⚠ Archived content — this site is no longer maintained.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

Timeline



Mar 18, 2017:

6:06 PM WebKitGTK/2.16.x edited by Michael Catanzaro
(diff)
5:58 PM Changeset in webkit [214142] by Michael Catanzaro
  • 2 edits in trunk/Source/WebCore/platform/gtk/po

Update Ukrainian translation
​https://bugs.webkit.org/show_bug.cgi?id=169812

Patch by Yuri Chornoivan <​yurchor@ukr.net> on 2017-03-18
Rubber-stamped by Michael Catanzaro.

  • uk.po:
5:58 PM Changeset in webkit [214141] by webkit@devinrousso.com
  • 2 edits in trunk/Source/WebInspectorUI

Web Inspector: Empty attribute added to a DOM tree outline element should not add whitespace within the tag
​https://bugs.webkit.org/show_bug.cgi?id=139623

Reviewed by Brian Burg.

  • UserInterface/Views/DOMTreeElement.js:

(WebInspector.DOMTreeElement.prototype._attributeEditingCommitted):
If the newText after editing has no content, remove the element for that attribute.

5:41 PM Changeset in webkit [214140] by webkit@devinrousso.com
  • 3 edits in trunk/Source/WebInspectorUI

Web Inspector: add keyboard shortcut to clear timeline records
​https://bugs.webkit.org/show_bug.cgi?id=132485

Reviewed by Brian Burg.

  • Localizations/en.lproj/localizedStrings.js:
  • UserInterface/Views/TimelineRecordingContentView.js:

(WebInspector.TimelineRecordingContentView):
(WebInspector.TimelineRecordingContentView.prototype.handleClearShortcut):

1:22 PM Changeset in webkit [214139] by jonlee@apple.com
  • 13 edits in trunk

Add support for ImplementedAs, Clamp, EnforceRange, TreatNullAs for dictionary members
​https://bugs.webkit.org/show_bug.cgi?id=169731

Reviewed by Alex Christensen.

LayoutTests/imported/w3c:

Rebaseline tests. The results seem to show a problem in the original tests.

  • web-platform-tests/WebCryptoAPI/encrypt_decrypt/aes_gcm.worker-expected.txt:
  • web-platform-tests/WebCryptoAPI/encrypt_decrypt/test_aes_gcm-expected.txt:

Source/WebCore:

  • bindings/scripts/CodeGeneratorJS.pm:

(GenerateDictionaryImplementationContent): Create a new $implementationAsKey. Look for
ImplementedAs, and if it exists, override the value with the alias. Set the parameters of
convert<>() using JSValueToNative.
(IsValidContextForJSValueToNative): Update to include IDLDictionaryMembers.
(JSValueToNative): Bypass parseEnumeration serialization for enums if the context is an
IDLDictionaryMember. Dictionary members need convert<IDLEnumeration> which throw a TypeError
or return the enum (and utilizes parseEnumeration). parseEnumeration, in contrast, returns
an optional.

  • bindings/scripts/test/TestObj.idl: Add test cases in TestDictionary.

Update test results.

  • bindings/scripts/test/JS/JSTestEventConstructor.cpp:
  • bindings/scripts/test/JS/JSTestObj.cpp:
  • bindings/scripts/test/JS/JSTestStandaloneDictionary.cpp:

LayoutTests:

Rebaseline. The enhancement progressed the results of these tests.

  • crypto/subtle/aes-gcm-encrypt-malformed-parameters-expected.txt:
  • crypto/subtle/aes-generate-key-malformed-parameters-expected.txt:
  • crypto/subtle/hmac-generate-key-malformed-parameters-expected.txt:
1:21 PM Changeset in webkit [214138] by gskachkov@gmail.com
  • 2 edits in trunk/Source/JavaScriptCore

[JSC] Remove unnecessary condition from needsDerivedConstructorInArrowFunctionLexicalEnvironment in BytecodeGenerator.cpp
​https://bugs.webkit.org/show_bug.cgi?id=169832

Reviewed by Mark Lam.

Remove already covered condition in needsDerivedConstructorInArrowFunctionLexicalEnvironment
function. Condition isConstructor() && constructorKind() == ConstructorKind::Extends is already
isClassContext.

  • bytecompiler/BytecodeGenerator.cpp:

(JSC::BytecodeGenerator::needsDerivedConstructorInArrowFunctionLexicalEnvironment):

1:03 PM Changeset in webkit [214137] by Simon Fraser
  • 3 edits in trunk/Source/WebCore

Use initializers for Document member variables
​https://bugs.webkit.org/show_bug.cgi?id=169848

Reviewed by Chris Dumez.

Move Document to use C++11 brace initializers.

No behavior change.

  • dom/Document.cpp:

(WebCore::Document::Document):

  • dom/Document.h:

(WebCore::Document::xmlStandaloneStatus):
(WebCore::Document::setHasXMLDeclaration):

12:51 PM Changeset in webkit [214136] by Chris Dumez
  • 8 edits in trunk

innerText setter inserts empty text node if value starts with newline
​https://bugs.webkit.org/show_bug.cgi?id=169600

Reviewed by Ryosuke Niwa.

LayoutTests/imported/w3c:

Re-sync tests with upstream to gain test coverage.

  • web-platform-tests/innerText/getter-expected.txt:
  • web-platform-tests/innerText/getter-tests.js:
  • web-platform-tests/innerText/setter-expected.txt:
  • web-platform-tests/innerText/setter-tests.js:
  • web-platform-tests/innerText/setter.html:

Source/WebCore:

innerText setter should not create empty text nodes as per:

No new tests, updated existing test.

  • html/HTMLElement.cpp:

(WebCore::textToFragment):

10:52 AM Changeset in webkit [214135] by Chris Dumez
  • 19 edits
    21 adds in trunk

Allow setting the prototype of cross-origin objects, as long as they don't change
​https://bugs.webkit.org/show_bug.cgi?id=169787

Reviewed by Mark Lam.

LayoutTests/imported/w3c:

Import test coverage from W3C web-platform-tests.

  • web-platform-tests/common/domain-setter.sub.html: Added.
  • web-platform-tests/common/test-setting-immutable-prototype.js: Added.

(self.testSettingImmutablePrototypeToNewValueOnly.test):
(self.testSettingImmutablePrototypeToNewValueOnly):
(self.testSettingImmutablePrototype.else.test):
(self.testSettingImmutablePrototype):

