⚠ Archived content — this site is no longer maintained.   Current WebKit documentation is at docs.webkit.org.

Timeline



Aug 8, 2017:

10:36 PM Changeset in webkit [220443] by Wenson Hsieh
  • 29 edits
    4 deletes in trunk

Unreviewed, rolling out r220393.

This patch regresses the performance of WebPage::editorState.

Reverted changeset:

"[iOS WK2] WKWebView schedules nonstop layout after pressing
cmb+b,i,u inside a contenteditable div"
​https://bugs.webkit.org/show_bug.cgi?id=175116
​http://trac.webkit.org/changeset/220393

10:15 PM Changeset in webkit [220442] by Chris Dumez
  • 28 edits
    17 adds in trunk

[Beacon] Add support for CORS-preflighting for WK2 / NETWORK_SESSION
​https://bugs.webkit.org/show_bug.cgi?id=175264
<rdar://problem/33547793>

Reviewed by Youenn Fablet.

Source/WebCore:

Pass additional information when creating a PingHandle so that the PingLoad
can deal with CORS-preflighting on Network process side.

Tests: http/wpt/beacon/cors/cors-preflight-arraybufferview-failure.html

http/wpt/beacon/cors/cors-preflight-arraybufferview-success.html
http/wpt/beacon/cors/cors-preflight-blob-failure.html
http/wpt/beacon/cors/cors-preflight-blob-success.html
http/wpt/beacon/cors/cors-preflight-cookie.html

  • WebCore.xcodeproj/project.pbxproj:
  • loader/CrossOriginAccessControl.cpp:

(WebCore::validatePreflightResponse):

  • loader/CrossOriginAccessControl.h:
  • loader/CrossOriginPreflightChecker.cpp:

(WebCore::CrossOriginPreflightChecker::validatePreflightResponse):

  • loader/CrossOriginPreflightResultCache.h:
  • loader/LoaderStrategy.h:
  • loader/PingLoader.cpp:

(WebCore::PingLoader::loadImage):
(WebCore::PingLoader::sendPing):
(WebCore::PingLoader::sendViolationReport):
(WebCore::PingLoader::startPingLoad):

  • loader/PingLoader.h:
  • loader/cache/CachedResource.cpp:

(WebCore::CachedResource::load):

  • page/SecurityOrigin.h:

Source/WebKit:

Implement CORS-preflighting for beacons with a payload that has a non
safelisted MIME type, as per:

CORS-preflighting is completely handled on Network Process side because
a beacon request can outlive its page and therefore its WebContent
process. This requires us to pass a little more information to the
Network process, in particular the source origin and the corsMode.

The current implementation does not currently deal with CORS preflights
needed upon a redirect. This will be added in a follow-up.

  • CMakeLists.txt:
  • NetworkProcess/NetworkCORSPreflightChecker.cpp: Added.

(WebKit::NetworkCORSPreflightChecker::NetworkCORSPreflightChecker):
(WebKit::NetworkCORSPreflightChecker::~NetworkCORSPreflightChecker):
(WebKit::NetworkCORSPreflightChecker::startPreflight):
(WebKit::NetworkCORSPreflightChecker::willPerformHTTPRedirection):
(WebKit::NetworkCORSPreflightChecker::didReceiveChallenge):
(WebKit::NetworkCORSPreflightChecker::didReceiveResponseNetworkSession):
(WebKit::NetworkCORSPreflightChecker::didReceiveData):
(WebKit::NetworkCORSPreflightChecker::didCompleteWithError):
(WebKit::NetworkCORSPreflightChecker::didSendData):
(WebKit::NetworkCORSPreflightChecker::wasBlocked):
(WebKit::NetworkCORSPreflightChecker::cannotShowURL):

  • NetworkProcess/NetworkCORSPreflightChecker.h: Added.
  • NetworkProcess/NetworkConnectionToWebProcess.cpp:

(WebKit::NetworkConnectionToWebProcess::loadPing):

  • NetworkProcess/NetworkConnectionToWebProcess.h:
  • NetworkProcess/NetworkResourceLoadParameters.cpp:

(WebKit::NetworkResourceLoadParameters::encode const):
(WebKit::NetworkResourceLoadParameters::decode):

  • NetworkProcess/NetworkResourceLoadParameters.h:
  • NetworkProcess/PingLoad.cpp: Added.

(WebKit::PingLoad::PingLoad):
(WebKit::PingLoad::~PingLoad):
(WebKit::PingLoad::startNetworkLoad):
(WebKit::PingLoad::willPerformHTTPRedirection):
(WebKit::PingLoad::didReceiveChallenge):
(WebKit::PingLoad::didReceiveResponseNetworkSession):
(WebKit::PingLoad::didReceiveData):
(WebKit::PingLoad::didCompleteWithError):
(WebKit::PingLoad::didSendData):
(WebKit::PingLoad::wasBlocked):
(WebKit::PingLoad::cannotShowURL):
(WebKit::PingLoad::timeoutTimerFired):
(WebKit::PingLoad::needsCORSPreflight const):
(WebKit::PingLoad::doCORSPreflight):

  • NetworkProcess/PingLoad.h:
  • WebKit.xcodeproj/project.pbxproj:
  • WebProcess/Network/WebLoaderStrategy.cpp:

(WebKit::WebLoaderStrategy::createPingHandle):

  • WebProcess/Network/WebLoaderStrategy.h:

Source/WebKitLegacy:

createPingHandle() now takes new parameters but there is currently no behavior
change on WebKit1.

  • WebCoreSupport/WebResourceLoadScheduler.cpp:

(WebResourceLoadScheduler::createPingHandle):

  • WebCoreSupport/WebResourceLoadScheduler.h:

LayoutTests:

Add layout test coverage.

  • http/wpt/beacon/cors/cors-preflight-arraybufferview-failure-expected.txt: Added.
  • http/wpt/beacon/cors/cors-preflight-arraybufferview-failure.html: Added.
  • http/wpt/beacon/cors/cors-preflight-arraybufferview-success-expected.txt: Added.
  • http/wpt/beacon/cors/cors-preflight-arraybufferview-success.html: Added.
  • http/wpt/beacon/cors/cors-preflight-blob-failure-expected.txt: Added.
  • http/wpt/beacon/cors/cors-preflight-blob-failure.html: Added.
  • http/wpt/beacon/cors/cors-preflight-blob-success-expected.txt: Added.
  • http/wpt/beacon/cors/cors-preflight-blob-success.html: Added.
  • http/wpt/beacon/cors/cors-preflight-cookie-expected.txt: Added.
  • http/wpt/beacon/cors/cors-preflight-cookie.html: Added.
  • http/wpt/beacon/resources/beacon-preflight.py: Added.

(respondToCORSPreflight):
(main):

  • http/wpt/beacon/resources/set-cookie.py: Added.

(main):

  • platform/mac-wk1/TestExpectations:
  • platform/mac-wk2/TestExpectations:
  • platform/win/TestExpectations:
8:48 PM Changeset in webkit [220441] by fpizlo@apple.com
  • 3 edits in trunk/Source/JavaScriptCore

ICs should do caging
​https://bugs.webkit.org/show_bug.cgi?id=175295

Reviewed by Saam Barati.

Adds the appropriate cage() calls in our inline caches.

  • bytecode/AccessCase.cpp:

(JSC::AccessCase::generateImpl):

  • bytecode/InlineAccess.cpp:

(JSC::InlineAccess::dumpCacheSizesAndCrash):
(JSC::InlineAccess::generateSelfPropertyAccess):
(JSC::InlineAccess::generateSelfPropertyReplace):
(JSC::InlineAccess::generateArrayLength):

8:24 PM Changeset in webkit [220440] by weinig@apple.com
  • 4 edits in trunk/Source/WebCore

Address review feedback from ​https://bugs.webkit.org/show_bug.cgi?id=175246.

  • bindings/js/DOMPromiseProxy.h:
  • css/FontFaceSet.cpp:
  • css/FontFaceSet.h:
8:18 PM Changeset in webkit [220439] by jmarcell@apple.com
  • 3 edits in branches/safari-604.1.38.1-branch/Source/WebKitLegacy/mac

Cherry-pick r220429. rdar://problem/33789343

8:18 PM Changeset in webkit [220438] by jmarcell@apple.com
  • 3 edits in branches/safari-604.1.38.1-branch

Cherry-pick r220289. rdar://problem/33789082

8:09 PM Changeset in webkit [220437] by jmarcell@apple.com
  • 3 edits in branches/safari-604.1.38.0-branch

Cherry-pick r220289. rdar://problem/33789085

6:55 PM Changeset in webkit [220436] by Devin Rousso
  • 12 edits
    2 adds in trunk

Web Inspector: Canvas: support editing WebGL shaders
​https://bugs.webkit.org/show_bug.cgi?id=124211
<rdar://problem/15448958>

Reviewed by Matt Baker.

Source/JavaScriptCore:

  • inspector/protocol/Canvas.json:

Add updateShader command that will change the given shader's source to the provided string,
recompile, and relink it to its associated program.
Drive-by: add description to requestShaderSource command.

Source/WebCore:

Test: inspector/canvas/updateShader.html

  • inspector/InspectorCanvasAgent.h:
  • inspector/InspectorCanvasAgent.cpp:

(WebCore::InspectorCanvasAgent::updateShader):

  • html/canvas/WebGLRenderingContextBase.h:
  • html/canvas/WebGLRenderingContextBase.cpp:

(WebCore::WebGLRenderingContextBase::linkProgram):
(WebCore::WebGLRenderingContextBase::linkProgramWithoutInvalidatingAttribLocations):
Normally, when a program is linked, it invalidates any WebGLUniformLocation associated with
the program by incrementing its linkCount. In order to allow live editing of shaders, we
need to be able to compile and link a shader without invalidating these locations. This
patch moves the shader linking logic to its own function that is called by linkProgram so
that InspectorCanvasAgent can compile and link without invalidation.

Source/WebInspectorUI:

  • UserInterface/Models/ShaderProgram.js:

(WI.ShaderProgram.prototype.updateVertexShader):
(WI.ShaderProgram.prototype.updateFragmentShader):
(WI.ShaderProgram.prototype._updateShader):

  • UserInterface/Views/ShaderProgramContentView.js:

(WI.ShaderProgramContentView):
(WI.ShaderProgramContentView.prototype._contentDidChange):

LayoutTests:

  • inspector/canvas/updateShader-expected.txt: Added.
  • inspector/canvas/updateShader.html: Added.
  • platform/win/TestExpectations:
6:39 PM Changeset in webkit [220435] by Megan Gardner
  • 3 edits in trunk/Source/WebKit

Remove old and unused pointIsInAssistedNode definition
​https://bugs.webkit.org/show_bug.cgi?id=175335

Stage removal here to allow removal from UIKit.
Replaced by textInteractionGesture:shouldBeginAtPoint:

Reviewed by Tim Horton.

  • Platform/spi/ios/UIKitSPI.h:
  • UIProcess/ios/WKContentViewInteraction.mm:

(-[WKContentView pointIsInAssistedNode:]): Deleted.

6:24 PM Changeset in webkit [220434] by commit-queue@webkit.org
  • 6 edits in trunk/Tools

Created Individual EWS Pages
​https://bugs.webkit.org/show_bug.cgi?id=174477

Patch by obinna obike <​oobike@apple.com> on 2017-08-08
Reviewed by Aakash Jain.

Created individual pages for each EWS. Changed the link within each status bubble
to go directly to the individual EWS page. Also added a link to original status page showing all EWSes.

  • QueueStatusServer/handlers/patch.py:

(Patch.get): Handles the case for individual ews queue as well as for all EWS queues.

  • QueueStatusServer/handlers/statusbubble.py:

(StatusBubble._build_bubble): Updated url for specific EWS.

  • QueueStatusServer/main.py: Added a link to the individual ews in the routes dictionary.
  • QueueStatusServer/templates/patch.html: Added a link to go directly to the page with

all ews queues.

  • QueueStatusServer/templates/statusbubble.html: If you click on a statusbubble it

takes you directly to the individual ews page.

6:01 PM Changeset in webkit [220433] by weinig@apple.com
  • 33 edits
    1 add
    5 deletes in trunk/Source/WebCore

[WebIDL] Add support for Promise<> attributes
​https://bugs.webkit.org/show_bug.cgi?id=175246

Reviewed by Yusuke Suzuki.

Introduces and adopts DOMPromiseProxy, a new class for binding promises in a more
straightforward and safe (from a wrapper world perspective) way.

  • CMakeLists.txt:
  • WebCore.xcodeproj/project.pbxproj:
  • bindings/js/JSFontFaceCustom.cpp: Removed.
  • bindings/js/JSFontFaceSetCustom.cpp: Removed.
  • bindings/js/JSMediaKeySessionCustom.cpp: Removed.
  • bindings/js/JSServiceWorkerContainerCustom.cpp: Removed.
  • bindings/js/JSWebGPUCommandBufferCustom.cpp: Removed.

Remove custom bindings only needed for promise attributes. Add DOMPromiseProxy.h

  • Modules/encryptedmedia/MediaKeySession.cpp:

(WebCore::MediaKeySession::sessionClosed):
(WebCore::MediaKeySession::registerClosedPromise): Deleted.

  • Modules/encryptedmedia/MediaKeySession.h:
  • Modules/encryptedmedia/MediaKeySession.idl:

Replace custom promise code using the registerPromise idiom with DOMPromiseProxy.

  • bindings/IDLTypes.h:

Add StorageType type alias to allow specifying a specific form for storage. This allows
use to use Ref<> for interfaces in DOMPromiseProxy.

(WebCore::IDLWrapper::convertToParameterType):

Also add convertToParameterType, an annoying hack to workaround the fact WTF::Function
can't return references / no

  • bindings/js/DOMPromiseProxy.h: Added.

(WebCore::DOMPromiseProxy<IDLType>):

DOMPromiseProxy is a new class (set of classes) to represent promises that works with the
standard toJS<> / JSDOMConvert infrastructure used by the bindings generator. From the
implementation perspective, it is similar to DOMPromiseDeferred, but instead of receiving
one as a parameter, you can initialize it directly in your constructor and start using it.

From the bindings perspective, a DOMPromiseProxy can be converted into a JS Promise by calling
its promise function, which will either create a new promise, or return a cached one.

To make this work, DOMPromiseProxy maintains a Vector of DeferredPromises that hold onto those
actualized promises. The reason it has a Vector, is that we need one DeferredPromises per
DOMWrapperWorld, just like we need one wrapper for an interface per DOMWrapperWorld in the
wrapper cache. In most cases, only the normal world will be in play, so the Vector has an inline
capacity of one.


In addition, DOMPromiseProxy maintains a Variant of either an Exception or the result value type.
This allows the DOMPromiseProxy to be resolved or rejected at anytime. Then, when the bindings
request the promise, the stored result / exception can be resolved / rejected into the promise.

(WebCore::DOMPromiseProxy<IDLVoid>):

For void promises, we need slightly different semantics (resolve takes no parameters, Value is
a bit indicating resolved state), so it is specialized.

(WebCore::DOMPromiseProxyWithResolveCallback<IDLType>):

In addition to the void specialization, we have DOMPromiseProxyWithResolveCallback. Instead of
storing the value of the resolution directly, this specialization allows the owner to specify
callback to be called when the resolved value is needed. This is needed to avoid reference
cycles when the resolved value is the owner such as is the case with FontFace and FontFaceSet.

  • bindings/js/JSDOMConvertPromise.h:

(WebCore::JSConverter<IDLPromise<T>>::convert):
Add converter that calls through to the DOMPromiseProxy function promise.

  • bindings/js/JSDOMPromiseDeferred.cpp:

(WebCore::DeferredPromise::callFunction):
(WebCore::DeferredPromise::reject):

  • bindings/js/JSDOMPromiseDeferred.h:

(WebCore::DeferredPromise::create):
(WebCore::DeferredPromise::DeferredPromise):

Add a new create function for DeferredPromise that creates the JSPromiseDeferred for you.
Also adds a mode to DeferredPromise to allow keeping the promise after resolution. Maintain
the old behavior as the default.

Change reject to take an Exception as sink rather than an r-value, to allow DOMPromiseProxy
to copy a Exception into multiple DeferredPromises. For callers already moving into reject,
this should be no change.

  • bindings/scripts/CodeGeneratorJS.pm:

(AddToIncludesForIDLType):
(GenerateParametersCheck):
(GenerateImplementationFunctionCall):
(NativeToJSValueDOMConvertNeedsState):
(NativeToJSValueDOMConvertNeedsGlobalObject):

Add support DOMPromiseProxy. For now, this means a new extended attribute, [PromiseProxy],
but once I convert all the existing promises over to using DOMPromiseProxy, it will be removed.


Also specify the right #includes for promises and necessary parameters for conversion.

  • bindings/scripts/IDLAttributes.json:


Add [PromiseProxy], a temporary extended attribute for using DOMPromiseProxy.

  • css/FontFace.cpp:
  • css/FontFace.h:
  • css/FontFace.idl:
  • css/FontFaceSet.cpp:
  • css/FontFaceSet.h:
  • css/FontFaceSet.idl:

Replace custom promise code using the registerPromise idiom with DOMPromiseProxyWithResolveCallback.
The callback is necessary for these two classes since the value the resolve is themselves, and using
a normal resolve would create a cycle.

