WebKit-https.git
3 years ago[CMake] Remove unused find modules
mcatanzaro@igalia.com [Wed, 9 Aug 2017 23:32:14 +0000 (23:32 +0000)]
[CMake] Remove unused find modules
https://bugs.webkit.org/show_bug.cgi?id=175363

Reviewed by Carlos Alberto Lopez Perez.

* Source/cmake/FindCFLite.cmake: Removed.
* Source/cmake/FindClang.cmake: Removed.
* Source/cmake/FindDBus.cmake: Removed.
* Source/cmake/FindDirectX.cmake: Removed.
* Source/cmake/FindEspeak.cmake: Removed.
* Source/cmake/FindQuickTimeSDK.cmake: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220498 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[Beacon][NetworkSession] Support CORS-preflighting on redirects
cdumez@apple.com [Wed, 9 Aug 2017 23:23:58 +0000 (23:23 +0000)]
[Beacon][NetworkSession] Support CORS-preflighting on redirects
https://bugs.webkit.org/show_bug.cgi?id=175386
<rdar://problem/33801370>

Reviewed by Youenn Fablet.

Source/WebCore:

Export a couple of WebCore symbols so I can use them in WebKit2.

Tests: http/wpt/beacon/cors/cors-preflight-redirect-failure.html
       http/wpt/beacon/cors/cors-preflight-redirect-from-crossorigin-to-sameorigin.html
       http/wpt/beacon/cors/cors-preflight-redirect-success.html

* loader/CrossOriginAccessControl.h:
* page/SecurityOrigin.h:

Source/WebKit:

Add support to Beacon for doing CORS-preflighting upon redirect to a different
domain.

* NetworkProcess/NetworkCORSPreflightChecker.h:
* NetworkProcess/PingLoad.cpp:
(WebKit::PingLoad::PingLoad):
(WebKit::PingLoad::~PingLoad):
(WebKit::PingLoad::loadRequest):
(WebKit::PingLoad::securityOrigin const):
(WebKit::PingLoad::willPerformHTTPRedirection):
(WebKit::PingLoad::didReceiveResponseNetworkSession):
(WebKit::PingLoad::needsCORSPreflight const):
(WebKit::PingLoad::doCORSPreflight):
* NetworkProcess/PingLoad.h:

LayoutTests:

Add layout test coverage.

* http/wpt/beacon/cors/cors-preflight-redirect-failure-expected.txt: Added.
* http/wpt/beacon/cors/cors-preflight-redirect-failure.html: Added.
* http/wpt/beacon/cors/cors-preflight-redirect-from-crossorigin-to-sameorigin-expected.txt: Added.
* http/wpt/beacon/cors/cors-preflight-redirect-from-crossorigin-to-sameorigin.html: Added.
* http/wpt/beacon/cors/cors-preflight-redirect-success-expected.txt: Added.
* http/wpt/beacon/cors/cors-preflight-redirect-success.html: Added.
* http/wpt/beacon/resources/beacon-preflight.py:
(main):
* http/wpt/beacon/resources/redirect.py: Added.
(main):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220497 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUse MPAVRoutingController instead of deprecated versions.
commit-queue@webkit.org [Wed, 9 Aug 2017 23:18:13 +0000 (23:18 +0000)]
Use MPAVRoutingController instead of deprecated versions.
https://bugs.webkit.org/show_bug.cgi?id=175063
Source/WebCore:

Patch by Jeremy Jones <jeremyj@apple.com> on 2017-08-09
Reviewed by Tim Horton.

No new tests because no behavior change. This uses a different platform class to present
an interface.

Remove deprecated MPAudioVideoRoutingPopoverController and MPAVRoutingSheet
Add MPMediaControlsViewController.

* platform/spi/ios/MediaPlayerSPI.h:

Source/WebKit:

rdar://problem/33301230

Patch by Jeremy Jones <jeremyj@apple.com> on 2017-08-09
Reviewed by Tim Horton.

Remove dependence on deprecated classes MPAVRoutingSheet and MPAudioVideoRoutingPopoverController.

* UIProcess/ios/forms/WKAirPlayRoutePicker.h:
* UIProcess/ios/forms/WKAirPlayRoutePicker.mm:
(-[WKAirPlayRoutePicker dealloc]):
(-[WKAirPlayRoutePicker show:fromRect:]):
(-[WKAirPlayRoutePicker popoverControllerDidDismissPopover:]): Deleted.
(-[WKAirPlayRoutePicker _presentAirPlayPopoverAnimated:fromRect:]): Deleted.
(-[WKAirPlayRoutePicker _windowWillRotate:]): Deleted.
(-[WKAirPlayRoutePicker _windowDidRotate:]): Deleted.
(-[WKAirPlayRoutePicker _dismissAirPlayRoutePickerIPad]): Deleted.
(-[WKAirPlayRoutePicker showAirPlayPickerIPad:fromRect:]): Deleted.
(-[WKAirPlayRoutePicker showAirPlayPickerIPhone:]): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220496 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUnreviewed, deflake http/wpt/beacon/keepalive-after-navigation.html
cdumez@apple.com [Wed, 9 Aug 2017 23:03:10 +0000 (23:03 +0000)]
Unreviewed, deflake http/wpt/beacon/keepalive-after-navigation.html

Temporarily restore previous timeout on this test to address flakiness on the
bots until I can rewrite it.

* http/wpt/beacon/keepalive-after-navigation.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220495 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoImport beacon/headers/header-content-type.html from upstream WPT
cdumez@apple.com [Wed, 9 Aug 2017 23:01:05 +0000 (23:01 +0000)]
Import beacon/headers/header-content-type.html from upstream WPT
https://bugs.webkit.org/show_bug.cgi?id=175395

Reviewed by Sam Weinig.

LayoutTests/imported/w3c:

* web-platform-tests/beacon/headers/header-content-type-expected.txt:
* web-platform-tests/beacon/headers/header-content-type.html:

LayoutTests:

Drop local test that has been upstreamed to WPT.

* http/wpt/beacon/headers/header-content-type-same-origin-expected.txt: Removed.
* http/wpt/beacon/headers/header-content-type-same-origin.html: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220494 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUnreviewed build fix.
msaboff@apple.com [Wed, 9 Aug 2017 22:59:32 +0000 (22:59 +0000)]
Unreviewed build fix.