  • web-platform-tests/common/w3c-import.log:
  • web-platform-tests/html/browsers/history/the-location-interface/location-prototype-setting-cross-origin-domain.sub-expected.txt: Added.
  • web-platform-tests/html/browsers/history/the-location-interface/location-prototype-setting-cross-origin-domain.sub.html: Added.
  • web-platform-tests/html/browsers/history/the-location-interface/location-prototype-setting-goes-cross-origin-domain.sub-expected.txt: Added.
  • web-platform-tests/html/browsers/history/the-location-interface/location-prototype-setting-goes-cross-origin-domain.sub.html: Added.
  • web-platform-tests/html/browsers/history/the-location-interface/location-prototype-setting-same-origin-domain.sub-expected.txt: Added.
  • web-platform-tests/html/browsers/history/the-location-interface/location-prototype-setting-same-origin-domain.sub.html: Added.
  • web-platform-tests/html/browsers/history/the-location-interface/location-prototype-setting-same-origin-expected.txt: Added.
  • web-platform-tests/html/browsers/history/the-location-interface/location-prototype-setting-same-origin.html: Added.
  • web-platform-tests/html/browsers/history/the-location-interface/w3c-import.log:
  • web-platform-tests/html/browsers/the-windowproxy-exotic-object/w3c-import.log: Added.
  • web-platform-tests/html/browsers/the-windowproxy-exotic-object/windowproxy-prototype-setting-cross-origin-domain.sub-expected.txt: Added.
  • web-platform-tests/html/browsers/the-windowproxy-exotic-object/windowproxy-prototype-setting-cross-origin-domain.sub.html: Added.
  • web-platform-tests/html/browsers/the-windowproxy-exotic-object/windowproxy-prototype-setting-goes-cross-origin-domain.sub-expected.txt: Added.
  • web-platform-tests/html/browsers/the-windowproxy-exotic-object/windowproxy-prototype-setting-goes-cross-origin-domain.sub.html: Added.
  • web-platform-tests/html/browsers/the-windowproxy-exotic-object/windowproxy-prototype-setting-same-origin-domain.sub-expected.txt: Added.
  • web-platform-tests/html/browsers/the-windowproxy-exotic-object/windowproxy-prototype-setting-same-origin-domain.sub.html: Added.
  • web-platform-tests/html/browsers/the-windowproxy-exotic-object/windowproxy-prototype-setting-same-origin-expected.txt: Added.
  • web-platform-tests/html/browsers/the-windowproxy-exotic-object/windowproxy-prototype-setting-same-origin.html: Added.

Source/JavaScriptCore:

  • runtime/JSGlobalObject.h:

Mark JS global object as an immutable prototype exotic object to match Window.

  • runtime/JSObject.cpp:

(JSC::JSObject::setPrototypeWithCycleCheck):
Update setPrototypeWithCycleCheck() for immutable prototype exotic objects in order
to align with:

In particular, we need to call GetPrototypeOf and return true if it returns the same
value as the new prototype. We really need to call GetPrototypeOf and not merely
getting the prototype slot via getPrototypeDirect() since Location and Window override
GetPrototypeOf to return null in the cross-origin case.

  • runtime/JSProxy.cpp:

(JSC::JSProxy::setPrototype):
Update JSProxy::setPrototype() to forward such calls to its target. This is needed so
we end up calling JSObject::setPrototypeWithCycleCheck() for the Window object.
Handling immutable prototype exotic objects in that method does the right thing for
Window.

Source/WebCore:

Allow setting the prototype of cross-origin objects, as long as they don't change:

Tests: imported/w3c/web-platform-tests/common/domain-setter.sub.html

imported/w3c/web-platform-tests/html/browsers/history/the-location-interface/location-prototype-setting-cross-origin-domain.sub.html
imported/w3c/web-platform-tests/html/browsers/history/the-location-interface/location-prototype-setting-goes-cross-origin-domain.sub.html
imported/w3c/web-platform-tests/html/browsers/history/the-location-interface/location-prototype-setting-same-origin-domain.sub.html
imported/w3c/web-platform-tests/html/browsers/history/the-location-interface/location-prototype-setting-same-origin.html
imported/w3c/web-platform-tests/html/browsers/the-windowproxy-exotic-object/windowproxy-prototype-setting-cross-origin-domain.sub.html
imported/w3c/web-platform-tests/html/browsers/the-windowproxy-exotic-object/windowproxy-prototype-setting-goes-cross-origin-domain.sub.html
imported/w3c/web-platform-tests/html/browsers/the-windowproxy-exotic-object/windowproxy-prototype-setting-same-origin-domain.sub.html
imported/w3c/web-platform-tests/html/browsers/the-windowproxy-exotic-object/windowproxy-prototype-setting-same-origin.html

  • bindings/js/JSLocationCustom.cpp:
  • page/Location.idl:

Drop custom setPrototype() implementation for Location and mark Location as an immutable
prototype exotic object now that JSC does the right thing for those.

LayoutTests:

  • http/tests/security/cross-frame-access-object-setPrototypeOf-expected.txt:
  • http/tests/security/cross-frame-access-object-setPrototypeOf.html:
  • js/dom/setPrototypeOf-location-window-expected.txt:
  • js/dom/setPrototypeOf-location-window.html:

Add checks for setting prototype of window and location objects to the same value
to make sure it does not throw, both in the same origin and the cross origin
cases.

  • fast/dom/Window/window-custom-prototype-crash-expected.txt:
  • http/tests/security/xss-DENIED-regular-propterty-with-iframe-proto-expected.txt:
  • js/object-literal-shorthand-construction-expected.txt:

Rebaseline tests now that the exception message is a bit different.

  • js/script-tests/prototype-assignment.js:

Remove special casing for Window, it now behaves like a regular immutable prototype
exotic object.

Mar 17, 2017:

9:32 PM Changeset in webkit [214134] by commit-queue@webkit.org
  • 6 edits in trunk/Source/WebCore

Stop sending media data as soon as peer connection is closed
​https://bugs.webkit.org/show_bug.cgi?id=169813

Patch by Youenn Fablet <​youenn@apple.com> on 2017-03-17
Reviewed by Eric Carlson.

Stopping sending of media data as soon as the peer connection is closed.
By stopping sending at source level, this may stop by a few frames/chunks earlier than before.
Making sure RealtimeOutgoingVideoSource is stopping observing the media source when being stopped and deleted.

  • Modules/mediastream/libwebrtc/LibWebRTCPeerConnectionBackend.cpp:

(WebCore::LibWebRTCPeerConnectionBackend::doStop):

  • platform/mediastream/mac/RealtimeOutgoingAudioSource.cpp:

(WebCore::RealtimeOutgoingAudioSource::stop):

  • platform/mediastream/mac/RealtimeOutgoingAudioSource.h:
  • platform/mediastream/mac/RealtimeOutgoingVideoSource.cpp:

(WebCore::RealtimeOutgoingVideoSource::stop):

  • platform/mediastream/mac/RealtimeOutgoingVideoSource.h:
8:49 PM Changeset in webkit [214133] by webkit@devinrousso.com
  • 4 edits in trunk/Source/WebInspectorUI

Web Inspector: wrong filename when saving resource from ResourceImageContentView
​https://bugs.webkit.org/show_bug.cgi?id=54454

Reviewed by Joseph Pecoraro.

  • UserInterface/Views/ResourceContentView.js:

(WebInspector.ResourceContentView.prototype.get supportsSave):
(WebInspector.ResourceContentView.prototype.get saveData):

  • UserInterface/Views/TextResourceContentView.js:

(WebInspector.TextResourceContentView.prototype.get supportsSave): Deleted.
Add save data getter to the parent class, so that both Images and Fonts can be saved.

  • UserInterface/Views/Resource.js:

(WebInspector.Resource.prototype.canRequestContent):
Drive-by: remove unused function.

7:36 PM Changeset in webkit [214132] by commit-queue@webkit.org
  • 13 edits
    1 copy
    1 add in trunk

Implement incoming webrtc data based on tracksCurr
​https://bugs.webkit.org/show_bug.cgi?id=169836

Patch by Youenn Fablet <​youenn@apple.com> on 2017-03-17
Reviewed by Eric Carlson.

Source/WebCore:

Test: webrtc/video-with-receiver.html

Constructing incoming tracks based on libwebrtc OnAddTrack.
Constructing incoming media streams based on libwebrtc OnAddStream.
Firing only addstream if legacy API flag is on.

