Timeline
Mar 9, 2015:
- 11:56 PM Changeset in webkit [181315] by
-
- 6 edits1 add in trunk/Source/JavaScriptCore
Yet another build fix for Windows.
https://webkit.org/b/142513>
Reviewed by Alex Christensen.
Looks like MSVC requires the function be explicitly declared in a header file
in order for it to be linkable from another file in the same project. This is
strange, but it seems to make MSVC happy.
Also fixed a typo in testapi.vcxproj.filters.
- API/tests/CompareAndSwapTest.cpp:
- API/tests/CompareAndSwapTest.h: Added.
- API/tests/testapi.c:
- JavaScriptCore.vcxproj/testapi/testapi.vcxproj:
- JavaScriptCore.vcxproj/testapi/testapi.vcxproj.filters:
- JavaScriptCore.xcodeproj/project.pbxproj:
- 11:54 PM Changeset in webkit [181314] by
-
- 2 edits in trunk/LayoutTests
Skip contentfiltering/block-after-add-data.html for now.
https://bugs.webkit.org/show_bug.cgi?id=142526
- 11:50 PM Changeset in webkit [181313] by
-
- 8 edits in trunk/Source
Unreviewed, rolling out r181307.
https://bugs.webkit.org/show_bug.cgi?id=142525
Broke ASan tests (Requested by ap on #webkit).
Reverted changeset:
"bmalloc: tryFastMalloc shouldn't crash"
https://bugs.webkit.org/show_bug.cgi?id=142443
http://trac.webkit.org/changeset/181307
- 11:35 PM Changeset in webkit [181312] by
-
- 5 edits in branches/safari-600.1.4.15-branch/Source
Versioning.
- 11:34 PM Changeset in webkit [181311] by
-
- 5 edits in branches/safari-600.5-branch/Source
Versioning.
- 10:56 PM March 2015 Meeting edited by
- (diff)
- 10:32 PM Changeset in webkit [181310] by
-
- 5 edits in trunk/Source
[iOS] Sweep all collected objects on critical memory pressure
https://bugs.webkit.org/show_bug.cgi?id=142457
<rdar://problem/20044440>
Reviewed by Geoffrey Garen.
Source/JavaScriptCore:
All fullSweep() API to IncrementalSweeper so that we can call it in the
memory pressure handler.
- heap/IncrementalSweeper.cpp:
(JSC::IncrementalSweeper::fullSweep):
- heap/IncrementalSweeper.h:
(JSC::IncrementalSweeper::hasWork):
Source/WebCore:
Do a full sweep of objects marked for destruction on critical memory
pressure to free up memory.
- platform/cocoa/MemoryPressureHandlerCocoa.mm:
(WebCore::MemoryPressureHandler::platformReleaseMemory):
- 10:24 PM March 2015 Meeting edited by
- (diff)
- 9:56 PM March 2015 Meeting edited by
- (diff)
- 9:41 PM Changeset in webkit [181309] by
-
- 2 edits in trunk/Source/JavaScriptCore
Another build fix for Windows.
https://webkit.org/b/142513>
Not reviewed.
- API/tests/CompareAndSwapTest.cpp:
- Added JS_EXPORT_PRIVATE attribute.
- 9:21 PM Changeset in webkit [181308] by
-
- 2 edits in trunk/Source/JavaScriptCore
Build fix for Windows after r181305.
https://webkit.org/b/142513>
Reviewed by Alex Christensen.
Windows doesn't like pthreads anymore. Changed test to use WTF threading.
- API/tests/CompareAndSwapTest.cpp:
(setBitThreadFunc):
(testCompareAndSwap):
- 9:10 PM Changeset in webkit [181307] by
-
- 8 edits in trunk/Source
bmalloc: tryFastMalloc shouldn't crash
https://bugs.webkit.org/show_bug.cgi?id=142443
Reviewed by Darin Adler.
Source/bmalloc:
Added support for tryMalloc.
We assume that non-x-large allocations always succeed, and we crash
otherwise, since normal allocation failure will just cause the next
non-try allocation or internal metadata allocation to fail, and it's
hard and not really useful to keep limping along after that. But
extra-large allocations can meaningfully fail, and we can recover.
- bmalloc/Heap.cpp:
(bmalloc::Heap::allocateXLarge):
(bmalloc::Heap::tryAllocateXLarge):
- bmalloc/Heap.h: Added support for non-crashy x-large allocation.
- bmalloc/VMAllocate.h:
(bmalloc::tryVMAllocate):
(bmalloc::vmAllocate): Added support for non-crashy VM allocation.
- bmalloc/bmalloc.h:
(bmalloc::api::tryMalloc):
(bmalloc::api::realloc):
(bmalloc::api::free): Tried to clarify our behavior with some comments.
Unfortunately, calling what we do "malloc" is still not quite right, since
malloc returns null on failure and we don't.
Source/WTF:
- wtf/FastMalloc.cpp:
(WTF::fastMalloc):
(WTF::fastRealloc):
(WTF::fastAlignedMalloc): Don't check for null. bmalloc automatically
crashes on allocation failure, and we'd rather not pay for an extra check.
(WTF::tryFastMalloc): Added an opt-out API to return null rather than
crashing, since some clients need this.
(WTF::tryFastRealloc): Deleted. Unused.
- wtf/FastMalloc.h:
- 8:25 PM March 2015 Meeting edited by
- (diff)
- 8:01 PM Changeset in webkit [181306] by
-
- 9 edits1 copy in trunk/Source/WebInspectorUI
Web Inspector: Inline Error / Warning message for Issues
https://bugs.webkit.org/show_bug.cgi?id=142520
Reviewed by Timothy Hatcher.
- Localizations/en.lproj/localizedStrings.js:
- UserInterface/Main.html:
New strings and new files.
- UserInterface/Controllers/IssueManager.js:
(WebInspector.IssueManager.prototype.issueWasAdded):
- UserInterface/Models/IssueMessage.js:
(WebInspector.IssueMessage):
(WebInspector.IssueMessage.prototype.get columnNumber):
- UserInterface/Protocol/ConsoleObserver.js:
(WebInspector.ConsoleObserver.prototype.messageAdded):
Correctly pass the column number into IssueMessage.
- UserInterface/Models/LineWidget.js:
(WebInspector.LineWidget):
(WebInspector.LineWidget.prototype.get codeMirrorLineWidget):
(WebInspector.LineWidget.prototype.get widgetElement):
Create a new Model class for a LineWidget. The root element
should never change, but its children can be updated.
- UserInterface/Views/SourceCodeTextEditor.css:
(.source-code.text-editor .CodeMirror-linewidget):
Override styles so that widgets can overlap line content.
(.source-code.text-editor > .CodeMirror .issue-widget):
(.source-code.text-editor > .CodeMirror .issue-widget.inline):
Float issue widgets to the right side of the editor.
(.source-code.text-editor > .CodeMirror .issue-widget > .arrow):
(.source-code.text-editor > .CodeMirror .issue-widget.inline > .arrow):
Pure CSS arrow for widgets on the same line as their issue.
(.source-code.text-editor > .CodeMirror .issue-widget > .icon):
(.source-code.text-editor > .CodeMirror .issue-widget > .icon.icon-warning):
(.source-code.text-editor > .CodeMirror .issue-widget > .icon.icon-error):
Issue icon styles for the different issue types.
(.source-code.text-editor > .CodeMirror .issue-widget.warning):
(.source-code.text-editor > .CodeMirror .issue-widget.inline.warning):
(.source-code.text-editor > .CodeMirror .issue-widget.inline.warning > .arrow):
(.source-code.text-editor > .CodeMirror .issue-widget.error):
(.source-code.text-editor > .CodeMirror .issue-widget.inline.error):
(.source-code.text-editor > .CodeMirror .issue-widget.inline.error > .arrow):
Different colors for the different issue types.
(.source-code.text-editor > .CodeMirror .issue-widget > .text):
(.source-code.text-editor > .CodeMirror .issue-widget.inline > .text):
Text positioning when on the same line or when expanded.
- UserInterface/Views/SourceCodeTextEditor.js:
(WebInspector.SourceCodeTextEditor.prototype._addIssue):
(WebInspector.SourceCodeTextEditor.prototype._iconClassNameForIssueLevel):
(WebInspector.SourceCodeTextEditor.prototype._updateIssueWidgetForIssues):
(WebInspector.SourceCodeTextEditor.prototype._isWidgetToggleable):
(WebInspector.SourceCodeTextEditor.prototype._handleWidgetClick):
Create widgets for issues.
(WebInspector.SourceCodeTextEditor.prototype._contentDidPopulate):
(WebInspector.SourceCodeTextEditor.prototype.textEditorUpdatedFormatting):
(WebInspector.SourceCodeTextEditor.prototype._clearWidgets):
(WebInspector.SourceCodeTextEditor.prototype._reinsertAllIssues):
Update all widgets in certain cases.
- UserInterface/Views/TextEditor.js:
(WebInspector.TextEditor.prototype.addStyleClassToLine):
This assertion is known to happen for issues added to an editor
before the content has loaded.
(WebInspector.TextEditor.prototype.createWidgetForLine):
Create a widget for a line.
- 7:18 PM Changeset in webkit [181305] by
-
- 7 edits1 add in trunk/Source
8-bit version of weakCompareAndSwap() can cause an infinite loop.
https://webkit.org/b/142513>
Reviewed by Filip Pizlo.
Source/JavaScriptCore:
Added a test that exercises the 8-bit CAS from multiple threads. The threads
will contend to set bits in a large array of bytes using the CAS function.
- API/tests/CompareAndSwapTest.cpp: Added.
(Bitmap::Bitmap):
(Bitmap::numBits):
(Bitmap::clearAll):
(Bitmap::concurrentTestAndSet):
(setBitThreadFunc):
(testCompareAndSwap):
- API/tests/testapi.c:
(main):
- JavaScriptCore.vcxproj/testapi/testapi.vcxproj:
- JavaScriptCore.vcxproj/testapi/testapi.vcxproj.filters:
- JavaScriptCore.xcodeproj/project.pbxproj:
Source/WTF:
Presently, Bitmap::concurrentTestAndSet() uses the 8-bit version of
weakCompareAndSwap() (which compares and swaps an uint8_t value).
Bitmap::concurrentTestAndSet() has a loop that checks if a bit in the
byte of interest has been set. If not, it will call the 8-bit CAS
function to set the bit.
Under the covers, for ARM, the 8-bit CAS function actually works with a
32-bit CAS. The 8-bit CAS will first fetch the 32-bit value in memory
that should contain the 8-bit value, and check if it contains the
expected byte. If the value in memory doesn't have the expected byte,
it will return early to its caller. The expectation is that the caller
will reload the byte from memory and call the 8-bit CAS again.
Unfortunately, this code path that returns early does not have a
compiler fence. Without a compiler fence, the C++ compiler can
optimize away the reloading of the expected byte value, leaving it
unchanged. As a result, we'll have a infinite loop here that checks a
value that will never change, and the loop will not terminate until the
value changes.
The fix is to eliminate the early return check in the 8-bit CAS, and
have it always call down to the 32-bit CAS. The 32-bit CAS has a
compiler fence which will prevent this issue.
- wtf/Atomics.h:
(WTF::weakCompareAndSwap):
- 7:00 PM Changeset in webkit [181304] by
-
- 2 edits in trunk/Source/JavaScriptCore
[Win] testapi project is unable to find the 'config.h' file.
Rubberstamped by Mark Lam.
- JavaScriptCore.vcxproj/testapi/testapiCommon.props: Add JavaScriptCore source directory
to the include path.
- 6:51 PM Changeset in webkit [181303] by
-
- 4 edits in trunk/LayoutTests
[Win] Rebaseline tests afer r181292.
- platform/win/TestExpectations:
- platform/win/fast/forms/search-vertical-alignment-expected.txt:
- platform/win/fast/forms/textfield-overflow-by-value-update-expected.txt:
- 6:38 PM Changeset in webkit [181302] by
-
- 4 edits in trunk/Source/WebCore
[Content Filtering] Give ContentFilter subclasses private constructors, and make them fast-allocated and non-copyable
https://bugs.webkit.org/show_bug.cgi?id=142515
Reviewed by Andreas Kling.
