WebKit-https.git
5 years agoWeb Replay: capture and replay mouse events
bburg@apple.com [Thu, 20 Mar 2014 22:14:41 +0000 (22:14 +0000)]
Web Replay: capture and replay mouse events
https://bugs.webkit.org/show_bug.cgi?id=129395

Reviewed by Joseph Pecoraro.

.:

Create a manual test for capture/replay of mouse events.
Copy over the crypto-md5.js library from SunSpider.

* ManualTests/inspector/replay-mouse-events.html: Added.
* ManualTests/inspector/resources/crypto-md5.js: Added.

Source/WebCore:

Add support for capturing and replaying mouse inputs that come from WebKit2.
Hook up the UserInputBridge to session state changes in the ReplayController so
that the bridge knows when to capture or deny mouse inputs.

Test: ManualTests/inspector/replay-mouse-events.html

* platform/PlatformEvent.h: Give explicit storage types to Modifiers and Type enums
so they can be forward-declared.
* platform/PlatformMouseEvent.h: Give an explicit storage type to enum MouseButton.
Add operator== and operator!= for MouseButton to work around an MSVC bug.

* replay/ReplayController.cpp: Perform session state changes in a helper function, and
at the same time change the state of the page's user input bridge.
(WebCore::ReplayController::setSessionState):
(WebCore::ReplayController::startCapturing):
(WebCore::ReplayController::stopCapturing):
(WebCore::ReplayController::startPlayback):
(WebCore::ReplayController::cancelPlayback):

* replay/ReplayInputDispatchMethods.cpp: Add dispatch methods for new inputs.
(WebCore::HandleMouseMove::dispatch):
(WebCore::HandleMousePress::dispatch):
(WebCore::HandleMouseRelease::dispatch):

* replay/SerializationMethods.cpp: Add helper macros so that encode/decode methods look
symmetric with one data member per line. This helps reduce unintentional inconsistencies.
(JSC::EncodingTraits<PlatformMouseEvent>::encodeValue): Added.
(JSC::EncodingTraits<PlatformMouseEvent>::decodeValue): Added.
* replay/SerializationMethods.h:

* replay/UserInputBridge.cpp: Fill in the bridge method implementations for mouse
events, adding helpers and macros as necessary to reduce boilerplate.
(WebCore::UserInputBridge::activeCursor): Added.
(WebCore::UserInputBridge::handleMousePressEvent):
(WebCore::UserInputBridge::handleMouseReleaseEvent):
(WebCore::UserInputBridge::handleMouseMoveEvent):
(WebCore::UserInputBridge::handleMouseMoveOnScrollbarEvent):

* replay/UserInputBridge.h: Add a bridge state enum along with getters and setters.
The enum value controls whether the bridge should capture commands, deny non-synthetic
commands (from the user), or allow anything to pass (the default).
(WebCore::UserInputBridge::setState): Added.
(WebCore::UserInputBridge::state): Added.

* replay/WebInputs.json: Add inputs HandleMouseMove, HandleMousePress, HandleMouseRelease.
Add enum definitions for PlatformEvent::Type, PlatformEvent::Modifiers, and PlatformMouseEvent::MouseButton.
Alphabetize the existing data type definitions.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@166006 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoAdd WebCore::IOSurface wrapper
timothy_horton@apple.com [Thu, 20 Mar 2014 22:06:38 +0000 (22:06 +0000)]
Add WebCore::IOSurface wrapper
https://bugs.webkit.org/show_bug.cgi?id=130529

Reviewed by Simon Fraser.

* WebCore.exp.in:
* WebCore.xcodeproj/project.pbxproj:
* platform/graphics/cg/ColorCG.cpp:
* platform/graphics/cg/GraphicsContextCG.h:
(WebCore::cachedCGColorSpace):
Move cachedCGColorSpace out so that other clients of GraphicsContextCG
can use its mapping from WebCore::ColorSpace -> CGColorSpaceRef.

* platform/graphics/cg/ImageBufferBackingStoreCache.cpp:
* platform/graphics/cg/ImageBufferCG.cpp:
Add FIXMEs for ImageBuffer to adopt WebCore::IOSurface.

* platform/graphics/cocoa/IOSurface.h: Added.
* platform/graphics/cocoa/IOSurface.mm: Added.
(IOSurface::create):
(IOSurface::createFromMachPort):
(IOSurface::createFromSurface):
(IOSurface::IOSurface):
(IOSurface::maximumSize):
(IOSurface::createMachPort):
(IOSurface::createImage):
(IOSurface::ensurePlatformContext):
(IOSurface::ensureGraphicsContext):
(IOSurface::state):
(IOSurface::isPurgeable):
(IOSurface::setIsPurgeable):
(IOSurface::inUse):
Add WebCore::IOSurface.

* Shared/mac/RemoteLayerBackingStore.h:
* Shared/mac/RemoteLayerBackingStore.mm:
(RemoteLayerBackingStore::encode):
(RemoteLayerBackingStore::decode):
(RemoteLayerBackingStore::display):
(RemoteLayerBackingStore::applyBackingStoreToLayer):
* UIProcess/ios/ViewGestureControllerIOS.mm:
(WebKit::ViewGestureController::beginSwipeGesture):
(WebKit::ViewGestureController::removeSwipeSnapshot):
* UIProcess/mac/ViewGestureController.h:
* UIProcess/mac/ViewGestureControllerMac.mm:
(WebKit::ViewGestureController::beginSwipeGesture):
(WebKit::ViewGestureController::removeSwipeSnapshot):
* UIProcess/mac/ViewSnapshotStore.h:
* UIProcess/mac/ViewSnapshotStore.mm:
(WebKit::createIOSurfaceFromImage):
(WebKit::ViewSnapshotStore::recordSnapshot):
Adopt WebCore::IOSurface.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@166005 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoiOS build fix after r165992.
psolanki@apple.com [Thu, 20 Mar 2014 21:53:03 +0000 (21:53 +0000)]
iOS build fix after r165992.

* generate-export-file:
(preprocessorMacros):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@166004 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[WebRTC] Moving RTCConfiguration and RTCIceServer to Modules/mediastream
thiago.lacerda@openbossa.org [Thu, 20 Mar 2014 21:19:03 +0000 (21:19 +0000)]
[WebRTC] Moving RTCConfiguration and RTCIceServer to Modules/mediastream
https://bugs.webkit.org/show_bug.cgi?id=130434

Reviewed by Eric Carlson.

Those classes are defined in WebRTC spec, so they should not be in platform layer.
Adding a platform implementation to be passed to handler classes.

* Modules/mediastream/RTCConfiguration.h: Copied from Source/WebCore/platform/mediastream/RTCConfiguration.h.
* Modules/mediastream/RTCIceServer.h: Copied from Source/WebCore/platform/mediastream/RTCIceServer.h.
* Modules/mediastream/RTCPeerConnection.cpp:
(WebCore::RTCPeerConnection::RTCPeerConnection):
(WebCore::RTCPeerConnection::updateIce):
* WebCore.xcodeproj/project.pbxproj:
* platform/mediastream/RTCConfigurationPrivate.h: Added.
* platform/mediastream/RTCIceServerPrivate.h: Added.
* platform/mediastream/RTCPeerConnectionHandler.h:
* platform/mock/RTCPeerConnectionHandlerMock.cpp:
(WebCore::RTCPeerConnectionHandlerMock::initialize):
(WebCore::RTCPeerConnectionHandlerMock::updateIce):
* platform/mock/RTCPeerConnectionHandlerMock.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@166003 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUpdate iOS input method code to use new cross-platform async functions
ap@apple.com [Thu, 20 Mar 2014 21:17:51 +0000 (21:17 +0000)]
Update iOS input method code to use new cross-platform async functions
https://bugs.webkit.org/show_bug.cgi?id=130530

Reviewed by Enrica Casucci.

Merge identical iOS and cross-platform functions.

Also tightened up ifdefs, and removed stub iOS implementations of sync methods,
which hopefully won't be ever needed. The only tricky case is sync cancelComposition,
will I'll investigate separately.

* UIProcess/WebPageProxy.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView setMarkedText:selectedRange:]):
(-[WKContentView unmarkText]):
(-[WKContentView insertText:]):
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::cancelComposition):
* UIProcess/mac/WebPageProxyMac.mm:
(WebKit::WebPageProxy::cancelComposition):
(WebKit::WebPageProxy::insertDictatedTextAsync):
(WebKit::WebPageProxy::attributedSubstringForCharacterRangeAsync):
(WebKit::WebPageProxy::attributedStringForCharacterRangeCallback):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/ios/WebPageIOS.mm:
* WebProcess/WebPage/mac/WebPageMac.mm:
(WebKit::WebPage::cancelComposition):
(WebKit::WebPage::insertDictatedTextAsync):
(WebKit::WebPage::attributedSubstringForCharacterRangeAsync):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@166002 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[WinCairo] Unreviewed speculative build after my r165996.
alex.christensen@flexsim.com [Thu, 20 Mar 2014 21:15:23 +0000 (21:15 +0000)]
[WinCairo] Unreviewed speculative build after my r165996.

* platform/network/soup/AuthenticationChallengeSoup.cpp:
* platform/network/soup/CertificateInfo.cpp:
* platform/network/soup/CookieJarSoup.cpp:
* platform/network/soup/ResourceRequestSoup.cpp:
* platform/network/soup/ResourceResponseSoup.cpp:
* platform/network/soup/SynchronousLoaderClientSoup.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@166001 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoASSERTION FAILED: event == [NSApp currentEvent] hitting Command-+
ap@apple.com [Thu, 20 Mar 2014 21:12:00 +0000 (21:12 +0000)]
ASSERTION FAILED: event == [NSApp currentEvent] hitting Command-+
https://bugs.webkit.org/show_bug.cgi?id=130526
<rdar://problem/16371612>

Reviewed by Anders Carlsson.

* UIProcess/API/mac/WKView.mm: (-[WKView performKeyEquivalent:]): Cut custom
processing short whenever we are redispatching an event, not just when the event
we are handling is the one being redispatched. There shouldn't be any other unrelated
events here, just the one being redispatched and synthetic ones generated from it by AppKit.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@166000 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoSource/JavaScriptCore: Crash beneath operationTearOffActivation running this JS compr...
barraclough@apple.com [Thu, 20 Mar 2014 21:05:49 +0000 (21:05 +0000)]
Source/JavaScriptCore: Crash beneath operationTearOffActivation running this JS compression demo
https://bugs.webkit.org/show_bug.cgi?id=130295
<rdar://problem/16332337>

Patch by Filip Pizlo <fpizlo@apple.com> on 2014-03-20
Reviewed by Oliver Hunt.

Make sure that we flush things as if we were at a terminal, if we are at a block with
no forward edges. This fixes infinitely loopy code with captured variables.

Make sure that the CFG simplifier adds explicit flushes whenever it jettisons a block.

Make it so that NodeIsFlushed is a thing. Previously only SSA used it and it computed
it by itself. Now it's an artifact of CPS rethreading.

Add a bunch of tests. All of them previously either crashed or returned bad output due
to memory corruption.

* bytecode/CodeBlock.cpp:
(JSC::CodeBlock::isCaptured):
* dfg/DFGByteCodeParser.cpp:
(JSC::DFG::ByteCodeParser::flushForTerminal):
(JSC::DFG::ByteCodeParser::flushForReturn):
(JSC::DFG::ByteCodeParser::flushIfTerminal):
(JSC::DFG::ByteCodeParser::branchData):
(JSC::DFG::ByteCodeParser::parseBlock):
* dfg/DFGCFGSimplificationPhase.cpp:
(JSC::DFG::CFGSimplificationPhase::keepOperandAlive):
* dfg/DFGCPSRethreadingPhase.cpp:
(JSC::DFG::CPSRethreadingPhase::run):
(JSC::DFG::CPSRethreadingPhase::computeIsFlushed):
(JSC::DFG::CPSRethreadingPhase::addFlushedLocalOp):
(JSC::DFG::CPSRethreadingPhase::addFlushedLocalEdge):
* dfg/DFGCSEPhase.cpp:
(JSC::DFG::CSEPhase::performNodeCSE):
* dfg/DFGGraph.cpp:
(JSC::DFG::Graph::clearFlagsOnAllNodes):
* dfg/DFGGraph.h:
* dfg/DFGNode.h:
* dfg/DFGNodeFlags.cpp:
(JSC::DFG::dumpNodeFlags):
* dfg/DFGNodeFlags.h:
* dfg/DFGSSAConversionPhase.cpp:
(JSC::DFG::SSAConversionPhase::run):
* tests/stress/activation-test-loop.js: Added.
(Inner.this.doStuff):
(Inner):
(foo.inner.isDone):
(foo):
* tests/stress/inferred-infinite-loop-that-uses-captured-variables.js: Added.
(bar):
(foo):
(noInline):
* tests/stress/infinite-loop-that-uses-captured-variables-before-throwing.js: Added.
(bar):
(foo):
(noInline):
* tests/stress/infinite-loop-that-uses-captured-variables-but-they-do-not-escape.js: Added.
(bar):
(foo):
(noInline):
* tests/stress/infinite-loop-that-uses-captured-variables-with-osr-entry.js: Added.
(bar):
(foo):
(noInline):
* tests/stress/infinite-loop-that-uses-captured-variables.js: Added.
(bar):
(foo):
(noInline):
* tests/stress/tricky-indirectly-inferred-infinite-loop-that-uses-captured-variables-and-creates-the-activation-outside-the-loop.js: Added.
(bar):
(fuzz):
(foo.f):
(foo):
* tests/stress/tricky-inferred-infinite-loop-that-uses-captured-variables-and-creates-the-activation-outside-the-loop.js: Added.
(bar):
(foo.f):
(foo):
* tests/stress/tricky-infinite-loop-that-uses-captured-variables-and-creates-the-activation-outside-the-loop.js: Added.
(bar):
(foo.f):
(foo):
* tests/stress/tricky-infinite-loop-that-uses-captured-variables.js: Added.
(bar):
(foo):
(noInline):

Source/WTF: Remove IdentifierTable typedef, isIdentifier()
https://bugs.webkit.org/show_bug.cgi?id=130533

Rubber stamped by Geoff Garen.

Code should use AtomicStringTable, isAtomic() directly.

* wtf/WTFThreadData.h:
(WTF::WTFThreadData::setCurrentAtomicStringTable):
(WTF::WTFThreadData::resetCurrentAtomicStringTable):
    - removed IdentifierTypedef.
* wtf/text/StringImpl.h:
(WTF::StringImpl::flagIsAtomic):
    - renamed flagIsIdentfier -> flagIsAtomic.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165999 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[iOS] asynchronous media callbacks should happen on the web thread
eric.carlson@apple.com [Thu, 20 Mar 2014 21:02:31 +0000 (21:02 +0000)]
[iOS] asynchronous media callbacks should happen on the web thread
https://bugs.webkit.org/show_bug.cgi?id=130536

Reviewed by Jer Noble.

* platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.mm:
(WebCore::MediaPlayerPrivateAVFoundationObjC::checkPlayability): Use callOnMainThread
    instead of dispatch_async(dispatch_get_main_queue()).

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165998 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoAdd a flushing mechanism for the WebVTTParser
bfulgham@apple.com [Thu, 20 Mar 2014 20:59:53 +0000 (20:59 +0000)]
Add a flushing mechanism for the WebVTTParser
https://bugs.webkit.org/show_bug.cgi?id=130532

Reviewed by Eric Carlson.

