Context Navigation

← Previous Period
Next Period →

Timeline

Sep 4, 2017:

9:55 PM Changeset in webkit [221608] by Chris Dumez

6 edits in trunk/Source/WebCore

Use StringView more in DOMFileSystem code
https://bugs.webkit.org/show_bug.cgi?id=176347

Reviewed by Darin Adler.

Modules/entriesapi/DOMFileSystem.cpp:

(WebCore::resolveRelativeVirtualPath):
(WebCore::DOMFileSystem::evaluatePath):

platform/FileSystem.h:
platform/glib/FileSystemGlib.cpp:

(WebCore::pathByAppendingComponents):

platform/posix/FileSystemPOSIX.cpp:

(WebCore::pathByAppendingComponents):

platform/win/FileSystemWin.cpp:

(WebCore::pathByAppendingComponents):

9:10 PM Changeset in webkit [221607] by sbarati@apple.com

21 edits
1 add in trunk

typeCheckHoistingPhase may emit a CheckStructure on the empty value which leads to a dereference of zero on 64 bit platforms
https://bugs.webkit.org/show_bug.cgi?id=176317

Reviewed by Keith Miller.

JSTests:

stress/dont-crash-when-hoist-check-structure-on-tdz.js: Added.

(Foo):

Source/JavaScriptCore:

It turns out that TypeCheckHoistingPhase may hoist a CheckStructure up to
the SetLocal of a particular value where the value is the empty JSValue.
On 64-bit platforms, the empty value is zero. This means that the empty value
passes a cell check. This will lead to a crash when we dereference null to load
the value's structure. This patch teaches TypeCheckHoistingPhase to be conservative
in the structure checks it hoists. On 64-bit platforms, instead of emitting a
CheckStructure node, we now emit a CheckStructureOrEmpty node. This node allows
the empty value to flow through. If the value isn't empty, it'll perform the normal
structure check that CheckStructure performs. For now, we only emit CheckStructureOrEmpty
on 64-bit platforms since a cell check on 32-bit platforms does not allow the empty
value to flow through.

dfg/DFGAbstractInterpreterInlines.h:

(JSC::DFG::AbstractInterpreter<AbstractStateType>::executeEffects):

dfg/DFGArgumentsEliminationPhase.cpp:
dfg/DFGClobberize.h:

(JSC::DFG::clobberize):

dfg/DFGConstantFoldingPhase.cpp:

(JSC::DFG::ConstantFoldingPhase::foldConstants):

dfg/DFGDoesGC.cpp:

(JSC::DFG::doesGC):

dfg/DFGFixupPhase.cpp:

(JSC::DFG::FixupPhase::fixupNode):

dfg/DFGNode.h:

(JSC::DFG::Node::convertCheckStructureOrEmptyToCheckStructure):
(JSC::DFG::Node::hasStructureSet):

dfg/DFGNodeType.h:
dfg/DFGObjectAllocationSinkingPhase.cpp:
dfg/DFGPredictionPropagationPhase.cpp:
dfg/DFGSafeToExecute.h:

(JSC::DFG::SafeToExecuteEdge::SafeToExecuteEdge):
(JSC::DFG::SafeToExecuteEdge::operator()):
(JSC::DFG::SafeToExecuteEdge::maySeeEmptyChild):
(JSC::DFG::safeToExecute):

dfg/DFGSpeculativeJIT.cpp:

(JSC::DFG::SpeculativeJIT::emitStructureCheck):
(JSC::DFG::SpeculativeJIT::compileCheckStructure):

dfg/DFGSpeculativeJIT.h:
dfg/DFGSpeculativeJIT32_64.cpp:

(JSC::DFG::SpeculativeJIT::compile):

dfg/DFGSpeculativeJIT64.cpp:

(JSC::DFG::SpeculativeJIT::compile):

dfg/DFGTypeCheckHoistingPhase.cpp:

(JSC::DFG::TypeCheckHoistingPhase::run):

dfg/DFGValidate.cpp:
ftl/FTLCapabilities.cpp:

(JSC::FTL::canCompile):

ftl/FTLLowerDFGToB3.cpp:

(JSC::FTL::DFG::LowerDFGToB3::compileNode):
(JSC::FTL::DFG::LowerDFGToB3::compileCheckStructureOrEmpty):

8:44 PM Changeset in webkit [221606] by eric.carlson@apple.com

13 edits
1 add in trunk

Switch HTMLMediaElement to release logging
https://bugs.webkit.org/show_bug.cgi?id=176065

Reviewed by Jer Noble.

Source/WebCore:

dom/Document.cpp:

(WebCore::Document::privateBrowsingStateDidChange): Disable the logger when private browsing
mode is enabled.
(WebCore::Document::logger const):

dom/Document.h:

Convert debug-only logging to configurable release logging.

html/HTMLMediaElement.cpp:

(PAL::LogArgument<WebCore::URL>::toString): Logger template for URL that returns the url as a
String when only when the LOG_DISABLED build flag is not defined. Returns "[url]" when it is.

html/HTMLMediaElement.h:

Source/WTF:

wtf/MediaTime.cpp:

(WTF::MediaTime::dump const): Use toString.
(WTF::MediaTime::toString const): New.

Tools:

TestWebKitAPI/Tests/WebCore/Logging.cpp:

(TestWebKitAPI::LogObserver::level const):

8:39 PM Changeset in webkit [221605] by commit-queue@webkit.org

11 edits in trunk/Source/WebCore

[Canvas] Move non-standard functionality/aliases back to CanvasRenderingContext2D.idl
https://bugs.webkit.org/show_bug.cgi?id=176340

Patch by Sam Weinig <sam@webkit.org> on 2017-09-04
Reviewed by Darin Adler.

In r221598 (https://webkit.org/b/176276), the CanvasRenderingContext2D was split out into
separate interfaces to the match the spec and prepare for implementing OffscreenCanvas. While
doing that, related non-standard functionality and aliases were moved as well. In retrospect,
that doesn't really makes sense, as we don't want to add these legacy properties to any new
interfaces that happen to implement/use the broken out interfaces. Since they are specific
to CanvasRenderingContext2D, let's move them back.

html/canvas/CanvasCompositing.idl:
html/canvas/CanvasDrawImage.idl:
html/canvas/CanvasFillStrokeStyles.idl:
html/canvas/CanvasImageData.idl:
html/canvas/CanvasImageSmoothing.idl:
html/canvas/CanvasPathDrawingStyles.idl:
html/canvas/CanvasShadowStyles.idl:
html/canvas/CanvasRenderingContext2D.idl:

Move non-standard operation and attributes back to CanvasRenderingContext2D.idl.

html/canvas/CanvasRenderingContext2D.h:

html/canvas/CanvasRenderingContext2D.cpp:

(WebCore::CanvasRenderingContext2D::stringForCanvasFillRule): Deleted.

Remove unused function.

(WebCore::CanvasRenderingContext2D::setAlpha): Deleted.
(WebCore::CanvasRenderingContext2D::setCompositeOperation): Deleted.

Removed. Instead, the IDL uses [ImplementedAs] to forward to the
underlying setGlobalAlpha/setGlobalCompositeOperation. This reduces
code and makes it clear they are strict aliases.

8:31 PM Changeset in webkit [221604] by yoav@yoav.ws

3 edits in trunk/LayoutTests

[iOS Simulator] http/tests/preload/viewport/meta-viewport-link-headers.php is a flaky failure.
https://bugs.webkit.org/show_bug.cgi?id=172518

This unflakes the test by downloading a "timer" image and kicking off the timer that runs the
test only once that image was downloaded, while potentially waiting for the preloaded images to arrive,
up to 1 second.
That means that slow test environments would have a longer timer than today, avoiding flakiness.

Reviewed by Youenn Fablet.

http/tests/preload/viewport/meta-viewport-link-headers.php: Add a "timer" image, which load event starts the test.
platform/ios/TestExpectations: Unflake the test in expectations.

8:29 PM Changeset in webkit [221603] by Darin Adler

13 edits in trunk

Fix a few minor problems found while working toward removing unneeded calls to updateStyle
https://bugs.webkit.org/show_bug.cgi?id=176279

Reviewed by Antti Koivisto.

Source/WebCore:

dom/Document.cpp:

(WebCore::Document::updateTitleElement): Fix the algorithm so that an SVG title won't
ever become the title when the document element is not an SVG element. We had this wrong
before but we were passing the test because SVGTitleElement::insertedInto had a bug that
hid the problem.
(WebCore::Document::titleElementAdded): Pass a reference instead of a pointer.
(WebCore::Document::titleElementRemoved): Pass a reference instead of nullptr since we
want to know which element is being removed.

dom/Document.h: Update the argument name and type for updateTitleElement.

html/HTMLTitleElement.cpp:

(WebCore::HTMLTitleElement::insertedInto): Call titleElementAdded unconditionally.
The checks to see if the title element is connected to the document and not in the
shadow tree are now handled by the Document class.
(WebCore::HTMLTitleElement::removedFrom): Same for titleElementRemoved.

platform/mock/MockRealtimeVideoSource.cpp:

(WebCore::MockRealtimeVideoSource::applySize): Removed incorrect caching of FontCascade
objects; these need to be created each time we draw.
(WebCore::MockRealtimeVideoSource::drawText): Create all the fonts here. Since this is
a mock for testing, there is no real performance concern with doing this.

platform/mock/MockRealtimeVideoSource.h: Removed the FontCascade data members.

rendering/RenderTreeAsText.cpp:

(WebCore::updateLayoutIgnoringPendingStylesheetsIncludingSubframes): Added. Used below
to update layout more thoroughly.
(WebCore::externalRepresentation): Update layout of all descendant frames, not just
the top level frame.

svg/SVGTitleElement.cpp:

(WebCore::SVGTitleElement::insertedInto): Removed bogus firstChild check, bogus isSVGDocument
check, and unneeded isConnected check and call titleElementAdded unconditionally. The checks
are now handled by the Document class.
(WebCore::SVGTitleElement::removedFrom): Same for titleElementRemoved.

testing/Internals.cpp:

(WebCore::Internals::layerTreeAsText const): Added missing updateLayoutIgnorePendingStylesheets,
since dumping the layer tree without first doing a layout will yield stale results.
(WebCore::Internals::layerIDForElement): Added missing updateLayoutIgnorePendingStylesheets,
since using a renderer to check if a layer exists without first doing a layout will yield stale results.
(WebCore::Internals::setElementUsesDisplayListDrawing): Ditto.
(WebCore::Internals::setElementTracksDisplayListReplay): Ditto.
(WebCore::Internals::displayListForElement): Ditto.
(WebCore::Internals::replayDisplayListForElement): Ditto.
(WebCore::Internals::isSelectPopupVisible): Added missing updateLayoutIgnorePendingStylesheets,
since working with the RenderMenuList without first doing a layout will yield stale results.
Also removed unneeded strict handling of null renderer, instead just returning false.
(WebCore::Internals::isPluginUnavailabilityIndicatorObscured): Removed unnneeded check of
renderer class, since the HTMLPlugInElement::isReplacementObscured already handles that correctly.
(WebCore::Internals::pageOverlayLayerTreeAsText const): Use updateLayoutIgnorePendingStylesheets
instead of the normal updateLayout for consistency with the behavior of the other functions in this
file, and because we need a real answer here.
(WebCore::Internals::scrollSnapOffsets): Ditto. Also rearranged the code so the renderer check is
done after updating the renderer tree.

Tools:

WebKitTestRunner/InjectedBundle/mac/AccessibilityUIElementMac.mm:

(WTR::AccessibilityUIElement::setSelectedChild const): Handle null here. The old code instead
would trigger an Objective-C exception unnecessarily.

LayoutTests:

svg/hittest/svg-tooltip.svg: Made this test easier to run outside of WebKitTestRunnner,

by removing some overzealous checks.

8:21 PM Changeset in webkit [221602] by sbarati@apple.com

41 edits in trunk/Source/JavaScriptCore

Support compiling catch in the FTL
https://bugs.webkit.org/show_bug.cgi?id=175396

Reviewed by Filip Pizlo.

This patch implements op_catch in the FTL. It extends the DFG implementation
by supporting multiple entrypoints in DFG-SSA. This patch implements this
by introducing an EntrySwitch node. When converting to SSA, we introduce a new
root block with an EntrySwitch that has the previous DFG entrypoints as its
successors. By convention, we pick the zeroth entry point index to be the
op_enter entrypoint. Like in B3, in DFG-SSA, EntrySwitch just acts like a
switch over the entrypoint index argument. DFG::EntrySwitch in the FTL
simply lowers to B3::EntrySwitch. The EntrySwitch in the root block that
SSAConversion creates can not exit because we would both not know where to exit
to in the program: we would not have valid OSR exit state. This design also
mandates that anything we hoist above EntrySwitch in the new root block
can not exit since they also do not have valid OSR exit state.

This patch also adds a new metadata node named InitializeEntrypointArguments.
InitializeEntrypointArguments is a metadata node that initializes the flush format for
the arguments at a given entrypoint. For a given entrypoint index, this node
tells AI and OSRAvailabilityAnalysis what the flush format for each argument
is. This allows each individual entrypoint to have an independent set of
argument types. Currently, this won't happen in practice because ArgumentPosition
unifies flush formats, but this is an implementation detail we probably want
to modify in the future. SSAConversion will add InitializeEntrypointArguments
to the beginning of each of the original DFG entrypoint blocks.