Ensuring that media stream and media stream tracks relationship is still correctly implemented.
For that, we keep a map that relates libwebrtc media streams with WebCore media streams.
Adding the ability to get a receiver related to the track on the track event.

Implementing the possibility to create a transceiver ahead of track arrival time.
Created transceivers that are not related to any real source are kept in the peer connection back end.
When a libwebrtc track is appearing, it is associated with the track source of the corresponding transceiver based on track type.

Added the ability to create empty real time sources and set their data libwebrtc track when being available.

  • Modules/mediastream/MediaStream.cpp:

(WebCore::MediaStream::addTrackFromPlatform):

  • Modules/mediastream/MediaStream.h:
  • Modules/mediastream/libwebrtc/LibWebRTCMediaEndpoint.cpp:

(WebCore::LibWebRTCMediaEndpoint::mediaStreamFromRTCStream):
(WebCore::LibWebRTCMediaEndpoint::addRemoteStream):
(WebCore::LibWebRTCMediaEndpoint::addRemoteTrack):
(WebCore::LibWebRTCMediaEndpoint::removeRemoteStream):
(WebCore::LibWebRTCMediaEndpoint::OnAddStream):
(WebCore::LibWebRTCMediaEndpoint::OnRemoveStream):
(WebCore::LibWebRTCMediaEndpoint::OnAddTrack):
(WebCore::LibWebRTCMediaEndpoint::stop):
(WebCore::createMediaStreamTrack): Deleted.
(WebCore::LibWebRTCMediaEndpoint::addStream): Deleted.

  • Modules/mediastream/libwebrtc/LibWebRTCMediaEndpoint.h:
  • Modules/mediastream/libwebrtc/LibWebRTCPeerConnectionBackend.cpp:

(WebCore::LibWebRTCPeerConnectionBackend::doStop):
(WebCore::createReceiverForSource):
(WebCore::createEmptySource):
(WebCore::LibWebRTCPeerConnectionBackend::createReceiver):
(WebCore::LibWebRTCPeerConnectionBackend::videoReceiver):
(WebCore::LibWebRTCPeerConnectionBackend::audioReceiver):
(WebCore::LibWebRTCPeerConnectionBackend::removeRemoteStream):
(WebCore::LibWebRTCPeerConnectionBackend::addRemoteStream):

  • Modules/mediastream/libwebrtc/LibWebRTCPeerConnectionBackend.h:
  • platform/mediastream/mac/RealtimeIncomingAudioSource.cpp:

(WebCore::RealtimeIncomingAudioSource::setSourceTrack):

  • platform/mediastream/mac/RealtimeIncomingAudioSource.h:
  • platform/mediastream/mac/RealtimeIncomingVideoSource.cpp:

(WebCore::RealtimeIncomingVideoSource::setSourceTrack):

  • platform/mediastream/mac/RealtimeIncomingVideoSource.h:

LayoutTests:

  • webrtc/video-with-receiver-expected.txt: Added.
  • webrtc/video-with-receiver.html: Copied from LayoutTests/webrtc/video.html.
  • webrtc/video.html:
5:39 PM Changeset in webkit [214131] by Ryan Haddad
  • 2 edits in trunk/LayoutTests

Remove TestExpectation for a test that is no longer in the tree.

Unreviewed test gardening.

  • platform/mac/TestExpectations:
5:34 PM Changeset in webkit [214130] by eric.carlson@apple.com
  • 2 edits in trunk/Source/WebCore

[MediaStream] Compensate for video capture orientation
​https://bugs.webkit.org/show_bug.cgi?id=169313
<rdar://problem/30994785>

Unreviewed, fix a crash in r214120 found by the bots.

  • platform/graphics/avfoundation/objc/MediaPlayerPrivateMediaStreamAVFObjC.mm:

(-[WebAVSampleBufferStatusChangeListener invalidate]): Drive-by removal of unnecessary code.
(-[WebAVSampleBufferStatusChangeListener observeValueForKeyPath:ofObject:change:context:]):
Actually retain self before callOnMainThread.
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::~MediaPlayerPrivateMediaStreamAVFObjC):
Invalidate the change listener first.

4:40 PM Changeset in webkit [214129] by dino@apple.com
  • 3 edits
    2 adds in trunk

MediaQueryList instances do not update for prefers-reduced-motion
​https://bugs.webkit.org/show_bug.cgi?id=169835
<rdar://problem/30339378>

Reviewed by Anders Carlsson.

Source/WebCore:

Tell the document it needs to evaluate the MediaQueryLists
if accessibility settings change.

Test: fast/media/mq-prefers-reduced-motion-matchMedia.html

  • page/Page.cpp:

(WebCore::Page::accessibilitySettingsDidChange):

LayoutTests:

  • fast/media/mq-prefers-reduced-motion-matchMedia-expected.txt: Added.
  • fast/media/mq-prefers-reduced-motion-matchMedia.html: Added.
4:33 PM Changeset in webkit [214128] by Ryan Haddad
  • 2 edits in trunk/LayoutTests

Mark media/modern-media-controls/airplay-placard/airplay-placard-text-section.html as flaky.
​https://bugs.webkit.org/show_bug.cgi?id=169654

Unreviewed test gardening.

  • platform/mac-wk1/TestExpectations:
4:23 PM Changeset in webkit [214127] by beidson@apple.com
  • 11 edits in trunk/Source

Make HTTPCookieStorage operate in the UIProcess in absence of a WebProcessPool.
<rdar://problem/31102330> and ​https://bugs.webkit.org/show_bug.cgi?id=169797

Reviewed by Alex Christensen.

Source/WebCore:

No new tests (No behavior change).

  • platform/network/NetworkStorageSession.h:
  • platform/network/cocoa/NetworkStorageSessionCocoa.mm:

(WebCore::NetworkStorageSession::flushCookieStore):

  • platform/network/soup/NetworkStorageSessionSoup.cpp:

(WebCore::NetworkStorageSession::flushCookieStore):

Source/WebKit2:

  • UIProcess/API/APIHTTPCookieStore.cpp:

(API::HTTPCookieStore::~HTTPCookieStore):
(API::HTTPCookieStore::cookies):
(API::HTTPCookieStore::setCookie):
(API::HTTPCookieStore::deleteCookie):
(API::HTTPCookieStore::registerObserver):
(API::HTTPCookieStore::unregisterObserver):
(API::HTTPCookieStore::cookieManagerDestroyed):
(API::HTTPCookieStore::registerForNewProcessPoolNotifications):

  • UIProcess/API/APIHTTPCookieStore.h:
  • UIProcess/WebProcessPool.cpp:

(WebKit::WebProcessPool::create):
(WebKit::WebProcessPool::notifyThisWebProcessPoolWasCreated):
(WebKit::m_hiddenPageThrottlingTimer):

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

(WebKit::WebsiteDataStore::processPoolForCookieStorageOperations):
(WebKit::WebsiteDataStore::processPoolForCookieStorageNotifications): Deleted.

  • UIProcess/WebsiteData/WebsiteDataStore.h:
2:59 PM Changeset in webkit [214126] by hyatt@apple.com
  • 4 edits
    2 adds in trunk

Disable per-region boxes for multicolumn
​https://bugs.webkit.org/show_bug.cgi?id=169830

Reviewed by Zalan Bujtas.

Source/WebCore:

Test: fast/multicol/float-adjacent-to-overflow-block.html

  • rendering/RenderBlockFlow.cpp:

(WebCore::RenderBlockFlow::determineLogicalLeftPositionForChild):

  • rendering/RenderBox.cpp:

(WebCore::RenderBox::borderBoxRectInRegion):
(WebCore::RenderBox::renderBoxRegionInfo):
Limit all of the per-region box code to RenderNamedFlowThreads.
This code should never be used by multicolumn layout.

