApplicationMemoryStats):
(WebKit::nextToken):
(WebKit::appendKeyValuePair):
(WebKit::sampleMemoryAllocatedForApplication):
(WebKit::WebMemorySampler::processName):
(WebKit::WebMemorySampler::sampleWebKit):
(WebKit::WebMemorySampler::sendMemoryPressureEvent):
- UIProcess/API/efl/ewk_context.cpp:
(_Ewk_Context::_Ewk_Context):
- 2:41 PM Changeset in webkit [127194] by
-
- 2 edits in trunk/Source/Platform
Chromium win build fix - fix typo in gypi
- Platform.gypi:
- 2:40 PM Changeset in webkit [127193] by
-
- 15 edits1 add in trunk
objc_msgSend and IMP should be cast appropriately before using
https://bugs.webkit.org/show_bug.cgi?id=95242
Reviewed by Benjamin Poulain.
Source/WebCore:
Use wtfObjcMsgSend and wtfCallIMP templates which do appropriate casts
to a function pointer with right types when calling objc_msgSend and an
IMP method directly.
No new tests because no functional changes.
- page/mac/EventHandlerMac.mm:
(WebCore::selfRetainingNSScrollViewScrollWheel):
- platform/mac/WebCoreObjCExtras.mm:
(deallocCallback):
Source/WebKit/mac:
Use wtfObjcMsgSend and wtfCallIMP templates which do appropriate casts
to a function pointer with right types when calling objc_msgSend and an
IMP method directly.
- WebCoreSupport/WebCachedFramePlatformData.h:
(WebCachedFramePlatformData::clear):
- WebCoreSupport/WebDeviceOrientationClient.mm:
(WebDeviceOrientationClient::getProvider):
- WebView/WebDelegateImplementationCaching.mm:
(CallDelegate):
(CallDelegateReturningBoolean):
(CallResourceLoadDelegateReturningBoolean):
(CallFormDelegate):
(CallFormDelegateReturningBoolean):
- WebView/WebHTMLView.mm:
(setCursor):
(setNeedsDisplayInRect):
Source/WebKit2:
Use wtfObjcMsgSend and wtfCallIMP templates which do appropriate casts
to a function pointer with right types when calling objc_msgSend and an
IMP method directly.
- UIProcess/API/mac/PDFViewController.mm:
(WebKit::PDFViewScrollView_scrollWheel):
Source/WTF:
Add new templates wtfObjcMsgSend and wtfCallIMP that do the appropriate
casts to correctly typed function pointers before calling objc_msgSend
and IMP methods directly.
- wtf/Functional.h:
(WTF::R): Use wtfObjcMsgSend.
- wtf/ObjcRuntimeExtras.h: Added.
(wtfObjcMsgSend):
(wtfCallIMP):
Tools:
Use wtfObjcMsgSend and wtfCallIMP templates which do appropriate casts
to a function pointer with right types when calling objc_msgSend and an
IMP method directly.
- DumpRenderTree/mac/DumpRenderTree.mm:
(drt_NSFontManager_availableFontFamilies):
- WebKitTestRunner/InjectedBundle/mac/ActivateFonts.mm:
(WTR::wtr_NSFontManager_availableFontFamilies):
- 2:38 PM Changeset in webkit [127192] by
-
- 2 edits in trunk/LayoutTests
CSS 2.1 failure: margin-collapse-clear-012 fails
https://bugs.webkit.org/show_bug.cgi?id=80394
Mark fast/block/float/024.hml as an expected text failure on Mac, not both an image and text
failure, since the bots are only seeing it as a text failure.
- platform/mac/TestExpectations:
- 2:23 PM Changeset in webkit [127191] by
-
- 296 edits7 deletes in trunk
Replace JSC::UString by WTF::String
https://bugs.webkit.org/show_bug.cgi?id=95271
Patch by Benjamin Poulain <bpoulain@apple.com> on 2012-08-30
Reviewed by Geoffrey Garen.
Source/JavaScriptCore:
Having JSC::UString and WTF::String increase the complexity of working on WebKit, and
add useless conversions in the bindings. It also cause some code bloat.
The performance advantages of UString have been ported over in previous patches. This patch
is the last step: getting rid of UString.
In addition to the simplified code, this also reduce the binary size by 15kb on x86_64.
- API/OpaqueJSString.cpp:
(OpaqueJSString::ustring):
- runtime/Identifier.h:
(JSC::Identifier::ustring):
To avoid changing everything at once, the function named ustring() were kept as is. They
will be renamed in a follow up patch.
- runtime/JSString.h:
(JSC::JSString::string):
(JSC::JSValue::toWTFString):
(JSC::inlineJSValueNotStringtoString):
(JSC::JSValue::toWTFStringInline):
Since JSValue::toString() already exist (and return the JSString), the direct accessor is renamed
to ::toWTFString(). We may change ::string() to ::jsString() and ::toWTFString() to ::toString()
in the future.
- runtime/StringPrototype.cpp:
(JSC::substituteBackreferencesSlow): Replace the use of UString::getCharacters<>() by String::getCharactersWithUpconvert<>().
Source/WebCore:
Update the code to use String instead of UString.
On x86_64, this reduces the binary size by 22kb.
Since it is no longer possible to differenciate JSC::jsString() and WebCore::jsString() by the input
types, WebCore::jsString() is renated to WebCore::jsStringWithCache().
Since the cache is using a PtrHash, JSC::jsString() is used in place of the old WebCore::jsString() when
the string is generated locally. This is because the cache can never match in those cases.
Source/WebKit/blackberry:
Replace UString by String.
- WebCoreSupport/ClientExtension.cpp:
- WebCoreSupport/PagePopupBlackBerry.cpp:
(WebCore::PagePopupBlackBerry::installDomFunction):
Source/WebKit/efl:
Replace UString by String.
- WebCoreSupport/DumpRenderTreeSupportEfl.cpp:
(DumpRenderTreeSupportEfl::sendWebIntentResponse):
- ewk/ewk_frame.cpp:
(ewk_frame_script_execute):
Source/WebKit/gtk:
Replace UString by String.
- gdom/ConvertToGCharPrivate.h:
(copyAsGchar):
Source/WebKit/mac:
Get rid of UString, replace it by String, and simplify the code when possible.
On x86_64, this reduces the binary size by 7kb.
- Plugins/Hosted/NetscapePluginHostProxy.mm:
(identifierFromIdentifierRep):
- Plugins/Hosted/NetscapePluginInstanceProxy.mm:
(WebKit::NetscapePluginInstanceProxy::addValueToArray):
(WebKit::NetscapePluginInstanceProxy::moveGlobalExceptionToExecState):
- Plugins/Hosted/ProxyInstance.mm:
(WebKit::ProxyRuntimeMethod::create):
(WebKit::ProxyRuntimeMethod::finishCreation):
(WebKit::ProxyInstance::getPropertyNames):
(WebKit::ProxyInstance::methodsNamed):
(WebKit::ProxyInstance::fieldNamed):
- WebView/WebFrame.mm:
(-[WebFrame _stringByEvaluatingJavaScriptFromString:forceUserGesture:]):
(-[WebFrame _stringByEvaluatingJavaScriptFromString:withGlobalObject:inScriptWorld:]):
- WebView/WebScriptDebugDelegate.mm:
(-[WebScriptCallFrame functionName]):
(-[WebScriptCallFrame evaluateWebScript:]):
- WebView/WebScriptDebugger.h:
(WTF):
(JSC):
(WebScriptDebugger):
- WebView/WebScriptDebugger.mm:
(toNSURL):
(WebScriptDebugger::sourceParsed):
- WebView/WebView.mm:
(aeDescFromJSValue):
Source/WebKit/qt:
Replace UString by String.
- Api/qwebelement.cpp:
(QWebElement::evaluateJavaScript):
Source/WebKit/win:
Replace UString by String.
- WebFrame.cpp:
(WebFrame::stringByEvaluatingJavaScriptInScriptWorld):
- WebView.cpp:
(WebView::stringByEvaluatingJavaScriptFromString):
Source/WebKit/wx:
Update the #includes to use the correct types.
- WebFrame.cpp:
- WebView.cpp:
Source/WebKit2:
Update to code to switch from UString to String.
- WebProcess/Plugins/Netscape/JSNPMethod.cpp:
(WebKit::JSNPMethod::finishCreation):
- WebProcess/Plugins/Netscape/JSNPMethod.h:
(WebKit::JSNPMethod::create):
(JSNPMethod):
- WebProcess/Plugins/Netscape/JSNPObject.cpp:
(WebKit::npIdentifierFromIdentifier):
- WebProcess/Plugins/Netscape/NPRuntimeObjectMap.cpp:
(WebKit::NPRuntimeObjectMap::evaluate):
(WebKit::NPRuntimeObjectMap::moveGlobalExceptionToExecState):
Source/WTF:
- wtf/Platform.h: Useless edit to force a full build. This is needed for some bots for some reason.
- wtf/text/WTFString.h: Export a symbol that was exported on UString and needed in WebCore.
Add String::getCharactersWithUpconvert<>(), which is similar to String::getCharacters<>() but with the same
behaviors as UString::getCharacters<>().
String::getCharactersWithUpconvert<>() is useful when manipulating multiple strings, it allow writting code
using 16bits characters if any of the input String is not 8bit.
Tools:
Get rid of UString.
- DumpRenderTree/efl/WorkQueueItemEfl.cpp:
- gdb/webkit.py:
(WTFStringPrinter.to_string):
(JSCIdentifierPrinter.to_string):
(JSCJSStringPrinter.to_string):
(add_pretty_printers):
Websites/webkit.org:
Update the coding style to avoid mentioning a class that no longer exist.
- coding/coding-style.html:
- 2:17 PM Changeset in webkit [127190] by
-
- 7 edits3 adds in trunk
Tools: nrwt should have TestExpectations everywhere we have Skipped files for apple ports
https://bugs.webkit.org/show_bug.cgi?id=95495
Reviewed by Ojan Vafai.
This is a follow-on to bug 95370 that adds in support for
platform/wk2/TestExpectations and
platform/{mac,win}-$version/TestExpectations.
- Scripts/webkitpy/layout_tests/port/apple.py:
(ApplePort.expectations_files):
- Scripts/webkitpy/layout_tests/port/base.py:
(Port._skipped_file_search_paths):
- Scripts/webkitpy/layout_tests/port/mac_unittest.py:
(MacTest.test_expectations_files):
- Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
(MockDRTPortTest.make_port):
- Scripts/webkitpy/layout_tests/port/win_unittest.py:
(WinPortTest.test_expectations_files):
LayoutTests: Remove uses of ClassInfo in StrictEq and CompareEq in the DFG
https://bugs.webkit.org/show_bug.cgi?id=93401
Patch by Mark Hahnenberg <mhahnenberg@apple.com> on 2012-08-24
Reviewed by Filip Pizlo.
New test to make sure the DFG watchpoint works correctly for these cases.
- fast/js/document-all-triggers-masquerades-watchpoint-expected.txt: Added.
- fast/js/document-all-triggers-masquerades-watchpoint.html: Added.
- fast/js/script-tests/document-all-triggers-masquerades-watchpoint.js: Added.
(f):
- 2:11 PM Changeset in webkit [127189] by
-
- 9 edits3 adds in trunk
Remove uses of ClassInfo in StrictEq and CompareEq in the DFG
https://bugs.webkit.org/show_bug.cgi?id=93401
Reviewed by Filip Pizlo.
Source/JavaScriptCore:
Another incremental step in removing the dependence on ClassInfo pointers in object headers.
- bytecode/SpeculatedType.h:
(JSC::isCellOrOtherSpeculation):
(JSC):
- dfg/DFGAbstractState.cpp: Updated the CFA to reflect the changes to the backend.
(JSC::DFG::AbstractState::execute):
- dfg/DFGNode.h:
(Node):
(JSC::DFG::Node::shouldSpeculateString): Added this new function since it was conspicuously absent.
(JSC::DFG::Node::shouldSpeculateNonStringCellOrOther): Also add this function for use in the CFA.
- dfg/DFGSpeculativeJIT.cpp: Refactored how we handle CompareEq and CompareStrictEq in the DFG. We now just
check for Strings by comparing the object's Structure to the global Structure for strings. We only
check for MasqueradesAsUndefined if the watchpoint has fired. These changes allow us to remove our
uses of the ClassInfo pointer for compiling these nodes.
(JSC::DFG::SpeculativeJIT::compilePeepHoleObjectEquality):
(JSC::DFG::SpeculativeJIT::compilePeepHoleBranch):
(JSC::DFG::SpeculativeJIT::compare):
(JSC::DFG::SpeculativeJIT::compileStrictEq):
- dfg/DFGSpeculativeJIT.h:
(SpeculativeJIT):
- dfg/DFGSpeculativeJIT32_64.cpp: Same changes for 32 bit as for 64 bit.
(JSC::DFG::SpeculativeJIT::compileObjectEquality):
(JSC::DFG::SpeculativeJIT::compileObjectToObjectOrOtherEquality):
(JSC::DFG::SpeculativeJIT::compilePeepHoleObjectToObjectOrOtherEquality):
- dfg/DFGSpeculativeJIT64.cpp:
(JSC::DFG::SpeculativeJIT::compileObjectEquality):
(JSC::DFG::SpeculativeJIT::compileObjectToObjectOrOtherEquality):
(JSC::DFG::SpeculativeJIT::compilePeepHoleObjectToObjectOrOtherEquality):
LayoutTests:
New test to make sure the DFG watchpoint works correctly for these cases.
- fast/js/document-all-triggers-masquerades-watchpoint-expected.txt: Added.
- fast/js/document-all-triggers-masquerades-watchpoint.html: Added.
- fast/js/script-tests/document-all-triggers-masquerades-watchpoint.js: Added.
(f):
- 2:06 PM Changeset in webkit [127188] by
-
- 3 edits1 delete in trunk/LayoutTests
Unreviewed rebaseline after r127163.
These tests need an image rebaseline for win to match the linux results.
- platform/chromium-win-xp/fast/block/margin-collapse: Removed.
- platform/chromium-win/fast/block/float/024-expected.png:
- platform/chromium-win/fast/block/margin-collapse/empty-clear-blocks-expected.png:
- 1:56 PM Changeset in webkit [127187] by
-
- 1 edit2 adds in trunk/LayoutTests
Add TestExpectations stubs for apple wk2 ports
Unreviewed, build fix.
This adds empty placeholder files so new-run-webkit-tests
doesn't get upset if they're missing.
- platform/mac-wk2/TestExpectations: Added.
- platform/win-wk2/TestExpectations: Added.
- 1:52 PM Changeset in webkit [127186] by
-
- 2 edits in trunk/Source/WTF
Vector::shrinkToFit should use realloc when suitable.
https://bugs.webkit.org/show_bug.cgi?id=94810
Reviewed by Benjamin Poulain.
Only tested on BlackBerry. So it is wrapped with PLATFORM(BLACKBERRY) in the mean time.
Use realloc to shrink buffer when inline buffer isn't involved and and canMoveWithMemcpy is true.
When running the test code attached to the bug, it gives 30-45% performance boost for the large blocks
(Every test cycle includes an extra pair of malloc/free, so the boost on shrinkToFit() is even bigger)
Performance impact on small blocks is not noticeable. (Tested on BlackBerry)
- wtf/Vector.h:
(WTF::VectorBufferBase::shouldReallocateBuffer):
(VectorBufferBase):
(WTF::VectorBufferBase::reallocateBuffer):
(VectorBuffer):
(WTF::VectorBuffer::shouldReallocateBuffer):
(WTF::VectorBuffer::reallocateBuffer):
(WTF::VectorBuffer::inlineBuffer):
(WTF::::shrinkCapacity):
- 1:50 PM Changeset in webkit [127185] by
-
- 3 edits in trunk/Source/JavaScriptCore
[BlackBerry] Implement IncrementalSweeper for PLATFORM(BLACKBERRY)
https://bugs.webkit.org/show_bug.cgi?id=95469
Reviewed by Rob Buis.
RIM PR# 200595.
Share most code with USE(CF) and implement timer-related methods
for PLATFORM(BLACKBERRY).
- heap/IncrementalSweeper.cpp:
(JSC):
(JSC::IncrementalSweeper::IncrementalSweeper):
(JSC::IncrementalSweeper::create):
(JSC::IncrementalSweeper::scheduleTimer):
(JSC::IncrementalSweeper::cancelTimer):
(JSC::IncrementalSweeper::doSweep):
- heap/IncrementalSweeper.h:
(IncrementalSweeper):
- 1:40 PM Changeset in webkit [127184] by
-
- 2 edits in trunk/LayoutTests
flexitem-stretch-image.html is flaky
https://bugs.webkit.org/show_bug.cgi?id=95366
Remove the failing test expectations in platform/mac, since this is showing up as passing on
all the mac bots.
- platform/mac/TestExpectations:
- 1:34 PM Changeset in webkit [127183] by
-
- 2 edits in trunk/Source/WebKit/blackberry
[BlackBerry] ASSERT failure in JSC::MarkedAllocator::allocateSlowCase
https://bugs.webkit.org/show_bug.cgi?id=95492
Reviewed by Yong Li.
Patch by Jacky Jiang <zhajiang@rim.com>
PR: 200724
Hold the JSLock for the current thread before we call toRef to avoid
the ASSERT failure.
- Api/WebPage.cpp:
(BlackBerry::WebKit::WebPagePrivate::executeJavaScript):
- 1:30 PM Changeset in webkit [127182] by
-
- 1 edit3 adds in trunk/LayoutTests
Unreviewed update of test expectations.
These tests just needed platform-specific results.
- platform/chromium-mac-snowleopard/fast/dom/shadow/shadowdom-for-textarea-complex-shadow-expected.png: Added.
- platform/chromium-mac/fast/dom/shadow: Added.
- platform/chromium-mac/fast/dom/shadow/shadowdom-for-textarea-complex-shadow-expected.png: Added.
- 1:24 PM Changeset in webkit [127181] by
-
- 3 edits in trunk/Tools
executive.run_in_parallel() hangs if given nothing to do
https://bugs.webkit.org/show_bug.cgi?id=95387
Reviewed by Ojan Vafai.
Check to make sure that Executive.run_in_parallel() requires
a non-empty list of commands to execute; passing an empty list
seems surely like a programming error.
- Scripts/webkitpy/common/system/executive.py:
(Executive.run_in_parallel):
- Scripts/webkitpy/common/system/executive_unittest.py:
(ExecutiveTest.test_run_in_parallel_assert_nonempty):
- 1:22 PM Changeset in webkit [127180] by
-
- 2 edits in trunk/Tools
decide what (and how) we should set the tolerance for ref test pixel compares and test for that
https://bugs.webkit.org/show_bug.cgi?id=94746
Reviewed by Ojan Vafai.
Add an assertion to diff_image() to check that we are passing
tolerance=0 explicitly when diffing ref test results.
- Scripts/webkitpy/layout_tests/port/test.py:
(TestPort.diff_image):
- 1:17 PM Changeset in webkit [127179] by
-
- 2 edits in trunk/Source/JavaScriptCore
Fix broken classic intrpreter build.
https://bugs.webkit.org/show_bug.cgi?id=95484.
Patch by Mark Lam <mark.lam@apple.com> on 2012-08-30
Reviewed by Filip Pizlo.
- interpreter/Interpreter.cpp:
(JSC::Interpreter::privateExecute):
- 1:14 PM Changeset in webkit [127178] by
-
- 4 edits in trunk
[CMake] Suppress ANGLE compilation warnings
https://bugs.webkit.org/show_bug.cgi?id=95377
Reviewed by Antonio Gomes.
.:
Change WEBKIT_SET_EXTRA_COMPILER_FLAGS so it has an option to suppress C++ warnings.
- Source/cmake/WebKitHelpers.cmake:
Source/WebCore:
Compile ANGLE sources in a static library, and make sure the compile flags suppress warnings.
- CMakeLists.txt:
- 1:10 PM Changeset in webkit [127177] by
-
- 2 edits in trunk/LayoutTests
Unreviewed update of TestExpectations.
These tests are awaiting a rebaseline after r127168 but have the wrong modifiers on them. They
are failing for Image but have Image+Text. Updating to fix that.
- platform/chromium/TestExpectations:
- 1:06 PM Changeset in webkit [127176] by
-
- 4 edits in trunk
[Chromium] Layout Test media/track/track-cue-rendering-snap-to-lines-not-set.html is flaky
https://bugs.webkit.org/show_bug.cgi?id=89167
Reviewed by Eric Carlson.
Source/WebCore:
Fix for rendering tracks when snap-to-lines not set.
No new tests. Removed from TestExpectations existing test.
- html/track/TextTrackCue.cpp:
(WebCore::TextTrackCue::TextTrackCue):
(WebCore::TextTrackCue::calculateDisplayParameters): Updated the place
where m_computedLinePosition is determined.
LayoutTests:
Removed possible flaky test from TestExceptations.
- platform/chromium/TestExpectations:
- 1:02 PM Changeset in webkit [127175] by
-
- 4 edits in trunk/Source/WebKit/chromium
[chromium] Make webkit_compositor specific unit test compilation conditional on gyp var
https://bugs.webkit.org/show_bug.cgi?id=95401
Reviewed by Dirk Pranke.
If use_libcc_for_compositor is set, these tests are being compiled and run elsewhere.
- WebKit.gyp:
- WebKit.gypi:
- WebKitUnitTests.gyp:
- 1:01 PM Changeset in webkit [127174] by
-
- 5 edits in trunk/Tools
NRWT should look in mac-wk2 for a TestExpecations file
https://bugs.webkit.org/show_bug.cgi?id=95370
Reviewed by Ojan Vafai.
Adds support for mac-wk2 and win-wk2 to expectations_files()
for the apple mac and apple win ports.
- Scripts/webkitpy/layout_tests/port/apple.py:
(ApplePort.expectations_files):
- Scripts/webkitpy/layout_tests/port/mac_unittest.py:
(MacTest.test_expectations_files):
- Scripts/webkitpy/layout_tests/port/win_unittest.py:
(WinTest.test_expectations_files):
- Scripts/webkitpy/layout_tests/port/port_testcase.py:
(PortTestCase.test_expectations_files):
- 12:58 PM Changeset in webkit [127173] by
-
- 4 edits in trunk/LayoutTests
[CSS Shaders] Update css3/filters/custom/effect-custom-transform-parameters.html to remove anti-aliasing issues
https://bugs.webkit.org/show_bug.cgi?id=95407
Reviewed by Dean Jackson.
The initial test had a white border to prevent anti-aliasing effects, but that doesn't seem to
help on some platforms. Removed the border in this patch.
- css3/filters/custom/effect-custom-transform-parameters-expected.html:
- css3/filters/custom/effect-custom-transform-parameters.html:
- platform/mac-lion/TestExpectations: Un-skipping the test.
- 12:52 PM Changeset in webkit [127172] by
-
- 14 edits2 adds in trunk/Source
[chromium] Add CompositorSupport interface for constructing compositor classes
https://bugs.webkit.org/show_bug.cgi?id=95040
Reviewed by Darin Fisher.
Source/Platform:
Adds a WebCompositorSupport interface as a tear-off of PlatformSupport so the embedder can handle provide
implementations of compositor interfaces.
Adds a new WebPassOwnPtr<> type for use when the caller of an API must take ownership of the provided parameter.
- Platform.gypi:
- chromium/public/Platform.h:
(WebKit):
(WebKit::Platform::compositorSupport):
(Platform):
- chromium/public/WebCompositorSupport.h: Added.
(WebKit):
(WebCompositorSupport):
(WebKit::WebCompositorSupport::~WebCompositorSupport):
Source/WebCore:
Uses WebCompositorSupport interfaces where appropriate to construct compositor types.
- platform/graphics/chromium/AnimationTranslationUtil.cpp:
(WebCore::createWebAnimation):
- platform/graphics/chromium/Canvas2DLayerBridge.cpp:
(WebCore::Canvas2DLayerBridge::Canvas2DLayerBridge):
- platform/graphics/chromium/DrawingBufferChromium.cpp:
(WebCore::DrawingBufferPrivate::DrawingBufferPrivate):
- platform/graphics/chromium/GraphicsLayerChromium.cpp:
(WebCore::GraphicsLayerChromium::GraphicsLayerChromium):
(WebCore::GraphicsLayerChromium::setContentsToImage):
(WebCore::GraphicsLayerChromium::setContentsTo):
(WebCore::GraphicsLayerChromium::addAnimation):
(WebCore::GraphicsLayerChromium::updateLayerPreserves3D):
Source/WebKit/chromium:
- src/LinkHighlight.cpp:
(WebKit::LinkHighlight::LinkHighlight):
(WebKit::LinkHighlight::startHighlightAnimation):
- src/WebMediaPlayerClientImpl.cpp:
(WebKit::WebMediaPlayerClientImpl::readyStateChanged):
- src/WebMediaPlayerClientImpl.h:
(WebKit):
- src/WebPluginContainerImpl.cpp:
(WebKit::WebPluginContainerImpl::setBackingTextureId):
(WebKit::WebPluginContainerImpl::setBackingIOSurfaceId):
- src/WebPluginContainerImpl.h:
(WebKit):
- 12:46 PM Changeset in webkit [127171] by
-
- 7 edits in trunk
[Mac] Add testRunner.dumpWebNotificationCallbacks() to DRT
https://bugs.webkit.org/show_bug.cgi?id=95232
<rdar://problem/12190776>
Reviewed by Alexey Proskuryakov.
Source/WebKit/mac:
- WebKit.exp: Export WebNotification class.
Tools:
Add additional output when running test, if testRunner.dumpWebNotificationCallbacks() is called.
- DumpRenderTree/TestRunner.cpp:
(TestRunner::TestRunner):
(dumpWebNotificationCallbacksCallback):
(TestRunner::staticFunctions):
- DumpRenderTree/TestRunner.h:
(TestRunner):
(TestRunner::dumpWebNotificationCallbacks):
(TestRunner::setDumpWebNotificationCallbacks):
- DumpRenderTree/mac/MockWebNotificationProvider.mm:
(-[WebNotification _drt_descriptionSuitableForTestResult]):
(-[MockWebNotificationProvider webView:didShowNotification:]):
(-[MockWebNotificationProvider webView:didClickNotification:]):
(-[MockWebNotificationProvider webView:didCloseNotifications:]):
- DumpRenderTree/mac/UIDelegate.mm:
(-[UIDelegate webView:decidePolicyForNotificationRequestFromOrigin:listener:]):
- 12:46 PM Changeset in webkit [127170] by
-
- 5 edits1 copy in trunk/Source/WebCore
[CG] ImageCG.cpp contains a mix of Image and BitmapImage functions
https://bugs.webkit.org/show_bug.cgi?id=95470
Reviewed by Darin Adler.
- WebCore.gypi: Added BitmapImageCG.cpp.
- WebCore.vcproj/WebCore.vcproj: Ditto.
- WebCore.xcodeproj/project.pbxproj: Ditto.
- platform/graphics/cg/BitmapImageCG.cpp: Copied from Source/WebCore/platform/graphics/cg/ImageCG.cpp
then deleted Image function implementations.
- platform/graphics/cg/ImageCG.cpp: Deleted BitmapImage and FrameData function implementations.
- 12:05 PM Changeset in webkit [127169] by
-
- 1 edit2 adds in trunk/LayoutTests
Unreviewed rebaseline after r127101.
The Mac 10.6 results are slightly different--decoding about 10% less of the image.
After some discussion with Adam Barth we determined to rebaseline these.
- platform/chromium-mac-snowleopard/http/tests/images: Added.
- platform/chromium-mac-snowleopard/http/tests/images/webp-partial-load-expected.png: Added.
- 11:53 AM Changeset in webkit [127168] by
-
- 6 edits in trunk
[Chromium] The CC button is not painted
https://bugs.webkit.org/show_bug.cgi?id=95395
Reviewed by Eric Carlson.
Source/WebCore:
The actual Chromium resource for the CC button was not used by the Chromium theme.
Existing track rendering tests will be rebaselined and contain the new CC button.
- rendering/RenderMediaControlsChromium.cpp:
(WebCore::paintMediaClosedCaptionsButton): Added for proper painting of the resource.
(WebCore):
(WebCore::RenderMediaControlsChromium::paintMediaControlsPart):
Changed to call paintMediaClosedCaptionsButton when the control is the CC button.
- rendering/RenderThemeChromiumSkia.cpp:
(WebCore::RenderThemeChromiumSkia::paintMediaToggleClosedCaptionsButton):
Implemented proper behaviour.
(WebCore):
- rendering/RenderThemeChromiumSkia.h:
(RenderThemeChromiumSkia):
LayoutTests:
No new tests, the existing ones will need to be rebaselined and include the CC button.
- platform/chromium/TestExpectations: Marked tests failing accordingly for rebaselining.
- 11:45 AM Changeset in webkit [127167] by
-
- 2 edits in trunk/Source/JavaScriptCore
Build warning : -Wsign-compare on DFGByteCodeParser.cpp.
https://bugs.webkit.org/show_bug.cgi?id=95418
Patch by Byungwoo Lee <bw80.lee@samsung.com> on 2012-08-30
Reviewed by Filip Pizlo.
There is a build warning '-Wsign-compare' on
findArgumentPositionForLocal() in DFGByteCodeParser.cpp.
For removing this warning, casting statement is added explicitly.
- dfg/DFGByteCodeParser.cpp:
(JSC::DFG::ByteCodeParser::findArgumentPositionForLocal):
(JSC::DFG::ByteCodeParser::findArgumentPosition):
- 11:41 AM Changeset in webkit [127166] by
-
- 2 edits in trunk/LayoutTests
Assertion failure in MessagePort::contextDestroyed in
http/tests/security/MessagePort/event-listener-context.html, usually attributed to later tests.
https://bugs.webkit.org/show_bug.cgi?id=94458
fast/events/message-port-constructor-for-deleted-document.html also asserts in
MessagePort::contextDestroyed, and it is usually attributed to later tests as well. Skip it.
- platform/mac/TestExpectations:
- 11:30 AM Changeset in webkit [127165] by
-
- 23 edits4 copies1 add in trunk
MediaStream API: Introduce MediaConstraints
https://bugs.webkit.org/show_bug.cgi?id=95198
Reviewed by Adam Barth.
Source/Platform:
Adds WebMediaConstraints.
- Platform.gypi:
- chromium/public/WebMediaConstraints.h: Copied from Source/Platform/chromium/public/WebRTCPeerConnectionHandler.h.
(WebCore):
(WebKit):
(WebMediaConstraints):
(WebKit::WebMediaConstraints::WebMediaConstraints):
(WebKit::WebMediaConstraints::~WebMediaConstraints):
(WebKit::WebMediaConstraints::operator=):
- chromium/public/WebRTCPeerConnectionHandler.h:
(WebKit):
(WebRTCPeerConnectionHandler):
Source/WebCore:
This introduces MediaConstraints together with relevant infrastructure, a chromium mock and LayoutTests.
Patch covered by expanded existing tests.
- CMakeLists.txt:
- GNUmakefile.list.am:
- Modules/mediastream/MediaConstraintsImpl.cpp: Added.
(WebCore):
(WebCore::MediaConstraintsImpl::create):
(WebCore::MediaConstraintsImpl::initialize):
(WebCore::MediaConstraintsImpl::~MediaConstraintsImpl):
(WebCore::MediaConstraintsImpl::getMandatoryConstraintNames):
(WebCore::MediaConstraintsImpl::getOptionalConstraintNames):
(WebCore::MediaConstraintsImpl::getMandatoryConstraintValue):
(WebCore::MediaConstraintsImpl::getOptionalConstraintValue):
- Modules/mediastream/MediaConstraintsImpl.h: Copied from Source/WebCore/platform/mediastream/RTCPeerConnectionHandler.h.
(WebCore):
(MediaConstraintsImpl):
(WebCore::MediaConstraintsImpl::MediaConstraintsImpl):
- Modules/mediastream/RTCPeerConnection.cpp:
(WebCore::RTCPeerConnection::create):
(WebCore::RTCPeerConnection::RTCPeerConnection):
- Modules/mediastream/RTCPeerConnection.h:
(WebCore):
(RTCPeerConnection):
- WebCore.gypi:
- bindings/js/Dictionary.cpp:
(WebCore::Dictionary::getOwnPropertyNames):
(WebCore):
- bindings/js/Dictionary.h:
(Dictionary):
- bindings/v8/Dictionary.cpp:
(WebCore::Dictionary::getOwnPropertyNames):
(WebCore):
- bindings/v8/Dictionary.h:
(Dictionary):
- platform/chromium/support/WebMediaConstraints.cpp: Copied from Source/WebCore/platform/mediastream/RTCPeerConnectionHandler.cpp.
(WebKit):
(WebKit::WebMediaConstraints::WebMediaConstraints):
(WebKit::WebMediaConstraints::assign):
(WebKit::WebMediaConstraints::reset):
(WebKit::WebMediaConstraints::isNull):
(WebKit::WebMediaConstraints::getMandatoryConstraintNames):
(WebKit::WebMediaConstraints::getOptionalConstraintNames):
(WebKit::WebMediaConstraints::getMandatoryConstraintValue):
(WebKit::WebMediaConstraints::getOptionalConstraintValue):
- platform/mediastream/MediaConstraints.h: Copied from Source/WebCore/platform/mediastream/RTCPeerConnectionHandler.h.
(WebCore):
(MediaConstraints):
(WebCore::MediaConstraints::~MediaConstraints):
(WebCore::MediaConstraints::MediaConstraints):
- platform/mediastream/RTCPeerConnectionHandler.cpp:
(RTCPeerConnectionHandlerDummy):
(WebCore::RTCPeerConnectionHandlerDummy::initialize):
- platform/mediastream/RTCPeerConnectionHandler.h:
(WebCore):
(RTCPeerConnectionHandler):
- platform/mediastream/chromium/RTCPeerConnectionHandlerChromium.cpp:
(WebCore::RTCPeerConnectionHandlerChromium::initialize):
- platform/mediastream/chromium/RTCPeerConnectionHandlerChromium.h:
(RTCPeerConnectionHandlerChromium):
Tools:
Extending the MockWebRTCPeerConnectionHandler with MediaConstraints functionality.
- DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
(MockWebRTCPeerConnectionHandler::MockWebRTCPeerConnectionHandler):
(isSupportedConstraint):
(isValidConstraint):
(MockWebRTCPeerConnectionHandler::initialize):
- DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h:
(MockWebRTCPeerConnectionHandler):
LayoutTests:
- fast/mediastream/RTCPeerConnection-expected.txt:
- fast/mediastream/RTCPeerConnection.html:
- 11:27 AM Changeset in webkit [127164] by
-
- 3 edits in trunk/LayoutTests
Fixing two inspector tests that fail on Mac WK1.
- inspector/timeline/timeline-animation-frame-expected.txt:
- inspector/timeline/timeline-paint-expected.txt:
- 11:15 AM Changeset in webkit [127163] by
-
- 17 edits20 adds in trunk
CSS 2.1 failure: margin-collapse-clear-012 fails
https://bugs.webkit.org/show_bug.cgi?id=80394
Reviewed by David Hyatt.
Source/WebCore:
CSS2.1 states: "If the top and bottom margins of an element with clearance are adjoining, its margins collapse with
the adjoining margins of following siblings but that resulting margin does not collapse with the bottom margin of the parent block."
This is a change in the spec since http://trac.webkit.org/changeset/47678, so prevent the margins of collapsed blocks from collapsing
with parent margins.
Also, in the case of self-collapsing blocks adjust the clearance so that it is equal to [height of float to clear] - margin-top of the
self-collapsing block. (See the 'Explanation' section near the end of http://www.w3.org/TR/CSS21/visuren.html#clearance). This allows
the correct value of any margins collapsed with subsequent siblings to contribute to the height of the parent. For example if a block
with margin-top of 40px has to clear a float of 100px, the clearance is now 60px so set that as the height of the parent. If a subsequent
sibling has a collapsed margin value of 140px (e.g. from a margin-top of 80px and a margin-bottom of 140px) then the height of the parent
becomes 200px by adding on that collapsed margin.
Tests: css2.1/20110323/margin-collapse-clear-012.htm
css2.1/20110323/margin-collapse-clear-013.htm
css2.1/20110323/margin-collapse-027.htm
fast/css/margin-collapse-013-reduction.html
fast/css/margin-collapse-top-margin-clearance.html
fast/css/margin-collapse-top-margin-clearance-with-sibling.html
- rendering/RenderBlock.cpp:
(WebCore::RenderBlock::clearFloatsIfNeeded):
(WebCore::RenderBlock::handleAfterSideOfBlock): Without this margin-collapse/block-inside-inline/025.html adds in the margin to
an anonymous block containing a block child. FF does this, Opera does not. The spec is not crystal-clear but Opera's behaviour
seems correct - it's totally unintuitive to think the margins of an anonymous block and its block-flow child should be considered adjoining.
Maybe this is a FIXME pending clarification in the spec?
- rendering/RenderBlock.h:
(WebCore::RenderBlock::MarginInfo::setCanCollapseMarginAfterWithChildren):
LayoutTests:
- css2.1/20110323/margin-collapse-027-expected.html: Added.
- css2.1/20110323/margin-collapse-027.htm: Added.
- css2.1/20110323/margin-collapse-clear-012-expected.html: Added.
- css2.1/20110323/margin-collapse-clear-012.htm: Added.
- css2.1/20110323/margin-collapse-clear-013-expected.html: Added.
- css2.1/20110323/margin-collapse-clear-013.htm: Added.
- css2.1/20110323/margin-collapse-clear-014-expected.html: Added.
- css2.1/20110323/margin-collapse-clear-014.htm: Added.
- css2.1/20110323/margin-collapse-clear-015-expected.html: Added.
- css2.1/20110323/margin-collapse-clear-015.htm: Added.
- css2.1/20110323/margin-collapse-clear-016-expected.html: Added.
- css2.1/20110323/margin-collapse-clear-016.htm: Added.
- css2.1/20110323/margin-collapse-clear-017-expected.html: Added.
- css2.1/20110323/margin-collapse-clear-017.htm: Added.
- fast/block/margin-collapse/empty-clear-blocks.html: Modified the test to reflect the new expected result. The text in the third paragraph avoids the float rather than sitting underneath it. The text in the third last paragraph is 20 px below the float rather than immediately beneath it. Both of these are a consequence of the revised treatment of clearance when margin collapsing. This is consistent with the spec and FF and Opera.
- fast/css/margin-collapse-013-reduction.html: Added.
Sourced from https://bugzilla.mozilla.org/show_bug.cgi?id=493380
- fast/css/margin-collapse-top-margin-clearance-with-sibling-expected.html: Added.
- fast/css/margin-collapse-top-margin-clearance-with-sibling.html: Added.
Ensure we only avoid collapsing the margins of a self-collapsing block with the parent if we don't
have a subsequent sibling with height.
- fast/css/margin-collapse-top-margin-clearance.html: Added.
- platform/chromium-linux/fast/block/float/024-expected.png:
- platform/chromium-linux/fast/block/margin-collapse/empty-clear-blocks-expected.png: These two progress to the same rendering used by FF and Opera. It is a consequence of calculating clearance as "clearance = [height of float] - margin-top" when margin-top is a negative value.
- platform/chromium-win/fast/block/float/024-expected.txt:
- platform/chromium-win/fast/block/margin-collapse/025-expected.txt:
- platform/chromium-win/fast/block/margin-collapse/block-inside-inline/025-expected.txt:
The change in result for these three is a consequence of self-collapsing blocks with clearance moving their top up past the float by the value
of their top margin. Unlike the block-inside-inline case the 'clear' has no intrinsic height so the uncollapsed margin value has
no effect on its height.
- platform/chromium-win/fast/block/margin-collapse/empty-clear-blocks-expected.txt: See the png result above for comment.
- 11:02 AM Changeset in webkit [127162] by
-
- 14 edits4 adds in trunk
Add support for blendmode to webkit rendering engine
https://bugs.webkit.org/show_bug.cgi?id=95258
Patch by Rik Cabanier <cabanier@adobe.com> on 2012-08-30
Reviewed by Simon Fraser.
Source/WebCore:
This code adds support for blendmodes to the WebCore engine. The CSS parser already
supported this keyword but didn't pass it along. Support for rendering blending will
be provided in subsequent patches
Test: css3/compositing/should-have-compositing-layer.html
- css/StyleBuilder.cpp:
(WebCore::StyleBuilder::StyleBuilder):
- css/StyleResolver.cpp:
(WebCore::StyleResolver::collectMatchingRulesForList):
- rendering/RenderBox.h:
- rendering/RenderBoxModelObject.h:
(RenderBoxModelObject):
(WebCore::RenderBoxModelObject::requiresLayer):
- rendering/RenderInline.h:
(WebCore::RenderInline::requiresLayer):
- rendering/RenderLayer.cpp:
(WebCore::RenderLayer::RenderLayer):
(WebCore):
(WebCore::RenderLayer::updateBlendMode):
(WebCore::RenderLayer::ensureBacking):
(WebCore::RenderLayer::shouldBeNormalFlowOnly):
(WebCore::RenderLayer::styleChanged):
- rendering/RenderLayer.h:
(RenderLayer):
(WebCore::RenderLayer::hasBlendMode):
- rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::requiresCompositingLayer):
(WebCore::RenderLayerCompositor::requiresOwnBackingStore):
(WebCore::RenderLayerCompositor::reasonForCompositing):
(WebCore::RenderLayerCompositor::requiresCompositingForIndirectReason):
(WebCore::RenderLayerCompositor::requiresCompositingForBlending):
(WebCore):
- rendering/RenderLayerCompositor.h:
(RenderLayerCompositor):
- rendering/RenderObject.h:
(RenderObject):
(WebCore::RenderObject::hasBlendMode):
(WebCore::RenderObject::createsGroup):
- rendering/RenderTableRow.h:
- rendering/style/RenderStyle.h:
LayoutTests:
Enable test to verify that blending modes introduce a new compositing layer
- css3/compositing/resources: Added.
- css3/compositing/resources/reference.png: Added.
- css3/compositing/should-have-compositing-layer-expected.txt: Added.
- css3/compositing/should-have-compositing-layer.html: Added.
- 10:59 AM Changeset in webkit [127161] by
-
- 4 edits in trunk/Source/WebKit/mac
[Mac] Registering web views for notifications is unbalanced
https://bugs.webkit.org/show_bug.cgi?id=95465
<rdar://problem/12206765>
Reviewed by Darin Adler.
Setting the notification provider for a WK1 WebView registers that web view to the
provider. When the WebView is closed, WebCore is essentially responsible for
unregistering the web view. This leads to an imbalance on Lion since notifications are
not supported, and the NotificationController responsible for making the unregistration
call doesn't exist.
Instead, the unregisterWebView method should be called when the web view is closed.
- WebView/WebViewPrivate.h: Remove _notificationControllerDestroyed, since it is
unneeded.
- WebCoreSupport/WebNotificationClient.mm:
(WebNotificationClient::notificationControllerDestroyed):
- WebView/WebView.mm:
(-[WebView _close]): Unregister the web view from the assigned provider here.
(-[WebView _setNotificationProvider:]): Only allow setting the provider once, to avoid
registering the web view to multiple providers.
- 10:47 AM Changeset in webkit [127160] by
-
- 2 edits in trunk/Source/WebKit/blackberry
[BlackBerry] Remove unneeded force-immediate-repaint from InRegionScroller::setLayerScrollPosition
https://bugs.webkit.org/show_bug.cgi?id=95476
PR #200704
Reviewed by Yong Li.
Patch by Antonio Gomes <agomes@rim.com>
This code is not needed at this point for neither the fast nor slow in-region
scroll codepaths, and it a huge performance beast as it forces all
containers to get full-repainted per scroll call (in webkit thread).
Patch also take this opportunity to remove an early-return we
have in the {i}frame slow scrolling code path so that we can
adjust the selection handles in this case as well.
- Api/InRegionScroller.cpp:
(BlackBerry::WebKit::InRegionScrollerPrivate::setLayerScrollPosition):
- 10:44 AM Changeset in webkit [127159] by
-
- 2 edits122 adds in trunk/LayoutTests
[Chromium-Android] Upstream layout test rebaselined expectations
https://bugs.webkit.org/show_bug.cgi?id=94885
Reviewed by Adam Barth.
We rebaselined tests that
- expect different behaviors on Android (but normally don't show FAIL)
- have different layout because of precision, font size, etc.
- have different layout because of different shadow DOM of the media player
We don't rebaseline tests that would expect to output FAIL.
- platform/chromium-android/compositing/layer-creation/overflow-scrolling-touch-expected.txt: Added.
- platform/chromium-android/compositing/layer-creation/overflow-scrolling-touch.html: Added.
- platform/chromium-android/fast/block/positioning/047-expected.txt: Added.
- platform/chromium-android/fast/css/font-family-pictograph-expected.txt: Added.
- platform/chromium-android/fast/css/line-height-determined-by-primary-font-expected.txt: Added.
- platform/chromium-android/fast/css/text-transform-select-expected.txt: Added.
- platform/chromium-android/fast/dom/Orientation/create-event-orientationchange-expected.txt: Added.
- platform/chromium-android/fast/dom/Window/window-lookup-precedence-expected.txt: Added.
- platform/chromium-android/fast/events/context-onmousedown-event-expected.txt: Added.
- platform/chromium-android/fast/events/mouseup-from-button2-expected.txt: Added.
- platform/chromium-android/fast/font-boosting/various-tricky-cases-expected.txt: Added.
- platform/chromium-android/fast/font-boosting/various-tricky-cases.html: Added.
- platform/chromium-android/fast/forms/HTMLOptionElement_label05-expected.txt: Added.
- platform/chromium-android/fast/forms/disabled-select-change-index-expected.txt: Added.
- platform/chromium-android/fast/forms/form-element-geometry-expected.txt: Added.
- platform/chromium-android/fast/forms/hidden-listbox-expected.txt: Added.
- platform/chromium-android/fast/forms/implicit-submission-expected.txt: Added.
- platform/chromium-android/fast/forms/label/labelable-elements-expected.txt: Added.
- platform/chromium-android/fast/forms/listbox-bidi-align-expected.txt: Added.
- platform/chromium-android/fast/forms/listbox-scrollbar-incremental-load-expected.txt: Added.
- platform/chromium-android/fast/forms/listbox-width-change-expected.txt: Added.
- platform/chromium-android/fast/forms/option-strip-whitespace-expected.txt: Added.
- platform/chromium-android/fast/forms/select-block-background-expected.txt: Added.
- platform/chromium-android/fast/forms/select-change-listbox-size-expected.txt: Added.
- platform/chromium-android/fast/forms/select-change-popup-to-listbox-expected.txt: Added.
- platform/chromium-android/fast/forms/select-initial-position-expected.txt: Added.
- platform/chromium-android/fast/forms/select-list-box-with-height-expected.txt: Added.
- platform/chromium-android/fast/forms/select-overflow-scroll-expected.txt: Added.
- platform/chromium-android/fast/forms/select-overflow-scroll-inherited-expected.txt: Added.
- platform/chromium-android/fast/hidpi/video-controls-in-hidpi-expected.txt: Added.
- platform/chromium-android/fast/layers/video-layer-expected.txt: Added.
- platform/chromium-android/fast/preloader/script-expected.txt: Added.
- platform/chromium-android/fast/repaint/moving-shadow-on-container-expected.txt: Added.
- platform/chromium-android/fast/repaint/moving-shadow-on-path-expected.txt: Added.
- platform/chromium-android/fast/replaced/replaced-breaking-expected.txt: Added.
- platform/chromium-android/fast/text/capitalize-boundaries-expected.txt: Added.
- platform/chromium-android/fast/text/cg-fallback-bolding-expected.txt: Added.
- platform/chromium-android/fast/text/complex-text-opacity-expected.txt: Added.
- platform/chromium-android/fast/text/drawBidiText-expected.txt: Added.
- platform/chromium-android/fast/text/international/bidi-listbox-atsui-expected.txt: Added.
- platform/chromium-android/fast/text/international/bidi-listbox-expected.txt: Added.
- platform/chromium-android/fast/text/international/hindi-spacing-expected.txt: Added.
- platform/chromium-android/fast/text/international/thai-line-breaks-expected.txt: Added.
- platform/chromium-android/fast/text/updateNewFont-expected.txt: Added.
- platform/chromium-android/http/tests/security/contentSecurityPolicy/media-src-blocked-expected.txt: Added.
- platform/chromium-android/media/W3C/audio/canPlayType/canPlayType_supported_but_no_codecs_parameter_2-expected.txt: Added.
- platform/chromium-android/media/W3C/video/canPlayType/canPlayType_codecs_order_2-expected.txt: Added.
- platform/chromium-android/media/W3C/video/canPlayType/canPlayType_codecs_order_3-expected.txt: Added.
- platform/chromium-android/media/W3C/video/canPlayType/canPlayType_supported_but_no_codecs_parameter_2-expected.txt: Added.
- platform/chromium-android/media/W3C/video/canPlayType/canPlayType_supported_but_no_codecs_parameter_3-expected.txt: Added.
- platform/chromium-android/media/W3C/video/canPlayType/canPlayType_two_implies_one_3-expected.txt: Added.
- platform/chromium-android/media/W3C/video/canPlayType/canPlayType_two_implies_one_4-expected.txt: Added.
- platform/chromium-android/media/audio-controls-rendering-expected.txt: Added.
- platform/chromium-android/media/controls-after-reload-expected.txt: Added.
- platform/chromium-android/media/controls-strict-expected.txt: Added.
- platform/chromium-android/media/controls-styling-expected.txt: Added.
- platform/chromium-android/media/controls-styling-strict-expected.txt: Added.
- platform/chromium-android/media/controls-without-preload-expected.txt: Added.
- platform/chromium-android/media/csp-blocks-video-expected.txt: Added.
- platform/chromium-android/media/encrypted-media/encrypted-media-can-play-type-expected.txt: Added.
- platform/chromium-android/media/media-can-play-mpeg-audio-expected.txt: Added.
- platform/chromium-android/media/media-can-play-mpeg4-video-expected.txt: Added.
- platform/chromium-android/media/media-can-play-ogg-expected.txt: Added.
- platform/chromium-android/media/media-controls-clone-expected.txt: Added.
- platform/chromium-android/media/media-document-audio-repaint-expected.txt: Added.
- platform/chromium-android/media/track/track-cue-rendering-horizontal-expected.txt: Added.
- platform/chromium-android/media/track/track-cue-rendering-vertical-expected.txt: Added.
- platform/chromium-android/media/video-controls-rendering-expected.txt: Added.
- platform/chromium-android/media/video-currentTime-set-expected.txt: Added.
- platform/chromium-android/media/video-display-toggle-expected.txt: Added.
- platform/chromium-android/media/video-empty-source-expected.txt: Added.
- platform/chromium-android/media/video-no-audio-expected.txt: Added.
- platform/chromium-android/media/video-playing-and-pause-expected.txt: Added.
- platform/chromium-android/media/video-source-error-expected.txt: Added.
- platform/chromium-android/media/video-zoom-controls-expected.txt: Added.
- platform/chromium-android/svg/W3C-SVG-1.1/paths-data-03-f-expected.txt: Added.
- platform/chromium-android/svg/W3C-SVG-1.1/paths-data-12-t-expected.txt: Added.
- platform/chromium-android/svg/css/composite-shadow-example-expected.txt: Added.
- platform/chromium-android/svg/css/composite-shadow-with-opacity-expected.txt: Added.
- platform/chromium-android/svg/css/stars-with-shadow-expected.txt: Added.
- platform/chromium-android/svg/custom/scrolling-embedded-svg-file-image-repaint-problem-expected.txt: Added.
- platform/chromium-android/svg/custom/use-on-symbol-inside-pattern-expected.txt: Added.
- platform/chromium-android/transforms/2d/hindi-rotated-expected.txt: Added.
- platform/chromium/TestExpectations:
- 10:41 AM Changeset in webkit [127158] by
-
- 2 edits in trunk/Source/WebKit2
JSNPObject doesn't always protect its data when calling into plugin code
https://bugs.webkit.org/show_bug.cgi?id=95394
Reviewed by Brady Eidson.
We need to use NPRuntimeObjectMap::PluginProtector when calling into plugin code since
there's no telling what the plugin will do, including destroying itself.
- WebProcess/Plugins/Netscape/JSNPObject.cpp:
(WebKit::JSNPObject::getOwnPropertySlot):
(WebKit::JSNPObject::getOwnPropertyDescriptor):
- 10:32 AM Changeset in webkit [127157] by
-
- 5 edits6 adds in trunk
Make RenderBox::computeInlineDirectionMargins const
https://bugs.webkit.org/show_bug.cgi?id=95255
Reviewed by Ojan Vafai.
Source/WebCore:
This is part of making computeLogical{Height,Width} return computed values rather than
mutating the RenderBox directly. This makes a submethod const.
This change is just a refactor, but I've added some tests to cover code that wasn't
previously covered by layout tests.
Tests: fast/block/margins-perpendicular-containing-block.html
fast/table/margins-flipped-text-direction.html
fast/table/margins-perpendicular-containing-block.html
- rendering/RenderBox.cpp:
(WebCore::RenderBox::computeLogicalWidthInRegion): Handle flipped text direction manually.
(WebCore::RenderBox::computeInlineDirectionMargins): Make const with out parameters.
No longer need to call setMargin{Start,End}ForChild.
(WebCore::shouldFlipBeforeAfterMargins): Helper function to figure out how to map a logical
writing mode direction to another logical writing mode direction.
(WebCore::RenderBox::computeLogicalHeight): Use const method. This also makes it more
obvious that when computing the height, we only modify the before/after margins.
- rendering/RenderBox.h:
(RenderBox): Make computeInlineDirectionMargins const with out parameters.
- rendering/RenderTable.cpp:
(WebCore::RenderTable::computeLogicalWidth): Same as RenderBox::comptueLogicalWidthInregion.
LayoutTests:
Add test cases for setting margins with perpendicular blocks or flipped text.
- fast/block/margins-perpendicular-containing-block-expected.txt: Added.
- fast/block/margins-perpendicular-containing-block.html: Added.
- fast/table/margins-flipped-text-direction-expected.txt: Added.
- fast/table/margins-flipped-text-direction.html: Added.
- fast/table/margins-perpendicular-containing-block-expected.txt: Added.
- fast/table/margins-perpendicular-containing-block.html: Added.
- 10:23 AM Changeset in webkit [127156] by
-
- 2 edits in trunk/Source/WebKit/chromium
[chromium] Disable CCLayerTreeHostTestScrollMultipleRedraw.runMultiThread
https://bugs.webkit.org/show_bug.cgi?id=95472
Unreviewed gardening.
When landing r127079, this test was re-enabled, and has not passed since. I'm disabling it, and created bug 95473 to track fixing this issue.
- tests/CCLayerTreeHostTest.cpp:
(WebKitTests::TEST_F):
- 10:20 AM EfficientStrings edited by
- (diff)
- 10:08 AM Changeset in webkit [127155] by
-
- 18 edits in trunk/Source/WebCore
Refactor WrapShape classes to BasicShape
https://bugs.webkit.org/show_bug.cgi?id=95461
Reviewed by Rob Buis.
This is a follow up patch of bug 95411. While the previous patch
just renamed the files, this patch renames the classes, enumerations
and functions.
Just refactoring of internal names. No new tests.
- css/BasicShapeFunctions.cpp:
(WebCore::valueForBasicShape):
(WebCore::basicShapeForValue):
- css/BasicShapeFunctions.h:
(WebCore):
- css/CSSBasicShapes.cpp:
(WebCore::CSSBasicShapeRectangle::cssText):
(WebCore::CSSBasicShapeCircle::cssText):
(WebCore::CSSBasicShapeEllipse::cssText):
(WebCore::CSSBasicShapePolygon::cssText):
- css/CSSBasicShapes.h:
(WebCore::CSSBasicShape::~CSSBasicShape):
(WebCore::CSSBasicShape::CSSBasicShape):
(WebCore::CSSBasicShapeRectangle::create):
(WebCore::CSSBasicShapeRectangle::type):
(WebCore::CSSBasicShapeRectangle::CSSBasicShapeRectangle):
(WebCore::CSSBasicShapeCircle::create):
(WebCore::CSSBasicShapeCircle::type):
(WebCore::CSSBasicShapeCircle::CSSBasicShapeCircle):
(WebCore::CSSBasicShapeEllipse::create):
(WebCore::CSSBasicShapeEllipse::type):
(WebCore::CSSBasicShapeEllipse::CSSBasicShapeEllipse):
(WebCore::CSSBasicShapePolygon::create):
(WebCore::CSSBasicShapePolygon::type):
(WebCore::CSSBasicShapePolygon::CSSBasicShapePolygon):
- css/CSSComputedStyleDeclaration.cpp:
(WebCore::CSSComputedStyleDeclaration::getPropertyCSSValue):
- css/CSSParser.cpp:
(WebCore::CSSParser::parseValue):
(WebCore::CSSParser::parseBasicShapeRectangle):
(WebCore::CSSParser::parseBasicShapeCircle):
(WebCore::CSSParser::parseBasicShapeEllipse):
(WebCore::CSSParser::parseBasicShapePolygon):
(WebCore::CSSParser::parseBasicShape):
- css/CSSParser.h:
(WebCore):
- css/CSSPrimitiveValue.cpp:
(WebCore::CSSPrimitiveValue::init):
- css/CSSPrimitiveValue.h:
(WebCore):
(WebCore::CSSPrimitiveValue::getShapeValue):
(CSSPrimitiveValue):
- css/StyleBuilder.cpp:
(WebCore):
(WebCore::ApplyPropertyWrapShape::setValue):
(WebCore::ApplyPropertyWrapShape::applyValue):
(WebCore::ApplyPropertyWrapShape::createHandler):
- rendering/RenderBlock.cpp:
(WebCore::RenderBlock::updateWrapShapeInfoAfterStyleChange):
- rendering/RenderBlock.h:
(RenderBlock):
- rendering/WrapShapeInfo.cpp:
(WebCore::WrapShapeInfo::isWrapShapeInfoEnabledForRenderBlock):
(WebCore::WrapShapeInfo::computeShapeSize):
- rendering/style/BasicShapes.cpp:
(WebCore::BasicShape::destroy):
- rendering/style/BasicShapes.h:
(WebCore::BasicShape::BasicShape):
(WebCore::BasicShapeRectangle::create):
(WebCore::BasicShapeRectangle::BasicShapeRectangle):
(WebCore::BasicShapeCircle::create):
(WebCore::BasicShapeCircle::BasicShapeCircle):
(WebCore::BasicShapeEllipse::create):
(WebCore::BasicShapeEllipse::BasicShapeEllipse):
(WebCore::BasicShapePolygon::create):
(WebCore::BasicShapePolygon::BasicShapePolygon):
- rendering/style/RenderStyle.h:
- rendering/style/StyleRareNonInheritedData.h:
(StyleRareNonInheritedData):
- 10:04 AM Changeset in webkit [127154] by
-
- 9 edits in trunk/Source/WebKit/chromium
[chromium] Clean up some webkit compositor unit tests
https://bugs.webkit.org/show_bug.cgi?id=95410
Patch by James Robinson <jamesr@chromium.org> on 2012-08-30
Reviewed by Adrienne Walker.
This adds includes that were implicitly being picked up, removes ones that weren't being used, removes
dead code, adds OVERRIDE on functions that OVERRIDE, and deinlines virtuals that the chromium clang style
plugin is unhappy about.
- tests/CCAnimationTestCommon.cpp:
(WebKitTests::FakeFloatAnimationCurve::duration):
(WebKitTests):
(WebKitTests::FakeFloatAnimationCurve::getValue):
(WebKitTests::FakeTransformTransition::duration):
(WebKitTests::FakeFloatTransition::duration):
(WebKitTests::FakeLayerAnimationControllerClient::id):
(WebKitTests::FakeLayerAnimationControllerClient::setOpacityFromAnimation):
(WebKitTests::FakeLayerAnimationControllerClient::opacity):
(WebKitTests::FakeLayerAnimationControllerClient::setTransformFromAnimation):
(WebKitTests::FakeLayerAnimationControllerClient::transform):
- tests/CCAnimationTestCommon.h:
(FakeFloatAnimationCurve):
(FakeTransformTransition):
(FakeFloatTransition):
(FakeLayerAnimationControllerClient):
- tests/CCTiledLayerTestCommon.cpp:
(WebKitTests::FakeLayerTextureUpdater::sampledTexelFormat):
(WebKitTests):
(WebKitTests::FakeTiledLayerWithScaledBounds::FakeTiledLayerWithScaledBounds):
(WebKitTests::FakeTiledLayerChromium::textureManager):
(WebKitTests::FakeTiledLayerChromium::textureUpdater):
(WebKitTests::FakeTiledLayerWithScaledBounds::contentBounds):
(WebKitTests::FakeTextureUploader::isBusy):
(WebKitTests::FakeTextureUploader::uploadTexture):
- tests/CCTiledLayerTestCommon.h:
(FakeLayerTextureUpdater):
(FakeTiledLayerChromium):
(FakeTiledLayerWithScaledBounds):
(FakeTextureUploader):
- tests/FakeWebCompositorOutputSurface.h:
- tests/TextureCopierTest.cpp:
(MockContext):
(TEST):
- tests/ThrottledTextureUploaderTest.cpp:
(WebKit::TEST):
- tests/WebLayerTreeViewTest.cpp:
- 9:52 AM Changeset in webkit [127153] by
-
- 3 edits2 adds in trunk
Fix a crash in SourceBufferList.remove().
https://bugs.webkit.org/show_bug.cgi?id=94950
Patch by Aaron Colwell <acolwell@chromium.org> on 2012-08-30
Reviewed by Eric Carlson.
Source/WebCore:
Move SourceBuffer::clear() call before the removal of the SourceBuffer from
SourceBufferList::m_list to avoid a use after free if m_list holds the last
reference.
Test: http/tests/media/media-source/video-media-source-sourcebufferlist-crash.html
- Modules/mediasource/SourceBufferList.cpp:
(WebCore::SourceBufferList::remove):
LayoutTests:
Added a layout test to verify that the application doesn't crash if a SourceBufferList is
holding the only references for its SourceBuffer objects when a MediaSource transitions to "closed".
- http/tests/media/media-source/video-media-source-sourcebufferlist-crash-expected.txt: Added.
- http/tests/media/media-source/video-media-source-sourcebufferlist-crash.html: Added.
- 9:38 AM Changeset in webkit [127152] by
-
- 2 edits in trunk/LayoutTests
[EFL] Remove redundant entries on the Skipped list
https://bugs.webkit.org/show_bug.cgi?id=95458
Unreviewed gardening.
Patch by Thiago Marcos P. Santos <thiago.santos@intel.com> on 2012-08-30
- platform/efl/Skipped:
- 9:31 AM Changeset in webkit [127151] by
-
- 2 edits in trunk/Source/JavaScriptCore
[BlackBerry] Set timer client on platform timer used in HeapTimer
https://bugs.webkit.org/show_bug.cgi?id=95464
Reviewed by Rob Buis.
Otherwise the timer won't work.
- heap/HeapTimer.cpp:
(JSC::HeapTimer::HeapTimer):
- 9:12 AM Changeset in webkit [127150] by
-
- 7 edits in trunk/Source/WebCore
[BlackBerry] Modifying how IP domains are handled in Cookies
https://bugs.webkit.org/show_bug.cgi?id=95381
Patch by Otto Derek Cheung <otcheung@rim.com> on 2012-08-30
Reviewed by Rob Buis.
Internally reviewed by Joe Mason.
Current implementation handles IP addresses like normal domains.
This makes invalid cross domain cookies possibe by setting cookie
domains to a suffix of the targeted IP. (ex. hackers on 11.121.61.97
can set cookies to 61.97, so they show up on the targeted website of
10.120.61.97)
New Implementation is to detect IP addresses and treat them without
exploding them with the delimiter ".". That way, IP addresses will
be stored as a whole and other IPs won't have access to it.
PR 130051
Manually tested by accessing a webpage via IP (hosted through
LAMP - ex:10.121.61.97) and tried to set cookies with domains that
are a suffix to the ip address (ex: .97, 121.61.97).
Also tried to set cookies to other ip addresses that "domain matches"
the IP address (ex. 11.121.61.97). Verified that they all failed.
Tested using the cookies test page: http://browser.swlab.rim.net/test/cookies
- platform/blackberry/CookieManager.cpp:
(WebCore::CookieManager::getRawCookies):
(WebCore::CookieManager::checkAndTreatCookie):
(WebCore::CookieManager::findOrCreateCookieMap):
- platform/blackberry/CookieManager.h:
- platform/blackberry/CookieParser.cpp:
(WebCore::CookieParser::CookieParser):
(WebCore::CookieParser::parseOneCookie):
- platform/blackberry/CookieParser.h:
(CookieParser):
- platform/blackberry/ParsedCookie.cpp:
(WebCore::ParsedCookie::ParsedCookie):
- platform/blackberry/ParsedCookie.h:
(WebCore::ParsedCookie::setDomain):
(WebCore::ParsedCookie::domainIsIPAddress):
(ParsedCookie):
- 9:08 AM Changeset in webkit [127149] by
-
- 2 edits in trunk/Source/WebCore
Web Inspector: [regression] ui: heap profiler: splitter between containment and retainment views has white background.
https://bugs.webkit.org/show_bug.cgi?id=95460
Reviewed by Yury Semikhatsky.
it was regressed at r122332.
- inspector/front-end/heapProfiler.css:
(.heap-snapshot-view .retainers-view-header):
- 9:03 AM Changeset in webkit [127148] by
-
- 3 edits in trunk/Source/JavaScriptCore
[sh4] Add missing implementation for JavaScriptCore JIT
https://bugs.webkit.org/show_bug.cgi?id=95452
Patch by Julien BRIANCEAU <jbrianceau@nds.com> on 2012-08-30
Reviewed by Oliver Hunt.
- assembler/MacroAssemblerSH4.h:
(JSC::MacroAssemblerSH4::isCompactPtrAlignedAddressOffset):
(MacroAssemblerSH4):
(JSC::MacroAssemblerSH4::add32):
(JSC::MacroAssemblerSH4::convertibleLoadPtr):
- assembler/SH4Assembler.h:
(JSC::SH4Assembler::labelIgnoringWatchpoints):
(SH4Assembler):
(JSC::SH4Assembler::replaceWithLoad):
(JSC::SH4Assembler::replaceWithAddressComputation):
- 8:58 AM Changeset in webkit [127147] by
-
- 45 edits in trunk/LayoutTests
Web Inspector: sort object properties when dumping them in tests
https://bugs.webkit.org/show_bug.cgi?id=95459
Reviewed by Vsevolod Vlasov.
add/dumpObject used to dump object properties in the natural enumerable order,
sorting by property name now.