Merged from Blink (patch by vcarbune@chromium.org):
https://chromium.googlesource.com/chromium/blink/+/1ac2b3081492ab8661429230004013a9a3623f0f
http://crbug.com/54203003
https://chromium.googlesource.com/chromium/blink/+/24343d32bc36d20d59ef3e10040faddde65929e6
http://crbug.com/59453002
https://chromium.googlesource.com/chromium/blink/+/ef2793f1d9d207389589ecf9a136ea5c7170b4af
http://crbug.com/75233002
https://chromium.googlesource.com/chromium/blink/+/a29f2f8324abcd4826b41e7dc34373dee2ec57a9
http://crbug.com/52713005
https://chromium.googlesource.com/chromium/blink/+/7ce003c8d47b39f116f1f2a592a5dfb4a9e77228
http://crbug.com/64303004
https://chromium.googlesource.com/chromium/blink/+/18f896b3498478311e880f782813d5dfc8c0c7d1
http://crbug.com/96933004
https://chromium.googlesource.com/chromium/blink/+/8670e8ecb13254a651f5493ec83f4484d18c154e
http://crbug.com/98783005
https://chromium.googlesource.com/chromium/blink/+/4ac55780a6af3d76e0159c1d145330964ad56647
http://crbug.com/104713002

Tests: http/tests/media/track/track-webvtt-slow-loading-2.html
       http/tests/media/track/track-webvtt-slow-loading.html

* CMakeLists.txt:
* GNUmakefile.list.am:
* WebCore.vcxproj/WebCore.vcxproj:
* WebCore.vcxproj/WebCore.vcxproj.filters:
* WebCore.xcodeproj/project.pbxproj:
* html/track/BufferedLineReader.cpp: Added.
* html/track/BufferedLineReader.h: Added.
* html/track/TextTrackRegion.cpp:
* html/track/VTTCue.cpp:
* html/track/WebVTTParser.cpp:
* html/track/WebVTTParser.h:
* loader/TextTrackLoader.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165997 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoPreparation for using Soup on Windows.
alex.christensen@flexsim.com [Thu, 20 Mar 2014 20:55:05 +0000 (20:55 +0000)]
Preparation for using Soup on Windows.
https://bugs.webkit.org/show_bug.cgi?id=130472

Reviewed by Brent Fulgham.

* WebCore.vcxproj/WebCore.vcxproj:
* WebCore.vcxproj/WebCore.vcxproj.filters:
Added Soup source to build on WinCairo.
* platform/network/curl/CookieJarCurl.cpp:
* platform/network/curl/CredentialStorageCurl.cpp:
* platform/network/curl/CurlCacheEntry.cpp:
* platform/network/curl/CurlCacheManager.cpp:
* platform/network/curl/CurlDownload.cpp:
* platform/network/curl/DNSCurl.cpp:
* platform/network/curl/FormDataStreamCurl.cpp:
* platform/network/curl/MultipartHandle.cpp:
* platform/network/curl/ProxyServerCurl.cpp:
* platform/network/curl/ResourceHandleCurl.cpp:
* platform/network/curl/ResourceHandleManager.cpp:
* platform/network/curl/SSLHandle.cpp:
* platform/network/curl/SocketStreamHandleCurl.cpp:
Protected with USE(CURL) to avoid duplicate symbols when linking.
* platform/network/soup/AuthenticationChallengeSoup.cpp:
* platform/network/soup/CertificateInfo.cpp:
* platform/network/soup/CookieJarSoup.cpp:
* platform/network/soup/CookieStorageSoup.cpp:
* platform/network/soup/CredentialStorageSoup.cpp:
* platform/network/soup/DNSSoup.cpp:
* platform/network/soup/NetworkStorageSessionSoup.cpp:
* platform/network/soup/ProxyResolverSoup.cpp:
* platform/network/soup/ProxyServerSoup.cpp:
* platform/network/soup/ResourceErrorSoup.cpp:
* platform/network/soup/ResourceHandleSoup.cpp:
* platform/network/soup/ResourceRequestSoup.cpp:
* platform/network/soup/ResourceResponseSoup.cpp:
* platform/network/soup/SocketStreamHandleSoup.cpp:
* platform/network/soup/SoupNetworkSession.cpp:
* platform/network/soup/SynchronousLoaderClientSoup.cpp:
Protected with USE(CURL) to avoid duplicate symbols when linking.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165996 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoCrash beneath operationTearOffActivation running this JS compression demo
fpizlo@apple.com [Thu, 20 Mar 2014 20:53:37 +0000 (20:53 +0000)]
Crash beneath operationTearOffActivation running this JS compression demo
https://bugs.webkit.org/show_bug.cgi?id=130295
<rdar://problem/16332337>

Reviewed by Oliver Hunt.

Make sure that we flush things as if we were at a terminal, if we are at a block with
no forward edges. This fixes infinitely loopy code with captured variables.

Make sure that the CFG simplifier adds explicit flushes whenever it jettisons a block.

Make it so that NodeIsFlushed is a thing. Previously only SSA used it and it computed
it by itself. Now it's an artifact of CPS rethreading.

Add a bunch of tests. All of them previously either crashed or returned bad output due
to memory corruption.

* bytecode/CodeBlock.cpp:
(JSC::CodeBlock::isCaptured):
* dfg/DFGByteCodeParser.cpp:
(JSC::DFG::ByteCodeParser::flushForTerminal):
(JSC::DFG::ByteCodeParser::flushForReturn):
(JSC::DFG::ByteCodeParser::flushIfTerminal):
(JSC::DFG::ByteCodeParser::branchData):
(JSC::DFG::ByteCodeParser::parseBlock):
* dfg/DFGCFGSimplificationPhase.cpp:
(JSC::DFG::CFGSimplificationPhase::keepOperandAlive):
* dfg/DFGCPSRethreadingPhase.cpp:
(JSC::DFG::CPSRethreadingPhase::run):
(JSC::DFG::CPSRethreadingPhase::computeIsFlushed):
(JSC::DFG::CPSRethreadingPhase::addFlushedLocalOp):
(JSC::DFG::CPSRethreadingPhase::addFlushedLocalEdge):
* dfg/DFGCSEPhase.cpp:
(JSC::DFG::CSEPhase::performNodeCSE):
* dfg/DFGGraph.cpp:
(JSC::DFG::Graph::clearFlagsOnAllNodes):
* dfg/DFGGraph.h:
* dfg/DFGNode.h:
* dfg/DFGNodeFlags.cpp:
(JSC::DFG::dumpNodeFlags):
* dfg/DFGNodeFlags.h:
* dfg/DFGSSAConversionPhase.cpp:
(JSC::DFG::SSAConversionPhase::run):
* tests/stress/activation-test-loop.js: Added.
(Inner.this.doStuff):
(Inner):
(foo.inner.isDone):
(foo):
* tests/stress/inferred-infinite-loop-that-uses-captured-variables.js: Added.
(bar):
(foo):
(noInline):
* tests/stress/infinite-loop-that-uses-captured-variables-before-throwing.js: Added.
(bar):
(foo):
(noInline):
* tests/stress/infinite-loop-that-uses-captured-variables-but-they-do-not-escape.js: Added.
(bar):
(foo):
(noInline):
* tests/stress/infinite-loop-that-uses-captured-variables-with-osr-entry.js: Added.
(bar):
(foo):
(noInline):
* tests/stress/infinite-loop-that-uses-captured-variables.js: Added.
(bar):
(foo):
(noInline):
* tests/stress/tricky-indirectly-inferred-infinite-loop-that-uses-captured-variables-and-creates-the-activation-outside-the-loop.js: Added.
(bar):
(fuzz):
(foo.f):
(foo):
* tests/stress/tricky-inferred-infinite-loop-that-uses-captured-variables-and-creates-the-activation-outside-the-loop.js: Added.
(bar):
(foo.f):
(foo):
* tests/stress/tricky-infinite-loop-that-uses-captured-variables-and-creates-the-activation-outside-the-loop.js: Added.
(bar):
(foo.f):
(foo):
* tests/stress/tricky-infinite-loop-that-uses-captured-variables.js: Added.
(bar):
(foo):
(noInline):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165995 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoBuild fixes for Win64.
alex.christensen@flexsim.com [Thu, 20 Mar 2014 20:49:01 +0000 (20:49 +0000)]
Build fixes for Win64.
https://bugs.webkit.org/show_bug.cgi?id=130481

Reviewed by Brent Fulgham.

* WebKit.vcxproj/WebKit.sln:
Don't build ANGLE projects for now, corrected build configurations,
and build WebInspectorUI.
* WebKit.vcxproj/WebKitExportGenerator/WebKitExports.def.in:
Added 64-bit linker symbol.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165994 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoFix build.
andersca@apple.com [Thu, 20 Mar 2014 20:41:03 +0000 (20:41 +0000)]
Fix build.

Rename WKPreferencesPrivate.h to WKPreferencesRefPrivate.h.

* UIProcess/API/C/WKPreferencesRefPrivate.h: Renamed from Source/WebKit2/UIProcess/API/C/WKPreferencesPrivate.h.
* UIProcess/API/Cocoa/WKPreferencesPrivate.h:
* WebKit2.xcodeproj/project.pbxproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165993 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoiOS build fix after r165979.
psolanki@apple.com [Thu, 20 Mar 2014 20:30:46 +0000 (20:30 +0000)]
iOS build fix after r165979.

* generate-export-file:
(preprocessorMacros):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165992 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[New Multicolumn] getClientRects returns wrong rectangle
hyatt@apple.com [Thu, 20 Mar 2014 20:26:28 +0000 (20:26 +0000)]
[New Multicolumn] getClientRects returns wrong rectangle
https://bugs.webkit.org/show_bug.cgi?id=128614

Reviewed by Sam Weinig.

Added fast/multicol/newmulticol/client-rects.html and beefed up existing
fast/multicol/client-rects.html to use more writing modes.

Source/WebCore:

* rendering/LayoutState.h:
Removed unused function.

* rendering/RenderFlowThread.h:
Make mapFlowToRegion virtual so that RenderMultiColumnFlowThread can
override it.

* rendering/RenderMultiColumnFlowThread.cpp:
(WebCore::RenderMultiColumnFlowThread::mapFromFlowToRegion):
* rendering/RenderMultiColumnFlowThread.h:
Implement mapFlowToRegion to correctly translate a flow thread position
into a column offset within a region.

* rendering/RenderMultiColumnSet.cpp:
(WebCore::RenderMultiColumnSet::columnTranslationForOffset):
(WebCore::RenderMultiColumnSet::adjustRegionBoundsFromFlowThreadPortionRect):
* rendering/RenderMultiColumnSet.h:
Refactored some code to make a columnTranslationForOffset helper that
can be used by RenderMultiColumnFlowThread.

LayoutTests:

* fast/multicol/client-rects.html:
* fast/multicol/newmulticol/client-rects-expected.html: Added.
* fast/multicol/newmulticol/client-rects.html: Added.
* platform/mac/fast/multicol/client-rects-expected.png:
* platform/mac/fast/multicol/client-rects-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165991 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoMerge AtomicString, Identifier
barraclough@apple.com [Thu, 20 Mar 2014 20:03:49 +0000 (20:03 +0000)]
Merge AtomicString, Identifier
https://bugs.webkit.org/show_bug.cgi?id=128624

Unreviewed build fix following r165982.

* WebProcess/Plugins/Netscape/NPJSObject.cpp:
    - update includes.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165990 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoIncorrect behavior when mutating a typed array during set.
oliver@apple.com [Thu, 20 Mar 2014 20:01:19 +0000 (20:01 +0000)]
Incorrect behavior when mutating a typed array during set.
https://bugs.webkit.org/show_bug.cgi?id=130428

Reviewed by Geoffrey Garen.

Source/JavaScriptCore:

This fixes a null derefence that occurs if a typed array
is mutated during the set() operation. The patch gets rid
of the "Quickly" version of setIndex that is assigning
JSValues of unknown type, as the numeric conversion can trigger
side effects that lead to neutering, and so we deref null.

* runtime/JSGenericTypedArrayView.h:
(JSC::JSGenericTypedArrayView::setIndex):
* runtime/JSGenericTypedArrayViewInlines.h:
(JSC::JSGenericTypedArrayView<Adaptor>::set):
(JSC::JSGenericTypedArrayView<Adaptor>::putByIndex):

LayoutTests:

Test behaviour of TypedArray.set(array) when we neuter the array.

* js/typed-array-mutated-during-set-expected.txt: Added.
* js/typed-array-mutated-during-set.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165989 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoWeb Inspector: DebuggerDashboardView looks scrunched debugging JSContext
commit-queue@webkit.org [Thu, 20 Mar 2014 19:59:26 +0000 (19:59 +0000)]
Web Inspector: DebuggerDashboardView looks scrunched debugging JSContext
https://bugs.webkit.org/show_bug.cgi?id=130527

Patch by Joseph Pecoraro <pecoraro@apple.com> on 2014-03-20
Reviewed by Timothy Hatcher.

Instead of using body.javascript to force a collapsed style,
actually use the toolbar's collapsed class.

* UserInterface/Views/DashboardContainerView.css:
(.toolbar.collapsed .dashboard-container):
* UserInterface/Views/Toolbar.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165988 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoWeb Inspector: probe sidebar should use overflow:scroll and split space evenly
commit-queue@webkit.org [Thu, 20 Mar 2014 19:46:53 +0000 (19:46 +0000)]
Web Inspector: probe sidebar should use overflow:scroll and split space evenly
https://bugs.webkit.org/show_bug.cgi?id=129627

Patch by Katie Madonna <madonnk@gmail.com> on 2014-03-20
Reviewed by Timothy Hatcher.

Use flexbox to split space evenly in the probe sidebar, which required changes to collapse
format, remove scrollbar styles from old timelapse branch.

* UserInterface/Views/ProbeDetailsSidebarPanel.css: Space elements evenly with flexbox
and make collapse correctly, remove scrollbar style.
* UserInterface/Views/ProbeSetDetailsSection.js: Removed inline style.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165987 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoPreserve original document URL for XSSAuditor reports
dbates@webkit.org [Thu, 20 Mar 2014 19:34:48 +0000 (19:34 +0000)]
Preserve original document URL for XSSAuditor reports

Merged from Blink (patch by Tom Sepez):
https://src.chromium.org/viewvc/blink?view=rev&revision=158406

Source/WebCore:

Test: http/tests/security/xssAuditor/report-script-tag-replace-state.html

* html/parser/XSSAuditor.cpp:
(WebCore::XSSAuditor::filterToken):
* html/parser/XSSAuditorDelegate.cpp:
(WebCore::buildConsoleError):
(WebCore::XSSAuditorDelegate::generateViolationReport):
(WebCore::XSSAuditorDelegate::didBlockScript):
* html/parser/XSSAuditorDelegate.h:
(WebCore::XSSInfo::XSSInfo):

LayoutTests:

* http/tests/security/xssAuditor/report-script-tag-replace-state-expected.txt: Added.
* http/tests/security/xssAuditor/report-script-tag-replace-state.html: Added.
* http/tests/security/xssAuditor/resources/echo-intertag.pl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165986 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[WebKit2] Implement NetworkingContext::sourceApplicationAuditData() for WebKit2
weinig@apple.com [Thu, 20 Mar 2014 19:32:50 +0000 (19:32 +0000)]
[WebKit2] Implement NetworkingContext::sourceApplicationAuditData() for WebKit2
https://bugs.webkit.org/show_bug.cgi?id=130528

Reviewed by Anders Carlsson.

* Platform/IPC/Connection.h:
* Platform/IPC/mac/ConnectionMac.cpp:
(IPC::Connection::getAuditToken):
Expose the audit data if available.

* NetworkProcess/mac/RemoteNetworkingContext.mm:
(WebKit::RemoteNetworkingContext::sourceApplicationAuditData):
* WebProcess/WebCoreSupport/mac/WebFrameNetworkingContext.mm:
(WebKit::WebFrameNetworkingContext::sourceApplicationAuditData):
Pass the audit data corresponding to the UIProcess.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165985 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoXSSAuditor bypass with frameset tags
dbates@webkit.org [Thu, 20 Mar 2014 19:32:16 +0000 (19:32 +0000)]
XSSAuditor bypass with frameset tags

Merged from Blink (patch by Tom Sepez):
https://src.chromium.org/viewvc/blink?view=rev&revision=157794

Source/WebCore:

Test: http/tests/security/xssAuditor/frameset-injection.html

* html/parser/XSSAuditor.cpp:
(WebCore::XSSAuditor::filterStartToken):
(WebCore::XSSAuditor::filterFrameToken):
* html/parser/XSSAuditor.h:

LayoutTests:

* http/tests/security/xssAuditor/frameset-injection-expected.txt: Added.
* http/tests/security/xssAuditor/frameset-injection.html: Added.
* http/tests/security/xssAuditor/resources/echo-intertag.pl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165984 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[Cocoa] Expose preference to enable telephone number detection
weinig@apple.com [Thu, 20 Mar 2014 19:31:27 +0000 (19:31 +0000)]
[Cocoa] Expose preference to enable telephone number detection
https://bugs.webkit.org/show_bug.cgi?id=130524

Reviewed by Anders Carlsson.