- platform/cocoa/NetworkExtensionContentFilter.h:
- platform/cocoa/ParentalControlsContentFilter.h:
- testing/MockContentFilter.h:
- 6:24 PM Changeset in webkit [181301] by
-
- 9 edits in trunk
[CG] Have Canvas use the IOSurfacePool
https://bugs.webkit.org/show_bug.cgi?id=142417
<rdar://problem/20044440>
Reviewed by Darin Adler.
PerformanceTests:
Lower the number of different canvas sizes from 1000 to 100 so that
the test does not require such a huge cache size. With 100, we now
get over 90% cache hit rate with the default IOSurfacePool size.
- Canvas/reuse.html:
Source/WebCore:
Have ImageBufferDataCG use the IOSurfacePool so that Canvas can
benefit from it. I see a ~75% progression on Canvas/reuse.html
performance test on my Macbook Pro with 1000 different canvas
sizes and ~110% progression with 100 different canvas sizes.
I also see a ~65% cache hit rate on the mobile version of
cnn.com.
Note that ImageData calls CGContextClearRect() after calling
IOSurface::create() so recycling IOSurfaces in this case should
be safe.
Performance test: Canvas/reuse.html
- platform/graphics/ImageBuffer.h:
(WebCore::ImageBuffer::baseTransform):
- platform/graphics/cg/ImageBufferCG.cpp:
(WebCore::ImageBuffer::ImageBuffer):
(WebCore::ImageBuffer::context):
(WebCore::ImageBuffer::copyImage):
(WebCore::ImageBuffer::copyNativeImage):
(WebCore::ImageBuffer::draw):
(WebCore::ImageBuffer::clip):
(WebCore::ImageBuffer::putByteArray):
(WebCore::ImageBuffer::toDataURL):
- platform/graphics/cg/ImageBufferDataCG.cpp:
(WebCore::ImageBufferData::~ImageBufferData):
(WebCore::ImageBufferData::getData):
(WebCore::ImageBufferData::putData):
(WebCore::ImageBufferData::ImageBufferData): Deleted.
- platform/graphics/cg/ImageBufferDataCG.h:
- platform/graphics/cocoa/IOSurface.h:
- platform/graphics/cocoa/IOSurface.mm:
(IOSurface::surfaceFromPool):
(IOSurface::create):
(IOSurface::createFromSendRight):
(IOSurface::createFromImage):
(IOSurface::setContextSize):
- 6:00 PM Changeset in webkit [181300] by
-
- 2 edits in trunk/Source/WebCore
Assertion in ScrollController::processWheelEventForScrollSnapOnAxis when scrolling with mechanical wheel mouse
https://bugs.webkit.org/show_bug.cgi?id=142517
Reviewed by Simon Fraser.
- platform/cocoa/ScrollController.mm:
(WebCore::toWheelEventStatus): Recognize stateless mechanical wheel events.
(WebCore::ScrollController::processWheelEventForScrollSnapOnAxis): If a mechanical wheel event is processed while
considering scroll snap logic, end any snap animations and clean up.
(WebCore::ScrollController::updateScrollSnapPoints): Don't begin tracking scroll snap state if there is not
scroll snap point defined in the markup.
- 5:34 PM Changeset in webkit [181299] by
-
- 1 copy in tags/Safari-600.1.4.15.12
New tag.
- 5:32 PM Changeset in webkit [181298] by
-
- 1 copy in tags/Safari-600.5.10
New tag.
- 5:09 PM Changeset in webkit [181297] by
-
- 17 edits1 copy1 add in trunk/Source
Stale entries in WeakGCMaps are keeping tons of WeakBlocks alive unnecessarily.
<https://webkit.org/b/142115>
<rdar://problem/19992268>
Reviewed by Geoffrey Garen.
Prune stale entries from WeakGCMaps as part of every full garbage collection.
This frees up tons of previously-stuck WeakBlocks that were only sitting around
with finalized handles waiting to die.
Note that WeakGCMaps register/unregister themselves with the GC heap in their
ctor/dtor, so creating one now requires passing the VM.
Average time spent in the PruningStaleEntriesFromWeakGCMaps GC phase appears
to be between 0.01ms and 0.3ms, though I've seen a few longer ones at ~1.2ms.
It seems somewhat excessive to do this on every Eden collection, so it's only
doing work in full collections for now.
Because the GC may now mutate WeakGCMap below object allocation, I've made it
so that the classic HashMap::add() optimization can't be used with WeakGCMap.
This caused intermittent test failures when originally landed due to having
an invalid iterator on the stack after add() inserted a new entry and we
proceeded to allocate the new object, triggering GC.
- API/JSWeakObjectMapRefInternal.h:
(OpaqueJSWeakObjectMap::create):
(OpaqueJSWeakObjectMap::OpaqueJSWeakObjectMap):
- API/JSWeakObjectMapRefPrivate.cpp:
- API/JSWrapperMap.mm:
(-[JSWrapperMap initWithContext:]):
(-[JSWrapperMap jsWrapperForObject:]): Pass VM to WeakGCMap constructor.
- JavaScriptCore.xcodeproj/project.pbxproj: Add WeakGCMapInlines.h and make
it project-private so WebCore clients can access it.
- heap/Heap.cpp:
(JSC::Heap::collect):
(JSC::Heap::pruneStaleEntriesFromWeakGCMaps): Added a new GC phase for pruning
stale entries from WeakGCMaps. This is only executed during full collections.
- heap/Heap.h:
- heap/HeapInlines.h:
(JSC::Heap::registerWeakGCMap):
(JSC::Heap::unregisterWeakGCMap): Added a mechanism for WeakGCMaps to register
themselves with the Heap and provide a pruning callback.
- runtime/PrototypeMap.h:
(JSC::PrototypeMap::PrototypeMap):
- runtime/Structure.cpp:
(JSC::StructureTransitionTable::add): Pass VM to WeakGCMap constructor.
- runtime/JSCInlines.h: Add "WeakGCMapInlines.h"
- runtime/JSGlobalObject.cpp: Include "WeakGCMapInlines.h" so this builds.
- runtime/JSString.cpp:
(JSC::jsStringWithCacheSlowCase):
- runtime/PrototypeMap.cpp:
(JSC::PrototypeMap::addPrototype):
(JSC::PrototypeMap::emptyObjectStructureForPrototype): Remove HashMap add()
optimization since it's not safe in the GC-managed WeakGCMap world.
- runtime/VM.cpp:
(JSC::VM::VM): Pass VM to WeakGCMap constructor.
- runtime/WeakGCMap.h:
(JSC::WeakGCMap::set):
(JSC::WeakGCMap::add):
(JSC::WeakGCMap::WeakGCMap): Deleted.
(JSC::WeakGCMap::gcMap): Deleted.
(JSC::WeakGCMap::gcMapIfNeeded): Deleted.
- runtime/WeakGCMapInlines.h: Added.
(JSC::WeakGCMap::WeakGCMap):
(JSC::WeakGCMap::~WeakGCMap):
(JSC::WeakGCMap::pruneStaleEntries): Moved ctor, dtor and pruning callback
to WeakGCMapInlines.h to fix interdependent header issues. Removed code that
prunes WeakGCMap at certain growth milestones and instead rely on the GC
callback for housekeeping.
- 4:57 PM Changeset in webkit [181296] by
-
- 6 edits in tags/Safari-601.1.21.1/Source
Merged patch for <rdar://problem/20080527>.
- 4:52 PM Changeset in webkit [181295] by
-
- 5 edits in tags/Safari-601.1.21.1/Source
Versioning.
- 4:49 PM Changeset in webkit [181294] by
-
- 1 copy in tags/Safari-601.1.21.1
New tag.
- 4:47 PM Changeset in webkit [181293] by
-
- 18 edits6 adds in trunk
Support extends and super keywords
https://bugs.webkit.org/show_bug.cgi?id=142200
Reviewed by Filip Pizlo.
Source/JavaScriptCore:
Added the support for ES6 class syntax inheritance.
Added ConstructorKind as well as boolean flags indicating the constructor kind to
various classes in UnlinkedCodeBlock as well as AST nodes.
Each method stores the associated class as its homeObjectPrivateName. This value is used to
make super calls.
- bytecode/UnlinkedCodeBlock.cpp:
(JSC::generateFunctionCodeBlock):
(JSC::UnlinkedFunctionExecutable::UnlinkedFunctionExecutable):
(JSC::UnlinkedCodeBlock::UnlinkedCodeBlock):
- bytecode/UnlinkedCodeBlock.h:
(JSC::ExecutableInfo::ExecutableInfo):
(JSC::UnlinkedFunctionExecutable::constructorKindIsDerived): Added.
(JSC::UnlinkedCodeBlock::constructorKindIsDerived): Added.
- bytecompiler/BytecodeGenerator.cpp:
(JSC::BytecodeGenerator::BytecodeGenerator): Don't emit op_create_this in a derived class
as the object is allocated by the highest base class's constructor. Also set "this" to null
and store the original value in m_newTargetRegister. "this" is supposed to be in TDZ but
that will be implemented in a separate patch.
(JSC::BytecodeGenerator::emitReturn): Allow "undefined" to be returned from a derived class.
In a derived class's constructor, not returning "undefined" or an object results in a type
error instead of "this" being returned.
(JSC::BytecodeGenerator::emitThrowTypeError): Added.
- bytecompiler/BytecodeGenerator.h:
(JSC::BytecodeGenerator::constructorKindIsDerived): Added.
(JSC::BytecodeGenerator::newTarget): Added.
- bytecompiler/NodesCodegen.cpp:
(JSC::SuperNode::emitBytecode): Added. Emits the code to obtain the callee's parent class.
(JSC::emitSuperBaseForCallee): Added. Emits the code to obtain the parent class's prototype.
(JSC::emitPutHomeObject): Added.
(JSC::PropertyListNode::emitBytecode): Stores the home object when adding methods.
(JSC::PropertyListNode::emitPutConstantProperty): Ditto.
(JSC::BracketAccessorNode::emitBytecode): Added the support for superfoo.
(JSC::DotAccessorNode::emitBytecode): Added the support for super.foo.
(JSC::FunctionCallValueNode::emitBytecode): Added the support for super().
(JSC::FunctionCallBracketNode::emitBytecode): Added the support for superfoo().
(JSC::FunctionCallDotNode::emitBytecode): Added the support for super.foo().
(JSC::DeleteBracketNode::emitBytecode): Forbid "delete super.foo".
(JSC::DeleteDotNode::emitBytecode): Forbid "delete superfoo".
(JSC::ClassExprNode::emitBytecode): Added the support for "classHeritage". This is the main
logic for inheritance. When a class B inherits from a class A, set B.proto to A and set
B.prototype.proto to A.prototype. Throw exceptions when either A or A.proto is not
an object.
- parser/ASTBuilder.h:
(JSC::ASTBuilder::superExpr): Added.
- parser/NodeConstructors.h:
(JSC::SuperNode::SuperNode): Added.
- parser/Nodes.cpp:
(JSC::FunctionBodyNode::FunctionBodyNode):
- parser/Nodes.h:
(JSC::ExpressionNode::isSuperNode):
(JSC::PropertyNode::type):
(JSC::PropertyNode::needsSuperBinding):
- parser/Parser.cpp:
(JSC::Parser<LexerType>::parseFunctionBody):
(JSC::Parser<LexerType>::parseFunctionInfo): Throw a parser error if super() is used outside
of class constructors.
(JSC::Parser<LexerType>::parseFunctionDeclaration):
(JSC::Parser<LexerType>::parseClass): ConstructorKind is "derived" if and only if the parent
class is specified in the declaration / expression.
(JSC::Parser<LexerType>::parseGetterSetter):
(JSC::Parser<LexerType>::parsePrimaryExpression):
(JSC::Parser<LexerType>::parseMemberExpression): Added the support for "super()", "super.foo",
and "superfoo". Throw a semantic error if "super" appears by itself.
- parser/Parser.h:
(JSC::Scope::Scope): Added m_hasDirectSuper. This variable keeps track of the use of "super()"
so that parseFunctionInfo can spit an error if it's used outside of class constructors.