- http/tests/inspector/inspector-test.js:
(initialize_InspectorTest.InspectorTest.addObject):
- http/tests/inspector/resource-har-conversion-expected.txt:
- http/tests/inspector/resource-har-headers-expected.txt:
- http/tests/inspector/resource-har-pages-expected.txt:
- http/tests/inspector/resource-parameters-expected.txt:
- http/tests/inspector/resources/extension-main.js:
- inspector/cookie-parser-expected.txt:
- inspector/debugger/script-extract-outline-expected.txt:
- inspector/extensions/extensions-api-expected.txt:
- inspector/extensions/extensions-audits-api-expected.txt:
- inspector/extensions/extensions-console-expected.txt:
- inspector/extensions/extensions-panel-expected.txt:
- inspector/extensions/extensions-resources-expected.txt:
- inspector/extensions/extensions-sidebar-expected.txt:
- inspector/filtered-item-selection-dialog-filtering-expected.txt:
- inspector/report-protocol-errors-expected.txt:
- inspector/runtime/runtime-getProperties-expected.txt:
- inspector/runtime/runtime-localStorage-getProperties-expected.txt:
- inspector/timeline/timeline-animation-frame-expected.txt:
- inspector/timeline/timeline-decode-resize-expected.txt:
- inspector/timeline/timeline-dom-content-loaded-event-expected.txt:
- inspector/timeline/timeline-enum-stability-expected.txt:
- inspector/timeline/timeline-event-dispatch-expected.txt:
- inspector/timeline/timeline-injected-script-eval-expected.txt:
- inspector/timeline/timeline-layout-expected.txt:
- inspector/timeline/timeline-load-event-expected.txt:
- inspector/timeline/timeline-mark-timeline-expected.txt:
- inspector/timeline/timeline-network-resource-expected.txt:
- inspector/timeline/timeline-paint-expected.txt:
- inspector/timeline/timeline-parse-html-expected.txt:
- inspector/timeline/timeline-recalculate-styles-expected.txt:
- inspector/timeline/timeline-script-tag-1-expected.txt:
- inspector/timeline/timeline-script-tag-2-expected.txt:
- inspector/timeline/timeline-time-stamp-expected.txt:
- inspector/timeline/timeline-timer-expected.txt:
- inspector/user-metrics-expected.txt:
- platform/chromium/http/tests/inspector/resource-har-conversion-expected.txt:
- platform/chromium/inspector/timeline/timeline-animation-frame-expected.txt:
- platform/chromium/inspector/timeline/timeline-event-dispatch-expected.txt:
- platform/chromium/inspector/timeline/timeline-layout-expected.txt:
- platform/chromium/inspector/timeline/timeline-network-resource-expected.txt:
- platform/chromium/inspector/timeline/timeline-paint-expected.txt:
- platform/chromium/inspector/timeline/timeline-parse-html-expected.txt:
- platform/chromium/inspector/timeline/timeline-timer-expected.txt:
- 8:53 AM Changeset in webkit [127146] by
-
- 2 edits in trunk/Source/WebKit/blackberry
[BlackBerry] node search does not work with elements on touch start listener
https://bugs.webkit.org/show_bug.cgi?id=95252
Patch by Hanna Ma <Hanma@rim.com> on 2012-08-30
Reviewed by Antonio Gomes.
Fix the node search function of web insepctor on elements with touch start listener.
Add methods from webPage to check if the node search functionality is enabled before handling touch events.
PR194107
- Api/WebPage.cpp:
(BlackBerry::WebKit::WebPage::touchEvent):
- 8:51 AM Changeset in webkit [127145] by
-
- 7 edits2 deletes in trunk
Unreviewed, rolling out r127131.
http://trac.webkit.org/changeset/127131
https://bugs.webkit.org/show_bug.cgi?id=95463
It makes the fast/regions/ test crash on Chromium (Requested
by jchaffraix on #webkit).
Patch by Sheriff Bot <webkit.review.bot@gmail.com> on 2012-08-30
Source/WebCore:
- css/StyleResolver.cpp:
(WebCore::StyleResolver::collectMatchingRulesForList):
- rendering/InlineBox.cpp:
- rendering/InlineBox.h:
(WebCore):
(InlineBox):
- rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::addToLine):
(WebCore::InlineFlowBox::addTextBoxVisualOverflow):
- rendering/RenderRegion.h:
(RenderRegion):
LayoutTests:
- fast/regions/region-style-text-shadow-expected.html: Removed.
- fast/regions/region-style-text-shadow.html: Removed.
- 8:46 AM Changeset in webkit [127144] by
-
- 2 edits in trunk/LayoutTests
Unreviewed, rolling out r127133.
http://trac.webkit.org/changeset/127133
https://bugs.webkit.org/show_bug.cgi?id=95462
These tests still fail (Requested by Ossy on #webkit).
Patch by Sheriff Bot <webkit.review.bot@gmail.com> on 2012-08-30
- platform/qt-5.0-wk2/Skipped:
- 8:30 AM Changeset in webkit [127143] by
-
- 2 edits in trunk/LayoutTests
[EFL] Remove tests from Skipped list that don't exist
https://bugs.webkit.org/show_bug.cgi?id=95457
Unreviewed gardening.
Patch by Thiago Marcos P. Santos <thiago.santos@intel.com> on 2012-08-30
- platform/efl/Skipped:
- 8:26 AM Changeset in webkit [127142] by
-
- 6 edits in trunk
Web Inspector: Some urls in CSS stylesheets cause errors when generating Computed Styles HTML
https://bugs.webkit.org/show_bug.cgi?id=95427
Reviewed by Vsevolod Vlasov.
Source/WebCore:
Pass the parent StylesSidebarPane into ComputedStylePropertiesSection to avoid referencing an undefined field.
Avoid duplicating the parent pane by externally setting the |data| field on style property sections.
- inspector/front-end/StylesSidebarPane.js:
(WebInspector.StylesSidebarPane.prototype._rebuildSectionsForStyleRules):
(WebInspector.StylesSidebarPane.prototype.addBlankSection):
(WebInspector.StylePropertiesSection.prototype.get pane):
(WebInspector.ComputedStylePropertiesSection):
(WebInspector.ComputedStylePropertiesSection.prototype.get pane):
(WebInspector.ComputedStylePropertiesSection.prototype.onpopulate):
(WebInspector.StylePropertyTreeElement.prototype.updateTitle.linkifyURL):
LayoutTests:
- inspector/styles/inject-stylesheet-expected.txt:
- inspector/styles/inject-stylesheet.html:
- platform/chromium/inspector/styles/inject-stylesheet-expected.txt:
- 8:06 AM Changeset in webkit [127141] by
-
- 2 edits in trunk/Source/WebKit/blackberry
[BlackBerry] One shot drawing sync flag never cleared when there's no layers
https://bugs.webkit.org/show_bug.cgi?id=95447
Reviewed by Antonio Gomes.
PR 199866
When we remove the layers, WebKit could tell us we need a one shot
drawing sync to repaint the BackingStore with the content that was
previously drawn by layers. We also sometimes set the one shot drawing
sync flag manually.
We would bail from commit if there were no layers and not actually
perform the osds, so it left the osds flag set forever and the
situation prevails, because nobody will ever clear the osds flag.
The BackingStore surrenders the responsibility of blitting to the AC
commit mechanism when a one shot drawing sync is pending, because we
want the operations render, commit and blit to happen in the correct
order with no intermediate blits.
So, no commit means no blit, and the result of regular rendering did
not show up on screen.
Fixed by not bailing from commit if there's no layers, instead clearing
the "needs commit" and osds flags and performing a blit.
Reviewed internally by Andrew Lo.
- Api/WebPage.cpp:
(BlackBerry::WebKit::WebPagePrivate::commitRootLayerIfNeeded):
- 7:56 AM Changeset in webkit [127140] by
-
- 11 edits6 adds in trunk
Tick marks don't match thumb when applying padding or border to input type=range
https://bugs.webkit.org/show_bug.cgi?id=93791
Reviewed by Kent Tamura.
Source/WebCore:
Tick marks don't match thumb when styling input type=range. This change
calculate tick mark positions from track element size. Bug 94915 handles
drawing the track inside the content area.
Tests: fast/forms/datalist/input-appearance-range-with-padding-with-datalist.html
fast/forms/datalist/input-appearance-range-with-transform.html
- html/HTMLInputElement.cpp:
(WebCore::HTMLInputElement::sliderTrackElement):
(WebCore):
- html/HTMLInputElement.h:
(HTMLInputElement):
- html/InputType.h:
(WebCore::InputType::sliderTrackElement):
- html/RangeInputType.cpp:
(WebCore::RangeInputType::sliderTrackElement):
(WebCore):
- html/RangeInputType.h:
(RangeInputType):
- rendering/RenderTheme.cpp:
(WebCore::RenderTheme::paintSliderTicks): rect's position is relative to
the transformed ancestor element. sliderBounds is absolute. We use them
to calculate the track element position relative to the transformed
ancestor element.
LayoutTests:
- fast/forms/datalist/input-appearance-range-with-datalist-rtl-expected.html:
- fast/forms/datalist/input-appearance-range-with-datalist-rtl.html:
- fast/forms/datalist/input-appearance-range-with-padding-with-datalist-expected.txt: Added.
- fast/forms/datalist/input-appearance-range-with-padding-with-datalist.html: Added.
- fast/forms/datalist/input-appearance-range-with-transform-expected.txt: Added.
- fast/forms/datalist/input-appearance-range-with-transform.html: Added.
- platform/chromium-mac/fast/forms/datalist/input-appearance-range-with-padding-with-datalist-expected.png: Added.
- platform/chromium-mac/fast/forms/datalist/input-appearance-range-with-transform-expected.png: Added.
- platform/chromium/TestExpectations:
- 7:48 AM Changeset in webkit [127139] by
-
- 2 edits in trunk/Source/WebCore
Remove the now-unneeded invalidations in RenderTable::removeCaption
https://bugs.webkit.org/show_bug.cgi?id=94889
Reviewed by Abhishek Arya.
Following bug 94842 and 95090, the invalidation code was pushed down to
RenderTableCaption. This made apparent that we did some invalidations that
were not needed.
Refactoring covered by existing tests.
- rendering/RenderTable.cpp:
(WebCore::RenderTable::removeCaption):
Removed setNeedsRecalcStyle() as it's unneeded now. It was probably needed back when
we didn't support multiple captions (see bug 58249) but the need was never documented
so you could wonder if it was really needed in the first place.
- 7:45 AM Changeset in webkit [127138] by
-
- 2 edits in trunk/Source/WebCore
Web Inspector: [regression] ui: selectors in heap profiler view have no arrows at the right side of the text.
https://bugs.webkit.org/show_bug.cgi?id=95455
Reviewed by Yury Semikhatsky.
In console we use a span which wraps selectors and have necessary background.
- inspector/front-end/HeapSnapshotView.js:
(WebInspector.HeapSnapshotView.prototype.get statusBarItems.appendArrowImage):
(WebInspector.HeapSnapshotView.prototype.get statusBarItems):
- 7:42 AM Changeset in webkit [127137] by
-
- 4 edits in trunk
[WK2][WTR] Add didReceiveServerRedirectForProvisionalLoadForFrame dumping
https://bugs.webkit.org/show_bug.cgi?id=95454
Patch by Mikhail Pozdnyakov <mikhail.pozdnyakov@intel.com> on 2012-08-30
Reviewed by Kenneth Rohde Christiansen.
Tools:
Added didReceiveServerRedirectForProvisionalLoadForFrame dumping to WTR.
- WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::didReceiveServerRedirectForProvisionalLoadForFrame):
LayoutTests:
Unskipped http/tests/loading/307-after-303-after-post.html.
Output for http/tests/loading/redirect-methods.html is improved.
- platform/wk2/Skipped:
- 7:40 AM Changeset in webkit [127136] by
-
- 2 edits in trunk/Tools
Update my email address.
Not reviewed.
- Scripts/webkitpy/common/config/committers.py:
- 7:31 AM Changeset in webkit [127135] by
-
- 2 edits in trunk/LayoutTests
REGRESSION(r127039): It broke inspector/timeline/timeline-load.html fails on many platforms
https://bugs.webkit.org/show_bug.cgi?id=95414
Unreviewed. Remove stray InspectorTest.completeTest() from a method that is called from several test cases.
- inspector/timeline/timeline-load.html:
- 7:05 AM Changeset in webkit [127134] by
-
- 2 edits in trunk/Source/WebCore
[BlackBerry] LayerTiler needs to respect new setting for prefill rect.
https://bugs.webkit.org/show_bug.cgi?id=95446
Patch by Andrew Lo <anlo@rim.com> on 2012-08-30
Reviewed by Rob Buis.
Internally reviewed by Arvid Nilsson.
Internal PR181637
Use new prefill rectangle setting to determine which tiles should be
prefilled in LayerTiler.
- platform/graphics/blackberry/LayerTiler.cpp:
(WebCore::LayerTiler::shouldPrefillTile):
- 7:05 AM Changeset in webkit [127133] by
-
- 2 edits in trunk/LayoutTests
[Qt][WK2] Unreviewed gardening, unskip now passing tests.
- platform/qt-5.0-wk2/Skipped:
- 6:56 AM WebInspector edited by
- (diff)
- 6:54 AM WebInspector edited by
- (diff)
- 6:44 AM Changeset in webkit [127132] by
-
- 13 edits6 moves in trunk/Source/WebCore
Refactor WrapShape to Shape/BasicShape
https://bugs.webkit.org/show_bug.cgi?id=95411
Reviewed by Andreas Kling.
The wrap shapes are currently specified by CSS3 Exclusions but are useful for other
CSS related proposals like CSS Masking as well. This is the first patch on a chain
of patches to refactor WrapShape to BasicShape. With this patch all relevant files
get renamend and the build systems updated. The classes will be renamend in a second
step.
Just renaming of files. No new tests.
- CMakeLists.txt:
- GNUmakefile.list.am:
- Target.pri:
- WebCore.gypi:
- WebCore.vcproj/WebCore.vcproj:
- WebCore.xcodeproj/project.pbxproj:
- css/BasicShapeFunctions.cpp: Renamed from Source/WebCore/css/WrapShapeFunctions.cpp.
(WebCore):
(WebCore::valueForWrapShape):
(WebCore::convertToLength):
(WebCore::wrapShapeForValue):
- css/BasicShapeFunctions.h: Renamed from Source/WebCore/css/WrapShapeFunctions.h.
(WebCore):
- css/CSSAllInOne.cpp:
- css/CSSBasicShapes.cpp: Renamed from Source/WebCore/css/CSSWrapShapes.cpp.
(WebCore):
(WebCore::CSSWrapShapeRectangle::cssText):
(WebCore::CSSWrapShapeCircle::cssText):
(WebCore::CSSWrapShapeEllipse::cssText):
(WebCore::CSSWrapShapePolygon::cssText):
- css/CSSBasicShapes.h: Renamed from Source/WebCore/css/CSSWrapShapes.h.
(WebCore):
(CSSWrapShape):
(WebCore::CSSWrapShape::~CSSWrapShape):
(WebCore::CSSWrapShape::CSSWrapShape):
(CSSWrapShapeRectangle):
(WebCore::CSSWrapShapeRectangle::create):
(WebCore::CSSWrapShapeRectangle::x):
(WebCore::CSSWrapShapeRectangle::y):
(WebCore::CSSWrapShapeRectangle::width):
(WebCore::CSSWrapShapeRectangle::height):
(WebCore::CSSWrapShapeRectangle::radiusX):
(WebCore::CSSWrapShapeRectangle::radiusY):
(WebCore::CSSWrapShapeRectangle::setX):
(WebCore::CSSWrapShapeRectangle::setY):
(WebCore::CSSWrapShapeRectangle::setWidth):
(WebCore::CSSWrapShapeRectangle::setHeight):
(WebCore::CSSWrapShapeRectangle::setRadiusX):
(WebCore::CSSWrapShapeRectangle::setRadiusY):
(WebCore::CSSWrapShapeRectangle::type):
(WebCore::CSSWrapShapeRectangle::CSSWrapShapeRectangle):
(CSSWrapShapeCircle):
(WebCore::CSSWrapShapeCircle::create):
(WebCore::CSSWrapShapeCircle::centerX):
(WebCore::CSSWrapShapeCircle::centerY):
(WebCore::CSSWrapShapeCircle::radius):
(WebCore::CSSWrapShapeCircle::setCenterX):
(WebCore::CSSWrapShapeCircle::setCenterY):
(WebCore::CSSWrapShapeCircle::setRadius):
(WebCore::CSSWrapShapeCircle::type):
(WebCore::CSSWrapShapeCircle::CSSWrapShapeCircle):
(CSSWrapShapeEllipse):
(WebCore::CSSWrapShapeEllipse::create):
(WebCore::CSSWrapShapeEllipse::centerX):
(WebCore::CSSWrapShapeEllipse::centerY):
(WebCore::CSSWrapShapeEllipse::radiusX):
(WebCore::CSSWrapShapeEllipse::radiusY):
(WebCore::CSSWrapShapeEllipse::setCenterX):
(WebCore::CSSWrapShapeEllipse::setCenterY):
(WebCore::CSSWrapShapeEllipse::setRadiusX):
(WebCore::CSSWrapShapeEllipse::setRadiusY):
(WebCore::CSSWrapShapeEllipse::type):
(WebCore::CSSWrapShapeEllipse::CSSWrapShapeEllipse):
(CSSWrapShapePolygon):
(WebCore::CSSWrapShapePolygon::create):
(WebCore::CSSWrapShapePolygon::appendPoint):
(WebCore::CSSWrapShapePolygon::getXAt):
(WebCore::CSSWrapShapePolygon::getYAt):
(WebCore::CSSWrapShapePolygon::values):
(WebCore::CSSWrapShapePolygon::setWindRule):
(WebCore::CSSWrapShapePolygon::windRule):
(WebCore::CSSWrapShapePolygon::type):
(WebCore::CSSWrapShapePolygon::CSSWrapShapePolygon):
- css/CSSComputedStyleDeclaration.cpp:
- css/CSSParser.cpp:
- css/CSSPrimitiveValue.cpp:
- css/StyleBuilder.cpp:
- rendering/style/BasicShapes.cpp: Renamed from Source/WebCore/rendering/style/WrapShapes.cpp.
(WebCore):
(WebCore::WrapShape::destroy):
- rendering/style/BasicShapes.h: Renamed from Source/WebCore/rendering/style/WrapShapes.h.
(WebCore):
(WrapShape):
(WebCore::WrapShape::deref):
(WebCore::WrapShape::type):
(WebCore::WrapShape::WrapShape):
(WrapShapeRectangle):
(WebCore::WrapShapeRectangle::create):
(WebCore::WrapShapeRectangle::x):
(WebCore::WrapShapeRectangle::y):
(WebCore::WrapShapeRectangle::width):
(WebCore::WrapShapeRectangle::height):
(WebCore::WrapShapeRectangle::cornerRadiusX):
(WebCore::WrapShapeRectangle::cornerRadiusY):
(WebCore::WrapShapeRectangle::setX):
(WebCore::WrapShapeRectangle::setY):
(WebCore::WrapShapeRectangle::setWidth):
(WebCore::WrapShapeRectangle::setHeight):
(WebCore::WrapShapeRectangle::setCornerRadiusX):
(WebCore::WrapShapeRectangle::setCornerRadiusY):
(WebCore::WrapShapeRectangle::WrapShapeRectangle):
(WrapShapeCircle):
(WebCore::WrapShapeCircle::create):
(WebCore::WrapShapeCircle::centerX):
(WebCore::WrapShapeCircle::centerY):
(WebCore::WrapShapeCircle::radius):
(WebCore::WrapShapeCircle::setCenterX):
(WebCore::WrapShapeCircle::setCenterY):
(WebCore::WrapShapeCircle::setRadius):
(WebCore::WrapShapeCircle::WrapShapeCircle):
(WrapShapeEllipse):
(WebCore::WrapShapeEllipse::create):
(WebCore::WrapShapeEllipse::centerX):
(WebCore::WrapShapeEllipse::centerY):
(WebCore::WrapShapeEllipse::radiusX):
(WebCore::WrapShapeEllipse::radiusY):
(WebCore::WrapShapeEllipse::setCenterX):
(WebCore::WrapShapeEllipse::setCenterY):
(WebCore::WrapShapeEllipse::setRadiusX):
(WebCore::WrapShapeEllipse::setRadiusY):
(WebCore::WrapShapeEllipse::WrapShapeEllipse):
(WrapShapePolygon):
(WebCore::WrapShapePolygon::create):
(WebCore::WrapShapePolygon::windRule):
(WebCore::WrapShapePolygon::values):
(WebCore::WrapShapePolygon::getXAt):
(WebCore::WrapShapePolygon::getYAt):
(WebCore::WrapShapePolygon::setWindRule):
(WebCore::WrapShapePolygon::appendPoint):
(WebCore::WrapShapePolygon::WrapShapePolygon):
- rendering/style/StyleRareNonInheritedData.h:
- 6:19 AM Changeset in webkit [127131] by
-
- 7 edits2 adds in trunk
[CSSRegions]Add support for text-shadow in region styling
https://bugs.webkit.org/show_bug.cgi?id=94472
Patch by Andrei Onea <onea@adobe.com> on 2012-08-30
Reviewed by David Hyatt.
Source/WebCore:
The CSSRegions spec allows region styling to be applied on the text-shadow property
as well. We need to also add this in WebKit.
Test: fast/regions/region-style-text-shadow.html
- css/StyleResolver.cpp:
(WebCore::StyleResolver::collectMatchingRulesForList):
Make sure we actually collect text-shadow property from parser in an @region rule.
- rendering/InlineBox.cpp:
(WebCore::InlineBox::styleInRegion):
(WebCore):
(WebCore::InlineBox::regionDuringLayout):
- rendering/InlineBox.h:
(WebCore):
(InlineBox):
- rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::addToLine):
Take into account region styling, so that "knownToHaveNoOverflow" is computed
properly.
(WebCore::InlineFlowBox::addTextBoxVisualOverflow):
Take into account region styling, so that the visual overflow rect is computed
properly.
- rendering/RenderRegion.h:
(RenderRegion):
Made computeStyleInRegion public.
LayoutTests:
Added test for region styling on the text shadow property.
- fast/regions/region-style-text-shadow-expected.html: Added.
- fast/regions/region-style-text-shadow.html: Added.
- 6:04 AM Changeset in webkit [127130] by
-
- 2 edits1 add in trunk/LayoutTests
Unreviewed GTK gardening.
Adding a platform-specific baseline that's required after r127117.
Fixing expectations for a couple of tests that were gardened in r127109.
They are both reftests so image failures are occurring.
- platform/gtk/TestExpectations:
- platform/gtk/fast/events/set-attribute-listener-window-onerror-crash-expected.txt: Added.
- 5:59 AM Changeset in webkit [127129] by
-
- 2 edits1 copy in trunk/LayoutTests
Update test expectations to follow common guidelines.
https://bugs.webkit.org/show_bug.cgi?id=95442
Patch by Anton Muhin <antonm@chromium.org> on 2012-08-30
Reviewed by Csaba Osztrogonác.
Follow up to https://trac.webkit.org/changeset/127117
- fast/events/set-attribute-listener-window-onerror-crash-expected.txt:
- platform/chromium/fast/events/set-attribute-listener-window-onerror-crash-expected.txt: Added.
- 5:37 AM Changeset in webkit [127128] by
-
- 3 edits in trunk/Source/JavaScriptCore
[BlackBerry] Eliminate build warnings
https://bugs.webkit.org/show_bug.cgi?id=95338
Reviewed by Filip Pizlo.
static_cast to the same type to eliminate the build time warnings.
- assembler/AssemblerBufferWithConstantPool.h:
(JSC::AssemblerBufferWithConstantPool::flushWithoutBarrier):
- assembler/MacroAssemblerARM.h:
(JSC::MacroAssemblerARM::branch32):
- 4:53 AM Changeset in webkit [127127] by
-
- 2 edits in trunk/LayoutTests
[EFL] Gardening after r127108
https://bugs.webkit.org/show_bug.cgi?id=95438
Unreviewed gardening.
Gardening of fast/dom/shadow/shadowdom-for-textarea-complex-shadow.html and
fast/dom/shadow/shadowdom-for-textarea-with-placeholder.html.
Patch by Mikhail Pozdnyakov <mikhail.pozdnyakov@intel.com> on 2012-08-30
- platform/efl/TestExpectations:
- 4:43 AM Changeset in webkit [127126] by
-
- 4 edits in trunk/Source/WebCore
Element: Share code between setAttributeNode() and other attribute setters.
<http://webkit.org/b/95328>
Reviewed by Antti Koivisto.
Removed the specialized ElementAttributeData::replaceAttribute() that was only used for
replacing an existing Attr node on an Element. Instead, just use Element::setAttributeInternal()
like all the other attribute setters.
- dom/Element.cpp:
(WebCore::Element::setAttributeNode):
- dom/ElementAttributeData.cpp:
- dom/ElementAttributeData.h:
(ElementAttributeData):
- 4:33 AM Changeset in webkit [127125] by
-
- 2 edits in trunk/Source/WebKit2
[QT][WK2] webview API doc
https://bugs.webkit.org/show_bug.cgi?id=81701
Patch by Mike Sierra <mike.sierra@nokia.com> on 2012-08-30
Reviewed by Simon Hausmann.
Various improvements and additions to the documentation of the QML WebView element.
- UIProcess/API/qt/qquickwebview.cpp:
- 4:18 AM Changeset in webkit [127124] by
-
- 3 edits in trunk/Source/WebCore
1.9.90 drops symbols, breaking compatibility
https://bugs.webkit.org/show_bug.cgi?id=93477
Patch by Xan Lopez <xlopez@igalia.com> on 2012-08-30
Reviewed by Martin Robinson.
Add a bunch of compatibility methods to the GObject DOM bindings
to cope with recent renames.
- bindings/gobject/WebKitDOMCustom.cpp:
(webkit_dom_html_element_get_class_list):
(webkit_dom_element_get_webkit_region_overflow):
(webkit_dom_webkit_named_flow_get_content_nodes):
(webkit_dom_webkit_named_flow_get_regions_by_content_node):
- bindings/gobject/WebKitDOMCustom.h:
- 4:07 AM Changeset in webkit [127123] by
-
- 9 edits in trunk/Source/WebCore
Cache and share parsed imported stylesheets
https://bugs.webkit.org/show_bug.cgi?id=95219
Reviewed by Andreas Kling.
We currently cache and share parsed data structures of stylesheets loaded with <link>. We should do
the same with stylesheets loaded using @import rules as they are also fairly common.
This patch adds support for caching and sharing stylesheets loaded using @import rules. Only leaf
stylesheets (that don't have @import rules themselves) can be cached for now.
- css/CSSImportRule.cpp:
(WebCore::CSSImportRule::reattachStyleSheetContents):
(WebCore):
- css/CSSImportRule.h:
(CSSImportRule):
- css/CSSStyleSheet.cpp:
(WebCore::CSSStyleSheet::willMutateRules):
(WebCore::CSSStyleSheet::reattachCSSOMWrappers):
- css/CSSStyleSheet.h:
(CSSStyleSheet):
- css/StyleRuleImport.cpp:
(WebCore::StyleRuleImport::setCSSStyleSheet):
(WebCore::StyleRuleImport::reattachStyleSheetContents):
(WebCore):
- css/StyleRuleImport.h:
(StyleRuleImport):
- loader/cache/CachedCSSStyleSheet.cpp:
(WebCore::CachedCSSStyleSheet::saveParsedStyleSheet):
- 4:00 AM Changeset in webkit [127122] by
-
- 3 edits in trunk/Tools
[NRWT] Add support for recognizing arbitrary process names in crash lines
https://bugs.webkit.org/show_bug.cgi?id=95435
Reviewed by Adam Barth.
When running layout tests in the chromium port using the content shell,
we want to be able to report sub-process crashes as well.
- Scripts/webkitpy/layout_tests/port/driver.py:
(Driver._check_for_driver_crash):
- Scripts/webkitpy/layout_tests/port/driver_unittest.py:
(DriverTest.test_check_for_driver_crash):
- 3:26 AM Changeset in webkit [127121] by
-
- 4 edits in trunk/Source/WebKit
Use ASCIILiteral for DEFINE_STATIC_LOCAL string
https://bugs.webkit.org/show_bug.cgi?id=95420
Reviewed by Benjamin Poulain.
As recommended by http://trac.webkit.org/wiki/EfficientStrings,
WebKit needs to use ASCIILiteral for the string of DEFINE_STATIC_LOCAL.
Source/WebKit/blackberry:
- Api/WebSettings.cpp:
(WebKit):
Source/WebKit/gtk:
- webkit/webkitwebsettings.cpp:
(webkitPlatform):
(webkitOSVersion):
- 3:16 AM Changeset in webkit [127120] by
-
- 2 edits in trunk/LayoutTests
[Qt] REGRESSION(r122175): fast/loader/document-destruction-within-unload.html makes the following test assert
https://bugs.webkit.org/show_bug.cgi?id=95441
Unreviewed gardening.
- platform/qt/Skipped: Skip fast/loader/document-destruction-within-unload.html to paint the bot green.
- 2:56 AM Changeset in webkit [127119] by
-
- 2 edits2 adds in trunk/LayoutTests
[EFL] Gardening for fast/forms/range/thumbslider-no-parent-slider.html after r126864
https://bugs.webkit.org/show_bug.cgi?id=95437
Unreviewed, gardening.
Slider thumb can be displayed without parent slider after r126864.
Patch by KwangYong Choi <ky0.choi@samsung.com> on 2012-08-30
- platform/efl/TestExpectations:
- platform/efl/fast/forms/range/thumbslider-no-parent-slider-expected.png: Added.
- platform/efl/fast/forms/range/thumbslider-no-parent-slider-expected.txt: Added.
- 2:38 AM Changeset in webkit [127118] by
-
- 4 edits in trunk/Source/WebCore
Fix compile warning when enable tiled backing store
https://bugs.webkit.org/show_bug.cgi?id=95422
Patch by Kangil Han <kangil.han@samsung.com> on 2012-08-30
Reviewed by Kentaro Hara.
Fixed compile warning messages when enabled tiled backing store.
In case of TiledBackingStore, it was first thought about static_cast<unsigned>.
However, if minus value is assigned to the comparison, it would be critical.
So, it was modified as using int value in tiled coordinate calculation.
- page/Frame.cpp:
(WebCore::Frame::tiledBackingStorePaintEnd): comparison between signed and unsigned integer expressions [-Wsign-compare]
- platform/graphics/TiledBackingStore.cpp:
(WebCore::TiledBackingStore::invalidate): comparison between signed and unsigned integer expressions [-Wsign-compare]
(WebCore::TiledBackingStore::paint): comparison between signed and unsigned integer expressions [-Wsign-compare]
(WebCore::TiledBackingStore::coverageRatio): comparison between signed and unsigned integer expressions [-Wsign-compare]
(WebCore::TiledBackingStore::createTiles): comparison between signed and unsigned integer expressions [-Wsign-compare]
- platform/graphics/cairo/GLContext.cpp:
(WebCore::GLContext::createOffscreenContext): no return statement in function returning non-void [-Wreturn-type]
- 2:33 AM Changeset in webkit [127117] by
-
- 3 edits2 adds in trunk
Heap-use-after-free in WebCore::ElementV8Internal::onclickAttrGetter
https://bugs.webkit.org/show_bug.cgi?id=94440
Reviewed by Adam Barth.
The problem appears due to onerror callback which resets onclick attribute.
As a part of changing onclick attribute value, previous event listener
gets deref which led to its destruction and hence use-after-free.
Refing it in ::getListenerObject helps to prevent this unfortunate scenario.
Source/WebCore:
Test: fast/events/set-attribute-listener-window-onerror-crash.html
- bindings/v8/V8AbstractEventListener.h:
(WebCore::V8AbstractEventListener::getListenerObject):
LayoutTests:
- fast/events/set-attribute-listener-window-onerror-crash-expected.txt: Added.
- fast/events/set-attribute-listener-window-onerror-crash.html: Added.
- 2:06 AM Changeset in webkit [127116] by
-
- 3 edits in trunk/Source/WebCore
Web Inspector: [Sources] Invisible right sidebar issue
https://bugs.webkit.org/show_bug.cgi?id=94924
Reviewed by Pavel Feldman.
Debugger sidebar resizer is now hidden when sidebar is hidden.
Debugger sidebar show button is moved to the upper right corner in this case.
- inspector/front-end/ScriptsPanel.js:
(WebInspector.ScriptsPanel.prototype._showDebuggerSidebar):
(WebInspector.ScriptsPanel.prototype.set _hideDebuggerSidebar):
- inspector/front-end/scriptsPanel.css:
(button.status-bar-item.scripts-debugger-show-hide-button):
(button.status-bar-item.scripts-debugger-show-hide-button:active):
(button.status-bar-item.scripts-debugger-show-hide-button.toggled-shown):
(button.status-bar-item.scripts-debugger-show-hide-button.toggled-shown:active):
(button.status-bar-item.scripts-debugger-show-hide-button.toggled-hidden):
(button.status-bar-item.scripts-debugger-show-hide-button.toggled-hidden:active):
- 2:01 AM Changeset in webkit [127115] by
-
- 8 edits1 add in trunk/Source/WebCore
Web Inspector: move GeneratedImage members into its own cpp file
https://bugs.webkit.org/show_bug.cgi?id=95351
Reviewed by Yury Semikhatsky.
It is trivial patch. The methods of GeneratedImage were in GeneratorGeneratedImage.cpp file.
It was Ok when it was a single method.
- CMakeLists.txt:
- GNUmakefile.list.am:
- Target.pri:
- WebCore.gypi:
- WebCore.vcproj/WebCore.vcproj:
- WebCore.xcodeproj/project.pbxproj:
- platform/graphics/GeneratedImage.cpp: Added.
(WebCore):
(WebCore::GeneratedImage::computeIntrinsicDimensions):
(WebCore::GeneratedImage::reportMemoryUsage):
- platform/graphics/GeneratorGeneratedImage.cpp:
- 1:38 AM Changeset in webkit [127114] by
-
- 2 edits in trunk/Source/WTF
Build fix for COMPILER(MSVC) && !CPU(X86) after r127001.
- wtf/MathExtras.h:
(lrint): Added additional parentheses to silence compiler warning.
- 1:25 AM Changeset in webkit [127113] by
-
- 2 edits in trunk/Source/WebCore
Build fix for WinCE after r126974.
- platform/graphics/BitmapImage.cpp:
(WebCore::BitmapImage::reportMemoryUsage):
- 1:20 AM Changeset in webkit [127112] by
-
- 8 edits in trunk/Source/WebCore
Replace uses of WTF::String::operator+= with StringBuilder
https://bugs.webkit.org/show_bug.cgi?id=95416
Reviewed by Benjamin Poulain.
WTF::String::operator+= appears to be a sandtrap for contributors. The
vast majority of the callers are using very inefficient string
patterns. This patch removes the use of operator+= in favor of
StringBuilder. Eventually, I'd like to remove operator+= so that more
code doesn't fall into this trap.
- Modules/websockets/WebSocketHandshake.cpp:
(WebCore::resourceName):
- html/HTMLAnchorElement.cpp:
(WebCore::appendServerMapMousePosition):
(WebCore::HTMLAnchorElement::handleClick):
- html/canvas/CanvasRenderingContext2D.cpp:
(WebCore::CanvasRenderingContext2D::font):
- platform/network/ResourceRequestBase.cpp:
(WebCore::ResourceRequestBase::addHTTPHeaderField):
- rendering/RenderLayerBacking.cpp:
(WebCore::RenderLayerBacking::nameForLayer):
- rendering/RenderTreeAsText.cpp:
(WebCore::RenderTreeAsText::writeRenderObject):
(WebCore::nodePosition):
- rendering/style/RenderStyle.cpp:
(WebCore::RenderStyle::setContent):
- 1:06 AM Changeset in webkit [127111] by
-
- 2 edits in trunk/LayoutTests
[EFL] Gardening after r127135, r127039
https://bugs.webkit.org/show_bug.cgi?id=95433
Unreviewed gardening.