LayoutTests:

  • fast/multicol/float-adjacent-to-overflow-block-expected.html: Added.
  • fast/multicol/float-adjacent-to-overflow-block.html: Added.
2:51 PM Changeset in webkit [214125] by Said Abou-Hallawa
  • 7 edits in trunk/Source

Time channel attack on SVG Filters
​https://bugs.webkit.org/show_bug.cgi?id=118689

Reviewed by Simon Fraser.

Source/WebCore:

The time channel attack can happen if the attacker applies FEColorMatrix
or FEConvolveMatrix and provides a matrix which is filled with subnormal
floating point values. Performing floating-point operations on subnormals
is very expensive unless the pixel in the source graphics is black (or
zero). By measuring the time a filter takes to be applied, the attacker
can know whether the pixel he wants to steal from an iframe is black or
white. By repeating the same process on all the pixels in the iframe, the
attacker can reconstruct the whole page of the iframe.

To fix this issue, the values in the matrices of these filters will clamped
to FLT_MIN. We do not want to consume too much time calculating filtered
pixels because of such tiny values. The difference between applying FLT_MIN
and applying a subnormal should not be even noticeable. Normalizing the
floating-point matrices should happen only at the beginning of the filter
platformApplySoftware().

  • platform/graphics/filters/FEColorMatrix.cpp:

(WebCore::FEColorMatrix::platformApplySoftware):

  • platform/graphics/filters/FEConvolveMatrix.cpp:

(WebCore::FEConvolveMatrix::fastSetInteriorPixels):
(WebCore::FEConvolveMatrix::fastSetOuterPixels):
(WebCore::FEConvolveMatrix::platformApplySoftware):

  • platform/graphics/filters/FEConvolveMatrix.h:
  • platform/graphics/filters/FilterEffect.h:

(WebCore::FilterEffect::normalizedFloats):

Source/WTF:

Performing arithmetic operations on subnormal floating-point numbers is
very expensive. Normalizing the floating-point number to the minimum normal
value should accelerate the calculations and there won't be a noticeable
difference in the result since all the subnormal values and the minimum
normal value are all very close to zero.

  • wtf/MathExtras.h:

(normalizedFloat):

2:31 PM Changeset in webkit [214124] by jer.noble@apple.com
  • 2 edits in trunk/Source/WebCore

Explicitly resize the audio buffer in RealtimeOutgoingAudioSource.
​https://bugs.webkit.org/show_bug.cgi?id=169814
<rdar://problem/30975496>

Reviewed by Eric Carlson.

Rather than just reserving sufficient capacity, actually change the buffer's underlying size to match.

  • platform/mediastream/mac/RealtimeOutgoingAudioSource.cpp:

(WebCore::RealtimeOutgoingAudioSource::pullAudioData):

2:28 PM Changeset in webkit [214123] by Chris Dumez
  • 2 edits in trunk/Source/WebKit2

[WK2] Disable background tab responsiveness detection on iOS
​https://bugs.webkit.org/show_bug.cgi?id=169824
<rdar://problem/30913776>

Reviewed by Dan Bernstein.

Disable background tab responsiveness detection on iOS where background tabs
often get suspended. This feature was not tested or meant to be enabled on
iOS, this was an oversight on my part.

  • UIProcess/BackgroundProcessResponsivenessTimer.cpp:

(WebKit::BackgroundProcessResponsivenessTimer::shouldBeActive):

2:18 PM Changeset in webkit [214122] by jiewen_tan@apple.com
  • 12 edits
    1 copy
    5 adds in trunk

[WebCrypto] Make sure all CryptoKey classes are structured clonable
​https://bugs.webkit.org/show_bug.cgi?id=169232
<rdar://problem/31106660>

Reviewed by Brent Fulgham.

Source/WebCore:

This patch does the following few things: 1) It makes CryptoKeyEC and CryptoKeyRaw
structured clonable; 2) It makes CryptoKeyAES and CryptoKeyHMAC to use a move importer
during deserialization; 3) It adds some tests to make sure CryptoKeyAES, CryptoKeyHMAC
and CryptoKeyRSA won't lose their internal data during serialization/deserialization.

Tests: crypto/workers/subtle/ec-postMessage-worker.html

crypto/workers/subtle/raw-postMessage-worker.html

  • bindings/js/SerializedScriptValue.cpp:

(WebCore::CloneSerializer::write):
(WebCore::CloneDeserializer::readHMACKey):
(WebCore::CloneDeserializer::readAESKey):
(WebCore::CloneDeserializer::readECKey):
(WebCore::CloneDeserializer::readRawKey):
(WebCore::CloneDeserializer::readCryptoKey):

  • crypto/keys/CryptoKeyEC.cpp:

(WebCore::CryptoKeyEC::namedCurveString):
(WebCore::CryptoKeyEC::isValidECAlgorithm):

  • crypto/keys/CryptoKeyEC.h:

LayoutTests:

  • crypto/workers/subtle/aes-postMessage-worker-expected.txt:
  • crypto/workers/subtle/aes-postMessage-worker.html:
  • crypto/workers/subtle/ec-postMessage-worker-expected.txt: Added.
  • crypto/workers/subtle/ec-postMessage-worker.html: Added.
  • crypto/workers/subtle/hmac-postMessage-worker-expected.txt:
  • crypto/workers/subtle/hmac-postMessage-worker.html:
  • crypto/workers/subtle/raw-postMessage-worker-expected.txt: Added.
  • crypto/workers/subtle/raw-postMessage-worker.html: Added.
  • crypto/workers/subtle/resources/ec-postMessage-worker.js: Added.
  • crypto/workers/subtle/resources/raw-postMessage-worker.js: Added.
  • crypto/workers/subtle/resources/rsa-postMessage-worker.js:
  • crypto/workers/subtle/rsa-postMessage-worker-expected.txt:
  • crypto/workers/subtle/rsa-postMessage-worker.html:
2:16 PM Changeset in webkit [214121] by msaboff@apple.com
  • 12 edits in trunk

Use USE_INTERNAL_SDK to compute ENABLE_FAST_JIT_PERMISSIONS instead of HAVE_INTERNAL_SDK
​https://bugs.webkit.org/show_bug.cgi?id=169817

Reviewed by Filip Pizlo.

Source/JavaScriptCore:

  • Configurations/FeatureDefines.xcconfig:

Source/WebCore:

  • Configurations/FeatureDefines.xcconfig:

Source/WebCore/PAL:

  • Configurations/FeatureDefines.xcconfig:

Source/WebKit/mac:

  • Configurations/FeatureDefines.xcconfig:

Source/WebKit2:

  • Configurations/FeatureDefines.xcconfig:

Tools:

  • TestWebKitAPI/Configurations/FeatureDefines.xcconfig:
2:14 PM Changeset in webkit [214120] by eric.carlson@apple.com
  • 9 edits in trunk/Source/WebCore

[MediaStream] Compensate for video capture orientation
​https://bugs.webkit.org/show_bug.cgi?id=169313
<rdar://problem/30994785>

Reviewed by Jer Noble.

No new tests, the mock video source doesn't support rotation. Test will be added when this
is fixed in ​https://bugs.webkit.org/show_bug.cgi?id=169822.

