WebKit-https.git
3 years ago[JSC][DFG] Make addShouldSpeculateAnyInt more conservative to avoid regression caused...
utatane.tea@gmail.com [Thu, 23 Mar 2017 22:53:54 +0000 (22:53 +0000)]
[JSC][DFG] Make addShouldSpeculateAnyInt more conservative to avoid regression caused by Double <-> Int52 conversions
https://bugs.webkit.org/show_bug.cgi?id=169998

Reviewed by Saam Barati.

JSTests:

* microbenchmarks/int52-back-and-forth.js: Added.
(shouldBe):
(num):

Source/JavaScriptCore:

Double <-> Int52 and JSValue <-> Int52 conversions are not so cheap. Thus, Int52Rep is super carefully emitted.
We make addShouldSpeculateAnyInt more conservative to avoid regressions caused by the above conversions.
We select ArithAdd(Int52, Int52) only when this calculation is beneficial compared to added Int52Rep conversions.

This patch tighten the conditions of addShouldSpeculateAnyInt.

1. Honor DoubleConstant.

When executing imaging-darkroom, we have a thing like that,

    132:< 2:loc36> DoubleConstant(Double|UseAsOther, AnyIntAsDouble, Double: 4607182418800017408, 1.000000, bc#114)
    1320:< 1:loc38>        Int52Rep(Check:Int32:@82, Int52|PureInt, Int32, Exits, bc#114)
    1321:< 1:loc39>        Int52Constant(Int52|PureInt, Boolint32Nonboolint32Int52, Double: 4607182418800017408, 1.000000, bc#114)
    133:<!3:loc39> ArithSub(Int52Rep:@1320<Int52>, Int52Rep:@1321<Int52>, Int52|MustGen, Int52, CheckOverflow, Exits, bc#114)

The LHS of ArithSub says predicting Boolint32, and the rhs says AnyIntAsDouble. Thus we select ArithSub(Int52, Int52) instead
of ArithSub(Double, Double). However, it soon causes OSR exits. In imaging-darkroom, LHS's Int32 prediction will be broken.
While speculating Int32 in the above situation is reasonable approach since the given LHS says predicting Int32, this causes
severe performance regression.

Previously, we always select ArithSub(Double, Double). So accidentally, we do not encounter this misprediction issue.

One thing can be found that we have DoubleConstant in the RHS. It means that we have `1.0` instead of `1` in the code.
We can see the code like `lhs - 1.0` instead of `lhs - 1` in imaging-darkroom. It offers good information that lhs and
the resulting value would be double. Handling the above ArithSub in double seems more appropriate rather than handling
it in Int52.

So, in this patch, we honor DoubleConstant. If we find DoubleConstant on one operand, we give up selecting
Arith[Sub,Add](Int52, Int52). This change removes OSR exits occurr in imaging-darkroom right now.

2. Two Int52Rep(Double) conversions are not desirable.

We allow AnyInt ArithAdd only when the one operand of the binary operation should be speculated AnyInt. It is a bit conservative
decision. This is because Double to Int52 conversion is not so cheap. Frequent back-and-forth conversions between Double and Int52
rather hurt the performance. If the one operand of the operation is already Int52, the cost for constructing ArithAdd becomes
cheap since only one Double to Int52 conversion could be required.
This recovers some regression in assorted tests while keeping kraken crypto improvements.

3. Avoid frequent Int52 to JSValue conversions.

Int52 to JSValue conversion is not so cheap. Thus, we would like to avoid such situations. So, in this patch, we allow
Arith(Int52, Int52) with AnyIntAsDouble operand only when the node is used as number. By doing so, we avoid the case like,
converting Int52, performing ArithAdd, and soon converting back to JSValue.

The above 3 changes recover the regression measured in microbenchmarks/int52-back-and-forth.js and assorted benchmarks.
And still it keeps kraken crypto improvements.

                                           baseline                  patched

imaging-darkroom                       201.112+-3.192      ^     189.532+-2.883         ^ definitely 1.0611x faster
stanford-crypto-pbkdf2                 103.953+-2.325            100.926+-2.396           might be 1.0300x faster
stanford-crypto-sha256-iterative        35.103+-1.071      ?      36.049+-1.143         ? might be 1.0270x slower

* dfg/DFGGraph.h:
(JSC::DFG::Graph::addShouldSpeculateAnyInt):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214323 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoSVG animations are not paused when inserted into a hidden page
cdumez@apple.com [Thu, 23 Mar 2017 22:40:50 +0000 (22:40 +0000)]
SVG animations are not paused when inserted into a hidden page
https://bugs.webkit.org/show_bug.cgi?id=170026
<rdar://problem/31228704>

Reviewed by Andreas Kling.

Source/WebCore:

SVG animations were not paused when inserted into a hidden page. We would pause
animations in a page when the page becomes hidden. However, new animations
inserted in the page after this point would start, despite the page being
hidden.

Tests:
- svg/animations/animations-paused-when-inserted-in-hidden-document.html
- svg/animations/animations-paused-when-inserted-in-hidden-document2.html

* dom/Document.cpp:
(WebCore::Document::accessSVGExtensions):
* svg/SVGDocumentExtensions.cpp:
(WebCore::SVGDocumentExtensions::SVGDocumentExtensions):
(WebCore::SVGDocumentExtensions::addTimeContainer):
(WebCore::reportMessage):
* svg/SVGDocumentExtensions.h:
* testing/Internals.cpp:
(WebCore::Internals::areSVGAnimationsPaused):
* testing/Internals.h:
* testing/Internals.idl:

LayoutTests:

Add layout test coverage.

* svg/animations/animations-paused-when-inserted-in-hidden-document-expected.txt: Added.
* svg/animations/animations-paused-when-inserted-in-hidden-document.html: Added.
* svg/animations/animations-paused-when-inserted-in-hidden-document2-expected.txt: Added.
* svg/animations/animations-paused-when-inserted-in-hidden-document2.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214322 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago== Rolled over to ChangeLog-2017-03-23 ==
mcatanzaro@igalia.com [Thu, 23 Mar 2017 22:39:57 +0000 (22:39 +0000)]
== Rolled over to ChangeLog-2017-03-23 ==

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214321 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoLet the UI process do memory kills and replace crashes with diagnostic logging.
akling@apple.com [Thu, 23 Mar 2017 22:21:31 +0000 (22:21 +0000)]
Let the UI process do memory kills and replace crashes with diagnostic logging.
<https://webkit.org/b/170016>
<rdar://problem/31091292>

Reviewed by Chris Dumez.

Source/WebCore:

Remove the memory kill code from WebCore that just crashes in favor of controlled
teardown from UI process side in WebKit2.

* page/DiagnosticLoggingKeys.cpp:
(WebCore::DiagnosticLoggingKeys::simulatedProcessCrashKey):
(WebCore::DiagnosticLoggingKeys::exceededActiveMemoryLimitKey):
(WebCore::DiagnosticLoggingKeys::exceededInactiveMemoryLimitKey):
(WebCore::DiagnosticLoggingKeys::exceededBackgroundCPULimitKey):
* page/DiagnosticLoggingKeys.h:
* page/MemoryRelease.cpp:
(WebCore::didExceedMemoryLimitAndFailedToRecover): Deleted.
* page/MemoryRelease.h:

Source/WebKit2:

Have the WebProcess memory kill callback send a notification to the UI process
instead of outright crashing. The UI process then sends a diagnostic logging message
before carefully taking down the web process.

Also add drive-by diagnostic logging for background CPU limit kills.

* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::simulateProcessCrash):
(WebKit::WebProcessProxy::didExceedActiveMemoryLimit):
(WebKit::WebProcessProxy::didExceedInactiveMemoryLimit):
(WebKit::WebProcessProxy::didExceedBackgroundCPULimit):
* UIProcess/WebProcessProxy.h:
* UIProcess/WebProcessProxy.messages.in:
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::initializeWebProcess):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214320 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[JSC] MachineThreads does not consider situation that one thread has multiple VMs
utatane.tea@gmail.com [Thu, 23 Mar 2017 21:54:58 +0000 (21:54 +0000)]
[JSC] MachineThreads does not consider situation that one thread has multiple VMs
https://bugs.webkit.org/show_bug.cgi?id=169819

Reviewed by Mark Lam.

The Linux port of PlatformThread suspend/resume mechanism relies on having a thread
specific singleton thread data, and was relying on MachineThreads::Thread to be this
thread specific singleton. But because MachineThreads::Thread is not a thread specific
singleton, we can get a deadlock in the GTK port's DatabaseProcess.

This patch fixes this issue by moving per thread data from MachineThreads::Thread to
MachineThreads::ThreadData, where there will only be one instance of
MachineThreads::ThreadData per thread. Each MachineThreads::Thread will now point to
the same MachineThreads::ThreadData for any given thread.

* heap/MachineStackMarker.cpp:
(pthreadSignalHandlerSuspendResume):
(JSC::threadData):
(JSC::MachineThreads::Thread::Thread):
(JSC::MachineThreads::Thread::createForCurrentThread):
(JSC::MachineThreads::Thread::operator==):
(JSC::MachineThreads::ThreadData::ThreadData):
(JSC::MachineThreads::ThreadData::~ThreadData):
(JSC::MachineThreads::ThreadData::suspend):
(JSC::MachineThreads::ThreadData::resume):
(JSC::MachineThreads::ThreadData::getRegisters):
(JSC::MachineThreads::ThreadData::Registers::stackPointer):
(JSC::MachineThreads::ThreadData::Registers::framePointer):
(JSC::MachineThreads::ThreadData::Registers::instructionPointer):
(JSC::MachineThreads::ThreadData::Registers::llintPC):
(JSC::MachineThreads::ThreadData::freeRegisters):
(JSC::MachineThreads::ThreadData::captureStack):
(JSC::MachineThreads::tryCopyOtherThreadStacks):
(JSC::MachineThreads::Thread::~Thread): Deleted.
(JSC::MachineThreads::Thread::suspend): Deleted.
(JSC::MachineThreads::Thread::resume): Deleted.
(JSC::MachineThreads::Thread::getRegisters): Deleted.
(JSC::MachineThreads::Thread::Registers::stackPointer): Deleted.
(JSC::MachineThreads::Thread::Registers::framePointer): Deleted.
(JSC::MachineThreads::Thread::Registers::instructionPointer): Deleted.
(JSC::MachineThreads::Thread::Registers::llintPC): Deleted.
(JSC::MachineThreads::Thread::freeRegisters): Deleted.
(JSC::MachineThreads::Thread::captureStack): Deleted.
* heap/MachineStackMarker.h:
(JSC::MachineThreads::Thread::operator!=):
(JSC::MachineThreads::Thread::suspend):
(JSC::MachineThreads::Thread::resume):
(JSC::MachineThreads::Thread::getRegisters):
(JSC::MachineThreads::Thread::freeRegisters):
(JSC::MachineThreads::Thread::captureStack):
(JSC::MachineThreads::Thread::platformThread):
(JSC::MachineThreads::Thread::stackBase):
(JSC::MachineThreads::Thread::stackEnd):
* runtime/SamplingProfiler.cpp:
(JSC::FrameWalker::isValidFramePointer):
* runtime/VMTraps.cpp:
(JSC::findActiveVMAndStackBounds):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214319 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agobuildbot: Fix mixing of tabs and spaces after r214172
clopez@igalia.com [Thu, 23 Mar 2017 21:45:08 +0000 (21:45 +0000)]
buildbot: Fix mixing of tabs and spaces after r214172
https://bugs.webkit.org/show_bug.cgi?id=169793

Unreviewed.

* BuildSlaveSupport/build.webkit.org-config/master.cfg:
(RunWebKitTests.evaluateCommand):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214318 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoClean RTCPeerConnection.idl after bug 169978
commit-queue@webkit.org [Thu, 23 Mar 2017 21:07:40 +0000 (21:07 +0000)]
Clean RTCPeerConnection.idl after bug 169978
https://bugs.webkit.org/show_bug.cgi?id=169989

Patch by Youenn Fablet <youenn@apple.com> on 2017-03-23
Reviewed by Eric Carlson.

No change of behavior.
Moving addTransceiver to overloaded method to union type to match spec.
Refactoring RTCRtpSender to take a String&&.

* Modules/mediastream/MediaEndpointPeerConnection.cpp:
(WebCore::MediaEndpointPeerConnection::setRemoteDescriptionTask):
* Modules/mediastream/RTCPeerConnection.cpp:
(WebCore::RTCPeerConnection::addTransceiver):
(WebCore::RTCPeerConnection::completeAddTransceiver):
* Modules/mediastream/RTCPeerConnection.h:
* Modules/mediastream/RTCPeerConnection.idl: Cosmetic changes except for addTransceiver.
* Modules/mediastream/RTCRtpSender.cpp:
(WebCore::RTCRtpSender::create):
(WebCore::RTCRtpSender::RTCRtpSender):
* Modules/mediastream/RTCRtpSender.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214317 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoAdd a test to make sure SVG animations are paused in detached iframes
cdumez@apple.com [Thu, 23 Mar 2017 20:57:06 +0000 (20:57 +0000)]
Add a test to make sure SVG animations are paused in detached iframes
https://bugs.webkit.org/show_bug.cgi?id=170020

Reviewed by Antti Koivisto.

* svg/animations/animations-paused-disconnected-iframe-expected.txt: Added.
* svg/animations/animations-paused-disconnected-iframe.html: Added.
* svg/animations/resources/frame-with-svg-animation.html: Added.
* svg/animations/resources/grandchild-frame-with-svg-animation.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214316 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoShare more code between ManifestGenerator and /api/triggerables
rniwa@webkit.org [Thu, 23 Mar 2017 20:56:36 +0000 (20:56 +0000)]
Share more code between ManifestGenerator and /api/triggerables
https://bugs.webkit.org/show_bug.cgi?id=169993

Reviewed by Chris Dumez.

Shared the code to fetch the list of triggerables from the database between ManifestGenerator
and /api/triggerables.

* public/api/triggerables.php:
(main):
* public/include/manifest-generator.php:
(ManifestGenerator::fetch_triggerables): Extracted as a static function. Also include the ID
in the triggerable data.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214315 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agocreate-test-group should allow a different set of repositories to be used in each...
rniwa@webkit.org [Thu, 23 Mar 2017 20:54:39 +0000 (20:54 +0000)]
create-test-group should allow a different set of repositories to be used in each configuration
https://bugs.webkit.org/show_bug.cgi?id=169992

Rubber-stamped by Antti Koivisto.

Added the support for new POST parameter, revisionSets, to /privileged-api/create-test-group.
This new parameter now specifies an array of repository id to revision dictionaries, and allows
different set of repositories' revisions to be specified in each dictionary.

We keep the old API for v2 UI and detect-changes.js compatibility for now.

* public/privileged-api/create-test-group.php:
(main):
(commit_sets_from_revision_sets): Added.
(ensure_commit_sets): Only fetch the top-level repository per r213788 and r213976.

* public/v3/models/test-group.js:
(TestGroup.createAndRefetchTestGroups): Use the newly added revisionSets parameter instead of
the now depreacted commitSets parameter.

* public/v3/pages/analysis-task-page.js:
(AnalysisTaskPage.prototype._createTestGroupAfterVerifyingCommitSetList): Simplified this code
by simply verifying the consistency of commit sets now that createAndRefetchTestGroups takes
an array of commit sets instead of a dictionary of repository name to a list of revisions.

* server-tests/privileged-api-create-test-group-tests.js: Added test cases for new parameter.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214314 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoClients of JSArray::tryCreateForInitializationPrivate() should do their own null...
mark.lam@apple.com [Thu, 23 Mar 2017 20:31:18 +0000 (20:31 +0000)]
Clients of JSArray::tryCreateForInitializationPrivate() should do their own null checks.
https://bugs.webkit.org/show_bug.cgi?id=169783

Reviewed by Saam Barati.

JSTests:

* stress/regress-169783.js: Added.

Source/JavaScriptCore:

Fixed clients of tryCreateForInitializationPrivate() to do a null check and throw
an OutOfMemoryError if allocation fails, or RELEASE_ASSERT that the allocation
succeeds.

* dfg/DFGOperations.cpp:
* ftl/FTLOperations.cpp:
(JSC::FTL::operationMaterializeObjectInOSR):
* runtime/ArrayPrototype.cpp:
(JSC::arrayProtoFuncSplice):
* runtime/CommonSlowPaths.cpp:
(JSC::SLOW_PATH_DECL):
* runtime/JSArray.cpp:
(JSC::JSArray::tryCreateForInitializationPrivate):
(JSC::JSArray::fastSlice):
* runtime/JSArray.h:
(JSC::constructArray):
(JSC::constructArrayNegativeIndexed):
* runtime/RegExpMatchesArray.cpp:
(JSC::createEmptyRegExpMatchesArray):
* runtime/RegExpMatchesArray.h:
(JSC::createRegExpMatchesArray):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214313 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoAdd logging in case libwebrtc.dylib cannot be opened
commit-queue@webkit.org [Thu, 23 Mar 2017 20:20:10 +0000 (20:20 +0000)]
Add logging in case libwebrtc.dylib cannot be opened
https://bugs.webkit.org/show_bug.cgi?id=170017

Patch by Youenn Fablet <youenn@apple.com> on 2017-03-23
Reviewed by Geoffrey Garen.

No change of behavior.

* platform/mediastream/libwebrtc/LibWebRTCProvider.cpp:
(WebCore::LibWebRTCProvider::webRTCAvailable):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214312 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoREGRESSION (r214047): LayoutTest fast/events/pageshow-pagehide-on-back-cached-with...
dbates@webkit.org [Thu, 23 Mar 2017 19:56:46 +0000 (19:56 +0000)]
REGRESSION (r214047): LayoutTest fast/events/pageshow-pagehide-on-back-cached-with-frames.html is a flaky failure
https://bugs.webkit.org/show_bug.cgi?id=169834

Reviewed by Alexey Proskuryakov.