Gardening of media/video-controls-captions.html and inspector/timeline/timeline-load.html.
Patch by Mikhail Pozdnyakov <mikhail.pozdnyakov@intel.com> on 2012-08-30
- platform/efl/TestExpectations:
- 12:57 AM Changeset in webkit [127110] by
-
- 2 edits in trunk/LayoutTests
[Qt] Unreviewed gardening, skip new failing tests.
- platform/qt/Skipped:
- 12:56 AM Changeset in webkit [127109] by
-
- 2 edits in trunk/LayoutTests
Unreviewed GTK gardening.
Adding failure expectations for media/video-controls-captions.html
and svg/custom/clamped-masking-clipping.svg, the tests have been failing
since introduced in r127035 and r126993, respectively.
Also adding a flakiness expectation for
fast/layers/scroll-no-visible-content-but-visible-descendant.html.
- platform/gtk/TestExpectations:
- 12:19 AM Changeset in webkit [127108] by
-
- 13 edits3 copies5 adds in trunk
AuthorShadowDOM support for textarea element.
https://bugs.webkit.org/show_bug.cgi?id=91485
Reviewed by Dimitri Glazkov.
Source/WebCore:
We add AuthorShadowDOM support for textarea element.
Unlike other replaced elements (e.g. meter, progress, img), we do not need to add
extra RenderBlock when we add AuthorShadowDOM. However, since inner element will not have
renderer when AuthorShadowDOM does not have any shadow insertion point, we have to check
the existence of the renderer of inner element.
Tests: fast/dom/shadow/shadowdom-for-textarea-with-attribute.html
fast/dom/shadow/shadowdom-for-textarea-with-placeholder.html
fast/dom/shadow/shadowdom-for-textarea-with-style.html
fast/dom/shadow/shadowdom-for-textarea.html
- dom/ShadowRoot.cpp:
(WebCore::allowsAuthorShadowRoot): Needs allow textarea to have AuthorShadowRoot.
- rendering/RenderTextControl.cpp: When AuthorShadowDOM does not have any insertion point,
innerTextElement() will not have any renderer. We have to tweak these renderers not to be crashed.
(WebCore::RenderTextControl::textBlockWidth):
(WebCore::RenderTextControl::updateFromElement):
(WebCore::RenderTextControl::computeLogicalHeight):
(WebCore::RenderTextControl::hitInnerTextElement):
(WebCore::RenderTextControl::computePreferredLogicalWidths):
LayoutTests:
We have the following tests.
(1) having only <shadow> insertion point.
(2) not having <shadow> insertion point.
(3) having <shadow> and <content> insertion point.
(4) with rows/cols attribute
(5) with placeholder attribute
- fast/dom/shadow/shadow-disable-expected.txt:
- fast/dom/shadow/shadow-disable.html:
- fast/dom/shadow/shadowdom-for-textarea-complex-shadow-expected.html:
- fast/dom/shadow/shadowdom-for-textarea-complex-shadow.html:
- fast/dom/shadow/shadowdom-for-textarea-expected.html: Added.
- fast/dom/shadow/shadowdom-for-textarea-only-shadow-expected.html:
- fast/dom/shadow/shadowdom-for-textarea-only-shadow.html:
- fast/dom/shadow/shadowdom-for-textarea-with-attribute-expected.html: Copied from LayoutTests/fast/dom/shadow/shadowdom-for-textarea-only-shadow-expected.html.
- fast/dom/shadow/shadowdom-for-textarea-with-attribute.html: Copied from LayoutTests/fast/dom/shadow/shadowdom-for-textarea-only-shadow.html.
- fast/dom/shadow/shadowdom-for-textarea-with-placeholder-expected.html: Added.
- fast/dom/shadow/shadowdom-for-textarea-with-placeholder.html: Added.
- fast/dom/shadow/shadowdom-for-textarea-with-style-expected.html: Added.
- fast/dom/shadow/shadowdom-for-textarea-with-style.html: Copied from LayoutTests/fast/dom/shadow/shadowdom-for-textarea-only-shadow.html.
- fast/dom/shadow/shadowdom-for-textarea-without-shadow.html:
- fast/dom/shadow/shadowdom-for-textarea.html: Added.
- platform/chromium/TestExpectations:
- platform/efl/TestExpectations:
Aug 29, 2012:
- 11:54 PM Changeset in webkit [127107] by
-
- 2 edits in trunk/LayoutTests
Layout Test fast/repaint/japanese-rl-selection-repaint-in-regions.html is failing after r126304
https://bugs.webkit.org/show_bug.cgi?id=94730
Patch by Andrei Bucur <abucur@adobe.com> on 2012-08-29
Reviewed by Dimitri Glazkov.
Re-enable the japanese-rl-selection-repaint-in-regions.html test for Chromium after the fix was deployed.
- platform/chromium/TestExpectations:
- 11:24 PM Changeset in webkit [127106] by
-
- 7 edits in trunk/Source/WebCore
[V8] ScriptController::matchesCurrentContext duplicates code from ScriptController::currentWorldContext
https://bugs.webkit.org/show_bug.cgi?id=95156
Reviewed by Eric Seidel.
matchesCurrentContext duplicated code from currentWorldContext in order
to avoid creating a new v8::Local handle in the (common) case that
we're already in the right context. This patch just exposes an accessor
for the underlying handle so that the bindings code can do this work
itself.
- bindings/scripts/CodeGeneratorV8.pm:
(GenerateToV8Converters):
- bindings/scripts/test/V8/V8TestActiveDOMObject.cpp:
(WebCore::V8TestActiveDOMObject::wrapSlow):
- bindings/scripts/test/V8/V8TestNode.cpp:
(WebCore::V8TestNode::wrapSlow):
- bindings/v8/ScriptController.cpp:
(WebCore::ScriptController::unsafeHandleToCurrentWorldContext):
(WebCore):
(WebCore::ScriptController::currentWorldContext):
- bindings/v8/ScriptController.h:
(ScriptController):
- bindings/v8/V8DOMWindowShell.h:
(WebCore::V8DOMWindowShell::context):
- 11:21 PM Changeset in webkit [127105] by
-
- 7 edits in trunk/LayoutTests
Unreviewed. [Chromium] Rebaseline fast/forms/time-multiple-fields/time-multiple-fields-appearance-basic.html for Chromium-Mac and Chromium-Win
after changes of https://bugs.webkit.org/show_bug.cgi?id=95285
- platform/chromium-mac-snowleopard/fast/forms/time-multiple-fields/time-multiple-fields-appearance-basic-expected.png:
- platform/chromium-mac/fast/forms/time-multiple-fields/time-multiple-fields-appearance-basic-expected.png:
- platform/chromium-mac/fast/forms/time-multiple-fields/time-multiple-fields-appearance-basic-expected.txt:
- platform/chromium-win/fast/forms/time-multiple-fields/time-multiple-fields-appearance-basic-expected.png:
- platform/chromium-win/fast/forms/time-multiple-fields/time-multiple-fields-appearance-basic-expected.txt:
- platform/chromium/TestExpectations:
- 11:19 PM Changeset in webkit [127104] by
-
- 2 edits in trunk/Tools
Random test-webkitpy failures on the buildbot
https://bugs.webkit.org/show_bug.cgi?id=95096
Reviewed by Dirk Pranke.
Suppress occasional errors when running test-webkitpy on GTK builders
by running these tests serially. The 64-bit Release builder is especially
prone to these as it can run up to 24 tests in parallel.
Despite the tests not being run in parallel, the testing only lasts a handful
of seconds more, so this is a worthy trade-off to avoid unnecessary
false-alarm redness on the GTK builders.
- BuildSlaveSupport/build.webkit.org-config/master.cfg:
(RunPythonTests.start):
- 10:50 PM Changeset in webkit [127103] by
-
- 2 edits in trunk/Source/WebKit/chromium
Fix Chromium builds (Win and Mac)
https://bugs.webkit.org/show_bug.cgi?id=95421
Reviewed by James Robinson.
This change replaces 'class WebRect' with 'struct WebRect' to fix build breaks
on the "Chromium Mac Release" and the "Chromium Win Release" bot caused by
r127095.
- public/WebViewClient.h:
(WebKit):
- 10:45 PM Changeset in webkit [127102] by
-
- 4 edits in trunk/Source/WebKit2
[EFL][WK2] Replace PageClientImpl with ewk view in constructor of EflViewportHandler.
https://bugs.webkit.org/show_bug.cgi?id=95408
Patch by Kangil Han <kangil.han@samsung.com> on 2012-08-29
Reviewed by Gyuyoung Kim.
To keep consistency of implementation, derived classes(from ewk view) should have view reference.
From this, derived classes would have less interference from changes that would happen in port specific classes, i.e. PageClientImpl.
- UIProcess/API/efl/EflViewportHandler.cpp:
(WebKit::EflViewportHandler::EflViewportHandler):
(WebKit::EflViewportHandler::drawingArea):
(WebKit):
(WebKit::EflViewportHandler::updateViewportSize):
- UIProcess/API/efl/EflViewportHandler.h:
(WebKit::EflViewportHandler::create):
(EflViewportHandler):
- UIProcess/API/efl/ewk_view.cpp:
(_ewk_view_initialize):
- 10:19 PM Changeset in webkit [127101] by
-
- 7 edits10 adds in trunk/LayoutTests
WebPImageDecoder progressive decodes fail to decode valid images
https://bugs.webkit.org/show_bug.cgi?id=74062
Reviewed by Adam Barth.
Add test for a partial image load, and a test for a complete image load with a
pause to cause a progressive/incremental webp image decode.
- fast/images/resources/large.webp: Added.
- http/tests/images/webp-partial-load-expected.txt: Added.
- http/tests/images/webp-partial-load.html: Added.
- http/tests/images/webp-progressive-load-expected.txt: Added.
- http/tests/images/webp-progressive-load.html: Added.
- platform/chromium-linux/http/tests/images/webp-partial-load-expected.png: Added.
Result differs because the linux DRT network layers returns partial encoded image
data from the cache layer in multiples of 2048 bytes until all data is received.
- platform/chromium/http/tests/images/webp-partial-load-expected.png: Added.
- platform/chromium/http/tests/images/webp-progressive-load-expected.png: Added.
- platform/efl/TestExpectations: This port does not support webp images.
- platform/gtk/TestExpectations: ditto.
- platform/mac/Skipped: ditto.
- platform/qt/Skipped: ditto.
- platform/win/Skipped: ditto.
- platform/wincairo/Skipped: ditto.
- 10:08 PM Changeset in webkit [127100] by
-
- 3 edits in trunk/LayoutTests
security/crypto-random-values-limits.html test the exact value that causes us to throw
https://bugs.webkit.org/show_bug.cgi?id=95403
Reviewed by Eric Seidel.
As requested by Darin Adler, this patch makes us test the exact length
threshold at which we start throwing an exception. This value is
defined in the spec, and we want to make sure we hit it exactly.
- security/crypto-random-values-limits-expected.txt:
- security/crypto-random-values-limits.html:
- 10:03 PM Changeset in webkit [127099] by
-
- 5 edits in trunk/Source
[chromium] setNeedsAnimate should not cause commitRequested to become true
https://bugs.webkit.org/show_bug.cgi?id=95393
Reviewed by James Robinson.
Source/WebCore:
We use the commitRequested state to determine if the page has been damaged, which
then is used by the input flow control logic to coalesce input events. However, we
actually have two notions of commitRequested. At the CCLayerTreeHost level, commit
being requested means "we've changed the tree in some way." At the proxy level, it
means "we've sent a commit request to the impl thread." Without this patch,
we use the latter state to answer ::commitRequested. That causes setNeedsAnimate
to incorrectly cause the commitRequested bit to be set.
This change separates the setNeedsCommit state from commitRequestSentToImplThread.
This allows us to correctly answer commitRequested in face of mixed animation and
invalidation requests.
- platform/graphics/chromium/cc/CCThreadProxy.cpp:
(WebCore::CCThreadProxy::CCThreadProxy):
(WebCore::CCThreadProxy::setNeedsAnimate):
(WebCore::CCThreadProxy::setNeedsCommit):
(WebCore::CCThreadProxy::beginFrame):
- platform/graphics/chromium/cc/CCThreadProxy.h:
(CCThreadProxy):
Source/WebKit/chromium:
- tests/CCLayerTreeHostTest.cpp:
(CCLayerTreeHostTestSetNeedsAnimateShouldNotSetCommitRequested):
(CCLayerTreeHostTestSetNeedsAnimateShouldNotSetCommitRequested::CCLayerTreeHostTestSetNeedsAnimateShouldNotSetCommitRequested):
- 9:53 PM Changeset in webkit [127098] by
-
- 9 edits in trunk/Source/WebKit2
[WK2] Use ASCIILiteral hotness for DEFINE_STATIC_LOCAL string
https://bugs.webkit.org/show_bug.cgi?id=95318
Reviewed by Benjamin Poulain.
As recommended by http://trac.webkit.org/wiki/EfficientStrings,
WebKit2 needs to use ASCIILiteral for the string of DEFINE_STATIC_LOCAL.
- Shared/WebError.cpp:
(WebKit::WebError::webKitErrorDomain):
- Shared/WebPreferencesStore.cpp:
(WebPreferencesKey):
- UIProcess/InspectorServer/WebInspectorServer.cpp:
(WebKit::WebInspectorServer::didReceiveWebSocketUpgradeHTTPRequest):
- UIProcess/InspectorServer/qt/WebInspectorServerQt.cpp:
(WebKit::remoteInspectorPagePath):
- UIProcess/WebDatabaseManagerProxy.cpp:
(WebKit::WebDatabaseManagerProxy::originKey):
(WebKit::WebDatabaseManagerProxy::originQuotaKey):
(WebKit::WebDatabaseManagerProxy::originUsageKey):
(WebKit::WebDatabaseManagerProxy::databaseDetailsKey):
(WebKit::WebDatabaseManagerProxy::databaseDetailsNameKey):
(WebKit::WebDatabaseManagerProxy::databaseDetailsDisplayNameKey):
(WebKit::WebDatabaseManagerProxy::databaseDetailsExpectedUsageKey):
(WebKit::WebDatabaseManagerProxy::databaseDetailsCurrentUsageKey):
- UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::executeEditCommand):
- WebProcess/WebCoreSupport/WebEditorClient.cpp:
(WebKit::WebEditorClient::didBeginEditing):
(WebKit::WebEditorClient::respondToChangedContents):
(WebKit::WebEditorClient::respondToChangedSelection):
(WebKit::WebEditorClient::didEndEditing):
- WebProcess/WebProcess.cpp:
(WebKit::getWebCoreMemoryCacheStatistics):
- 9:47 PM Changeset in webkit [127097] by
-
- 3 edits in trunk/LayoutTests
[Qt][WK1] New test css3/filters/null-effect-check.html added in r126927 fails
https://bugs.webkit.org/show_bug.cgi?id=95308
Reviewed by Simon Hausmann.
- css3/filters/null-effect-check.html: Modify this reftest not to be dumpAsText() test.
- platform/qt-5.0-wk1/Skipped: Unskip the now passing test.
- 9:33 PM Changeset in webkit [127096] by
-
- 2 edits in trunk/Source/WebCore
[Qt] REGRESSION(r126694): It broke the debug build
https://bugs.webkit.org/show_bug.cgi?id=95037
Unreviewed trivial build fix for debug builds.
Don't pass WTFStrings through printf, use .ascii().data().
Patch by Simon Hausmann <simon.hausmann@nokia.com> on 2012-08-29
- platform/graphics/texmap/TextureMapperShaderManager.cpp:
(WebCore::StandardFilterProgram::StandardFilterProgram):
- 9:08 PM Changeset in webkit [127095] by
-
- 5 edits3 adds in trunk/Source
[chromium] Implement disambiguation popup (a.k.a. Link Preview)
https://bugs.webkit.org/show_bug.cgi?id=94182
Patch by Tien-Ren Chen <trchen@chromium.org> on 2012-08-29
Reviewed by Adam Barth.
In this new implementation, we add a new WebViewClient::handleDisambiguationPopup delegate.
The disambiguation sequence will be initiated by the gesture event handler
in WebViewImpl if an ambiguous tap is detected, then
m_client->handleDisambiguationPopup will be called, so the embedder can
decide whether to swallow the touch event and show a popup.
New test: WebFrameTest.DisambiguationPopupTest
- WebKit.gyp:
- features.gypi:
- public/WebInputEvent.h:
(WebGestureEvent):
(WebKit::WebGestureEvent::WebGestureEvent):
- public/WebTouchCandidatesInfo.h: Removed.
- public/WebView.h:
(WebKit):
- public/WebViewClient.h:
(WebKit):
(WebViewClient):
(WebKit::WebViewClient::triggersLinkPreview):
- src/WebInputEvent.cpp:
(SameSizeAsWebGestureEvent):
- src/WebViewImpl.cpp:
(WebKit):
(WebKit::WebViewImpl::handleGestureEventWithLinkPreview):
(WebKit::WebViewImpl::handleGestureEvent):
- src/WebViewImpl.h:
(WebViewImpl):
- 9:03 PM Changeset in webkit [127094] by
-
- 2 edits in trunk/LayoutTests
[Chromium-Android] Upstream layout test expectations (Part 3)
https://bugs.webkit.org/show_bug.cgi?id=95392
Reviewed by Adam Barth.
This part include:
- media tests
- slow tests
- other misc updates
- platform/chromium/TestExpectations:
- 8:56 PM Changeset in webkit [127093] by
-
- 4 edits in trunk
REGRESSION(r126780): Crash using StringImpl::is8Bit before checking if there is an impl
https://bugs.webkit.org/show_bug.cgi?id=95380
Patch by Benjamin Poulain <bpoulain@apple.com> on 2012-08-29
Reviewed by Michael Saboff.
Source/WTF:
Blindly copying code from UString in r126780 was stupid. I just brought over a bug.
This patch adds the zero length branch back so that null strings are handled correctly.
- wtf/text/WTFString.cpp:
(WTF::String::ascii): Return a empty CString if the String is null or empty.
Tools:
- TestWebKitAPI/Tests/WTF/WTFString.cpp:
Add very basic tests for String::ascii(). This covers the case of null strings that caused
the crash.
- 8:48 PM Changeset in webkit [127092] by
-
- 2 edits in trunk/Source/WebCore
Fix assertion on Document::recalcStyle to not recalc style while painting
https://bugs.webkit.org/show_bug.cgi?id=95386
Reviewed by Eric Seidel.
- dom/Document.cpp: Move assertion outside the if to reflect that safety check.
- 8:30 PM Changeset in webkit [127091] by
-
- 4 edits in trunk
[Qt][WK2] ApplicationCache LayoutTests failed
https://bugs.webkit.org/show_bug.cgi?id=69541
Patch by Luciano Wolf <Luciano Miguel Wolf> on 2012-08-29
Reviewed by Kenneth Rohde Christiansen.
Source/WebKit2:
Returns defaultDiskCacheDirectory when no cache directory was provided.
It's used by setOfflineWebApplicationCacheEnabled method that won't work
with an invalid cache directory.
- UIProcess/qt/WebContextQt.cpp:
(WebKit::WebContext::applicationCacheDirectory):
LayoutTests:
Unskip http/appcache tests for qt-5.0-wk2.
- platform/qt-5.0-wk2/Skipped:
- 8:23 PM Changeset in webkit [127090] by
-
- 2 edits in trunk/Source/JavaScriptCore
Remove use of ClassInfo from compileGetByValOnArguments and compileGetArgumentsLength
https://bugs.webkit.org/show_bug.cgi?id=95131
Reviewed by Filip Pizlo.
- dfg/DFGSpeculativeJIT.cpp:
(JSC::DFG::SpeculativeJIT::compileGetByValOnArguments): We don't need this speculation check. We can replace it
with an assert to guarantee this.
- 8:22 PM Changeset in webkit [127089] by
-
- 5 edits in trunk/LayoutTests
[Tests] Add test case for right-to-left rendering of multiple fields time input UI
https://bugs.webkit.org/show_bug.cgi?id=95285
Reviewed by Kent Tamura.
This patch adds a new test case of appearance of dir="rtl" case to
multiple fields time input UI.
This patch affects ports which enable both ENABLE_INPUT_TYPE_TIME and
ENABLE_INPUT_TYPE_TIME_MULTIPLE_FIELDS.
Note: I'll do rebaseline for Chromium-Mac and Chromium-Win.
- fast/forms/time-multiple-fields/time-multiple-fields-appearance-basic.html: Added an input element with dir="rtl".
- platform/chromium-linux/fast/forms/time-multiple-fields/time-multiple-fields-appearance-basic-expected.png: Updated for a new input element.
- platform/chromium-linux/fast/forms/time-multiple-fields/time-multiple-fields-appearance-basic-expected.txt: ditto.
- platform/chromium/TestExpectations: Disabled time-multiple-fields-appearance-basic.html for Chromium-Mac and Chromium-Win.
- 8:12 PM Changeset in webkit [127088] by
-
- 11 edits in trunk
Add WebKit prefix to MediaSource, SourceBuffer, & SourceBufferList DOMWindow constructor attributes.
https://bugs.webkit.org/show_bug.cgi?id=95247
Patch by Aaron Colwell <acolwell@chromium.org> on 2012-08-29
Reviewed by Eric Carlson.
Source/WebCore:
Add WebKit prefix to MediaSource, SourceBuffer, and SourceBufferList object constructor attributes.
Covered by existing layout tests..
- page/DOMWindow.idl:
LayoutTests:
Update MediaSource tests to include the WebKit prefix on MediaSource, SourceBuffer, & SourceBufferTest objects.
- http/tests/media/media-source/media-source.js:
- http/tests/media/media-source/video-media-source-add-and-remove-buffers.html:
- http/tests/media/media-source/video-media-source-async-events.html:
- http/tests/media/media-source/video-media-source-event-attributes.html:
- http/tests/media/media-source/video-media-source-objects-expected.txt:
- http/tests/media/media-source/video-media-source-objects.html:
- http/tests/media/media-source/video-media-source-play.html:
- http/tests/media/media-source/video-media-source-state-changes.html:
- 7:58 PM Changeset in webkit [127087] by
-
- 3 edits in trunk/Source/WebCore
ProgressTracker never completes if iframe detached during parsing
https://bugs.webkit.org/show_bug.cgi?id=92272
Reviewed by Adam Barth.
Add a simple helper class to FrameLoader to ensure progressStarted/progressCompleted calls are matched,
and balance the calls when the Frame is detached.
No new tests, as this behavior has only been producing reliably by setting a breakpoint in a specific place.
- loader/FrameLoader.cpp:
(FrameLoader::FrameProgressTracker):
(WebCore::FrameLoader::FrameProgressTracker::create):
(WebCore::FrameLoader::FrameProgressTracker::~FrameProgressTracker):
(WebCore::FrameLoader::FrameProgressTracker::progressStarted):
(WebCore::FrameLoader::FrameProgressTracker::progressCompleted):
(WebCore::FrameLoader::FrameProgressTracker::FrameProgressTracker):
(WebCore):
(WebCore::FrameLoader::init):
(WebCore::FrameLoader::prepareForLoadStart):
(WebCore::FrameLoader::clearProvisionalLoad):
(WebCore::FrameLoader::checkLoadCompleteForThisFrame):
(WebCore::FrameLoader::detachFromParent):
- loader/FrameLoader.h:
(FrameLoader):
- 7:49 PM Changeset in webkit [127086] by
-
- 2 edits in trunk/Source/WebKit/blackberry
[BlackBerry] Do not use the shadow tree when retrieving the underlying element for FatFinger.
https://bugs.webkit.org/show_bug.cgi?id=95372
By using the shadow tree we were getting a mismatch between elements of the
current element under focus and the one return to us from FatFingers.
Passing ShadowContentNotAllowed to get the right handle.
Patch by Nima Ghanavatian <nghanavatian@rim.com> on 2012-08-29
Reviewed by Antonio Gomes.
- WebKitSupport/TouchEventHandler.cpp:
(BlackBerry::WebKit::TouchEventHandler::handleTouchPoint):
- 7:24 PM Changeset in webkit [127085] by
-
- 10 edits in trunk/Source/WebCore
[Forms] Rename DateTimeFieldElement::FieldEventHandler to FieldOwner
https://bugs.webkit.org/show_bug.cgi?id=95280
Reviewed by Kent Tamura.
This patch renames DateTimeFieldElement::FieldEventHandler to FieldOwner
for matching functionaly of class and class name to add functions like
focusOnNextField().
This patch affects ports which enables both ENABLE_INPUT_TYPE_TIME and
ENABLE_INPUT_TYPE_TIME_MULTIPLE_FIELDS.
This patch is a part of changing Shift+Tab focus navigation of
multiple fields input time UI, bug 95168.
No new tests. This patch doesn't change behavior.
- html/shadow/DateTimeEditElement.h:
(DateTimeEditElement):
- html/shadow/DateTimeFieldElement.cpp:
(WebCore::DateTimeFieldElement::FieldOwner::~FieldOwner):
(WebCore::DateTimeFieldElement::DateTimeFieldElement):
(WebCore::DateTimeFieldElement::focusOnNextField):
(WebCore::DateTimeFieldElement::updateVisibleValue):
- html/shadow/DateTimeFieldElement.h:
(FieldOwner):
(WebCore::DateTimeFieldElement::removeEventHandler):
(DateTimeFieldElement):
- html/shadow/DateTimeFieldElements.cpp:
(WebCore::DateTimeAMPMFieldElement::DateTimeAMPMFieldElement):
(WebCore::DateTimeAMPMFieldElement::create):
(WebCore::DateTimeHourFieldElement::DateTimeHourFieldElement):
(WebCore::DateTimeHourFieldElement::create):
(WebCore::DateTimeMillisecondFieldElement::DateTimeMillisecondFieldElement):
(WebCore::DateTimeMillisecondFieldElement::create):
(WebCore::DateTimeMinuteFieldElement::DateTimeMinuteFieldElement):
(WebCore::DateTimeMinuteFieldElement::create):
(WebCore::DateTimeSecondFieldElement::DateTimeSecondFieldElement):
(WebCore::DateTimeSecondFieldElement::create):
- html/shadow/DateTimeFieldElements.h:
(DateTimeAMPMFieldElement):
(DateTimeHourFieldElement):
(DateTimeMillisecondFieldElement):
(DateTimeMinuteFieldElement):
(DateTimeSecondFieldElement):
- html/shadow/DateTimeNumericFieldElement.cpp:
(WebCore::DateTimeNumericFieldElement::DateTimeNumericFieldElement):
- html/shadow/DateTimeNumericFieldElement.h:
(DateTimeNumericFieldElement):
- html/shadow/DateTimeSymbolicFieldElement.cpp:
(WebCore::DateTimeSymbolicFieldElement::DateTimeSymbolicFieldElement):
- html/shadow/DateTimeSymbolicFieldElement.h:
(DateTimeSymbolicFieldElement):
- 7:14 PM Changeset in webkit [127084] by
-
- 16 edits2 moves3 adds in trunk
AX: Canvas should have a distinct role
https://bugs.webkit.org/show_bug.cgi?id=95248
Reviewed by Chris Fleizach.
Source/WebCore:
Add new role for a canvas element, and a method to determine if
a canvas has fallback content, so each platform can decide on the
appropriate role mapping to use.
Test: accessibility/canvas-description-and-role.html
- accessibility/AccessibilityNodeObject.cpp:
(WebCore::AccessibilityNodeObject::canvasHasFallbackContent):
(WebCore):
- accessibility/AccessibilityNodeObject.h:
(AccessibilityNodeObject):
- accessibility/AccessibilityObject.h:
(WebCore::AccessibilityObject::isCanvas):
(WebCore::AccessibilityObject::canvasHasFallbackContent):
- accessibility/AccessibilityRenderObject.cpp:
(WebCore::AccessibilityRenderObject::accessibilityIsIgnored):
(WebCore::AccessibilityRenderObject::determineAccessibilityRole):
(WebCore::AccessibilityRenderObject::canHaveChildren):
- accessibility/gtk/WebKitAccessibleWrapperAtk.cpp:
(atkRole):
- accessibility/mac/WebAccessibilityObjectWrapper.mm:
(createAccessibilityRoleMap):
(-[WebAccessibilityObjectWrapper role]):
Source/WebKit/chromium:
Add support for canvas accessibility role.
- public/WebAccessibilityRole.h:
- src/AssertMatchingEnums.cpp:
Source/WebKit/win:
Map new CanvasRole to the same as ImageRole.
- AccessibleBase.cpp:
(MSAARole):
Tools:
Add support for canvas accessibility role.
- DumpRenderTree/chromium/TestRunner/AccessibilityUIElementChromium.cpp:
(roleToString):
LayoutTests:
Add new tests for canvas role.
- accessibility/canvas.html: Deleted.
- accessibility/canvas-expected.txt: Deleted.
- accessibility/canvas-description-and-role.html: Added.
- platform/chromium/accessibility/canvas-description-and-role-expected.txt: Added.
- platform/gtk/TestExpectations:
- platform/mac/accessibility/canvas.html: Added.
- platform/mac/accessibility/canvas-expected.txt: Added.
- platform/mac/accessibility/canvas-description-and-role-expected.txt: Added.
- 7:07 PM Changeset in webkit [127083] by
-
- 4 edits in trunk/Source/WebKit/blackberry
[BlackBerry] Move the handling of UserMedia requests to the client
https://bugs.webkit.org/show_bug.cgi?id=95317
PR #197283
Patch by Robin Cao <robin.cao@torchmobile.com.cn> on 2012-08-29
Reviewed by George Staikos.
Internally reviewed by George Staikos.
Move the logic for user media requests handling to the client, as it requires
inputs from users. Also adds a origin field to WebUserMediaRequest.
- Api/WebPageClient.h:
(Platform):
- WebCoreSupport/UserMediaClientImpl.cpp:
(WebCore::UserMediaClientImpl::UserMediaClientImpl):
(WebCore::UserMediaClientImpl::~UserMediaClientImpl):
(WebCore::UserMediaClientImpl::requestUserMedia):
(WebCore::UserMediaClientImpl::cancelUserMediaRequest):
- WebCoreSupport/UserMediaClientImpl.h:
(UserMediaClientImpl):
- 7:05 PM Changeset in webkit [127082] by
-
- 3 edits4 adds in trunk
use after free in WebCore::FileReader::doAbort
https://bugs.webkit.org/show_bug.cgi?id=91004
Reviewed by Jian Li.
Source/WebCore:
Added check in FileReader::abort to not process the abort if we aren't in the LOADING
state. This is per the FileAPI spec section 8.5.6 step #1.
Tests: fast/files/file-reader-immediate-abort.html
fast/files/file-reader-done-reading-abort.html
- fileapi/FileReader.cpp:
(WebCore::FileReader::abort):
LayoutTests:
New tests to check that FileReader::abort doesn't crash or create events before
or after reading.
- fast/files/file-reader-done-reading-abort-expected.txt: Added.
- fast/files/file-reader-done-reading-abort.html: Added.
- fast/files/file-reader-immediate-abort-expected.txt: Added.
- fast/files/file-reader-immediate-abort.html: Added.
- 7:02 PM Changeset in webkit [127081] by
-
- 4 edits in trunk/Source
[chromium] CCLayerTreeHost::finishCommitOnImplThread wrong setter order
https://bugs.webkit.org/show_bug.cgi?id=94828
Patch by Alex Sakhartchouk <alexst@chromium.org> on 2012-08-29
Reviewed by Adrienne Walker.
Source/WebCore:
setDeviceScaleFactor affects maxScrollPosition, making sure it's properly updated.
This also removes setter order dependency.
- platform/graphics/chromium/cc/CCLayerTreeHostImpl.cpp:
(WebCore::CCLayerTreeHostImpl::setDeviceScaleFactor):
Source/WebKit/chromium:
Testing that setDeviceScaleFactor properly changes maxScrollPosition
- tests/CCLayerTreeHostImplTest.cpp:
- 6:59 PM Changeset in webkit [127080] by
-
- 2 edits in trunk/LayoutTests
ASSERTION FAILURE in JSC::JSGlobalData::float32ArrayDescriptor when running fast/js/dfg-float64array.html
https://bugs.webkit.org/show_bug.cgi?id=95398
Skip the tests affected by this assertion failure.
- platform/mac-wk2/Skipped:
- 6:58 PM Changeset in webkit [127079] by
-
- 11 edits in trunk/Source
Unreviewed, rolling out r126956.
http://trac.webkit.org/changeset/126956
https://bugs.webkit.org/show_bug.cgi?id=94721
Source/WebCore:
Breaks several unit tests - see https://bugs.webkit.org/show_bug.cgi?id=95358 for instance.
- platform/graphics/chromium/cc/CCScheduler.cpp:
(WebCore::CCScheduler::CCScheduler):
(WebCore::CCScheduler::beginFrameComplete):
(WebCore::CCScheduler::vsyncTick):
(WebCore::CCScheduler::processScheduledActions):
- platform/graphics/chromium/cc/CCScheduler.h:
(CCSchedulerClient):
(CCScheduler):
- platform/graphics/chromium/cc/CCTextureUpdateController.cpp:
(WebCore::CCTextureUpdateController::CCTextureUpdateController):
(WebCore::CCTextureUpdateController::hasMoreUpdates):
(WebCore):
(WebCore::CCTextureUpdateController::updateMoreTextures):
(WebCore::CCTextureUpdateController::onTimerFired):
(WebCore::CCTextureUpdateController::updateMoreTexturesIfEnoughTimeRemaining):
- platform/graphics/chromium/cc/CCTextureUpdateController.h:
(WebCore):
(WebCore::CCTextureUpdateController::create):
(CCTextureUpdateController):
- platform/graphics/chromium/cc/CCThreadProxy.cpp:
(WebCore::CCThreadProxy::didLoseContextOnImplThread):
(WebCore::CCThreadProxy::beginFrameCompleteOnImplThread):
(WebCore::CCThreadProxy::hasMoreResourceUpdates):
(WebCore):
(WebCore::CCThreadProxy::scheduledActionCommit):
- platform/graphics/chromium/cc/CCThreadProxy.h:
(WebCore):
Source/WebKit/chromium:
Breaks several unit tests
- tests/CCSchedulerTest.cpp:
(WebKitTests::FakeCCSchedulerClient::reset):
(WebKitTests::FakeCCSchedulerClient::setHasMoreResourceUpdates):
(WebKitTests::TEST):
- tests/CCTextureUpdateControllerTest.cpp:
- 6:44 PM Changeset in webkit [127078] by
-
- 1 edit1 add in trunk/LayoutTests
Mountain Lion missing results for media/controls-without-preload.html
https://bugs.webkit.org/show_bug.cgi?id=95417
Add the text results the bots were seeing to get the bots greener.
