Fix CMake build.

Source/WebKit2:

* PlatformMac.cmake:

Tools:

* DumpRenderTree/PlatformMac.cmake:
Define NS_RETURNS_RETAINED as nothing for now.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209786 268f45cc-cd09-0410-ab3c-d52691b4dbfc

WebAssembly: implement the elements section
https://bugs.webkit.org/show_bug.cgi?id=165715

Reviewed by Keith Miller.

JSTests:

* wasm/Builder.js:
(export.default.Builder.prototype._registerSectionBuilders.switch.case.string_appeared_here.this.section):
(export.default.Builder.prototype._registerSectionBuilders.switch):
* wasm/Builder_WebAssemblyBinary.js:
(const.emitters.Element):
* wasm/function-tests/basic-element.js: Added.
* wasm/js-api/element.js: Added.
(assertBadBinary):
(assertBadBinary.badInstantiation):

Source/JavaScriptCore:

This is a straight forward implementation of the Element
section in the Wasm spec:
https://github.com/WebAssembly/design/blob/master/BinaryEncoding.md#element-section

There are a few ambiguities I encountered when implementing this, so I've
filed bugs against the Wasm design repo, and corresponding bugzilla bugs
for us to address after they've been discussed by the various Wasm folks:
- https://bugs.webkit.org/show_bug.cgi?id=165827
- https://bugs.webkit.org/show_bug.cgi?id=165826
- https://bugs.webkit.org/show_bug.cgi?id=165825

* wasm/WasmFormat.h:
* wasm/WasmModuleParser.cpp:
(JSC::Wasm::ModuleParser::parseElement):
(JSC::Wasm::ModuleParser::parseInitExpr):
(JSC::Wasm::ModuleParser::parseData):
* wasm/WasmModuleParser.h:
* wasm/js/WebAssemblyModuleRecord.cpp:
(JSC::WebAssemblyModuleRecord::evaluate):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209785 268f45cc-cd09-0410-ab3c-d52691b4dbfc

REGRESSION(r204163): Web Inspector: Page crashes when Inspector tries to load insecure SourceMap
https://bugs.webkit.org/show_bug.cgi?id=165806
<rdar://problem/28169025>

Patch by Joseph Pecoraro <pecoraro@apple.com> on 2016-12-13
Reviewed by Alex Christensen.

Source/WebCore:

Test: http/tests/inspector/network/loadResource-insecure-resource.html

* inspector/InspectorNetworkAgent.cpp:
(WebCore::InspectorNetworkAgent::loadResource):
DocumentThreadableLoader now triggers the ThreadableLoaderClient's didFail
callback, so this didFailLoaderCreation is now redundent and incorrect.

LayoutTests:

* http/tests/inspector/network/loadResource-insecure-resource-expected.txt: Added.
* http/tests/inspector/network/loadResource-insecure-resource.html: Added.
* http/tests/inspector/network/resource-timing.html:
* http/tests/inspector/resources/inspector-test.js:
(runTestHTTPS):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209784 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[CSS Parser] Enhance fast path translate transforms to allow percentages
https://bugs.webkit.org/show_bug.cgi?id=165822

Reviewed by Simon Fraser.

* css/parser/CSSParserFastPaths.cpp:
(WebCore::parseTransformTranslateArguments):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209783 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fullscreen in WebKit2 does not restore topContentInset upon exiting; leaves top of page not visible
https://bugs.webkit.org/show_bug.cgi?id=165697

Source/WebKit2:

Delegate the values of topContentInset() from WebViewImpl (used by WKWebView and WKView) to the
WebPageProxy, so that setting the topContentInset() on WebPageProxy is reflected in the getters
for the two view classes.

Reviewed by Tim Horton.

* UIProcess/Cocoa/WebViewImpl.h:
(WebKit::WebViewImpl::topContentInset): Deleted.
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::updateContentInsetsIfAutomatic):
(WebKit::WebViewImpl::topContentInset):
(WebKit::WebViewImpl::setTopContentInset):
(WebKit::WebViewImpl::dispatchSetTopContentInset):
* UIProcess/mac/WKFullScreenWindowController.mm:
(-[WKFullScreenWindowController enterFullScreen:]):
(-[WKFullScreenWindowController finishedEnterFullScreenAnimation:]):
(-[WKFullScreenWindowController finishedExitFullScreenAnimation:]):
(-[WKFullScreenWindowController _saveTopContentInset]): Deleted.
(-[WKFullScreenWindowController _restoreTopContentInset]): Deleted.

Tools:

Reviewed by Tim Horton.

* TestWebKitAPI/Tests/WebKit2Cocoa/FullscreenTopContentInset.mm:
(TestWebKitAPI::TEST):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209782 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Rebaseline js/dom/global-constructors-attributes.html after r209777.

Unreviewed test gardening.

* platform/mac-wk1/js/dom/global-constructors-attributes-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209781 268f45cc-cd09-0410-ab3c-d52691b4dbfc

:hover rule causes a single tap to not activate a slotted anchor element
https://bugs.webkit.org/show_bug.cgi?id=165551

Reviewed by Antti Koivisto.

Source/WebCore:

Fixed a bug in ancestorRespondingToClickEvents that we were traversing the ancestor nodes without taking
shadow roots and slots into account. This prevented tapping on a text node assigned to a slot inside
an anchor element to activate the hyperlink on iOS.

This bug was supposed to be fixed in r206605, and it was still broken on iOS due to the bug in
ancestorRespondingToClickEvents. It is now tested by click-text-inside-linked-slot.html.

Tests: fast/shadow-dom/click-on-slotted-anchor-with-hover.html
       fast/shadow-dom/click-text-inside-linked-slot.html

* page/ios/FrameIOS.mm:
(WebCore::ancestorRespondingToClickEvents):
(WebCore::Frame::qualifyingNodeAtViewportLocation):

LayoutTests:

Added a test for tapping on an anchor element assigned to a slot, which has been fixed in r209065.

Also added a new helper JS wrapepr, UIHelper, defined inside LayoutTests/resources/js-helper.js
to provide an abstraction around EventSender and UIScriptController.

Fixed click-text-inside-linked-slot.html on iOS using UIHelper.

* fast/shadow-dom/click-on-slotted-anchor-with-hover-expected.txt: Added.
* fast/shadow-dom/click-on-slotted-anchor-with-hover.html: Added.
* fast/shadow-dom/click-text-inside-linked-slot.html:
* platform/ios-simulator/fast/shadow-dom/click-text-inside-linked-slot-expected.txt: Added.
* resources/ui-helper.js: Added.
(window.UIHelper.isIOS):
(window.UIHelper.activateAt.return.new.Promise):
(window.UIHelper.activateAt):
(window.UIHelper.wait):
(window.UIHelper):
* platform/ios-simulator-wk2/TestExpectations: Skip the test in the open source iOS's WebKit2.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209780 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Mac][WK2] Tighten Keychain directory access
https://bugs.webkit.org/show_bug.cgi?id=165818
<rdar://problem/16863857>

Reviewed by Anders Carlsson.

Lock down Keychain directory access to just the file-read-data, file-read-metadata, and
file-write-data operations we actually need.

* NetworkProcess/mac/com.apple.WebKit.NetworkProcess.sb.in:
* PluginProcess/mac/com.apple.WebKit.plugin-common.sb.in:
* WebProcess/com.apple.WebProcess.sb.in:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209779 268f45cc-cd09-0410-ab3c-d52691b4dbfc

std::expected: fix rvalue forwarding issues
https://bugs.webkit.org/show_bug.cgi?id=165812

Reviewed by Mark Lam.

* wtf/Expected.h:
(WTF::UnexpectedType::UnexpectedType):
(WTF::ExpectedDetail::Base::Base):
(WTF::Expected::Expected):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209778 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[CSS Parser] Eliminate SVGPaint and SVGColor
https://bugs.webkit.org/show_bug.cgi?id=165819

Reviewed by Dean Jackson.

Source/WebCore:

Remove SVGColor and SVGPaint CSS values from the tree. The new parser
already stopped making these values when parsing, so all that's left
is to make sure computed style doesn't use these values either.

We are the only browser engine to support these CSS values, they see no use
on the Web, and they're already gone at the parsing level, so it's time
to remove these interfaces from our tree.

* CMakeLists.txt:
* DerivedSources.make:
* WebCore.xcodeproj/project.pbxproj:
* bindings/js/JSCSSValueCustom.cpp:
(WebCore::toJSNewlyCreated):
* css/CSSComputedStyleDeclaration.h:
* css/CSSValue.cpp:
(WebCore::CSSValue::equals):
(WebCore::CSSValue::cssText):
(WebCore::CSSValue::destroy):
(WebCore::CSSValue::cloneForCSSOM):
* css/CSSValue.h:
(WebCore::CSSValue::isSubtypeExposedToCSSOM):
(WebCore::CSSValue::isSVGColor): Deleted.
(WebCore::CSSValue::isSVGPaint): Deleted.
* css/SVGCSSComputedStyleDeclaration.cpp:
(WebCore::ComputedStyleExtractor::adjustSVGPaintForCurrentColor):
(WebCore::ComputedStyleExtractor::svgPropertyValue):
* css/StyleBuilderConverter.h:
(WebCore::StyleBuilderConverter::convertSVGColor):
* css/StyleBuilderCustom.h:
(WebCore::StyleBuilderCustom::applyValueFill):
(WebCore::StyleBuilderCustom::applyValueStroke):
* page/animation/CSSPropertyAnimation.cpp:
(WebCore::PropertyWrapperSVGPaint::PropertyWrapperSVGPaint):
* rendering/style/RenderStyle.h:
(WebCore::RenderStyle::fillPaintType):
(WebCore::RenderStyle::setFillPaintColor):
(WebCore::RenderStyle::strokePaintType):
(WebCore::RenderStyle::setStrokePaintColor):
* rendering/style/SVGRenderStyle.h:
(WebCore::SVGRenderStyle::initialFillPaintType):
(WebCore::SVGRenderStyle::initialStrokePaintType):
(WebCore::SVGRenderStyle::setFillPaint):
(WebCore::SVGRenderStyle::setStrokePaint):
(WebCore::SVGRenderStyle::fillPaintType):
(WebCore::SVGRenderStyle::strokePaintType):
(WebCore::SVGRenderStyle::visitedLinkFillPaintType):
(WebCore::SVGRenderStyle::visitedLinkStrokePaintType):
(WebCore::SVGRenderStyle::hasStroke):
(WebCore::SVGRenderStyle::hasFill):
* rendering/style/SVGRenderStyleDefs.h:
* rendering/svg/RenderSVGResource.cpp:
(WebCore::requestPaintingResource):
* rendering/svg/SVGResources.cpp:
(WebCore::paintingResourceFromSVGPaint):
* svg/SVGAllInOne.cpp:
* svg/SVGAnimatedColor.cpp:
(WebCore::SVGAnimatedColorAnimator::constructFromString):
(WebCore::parseColorFromString):
(WebCore::SVGAnimatedColorAnimator::calculateDistance):
* svg/SVGAnimatedType.cpp:
(WebCore::SVGAnimatedType::setValueAsString):
* svg/SVGAnimatedType.h:
* svg/SVGColor.cpp: Removed.
* svg/SVGColor.h: Removed.
* svg/SVGColor.idl: Removed.
* svg/SVGFEDiffuseLightingElement.cpp:
* svg/SVGFESpecularLightingElement.cpp:
* svg/SVGPaint.cpp: Removed.
* svg/SVGPaint.h: Removed.
* svg/SVGPaint.idl: Removed.

LayoutTests:

* fast/dom/Window/resources/window-properties.js:
* platform/mac-yosemite/js/dom/global-constructors-attributes-expected.txt:
* platform/mac/fast/dom/Window/window-lookup-precedence-expected.txt:
* platform/mac/js/dom/global-constructors-attributes-expected.txt:
* svg/animations/additive-type-by-animation-expected.txt:
* svg/animations/animate-color-calcMode-discrete-expected.txt:
* svg/animations/animate-color-fill-currentColor-expected.txt:
* svg/animations/animate-color-fill-from-by-expected.txt:
* svg/animations/animate-color-transparent-expected.txt:
* svg/animations/animate-currentColor-expected.txt:
* svg/animations/animate-dynamic-update-attributeName-expected.txt:
* svg/animations/animate-elem-03-t-drt-expected.txt:
* svg/animations/animate-inherit-css-property-expected.txt:
* svg/animations/attributeTypes-expected.txt:
* svg/animations/resources/SVGAnimationTestCase.js:
(expectColor):
* svg/animations/script-tests/animate-color-fill-currentColor.js:
(sample1):
(sample2):
(sample3):
* svg/animations/script-tests/animate-color-fill-from-by.js:
(sample1):
(sample2):
(sample3):
* svg/css/getComputedStyle-basic-expected.txt:
* svg/css/rect-system-color-expected.txt:
* svg/css/rect-system-color.xhtml:
* svg/custom/global-constructors-expected.txt:
* svg/custom/js-svg-constructors-expected.txt:
* svg/custom/js-svg-constructors.svg:
* svg/custom/script-tests/global-constructors.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209777 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Restore NSURLRequest's default time interval to match behavior before NSURLSession adoption
https://bugs.webkit.org/show_bug.cgi?id=165821
<rdar://problem/28492939>

Reviewed by Brady Eidson.

Before adopting NSURLSession, iOS used CFURLConnection, not NSURLConnection.
iOS used to have a default timeout of INT_MAX and it now has a default timeout of 0, which means use the
default NSURLRequest timeout, which is 60 seconds.  This is not enough for some slow mobile networks,
so we want to match behavior of our CFURLConnection code here.

* platform/network/ResourceRequestBase.cpp:
Use INT_MAX as the default timeout of requests on iOS.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209776 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Cocoa] Implement -shouldInsertText: on WKWebProcessPlugInEditingDelegate
https://bugs.webkit.org/show_bug.cgi?id=165785
<rdar://problem/26929398>

Reviewed by Darin Adler.

Source/WebKit2:

* WebProcess/InjectedBundle/API/Cocoa/WKWebProcessPlugInEditingDelegate.h: Defined
WKEditorInsertAction and declared
-_webProcessPlugInBrowserContextController:shouldInsertText:replacingRange:givenAction:.
* WebProcess/InjectedBundle/API/mac/WKWebProcessPlugInBrowserContextController.mm:
(toWK): Added. Mapped a WebCore::EditorInsertAction to a WKEditorInsertAction.
(-[WKWebProcessPlugInBrowserContextController _setEditingDelegate:]): Implemented
EditorClient::shouldInsertText() by calling
-_webProcessPlugInBrowserContextController:shouldInsertText:replacingRange:givenAction: with
the text, wrapped range handle, and action. Added shouldInsertText to DelegateMethods and
initialized it according to whether the delegate responds to shouldInsertText:.

Tools:

* TestWebKitAPI/Tests/WebKit2Cocoa/BundleEditingDelegate.mm:
(-[BundleEditingDelegateRemoteObject shouldInsertText:replacingRange:givenAction:]): Added
expectations for the text, range, and action.
(TEST): Set the "EditingDelegateShouldInsertText" bundle parameter to NO so that the plug-in
will return NO in -_webProcessPlugInBrowserContextController:willInsertText:. Added a
plain-text string to the pasteboard, executed a paste action, and verified that the selected
text in the web view did not change.
* TestWebKitAPI/Tests/WebKit2Cocoa/BundleEditingDelegatePlugIn.mm:
(-[BundleEditingDelegatePlugIn webProcessPlugIn:didCreateBrowserContextController:]): Set
_editingDelegateShouldInsertText according to the "EditingDelegateShouldInsertText" bundle
parameter.
(-[BundleEditingDelegatePlugIn _webProcessPlugInBrowserContextController:shouldInsertText:replacingRange:givenAction:]):
Added. Called -shouldInsertText:replacingRange:givenAction: on the remote object proxy
and returned the value of _editingDelegateShouldInsertText.
* TestWebKitAPI/Tests/WebKit2Cocoa/BundleEditingDelegateProtocol.h: Declared
-shouldInsertText:replacingRange:givenAction:.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209775 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Rebaseline js/dom/global-constructors-attributes.html after r209760.

Unreviewed test gardening.

* platform/mac/js/dom/global-constructors-attributes-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209774 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, rolling out r209544.

Looks like r209489 did not cause the performance regression
after all

Reverted changeset:

"Unreviewed, rolling out r209489."
https://bugs.webkit.org/show_bug.cgi?id=165550
http://trac.webkit.org/changeset/209544

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209773 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Add a PerfTest targeting IDBObjectStore.get().
https://bugs.webkit.org/show_bug.cgi?id=165816

Reviewed by Alex Christensen.

* IndexedDB/objectstore-get.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209772 268f45cc-cd09-0410-ab3c-d52691b4dbfc

WebAssembly: implement the table section and table import
https://bugs.webkit.org/show_bug.cgi?id=165716

Reviewed by Keith Miller.

JSTests:

* wasm/Builder.js:
(const._importMemoryContinuation):
(const._importTableContinuation):
(export.default.Builder.prototype._registerSectionBuilders.switch.case.string_appeared_here.this.section):
(const._importMemoryContinuation.section): Deleted.
(const): Deleted.
(const._importMemoryContinuation.assert): Deleted.
* wasm/Builder_WebAssemblyBinary.js:
(const.putResizableLimits):
(const.putTable):
(const.emitters.Import):
(const.emitters.Table):
* wasm/function-tests/call-indirect-params.js:
* wasm/function-tests/call-indirect.js:
* wasm/function-tests/table-basic-2.js: Added.
(import.Builder.from.string_appeared_here.import.as.assert.from.string_appeared_here.makeInstance):
(func):
* wasm/function-tests/table-basic.js: Added.
(import.Builder.from.string_appeared_here.import.as.assert.from.string_appeared_here.makeInstance):
* wasm/js-api/call-indirect-results.js:
(const.wasmModuleWhichImportJS): Deleted.
(MonomorphicImport): Deleted.
* wasm/js-api/call-indirect.js:
(const.wasmModuleWhichImportJS):
(const.makeTable):
(Polyphic2Import):
(VirtualImport):
(MonomorphicImport): Deleted.
* wasm/js-api/table.js: Added.
(assertBadBinary):
(assert.truthy):
(assertBadTable):
(assertBadTableImport):
(assertBadBinary.assertBadTableInstance):
(assertBadTableInstance):
(new.WebAssembly.Table):
* wasm/js-api/test_basic_api.js:
(const.c.in.constructorProperties.switch):

Source/JavaScriptCore:

This patch implements the Table space for wasm:
https://github.com/WebAssembly/design/blob/master/BinaryEncoding.md#table-section

It only implements defining and importing a table. The bulk
of this patch is implementing the various wasm Table prototype
methods and the underlying Table object:
https://github.com/WebAssembly/design/blob/master/JS.md#webassemblytable-constructor

This patch also fixes a bug in our implementation with call_indirect.
We initially implemented call_indirect as a way to call functions that
are imported or defined in the module. This was the wrong
interpretation of the spec. Instead, call_indirect can only index into
the table index space.