This patch also adds the ability to specify custom prologue code generators in Air.
This allows the FTL to specify a custom prologue for catch entrypoints that
matches the op_catch OSR entry calling convention that the DFG uses. This way,
the baseline JIT code OSR enters into op_catch the same way both in the DFG
and the FTL. In the future, we can use this same mechanism to perform stack
overflow checks instead of using a patchpoint.

b3/air/AirCode.cpp:

(JSC::B3::Air::Code::isEntrypoint):
(JSC::B3::Air::Code::entrypointIndex):

b3/air/AirCode.h:

(JSC::B3::Air::Code::setPrologueForEntrypoint):
(JSC::B3::Air::Code::prologueGeneratorForEntrypoint):

b3/air/AirGenerate.cpp:

(JSC::B3::Air::generate):

dfg/DFGAbstractInterpreterInlines.h:

(JSC::DFG::AbstractInterpreter<AbstractStateType>::executeEffects):

dfg/DFGBasicBlock.h:
dfg/DFGByteCodeParser.cpp:

(JSC::DFG::ByteCodeParser::parseBlock):
(JSC::DFG::ByteCodeParser::parse):

dfg/DFGCFG.h:

(JSC::DFG::selectCFG):

dfg/DFGClobberize.h:

(JSC::DFG::clobberize):

dfg/DFGClobbersExitState.cpp:

(JSC::DFG::clobbersExitState):

dfg/DFGCommonData.cpp:

(JSC::DFG::CommonData::shrinkToFit):
(JSC::DFG::CommonData::finalizeCatchEntrypoints):

dfg/DFGCommonData.h:

(JSC::DFG::CommonData::catchOSREntryDataForBytecodeIndex):
(JSC::DFG::CommonData::appendCatchEntrypoint):

dfg/DFGDoesGC.cpp:

(JSC::DFG::doesGC):

dfg/DFGFixupPhase.cpp:

(JSC::DFG::FixupPhase::fixupNode):

dfg/DFGGraph.cpp:

(JSC::DFG::Graph::dump):
(JSC::DFG::Graph::invalidateCFG):
(JSC::DFG::Graph::ensureCPSCFG):
(JSC::DFG::Graph::methodOfGettingAValueProfileFor):

dfg/DFGGraph.h:

(JSC::DFG::Graph::isEntrypoint):

dfg/DFGInPlaceAbstractState.cpp:

(JSC::DFG::InPlaceAbstractState::initialize):
(JSC::DFG::InPlaceAbstractState::mergeToSuccessors):

dfg/DFGJITCode.cpp:

(JSC::DFG::JITCode::shrinkToFit):
(JSC::DFG::JITCode::finalizeOSREntrypoints):

dfg/DFGJITCode.h:

(JSC::DFG::JITCode::catchOSREntryDataForBytecodeIndex): Deleted.
(JSC::DFG::JITCode::appendCatchEntrypoint): Deleted.

dfg/DFGJITCompiler.cpp:

(JSC::DFG::JITCompiler::noticeCatchEntrypoint):
(JSC::DFG::JITCompiler::makeCatchOSREntryBuffer):

dfg/DFGMayExit.cpp:
dfg/DFGNode.h:

(JSC::DFG::Node::isEntrySwitch):
(JSC::DFG::Node::isTerminal):
(JSC::DFG::Node::entrySwitchData):
(JSC::DFG::Node::numSuccessors):
(JSC::DFG::Node::successor):
(JSC::DFG::Node::entrypointIndex):

dfg/DFGNodeType.h:
dfg/DFGOSRAvailabilityAnalysisPhase.cpp:

(JSC::DFG::OSRAvailabilityAnalysisPhase::run):
(JSC::DFG::LocalOSRAvailabilityCalculator::executeNode):

dfg/DFGOSREntry.cpp:

(JSC::DFG::prepareCatchOSREntry):

dfg/DFGOSREntry.h:
dfg/DFGOSREntrypointCreationPhase.cpp:

(JSC::DFG::OSREntrypointCreationPhase::run):

dfg/DFGPredictionPropagationPhase.cpp:
dfg/DFGSSAConversionPhase.cpp:

(JSC::DFG::SSAConversionPhase::SSAConversionPhase):
(JSC::DFG::SSAConversionPhase::run):

dfg/DFGSafeToExecute.h:

(JSC::DFG::safeToExecute):

dfg/DFGSpeculativeJIT.cpp:

(JSC::DFG::SpeculativeJIT::linkOSREntries):

dfg/DFGSpeculativeJIT32_64.cpp:

(JSC::DFG::SpeculativeJIT::compile):

dfg/DFGSpeculativeJIT64.cpp:

(JSC::DFG::SpeculativeJIT::compile):

dfg/DFGStaticExecutionCountEstimationPhase.cpp:

(JSC::DFG::StaticExecutionCountEstimationPhase::run):

dfg/DFGValidate.cpp:
ftl/FTLCapabilities.cpp:

(JSC::FTL::canCompile):

ftl/FTLCompile.cpp:

(JSC::FTL::compile):

ftl/FTLLowerDFGToB3.cpp:

(JSC::FTL::DFG::LowerDFGToB3::lower):
(JSC::FTL::DFG::LowerDFGToB3::compileNode):
(JSC::FTL::DFG::LowerDFGToB3::compileExtractCatchLocal):
(JSC::FTL::DFG::LowerDFGToB3::compileGetStack):
(JSC::FTL::DFG::LowerDFGToB3::compileEntrySwitch):
(JSC::FTL::DFG::LowerDFGToB3::speculate):
(JSC::FTL::DFG::LowerDFGToB3::appendOSRExitDescriptor):
(JSC::FTL::DFG::LowerDFGToB3::appendOSRExit):
(JSC::FTL::DFG::LowerDFGToB3::blessSpeculation):

ftl/FTLOutput.cpp:

(JSC::FTL::Output::entrySwitch):

ftl/FTLOutput.h:
jit/JITOperations.cpp:

8:03 PM Changeset in webkit [221601] by Yusuke Suzuki

23 edits
10 adds in trunk

[DFG][FTL] Efficiently execute number#toString()
https://bugs.webkit.org/show_bug.cgi?id=170007

Reviewed by Keith Miller.

JSTests:

microbenchmarks/number-to-string-strength-reduction.js: Added.

(test):

microbenchmarks/number-to-string-with-radix-10.js: Added.

(test):

microbenchmarks/number-to-string-with-radix-cse.js: Added.

(test):

microbenchmarks/number-to-string-with-radix.js: Added.

(test):

stress/number-to-string-strength-reduction.js: Added.

(shouldBe):
(test):

stress/number-to-string-with-radix-10.js: Added.

(shouldBe):
(test):

stress/number-to-string-with-radix-cse.js: Added.

(shouldBe):
(test):

stress/number-to-string-with-radix-invalid.js: Added.

(shouldThrow):

stress/number-to-string-with-radix-watchpoint.js: Added.

(shouldBe):
(test):
(i.i.1e3.Number.prototype.toString):

stress/number-to-string-with-radix.js: Added.

(shouldBe):
(test):

Source/JavaScriptCore:

In JS, the natural way to convert number to string with radix is number.toString(radix).
However, our IC only cares about cells. If the base value is a number, it always goes to the slow path.

While extending our IC for number and boolean, the most meaningful use of this IC is calling number.toString(radix).
So, in this patch, we first add a fast path for this in DFG by using watchpoint. We set up a watchpoint for
Number.prototype.toString. And if this watchpoint is kept alive and GetById(base, "toString")'s base should be
speculated as Number, we emit Number related Checks and convert GetById to Number.prototype.toString constant.
It removes costly GetById slow path, and makes it non-clobbering node (JSConstant).

In addition, we add NumberToStringWithValidRadixConstant node. We have NumberToStringWithRadix node, but it may
throw an error if the valid value is incorrect (for example, number.toString(2000)). So its clobbering rule is
conservatively use read(World)/write(Heap). But in reality, number.toString is mostly called with the constant
radix, and we can easily figure out this radix is valid (2 <= radix && radix < 32).
We add a rule to the constant folding phase to convert NumberToStringWithRadix to NumberToStringWithValidRadixConstant.
It ensures that it has valid constant radix. And we relax our clobbering rule for NumberToStringWithValidRadixConstant.

Added microbenchmarks show performance improvement.

baseline patched

number-to-string-with-radix-cse 43.8312+-1.3017 ^{7.4930+-0.5105} definitely 5.8496x faster
number-to-string-with-radix-10 7.2775+-0.5225 ^{2.1906+-0.1864} definitely 3.3222x faster
number-to-string-with-radix 39.7378+-1.4921 ^{16.6137+-0.7776} definitely 2.3919x faster
number-to-string-strength-reduction 94.9667+-2.7157 ^{9.3060+-0.7202} definitely 10.2049x faster

dfg/DFGAbstractInterpreterInlines.h:

(JSC::DFG::AbstractInterpreter<AbstractStateType>::executeEffects):

dfg/DFGClobberize.h:

(JSC::DFG::clobberize):

dfg/DFGConstantFoldingPhase.cpp:

(JSC::DFG::ConstantFoldingPhase::foldConstants):

dfg/DFGDoesGC.cpp:

(JSC::DFG::doesGC):

dfg/DFGFixupPhase.cpp:

(JSC::DFG::FixupPhase::fixupNode):

dfg/DFGGraph.h:

(JSC::DFG::Graph::isWatchingGlobalObjectWatchpoint):
(JSC::DFG::Graph::isWatchingArrayIteratorProtocolWatchpoint):
(JSC::DFG::Graph::isWatchingNumberToStringWatchpoint):

dfg/DFGNode.h:

(JSC::DFG::Node::convertToNumberToStringWithValidRadixConstant):
(JSC::DFG::Node::hasValidRadixConstant):
(JSC::DFG::Node::validRadixConstant):

dfg/DFGNodeType.h:
dfg/DFGPredictionPropagationPhase.cpp:
dfg/DFGSafeToExecute.h:

(JSC::DFG::safeToExecute):

dfg/DFGSpeculativeJIT.cpp:

(JSC::DFG::SpeculativeJIT::compileToStringOrCallStringConstructor):
(JSC::DFG::SpeculativeJIT::compileNumberToStringWithValidRadixConstant):
(JSC::DFG::SpeculativeJIT::compileToStringOrCallStringConstructorOnNumber): Deleted.

dfg/DFGSpeculativeJIT.h:
dfg/DFGSpeculativeJIT32_64.cpp:

(JSC::DFG::SpeculativeJIT::compile):

dfg/DFGSpeculativeJIT64.cpp:

(JSC::DFG::SpeculativeJIT::compile):

dfg/DFGStrengthReductionPhase.cpp:

(JSC::DFG::StrengthReductionPhase::handleNode):

ftl/FTLCapabilities.cpp:

(JSC::FTL::canCompile):

ftl/FTLLowerDFGToB3.cpp:

(JSC::FTL::DFG::LowerDFGToB3::compileNode):
(JSC::FTL::DFG::LowerDFGToB3::compileNumberToStringWithValidRadixConstant):

runtime/JSGlobalObject.cpp:

(JSC::JSGlobalObject::JSGlobalObject):
(JSC::JSGlobalObject::init):
(JSC::JSGlobalObject::visitChildren):

runtime/JSGlobalObject.h:

(JSC::JSGlobalObject::numberToStringWatchpoint):
(JSC::JSGlobalObject::numberProtoToStringFunction const):

runtime/NumberPrototype.cpp:

(JSC::NumberPrototype::finishCreation):
(JSC::toStringWithRadixInternal):
(JSC::toStringWithRadix):
(JSC::int32ToStringInternal):
(JSC::numberToStringInternal):

runtime/NumberPrototype.h:

6:48 PM Changeset in webkit [221600] by Yusuke Suzuki

13 edits in trunk/Source

Remove OS(SOLARIS) support
https://bugs.webkit.org/show_bug.cgi?id=176341

Reviewed by Sam Weinig.

Source/WebCore:

bindings/scripts/preprocessor.pm:

(applyPreprocessor):

dom/make_names.pl:
inspector/InspectorFrontendHost.cpp:

(WebCore::InspectorFrontendHost::platform):

page/Page.h:

Source/WTF:

WebKit project does not have stake holders supporting Solaris right now.
And Solaris + SPARC in 64bit environment does not work well since its
address space includes non-48bit area. It breaks our JSVALUE64 in JSC.
In addition, Solaris a bit complicates our threading implementation
because of its special threading stack.

This patch removes OS(SOLARIS) and COMPILER(SUNCC) support from WebKit.

wtf/Compiler.h:
wtf/InlineASM.h:
wtf/MathExtras.h:

(std::isfinite): Deleted.
(std::signbit): Deleted.
(std::isinf): Deleted.

wtf/NumberOfCores.cpp:

(WTF::numberOfProcessorCores):

wtf/Platform.h:
wtf/StackBounds.cpp:
wtf/ThreadingPthreads.cpp:

6:13 PM Changeset in webkit [221599] by Yusuke Suzuki

2 edits in trunk/Source/WebCore

Unreviewed, fixing unused constant for initial buffer size
https://bugs.webkit.org/show_bug.cgi?id=176310

Follow-up patch after r221583. Pointed by darin@.

platform/graphics/win/FontPlatformDataWin.cpp:

(WebCore::FontPlatformData::FontPlatformData):

5:23 PM Changeset in webkit [221598] by weinig@apple.com

39 edits
3 moves
29 adds in trunk

[Canvas] Split CanvasRenderingContext2D.idl into separate IDLs to match current HTML spec
https://bugs.webkit.org/show_bug.cgi?id=176276

Reviewed by Dean Jackson.

Source/WebCore:

Splits CanvasRenderingContext2D.idl out into:

CanvasState
CanvasTransform
CanvasCompositing
CanvasImageSmoothing
CanvasFillStrokeStyles
CanvasShadowStyles
CanvasFilters
CanvasRect
CanvasDrawPath
CanvasUserInterface
CanvasText
CanvasDrawImage
CanvasImageData
CanvasPathDrawingStyles
CanvasTextDrawingStyles

Renames CanvasWindingRule enum to CanvasFillRule, and moves it to its own file so it can be reused by multiple IDLs.
Renames DOMPath to Path2D, which is what it is supposed to be called, and doesn't conflict with any WebCore names.
Converts lineCap, lineJoin, textAlign, textBaseline, direction to use enums, rather than DOMStrings, allowing us to remove a bunch of custom parsing.
Non-standard canvas API was grouped with related functionality in the broken out IDLs and clearly marked as non-standard.
Comments were added for places where we break with the standard, mostly around use of float where we should be using double.

CMakeLists.txt:
DerivedSources.cpp:
DerivedSources.make:
WebCore.xcodeproj/project.pbxproj:

Add new files.

bindings/js/CallTracerTypes.h:
inspector/InspectorCanvas.cpp:

(WebCore::InspectorCanvas::buildInitialState):
(WebCore::InspectorCanvas::buildAction):

Update for renames, and use new convertEnumerationToString to
serialize enumerations.

bindings/scripts/CodeGeneratorJS.pm:

(GenerateEnumerationImplementationContent):
(GenerateEnumerationHeaderContent):

Split string conversion out of convertEnumerationToJS, and into
its own function, convertEnumerationToString, to allow serialization
to string outside of the bindings context, in this case, for the
InspectorCanvas.