- platform/mac/media/controls-without-preload-expected.txt: Added.
- 6:34 PM Changeset in webkit [127077] by
-
- 9 edits2 adds in trunk
[chromium] Register/unregister contents layers with GraphicsLayerChromium
https://bugs.webkit.org/show_bug.cgi?id=95379
Reviewed by Adrienne Walker.
Source/WebCore:
Several composited layer types in WebCore are represented by a painted layer and a child "contents layer" that
represents some non-painted specific content type. For example, a composited video has a WebCore-painted layer
for CSS background and border effects and a child platform video layer backed by a WebVideoLayer with the output
of the video decoding pipeline. Cross-platform code associates the PlatformLayer from the various composited
systems with the right GraphicsLayer, but the object owning the layer and the GraphicsLayer holding the pointer
otherwise have no relationship. This makes shutdown a bit tricky since the object destroying the contents layer
has no direct way to notify the GraphicsLayer holding the contents layer pointer that it is going away. The
GraphicsLayer will be notified after the next style recalc that its contents layer is gone, but may need to do
any number of bookkeeping operations before that happens.
On most platforms the PlatformLayer is refcounted, so the GraphicsLayer simply holds a ref to its contents layer
from the time it is orphaned until the next style recalc and compositing tree rebuild. In Chromium, however,
PlatformLayer is not refcounted. This adds an explicit registration mechanism for layers that may be contents
layers. A layer has to be registered with GraphicsLayerChromium before it can be used as a contents layer -
typically this is just done at creation - and unregistered before it is destroyed.
Tests: fast/canvas/transformed-canvas-reset.html
platform/chromium/virtual/gpu/fast/canvas/transformed-canvas-reset.html
- page/scrolling/chromium/ScrollingCoordinatorChromium.cpp:
(WebCore::ScrollingCoordinatorPrivate::~ScrollingCoordinatorPrivate):
(WebCore::createScrollbarLayer):
- platform/graphics/chromium/Canvas2DLayerBridge.cpp:
(WebCore::Canvas2DLayerBridge::Canvas2DLayerBridge):
(WebCore::Canvas2DLayerBridge::~Canvas2DLayerBridge):
- platform/graphics/chromium/DrawingBufferChromium.cpp:
(WebCore::DrawingBufferPrivate::DrawingBufferPrivate):
(WebCore::DrawingBufferPrivate::~DrawingBufferPrivate):
- platform/graphics/chromium/GraphicsLayerChromium.cpp:
(WebCore::GraphicsLayerChromium::updateNames):
(WebCore::GraphicsLayerChromium::clearBackgroundColor):
(WebCore::GraphicsLayerChromium::setContentsNeedsDisplay):
(WebCore::GraphicsLayerChromium::setContentsToImage):
(WebCore):
(WebCore::GraphicsLayerChromium::registerContentsLayer):
(WebCore::GraphicsLayerChromium::unregisterContentsLayer):
(WebCore::GraphicsLayerChromium::clearContentsLayerIfUnregistered):
(WebCore::GraphicsLayerChromium::setContentsTo):
(WebCore::GraphicsLayerChromium::updateChildList):
(WebCore::GraphicsLayerChromium::updateLayerIsDrawable):
(WebCore::GraphicsLayerChromium::updateLayerBackgroundColor):
(WebCore::GraphicsLayerChromium::updateContentsRect):
- platform/graphics/chromium/GraphicsLayerChromium.h:
(GraphicsLayerChromium):
Source/WebKit/chromium:
- src/WebMediaPlayerClientImpl.cpp:
(WebKit::WebMediaPlayerClientImpl::~WebMediaPlayerClientImpl):
(WebKit::WebMediaPlayerClientImpl::readyStateChanged):
- src/WebPluginContainerImpl.cpp:
(WebKit::WebPluginContainerImpl::setBackingTextureId):
(WebKit::WebPluginContainerImpl::setBackingIOSurfaceId):
(WebKit::WebPluginContainerImpl::~WebPluginContainerImpl):
- 6:23 PM Changeset in webkit [127076] by
-
- 5 edits in trunk/Source/WebKit/chromium
[chromium] Add WebSettings support for flag to enable/disable gesture tap highlights.
https://bugs.webkit.org/show_bug.cgi?id=95119
Reviewed by Adam Barth.
Add support to WebSettings for flag to enable/disable gesture tap highlights. Relies on existing tests.
- public/WebSettings.h:
- src/WebSettingsImpl.cpp:
(WebKit::WebSettingsImpl::WebSettingsImpl):
(WebKit::WebSettingsImpl::setGestureTapHighlightEnabled):
(WebKit):
- src/WebSettingsImpl.h:
(WebSettingsImpl):
(WebKit::WebSettingsImpl::gestureTapHighlightEnabled):
- src/WebViewImpl.cpp:
(WebKit::WebViewImpl::handleGestureEvent):
- 6:18 PM EfficientStrings edited by
- (diff)
- 6:16 PM Changeset in webkit [127075] by
-
- 5 edits in trunk/LayoutTests
Unreviewed update of circle-in-mask-with-shadow test expectation.
- platform/chromium-linux/svg/css/circle-in-mask-with-shadow-expected.png:
- platform/chromium-mac/svg/css/circle-in-mask-with-shadow-expected.png: Modified property svn:mime-type.
- platform/chromium-win/svg/css/circle-in-mask-with-shadow-expected.png:
- platform/chromium/TestExpectations:
- 6:14 PM Changeset in webkit [127074] by
-
- 4 edits in branches/safari-536.26-branch/Source
Versioning.
- 6:13 PM Changeset in webkit [127073] by
-
- 1 copy in tags/Safari-536.26.11
New tag.
- 6:06 PM Changeset in webkit [127072] by
-
- 2 edits in trunk/Source/WebCore
[Texmap] CSS Transform flicks at the end of animation
https://bugs.webkit.org/show_bug.cgi?id=95347
Reviewed by Noam Rosenthal.
- platform/graphics/GraphicsLayerAnimation.cpp: Check if the last loop has
been completed and then use 1.0 as normalized value for the progress, otherwise
it would work as if there was a new loop forward and then cycle the progress value.
- 5:42 PM Changeset in webkit [127071] by
-
- 3 edits2 adds in trunk
Crash in WebCore::StyleSheetContents::checkLoadCompleted.
https://bugs.webkit.org/show_bug.cgi?id=95106
Reviewed by Antti Koivisto.
Source/WebCore:
RefPtr StyleSheetContents since it can get blown away in script execution inside
sheetLoaded().
Test: fast/css/style-element-process-crash.html
- css/StyleSheetContents.cpp:
(WebCore::StyleSheetContents::checkLoadCompleted):
LayoutTests:
- fast/css/style-element-process-crash-expected.txt: Added.
- fast/css/style-element-process-crash.html: Added.
- 5:34 PM Changeset in webkit [127070] by
-
- 20 edits6 adds in trunk
Source/WebCore: [Gtk] Process Gtk 3.4 smooth scroll events properly.
https://bugs.webkit.org/show_bug.cgi?id=88070
Gtk 3.3.18 added smooth scroll events, adding a new scroll direction that
provides detailed delta information.
Added GDK_SMOOTH_SCROLL_MASK to the events listened, and added
code to process properly the new direction GDK_SCROLL_SMOOTH and
its deltas.
Patch by José Dapena Paz <jdapena@igalia.com> on 2012-08-29
Reviewed by Martin Robinson.
Test: fast/events/continuous-platform-wheelevent-in-scrolling-div.html
- platform/gtk/PlatformWheelEventGtk.cpp:
(WebCore::PlatformWheelEvent::PlatformWheelEvent):
Source/WebKit/gtk: [Gtk] Process Gtk 3.4 smooth scroll events properly.
https://bugs.webkit.org/show_bug.cgi?id=88070
Gtk 3.3.18 added smooth scroll events, adding a new scroll direction that
provides detailed delta information.
Added GDK_SMOOTH_SCROLL_MASK to the events listened, and added
code to process properly the new direction GDK_SCROLL_SMOOTH and
its deltas.
Patch by José Dapena Paz <jdapena@igalia.com> on 2012-08-29
Reviewed by Martin Robinson.
- webkit/webkitwebview.cpp:
(webkit_web_view_realize):
Source/WebKit2: [Gtk] Process Gtk 3.4 smooth scroll events properly.
https://bugs.webkit.org/show_bug.cgi?id=88070
Gtk 3.3.18 added smooth scroll events, adding a new scroll direction that
provides detailed delta information.
Added GDK_SMOOTH_SCROLL_MASK to the events listened, and added
code to process properly the new direction GDK_SCROLL_SMOOTH and
its deltas.
Patch by José Dapena Paz <jdapena@igalia.com> on 2012-08-29
Reviewed by Martin Robinson.
- Shared/gtk/WebEventFactory.cpp:
(WebKit::WebEventFactory::createWebWheelEvent):
- UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseRealize):
Tools: [Gtk] Process Gtk 3.4 smooth scroll events properly
https://bugs.webkit.org/show_bug.cgi?id=88070
Added continousMouseScrollBy support in WebKitTestRunner, and added
implementation for gtk, and stub for Qt, mac and EFL.
Added layout tests support for smooth scroll in Gtk 3.4, and use smooth
scroll for emulating multi-tick mouseScrollBy events.
Patch by José Dapena Paz <jdapena@igalia.com> on 2012-08-29
Reviewed by Martin Robinson.
- DumpRenderTree/gtk/EventSender.cpp:
(mouseScrollByCallback):
(continuousMouseScrollByCallback):
- WebKitTestRunner/EventSenderProxy.h:
(EventSenderProxy):
- WebKitTestRunner/InjectedBundle/Bindings/EventSendingController.idl:
- WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
(WTR::EventSendingController::mouseScrollBy):
(WTR):
(WTR::EventSendingController::continuousMouseScrollBy):
- WebKitTestRunner/InjectedBundle/EventSendingController.h:
(EventSendingController):
- WebKitTestRunner/TestController.cpp:
(WTR::TestController::didReceiveSynchronousMessageFromInjectedBundle):
- WebKitTestRunner/gtk/EventSenderProxyGtk.cpp:
(WTR):
(WTR::EventSenderProxy::mouseScrollBy):
(WTR::EventSenderProxy::continuousMouseScrollBy):
- WebKitTestRunner/efl/EventSenderProxyEfl.cpp:
(WTR):
(WTR::EventSenderProxy::continuousMouseScrollBy):
- WebKitTestRunner/mac/EventSenderProxy.mm:
(WTR::EventSenderProxy::continuousScrollBy):
- WebKitTestRunner/qt/EventSenderProxyQt.cpp:
(WTR::EventSenderProxy::continuousMouseScrollBy):
(WTR):
LayoutTests: [Gtk] Process Gtk 3.4 smooth scroll events properly.
https://bugs.webkit.org/show_bug.cgi?id=88070
Added layout tests support for smooth scroll in Gtk 3.4, and use smooth
scroll for emulating multi-tick mouseScrollBy events.
Patch by José Dapena Paz <jdapena@igalia.com> on 2012-08-29
Reviewed by Martin Robinson.
- platform/gtk-wk2/fast/events/wheelevent-in-horizontal-scrollbar-in-rtl-expected.txt: Added.
- platform/gtk-wk2/fast/events/wheelevent-in-vertical-scrollbar-in-rtl-expected.txt: Added.
- platform/gtk/TestExpectations:
- platform/gtk/fast/events/wheelevent-in-horizontal-scrollbar-in-rtl-expected.txt: Added.
- platform/gtk/fast/events/wheelevent-in-vertical-scrollbar-in-rtl-expected.txt: Added.
- 5:28 PM Changeset in webkit [127069] by
-
- 2 edits in trunk/LayoutTests
Unreviewed gardening for Mac Lion after r127046.
- platform/mac-lion/TestExpectations:
- 5:27 PM Changeset in webkit [127068] by
-
- 17 edits in trunk/Source/JavaScriptCore
Refactoring LLInt::Data.
https://bugs.webkit.org/show_bug.cgi?id=95316.
Patch by Mark Lam <mark.lam@apple.com> on 2012-08-29
Reviewed by Geoff Garen.
This change allows its opcodeMap to be easily queried from any function
without needing to go through a GlobalData object. It also introduces
the LLInt::getCodePtr() methods that will be used by the LLInt C loop
later to redefine how llint symbols (opcodes and trampoline glue
labels) get resolved.
- assembler/MacroAssemblerCodeRef.h:
(MacroAssemblerCodePtr):
(JSC::MacroAssemblerCodePtr::createLLIntCodePtr):
(MacroAssemblerCodeRef):
(JSC::MacroAssemblerCodeRef::createLLIntCodeRef):
- bytecode/CodeBlock.cpp:
(JSC::CodeBlock::adjustPCIfAtCallSite):
(JSC::CodeBlock::bytecodeOffset):
- bytecode/Opcode.h:
Remove the 'const' to simplify things and avoid having to do
additional casts and #ifdefs in many places.
- bytecode/ResolveGlobalStatus.cpp:
(JSC::computeForLLInt):
- bytecompiler/BytecodeGenerator.cpp:
(JSC::BytecodeGenerator::generate):
- interpreter/Interpreter.cpp:
(JSC::Interpreter::initialize):
- interpreter/Interpreter.h:
(Interpreter):
- jit/JITExceptions.cpp:
(JSC::genericThrow):
- llint/LLIntData.cpp:
(LLInt):
(JSC::LLInt::initialize):
- llint/LLIntData.h:
(JSC):
(LLInt):
(Data):
(JSC::LLInt::exceptionInstructions):
(JSC::LLInt::opcodeMap):
(JSC::LLInt::getOpcode):
(JSC::LLInt::getCodePtr):
(JSC::LLInt::Data::performAssertions):
- llint/LLIntExceptions.cpp:
(JSC::LLInt::returnToThrowForThrownException):
(JSC::LLInt::returnToThrow):
(JSC::LLInt::callToThrow):
- llint/LLIntSlowPaths.cpp:
(JSC::LLInt::LLINT_SLOW_PATH_DECL):
(JSC::LLInt::handleHostCall):
- runtime/InitializeThreading.cpp:
(JSC::initializeThreadingOnce): Initialize the singleton LLInt data.
- runtime/JSGlobalData.cpp:
(JSC::JSGlobalData::JSGlobalData):
- runtime/JSGlobalData.h:
(JSGlobalData): Removed the now unneeded LLInt::Data instance in
JSGlobalData.
- runtime/JSValue.h:
(JSValue):
- 5:16 PM Changeset in webkit [127067] by
-
- 2 edits in trunk/LayoutTests
Unreviewed Gardening. platform/win/accessibility/single-select-children.html times out on Windows.
https://bugs.webkit.org/show_bug.cgi?id=95405
This test times out, probably for a similar reason as https://bugs.webkit.org/show_bug.cgi?id=93667.
DRT on Windows seems to have a tough time figuring out to continue the test when a select menu has gained focus / is opened.
Adding to skip list for now to get bots greener.
- platform/win/Skipped:
- 4:52 PM Changeset in webkit [127066] by
-
- 3 edits in trunk/Source/JavaScriptCore
PutById uses DataLabel32, not DataLabelCompact
https://bugs.webkit.org/show_bug.cgi?id=95245
Reviewed by Geoff Garen.
JIT::resetPatchPutById calls the the wrong thing on x86-64 – this is moot right now,
since they currently both do the same thing, but if we were to ever make compact mean
8-bit this could be a real problem. Also, relying on the object still being in eax
on entry to the transition stub isn't very robust - added nonArgGPR1 to at least make
this explicit.
- jit/JITPropertyAccess.cpp:
(JSC::JIT::emitSlow_op_put_by_id):
- copy regT0 to nonArgGPR1
(JSC::JIT::privateCompilePutByIdTransition):
- DataLabelCompact -> DataLabel32
(JSC::JIT::resetPatchPutById):
- reload regT0 from nonArgGPR1
- jit/JSInterfaceJIT.h:
(JSInterfaceJIT):
- added nonArgGPR1
- 4:22 PM Changeset in webkit [127065] by
-
- 16 edits2 deletes in trunk/LayoutTests
Unreviewed rebaseline after r127035
Updated base expectations:
- platform/chromium-linux/media/track/track-cue-rendering-vertical-expected.txt:
- platform/chromium-mac/media/track/track-cue-rendering-horizontal-expected.txt:
- platform/chromium-mac/media/track/track-cue-rendering-vertical-expected.txt:
- platform/chromium-win/media/track/track-cue-rendering-horizontal-expected.txt:
- platform/chromium-win/media/track/track-cue-rendering-vertical-expected.txt:
Removed unnecessary expectations:
- platform/chromium-mac-snowleopard/media/track/track-cue-rendering-horizontal-expected.txt: Removed.
- platform/chromium-mac-snowleopard/media/track/track-cue-rendering-vertical-expected.txt: Removed.
Updated expectations and added mime types:
- platform/chromium-linux/media/track/track-cue-rendering-horizontal-expected.png: Added property svn:mime-type.
- platform/chromium-linux/media/track/track-cue-rendering-vertical-expected.png: Added property svn:mime-type.
- platform/chromium-mac-snowleopard/media/track/track-cue-rendering-horizontal-expected.png: Added property svn:mime-type.
- platform/chromium-mac-snowleopard/media/track/track-cue-rendering-vertical-expected.png: Added property svn:mime-type.
- platform/chromium-mac/media/track/track-cue-rendering-horizontal-expected.png: Added property svn:mime-type.
- platform/chromium-win-xp/media/track/track-cue-rendering-vertical-expected.png: Added property svn:mime-type.
- platform/chromium-win/media/track/track-cue-rendering-horizontal-expected.png: Added property svn:mime-type.
- platform/chromium-win/media/track/track-cue-rendering-vertical-expected.png: Added property svn:mime-type.
Updated TestExpectations:
- platform/chromium/TestExpectations:
- 4:20 PM Changeset in webkit [127064] by
-
- 8 edits1 add in trunk/Source
[chromium] Remove HUD layer when rootLayer is set to null
https://bugs.webkit.org/show_bug.cgi?id=95257
Reviewed by James Robinson.
Source/WebCore:
When the rootLayer changes, remove the HUD layer from the old
root layer immediately. Then, during commit, set the hud layer
on the impl side only if there is a HUD layer present, and if
there is a rootLayer present for it to be a child of.
Tests: CCHudWithRootLayerChange
- platform/graphics/chromium/cc/CCLayerTreeHost.cpp:
(WebCore::CCLayerTreeHost::finishCommitOnImplThread):
(WebCore::CCLayerTreeHost::willCommit):
(WebCore::CCLayerTreeHost::setRootLayer):
- platform/graphics/chromium/cc/CCLayerTreeHost.h:
(WebCore::CCLayerTreeHost::hudLayer):
(CCLayerTreeHost):
Source/WebKit/chromium:
- WebKit.gypi:
- tests/CCHeadsUpDisplayTest.cpp: Added.
(CCHeadsUpDisplayTest):
(DrawsContentLayerChromium):
(DrawsContentLayerChromium::create):
(DrawsContentLayerChromium::DrawsContentLayerChromium):
(CCHudWithRootLayerChange):
(CCHudWithRootLayerChange::CCHudWithRootLayerChange):
(TEST_F):
- tests/CCLayerTreeHostTest.cpp:
- tests/CCThreadedTest.cpp:
(WebKitTests::CCThreadedTest::runTest):
- tests/CCThreadedTest.h:
(WebKitTests::CCThreadedTest::initializeSettings):
(CCThreadedTest):
- 4:12 PM Changeset in webkit [127063] by
-
- 6 edits in trunk/Source/WebCore
[Texmap] Move TextureMapperGL to use GraphicsContext3D
https://bugs.webkit.org/show_bug.cgi?id=78672
Patch by Roland Takacs <rtakacs@inf.u-szeged.hu>, Helder Correia <Helder Correia> on 2012-08-29
Reviewed by Noam Rosenthal.
It is based on a previous patch by Helder Correia.
TextureMapperGL (TMGL) includes direct GL calls and
GraphicsContext3D (GC3D) offers many conveniences over the
former approach: using existing CSS shader code, ANGLE for
shader compilation, reusing WebCore::Texture, having shaders and
textures that can delete themselves.
A GC3D object is created by TMGL with the newly introduced
builder createForCurrentGLContext(), which in turn uses
the new RenderToCurrentGLContext flag underneath.
TMGL's dependency on OpenGLShims.h was completely removed.
However, GC3D does not map every single GL constant. Thus, it's
important to document the following:
- GL_FALSE was mapped to false.
- GL_UNPACK_ROW_LENGTH, GL_UNPACK_SKIP_PIXELS, GL_UNPACK_SKIP_ROWS, GL_TEXTURE_RECTANGLE_ARB, and GL_UNSIGNED_INT_8_8_8_8_REV were locally defined in TMGL.
The patch was originally developed by Helder Correia and finished
by Roland Takacs.
No new tests, refactoring.
- platform/graphics/GraphicsContext3D.h:
- platform/graphics/texmap/TextureMapperGL.cpp:
(SharedGLData):
(WebCore::TextureMapperGLData::SharedGLData::currentSharedGLData):
(WebCore::TextureMapperGLData::SharedGLData::SharedGLData):
(WebCore::TextureMapperGLData::sharedGLData):
(WebCore::TextureMapperGLData::TextureMapperGLData):
(TextureMapperGLData):
(WebCore::scissorClip):
(WebCore::TextureMapperGL::ClipStack::apply):
(WebCore::TextureMapperGLData::initializeStencil):
(WebCore::TextureMapperGL::TextureMapperGL):
(WebCore::TextureMapperGL::beginPainting):
(WebCore::TextureMapperGL::endPainting):
(WebCore::TextureMapperGL::drawQuad):
(WebCore::TextureMapperGL::drawBorder):
(WebCore):
(WebCore::TextureMapperGL::drawTextureRectangleARB):
(WebCore::TextureMapperGL::drawTexture):
(WebCore::viewportMatrix):
(WebCore::TextureMapperGL::drawTextureWithAntialiasing):
(WebCore::TextureMapperGL::drawTexturedQuadWithProgram):
(WebCore::BitmapTextureGL::didReset):
(WebCore::BitmapTextureGL::updateContents):
(WebCore::TextureMapperGL::drawFiltered):
(WebCore::BitmapTextureGL::initializeStencil):
(WebCore::BitmapTextureGL::clearIfNeeded):
(WebCore::BitmapTextureGL::createFboIfNeeded):
(WebCore::BitmapTextureGL::bind):
(WebCore::BitmapTextureGL::~BitmapTextureGL):
(WebCore::TextureMapperGL::bindDefaultSurface):
(WebCore::TextureMapperGL::beginScissorClip):
(WebCore::TextureMapperGL::beginClip):
(WebCore::TextureMapperGL::endClip):
(WebCore::TextureMapperGL::createTexture):
- platform/graphics/texmap/TextureMapperGL.h:
(WebCore::TextureMapperGL::graphicsContext3D):
(TextureMapperGL):
(ClipStack):
(WebCore::BitmapTextureGL::textureTarget):
(BitmapTextureGL):
(WebCore::BitmapTextureGL::BitmapTextureGL):
- platform/graphics/texmap/TextureMapperShaderManager.cpp:
(WebCore):
(WebCore::TextureMapperShaderManager::getShaderProgram):
(WebCore::TextureMapperShaderProgram::TextureMapperShaderProgram):
(WebCore::TextureMapperShaderProgram::initializeProgram):
(WebCore::TextureMapperShaderProgram::getUniformLocation):
(WebCore::TextureMapperShaderProgram::~TextureMapperShaderProgram):
(WebCore::TextureMapperShaderProgramSimple::TextureMapperShaderProgramSimple):
(WebCore::TextureMapperShaderProgramSolidColor::TextureMapperShaderProgramSolidColor):
(WebCore::TextureMapperShaderProgramRectSimple::TextureMapperShaderProgramRectSimple):
(WebCore::TextureMapperShaderProgramOpacityAndMask::TextureMapperShaderProgramOpacityAndMask):
(WebCore::TextureMapperShaderProgramRectOpacityAndMask::TextureMapperShaderProgramRectOpacityAndMask):
(WebCore::TextureMapperShaderProgramAntialiasingNoMask::TextureMapperShaderProgramAntialiasingNoMask):
(WebCore::TextureMapperShaderManager::TextureMapperShaderManager):
(WebCore::StandardFilterProgram::~StandardFilterProgram):
(WebCore::StandardFilterProgram::StandardFilterProgram):
(WebCore::StandardFilterProgram::create):
(WebCore::StandardFilterProgram::prepare):
(WebCore::TextureMapperShaderManager::getShaderForFilter):
- platform/graphics/texmap/TextureMapperShaderManager.h:
(WebCore::TextureMapperShaderProgram::id):
(WebCore::TextureMapperShaderProgram::vertexAttrib):
(TextureMapperShaderProgram):
(WebCore::TextureMapperShaderProgram::matrixLocation):
(WebCore::TextureMapperShaderProgram::flipLocation):
(WebCore::TextureMapperShaderProgram::textureSizeLocation):
(WebCore::TextureMapperShaderProgram::sourceTextureLocation):
(WebCore::TextureMapperShaderProgram::maskTextureLocation):
(WebCore::TextureMapperShaderProgram::opacityLocation):
(WebCore::TextureMapperShaderProgram::isValidUniformLocation):
(StandardFilterProgram):
(WebCore::StandardFilterProgram::vertexAttrib):
(WebCore::StandardFilterProgram::texCoordAttrib):
(WebCore::StandardFilterProgram::textureUniform):
(WebCore::TextureMapperShaderProgramSimple::create):
(TextureMapperShaderProgramSimple):
(WebCore::TextureMapperShaderProgramRectSimple::create):
(TextureMapperShaderProgramRectSimple):
(WebCore::TextureMapperShaderProgramOpacityAndMask::create):
(TextureMapperShaderProgramOpacityAndMask):
(WebCore::TextureMapperShaderProgramRectOpacityAndMask::create):
(TextureMapperShaderProgramRectOpacityAndMask):
(WebCore::TextureMapperShaderProgramSolidColor::create):
(WebCore::TextureMapperShaderProgramSolidColor::colorLocation):
(TextureMapperShaderProgramSolidColor):
(WebCore::TextureMapperShaderProgramAntialiasingNoMask::create):
(WebCore::TextureMapperShaderProgramAntialiasingNoMask::expandedQuadVerticesInTextureCoordinatesLocation):
(WebCore::TextureMapperShaderProgramAntialiasingNoMask::expandedQuadEdgesInScreenSpaceLocation):
(TextureMapperShaderProgramAntialiasingNoMask):
(WebCore::TextureMapperShaderManager::TextureMapperShaderManager):
(TextureMapperShaderManager):
- 4:04 PM Changeset in webkit [127062] by
-
- 24 edits in trunk/Source/WebCore
Convert more static Strings to use ASCIILiteral
https://bugs.webkit.org/show_bug.cgi?id=95313
Reviewed by Eric Seidel.
This patch converts another swath of static strings to use ASCIILiteral
as recommended by http://trac.webkit.org/wiki/EfficientStrings.
- Modules/mediasource/MediaSource.cpp:
(WebCore::MediaSource::openKeyword):
(WebCore):
(WebCore::MediaSource::closedKeyword):
(WebCore::MediaSource::endedKeyword):
- Modules/mediasource/MediaSource.h:
(MediaSource):
- Modules/navigatorcontentutils/NavigatorContentUtils.cpp:
(WebCore::customHandlersStateString):
- Modules/notifications/Notification.cpp:
(WebCore::Notification::permissionString):
- accessibility/AccessibilityMediaControls.cpp:
(WebCore::AccessibilityMediaControl::controlTypeName):
(WebCore::AccessibilityMediaControl::title):
(WebCore::AccessibilityMediaControlsContainer::elementTypeName):
(WebCore::AccessibilityMediaTimeline::helpText):
- bindings/js/JSInspectorFrontendHostCustom.cpp:
(WebCore::JSInspectorFrontendHost::platform):
(WebCore::JSInspectorFrontendHost::port):
- bindings/js/ScriptEventListener.cpp:
(WebCore::eventParameterName):
- bindings/v8/ScriptEventListener.cpp:
(WebCore::eventParameterName):
- css/CSSWrapShapes.cpp:
(WebCore::CSSWrapShapeRectangle::cssText):
(WebCore::CSSWrapShapeCircle::cssText):
(WebCore::CSSWrapShapeEllipse::cssText):
(WebCore::CSSWrapShapePolygon::cssText):
- css/StyleSheetContents.cpp:
(WebCore::StyleSheetContents::parseAuthorStyleSheet):
- dom/Document.cpp:
(WebCore::Document::readyState):
- editing/MarkupAccumulator.cpp:
(WebCore::appendCharactersReplacingEntities):
- editing/markup.cpp:
(WebCore::StyledMarkupAccumulator::appendStyleNodeOpenTag):
(WebCore::StyledMarkupAccumulator::styleNodeCloseTag):
(WebCore::StyledMarkupAccumulator::appendElement):
(WebCore::createMarkup):
- html/track/TextTrackCue.cpp:
(WebCore::startKeyword):
(WebCore::middleKeyword):
(WebCore::endKeyword):
(WebCore::verticalGrowingLeftKeyword):
(WebCore::verticalGrowingRightKeyword):
(WebCore::TextTrackCue::updateDisplayTree):
(WebCore::TextTrackCue::settingName):
- page/DiagnosticLoggingKeys.cpp:
(WebCore::DiagnosticLoggingKeys::mediaLoadedKey):
(WebCore::DiagnosticLoggingKeys::mediaLoadingFailedKey):
(WebCore::DiagnosticLoggingKeys::pluginLoadedKey):
(WebCore::DiagnosticLoggingKeys::pluginLoadingFailedKey):
(WebCore::DiagnosticLoggingKeys::pageContainsPluginKey):
(WebCore::DiagnosticLoggingKeys::pageContainsAtLeastOnePluginKey):
(WebCore::DiagnosticLoggingKeys::passKey):
(WebCore::DiagnosticLoggingKeys::failKey):
(WebCore::DiagnosticLoggingKeys::noopKey):
- page/PageVisibilityState.cpp:
(WebCore::pageVisibilityStateString):
- page/UserContentURLPattern.cpp:
(WebCore::UserContentURLPattern::parse):
- platform/KURLWTFURL.cpp:
(WebCore::KURL::string):
- platform/MIMETypeRegistry.cpp:
(WebCore::defaultMIMEType):
(WebCore::mimeTypeAssociationMap):
- platform/network/ContentTypeParser.cpp:
(WebCore::ContentTypeParser::parse):
- platform/network/win/DownloadBundleWin.cpp:
(WebCore::DownloadBundle::fileExtension):
- storage/StorageTracker.cpp:
(WebCore::StorageTracker::syncFileSystemAndTrackerDatabase):
- svg/SVGAnimatedBoolean.cpp:
(WebCore::SVGAnimatedBooleanAnimator::constructFromString):
- 3:59 PM Changeset in webkit [127061] by
-
- 24 edits in trunk
[chromium] Support high DIP pixel tests with DumpRenderTree
https://bugs.webkit.org/show_bug.cgi?id=94935
Patch by Alex Sakhartchouk <alexst@chromium.org> on 2012-08-29
Reviewed by James Robinson.
Tools:
Adding device scale factor into the drt image dumping code
- DumpRenderTree/chromium/DRTTestRunner.cpp:
(DRTTestRunner::reset):
(DRTTestRunner::setBackingScaleFactor):
- DumpRenderTree/chromium/WebViewHost.cpp:
(WebViewHost::setDeviceScaleFactor):
(WebViewHost::paintInvalidatedRegion):
(WebViewHost::canvas):
- DumpRenderTree/chromium/WebViewHost.h:
(WebViewHost):
LayoutTests:
Introducing new expected test results in full resolution and temporarily
marking the files on other platforms as expected to fail until I can update them later
- platform/chromium-linux/fast/hidpi/broken-image-icon-hidpi-expected.png:
- platform/chromium-linux/fast/hidpi/broken-image-with-size-hidpi-expected.png:
- platform/chromium-linux/fast/hidpi/clip-text-in-hidpi-expected.png:
- platform/chromium-linux/fast/hidpi/device-scale-factor-paint-expected.png:
- platform/chromium-linux/fast/hidpi/focus-rings-expected.png:
- platform/chromium-linux/fast/hidpi/image-set-as-background-expected.png:
- platform/chromium-linux/fast/hidpi/image-set-background-dynamic-expected.png:
- platform/chromium-linux/fast/hidpi/image-set-background-repeat-expected.png:
- platform/chromium-linux/fast/hidpi/image-set-background-repeat-without-size-expected.png:
- platform/chromium-linux/fast/hidpi/image-set-border-image-comparison-expected.png:
- platform/chromium-linux/fast/hidpi/image-set-border-image-dynamic-expected.png:
- platform/chromium-linux/fast/hidpi/image-set-border-image-simple-expected.png:
- platform/chromium-linux/fast/hidpi/image-set-in-content-dynamic-expected.png:
- platform/chromium-linux/fast/hidpi/image-set-out-of-order-expected.png:
- platform/chromium-linux/fast/hidpi/image-set-simple-expected.png:
- platform/chromium-linux/fast/hidpi/image-set-without-specified-width-expected.png:
- platform/chromium-linux/fast/hidpi/resize-corner-hidpi-expected.png:
- platform/chromium-linux/fast/hidpi/video-controls-in-hidpi-expected.png:
- platform/chromium/TestExpectations:
- 3:48 PM Changeset in webkit [127060] by
-
- 2 edits in trunk/LayoutTests
Unreviewed Gardening: editing/delete/br-004(5)(6) failing on Windows (old run webkit tests)
https://bugs.webkit.org/show_bug.cgi?id=95391
Looks like NRWT expects different results for these tests than ORWT, as per http://trac.webkit.org/changeset/90490.