Add 'orientation' and 'mirrored' attributes to MediaSample

  • platform/MediaSample.h:

(WebCore::MediaSample::videoOrientation):
(WebCore::MediaSample::videoMirrored):

  • platform/graphics/avfoundation/MediaSampleAVFObjC.h:

A video sample can be rotated and/or mirrored, so the video layer may need to be rotated
and resized for display. We don't want expose this information to the renderer, so allocate
return a generic CALayer as the player's platforLayer, and add the video layer as a sublayer
so we can adjust it to display correctly. Add an enum for playback state as well as display
mode so we correctly display a black frame when video frames are available but playback has
not yet started.

  • platform/graphics/avfoundation/objc/MediaPlayerPrivateMediaStreamAVFObjC.h:
  • platform/graphics/avfoundation/objc/MediaPlayerPrivateMediaStreamAVFObjC.mm:

(-[WebAVSampleBufferStatusChangeListener initWithParent:]):
(-[WebAVSampleBufferStatusChangeListener invalidate]):
(-[WebAVSampleBufferStatusChangeListener beginObservingLayers]):
(-[WebAVSampleBufferStatusChangeListener stopObservingLayers]): Ditto.
(-[WebAVSampleBufferStatusChangeListener observeValueForKeyPath:ofObject:change:context:]):
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::videoTransformationMatrix):
(WebCore::runWithoutAnimations):
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::enqueueVideoSample):
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::layerErrorDidChange):
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::ensureLayers):
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::destroyLayers):
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::cancelLoad):
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::platformLayer):
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::displayLayer):
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::backgroundLayer):
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::currentDisplayMode):
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::updateDisplayMode):
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::play):
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::pause):
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::currentMediaTime):
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::activeStatusChanged):
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::updateRenderingMode):
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::checkSelectedVideoTrack):
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::paintCurrentFrameInContext):
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::acceleratedRenderingStateChanged):
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::setShouldBufferData):
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::backgroundLayerBoundsChanged):
(-[WebAVSampleBufferStatusChangeListener beginObservingLayer:]): Deleted.
(-[WebAVSampleBufferStatusChangeListener stopObservingLayer:]): Deleted.
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::ensureLayer): Deleted.
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::destroyLayer): Deleted.
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::paused): Deleted.
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::shouldBePlaying): Deleted.

  • platform/mediastream/mac/AVVideoCaptureSource.h:
  • platform/mediastream/mac/AVVideoCaptureSource.mm:

(WebCore::AVVideoCaptureSource::processNewFrame): Add connection parameter so we can get the
video orientation.
(WebCore::AVVideoCaptureSource::captureOutputDidOutputSampleBufferFromConnection):

Pass sample orientation to libwebrtc.

  • platform/mediastream/mac/RealtimeOutgoingVideoSource.cpp:

(WebCore::RealtimeOutgoingVideoSource::sendFrame):
(WebCore::RealtimeOutgoingVideoSource::videoSampleAvailable):

  • platform/mediastream/mac/RealtimeOutgoingVideoSource.h:
2:12 PM Changeset in webkit [214119] by Alan Bujtas
  • 3 edits
    2 adds in trunk

Fix the flow thread state on the descendants of out of flow positioned replaced elements.
​https://bugs.webkit.org/show_bug.cgi?id=169821
<rdar://problem/30964017>

Reviewed by Simon Fraser.

Source/WebCore:

Descendants of a replaced out of flow elmement should inherit the flowthread state
from the replaced element and not from the replaced element's parent.

Test: fast/multicol/fix-inherit-when-container-is-replaced.html

  • rendering/RenderObject.cpp:

(WebCore::RenderObject::computedFlowThreadState):

LayoutTests:

  • fast/multicol/fix-inherit-when-container-is-replaced-expected.txt: Added.
  • fast/multicol/fix-inherit-when-container-is-replaced.html: Added.
1:59 PM Changeset in webkit [214118] by Jon Davis
  • 2 edits in trunk/Source/WebCore

Update feature status for outdated entries: CSS Shapes, Force Click Events
​https://bugs.webkit.org/show_bug.cgi?id=169829

Reviewed by Matt Baker.

  • features.json:
1:45 PM Changeset in webkit [214117] by achristensen@apple.com
  • 2 edits in trunk/Source/WebCore

Make SocketStreamError a private header.
​https://bugs.webkit.org/show_bug.cgi?id=169831

  • WebCore.xcodeproj/project.pbxproj:

I'll need this from WebKit2.

1:45 PM Changeset in webkit [214116] by commit-queue@webkit.org
  • 2 edits in trunk/LayoutTests

Update webrtc tests expectations
​https://bugs.webkit.org/show_bug.cgi?id=169828

Unreviewed.

Patch by Youenn Fablet <​youenn@apple.com> on 2017-03-17

1:42 PM Changeset in webkit [214115] by achristensen@apple.com
  • 2 edits in trunk/Source/WebCore

Add encoder for SocketStreamError
​https://bugs.webkit.org/show_bug.cgi?id=169831

Reviewed by Andy Estes.

This isn't used yet, but we'll soon need to send these across IPC.

  • platform/network/SocketStreamError.h:

(WebCore::SocketStreamError::encode):
(WebCore::SocketStreamError::decode):

1:30 PM Changeset in webkit [214114] by achristensen@apple.com
  • 11 edits
    1 add in trunk/Source/WebCore

Make SocketStreamHandle virtual functions asynchronous
​https://bugs.webkit.org/show_bug.cgi?id=169818

Reviewed by Andy Estes.

No change in behavior. Instead of returning immediately, call a completion handler.
This is in preparation for making them truly asynchronous.

I still need to be able to synchronously get the number of buffered bytes, but the
buffer itself will soon be in the NetworkProcess with a new subclass of SocketStreamHandle
that messages across IPC. The number of buffered bytes will still be stored in the WebProcess,
and when it updates, the message from SocketStreamHandleClient::didUpdateBufferedAmount will update it.

  • CMakeLists.txt:
  • WebCore.xcodeproj/project.pbxproj:
  • platform/network/SocketStreamHandle.cpp:

(WebCore::SocketStreamHandle::send):
(WebCore::SocketStreamHandle::close):
(WebCore::SocketStreamHandle::sendPendingData): Deleted.

  • platform/network/SocketStreamHandle.h:

(WebCore::SocketStreamHandle::bufferedAmount): Deleted.

  • platform/network/SocketStreamHandleImpl.cpp: Added.

(WebCore::SocketStreamHandleImpl::platformSend):
(WebCore::SocketStreamHandleImpl::sendPendingData):
(WebCore::SocketStreamHandleImpl::bufferedAmount):
m_buffer was moved from SocketStreamHandle to SocketStreamHandleImpl, so some of the logic must move with it.
I moved as much logic that was shared in the superclass to a new shared location for code shared among the subclass implementations.

  • platform/network/cf/SocketStreamHandleImpl.h:
  • platform/network/cf/SocketStreamHandleImplCFNet.cpp:

(WebCore::SocketStreamHandleImpl::platformSendInternal):
(WebCore::SocketStreamHandleImpl::platformSend): Deleted.

  • platform/network/curl/SocketStreamHandleImpl.h:
  • platform/network/curl/SocketStreamHandleImplCurl.cpp:

(WebCore::SocketStreamHandleImpl::platformSendInternal):
(WebCore::SocketStreamHandleImpl::platformSend): Deleted.