(GenerateParametersCheck):

Improve error message to make it clear which operation has the issue
in order to make debugging easier.

html/ImageData.idl:

Re-order constructors to match specification, and add FIXME for incorrect
default argument.

html/TextMetrics.idl:

Add spec'd exposure and spec comments, as well a FIXME to convert the
unrestricted float parameters should be doubles.

html/canvas/CanvasCompositing.idl: Added.
html/canvas/CanvasDirection.h: Added.
html/canvas/CanvasDirection.idl: Added.
html/canvas/CanvasDrawImage.idl: Added.
html/canvas/CanvasDrawPath.idl: Added.
html/canvas/CanvasFillRule.h: Added.
html/canvas/CanvasFillRule.idl: Added.
html/canvas/CanvasFillStrokeStyles.idl: Added.
html/canvas/CanvasFilters.idl: Added.
html/canvas/CanvasImageData.idl: Added.
html/canvas/CanvasImageSmoothing.idl: Added.
html/canvas/CanvasLineCap.h: Added.
html/canvas/CanvasLineCap.idl: Added.
html/canvas/CanvasLineJoin.h: Added.
html/canvas/CanvasLineJoin.idl: Added.
html/canvas/CanvasPathDrawingStyles.idl: Added.
html/canvas/CanvasRect.idl: Added.
html/canvas/CanvasShadowStyles.idl: Added.
html/canvas/CanvasState.idl: Added.
html/canvas/CanvasText.idl: Added.
html/canvas/CanvasTextAlign.h: Added.
html/canvas/CanvasTextAlign.idl: Added.
html/canvas/CanvasTextBaseline.h: Added.
html/canvas/CanvasTextBaseline.idl: Added.
html/canvas/CanvasTextDrawingStyles.idl: Added.
html/canvas/CanvasTransform.idl: Added.
html/canvas/CanvasUserInterface.idl: Added.
html/canvas/ImageSmoothingQuality.h: Added.
html/canvas/ImageSmoothingQuality.idl: Added.

Move CanvasRenderingContext2D.idl content into separate files to
match the specification.

html/canvas/CanvasGradient.idl:
html/canvas/CanvasPath.idl:
html/canvas/CanvasPattern.idl:

Add exposure and FIXMEs on changes need to match the spec.

html/canvas/CanvasRenderingContext2D.cpp:
html/canvas/CanvasRenderingContext2D.h:

Switch to using enums for lineCap, lineJoin, textAlign, textBaseline, and direction.
Rather than using the parsing / stringifying functions from GraphicsTypes, add simple
switch based to/from converters to convert to/from the canvas type to the platform
type. Also update for rename of DOMPath -> Path2D and WindingRule -> CanvasFillRule.

Unfortunately, not all custom parsing could be removed. To support the legacy functions
setLineCap and setLineJoin, which take Strings, two simply parse paths had to be
brought back. They can't use the enums, because the rules for enums as parameters
is stricter that what was implemented in the custom code (e.g. the bindings would
have thrown for strings not in the enum set).

html/canvas/CanvasRenderingContext2D.idl:

Move most of the functions into the new files, which are then 'implemented' by
this.

html/canvas/DOMPath.cpp: Removed.
html/canvas/DOMPath.h: Removed.
html/canvas/DOMPath.idl: Removed.
html/canvas/Path2D.cpp: Copied from Source/WebCore/html/canvas/DOMPath.cpp.
html/canvas/Path2D.h: Copied from Source/WebCore/html/canvas/DOMPath.h.
html/canvas/Path2D.idl: Copied from Source/WebCore/html/canvas/DOMPath.idl.

Rename DOMPath to Path2D to match the spec..

platform/graphics/GraphicsTypes.cpp:

(WebCore::parseLineCap): Deleted.
(WebCore::lineCapName): Deleted.
(WebCore::parseLineJoin): Deleted.
(WebCore::lineJoinName): Deleted.
(WebCore::textAlignName): Deleted.
(WebCore::parseTextAlign): Deleted.
(WebCore::textBaselineName): Deleted.
(WebCore::parseTextBaseline): Deleted.

platform/graphics/GraphicsTypes.h:

Remove now unused parsing / stringifying functions.

testing/Internals.cpp:

Remove unused #include.

bindings/scripts/test/JS/JSTestCallbackInterface.cpp:
bindings/scripts/test/JS/JSTestCallbackInterface.h:
bindings/scripts/test/JS/JSTestObj.cpp:
bindings/scripts/test/JS/JSTestObj.h:
bindings/scripts/test/JS/JSTestStandaloneDictionary.cpp:
bindings/scripts/test/JS/JSTestStandaloneDictionary.h:
bindings/scripts/test/JS/JSTestStandaloneEnumeration.cpp:
bindings/scripts/test/JS/JSTestStandaloneEnumeration.h:

Update test results to add convertEnumerationToString.

LayoutTests:

fast/canvas/canvas-clip-path-expected.txt:
fast/canvas/canvas-fill-path-expected.txt:
fast/canvas/canvas-path-addPath-expected.txt:
fast/canvas/canvas-path-isPointInPath-expected.txt:
fast/canvas/canvas-path-isPointInStroke-expected.txt:
fast/canvas/canvas-stroke-path-expected.txt:
fast/canvas/winding-enumeration-expected.txt:
inspector/canvas/recording-2d-expected.txt:
js/dom/global-constructors-attributes-dedicated-worker-expected.txt:

Update results.

5:11 PM Changeset in webkit [221597] by Yusuke Suzuki

2 edits in trunk/Source/JavaScriptCore

[DFG] Consider increasing the number of DFG worklist threads
https://bugs.webkit.org/show_bug.cgi?id=176222

Reviewed by Saam Barati.

Attempt to add one more thread to DFG worklist. DFG compiler sometimes takes
very long time if the target function is very large. However, DFG worklist
has only one thread before this patch. Therefore, one function that takes
too much time to be compiled can prevent the other functions from being
compiled in DFG or upper tiers.

One example is Octane/zlib. In zlib, compiling "a1" function in DFG takes
super long time (447 ms) because of its super large size of the function.
While this function never gets compiled in FTL due to its large size,
it can be compiled in DFG and takes super long time. Subsequent "a8" function
compilation in DFG is blocked by this "a1". As a consequence, the benchmark
takes very long time in a1/Baseline code, which is slower than DFG of course.

While FTL has a bit more threads, DFG worklist has only one thread. This patch
adds one more thread to DFG worklist to alleviate the above situation. This
change significantly improves Octane/zlib performance.

baseline patched

zlib x2 482.32825+-6.07640 ^{408.66072+-14.03856} definitely 1.1803x faster

runtime/Options.h:

4:50 PM Changeset in webkit [221596] by Darin Adler

3 edits in trunk/Source/WebCore

Follow up FrameView::updateLayoutAndStyleIfNeededRecursive changes with related improvements
https://bugs.webkit.org/show_bug.cgi?id=176277

Reviewed by Antti Koivisto.

page/FrameView.cpp:

(WebCore::FrameView::needsStyleRecalcOrLayout): Deleted. This function was only used
by an assertion inside updateLayoutAndStyleIfNeededRecursive, and thus there is no reason
for it to be in the header file, or for it to be a public member function.
(WebCore::appendRenderedChildren): Deleted. This function was only used inside
updateLayoutAndStyleIfNeededRecursive, and it is now packaged in an even better way
for efficient use inside that function.
(WebCore::FrameView::renderedChildFrameViews): Deleted. This function was only used
inside needsStyleRecalcOrLayout, and it's now packaged in a better way inside that function.
(WebCore::FrameView::updateLayoutAndStyleIfNeededRecursive): Created a new lambda called
nextRendereredDescendant that packages up the process of repeatedly iterating this view
and all of its descendants in an easy-to-use way. Replaces both of the functions above.
Rewrote to use it; it made the logic clear enough that it was good to get rid of the
updateOneFrame lambda, too. Added two separate functions, one that checks for needed
style recalculation and a separate one that checked for needed layout. Using those,
replaced the old single assertion with two separate assertions.

page/FrameView.h: Removed needsStyleRecalcOrLayout, renderedChildFrameViews, and

FrameViewList.

4:36 PM Changeset in webkit [221595] by Wenson Hsieh

14 edits
3 adds in trunk

[iOS DnD] Refactor drag and drop logic in WKContentView in preparation for supporting multiple drag items in a drag session
https://bugs.webkit.org/show_bug.cgi?id=176264
<rdar://problem/31144674>

Reviewed by Darin Adler.

Source/WebCore:

Makes some small adjustments to WebItemProviderPasteboard. Rather than just -setItemProviders: on the
WebItemProviderPasteboard when initiating a drag, also stage the WebItemProviderRegistrationList that will be
used to generate an item provider. While it would be cleaner to avoid directly setting item providers so we
don't overwrite item providers on the WebItemProviderPasteboard when adding items to an existing drag session,
this isn't possible without breaking binary compability with older UIKit versions.

Importantly, WKContentView will now ignore any item providers that have been set on the
WebItemProviderPasteboard when initiating a drag or adding items to an existing drag session, and instead only
consider the staged registration list when generating item providers for dragging. This only has the drawback of
generating an unnecessary item provider, but otherwise maintains backwards compatibility while allowing us to
provide WebKit2 support for multiple items per drag session.

Tests: DataInteractionTests.DragEventClientCoordinatesBasic

DataInteractionTests.DragEventClientCoordinatesWithScrollOffset
DataInteractionTests.DragEventPageCoordinatesBasic
DataInteractionTests.DragEventPageCoordinatesWithScrollOffset

platform/ios/AbstractPasteboard.h:
platform/ios/PlatformPasteboardIOS.mm:

(WebCore::registerItemToPasteboard):

Changed to only stage registration info on the item provider pasteboard, if possible. This has no effect on the
copy/paste codepath, since it uses a UIPasteboard.

platform/ios/WebItemProviderPasteboard.h:
platform/ios/WebItemProviderPasteboard.mm:

Rename _registrationInfoLists to _stagedRegistrationInfoList, and change it from an array of registration info
lists to a single registration info list. This could be updated in the future to be an array of registration
lists, but currently, it serves no purpose and makes coordination with DragItem info more difficult. This would
need to support multiple registration lists if we are to add a way to begin a drag containing multiple items in
vanilla web content, such as dragging multiple selections.

(-[WebItemProviderPasteboard init]):
(-[WebItemProviderPasteboard setItemProviders:]):
(-[WebItemProviderPasteboard stageRegistrationList:]):

Sets the staged item provider registration list.

(-[WebItemProviderPasteboard takeRegistrationList]):

Removes the staged item provider registration list from the WebItemProviderPasteboard and also returns it.

(-[WebItemProviderPasteboard registrationInfoAtIndex:]): Deleted.
(-[WebItemProviderPasteboard setRegistrationInfoLists:]): Deleted.

Source/WebKit:

Move DataInteractionState from WKContentViewInteraction.h to DragDropInteractionState.h, and also rename it to
DragDropInteractionState. Additionally, refactor drag and drop state in the UI process to capture metadata about
the dragged element in a separate DragSourceState struct. This patch also moves drag and drop state transition
logic that doesn't involve WKContentView internals out of WKContentView, and into the implementation of
DragDropInteractionState.

To support multiple drag items per session, we also introduce a simple mechanism to trace a UIDragItem back to
the DragSourceState used to generate it. When generating a DragSourceState, we assign it a unique identifier,
which we also set as the privateLocalContext of all UIDragItems generated when beginning the drag (this
includes drag items returned by an internal client that uses one of the SPI hooks to augment drag items when
starting a drag). This is subsequently used in the implementation of lift and cancellation preview delegate
methods to supply the appropriate drag preview for each UIDragItem.

Lastly, fix a bug wherein the pageX and pageY of mouse drag events are inconsistent with other synthetic mouse
events, such as synthetic clicks. For synthetic clicks, the PlatformMouseEvent is initialized with the same
position and globalPosition. Whether this is really intended is unclear (see http://webkit.org/b/173855), but
it's a trivial change for now to keep mouse events on iOS consistent by tweaking the behavior during drag and
drop. See Tools/ChangeLog for some more information.

Platform/spi/ios/UIKitSPI.h:

Add -[UIDragItem privateLocalContext].

UIProcess/ios/DragDropInteractionState.h: Added.

(WebKit::DragDropInteractionState::stagedDragSource const):
(WebKit::DragDropInteractionState::dropSessionDidExit):
(WebKit::DragDropInteractionState::dropSessionWillPerformDrop):
(WebKit::DragDropInteractionState::adjustedPositionForDragEnd const):
(WebKit::DragDropInteractionState::didBeginDragging const):
(WebKit::DragDropInteractionState::isPerformingDrop const):
(WebKit::DragDropInteractionState::dragSession const):
(WebKit::DragDropInteractionState::dropSession const):

Wrap private drag/drop state member variables behind const getters, and move drag and drop logic that involves
only the DragDropInteractionState into helper methods on DragDropInteractionState.

(WebKit::DragDropInteractionState::BlockPtr<void):

UIProcess/ios/DragDropInteractionState.mm: Added.

(WebKit::dragItemMatchingIdentifier):
(WebKit::createTargetedDragPreview):
(WebKit::uiImageForImage):

Move drag preview creation logic here, from WKContentViewInteraction.mm.

(WebKit::shouldUseTextIndicatorToCreatePreviewForDragAction):
(WebKit::DragDropInteractionState::activeDragSourceForItem const):
(WebKit::DragDropInteractionState::anyActiveDragSourceIs const):
(WebKit::DragDropInteractionState::prepareForDragSession):
(WebKit::DragDropInteractionState::dragSessionWillBegin):
(WebKit::DragDropInteractionState::previewForDragItem const):
(WebKit::DragDropInteractionState::dragSessionWillDelaySetDownAnimation):
(WebKit::DragDropInteractionState::dropSessionDidEnterOrUpdate):
(WebKit::DragDropInteractionState::stageDragItem):
(WebKit::DragDropInteractionState::hasStagedDragSource const):
(WebKit::DragDropInteractionState::clearStagedDragSource):
(WebKit::DragDropInteractionState::dragAndDropSessionsDidEnd):
(WebKit::DragDropInteractionState::updatePreviewsForActiveDragSources):

UIProcess/ios/WKContentViewInteraction.h:

Move drag-and-drop-related state tied to the WKContentView here, from DataInteractionState (for instance, the
UIView for the drop caret, the WKContentView snapshot when dropping, and a flag use to keep track of hiding the
callout bar when dragging a text selection).

(): Deleted.

UIProcess/ios/WKContentViewInteraction.mm:

(-[WKContentView _startAssistingNode:userIsInteracting:blurPreviousNode:userObject:]):
(-[WKContentView actionSheetAssistant:showCustomSheetForElement:]):
(-[WKContentView _didChangeDragInteractionPolicy]):
(-[WKContentView setupDataInteractionDelegates]):
(-[WKContentView teardownDataInteractionDelegates]):
(-[WKContentView _startDrag:item:]):
(-[WKContentView _didHandleStartDataInteractionRequest:]):
(-[WKContentView computeClientAndGlobalPointsForDropSession:outClientPoint:outGlobalPoint:]):
(-[WKContentView cleanUpDragSourceSessionState]):
(-[WKContentView _didConcludeEditDataInteraction:]):
(-[WKContentView _didPerformDataInteractionControllerOperation:]):
(-[WKContentView _didChangeDataInteractionCaretRect:currentRect:]):
(-[WKContentView currentDragOrDropSession]):
(-[WKContentView _restoreCalloutBarIfNeeded]):
(-[WKContentView _dragInteraction:prepareForSession:completion:]):
(-[WKContentView dragInteraction:itemsForBeginningSession:]):
(-[WKContentView dragInteraction:previewForLiftingItem:session:]):
(-[WKContentView dragInteraction:willAnimateLiftWithAnimator:session:]):
(-[WKContentView dragInteraction:sessionWillBegin:]):
(-[WKContentView dragInteraction:session:didEndWithOperation:]):
(-[WKContentView dragInteraction:previewForCancellingItem:withDefault:]):
(-[WKContentView _dragInteraction:item:shouldDelaySetDownAnimationWithCompletion:]):
(-[WKContentView dragInteraction:item:willAnimateCancelWithAnimator:]):
(-[WKContentView dropInteraction:sessionDidEnter:]):
(-[WKContentView dropInteraction:sessionDidUpdate:]):
(-[WKContentView dropInteraction:sessionDidExit:]):
(-[WKContentView dropInteraction:performDrop:]):
(-[WKContentView dropInteraction:sessionDidEnd:]):

Pull out logic that mutates drag and drop state into DragDropInteractionState. Adjust places that previously
accessed DataInteractionState's members directly with corresponding getters in DragDropInteractionState.