  • html/canvas/WebGPUCommandBuffer.cpp:
  • html/canvas/WebGPUCommandBuffer.h:
  • html/canvas/WebGPUCommandBuffer.idl:
  • platform/graphics/cocoa/GPUCommandBufferMetal.mm:
  • platform/graphics/gpu/GPUCommandBuffer.cpp:
  • platform/graphics/gpu/GPUCommandBuffer.h:

Replace custom promise code using the registerPromise idiom with DOMPromiseProxy.

  • page/NavigatorBase.cpp:

(WebCore::NavigatorBase::NavigatorBase):

  • page/NavigatorBase.h:

Forward declare ServiceWorkerContainer and add an explicit constructor to allow
removing #include of ServiceWorkerContainer.h in the header, avoiding an #include cycle.

  • workers/ServiceWorkerContainer.cpp:
  • workers/ServiceWorkerContainer.h:
  • workers/ServiceWorkerContainer.idl:

Replace custom promise code using the registerPromise idiom with DOMPromiseProxy.

  • bindings/scripts/test/JS/JSTestObj.cpp:
  • bindings/scripts/test/JS/JSTestPromiseRejectionEvent.cpp:
  • bindings/scripts/test/TestObj.idl:

Add / update tests.

5:26 PM Changeset in webkit [220432] by commit-queue@webkit.org
  • 4 edits in trunk/Source/JavaScriptCore

Make JSC_validateExceptionChecks=1 succeed on JSTests/slowMicrobenchmarks/spread-small-array.js.
​https://bugs.webkit.org/show_bug.cgi?id=175347

Patch by Robin Morisset <​rmorisset@apple.com> on 2017-08-08
Reviewed by Saam Barati.

This is done by making finishCreation explicitely check for exceptions after setConstantRegister and setConstantIdentifiersSetRegisters.
I chose to have this check replace the boolean returned previously by these functions for readability. The performance impact should be
negligible considering how much more finishCreation does.
This fix then caused another issue to appear as it was now clear that finishCreation can throw. And since it is called by ProgramCodeBlock::create(),
FunctionCodeBlock::create() and friends, that are in turn called by ScriptExecutable::newCodeBlockFor, this last function also required a few tweaks.

  • bytecode/CodeBlock.cpp:

(JSC::CodeBlock::finishCreation):
(JSC::CodeBlock::setConstantIdentifierSetRegisters):
(JSC::CodeBlock::setConstantRegisters):

  • bytecode/CodeBlock.h:
  • runtime/ScriptExecutable.cpp:

(JSC::ScriptExecutable::newCodeBlockFor):

5:18 PM Changeset in webkit [220431] by commit-queue@webkit.org
  • 2 edits
    1 add in trunk/Tools

Separate jsc stress test script writer from run-jsc-stress-tests
​https://bugs.webkit.org/show_bug.cgi?id=175216

Patch by Stephan Szabo <​stephan.szabo@sony.com> on 2017-08-08
Reviewed by Mark Lam.

  • Scripts/run-jsc-stress-tests:
  • Scripts/webkitruby/jsc-stress-test-writer-default.rb: Added.
5:03 PM Changeset in webkit [220430] by wilander@apple.com
  • 2 edits in trunk/Source/WebCore

Resource Load Statistics: Merge of dataRecordsRemoved should use std::max()
​https://bugs.webkit.org/show_bug.cgi?id=175348
<rdar://problem/33785695>

Reviewed by Chris Dumez.

No new tests. Online change of a statistic that doesn't affect logic flow.

  • loader/ResourceLoadStatistics.cpp:

(WebCore::ResourceLoadStatistics::merge):

Now uses std::max() instead of addition.

4:53 PM Changeset in webkit [220429] by beidson@apple.com
  • 3 edits in trunk/Source/WebKitLegacy/mac

Don't enable default icon loading in WK1 for apps linked against old SDKs.
<rdar://problem/33724060> and ​https://bugs.webkit.org/show_bug.cgi?id=175342

Reviewed by Andy Estes.

  • Misc/WebKitVersionChecks.h:
  • WebCoreSupport/WebFrameLoaderClient.mm:

(WebFrameLoaderClient::getLoadDecisionForIcons): Don't do any icon loading unless

the app is linked against new WebKit with the new behavior.

4:36 PM Changeset in webkit [220428] by Ryan Haddad
  • 2 edits in trunk/LayoutTests

Mark media/modern-media-controls/css/webkit-cursor-visibility-auto-hide.html as flaky.
​https://bugs.webkit.org/show_bug.cgi?id=175195

Unreviewed test gardening.

  • platform/mac-wk1/TestExpectations:
4:27 PM Changeset in webkit [220427] by Brent Fulgham
  • 4 edits
    3 adds in trunk

Sandbox flags do not support document.domain control
​https://bugs.webkit.org/show_bug.cgi?id=175281
<rdar://problem/33778936>

Reviewed by Chris Dumez.

Source/WebCore:

Update the 'setDomain' logic to honor the sandbox properties as defined in the current
HTML5 specification. This brings us in line with how Chrome and other browsers have
worked for some time.

Test: fast/frames/sandboxed-iframe-domain.html

  • dom/Document.cpp:

(WebCore::Document::setDomain): Add check for sandbox flag (with appropriate error message)

  • dom/SecurityContext.h:

LayoutTests:

  • fast/frames/resources/sandboxed-iframe-set-domain.html: Added.
  • fast/frames/sandboxed-iframe-domain.html: Added.
  • fast/frames/sandboxed-iframe-domain-expected.txt: Added.
4:24 PM Changeset in webkit [220426] by Matt Lewis
  • 3 edits in trunk/LayoutTests

Skipping imported/w3c/IndexedDB-private-browsing/idbfactory_open12.html
<rdar://problem/33555052>

Unreviewed test gardening.

  • platform/ios-wk2/TestExpectations:
  • platform/mac-highsierra/TestExpectations:
4:03 PM Changeset in webkit [220425] by Michael Catanzaro
  • 9 edits in trunk

Unreviewed, fix Ubuntu LTS build
​https://bugs.webkit.org/show_bug.cgi?id=174490

Source/JavaScriptCore:

  • inspector/remote/glib/RemoteInspectorGlib.cpp:
  • inspector/remote/glib/RemoteInspectorServer.cpp:

Source/WebDriver:

  • glib/SessionHostGlib.cpp:

Source/WebKit:

  • UIProcess/glib/RemoteInspectorClient.cpp:

(WebKit::RemoteInspectorClient::RemoteInspectorClient):

Tools:

  • TestWebKitAPI/Tests/WebKitGLib/TestAutomationSession.cpp:
3:53 PM Changeset in webkit [220424] by jmarcell@apple.com
  • 7 edits in branches/safari-604.1.38.0-branch/Source

Versioning.

3:50 PM Changeset in webkit [220423] by jmarcell@apple.com
  • 7 edits in branches/safari-604.1.38.1-branch/Source

Versioning.

3:46 PM Changeset in webkit [220422] by jmarcell@apple.com
  • 1 copy in tags/Safari-604.1.38.1.2

Tag Safari-604.1.38.1.2.

3:45 PM Changeset in webkit [220421] by jmarcell@apple.com
  • 1 copy in tags/Safari-604.1.38.0.2

Tag Safari-604.1.38.0.2.

2:38 PM WebKitGTK/Gardening/Calendar edited by cturner@igalia.com
(diff)
2:06 PM Changeset in webkit [220420] by Wenson Hsieh
  • 2 edits in trunk/Source/WebKit

Unreviewed, fix the Mac debug build after r220418.

  • UIProcess/Cocoa/WebViewImpl.mm:

(WebKit::WebViewImpl::prepareForMoveToWindow):

1:55 PM Changeset in webkit [220419] by Ryan Haddad
  • 2 edits in trunk/LayoutTests

Skip plugins/crash-restoring-plugin-page-from-page-cache.html on macOS debug.
​https://bugs.webkit.org/show_bug.cgi?id=175270

Unreviewed test gardening.

  • platform/mac-wk2/TestExpectations:
1:53 PM WebKitGTK/Gardening/Calendar edited by cturner@igalia.com
(diff)
1:52 PM WebKitGTK/Gardening/Calendar edited by cturner@igalia.com
(diff)
1:45 PM Changeset in webkit [220418] by Wenson Hsieh
  • 6 edits
    2 moves in trunk

[Mac WK2] WebViewImpl should hold a weak pointer to its WKWebView or WKView
​https://bugs.webkit.org/show_bug.cgi?id=175302
<rdar://problem/33762116>

Reviewed by Tim Horton.

Source/WebKit:

Refactor WebViewImpl's m_view from a raw pointer to a WeakObjCPtr to ensure that it is automatically cleared out
when the WKWebView or WKView are destroyed, and also changes several places where we implicitly block capture
the WebViewImpl to only capture a WeakPtr to the WebViewImpl. Most of the changes here are boilerplate
adjustments for how m_view is now a weak pointer.

Test: WKWebViewMacEditingTests.DoNotCrashWhenInterpretingKeyEventWhileDeallocatingView.

  • Shared/mac/WeakObjCPtr.h:

(WebKit::WeakObjCPtr::operator ValueType * const):

  • UIProcess/Cocoa/WebViewImpl.h:
  • UIProcess/Cocoa/WebViewImpl.mm:

(WebKit::WebViewImpl::updateTouchBar):
(WebKit::WebViewImpl::updateMediaTouchBar):
(WebKit::WebViewImpl::WebViewImpl):
(WebKit::WebViewImpl::~WebViewImpl):
(WebKit::WebViewImpl::window):
(WebKit::WebViewImpl::resignFirstResponder):
(WebKit::WebViewImpl::isFocused const):
(WebKit::WebViewImpl::renewGState):
(WebKit::WebViewImpl::setFrameAndScrollBy):
(WebKit::WebViewImpl::updateWindowAndViewFrames):
(WebKit::WebViewImpl::updateLayer):
(WebKit::WebViewImpl::printOperationWithPrintInfo):
(WebKit::WebViewImpl::updateContentInsetsIfAutomatic):
(WebKit::WebViewImpl::setLayoutMode):
(WebKit::WebViewImpl::intrinsicDeviceScaleFactor const):
(WebKit::WebViewImpl::windowDidBecomeKey):
(WebKit::WebViewImpl::windowDidResignKey):
(WebKit::WebViewImpl::windowDidChangeScreen):
(WebKit::WebViewImpl::mightBeginDragWhileInactive):
(WebKit::WebViewImpl::acceptsFirstMouse):
(WebKit::WebViewImpl::shouldDelayWindowOrderingForEvent):
(WebKit::WebViewImpl::windowResizeMouseLocationIsInVisibleScrollerThumb):
(WebKit::WebViewImpl::viewWillMoveToWindow):
(WebKit::WebViewImpl::viewDidMoveToWindow):
(WebKit::WebViewImpl::viewDidChangeBackingProperties):
(WebKit::WebViewImpl::hitTest):
(WebKit::WebViewImpl::postFakeMouseMovedEventForFlagsChangedEvent):
(WebKit::WebViewImpl::colorSpace):
(WebKit::WebViewImpl::prepareForMoveToWindow):
(WebKit::WebViewImpl::notifyInputContextAboutDiscardedComposition):
(WebKit::WebViewImpl::pressureChangeWithEvent):
(WebKit::WebViewImpl::fullScreenWindowController):
(WebKit::WebViewImpl::executeSavedCommandBySelector):
(WebKit::WebViewImpl::validRequestorForSendAndReturnTypes):
(WebKit::WebViewImpl::handleRequestedCandidates):
(WebKit::WebViewImpl::preferencesDidChange):
(WebKit::WebViewImpl::setTextIndicator):
(WebKit::WebViewImpl::dismissContentRelativeChildWindowsFromViewOnly):
(WebKit::WebViewImpl::accessibilityRegisterUIProcessTokens):
(WebKit::WebViewImpl::sendToolTipMouseExited):
(WebKit::WebViewImpl::sendToolTipMouseEntered):
(WebKit::WebViewImpl::toolTipChanged):
(WebKit::WebViewImpl::setAcceleratedCompositingRootLayer):
(WebKit::WebViewImpl::updateThumbnailViewLayer):
(WebKit::WebViewImpl::setInspectorAttachmentView):
(WebKit::WebViewImpl::inspectorAttachmentView):
(WebKit::WebViewImpl::draggedImage):
(WebKit::WebViewImpl::draggingEntered):
(WebKit::WebViewImpl::draggingUpdated):
(WebKit::WebViewImpl::draggingExited):
(WebKit::WebViewImpl::performDragOperation):
(WebKit::WebViewImpl::hitTestForDragTypes):
(WebKit::WebViewImpl::startWindowDrag):
(WebKit::WebViewImpl::dragImageForView):
(WebKit::WebViewImpl::setPromisedDataForImage):
(WebKit::WebViewImpl::setPromisedDataForAttachment):
(WebKit::WebViewImpl::takeViewSnapshot):
(WebKit::WebViewImpl::scrollWheel):
(WebKit::WebViewImpl::magnifyWithEvent):
(WebKit::WebViewImpl::rotateWithEvent):
(WebKit::WebViewImpl::doneWithKeyEvent):
(WebKit::WebViewImpl::interpretKeyEvent):
(WebKit::WebViewImpl::firstRectForCharacterRange):
(WebKit::WebViewImpl::characterIndexForPoint):
(WebKit::WebViewImpl::performKeyEquivalent):
(WebKit::WebViewImpl::keyUp):
(WebKit::WebViewImpl::keyDown):
(WebKit::WebViewImpl::mouseMoved):
(WebKit::WebViewImpl::windowIsFrontWindowUnderMouse):
(WebKit::WebViewImpl::userInterfaceLayoutDirection):

Tools:

Adds a new unit test checking that delayed event processing in text input context does not cause unwanted behaviors.
See WebKit ChangeLog for more details.

  • TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
  • TestWebKitAPI/Tests/TestWebKitAPI/mac/AppKitSPI.h: Renamed from Tools/TestWebKitAPI/Tests/TestWebKitAPI/mac/NSTextInputClientSPI.h.
  • TestWebKitAPI/Tests/mac/WKWebViewMacEditingTests.mm: Renamed from Tools/TestWebKitAPI/Tests/mac/WKWebViewSelectionTests.mm.

(-[SlowTextInputContext handleEventByInputMethod:completionHandler:]):
(-[SlowTextInputContext handleEvent:completionHandler:]):
(-[SlowInputWebView inputContext]):
(-[SlowInputWebView _web_superInputContext]):
(TEST):

1:42 PM WebKitGTK/Gardening/Calendar edited by cturner@igalia.com
(diff)
12:58 PM Changeset in webkit [220417] by Ryan Haddad
  • 2 edits in trunk/LayoutTests

Mark fast/images/async-image-body-background-image.html as flaky on macOS.
​https://bugs.webkit.org/show_bug.cgi?id=175193

Unreviewed test gardening.

  • platform/mac/TestExpectations:
12:44 PM Changeset in webkit [220416] by fpizlo@apple.com
  • 14 edits in trunk/Source

Baseline JIT should do caging
​https://bugs.webkit.org/show_bug.cgi?id=175037

Reviewed by Mark Lam.
Source/bmalloc:


This centralizes the notion of permanently enabling the primitive gigacage, which we only do in jsc
and WebProcess.

This saves the baseline JIT from emitting some code. Otherwise it would always have to emit enabled
checks on each typed array access.

  • bmalloc/Gigacage.cpp:

(Gigacage::primitiveGigacageDisabled):
(Gigacage::disableDisablingPrimitiveGigacageIfShouldBeEnabled):
(Gigacage::isDisablingPrimitiveGigacageDisabled):

  • bmalloc/Gigacage.h:

(Gigacage::isPrimitiveGigacagePermanentlyEnabled):
(Gigacage::canPrimitiveGigacageBeDisabled):

Source/JavaScriptCore:


Adds a AssemblyHelpers::cage and cageConditionally. Uses it in the baseline JIT.

Also modifies FTL caging to be more defensive when caging is disabled.

Relanded with fixed AssemblyHelpers::cageConditionally().

  • bytecode/AccessCase.cpp:

(JSC::AccessCase::generateImpl):

  • bytecode/InlineAccess.cpp:

(JSC::InlineAccess::dumpCacheSizesAndCrash):
(JSC::InlineAccess::generateSelfPropertyAccess):
(JSC::InlineAccess::generateSelfPropertyReplace):
(JSC::InlineAccess::generateArrayLength):

  • ftl/FTLLowerDFGToB3.cpp:

(JSC::FTL::DFG::LowerDFGToB3::caged):

  • jit/AssemblyHelpers.h:

(JSC::AssemblyHelpers::cage):
(JSC::AssemblyHelpers::cageConditionally):

  • jit/JITPropertyAccess.cpp:

(JSC::JIT::emitDoubleLoad):
(JSC::JIT::emitContiguousLoad):
(JSC::JIT::emitArrayStorageLoad):
(JSC::JIT::emitGenericContiguousPutByVal):
(JSC::JIT::emitArrayStoragePutByVal):
(JSC::JIT::emit_op_get_from_scope):
(JSC::JIT::emit_op_put_to_scope):
(JSC::JIT::emitIntTypedArrayGetByVal):
(JSC::JIT::emitFloatTypedArrayGetByVal):
(JSC::JIT::emitIntTypedArrayPutByVal):
(JSC::JIT::emitFloatTypedArrayPutByVal):

  • jsc.cpp:

(jscmain):
(primitiveGigacageDisabled): Deleted.

Source/WebKit:


Use a better API to disable disabling the primitive gigacage.

  • WebProcess/WebProcess.cpp:

(WebKit::m_webSQLiteDatabaseTracker):
(WebKit::primitiveGigacageDisabled): Deleted.