Ensure that the subframes are loaded in a consistent order between test runs.

Currently the test fast/events/pageshow-pagehide-on-back-cached-with-frames.html assumes
that subframes will load in tree order. This assumption is incorrect. Therefore, we need
to explicitly manage the loading of the subframes to ensure this invariant.

* fast/events/pageshow-pagehide-on-back-cached-with-frames-expected.txt:
* fast/events/pageshow-pagehide-on-back-cached-with-frames.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214311 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUnreviewed, a speculative build fix
jiewen_tan@apple.com [Thu, 23 Mar 2017 19:53:02 +0000 (19:53 +0000)]
Unreviewed, a speculative build fix
<rdar://problem/31225378>

* crypto/CommonCryptoUtilities.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214310 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoWebSQL databases should not openable in private browsing.
beidson@apple.com [Thu, 23 Mar 2017 19:06:31 +0000 (19:06 +0000)]
WebSQL databases should not openable in private browsing.
<rdar://problem/30383335> and https://bugs.webkit.org/show_bug.cgi?id=170013

Reviewed by Alex Christensen.

Source/WebCore:

Test: storage/websql/private-browsing-open-disabled.html

* Modules/webdatabase/DatabaseManager.cpp:
(WebCore::DatabaseManager::openDatabaseBackend):
(WebCore::DatabaseManager::tryToOpenDatabaseBackend): Throw an exception if in private browsing.
* Modules/webdatabase/DatabaseManager.h:

LayoutTests:

* storage/websql/private-browsing-open-disabled-expected.txt: Added.
* storage/websql/private-browsing-open-disabled.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214309 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoWeb Inspector: Editing local storage keys/values starts with truncated strings
nvasilyev@apple.com [Thu, 23 Mar 2017 18:47:49 +0000 (18:47 +0000)]
Web Inspector: Editing local storage keys/values starts with truncated strings
https://bugs.webkit.org/show_bug.cgi?id=169896

Reviewed by Brian Burg.

* UserInterface/Views/DataGrid.css:
(.data-grid td.editing > div):
* UserInterface/Views/DataGrid.js:
(WebInspector.DataGrid.prototype._editingCancelled):
div.cell-content is removed because the <td> itself is contenteditable (using -webkit-user-modify:read-write-plaintext-only),
not the inner div.cell-content. Re-create div.cell-content after commiting (by pressing Enter) or canceling (by pressing Esc)
since it may get removed while editing.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214308 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoRemove pass expectation for fast/forms/range/range-remove-on-drag.html as it relies...
ryanhaddad@apple.com [Thu, 23 Mar 2017 18:45:50 +0000 (18:45 +0000)]
Remove pass expectation for fast/forms/range/range-remove-on-drag.html as it relies on touch events.

Unreviewed test gardening.

* platform/ios-simulator-wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214307 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[GTK] Mark compositing and css3/blending tests asserting on debug builds.
clopez@igalia.com [Thu, 23 Mar 2017 18:22:10 +0000 (18:22 +0000)]
[GTK] Mark compositing and css3/blending tests asserting on debug builds.

Unreviewed test gardening.

* platform/gtk/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214306 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoNeverDestroyed<MediaQueryEvaluator> must explicitly construct with a String
dino@apple.com [Thu, 23 Mar 2017 18:15:58 +0000 (18:15 +0000)]
NeverDestroyed<MediaQueryEvaluator> must explicitly construct with a String
https://bugs.webkit.org/show_bug.cgi?id=169987
<rdar://problem/31211087>

Reviewed by Alex Christensen.

CSSDefaultStyleSheets creates a static MediaQueryEvaluator, but thanks
to the template magic of NeverDestroyed, it was converting the char*
argument into a bool, and calling the wrong constructor.

Unfortunately this is difficult to test because it only affects
the default UA style sheets, and they currently don't have
and @media rules (which would always evaluate to true given
the bug). I don't want to put in a useless rule just to check
if the bug is fixed. When one is added for bug 168447, this change
will be exercised.

* css/CSSDefaultStyleSheets.cpp: Explicitly construct with a String
rather than a char*.
(WebCore::screenEval):
(WebCore::printEval):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214305 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoVersioning.
jmarcell@apple.com [Thu, 23 Mar 2017 16:13:18 +0000 (16:13 +0000)]
Versioning.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214304 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[MediaStream] Make mock video source work on iOS
eric.carlson@apple.com [Thu, 23 Mar 2017 15:54:48 +0000 (15:54 +0000)]
[MediaStream] Make mock video source work on iOS
https://bugs.webkit.org/show_bug.cgi?id=169923

Reviewed by Youenn Fablet.

Generate IOSurface-backed CVPixelBuffers. Allocate them from a
CVPixelBufferPool for efficiency.

No new tests, no functional change.

* platform/cocoa/CoreVideoSoftLink.cpp:
* platform/cocoa/CoreVideoSoftLink.h:
* platform/graphics/avfoundation/objc/MediaPlayerPrivateMediaStreamAVFObjC.mm:
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::videoTransformationMatrix):
* platform/mediastream/mac/MockRealtimeVideoSourceMac.h:
* platform/mediastream/mac/MockRealtimeVideoSourceMac.mm:
(WebCore::MockRealtimeVideoSourceMac::pixelBufferFromCGImage):
(WebCore::MockRealtimeVideoSourceMac::applySize):
* platform/mock/MockRealtimeVideoSource.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214302 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoTest DataDetectors supported schemes in lowercase
timothy_horton@apple.com [Thu, 23 Mar 2017 11:33:04 +0000 (11:33 +0000)]
Test DataDetectors supported schemes in lowercase
https://bugs.webkit.org/show_bug.cgi?id=169994
<rdar://problem/30788920>

Reviewed by Wenson Hsieh.

* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _actionForLongPress]):
(-[WKContentView _interactionShouldBeginFromPreviewItemController:forPosition:]):
Lower-case the scheme before looking in tapAndHoldSchemes, which is
guaranteed to be in lower-case. WebCore already does this in places
that it confers with tapAndHoldSchemes.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214301 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[jsc] Add MacroAssemblerMIPS::storeFence()
commit-queue@webkit.org [Thu, 23 Mar 2017 09:25:06 +0000 (09:25 +0000)]
[jsc] Add MacroAssemblerMIPS::storeFence()
https://bugs.webkit.org/show_bug.cgi?id=169705

Patch by Guillaume Emont <guijemont@igalia.com> on 2017-03-23
Reviewed by Yusuke Suzuki.

There doesn't seem to be anything more fine grained than "sync" that
guarantees that all memory operations following it are going to happen
after all stores before it, so we just use sync.

* assembler/MIPSAssembler.h:
(JSC::MIPSAssembler::sync): Added a FIXME about SYNC_MB.
* assembler/MacroAssemblerMIPS.h:
(JSC::MacroAssemblerMIPS::storeFence): Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214300 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoFix uninitialized public members in WTF
tpopela@redhat.com [Thu, 23 Mar 2017 09:10:36 +0000 (09:10 +0000)]
Fix uninitialized public members in WTF
https://bugs.webkit.org/show_bug.cgi?id=169676

Reviewed by Michael Catanzaro.

Found by Coverity scan.

* wtf/Bag.h:
* wtf/HashTable.h:
* wtf/HashTraits.h:
* wtf/Insertion.h:
* wtf/OrderMaker.h:
* wtf/ParallelJobsGeneric.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214299 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUnreviewed GTK+ gardening. Mark webrtc/connection-state.html as failing.
carlosgc@webkit.org [Thu, 23 Mar 2017 07:38:53 +0000 (07:38 +0000)]
Unreviewed GTK+ gardening. Mark webrtc/connection-state.html as failing.

* platform/gtk/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214298 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[Cocoa] Add an option to exclude overflow when snapshotting a WKWebProcessPlugInNodeH...
aestes@apple.com [Thu, 23 Mar 2017 06:21:25 +0000 (06:21 +0000)]
[Cocoa] Add an option to exclude overflow when snapshotting a WKWebProcessPlugInNodeHandle
https://bugs.webkit.org/show_bug.cgi?id=169991
<rdar://problem/30141083>

Reviewed by Tim Horton.

Source/WebKit2:

When snapshotting an element using -[WKWebProcessPlugInNodeHandle renderedImageWithOptions:],
the image is sized to encompass all descendants, even if the element is styled with
overflow:hidden. This patch adds a WKSnapshotOption to exclude overflow, resulting in an
image that's always sized to the node's bounding box.

* Shared/API/c/WKImage.h: Defined kWKSnapshotOptionsExcludeOverflow.
* WebProcess/InjectedBundle/API/Cocoa/WKWebProcessPlugInNodeHandle.mm:
(-[WKWebProcessPlugInNodeHandle renderedImageWithOptions:]): Converted the
WKSnapshotOptions argument to a SnapshotOptions using toSnapshotOptions(), and called
InjectedBundleNodeHandle::renderedImage() with shouldExcludeOverflow set to true if the
WKSnapshotOptions included kWKSnapshotOptionsExcludeOverflow.
* WebProcess/InjectedBundle/API/c/WKBundleNodeHandle.cpp:
(WKBundleNodeHandleCopySnapshotWithOptions): Ditto.
* WebProcess/InjectedBundle/DOM/InjectedBundleNodeHandle.cpp:
(WebKit::InjectedBundleNodeHandle::renderedImage): If shouldExcludeOverflow is true, use
RenderObject::absoluteBoundingBoxRectIgnoringTransforms() as the paintingRect for
imageForRect().
* WebProcess/InjectedBundle/DOM/InjectedBundleNodeHandle.h:

Tools:

* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/Tests/WebKit2Cocoa/RenderedImageWithOptions.mm: Added.
(-[RenderedImageWithOptionsObject didRenderImageWithSize:]):
(TEST):
* TestWebKitAPI/Tests/WebKit2Cocoa/RenderedImageWithOptionsPlugIn.mm: Added.
(-[RenderedImageWithOptionsPlugIn webProcessPlugIn:didCreateBrowserContextController:]):
(-[RenderedImageWithOptionsPlugIn webProcessPlugInBrowserContextController:didFinishLoadForFrame:]):
* TestWebKitAPI/Tests/WebKit2Cocoa/RenderedImageWithOptionsProtocol.h: Added.
* TestWebKitAPI/Tests/WebKit2Cocoa/rendered-image-excluding-overflow.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214297 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[JSC][DFG] Propagate AnyIntAsDouble information carefully to utilize it in fixup
utatane.tea@gmail.com [Thu, 23 Mar 2017 06:01:39 +0000 (06:01 +0000)]
[JSC][DFG] Propagate AnyIntAsDouble information carefully to utilize it in fixup
https://bugs.webkit.org/show_bug.cgi?id=169914

Reviewed by Saam Barati.

JSTests:

* stress/any-int-as-double-add.js: Added.
(shouldBe):
(test):
* stress/to-this-numbers.js: Added.
(shouldBe):
(Number.prototype.toThis):

Source/JavaScriptCore:

In DFG prediction propagation phase, we pollute the prediction of GetByVal for Array::Double
as SpecDoubleReal even if the heap prediction says the proper prediction is SpecAnyIntAsDouble.
Thus, the following nodes just see the result of GetByVal(Array::Double) as double value,
and select suboptimal edge filters in fixup phase. For example, if the result of GetByVal is
SpecAnyIntAsDouble, we can see the node like ArithAdd(SpecAnyIntAsDouble, Int52) and we should
have a chance to make it ArithAdd(Check:Int52, Int52) instead of ArithAdd(Double, Double).

This patch propagates SpecAnyIntAsDouble in GetByVal(Array::Double) properly. And ValueAdd,
ArithAdd and ArithSub select AnyInt edge filters for SpecAnyIntAsDouble values. It finally
produces a Int52 specialized DFG node. And subsequent nodes using the produced one also
become Int52 specialized.

One considerable problem is that the heap prediction misses the non any int doubles. In that case,
if Int52 edge filter is used, BadType exit will occur. It updates the prediction of the value profile
of GetByVal. So, in the next time, GetByVal(Array::Double) produces more conservative predictions
and avoids exit-and-recompile loop correctly.

This change is very sensitive to the correct AI and appropriate predictions. Thus, this patch finds
and fixes some related issues. One is incorrect prediction of ToThis and another is incorrect
AI logic for Int52Rep.

This change dramatically improves kraken benchmarks' crypto-pbkdf2 and crypto-sha256-iterative
by 42.0% and 30.7%, respectively.

                                             baseline                  patched
Kraken:
ai-astar                                  158.851+-4.132      ?     159.433+-5.176         ?
audio-beat-detection                       53.193+-1.621      ?      53.391+-2.072         ?
audio-dft                                 103.589+-2.277      ?     104.902+-1.924         ? might be 1.0127x slower
audio-fft                                  40.491+-1.102             39.854+-0.755           might be 1.0160x faster
audio-oscillator                           68.504+-1.721      ?      68.957+-1.725         ?
imaging-darkroom                          118.367+-2.171      ?     119.581+-2.310         ? might be 1.0103x slower
imaging-desaturate                         71.443+-1.461      ?      72.398+-1.918         ? might be 1.0134x slower
imaging-gaussian-blur                     110.648+-4.035            109.184+-3.373           might be 1.0134x faster
json-parse-financial                       60.363+-1.628      ?      61.936+-1.585         ? might be 1.0261x slower
json-stringify-tinderbox                   37.903+-0.869      ?      39.559+-1.607         ? might be 1.0437x slower
stanford-crypto-aes                        56.313+-1.512      ?      56.675+-1.715         ?
stanford-crypto-ccm                        51.564+-1.900      ?      53.456+-2.548         ? might be 1.0367x slower
stanford-crypto-pbkdf2                    129.546+-2.738      ^      91.214+-2.027         ^ definitely 1.4202x faster
stanford-crypto-sha256-iterative           43.515+-0.730      ^      33.292+-0.653         ^ definitely 1.3071x faster

<arithmetic>                               78.878+-0.528      ^      75.988+-0.621         ^ definitely 1.0380x faster

* dfg/DFGAbstractInterpreterInlines.h:
(JSC::DFG::AbstractInterpreter<AbstractStateType>::executeEffects):
* dfg/DFGGraph.h:
(JSC::DFG::Graph::addShouldSpeculateAnyInt):
* dfg/DFGPredictionPropagationPhase.cpp:
* ftl/FTLLowerDFGToB3.cpp:
(JSC::FTL::DFG::LowerDFGToB3::compileArithNegate):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214296 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoWKScrollView should not require data interaction gestures to fail before panning
wenson_hsieh@apple.com [Thu, 23 Mar 2017 04:51:47 +0000 (04:51 +0000)]
WKScrollView should not require data interaction gestures to fail before panning
https://bugs.webkit.org/show_bug.cgi?id=169986

Reviewed by Tim Horton.

This workaround is no longer needed.

* UIProcess/ios/WKScrollView.mm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214295 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoLet MemoryPressureHandler track whether the process is active or inactive.
akling@apple.com [Thu, 23 Mar 2017 04:14:03 +0000 (04:14 +0000)]
Let MemoryPressureHandler track whether the process is active or inactive.
<https://webkit.org/b/169990>

Reviewed by Antti Koivisto.

Source/WebCore:

Have PerformanceMonitor inform MemoryPressureHandler about relevant
changes to the page activity states.

Also change the "1 hour after last page load" policy to "1 hour after
last being visible in an active window."

* page/MainFrame.cpp:
(WebCore::MainFrame::didCompleteLoad):
* page/MainFrame.h:
* page/MemoryRelease.cpp:
(WebCore::processIsEligibleForMemoryKill): Deleted.
* page/MemoryRelease.h:
* page/Page.h:
(WebCore::Page::performanceMonitor):
* page/PerformanceMonitor.cpp:
(WebCore::PerformanceMonitor::PerformanceMonitor):
(WebCore::PerformanceMonitor::activityStateChanged):
(WebCore::PerformanceMonitor::processMayBecomeInactiveTimerFired):
(WebCore::PerformanceMonitor::updateProcessStateForMemoryPressure):
* page/PerformanceMonitor.h:

Source/WebKit2:

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

Source/WTF:

An active process is a process that meets any of these criteria:

    - Has had a visible Page in an active window in the last hour
    - Is playing audio

This replaces the old mechanism where MemoryPressureHandler would invoke
a callback to know whether the process was eligible for a memory kill.
Instead we now plumb the relevant information down from WebCore and kill
based on this activity state.

* wtf/MemoryPressureHandler.cpp:
(WTF::MemoryPressureHandler::measurementTimerFired):
(WTF::MemoryPressureHandler::setProcessState):
* wtf/MemoryPressureHandler.h:
(WTF::MemoryPressureHandler::setMemoryKillCallback):
(WTF::MemoryPressureHandler::processState):
(WTF::MemoryPressureHandler::setProcessIsEligibleForMemoryKillCallback): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214294 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoSupport RTCPeerConnectionState
commit-queue@webkit.org [Thu, 23 Mar 2017 02:44:57 +0000 (02:44 +0000)]
Support RTCPeerConnectionState
https://bugs.webkit.org/show_bug.cgi?id=169978

Patch by Youenn Fablet <youenn@apple.com> on 2017-03-22
Reviewed by Jon Lee.

LayoutTests/imported/w3c:

* web-platform-tests/webrtc/rtcpeerconnection/rtcpeerconnection-constructor-expected.txt:

Source/WebCore:

Test: webrtc/connection-state.html

Implementing https://www.w3.org/TR/webrtc/#rtcpeerconnectionstate-enum.
Its state and event is based on changes made to ice gathering state and ice connection state.