(-[WKContentView _simulateDataInteractionEntered:]):
(-[WKContentView _simulateDataInteractionUpdated:]):
(-[WKContentView _simulateDataInteractionEnded:]):
(-[WKContentView _simulateDataInteractionPerformOperation:]):
(-[WKContentView _simulateDataInteractionSessionDidEnd:]):
(-[WKContentView _simulateWillBeginDataInteractionWithSession:]):
(-[WKContentView _simulatedItemsForSession:]):
(-[WKContentView _simulatePrepareForDataInteractionSession:completion:]):

Rename _dataInteraction => _dragInteraction and _dataOperation => _dropInteraction.

(uiImageForImage): Deleted.
(shouldUseTextIndicatorToCreatePreviewForDragAction): Deleted.
(-[WKContentView dragPreviewForImage:frameInRootViewCoordinates:clippingRectsInFrameCoordinates:backgroundColor:]): Deleted.
(-[WKContentView dragPreviewForCurrentDataInteractionState]): Deleted.
(-[WKContentView _transitionDragPreviewToImageIfNecessary:]): Deleted.

WebKit.xcodeproj/project.pbxproj:

Tools:

Adds two new iOS drag and drop tests to check that the clientX and clientY attributes of mouse events propagated
to the page during drop are correct. Each test drags from an image element and drops into three custom-drop-
handling elements; dragenter, dragover, and drop event listeners on the body then use the clientX and
clientY event attributes to hit-test for drop target elements. The first test is suffixed with "-Basic"; the
second test, suffixed with "-WithScrollOffset", makes the document scrollable to check that clientY is correct
when scrolled.

TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
TestWebKitAPI/Tests/WebKitCocoa/drop-targets.html: Added.
TestWebKitAPI/Tests/ios/DataInteractionTests.mm:

(TestWebKitAPI::testDragAndDropOntoTargetElements):
(TestWebKitAPI::TEST):

TestWebKitAPI/cocoa/TestWKWebView.mm:

(-[TestWKWebView stringByEvaluatingJavaScript:]):

Log a warning message when an API test fails due to JavaScript evaluation in TestWKWebView.

1:24 PM Changeset in webkit [221594] by Nikita Vasilyev

5 edits
1 copy
1 move
4 adds
1 delete in trunk/Source/WebInspectorUI

Web Inspector: Styles Redesign: hook up real data to spreadsheet style editor
https://bugs.webkit.org/show_bug.cgi?id=175343
<rdar://problem/33784793>

Reviewed by Devin Rousso.

Replace static HTML added in <https://webkit.org/b/174838> with actual data.

Addressed in this patch:

CSS rules can be copied.
Source links work as expected.
Nonmatching selectors are grayed out.

Known limitations:

Neither CSS selectors nor CSS properties are editable.
"Inherited from" and media query section headers are missing.
No syntax highlighting of complex CSS values, e.g. hsl(0, 0%, 70%).

UserInterface/Main.html:
UserInterface/Models/CSSStyleDeclaration.js:
UserInterface/Views/CSSStyleDeclarationTextEditor.js:

(WI.CSSStyleDeclarationTextEditor):

UserInterface/Views/CSSStyleDetailsSidebarPanel.js:

(WI.CSSStyleDetailsSidebarPanel):

UserInterface/Views/RulesStyleSpreadsheetDetailsPanel.js: Removed.

Rename RulesStyleSpreadsheetDetailsPanel to SpreadsheetRulesStyleDetailsPanel. All files related to the spreadsheet editor start with "Spreadsheet" now.

UserInterface/Views/SpreadsheetCSSStyleDeclarationEditor.css: Added.

(.spreadsheet-style-declaration-editor):
(.spreadsheet-style-declaration-editor .name):
(.spreadsheet-style-declaration-editor .value):
(.spreadsheet-style-declaration-editor.no-properties):

UserInterface/Views/SpreadsheetCSSStyleDeclarationEditor.js: Added.

(WI.SpreadsheetCSSStyleDeclarationEditor):
(WI.SpreadsheetCSSStyleDeclarationEditor.prototype.layout):
(WI.SpreadsheetCSSStyleDeclarationEditor.prototype.get delegate):
(WI.SpreadsheetCSSStyleDeclarationEditor.prototype.set delegate):
(WI.SpreadsheetCSSStyleDeclarationEditor.prototype.get style):
(WI.SpreadsheetCSSStyleDeclarationEditor.prototype.set style):
(WI.SpreadsheetCSSStyleDeclarationEditor.prototype.get _propertiesToRender):
Similar condition is commonly used in CSSStyleDeclarationTextEditor. Abstact it out to its own method.

(WI.SpreadsheetCSSStyleDeclarationEditor.prototype._renderProperty):

UserInterface/Views/SpreadsheetCSSStyleDeclarationSection.css: Copied from Source/WebInspectorUI/UserInterface/Views/RulesStyleSpreadsheetDetailsPanel.css.

(.spreadsheet-css-declaration):
(.spreadsheet-css-declaration .origin):
(.spreadsheet-css-declaration.locked .origin::after):
(.spreadsheet-css-declaration .origin .go-to-link):
(.spreadsheet-css-declaration .origin .go-to-link:hover):
(.spreadsheet-css-declaration .styles-source):
(.spreadsheet-css-declaration .selector.style-attribute):
(.spreadsheet-css-declaration .selector > span):
(.spreadsheet-css-declaration .selector > .matched):
(.spreadsheet-css-declaration .properties):
(.spreadsheet-css-declaration.locked):
(.spreadsheet-css-declaration .locked-icon):

UserInterface/Views/SpreadsheetCSSStyleDeclarationSection.js: Added.

(WI.SpreadsheetCSSStyleDeclarationSection):
(WI.SpreadsheetCSSStyleDeclarationSection.prototype.get element):
(WI.SpreadsheetCSSStyleDeclarationSection.prototype.get style):
(WI.SpreadsheetCSSStyleDeclarationSection.prototype.initialLayout):
(WI.SpreadsheetCSSStyleDeclarationSection.prototype.layout):
(WI.SpreadsheetCSSStyleDeclarationSection.prototype.cssStyleDeclarationTextEditorFocused):
(WI.SpreadsheetCSSStyleDeclarationSection.prototype.get locked):
(WI.SpreadsheetCSSStyleDeclarationSection.prototype.get selectorEditable):

UserInterface/Views/SpreadsheetRulesStyleDetailsPanel.css: Renamed from Source/WebInspectorUI/UserInterface/Views/RulesStyleSpreadsheetDetailsPanel.css.
UserInterface/Views/SpreadsheetRulesStyleDetailsPanel.js: Added.

(WI.SpreadsheetRulesStyleDetailsPanel):
(WI.SpreadsheetRulesStyleDetailsPanel.prototype.refresh):

UserInterface/Views/StyleDetailsPanel.js:

(WI.StyleDetailsPanel.prototype.filterDidChange):

UserInterface/Views/View.js:

(WI.View.prototype.removeAllSubviews):
Add a method to remove all subviews more efficiently.

1:15 PM Changeset in webkit [221593] by fpizlo@apple.com

15 edits
1 add in trunk/Tools

WSL IntLiteral should have variable type so that it can unify with things like uint
https://bugs.webkit.org/show_bug.cgi?id=176209

Reviewed by Mylex Maxfield.

This introduces IntLiteralType, which can unify with any number type that can represent the
int literal.

This means that all of the following are legal:

double x = 0; 0 unifies with double
double y = x + 5; should resolve to double+double
double z = array[0]; 0 unifies with uint

We can do similar things with float literals, so that you don't have to suffix "f" all the
time.

I think that this is very different from what C does. C just supports implicit type
conversions. That's why "double x = 0" works in C. We don't do type conversions. Instead, we
let each literal unify itself with one type.

Also updated Test.html so that it runs again and is better at printing things.

WebGPUShadingLanguageRI/All.js:
WebGPUShadingLanguageRI/Checker.js:
WebGPUShadingLanguageRI/IntLiteral.js:

(IntLiteral):
(IntLiteral.prototype.toString):

WebGPUShadingLanguageRI/IntLiteralType.js: Added.

(IntLiteralType):
(IntLiteralType.prototype.get origin):
(IntLiteralType.prototype.get value):
(IntLiteralType.prototype.get isPrimitive):
(IntLiteralType.prototype.get isUnifiable):
(IntLiteralType.prototype.typeVariableUnify):
(IntLiteralType.prototype.unifyImpl):
(IntLiteralType.prototype.verifyAsArgument):
(IntLiteralType.prototype.verifyAsParameter):
(IntLiteralType.prototype.commitUnification):
(IntLiteralType.prototype.toString):

WebGPUShadingLanguageRI/Intrinsics.js:

(Intrinsics.):
(Intrinsics):

WebGPUShadingLanguageRI/LiteralTypeChecker.js:

(LiteralTypeChecker.prototype.visitIntLiteralType):
(LiteralTypeChecker):

WebGPUShadingLanguageRI/NativeType.js:

(NativeType):
(NativeType.prototype.get isNumber):
(NativeType.prototype.set isNumber):
(NativeType.prototype.get isInt):
(NativeType.prototype.set isInt):
(NativeType.prototype.get isFloat):
(NativeType.prototype.set isFloat):

WebGPUShadingLanguageRI/Prepare.js:

(prepare):

WebGPUShadingLanguageRI/Rewriter.js:

(Rewriter.prototype.visitIntLiteral):
(Rewriter.prototype.visitIntLiteralType):

WebGPUShadingLanguageRI/StandardLibrary.js:
WebGPUShadingLanguageRI/Test.html:
WebGPUShadingLanguageRI/Test.js:

(this.window.this.print):
(doPrep):
(doLex):
(checkNumber):
(checkInt):
(TEST_nameResolutionFailure):
(TEST_threadArrayLoadIntLiteral):
(TEST_deviceArrayStoreIntLiteral):
(TEST_returnIntLiteralDouble):

WebGPUShadingLanguageRI/Type.js:

(Type.prototype.get isNumber):
(Type.prototype.get isInt):
(Type.prototype.get isFloat):

WebGPUShadingLanguageRI/Visitor.js:

(Visitor.prototype.visitIntLiteral):
(Visitor.prototype.visitIntLiteralType):

1:14 PM Changeset in webkit [221592] by fpizlo@apple.com

7 edits in trunk/Tools

Add more tests for null behavior in WSL
https://bugs.webkit.org/show_bug.cgi?id=176318

Reviewed by Myles Maxfield.

This adds more tests for the unification behavior of null. In the process, I found a couple
small bugs and fixed them.

WebGPUShadingLanguageRI/Checker.js:
WebGPUShadingLanguageRI/FuncInstantiator.js:

(FuncInstantiator.prototype.getUnique.InstantiationSubstitution.prototype.visitCallExpression):
(FuncInstantiator.prototype.getUnique.InstantiationSubstitution):

WebGPUShadingLanguageRI/Substitution.js:

(Substitution.prototype.visitTypeRef):

WebGPUShadingLanguageRI/Test.js:

(TEST_passNullAndNotNullFullPoly):
(TEST_passNullAndNotNullFullPolyReverse):
(TEST_nullTypeVariableUnify.recurse):
(TEST_nullTypeVariableUnify.everyOrder):
(TEST_nullTypeVariableUnify.everyPair):
(TEST_nullTypeVariableUnify):

WebGPUShadingLanguageRI/TypeRef.js:

(TypeRef.wrap):

WebGPUShadingLanguageRI/TypeVariable.js:

(TypeVariable.prototype.verifyAsArgument):

12:34 PM Changeset in webkit [221591] by commit-queue@webkit.org

3 edits in trunk/LayoutTests

Loosen the check in video-canvas-drawing-output.html.
https://bugs.webkit.org/show_bug.cgi?id=127106

Patch by Ms2ger <Ms2ger@igalia.com> on 2017-09-04
Reviewed by Xabier Rodriguez-Calvar.

GStreamer needs the extra tolerance since version 1.2.1.

media/video-canvas-drawing-output.html:
platform/gtk/TestExpectations:

10:54 AM Changeset in webkit [221590] by commit-queue@webkit.org

13 edits
1 delete in trunk

Unreviewed, rolling out r221494 and r221500.
https://bugs.webkit.org/show_bug.cgi?id=176258

This caused the windows build to fail. (Requested by mlewis13
on #webkit).

Reverted changesets:

"Switch HTMLMediaElement to release logging"
https://bugs.webkit.org/show_bug.cgi?id=176065
http://trac.webkit.org/changeset/221494

"Switch HTMLMediaElement to release logging"
https://bugs.webkit.org/show_bug.cgi?id=176065
http://trac.webkit.org/changeset/221500

9:27 AM Changeset in webkit [221589] by Darin Adler

5 edits in trunk/Tools

REGRESSION (r220052): [Sierra] API test VideoControlsManager.VideoControlsManagerFullSizeVideoInWideMainFrame is a flaky failure
https://bugs.webkit.org/show_bug.cgi?id=175329

Reviewed by Wenson Hsieh.

These tests rely on using setTimeout to wait a run loop iteration for the
video controls to appear. With changes to how loading occurs, it can now
take two run loop iterations. The concept here is inherently a bit flaky
and racy -- end users will not be able to tell how many run loop cycles it
takes for controls to come up, and websites can't tell either -- but for now
changing the tests to wait one more run loop should get rid of the flakiness.

TestWebKitAPI/Tests/WebKitCocoa/VideoControlsManager.mm: Re-enable test.

TestWebKitAPI/Tests/WebKitCocoa/full-size-autoplaying-video-with-audio.html: Add a

second setTimeout and a comment.

TestWebKitAPI/Tests/WebKitCocoa/skinny-autoplaying-video-with-audio.html: Ditto.
TestWebKitAPI/Tests/WebKitCocoa/wide-autoplaying-video-with-audio.html: Ditto.