* JavaScriptCore.xcodeproj/project.pbxproj:
* wasm/WasmB3IRGenerator.cpp:
(JSC::Wasm::B3IRGenerator::B3IRGenerator):
(JSC::Wasm::B3IRGenerator::addCallIndirect):
(JSC::Wasm::parseAndCompile):
* wasm/WasmFormat.h:
(JSC::Wasm::TableInformation::TableInformation):
(JSC::Wasm::TableInformation::operator bool):
(JSC::Wasm::TableInformation::isImport):
(JSC::Wasm::TableInformation::initial):
(JSC::Wasm::TableInformation::maximum):
(JSC::Wasm::CallableFunction::CallableFunction):
* wasm/WasmFunctionParser.h:
(JSC::Wasm::FunctionParser<Context>::parseExpression):
* wasm/WasmModuleParser.cpp:
(JSC::Wasm::ModuleParser::parseImport):
(JSC::Wasm::ModuleParser::parseResizableLimits):
(JSC::Wasm::ModuleParser::parseTableHelper):
(JSC::Wasm::ModuleParser::parseTable):
(JSC::Wasm::ModuleParser::parseMemoryHelper):
(JSC::Wasm::ModuleParser::parseExport):
* wasm/WasmModuleParser.h:
* wasm/js/JSWebAssemblyHelpers.h: Added.
(JSC::toNonWrappingUint32):
* wasm/js/JSWebAssemblyInstance.cpp:
(JSC::JSWebAssemblyInstance::visitChildren):
* wasm/js/JSWebAssemblyInstance.h:
(JSC::JSWebAssemblyInstance::table):
(JSC::JSWebAssemblyInstance::setTable):
(JSC::JSWebAssemblyInstance::offsetOfTable):
* wasm/js/JSWebAssemblyTable.cpp:
(JSC::JSWebAssemblyTable::create):
(JSC::JSWebAssemblyTable::JSWebAssemblyTable):
(JSC::JSWebAssemblyTable::visitChildren):
(JSC::JSWebAssemblyTable::grow):
(JSC::JSWebAssemblyTable::clearFunction):
(JSC::JSWebAssemblyTable::setFunction):
* wasm/js/JSWebAssemblyTable.h:
(JSC::JSWebAssemblyTable::maximum):
(JSC::JSWebAssemblyTable::size):
(JSC::JSWebAssemblyTable::getFunction):
(JSC::JSWebAssemblyTable::offsetOfSize):
(JSC::JSWebAssemblyTable::offsetOfFunctions):
(JSC::JSWebAssemblyTable::isValidSize):
* wasm/js/WebAssemblyFunction.cpp:
(JSC::WebAssemblyFunction::call):
(JSC::WebAssemblyFunction::create):
(JSC::WebAssemblyFunction::visitChildren):
(JSC::WebAssemblyFunction::finishCreation):
* wasm/js/WebAssemblyFunction.h:
(JSC::WebAssemblyFunction::signature):
(JSC::WebAssemblyFunction::wasmEntrypoint):
(JSC::WebAssemblyFunction::webAssemblyCallee): Deleted.
* wasm/js/WebAssemblyInstanceConstructor.cpp:
(JSC::constructJSWebAssemblyInstance):
* wasm/js/WebAssemblyMemoryConstructor.cpp:
(JSC::constructJSWebAssemblyMemory):
* wasm/js/WebAssemblyModuleRecord.cpp:
(JSC::WebAssemblyModuleRecord::finishCreation):
(JSC::WebAssemblyModuleRecord::link):
* wasm/js/WebAssemblyTableConstructor.cpp:
(JSC::constructJSWebAssemblyTable):
* wasm/js/WebAssemblyTablePrototype.cpp:
(JSC::getTable):
(JSC::webAssemblyTableProtoFuncLength):
(JSC::webAssemblyTableProtoFuncGrow):
(JSC::webAssemblyTableProtoFuncGet):
(JSC::webAssemblyTableProtoFuncSet):
(JSC::WebAssemblyTablePrototype::create):
(JSC::WebAssemblyTablePrototype::finishCreation):
* wasm/js/WebAssemblyTablePrototype.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209771 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remove ios-simulator-wk2 pass expectation for a directory that requires touch events.
https://bugs.webkit.org/show_bug.cgi?id=165787

Unreviewed test gardening.

* platform/ios-simulator-wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209770 268f45cc-cd09-0410-ab3c-d52691b4dbfc

NSArray leaks seen in Safari, allocated under WKIconDatabaseTryCopyCGImageArrayForURL
https://bugs.webkit.org/show_bug.cgi?id=165809

Patch by Joseph Pecoraro <pecoraro@apple.com> on 2016-12-13
Reviewed by Dan Bernstein.

* UIProcess/API/C/cg/WKIconDatabaseCG.cpp:
(WKIconDatabaseTryCopyCGImageArrayForURL):
Avoid an extra retain on already newly created array.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209769 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Marking two scroll-to-fragid tests as flaky on ios-simulator.

Unreviewed test gardening.

* platform/ios-simulator-wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209768 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Add null checks to opaque root APIs.

Rubber stamped by Saam Barati.

If we got a crash report about null in the opaque root HashSet, we would probably not
celebrate how great it is that we found out about a new race - instead we would probably
be annoyed that null wasn't just silently ignored.

* heap/SlotVisitor.cpp:
(JSC::SlotVisitor::addOpaqueRoot):
(JSC::SlotVisitor::containsOpaqueRoot):
(JSC::SlotVisitor::containsOpaqueRootTriState):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209767 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Make opaque root scanning truly constraint-based
https://bugs.webkit.org/show_bug.cgi?id=165760

Reviewed by Saam Barati.
Source/JavaScriptCore:

We have bugs when visitChildren() changes its mind about what opaque root to add, since
we don't have barriers on opaque roots. This supposedly once worked for generational GC,
and I started adding more barriers to support concurrent GC. But I think that the real
bug here is that we want the JSObject->OpaqueRoot to be evaluated as a constraint that
participates in the fixpoint. A constraint is different from the normal visiting in that
the GC will not wait for a barrier to rescan the object.

So, it's now possible for any visitChildren() method to become a constraint by calling
slotVisitor.rescanAsConstraint(). Because opaque roots are constraints, addOpaqueRoot()
does rescanAsConstraint() for you.

The constraint set is simply a HashSet<JSCell*> that accumulates with every
rescanAsConstraint() call and is only cleared at the start of full GC. This trivially
resolves most classes of GC bugs that would have arisen from opaque roots being changed
in a way that the GC did not anticipate.

Looks like this is perf-neutral.

* heap/Heap.cpp:
(JSC::Heap::markToFixpoint):
(JSC::Heap::setMutatorShouldBeFenced):
(JSC::Heap::writeBarrierOpaqueRootSlow): Deleted.
(JSC::Heap::addMutatorShouldBeFencedCache): Deleted.
* heap/Heap.h:
* heap/HeapInlines.h:
(JSC::Heap::writeBarrierOpaqueRoot): Deleted.
* heap/MarkedSpace.cpp:
(JSC::MarkedSpace::visitWeakSets):
* heap/MarkedSpace.h:
* heap/SlotVisitor.cpp:
(JSC::SlotVisitor::visitChildren):
(JSC::SlotVisitor::visitSubsequently):
(JSC::SlotVisitor::drain):
(JSC::SlotVisitor::addOpaqueRoot):
(JSC::SlotVisitor::rescanAsConstraint):
(JSC::SlotVisitor::mergeIfNecessary):
(JSC::SlotVisitor::mergeOpaqueRootsAndConstraints):
(JSC::SlotVisitor::mergeOpaqueRootsIfNecessary): Deleted.
* heap/SlotVisitor.h:
* heap/SlotVisitorInlines.h:
(JSC::SlotVisitor::reportExtraMemoryVisited):
(JSC::SlotVisitor::reportExternalMemoryVisited):
(JSC::SlotVisitor::didNotRace):
* heap/WeakBlock.cpp:
(JSC::WeakBlock::specializedVisit):
(JSC::WeakBlock::visit):
* heap/WeakBlock.h:
* heap/WeakSet.h:
(JSC::WeakSet::visit):

Source/WebCore:

No new tests yet. I think that writing tests for this is a big investigation:
https://bugs.webkit.org/show_bug.cgi?id=165808

Remove the previous advancing wavefront DOM write barrier. I don't think this will scale
very well. It's super confusing.

This change makes it so that visitChildren can become a GC constraint that executes as
part of the fixpoint. This changes all WebCore visitChildren methods that do opaque
roots into constraints.

* bindings/js/CommonVM.cpp:
(WebCore::commonVMSlow):
(WebCore::writeBarrierOpaqueRootSlow): Deleted.
* bindings/js/CommonVM.h:
(WebCore::writeBarrierOpaqueRoot): Deleted.
* bindings/js/JSAttrCustom.cpp:
(WebCore::JSAttr::visitAdditionalChildren):
* bindings/js/JSDOMWindowCustom.cpp:
(WebCore::JSDOMWindow::visitAdditionalChildren):
* bindings/js/JSIDBCursorCustom.cpp:
(WebCore::JSIDBCursor::visitAdditionalChildren):
* bindings/js/JSMessageChannelCustom.cpp:
(WebCore::JSMessageChannel::visitAdditionalChildren):
* bindings/js/JSMessagePortCustom.cpp:
(WebCore::JSMessagePort::visitAdditionalChildren):
* bindings/js/JSNodeIteratorCustom.cpp:
(WebCore::JSNodeIterator::visitAdditionalChildren):
* bindings/js/JSTextTrackCueCustom.cpp:
(WebCore::JSTextTrackCue::visitAdditionalChildren):
* bindings/js/JSTreeWalkerCustom.cpp:
(WebCore::JSTreeWalker::visitAdditionalChildren):
* bindings/js/JSWorkerGlobalScopeCustom.cpp:
(WebCore::JSWorkerGlobalScope::visitAdditionalChildren):
* bindings/js/JSXMLHttpRequestCustom.cpp:
(WebCore::JSXMLHttpRequest::visitAdditionalChildren):
* bindings/js/JSXPathResultCustom.cpp:
(WebCore::JSXPathResult::visitAdditionalChildren):
* dom/ContainerNodeAlgorithms.cpp:
(WebCore::notifyChildNodeInserted):
(WebCore::notifyChildNodeRemoved):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209766 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Add a PerfTest targeting IDBIndex.get().
https://bugs.webkit.org/show_bug.cgi?id=165803

Reviewed by Alex Christensen.

* IndexedDB/index-get.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209765 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, rolling out r209725.
https://bugs.webkit.org/show_bug.cgi?id=165811

"Broke ARMv7 builds" (Requested by msaboff on #webkit).

Reverted changeset:

"REGRESSION(r209653): speedometer crashes making virtual slow
path tailcalls"
https://bugs.webkit.org/show_bug.cgi?id=165748
http://trac.webkit.org/changeset/209725

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209764 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, revert the collectorPermittedIdleRatio back to 0 because of 100MB
regression on membuster. Also, it didn't seem to help perf.

* runtime/Options.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209763 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Skip fast/events/page-visibility-onvisibilitychange.html on ios-simulator.
https://bugs.webkit.org/show_bug.cgi?id=165799

Unreviewed test gardening.

* platform/ios-simulator/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209762 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[WTF] Turn tryMakeString(), makeString() into variadic templates
https://bugs.webkit.org/show_bug.cgi?id=147142

Reviewed by Mark Lam.

Source/JavaScriptCore:

* runtime/JSStringBuilder.h:
(JSC::jsMakeNontrivialString): remove WTF:: prefix, it isn't needed anymore
* runtime/Lookup.cpp:
(JSC::reifyStaticAccessor): remove WTF:: prefix, it isn't needed anymore
* runtime/ObjectPrototype.cpp:
(JSC::objectProtoFuncToString): remove WTF:: prefix, it isn't needed anymore

Source/WTF:

I wrote this patch while improving WebAssembly's error messages,
and only found this bug afterwards. My implementation does the
bare minimum to make this code variadic without changing
behavior. I think it's better to go with this baby step first, and
improve the code later.

Notable, for my WebAssembly patch I also taught the code to handle
integers and other types (including WebAssembly types). A
follow-up could rely on ADL magic to pretty-format these other
types.

* wtf/text/StringConcatenate.h:
(WTF::sumWithOverflow): This unconditionally does the sum for all
inputs, which the compiler is more likely to appreciate (because
it's the common case) compared to testing for overflow and bailing
on each addition
(WTF::are8Bit): are: the plural of is!
(WTF::makeStringAccumulator): accumulate strings
(WTF::tryMakeStringFromAdapters): a small helper which creates the string adapters
(WTF::tryMakeString): expose out of WTF, since it's part of this file's used API
(WTF::makeString): make it variadic

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209761 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[WebIDL] Remove use of Dictionary in ApplePaySession
https://bugs.webkit.org/show_bug.cgi?id=165787

Reviewed by Anders Carlsson.

Source/WebCore:

First take at generating the bindings for ApplePaySession and removing
all use of Dictionary.

- Removes all use of Dictionary!
- Removes use of logging errors to the console with improved Exception messages.
- Use ExceptionOr extensively to pass exception state.

Still to do:
    - Reconcile / merge ApplePaySession::PaymentRequest with WebCore::PaymentRequest
      and all the sub-objects held within.
    - Remove PaymentRequestValidator entirely, merging validation into the validation
      we already do in ApplePaySession.cpp
    - Make ApplePayEvents use generated dictionary creation code.

Test: http/tests/ssl/applepay/ApplePaySession.html

* Modules/applepay/ApplePaySession.cpp:
(WebCore::convertAndValidate):
(WebCore::canCallApplePaySessionAPIs):
(WebCore::ApplePaySession::create):
(WebCore::ApplePaySession::ApplePaySession):
(WebCore::ApplePaySession::supportsVersion):
(WebCore::ApplePaySession::canMakePayments):
(WebCore::ApplePaySession::canMakePaymentsWithActiveCard):
(WebCore::ApplePaySession::openPaymentSetup):
(WebCore::ApplePaySession::begin):
(WebCore::ApplePaySession::completeMerchantValidation):
(WebCore::ApplePaySession::completeShippingMethodSelection):
(WebCore::ApplePaySession::completeShippingContactSelection):
(WebCore::ApplePaySession::completePaymentMethodSelection):
(WebCore::ApplePaySession::didSelectShippingMethod):
(WebCore::createContactFields): Deleted.
(WebCore::toLineItemType): Deleted.
(WebCore::isValidLineItemPropertyName): Deleted.
(WebCore::createLineItem): Deleted.
(WebCore::createLineItems): Deleted.
(WebCore::createMerchantCapabilities): Deleted.
(WebCore::createSupportedNetworks): Deleted.
(WebCore::toShippingType): Deleted.
(WebCore::isValidShippingMethodPropertyName): Deleted.
(WebCore::createShippingMethod): Deleted.
(WebCore::createShippingMethods): Deleted.
(WebCore::isValidPaymentRequestPropertyName): Deleted.
(WebCore::createPaymentRequest): Deleted.
* Modules/applepay/ApplePaySession.h:
Replace hand written dictionary extraction code with autogenerated code
and hand written additional validation.

* Modules/applepay/ApplePaySession.idl:
Add helper dictionaries and enums.

* Modules/applepay/PaymentRequestValidator.cpp:
(WebCore::PaymentRequestValidator::validate):
(WebCore::PaymentRequestValidator::validateTotal):
(WebCore::validateCountryCode):
(WebCore::validateCurrencyCode):
(WebCore::validateMerchantCapabilities):
(WebCore::validateSupportedNetworks):
(WebCore::validateShippingMethod):
(WebCore::validateShippingMethods):
(WebCore::PaymentRequestValidator::PaymentRequestValidator): Deleted.
(WebCore::PaymentRequestValidator::~PaymentRequestValidator): Deleted.
(WebCore::PaymentRequestValidator::validateCountryCode): Deleted.
(WebCore::PaymentRequestValidator::validateCurrencyCode): Deleted.
(WebCore::PaymentRequestValidator::validateMerchantCapabilities): Deleted.
(WebCore::PaymentRequestValidator::validateShippingMethod): Deleted.
(WebCore::PaymentRequestValidator::validateSupportedNetworks): Deleted.
(WebCore::PaymentRequestValidator::validateShippingMethods): Deleted.
* Modules/applepay/PaymentRequestValidator.h:
Instead of reporting the validation errors to the console, pass them in the exception.
This allows us to remove the window member, and move the helpers to the implementation
file as statics.

* WebCore.xcodeproj/project.pbxproj:
Remove file.

* bindings/generic/IDLTypes.h:
(WebCore::IDLObject::nullValue):
(WebCore::IDLObject::isNullValue):
(WebCore::IDLObject::extractValueFromNullable):
Add nullability traits for IDLObject.

* bindings/js/JSApplePaySessionCustom.cpp:
Removed.

Tools:

* WebKitTestRunner/cocoa/TestControllerCocoa.mm:
(WTR::initializeWebViewConfiguration):
(WTR::TestController::platformCreateWebView):
Enable ApplePay for the tests (where supported).