* UIProcess/API/Cocoa/WKPreferences.mm:
(-[WKPreferences _telephoneNumberDetectionIsEnabled]):
(-[WKPreferences _setTelephoneNumberDetectionIsEnabled:]):
* UIProcess/API/Cocoa/WKPreferencesInternal.h:
* UIProcess/API/Cocoa/WKPreferencesPrivate.h: Added.
* WebKit2.xcodeproj/project.pbxproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165983 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoMerge AtomicString, Identifier
barraclough@apple.com [Thu, 20 Mar 2014 19:26:31 +0000 (19:26 +0000)]
Merge AtomicString, Identifier
https://bugs.webkit.org/show_bug.cgi?id=128624

Reviewed by Geoff Garen.

Source/JavaScriptCore:

WTF::StringImpl currently supports two uniquing mechanism - AtomicString and
Identifer - that is one too many.

Remove Identifier in favour of AtomicString. Identifier had two interesting
mechanisms that we preserve.

(1) JSC API VMs each get their own string table, switch the string table on
    API entry/exit.
(2) JSC caches a pointer to the string table on the VM to avoid a thread
    specific access. Adds a new AtomicString::add method to support this.

* API/JSAPIWrapperObject.mm:
    - updated includes.
* JavaScriptCore.xcodeproj/project.pbxproj:
    - added IdentifierInlines.h.
* inspector/JSInjectedScriptHostPrototype.cpp:
* inspector/JSJavaScriptCallFramePrototype.cpp:
    - updated includes.
* interpreter/CallFrame.h:
(JSC::ExecState::atomicStringTable):
    - added, used via AtomicString::add to avoid thread-specific access.
* runtime/ConsolePrototype.cpp:
    - updated includes.
* runtime/Identifier.cpp:
(JSC::Identifier::add):
(JSC::Identifier::add8):
    - vm->smallStrings.singleCharacterStringRep now returns Atomic strings, use AtomicString::add.
* runtime/Identifier.h:
(JSC::Identifier::Identifier):
    - added ASSERTS.
(JSC::Identifier::add):
    - vm->smallStrings.singleCharacterStringRep now returns Atomic strings, use AtomicString::add.
* runtime/IdentifierInlines.h: Added.
(JSC::Identifier::add):
    - moved from Identifier.h, use AtomicString::add.
* runtime/JSCInlines.h:
    - added IdentifierInlines.h.
* runtime/JSLock.h:
    - removed IdentifierTable.
* runtime/PropertyNameArray.cpp:
    - updated includes.
* runtime/SmallStrings.cpp:
(JSC::SmallStringsStorage::SmallStringsStorage):
    - ensure all single character strings are Atomic.
* runtime/VM.cpp:
(JSC::VM::VM):
    - instantiate CommonIdentifiers with the correct AtomicStringTable set on thread data.
* runtime/VM.h:
(JSC::VM::atomicStringTable):
    - added, used via AtomicString::add to avoid thread-specific access.

Source/WebKit:

* WebKit.vcxproj/WebKitExportGenerator/WebKitExports.def.in:
    - exports no longer needed.

Source/WTF:

WTF::StringImpl currently supports two uniquing mechanism - AtomicString and
Identifer - that is one too many.

Remove Identifier in favour of AtomicString. Identifier had two interesting
mechanisms that we preserve.

(1) JSC API VMs each get their own string table, switch the string table on
    API entry/exit.
(2) JSC caches a pointer to the string table on the VM to avoid a thread
    specific access. Adds a new AtomicString::add method to support this.

* wtf/WTFThreadData.cpp:
(WTF::WTFThreadData::WTFThreadData):
    - remove allocation of IdentifierTable.
(WTF::WTFThreadData::~WTFThreadData):
    - remove deletion of IdentifierTable.
* wtf/WTFThreadData.h:
(WTF::WTFThreadData::atomicStringTable):
    - table is now switched by JSC API, return the current table.
(WTF::WTFThreadData::currentIdentifierTable):
    - now returns the current AtomicStringTable.
(WTF::WTFThreadData::setCurrentIdentifierTable):
    - now sets the current AtomicStringTable.
(WTF::WTFThreadData::resetCurrentIdentifierTable):
    - now resets the AtomicStringTable.
* wtf/text/AtomicString.cpp:
(WTF::AtomicString::addSlowCase):
    - add without thread-specific access to access string table.
* wtf/text/AtomicString.h:
(WTF::AtomicString::addWithStringTableProvider):
    - add without thread-specific access (used by JSC, string table provided by VM or ExecState).
* wtf/text/AtomicStringTable.cpp:
(WTF::AtomicStringTable::create):
    - renamed m_atomicStringTable -> m_defaultAtomicStringTable.
(WTF::AtomicStringTable::~AtomicStringTable):
(WTF::AtomicStringTable::destroy):
    - clearing of table moved from AtomicStringTable::destroy to destructor.
* wtf/text/AtomicStringTable.h:
    - added destructor.
* wtf/text/StringImpl.cpp:
(WTF::StringImpl::~StringImpl):
    - Identifers no longer exist; no need to remove them from IdentifierTable.
* wtf/text/StringImpl.h:
(WTF::StringImpl::StringImpl):
    - removed s_hashFlagIsIdentifier.
(WTF::StringImpl::flagIsIdentifier):
    - s_hashFlagIsIdentifier -> s_hashFlagIsAtomic.
(WTF::StringImpl::isIdentifier):
    - now synonymous to isAtomic().
* wtf/text/StringStatics.cpp:
(WTF::StringImpl::hashAndFlagsForEmptyUnique):
    - removed s_hashFlagIsIdentifier.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165982 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoWeb Replay: fix some oversights in the replay backend
bburg@apple.com [Thu, 20 Mar 2014 18:49:59 +0000 (18:49 +0000)]
Web Replay: fix some oversights in the replay backend
https://bugs.webkit.org/show_bug.cgi?id=130435

Reviewed by Andreas Kling.

This functionality will be covered by tests in https://webkit.org/b/130145.

* inspector/InspectorInstrumentation.cpp:
(WebCore::InspectorInstrumentation::playbackFinishedImpl):
* inspector/InspectorInstrumentation.h:
(WebCore::InspectorInstrumentation::playbackFinished):
* inspector/InspectorReplayAgent.cpp:
(WebCore::buildInspectorObjectForSegment):
(WebCore::InspectorReplayAgent::playbackFinished):
* inspector/InspectorReplayAgent.h: Add a playbackFinished event to Inspector
instrumentation. This is fired by ReplayController when playback is cancelled or
playback of the final segment has completed.

* replay/EventLoopInputDispatcher.cpp:
(WebCore::EventLoopInputDispatcher::dispatchInputSoon): Fix wrong cast.
(WebCore::EventLoopInputDispatcher::dispatchInput): Ensure that m_dispatching is
false when m_running is false.
* replay/EventLoopInputDispatcher.h: Distinguish 'dispaching' and 'running' in comments.
(WebCore::EventLoopInputDispatcher::isRunning): Added.

* replay/FunctorInputCursor.h: Since the base class InputCursor is RefCounted, the
Functor subclass cannot be put into a unique_ptr. Add create() method instead.

* replay/ReplayController.cpp:
(WebCore::ReplayController::loadSegmentAtIndex): Renamed from loadSegment(). Callers
now pass in the desired segment index rather than the segment itself. The index is
necessary to properly reset m_position after the segment is loaded.

(WebCore::ReplayController::startPlayback): Set the dispatcher's speed before run().
(WebCore::ReplayController::pausePlayback): Only pause the dispatcher if it's running.
(WebCore::ReplayController::cancelPlayback): Add playbackFinished instrumentation.
(WebCore::ReplayController::replayToPosition):
(WebCore::ReplayController::didDispatchFinalInput): If there are no more segments to load,
manually adjust the final m_position and defer to cancelPlayback to unload the segment.
* replay/ReplayController.h:
* replay/ReplaySessionSegment.cpp:
(WebCore::ReplaySessionSegment::createFunctorCursor):
* replay/ReplaySessionSegment.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165981 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[CSS Regions] Cleanup scrollable regions tests
stavila@adobe.com [Thu, 20 Mar 2014 18:44:40 +0000 (18:44 +0000)]
[CSS Regions] Cleanup scrollable regions tests
https://bugs.webkit.org/show_bug.cgi?id=130505

Reviewed by Mihnea Ovidenie.

Tests should only display red in case of failure. Purpose of the test should be very clearly stated.

* fast/regions/scrollable-last-region-expected.html:
* fast/regions/scrollable-last-region.html:
* fast/regions/scrollable-regions-hit-testing-inside-expected.html:
* fast/regions/scrollable-regions-hit-testing-inside.html:
* fast/regions/scrollable-single-region-bt-expected.html:
* fast/regions/scrollable-single-region-bt.html:
* fast/regions/scrollable-single-region-expected.html:
* fast/regions/scrollable-single-region-lr-expected.html:
* fast/regions/scrollable-single-region-lr.html:
* fast/regions/scrollable-single-region-relative-element-expected.html:
* fast/regions/scrollable-single-region-relative-element.html:
* fast/regions/scrollable-single-region-rl-expected.html:
* fast/regions/scrollable-single-region-rl.html:
* fast/regions/scrollable-single-region.html:
* fast/regions/wheel-scroll-abspos-expected.html:
* fast/regions/wheel-scroll-abspos.html:
* fast/regions/wheel-scroll-expected.html:
* fast/regions/wheel-scroll.html:
* fast/regions/wheel-scrollable-single-region-expected.html:
* fast/regions/wheel-scrollable-single-region.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165980 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[iOS] Get code to compile on older iOS versions
psolanki@apple.com [Thu, 20 Mar 2014 18:40:36 +0000 (18:40 +0000)]
[iOS] Get code to compile on older iOS versions
https://bugs.webkit.org/show_bug.cgi?id=130142
<rdar://problem/16302908>

Reviewed by Darin Adler.

Source/WebCore:

* WebCore.exp.in:
* platform/ios/WebVideoFullscreenControllerAVKit.mm:
(-[WebVideoFullscreenController WebCore::]):
(-[WebVideoFullscreenController enterFullscreen:]):
(-[WebVideoFullscreenController exitFullscreen]):
* platform/ios/WebVideoFullscreenInterfaceAVKit.h:
* platform/ios/WebVideoFullscreenInterfaceAVKit.mm:
* platform/mac/HTMLConverter.mm:
(_dateForString):
* platform/network/cf/CookieJarCFNet.cpp:
(WebCore::copyCookiesForURLWithFirstPartyURL):
* platform/text/ios/LocalizedDateCache.mm:
(WebCore::LocalizedDateCache::calculateMaximumWidth):
* platform/text/mac/LocaleMac.mm:
(WebCore::LocaleMac::LocaleMac):

Source/WebKit/mac:

* History/WebHistory.mm:
(getDayBoundaries):
* Plugins/WebPluginController.h:
* Plugins/WebPluginController.mm:
(+[WebPluginController plugInViewWithArguments:fromPluginPackage:]):
* WebCoreSupport/WebFrameLoaderClient.mm:
(pluginView):
* WebView/WebPreferences.mm:
(-[WebPreferences _setAllowCompositingLayerVisualDegradation:]):
* WebView/WebPreferencesPrivate.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165979 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoFollowup build fix: AnimationBase.h needs to be private because
simon.fraser@apple.com [Thu, 20 Mar 2014 18:34:51 +0000 (18:34 +0000)]
Followup build fix: AnimationBase.h needs to be private because
it's included by AnimationController.h.

* WebCore.xcodeproj/project.pbxproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165978 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoA completed fill-forwards animation should not disable overlap testing
simon.fraser@apple.com [Thu, 20 Mar 2014 18:32:13 +0000 (18:32 +0000)]
A completed fill-forwards animation should not disable overlap testing
https://bugs.webkit.org/show_bug.cgi?id=130522
<rdar://problem/15862395>

Source/WebCore:

Reviewed by Dean Jackson.

Previously, if -webkit-transform was being keyframe-animated, and the
animation had fill-forwards, then we would continue to think that the
animation is running and turn off compositing overlap testing. This
caused some sites to keep too much backing store around.

Fix by having isRunning{Accelerated}AnimationOnRenderer take some flags
so that more specific questions about the running state can be asked.
For layer creation, keep using the same criteria as before (for now)
which includes paused and fill-forwards animations. For overlap testing,
don't include the fill-forwards test.

Tests: compositing/animation/filling-animation-overlap-at-end.html
       compositing/animation/filling-animation-overlap.html
       compositing/animation/layer-for-filling-animation.html

* page/animation/AnimationBase.h:
(WebCore::AnimationBase::fillingForwards):
(WebCore::AnimationBase::inPausedState):
(WebCore::AnimationBase::isAnimatingProperty):
* page/animation/AnimationController.cpp:
(WebCore::AnimationControllerPrivate::isRunningAnimationOnRenderer):
(WebCore::AnimationControllerPrivate::isRunningAcceleratedAnimationOnRenderer):
(WebCore::AnimationController::isRunningAnimationOnRenderer):
(WebCore::AnimationController::isRunningAcceleratedAnimationOnRenderer):
* page/animation/AnimationController.h:
* page/animation/AnimationControllerPrivate.h:
* page/animation/CompositeAnimation.cpp:
(WebCore::CompositeAnimation::isAnimatingProperty):
* page/animation/CompositeAnimation.h:
* rendering/RenderLayerBacking.cpp:
(WebCore::RenderLayerBacking::updateGraphicsLayerGeometry):
* rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::requiresCompositingForAnimation):
(WebCore::RenderLayerCompositor::isRunningAcceleratedTransformAnimation):

LayoutTests:

Reviewed by Dean Jackson.

Tests that dump layer trees when a fill-forwards animation has finished,
both when the final keyframe causes overlap, and when it does not.

* compositing/animation/filling-animation-overlap-at-end-expected.txt: Added.
* compositing/animation/filling-animation-overlap-at-end.html: Added.
* compositing/animation/filling-animation-overlap-expected.txt: Added.
* compositing/animation/filling-animation-overlap.html: Added.
* compositing/animation/layer-for-filling-animation-expected.txt: Added.
* compositing/animation/layer-for-filling-animation.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165977 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoCanvas methods clip/fill/stroke should not except 0 argument
krit@webkit.org [Thu, 20 Mar 2014 18:24:34 +0000 (18:24 +0000)]
Canvas methods clip/fill/stroke should not except 0 argument
https://bugs.webkit.org/show_bug.cgi?id=130497

Reviewed by Dean Jackson.

Source/WebCore:

Do not allow nullable argument for fill/stroke/clip.

No new tests. Updated existing tests.

* html/canvas/CanvasRenderingContext2D.idl:

LayoutTests:

Add more exception tests for fill/stroke/clip.

* fast/canvas/canvas-clip-path-expected.txt:
* fast/canvas/canvas-fill-path-expected.txt:
* fast/canvas/canvas-stroke-path-expected.txt:
* fast/canvas/script-tests/canvas-clip-path.js:
* fast/canvas/script-tests/canvas-fill-path.js:
* fast/canvas/script-tests/canvas-stroke-path.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165976 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[GTK][CMake] Add support for building with Clang
zandobersek@gmail.com [Thu, 20 Mar 2014 18:22:54 +0000 (18:22 +0000)]
[GTK][CMake] Add support for building with Clang
https://bugs.webkit.org/show_bug.cgi?id=130260

Reviewed by Martin Robinson.

