WebKit-https.git
5 months agoWebProcessCache should keep track of processes being added
cdumez@apple.com [Mon, 11 Mar 2019 19:30:42 +0000 (19:30 +0000)]
WebProcessCache should keep track of processes being added
https://bugs.webkit.org/show_bug.cgi?id=195538

Reviewed by Geoffrey Garen.

WebProcessCache should keep track of processes being added, while they are being
checked for responsiveness. This is useful so that:
- Requests to clear the cache also clear processes being added
- Requests to remove a given process from the cache (either because it crashed
  or because it is being used for a history navigation) actually remove the
  process if it is still being checked for responsiveness.
- The cached process eviction timer applies to such processes in case something
  goes wrong with the code and the pending request does not get processed.

* UIProcess/WebProcessCache.cpp:
(WebKit::generateAddRequestIdentifier):
(WebKit::WebProcessCache::addProcessIfPossible):
(WebKit::WebProcessCache::addProcess):
(WebKit::WebProcessCache::clear):
(WebKit::WebProcessCache::clearAllProcessesForSession):
(WebKit::WebProcessCache::removeProcess):
(WebKit::WebProcessCache::CachedProcess::evictionTimerFired):
(WebKit::WebProcessCache::evictProcess): Deleted.
* UIProcess/WebProcessCache.h:
(WebKit::WebProcessCache::size const):
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::processForNavigationInternal):
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::processDidTerminateOrFailedToLaunch):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242726 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoUnreviewed. Manually rolling out r242701 and r242703 since the changes
zandobersek@gmail.com [Mon, 11 Mar 2019 19:21:48 +0000 (19:21 +0000)]
Unreviewed. Manually rolling out r242701 and r242703 since the changes
are causing test timeouts and crashes on GTK and WPE.

* platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
(WebCore::MediaPlayerPrivateGStreamer::createGSTPlayBin):
* platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.cpp:
(WebCore::MediaPlayerPrivateGStreamerBase::~MediaPlayerPrivateGStreamerBase):
(WebCore::MediaPlayerPrivateGStreamerBase::flushCurrentBuffer):
(WebCore::MediaPlayerPrivateGStreamerBase::createGLAppSink):
* platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.h:
* platform/graphics/texmap/TextureMapperPlatformLayerProxy.cpp:
(WebCore::TextureMapperPlatformLayerProxy::pushNextBuffer):
(WebCore::TextureMapperPlatformLayerProxy::dropCurrentBufferWhilePreservingTexture):
* platform/graphics/texmap/TextureMapperPlatformLayerProxy.h:
(): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242725 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[CMake] Build 32bit binaries on Linux/64bit when the --32-bit is passed to build-jsc
commit-queue@webkit.org [Mon, 11 Mar 2019 19:11:09 +0000 (19:11 +0000)]
[CMake] Build 32bit binaries on Linux/64bit when the --32-bit is passed to build-jsc
https://bugs.webkit.org/show_bug.cgi?id=194147

Patch by Xan Lopez <xan@igalia.com> on 2019-03-11
Reviewed by Michael Saboff.

.:

* CMakeLists.txt: set WTF_CPU properly if FORCE_32BIT is set in
build-jsc.

Tools:

To make --32-bit work correctly on Linux/64bit we need to:

- Set FORCE_32BIT on, which will be read by CMake to set WTF_CPU
correctly. Ideally we'd just redefine CMAKE_SYSTEM_PROCESSOR, but
unfortunately CMake only allows us to do this during
crosscompilation, which is overkill here.
- Set CMAKE_PREFIX_PATH and CMAKE_LIBRARY_ARCHITECTURE so that the
pkg-config detection module uses the x86 .pc files instead of the
x86_64 ones.
- Set the -m32 flags for the compiler.

* Scripts/webkitdirs.pm:
(generateBuildSystemFromCMakeProject):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242724 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoREGRESSION: ( r240978-r240985 ) [ iOS Release ] Layout Test imported/w3c/web-platform...
commit-queue@webkit.org [Mon, 11 Mar 2019 19:07:38 +0000 (19:07 +0000)]
REGRESSION: ( r240978-r240985 ) [ iOS Release ] Layout Test imported/w3c/web-platform-tests/xhr/send-redirect-post-upload.htm is crashing
https://bugs.webkit.org/show_bug.cgi?id=194523

Patch by Alex Christensen <achristensen@webkit.org> on 2019-03-11
Reviewed by Alexey Proskuryakov.

Attempt another workaround to prevent crashes.

* NetworkProcess/cocoa/NetworkDataTaskCocoa.h:
* NetworkProcess/cocoa/NetworkDataTaskCocoa.mm:
(WebKit::NetworkDataTaskCocoa::~NetworkDataTaskCocoa):
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(-[WKNetworkSessionDelegate URLSession:task:needNewBodyStream:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242723 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[JSC] BuiltinExecutables should behave like a WeakSet instead of generic WeakHandleOw...
ysuzuki@apple.com [Mon, 11 Mar 2019 18:59:23 +0000 (18:59 +0000)]
[JSC] BuiltinExecutables should behave like a WeakSet instead of generic WeakHandleOwner for memory footprint
https://bugs.webkit.org/show_bug.cgi?id=195508

Reviewed by Darin Adler.

Weak<> is not cheap in terms of memory footprint. We allocate WeakBlock (256 bytes) for book-keeping Weak<>.
Currently BuiltinExecutables has 203 Weak<> members and many WeakBlocks are actually allocated because
many UnlinkedFunctionExecutables in BuiltinExecutables are allocated during JSGlobalObject initialization process.

This patch changes two things in BuiltinExecutables.

1. Previously we have m_xxxSourceCode fields too. But we do not need to keep it since we know how to produce it when it is required.
   We generate SourceCode in xxxSourceCode() method instead of just returning m_xxxSourceCode. This reduces sizeof(BuiltinExecutables) 24 x 203 = 4KB.

2. Instead of using Weak<>, BuiltinExecutables holds raw array of UnlinkedFunctionExecutable*. And Heap::finalizeUnconditionalFinalizers() correctly clears dead executables.
   This is similar to JSWeakSet implementation. And it saves WeakBlock allocations.

* builtins/BuiltinExecutables.cpp:
(JSC::BuiltinExecutables::BuiltinExecutables):
(JSC::BuiltinExecutables::finalizeUnconditionally):
(JSC::JSC_FOREACH_BUILTIN_CODE): Deleted.
(JSC::BuiltinExecutables::finalize): Deleted.
* builtins/BuiltinExecutables.h:
(JSC::BuiltinExecutables::static_cast<unsigned>):
(): Deleted.
* heap/Heap.cpp:
(JSC::Heap::finalizeUnconditionalFinalizers):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242722 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoAdd MotionMark-1.1 plan file for run-benchmark script.
dewei_zhu@apple.com [Mon, 11 Mar 2019 18:45:52 +0000 (18:45 +0000)]
Add MotionMark-1.1 plan file for run-benchmark script.
https://bugs.webkit.org/show_bug.cgi?id=195481

Rubber-stamped by Darin Adler.

Run-benchmark script should support MontionMark-1.1.

* Scripts/webkitpy/benchmark_runner/data/plans/motionmark1.1.plan: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242721 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoUse AVContentKeySession for "com.apple.fps.2_0" CDM version when AVStreamSession...
jer.noble@apple.com [Mon, 11 Mar 2019 18:30:11 +0000 (18:30 +0000)]
Use AVContentKeySession for "com.apple.fps.2_0" CDM version when AVStreamSession is absent
https://bugs.webkit.org/show_bug.cgi?id=195462
<rdar://problem/48712306>

Reviewed by Eric Carlson.

The difference between "com.apple.fps.2_0" and "3_0" is a protocol difference more than an
implementation difference. In "2_0", the "EME nitialization" data comes in the form of a "content
identifier", while the true initialization data is retrieved through a side channel directly from
the attached element. In "3_0", the "EME initialization data" is the exact initialization data
given by the parser, with no "content identifier" at all.

In the original implementation, the "2_0" used AVStreamSession, and "3_0" used AVContentKeySession,
but in the absense of AVStreamSession, those protocol differences are minor and can be implemented
using AVContentKeySession.

Changes:

- Add a new helper struct in CDMPrivateMediaSourceAVFObjC that represents the parsed parameters
  of the CDM string.
- Add an "initData()" accessor to SourceBufferPrivateAVFObjC so that the "2_0" path can implement
  the side channel access to the necessary initialization data.
- Refactor some of the SPI code to not re-declare unnecessary APIs.
- In CDMSessionAVContentKeySession::generateKeyRequest(), this function can never be called twice
  so it is a logical impossibility to have a certificate at this point. Remove all this if() code.

* Modules/encryptedmedia/legacy/LegacyCDM.cpp:
* platform/graphics/avfoundation/CDMPrivateMediaSourceAVFObjC.h:
* platform/graphics/avfoundation/CDMPrivateMediaSourceAVFObjC.mm:
(WebCore::CDMPrivateMediaSourceAVFObjC::parseKeySystem):
(WebCore::queryDecoderAvailability):
(WebCore::CDMPrivateMediaSourceAVFObjC::supportsKeySystem):
(WebCore::CDMPrivateMediaSourceAVFObjC::createSession):
(WebCore::validKeySystemRE): Deleted.
* platform/graphics/avfoundation/objc/CDMSessionAVContentKeySession.h:
* platform/graphics/avfoundation/objc/CDMSessionAVContentKeySession.mm:
(WebCore::CDMSessionAVContentKeySession::CDMSessionAVContentKeySession):
(WebCore::CDMSessionAVContentKeySession::generateKeyRequest):
(WebCore::CDMSessionAVContentKeySession::update):
(WebCore::CDMSessionAVContentKeySession::addParser):
(WebCore::CDMSessionAVContentKeySession::removeParser):
(WebCore::CDMSessionAVContentKeySession::contentKeySession):
* platform/graphics/avfoundation/objc/CDMSessionAVStreamSession.h:
* platform/graphics/avfoundation/objc/CDMSessionAVStreamSession.mm:
(WebCore::CDMSessionAVStreamSession::CDMSessionAVStreamSession):
* platform/graphics/avfoundation/objc/MediaPlayerPrivateMediaSourceAVFObjC.h:
* platform/graphics/avfoundation/objc/MediaPlayerPrivateMediaSourceAVFObjC.mm:
(WebCore::MediaPlayerPrivateMediaSourceAVFObjC::setCDMSession):
* platform/graphics/avfoundation/objc/SourceBufferPrivateAVFObjC.h:
* platform/graphics/avfoundation/objc/SourceBufferPrivateAVFObjC.mm:
(WebCore::SourceBufferPrivateAVFObjC::didProvideContentKeyRequestInitializationDataForTrackID):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242720 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[ews-app] Add support for submit-to-ews url
aakash_jain@apple.com [Mon, 11 Mar 2019 18:22:02 +0000 (18:22 +0000)]
[ews-app] Add support for submit-to-ews url
https://bugs.webkit.org/show_bug.cgi?id=195477

Reviewed by Lucas Forschler.

* BuildSlaveSupport/ews-app/ews/fetcher.py:
* BuildSlaveSupport/ews-app/ews/templates/statusbubble.html:
* BuildSlaveSupport/ews-app/ews/templates/submittoews.html: Copied from QueueStatusServer/templates/submittoews.html.
* BuildSlaveSupport/ews-app/ews/urls.py:
* BuildSlaveSupport/ews-app/ews/views/submittoews.py: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242719 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoIntlDateTimeFormat can be shrunk by 32 bytes
rmorisset@apple.com [Mon, 11 Mar 2019 17:54:59 +0000 (17:54 +0000)]
IntlDateTimeFormat can be shrunk by 32 bytes
https://bugs.webkit.org/show_bug.cgi?id=195504

Reviewed by Darin Adler.

* runtime/IntlDateTimeFormat.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242718 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoIntlCollator can be shrunk by 16 bytes
rmorisset@apple.com [Mon, 11 Mar 2019 17:39:35 +0000 (17:39 +0000)]
IntlCollator can be shrunk by 16 bytes
https://bugs.webkit.org/show_bug.cgi?id=195503

Reviewed by Darin Adler.

* runtime/IntlCollator.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242717 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoIntlNumberFormat can be shrunk by 16 bytes
rmorisset@apple.com [Mon, 11 Mar 2019 17:27:47 +0000 (17:27 +0000)]
IntlNumberFormat can be shrunk by 16 bytes
https://bugs.webkit.org/show_bug.cgi?id=195505

Reviewed by Darin Adler.

* runtime/IntlNumberFormat.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242716 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[ESNext][BigInt] Implement "~" unary operation
ticaiolima@gmail.com [Mon, 11 Mar 2019 17:21:41 +0000 (17:21 +0000)]
[ESNext][BigInt] Implement "~" unary operation
https://bugs.webkit.org/show_bug.cgi?id=182216

Reviewed by Keith Miller.

JSTests:

* stress/big-int-bit-not-general.js: Added.
* stress/big-int-bitwise-not-jit.js: Added.
* stress/big-int-bitwise-not-wrapped-value.js: Added.
* stress/bit-op-with-object-returning-int32.js:
* stress/bitwise-not-fixup-rules.js: Added.
* stress/value-bit-not-ai-rule.js: Added.

PerformanceTests:

* BigIntBench/big-int-simple-bit-not.js: Added.

Source/JavaScriptCore:

This patch is adding support of BigInt into op_bitnot operations. In
addition, we are changing ArithBitNot to handle only Number operands,
while introducing a new node named ValueBitNot to handle Untyped and
BigInt. This node follows the same approach we are doing into other
arithimetic operations into DFG.

* dfg/DFGAbstractInterpreterInlines.h:
(JSC::DFG::AbstractInterpreter<AbstractStateType>::executeEffects):

It is possible that fixup and prediction propagation don't convert a
ValueBitNot(ConstInt32) into ArithBitNot(ConstInt32) because these
analysis are conservative. In such case, we are adding constant
folding rules to ValueBitNot AI.

* dfg/DFGBackwardsPropagationPhase.cpp:
(JSC::DFG::BackwardsPropagationPhase::propagate):

ValueBitNot has same rules as ArithBitNot on backwards propagation.

* dfg/DFGByteCodeParser.cpp:
(JSC::DFG::ByteCodeParser::parseBlock):

We can emit ArithBitNot if we know that operand of op_bitnot is a
Number or any int. Otherwise we fallback to ValueBitNot and rely on
fixup to convert the node to ArithBitNot when it is possible.
ValueBitNot uses heap prediction on prediction propagation and we
collect its type from op_bitnot's value profiler.

* dfg/DFGClobberize.h:
(JSC::DFG::clobberize):

When we have the case with ValueBitNot(BigInt), we don't clobberize
world.

* dfg/DFGDoesGC.cpp:
(JSC::DFG::doesGC):

ValueBitNot can GC on BigIntUse because, right now, all bitNot
operation allocates temporary BigInts to perform calculations and it
can potentially trigger GC.

* dfg/DFGFixupPhase.cpp:
(JSC::DFG::FixupPhase::fixupNode):

ValueBitNot is responsible do handle BigIntUse and UntypedUse. To all
other uses, we fallback to ArithBitNot.

* dfg/DFGNode.h:
(JSC::DFG::Node::hasHeapPrediction):
* dfg/DFGNodeType.h:
* dfg/DFGOperations.cpp:
(JSC::DFG::bitwiseBinaryOp):

This template function is abstracting the new semantics of numeric
values operations on bitwise operations. These operations usually
folow these steps:

    1. rhsNumeric = GetInt32OrBigInt(rhs)
    2. lhsNumeric = GetInt32OrBigInt(lhs)
    3. trhow error if TypeOf(rhsNumeric) != TypeOf(lhsNumeric)
    4. return BigInt::bitwiseOp(bitOp, rhs, lhs) if TypeOf(lhsNumeric) == BigInt
    5. return rhs <int32BitOp> lhs

Since we have almost the same code for every bitwise op,
we use such template to avoid code duplication. The template receives
Int32 and BigInt operations as parameter. Error message is received as
`const char*` instead of `String&` to avoid String allocation even when
there is no error to throw.

* dfg/DFGOperations.h:
* dfg/DFGPredictionPropagationPhase.cpp:
* dfg/DFGSafeToExecute.h:
(JSC::DFG::safeToExecute):
* dfg/DFGSpeculativeJIT.cpp:
(JSC::DFG::SpeculativeJIT::compileValueBitNot):

ValueBitNot generates speculative code for BigIntUse and this code is a
call to `operationBitNotBigInt`. This operation is faster than
`operationValueBitNot` because there is no need to check types of
operands and execute properly operation. We still need to check
exceptions after `operationBitNotBigInt` because it can throw OOM.

(JSC::DFG::SpeculativeJIT::compileBitwiseNot):
* dfg/DFGSpeculativeJIT.h:
* dfg/DFGSpeculativeJIT32_64.cpp:
(JSC::DFG::SpeculativeJIT::compile):
* dfg/DFGSpeculativeJIT64.cpp:
(JSC::DFG::SpeculativeJIT::compile):
* ftl/FTLCapabilities.cpp:
(JSC::FTL::canCompile):
* ftl/FTLLowerDFGToB3.cpp:
(JSC::FTL::DFG::LowerDFGToB3::compileNode):
(JSC::FTL::DFG::LowerDFGToB3::compileValueBitNot):
(JSC::FTL::DFG::LowerDFGToB3::compileArithBitNot):
* runtime/CommonSlowPaths.cpp:
(JSC::SLOW_PATH_DECL):
* runtime/JSBigInt.cpp:
(JSC::JSBigInt::bitwiseNot):
* runtime/JSBigInt.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242715 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoUnreviewed, rolling out r242688, r242643, r242624.
ryanhaddad@apple.com [Mon, 11 Mar 2019 17:19:20 +0000 (17:19 +0000)]
Unreviewed, rolling out r242688, r242643, r242624.

Caused multiple layout test failures and crashes on iOS and macOS.