(JSC::Scope::hasDirectSuper): Added.
(JSC::Scope::setHasDirectSuper): Added.
- parser/ParserModes.h:
(JSC::ConstructorKind): Added.
- parser/SyntaxChecker.h:
(JSC::SyntaxChecker::superExpr): Added.
- runtime/CommonIdentifiers.h: Added homeObjectPrivateName.
- runtime/Executable.h:
(JSC::EvalExecutable::executableInfo):
(JSC::ProgramExecutable::executableInfo):
LayoutTests:
Added tests for "extends" and "super" keywords.
- TestExpectations:
- js/class-syntax-extends-expected.txt: Added.
- js/class-syntax-extends.html: Added.
- js/class-syntax-super-expected.txt: Added.
- js/class-syntax-super.html: Added.
- js/script-tests/class-syntax-extends.js: Added.
- js/script-tests/class-syntax-super.js: Added.
- 4:43 PM Changeset in webkit [181292] by
-
- 8 edits4 adds in trunk
REGRESSION(r176978): Inline-blocks with overflowing contents have ascents that are too large
https://bugs.webkit.org/show_bug.cgi?id=141783
Reviewed by David Hyatt.
Source/WebCore:
When we have an inline-block element, and we want to find its baseline (to lay out other
elements on the same line) we loop through the element's children and ask them what their
baselines are. The children use the location of the top of their last line to compute this
value. However, if the child has overflow-y, this might not be the correct calculation.
This behavior is in the spec: "The baseline of an 'inline-block' is the baseline of its last
line box in the normal flow, unless it has either no in-flow line boxes or if its 'overflow'
property has a computed value other than 'visible', in which case the baseline is the bottom
margin edge."
However, we believe that a better policy is, when overflow is not "visible," to place the
baseline at the bottom of the block if the contents overflowed in the Y direction, and to place
it at the bottom of the last line if the contents did not overflow in the Y direction. This is
partially consistent with previous behavior, and isn't too far from the spec to cause too many
breakages.
Test: fast/css/inline-block-tricky-baselines.html
fast/text/baseline-inline-block.html
- rendering/RenderBlockFlow.cpp:
(WebCore::RenderBlockFlow::inlineBlockBaseline):
LayoutTests:
Update expected results.
- css3/flexbox/child-overflow-expected.html:
- css3/flexbox/child-overflow.html:
- fast/css/inline-block-tricky-baselines-expected.html: Added.
- fast/css/inline-block-tricky-baselines.html: Added.
- fast/forms/textfield-overflow-by-value-update-expected.txt:
- fast/text/baseline-inline-block-expected.html: Added.
- fast/text/baseline-inline-block.html: Added.
- platform/mac/fast/forms/search-vertical-alignment-expected.txt:
- 4:39 PM Changeset in webkit [181291] by
-
- 13 edits3 copies16 adds in trunk
[Content Filtering] Add tests
https://bugs.webkit.org/show_bug.cgi?id=142475
Reviewed by Andreas Kling.
Source/WebCore:
Added a mock content filter that can be configured by window.internals. A test can control whether the filter
is enabled, what to display when a page is blocked, at what point the allow/block decision is made, and what
decision to make.
Added tests to cover allowing and blocking content at each decision point.
Tests: contentfiltering/allow-after-add-data.html
contentfiltering/allow-after-finished-adding-data.html
contentfiltering/allow-after-response.html
contentfiltering/block-after-add-data.html
contentfiltering/block-after-finished-adding-data.html
contentfiltering/block-after-response.html
- DerivedSources.make: Added MockContentFilterSettings.idl.
- WebCore.xcodeproj/project.pbxproj:
- bindings/js/JSMockContentFilterSettingsCustom.cpp: Added.
(WebCore::JSMockContentFilterSettings::decisionPoint): Returned the decision point's JSValue.
(WebCore::JSMockContentFilterSettings::setDecisionPoint): Set the decision point; threw an exception for invalid values.
(WebCore::JSMockContentFilterSettings::decision): Returned the decision's JSValue.
(WebCore::JSMockContentFilterSettings::setDecision): Set the decision; threw an exception for invalid values.
- platform/ContentFilter.cpp:
(WebCore::ContentFilter::types): Renamed from contentFilterTypes.
(WebCore::ContentFilter::createIfNeeded):
(WebCore::contentFilterType): Deleted.
(WebCore::contentFilterTypes): Deleted.
- platform/ContentFilter.h:
(WebCore::ContentFilter::addType): Used to register a new content filter type.
(WebCore::ContentFilter::type): Renamed from contentFilterType.
- platform/cocoa/NetworkExtensionContentFilter.h: Moved definition of HAVE_NE_FILTER_SOURCE to here.
- testing/Internals.cpp:
(WebCore::Internals::resetToConsistentState): Called MockContentFilterSettings::reset().
(WebCore::Internals::Internals): Called MockContentFilter::ensureInstalled().
(WebCore::Internals::mockContentFilterSettings): Returned a reference to MockContentFilterSettings::singleton().
- testing/Internals.h:
- testing/Internals.idl: Added attribute mockContentFilterSettings.
- testing/MockContentFilter.cpp: Added.
(WebCore::MockContentFilter::ensureInstalled): Called MockContentFilter::addType().
(WebCore::MockContentFilter::canHandleResponse): Returned whether MockContentFilterSettings is enabled.
(WebCore::MockContentFilter::create): Returned a new MockContentFilter.
(WebCore::MockContentFilter::MockContentFilter): Called maybeDetermineStatus().
(WebCore::MockContentFilter::addData): Copied data into m_replacementData and called maybeDetermineStatus().
(WebCore::MockContentFilter::finishedAddingData): Called maybeDetermineStatus().
(WebCore::MockContentFilter::needsMoreData): Returned whether m_status equals NeedsMoreData.
(WebCore::MockContentFilter::didBlockData): Returned whether m_status equals Blocked.
(WebCore::MockContentFilter::getReplacementData): Returned a pointer to the start of m_replacementData.
(WebCore::MockContentFilter::unblockHandler): Returned an empty ContentFilterUnblockHandler.
(WebCore::MockContentFilter::maybeDetermineStatus): If m_status is NeedsMoreData and the current decision point
equals MockContentFilterSettings's decision point, set m_status according to MockContentFilterSettings's
decision. If m_status becomes Blocked, clear the original page data from m_replacementData and copy
MockContentFilterSettings's blocked string instead.
- testing/MockContentFilter.h: Copied from Source/WebCore/platform/cocoa/NetworkExtensionContentFilter.h.
- testing/MockContentFilterSettings.cpp: Copied from Source/WebCore/platform/ContentFilter.h.
(WebCore::MockContentFilterSettings::singleton):
(WebCore::MockContentFilterSettings::reset): Reset settings to their initial values.
(WebCore::MockContentFilterSettings::MockContentFilterSettings):
- testing/MockContentFilterSettings.h: Copied from Source/WebCore/platform/cocoa/NetworkExtensionContentFilter.h.
(WebCore::MockContentFilterSettings::ref): Added an empty implementation to make the bindings think this is RefCounted.
(WebCore::MockContentFilterSettings::deref): Ditto.
(WebCore::MockContentFilterSettings::enabled):
(WebCore::MockContentFilterSettings::setEnabled):
(WebCore::MockContentFilterSettings::blockedString):
(WebCore::MockContentFilterSettings::setBlockedString):
(WebCore::MockContentFilterSettings::decisionPoint):
(WebCore::MockContentFilterSettings::setDecisionPoint):
(WebCore::MockContentFilterSettings::decision):
(WebCore::MockContentFilterSettings::setDecision):
- testing/MockContentFilterSettings.idl: Copied from Source/WebCore/platform/ContentFilter.h.
LayoutTests:
- contentfiltering/allow-after-add-data-expected.html: Added.
- contentfiltering/allow-after-add-data.html: Added.
- contentfiltering/allow-after-finished-adding-data-expected.html: Added.
- contentfiltering/allow-after-finished-adding-data.html: Added.
- contentfiltering/allow-after-response-expected.html: Added.
- contentfiltering/allow-after-response.html: Added.
- contentfiltering/block-after-add-data-expected.html: Added.
- contentfiltering/block-after-add-data.html: Added.
- contentfiltering/block-after-finished-adding-data-expected.html: Added.
- contentfiltering/block-after-finished-adding-data.html: Added.
- contentfiltering/block-after-response-expected.html: Added.
- contentfiltering/block-after-response.html: Added.
- 4:30 PM Changeset in webkit [181290] by
-
- 9 edits in trunk/Source/WebCore
Simple line layout should not be limited to RenderText.
https://bugs.webkit.org/show_bug.cgi?id=142504
Reviewed by Antti Koivisto.
This is in transition to support br element.
No change in functionality.
- rendering/RenderText.cpp:
(WebCore::RenderText::absoluteRects):
(WebCore::RenderText::absoluteQuadsClippedToEllipsis):
(WebCore::RenderText::absoluteQuads):
(WebCore::RenderText::firstRunLocation):
(WebCore::RenderText::linesBoundingBox):
(WebCore::RenderText::caretMinOffset):
(WebCore::RenderText::caretMaxOffset):
(WebCore::RenderText::containsCaretOffset):
- rendering/SimpleLineLayoutFlowContents.cpp:
(WebCore::SimpleLineLayout::FlowContents::segmentForRenderer):
- rendering/SimpleLineLayoutFlowContents.h:
- rendering/SimpleLineLayoutFunctions.cpp:
(WebCore::SimpleLineLayout::computeBoundingBox):
(WebCore::SimpleLineLayout::computeFirstRunLocation):
(WebCore::SimpleLineLayout::collectAbsoluteRects):
(WebCore::SimpleLineLayout::collectAbsoluteQuads):
(WebCore::SimpleLineLayout::computeTextBoundingBox): Deleted.
(WebCore::SimpleLineLayout::computeTextFirstRunLocation): Deleted.
(WebCore::SimpleLineLayout::collectTextAbsoluteRects): Deleted.
(WebCore::SimpleLineLayout::collectTextAbsoluteQuads): Deleted.
- rendering/SimpleLineLayoutFunctions.h:
(WebCore::SimpleLineLayout::findCaretMinimumOffset):
(WebCore::SimpleLineLayout::findCaretMaximumOffset):
(WebCore::SimpleLineLayout::containsCaretOffset):
(WebCore::SimpleLineLayout::findTextCaretMinimumOffset): Deleted.
(WebCore::SimpleLineLayout::findTextCaretMaximumOffset): Deleted.
(WebCore::SimpleLineLayout::containsTextCaretOffset): Deleted.
- rendering/SimpleLineLayoutResolver.cpp:
(WebCore::SimpleLineLayout::RunResolver::Run::text):
(WebCore::SimpleLineLayout::RunResolver::rangeForRenderer):
- rendering/SimpleLineLayoutResolver.h:
- rendering/SimpleLineLayoutTextFragmentIterator.cpp:
(WebCore::SimpleLineLayout::TextFragmentIterator::textWidth):
- 3:51 PM Changeset in webkit [181289] by
-
- 2 edits in branches/safari-600.5-branch/Source/WebCore
Merged r179958. rdar://problem/20096968
- 3:49 PM Changeset in webkit [181288] by
-
- 2 edits in branches/safari-600.5-branch/Source/WebCore
Merged r179895. rdar://problem/20096968
- 3:48 PM Changeset in webkit [181287] by
-
- 2 edits in branches/safari-600.5-branch/Source/WebCore
Merged r179880. rdar://problem/20096968
- 3:34 PM Changeset in webkit [181286] by
-
- 1 edit2 adds in trunk/WebKitLibraries
[iOS] Add WebKitSystemInterface for iOS 8.2
https://bugs.webkit.org/show_bug.cgi?id=142505
Patch by Daniel Bates <dabates@apple.com> on 2015-03-09
Reviewed by Andy Estes.
- libWebKitSystemInterfaceIOSDevice8.2.a: Added.
- libWebKitSystemInterfaceIOSSimulator8.2.a: Added.