Added missing files.

* RexBench/Octane2/benchmark.js: Added.
(Octane2RegExpBenchmark):
(Octane2RegExpBenchmark.prototype.setup.):
(Octane2RegExpBenchmark.prototype.setup.Math.random):
(Octane2RegExpBenchmark.prototype.setup):
(Octane2RegExpBenchmark.prototype.runOnce):
(Octane2RegExpBenchmark.prototype.tearDown):
* RexBench/SunSpider/benchmark.js: Added.
(RegexDNABenchmark):
(RegexDNABenchmark.prototype.runOnce):
(RegexDNABenchmark.prototype.validate):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220493 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoMark font-display as in development
mmaxfield@apple.com [Wed, 9 Aug 2017 22:42:59 +0000 (22:42 +0000)]
Mark font-display as in development
https://bugs.webkit.org/show_bug.cgi?id=175387

Unreviewed.

* features.json:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220488 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUpdate for https://bugs.webkit.org/show_bug.cgi?id=175244.
weinig@apple.com [Wed, 9 Aug 2017 22:40:53 +0000 (22:40 +0000)]
Update for https://bugs.webkit.org/show_bug.cgi?id=175244.

Remove one addition of null Function test.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220487 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoLayoutTest accessibility/presentation-role-iframe.html is a flaky failure
n_wang@apple.com [Wed, 9 Aug 2017 22:36:04 +0000 (22:36 +0000)]
LayoutTest accessibility/presentation-role-iframe.html is a flaky failure
https://bugs.webkit.org/show_bug.cgi?id=175194

Reviewed by Chris Fleizach.

In the test we have no way to control which iframe will be loaded first.
Output the logs at the end of the test so that they are in order.

* accessibility/presentation-role-iframe.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220486 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoRex Bench doesn't work from the command line
msaboff@apple.com [Wed, 9 Aug 2017 22:33:27 +0000 (22:33 +0000)]
Rex Bench doesn't work from the command line
https://bugs.webkit.org/show_bug.cgi?id=175349

Reviewed by Saam Barati.

Fixed cli.js to have the same set of benchmarks and .js files as the web
page driver.

Refactored the way benchmarks are run using a new Benchmark base class.  That
class will run the benchmark as well.  Eliminated the various runBenchmark()
functions that were used for individual tests.  This refactoring was threaded
into the web page driver as well.  Changed the iteration counts on many of the
individual benchmarks.  Also changed the inner loop count for the flight
planner test.

Eliminated Octane2/base.js as it isn't needed to run Octane2/RegExp with the
new test harness.  Eliminated other dead code in Octane2/regexp.js.

* RexBench/Basic/benchmark.js:
(BasicBenchmark):
(BasicBenchmark.prototype.runIteration): Deleted.
(runBenchmark): Deleted.
* RexBench/FlightPlanner/benchmark.js:
(FlightPlannerBenchmark):
(FlightPlannerBenchmark.prototype.runOnce):
(this.performance.performance.now.currentTime): Deleted.
(else.this.preciseTime.currentTime): Deleted.
(else.currentTime): Deleted.
(FlightPlannerBenchmark.prototype.runIteration): Deleted.
(FlightPlannerBenchmark.prototype.checkResults): Deleted.
(runBenchmark): Deleted.
* RexBench/Octane2/base.js: Removed.
* RexBench/Octane2/regexp.js:
* RexBench/basic_benchmark.js:
* RexBench/benchmark.js: Added.
(this.performance.performance.now.currentTime):
(else.this.preciseTime.currentTime):
(else.currentTime):
(Benchmark):
(Benchmark.prototype.runIterations):
(Benchmark.prototype.setup):
(Benchmark.prototype.runOnce):
(Benchmark.prototype.validate):
(Benchmark.prototype.tearDown):
* RexBench/cli.js:
* RexBench/flightplan_benchmark.js:
* RexBench/flightplan_unicode_benchmark.js:
* RexBench/octane2_benchmark.js:
* RexBench/stats.js:
(Stats.prototype.toString.return.span.0.span.span):
* RexBench/sunspider_benchmark.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220485 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUnreviewed, rolling out r219334.
commit-queue@webkit.org [Wed, 9 Aug 2017 22:08:35 +0000 (22:08 +0000)]
Unreviewed, rolling out r219334.
https://bugs.webkit.org/show_bug.cgi?id=175398

Caused bug 175023 (Requested by ap on #webkit).

Reverted changeset:

"[SVG] Leak in SVGAnimatedListPropertyTearOff"
https://bugs.webkit.org/show_bug.cgi?id=172545
http://trac.webkit.org/changeset/219334

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220484 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoAllow nested timeouts in webkitpy
jbedard@apple.com [Wed, 9 Aug 2017 21:58:18 +0000 (21:58 +0000)]
Allow nested timeouts in webkitpy
https://bugs.webkit.org/show_bug.cgi?id=175390
<rdar://problem/33803003>

Reviewed by David Kilzer.

We need to be able to nest timeouts in webkitpy. In particular, we have a few cases where functions
which use timeouts also call the executive. For on-device testing, we need to have timeouts inside
the executive to detect and recover from any issues connecting with devices.