Reverted changeset:

"requestAnimationFrame should execute before the next frame"
https://bugs.webkit.org/show_bug.cgi?id=177484
https://trac.webkit.org/changeset/242624/webkit

Source/WebCore:

* Sources.txt:
* WebCore.xcodeproj/project.pbxproj:
* accessibility/mac/AXObjectCacheMac.mm:
(WebCore::AXObjectCache::platformHandleFocusedUIElementChanged):
* animation/DocumentAnimationScheduler.cpp: Added.
(WebCore::DocumentAnimationScheduler::create):
(WebCore::DocumentAnimationScheduler::DocumentAnimationScheduler):
(WebCore::DocumentAnimationScheduler::detachFromDocument):
(WebCore::DocumentAnimationScheduler::scheduleWebAnimationsResolution):
(WebCore::DocumentAnimationScheduler::unscheduleWebAnimationsResolution):
(WebCore::DocumentAnimationScheduler::scheduleScriptedAnimationResolution):
(WebCore::DocumentAnimationScheduler::displayRefreshFired):
(WebCore::DocumentAnimationScheduler::windowScreenDidChange):
(WebCore::DocumentAnimationScheduler::createDisplayRefreshMonitor const):
* animation/DocumentAnimationScheduler.h: Renamed from Source/WebCore/page/RenderingUpdateScheduler.h.
(WebCore::DocumentAnimationScheduler::lastTimestamp):
(WebCore::DocumentAnimationScheduler::isFiring const):
* animation/DocumentTimeline.cpp:
(WebCore::DocumentTimeline::DocumentTimeline):
(WebCore::DocumentTimeline::updateThrottlingState):
(WebCore::DocumentTimeline::suspendAnimations):
(WebCore::DocumentTimeline::resumeAnimations):
(WebCore::DocumentTimeline::liveCurrentTime const):
(WebCore::DocumentTimeline::currentTime):
(WebCore::DocumentTimeline::cacheCurrentTime):
(WebCore::DocumentTimeline::scheduleAnimationResolutionIfNeeded):
(WebCore::DocumentTimeline::animationTimingDidChange):
(WebCore::DocumentTimeline::scheduleAnimationResolution):
(WebCore::DocumentTimeline::unscheduleAnimationResolution):
(WebCore::DocumentTimeline::animationResolutionTimerFired):
(WebCore::DocumentTimeline::updateAnimationsAndSendEvents):
(WebCore::DocumentTimeline::scheduleNextTick):
(WebCore::DocumentTimeline::updateListOfElementsWithRunningAcceleratedAnimationsForElement):
(WebCore::DocumentTimeline::resolveAnimationsForElement):
(WebCore::DocumentTimeline::internalUpdateAnimationsAndSendEvents): Deleted.
* animation/DocumentTimeline.h:
* dom/Document.cpp:
(WebCore::Document::resolveStyle):
(WebCore::Document::prepareForDestruction):
(WebCore::Document::windowScreenDidChange):
(WebCore::Document::updateIntersectionObservations):
(WebCore::Document::scheduleForcedIntersectionObservationUpdate):
(WebCore::Document::animationScheduler):
(WebCore::Document::updateAnimationsAndSendEvents): Deleted.
(WebCore::Document::serviceRequestAnimationFrameCallbacks): Deleted.
* dom/Document.h:
* dom/ScriptedAnimationController.cpp:
(WebCore::ScriptedAnimationController::serviceScriptedAnimations):
(WebCore::ScriptedAnimationController::scheduleAnimation):
(WebCore::ScriptedAnimationController::animationTimerFired):
(WebCore::ScriptedAnimationController::documentAnimationSchedulerDidFire):
(WebCore::ScriptedAnimationController::serviceRequestAnimationFrameCallbacks): Deleted.
* dom/ScriptedAnimationController.h:
* page/FrameView.cpp:
(WebCore::FrameView::viewportContentsChanged):
* page/IntersectionObserver.cpp:
(WebCore::IntersectionObserver::observe):
* page/Page.cpp:
(WebCore::Page::Page):
(WebCore::Page::willDisplayPage):
(WebCore::Page::addDocumentNeedingIntersectionObservationUpdate):
(WebCore::Page::updateIntersectionObservations):
(WebCore::Page::scheduleForcedIntersectionObservationUpdate):
(WebCore::Page::layoutIfNeeded): Deleted.
(WebCore::Page::renderingUpdate): Deleted.
(WebCore::Page::renderingUpdateScheduler): Deleted.
* page/Page.h:
* page/PageOverlayController.cpp:
(WebCore::PageOverlayController::didChangeViewExposedRect):
(WebCore::PageOverlayController::notifyFlushRequired):
* page/RenderingUpdateScheduler.cpp: Removed.
* page/ios/ContentChangeObserver.h:
* page/mac/ServicesOverlayController.mm:
(WebCore::ServicesOverlayController::Highlight::notifyFlushRequired):
* rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::scheduleLayerFlushNow):

Source/WebKit:

* WebProcess/WebPage/CoordinatedGraphics/DrawingAreaCoordinatedGraphics.cpp:
(WebKit::DrawingAreaCoordinatedGraphics::scheduleCompositingLayerFlush):
(WebKit::DrawingAreaCoordinatedGraphics::updateBackingStoreState):
(WebKit::DrawingAreaCoordinatedGraphics::display):
* WebProcess/WebPage/CoordinatedGraphics/LayerTreeHost.cpp:
(WebKit::LayerTreeHost::layerFlushTimerFired):
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.mm:
(WebKit::RemoteLayerTreeDrawingArea::flushLayers):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::layoutIfNeeded):
(WebKit::WebPage::willDisplayPage):
(WebKit::WebPage::renderingUpdate): Deleted.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.mm:
(WebKit::TiledCoreAnimationDrawingArea::flushLayers):

Source/WebKitLegacy/mac:

* WebView/WebView.mm:
(-[WebView _viewWillDrawInternal]):

Source/WebKitLegacy/win:

* WebView.cpp:
(WebView::updateBackingStore):
(WebView::flushPendingGraphicsLayerChangesSoon):
(WebView::flushPendingGraphicsLayerChanges):

Source/WTF:

* wtf/SystemTracing.h:

Tools:

* Tracing/SystemTracePoints.plist:

LayoutTests:

* TestExpectations:
* animations/animation-multiple-callbacks-timestamp.html:
* animations/no-style-recalc-during-accelerated-animation-expected.txt:
* animations/no-style-recalc-during-accelerated-animation.html:
* platform/mac-wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242714 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoSpecify fixed precision explicitly to prepare to change String::number and StringBuil...
darin@apple.com [Mon, 11 Mar 2019 17:11:13 +0000 (17:11 +0000)]
Specify fixed precision explicitly to prepare to change String::number and StringBuilder::appendNumber floating point behavior
https://bugs.webkit.org/show_bug.cgi?id=195533

Reviewed by Brent Fulgham.

Source/JavaScriptCore:

* API/tests/ExecutionTimeLimitTest.cpp:
(testExecutionTimeLimit): Use appendFixedPrecisionNumber.
* runtime/NumberPrototype.cpp:
(JSC::numberProtoFuncToPrecision): Use numberToStringFixedPrecision.
* runtime/Options.cpp:
(JSC::Option::dump const): Use appendFixedPrecisionNumber.

Source/WebCore:

* accessibility/AccessibilityNodeObject.cpp:
(WebCore::AccessibilityNodeObject::changeValueByStep): Use numberToStringFixedPrecision.
(WebCore::AccessibilityNodeObject::changeValueByPercent): Ditto.
* accessibility/AccessibilityScrollbar.cpp:
(WebCore::AccessibilityScrollbar::setValue): Ditto.
* css/CSSFontVariationValue.cpp:
(WebCore::CSSFontVariationValue::customCSSText const): Use appendFixedPrecisionNumber.
* css/CSSGradientValue.cpp:
(WebCore::CSSLinearGradientValue::customCSSText const): Ditto.
(WebCore::CSSRadialGradientValue::customCSSText const): Ditto.
* css/CSSKeyframeRule.cpp:
(WebCore::StyleRuleKeyframe::keyText const): Ditto.
* css/CSSTimingFunctionValue.cpp:
(WebCore::CSSCubicBezierTimingFunctionValue::customCSSText const): Ditto.
(WebCore::CSSSpringTimingFunctionValue::customCSSText const): Ditto.
* css/parser/CSSParserToken.cpp:
(WebCore::CSSParserToken::serialize const): Ditto.
* html/HTMLImageElement.cpp:
(WebCore::HTMLImageElement::completeURLsInAttributeValue const): Ditto.
* inspector/InspectorOverlay.cpp:
(WebCore::InspectorOverlay::drawRulers): Use numberToStringFixedPrecision.
* loader/ResourceLoadStatistics.cpp:
(WebCore::ResourceLoadStatistics::toString const): Use appendFixedPrecisionNumber.
* page/PrintContext.cpp:
(WebCore::PrintContext::pageProperty): Use numberToStringFixedPrecision.
* page/cocoa/ResourceUsageOverlayCocoa.mm:
(WebCore::gcTimerString): Use numberToStringFixedPrecision.
* platform/LayoutUnit.h:
(WTF::ValueToString<WebCore::LayoutUnit>::string): Ditto.
* platform/graphics/Color.cpp:
(WebCore::Color::cssText const): Use appendFixedPrecisionNumber.
* platform/graphics/ExtendedColor.cpp:
(WebCore::ExtendedColor::cssText const): Ditto.
* svg/SVGAngleValue.cpp:
(WebCore::SVGAngleValue::valueAsString const): Use numberToStringFixedPrecision.
* svg/SVGNumberListValues.cpp:
(WebCore::SVGNumberListValues::valueAsString const): Use appendFixedPrecisionNumber.
* svg/SVGPathStringBuilder.cpp:
(WebCore::appendNumber): Ditto.
(WebCore::appendPoint): Ditto.
* svg/SVGPointListValues.cpp:
(WebCore::SVGPointListValues::valueAsString const): Ditto.
* svg/SVGTransformValue.cpp:
(WebCore::SVGTransformValue::valueAsString const): Ditto.
* svg/properties/SVGPropertyTraits.h:
(WebCore::SVGPropertyTraits<float>::toString): Use numberToStringFixedPrecision.
(WebCore::SVGPropertyTraits<FloatPoint>::toString): Use appendFixedPrecisionNumber.
(WebCore::SVGPropertyTraits<FloatRect>::toString): Ditto.
* testing/Internals.cpp:
(WebCore::Internals::dumpMarkerRects): Use appendFixedPrecisionNumber.
(WebCore::Internals::getCurrentCursorInfo): Ditto.
* xml/XPathValue.cpp:
(WebCore::XPath::Value::toString const): Use numberToStringFixedPrecision.

Source/WebKit:

* NetworkProcess/cache/NetworkCache.cpp:
(WebKit::NetworkCache::Cache::dumpContentsToFile): Use appendFixedPrecisionNumber.
* NetworkProcess/cache/NetworkCacheEntry.cpp:
(WebKit::NetworkCache::Entry::asJSON const): Ditto.
* Shared/Gamepad/GamepadData.cpp:
(WebKit::GamepadData::loggingString const): Ditto.
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::logDiagnosticMessageWithValue): Use numberToStringFixedPrecision.

Source/WTF:

Soon, we will change String::number and StringBuilder::appendNumber for floating
point to use "shortest form" serialization instead of the current default, which is
"6-digit fixed precision stripping trailing zeros". To prepare to do this safely
without accidentally changing any behavior, changing callers to call the explicit
versions. Later, we may want to return and change many of them to use shortest form
instead, but that may require rebaselining tests, and in some extreme cases, getting
rid of flawed logic that converts between different single and double precision
floating point; such problems may be hidden by fixed precision serialization.

Since "shortest form" is already the behavior for AtomicString::number and
for makeString, no changes required for clients of either of those.

* wtf/Logger.h:
(WTF::LogArgument::toString): Use numberToStringFixedPrecision.
* wtf/MediaTime.cpp:
(WTF::MediaTime::toString const): Use appendFixedPrecisionNumber.
* wtf/text/ValueToString.h:
(WTF::ValueToString<float>::string): Use numberToStringFixedPrecision.
(WTF::ValueToString<double>::string): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242713 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoResource Load Statistics: Make it possible exclude localhost from classification
wilander@apple.com [Mon, 11 Mar 2019 17:01:27 +0000 (17:01 +0000)]
Resource Load Statistics: Make it possible exclude localhost from classification
https://bugs.webkit.org/show_bug.cgi?id=195474
<rdar://problem/47520577>

Reviewed by Brent Fulgham.

Source/WebKit:

This patch allows for localhost to be excluded from classification and
treatment as a prevalent resource.

* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.cpp:
(WebKit::ResourceLoadStatisticsDatabaseStore::ResourceLoadStatisticsDatabaseStore):
(WebKit::ResourceLoadStatisticsDatabaseStore::reclassifyResources):
    Makes use of the new ResourceLoadStatisticsMemoryStore::shouldSkip().
(WebKit::ResourceLoadStatisticsDatabaseStore::setPrevalentResource):
    Makes use of the new ResourceLoadStatisticsMemoryStore::shouldSkip().
(WebKit::ResourceLoadStatisticsDatabaseStore::isPrevalentResource const):
    Makes use of the new ResourceLoadStatisticsMemoryStore::shouldSkip().
(WebKit::ResourceLoadStatisticsDatabaseStore::isVeryPrevalentResource const):
    Makes use of the new ResourceLoadStatisticsMemoryStore::shouldSkip().
(WebKit::ResourceLoadStatisticsDatabaseStore::setVeryPrevalentResource):
    Makes use of the new ResourceLoadStatisticsMemoryStore::shouldSkip().
* NetworkProcess/Classifier/ResourceLoadStatisticsDatabaseStore.h:
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.cpp:
(WebKit::ResourceLoadStatisticsMemoryStore::ResourceLoadStatisticsMemoryStore):
(WebKit::ResourceLoadStatisticsMemoryStore::classifyPrevalentResources):
    Makes use of the new ResourceLoadStatisticsMemoryStore::shouldSkip().
(WebKit::ResourceLoadStatisticsMemoryStore::setPrevalentResource):
    Makes use of the new ResourceLoadStatisticsMemoryStore::shouldSkip().
(WebKit::ResourceLoadStatisticsMemoryStore::isPrevalentResource const):
    Makes use of the new ResourceLoadStatisticsMemoryStore::shouldSkip().
(WebKit::ResourceLoadStatisticsMemoryStore::isVeryPrevalentResource const):
    Makes use of the new ResourceLoadStatisticsMemoryStore::shouldSkip().
(WebKit::ResourceLoadStatisticsMemoryStore::setVeryPrevalentResource):
    Makes use of the new ResourceLoadStatisticsMemoryStore::shouldSkip().
* NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.h:
* NetworkProcess/Classifier/ResourceLoadStatisticsStore.cpp:
(WebKit::ResourceLoadStatisticsStore::ResourceLoadStatisticsStore):
    Now takes a ShouldIncludeLocalhost parameter.
(WebKit::ResourceLoadStatisticsStore::shouldSkip const):
    Convenience function, currently supporting the localhost exclusion.
(WebKit::ResourceLoadStatisticsStore::setIsRunningTest):
    Test infrastructure.
* NetworkProcess/Classifier/ResourceLoadStatisticsStore.h:
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:
(WebKit::WebResourceLoadStatisticsStore::setIsRunningTest):
    Test infrastructure.
(WebKit::WebResourceLoadStatisticsStore::WebResourceLoadStatisticsStore):
    Now takes a ShouldIncludeLocalhost parameter.
* NetworkProcess/Classifier/WebResourceLoadStatisticsStore.h:
    Defines the new ShouldIncludeLocalhost boolean enum.
* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::setIsRunningResourceLoadStatisticsTest):
    Test infrastructure.
* NetworkProcess/NetworkProcess.h:
* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/NetworkSession.cpp:
(WebKit::NetworkSession::setResourceLoadStatisticsEnabled):
    Forwards the localhost setting to the create function.
* NetworkProcess/NetworkSession.h:
* NetworkProcess/NetworkSessionCreationParameters.cpp:
(WebKit::NetworkSessionCreationParameters::encode const):
(WebKit::NetworkSessionCreationParameters::decode):
* NetworkProcess/NetworkSessionCreationParameters.h:
    New parameter called shouldIncludeLocalhostInResourceLoadStatistics.
* NetworkProcess/cocoa/NetworkSessionCocoa.mm:
(WebKit::NetworkSessionCocoa::NetworkSessionCocoa):
    Picks up the localhost setting from the parameters.
* UIProcess/API/C/WKWebsiteDataStoreRef.cpp:
(WKWebsiteDataStoreSetStatisticsIsRunningTest):
    Test infrastructure.
* UIProcess/API/C/WKWebsiteDataStoreRef.h:
* UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::setIsRunningResourceLoadStatisticsTest):
    Test infrastructure.
* UIProcess/Network/NetworkProcessProxy.h:
* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::ensureNetworkProcess):
    Picks up the localhost setting from the WebsiteDataStore parameters.
* UIProcess/WebsiteData/Cocoa/WebsiteDataStoreCocoa.mm:
(WebKit::WebsiteDataStore::parameters):
    Makes sure Safari does not exclude localhost.
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::setIsRunningResourceLoadStatisticsTest):
    Test infrastructure.
* UIProcess/WebsiteData/WebsiteDataStore.h:

Tools:

This patch allows for localhost to be excluded from classification and
treatment as a prevalent resource.

The WebKit Tools change adds a new function called
testRunner.setStatisticsIsRunningTest() which can be used to control this
behavior.

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

LayoutTests:

This patch makes sure that all test cases that need to, call the enableFeature()
function in http/tests/resourceLoadStatistics/resources/util.js.

The enableFeature() now calls the new function testRunner.setStatisticsIsRunningTest().

* http/tests/resourceLoadStatistics/do-not-block-top-level-navigation-redirect.html:
* http/tests/resourceLoadStatistics/non-sandboxed-iframe-redirect-ip-to-localhost-to-ip.html:
* http/tests/resourceLoadStatistics/prevalent-resource-with-user-interaction.html:
* http/tests/resourceLoadStatistics/resources/set-cookie-on-redirect.php:
* http/tests/resourceLoadStatistics/resources/util.js:
(setEnableFeature):
* http/tests/storageAccess/has-storage-access-from-prevalent-domain-with-user-interaction.html:
* http/tests/storageAccess/request-and-grant-access-cross-origin-sandboxed-iframe-from-prevalent-domain-with-user-interaction.html:
* http/tests/storageAccess/request-and-grant-access-cross-origin-sandboxed-iframe-from-prevalent-domain-without-user-interaction.html:
* http/tests/storageAccess/request-and-grant-access-cross-origin-sandboxed-iframe.html:
* http/tests/storageAccess/request-storage-access-cross-origin-sandboxed-iframe-with-unique-origin.html:
* http/tests/storageAccess/request-storage-access-same-origin-iframe.html:
* http/tests/storageAccess/request-storage-access-same-origin-sandboxed-iframe.html:
* http/tests/storageAccess/request-storage-access-top-frame.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242712 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[ews-build] Disable waterfall and console view for ews-build.webkit.org
aakash_jain@apple.com [Mon, 11 Mar 2019 16:56:57 +0000 (16:56 +0000)]
[ews-build] Disable waterfall and console view for ews-build.webkit.org
https://bugs.webkit.org/show_bug.cgi?id=195560

Reviewed by Lucas Forschler.

* BuildSlaveSupport/ews-build/master.cfg:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242711 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoUnreviewed, rolling out r242698.
achristensen@apple.com [Mon, 11 Mar 2019 16:50:32 +0000 (16:50 +0000)]
Unreviewed, rolling out r242698.

API test crashes on bots.

Reverted changeset:

"Add a WKContentRuleList variant that uses copied memory
instead of mmap'd shared memory for class A containerized
apps"
https://bugs.webkit.org/show_bug.cgi?id=195511
https://trac.webkit.org/changeset/242698

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242710 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[WPE] Enable web process sandbox
mcatanzaro@igalia.com [Mon, 11 Mar 2019 16:47:21 +0000 (16:47 +0000)]
[WPE] Enable web process sandbox
https://bugs.webkit.org/show_bug.cgi?id=195169

Reviewed by Daniel Bates.

.:

* Source/cmake/BubblewrapSandboxChecks.cmake: Added.
* Source/cmake/OptionsGTK.cmake:
* Source/cmake/OptionsWPE.cmake:

Source/WebKit:

* PlatformWPE.cmake:
* UIProcess/Launcher/glib/BubblewrapLauncher.cpp:
(WebKit::bubblewrapSpawn):

Tools:

* wpe/install-dependencies:
* wpe/jhbuild.modules:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242709 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[ews-app] Use port 17000 for worker communication
aakash_jain@apple.com [Mon, 11 Mar 2019 16:40:55 +0000 (16:40 +0000)]
[ews-app] Use port 17000 for worker communication
https://bugs.webkit.org/show_bug.cgi?id=195558

Reviewed by Lucas Forschler.

* BuildSlaveSupport/ews-build/master.cfg:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242708 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[ews-build] unit-tests fail when passwords.json is missing
aakash_jain@apple.com [Mon, 11 Mar 2019 16:38:06 +0000 (16:38 +0000)]
[ews-build] unit-tests fail when passwords.json is missing
https://bugs.webkit.org/show_bug.cgi?id=195557

Reviewed by Lucas Forschler.

* BuildSlaveSupport/ews-build/loadConfig.py:
(loadBuilderConfig):
* BuildSlaveSupport/ews-build/loadConfig_unittest.py:
(ConfigDotJSONTest.test_configuration):
* BuildSlaveSupport/ews-build/master.cfg:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242707 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoUnreviewed, rolling out r242702.
tsavell@apple.com [Mon, 11 Mar 2019 16:25:41 +0000 (16:25 +0000)]
Unreviewed, rolling out r242702.

Broke High Sierra builders.

Reverted changeset:

"Add utility function to allow easy reverse range-based
iteration of a container"
https://bugs.webkit.org/show_bug.cgi?id=195542
https://trac.webkit.org/changeset/242702

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242706 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoUnreviewed, rolling out r242697.
tsavell@apple.com [Mon, 11 Mar 2019 15:55:07 +0000 (15:55 +0000)]
Unreviewed, rolling out r242697.