* CMakeLists.txt: Adding RTCPeerConnectionState idl.
* DerivedSources.make: Ditto.
* Modules/mediastream/RTCPeerConnection.cpp: Splitting close in doClose/doStop so that we can send closed event
in case close is called, but not if stopped.
(WebCore::RTCPeerConnection::doClose):
(WebCore::RTCPeerConnection::close):
(WebCore::RTCPeerConnection::stop):
(WebCore::RTCPeerConnection::doStop):
(WebCore::RTCPeerConnection::updateIceGatheringState):
(WebCore::RTCPeerConnection::updateIceConnectionState):
(WebCore::RTCPeerConnection::updateConnectionState):
* Modules/mediastream/RTCPeerConnection.h:
* Modules/mediastream/RTCPeerConnection.idl: Fixing IDL and minor cosmetic changes
* Modules/mediastream/RTCPeerConnection.js:
(setLocalDescription): Cosmetic change.
(setRemoteDescription):
* Modules/mediastream/RTCPeerConnectionState.idl: Added.
* Modules/mediastream/libwebrtc/LibWebRTCMediaEndpoint.cpp:
(WebCore::LibWebRTCMediaEndpoint::OnIceGatheringChange): Adding 'gathering' state
* WebCore.xcodeproj/project.pbxproj:
* dom/EventNames.h:
* platform/mediastream/RTCPeerConnectionState.h: Added.

LayoutTests:

* webrtc/connection-state-expected.txt: Added.
* webrtc/connection-state.html: Added.
* webrtc/rtcpeerconnection-error-messages-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214293 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[GTK] Enable CSS filters related tests.
clopez@igalia.com [Thu, 23 Mar 2017 02:37:08 +0000 (02:37 +0000)]
[GTK] Enable CSS filters related tests.
https://bugs.webkit.org/show_bug.cgi?id=99026

Unreviewed test gardening.

* platform/gtk/TestExpectations:
* platform/gtk/css3/filters/composited-during-transition-layertree-expected.txt: Added. Rebaselined.
* platform/gtk/css3/filters/effect-blur-expected.png: Added. Rebaselined.
* platform/gtk/css3/filters/effect-blur-expected.txt: Added. Rebaselined.
* platform/gtk/css3/filters/effect-blur-hw-expected.png: Added. Rebaselined.
* platform/gtk/css3/filters/effect-blur-hw-expected.txt: Added. Rebaselined.
* platform/gtk/css3/filters/effect-combined-expected.png: Added. Rebaselined.
* platform/gtk/css3/filters/effect-combined-expected.txt: Added. Rebaselined.
* platform/gtk/css3/filters/effect-combined-hw-expected.png: Added. Rebaselined.
* platform/gtk/css3/filters/effect-combined-hw-expected.txt: Added. Rebaselined.
* platform/gtk/css3/filters/effect-contrast-expected.png: Added. Rebaselined.
* platform/gtk/css3/filters/effect-contrast-expected.txt: Added. Rebaselined.
* platform/gtk/css3/filters/effect-contrast-hw-expected.png: Added. Rebaselined.
* platform/gtk/css3/filters/effect-contrast-hw-expected.txt: Added. Rebaselined.
* platform/gtk/css3/filters/effect-grayscale-expected.png: Added. Rebaselined.
* platform/gtk/css3/filters/effect-grayscale-expected.txt: Added. Rebaselined.
* platform/gtk/css3/filters/effect-grayscale-hw-expected.png: Added. Rebaselined.
* platform/gtk/css3/filters/effect-grayscale-hw-expected.txt: Added. Rebaselined.
* platform/gtk/css3/filters/effect-hue-rotate-expected.png: Added. Rebaselined.
* platform/gtk/css3/filters/effect-hue-rotate-expected.txt: Added. Rebaselined.
* platform/gtk/css3/filters/effect-hue-rotate-hw-expected.png: Added. Rebaselined.
* platform/gtk/css3/filters/effect-hue-rotate-hw-expected.txt: Added. Rebaselined.
* platform/gtk/css3/filters/effect-invert-expected.png: Added. Rebaselined.
* platform/gtk/css3/filters/effect-invert-expected.txt: Added. Rebaselined.
* platform/gtk/css3/filters/effect-invert-hw-expected.png: Added. Rebaselined.
* platform/gtk/css3/filters/effect-invert-hw-expected.txt: Added. Rebaselined.
* platform/gtk/css3/filters/effect-opacity-expected.png: Added. Rebaselined.
* platform/gtk/css3/filters/effect-opacity-expected.txt: Added. Rebaselined.
* platform/gtk/css3/filters/effect-opacity-hw-expected.png: Added. Rebaselined.
* platform/gtk/css3/filters/effect-opacity-hw-expected.txt: Added. Rebaselined.
* platform/gtk/css3/filters/effect-reference-composite-expected.png: Added. Rebaselined.
* platform/gtk/css3/filters/effect-reference-composite-expected.txt: Added. Rebaselined.
* platform/gtk/css3/filters/effect-reference-composite-hw-expected.png: Added. Rebaselined.
* platform/gtk/css3/filters/effect-reference-composite-hw-expected.txt: Added. Rebaselined.
* platform/gtk/css3/filters/effect-reference-expected.png: Added. Rebaselined.
* platform/gtk/css3/filters/effect-reference-expected.txt: Added. Rebaselined.
* platform/gtk/css3/filters/effect-reference-hw-expected.png: Added. Rebaselined.
* platform/gtk/css3/filters/effect-reference-hw-expected.txt: Added. Rebaselined.
* platform/gtk/css3/filters/effect-reference-ordering-expected.png: Added. Rebaselined.
* platform/gtk/css3/filters/effect-reference-ordering-expected.txt: Added. Rebaselined.
* platform/gtk/css3/filters/effect-reference-ordering-hw-expected.png: Added. Rebaselined.
* platform/gtk/css3/filters/effect-reference-ordering-hw-expected.txt: Added. Rebaselined.
* platform/gtk/css3/filters/effect-saturate-expected.png: Added. Rebaselined.
* platform/gtk/css3/filters/effect-saturate-expected.txt: Added. Rebaselined.
* platform/gtk/css3/filters/effect-saturate-hw-expected.png: Added. Rebaselined.
* platform/gtk/css3/filters/effect-saturate-hw-expected.txt: Added. Rebaselined.
* platform/gtk/css3/filters/effect-sepia-expected.png: Added. Rebaselined.
* platform/gtk/css3/filters/effect-sepia-expected.txt: Added. Rebaselined.
* platform/gtk/css3/filters/effect-sepia-hw-expected.png: Added. Rebaselined.
* platform/gtk/css3/filters/effect-sepia-hw-expected.txt: Added. Rebaselined.
* platform/gtk/css3/filters/regions-expanding-expected.png: Added. Rebaselined.
* platform/gtk/css3/filters/regions-expanding-expected.txt: Added. Rebaselined.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214292 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoASSERT_WITH_SECURITY_IMPLICATION hit when removing an <input type="range"> while...
jiewen_tan@apple.com [Thu, 23 Mar 2017 02:21:58 +0000 (02:21 +0000)]
ASSERT_WITH_SECURITY_IMPLICATION hit when removing an <input type="range"> while dragging on iOS
https://bugs.webkit.org/show_bug.cgi?id=165535
<rdar://problem/29559749>

Reviewed by Ryosuke Niwa.

Source/WebCore:

Utimately we should prevent SliderThumbElement::unregisterForTouchEvents() being called while
updating render tree. A quick fix for this is to move dispatchFormControlChangeEvent for input
from stopDragging up to the callers which really needs to dispatch this event, i.e., finishing
dragging the slider. It is clear that not every caller of stopDragging wants to
dispatchFormControlChangeEvent.

Test: fast/forms/range/range-remove-on-drag.html

* html/shadow/SliderThumbElement.cpp:
(WebCore::SliderThumbElement::stopDragging):
(WebCore::SliderThumbElement::defaultEventHandler):
(WebCore::SliderThumbElement::handleTouchEndAndCancel):

LayoutTests:

This test case is only enabled in iOS simulator WK2.

* TestExpectations:
* fast/forms/range/range-remove-on-drag-expected.txt: Added.
* fast/forms/range/range-remove-on-drag.html: Added.
* platform/ios-simulator-wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214291 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoDynamically applied :empty pseudo class with display:none does not get unapplied
antti@apple.com [Thu, 23 Mar 2017 01:16:07 +0000 (01:16 +0000)]
Dynamically applied :empty pseudo class with display:none does not get unapplied
https://bugs.webkit.org/show_bug.cgi?id=169907

Reviewed by Ryosuke Niwa.

Source/WebCore:

We improperly reset the styleAffectedByEmpty bit when removing the renderer when :empty starts
applying. We then fail to invalidate the style when the element becomes non-empty again.

Fix by resetting the style relation bits only when computing the style.

Test: fast/css/empty-display-none-invalidation.html

* dom/Element.cpp:
(WebCore::Element::resetStyleRelations):

    Expose this separately.

(WebCore::Element::clearStyleDerivedDataBeforeDetachingRenderer):

    Don't reset style relation bits when removing renderers.

* dom/Element.h:
* dom/ElementRareData.h:
(WebCore::ElementRareData::resetComputedStyle):
(WebCore::ElementRareData::resetStyleRelations):

    Reset all these bits in one function.

(WebCore::ElementRareData::resetDynamicRestyleObservations): Deleted.
* style/StyleTreeResolver.cpp:
(WebCore::Style::resetStyleForNonRenderedDescendants):
(WebCore::Style::TreeResolver::resolveComposedTree):

    Call the explicit style relation reset function when recomputing style.

LayoutTests:

* fast/css/empty-display-none-invalidation-expected.html: Added.
* fast/css/empty-display-none-invalidation.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214290 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoAdd support for Error.stackTraceLimit.
mark.lam@apple.com [Thu, 23 Mar 2017 01:14:06 +0000 (01:14 +0000)]
Add support for Error.stackTraceLimit.
https://bugs.webkit.org/show_bug.cgi?id=169904

Reviewed by Saam Barati.

JSTests:

* stress/error-stack-trace-limit.js: Added.

Source/JavaScriptCore:

Since there's no standard for this yet, we'll implement Error.stackTraceLimit
based on how Chrome does it.  This includes some idiosyncrasies like:
1. If we set Error.stackTraceLimit = 0, then new Error().stack yields an empty
   stack trace (Chrome has a title with no stack frame entries).
2. If we set Error.stackTraceLimit = {] (i.e. to a non-number value), then
   new Error().stack is undefined.

Chrome and IE defaults their Error.stackTraceLimit to 10.  We'll default ours to
100 because 10 may be a bit too skimpy and it is not that costly to allow up to
100 frames instead of 10.

The default value for Error.stackTraceLimit is specified by
Options::defaultErrorStackTraceLimit().

Also, the Exception object now limits the number of stack trace frames it captures
to the limit specified by Options::exceptionStackTraceLimit().

Note: the Exception object captures a stack trace that is not necessarily the
same as the one in an Error object being thrown:

- The Error object captures the stack trace at the point of object creation.

- The Exception object captures the stack trace at the point that the exception
  is thrown.  This stack trace is captured even when throwing a value that is not
  an Error object e.g. a primitive value.  The Exception object stack trace is
  only used by WebInspector to identify where a value is thrown from.  Hence,
  it does not necessary make sense the Exception object stack trace limited by
  Error.stackTraceLimit.  Instead, we have it use own Options::exceptionStackTraceLimit().

* interpreter/Interpreter.cpp:
(JSC::Interpreter::unwind):
* jsc.cpp:
(dumpException):
* runtime/CommonIdentifiers.h:
* runtime/Error.cpp:
(JSC::addErrorInfoAndGetBytecodeOffset):
* runtime/ErrorConstructor.cpp:
(JSC::ErrorConstructor::finishCreation):
(JSC::ErrorConstructor::put):
(JSC::ErrorConstructor::deleteProperty):
* runtime/ErrorConstructor.h:
(JSC::ErrorConstructor::stackTraceLimit):
* runtime/Exception.cpp:
(JSC::Exception::finishCreation):
* runtime/Options.h:

LayoutTests:

Rebased test.

* js/Object-getOwnPropertyNames-expected.txt:
* js/script-tests/Object-getOwnPropertyNames.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214289 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoFix warnings in libwebrtc
achristensen@apple.com [Thu, 23 Mar 2017 00:57:27 +0000 (00:57 +0000)]
Fix warnings in libwebrtc
https://bugs.webkit.org/show_bug.cgi?id=169973

Reviewed by Geoffrey Garen.

* Configurations/boringssl.xcconfig:
* Configurations/libsrtp.xcconfig:
* Configurations/libwebrtc.xcconfig:
* Configurations/libwebrtcpcrtc.xcconfig:
* Configurations/libyuv.xcconfig:
* Configurations/opus.xcconfig:
* Configurations/usrsctp.xcconfig:
Build with more warnings.
opus still needs some incompatible pointer warnings disabled because it converts
const opus_int16 * to const opus_val16 * and opus_int32 * to opus_val32 *
and that's ok because its a codec and that's what codecs do.
* Source/webrtc/base/logging.cc:
* Source/webrtc/base/logging.h:
* Source/webrtc/base/neverdestroyed.h: Added.
(webrtc::NeverDestroyed::NeverDestroyed):
(webrtc::NeverDestroyed::operator T&):
(webrtc::NeverDestroyed::get):
(webrtc::NeverDestroyed::operator&):
(webrtc::NeverDestroyed::asPtr):
Added webrtc::NeverDestroyed which may or may not be based on WTF::NeverDestroyed.
This allows us to avoid exit time destructors, which would slow down program termination for no reason.
* Source/webrtc/base/virtualsocketserver.cc:
* Source/webrtc/modules/audio_coding/codecs/builtin_audio_decoder_factory_internal.cc:
Adopt NeverDestroyed in function scope so we don't have global constructors or destructors.
* Source/webrtc/modules/audio_processing/beamformer/array_util.h:
(webrtc::DegreesToRadians):
(webrtc::RadiansToDegrees):
Add constexpr so we can calculate values at compile time instead of launch time.
* Source/webrtc/modules/rtp_rtcp/source/rtp_utility.cc:
* Source/webrtc/system_wrappers/source/clock.cc:
* libwebrtc.xcodeproj/project.pbxproj:
Don't compile ssl_test.cc.  We don't need it.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214288 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago/api/uploaded-file should return createdAt as a POSIX timestamp
rniwa@webkit.org [Thu, 23 Mar 2017 00:27:01 +0000 (00:27 +0000)]
/api/uploaded-file should return createdAt as a POSIX timestamp
https://bugs.webkit.org/show_bug.cgi?id=169980

Rubber-stamped by Antti Koivisto.

Call Database::to_js_time on createdAt to return it as a POSIX timestamp.

* public/include/uploaded-file-helpers.php:
(format_uploaded_file): Fixed the bug.
* server-tests/api-manifest-tests.js: Renamed from api-manifest.js.
* server-tests/api-uploaded-file-tests.js: Renamed from api-uploaded-file.js. Added a test case.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214287 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[GTK] Update layout test expectations file.
clopez@igalia.com [Thu, 23 Mar 2017 00:16:37 +0000 (00:16 +0000)]
[GTK] Update layout test expectations file.

Unreviewed test gardening.

* platform/gtk/TestExpectations: Update the list of expected failures.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214286 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUploadedFile should support a callback for upload progress
rniwa@webkit.org [Wed, 22 Mar 2017 23:43:32 +0000 (23:43 +0000)]
UploadedFile should support a callback for upload progress
https://bugs.webkit.org/show_bug.cgi?id=169977

Reviewed by Andreas Kling.

Added a new option dictionary to CommonRemoteAPI.sendHttpRequest with uploadProgressCallback

Moved request headers and responseHandler callback in NodeRemoteAPI to this dictionary,
and updated the tests which relied on this code.

* public/shared/common-remote.js:
(CommonRemoteAPI.prototype.postJSON):
(CommonRemoteAPI.prototype.postJSONWithStatus):
(CommonRemoteAPI.prototype.postFormData):
(CommonRemoteAPI.prototype.postFormDataWithStatus):
* public/v3/privileged-api.js:
(PrivilegedAPI.prototype.sendRequest):
* public/v3/remote.js:
(BrowserRemoteAPI.prototype.sendHttpRequest):
(BrowserRemoteAPI.prototype.sendHttpRequestWithFormData):
(BrowserRemoteAPI):
* server-tests/api-uploaded-file.js:
* tools/js/remote.js:
(NodeRemoteAPI.prototype.sendHttpRequest):
(NodeRemoteAPI.prototype.sendHttpRequestWithFormData):
(NodeRemoteAPI):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214285 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoAdd support for showing the platform data interaction caret
wenson_hsieh@apple.com [Wed, 22 Mar 2017 23:21:38 +0000 (23:21 +0000)]
Add support for showing the platform data interaction caret
https://bugs.webkit.org/show_bug.cgi?id=169898
<rdar://problem/31146295>

Reviewed by Beth Dakin.

Source/WebKit/mac:

Plumb the dataInteractionCaretRect through WebView SPI to WebKit1 clients.

* WebView/WebView.mm:
(-[WebView _dataInteractionCaretRect]):
* WebView/WebViewPrivate.h:

Source/WebKit2:

Add support for showing the caret rect when data interacting over an editable target. To do this, we add
boilerplate plumbing to inform the WKContentView whenever the WebPageProxy's current caret rect changes.

* Platform/spi/ios/UIKitSPI.h:

Define IPI to retrieve the default insertion caret color.