- 3:09 PM Changeset in webkit [181285] by
-
- 5 edits in trunk
svg/custom/svg-fonts-without-missing-glyph.xhtml fails after fonts/font-fallback-prefers-pictographs.html
https://bugs.webkit.org/show_bug.cgi?id=142470
Reviewed by Anders Carlsson.
Source/WebCore:
After each test, reset all internal settings.
Covered by svg/custom/svg-fonts-without-missing-glyph.xhtml
- testing/InternalSettings.cpp:
(WebCore::InternalSettings::Backup::Backup):
(WebCore::InternalSettings::Backup::restoreTo):
- testing/InternalSettings.h:
LayoutTests:
- platform/mac/TestExpectations:
- 3:08 PM Changeset in webkit [181284] by
-
- 3 edits in trunk/Source/WebCore
Unreviewed, rolling out r181268.
https://bugs.webkit.org/show_bug.cgi?id=142508
Broke tests (Requested by ap on #webkit).
Reverted changeset:
"Simple line layout: Merge
TextFragmentIterator::findNextBreakablePosition() and
TextFragmentIterator::findNextNonWhitespacePosition()."
https://bugs.webkit.org/show_bug.cgi?id=142344
http://trac.webkit.org/changeset/181268
- 2:10 PM Changeset in webkit [181283] by
-
- 3 edits8 adds in trunk
CSS JIT: add aliases between :nth-child()/:nth-last-child() and :first-child/:last-child
https://bugs.webkit.org/show_bug.cgi?id=142472
Reviewed by Andreas Kling.
Source/WebCore:
The pseudo class :first-child has weaker tree marking than :nth-child(1).
This patch aliases :nth-child(1) to :first-child in the CSS JIT to take
advantage of that.
The strength of :last-child and :nth-last-child(1) are pretty much identical
but :last-child is a bit simpler so I changed it too. It is also easier
to handle both the same.
Tests: fast/selectors/nth-child-matching-first-on-root.html
fast/selectors/nth-child-matching-first.html
fast/selectors/nth-last-child-matching-first-on-root.html
fast/selectors/nth-last-child-matching-first.html
- cssjit/SelectorCompiler.cpp:
(WebCore::SelectorCompiler::addNthChildType):
(WebCore::SelectorCompiler::addPseudoClassType):
LayoutTests:
- fast/selectors/nth-child-matching-first-expected.txt: Added.
- fast/selectors/nth-child-matching-first-on-root-expected.txt: Added.
- fast/selectors/nth-child-matching-first-on-root.html: Added.
- fast/selectors/nth-child-matching-first.html: Added.
- fast/selectors/nth-last-child-matching-first-expected.txt: Added.
- fast/selectors/nth-last-child-matching-first-on-root-expected.txt: Added.
- fast/selectors/nth-last-child-matching-first-on-root.html: Added.
- fast/selectors/nth-last-child-matching-first.html: Added.
- 2:07 PM Changeset in webkit [181282] by
-
- 5 edits in trunk/Source/WebCore
Merge TrivialAtom and CharacterSet into a Term abstraction, prepare Term for composition
https://bugs.webkit.org/show_bug.cgi?id=142429
Patch by Benjamin Poulain <bpoulain@apple.com> on 2015-03-09
Reviewed by Darin Adler.
This patch merges CharacterSet and Trivial atom into a new class: Term. A Term is
a combination of an Atom and one Quantifier.
With term being the basic block, we can use the PrefixTree for any construct,
greatly reducing the size of the NFA graph.
Term is built on top of an union holding the Atom storage. This is done in preparation
for more complicated Atoms like a disjunction.
Everything else is pretty much the same. BuildMode is gone since we use the prefix
tree for everything. FloatingAtomType is gone, a TrivialAtom is now represented
by a single character CharacterSet (or two for case insensitive).
- contentextensions/ContentExtensionParser.cpp:
(WebCore::ContentExtensions::parseRuleList):
- contentextensions/NFA.cpp:
(WebCore::ContentExtensions::NFA::addRuleId):
- contentextensions/URLFilterParser.cpp:
(WebCore::ContentExtensions::Term::Term):
(WebCore::ContentExtensions::Term::~Term):
(WebCore::ContentExtensions::Term::isValid):
(WebCore::ContentExtensions::Term::addCharacter):
(WebCore::ContentExtensions::Term::quantify):
(WebCore::ContentExtensions::Term::quantifier):
(WebCore::ContentExtensions::Term::isUniversalTransition):
(WebCore::ContentExtensions::Term::visitSimpleTransitions):
(WebCore::ContentExtensions::Term::operator=):
(WebCore::ContentExtensions::Term::operator==):
(WebCore::ContentExtensions::Term::hash):
(WebCore::ContentExtensions::Term::isEmptyValue):
(WebCore::ContentExtensions::Term::isDeletedValue):
(WebCore::ContentExtensions::Term::destroy):
(WebCore::ContentExtensions::Term::CharacterSet::operator==):
(WebCore::ContentExtensions::Term::CharacterSet::hash):
(WebCore::ContentExtensions::TermHash::hash):
(WebCore::ContentExtensions::TermHash::equal):
(WebCore::ContentExtensions::GraphBuilder::GraphBuilder):
(WebCore::ContentExtensions::GraphBuilder::finalize):
(WebCore::ContentExtensions::GraphBuilder::atomPatternCharacter):
(WebCore::ContentExtensions::GraphBuilder::atomBuiltInCharacterClass):
(WebCore::ContentExtensions::GraphBuilder::quantifyAtom):
(WebCore::ContentExtensions::GraphBuilder::atomCharacterClassBegin):
(WebCore::ContentExtensions::GraphBuilder::atomCharacterClassAtom):
(WebCore::ContentExtensions::GraphBuilder::atomCharacterClassRange):
(WebCore::ContentExtensions::GraphBuilder::addTransitions):
(WebCore::ContentExtensions::GraphBuilder::sinkFloatingTerm):
(WebCore::ContentExtensions::GraphBuilder::sinkFloatingTermIfNecessary):
(WebCore::ContentExtensions::URLFilterParser::URLFilterParser):
(WebCore::ContentExtensions::URLFilterParser::~URLFilterParser):
(WebCore::ContentExtensions::URLFilterParser::addPattern):
(WebCore::ContentExtensions::trivialAtomFromASCIICharacter): Deleted.
(WebCore::ContentExtensions::quantifyTrivialAtom): Deleted.
(WebCore::ContentExtensions::trivialAtomQuantifier): Deleted.
(WebCore::ContentExtensions::trivialAtomForNewlineClassIDBuiltin): Deleted.
(WebCore::ContentExtensions::GraphBuilder::sinkAtom): Deleted.
(WebCore::ContentExtensions::GraphBuilder::generateTransition): Deleted.
(WebCore::ContentExtensions::GraphBuilder::sinkTrivialAtom): Deleted.
(WebCore::ContentExtensions::GraphBuilder::sinkCharacterSet): Deleted.
(WebCore::ContentExtensions::GraphBuilder::sinkPendingAtomIfNecessary): Deleted.
- contentextensions/URLFilterParser.h:
- 1:51 PM Changeset in webkit [181281] by
-
- 2 edits in trunk/Source/WebCore
Fix the !ENABLE(WEBGL) build after r180609
https://bugs.webkit.org/show_bug.cgi?id=142057
Reviewed by Darin Adler.
- platform/graphics/opengl/GraphicsContext3DOpenGLCommon.cpp:
(WebCore::GraphicsContext3D::forceContextLost):
- 1:51 PM Changeset in webkit [181280] by
-
- 3 edits in trunk/Tools
Add an option to run-webkit-tests to override the LayoutTests/ directory
https://bugs.webkit.org/show_bug.cgi?id=142418
Reviewed by David Kilzer.
Add an arugment to run-webkit-tests which, when set, overrides the port's default LayoutTests
directory. The base port will parse the options during initialization and store the override
location if present. layout_tests_dir() will return this overridden location if set.
- Scripts/webkitpy/layout_tests/run_webkit_tests.py:
(parse_args):
- Scripts/webkitpy/port/base.py:
(Port.init):
(Port.layout_tests_dir):
- 1:22 PM Changeset in webkit [181279] by
-
- 3 edits in trunk/Source/WebCore
Only call showControls if controls are hidden.
https://bugs.webkit.org/show_bug.cgi?id=142496.
<rdar://problem/20083732>
Reviewed by Dean Jackson.
- Modules/mediacontrols/mediaControlsApple.js:
(Controller.prototype.handleWrapperMouseMove):
(Controller.prototype.controlsAreHidden):
Controls are hidden if the ‘show’ class name is not present,
not if the ‘hidden’ class name is.
- 12:58 PM Changeset in webkit [181278] by
-
- 3 edits3 adds3 deletes in trunk
Work around a Cocoa font parsing bug
https://bugs.webkit.org/show_bug.cgi?id=142446
Source/WebCore:
<rdar://problem/20086223>
Reviewed by Simon Fraser.
Work around a bug in Cocoa regarding font parsing.
- svg/SVGToOTFFontConversion.cpp:
LayoutTests:
Reviewed by Simon Fraser.
Adding a few extra tests for bits of SVG fonts that could use some more tests.
- svg/fonts/resources/svg-font-general.svg: Updated and moved from svg-font-horiz-origin-font.svg
- svg/fonts/svg-font-general.html: Updated and moved from svg-font-horiz-origin.html
- svg/fonts/svg-font-general-expected.html: Updated and moved from svg-font-horiz-origin-expected.html
- 12:39 PM Changeset in webkit [181277] by
-
- 2 edits2 deletes in trunk/LayoutTests
[SVG -> OTF Converter] x-height cannot be specified
https://bugs.webkit.org/show_bug.cgi?id=142465
Reviewed by Simon Fraser.
For all non-system fonts, OS X disregards the font's reported x-height. Therefore,
it is impossible to satisfy this test.
- fast/css/ex-unit-with-no-x-height-expected.txt: Removed.
- fast/css/ex-unit-with-no-x-height.html: Removed.
- platform/mac/TestExpectations:
- 12:19 PM Changeset in webkit [181276] by
-
- 2 edits in trunk/LayoutTests
Unreviewed GTK+ Gardening
https://bugs.webkit.org/show_bug.cgi?id=142499
Patch by Marcos Chavarría Teijeiro <chavarria1991@gmail.com> on 2015-03-09
- platform/gtk/TestExpectations:
- 12:02 PM Changeset in webkit [181275] by
-
- 3 edits in trunk/Tools
[CMake] Enable WorkQueue tests for CMake ports
https://bugs.webkit.org/show_bug.cgi?id=142473
Patch by Martin Robinson <mrobinson@igalia.com> on 2015-03-09
Reviewed by Darin Adler.
- TestWebKitAPI/CMakeLists.txt: Add the WorkQueue tests source file to
the list of source files.
- Tools/TestWebKitAPI/PlatformEfl.cmake: Add the DispatchQueueEfl directory
to the list of include directories for EFL.
- 12:01 PM Changeset in webkit [181274] by
-
- 2 edits in trunk/Source/WebCore
Fix the build after r181153
https://bugs.webkit.org/show_bug.cgi?id=142497
Reviewed by Philippe Normand.
- platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
(WebCore::MediaPlayerPrivateGStreamer::load):
- 11:58 AM Changeset in webkit [181273] by
-
- 4 edits in trunk/Source/WebKit
[iOS] Fix build when compiling with dead code stripping disabled
<http://webkit.org/b/142467>
Reviewed by Simon Fraser.
Source/WebKit/ios:
- WebCoreSupport/WebInspectorClientIOS.mm:
(WebInspectorFrontendClient::save): Add missing method.
Source/WebKit/mac:
- Configurations/WebKitLegacy.xcconfig:
(EXCLUDED_SOURCE_FILE_NAMES[sdk=iphone*]): Disable mig code
generation and compilation for out-of-process plugin support.
This was never used on iOS.
- 11:55 AM WebKitGTK/Gardening/Calendar edited by
- (diff)
- 11:43 AM Changeset in webkit [181272] by
-
- 4 edits in trunk/Source/WebKit2
bmalloc: tryFastMalloc shouldn't crash
https://bugs.webkit.org/show_bug.cgi?id=142443
Reviewed by Anders Carlsson.