LayoutTests:

* http/tests/ssl/applepay: Added.
* http/tests/ssl/applepay/ApplePaySession-expected.txt: Added.
* http/tests/ssl/applepay/ApplePaySession.html: Added.
Add a compressive test of the handling of the payment request to the ApplePaySession
constructor. For further testing, we will need to mock out bits of the underlying
coordinator, but this allows for testing the aspect I am changing. The applepay
directory is in the ssl directory, as all use of ApplePay requires https.

* TestExpectations:
* platform/ios-simulator-wk2/TestExpectations:
* platform/mac-wk2/TestExpectations:
Only enable the tests on iOS 10.10 and greater, and macOS 10.12 and greater, WebKit2.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209760 268f45cc-cd09-0410-ab3c-d52691b4dbfc

CSP: Teach the preload scanner about the 'nonce' attribute
https://bugs.webkit.org/show_bug.cgi?id=161192
<rdar://problem/28010354>

Reviewed by Darin Adler.

Source/WebCore:

This patch was inspired by a similar Blink change:
<https://chromium.googlesource.com/chromium/src/+/dde5487f380cf774e4c0e96ba7f88ea68e723907>

Preload external scripts and stylesheets whose HTML script and link elements have a nonce
attribute that is listed in the Content Security Policy (CSP) of the page.

Currently the preload scanner ignores the nonce attribute on HTML script and link elements.
So, WebKit does not preload their associated subresources unless the value of the src
attribute or href attribute is whitelisted in the CSP of the page for script and link
elements, respectively. Instead the preload scanner should recognize the nonce attribute on
script and link elements and query the CSP of the page with it. If the nonce attribute is
whitelisted then the request should be preloaded.

Tests: http/tests/loading/do-not-preload-css-blocked-by-csp.html
       http/tests/loading/do-not-preload-script-src-blocked-by-csp.html
       http/tests/loading/preload-css-with-csp-nonce.html
       http/tests/loading/preload-script-src-with-csp-nonce.html

* html/parser/HTMLPreloadScanner.cpp:
(WebCore::TokenPreloadScanner::StartTagScanner::createPreloadRequest): Set the nonce on the
PreloadRequest to the nonce that we found during the scan.
(WebCore::TokenPreloadScanner::StartTagScanner::processAttribute): For script and link tag names,
save the value of the nonce attribute (if it has one).
* html/parser/HTMLResourcePreloader.cpp:
(WebCore::PreloadRequest::resourceRequest): Skip CSP policy check if the nonce is listed in
the CSP of the page.
* html/parser/HTMLResourcePreloader.h:
(WebCore::PreloadRequest::setNonce): Added.

LayoutTests:

Add tests to ensure that we preload <script>s and <link>s whose nonce is allowed by the
Content Security Policy of the page.

* http/tests/loading/do-not-preload-css-blocked-by-csp-expected.txt: Added.
* http/tests/loading/do-not-preload-css-blocked-by-csp.html: Added.
* http/tests/loading/do-not-preload-script-src-blocked-by-csp-expected.txt: Added.
* http/tests/loading/do-not-preload-script-src-blocked-by-csp.html: Added.
* http/tests/loading/preload-css-with-csp-nonce-expected.txt: Added.
* http/tests/loading/preload-css-with-csp-nonce.html: Added.
* http/tests/loading/preload-script-src-with-csp-nonce-expected.txt: Added.
* http/tests/loading/preload-script-src-with-csp-nonce.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209759 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[CSS Parser] Rename CSSPrimitiveValue::UnitTypes to CSSPrimitiveValue::UnitType
https://bugs.webkit.org/show_bug.cgi?id=165776

Reviewed by Zalan Bujtas.

* css/CSSCalculationValue.cpp:
(WebCore::unitCategory):
(WebCore::hasDoubleValue):
(WebCore::determineCategory):
(WebCore::CSSCalcExpressionNodeParser::parseValue):
* css/CSSCalculationValue.h:
* css/CSSPrimitiveValue.cpp:
(WebCore::isValidCSSUnitTypeForDoubleConversion):
(WebCore::isStringType):
(WebCore::CSSPrimitiveValue::unitCategory):
(WebCore::CSSPrimitiveValue::CSSPrimitiveValue):
(WebCore::CSSPrimitiveValue::cleanup):
(WebCore::CSSPrimitiveValue::computeLengthDouble):
(WebCore::CSSPrimitiveValue::computeNonCalcLengthDouble):
(WebCore::CSSPrimitiveValue::conversionToCanonicalUnitsScaleFactor):
(WebCore::CSSPrimitiveValue::getFloatValue):
(WebCore::CSSPrimitiveValue::doubleValue):
(WebCore::CSSPrimitiveValue::canonicalUnitTypeForCategory):
(WebCore::CSSPrimitiveValue::doubleValueInternal):
(WebCore::CSSPrimitiveValue::cloneForCSSOM):
* css/CSSPrimitiveValue.h:
(WebCore::CSSPrimitiveValue::isFontRelativeLength):
(WebCore::CSSPrimitiveValue::isLength):
(WebCore::CSSPrimitiveValue::isResolution):
(WebCore::CSSPrimitiveValue::createAllowingMarginQuirk):
* css/CSSValue.h:
* css/CSSValuePool.cpp:
(WebCore::CSSValuePool::createValue):
* css/CSSValuePool.h:
(WebCore::CSSValuePool::createValue):
* css/MediaQueryExp.cpp:
(WebCore::featureWithValidDensity):
(WebCore::MediaQueryExpression::MediaQueryExpression):
* css/parser/CSSParserFastPaths.cpp:
(WebCore::parseSimpleLength):
(WebCore::parseSimpleLengthValue):
(WebCore::parseColorIntOrPercentage):
(WebCore::fastParseColorInternal):
(WebCore::parseTransformTranslateArguments):
(WebCore::parseTransformNumberArguments):
* css/parser/CSSParserToken.cpp:
(WebCore::cssPrimitiveValueUnitFromTrie):
(WebCore::stringToUnitType):
(WebCore::CSSParserToken::CSSParserToken):
(WebCore::CSSParserToken::convertToPercentage):
* css/parser/CSSParserToken.h:
(WebCore::CSSParserToken::unitType):
* css/parser/CSSPropertyParser.cpp:
(WebCore::CSSPropertyParser::consumeTransformOrigin):
(WebCore::consumeCounter):
(WebCore::consumeAnimationName):
(WebCore::consumePerspective):
(WebCore::consumePositionLonghand):
(WebCore::consumeCounterContent):
(WebCore::consumeReflect):
(WebCore::consumeGridBreadth):
(WebCore::CSSPropertyParser::consumeFlex):
* css/parser/CSSPropertyParserHelpers.cpp:
(WebCore::CSSPropertyParserHelpers::CalcParser::consumeNumber):
(WebCore::CSSPropertyParserHelpers::consumeInteger):
(WebCore::CSSPropertyParserHelpers::consumeLength):
(WebCore::CSSPropertyParserHelpers::consumePercent):
(WebCore::CSSPropertyParserHelpers::consumeAngle):
(WebCore::CSSPropertyParserHelpers::consumeTime):
(WebCore::CSSPropertyParserHelpers::consumeCustomIdent):
(WebCore::CSSPropertyParserHelpers::consumeString):
(WebCore::CSSPropertyParserHelpers::consumeUrl):
(WebCore::CSSPropertyParserHelpers::consumeDeprecatedGradientPoint):
(WebCore::CSSPropertyParserHelpers::consumeDeprecatedGradientColorStop):
(WebCore::CSSPropertyParserHelpers::consumeCrossFade):
(WebCore::CSSPropertyParserHelpers::consumeImageSet):
(WebCore::CSSPropertyParserHelpers::consumeFilterFunction):
* css/parser/MediaQueryParser.cpp:
(WebCore::MediaQueryParser::readFeatureValue):
* css/parser/SizesAttributeParser.cpp:
(WebCore::SizesAttributeParser::computeLength):
* css/parser/SizesAttributeParser.h:
* dom/StyledElement.cpp:
(WebCore::StyledElement::setInlineStyleProperty):
(WebCore::StyledElement::addPropertyToPresentationAttributeStyle):
* dom/StyledElement.h:
* svg/SVGLengthValue.cpp:
(WebCore::SVGLengthValue::toCSSPrimitiveValue):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209758 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Annotate MediaStream and WebRTC idl with EnabledAtRuntime flag
https://bugs.webkit.org/show_bug.cgi?id=165251

Reviewed by Youenn Fablet.

Source/WebCore:

No new tests, fixes an API test.

* page/Settings.in: Delete mediaStreamEnabled and peerConnectionEnabled, we don't need a
setting and a runtime feature.

Source/WebKit2:

* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updatePreferences): Update the media stream and peer connection runtime
feature instead of the setting.

Tools:

* TestWebKitAPI/Tests/WebKit2/EnumerateMediaDevices.cpp:
(TestWebKitAPI::TEST): Enable MediaStream, create web view from the page group upon which
the preferences have been set.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209757 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Source/WebCore:
REGRESSION (r198990): Cannot edit content inside <details> in wysiwyg editor
https://bugs.webkit.org/show_bug.cgi?id=165757

Reviewed by Andreas Kling.

Test: fast/html/details-edit.html

-webkit-user-modify is reset on shadow boundary so it doesn't go through <details> shadow tree.

* css/StyleResolver.cpp:
(WebCore::StyleResolver::styleForElement):
(WebCore::StyleResolver::pseudoStyleForElement):
(WebCore::StyleResolver::styleForPage):
(WebCore::StyleResolver::applyMatchedProperties):
* dom/Node.cpp:
(WebCore::computeEditabilityFromComputedStyle):
(WebCore::Node::computeEditability):

    Make -webkit-user-modify (which we would want to get rid of completely eventually) have no effect in shadow trees.
    Check for contenteditable directly instead.

* html/HTMLInputElement.cpp:
(WebCore::HTMLInputElement::createInnerTextStyle):
* html/HTMLTextAreaElement.cpp:
(WebCore::HTMLTextAreaElement::createInnerTextStyle):
* html/shadow/TextControlInnerElements.cpp:
* rendering/RenderFlowThread.cpp:
(WebCore::RenderFlowThread::createFlowThreadStyle):
* rendering/RenderLayer.cpp:
(WebCore::RenderLayer::calculateClipRects):
* rendering/RenderListItem.cpp:
(WebCore::RenderListItem::styleDidChange):
* rendering/style/RenderStyle.cpp:
(WebCore::RenderStyle::createAnonymousStyleWithDisplay):
(WebCore::RenderStyle::createStyleInheritingFromPseudoStyle):
(WebCore::RenderStyle::inheritFrom):

    Let -webkit-user-modify inherit through shadow boundary as normal.

* rendering/style/RenderStyle.h:

LayoutTests:
REGRESSION (r198990): Safari - Cannot edit content inside <details> in wysiwyg editor
https://bugs.webkit.org/show_bug.cgi?id=165757

Reviewed by Andreas Kling.

* editing/execCommand/justify-right-then-indent-with-problematic-body-expected.txt:
* fast/html/details-edit-expected.txt: Added.
* fast/html/details-edit.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209756 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, silence -Wsuggest-attribute for GCC with pragmas

* wtf/PrintStream.cpp:
(WTF::PrintStream::printfVariableFormat):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209755 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed test gardening.

Update test expectations on Windows for failing tests. These failures are tracked in
https://bugs.webkit.org/show_bug.cgi?id=160447.

* platform/win/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209754 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fix GTK build after r209733
https://bugs.webkit.org/show_bug.cgi?id=165794

Document::visibilityState() now returns an IDL string enumeration.

Reviewed by Žan Doberšek.

* WebProcess/InjectedBundle/API/gtk/DOM/WebKitDOMDocument.cpp:
(webkit_dom_document_get_visibility_state):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209753 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remove bindings generation support for legacy WebCore::Dictionary
https://bugs.webkit.org/show_bug.cgi?id=165762

Reviewed by Sam Weinig.

LayoutTests/imported/w3c:

* web-platform-tests/webrtc/datachannel-emptystring-expected.txt:
* web-platform-tests/webrtc/no-media-call-expected.txt:
* web-platform-tests/webrtc/promises-call-expected.txt:
* web-platform-tests/webrtc/rtcpeerconnection/rtcpeerconnection-constructor-expected.txt:
* web-platform-tests/webrtc/rtcpeerconnection/rtcpeerconnection-idl-expected.txt:
Updated to expect error messages with "dictionary" with a lowercase D.

Source/WebCore:

After this patch, all use of legacy WebCore::Dictionary is within custom bindings or
inside the DOM code itself. Bindings generation machinery no longer has a special
type named "Dictionary" but has increasingly good support for IDL-defined dictionaries.

* Modules/applepay/ApplePaySession.cpp:
(WebCore::ApplePaySession::create): Take ExecState and a JSValue rather than a Dictionary.
(WebCore::ApplePaySession::completeMerchantValidation): Ditto.
* Modules/applepay/ApplePaySession.h: Updated for the above changes.
* Modules/applepay/ApplePaySession.idl: For this last file using Dictionary in IDL, move
to "[CallWith=ScriptState]" and "any" as a stopgap. Later we can return and use IDL
dictionaries or other more modern solutions.

* Modules/mediastream/MediaDevices.idl: For the custom binding written in JavaScript,
use the type mentioned in the specification rather than Dictionary. The bindings generator
ignores the type, so this is more like changing a comment rather than changing source code.

* Modules/mediastream/RTCPeerConnection.js:
(initializeRTCPeerConnection): Refer to a non-object as not a valid "dictionary", lowercase
rather than "Dictionary".
* Modules/mediastream/RTCPeerConnectionInternals.js:
(callbacksAndDictionaryOverload): Ditto.

* bindings/generic/IDLTypes.h: Removed IDLLegacyDictionary.

* bindings/js/JSCryptoAlgorithmDictionary.cpp:
(WebCore::getHashAlgorithm): Fixed a comment.

* bindings/js/JSDOMConvert.h:
(WebCore::Converter<IDLLegacyDictionary<T>>::convert): Deleted.

* bindings/scripts/CodeGenerator.pm:
(IsBuiltinType): Removed special case for "Dictionary".
* bindings/scripts/CodeGeneratorJS.pm:
(AddToIncludesForIDLType): Ditto.
(AreTypesDistinguishableForOverloadResolution): Ditto.
(GenerateOverloadedFunctionOrConstructor): Ditto.
(GenerateParametersCheck): Ditto.
(GetBaseIDLType): Ditto.

* bindings/scripts/test/JS/JSTestObj.cpp: Regenerated results.
* bindings/scripts/test/TestObj.idl: Removed test case for Dictionary.

LayoutTests:

* fast/mediastream/RTCPeerConnection-expected.txt:
* fast/mediastream/RTCPeerConnection-overloaded-operations-params-expected.txt:
* fast/mediastream/RTCPeerConnection-overloaded-operations-params.html:
Updated to expect error messages with "dictionary" with a lowercase D.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209752 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Drop unnecessary null check in removeDetachedChildrenInContainer()
https://bugs.webkit.org/show_bug.cgi?id=165778

Reviewed by Alex Christensen.

Drop unnecessary null check in removeDetachedChildrenInContainer().
Node cannot be null but we do an implicit null check.

No new tests, no Web-exposed behavior change.

* dom/ContainerNodeAlgorithms.cpp:
(WebCore::removeDetachedChildrenInContainer):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209751 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Add Document.onvisibilitychange event handler attribute
https://bugs.webkit.org/show_bug.cgi?id=165784

Reviewed by Daniel Bates.

Source/WebCore:

Add Document.onvisibilitychange event handler attribute as per Page
Visibility Level 2 API:
- http://w3c.github.io/page-visibility/#onvisiblitychange-event-handler

Test: fast/events/page-visibility-onvisibilitychange.html

* dom/Document.idl:

LayoutTests:

Add layout test coverage.

* fast/events/page-visibility-onvisibilitychange-expected.txt: Added.
* fast/events/page-visibility-onvisibilitychange.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209750 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[WebCrypto] Migrate some tests from webkitSubtle to subtle
https://bugs.webkit.org/show_bug.cgi?id=165554

Reviewed by Brent Fulgham.

To improve test coverage for subtle.

* crypto/crypto-key-algorithm-gc-expected.txt: Renamed from LayoutTests/crypto/webkitSubtle/crypto-key-algorithm-gc-expected.txt.
* crypto/crypto-key-algorithm-gc.html: Renamed from LayoutTests/crypto/webkitSubtle/crypto-key-algorithm-gc.html.
* crypto/crypto-key-usages-gc-expected.txt: Renamed from LayoutTests/crypto/webkitSubtle/crypto-key-usages-gc-expected.txt.
* crypto/crypto-key-usages-gc.html: Renamed from LayoutTests/crypto/webkitSubtle/crypto-key-usages-gc.html.
* crypto/subtle/resources/rsa-indexeddb-non-exportable.js: Added.
* crypto/subtle/resources/rsa-indexeddb.js: Added.
* crypto/subtle/rsa-indexeddb-expected.txt: Added.
* crypto/subtle/rsa-indexeddb-non-exportable-expected.txt: Added.
* crypto/subtle/rsa-indexeddb-non-exportable-private-expected.txt: Added.
* crypto/subtle/rsa-indexeddb-non-exportable-private.html: Added.
* crypto/subtle/rsa-indexeddb-non-exportable.html: Added.
* crypto/subtle/rsa-indexeddb-private-expected.txt: Added.
* crypto/subtle/rsa-indexeddb-private.html: Added.
* crypto/subtle/rsa-indexeddb.html: Added.
* crypto/subtle/rsa-oaep-plaintext-length-expected.txt: Added.
* crypto/subtle/rsa-oaep-plaintext-length.html: Added.
* crypto/subtle/rsassa-pkcs1-v1_5-generate-key-with-leading-zeroes-in-exponent-expected.txt: Added.
* crypto/subtle/rsassa-pkcs1-v1_5-generate-key-with-leading-zeroes-in-exponent.html: Added.
* platform/ios-simulator-wk1/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209749 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[GTK] Web Inspector: Add NavigationItemCodeCoverage.svg icon
https://bugs.webkit.org/show_bug.cgi?id=165698

Reviewed by Joseph Pecoraro.