Source/WTF:

  • wtf/Gigacage.h:

(Gigacage::disableDisablingPrimitiveGigacageIfShouldBeEnabled):
(Gigacage::isDisablingPrimitiveGigacageDisabled):
(Gigacage::isPrimitiveGigacagePermanentlyEnabled):
(Gigacage::canPrimitiveGigacageBeDisabled):
(Gigacage::basePtr):

12:16 PM Changeset in webkit [220415] by Yusuke Suzuki
  • 2 edits in trunk/JSTests

Unreviewed, gardening test262 for Promise resolve / reject function length
​https://bugs.webkit.org/show_bug.cgi?id=175333

  • test262.yaml:
12:03 PM Changeset in webkit [220414] by commit-queue@webkit.org
  • 7 edits in trunk

Change fast seek logic to prevent ping-ponging.
​https://bugs.webkit.org/show_bug.cgi?id=175167
rdar://problem/32936116

Patch by Jeremy Jones <​jeremyj@apple.com> on 2017-08-08
Reviewed by Jer Noble.

Source/WebCore:

Updated test: media/video-fast-seek.html

Changed fast seek behavior to always seek early than seek time; and after current time when seeking forward.
This prevents ping-ponging, where a fast seek overshoots the target looking for a keyframe.
A subsequent seek will suddenly be on the other side of current time and
will search in the opposite direction for a key frame; repeat.

  • html/HTMLMediaElement.cpp:

(WebCore::HTMLMediaElement::fastSeek):

LayoutTests:

Test now requires forward fast seek result to be between current time and the target time,
and a backward seek to be before the seek time.

  • media/video-fast-seek.html:
  • media/media-source/media-source-fastseek.html:
11:20 AM WebKitGTK/Gardening/Calendar edited by cturner@igalia.com
(diff)
11:17 AM WebKitGTK/Gardening/Calendar edited by cturner@igalia.com
(diff)
10:22 AM Changeset in webkit [220413] by commit-queue@webkit.org
  • 2 edits in trunk/LayoutTests

Remove obsolete failure expectations for event_loadedmetadata.html.
​https://bugs.webkit.org/show_bug.cgi?id=175027

Unreviewed test gardening.

Patch by Ms2ger <​Ms2ger@igalia.com> on 2017-08-08

10:18 AM Changeset in webkit [220412] by wilander@apple.com
  • 20 edits
    2 adds in trunk

Resource Load Statistics: Clear web processes' state on history removal
​https://bugs.webkit.org/show_bug.cgi?id=175271
<rdar://problem/33756560>

Reviewed by Chris Dumez.

Source/WebKit:

​https://bugs.webkit.org/show_bug.cgi?id=175090 introduced state
in the web process and that state needs to be cleared when the
UI process' statistics are cleared as part of history removal.

  • UIProcess/WebProcessPool.cpp:

(WebKit::WebProcessPool::clearResourceLoadStatistics):

  • UIProcess/WebProcessPool.h:
  • UIProcess/WebsiteData/WebsiteDataStore.cpp:

(WebKit::WebsiteDataStore::removeData):
(WebKit::WebsiteDataStore::clearResourceLoadStatisticsInWebProcesses):

  • UIProcess/WebsiteData/WebsiteDataStore.h:
  • WebProcess/WebProcess.cpp:

(WebKit::WebProcess::clearResourceLoadStatistics):

  • WebProcess/WebProcess.h:
  • WebProcess/WebProcess.messages.in:

Tools:

Test infrastructure for asynchronous tests of website data
removal for _WKWebsiteDataTypeResourceLoadStatistics.

  • WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
  • WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:

(WTR::InjectedBundle::didReceiveMessageToPage):

  • WebKitTestRunner/InjectedBundle/TestRunner.cpp:

(WTR::TestRunner::statisticsClearThroughWebsiteDataRemoval):
(WTR::TestRunner::statisticsCallClearThroughWebsiteDataRemovalCallback):

  • WebKitTestRunner/InjectedBundle/TestRunner.h:
  • WebKitTestRunner/TestController.cpp:

(WTR::TestController::statisticsClearThroughWebsiteDataRemoval):

  • WebKitTestRunner/TestController.h:
  • WebKitTestRunner/TestInvocation.cpp:

(WTR::TestInvocation::didReceiveSynchronousMessageFromInjectedBundle):
(WTR::TestInvocation::didClearStatisticsThroughWebsiteDataRemoval):

  • WebKitTestRunner/TestInvocation.h:
  • WebKitTestRunner/cocoa/TestControllerCocoa.mm:

(WTR::TestController::statisticsClearThroughWebsiteDataRemoval):

LayoutTests:

  • http/tests/loading/resourceLoadStatistics/user-interaction-reported-after-website-data-removal-expected.txt: Added.
  • http/tests/loading/resourceLoadStatistics/user-interaction-reported-after-website-data-removal.html: Added.
  • platform/mac-wk2/TestExpectations:

user-interaction-reported-after-website-data-removal.html marked as [ Pass ].

10:15 AM Changeset in webkit [220411] by commit-queue@webkit.org
  • 2 edits in trunk/Source/WebCore

Unreviewed. Correct the function name in a debug message.
​https://bugs.webkit.org/show_bug.cgi?id=175262

No new tests: debug logs are not tested.

Patch by Ms2ger <​Ms2ger@igalia.com> on 2017-08-08

  • platform/audio/PlatformMediaSessionManager.cpp:

(WebCore::PlatformMediaSessionManager::applicationDidBecomeActive const):

10:14 AM Changeset in webkit [220410] by Ryan Haddad
  • 2 edits in trunk/LayoutTests

Fix baseline for http/tests/security/set-domain-remove-subdomain-for-ip-address.html.

Unreviewed test gardening.

  • http/tests/security/set-domain-remove-subdomain-for-ip-address-expected.txt:
10:00 AM Changeset in webkit [220409] by Matt Lewis
  • 1 edit
    1 move in trunk/LayoutTests

Fixed rebaselined expectations for js/dom/global-constructors-attributes.html.
​https://bugs.webkit.org/show_bug.cgi?id=175201

Unreviewed test gardening.

  • platform/mac-elcapitan-wk2/js/dom/global-constructors-attributes-expected.txt: Renamed from LayoutTests/platform/mac-elcapitan-wk2/js/dom/global-constructors-attributes-dedicated-worker-expected.txt.
9:49 AM Changeset in webkit [220408] by Adrian Perez de Castro
  • 4 edits
    1 move in trunk/Source/WebKit

[WPE] Implement WebsiteDataStore::defaultApplicationCacheDirectory() and friends
​https://bugs.webkit.org/show_bug.cgi?id=175322

Reviewed by Carlos Garcia Campos.

This makes WPE use the same implementation as the GTK+ port, with a preprocessor switch to choose the name of
the base directory inside the user XDG cache directory.

  • PlatformGTK.cmake: Add APIWebsiteDataStoreGLib.cpp to the build sources and remove APIWebsiteDataStoreGtk.cpp.
  • PlatformWPE.cmake: Add APIWebsiteDataStoreGLib.cpp to the built sources.
  • UIProcess/API/APIWebsiteDataStore.cpp: Remove now dead code.
  • UIProcess/API/glib/APIWebsiteDataStoreGLib.cpp: Renamed from Source/WebKit/UIProcess/API/gtk/APIWebsiteDataStoreGtk.cpp.

(API::WebsiteDataStore::defaultApplicationCacheDirectory):
(API::WebsiteDataStore::defaultIndexedDBDatabaseDirectory):
(API::WebsiteDataStore::defaultLocalStorageDirectory):
(API::WebsiteDataStore::defaultMediaKeysStorageDirectory):
(API::WebsiteDataStore::defaultWebSQLDatabaseDirectory):
(API::WebsiteDataStore::defaultResourceLoadStatisticsDirectory):

9:41 AM Changeset in webkit [220407] by Michael Catanzaro
  • 2 edits in trunk/Tools

Add Alicia Boya García as contributor
​https://bugs.webkit.org/show_bug.cgi?id=175326

Patch by Alicia Boya García <​aboya@igalia.com> on 2017-08-08
Reviewed by Michael Catanzaro.

  • Scripts/webkitpy/common/config/contributors.json:
9:24 AM Changeset in webkit [220406] by Brent Fulgham
  • 5 edits in trunk/LayoutTests

Unreviewed test results fix after r220376.

Rebaseline for new SecurityException message text.

  • fast/dom/Document/invalid-domain-change-throws-exception-expected.txt:
  • http/tests/dom/document-attributes-null-handling-expected.txt:
  • http/tests/security/set-domain-remove-subdomain-for-ip-address-expected.txt:
  • http/tests/security/xss-DENIED-invalid-domain-change-expected.txt:
9:13 AM Changeset in webkit [220405] by Brent Fulgham
  • 2 edits in trunk/Source/WebCore

Unreviewed build fix after r220376.

Don't attempt to use isPublicSuffix when building without that
feature enabled.

  • dom/Document.cpp:

(WebCore::Document::domainIsRegisterable const):

9:00 AM Changeset in webkit [220404] by Ryan Haddad
  • 14 edits in trunk/Source

Unreviewed, rolling out r220368.

This change caused WK1 tests to exit early with crashes.

Reverted changeset:

"Baseline JIT should do caging"
​https://bugs.webkit.org/show_bug.cgi?id=175037
​http://trac.webkit.org/changeset/220368

8:03 AM Changeset in webkit [220403] by Michael Catanzaro
  • 46 edits
    1 add in trunk

[CMake] Properly test if compiler supports compiler flags
​https://bugs.webkit.org/show_bug.cgi?id=174490

Reviewed by Konstantin Tokarev.

.:

This turned out to be a massive pain. I didn't want to merely check options before using
them: I also wanted to organize the code to avoid setting similar flags in different places.
Right now we set a bunch of global flags in OptionsCommon.cmake, and a bunch more flags in
WEBKIT_SET_EXTRA_COMPILER_FLAGS on a per-target basis.

Setting flags per-target seems better in general, e.g. because it makes it very easy to
disable warnings for particular ThirdParty targets. But it turns out that all the flags set
on a per-target basis get passed to both the C compiler and the C++ compiler, so it's
impossible to pass C++-only flags there. That's terrible. It's possible to make the flags
language-conditional using generator expressions, but that doesn't work for the Visual
Studio backend, so we would have to drop support for that (not going to happen). The CMake
documentation suggests that C and C++ files ought to be built in separate targets to avoid
this. It's a mess, basically.

So I've wound up removing WEBKIT_SET_EXTRA_COMPILER_FLAGS and adding most of those flags to
CMAKE_C_FLAGS and CMAKE_CXX_FLAGS instead. Really the only disadvantage of this is we now
have to suppress individual warnings when building ANGLESupport in WebCore. That's not the
end of the world. The only remaining useful feature of WEBKIT_SET_EXTRA_COMPILER_FLAGS was
to add -fPIC to static library targets, but turns out CMake does that for us if we just set
the variable CMAKE_POSITION_INDEPENDENT_CODE, so we can get rid of it completely.

Of course there are also macros for setting target-specific compiler flags, which we
frequently need in order to suppress specific warnings, particularly warnings coming from
third-party libraries like ANGLE and gtest. But remember the footgun: these macros will test
the flag against only one compiler, but must work with both C and C++ compilers unless the
build target exclusively contains targets built with just one of those compilers. Yuck.

  • CMakeLists.txt:
  • Source/CMakeLists.txt:
  • Source/PlatformGTK.cmake:
  • Source/cmake/OptionsCommon.cmake:
  • Source/cmake/WebKitCommon.cmake:
  • Source/cmake/WebKitCompilerFlags.cmake: Added.
  • Source/cmake/WebKitMacros.cmake:

Source/JavaScriptCore:

  • API/tests/PingPongStackOverflowTest.cpp:

(testPingPongStackOverflow):

  • API/tests/testapi.c:
  • b3/testb3.cpp:

(JSC::B3::testPatchpointLotsOfLateAnys):

Source/ThirdParty:

  • brotli/CMakeLists.txt:
  • gtest/CMakeLists.txt:
  • woff2/CMakeLists.txt:
  • xdgmime/CMakeLists.txt:

Source/WebCore:

  • CMakeLists.txt:
  • PlatformGTK.cmake:
  • PlatformWPE.cmake:

Source/WebDriver:

  • WebDriverService.cpp:

(WebDriver::WebDriverService::run):

  • glib/SessionHostGlib.cpp:

Source/WebKit:

  • CMakeLists.txt:
  • PlatformGTK.cmake:

Source/WTF:

  • wtf/Compiler.h:

Tools:

  • DumpRenderTree/TestNetscapePlugIn/CMakeLists.txt:
  • MiniBrowser/gtk/CMakeLists.txt:
  • TestRunnerShared/Bindings/JSWrapper.cpp:

(WTR::JSWrapper::initialize):

  • TestWebKitAPI/CMakeLists.txt:
  • TestWebKitAPI/PlatformGTK.cmake:
  • TestWebKitAPI/Tests/WTF/CheckedArithmeticOperations.cpp:

(TestWebKitAPI::CheckedArithmeticTester::run):

  • TestWebKitAPI/Tests/WebKitGLib/TestAutomationSession.cpp:
  • TestWebKitAPI/Tests/WebKitGLib/TestWebExtensions.cpp:
  • TestWebKitAPI/Tests/WebKitGLib/WebExtensionTest.cpp:

(formControlsAssociatedCallback):

  • TestWebKitAPI/glib/CMakeLists.txt:
  • TestWebKitAPI/glib/WebKitGLib/TestMain.h:

(Test::getResourcesDir):

  • WebKitTestRunner/CMakeLists.txt:
  • WebKitTestRunner/InjectedBundle/EventSendingController.cpp:

(WTR::menuItemClickCallback):
(WTR::staticConvertMenuItemToType):

  • WebKitTestRunner/InjectedBundle/TestRunner.cpp:

(WTR::TestRunner::setUseDashboardCompatibilityMode):

  • WebKitTestRunner/InjectedBundle/atk/AccessibilityNotificationHandlerAtk.cpp:

(WTR::AccessibilityNotificationHandler::disconnectAccessibilityCallbacks):

  • WebKitTestRunner/InjectedBundle/atk/AccessibilityUIElementAtk.cpp:

(WTR::AccessibilityUIElement::helpText const):
(WTR::AccessibilityUIElement::attributedStringForRange):

  • WebKitTestRunner/gtk/EventSenderProxyGtk.cpp:

(WTR::EventSenderProxy::updateTouchPoint):
(WTR::EventSenderProxy::releaseTouchPoint):

6:29 AM Changeset in webkit [220402] by zandobersek@gmail.com
  • 3 edits in trunk/Source/WebCore

[GStreamer] Don't use GraphicsContext3D in VideoTextureCoperGStreamer
​https://bugs.webkit.org/show_bug.cgi?id=174774

Reviewed by Carlos Garcia Campos.

VideoTextureCoperGStreamer class creates a new GraphicsContext3D internally
that's used to render to whichever OpenGL context is current. Such usage
doesn't add anything to the user since there's no underlying offscreen-based
OpenGL context created, and instead all GraphicsContext3D calls are directly
translated to usual OpenGL API calls. We should avoid using GraphicsContext3D
in such cases and instead use direct OpenGL API calls.

This patch achieves that for the VideoTextureCoperGStreamer class. A
GraphicsContext3D object is still created because it's needed to construct an
instance of the TextureMapperShaderProgram class, but this will be removed as
soon as TextureMapperShaderProgram drops the GraphicsContext3D dependence.

No new tests -- no change in behavior.

  • platform/graphics/gstreamer/VideoTextureCopierGStreamer.cpp:

(WebCore::VideoTextureCopierGStreamer::VideoTextureCopierGStreamer):
(WebCore::VideoTextureCopierGStreamer::~VideoTextureCopierGStreamer):
(WebCore::VideoTextureCopierGStreamer::copyVideoTextureToPlatformTexture):

  • platform/graphics/gstreamer/VideoTextureCopierGStreamer.h:

(WebCore::VideoTextureCopierGStreamer::resultTexture):

6:11 AM Changeset in webkit [220401] by Yusuke Suzuki
  • 2 edits in trunk/Source/JavaScriptCore

[Linux] Clear WasmMemory with madvice instead of memset
​https://bugs.webkit.org/show_bug.cgi?id=175150

Reviewed by Filip Pizlo.

In Linux, zeroing pages with memset populates backing store.
Instead, we should use madvise with MADV_DONTNEED. It discards
pages. And if you access these pages, on-demand-zero-pages will
be shown.

We also commit grown pages in all OSes.

  • wasm/WasmMemory.cpp:

(JSC::Wasm::commitZeroPages):
(JSC::Wasm::Memory::create):
(JSC::Wasm::Memory::grow):

6:05 AM Changeset in webkit [220400] by zandobersek@gmail.com
  • 5 edits in trunk/Source/WebCore

[TexMap] Drop GC3D* type usage from TextureMapperPlatformLayer{Buffer,Proxy}
​https://bugs.webkit.org/show_bug.cgi?id=175314

Reviewed by Carlos Garcia Campos.

Switch to using OpenGL types in the TextureMapperPlatformLayerBuffer and
TextureMapperPlatformLayerProxy classes, instead of the equivalent GC3D*
types provided in the GraphicsTypes3D.h header.

No new tests -- no change in behavior.