.:

* Source/PlatformGTK.cmake: Run the generate-gtkdoc script with the CFLAGS env set
to the list C compiler flags as constructed by CMake.
* Source/cmake/OptionsCommon.cmake: When building with Clang, use the -Qunused-arguments
compiler option to suppress verbose warnings about arguments that went unused by the driver.
* Source/cmake/OptionsGTK.cmake: Use -fno-omit-frame-pointer for both GCC and Clang.
-fno-tree-dce should be used only for GCC.
* Source/cmake/gtest/CMakeLists.txt: When building with Clang the GTest framework
should be built with GTEST_HAS_TR1_TUPLE macro defined to 0 to avoid build failures.

Tools:

* TestWebKitAPI/CMakeLists.txt: When building with Clang TestWebKitAPI should be
built with GTEST_HAS_TR1_TUPLE defined to 0 to avoid build failures occuring in the
included GTest framework headers.
* gtk/generate-gtkdoc:
(prepare_environment_for_gtkdoc_generation): Don't modify the CFLAGS env anymore
in case of using Clang since that is now set by the CMake rule.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165975 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUnreviewed GTK Gardening.
mrobinson@webkit.org [Thu, 20 Mar 2014 18:18:32 +0000 (18:18 +0000)]
Unreviewed GTK Gardening.

Update expectation of accessibility/table-detection.html after r165535.
Update expectation of accessibility/aria-hidden-false-works-in-subtrees.html.

Patch by Carlos Alberto Lopez Perez <clopez@igalia.com> on 2014-03-20

* platform/gtk/accessibility/aria-hidden-false-works-in-subtrees-expected.txt:
* platform/gtk/accessibility/table-detection-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165974 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUnreviewed Clang build fix for the GTK port after r165952.
zandobersek@gmail.com [Thu, 20 Mar 2014 18:09:52 +0000 (18:09 +0000)]
Unreviewed Clang build fix for the GTK port after r165952.

* wtf/gobject/GMainLoopSource.h: Include the <functional> header for the std::function<> objects.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165973 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[Mac] Support asynchronous NSTextInputClient
ap@apple.com [Thu, 20 Mar 2014 17:45:04 +0000 (17:45 +0000)]
[Mac] Support asynchronous NSTextInputClient
https://bugs.webkit.org/show_bug.cgi?id=130479

Reviewed by Anders Carlsson.

The implementation is currently disabled, pending lower level support.
Most of the code is not under compile time guard however, to facilitate cross-platform
reuse, or at least under a PLATFORM(COCOA) guard to share the code with iOS.

* UIProcess/API/mac/WKView.mm: Added a compile time branch for USE(ASYNC_NSTEXTINPUTCLIENT).
We still implement sync NSTextInputClient here, in order to get assertions when
its methods are unexpectedly called.
The new code first sends an event to input method asynchronously, handling any callbacks
that may arrive. During this time, we no longer care about WKViewInterpretKeyEventsParameters
at all. Once done, we interpret key bindings synchronously, collecting them into
a vector.

* UIProcess/API/mac/WKViewInternal.h: We no longer expose _interpretKeyEvent outside
WKView.

* UIProcess/WebPageProxy.cpp:
* UIProcess/WebPageProxy.h:
Added async calls and callbacks. Removed unnecessary and slightly harmful .get() when moving
a callback pointer into map. Moved insertDictatedText() and getAttributedSubstringFromRange()
from PLATFORM(COCOA) to PLATFORM(MAC), because they are unused and unimplemented on
iOS, and unlikely to be needed any time soon. Changed USE(APPKIT) to PLATFORM(MAC),
because that's more accurate in this case (nothing depends on AppKit, it's just code
that we only need on Mac).

* UIProcess/WebPageProxy.messages.in: Added messages for new async IM responses.

* UIProcess/ios/WebPageProxyIOS.mm: Removed insertDictatedText() and getAttributedSubstringFromRange().

* UIProcess/mac/WebPageProxyMac.mm:
(WebKit::WebPageProxy::insertDictatedTextAsync):
(WebKit::WebPageProxy::attributedSubstringForCharacterRangeAsync):
(WebKit::WebPageProxy::attributedStringForCharacterRangeCallback):
Added async calls and callbacks that are Mac only.

* WebProcess/WebPage/WebPage.cpp:
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/mac/WebPageMac.mm:
Added async implementations (which are essentially the same as sync ones, sadly).

* WebProcess/WebPage/WebPage.messages.in: Added async messages, moved some messages
under PLATFORM(MAC).

* WebProcess/WebPage/ios/WebPageIOS.mm: More of deleting functions that are Mac only,
and cannot be easily implemented in WebPage.cpp with shared code.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165972 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[iOS] Support MediaDocument when building with PLUGIN_PROXY_FOR_VIDEO
eric.carlson@apple.com [Thu, 20 Mar 2014 17:37:57 +0000 (17:37 +0000)]
[iOS] Support MediaDocument when building with PLUGIN_PROXY_FOR_VIDEO
https://bugs.webkit.org/show_bug.cgi?id=130520

Reviewed by Jer Noble.

* dom/DOMImplementation.cpp:
(WebCore::DOMImplementation::createDocument): Check the isVideoPluginProxyEnabled setting
    when building with ENABLE_PLUGIN_PROXY_FOR_VIDEO.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165971 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[CSS Blending] Blending operation is not isolated when setting z-index on parent...
commit-queue@webkit.org [Thu, 20 Mar 2014 17:37:16 +0000 (17:37 +0000)]
[CSS Blending] Blending operation is not isolated when setting z-index on parent from javascript
https://bugs.webkit.org/show_bug.cgi?id=129307

Patch by Mihai Tica <mitica@adobe.com> on 2014-03-20
Reviewed by Dean Jackson.

Source/WebCore:

Reimplement the logic behind isolating blend modes. The blending operation should be limited to the
contents of the parent stacking context. The proposed solution implies adding logic in updateDescendantDependentFlags,
to track child elements that have blending and create a transparency layer that triggers isolation.

Tests: css3/compositing/blend-mode-background.html
       fast/repaint/blend-mode-isolate-stacking-context.html

* rendering/RenderLayer.cpp:
(WebCore::RenderLayer::RenderLayer): Remove old members, add new members to constructor.
(WebCore::RenderLayer::updateBlendMode): Update parent stacking context isolation on blend mode change.
(WebCore::RenderLayer::updateNonCompositedParentStackingContextHasBlendedChild):
- Set m_hasBlendedElementInChildStackingContext for the non composited parent stacking context.
(WebCore::RenderLayer::dirtyAncestorParentStackingContextHasBlendedElement):
- Set flag so isolation should be considered in updateDescendantDependentFlags.
(WebCore::RenderLayer::nonCompositedParentStackingContextHasBlendedChild):
- Return whether the non composited parent stacking context isolates blending.
(WebCore::RenderLayer::updateDescendantDependentFlags): Add logic for isolating blending.
(WebCore::RenderLayer::addChild): Call to updateNonCompositedParentStackingContextHasBlendedChild if needed.
(WebCore::RenderLayer::removeChild): Call to dirtyAncestorParentStackingContextHasBlendedElement if needed.
(WebCore::RenderLayer::updateStackingContextsAfterStyleChange): Update isolation when a stacking context was set or unset.
(WebCore::RenderLayer::styleChanged): Removed oldStyle parameter from updateBlendMode call.
 Update parent stacking context isolation on blend mode change.
* rendering/RenderLayer.h:
- Add members, getters and setters for tracking whether the element has a blended child:
    - m_hasBlendedElementInChildStackingContext
    - m_hasBlendedElementInChildStackingContextStatusDirty
- Removed m_updateParentStackingContextShouldIsolateBlendingDirty (replaced with m_hasBlendedElementInChildStackingContextStatusDirty)
- Removed m_isolatesBlending (replaced with m_hasBlendedElementInChildStackingContext)
- Updated the isolatesBlending method to return true only when the element is a stacking context and has a blended child.
* rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::computeCompositingRequirements): Removed layer.updateParentStackingContextShouldIsolateBlending call.

LayoutTests:

Test if setting or unsetting stacking context or blend mode from script correctly updates blending
isolation.

* css3/compositing/blend-mode-background.html: Added.
* fast/repaint/blend-mode-isolate-stacking-context-expected.txt: Added.
* fast/repaint/blend-mode-isolate-stacking-context.html: Added.
* platform/mac/css3/compositing/blend-mode-background-expected.png: Added.
* platform/mac/css3/compositing/blend-mode-background-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165970 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[CSS Regions] Remove "visibility:hidden" from overflow-scrollable* tests
stavila@adobe.com [Thu, 20 Mar 2014 17:28:53 +0000 (17:28 +0000)]
[CSS Regions] Remove "visibility:hidden" from overflow-scrollable* tests
https://bugs.webkit.org/show_bug.cgi?id=130510

Reviewed by Andrei Bucur.

The visibility hidden option was needed to prevent these tests from failing before the
visual overflow was properly implemented (bug 118665). It is now no longer required.

* fast/regions/overflow-scrollable-rel-pos-fragment-expected.html:
* fast/regions/overflow-scrollable-rel-pos-fragment.html:
* fast/regions/overflow-scrollable-rotated-fragment-expected.html:
* fast/regions/overflow-scrollable-rotated-fragment.html:
* fast/regions/overflow-scrollable-unsplittable-fragment-expected.html:
* fast/regions/overflow-scrollable-unsplittable-fragment.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165969 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoRemove remaining GTK+ unused parameter warnings from WebKit2
mrobinson@webkit.org [Thu, 20 Mar 2014 17:20:58 +0000 (17:20 +0000)]
Remove remaining GTK+ unused parameter warnings from WebKit2
https://bugs.webkit.org/show_bug.cgi?id=130518

Reviewed by Sergio Villar Senin.

Fix almost all unused parameter warnings in the WebKit2GTK+ build.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165968 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[GTK] [CMake] Do not compile ProxyResolverSoup.cpp
mrobinson@webkit.org [Thu, 20 Mar 2014 17:05:48 +0000 (17:05 +0000)]
[GTK] [CMake] Do not compile ProxyResolverSoup.cpp
https://bugs.webkit.org/show_bug.cgi?id=130516

Reviewed by Sergio Villar Senin.

* PlatformGTK.cmake: Remove the file from the source list.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165967 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUnreviewed, rolling out r165962.
commit-queue@webkit.org [Thu, 20 Mar 2014 15:34:56 +0000 (15:34 +0000)]
Unreviewed, rolling out r165962.
https://bugs.webkit.org/show_bug.cgi?id=130512

It broke the build (Requested by Ossy on #webkit).

Reverted changeset:

.:

"[EFL][GTK] Get CMake to find Freetype2 properly"
https://bugs.webkit.org/show_bug.cgi?id=130150
http://trac.webkit.org/changeset/165962

Source/WebCore:

"[EFL][GTK] Get CMake to find Freetype2 properly"
https://bugs.webkit.org/show_bug.cgi?id=130150
http://trac.webkit.org/changeset/165962

Source/WebKit:

"[EFL][GTK] Get CMake to find Freetype2 properly"
https://bugs.webkit.org/show_bug.cgi?id=130150
http://trac.webkit.org/changeset/165962

Source/WebKit2:

"[EFL][GTK] Get CMake to find Freetype2 properly"
https://bugs.webkit.org/show_bug.cgi?id=130150
http://trac.webkit.org/changeset/165962

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165966 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago"input" event is not fired when changing a select element's value
commit-queue@webkit.org [Thu, 20 Mar 2014 15:09:33 +0000 (15:09 +0000)]
"input" event is not fired when changing a select element's value
https://bugs.webkit.org/show_bug.cgi?id=117526

Patch by Lorenzo Tilve <ltilve@igalia.com> on 2014-03-20
Reviewed by Antti Koivisto.

Source/WebCore:

Add a call to dispatch input event when listbox or menulist
values are modified.

According to the spec: "When the user agent is to send select update
notifications, queue a task to first fire a simple even that bubbles
named input at the select element, and then fire a simple event that
bubbles named change at the select element, using the user interaction
task source as the task source."

Tests: fast/forms/select/listbox-oninput-fired.html
       fast/forms/select/menulist-oninput-fired.html

* html/HTMLSelectElement.cpp:
(WebCore::HTMLSelectElement::listBoxOnChange):
(WebCore::HTMLSelectElement::dispatchChangeEventForMenuList):

LayoutTests:

Add tests to verify that input event is dispatched when listbox
or menulist values are modified.

* fast/forms/select/listbox-oninput-fired-expected.txt: Added.
* fast/forms/select/listbox-oninput-fired.html: Added.
* fast/forms/select/menulist-oninput-fired-expected.txt: Added.
* fast/forms/select/menulist-oninput-fired.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165965 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[CSS Regions] The height of a scrollable element flowed inside a scrollable region...
stavila@adobe.com [Thu, 20 Mar 2014 15:01:12 +0000 (15:01 +0000)]
[CSS Regions] The height of a scrollable element flowed inside a scrollable region is not computed correctly
https://bugs.webkit.org/show_bug.cgi?id=129487

Reviewed by Andrei Bucur.

Source/WebCore:

When propagating the layout overflow from the flowed element up to the flow thread itself,
the region's logical bottom padding and the element's logical bottom margin must be taken into consideration.
However, if the flowed element has overflow, the overflow will be painted over the region's padding
and the element's margin, in order to be compatible with the non-regions case.
More info can be found here: http://lists.w3.org/Archives/Public/www-style/2014Jan/0089.html

Tests: fast/regions/scrollable-region-content-size-horiz-bt.html
       fast/regions/scrollable-region-content-size-no-overflow-horiz-bt.html
       fast/regions/scrollable-region-content-size-no-overflow-vert-rl.html
       fast/regions/scrollable-region-content-size-no-overflow.html
       fast/regions/scrollable-region-content-size-relative.html
       fast/regions/scrollable-region-content-size-vert-rl.html
       fast/regions/scrollable-region-content-size-vertical-lr.html
       fast/regions/scrollable-region-content-size.html
       fast/regions/scrollable-region-scrollable-content.html

* rendering/RenderFlowThread.cpp:
(WebCore::RenderFlowThread::addRegionsOverflowFromChild):
* rendering/RenderRegion.cpp:
(WebCore::RenderRegion::rectFlowPortionForBox):

LayoutTests:

Added tests for different situations using scrollable regions and fixed a couple existing tests.

* fast/regions/clip-to-padding-box-margins.html: overflow-x and overflow-y cannot be hidden & visible.
* fast/regions/scrollable-last-region-expected.html: "hack" no longer needed now that the scroll content size is properly computed.
* fast/regions/scrollable-region-content-size-expected.html: Added.
* fast/regions/scrollable-region-content-size-horiz-bt-expected.html: Added.
* fast/regions/scrollable-region-content-size-horiz-bt.html: Added.
* fast/regions/scrollable-region-content-size-no-overflow-expected.html: Added.
* fast/regions/scrollable-region-content-size-no-overflow-horiz-bt-expected.html: Added.
* fast/regions/scrollable-region-content-size-no-overflow-horiz-bt.html: Added.
* fast/regions/scrollable-region-content-size-no-overflow-vert-rl-expected.html: Added.
* fast/regions/scrollable-region-content-size-no-overflow-vert-rl.html: Added.
* fast/regions/scrollable-region-content-size-no-overflow.html: Added.
* fast/regions/scrollable-region-content-size-relative-expected.html: Added.
* fast/regions/scrollable-region-content-size-relative.html: Added.
* fast/regions/scrollable-region-content-size-vert-rl-expected.html: Added.
* fast/regions/scrollable-region-content-size-vert-rl.html: Added.
* fast/regions/scrollable-region-content-size-vertical-lr-expected.html: Added.
* fast/regions/scrollable-region-content-size-vertical-lr.html: Added.
* fast/regions/scrollable-region-content-size.html: Added.
* fast/regions/scrollable-region-scrollable-content-expected.html: Added.
* fast/regions/scrollable-region-scrollable-content.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165964 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoSubpixel rendering: Nested layers with subpixel accumulation paint to wrong position.
zalan@apple.com [Thu, 20 Mar 2014 14:20:22 +0000 (14:20 +0000)]
Subpixel rendering: Nested layers with subpixel accumulation paint to wrong position.
https://bugs.webkit.org/show_bug.cgi?id=130153

Reviewed by Simon Fraser.