Broke internal builders.

Reverted changeset:

"Optimizing loads when creating new pages"
https://bugs.webkit.org/show_bug.cgi?id=195516
https://trac.webkit.org/changeset/242697

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242705 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoUnreviewed GTK test gardening
jfernandez@igalia.com [Mon, 11 Mar 2019 15:38:13 +0000 (15:38 +0000)]
Unreviewed GTK test gardening
https://bugs.webkit.org/show_bug.cgi?id=195551

Unreviewed test gardening.

Mark the text-transform-capitilize-026.html as flaky for all platforms.

* TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242704 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoUnreviewed, Non-GStreamer-GL build fix after r242701.
philn@webkit.org [Mon, 11 Mar 2019 15:26:04 +0000 (15:26 +0000)]
Unreviewed, Non-GStreamer-GL build fix after r242701.

* platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.cpp:
(WebCore::MediaPlayerPrivateGStreamerBase::~MediaPlayerPrivateGStreamerBase):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242703 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoAdd utility function to allow easy reverse range-based iteration of a container
commit-queue@webkit.org [Mon, 11 Mar 2019 15:25:08 +0000 (15:25 +0000)]
Add utility function to allow easy reverse range-based iteration of a container
https://bugs.webkit.org/show_bug.cgi?id=195542

Patch by Sam Weinig <sam@webkit.org> on 2019-03-11
Reviewed by Antti Koivisto.

Source/WTF:

Add functions to create an IteratorRange<T> that will iterate a container backwards. It
works with any container that is compatible with std::rbegin() and std::rend(). It is
expected to be used in conjunction with range-based for-loops like so:

for (auto& value : WTF::makeReversedRange(myContainer))
    ...

* wtf/IteratorRange.h:
(WTF::makeReversedRange):

Tools:

* TestWebKitAPI/CMakeLists.txt:
* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/Tests/WTF/IteratorRange.cpp: Added.
(TestWebKitAPI::TEST):
Add test to ensure WTF::makeReversedRange() works correctly and uses the correct types.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242702 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[GStreamer][v4l2] Synchronous video texture flushing support
philn@webkit.org [Mon, 11 Mar 2019 14:41:39 +0000 (14:41 +0000)]
[GStreamer][v4l2] Synchronous video texture flushing support
https://bugs.webkit.org/show_bug.cgi?id=195453

Reviewed by Xabier Rodriguez-Calvar.

The v4l2 video decoder currently requires that downstream users of
the graphics resources complete any pending draw call and release
resources before returning from the DRAIN query.

To accomplish this the player monitors the pipeline and whenever a
v4l2 decoder is added, synchronous video texture flushing support
is enabled. Additionally and for all decoder configurations, a
flush is performed before disposing of the player.

* platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
(WebCore::MediaPlayerPrivateGStreamer::playbinDeepElementAddedCallback):
Monitor elements added to the decodebin bin.
(WebCore::MediaPlayerPrivateGStreamer::decodebinElementAdded): Set
a flag if a v4l2 decoder was added in decodebin.
(WebCore::MediaPlayerPrivateGStreamer::createGSTPlayBin): Connect
to the deep-element-added signal so as to monitor pipeline
topology updates.
* platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.h:
* platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.cpp:
(WebCore::MediaPlayerPrivateGStreamerBase::~MediaPlayerPrivateGStreamerBase):
Flush video texture before disposing of the player.
(WebCore::MediaPlayerPrivateGStreamerBase::flushCurrentBuffer):
Synchronously flush if the pipeline contains a v4l2 decoder.
(WebCore::MediaPlayerPrivateGStreamerBase::createGLAppSink): Monitor push events only.
* platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.h:
* platform/graphics/texmap/TextureMapperPlatformLayerProxy.cpp:
(WebCore::TextureMapperPlatformLayerProxy::pushNextBuffer): New
boolean flag used mostly to trigger synchronous flush conditions.
(WebCore::TextureMapperPlatformLayerProxy::dropCurrentBufferWhilePreservingTexture):
Optionally drop the current buffer in a synchronous manner. By
default the method keeps operating asynchronously.
* platform/graphics/texmap/TextureMapperPlatformLayerProxy.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242701 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoRename contentOffsetInCompostingLayer to contentOffsetInCompositingLayer
antti@apple.com [Mon, 11 Mar 2019 12:58:57 +0000 (12:58 +0000)]
Rename contentOffsetInCompostingLayer to contentOffsetInCompositingLayer
https://bugs.webkit.org/show_bug.cgi?id=195553

Reviewed by Simon Fraser.

Less composting, more compositing.

* rendering/RenderLayerBacking.cpp:
(WebCore::RenderLayerBacking::updateBackdropFiltersGeometry):
(WebCore::RenderLayerBacking::resetContentsRect):
(WebCore::RenderLayerBacking::updateChildClippingStrategy):
(WebCore::RenderLayerBacking::updateImageContents):
(WebCore::RenderLayerBacking::contentOffsetInCompositingLayer const):
(WebCore::RenderLayerBacking::contentsBox const):
(WebCore::RenderLayerBacking::backgroundBoxForSimpleContainerPainting const):
(WebCore::RenderLayerBacking::contentOffsetInCompostingLayer const): Deleted.
* rendering/RenderLayerBacking.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242700 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoInvalid flags in a RegExp literal should be an early SyntaxError
ross.kirsling@sony.com [Mon, 11 Mar 2019 06:20:53 +0000 (06:20 +0000)]
Invalid flags in a RegExp literal should be an early SyntaxError
https://bugs.webkit.org/show_bug.cgi?id=195514

Reviewed by Darin Adler.

JSTests:

* test262/expectations.yaml:
Mark 4 test cases as passing.

* stress/regexp-syntax-error-invalid-flags.js:
* stress/regress-161995.js: Removed.
Update existing test, merging in an older test for the same behavior.

Source/JavaScriptCore:

Currently we're throwing a *runtime* SyntaxError; this should occur at parse time.

  12.2.8.1 Static Semantics: Early Errors
    PrimaryExpression : RegularExpressionLiteral
      - It is a Syntax Error if BodyText of RegularExpressionLiteral cannot be recognized
        using the goal symbol Pattern of the ECMAScript RegExp grammar specified in 21.2.1.
      - It is a Syntax Error if FlagText of RegularExpressionLiteral contains any code points
        other than "g", "i", "m",  "s", "u", or "y", or if it contains the same code point more than once.

In fixing this, let's also move flag handling from runtime/ to yarr/.

* yarr/YarrSyntaxChecker.cpp:
(JSC::Yarr::checkSyntax):
Check flags before checking pattern.

* CMakeLists.txt:
* JavaScriptCore.xcodeproj/project.pbxproj:
* Sources.txt:
* bytecompiler/NodesCodegen.cpp:
(JSC::RegExpNode::emitBytecode):
* inspector/ContentSearchUtilities.cpp:
(Inspector::ContentSearchUtilities::findMagicComment):
* runtime/CachedTypes.cpp:
* runtime/RegExp.cpp:
(JSC::RegExp::RegExp):
(JSC::RegExp::createWithoutCaching):
(JSC::RegExp::create):
(JSC::regExpFlags): Deleted.
* runtime/RegExp.h:
* runtime/RegExpCache.cpp:
(JSC::RegExpCache::lookupOrCreate):
(JSC::RegExpCache::ensureEmptyRegExpSlow):
* runtime/RegExpCache.h:
* runtime/RegExpConstructor.cpp:
(JSC::toFlags):
(JSC::regExpCreate):
(JSC::constructRegExp):
* runtime/RegExpKey.h:
(JSC::RegExpKey::RegExpKey):
(WTF::HashTraits<JSC::RegExpKey>::constructDeletedValue):
(WTF::HashTraits<JSC::RegExpKey>::isDeletedValue):
(): Deleted.
* runtime/RegExpPrototype.cpp:
(JSC::regExpProtoFuncCompile):
* testRegExp.cpp:
(parseRegExpLine):
* yarr/RegularExpression.cpp:
(JSC::Yarr::RegularExpression::Private::compile):
* yarr/YarrFlags.cpp: Added.
(JSC::Yarr::parseFlags):
* yarr/YarrFlags.h: Added.
* yarr/YarrInterpreter.h:
(JSC::Yarr::BytecodePattern::ignoreCase const):
(JSC::Yarr::BytecodePattern::multiline const):
(JSC::Yarr::BytecodePattern::sticky const):
(JSC::Yarr::BytecodePattern::unicode const):
(JSC::Yarr::BytecodePattern::dotAll const):
* yarr/YarrPattern.cpp:
(JSC::Yarr::YarrPattern::compile):
(JSC::Yarr::YarrPattern::YarrPattern):
(JSC::Yarr::YarrPattern::dumpPattern):
* yarr/YarrPattern.h:
(JSC::Yarr::YarrPattern::global const):
(JSC::Yarr::YarrPattern::ignoreCase const):
(JSC::Yarr::YarrPattern::multiline const):
(JSC::Yarr::YarrPattern::sticky const):
(JSC::Yarr::YarrPattern::unicode const):
(JSC::Yarr::YarrPattern::dotAll const):
Move flag handling to Yarr and modernize API.

Source/WebCore:

* bindings/js/SerializedScriptValue.cpp:
(WebCore::CloneDeserializer::readTerminal):
Consume YarrFlags.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242699 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoAdd a WKContentRuleList variant that uses copied memory instead of mmap'd shared...
commit-queue@webkit.org [Mon, 11 Mar 2019 06:04:00 +0000 (06:04 +0000)]
Add a WKContentRuleList variant that uses copied memory instead of mmap'd shared memory for class A containerized apps
https://bugs.webkit.org/show_bug.cgi?id=195511
<rdar://problem/44873269>

Patch by Alex Christensen <achristensen@webkit.org> on 2019-03-10
Reviewed by Darin Adler.

Source/WebKit:

* NetworkProcess/NetworkContentRuleListManager.cpp:
(WebKit::NetworkContentRuleListManager::addContentRuleLists):
* NetworkProcess/NetworkContentRuleListManager.h:
* NetworkProcess/cache/NetworkCacheFileSystem.cpp:
(WebKit::NetworkCache::pathRegisteredAsUnsafeToMemoryMapForTesting):
(WebKit::NetworkCache::registerPathAsUnsafeToMemoryMapForTesting):
(WebKit::NetworkCache::isSafeToUseMemoryMapForPath):
* NetworkProcess/cache/NetworkCacheFileSystem.h:
* Shared/WebCompiledContentRuleList.cpp:
(WebKit::WebCompiledContentRuleList::usesCopiedMemory const):
(WebKit::WebCompiledContentRuleList::conditionsApplyOnlyToDomain const):
(WebKit::WebCompiledContentRuleList::filtersWithoutConditionsBytecode const):
(WebKit::WebCompiledContentRuleList::filtersWithConditionsBytecode const):
(WebKit::WebCompiledContentRuleList::topURLFiltersBytecode const):
(WebKit::WebCompiledContentRuleList::actions const):
* Shared/WebCompiledContentRuleList.h:
* Shared/WebCompiledContentRuleListData.cpp:
(WebKit::WebCompiledContentRuleListData::size const):
(WebKit::WebCompiledContentRuleListData::dataPointer const):
(WebKit::WebCompiledContentRuleListData::encode const):
(WebKit::WebCompiledContentRuleListData::decode):
* Shared/WebCompiledContentRuleListData.h:
(WebKit::WebCompiledContentRuleListData::WebCompiledContentRuleListData):
* UIProcess/API/APIContentRuleList.cpp:
(API::ContentRuleList::usesCopiedMemory const):
* UIProcess/API/APIContentRuleList.h:
* UIProcess/API/APIContentRuleListStore.cpp:
(API::getData):
(API::decodeContentRuleListMetaData):
(API::ContentRuleListStore::readContentsOfFile):
(API::MappedOrCopiedData::dataPointer const):
(API::openAndMapOrCopyContentRuleList):
(API::compiledToFile):
(API::createExtension):
(API::ContentRuleListStore::lookupContentRuleList):
(API::ContentRuleListStore::compileContentRuleList):
(API::ContentRuleListStore::getContentRuleListSource):
(API::openAndMapContentRuleList): Deleted.
* UIProcess/API/APIContentRuleListStore.h:
* UIProcess/API/Cocoa/APIContentRuleListStoreCocoa.mm:
(API::ContentRuleListStore::readContentsOfFile):
* UIProcess/API/Cocoa/WKContentRuleListStore.mm:
(+[WKContentRuleListStore _registerPathAsUnsafeToMemoryMapForTesting:]):
* UIProcess/API/Cocoa/WKContentRuleListStorePrivate.h:
* UIProcess/API/Cocoa/_WKUserContentFilter.mm:
(-[_WKUserContentFilter usesCopiedMemory]):
* UIProcess/API/Cocoa/_WKUserContentFilterPrivate.h:
* WebProcess/UserContent/WebUserContentController.cpp:
(WebKit::WebUserContentController::addContentRuleLists):
* WebProcess/UserContent/WebUserContentController.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::m_hostFileDescriptor):

Tools:

* TestWebKitAPI/Tests/WebKitCocoa/WKContentExtensionStore.mm:
(-[TestSchemeHandlerSubresourceShouldBeBlocked webView:startURLSchemeTask:]):
(-[TestSchemeHandlerSubresourceShouldBeBlocked webView:stopURLSchemeTask:]):
(TEST_F):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242698 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoOptimizing loads when creating new pages
jiewen_tan@apple.com [Mon, 11 Mar 2019 03:41:13 +0000 (03:41 +0000)]
Optimizing loads when creating new pages
https://bugs.webkit.org/show_bug.cgi?id=195516
<rdar://problem/48738086>

Reviewed by Darin Adler.

This patch adds hooks in WebPageProxy::createNewPage to optimize loads, and moves the creationParameters
of API::NavigationAction from UI clients to WebPageProxy::createNewPage. Also, we now pass the whole
API::NavigationAction to the load optimizer instead of the request within.

* UIProcess/API/APINavigationAction.h:
* UIProcess/API/APIUIClient.h:
(API::UIClient::createNewPage):
* UIProcess/API/C/WKPage.cpp:
(WKPageSetPageUIClient):
* UIProcess/API/glib/WebKitUIClient.cpp:
* UIProcess/Cocoa/NavigationState.mm:
(WebKit::tryOptimizingLoad):
(WebKit::tryInterceptNavigation):
(WebKit::NavigationState::NavigationClient::decidePolicyForNavigationAction):
* UIProcess/Cocoa/UIDelegate.h:
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::UIDelegate::UIClient::createNewPage):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::decidePolicyForNavigationAction):
(WebKit::tryOptimizingLoad):
(WebKit::WebPageProxy::createNewPage):
* UIProcess/WebPageProxy.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242697 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoAdd SPI to retrieve the set of text inputs in a given rect, and later focus one
timothy_horton@apple.com [Mon, 11 Mar 2019 03:32:50 +0000 (03:32 +0000)]
Add SPI to retrieve the set of text inputs in a given rect, and later focus one
https://bugs.webkit.org/show_bug.cgi?id=195499

Reviewed by Darin Adler.

Source/WebCore:

New API tests: WebKit.RequestTextInputContext and WebKit.FocusTextInputContext

* WebCore.xcodeproj/project.pbxproj:
* dom/Document.cpp:
(WebCore::Document::identifierForElement):
(WebCore::Document::elementWithIdentifier):
(WebCore::Document::identifiedElementWasRemovedFromDocument):
* dom/Document.h:
Add a mechanism where Document will vend an ObjectIdentifier for a given
element, and can (if possible) retrieve that element later.

* dom/Element.cpp:
(WebCore::Element::removedFromAncestor):
If an Element has an identifier created for it, inform Document to remove
it from the identifier map when the element is detached.

(WebCore::Element::createElementIdentifier):
* dom/Element.h:
* dom/ElementIdentifier.h: Added.
* dom/ElementRareData.cpp:
* dom/ElementRareData.h:
(WebCore::ElementRareData::hasElementIdentifier const):
(WebCore::ElementRareData::setHasElementIdentifier):
(WebCore::ElementRareData::ElementRareData):
Store a bit indicating if the Element has had a identifier created for it,
so that we can avoid a hash lookup on every Element removal.

* dom/Node.h:
* html/HTMLTextFormControlElement.h:

Source/WebKit:

* Scripts/webkit/messages.py:
* Shared/TextInputContext.cpp: Added.
(IPC::ArgumentCoder<WebKit::TextInputContext>::encode):
(IPC::ArgumentCoder<WebKit::TextInputContext>::decode):
* Shared/TextInputContext.h: Added.
(WebKit::TextInputContext::operator== const):
Add TextInputContext, which represents a minimal set of information
about a text field.

* Sources.txt:
* SourcesCocoa.txt:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _requestTextInputContextsInRect:completionHandler:]):
(-[WKWebView _focusTextInputContext:completionHandler:]):
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
Add SPI that allows clients to asynchronously request text input
contexts for a given rect, and later focus a given context.

* UIProcess/API/Cocoa/_WKTextInputContext.h: Added.
* UIProcess/API/Cocoa/_WKTextInputContext.mm: Added.
(-[_WKTextInputContext _initWithTextInputContext:]):
(-[_WKTextInputContext boundingRect]):
(-[_WKTextInputContext _textInputContext]):
(-[_WKTextInputContext isEqual:]):
(-[_WKTextInputContext hash]):
(-[_WKTextInputContext copyWithZone:]):
* UIProcess/API/Cocoa/_WKTextInputContextInternal.h: Added.
Add an SPI object that exposes a read-only window on a TextInputContext to clients.

* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::textInputContextsInRect):
(WebKit::WebPageProxy::focusTextInputContext):
* UIProcess/WebPageProxy.h:
Plumbing from WKWebView<->WebPage.

* WebKit.xcodeproj/project.pbxproj:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::elementRectInWindowCoordinates):
(WebKit::isEditableTextInputElement):
(WebKit::WebPage::textInputContextsInRect):
Search the DOM for text input contexts: <input type='text'> (or other
form fields that fall back on text field behavior), <textarea>, and
contenteditable roots. Store the WebPage, Document, and Element identifiers
so that we can find the element again later.

(WebKit::WebPage::focusTextInputContext):
Find the element for a given (web page, document, element) triple and focus it,
if it's still available.

* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:

Tools:

* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/Tests/WebKitCocoa/RequestTextInputContext.mm: Added.
(-[WKWebView synchronouslyRequestTextInputContextsInRect:]):
(-[WKWebView synchronouslyFocusTextInputContext:]):
(applyStyle):
(applyIframe):
(TEST):
Add some tests for this SPI.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242696 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoiOS: Using ⌥ to scroll horizontally is no different than arrow key
timothy_horton@apple.com [Mon, 11 Mar 2019 02:54:55 +0000 (02:54 +0000)]
iOS: Using ⌥ to scroll horizontally is no different than arrow key
https://bugs.webkit.org/show_bug.cgi?id=195268
<rdar://problem/48326682>

Reviewed by Brent Fulgham.

* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView keyboardScrollViewAnimator:distanceForIncrement:inDirection:]):
(-[WKContentView keyboardScrollViewAnimator:distanceForIncrement:]): Deleted.
* UIProcess/ios/WKKeyboardScrollingAnimator.h:
* UIProcess/ios/WKKeyboardScrollingAnimator.mm:
(-[WKKeyboardScrollingAnimator keyboardScrollForEvent:]):
(-[WKKeyboardScrollViewAnimator setDelegate:]):
(-[WKKeyboardScrollViewAnimator distanceForIncrement:inDirection:]):
(-[WKKeyboardScrollViewAnimator distanceForIncrement:]): Deleted.
Add a direction parameter to -distanceForIncrement:, so the client
can return a different page/document size based on the scrolling axis.
Adopt it both in the default implementation and in WKContentViewInteraction.
Make the option key scroll by page when scrolling horizontally, like it
does when scrolling vertically.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242695 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[WTF] Align assumption in RunLoopWin to the other platform's RunLoop
Hironori.Fujii@sony.com [Mon, 11 Mar 2019 02:36:23 +0000 (02:36 +0000)]
[WTF] Align assumption in RunLoopWin to the other platform's RunLoop
https://bugs.webkit.org/show_bug.cgi?id=181151

Source/WTF:

Reviewed by Don Olmstead.