  • platform/graphics/texmap/TextureMapperPlatformLayerBuffer.cpp:

(WebCore::TextureMapperPlatformLayerBuffer::TextureMapperPlatformLayerBuffer):
(WebCore::TextureMapperPlatformLayerBuffer::canReuseWithoutReset):

  • platform/graphics/texmap/TextureMapperPlatformLayerBuffer.h:
  • platform/graphics/texmap/TextureMapperPlatformLayerProxy.cpp:

(WebCore::TextureMapperPlatformLayerProxy::getAvailableBuffer):

  • platform/graphics/texmap/TextureMapperPlatformLayerProxy.h:
6:04 AM Changeset in webkit [220399] by zandobersek@gmail.com
  • 5 edits in trunk/Source/WebCore

[TexMap] Don't use GraphicsContext3D in ClipStack
​https://bugs.webkit.org/show_bug.cgi?id=174776

Reviewed by Carlos Garcia Campos.

Any GraphicsContext3D object that's passed to ClipStack methods is of the
render-to-current-context nature, meaning there's no internally owned GL
context that has to be properly handled and all calls are simply passed to
OpenGL APIs. We should drop such (non-)usage of GraphicsContext3D in favor
of direct OpenGL API invocations.

This patch covers TextureMapper's ClipStack. Call sites to the apply() and
applyIfNeeded() are modified to not pass a reference to any
GraphicsContext3D object. Internally, OpenGL API entrypoints and constants
are used instead of GraphicsContext3D invocations.

No new tests -- no change in behavior.

  • platform/graphics/texmap/BitmapTextureGL.cpp:

(WebCore::BitmapTextureGL::clearIfNeeded):
(WebCore::BitmapTextureGL::bindAsSurface):

  • platform/graphics/texmap/ClipStack.cpp:

(WebCore::ClipStack::apply):
(WebCore::ClipStack::applyIfNeeded):

  • platform/graphics/texmap/ClipStack.h:
  • platform/graphics/texmap/TextureMapperGL.cpp:

(WebCore::TextureMapperGL::bindDefaultSurface):
(WebCore::TextureMapperGL::beginScissorClip):
(WebCore::TextureMapperGL::beginClip):
(WebCore::TextureMapperGL::endClip):

5:22 AM Changeset in webkit [220398] by jfernandez@igalia.com
  • 3 edits
    16 adds in trunk

Not possible to remove the 'li' element inside the table cell
​https://bugs.webkit.org/show_bug.cgi?id=173148

Reviewed by Ryosuke Niwa.

Source/WebCore:

We need to add a new case for breaking out empty list items when they are
at the start of an editable area. Since list items can be also inside
table cells, we need to consider this kind of elements as well.

Tests: editing/deleting/delete-list-items-in-table-cell-1.html

editing/deleting/delete-list-items-in-table-cell-2.html
editing/deleting/delete-list-items-in-table-cell-3.html
editing/deleting/delete-list-items-in-table-cell-4.html
editing/deleting/delete-list-items-in-table-cell-5.html
editing/deleting/delete-list-items-in-table-cell-6.html
editing/deleting/delete-list-items-in-table-cell-7.html
editing/deleting/delete-list-items-in-table-cell-8.html

  • editing/TypingCommand.cpp:

(WebCore::TypingCommand::deleteKeyPressed):

LayoutTests:

Regression tests for different scenarios of list items removal.

  • editing/deleting/delete-list-items-in-table-cell-1-expected.txt: Added.
  • editing/deleting/delete-list-items-in-table-cell-1.html: Added.
  • editing/deleting/delete-list-items-in-table-cell-2-expected.txt: Added.
  • editing/deleting/delete-list-items-in-table-cell-2.html: Added.
  • editing/deleting/delete-list-items-in-table-cell-3-expected.txt: Added.
  • editing/deleting/delete-list-items-in-table-cell-3.html: Added.
  • editing/deleting/delete-list-items-in-table-cell-4-expected.txt: Added.
  • editing/deleting/delete-list-items-in-table-cell-4.html: Added.
  • editing/deleting/delete-list-items-in-table-cell-5-expected.txt: Added.
  • editing/deleting/delete-list-items-in-table-cell-5.html: Added.
  • editing/deleting/delete-list-items-in-table-cell-6-expected.txt: Added.
  • editing/deleting/delete-list-items-in-table-cell-6.html: Added.
  • editing/deleting/delete-list-items-in-table-cell-7-expected.txt: Added.
  • editing/deleting/delete-list-items-in-table-cell-7.html: Added.
  • editing/deleting/delete-list-items-in-table-cell-8-expected.txt: Added.
  • editing/deleting/delete-list-items-in-table-cell-8.html: Added.
3:27 AM Changeset in webkit [220397] by zandobersek@gmail.com
  • 9 edits
    1 add in trunk/Source/WebCore

[TexMap] Isolate the TextureMapperPlatformLayerProxyProvider class
​https://bugs.webkit.org/show_bug.cgi?id=175316

Reviewed by Carlos Garcia Campos.

Move the TextureMapperPlatformLayerProxyProvider class (which is the type
aliased to PlatformLayer for threaded CoordGraphics) into its own header
file. This prevents including the TextureMapperPlatformLayerProxy.h header
file in MediaPlayerPrivateGStreamerBase.h, avoiding spilling OpenGL types
and function declarations before the GStreamer GL headers include them
later in the MediaPlayerPrivateGStreamerBase implementation file.

In the MediaPlayerPrivateGStreamerBase.h header file, only the new header
is included, and a forward declaration of the TextureMapperPlatformLayerProxy
class is used. proxy() and swapBuffersIfNeeded() methods are moved into
the implementation file to avoid requiring the full definition of the
TextureMapperPlatformLayerProxy class.

Similar is done for the TextureMapperGC3DPlatformLayer class and the
Cairo-specific implementation of the ImageBufferData class. The
CoordinatedGraphicsLayer implementation file also gains an include of the
TextureMapperPlatformLayerProxyProvider.h header since it requires the full
definition of that class.

No new tests -- no change in behavior.

  • platform/graphics/cairo/ImageBufferCairo.cpp:

(WebCore::ImageBufferData::proxy const):

  • platform/graphics/cairo/ImageBufferDataCairo.h:
  • platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.cpp:

(WebCore::MediaPlayerPrivateGStreamerBase::updateTexture):
(WebCore::MediaPlayerPrivateGStreamerBase::proxy const):
(WebCore::MediaPlayerPrivateGStreamerBase::swapBuffersIfNeeded):

  • platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.h:
  • platform/graphics/texmap/TextureMapperGC3DPlatformLayer.cpp:
  • platform/graphics/texmap/TextureMapperGC3DPlatformLayer.h:
  • platform/graphics/texmap/TextureMapperPlatformLayerProxy.h:
  • platform/graphics/texmap/TextureMapperPlatformLayerProxyProvider.h: Added.
  • platform/graphics/texmap/coordinated/CoordinatedGraphicsLayer.cpp:
3:02 AM WebKitGTK/2.16.x edited by cturner@igalia.com
(diff)
2:45 AM WebKitGTK/2.16.x edited by magomez@igalia.com
(diff)
2:03 AM Changeset in webkit [220396] by zandobersek@gmail.com
  • 2 edits in trunk/Source/WebCore

Unreviewed. Follow-up to r220392 that fixes build on configurations
that disable USE(GSTREAMER_GL).

  • platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.cpp:

(WebCore::MediaPlayerPrivateGStreamerBase::pushTextureToCompositor):
When creating the BitmapTextureGL object, also provide an initialized
TextureMapperContextAttributes object.

1:36 AM Changeset in webkit [220395] by zandobersek@gmail.com
  • 2 edits
    1 add in trunk/Source/WebCore

[TexMap] Group GL header file inclusions in a single header file
​https://bugs.webkit.org/show_bug.cgi?id=175313

Reviewed by Carlos Garcia Campos.

Add a helper header file to the TextureMapper subsystem that includes
the OpenGL headers, properly addressing the build configuration to
include headers as provided by either libepoxy, the OpenGL ES library,
or the OpenGL library.

TextureMapperContextAttributes implementation file is the only one
that can leverage the header at the moment, but more will follow.

  • platform/graphics/texmap/TextureMapperContextAttributes.cpp:
  • platform/graphics/texmap/TextureMapperGLHeaders.h: Copied from Source/WebCore/platform/graphics/texmap/TextureMapperContextAttributes.cpp.
1:29 AM Changeset in webkit [220394] by Carlos Garcia Campos
  • 11 edits in trunk/Source

Web Automation: setUserInputForCurrentJavaScriptPrompt should fail if current dialog is not a prompt
​https://bugs.webkit.org/show_bug.cgi?id=175261

Reviewed by Brian Burg.

Source/WebDriver:

  • CommandResult.cpp:

(WebDriver::CommandResult::CommandResult): Handle ElementNotInteractable protocol error.

Source/WebKit:

According to the spec, send alert text command should fail if the current dialog is not a prompt. This patch
adds JavaScriptDialogType enum to API::AutomationSessionClient and a new virtual method to ask the client about
the type of the current dialog. WebAutomationSession::setUserInputForCurrentJavaScriptPrompt() uses the new
client method to check the type of the current dialog and fail in case it's not a prompt. Cocoa needs an
implementation, for now it always returns Prompt as the type to keep compatibility.

18.4 Send Alert Text.
​https://w3c.github.io/webdriver/webdriver-spec.html#send-alert-text

This fixes selenium test testSettingTheValueOfAnAlertThrows.

  • UIProcess/API/APIAutomationSessionClient.h:

(API::AutomationSessionClient::typeOfCurrentJavaScriptDialogOnPage):

  • UIProcess/API/glib/WebKitAutomationSession.cpp:
  • UIProcess/API/glib/WebKitWebView.cpp:

(webkitWebViewGetCurrentScriptDialogType):

  • UIProcess/API/glib/WebKitWebViewPrivate.h:
  • UIProcess/Automation/Automation.json:
  • UIProcess/Automation/WebAutomationSession.cpp:

(WebKit::WebAutomationSession::setUserInputForCurrentJavaScriptPrompt):

  • UIProcess/Cocoa/AutomationSessionClient.h:
  • UIProcess/Cocoa/AutomationSessionClient.mm:

(WebKit::AutomationSessionClient::typeOfCurrentJavaScriptDialogOnPage):

1:10 AM Changeset in webkit [220393] by Wenson Hsieh
  • 29 edits
    4 adds in trunk

[iOS WK2] WKWebView schedules nonstop layout after pressing cmb+b,i,u inside a contenteditable div
​https://bugs.webkit.org/show_bug.cgi?id=175116
<rdar://problem/28279301>

Reviewed by Darin Adler and Ryosuke Niwa.

Source/WebCore:

WebCore support for WebPage::editorState refactoring. See WebKit ChangeLogs for more detail.

Tests: EditorStateTests.TypingAttributesBold

EditorStateTests.TypingAttributesItalic
EditorStateTests.TypingAttributesUnderline
EditorStateTests.TypingAttributesTextAlignmentAbsoluteAlignmentOptions
EditorStateTests.TypingAttributesTextAlignmentStartEnd
EditorStateTests.TypingAttributesTextAlignmentDirectionalText
EditorStateTests.TypingAttributesTextColor
EditorStateTests.TypingAttributesMixedStyles
EditorStateTests.TypingAttributesLinkColor

  • css/StyleProperties.cpp:

(WebCore::StyleProperties::propertyAsColor const):
(WebCore::StyleProperties::propertyAsValueID const):

Introduces some helper functions in StyleProperties to convert CSS property values to Color or a CSSValueID.

  • css/StyleProperties.h:
  • editing/EditingStyle.cpp:

(WebCore::EditingStyle::hasStyle):

Pull out logic in selectionStartHasStyle that asks for a style TriState into EditingStyle::hasStyle. This is
because WebPage::editorState will now query for multiple styles at the selection start, but
selectionStartHasStyle currently recomputes styleAtSelectionStart every time it is called. To prevent extra work
from being done, we can just call selectionStartHasStyle once and use ask for EditingStyle::hasStyle on the
computed EditingStyle at selection start.

  • editing/EditingStyle.h:
  • editing/Editor.cpp:

(WebCore::Editor::selectionStartHasStyle const):

Source/WebKit:

Refactors WebPage::editorState to only use the StyleProperties derived from EditingStyle, instead of inserting
and removing a temporary node to figure out the style. Also adds hooks to notify the UI delegate of EditorState
changes.

  • UIProcess/API/Cocoa/WKUIDelegatePrivate.h:
  • UIProcess/API/Cocoa/WKWebView.mm:

(nsTextAlignment):
(dictionaryRepresentationForEditorState):
(-[WKWebView _didChangeEditorState]):

Alerts the private UI delegate of UI-side EditorState updates.

(-[WKWebView _web_editorStateDidChange]):
(-[WKWebView _executeEditCommand:argument:completion:]):

  • UIProcess/API/Cocoa/WKWebViewInternal.h:
  • UIProcess/API/Cocoa/WKWebViewPrivate.h:
  • UIProcess/API/mac/WKView.mm:

(-[WKView _web_editorStateDidChange]):

  • UIProcess/Cocoa/WebViewImpl.h:
  • UIProcess/Cocoa/WebViewImpl.mm:

(WebKit::WebViewImpl::selectionDidChange):

  • UIProcess/WebPageProxy.cpp:

(WebKit::WebPageProxy::executeEditCommand):

Change executeEditCommand(name, callback) to executeEditCommand(name, argument, callback) and lift out of iOS
platform code and into WebPage.cpp.

  • UIProcess/WebPageProxy.h:
  • UIProcess/ios/WKContentViewInteraction.mm:

(-[WKContentView executeEditCommandWithCallback:]):
(-[WKContentView _selectionChanged]):

  • UIProcess/ios/WebPageProxyIOS.mm:

(WebKit::WebPageProxy::executeEditCommand): Deleted.

Move the iOS-specific implementation of executeEditCommand that invokes a callback when the web process responds
out of WebPageProxyIOS, and into cross-platform WebPageProxy code. Additionally, add a parameter for the edit
command's argument.

  • WebProcess/WebPage/WebPage.cpp:

(WebKit::WebPage::editorState const):

Use EditingStyle::styleAtSelectionStart instead of Editor::styleForSelectionStart when computing an EditorState.
Tweak bold, italic and underline to use EditingStyle TriStates.

(WebKit::shouldEnsureEditorStateUpdateAfterExecutingCommand):
(WebKit::WebPage::executeEditCommandWithCallback):

  • WebProcess/WebPage/WebPage.h:
  • WebProcess/WebPage/WebPage.messages.in:
  • WebProcess/WebPage/ios/WebPageIOS.mm:

(WebKit::WebPage::executeEditCommandWithCallback): Deleted.

Tools:

Introduces new testing infrastructure and API tests to test EditorState updates in the UI process. The new
EditorStateTests run on both iOS and Mac.

  • TestWebKitAPI/EditingTestHarness.h: Added.
  • TestWebKitAPI/EditingTestHarness.mm: Added.

EditingTestHarness is a helper object that API tests may use to apply editing commands and store EditorState
history. This test harness adds sugaring around various editing commands, and simplifies the process of checking
the state of the latest observed EditorState.

(-[EditingTestHarness initWithWebView:]):
(-[EditingTestHarness dealloc]):
(-[EditingTestHarness webView]):
(-[EditingTestHarness latestEditorState]):
(-[EditingTestHarness editorStateHistory]):
(-[EditingTestHarness insertText:andExpectEditorStateWith:]):
(-[EditingTestHarness insertHTML:andExpectEditorStateWith:]):
(-[EditingTestHarness selectAllAndExpectEditorStateWith:]):
(-[EditingTestHarness moveBackwardAndExpectEditorStateWith:]):
(-[EditingTestHarness moveWordBackwardAndExpectEditorStateWith:]):
(-[EditingTestHarness toggleBold]):
(-[EditingTestHarness toggleItalic]):
(-[EditingTestHarness toggleUnderline]):
(-[EditingTestHarness setForegroundColor:]):
(-[EditingTestHarness alignJustifiedAndExpectEditorStateWith:]):
(-[EditingTestHarness alignLeftAndExpectEditorStateWith:]):
(-[EditingTestHarness alignCenterAndExpectEditorStateWith:]):
(-[EditingTestHarness alignRightAndExpectEditorStateWith:]):
(-[EditingTestHarness insertParagraphAndExpectEditorStateWith:]):
(-[EditingTestHarness deleteBackwardAndExpectEditorStateWith:]):
(-[EditingTestHarness _execCommand:argument:expectEntries:]):

Dispatches an editing command to the web process, and blocks until a response is received. If an expected
entries dictionary is given, this will additionally verify that the latest EditorState contains all the expected
keys and values.

(-[EditingTestHarness latestEditorStateContains:]):
(-[EditingTestHarness _webView:editorStateDidChange:]):

  • TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
  • TestWebKitAPI/Tests/WebKit2Cocoa/EditorStateTests.mm: Added.

(TestWebKitAPI::setUpEditorStateTestHarness):
(TestWebKitAPI::TEST):

  • TestWebKitAPI/Tests/WebKit2Cocoa/editor-state-test-harness.html: Added.

LayoutTests:

Rebaseline some iOS WK2 LayoutTest expectations. These tests currently expect an empty anonymous RenderBlock to
be inserted into the render tree, but this is only a result of us adding and removing a temporary <span> when
computing a RenderStyle in WebPage::editorState -- this patch removes these empty RenderBlocks, making these
expectations' RenderTrees consistent with WebKit1.