7:01 AM WebKitGTK/2.18.x edited by Adrian Perez de Castro

(diff)

7:01 AM Changeset in webkit [221588] by Adrian Perez de Castro

2 edits in trunk/Source/bmalloc

Unreviewed build fix for Clang with libc++

Fixes a build failure when building with Clang, -stdlib=libc++, and gigacage
support enabled, which resulted in "stderr" being undefined.

bmalloc/Gigacage.cpp: Add missing <ctsdio> include to pull the definition.

6:38 AM WebKitGTK/2.18.x edited by Carlos Garcia Campos

(diff)

6:38 AM WebKitGTK/2.18.x edited by Carlos Garcia Campos

(diff)

4:45 AM Changeset in webkit [221587] by Carlos Garcia Campos

1 copy in releases/WebKitGTK/webkit-2.17.92

WebKitGTK+ 2.17.92

4:44 AM Changeset in webkit [221586] by Carlos Garcia Campos

4 edits in releases/WebKitGTK/webkit-2.18

Unreviewed. Update OptionsGTK.cmake and NEWS for 2.17.92 release.

Source/cmake/OptionsGTK.cmake: Bump version numbers.

Source/WebKit:

gtk/NEWS: Add release notes for 2.17.92.

3:50 AM Changeset in webkit [221585] by commit-queue@webkit.org

10 edits
3 adds in trunk/Source

Source/JavaScriptCore:
[WebIDL] Unify and simplify EnableBySettings with the rest of the runtime settings
https://bugs.webkit.org/show_bug.cgi?id=176312

Patch by Sam Weinig <sam@webkit.org> on 2017-09-04
Reviewed by Darin Adler.

runtime/CommonIdentifiers.h:

Remove WebCore specific identifiers from CommonIdentifiers. They have been moved
to WebCoreBuiltinNames in WebCore.

Source/WebCore:
[WebIDL] Simplify EnabledBySettings generation
https://bugs.webkit.org/show_bug.cgi?id=176312

Patch by Sam Weinig <sam@webkit.org> on 2017-09-04
Reviewed by Darin Adler.

bindings/scripts/CodeGeneratorJS.pm:

(NeedsRuntimeCheck):
(OperationShouldBeOnInstance):
(GeneratePropertiesHashTable):
(GenerateRuntimeEnableConditionalString):
(GenerateImplementation):
(GenerateAttributeGetterBodyDefinition):

Simplify the [EnabledBySetting] code generation by merging it with the other runtime checks (Exposed, EnabledAtRuntime, EnabledForWorld, and SecureContext).

Switches to using WebCoreBuiltinNames for identifiers, rather than relying on the appropriate identifiers being available on JavaScriptCore's CommonIdentifiers which was an unnecessary layering violation. There are still problems with this as we continue to be inconsistent about which identifiers deserve to be in WebCoreBuiltinNames, and which ones can be created each time and there is no automation of WebCoreBuiltinNames from the IDLs.

Enforce at code generation time that [EnabledBySetting] can only be used on interfaces that are uniquely exposed to the window environment. This is a condition we should find a way to remove (e.g. propagating necessary settings to an object the ScriptExecutionContext owns), but for now, Settings is a main thread only object.

bindings/js/JSDOMWindowCustom.cpp:

(WebCore::jsDOMWindowGetOwnPropertySlotRestrictedAccess):
(WebCore::JSDOMWindow::getOwnPropertySlot):
(WebCore::JSDOMWindow::put):
(WebCore::addCrossOriginWindowPropertyNames):
(WebCore::addCrossOriginWindowOwnPropertyNames):

Switch to using WebCoreBuiltinNames for WebCore specific identifiers.

bindings/js/WebCoreBuiltinNames.h:

Add WebCore specific identifiers from JavaScriptCore's CommonIdentifiers
as well as some new ones need by the [EnableBySetting] unification.

bindings/scripts/test/JS/JSTestEnabledBySetting.cpp: Added.
bindings/scripts/test/JS/JSTestEnabledBySetting.h: Added.
bindings/scripts/test/TestEnabledBySetting.idl: Added.

Add new test for [EnabledBySetting].

bindings/scripts/test/JS/JSTestGlobalObject.cpp:
bindings/scripts/test/JS/JSTestObj.cpp:
bindings/scripts/test/TestObj.idl:

Remove use of [EnabledBySetting] from TestObj, as it is no longer
valid to use [EnabledBySetting] on interfaces that are no uniquely
exposed to the window environment.

2:24 AM Changeset in webkit [221584] by Carlos Garcia Campos

2 edits in releases/WebKitGTK/webkit-2.18/Source/WebKit

Merge r221574 - [Threaded Compositor] Deadlock in ThreadedDisplayRefreshMonitor
https://bugs.webkit.org/show_bug.cgi?id=176296

Reviewed by Žan Doberšek.

The main thread tries to schedule a display refresh monitor from RAF, requestDisplayRefreshMonitorUpdate() tries
to take the attributes lock to update coordinateUpdateCompletionWithClient. The compositing thread has finished
a frame and sceneUpdateFinished() takes the attributes lock to update clientRendersNextFrame and then asks the
display refresh monitor if it requires a new callback. ThreadedDisplayRefreshMonitor::requiresDisplayRefreshCallback()
takes its lock to check the state, and requestDisplayRefreshMonitorUpdate() was called by
ThreadedDisplayRefreshMonitor::requestRefreshCallback() with the display refresh monitor lock taken. So,
the compositing thread has the attributes lock and wants the display refresh monitor one, and main thread has the
display refresh monitor lock and wants the attributes one.

Shared/CoordinatedGraphics/threadedcompositor/ThreadedDisplayRefreshMonitor.cpp:

(WebKit::ThreadedDisplayRefreshMonitor::requestRefreshCallback): Release the mutex before calling
ThreadedCompositor::requestDisplayRefreshMonitorUpdate().

2:24 AM Changeset in webkit [221583] by Yusuke Suzuki

11 edits in trunk/Source

Remove "malloc" and "free" use
https://bugs.webkit.org/show_bug.cgi?id=176310

Reviewed by Darin Adler.

Source/JavaScriptCore:

Use Vector instead.

API/JSWrapperMap.mm:

(selectorToPropertyName):

Source/WebCore:

Use MallocPtr<>, fastMalloc/fastFree, or Vector instead of manual call of system malloc/free.
In this patch, we apply the above change if we can easily find the pair of malloc/free.
And we do not touch plugin directory since the external code could call free() onto the
fastMalloc-ed memory.

Also, we still use malloc if the system adopts the allocated memory. Later, the system
will deallocate it by calling the system "free".

platform/audio/mac/FFTFrameMac.cpp:

(WebCore::FFTFrame::fftSetupForSize):
(WebCore::FFTFrame::cleanup):

platform/graphics/win/FontCacheWin.cpp:

(WebCore::getLinkedFonts):

platform/graphics/win/FontCustomPlatformData.cpp:

(WebCore::FontCustomPlatformData::fontPlatformData):

platform/graphics/win/FontPlatformDataWin.cpp:

(WebCore::FontPlatformData::FontPlatformData):

platform/mac/WebCoreNSURLExtras.mm:

(WebCore::URLByTruncatingOneCharacterBeforeComponent):
(WebCore::dataForURLComponentType):
(WebCore::URLByRemovingComponentAndSubsequentCharacter):
(WebCore::createStringWithEscapedUnsafeCharacters):
(WebCore::userVisibleString):

platform/win/ClipboardUtilitiesWin.cpp:

(WebCore::markupToCFHTML):

Source/WTF:

Use Vector instead.

wtf/Assertions.cpp:

2:24 AM Changeset in webkit [221582] by Carlos Garcia Campos

4 edits
1 copy in releases/WebKitGTK/webkit-2.18/Source

Merge r221572 - [GTK] Libgcrypt warning: missing initialization - please fix the application
https://bugs.webkit.org/show_bug.cgi?id=171230

Reviewed by Michael Catanzaro.

Source/WebCore/PAL:

Add the Initialization.h header, containing the single initialize() static
function that initializes the libgcrypt library. This header only includes
the default libgcrypt header, and nothing else.

It's not added to the existing Utilities.h header because that one pulls in
other unnecessary headers that for instance use the new() operator, but that
collides with the implementation files that define WebKit process entrypoints
and intentionally don't include the config.h that brings in FastMalloc to
properly define the new() and delete() operators, resulting in compilation
errors due to the new and delete operators being defined to warning strings
in always-included WebKit2Prefix.h.

pal/crypto/gcrypt/Initialization.h: Copied from Source/WebKit/WebProcess/EntryPoint/unix/WebProcessMain.cpp.

(PAL::GCrypt::initialize):

Source/WebKit:

Replicate the libgcrypt initialization in the NetworkProcess entrypoint.
This is required due to libgcrypt being used through digest operations
originating from the certificate hash computation in SoupNetworkSession.

The initialization is moved to the PAL library, and the initialize()
function is now leveraged in both NetworkProcess and WebProcess entrypoints.

NetworkProcess/EntryPoint/unix/NetworkProcessMain.cpp:

(main):

WebProcess/EntryPoint/unix/WebProcessMain.cpp:

(main):

2:22 AM Changeset in webkit [221581] by Carlos Garcia Campos

5 edits in releases/WebKitGTK/webkit-2.18/Source

Merge r221562 - String#utf8() allocates new CString
https://bugs.webkit.org/show_bug.cgi?id=176302

Reviewed by Sam Weinig.

Source/WebCore:

Several places uses String#utf8() like an accessor. In reality, it allocates new CString per call.
It is very costly. Furthermore, some places uses this incorrectly. For example,
std::vector<char>(srtpAuth.utf8().data(), srtpAuth.utf8().data() + srtpAuth.utf8().length());
is incorrect since each time strpAuth.utf8() allocates different CString.

This patch calls utf8() first and use it.

Modules/websockets/WebSocketChannel.cpp:

(WebCore::WebSocketChannel::startClosingHandshake):

platform/wpe/PlatformPasteboardWPE.cpp:

(WebCore::PlatformPasteboard::write):

Source/WebKit:

NetworkProcess/webrtc/NetworkRTCSocket.cpp:

(WebKit::NetworkRTCSocket::sendTo):

2:22 AM Changeset in webkit [221580] by Carlos Garcia Campos

2 edits in releases/WebKitGTK/webkit-2.18/Source/WebCore

Merge r221561 - [SOUP] Use fastMalloced SoupBuffer in ResourcehandleSoup
https://bugs.webkit.org/show_bug.cgi?id=176311

Reviewed by Michael Catanzaro.

Use fastMalloced memory instead of g_malloc for SoupBuffer.

platform/network/soup/ResourceHandleSoup.cpp:

(WebCore::ResourceHandle::ensureReadBuffer):

2:14 AM Changeset in webkit [221579] by Carlos Garcia Campos

9 edits in releases/WebKitGTK/webkit-2.18

Merge r221542 - Wrong getComputedStyle result for pseudo-elements in display: none subtrees.
https://bugs.webkit.org/show_bug.cgi?id=176259

Patch by Emilio Cobos Álvarez <emilio> on 2017-09-02
Reviewed by Antti Koivisto.

LayoutTests/imported/w3c:

web-platform-tests/cssom/getComputedStyle-pseudo-expected.txt:

Source/WebCore:

Before this patch, we were wrongly returning the parent style when there wasn't
a cached pseudo style.

With this, the only remaining failures in getComputedStyle-pseudo is a
serialisation bug for the "content" property, and the display: contents test,
which is because we don't support pseudo-elements of display: contents elements
yet.

Tests: imported/w3c/web-platform-tests/cssom/getComputedStyle-pseudo.html

dom/Document.cpp:

(WebCore::Document::styleForElementIgnoringPendingStylesheets):

dom/Document.h:
dom/Element.cpp:

(WebCore::Element::resolvePseudoElementStyle):
(WebCore::Element::computedStyle):

dom/Element.h:

1:50 AM Changeset in webkit [221578] by Carlos Garcia Campos

7 edits in releases/WebKitGTK/webkit-2.18/Source/WebCore

Merge r221539 - Performance / PerformanceObserver leaks seen on bots
https://bugs.webkit.org/show_bug.cgi?id=176269

Patch by Joseph Pecoraro <Joseph Pecoraro> on 2017-09-02
Reviewed by Darin Adler.

page/DOMWindow.cpp:

(WebCore::DOMWindow::removeAllEventListeners):

workers/WorkerGlobalScope.cpp:

(WebCore::WorkerGlobalScope::removeAllEventListeners):
Also use this opportunity to break up the reference cycle
between PerformanceObservers and the Performance object.

page/Performance.h:
page/Performance.cpp:

(WebCore::Performance::removeAllObservers):
Disassociate and clear observers.

page/PerformanceObserver.h:
page/PerformanceObserver.cpp:

(WebCore::PerformanceObserver::disassociate):
Clear the back-reference to the Performance object to
break the cycle.

1:49 AM Changeset in webkit [221577] by Carlos Garcia Campos

5 edits in releases/WebKitGTK/webkit-2.18/Source

Merge r221521 - [GTK][Wayland] Use fast malloc to allocate backing store cairo surfaces data
https://bugs.webkit.org/show_bug.cgi?id=176226

Reviewed by Michael Catanzaro.

Source/WebCore:

It reduces the memory consumption and improves the performance. Also, since performance is better we don't need
to keep the scroll surface allocated all the time, we can create it on demand and keep it allocated only while
scrolling too fast.

platform/graphics/cairo/BackingStoreBackendCairoImpl.cpp:

(WebCore::createCairoImageSurfaceWithFastMalloc): Helper to create a cairo image surface using fast malloc.
(WebCore::BackingStoreBackendCairoImpl::BackingStoreBackendCairoImpl): Create the main surface, but no the
scroll one. Initialize the hysteresis activity to discard the scroll surface after scrolling.
(WebCore::BackingStoreBackendCairoImpl::scroll): Create the scroll surface only if it's nullptr, and notify the
hysteresis activity.

platform/graphics/cairo/BackingStoreBackendCairoImpl.h:

Source/WebKit:

Do not create the cairo surface for BackingStoreBackendCairoImpl. When not using X11 we are going to use an
image surface anyway, so we don't really need to use the create similar. This way we only need to pass the size
and device scale factor to BackingStoreBackendCairoImpl that will create the image surface using fast malloc to
allocate the data.

UIProcess/cairo/BackingStoreCairo.cpp:

(WebKit::BackingStore::createBackend):

1:49 AM Changeset in webkit [221576] by Carlos Garcia Campos

5 edits in releases/WebKitGTK/webkit-2.18/Source/WebKit

Merge r221514 - [GTK][Wayland] Opening FedoraProject's pastebin chews CPU
https://bugs.webkit.org/show_bug.cgi?id=175942

Reviewed by Žan Doberšek.