Part 1: Stop using tryFastRealloc.
- Shared/ShareableBitmap.cpp:
(WebKit::ShareableBitmap::resize): Deleted.
- Shared/ShareableBitmap.h: Removed the resize function because it has
no clients.
- WebProcess/Plugins/PluginProxy.cpp:
(WebKit::PluginProxy::updateBackingStore): Changed to allocate a new
backing store instead of resizing the old one. This has three advantages:
(1) Might be more memory-efficient, since you don't have to keep the old
one around while allocating the new one.
(2) Avoids the overhead of realloc() copying the contents of the old
backing store even though we only want uninitialized memory.
(3) Makes resize failure consistent with initial allocation failure.
Previously, while initial allocation failure would set the backing store
to null, resize failure would keep the old wrong backing store and then
tell it not to paint. Now, resize failure also sets the backing store to
null.
- 10:49 AM Changeset in webkit [181271] by
-
- 4 edits3 moves in trunk/Source
[EFL] Move DispatchQueue to WTF
https://bugs.webkit.org/show_bug.cgi?id=142493
Patch by Martin Robinson <mrobinson@igalia.com> on 2015-03-09
Reviewed by Csaba Osztrogonác.
Source/WebKit2:
- PlatformEfl.cmake: Remove DispatchQueue from the source list.
Source/WTF:
- wtf/PlatformEfl.cmake: Add DispatchQueue to the source list.
- wtf/efl/DispatchQueueEfl.cpp: Renamed from Source/WebKit2/Platform/efl/DispatchQueueEfl.cpp.
- wtf/efl/DispatchQueueEfl.h: Renamed from Source/WebKit2/Platform/efl/DispatchQueueEfl.h.
- wtf/efl/DispatchQueueWorkItemEfl.h: Renamed from Source/WebKit2/Platform/efl/DispatchQueueWorkItemEfl.h.
- 10:35 AM Changeset in webkit [181270] by
-
- 2 edits in trunk/Source/WebCore
Unreviewed. Re-add CSS property that was unintentionally removed in r180893
<rdar://problem/20083715>
- Modules/mediacontrols/mediaControlsApple.css:
(audio::-webkit-media-controls-panel):
- 10:19 AM Changeset in webkit [181269] by
-
- 5 edits in branches/safari-600.5-branch/Source
Versioning.
- 10:09 AM Changeset in webkit [181268] by
-
- 3 edits in trunk/Source/WebCore
Simple line layout: Merge TextFragmentIterator::findNextBreakablePosition() and TextFragmentIterator::findNextNonWhitespacePosition().
https://bugs.webkit.org/show_bug.cgi?id=142344
Reviewed by Antti Koivisto.
This patch merges findNextBreakablePosition() and findNextNonWhitespacePosition() so that
the segment looping and position handling logic are no longer duplicated. It also unifies
the static next*() functions' signature.
No change in functionality.
- rendering/SimpleLineLayoutTextFragmentIterator.cpp:
(WebCore::SimpleLineLayout::TextFragmentIterator::nextTextFragment):
(WebCore::SimpleLineLayout::nextBreakablePosition):
(WebCore::SimpleLineLayout::nextNonWhitespacePosition):
(WebCore::SimpleLineLayout::TextFragmentIterator::skipToNextPosition):
(WebCore::SimpleLineLayout::TextFragmentIterator::findNextBreakablePosition): Deleted.
(WebCore::SimpleLineLayout::findNextNonWhitespace): Deleted.
(WebCore::SimpleLineLayout::TextFragmentIterator::findNextNonWhitespacePosition): Deleted.
- rendering/SimpleLineLayoutTextFragmentIterator.h:
- 7:27 AM Changeset in webkit [181267] by
-
- 2 edits in trunk/Tools
Attempt to fix timeouts on iOS Simulator layout tests
- Scripts/webkitpy/port/ios.py:
(IOSSimulatorPort.setup_test_run): Most layout test runs on iOS
Simulator end after the first 50 tests time out. Some analysis
this past weekend shows that the 'Booted' status in the output
ofxcrun simctl listis not sufficient (and watching the test
run shows that SpringBoard hasn't fully started to the point
where it's displaying icons on the home screen). The workaround
until we have a better fix is to add an arbitrary timeout to let
SpringBoard finish launching. In practice, 15 seconds is
probably enough, but 30 seconds should nearly always work.
- 6:57 AM Changeset in webkit [181266] by
-
- 2 edits in trunk/Tools
Unreviewed. Add myself as a committer.
- Scripts/webkitpy/common/config/contributors.json:
- 5:57 AM Changeset in webkit [181265] by
-
- 2 edits in trunk/Tools
Attempt to fix WTF_WorkQueue.TwoQueues timeout test failure
- TestWebKitAPI/Tests/WTF/WorkQueue.cpp:
(TestWebKitAPI::TEST): Change delay from 50ns to 50ms, and add
an expectation that will fail if we hit the race condition.
This test is inherently racy and will deadlock if the first job
for queue2 takes the mutex lock before the second job for
queue1.
- 4:36 AM Changeset in webkit [181264] by
-
- 3 edits in trunk/Source/WebCore
ReadableStreamJSSource should use JSC::Strong UnspecifiedBoolType operator
https://bugs.webkit.org/show_bug.cgi?id=142397
Reviewed by Benjamin Poulain.
No behavior change. Current tests cover the changes.
- bindings/js/ReadableStreamJSSource.cpp:
- bindings/js/ReadableStreamJSSource.h:
- 4:31 AM Changeset in webkit [181263] by
-
- 2 edits in trunk/LayoutTests
Unreviewed EFL gardening on 9th Mar.
Mark svg/text/select-textLength-spacingAndGlyphs-foo.html to failure.
Update new failling tests to existing bugs.
- platform/efl/TestExpectations:
- 4:31 AM Changeset in webkit [181262] by
-
- 3 edits in trunk/Source/WebCore
ReadableStream does not not need to pass itself as callback parameter
https://bugs.webkit.org/show_bug.cgi?id=142399
Reviewed by Darin Adler.
No change in behavior.
- Modules/streams/ReadableStream.h: Remove ReadableStream parameter from SuccessCallback definition.
- bindings/js/JSReadableStreamCustom.cpp:
(WebCore::JSReadableStream::ready): Capture JSReadableStream instance instead.
(WebCore::JSReadableStream::closed): Capture JSReadableStream instance instead.
- 3:16 AM WebKitGTK/Gardening/Calendar edited by
- (diff)
- 12:28 AM Changeset in webkit [181261] by
-
- 2 edits in releases/WebKitGTK/webkit-2.8/Source/WebCore
Merge r181256 - Crash in WebCore::NotificationCenter::stop()
https://bugs.webkit.org/show_bug.cgi?id=142444
<rdar://problem/20082520>
Reviewed by Darin Adler.
Rework the patch in r181219 so that we do not need a Ref<NotificationCenter> protector
in NotificationCenter::stop(). Instead, we put the client in a local variable and null
out m_client *before* calling NotificationClient::clearNotifications().
No new tests, already covered by:
http/tests/notifications/event-listener-crash.html
- Modules/notifications/NotificationCenter.cpp:
(WebCore::NotificationCenter::stop):
Mar 8, 2015:
- 1:47 AM Changeset in webkit [181235] by
-
- 2 edits in releases/WebKitGTK/webkit-2.8/Tools
Merge r181075 - Unreviewed. Fix /webkit2/WebKitWebResource/mime-type after r180927.
In r180927 we updated the libsoup version used by the jhbuild. In
this new version the sniffer uses image/x-icon instead of
image/vnd.microsoft.icon for blank.ico resource.
- TestWebKitAPI/Tests/WebKit2Gtk/TestResources.cpp:
(testWebResourceMimeType):
- 1:46 AM Changeset in webkit [181234] by
-
- 4 edits in releases/WebKitGTK/webkit-2.8
Merge r181074 - [SOUP] Check TLS errors as soon as they are set in the SoupMessage
https://bugs.webkit.org/show_bug.cgi?id=142244
Reviewed by Sergio Villar Senin.
Source/WebCore:
Connect to the notify::tls-errors signal of SoupMessage to cancel
the load earlier in case of TLS failure, preventing any private
data from being sent to the server before the TLS errors are checked.
- platform/network/soup/ResourceHandleSoup.cpp:
(WebCore::tlsErrorsChangedCallback):
(WebCore::gotHeadersCallback):
(WebCore::createSoupMessageForHandleAndRequest):
Tools:
Check that the SSL server doesn't process any request in case of
TLS errors when the policy is set to FAIL.
- TestWebKitAPI/Tests/WebKit2Gtk/TestSSL.cpp:
(testTLSErrorsPolicy):
(testTLSErrorsRedirect):
(testTLSErrorsHTTPAuth):
(testLoadFailedWithTLSErrors):
(testSubresourceLoadFailedWithTLSErrors):
(httpsServerCallback):
- 1:44 AM Changeset in webkit [181233] by
-
- 3 edits in releases/WebKitGTK/webkit-2.8/Source/WebKit2
Merge r181073 - [SOUP] Disable RC4
https://bugs.webkit.org/show_bug.cgi?id=140014
Patch by Michael Catanzaro <Michael Catanzaro> on 2015-03-05
Reviewed by Carlos Garcia Campos.
Disallow RC4-based ciphersuites when performing TLS negotiation,
because it is no longer considered secure.
- NetworkProcess/EntryPoint/unix/NetworkProcessMain.cpp:
(main):
- WebProcess/EntryPoint/unix/WebProcessMain.cpp:
(main):
- 1:43 AM Changeset in webkit [181232] by
-
- 4 edits in releases/WebKitGTK/webkit-2.8/Source/WebCore
Merge r181072 - TextCheckingParagraph::isEmpty() is sufficient to check whether paragraph is empty
https://bugs.webkit.org/show_bug.cgi?id=142276
Reviewed by Darin Adler.
TextCheckingParagraph::isEmpty() already calls TextCheckingParagraph::isRangeEmpty().
There is no need to call them both at the caller site.
No new tests. No behavior change.
- editing/Editor.cpp:
(WebCore::Editor::markAllMisspellingsAndBadGrammarInRanges):
Update caller site.
- editing/TextCheckingHelper.cpp:
(WebCore::TextCheckingParagraph::isEmpty):
Avoid using helepers here to get rid of them as they are
no longer needed outside TextCheckingParagraph.
- editing/TextCheckingHelper.h:
(WebCore::TextCheckingParagraph::isTextEmpty): Deleted.
(WebCore::TextCheckingParagraph::isRangeEmpty): Deleted.
Do not expose isTextEmpty() and isRangeEmpty().
- 1:42 AM Changeset in webkit [181231] by
-
- 5 edits in releases/WebKitGTK/webkit-2.8/Source/JavaScriptCore
Merge r181060 - GC should compute stack bounds and dump registers at the earliest opportunity.
<https://webkit.org/b/142310>
<rdar://problem/20045624>
Reviewed by Geoffrey Garen.
Make Heap::collect() a wrapper function around a collectImpl() where the work is actually done.
The wrapper function that grabs a snapshot of the current stack boundaries and register values
on entry, and sanitizes the stack on exit.
This is a speculative fix for what appears to be overly conservative behavior in the garbage
collector following r178364 which caused a measurable regression in memory usage on Membuster.
The theory being that we were putting pointers to dead things on the stack before scanning it,
and by doing that ended up marking things that we'd otherwise discover to be garbage.
- heap/Heap.cpp:
(JSC::Heap::markRoots):
(JSC::Heap::gatherStackRoots):
(JSC::Heap::collect):
(JSC::Heap::collectImpl):
- heap/Heap.h:
- heap/MachineStackMarker.cpp:
(JSC::MachineThreads::gatherFromCurrentThread):
(JSC::MachineThreads::gatherConservativeRoots):
- heap/MachineStackMarker.h:
- 1:20 AM Changeset in webkit [181230] by
-
- 2 edits in releases/WebKitGTK/webkit-2.8/Source/JavaScriptCore
Merge r181059 - Silence GCC's -Wstrict-prototypes
https://bugs.webkit.org/show_bug.cgi?id=142278
Patch by Debarshi Ray <debarshir@gnome.org> on 2015-03-04
Reviewed by Alexey Proskuryakov.