* Scripts/webkitpy/benchmark_runner/utils.py:
(TimeoutError): Deleted.
(timeout): Deleted.
* Scripts/webkitpy/common/timeout_context.py: Added.
(Timeout): A timeout context designed to be nested.
(Timeout.TimeoutData): The data required to construct an alarm for a given timeout.
(Timeout.TimeoutData.__init__):
(Timeout.default_handler): Timeout handler used if none is specified.
(Timeout.current): Access data about the most urgent timeout.
(Timeout.__init__): Construct a Timeout object with seconds and an optional handler.
(Timeout._bind_timeout_data_to_alarm): Given data about a timeout, initialize an alarm for that timeout.
(Timeout.__enter__): Un-bind all alarms. Add data for this timeout to the ordered list and bind the most
urgent timeout data.
(Timeout.__exit__): Un-bind all alarms. Remove data for this timeout from the ordered list and bind the
most urgent timeout data, if such data exists.
* Scripts/webkitpy/common/timeout_context_unittest.py: Added.
(TimeoutContextTests):
(TimeoutContextTests.test_current_timeout): Test that accessing the nearest timeout works as expected.
(TimeoutContextTests.test_invalid_timeout): Test a timeout of 0.
(TimeoutContextTests.test_timeout_data): Confirm that timeouts are constructed correctly.
(TimeoutContextTests.test_nested_inner_precedence): Check that a more urgent inner timeout takes precedence
over a less urgent outer timeout.
(TimeoutContextTests.test_nested_outer_precedence): Check that a more urgent outer timeout takes precedence
over a less urgent inner timeout.
(TimeoutContextTests.test_no_timeout): Test a block of code without a timeout.
(TimeoutContextTests.test_basic_timeout): Test a block of code expected to timeout.
(TimeoutContextTests.test_exception_constructor_timeout): Test a timeout where the handler is an exception.
(TimeoutContextTests.test_nested_inner_timeout): Confirm that a more urgent inner timeout is triggered.
(TimeoutContextTests.test_nested_outer_timeout): Confirm that a more urgent outer timeout is triggered.
* Scripts/webkitpy/port/simulator_process.py:
(SimulatorProcess._start): Use Timeout class.
* Scripts/webkitpy/xcode/simulated_device.py:
(SimulatedDevice.launch_app._log_debug_error): Use Timeout class.
(SimulatedDevice.launch_app): Ditto.
(SimulatedDevice.launch_app._install_timeout): Deleted.
* Scripts/webkitpy/xcode/simulator.py:
(Simulator.wait_until_device_is_booted):Use Timeout class.
(Simulator.wait_until_device_is_in_state): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220483 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoEarly error on ANY operator before new.target
caitp@igalia.com [Wed, 9 Aug 2017 21:34:20 +0000 (21:34 +0000)]
Early error on ANY operator before new.target
https://bugs.webkit.org/show_bug.cgi?id=157970

Reviewed by Saam Barati.

Instead of throwing if any unary operator precedes new.target, only
throw if the unary operator updates the reference.

The following become legal in JSC:

```
!new.target
~new.target
typeof new.target
delete new.target
void new.target
```

All of which are legal in v8 and SpiderMonkey in strict and sloppy mode

JSTests:

* stress/new-target-syntax-errors.js:
* stress/new-target.js:

Source/JavaScriptCore:

* parser/Parser.cpp:
(JSC::Parser<LexerType>::parseUnaryExpression):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220481 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[Repaint from Layout Removal] Move layer repaint rects into a map
hyatt@apple.com [Wed, 9 Aug 2017 21:03:03 +0000 (21:03 +0000)]
[Repaint from Layout Removal] Move layer repaint rects into a map
https://bugs.webkit.org/show_bug.cgi?id=175393

Reviewed by Zalan Bujtas.

Move the two repaint rects held by RenderLayer into a RenderLayerModelObject -> RepaintLayoutRects hash map.
RepaintLayoutRects is a new struct that holds both rects.

Eventually more objects than just self painting layers will be caching repaint rects, so this takes a first
step towards having a common cache for these repaint rects. In addition this change saves memory, since
layers that aren't self-painting no longer have empty repaint rects taking up space in RenderLayer.

* rendering/RenderBlockLineLayout.cpp:
(WebCore::RenderBlockFlow::layoutRunsAndFloats):
* rendering/RenderLayer.cpp:
(WebCore::RenderLayer::RenderLayer):
(WebCore::RenderLayer::updateLayerPositions):
(WebCore::RenderLayer::repaintRectIncludingNonCompositingDescendants):
(WebCore::RenderLayer::computeRepaintRects):
(WebCore::RenderLayer::clearRepaintRects):
(WebCore::RenderLayer::updateLayerPositionsAfterScroll):
(WebCore::RenderLayer::scrollTo):
* rendering/RenderLayer.h:
* rendering/RenderLayerModelObject.cpp:
(WebCore::RepaintLayoutRects::RepaintLayoutRects):
(WebCore::RenderLayerModelObject::willBeDestroyed):
(WebCore::RenderLayerModelObject::destroyLayer):
(WebCore::RenderLayerModelObject::styleDidChange):
(WebCore::RenderLayerModelObject::hasRepaintLayoutRects):
(WebCore::RenderLayerModelObject::setRepaintLayoutRects):
(WebCore::RenderLayerModelObject::clearRepaintLayoutRects):
(WebCore::RenderLayerModelObject::repaintLayoutRects):
(WebCore::RenderLayerModelObject::computeRepaintLayoutRects):
* rendering/RenderLayerModelObject.h:
(WebCore::RepaintLayoutRects::RepaintLayoutRects):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220479 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUpdate reviewer
weinig@apple.com [Wed, 9 Aug 2017 21:02:23 +0000 (21:02 +0000)]
Update reviewer

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220478 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoWTF::Function does not allow for reference / non-default constructible return types
commit-queue@webkit.org [Wed, 9 Aug 2017 20:57:51 +0000 (20:57 +0000)]
WTF::Function does not allow for reference / non-default constructible return types
https://bugs.webkit.org/show_bug.cgi?id=175244

Patch by Sam Weinig <sam@webkit.org> on 2017-08-09
Reviewed by Sam Weinig.

Source/JavaScriptCore:

* runtime/ArrayBuffer.cpp:
(JSC::ArrayBufferContents::transferTo):
Call reset(), rather than clear() to avoid the call to destroy() in clear(). The
destroy call needed to be a no-op anyway, since the data is being moved.

Source/WebCore:

* bindings/js/JSCustomElementInterface.h:
(WebCore::JSCustomElementInterface::invokeCallback):
Update the default value for the addArguments parameter to be an empty lambda, rather than
default initialization, which leads to a null WTF::Function. This allows us to remove support
for calling null WTF::Function. No change in behavior.

Source/WebKit:

* UIProcess/WebResourceLoadStatisticsStore.h:
Update the default value for the updateCookiePartitioningForDomainsHandler parameter to be an
empty lambda, rather than default initialization, which leads to a null WTF::Function. This allows
us to remove support for calling null WTF::Function. No change in behavior.

Source/WTF:

When Function, then NoncopyableFunction, was templatized to allow non-void return values
in r201493, it maintained the behavior of being callable even if the Function was null.
To accomplish this, when null, the default construction of the return parameter was used.
This means Function can't be used with return types that are not default constructible,
such as reference types and Ref.