Subpixels (fractional device pixels here) can accumulate through nested layers. Subpixels
need to be propagated through the layer tree so that painting coordinates match layout coordinates.

Subpixel accumulation through nesting (absolute positioning, 2x display):
non-compositing case:
(nested boxes)  (layout pos) (norm.paint pos) (translate accumulation, subpixel accumulation, final paint pos)
div -> top: 1.3px   1.3px     1.5px      1.5px 0.2px -> snapped 0.0px -> 1.5px
 div -> top: 1.3px   2.6px     2.5px      3.0px 0.4px -> snapped 0.5px -> 2.5px
  div -> top: 1.3px   3.9px     4.0px      4.5px 0.6px -> snapped 0.5px -> 4.0px
   div -> top: 1.3px   5.2px     5.0px      6.0px 0.8px -> snapped 1.0px -> 5.0px

compositing case:
(nested boxes)  (layout pos) (norm.paint pos) (device pixel offset + fractional offset, final pos)
div -> top: 1.3px   1.3px      1.5px      1.0px + 0.3px -> snapped -> 1.5px
 div -> top: 1.3px   2.6px      2.5px      2.5px + 0.1px -> snapped -> 2.5px
  div -> top: 1.3px   3.9px      4.0px      3.5px + 0.4px -> snapped -> 4.0px
   div -> top: 1.3px   5.2px      5.0px      5.0px + 0.2px -> snapped -> 5.0px

Source/WebCore:

Tests: compositing/hidpi-nested-compositing-layers-with-subpixel-accumulation.html
       fast/layers/hidpi-nested-layers-with-subpixel-accumulation.html

* rendering/RenderLayer.cpp:
(WebCore::RenderLayer::paintLayerByApplyingTransform):
* rendering/RenderLayerBacking.cpp:
(WebCore::RenderLayerBacking::updateGraphicsLayerGeometry):
* rendering/RenderLayerBacking.h:
(WebCore::RenderLayerBacking::devicePixelFractionFromRenderer):

LayoutTests:

* compositing/hidpi-nested-compositing-layers-with-subpixel-accumulation-expected.html: Added.
* compositing/hidpi-nested-compositing-layers-with-subpixel-accumulation.html: Added.
* fast/layers/hidpi-nested-layers-with-subpixel-accumulation-expected.html: Added.
* fast/layers/hidpi-nested-layers-with-subpixel-accumulation.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165963 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[EFL][GTK] Get CMake to find Freetype2 properly
thiago.lacerda@openbossa.org [Thu, 20 Mar 2014 14:19:00 +0000 (14:19 +0000)]
[EFL][GTK] Get CMake to find Freetype2 properly
https://bugs.webkit.org/show_bug.cgi?id=130150

Reviewed by Csaba Osztrogon√°c.

Newer versions of CMake are not able to find Freetype2 correctly.
FindFreetype2.cmake checks for the version number in freetype.h header, since it seems that bug exists in
freetype2 repo, because a completely different version number is found in freetype2.pc.

.:

* Source/cmake/FindFreetype2.cmake: Added.
* Source/cmake/OptionsEfl.cmake:
* Source/cmake/OptionsGTK.cmake:

Source/WebCore:

* PlatformEfl.cmake:
* PlatformGTK.cmake:

Source/WebKit:

* PlatformEfl.cmake:

Source/WebKit2:

* PlatformEfl.cmake:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165962 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[ARM64] Fix assembler build issues and add cacheFlush support for Linux
rgabor@webkit.org [Thu, 20 Mar 2014 14:03:43 +0000 (14:03 +0000)]
[ARM64] Fix assembler build issues and add cacheFlush support for Linux
https://bugs.webkit.org/show_bug.cgi?id=130502

Reviewed by Michael Saboff.

Add limits.h for INT_MIN in ARM64Assembler(). Delete shouldBlindForSpecificArch(uintptr_t)
because on ARM64 uint64_t and uintptr_t is the same with GCC and Clang as well.
Add cacheFlush support for Linux.

* assembler/ARM64Assembler.h:
(JSC::ARM64Assembler::linuxPageFlush):
(JSC::ARM64Assembler::cacheFlush):
* assembler/MacroAssemblerARM64.h:
(JSC::MacroAssemblerARM64::shouldBlindForSpecificArch):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165961 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[EFL][WK2] test_ewk2_application_cache_manager is flaky
ryuan.choi@samsung.com [Thu, 20 Mar 2014 11:04:22 +0000 (11:04 +0000)]
[EFL][WK2] test_ewk2_application_cache_manager is flaky
https://bugs.webkit.org/show_bug.cgi?id=130501

Reviewed by Gyuyoung Kim.

* UIProcess/API/efl/tests/test_ewk2_application_cache_manager.cpp:
(timerCallback): Return ECORE_CALLBACK_CANCEL to clear the timer.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165959 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUnreviewed EFL gardening
k.czech@samsung.com [Thu, 20 Mar 2014 10:55:45 +0000 (10:55 +0000)]
Unreviewed EFL gardening

Update expectation of table-detection.html after r165535.

* platform/efl-wk2/accessibility/table-detection-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165958 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoVersioning.
lforschler@apple.com [Thu, 20 Mar 2014 10:19:51 +0000 (10:19 +0000)]
Versioning.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165957 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoRefactor checking log channel state in MediaPlayerPrivateGStreamerBase.
pmolnar.u-szeged@partner.samsung.com [Thu, 20 Mar 2014 10:07:01 +0000 (10:07 +0000)]
Refactor checking log channel state in MediaPlayerPrivateGStreamerBase.
https://bugs.webkit.org/show_bug.cgi?id=130241

Reviewed by Philippe Normand.

The channel was passed around only to check its state. Cleaned this up
by adding a method to check whether a log channel is enabled.

* platform/Logging.cpp:
(WebCore::isLogChannelEnabled):
* platform/Logging.h:
* platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.cpp:
(WebCore::MediaPlayerPrivateGStreamerBase::createVideoSink):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165955 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoDispatch input event for Range element keyboard event
commit-queue@webkit.org [Thu, 20 Mar 2014 08:56:36 +0000 (08:56 +0000)]
Dispatch input event for Range element keyboard event
https://bugs.webkit.org/show_bug.cgi?id=129929

Patch by Lorenzo Tilve <ltilve@igalia.com> on 2014-03-20
Reviewed by Kent Tamura.

For any value update, dispatch both input and change event.
Also removes extra dispatch change event, as event is dispatched
from TextFieldInputType:setValue.

Merged from Blink (patch by Habib Virji):
https://src.chromium.org/viewvc/blink?revision=168558&view=revision

Source/WebCore:

Test: fast/forms/range/range-keyboard-oninput-event.html

* html/RangeInputType.cpp:
(WebCore::RangeInputType::handleKeydownEvent):

LayoutTests:

* fast/forms/range/range-keyboard-oninput-event-expected.txt: Added.
* fast/forms/range/range-keyboard-oninput-event.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165954 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago Rollout r165944. Build break.
lforschler@apple.com [Thu, 20 Mar 2014 08:28:32 +0000 (08:28 +0000)]
Rollout r165944.  Build break.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165953 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[GLIB] Add GMainLoopSource class to wrap idle and timeout sources
carlosgc@webkit.org [Thu, 20 Mar 2014 07:56:38 +0000 (07:56 +0000)]
[GLIB] Add GMainLoopSource class to wrap idle and timeout sources
https://bugs.webkit.org/show_bug.cgi?id=130027

Reviewed by Martin Robinson.

GLib main loop sources like idle and timeouts are sometimes
unconvenient to use and it's very common to forget canceling the
source when the object is destroyed or reset the source ID in the
callback when called. GMainLoopSource is a wrapper class to make it
easier to handle sources and also to avoid those typical mistakes.
It forces to create sources with a name and the user doesn't have
to deal with the source ID anymore. The source is cancelled when
the object is deleted or when a new source is scheduled. It uses
std::function for callbacks so that we no longer need to use the
"proxy" static callbacks either. We can use std::bind to use a
function pointer or a member or even lambda functions. It also handles
repeating sources automatically depending on whether the given
function returns a bool or not.

* GNUmakefile.list.am:
* wtf/PlatformEfl.cmake:
* wtf/PlatformGTK.cmake:
* wtf/RunLoop.h:
* wtf/gobject/GMainLoopSource.cpp: Added.
(WTF::GMainLoopSource::createAndDeleteOnDestroy):
(WTF::GMainLoopSource::GMainLoopSource):
(WTF::GMainLoopSource::~GMainLoopSource):
(WTF::GMainLoopSource::cancel):
(WTF::GMainLoopSource::reset):
(WTF::GMainLoopSource::scheduleIdleSource):
(WTF::GMainLoopSource::schedule):
(WTF::GMainLoopSource::scheduleTimeoutSource):
(WTF::GMainLoopSource::scheduleAfterDelay):
(WTF::GMainLoopSource::voidCallback):
(WTF::GMainLoopSource::boolCallback):
(WTF::GMainLoopSource::socketCallback):
(WTF::GMainLoopSource::destroy):
(WTF::GMainLoopSource::voidSourceCallback):
(WTF::GMainLoopSource::boolSourceCallback):
(WTF::GMainLoopSource::socketSourceCallback):
* wtf/gobject/GMainLoopSource.h: Added.
(WTF::GMainLoopSource::isScheduled):
* wtf/gtk/MainThreadGtk.cpp:
(WTF::scheduleDispatchFunctionsOnMainThread):
* wtf/gtk/RunLoopGtk.cpp:
(WTF::RunLoop::wakeUp):
(WTF::RunLoop::TimerBase::TimerBase):
(WTF::RunLoop::TimerBase::start):
(WTF::RunLoop::TimerBase::stop):
(WTF::RunLoop::TimerBase::isActive):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165952 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[CSSRegions] Content flowed in 0px height region not displayed
mihnea@adobe.com [Thu, 20 Mar 2014 07:14:15 +0000 (07:14 +0000)]
[CSSRegions] Content flowed in 0px height region not displayed
https://bugs.webkit.org/show_bug.cgi?id=98265

After the past work on overflow, this looks fixed. Add a test to make sure
we are covered for future regressions.

Reviewed by Andrei Bucur.

* fast/regions/overflow-content-zero-height-region-expected.html: Added.
* fast/regions/overflow-content-zero-height-region.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165951 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoRefine BatteryStatus module
jinwoo7.song@samsung.com [Thu, 20 Mar 2014 07:11:32 +0000 (07:11 +0000)]
Refine BatteryStatus module
https://bugs.webkit.org/show_bug.cgi?id=130495

Reviewed by Andreas Kling.

Small refinements:
 - Return PassRef instead of PassRefPtr in create method
 - Switched to nullptr instead of 0 where appropriate.
 - Removed unused function and header file.
 - Removed unnecessary empty lines.

* Modules/battery/BatteryController.h:
* Modules/battery/BatteryManager.cpp:
(WebCore::BatteryManager::create): Make return passRef.
* Modules/battery/BatteryManager.h:
(WebCore::BatteryManager::batteryControllerDestroyed): Use nullptr.
* Modules/battery/BatteryStatus.cpp:
(WebCore::BatteryStatus::create): Make return passRef.
* Modules/battery/BatteryStatus.h:
* Modules/battery/NavigatorBattery.cpp:
(WebCore::NavigatorBattery::webkitBattery): Remove batteryManager().
* Modules/battery/NavigatorBattery.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165950 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agohttps://bugs.webkit.org/show_bug.cgi?id=130494
barraclough@apple.com [Thu, 20 Mar 2014 06:21:49 +0000 (06:21 +0000)]
https://bugs.webkit.org/show_bug.cgi?id=130494
EmptyUnique strings are Identifiers/Atomic

Reviewed by Geoff Garen.

EmptyUnique strings should set the Identifier/Atomic flag.

Source/JavaScriptCore:

This fixes an unreproducible bug we believe exists in Identifier handling.
Expected behaviour is that while Identifiers may reference EmptyUniques
(StringImpls allocated as UIDs for PrivateNames), these are not created
through the main Identifier constructor, the Identifier flag is not set
on PrivateNames, and we should never lookup EmptyUnique strings in the
IdentifierTable.

Unfortunately that was happening. Some tables used to implement property
access in the JIT hold StringImpl*s, and turn these back into Identifiers
using the identfiier constructor. Since the code generator will now plant
by-id (cachable) accesses to PrivateNames we can end up passing an
EmptyUnique to Identifier::add, potentially leading to PrivateNames being
uniqued together (though hard to prove, since the hash codes are random).

* runtime/PropertyName.h:
(JSC::PropertyName::PropertyName):
(JSC::PropertyName::uid):
(JSC::PropertyName::publicName):
(JSC::PropertyName::asIndex):
    - PropertyName assumed that PrivateNames are not Identifiers - instead check isEmptyUnique().
* runtime/Structure.cpp:
(JSC::Structure::getPropertyNamesFromStructure):
    - Structure assumed that PrivateNames are not Identifiers - instead check isEmptyUnique().

Source/WTF:

* wtf/text/AtomicString.h:
(WTF::AtomicString::add):
    - Previously we assumed the only StringImpl that was validly allowed to claim to be
      Atomic but not be in a table was the canonical empty string. Now that EmptyUniques
      are also marked Atomic, all empty strings may pass this condition.
* wtf/text/StringImpl.cpp:
(WTF::StringImpl::~StringImpl):
    - EmptyUnique strings are not in the Atomic/Identfiier tabels, so don't need removing.
* wtf/text/StringImpl.h:
(WTF::StringImpl::StringImpl):
    - Change EmptyUnique constructor to call hashAndFlagsForEmptyUnique.
* wtf/text/StringStatics.cpp:
(WTF::StringImpl::hashAndFlagsForEmptyUnique):
    - Allocate a sequential hash code (this should be just as good for distribution & better
      for debugging than the random value) and set flags, now including Atomic & Identifier.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165946 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[iOS] Get code to compile on older iOS versions
psolanki@apple.com [Thu, 20 Mar 2014 05:10:51 +0000 (05:10 +0000)]
[iOS] Get code to compile on older iOS versions
https://bugs.webkit.org/show_bug.cgi?id=130142
<rdar://problem/16302908>

Reviewed by Darin Adler.

Source/WebCore:

* WebCore.exp.in:
* platform/ios/WebVideoFullscreenControllerAVKit.mm:
(-[WebVideoFullscreenController WebCore::]):
(-[WebVideoFullscreenController enterFullscreen:]):
(-[WebVideoFullscreenController exitFullscreen]):
* platform/ios/WebVideoFullscreenInterfaceAVKit.h:
* platform/ios/WebVideoFullscreenInterfaceAVKit.mm:
* platform/mac/HTMLConverter.mm:
(_dateForString):
* platform/network/cf/CookieJarCFNet.cpp:
(WebCore::copyCookiesForURLWithFirstPartyURL):
* platform/text/ios/LocalizedDateCache.mm:
(WebCore::LocalizedDateCache::calculateMaximumWidth):
* platform/text/mac/LocaleMac.mm:
(WebCore::LocaleMac::LocaleMac):

Source/WebKit/mac:

* History/WebHistory.mm:
(getDayBoundaries):
* Plugins/WebPluginController.h:
* Plugins/WebPluginController.mm:
(+[WebPluginController plugInViewWithArguments:fromPluginPackage:]):
* WebCoreSupport/WebFrameLoaderClient.mm:
(pluginView):
* WebView/WebPreferences.mm:
(-[WebPreferences _setAllowCompositingLayerVisualDegradation:]):
* WebView/WebPreferencesPrivate.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165944 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoFix "WEBKIT_WEBGL_compressed_texture_pvrtc" extension support
commit-queue@webkit.org [Thu, 20 Mar 2014 03:12:51 +0000 (03:12 +0000)]
Fix "WEBKIT_WEBGL_compressed_texture_pvrtc" extension support
https://bugs.webkit.org/show_bug.cgi?id=130292

Patch by Byungseon Shin <sun.shin@lge.com> on 2014-03-19
Reviewed by Dean Jackson.