This patch fixes RunLoop in Windows to align it to the implementations in the other platforms
to use RunLoop more aggressively.

* wtf/RunLoop.h:
(WTF::RunLoop::Timer::Timer):
* wtf/win/MainThreadWin.cpp:
(initializeMainThreadPlatform): Call RunLoop::registerRunLoopMessageWindowClass.
* wtf/win/RunLoopWin.cpp:
(WTF::RunLoop::wndProc):
(WTF::RunLoop::iterate):
(WTF::RunLoop::stop):
PostQuitMessage is only available in the RunLoop's thread. We should post a message and call
it inside this task.

(WTF::RunLoop::registerRunLoopMessageWindowClass):
Changed the return type from bool to void, and added RELEASE_ASSERT to check the return value of RegisterClass.

(WTF::RunLoop::~RunLoop):
When the RunLoop's thread is freed, its associated window is freed. We do not need to do here.

(WTF::RunLoop::TimerBase::timerFired):
(WTF::RunLoop::TimerBase::TimerBase):
(WTF::RunLoop::TimerBase::start):
(WTF::RunLoop::TimerBase::stop):
(WTF::RunLoop::TimerBase::isActive const):
(WTF::RunLoop::TimerBase::secondsUntilFire const):
(WTF::generateTimerID): Deleted.
We can use TimerBase's pointer as ID since it is uintptr_t.

Tools:

Patch by Yusuke Suzuki <utatane.tea@gmail.com> on 2019-03-10
Reviewed by Don Olmstead.

* TestWebKitAPI/CMakeLists.txt:
* TestWebKitAPI/PlatformWin.cmake:
Enable TestWTF RunLoop tests in all platforms.

* TestWebKitAPI/Tests/WTF/RunLoop.cpp:
(TestWebKitAPI::DerivedOneShotTimer::DerivedOneShotTimer):
(TestWebKitAPI::DerivedOneShotTimer::fired):
(TestWebKitAPI::TEST):
Only a few platforms support nested RunLoop.

(TestWebKitAPI::DerivedRepeatingTimer::DerivedRepeatingTimer):
(TestWebKitAPI::DerivedRepeatingTimer::fired):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242694 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoASSERT(m_downloads.isEmpty()) fails in DownloadProxyMap::~DownloadProxyMap()
david_quesada@apple.com [Mon, 11 Mar 2019 00:27:22 +0000 (00:27 +0000)]
ASSERT(m_downloads.isEmpty()) fails in DownloadProxyMap::~DownloadProxyMap()
https://bugs.webkit.org/show_bug.cgi?id=152480

Reviewed by Chris Dumez.

Source/WebKit:

* UIProcess/Downloads/DownloadProxyMap.cpp:
(WebKit::DownloadProxyMap::downloadFinished):
    If the DownloadProxy is holding the last reference to the process pool, then
    invalidating the proxy will cause the process pool, the network process proxy,
    and this DownloadProxyMap to deallocate. Ensure that doesn't happen until this
    method has done everything it wants to do to clean up.

Tools:

Add a unit test based on Daniel Bates's test case that starts a download, ensures
there are no additional references to the process pool besides the one held by
the download, waits for the download to finish (in the sense that the
DownloadProxyMap is done tracking the DownloadProxy), and doesn't crash. For good
measure, also check that the process pool has been deallocated at the end of the
test. The test wouldn't be meaningful if the process pool were still alive.

* TestWebKitAPI/Tests/WebKitCocoa/Download.mm:
(-[WaitUntilDownloadCanceledDelegate _downloadDidStart:]):
(-[WaitUntilDownloadCanceledDelegate _downloadDidCancel:]):
    The download will be canceled because the delegate does not implement the
    method to decide the download's destination, so this is where we know the
    DownloadProxyMap is done with the DownloadProxy.
(TEST):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242693 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoFix some misleading function and variable names in WKContentViewInteraction.mm
wenson_hsieh@apple.com [Sun, 10 Mar 2019 23:49:10 +0000 (23:49 +0000)]
Fix some misleading function and variable names in WKContentViewInteraction.mm
https://bugs.webkit.org/show_bug.cgi?id=195536

Reviewed by Tim Horton.

* UIProcess/ios/WKContentViewInteraction.mm:
(shouldDeferZoomingToSelectionWhenRevealingFocusedElement):

Rename shouldZoomToRevealSelectionRect to shouldDeferZoomingToSelectionWhenRevealingFocusedElement; this
function is used to determine whether we should zoom to the selection rect when revealing the focused element,
and therefore affects whether we need to defer zooming until we get selection information.

(rectToRevealWhenZoomingToFocusedElement):
(-[WKContentView _elementDidFocus:userIsInteracting:blurPreviousNode:changingActivityState:userObject:]):

Rename shouldShowKeyboard to shouldShowInputView; this boolean indicates whether we should show not only the
keyboard, but other UI for user input such as date and select pickers.

(-[WKContentView _didReceiveEditorStateUpdateAfterFocus]):
(shouldZoomToRevealSelectionRect): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242690 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[ContentChangeObserver] Fix failing test cases
zalan@apple.com [Sun, 10 Mar 2019 20:08:54 +0000 (20:08 +0000)]
[ContentChangeObserver] Fix failing test cases
https://bugs.webkit.org/show_bug.cgi?id=195524
<rdar://problem/48745101>

Reviewed by Simon Fraser.

Source/WebCore:

1. Do not start DOM timer install observation when we already detected change at touchstart.
2. hasPendingActivity() should only care about ContentChangeObserver flags.
3. Do not try to notify the client when we are in the mouseMoved dispatch call (currently it could happen
when a timer gets intalled and removed right away).

* page/ios/ContentChangeObserver.cpp:
(WebCore::ContentChangeObserver::adjustObservedState):
(WebCore::ContentChangeObserver::isNotifyContentChangeAllowed const): Deleted.
* page/ios/ContentChangeObserver.h:
(WebCore::ContentChangeObserver::hasPendingActivity const):
(WebCore::ContentChangeObserver::isObservationTimeWindowActive const):

LayoutTests:

They've been failing ever since the 32ms fixed time window was introduced.

* fast/events/touch/ios/content-observation/click-instead-of-hover-simple.html:
* fast/events/touch/ios/content-observation/stuck-with-hover-state.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242689 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoMark two tests as failing after r242624
simon.fraser@apple.com [Sun, 10 Mar 2019 18:10:43 +0000 (18:10 +0000)]
Mark two tests as failing after r242624
https://bugs.webkit.org/show_bug.cgi?id=195531

Unreviewed test gardening.

These tests are failing:
accessibility/mac/selection-notification-focus-change.html [ Failure ]
compositing/video/video-clip-change-src.html [ ImageOnlyFailure ]

* platform/mac-wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242688 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoScrollingTree should have the final say on where layers go
simon.fraser@apple.com [Sun, 10 Mar 2019 18:03:42 +0000 (18:03 +0000)]
ScrollingTree should have the final say on where layers go
https://bugs.webkit.org/show_bug.cgi?id=195507

Reviewed by Antti Koivisto.

Source/WebCore:

Main thread layer flushing can race with scrolling tree layer changes on macOS, causing
flashing as layers jump around sometimes. We go to some lengths to avoid this by trying
not to touch properties on layers that are being interacted with (scrollableArea->setIsUserScroll in
updateScrollPositionAfterAsyncScroll()), but that's fragile.

This patch adds ScrollingTree::applyScrollingTreeLayerPositions(), which enters
ScrollingTree::applyLayerPositions() on the main thread/UI process. This traverses
the tree allowing each node to run their layer positioning logic.

For macOS WK2, this is called from TiledCoreAnimationDrawingArea::flushLayers() after flushCompositingStateIncludingSubframes().
For macOS WK2 with UI-side compositing, RemoteLayerTreeDrawingAreaProxy::commitLayerTree()
calls m_webPageProxy.scrollingCoordinatorProxy()->applyScrollingTreeLayerPositions().
iOS WK2 is unchanged, using viewportChangedViaDelegatedScrolling() which does the same thing, allowing
for the dynamic viewport changes that happen when zooming on iOS.

Testing this requires infrastructure that we don't have yet.

* page/scrolling/AsyncScrollingCoordinator.cpp:
(WebCore::AsyncScrollingCoordinator::applyScrollingTreeLayerPositions):
* page/scrolling/AsyncScrollingCoordinator.h:
* page/scrolling/ScrollingCoordinator.h:
(WebCore::ScrollingCoordinator::applyScrollingTreeLayerPositions):
* page/scrolling/ScrollingTree.cpp:
(WebCore::ScrollingTree::handleWheelEvent):
(WebCore::ScrollingTree::commitTreeState):
(WebCore::ScrollingTree::applyLayerPositions):
(WebCore::ScrollingTree::applyLayerPositionsRecursive):
* page/scrolling/ScrollingTree.h:
* page/scrolling/ScrollingTreeFrameHostingNode.cpp:
(WebCore::ScrollingTreeFrameHostingNode::applyLayerPositions):
* page/scrolling/ScrollingTreeFrameHostingNode.h:
* page/scrolling/ScrollingTreeNode.h:
* page/scrolling/ScrollingTreeScrollingNode.cpp:
(WebCore::ScrollingTreeScrollingNode::applyLayerPositions):
* page/scrolling/ScrollingTreeScrollingNode.h:
* page/scrolling/cocoa/ScrollingTreeFixedNode.h:
* page/scrolling/cocoa/ScrollingTreeFixedNode.mm:
(WebCore::ScrollingTreeFixedNode::relatedNodeScrollPositionDidChange):
* page/scrolling/cocoa/ScrollingTreeStickyNode.h:
* page/scrolling/cocoa/ScrollingTreeStickyNode.mm:
(WebCore::ScrollingTreeStickyNode::applyLayerPositions):
(WebCore::ScrollingTreeStickyNode::relatedNodeScrollPositionDidChange):
* page/scrolling/mac/ScrollingTreeFrameScrollingNodeMac.h:
* page/scrolling/mac/ScrollingTreeFrameScrollingNodeMac.mm:
(WebCore::ScrollingTreeFrameScrollingNodeMac::applyLayerPositions):

Source/WebKit:

* UIProcess/RemoteLayerTree/RemoteLayerTreeDrawingAreaProxy.mm:
(WebKit::RemoteLayerTreeDrawingAreaProxy::commitLayerTree):
* UIProcess/RemoteLayerTree/RemoteScrollingCoordinatorProxy.cpp:
(WebKit::RemoteScrollingCoordinatorProxy::applyScrollingTreeLayerPositions):
* UIProcess/RemoteLayerTree/RemoteScrollingCoordinatorProxy.h:
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.mm:
(WebKit::TiledCoreAnimationDrawingArea::flushLayers):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242687 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[Cocoa] Code signing fails because services are copied into XPCServices after the...
darin@apple.com [Sun, 10 Mar 2019 17:13:52 +0000 (17:13 +0000)]
[Cocoa] Code signing fails because services are copied into XPCServices after the framework is signed
https://bugs.webkit.org/show_bug.cgi?id=195523

Reviewed by Dan Bernstein.

* WebKit.xcodeproj/project.pbxproj: Make symbolic links while building the WebKit
framework instead of copying in the services after the framework is built.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242686 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[iOS Simulator] LayoutTest/scrollingcoordinator/ios/nested-fixed-layer-positions...
simon.fraser@apple.com [Sun, 10 Mar 2019 05:52:26 +0000 (05:52 +0000)]
[iOS Simulator] LayoutTest/scrollingcoordinator/ios/nested-fixed-layer-positions.html is flaky fail
https://bugs.webkit.org/show_bug.cgi?id=171628

Unreviewed test gardening.

I can't reproduce flakiness for this test. Let's see if it's fixed.

* platform/ios-wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242685 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[ios-simulator WK2] LayoutTest scrollingcoordinator/ios/non-stable-viewport-scroll...
simon.fraser@apple.com [Sun, 10 Mar 2019 05:52:24 +0000 (05:52 +0000)]
[ios-simulator WK2] LayoutTest scrollingcoordinator/ios/non-stable-viewport-scroll.html is a flaky failure
https://bugs.webkit.org/show_bug.cgi?id=168924

Unreviewed test gardening.

Let's try a new expectation for this test (it's non-flakey for me for 100 iterations).

* platform/ios-wk2/TestExpectations:
* scrollingcoordinator/ios/non-stable-viewport-scroll-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242684 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoAdd a test for scrolling tree adjustment of fixed layers (r242601)
simon.fraser@apple.com [Sun, 10 Mar 2019 05:52:21 +0000 (05:52 +0000)]
Add a test for scrolling tree adjustment of fixed layers (r242601)
https://bugs.webkit.org/show_bug.cgi?id=195521

Reviewed by Sam Weinig.

This test sets the "unstable" scrolling state (as if the user is actively scrolling) and scrolls,
then changes style to trigger a scrolling tree commit with a changed offset for a fixed layer.
The test shows the wrong fixed layer position before r242601.

Helpers are added to UIHelper to do the scroll, and to wrap setTimeout in an async function.
Sadly we have to wait about 120ms for the scrollbars to fade out.

* resources/ui-helper.js:
(window.UIHelper.async.delayFor):
(window.UIHelper.async.immediateScrollTo):
(window.UIHelper.async.immediateUnstableScrollTo):
* scrollingcoordinator/ios/ui-scroll-fixed-expected.html: Added.
* scrollingcoordinator/ios/ui-scroll-fixed.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242683 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoREGRESSION (r242551): Sporadic hangs when tapping to change selection on iOS
wenson_hsieh@apple.com [Sun, 10 Mar 2019 05:43:36 +0000 (05:43 +0000)]
REGRESSION (r242551): Sporadic hangs when tapping to change selection on iOS
https://bugs.webkit.org/show_bug.cgi?id=195475
<rdar://problem/48721153>

Reviewed by Chris Dumez.

Source/WebKit:

r242551 refactored synchronous autocorrection context requests to send an async IPC message and then use
waitForAndDispatchImmediately, instead of calling sendSync. However, this exposes a couple of existing issues in
the implementation of waitForAndDispatchImmediately that causes sporadic IPC deadlocks when changing selection.

First, passing in InterruptWaitingIfSyncMessageArrives when synchronously waiting for an IPC message currently
does not fulfill its intended behavior of interrupting waiting when a sync message arrives. This is because sync
IPC messages, by default, may be dispatched while the receiver is waiting for a sync reply. This means that the
logic in Connection::SyncMessageState::processIncomingMessage to dispatch an incoming sync message on the main
thread will attempt to handle the incoming message by enqueueing it on the main thread, and then waking up the
client runloop (i.e. signaling m_waitForSyncReplySemaphore). This works in the case of sendSync since the sync
reply semaphore is used to block the main thread, but in the case of waitForAndDispatchImmediately, a different
m_waitForMessageCondition is used instead, so SyncMessageState::processIncomingMessage will only enqueue the
incoming sync message on the main thread, and not actually invoke it.

To fix this first issue, observe that there is pre-existing logic to enqueue the incoming message and signal
m_waitForMessageCondition in Connection::processIncomingMessage. This codepath is currently not taken because we
end up bailing early in the call to SyncMessageState::processIncomingMessage. Instead, we can move this early
return further down in the function, such that if there is an incoming sync message and we're waiting with the
InterruptWaitingIfSyncMessageArrives option, we will correctly enqueue the incoming message, wake the runloop,
and proceed to handle the incoming message.

The second issue is more subtle; consider the scenario in which we send a sync message A from the web process to
the UI process, and simultaneously, in the UI process, we schedule some work to be done on the main thread.
Let's additionally suppose that this scheduled work will send an IPC message B to the web process and
synchronously wait for a reply (in the case of this particular bug, this is the sync autocorrection context
request). What happens upon receiving sync message A is that the IPC thread in the UI process will schedule A on
the main thread; however, before the scheduled response to A is invoked, we will first invoke previously
scheduled work that attempts to block synchronously until a message B is received. In summary:

1. (Web process)    sends sync IPC message A to UI process.
2. (UI process)     schedules some main runloop task that will block synchronously on IPC message B.
3. (UI process)     receives sync IPC message A and schedules A on the main runloop.
4. (UI process)     carry out the task scheduled in (2) and block on B.

...and then, the UI process and web process are now deadlocked because the UI process is waiting for B to
arrive, but the web process can't send B because it's waiting for a reply for IPC message A! To fix this second
deadlock, we first make an important observation: when using sendSync, we don't run into this problem because
immediately before sending sync IPC, we will attempt to handle any incoming sync IPC messages that have been
queued up. However, when calling waitForAndDispatchImmediately, we don't have this extra step, so a deadlock may
occur in the manner described above. To fix this, we make waitForAndDispatchImmediately behave more like
sendSync, by handling all incoming sync messages prior to blocking on an IPC response.

Test: editing/selection/ios/change-selection-by-tapping.html

* Platform/IPC/Connection.cpp:
(IPC::Connection::waitForMessage):
(IPC::Connection::processIncomingMessage):

LayoutTests:

Add a new layout test that taps to change selection 20 times in a contenteditable area and additionally
disables IPC timeout, to ensure that any IPC deadlocks will result in the test failing due to timing out.

* editing/selection/ios/change-selection-by-tapping-expected.txt: Added.
* editing/selection/ios/change-selection-by-tapping.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242682 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[Apple Pay] CanMakePaymentsWithActiveCard and OpenPaymentSetup should be async messages
aestes@apple.com [Sun, 10 Mar 2019 05:29:42 +0000 (05:29 +0000)]
[Apple Pay] CanMakePaymentsWithActiveCard and OpenPaymentSetup should be async messages
https://bugs.webkit.org/show_bug.cgi?id=195526
<rdar://problem/48745636>

Reviewed by Chris Dumez.

Source/WebCore:

* Modules/applepay/PaymentCoordinatorClient.h:
* loader/EmptyClients.cpp:
* testing/MockPaymentCoordinator.cpp:
(WebCore::MockPaymentCoordinator::canMakePaymentsWithActiveCard):
(WebCore::MockPaymentCoordinator::openPaymentSetup):
* testing/MockPaymentCoordinator.h:

Source/WebKit:

* Shared/ApplePay/WebPaymentCoordinatorProxy.cpp:
(WebKit::WebPaymentCoordinatorProxy::canMakePaymentsWithActiveCard):
(WebKit::WebPaymentCoordinatorProxy::openPaymentSetup):
* Shared/ApplePay/WebPaymentCoordinatorProxy.h:
* Shared/ApplePay/WebPaymentCoordinatorProxy.messages.in:
* WebProcess/ApplePay/WebPaymentCoordinator.cpp:
(WebKit::WebPaymentCoordinator::canMakePaymentsWithActiveCard):
(WebKit::WebPaymentCoordinator::openPaymentSetup):
(WebKit::generateCanMakePaymentsWithActiveCardReplyID): Deleted.
(WebKit::generateOpenPaymentSetupReplyID): Deleted.
(WebKit::WebPaymentCoordinator::canMakePaymentsWithActiveCardReply): Deleted.
(WebKit::WebPaymentCoordinator::openPaymentSetupReply): Deleted.
* WebProcess/ApplePay/WebPaymentCoordinator.h:
* WebProcess/ApplePay/WebPaymentCoordinator.messages.in:

Source/WebKitLegacy/mac:

* WebCoreSupport/WebPaymentCoordinatorClient.h:
* WebCoreSupport/WebPaymentCoordinatorClient.mm:
(WebPaymentCoordinatorClient::canMakePaymentsWithActiveCard):
(WebPaymentCoordinatorClient::openPaymentSetup):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242681 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[iOS] Remove unneeded entitlements and sandbox allowances from the Networking service
aestes@apple.com [Sun, 10 Mar 2019 05:28:48 +0000 (05:28 +0000)]
[iOS] Remove unneeded entitlements and sandbox allowances from the Networking service
https://bugs.webkit.org/show_bug.cgi?id=195527

Reviewed by Eric Carlson.