This behavior of returning something when null is surprising, as this is not how normal
functions behave, and not very useful. Instead, we now assert that the function is not
null when being called.

* wtf/Function.h:
(WTF::Function operator(...)):
Instead of allowing a null callable wrapper by returning the default construction of
the return type, assert that the wrapper is there when calling a Function.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220477 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoImport CSS-UI-3 Web Platform Tests
dbates@webkit.org [Wed, 9 Aug 2017 20:28:01 +0000 (20:28 +0000)]
Import CSS-UI-3 Web Platform Tests
https://bugs.webkit.org/show_bug.cgi?id=175294

Reviewed by Zalan Bujtas.

LayoutTests/imported/w3c:

* resources/import-expectations.json:
* resources/resource-files.json:
* web-platform-tests/css/css-ui-3/OWNERS: Added.
* web-platform-tests/css/css-ui-3/box-sizing-001-expected.html: Added.
* web-platform-tests/css/css-ui-3/box-sizing-001.html: Added.
* web-platform-tests/css/css-ui-3/box-sizing-003-expected.html: Added.
* web-platform-tests/css/css-ui-3/box-sizing-003.html: Added.
* web-platform-tests/css/css-ui-3/box-sizing-005-expected.html: Added.
* web-platform-tests/css/css-ui-3/box-sizing-005.html: Added.
* web-platform-tests/css/css-ui-3/box-sizing-007-expected.html: Added.
* web-platform-tests/css/css-ui-3/box-sizing-007.html: Added.
[truncated]

LayoutTests:

Mark tests that we currently failure.

* TestExpectations:
* platform/ios/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220476 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoTeach ScriptExecutionContexts about their SessionID.
beidson@apple.com [Wed, 9 Aug 2017 20:25:57 +0000 (20:25 +0000)]
Teach ScriptExecutionContexts about their SessionID.
https://bugs.webkit.org/show_bug.cgi?id=175391

Reviewed by Andy Estes.

No new tests (No current change in behavior).

Turns out that Workers need to know their SessionID for future feature development.
So let's teach it to them.

* dom/ScriptExecutionContext.h: Expose a pure virtual SessionID accessor.

* dom/Document.cpp:
(WebCore::Document::sessionID const):
* dom/Document.h:

* page/SessionID.cpp:
(WebCore::SessionID::isolatedCopy const):
* page/SessionID.h:

* workers/DedicatedWorkerGlobalScope.cpp:
(WebCore::DedicatedWorkerGlobalScope::create):
(WebCore::DedicatedWorkerGlobalScope::DedicatedWorkerGlobalScope):
* workers/DedicatedWorkerGlobalScope.h:

* workers/DedicatedWorkerThread.cpp:
(WebCore::DedicatedWorkerThread::DedicatedWorkerThread):
(WebCore::DedicatedWorkerThread::createWorkerGlobalScope):
* workers/DedicatedWorkerThread.h:

* workers/Worker.cpp:
(WebCore::Worker::notifyFinished):

* workers/WorkerGlobalScope.cpp:
(WebCore::WorkerGlobalScope::WorkerGlobalScope):
* workers/WorkerGlobalScope.h:

* workers/WorkerGlobalScopeProxy.h:

* workers/WorkerMessagingProxy.cpp:
(WebCore::WorkerMessagingProxy::startWorkerGlobalScope):
* workers/WorkerMessagingProxy.h:

* workers/WorkerThread.cpp:
(WebCore::WorkerThreadStartupData::WorkerThreadStartupData):
(WebCore::WorkerThread::WorkerThread):
(WebCore::WorkerThread::workerThread):
* workers/WorkerThread.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220475 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoMarked http/tests/media/video-buffered-range-contains-currentTime.html as flaky on...
jlewis3@apple.com [Wed, 9 Aug 2017 20:08:01 +0000 (20:08 +0000)]
Marked http/tests/media/video-buffered-range-contains-currentTime.html as flaky on High Sierra
<rdar://problem/33555759>

Unreviewed test gardening.

* platform/mac-highsierra-wk2/TestExpectations:
* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220474 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[iOS DnD] ENABLE_DRAG_SUPPORT should be turned off for iOS 10 and enabled by default
wenson_hsieh@apple.com [Wed, 9 Aug 2017 19:15:37 +0000 (19:15 +0000)]
[iOS DnD] ENABLE_DRAG_SUPPORT should be turned off for iOS 10 and enabled by default
https://bugs.webkit.org/show_bug.cgi?id=175392
<rdar://problem/33783207>

Reviewed by Tim Horton and Megan Gardner.

Tweak FeatureDefines to enable drag and drop by default, and disable only on unsupported platforms (i.e. iOS 10).
Source/JavaScriptCore:

* Configurations/FeatureDefines.xcconfig:

Source/WebCore:

No change in behavior.

* Configurations/FeatureDefines.xcconfig:

Source/WebCore/PAL:

* Configurations/FeatureDefines.xcconfig:

Source/WebKit:

* Configurations/FeatureDefines.xcconfig:

Source/WebKitLegacy/mac:

* Configurations/FeatureDefines.xcconfig:

Tools:

* TestWebKitAPI/Configurations/FeatureDefines.xcconfig:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220473 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoOld subtitle track is not deleted on 'src' attribute change event
commit-queue@webkit.org [Wed, 9 Aug 2017 19:13:51 +0000 (19:13 +0000)]
Old subtitle track is not deleted on 'src' attribute change event
https://bugs.webkit.org/show_bug.cgi?id=175031

Patch by Kirill Ovchinnikov <kirill.ovchinnikov@lge.com> on 2017-08-09
Reviewed by Eric Carlson.

Source/WebCore:

Test: media/track/text-track-src-change.html

* html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::textTrackReadyStateChanged):
* html/HTMLTrackElement.cpp:
(WebCore::HTMLTrackElement::loadTimerFired):
* html/shadow/MediaControls.cpp:
(WebCore::MediaControls::clearTextDisplayContainer):
* html/shadow/MediaControls.h:

LayoutTests:

* media/track/text-track-src-change-expected.txt: Added.
* media/track/text-track-src-change.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220472 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoMake JSC_validateExceptionChecks=1 succeed on JSTests/stress/v8-deltablue-strict.js.
commit-queue@webkit.org [Wed, 9 Aug 2017 18:31:24 +0000 (18:31 +0000)]
Make JSC_validateExceptionChecks=1 succeed on JSTests/stress/v8-deltablue-strict.js.
https://bugs.webkit.org/show_bug.cgi?id=175358

Patch by Robin Morisset <rmorisset@apple.com> on 2017-08-09
Reviewed by Mark Lam.