- API/JSContextRef.h:
- 1:17 AM Changeset in webkit [181229] by
-
- 2 edits in releases/WebKitGTK/webkit-2.8/Source/WebCore
Merge r181038 - REGRESSION (r179597): Can't see power saver banner for plugins
https://bugs.webkit.org/show_bug.cgi?id=142312
<rdar://problem/20040517>
Reviewed by Brent Fulgham.
We were being a bit too restrictive when deciding a child
should not create a renderer. All shadow root children
of the snapshotted plugin need one.
- html/HTMLPlugInImageElement.cpp:
(WebCore::HTMLPlugInImageElement::childShouldCreateRenderer):
Test if we're part of the shadow tree.
- 1:12 AM Changeset in webkit [181228] by
-
- 4 edits in releases/WebKitGTK/webkit-2.8
Merge r180998 - [GTK] WebView should hold a reference on WebContext because non-default contexts are a reality
https://bugs.webkit.org/show_bug.cgi?id=142225
Patch by Debarshi Ray <debarshir@gnome.org> on 2015-03-04
Reviewed by Carlos Garcia Campos.
Source/WebKit2:
- UIProcess/API/gtk/WebKitWebView.cpp:
(webkitWebViewRequestFavicon):
(webkitWebViewWatchForChangesInFavicon):
(webkitWebViewDisconnectFaviconDatabaseSignalHandlers):
(webkitWebViewConstructed):
(webkitWebViewGetProperty):
(webkitWebViewDispose):
(webkitWebViewLoadChanged):
(webkitWebViewLoadFailedWithTLSErrors):
(webkit_web_view_get_context):
(webkit_web_view_download_uri):
Tools:
- TestWebKitAPI/Tests/WebKit2Gtk/TestWebKitWebView.cpp:
(testWebViewWebContextLifetime):
(beforeAll):
- 1:11 AM Changeset in webkit [181227] by
-
- 2 edits in releases/WebKitGTK/webkit-2.8/Source/WebCore
Merge r180997 - [GStreamer] the GST_SCHEDULING_FLAG_BANDWIDTH_LIMITED should be wrapped by a ifdef
https://bugs.webkit.org/show_bug.cgi?id=142274
Patch by Philippe Normand <pnormand@igalia.com> on 2015-03-04
Reviewed by Carlos Garcia Campos.
Don't handle scheduling queries if building against versions of
GStreamer older than 1.2.0.
- platform/graphics/gstreamer/WebKitWebSourceGStreamer.cpp:
(webKitWebSrcQueryWithParent):
- 1:06 AM Changeset in webkit [181226] by
-
- 4 edits2 adds in releases/WebKitGTK/webkit-2.8
Merge r181168 - Setting any of the <object> element plugin controlling attributes does not have any affect.
https://bugs.webkit.org/show_bug.cgi?id=141936.
Patch by Said Abou-Hallawa <sabouhallawa@apple.com> on 2015-03-06
Reviewed by Simon Fraser.
Source/WebCore:
When setting any of the <object> element plugin controlling attributes
dynamically we need to mark the the element to be dirty by calling
setNeedsStyleRecalc(), so it has to recreate its renderer when needed.
Tests: fast/css/image-object-hover-inherit.html
svg/as-object/svg-in-object-dynamic-attribute-change.html
- dom/Element.h: Delete unimplemented function.
- html/HTMLObjectElement.cpp:
(WebCore::HTMLObjectElement::parseAttribute): Mark the element dirty by
calling setNeedsStyleRecalc() when one of the plugin controlling attributes
gets changed. We have to clear m_useFallbackContent because the attribute's
new value might fix the object rendering.
- html/HTMLObjectElement.h: Add a function to clear m_useFallbackContent.
LayoutTests:
- fast/css/image-object-hover-inherit-expected.html: Added.
- fast/css/image-object-hover-inherit.html: Added.
A guarding test to catch the case of reconstructing the image <object>
renderer while performing a synchronous resolveTree() followed by page
rendering or dump render tree.
- svg/as-object/resources/lime100x100.html: Added.
- svg/as-object/resources/lime100x100.png: Added.
- svg/as-object/resources/lime100x100.svg: Added.
- svg/as-object/resources/red100x100.svg: Added.
- svg/as-object/svg-in-object-dynamic-attribute-change-expected.html: Added.
- svg/as-object/svg-in-object-dynamic-attribute-change.html: Added.
Ensure that changing the 'type' and the 'data' attributes of the <object>
element will have the expected outcome. Also make sure that the <object>
element renderer falls back correctly when setting any of the attributes
to some unexpected value.
- 12:49 AM Changeset in webkit [181225] by
-
- 2 edits in releases/WebKitGTK/webkit-2.8/Source/bmalloc
Merge r180964 - bmalloc: Don't branch when setting the owner of a large object
https://bugs.webkit.org/show_bug.cgi?id=142241
Reviewed by Andreas Kling.
- bmalloc/BoundaryTag.h:
(bmalloc::BoundaryTag::owner):
(bmalloc::BoundaryTag::setOwner):
- 12:47 AM Changeset in webkit [181224] by
-
- 6 edits in releases/WebKitGTK/webkit-2.8/Source/bmalloc
Merge r180953 - bmalloc: Miscellaneous cleanup
https://bugs.webkit.org/show_bug.cgi?id=142231
Reviewed by Andreas Kling.
No performance change -- maybe a tiny reduction in memory use.
- bmalloc/Heap.cpp: Moved the sleep function into StaticMutex, since
it's a helper for working with mutexes.
(bmalloc::Heap::scavenge): Make sure to wait before we start any
scavenging, since individual scavenging functions now always scavenge
at least one page before waiting themselves.
(bmalloc::Heap::scavengeSmallPages):
(bmalloc::Heap::scavengeMediumPages):
(bmalloc::Heap::scavengeLargeObjects): Use the new wait helper to
simplify this code. Also, we now require our caller to wait until at
least one deallocation is desirable. This simplifies our loop.
(bmalloc::Heap::allocateSmallPage):
(bmalloc::Heap::allocateMediumPage):
(bmalloc::Heap::allocateXLarge):
(bmalloc::Heap::allocateLarge): Don't freak out any time the heap does
an allocation. Only consider the heap to be growing if it actually needs
to allocate new VM. This allows us to shrink the heap back down from a
high water mark more reliably even if heap activity continues.
(bmalloc::sleep): Deleted.
(bmalloc::Heap::scavengeLargeRanges): Renamed to match our use of
"LargeObject".
- bmalloc/Heap.h:
- bmalloc/LargeObject.h:
(bmalloc::LargeObject::operator bool): Added to simplify a while loop.
- bmalloc/StaticMutex.h:
(bmalloc::sleep):
(bmalloc::waitUntilFalse): New helper for waiting until a condition
becomes reliably false.
- bmalloc/Vector.h:
(bmalloc::Vector<T>::~Vector): Oops! Don't deallocate the null pointer.
We don't actually run any Vector destructors, but an iteration of this
patch did, and then crashed. So, let's fix that.
- 12:20 AM Changeset in webkit [181223] by
-
- 2 edits in releases/WebKitGTK/webkit-2.8/Source/JavaScriptCore
Merge r180938 - [Win64] JSC compile error.
https://bugs.webkit.org/show_bug.cgi?id=142216
Patch by peavo@outlook.com <peavo@outlook.com> on 2015-03-03
Reviewed by Mark Lam.
There is missing a version of setupArgumentsWithExecState when NUMBER_OF_ARGUMENT_REGISTERS == 4.
- jit/CCallHelpers.h:
(JSC::CCallHelpers::setupArgumentsWithExecState):
- 12:10 AM Changeset in webkit [181222] by
-
- 7 edits in trunk
Remove willDestroyFrame in WKBundlePageLoaderClient
https://bugs.webkit.org/show_bug.cgi?id=142389
Reviewed by Anders Carlsson.
Source/WebKit2:
I added willDestroyFrame in r154540 but I realized recently that
it has never actually worked. Adding the callback to
WKBundlePageLoaderClient was the first mistake, since frames are
handled by WebProcess and DidDestroyFrame message is indeed a
WebProcessProxy not a WebPageProxy one. Second mistake was calling
the callback from WebFrameLoaderClient::frameLoaderDestroyed(),
since at that point the frame has already been detached from the
page and so WebFrame::page() always returns nullptr. So, this in
fact dead code and since r180211 nobody is using it.
- WebProcess/InjectedBundle/API/c/WKBundlePageLoaderClient.h:
- WebProcess/InjectedBundle/API/gtk/WebKitWebPage.cpp:
(webkitWebPageCreate): Use verion 6 instead of version 7 of loader client.
- WebProcess/InjectedBundle/InjectedBundlePageLoaderClient.cpp:
(WebKit::InjectedBundlePageLoaderClient::willDestroyFrame): Deleted.
- WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::frameLoaderDestroyed):
Tools:
- WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::InjectedBundlePage):
Mar 7, 2015:
- 10:32 PM Changeset in webkit [181260] by
-
- 16 edits in trunk/LayoutTests
Last test gardening after r177774 (for reals this time)
https://bugs.webkit.org/show_bug.cgi?id=139968
Unreviewed.
Patch by Myles C. Maxfield <mmaxfield@apple.com> on 2015-03-08
- platform/mac/TestExpectations:
- platform/mac/svg/text/select-textLength-spacingAndGlyphs-squeeze-1-expected.txt:
- platform/mac/svg/text/select-textLength-spacingAndGlyphs-squeeze-2-expected.txt:
- platform/mac/svg/text/select-textLength-spacingAndGlyphs-squeeze-3-expected.txt:
- platform/mac/svg/text/select-textLength-spacingAndGlyphs-squeeze-4-expected.txt:
- platform/mac/svg/text/select-textLength-spacingAndGlyphs-stretch-1-expected.txt:
- platform/mac/svg/text/select-textLength-spacingAndGlyphs-stretch-2-expected.txt:
- platform/mac/svg/text/select-textLength-spacingAndGlyphs-stretch-3-expected.txt:
- svg/text/select-textLength-spacingAndGlyphs-squeeze-1.svg:
- svg/text/select-textLength-spacingAndGlyphs-squeeze-2.svg:
- svg/text/select-textLength-spacingAndGlyphs-squeeze-3.svg:
- svg/text/select-textLength-spacingAndGlyphs-squeeze-4.svg:
- svg/text/select-textLength-spacingAndGlyphs-stretch-1.svg:
- svg/text/select-textLength-spacingAndGlyphs-stretch-2.svg:
- svg/text/select-textLength-spacingAndGlyphs-stretch-3.svg:
- 9:55 PM Changeset in webkit [181259] by
-
- 12 edits in trunk/Source
[iOS Media] Video buffer progress missing on iPhones
https://bugs.webkit.org/show_bug.cgi?id=142462
Reviewed by Eric Carlson.
Source/WebCore:
The fullscreen controller on the iPhone idiom was using
the media duration as its maximum loaded time, meaning the
download progress was missing (it looked like everything
was already downloaded). The fix is to calculate
what the maximum buffered time is from the HTMLMediaElement,
and send that to the fullscreen controller.
- Modules/mediacontrols/mediaControlsiOS.js:
(ControllerIOS.prototype.drawTimelineBackground): While implementing
the native function I noticed that the JS controls forgot that
the buffered time ranges are ordered, and thus they don't need
to search through the array. The last range will be the highest.
- html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::maxBufferedTime): New helper function
to calculate the highest buffered time in the media source.
- html/HTMLMediaElement.h:
- platform/ios/WebVideoFullscreenInterface.h: New setBufferedTime method.
- platform/ios/WebVideoFullscreenInterfaceAVKit.h:
- platform/ios/WebVideoFullscreenInterfaceAVKit.mm:
(WebVideoFullscreenInterfaceAVKit::setDuration): No longer set
the loadedTimeRanges.
(WebVideoFullscreenInterfaceAVKit::setBufferedTime): Calculate a normalized
value of the maximum loaded/buffered time, and tell the platform controller
to show that range.