* UIProcess/PageClient.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::didPerformDragControllerAction):
(WebKit::WebPageProxy::dragEnded):
(WebKit::WebPageProxy::didEndDragging):
(WebKit::WebPageProxy::resetCurrentDragInformation):
(WebKit::WebPageProxy::setDragCaretRect):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/ios/PageClientImplIOS.h:
* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::didChangeDataInteractionCaretRect):
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView cleanupInteraction]):
(-[WKContentView insertionPointColor]):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::dragEnded):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214284 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[GTK] Honor GTK+ font settings
mcatanzaro@igalia.com [Wed, 22 Mar 2017 23:18:54 +0000 (23:18 +0000)]
[GTK] Honor GTK+ font settings
https://bugs.webkit.org/show_bug.cgi?id=82889

Reviewed by Carlos Garcia Campos.

After much discussion with Behdad and Martin (who is still not completely convinced I think
:) I want to merge cairo font options into the Fontconfig pattern used for rendering using
cairo_ft_font_options_substitute(). This is how the API was designed to be used anyway.
Fontconfig will still have final say over whether to actually respect the desktop settings
or not, so it can still choose to ignore the desktop's settings, but I don't think it makes
sense to have desktop-wide font settings and not tell Fontconfig about them, especially when
the whole point of WebKitGTK+ is desktop integration. This should also reduce complaints
that we're not following desktop settings and that we're drawing fonts differently than
Firefox.

* PlatformGTK.cmake:
* platform/graphics/cairo/CairoUtilities.cpp:
(WebCore::getDefaultCairoFontOptions):
* platform/graphics/cairo/CairoUtilities.h:
* platform/graphics/freetype/FontCacheFreeType.cpp:
(WebCore::createFontConfigPatternForCharacters):
(WebCore::strongAliasesForFamily):
(WebCore::FontCache::createFontPlatformData):
* platform/graphics/freetype/FontPlatformDataFreeType.cpp:
(WebCore::getDefaultFontconfigOptions):
(WebCore::getDefaultCairoFontOptions): Deleted.
* platform/graphics/gtk/GdkCairoUtilities.cpp:
(getDefaultCairoFontOptions):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214283 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[Modern Media Controls] Controls size lags behind media size
commit-queue@webkit.org [Wed, 22 Mar 2017 22:51:59 +0000 (22:51 +0000)]
[Modern Media Controls] Controls size lags behind media size
https://bugs.webkit.org/show_bug.cgi?id=169962
<rdar://problem/30340293>

Patch by Antoine Quint <graouts@apple.com> on 2017-03-22
Reviewed by Dean Jackson.

Source/WebCore:

We used to dispatch the "resize" event to the shadow root on a timer and commit changes to the DOM
when handling it inside of a requestAnimationFrame() callback. We now dispatch the "resize" event
as a post-layout task and commit to the DOM right away which ensures immediate size changes of the
controls.

* Modules/modern-media-controls/controls/ios-inline-media-controls.js:
(IOSInlineMediaControls.prototype.layout):
* Modules/modern-media-controls/controls/macos-fullscreen-media-controls.js:
(MacOSFullscreenMediaControls.prototype.layout):
* Modules/modern-media-controls/controls/macos-inline-media-controls.js:
(MacOSInlineMediaControls.prototype.layout):
Making some layout() safer by returning early in case they're called during the object construction
phase, due to the LayoutNode "width" setter now calling that method.

* Modules/modern-media-controls/controls/media-controls.js:
(MediaControls.prototype.get width):
(MediaControls.prototype.set width):
Calling layout() in the "width" setter to ensure that the inline controls layout is invalidated as
soon as the controls width changes instead of waiting on the next rAF to update the layout.

* Modules/modern-media-controls/controls/scheduler.js:
(const.scheduler.new.prototype.flushScheduledLayoutCallbacks):
Add a new method to immediately call queued layouts.

* Modules/modern-media-controls/media/media-controller.js:
(MediaController.prototype.handleEvent):
Flush all queued layouts when handling a "resize" event.

* html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::layoutSizeChanged):
Queue the "resize" event as a post-layout task.

(WebCore::HTMLMediaElement::contextDestroyed):
* html/HTMLMediaElement.h:
Remove the queue we no longer use.

LayoutTests:

Updating an existing test to use a "resize" event handler to check that a size change
on the element is reflected on the shadow root. This would have failed prior to this
patch. Also rebaselining a test which now reports correct values.

* media/modern-media-controls/media-controller/media-controller-resize-expected.txt:
* media/modern-media-controls/media-controller/media-controller-resize.html:
* media/modern-media-controls/media-controller/media-controller-scale-factor-audio-expected.txt:
* media/modern-media-controls/media-controller/media-controller-scale-factor-audio.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214282 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoActivate remaining webrtc tests
commit-queue@webkit.org [Wed, 22 Mar 2017 22:51:22 +0000 (22:51 +0000)]
Activate remaining webrtc tests
https://bugs.webkit.org/show_bug.cgi?id=169957

Unreviewed.

Patch by Youenn Fablet <youenn@apple.com> on 2017-03-22

* platform/mac-wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214281 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoComponentBase should enqueue itself to render when it becomes connected
rniwa@webkit.org [Wed, 22 Mar 2017 22:50:57 +0000 (22:50 +0000)]
ComponentBase should enqueue itself to render when it becomes connected
https://bugs.webkit.org/show_bug.cgi?id=169905

Reviewed by Antti Koivisto.

When a component becomes connected to a document, enqueue itself to render automatically.
Also added the support for boolean attribute to ComponentBase.createElement.

* ReadMe.md: Added an instruction to raise the upload limit per r214065.
* browser-tests/component-base-tests.js: Added tests for the new behavior and createElement. Also moved
the tests related to enqueueToRenderOnResize out of defineElement tests.

* browser-tests/index.html:
(BrowsingContext.prototype.constructor): Override requestAnimationFrame so that the callback would be
involved immediately durign testing.

* public/v3/components/base.js:
(ComponentBase): Enqueue itself to render during construction if custom elements is not available.
(ComponentBase.defineElement):
(ComponentBase.defineElement.elementClass.prototype.connectedCallback): Enqueue itself to render when
the component's element became connected.
(ComponentBase.createElement): Use Array.isArray instead of instanceof to make it work with arrays made
in other realms (global objects) during testing. Added the support for boolean attributes. Setting an
attribute value to true would set the attribute, and setting it to false would not set the attribute.
(ComponentBase.useNativeCustomElements): Added. True iff window.customElements is defined.

* public/v3/components/chart-pane-base.js:
(ChartPaneBase.prototype.render): No longer need to call enqueueToRender on the commit log viewer.

* public/v3/components/commit-log-viewer.js:
(CommitLogViewer.prototype.render): No longer need to call enqueueToRender on the spinner icon.

* public/v3/models/time-series.js:
(TimeSeries): Made this a proper class declaration now that we don't include data.js after r213300.

* public/v3/pages/chart-pane.js:
(ChartPane.prototype._renderActionToolbar): No longer need to call enqueueToRender on the close icon.

* public/v3/pages/summary-page.js:
(SummaryPage.prototype._renderCell): No longer need to call enqueueToRender on the spinner icon.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214280 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoTry to fix the Mac CMake build after r214266.
aestes@apple.com [Wed, 22 Mar 2017 22:25:09 +0000 (22:25 +0000)]
Try to fix the Mac CMake build after r214266.

* PlatformMac.cmake:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214279 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoOnly run volume-related autoplay policy tests on Mac.
mrajca@apple.com [Wed, 22 Mar 2017 22:22:11 +0000 (22:22 +0000)]
Only run volume-related autoplay policy tests on Mac.
https://bugs.webkit.org/show_bug.cgi?id=169969

Reviewed by Alex Christensen.

iOS does not support volume changes to media elements so these tests are timing out.

* TestWebKitAPI/Tests/WebKit2Cocoa/WebsitePolicies.mm:
(TEST):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214278 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoWebKit should disallow beforeunload alerts from web pages users have never interacted...
cdumez@apple.com [Wed, 22 Mar 2017 21:14:23 +0000 (21:14 +0000)]
WebKit should disallow beforeunload alerts from web pages users have never interacted with
https://bugs.webkit.org/show_bug.cgi?id=169936
<rdar://problem/23798897>

Reviewed by Brent Fulgham.

LayoutTests/imported/w3c:

* web-platform-tests/html/browsers/browsing-the-web/unloading-documents/beforeunload-canceling-expected.txt:
* web-platform-tests/html/webappapis/scripting/events/compile-event-handler-settings-objects-expected.txt:
Rebaseline now that the CONFIRM MESSAGE lines are now longer shown. This is because there is no user interaction
with the page.

Source/WebCore:

WebKit should disallow beforeunload alerts from web pages users have never interacted with.
This reduces the risk of annoyance to the user and is allowed by the specification:
- https://html.spec.whatwg.org/multipage/browsers.html#prompt-to-unload-a-document (Step 8):
which says:
"""
The user agent is encouraged to avoid asking the user for confirmation if it judges that doing
so would be annoying, deceptive, or pointless. A simple heuristic might be that if the user
has not interacted with the document, the user agent would not ask for confirmation before
unloading it.
"""

Firefox already implements this, Chrome does not.

Tests: fast/events/beforeunload-alert-no-user-interaction.html
       fast/events/beforeunload-alert-user-interaction.html
       fast/events/beforeunload-alert-user-interaction2.html

* loader/FrameLoader.cpp:
(WebCore::shouldAskForNavigationConfirmation):
(WebCore::FrameLoader::dispatchBeforeUnloadEvent):

LayoutTests:

* fast/events/before-unload-return-string-conversion-expected.txt:
* fast/events/before-unload-returnValue-expected.txt:
Rebaseline now that the CONFIRM MESSAGE is no longer shown. This is because there is
no user interaction with the page.

* fast/events/beforeunload-alert-no-user-interaction-expected.txt: Added.
* fast/events/beforeunload-alert-no-user-interaction.html: Added.
* fast/events/beforeunload-alert-user-interaction-expected.txt: Added.
* fast/events/beforeunload-alert-user-interaction.html: Added.
* fast/events/beforeunload-alert-user-interaction2-expected.txt: Added.
* fast/events/beforeunload-alert-user-interaction2.html: Added.
Add layout test coverage.

* fast/loader/form-submission-after-beforeunload-cancel.html:
* fast/loader/show-only-one-beforeunload-dialog.html:
* http/tests/misc/iframe-beforeunload-dialog-matching-ancestor-securityorigin.html:
* http/tests/misc/iframe-beforeunload-dialog-not-matching-ancestor-securityorigin.html:
Simulate user interaction with the page so that the CONFIRM MESSAGE log lines are still
shown.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214277 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoASan violation in IconLoader::stopLoading
bfulgham@apple.com [Wed, 22 Mar 2017 20:46:31 +0000 (20:46 +0000)]
ASan violation in IconLoader::stopLoading
https://bugs.webkit.org/show_bug.cgi?id=169960
<rdar://problem/30577691>

Reviewed by David Kilzer.

DocumentLoader::finishLoadingIcon handles the life cycle of the IconLoader. Once this method is called,
we should return immediately rather than attempt to make further modifications to the IconLoader.

No new tests due to lack of test features (see https://bugs.webkit.org/show_bug.cgi?id=164895). Easily
tested in MiniBrowser under ASan visiting websites with icons.

* loader/icon/IconLoader.cpp:
(WebCore::IconLoader::notifyFinished):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214276 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoAX: WebKit is returning the wrong rangeForLine
n_wang@apple.com [Wed, 22 Mar 2017 20:30:41 +0000 (20:30 +0000)]
AX: WebKit is returning the wrong rangeForLine
https://bugs.webkit.org/show_bug.cgi?id=169940

Reviewed by Chris Fleizach.

Source/WebCore:

The AXRangeForLine is being calculated using VisiblePostition, so
when we try to use the index we should validate it using VisiblePosition.

Changes are covered in the modified test.

* accessibility/AXObjectCache.cpp:
(WebCore::AXObjectCache::traverseToOffsetInRange):
(WebCore::AXObjectCache::characterOffsetForIndex):

LayoutTests:

* accessibility/mac/range-for-contenteditable-newline-expected.txt:
* accessibility/mac/range-for-contenteditable-newline.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214274 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[iOS] downloadableBinaryFontsEnabled is set to false in -initSimpleHTMLDocumentWithSt...
mitz@apple.com [Wed, 22 Mar 2017 20:30:11 +0000 (20:30 +0000)]
[iOS] downloadableBinaryFontsEnabled is set to false in -initSimpleHTMLDocumentWithStyle:frame:preferences:groupName:
https://bugs.webkit.org/show_bug.cgi?id=169967
<rdar://problem/30406898>

Reviewed by Anders Carlsson.

* WebView/WebView.mm:
(-[WebView initSimpleHTMLDocumentWithStyle:frame:preferences:groupName:]): Removed call to
  Settings::downloadableBinaryFontsEnabled. The default is true.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214273 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[JSC] Use jsNontrivialString for Number toString operations
utatane.tea@gmail.com [Wed, 22 Mar 2017 20:07:21 +0000 (20:07 +0000)]
[JSC] Use jsNontrivialString for Number toString operations
https://bugs.webkit.org/show_bug.cgi?id=169965

Reviewed by Mark Lam.

JSTests:

* stress/to-string-int32.js: Added.
(shouldBe):
(toString10):
(expected):

Source/JavaScriptCore:

After single character check, produced string is always longer than 1.
Thus, we can use jsNontrivialString.

* runtime/NumberPrototype.cpp:
(JSC::int32ToStringInternal):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214272 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoResource Load Statistics: Fix partitioning bug for client-side cookie access
wilander@apple.com [Wed, 22 Mar 2017 19:43:25 +0000 (19:43 +0000)]
Resource Load Statistics: Fix partitioning bug for client-side cookie access
https://bugs.webkit.org/show_bug.cgi?id=169906
<rdar://problem/31164456>

Reviewed by Alex Christensen.

Source/WebCore:

The existing test case was expanded to cover this change.

* platform/network/NetworkStorageSession.h:
    Moved the two cookieStoragePartition() functions into the class.
    Also declared them const.
* platform/network/cf/NetworkStorageSessionCFNet.cpp:
(WebCore::NetworkStorageSession::cookieStoragePartition):
    Now checks whether it should partition or not.
(WebCore::getPartitioningDomain):
    Inline convenience function.
(WebCore::NetworkStorageSession::shouldPartitionCookies):
    Renamed since it now receives a top privately controlled domain
    instead of a host.
(WebCore::cookieStoragePartition): Deleted.
    This moved into NetworkStorageSession.
(WebCore::hostIsInDomain): Deleted.
    No longer needed since we generate the top privately controlled
    domain for both the resource and the partition.
(WebCore::NetworkStorageSession::shouldPartitionCookiesForHost): Deleted.
    Renamed WebCore::NetworkStorageSession::shouldPartitionCookies().
* platform/network/mac/CookieJarMac.mm:
(WebCore::cookiesInPartitionForURL):
    Now calls WebCore::NetworkStorageSession::cookieStoragePartition().
(WebCore::setCookiesFromDOM):
    Now calls WebCore::NetworkStorageSession::cookieStoragePartition().
* platform/network/mac/ResourceHandleMac.mm:
(WebCore::ResourceHandle::createNSURLConnection):
    Now calls WebCore::NetworkStorageSession::cookieStoragePartition().

Source/WebKit2:

* NetworkProcess/cocoa/NetworkDataTaskCocoa.mm:
(WebKit::NetworkDataTaskCocoa::NetworkDataTaskCocoa):
    This function no longer asks the NetworkStorageSession
    whether it should partition or not. That decision is
    now made as part of
    WebCore::NetworkStorageSession::cookieStoragePartition().
    The empty partition signals no partitioning.

LayoutTests:

* http/tests/loading/resourceLoadStatistics/partitioned-cookies-with-and-without-user-interaction-expected.txt:
* http/tests/loading/resourceLoadStatistics/resources/get-cookies.php:
    Now also outputs document.cookie client-side.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214270 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoMedia files served without an extension will not load in <video> tag.
jer.noble@apple.com [Wed, 22 Mar 2017 19:32:38 +0000 (19:32 +0000)]
Media files served without an extension will not load in <video> tag.
https://bugs.webkit.org/show_bug.cgi?id=169895

Reviewed by Eric Carlson.

Source/WebCore:

Test: http/tests/media/video-served-as-text-with-type.html

Media files served without an extension will commonly have no MIME type (or incorrect MIME type) information
delivered in the HTTP header. In situations like these, if media referenced by a <source> tag with a type=
attribute containing MIME type (and optionally, codec) information, that type should be used preferentially.

* platform/graphics/MediaPlayer.h:
(WebCore::MediaPlayer::contentMIMEType):
(WebCore::MediaPlayer::contentTypeCodecs):
(WebCore::MediaPlayer::contentMIMETypeWasInferredFromExtension):
* platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.mm:
(WebCore::MediaPlayerPrivateAVFoundationObjC::createAVAssetForURL):

LayoutTests:

* http/tests/media/video-served-as-text-with-type-expected.txt: Added.
* http/tests/media/video-served-as-text-with-type.html: Added.
* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214269 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[QuickLook] Move QuickLook.{h,mm} to platform/ios/
aestes@apple.com [Wed, 22 Mar 2017 19:23:12 +0000 (19:23 +0000)]
[QuickLook] Move QuickLook.{h,mm} to platform/ios/
https://bugs.webkit.org/show_bug.cgi?id=169963

Reviewed by Andreas Kling.