* jit/JITOperations.cpp:
* runtime/JSObjectInlines.h:
(JSC::JSObject::putInlineForJSObject):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220471 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoServiceWorker experimental feature should not be on by default (definitely not ready).
beidson@apple.com [Wed, 9 Aug 2017 18:05:24 +0000 (18:05 +0000)]
ServiceWorker experimental feature should not be on by default (definitely not ready).
https://bugs.webkit.org/show_bug.cgi?id=175389

Rubberstamped by Andy Estes.

* Shared/WebPreferencesDefinitions.h: false, not DEFAULT_EXPERIMENTAL_FEATURES_ENABLED

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220467 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUpdate for https://bugs.webkit.org/show_bug.cgi?id=175244.
weinig@apple.com [Wed, 9 Aug 2017 18:03:11 +0000 (18:03 +0000)]
Update for https://bugs.webkit.org/show_bug.cgi?id=175244.

* TestWebKitAPI/Tests/WTF/Function.cpp:
Remove tests of null Functions, as they are no longer supported.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220466 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUnreviewed, rolling out r220457.
ryanhaddad@apple.com [Wed, 9 Aug 2017 17:58:12 +0000 (17:58 +0000)]
Unreviewed, rolling out r220457.

This change introduced API test failures.

Reverted changeset:

"WTF::Function does not allow for reference / non-default
constructible return types"
https://bugs.webkit.org/show_bug.cgi?id=175244
http://trac.webkit.org/changeset/220457

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220465 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[WinCairo] Fix build after rev 220399
don.olmstead@sony.com [Wed, 9 Aug 2017 17:53:09 +0000 (17:53 +0000)]
[WinCairo] Fix build after rev 220399
https://bugs.webkit.org/show_bug.cgi?id=175360

Reviewed by Žan Doberšek.

No new tests. No change in behavior.

* platform/graphics/texmap/TextureMapperGLHeaders.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220464 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoAX: crash at WebCore::AccessibilityObject::supportsARIALiveRegion() const + 24
n_wang@apple.com [Wed, 9 Aug 2017 17:52:36 +0000 (17:52 +0000)]
AX: crash at WebCore::AccessibilityObject::supportsARIALiveRegion() const + 24
https://bugs.webkit.org/show_bug.cgi?id=175340
<rdar://problem/33782159>

Reviewed by Chris Fleizach.

Source/WebCore:

When adding a psuedo element child to a RenderBlockFlow element, there might be a chance where
the element has already been layed out but we are still holding onto its stale children. Fixed it
by notifying AX correctly when inserting/removing children during layout.

Test: accessibility/add-children-pseudo-element.html

* rendering/RenderBlockFlow.cpp:
(WebCore::RenderBlockFlow::insertFloatingObject):
(WebCore::RenderBlockFlow::removeFloatingObject):

LayoutTests:

* accessibility/add-children-pseudo-element-expected.txt: Added.
* accessibility/add-children-pseudo-element.html: Added.
* accessibility/resources/svg-circle.svg: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220463 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoWeb Inspector: Styles: Add pre-populated data to spreadsheet-style view
nvasilyev@apple.com [Wed, 9 Aug 2017 17:46:27 +0000 (17:46 +0000)]
Web Inspector: Styles: Add pre-populated data to spreadsheet-style view
https://bugs.webkit.org/show_bug.cgi?id=174838
<rdar://problem/33523220>

Reviewed by Matt Baker.

This patch adds static content to the experimental RulesStyleSpreadsheetDetailsPanel.
The static content will be replaced with actual data in the follow up patches.

* UserInterface/Main.html:
* UserInterface/Views/RulesStyleSpreadsheetDetailsPanel.css: Added.
(.style-spreadsheet .style-rule):
(.style-spreadsheet .selector-line):
(.style-spreadsheet .selector):
(.style-spreadsheet :matches(.selector .matched, .selector:focus)):
(.style-spreadsheet .name):
(.style-spreadsheet .value):
(.style-spreadsheet :matches(.selector, .name, .value)):
(.style-spreadsheet :matches(.selector, .name, .value):focus):
(.style-spreadsheet .selector-inline):
(.style-spreadsheet .property):
(.style-spreadsheet .property-disabled,):
(.style-spreadsheet .property-disabled :matches(.name, .value)):
(.style-spreadsheet .styles-source):
(.style-spreadsheet :matches(a, .node-link)):
(.style-spreadsheet :matches(a, .node-link):hover):
(.style-spreadsheet .inherited):
(.style-spreadsheet .declarations):
(.style-spreadsheet input[type="checkbox"]):
(.style-spreadsheet input[type="checkbox"]:not(:checked)):
(.style-spreadsheet .declarations:hover input[type="checkbox"]):
* UserInterface/Views/RulesStyleSpreadsheetDetailsPanel.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220462 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[GStreamer][MSE] Add missing lock around getStreamByTrackId
commit-queue@webkit.org [Wed, 9 Aug 2017 17:23:16 +0000 (17:23 +0000)]
[GStreamer][MSE] Add missing lock around getStreamByTrackId
https://bugs.webkit.org/show_bug.cgi?id=175323

Patch by Charlie Turner <cturner@igalia.com> on 2017-08-09
Reviewed by Xabier Rodriguez-Calvar.

Covered by existing tests.

* platform/graphics/gstreamer/mse/PlaybackPipeline.cpp:
(WebCore::PlaybackPipeline::enqueueSample): Ensure the media source
lock is taken before modifiying the track queues indirectly via the
iterator use in getTrackId.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220461 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoWPE-focused test gardening.
commit-queue@webkit.org [Wed, 9 Aug 2017 16:41:56 +0000 (16:41 +0000)]
WPE-focused test gardening.
https://bugs.webkit.org/show_bug.cgi?id=175385

Unreviewed test gardening.