* UserInterface/Images/gtk/NavigationItemCodeCoverage.svg: Added.
Match in style the existing NavigationItemTypes.svg icon.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209748 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[iOS] MediaDocument "Done" button should navigate the page back
https://bugs.webkit.org/show_bug.cgi?id=165779

Reviewed by Sam Weinig.

Source/WebCore:

Detect if the exit from fullscreen was caused by the Done button,
and if so, tell the page to navigate back.

Unfortunately this is not yet testable. It's waiting on the
UI-based device testing in development by the media team.

* platform/cocoa/WebVideoFullscreenModel.h: Add a finishedWithMedia parameter to
requestFullscreenMode, to indicate if the change in mode is associated with
the closing of the media document.
* platform/cocoa/WebVideoFullscreenModelVideoElement.h:
* platform/cocoa/WebVideoFullscreenModelVideoElement.mm:
(WebVideoFullscreenModelVideoElement::requestFullscreenMode): If we're a media
document, and we say we're finished with the media, tell the document
to navigate back a page.
* platform/ios/WebVideoFullscreenControllerAVKit.mm:
(WebVideoFullscreenControllerContext::requestFullscreenMode):
* platform/ios/WebVideoFullscreenInterfaceAVKit.mm:
(WebVideoFullscreenInterfaceAVKit::shouldExitFullscreenWithReason):

Source/WebKit2:

* UIProcess/Cocoa/WebVideoFullscreenManagerProxy.h:
* UIProcess/Cocoa/WebVideoFullscreenManagerProxy.mm:
(WebKit::WebVideoFullscreenModelContext::requestFullscreenMode):
(WebKit::WebVideoFullscreenManagerProxy::requestFullscreenMode):
* WebProcess/cocoa/WebVideoFullscreenManager.h:
* WebProcess/cocoa/WebVideoFullscreenManager.messages.in:
* WebProcess/cocoa/WebVideoFullscreenManager.mm:
(WebKit::WebVideoFullscreenManager::requestFullscreenMode):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209747 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remove unused workaround for Silverlight
https://bugs.webkit.org/show_bug.cgi?id=165773

Patch by Alex Christensen <achristensen@webkit.org> on 2016-12-12
Reviewed by Dean Jackson.

* Plugins/WebNetscapePluginView.mm:
(-[WebNetscapePluginView _workaroundSilverlightFullscreenBug:]): Deleted.
This code was introduced in January 2010.  Microsoft fixed their bug shortly thereafter.
Allowing 6 years for users to update Silverlight is more than enough.
Safari doesn't even use WebKit1 for this any more.
I need this to be removed because this is the only non-WebGL code that needs to link against
the OpenGL framework, and I'm changing how that is done to link only through ANGLE instead.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209746 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Infinite recursion when viewport is set to the size of the content but the content overflows the viewport.
https://bugs.webkit.org/show_bug.cgi?id=165775
rdar://problem/29366628

Reviewed by Simon Fraser.

In certain cases when the viewport is sized to accomodate the content and
the content always overflows the viewport, we might end up in recursive FrameView::layout calls.
This is specific to content with viewport units, because we always invalidate elements with vw/vh units on
viewport size change. However if this viewport size change is in response to content size change (layout),
invalidating elements could trigger synchronous layout, while we are still inside this current layout.
This is very similar to the m_setNeedsLayoutWasDeferred case and they should eventually be merged.
It also means that we might be behind by one layout on elements with vw/vh units (fixed layout only though).

Currently not testable.

* page/FrameView.cpp:
(WebCore::FrameView::availableContentSizeChanged):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209745 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Memory warning logging appears to capture resident footprint, missing compress/swap.
https://bugs.webkit.org/show_bug.cgi?id=165533
<rdar://problem/29318410>

Reviewed by Andreas Kling.

Have platformMemoryUsage return both resident and resident + swapped
values. This is now returned as a struct in a std::optional to better
support cases where values cannot or are not returned. Report these
values in logMemoryUsageChange.

Remove most of the ReliefLogger instances and their messages. We no
longer free up fastMalloc memory after each memory-release operation,
so there are no memory regions returned to the OS. Instead, we now
free up fastMalloc memory regions at the end of all memory-releasing
operations in one fell swoop and report a grand total of memory
returned.

No new tests -- no changes to user accessible functionality.

* page/MemoryRelease.cpp:
(WebCore::releaseNoncriticalMemory):
(WebCore::releaseCriticalMemory):
(WebCore::releaseMemory):
* page/cocoa/MemoryReleaseCocoa.mm:
(WebCore::platformReleaseMemory):
* platform/MemoryPressureHandler.cpp:
(WebCore::MemoryPressureHandler::releaseMemory):
(WebCore::MemoryPressureHandler::ReliefLogger::logMemoryUsageChange):
(WebCore::MemoryPressureHandler::ReliefLogger::platformMemoryUsage):
* platform/MemoryPressureHandler.h:
(WebCore::MemoryPressureHandler::ReliefLogger::ReliefLogger):
(WebCore::MemoryPressureHandler::ReliefLogger::~ReliefLogger):
(WebCore::MemoryPressureHandler::ReliefLogger::loggingEnabled):
* platform/cocoa/MemoryPressureHandlerCocoa.mm:
(WebCore::MemoryPressureHandler::platformReleaseMemory):
(WebCore::MemoryPressureHandler::ReliefLogger::platformMemoryUsage):
* platform/linux/MemoryPressureHandlerLinux.cpp:
* platform/win/MemoryPressureHandlerWin.cpp:
(WebCore::MemoryPressureHandler::ReliefLogger::platformMemoryUsage):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209744 268f45cc-cd09-0410-ab3c-d52691b4dbfc

More IndexedDB perf tests.
https://bugs.webkit.org/show_bug.cgi?id=165634

Reviewed by Sam Weinig.

* IndexedDB/index-multientry.html:
* IndexedDB/large-number-of-inserts-responsiveness.html: Added.
* IndexedDB/large-number-of-inserts.html: Added.
* IndexedDB/objectstore-cursor.html: Added. Runtime cut in 1/4th compared to r209672.

* resources/runner.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209738 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Provide SPI to avoid blocking on painting when coming into view
https://bugs.webkit.org/show_bug.cgi?id=165780
<rdar://problem/29009559>

Reviewed by Simon Fraser.

By default, when parenting a WKWebView which was previously in-window,
we block on painting to ensure that we don't flash stale content.
We added SPI to disable this behavior to WKWebViewConfiguration, but
some clients might want the behavior most of the time, but sometimes
instead want to not block the UI process main thread, but instead do
their own work (e.g. removing a snapshot) when painting eventually occurs.

* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _doAfterNextPresentationUpdateWithoutWaitingForPainting:]):
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
Tell WebPageProxy not to block on painting next time we are parented,
and use doAfterNextPresentationUpdate to call the block when the painting
does eventually happen. This SPI must be called before - but in the same
runloop cycle - the WKWebView is brought back in-window to work correctly.

* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::dispatchActivityStateChange):
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::skipWaitingForPaintAfterNextViewDidMoveToWindow):
If we would block because we're coming in-window, but the bit to skip
blocking the next time we get brought in-window is set, reset the bit,
and avoid blocking.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209736 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, follow-up after r209706.

* inspector/debugger/breakpoint-action-log.html:
Add missing test case description.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209735 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: Add tests for Fetch API Network Data (CORs, Opaque Responses, Filtered Headers)
https://bugs.webkit.org/show_bug.cgi?id=165683

Reviewed by Brian Burg.

* http/tests/inspector/network/fetch-network-data-expected.txt: Added.
* http/tests/inspector/network/fetch-network-data.html: Added.
* http/tests/inspector/network/resources/cors-data.pl: Added.
* http/tests/inspector/resources/inspector-test.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209734 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Document.visibilityState should use an IDL string enumeration
https://bugs.webkit.org/show_bug.cgi?id=165774

Reviewed by Daniel Bates.

Source/WebCore:

Document.visibilityState should use an IDL string enumeration:
- http://w3c.github.io/page-visibility/#extensions-to-the-document-interface

No new tests, there should be no Web-exposed behavior change.

* CMakeLists.txt:
* Modules/vibration/NavigatorVibration.cpp:
(WebCore::NavigatorVibration::vibrate):
* WebCore.xcodeproj/project.pbxproj:
* dom/Document.cpp:
(WebCore::Document::hidden):
(WebCore::Document::pageVisibilityState): Deleted.
(WebCore::Document::visibilityState): Deleted.
* dom/Document.h:
* dom/Document.idl:
* page/Page.cpp:
(WebCore::Page::visibilityState):
* page/PageVisibilityState.cpp: Removed.
* page/PageVisibilityState.h:
(): Deleted.

Source/WebKit/mac:

* DOM/DOMDocument.mm:
(-[DOMDocument visibilityState]):
* WebView/WebView.mm:
(kit):

Source/WebKit2:

* Shared/API/c/WKSharedAPICast.h:
(WebKit::toPageVisibilityState):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209733 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remove implementation of legacy Mozilla-based Fullscreen API.
https://bugs.webkit.org/show_bug.cgi?id=165689
Source/WebCore:

Reviewed by Eric Carlson.

Alias the legacy webkitRequestFull_S_creen() function on Element to webkitRequestFull_s_creen().
Remove the custom behavior in Document::requestFullScreenForElement() to handle the legacy path.

* dom/Document.cpp:
(WebCore::Document::requestFullScreenForElement):
* dom/Document.h:
* dom/Element.cpp:
(WebCore::Element::webkitRequestFullscreen):
(WebCore::Element::webkitRequestFullScreen): Deleted.
* dom/Element.h:
* dom/Element.idl:
* html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::enterFullscreen):
* html/shadow/MediaControlElements.cpp:
(WebCore::MediaControlFullscreenButtonElement::defaultEventHandler):

Source/WebKit/mac:

<rdar://problem/29603053>

Reviewed by Eric Carlson.

* DOM/DOMElement.mm:
(-[DOMElement webkitRequestFullScreen:]):

LayoutTests:

Reviewed by Eric Carlson.

* fullscreen/full-screen-css-expected.txt:
* fullscreen/full-screen-css.html:
* fullscreen/full-screen-keyboard-disabled-expected.txt: Removed.
* fullscreen/full-screen-keyboard-disabled.html: Removed.
* fullscreen/full-screen-twice-expected.txt:
* fullscreen/full-screen-twice.html:
* js/dom/dom-static-property-for-in-iteration-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209730 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Rename BytecodeGenerator's ControlFlowContext to ControlFlowScope.
https://bugs.webkit.org/show_bug.cgi?id=165777

Reviewed by Keith Miller.

The existing code sometimes refer to ControlFlowContext (and associated references)
as context, and sometimes as scope.  Let's be consistent and always call it a scope.

Also renamed push/popScopedControlFlowContext() to push/popLocalControlFlowScope()
because these are only used when we inc/dec the m_localScopeDepth.

* bytecompiler/BytecodeGenerator.cpp:
(JSC::BytecodeGenerator::initializeVarLexicalEnvironment):
(JSC::BytecodeGenerator::pushLexicalScopeInternal):
(JSC::BytecodeGenerator::popLexicalScopeInternal):
(JSC::BytecodeGenerator::emitPushWithScope):
(JSC::BytecodeGenerator::emitPopWithScope):
(JSC::BytecodeGenerator::pushFinallyControlFlowScope):
(JSC::BytecodeGenerator::pushIteratorCloseControlFlowScope):
(JSC::BytecodeGenerator::popFinallyControlFlowScope):
(JSC::BytecodeGenerator::popIteratorCloseControlFlowScope):
(JSC::BytecodeGenerator::emitComplexPopScopes):
(JSC::BytecodeGenerator::emitPopScopes):
(JSC::BytecodeGenerator::pushLocalControlFlowScope):
(JSC::BytecodeGenerator::popLocalControlFlowScope):
(JSC::BytecodeGenerator::emitEnumeration):
(JSC::BytecodeGenerator::pushFinallyContext): Deleted.
(JSC::BytecodeGenerator::pushIteratorCloseContext): Deleted.
(JSC::BytecodeGenerator::popFinallyContext): Deleted.
(JSC::BytecodeGenerator::popIteratorCloseContext): Deleted.
(JSC::BytecodeGenerator::pushScopedControlFlowContext): Deleted.
(JSC::BytecodeGenerator::popScopedControlFlowContext): Deleted.
* bytecompiler/BytecodeGenerator.h:
* bytecompiler/NodesCodegen.cpp:
(JSC::TryNode::emitBytecode):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209728 268f45cc-cd09-0410-ab3c-d52691b4dbfc

GC scheduler should avoid consecutive pauses
https://bugs.webkit.org/show_bug.cgi?id=165758

Reviewed by Michael Saboff.

This factors out the scheduler from lambdas in Heap::markToFixpoint to an actual class.
It's called the SpaceTimeScheduler because it is a linear controller that ties the
amount of time you spend on things to the amount of space you are using.

This patch uses this refactoring to fix a bug where the GC would pause even though we
still had time during a mutator timeslice. This is a 15% improvement on
JetStream/splay-latency. Seems neutral on everything else. However, it's not at all
clear if this is the right policy or not since retreating wavefront can sometimes be so
sensitive to scheduling decisions. For this reason, there is a tunable option that lets
you decide how long the GC will sit idle before the start of its timeslice.

So, we can revert this policy change in this patch without reverting the patch.

* CMakeLists.txt:
* JavaScriptCore.xcodeproj/project.pbxproj:
* heap/Heap.cpp:
(JSC::Heap::markToFixpoint):
* heap/Heap.h:
* heap/SpaceTimeScheduler.cpp: Added.
(JSC::SpaceTimeScheduler::Decision::targetMutatorUtilization):
(JSC::SpaceTimeScheduler::Decision::targetCollectorUtilization):
(JSC::SpaceTimeScheduler::Decision::elapsedInPeriod):
(JSC::SpaceTimeScheduler::Decision::phase):
(JSC::SpaceTimeScheduler::Decision::shouldBeResumed):
(JSC::SpaceTimeScheduler::Decision::timeToResume):
(JSC::SpaceTimeScheduler::Decision::timeToStop):
(JSC::SpaceTimeScheduler::SpaceTimeScheduler):
(JSC::SpaceTimeScheduler::snapPhase):
(JSC::SpaceTimeScheduler::currentDecision):
* heap/SpaceTimeScheduler.h: Added.
(JSC::SpaceTimeScheduler::Decision::Decision):
(JSC::SpaceTimeScheduler::Decision::operator bool):
* runtime/Options.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209727 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[CSS Parser] MediaQueryParser should use Ref and not RefPtr
https://bugs.webkit.org/show_bug.cgi?id=165771

Reviewed by Simon Fraser.

* css/MediaList.cpp:
(WebCore::MediaQuerySet::create):
* css/parser/CSSParserImpl.cpp:
(WebCore::CSSParserImpl::consumeImportRule):
(WebCore::CSSParserImpl::consumeMediaRule):
* css/parser/MediaQueryParser.cpp:
(WebCore::MediaQueryParser::parseMediaQuerySet):
(WebCore::MediaQueryParser::parseMediaCondition):
(WebCore::MediaQueryParser::parseInternal):
* css/parser/MediaQueryParser.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209726 268f45cc-cd09-0410-ab3c-d52691b4dbfc

REGRESSION(r209653): speedometer crashes making virtual slow path tailcalls
https://bugs.webkit.org/show_bug.cgi?id=165748

Reviewed by Filip Pizlo.

JSTests:

New regression test.

* stress/regress-165748.js: Added.
(sum1):
(sum2):
(sum3):
(sum4):
(sum5):
(sum6):
(tailCaller):
(test):

Source/JavaScriptCore:

The virtual slow path for tailcalls always passes arguments on the stack.
The fix here is to link to the stack argument entrypoint instead of a register
argument entrypoint.

While fixing this bug, I found that we weren't clearing the code origin when
shuffling the call frame for a register argument tailcall.

Also rolling back in r209653, r209654, r209663, and r209673.

* jit/CallFrameShuffler.cpp:
(JSC::CallFrameShuffler::prepareAny):
* jit/ThunkGenerators.cpp:
(JSC::virtualThunkFor):

Source/WTF:

Rolling back in r209653, r209654, r209663, and r209673.

* wtf/Platform.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209725 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Debug JSC test timeout: stress/string-prototype-replace-should-throw-out-of-memory-error-when-using-too-much-memory.js.ftl-no-cjit.
https://bugs.webkit.org/show_bug.cgi?id=164630

Reviewed by Keith Miller.

Speculative fix for timeout issue.  It looks like this test just runs a bit
slower on debug builds.  So, we'll give it a little more time to run.

* stress/string-prototype-replace-should-throw-out-of-memory-error-when-using-too-much-memory.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209724 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Rename BytecodeGenerator's m_symbolTableStack to m_lexicalScopeStack.
https://bugs.webkit.org/show_bug.cgi?id=165768

Reviewed by Saam Barati.

The lexical scope in "m_lexicalScopeStack" here refers to a pair of { } in the
source code that bounds the scope of variables.

There are 4 places in the code where we call m_symbolTableStack.append() to
append a new stack entry.  In only 3 of the 4 cases, a symbol table is provided
in the new stack entry.  In all 4 cases, a scope register is provided in the new
stack entry.

Also, 3 of the 4 functions that appends an entry to this stack are named:
1. initializeVarLexicalEnvironment()
2. pushLexicalScopeInternal()
3. emitPushWithScope()

The 4th function is the BytecodeGenerator constructor where it pushes the scope
for a module environment.

Based on these details, m_lexicalScopeStack is a better name for this stack than
m_symbolTableStack.