* WebCore.xcodeproj/project.pbxproj:
* platform/ios/QuickLook.h: Renamed from Source/WebCore/loader/ios/QuickLook.h.
* platform/ios/QuickLook.mm: Renamed from Source/WebCore/loader/ios/QuickLook.mm.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214268 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoRemove duplicate TestExpectations entry.
ryanhaddad@apple.com [Wed, 22 Mar 2017 18:20:56 +0000 (18:20 +0000)]
Remove duplicate TestExpectations entry.

Unreviewed test gardening.

* TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214267 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[QuickLook] Rename QuickLookHandle to PreviewLoader
aestes@apple.com [Wed, 22 Mar 2017 18:11:08 +0000 (18:11 +0000)]
[QuickLook] Rename QuickLookHandle to PreviewLoader
https://bugs.webkit.org/show_bug.cgi?id=169961

Reviewed by Andreas Kling.

Source/WebCore:

* WebCore.xcodeproj/project.pbxproj:
* loader/EmptyClients.cpp:
* loader/FrameLoaderClient.h:
* loader/ResourceLoader.cpp:
(WebCore::ResourceLoader::isQuickLookResource):
* loader/ResourceLoader.h:
* loader/SubresourceLoader.cpp:
(WebCore::SubresourceLoader::shouldCreatePreviewLoaderForResponse):
(WebCore::SubresourceLoader::didReceiveResponse):
(WebCore::SubresourceLoader::didReceiveData):
(WebCore::SubresourceLoader::didReceiveBuffer):
(WebCore::SubresourceLoader::didFinishLoading):
(WebCore::SubresourceLoader::didFail):
(WebCore::SubresourceLoader::shouldCreateQuickLookHandleForResponse): Deleted.
* loader/SubresourceLoader.h:
* loader/ios/PreviewLoader.h: Copied from Source/WebCore/loader/ios/QuickLook.h.
* loader/ios/PreviewLoader.mm: Added.
(testingClient):
(emptyClient):
(-[WebPreviewLoader initWithResourceLoader:resourceResponse:]):
(-[WebPreviewLoader appendDataArray:]):
(-[WebPreviewLoader finishedAppending]):
(-[WebPreviewLoader failed]):
(-[WebPreviewLoader _sendDidReceiveResponseIfNecessary]):
(-[WebPreviewLoader connection:didReceiveData:lengthReceived:]):
(-[WebPreviewLoader connectionDidFinishLoading:]):
(isQuickLookPasswordError):
(-[WebPreviewLoader connection:didFailWithError:]):
(WebCore::PreviewLoader::PreviewLoader):
(WebCore::PreviewLoader::~PreviewLoader):
(WebCore::PreviewLoader::shouldCreateForMIMEType):
(WebCore::PreviewLoader::create):
(WebCore::PreviewLoader::didReceiveData):
(WebCore::PreviewLoader::didReceiveBuffer):
(WebCore::PreviewLoader::didFinishLoading):
(WebCore::PreviewLoader::didFail):
(WebCore::PreviewLoader::setClientForTesting):
* loader/ios/QuickLook.h:
(): Deleted.
* loader/ios/QuickLook.mm:
(WebCore::QLPreviewGetSupportedMIMETypesSet):
(WebCore::removeQLPreviewConverterForURL):
(WebCore::registerQLPreviewConverterIfNeeded):
(WebCore::QLPreviewProtocol):
(WebCore::isQuickLookPreviewURL):
(qlPreviewConverterDictionaryMutex): Deleted.
(QLPreviewConverterDictionary): Deleted.
(QLContentDictionary): Deleted.
(addQLPreviewConverterWithFileForURL): Deleted.
(createQLPreviewProtocol): Deleted.
(testingClient): Deleted.
(emptyClient): Deleted.
(-[WebPreviewLoader initWithResourceLoader:resourceResponse:quickLookHandle:]): Deleted.
(-[WebPreviewLoader appendDataArray:]): Deleted.
(-[WebPreviewLoader finishedAppending]): Deleted.
(-[WebPreviewLoader failed]): Deleted.
(-[WebPreviewLoader _sendDidReceiveResponseIfNecessary]): Deleted.
(-[WebPreviewLoader connection:didReceiveData:lengthReceived:]): Deleted.
(-[WebPreviewLoader connectionDidFinishLoading:]): Deleted.
(isQuickLookPasswordError): Deleted.
(-[WebPreviewLoader connection:didFailWithError:]): Deleted.
(WebCore::QuickLookHandle::QuickLookHandle): Deleted.
(WebCore::QuickLookHandle::~QuickLookHandle): Deleted.
(WebCore::QuickLookHandle::shouldCreateForMIMEType): Deleted.
(WebCore::QuickLookHandle::create): Deleted.
(WebCore::QuickLookHandle::didReceiveData): Deleted.
(WebCore::QuickLookHandle::didReceiveBuffer): Deleted.
(WebCore::QuickLookHandle::didFinishLoading): Deleted.
(WebCore::QuickLookHandle::didFail): Deleted.
(WebCore::QuickLookHandle::setClientForTesting): Deleted.
* platform/network/ResourceHandle.h:
* platform/network/ios/PreviewLoaderClient.h: Renamed from Source/WebCore/platform/network/ios/QuickLookHandleClient.h.
(WebCore::PreviewLoaderClient::~PreviewLoaderClient):
* testing/Internals.cpp:
(WebCore::Internals::resetToConsistentState):
(WebCore::Internals::setQuickLookPassword):
* testing/MockPreviewLoaderClient.cpp: Renamed from Source/WebCore/testing/MockQuickLookHandleClient.cpp.
(WebCore::MockPreviewLoaderClient::singleton):
(WebCore::MockPreviewLoaderClient::didRequestPassword):
* testing/MockPreviewLoaderClient.h: Renamed from Source/WebCore/testing/MockQuickLookHandleClient.h.

Source/WebKit/mac:

* WebCoreSupport/WebFrameLoaderClient.h:
* WebCoreSupport/WebFrameLoaderClient.mm:
(WebFrameLoaderClient::createPreviewLoaderClient):
(WebFrameLoaderClient::createQuickLookHandleClient): Deleted.

Source/WebKit2:

* WebKit2.xcodeproj/project.pbxproj:
* WebProcess/WebCoreSupport/WebFrameLoaderClient.h:
* WebProcess/WebCoreSupport/ios/WebFrameLoaderClientIOS.mm:
(WebKit::WebFrameLoaderClient::createPreviewLoaderClient):
(WebKit::WebFrameLoaderClient::createQuickLookHandleClient): Deleted.
* WebProcess/WebCoreSupport/ios/WebPreviewLoaderClient.cpp: Renamed from Source/WebKit2/WebProcess/WebCoreSupport/ios/WebQuickLookHandleClient.cpp.
(WebKit::WebPreviewLoaderClient::WebPreviewLoaderClient):
(WebKit::WebPreviewLoaderClient::~WebPreviewLoaderClient):
(WebKit::WebPreviewLoaderClient::didReceiveDataArray):
(WebKit::WebPreviewLoaderClient::didFinishLoading):
(WebKit::WebPreviewLoaderClient::didFail):
(WebKit::WebPreviewLoaderClient::didRequestPassword):
(WebKit::WebPreviewLoaderClient::didReceivePassword):
* WebProcess/WebCoreSupport/ios/WebPreviewLoaderClient.h: Renamed from Source/WebKit2/WebProcess/WebCoreSupport/ios/WebQuickLookHandleClient.h.
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::didReceivePasswordForQuickLookDocument):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214266 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[Readable Streams API] Implement ReadableStreamBYOBRequest respond() (readable stream...
commit-queue@webkit.org [Wed, 22 Mar 2017 18:07:49 +0000 (18:07 +0000)]
[Readable Streams API] Implement ReadableStreamBYOBRequest respond() (readable stream state)
https://bugs.webkit.org/show_bug.cgi?id=169759

Patch by Romain Bellessort <romain.bellessort@crf.canon.fr> on 2017-03-22
Reviewed by Youenn Fablet.

LayoutTests/imported/w3c:

Updated test expectations.

* web-platform-tests/streams/readable-byte-streams/general-expected.txt:
* web-platform-tests/streams/readable-byte-streams/general.dedicatedworker-expected.txt:

Source/WebCore:

Implemented readable state part of respond() function. Parts of code cannot
currently be reached as they require ReadableStreamBYOBReader to be implemented.

Added new tests and updated expectations.

* Modules/streams/ReadableByteStreamInternals.js:
(readableByteStreamControllerEnqueue): Name shortened.
(readableByteStreamControllerRespondInternal): Updated with readable state case.
(cloneArrayBuffer): Added.
(readableByteStreamControllerRespondInReadableState): Added.
(readableByteStreamControllerRespondInClosedState): Updated.
(readableByteStreamControllerProcessPullDescriptors): Added.
(readableByteStreamControllerFillDescriptorFromQueue): Added.
(readableByteStreamControllerShiftPendingDescriptor): Name shortened.
(readableByteStreamControllerCommitDescriptor): Name shortened.
(readableByteStreamControllerConvertDescriptor): Name shortened.

LayoutTests:

Added new tests to check code that can currently be reached.

* streams/readable-stream-byob-request-expected.txt: Updated.
* streams/readable-stream-byob-request.js: Updated.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214265 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[WK2] Add SPI to notify clients when data interaction is beginning or ending
wenson_hsieh@apple.com [Wed, 22 Mar 2017 18:04:19 +0000 (18:04 +0000)]
[WK2] Add SPI to notify clients when data interaction is beginning or ending
https://bugs.webkit.org/show_bug.cgi?id=169952
<rdar://problem/31187804>

Reviewed by Beth Dakin.

Adds two new UI delegate SPI hooks.

* UIProcess/API/Cocoa/WKUIDelegatePrivate.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214264 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoMove LayoutTests/webrtc/rtcconfiguration-icecandidatepoolsize.html to web-platform...
commit-queue@webkit.org [Wed, 22 Mar 2017 18:02:31 +0000 (18:02 +0000)]
Move LayoutTests/webrtc/rtcconfiguration-icecandidatepoolsize.html to web-platform-tests
https://bugs.webkit.org/show_bug.cgi?id=169958

Patch by Youenn Fablet <youenn@apple.com> on 2017-03-22
Reviewed by Jon Lee.

LayoutTests/imported/w3c:

* web-platform-tests/webrtc/rtcpeerconnection/rtcconfiguration-icecandidatepoolsize-expected.txt: Added.
* web-platform-tests/webrtc/rtcpeerconnection/rtcconfiguration-icecandidatepoolsize.html: Added.

LayoutTests:

* webrtc/rtcconfiguration-icecandidatepoolsize-expected.txt: Removed.
* webrtc/rtcconfiguration-icecandidatepoolsize.html: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214263 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoRTCPeerConnection is crashing if no backend is created
commit-queue@webkit.org [Wed, 22 Mar 2017 18:00:22 +0000 (18:00 +0000)]
RTCPeerConnection is crashing if no backend is created
https://bugs.webkit.org/show_bug.cgi?id=169938

Patch by Youenn Fablet <youenn@apple.com> on 2017-03-22
Reviewed by Geoffrey Garen.

RTCPeerConnection should not be visible if it is not able to create a backend.
In case when this happens anyway, it should be able to be stopped and destroyed without crashing.
To do so, we need to set its state to closed.

* Modules/mediastream/RTCPeerConnection.cpp:
(WebCore::RTCPeerConnection::RTCPeerConnection):
(WebCore::RTCPeerConnection::close): Cosmetic change, just in case stop becomes reentrant at some point.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214262 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoWebAssembly: name ExecState consistently
jfbastien@apple.com [Wed, 22 Mar 2017 17:56:29 +0000 (17:56 +0000)]
WebAssembly: name ExecState consistently
https://bugs.webkit.org/show_bug.cgi?id=169954

Reviewed by Saam Barati.

No functional change.

* wasm/js/JSWebAssemblyCompileError.cpp:
(JSC::JSWebAssemblyCompileError::create):
(JSC::createJSWebAssemblyCompileError):
* wasm/js/JSWebAssemblyCompileError.h:
(JSC::JSWebAssemblyCompileError::create):
* wasm/js/JSWebAssemblyLinkError.cpp:
(JSC::JSWebAssemblyLinkError::create):
(JSC::createJSWebAssemblyLinkError):
* wasm/js/JSWebAssemblyLinkError.h:
(JSC::JSWebAssemblyLinkError::create):
* wasm/js/JSWebAssemblyRuntimeError.cpp:
(JSC::JSWebAssemblyRuntimeError::create):
* wasm/js/JSWebAssemblyRuntimeError.h:
(JSC::JSWebAssemblyRuntimeError::create):
* wasm/js/WebAssemblyInstanceConstructor.cpp:
(JSC::callJSWebAssemblyInstance):
* wasm/js/WebAssemblyMemoryConstructor.cpp:
(JSC::callJSWebAssemblyMemory):
* wasm/js/WebAssemblyModuleConstructor.cpp:
(JSC::callJSWebAssemblyModule):
(JSC::WebAssemblyModuleConstructor::createModule):
* wasm/js/WebAssemblyModuleRecord.cpp:
(JSC::WebAssemblyModuleRecord::link):
(JSC::dataSegmentFail):
(JSC::WebAssemblyModuleRecord::evaluate):
* wasm/js/WebAssemblyPrototype.cpp:
(JSC::webAssemblyFunctionValidate):
(JSC::webAssemblyFunctionCompile):
* wasm/js/WebAssemblyTableConstructor.cpp:
(JSC::callJSWebAssemblyTable):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214261 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoWebAssembly: test module namespace object for WebAssembly.Instance
jfbastien@apple.com [Wed, 22 Mar 2017 17:54:36 +0000 (17:54 +0000)]
WebAssembly: test module namespace object for WebAssembly.Instance
https://bugs.webkit.org/show_bug.cgi?id=169951

Reviewed by Saam Barati.

* wasm/js-api/test_basic_api.js:
(const.c.in.constructorProperties.switch):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214260 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoWebAssembly: constructors without new don't throw
jfbastien@apple.com [Wed, 22 Mar 2017 17:52:55 +0000 (17:52 +0000)]
WebAssembly: constructors without new don't throw
https://bugs.webkit.org/show_bug.cgi?id=165995

Reviewed by Saam Barati.

JSTests:

* wasm/js-api/test_basic_api.js:
(const.c.in.constructorProperties.switch):

Source/JavaScriptCore:

* wasm/js/WebAssemblyCompileErrorConstructor.cpp:
(JSC::constructJSWebAssemblyCompileError):
(JSC::callJSWebAssemblyCompileError):
* wasm/js/WebAssemblyLinkErrorConstructor.cpp:
(JSC::constructJSWebAssemblyLinkError):
(JSC::callJSWebAssemblyLinkError):
* wasm/js/WebAssemblyRuntimeErrorConstructor.cpp:
(JSC::constructJSWebAssemblyRuntimeError):
(JSC::callJSWebAssemblyRuntimeError):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214259 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[Modern Media Controls] Rendering glitches for fullscreen controls while dragging
graouts@webkit.org [Wed, 22 Mar 2017 17:51:10 +0000 (17:51 +0000)]
[Modern Media Controls] Rendering glitches for fullscreen controls while dragging
https://bugs.webkit.org/show_bug.cgi?id=169949
<rdar://problem/30687803>

Reviewed by Eric Carlson.

Source/WebCore:

Clipping the controls bar has some unwelcome side effects and isn't necessary, so
we simply remove it. An existing test was amended to check that there is no clipping
applied to the controls bar.

* Modules/modern-media-controls/controls/macos-fullscreen-media-controls.css:
(.media-controls.mac.fullscreen > .controls-bar):

LayoutTests:

Add an assertion that checks that the controls bar is not clipped.

* media/modern-media-controls/macos-fullscreen-media-controls/macos-fullscreen-media-controls-controls-bar-styles-expected.txt:
* media/modern-media-controls/macos-fullscreen-media-controls/macos-fullscreen-media-controls-controls-bar-styles.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214258 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoAdd back the null check in TextureCacheCV::textureFromImage()
wilander@apple.com [Wed, 22 Mar 2017 17:48:51 +0000 (17:48 +0000)]
Add back the null check in TextureCacheCV::textureFromImage()
https://bugs.webkit.org/show_bug.cgi?id=169931
<rdar://problem/27139820>

Reviewed by Alex Christensen.

No new tests. Just adding back a check that was previously there.

* platform/graphics/cv/TextureCacheCV.mm:
(WebCore::TextureCacheCV::textureFromImage):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214257 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoWeb Inspector: Clicking DOM breakpoint marker should enable/disable breakpoints
mattbaker@apple.com [Wed, 22 Mar 2017 17:41:11 +0000 (17:41 +0000)]
Web Inspector: Clicking DOM breakpoint marker should enable/disable breakpoints
https://bugs.webkit.org/show_bug.cgi?id=169856
<rdar://problem/31133090>

Reviewed by Joseph Pecoraro.

Clicking DOM breakpoint marker should enable/disable breakpoints,
matching the behavior of the marker's context menu:

- If one or more breakpoints are disabled, click enables all
- If all breakpoints are enabled, click disables all

* UserInterface/Views/DOMTreeElement.js:
(WebInspector.DOMTreeElement.prototype._updateBreakpointStatus):
(WebInspector.DOMTreeElement.prototype._statusImageClicked):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214256 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUse AtomicString in RuleSet and RuleFeature
antti@apple.com [Wed, 22 Mar 2017 17:19:28 +0000 (17:19 +0000)]
Use AtomicString in RuleSet and RuleFeature
https://bugs.webkit.org/show_bug.cgi?id=119310
<rdar://problem/28214658>

Reviewed by Andreas Kling.

..instead of the plain AtomicStringImpl*. This introduces some ref churn but not too much.