Patch by Ms2ger <Ms2ger@igalia.com> on 2017-08-09

* platform/wpe/TestExpectations: Marked failing tests.
* platform/wpe/fast/events/touch/basic-single-touch-events-expected.txt: Added platform-specific expectation.
* platform/wpe/js/date-timeClip-large-values-expected.txt: Removed (the platform-neutral expectation is fine).
* platform/wpe/js/dom/global-constructors-attributes-dedicated-worker-expected.txt: Updated for Cache enabling.
* transitions/clip-path-path-transitions-expected.txt: Updated for test change in r219243.
* transitions/resources/transition-test-helpers.js:
(parseClipPath): Updated to support the quotes in clip-path-path-transitions.html

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220460 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoREGRESSION (r219013): OAuth flows are broken when redirecting back to application...
dbates@webkit.org [Wed, 9 Aug 2017 16:30:41 +0000 (16:30 +0000)]
REGRESSION (r219013): OAuth flows are broken when redirecting back to application after authentication
https://bugs.webkit.org/show_bug.cgi?id=175247
<rdar://problem/33679804>

Reviewed by Brady Eidson.

Source/WebCore:

Add SPI so that Safari can differentiate between a form submission and a redirected form submission
and have PolicyCheck notify the frame loader client if the navigation was in response to receiving a
redirect response. This is the WebKit portion to fix an issue when a native app makes use of an OAuth
OAuth flow that bounces to Safari for user login and then bounce back to the app. Microsoft Graph's
OAuth flow is one example.

Safari was differentiating between a form submission and a redirected form submission based on the
nullity of WKNavigationAction.sourceFrame because in both cases the navigation type was WKNavigationTypeFormSubmitted.
The navigation type is the same for both navigations because WebKit always used the navigation
action from the original request for the redirect request when the original request redirected.
Prior to r219013, WKNavigationAction.sourceFrame would be nil for a form submission that redirects.
Following r219013, WKNavigationAction.sourceFrame is non-nil unless the navigation was initiated by
API. In particular, WKNavigationAction.sourceFrame is non-nil for the redirect navigation corresponding
to a form submission that redirects.

* loader/EmptyClients.cpp:
(WebCore::EmptyFrameLoaderClient::dispatchDecidePolicyForNavigationAction):
* loader/FrameLoaderClient.h:
Have dispatchDecidePolicyForNavigationAction() take a boolean as to whether the navigation was in
response to receiving a redirect response.
* loader/PolicyChecker.cpp:
(WebCore::PolicyChecker::checkNavigationPolicy): Notify the frame loader client whether the navigation
is in response to receiving a redirect response.

Source/WebKit:

Add SPI WKNavigationAction._isRedirect to query whether the navigation was in response to receiving
a redirect response. The majority of the WebKit change is plumbing this knowledge through to connect
it with the SPI.

* Shared/NavigationActionData.cpp:
(WebKit::NavigationActionData::encode const):
(WebKit::NavigationActionData::decode):
Encode and decode the boolean NavigationActionData::isRedirect.
* Shared/NavigationActionData.h:
* UIProcess/API/APINavigationAction.h:
* UIProcess/API/Cocoa/WKNavigationAction.mm:
(-[WKNavigationAction _isRedirect]): Added.
* UIProcess/API/Cocoa/WKNavigationActionPrivate.h:
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::dispatchDecidePolicyForNavigationAction): Set NavigationActionData::isRedirect
depending on whether the navigation is in response to receiving a redirect response.
* WebProcess/WebCoreSupport/WebFrameLoaderClient.h:

Source/WebKitLegacy/mac:

Plumb knowledge of whether a navigation was in response to receiving a redirect response.
We do not actually make use of this knowledge in WebKitLegacy because we do not know of any
clients that need to make use of this information at this time. If such a needs comes up
then we can expose API/SPI similar to what we do for WebKit.

* WebCoreSupport/WebFrameLoaderClient.h:
* WebCoreSupport/WebFrameLoaderClient.mm:
(WebFrameLoaderClient::dispatchDecidePolicyForNavigationAction):

Source/WebKitLegacy/win:

Plumb knowledge of whether a navigation was in response to receiving a redirect response.
We do not actually make use of this knowledge in WebKitLegacy because we do not know of any
clients that need to make use of this information at this time. If such a needs comes up
then we can expose API/SPI similar to what we do for WebKit.

* WebCoreSupport/WebFrameLoaderClient.cpp:
(WebFrameLoaderClient::dispatchDecidePolicyForNavigationAction):
* WebCoreSupport/WebFrameLoaderClient.h:

Tools:

Add tests for redirects.

* TestWebKitAPI/Tests/WebKit2Cocoa/DecidePolicyForNavigationAction.mm:
(TEST):
* TestWebKitAPI/cocoa/TestProtocol.mm:
(createRedirectURL):
(-[TestProtocol startLoading]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220459 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoSkip failing JSC tests stress/regress-169783.js and wasm.yaml/wasm/stress/oom.js.
ryanhaddad@apple.com [Wed, 9 Aug 2017 16:02:10 +0000 (16:02 +0000)]
Skip failing JSC tests stress/regress-169783.js and wasm.yaml/wasm/stress/oom.js.
https://bugs.webkit.org/show_bug.cgi?id=175255

Unreviewed test gardening.

* stress/regress-169783.js:
* wasm/stress/oom.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220458 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoWTF::Function does not allow for reference / non-default constructible return types
commit-queue@webkit.org [Wed, 9 Aug 2017 15:59:01 +0000 (15:59 +0000)]
WTF::Function does not allow for reference / non-default constructible return types
https://bugs.webkit.org/show_bug.cgi?id=175244

Patch by Sam Weinig <sam@webkit.org> on 2017-08-09
Reviewed by Chris Dumez.

Source/JavaScriptCore:

* runtime/ArrayBuffer.cpp:
(JSC::ArrayBufferContents::transferTo):
Call reset(), rather than clear() to avoid the call to destroy() in clear(). The
destroy call needed to be a no-op anyway, since the data is being moved.

Source/WebCore:

* bindings/js/JSCustomElementInterface.h:
(WebCore::JSCustomElementInterface::invokeCallback):
Update the default value for the addArguments parameter to be an empty lambda, rather than
default initialization, which leads to a null WTF::Function. This allows us to remove support
for calling null WTF::Function. No change in behavior.