* Configurations/Network-iOS.entitlements:
* Resources/SandboxProfiles/ios/com.apple.WebKit.Networking.sb:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242680 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[ContentChangeObserver] Click event fires immediately on hover menu at seriouseats.com
zalan@apple.com [Sun, 10 Mar 2019 04:48:34 +0000 (04:48 +0000)]
[ContentChangeObserver] Click event fires immediately on hover menu at seriouseats.com
https://bugs.webkit.org/show_bug.cgi?id=195520
<rdar://problem/48740098>

Reviewed by Simon Fraser.

Source/WebCore:

Unfortunately seriouseats has a 300ms hover intent delay to deal with accidental menupane pop-ups. This page also hides this
non-fixed width menupane using absolute positioning and negative left.

Test: fast/events/touch/ios/content-observation/move-content-from-offscreen.html

* page/ios/ContentChangeObserver.cpp:
(WebCore::ContentChangeObserver::didInstallDOMTimer):
(WebCore::ContentChangeObserver::StyleChangeScope::~StyleChangeScope):
(WebCore::ContentChangeObserver::StyleChangeScope::isConsideredHidden const): Content auhtors tend to use x - 1 values (where x = 10^y)

LayoutTests:

* fast/events/touch/ios/content-observation/move-content-from-offscreen-expected.txt: Added.
* fast/events/touch/ios/content-observation/move-content-from-offscreen.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242679 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoUse modern async IPC with reply for device orientation permission
cdumez@apple.com [Sun, 10 Mar 2019 04:34:19 +0000 (04:34 +0000)]
Use modern async IPC with reply for device orientation permission
https://bugs.webkit.org/show_bug.cgi?id=195529

Reviewed by Ryosuke Niwa.

* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::requestDeviceOrientationAndMotionAccess):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::shouldAllowDeviceOrientationAndMotionAccess):
(WebKit::nextDeviceOrientationAndMotionPermissionCallbackID): Deleted.
(WebKit::WebPage::didReceiveDeviceOrientationAndMotionAccessDecision): Deleted.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242678 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoAdd assertions to help debug crash under DOMWindowExtension::suspendForPageCache()
cdumez@apple.com [Sun, 10 Mar 2019 03:26:10 +0000 (03:26 +0000)]
Add assertions to help debug crash under DOMWindowExtension::suspendForPageCache()
https://bugs.webkit.org/show_bug.cgi?id=195488

Reviewed by Ryosuke Niwa.

Try and figure out how the document can be detached from its frame while we're suspending
DOMWindowExtensions.

* page/DOMWindow.cpp:
(WebCore::DOMWindow::willDetachDocumentFromFrame):
(WebCore::DOMWindow::suspendForPageCache):
* page/DOMWindow.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242677 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoSimplify DOMWindowProperty code / handling
cdumez@apple.com [Sun, 10 Mar 2019 02:25:42 +0000 (02:25 +0000)]
Simplify DOMWindowProperty code / handling
https://bugs.webkit.org/show_bug.cgi?id=195495

Reviewed by Ryosuke Niwa.

DOMWindowProperty code was unnecessarily complex because DOMWindowExtension inherited
from it and DOMWindowExtension needs a lot of information about the global object's
lifetime to communicate to the injected bbundle client. This subclassing is also
very confusing because a DOMWindowExtension is not a *property* on the Window object.

This patch updates DOMWindowExtension to stop subclassing DOMWindowProperty and
moves all the complexity from DOMWindowProperty to DOMWindowExtension.
DOMWindowProperty is now a very simple base class which merely has a WeakPtr to
the window and getters for the window and the frame.

* Modules/indexeddb/DOMWindowIndexedDatabase.cpp:
(WebCore::DOMWindowIndexedDatabase::DOMWindowIndexedDatabase):
(WebCore::DOMWindowIndexedDatabase::indexedDB):
* Modules/indexeddb/DOMWindowIndexedDatabase.h:
There is no reason for DOMWindowIndexedDatabase to move its IDBFactory to a
separate data member which in PageCache. Script do not run while in PageCache.
Also, frames are nulled out while in the PageCache so the indexedDB() getter
would return null anyway while in PageCache.

* css/StyleMedia.idl:
* loader/appcache/ApplicationCacheHost.cpp:
(WebCore::ApplicationCacheHost::setDOMApplicationCache):
Store a WeakPtr to the DOMApplicationCache for safety.

(WebCore::ApplicationCacheHost::dispatchDOMEvent):
Do not fire events on the DOMApplicationCache if it is frameless to maintain
previous behavior. Previously, the DOMApplicationCache would have been nulled
out when detached from its frame so we would not have fired events.

* loader/appcache/ApplicationCacheHost.h:
* loader/appcache/DOMApplicationCache.cpp:
* loader/appcache/DOMApplicationCache.h:
Remove some unnecessary complexity. The ApplicationCacheHost is owned by the
DocumentLoader, which changes on navigation. There is therefore no reason to
null out the DOMApplicationCache on the ApplicationCacheHost when its gets
detached from its frame or enters PageCache.

* page/BarProp.idl:
* page/DOMSelection.idl:
* page/DOMWindow.cpp:
(WebCore::DOMWindow::willDestroyCachedFrame):
(WebCore::DOMWindow::willDestroyDocumentInFrame):
(WebCore::DOMWindow::willDetachDocumentFromFrame):
(WebCore::DOMWindow::registerExtension):
(WebCore::DOMWindow::unregisterExtension):

(WebCore::DOMWindow::resetDOMWindowProperties): Removed.
Stop clearing some of the DOMWindow's properties when the document gets destroyed or when
the Window for the initial empty document gets reused on navigation. I think we used to
need this because DOMWindowProperty used to hold pointers to their frame. However, this
is no longer the case nowadays as DOMWindowProperty objects get their frame from their
Window.

(WebCore::DOMWindow::resetUnlessSuspendedForDocumentSuspension):
(WebCore::DOMWindow::suspendForPageCache):
(WebCore::DOMWindow::resumeFromPageCache):
* page/DOMWindow.h:
* page/DOMWindowExtension.cpp:
(WebCore::DOMWindowExtension::DOMWindowExtension):
(WebCore::DOMWindowExtension::~DOMWindowExtension):
(WebCore::DOMWindowExtension::frame const):
(WebCore::DOMWindowExtension::suspendForPageCache):
(WebCore::DOMWindowExtension::resumeFromPageCache):
(WebCore::DOMWindowExtension::willDestroyGlobalObjectInCachedFrame):
(WebCore::DOMWindowExtension::willDestroyGlobalObjectInFrame):
(WebCore::DOMWindowExtension::willDetachGlobalObjectFromFrame):
* page/DOMWindowExtension.h:
* page/DOMWindowProperty.cpp:
(WebCore::DOMWindowProperty::DOMWindowProperty):
* page/DOMWindowProperty.h:
(WebCore::DOMWindowProperty::window const):
* page/History.idl:
* page/Location.idl:
* page/Navigator.cpp:
(WebCore::Navigator::plugins):
(WebCore::Navigator::mimeTypes):
* page/PerformanceNavigation.idl:
* page/PerformanceTiming.idl:
* page/Screen.idl:
* page/VisualViewport.cpp:
(WebCore::VisualViewport::scriptExecutionContext const):
* plugins/DOMMimeTypeArray.idl:
* plugins/DOMPluginArray.idl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242676 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[ContentChangeObserver] Start observing for content change between touchEnd and mouse...
zalan@apple.com [Sun, 10 Mar 2019 00:09:05 +0000 (00:09 +0000)]
[ContentChangeObserver] Start observing for content change between touchEnd and mouseMoved start
https://bugs.webkit.org/show_bug.cgi?id=195510
<rdar://problem/48735695>

Reviewed by Simon Fraser.

Source/WebCore:

This patch covers the observation of async changes triggered by touchStart/touchEnd (animations, timers, style recalcs).

Test: fast/events/touch/ios/content-observation/visibility-change-after-touch-end.html

* page/ios/ContentChangeObserver.cpp:
(WebCore::ContentChangeObserver::didCancelTouchEvent):
(WebCore::ContentChangeObserver::adjustObservedState):
* page/ios/ContentChangeObserver.h:
(WebCore::ContentChangeObserver::setIsInBetweenTouchEndAndMouseMoved):
(WebCore::ContentChangeObserver::isInBetweenTouchEndAndMouseMoved const):
(WebCore::ContentChangeObserver::isObservingContentChanges const):

Source/WebKit:

* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::cancelPotentialTapInFrame):

LayoutTests:

* fast/events/touch/ios/content-observation/visibility-change-after-touch-end-expected.txt: Added.
* fast/events/touch/ios/content-observation/visibility-change-after-touch-end.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242675 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoCompilation can be shrunk by 8 bytes
rmorisset@apple.com [Sat, 9 Mar 2019 17:29:35 +0000 (17:29 +0000)]
Compilation can be shrunk by 8 bytes
https://bugs.webkit.org/show_bug.cgi?id=195500

Reviewed by Mark Lam.

* profiler/ProfilerCompilation.cpp:
(JSC::Profiler::Compilation::Compilation):
* profiler/ProfilerCompilation.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242674 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoBinarySwitch can be shrunk by 8 bytes
rmorisset@apple.com [Sat, 9 Mar 2019 17:25:11 +0000 (17:25 +0000)]
BinarySwitch can be shrunk by 8 bytes
https://bugs.webkit.org/show_bug.cgi?id=195493

Reviewed by Mark Lam.

* jit/BinarySwitch.cpp:
(JSC::BinarySwitch::BinarySwitch):
* jit/BinarySwitch.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242673 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoAsyncStackTrace can be shrunk by 8 bytes
rmorisset@apple.com [Sat, 9 Mar 2019 17:24:16 +0000 (17:24 +0000)]
AsyncStackTrace can be shrunk by 8 bytes
https://bugs.webkit.org/show_bug.cgi?id=195491

Reviewed by Mark Lam.

* inspector/AsyncStackTrace.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242672 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoMake it clearer which data is protected by the two locks in ScrollingTree
simon.fraser@apple.com [Sat, 9 Mar 2019 02:35:57 +0000 (02:35 +0000)]
Make it clearer which data is protected by the two locks in ScrollingTree
https://bugs.webkit.org/show_bug.cgi?id=195501

Reviewed by Tim Horton.

Gather ScrollingTree member variables into two structs, and name the struct
members and the locks to make it clear which data is protected by each lock.

We only need to protect data read by multiple threads; these are the scrolling
thread, the event handling thread (which runs ThreadedScrollingTree::tryToHandleWheelEvent()),
and the main thread, which pokes various bits of pin/rubber-banding state.
Ideally the main thread would always push data to the scrolling thread via a commit,
but that's not what happens now.

Suspiciously, ScrollingTree::shouldHandleWheelEventSynchronously() uses the root node,
so should probably hold a lock shared with the scrolling thread (webkit.org/b/195502).

* page/scrolling/ScrollingTree.cpp:
(WebCore::ScrollingTree::shouldHandleWheelEventSynchronously):
(WebCore::ScrollingTree::commitTreeState):
(WebCore::ScrollingTree::setAsyncFrameOrOverflowScrollingEnabled):
(WebCore::ScrollingTree::setMainFrameScrollPosition):
(WebCore::ScrollingTree::eventTrackingTypeForPoint):
(WebCore::ScrollingTree::isRubberBandInProgress):
(WebCore::ScrollingTree::setMainFrameIsRubberBanding):
(WebCore::ScrollingTree::isScrollSnapInProgress):
(WebCore::ScrollingTree::setMainFrameIsScrollSnapping):
(WebCore::ScrollingTree::setMainFramePinState):
(WebCore::ScrollingTree::setCanRubberBandState):
(WebCore::ScrollingTree::setScrollPinningBehavior):
(WebCore::ScrollingTree::scrollPinningBehavior):
(WebCore::ScrollingTree::willWheelEventStartSwipeGesture):
(WebCore::ScrollingTree::latchedNode):
(WebCore::ScrollingTree::setLatchedNode):
(WebCore::ScrollingTree::clearLatchedNode):
(WebCore::ScrollingTree::scrollingTreeAsText):
(WebCore::ScrollingTree::touchActionDataAtPoint const):
(WebCore::ScrollingTree::mainFrameScrollPosition): Deleted.
(WebCore::ScrollingTree::mainFrameLayoutViewport): Deleted.
(WebCore::ScrollingTree::rubberBandsAtLeft): Deleted.
(WebCore::ScrollingTree::rubberBandsAtRight): Deleted.
(WebCore::ScrollingTree::rubberBandsAtBottom): Deleted.
(WebCore::ScrollingTree::rubberBandsAtTop): Deleted.
* page/scrolling/ScrollingTree.h:
(WebCore::ScrollingTree::hasLatchedNode const):
* page/scrolling/mac/ScrollingTreeScrollingNodeDelegateMac.mm:
(WebCore::ScrollingTreeScrollingNodeDelegateMac::stretchAmount):
* platform/graphics/FloatPoint.h:
(WebCore::FloatPoint::isZero const):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242670 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoShare some code that sets CALayer positions
simon.fraser@apple.com [Sat, 9 Mar 2019 02:35:54 +0000 (02:35 +0000)]
Share some code that sets CALayer positions
https://bugs.webkit.org/show_bug.cgi?id=195485

Reviewed by Zalan Bujtas.

Share some code between ScrollingTreeStickyNode and ScrollingTreeFixedNode that sets the position
of a CALayer given the top-left location.

* page/scrolling/cocoa/ScrollingTreeFixedNode.mm:
(WebCore::ScrollingTreeFixedNode::relatedNodeScrollPositionDidChange):
(WebCore::ScrollingTreeFixedNodeInternal::operator*): Deleted.
* page/scrolling/cocoa/ScrollingTreeStickyNode.mm:
(WebCore::ScrollingTreeStickyNode::relatedNodeScrollPositionDidChange):
(WebCore::ScrollingTreeStickyNodeInternal::operator*): Deleted.
* platform/graphics/cocoa/WebCoreCALayerExtras.h:
* platform/graphics/cocoa/WebCoreCALayerExtras.mm:
(-[CALayer _web_setLayerTopLeftPosition:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242669 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoStack overflow crash in JSC::JSObject::hasInstance.
mark.lam@apple.com [Sat, 9 Mar 2019 01:10:33 +0000 (01:10 +0000)]
Stack overflow crash in JSC::JSObject::hasInstance.
https://bugs.webkit.org/show_bug.cgi?id=195458
<rdar://problem/48710195>

Reviewed by Yusuke Suzuki.

JSTests:

* stress/stack-overflow-in-custom-hasInstance.js: Added.

Source/JavaScriptCore:

* runtime/JSObject.cpp:
(JSC::JSObject::hasInstance):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242667 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoEnable ProcessAssertions on iOS Simulator
cdumez@apple.com [Sat, 9 Mar 2019 00:59:40 +0000 (00:59 +0000)]
Enable ProcessAssertions on iOS Simulator
https://bugs.webkit.org/show_bug.cgi?id=195479

Reviewed by Alexey Proskuryakov.

Enable ProcessAssertions on iOS Simulator. We have some layout tests timing out in the
iOS Simulator due to the WebContent process suspending. Turning on ProcessAssertions in
the iOS Simulator seems to make those tests pass.

* UIProcess/ProcessAssertion.cpp:
* UIProcess/ProcessAssertion.h:
* UIProcess/ios/ProcessAssertionIOS.mm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242666 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[ iOS Simulator ] REGRESSION (r237087) Layout Test fast/viewport/ios/*-width-viewport...
ryanhaddad@apple.com [Sat, 9 Mar 2019 00:53:46 +0000 (00:53 +0000)]
[ iOS Simulator ] REGRESSION (r237087) Layout Test fast/viewport/ios/*-width-viewport-after-changing-view-scale.html is a flaky failure
https://bugs.webkit.org/show_bug.cgi?id=195341

Unreviewed test gardening.

* platform/ios-wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242665 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoAdd support for Device Orientation / Motion permission API
cdumez@apple.com [Sat, 9 Mar 2019 00:30:45 +0000 (00:30 +0000)]
Add support for Device Orientation / Motion permission API
https://bugs.webkit.org/show_bug.cgi?id=195329
<rdar://problem/47645367>

Reviewed by Geoffrey Garen.

Source/WebCore:

Add support for Device Orientation / Motion permission API:
- https://github.com/w3c/deviceorientation/issues/57

Pages can add event listeners for 'deviceorientation' / 'devicemotion' events but
such events will not be fired until the page's JavaScript calls
DeviceOrientationEvent.requestPermission() / DeviceMotionEvent.requestPermission()
and the user grants the request.

The feature is currently behind an experimental feature flag, off by default.

Tests: fast/device-orientation/device-motion-request-permission-denied.html
       fast/device-orientation/device-motion-request-permission-granted.html
       fast/device-orientation/device-motion-request-permission-user-gesture.html
       fast/device-orientation/device-orientation-request-permission-denied.html
       fast/device-orientation/device-orientation-request-permission-granted.html
       fast/device-orientation/device-orientation-request-permission-user-gesture.html

* CMakeLists.txt:
* DerivedSources-input.xcfilelist:
* DerivedSources-output.xcfilelist:
* DerivedSources.make:
* Sources.txt:
* WebCore.xcodeproj/project.pbxproj:
* dom/DeviceMotionEvent.h:
* dom/DeviceMotionEvent.idl:
* dom/DeviceOrientationAndMotionAccessController.cpp: Added.
(WebCore::DeviceOrientationAndMotionAccessController::DeviceOrientationAndMotionAccessController):
(WebCore::DeviceOrientationAndMotionAccessController::shouldAllowAccess):
(WebCore::DeviceOrientationAndMotionAccessController::setAccessState):
* dom/DeviceOrientationAndMotionAccessController.h: Added.
(WebCore::DeviceOrientationAndMotionAccessController::accessState const):
* dom/DeviceOrientationEvent.h:
* dom/DeviceOrientationEvent.idl:
* dom/DeviceOrientationOrMotionEvent.cpp: Added.
(WebCore::DeviceOrientationOrMotionEvent::requestPermission):
* dom/DeviceOrientationOrMotionEvent.h: Added.
* dom/DeviceOrientationOrMotionEvent.idl: Added.
* dom/DeviceOrientationOrMotionPermissionState.h: Added.
* dom/DeviceOrientationOrMotionPermissionState.idl: Added.
* dom/Document.cpp:
(WebCore::Document::deviceOrientationAndMotionAccessController):
* dom/Document.h:
* dom/Event.cpp:
* dom/MessagePort.cpp:
* dom/Microtasks.cpp:
* page/ChromeClient.h:
* page/DOMWindow.cpp:
(WebCore::DOMWindow::addEventListener):
(WebCore::DOMWindow::deviceOrientationController const):
(WebCore::DOMWindow::deviceMotionController const):
(WebCore::DOMWindow::isAllowedToUseDeviceMotionOrientation const):
(WebCore::DOMWindow::isAllowedToAddDeviceMotionOrientationListener const):
(WebCore::DOMWindow::startListeningForDeviceOrientationIfNecessary):
(WebCore::DOMWindow::stopListeningForDeviceOrientationIfNecessary):
(WebCore::DOMWindow::startListeningForDeviceMotionIfNecessary):
(WebCore::DOMWindow::stopListeningForDeviceMotionIfNecessary):
(WebCore::DOMWindow::removeEventListener):
(WebCore::DOMWindow::removeAllEventListeners):
* page/DOMWindow.h:
* page/DeviceController.cpp:
(WebCore::DeviceController::hasDeviceEventListener const):
* page/DeviceController.h:
* page/Settings.yaml:

Source/WebKit:

Add support for Device Orientation / Motion permission API:
- https://github.com/w3c/deviceorientation/issues/57

This adds new SPI to WKUIDelegatePrivate, until we can make this API.