This regressed when we introduced the display refresh monitor. The monitor schedules updates immediately,
because we removed the option to not do frame sync in X11 to let swapBuffers do the throttling, but that's
not happening in Wayland because the nested compositor is dispatching frame callbacks on surface commit.
We need to ensure that frame callbacks are dispatched on every monitor refresh, because swapBuffers waits for
frame callbacks to be queued on display.

UIProcess/gtk/WaylandCompositor.cpp:

(WebKit::WaylandCompositor::Surface::~Surface): Destroy pending frame callbacks too.
(WebKit::WaylandCompositor::Surface::setWebPage): Add a tick callback to the web view widget to flush all
committed frame callbacks on every frame update.
(WebKit::WaylandCompositor::Surface::requestFrame): Add the callbacks to m_pendingFrameCallbackList.
(WebKit::WaylandCompositor::Surface::flushFrameCallbacks): Dispatch all committed frame callabcks.
(WebKit::WaylandCompositor::Surface::flushPendingFrameCallbacks): Dispatch all pending frame callbacks.
(WebKit::WaylandCompositor::Surface::commit): Do not dispatch frame callbacks here, move them to the list of
committed frame callbacks that will be dispatched on the next frame clock update.

UIProcess/gtk/WaylandCompositor.h:

(WebKit::WaylandCompositor::Surface::setWebPage): Moved to the cpp.

WebProcess/WebPage/gtk/AcceleratedSurfaceWayland.cpp:

(WebKit::AcceleratedSurfaceWayland::AcceleratedSurfaceWayland): Move surface initialization and destruction to
the compositing thread.
(WebKit::AcceleratedSurfaceWayland::initialize):
(WebKit::AcceleratedSurfaceWayland::finalize):
(WebKit::AcceleratedSurfaceWayland::~AcceleratedSurfaceWayland): Deleted.

WebProcess/WebPage/gtk/AcceleratedSurfaceWayland.h:

1:04 AM Changeset in webkit [221575] by Yusuke Suzuki

2 edits in trunk/Source/WTF

Unreviewed, support libstdc++ use with clang
https://bugs.webkit.org/show_bug.cgi?id=176301

wtf/FastMalloc.h:

Sep 3, 2017:

11:52 PM Changeset in webkit [221574] by Carlos Garcia Campos

2 edits in trunk/Source/WebKit

[Threaded Compositor] Deadlock in ThreadedDisplayRefreshMonitor
https://bugs.webkit.org/show_bug.cgi?id=176296

Reviewed by Žan Doberšek.

Shared/CoordinatedGraphics/threadedcompositor/ThreadedDisplayRefreshMonitor.cpp:

(WebKit::ThreadedDisplayRefreshMonitor::requestRefreshCallback): Release the mutex before calling
ThreadedCompositor::requestDisplayRefreshMonitorUpdate().

11:09 PM WebKitGTK/2.18.x edited by zandobersek@gmail.com

(diff)

11:07 PM Changeset in webkit [221573] by Chris Dumez

3 edits in trunk/Source/WebCore

Use StringView in DOMFileSystem::evaluatePath()
https://bugs.webkit.org/show_bug.cgi?id=176324

Reviewed by Darin Adler.

Use StringView in DOMFileSystem::evaluatePath() for performance.

Modules/entriesapi/DOMFileSystem.cpp:

(WebCore::resolveRelativeVirtualPath):
(WebCore::DOMFileSystem::evaluatePath):

Modules/entriesapi/DOMFileSystem.h:

11:04 PM Changeset in webkit [221572] by zandobersek@gmail.com

4 edits
1 add in trunk/Source

[GTK] Libgcrypt warning: missing initialization - please fix the application
https://bugs.webkit.org/show_bug.cgi?id=171230

Reviewed by Michael Catanzaro.

Source/WebCore/PAL:

Add the Initialization.h header, containing the single initialize() static
function that initializes the libgcrypt library. This header only includes
the default libgcrypt header, and nothing else.

pal/crypto/gcrypt/Initialization.h: Copied from Source/WebKit/WebProcess/EntryPoint/unix/WebProcessMain.cpp.

(PAL::GCrypt::initialize):

Source/WebKit:

The initialization is moved to the PAL library, and the initialize()
function is now leveraged in both NetworkProcess and WebProcess entrypoints.

NetworkProcess/EntryPoint/unix/NetworkProcessMain.cpp:

(main):

WebProcess/EntryPoint/unix/WebProcessMain.cpp:

(main):

11:03 PM Changeset in webkit [221571] by zandobersek@gmail.com

2 edits in trunk/Source/WebKit

[WPE] Use the device scale factor value in WebPageCreationParameters
https://bugs.webkit.org/show_bug.cgi?id=175377

Reviewed by Michael Catanzaro.

WebProcess/WebPage/AcceleratedDrawingArea.cpp:

(WebKit::AcceleratedDrawingArea::AcceleratedDrawingArea):
WebPageCreationParameters::deviceScaleFactor value should be used
to set the device scale factor on the given WebPage object. This
enables correctly leveraging the scale factor for the WPE port
which enters accelerated mode unconditionally and as such doesn't
wait for the backing store update that would provide that value
otherwise.

10:58 PM Changeset in webkit [221570] by Brent Fulgham

2 edits in trunk/Source/WebKit

[WK2][iOS] Remove unnecessary IOHIDEventServiceFastPathUserClient access
https://bugs.webkit.org/show_bug.cgi?id=175947
<rdar://problem/33353627>

Reviewed by Darin Adler.

Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:

9:08 PM Changeset in webkit [221569] by mitz@apple.com

2 edits in trunk/Source/WebCore/PAL

[Cocoa] Tidy up PassKitSPI.h
https://bugs.webkit.org/show_bug.cgi?id=176323

Reviewed by Darin Adler.

pal/spi/cocoa/PassKitSPI.h: Moved the declarations of requiredShippingContactFields and requiredBillingContactFields to where other PKPaymentRequest properties are declared, and changed the guard to relate to the SDK version. Replaced use of has_include with checking for the version of the Apple internal SDK in which the typedefs first appeared.

8:33 PM Changeset in webkit [221568] by mitz@apple.com

2 edits in trunk/Source/WebKit

WKContentViewInteraction.mm includes redundant declaration of UIPreviewItemController's presentationSecondaryGestureRecognizer
https://bugs.webkit.org/show_bug.cgi?id=176322

Reviewed by Darin Adler.

UIProcess/ios/WKContentViewInteraction.mm: Removed the declaration, which is already made in the SDK or in UIKitSPI.h.

8:27 PM Changeset in webkit [221567] by commit-queue@webkit.org

2 edits in trunk/Tools

Make webkit-patch post --no-review enable EWS just like webkit-patch post-commits --no-review
https://bugs.webkit.org/show_bug.cgi?id=176321

Patch by Sam Weinig <sam@webkit.org> on 2017-09-03
Reviewed by Darin Adler.

Scripts/webkitpy/tool/commands/upload.py:

Add the SubmitToEWS step to Post command

8:01 PM Changeset in webkit [221566] by timothy_horton@apple.com

2 edits in trunk/Source/WebKit

Try to fix the iOS build

UIProcess/RemoteLayerTree/ios/RemoteLayerTreeHostIOS.mm:

(-[WKRemoteView initWithFrame:contextID:]):

7:51 PM Changeset in webkit [221565] by Darin Adler

2 edits in trunk/Source/JavaScriptCore

Try to fix Windows build.

runtime/JSGlobalObjectFunctions.cpp: #include <unicode/utf8.h>.

7:12 PM Changeset in webkit [221564] by mitz@apple.com

2 edits in trunk/Source/WebCore/PAL

[Cocoa] Remove use of has_include in CoreAudioSPI.h
https://bugs.webkit.org/show_bug.cgi?id=176320

Reviewed by Tim Horton.

pal/spi/cf/CoreAudioSPI.h: AudioHardwarePriv.h is available in all Apple internal SDKs and only in Apple internal SDKs. AudioHardwareBase.h is available only in the macOS SDK, where it is included by AudioHardware.h.

6:40 PM Changeset in webkit [221563] by timothy_horton@apple.com

13 edits in trunk/Source

Fix some style issues in the Remote Layer Tree code
https://bugs.webkit.org/show_bug.cgi?id=176298

Reviewed by Dan Bernstein.

Source/WebCore/PAL:

pal/spi/cocoa/QuartzCoreSPI.h:

Source/WebKit:

In bugs 176273 and 176297, the style bot noticed some existing style
issues in the UI-side compositing code. Fix them, because there aren't
many and they're straightforward.

Shared/RemoteLayerTree/RemoteLayerBackingStore.h:
Shared/RemoteLayerTree/RemoteLayerBackingStoreCollection.h:
Shared/RemoteLayerTree/RemoteLayerTreeTransaction.h:
Shared/RemoteLayerTree/RemoteScrollingCoordinatorTransaction.h:
UIProcess/RemoteLayerTree/RemoteLayerTreeDrawingAreaProxy.mm:
UIProcess/RemoteLayerTree/RemoteLayerTreeScrollingPerformanceData.h:
UIProcess/RemoteLayerTree/RemoteScrollingCoordinatorProxy.h:
UIProcess/RemoteLayerTree/RemoteScrollingTree.h:
UIProcess/RemoteLayerTree/ios/RemoteLayerTreeHostIOS.mm:
UIProcess/RemoteLayerTree/ios/ScrollingTreeOverflowScrollingNodeIOS.h:

6:38 PM Changeset in webkit [221562] by Yusuke Suzuki

5 edits in trunk/Source

String#utf8() allocates new CString
https://bugs.webkit.org/show_bug.cgi?id=176302

Reviewed by Sam Weinig.

Source/WebCore:

This patch calls utf8() first and use it.

Modules/websockets/WebSocketChannel.cpp:

(WebCore::WebSocketChannel::startClosingHandshake):

platform/wpe/PlatformPasteboardWPE.cpp:

(WebCore::PlatformPasteboard::write):

Source/WebKit:

NetworkProcess/webrtc/NetworkRTCSocket.cpp:

(WebKit::NetworkRTCSocket::sendTo):

6:37 PM Changeset in webkit [221561] by Yusuke Suzuki

2 edits in trunk/Source/WebCore

[SOUP] Use fastMalloced SoupBuffer in ResourcehandleSoup
https://bugs.webkit.org/show_bug.cgi?id=176311

Reviewed by Michael Catanzaro.

Use fastMalloced memory instead of g_malloc for SoupBuffer.

platform/network/soup/ResourceHandleSoup.cpp:

(WebCore::ResourceHandle::ensureReadBuffer):

6:00 PM Changeset in webkit [221560] by Yusuke Suzuki

4 edits in trunk/Source

[WTF] Add C++03 allocator interface for GCC < 6
https://bugs.webkit.org/show_bug.cgi?id=176301

Reviewed by Darin Adler.

Source/JavaScriptCore:

dfg/DFGObjectAllocationSinkingPhase.cpp:

Source/WTF:

Unfortunately, std::list in GCC < 6 does not support C++11 allocator interface.
This patch adds C++03 allocator interface to FastAllocator to make it usable
for std::list. It also allows us to use FastAllocator for data structures that
only support C++03 interface.

wtf/FastMalloc.h:

(WTF::FastAllocator::allocate):
(WTF::FastAllocator::construct):
(WTF::FastAllocator::destroy):
(WTF::FastAllocator::max_size const):
(WTF::FastAllocator::select_on_container_copy_construction const):

3:28 PM Changeset in webkit [221559] by Darin Adler

6 edits in trunk/Source/WebKitLegacy/mac

[Cocoa] Tidy a few things in legacy WebHTMLView
https://bugs.webkit.org/show_bug.cgi?id=176280

Reviewed by Dan Bernstein.

WebCoreSupport/WebChromeClient.mm:

(WebChromeClient::attachRootGraphicsLayer): Made this function Mac-only, because the
methods it calls on WebHTMLView were doing nothing on iOS.

WebView/WebFrame.mm:

(-[WebFrame _paintBehaviorForDestinationContext:]): Changed iOS code so that it no longer
depends on -[WebHTMLView _web_isDrawingIntoLayer], which always returned YES on iOS.

WebView/WebHTMLView.mm: Define WebEvent on both iOS and Mac so more event handling