* bytecompiler/BytecodeGenerator.cpp:
(JSC::BytecodeGenerator::BytecodeGenerator):
(JSC::BytecodeGenerator::initializeArrowFunctionContextScopeIfNeeded):
(JSC::BytecodeGenerator::initializeVarLexicalEnvironment):
(JSC::BytecodeGenerator::pushLexicalScopeInternal):
(JSC::BytecodeGenerator::initializeBlockScopedFunctions):
(JSC::BytecodeGenerator::hoistSloppyModeFunctionIfNecessary):
(JSC::BytecodeGenerator::popLexicalScopeInternal):
(JSC::BytecodeGenerator::prepareLexicalScopeForNextForLoopIteration):
(JSC::BytecodeGenerator::variable):
(JSC::BytecodeGenerator::resolveType):
(JSC::BytecodeGenerator::emitResolveScope):
(JSC::BytecodeGenerator::emitPushWithScope):
(JSC::BytecodeGenerator::emitPopWithScope):
(JSC::BytecodeGenerator::pushFinallyContext):
(JSC::BytecodeGenerator::pushIteratorCloseContext):
(JSC::BytecodeGenerator::emitComplexPopScopes):
(JSC::BytecodeGenerator::popTryAndEmitCatch):
(JSC::BytecodeGenerator::emitPushFunctionNameScope):
* bytecompiler/BytecodeGenerator.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209723 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed. Try to fix the cloop build.

* interpreter/StackVisitor.cpp:
(JSC::StackVisitor::Frame::calleeSaveRegisters):
* interpreter/StackVisitor.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209722 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[css-grid] Add some test cases that were removed in r209562
https://bugs.webkit.org/show_bug.cgi?id=165750

Reviewed by Sergio Villar Senin.

Some valid test cases were removed in r209562, this patch adds them back.

* fast/css-grid-layout/grid-template-shorthand-get-set.html: The specified style is now preserved,
so we've to use a different value for it (as it doesn't merge the line names).
* fast/css-grid-layout/resources/grid-columns-rows-get-set.js: Flexible sizes must be non-negative,
but 0fr should be accepted (this is new as it was not working like that with the old parser).
A space between the dimension and the unit is wrong for flexible sizes.
* fast/css-grid-layout/resources/non-grid-columns-rows-get-set.js: Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209721 268f45cc-cd09-0410-ab3c-d52691b4dbfc

The CSS 'columns' property when set on the <body> element makes short columns
https://bugs.webkit.org/show_bug.cgi?id=164891

Reviewed by Darin Adler and David Hyatt.

Source/WebCore:

Currently when the body has auto height, we use the height of the viewport as
the available height for the columns defined on the body element.
This is the desired behaviour for paginated content.
However in case of a simple, non-paginated multicolumn context, we should let the columns grow beyond
the viewport vertically. This is also what other browsers do.
This patch sets the available height on columns (in non-paginated context) only if the logical height
is defined (we are right before layout, so applying non-defined values does not make much sense anyway).
The reason why it is only broken on the <body> is because it stretches to the viewport by
default (see stretchesToViewport()) while other block renderers do not.

Test: fast/multicol/columns-on-body.html

* rendering/RenderBlockFlow.cpp:
(WebCore::RenderBlockFlow::checkForPaginationLogicalHeightChange):

LayoutTests:

* fast/multicol/columns-on-body-expected.html: Added.
* fast/multicol/columns-on-body.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209719 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[CSS Parser] Implement deferred parsing of properties, @media, @supports and @keyframes
https://bugs.webkit.org/show_bug.cgi?id=165743

Reviewed by Darin Adler.

Source/WebCore:

Added new tests in fast/css/deferred-parsing.

This patch implements deferred parsing in CSS. By default, when parsing
author stylesheets, the parser goes into a deferred mode where it leaves
the declaration block of all style rules unparsed initially. Instead of
creating and setting an ImmutableStyleProperties object on the StyleRule,
the parser creates a DeferredStyleProperties object instead that is
capable of calling into CSSDeferredParser to parse the properties and values
only when the properties are finally asked for.

In addition, this patch defers the parsing of the interior of @media,
@supports, and @keyframes rules. @media blocks that are never matched will
never parse any of the rules inside. @supports conditions for unsupported
features will also not be parsed. For @keyframes, if the animation is never
referenced/used, then the @keyframes child rules will never be parsed.

* CMakeLists.txt:
* WebCore.xcodeproj/project.pbxproj:
Add CSSDeferredParser to the build.

* css/CSSFontFaceSet.cpp:
(WebCore::CSSFontFaceSet::matchingFaces):
Patched to drop the StyleSheetContents* argument from parseValue, since it
is no longer needed.

* css/CSSKeyframesRule.cpp:
(WebCore::StyleRuleKeyframes::StyleRuleKeyframes):
(WebCore::StyleRuleKeyframes::parseDeferredRulesIfNeeded):
(WebCore::StyleRuleKeyframes::keyframes):
(WebCore::StyleRuleKeyframes::wrapperAppendKeyframe):
(WebCore::StyleRuleKeyframes::wrapperRemoveKeyframe):
(WebCore::StyleRuleKeyframes::findKeyframeIndex):
* css/CSSKeyframesRule.h:
Keyframes now hold both a keyframes Vector and a DeferredStyleGroupRuleList
object. The keyframes vector won't be populated until it is needed, and the
population is done by DeferredStyleGroupRuleList, which holds both the original
CSSParserTokenRange of the keyframes rule interior, and a references to the
CSSDeferredParser that will do the parsing.

* css/DOMCSSNamespace.cpp:
(WebCore::DOMCSSNamespace::supports):
Patched to drop the StyleSheetContents* argument from parseValue, since it
is no longer needed.

* css/ElementRuleCollector.cpp:
(WebCore::ElementRuleCollector::collectMatchingRulesForList):
ElementRuleCollector tests to see if a rule is empty by counting properties(),
but we don't want to initiate deferred parsing just to check this. Instead we
invoke propertiesWithoutDeferredParsing, which will hand back a nullptr if
the properties haven't been parsed yet. In order to ensure this optimization
continues to work, any rules whose interiors consist of only whitespace do
not get deferred.

* css/FontFace.cpp:
(WebCore::FontFace::setVariant):
* css/PropertySetCSSStyleDeclaration.cpp:
(WebCore::PropertySetCSSStyleDeclaration::setProperty):
(WebCore::PropertySetCSSStyleDeclaration::setPropertyInternal):
Patched to drop the StyleSheetContents* argument from parseValue, since it
is no longer needed.

* css/StyleProperties.cpp:
(WebCore::MutableStyleProperties::MutableStyleProperties):
(WebCore::MutableStyleProperties::setProperty):
(WebCore::DeferredStyleProperties::create):
(WebCore::DeferredStyleProperties::DeferredStyleProperties):
(WebCore::DeferredStyleProperties::~DeferredStyleProperties):
(WebCore::DeferredStyleProperties::parseDeferredProperties):
* css/StyleProperties.h:
(WebCore::StylePropertiesBase::type):
(WebCore::StylePropertiesBase::cssParserMode):
(WebCore::StylePropertiesBase::StylePropertiesBase):
(WebCore::StyleProperties::isMutable):
(WebCore::StyleProperties::StyleProperties):
(WebCore::StylePropertiesBase::deref):
(isType):
(WebCore::StyleProperties::cssParserMode): Deleted.
(WebCore::StyleProperties::deref): Deleted.
Right now StyleProperties has two subclasses, ImmutableStyleProperties (which
is created by the parser when not deferring), and MutableStyleProperties (used
when making changes and doing DOM Manipulations).

This patch adds a new base class called StylePropertiesBase, and StyleProperties
now derives from it. DeferredStyleProperties derives from this base. The members
of StyleProperties have been moved up into StylePropertiesBase.

The m_isMutable bit, used to determine the type for downcasting has been enhanced
to m_type (2 bits), since there are now three possible subclasses.

* css/StyleRule.cpp:
(WebCore::StyleRule::StyleRule):
(WebCore::StyleRule::properties):
(WebCore::StyleRule::mutableProperties):
(WebCore::StyleRule::splitIntoMultipleRulesWithMaximumSelectorComponentCount):
Patched to ensure that when properties() are accessed, that if the current
stored StylePropertyBase in the rule is a DeferredStyleProperties, we then
parse and replace DeferredStyleProperties with ImmutableStyleProperties on the
rule. By using a common base class, we avoid increasing memory usage in
StyleRule.

(WebCore::DeferredStyleGroupRuleList::create):
(WebCore::DeferredStyleGroupRuleList::DeferredStyleGroupRuleList):
(WebCore::DeferredStyleGroupRuleList::parseDeferredRules):
(WebCore::DeferredStyleGroupRuleList::parseDeferredKeyframes):
DeferredStyleGroupList is used to handle deferred parsing of objects with
child rules. These include @media, @supports and @keyframes.

(WebCore::StyleRuleGroup::StyleRuleGroup):
(WebCore::StyleRuleGroup::childRules):
(WebCore::StyleRuleGroup::wrapperInsertRule):
(WebCore::StyleRuleGroup::wrapperRemoveRule):
(WebCore::StyleRuleGroup::parseDeferredRulesIfNeeded):
(WebCore::StyleRuleMedia::StyleRuleMedia):
(WebCore::StyleRuleSupports::StyleRuleSupports):
StyleRuleGroup now holds both a child rules Vector and a
DeferredStyleGroupRuleList. If DeferredStyleGroupRuleList is set, then
when the childRules are finally accessed, we will populate the childRules
Vector and then null out the DeferredStyleGroupRuleList.

* css/StyleRule.h:
(WebCore::StyleRule::propertiesWithoutDeferredParsing):
(WebCore::StyleRuleGroup::childRulesWithoutDeferredParsing):
(WebCore::StyleRuleGroup::childRules): Deleted.
Add new accesors for fetching properties and child rules without causing
deferred parsing to happen.

* css/StyleSheetContents.cpp:
(WebCore::StyleSheetContents::parseAuthorStyleSheet):
(WebCore::StyleSheetContents::parseString):
Patched to turn on deferred parsing for author sheets and for all
non-UA stylesheets parsed from a string (e.g., <style>).

(WebCore::traverseSubresourcesInRules):
The broken resource check was aggressively accessing properties(). Patch
it to call propertiesWithoutDeferredParsing() instead so that it doesn't
defeat the new optimization.

* css/WebKitCSSMatrix.cpp:
(WebCore::WebKitCSSMatrix::setMatrixValue):
Patched to drop the StyleSheetContents* argument from parseValue, since it
is no longer needed.

* css/parser/CSSDeferredParser.cpp: Added.
(WebCore::CSSDeferredParser::CSSDeferredParser):
(WebCore::CSSDeferredParser::parseDeclaration):
(WebCore::CSSDeferredParser::parseRuleList):
(WebCore::CSSDeferredParser::parseKeyframeList):
* css/parser/CSSDeferredParser.h: Added.
(WebCore::CSSDeferredParser::create):
(WebCore::CSSDeferredParser::tokenizer):
(WebCore::CSSDeferredParser::mode):
(WebCore::CSSDeferredParser::context):
(WebCore::CSSDeferredParser::styleSheet):
The CSSDeferredParser. It caches the original CSSTokenizer used to parse
the sheet, which in turn retains all the tokens and escaped strings. Because
the CSSDeferredParser is referenced by all the rules that have yet to parse,
it will stay alive until all rules in the stylesheet has been fully parsed.

The parser receives CSSParserTokenRanges from DeferredStyleProperties or
DeferredStyleGroupRuleList and feeds them to CSSParserImpl, which then
is able to do the parsing as it would have occurred originally had the
range not been deferred.

* css/parser/CSSParser.cpp:
(WebCore::CSSParser::parseSheet):
(WebCore::CSSParser::parseSupportsCondition):
(WebCore::CSSParser::parseSingleValue):
(WebCore::CSSParser::parseValue):
* css/parser/CSSParser.h:
* css/parser/CSSParserFastPaths.cpp:
(WebCore::CSSParserFastPaths::isValidKeywordPropertyAndValue):
(WebCore::parseKeywordValue):
(WebCore::CSSParserFastPaths::maybeParseValue):
* css/parser/CSSParserFastPaths.h:
Drop the StyleSheetContents* argument from the methods that no longer need
them.

* css/parser/CSSParserImpl.cpp:
(WebCore::CSSParserImpl::CSSParserImpl):
(WebCore::CSSParserImpl::tokenizer):
(WebCore::CSSParserImpl::parseValue):
(WebCore::CSSParserImpl::parseCustomPropertyValue):
(WebCore::CSSParserImpl::parseDeclarationList):
(WebCore::CSSParserImpl::parseInlineStyleDeclaration):
(WebCore::CSSParserImpl::parseRule):
(WebCore::CSSParserImpl::parseStyleSheet):
(WebCore::CSSParserImpl::parseDeclarationListForInspector):
(WebCore::CSSParserImpl::parseStyleSheetForInspector):
(WebCore::CSSParserImpl::consumeQualifiedRule):
(WebCore::CSSParserImpl::consumeMediaRule):
(WebCore::CSSParserImpl::consumeSupportsRule):
(WebCore::CSSParserImpl::consumeKeyframesRule):
(WebCore::CSSParserImpl::consumeStyleRule):
Drop the StyleSheetContents* argument from the methods that no longer need
them.

(WebCore::CSSParserImpl::createDeferredStyleProperties):
(WebCore::CSSParserImpl::parseDeferredDeclaration):
(WebCore::CSSParserImpl::parseDeferredRuleList):
(WebCore::CSSParserImpl::parseDeferredKeyframeList):
The methods that handle deferred parsing. They make sure to initialize
new CSSParserImpls with the original deferred parser, and then call existing
methods to do the parsing.

* css/parser/CSSParserImpl.h:
Patched to remove StyleSheetContents* arguments from some methods and
to add the new deferred parsing methods.

* css/parser/CSSParserTokenRange.cpp:
(WebCore::CSSParserTokenRange::consumeBlockCheckingForEditability):
* css/parser/CSSParserTokenRange.h:
In order to preserve the editability optimization from r180867, when
a style rule's block is consumed, we use a special method that looks
for -webkit-user-modify tokens. If one is seen, then the editability
optimization is turned off. Because we do this at the token level,
the optimization keeps working even if properties don't get parsed.

* css/parser/CSSPropertyParser.cpp:
(WebCore::CSSPropertyParser::parseSingleValue):
(WebCore::CSSPropertyParser::parseFontFaceDescriptor):
(WebCore::CSSPropertyParser::consumeFont):
(WebCore::CSSPropertyParser::parseShorthand):
Patched to remove StyleSheetContents* arguments now that the editability
check is in CSSTokenizer instead.

* css/parser/CSSTokenizer.cpp:
(WebCore::CSSTokenizer::tokenRange):
* css/parser/CSSTokenizer.h:
Make tokenRange() const.

* html/canvas/CanvasRenderingContext2D.cpp:
(WebCore::CanvasRenderingContext2D::setFont):
Patched to drop the StyleSheetContents* argument.

* testing/Internals.cpp:
(WebCore::deferredStyleRulesCountForList):
(WebCore::Internals::deferredStyleRulesCount):
(WebCore::deferredGroupRulesCountForList):
(WebCore::Internals::deferredGroupRulesCount):
(WebCore::deferredKeyframesRulesCountForList):
(WebCore::Internals::deferredKeyframesRulesCount):
* testing/Internals.h:
* testing/Internals.idl:
Add new testing methods that can count the deferred style rules and
group child rule lists. These methods are used by the new layout tests.

LayoutTests:

* fast/css/deferred-parsing: Added.
* fast/css/deferred-parsing/dynamic-external-style-expected.txt: Added.
* fast/css/deferred-parsing/dynamic-external-style.html: Added.
* fast/css/deferred-parsing/dynamic-style-in-document-expected.txt: Added.
* fast/css/deferred-parsing/dynamic-style-in-document.html: Added.
* fast/css/deferred-parsing/hover-test-expected.txt: Added.
* fast/css/deferred-parsing/hover-test.html: Added.
* fast/css/deferred-parsing/keyframes-rule-expected.txt: Added.
* fast/css/deferred-parsing/keyframes-rule.html: Added.
* fast/css/deferred-parsing/media-print-expected.txt: Added.
* fast/css/deferred-parsing/media-print.html: Added.
* fast/css/deferred-parsing/nth-of-type-expected.txt: Added.
* fast/css/deferred-parsing/nth-of-type.html: Added.
* fast/css/deferred-parsing/resources: Added.
* fast/css/deferred-parsing/resources/basic-sheet.css: Added.
* fast/css/deferred-parsing/simple-external-style-expected.txt: Added.
* fast/css/deferred-parsing/simple-external-style.html: Added.
* fast/css/deferred-parsing/simple-style-in-document-expected.txt: Added.
* fast/css/deferred-parsing/simple-style-in-document.html: Added.
* fast/css/deferred-parsing/supports-rule-expected.txt: Added.
* fast/css/deferred-parsing/supports-rule.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209718 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Enable network cache speculative revalidation in MiniBrowser
https://bugs.webkit.org/show_bug.cgi?id=165616

Reviewed by Darin Adler.

Match Safari. Also add a menu item to disable it.

* MiniBrowser/mac/AppDelegate.m:
(defaultConfiguration):
* MiniBrowser/mac/SettingsController.h:
* MiniBrowser/mac/SettingsController.m:
(-[SettingsController _populateMenu]):
(-[SettingsController validateMenuItem:]):
(-[SettingsController networkCacheSpeculativeRevalidationDisabled]):
(-[SettingsController toggleNetworkCacheSpeculativeRevalidationDisabled:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209717 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, rolling out r209672.
https://bugs.webkit.org/show_bug.cgi?id=165766

IndexedDB/objectstore-cursor.html perf test always times out
(Requested by ap on #webkit).

Reverted changeset:

"More IndexedDB perf tests."
https://bugs.webkit.org/show_bug.cgi?id=165634
http://trac.webkit.org/changeset/209672

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209716 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Autoplay isn't working in iPhone MediaDocuments
https://bugs.webkit.org/show_bug.cgi?id=165710
<rdar://problems/29606761>

Reviewed by Darin Adler.
Source/WebCore:

Our silent video autoplay detection disabled autoplay
of videos in MediaDocuments.