* Shared/WebPreferences.yaml:
* UIProcess/API/APIUIClient.h:
(API::UIClient::shouldAllowDeviceOrientationAndMotionAccess):
* UIProcess/API/C/WKPage.cpp:
(WKPageSetPageUIClient):
* UIProcess/API/C/WKPageUIClient.h:
* UIProcess/API/Cocoa/WKUIDelegate.h:
* UIProcess/API/Cocoa/WKUIDelegatePrivate.h:
* UIProcess/Cocoa/UIDelegate.h:
* UIProcess/Cocoa/UIDelegate.mm:
(WebKit::UIDelegate::setDelegate):
(WebKit::UIDelegate::UIClient::shouldAllowDeviceOrientationAndMotionAccess):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::requestDeviceOrientationAndMotionAccess):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::shouldAllowDeviceOrientationAndMotionAccess):
* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::nextDeviceOrientationAndMotionPermissionCallbackID):
(WebKit::WebPage::shouldAllowDeviceOrientationAndMotionAccess):
(WebKit::WebPage::didReceiveDeviceOrientationAndMotionAccessDecision):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:

Tools:

Add test infrastructure to help test the Device Orientation / Motion permission API.

* WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
* WebKitTestRunner/InjectedBundle/TestRunner.cpp:
(WTR::TestRunner::setShouldAllowDeviceOrientationAndMotionAccess):
* WebKitTestRunner/InjectedBundle/TestRunner.h:
* WebKitTestRunner/TestController.cpp:
(WTR::shouldAllowDeviceOrientationAndMotionAccess):
(WTR::TestController::createWebViewWithOptions):
(WTR::TestController::resetStateToConsistentValues):
(WTR::TestController::handleDeviceOrientationAndMotionAccessRequest):
* WebKitTestRunner/TestController.h:
(WTR::TestController::setShouldAllowDeviceOrientationAndMotionAccess):
* WebKitTestRunner/TestInvocation.cpp:
(WTR::TestInvocation::didReceiveMessageFromInjectedBundle):

LayoutTests:

Add layout test coverage.

* TestExpectations:
* fast/device-orientation/device-motion-request-permission-denied-expected.txt: Added.
* fast/device-orientation/device-motion-request-permission-denied.html: Added.
* fast/device-orientation/device-motion-request-permission-granted-expected.txt: Added.
* fast/device-orientation/device-motion-request-permission-granted.html: Added.
* fast/device-orientation/device-motion-request-permission-user-gesture-expected.txt: Added.
* fast/device-orientation/device-motion-request-permission-user-gesture.html: Added.
* fast/device-orientation/device-orientation-request-permission-denied-expected.txt: Added.
* fast/device-orientation/device-orientation-request-permission-denied.html: Added.
* fast/device-orientation/device-orientation-request-permission-granted-expected.txt: Added.
* fast/device-orientation/device-orientation-request-permission-granted.html: Added.
* fast/device-orientation/device-orientation-request-permission-user-gesture-expected.txt: Added.
* fast/device-orientation/device-orientation-request-permission-user-gesture.html: Added.
* http/tests/events/device-orientation-motion-non-secure-context.html:
* http/tests/events/device-orientation-motion-secure-context-expected.txt:
* http/tests/events/device-orientation-motion-secure-context.html:
* platform/ios-wk2/TestExpectations:
* platform/ios/http/tests/events/device-orientation-motion-non-secure-context-expected.txt:
* platform/ios/http/tests/events/device-orientation-motion-secure-context-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242664 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoIntegerCheckCombiningPhase::Range can be shrunk by 8 bytes
rmorisset@apple.com [Sat, 9 Mar 2019 00:17:37 +0000 (00:17 +0000)]
IntegerCheckCombiningPhase::Range can be shrunk by 8 bytes
https://bugs.webkit.org/show_bug.cgi?id=195487

Reviewed by Saam Barati.

* dfg/DFGIntegerCheckCombiningPhase.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242663 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoHave the UIProcess take the UnboundedNetworking assertion when downloads are in progress.
beidson@apple.com [Fri, 8 Mar 2019 23:58:52 +0000 (23:58 +0000)]
Have the UIProcess take the UnboundedNetworking assertion when downloads are in progress.
https://bugs.webkit.org/show_bug.cgi?id=195468

Reviewed by Andy Estes.

* UIProcess/Downloads/DownloadProxyMap.cpp:
(WebKit::DownloadProxyMap::DownloadProxyMap):
(WebKit::DownloadProxyMap::createDownloadProxy): If this is the first download, and the process has the entitlement,
  take the assertion.
(WebKit::DownloadProxyMap::downloadFinished):
(WebKit::DownloadProxyMap::processDidClose):
* UIProcess/Downloads/DownloadProxyMap.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242662 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[ContentChangeObserver] Expand "isConsideredClickable" to descendants
zalan@apple.com [Fri, 8 Mar 2019 23:51:05 +0000 (23:51 +0000)]
[ContentChangeObserver] Expand "isConsideredClickable" to descendants
https://bugs.webkit.org/show_bug.cgi?id=195478
<rdar://problem/48724935>

Reviewed by Simon Fraser.

Source/WebCore:

In StyleChangeScope we try to figure out whether newly visible content should stick (menu panes etc) by checking if it is clickable.
This works fine as long as all the visible elements are gaining new renderers through this style update processs.
However when an element becomes visible by a change other than display: (not)none, it's not sufficient to just check the element itself,
since it might not respond to click at all, while its descendants do.
A concrete example is a max-height value change on usps.com, where the max-height is on a container (menu pane).
This container itself is not clickable while most of its children are (menu items).

Test: fast/events/touch/ios/content-observation/clickable-content-is-inside-a-container.html

* page/ios/ContentChangeObserver.cpp:
(WebCore::ContentChangeObserver::StyleChangeScope::StyleChangeScope):
(WebCore::ContentChangeObserver::StyleChangeScope::~StyleChangeScope):
(WebCore::ContentChangeObserver::StyleChangeScope::isConsideredHidden const):
(WebCore::ContentChangeObserver::StyleChangeScope::isConsideredClickable const):
(WebCore::isConsideredHidden): Deleted.
* page/ios/ContentChangeObserver.h:

LayoutTests:

* fast/events/touch/ios/content-observation/clickable-content-is-inside-a-container-expected.txt: Added.
* fast/events/touch/ios/content-observation/clickable-content-is-inside-a-container.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242661 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoTypeLocation can be shrunk by 8 bytes
rmorisset@apple.com [Fri, 8 Mar 2019 23:31:38 +0000 (23:31 +0000)]
TypeLocation can be shrunk by 8 bytes
https://bugs.webkit.org/show_bug.cgi?id=195483

Reviewed by Mark Lam.

* bytecode/TypeLocation.h:
(JSC::TypeLocation::TypeLocation):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242660 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoGetByIdStatus can be shrunk by 16 bytes
rmorisset@apple.com [Fri, 8 Mar 2019 23:24:25 +0000 (23:24 +0000)]
GetByIdStatus can be shrunk by 16 bytes
https://bugs.webkit.org/show_bug.cgi?id=195480

Reviewed by Saam Barati.

8 bytes from reordering fields
8 more bytes by making the enum State only use 1 byte.

* bytecode/GetByIdStatus.cpp:
(JSC::GetByIdStatus::GetByIdStatus):
* bytecode/GetByIdStatus.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242659 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoAdd a compare-results script to compare benchmark results
sbarati@apple.com [Fri, 8 Mar 2019 23:20:15 +0000 (23:20 +0000)]
Add a compare-results script to compare benchmark results
https://bugs.webkit.org/show_bug.cgi?id=195486
<rdar://problem/48723397>

Reviewed by Geoffrey Garen.

This patch adds a script to compare benchmark results using Welch's two-tailed t test.
Initially, this patch only reasons about PLT5/JetStream2/Speedometer2. It will be easy
to extend it to learn about our other benchmarks.

* Scripts/compare-results: Added.
(readJSONFile):
(detectJetStream2):
(JetStream2Results):
(detectSpeedometer2):
(Speedometer2Results):
(detectPLT5):
(PLT5Results):
(detectBenchmark):
(biggerIsBetter):
(ttest):
(getOptions):
(main):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242658 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoPutByIdVariant can be shrunk by 8 bytes
rmorisset@apple.com [Fri, 8 Mar 2019 23:19:27 +0000 (23:19 +0000)]
PutByIdVariant can be shrunk by 8 bytes
https://bugs.webkit.org/show_bug.cgi?id=195482

Reviewed by Mark Lam.

* bytecode/PutByIdVariant.h:
(JSC::PutByIdVariant::PutByIdVariant):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242657 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[ContentChangeObserver] Cleanup adjustObservedState
zalan@apple.com [Fri, 8 Mar 2019 22:12:46 +0000 (22:12 +0000)]
[ContentChangeObserver] Cleanup adjustObservedState
https://bugs.webkit.org/show_bug.cgi?id=195470
<rdar://problem/48717823>

Reviewed by Simon Fraser.

This is in preparation for introducing an observation window from touchStart -> mouseMoved.
1. Cancel pending activities (future timers, pending stylesheet recalcs) when visible content change is detected.
2. The fixed time window takes care of notifying the client -timers, style recalcs during the window should not signal themselves.
3. Reset m_isObservingPendingStyleRecalc at StartedStyleRecalc instead of EndedStyleRecalc.

* page/ios/ContentChangeObserver.cpp:
(WebCore::ContentChangeObserver::domTimerExecuteDidFinish):
(WebCore::ContentChangeObserver::styleRecalcDidStart):
(WebCore::ContentChangeObserver::styleRecalcDidFinish):
(WebCore::ContentChangeObserver::adjustObservedState):
* page/ios/ContentChangeObserver.h:
(WebCore::ContentChangeObserver::hasPendingActivity const):
(WebCore::ContentChangeObserver::isObservationTimeWindowActive const):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242656 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoUnreviewed, follow-up after r242568
ysuzuki@apple.com [Fri, 8 Mar 2019 21:59:34 +0000 (21:59 +0000)]
Unreviewed, follow-up after r242568

Robin pointed that calculation of `numberOfChildren` and `nonEmptyIndex` is unnecessary.

* dfg/DFGAbstractInterpreterInlines.h:
(JSC::DFG::AbstractInterpreter<AbstractStateType>::executeEffects):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242655 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[ContentChangeObserver] Add StartedDOMTimerExecution and StartedStyleRecalc
zalan@apple.com [Fri, 8 Mar 2019 21:11:41 +0000 (21:11 +0000)]
[ContentChangeObserver] Add StartedDOMTimerExecution and StartedStyleRecalc
https://bugs.webkit.org/show_bug.cgi?id=195463
<rdar://problem/48714762>

Reviewed by Simon Fraser.

This is in preparation for introducing m_isObservingContentChanges flag to track observing state across events (touchStart -> mouseMoved).

* page/ios/ContentChangeObserver.cpp:
(WebCore::ContentChangeObserver::domTimerExecuteDidStart):
(WebCore::ContentChangeObserver::styleRecalcDidStart):
(WebCore::ContentChangeObserver::styleRecalcDidFinish):
(WebCore::ContentChangeObserver::setShouldObserveNextStyleRecalc):
(WebCore::ContentChangeObserver::adjustObservedState):
* page/ios/ContentChangeObserver.h:
(WebCore::ContentChangeObserver::isObservingPendingStyleRecalc const):
(WebCore::ContentChangeObserver::isObservingStyleRecalc const): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242654 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoEnsure old tab state is cleared between iterations of run-benchmark
slewis@apple.com [Fri, 8 Mar 2019 20:52:56 +0000 (20:52 +0000)]
Ensure old tab state is cleared between iterations of run-benchmark
https://bugs.webkit.org/show_bug.cgi?id=195393
<rdar://problem/46885583>

Reviewed by Dewei Zhu.

a) ensure the default to restore state is not set
b) terminate Safari correctly
c) Set the system default to ignore Persistent State in Safari

* Scripts/webkitpy/benchmark_runner/browser_driver/osx_browser_driver.py:
(OSXBrowserDriver): fix terminate processes to call terminate before killing.
(OSXBrowserDriver.prepare_env):
(OSXBrowserDriver.restore_env):
(OSXBrowserDriver.close_browsers):
(OSXBrowserDriver._terminate_processes):
* Scripts/webkitpy/benchmark_runner/browser_driver/osx_chrome_driver.py: add bundle id
(OSXChromeDriver):
(OSXChromeCanaryDriver):
* Scripts/webkitpy/benchmark_runner/browser_driver/osx_firefox_driver.py: ditto
(OSXFirefoxDriver):
(OSXFirefoxNightlyDriver):
* Scripts/webkitpy/benchmark_runner/browser_driver/osx_safari_driver.py: add preferences to not restore state, bundle id
(OSXSafariDriver):
(OSXSafariDriver.prepare_env):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242653 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoAdd assertions to help debug a WebProcessCache crash
cdumez@apple.com [Fri, 8 Mar 2019 19:54:02 +0000 (19:54 +0000)]
Add assertions to help debug a WebProcessCache crash
https://bugs.webkit.org/show_bug.cgi?id=195469

Reviewed by Brady Eidson.

I suspect the process's registrableDomain in null when evictProcess() gets
called, thus crashing when lookup it up in the HashMap. Confirm this and
how this could happen via assertions.

* UIProcess/WebProcessCache.cpp:
(WebKit::WebProcessCache::evictProcess):
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::didStartProvisionalLoadForMainFrame):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242652 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoAX: AOM accessibleclick does not work on iOS
cfleizach@apple.com [Fri, 8 Mar 2019 19:51:07 +0000 (19:51 +0000)]
AX: AOM accessibleclick does not work on iOS
https://bugs.webkit.org/show_bug.cgi?id=195423
<rdar://problem/48682110>

Reviewed by Joanmarie Diggs.

Return this value of this method so it can be surfaced to a higher level.

* accessibility/ios/WebAccessibilityObjectWrapperIOS.mm:
(-[WebAccessibilityObjectWrapper _accessibilityActivate]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242651 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[JSC] We should have more WithoutTransition functions which are usable for JSGlobalOb...
ysuzuki@apple.com [Fri, 8 Mar 2019 19:33:51 +0000 (19:33 +0000)]
[JSC] We should have more WithoutTransition functions which are usable for JSGlobalObject initialization
https://bugs.webkit.org/show_bug.cgi?id=195447

Reviewed by Filip Pizlo.

This patch reduces # of unnecessary structure transitions in JSGlobalObject initialization to avoid unnecessary allocations
caused by Structure transition. One example is WeakBlock allocation for StructureTransitionTable.
To achieve this, we (1) add putDirectNonIndexAccessorWithoutTransition and putDirectNativeIntrinsicGetterWithoutTransition
to add accessor properties without transition, and (2) add NameAdditionMode::WithoutStructureTransition mode to InternalFunction::finishCreation
to use `putDirectWithoutTransition` instead of `putDirect`.

* inspector/JSInjectedScriptHostPrototype.cpp:
(Inspector::JSInjectedScriptHostPrototype::finishCreation):
* inspector/JSJavaScriptCallFramePrototype.cpp:
(Inspector::JSJavaScriptCallFramePrototype::finishCreation):
* runtime/ArrayConstructor.cpp:
(JSC::ArrayConstructor::finishCreation):
* runtime/AsyncFunctionConstructor.cpp:
(JSC::AsyncFunctionConstructor::finishCreation):
* runtime/AsyncGeneratorFunctionConstructor.cpp:
(JSC::AsyncGeneratorFunctionConstructor::finishCreation):
* runtime/BigIntConstructor.cpp:
(JSC::BigIntConstructor::finishCreation):
* runtime/BooleanConstructor.cpp:
(JSC::BooleanConstructor::finishCreation):
* runtime/DateConstructor.cpp:
(JSC::DateConstructor::finishCreation):
* runtime/ErrorConstructor.cpp:
(JSC::ErrorConstructor::finishCreation):
* runtime/FunctionConstructor.cpp:
(JSC::FunctionConstructor::finishCreation):
* runtime/FunctionPrototype.cpp:
(JSC::FunctionPrototype::finishCreation):
(JSC::FunctionPrototype::addFunctionProperties):
(JSC::FunctionPrototype::initRestrictedProperties):
* runtime/FunctionPrototype.h:
* runtime/GeneratorFunctionConstructor.cpp:
(JSC::GeneratorFunctionConstructor::finishCreation):
* runtime/InternalFunction.cpp:
(JSC::InternalFunction::finishCreation):
* runtime/InternalFunction.h:
* runtime/IntlCollatorConstructor.cpp:
(JSC::IntlCollatorConstructor::finishCreation):
* runtime/IntlDateTimeFormatConstructor.cpp:
(JSC::IntlDateTimeFormatConstructor::finishCreation):
* runtime/IntlNumberFormatConstructor.cpp:
(JSC::IntlNumberFormatConstructor::finishCreation):
* runtime/IntlPluralRulesConstructor.cpp:
(JSC::IntlPluralRulesConstructor::finishCreation):
* runtime/JSArrayBufferConstructor.cpp:
(JSC::JSGenericArrayBufferConstructor<sharingMode>::finishCreation):
* runtime/JSArrayBufferPrototype.cpp:
(JSC::JSArrayBufferPrototype::finishCreation):
* runtime/JSGenericTypedArrayViewConstructorInlines.h:
(JSC::JSGenericTypedArrayViewConstructor<ViewClass>::finishCreation):
* runtime/JSGlobalObject.cpp:
(JSC::JSGlobalObject::init):
* runtime/JSObject.cpp:
(JSC::JSObject::putDirectNonIndexAccessorWithoutTransition):
(JSC::JSObject::putDirectNativeIntrinsicGetterWithoutTransition):
* runtime/JSObject.h:
* runtime/JSPromiseConstructor.cpp:
(JSC::JSPromiseConstructor::finishCreation):
* runtime/JSTypedArrayViewConstructor.cpp:
(JSC::JSTypedArrayViewConstructor::finishCreation):
* runtime/JSTypedArrayViewPrototype.cpp:
(JSC::JSTypedArrayViewPrototype::finishCreation):
* runtime/MapConstructor.cpp:
(JSC::MapConstructor::finishCreation):
* runtime/MapPrototype.cpp:
(JSC::MapPrototype::finishCreation):
* runtime/NativeErrorConstructor.cpp:
(JSC::NativeErrorConstructorBase::finishCreation):
* runtime/NullGetterFunction.h:
* runtime/NullSetterFunction.h:
* runtime/NumberConstructor.cpp:
(JSC::NumberConstructor::finishCreation):
* runtime/ObjectConstructor.cpp:
(JSC::ObjectConstructor::finishCreation):
* runtime/ProxyConstructor.cpp:
(JSC::ProxyConstructor::finishCreation):
* runtime/RegExpConstructor.cpp:
(JSC::RegExpConstructor::finishCreation):
* runtime/RegExpPrototype.cpp:
(JSC::RegExpPrototype::finishCreation):
* runtime/SetConstructor.cpp:
(JSC::SetConstructor::finishCreation):
* runtime/SetPrototype.cpp:
(JSC::SetPrototype::finishCreation):
* runtime/StringConstructor.cpp:
(JSC::StringConstructor::finishCreation):
* runtime/SymbolConstructor.cpp:
(JSC::SymbolConstructor::finishCreation):
* runtime/WeakMapConstructor.cpp:
(JSC::WeakMapConstructor::finishCreation):
* runtime/WeakSetConstructor.cpp:
(JSC::WeakSetConstructor::finishCreation):
* wasm/js/WebAssemblyCompileErrorConstructor.cpp:
(JSC::WebAssemblyCompileErrorConstructor::finishCreation):
* wasm/js/WebAssemblyInstanceConstructor.cpp:
(JSC::WebAssemblyInstanceConstructor::finishCreation):
* wasm/js/WebAssemblyLinkErrorConstructor.cpp:
(JSC::WebAssemblyLinkErrorConstructor::finishCreation):
* wasm/js/WebAssemblyMemoryConstructor.cpp:
(JSC::WebAssemblyMemoryConstructor::finishCreation):
* wasm/js/WebAssemblyModuleConstructor.cpp:
(JSC::WebAssemblyModuleConstructor::finishCreation):
* wasm/js/WebAssemblyRuntimeErrorConstructor.cpp:
(JSC::WebAssemblyRuntimeErrorConstructor::finishCreation):
* wasm/js/WebAssemblyTableConstructor.cpp:
(JSC::WebAssemblyTableConstructor::finishCreation):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242650 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoop_check_tdz does not def its argument
tzagallo@apple.com [Fri, 8 Mar 2019 19:18:18 +0000 (19:18 +0000)]
op_check_tdz does not def its argument
https://bugs.webkit.org/show_bug.cgi?id=192880
<rdar://problem/46221598>

Reviewed by Saam Barati.