WebGLRenderingContext::getExtension was not returning WEBGL_compressed_texture_pvrtc extension.

* html/canvas/WebGLRenderingContext.cpp:
(WebCore::WebGLRenderingContext::getExtension):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165943 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoMerge Misc. WebVTT Updates from Blink.
bfulgham@apple.com [Thu, 20 Mar 2014 02:53:59 +0000 (02:53 +0000)]
Merge Misc. WebVTT Updates from Blink.
https://bugs.webkit.org/show_bug.cgi?id=130491

Reviewed by Eric Carlson.

Source/WebCore:

Test: media/track/track-webvtt-tc030-interspersed-non-cue.html

Merged from Blink (patch by vcarbune@chromium.org):
https://chromium.googlesource.com/chromium/blink/+/fb62b2b54944bd19026192c049a1db250756fc12
http://crbug.com/40323003
https://chromium.googlesource.com/chromium/blink/+/29176fb0b394a97e85f4dabc07e18329f3ceae41
http://crbug.com/54463002
https://chromium.googlesource.com/chromium/blink/+/29476f839a08cb17c5a3fc713f7e454e7cea9c60
http://crbug.com/55853002
https://chromium.googlesource.com/chromium/blink/+/ba2378ef873e786cf2937034e012669cdd37dfe8
http://crbug.com/65343003
https://chromium.googlesource.com/chromium/blink/+/1915b9aa1bba45aebf28531b40ae627ecf9a3d59
http://crbug.com/62833005
https://chromium.googlesource.com/chromium/blink/+/ea46dbd0f1f29779a1cd4f59b19361b803f09a9b
http://crbug.com/64273013
https://chromium.googlesource.com/chromium/blink/+/9eb4bb6b627357bb94d66386a4123ec004cccfc4
http://crbug.com/101513002
https://chromium.googlesource.com/chromium/blink/+/192a843b32eafb77bf92c50ec50d334ba935028a
http://crbug.com/102403002
https://chromium.googlesource.com/chromium/blink/+/15f9efcc952fcf73184e28fb9db79633c1fa6653
http://crbug.com/104443002
https://chromium.googlesource.com/chromium/blink/+/74fd66b451a464683f7cb748855b1740a54b2c16
http://crbug.com/75243004

* html/track/TextTrackRegion.cpp:
(WebCore::TextTrackRegion::parseSettingValue):
* html/track/VTTCue.cpp:
(WebCore::VTTCue::markFutureAndPastNodes):
(WebCore::VTTCue::setCueSettings):
* html/track/WebVTTParser.cpp:
(WebCore::scanDigits):
(WebCore::WebVTTParser::collectDigitsToInt):
(WebCore::WebVTTParser::parseFloatPercentageValue):
(WebCore::WebVTTParser::parseBytes):
(WebCore::WebVTTParser::collectMetadataHeader):
(WebCore::WebVTTParser::collectTimingsAndSettings):
(WebCore::WebVTTParser::createDocumentFragmentFromCueText):
(WebCore::WebVTTParser::createNewCue):
(WebCore::WebVTTParser::createNewRegion):
(WebCore::WebVTTParser::collectTimeStamp):
(WebCore::tokenToNodeType):
(WebCore::WebVTTParser::constructTreeFromToken):
* html/track/WebVTTParser.h:
* html/track/WebVTTToken.h:
(WebCore::WebVTTToken::name):
(WebCore::WebVTTToken::characters):
(WebCore::WebVTTToken::appendToCharacter):
(WebCore::WebVTTToken::addNewClass):
(WebCore::WebVTTToken::classes):
(WebCore::WebVTTToken::addNewAnnotation):
(WebCore::WebVTTToken::annotation):
* html/track/WebVTTTokenizer.cpp:
(WebCore::equalLiteral):
(WebCore::WebVTTTokenizer::WebVTTTokenizer):
(WebCore::WebVTTTokenizer::nextToken):
* html/track/WebVTTTokenizer.h:

LayoutTests:

Merged from Blink (patch by vcarbune@chromium.org):
https://chromium.googlesource.com/chromium/blink/+/fb62b2b54944bd19026192c049a1db250756fc12
http://crbug.com/40323003
https://chromium.googlesource.com/chromium/blink/+/29176fb0b394a97e85f4dabc07e18329f3ceae41
http://crbug.com/54463002
https://chromium.googlesource.com/chromium/blink/+/29476f839a08cb17c5a3fc713f7e454e7cea9c60
http://crbug.com/55853002
https://chromium.googlesource.com/chromium/blink/+/ba2378ef873e786cf2937034e012669cdd37dfe8
http://crbug.com/65343003
https://chromium.googlesource.com/chromium/blink/+/1915b9aa1bba45aebf28531b40ae627ecf9a3d59
http://crbug.com/62833005
https://chromium.googlesource.com/chromium/blink/+/ea46dbd0f1f29779a1cd4f59b19361b803f09a9b
http://crbug.com/64273013
https://chromium.googlesource.com/chromium/blink/+/9eb4bb6b627357bb94d66386a4123ec004cccfc4
http://crbug.com/101513002
https://chromium.googlesource.com/chromium/blink/+/192a843b32eafb77bf92c50ec50d334ba935028a
http://crbug.com/102403002
https://chromium.googlesource.com/chromium/blink/+/15f9efcc952fcf73184e28fb9db79633c1fa6653
http://crbug.com/104443002
https://chromium.googlesource.com/chromium/blink/+/74fd66b451a464683f7cb748855b1740a54b2c16
http://crbug.com/75243004

* media/track/captions-webvtt/tc030-interspersed-non-cue.vtt: Added.
* media/track/track-webvtt-tc027-empty-cue-expected.txt: Mark one test as expected to fail, since
the test (incorrectly) expects to fail on empty cues. The test will be updated in a future patch.
* media/track/track-webvtt-tc030-interspersed-non-cue-expected.txt: Added.
* media/track/track-webvtt-tc030-interspersed-non-cue.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165942 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUnreviewed, revert the DFGCommon.h change in r165938. It was not intentional.
fpizlo@apple.com [Thu, 20 Mar 2014 02:00:40 +0000 (02:00 +0000)]
Unreviewed, revert the DFGCommon.h change in r165938. It was not intentional.

* dfg/DFGCommon.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165941 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoGC timer should intelligently choose between EdenCollections and FullCollections
mhahnenberg@apple.com [Thu, 20 Mar 2014 01:56:19 +0000 (01:56 +0000)]
GC timer should intelligently choose between EdenCollections and FullCollections
https://bugs.webkit.org/show_bug.cgi?id=128261

Reviewed by Geoffrey Garen.

Source/JavaScriptCore:

Most of the GCs while browsing the web are due to the GC timer. Currently the GC timer
always does FullCollections. To reduce the impact of the GC timer on the system this patch
changes Heap so that it has two timers, one for each type of collection. The FullCollection
timer is notified at the end of EdenCollections how much the Heap has grown since the last
FullCollection and when somebody notifies the Heap of abandoned memory (which usually wouldn't
be detected by an EdenCollection).

* CMakeLists.txt:
* GNUmakefile.list.am:
* JavaScriptCore.vcxproj/JavaScriptCore.vcxproj:
* JavaScriptCore.vcxproj/JavaScriptCore.vcxproj.filters:
* JavaScriptCore.xcodeproj/project.pbxproj:
* heap/EdenGCActivityCallback.cpp: Added.
(JSC::EdenGCActivityCallback::EdenGCActivityCallback):
(JSC::EdenGCActivityCallback::doCollection):
(JSC::EdenGCActivityCallback::lastGCLength):
(JSC::EdenGCActivityCallback::deathRate):
(JSC::EdenGCActivityCallback::gcTimeSlice):
* heap/EdenGCActivityCallback.h: Added.
(JSC::GCActivityCallback::createEdenTimer):
* heap/FullGCActivityCallback.cpp: Added.
(JSC::FullGCActivityCallback::FullGCActivityCallback):
(JSC::FullGCActivityCallback::doCollection):
(JSC::FullGCActivityCallback::lastGCLength):
(JSC::FullGCActivityCallback::deathRate):
(JSC::FullGCActivityCallback::gcTimeSlice):
* heap/FullGCActivityCallback.h: Added.
(JSC::GCActivityCallback::createFullTimer):
* heap/GCActivityCallback.cpp:
(JSC::GCActivityCallback::GCActivityCallback):
(JSC::GCActivityCallback::doWork):
(JSC::GCActivityCallback::scheduleTimer):
(JSC::GCActivityCallback::cancelTimer):
(JSC::GCActivityCallback::didAllocate):
(JSC::GCActivityCallback::willCollect):
(JSC::GCActivityCallback::cancel):
* heap/GCActivityCallback.h:
* heap/Heap.cpp:
(JSC::Heap::Heap):
(JSC::Heap::reportAbandonedObjectGraph):
(JSC::Heap::didAbandon):
(JSC::Heap::collectAllGarbage):
(JSC::Heap::collect):
(JSC::Heap::willStartCollection):
(JSC::Heap::updateAllocationLimits):
(JSC::Heap::didFinishCollection):
(JSC::Heap::setFullActivityCallback):
(JSC::Heap::setEdenActivityCallback):
(JSC::Heap::fullActivityCallback):
(JSC::Heap::edenActivityCallback):
(JSC::Heap::setGarbageCollectionTimerEnabled):
(JSC::Heap::didAllocate):
(JSC::Heap::shouldDoFullCollection):
* heap/Heap.h:
(JSC::Heap::lastFullGCLength):
(JSC::Heap::lastEdenGCLength):
(JSC::Heap::increaseLastFullGCLength):
(JSC::Heap::sizeBeforeLastEdenCollection):
(JSC::Heap::sizeAfterLastEdenCollection):
(JSC::Heap::sizeBeforeLastFullCollection):
(JSC::Heap::sizeAfterLastFullCollection):
* heap/HeapOperation.h:
* heap/HeapStatistics.cpp:
(JSC::HeapStatistics::showObjectStatistics):
* heap/HeapTimer.cpp:
(JSC::HeapTimer::timerDidFire):
* jsc.cpp:
(functionFullGC):
(functionEdenGC):
* runtime/Options.h:

Source/WebCore:

No new tests.

Updated WebSafeGCActivityCallbacks for both Eden and Full timers.

* bindings/js/JSDOMWindowBase.cpp:
(WebCore::JSDOMWindowBase::commonVM):
* platform/ios/WebSafeGCActivityCallbackIOS.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165940 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoGenerated hidpi- images (actual) on ML WK2 Debug are blurry/color shifted.
zalan@apple.com [Thu, 20 Mar 2014 01:54:21 +0000 (01:54 +0000)]
Generated hidpi- images (actual) on ML WK2 Debug are blurry/color shifted.

Unreviewed TestExpectations update on ML WK2 debug.

* platform/mac-wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165939 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoFix JavaScriptCore ChangeLog
fpizlo@apple.com [Thu, 20 Mar 2014 01:49:03 +0000 (01:49 +0000)]
Fix JavaScriptCore ChangeLog

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165938 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agomedia/video-remote-control-playpause.html frequently fails on the bots
eric.carlson@apple.com [Thu, 20 Mar 2014 01:05:37 +0000 (01:05 +0000)]
media/video-remote-control-playpause.html frequently fails on the bots
https://bugs.webkit.org/show_bug.cgi?id=130490

Unreviewed TestExpectations update.

* platform/mac/TestExpectations: Mark test as flakey while I try to reproduce it.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165937 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoAddress late review comment and style nit after r165934.
timothy_horton@apple.com [Thu, 20 Mar 2014 00:50:26 +0000 (00:50 +0000)]
Address late review comment and style nit after r165934.

* UIProcess/ios/ViewGestureControllerIOS.mm:
(WebKit::ViewGestureController::beginSwipeGesture):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165936 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoFix the iOS build after r165917.
aestes@apple.com [Thu, 20 Mar 2014 00:37:46 +0000 (00:37 +0000)]
Fix the iOS build after r165917.

* platform/ios/WebCoreSystemInterfaceIOS.mm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165935 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[iOS][WebKit2] Back-forward swipe doesn't always complete with fingers down
timothy_horton@apple.com [Thu, 20 Mar 2014 00:32:32 +0000 (00:32 +0000)]
[iOS][WebKit2] Back-forward swipe doesn't always complete with fingers down
https://bugs.webkit.org/show_bug.cgi?id=130489
<rdar://problem/16027007>

Reviewed by Simon Fraser.

* UIProcess/ios/ViewGestureControllerIOS.mm:
(WebKit::ViewGestureController::beginSwipeGesture):
In order to have fingers-down swipe completion work correctly, we need
to give UIKit an empty interactive update block.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165934 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUnreviewed, rolling out r165926.
commit-queue@webkit.org [Thu, 20 Mar 2014 00:18:26 +0000 (00:18 +0000)]
Unreviewed, rolling out r165926.
https://bugs.webkit.org/show_bug.cgi?id=130488

broke the iOS build (Requested by estes on #webkit).

Reverted changeset:

Source/JavaScriptCore:

"GC timer should intelligently choose between EdenCollections
and FullCollections"
https://bugs.webkit.org/show_bug.cgi?id=128261
http://trac.webkit.org/changeset/165926

Source/WebCore:

"GC timer should intelligently choose between EdenCollections
and FullCollections"
https://bugs.webkit.org/show_bug.cgi?id=128261
http://trac.webkit.org/changeset/165926

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165933 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUnreviewed, rolling out r165929.
commit-queue@webkit.org [Thu, 20 Mar 2014 00:13:25 +0000 (00:13 +0000)]
Unreviewed, rolling out r165929.
https://bugs.webkit.org/show_bug.cgi?id=130487

in order to roll out r165926 (Requested by estes on #webkit).

Reverted changeset:

http://trac.webkit.org/changeset/165929

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165932 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoGet rid of two IPC::Connection functions by folding them into lambdas
andersca@apple.com [Wed, 19 Mar 2014 23:45:46 +0000 (23:45 +0000)]
Get rid of two IPC::Connection functions by folding them into lambdas
https://bugs.webkit.org/show_bug.cgi?id=130482

Reviewed by Tim Horton.

* Platform/IPC/Connection.cpp:
(IPC::Connection::addWorkQueueMessageReceiver):
(IPC::Connection::removeWorkQueueMessageReceiver):
* Platform/IPC/Connection.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165931 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoExpose text and page zoom in WK2 SPI, and hook them up in MiniBrowser
simon.fraser@apple.com [Wed, 19 Mar 2014 23:44:39 +0000 (23:44 +0000)]
Expose text and page zoom in WK2 SPI, and hook them up in MiniBrowser
https://bugs.webkit.org/show_bug.cgi?id=130476

Source/WebKit2:

Reviewed by Anders Carlsson.

Expose page and text zoom as SPI.

* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _supportsTextZoom]):
(-[WKWebView _textZoomFactor]):
(-[WKWebView _setTextZoomFactor:]):
(-[WKWebView _pageZoomFactor]):
(-[WKWebView _setPageZoomFactor:]):
* UIProcess/API/Cocoa/WKWebViewPrivate.h:

Tools:

Reviewed by Anders Carlsson.

Hook up page/text zoom in MiniBrowser, make reload work, remove some logging.