  • platform/ios-wk2/editing/inserting/insert-div-024-expected.txt:
  • platform/ios-wk2/editing/inserting/insert-div-026-expected.txt:
  • platform/ios-wk2/editing/style/5084241-expected.txt:
  • platform/ios-wk2/editing/style/unbold-in-bold-expected.txt:
12:46 AM Changeset in webkit [220392] by zandobersek@gmail.com
  • 8 edits
    2 adds in trunk/Source/WebCore

[TexMap] Add TextureMapperContextAttributes
​https://bugs.webkit.org/show_bug.cgi?id=175311

Reviewed by Carlos Garcia Campos.

Add and use TextureMapperContextAttributes, a slim struct that holds
information about the type and extensions supported by the OpenGL
context that's being used for one specific TextureMapperGL instance.

This struct is initialized in the TextureMapperGL constructor for the
OpenGL context that's been made current on that thread, and that will
be used for that TextureMapperGL instance through its lifetime. The
struct is then also copied into any BitmapTextureGL objects that have
been created through this TextureMapperGL (or its texture pool).

The struct is initialized with information about GLES2 support (which
is at this point done through the USE_OPENGL_ES_2 macro) and about
specific OpenGL extensions that are supported. These are then used in
TextureMapperGL (NPOT textures) and BitmapTextureGL (BGRA support,
sub-image unpacking) to deploy specific optimizations or workarounds.

This ultimately serves as a replacement for clunky static functions
that need to operate on GraphicsContext3D objects to retrieve this
information.

No new tests -- no change in behavior.

  • platform/TextureMapper.cmake:
  • platform/graphics/texmap/BitmapTextureGL.cpp:

(WebCore::BitmapTextureGL::BitmapTextureGL):
(WebCore::BitmapTextureGL::updateContentsNoSwizzle):
(WebCore::BitmapTextureGL::updateContents):
(WebCore::driverSupportsSubImage): Deleted.

  • platform/graphics/texmap/BitmapTextureGL.h:

(WebCore::BitmapTextureGL::create):

  • platform/graphics/texmap/BitmapTexturePool.cpp:

(WebCore::BitmapTexturePool::BitmapTexturePool):
(WebCore::BitmapTexturePool::createTexture):

  • platform/graphics/texmap/BitmapTexturePool.h:
  • platform/graphics/texmap/TextureMapperContextAttributes.cpp: Added.

(WebCore::TextureMapperContextAttributes::initialize):

  • platform/graphics/texmap/TextureMapperContextAttributes.h: Added.
  • platform/graphics/texmap/TextureMapperGL.cpp:

(WebCore::TextureMapperGL::TextureMapperGL):
(WebCore::TextureMapperGL::drawTexture):
(WebCore::TextureMapperGL::drawTexturedQuadWithProgram):
(WebCore::TextureMapperGL::createTexture):
(WebCore::driverSupportsNPOTTextures): Deleted.

  • platform/graphics/texmap/TextureMapperGL.h:
12:44 AM Changeset in webkit [220391] by zandobersek@gmail.com
  • 6 edits in trunk/Source

[TexMap] Don't expose GraphicsContext3D object
​https://bugs.webkit.org/show_bug.cgi?id=175310

Reviewed by Carlos Garcia Campos.

Source/WebCore:

Remove the GraphicsContext3D getter from the TextureMapperGL class. Instead,
the clearColor() method is added that's to be used by the CoordinatedGraphicsScene
class which was accessing the GraphicsContext3D object for this purpose.

  • platform/graphics/texmap/TextureMapper.h:
  • platform/graphics/texmap/TextureMapperGL.cpp:

(WebCore::TextureMapperGL::clearColor):

  • platform/graphics/texmap/TextureMapperGL.h:

(WebCore::TextureMapperGL::graphicsContext3D const): Deleted.

Source/WebKit:

  • Shared/CoordinatedGraphics/CoordinatedGraphicsScene.cpp:

(WebKit::CoordinatedGraphicsScene::paintToCurrentGLContext):
Stop using GraphicsContext3D directly and instead use the
TextureMapper::clearColor() method to achieve the same result.

12:29 AM Changeset in webkit [220390] by jmarcell@apple.com
  • 2 edits in branches/safari-604-branch/Source/WebCore

Cherry-pick r220248. rdar://problem/33754458

12:29 AM Changeset in webkit [220389] by jmarcell@apple.com
  • 4 edits in branches/safari-604-branch/Source/WebCore

Cherry-pick r220247. rdar://problem/33754443

Aug 7, 2017:

11:43 PM Changeset in webkit [220388] by Carlos Garcia Campos
  • 7 edits in trunk/Source/WebDriver

WebDriver: implement unhandled prompt behavior
​https://bugs.webkit.org/show_bug.cgi?id=175184

Reviewed by Brian Burg.

Handle user prompts before running some of the commands according to the specification.

  • Capabilities.h: Add UnhandledPromptBehavior capability.
  • CommandResult.cpp:

(WebDriver::CommandResult::httpStatusCode const): Add UnexpectedAlertOpen error.
(WebDriver::CommandResult::errorString const): Ditto.

  • CommandResult.h:

(WebDriver::CommandResult::setAdditonalErrorData): New method to set an additional data object that will be sent
as part of the result error message.
(WebDriver::CommandResult::additionalErrorData const): Return the additional data object.

  • Session.cpp:

(WebDriver::Session::handleUserPrompts): Check if there's an active JavaScript dialog and deal with it depeding
on the unhandled prompt behavior.
(WebDriver::Session::reportUnexpectedAlertOpen): Generate an error message with UnexpectedAlertOpen error and
including the alert text as additional error data.
(WebDriver::Session::go): Handle user prompts before running the command.
(WebDriver::Session::getCurrentURL): Ditto.
(WebDriver::Session::back): Ditto.
(WebDriver::Session::forward): Ditto.
(WebDriver::Session::refresh): Ditto.
(WebDriver::Session::getTitle): Ditto.
(WebDriver::Session::closeWindow): Ditto.
(WebDriver::Session::switchToFrame): Ditto.
(WebDriver::Session::switchToParentFrame): Ditto.
(WebDriver::Session::isElementSelected): Ditto.
(WebDriver::Session::getElementText): Ditto.
(WebDriver::Session::getElementTagName): Ditto.
(WebDriver::Session::getElementRect): Ditto.
(WebDriver::Session::isElementEnabled): Ditto.
(WebDriver::Session::isElementDisplayed): Ditto.
(WebDriver::Session::getElementAttribute): Ditto.
(WebDriver::Session::elementSendKeys): Ditto.
(WebDriver::Session::elementSubmit): Ditto.
(WebDriver::Session::executeScript): Ditto.

  • Session.h:
  • WebDriverService.cpp:

(WebDriver::WebDriverService::sendResponse const): Send data object as part of the result error message if present.
(WebDriver::deserializeUnhandledPromptBehavior):
(WebDriver::WebDriverService::parseCapabilities const):
(WebDriver::WebDriverService::validatedCapabilities const):
(WebDriver::WebDriverService::newSession):

11:25 PM Changeset in webkit [220387] by Carlos Garcia Campos
  • 12 edits
    1 copy
    1 add in trunk/Source

[GTK] Implement JavaScript dialog methods of API::AutomationSessionClient
​https://bugs.webkit.org/show_bug.cgi?id=175259

Reviewed by Michael Catanzaro.

Source/WebCore/platform/gtk/po:

  • POTFILES.in: Add WebKitScriptDialogGtk.cpp.

Source/WebKit:

Move the default implementation of WebKitScriptDialog from WebKitWebView platform specific files to their own
files. Implement all JavaScript dialog methods of API::AutomationSessionClient in WebKitAutomationSession. For
now it only works when the user doesn't override WebKitWebView::script-dialog signal and default implementation
is used.

  • PlatformGTK.cmake: Add new files to compilation.
  • PlatformWPE.cmake: Ditto.
  • UIProcess/API/glib/WebKitAutomationSession.cpp:

(webkitAutomationSessionCreate): Pass the WebKitWebContext to the constructor and keep a pointer to it in session.

  • UIProcess/API/glib/WebKitAutomationSessionPrivate.h:
  • UIProcess/API/glib/WebKitScriptDialogPrivate.h:
  • UIProcess/API/glib/WebKitWebContext.cpp:
  • UIProcess/API/glib/WebKitWebView.cpp:

(webkitWebViewRunJavaScriptAlert): Set the currently script dialog for the scope of the function.
(webkitWebViewRunJavaScriptConfirm): Ditto.
(webkitWebViewRunJavaScriptPrompt): Ditto.
(webkitWebViewRunJavaScriptBeforeUnloadConfirm): Ditto.
(webkitWebViewIsShowingScriptDialog): Ask current dialog if there's one.
(webkitWebViewGetCurrentScriptDialogMessage): Ditto.
(webkitWebViewSetCurrentScriptDialogUserInput): Ditto.
(webkitWebViewAcceptCurrentScriptDialog): Ditto.
(webkitWebViewDismissCurrentScriptDialog): Ditto.

  • UIProcess/API/glib/WebKitWebViewPrivate.h:
  • UIProcess/API/gtk/WebKitScriptDialogGtk.cpp: Added.

(webkitWebViewCreateJavaScriptDialog): Moved from WebKitWebViewGtk.cpp.
(webkitScriptDialogRun): Run the dialog, this code is moved from WebKitWebViewGtk.cpp.
(webkitScriptDialogIsRunning): Return true if the script dialog has a native dialog running.
(webkitScriptDialogAccept): Send Ok or Close response to the native dialog depending on the dialog type.
(webkitScriptDialogDismiss): Send Close response to the native dialog.
(webkitScriptDialogSetUserInput): Set the given text on the prompt dialog entry.

  • UIProcess/API/gtk/WebKitWebViewGtk.cpp:

(webkitWebViewScriptDialog): Simply call webkitScriptDialogRun() now.

  • UIProcess/API/wpe/WebKitScriptDialogWPE.cpp: Copied from Source/WebKit/UIProcess/API/glib/WebKitAutomationSessionPrivate.h.

(webkitScriptDialogRun):
(webkitScriptDialogIsRunning):
(webkitScriptDialogAccept):
(webkitScriptDialogDismiss):
(webkitScriptDialogSetUserInput):

11:22 PM Changeset in webkit [220386] by Carlos Garcia Campos
  • 7 edits in trunk/Source/WebDriver

WebDriver: implement user prompt commands
​https://bugs.webkit.org/show_bug.cgi?id=174614

Reviewed by Brian Burg.

  • CommandResult.cpp:

(WebDriver::CommandResult::CommandResult): Handle NoJavaScriptDialog protocol error.
(WebDriver::CommandResult::httpStatusCode const): Add NoSuchAlert.
(WebDriver::CommandResult::errorString const): Ditto.

  • CommandResult.h:
  • Session.cpp:

(WebDriver::Session::dismissAlert):
(WebDriver::Session::acceptAlert):
(WebDriver::Session::getAlertText):
(WebDriver::Session::sendAlertText):

  • Session.h:
  • WebDriverService.cpp:

(WebDriver::WebDriverService::dismissAlert):
(WebDriver::WebDriverService::acceptAlert):
(WebDriver::WebDriverService::getAlertText):
(WebDriver::WebDriverService::sendAlertText):

  • WebDriverService.h:
10:48 PM Changeset in webkit [220385] by jmarcell@apple.com
  • 2 edits in branches/safari-604.1.38.1-branch/Source/WebCore

Cherry-pick r220333. rdar://problem/33601173

8:20 PM Changeset in webkit [220384] by msaboff@apple.com
  • 2 edits in trunk/PerformanceTests

Use more accurate Russian keywords for RexBench/FlightPlanner Unicode mode
​https://bugs.webkit.org/show_bug.cgi?id=175298

Reviewed by Saam Barati.

With the help of Alexey Proskuryakov updated the Russian keywords.

  • RexBench/FlightPlanner/use_unicode.js:
7:52 PM Changeset in webkit [220383] by Simon Fraser
  • 3 edits in trunk/Source/WebCore

RenderStyle:diff() was inadvertently doing deep compares of StyleRareNonInheritedData etc
​https://bugs.webkit.org/show_bug.cgi?id=175304

Reviewed by Tim Horton.

r210758 changed DataRef::get() from returning a pointer to a reference. This caused all the places
in RenderStyle::diff() and related functions, where we intended to do a quick pointer comparison,
to doing deep compares via operator!=. This made the code slightly slower.

Fix by exposing ptr() on DataRef and using it wherever we wish to do pointer comparison.

  • rendering/style/DataRef.h:

(WebCore::DataRef::ptr const):

  • rendering/style/RenderStyle.cpp:

(WebCore::RenderStyle::inheritedDataShared const):
(WebCore::RenderStyle::changeAffectsVisualOverflow const):
(WebCore::RenderStyle::changeRequiresLayout const):
(WebCore::RenderStyle::changeRequiresRecompositeLayer const):
(WebCore::RenderStyle::listStyleImage const): Expand the function onto multiple lines.

7:52 PM Changeset in webkit [220382] by Simon Fraser
  • 2 edits in trunk/Source/WebCore

Add a fast path for rotate() and rotateZ() transform parsing
​https://bugs.webkit.org/show_bug.cgi?id=175308

Reviewed by Zalan Bujtas.

Fast paths only existed for translate-related functions, matrix3d() and scale3d(). Add
rotate() and rotateX(), which gives a small boost to one of the MotionMark tests.

  • css/parser/CSSParserFastPaths.cpp:

(WebCore::parseSimpleAngle):
(WebCore::parseTransformAngleArgument):
(WebCore::parseSimpleTransformValue):
(WebCore::transformCanLikelyUseFastPath):

7:52 PM Changeset in webkit [220381] by Simon Fraser
  • 2 edits in trunk/Source/WebCore

Re-order the tests in RenderLayerCompositor::requiresCompositingLayer() for performance
​https://bugs.webkit.org/show_bug.cgi?id=175306

Reviewed by Tim Horton.

Re-order the tests for compositing reasons so that reasons more likely to happen are higher in the list.

  • rendering/RenderLayerCompositor.cpp:

(WebCore::RenderLayerCompositor::requiresCompositingLayer const):
(WebCore::RenderLayerCompositor::requiresOwnBackingStore const):

7:52 PM Changeset in webkit [220380] by Simon Fraser
  • 2 edits in trunk/Source/WebCore

Avoid a hash lookup in FilterInfo::remove()
​https://bugs.webkit.org/show_bug.cgi?id=175301

Reviewed by Sam Weinig.

FilterInfo::remove() always called map().remove(&layer)), even when layer.m_hasFilterInfo
was false (and even asserted that layer.m_hasFilterInfo == map().contains(&layer)).

So we can early return if layer.m_hasFilterInfo is false.

  • rendering/RenderLayerFilterInfo.cpp:

(WebCore::RenderLayer::FilterInfo::getIfExists):
(WebCore::RenderLayer::FilterInfo::remove):

7:46 PM Changeset in webkit [220379] by Simon Fraser
  • 14 edits in trunk/Source/WebCore

Make TransformOperation::type() non-virtual
​https://bugs.webkit.org/show_bug.cgi?id=175297

Reviewed by Sam Weinig.

Store the OperationType in the base class so that type() and isSameType() can
be non-virtual.

Small perf win on some benchmarks.

  • platform/graphics/transforms/IdentityTransformOperation.h:
  • platform/graphics/transforms/Matrix3DTransformOperation.h:
  • platform/graphics/transforms/MatrixTransformOperation.h:
  • platform/graphics/transforms/PerspectiveTransformOperation.h:
  • platform/graphics/transforms/RotateTransformOperation.cpp:

(WebCore::RotateTransformOperation::blend):

  • platform/graphics/transforms/RotateTransformOperation.h:
  • platform/graphics/transforms/ScaleTransformOperation.cpp:

(WebCore::ScaleTransformOperation::blend):

  • platform/graphics/transforms/ScaleTransformOperation.h:
  • platform/graphics/transforms/SkewTransformOperation.cpp:

(WebCore::SkewTransformOperation::blend):

  • platform/graphics/transforms/SkewTransformOperation.h:
  • platform/graphics/transforms/TransformOperation.h:

(WebCore::TransformOperation::TransformOperation):
(WebCore::TransformOperation::type const):
(WebCore::TransformOperation::isSameType const):

  • platform/graphics/transforms/TranslateTransformOperation.cpp:

(WebCore::TranslateTransformOperation::blend):

  • platform/graphics/transforms/TranslateTransformOperation.h:
7:46 PM Changeset in webkit [220378] by Simon Fraser
  • 4 edits in trunk/Source/WebCore

Inline ~Color and Color::isExtended()
​https://bugs.webkit.org/show_bug.cgi?id=175293

Reviewed by Zalan Bujtas.

The Color destructor and Color::isExtended() show up on profiles, so inline them.
Before r207265 the destructor was inlined.

Also make sure that LengthSize::operator== is inlined, which it was not (according
to profiles).

  • platform/LengthSize.h:

(WebCore::operator==):

  • platform/graphics/Color.cpp:

(WebCore::Color::~Color): Deleted.
(WebCore::Color::isExtended const): Deleted.

  • platform/graphics/Color.h:

(WebCore::Color::~Color):
(WebCore::Color::isExtended const):

7:29 PM Changeset in webkit [220377] by commit-queue@webkit.org
  • 3 edits
    1 add in trunk

GetOwnProperty of TypedArray indexed fields is wrongly configurable
​https://bugs.webkit.org/show_bug.cgi?id=175307

Patch by Robin Morisset <​rmorisset@apple.com> on 2017-08-07
Reviewed by Saam Barati.