JSTests:

* microbenchmarks/let-for-in.js: Added.
(foo):

Source/JavaScriptCore:

This prevented the for-in loop optimization in the bytecode generator, since
the analysis sees a redefinition of the loop variable.

* bytecode/BytecodeUseDef.h:
(JSC::computeDefsForBytecodeOffset):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242649 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoRename AssertionState::Download.
beidson@apple.com [Fri, 8 Mar 2019 19:01:31 +0000 (19:01 +0000)]
Rename AssertionState::Download.
https://bugs.webkit.org/show_bug.cgi?id=195465

Reviewed by Andy Estes.

It's (currently) about uploads and downloads.
Let's call it "UnboundedNetworking"

* NetworkProcess/Downloads/DownloadMap.cpp:
(WebKit::DownloadMap::add):

* UIProcess/ProcessAssertion.h:

* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::didSetAssertionState):

* UIProcess/ios/ProcessAssertionIOS.mm:
(WebKit::flagsForState):
(WebKit::reasonForState):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242646 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[macOS UI-side compositing] Mouse handling can trigger a crash before we have a scrol...
simon.fraser@apple.com [Fri, 8 Mar 2019 18:34:23 +0000 (18:34 +0000)]
[macOS UI-side compositing] Mouse handling can trigger a crash before we have a scrolling tree root
https://bugs.webkit.org/show_bug.cgi?id=195467

Reviewed by Antti Koivisto.

When launching MiniBrowser with UI-side compositing enabled in a state where the window
appears under the mouse, we can hit RemoteScrollingTree::handleMouseEvent() for a mouseEnter
event before we have a scrolling tree root node, so add a null check.

* UIProcess/RemoteLayerTree/RemoteScrollingTree.cpp:
(WebKit::RemoteScrollingTree::handleMouseEvent):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242645 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago(r242595) Layout Tests in imported/w3c/web-platform-tests/html/semantics/embedded...
tsavell@apple.com [Fri, 8 Mar 2019 18:24:10 +0000 (18:24 +0000)]
(r242595) Layout Tests in imported/w3c/web-platform-tests/html/semantics/embedded-content/media-elements/* are failing
https://bugs.webkit.org/show_bug.cgi?id=195466

Unreviewed test gardening.

* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242644 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoREGRESSION(r242624): [GTK] New rAF code path assumes AC mode
commit-queue@webkit.org [Fri, 8 Mar 2019 18:13:29 +0000 (18:13 +0000)]
REGRESSION(r242624): [GTK] New rAF code path assumes AC mode
https://bugs.webkit.org/show_bug.cgi?id=195459

Patch by Carlos Garcia Campos <cgarcia@igalia.com> on 2019-03-08
Reviewed by Žan Doberšek.

* WebProcess/WebPage/CoordinatedGraphics/DrawingAreaCoordinatedGraphics.cpp:
(WebKit::DrawingAreaCoordinatedGraphics::scheduleCompositingLayerFlush):
The refresh monitor now calls scheduleLayerFlush() but when AC
mode is disabled this method does nothing, so setNeedsDisplay()
needs to be called instead.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242643 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoLayout Test http/tests/referrer-policy-iframe/no-referrer/cross-origin-http-http...
pvollan@apple.com [Fri, 8 Mar 2019 17:32:02 +0000 (17:32 +0000)]
Layout Test http/tests/referrer-policy-iframe/no-referrer/cross-origin-http-http.html is failing
https://bugs.webkit.org/show_bug.cgi?id=195461

Unreviewed test gardening.

* platform/win/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242642 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoimported/w3c/web-platform-tests/FileAPI/reading-data-section/filereader_error.html...
cdumez@apple.com [Fri, 8 Mar 2019 17:02:16 +0000 (17:02 +0000)]
imported/w3c/web-platform-tests/FileAPI/reading-data-section/filereader_error.html is a flaky crash
https://bugs.webkit.org/show_bug.cgi?id=195441
<rdar://problem/43437394>

Reviewed by Alexey Proskuryakov.

FileReader is an ActiveDOMObject, which means that FileReader::stop() gets called when
its script execution context is about to get destroyed. FileReader::stop() sets m_state
to DONE. FileReader::abort() would schedule an asynchronous task and then ASSERT that
m_state is not DONE, which would hit if FileReader::stop() had been called in between
the task being scheduled and its execution. To address the issue, have the task abort
early if isContextStopped() returns true.

Also replace calls to setPendingActivity() / unsetPendingActivity() with a
PendingActivity data member as mismatched call to those can lead to leaks.

* fileapi/FileReader.cpp:
(WebCore::FileReader::canSuspendForDocumentSuspension const):
No reason not to suspend if there is no pending read.

(WebCore::FileReader::stop):
(WebCore::FileReader::readInternal):
(WebCore::FileReader::abort):
(WebCore::FileReader::didFinishLoading):
(WebCore::FileReader::didFail):
* fileapi/FileReader.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242641 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoGLContextEGL: desired EGL config should search for 8-bit components by default
zandobersek@gmail.com [Fri, 8 Mar 2019 11:57:54 +0000 (11:57 +0000)]
GLContextEGL: desired EGL config should search for 8-bit components by default
https://bugs.webkit.org/show_bug.cgi?id=195413

Reviewed by Carlos Garcia Campos.

The EGL config search in GLContextEGL should by default look for
RGBA8888 configurations while allowing RGB565 as an alternative.
This prevents from accidentally landing on an RGBA1010102
configuration that is available with some graphics stacks, and which is
not expected in e.g. window snapshotting that's done for layout test
output comparison.

* platform/graphics/egl/GLContextEGL.cpp:
(WebCore::GLContextEGL::getEGLConfig): EGL config search should by
default request 8-bit color channels.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242640 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoUse a thread safe refcounter for FilterOperation.
magomez@igalia.com [Fri, 8 Mar 2019 11:26:18 +0000 (11:26 +0000)]
Use a thread safe refcounter for FilterOperation.
https://bugs.webkit.org/show_bug.cgi?id=194149

Reviewed by Carlos Garcia Campos.

Use a thread safe refcounter for FilterOperation.

* platform/graphics/filters/FilterOperation.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242639 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[GTK] Several InputMethodFilter tests are failing
dpino@igalia.com [Fri, 8 Mar 2019 10:18:28 +0000 (10:18 +0000)]
[GTK] Several InputMethodFilter tests are failing
https://bugs.webkit.org/show_bug.cgi?id=195408

Reviewed by Carlos Garcia Campos.

* TestWebKitAPI/Tests/WebKit/gtk/InputMethodFilter.cpp:
Regressed in r241751.
Hexadecimal values of keycodes are expressed now in upper letters.
Composited characters should be cast to UTF8 strings.

(TestWebKitAPI::TEST):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242638 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[GTK] Make Tools/gtkdoc python3 compatible
tpopela@redhat.com [Fri, 8 Mar 2019 07:11:39 +0000 (07:11 +0000)]
[GTK] Make Tools/gtkdoc python3 compatible
https://bugs.webkit.org/show_bug.cgi?id=195359

Reviewed by Carlos Garcia Campos.

* gtkdoc/generate-gtkdoc:
ConfigParser was reworked in Python 3.2 so we have adapt the code to
work with Python 2 and 3.
(get_gtkdoc_module_paths):
The iteritems() was removed in Python 3, so let's use items() that's
available in Python 2 and 3.
(get_generator_for_config):
* gtkdoc/gtkdoc.py:
(GTKDoc._run_command):
The sys.stdout.write() is expecting str in Python 3 and not bytes
(that are coming from stdout.encode()). Use sys.stdout.buffer.write()
for passing the bytes there.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242637 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[JSC] Make more fields lazy in JSGlobalObject
ysuzuki@apple.com [Fri, 8 Mar 2019 06:17:32 +0000 (06:17 +0000)]
[JSC] Make more fields lazy in JSGlobalObject
https://bugs.webkit.org/show_bug.cgi?id=195449

Reviewed by Mark Lam.

Source/JavaScriptCore:

This patch makes more fields lazy-allocated in JSGlobalObject to save memory.

1. Some minor structures like moduleRecordStructure.
2. Some functions like parseInt / parseFloat. While they are eagerly created in JIT mode anyway to materialize NumberConstructor, we can lazily allocate them in non JIT mode.
3. ArrayBuffer constructor. While it is eagerly allocated in WebCore, we can make lazily allocated in JSC.

* interpreter/Interpreter.cpp:
(JSC::Interpreter::execute):
* runtime/JSArrayBufferPrototype.h:
* runtime/JSGlobalObject.cpp:
(JSC::JSGlobalObject::init):
(JSC::JSGlobalObject::visitChildren):
* runtime/JSGlobalObject.h:
(JSC::JSGlobalObject::parseIntFunction const):
(JSC::JSGlobalObject::parseFloatFunction const):
(JSC::JSGlobalObject::evalFunction const):
(JSC::JSGlobalObject::strictEvalActivationStructure const):
(JSC::JSGlobalObject::moduleRecordStructure const):
(JSC::JSGlobalObject::moduleNamespaceObjectStructure const):
(JSC::JSGlobalObject::proxyObjectStructure const):
(JSC::JSGlobalObject::callableProxyObjectStructure const):
(JSC::JSGlobalObject::proxyRevokeStructure const):
(JSC::JSGlobalObject::arrayBufferConstructor const):
(JSC::JSGlobalObject::arrayBufferPrototype const):
(JSC::JSGlobalObject::arrayBufferStructure const):
* runtime/ProxyObject.h:
* runtime/StrictEvalActivation.cpp:
(JSC::StrictEvalActivation::StrictEvalActivation):
* runtime/StrictEvalActivation.h:
* wasm/js/JSWebAssemblyMemory.cpp:
(JSC::JSWebAssemblyMemory::buffer):
* wasm/js/WebAssemblyModuleConstructor.cpp:
(JSC::webAssemblyModuleCustomSections):

Source/WebCore:

Use arrayBufferConstructor() since getDirect does not work with lazy property.

* bindings/js/JSDOMGlobalObject.cpp:
(WebCore::JSDOMGlobalObject::addBuiltinGlobals):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242636 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoSource/WebCore:
zalan@apple.com [Fri, 8 Mar 2019 06:11:40 +0000 (06:11 +0000)]
Source/WebCore:
[ContentChangeObserver][REGRESSION] Check if visibility change happens while dispatching mouseMoved
https://bugs.webkit.org/show_bug.cgi?id=195421
<rdar://problem/48682004>

Reviewed by Simon Fraser.

Visibility change might be triggered synchronously while dispatching mouseMoved event.

Test: fast/events/touch/ios/content-observation/visibility-change-happens-while-in-mousemoved.html

* page/ios/ContentChangeObserver.cpp:
(WebCore::ContentChangeObserver::mouseMovedDidStart):
(WebCore::ContentChangeObserver::mouseMovedDidFinish):
(WebCore::ContentChangeObserver::isNotifyContentChangeAllowed const):
* page/ios/ContentChangeObserver.h:
(WebCore::ContentChangeObserver::isObservingContentChanges const):

LayoutTests:
[ContentChangeObserver] Check if visibility change happens while dispatching mouseMoved
https://bugs.webkit.org/show_bug.cgi?id=195421
<rdar://problem/48682004>

Reviewed by Simon Fraser.

* fast/events/touch/ios/content-observation/visibility-change-happens-while-in-mousemoved-expected.txt: Added.
* fast/events/touch/ios/content-observation/visibility-change-happens-while-in-mousemoved.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242635 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[ContentChangeObserver] Check if max-height change triggers visible content change.
zalan@apple.com [Fri, 8 Mar 2019 06:08:59 +0000 (06:08 +0000)]
[ContentChangeObserver] Check if max-height change triggers visible content change.
https://bugs.webkit.org/show_bug.cgi?id=195417
<rdar://problem/48680631>

Reviewed by Simon Fraser.

Source/WebCore:

A fixed max-height non-zero value could indicate visible content change. usps.com uses this technique to show menu panes.

Test: fast/events/touch/ios/content-observation/visibility-change-is-max-height-change.html

* page/ios/ContentChangeObserver.cpp:
(WebCore::isConsideredHidden):
(WebCore::ContentChangeObserver::StyleChangeScope::StyleChangeScope):
(WebCore::ContentChangeObserver::StyleChangeScope::~StyleChangeScope):
(WebCore::elementImplicitVisibility): Deleted.
* page/ios/ContentChangeObserver.h:

LayoutTests:

* fast/events/touch/ios/content-observation/visibility-change-is-max-height-change-expected.txt: Added.
* fast/events/touch/ios/content-observation/visibility-change-is-max-height-change.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242634 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoUnreviewed, fix failing EWS build for ios-sim
ysuzuki@apple.com [Fri, 8 Mar 2019 04:51:12 +0000 (04:51 +0000)]
Unreviewed, fix failing EWS build for ios-sim

* page/ios/ContentChangeObserver.cpp:
(WebCore::ContentChangeObserver::touchEventDidStart):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242633 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[Win] Remove -DUCHAR_TYPE=wchar_t stopgap and learn to live with char16_t.
Hironori.Fujii@sony.com [Fri, 8 Mar 2019 04:39:45 +0000 (04:39 +0000)]
[Win] Remove -DUCHAR_TYPE=wchar_t stopgap and learn to live with char16_t.
https://bugs.webkit.org/show_bug.cgi?id=195346
<rdar://problem/48667048>

Unreviewed build fix for clang-cl.

* DumpRenderTree/win/DumpRenderTree.cpp:
(setApplicationId): Replaced calling String::charactersWithNullTermination with String::wideCharacters to get wchar_t.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242632 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[MSE] Adopt new AVSampleBufferDisplayLayer SPI
eric.carlson@apple.com [Fri, 8 Mar 2019 04:28:48 +0000 (04:28 +0000)]
[MSE] Adopt new AVSampleBufferDisplayLayer SPI
https://bugs.webkit.org/show_bug.cgi?id=195445
<rdar://problem/48480516>

Reviewed by Jer Noble.

No new tests, no functional change.

* platform/graphics/avfoundation/objc/SourceBufferPrivateAVFObjC.mm:
(WebCore::SourceBufferPrivateAVFObjC::SourceBufferPrivateAVFObjC): Only register for
kCMSampleBufferConsumerNotification_BufferConsumed notiication when
-[AVSampleBufferDisplayLayer prerollDecodeWithCompletionHandler:] isn't available.
(WebCore::SourceBufferPrivateAVFObjC::~SourceBufferPrivateAVFObjC): Ditto, for unregistering.
(WebCore::SourceBufferPrivateAVFObjC::enqueueSample): Use
-[AVSampleBufferDisplayLayer prerollDecodeWithCompletionHandler:] when possible.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242631 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[ContentChangeObserver] Click event fires immediately on hover menu at Ebbets.com
zalan@apple.com [Fri, 8 Mar 2019 04:24:40 +0000 (04:24 +0000)]
[ContentChangeObserver] Click event fires immediately on hover menu at Ebbets.com
https://bugs.webkit.org/show_bug.cgi?id=195397

Reviewed by Simon Fraser.

Source/WebCore:

This patch introduces TouchEventScope to track changes triggered by touch start.

Test: fast/events/touch/ios/content-observation/visibility-change-on-touch-start-simple.html

* page/ios/ContentChangeObserver.cpp:
(WebCore::ContentChangeObserver::touchEventDidStart):
(WebCore::ContentChangeObserver::touchEventDidFinish):
(WebCore::ContentChangeObserver::mouseMovedDidStart):
(WebCore::ContentChangeObserver::mouseMovedDidFinish):
(WebCore::ContentChangeObserver::adjustObservedState):
(WebCore::ContentChangeObserver::TouchEventScope::TouchEventScope):
(WebCore::ContentChangeObserver::TouchEventScope::~TouchEventScope):
* page/ios/ContentChangeObserver.h:
(WebCore::ContentChangeObserver::isObservingContentChanges const):

LayoutTests:

* fast/events/touch/ios/content-observation/visibility-change-on-touch-start-simple-expected.txt: Added.
* fast/events/touch/ios/content-observation/visibility-change-on-touch-start-simple.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242630 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoCrash under RemoteLayerTreePropertyApplier::applyProperties
timothy_horton@apple.com [Fri, 8 Mar 2019 03:26:38 +0000 (03:26 +0000)]
Crash under RemoteLayerTreePropertyApplier::applyProperties
https://bugs.webkit.org/show_bug.cgi?id=195448
<rdar://problem/48588226>

Reviewed by Simon Fraser.

* UIProcess/RemoteLayerTree/RemoteLayerTreeHost.mm:
(WebKit::RemoteLayerTreeHost::updateLayerTree):
Under some currently-unknown circumstances, the UI process is receiving
commits referring to layers that it does not know about.
One understood case of this was fixed in r241899, but there seem to be
cases remaining that are not understood.
Also, add a release log so that we can identify any downstream effects.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242629 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[ContentChangeObserver] Introduce fixed duration content observation
zalan@apple.com [Fri, 8 Mar 2019 03:21:14 +0000 (03:21 +0000)]
[ContentChangeObserver] Introduce fixed duration content observation
https://bugs.webkit.org/show_bug.cgi?id=195295
<rdar://problem/48579913>

Reviewed by Simon Fraser.

Source/WebCore:

Some pages have a runloop-like scheduling setup where the content triggering change happens at a nested timer firing.
This patch helps finding cases like that using a 32ms long fixed window. Currently nested timers get dropped on the floor and
we stop observing for content changes before they even get fired.

Test: fast/events/touch/ios/content-observation/visibility-change-happens-on-timer-hops.html

* page/ios/ContentChangeObserver.cpp:
(WebCore::ContentChangeObserver::ContentChangeObserver):
(WebCore::ContentChangeObserver::startContentObservationForDuration):
(WebCore::ContentChangeObserver::completeDurationBasedContentObservation):
(WebCore::ContentChangeObserver::didInstallDOMTimer):
(WebCore::ContentChangeObserver::didRemoveDOMTimer):
(WebCore::ContentChangeObserver::domTimerExecuteDidStart):
(WebCore::ContentChangeObserver::domTimerExecuteDidFinish):
(WebCore::ContentChangeObserver::styleRecalcDidStart):
(WebCore::ContentChangeObserver::styleRecalcDidFinish):
(WebCore::ContentChangeObserver::cancelPendingActivities):
(WebCore::ContentChangeObserver::didSuspendActiveDOMObjects):
(WebCore::ContentChangeObserver::willDetachPage):
(WebCore::ContentChangeObserver::contentVisibilityDidChange):
(WebCore::ContentChangeObserver::setShouldObserveNextStyleRecalc):
(WebCore::ContentChangeObserver::adjustObservedState):
(WebCore::ContentChangeObserver::StyleChangeScope::StyleChangeScope):
(WebCore::ContentChangeObserver::StyleChangeScope::~StyleChangeScope):
(WebCore::ContentChangeObserver::clearTimersAndReportContentChange): Deleted.
* page/ios/ContentChangeObserver.h:
(WebCore::ContentChangeObserver::clearObservedDOMTimers):
(WebCore::ContentChangeObserver::isObservingContentChanges const):
(WebCore::ContentChangeObserver::hasPendingActivity const):

Source/WebKit:

* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::handleSyntheticClick):
(WebKit::WebPage::completePendingSyntheticClickForContentChangeObserver):

LayoutTests:

* fast/events/touch/ios/content-observation/visibility-change-happens-on-timer-hops-expected.txt: Added.
* fast/events/touch/ios/content-observation/visibility-change-happens-on-timer-hops.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242628 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[JSC] Remove merging must handle values into proven types in CFA
ysuzuki@apple.com [Fri, 8 Mar 2019 02:54:17 +0000 (02:54 +0000)]
[JSC] Remove merging must handle values into proven types in CFA
https://bugs.webkit.org/show_bug.cgi?id=195444

Reviewed by Saam Barati.