code can be shared without #if below. Use #if PLATFORM(MAC) instead of #if !PLATFORM(IOS)
in a lot more places. Move many of the SPI declarations up to the top in alphabetical
order. Removed all sorts of unused things. Made more code and fields Mac-only where
what was compiled was doing nothing on iOS and it makes the resulting code simpler.
Use RetainPtr instead of writing retain/release wherever possible.
(+[WebMenuTarget sharedMenuTarget]): Moved global inside this method and used C++
initialization for it. Removed various unused methods and declarations of methods.
(promisedDataClient): Use NeverDestroyed instead of just doing a "new" on the heap.
(-[WebHTMLView _imageExistsAtPaths:]): Deleted.
(-[WebHTMLView _documentFragmentWithPaths:]): Use RetainPtr and modern for loop.
(+[WebHTMLView _excludedElementsForAttributedStringConversion]): Use C++ initialization,
and remove CFRetain no longer needed since we don't support garbage collection any more.
(uniqueURLWithRelativePart): Simplified using [NSUUID UUID].
(-[WebHTMLView _plainTextFromPasteboard:]): Simplified with RetainPtr.
(-[WebHTMLView _pasteWithPasteboard:allowPlainText:]): Ditto.
(-[WebHTMLView _pasteAsPlainTextWithPasteboard:]): Ditto.
(-[WebHTMLView _removeWindowObservers]): Removed code dealing with empty
methods windowWillOrderOnScreen: and windowWillOrderOffScreen:.
(-[WebHTMLView _shouldDeleteRange:]): Deleted. Was unused.
(-[WebHTMLView _writeSelectionWithPasteboardTypes:toPasteboard:cachedAttributedString:]):
Simplified with stringByReplacingOccurrencesOfString and modern string literal syntax.
(-[WebHTMLView _setMouseDownEvent:]): Ditto.
(-[WebHTMLView _web_setPrintingModeRecursive:adjustViewSize:]): Added. Helper method so
the next three methhods can be super-simple. Used RetainPtr and a modern for loop to make
this method simpler than the functions it replaced. Also got rid of the enumeratingSubviews
debug flag; we haven't used it for years and it's not clearly helpful.
(-[WebHTMLView _web_setPrintingModeRecursive]): Turned into a one-liner with the above.
(-[WebHTMLView _web_clearPrintingModeRecursive]): Ditto.
(-[WebHTMLView _web_setPrintingModeRecursiveAndAdjustViewSize]): Ditto.
(-[WebHTMLView _web_makePluginSubviewsPerformSelector:withObject:]): Moved from the
WebInternal category to the WebHTMLViewFileInternal category so it can be called by
two additional methods in the WebPrivate category and still be called by the two
methods in the WebInternal category that were already using it.
(-[WebHTMLView didAddSubview:]): Deleted.
(-[WebHTMLView _insideAnotherHTMLView]): Deleted.
(-[WebHTMLView _clearLastHitViewIfSelf]): Deleted. Inlined the implementation in the one
place this was called, the close method.
(-[WebHTMLView _setToolTip:]): Simplified with RetainPtr.
(-[WebHTMLView pasteboard:provideDataForType:]): Ditto.
(-[WebHTMLView _autoscroll]): Simplified by making this Mac-only.
(-[WebHTMLView _hasHTMLDocument]): Deleted. The one caller was removed a long time ago.
(-[WebHTMLView _web_documentFragmentFromPasteboard:pasteboardType:imageMIMEType:]):
Added helper method to make the one below simpler.
(-[WebHTMLView _documentFragmentFromPasteboard:forType:inContext:subresources:]):
Compared pasteboard type strings using isEqualToString: rather than using == since
we want to actually compare the string contents. Also updated to use modern for loop,
renamed a local variable from "s" to "localSubresources", and factored out the handling
of three image types to use the helper method above.
(-[WebHTMLView _pauseNullEventsForAllNetscapePlugins]): Rewrite using the
_web_makePluginSubviewsPerformSelector:withObject: method.
(-[WebHTMLView _resumeNullEventsForAllNetscapePlugins]): Ditto.
(-[NSView _web_addDescendentWebHTMLViewsToArray:]): Fixed spelling to use the adjective
"descendent" rather than the noun "descendant". Also removed Mac-only optimization to
not call the -[NSView subviews] method to make this identical on both platforms and used
a modern for loop.
(-[NSMutableDictionary _web_setObjectIfNotNil:forKey:]): Deleted.
(-[WebHTMLView acceptsFirstResponder]): Reduced amount of platform-specific code by
defining WebEventMouseDown as a synonym for NSEventTypeLeftMouseDown on Mac.
(-[WebHTMLView addWindowObservers]): Removed code dealing with empty
methods windowWillOrderOnScreen: and windowWillOrderOffScreen:.
(-[WebHTMLView willRemoveSubview:]): Removed the enumeratingSubviews check.
(-[WebHTMLView drawSingleRect:]): Rearranged this method to make it clearer that most
of it is Mac-only.
(-[WebHTMLView windowWillOrderOnScreen:]): Deleted. All the current version of this
method did was call itself on the main thread, and do nothing if called on main thread.
(-[WebHTMLView windowWillOrderOffScreen:]): Deleted. Ditto.
(-[WebHTMLView _isSelectionEvent:]): Merged macOS and iOS versions.
(-[WebHTMLView resignFirstResponder]): Removed iOS-specific code in the macOS-specific
half of this method.
(-[WebHTMLView setDataSource:]): Simplified using RetainPtr.
(-[WebHTMLView _setPrinting:minimumPageLogicalWidth:logicalHeight:originalPageWidth:originalPageHeight:maximumShrinkRatio:adjustViewSize:paginateScreenContent:]):
Simplified using a modern for loop and RetainPtr.
(-[WebHTMLView knowsPageRange:]): Simplified using RetainPtr.
(-[WebHTMLView keyDown:]): Ditto.
(-[WebHTMLView _selectionStartFontAttributesAsRTF]): Ditto.
(-[WebHTMLView performKeyEquivalent:]): Ditto.
(-[WebHTMLView _styleForAttributeChange:]): Ditto.
(-[WebHTMLView complete:]): Ditto.
(-[WebHTMLView _windowChangedKeyState]): Removed iOS-specific code since this method is
now macOS-only.
(-[WebHTMLView _selectionChanged]): Removed call to _updateSelectionForInputManager on
iOS since that is now macOS-only.
(-[WebHTMLView promisedDragTIFFDataSource]): Deleted.
(-[WebHTMLView _startAutoscrollTimer:]): Moved this into an internal category instead of
a private one because it is not used or needed outside WebKit. Also made it macOS-only
and reorganized to use early return.
(-[WebHTMLView _stopAutoscrollTimer]): Moved into internal category. It remains a no-op
on iOS, but kept there so we don't have to do #if at every call site.
(-[WebHTMLView _pluginController]): Moved into internal category.
(-[WebHTMLView validAttributesForMarkedText]): Simplified by removing CFRetain that was
there only for GC, using C++ initialization, removing obsolete comment.
(-[WebHTMLView _updateSelectionForInputManager]): Made this macOS-only and removed
peculiar iOS-only comment.
(+[NSURL _web_uniqueWebDataURL]): Simplified using [NSUUID UUID].

WebView/WebHTMLViewInternal.h: Removed unused promisedDragTIFFDataSource method.

Made attachRootLayer:, detachRootLayer, _web_isDrawingIntoLayer,
_web_isDrawingIntoAcceleratedLayer all macOS-only. Moved _startAutoscrollTimer:,
_stopAutoscrollTimer, and _pluginController here from SPI header since they aren't
needed or used as SPI. And made _startAutoscrollTimer macOS-only.

WebView/WebHTMLViewPrivate.h: Removed unneeded and unused SPI: _insideAnotherHTMLView,

_clearLastHitViewIfSelf, _pluginController, _startAutoscrollTimer:, and _stopAutoscrollTimer.
Made _transparentBackground and _setTransparentBackground: macOS-only. These probably
are not needed on macOS either, but I wasn't able to make absolutely sure I could remove
them at this time.

3:03 PM Changeset in webkit [221558] by Chris Dumez

13 edits in trunk

Have getFileMetadata() return a std::optional<FileMetadata>
https://bugs.webkit.org/show_bug.cgi?id=176293

Reviewed by Darin Adler.

Source/WebCore:

Have getFileMetadata() return a std::optional<FileMetadata> instead of
using an out parameter for the metadata.

Modules/entriesapi/DOMFileSystem.cpp:

(WebCore::listDirectoryWithMetadata):
(WebCore::validatePathIsExpectedType):
(WebCore::DOMFileSystem::getEntry):

html/FileListCreator.cpp:

(WebCore::appendDirectoryFiles):

platform/FileSystem.cpp:

(WebCore::fileIsDirectory):

platform/FileSystem.h:
platform/glib/FileSystemGlib.cpp:

(WebCore::getFileMetadata):

platform/network/BlobDataFileReference.cpp:

(WebCore::BlobDataFileReference::startTrackingModifications):

platform/network/mac/BlobDataFileReferenceMac.mm:

(WebCore::BlobDataFileReference::generateReplacementFile):

platform/posix/FileSystemPOSIX.cpp:

(WebCore::getFileMetadata):

platform/win/FileSystemWin.cpp:

(WebCore::getFileMetadata):

Tools:

TestWebKitAPI/Tests/WebCore/FileSystem.cpp:

(TestWebKitAPI::TEST_F):

2:55 PM Changeset in webkit [221557] by Chris Dumez

2 edits in trunk/Source/JavaScriptCore

Unreviewed, rolling out r221555.

Did not fix Windows build

Reverted changeset:

"Unreviewed attempt to fix Windows build."
http://trac.webkit.org/changeset/221555

2:49 PM Changeset in webkit [221556] by Chris Dumez

6 edits
5 copies
25 moves
8 adds
3 deletes in trunk/LayoutTests

Consolidate entries-api tests under entries-api/ folders
https://bugs.webkit.org/show_bug.cgi?id=176315

Reviewed by Darin Adler.

Consolidate entries-api tests under entries-api/ folders, for clarity.
Also drop an older test for directory upload which did not pass because
it was missing its resources. We now have good coverage for this feature
so this test is not needed.

editing/pasteboard/enties-api/datatransfer-items-drop-directoryReader-error-expected.txt: Renamed from LayoutTests/editing/pasteboard/datatransfer-items-drop-directoryReader-error-expected.txt.
editing/pasteboard/enties-api/datatransfer-items-drop-directoryReader-error.html: Renamed from LayoutTests/editing/pasteboard/datatransfer-items-drop-directoryReader-error.html.
editing/pasteboard/enties-api/datatransfer-items-drop-directoryReader-expected.txt: Renamed from LayoutTests/editing/pasteboard/datatransfer-items-drop-directoryReader-expected.txt.
editing/pasteboard/enties-api/datatransfer-items-drop-directoryReader-root-expected.txt: Renamed from LayoutTests/editing/pasteboard/datatransfer-items-drop-directoryReader-root-expected.txt.
editing/pasteboard/enties-api/datatransfer-items-drop-directoryReader-root.html: Renamed from LayoutTests/editing/pasteboard/datatransfer-items-drop-directoryReader-root.html.
editing/pasteboard/enties-api/datatransfer-items-drop-directoryReader.html: Renamed from LayoutTests/editing/pasteboard/datatransfer-items-drop-directoryReader.html.
editing/pasteboard/enties-api/datatransfer-items-drop-fileEntry-file-expected.txt: Renamed from LayoutTests/editing/pasteboard/datatransfer-items-drop-fileEntry-file-expected.txt.
editing/pasteboard/enties-api/datatransfer-items-drop-fileEntry-file.html: Renamed from LayoutTests/editing/pasteboard/datatransfer-items-drop-fileEntry-file.html.
editing/pasteboard/enties-api/datatransfer-items-drop-getAsEntry-expected.txt: Renamed from LayoutTests/editing/pasteboard/datatransfer-items-drop-getAsEntry-expected.txt.
editing/pasteboard/enties-api/datatransfer-items-drop-getAsEntry.html: Renamed from LayoutTests/editing/pasteboard/datatransfer-items-drop-getAsEntry.html.
editing/pasteboard/enties-api/datatransfer-items-drop-getDirectory-expected.txt: Renamed from LayoutTests/editing/pasteboard/datatransfer-items-drop-getDirectory-expected.txt.
editing/pasteboard/enties-api/datatransfer-items-drop-getDirectory.html: Renamed from LayoutTests/editing/pasteboard/datatransfer-items-drop-getDirectory.html.
editing/pasteboard/enties-api/datatransfer-items-drop-getFile-expected.txt: Renamed from LayoutTests/editing/pasteboard/datatransfer-items-drop-getFile-expected.txt.
editing/pasteboard/enties-api/datatransfer-items-drop-getFile.html: Renamed from LayoutTests/editing/pasteboard/datatransfer-items-drop-getFile.html.
editing/pasteboard/enties-api/datatransfer-items-drop-getParent-expected.txt: Renamed from LayoutTests/editing/pasteboard/datatransfer-items-drop-getParent-expected.txt.
editing/pasteboard/enties-api/datatransfer-items-drop-getParent-root-expected.txt: Renamed from LayoutTests/editing/pasteboard/datatransfer-items-drop-getParent-root-expected.txt.
editing/pasteboard/enties-api/datatransfer-items-drop-getParent-root.html: Renamed from LayoutTests/editing/pasteboard/datatransfer-items-drop-getParent-root.html.
editing/pasteboard/enties-api/datatransfer-items-drop-getParent.html: Renamed from LayoutTests/editing/pasteboard/datatransfer-items-drop-getParent.html.
editing/pasteboard/enties-api/datatransfer-items-drop-getParent2-expected.txt: Renamed from LayoutTests/editing/pasteboard/datatransfer-items-drop-getParent2-expected.txt.
editing/pasteboard/enties-api/datatransfer-items-drop-getParent2.html: Renamed from LayoutTests/editing/pasteboard/datatransfer-items-drop-getParent2.html.
fast/forms/file/entries-api/resources/testFiles/file1.txt: Renamed from LayoutTests/fast/forms/file/resources/testFiles/file1.txt.
fast/forms/file/entries-api/resources/testFiles/file2.txt: Renamed from LayoutTests/fast/forms/file/resources/testFiles/file2.txt.
fast/forms/file/entries-api/resources/testFiles/subfolder1/file3.txt: Renamed from LayoutTests/fast/forms/file/resources/testFiles/subfolder1/file3.txt.
fast/forms/file/entries-api/resources/testFiles/subfolder2/file4.txt: Renamed from LayoutTests/fast/forms/file/resources/testFiles/subfolder2/file4.txt.
fast/forms/file/entries-api/resources/testFiles/subfolder2/subfolder2a/file5.txt: Renamed from LayoutTests/fast/forms/file/resources/testFiles/subfolder2/subfolder2a/file5.txt.
fast/forms/file/entries-api/webkitdirectory-drag-folder-expected.txt: Renamed from LayoutTests/fast/forms/file/webkitdirectory-drag-folder-expected.txt.
fast/forms/file/entries-api/webkitdirectory-drag-folder.html: Renamed from LayoutTests/fast/forms/file/webkitdirectory-drag-folder.html.
fast/forms/file/entries-api/webkitdirectory-open-panel-expected.txt: Renamed from LayoutTests/fast/forms/file/webkitdirectory-open-panel-expected.txt.
fast/forms/file/entries-api/webkitdirectory-open-panel.html: Renamed from LayoutTests/fast/forms/file/webkitdirectory-open-panel.html.
fast/forms/file/input-file-directory-upload-expected.txt: Removed.
fast/forms/file/input-file-directory-upload.html: Removed.
platform/ios/TestExpectations:
platform/mac-wk1/fast/forms/file/entries-api/webkitdirectory-open-panel-expected.txt: Renamed from LayoutTests/platform/mac-wk1/fast/forms/file/webkitdirectory-open-panel-expected.txt.
platform/mac/TestExpectations:
platform/win/TestExpectations:
platform/wk2/TestExpectations:

1:38 PM Changeset in webkit [221555] by Chris Dumez

2 edits in trunk/Source/JavaScriptCore

Unreviewed attempt to fix Windows build.

runtime/JSGlobalObjectFunctions.cpp:

12:08 PM Changeset in webkit [221554] by timothy_horton@apple.com

3 edits
2 copies
8 moves
1 add
1 delete in trunk/Source/WebKit

Move shared part of the Remote Layer Tree code out of a Mac-specific directory
https://bugs.webkit.org/show_bug.cgi?id=176297

Reviewed by Dan Bernstein.

UI-side compositing works on all Cocoa platforms, so the code should
not live inside Mac/. Also merge the Scrolling directory in here,
since those files are very related.