* css/DocumentRuleSets.cpp:
(WebCore::DocumentRuleSets::ancestorClassRules):
(WebCore::DocumentRuleSets::ancestorAttributeRulesForHTML):
* css/DocumentRuleSets.h:
* css/ElementRuleCollector.cpp:
(WebCore::ElementRuleCollector::collectMatchingRules):
(WebCore::ElementRuleCollector::collectMatchingShadowPseudoElementRules):
* css/RuleFeature.cpp:
(WebCore::RuleFeatureSet::recursivelyCollectFeaturesFromSelector):
(WebCore::makeAttributeSelectorKey):
(WebCore::RuleFeatureSet::collectFeatures):
* css/RuleFeature.h:
* css/RuleSet.cpp:
(WebCore::RuleSet::addToRuleSet):
(WebCore::rulesCountForName):
(WebCore::RuleSet::addRule):
* css/RuleSet.h:
(WebCore::RuleSet::idRules):
(WebCore::RuleSet::classRules):
(WebCore::RuleSet::shadowPseudoElementRules):
(WebCore::RuleSet::tagRules):
* css/StyleResolver.h:
(WebCore::StyleResolver::hasSelectorForAttribute):
(WebCore::StyleResolver::hasSelectorForClass):
(WebCore::StyleResolver::hasSelectorForId):
* style/AttributeChangeInvalidation.cpp:
(WebCore::Style::mayBeAffectedByAttributeChange):
(WebCore::Style::AttributeChangeInvalidation::invalidateStyle):
* style/IdChangeInvalidation.cpp:
(WebCore::Style::mayBeAffectedByHostRules):
(WebCore::Style::mayBeAffectedBySlottedRules):
(WebCore::Style::IdChangeInvalidation::invalidateStyle):
* style/StyleSharingResolver.cpp:
(WebCore::Style::SharingResolver::resolve):
(WebCore::Style::SharingResolver::canShareStyleWithElement):
(WebCore::Style::SharingResolver::classNamesAffectedByRules):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214255 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoSafari sends empty "Access-Control-Request-Headers" in preflight request
commit-queue@webkit.org [Wed, 22 Mar 2017 17:18:27 +0000 (17:18 +0000)]
Safari sends empty "Access-Control-Request-Headers" in preflight request
https://bugs.webkit.org/show_bug.cgi?id=169851

Patch by Youenn Fablet <youenn@apple.com> on 2017-03-22
Reviewed by Chris Dumez.

LayoutTests/imported/w3c:

* web-platform-tests/fetch/api/cors/cors-preflight-expected.txt:
* web-platform-tests/fetch/api/cors/cors-preflight.js:

Source/WebCore:

Covered by updated test.

* loader/CrossOriginAccessControl.cpp:
(WebCore::createAccessControlPreflightRequest): Not adding "Access-Control-Request-Headers" to
request header if value is empty.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214254 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[DFG] Don't use ArraySlice intrinsic on MIPS
commit-queue@webkit.org [Wed, 22 Mar 2017 17:15:17 +0000 (17:15 +0000)]
[DFG] Don't use ArraySlice intrinsic on MIPS
https://bugs.webkit.org/show_bug.cgi?id=169721

Patch by Guillaume Emont <guijemont@igalia.com> on 2017-03-22
Reviewed by Yusuke Suzuki.

Like on x86, we don't have enough registers available for this.

* assembler/CPU.h:
(JSC::isMIPS): Added.
* dfg/DFGByteCodeParser.cpp:
(JSC::DFG::ByteCodeParser::handleIntrinsicCall):
Don't use the ArraySlice intrinsic on MIPS.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214253 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoXMLHttpRequest: getAllResponseHeaders() should lowercase header names before sorting
commit-queue@webkit.org [Wed, 22 Mar 2017 17:14:36 +0000 (17:14 +0000)]
XMLHttpRequest: getAllResponseHeaders() should lowercase header names before sorting
https://bugs.webkit.org/show_bug.cgi?id=169286

Patch by Youenn Fablet <youenn@apple.com> on 2017-03-22
Reviewed by Chris Dumez.

LayoutTests/imported/w3c:

* web-platform-tests/XMLHttpRequest/getallresponseheaders-cl-expected.txt:
* web-platform-tests/XMLHttpRequest/getallresponseheaders-expected.txt:

Source/WebCore:

Covered by rebased tests.

https://xhr.spec.whatwg.org/#the-getallresponseheaders()-method now mandates to lowercase the header names and
sort them before appending them. This ensures that differences in the way headers are processed by browsers are
not visible to web pages. This also moves towards lowercased/case-insensitive header processing which should be
the norm.

* xml/XMLHttpRequest.cpp:
(WebCore::XMLHttpRequest::getAllResponseHeaders): Lower-casing, then sorting, then serializing headers.
* xml/XMLHttpRequest.h: Storing getAllResponseHeaders result if case needed again.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214252 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoImplement the "noopener" feature for window.open()
cdumez@apple.com [Wed, 22 Mar 2017 16:15:16 +0000 (16:15 +0000)]
Implement the "noopener" feature for window.open()
https://bugs.webkit.org/show_bug.cgi?id=163960
<rdar://problem/28941679>

Reviewed by Sam Weinig.

LayoutTests/imported/w3c:

Import test coverage from WPT upstream.

* resources/import-expectations.json:
* web-platform-tests/common/PrefixedLocalStorage.js: Added.
(PrefixedLocalStorage):
(PrefixedLocalStorage.prototype.clear):
(PrefixedLocalStorage.prototype.url):
(PrefixedLocalStorage.prototype.prefixedKey):
(PrefixedLocalStorage.prototype.setItem):
(PrefixedLocalStorage.prototype.onSet):
(PrefixedLocalStorageTest):
(PrefixedLocalStorageTest.prototype.cleanup):
(PrefixedLocalStorageResource):
* web-platform-tests/common/w3c-import.log:
* web-platform-tests/html/browsers/windows/auxiliary-browsing-contexts/opener-closed-expected.txt: Added.
* web-platform-tests/html/browsers/windows/auxiliary-browsing-contexts/opener-closed.html: Added.
* web-platform-tests/html/browsers/windows/auxiliary-browsing-contexts/opener-expected.txt: Added.
* web-platform-tests/html/browsers/windows/auxiliary-browsing-contexts/opener-multiple-expected.txt: Added.
* web-platform-tests/html/browsers/windows/auxiliary-browsing-contexts/opener-multiple.html: Added.
* web-platform-tests/html/browsers/windows/auxiliary-browsing-contexts/opener-noopener-expected.txt: Added.
* web-platform-tests/html/browsers/windows/auxiliary-browsing-contexts/opener-noopener.html: Added.
* web-platform-tests/html/browsers/windows/auxiliary-browsing-contexts/opener-noreferrer-expected.txt: Added.
* web-platform-tests/html/browsers/windows/auxiliary-browsing-contexts/opener-noreferrer.html: Added.
* web-platform-tests/html/browsers/windows/auxiliary-browsing-contexts/opener.html: Added.
* web-platform-tests/html/browsers/windows/auxiliary-browsing-contexts/resources/close-opener.html: Added.
* web-platform-tests/html/browsers/windows/auxiliary-browsing-contexts/resources/message-window-opener.html: Added.
* web-platform-tests/html/browsers/windows/auxiliary-browsing-contexts/resources/multiple-opener.html: Added.
* web-platform-tests/html/browsers/windows/auxiliary-browsing-contexts/resources/no-opener.html: Added.
* web-platform-tests/html/browsers/windows/auxiliary-browsing-contexts/resources/open-closer.html: Added.
* web-platform-tests/html/browsers/windows/auxiliary-browsing-contexts/resources/w3c-import.log: Copied from LayoutTests/imported/w3c/web-platform-tests/html/browsers/windows/auxiliary-browsing-contexts/w3c-import.log.
* web-platform-tests/html/browsers/windows/auxiliary-browsing-contexts/w3c-import.log:

Source/WebCore:

Implement the "noopener" feature for window.open(). If 'noopener' is passed to window.open()
then the new Window's opener property will be null.

Tests: imported/w3c/web-platform-tests/html/browsers/windows/auxiliary-browsing-contexts/opener-closed.html
       imported/w3c/web-platform-tests/html/browsers/windows/auxiliary-browsing-contexts/opener-multiple.html
       imported/w3c/web-platform-tests/html/browsers/windows/auxiliary-browsing-contexts/opener-noopener.html
       imported/w3c/web-platform-tests/html/browsers/windows/auxiliary-browsing-contexts/opener-noreferrer.html
       imported/w3c/web-platform-tests/html/browsers/windows/auxiliary-browsing-contexts/opener.html

* page/DOMWindow.cpp:
(WebCore::DOMWindow::createWindow):
* page/WindowFeatures.cpp:
(WebCore::parseWindowFeatures):
(WebCore::setWindowFeature):
* page/WindowFeatures.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214251 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[Linux] determineArchitecture is not cross-compile aware
clopez@igalia.com [Wed, 22 Mar 2017 14:36:32 +0000 (14:36 +0000)]
[Linux] determineArchitecture is not cross-compile aware
https://bugs.webkit.org/show_bug.cgi?id=169886

Reviewed by Michael Catanzaro.

Try to detect the target architecture name from the GCC triplet
when cross-building.

The aarch64 architecture is renamed to arm64 (this is coherent with
the rest of the perl tooling that consults the architecture determined
in determineArchitecture(), the check for this arch is always done
with the arm64 name).

* Scripts/webkitdirs.pm: Remove the isARM() function: its dead code
and not cross-build aware.
(determineArchitecture): When crossbuilding, try to detect the target
architecture name from the GCC triplet.
Remove also the fallback option to run the arch command, because this
command doesn't print the expected on Mac/BSD, and on Linux is the
same than uname -m. See https://bugs.webkit.org/show_bug.cgi?id=152958#c6
(isCrossCompilation): Some cross-compilers (buildroot one for example)
don't define the --host option. Add another option to detect that
we are cross-building by building a dummy program and checking if
we can run it.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214250 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUnreviewed GTK+ gardening. Rebaseline fast/text/hyphenate-limit-lines.html.
carlosgc@webkit.org [Wed, 22 Mar 2017 11:59:33 +0000 (11:59 +0000)]
Unreviewed GTK+ gardening. Rebaseline fast/text/hyphenate-limit-lines.html.

* platform/gtk/fast/text/hyphenate-limit-lines-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214249 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUnreviewed GTK+ gardening. Remove platform expectation after r214246.
carlosgc@webkit.org [Wed, 22 Mar 2017 11:55:27 +0000 (11:55 +0000)]
Unreviewed GTK+ gardening. Remove platform expectation after r214246.

* platform/gtk/http/tests/loading/server-redirect-for-provisional-load-caching-expected.txt: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214248 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUnreviewed GTK+ gardening. Rebaseline several tests after r214244.
carlosgc@webkit.org [Wed, 22 Mar 2017 11:50:54 +0000 (11:50 +0000)]
Unreviewed GTK+ gardening. Rebaseline several tests after r214244.

* platform/gtk/accessibility/file-upload-button-stringvalue-expected.txt:
* platform/gtk/fast/forms/box-shadow-override-expected.png:
* platform/gtk/fast/forms/box-shadow-override-expected.txt:
* platform/gtk/fast/forms/file/file-input-direction-expected.png:
* platform/gtk/fast/forms/file/file-input-direction-expected.txt:
* platform/gtk/fast/forms/file/file-input-disabled-expected.png:
* platform/gtk/fast/forms/file/file-input-disabled-expected.txt:
* platform/gtk/fast/forms/form-element-geometry-expected.png:
* platform/gtk/fast/forms/form-element-geometry-expected.txt:
* platform/gtk/fast/forms/input-appearance-height-expected.png:
* platform/gtk/fast/forms/input-appearance-height-expected.txt:
* platform/gtk/fast/forms/input-value-expected.txt:
* platform/gtk/media/video-controls-captions-trackmenu-localized-expected.txt:
* platform/gtk/media/video-controls-captions-trackmenu-sorted-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214247 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[Soup] "Only from websites I visit" cookie policy is broken
svillar@igalia.com [Wed, 22 Mar 2017 10:09:23 +0000 (10:09 +0000)]
[Soup] "Only from websites I visit" cookie policy is broken
https://bugs.webkit.org/show_bug.cgi?id=168912

Reviewed by Carlos Garcia Campos.

Source/WebCore:

Do not reset the first party for cookies on redirects. That's properly done for the main
resource in DocumentLoader::willSendRequest and, in the case of subresources, is absolutely
wrong (which is what we were doing since r143931).

The most notable effect was that subresources loaded via redirects were effectively
bypassing the "no third party" policy for cookies.

Test: http/tests/security/cookies/third-party-cookie-blocking-redirect.html

* platform/network/soup/ResourceHandleSoup.cpp:
(WebCore::doRedirect):

Source/WebKit2:

Do not reset the first party for cookies on redirects. That's properly done for the main
resource in DocumentLoader::willSendRequest and, in the case of subresources, is absolutely
wrong (which is what we were doing since r143931).

The most notable effect was that subresources loaded via redirects were effectively
bypassing the "no third party" policy for cookies.

* NetworkProcess/soup/NetworkDataTaskSoup.cpp:
(WebKit::NetworkDataTaskSoup::continueHTTPRedirection):

LayoutTests:

* http/tests/security/cookies/third-party-cookie-blocking-redirect-expected.txt: Added.
* http/tests/security/cookies/third-party-cookie-blocking-redirect.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214246 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoAll http/ tests with PHP fail in Debian unstable with php7
svillar@igalia.com [Wed, 22 Mar 2017 10:07:14 +0000 (10:07 +0000)]
All http/ tests with PHP fail in Debian unstable with php7
https://bugs.webkit.org/show_bug.cgi?id=169913

Reviewed by Michael Catanzaro.

The problem is that the configuration file was still checking for php5 instead of php7.

* http/conf/debian-httpd-2.4-php7.conf:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214245 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoMake it possible to use WEB_UI_STRING macros to mark translatable strings in glib...
carlosgc@webkit.org [Wed, 22 Mar 2017 09:55:34 +0000 (09:55 +0000)]
Make it possible to use WEB_UI_STRING macros to mark translatable strings in glib based ports
https://bugs.webkit.org/show_bug.cgi?id=169672

Reviewed by Michael Catanzaro.

Source/WebCore:

WEB_UI_STRING macros are currently used by Cocoa based ports and Windows port. For other ports it simply returns
the given string, but it's not translated. In the GTK+ port we have all translatable strings from
LocalizedStrings.cpp copied in LocalizedStringsGtk.cpp to use the glib macros to mark them as translatable. In
most of the cases the strings are exactly in the same, in some others (the context menu ones) the string only
differs in the mnemonics used by GTK, and only in a few cases the strings are different. We could remove most of
that duplication if strings used by WEB_UI_STRING were marked as translatable in glib based ports too.

This patch makes that possible with the following changes:

  - Add WEB_UI_STRING_WITH_MNEMONIC() to handle the cases where the strings only differ in the mnemonics. It
    receives the string with mnemonics that is ignored in all ports except GTK+.
  - All platform specific strings have been moved to the platform files, to ensure scanners don't include
    strings from other ports.
  - For the same reason a new macro WEB_UI_CFSTRING has been also added, just to make sure that CF strings are
    only used by CF based ports. The implementation is the same as WEB_UI_STRING, it only changes the name to
    not confuse the scanners.
  - And again for the same reason, for non-CF and non-Glib cases WEB_UI_STRING is not used, since in those cases
    strings are not really translated and String::fromUTF8() is actually used. So, now we use String::fromUTF8()
    directly instead of WEB_UI_STRING to not confuse the scanners.