  • platform/network/soup/SocketStreamHandleImpl.h:
  • platform/network/soup/SocketStreamHandleImplSoup.cpp:

(WebCore::SocketStreamHandleImpl::platformSendInternal):
(WebCore::SocketStreamHandleImpl::platformSend): Deleted.

12:52 PM Changeset in webkit [214113] by Antti Koivisto
  • 33 edits
    3 adds in trunk

Add a reload policy where only expired subresources are revalidated
​https://bugs.webkit.org/show_bug.cgi?id=169756

Reviewed by Andreas Kling.

Source/WebCore:

Test: http/tests/cache/reload-expired-only.html

The default reload behavior revalidates all resources on the page. This patch adds
a new policy that revalidates expired (and uncacheable) resources only. Using this
policy can speed up reloads significantly and also reduce network traffic and server
load.

  • history/PageCache.cpp:

(WebCore::canCachePage):

  • inspector/InspectorPageAgent.cpp:

(WebCore::InspectorPageAgent::reload):

  • loader/FrameLoader.cpp:

(WebCore::isBackForwardLoadType):
(WebCore::isReload):

Add a helper function.

(WebCore::FrameLoader::loadURL):
(WebCore::logNavigation):
(WebCore::FrameLoader::reload):

Replace bool arguments with OptionSet<ReloadOption>. Most of the patch is about adapting to this.

(WebCore::FrameLoader::transitionToCommitted):
(WebCore::FrameLoader::subresourceCachePolicy):

Return CachePolicyVerify for 'ReloadExpiredOnly' frame load type. This is the substantive
change that causes the behavior difference.

(WebCore::FrameLoader::checkLoadCompleteForThisFrame):
(WebCore::FrameLoader::defaultRequestCachingPolicy):

Flip the isMainResource condition for better readability.

(WebCore::FrameLoader::shouldPerformFragmentNavigation):
(WebCore::FrameLoader::loadDifferentDocumentItem):

  • loader/FrameLoader.h:

(WebCore::FrameLoader::reload):

  • loader/FrameLoaderTypes.h:
  • loader/HistoryController.cpp:

(WebCore::HistoryController::restoreDocumentState):

  • page/DiagnosticLoggingKeys.cpp:

(WebCore::DiagnosticLoggingKeys::reloadRevalidatingExpiredKey):

  • page/DiagnosticLoggingKeys.h:
  • replay/UserInputBridge.cpp:

(WebCore::UserInputBridge::reloadFrame):

  • replay/UserInputBridge.h:
  • testing/Internals.cpp:

(WebCore::Internals::forceReload):
(WebCore::Internals::reloadExpiredOnly):

Testing support.

  • testing/Internals.h:
  • testing/Internals.idl:

Source/WebKit/mac:

  • WebView/WebFrame.mm:

(toWebFrameLoadType):
(-[WebFrame reload]):
(-[WebFrame reloadFromOrigin]):

Source/WebKit2:

  • UIProcess/API/C/WKPage.cpp:

(WKPageReload):
(WKPageReloadWithoutContentBlockers):
(WKPageReloadFromOrigin):
(WKPageReloadExpiredOnly):

  • UIProcess/API/C/WKPage.h:
  • UIProcess/API/Cocoa/WKBrowsingContextController.mm:

(-[WKBrowsingContextController reload]):
(-[WKBrowsingContextController reloadFromOrigin]):

  • UIProcess/API/Cocoa/WKWebView.mm:

(-[WKWebView reload]):
(-[WKWebView reloadFromOrigin]):
(-[WKWebView _reloadWithoutContentBlockers]):
(-[WKWebView _reloadExpiredOnly]):

  • UIProcess/API/Cocoa/WKWebViewPrivate.h:
  • UIProcess/Automation/WebAutomationSession.cpp:

(WebKit::WebAutomationSession::reloadBrowsingContext):

  • UIProcess/WebFrameProxy.cpp:

(WebKit::WebFrameProxy::didHandleContentFilterUnblockNavigation):

  • UIProcess/WebPageProxy.cpp:

(WebKit::WebPageProxy::reload):

  • UIProcess/WebPageProxy.h:
  • WebProcess/WebPage/WebPage.cpp:

(WebKit::WebPage::reload):
(WebKit::shouldReuseCommittedSandboxExtension):

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

LayoutTests:

  • http/tests/cache/reload-expired-only-expected.txt: Added.
  • http/tests/cache/reload-expired-only.html: Added.
  • http/tests/cache/resources/document-with-cached-unique-script.html: Added.
  • http/tests/cache/resources/random-cached.cgi:
12:48 PM Changeset in webkit [214112] by n_wang@apple.com
  • 4 edits
    4 adds in trunk

AX: VoiceOver no longer works corectly with editable text in the web
​https://bugs.webkit.org/show_bug.cgi?id=169801

Reviewed by Chris Fleizach.

Source/WebCore:

The issue is that since we are entering text controls when getting the text marker,
the end text marker of the input tag might associate to the placeholder div instead of
the input tag itself.
Fixed the issue by checking the shadow elements using shadowHost() instead of isShadowRoot().
Also, only enter text controls when navigating by characters/indexes, so that both Mac and
iOS will get the correct text marker ranges for shadow host elements.

Tests: accessibility/ios-simulator/element-text-range-for-text-control.html

accessibility/mac/text-markers-for-input-with-placeholder.html

  • accessibility/AXObjectCache.cpp:

(WebCore::AXObjectCache::traverseToOffsetInRange):
(WebCore::AXObjectCache::startOrEndCharacterOffsetForRange):
(WebCore::AXObjectCache::characterOffsetForIndex):

  • accessibility/AXObjectCache.h:

LayoutTests:

  • accessibility/ios-simulator/element-text-range-for-text-control-expected.txt: Added.
  • accessibility/ios-simulator/element-text-range-for-text-control.html: Added.
  • accessibility/mac/text-markers-for-input-with-placeholder-expected.txt: Added.
  • accessibility/mac/text-markers-for-input-with-placeholder.html: Added.
12:30 PM Changeset in webkit [214111] by Wenson Hsieh
  • 11 edits in trunk/Source

[WK1] Support animated transitions when performing a data interaction operation
​https://bugs.webkit.org/show_bug.cgi?id=169805
<rdar://problem/31045767>

Reviewed by Beth Dakin and Megan Gardner.

Source/WebCore:

When performing a snapshot excluding the current selection, use ScrollableArea::visibleContentRect to compute
the visible content rect rather than FrameView::visualViewportRect. In WebKit1, the latter will return the
entire content view's bounds.

  • page/TextIndicator.cpp:

(WebCore::takeSnapshots):

Source/WebKit/mac:

Adds WebKit1 support for animated transitions when performing a data interaction operation. See per-method changes
for more details.

  • DOM/DOMElement.h:
  • DOM/DOMElement.mm:

(-[DOMElement boundsInRootViewSpace]):

Introduce -[DOMElement boundsInRootViewSpace], which returns the bounds of the element in root view coordinates.

  • WebCoreSupport/WebDragClient.h:

(): Deleted.

  • WebCoreSupport/WebDragClient.mm:

(WebDragClient::didConcludeEditDrag):

Add boilerplate plumbing.

  • WebView/WebView.mm:

(-[WebUITextIndicatorData initWithImage:textIndicatorData:scale:]):

Add contentImageWithoutSelection and contentImageWithoutSelectionRectInRootViewCoordinates to WebUITextIndicatorData.