* MiniBrowser/mac/WK1BrowserWindowController.m:
(-[WK1BrowserWindowController webView:decidePolicyForNavigationAction:request:frame:decisionListener:]):
Remove log.
* MiniBrowser/mac/WK2BrowserWindowController.m:
(-[WK2BrowserWindowController awakeFromNib]):
(-[WK2BrowserWindowController validateMenuItem:]):
(-[WK2BrowserWindowController reload:]):
Hook up reload.
(-[WK2BrowserWindowController toggleZoomMode:]):
(-[WK2BrowserWindowController resetZoom:]):
(-[WK2BrowserWindowController canResetZoom]):
(-[WK2BrowserWindowController currentZoomFactor]):
(-[WK2BrowserWindowController setCurrentZoomFactor:]):
(-[WK2BrowserWindowController canZoomIn]):
(-[WK2BrowserWindowController canZoomOut]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165930 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoBuild fix for iOS after r165926
mhahnenberg@apple.com [Wed, 19 Mar 2014 23:27:50 +0000 (23:27 +0000)]
Build fix for iOS after r165926

* bindings/js/JSDOMWindowBase.cpp:
(WebCore::JSDOMWindowBase::commonVM):
* platform/ios/WebSafeGCActivityCallbackIOS.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165929 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoAllow media engine to update HTMLMediaElement state when playing to external device
eric.carlson@apple.com [Wed, 19 Mar 2014 22:42:40 +0000 (22:42 +0000)]
Allow media engine to update HTMLMediaElement state when playing to external device
https://bugs.webkit.org/show_bug.cgi?id=130480

Reviewed by Jer Noble.

* html/HTMLMediaElement.h: Add mediaPlayerHandlePlaybackCommand.

* platform/graphics/MediaPlayer.cpp:
(WebCore::MediaPlayer::handlePlaybackCommand): New.
* platform/graphics/MediaPlayer.h:
(WebCore::MediaPlayerClient::mediaPlayerHandlePlaybackCommand):

* platform/graphics/avfoundation/MediaPlayerPrivateAVFoundation.cpp:
(WebCore::MediaPlayerPrivateAVFoundation::rateChanged): Send Play and Pause commands when
    playing to an external device.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165928 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoImplement isPointInPath and isPointInStroke with Path2D argument
krit@webkit.org [Wed, 19 Mar 2014 22:39:41 +0000 (22:39 +0000)]
Implement isPointInPath and isPointInStroke with Path2D argument
https://bugs.webkit.org/show_bug.cgi?id=130473

Reviewed by Dean Jackson.

Source/WebCore:

Implement isPointInPath and isPointInStroke for 2D Canvas contxts
with Path2D argument.

Tests: fast/canvas/canvas-path-isPointInPath.html
       fast/canvas/canvas-path-isPointInStroke.html

* html/canvas/CanvasRenderingContext2D.cpp:
(WebCore::CanvasRenderingContext2D::isPointInPath):
(WebCore::CanvasRenderingContext2D::isPointInStroke):
(WebCore::CanvasRenderingContext2D::isPointInPathInternal):
(WebCore::CanvasRenderingContext2D::isPointInStrokeInternal):
* html/canvas/CanvasRenderingContext2D.h:
* html/canvas/CanvasRenderingContext2D.idl:

LayoutTests:

Test isPointInPath and isPointInStroke with Path2D argument.

* fast/canvas/canvas-path-isPointInPath-expected.txt: Added.
* fast/canvas/canvas-path-isPointInPath.html: Added.
* fast/canvas/canvas-path-isPointInStroke-expected.txt: Added.
* fast/canvas/canvas-path-isPointInStroke.html: Added.
* fast/canvas/script-tests/canvas-path-isPointInPath.js: Added.
* fast/canvas/script-tests/canvas-path-isPointInStroke.js: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165927 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoGC timer should intelligently choose between EdenCollections and FullCollections
mhahnenberg@apple.com [Wed, 19 Mar 2014 22:21:51 +0000 (22:21 +0000)]
GC timer should intelligently choose between EdenCollections and FullCollections
https://bugs.webkit.org/show_bug.cgi?id=128261

Reviewed by Geoffrey Garen.

Source/JavaScriptCore:

Most of the GCs while browsing the web are due to the GC timer. Currently the GC timer
always does FullCollections. To reduce the impact of the GC timer on the system this patch
changes Heap so that it has two timers, one for each type of collection. The FullCollection
timer is notified at the end of EdenCollections how much the Heap has grown since the last
FullCollection and when somebody notifies the Heap of abandoned memory (which wouldn't be
detected by an EdenCollection).

* heap/GCActivityCallback.cpp:
(JSC::GCActivityCallback::GCActivityCallback):
(JSC::GCActivityCallback::doWork):
(JSC::FullGCActivityCallback::FullGCActivityCallback):
(JSC::FullGCActivityCallback::doCollection):
(JSC::EdenGCActivityCallback::EdenGCActivityCallback):
(JSC::EdenGCActivityCallback::doCollection):
(JSC::GCActivityCallback::scheduleTimer):
(JSC::GCActivityCallback::cancelTimer):
(JSC::GCActivityCallback::didAllocate):
(JSC::GCActivityCallback::willCollect):
(JSC::GCActivityCallback::cancel):
* heap/GCActivityCallback.h:
(JSC::GCActivityCallback::GCActivityCallback):
(JSC::GCActivityCallback::createFullTimer):
(JSC::GCActivityCallback::createEdenTimer):
* heap/Heap.cpp:
(JSC::Heap::Heap):
(JSC::Heap::didAbandon):
(JSC::Heap::willStartCollection):
(JSC::Heap::updateAllocationLimits):
(JSC::Heap::setFullActivityCallback):
(JSC::Heap::setEdenActivityCallback):
(JSC::Heap::fullActivityCallback):
(JSC::Heap::edenActivityCallback):
(JSC::Heap::setGarbageCollectionTimerEnabled):
(JSC::Heap::didAllocate):
* heap/Heap.h:
* heap/HeapTimer.cpp:
(JSC::HeapTimer::timerDidFire):

Source/WebCore:

No new tests.

* platform/ios/WebSafeGCActivityCallbackIOS.h: Update WebSafeGCActivityCallback
appropriately for the introduction of another type of GCActivityCallback.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165926 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[iOS][WK2] Reduce the extended background's opacity when WKContentView is smaller...
benjamin@webkit.org [Wed, 19 Mar 2014 21:57:01 +0000 (21:57 +0000)]
[iOS][WK2] Reduce the extended background's opacity when WKContentView is smaller than its WKWebView
https://bugs.webkit.org/show_bug.cgi?id=130436

Patch by Benjamin Poulain <bpoulain@apple.com> on 2014-03-19
Reviewed by Beth Dakin.

* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _updateScrollViewBackground]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165925 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[iOS][WebKit2] <rdar://problem/16330387> Do not use the keyboard rect for computing...
benjamin@webkit.org [Wed, 19 Mar 2014 21:55:11 +0000 (21:55 +0000)]
[iOS][WebKit2] <rdar://problem/16330387> Do not use the keyboard rect for computing the general unobscuredRect, only use it for centering
https://bugs.webkit.org/show_bug.cgi?id=130382

Patch by Benjamin Poulain <bpoulain@apple.com> on 2014-03-19
Reviewed by Simon Fraser.

Always accumulating the keyboard rect out of the exposedRect caused some issues (see <rdar://problem/16330387>).
This patch split the rect in two: unobscured rect for WebCore, and unobscuredContentRectIncludingKeyboard for centering
UI Elements in view.

* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _zoomToRect:WebCore::atScale:origin:WebCore::]):
(-[WKWebView _scrollToRect:WebCore::origin:WebCore::minimumScrollDistance:]):
(-[WKWebView _zoomToRect:WebCore::withOrigin:WebCore::fitEntireRect:minimumScale:maximumScale:minimumScrollDistance:]):
(-[WKWebView _contentRectForUserInteraction]):
(-[WKWebView _updateVisibleContentRects]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165924 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoAttempt to fix the Production build.
timothy@apple.com [Wed, 19 Mar 2014 21:50:16 +0000 (21:50 +0000)]
Attempt to fix the Production build.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165923 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoGet rid of WKProcessClass
andersca@apple.com [Wed, 19 Mar 2014 21:48:49 +0000 (21:48 +0000)]
Get rid of WKProcessClass
https://bugs.webkit.org/show_bug.cgi?id=130478

Reviewed by Dan Bernstein.

* UIProcess/API/Cocoa/WKProcessClass.h: Removed.
* UIProcess/API/Cocoa/WKProcessClass.mm: Removed.
* UIProcess/API/Cocoa/WKProcessClassConfiguration.h: Removed.
* UIProcess/API/Cocoa/WKProcessClassConfiguration.mm: Removed.
* UIProcess/API/Cocoa/WKProcessClassConfigurationPrivate.h: Removed.
* UIProcess/API/Cocoa/WKProcessClassPrivate.h: Removed.
* UIProcess/API/Cocoa/WKWebViewConfigurationToBeRemoved.mm: Removed.
* WebKit2.xcodeproj/project.pbxproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165922 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoCrash with long selector list
antti@apple.com [Wed, 19 Mar 2014 21:43:52 +0000 (21:43 +0000)]
Crash with long selector list
https://bugs.webkit.org/show_bug.cgi?id=123006

Reviewed by Andreas Kling.

Source/WebCore:

Test: fast/css/long-selector-list-crash.html

* css/CSSSelectorList.cpp:
(WebCore::CSSSelectorList::CSSSelectorList):
(WebCore::CSSSelectorList::adoptSelectorVector):
(WebCore::CSSSelectorList::operator=):
* css/StyleRule.cpp:
(WebCore::StyleRule::create):

    Add a bunch of asserts.

(WebCore::StyleRule::splitIntoMultipleRulesWithMaximumSelectorComponentCount):

    This could produce a zero-length selector list.

LayoutTests:

* fast/css/long-selector-list-crash-expected.txt: Added.
* fast/css/long-selector-list-crash.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165921 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoRebaseline bindings tests.
akling@apple.com [Wed, 19 Mar 2014 21:42:54 +0000 (21:42 +0000)]
Rebaseline bindings tests.

* bindings/scripts/test/JS/*:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165920 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[iOS][MediaControls] Reverse-pinching using the built-in controls should enter full...
jer.noble@apple.com [Wed, 19 Mar 2014 21:38:51 +0000 (21:38 +0000)]
[iOS][MediaControls] Reverse-pinching using the built-in controls should enter full screen
https://bugs.webkit.org/show_bug.cgi?id=129707

Reviewed by Eric Carlson.

The default controls on iOS should respond to a pinch gesture by entering fullscreen. Add a gesture
event listener, and enter full screen if the user does a quick (>2 velocity) reverse (>1.25 scale) pinch.

* Modules/mediacontrols/mediaControlsiOS.js:
(ControllerIOS.prototype.createBase):
(ControllerIOS.prototype.handlePlayButtonTouchEnd):
(ControllerIOS.prototype.handleBaseGestureStart):
(ControllerIOS.prototype.handleBaseGestureChange):
(ControllerIOS.prototype.handleBaseGestureEnd):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165919 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoREGRESSION(r156316): Scrollbar on a large document is broken
rniwa@webkit.org [Wed, 19 Mar 2014 21:36:47 +0000 (21:36 +0000)]
REGRESSION(r156316): Scrollbar on a large document is broken
https://bugs.webkit.org/show_bug.cgi?id=130474
<rdar://problem/16371523>

Reviewed by Anders Carlsson.

The bug was caused by updateArrowPlacement() doing the work even though Scrollbar Overlay API is available.
Prior to r156316, this function exited early:
http://trac.webkit.org/browser/trunk/Source/WebCore/platform/mac/ScrollbarThemeMac.mm?rev=156315#L147

Fixed the bug by deleting updateArrowPlacement() entirely.

* platform/mac/ScrollbarThemeMac.mm:
(WebCore::ScrollbarThemeMac::preferencesChanged):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165918 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[Mac][EME] Query whether decoder support exists before creating a new MediaKeys object.
jer.noble@apple.com [Wed, 19 Mar 2014 21:27:10 +0000 (21:27 +0000)]
[Mac][EME] Query whether decoder support exists before creating a new MediaKeys object.
https://bugs.webkit.org/show_bug.cgi?id=130459

Reviewed by Eric Carlson.

Source/WebCore:

* WebCore.exp.in: Export the new WCSI method.
* platform/graphics/avfoundation/objc/MediaPlayerPrivateMediaSourceAVFObjC.mm:
(WebCore::MediaPlayerPrivateMediaSourceAVFObjC::supportsKeySystem): Bail early if we have no decoder support.
* platform/mac/WebCoreSystemInterface.h: Provide a WCSI wrapper for WKSI.
* platform/mac/WebCoreSystemInterface.mm: Ditto.

Source/WebKit/mac:

* WebCoreSupport/WebSystemInterface.mm:
(InitWebCoreSystemInterface): Initialize the new WCSI method.

Source/WebKit2:

* WebProcess/WebCoreSupport/mac/WebSystemInterface.mm:
(InitWebCoreSystemInterface): Initialize the new WCSI method.

WebKitLibraries:

* WebKitSystemInterface.h:
* libWebKitSystemInterfaceMavericks.a:
* libWebKitSystemInterfaceMountainLion.a:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165917 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoMake Remote Control commands testable
eric.carlson@apple.com [Wed, 19 Mar 2014 21:08:24 +0000 (21:08 +0000)]
Make Remote Control commands testable
https://bugs.webkit.org/show_bug.cgi?id=130358

Reviewed by Jer Noble.

Source/WebCore:

Test: media/video-remote-control-playpause.html

* html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::mediaPlayerTimeChanged): Call session->clientWillPausePlayback.
(WebCore::HTMLMediaElement::didReceiveRemoteControlCommand): Log the command.

* platform/audio/MediaSession.cpp:
(WebCore::MediaSessionClient::mediaSessionTitle): Move from header file.
(WebCore::MediaSessionClient::mediaSessionDuration): Ditto.
(WebCore::MediaSessionClient::mediaSessionCurrentTime): Ditto.
* platform/audio/MediaSession.h:

* platform/audio/MediaSessionManager.cpp:
(WebCore::MediaSessionManager::MediaSessionManager): Get rid of m_activeSession.
(WebCore::MediaSessionManager::removeSession): No need to clear m_activeSession because
    we use the m_sessions vector.
(WebCore::MediaSessionManager::sessionWillBeginPlayback): setCurrentSession takes a
    MediaSession&.
(WebCore::MediaSessionManager::sessionWillEndPlayback): Keep m_sessions sorted so playing
    sessions are before paused sessions.
(WebCore::MediaSessionManager::setCurrentSession): Move the session that is beginning to
    play to the head of m_sessions.
(WebCore::MediaSessionManager::currentSession): Return m_sessions[0] or nullptr.
(WebCore::MediaSessionManager::didReceiveRemoteControlCommand):
* platform/audio/MediaSessionManager.h:

* platform/audio/ios/MediaSessionManagerIOS.mm: Include MediaPlayer.h.

* testing/Internals.cpp:
(WebCore::Internals::postRemoteControlCommand): New.
* testing/Internals.h:
* testing/Internals.idl:

LayoutTests:

* media/video-remote-control-playpause-expected.txt: Added.
* media/video-remote-control-playpause.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165916 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoImproving webkitGetUserMedia error handling and error messages
thiago.lacerda@openbossa.org [Wed, 19 Mar 2014 21:06:38 +0000 (21:06 +0000)]
Improving webkitGetUserMedia error handling and error messages
https://bugs.webkit.org/show_bug.cgi?id=130452

Reviewed by Eric Carlson.

Checking if first argument of webkitGetUserMedia is a valid Dictionary. If not, throw an exception with a
clearer message of the error.

Source/WebCore:

Test updated.

* Modules/mediastream/NavigatorMediaStream.idl:
* UseJSC.cmake:
* bindings/js/JSNavigatorCustom.cpp: Added.

LayoutTests:

* fast/mediastream/argument-types-expected.txt:
* fast/mediastream/script-tests/argument-types.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165915 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoJSC bindings should use the passed-in global object for wrapper caching.
akling@apple.com [Wed, 19 Mar 2014 20:42:55 +0000 (20:42 +0000)]
JSC bindings should use the passed-in global object for wrapper caching.
<https://webkit.org/b/130458>

Take a shorter path to the DOMWrapperWorld by using the passed-in
global object instead of getting the lexical global from ExecState.
Removed an old FIXME that no longer applies since JSObjects can
always find the right global object these days.