* CMakeLists.txt: Add LocalizedStrings.cpp to the compilation.
* PlatformGTK.cmake: Move LocalizedStringsGtk.cpp to the list of sources not depending on GTK+, since it no
longer uses GTK+.
* extract-localizable-strings.pl: Update it to handle the new macros.
* platform/LocalizedStrings.cpp:
(WebCore::formatLocalizedString):
(WebCore::contextMenuItemTagOpenLinkInNewWindow):
(WebCore::contextMenuItemTagDownloadLinkToDisk):
(WebCore::contextMenuItemTagOpenImageInNewWindow):
(WebCore::contextMenuItemTagCopyImageToClipboard):
(WebCore::contextMenuItemTagOpenFrameInNewWindow):
(WebCore::contextMenuItemTagCopy):
(WebCore::contextMenuItemTagGoBack):
(WebCore::contextMenuItemTagGoForward):
(WebCore::contextMenuItemTagStop):
(WebCore::contextMenuItemTagReload):
(WebCore::contextMenuItemTagCut):
(WebCore::contextMenuItemTagPaste):
(WebCore::contextMenuItemTagIgnoreSpelling):
(WebCore::contextMenuItemTagLearnSpelling):
(WebCore::contextMenuItemTagLookUpInDictionary):
(WebCore::contextMenuItemTagOpenLink):
(WebCore::contextMenuItemTagIgnoreGrammar):
(WebCore::contextMenuItemTagSpellingMenu):
(WebCore::contextMenuItemTagShowSpellingPanel):
(WebCore::contextMenuItemTagCheckSpelling):
(WebCore::contextMenuItemTagCheckSpellingWhileTyping):
(WebCore::contextMenuItemTagCheckGrammarWithSpelling):
(WebCore::contextMenuItemTagFontMenu):
(WebCore::contextMenuItemTagBold):
(WebCore::contextMenuItemTagItalic):
(WebCore::contextMenuItemTagUnderline):
(WebCore::contextMenuItemTagOutline):
(WebCore::contextMenuItemTagOpenVideoInNewWindow):
(WebCore::contextMenuItemTagOpenAudioInNewWindow):
(WebCore::contextMenuItemTagDownloadVideoToDisk):
(WebCore::contextMenuItemTagDownloadAudioToDisk):
(WebCore::contextMenuItemTagExitVideoFullscreen):
(WebCore::contextMenuItemTagMediaPlay):
(WebCore::contextMenuItemTagMediaPause):
(WebCore::contextMenuItemTagMediaMute):
(WebCore::contextMenuItemTagInspectElement):
(WebCore::AXAutoFillContactsLabel):
(WebCore::imageTitle):
(WebCore::validationMessageRangeUnderflowText):
(WebCore::validationMessageRangeOverflowText):
(WebCore::textTrackCountryAndLanguageMenuItemText):
(WebCore::textTrackLanguageMenuItemText):
(WebCore::closedCaptionTrackMenuItemText):
(WebCore::sdhTrackMenuItemText):
(WebCore::easyReaderTrackMenuItemText):
(WebCore::forcedTrackMenuItemText):
(WebCore::audioDescriptionTrackSuffixText):
(WebCore::webCryptoMasterKeyKeychainLabel):
* platform/LocalizedStrings.h:
* platform/cocoa/LocalizedStringsCocoa.mm:
(WebCore::copyImageUnknownFileLabel):
(WebCore::contextMenuItemTagSearchInSpotlight):
(WebCore::contextMenuItemTagSearchWeb):
(WebCore::contextMenuItemTagShowFonts):
(WebCore::contextMenuItemTagStyles):
(WebCore::contextMenuItemTagShowColors):
(WebCore::contextMenuItemTagSpeechMenu):
(WebCore::contextMenuItemTagStartSpeaking):
(WebCore::contextMenuItemTagStopSpeaking):
(WebCore::contextMenuItemTagCorrectSpellingAutomatically):
(WebCore::contextMenuItemTagSubstitutionsMenu):
(WebCore::contextMenuItemTagShowSubstitutions):
(WebCore::contextMenuItemTagSmartCopyPaste):
(WebCore::contextMenuItemTagSmartQuotes):
(WebCore::contextMenuItemTagSmartDashes):
(WebCore::contextMenuItemTagSmartLinks):
(WebCore::contextMenuItemTagTextReplacement):
(WebCore::contextMenuItemTagTransformationsMenu):
(WebCore::contextMenuItemTagMakeUpperCase):
(WebCore::contextMenuItemTagMakeLowerCase):
(WebCore::contextMenuItemTagCapitalize):
(WebCore::contextMenuItemTagChangeBack):
(WebCore::contextMenuItemTagEnterVideoEnhancedFullscreen):
(WebCore::contextMenuItemTagExitVideoEnhancedFullscreen):
(WebCore::AXARIAContentGroupText):
(WebCore::AXHorizontalRuleDescriptionText):
(WebCore::AXMarkText):
(WebCore::AXMeterGaugeRegionOptimumText):
(WebCore::AXMeterGaugeRegionSuboptimalText):
(WebCore::AXMeterGaugeRegionLessGoodText):
(WebCore::builtInPDFPluginName):
(WebCore::pdfDocumentTypeDescription):
(WebCore::postScriptDocumentTypeDescription):
(WebCore::keygenMenuItem2048):
(WebCore::keygenKeychainItemName):
(WebCore::htmlSelectMultipleItems):
(WebCore::fileButtonChooseMediaFileLabel):
(WebCore::fileButtonChooseMultipleMediaFilesLabel):
(WebCore::fileButtonNoMediaFileSelectedLabel):
(WebCore::fileButtonNoMediaFilesSelectedLabel):
(WebCore::insertListTypeNone):
(WebCore::insertListTypeBulleted):
(WebCore::insertListTypeBulletedAccessibilityTitle):
(WebCore::insertListTypeNumbered):
(WebCore::insertListTypeNumberedAccessibilityTitle):
(WebCore::exitFullScreenButtonAccessibilityTitle):
* platform/gtk/LocalizedStringsGtk.cpp:
(WebCore::contextMenuItemTagDelete):
(WebCore::contextMenuItemTagSelectAll):
* platform/win/LocalizedStringsWin.cpp:
(WebCore::contextMenuItemTagSearchWeb):
(WebCore::uploadFileText):
(WebCore::allFilesText):

Source/WebCore/platform/gtk/po:

* CMakeLists.txt: Add keyword parameters for WEB_UI_STRING macros.
* POTFILES.in: Add LocalizedStrings.cpp to the list.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214244 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[GTK] Several WPT tests are failing since they were added in the last update in r213882
carlosgc@webkit.org [Wed, 22 Mar 2017 08:41:44 +0000 (08:41 +0000)]
[GTK] Several WPT tests are failing since they were added in the last update in r213882
https://bugs.webkit.org/show_bug.cgi?id=169915

Reviewed by Michael Catanzaro.

Tools:

Add libsoup patch to not fail loads when there's no empty line after the headers in responses with no body.

* gtk/jhbuild.modules:
* gtk/patches/libsoup-soup-message-io-Do-not-fail-when-there-s-no-empty-li.patch: Added.

LayoutTests:

They fail because libsoup fails to load the requests because the response doesn't include the empty line after the
headers while other browsers allow that. Even with libsoup patched we still need platform specific expectations
for imported/w3c/web-platform-tests/XMLHttpRequest/getallresponseheaders.htm, because libsoup caches all header
names, so the first time foo-test header is cached and then always used no matter what case is used. The results
are still valid in any case.

* platform/gtk/imported/w3c/web-platform-tests/XMLHttpRequest/getallresponseheaders-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214243 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[iOS] Migrate off of CTFontCreateForCSS
mmaxfield@apple.com [Wed, 22 Mar 2017 07:15:16 +0000 (07:15 +0000)]
[iOS] Migrate off of CTFontCreateForCSS
https://bugs.webkit.org/show_bug.cgi?id=168894

Reviewed by Antti Koivisto.

Source/WebCore:

Use the new font selection algorithm introduced in r213163 on iOS.

Test: fast/text/font-weights.html

* platform/graphics/cocoa/FontCacheCoreText.cpp:
* platform/graphics/cocoa/FontCocoa.mm: On iOS only, CTFontCreateForCSS()
will return a 0-sized font if the passed size parameter is 0. However, all
other CT APIs will create a 12-px-sized font if the passed size parameter is
0. Most of our code knows and understands this (because this is how it works
on the Mac). However, this one place needs to be updated to understand this.

LayoutTests:

* platform/ios-simulator/fast/text/font-weights-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214242 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoMake non-decoded image flashing less intrusive
simon.fraser@apple.com [Wed, 22 Mar 2017 03:49:50 +0000 (03:49 +0000)]
Make non-decoded image flashing less intrusive
https://bugs.webkit.org/show_bug.cgi?id=169942

Reviewed by Said Abou-Hallawa.

When compositing borders are enabled, we paint yellow for images that paint but don't yet have a decoded
frame. Give this yellow 50% alpha to make it less intrusive

Also log to the images log channel when we destroy decoded data.

* platform/graphics/BitmapImage.cpp:
(WebCore::BitmapImage::destroyDecodedData):
(WebCore::BitmapImage::draw):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214241 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoThe DFG Integer Check Combining phase should force an OSR exit for CheckInBounds...
mark.lam@apple.com [Wed, 22 Mar 2017 01:55:00 +0000 (01:55 +0000)]
The DFG Integer Check Combining phase should force an OSR exit for CheckInBounds on a negative constant min bound.
https://bugs.webkit.org/show_bug.cgi?id=169933
<rdar://problem/31105125>

Reviewed by Filip Pizlo and Geoffrey Garen.

Also fixed the bit-rotted RangeKey::dump() function.

* dfg/DFGIntegerCheckCombiningPhase.cpp:
(JSC::DFG::IntegerCheckCombiningPhase::handleBlock):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214240 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoLoosen audio playback restrictions when the volume is 0.
mrajca@apple.com [Wed, 22 Mar 2017 01:44:21 +0000 (01:44 +0000)]
Loosen audio playback restrictions when the volume is 0.
https://bugs.webkit.org/show_bug.cgi?id=169897

Reviewed by Eric Carlson.

Source/WebCore:

Added API tests.

* html/MediaElementSession.cpp:
(WebCore::MediaElementSession::playbackPermitted):

Tools:

* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/Tests/WebKit2/autoplay-zero-volume-check.html: Added.
* TestWebKitAPI/Tests/WebKit2Cocoa/WebsitePolicies.mm:
(TEST):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214239 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoEWS fails to parse_bug_id_from_attachment_page after bugzilla upgrade
aakash_jain@apple.com [Wed, 22 Mar 2017 00:45:15 +0000 (00:45 +0000)]
EWS fails to parse_bug_id_from_attachment_page after bugzilla upgrade
https://bugs.webkit.org/show_bug.cgi?id=169937
<rdar://problem/31181859>

Reviewed by Alexey Proskuryakov.

* Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
(Bugzilla._parse_bug_id_from_attachment_page): Since rel tag is no longer present, parsing the bug id from bug_title.
* Scripts/webkitpy/common/net/bugzilla/bugzilla_unittest.py: Fixed unit test.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214238 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoDisable all virtual tables.
beidson@apple.com [Wed, 22 Mar 2017 00:08:40 +0000 (00:08 +0000)]
Disable all virtual tables.
<rdar://problem/31081972> and https://bugs.webkit.org/show_bug.cgi?id=169928
Source/WebCore:

Reviewed by Jer Noble.

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

* Modules/webdatabase/DatabaseAuthorizer.cpp:
(WebCore::DatabaseAuthorizer::createVTable):
(WebCore::DatabaseAuthorizer::dropVTable):

LayoutTests:

Reviewed by Jer Noble.

* storage/websql/test-authorizer-expected.txt:
* storage/websql/test-authorizer.js:
(createStatementsCallback):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214237 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoRemove bogus availability annotations from DOM SPI headers.
andersca@apple.com [Tue, 21 Mar 2017 23:40:26 +0000 (23:40 +0000)]
Remove bogus availability annotations from DOM SPI headers.
rdar://problem/31015675

Rubber-stamped by Dan Bernstein.

Source/WebCore:

* platform/cocoa/WebKitAvailability.h:
Remove __NSi_9876_5 #define.

Source/WebKit/mac:

* DOM/DOMHTMLMediaElement.h:
* DOM/DOMHTMLVideoElement.h:
* DOM/DOMMediaError.h:
* DOM/DOMTextEvent.h:
* DOM/DOMTimeRanges.h:
* DOM/DOMTokenList.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214235 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoews should indicate in logs when it fails to parse_bug_id_from_attachment_page
aakash_jain@apple.com [Tue, 21 Mar 2017 22:33:30 +0000 (22:33 +0000)]
ews should indicate in logs when it fails to parse_bug_id_from_attachment_page
https://bugs.webkit.org/show_bug.cgi?id=169927

Reviewed by  Alexey Proskuryakov.

* Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
(Bugzilla._parse_bug_id_from_attachment_page): Added log statement.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214233 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago2017-03-21 Zalan Bujtas <zalan@apple.com>
zalan@apple.com [Tue, 21 Mar 2017 21:39:01 +0000 (21:39 +0000)]
2017-03-21  Zalan Bujtas  <zalan@apple.com>

        Tear down descendant renderers when <slot>'s display value is set to no "contents".
        https://bugs.webkit.org/show_bug.cgi?id=169921
        <rdar://problem/30336417>

        Reviewed by Antti Koivisto.

        Since "display: contents" does not generate a renderer, when an element's display value is
        changed to something other than "contents", we not only create a renderer but also reparent its descendant
        subtree (e.g from slot's parent to the newly constructed slot renderer). During this reparenting, we
        need to tear down the descendant subtree tree and build it up again to reflect the new rendering context.

        Test: fast/shadow-dom/slot-with-continuation-descendants.html

        * style/RenderTreeUpdater.cpp:
        (WebCore::RenderTreeUpdater::updateElementRenderer):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214232 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoTeach TextIndicator to estimate the background color of the given Range
wenson_hsieh@apple.com [Tue, 21 Mar 2017 21:25:47 +0000 (21:25 +0000)]
Teach TextIndicator to estimate the background color of the given Range
https://bugs.webkit.org/show_bug.cgi?id=169869
<rdar://problem/31127272>

Reviewed by Anders Carlsson and Simon Fraser.

Source/WebCore:

Introduces a simple heuristic to compute a background color that is appropriate to use as a border surrounding
the snapshot of the given Range. This work is only carried out if TextIndicatorOptionComputeEstimatedBackgroundColor
is specified. The details of how this background color is estimated (as well as when to fall back) can be improved
in several ways (one idea is to sample colors along the edge of the snapshot). For the time being, this patch
naively walks up the render tree in search of enclosing parent renderers that have background colors. If any
renderers have a style that is deemed too complex to capture in a single background color, then fall back to the
base document background color; otherwise, the estimated background color is the result of blending these
background colors together.

* page/TextIndicator.cpp:
(WebCore::styleContainsComplexBackground):

Bail out of the background color codepath if the renderer has backdrop filters, has a background image, or uses
a non-normal blend mode.

(WebCore::fallbackBackgroundColorForTextSelection):
(WebCore::estimatedBackgroundColorForRange):
(WebCore::initializeIndicator):
* page/TextIndicator.h:

Source/WebKit/mac:

Plumb the estimated background color for WebKit1 clients through a new property in WebUITextIndicatorData.

* WebView/WebView.mm:
(-[WebUITextIndicatorData initWithImage:textIndicatorData:scale:]):
(-[WebUITextIndicatorData dealloc]):
* WebView/WebViewPrivate.h:

Source/WebKit2:

Send the estimated background color over XPC, and adopt TextIndicatorOptionComputeEstimatedBackgroundColor when
snapshotting after performing an edit data interaction.

* Shared/WebCoreArgumentCoders.cpp:
(IPC::ArgumentCoder<TextIndicatorData>::encode):
(IPC::ArgumentCoder<TextIndicatorData>::decode):
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::didConcludeEditDataInteraction):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214231 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[WK2] Call processDidCrash() right away when terminating a process that exceeded...
cdumez@apple.com [Tue, 21 Mar 2017 21:08:21 +0000 (21:08 +0000)]
[WK2] Call processDidCrash() right away when terminating a process that exceeded background CPU limit
https://bugs.webkit.org/show_bug.cgi?id=169573
<rdar://problem/30960968>

Reviewed by Andreas Kling.

Call processDidCrash() right away when terminating a process that exceeded background CPU limit
and let the client decide when to reload the content.

* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::reattachToWebProcess):
(WebKit::WebPageProxy::dispatchActivityStateChange):
(WebKit::WebPageProxy::terminateProcess):
* UIProcess/WebPageProxy.h:
* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::didExceedBackgroundCPULimit):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214230 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[GTK][Wayland] Enable WebGL, 3D and compositing layout tests
clopez@igalia.com [Tue, 21 Mar 2017 20:52:51 +0000 (20:52 +0000)]
[GTK][Wayland] Enable WebGL, 3D and compositing layout tests
https://bugs.webkit.org/show_bug.cgi?id=71849

Unreviewed test gardening.

* platform/gtk-wayland/TestExpectations: Add the diff of native Wayland compared to Xvfb.
* platform/gtk/TestExpectations: This two tests seem flaky.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214229 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoCrash in WebCore: WebCore::CARingBuffer::getCurrentFrameBounds + 28
jer.noble@apple.com [Tue, 21 Mar 2017 19:45:38 +0000 (19:45 +0000)]
Crash in WebCore: WebCore::CARingBuffer::getCurrentFrameBounds + 28
https://bugs.webkit.org/show_bug.cgi?id=169887
<rdar://problem/23648082>

Reviewed by Eric Carlson.

Because AudioSourceProviderAVFObjC::prepareCallback() can concievably be called after the AudioSourceProviderAVFObjC
it refers to has been destroyed, add an extra layer of indirection between the tap and the provider, and invalidate
that level of indirection before the AudioSourceProviderAVFObjC is destroyed.

* platform/graphics/avfoundation/AudioSourceProviderAVFObjC.h:
* platform/graphics/avfoundation/AudioSourceProviderAVFObjC.mm:
(WebCore::AudioSourceProviderAVFObjC::initCallback):
(WebCore::AudioSourceProviderAVFObjC::finalizeCallback):
(WebCore::AudioSourceProviderAVFObjC::prepareCallback):
(WebCore::AudioSourceProviderAVFObjC::unprepareCallback):
(WebCore::AudioSourceProviderAVFObjC::processCallback):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214227 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoRemove logging left in by mistake.
simon.fraser@apple.com [Tue, 21 Mar 2017 18:18:35 +0000 (18:18 +0000)]
Remove logging left in by mistake.

* rendering/RenderNamedFlowThread.cpp:
(WebCore::RenderNamedFlowThread::willBeDestroyed):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214226 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoCommit local changes for Bugzilla 5.0.3
ddkilzer@apple.com [Tue, 21 Mar 2017 16:54:04 +0000 (16:54 +0000)]
Commit local changes for Bugzilla 5.0.3

* images/favicon.ico: Update favicon from Dec 11, 2015.
* data/params.json: Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214225 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUpgrade to Bugzilla 5.0.3.
ddkilzer@apple.com [Tue, 21 Mar 2017 16:27:49 +0000 (16:27 +0000)]
Upgrade to Bugzilla 5.0.3.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214224 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[GTK] Enable WebGL, 3D and compositing layout tests
clopez@igalia.com [Tue, 21 Mar 2017 15:52:28 +0000 (15:52 +0000)]
[GTK] Enable WebGL, 3D and compositing layout tests
https://bugs.webkit.org/show_bug.cgi?id=71849

Unreviewed test gardeninng.