(floatRectsForCGRectArray):
(-[WebView _createImageWithPlatterForImage:boundingRect:contentScaleFactor:clippingRects:]):

Introduces a SPI helper method in WebView that clients can use to generate an image of a text platter given an
image of some text, as well as clipping rects to apply to the image.

(-[WebView _dataOperationTextIndicator]):

Add SPI to fetch the text indicator used to drive animations for the current data operation.

(-[WebView _didConcludeEditDataInteraction]):

  • WebView/WebViewData.h:
  • WebView/WebViewInternal.h:
  • WebView/WebViewPrivate.h:
11:54 AM Changeset in webkit [214110] by hyatt@apple.com
  • 12 edits
    8 adds in trunk

Initial letter does not paginate properly.
​https://bugs.webkit.org/show_bug.cgi?id=169795
<rdar://problem/23715770>

Reviewed by Zalan Bujtas.

Source/WebCore:

Tests: fast/css-generated-content/initial-letter-pagination-raised.html

fast/css-generated-content/initial-letter-pagination-raised-rl.html
fast/css-generated-content/initial-letter-pagination-sunken.html
fast/css-generated-content/initial-letter-pagination-sunken-rl.html

  • rendering/FloatingObjects.cpp:

(WebCore::FloatingObject::FloatingObject):
(WebCore::FloatingObject::copyToNewContainer):
(WebCore::FloatingObject::cloneForNewParent):
(WebCore::FloatingObject::translationOffsetToAncestor):

  • rendering/FloatingObjects.h:

(WebCore::FloatingObject::setMarginOffset):
(WebCore::FloatingObject::locationOffsetOfBorderBox):
(WebCore::FloatingObject::marginOffset):
Add the margin offset to the FloatingObject so that we can control the
placement of the border box relative to the margin box without relying
on only the CSS-specified margins. This allows for sunken initial letter
FloatingObjects to be created and propagated out to ancestor blocks for painting
without altering the CSS margins of the sunken letter.

  • rendering/RenderBlockFlow.cpp:

(WebCore::RenderBlockFlow::adjustForUnsplittableChild):
Update the logical height for initial letter containers after the initial
letter shifts to the next page, since the lines that follow need to lay out
relative to the initial letter, i.e., they shift down if it shifts down.

(WebCore::RenderBlockFlow::addOverflowFromFloats):
(WebCore::RenderBlockFlow::paintFloats):
(WebCore::RenderBlockFlow::clipOutFloatingObjects):
Patched to use the new helpers on FloatingObject rather than
xPositionForFloatIncludingMargins and yPositionForFloatIncludingMargins.

(WebCore::RenderBlockFlow::computeLogicalLocationForFloat):
Updated to set the placement and margins of the floating object before
returning. The initial letter adjustments have been factored out into their
own function.

(WebCore::RenderBlockFlow::adjustInitialLetterPosition):
A helper function called by computeLogicalLocationForFloat that handles
the placement and margin adjustments for initial letters. Note that sunken
letters no longer "hack" the top margin of the child but instead rely on
the fact that the FloatingObject has its own independent margin offset now
that can be set to a value that is distinct from the CSS supplied margin.

(WebCore::RenderBlockFlow::positionNewFloats):
Revised to no longer do the placement, but to let computeLogicalLocationForFloat
do it. The height is still set here however, since we don't know that until after
the child has received a layout.

(WebCore::RenderBlockFlow::addOverhangingFloats):
(WebCore::RenderBlockFlow::flipFloatForWritingModeForChild):
(WebCore::RenderBlockFlow::hitTestFloats):
(WebCore::RenderBlockFlow::adjustForBorderFit):
Patched to use the new FloatingObject helpers rather than
xPositionForFloatIncludingMargins and yPositionForFloatIncludingMargins.

  • rendering/RenderBlockFlow.h:

(WebCore::RenderBlockFlow::setLogicalMarginsForFloat):
New helper function for setting the margin offsets on a FloatingObject.

(WebCore::RenderBlockFlow::xPositionForFloatIncludingMargin): Deleted.
(WebCore::RenderBlockFlow::yPositionForFloatIncludingMargin): Deleted.
Deleted in favor of new FloatingObject helpers.

  • rendering/RenderBox.cpp:

(WebCore::RenderBox::isUnsplittableForPagination):
Make initial letter unsplittable as far as pagination is concerned.

  • rendering/RootInlineBox.cpp:

(WebCore::RootInlineBox::ascentAndDescentForBox):
(WebCore::RootInlineBox::includeFontForBox):
(WebCore::RootInlineBox::includeGlyphsForBox):
(WebCore::RootInlineBox::fitsToGlyphs):
Turn off the restriction that you can't fit to vertical glyphs. The numbers returned
are fine now. This allows vertical initial letters to work.

LayoutTests:

  • fast/css-generated-content/initial-letter-pagination-raised-expected.html: Added.
  • fast/css-generated-content/initial-letter-pagination-raised.html: Added.
  • fast/css-generated-content/initial-letter-pagination-sunken-expected.html: Added.
  • fast/css-generated-content/initial-letter-pagination-sunken.html: Added.
  • platform/ios-simulator/TestExpectations:
10:58 AM Changeset in webkit [214109] by fpizlo@apple.com
  • 12 edits
    2 adds in trunk/Source

Air should be powerful enough to support Tmp-splitting
​https://bugs.webkit.org/show_bug.cgi?id=169515

Reviewed by Saam Barati.

Source/JavaScriptCore:

In the process of implementing the Tmp-splitting optimization, I made some small
clean-ups. They don't affect anything - it's basically moving code around and adding
utility functions.

  • CMakeLists.txt:
  • JavaScriptCore.xcodeproj/project.pbxproj:
  • assembler/LinkBuffer.cpp:

(JSC::LinkBuffer::allocate): testb3 was sometimes failing its checkDoesNotUseInstruction check because of uninitialized memory. This initializes the internal fragmentation slop of every JIT allocation.

  • b3/air/AirAllocateRegistersByGraphColoring.cpp:
  • b3/air/AirAllocateRegistersByGraphColoring.h:

(JSC::B3::Air::useIRC): It's useful to be able to query which register allocator we're using.

  • b3/air/AirArg.cpp:

(WTF::printInternal):

  • b3/air/AirArg.h:

(JSC::B3::Air::Arg::temperature): The temperature of a role is a useful concept to have factored out.

  • b3/air/AirBreakCriticalEdges.cpp: Added.

(JSC::B3::Air::breakCriticalEdges): I was surprised that we didn't have this already. It's a pretty fundamental CFG utility.

  • b3/air/AirBreakCriticalEdges.h: Added.
  • b3/air/AirGenerate.cpp:
  • b3/air/AirInsertionSet.h: You can't use & if you want copy-constructibility, which seems to be a prerequisite to IndexMap<BasicBlock, InsertionSet>.

(JSC::B3::Air::InsertionSet::InsertionSet):
(JSC::B3::Air::InsertionSet::code):

  • b3/air/AirLiveness.h: Teach Liveness to track only warm liveness.

(JSC::B3::Air::TmpLivenessAdapter::acceptsRole):
(JSC::B3::Air::StackSlotLivenessAdapter::acceptsRole):
(JSC::B3::Air::RegLivenessAdapter::acceptsRole):
(JSC::B3::Air::AbstractLiveness::AbstractLiveness):
(JSC::B3::Air::AbstractLiveness::LocalCalc::execute):

Source/WTF:

Teach IndexMap how to handle types that don't have a default constructor. This makes it
easy to create a IndexMap<BasicBlock, InsertionSet>, which is a super powerful
construct.