- platform/ios/WebVideoFullscreenModelVideoElement.mm:
(WebVideoFullscreenModelVideoElement::updateForEventName): Send the maximum
buffered time.
Source/WebKit2:
Coordinate the SetBufferedTime message between the Web and UI
processes.
- UIProcess/ios/WebVideoFullscreenManagerProxy.messages.in: New message SetBufferedTime.
- WebProcess/ios/WebVideoFullscreenManager.h:
- WebProcess/ios/WebVideoFullscreenManager.mm:
(WebKit::WebVideoFullscreenManager::setBufferedTime):
- 9:25 PM Changeset in webkit [181258] by
-
- 3 edits in trunk/LayoutTests
Fix the tests for the new double-child descendant in css-set-selector-text
https://bugs.webkit.org/show_bug.cgi?id=141882
Reviewed by Andreas Kling.
In 178592, I copied the rules from css-selector-text.html without removing
the declarations.
Kudos to Dhi Aurrahman for catching this.
- fast/css/css-set-selector-text-expected.txt:
- fast/css/css-set-selector-text.html:
- 9:15 PM Changeset in webkit [181257] by
-
- 4 edits in trunk
[Win] WTF::WorkQueue does not balance ref/deref properly
https://bugs.webkit.org/show_bug.cgi?id=142471
Reviewed by Antti Koivisto.
Source/WTF:
Make sure we deref the count when we execute a function in the
WorkQueue.
- wtf/win/WorkQueueWin.cpp:
(WTF::WorkQueue::performWorkOnRegisteredWorkThread):
Tools:
- TestWebKitAPI/Tests/WTF/WorkQueue.cpp:
(TestWebKitAPI::TEST): Check that the refcount is correct.
- 7:55 PM Changeset in webkit [181256] by
-
- 2 edits in trunk/Source/WebCore
Crash in WebCore::NotificationCenter::stop()
https://bugs.webkit.org/show_bug.cgi?id=142444
<rdar://problem/20082520>
Reviewed by Darin Adler.
Rework the patch in r181219 so that we do not need a Ref<NotificationCenter> protector
in NotificationCenter::stop(). Instead, we put the client in a local variable and null
out m_client *before* calling NotificationClient::clearNotifications().
No new tests, already covered by:
http/tests/notifications/event-listener-crash.html
- Modules/notifications/NotificationCenter.cpp:
(WebCore::NotificationCenter::stop):
- 7:07 PM Changeset in webkit [181255] by
-
- 2 edits in trunk/Tools
Do not show passing test result on output of EFL layout test
https://bugs.webkit.org/show_bug.cgi?id=142380
Reviewed by Darin Adler.
To show passing tests has caused inconvenience when we detect problem tests
in output of layout test. So this patch applies '--debug-rwt-logging' except for EFL port
in order not to show passing tests in the output of EFL layout test.
- BuildSlaveSupport/build.webkit.org-config/master.cfg:
(RunWebKitTests.start):
- 6:58 PM Changeset in webkit [181254] by
-
- 2 edits in trunk/Tools
Unreviewed build fix.
- TestWebKitAPI/Tests/WTF/WorkQueue.cpp:
(TestWebKitAPI::TEST): Remove two unused variables.
- 6:44 PM Changeset in webkit [181253] by
-
- 2 edits in trunk/LayoutTests
Change bug number for svg/custom/svg-fonts-without-missing-glyph.xhtml
Unreviewed.
Patch by Myles C. Maxfield <mmaxfield@apple.com> on 2015-03-08
- platform/mac/TestExpectations:
- 6:30 PM Changeset in webkit [181252] by
-
- 7 edits7 adds in trunk/Source/WebKit2
[Content Extensions] Add scaffolding for the WKUserContentExtensionStore
https://bugs.webkit.org/show_bug.cgi?id=142464
Reviewed by Dan Bernstein.
- Shared/API/APIObject.h:
Add a UserContentExtensionStore type and move the UserContentExtension type into the correct section (UIProcess).
- Shared/API/c/WKBase.h:
Add the WKUserContentExtensionStoreRef C-SPI type.
- Shared/Cocoa/APIObject.mm:
(API::Object::newObject):
Add allocation support for UserContentExtensionStore (and UserContentExtension while we are there, it will be
needed shortly).
- UIProcess/API/APIUserContentExtensionStore.cpp: Added.
(API::UserContentExtensionStore::UserContentExtensionStore):
(API::UserContentExtensionStore::~UserContentExtensionStore):
- UIProcess/API/APIUserContentExtensionStore.h: Added.
Stub out the UserContentExtensionStore.
- UIProcess/API/C/WKAPICast.h:
Add casting for the UserContentExtensionStore.
- UIProcess/API/C/WKUserContentExtensionStoreRef.cpp: Added.
(WKUserContentExtensionStoreGetTypeID):
- UIProcess/API/C/WKUserContentExtensionStoreRef.h: Added.
- UIProcess/API/Cocoa/_WKUserContentExtensionStore.h: Added.
- UIProcess/API/Cocoa/_WKUserContentExtensionStore.mm: Added.
(-[_WKUserContentExtensionStore init]):
(-[_WKUserContentExtensionStore dealloc]):
(-[_WKUserContentExtensionStore _apiObject]):
- UIProcess/API/Cocoa/_WKUserContentExtensionStoreInternal.h: Added.
(WebKit::wrapper):
Stub out SPI for the UserContentExtensionStore.
- WebKit2.xcodeproj/project.pbxproj:
- CMakeLists.txt:
Add new files.
- 6:17 PM Changeset in webkit [181251] by
-
- 5 edits in trunk/Source/WebKit2
AX: Expose a frame's accessible name through WK2 C SPI
<rdar://problem/19653068>
https://bugs.webkit.org/show_bug.cgi?id=142468
Reviewed by Dan Bernstein.
- WebProcess/InjectedBundle/API/c/WKBundleFrame.cpp:
(WKBundleFrameSetAccessibleName):
- WebProcess/InjectedBundle/API/c/WKBundleFrame.h:
- WebProcess/WebPage/WebFrame.cpp:
(WebKit::WebFrame::setAccessibleName):
- WebProcess/WebPage/WebFrame.h:
Add SPI to set a frame's accessible name. This matches the Legacy WebKit SPI -[WebFrame setAccessibleName:].
- 4:58 PM Changeset in webkit [181250] by
-
- 3 edits in trunk/Source/JavaScriptCore
JITThunks keeps finalized Weaks around, pinning WeakBlocks.
<https://webkit.org/b/142454>
Reviewed by Darin Adler.
Make JITThunks a WeakHandleOwner so it can keep its host function map free of stale entries.
This fixes an issue I was seeing where a bunch of WeakBlocks stuck around with nothing but
finalized Weak<NativeExecutable> entries.
- jit/JITThunks.h:
- jit/JITThunks.cpp:
(JSC::JITThunks::finalize): Make JITThunks inherit from WeakHandleOwner so it can receive
a callback when the NativeExecutables get garbage collected.
(JSC::JITThunks::hostFunctionStub): Pass 'this' as the handle owner when creating Weaks.
- 4:55 PM Changeset in webkit [181249] by
-
- 4 edits1 add in trunk/Tools
Add some WorkQueue tests to TestWebKitAPI
https://bugs.webkit.org/show_bug.cgi?id=142447
Reviewed by Sam Weinig.
- TestWebKitAPI/TestWebKitAPI.vcxproj/TestWebKitAPI.vcxproj:
- TestWebKitAPI/TestWebKitAPI.vcxproj/TestWebKitAPI.vcxproj.filters:
- TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
- TestWebKitAPI/Tests/WTF/WorkQueue.cpp: Added.
(TestWebKitAPI::simpleTest):
(TestWebKitAPI::longTest):
(TestWebKitAPI::thirdTest):
(TestWebKitAPI::finishedTests):
(TestWebKitAPI::TEST):
(TestWebKitAPI::longTest2):
- 4:47 PM Changeset in webkit [181248] by
-
- 3 edits in trunk/Source/JavaScriptCore
BuiltinExecutables keeps finalized Weaks around, pinning WeakBlocks.
<https://webkit.org/b/142460>
Reviewed by Geoffrey Garen.
Make BuiltinExecutables a WeakHandleOwner so it can clear out its respective Weak members
if and when their pointees get garbage collected.
This fixes an issue I've seen locally where a WeakBlock is pinned down by a single one of
these Weak<BuiltinExecutables>.
- builtins/BuiltinExecutables.h: Make BuiltinExecutables inherit from WeakHandleOwner.
- builtins/BuiltinExecutables.cpp:
(JSC::BuiltinExecutables::finalize): Clear out the relevant member pointer when it's been
garbage collected. We use the WeakImpl's "context" field to pass the address of the member.
- 4:39 PM Changeset in webkit [181247] by
-
- 3 edits in trunk/Source/WebCore
Don't call RenderLayerCompositor::needsToBeComposited() so many times
https://bugs.webkit.org/show_bug.cgi?id=142441
Reviewed by Brent Fulgham.
Save one call to needsToBeComposited() per RenderLayer while rebuilding compositing
layers, and a call to isRunningAcceleratedTransformAnimation() in some cases.
computeCompositingRequirements() already knows if the layer will be composited, so
pass that to updateBacking() in the form of a "yes/no/don't" know enum. However,
since updateBacking() no longer calls needsToBeComposited(), we need to call
requiresCompositingForPosition() to get the viewportConstrainedNotCompositedReason flags.
isRunningAcceleratedTransformAnimation() will only be true for a layer which has a
direct compositing reason, so only call isRunningAcceleratedTransformAnimation()
in that case.
- rendering/RenderLayerCompositor.cpp:
(WebCore::rendererForCompositingTests):
(WebCore::RenderLayerCompositor::updateBacking):
(WebCore::RenderLayerCompositor::computeCompositingRequirements):
(WebCore::RenderLayerCompositor::requiresCompositingLayer): Use references.
(WebCore::RenderLayerCompositor::reasonsForCompositing): Use references.
- rendering/RenderLayerCompositor.h:
- 2:32 PM Changeset in webkit [181246] by
-
- 9 edits in trunk/Source/WebKit2
[Content Extensions] Pass compiled content extensions as read-only shared memory rather than copying them
https://bugs.webkit.org/show_bug.cgi?id=142458
Reviewed by Dan Bernstein.
- Shared/WebCompiledContentExtensionData.cpp:
(WebKit::WebCompiledContentExtensionData::encode):
(WebKit::WebCompiledContentExtensionData::decode):
- Shared/WebCompiledContentExtensionData.h:
(WebKit::WebCompiledContentExtensionData::WebCompiledContentExtensionData):
Store the compiled data as a SharedMemory block with offset/size pairs for the
bytecode and actions. While the bytecode offset is currently always 0, this won't
be the case when we map these in from files due to metadata that will be placed
at the top of the files.
- Shared/WebCompiledContentExtension.h:
- Shared/WebCompiledContentExtension.cpp:
(WebKit::WebCompiledContentExtension::createFromCompiledContentExtensionData):
Temporary creation helper. In subsequent patches, this will be removed and replaced
by always passing in shared/mapped data.
(WebKit::WebCompiledContentExtension::create):
(WebKit::WebCompiledContentExtension::WebCompiledContentExtension):
(WebKit::WebCompiledContentExtension::bytecode):
(WebKit::WebCompiledContentExtension::bytecodeLength):
(WebKit::WebCompiledContentExtension::actions):
(WebKit::WebCompiledContentExtension::actionsLength):
Convert to work with the new WebCompiledContentExtension format.
- UIProcess/API/C/WKUserContentFilterRef.cpp:
(WKUserContentFilterCreate):
- UIProcess/API/Cocoa/_WKUserContentFilter.mm:
(-[_WKUserContentFilter initWithName:serializedRules:]):
Switch to using WebCompiledContentExtension::createFromCompiledContentExtensionData().
- WebProcess/UserContent/WebUserContentController.cpp:
(WebKit::WebUserContentController::addUserContentExtensions):
- WebProcess/WebPage/WebPageGroupProxy.cpp:
(WebKit::WebPageGroupProxy::addUserContentExtension):
Update to pass the WebCompiledContentExtensionData directly to the WebCompiledContentExtension.