* http/tests/webgl/1.0.2/origin-clean-conformance-expected.txt: Rebaseline after r208049.
* platform/gtk/TestExpectations: Mark new expected failures.
* platform/gtk/compositing/animation/state-at-end-event-transform-layer-expected.png: Added. Rebaselined.
* platform/gtk/compositing/animation/state-at-end-event-transform-layer-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/backing/backface-visibility-in-3dtransformed-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/backing/child-layer-no-backing-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/backing/inline-block-no-backing-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/backing/no-backing-for-clip-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/backing/no-backing-for-clip-overhang-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/backing/no-backing-for-clip-overlap-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/backing/replaced-child-no-backing-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/color-matching/image-color-matching-expected.png: Added. Rebaselined.
* platform/gtk/compositing/color-matching/image-color-matching-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/color-matching/pdf-image-match-expected.png: Added. Rebaselined.
* platform/gtk/compositing/color-matching/pdf-image-match-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/compositing-visible-descendant-expected.png: Added. Rebaselined.
* platform/gtk/compositing/compositing-visible-descendant-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/direct-image-compositing-expected.png: Added. Rebaselined.
* platform/gtk/compositing/direct-image-compositing-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/generated-content-expected.png: Added. Rebaselined.
* platform/gtk/compositing/generated-content-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/geometry/abs-position-inside-opacity-expected.png: Added. Rebaselined.
* platform/gtk/compositing/geometry/abs-position-inside-opacity-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/geometry/clipping-foreground-expected.png: Added. Rebaselined.
* platform/gtk/compositing/geometry/clipping-foreground-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/geometry/composited-html-size-expected.png: Added. Rebaselined.
* platform/gtk/compositing/geometry/composited-html-size-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/geometry/composited-in-columns-expected.png: Added. Rebaselined.
* platform/gtk/compositing/geometry/composited-in-columns-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/geometry/fixed-in-composited-expected.png: Added. Rebaselined.
* platform/gtk/compositing/geometry/fixed-in-composited-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/geometry/fixed-position-expected.png: Added. Rebaselined.
* platform/gtk/compositing/geometry/fixed-position-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/geometry/horizontal-scroll-composited-expected.png: Added. Rebaselined.
* platform/gtk/compositing/geometry/horizontal-scroll-composited-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/geometry/layer-due-to-layer-children-deep-expected.png: Added. Rebaselined.
* platform/gtk/compositing/geometry/layer-due-to-layer-children-deep-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/geometry/layer-due-to-layer-children-expected.png: Added. Rebaselined.
* platform/gtk/compositing/geometry/layer-due-to-layer-children-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/geometry/outline-change-expected.png: Added. Rebaselined.
* platform/gtk/compositing/geometry/outline-change-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/geometry/partial-layout-update-expected.png: Added. Rebaselined.
* platform/gtk/compositing/geometry/partial-layout-update-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/geometry/root-layer-update-expected.png: Added. Rebaselined.
* platform/gtk/compositing/geometry/root-layer-update-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/geometry/transfrom-origin-on-zero-size-layer-expected.png: Added. Rebaselined.
* platform/gtk/compositing/geometry/transfrom-origin-on-zero-size-layer-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/geometry/vertical-scroll-composited-expected.png: Added. Rebaselined.
* platform/gtk/compositing/geometry/vertical-scroll-composited-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/geometry/video-fixed-scrolling-expected.png: Added. Rebaselined.
* platform/gtk/compositing/geometry/video-fixed-scrolling-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/geometry/video-opacity-overlay-expected.png: Added. Rebaselined.
* platform/gtk/compositing/geometry/video-opacity-overlay-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/iframes/composited-iframe-alignment-expected.png: Added. Rebaselined.
* platform/gtk/compositing/iframes/composited-iframe-alignment-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/iframes/iframe-copy-on-scroll-expected.png: Added. Rebaselined.
* platform/gtk/compositing/iframes/iframe-copy-on-scroll-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/iframes/iframe-in-composited-layer-expected.png: Added. Rebaselined.
* platform/gtk/compositing/iframes/iframe-in-composited-layer-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/iframes/invisible-nested-iframe-show-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/iframes/resizer-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/images/direct-image-background-color-expected.png: Added. Rebaselined.
* platform/gtk/compositing/images/direct-image-background-color-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/layer-creation/fixed-position-scroll-expected.png: Added. Rebaselined.
* platform/gtk/compositing/layer-creation/fixed-position-scroll-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/layer-creation/overlap-animation-container-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/layer-creation/overlap-animation-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/layer-creation/will-change-layer-creation-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/layer-creation/will-change-on-normal-flow-content-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/layers-inside-overflow-scroll-expected.png: Added. Rebaselined.
* platform/gtk/compositing/layers-inside-overflow-scroll-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/masks/direct-image-mask-expected.png: Added. Rebaselined.
* platform/gtk/compositing/masks/direct-image-mask-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/masks/masked-ancestor-expected.png: Added. Rebaselined.
* platform/gtk/compositing/masks/masked-ancestor-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/masks/multiple-masks-expected.png: Added. Rebaselined.
* platform/gtk/compositing/masks/multiple-masks-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/masks/simple-composited-mask-expected.png: Added. Rebaselined.
* platform/gtk/compositing/masks/simple-composited-mask-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/media-controls-bar-appearance-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/overflow/composited-scrolling-creates-a-stacking-container-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/overflow/composited-scrolling-paint-phases-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/overflow/fixed-position-ancestor-clip-expected.png: Added. Rebaselined.
* platform/gtk/compositing/overflow/fixed-position-ancestor-clip-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/overflow/nested-scrolling-expected.png: Added. Rebaselined.
* platform/gtk/compositing/overflow/nested-scrolling-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/overflow/overflow-compositing-descendant-expected.png: Added. Rebaselined.
* platform/gtk/compositing/overflow/overflow-compositing-descendant-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/overflow/overflow-positioning-expected.png: Added. Rebaselined.
* platform/gtk/compositing/overflow/overflow-positioning-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/overflow/overflow-scroll-expected.png: Added. Rebaselined.
* platform/gtk/compositing/overflow/overflow-scroll-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/overflow/parent-overflow-expected.png: Added. Rebaselined.
* platform/gtk/compositing/overflow/parent-overflow-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/overflow/remove-overflow-crash2-expected.png: Added. Rebaselined.
* platform/gtk/compositing/overflow/remove-overflow-crash2-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/overflow/scroll-ancestor-update-expected.png: Added. Rebaselined.
* platform/gtk/compositing/overflow/scroll-ancestor-update-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/overflow/scrollbar-painting-expected.png: Added. Rebaselined.
* platform/gtk/compositing/overflow/scrollbar-painting-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/overflow/theme-affects-visual-overflow-expected.png: Added. Rebaselined.
* platform/gtk/compositing/overflow/theme-affects-visual-overflow-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/reflections/reflection-in-composited-expected.png: Added. Rebaselined.
* platform/gtk/compositing/reflections/reflection-in-composited-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/repaint/become-overlay-composited-layer-expected.png: Added. Rebaselined.
* platform/gtk/compositing/repaint/become-overlay-composited-layer-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/repaint/composited-document-element-expected.png: Added. Rebaselined.
* platform/gtk/compositing/repaint/composited-document-element-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/repaint/content-into-overflow-expected.png: Added. Rebaselined.
* platform/gtk/compositing/repaint/content-into-overflow-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/repaint/layer-repaint-expected.png: Added. Rebaselined.
* platform/gtk/compositing/repaint/layer-repaint-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/repaint/layer-repaint-rects-expected.png: Added. Rebaselined.
* platform/gtk/compositing/repaint/layer-repaint-rects-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/repaint/newly-composited-on-scroll-expected.png: Added. Rebaselined.
* platform/gtk/compositing/repaint/newly-composited-on-scroll-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/repaint/opacity-between-absolute-expected.png: Added. Rebaselined.
* platform/gtk/compositing/repaint/opacity-between-absolute-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/repaint/opacity-between-absolute2-expected.png: Added. Rebaselined.
* platform/gtk/compositing/repaint/opacity-between-absolute2-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/repaint/overflow-into-content-expected.png: Added. Rebaselined.
* platform/gtk/compositing/repaint/overflow-into-content-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/rtl/rtl-iframe-absolute-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/rtl/rtl-iframe-relative-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/rtl/rtl-overflow-scrolling-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/rtl/rtl-scrolling-with-transformed-descendants-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/self-painting-layers-expected.png: Added. Rebaselined.
* platform/gtk/compositing/self-painting-layers-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/shadows/shadow-drawing-expected.png: Added. Rebaselined.
* platform/gtk/compositing/shadows/shadow-drawing-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/sibling-positioning-expected.png: Added. Rebaselined.
* platform/gtk/compositing/sibling-positioning-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/text-on-large-layer-expected.png: Added. Rebaselined.
* platform/gtk/compositing/text-on-large-layer-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/tiling/huge-layer-add-remove-child-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/tiling/visiblerect-accumulated-offset-expected.png: Added. Rebaselined.
* platform/gtk/compositing/tiling/visiblerect-accumulated-offset-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/visibility/layer-visible-content-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/webgl/webgl-background-color-expected.png: Added. Rebaselined.
* platform/gtk/compositing/webgl/webgl-background-color-expected.txt: Added. Rebaselined.
* platform/gtk/compositing/webgl/webgl-nonpremultiplied-blend-expected.png: Added. Rebaselined.
* platform/gtk/compositing/webgl/webgl-nonpremultiplied-blend-expected.txt: Added. Rebaselined.
* platform/gtk/css3/blending/background-blend-mode-default-value-expected.png: Added. Rebaselined.
* platform/gtk/css3/blending/background-blend-mode-default-value-expected.txt: Added. Rebaselined.
* platform/gtk/css3/blending/background-blend-mode-different-image-formats-expected.png: Added. Rebaselined.
* platform/gtk/css3/blending/background-blend-mode-different-image-formats-expected.txt: Added. Rebaselined.
* platform/gtk/css3/blending/background-blend-mode-gif-color-2-expected.png: Added. Rebaselined.
* platform/gtk/css3/blending/background-blend-mode-gif-color-2-expected.txt: Added. Rebaselined.
* platform/gtk/css3/blending/background-blend-mode-gif-color-expected.png: Added. Rebaselined.
* platform/gtk/css3/blending/background-blend-mode-gif-color-expected.txt: Added. Rebaselined.
* platform/gtk/css3/blending/background-blend-mode-image-color-expected.png: Added. Rebaselined.
* platform/gtk/css3/blending/background-blend-mode-image-color-expected.txt: Added. Rebaselined.
* platform/gtk/css3/blending/background-blend-mode-image-image-expected.png: Added. Rebaselined.
* platform/gtk/css3/blending/background-blend-mode-image-image-expected.txt: Added. Rebaselined.
* platform/gtk/css3/blending/background-blend-mode-single-layer-no-blending-expected.png: Added. Rebaselined.
* platform/gtk/css3/blending/background-blend-mode-single-layer-no-blending-expected.txt: Added. Rebaselined.
* platform/gtk/css3/blending/blend-mode-background-expected.png: Added. Rebaselined.
* platform/gtk/css3/blending/blend-mode-background-expected.txt: Added. Rebaselined.
* platform/gtk/css3/blending/blend-mode-isolation-flags-remove-non-stacking-context-blending-expected.png: Added. Rebaselined.
* platform/gtk/css3/blending/blend-mode-isolation-flags-remove-non-stacking-context-blending-expected.txt: Added. Rebaselined.
* platform/gtk/css3/blending/blend-mode-isolation-flags-remove-stacking-context-blending-expected.png: Added. Rebaselined.
* platform/gtk/css3/blending/blend-mode-isolation-flags-remove-stacking-context-blending-expected.txt: Added. Rebaselined.
* platform/gtk/css3/blending/svg-blend-layer-blend-expected.png: Added. Rebaselined.
* platform/gtk/css3/blending/svg-blend-layer-blend-expected.txt: Added. Rebaselined.
* platform/gtk/css3/blending/svg-blend-layer-filter-expected.png: Added. Rebaselined.
* platform/gtk/css3/blending/svg-blend-layer-filter-expected.txt: Added. Rebaselined.
* platform/gtk/css3/blending/svg-blend-layer-mask-expected.png: Added. Rebaselined.
* platform/gtk/css3/blending/svg-blend-layer-mask-expected.txt: Added. Rebaselined.
* platform/gtk/css3/blending/svg-blend-layer-opacity-expected.png: Added. Rebaselined.
* platform/gtk/css3/blending/svg-blend-layer-opacity-expected.txt: Added. Rebaselined.
* platform/gtk/css3/blending/svg-blend-layer-shadow-expected.png: Added. Rebaselined.
* platform/gtk/css3/blending/svg-blend-layer-shadow-expected.txt: Added. Rebaselined.
* platform/gtk/css3/blending/svg-blend-multiply-alpha-expected.png: Added. Rebaselined.
* platform/gtk/css3/blending/svg-blend-multiply-alpha-expected.txt: Added. Rebaselined.
* platform/gtk/css3/blending/svg-blend-normal-expected.png: Added. Rebaselined.
* platform/gtk/css3/blending/svg-blend-normal-expected.txt: Added. Rebaselined.
* platform/gtk/css3/blending/svg-blend-plus-darker-expected.png: Added. Rebaselined.
* platform/gtk/css3/blending/svg-blend-plus-darker-expected.txt: Added. Rebaselined.
* platform/gtk/css3/blending/svg-blend-plus-lighter-expected.png: Added. Rebaselined.
* platform/gtk/css3/blending/svg-blend-plus-lighter-expected.txt: Added. Rebaselined.
* platform/gtk/fast/canvas/webgl/css-webkit-canvas-expected.png: Added. Rebaselined.
* platform/gtk/fast/canvas/webgl/css-webkit-canvas-expected.txt: Added. Rebaselined.
* platform/gtk/fast/canvas/webgl/css-webkit-canvas-repaint-expected.png: Added. Rebaselined.
* platform/gtk/fast/canvas/webgl/css-webkit-canvas-repaint-expected.txt: Added. Rebaselined.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214223 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[Planet WebKit] Use HTTPS instead of HTTP for Igalia blog
rego@igalia.com [Tue, 21 Mar 2017 14:34:36 +0000 (14:34 +0000)]
[Planet WebKit] Use HTTPS instead of HTTP for Igalia blog

Unreviewed patch just updating the Igalia WebKit blog configuration.

* config.ini:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214222 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUnreviewed GTK+ gardening. Remove expectations for imported/w3c/web-platform-tests...
carlosgc@webkit.org [Tue, 21 Mar 2017 14:12:09 +0000 (14:12 +0000)]
Unreviewed GTK+ gardening. Remove expectations for imported/w3c/web-platform-tests/XMLHttpRequest/setrequestheader-content-type.htm.

It no longer works with current version of wptserve and it's already skipped globally.

* platform/gtk/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214221 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[ARM] Add missing MacroAssembler functions after r214187
ossy@webkit.org [Tue, 21 Mar 2017 11:42:36 +0000 (11:42 +0000)]
[ARM] Add missing MacroAssembler functions after r214187
https://bugs.webkit.org/show_bug.cgi?id=169912

Reviewed by Yusuke Suzuki.

* assembler/MacroAssemblerARM.h:
(JSC::MacroAssemblerARM::loadFloat):
(JSC::MacroAssemblerARM::storeFloat):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214220 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[JSC] Optimize Number.prototype.toString on Int32 / Int52 / Double
utatane.tea@gmail.com [Tue, 21 Mar 2017 11:31:43 +0000 (11:31 +0000)]
[JSC] Optimize Number.prototype.toString on Int32 / Int52 / Double
https://bugs.webkit.org/show_bug.cgi?id=167454

Reviewed by Saam Barati.

JSTests:

* stress/number-to-string-abstract-operation.js: Added.
(shouldBe):
(int32ToString):
(shouldBe.int32ToString.new.Number.int52ToString):
(shouldBe.int32ToString.new.Number):
(shouldBe.doubleToString):
* stress/number-to-string-radix.js: Added.
(shouldBe):
(int32ToString):
(shouldBe.int32ToString.new.Number.int52ToString):
(shouldBe.int32ToString.new.Number):
(shouldBe.doubleToString):
* stress/number-to-string.js: Added.
(shouldBe):
(int32ToString):
(shouldBe.int32ToString.new.Number.int52ToString):
(shouldBe.int32ToString.new.Number):
(shouldBe.doubleToString):

Source/JavaScriptCore:

This patch improves Number.toString(radix) performance
by introducing NumberToStringWithRadix DFG node. It directly
calls the operation and it always returns String.

                                               baseline                  patched

    stanford-crypto-sha256-iterative        45.130+-0.928             44.032+-1.184           might be 1.0250x faster

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214219 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[JSC] Add JSPromiseDeferred::reject(ExecState*, Exception*) interface
utatane.tea@gmail.com [Tue, 21 Mar 2017 11:16:47 +0000 (11:16 +0000)]
[JSC] Add JSPromiseDeferred::reject(ExecState*, Exception*) interface
https://bugs.webkit.org/show_bug.cgi?id=169908

Reviewed by Sam Weinig.

To avoid calling reject(ExecState*, JSValue) with Exception* accidentally,
we add a new interface reject(ExecState*, Exception*).
Such an interface is already added in DOMPromise in WebCore.

* runtime/JSInternalPromiseDeferred.cpp:
(JSC::JSInternalPromiseDeferred::reject):
* runtime/JSInternalPromiseDeferred.h:
* runtime/JSPromiseDeferred.cpp:
(JSC::JSPromiseDeferred::reject):
* runtime/JSPromiseDeferred.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@214218 268f45cc-cd09-0410-ab3c-d52691b4dbfc