Reviewed by Geoff Garen.

* bindings/js/JSCSSRuleCustom.cpp:
(WebCore::toJS):
* bindings/js/JSCSSStyleDeclarationCustom.cpp:
(WebCore::JSCSSStyleDeclaration::getPropertyCSSValue):
* bindings/js/JSCSSValueCustom.cpp:
(WebCore::toJS):
* bindings/js/JSDOMBinding.h:
(WebCore::createWrapper):
(WebCore::wrap):
(WebCore::getExistingWrapper):
(WebCore::createNewWrapper):
(WebCore::toJS):
* bindings/js/JSDOMWindowCustom.cpp:
(WebCore::JSDOMWindow::setTimeout):
(WebCore::JSDOMWindow::setInterval):
(WebCore::JSDOMWindow::addEventListener):
(WebCore::JSDOMWindow::removeEventListener):
* bindings/js/JSDocumentCustom.cpp:
(WebCore::JSDocument::location):
(WebCore::toJS):
* bindings/js/JSElementCustom.cpp:
(WebCore::toJSNewlyCreated):
* bindings/js/JSEventCustom.cpp:
(WebCore::toJS):
* bindings/js/JSHTMLCollectionCustom.cpp:
(WebCore::toJS):
* bindings/js/JSHTMLTemplateElementCustom.cpp:
(WebCore::JSHTMLTemplateElement::content):
* bindings/js/JSImageDataCustom.cpp:
(WebCore::toJS):
* bindings/js/JSNodeCustom.cpp:
(WebCore::createWrapperInline):
* bindings/js/JSNodeCustom.h:
(WebCore::toJS):
* bindings/js/JSSVGPathSegCustom.cpp:
(WebCore::toJS):
* bindings/js/JSStyleSheetCustom.cpp:
(WebCore::toJS):
* bindings/js/JSTextTrackCueCustom.cpp:
(WebCore::toJS):
* bindings/js/JSTrackCustom.cpp:
(WebCore::toJS):
* bindings/js/JSWorkerGlobalScopeCustom.cpp:
(WebCore::JSWorkerGlobalScope::setTimeout):
(WebCore::JSWorkerGlobalScope::setInterval):
* bindings/scripts/CodeGeneratorJS.pm:
(GenerateImplementation):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165914 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoPage::setPageScaleFactor should not force the scroll view position when using delegat...
benjamin@webkit.org [Wed, 19 Mar 2014 20:41:46 +0000 (20:41 +0000)]
Page::setPageScaleFactor should not force the scroll view position when using delegatesScrolling()
https://bugs.webkit.org/show_bug.cgi?id=130429

Patch by Benjamin Poulain <bpoulain@apple.com> on 2014-03-19
Reviewed by Simon Fraser.

When setting the same scale mutliple time (for example on load), Page was asking the top scrollview
to scroll to the origin. This is bogus since the scroll view is driving Page::setDeviceScaleFactor,
not the other way around.

Simon fixed the general case in http://trac.webkit.org/changeset/165652/trunk/Source/WebCore/page/Page.cpp
but this branch was still scrolling content.

* page/Page.cpp:
(WebCore::Page::setPageScaleFactor):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165913 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoREGRESSION(r165459): It broke 109 jsc stress test on ARM Thumb2 and Mac 32 bit
fpizlo@apple.com [Wed, 19 Mar 2014 20:36:45 +0000 (20:36 +0000)]
REGRESSION(r165459): It broke 109 jsc stress test on ARM Thumb2 and Mac 32 bit
https://bugs.webkit.org/show_bug.cgi?id=130134

Reviewed by Mark Hahnenberg.

* dfg/DFGFixupPhase.cpp:
(JSC::DFG::FixupPhase::fixupNode): Can't do some optimizations if you don't have a lot of registers.
* dfg/DFGSpeculativeJIT32_64.cpp:
(JSC::DFG::SpeculativeJIT::cachedGetById): Move stuff around before going into the IC code to ensure that we give the IC code the invariants it needs. This only happens in case of GetByIdFlush, where we are forced into using weird combinations of registers because the results have to be in t0/t1.
(JSC::DFG::SpeculativeJIT::compile): For a normal GetById, the register allocator should just do the right thing so nobody has to move anything around.
* jit/JITInlineCacheGenerator.cpp:
(JSC::JITGetByIdGenerator::JITGetByIdGenerator): Assert the things we want.
* jit/JITInlineCacheGenerator.h:
* jit/Repatch.cpp:
(JSC::generateGetByIdStub): Remove a previous incomplete hack to try to work around the DFG's problem.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165912 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoFilter out editable services from image menu for non-contenteditable images
beidson@apple.com [Wed, 19 Mar 2014 20:12:24 +0000 (20:12 +0000)]
Filter out editable services from image menu for non-contenteditable images
<rdar://problem/16319645> and https://bugs.webkit.org/show_bug.cgi?id=130437

Reviewed by Tim Horton.

Source/WebCore:

* WebCore.exp.in:

Source/WebKit/mac:

* Misc/WebSharingServicePickerController.h:
* Misc/WebSharingServicePickerController.mm:
(-[WebSharingServicePickerController sharingServicePicker:sharingServicesForItems:mask:proposedSharingServices:]):
  Filter out editing services based on the includeEditorServices flag.
* WebCoreSupport/WebContextMenuClient.mm:
(WebContextMenuClient::contextMenuForEvent):

Source/WebKit2:

* UIProcess/mac/WebContextMenuProxyMac.h:
* UIProcess/mac/WebContextMenuProxyMac.mm:
(-[WKSharingServicePickerDelegate setIncludeEditorServices:]):
(-[WKSharingServicePickerDelegate sharingServicePicker:sharingServicesForItems:mask:proposedSharingServices:]):
  Filter out editing services based on the includeEditorServices flag.
(WebKit::WebContextMenuProxyMac::setupImageServicesMenu):
(WebKit::WebContextMenuProxyMac::populate):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165911 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoImplement method addPath for Path2D
krit@webkit.org [Wed, 19 Mar 2014 19:54:42 +0000 (19:54 +0000)]
Implement method addPath for Path2D
https://bugs.webkit.org/show_bug.cgi?id=130461

Reviewed by Dean Jackson.

Source/WebCore:

Add support for the Path2D method addPath. The method allows
to append and existing Path2D object to another one. It can be
combined with a transformation matrix SVGMatrix.

Test: fast/canvas/canvas-path-addPath.html

* html/canvas/DOMPath.h: Add new method addPath().
(WebCore::DOMPath::addPath):
* html/canvas/DOMPath.idl:
* platform/graphics/Path.h: Add support for appending paths.
* platform/graphics/cairo/PathCairo.cpp:
(WebCore::Path::addPath):
* platform/graphics/cg/PathCG.cpp:
(WebCore::Path::addPath):
* platform/graphics/wince/PathWinCE.cpp:
(WebCore::Path::addPath):

LayoutTests:

Test the addPath Path2D method.

* platform/efl/TestExpectations:
* platform/gtk/TestExpectations:
* fast/canvas/canvas-path-addPath-expected.txt: Added.
* fast/canvas/canvas-path-addPath.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165910 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoNormalize some of the older JSC options
mhahnenberg@apple.com [Wed, 19 Mar 2014 19:32:43 +0000 (19:32 +0000)]
Normalize some of the older JSC options
https://bugs.webkit.org/show_bug.cgi?id=128753

Reviewed by Michael Saboff.

* runtime/Options.cpp:
(JSC::Options::initialize):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165909 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[iOS] WebKit2 Quicklook.
commit-queue@webkit.org [Wed, 19 Mar 2014 19:24:30 +0000 (19:24 +0000)]
[iOS] WebKit2 Quicklook.
https://bugs.webkit.org/show_bug.cgi?id=130360

Source/WebCore:

Patch by Yongjun Zhang <yongjun_zhang@apple.com> on 2014-03-19
Reviewed by Tim Horton.

Since ResourceLoaders using WebKit2 network process don't have ResourceHandle any more, we
need to make it possible to create QuickLookHandle from ResourceLoader.

* WebCore.exp.in: Make some QuickLookHandle methods visible from WebKit2.
* loader/ResourceLoader.h: Add m_quickLookHandle data member to ResourceLoader.
(WebCore::ResourceLoader::quickLookHandle):
(WebCore::ResourceLoader::setQuickLookHandle):
* platform/network/ios/QuickLook.h: Add a new method to create QuickLookHandle from ResourceLoader.
* platform/network/ios/QuickLook.mm:
(WebCore::QuickLookHandle::create):

Source/WebKit2:

For WebKit2 resource loads, if we detect a quicklook content that QLConverter is able to
convert, handle the response data to QLConverter and use its delegate WKWebResourceQuickLookDelegate
to rounte the converted results (HTML) back into WebCore.  Thie similiar logic has been in place
for WebKit1.

Patch by Yongjun Zhang <yongjun_zhang@apple.com> on 2014-03-19
Reviewed by Tim Horton.

* WebKit2.xcodeproj/project.pbxproj: Add WebResourceLoaderIOS.mm to the project.
* WebProcess/Network/WebResourceLoadScheduler.cpp:
* WebProcess/ios/WebResourceLoaderIOS.mm: Added.
(-[WKWebResourceQuickLookDelegate initWithWebResourceLoader:WebKit::]): WKWebResourceQuickLookDelegate holds
    the original WebResourceLoader so that we could send converted result to WebCore.
(-[WKWebResourceQuickLookDelegate connection:didReceiveDataArray:]):
(-[WKWebResourceQuickLookDelegate connection:didReceiveData:lengthReceived:]):
(-[WKWebResourceQuickLookDelegate connectionDidFinishLoading:]):
(-[WKWebResourceQuickLookDelegate connection:didFailWithError:]):
(-[WKWebResourceQuickLookDelegate clearHandle]):
(WebKit::WebResourceLoader::setUpQuickLookHandleIfNeeded):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165908 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[iOS] Disable the video plugin proxy by default on WK1.
jer.noble@apple.com [Wed, 19 Mar 2014 19:21:03 +0000 (19:21 +0000)]
[iOS] Disable the video plugin proxy by default on WK1.
https://bugs.webkit.org/show_bug.cgi?id=130460

Reviewed by Eric Carlson.

Update the defaults to universally enable the AVFoundationEnabled preference, disable the
VideoPluginProxyEnabled preference, and on iOS only enable the AVKitEnabled preference.

* WebView/WebPreferences.mm:
(+[WebPreferences initialize]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165907 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoSmall cleanup of empty string
barraclough@apple.com [Wed, 19 Mar 2014 19:00:51 +0000 (19:00 +0000)]
Small cleanup of empty string
https://bugs.webkit.org/show_bug.cgi?id=130438

Reviewed by Anders Carlson.

The empty string is weird becuase it is used as the impl for an AtomicString, but isAtomic()
returns false.
Also, we set a valid pointer for the characters pointer due to an old PCRE bug; remove this.
Also, we currently only create one 'static' string, the empty string. Adding more might break
assumtions, inhibit the constructor from being used to create other static strings for now.

* wtf/text/AtomicString.h:
(WTF::AtomicString::add):
    - update assert.
* wtf/text/StringImpl.h:
(WTF::StringImpl::StringImpl):
    - Merge ConstructStaticString constructors, only allow the empty static string to be constructed.
* wtf/text/StringStatics.cpp:
(WTF::StringImpl::empty):
    - ConstructStaticString -> ConstructEmptyString, remove arguments, DEPRECATED_DEFINE_STATIC_LOCAL -> NeverDestroyed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165906 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoFix cue rendering test and include support for left/right alignment
bfulgham@apple.com [Wed, 19 Mar 2014 19:00:14 +0000 (19:00 +0000)]
Fix cue rendering test and include support for left/right alignment
https://bugs.webkit.org/show_bug.cgi?id=130406

Reviewed by Eric Carlson.

Merged from Blink (patch by vcarbune@chromium.org):
https://chromium.googlesource.com/chromium/blink/+/e695703cb603421c741e59a557a7989113c13b94
http://crbug.com/25155003

Source/WebCore:

No new tests. Covered by media/track/track-cue-rendering-snap-to-lines-not-set.html.

* html/track/VTTCue.cpp:
(WebCore::leftKeyword): Added.
(WebCore::rightKeyword): Added.
(WebCore::VTTCueBox::applyCSSProperties): Added.
(WebCore::VTTCue::VTTCue):
(WebCore::VTTCue::align): Handle left/right alignment.
(WebCore::VTTCue::setAlign): Ditto.
(WebCore::VTTCue::calculateDisplayParameters): Refactored and revised for left/right tokens.
(WebCore::VTTCue::setCueSettings): Ditto.
(WebCore::VTTCue::getCSSAlignment): Added.
(WebCore::VTTCue::getCSSWritingMode): Added.
* html/track/VTTCue.h:

LayoutTests:

* media/track/captions-webvtt/captions-snap-to-lines-not-set.vtt: Revised for left/right alignment tags.
* media/track/track-cue-rendering-snap-to-lines-not-set-expected.txt: Ditto.
* media/track/track-cue-rendering-snap-to-lines-not-set.html: Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165905 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoWebKit2 View Gestures: Swipe gesture can track vertical movement instead of horizonta...
timothy_horton@apple.com [Wed, 19 Mar 2014 17:56:55 +0000 (17:56 +0000)]
WebKit2 View Gestures: Swipe gesture can track vertical movement instead of horizontal movement
https://bugs.webkit.org/show_bug.cgi?id=130441
<rdar://problem/16337279>

Reviewed by Simon Fraser.

* UIProcess/mac/ViewGestureControllerMac.mm:
(WebKit::scrollEventCanBecomeSwipe):
Factor logic that determines if a scroll event can become a swipe
out into scrollEventCanBecomeSwipe.
Use <= instead of < when comparing the axes, because we don't want
to start a swipe on a completely diagonal scroll either (because AppKit
can start tracking a vertical swipe in that case).

(WebKit::ViewGestureController::handleScrollWheelEvent):
(WebKit::ViewGestureController::wheelEventWasNotHandledByWebCore):
Make use of scrollEventCanBecomeSwipe.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165903 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUpdate type of local vars to match the type of String length.
mark.lam@apple.com [Wed, 19 Mar 2014 17:53:08 +0000 (17:53 +0000)]
Update type of local vars to match the type of String length.
<https://webkit.org/b/130077>

Reviewed by Geoffrey Garen.

* runtime/JSStringJoiner.cpp:
(JSC::JSStringJoiner::join):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165902 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoAdd new functionality that will be required by the history item decoding
andersca@apple.com [Wed, 19 Mar 2014 17:36:37 +0000 (17:36 +0000)]
Add new functionality that will be required by the history item decoding
https://bugs.webkit.org/show_bug.cgi?id=130433

Reviewed by Sam Weinig.

Source/WebCore:

* platform/KeyedCoding.h:
(WebCore::KeyedDecoder::decodeConditionalObject):

Source/WebKit2:

* Shared/cf/KeyedDecoder.cpp:
(WebKit::KeyedDecoder::decodeUInt32):
(WebKit::KeyedDecoder::decodeInt32):
(WebKit::KeyedDecoder::decodeFloat):
* Shared/cf/KeyedDecoder.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165900 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoREGRESSION (OS X 10.9.2): PageVisibilityStateWithWindowChanges.WebKit2 API test fails
ap@apple.com [Wed, 19 Mar 2014 17:27:24 +0000 (17:27 +0000)]
REGRESSION (OS X 10.9.2): PageVisibilityStateWithWindowChanges.WebKit2 API test fails
https://bugs.webkit.org/show_bug.cgi?id=130463

* TestWebKitAPI/Tests/mac/PageVisibilityStateWithWindowChanges.mm: Disable the test.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@165899 268f45cc-cd09-0410-ab3c-d52691b4dbfc