JSTests:

  • stress/typedarray-getownproperty-not-configurable.js: Added.

(assert):
(foo):

Source/JavaScriptCore:

`
let a = new Uint8Array(10);
let b = Object.getOwnPropertyDescriptor(a, 0);
assert(b.configurable === false);
`
should not fail: by section 9.4.5.1 (​https://tc39.github.io/ecma262/#sec-integer-indexed-exotic-objects-getownproperty-p)
that applies to integer indexed exotic objects, and section 22.2.7 (​https://tc39.github.io/ecma262/#sec-properties-of-typedarray-instances)
that says that typed arrays are integer indexed exotic objects.

  • runtime/JSGenericTypedArrayViewInlines.h:

(JSC::JSGenericTypedArrayView<Adaptor>::getOwnPropertySlotByIndex):

7:28 PM Changeset in webkit [220376] by Brent Fulgham
  • 5 edits in trunk

Source/WebCore:
Implement the HTML5 same-origin restriction specification
​https://bugs.webkit.org/show_bug.cgi?id=175226
<rdar://problem/11079948>

Reviewed by Chris Dumez.

Follow the algorithms defined in the HTML5 specification for relaxing
the same-origin restriction. We were missing a few steps related to
checking for public suffix and presence of a browsing context.

Tested by new TestWebKitAPI tests.

  • dom/Document.cpp:

(WebCore::Document::domainIsRegisterable): Added helper function.
(WebCore::Document::setDomain):

Tools:
Prevent domain from being set to a TLD
​https://bugs.webkit.org/show_bug.cgi?id=175226
<rdar://problem/11079948>

Reviewed by Chris Dumez.

Extend the public suffix tests to include cases used by the
Public Domain 'Public Suffix List'.

  • TestWebKitAPI/Tests/mac/PublicSuffix.cpp:

(TestWebKitAPI::TEST):

6:05 PM Changeset in webkit [220375] by Matt Lewis
  • 2 edits in trunk/LayoutTests

Marked media/modern-media-controls/fullscreen-support/fullscreen-support-press.html as flaky.
​https://bugs.webkit.org/show_bug.cgi?id=173946

Unreviewed test gardening.

  • platform/mac-wk2/TestExpectations:
5:53 PM Changeset in webkit [220374] by Matt Lewis
  • 1 edit
    3 adds in trunk/LayoutTests

Rebaselined js/dom/global-constructors-attributes.html on El Capitan.
​https://bugs.webkit.org/show_bug.cgi?id=175201

Unreviewed test gardening.

  • platform/mac-elcapitan-wk2/js/dom/global-constructors-attributes-dedicated-worker-expected.txt: Added.
5:47 PM Changeset in webkit [220373] by commit-queue@webkit.org
  • 11 edits
    10 adds
    2 deletes in trunk/PerformanceTests

Speedometer: Update to modern production Ember version
​https://bugs.webkit.org/show_bug.cgi?id=175278

Patch by Mathias Bynens <​mathias@qiwi.be> on 2017-08-07
Reviewed by Ryosuke Niwa.

  • Speedometer/resources/todomvc/architecture-examples/emberjs/README.md: Added.
  • Speedometer/resources/todomvc/architecture-examples/emberjs/index.html: Made title consistent.
  • Speedometer/resources/todomvc/architecture-examples/emberjs/source/README.md: Documented build process.
  • Speedometer/resources/todomvc/architecture-examples/emberjs/source/app/components/.gitkeep: Added.
  • Speedometer/resources/todomvc/architecture-examples/emberjs/source/app/index.html: Made title consistent.
  • Speedometer/resources/todomvc/architecture-examples/emberjs/source/app/instance-initializers/global.js: Removed.
  • Speedometer/resources/todomvc/architecture-examples/emberjs/source/app/models/.gitkeep: Added.
  • Speedometer/resources/todomvc/architecture-examples/emberjs/source/app/router.js: Update to latest version.
  • Speedometer/resources/todomvc/architecture-examples/emberjs/source/app/routes/.gitkeep: Added.
  • Speedometer/resources/todomvc/architecture-examples/emberjs/source/app/services/repo.js: Update to latest version
  • Speedometer/resources/todomvc/architecture-examples/emberjs/source/app/styles/app.css: Added.
  • Speedometer/resources/todomvc/architecture-examples/emberjs/source/app/templates/components/.gitkeep: Added.
  • Speedometer/resources/todomvc/architecture-examples/emberjs/source/bower.json: Update to latest version.
  • Speedometer/resources/todomvc/architecture-examples/emberjs/source/config/environment.js: Update to latest version.
  • Speedometer/resources/todomvc/architecture-examples/emberjs/source/dist/*: Added generated files per build instructions.
  • Speedometer/resources/todomvc/architecture-examples/emberjs/source/package-lock.json: Pinned dependencies.
  • Speedometer/resources/todomvc/architecture-examples/emberjs/source/package.json: Update to latest version.
  • Speedometer/resources/todomvc/architecture-examples/emberjs/source/tests/.jshintrc: Removed.
  • Speedometer/resources/todomvc/architecture-examples/emberjs/source/tests/helpers/module-for-acceptance.js: Updated to latest version.
  • Speedometer/resources/todomvc/architecture-examples/emberjs/source/tests/helpers/start-app.js: Updated to latest version.
  • Speedometer/resources/todomvc/architecture-examples/emberjs/source/tests/index.html: Updated to latest version.
  • Speedometer/resources/todomvc/architecture-examples/emberjs/source/tests/integration/.gitkeep: Added.
  • Speedometer/resources/todomvc/architecture-examples/emberjs/source/tests/unit/.gitkeep: Added.
  • Speedometer/resources/todomvc/architecture-examples/emberjs/source/vendor/.gitkeep: Added.
5:25 PM Changeset in webkit [220372] by commit-queue@webkit.org
  • 8 edits
    1 add in trunk/Tools

[XCode] webkit-patch should run sort-Xcode-project-file
​https://bugs.webkit.org/show_bug.cgi?id=174036
<rdar://problem/33732709>

Patch by Stephan Szabo <​stephan.szabo@sony.com> on 2017-08-07
Reviewed by Simon Fraser.

  • Scripts/webkitpy/common/config/ports.py:
  • Scripts/webkitpy/tool/commands/download.py:
  • Scripts/webkitpy/tool/commands/download_unittest.py:
  • Scripts/webkitpy/tool/commands/upload.py:
  • Scripts/webkitpy/tool/commands/upload_unittest.py:
  • Scripts/webkitpy/tool/steps/init.py:
  • Scripts/webkitpy/tool/steps/options.py:
  • Scripts/webkitpy/tool/steps/sortxcodeprojectfiles.py: Added.
4:56 PM Changeset in webkit [220371] by Matt Lewis
  • 2 edits in trunk/LayoutTests

Changed expectations for webrtc/video-rotation.html.
​https://bugs.webkit.org/show_bug.cgi?id=175305

Unreviewed test gardening.

4:55 PM Changeset in webkit [220370] by Devin Rousso
  • 36 edits
    8 adds in trunk

Web Inspector: Preview Canvas path when viewing a recording
​https://bugs.webkit.org/show_bug.cgi?id=174967

Reviewed by Brian Burg.

Source/WebCore:

Tests: fast/canvas/2d.currentPoint.html

fast/canvas/2d.getPath.modification.html
fast/canvas/2d.getPath.newobject.html
fast/canvas/2d.setPath.html

  • html/canvas/CanvasPath.idl:
  • html/canvas/CanvasPath.h:
  • html/canvas/CanvasPath.cpp:

(WebCore::CanvasPath::currentX const):
(WebCore::CanvasPath::currentY const):

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

(WebCore::CanvasRenderingContext2D::setPath):
(WebCore::CanvasRenderingContext2D::getPath const):

  • page/RuntimeEnabledFeatures.h:

(WebCore::RuntimeEnabledFeatures::setInspectorAdditionsEnabled):
(WebCore::RuntimeEnabledFeatures::inspectorAdditionsEnabled const):
Add runtime flag for added IDL items above so that they are only usable within the inspector
process. The runtime flag is not enabled from anywhere else as of now.

  • inspector/InspectorCanvas.cpp:

(WebCore::InspectorCanvas::buildInitialState):
Send current path as part of the InitialState.
Drive-by: deduplicate more string values.

Source/WebInspectorUI:

  • Localizations/en.lproj/localizedStrings.js:
  • UserInterface/Base/Setting.js:
  • UserInterface/Views/RecordingContentView.js:

(WI.RecordingContentView):
(WI.RecordingContentView.supportsCanvasPathDebugging):
(WI.RecordingContentView.prototype.get navigationItems):
(WI.RecordingContentView.prototype.shown):
(WI.RecordingContentView.prototype._generateContentCanvas2D):
(WI.RecordingContentView.prototype._actionModifiesPath):
(WI.RecordingContentView.prototype._updateCanvasPath):
(WI.RecordingContentView.prototype._showPathButtonClicked):
Show each segment of the current path as an overlay when the setting is enabled.
Drive-by: fix forgotten function rename.

  • UserInterface/Views/RecordingContentView.css:

(.content-view:not(.tab).recording canvas.path):

  • UserInterface/Views/RecordingStateDetailsSidebarPanel.js:

(WI.RecordingStateDetailsSidebarPanel.prototype._generateDetailsCanvas2D):
Show the currentX/currentY in the current state.

  • UserInterface/Models/RecordingAction.js:

(WI.RecordingAction.isFunctionForType):
(WI.RecordingAction.prototype.swizzle):
Use Sets for better performance.

Source/WebKit:

  • Shared/WebPreferencesDefinitions.h:
  • UIProcess/API/C/WKPreferencesRefPrivate.h:
  • UIProcess/API/C/WKPreferences.cpp:

(WKPreferencesSetInspectorAdditionsEnabled):
(WKPreferencesGetInspectorAdditionsEnabled):

  • WebProcess/WebPage/WebPage.cpp:

(WebKit::WebPage::updatePreferences):
Add plumbing for new InspectorAdditions runtime flag.

  • WebProcess/WebPage/WebInspectorUI.cpp:

(WebKit::WebInspectorUI::WebInspectorUI):
Enable InspectorAdditions by default in the WebInspector page.

Source/WebKitLegacy/mac:

  • WebView/WebPreferenceKeysPrivate.h:
  • WebView/WebPreferences.mm:

(+[WebPreferences initialize]):
(-[WebPreferences inspectorAdditionsEnabled]):
(-[WebPreferences setInspectorAdditionsEnabled:]):

  • WebView/WebPreferencesPrivate.h:
  • WebView/WebView.mm:

(-[WebView _preferencesChanged:]):
Add plumbing for new InspectorAdditions runtime flag.

Tools:

  • DumpRenderTree/TestOptions.h:
  • DumpRenderTree/TestOptions.mm:

(TestOptions::TestOptions):

  • DumpRenderTree/mac/DumpRenderTree.mm:

(setWebPreferencesForTestOptions):

  • WebKitTestRunner/TestController.cpp:

(WTR::TestController::resetPreferencesToConsistentValues):
(WTR::updateTestOptionsFromTestHeader):

  • WebKitTestRunner/TestOptions.h:

(WTR::TestOptions::hasSameInitializationOptions const):
Add plumbing for new InspectorAdditions runtime flag.

LayoutTests:

  • fast/canvas/2d.currentPoint-expected.txt: Added.
  • fast/canvas/2d.currentPoint.html: Added.
  • fast/canvas/2d.getPath.modification-expected.txt: Added.
  • fast/canvas/2d.getPath.modification.html: Added.
  • fast/canvas/2d.getPath.newobject-expected.txt: Added.
  • fast/canvas/2d.getPath.newobject.html: Added.
  • fast/canvas/2d.setPath-expected.txt: Added.
  • fast/canvas/2d.setPath.html: Added.
  • inspector/canvas/recording-2d-expected.txt:
  • inspector/canvas/recording-2d.html:

Updated for additional deduplication in InitialState.

4:33 PM Changeset in webkit [220369] by commit-queue@webkit.org
  • 3 edits in trunk/Tools

Sorted EWS Queues on Patch page
​https://bugs.webkit.org/show_bug.cgi?id=173667

Patch by obinna obike <​oobike@apple.com> on 2017-08-07
Reviewed by Aakash Jain.

  • QueueStatusServer/handlers/patch.py:

(Patch.get): Sorted queue_status.

  • QueueStatusServer/templates/patch.html:

Changed queue_status.items to queue_status because it's a list.

4:30 PM Changeset in webkit [220368] by fpizlo@apple.com
  • 14 edits in trunk/Source

Baseline JIT should do caging
​https://bugs.webkit.org/show_bug.cgi?id=175037

Reviewed by Mark Lam.
Source/bmalloc:


This centralizes the notion of permanently enabling the primitive gigacage, which we only do in jsc
and WebProcess.

This saves the baseline JIT from emitting some code. Otherwise it would always have to emit enabled
checks on each typed array access.

  • bmalloc/Gigacage.cpp:

(Gigacage::primitiveGigacageDisabled):
(Gigacage::disableDisablingPrimitiveGigacageIfShouldBeEnabled):
(Gigacage::isDisablingPrimitiveGigacageDisabled):

  • bmalloc/Gigacage.h:

(Gigacage::isPrimitiveGigacagePermanentlyEnabled):
(Gigacage::canPrimitiveGigacageBeDisabled):

Source/JavaScriptCore:


Adds a AssemblyHelpers::cage and cageConditionally. Uses it in the baseline JIT.

Also modifies FTL caging to be more defensive when caging is disabled.

  • ftl/FTLLowerDFGToB3.cpp:

(JSC::FTL::DFG::LowerDFGToB3::caged):

  • jit/AssemblyHelpers.h:

(JSC::AssemblyHelpers::cage):
(JSC::AssemblyHelpers::cageConditionally):

  • jit/JITPropertyAccess.cpp:

(JSC::JIT::emitDoubleLoad):
(JSC::JIT::emitContiguousLoad):
(JSC::JIT::emitArrayStorageLoad):
(JSC::JIT::emitGenericContiguousPutByVal):
(JSC::JIT::emitArrayStoragePutByVal):
(JSC::JIT::emit_op_get_from_scope):
(JSC::JIT::emit_op_put_to_scope):
(JSC::JIT::emitIntTypedArrayGetByVal):
(JSC::JIT::emitFloatTypedArrayGetByVal):
(JSC::JIT::emitIntTypedArrayPutByVal):
(JSC::JIT::emitFloatTypedArrayPutByVal):

  • jsc.cpp:

(jscmain):
(primitiveGigacageDisabled): Deleted.

Source/WebKit:


Use a better API to disable disabling the primitive gigacage.

  • WebProcess/WebProcess.cpp:

(WebKit::m_webSQLiteDatabaseTracker):
(WebKit::primitiveGigacageDisabled): Deleted.

Source/WTF:

  • wtf/Gigacage.h:

(Gigacage::disableDisablingPrimitiveGigacageIfShouldBeEnabled):
(Gigacage::isDisablingPrimitiveGigacageDisabled):
(Gigacage::isPrimitiveGigacagePermanentlyEnabled):
(Gigacage::canPrimitiveGigacageBeDisabled):

4:25 PM Changeset in webkit [220367] by commit-queue@webkit.org
  • 3 edits in trunk/Tools

check-webkit-style: NS_ERROR_ENUM and NS_OPTIONS should not be handled as function definitions.
​https://bugs.webkit.org/show_bug.cgi?id=175286

Patch by Yoshiaki Jitsukawa <​Yoshiaki.Jitsukawa@sony.com> on 2017-08-07
Reviewed by Myles C. Maxfield.

  • Scripts/webkitpy/style/checkers/cpp.py:

(check_braces):

  • Scripts/webkitpy/style/checkers/cpp_unittest.py:

(WebKitStyleTest.test_braces.NS_ERROR_ENUM):
(WebKitStyleTest.test_braces):
(WebKitStyleTest.test_braces.NS_OPTIONS):

4:22 PM Changeset in webkit [220366] by Chris Dumez
  • 2 edits in trunk/Source/WebCore

Update sendBeacon() to rely on FetchBody instead of the whole FetchRequest
​https://bugs.webkit.org/show_bug.cgi?id=175280

Reviewed by Youenn Fablet.

Update sendBeacon() to rely on FetchBody instead of the whole FetchRequest. FetchBody
for data extraction is really the only thing we need at the moment.
The new code also properly sets the CORS mode, which will be needed for Bug 175264.

  • Modules/beacon/NavigatorBeacon.cpp:

(WebCore::NavigatorBeacon::sendBeacon):

4:18 PM Changeset in webkit [220365] by jmarcell@apple.com
  • 7 edits in branches/safari-604.1.38.0-branch/Source

Versioning.

4:16 PM Changeset in webkit [220364] by jmarcell@apple.com
  • 7 edits in branches/safari-604.1.38.1-branch/Source

Versioning.

4:12 PM Changeset in webkit [220363] by jmarcell@apple.com
  • 1 copy in tags/Safari-604.1.38.0.1

Tag Safari-604.1.38.0.1.

4:10 PM Changeset in webkit [220362] by jmarcell@apple.com
  • 1 copy in tags/Safari-604.1.38.1.1

Tag Safari-604.1.38.1.1.

3:43 PM Changeset in webkit [220361] by msaboff@apple.com
  • 2 edits in trunk/PerformanceTests

REGRESSION(r220307): Perf bot failure trying to run RexBench tests
​https://bugs.webkit.org/show_bug.cgi?id=175289

Reviewed by Filip Pizlo.