  • wtf/IndexMap.h:

(WTF::IndexMap::IndexMap):
(WTF::IndexMap::resize):
(WTF::IndexMap::clear):

10:16 AM Changeset in webkit [214108] by commit-queue@webkit.org
  • 3 edits
    2 adds in trunk

RealtimeIncomingAudioSource is not stopping properly
​https://bugs.webkit.org/show_bug.cgi?id=169807

Patch by Youenn Fablet <​youenn@apple.com> on 2017-03-17
Reviewed by Eric Carlson.

Source/WebCore:

Test: webrtc/release-after-getting-track.html

  • platform/mediastream/mac/RealtimeIncomingAudioSource.cpp:

(WebCore::RealtimeIncomingAudioSource::~RealtimeIncomingAudioSource):
(WebCore::RealtimeIncomingAudioSource::stopProducingData):

LayoutTests:

  • webrtc/release-after-getting-track-expected.txt: Added.
  • webrtc/release-after-getting-track.html: Added.
10:09 AM Changeset in webkit [214107] by bshafiei@apple.com
  • 5 edits in branches/safari-603-branch/Source

Versioning.

8:57 AM Changeset in webkit [214106] by achristensen@apple.com
  • 8 edits in trunk/Source

Use completion handlers instead of return values for sending websocket data
​https://bugs.webkit.org/show_bug.cgi?id=169802

Reviewed by Carlos Garcia Campos.

Source/WebCore:

No change in behavior, but this is a baby step towards making things more asynchronous.

  • Modules/websockets/WebSocketChannel.cpp:

(WebCore::WebSocketChannel::didOpenSocketStream):
(WebCore::WebSocketChannel::processOutgoingFrameQueue):
(WebCore::WebSocketChannel::sendFrame):

  • Modules/websockets/WebSocketChannel.h:
  • platform/network/SocketStreamHandle.cpp:

(WebCore::SocketStreamHandle::send):

  • platform/network/SocketStreamHandle.h:
  • platform/network/cf/SocketStreamHandleImpl.h:

Source/WebKit2:

  • UIProcess/InspectorServer/WebSocketServerConnection.cpp:

(WebKit::WebSocketServerConnection::sendWebSocketMessage):
(WebKit::WebSocketServerConnection::sendHTTPResponseHeader):
(WebKit::WebSocketServerConnection::sendRawData):

7:11 AM Changeset in webkit [214105] by zandobersek@gmail.com
  • 7 edits in trunk/Source/WebCore

WebGPU: build fixes for non-Cocoa platforms
​https://bugs.webkit.org/show_bug.cgi?id=169781

Reviewed by Alex Christensen.

Enable building WebGPU code for other ports and fix various build
failures. This still leaves the GPUDevice::platformLayer() method
missing for other ports, as well as the general implementation of
the WebGPU platform abstractions under platform/graphics/gpu/.

  • CMakeLists.txt: Move the WebGPU build targets here.
  • PlatformMac.cmake:
  • html/canvas/WebGPULibrary.h: Explicitly include the WTFString.h

header to suppress missing WTF::String definition errors.

  • platform/graphics/gpu/GPUBuffer.cpp:

(WebCore::GPUBuffer::contents): Avoid ambiguous calls to
ArrayBuffer::create() by passing nullptr as the first argument.

  • platform/graphics/gpu/GPUDevice.h: Include the Forward.h header

to forward-declare the WTF::String class. Also remove the unused
PlatformGPUDevice typedef for non-Cocoa platforms.
(WebCore::GPUDevice::layer): Move this and the related m_layer
member variable under the PLATFORM(COCOA) guard.

  • platform/graphics/gpu/GPUFunction.cpp:

(WebCore::GPUFunction::GPUFunction): Remove the GPULibrary
parameter checks in this no-op constructor that's used on
non-Cocoa platforms.

6:13 AM Changeset in webkit [214104] by tpopela@redhat.com
  • 5 edits in trunk/Source/WebKit2

Fix uninitialized public members in WebProcess
​https://bugs.webkit.org/show_bug.cgi?id=169604

Reviewed by Michael Catanzaro.

Found by Coverity scan.

  • WebProcess/Databases/IndexedDB/WebIDBConnectionToServer.h:
  • WebProcess/InjectedBundle/InjectedBundleNavigationAction.h:
  • WebProcess/Plugins/Netscape/NetscapePlugin.h:
  • WebProcess/Plugins/Netscape/x11/NetscapePluginX11.cpp:

(WebKit::NetscapePluginX11::NetscapePluginX11):

5:22 AM Changeset in webkit [214103] by magomez@igalia.com
  • 7 edits
    2 adds in trunk

Follow-up (r213833): write a layout test for 169199
​https://bugs.webkit.org/show_bug.cgi?id=169576
<rdar://problem/31024766>

Reviewed by Said Abou-Hallawa.

Source/WebCore:

Add a function to Internals to force the deletion of the ImageDecoder just after
requesting the async decoding of a frame, and implement that behaviour in BitmapImage.

Test: fast/images/clear-animation-decoder.html

  • platform/graphics/BitmapImage.cpp:

(WebCore::BitmapImage::internalStartAnimation):

  • platform/graphics/BitmapImage.h:
  • testing/Internals.cpp:

(WebCore::Internals::setClearDecoderAfterAsyncFrameRequestForTesting):

  • testing/Internals.h:
  • testing/Internals.idl:

LayoutTests:

Add a test that ensures there's no crash when deleting the ImageDecoder after requesting
an async decoding operation.

  • fast/images/clear-animation-decoder-expected.txt: Added.
  • fast/images/clear-animation-decoder.html: Added.
2:46 AM Changeset in webkit [214102] by Carlos Garcia Campos
  • 2 edits in trunk/LayoutTests

Unreviewed GTK+ gardening. Mark mediastream and webrtc new failures.

  • platform/gtk/TestExpectations:
2:45 AM Changeset in webkit [214101] by tpopela@redhat.com
  • 3 edits in trunk/Source/WebKit2

Fix uninitialized public members in NetworkProcess
​https://bugs.webkit.org/show_bug.cgi?id=169598

Reviewed by Alex Christensen.

Found by Coverity scan.

  • NetworkProcess/NetworkProcessCreationParameters.h:
  • NetworkProcess/cache/NetworkCacheStorage.cpp:
1:22 AM Changeset in webkit [214100] by Carlos Garcia Campos
  • 4 edits in trunk/Source/WebCore

[Cairo] Handle the blend mode in GraphicsContext::drawPattern
​https://bugs.webkit.org/show_bug.cgi?id=169746

Reviewed by Žan Doberšek.

We are not taking into account the blend mode when passing the cairo operator to drawPatternToCairoContext().
This is based on patch by Žan Doberšek, just adding the toCairoOperator changes to make it easier to handle
it. Instead of checking everywhere if blend mode is Normal to decide whether to use toCairoOperator with
CompositeOperator or BlendMode, there's no a single toCairoOperator that receives both parameters, but BlendMode
is optional and defaults to Normal.

  • platform/graphics/cairo/CairoUtilities.cpp:

(WebCore::toCairoCompositeOperator):
(WebCore::toCairoOperator):

  • platform/graphics/cairo/CairoUtilities.h:
  • platform/graphics/cairo/GraphicsContextCairo.cpp:

(WebCore::GraphicsContext::setPlatformCompositeOperation):
(WebCore::GraphicsContext::drawPattern):

Note: See TracTimeline for information about the timeline view.