Source/WebKit:

* UIProcess/WebResourceLoadStatisticsStore.h:
Update the default value for the updateCookiePartitioningForDomainsHandler parameter to be an
empty lambda, rather than default initialization, which leads to a null WTF::Function. This allows
us to remove support for calling null WTF::Function. No change in behavior.

Source/WTF:

When Function, then NoncopyableFunction, was templatized to allow non-void return values
in r201493, it maintained the behavior of being callable even if the Function was null.
To accomplish this, when null, the default construction of the return parameter was used.
This means Function can't be used with return types that are not default constructible,
such as reference types and Ref.

This behavior of returning something when null is surprising, as this is not how normal
functions behave, and not very useful. Instead, we now assert that the function is not
null when being called.

* wtf/Function.h:
(WTF::Function operator(...)):
Instead of allowing a null callable wrapper by returning the default construction of
the return type, assert that the wrapper is there when calling a Function.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220457 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[QuickLook] Use case-insensitive comparison of preview MIME types
aestes@apple.com [Wed, 9 Aug 2017 15:52:19 +0000 (15:52 +0000)]
[QuickLook] Use case-insensitive comparison of preview MIME types
https://bugs.webkit.org/show_bug.cgi?id=175350
<rdar://problem/33761159>

Reviewed by Brady Eidson.

Source/WebCore:

MIME types are case-insensitive, but PreviewLoader was using a case-sensitive comparison to
determine if a MIME type was supported by QLPreviewConverter. As a result we would fail to
preview Excel macro-enabled workbooks, since CFNetwork would sniff a MIME type of
"application/vnd.ms-excel.sheet.macroEnabled.12" but QuickLook contained
"application/vnd.ms-excel.sheet.macroenabled.12" in its set of supported MIME type strings.

Fix this by copying the QuickLook supported MIME type set into a HashSet using
ASCIICaseInsensitiveHash and using that HashSet for MIME type checks.

Test: quicklook/excel-macro-enabled.html

* loader/ios/PreviewLoader.mm:
(WebCore::PreviewLoader::shouldCreateForMIMEType):

LayoutTests:

* platform/ios-11/quicklook/excel-macro-enabled-expected.txt: Added.
* quicklook/excel-macro-enabled-expected.txt: Added.
* quicklook/excel-macro-enabled.html: Added.
* quicklook/resources/excel-macro-enabled.xlsm: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220456 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUnreviewed WPE gardening.
zandobersek@gmail.com [Wed, 9 Aug 2017 13:25:49 +0000 (13:25 +0000)]
Unreviewed WPE gardening.

* platform/wpe/TestExpectations: Stop skipping the whole fast/
directory and then unskipping specific subdirectories since
that overrides any expectations coming from the generic WK2
TestExpectations file. Specific subdirectories that we haven't
yet generated results for are now skipped instead.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220455 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoRemove unnecessary copies of PlatformMediaSessionManager::m_sessions
commit-queue@webkit.org [Wed, 9 Aug 2017 13:06:11 +0000 (13:06 +0000)]
Remove unnecessary copies of PlatformMediaSessionManager::m_sessions
https://bugs.webkit.org/show_bug.cgi?id=175375

Patch by Ms2ger <Ms2ger@igalia.com> on 2017-08-09
Reviewed by Sam Weinig.

Remove unnecessary copies of PlatformMediaSessionManager::m_sessions.

These should have been removed in r207688.

No new tests: no behavior change.

* platform/audio/PlatformMediaSessionManager.cpp:
(WebCore::PlatformMediaSessionManager::applicationWillBecomeInactive const):
(WebCore::PlatformMediaSessionManager::applicationDidBecomeActive const):
(WebCore::PlatformMediaSessionManager::applicationDidEnterBackground const):
(WebCore::PlatformMediaSessionManager::applicationWillEnterForeground const):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220454 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoReinstate active flag for iterators
cdumez@apple.com [Wed, 9 Aug 2017 13:04:39 +0000 (13:04 +0000)]
Reinstate active flag for iterators
https://bugs.webkit.org/show_bug.cgi?id=175312

Reviewed by Sam Weinig.

LayoutTests/imported/w3c:

Resync WPT tests from upstream to gain test coverage.

* web-platform-tests/dom/traversal/NodeIterator-expected.txt:
* web-platform-tests/dom/traversal/NodeIterator.html:
* web-platform-tests/dom/traversal/TreeWalker-expected.txt:
* web-platform-tests/dom/traversal/TreeWalker.html:

Source/WebCore:

NodeIterator / TreeWalker should no longer allow recursive filters
after the following change to the DOM specification:
- https://github.com/whatwg/dom/pull/359

This patch aligns our behavior with the latest specification.

No new tests, updated existing tests.

* dom/NodeIterator.cpp:
(WebCore::NodeIterator::nextNode):
(WebCore::NodeIterator::previousNode):
Note that we now also call m_candidateNode.clear() before returning an
exception. This was a pre-existing bug that we failed to do so in the
exception case but it became more obvious after this change now that
we throw. This was causing traversal/moz-bug559526.html to fail
otherwise (the filter was called one too many times). The test case
is passing in Firefox (The filter is called 4 times and they throw
each time).

* dom/Traversal.cpp:
(WebCore::NodeIteratorBase::NodeIteratorBase):
(WebCore::NodeIteratorBase::acceptNode):
* dom/Traversal.h:
* dom/TreeWalker.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220453 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[GTK] [WPE] Update expectations for global-constructors-attributes.html.
commit-queue@webkit.org [Wed, 9 Aug 2017 09:22:21 +0000 (09:22 +0000)]
[GTK] [WPE] Update expectations for global-constructors-attributes.html.
https://bugs.webkit.org/show_bug.cgi?id=175373

Unreviewed test gardening.

Patch by Ms2ger <Ms2ger@igalia.com> on 2017-08-09

* platform/gtk/js/dom/global-constructors-attributes-expected.txt:
* platform/wpe/js/dom/global-constructors-attributes-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220450 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoREGRESSION: 2 test262/test/language/statements/async-function failures
gskachkov@gmail.com [Wed, 9 Aug 2017 09:17:44 +0000 (09:17 +0000)]
REGRESSION: 2 test262/test/language/statements/async-function failures
https://bugs.webkit.org/show_bug.cgi?id=175334

Reviewed by Yusuke Suzuki.