Add RexBench to the list of benchmarks we skip on the perf bots.

  • Skipped:
3:31 PM Changeset in webkit [220360] by Jon Davis
  • 2 edits in trunk/Websites/webkit.org

Fixed superscript rendering for blog posts
​https://bugs.webkit.org/show_bug.cgi?id=175285

Reviewed by Filip Pizlo.

  • wp-content/themes/webkit/style.css:

(sup): Added.

3:26 PM Changeset in webkit [220359] by matthew_hanson@apple.com
  • 7 edits in branches/safari-604.1.38.1-branch/Source

Versioning.

3:24 PM Changeset in webkit [220358] by matthew_hanson@apple.com
  • 1 copy in branches/safari-604.1.38.1-branch

New branch.

3:22 PM Changeset in webkit [220357] by matthew_hanson@apple.com
  • 2 edits in branches/safari-604.1.38.0-branch/Source/WebKit

Cherry-pick r220273. rdar://problem/33732757

3:20 PM Changeset in webkit [220356] by matthew_hanson@apple.com
  • 7 edits in branches/safari-604.1.38.0-branch/Source

Versioning.

3:09 PM Changeset in webkit [220355] by matthew_hanson@apple.com
  • 1 copy in branches/safari-604.1.38.0-branch

New branch.

3:01 PM Changeset in webkit [220354] by Michael Catanzaro
  • 2 edits in trunk/Source/WebCore

-Wimplicit-fallthrough warning in ComputedStyleExtractor::propertyValue
​https://bugs.webkit.org/show_bug.cgi?id=174469
<rdar://problem/33311638>

Reviewed by Simon Fraser.

Add a RELEASE_ASSERT_NOT_REACHED to silence this warning.

  • css/CSSComputedStyleDeclaration.cpp:

(WebCore::ComputedStyleExtractor::propertyValue):

2:53 PM Changeset in webkit [220353] by fpizlo@apple.com
  • 3 edits in trunk/Source/WTF

Unreviewed, try to fix Windows build.

  • wtf/Gigacage.cpp:
  • wtf/Gigacage.h:
2:31 PM Changeset in webkit [220352] by fpizlo@apple.com
  • 54 edits
    1 move
    1 add in trunk/Source

Primitive auxiliaries and JSValue auxiliaries should have separate gigacages
​https://bugs.webkit.org/show_bug.cgi?id=174919

Reviewed by Keith Miller.
Source/bmalloc:


This introduces two kinds of Gigacage, Primitive and JSValue. This translates to two kinds of
HeapKind, PrimitiveGigacage and JSValueGigacage.

The new support functionality required turning Inline.h into BInline.h, and INLINE into BINLINE, and
NO_INLINE into BNO_INLINE.

  • bmalloc.xcodeproj/project.pbxproj:
  • bmalloc/Allocator.cpp:

(bmalloc::Allocator::refillAllocatorSlowCase):
(bmalloc::Allocator::refillAllocator):
(bmalloc::Allocator::allocateLarge):
(bmalloc::Allocator::allocateLogSizeClass):

  • bmalloc/AsyncTask.h:
  • bmalloc/BInline.h: Copied from Source/bmalloc/bmalloc/Inline.h.
  • bmalloc/Cache.cpp:

(bmalloc::Cache::tryAllocateSlowCaseNullCache):
(bmalloc::Cache::allocateSlowCaseNullCache):
(bmalloc::Cache::deallocateSlowCaseNullCache):
(bmalloc::Cache::reallocateSlowCaseNullCache):

  • bmalloc/Deallocator.cpp:
  • bmalloc/Gigacage.cpp:

(Gigacage::PrimitiveDisableCallbacks::PrimitiveDisableCallbacks):
(Gigacage::ensureGigacage):
(Gigacage::disablePrimitiveGigacage):
(Gigacage::addPrimitiveDisableCallback):
(Gigacage::removePrimitiveDisableCallback):
(Gigacage::Callbacks::Callbacks): Deleted.
(Gigacage::disableGigacage): Deleted.
(Gigacage::addDisableCallback): Deleted.
(Gigacage::removeDisableCallback): Deleted.

  • bmalloc/Gigacage.h:

(Gigacage::name):
(Gigacage::basePtr):
(Gigacage::forEachKind):
(Gigacage::caged):
(Gigacage::isCaged):

  • bmalloc/Heap.cpp:

(bmalloc::Heap::Heap):
(bmalloc::Heap::usingGigacage):
(bmalloc::Heap::gigacageBasePtr):

  • bmalloc/Heap.h:
  • bmalloc/HeapKind.h:

(bmalloc::isGigacage):
(bmalloc::gigacageKind):
(bmalloc::heapKind):

  • bmalloc/Inline.h: Removed.
  • bmalloc/Map.h:
  • bmalloc/PerProcess.h:

(bmalloc::PerProcess<T>::getFastCase):
(bmalloc::PerProcess<T>::get):
(bmalloc::PerProcess<T>::getSlowCase):

  • bmalloc/PerThread.h:

(bmalloc::PerThread<T>::getFastCase):

  • bmalloc/Vector.h:

(bmalloc::Vector<T>::push):
(bmalloc::Vector<T>::shrinkCapacity):
(bmalloc::Vector<T>::growCapacity):

Source/JavaScriptCore:


This adapts JSC to there being two gigacages.

To make matters simpler, this turns AlignedMemoryAllocators into per-VM instances rather than
singletons. I don't think we were gaining anything by making them be singletons.

This makes it easy to teach GigacageAlignedMemoryAllocator that there are multiple kinds of
gigacages. We'll have one of those allocators per cage.

From there, this change teaches everyone who previously knew about cages that there are two cages.
This means having to specify either Gigacage::Primitive or Gigacage::JSValue. In most places, this is
easy: typed arrays are Primitive and butterflies are JSValue. But there are a few places where it's
not so obvious, so this change introduces some helpers to make it easy to define what cage you want
to use in one place and refer to it abstractly. We do this in DirectArguments and GenericArguments.h

A lot of the magic of this change is due to CagedBarrierPtr, which combines AuxiliaryBarrier and
CagedPtr. This removes one layer of "get()" calls from a bunch of places.

(JSC::AccessCase::generateImpl):

  • dfg/DFGSpeculativeJIT.cpp:

(JSC::DFG::SpeculativeJIT::emitAllocateRawObject):
(JSC::DFG::SpeculativeJIT::compileAllocatePropertyStorage):
(JSC::DFG::SpeculativeJIT::compileReallocatePropertyStorage):
(JSC::DFG::SpeculativeJIT::compileNewTypedArray):
(JSC::DFG::SpeculativeJIT::emitAllocateButterfly):

  • ftl/FTLLowerDFGToB3.cpp:

(JSC::FTL::DFG::LowerDFGToB3::compileGetButterfly):
(JSC::FTL::DFG::LowerDFGToB3::compileGetIndexedPropertyStorage):
(JSC::FTL::DFG::LowerDFGToB3::compileNewTypedArray):
(JSC::FTL::DFG::LowerDFGToB3::compileGetDirectPname):
(JSC::FTL::DFG::LowerDFGToB3::compileMaterializeNewObject):
(JSC::FTL::DFG::LowerDFGToB3::allocatePropertyStorageWithSizeImpl):
(JSC::FTL::DFG::LowerDFGToB3::allocateJSArray):
(JSC::FTL::DFG::LowerDFGToB3::caged):

  • heap/FastMallocAlignedMemoryAllocator.cpp:

(JSC::FastMallocAlignedMemoryAllocator::instance): Deleted.

  • heap/FastMallocAlignedMemoryAllocator.h:
  • heap/GigacageAlignedMemoryAllocator.cpp:

(JSC::GigacageAlignedMemoryAllocator::GigacageAlignedMemoryAllocator):
(JSC::GigacageAlignedMemoryAllocator::tryAllocateAlignedMemory):
(JSC::GigacageAlignedMemoryAllocator::freeAlignedMemory):
(JSC::GigacageAlignedMemoryAllocator::dump const):
(JSC::GigacageAlignedMemoryAllocator::instance): Deleted.

  • heap/GigacageAlignedMemoryAllocator.h:
  • jsc.cpp:

(primitiveGigacageDisabled):
(jscmain):
(gigacageDisabled): Deleted.

  • llint/LowLevelInterpreter64.asm:
  • runtime/ArrayBuffer.cpp:

(JSC::ArrayBufferContents::tryAllocate):
(JSC::ArrayBuffer::createAdopted):
(JSC::ArrayBuffer::createFromBytes):

  • runtime/AuxiliaryBarrier.h:
  • runtime/ButterflyInlines.h:

(JSC::Butterfly::createUninitialized):
(JSC::Butterfly::tryCreate):
(JSC::Butterfly::growArrayRight):

  • runtime/CagedBarrierPtr.h: Added.

(JSC::CagedBarrierPtr::CagedBarrierPtr):
(JSC::CagedBarrierPtr::clear):
(JSC::CagedBarrierPtr::set):
(JSC::CagedBarrierPtr::get const):
(JSC::CagedBarrierPtr::getMayBeNull const):
(JSC::CagedBarrierPtr::operator== const):
(JSC::CagedBarrierPtr::operator!= const):
(JSC::CagedBarrierPtr::operator bool const):
(JSC::CagedBarrierPtr::setWithoutBarrier):
(JSC::CagedBarrierPtr::operator* const):
(JSC::CagedBarrierPtr::operator-> const):
(JSC::CagedBarrierPtr::operator[] const):

  • runtime/DirectArguments.cpp:

(JSC::DirectArguments::overrideThings):
(JSC::DirectArguments::unmapArgument):

  • runtime/DirectArguments.h:

(JSC::DirectArguments::isMappedArgument const):

  • runtime/GenericArguments.h:
  • runtime/GenericArgumentsInlines.h:

(JSC::GenericArguments<Type>::initModifiedArgumentsDescriptor):
(JSC::GenericArguments<Type>::setModifiedArgumentDescriptor):
(JSC::GenericArguments<Type>::isModifiedArgumentDescriptor):

  • runtime/HashMapImpl.cpp:

(JSC::HashMapImpl<HashMapBucket>::visitChildren):

  • runtime/HashMapImpl.h:

(JSC::HashMapBuffer::create):
(JSC::HashMapImpl::buffer const):
(JSC::HashMapImpl::rehash):

  • runtime/JSArray.cpp:

(JSC::JSArray::tryCreateUninitializedRestricted):
(JSC::JSArray::unshiftCountSlowCase):
(JSC::JSArray::setLength):
(JSC::JSArray::pop):
(JSC::JSArray::push):
(JSC::JSArray::fastSlice):
(JSC::JSArray::shiftCountWithArrayStorage):
(JSC::JSArray::shiftCountWithAnyIndexingType):
(JSC::JSArray::unshiftCountWithAnyIndexingType):
(JSC::JSArray::fillArgList):
(JSC::JSArray::copyToArguments):

  • runtime/JSArray.h:

(JSC::JSArray::tryCreate):

  • runtime/JSArrayBufferView.cpp:

(JSC::JSArrayBufferView::ConstructionContext::ConstructionContext):
(JSC::JSArrayBufferView::finalize):

  • runtime/JSLock.cpp:

(JSC::JSLock::didAcquireLock):

  • runtime/JSObject.cpp:

(JSC::JSObject::heapSnapshot):
(JSC::JSObject::getOwnPropertySlotByIndex):
(JSC::JSObject::putByIndex):
(JSC::JSObject::enterDictionaryIndexingMode):
(JSC::JSObject::createInitialIndexedStorage):
(JSC::JSObject::createArrayStorage):
(JSC::JSObject::convertUndecidedToInt32):
(JSC::JSObject::convertUndecidedToDouble):
(JSC::JSObject::convertUndecidedToContiguous):
(JSC::JSObject::constructConvertedArrayStorageWithoutCopyingElements):
(JSC::JSObject::convertUndecidedToArrayStorage):
(JSC::JSObject::convertInt32ToDouble):
(JSC::JSObject::convertInt32ToContiguous):
(JSC::JSObject::convertInt32ToArrayStorage):
(JSC::JSObject::convertDoubleToContiguous):
(JSC::JSObject::convertDoubleToArrayStorage):
(JSC::JSObject::convertContiguousToArrayStorage):
(JSC::JSObject::setIndexQuicklyToUndecided):
(JSC::JSObject::ensureArrayStorageExistsAndEnterDictionaryIndexingMode):
(JSC::JSObject::deletePropertyByIndex):
(JSC::JSObject::getOwnPropertyNames):
(JSC::JSObject::putIndexedDescriptor):
(JSC::JSObject::defineOwnIndexedProperty):
(JSC::JSObject::putByIndexBeyondVectorLengthWithoutAttributes):
(JSC::JSObject::putDirectIndexSlowOrBeyondVectorLength):
(JSC::JSObject::getNewVectorLength):
(JSC::JSObject::ensureLengthSlow):
(JSC::JSObject::reallocateAndShrinkButterfly):
(JSC::JSObject::allocateMoreOutOfLineStorage):
(JSC::JSObject::getEnumerableLength):

  • runtime/JSObject.h:

(JSC::JSObject::getArrayLength const):
(JSC::JSObject::getVectorLength):
(JSC::JSObject::putDirectIndex):
(JSC::JSObject::canGetIndexQuickly):
(JSC::JSObject::getIndexQuickly):
(JSC::JSObject::tryGetIndexQuickly const):
(JSC::JSObject::canSetIndexQuickly):
(JSC::JSObject::setIndexQuickly):
(JSC::JSObject::initializeIndex):
(JSC::JSObject::initializeIndexWithoutBarrier):
(JSC::JSObject::hasSparseMap):
(JSC::JSObject::inSparseIndexingMode):
(JSC::JSObject::butterfly const):
(JSC::JSObject::butterfly):
(JSC::JSObject::outOfLineStorage const):
(JSC::JSObject::outOfLineStorage):
(JSC::JSObject::ensureInt32):
(JSC::JSObject::ensureDouble):
(JSC::JSObject::ensureContiguous):
(JSC::JSObject::ensureArrayStorage):
(JSC::JSObject::arrayStorage):
(JSC::JSObject::arrayStorageOrNull):
(JSC::JSObject::ensureLength):

  • runtime/RegExpMatchesArray.h:

(JSC::tryCreateUninitializedRegExpMatchesArray):

  • runtime/VM.cpp:

(JSC::VM::VM):
(JSC::VM::~VM):
(JSC::VM::primitiveGigacageDisabledCallback):
(JSC::VM::primitiveGigacageDisabled):
(JSC::VM::gigacageDisabledCallback): Deleted.
(JSC::VM::gigacageDisabled): Deleted.

  • runtime/VM.h:

(JSC::VM::gigacageAuxiliarySpace):
(JSC::VM::firePrimitiveGigacageEnabledIfNecessary):
(JSC::VM::primitiveGigacageEnabled):
(JSC::VM::fireGigacageEnabledIfNecessary): Deleted.
(JSC::VM::gigacageEnabled): Deleted.

  • wasm/WasmMemory.cpp:

(JSC::Wasm::Memory::create):
(JSC::Wasm::Memory::~Memory):
(JSC::Wasm::Memory::grow):

Source/WebCore:

No new tests because no change in behavior.

Adapting to API changes - we now specify the AlignedMemoryAllocator differently and we need to be
specific about which Gigacage we're using.

  • bindings/js/WebCoreJSClientData.cpp:

(WebCore::JSVMClientData::JSVMClientData):

  • platform/graphics/cocoa/GPUBufferMetal.mm:

(WebCore::GPUBuffer::GPUBuffer):

Source/WebKit:


The disable callback is all about the primitive gigacage.

  • WebProcess/WebProcess.cpp:

(WebKit::primitiveGigacageDisabled):
(WebKit::m_webSQLiteDatabaseTracker):
(WebKit::gigacageDisabled): Deleted.

Source/WTF:


This mirrors the changes from bmalloc/Gigacage.h.

Also it teaches CagedPtr how to reason about multiple gigacages.

  • wtf/CagedPtr.h:

(WTF::CagedPtr::get const):
(WTF::CagedPtr::operator[] const):

  • wtf/Gigacage.cpp:

(Gigacage::tryMalloc):
(Gigacage::tryAllocateVirtualPages):
(Gigacage::freeVirtualPages):
(Gigacage::tryAlignedMalloc):
(Gigacage::alignedFree):
(Gigacage::free):

  • wtf/Gigacage.h:

(Gigacage::disablePrimitiveGigacage):
(Gigacage::addPrimitiveDisableCallback):
(Gigacage::removePrimitiveDisableCallback):
(Gigacage::name):
(Gigacage::basePtr):
(Gigacage::caged):
(Gigacage::isCaged):
(Gigacage::tryAlignedMalloc):
(Gigacage::alignedFree):
(Gigacage::free):
(Gigacage::disableGigacage): Deleted.
(Gigacage::addDisableCallback): Deleted.
(Gigacage::removeDisableCallback): Deleted.

1:58 PM Changeset in webkit [220351] by commit-queue@webkit.org
  • 4 edits
    1 add in trunk/Source/WebCore

[Curl] Add abstraction layer of cookie jar implementation for Curl port
​https://bugs.webkit.org/show_bug.cgi?id=174943

Patch by Basuke Suzuki <Basuke Suzuki> on 2017-08-07
Reviewed by Brent Fulgham.