Test: media/ios/autoplay-only-in-main-document.html

* html/MediaElementSession.cpp:
(WebCore::MediaElementSession::playbackPermitted):
Return true if we're in a top-level media document.

LayoutTests:

* media/ios/autoplay-only-in-main-document.html: Make sure that
an iframe pointing to media is not autoplayed.
* TestExpectations:
* platform/ios-simulator/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209715 268f45cc-cd09-0410-ab3c-d52691b4dbfc

ChangeLogs become readonly after resolving a conflict
https://bugs.webkit.org/show_bug.cgi?id=165440
rdar://problem/29491002

Reviewed by Darin Adler.

* Scripts/resolve-ChangeLogs: (resolveChangeLog): Set correct permissions when
moving the file.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209714 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Marking http/tests/security/contentSecurityPolicy/block-all-mixed-content/insecure-iframe-in-main-frame.html as flaky on ios-simulator.
https://bugs.webkit.org/show_bug.cgi?id=165691

Unreviewed test gardening.

* platform/ios-simulator/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209713 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Rebaseline js/dom/global-constructors-attributes.html for mac-wk1 after r209666.

Unreviewed test gardening.

* platform/mac-wk1/js/dom/global-constructors-attributes-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209712 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: Cleanup HierarchicalPathComponent
https://bugs.webkit.org/show_bug.cgi?id=165745

Reviewed by Brian Burg.

Prefer toggle for style class names, remove single use CSS class name
constants, and back all properties by the model not the DOM.

* UserInterface/Views/HierarchicalPathComponent.js:
(WebInspector.HierarchicalPathComponent):
(WebInspector.HierarchicalPathComponent.prototype.get element):
(WebInspector.HierarchicalPathComponent.prototype.get representedObject):
(WebInspector.HierarchicalPathComponent.prototype.get minimumWidth):
(WebInspector.HierarchicalPathComponent.prototype.get forcedWidth):
(WebInspector.HierarchicalPathComponent.prototype.set forcedWidth):
(WebInspector.HierarchicalPathComponent.prototype.get hidden):
(WebInspector.HierarchicalPathComponent.prototype.set hidden):
(WebInspector.HierarchicalPathComponent.prototype.get collapsed):
(WebInspector.HierarchicalPathComponent.prototype.set collapsed):
(WebInspector.HierarchicalPathComponent.prototype.get selectorArrows):
(WebInspector.HierarchicalPathComponent.prototype.set selectorArrows):
(WebInspector.HierarchicalPathComponent.prototype.get previousSibling):
(WebInspector.HierarchicalPathComponent.prototype.set previousSibling):
(WebInspector.HierarchicalPathComponent.prototype.get nextSibling):
(WebInspector.HierarchicalPathComponent.prototype.set nextSibling):
(WebInspector.HierarchicalPathComponent.prototype._updateElementTitleAndText):
(WebInspector.HierarchicalPathComponent.prototype._updateSelectElement.createOption):
(WebInspector.HierarchicalPathComponent.prototype._updateSelectElement):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209711 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Re-enable collectContinuously since it doesn't crash anymore

Rubber stamped by Michael Saboff.

* Scripts/run-jsc-stress-tests:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209710 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: Allow keyboard navigation in Heap Snapshot data grids
https://bugs.webkit.org/show_bug.cgi?id=165752

Reviewed by Brian Burg.

* UserInterface/Views/DOMTreeOutline.css:
(.tree-outline.dom:focus li.selected > span::after):
Use new color variable.

* UserInterface/Views/HeapSnapshotInstanceDataGridNode.js:
(WebInspector.HeapSnapshotInstanceDataGridNode.prototype.get selectable): Deleted.
Should be selectable.

* UserInterface/Views/HeapSnapshotInstancesContentView.css:
(.heap-snapshot > .data-grid:focus tr.selected > td .sub-retained):
(.heap-snapshot > .data-grid:focus tr.selected .object-id,):
(.heap-snapshot > .data-grid:focus tr.selected .formatted-string,):
(.heap-snapshot > .data-grid:focus tr.selected td .location):
Selected row text should just be white, except for values that are usually
gray (object ID, object size, and sub-retained size values). These should
be styled like the "= $0" marker in the DOM tree outline.

(.heap-snapshot > .data-grid:matches(:focus, .force-focus) tr.selected td .location): Deleted.
Removed forced-focus reference. Only used by TreeOutlineDataGridSynchronizer.

* UserInterface/Views/Variables.css:
(:root):
Added --selected-secondary-text-color.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209709 268f45cc-cd09-0410-ab3c-d52691b4dbfc

FTL: Dumping disassembly requires that code origin is set when making polymorphic tail calls.
https://bugs.webkit.org/show_bug.cgi?id=165747

Reviewed by Filip Pizlo.

Setting the code origin needs to be done for both the fast and slow path as we might need
it when linking a polymorphic or virtual call stub.

* ftl/FTLLowerDFGToB3.cpp:
(JSC::FTL::DFG::LowerDFGToB3::compileTailCall):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209708 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed test gardening.

Update test expectations on Windows for failing tests. These failures are tracked in
https://bugs.webkit.org/show_bug.cgi?id=160447.

* platform/win/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209707 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: Breakpoint Log action should support template literals
https://bugs.webkit.org/show_bug.cgi?id=165116
<rdar://problem/29464765>

Reviewed by Joseph Pecoraro.

Source/WebInspectorUI:

* Localizations/en.lproj/localizedStrings.js:
New string "${expr} = expression" for breakpoint popover.

* UserInterface/Controllers/BreakpointLogMessageLexer.js: Added.
To support template literal placeholders in message text, it is necessary
to distinguish plain text parts of the message from placeholders, since
plain text must be escaped.

(WebInspector.BreakpointLogMessageLexer):
(WebInspector.BreakpointLogMessageLexer.prototype.tokenize):
(WebInspector.BreakpointLogMessageLexer.prototype.reset):
(WebInspector.BreakpointLogMessageLexer.prototype._finishPlainText):
(WebInspector.BreakpointLogMessageLexer.prototype._finishExpression):
(WebInspector.BreakpointLogMessageLexer.prototype._appendToken):
(WebInspector.BreakpointLogMessageLexer.prototype._consume):
(WebInspector.BreakpointLogMessageLexer.prototype._peek):
(WebInspector.BreakpointLogMessageLexer.prototype._expression):
(WebInspector.BreakpointLogMessageLexer.prototype._plainText):
(WebInspector.BreakpointLogMessageLexer.prototype._possiblePlaceholder):
(WebInspector.BreakpointLogMessageLexer.prototype._regExpOrStringLiteral):

* UserInterface/Controllers/DebuggerManager.js:
(WebInspector.DebuggerManager.prototype._debuggerBreakpointOptions):
Build breakpoint "options" object for DebuggerAgent. If a Log action
contains template literal placeholders it is changed to an Evaluate
action, which calls console.log with a template literal.

(WebInspector.DebuggerManager.prototype._setBreakpoint):
Use converted breakpoint options.

* UserInterface/Main.html:
* UserInterface/Test.html:
New file.

* UserInterface/Views/BreakpointActionView.css:
(.breakpoint-action-block-body > .description):
Styles for breakpoint Log action hint text.

* UserInterface/Views/BreakpointActionView.js:
(WebInspector.BreakpointActionView.prototype._updateBody):
Add Log action hint text element.

LayoutTests:

* inspector/debugger/breakpoint-action-log-expected.txt: Added.
* inspector/debugger/breakpoint-action-log.html: Added.
Check that breakpoint log actions support template literal placeholders.

* inspector/unit-tests/breakpoint-log-message-lexer-expected.txt: Added.
* inspector/unit-tests/breakpoint-log-message-lexer.html: Added.
Check that the lexer correctly scans breakpoint log action text.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209706 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed. Fixing build breakage in GStreamer's WebKitCommonEncryptionDecryptorGStreamer.

* platform/graphics/gstreamer/eme/WebKitCommonEncryptionDecryptorGStreamer.cpp:
(webkitMediaCommonEncryptionDecryptTransformInPlace):
WTF::Seconds should be used instead of std::chrono::seconds.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209705 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, rolling out r209703.
https://bugs.webkit.org/show_bug.cgi?id=165749

Broke Apple builds (Requested by annulen|home on #webkit).

Reverted changeset:

"Unreviewed, use WTF_ATTRIBUTE_PRINTF instead of clang pragma"
http://trac.webkit.org/changeset/209703

Patch by Commit Queue <commit-queue@webkit.org> on 2016-12-12

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209704 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, use WTF_ATTRIBUTE_PRINTF instead of clang pragma

* wtf/PrintStream.cpp:
(WTF::PrintStream::printfVariableFormat):
* wtf/PrintStream.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209703 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed. Try to fix the linux build.

* runtime/StackFrame.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209702 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, guarded clang pragma with COMPILER(CLANG) to fix -Werror

* wtf/PrintStream.cpp:
(WTF::PrintStream::printfVariableFormat):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209701 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed attempt to fix EFL and Mac/cmake builds after r209665.
https://bugs.webkit.org/show_bug.cgi?id=165686

Source/WebKit2:

* PlatformEfl.cmake: WebKit2 should precede other target so that right
config.h comes first in include path.

Tools:

* DumpRenderTree/CMakeLists.txt: WebKit2 should precede other target
so that right config.h comes first in include path.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209700 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[CSS Parser] Make sure content extensions initialize AtomicString
https://bugs.webkit.org/show_bug.cgi?id=165737

Reviewed by Darin Adler.

* TestWebKitAPI/Tests/WebCore/ContentExtensions.cpp:
(TestWebKitAPI::makeBackend):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209699 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: ThreadTreeElement should have a "Resume" status button when paused
https://bugs.webkit.org/show_bug.cgi?id=165581

Reviewed by Joseph Pecoraro.

* UserInterface/Images/Resume.svg:
Fill/stroke should be unspecified so that both can be styled in CSS.

* UserInterface/Views/DebuggerDashboardView.css:
(.dashboard.debugger .navigation-bar .item.debugger-dashboard-pause.activated):
* UserInterface/Views/DebuggerSidebarPanel.css:
(.sidebar > .panel.navigation.debugger > .navigation-bar .debugger-pause-resume):
(.sidebar > .panel.navigation.debugger > .navigation-bar .debugger-pause-resume.activated):
Recreate original style which relied on 'fill="none"' being specified in the SVG.

* UserInterface/Views/ThreadTreeElement.css:
(.tree-outline > .item.thread .icon):
(.tree-outline > .item.thread .status-button.resume):
(.tree-outline > .item.thread .status-button.resume:active):
(.tree-outline:matches(:focus, .force-focus) > .item.thread.selected .status-button.resume):
(.tree-outline > .item.thread.selected .status-button.resume,):
(.details-section.call-stack .thread .icon): Deleted.
Status button styles. Colors match those of the goto-arrow button.

* UserInterface/Views/ThreadTreeElement.js:
(WebInspector.ThreadTreeElement.prototype.refresh):
Update status icon.
(WebInspector.ThreadTreeElement.prototype.oncontextmenu):
(WebInspector.ThreadTreeElement.prototype._updateStatus):
Add/remove status button based on paused state.
Stop propagation of "mousedown" events on the status button, to
prevent button press from selecting the tree element.
(WebInspector.ThreadTreeElement):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209698 268f45cc-cd09-0410-ab3c-d52691b4dbfc

REGRESSION (r200283): Transform, overflow hidden and filter combination completely hides the element
https://bugs.webkit.org/show_bug.cgi?id=161509

Reviewed by David Hyatt.
Source/WebCore:

When the filter painting code needs to recompute a paintDirtyRect, it was using selfClipRect()
which is obviously wrong because it returns a rect in absolute coordinates. Use code factored
out of localClipRect() instead, which returns a rect relative to the painting root.

Test: css3/filters/filter-on-overflow-hidden.html

* rendering/RenderLayer.cpp:
(WebCore::RenderLayer::paintLayerContents):

LayoutTests:

* css3/filters/filter-on-overflow-hidden-expected.html: Added.
* css3/filters/filter-on-overflow-hidden.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209697 268f45cc-cd09-0410-ab3c-d52691b4dbfc

We should be able to throw exceptions from Wasm code and when Wasm frames are on the stack
https://bugs.webkit.org/show_bug.cgi?id=165429

Reviewed by Keith Miller.

JSTests:

* wasm/function-tests/trap-load.js: Added.
(assert):
(wasmFrameCountFromError):
(i.catch):
(assert.continuation):
* wasm/function-tests/trap-store.js: Added.
(import.Builder.from.string_appeared_here.assert):
(i.catch):
(assert.continuation):
(assert):
* wasm/js-api/test_memory_constructor.js:
(assert):

Source/JavaScriptCore:

This patch teaches the stack walking runtime about wasm.
To do this, I taught StackVisitor that a callee is not
always an object.

To be able to unwind callee save registers properly, I've given
JSWebAssemblyCallee a list of RegisterAtOffsetList for the callee
saves that B3 saved in the prologue. Also, because we have two
B3Compilations per wasm function, one for wasm entrypoint, and
one for the JS entrypoint, I needed to create a callee for each
because they each might spill callee save registers.

I also fixed a bug inside the Wasm::Memory constructor where we
were trying to mmap the same number of bytes even after the first
mmap failed. We should start by trying to mmap the maximum bytes,
and if that fails, fall back to the specified initial bytes. However,
the code was just mmapping the maximum twice. I've fixed that and
also added a RELEASE_ASSERT_NOT_REACHED() for when the second mmap
fails along with a FIXME to throw an OOM error.

There was a second bug I fixed where JSModuleRecord was calling
visitWeak on its CallLinkInfos inside ::visitChldren(). It needs
to do this after marking. I changed JSModuleRecord to do what
CodeBlock does and call visitWeak on its CallLinkInfos inside
an UnconditionalFinalizer.

* API/JSContextRef.cpp:
(BacktraceFunctor::operator()):
* inspector/ScriptCallStackFactory.cpp:
(Inspector::createScriptCallStackFromException):
* interpreter/CallFrame.cpp:
(JSC::CallFrame::vmEntryGlobalObject):
* interpreter/CallFrame.h:
(JSC::ExecState::callee):
* interpreter/Interpreter.cpp:
(JSC::GetStackTraceFunctor::operator()):
(JSC::UnwindFunctor::operator()):
(JSC::UnwindFunctor::copyCalleeSavesToVMEntryFrameCalleeSavesBuffer):
* interpreter/Interpreter.h:
* interpreter/ShadowChicken.cpp:
(JSC::ShadowChicken::update):
* interpreter/StackVisitor.cpp:
(JSC::StackVisitor::StackVisitor):
(JSC::StackVisitor::readFrame):
(JSC::StackVisitor::readNonInlinedFrame):
(JSC::StackVisitor::readInlinedFrame):
(JSC::StackVisitor::Frame::isWasmFrame):
(JSC::StackVisitor::Frame::codeType):
(JSC::StackVisitor::Frame::calleeSaveRegisters):
(JSC::StackVisitor::Frame::functionName):
(JSC::StackVisitor::Frame::sourceURL):
(JSC::StackVisitor::Frame::toString):
(JSC::StackVisitor::Frame::hasLineAndColumnInfo):
(JSC::StackVisitor::Frame::setToEnd):
* interpreter/StackVisitor.h:
(JSC::StackVisitor::Frame::callee):
(JSC::StackVisitor::Frame::isNativeFrame):
(JSC::StackVisitor::Frame::isJSFrame): Deleted.
* jsc.cpp:
(callWasmFunction):
(functionTestWasmModuleFunctions):
* runtime/Error.cpp:
(JSC::addErrorInfoAndGetBytecodeOffset):
* runtime/JSCell.cpp:
(JSC::JSCell::isAnyWasmCallee):
* runtime/JSCell.h:
* runtime/JSFunction.cpp:
(JSC::RetrieveArgumentsFunctor::operator()):
(JSC::RetrieveCallerFunctionFunctor::operator()):
* runtime/StackFrame.cpp:
(JSC::StackFrame::sourceID):
(JSC::StackFrame::sourceURL):
(JSC::StackFrame::functionName):
(JSC::StackFrame::computeLineAndColumn):
(JSC::StackFrame::toString):
* runtime/StackFrame.h:
(JSC::StackFrame::StackFrame):
(JSC::StackFrame::hasLineAndColumnInfo):
(JSC::StackFrame::hasBytecodeOffset):
(JSC::StackFrame::bytecodeOffset):
(JSC::StackFrame::isNative): Deleted.
* runtime/VM.h:
* wasm/WasmB3IRGenerator.cpp:
(JSC::Wasm::B3IRGenerator::B3IRGenerator):
(JSC::Wasm::createJSToWasmWrapper):
(JSC::Wasm::parseAndCompile):
* wasm/WasmCallingConvention.h:
(JSC::Wasm::CallingConvention::setupFrameInPrologue):
* wasm/WasmFormat.h:
* wasm/WasmMemory.cpp:
(JSC::Wasm::Memory::Memory):
* wasm/WasmMemory.h:
(JSC::Wasm::Memory::isValid):
* wasm/WasmPlan.cpp:
(JSC::Wasm::Plan::run):
(JSC::Wasm::Plan::initializeCallees):
* wasm/WasmPlan.h:
(JSC::Wasm::Plan::jsToWasmEntryPointForFunction): Deleted.
* wasm/js/JSWebAssemblyCallee.cpp:
(JSC::JSWebAssemblyCallee::finishCreation):
* wasm/js/JSWebAssemblyCallee.h:
(JSC::JSWebAssemblyCallee::create):
(JSC::JSWebAssemblyCallee::entrypoint):
(JSC::JSWebAssemblyCallee::calleeSaveRegisters):
(JSC::JSWebAssemblyCallee::jsToWasmEntryPoint): Deleted.
* wasm/js/JSWebAssemblyModule.cpp:
(JSC::JSWebAssemblyModule::JSWebAssemblyModule):
(JSC::JSWebAssemblyModule::visitChildren):
(JSC::JSWebAssemblyModule::UnconditionalFinalizer::finalizeUnconditionally):
* wasm/js/JSWebAssemblyModule.h:
(JSC::JSWebAssemblyModule::jsEntrypointCalleeFromFunctionIndexSpace):
(JSC::JSWebAssemblyModule::wasmEntrypointCalleeFromFunctionIndexSpace):
(JSC::JSWebAssemblyModule::setJSEntrypointCallee):
(JSC::JSWebAssemblyModule::setWasmEntrypointCallee):
(JSC::JSWebAssemblyModule::allocationSize):
(JSC::JSWebAssemblyModule::calleeFromFunctionIndexSpace): Deleted.
* wasm/js/JSWebAssemblyRuntimeError.h:
* wasm/js/WebAssemblyFunction.cpp:
(JSC::WebAssemblyFunction::call):
* wasm/js/WebAssemblyInstanceConstructor.cpp:
(JSC::constructJSWebAssemblyInstance):
* wasm/js/WebAssemblyMemoryConstructor.cpp:
(JSC::constructJSWebAssemblyMemory):
* wasm/js/WebAssemblyModuleConstructor.cpp:
(JSC::constructJSWebAssemblyModule):
* wasm/js/WebAssemblyModuleRecord.cpp:
(JSC::WebAssemblyModuleRecord::link):

Source/WebCore:

* bindings/js/JSDOMBinding.cpp:
(WebCore::GetCallerGlobalObjectFunctor::operator()):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209696 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remove uses of Dictionary in WebRTC IDL files
https://bugs.webkit.org/show_bug.cgi?id=165736

Reviewed by Sam Weinig.