PlatformMac.cmake:
Shared/RemoteLayerTree/RemoteLayerBackingStore.h: Renamed from Shared/mac/RemoteLayerBackingStore.h.
Shared/RemoteLayerTree/RemoteLayerBackingStore.mm: Renamed from Shared/mac/RemoteLayerBackingStore.mm.
Shared/RemoteLayerTree/RemoteLayerBackingStoreCollection.h: Renamed from Shared/mac/RemoteLayerBackingStoreCollection.h.
Shared/RemoteLayerTree/RemoteLayerBackingStoreCollection.mm: Renamed from Shared/mac/RemoteLayerBackingStoreCollection.mm.
Shared/RemoteLayerTree/RemoteLayerTreePropertyApplier.h: Renamed from Shared/mac/RemoteLayerTreePropertyApplier.h.
Shared/RemoteLayerTree/RemoteLayerTreePropertyApplier.mm: Renamed from Shared/mac/RemoteLayerTreePropertyApplier.mm.
Shared/RemoteLayerTree/RemoteLayerTreeTransaction.h: Renamed from Shared/mac/RemoteLayerTreeTransaction.h.
Shared/RemoteLayerTree/RemoteLayerTreeTransaction.mm: Renamed from Shared/mac/RemoteLayerTreeTransaction.mm.
Shared/RemoteLayerTree/RemoteScrollingCoordinatorTransaction.cpp: Renamed from Shared/Scrolling/RemoteScrollingCoordinatorTransaction.cpp.
Shared/RemoteLayerTree/RemoteScrollingCoordinatorTransaction.h: Renamed from Shared/Scrolling/RemoteScrollingCoordinatorTransaction.h.
WebKit.xcodeproj/project.pbxproj:

11:54 AM Changeset in webkit [221553] by Chris Dumez

4 edits in trunk/Source

Unreviewed, rolling out r221552.

Broke the build

Reverted changeset:

"[WTF] Add C++03 allocator interface for GCC < 6"
https://bugs.webkit.org/show_bug.cgi?id=176301
http://trac.webkit.org/changeset/221552

11:49 AM Changeset in webkit [221552] by Yusuke Suzuki

4 edits in trunk/Source

[WTF] Add C++03 allocator interface for GCC < 6
https://bugs.webkit.org/show_bug.cgi?id=176301

Reviewed by Darin Adler.

Source/JavaScriptCore:

dfg/DFGObjectAllocationSinkingPhase.cpp:

Source/WTF:

wtf/FastMalloc.h:

10:51 AM Changeset in webkit [221551] by Yusuke Suzuki

5 edits in trunk/Source/JavaScriptCore

[JSC] Clean up BytecodeLivenessAnalysis
https://bugs.webkit.org/show_bug.cgi?id=176295

Reviewed by Saam Barati.

Previously, computeDefsForBytecodeOffset was a bit customizable.
This is used for try-catch handler's liveness analysis. But after
careful generatorification implementation, it is now not necessary.
This patch drops this customizability.

bytecode/BytecodeGeneratorification.cpp:

(JSC::GeneratorLivenessAnalysis::computeDefsForBytecodeOffset): Deleted.
(JSC::GeneratorLivenessAnalysis::computeUsesForBytecodeOffset): Deleted.

bytecode/BytecodeLivenessAnalysis.cpp:

(JSC::BytecodeLivenessAnalysis::computeKills):
(JSC::BytecodeLivenessAnalysis::computeDefsForBytecodeOffset): Deleted.
(JSC::BytecodeLivenessAnalysis::computeUsesForBytecodeOffset): Deleted.

bytecode/BytecodeLivenessAnalysis.h:
bytecode/BytecodeLivenessAnalysisInlines.h:

(JSC::BytecodeLivenessPropagation::stepOverInstruction):
(JSC::BytecodeLivenessPropagation::computeLocalLivenessForBytecodeOffset):
(JSC::BytecodeLivenessPropagation::computeLocalLivenessForBlock):
(JSC::BytecodeLivenessPropagation::getLivenessInfoAtBytecodeOffset):
(JSC::BytecodeLivenessPropagation::runLivenessFixpoint):
(JSC::BytecodeLivenessPropagation<DerivedAnalysis>::stepOverInstruction): Deleted.
(JSC::BytecodeLivenessPropagation<DerivedAnalysis>::computeLocalLivenessForBytecodeOffset): Deleted.
(JSC::BytecodeLivenessPropagation<DerivedAnalysis>::computeLocalLivenessForBlock): Deleted.
(JSC::BytecodeLivenessPropagation<DerivedAnalysis>::getLivenessInfoAtBytecodeOffset): Deleted.
(JSC::BytecodeLivenessPropagation<DerivedAnalysis>::runLivenessFixpoint): Deleted.

10:46 AM Changeset in webkit [221550] by fpizlo@apple.com

25 edits
3 adds in trunk/Tools

WSL should support structs
https://bugs.webkit.org/show_bug.cgi?id=176290

Reviewed by Mylex Maxfield.

This adds parsing of struct types and makes them work.

This required adding a struct layout phase, which in turn required being a bit more careful
about how we instantiate functions and types. We cannot have any instantiation TypeRefs left
once we're doing struct layout.

WebGPUShadingLanguageRI/All.js:
WebGPUShadingLanguageRI/CallFunction.js:

(callFunction):

WebGPUShadingLanguageRI/Checker.js:

(Checker.prototype.visitProtocolDecl.set throw):

WebGPUShadingLanguageRI/DotExpression.js: Added.

(DotExpression):
(DotExpression.prototype.get struct):
(DotExpression.prototype.get fieldName):
(DotExpression.prototype.get isLValue):
(DotExpression.prototype.toString):

WebGPUShadingLanguageRI/EBufferBuilder.js:

(EBufferBuilder.prototype._createEPtr):

WebGPUShadingLanguageRI/EPtr.js:

(EPtr.prototype.copyFrom): Deleted.
(EPtr.prototype.toString): Deleted.

WebGPUShadingLanguageRI/Evaluator.js:

(Evaluator.prototype.visitAssignment):
(Evaluator.prototype.visitDotExpression):

WebGPUShadingLanguageRI/Field.js:

(Field):

WebGPUShadingLanguageRI/FuncDef.js:

(FuncDef.prototype.rewrite):

WebGPUShadingLanguageRI/FuncInstantiator.js:

(FuncInstantiator.prototype.getUnique):
(FuncInstantiator):

WebGPUShadingLanguageRI/Inline.js:

(inline):
(_inlineFunction):
(resolveInlinedFunction):

WebGPUShadingLanguageRI/Inliner.js:

(Inliner.prototype.visitDotExpression):

WebGPUShadingLanguageRI/InstantiateImmediates.js:

(InstantiateImmediates.prototype.visitTypeRef):
(InstantiateImmediates.prototype.visitVariableDecl):
(InstantiateImmediates):

WebGPUShadingLanguageRI/NameResolver.js:

(NameResolver.prototype._resolveTypeArguments):
(NameResolver.prototype.visitTypeRef):

WebGPUShadingLanguageRI/Node.js:

(Node.prototype.visit):

WebGPUShadingLanguageRI/OverloadResolutionFailure.js: Added.

(OverloadResolutionFailure):
(OverloadResolutionFailure.prototype.get func):
(OverloadResolutionFailure.prototype.get reason):
(OverloadResolutionFailure.prototype.toString):

WebGPUShadingLanguageRI/Parse.js:

(parsePossibleSuffix):
(parseField):
(parseStructType):
(parse):

WebGPUShadingLanguageRI/ResolveOverloadImpl.js:

(resolveOverloadImpl):

WebGPUShadingLanguageRI/Rewriter.js:

(Rewriter.prototype.visitDotExpression):
(Rewriter.prototype.visitVariableRef):

WebGPUShadingLanguageRI/StructLayoutBuilder.js: Added.

(StructLayoutBuilder):
(StructLayoutBuilder.prototype.visitStructType):
(StructLayoutBuilder.prototype.get offset):
(StructLayoutBuilder.prototype.visitField):
(StructLayoutBuilder.prototype.visitVariableRef):
(StructLayoutBuilder.prototype.visitVariableDecl):

WebGPUShadingLanguageRI/StructType.js:

(StructType.prototype.add):
(StructType.prototype.instantiate):
(StructType.prototype.populateDefaultValue):

WebGPUShadingLanguageRI/Substitution.js:

(Substitution.prototype.visitTypeRef):
(Substitution.prototype.visitVariableRef):
(Substitution):

WebGPUShadingLanguageRI/Test.js:

(TEST_genericStructInstance):

WebGPUShadingLanguageRI/Type.js:

(Type.prototype.inherits):

WebGPUShadingLanguageRI/TypeRef.js:

(TypeRef.instantiate):

WebGPUShadingLanguageRI/VariableRef.js:

(VariableRef.wrap):

WebGPUShadingLanguageRI/Visitor.js:

(Visitor.prototype.visitTypeRef):
(Visitor.prototype.visitDotExpression):
(Visitor.prototype.visitVariableRef):

10:43 AM Changeset in webkit [221549] by fpizlo@apple.com

21 edits
4 adds in trunk/Tools

WSL NullType should be a special type variable so it can unify with any pointer type
https://bugs.webkit.org/show_bug.cgi?id=176235

Reviewed by Myles Maxfield.

Null is just a type variable that plays by weird rules:

Other type variables don't recognize it, so if it's unified with T, it fails.
It behaves like a type variable for pointer and array reference types.
It arises only in argument (RHS) position.
Once Checker finds it to be equals to something, we need to remember what that was.

I refactored some of the type system code to allow for the possibility of type variables being
unified during equals() tests. It turns out that we were playing fast and loose here; I needed
to beef up UnificationContext.prototype.verify. I added tests for this.

I added an equalsWithCommit() that triggers commitUnification(), which allows NullType to
remember what type it got unified with.

NullLiteral then just instantiates its own NullType, which is what Checker uses as the type of
the NullLiteral. Everything else just works.

This adds a bunch of tests to make sure we generate type errors in various weird cases while
also accepting a number of safe cases.

I also added a way to run the tests in-browser.

WebGPUShadingLanguageRI/All.js:
WebGPUShadingLanguageRI/ArrayRefType.js:

(ArrayRefType.prototype.get isArrayRef):

WebGPUShadingLanguageRI/Checker.js:

(Checker.prototype.visitProtocolDecl.set throw):

WebGPUShadingLanguageRI/ConstexprTypeParameter.js:

(ConstexprTypeParameter.prototype.typeVariableUnify):

WebGPUShadingLanguageRI/Evaluator.js:

(Evaluator.prototype.visitDereferenceExpression):
(Evaluator.prototype.visitNullLiteral):

WebGPUShadingLanguageRI/Func.js:

(Func.prototype.toString):
(Func):

WebGPUShadingLanguageRI/Intrinsics.js:

(Intrinsics):

WebGPUShadingLanguageRI/NativeFunc.js:

(NativeFunc.prototype.toDeclString):
(NativeFunc):
(NativeFunc.prototype.toString): Deleted.

WebGPUShadingLanguageRI/NativeFuncInstance.js:

(NativeFuncInstance.prototype.toDeclString):
(NativeFuncInstance):
(NativeFuncInstance.prototype.toString): Deleted.

WebGPUShadingLanguageRI/Node.js:

(Node.unify):
(Node.prototype._typeVariableUnifyImpl):
(Node.prototype.equals):

WebGPUShadingLanguageRI/NullLiteral.js: Added.

(NullLiteral):
(NullLiteral.prototype.toString):

WebGPUShadingLanguageRI/NullType.js:

(NullType):
(NullType.prototype.get isPrimitive):
(NullType.prototype.get isUnifiable):
(NullType.prototype.typeVariableUnify):
(NullType.prototype.unifyImpl):
(NullType.prototype.verifyAsArgument):
(NullType.prototype.verifyAsParameter):
(NullType.prototype.toString):

WebGPUShadingLanguageRI/Parse.js:

(parseTerm):
(parseType.getAddressSpace):
(parseType):

WebGPUShadingLanguageRI/Rewriter.js:

(Rewriter.prototype.visitNullLiteral):

WebGPUShadingLanguageRI/Test.html: Added.
WebGPUShadingLanguageRI/Test.js:

(this.window.this.print):
(TEST_doubleGenericCallsSingleGeneric):
(TEST_dereferenceDefaultNull):
(TEST_dereferenceDefaultNullArrayRef):
(doTest):
(let.s.in.this.s.startsWith.string_appeared_here.s.match): Deleted.

WebGPUShadingLanguageRI/Type.js:

(Type.prototype.get isArrayRef):

WebGPUShadingLanguageRI/TypeVariable.js:

(TypeVariable.prototype.typeVariableUnify):

WebGPUShadingLanguageRI/UnificationContext.js:

(UnificationContext.prototype.verify):
(UnificationContext):
(UnificationContext.prototype.isUnunified): Deleted.

WebGPUShadingLanguageRI/VariableRef.js:

(VariableRef.prototype.get unifyNode):

WebGPUShadingLanguageRI/Visitor.js:

(Visitor.prototype.visitNullLiteral):

3:31 AM Changeset in webkit [221548] by Yusuke Suzuki

3 edits in trunk/Source/bmalloc

Large virtual memory region allocation requires MMAP_NORESERVE in Linux
https://bugs.webkit.org/show_bug.cgi?id=176211

Reviewed by Geoffrey Garen.

In Linux, we cannot allocate very large memory region without MMAP_NORESERVE.
Linux kernel needs to reserve swap area for allocated memory region. If the
swap area is exhausted, kernel fails to allocate the memory region with ENOMEM.

This patch adds MMAP_NORESERVE to mmap flags in Linux. By adding this flag,
mmap does not need to reserve swap area for the reserved memory region.
This allows us to reserve very large memory region that is necessary for Gigacage.

bmalloc/BPlatform.h:
bmalloc/VMAllocate.h:

(bmalloc::tryVMAllocate):

1:42 AM Changeset in webkit [221547] by commit-queue@webkit.org

25 edits
3 deletes in trunk

Remove CanvasProxy
https://bugs.webkit.org/show_bug.cgi?id=176288

Patch by Sam Weinig <sam@webkit.org> on 2017-09-03
Reviewed by Yusuke Suzuki.

CanvasProxy does not appear to be in any current HTML spec
and was disabled and unimplemented in our tree. Time to
get rid of it.

Source/cmake/OptionsMac.cmake:
Source/cmake/OptionsWin.cmake:
Source/cmake/WebKitFeatures.cmake:
Source/cmake/tools/vsprops/FeatureDefines.props:
Source/cmake/tools/vsprops/FeatureDefinesCairo.props:

Source/JavaScriptCore:

Configurations/FeatureDefines.xcconfig:

Source/WebCore:

CMakeLists.txt:
Configurations/FeatureDefines.xcconfig:
DerivedSources.cpp:
DerivedSources.make:
WebCore.xcodeproj/project.pbxproj:
html/canvas/CanvasProxy.cpp: Removed.
html/canvas/CanvasProxy.h: Removed.
html/canvas/CanvasProxy.idl: Removed.

Source/WebCore/PAL:

Configurations/FeatureDefines.xcconfig:

Source/WebKit:

Configurations/FeatureDefines.xcconfig:

Source/WebKitLegacy/mac:

Configurations/FeatureDefines.xcconfig:

Source/WTF:

wtf/FeatureDefines.h:

Tools:

Scripts/webkitperl/FeatureList.pm:
TestWebKitAPI/Configurations/FeatureDefines.xcconfig:

Note: See TracTimeline for information about the timeline view.

Download in other formats:

RSS Feed