  • platform/network/curl/CookieJarCurl.cpp:

(WebCore::CookieJarCurlFileSystem::setCookiesFromDOM):
(WebCore::CookieJarCurlFileSystem::cookiesForDOM):
(WebCore::CookieJarCurlFileSystem::cookieRequestHeaderFieldValue):
(WebCore::CookieJarCurlFileSystem::cookiesEnabled):
(WebCore::CookieJarCurlFileSystem::getRawCookies):
(WebCore::CookieJarCurlFileSystem::deleteCookie):
(WebCore::CookieJarCurlFileSystem::getHostnamesWithCookies):
(WebCore::CookieJarCurlFileSystem::deleteCookiesForHostnames):
(WebCore::CookieJarCurlFileSystem::deleteAllCookies):
(WebCore::CookieJarCurlFileSystem::deleteAllCookiesModifiedSince):
(WebCore::cookiesForDOM):
(WebCore::setCookiesFromDOM):
(WebCore::cookieRequestHeaderFieldValue):
(WebCore::cookiesEnabled):
(WebCore::getRawCookies):
(WebCore::deleteCookie):
(WebCore::getHostnamesWithCookies):
(WebCore::deleteCookiesForHostnames):
(WebCore::deleteAllCookies):
(WebCore::deleteAllCookiesModifiedSince):

  • platform/network/curl/CookieJarCurl.h: Added.
  • platform/network/curl/CurlContext.cpp:
  • platform/network/curl/CurlContext.h:

(WebCore::CurlContext::cookieJar):

1:30 PM Changeset in webkit [220350] by Ryan Haddad
  • 2 edits in trunk/LayoutTests

Skip workers/wasm-hashset-many.html and workers/wasm-hashset-many-2.html on El Capitan.
​https://bugs.webkit.org/show_bug.cgi?id=175102

Unreviewed test gardening.

  • platform/mac/TestExpectations:
11:58 AM Changeset in webkit [220349] by matthew_hanson@apple.com
  • 2 edits in branches/safari-604.1-branch/Source/WebCore

Cherry-pick r220248. rdar://problem/33754446

11:55 AM Changeset in webkit [220348] by matthew_hanson@apple.com
  • 2 edits in branches/safari-604.1-branch/Source/WebKit

Cherry-pick r220273. rdar://problem/33732754

11:55 AM Changeset in webkit [220347] by matthew_hanson@apple.com
  • 4 edits in branches/safari-604.1-branch/Source/WebCore

Cherry-pick r220247. rdar://problem/33754440

11:42 AM Changeset in webkit [220346] by commit-queue@webkit.org
  • 3 edits in trunk/Source/JavaScriptCore

Unreviewed, rolling out r220144.
​https://bugs.webkit.org/show_bug.cgi?id=175276

"It did not actually speed things up in the way I expected"
(Requested by saamyjoon on #webkit).

Reverted changeset:

"On memory-constrained iOS devices, reduce the rate at which
the JS heap grows before a GC to try to keep more memory
available for the system"
​https://bugs.webkit.org/show_bug.cgi?id=175041
​http://trac.webkit.org/changeset/220144

11:25 AM Changeset in webkit [220345] by Ryan Haddad
  • 2 edits in trunk/LayoutTests

Marking imported/w3c/web-platform-tests/html/browsers/windows/noreferrer-window-name.html as failing on iOS.
​https://bugs.webkit.org/show_bug.cgi?id=175273

Unreviewed test gardening.

  • platform/ios/TestExpectations:
11:22 AM Changeset in webkit [220344] by beidson@apple.com
  • 62 edits
    4 copies in trunk

Implement most of ServiceWorkerContainer::addRegistration.
​https://bugs.webkit.org/show_bug.cgi?id=175237

Reviewed by Andy Estes.

LayoutTests/imported/w3c:

  • web-platform-tests/FileAPI/historical.https-expected.txt:
  • web-platform-tests/background-fetch/interfaces-worker.https-expected.txt:
  • web-platform-tests/fetch/api/policies/referrer-no-referrer-service-worker.https-expected.txt:
  • web-platform-tests/fetch/api/policies/referrer-origin-service-worker.https-expected.txt:
  • web-platform-tests/fetch/api/policies/referrer-origin-when-cross-origin-service-worker.https-expected.txt:
  • web-platform-tests/fetch/api/policies/referrer-unsafe-url-service-worker.https-expected.txt:
  • web-platform-tests/html/webappapis/scripting/events/messageevent-constructor.https-expected.txt:
  • web-platform-tests/html/webappapis/scripting/processing-model-2/integration-with-the-javascript-agent-formalism/canblock-serviceworker.https-expected.txt:
  • web-platform-tests/service-workers/cache-storage/serviceworker/cache-add.https-expected.txt:
  • web-platform-tests/service-workers/cache-storage/serviceworker/cache-delete.https-expected.txt:
  • web-platform-tests/service-workers/cache-storage/serviceworker/cache-keys.https-expected.txt:
  • web-platform-tests/service-workers/cache-storage/serviceworker/cache-match.https-expected.txt:
  • web-platform-tests/service-workers/cache-storage/serviceworker/cache-matchAll.https-expected.txt:
  • web-platform-tests/service-workers/cache-storage/serviceworker/cache-put.https-expected.txt:
  • web-platform-tests/service-workers/cache-storage/serviceworker/cache-storage-keys.https-expected.txt:
  • web-platform-tests/service-workers/cache-storage/serviceworker/cache-storage-match.https-expected.txt:
  • web-platform-tests/service-workers/cache-storage/serviceworker/cache-storage.https-expected.txt:
  • web-platform-tests/streams/byte-length-queuing-strategy.serviceworker.https-expected.txt:
  • web-platform-tests/streams/count-queuing-strategy.serviceworker.https-expected.txt:
  • web-platform-tests/streams/piping/close-propagation-backward.serviceworker.https-expected.txt:
  • web-platform-tests/streams/piping/close-propagation-forward.serviceworker.https-expected.txt:
  • web-platform-tests/streams/piping/error-propagation-backward.serviceworker.https-expected.txt:
  • web-platform-tests/streams/piping/error-propagation-forward.serviceworker.https-expected.txt:
  • web-platform-tests/streams/piping/flow-control.serviceworker.https-expected.txt:
  • web-platform-tests/streams/piping/general.serviceworker.https-expected.txt:
  • web-platform-tests/streams/piping/multiple-propagation.serviceworker.https-expected.txt:
  • web-platform-tests/streams/piping/pipe-through.serviceworker.https-expected.txt:
  • web-platform-tests/streams/piping/transform-streams.serviceworker.https-expected.txt:
  • web-platform-tests/streams/readable-byte-streams/general.serviceworker.https-expected.txt:
  • web-platform-tests/streams/readable-streams/bad-strategies.serviceworker.https-expected.txt:
  • web-platform-tests/streams/readable-streams/bad-underlying-sources.serviceworker.https-expected.txt:
  • web-platform-tests/streams/readable-streams/brand-checks.serviceworker.https-expected.txt:
  • web-platform-tests/streams/readable-streams/cancel.serviceworker.https-expected.txt:
  • web-platform-tests/streams/readable-streams/count-queuing-strategy-integration.serviceworker.https-expected.txt:
  • web-platform-tests/streams/readable-streams/default-reader.serviceworker.https-expected.txt:
  • web-platform-tests/streams/readable-streams/floating-point-total-queue-size.serviceworker.https-expected.txt:
  • web-platform-tests/streams/readable-streams/garbage-collection.serviceworker.https-expected.txt:
  • web-platform-tests/streams/readable-streams/general.serviceworker.https-expected.txt:
  • web-platform-tests/streams/readable-streams/pipe-through.serviceworker.https-expected.txt:
  • web-platform-tests/streams/readable-streams/readable-stream-reader.serviceworker.https-expected.txt:
  • web-platform-tests/streams/readable-streams/tee.serviceworker.https-expected.txt:
  • web-platform-tests/streams/readable-streams/templated.serviceworker.https-expected.txt:

Source/WebCore:

No new tests (Covered by changes to existing tests).

There's still so much supporting infrastructure to add with these early patches
that I'm still moving them in baby steps for now, hence not implementing
register() all in one shot.

Things will start moving very quickly once we no longer need to add lots of new
primitives in each change.

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

(WebCore::DOMWindow::navigator const):

  • page/Navigator.cpp:

(WebCore::Navigator::Navigator): Make the constructor take a ScriptExecutionContext for

creation of objects where its import (e.g. ServiceWorkerContainer).

  • page/Navigator.h:
  • page/WorkerNavigator.cpp:

(WebCore::WorkerNavigator::WorkerNavigator): Ditto.

  • page/WorkerNavigator.h:
  • page/NavigatorBase.cpp:

(WebCore::NavigatorBase::NavigatorBase): Create the ServiceWorkerContainer upfront with

the passed-in ScriptExecutionContext.

(WebCore::NavigatorBase::serviceWorker):

  • page/NavigatorBase.h:

Make ServiceWorkerContainer into an ActiveDOMObject. This will eventually be necessary for
Document suspension reasons, but is also necessary because it also needs to be a
ContextDestructionObserver (which ActiveDOMObject is):

  • workers/ServiceWorkerContainer.cpp:

(WebCore::ServiceWorkerContainer::ServiceWorkerContainer):
(WebCore::ServiceWorkerContainer::ready):
(WebCore::ServiceWorkerContainer::addRegistration):
(WebCore::ServiceWorkerContainer::getRegistration):
(WebCore::ServiceWorkerContainer::getRegistrations):
(WebCore::ServiceWorkerContainer::activeDOMObjectName const):
(WebCore::ServiceWorkerContainer::canSuspendForDocumentSuspension const):
(WebCore::rejectLater): Deleted.
(WebCore::ServiceWorkerContainer::eventTargetInterface const): Deleted.
(WebCore::ServiceWorkerContainer::scriptExecutionContext const): Deleted.

  • workers/ServiceWorkerContainer.h:
  • workers/ServiceWorkerContainer.idl:

Add updateViaCache, etc:

  • workers/ServiceWorkerRegistration.cpp:

(WebCore::ServiceWorkerRegistration::updateViaCache const):
(WebCore::ServiceWorkerRegistration::update):
(WebCore::ServiceWorkerRegistration::unregister):

  • workers/ServiceWorkerRegistration.h:
  • workers/ServiceWorkerRegistration.idl:
  • workers/ServiceWorkerUpdateViaCache.h: Copied from Source/WebCore/workers/ServiceWorkerRegistration.idl.
  • workers/ServiceWorkerUpdateViaCache.idl: Copied from Source/WebCore/workers/ServiceWorkerRegistration.idl.
  • workers/WorkerGlobalScope.cpp:

(WebCore::WorkerGlobalScope::navigator):
(WebCore::WorkerGlobalScope::navigator const): Deleted.

  • workers/WorkerGlobalScope.h:
  • workers/WorkerType.h: Copied from Source/WebCore/workers/ServiceWorkerRegistration.idl.
  • workers/WorkerType.idl: Copied from Source/WebCore/workers/ServiceWorkerRegistration.idl.
10:06 AM Changeset in webkit [220343] by Ryan Haddad
  • 3 edits in trunk/LayoutTests

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

Unreviewed test gardening.

  • platform/mac-elcapitan/js/dom/global-constructors-attributes-expected.txt:
  • platform/mac/js/dom/global-constructors-attributes-expected.txt:
9:44 AM Changeset in webkit [220342] by Ryan Haddad
  • 12 edits
    1 delete in trunk

Unreviewed, rolling out r220299.

This change caused LayoutTest inspector/dom-debugger/dom-
breakpoints.html to fail.

Reverted changeset:

"Web Inspector: capture async stack trace when workers/main
context posts a message"
​https://bugs.webkit.org/show_bug.cgi?id=167084
​http://trac.webkit.org/changeset/220299

9:42 AM Changeset in webkit [220341] by matthew_hanson@apple.com
  • 2 edits in branches/safari-604-branch/Source/WebKit

Cherry-pick r220273. rdar://problem/33732762

9:42 AM Changeset in webkit [220340] by matthew_hanson@apple.com
  • 6 edits in branches/safari-604-branch

Cherry-pick r220185. rdar://problem/33713534

9:42 AM Changeset in webkit [220339] by matthew_hanson@apple.com
  • 4 edits
    3 adds in branches/safari-604-branch

Cherry-pick r220163. rdar://problem/33711032

9:42 AM Changeset in webkit [220338] by matthew_hanson@apple.com
  • 2 edits in branches/safari-604-branch/Source/WebCore

Cherry-pick r220153. rdar://problem/33711047

9:32 AM Changeset in webkit [220337] by BJ Burg
  • 6 edits in trunk/Source/WebCore

Cleanup: simplify WebSockets code for RuntimeEnabledFeatures
​https://bugs.webkit.org/show_bug.cgi?id=175190

Reviewed by Sam Weinig.

This runtime enabled flag is not used by anything. It was added
for V8 in <​https://bugs.webkit.org/show_bug.cgi?id=29896>.

  • Modules/websockets/WebSocket.cpp:

(WebCore::WebSocket::setIsAvailable): Deleted.
(WebCore::WebSocket::isAvailable): Deleted.

  • Modules/websockets/WebSocket.h:
  • Modules/websockets/WebSocket.idl:
  • page/RuntimeEnabledFeatures.cpp:

(WebCore::RuntimeEnabledFeatures::webSocketEnabled const): Deleted.

  • page/RuntimeEnabledFeatures.h:
9:29 AM Changeset in webkit [220336] by BJ Burg
  • 24 edits in trunk

Remove CANVAS_PATH compilation guard
​https://bugs.webkit.org/show_bug.cgi?id=175207

Reviewed by Sam Weinig.

.:

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

Source/JavaScriptCore:

  • Configurations/FeatureDefines.xcconfig:

Source/WebCore:

  • Configurations/FeatureDefines.xcconfig:
  • html/canvas/DOMPath.h:
  • html/canvas/DOMPath.idl:

Source/WebCore/PAL:

  • Configurations/FeatureDefines.xcconfig:

Source/WebKit:

  • Configurations/FeatureDefines.xcconfig:

Source/WebKitLegacy/mac:

  • Configurations/FeatureDefines.xcconfig:

Source/WTF:

  • wtf/FeatureDefines.h:

Tools:

  • Scripts/webkitperl/FeatureList.pm:
  • TestWebKitAPI/Configurations/FeatureDefines.xcconfig:
9:22 AM Changeset in webkit [220335] by pvollan@apple.com
  • 2 edits in trunk/Tools

[Win] Use Visual Studio 2017 if installed.
​https://bugs.webkit.org/show_bug.cgi?id=175188

Reviewed by Sam Weinig.

If installed, use VS2017 to build WebKit. If not, fall back to VS2015.

  • Scripts/webkitdirs.pm:

(visualStudioInstallDir):
(msBuildInstallDir):

8:44 AM Changeset in webkit [220334] by clopez@igalia.com
  • 4 edits in trunk/Source

[WPE][CMake] Only pass pkg-config CFLAGS from (E)GL when not using libepoxy.
​https://bugs.webkit.org/show_bug.cgi?id=175125

Reviewed by Michael Catanzaro.

Source/WebCore:

No change of behavior, covered by existing tests.

  • CMakeLists.txt: Instead of negating 3 times the use of libepoxy

move the logic for (E)GL detection inside an else() block.

Source/WebKit:

  • CMakeLists.txt: When using libepoxy avoid passing the CFLAGS from (E)GL,

and instead pass the libepoxy ones.

6:14 AM Changeset in webkit [220333] by Antti Koivisto
  • 2 edits in trunk/Source/WebCore

REGRESSION (r219121): Airmail 3 prints header part only.
​https://bugs.webkit.org/show_bug.cgi?id=175258
<rdar://problem/33601173>

Reviewed by Andreas Kling.

When a WK1 WebViw is printed via AppKit view hierarchy it won't explictly set the page width
but uses the existing width. r219121 assumes that all printing code paths set the page width.

No test, there appears to be no good way to test AppKit printing behaviors without adding complicated
new testing infrastructure.

  • rendering/RenderView.cpp:

(WebCore::RenderView::layout):

If we are in printing layout and don't have page width set yet then use the current view width.
This matches the behavior prior r219121.

6:08 AM Changeset in webkit [220332] by commit-queue@webkit.org
  • 2 edits in trunk/LayoutTests

Remove obsolete failure expectations for wpt browsers tests.
​https://bugs.webkit.org/show_bug.cgi?id=175073

Unreviewed test gardening.

Patch by Ms2ger <​Ms2ger@igalia.com> on 2017-08-07

3:03 AM WebKitGTK/2.16.x edited by clopez@igalia.com
(diff)
2:53 AM Changeset in webkit [220331] by clopez@igalia.com
  • 2 edits in trunk/Source/WebKit

[GTK][WPE] CFLAGS from pkg-config for (E)GL are not passed to WebKit
​https://bugs.webkit.org/show_bug.cgi?id=175125

Unreviewed follow-up fix after r220326.

  • CMakeLists.txt: Move the block appending to WebKit2_LIBRARIES

below where the list itself is created.

12:03 AM Changeset in webkit [220330] by keith_miller@apple.com
  • 2 edits in trunk/Source/JavaScriptCore

REGRESSION: wasm.yaml/wasm/js-api/dont-mmap-zero-byte-memory.js failing on JSC Debug bots
​https://bugs.webkit.org/show_bug.cgi?id=175256

Reviewed by Saam Barati.

The check in createFromBytes just needed to check that the buffer was not null before
calling isCaged.

  • runtime/ArrayBuffer.cpp:

(JSC::ArrayBuffer::createFromBytes):

Note: See TracTimeline for information about the timeline view.