Previously, we are merging must handle values as a proven constant in CFA. This is OK as long as this proven AbstractValue is blurred by merging the other legit AbstractValues
from the successors. But let's consider the following code, this is actually generated DFG graph from the attached test in r242626.

    Block #2 (loop header) succ #3, #4
    ...
    1: ForceOSRExit
    ...
    2: JSConstant(0)
    3: SetLocal(@2, loc6)
    ...
    4: Branch(#3, #4)

    Block #3 (This is OSR entry target) pred #2, #3, must handle value for loc6 => JSConstant(Int32, 31)
    ...
    5: GetLocal(loc6)
    6: StringFromCharCode(@5)
    ...

Block #3 is OSR entry target. So we have must handle value for loc6 and it is Int32 constant 31. Then we merge this constant as a proven value in #3's loc6 AbstractValue.
If the value from #2 blurs the value, it is OK. However, #2 has ForceOSRExit. So must handle value suddenly becomes the only source of loc6 in #3. Then we use this constant
as a proven value. But this is not expected behavior since must handle value is just a snapshot of the locals when we kick off the concurrent compilation. In the above example,
we assume that loop index is an constant 31, but it is wrong, and OSR entry fails. Because there is no strong assumption that the must handle value is the proven type or value,
we should not merge it in CFA.

Since (1) this is just an optimization, (2) type information is already propagated in prediction injection phase, and (3) the must handle value does not show the performance
progression in r211461 and we no longer see type misprediction in marsaglia-osr-entry.js, this patch simply removes must handle value type widening in CFA.

* dfg/DFGCFAPhase.cpp:
(JSC::DFG::CFAPhase::run):
(JSC::DFG::CFAPhase::performBlockCFA):
(JSC::DFG::CFAPhase::injectOSR): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242627 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[JSC] StringFromCharCode fast path should accept 0xff in DFG and FTL
ysuzuki@apple.com [Fri, 8 Mar 2019 01:53:35 +0000 (01:53 +0000)]
[JSC] StringFromCharCode fast path should accept 0xff in DFG and FTL
https://bugs.webkit.org/show_bug.cgi?id=195429

Reviewed by Saam Barati.

JSTests:

* stress/must-handled-values-should-not-be-used-as-proven-constants-in-cfa.js: Added.
(foo):
* stress/string-from-char-code-255.js: Added.

Source/JavaScriptCore:

We can create single characters without allocation up to 0xff character code. But currently, DFGSpeculativeJIT and FTLLowerDFGToB3 go to the slow path
for 0xff case. On the other hand, DFG DoesGC phase says GC won't happen if the child is int32 constant and it is <= 0xff. So, if you have `String.fromCharCode(0xff)`,
this breaks the assumption in DFG DoesGC. The correct fix is changing the check in DFGSpeculativeJIT and FTLLowerDFGToB3 from AboveOrEqual to Above.
Note that ThunkGenerators's StringFromCharCode thunk was correct.

* dfg/DFGSpeculativeJIT.cpp:
(JSC::DFG::SpeculativeJIT::compileFromCharCode):
* ftl/FTLLowerDFGToB3.cpp:
(JSC::FTL::DFG::LowerDFGToB3::compileStringFromCharCode):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242626 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[WinCairo][WebKit] Nothing is drawn after Web process crashed
Hironori.Fujii@sony.com [Fri, 8 Mar 2019 01:35:23 +0000 (01:35 +0000)]
[WinCairo][WebKit] Nothing is drawn after Web process crashed
https://bugs.webkit.org/show_bug.cgi?id=195399

Reviewed by Alex Christensen.

PageClientImpl::viewSize returned 0x0 size after Web process
crashed. Restored the original implementation using GetClientRect
(Bug 52175).

* UIProcess/win/PageClientImpl.cpp:
(WebKit::PageClientImpl::viewSize):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242625 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agorequestAnimationFrame should execute before the next frame
commit-queue@webkit.org [Fri, 8 Mar 2019 00:52:57 +0000 (00:52 +0000)]
requestAnimationFrame should execute before the next frame
https://bugs.webkit.org/show_bug.cgi?id=177484

Patch by Said Abou-Hallawa <sabouhallawa@apple.com> on 2019-03-07
Reviewed by Simon Fraser.

Source/WebCore:

This change fixes two issues with animation timing:

1. Calling the requestAnimationFrame callbacks would have happened when
   the DisplayLink fires. This may have happened even if the frame is
   missed and no display is committed.

2. Style changes and layout triggered by script could trigger painting
   at more than 60fps. CoreAnimation commits could happen at more than
   60fps, although WindowServer will throttle those, and only some will
   be shown on the screen.

This change introduces a new paint scheduling model where painting is
driven by a "RenderingUpdateScheduler", which only triggers paints once
per 16.7ms frame.

Code that previously scheduled a compositing layer flush now schedules a
"RenderingUpdate", and that update is driven by a DisplayRefreshMonitor
callback. When the render happens, we service requestAnimationFrame callbacks,
Web Animations and intersection observations per the "Update the rendering"
step of the HTML Event Loop specification
<https://html.spec.whatwg.org/multipage/webappapis.html#update-the-rendering>.

In the future, more rendering steps will be added to this code.

* Sources.txt:
* WebCore.xcodeproj/project.pbxproj:
* accessibility/mac/AXObjectCacheMac.mm:
Fix layout tests by adding null check.

* animation/DocumentAnimationScheduler.cpp: Removed.
* animation/DocumentAnimationScheduler.h: Removed.
* animation/DocumentTimeline.cpp:
(WebCore::DocumentTimeline::DocumentTimeline):
(WebCore::DocumentTimeline::updateThrottlingState):
(WebCore::DocumentTimeline::resumeAnimations):
(WebCore::DocumentTimeline::liveCurrentTime const):
(WebCore::DocumentTimeline::currentTime):
(WebCore::DocumentTimeline::animationTimingDidChange):
(WebCore::DocumentTimeline::scheduleAnimationResolution):
(WebCore::DocumentTimeline::unscheduleAnimationResolution):
(WebCore::DocumentTimeline::updateAnimationsAndSendEvents):
(WebCore::DocumentTimeline::internalUpdateAnimationsAndSendEvents):
(WebCore::DocumentTimeline::scheduleNextTick):
(WebCore::DocumentTimeline::updateListOfElementsWithRunningAcceleratedAnimationsForElement):
Simplify this function by handling the case of no-animations separately.

(WebCore::DocumentTimeline::resolveAnimationsForElement):
Simplify the loop and delete hasPendingAcceleratedAnimations because it
is initialized to true and is not changed inside the loop.

(WebCore::DocumentTimeline::scheduleAnimationResolutionIfNeeded): Deleted.
(WebCore::DocumentTimeline::animationResolutionTimerFired): Deleted.
* animation/DocumentTimeline.h:
* dom/Document.cpp:
(WebCore::Document::resolveStyle):
There is no need to force update in resolveStyle(). notifyFlushRequired()
will be called eventually which will scheduleRenderingUpdate().

(WebCore::Document::prepareForDestruction):
(WebCore::Document::updateAnimationsAndSendEvents):
(WebCore::Document::serviceRequestAnimationFrameCallbacks):
(WebCore::Document::windowScreenDidChange):
(WebCore::Document::updateIntersectionObservations):
(WebCore::Document::scheduleForcedIntersectionObservationUpdate): Deleted.
(WebCore::Document::animationScheduler): Deleted.
* dom/Document.h:
(WebCore::Document::numberOfIntersectionObservers const):
* dom/ScriptedAnimationController.cpp:
(WebCore::ScriptedAnimationController::serviceRequestAnimationFrameCallbacks):
(WebCore::ScriptedAnimationController::scheduleAnimation):
(WebCore::ScriptedAnimationController::animationTimerFired):
(WebCore::ScriptedAnimationController::serviceScriptedAnimations): Deleted.
(WebCore::ScriptedAnimationController::documentAnimationSchedulerDidFire): Deleted.
* dom/ScriptedAnimationController.h:
* page/FrameView.cpp:
(WebCore::FrameView::viewportContentsChanged):
* page/IntersectionObserver.cpp:
(WebCore::IntersectionObserver::observe):
* page/Page.cpp:
(WebCore::Page::Page):
(WebCore::Page::layoutIfNeeded):
(WebCore::Page::renderingUpdate):
(WebCore::Page::renderingUpdateScheduler):
(WebCore::Page::willDisplayPage): Deleted.
(WebCore::Page::addDocumentNeedingIntersectionObservationUpdate): Deleted.
(WebCore::Page::updateIntersectionObservations): Deleted.
(WebCore::Page::scheduleForcedIntersectionObservationUpdate): Deleted.
* page/Page.h:
* page/PageOverlayController.cpp:
(WebCore::PageOverlayController::didChangeViewExposedRect):
(WebCore::PageOverlayController::notifyFlushRequired):
* page/ResourceUsageData.h:
Include header files that become missing because of adding
RenderingUpdateScheduler.cpp.

* page/RenderingUpdateScheduler.cpp: Added.
(WebCore::RenderingUpdateScheduler::RenderingUpdateScheduler):
(WebCore::RenderingUpdateScheduler::scheduleRenderingUpdate):
(WebCore::RenderingUpdateScheduler::startTimer):
(WebCore::RenderingUpdateScheduler::clearTimer):
(WebCore::RenderingUpdateScheduler::windowScreenDidChange):
(WebCore::RenderingUpdateScheduler::createDisplayRefreshMonitor const):
(WebCore::RenderingUpdateScheduler::displayRefreshFired):
* page/RenderingUpdateScheduler.h: Added.
(WebCore::RenderingUpdateScheduler::create):
* page/ios/ContentChangeObserver.h:
Include header files that become missing because of adding
RenderingUpdateScheduler.cpp.

* page/mac/ServicesOverlayController.mm:
(WebCore::ServicesOverlayController::Highlight::notifyFlushRequired):
* rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::scheduleLayerFlushNow):

Source/WebKit:

Replace the calls to Page::layoutIfNeeded() and willDisplayPage() by
a single call to Page::renderingUpdate(). This new function implements
"Update the rendering" step of the HTML Event Loop specification
<https://html.spec.whatwg.org/multipage/webappapis.html#update-the-rendering>.

* WebProcess/WebPage/AcceleratedDrawingArea.cpp:
(WebKit::AcceleratedDrawingArea::updateBackingStoreState):
* WebProcess/WebPage/CoordinatedGraphics/LayerTreeHost.cpp:
(WebKit::LayerTreeHost::layerFlushTimerFired):
* WebProcess/WebPage/DrawingAreaImpl.cpp:
(WebKit::DrawingAreaImpl::display):
* WebProcess/WebPage/RemoteLayerTree/RemoteLayerTreeDrawingArea.mm:
(WebKit::RemoteLayerTreeDrawingArea::flushLayers):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::layoutIfNeeded):
(WebKit::WebPage::renderingUpdate):
(WebKit::WebPage::willDisplayPage): Deleted.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.mm:
(WebKit::TiledCoreAnimationDrawingArea::flushLayers):

Source/WebKitLegacy/mac:

* WebView/WebView.mm:
(-[WebView _viewWillDrawInternal]):
Call Page::renderingUpdate() which implements  "Update the rendering"
step of the HTML Event Loop specification.

Source/WebKitLegacy/win:

* WebView.cpp:
(WebView::updateBackingStore):
(WebView::flushPendingGraphicsLayerChangesSoon):
(WebView::flushPendingGraphicsLayerChanges):
Call Page::renderingUpdate() which implements  "Update the rendering"
step of the HTML Event Loop specification.

Source/WTF:

Add trace points for the page RenderingUpdate.

* wtf/SystemTracing.h:

Tools:

Add trace points for the page RenderingUpdate.

* Tracing/SystemTracePoints.plist:

LayoutTests:

* TestExpectations:
There is a slight difference between the actual DRT and the expected DRT
due to animation timing change. But these two tests are not animating
correctly if they are opened in Safari with web animation turned on.

* animations/animation-multiple-callbacks-timestamp.html:
Fix variable names used by an error message.
* animations/no-style-recalc-during-accelerated-animation-expected.txt:
* animations/no-style-recalc-during-accelerated-animation.html:
One extra styleReclc was incurred due to the document styleRecalcTimer.
I think this timer is not needed anymore. I will look at removing it in
a separate patch.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242624 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoSkip compilation of unused audio device files for Mac and iOS
youenn@apple.com [Fri, 8 Mar 2019 00:44:30 +0000 (00:44 +0000)]
Skip compilation of unused audio device files for Mac and iOS
https://bugs.webkit.org/show_bug.cgi?id=195412

Reviewed by Eric Carlson.

Stop compiling audio_device_mac.cc, audio_mixer_manager_mac.cc and voice_processing_audio_unit.mm
as unused in WebKit.
* libwebrtc.xcodeproj/project.pbxproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242623 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months agoWeb Inspector: Styles: overridden CSS property should have go-to button to jump to...
nvasilyev@apple.com [Fri, 8 Mar 2019 00:40:02 +0000 (00:40 +0000)]
Web Inspector: Styles: overridden CSS property should have go-to button to jump to effective property
https://bugs.webkit.org/show_bug.cgi?id=185930
<rdar://problem/40506252>

Reviewed by Matt Baker.

Introduce a new experimental jump to effective property button. The button is a small arrow button
next to an overridden CSS property. Clicking the button scrolls to the effective CSS property and
selects it.

* Localizations/en.lproj/localizedStrings.js:
* UserInterface/Base/Setting.js:
* UserInterface/Models/CSSProperty.js:
(WI.CSSProperty):
(WI.CSSProperty.prototype.update):
(WI.CSSProperty.prototype.get overridingProperty):
(WI.CSSProperty.prototype.set overridingProperty):
* UserInterface/Models/DOMNodeStyles.js:
(WI.DOMNodeStyles.prototype._markOverriddenProperties):
* UserInterface/Views/SettingsTabContentView.js:
(WI.SettingsTabContentView.prototype._createExperimentalSettingsView):
* UserInterface/Views/SpreadsheetCSSStyleDeclarationEditor.css:
(.spreadsheet-style-declaration-editor .property.overridden .select-effective-property):
(.spreadsheet-style-declaration-editor .property.overridden:hover .select-effective-property,):
(.spreadsheet-style-declaration-editor .property.overridden:hover .select-effective-property::after,):
(@media (prefers-color-scheme: dark)):
* UserInterface/Views/SpreadsheetCSSStyleDeclarationEditor.js:
(WI.SpreadsheetCSSStyleDeclarationEditor.prototype.spreadsheetStylePropertySelectByProperty):
* UserInterface/Views/SpreadsheetCSSStyleDeclarationSection.js:
(WI.SpreadsheetCSSStyleDeclarationSection.prototype.spreadsheetCSSStyleDeclarationEditorSelectProperty):
* UserInterface/Views/SpreadsheetRulesStyleDetailsPanel.js:
(WI.SpreadsheetRulesStyleDetailsPanel.prototype.spreadsheetCSSStyleDeclarationSectionSelectProperty):
* UserInterface/Views/SpreadsheetStyleProperty.js:
(WI.SpreadsheetStyleProperty.prototype.updateStatus):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242622 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 months ago[ContentChangeObserver] Add a setting to be able to turn content change observation...
zalan@apple.com [Fri, 8 Mar 2019 00:36:42 +0000 (00:36 +0000)]
[ContentChangeObserver] Add a setting to be able to turn content change observation on/off
https://bugs.webkit.org/show_bug.cgi?id=195353
<rdar://problem/48626394>

Reviewed by Simon Fraser.

Source/WebCore:

Move content observation tests to a dedicated directory.

Tests: fast/events/touch/ios/content-observation/click-instead-of-hover-simple.html
       fast/events/touch/ios/content-observation/hover-when-style-change-is-async.html
       fast/events/touch/ios/content-observation/stuck-with-hover-state.html
       fast/events/touch/ios/content-observation/style-recalc-schedule-and-force-relalc.html
       fast/events/touch/ios/content-observation/visibility-change-happens-at-the-second-timer.html

* page/Settings.yaml:
* page/ios/ContentChangeObserver.cpp:
(WebCore::ContentChangeObserver::didInstallDOMTimer):
(WebCore::ContentChangeObserver::isNotifyContentChangeAllowed const):
* page/ios/ContentChangeObserver.h:
(WebCore::ContentChangeObserver::isNotifyContentChangeAllowed const): Deleted.
* testing/InternalSettings.cpp:
(WebCore::InternalSettings::resetToConsistentState):

Source/WebKit:

* Shared/WebPreferences.yaml:
* UIProcess/API/C/WKPreferences.cpp:
(WKPreferencesSetContentChangeObserverEnabled):
(WKPreferencesGetContentChangeObserverEnabled):
* UIProcess/API/C/WKPreferencesRefPrivate.h:
* UIProcess/API/Cocoa/WKPreferences.mm:
(-[WKPreferences _contentChangeObserverEnabled]):
(-[WKPreferences _setContentChangeObserverEnabled:]):
* UIProcess/API/Cocoa/WKPreferencesPrivate.h:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::handleSyntheticClick):

Source/WebKitLegacy/mac:

* WebView/WebPreferenceKeysPrivate.h:
* WebView/WebPreferences.mm:
(+[WebPreferences initialize]):
(-[WebPreferences contentChangeObserverEnabled]):
(-[WebPreferences setContentChangeObserverEnabled:]):
* WebView/WebPreferencesPrivate.h:
* WebView/WebView.mm:
(-[WebView _preferencesChanged:]):

Tools:

* WebKitTestRunner/ios/TestControllerIOS.mm:
(WTR::TestController::platformResetPreferencesToConsistentValues):

LayoutTests:

* fast/events/touch/ios/content-observation/click-instead-of-hover-simple-expected.txt: Renamed from LayoutTests/fast/events/touch/ios/click-instead-of-hover-simple-expected.txt.
* fast/events/touch/ios/content-observation/click-instead-of-hover-simple.html: Renamed from LayoutTests/fast/events/touch/ios/click-instead-of-hover-simple.html.
* fast/events/touch/ios/content-observation/hover-when-style-change-is-async-expected.txt: Renamed from LayoutTests/fast/events/touch/ios/hover-when-style-change-is-async-expected.txt.
* fast/events/touch/ios/content-observation/hover-when-style-change-is-async.html: Renamed from LayoutTests/fast/events/touch/ios/hover-when-style-change-is-async.html.
* fast/events/touch/ios/content-observation/stuck-with-hover-state-expected.txt: Renamed from LayoutTests/fast/events/touch/ios/stuck-with-hover-state-expected.txt.
* fast/events/touch/ios/content-observation/stuck-with-hover-state.html: Renamed from LayoutTests/fast/events/touch/ios/stuck-with-hover-state.html.
* fast/events/touch/ios/content-observation/style-recalc-schedule-and-force-relalc-expected.txt: Renamed from LayoutTests/fast/events/touch/ios/style-recalc-schedule-and-force-relalc-expected.txt.
* fast/events/touch/ios/content-observation/style-recalc-schedule-and-force-relalc.html: Renamed from LayoutTests/fast/events/touch/ios/style-recalc-schedule-and-force-relalc.html.
* fast/events/touch/ios/content-observation/visibility-change-happens-at-the-second-timer-expected.txt: Renamed from LayoutTests/fast/events/touch/ios/visibility-change-happens-at-the-second-timer-expected.txt.
* fast/events/touch/ios/content-observation/visibility-change-happens-at-the-second-timer.html: Renamed from LayoutTests/fast/events/touch/ios/visibility-change-happens-at-the-second-timer.html.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@242621 268f45cc-cd09-0410-ab3c-d52691b4dbfc