Source/WebCore:

Also removed quite a bit of unused code. There were some mocks that were out of date and
no longer matched the types used in the real code, that also were no longer hooked up,
and other types of dead code. We will have to implement anew when we want to restore tests
like the ones these were intended to enable.

* CMakeLists.txt: Updated for all the removed files.

* Modules/mediastream/MediaEndpointPeerConnection.cpp: Moved some types in here
that don't need to be in the header.
(WebCore::MediaEndpointPeerConnection::MediaEndpointPeerConnection): Use make_unique.
(WebCore::MediaEndpointPeerConnection::setConfiguration): Changed argument type,
obviating the need for most of the code that was here.
* Modules/mediastream/MediaEndpointPeerConnection.h: Updated for the above. Made more
things private and final. Marked the constructor explicit. Removed unneeded includes.

* Modules/mediastream/PeerConnectionBackend.h: Removed unneeded includes and forward
declarations. Changed the argument type for setConfiguration (see above).

* Modules/mediastream/RTCConfiguration.cpp: Removed.
None of the code here was needed except for the ICE server validation, and that was
moved into RTCPeerConnection::setConfiguration.
* Modules/mediastream/RTCConfiguration.h: Changed this from a class to a struct since
this is now a dictionary rather than an interface.
* Modules/mediastream/RTCConfiguration.idl: Changed this from an interface to a dictionary.

* Modules/mediastream/RTCDTMFSender.cpp: Removed some of the code from this file. This
class currently isn't implemented, but was depending on RTCPeerConnectionHandler. I removed
some of the dead code. Someone will have to straighten this out so we can turn it back on.
* Modules/mediastream/RTCDTMFSender.h: Ditto.

* Modules/mediastream/RTCDataChannel.cpp: Updated includes.

* Modules/mediastream/RTCIceCandidate.cpp:
(WebCore::RTCIceCandidate::create): Removed most of the code, since this now takes
a structure rather than a WebCore::Dictionary, and so the bindings take care of the work.
* Modules/mediastream/RTCIceCandidate.h: Updated for the above.
* Modules/mediastream/RTCIceCandidate.idl: Changed the constructor to take a
RTCIceCandidateInit instead of a Dictionary.

* Modules/mediastream/RTCIceServer.h: Changed this from a class to a struct since
this is now a dictionary rather than an interface.
* Modules/mediastream/RTCIceServer.idl: Changed this from an interface to a dictionary.

* Modules/mediastream/RTCPeerConnection.cpp:
(WebCore::RTCPeerConnection::initializeWith): Take an RTCConfiguration rather than
a Dictionary.
(WebCore::RTCPeerConnection::getConfiguration): Moved to header, now an inline.
(WebCore::RTCPeerConnection::setConfiguration): Use the configuration dictionary now instead
of the configuration class. Also moved validation of server URLs here, formerly in the
RTCConfiguration RTCConfiguration::initialize function. Also moved code to convert from
RTCConfiguration to MediaEndpointConfiguration here, formerly in
MediaEndpointPeerConnection::setConfiguration.
* Modules/mediastream/RTCPeerConnection.h: Updated for the above.
* Modules/mediastream/RTCPeerConnection.idl: Changed argument to initializeWith and
setConfiguration from Dictionary to RTCConfiguration.

* Modules/mediastream/RTCSessionDescription.cpp:
(WebCore::parseTypeString): Deleted.
(WebCore::RTCSessionDescription::create): Changed to take struct instead of Dictionary.
* Modules/mediastream/RTCSessionDescription.h: Added Init struct and changed create to
take it.
* Modules/mediastream/RTCSessionDescription.idl: Changed constructor argument to take
RTCSessionDescriptionInit dictionary instead of Dictionary.

* Modules/webaudio/MediaStreamAudioDestinationNode.cpp: Updated includes.

* WebCore.xcodeproj/project.pbxproj: Updated for all the removed files.

* bindings/scripts/CodeGeneratorJS.pm:
(GenerateDictionaryImplementationContent): Fixed convertDictionaryToJS to work with other
nullable types besides RefPtr.
* bindings/scripts/test/JS/JSTestObj.cpp: Regenerated.

* platform/mediastream/RTCConfigurationPrivate.h: Removed.
* platform/mediastream/RTCIceServerPrivate.h: Removed.
* platform/mediastream/RTCPeerConnectionHandler.cpp: Removed.
* platform/mediastream/RTCPeerConnectionHandler.h: Removed.
* platform/mock/RTCDTMFSenderHandlerMock.cpp: Removed.
* platform/mock/RTCDTMFSenderHandlerMock.h: Removed.

* platform/mock/RTCDataChannelHandlerMock.h: Updated includes, made more things private and final.

* platform/mock/RTCPeerConnectionHandlerMock.cpp: Removed.
* platform/mock/RTCPeerConnectionHandlerMock.h: Removed.

* testing/Internals.cpp:
(WebCore::Internals::Internals): Removed call to enableMockRTCPeerConnectionHandler.
(WebCore::Internals::enableMockRTCPeerConnectionHandler): Deleted.

LayoutTests:

* fast/mediastream/RTCIceCandidate-expected.txt: Regenerated; no new failures.
* fast/mediastream/RTCIceCandidate.html: Updated test to expect the WebIDL rules for parsing
unsigned short for sdpMLineIndex. The old code expected additional range checking which is
not expected behavior.

* fast/mediastream/RTCPeerConnection-expected.txt: Regenerated; no new failures.
* fast/mediastream/RTCPeerConnection.html: Updated test to not expect an exception when
iceServers is entirely omitted or when it is an empty array. This is a progression and is
closer to the specification.

* fast/mediastream/RTCPeerConnection-getConfiguration-expected.txt: Regenerated; no new failures.
* fast/mediastream/RTCPeerConnection-getConfiguration.html: Updated test to expect the username
and credential fields to round trip as missing rather than turning from missing into empty strings.
Also updated the tests to pass in sequences of strings for "urls" since they expect sequences on
the way back out. We could also test round tripping of single strings, which should go in and come
back out as individual strings, but for now I did not bother adding that to this test.

* fast/mediastream/RTCSessionDescription-expected.txt: Regenerated; some exception messages are
now more specific than they were before.

* imported/w3c/web-platform-tests/webrtc/rtcpeerconnection/rtcpeerconnection-constructor-expected.txt:
Updated expected results. The new code correctly allows callers of the RTCPeerConnection constructor
to omit iceServers, pass undefined, or pass an empty array, as specified. Thuis change leads to
a couple tests that now correctly pass instead of failing. And it also leads to some other tests
failing that were passing before. These were passing because they expected an exception for some
other reason, however the exception was actually occurring because of incorrect handling of iceServers.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209695 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Re-enable concurrent GC.

Rubber stampted by Saam Barati.

This change actually landed in r209692 by accident.

* runtime/Options.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209694 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Change to use #pragma once (requested by Darin Adler).

* wtf/RecursiveLockAdapter.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209693 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Change to use #pragma once (requested by Darin Adler).

* wtf/LockedPrintStream.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209692 268f45cc-cd09-0410-ab3c-d52691b4dbfc

MarkedBlock::marksConveyLivenessDuringMarking should take into account collection scope
https://bugs.webkit.org/show_bug.cgi?id=165741

Reviewed by Saam Barati.

Source/JavaScriptCore:

MarkedBlock::marksConveyLivenessDuringMarking thought that the off-by-one marking
version indicated liveness during any collection when it's just during full collection.
One of its users - MarkedBlock::sweep - knew this and had a special case, but the other
one - MarkedBlock::isLive - didn't. So, I moved the special case into
marksConveyLivenessDuringMarking.

Also, this cleans up some remaining bitvector races.

To find this bug, I significantly strengthened our assertions.

* CMakeLists.txt:
* JavaScriptCore.xcodeproj/project.pbxproj:
* heap/CellContainer.cpp: Added.
(JSC::CellContainer::isNewlyAllocated):
* heap/CellContainer.h:
* heap/MarkedAllocator.cpp:
(JSC::MarkedAllocator::addBlock):
(JSC::MarkedAllocator::removeBlock):
(JSC::MarkedAllocator::dumpBits):
* heap/MarkedAllocator.h:
(JSC::MarkedAllocator::forEachBitVector):
(JSC::MarkedAllocator::forEachBitVectorWithName):
* heap/MarkedBlock.cpp:
(JSC::MarkedBlock::tryCreate):
(JSC::MarkedBlock::Handle::~Handle):
(JSC::MarkedBlock::MarkedBlock):
(JSC::MarkedBlock::Handle::specializedSweep):
(JSC::MarkedBlock::Handle::sweepHelperSelectMarksMode):
(JSC::MarkedBlock::Handle::stopAllocating):
(JSC::MarkedBlock::Handle::resumeAllocating):
(JSC::MarkedBlock::aboutToMarkSlow):
(JSC::MarkedBlock::Handle::didConsumeFreeList):
(JSC::MarkedBlock::Handle::dumpState):
* heap/MarkedBlock.h:
(JSC::MarkedBlock::markingVersion):
(JSC::MarkedBlock::isMarkedRaw):
(JSC::MarkedBlock::isMarked):
* heap/MarkedBlockInlines.h:
(JSC::MarkedBlock::marksConveyLivenessDuringMarking):
* heap/SlotVisitor.cpp:
(JSC::SlotVisitor::appendJSCellOrAuxiliary):
* runtime/Options.cpp:
(JSC::recomputeDependentOptions):
* runtime/StructureIDTable.h:
(JSC::StructureIDTable::size):
(JSC::StructureIDTable::get):

Source/WTF:

To find this bug, I needed to seriously beef up our logging infrastructure.

It's now the case that:

    dataLog(...);

will print its output atomically. This happens with some careful magic:

- dataFile() is now a LockedPrintStream that locks around print().

- The lock is a recursive lock via RecursiveLockAdapter<>, so if the dump methods end
  up calling back into dataLog() then it just works. This is important: say the dump()
  calls a getter that itself does logging, maybe because it's encountering badness and
  wants to report it before crashing).

- The lock is a WordLock so that ParkingLot and Lock can keep using dataLog() for
  debugging. We probably won't need to debug WordLock anytime soon - the algorithm is
  so simple.

- LockedPrintStream::print(...) causes the print callbacks of its arguments to run on
  the underlying PrintStream, so that you don't need to do recursive lock acquisition
  on each individual argument and whatever printing it does recursively.

* WTF.xcodeproj/project.pbxproj:
* wtf/CMakeLists.txt:
* wtf/DataLog.cpp:
(WTF::initializeLogFileOnce):
(WTF::initializeLogFile):
(WTF::dataFile):
* wtf/DataLog.h:
* wtf/LockedPrintStream.cpp: Added.
(WTF::LockedPrintStream::LockedPrintStream):
(WTF::LockedPrintStream::~LockedPrintStream):
(WTF::LockedPrintStream::vprintf):
(WTF::LockedPrintStream::flush):
(WTF::LockedPrintStream::begin):
(WTF::LockedPrintStream::end):
* wtf/LockedPrintStream.h: Added.
* wtf/PrintStream.cpp:
(WTF::PrintStream::printfVariableFormat):
(WTF::PrintStream::begin):
(WTF::PrintStream::end):
* wtf/PrintStream.h:
(WTF::PrintStream::atomically):
(WTF::PrintStream::print):
(WTF::PrintStream::println):
(WTF::PrintStream::printImpl):
(WTF::>::unpack):
(WTF::FormatImpl::FormatImpl):
(WTF::FormatImpl::dump):
(WTF::format):
(WTF::printInternal):
* wtf/RecursiveLockAdapter.h: Added.
(WTF::RecursiveLockAdapter::RecursiveLockAdapter):
(WTF::RecursiveLockAdapter::lock):
(WTF::RecursiveLockAdapter::unlock):
(WTF::RecursiveLockAdapter::tryLock):
(WTF::RecursiveLockAdapter::isLocked):
* wtf/WordLock.cpp:
* wtf/WordLock.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209691 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Quarter-second stalls scrolling images that are links because of sync getPositionInformation
https://bugs.webkit.org/show_bug.cgi?id=165707

Reviewed by Simon Fraser.

* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView positionInformationForActionSheetAssistant:]):
(-[WKContentView updatePositionInformationForActionSheetAssistant:]):
For now, synchronously block when presenting the action sheet if we
don't have a snapshot ready.

* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::getPositionInformation):
Respect includeSnapshot in another place in getPositionInformation,
so that many getPositionInformations don't have to do extra snapshotting work.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209690 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Use std::vsnprintf instead of vasprintf
https://bugs.webkit.org/show_bug.cgi?id=165740

Reviewed by Sam Weinig.

Source/WebCore:

* platform/FileHandle.cpp:
(WebCore::FileHandle::printf): Use vsnprintf, including StringExtras.h to
ensure compatibility with older versions of the Visual Studio library,
and Vector for the buffer. Use inline capacity in the vector so we normally
don't need to allocate any memory on the heap.
* xml/XSLTUnicodeSort.cpp:
(xsltTransformErrorTrampoline): Ditto.
* xml/parser/XMLDocumentParserLibxml2.cpp:
(WebCore::XMLDocumentParser::error): Ditto.

Source/WTF:

* wtf/Platform.h: Remove HAVE_VASPRINTF.
* wtf/StringExtras.h: Change the vsnprintf workaround to be used only
in older versions of Visual Studio, since the problem it works around
was resolved in Visual Studio 2015.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209689 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Make some refinements to HTMLPlugInImageElement
https://bugs.webkit.org/show_bug.cgi?id=165742

Reviewed by Sam Weinig.

* html/HTMLPlugInImageElement.cpp: Removed many unneeded includes.
(WebCore::titleText): Use HashMap::ensure, correct argument types.
(WebCore::subtitleText): Ditto.
(WebCore::HTMLPlugInImageElement::HTMLPlugInImageElement): Moved
initialization of most scalars to the class definition.
(WebCore::HTMLPlugInImageElement::isImageType): Use auto.
(WebCore::HTMLPlugInImageElement::wouldLoadAsPlugIn): Removed
unneeded local variable and if statement.
(WebCore::HTMLPlugInImageElement::willDetachRenderers): Use auto.
(WebCore::HTMLPlugInImageElement::updateSnapshot): Do not call
renderer twice unnecessarily.
(WebCore::plugInImageElementIsolatedWorld): Use auto.
(WebCore::HTMLPlugInImageElement::didAddUserAgentShadowRoot): Use auto.
(WebCore::HTMLPlugInImageElement::partOfSnapshotOverlay): Use auto.
(WebCore::HTMLPlugInImageElement::restartSimilarPlugIns): Got rid of
unnecessary typedef.
(WebCore::HTMLPlugInImageElement::userDidClickSnapshot): Take a
reference rather than PassRefPtr.
(WebCore::documentHadRecentUserGesture): Removed unneeded if.
(WebCore::HTMLPlugInImageElement::checkSizeChangeForSnapshotting): Use auto.
(WebCore::isSmallerThanTinySizingThreshold): Use auto.
(WebCore::HTMLPlugInImageElement::isTopLevelFullPagePlugin): Use auto.
(WebCore::HTMLPlugInImageElement::checkSnapshotStatus): Use auto.
(WebCore::HTMLPlugInImageElement::subframeLoaderWillCreatePlugIn): Got rid
of unnecessary local variable. Use auto.
(WebCore::HTMLPlugInImageElement::requestObject): Got rid of unnecessary
local variable.
* html/HTMLPlugInImageElement.h: Updated for the above. Removed incorrect
use of const. Initialized all scalars. Made function private and final.

* rendering/RenderSnapshottedPlugIn.cpp:
(WebCore::RenderSnapshottedPlugIn::handleEvent): Pass reference instead
of pointer.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209688 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: Cleanup some InspectorInstrumentation network loading notifications
https://bugs.webkit.org/show_bug.cgi?id=165688

Patch by Joseph Pecoraro <pecoraro@apple.com> on 2016-12-11
Reviewed by Brian Burg.

Eliminate InspectorInstrumentation::willReceiveResourceResponse by folding
it into InspectorInstrumentation::didReceiveResourceResponse. Simplify other
related InspectorInstrumentation calls by using references and using more
consistent InspectorInstrumentation patterns.