Adding these three failing tests to skip list to get bots greener.
- platform/win/Skipped:
- 3:47 PM Changeset in webkit [127059] by
-
- 2 edits in trunk/Tools
Fix check-webkit-style (and probably others) when WebKit is in a git submodule
https://bugs.webkit.org/show_bug.cgi?id=95177
Patch by Kevin Funk <kevin.funk@kdab.com> on 2012-08-29
Reviewed by Dirk Pranke.
Find the real checkout root by using 'git rev-parse --show-toplevel' instead of '--git-dir'
- Scripts/webkitpy/common/checkout/scm/git.py:
- 3:35 PM Changeset in webkit [127058] by
-
- 4 edits in trunk/Tools
webkit-patch rebaseline-expectations wrongly touches other platforms' TestExpectations
https://bugs.webkit.org/show_bug.cgi?id=95222
Reviewed by Adam Barth.
The code we had to remove REBASELINE lines from the
TestExpectations files did not handle multiple files (in a
cascade) correctly; we weren't limiting lines to a particular
file correctly in without_rebaseline_modifier(). This patch
fixes that and corrects the tests (which weren't correct and
didn't cover things very well).
Note also that the webkit-patch rebaseline-commands are still
using the include_overrides=False option in a couple places;
this is never correct or needed at this point. This patch fixes
the usage for webkit-patch rebaseline-expectations, but I still
need to fix rebaseline-test-internal, which I will do in a
separate patch (see bug 95268).
- Scripts/webkitpy/layout_tests/models/test_expectations.py:
(TestExpectations.remove_rebaselined_tests.without_rebaseline_modifier):
- Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
(RebaseliningTest.assertRemove):
(RebaseliningTest.test_remove):
- Scripts/webkitpy/tool/commands/rebaseline.py:
(RebaselineExpectations._update_expectations_files):
- 3:23 PM Changeset in webkit [127057] by
-
- 3 edits in trunk/Tools
webkit-patch rebaseline-expectations hangs
https://bugs.webkit.org/show_bug.cgi?id=95243
Reviewed by Tony Chang.
Don't try to run no commands in parallel; report an error
instead that we didn't find any tests to rebaseline.
Filed bug 95387 as well to make sure run_in_parallel() doesn't hang forever.
- Scripts/webkitpy/tool/commands/rebaseline.py:
(RebaselineExpectations.execute):
- Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
(_assert_command):
(test_rebaseline_expectations_noop):
- 3:08 PM Changeset in webkit [127056] by
-
- 3 edits in trunk/LayoutTests
Unreviewed, adjusting reftest mask to avoid spurious Mac failures.
- svg/custom/clamped-masking-clipping-expected.svg:
- svg/custom/clamped-masking-clipping.svg:
- 3:01 PM Changeset in webkit [127055] by
-
- 2 edits in trunk/Source/WebKit/chromium
[chromium] Add more unit test coverage for semantics of drawableContentRect and visibleContentRect
https://bugs.webkit.org/show_bug.cgi?id=94542
Reviewed by Adrienne Walker.
To make upcoming refactors safer, it was appropriate to (finally)
add unit tests that check the behavior of drawableContentRect and
visibleContentRect computations in calculateDrawTransforms.
- tests/CCLayerTreeHostCommonTest.cpp:
- 2:58 PM Changeset in webkit [127054] by
-
- 2 edits in trunk/LayoutTests
Unreviewed, mark css3/flexbox/flexitem.html as failing on efl. There
are differences with form controls.
- platform/efl/TestExpectations:
- 2:52 PM Changeset in webkit [127053] by
-
- 2 edits in trunk/Source/WebKit/blackberry
[BlackBerry] Replace blackberry specific HitTestRequest::FingerUp by TouchEvent + Release
https://bugs.webkit.org/show_bug.cgi?id=95383
Reviewed by Rob Buis.
Patch by Antonio Gomes <agomes@rim.com>
- WebKitSupport/TouchEventHandler.cpp:
(BlackBerry::WebKit::TouchEventHandler::touchEventCancel):
- 2:49 PM Changeset in webkit [127052] by
-
- 34 edits16 adds68 deletes in trunk/LayoutTests
Final unreviewed rebaselining after r126683.
This change is identical to r126944, r126832, r126818, r126793 and r126957 except that it touches around
68 files. If you are interested in the details, it's better to check one of those as they are easier to
follow.
This change is massive as it optimizes the baselines from the different ports. The apparent sizes change in the
common baselines are expected (I checked them against the other baselines). They don't represent real size changes
following r126683 (which shouldn't happen), just that some older change made the size change and we are propagating
it to the common baseline.
- platform/chromium/TestExpectations
Removed all the entries related to BUGWK84286.
[Snipped the long list of touched file]
- 2:27 PM Changeset in webkit [127051] by
-
- 7 edits in trunk/Source/WebCore
[New Multicolumn] Rename methods to prepare for proper pagination of columns
https://bugs.webkit.org/show_bug.cgi?id=95375
Reviewed by Simon Fraser.
This patch is doing some renaming and refactoring to prepare for proper pagination of columns. Most of the renames
involve changing RenderFlowThread functions to exactly match the names of their RenderBlock callers. These names
end up being more accurate once the top of a page and remaining height on a page no longer have a 1:1 mapping to
the RenderRegion's dimensions.
The renames/additions include:
renderRegionForLine -> regionAtBlockOffset
Justification: The block method is already called regionAtBlockOffset. No lines are involved, so line was
never the correct term to be passing in.
regionLogicalXXXForLine -> pageLogicalXXXForOffset
Justification: Matches the RenderBlock callers, and it's more accurate to talk in terms of "pages" now that
we have RenderRegionSets that can contain multiple "pages" in a single region.
logicalWidthForFlowThreadContent/logicalHeightForFlowThreadContent -> pageLogicalWidth/Height.
Justification: Makes it more clear we're talking about the width and height of a single page/column rather
than the width and height of the region itself.
logicalHeightOfAllFlowThreadContent
This method is new and represents the total flow thread logical height that is consumed by the region.
It has to be distinguished from the pageLogicalHeight for a region since sets can have multiple pages/columns.
Note with this patch we're essentially adopting the convention used by all of the pagination code besides regions
of referring to anything we paginate as a "page." I continue to believe this is the simplest way to talk about
these objects in code that is generically dealing with all three (like the breaking code in RenderBlock).
Eventually we might adopt the fragment terminology in the latest CSS draft, in which case
RenderRegion would become RenderFragment, RenderFlowThread would become RenderFragmentedBlock, and uses of
the generic "page" would become "fragment" instead, but we'll wait for that draft's terminology to stabilize first
before switching away from the current names.
- rendering/RenderBlock.cpp:
(WebCore::RenderBlock::hasNextPage):
(WebCore::RenderBlock::pageLogicalTopForOffset):
(WebCore::RenderBlock::pageLogicalHeightForOffset):
(WebCore::RenderBlock::pageRemainingLogicalHeightForOffset):
(WebCore::RenderBlock::regionAtBlockOffset):
- rendering/RenderFlowThread.cpp:
(WebCore::RenderFlowThread::layout):
(WebCore::RenderFlowThread::computeLogicalWidth):
(WebCore::RenderFlowThread::computeLogicalHeight):
(WebCore::RenderFlowThread::regionAtBlockOffset):
(WebCore::RenderFlowThread::pageLogicalTopForOffset):
(WebCore::RenderFlowThread::pageLogicalWidthForOffset):
(WebCore::RenderFlowThread::pageLogicalHeightForOffset):
(WebCore::RenderFlowThread::pageRemainingLogicalHeightForOffset):
(WebCore::RenderFlowThread::mapFromFlowToRegion):
(WebCore::RenderFlowThread::setRegionRangeForBox):
- rendering/RenderFlowThread.h:
- rendering/RenderMultiColumnSet.h:
- rendering/RenderRegion.cpp:
(WebCore::RenderRegion::pageLogicalWidth):
(WebCore::RenderRegion::pageLogicalHeight):
(WebCore):
(WebCore::RenderRegion::logicalHeightOfAllFlowThreadContent):
(WebCore::RenderRegion::layout):
- rendering/RenderRegion.h:
(RenderRegion):
- 2:23 PM Changeset in webkit [127050] by
-
- 7 edits2 adds in trunk/LayoutTests
r126257 broke css3/flexbox/flexitem.html
https://bugs.webkit.org/show_bug.cgi?id=94720
Reviewed by Ojan Vafai.
Move the failing test case into a new file so we can re-enable the
rest of the tests. We probably need to add some debugging information
to the failing test case to identify the cause of flakiness.
- css3/flexbox/flexitem-expected.txt: Remove flaky test case.
- css3/flexbox/flexitem-stretch-image-expected.txt: Added.
- css3/flexbox/flexitem-stretch-image.html: Added.
- css3/flexbox/flexitem.html: Remove flaky test case.
- platform/chromium/TestExpectations: Change failure to new test.
- platform/efl/TestExpectations: Change failure to new test.
- platform/mac/Skipped: Remove so we can get flakiness information.
- platform/mac/TestExpectations: Mark new test as flaky.
- 2:20 PM EfficientStrings edited by
- (diff)
- 2:17 PM Changeset in webkit [127049] by
-
- 4 edits3 adds in trunk
IndexedDB: Throw TypeError for invalid version parameters
https://bugs.webkit.org/show_bug.cgi?id=95143
Reviewed by Tony Chang.
Source/WebCore:
Still need to throw for -1, but developers are running into the string
scenario so handling that is more urgent. See
https://groups.google.com/a/chromium.org/forum/?fromgroups=#!topic/chromium-html5/QvjsPbBdP4M
Test: storage/indexeddb/intversion-bad-parameters.html
- Modules/indexeddb/IDBFactory.cpp:
(WebCore::IDBFactory::open):
LayoutTests:
- http/tests/inspector/indexeddb/indexeddb-test.js:
Remove 0 as a version parameter now that it will throw TypeError. It
should have been a no-op anyway.
- storage/indexeddb/intversion-bad-parameters-expected.txt: Added.
- storage/indexeddb/intversion-bad-parameters.html: Added.
- storage/indexeddb/resources/intversion-bad-parameters.js: Added.
(test):
(deleteSuccess):
- 2:05 PM Changeset in webkit [127048] by
-
- 3 edits in trunk/LayoutTests
ASSERTION FAILED: enclosingIntRect(rendererMappedResult) == enclosingIntRect(FloatQuad(result).boundingBox()) : WebCore::FloatRect WebCore::RenderGeometryMap::absoluteRect(const WebCore::FloatRect &) const
https://bugs.webkit.org/show_bug.cgi?id=92464
Saw an instance of fast/events/tabindex-focus-blur-all.html hitting the assertion on Debug
WK1 as well. Since it doesn't always happen on Debug WK1, mark it as flakey.
- platform/mac-wk2/Skipped:
Move handling of the test from here ...
- platform/mac/TestExpectations:
... to here.
- 1:58 PM Changeset in webkit [127047] by
-
- 9 edits3 adds in trunk
REGRESSION: Not sending NPP_SetWindow is causing Flash to not throttle itself
<rdar://problem/12133021> and https://bugs.webkit.org/show_bug.cgi?id=95274
Reviewed by Sam Weinig.
Source/WebKit2:
- UIProcess/API/mac/WKView.mm:
(-[WKView viewDidMoveToWindow]): Previously we'd only update window visibility when the window is hidden.
Now we also update window visibility when the window is shown.
- WebProcess/Plugins/Netscape/NetscapePlugin.cpp:
(WebKit::NetscapePlugin::callSetWindowInvisible): Call set window with a manufactured empty clip rect to tell
the plug-in that it is complete hidden.
(WebKit):
- WebProcess/Plugins/Netscape/NetscapePlugin.h:
(NetscapePlugin):
- WebProcess/Plugins/Netscape/mac/NetscapePluginMac.mm:
(WebKit::NetscapePlugin::windowVisibilityChanged): Call "callSetWindow" or "callSetWindowInvisible" as appropriate.
- WebProcess/Plugins/PluginView.cpp:
(WebKit::PluginView::setWindowIsVisible): Tell the plugin that its visibility changed.
(WebKit::PluginView::viewGeometryDidChange): Grab a clip rect that - although incorrect - at least is correct if
the PluginView is completely hidden.
Tools:
Add a test plug-in that calls back into the page with info on the NPWindow passed
in to NPP_SetWindow.
- DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
- DumpRenderTree/TestNetscapePlugIn/Tests/LogNPPSetWindow.cpp: Added.
(LogNPPSetWindow):
(LogNPPSetWindow::LogNPPSetWindow):
(LogNPPSetWindow::NPP_SetWindow):
LayoutTests:
- platform/mac-wk2/plugins/npp-setwindow-called-on-scroll-expected.txt: Added.
- platform/mac-wk2/plugins/npp-setwindow-called-on-scroll.html: Added.
- 1:55 PM Changeset in webkit [127046] by
-
- 17 edits1 copy3 adds in trunk
[CSS Shaders] Use CSS transform parsing code within CSS Shader
https://bugs.webkit.org/show_bug.cgi?id=71401
Reviewed by Dean Jackson.
Source/WebCore:
Added computed style for the transform parameters of the custom() filter function.
Implemented the FECustomFilter bindings needed to push the value to an actual matrix for the CSS Shader.
Note that transform parameters animations support will come in a different patch:
https://bugs.webkit.org/show_bug.cgi?id=94980.
Test: css3/filters/custom/effect-custom-transform-parameters.html
- GNUmakefile.list.am:
- Target.pri:
- WebCore.gypi:
- WebCore.vcproj/WebCore.vcproj:
- WebCore.xcodeproj/project.pbxproj:
- css/CSSComputedStyleDeclaration.cpp:
(WebCore::matrixTransformValue): Extracted code from computedTransform into a function, so that we
can reuse it.
(WebCore):
(WebCore::computedTransform):
(WebCore::valueForCustomFilterNumberParameter): Made all the custom filter related functions static to match
most of the other functions in this file.
(WebCore::valueForCustomFilterTransformParameter):
(WebCore::valueForCustomFilterParameter):
(WebCore::CSSComputedStyleDeclaration::valueForFilter): Needed the object size to compute the transform.
(WebCore::CSSComputedStyleDeclaration::valueForShadow): Added the const keyword.
(WebCore::CSSComputedStyleDeclaration::getPropertyCSSValue): Passed in the object to the valueForFilter.
- css/CSSComputedStyleDeclaration.h:
(WebCore):
(CSSComputedStyleDeclaration):
- css/StyleResolver.cpp:
(StyleResolver::parseCustomFilterTransformParameter):
(StyleResolver::parseCustomFilterParameter):
(StyleResolver::parseCustomFilterParameterList):
- css/StyleResolver.h:
(StyleResolver):
- platform/graphics/filters/CustomFilterParameter.h:
- platform/graphics/filters/CustomFilterTransformParameter.h: Filter parameter wrapper for the FilterOperations.
(WebCore):
(CustomFilterTransformParameter):
(WebCore::CustomFilterTransformParameter::create):
(WebCore::CustomFilterTransformParameter::blend): Animations will come in future patch.
(WebCore::CustomFilterTransformParameter::operator==):
(WebCore::CustomFilterTransformParameter::applyTransform):
(WebCore::CustomFilterTransformParameter::operations):
(WebCore::CustomFilterTransformParameter::setOperations):
(WebCore::CustomFilterTransformParameter::CustomFilterTransformParameter):
- platform/graphics/filters/FECustomFilter.cpp:
(WebCore::FECustomFilter::bindProgramTransformParameter):
(WebCore):
(WebCore::FECustomFilter::bindProgramParameters):
- platform/graphics/filters/FECustomFilter.h:
(WebCore):
(FECustomFilter):
LayoutTests:
Added test to check for the computed style of the transform parameters of the
custom() function. Also, added a test file to compare the result of applying the transform
using CSS Custom Filters and using CSS 3D Transforms.
- css3/filters/custom/custom-filter-property-computed-style-expected.txt:
- css3/filters/custom/effect-custom-transform-parameters-expected.html: Added.
- css3/filters/custom/effect-custom-transform-parameters.html: Added.
- css3/filters/resources/vertex-transform-parameter.vs: Added.
- css3/filters/script-tests/custom-filter-property-computed-style.js:
(description):
- platform/chromium/css3/filters/custom/custom-filter-property-computed-style-expected.txt:
- 1:46 PM Changeset in webkit [127045] by
-
- 2 edits in trunk/Source/WebCore
Use Animation::IterationCountInfinite instead of -1 when setting iteration count
https://bugs.webkit.org/show_bug.cgi?id=95339
Reviewed by Andreas Kling.
- css/CSSToStyleMap.cpp: Replace -1 for the proper enum.
- 1:41 PM Changeset in webkit [127044] by
-
- 4 edits in trunk/Source/WebCore
[TexMap][cairo] Add GC3D::RenderToCurrentGLContext support
https://bugs.webkit.org/show_bug.cgi?id=92441
Reviewed by Noam Rosenthal.
Add a RenderToCurrentGLContext for the Cairo GraphicsContext3D. This will allow
TextureMapperGL to be rewritten on top of GraphicsContext3D by exposing the GC3D
interface for the widget's GL context.
No new tests. This will be covered by the existing AC tests after the patch on
bug 78672 lands.
- platform/graphics/cairo/GraphicsContext3DCairo.cpp:
(WebCore::GraphicsContext3D::GraphicsContext3D): Only create the offscreen
rendering buffers if we are rendering offscreen. Pass the rendering style to
the private data factory.
(WebCore::GraphicsContext3D::~GraphicsContext3D): Only destroy the offscreen
rendering buffers if we are rendering offscreen.
- platform/graphics/cairo/GraphicsContext3DPrivate.cpp:
(WebCore::GraphicsContext3DPrivate::create): Pass the rendering style through.
(WebCore::GraphicsContext3DPrivate::GraphicsContext3DPrivate): If we are using
a "current GL context" rendering style, we don't need to create a GL context.
We'll always just use the one that's currently active.
(WebCore::GraphicsContext3DPrivate::paintToTextureMapper): Assert that we only
do this with the offscreen rendering style.
- platform/graphics/cairo/GraphicsContext3DPrivate.h: Update method definitions
and expose the rendering style member GraphicsContext3D.
- 1:40 PM Changeset in webkit [127043] by
-
- 2 edits in trunk/Source/WebCore
[Qt] Implement CompositeDifference
https://bugs.webkit.org/show_bug.cgi?id=77355
Patch by Martin Leutelt <martin.leutelt@basyskom.com> on 2012-08-29
Reviewed by Noam Rosenthal.
Add mapping for difference composite mode for
future use.
- platform/graphics/qt/GraphicsContextQt.cpp:
(WebCore::toQtCompositionMode):
- 1:32 PM Changeset in webkit [127042] by
-
- 5 edits2 adds in trunk/Tools
[Mac] Basic DRT support for web notifications
https://bugs.webkit.org/show_bug.cgi?id=79492
<rdar://problem/10357639>
Reviewed by Alexey Proskuryakov.
This patch implements the stubs for web notification support on DRT. Assume that when
Notification.requestPermission() is called, the user always chooses to allow it. This can be easily
overridden in layout tests with testRunner.denyWebNotificationPermission() when needed.
- DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj: Add MockWebNotificationProvider.
- DumpRenderTree/mac/DumpRenderTree.mm:
(createWebViewAndOffscreenWindow): Set MockWebNotificationProvider singleton as the provider for the
web view.
(resetWebViewToConsistentStateBeforeTesting): Make sure to reset the provider's state between each test.
- DumpRenderTree/mac/TestRunnerMac.mm:
(TestRunner::grantWebNotificationPermission): Set permission to granted for specified origin.
(TestRunner::denyWebNotificationPermission): Set permission to denied for specified origin.
(TestRunner::removeAllWebNotificationPermissions):
(TestRunner::simulateWebNotificationClick): Retrieve the notification ID from the JS notification object, and
simulate the click. This is similar to how WTR performs the click.
- DumpRenderTree/mac/UIDelegate.mm:
(-[UIDelegate webView:decidePolicyForNotificationRequestFromOrigin:listener:]): Assume when asked, the
user allows web notifications.
- DumpRenderTree/mac/MockWebNotificationProvider.h: Added. Maintains a list of registered web views, known permissions,
a map of notification IDs to WebNotification instances, and a map of the WebViews from which the notifications were dispatched.
- DumpRenderTree/mac/MockWebNotificationProvider.mm: Added.
(+[MockWebNotificationProvider shared]): The provider is a singleton.
(-[MockWebNotificationProvider init]): Instantiate the permissions map.
(-[MockWebNotificationProvider registerWebView:]):
(-[MockWebNotificationProvider unregisterWebView:]):
(-[MockWebNotificationProvider showNotification:fromWebView:]): Add the notification to the maps. Tell the web view that the
notification did show.
(-[MockWebNotificationProvider cancelNotification:]): Tell the web view the notification did close.
(-[MockWebNotificationProvider notificationDestroyed:]): Remove the notification from the maps. Here we don't make a callback.
(-[MockWebNotificationProvider clearNotifications:]): Remove the specified notifications from the maps. Here we don't make a callback.
(-[MockWebNotificationProvider webView:didShowNotification:]): Delegate callback. Dispatch the show event.
(-[MockWebNotificationProvider webView:didClickNotification:]): Delegate callback. Dispatch the click event.
(-[MockWebNotificationProvider webView:didCloseNotifications:]): Delegate callback. Dispatch the close event.
(-[MockWebNotificationProvider simulateWebNotificationClick:]): Tell the web view we clicked the notification.
(-[MockWebNotificationProvider policyForOrigin:]): Look in the permission dictionary.
(-[MockWebNotificationProvider setWebNotificationOrigin:permission:]): Set the permission for the origin in the dictionary.
(-[MockWebNotificationProvider removeAllWebNotificationPermissions]): Clear the permission map.
(-[MockWebNotificationProvider reset]): Remove all mappings and permissions.
- 1:28 PM Changeset in webkit [127041] by
-
- 2 edits in trunk/Source/WebKit2
Build patch for Qt
- Target.pri: Missing WKMutableArray.cpp.
- 1:26 PM Changeset in webkit [127040] by
-
- 20 edits in trunk
[BlackBerry] Adjust wtf include header style
https://bugs.webkit.org/show_bug.cgi?id=95368
Reviewed by Yong Li.
Switch to #include <wtf/...> like the other ports.
Source/WebCore:
- platform/graphics/blackberry/LayerFilterRenderer.h:
Source/WebKit/blackberry:
- Api/BlackBerryGlobal.cpp:
- Api/WebPage.cpp:
- WebCoreSupport/ChromeClientBlackBerry.cpp:
- WebCoreSupport/GeolocationControllerClientBlackBerry.h:
- WebCoreSupport/SelectPopupClient.h:
- WebKitSupport/AboutData.cpp:
- WebKitSupport/DOMSupport.cpp:
- WebKitSupport/GLES2Context.cpp:
- WebKitSupport/InPageSearchManager.h:
- WebKitSupport/InputHandler.cpp:
Tools:
- DumpRenderTree/blackberry/DumpRenderTree.cpp:
- DumpRenderTree/blackberry/PNGImageEncoder.cpp:
- DumpRenderTree/blackberry/PNGImageEncoder.h:
- DumpRenderTree/blackberry/PixelDumpSupportBlackBerry.cpp:
- DumpRenderTree/blackberry/PixelDumpSupportBlackBerry.h:
- DumpRenderTree/blackberry/WorkQueueItemBlackBerry.cpp:
- 1:23 PM Changeset in webkit [127039] by
-
- 11 edits2 deletes in trunk
Web Inspector: Timeline: avoid "IPC message too big" on save/load
https://bugs.webkit.org/show_bug.cgi?id=91991
Patch by Eugene Klyuchnikov <eustas.bug@gmail.com> on 2012-08-29
Reviewed by Alexander Pavlov.
Source/WebCore:
Motivation: Now timeline tries to save all data with one chunk.
Sometimes this causes "IPC message too big" error.
Solution: Reuse Profiles/Heap save/load code.
- inspector/front-end/FileUtils.js:
(WebInspector.OutputStream): Moved from HeapSnapshotView.js
(WebInspector.findBalancedCurlyBrackets): Moved from HeapSnapshotLoader.js
(WebInspector.ChunkedXHRReader): Added.
(WebInspector.ChunkedFileWriter): Moved from HeapSnapshotView.js
- inspector/front-end/HeapSnapshotLoader.js: Utility method moved to FileUtils.js
- inspector/front-end/HeapSnapshotView.js: IO specific classes moved to FileUtils.js
- inspector/front-end/HeapSnapshotWorker.js: Added dependency to FileUtils.js
- inspector/front-end/TimelineModel.js: Adopted chunked IO API.
(WebInspector.TimelineModelLoader): Class for chunked deserialization.
(WebInspector.TimelineModelWriteToFileDelegate): Chunked serializer.
LayoutTests:
Refactored to adopt workflow.
- inspector/timeline/timeline-load.html: Turned into test suite to avoid code
duplication. Added "malformed input" case.
- 1:13 PM Changeset in webkit [127038] by
-
- 2 edits in trunk/Tools
Update cygwin-downloader.zip to match modified downloader script.
https://bugs.webkit.org/show_bug.cgi?id=76936
Rubber Stamped by Jon Honeycutt.
cygwin-downloader.py was modified in http://trac.webkit.org/changeset/126948.
Updating zip to reflect changes.
- CygwinDownloader/cygwin-downloader.zip:
- 1:10 PM Changeset in webkit [127037] by
-
- 4 edits in trunk/Source/WebCore
[New Multicolumn] Implement hit testing for columns.
https://bugs.webkit.org/show_bug.cgi?id=95367
Reviewed by Simon Fraser.
Add an implementation of nodeAtPoint for RenderMultiColumnSet that works similarly to painting.
Multiple calls get made as you walk through the columns to hitTestFlowThreadPortionInRegion.
- rendering/RenderMultiColumnSet.cpp:
(WebCore::RenderMultiColumnSet::nodeAtPoint):
(WebCore):
- rendering/RenderMultiColumnSet.h:
(RenderMultiColumnSet):
- rendering/RenderRegion.cpp:
(WebCore::RenderRegion::nodeAtPoint):
- 1:05 PM Changeset in webkit [127036] by
-
- 3 edits in trunk/Tools
Stylechecker warns about comparison to zero when comparing to 0.5
https://bugs.webkit.org/show_bug.cgi?id=94913
Reviewed by Dirk Pranke.
According to the python documentation, \W is the character group with all
non-alphanumeric characters, equivalent to [a-zA-Z0-9_], which is equal to
to [\w]. We need to exclude the dot "." as well, so that floating point
values do not trigger this warning. Adding a unit test that shows the problem.
- Scripts/webkitpy/style/checkers/cpp.py:
(check_for_comparisons_to_zero): Modifying the regex to not get triggered by comparing to floats.
- Scripts/webkitpy/style/checkers/cpp_unittest.py:
(WebKitStyleTest.test_null_false_zero): Unit test exposing this issue.
- 1:03 PM Changeset in webkit [127035] by
-
- 9 edits2 adds in trunk
Create a toggle button for closed captions.
https://bugs.webkit.org/show_bug.cgi?id=94395
Reviewed by Eric Carlson.
This patch will create a button that toggles any captions or subtitles on or off.
Source/WebCore:
Test: media/video-controls-captions.html
- css/mediaControlsChromium.css:
(audio::-webkit-media-controls-toggle-closed-captions-button, video::-webkit-media-controls-toggle-closed-captions-button):
New style for new button.
- html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::HTMLMediaElement):
Initialize new variable for keeping track of user (button) disabled captions.
(WebCore::HTMLMediaElement::loadTimerFired):
Rename configureNewTextTracks() to configureTextTracks().
(WebCore::HTMLMediaElement::prepareForLoad):
Do not set closedCaptionsVisible to false, we should instead honor the
checks for default out-of-band tracks.
(WebCore::HTMLMediaElement::textTrackModeChanged):
Rename configureNewTextTracks() to configureTextTracks().
(WebCore::HTMLMediaElement::userIsInterestedInThisTrackKind):
Add checks for when user has requested to see or not see captions.
(WebCore::HTMLMediaElement::configureTextTracks):
Rename configureNewTextTracks() to configureTextTracks().
(WebCore::HTMLMediaElement::hasClosedCaptions):
Return true if we have any caption or subtitle text tracks.
(WebCore::HTMLMediaElement::setClosedCaptionsVisible):
Update the text track display and the closed captions button when the
closed captions button is toggled.
(WebCore::HTMLMediaElement::configureTextTrackDisplay):
If the visibility of any text tracks has changed, update the display and
the controls.
(WebCore::HTMLMediaElement::updateClosedCaptionsControls):
New function that updates both the text track display and the closed
captions button.
- html/HTMLMediaElement.h:
(HTMLMediaElement):
New function updateClosedCaptionsControls()
Adding the button:
- html/shadow/MediaControlRootElementChromium.cpp:
(WebCore::MediaControlRootElementChromium::MediaControlRootElementChromium):
(WebCore::MediaControlRootElementChromium::initializeControls):
(WebCore::MediaControlRootElementChromium::setMediaController):
(WebCore::MediaControlRootElementChromium::reset):
(WebCore::MediaControlRootElementChromium::reportedError):
(WebCore::MediaControlRootElementChromium::changedClosedCaptionsVisibility):
(WebCore::MediaControlRootElementChromium::createTextTrackDisplay):
- html/shadow/MediaControlRootElementChromium.h:
(WebCore):
(MediaControlRootElementChromium):
Ensure that Linux and Windows render themes will support closed captions:
- rendering/RenderThemeChromiumSkia.cpp:
(WebCore):
(WebCore::supportsClosedCaptioning):
- rendering/RenderThemeChromiumSkia.h:
(RenderThemeChromiumSkia):
LayoutTests:
- media/video-controls-captions-expected.txt: Added.
- media/video-controls-captions.html: Added.
- 12:53 PM Changeset in webkit [127034] by
-
- 3 edits in trunk/Source/JavaScriptCore
ExecutableAllocator should be destructed after Heap
https://bugs.webkit.org/show_bug.cgi?id=95244
Reviewed by Rob Buis.
RIM PR# 199364.
Make ExecutableAllocator the first member in JSGlobalData.
Existing Web Worker tests can show the issue.
- runtime/JSGlobalData.cpp:
(JSC::JSGlobalData::JSGlobalData):
- runtime/JSGlobalData.h:
(JSGlobalData):
- 12:53 PM Changeset in webkit [127033] by
-
- 2 edits in trunk/LayoutTests
Unreviewed, chromium test expectations cleanup.
- platform/chromium/TestExpectations:
- 12:50 PM Changeset in webkit [127032] by
-
- 8 edits in trunk/Source/WebCore
run-bindings-tests failing on Apple Mountain Lion Testers.
https://bugs.webkit.org/show_bug.cgi?id=95354
Reviewed by Eric Seidel.
The binding tests were expecting incorrect results on Mac. The versions using the @property
syntax are correct for Leopard and above. CodeGeneratorObjC.pm was determining which syntax
to use based on the value of MACOSX_DEPLOYMENT_TARGET, which might not be set in the
environment the tests get run in but is correctly set by xcodebuild.
- bindings/scripts/CodeGeneratorObjC.pm:
(GenerateHeader):
Remove the code to support Tiger and earlier and the reliance on MACOSX_DEPLOYMENT_TARGET.
- bindings/scripts/test/ObjC/DOMTestActiveDOMObject.h:
Update the expectations to expect the @property syntax (done with --reset-results).
- bindings/scripts/test/ObjC/DOMTestEventConstructor.h:
Ditto.
- bindings/scripts/test/ObjC/DOMTestException.h:
Ditto.
- bindings/scripts/test/ObjC/DOMTestInterface.h:
Ditto.
- bindings/scripts/test/ObjC/DOMTestObj.h:
Ditto.
- bindings/scripts/test/ObjC/DOMTestSerializedScriptValueInterface.h:
Ditto.
- 12:49 PM Changeset in webkit [127031] by
-
- 3 edits in trunk/Source/WebCore
Web Inspector: Refactor InspectorFrontendClientLocal to remove ScriptState as member
https://bugs.webkit.org/show_bug.cgi?id=95343
Patch by Vivek Galatage <vivekgalatage@gmail.com> on 2012-08-29
Reviewed by Yury Semikhatsky.
The ScriptState is not required to be stored as a member to
InspectorFrontendClientLocal hence removing it as member.
No new tests due to code refactoring.
- inspector/InspectorFrontendClientLocal.cpp:
(WebCore::InspectorFrontendClientLocal::InspectorFrontendClientLocal):
(WebCore::InspectorFrontendClientLocal::~InspectorFrontendClientLocal):
(WebCore::InspectorFrontendClientLocal::windowObjectCleared):
- inspector/InspectorFrontendClientLocal.h:
(WebCore):
(InspectorFrontendClientLocal):
- 12:43 PM Changeset in webkit [127030] by
-
- 2 edits in trunk/Source/WebKit2
Build patch for Qt.
- Target.pri: Export WKArray and WKMutableArray for Qt.
- 12:35 PM Changeset in webkit [127029] by
-
- 2 edits in trunk/Source/WebKit2
Added missing exports for Windows.
- win/WebKit2Generated.make:
- 12:23 PM Changeset in webkit [127028] by
-
- 6 edits in trunk/Tools
WTR build fixes.
- WebKitTestRunner/CMakeLists.txt:
- WebKitTestRunner/GNUmakefile.am:
- WebKitTestRunner/Target.pri:
- WebKitTestRunner/WebNotificationProvider.cpp:
(WTR::WebNotificationProvider::showWebNotification): Make ML happier.
- WebKitTestRunner/win/WebKitTestRunner.vcproj:
- 12:16 PM Changeset in webkit [127027] by
-
- 2 edits in trunk/Source/WebKit2
Windows build fix.