- 10:02 AM Changeset in webkit [181245] by
-
- 3 edits in trunk/Source/WebCore
In RenderLayerCompositor, track layer bounds and the haveComputedBounds flag together in a struct
https://bugs.webkit.org/show_bug.cgi?id=142449
Reviewed by Brent Fulgham.
Rather than tracking 'layerBounds' and 'haveComputedBounds' separately, group them together
in an OverlapExtent struct and add a function to compute the bounds, which removes
a copy of this code. OverlapExtent will be extended in a future patch.
No behavior change.
- rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::OverlapMap::geometryMap): Allow const access to
the geometryMap, used by computeExtent().
(WebCore::RenderLayerCompositor::computeExtent):
(WebCore::RenderLayerCompositor::addToOverlapMap):
(WebCore::RenderLayerCompositor::addToOverlapMapRecursive):
(WebCore::RenderLayerCompositor::computeCompositingRequirements):
- rendering/RenderLayerCompositor.h:
- 3:50 AM Changeset in webkit [181244] by
-
- 2 edits in releases/WebKitGTK/webkit-2.8/Source/WebCore
Merge r181219 - Crash in WebCore::NotificationCenter::stop()
https://bugs.webkit.org/show_bug.cgi?id=142444
<rdar://problem/20082520>
Reviewed by Andreas Kling.
A use-after-free would sometimes cause us to crash in NotificationCenter::stop().
After investigation, it turns out that NotificationCenter::stop() calls
NotificationClient::clearNotifications() which will destroy the Notification
objects, all of which hold a strong reference to the NotificationCenter. If at
this point, only Notifications are ref'ing the NotificationCenter, this means
that the NotificationCenter will get destroyed right after the call to
NotificationClient::clearNotifications(). However, we reset m_client to null
after calling clearNotifications() and it causes us to crash in this case.
The issue is addressed by adding a Ref<NotificationCenter> protector in
NotificationCenter::stop() so that we make sure the NotificationCenter lives
at least until the end of the method execution.
I was able to consistently reproduce the crash by doing:
Tools/Scripts/run-webkit-tests -1 --debug --repeat-each=30 -g http/tests/notifications/event-listener-crash.html
No new tests, already covered by:
http/tests/notifications/event-listener-crash.html
- Modules/notifications/NotificationCenter.cpp:
(WebCore::NotificationCenter::stop):
- 3:49 AM Changeset in webkit [181243] by
-
- 3 edits2 adds in releases/WebKitGTK/webkit-2.8
Merge r181194 - Setting scroll-snap-desination to (100% 100%) locks up WebKit
https://bugs.webkit.org/show_bug.cgi?id=142414
<rdar://problem/20077275>
Reviewed by Dean Jackson.
Source/WebCore:
Tested by css3/scroll-snap/scroll-snap-desination-lock-up.html.
Correct an infinite loop that is triggered when you combine a repeating (100%)
scroll-snap-point-{x,y} along with a 100% scroll-snap-destination value.
- page/scrolling/AxisScrollSnapOffsets.cpp:
(WebCore::updateFromStyle): Make sure we break out of the loop properly when
the scroll-snap-point-{x,y} step is the same as the scroll-snap-destination.
LayoutTests:
- css3/scroll-snap/scroll-snap-desination-lock-up.html: Added.
- css3/scroll-snap/scroll-snap-desination-lock-up-expected.txt: Added.
- 3:27 AM Changeset in webkit [181242] by
-
- 11 edits5 adds5 deletes in trunk/LayoutTests
[iOS] Gardening to skip new tests, rebaseline existing tests
- platform/ios-simulator-wk1/TestExpectations:
- Mark test as flakey.
- platform/ios-simulator-wk2/fast/css/text-overflow-input-expected.txt: Remove.
- platform/ios-simulator-wk2/fast/forms/input-disabled-color-expected.txt: Remove.
- platform/ios-simulator-wk2/fast/forms/input-text-word-wrap-expected.txt: Remove.
- platform/ios-simulator-wk2/fast/forms/minWidthPercent-expected.txt: Remove.
- Superceded by new platform/ios-simulator results.
- platform/ios-simulator/TestExpectations:
- Skip two tests that will always time out on iOS.
- Mark two tests as ImageOnlyFailure that fail consistently.
- platform/ios-simulator/compositing/media-controls-bar-appearance-expected.txt: Add.
- Add result for new test from r180965.
- platform/ios-simulator/fast/attachment/attachment-label-highlight-expected.txt: Add.
- platform/ios-simulator/fast/attachment/attachment-progress-expected.txt: Add.
- platform/ios-simulator/fast/attachment/attachment-subtitle-expected.txt: Add.
- platform/ios-simulator/fast/attachment/attachment-title-expected.txt: Add.
- Add results for new attachment tests.
- platform/ios-simulator/fast/css/text-overflow-input-expected.txt: Update.
- platform/ios-simulator/fast/forms/input-disabled-color-expected.txt: Update.
- platform/ios-simulator/fast/forms/input-text-scroll-left-on-blur-expected.txt: Update.
- platform/ios-simulator/fast/forms/input-text-word-wrap-expected.txt: Update.
- platform/ios-simulator/fast/forms/minWidthPercent-expected.txt: Update.
- platform/ios-simulator/tables/mozilla/bugs/bug59354-expected.txt: Update.
- platform/ios-simulator/tables/mozilla/bugs/bug96334-expected.txt: Update.
- platform/ios-simulator/tables/mozilla/other/move_row-expected.txt: Update.
- Rebaseline after other bug fixes like r180815.
- 3:19 AM Changeset in webkit [181241] by
-
- 2 edits in releases/WebKitGTK/webkit-2.8/Tools
Merge r181159 - [GTK] Test /webkit2/WebKitWebView/sync-request-on-max-conns might fail after finished
https://bugs.webkit.org/show_bug.cgi?id=142385
Reviewed by Sergio Villar Senin.
Use stack allocated GMainLoopSources to make sure they are
cancelled automatically if the test finishes before they have
been processed.
- TestWebKitAPI/Tests/WebKit2Gtk/TestResources.cpp:
(testWebViewSyncRequestOnMaxConns):
- 3:18 AM Changeset in webkit [181240] by
-
- 3 edits in releases/WebKitGTK/webkit-2.8/Source/WebKit2
Merge r181138 - REGRESSION(r180924): ASSERTION FAILED: !from.isEmpty() in WebCore::TransformationMatrix::rectToRect
https://bugs.webkit.org/show_bug.cgi?id=142345
Reviewed by Martin Robinson.
This was caused by r180924 that postpones the creation of the
TextureMapper, which could cause that a layer has not yet a size
when TextureMapper::paint() is called. This patch moves the
creation of the TextureMapper to
LayerTreeHostGtk::setNativeSurfaceHandleForCompositing(), so that
it's created as soon as it's possible to create. This method is
called by the drawing area right after creating the
LayerTreeHostGtk if it already have a handler, or when the handle
is received from the UI process.
- WebProcess/WebPage/gtk/LayerTreeHostGtk.cpp:
(WebKit::LayerTreeHostGtk::initialize): Remove the
nsureTextureMapper call because at this point the layer context ID
is always 0, so it's impossible to create the TextureMapper.
(WebKit::LayerTreeHostGtk::compositeLayersToContext): Remove the
ensureTextureMapper call from here too, since at this point, if we
have a context, we should also have a TextureMapper. Add an ASSERT
right before using the TextureMapper.
(WebKit::LayerTreeHostGtk::setNativeSurfaceHandleForCompositing):
Create the TextureMapper here.
(WebKit::LayerTreeHostGtk::ensureTextureMapper): Deleted.
- WebProcess/WebPage/gtk/LayerTreeHostGtk.h:
- 3:15 AM Changeset in webkit [181239] by
-
- 2 edits in releases/WebKitGTK/webkit-2.8/Source/WebCore
Merge r181133 - Document::recalcStyle() shouldn't call viewportContentsChanged() if there is a pending layout
https://bugs.webkit.org/show_bug.cgi?id=142140
Reviewed by Darin Adler.
Stop calling FrameView::viewportContentsChanged() in Document::recalcStyle()
if there is a layout pending to avoid doing unncessary work. If there is a
layout pending, we don't need to do anything because viewportContentsChanged()
will be called after layout.
We only need to call FrameView::viewportContentsChanged() in
Document::recalcStyle() if a style recalc does not cause a layout. For e.g.
a '-webkit-transform' could make an animated GIF visible without causing a
layout, in which case we need to resume the animated GIF after style recalc.
No new tests, already covered by:
fast/images/animated-gif-webkit-transform.html
- dom/Document.cpp:
(WebCore::Document::recalcStyle):
- 3:13 AM Changeset in webkit [181238] by
-
- 2 edits in releases/WebKitGTK/webkit-2.8/Source/WebCore
Merge r181111 - Unreviewed, rolling out r180846.
https://bugs.webkit.org/show_bug.cgi?id=142368
Caused missing image banners in iTunes store pages (Requested
by smfr on #webkit).
Reverted changeset:
"FrameView::layoutTimerFired() should update style if needed
before doing layout"
https://bugs.webkit.org/show_bug.cgi?id=141688
http://trac.webkit.org/changeset/180846
- 3:12 AM Changeset in webkit [181237] by
-
- 3 edits in releases/WebKitGTK/webkit-2.8/Source/WebCore
Merge r181110 - Fix the build when NEON_INTRINSICS is enabled
https://bugs.webkit.org/show_bug.cgi?id=142361
Patch by Sandy Perez <sperez@indaba.es> on 2015-03-05
Reviewed by Csaba Osztrogonác.
- platform/graphics/cpu/arm/filters/FEBlendNEON.h:
(WebCore::FEBlend::platformApplySoftware):
- platform/graphics/filters/FEGaussianBlur.cpp:
(WebCore::standardBoxBlur):
- 3:11 AM Changeset in webkit [181236] by
-
- 11 edits in releases/WebKitGTK/webkit-2.8
Merge r181105 - Regression(r173761): ASSERTION FAILED: !is8Bit() in StringImpl::characters16()
https://bugs.webkit.org/show_bug.cgi?id=142350
Patch by Chris Dumez <Chris Dumez> on 2015-03-05
Reviewed by Michael Saboff and Benjamin Poulain.
Source/JavaScriptCore:
Call WTFString::hasInfixStartingAt() / hasInfixEndingAt() now that these
methods have been renamed for clarity.
- runtime/StringPrototype.cpp:
(JSC::stringProtoFuncStartsWith):
(JSC::stringProtoFuncEndsWith):
Source/WTF:
Fix ASSERTION FAILED: !is8Bit() in StringImpl::characters16() from
WTF::equalInner() after r173761. The code was incorrectly assuming that
if stringImpl is 16-bit, then matchString is 16-bit too, which is not
correct.
Also rename WTFString::startsWith() / endsWith() taking an offset to
hasInfixStartingAt() / hasInfixEndingAt() for clarity. It seems odd
to call it startsWith even though it won't technically *start* with
the pattern if the input offset is greater than zero.
Also drop the caseSensitive argument as it is never used (always true
at call sites.
- wtf/text/StringImpl.cpp:
(WTF::equalInner):
(WTF::StringImpl::hasInfixStartingAt):
(WTF::StringImpl::hasInfixEndingAt):
(WTF::StringImpl::startsWith): Deleted.
(WTF::StringImpl::endsWith): Deleted.
- wtf/text/StringImpl.h:
- wtf/text/WTFString.h:
(WTF::String::hasInfixStartingAt):
(WTF::String::hasInfixEndingAt):
(WTF::String::startsWith): Deleted.
(WTF::String::endsWith): Deleted.
Tools:
Add API test for WTFString::hasInfixStartingAt() to make sure it doesn't
crash if the string is 8-bit but the pattern is 16-bit (and vice-versa).
- TestWebKitAPI/Tests/WTF/WTFString.cpp:
(TestWebKitAPI::TEST):
LayoutTests:
Update String.startsWith() / endsWith() test to cover cases where the
input string is 8-bit and the pattern is 16-bit, and vice-versa.
- js/script-tests/string-includes.js:
- js/string-includes-expected.txt:
