WebKit-https.git
10 months ago[JSC] Remove gcc warnings on mips and armv7
guijemont@igalia.com [Fri, 12 Oct 2018 02:51:45 +0000 (02:51 +0000)]
[JSC] Remove gcc warnings on mips and armv7
https://bugs.webkit.org/show_bug.cgi?id=188598

Reviewed by Mark Lam.

Source/bmalloc:

Add bitwise_cast (from WTF) and use it instead of reinterpret_cast in
a couple places where reinterpret_cast triggers a warning about
alignment even though we know that alignment is correct.

* bmalloc/Algorithm.h:
(bmalloc::bitwise_cast): Copied from WTF/wtf/StdLibextras.h
* bmalloc/IsoDirectoryPageInlines.h:
(bmalloc::IsoDirectoryPage<Config>::pageFor):
* bmalloc/IsoPageInlines.h:
(bmalloc::IsoPage<Config>::startAllocating):

Source/JavaScriptCore:

Fix many gcc/clang warnings that are false positives, mostly alignment
issues.

* assembler/MacroAssemblerPrinter.cpp:
(JSC::Printer::printMemory):
Use bitwise_cast instead of reinterpret_cast.
* assembler/testmasm.cpp:
(JSC::floatOperands):
marked as potentially unused as it is not used on all platforms.
(JSC::testProbeModifiesStackValues):
modifiedFlags is not used on mips, so don't declare it.
* bytecode/CodeBlock.h:
Make ScriptExecutable::prepareForExecution() return an
std::optional<Exception*> instead of a JSObject*.
* interpreter/Interpreter.cpp:
(JSC::Interpreter::executeProgram):
(JSC::Interpreter::executeCall):
(JSC::Interpreter::executeConstruct):
(JSC::Interpreter::prepareForRepeatCall):
(JSC::Interpreter::execute):
(JSC::Interpreter::executeModuleProgram):
Update calling code for the prototype change of
ScriptExecutable::prepareForExecution().
* jit/JITOperations.cpp: Same as for Interpreter.cpp.
* llint/LLIntSlowPaths.cpp:
(JSC::LLInt::setUpCall): Same as for Interpreter.cpp.
* runtime/JSBigInt.cpp:
(JSC::JSBigInt::dataStorage):
Use bitwise_cast instead of reinterpret_cast.
* runtime/ScriptExecutable.cpp:
* runtime/ScriptExecutable.h:
Make ScriptExecutable::prepareForExecution() return an
std::optional<Exception*> instead of a JSObject*.
* tools/JSDollarVM.cpp:
(JSC::codeBlockFromArg): Use bitwise_cast instead of reinterpret_cast.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237063 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoAdd Houdini specs to features.json
don.olmstead@sony.com [Fri, 12 Oct 2018 02:45:23 +0000 (02:45 +0000)]
Add Houdini specs to features.json
https://bugs.webkit.org/show_bug.cgi?id=190494
<rdar://problem/45211879>

Unreviewed fix.

Specifications should not reference other specifications. Also
move Worklet into the specifications section.

* features.json:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237062 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoWeb Inspector: Network: detail view reverts to "Response" when new requests are added
drousso@apple.com [Fri, 12 Oct 2018 02:30:18 +0000 (02:30 +0000)]
Web Inspector: Network: detail view reverts to "Response" when new requests are added
https://bugs.webkit.org/show_bug.cgi?id=190443

Reviewed by Joseph Pecoraro.

* UserInterface/Views/NetworkTableContentView.js:
(WI.NetworkTableContentView.prototype._showDetailView):
Return early if we are already showing a detail view for the selected object.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237061 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoIOS 12 - Service worker cache not shared when added to homescreen
youenn@apple.com [Fri, 12 Oct 2018 01:33:26 +0000 (01:33 +0000)]
IOS 12 - Service worker cache not shared when added to homescreen
https://bugs.webkit.org/show_bug.cgi?id=190269
<rdar://problem/45009961>

Reviewed by Alex Christensen.

Source/WebCore:

Expose Cache Storage API when Service Worker API is exposed.
This is used for API tests.

* dom/ScriptExecutionContext.h: Make hasServiceWorkerScheme
available outside SERVICE_WORKER compilation flag since used by DOMWindowCaches.
* dom/ScriptExecutionContext.cpp: Ditto.
* Modules/cache/DOMWindowCaches.idl:

Source/WebKit:

NetworkProcessProxy does not always have all its stores in its hash map.
Make sure to check for the default web process pool store when grabbing cache storage parameters from a given session ID.

Add a way to remove the default website data for testing.
This is used in the new API test.

* UIProcess/API/APIWebsiteDataStore.cpp:
(API::WebsiteDataStore::deleteDefaultDataStoreForTesting):
* UIProcess/API/APIWebsiteDataStore.h:
* UIProcess/API/Cocoa/WKWebsiteDataStore.mm:
(+[WKWebsiteDataStore _deleteDefaultDataStoreForTesting]):
* UIProcess/API/Cocoa/WKWebsiteDataStorePrivate.h:
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::retrieveCacheStorageParameters):

Tools:

* TestWebKitAPI/Tests/WebKitCocoa/ServiceWorkerBasic.mm:
Add test to ensure that CacheStorage use the default website data store parameters.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237060 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoUse currentStackPointer more
yusukesuzuki@slowstart.org [Fri, 12 Oct 2018 01:32:19 +0000 (01:32 +0000)]
Use currentStackPointer more
https://bugs.webkit.org/show_bug.cgi?id=190503

Reviewed by Saam Barati.

Source/JavaScriptCore:

* runtime/VM.cpp:
(JSC::VM::committedStackByteCount):

Source/WTF:

Use WTF::currentStackPointer more in WebKit to adopt ASAN detect_stack_use_after_return option.

* wtf/StackBounds.cpp:
(WTF::testStackDirection2):
(WTF::testStackDirection):
* wtf/ThreadingPthreads.cpp:
(WTF::Thread::signalHandlerSuspendResume):
(WTF::getApproximateStackPointer): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237059 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoHide RenderLayer z-order and normal flow lists behind iterators
simon.fraser@apple.com [Fri, 12 Oct 2018 00:23:08 +0000 (00:23 +0000)]
Hide RenderLayer z-order and normal flow lists behind iterators
https://bugs.webkit.org/show_bug.cgi?id=190457

Source/WebCore:

Reviewed by Zalan Bujtas.

Expose the positive z-order, negative z-order and normal flow lists
from RenderLayer as iterators rather than vectors of raw pointers.

This hides the fact that the vectors can be null, and allows for easier casting in future.

* rendering/RenderLayer.cpp:
(WebCore::RenderLayer::update3DTransformedDescendantStatus):
(WebCore::RenderLayer::paintLayerContents):
(WebCore::RenderLayer::paintList):
(WebCore::RenderLayer::hitTestLayer):
(WebCore::RenderLayer::hitTestList):
(WebCore::RenderLayer::calculateClipRects const):
* rendering/RenderLayer.h:
* rendering/RenderLayerBacking.cpp:
(WebCore::traverseVisibleNonCompositedDescendantLayers):
* rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::addToOverlapMapRecursive):
(WebCore::RenderLayerCompositor::computeCompositingRequirements):
(WebCore::RenderLayerCompositor::rebuildCompositingLayerTree):
(WebCore::RenderLayerCompositor::updateLayerTreeGeometry):
(WebCore::RenderLayerCompositor::updateCompositingDescendantGeometry):
(WebCore::RenderLayerCompositor::recursiveRepaintLayer):
(WebCore::RenderLayerCompositor::needsContentsCompositingLayer const):
(WebCore::RenderLayerCompositor::layerHas3DContent const):
* rendering/RenderTreeAsText.cpp:
(WebCore::writeLayers):

Source/WebKit:

Reviewed by Zalan Bujtas.

Expose the positive z-order, negative z-order and normal flow lists
from RenderLayer as iterators rather than vectors of raw pointers.

Use a lambda function to get access to the private constructor, while not having
to refer to the nested RenderLayer::LayerIterator class in the header.

* Shared/WebRenderLayer.cpp:
(WebKit::WebRenderLayer::WebRenderLayer):
(WebKit::WebRenderLayer::createArrayFromLayerList): Deleted.
* Shared/WebRenderLayer.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237058 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoAdd Houdini specs to features.json
don.olmstead@sony.com [Fri, 12 Oct 2018 00:08:39 +0000 (00:08 +0000)]
Add Houdini specs to features.json
https://bugs.webkit.org/show_bug.cgi?id=190494

Reviewed by Simon Fraser.

Add CSS Typed OM Level 1, CSS Layout API Level 1, CSS Animation
Worklet, and group them under a Houdini feature set.

* features.json:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237057 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoiOS: Scrolling using the arrow keys doesn't show the scroll indicator like it does...
timothy_horton@apple.com [Thu, 11 Oct 2018 23:51:08 +0000 (23:51 +0000)]
iOS: Scrolling using the arrow keys doesn't show the scroll indicator like it does on macOS
https://bugs.webkit.org/show_bug.cgi?id=190478
<rdar://problem/22194031>

Reviewed by Megan Gardner.

* Platform/spi/ios/UIKitSPI.h:
Add some SPI.

* UIProcess/ios/WKKeyboardScrollingAnimator.mm:
(-[WKKeyboardScrollViewAnimator scrollToContentOffset:animated:]):
Do a persistent flash when scrolling continuously.

(-[WKKeyboardScrollViewAnimator scrollWithScrollToExtentAnimationTo:]):
Do a single flash when doing a scroll-to-top/bottom.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237056 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoClicking on bookmarklet crashes at NavigationState::willRecordNavigationSnapshot
timothy_horton@apple.com [Thu, 11 Oct 2018 23:48:19 +0000 (23:48 +0000)]
Clicking on bookmarklet crashes at NavigationState::willRecordNavigationSnapshot
https://bugs.webkit.org/show_bug.cgi?id=190476
<rdar://problem/32757191>

Reviewed by Megan Gardner.

Source/WebKit:

* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _saveBackForwardSnapshotForItem:]):
The WKBackForwardListItem can be nil. Then we retrieve a reference from it.
An Objective-C method that returns a reference is a great way to get
a null reference. Other code then assumes it's not null and crashes.
Add a nil check.

Tools:

* TestWebKitAPI/Tests/WebKitCocoa/SnapshotStore.mm:
(TEST):
Add a test that we don't crash.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237055 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago[JSC] JSC should have "parseFunction" to optimize Function constructor
yusukesuzuki@slowstart.org [Thu, 11 Oct 2018 23:43:58 +0000 (23:43 +0000)]
[JSC] JSC should have "parseFunction" to optimize Function constructor
https://bugs.webkit.org/show_bug.cgi?id=190340

Reviewed by Mark Lam.

JSTests:

This patch fixes the line number of syntax errors raised by the Function constructor,
since we now parse the final code only once. And we no longer use block statement
for Function constructor's parsing.

* ChakraCore/test/Function/FuncBodyES5.baseline-jsc:
* stress/function-cache-with-parameters-end-position.js: Added.
(shouldBe):
(shouldThrow):
(i.anonymous):
* stress/function-constructor-name.js: Added.
(shouldBe):
(GeneratorFunction):
(AsyncFunction.async):
(AsyncGeneratorFunction.async):
(anonymous):
(async.anonymous):
* test262/expectations.yaml:

LayoutTests/imported/w3c:

* web-platform-tests/html/webappapis/scripting/events/inline-event-handler-ordering-expected.txt:
* web-platform-tests/html/webappapis/scripting/events/invalid-uncompiled-raw-handler-compiled-late-expected.txt:
* web-platform-tests/html/webappapis/scripting/processing-model-2/compile-error-in-attribute-expected.txt:
* web-platform-tests/html/webappapis/scripting/processing-model-2/compile-error-in-body-onerror-expected.txt:

Source/JavaScriptCore:

The current Function constructor is suboptimal. We parse the piece of the same code three times to meet
the spec requirement. (1) check parameters syntax, (2) check body syntax, and (3) parse the entire function.
And to parse 1-3 correctly, we create two strings, the parameters and the entire function. This operation
is really costly and ideally we should meet the above requirement by the one time parsing.

To meet the above requirement, we add a special function for Parser, parseSingleFunction. This function
takes `std::optional<int> functionConstructorParametersEndPosition` and check this end position is correct in the parser.
For example, if we run the code,

    Function('/*', '*/){')

According to the spec, this should produce '/*' parameter string and '*/){' body string. And parameter
string should be syntax-checked by the parser, and raise the error since it is incorrect. Instead of doing
that, in our implementation, we first create the entire string.

    function anonymous(/*) {
        */){
    }

And we parse it. At that time, we also pass the end position of the parameters to the parser. In the above case,
the position of the `function anonymous(/*)' <> is passed. And in the parser, we check that the last token
offset of the parameters is the given end position. This check allows us to raise the error correctly to the
above example while we parse the entire function only once. And we do not need to create two strings too.

This improves the performance of the Function constructor significantly. And web-tooling-benchmark/uglify-js is
significantly sped up (28.2%).

Before:
    uglify-js:  2.94 runs/s
After:
    uglify-js:  3.77 runs/s

* bytecode/UnlinkedFunctionExecutable.cpp:
(JSC::UnlinkedFunctionExecutable::fromGlobalCode):
* bytecode/UnlinkedFunctionExecutable.h:
* parser/Parser.cpp:
(JSC::Parser<LexerType>::parseInner):
(JSC::Parser<LexerType>::parseSingleFunction):
(JSC::Parser<LexerType>::parseFunctionInfo):
(JSC::Parser<LexerType>::parseFunctionDeclaration):
(JSC::Parser<LexerType>::parseAsyncFunctionDeclaration):
(JSC::Parser<LexerType>::parseClass):
(JSC::Parser<LexerType>::parsePropertyMethod):
(JSC::Parser<LexerType>::parseGetterSetter):
(JSC::Parser<LexerType>::parseFunctionExpression):
(JSC::Parser<LexerType>::parseAsyncFunctionExpression):
(JSC::Parser<LexerType>::parseArrowFunctionExpression):
* parser/Parser.h:
(JSC::Parser<LexerType>::parse):
(JSC::parse):
(JSC::parseFunctionForFunctionConstructor):
* parser/ParserModes.h:
* parser/ParserTokens.h:
(JSC::JSTextPosition::JSTextPosition):
(JSC::JSTokenLocation::JSTokenLocation): Deleted.
* parser/SourceCodeKey.h:
(JSC::SourceCodeKey::SourceCodeKey):
(JSC::SourceCodeKey::operator== const):
* runtime/CodeCache.cpp:
(JSC::CodeCache::getUnlinkedGlobalCodeBlock):
(JSC::CodeCache::getUnlinkedGlobalFunctionExecutable):
* runtime/CodeCache.h:
* runtime/FunctionConstructor.cpp:
(JSC::constructFunctionSkippingEvalEnabledCheck):
* runtime/FunctionExecutable.cpp:
(JSC::FunctionExecutable::fromGlobalCode):
* runtime/FunctionExecutable.h:

LayoutTests:

* fast/dom/attribute-event-listener-errors-expected.txt:
* fast/events/attribute-listener-deletion-crash-expected.txt:
* fast/events/window-onerror-syntax-error-in-attr-expected.txt:
* js/dom/invalid-syntax-for-function-expected.txt:
* js/dom/script-start-end-locations-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237054 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoHardware keyboard arrow keys/spacebar don't scroll PDFs (works for web content)
timothy_horton@apple.com [Thu, 11 Oct 2018 23:41:18 +0000 (23:41 +0000)]
Hardware keyboard arrow keys/spacebar don't scroll PDFs (works for web content)
https://bugs.webkit.org/show_bug.cgi?id=190495
<rdar://problem/22734616>

Reviewed by Andy Estes.

Hook up WKKeyboardScrollViewAnimator to the scroll view when
WKPDFView is installed.

* Platform/spi/ios/UIKitSPI.h:
Move _inputFlags to the IPI section so we can remove it
from WKContentViewInteraction.

* SourcesCocoa.txt:
* UIProcess/ios/WKWebEvent.h:
* UIProcess/ios/WKWebEvent.mm:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView handleKeyEvent:]):
(-[WKWebEvent dealloc]): Deleted.
* WebKit.xcodeproj/project.pbxproj:
Move WKWebEvent out of WKContentViewInteraction.
We tend to prefer one file per class.
Also, move the code to make a WKWebEvent from a UIEvent into
-initWithEvent:, instead of being ad-hoc in WKContentViewInteraction.
Adopt RetainPtr for the WKWebEvent's uiEvent property.

* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _handleKeyUIEvent:]):
* UIProcess/Cocoa/WKWebViewContentProvider.h:
Plumb hardware keyboard events to custom content views, if they want them.

* UIProcess/ios/WKPDFView.mm:
(-[WKPDFView dealloc]):
(-[WKPDFView web_handleKeyEvent:]):
(-[WKPDFView web_initWithFrame:webView:mimeType:]):
Install a WKKeyboardScrollViewAnimator and plumb key events to it.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237053 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago[WTF] Semaphore.h conflicts with POSIX header
ross.kirsling@sony.com [Thu, 11 Oct 2018 22:52:43 +0000 (22:52 +0000)]
[WTF] Semaphore.h conflicts with POSIX header
https://bugs.webkit.org/show_bug.cgi?id=190486

Reviewed by Yusuke Suzuki.

Source/WebCore:

* platform/graphics/avfoundation/objc/SourceBufferPrivateAVFObjC.mm:

Source/WTF:

Rename Semaphore.h to WTFSemaphore.h to avoid conflict with POSIX semaphore.h on case-insensitive file systems.

* WTF.xcodeproj/project.pbxproj:
* wtf/CMakeLists.txt:
* wtf/WTFSemaphore.h: Renamed from wtf/Semaphore.h.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237052 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoFix non-existent define `CPU(JSVALUE64)`
ross.kirsling@sony.com [Thu, 11 Oct 2018 22:06:42 +0000 (22:06 +0000)]
Fix non-existent define `CPU(JSVALUE64)`
https://bugs.webkit.org/show_bug.cgi?id=190479

Reviewed by Yusuke Suzuki.

* jit/CCallHelpers.h:
(JSC::CCallHelpers::setupArgumentsImpl):
Correct CPU(JSVALUE64) to USE(JSVALUE64).

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237051 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoWeb Inspector: remove unused TreeOutline style .force-focus
mattbaker@apple.com [Thu, 11 Oct 2018 21:46:43 +0000 (21:46 +0000)]
Web Inspector: remove unused  TreeOutline style .force-focus
https://bugs.webkit.org/show_bug.cgi?id=190480
<rdar://problem/45203484>

Reviewed by Joseph Pecoraro.

* UserInterface/Views/CallFrameTreeElement.css:
(.tree-outline:focus .item.call-frame.selected .status > .status-image):
(.tree-outline:matches(:focus, .force-focus) .item.call-frame.selected .status > .status-image): Deleted.

* UserInterface/Views/CanvasSidebarPanel.css:
(.sidebar > .panel.navigation.canvas > .content > .recording-content > .tree-outline:focus .item.processing.selected .subtitle > progress):
(.sidebar > .panel.navigation.canvas > .content > .recording-content > .tree-outline:matches(:focus, .force-focus) .item.processing.selected .subtitle > progress): Deleted.

* UserInterface/Views/DataGrid.css:
(body[dir=ltr] .data-grid:focus tr.selected td:not(:last-child)):
(body[dir=rtl] .data-grid:focus tr.selected td:not(:last-child)):
(.data-grid:focus tr.parent.selected td.disclosure::before):
(.data-grid:focus tr.parent.expanded.selected td.disclosure::before):
(.data-grid:focus tr.selected):
(.data-grid:focus tr.selected td .subtitle):
(body:not(.window-inactive, .window-docked-inactive) .data-grid:focus tr.editable.selected .cell-content > input):
(@media (prefers-dark-interface)):
(body[dir=ltr] .data-grid:matches(:focus, .force-focus) tr.selected td:not(:last-child)): Deleted.
(body[dir=rtl] .data-grid:matches(:focus, .force-focus) tr.selected td:not(:last-child)): Deleted.
(.data-grid:matches(:focus, .force-focus) tr.parent.selected td.disclosure::before): Deleted.
(.data-grid:matches(:focus, .force-focus) tr.parent.expanded.selected td.disclosure::before): Deleted.
(.data-grid:matches(:focus, .force-focus) tr.selected): Deleted.
(.data-grid:matches(:focus, .force-focus) tr.selected td .subtitle): Deleted.
(body:not(.window-inactive, .window-docked-inactive) .data-grid:matches(:focus, .force-focus) tr.editable.selected .cell-content > input): Deleted.

* UserInterface/Views/Main.css:
(:focus .selected .go-to-arrow):
(:focus .selected .go-to-arrow:active):
(:matches(:focus, .force-focus) .selected .go-to-arrow): Deleted.
(:matches(:focus, .force-focus) .selected .go-to-arrow:active): Deleted.

* UserInterface/Views/ProfileView.css:
(.profile > .data-grid:focus tr.selected td .location):
(.profile > .data-grid:matches(:focus, .force-focus) tr.selected td .location): Deleted.

* UserInterface/Views/RecordingActionTreeElement.css:
(.tree-outline:focus .item.action.selected:not(.initial-state, .invalid) > .icon):
(body:not(.window-inactive, .window-docked-inactive) .tree-outline:focus .item.action.selected > .titles .parameter.swizzled,):
(.tree-outline:matches(:focus, .force-focus) .item.action.selected:not(.initial-state, .invalid) > .icon): Deleted.
(body:not(.window-inactive, .window-docked-inactive) .tree-outline:matches(:focus, .force-focus) .item.action.selected > .titles .parameter.swizzled,): Deleted.

* UserInterface/Views/ScriptDetailsTimelineView.css:
(.tree-outline:focus .item.selected .alternate-subtitle):
(.tree-outline:matches(:focus, .force-focus) .item.selected .alternate-subtitle): Deleted.

* UserInterface/Views/ShaderProgramTreeElement.css:
(.tree-outline:focus .item.shader-program.selected .status > img):
(.tree-outline:matches(:focus, .force-focus) .item.shader-program.selected .status > img): Deleted.

* UserInterface/Views/ThreadTreeElement.css:
(.tree-outline:focus > .item.thread.selected .status-button.resume):
(.tree-outline:matches(:focus, .force-focus) > .item.thread.selected .status-button.resume): Deleted.

* UserInterface/Views/TimelineRecordBar.css:
(:focus .selected .timeline-record-bar > .segment):
(:focus .selected .timeline-record-bar > .segment.inactive):
(body[dir=ltr] :focus .selected .timeline-record-bar.has-inactive-segment > .segment:not(.inactive)):
(body[dir=rtl] :focus .selected .timeline-record-bar.has-inactive-segment > .segment:not(.inactive)):
(:matches(:focus, .force-focus) .selected .timeline-record-bar > .segment): Deleted.
(:matches(:focus, .force-focus) .selected .timeline-record-bar > .segment.inactive): Deleted.
(body[dir=ltr] :matches(:focus, .force-focus) .selected .timeline-record-bar.has-inactive-segment > .segment:not(.inactive)): Deleted.
(body[dir=rtl] :matches(:focus, .force-focus) .selected .timeline-record-bar.has-inactive-segment > .segment:not(.inactive)): Deleted.

* UserInterface/Views/TreeElementStatusButton.css:
(:focus .item.selected > .status > .status-button):
(:matches(:focus, .force-focus) .item.selected > .status > .status-button): Deleted.

* UserInterface/Views/TreeOutline.css:
(.tree-outline:focus .item.selected .disclosure-button):
(.tree-outline:focus .item.selected.expanded .disclosure-button):
(.tree-outline:focus .item.selected):
(.tree-outline:focus .item.selected .subtitle):
(.tree-outline:not(.large):focus .item.selected .status .indeterminate-progress-spinner):
(.tree-outline:matches(:focus, .force-focus) .item.selected .disclosure-button): Deleted.
(.tree-outline:matches(:focus, .force-focus) .item.selected.expanded .disclosure-button): Deleted.
(.tree-outline:matches(:focus, .force-focus) .item.selected): Deleted.
(.tree-outline:matches(:focus, .force-focus) .item.selected .subtitle): Deleted.
(.tree-outline:not(.large):matches(:focus, .force-focus) .item.selected .status .indeterminate-progress-spinner): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237050 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago[Web Animations] Allow iOS to also control toggling Web Animations CSS Integration
graouts@webkit.org [Thu, 11 Oct 2018 21:43:22 +0000 (21:43 +0000)]
[Web Animations] Allow iOS to also control toggling Web Animations CSS Integration
https://bugs.webkit.org/show_bug.cgi?id=190489

Reviewed by Tim Horton.

* UIProcess/API/Cocoa/WKPreferences.mm:
(-[WKPreferences _setWebAnimationsCSSIntegrationEnabled:]):
(-[WKPreferences _webAnimationsCSSIntegrationEnabled]):
* UIProcess/API/Cocoa/WKPreferencesPrivate.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237049 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago[GStreamer] Support arbitrary video resolution in getUserMedia API
commit-queue@webkit.org [Thu, 11 Oct 2018 21:30:33 +0000 (21:30 +0000)]
[GStreamer] Support arbitrary video resolution in getUserMedia API
https://bugs.webkit.org/show_bug.cgi?id=189734

Source/WebCore:

Patch by Thibault Saunier <tsaunier@igalia.com> on 2018-10-11
Reviewed by Xabier Rodriguez-Calvar.

Implement arbitrary video resolution for the getUserMedia API in GStreamer.

Fix the MockRealtimeMediaSource device list to make devices properties match
test expectations.

Reactivate tests that were failling because of that.

* platform/mediastream/RealtimeVideoSource.cpp:
(WebCore::RealtimeVideoSource::standardVideoSizes):
(WebCore::standardVideoSizes): Deleted.
* platform/mediastream/RealtimeVideoSource.h:
* platform/mediastream/VideoPreset.h:
* platform/mediastream/gstreamer/GStreamerVideoCaptureSource.cpp:
(WebCore::GStreamerVideoPreset::create):
(WebCore::GStreamerVideoPreset::GStreamerVideoPreset):
(WebCore::GStreamerVideoCaptureSource::GStreamerVideoCaptureSource):
(WebCore::GStreamerVideoCaptureSource::capabilities):
(WebCore::GStreamerVideoCaptureSource::generatePresets):
* platform/mediastream/gstreamer/GStreamerVideoCaptureSource.h:
* platform/mock/MockRealtimeMediaSourceCenter.cpp:
(WebCore::defaultDevices):

LayoutTests:

Implement arbitrary video resolution for the getUserMedia API in GStreamer.

Fix the MockRealtimeMediaSource device list to make devices properties match
test expectations.

Reactivate tests that were failling because of that.

Patch by Thibault Saunier <tsaunier@igalia.com> on 2018-10-11
Reviewed by Xabier Rodriguez-Calvar.

* platform/gtk/TestExpectations:
* platform/wpe/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237048 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoCURRENT_ARCH should not be used in Run Script phase.
krollin@apple.com [Thu, 11 Oct 2018 21:22:40 +0000 (21:22 +0000)]
CURRENT_ARCH should not be used in Run Script phase.
https://bugs.webkit.org/show_bug.cgi?id=190407
<rdar://problem/45133556>

Reviewed by Alexey Proskuryakov.

CURRENT_ARCH is used in a number of Xcode Run Script phases. However,
CURRENT_ARCH is not well-defined during this phase (and may even have
the value "undefined") since this phase is run just once per build
rather than once per supported architecture. Migrate away from
CURRENT_ARCH in favor of ARCHS, either by iterating over ARCHS and
performing an operation for each value, or by picking the first entry
in ARCHS and using that as a representative value.

Source/JavaScriptCore:

* JavaScriptCore.xcodeproj/project.pbxproj: Store
LLIntDesiredOffsets.h into a directory with a name based on ARCHS
rather than CURRENT_ARCH.

Source/WebCore:

No new tests -- no functional changes.

* DerivedSources.make: When forming TARGET_TRIPLE_FLAGS, grab the
first entry in ARCHS rather than use CURRENT_ARCH.

Source/WebKit:

* DerivedSources.make: When forming TARGET_TRIPLE_FLAGS, grab the
first entry in ARCHS rather than use CURRENT_ARCH.

Source/WebKitLegacy:

* WebKitLegacy.xcodeproj/project.pbxproj: When generating
WebKitLegacy.*.exp, generate both 32- and 64-bit versions for
macosx platforms.

Source/WebKitLegacy/mac:

* Configurations/WebKitLegacy.xcconfig: When generating
WebKitLegacy.*.exp, generate both 32- and 64-bit versions for
macosx platforms.

* MigrateHeaders.make:
* migrate-headers.sh: When generating ReexportedWebCoreSymbols_*.exp,
generate one for each architecture in ARCHS.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237047 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoSupport building WebKit for macOS Mojave using a newer SDK
dbates@webkit.org [Thu, 11 Oct 2018 20:30:54 +0000 (20:30 +0000)]
Support building WebKit for macOS Mojave using a newer SDK
https://bugs.webkit.org/show_bug.cgi?id=190431

Reviewed by Andy Estes.

* platform/network/cocoa/CookieCocoa.mm:
(WebCore::coreSameSitePolicy):
(WebCore::nsSameSitePolicy):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237046 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago[iOS] Add typedef for WebEvent keyboard flags
dbates@webkit.org [Thu, 11 Oct 2018 20:29:57 +0000 (20:29 +0000)]
[iOS] Add typedef for WebEvent keyboard flags
https://bugs.webkit.org/show_bug.cgi?id=190435

Reviewed by Wenson Hsieh.

Currently WebEvent and UIKit SPI are intertwined when it comes to keyboard flags. It seems sufficient
to have WebEvent defined its own keyboard flags so that callers do not need to be aware of UIKit SPI.

No functionality changed. So, no new tests.

* platform/ios/WebEvent.h:
* platform/ios/WebEvent.mm:
(-[WebEvent keyboardFlags]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237045 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoREGRESSION (r237037): Cannot use webkit-patch upload, always fails with unknown ...
timothy_horton@apple.com [Thu, 11 Oct 2018 19:32:56 +0000 (19:32 +0000)]
REGRESSION (r237037): Cannot use webkit-patch upload, always fails with unknown 'cc_radar'

* Scripts/webkitpy/tool/steps/options.py:
(Options):
Fix the name of the variable that the --cc-radar option writes into.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237044 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoWeb Inspector: detail view is not re-shown after sorting the Network table
drousso@apple.com [Thu, 11 Oct 2018 19:26:27 +0000 (19:26 +0000)]
Web Inspector: detail view is not re-shown after sorting the Network table
https://bugs.webkit.org/show_bug.cgi?id=190330
<rdar://problem/45089607>

Reviewed by Matt Baker.

* UserInterface/Views/NetworkTableContentView.js:
(WI.NetworkTableContentView.prototype._restoreSelectedRow):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237043 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoChanges towards allowing use of the ASAN detect_stack_use_after_return option.
mark.lam@apple.com [Thu, 11 Oct 2018 19:19:18 +0000 (19:19 +0000)]
Changes towards allowing use of the ASAN detect_stack_use_after_return option.
https://bugs.webkit.org/show_bug.cgi?id=190405
<rdar://problem/45131464>

Reviewed by Michael Saboff.

Source/JavaScriptCore:

The ASAN detect_stack_use_after_return option checks for use of stack variables
after they have been freed.  It does this by allocating relevant stack variables
in heap memory (instead of on the stack) if the code ever takes the address of
those stack variables.  Unfortunately, this is a common idiom that we use to
compute the approximate stack pointer value.  As a result, on such ASAN runs, the
computed approximate stack pointer value will point into the heap instead of the
stack.  This breaks the VM's expectations and wreaks havoc.

To fix this, we use the newly introduced WTF::currentStackPointer() instead of
taking the address of stack variables.

We also need to enhance ExceptionScopes to be able to work with ASAN
detect_stack_use_after_return which will allocated the scope in the heap.  We
work around this by passing the current stack pointer of the instantiating calling
frame into the scope constructor, and using that for the position check in
~ThrowScope() instead.

The above is only a start towards enabling ASAN detect_stack_use_after_return on
the VM.  There are still other issues to be resolved before we can run with this
ASAN option.

* runtime/CatchScope.h:
* runtime/ExceptionEventLocation.h:
(JSC::ExceptionEventLocation::ExceptionEventLocation):
* runtime/ExceptionScope.h:
(JSC::ExceptionScope::stackPosition const):
* runtime/JSLock.cpp:
(JSC::JSLock::didAcquireLock):
* runtime/ThrowScope.cpp:
(JSC::ThrowScope::~ThrowScope):
* runtime/ThrowScope.h:
* runtime/VM.h:
(JSC::VM::needExceptionCheck const):
(JSC::VM::isSafeToRecurse const):
* wasm/js/WebAssemblyFunction.cpp:
(JSC::callWebAssemblyFunction):
* yarr/YarrPattern.cpp:
(JSC::Yarr::YarrPatternConstructor::isSafeToRecurse const):

Source/WTF:

Introduce WTF::currentStackPointer() which computes its caller's stack pointer value.

* WTF.xcodeproj/project.pbxproj:
* wtf/CMakeLists.txt:
* wtf/StackBounds.h:
(WTF::StackBounds::checkConsistency const):
* wtf/StackPointer.cpp: Added.
(WTF::currentStackPointer):
* wtf/StackPointer.h: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237042 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoUse finer grained locking in FontDatabase
antti@apple.com [Thu, 11 Oct 2018 18:39:05 +0000 (18:39 +0000)]
Use finer grained locking in FontDatabase
https://bugs.webkit.org/show_bug.cgi?id=190467

Reviewed by Alex Christensen.

* platform/graphics/FontCache.h:

Also use ListHashSet for prewarming info so we can prewarm in the same order the fonts were
seen last time.

* platform/graphics/cocoa/FontCacheCoreText.cpp:
(WebCore::FontDatabase::collectionForFamily):

Only hold the lock when accessing the hashmap. There is no need to hold it during font construction
which can take a long time.

(WebCore::FontDatabase::fontForPostScriptName):

This is currently not prewarmed from a thread so no need for locking.

(WebCore::FontDatabase::clear):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237040 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoAdd --cc-radar option to webkit-patch bug creation
dino@apple.com [Thu, 11 Oct 2018 17:32:31 +0000 (17:32 +0000)]
Add --cc-radar option to webkit-patch bug creation
https://bugs.webkit.org/show_bug.cgi?id=190451
<rdar://problem/45176671>

Reviewed by Antoine Quint.

Add a new command line option to automatically
CC radar to new or updated bugs.

* Scripts/webkitpy/tool/commands/commandtest.py:
(CommandsTest.assert_execute_outputs):
* Scripts/webkitpy/tool/commands/upload.py:
(CreateBug.__init__):
(CreateBug.execute):
* Scripts/webkitpy/tool/commands/upload_unittest.py:
(test_prepare_with_cc):
(test_prepare_with_radar):
(test_prepare_with_cc_and_radar):
* Scripts/webkitpy/tool/steps/createbug.py:
(CreateBug.options):
(CreateBug.run):
* Scripts/webkitpy/tool/steps/options.py:
(Options):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237037 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago[GStreamer] Fix race condition in GStreamerVideoDecoder
commit-queue@webkit.org [Thu, 11 Oct 2018 17:29:30 +0000 (17:29 +0000)]
[GStreamer] Fix race condition in GStreamerVideoDecoder
https://bugs.webkit.org/show_bug.cgi?id=190470

The GStreamerVideoDecoder.m_dtsPtsMap filed is accessed from
the main thread and some GStreamer streaming thread, make sure
to protect its access.

And use WTF::StdMap instead of std::map.

Patch by Thibault Saunier <tsaunier@igalia.com> on 2018-10-11
Reviewed by Philippe Normand.

Manually tested and a random crash is gone.

* platform/mediastream/libwebrtc/GStreamerVideoDecoderFactory.cpp:
(WebCore::GStreamerVideoDecoder::newSampleCallback):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237036 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoRegression(PSON): Deal with the drawing area potentially being null under WebChromeCl...
cdumez@apple.com [Thu, 11 Oct 2018 17:25:18 +0000 (17:25 +0000)]
Regression(PSON): Deal with the drawing area potentially being null under WebChromeClient::contentsSizeChanged()
https://bugs.webkit.org/show_bug.cgi?id=190474

Reviewed by Geoffrey Garen.

* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::contentsSizeChanged const):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237035 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoLayout Test fast/forms/fieldset/fieldset-elements-htmlcollection.html is failing
pvollan@apple.com [Thu, 11 Oct 2018 15:54:59 +0000 (15:54 +0000)]
Layout Test fast/forms/fieldset/fieldset-elements-htmlcollection.html is failing
https://bugs.webkit.org/show_bug.cgi?id=190472

Unreviewed test gardening.

* platform/win/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237034 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago[GStreamer][MSE] Fix height calculation for streams with source aspect ratio
eocanha@igalia.com [Thu, 11 Oct 2018 11:10:14 +0000 (11:10 +0000)]
[GStreamer][MSE] Fix height calculation for streams with source aspect ratio
https://bugs.webkit.org/show_bug.cgi?id=190464

Reviewed by Xabier Rodriguez-Calvar.

This patch is authored by Nikola Veljkovic <Nikola.Veljkovic@zenterio.com>

* platform/graphics/gstreamer/GStreamerCommon.cpp:
(WebCore::getVideoResolutionFromCaps): Reverse the SAR adjustment to "undo" it
instead of applying it twice.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237033 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago[GStreamer] Do not build gstgtk plugins
csaavedra@igalia.com [Thu, 11 Oct 2018 10:41:43 +0000 (10:41 +0000)]
[GStreamer] Do not build gstgtk plugins
https://bugs.webkit.org/show_bug.cgi?id=190463

Reviewed by Xabier Rodriguez-Calvar.

These plugins pull a GTK+ dependency that, at least in WPE,
is unnecessary. The GTK+ port doesn't need it either, as far
as I can tell.

* gstreamer/jhbuild.modules: Disable gstgtk plugins

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237032 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago[GTK][WPE] Add mediaDevices.enumerateDevices support
commit-queue@webkit.org [Thu, 11 Oct 2018 10:26:52 +0000 (10:26 +0000)]
[GTK][WPE] Add mediaDevices.enumerateDevices support
https://bugs.webkit.org/show_bug.cgi?id=185761

Patch by Alejandro G. Castro <alex@igalia.com> on 2018-10-11
Reviewed by Youenn Fablet.

Source/WebCore:

We are adopting the same policy COCOA is using when returning the
list of media devices if the user does not have persistent
access. Namely, we just return the first media device for audio
and video capture.

* Modules/mediastream/MediaDevicesRequest.cpp:
(WebCore::MediaDevicesRequest::filterDeviceList): Add support for
other platforms when filtering devices if there is no persistent
access to the origin.

Source/WebKit:

Implemented the enumerateDevices API using a new WebKit class
(DeviceIdHashSaltStorage) that generates and stores the device ids
hash salts per origin, used to avoid fingerprinting in the
webpages.

The patch also adds a new type of data for the WebsiteDataStore.
That way the users can call the WebsiteDataStore to show what
origins have hash salt generated and remove them at some point.

For the moment just GTK+ and WPE ports are using this class to
generate the hash salts. The patch adds code to the
checkUserMediaPermissionForOrigin API implementation, it was empty
until now for these ports. In this function we create an instance
of a new class WebKitDeviceInfoPermissionRequest that implements
the WebKitPermissionRequestIface interface, that allows the ports
to use the current permission managers implemented in the
embedders to handle this new kind of request the way they like
it. The default implementation is deny.

The class WebKitDeviceInfoPermissionRequest takes care of
contacting the DeviceIdHashSaltStorage and request/regenerate the
hash salts accordingly.

Persistency is still not implemented, we will add it in a
future patch.

* Shared/WebsiteData/WebsiteDataType.h: Added the new type of
webside data: DeviceIdHashSalt.
* Sources.txt: Added new files to compilation.
* SourcesGTK.txt: Ditto.
* SourcesWPE.txt: Ditto.
* UIProcess/API/glib/WebKitDeviceInfoPermissionRequest.cpp: Added
this class to represent a request of permission to access the
devices information. This is specific for glib ports and implements
the policies regarding the hash salts when the embedder allows or
denies the access using the DeviceIdHashSaltStorage class.
(webkitDeviceInfoPermissionRequestAllow): Get the device hash salt
when the origin is allowed and set it in the original request.
(webkitDeviceInfoPermissionRequestDeny): Regenerate the device id
hash salt when the user does not allow to access the device information.
(webkit_permission_request_interface_init):
(webkitDeviceInfoPermissionRequestDispose):
(webkit_device_info_permission_request_class_init):
(webkitDeviceInfoPermissionRequestCreate): Create the class using
the proxy request from the webprocess and a reference to the
DeviceIdHashSaltStorage.
* UIProcess/API/glib/WebKitDeviceInfoPermissionRequestPrivate.h:
Ditto.
* UIProcess/API/glib/WebKitUIClient.cpp: Added the implementation
for the checkUserMediaPermissionForOrigin API for the glib ports,
it creates the device info request and calls the permission
request API.
* UIProcess/API/glib/WebKitWebsiteData.cpp:
(recordContainsSupportedDataTypes): Added the DeviceIdHashSalt
type.
(toWebKitWebsiteDataTypes): Added the conversion from the WebKit
types to the glib types for the WebsiteDataType::DeviceIdHashSalt.
* UIProcess/API/glib/WebKitWebsiteDataManager.cpp:
(toWebsiteDataTypes): Added the conversion from the glib type
WEBKIT_WEBSITE_DATA_DEVICE_ID_HASH_SALT to the WebKit type.
(webkit_website_data_manager_remove): Make sure we remote the
DeviceIdHashSalt if the Cookies are selected.
* UIProcess/API/gtk/WebKitDeviceInfoPermissionRequest.h: Added
this new class that represents a permission request for the device
information in the GTK+ port.
* UIProcess/API/gtk/WebKitWebsiteData.h: Added the new type of
website that in the GTK+ port:
WEBKIT_WEBSITE_DATA_DEVICE_ID_HASH_SALT.
* UIProcess/API/wpe/WebKitDeviceInfoPermissionRequest.h: Added
this new class that represents a permission request for the device
information in the WPE port.
* UIProcess/API/wpe/WebKitWebsiteData.h: Added the new type of
website that in the WPE port:
WEBKIT_WEBSITE_DATA_DEVICE_ID_HASH_SALT.
* UIProcess/DeviceIdHashSaltStorage.cpp: Added this new class that
handles how to generate and store the hash salts inside
WebKit. Persistency is still not implemented, we will add it in a
future patch.
(WebKit::DeviceIdHashSaltStorage::create):
(WebKit::DeviceIdHashSaltStorage::deviceIdentifierHashSaltForOrigin):
Check the map to get the hash salt for an origin, if there is none
create a new random one.
(WebKit::DeviceIdHashSaltStorage::regenerateDeviceIdentifierHashSaltForOrigin):
Delete the hash salt of an origin if it already exists and create a new one.
(WebKit::DeviceIdHashSaltStorage::getDeviceIdHashSaltOrigins):
Returns the list of origins that have a hash salt generated for
them. It is used in the WebsiteDataStore.
(WebKit::DeviceIdHashSaltStorage::deleteDeviceIdHashSaltForOrigins):
Deletes the hash salts in the map for an origin.
(WebKit::DeviceIdHashSaltStorage::deleteDeviceIdHashSaltOriginsModifiedSince):
Delete the origins that have not been modified since a time.
* UIProcess/DeviceIdHashSaltStorage.h: Ditto.
* UIProcess/UserMediaPermissionCheckProxy.h: Remove uneeded class
definition.
* UIProcess/WebsiteData/WebsiteDataStore.cpp: Added support to
return and remove the origins with a generate hash salt used to
generate the device ids.
(WebKit::WebsiteDataStore::WebsiteDataStore): Added the new
reference to the DeviceIdHashSaltStorage class, used to handle the
hash salts.
(WebKit::WebsiteDataStore::fetchDataAndApply): Get the list of
origins with a hash salts in the DeviceIdHashSaltStorage.
(WebKit::WebsiteDataStore::removeData): Remove the hash salts in
DeviceIdHashSaltStorage for an origin.
* UIProcess/WebsiteData/WebsiteDataStore.h: Ditto.
(WebKit::WebsiteDataStore::deviceIdHashSaltStorage):
* WebKit.xcodeproj/project.pbxproj: Added the
DeviceIdHashSaltStorage to the compilation.

Tools:

Added new API test for the mediaDevices.enumerateDevices:
usermedia-enumeratedevices-permission-check. And a new API test
for the WebsiteDataStore: testWebsiteDataDeviceIdHashSalt.

* MiniBrowser/gtk/main.c:
(gotWebsiteDataCallback): Added a new section to the about:data
webpage to include the information about the hash salt.
* TestWebKitAPI/Tests/WebKitGLib/TestUIClient.cpp:
(testWebViewUserMediaEnumerateDevicesPermissionCheck): Added the
new test checking the enumerateDevices API when permission is
denied and when permission is allowed for the origin.
(beforeAll): Defined the new test.
* TestWebKitAPI/Tests/WebKitGLib/TestWebsiteData.cpp:
(serverCallback): Register a new URI for the enumerateDevices.
(testWebsiteDataConfiguration): Remove the hash salts from the
directories.
(testWebsiteDataDeviceIdHashSalt): New test cheking the
enumerateDevices API.
(beforeAll): Added the new test testWebsiteDataDeviceIdHashSalt to
the init structure.
* TestWebKitAPI/glib/WebKitGLib/WebViewTest.cpp:
(WebViewTest::initializeWebView): Make sure the media-stream is
activated in the webView.

LayoutTests:

Remove the failure expectation for the test
media-devices-enumerate-devices.html. Just for the record we have
also checked the following tests.

fast/mediastream/media-devices-enumerate-devices.html
fast/mediastream/resources/enumerate-devices-frame.html
http/tests/media/media-stream/enumerate-devices-source-id.html
http/tests/media/media-stream/enumerate-devices-source-id-persistent.html
http/tests/media/media-stream/resources/enumerate-devices-source-id-frame.html

* platform/gtk/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237031 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoRename a couple of DOMWindowProperty virtual functions
cdumez@apple.com [Thu, 11 Oct 2018 04:27:08 +0000 (04:27 +0000)]
Rename a couple of DOMWindowProperty virtual functions
https://bugs.webkit.org/show_bug.cgi?id=190458

Reviewed by Geoffrey Garen.

Rename a couple of DOMWindowProperty virtual functions as the current naming no longer makes
sense after recent refactoring.

disconnectFrameForDocumentSuspension() was renamed to suspendForPageCache(), and
reconnectFrameFromDocumentSuspension(Frame*) was renamed to resumeFromPageCache().
DOMWindowProperty objects no longer need to disconnect / reconnect from their
frame since they now get their frame from their associated Window. However, some
DOMWindowProperty subclasses do have some page cache suspension / resuming logic
implemented of overrides of these functions.

Also drop the disconnectDOMWindowProperties() / reconnectDOMWindowProperties()
methods in DOMWindow. The naming made little sense as it does not really
disconnect those properties in any way. Instead, inline them in
DOMWindow's suspendForPageCache() / resumeFromPageCache() since these are
the only callers.

* Modules/indexeddb/DOMWindowIndexedDatabase.cpp:
(WebCore::DOMWindowIndexedDatabase::suspendForPageCache):
(WebCore::DOMWindowIndexedDatabase::resumeFromPageCache):
(WebCore::DOMWindowIndexedDatabase::disconnectFrameForDocumentSuspension): Deleted.
(WebCore::DOMWindowIndexedDatabase::reconnectFrameFromDocumentSuspension): Deleted.
* Modules/indexeddb/DOMWindowIndexedDatabase.h:
* history/CachedFrame.cpp:
(WebCore::CachedFrame::CachedFrame):
* loader/FrameLoader.cpp:
(WebCore::FrameLoader::open):
* loader/appcache/DOMApplicationCache.cpp:
(WebCore::DOMApplicationCache::suspendForPageCache):
(WebCore::DOMApplicationCache::resumeFromPageCache):
(WebCore::DOMApplicationCache::disconnectFrameForDocumentSuspension): Deleted.
(WebCore::DOMApplicationCache::reconnectFrameFromDocumentSuspension): Deleted.
* loader/appcache/DOMApplicationCache.h:
* page/DOMWindow.cpp:
(WebCore::DOMWindow::suspendForPageCache):
(WebCore::DOMWindow::resumeFromPageCache):
(WebCore::DOMWindow::suspendForDocumentSuspension): Deleted.
(WebCore::DOMWindow::resumeFromDocumentSuspension): Deleted.
(WebCore::DOMWindow::disconnectDOMWindowProperties): Deleted.
(WebCore::DOMWindow::reconnectDOMWindowProperties): Deleted.
* page/DOMWindow.h:
* page/DOMWindowExtension.cpp:
(WebCore::DOMWindowExtension::suspendForPageCache):
(WebCore::DOMWindowExtension::resumeFromPageCache):
(WebCore::DOMWindowExtension::disconnectFrameForDocumentSuspension): Deleted.
(WebCore::DOMWindowExtension::reconnectFrameFromDocumentSuspension): Deleted.
* page/DOMWindowExtension.h:
* page/DOMWindowProperty.cpp:
(WebCore::DOMWindowProperty::suspendForPageCache):
(WebCore::DOMWindowProperty::resumeFromPageCache):
(WebCore::DOMWindowProperty::disconnectFrameForDocumentSuspension): Deleted.
(WebCore::DOMWindowProperty::reconnectFrameFromDocumentSuspension): Deleted.
* page/DOMWindowProperty.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237029 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoWeb Inspector: create special Network waterfall for media events
drousso@apple.com [Thu, 11 Oct 2018 04:13:17 +0000 (04:13 +0000)]
Web Inspector: create special Network waterfall for media events
https://bugs.webkit.org/show_bug.cgi?id=189773
<rdar://problem/44626605>

Reviewed by Joseph Pecoraro.

Source/JavaScriptCore:

* inspector/protocol/DOM.json:
Add `didFireEvent` event that is fired when specific event listeners added by
`InspectorInstrumentation::addEventListenersToNode` are fired.

Source/WebCore:

Test: http/tests/inspector/dom/didFireEvent.html

* html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::HTMLMediaElement):

* inspector/InspectorInstrumentation.h:
(WebCore::InspectorInstrumentation::addEventListenersToNode): Added.
* inspector/InspectorInstrumentation.cpp:
(WebCore::InspectorInstrumentation::addEventListenersToNodeImpl): Added.

* inspector/agents/InspectorDOMAgent.h:
* inspector/agents/InspectorDOMAgent.cpp:
(WebCore::EventFiredCallback): Added.
(WebCore::EventFiredCallback::create): Added.
(WebCore::EventFiredCallback::operator==): Added.
(WebCore::EventFiredCallback::handleEvent): Added.
(WebCore::EventFiredCallback::EventFiredCallback): Added.
(WebCore::InspectorDOMAgent::didCreateFrontendAndBackend):
(WebCore::InspectorDOMAgent::addEventListenersToNode): Added.

Source/WebInspectorUI:

* Localizations/en.lproj/localizedStrings.js:
* UserInterface/Main.html:
* UserInterface/Base/Utilities.js:

* UserInterface/Protocol/DOMObserver.js:
(WI.DOMObserver.prototype.didFireEvent): Added.
* UserInterface/Controllers/DOMManager.js:
(WI.DOMManager.prototype.didFireEvent): Added.
* UserInterface/Models/DOMNode.js:
(WI.DOMNode):
(WI.DOMNode.prototype.get domEvents): Added.
(WI.DOMNode.prototype.didFireEvent): Added.
(WI.DOMNode.prototype._addDOMEvent): Added.

* UserInterface/Views/NetworkTableContentView.js:
(WI.NetworkTableContentView):
(WI.NetworkTableContentView.prototype.shown):
(WI.NetworkTableContentView.prototype.hidden):
(WI.NetworkTableContentView.prototype.closed):
(WI.NetworkTableContentView.prototype.reset):
(WI.NetworkTableContentView.prototype.showRepresentedObject):
(WI.NetworkTableContentView.prototype.networkDetailViewClose): Added.
(WI.NetworkTableContentView.prototype.tableSortChanged):
(WI.NetworkTableContentView.prototype.tableSelectionDidChange):
(WI.NetworkTableContentView.prototype._populateNameCell):
(WI.NetworkTableContentView.prototype._populateWaterfallGraph.positionByStartOffset): Added.
(WI.NetworkTableContentView.prototype._populateWaterfallGraph.setWidthForDuration): Added.
(WI.NetworkTableContentView.prototype._populateWaterfallGraph.createDOMEventLine): Added.
(WI.NetworkTableContentView.prototype._populateWaterfallGraph.appendBlock):
(WI.NetworkTableContentView.prototype._populateWaterfallGraph):
(WI.NetworkTableContentView.prototype._processPendingEntries):
(WI.NetworkTableContentView.prototype._rowIndexForRepresentedObject): Added.
(WI.NetworkTableContentView.prototype._updateEntryForResource):
(WI.NetworkTableContentView.prototype._hideDetailView): Added.
(WI.NetworkTableContentView.prototype._showDetailView): Added.
(WI.NetworkTableContentView.prototype._positionDetailView): Added.
(WI.NetworkTableContentView.prototype._resourceTransferSizeDidChange):
(WI.NetworkTableContentView.prototype._tryLinkResourceToDOMNode):
(WI.NetworkTableContentView.prototype._handleNodeDidFireEvent): Added.
(WI.NetworkTableContentView.prototype._updateFilteredEntries):
(WI.NetworkTableContentView.prototype._typeFilterScopeBarSelectionChanged):
(WI.NetworkTableContentView.prototype._urlFilterDidChange):
(WI.NetworkTableContentView.prototype._restoreSelectedRow):
(WI.NetworkTableContentView.prototype._waterfallPopoverContent): Added.
(WI.NetworkTableContentView.prototype._waterfallPopoverContentForResourceEntry): Added.
(WI.NetworkTableContentView.prototype._waterfallPopoverContentForNodeEntry): Added.
(WI.NetworkTableContentView.prototype._handleResourceEntryMousedownWaterfall): Added.
(WI.NetworkTableContentView.prototype._handleNodeEntryMousedownWaterfall): Added.
(WI.NetworkTableContentView.prototype._handleMousedownWaterfall): Added.
(WI.NetworkTableContentView.prototype.networkResourceDetailViewClose): Deleted.
(WI.NetworkTableContentView.prototype._rowIndexForResource): Deleted.
(WI.NetworkTableContentView.prototype._hideResourceDetailView): Deleted.
(WI.NetworkTableContentView.prototype._showResourceDetailView): Deleted.
(WI.NetworkTableContentView.prototype._waterfallPopoverContentForResource): Deleted.
* UserInterface/Views/NetworkTableContentView.css:
(.content-view.network .network-table): Added.
(.network-table :not(.header) .cell.waterfall .waterfall-container > .dom-event): Added.
(.network-table :not(.header) .cell.waterfall .waterfall-container > .dom-activity): Added.
(.network-table :not(.header) .cell.waterfall .waterfall-container > .dom-activity.playing): Added.

* UserInterface/Views/NetworkDOMNodeDetailView.js: Added.
(WI.NetworkDOMNodeDetailView):
(WI.NetworkDOMNodeDetailView.prototype.initialLayout):
(WI.NetworkDOMNodeDetailView.prototype.showContentViewForIdentifier):

* UserInterface/Views/NetworkResourceDetailView.css:
(.content-view.resource-details):
(.network-resource-detail): Deleted.
(.network-resource-detail .navigation-bar): Deleted.
(.network-resource-detail .item.close > .glyph): Deleted.
(.network-resource-detail .item.close > .glyph:hover): Deleted.
(.network-resource-detail .item.close > .glyph:active): Deleted.
(.network .network-resource-detail .navigation-bar .item.radio.button.text-only): Deleted.
(.network .network-resource-detail .navigation-bar .item.radio.button.text-only.selected): Deleted.
(.network-resource-detail > .content-browser): Deleted.
(@media (prefers-dark-interface)): Deleted.
* UserInterface/Views/NetworkResourceDetailView.js:
(WI.NetworkResourceDetailView):
(WI.NetworkResourceDetailView.prototype.shown):
(WI.NetworkResourceDetailView.prototype.headersContentViewGoToRequestData):
(WI.NetworkResourceDetailView.prototype.sizesContentViewGoToHeaders):
(WI.NetworkResourceDetailView.prototype.sizesContentViewGoToRequestBody):
(WI.NetworkResourceDetailView.prototype.sizesContentViewGoToResponseBody):
(WI.NetworkResourceDetailView.prototype.initialLayout):
(WI.NetworkResourceDetailView.prototype.showContentViewForIdentifier):
(WI.NetworkResourceDetailView.prototype.get resource): Deleted.
(WI.NetworkResourceDetailView.prototype.hidden): Deleted.
(WI.NetworkResourceDetailView.prototype.dispose): Deleted.
(WI.NetworkResourceDetailView.prototype.willShowWithCookie): Deleted.
(WI.NetworkResourceDetailView.prototype.initialLayout): Deleted.
(WI.NetworkResourceDetailView.prototype._showPreferredContentView): Deleted.
(WI.NetworkResourceDetailView.prototype._showContentViewForNavigationItem): Deleted.
(WI.NetworkResourceDetailView.prototype._navigationItemSelected): Deleted.
(WI.NetworkResourceDetailView.prototype._handleCloseButton): Deleted.

* UserInterface/Views/NetworkDetailView.js: Added.
(WI.NetworkDetailView):
(WI.NetworkDetailView.prototype.get representedObject):
(WI.NetworkDetailView.prototype.shown):
(WI.NetworkDetailView.prototype.hidden):
(WI.NetworkDetailView.prototype.dispose):
(WI.NetworkDetailView.prototype.willShowWithCookie):
(WI.NetworkDetailView.prototype.initialLayout):
(WI.NetworkDetailView.prototype.createDetailNavigationItem):
(WI.NetworkDetailView.prototype.detailNavigationItemForIdentifier):
(WI.NetworkDetailView.prototype.showContentViewForIdentifier):
(WI.NetworkDetailView.prototype._showPreferredContentView):
(WI.NetworkDetailView.prototype._navigationItemSelected):
(WI.NetworkDetailView.prototype._handleCloseButton):
* UserInterface/Views/NetworkDetailView.css: Added.
(.network-detail):
(.network-detail .navigation-bar):
(.network-detail .item.close > .glyph):
(.network-detail .item.close > .glyph:hover):
(.network-detail .item.close > .glyph:active):
(.network .network-detail .navigation-bar .item.radio.button.text-only):
(.network .network-detail .navigation-bar .item.radio.button.text-only.selected):
(.network-detail > .content-browser):
(@media (prefers-dark-interface)):
Create base class for detail views shown in the Network tab.

* UserInterface/Views/DOMNodeEventsContentView.js: Added.
(WI.DOMNodeEventsContentView):
(WI.DOMNodeEventsContentView.prototype.initialLayout):
(WI.DOMNodeEventsContentView.prototype.closed):
(WI.DOMNodeEventsContentView.prototype._handleDOMNodeDidFireEvent):
* UserInterface/Views/DOMNodeEventsContentView.css: Added.
(.dom-node-details.dom-events):

* UserInterface/Views/DOMEventsBreakdownView.js: Added.
(WI.DOMEventsBreakdownView):
(WI.DOMEventsBreakdownView.prototype.addEvent):
(WI.DOMEventsBreakdownView.prototype.initialLayout):
(WI.DOMEventsBreakdownView.prototype._populateTable.percentOfTotalTime):
(WI.DOMEventsBreakdownView.prototype._populateTable):
* UserInterface/Views/DOMEventsBreakdownView.css: Added.
(.waterfall-popover-content .dom-events-breakdown):
(.dom-events-breakdown):
(.dom-events-breakdown table):
(.dom-events-breakdown tr > :matches(th, td)):
(.dom-events-breakdown tbody > tr):
(.dom-events-breakdown .graph):
(.dom-events-breakdown .graph > :matches(.point, .area)):
(.dom-events-breakdown .graph > .point):
(.dom-events-breakdown .time):

* UserInterface/Views/ResourceTimingBreakdownView.css:
(.resource-timing-breakdown > table > tr.header:not(.total-row) > td): Added.
(.popover.waterfall-popover): Deleted.

LayoutTests:

* http/tests/inspector/dom/didFireEvent-expected.txt: Added.
* http/tests/inspector/dom/didFireEvent.html: Added.

* inspector/unit-tests/array-utilities-expected.txt:
* inspector/unit-tests/array-utilities.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237028 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago[iOS] Cleanup -[WAKView _selfHandleEvent:] and -[WAKWindow sendEventSynchronously:]
dbates@webkit.org [Thu, 11 Oct 2018 03:08:34 +0000 (03:08 +0000)]
[iOS] Cleanup -[WAKView _selfHandleEvent:] and -[WAKWindow sendEventSynchronously:]
https://bugs.webkit.org/show_bug.cgi?id=190402

Reviewed by Anders Carlsson.

Unindent case statements. Remove use of default case statements with body ASSERT_NOT_REACHED()
to catch the cases of a missing enumerator. By omitting the default case statement we turn
such a scenario into a compile-time error instead of handling it as a runtime error.

* platform/ios/wak/WAKView.mm:
(-[WAKView _selfHandleEvent:]):
* platform/ios/wak/WAKWindow.mm:
(-[WAKWindow sendEventSynchronously:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237027 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoSafari crashes when attempting to share a YouTube video while in full-screen
timothy_horton@apple.com [Thu, 11 Oct 2018 03:03:58 +0000 (03:03 +0000)]
Safari crashes when attempting to share a YouTube video while in full-screen
https://bugs.webkit.org/show_bug.cgi?id=190456
<rdar://problem/45130748>

Reviewed by Wenson Hsieh.

* Platform/spi/ios/UIKitSPI.h:
* UIProcess/Cocoa/WKShareSheet.mm:
(-[WKShareSheet presentWithParameters:completionHandler:]):
Allow the share sheet to present centered, since we don't have enough
context to put it in a more sensible place.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237026 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoRename MarkupAccumulator::appendStartTag, appendElement, etc... for clarity
rniwa@webkit.org [Thu, 11 Oct 2018 02:27:18 +0000 (02:27 +0000)]
Rename MarkupAccumulator::appendStartTag, appendElement, etc... for clarity
https://bugs.webkit.org/show_bug.cgi?id=190308

Reviewed by Darin Adler.

Renamed appendStartTag and appendEndTag to startAppendingNode and endAppendingNode since serialize any node,
not just elements which produce start and end tags.

Renamed appendElement and appendEndElement to appendStartTag and appendEndTag since that's what they do.

* editing/MarkupAccumulator.cpp:
(WebCore::elementCannotHaveEndTag): Made this a static local function.
(WebCore::shouldSelfClose): Ditto.
(WebCore::MarkupAccumulator::serializeNodesWithNamespaces):
(WebCore::MarkupAccumulator::startAppendingNode): Renamed from appendStartTag.
(WebCore::MarkupAccumulator::endAppendingNode): Renamed from appendEndElement and merged appendEndMarkup here.
(WebCore::MarkupAccumulator::appendTextSubstring): Deleted. This was only used by StyledMarkupAccumulator.
(WebCore::MarkupAccumulator::appendStringView): Added.
(WebCore::MarkupAccumulator::appendStartTag): Renamed from appendElement.
(WebCore::MarkupAccumulator::appendCloseTag):
(WebCore::MarkupAccumulator::appendNonElementNode): Renamed from appendStartMarkup. No longer serializes
an element. StyledMarkupAccumulator had a check before calling this function already so this clarifies
the purpose of this function.
(WebCore::MarkupAccumulator::appendElement): Deleted.
(WebCore::MarkupAccumulator::appendEndMarkup): Deleted. Merged into appendEndTag.
* editing/MarkupAccumulator.h:
(WebCore::MarkupAccumulator::appendNodeEnd): Renamed from appendEndTag.
* editing/markup.cpp:
(WebCore::StyledMarkupAccumulator::wrapWithNode):
(WebCore::StyledMarkupAccumulator::appendStartTag): Renamed from appendElement.
(WebCore::StyledMarkupAccumulator::appendEndTag): Renamed from appendEndElement.
(WebCore::StyledMarkupAccumulator::traverseNodesForSerialization):
(WebCore::StyledMarkupAccumulator::appendNodeToPreserveMSOList): Use String::substring directly instead of
going through appendTextSubstring which has been deleted.
* page/PageSerializer.cpp:
(WebCore::PageSerializer::SerializerMarkupAccumulator::appendStartTag): Renamed from appendElement.
(WebCore::PageSerializer::SerializerMarkupAccumulator::appendEndTag): Renamed from appendEndElement.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237025 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoIncrease executable memory pool from 64MB to 128MB for ARM64
msaboff@apple.com [Thu, 11 Oct 2018 01:39:27 +0000 (01:39 +0000)]
Increase executable memory pool from 64MB to 128MB for ARM64
https://bugs.webkit.org/show_bug.cgi?id=190453

Reviewed by Saam Barati.

* jit/ExecutableAllocator.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237024 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoAdd WinCairo to the flakiness dashboard
Hironori.Fujii@sony.com [Thu, 11 Oct 2018 01:25:05 +0000 (01:25 +0000)]
Add WinCairo to the flakiness dashboard
https://bugs.webkit.org/show_bug.cgi?id=188345

Reviewed by Lucas Forschler.

* TestResultServer/static-dashboards/builders.jsonp: Regenerated by generate_builders_json.py.
* TestResultServer/static-dashboards/flakiness_dashboard.js: Added a new platform 'WINCAIRO'.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237023 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago[WHSL -> MSL] Annotate semantics correctly in generated MSL
dino@apple.com [Thu, 11 Oct 2018 01:09:55 +0000 (01:09 +0000)]
[WHSL -> MSL] Annotate semantics correctly in generated MSL
https://bugs.webkit.org/show_bug.cgi?id=190452
<rdar://problem/45178272>

Reviewed by Myles Maxfield.

Correctly label the appropriate semantics as [[position]]
and [[color(N)]].

* WebGPUShadingLanguageRI/Metal/MSLBackend.js: Identify position
and color attributes. Also use the semantic's index rather than
a counter.
* WebGPUShadingLanguageRI/Metal/WhlslToMsl.html: Use a <pre> for
output, and only include the actual shader code.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237022 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoUnreviewed, attempt to fix the build with current SDKs.
ryanhaddad@apple.com [Thu, 11 Oct 2018 00:47:28 +0000 (00:47 +0000)]
Unreviewed, attempt to fix the build with current SDKs.

* crypto/mac/CryptoKeyRSAMac.cpp:
(WebCore::getPublicKeyComponents):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237021 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoREGRESSION (r236935): Layout test fast/events/ios/keyboard-scrolling-distance.html...
timothy_horton@apple.com [Wed, 10 Oct 2018 22:41:26 +0000 (22:41 +0000)]
REGRESSION (r236935): Layout test fast/events/ios/keyboard-scrolling-distance.html is timing out
https://bugs.webkit.org/show_bug.cgi?id=190444
<rdar://problem/45110698>

Reviewed by Simon Fraser.

There's a race in WKKeyboardScrollingAnimator that's exacerbated by HIDEventGenerator
being much faster than a human finger. We get our "begin" events from interpretKeyEvent,
after the Web Content process has had its way with it, but currently the
back-channel "handle" events (e.g. for key up, which doesn't go to interpretKeyEvent)
are retrieved from handleKeyWebEvent in the UI process, which is *before*
the Web Content process has had a swing at it.

If you lose the race (an insanely short tap like you get from HIDEventGenerator,
or with a very busy Web Content process), we see handle(keyDown), handle(keyUp), begin(keyDown),
and get stuck scrolling!

Instead, retrieve the out-of-band "handle" events from _didHandleKeyEvent,
so that they're sensibly and strictly ordered with respect to the timing of interpretKeyEvent/"begin".

Also, hook up didFinishScrolling, so that UIScriptController's callbacks work correctly.

* UIProcess/API/Cocoa/WKWebViewInternal.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView handleKeyWebEvent:]):
(-[WKContentView _didHandleKeyEvent:eventWasHandled:]):
(-[WKContentView keyboardScrollViewAnimatorDidFinishScrolling:]):
* UIProcess/ios/WKKeyboardScrollingAnimator.h:
* UIProcess/ios/WKKeyboardScrollingAnimator.mm:
(-[WKKeyboardScrollingAnimator handleKeyEvent:]):
(-[WKKeyboardScrollingAnimator displayLinkFired:]):
(-[WKKeyboardScrollViewAnimator setDelegate:]):
(-[WKKeyboardScrollViewAnimator handleKeyEvent:]):
(-[WKKeyboardScrollViewAnimator didFinishScrolling]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237020 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoWeb Inspector: REGRESSION(r236853): Uncaught Exception: undefined is not an object...
drousso@apple.com [Wed, 10 Oct 2018 22:41:25 +0000 (22:41 +0000)]
Web Inspector: REGRESSION(r236853): Uncaught Exception: undefined is not an object (evaluating 'entry.resource')
https://bugs.webkit.org/show_bug.cgi?id=190442

Reviewed by Joseph Pecoraro.

* UserInterface/Views/NetworkTableContentView.js:
(WI.NetworkTableContentView.prototype.reset):
It's unnecessary to deselect rows in the `WI.Table` when we're about to remove them.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237019 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoOnly report the supported WebGL version
bfulgham@apple.com [Wed, 10 Oct 2018 22:25:17 +0000 (22:25 +0000)]
Only report the supported WebGL version
https://bugs.webkit.org/show_bug.cgi?id=190434
<rdar://problem/45024677>

Reviewed by Dean Jackson.

Source/WebCore:

Tested by fast/canvas/webgl/gl-getstring.html

Revise getParameter(gl.VERSION) to only return the WebGL version without the hardware and
driver-specific details available through the low-level OpenGL driver interface. These details
are not needed for WebGL use and expose information about the user's system that we do not
need to share.

* html/canvas/WebGL2RenderingContext.cpp:
(WebCore::WebGL2RenderingContext::getParameter):
* html/canvas/WebGLRenderingContext.cpp:
(WebCore::WebGLRenderingContext::getParameter):

LayoutTests:

* fast/canvas/webgl/gl-getstring-expected.txt:
* fast/canvas/webgl/gl-getstring.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237018 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoUnreviewed, add missing headers for inline functions
yusukesuzuki@slowstart.org [Wed, 10 Oct 2018 21:23:42 +0000 (21:23 +0000)]
Unreviewed, add missing headers for inline functions
https://bugs.webkit.org/show_bug.cgi?id=190429

* platform/graphics/avfoundation/cf/MediaPlayerPrivateAVFoundationCF.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237017 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoWeb Inspector: REGRESSION: selection in network table is lost when new entries are...
drousso@apple.com [Wed, 10 Oct 2018 20:52:12 +0000 (20:52 +0000)]
Web Inspector: REGRESSION: selection in network table is lost when new entries are added
https://bugs.webkit.org/show_bug.cgi?id=190362

Reviewed by Matt Baker.

The `_selectedRows` list is cleared each time `reloadData` is called, meaning that
`WI.Table` expects its "owner" to restore the selection (since it doesn't know how). As a
result, `WI.NetworkTableContentView` needs to call `selectRow` _after_ `reloadData` is
called, not before.

* UserInterface/Views/NetworkTableContentView.js:
(WI.NetworkTableContentView.prototype.tableSortChanged):
(WI.NetworkTableContentView.prototype._populateNameCell):
(WI.NetworkTableContentView.prototype._processPendingEntries):
(WI.NetworkTableContentView.prototype._insertResourceAndReloadTable):
(WI.NetworkTableContentView.prototype._updateSort): Added.
(WI.NetworkTableContentView.prototype._updateFilteredEntries):
(WI.NetworkTableContentView.prototype._reloadTable): Added.
(WI.NetworkTableContentView.prototype._resetFilters):
(WI.NetworkTableContentView.prototype._typeFilterScopeBarSelectionChanged):
(WI.NetworkTableContentView.prototype._handleGroupByDOMNodeCheckedDidChange):
(WI.NetworkTableContentView.prototype._urlFilterDidChange):
(WI.NetworkTableContentView.prototype._updateSortAndFilteredEntries): Deleted.

* UserInterface/Views/Table.js:
(WI.Table.prototype.reloadData):
Ensure that the `_selectedRowIndex` is also reset.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237016 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agobuild-webkit --inspector-frontend shouldn't have a nonzero exit code
ross.kirsling@sony.com [Wed, 10 Oct 2018 20:29:01 +0000 (20:29 +0000)]
build-webkit --inspector-frontend shouldn't have a nonzero exit code
https://bugs.webkit.org/show_bug.cgi?id=190430

Reviewed by Brian Burg.

* Scripts/webkitdirs.pm:
(argumentsForConfiguration):
--inspector-frontend should only be popped off of argv by determineIsInspectorFrontend;
it shouldn't be re-inserted as a CLI option by argumentsForConfiguration.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237015 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoVersioning.
kocsen_chung@apple.com [Wed, 10 Oct 2018 20:06:05 +0000 (20:06 +0000)]
Versioning.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237014 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoShare more WKShareSheet code between macOS and iOS, and fix a few bugs
timothy_horton@apple.com [Wed, 10 Oct 2018 19:05:03 +0000 (19:05 +0000)]
Share more WKShareSheet code between macOS and iOS, and fix a few bugs
https://bugs.webkit.org/show_bug.cgi?id=190420

Reviewed by Simon Fraser.

Source/WebCore:

Test: fast/web-share/share-with-no-url.html

* page/Navigator.cpp:
(WebCore::Navigator::share):
* page/ShareData.h:
Make 'url' an optional field on ShareDataWithParsedURL, and don't
reject the share() promise if the URL is totally empty (as opposed to invalid).

Source/WebKit:

* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _resolutionForShareSheetImmediateCompletionForTesting]):
(-[WKWebView _setShareSheetCompletesImmediatelyWithResolutionForTesting:]):
(-[WKWebView _invokeShareSheetWithResolution:]): Deleted.
Add a new mechanism to tell the share sheet up-front that it should
immediately dismiss with a particular resolution, for testing.
See the Tools ChangeLog for justification.
We keep the bit on WKWebView instead of WKShareSheet so that it can be
called prior to invoking the share sheet.

* UIProcess/API/Cocoa/WKWebViewInternal.h:
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/Cocoa/WKShareSheet.h:
* UIProcess/Cocoa/WKShareSheet.mm:
(-[WKShareSheet initWithView:]):
(-[WKShareSheet presentWithParameters:completionHandler:]):
(-[WKShareSheet sharingServicePicker:didChooseSharingService:]):
(-[WKShareSheet _didCompleteWithSuccess:]):
(-[WKShareSheet dismiss]):
(-[WKShareSheet dispatchDidDismiss]):
(-[WKShareSheet initWithView:initWithView:]): Deleted.
(-[WKShareSheet _dispatchDidDismiss]): Deleted.
(-[WKShareSheet _cancel]): Deleted.
(-[WKShareSheet _dismissDisplayAnimated:]): Deleted.
(-[WKShareSheet _presentFullscreenViewController:animated:]): Deleted.
(-[WKShareSheet invokeShareSheetWithResolution:]): Deleted.
Reduce the number of extraneous members, methods, and #ifdefs.
Try to share as much of WKShareSheet between the platforms as makes sense.
The interface is now totally identical on the two platforms.

* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::showShareSheet):
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _showShareSheet:completionHandler:]):
(-[WKContentView invokeShareSheetWithResolution:]): Deleted.
Instead of silently failing (without even calling the completion handler!)
if we try to present one share sheet on top of another, dismiss the first
and present the new one. This is important because there are (separately
filed) bugs on both platforms where the completion handler is not called;
now, instead of a tab getting stuck forever, we'll reject the promise
when another share takes place.

Tools:

* DumpRenderTree/ios/UIScriptControllerIOS.mm:
(WTR::UIScriptController::setShareSheetCompletesImmediatelyWithResolution):
(WTR::UIScriptController::invokeShareSheetWithResolution): Deleted.
* DumpRenderTree/mac/UIScriptControllerMac.mm:
(WTR::UIScriptController::setShareSheetCompletesImmediatelyWithResolution):
* TestRunnerShared/UIScriptContext/Bindings/UIScriptController.idl:
* TestRunnerShared/UIScriptContext/UIScriptController.cpp:
(WTR::UIScriptController::setShareSheetCompletesImmediatelyWithResolution):
(WTR::UIScriptController::invokeShareSheetWithResolution): Deleted.
* TestRunnerShared/UIScriptContext/UIScriptController.h:
* WebKitTestRunner/ios/UIScriptControllerIOS.mm:
(WTR::UIScriptController::setShareSheetCompletesImmediatelyWithResolution):
(WTR::UIScriptController::invokeShareSheetWithResolution): Deleted.
* WebKitTestRunner/mac/UIScriptControllerMac.mm:
(WTR::UIScriptController::setShareSheetCompletesImmediatelyWithResolution):
Change the testing mechanism for the share sheet such that we up-front
set that it should immediately return with a particular state, instead
of programmatically dismissing the sheet.

Otherwise, running the tests on macOS doesn't work, because we pop
a context menu, which spins the runloop and prevents dismissal from
going through.

LayoutTests:

* fast/events/ios/share-expected.txt: Removed.
* fast/events/ios/share.html: Removed.
* fast/web-share/share.html: Added.
* fast/web-share/share-expected.txt: Added.
* resources/ui-helper.js:
(window.UIHelper.setShareSheetCompletesImmediatelyWithResolution):
(window.UIHelper.invokeShareSheetWithResolution): Deleted.
Move the existing share sheet test to fast/web-share so it runs on both macOS and iOS.

* fast/web-share/share-with-no-url.html: Added.
* fast/web-share/share-with-no-url-expected.txt: Added.
Add a new test that the sheet is successfully invoked even when a URL is not provided.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237012 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoREGRESSION (r236678): Keyboard scrolling with arrow keys doesn't work on iOS
timothy_horton@apple.com [Wed, 10 Oct 2018 18:43:20 +0000 (18:43 +0000)]
REGRESSION (r236678): Keyboard scrolling with arrow keys doesn't work on iOS
https://bugs.webkit.org/show_bug.cgi?id=190433
<rdar://problem/45111986>

Reviewed by Simon Fraser.

* UIProcess/ios/WKKeyboardScrollingAnimator.mm:
(-[WKKeyboardScrollingAnimator keyboardScrollForEvent:]):
(-[WKKeyboardScrollingAnimator beginWithEvent:]): Deleted.
(-[WKKeyboardScrollViewAnimator beginWithEvent:]): Deleted.
r236678 changes the characters in charactersIgnoringModifiers to match AppKit.
We similarly need to update WKKeyboardScrollingAnimator's comparison.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237011 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoWeb Inspector: notify the frontend when a canvas has started recording via console...
drousso@apple.com [Wed, 10 Oct 2018 18:34:49 +0000 (18:34 +0000)]
Web Inspector: notify the frontend when a canvas has started recording via console.record
https://bugs.webkit.org/show_bug.cgi?id=190306

Reviewed by Brian Burg.

Source/JavaScriptCore:

* inspector/protocol/Canvas.json:
Add `recordingStarted` event.

* inspector/protocol/Recording.json:
Add `Initiator` enum for determining who started the recording.

Source/WebCore:

Updated existing tests: LayoutTests/inspector/canvas/recording-2d.html
                        LayoutTests/inspector/canvas/recording-bitmaprenderer.html
                        LayoutTests/inspector/canvas/recording-webgl-snapshots.html
                        LayoutTests/inspector/canvas/recording-webgl.html

* inspector/agents/InspectorCanvasAgent.cpp:
(WebCore::InspectorCanvasAgent::startRecording):
(WebCore::InspectorCanvasAgent::consoleStartRecordingCanvas):

* inspector/InspectorCanvas.h:
Default `singleFrame` to false to better match the default behaviour of other `console`
functions (`profile` doesn't stop until `profileEnd` is called, so the same should be true
of `record` and `recordEnd`).

Source/WebInspectorUI:

* UserInterface/Protocol/CanvasObserver.js:
(WI.CanvasObserver.prototype.recordingStarted): Added.

* UserInterface/Protocol/CanvasManager.js:
(WI.CanvasManager.prototype.recordingStarted): Added.

* UserInterface/Models/Canvas.js:
(WI.Canvas.prototype.startRecording):
(WI.Canvas.prototype.recordingStarted): Added.
(WI.Canvas.prototype.recordingFinished):

LayoutTests:

* inspector/canvas/recording-2d.html:
* inspector/canvas/recording-bitmaprenderer.html:
* inspector/canvas/recording-webgl-snapshots.html:
* inspector/canvas/recording-webgl.html:
* inspector/canvas/resources/recording-utilities.js:
(TestPage.registerInitializer.window.startRecording):
Wait to complete the test until the "LastFrame" event is fired, ensuring that any running
commands don't affect the next test case.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237010 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago[JSC] Rename createXXX to tryCreateXXX if it can return RefPtr
yusukesuzuki@slowstart.org [Wed, 10 Oct 2018 18:29:44 +0000 (18:29 +0000)]
[JSC] Rename createXXX to tryCreateXXX if it can return RefPtr
https://bugs.webkit.org/show_bug.cgi?id=190429

Reviewed by Saam Barati.

Source/JavaScriptCore:

Some createXXX functions can fail. But sometimes the caller does not perform error checking.
To make it explicit that these functions can fail, we rename these functions from createXXX
to tryCreateXXX. In this patch, we focus on non-JS-managed factory functions. If the factory
function does not fail, it should return Ref<>. Otherwise, it should be named as tryCreateXXX
and it should return RefPtr<>.

This patch mainly focuses on TypedArray factory functions. Previously, these functions are
`RefPtr<XXXArray> create(...)`. This patch changes them to `RefPtr<XXXArray> tryCreate(...)`.
And we also introduce `Ref<XXXArray> create(...)` function which internally performs
RELEASE_ASSERT on the result of `tryCreate(...)`.

And we also convert OpaqueJSString::create to OpaqueJSString::tryCreate since it can fail.

This change actually finds one place which does not perform any null checkings while it uses
`RefPtr<> create(...)` function.

* API/JSCallbackObjectFunctions.h:
(JSC::JSCallbackObject<Parent>::getOwnPropertySlot):
(JSC::JSCallbackObject<Parent>::put):
(JSC::JSCallbackObject<Parent>::putByIndex):
(JSC::JSCallbackObject<Parent>::deleteProperty):
(JSC::JSCallbackObject<Parent>::callbackGetter):
* API/JSClassRef.h:
(StaticValueEntry::StaticValueEntry):
* API/JSContext.mm:
(-[JSContext evaluateScript:withSourceURL:]):
(-[JSContext setName:]):
* API/JSContextRef.cpp:
(JSGlobalContextCopyName):
(JSContextCreateBacktrace):
* API/JSObjectRef.cpp:
(JSObjectCopyPropertyNames):
* API/JSScriptRef.cpp:
* API/JSStringRef.cpp:
(JSStringCreateWithCharactersNoCopy):
* API/JSValue.mm:
(+[JSValue valueWithNewRegularExpressionFromPattern:flags:inContext:]):
(+[JSValue valueWithNewErrorFromMessage:inContext:]):
(+[JSValue valueWithNewSymbolFromDescription:inContext:]):
(performPropertyOperation):
(-[JSValue invokeMethod:withArguments:]):
(containerValueToObject):
(objectToValueWithoutCopy):
(objectToValue):
* API/JSValueRef.cpp:
(JSValueCreateJSONString):
(JSValueToStringCopy):
* API/OpaqueJSString.cpp:
(OpaqueJSString::tryCreate):
(OpaqueJSString::create): Deleted.
* API/OpaqueJSString.h:
* API/glib/JSCContext.cpp:
(evaluateScriptInContext):
* API/glib/JSCValue.cpp:
(jsc_value_new_string_from_bytes):
* ftl/FTLLazySlowPath.h:
(JSC::FTL::LazySlowPath::createGenerator):
* ftl/FTLLazySlowPathCall.h:
(JSC::FTL::createLazyCallGenerator):
* ftl/FTLOSRExit.cpp:
(JSC::FTL::OSRExitDescriptor::emitOSRExit):
(JSC::FTL::OSRExitDescriptor::emitOSRExitLater):
(JSC::FTL::OSRExitDescriptor::prepareOSRExitHandle):
* ftl/FTLOSRExit.h:
* ftl/FTLPatchpointExceptionHandle.cpp:
(JSC::FTL::PatchpointExceptionHandle::create):
(JSC::FTL::PatchpointExceptionHandle::createHandle):
* ftl/FTLPatchpointExceptionHandle.h:
* heap/EdenGCActivityCallback.h:
(JSC::GCActivityCallback::tryCreateEdenTimer):
(JSC::GCActivityCallback::createEdenTimer): Deleted.
* heap/FullGCActivityCallback.h:
(JSC::GCActivityCallback::tryCreateFullTimer):
(JSC::GCActivityCallback::createFullTimer): Deleted.
* heap/GCActivityCallback.h:
* heap/Heap.cpp:
(JSC::Heap::Heap):
* inspector/AsyncStackTrace.cpp:
(Inspector::AsyncStackTrace::create):
* inspector/AsyncStackTrace.h:
* jsc.cpp:
(fillBufferWithContentsOfFile):
* runtime/ArrayBuffer.h:
* runtime/GenericTypedArrayView.h:
* runtime/GenericTypedArrayViewInlines.h:
(JSC::GenericTypedArrayView<Adaptor>::create):
(JSC::GenericTypedArrayView<Adaptor>::tryCreate):
(JSC::GenericTypedArrayView<Adaptor>::createUninitialized):
(JSC::GenericTypedArrayView<Adaptor>::tryCreateUninitialized):
(JSC::GenericTypedArrayView<Adaptor>::subarray const):
* runtime/JSArrayBufferView.cpp:
(JSC::JSArrayBufferView::possiblySharedImpl):
* runtime/JSGenericTypedArrayViewInlines.h:
(JSC::JSGenericTypedArrayView<Adaptor>::possiblySharedTypedImpl):
(JSC::JSGenericTypedArrayView<Adaptor>::unsharedTypedImpl):
* wasm/WasmMemory.cpp:
(JSC::Wasm::Memory::create):
(JSC::Wasm::Memory::tryCreate):
* wasm/WasmMemory.h:
* wasm/WasmTable.cpp:
(JSC::Wasm::Table::tryCreate):
(JSC::Wasm::Table::create): Deleted.
* wasm/WasmTable.h:
* wasm/js/JSWebAssemblyInstance.cpp:
(JSC::JSWebAssemblyInstance::create):
* wasm/js/JSWebAssemblyMemory.cpp:
(JSC::JSWebAssemblyMemory::JSWebAssemblyMemory):
* wasm/js/WebAssemblyMemoryConstructor.cpp:
(JSC::constructJSWebAssemblyMemory):
* wasm/js/WebAssemblyModuleRecord.cpp:
(JSC::WebAssemblyModuleRecord::link):
* wasm/js/WebAssemblyTableConstructor.cpp:
(JSC::constructJSWebAssemblyTable):

Source/WebCore:

Use `tryCreate` or new `create`. If we use new `create`, we can use Ref<> and remove null check.

* Modules/encryptedmedia/legacy/LegacyCDMSessionClearKey.cpp:
(WebCore::CDMSessionClearKey::cachedKeyForKeyID const):
* Modules/plugins/QuickTimePluginReplacement.mm:
(WebCore::jsValueWithDictionaryInContext):
* Modules/webaudio/AudioBuffer.cpp:
(WebCore::AudioBuffer::AudioBuffer):
(WebCore::AudioBuffer::getChannelData):
* Modules/webvr/VREyeParameters.cpp:
(WebCore::VREyeParameters::offset const):
* Modules/webvr/VRFrameData.cpp:
(WebCore::matrixToArray):
* Modules/webvr/VRPose.cpp:
(WebCore::optionalFloat3ToJSCArray):
(WebCore::VRPose::position const):
(WebCore::VRPose::orientation const):
* Modules/webvr/VRStageParameters.cpp:
(WebCore::VRStageParameters::sittingToStandingTransform const):
* bindings/js/ReadableStreamDefaultController.h:
(WebCore::ReadableStreamDefaultController::enqueue):
* bindings/js/SerializedScriptValue.cpp:
(WebCore::CloneDeserializer::readArrayBufferView):
* crypto/gcrypt/CryptoKeyRSAGCrypt.cpp:
(WebCore::CryptoKeyRSA::algorithm const):
* crypto/mac/CryptoKeyRSAMac.cpp:
(WebCore::CryptoKeyRSA::algorithm const):
* css/DOMMatrixReadOnly.cpp:
(WebCore::DOMMatrixReadOnly::toFloat32Array const):
(WebCore::DOMMatrixReadOnly::toFloat64Array const):
* css/FontFace.cpp:
(WebCore::FontFace::create):
* dom/TextEncoder.cpp:
(WebCore::TextEncoder::encode const):
* html/ImageData.cpp:
(WebCore::ImageData::ImageData):
* html/ImageData.h:
(WebCore::ImageData::data const):
(): Deleted.
* html/canvas/WebGL2RenderingContext.cpp:
(WebCore::WebGL2RenderingContext::bufferData):
(WebCore::WebGL2RenderingContext::bufferSubData):
(WebCore::WebGL2RenderingContext::getInternalformatParameter):
(WebCore::WebGL2RenderingContext::getParameter):
* html/canvas/WebGLRenderingContext.cpp:
(WebCore::WebGLRenderingContext::getParameter):
* html/canvas/WebGLRenderingContextBase.cpp:
(WebCore::WebGLRenderingContextBase::getUniform):
(WebCore::WebGLRenderingContextBase::getVertexAttrib):
(WebCore::WebGLRenderingContextBase::getWebGLFloatArrayParameter):
(WebCore::WebGLRenderingContextBase::getWebGLIntArrayParameter):
* platform/graphics/avfoundation/MediaPlayerPrivateAVFoundation.cpp:
(WebCore::MediaPlayerPrivateAVFoundation::extractKeyURIKeyIDAndCertificateFromInitData):
* platform/graphics/avfoundation/cf/CDMSessionAVFoundationCF.cpp:
(WebCore::CDMSessionAVFoundationCF::generateKeyRequest):
* platform/graphics/avfoundation/cf/MediaPlayerPrivateAVFoundationCF.cpp:
(WebCore::AVFWrapper::shouldWaitForLoadingOfResource):
* platform/graphics/avfoundation/objc/CDMSessionAVContentKeySession.mm:
(WebCore::CDMSessionAVContentKeySession::generateKeyRequest):
(WebCore::CDMSessionAVContentKeySession::releaseKeys):
(WebCore::CDMSessionAVContentKeySession::update):
(WebCore::CDMSessionAVContentKeySession::generateKeyReleaseMessage):
* platform/graphics/avfoundation/objc/CDMSessionAVFoundationObjC.mm:
(WebCore::CDMSessionAVFoundationObjC::generateKeyRequest):
* platform/graphics/avfoundation/objc/CDMSessionAVStreamSession.mm:
(WebCore::CDMSessionAVStreamSession::generateKeyRequest):
(WebCore::CDMSessionAVStreamSession::releaseKeys):
(WebCore::CDMSessionAVStreamSession::generateKeyReleaseMessage):
* platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.mm:
(WebCore::MediaPlayerPrivateAVFoundationObjC::shouldWaitForLoadingOfResource):
* platform/graphics/avfoundation/objc/MediaSampleAVFObjC.mm:
(WebCore::MediaSampleAVFObjC::getRGBAImageData const):
* platform/graphics/avfoundation/objc/SourceBufferPrivateAVFObjC.mm:
(WebCore::SourceBufferPrivateAVFObjC::didProvideContentKeyRequestInitializationDataForTrackID):
* platform/graphics/cairo/ImageBufferCairo.cpp:
(WebCore::getImageData):
* platform/graphics/cg/ImageBufferDataCG.cpp:
(WebCore::ImageBufferData::getData const):
* platform/graphics/filters/FEGaussianBlur.cpp:
(WebCore::FEGaussianBlur::platformApplySoftware):
* platform/graphics/filters/FilterEffect.cpp:
(WebCore::FilterEffect::unmultipliedResult):
(WebCore::FilterEffect::premultipliedResult):
(WebCore::FilterEffect::copyUnmultipliedResult):
(WebCore::FilterEffect::copyPremultipliedResult):
(WebCore::FilterEffect::createUnmultipliedImageResult):
(WebCore::FilterEffect::createPremultipliedImageResult):
* platform/graphics/win/ImageBufferDataDirect2D.cpp:
(WebCore::ImageBufferData::getData const):
* platform/mac/SerializedPlatformRepresentationMac.mm:
(WebCore::jsValueWithDictionaryInContext):
* platform/mock/mediasource/MockBox.cpp:
(WebCore::MockBox::peekType):
(WebCore::MockBox::peekLength):
(WebCore::MockTrackBox::MockTrackBox):
(WebCore::MockInitializationBox::MockInitializationBox):
(WebCore::MockSampleBox::MockSampleBox):
* rendering/shapes/Shape.cpp:
(WebCore::Shape::createRasterShape):
* testing/LegacyMockCDM.cpp:
(WebCore::initDataPrefix):
(WebCore::keyPrefix):
(WebCore::keyRequest):

Source/WebKit:

Use `tryCreate` or new `create`. If we use new `create`, we can use Ref<> and remove null check.

* Shared/API/c/WKString.cpp:
(WKStringCopyJSString):
* WebProcess/Automation/WebAutomationSessionProxy.cpp:
(WebKit::toJSValue):
(WebKit::callPropertyFunction):
(WebKit::WebAutomationSessionProxy::scriptObjectForFrame):
(WebKit::WebAutomationSessionProxy::evaluateJavaScriptFunction):
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::runScriptsInPDFDocument):

Source/WebKitLegacy/ios:

Use `tryCreate` or new `create`. If we use new `create`, we can use Ref<> and remove null check.

* WebView/WebPDFViewIOS.mm:
(-[WebPDFView finishedLoadingWithDataSource:]):
* WebView/WebPDFViewPlaceholder.mm:
(-[WebPDFViewPlaceholder _evaluateJSForDocument:]):

Source/WebKitLegacy/mac:

Use `tryCreate` or new `create`. If we use new `create`, we can use Ref<> and remove null check.

* WebView/WebPDFRepresentation.mm:
(-[WebPDFRepresentation finishedLoadingWithDataSource:]):

Tools:

* TestWebKitAPI/Tests/WebCore/mac/GPUBuffer.mm:
(TestWebKitAPI::TEST_F):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237009 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoRegression(PSON): Assertion hit under WebPageProxy::didNavigateWithNavigationData()
cdumez@apple.com [Wed, 10 Oct 2018 18:21:11 +0000 (18:21 +0000)]
Regression(PSON): Assertion hit under WebPageProxy::didNavigateWithNavigationData()
https://bugs.webkit.org/show_bug.cgi?id=190418
<rdar://problem/45059769>

Reviewed by Geoffrey Garen.

Source/WebKit:

When process swapping and "suspending" the previous WebProcess in a SuspendedPageProxy,
we need to keep around the main frame's ID that still exists on in this process. This
is needed so that we can re-create a UI-side WebFrameProxy for the WebFrame that exists
in the WebProcess, if we ever swap back to this suspended process (see login in
WebPageProxy::swapToWebProcess()).

The bug was that the main frame ID was stored on the WebPageProxy via m_mainFrameID instead of the
SuspendedPageProxy. This means that m_mainFrameID would get overriden when navigating in the new
WebProcess with the value 1 (because WebFrame identifiers start at 1 and are per-WebProcess).
This would lead to us constructing a WebFrameProxy with the wrong frame identifier in
WebPageProxy::swapToWebProcess(), which would override an existing unrelated WebFrame in the
WebProcessProxy's HashMap of frames. This would lead to crashes later on as the WebFrame
would not be associated to the WebPageProxy we expect.

* UIProcess/SuspendedPageProxy.cpp:
(WebKit::SuspendedPageProxy::SuspendedPageProxy):
* UIProcess/SuspendedPageProxy.h:
(WebKit::SuspendedPageProxy::mainFrameID const):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::maybeCreateSuspendedPage):
(WebKit::WebPageProxy::swapToWebProcess):
(WebKit::WebPageProxy::continueNavigationInNewProcess):
(WebKit::WebPageProxy::didCreateMainFrame):
* UIProcess/WebPageProxy.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::suspendWebPageProxy):
* UIProcess/WebProcessProxy.h:

Tools:

Add API test coverage.

* TestWebKitAPI/Tests/WebKitCocoa/ProcessSwapOnNavigation.mm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237008 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago[iOS] Compare input string to UIKeyInput constants using string comparison instead...
dbates@webkit.org [Wed, 10 Oct 2018 18:20:08 +0000 (18:20 +0000)]
[iOS] Compare input string to UIKeyInput constants using string comparison instead of pointer comparison
https://bugs.webkit.org/show_bug.cgi?id=190432

Reviewed by Tim Horton.

Pointer comparision is brittle. We should be more forgiving and perform string comparision
of an input string to a UIKeyInput constant.

* platform/ios/WebEvent.mm:
(normalizedStringWithAppKitCompatibilityMapping):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237007 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoWeb Inspector: indent all network entries when "Group by Node" is checked
drousso@apple.com [Wed, 10 Oct 2018 18:01:25 +0000 (18:01 +0000)]
Web Inspector: indent all network entries when "Group by Node" is checked
https://bugs.webkit.org/show_bug.cgi?id=190388

Reviewed by Timothy Hatcher.

* UserInterface/Views/NetworkTableContentView.js:
(WI.NetworkTableContentView.prototype._populateNameCell):
(WI.NetworkTableContentView.prototype._handleGroupByDOMNodeCheckedDidChange):

* UserInterface/Views/NetworkTableContentView.css:
(.network-table.grouped .data-container .cell.name): Added.
(.network-table.grouped .data-container .cell:not(.parent).name): Added.
(.network-table.grouped .data-container .cell.child.name): Added.
(.network-table .cell.grouped-by-node.name): Deleted.
(body[dir=ltr] .network-table .cell.grouped-by-node.name): Deleted.
(body[dir=rtl] .network-table .cell.grouped-by-node.name): Deleted.
Apply a padding to all nodes when the `WI.Table` is grouped.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237006 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoDo domain prewarming for processes for new tabs
antti@apple.com [Wed, 10 Oct 2018 17:41:55 +0000 (17:41 +0000)]
Do domain prewarming for processes for new tabs
https://bugs.webkit.org/show_bug.cgi?id=190425

Reviewed by Chris Dumez.

We currently only do domain prewarming for navigation process swaps.

* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::processForNavigationInternal):

Do domain prewarming when using an existing but so far unused process.

(WebKit::WebProcessPool::tryPrewarmWithDomainInformation):

Factor into a function.

* UIProcess/WebProcessPool.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::close):

This didn't usually get invoked.

(WebKit::WebPage::didReceivePolicyDecision):
(WebKit::WebPage::didFinishLoad):

Update prewarm information when a top level load finishes.

* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::sendPrewarmInformation):
* WebProcess/WebProcess.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237004 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoUnreviewed, rolling out r236802.
cdumez@apple.com [Wed, 10 Oct 2018 16:30:07 +0000 (16:30 +0000)]
Unreviewed, rolling out r236802.

Working on getting the HTML spec updated instead
(https://github.com/whatwg/html/pull/4079)

Reverted changeset:

"Passing noopener=NOOPENER to window.open() should cause the
new window to not have an opener"
https://bugs.webkit.org/show_bug.cgi?id=190251
https://trac.webkit.org/changeset/236802

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237002 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago[JSCOnly Add an armv7 JSCOnly EWS that runs tests
guijemont@igalia.com [Wed, 10 Oct 2018 13:33:03 +0000 (13:33 +0000)]
[JSCOnly Add an armv7 JSCOnly EWS that runs tests
https://bugs.webkit.org/show_bug.cgi?id=190191

Reviewed by Michael Catanzaro.

* QueueStatusServer/config/queues.py:
* Scripts/webkitpy/common/config/ews.json:
* Scripts/webkitpy/common/config/ports.py:
(DeprecatedPort.run_javascriptcore_tests_command):
Add the possibility to pass arguments to run-javascriptcore-tests with
the JSCTESTS_OPTIONS environment variable
* Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
(test_ews_name):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237001 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoSkip JSC test stress/sampling-profiler-richards.js on armv7/linux
guijemont@igalia.com [Wed, 10 Oct 2018 11:39:21 +0000 (11:39 +0000)]
Skip JSC test stress/sampling-profiler-richards.js on armv7/linux
https://bugs.webkit.org/show_bug.cgi?id=190426

Unreviewed gardening.

* stress/sampling-profiler-richards.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@237000 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoXMLHttpRequest should use reportExtraMemoryAllocated/reportExtraMemoryVisited instead...
yusukesuzuki@slowstart.org [Wed, 10 Oct 2018 08:48:57 +0000 (08:48 +0000)]
XMLHttpRequest should use reportExtraMemoryAllocated/reportExtraMemoryVisited instead of deprecatedReportExtraMemory
https://bugs.webkit.org/show_bug.cgi?id=190279

Reviewed by Ryosuke Niwa.

This patch switches deprecatedReportExtraMemory to reportExtraMemoryAllocated/reportExtraMemoryVisited
in XMLHttpRequest. We report extra memory allocation when the readyState becomes DONE. And memoryCost
function returns the memory cost which is based on the readyState and m_responseBuilder.
We annotate XMLHttpRequest with ReportExtraMemoryCost to use reportExtraMemoryVisited automatically with
memoryCost() function.

* xml/XMLHttpRequest.cpp:
(WebCore::XMLHttpRequest::changeState):
(WebCore::XMLHttpRequest::abort):
(WebCore::XMLHttpRequest::internalAbort):
(WebCore::XMLHttpRequest::networkErrorTimerFired):
(WebCore::XMLHttpRequest::memoryCost const):
(WebCore::XMLHttpRequest::didFinishLoading):
(WebCore::XMLHttpRequest::didReachTimeout):
(WebCore::XMLHttpRequest::dropProtection): Deleted.
* xml/XMLHttpRequest.h:
* xml/XMLHttpRequest.idl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236999 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoRemove the frames() timing function
graouts@webkit.org [Wed, 10 Oct 2018 06:55:24 +0000 (06:55 +0000)]
Remove the frames() timing function
https://bugs.webkit.org/show_bug.cgi?id=190034
<rdar://problem/44827544>

Reviewed by Dean Jackson.

LayoutTests/imported/w3c:

* web-platform-tests/css-timing-1/frames-timing-functions-output-expected.txt:
* web-platform-tests/css-timing-1/frames-timing-functions-syntax-expected.txt:
* web-platform-tests/web-animations/animation-model/keyframe-effects/effect-value-transformed-distance-expected.txt:
* web-platform-tests/web-animations/interfaces/AnimationEffectTiming/easing-expected.txt:
* web-platform-tests/web-animations/timing-model/time-transformations/transformed-progress-expected.txt:

Source/WebCore:

* css/CSSComputedStyleDeclaration.cpp:
(WebCore::createTimingFunctionValue):
* css/CSSTimingFunctionValue.cpp:
(WebCore::CSSFramesTimingFunctionValue::customCSSText const): Deleted.
(WebCore::CSSFramesTimingFunctionValue::equals const): Deleted.
* css/CSSTimingFunctionValue.h:
* css/CSSToStyleMap.cpp:
(WebCore::CSSToStyleMap::mapAnimationTimingFunction):
* css/CSSValue.cpp:
(WebCore::CSSValue::equals const):
(WebCore::CSSValue::cssText const):
(WebCore::CSSValue::destroy):
* css/CSSValue.h:
(WebCore::CSSValue::isStepsTimingFunctionValue const):
(WebCore::CSSValue::isFramesTimingFunctionValue const): Deleted.
* css/CSSValueKeywords.in:
* css/parser/CSSPropertyParser.cpp:
(WebCore::consumeAnimationTimingFunction):
(WebCore::consumeFrames): Deleted.
* platform/animation/TimingFunction.cpp:
(WebCore::operator<<):
(WebCore::TimingFunction::transformTime const):
(WebCore::TimingFunction::createFromCSSValue):
* platform/animation/TimingFunction.h:
(WebCore::TimingFunction::isStepsTimingFunction const):
(WebCore::TimingFunction::isFramesTimingFunction const): Deleted.
* platform/graphics/ca/GraphicsLayerCA.cpp:
(WebCore::GraphicsLayerCA::animationCanBeAccelerated const):
(WebCore::animationHasFramesTimingFunction): Deleted.

Source/WebKit:

* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<FramesTimingFunction>::encode): Deleted.
(IPC::ArgumentCoder<FramesTimingFunction>::decode): Deleted.
* Shared/WebCoreArgumentCoders.h:
* WebProcess/WebPage/RemoteLayerTree/PlatformCAAnimationRemote.mm:
(WebKit::PlatformCAAnimationRemote::Properties::encode const):
(WebKit::PlatformCAAnimationRemote::Properties::decode):

LayoutTests:

* legacy-animation-engine/transitions/frames-timing-function-expected.txt:
* legacy-animation-engine/transitions/transitions-parsing-expected.txt:
* transitions/frames-timing-function-expected.txt:
* transitions/transitions-parsing-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236998 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoPerformance dashboard should show added iterations due failed build requests.
dewei_zhu@apple.com [Wed, 10 Oct 2018 03:43:47 +0000 (03:43 +0000)]
Performance dashboard should show added iterations due failed build requests.
https://bugs.webkit.org/show_bug.cgi?id=190419

Reviewed by Ryosuke Niwa.

Added UI to show added build requests due to previous failed build requests.
Updated the retry/bisect button to default to initial repetition count when test group is created.

* public/v3/pages/analysis-task-page.js: Added a div to show added build requests.
Retry and bisect button should default to initial repetion count.
(AnalysisTaskTestGroupPane.prototype._renderCurrentTestGroup):
(AnalysisTaskTestGroupPane.cssTemplate):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236997 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoManifestGenerator shouldn't need more than 1GB of memory or run for 30 seconds
rniwa@webkit.org [Wed, 10 Oct 2018 03:15:39 +0000 (03:15 +0000)]
ManifestGenerator shouldn't need more than 1GB of memory or run for 30 seconds
https://bugs.webkit.org/show_bug.cgi?id=190393

Reviewed by Antti Koivisto and unofficially reviewed by Dewei Zhu.

This patch reduces the runtime of /api/manifest from 13s to 7s and reduces the memory requirement from
1GB to 400MB for the internal dashboard in my local testing.

The biggest perf win comes from avoid running a complex query over test_configurations to compute
the latest modified date across different test configuration types ("current" vs. "baseline").
Instead, we now fetch the entire table row by row and compute the latest modified date in memory.

Also call intval in many more places to avoid generating double quotes, which is a pretty significant
proportion of the JSON file size at this point.

Finally, use references in more places to avoid deep copying of arrays.

* public/include/manifest-generator.php:
(ManifestGenerator::generate): Skip the generation of "dashboard" since it's only used by v1 UI.
(ManifestGenerator::tests): Don't copy each row.
(ManifestGenerator::metrics): Ditto.
(ManifestGenerator::platforms): Implement the aforementioned optimization. Instead of grouping
test configurations for a given metric and platform together, fetch them all and do in-memory
processing in PHP. Avoid more copying as well.
(ManifestGenerator::repositories): Avoid more copying.
(ManifestGenerator::bug_trackers): Ditto.
(ManifestGenerator::fetch_triggerables): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236996 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoWeb Inspector: show redirect requests in Network and Timelines tabs
drousso@apple.com [Wed, 10 Oct 2018 02:49:52 +0000 (02:49 +0000)]
Web Inspector: show redirect requests in Network and Timelines tabs
https://bugs.webkit.org/show_bug.cgi?id=150005
<rdar://problem/5378164>

Reviewed by Joseph Pecoraro.

Source/JavaScriptCore:

* inspector/protocol/Network.json:
Add missing fields to `ResourceTiming`.

Source/WebCore:

Updated existing test http/tests/inspector/network/resource-timing.html.

* inspector/agents/InspectorNetworkAgent.cpp:
(WebCore::InspectorNetworkAgent::buildObjectForTiming):
(WebCore::InspectorNetworkAgent::didFinishLoading):
Add missing fields for `Network.types.ResourceTiming`.

Source/WebInspectorUI:

* Localizations/en.lproj/localizedStrings.js:
* UserInterface/Views/Variables.css:
* UserInterface/Main.html:
* UserInterface/Test.html:

* UserInterface/Controllers/NetworkManager.js:
(WI.NetworkManager.prototype.resourceRequestWillBeSent):

* UserInterface/Models/Resource.js:
(WI.Resource):
(WI.Resource.prototype.get redirects): Added.
(WI.Resource.prototype.get lastRedirectReceivedTimestamp):
(WI.Resource.prototype.updateForRedirectResponse):
Save each redirect in an array instead of just remembering the last timestamp.

* UserInterface/Models/ResourceTimingData.js:
(WI.ResourceTimingData):
(WI.ResourceTimingData.fromPayload.offsetToTimestamp):
(WI.ResourceTimingData.fromPayload):
(WI.ResourceTimingData.prototype.get redirectStart): Added.
(WI.ResourceTimingData.prototype.get redirectEnd): Added.
(WI.ResourceTimingData.prototype.get fetchStart): Added.
Add missing fields for `Network.types.ResourceTiming`.

* UserInterface/Models/Redirect.js: Added.
(WI.Redirect):
(WI.Redirect.prototype.get url):
(WI.Redirect.prototype.get requestMethod):
(WI.Redirect.prototype.get requestHeaders):
(WI.Redirect.prototype.get responseStatusCode):
(WI.Redirect.prototype.get responseStatusText):
(WI.Redirect.prototype.get responseHeaders):
(WI.Redirect.prototype.get timestamp):
(WI.Redirect.prototype.get urlComponents):

* UserInterface/Views/ResourceHeadersContentView.js:
(WI.ResourceHeadersContentView):
(WI.ResourceHeadersContentView.prototype.initialLayout):
(WI.ResourceHeadersContentView.prototype.layout):
(WI.ResourceHeadersContentView.prototype._refreshRedirectHeadersSections): Added.
(WI.ResourceHeadersContentView.prototype._resourceRequestHeadersDidChange):
* UserInterface/Views/ResourceHeadersContentView.css:
(body[dir] .resource-headers > section.summary > .details): Added.
(body[dir] .resource-headers > section:matches(.redirect, .headers) > .details): Added.
(.resource-headers .details .key):
(.resource-headers .summary .key):
(body[dir] .resource-headers > section > .details): Deleted.
(body[dir] .resource-headers > section.headers > .details): Deleted.
(.resource-headers .value): Deleted.
Add a request/response header section for each redirect.

* UserInterface/Views/NetworkTableContentView.js:
(WI.NetworkTableContentView.prototype._populateWaterfallGraph.appendBlock):
(WI.NetworkTableContentView.prototype._populateWaterfallGraph):
(WI.NetworkTableContentView.prototype._checkURLFilterAgainstResource):
(WI.NetworkTableContentView.prototype._waterfallPopoverContentForResource):
* UserInterface/Views/NetworkTableContentView.css:
(.waterfall .block.redirect): Added.
(.waterfall .block.queue):
* UserInterface/Views/ResourceTimelineDataGridNode.js:
(WI.ResourceTimelineDataGridNode.prototype._mouseoverRecordBar):
* UserInterface/Views/ResourceTimingBreakdownView.js:
(WI.ResourceTimingBreakdownView.prototype.initialLayout):
Add timeline/waterfall entries for total redirect time.

LayoutTests:

* http/tests/inspector/network/resource-timing-expected.txt:
* http/tests/inspector/network/resource-timing.html:

* http/tests/inspector/network/resources/delay.php: Added.
* http/tests/inspector/network/resources/redirect.php: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236995 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoASSERTION FAILED: m_cachesLocks.contains(sessionID) in WebKit::CacheStorageEngineConn...
youenn@apple.com [Wed, 10 Oct 2018 00:31:33 +0000 (00:31 +0000)]
ASSERTION FAILED: m_cachesLocks.contains(sessionID) in WebKit::CacheStorageEngineConnection::dereference(PAL::SessionID, unsigned long long)
https://bugs.webkit.org/show_bug.cgi?id=190373
<rdar://problem/45106402>

Reviewed by Alex Christensen.

This assertion can be hit if:
- WebProcess has a DOMCache
- NetworkProcess crashes
- WebProcess navigates and dereference DOMCache.
Remove this assertion as it is hit when a test is using testRunner.terminateNetworkProcess().

* NetworkProcess/cache/CacheStorageEngineConnection.cpp:
(WebKit::CacheStorageEngineConnection::dereference):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236994 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoAllow run-javascriptcore-tests to distribute tests between devices
guijemont@igalia.com [Wed, 10 Oct 2018 00:28:14 +0000 (00:28 +0000)]
Allow run-javascriptcore-tests to distribute tests between devices
https://bugs.webkit.org/show_bug.cgi?id=190190

Reviewed by Michael Catanzaro.

* Scripts/run-jsc-stress-tests:
* Scripts/webkitruby/jsc-stress-test-writer-default.rb:
* Scripts/webkitruby/jsc-stress-test-writer-ruby.rb:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236993 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoUpdate WHLSL to Metal tester with semantics
dino@apple.com [Wed, 10 Oct 2018 00:23:56 +0000 (00:23 +0000)]
Update WHLSL to Metal tester with semantics
https://bugs.webkit.org/show_bug.cgi?id=190416
<rdar://problem/45145139>

Reviewed by Myles Maxfield.

A few small changes to the WHLSL to Metal tools:
- Make it compile by adding some missing JS files :)
- Add semantics to the default shader
- Add FIXMEs to the other shaders
- Create an in-browser test for the WHLSLToMetal codepath
- Fix typos in the semantics checker

* WebGPUShadingLanguageRI/Checker.js:
(Checker.prototype._checkSemantics.checkSemanticTypes):
(Checker.prototype._checkSemantics.checkSemanticForShaderType):
* WebGPUShadingLanguageRI/Metal/WHLSL Tests/Core/Compiler.m:
* WebGPUShadingLanguageRI/Metal/WHLSL Tests/Core/Demo shaders/Default.whlsl:
* WebGPUShadingLanguageRI/Metal/WHLSL Tests/Core/Demo shaders/Julia.whlsl:
* WebGPUShadingLanguageRI/Metal/WHLSL Tests/Core/Demo shaders/Mandelbrot.whlsl:
* WebGPUShadingLanguageRI/Metal/WHLSL Tests/WHLSL Tests.xcodeproj/project.pbxproj:
* WebGPUShadingLanguageRI/Metal/WhlslToMsl.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236992 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoREGRESSION(r234620): SVGLangSpace::svgAttributeChanged() should invalidate the render...
said@apple.com [Tue, 9 Oct 2018 23:59:52 +0000 (23:59 +0000)]
REGRESSION(r234620): SVGLangSpace::svgAttributeChanged() should invalidate the renderer of the SVGGeometryElement descendant only
https://bugs.webkit.org/show_bug.cgi?id=190411

Reviewed by Simon Fraser.

Source/WebCore:

Test: svg/dynamic-updates/SVGStopElement-dom-xml-lang-attrr.html

When changing the attributes of the SVGLangSpace, we should invalidate
the renderer of the SVGGeometryElement descendant only. Renderer of other
elements, like SVGStopElement, should not be invalidated because they do
not have geometry and they can be used as resources for drawing another
SVGGeometryElement.

* svg/SVGElement.h:
(WebCore::SVGElement::isSVGGeometryElement const):
* svg/SVGGeometryElement.h:
(isType):
* svg/SVGLangSpace.cpp:
(WebCore::SVGLangSpace::svgAttributeChanged):

LayoutTests:

* svg/dynamic-updates/SVGStopElement-dom-xml-lang-attrr-expected.txt: Added.
* svg/dynamic-updates/SVGStopElement-dom-xml-lang-attrr.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236991 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoCrash when closing WKWebView during enter fullscreen animation
jer.noble@apple.com [Tue, 9 Oct 2018 23:46:12 +0000 (23:46 +0000)]
Crash when closing WKWebView during enter fullscreen animation
https://bugs.webkit.org/show_bug.cgi?id=190412

Reviewed by Brian Burg.

Source/WebKit:

We need a valid _webView if, during -[NSWindow close], we catch a notification saying we failed to enter
fullscreen and try to replace the placeholder with the actual _webView.

* UIProcess/mac/WKFullScreenWindowController.mm:
(-[WKFullScreenWindowController close]):

Tools:

* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/Tests/WebKitCocoa/CloseWebViewDuringEnterFullscreen.mm: Added.
(-[CloseWebViewDuringEnterFullscreenUIDelegate _webViewWillEnterFullscreen:]):
(-[CloseWebViewDuringEnterFullscreenUIDelegate _webViewWillExitFullscreen:]):
(TestWebKitAPI::TEST):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236990 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoREGRESSION(r231663) loading in hidden WKWebViews stalls because WebProcess is suspended
achristensen@apple.com [Tue, 9 Oct 2018 23:40:21 +0000 (23:40 +0000)]
REGRESSION(r231663) loading in hidden WKWebViews stalls because WebProcess is suspended
https://bugs.webkit.org/show_bug.cgi?id=190417
<rdar://problem/43391014>

Reviewed by Chris Dumez.

* UIProcess/Cocoa/NavigationState.mm:
(WebKit::NavigationState::didChangeIsLoading):
Revert r321663, but keep the change to TestWKWebView.mm because that made the tests more reliable.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236989 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoAllow behavior when the parent process IPC::Connection closes to be overridden by...
cdumez@apple.com [Tue, 9 Oct 2018 23:25:46 +0000 (23:25 +0000)]
Allow behavior when the parent process IPC::Connection closes to be overridden by ChildProcess subclasses
https://bugs.webkit.org/show_bug.cgi?id=190294

Reviewed by Geoffrey Garen.

Allow behavior when the parent process IPC::Connection closes to be overridden by ChildProcess subclasses.
This will be useful to allow the NetworkProcess to not exit if it still has pending downloads.

* NetworkProcess/NetworkProcess.cpp:
(WebKit::callExitSoon):
(WebKit::NetworkProcess::initializeConnection):
* NetworkProcess/NetworkProcess.h:
* PluginProcess/PluginProcess.cpp:
(WebKit::callExit):
(WebKit::PluginProcess::initializeConnection):
* PluginProcess/PluginProcess.h:
* Shared/ChildProcess.cpp:
(WebKit::ChildProcess::didClose):
(WebKit::ChildProcess::initialize):
(WebKit::callExitNow): Deleted.
(WebKit::callExitSoon): Deleted.
* Shared/ChildProcess.h:
(WebKit::ChildProcess::shouldCallExitWhenConnectionIsClosed const): Deleted.
* WebProcess/WebProcess.cpp:
(WebKit::callExit):
(WebKit::WebProcess::initializeConnection):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236988 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoAnchor target should be ignored on activation when the download attribute is set
cdumez@apple.com [Tue, 9 Oct 2018 23:21:18 +0000 (23:21 +0000)]
Anchor target should be ignored on activation when the download attribute is set
https://bugs.webkit.org/show_bug.cgi?id=190408

Reviewed by Geoffrey Garen.

Source/WebCore:

Anchor target should be ignored on activation when the download attribute is set:
- https://html.spec.whatwg.org/#the-a-element:downloading-hyperlinks-2

When the download attribute is set, we should use the "download the hyperlink" algorithm [1]
instead of the "follow the hyperlink" [2] algorithm.

Note that the "download the hyperlink" triggers a download and ignores the target attribute
entirely.

This is important as an anchor element with target=_blank and the download attribute set may
fail because of Safari's popup blocker if we do not disregard the anchor target.

[1] https://html.spec.whatwg.org/#downloading-hyperlinks
[2] https://html.spec.whatwg.org/#following-hyperlinks-2

Tests: fast/dom/HTMLAnchorElement/anchor-file-blob-download-blank-base-target-popup-not-allowed.html
       fast/dom/HTMLAnchorElement/anchor-file-blob-download-blank-target-popup-not-allowed.html

* loader/FrameLoader.cpp:
(WebCore::FrameLoader::loadURL):

Tools:

Update WebKitTestRunner to block popups by default and add support for testRunner.setCanOpenWindows()
for tests that need popups. This aligns WebKitTestRunner with DumpRenderTree and allows testing
popup blocking.

* WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
* WebKitTestRunner/InjectedBundle/TestRunner.cpp:
(WTR::TestRunner::setCanOpenWindows):
* WebKitTestRunner/InjectedBundle/TestRunner.h:
* WebKitTestRunner/TestController.cpp:
(WTR::TestController::createOtherPage):
* WebKitTestRunner/TestController.h:
* WebKitTestRunner/TestInvocation.cpp:
(WTR::TestInvocation::didReceiveSynchronousMessageFromInjectedBundle):
* WebKitTestRunner/TestInvocation.h:

LayoutTests:

* fast/dom/HTMLAnchorElement/anchor-file-blob-download-blank-base-target-popup-not-allowed-expected.txt: Added.
* fast/dom/HTMLAnchorElement/anchor-file-blob-download-blank-base-target-popup-not-allowed.html: Added.
* fast/dom/HTMLAnchorElement/anchor-file-blob-download-blank-target-popup-not-allowed-expected.txt: Added.
* fast/dom/HTMLAnchorElement/anchor-file-blob-download-blank-target-popup-not-allowed.html: Added.
Add layout test coverage.

* http/tests/download/anchor-load-after-download.html:
* http/tests/workers/service/resources/registration-task-queue-scheduling-1.js:
Update existing WK2-only layout tests to call testRunner.setCanOpenWindows() since they rely on
opening a new window. We failed to notice the issue because these tests are not run on WK1.

* platform/ios-wk2/TestExpectations:
* platform/mac-wk1/TestExpectations:
* platform/win/TestExpectations:
Skip new tests on WebKit1 since the download attribute is only supported on WebKit2.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236987 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoWeb Inspector: Canvas Tab: grayed out Record button in navigator is nearly invisible
drousso@apple.com [Tue, 9 Oct 2018 22:31:55 +0000 (22:31 +0000)]
Web Inspector: Canvas Tab: grayed out Record button in navigator is nearly invisible
https://bugs.webkit.org/show_bug.cgi?id=190365
<rdar://problem/45097739>

Reviewed by Brian Burg.

* UserInterface/Views/CanvasSidebarPanel.css:
(@media (prefers-dark-interface)): Added.
(.sidebar > .panel.navigation.canvas > .navigation-bar > .item.record-start-stop.disabled): Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236986 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago[WPE][GTK] Complex text crashes with harfbuzz 1.8.8
mcatanzaro@igalia.com [Tue, 9 Oct 2018 21:59:12 +0000 (21:59 +0000)]
[WPE][GTK] Complex text crashes with harfbuzz 1.8.8
https://bugs.webkit.org/show_bug.cgi?id=190409

Reviewed by Žan Doberšek.

We discovered that harfbuzz 1.8.8 breaks WebKit by changing the ownership semantics of
hb_icu_get_unicode_funcs() from (transfer full) to (transfer none). Our code was actually
expecting (transfer none), so it's not immediately clear to me why it's crashing now, but
doesn't matter as Behdad recommends removing this line of code because it hasn't been
been needed for many years.

This should be covered by all our complex text tests if the bots were upgraded to the newer
harfbuzz.

* platform/graphics/harfbuzz/ComplexTextControllerHarfBuzz.cpp:
(WebCore::ComplexTextController::collectComplexTextRunsForCharacters):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236985 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoWebDriver: thrown ObjC exception under -[WKFullScreenWindowController windowDidFailTo...
jer.noble@apple.com [Tue, 9 Oct 2018 21:45:14 +0000 (21:45 +0000)]
WebDriver: thrown ObjC exception under -[WKFullScreenWindowController windowDidFailToEnterFullScreen:] when session is terminated
https://bugs.webkit.org/show_bug.cgi?id=190098
<rdar://problem/42822671>

Reviewed by Brian Burg.

Source/WebKit:

Forcibly exit fullscreen in resetState(), while the VideoFullscreenManagerProxy is still valid.

* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::resetState):

Tools:

* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/Tests/WebKitCocoa/CloseWebViewAfterEnterFullscreen.mm: Added.
(-[CloseWebViewAfterEnterFullscreenUIDelegate _webViewDidEnterFullscreen:]):
(-[CloseWebViewAfterEnterFullscreenUIDelegate _webViewDidExitFullscreen:]):
(TestWebKitAPI::TEST):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236984 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago[iOS] Cleanup EventHandler::passSubframeEventToSubframe()
dbates@webkit.org [Tue, 9 Oct 2018 21:31:35 +0000 (21:31 +0000)]
[iOS] Cleanup EventHandler::passSubframeEventToSubframe()
https://bugs.webkit.org/show_bug.cgi?id=190390

Reviewed by Wenson Hsieh.

Unindent case statements in switch block and use more auto.

* page/ios/EventHandlerIOS.mm:
(WebCore::EventHandler::passSubframeEventToSubframe):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236983 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoCleanup WebEvent.mm
dbates@webkit.org [Tue, 9 Oct 2018 21:30:53 +0000 (21:30 +0000)]
Cleanup WebEvent.mm
https://bugs.webkit.org/show_bug.cgi?id=190391

Reviewed by Wenson Hsieh.

Unindent case statements in switch blocks.

* platform/ios/WebEvent.mm:
(-[WebEvent _typeDescription]):
(-[WebEvent _modiferFlagsDescription]):
(-[WebEvent _touchPhaseDescription:]):
(-[WebEvent _eventDescription]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236982 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago[iOS] Ignore some deprecated declarations
dbates@webkit.org [Tue, 9 Oct 2018 21:29:31 +0000 (21:29 +0000)]
[iOS] Ignore some deprecated declarations
https://bugs.webkit.org/show_bug.cgi?id=190371

Reviewed by Dan Bernstein.

* UIProcess/ios/WKActionSheet.mm:
(-[WKActionSheet init]):
(-[WKActionSheet presentSheet:]):
* UIProcess/ios/WKActionSheetAssistant.mm:
(-[WKActionSheetAssistant presentSheet]):
* UIProcess/ios/WKContentViewInteraction.mm:
* UIProcess/ios/forms/WKFormSelectPicker.mm:
(-[WKMultipleSelectPicker initWithView:]):
* UIProcess/ios/forms/WKFormSelectPopover.mm:
(-[WKSelectTableViewController initWithView:hasGroups:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236981 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoUnreviewed, skip fast/dom/HTMLAnchorElement/anchor-file-blob-download-blank-target...
cdumez@apple.com [Tue, 9 Oct 2018 21:12:54 +0000 (21:12 +0000)]
Unreviewed, skip fast/dom/HTMLAnchorElement/anchor-file-blob-download-blank-target.html on iOS.

The download attribute is not supported on iOS yet.

* platform/ios-wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236980 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoResolve inset properties to computed style when there is overconstraintment
commit-queue@webkit.org [Tue, 9 Oct 2018 20:11:04 +0000 (20:11 +0000)]
Resolve inset properties to computed style when there is overconstraintment
https://bugs.webkit.org/show_bug.cgi?id=188711

Patch by Oriol Brufau <obrufau@igalia.com> on 2018-10-09
Reviewed by Manuel Rego Casasnovas.

This patch makes WebKit behave closer to Blink. Specifically,
 - In overconstrained relative or absolute positioning, inset properties resolve
   to the computed value (absolutizing percentages) instead of to the used value.
 - In fixed positioning, the resolved value of non-'auto' values is no longer
   increased by the border of the containg block.

LayoutTests/imported/w3c:

Some tests still have failures due to
 - https://bugs.webkit.org/show_bug.cgi?id=189513
 - https://bugs.webkit.org/show_bug.cgi?id=189518
 - https://bugs.webkit.org/show_bug.cgi?id=189549

This patch can slighlty alter the resolved value if it's a long decimal number,
that's why test expectations for the timing functions test changed.

* web-platform-tests/css-timing-1/cubic-bezier-timing-functions-output-expected.txt:
* web-platform-tests/css/cssom/getComputedStyle-insets-absolute-expected.txt: Added.
* web-platform-tests/css/cssom/getComputedStyle-insets-absolute.html: Added.
* web-platform-tests/css/cssom/getComputedStyle-insets-fixed-expected.txt: Added.
* web-platform-tests/css/cssom/getComputedStyle-insets-fixed.html: Added.
* web-platform-tests/css/cssom/getComputedStyle-insets-nobox-expected.txt: Added.
* web-platform-tests/css/cssom/getComputedStyle-insets-nobox.html: Added.
* web-platform-tests/css/cssom/getComputedStyle-insets-relative-expected.txt: Added.
* web-platform-tests/css/cssom/getComputedStyle-insets-relative.html: Added.
* web-platform-tests/css/cssom/getComputedStyle-insets-static-expected.txt: Added.
* web-platform-tests/css/cssom/getComputedStyle-insets-static.html: Added.
* web-platform-tests/css/cssom/getComputedStyle-insets-sticky-expected.txt: Added.
* web-platform-tests/css/cssom/getComputedStyle-insets-sticky.html: Added.
* web-platform-tests/css/cssom/support/getComputedStyle-insets.js: Added.
(serialize):
(wmName):
(checkStyle):
(runTestsWithWM):
(export.runTests):

Source/WebCore:

This patch can slighlty alter the resolved value if it's a long decimal number.

Tests: imported/w3c/web-platform-tests/css/cssom/getComputedStyle-insets-absolute.html
       imported/w3c/web-platform-tests/css/cssom/getComputedStyle-insets-fixed.html
       imported/w3c/web-platform-tests/css/cssom/getComputedStyle-insets-nobox.html
       imported/w3c/web-platform-tests/css/cssom/getComputedStyle-insets-relative.html
       imported/w3c/web-platform-tests/css/cssom/getComputedStyle-insets-static.html
       imported/w3c/web-platform-tests/css/cssom/getComputedStyle-insets-sticky.html

* css/CSSComputedStyleDeclaration.cpp:
(WebCore::positionOffsetValue):
* rendering/RenderBox.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236979 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago[WebAuthN] Import CTAP HID message and packet structure from Chromium
jiewen_tan@apple.com [Tue, 9 Oct 2018 19:32:41 +0000 (19:32 +0000)]
[WebAuthN] Import CTAP HID message and packet structure from Chromium
https://bugs.webkit.org/show_bug.cgi?id=189289
<rdar://problem/44120310>

Reviewed by Brent Fulgham.

Source/WebCore:

This patch imports CTAP HID message and packet structure:
https://fidoalliance.org/specs/fido-v2.0-ps-20170927/fido-client-to-authenticator-protocol-v2.0-ps-20170927.html#message-and-packet-structure
from Chromium. With this library, WebKit can now turn binaries into messages that CTAP devices could understand.
This patch contains the following Chromium files and modifies them to fit into WebKit:
https://cs.chromium.org/chromium/src/device/fido/fido_constants.cc?l=1&rcl=1efcfbeaf4e4cedf58716e1982b5702770571a75
https://cs.chromium.org/chromium/src/device/fido/fido_constants.h?l=1&rcl=1efcfbeaf4e4cedf58716e1982b5702770571a75
https://cs.chromium.org/chromium/src/device/fido/hid/fido_hid_message.cc?l=1&rcl=387f3725de2842e0e6b7175a9b2ed472b0cf781a
https://cs.chromium.org/chromium/src/device/fido/hid/fido_hid_message.h?rcl=1efcfbeaf4e4cedf58716e1982b5702770571a75
https://cs.chromium.org/chromium/src/device/fido/hid/fido_hid_packet.cc?rcl=1efcfbeaf4e4cedf58716e1982b5702770571a75
https://cs.chromium.org/chromium/src/device/fido/hid/fido_hid_packet.h?rcl=1efcfbeaf4e4cedf58716e1982b5702770571a75
https://cs.chromium.org/chromium/src/device/fido/hid/fido_hid_message_unittest.cc?rcl=1efcfbeaf4e4cedf58716e1982b5702770571a75

Covered by API tests.

* Modules/webauthn/fido/FidoConstants.cpp: Added.
(fido::isFidoHidDeviceCommand):
* Modules/webauthn/fido/FidoConstants.h: Added.
* Modules/webauthn/fido/FidoHidMessage.cpp: Added.
(fido::FidoHidMessage::create):
(fido::FidoHidMessage::createFromSerializedData):
(fido::FidoHidMessage::messageComplete const):
(fido::FidoHidMessage::getMessagePayload const):
(fido::FidoHidMessage::popNextPacket):
(fido::FidoHidMessage::addContinuationPacket):
(fido::FidoHidMessage::numPackets const):
(fido::FidoHidMessage::FidoHidMessage):
* Modules/webauthn/fido/FidoHidMessage.h: Added.
* Modules/webauthn/fido/FidoHidPacket.cpp: Added.
(fido::FidoHidPacket::FidoHidPacket):
(fido::FidoHidInitPacket::createFromSerializedData):
(fido::FidoHidInitPacket::FidoHidInitPacket):
(fido::FidoHidInitPacket::getSerializedData const):
(fido::FidoHidContinuationPacket::createFromSerializedData):
(fido::FidoHidContinuationPacket::FidoHidContinuationPacket):
(fido::FidoHidContinuationPacket::getSerializedData const):
* Modules/webauthn/fido/FidoHidPacket.h: Added.
* Modules/webauthn/fido/FidoParsingUtils.cpp: Added.
(fido::getInitPacketData):
(fido::getContinuationPacketData):
* Modules/webauthn/fido/FidoParsingUtils.h: Added.
* Sources.txt:
* WebCore.xcodeproj/project.pbxproj:

Tools:

* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/Tests/WebCore/FidoHidMessageTest.cpp: Added.
(TestWebKitAPI::TEST):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236976 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago[WPE] Explicitly link against gmodule where used
csaavedra@igalia.com [Tue, 9 Oct 2018 19:25:18 +0000 (19:25 +0000)]
[WPE] Explicitly link against gmodule where used
https://bugs.webkit.org/show_bug.cgi?id=190398

Reviewed by Michael Catanzaro.

* PlatformWPE.cmake:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236975 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago[CoordGraphics] Remove the 'previous backing store' logic
commit-queue@webkit.org [Tue, 9 Oct 2018 19:23:21 +0000 (19:23 +0000)]
[CoordGraphics] Remove the 'previous backing store' logic
https://bugs.webkit.org/show_bug.cgi?id=188838

Patch by Carlos Eduardo Ramalho <cadubentzen@gmail.com> on 2018-10-09
Reviewed by Žan Doberšek.

Remove previousBackingStore from LayerState, since it's not needed anymore.
When there is scale adjustment, we now simply discard mainBackingStore and
recreate a new one.

No new tests required.

* platform/graphics/nicosia/texmap/NicosiaBackingStoreTextureMapperImpl.h:
* platform/graphics/texmap/coordinated/CoordinatedGraphicsLayer.cpp:
(WebCore::CoordinatedGraphicsLayer::flushCompositingStateForThisLayerOnly):
(WebCore::CoordinatedGraphicsLayer::updateContentBuffers):
(WebCore::CoordinatedGraphicsLayer::purgeBackingStores):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236974 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoPSON: Doing a cross-site navigation via the URL bar does not swap process on iOS
cdumez@apple.com [Tue, 9 Oct 2018 19:04:44 +0000 (19:04 +0000)]
PSON: Doing a cross-site navigation via the URL bar does not swap process on iOS
https://bugs.webkit.org/show_bug.cgi?id=190378
<rdar://problem/45059466>

Reviewed by Geoffrey Garen.

Source/WebKit:

Process swapping was sometimes not happening via URL bar navigation on iOS due to top-hit preloading,
which would use a new WKWebView for the speculative load and rely on the _relatedWebView SPI to use
the same WebContent process as the view currently on screen.

To address the issue, if the source URL is empty and the page has a related page, use the related
page's URL as source URL when doing the process-swap decision.

* UIProcess/API/APIPageConfiguration.cpp:
(API::PageConfiguration::relatedPage const):
(API::PageConfiguration::relatedPage): Deleted.
* UIProcess/API/APIPageConfiguration.h:
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::processForNavigationInternal):

Tools:

Add API test coverage.

* TestWebKitAPI/Tests/WebKitCocoa/ProcessSwapOnNavigation.mm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236973 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoWeb Inspector: REGRESSION: fix canvas test failures after r236952 and r236954
drousso@apple.com [Tue, 9 Oct 2018 19:04:18 +0000 (19:04 +0000)]
Web Inspector: REGRESSION: fix canvas test failures after r236952 and r236954
https://bugs.webkit.org/show_bug.cgi?id=190403

Reviewed by Joseph Pecoraro.

Source/WebInspectorUI:

* UserInterface/Controllers/CanvasManager.js:
(WI.CanvasManager.prototype._removeCanvas):
Clear the `shaderProgramCollection` when a `WI.Canvas` is removed so that a remove event is
fired for each `WI.ShaderProgram`.

LayoutTests:

* inspector/canvas/create-context-webmetal.html:
Fix missing parenthesis.

* inspector/canvas/resources/shaderProgram-utilities.js:
(createProgram):
(TestPage.registerInitializer.awaitProgramAdded):
(TestPage.registerInitializer):
(TestPage.registerInitializer.window.initializeTestSuite):
Replace the `WI.CanvasManager` events (which were removed) with listeners for events on the
specific `WI.Canvas`.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236972 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago[PSON] Prewarm system UI font
antti@apple.com [Tue, 9 Oct 2018 18:38:47 +0000 (18:38 +0000)]
[PSON] Prewarm system UI font
https://bugs.webkit.org/show_bug.cgi?id=190397

Reviewed by Geoffrey Garen.

Cache system UI font fallbacks. Almost every web process needs these.

* page/ProcessWarming.cpp:
(WebCore::ProcessWarming::prewarmGlobally):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236971 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoREGRESSION (Safari 12): Download of Blob URL fails
cdumez@apple.com [Tue, 9 Oct 2018 18:25:29 +0000 (18:25 +0000)]
REGRESSION (Safari 12): Download of Blob URL fails
https://bugs.webkit.org/show_bug.cgi?id=190351
<rdar://problem/45091181>

Reviewed by Geoffrey Garen.

Source/WebCore:

When using both the download attribute and target="_blank" on an anchor element, we would
mistakenly drop the download attribute after the "new window" policy decision has been made.
As a result, we would try to load the blob instead of downloading it.

Test: fast/dom/HTMLAnchorElement/anchor-file-blob-download-blank-target.html

* loader/FrameLoader.cpp:
(WebCore::FrameLoader::continueLoadAfterNewWindowPolicy):

LayoutTests:

Add layout test coverage.

* fast/dom/HTMLAnchorElement/anchor-file-blob-download-blank-target-expected.txt: Added.
* fast/dom/HTMLAnchorElement/anchor-file-blob-download-blank-target.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236970 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoStringTypeAdapter constructor is not properly enforcing String::MaxLength.
mark.lam@apple.com [Tue, 9 Oct 2018 18:15:56 +0000 (18:15 +0000)]
StringTypeAdapter constructor is not properly enforcing String::MaxLength.
https://bugs.webkit.org/show_bug.cgi?id=190392
<rdar://problem/45116210>

Reviewed by Saam Barati.

Previously, the StringTypeAdapter constructor for a UChar* string was summing the
unsigned length of the source string without an overflow check.  We now make that
length a size_t which removes this issue, and assert that it's within
String::MaxLength thereafter.

Also made the StringTypeAdapter constructor for a LChar* string behave in an
equivalent manner for consistency.  In both cases, we'll crash in a RELEASE_ASSERT
if the source string length exceeds String::MaxLength.

* wtf/text/StringConcatenate.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236969 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago[iOS] Replace @"UIPreviewDataAttachmentListIsContentManaged" with a UIKit constant
aestes@apple.com [Tue, 9 Oct 2018 17:56:58 +0000 (17:56 +0000)]
[iOS] Replace @"UIPreviewDataAttachmentListIsContentManaged" with a UIKit constant
https://bugs.webkit.org/show_bug.cgi?id=190400
<rdar://problem/35442879>

Reviewed by Wenson Hsieh.

Also fixed the iOS version check in WKContentViewInteraction.mm.

* Platform/spi/ios/UIKitSPI.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _dataForPreviewItemController:atPosition:type:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236968 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago[MSE][GStreamer] r236735 has some dead ASSERTs that need to be moved
aboya@igalia.com [Tue, 9 Oct 2018 17:40:01 +0000 (17:40 +0000)]
[MSE][GStreamer] r236735 has some dead ASSERTs that need to be moved
https://bugs.webkit.org/show_bug.cgi?id=190394

Reviewed by Xabier Rodriguez-Calvar.

* platform/graphics/gstreamer/mse/AppendPipeline.cpp:
(WebCore::createOptionalParserForFormat):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236967 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoREGRESSION (r232416): Can not scroll after swiping back on quoteunquoteapps.com
timothy_horton@apple.com [Tue, 9 Oct 2018 17:18:58 +0000 (17:18 +0000)]
REGRESSION (r232416): Can not scroll after swiping back on quoteunquoteapps.com
https://bugs.webkit.org/show_bug.cgi?id=190377
<rdar://problem/45108222>

Reviewed by Andy Estes.

Introduce the notion of 'pausing' to SnapshotRemovalTracker.
Reimplement r232416 in terms of this: the SnapshotRemovalTracker
starts out paused (not accepting events), and un-pauses when
we get a provisional load or same-document navigation.
This way, we don't lose the watchdog timer in cases where we get
no provisional load, same-document navigation, or main frame load
(which is the separate root cause for this bug -- this just papers
over it with a timeout).

* UIProcess/Cocoa/ViewGestureController.cpp:
(WebKit::ViewGestureController::didStartProvisionalLoadForMainFrame):
(WebKit::ViewGestureController::didSameDocumentNavigationForMainFrame):
Resume the snapshot removal tracker.

(WebKit::ViewGestureController::didReachMainFrameLoadTerminalState):
If we didn't see a provisional load or same document navigation,
but somehow got to the terminal load state, immediately remove the snapshot.

(WebKit::ViewGestureController::SnapshotRemovalTracker::resume):
(WebKit::ViewGestureController::SnapshotRemovalTracker::start):
Start the SnapshotRemovalTracker out in the paused state; it will be
resumed in the same places we previously would call the
provisionalOrSameDocumentLoadCallback.