- win/WebKit2Generated.make: Export WKNotificationManager.h.
- 11:54 AM Changeset in webkit [127026] by
-
- 10 edits2 adds in trunk/Tools
[WK2] Basic WTR support for web notifications
https://bugs.webkit.org/show_bug.cgi?id=95154
<rdar://problem/12184492>
Reviewed by Alexey Proskuryakov.
Implement WTR support, without platform event output (bug 95233).
Legacy APIs will not be supported, since many of the tests are skipped
on ports that do test notifications, and the issue with using file:// as the origin means a lot of the permissions
tests won't work. Bugs 81048 and 81697 will track migrating the existing tests to use the new testRunner API, and move
the tests to http/tests/notifications.
Permissions will be handled only in the injected bundle. This allows tests to set permission synchronously, and not
require a chain of setTimeout()'s when writing a notification test. Clicking a notification, however, should
be invoked from the UIProcess.
- WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
- WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
(WTR::InjectedBundle::didReceiveMessage): When told to "Reset", clear out all of the permissions.
(WTR::InjectedBundle::postSimulateWebNotificationClick): Tell the bundle client to simulate a click based on the
notification's internal ID.
- WebKitTestRunner/InjectedBundle/InjectedBundle.h:
- WebKitTestRunner/InjectedBundle/TestRunner.cpp:
(WTR::TestRunner::grantWebNotificationPermission): Manually set the permission using WKBundle SPI.
(WTR::TestRunner::denyWebNotificationPermission): Manually set the permission using WKBundle SPI.
(WTR::TestRunner::removeAllWebNotificationPermissions):
(WTR::TestRunner::simulateWebNotificationClick): Post a message through the injected bundle.
- WebKitTestRunner/InjectedBundle/TestRunner.h:
Because supporting web notifications goes beyond a couple functions, move all of the logic into a WebNotificationProvider
class.
- WebKitTestRunner/TestInvocation.cpp:
(WTR::TestInvocation::didReceiveMessageFromInjectedBundle): When the injected bundle sends a message to simulate a click,
TestInvocation forwards that request to the TestController.
- WebKitTestRunner/TestController.cpp:
(WTR::TestController::initialize): Set the provider.
(WTR::TestController::resetStateToConsistentValues): Tell the provider to reset.
(WTR::TestController::simulateWebNotificationClick): Tell the provider to simulate a user clicking on a platform notification.
- WebKitTestRunner/TestController.h: Holds an instance to the provider.
The provider maintains a pointer to one notification manager (presumably the web process associated with WTR), and a set
of notification IDs that have been "shown" by the platform. This allows us to simulate a click on a notification.
- WebKitTestRunner/WebNotificationProvider.h: Added.
- WebKitTestRunner/WebNotificationProvider.cpp: Added.
(WTR::WebNotificationProvider::showWebNotification): Adds the notification to set of shown notifications. Notifies provider that
the notification got shown.
(WTR::WebNotificationProvider::closeWebNotification): Removes the notification from the set. Notifies provider that the notification
got closed.
(WTR::WebNotificationProvider::addNotificationManager): Maintains one manager, which is fine for testing purposes.
(WTR::WebNotificationProvider::removeNotificationManager): Maintains one manager.
(WTR::WebNotificationProvider::notificationPermissions): When the web process starts, it maintains an initial set of known
permissions. For testing purposes, we always want this set to be empty.
(WTR::WebNotificationProvider::simulateWebNotificationClick): Notifies provider that the notification got "clicked".
(WTR::WebNotificationProvider::reset): To reset the state, we pretend that we closed all of the platform notifications.
- WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj: Add WebNotificationProvider.
- 11:50 AM Changeset in webkit [127025] by
-
- 2 edits in trunk
Tell git-archive to not export .gitattributes and .gitignore
Patch by Thiago Macieira <thiago.macieira@intel.com> on 2012-08-29
Reviewed by Simon Hausmann.
Exclude git specific files from archives created via git-archive.
- .gitattributes:
- 11:16 AM Changeset in webkit [127024] by
-
- 8 edits in trunk
[EFL] Add setting API for allow universal/file access from file URLs.
https://bugs.webkit.org/show_bug.cgi?id=83121
Patch by Kamil Blank <k.blank@samsung.com> on 2012-08-29
Reviewed by Eric Seidel.
Source/WebKit/efl:
Make it possible to enable allow universal/file access from file URLs.
Default value for both settings is true.
- ewk/ewk_view.cpp:
(_Ewk_View_Private_Data):
(_ewk_view_priv_new):
(ewk_view_setting_allow_universal_access_from_file_urls_set): Function sets if locally loaded documents
are allowed to access remote urls.
(ewk_view_setting_allow_universal_access_from_file_urls_get):
(ewk_view_setting_allow_file_access_from_file_urls_set): Function sets if locally loaded documents
are allowed to access other local urls.
(ewk_view_setting_allow_file_access_from_file_urls_get):
- ewk/ewk_view.h:
Tools:
Implementation of setAllowUniversalAccessFromFileURLs and setAllowFileAccessFromFileURLs.
- DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
(DumpRenderTreeChrome::resetDefaultsToConsistentValues):
- DumpRenderTree/efl/TestRunnerEfl.cpp:
(TestRunner::setAllowUniversalAccessFromFileURLs):
(TestRunner::setAllowFileAccessFromFileURLs):
LayoutTests:
Enable test connected with setAllowUniversalAccessFromFileURLs and setAllowFileAccessFromFileURLs.
- platform/efl/Skipped:
- 11:14 AM Changeset in webkit [127023] by
-
- 3 edits4 adds in trunk
[chromium] Link highlight should clear on page navigation.
https://bugs.webkit.org/show_bug.cgi?id=95129
Reviewed by James Robinson.
Modified WebViewImpl to clear link highlight when navigating to new url.
Source/WebKit/chromium:
- src/WebViewImpl.cpp:
(WebKit::WebViewImpl::didCommitLoad):
LayoutTests:
- platform/chromium-linux/compositing/gestures/gesture-tapHighlight-simple-navigate.html: Added.
- platform/chromium-linux/compositing/gestures/resources/gesture-tapHighlight-simple-navigate-destination.html: Added.
- platform/chromium-linux/platform/chromium-linux/compositing/gestures/gesture-tapHighlight-simple-navigate-expected.png: Added.
- platform/chromium-linux/platform/chromium-linux/compositing/gestures/gesture-tapHighlight-simple-navigate-expected.txt: Added.
- 11:09 AM Changeset in webkit [127022] by
-
- 2 edits in trunk/LayoutTests
ASSERTION FAILED: enclosingIntRect(rendererMappedResult) == enclosingIntRect(FloatQuad(result).boundingBox()) : WebCore::FloatRect WebCore::RenderGeometryMap::absoluteRect(const WebCore::FloatRect &) const
https://bugs.webkit.org/show_bug.cgi?id=92464
Skip fast/events/tabindex-focus-blur-all.html since it is manifesting as a crash on the WK2
Debug Testers.
- platform/mac-wk2/Skipped:
- 11:08 AM Changeset in webkit [127021] by
-
- 2 edits in branches/safari-536.26-branch/Source/WebCore
Merge r126921 for <rdar://problem/12117040>.
- 10:59 AM Changeset in webkit [127020] by
-
- 2 edits in trunk/Source/WebKit/chromium
[chromium] CCThreadImpl / WebCompositorImpl shouldn't compile from webkit when use_libcc_for_compositor=1
https://bugs.webkit.org/show_bug.cgi?id=94995
Reviewed by Adam Barth.
- WebKit.gyp:
- 10:47 AM Changeset in webkit [127019] by
-
- 8 edits in trunk/Source/WebKit2
[WK2] Add SPI to retrieve internal IDs for web notifications
https://bugs.webkit.org/show_bug.cgi?id=95100
<rdar://problem/12180208>
Reviewed by Alexey Proskuryakov.
This is work toward providing Mac support for web notifications in DRT and WTR (77969).
Add function to retrieve the internal ID for a notification. This is needed by tests to support simulating
a user click on a notification.
- DerivedSources.pri: Expose JSNotification.h as private header.
- WebProcess/InjectedBundle/API/c/WKBundle.cpp:
(WKBundleGetWebNotificationID): Calls into notification manager to get internal ID.
- WebProcess/InjectedBundle/API/c/WKBundlePrivate.h: Add WKBundleGetWebNotificationID() to be able
retrieve notification ID.
- WebProcess/InjectedBundle/InjectedBundle.cpp:
(WebKit::InjectedBundle::webNotificationID):
(WebKit):
- WebProcess/InjectedBundle/InjectedBundle.h:
(InjectedBundle):
- WebProcess/Notifications/WebNotificationManager.cpp:
(WebKit::WebNotificationManager::notificationIDForTesting):
(WebKit):
- WebProcess/Notifications/WebNotificationManager.h:
(WebNotificationManager):
- 10:42 AM Changeset in webkit [127018] by
-
- 2 edits in trunk/Tools
Update TestRunner API for web notifications
https://bugs.webkit.org/show_bug.cgi?id=95093
<rdar://problem/12179649>
Reviewed by Alexey Proskuryakov.
As it turns out the patch for this accidentally got squashed into the commit for
bug 95099 (r126909). This amendment patch adds some comments about the renaming
of the older TestRunner API for web notifications.
- DumpRenderTree/TestRunner.h:
(TestRunner):
- 10:40 AM Changeset in webkit [127017] by
-
- 5 edits in trunk/Source
[chromium] Do not clip root layer's subtree to viewport
https://bugs.webkit.org/show_bug.cgi?id=95235
Reviewed by Adrienne Walker.
Source/WebCore:
The root layer's renderSurface already correctly clips everything
to the viewport's bounds. There are some useful reasons that we
should not cause the root layer itself to clip the subtree any
further, in particular so that surfaces can remain more cacheable,
and to make root layer semantics more homogeneous.
Existing tests updated, and otherwise this change is already
covered by layout and unit tests.
- platform/graphics/chromium/cc/CCLayerTreeHostCommon.cpp:
(WebCore::calculateDrawTransformsInternal):
Source/WebKit/chromium:
Existing tests updated, and otherwise this change is already
covered by layout and unit tests.
- tests/CCLayerTreeHostCommonTest.cpp:
- tests/CCOcclusionTrackerTest.cpp:
(WebKitTests::CCOcclusionTrackerTestChildInRotatedChild::runMyTest):
(WebKitTests::CCOcclusionTrackerTestVisitTargetTwoTimes::runMyTest):
(WebKitTests::CCOcclusionTrackerTestSurfaceWithTwoOpaqueChildren::runMyTest):
(WebKitTests::CCOcclusionTrackerTestOverlappingSurfaceSiblings::runMyTest):
(WebKitTests::CCOcclusionTrackerTestOverlappingSurfaceSiblingsWithTwoTransforms::runMyTest):
(WebKitTests::CCOcclusionTrackerTestFilters::runMyTest):
(WebKitTests::CCOcclusionTrackerTestReplicaWithClipping::runMyTest):
(WebKitTests::CCOcclusionTrackerTestLargePixelsOccludeInsideClipRect::runMyTest):
(WebKitTests::CCOcclusionTrackerTestSurfaceOcclusionTranslatesWithClipping::runMyTest):
(WebKitTests::CCOcclusionTrackerTestSurfaceChildOfClippingSurface::runMyTest):
- 10:38 AM Changeset in webkit [127016] by
-
- 2 edits in trunk/Source/WebKit/blackberry
[BlackBerry] Remove unused tapGesture in WebPage.cpp
https://bugs.webkit.org/show_bug.cgi?id=95357
Reviewed by Rob Buis.
Patch by Jacky Jiang <zhajiang@rim.com>
Remove unused variable tapGesture in WebPage.cpp.
Internally reviewed by Gen Mak.
- Api/WebPage.cpp:
(BlackBerry::WebKit::WebPage::touchEvent):
- 10:32 AM Changeset in webkit [127015] by
-
- 2 edits in trunk/Source/JavaScriptCore
2012-08-29 Geoffrey Garen <ggaren@apple.com>
Try to fix the Windows build.
- JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def: Export!
- 10:24 AM Changeset in webkit [127014] by
-
- 2 edits in trunk/LayoutTests
[EFL] Unskip 2 willSendRequest tests that are passing consistently
https://bugs.webkit.org/show_bug.cgi?id=95289
Unreviewed EFL gardening.
Unskip the following tests are EFL port:
fast/loader/onload-willSendRequest-null-for-script.html
fast/loader/willSendRequest-null-for-preload.html
Patch by Christophe Dumez <Christophe Dumez> on 2012-08-29
- platform/efl/Skipped:
- 10:22 AM Changeset in webkit [127013] by
-
- 2 edits in trunk/Source/WebKit/chromium
[Chromium] CCLayerTreeHostTestScrollChildLayer makes the wrong assumptions
https://bugs.webkit.org/show_bug.cgi?id=95358
Unreviewed gardening.
- tests/CCLayerTreeHostTest.cpp:
Disabled the test until it is fixed.
- 10:18 AM Changeset in webkit [127012] by
-
- 5 edits in trunk/LayoutTests
'\r's in test expectations files confuse chromium-android layout test driver
https://bugs.webkit.org/show_bug.cgi?id=95353
Reviewed by Adam Barth.
Modified the test cases not to output '\r's.
Removed the '\r's from expectation files.
- fast/dom/Window/window-property-clearing-expected.txt:
- fast/dom/Window/window-property-clearing.html:
- fast/dom/simultaneouslyRegsiteredTimerFireOrder-expected.txt:
- fast/dom/simultaneouslyRegsiteredTimerFireOrder.html:
- 10:15 AM Changeset in webkit [127011] by
-
- 1 edit in trunk/Source/WebKit/mac/ChangeLog
Fix ChangeLog
- 10:13 AM Changeset in webkit [127010] by
-
- 16 edits2 adds in trunk/Source/JavaScriptCore
Introduced JSWithScope, making all scope objects subclasses of JSScope
https://bugs.webkit.org/show_bug.cgi?id=95295
Reviewed by Filip Pizlo.
This is a step toward removing ScopeChainNode. With a uniform representation
for objects in the scope chain, we can move data from ScopeChainNode
into JSScope.
- CMakeLists.txt:
- GNUmakefile.list.am:
- JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.vcproj:
- JavaScriptCore.xcodeproj/project.pbxproj:
- Target.pri: Build!
- interpreter/Interpreter.cpp:
(JSC::Interpreter::privateExecute):
- jit/JITStubs.cpp:
(JSC::DEFINE_STUB_FUNCTION):
- llint/LLIntSlowPaths.cpp:
(JSC::LLInt::LLINT_SLOW_PATH_DECL): Use an explicit JSWithScope object
for 'with' statements. Since 'with' can put any object in the scope
chain, we'll need an adapter object to hold the data ScopeChainNode
currently holds.
- runtime/JSGlobalData.cpp:
(JSC::JSGlobalData::JSGlobalData):
- runtime/JSGlobalData.h:
(JSGlobalData): Support for JSWithScope.
- runtime/JSScope.cpp:
(JSC::JSScope::objectAtScope):
- runtime/JSScope.h: Check for and unwrap JSWithScope.
- runtime/JSType.h: Support for JSWithScope.
- runtime/StrictEvalActivation.cpp:
(JSC::StrictEvalActivation::StrictEvalActivation):
- runtime/StrictEvalActivation.h:
(StrictEvalActivation): Inherit from JSScope, to make the scope chain uniform.
- runtime/JSWithScope.cpp: Added.
(JSC::JSWithScope::visitChildren):
- runtime/JSWithScope.h: Added.
(JSWithScope):
(JSC::JSWithScope::create):
(JSC::JSWithScope::object):
(JSC::JSWithScope::createStructure):
(JSC::JSWithScope::JSWithScope): New adapter object. Since this object
is never exposed to scripts, it doesn't need any meaningful implementation
of property access or other callbacks.
- 10:12 AM Changeset in webkit [127009] by
-
- 3 edits in trunk/Source/WebKit/mac
[Mac] Add iconURL to WebNotification
https://bugs.webkit.org/show_bug.cgi?id=95249
<rdar://problem/12192060>
Reviewed by Jessie Berlin.
- WebView/WebNotification.h: Expose iconURL.
- WebView/WebNotification.mm: Remove unnecessary ASSERTs.
(-[WebNotification title]):
(-[WebNotification body]):
(-[WebNotification tag]):
(-[WebNotification iconURL]): Added.
(-[WebNotification origin]):
(-[WebNotification notificationID]):
(-[WebNotification dispatchShowEvent]):
(-[WebNotification dispatchCloseEvent]):
(-[WebNotification dispatchClickEvent]):
(-[WebNotification dispatchErrorEvent]):
Reviewed by Jessie Berlin.
- WebView/WebNotification.h: Expose iconURL.
- WebView/WebNotification.mm:
(-[WebNotification iconURL]):
- 10:09 AM Changeset in webkit [127008] by
-
- 6 edits in trunk/Source/WebCore
[New Multicolumn] Implement column contents painting.
https://bugs.webkit.org/show_bug.cgi?id=95251
Reviewed by Simon Fraser.
This patch implements paintColumnContents for the new multicolumn blocks. There are a number of
improvements made over the current multi-column implementation. They include the fact that left
and right columns will now be unclipped, so contents of columns can actually spill out of the block now.
Outlines now also work properly. The contents of the first column and last column can also now overflow
visually out of the top and bottom of those columns respectively.
- rendering/RenderMultiColumnSet.cpp:
(WebCore::RenderMultiColumnSet::flowThreadPortionRectAt):
This method returns the exact portion of the flow thread that matches the column dimensions.
The width and height are the column width and height. It is equivalent to RenderRegion::flowThreadPortionRect,
but is applied only for a specific column in the set.
(WebCore::RenderMultiColumnSet::flowThreadPortionOverflowRect):
This method expands the flowThreadPortionRect to unclip the edges of left and right columns, the top and
bottom edges of first and last columns, and it also expands the painting to go into half of the column
gap. It is analogous to RenderRegion::flowThreadPortionOverflowRect, but it's doing the right thing
for each individual column.
(WebCore::RenderMultiColumnSet::paintColumnContents):
The implementation of column contents painting. Each column is iterated over, and the two rects above
are computed and passed in to RenderFlowThread::paintFlowThreadPortionInRegion. This code does the right
thing with offsets and clipping when handed these two rects and handles all of the rest.
- rendering/RenderMultiColumnSet.h:
Declare the new functions that return the flow thread portion rects.
- rendering/RenderRegion.cpp:
(WebCore::RenderRegion::flowThreadPortionOverflowRect):
(WebCore):
(WebCore::RenderRegion::overflowRectForFlowThreadPortion):
- rendering/RenderRegion.h:
(RenderRegion):
- rendering/RenderRegionSet.cpp:
Refactor this so that it can be invoked by column sets and operate on the first and last columns instead
of only being tied to the first and last regions.
(WebCore::RenderRegionSet::expandToEncompassFlowThreadContentsIfNeeded):
Fix a bug here where the flowThread.y() should not have been subtracted. We're in local flow thread
coordinates.
- 9:46 AM Changeset in webkit [127007] by
-
- 1 edit1 add in trunk/LayoutTests
Unreviewed gardening.
- platform/chromium-linux/fast/css/absolute-child-with-percent-height-inside-relative-parent-expected.png: Added.
Added missing baseline for this test.
- 9:25 AM Changeset in webkit [127006] by
-
- 16 edits2 moves3 deletes in trunk
Unreviewed, rolling out r126972.
http://trac.webkit.org/changeset/126972
https://bugs.webkit.org/show_bug.cgi?id=95349
accessibility/canvas-description-and-role.html has been
failing consistently on several bots and Dominic needs some
time to investigate why (Requested by jchaffraix on #webkit).
Patch by Sheriff Bot <webkit.review.bot@gmail.com> on 2012-08-29
Source/WebCore:
- accessibility/AccessibilityNodeObject.cpp:
- accessibility/AccessibilityNodeObject.h:
- accessibility/AccessibilityObject.h:
(AccessibilityObject):
- accessibility/AccessibilityRenderObject.cpp:
(WebCore::AccessibilityRenderObject::accessibilityDescription):
(WebCore::AccessibilityRenderObject::accessibilityIsIgnored):
(WebCore::AccessibilityRenderObject::determineAccessibilityRole):
(WebCore::AccessibilityRenderObject::canHaveChildren):
- accessibility/gtk/WebKitAccessibleWrapperAtk.cpp:
(atkRole):
- accessibility/mac/WebAccessibilityObjectWrapper.mm:
(createAccessibilityRoleMap):
(-[WebAccessibilityObjectWrapper role]):
Source/WebKit/chromium:
- public/WebAccessibilityRole.h:
- src/AssertMatchingEnums.cpp:
Source/WebKit/win:
- AccessibleBase.cpp:
(MSAARole):
Tools:
- DumpRenderTree/chromium/TestRunner/AccessibilityUIElementChromium.cpp:
(roleToString):
LayoutTests:
- accessibility/canvas-description-and-role.html: Removed.
- accessibility/canvas-expected.txt: Renamed from LayoutTests/platform/mac/accessibility/canvas-expected.txt.
- accessibility/canvas.html: Renamed from LayoutTests/platform/mac/accessibility/canvas.html.
- platform/chromium/accessibility/canvas-description-and-role-expected.txt: Removed.
- platform/gtk/TestExpectations:
- platform/mac/accessibility/canvas-description-and-role-expected.txt: Removed.
- 9:17 AM Changeset in webkit [127005] by
-
- 2 edits in trunk/Source/WebCore
Web Inspector: Minor ResourceScriptMapping polish.
https://bugs.webkit.org/show_bug.cgi?id=95350
Reviewed by Alexander Pavlov.
Added some compiler annotations.
ResourceScriptMapping is no longer a UISourceCodeProvider, removed unused methods and obsolete compiler annotations.
- inspector/front-end/ResourceScriptMapping.js:
(WebInspector.ResourceScriptMapping):
- 9:12 AM Changeset in webkit [127004] by
-
- 3 edits in trunk/LayoutTests
Web Inspector: Minor debugger-test polish.
https://bugs.webkit.org/show_bug.cgi?id=95348
Reviewed by Alexander Pavlov.
Added meaningful assertion messages to facilitate debugging.
Added meaningful mimeType to script mock.
- http/tests/inspector/debugger-test.js:
(initialize_DebuggerTest.):
(initialize_DebuggerTest):
- inspector/debugger/raw-source-code-expected.txt:
- 9:06 AM Changeset in webkit [127003] by
-
- 2 edits in trunk/Tools
[EFL] Resolve CMake warnings on overlapping search paths for EFL jhbuild-enabled build
https://bugs.webkit.org/show_bug.cgi?id=84707
Reviewed by Gustavo Noronha Silva.
Most of the warnings have been fixed now that FindCairo.cmake and
FindGLIB.cmake have been rewritten. The remaining one, related to
FindFontconfig.cmake, shows up when building on a 64-bit
Debian-based systems.
jhbuild installs libraries into lib64/ by default on 64-bit Linux
installations, while CMake does not look for libraries in lib64/
when /etc/debian_version exists on the system. The FIND_LIBRARY()
would then sometimes end up using the system installation instead of
the jhbuild one, causing mismatches and, when pkg-config is not used
at all, failing to find libraries altogether.
- efl/jhbuildrc: Set the CMAKE_LIBRARY_PATH environment variable
when use_lib64 is set to force CMake to look into lib64/ regardless
of the presence of /etc/debian_version.
- 8:39 AM Changeset in webkit [127002] by
-
- 8 edits1 copy in trunk/Source/WebCore
Web Inspector: Extract StylesSourceMapping from StylesUISourceCodeProvider.
https://bugs.webkit.org/show_bug.cgi?id=95345
Reviewed by Alexander Pavlov.
- WebCore.gypi:
- WebCore.vcproj/WebCore.vcproj:
- inspector/compile-front-end.py:
- inspector/front-end/ResourceScriptMapping.js:
(WebInspector.ResourceScriptMapping): Drive-by: ResourceScriptMapping now uses workspace object passed in parameters, not the global one.
(WebInspector.ResourceScriptMapping.prototype._uiSourceCodeAdded):
(WebInspector.ResourceScriptMapping.prototype._uiSourceCodeReplaced):
- inspector/front-end/StylesSourceMapping.js: Copied from Source/WebCore/inspector/front-end/StylesUISourceCodeProvider.js.
(WebInspector.StylesSourceMapping):
(WebInspector.StylesSourceMapping.prototype.addUISourceCode):
(WebInspector.StylesSourceMapping.prototype.rawLocationToUILocation):
(WebInspector.StylesSourceMapping.prototype.uiLocationToRawLocation):
(WebInspector.StylesSourceMapping.prototype.reset):
- inspector/front-end/StylesUISourceCodeProvider.js:
(WebInspector.StylesUISourceCodeProvider):
(WebInspector.StylesUISourceCodeProvider.prototype._resourceAdded):
(WebInspector.StylesUISourceCodeProvider.prototype._reset): Drive-by: _populate is now called with setTimeout.
- inspector/front-end/WebKit.qrc:
- inspector/front-end/inspector.html:
- 8:33 AM Changeset in webkit [127001] by
-
- 9 edits1 add in trunk
The 2d.imageData.object.round canvas test is failing
https://bugs.webkit.org/show_bug.cgi?id=40272
Reviewed by Benjamin Poulain.
Source/WTF:
Updating previous patch to address Benjamin's comments.
#ifdef in Uint8ClampedArray removed, fallback implementation for MSVC on non-X86 added.
- wtf/MathExtras.h:
(lrint): Fallback implementation for non-X86 & MSVC case.
- wtf/Uint8ClampedArray.h: Removed #ifdef.
Tools:
Updating patch to address Benjamin's review comments.
Adding a WTF test to test lrint implementation.
- TestWebKitAPI/CMakeLists.txt: Added MathExtras.cpp test file.
- TestWebKitAPI/GNUmakefile.am: Added MathExtras.cpp test file.
- TestWebKitAPI/TestWebKitAPI.gypi: Added MathExtras.cpp test file.
- TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added MathExtras.cpp test file.
- TestWebKitAPI/Tests/WTF/MathExtras.cpp: Added this test file containing a test for lrint().
(TestWebKitAPI):
(TestWebKitAPI::TEST):
- TestWebKitAPI/win/TestWebKitAPI.vcproj: Added MathExtras.cpp test file.
- 7:48 AM Changeset in webkit [127000] by
-
- 3 edits4 adds in trunk
Web Inspector: Page with @import and :last-child in an edited stylesheet will crash
https://bugs.webkit.org/show_bug.cgi?id=95324
Reviewed by Antti Koivisto.
Source/WebCore:
Ensure the destroyed StyleRules removal from StyleResolver by creating a separate RuleMutationScope for clearing the StyleSheetContents.
Test: inspector/styles/import-pseudoclass-crash.html
- inspector/InspectorStyleSheet.cpp:
(WebCore::InspectorStyleSheet::reparseStyleSheet):
LayoutTests:
- inspector/styles/import-pseudoclass-crash-expected.txt: Added.
- inspector/styles/import-pseudoclass-crash.html: Added.
- inspector/styles/resources/import-pseudoclass-crash-empty.css: Added.
- inspector/styles/resources/import-pseudoclass-crash.css: Added.
(:last-child):
- 6:58 AM WebInspector edited by
- (diff)
- 6:52 AM Changeset in webkit [126999] by
-
- 17 edits in trunk
Web Inspector: Turn workspace into a container of UiSourceCodes put in different projects.
https://bugs.webkit.org/show_bug.cgi?id=95335
Reviewed by Pavel Feldman.
Source/WebCore:
Workspace now contains a project that could be filled with uiSourceCodes.
Workspace project is still filled by script / style mappings as before.
The next step would be to extract NetworkUISourceCodeProvider.
- inspector/front-end/BreakpointManager.js:
(WebInspector.BreakpointManager):
- inspector/front-end/CompilerScriptMapping.js:
(WebInspector.CompilerScriptMapping):
(WebInspector.CompilerScriptMapping.prototype._reset):
- inspector/front-end/DebuggerScriptMapping.js:
(WebInspector.DebuggerScriptMapping):
- inspector/front-end/ResourceScriptMapping.js:
(WebInspector.ResourceScriptMapping):
(WebInspector.ResourceScriptMapping.prototype._uiSourceCodeAdded):
(WebInspector.ResourceScriptMapping.prototype._uiSourceCodeReplaced):
(WebInspector.ResourceScriptMapping.prototype._reset):
- inspector/front-end/RevisionHistoryView.js:
(WebInspector.RevisionHistoryView):
- inspector/front-end/SASSSourceMapping.js:
(WebInspector.SASSSourceMapping):
(_bindUISourceCode):
(rawLocationToUILocation):
(_reset):
- inspector/front-end/ScriptSnippetModel.js:
(WebInspector.ScriptSnippetModel):
(WebInspector.ScriptSnippetModel.prototype._addScriptSnippet):
(WebInspector.ScriptSnippetModel.prototype.deleteScriptSnippet):
(WebInspector.SnippetScriptMapping.prototype.addScript):
- inspector/front-end/ScriptsPanel.js:
(WebInspector.ScriptsPanel):
- inspector/front-end/StylesUISourceCodeProvider.js:
(WebInspector.StylesUISourceCodeProvider):
(WebInspector.StylesUISourceCodeProvider.prototype._resourceAdded):
(WebInspector.StylesUISourceCodeProvider.prototype._reset):
- inspector/front-end/Workspace.js:
(WebInspector.WorkspaceController):
(WebInspector.WorkspaceController.prototype._mainFrameNavigated):
(WebInspector.Project):
(WebInspector.Project.prototype.reset):
(WebInspector.Project.prototype.addUISourceCode):
(WebInspector.Project.prototype.replaceUISourceCode):
(WebInspector.Project.prototype.removeUISourceCode):
(WebInspector.Project.prototype.uiSourceCodeForURL):
(WebInspector.Project.prototype.uiSourceCodes):
(WebInspector.Workspace):
(WebInspector.Workspace.prototype.uiSourceCodeForURL):
(WebInspector.Workspace.prototype.project):
(WebInspector.Workspace.prototype.uiSourceCodes):
- inspector/front-end/inspector.js:
LayoutTests:
- http/tests/inspector/compiler-script-mapping.html:
- inspector/debugger/breakpoint-manager.html:
- inspector/debugger/script-snippet-model.html:
- inspector/debugger/scripts-panel.html:
- 6:40 AM Changeset in webkit [126998] by
-
- 4 edits in trunk/Source
[Qt] Fix doc generation with make docs
https://bugs.webkit.org/show_bug.cgi?id=95340
Reviewed by Kenneth Rohde Christiansen.
Fix doc target, similar to what the other Qt 5 modules are using and comment out the indexes
line, because cross-referencing doesn't work right now and the use of the QTDIR environment
is wrong, too.
- docs/docs.pri:
- docs/qtwebkit.qdocconf:
- 6:15 AM Changeset in webkit [126997] by
-
- 2 edits in trunk/Tools
Unreviewed, updating Kelly Norton's email address at his request.
- Scripts/webkitpy/common/config/committers.py:
- 6:00 AM Changeset in webkit [126996] by
-
- 2 edits in trunk/Source/WebKit/blackberry
[BlackBerry] crash while trying to scroll any inner frame
https://bugs.webkit.org/show_bug.cgi?id=95287
PR #198510
Reviewed by George Staikos.
Patch by Antonio Gomes <agomes@rim.com>
Set the starting point of scrolling (slow path only still)
for inner frames. It was left over as set to 0 by r126474.
In practice, patch fixes a crash on gmail.com desktop edition.
- WebKitSupport/InRegionScrollableArea.cpp:
(BlackBerry::WebKit::InRegionScrollableArea::InRegionScrollableArea):
- 5:40 AM Changeset in webkit [126995] by
-
- 2 edits in trunk/LayoutTests
[WK2] New fast/events/tab-focus-link-in-canvas fails from r126908
https://bugs.webkit.org/show_bug.cgi?id=95329
Patch by Szilard Ledan <Szilárd LEDÁN> on 2012-08-29
Reviewed by Csaba Osztrogonác.
- platform/wk2/Skipped:
- 5:27 AM Changeset in webkit [126994] by
-
- 2 edits in trunk/LayoutTests
[EFL] Unreviewed EFL gardening.
Moving fast/forms/color/input-value-sanitization-color.html out as flaky.
- platform/efl-wk2/TestExpectations: Flakily crashing test marked.
- 5:24 AM Changeset in webkit [126993] by
-
- 3 edits2 adds in trunk
Incorrect large-area clipping
https://bugs.webkit.org/show_bug.cgi?id=95197
Reviewed by Nikolas Zimmermann.
Source/WebCore:
ImageBuffers allocated for clipping and masking are clamped to kMaxImageBufferSize max
(4096x4096). In order to properly account for the scaling factor introduced by this
clamping, the repaintRect translation component needs to be pushed after the scaling
transform.
Tests: svg/custom/clamped-masking-clipping-expected.svg
svg/custom/clamped-masking-clipping.svg
- rendering/svg/SVGRenderingContext.cpp:
(WebCore::SVGRenderingContext::createImageBuffer):
LayoutTests:
- svg/custom/clamped-masking-clipping-expected.svg: Added.
- svg/custom/clamped-masking-clipping.svg: Added.
- 5:18 AM Changeset in webkit [126992] by
-
- 3 edits in trunk/Tools
[NRWT] The nrwt should check the contents of the skipped files with --lint-test-files
https://bugs.webkit.org/show_bug.cgi?id=93723
Reviewed by Dirk Pranke.
- Scripts/webkitpy/layout_tests/models/test_expectations.py:
(TestExpectationParser.expectation_for_skipped_test):
- Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
(SkippedTests.test_skipped_entry_dont_exist):
- 5:08 AM Changeset in webkit [126991] by
-
- 2 edits in trunk/LayoutTests
[EFL] Move known crash issue to TestExpectations
https://bugs.webkit.org/show_bug.cgi?id=95327
Unreviewed, EFL gardening.
Moving out x-frame-options-deny-meta-tag-in-body.html with details discussed in bug 94458.
- platform/efl/TestExpectations: Marked crash issue, discussed in bug 94458.
- 4:54 AM Changeset in webkit [126990] by
-
- 2 edits in trunk/Source/JavaScriptCore
Unreviewed. Build fix for !ENABLE(JIT) after r126962.