* inspector/InspectorInstrumentation.cpp:
(WebCore::InspectorInstrumentation::didReceiveResourceResponseImpl):
(WebCore::InspectorInstrumentation::willReceiveResourceResponseImpl): Deleted.
(WebCore::InspectorInstrumentation::didReceiveResourceResponseButCanceledImpl): Deleted.
(WebCore::InspectorInstrumentation::continueAfterXFrameOptionsDeniedImpl): Deleted.
(WebCore::InspectorInstrumentation::continueWithPolicyDownloadImpl): Deleted.
(WebCore::InspectorInstrumentation::continueWithPolicyIgnoreImpl): Deleted.
* inspector/InspectorInstrumentation.h:
(WebCore::InspectorInstrumentation::didReceiveResourceResponse):
(WebCore::InspectorInstrumentation::continueAfterXFrameOptionsDenied):
(WebCore::InspectorInstrumentation::continueWithPolicyDownload):
(WebCore::InspectorInstrumentation::continueWithPolicyIgnore):
(WebCore::InspectorInstrumentation::willReceiveResourceResponse): Deleted.
* loader/CrossOriginPreflightChecker.cpp:
(WebCore::CrossOriginPreflightChecker::validatePreflightResponse):
* loader/DocumentLoader.cpp:
(WebCore::DocumentLoader::stopLoadingAfterXFrameOptionsOrContentSecurityPolicyDenied):
(WebCore::DocumentLoader::continueAfterContentPolicy):
* loader/ResourceLoadNotifier.cpp:
(WebCore::ResourceLoadNotifier::dispatchDidReceiveResponse):
* loader/appcache/ApplicationCacheGroup.cpp:
(WebCore::ApplicationCacheGroup::didReceiveResponse):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209687 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: Move MainTarget and WorkerTarget to their own files
https://bugs.webkit.org/show_bug.cgi?id=165701

Patch by Joseph Pecoraro <pecoraro@apple.com> on 2016-12-11
Reviewed by Brian Burg.

* UserInterface/Controllers/DebuggerManager.js:
(WebInspector.DebuggerManager.prototype.scriptDidParse):
* UserInterface/Main.html:
* UserInterface/Protocol/MainTarget.js: Added.
(WebInspector.MainTarget):
(WebInspector.MainTarget.prototype.get displayName):
(WebInspector.MainTarget.prototype.get mainResource):
* UserInterface/Protocol/Target.js:
(WebInspector.Target):
(WebInspector.MainTarget): Deleted.
(WebInspector.MainTarget.prototype.get displayName): Deleted.
(WebInspector.MainTarget.prototype.get mainResource): Deleted.
(WebInspector.MainTarget.prototype.initialize): Deleted.
(WebInspector.WorkerTarget): Deleted.
(WebInspector.WorkerTarget.prototype.get displayName): Deleted.
(WebInspector.WorkerTarget.prototype.initialize): Deleted.
* UserInterface/Protocol/WorkerTarget.js: Added.
(WebInspector.WorkerTarget):
(WebInspector.WorkerTarget.prototype.get displayName):
* UserInterface/Test.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209686 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[MediaStream] Protect MediaDevicesRequest during callback
https://bugs.webkit.org/show_bug.cgi?id=165711
<rdar://problem/28400468>

Reviewed by Sam Weinig.

No new tests, I was unable to create a reproducible test but this fix avoids
an occasional crash in existing tests.

* Modules/mediastream/MediaDevicesEnumerationRequest.cpp:
(WebCore::MediaDevicesEnumerationRequest::start): Take a reference to the object
before calling the controller in case the completion handler is called synchronously.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209685 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[WebCrypto] Add tests for wrap/unwrap SPKI/PKCS8 keys
https://bugs.webkit.org/show_bug.cgi?id=165557

Reviewed by Darin Adler.

* crypto/subtle/aes-cbc-import-key-unwrap-pkcs8-key-expected.txt: Added.
* crypto/subtle/aes-cbc-import-key-unwrap-pkcs8-key.html: Added.
* crypto/subtle/aes-cbc-import-key-unwrap-spki-key-expected.txt: Added.
* crypto/subtle/aes-cbc-import-key-unwrap-spki-key.html: Added.
* crypto/subtle/aes-cbc-import-key-wrap-pkcs8-key-expected.txt: Added.
* crypto/subtle/aes-cbc-import-key-wrap-pkcs8-key.html: Added.
* crypto/subtle/aes-cbc-import-key-wrap-spki-key-expected.txt: Added.
* crypto/subtle/aes-cbc-import-key-wrap-spki-key.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209684 268f45cc-cd09-0410-ab3c-d52691b4dbfc

The DOM should have an advancing wavefront opaque root barrier
https://bugs.webkit.org/show_bug.cgi?id=165712

Reviewed by Yusuke Suzuki.
Source/JavaScriptCore:

This exposes the ability to fire an advancing wavefront barrier on opaque roots. It also
gives clients the ability to maintain their own cache of whether that barrier needs to
be enabled.

The DOM uses this to enable a very cheap barrier on the DOM. This is neutral on
Speedometer and fixes another concurrent GC crash.

* heap/Heap.cpp:
(JSC::Heap::beginMarking):
(JSC::Heap::endMarking):
(JSC::Heap::writeBarrierOpaqueRootSlow):
(JSC::Heap::addMutatorShouldBeFencedCache):
(JSC::Heap::setMutatorShouldBeFenced):
* heap/Heap.h:
* heap/HeapInlines.h:
(JSC::writeBarrierOpaqueRoot):

Source/WebCore:

No new tests because this was covered by crashing tests.

Consider these two cases:

   Removal:
   1) DOM at start: D->X->Y
   2) Mark X, X->visitChildren, addOpaqueRoot(D)
   3) remove X
   4) Y thinks it's not reachable (its opaque root, X, is not in the set).

   Insertion:
   1) DOM at start: D, X->Y
   2) Mark X, X->visitChildren, addOpaqueRoot(X)
   3) insert X into D
   4) Y thinks it's not reachable (its opaque root, D, is not in the set).

We can fix this with two barriers:

   Removal: add X (the removed child) to the opaque root set.
   Insertion: add D (the insertion point) to the opaque root set.

Thanks Rysosuke for coming up with this idea!

Both barriers advance the wavefront. We could consider retreating wavefront barriers in
the future (where we cause visitChildren to be called again on wrappers that belonged to
roots that got affected by insertion/removal) but those would probably require more
bookkeeping.

To make this barrier very fast, the WebCore caches the JSC VM's barrier state in
its own global variable for very fast access. This variable will be false most of the
time. It's false when there is no VM, so triggering the barrier won't cause the VM to be
created. It's only true when GC is running, which is rare by design.

To make that caching more sensible, I finally gave WebCore a central header for
the common VM (CommonVM.h).

* CMakeLists.txt:
* Modules/mediastream/SDPProcessor.cpp:
(WebCore::SDPProcessor::callScript):
* WebCore.xcodeproj/project.pbxproj:
* bindings/js/CommonVM.cpp: Added.
(WebCore::commonVMSlow):
(WebCore::writeBarrierOpaqueRootSlow):
* bindings/js/CommonVM.h: Added.
(WebCore::commonVM):
(WebCore::writeBarrierOpaqueRoot):
* bindings/js/DOMWrapperWorld.cpp:
(WebCore::mainThreadNormalWorld):
* bindings/js/GCController.cpp:
(WebCore::collect):
(WebCore::GCController::garbageCollectSoon):
(WebCore::GCController::garbageCollectNow):
(WebCore::GCController::garbageCollectNowIfNotDoneRecently):
(WebCore::GCController::setJavaScriptGarbageCollectorTimerEnabled):
(WebCore::GCController::deleteAllCode):
(WebCore::GCController::deleteAllLinkedCode):
* bindings/js/JSCustomXPathNSResolver.cpp:
(WebCore::JSCustomXPathNSResolver::lookupNamespaceURI):
* bindings/js/JSDOMBinding.cpp:
(WebCore::addImpureProperty):
* bindings/js/JSDOMWindowBase.cpp:
(WebCore::JSDOMWindowBase::fireFrameClearedWatchpointsForWindow):
(WebCore::JSDOMWindowBase::commonVM): Deleted.
* bindings/js/JSDOMWindowBase.h:
* bindings/js/JSDOMWindowShell.cpp:
(WebCore::JSDOMWindowShell::setWindow):
* bindings/js/JSNodeCustom.h:
(WebCore::root):
* bindings/js/ScriptCachedFrameData.cpp:
(WebCore::ScriptCachedFrameData::ScriptCachedFrameData):
(WebCore::ScriptCachedFrameData::restore):
(WebCore::ScriptCachedFrameData::clear):
* bindings/js/ScriptController.cpp:
(WebCore::ScriptController::~ScriptController):
(WebCore::ScriptController::createWorld):
(WebCore::ScriptController::getAllWorlds):
(WebCore::ScriptController::clearWindowShell):
(WebCore::ScriptController::cacheableBindingRootObject):
(WebCore::ScriptController::bindingRootObject):
(WebCore::ScriptController::windowScriptNPObject):
(WebCore::ScriptController::jsObjectForPluginElement):
(WebCore::ScriptController::clearScriptObjects):
* dom/CollectionIndexCache.cpp:
(WebCore::reportExtraMemoryAllocatedForCollectionIndexCache):
* dom/ContainerNode.cpp:
* dom/ContainerNodeAlgorithms.cpp:
(WebCore::notifyChildNodeInserted):
(WebCore::notifyChildNodeRemoved):
* dom/Document.cpp:
(WebCore::Document::shouldBypassMainWorldContentSecurityPolicy):
* dom/Node.h:
(WebCore::Node::opaqueRoot):
* dom/ScriptExecutionContext.cpp:
(WebCore::ScriptExecutionContext::vm):
* html/HTMLImageLoader.cpp:
(WebCore::HTMLImageLoader::notifyFinished):
* html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::pauseAfterDetachedTask):
(WebCore::HTMLMediaElement::ensureIsolatedWorld):
* html/HTMLPlugInImageElement.cpp:
(WebCore::plugInImageElementIsolatedWorld):
* inspector/InspectorController.cpp:
(WebCore::InspectorController::vm):
* inspector/PageScriptDebugServer.cpp:
(WebCore::PageScriptDebugServer::PageScriptDebugServer):
* page/PerformanceLogging.cpp:
(WebCore::PerformanceLogging::memoryUsageStatistics):
(WebCore::PerformanceLogging::javaScriptObjectCounts):
* page/ResourceUsageThread.cpp:
(WebCore::ResourceUsageThread::createThreadIfNeeded):
* svg/graphics/SVGImage.cpp:
(WebCore::SVGImage::reportApproximateMemoryCost):
* testing/MemoryInfo.h:
(WebCore::MemoryInfo::MemoryInfo):

Source/WebKit/mac:

Propagate the JSDOMWindowBase::commonVM() -> commonVM() change.

* Misc/WebCoreStatistics.mm:
(+[WebCoreStatistics javaScriptObjectsCount]):
(+[WebCoreStatistics javaScriptGlobalObjectsCount]):
(+[WebCoreStatistics javaScriptProtectedObjectsCount]):
(+[WebCoreStatistics javaScriptProtectedGlobalObjectsCount]):
(+[WebCoreStatistics javaScriptProtectedObjectTypeCounts]):
(+[WebCoreStatistics javaScriptObjectTypeCounts]):
(+[WebCoreStatistics shouldPrintExceptions]):
(+[WebCoreStatistics setShouldPrintExceptions:]):
(+[WebCoreStatistics memoryStatistics]):
(+[WebCoreStatistics javaScriptReferencedObjectsCount]):
* Plugins/Hosted/NetscapePluginHostProxy.mm:
(identifierFromIdentifierRep):
* Plugins/Hosted/ProxyInstance.mm:
(WebKit::ProxyInstance::getPropertyNames):
* Plugins/WebNetscapePluginStream.mm:
(WebNetscapePluginStream::wantsAllStreams):
* Plugins/WebNetscapePluginView.mm:
(-[WebNetscapePluginView sendEvent:isDrawRect:]):
(-[WebNetscapePluginView privateBrowsingModeDidChange]):
(-[WebNetscapePluginView setWindowIfNecessary]):
(-[WebNetscapePluginView createPluginScriptableObject]):
(-[WebNetscapePluginView getFormValue:]):
(-[WebNetscapePluginView evaluateJavaScriptPluginRequest:]):
(-[WebNetscapePluginView webFrame:didFinishLoadWithReason:]):
(-[WebNetscapePluginView loadPluginRequest:]):
(-[WebNetscapePluginView _printedPluginBitmap]):
* Plugins/WebPluginController.mm:
(-[WebPluginController plugInViewWithArguments:fromPluginPackage:]):
(-[WebPluginController stopOnePlugin:]):
(-[WebPluginController stopOnePluginForPageCache:]):
(-[WebPluginController destroyOnePlugin:]):
(-[WebPluginController startAllPlugins]):
(-[WebPluginController addPlugin:]):

Source/WebKit/win:

Propagate the JSDOMWindowBase::commonVM() -> commonVM() change.

* Plugins/PluginView.cpp:
(WebCore::PluginView::start):
(WebCore::PluginView::stop):
(WebCore::PluginView::performRequest):
(WebCore::PluginView::npObject):
(WebCore::PluginView::privateBrowsingStateChanged):
* Plugins/PluginViewWin.cpp:
(WebCore::PluginView::dispatchNPEvent):
(WebCore::PluginView::handleKeyboardEvent):
(WebCore::PluginView::handleMouseEvent):
(WebCore::PluginView::setNPWindowRect):
* WebCoreStatistics.cpp:
(WebCoreStatistics::javaScriptObjectsCount):
(WebCoreStatistics::javaScriptGlobalObjectsCount):
(WebCoreStatistics::javaScriptProtectedObjectsCount):
(WebCoreStatistics::javaScriptProtectedGlobalObjectsCount):
(WebCoreStatistics::javaScriptProtectedObjectTypeCounts):
(WebCoreStatistics::javaScriptObjectTypeCounts):
(WebCoreStatistics::shouldPrintExceptions):
(WebCoreStatistics::setShouldPrintExceptions):
(WebCoreStatistics::memoryStatistics):
* WebJavaScriptCollector.cpp:
(WebJavaScriptCollector::objectCount):

Source/WebKit2:

Propagate the JSDOMWindowBase::commonVM() -> commonVM() change.

* Shared/linux/WebMemorySamplerLinux.cpp:
(WebKit::WebMemorySampler::sampleWebKit):
* Shared/mac/WebMemorySampler.mac.mm:
(WebKit::WebMemorySampler::sampleWebKit):
* WebProcess/InjectedBundle/InjectedBundle.cpp:
(WebKit::InjectedBundle::javaScriptObjectsCount):
* WebProcess/Plugins/Netscape/JSNPObject.cpp:
(WebKit::JSNPObject::callMethod):
(WebKit::JSNPObject::callObject):
(WebKit::JSNPObject::callConstructor):
(WebKit::JSNPObject::put):
(WebKit::JSNPObject::deleteProperty):
(WebKit::JSNPObject::getOwnPropertyNames):
(WebKit::JSNPObject::propertyGetter):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::runJavaScriptInMainFrame):
(WebKit::WebPage::getBytecodeProfile):
(WebKit::WebPage::getSamplingProfilerOutput):
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::getWebCoreStatistics):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209683 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Cocoa] NSAttributedString representation of text copied from -webkit-nbsp-mode:space element contains non-breaking space characters, but shouldn’t
https://bugs.webkit.org/show_bug.cgi?id=165515
<rdar://problem/4108460>

Reviewed by Darin Adler.

Source/WebCore:

Test: platform/mac/fast/text/attributed-substring-from-range.html

* editing/cocoa/HTMLConverter.mm:
(HTMLConverter::_processText): Emit a space instead of a non-breaking space if the text node
  is styled with -webkit-nbsp-mode:space.
(WebCore::editingAttributedStringFromRange): Replace all non-breaking spaces with spaces if
  they come from a text node with -webkit-nbsp-mode:space.

LayoutTests:

* platform/mac/fast/text/attributed-substring-from-range-expected.txt: Updated.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209682 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed build fix for EFL after r209665

* UIProcess/efl/TextCheckerClientEfl.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209681 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, add KHR include dir to fix ANGLE build after r209665
https://bugs.webkit.org/show_bug.cgi?id=165686

* CMakeLists.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209680 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Rolling out 165737, since it broke layout tests. I need to find a
different place to put the init.

* contentextensions/ContentExtensionParser.cpp:
(WebCore::ContentExtensions::isValidSelector):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209679 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, rolling out r209653, r209654, r209663, and
r209673.
https://bugs.webkit.org/show_bug.cgi?id=165739

speedometer crashes (Requested by pizlo on #webkit).

Reverted changesets:

"JSVALUE64: Pass arguments in platform argument registers when
making JavaScript calls"
https://bugs.webkit.org/show_bug.cgi?id=160355
http://trac.webkit.org/changeset/209653

"Unreviewed build fix for 32 bit builds."
http://trac.webkit.org/changeset/209654

"Unreviewed build fix for the CLOOP after r209653"
http://trac.webkit.org/changeset/209663

"REGRESSION(r209653) Crash in CallFrameShuffler::snapshot()"
https://bugs.webkit.org/show_bug.cgi?id=165728
http://trac.webkit.org/changeset/209673

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209678 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Support the deprecated dictionary constructor for DOMPointReadOnly and DOMPoint
https://bugs.webkit.org/show_bug.cgi?id=165732

Reviewed by Sam Weinig.
Source/WebCore:

For compatibility with other browsers, support the DOMPointInit constructor to
DOMPoint and DOMPointReadOnly per <https://www.w3.org/TR/geometry-1/#DOMPoint>

Extended geometry/DOMPoint-001.html to test.

* dom/DOMPoint.h:
* dom/DOMPoint.idl:
* dom/DOMPointReadOnly.h:
(WebCore::DOMPointReadOnly::create):
* dom/DOMPointReadOnly.idl:

LayoutTests:

* geometry/DOMPoint-001-expected.txt:
* geometry/DOMPoint-001.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209677 268f45cc-cd09-0410-ab3c-d52691b4dbfc