(WebKit::ViewGestureController::SnapshotRemovalTracker::stopWaitingForEvent):
Ignore (but debug log) incoming events while paused.

* UIProcess/Cocoa/ViewGestureController.h:
(WebKit::ViewGestureController::SnapshotRemovalTracker::pause):
(WebKit::ViewGestureController::SnapshotRemovalTracker::isPaused const):
Add the pausing bit.

* UIProcess/ios/ViewGestureControllerIOS.mm:
(WebKit::ViewGestureController::endSwipeGesture):
* UIProcess/mac/ViewGestureControllerMac.mm:
(WebKit::ViewGestureController::endSwipeGesture):
Remove m_provisionalOrSameDocumentLoadCallback.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236966 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoHave DOMWindow get its frame from its document
cdumez@apple.com [Tue, 9 Oct 2018 17:15:21 +0000 (17:15 +0000)]
Have DOMWindow get its frame from its document
https://bugs.webkit.org/show_bug.cgi?id=190389

Reviewed by Geoff Garen.

Have DOMWindow get its frame from its document instead of having its own m_frame which can potentially
be out-of-sync.

* dom/Document.cpp:
(WebCore::Document::frameDestroyed):
(WebCore::Document::willDetachPage):
(WebCore::Document::attachToCachedFrame):
* dom/Document.h:
* page/DOMWindow.cpp:
(WebCore::DOMWindow::allowPopUp):
(WebCore::ContextDestructionObserver):
(WebCore::DOMWindow::didSecureTransitionTo):
(WebCore::DOMWindow::frameDestroyed):
(WebCore::DOMWindow::willDetachDocumentFromFrame):
(WebCore::DOMWindow::reconnectDOMWindowProperties):
(WebCore::DOMWindow::isCurrentlyDisplayedInFrame const):
(WebCore::DOMWindow::collectMatchingElementsInFlatTree):
(WebCore::DOMWindow::matchingElementInFlatTree):
(WebCore::DOMWindow::orientation const):
(WebCore::DOMWindow::console const):
(WebCore::DOMWindow::shouldHaveWebKitNamespaceForWorld):
(WebCore::DOMWindow::webkitNamespace):
(WebCore::DOMWindow::postMessage):
(WebCore::DOMWindow::postMessageTimerFired):
(WebCore::DOMWindow::frameElement const):
(WebCore::DOMWindow::focus):
(WebCore::DOMWindow::blur):
(WebCore::DOMWindow::close):
(WebCore::DOMWindow::print):
(WebCore::DOMWindow::stop):
(WebCore::DOMWindow::alert):
(WebCore::DOMWindow::confirm):
(WebCore::DOMWindow::prompt):
(WebCore::DOMWindow::find const):
(WebCore::DOMWindow::outerHeight const):
(WebCore::DOMWindow::outerWidth const):
(WebCore::DOMWindow::innerHeight const):
(WebCore::DOMWindow::innerWidth const):
(WebCore::DOMWindow::screenX const):
(WebCore::DOMWindow::screenY const):
(WebCore::DOMWindow::scrollX const):
(WebCore::DOMWindow::scrollY const):
(WebCore::DOMWindow::closed const):
(WebCore::DOMWindow::length const):
(WebCore::DOMWindow::name const):
(WebCore::DOMWindow::setName):
(WebCore::DOMWindow::setStatus):
(WebCore::DOMWindow::setDefaultStatus):
(WebCore::DOMWindow::self const):
(WebCore::DOMWindow::opener const):
(WebCore::DOMWindow::disownOpener):
(WebCore::DOMWindow::parent const):
(WebCore::DOMWindow::top const):
(WebCore::DOMWindow::getMatchedCSSRules const):
(WebCore::DOMWindow::devicePixelRatio const):
(WebCore::DOMWindow::scrollBy const):
(WebCore::DOMWindow::scrollTo const):
(WebCore::DOMWindow::allowedToChangeWindowGeometry const):
(WebCore::DOMWindow::moveBy const):
(WebCore::DOMWindow::moveTo const):
(WebCore::DOMWindow::resizeBy const):
(WebCore::DOMWindow::resizeTo const):
(WebCore::DOMWindow::clearTimeout):
(WebCore::DOMWindow::isSameSecurityOriginAsMainFrame const):
(WebCore::DOMWindow::finishedLoading):
(WebCore::DOMWindow::setLocation):
(WebCore::DOMWindow::open):
(WebCore::DOMWindow::showModalDialog):
(WebCore::DOMWindow::frame const):
(WebCore::FrameDestructionObserver): Deleted.
(WebCore::DOMWindow::willDetachPage): Deleted.
(WebCore::DOMWindow::detachFromFrame): Deleted.
(WebCore::DOMWindow::attachToFrame): Deleted.
* page/DOMWindow.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236965 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoISOTrackEncryptionBox returns incorrect defaultKeyID
jer.noble@apple.com [Tue, 9 Oct 2018 16:54:32 +0000 (16:54 +0000)]
ISOTrackEncryptionBox returns incorrect defaultKeyID
https://bugs.webkit.org/show_bug.cgi?id=190368

Reviewed by Eric Carlson.

Source/WebCore:

Test: TestWebKitAPI.ISOBox.ISOProtectionSchemeInfoBox

ISOTrackEncryptionBox::parse() increments the data offset by an incorrect amount.

Drive-by fix: add EXPORT macros to all the ISO box classes so that tests can be written in TestWebKitAPI.

* WebCore.xcodeproj/project.pbxproj:
* platform/graphics/iso/ISOBox.h:
* platform/graphics/iso/ISOOriginalFormatBox.h:
* platform/graphics/iso/ISOProtectionSchemeInfoBox.h:
* platform/graphics/iso/ISOProtectionSystemSpecificHeaderBox.h:
* platform/graphics/iso/ISOSchemeInformationBox.h:
* platform/graphics/iso/ISOSchemeTypeBox.h:
* platform/graphics/iso/ISOTrackEncryptionBox.cpp:
(WebCore::ISOTrackEncryptionBox::parse):
* platform/graphics/iso/ISOTrackEncryptionBox.h:

Tools:

* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/Tests/WebCore/ISOBox.cpp: Added.
(TestWebKitAPI::TEST):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236964 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoAdd support for IceCandidate stats
youenn@apple.com [Tue, 9 Oct 2018 15:44:34 +0000 (15:44 +0000)]
Add support for IceCandidate stats
https://bugs.webkit.org/show_bug.cgi?id=190329

Reviewed by Eric Carlson.

Source/ThirdParty/libwebrtc:

Export new stats kType values.

* Configurations/libwebrtc.iOS.exp:
* Configurations/libwebrtc.iOSsim.exp:
* Configurations/libwebrtc.mac.exp:

Source/WebCore:

Convert IceCandidate stats gathered by libwebrtc.
Since networkType might be sensitive information, we currently do not expose it.
We do not expose address either if it is a host or prflx candidate.

Test: webrtc/candidate-stats.html

* Modules/mediastream/RTCStatsReport.h:
* Modules/mediastream/RTCStatsReport.idl:
* Modules/mediastream/libwebrtc/LibWebRTCStatsCollector.cpp:
(WebCore::iceCandidateState):
(WebCore::fillRTCIceCandidateStats):
(WebCore::LibWebRTCStatsCollector::OnStatsDelivered):

LayoutTests:

* webrtc/candidate-stats-expected.txt: Added.
* webrtc/candidate-stats.html: Added.
* webrtc/routines.js:
* webrtc/video-stats.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236963 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoRevert temporary asserts for debugging a mysterious ASAN bot crash.
mark.lam@apple.com [Tue, 9 Oct 2018 15:31:31 +0000 (15:31 +0000)]
Revert temporary asserts for debugging a mysterious ASAN bot crash.
https://bugs.webkit.org/show_bug.cgi?id=190396

Reviewed by Yusuke Suzuki.

* wtf/StackBounds.cpp:
(WTF::StackBounds::newThreadStackBounds):
* wtf/StackBounds.h:
(WTF::StackBounds::checkConsistency const):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236962 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago[GStreamer] Stealing cross-origin video pixel with HLS
philn@webkit.org [Tue, 9 Oct 2018 15:10:17 +0000 (15:10 +0000)]
[GStreamer] Stealing cross-origin video pixel with HLS
https://bugs.webkit.org/show_bug.cgi?id=190003

Reviewed by Xabier Rodriguez-Calvar.

Source/WebCore:

Report the SecurityOrigin of downloaded adaptivedemux (HLS, DASH,
SmoothStreaming) fragments as tainted if their origin differs from
the manifest SecurityOrigin. SecurityOrigins are stored in the
CachedResourceStreamingClient implemented in the internal
GStreamer HTTP(S) source element.

The implementation is not ideal yet because the fragments download
is performed by the WebProcess, until bug 189967 is fixed. When
this bug is fixed, the m_hasTaintedOrigin member variable should
be removed and all checks be done unconditionally to the
webkithttpsrc element which will manage the download of the
manifests and fragments.

* platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
(WebCore::MediaPlayerPrivateGStreamer::loadFull): Reset the m_hasTaintedOrigin value.
(WebCore::MediaPlayerPrivateGStreamer::handleMessage): Get the
fragment URL from the adaptivedemux stats message and check if its
origin is tainted.
(WebCore::MediaPlayerPrivateGStreamer::wouldTaintOrigin const):
Initial implementation by checking the m_hasTaintedOrigin member
variable value.
* platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.h:
* platform/graphics/gstreamer/WebKitWebSourceGStreamer.cpp:
(CachedResourceStreamingClient::responseReceived): Store the
resource origin internally so it can be checked later on by
webKitSrtcWouldTaintOrigin().
(webKitSrcWouldTaintOrigin): Check given origin against cached
origins. This implementation is similar to Cocoa's
WebCoreNSURLSession implementation.
* platform/graphics/gstreamer/WebKitWebSourceGStreamer.h:

LayoutTests:

* platform/gtk/TestExpectations: Unflag now-passing test.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236961 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago[JHBuild] Update to GStreamer 1.14.4
philn@webkit.org [Tue, 9 Oct 2018 13:54:57 +0000 (13:54 +0000)]
[JHBuild] Update to GStreamer 1.14.4
https://bugs.webkit.org/show_bug.cgi?id=190395

Reviewed by Xabier Rodriguez-Calvar.

* gstreamer/jhbuild.modules: Remove patches merged in 1.14.4 and bump version of GStreamer modules.
* gstreamer/patches/gst-plugins-good-0007-matroskademux-Allow-Matroska-headers-to-be-read-more.patch: Removed.
* gstreamer/patches/gst-plugins-good-0008-matroskademux-Refactor-track-parsing-out-from-adding.patch: Removed.
* gstreamer/patches/gst-plugins-good-0009-matroskademux-Parse-successive-Tracks-elements.patch: Removed.
* gstreamer/patches/gst-plugins-good-0010-matroskademux-Emit-no-more-pads-after-parsing-Tracks.patch: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236960 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoPrewarm FontDatabase on process swap
antti@apple.com [Tue, 9 Oct 2018 09:14:37 +0000 (09:14 +0000)]
Prewarm FontDatabase on process swap
https://bugs.webkit.org/show_bug.cgi?id=190312

Reviewed by Chris Dumez.

Source/WebCore:

Implement basic prewarming of FontDatabase. When the domain of the first page load on
a new process is known we pass the list of font families previous used by that domain to the
process. This is used to prewarm CoreText font database. Initialization (which involves lots of
blocking IPC) happens outside the main thread so the fonts are ready to use when needed.

* WebCore.xcodeproj/project.pbxproj:
* page/PrewarmInformation.h: Added.
(WebCore::PrewarmInformation::encode const):
(WebCore::PrewarmInformation::decode):

Add data structure for prewarm information with encode/decode support.

* page/ProcessWarming.cpp:
(WebCore::ProcessWarming::collectPrewarmInformation):
(WebCore::ProcessWarming::prewarmWithInformation):

Prewarming interface to be used from WebKit.

* page/ProcessWarming.h:
* platform/graphics/FontCache.cpp:
(WebCore::FontCache::collectPrewarmInformation const):
(WebCore::FontCache::prewarm):
* platform/graphics/FontCache.h:
* platform/graphics/cocoa/FontCacheCoreText.cpp:
(WebCore::FontDatabase::collectionForFamily):
(WebCore::FontDatabase::fontForPostScriptName):
(WebCore::FontDatabase::clear):

Add mutex for thread safe access.

(WebCore::FontCache::createFontPlatformData):
(WebCore::FontCache::collectPrewarmInformation const):

Collect font families seen by this process.

(WebCore::FontCache::prewarm):

Prewarm FontDatabase in a dispatch queue.

Source/WebKit:

Add mechanism for caching prewarm information per domain on UI process side. This information
is then passed to newly initialized web processes that are being used for the same domain.

* UIProcess/WebPageProxy.h:
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::processForNavigationInternal):

Pass prewarm information on process swap to the new process.

(WebKit::WebProcessPool::didCollectPrewarmInformation):

UI process side cache.

* UIProcess/WebProcessPool.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::didCollectPrewarmInformation):
* UIProcess/WebProcessProxy.h:
* UIProcess/WebProcessProxy.messages.in:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::close):
(WebKit::WebPage::didReceivePolicyDecision):

Collect and cache prewarm information for the previous process on process swap.

* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::prewarmGlobally):

Renamed for clarity.

(WebKit::WebProcess::prewarmWithDomainInformation):
(WebKit::WebProcess::sendPrewarmInformation):
* WebProcess/WebProcess.h:
* WebProcess/WebProcess.messages.in:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236959 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago[macOS] API test DragAndDropTests.DropColor is failing on macOS 10.12
wenson_hsieh@apple.com [Tue, 9 Oct 2018 03:28:31 +0000 (03:28 +0000)]
[macOS] API test DragAndDropTests.DropColor is failing on macOS 10.12
https://bugs.webkit.org/show_bug.cgi?id=190386

Reviewed by Ryosuke Niwa.

Unlike macOS 10.13 and 10.14, -[NSColor redColor] is not equivalent to a fully red color in sRGB on macOS 10.12.
Fix the test by dragging and dropping an NSColor that is equivalent to rgb(255, 0, 0) on all relevant versions
of macOS.

* TestWebKitAPI/Tests/mac/DragAndDropTestsMac.mm:
(TEST):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236958 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months agoFix inspector/canvas/recording.html after r236952
drousso@apple.com [Tue, 9 Oct 2018 02:37:24 +0000 (02:37 +0000)]
Fix inspector/canvas/recording.html after r236952

Unreviewed. Test gardening.

* inspector/canvas/recording-expected.txt:
* inspector/canvas/recording.html:
Recordings default to only capturing a single frame, so add a parameter to ensure that we
are capturing multiple frames.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236957 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago/api/report takes 15+ minutes submitting some test results
rniwa@webkit.org [Tue, 9 Oct 2018 02:17:42 +0000 (02:17 +0000)]
/api/report takes 15+ minutes submitting some test results
https://bugs.webkit.org/show_bug.cgi?id=190382

Rubber-stamped by Alexey Proskuryakov and unofficially reviewed by Dewei Zhu.

The issue was that recursively_ensure_tests would issue thousands of dependent queries.

Since most reporting of results would be happening after all the tests and test metrics had been created,
simply fetch the list of all tests and test metrics upfront in memory.

This would slow-down the reporting of other test results but in practice doesn't matter (an extra few seconds
of overhead) but for slow reporting, it can significantly reduce the runtime from ~10 minutes to ~10 seconds.

* public/include/report-processor.php:
(ReportProcessor::process): Call fetch_tests before recursively_ensure_tests.
(ReportProcessor::fetch_tests): Added. This builds up a mapping of tests based on a parent test ID as well as
a mapping of metrics per test based on its name and then its aggregator's name.
(ReportProcessor::recursively_ensure_tests): Added. Use the in-memory maps built in fetch_tests when possible.
(ReportProcessor::aggregator_list_if_exists): Take a reference instead of passing it by value.
(TestRunsGenerator::add_aggregated_metric): Now takes metric ID. The code to add the test metric had been moved
to recursively_ensure_tests.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236956 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 months ago[Cocoa] [WK2] Add support for text alignment and text lists in font attributes
wenson_hsieh@apple.com [Tue, 9 Oct 2018 01:52:49 +0000 (01:52 +0000)]
[Cocoa] [WK2] Add support for text alignment and text lists in font attributes
https://bugs.webkit.org/show_bug.cgi?id=190342
<rdar://problem/44767118>

Reviewed by Tim Horton.

Source/WebCore:

Adds support for computing and encoding information about text alignment and enclosing list elements in the font
attributes dictionary, exposed to the UI delegate on Cocoa platforms. This is exposed through NSParagraphStyle,
which contains properties for both `NSTextAlignment` and an array of enclosing `NSTextList` objects.

Test:   FontAttributes.NestedTextListsWithHorizontalAlignment
        FontAttributes.FontAttributesAfterChangingSelection

* editing/Editor.cpp:
(WebCore::editableTextListsAtPositionInDescendingOrder):

Add a helper function to ascend the DOM, starting at the given position, in search of enclosing list elements.
For each enclosing list element we find within the scope of the editable root, we create a `TextList`
representing the list element.

(WebCore::Editor::fontAttributesAtSelectionStart const):

Compute and set enclosing text lists and text alignment. For text alignment, we convert the initial text-align
value, `start`, to NSTextAlignmentNatural (the default text alignment type on Cocoa platforms); other values
then map to Left, Right, Center, and Justified as expected (taking direction into account).

* editing/FontAttributes.h:
(WebCore::TextList::encode const):
(WebCore::TextList::decode):

Introduce TextList, a platform-agnostic representation of an unordered or ordered list. On Cocoa, this can be
used to construct a corresponding NSTextList.

* editing/cocoa/FontAttributesCocoa.mm:
(WebCore::cocoaTextListMarkerName):

Attempt to map a WebCore list style type to a `NSTextListMarker*` constant. While most of the Cocoa marker
formats have a corresponding web-exposed list style type, many web-exposed types have no Cocoa equivalent; as
such, fall back to default marker formats: "{disc}" for unordered lists and "{decimal}" for ordered lists.

(WebCore::TextList::createTextList const):
(WebCore::FontAttributes::createDictionary const):

Include an NSParagraphStyle in the dictionary of font attributes that includes information about text alignment
and enclosing text lists (per Cocoa convention, in order from outermost list to innermost list).

Source/WebCore/PAL:

* pal/ios/UIKitSoftLink.h:
* pal/ios/UIKitSoftLink.mm:

Add NSParagraphStyle and NSTextList to the UIKit soft link header.

* pal/spi/cocoa/NSAttributedStringSPI.h:
* pal/spi/ios/UIKitSPI.h:

Add some SPI declarations for NSMutableParagraphStyle and NSTextList when building with a non-internal iOS SDK,
and import <UIKit/NSParagraphStyle_Private.h> and <UIKit/NSTextList.h> when building with an internal iOS SDK.
Additionally, define some internal UIFoundation string constants that denote NSTextList marker formats. These
constants are API on macOS, but are neither exposed as API nor SPI on iOS.

Source/WebKit:

Add encoding and decoding for text alignment and text lists in FontAttributes.

* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<FontAttributes>::encode):
(IPC::ArgumentCoder<FontAttributes>::decode):

Tools:

* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/Tests/WebKitCocoa/FontAttributes.mm:

Add a new API test to check that the font attributes dictionary contains the correct NSParagraphStyles when
moving the selection around text inside nested list elements.

(TestWebKitAPI::checkParagraphStyles):

Add a new helper to check an NSParagraphStyle against expected results.

(TestWebKitAPI::webViewForTestingFontAttributes):

Allow each test to pass in the name of the test page to load.

(TestWebKitAPI::TEST):
* TestWebKitAPI/Tests/WebKitCocoa/nested-lists.html: Added.

Add a new test page that contains text enclosed in multiple levels of unordered and ordered lists.

* TestWebKitAPI/Tests/WebKitCocoa/rich-text-attributes.html:

Remove `text-align: left` from one of these elements, to test for NSTextAlignmentNatural.

LayoutTests:

* editing/mac/attributed-string/attributed-string-for-typing-expected.txt:
* editing/mac/attributed-string/attributed-string-for-typing-with-color-filter-expected.txt:
* platform/mac-sierra/editing/mac/attributed-string/attributed-string-for-typing-expected.txt:
* platform/mac-sierra/editing/mac/attributed-string/attributed-string-for-typing-with-color-filter-expected.txt:

Rebaseline these two layout tests, now that typing attributes include an NSParagraphStyle.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@236955 268f45cc-cd09-0410-ab3c-d52691b4dbfc