- interpreter/Interpreter.cpp:
(JSC::Interpreter::privateExecute):
- 4:53 AM Changeset in webkit [126989] by
-
- 2 edits in trunk/Source/WebCore
Web Inspector: unsafe static_cast in RetainedDOMInfo::IsEquivalent
https://bugs.webkit.org/show_bug.cgi?id=95315
Reviewed by Vsevolod Vlasov.
- bindings/v8/RetainedDOMInfo.cpp:
(WebCore::RetainedDOMInfo::IsEquivalent): compare retained object info
labels before casting v8::RetainedObjectInfo to WebCore::RetainedObjectInfo
as the |other| object may not be a descendant of WebCore::RetainedObjectInfo.
- 4:29 AM Changeset in webkit [126988] by
-
- 1 edit1 add in trunk/LayoutTests
[EFL] Unreviewed EFL gardening.
EFL requires a newline at the end of the test expectations here.
- platform/efl/fast/events/overflow-viewport-renderer-deleted-expected.txt: Added.
- 4:24 AM Changeset in webkit [126987] by
-
- 4 edits in trunk/Source/WebKit/efl
[EFL] Add *explicit* keyword to constructors in WebCoreSupport
https://bugs.webkit.org/show_bug.cgi?id=95307
Reviewed by Kentaro Hara.
Added explicit keyword in constructors in order to avoid implicit type conversion.
- WebCoreSupport/ColorChooserEfl.h:
(ColorChooserEfl):
- WebCoreSupport/PopupMenuEfl.h:
(PopupMenuEfl):
- WebCoreSupport/SearchPopupMenuEfl.h:
(SearchPopupMenuEfl):
- 4:24 AM Changeset in webkit [126986] by
-
- 3 edits in trunk/Source/WebCore
REGRESSION(r126816): Missing includes when compiling without SVG
https://bugs.webkit.org/show_bug.cgi?id=95312
Reviewed by Eric Seidel.
- rendering/FilterEffectRenderer.h:
- rendering/RenderLayerFilterInfo.h:
- 4:19 AM Changeset in webkit [126985] by
-
- 2 edits in trunk/Tools
Unreviewed, rolling out r120113.
http://trac.webkit.org/changeset/120113
https://bugs.webkit.org/show_bug.cgi?id=95320
Wrong fix for the problem, experimentally rolling it out for
bug 95237. (Requested by rakuco on #webkit).
Patch by Sheriff Bot <webkit.review.bot@gmail.com> on 2012-08-29
- efl/jhbuildrc:
- 4:16 AM Changeset in webkit [126984] by
-
- 7 edits in trunk/Source/WebCore
Improve string efficiency using StringBuilder and StringOperations
https://bugs.webkit.org/show_bug.cgi?id=95304
Reviewed by Eric Seidel.
As recommended by http://trac.webkit.org/wiki/EfficientStrings.
- css/CSSLineBoxContainValue.cpp:
(WebCore::CSSLineBoxContainValue::customCssText):
- css/CSSPropertySourceData.cpp:
(WebCore::CSSPropertySourceData::toString):
- css/MediaList.cpp:
(WebCore::MediaQuerySet::mediaText):
- css/ShadowValue.cpp:
(WebCore::ShadowValue::customCssText):
- dom/MicroDataItemList.cpp:
(WebCore::MicroDataItemList::undefinedItemType):
- editing/HTMLInterchange.cpp:
(WebCore::convertedSpaceString):
- 4:13 AM Changeset in webkit [126983] by
-
- 4 edits11 adds in trunk/LayoutTests
[EFL] Clean up Skipped file
https://bugs.webkit.org/show_bug.cgi?id=95306
Unreviewed EFL gardening.
Remove from Skipped list several test cases that are now passing
consistently on EFL port. Some other tests have been moved to
TestExpectations since we need to get rid of the Skipped list
progressively.
Patch by Christophe Dumez <Christophe Dumez> on 2012-08-29
- platform/efl-wk2/TestExpectations:
- platform/efl/Skipped:
- platform/efl/TestExpectations:
- platform/efl/editing/selection/collapse-selection-in-bidi-expected.txt: Added.
- platform/efl/editing/selection/extend-selection-bidi-expected.png: Added.
- platform/efl/editing/selection/extend-selection-bidi-expected.txt: Added.
- platform/efl/fast/box-shadow/shadow-tiling-artifact-expected.txt: Added.
- platform/efl/fast/forms/color/input-color-onchange-event-expected.txt: Added.
- platform/efl/fast/repaint/no-caret-repaint-in-non-content-editable-element-expected.png: Added.
- platform/efl/fast/repaint/no-caret-repaint-in-non-content-editable-element-expected.txt: Added.
- platform/efl/fast/replaced/object-with-embed-url-param-expected.txt: Added.
- platform/efl/fast/text/complex-synthetic-bold-space-width-expected.png: Added.
- platform/efl/fast/text/complex-synthetic-bold-space-width-expected.txt: Added.
- 4:00 AM Changeset in webkit [126982] by
-
- 3 edits in trunk/LayoutTests
[EFL] Rebaseline after r126864
https://bugs.webkit.org/show_bug.cgi?id=95278
Unreviewed, rebaseline.
Updated range with datalist pixel test expectations because slider theme is changed.
Patch by KwangYong Choi <ky0.choi@samsung.com> on 2012-08-29
- platform/efl/fast/forms/datalist/input-appearance-range-with-datalist-expected.png:
- platform/efl/fast/forms/datalist/input-appearance-range-with-datalist-zoomed-expected.png:
- 3:52 AM Changeset in webkit [126981] by
-
- 3 edits in trunk/Source/WebCore
[chromium] Use floating point literals in expressions that initialize floats
https://bugs.webkit.org/show_bug.cgi?id=95293
Patch by James Robinson <jamesr@chromium.org> on 2012-08-29
Reviewed by Adam Barth.
Visual studio's C4305, which is on for some chromium code, complains about these.
- platform/graphics/chromium/cc/CCLayerSorter.cpp:
(WebCore::CCLayerSorter::createGraphEdges):
- platform/graphics/chromium/cc/CCLayerTreeHostImpl.cpp:
(WebCore::CCLayerTreeHostImpl::scrollBy):
(WebCore::CCLayerTreeHostImpl::computePinchZoomDeltas):
- 3:46 AM Changeset in webkit [126980] by
-
- 2 edits in trunk/Source/WebKit/chromium
Unreviewed. Rolled DEPS.
- DEPS:
- 3:31 AM Changeset in webkit [126979] by
-
- 7 edits1 copy in trunk/Source/WebCore
Web Inspector: Extract StylesUISourceCodeProvider to separate file.
https://bugs.webkit.org/show_bug.cgi?id=95319
Reviewed by Alexander Pavlov.
- WebCore.gypi:
- WebCore.vcproj/WebCore.vcproj: Also added WebGLProfiler files forgotten before.
- inspector/compile-front-end.py:
- inspector/front-end/StyleSource.js:
- inspector/front-end/StylesUISourceCodeProvider.js: Copied from Source/WebCore/inspector/front-end/StyleSource.js.
(WebInspector.StylesUISourceCodeProvider):
(WebInspector.StylesUISourceCodeProvider.prototype.uiSourceCodes):
(WebInspector.StylesUISourceCodeProvider.prototype.rawLocationToUILocation):
(WebInspector.StylesUISourceCodeProvider.prototype.uiLocationToRawLocation):
(WebInspector.StylesUISourceCodeProvider.prototype._populate):
(WebInspector.StylesUISourceCodeProvider.prototype._resourceAdded):
(WebInspector.StylesUISourceCodeProvider.prototype.reset):
- inspector/front-end/WebKit.qrc:
- inspector/front-end/inspector.html:
- 3:27 AM Changeset in webkit [126978] by
-
- 2 edits in branches/chromium/1229
Merge 126959 - style->fontMetrics() should be available when setting line-height
https://bugs.webkit.org/show_bug.cgi?id=93327
Reviewed by Darin Adler.
Source/WebCore:
Setting line-height assumes the fontMetrics are available for the affected font, but
the fontMetrics won't be available immediately after setting other properties like
font-size. Call styleResolver->updateFont() before setting line-height to update fontMetrics.
Added a test case to fast/canvas/crash-set-font.html.
- html/canvas/CanvasRenderingContext2D.cpp:
(WebCore::CanvasRenderingContext2D::setFont):
Call styleResolver->updateFont() before styleResolver->applyPropertyToCurrentStyle(CSSPropertyLineHeight,...)
LayoutTests:
- fast/canvas/crash-set-font.html: Add a test case that sets both font-size and line-height.
TBR=bashi@chromium.org
Review URL: https://chromiumcodereview.appspot.com/10885027
- 3:17 AM Changeset in webkit [126977] by
-
- 3 edits2 adds in trunk
Use SVGImage instead of cached image when drawing without a render tree.
https://bugs.webkit.org/show_bug.cgi?id=95002
Reviewed by Nikolas Zimmermann.
Source/WebCore:
Previously if we tried to use canvas.context2d.drawImage() with an SVG image
that was not in the render tree, we would crash. This patch changes this behavior
so that we use SVGImage::draw() to draw images that are not in the render tree.
Test: svg/as-image/svg-canvas-draw-image-detached.html
- svg/graphics/SVGImageCache.cpp:
(WebCore::SVGImageCache::requestedSizeAndScales):
(WebCore::SVGImageCache::lookupOrCreateBitmapImageForRenderer):
LayoutTests:
- svg/as-image/svg-canvas-draw-image-detached-expected.txt: Added.
- svg/as-image/svg-canvas-draw-image-detached.html: Added.
- 3:15 AM Changeset in webkit [126976] by
-
- 9 edits in trunk
[Qt] REGRESSION(r125428): fast/profiler/nested-start-and-stop-profiler.html fails
https://bugs.webkit.org/show_bug.cgi?id=93897
Reviewed by Kenneth Rohde Christiansen.
Source/WebCore:
Before r125428 run-time methods (wrapped signals, slots or invokable
functions) were subclasses of JSInternalFunction and therefore real
function objects in the JavaScript sense. r125428 changed them to be
just callable objects, but they did not have Function.prototype as
prototype anymore for example nor was their name correct (resulting in
a layout test failure).
This patch changes run-time methods back to being real function objects
that have a correct name and have Function.prototype in their prototype
change
The objects returned by JSObjectMakeFunctionWithCallbackInjected are
light-weight internal function objects that do not support
JSObject{Set/Get}Private. Therefore we inject our own prototype right
before the Function.prototype prototype, which uses private data to
store a pointer to our C++ QtRuntimeMethod object. This complicates
the retrieval of the pointer to that instance slightly, which is why
this patch introduces the toRuntimeMethod convenience function that
looks up our prototype first and does a check for type-safety.
At the same time the patch removes the length properties from the
run-time method itself as well as connect/disconnect. The length
property on a function signifies the number of arguments, but in all
three cases that number is actually variable, because of overloading.
That is why we choose not to expose it in the first place.
In QtInstance we cache the JS wrapper objects for QtRuntimeMethod in a
JSWeakObjectMap. JSWeakObjectMap requires the stored objects to be
either the result of JSObjectMake or the global object of a context ref
(AFAICS), which is ensured using an ASSERT. Objects created via
JSObjectMakeFunctionWithCalllback do not fall into the required
category, cause a failing assertion and can therefore not be stored in
the weak object map.
Consequently this patch removes the use of JSWeakObjectMap again and
goes back to the old way of using the internal Weak<> API, for the time
being. In a future patch the storage will be simplified to not require
the use of a weak object map cache for the run-time methods anymore.
- bridge/qt/qt_instance.cpp: Remove unused WeakMap code.
- bridge/qt/qt_instance.h: Remove method cache.
(QtInstance):
- bridge/qt/qt_runtime.cpp:
(JSC::Bindings::prototypeForSignalsAndSlots):
(JSC::Bindings::QtRuntimeMethod::call):
(JSC::Bindings::QtRuntimeMethod::jsObjectRef):
(JSC::Bindings::QtRuntimeMethod::toRuntimeMethod):
(Bindings):
(JSC::Bindings::QtRuntimeMethod::connectOrDisconnect):
- bridge/qt/qt_runtime.h:
(QtRuntimeMethod): Remove unused member variables.
Source/WebKit/qt:
Fixed some test expectations.
- tests/qobjectbridge/tst_qobjectbridge.cpp:
(tst_QObjectBridge::objectDeleted): Since runtime methods are real function objects again, we
can go back to testing Function.prototype.call, as it was done before r125428.
(tst_QObjectBridge::introspectQtMethods_data): Removed tests for the length property.
(tst_QObjectBridge::introspectQtMethods): Changed test expectation of the properties of
run-time methods back to being non-configurable, as before r125428.
LayoutTests:
- platform/qt/Skipped: Unskip test that is now passing.
- 2:37 AM Changeset in webkit [126975] by
-
- 2 edits in trunk/Source/WebCore
Unreviewed: Single line build fix.
- platform/graphics/BitmapImage.cpp:
(WebCore::BitmapImage::reportMemoryUsage):
- 2:28 AM Changeset in webkit [126974] by
-
- 15 edits1 delete in trunk/Source/WebCore
Web Inspector: NMI: Instrument WebCore part of the Image class hierarchy
https://bugs.webkit.org/show_bug.cgi?id=94959
Drive by fix: remove unused GeneratedImage.cpp
Reviewed by Yury Semikhatsky.
- html/HTMLImageElement.cpp:
(WebCore::HTMLImageElement::reportMemoryUsage):
(WebCore):
- html/HTMLImageElement.h:
(HTMLImageElement):
- loader/cache/CachedImage.cpp:
(WebCore::CachedImage::reportMemoryUsage):
- platform/graphics/BitmapImage.cpp:
(WebCore::BitmapImage::reportMemoryUsage):
(WebCore):
- platform/graphics/BitmapImage.h:
(FrameData):
(BitmapImage):
- platform/graphics/CrossfadeGeneratedImage.cpp:
(WebCore::CrossfadeGeneratedImage::reportMemoryUsage):
(WebCore):
- platform/graphics/CrossfadeGeneratedImage.h:
(WebCore):
(CrossfadeGeneratedImage):
- platform/graphics/GeneratedImage.cpp: Removed.
- platform/graphics/GeneratedImage.h:
(GeneratedImage):
- platform/graphics/GeneratorGeneratedImage.cpp:
(WebCore::GeneratedImage::reportMemoryUsage):
(WebCore):
(WebCore::GeneratorGeneratedImage::reportMemoryUsage):
- platform/graphics/GeneratorGeneratedImage.h:
(GeneratorGeneratedImage):
- platform/graphics/Image.cpp:
(WebCore::Image::reportMemoryUsage):
(WebCore):
- platform/graphics/Image.h:
(WebCore):
(Image):
- svg/graphics/SVGImage.cpp:
(WebCore::SVGImage::reportMemoryUsage):
(WebCore):
- svg/graphics/SVGImage.h:
(SVGImage):
- 2:02 AM WebKit Team edited by
- Adding myself to committers list. (diff)
- 1:56 AM EfficientStrings edited by
- Some more hints (diff)
- 1:21 AM Changeset in webkit [126973] by
-
- 2 edits in trunk/LayoutTests
[Qt][WK1] Unreviewd gardening. Skip failing css3 filter test.
https://bugs.webkit.org/show_bug.cgi?id=95308
Patch by Zoltan Arvai <zarvai@inf.u-szeged.hu> on 2012-08-29
- platform/qt-5.0-wk1/Skipped: skip css3/filters/null-effect-check.html after r126927.
- 1:05 AM Changeset in webkit [126972] by
-
- 16 edits2 moves3 adds in trunk
AX: Canvas should have a distinct role
https://bugs.webkit.org/show_bug.cgi?id=95248
Reviewed by Chris Fleizach.
Source/WebCore:
Add new role for a canvas element, and a method to determine if
a canvas has fallback content, so each platform can decide on the
appropriate role mapping to use.
Test: accessibility/canvas-description-and-role.html
- accessibility/AccessibilityNodeObject.cpp:
(WebCore::AccessibilityNodeObject::canvasHasFallbackContent):
(WebCore):
- accessibility/AccessibilityNodeObject.h:
(AccessibilityNodeObject):
- accessibility/AccessibilityObject.h:
(WebCore::AccessibilityObject::isCanvas):
(WebCore::AccessibilityObject::canvasHasFallbackContent):
- accessibility/AccessibilityRenderObject.cpp:
(WebCore::AccessibilityRenderObject::accessibilityIsIgnored):
(WebCore::AccessibilityRenderObject::determineAccessibilityRole):
(WebCore::AccessibilityRenderObject::canHaveChildren):
- accessibility/gtk/WebKitAccessibleWrapperAtk.cpp:
(atkRole):
- accessibility/mac/WebAccessibilityObjectWrapper.mm:
(createAccessibilityRoleMap):
(-[WebAccessibilityObjectWrapper role]):
Source/WebKit/chromium:
Add support for canvas accessibility role.
- public/WebAccessibilityRole.h:
- src/AssertMatchingEnums.cpp:
Source/WebKit/win:
Map new CanvasRole to the same as ImageRole.
- AccessibleBase.cpp:
(MSAARole):
Tools:
Add support for canvas accessibility role.
- DumpRenderTree/chromium/TestRunner/AccessibilityUIElementChromium.cpp:
(roleToString):
LayoutTests:
Add new tests for canvas role.
- accessibility/canvas.html: Deleted.
- accessibility/canvas-expected.txt: Deleted.
- accessibility/canvas-description-and-role.html: Added.
- platform/chromium/accessibility/canvas-description-and-role-expected.txt: Added.
- platform/gtk/TestExpectations:
- platform/mac/accessibility/canvas.html: Added.
- platform/mac/accessibility/canvas-expected.txt: Added.
- platform/mac/accessibility/canvas-description-and-role-expected.txt: Added.
- 1:02 AM Changeset in webkit [126971] by
-
- 6 edits6 moves in trunk/Source
[EFL] Move several files to remove webkit1 dependency from WebCore.
https://bugs.webkit.org/show_bug.cgi?id=95264
Reviewed by Gyuyoung Kim.
Source/WebCore:
ColorChooserEfl.cpp, PopupMenuEfl.cpp and SearchPopupMenuEfl.cpp are WebKit1/Efl
specific files and they can not be shared with WebKit2/Efl.
In order to remove WebKit dependency from WebCore, this patch moves them
to WebKit/efl/WebCoreSupport.
No behavior change. Just a refactoring.
- PlatformEfl.cmake: Removed files which moved from sources and WebKit/efl/WebCoreSupport from includes.
- WebCore.gypi: Ditto.
Source/WebKit:
- PlatformEfl.cmake: Added files which is moved.
Source/WebKit/efl:
ColorChooserEfl.cpp, PopupMenuEfl.cpp and SearchPopupMenuEfl.cpp are WebKit1/Efl
specific files and they can not be shared with WebKit2/Efl.
In order to remove WebKit dependency from WebCore, this patch moves them
to WebKit/efl/WebCoreSupport.
- WebCoreSupport/ColorChooserEfl.cpp: Renamed from Source/WebCore/platform/efl/ColorChooserEfl.cpp.
(WebCore):
(WebCore::ColorChooserEfl::ColorChooserEfl):
(WebCore::ColorChooserEfl::~ColorChooserEfl):
(WebCore::ColorChooserEfl::setSelectedColor):
(WebCore::ColorChooserEfl::endChooser):
- WebCoreSupport/ColorChooserEfl.h: Renamed from Source/WebCore/platform/efl/ColorChooserEfl.h.
(WebCore):
(ColorChooserEfl):
- WebCoreSupport/PopupMenuEfl.cpp: Renamed from Source/WebCore/platform/efl/PopupMenuEfl.cpp.
(WebCore):
(WebCore::PopupMenuEfl::PopupMenuEfl):
(WebCore::PopupMenuEfl::~PopupMenuEfl):
(WebCore::PopupMenuEfl::show):
(WebCore::PopupMenuEfl::hide):
(WebCore::PopupMenuEfl::updateFromElement):
(WebCore::PopupMenuEfl::disconnectClient):
- WebCoreSupport/PopupMenuEfl.h: Renamed from Source/WebCore/platform/efl/PopupMenuEfl.h.
(WebCore):
(PopupMenuEfl):
(WebCore::PopupMenuEfl::client):
- WebCoreSupport/SearchPopupMenuEfl.cpp: Renamed from Source/WebCore/platform/efl/SearchPopupMenuEfl.cpp.
(WebCore):
(WebCore::SearchPopupMenuEfl::SearchPopupMenuEfl):
(WebCore::SearchPopupMenuEfl::popupMenu):
(WebCore::SearchPopupMenuEfl::saveRecentSearches):
(WebCore::SearchPopupMenuEfl::loadRecentSearches):
(WebCore::SearchPopupMenuEfl::enabled):
- WebCoreSupport/SearchPopupMenuEfl.h: Renamed from Source/WebCore/platform/efl/SearchPopupMenuEfl.h.
(WebCore):
(SearchPopupMenuEfl):
- 12:56 AM Changeset in webkit [126970] by
-
- 11 edits2 adds in trunk
AX: Focusable elements without a role should not be ignored
https://bugs.webkit.org/show_bug.cgi?id=94302
Reviewed by Chris Fleizach.
Source/WebCore:
Changes the accessibility logic so that a generic element that's focusable is
not ignored for accessibility, and returns its inner text as its title. That way
if you Tab to the element, a reasonable accessibility notification is generated.
One exception is the body element, because focusing the body is equivalent to
blurring the current focused element and does not result in a "focus" accessibility
notification.
Also fixes logic that determined if an element was contentEditable by making
sure it catches the case with no attribute value (e.g. <div contentEditable>),
which also implies contentEditable=true according to the spec.
Test: accessibility/focusable-div.html
- accessibility/AccessibilityRenderObject.cpp:
(WebCore):
(WebCore::nodeHasContentEditableAttributeSet):
(WebCore::AccessibilityRenderObject::title):
(WebCore::AccessibilityRenderObject::accessibilityIsIgnored):
LayoutTests:
Adds a new test to make sure that a generic focusable element (like a div with tabindex=0)
can get focus and return an appropriate title, just like a form control or element with
an ARIA role.
Modifies three existing tests that were previously assuming that a focusable node
with no role would be ignored for accessibility ("accessibilityIsIgnored").
- accessibility/editable-webarea-context-menu-point.html:
- accessibility/focusable-div-expected.txt: Added.
- accessibility/focusable-div.html: Added.
- accessibility/table-detection.html:
- platform/mac/accessibility/listbox-hit-test.html:
- 12:48 AM Changeset in webkit [126969] by
-
- 2 edits in trunk/LayoutTests
[EFL] Unreviewed gardening on the 64bit build bot.
The following test cases don't crash so they should be removed from
WK2 TestExpectations file.
inspector/elements/event-listeners-about-blank.html
inspector/console/console-assert.html
inspector/debugger/debugger-step-out.html
inspector/debugger/debugger-suspend-active-dom-objects.html
fast/forms/fieldset/fieldset-crash.html
fast/events/keyevent-iframe-removed-crash.html
The following test cases pass only in WK2 so marked them as pass.
fast/loader/opaque-base-url.html
canvas/philip/tests/2d.text.draw.fontface.notinpage.html
editing/execCommand/indent-paragraphs.html
fast/animation/request-animation-frame-during-modal.html
fast/forms/mailto/formenctype-attribute-button-html.html
fast/forms/mailto/formenctype-attribute-input-html.html
fast/frames/seamless/seamless-inherited-document-style.html
fast/js/names.html
http/tests/security/cross-frame-access-call.html
http/tests/security/frameNavigation/inactive-function-in-popup-navigate-child.html
The following test case is marked as known failure.
fast/events/tab-focus-link-in-canvas.html
- platform/efl-wk2/TestExpectations:
- 12:38 AM Changeset in webkit [126968] by
-
- 38 edits in trunk/Source/WebCore
Deploy ASCIILiteral hotness throughout WebCore
https://bugs.webkit.org/show_bug.cgi?id=95282
Reviewed by Eric Seidel.
As recommended by http://trac.webkit.org/wiki/EfficientStrings.
This patch converts all the DEFINE_STATIC_LOCAL Strings.
- Modules/indexeddb/IDBDatabase.cpp:
(WebCore::IDBDatabase::transaction):
- Modules/indexeddb/IDBIndex.cpp:
(WebCore::IDBIndex::openCursor):
(WebCore::IDBIndex::openKeyCursor):
- Modules/indexeddb/IDBObjectStore.cpp:
(WebCore::IDBObjectStore::openCursor):
- Modules/mediastream/MediaStreamTrack.cpp:
(WebCore::MediaStreamTrack::kind):
- css/CSSPropertySourceData.cpp:
(WebCore::CSSPropertySourceData::toString):
- dom/Document.cpp:
(WebCore::Document::processHttpEquiv):
- dom/MicroDataItemList.cpp:
(WebCore::MicroDataItemList::undefinedItemType):
- dom/ScriptElement.cpp:
(WebCore::ScriptElement::notifyFinished):
- editing/MarkupAccumulator.cpp:
(WebCore::MarkupAccumulator::shouldAddNamespaceElement):
- html/FormController.cpp:
(WebCore::formStateSignature):
- html/HTMLTrackElement.cpp:
(WebCore::HTMLTrackElement::canLoadUrl):
- html/ImageInputType.cpp:
(WebCore::ImageInputType::appendFormData):
- html/canvas/CanvasRenderingContext2D.cpp:
(WebCore::CanvasRenderingContext2D::getImageData):
- html/parser/XSSAuditor.cpp:
(WebCore::XSSAuditor::init):
(WebCore::XSSAuditor::filterToken):
(WebCore::XSSAuditor::eraseDangerousAttributesIfInjected):
- inspector/InspectorCSSAgent.cpp:
(WebCore::computePseudoClassMask):
- inspector/InspectorDebuggerAgent.cpp:
(WebCore::InspectorDebuggerAgent::sourceMapURLForScript):
- inspector/InspectorIndexedDBAgent.cpp:
(WebCore):
- inspector/InspectorOverlay.cpp:
- inspector/InspectorStyleSheet.cpp:
(WebCore::InspectorStyle::setPropertyText):
(WebCore::InspectorStyle::newLineAndWhitespaceDelimiters):
- loader/ImageLoader.cpp:
(WebCore::ImageLoader::notifyFinished):
- loader/MainResourceLoader.cpp:
(WebCore::MainResourceLoader::didReceiveResponse):
- loader/TextTrackLoader.cpp:
(WebCore::TextTrackLoader::corsPolicyPreventedLoad):
- loader/icon/IconDatabase.cpp:
(WebCore::IconDatabase::defaultDatabaseFilename):
- page/ContentSecurityPolicy.cpp:
(WebCore::CSPDirectiveList::allowJavaScriptURLs):
(WebCore::CSPDirectiveList::allowInlineEventHandlers):
(WebCore::CSPDirectiveList::allowInlineScript):
(WebCore::CSPDirectiveList::allowInlineStyle):
(WebCore::CSPDirectiveList::allowEval):
(WebCore::CSPDirectiveList::allowScriptNonce):
(WebCore::CSPDirectiveList::allowScriptFromSource):
(WebCore::CSPDirectiveList::allowObjectFromSource):
(WebCore::CSPDirectiveList::allowChildFrameFromSource):
(WebCore::CSPDirectiveList::allowImageFromSource):
(WebCore::CSPDirectiveList::allowStyleFromSource):
(WebCore::CSPDirectiveList::allowFontFromSource):
(WebCore::CSPDirectiveList::allowMediaFromSource):
(WebCore::CSPDirectiveList::allowConnectToSource):
(WebCore::CSPDirectiveList::allowFormAction):
(WebCore::CSPDirectiveList::addDirective):
- page/Page.cpp:
(WebCore::Page::groupName):
- platform/animation/Animation.cpp:
(WebCore::Animation::initialAnimationName):
- platform/blackberry/CookieDatabaseBackingStore/CookieDatabaseBackingStore.cpp:
(WebCore::CookieDatabaseBackingStore::upgradeTableIfNeeded):
- platform/blackberry/RenderThemeBlackBerry.cpp:
(WebCore::RenderThemeBlackBerry::defaultGUIFont):
- platform/chromium/ClipboardUtilitiesChromium.cpp:
(WebCore::replaceNewlinesWithWindowsStyleNewlines):
- platform/efl/RenderThemeEfl.cpp:
(WebCore::RenderThemeEfl::systemFont):
- platform/graphics/filters/CustomFilterCompiledProgram.cpp:
(WebCore):
- platform/win/ClipboardUtilitiesWin.cpp:
(WebCore::replaceNewlinesWithWindowsStyleNewlines):
- rendering/RenderThemeChromiumSkia.cpp:
(WebCore::RenderThemeChromiumSkia::defaultGUIFont):
- svg/SVGAngle.cpp:
(WebCore::SVGAngle::valueAsString):
- svg/SVGTransform.cpp:
(WebCore::SVGTransform::transformTypePrefixForParsing):
- xml/parser/XMLTokenizer.cpp:
(WebCore::XMLTokenizer::nextToken):
- xml/parser/XMLTreeBuilder.cpp:
(WebCore::XMLTreeBuilder::processXMLEntity):
- 12:32 AM Changeset in webkit [126967] by
-
- 22 edits in trunk/Source/WebCore
Deploy ASCIILiteral and StringBuilder in more places in WebCore
https://bugs.webkit.org/show_bug.cgi?id=95291
Reviewed by Benjamin Poulain.
I wanted to deploy ASCIILiteral in more places in WebCore, but there's
a bunch of code that should be using StringBuilder, which I couldn't
resist deploying as well.
- Modules/mediastream/PeerConnection00.cpp:
(WebCore::PeerConnection00::createIceOptions):
- bindings/js/JSBlobCustom.cpp:
(WebCore::JSBlobConstructor::constructJSBlob):
- bindings/v8/V8LazyEventListener.cpp:
(WebCore::V8LazyEventListener::prepareListenerObject):
- bindings/v8/custom/V8BlobCustom.cpp:
(WebCore::V8Blob::constructorCallback):
- css/CSSCanvasValue.cpp:
(WebCore::CSSCanvasValue::customCssText):
- css/CSSCrossfadeValue.cpp:
(WebCore::CSSCrossfadeValue::customCssText):
- css/CSSPageRule.cpp:
(WebCore::CSSPageRule::selectorText):
- css/CSSSelector.cpp:
(WebCore::CSSSelector::selectorText):
- editing/EditorCommand.cpp:
(WebCore::executeToggleStyleInList):
- editing/SmartReplaceICU.cpp:
(WebCore::getSmartSet):
- editing/markup.cpp:
(WebCore::fillContainerFromString):
- html/HTMLTitleElement.cpp:
(WebCore::HTMLTitleElement::text):
- inspector/InspectorConsoleAgent.cpp:
(WebCore::InspectorConsoleAgent::didFailLoading):
- page/EventSource.cpp:
(WebCore::EventSource::didReceiveResponse):
- platform/blackberry/LocalizedStringsBlackBerry.cpp:
(WebCore::platformLanguage):
- platform/graphics/GraphicsLayer.cpp:
(WebCore::GraphicsLayer::animationNameForTransition):
- platform/network/blackberry/rss/RSSAtomParser.cpp:
(WebCore::RSSAtomParser::parseContent):
- platform/sql/SQLiteDatabase.cpp:
(WebCore::SQLiteDatabase::open):
(WebCore::SQLiteDatabase::setFullsync):
(WebCore::SQLiteDatabase::maximumSize):
(WebCore::SQLiteDatabase::pageSize):
(WebCore::SQLiteDatabase::freeSpaceSize):
(WebCore::SQLiteDatabase::totalSize):
(WebCore::SQLiteDatabase::tableExists):
(WebCore::SQLiteDatabase::clearAllTables):
(WebCore::SQLiteDatabase::runVacuumCommand):
(WebCore::SQLiteDatabase::runIncrementalVacuumCommand):
(WebCore::SQLiteDatabase::turnOnIncrementalAutoVacuum):
- platform/win/FileSystemWin.cpp:
(WebCore::bundleName):
(WebCore::storageDirectory):
- plugins/PluginStream.cpp:
(WebCore::PluginStream::startStream):
- rendering/RenderMenuList.cpp:
(WebCore::RenderMenuList::setTextFromOption):
- 12:12 AM Changeset in webkit [126966] by
-
- 6 edits in trunk/Source/WebCore
[Forms] Make HTMLInputElement::blur()/focus() override-able by input type
https://bugs.webkit.org/show_bug.cgi?id=95279
Reviewed by Hajime Morrita.
This patch allows HTMLInputElement::focus() and blur() functions to be
override-able.
This patch is part of Shift+Tab focus navigation change for multiple
field time input UI, bug 95168.
No new tests. This patch doesn't change behavior.
- dom/Element.h:
(WebCore::Element): Changed blur() to virtual function.
- html/HTMLInputElement.cpp:
(WebCore::HTMLInputElement::blur): Added to call InputType::blur().
(WebCore::HTMLInputElement::defaultBlur): Added for default implementation of blur().
(WebCore::HTMLInputElement::defaultFocus): Added for default implementation of focus().
(WebCore::HTMLInputElement::focus): Added to call InputType::focus()
- html/HTMLInputElement.h:
(HTMLInputElement): Added declarations of blur(), defaultBlur(), defaultFocus() and focus().
- html/InputType.cpp:
(WebCore::InputType::blur): Added to call HTMLInputElement::defaultBlur() as default implementation.
(WebCore::InputType::focus): Added to call HTMLInputElement::defaultFocus() as default implementation.
- html/InputType.h:
(InputType): Added declarations of blur(), and focus().
- 12:03 AM Changeset in webkit [126965] by
-
- 5 edits in trunk/Source/WebCore
Unreviewed, rolling out r126963.
http://trac.webkit.org/changeset/126963
https://bugs.webkit.org/show_bug.cgi?id=95298
Does not compile with clang (Requested by abarth on #webkit).
Patch by Sheriff Bot <webkit.review.bot@gmail.com> on 2012-08-29
- html/HTMLInputElement.cpp:
- html/HTMLInputElement.h:
(HTMLInputElement):
- html/InputType.cpp:
- html/InputType.h:
(InputType):