JSTests:

Add @skip parameters to tests, and remove test for async iterator from
async await syntax test because it is already covered by async-iterator-syntax.js

* stress/async-await-syntax.js:
* stress/async-iteration-syntax.js:

Source/JavaScriptCore:

Switch off useAsyncIterator by default

* runtime/Options.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220449 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoRenderQuote should not mutate render tree
antti@apple.com [Wed, 9 Aug 2017 07:36:45 +0000 (07:36 +0000)]
RenderQuote should not mutate render tree
https://bugs.webkit.org/show_bug.cgi?id=175328

Reviewed by Zalan Bujtas.

RenderQuote text renderers are currently created and deleted in a quirky fashion using a linked list.
This patch moves to a simpler model that guarantees the mutations are always done in controlled fashion
during render tree update.

* dom/Document.cpp:
(WebCore::Document::updateTextRenderer):

    Move text renderer updating to Document so we can set the inRenderTreeUpdate bit for it too.

* dom/Document.h:
* dom/Text.cpp:
(WebCore::Text::updateRendererAfterContentChange):
* rendering/RenderDescendantIterator.h:
(WebCore::RenderDescendantIteratorAdapter<T>::at):
(WebCore::RenderDescendantConstIteratorAdapter<T>::at const):

    Add at() function for starting iteration from a specified renderer.

* rendering/RenderQuote.cpp:
(WebCore::RenderQuote::insertedIntoTree):
(WebCore::RenderQuote::willBeRemovedFromTree):

    Register and unregister quotes to RenderView.
    Don't do any mutations.

(WebCore::RenderQuote::styleDidChange):

    Invalidate the text renderer but don't mutate it.

(WebCore::RenderQuote::updateTextRenderer):
(WebCore::RenderQuote::computeText const):
(WebCore::RenderQuote::updateRenderers):

    Compute depth of all render quotes and update the text renderer as needed.

(WebCore::RenderQuote::willBeDestroyed): Deleted.
(WebCore::RenderQuote::attachQuote): Deleted.
(WebCore::RenderQuote::detachQuote): Deleted.
(WebCore::RenderQuote::updateDepth): Deleted.

    Get rid of the linked list.

* rendering/RenderQuote.h:
* rendering/RenderView.cpp:
(WebCore::RenderView::registerQuote):
(WebCore::RenderView::unregisterQuote):

    Maintain a render tree order ListHashSet of RenderQuotes.

(WebCore::RenderView::updateSpecialRenderers):

    Add a function for making additional render tree mutations at the end of a render tree update.
    Currently this just invokes RenderQuote::updateRenderers.

* rendering/RenderView.h:
* style/RenderTreeUpdater.cpp:
(WebCore::RenderTreeUpdater::commit):

    Call RenderView::updateSpecialRenderers after committing all other changes.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220447 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[WPE] Enable SUBTLE_CRYPTO in release builds
zandobersek@gmail.com [Wed, 9 Aug 2017 07:16:56 +0000 (07:16 +0000)]
[WPE] Enable SUBTLE_CRYPTO in release builds
https://bugs.webkit.org/show_bug.cgi?id=175371

Reviewed by Carlos Garcia Campos.

* Source/cmake/OptionsWPE.cmake:
Use ON as the default port value for ENABLE_SUBTLE_CRYPTO in WPE.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220446 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[GTK] Enable SUBTLE_CRYPTO in GTK+ releases
carlosgc@webkit.org [Wed, 9 Aug 2017 07:13:05 +0000 (07:13 +0000)]
[GTK] Enable SUBTLE_CRYPTO in GTK+ releases
https://bugs.webkit.org/show_bug.cgi?id=172758

Reviewed by Žan Doberšek.

Add ENABLE_WEB_CRYPTO public option and make ENABLE_SUBTLE_CRYPTO depend on it.

* Source/cmake/OptionsGTK.cmake:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220445 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[Soup] Incorrect conversion in msToSoupDate()
zandobersek@gmail.com [Wed, 9 Aug 2017 07:00:34 +0000 (07:00 +0000)]
[Soup] Incorrect conversion in msToSoupDate()
https://bugs.webkit.org/show_bug.cgi?id=175320

Reviewed by Carlos Garcia Campos.

Original patch by VaL Doroshchuk.

* platform/network/soup/NetworkStorageSessionSoup.cpp:
(WebCore::msToSoupDate): monthFromDayInYear() returns a value in the [0,11] range,
while soup_date_new() expects a month value in the [1,12] range. The return value
must be thus incremented by one to match that expectation.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220444 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUnreviewed, rolling out r220393.
wenson_hsieh@apple.com [Wed, 9 Aug 2017 05:36:57 +0000 (05:36 +0000)]
Unreviewed, rolling out r220393.

This patch regresses the performance of WebPage::editorState.

Reverted changeset:

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

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220443 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[Beacon] Add support for CORS-preflighting for WK2 / NETWORK_SESSION
cdumez@apple.com [Wed, 9 Aug 2017 05:15:47 +0000 (05:15 +0000)]
[Beacon] Add support for CORS-preflighting for WK2 / NETWORK_SESSION
https://bugs.webkit.org/show_bug.cgi?id=175264
<rdar://problem/33547793>

Reviewed by Youenn Fablet.

Source/WebCore:

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

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

* WebCore.xcodeproj/project.pbxproj:
* loader/CrossOriginAccessControl.cpp:
(WebCore::validatePreflightResponse):
* loader/CrossOriginAccessControl.h:
* loader/CrossOriginPreflightChecker.cpp:
(WebCore::CrossOriginPreflightChecker::validatePreflightResponse):
* loader/CrossOriginPreflightResultCache.h:
* loader/LoaderStrategy.h:
* loader/PingLoader.cpp:
(WebCore::PingLoader::loadImage):
(WebCore::PingLoader::sendPing):
(WebCore::PingLoader::sendViolationReport):
(WebCore::PingLoader::startPingLoad):
* loader/PingLoader.h:
* loader/cache/CachedResource.cpp:
(WebCore::CachedResource::load):
* page/SecurityOrigin.h:

Source/WebKit:

Implement CORS-preflighting for beacons with a payload that has a non
safelisted MIME type, as per:
- https://w3c.github.io/beacon/#privacy
- https://www.w3.org/TR/beacon/#sec-processing-model

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

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

* CMakeLists.txt:
* NetworkProcess/NetworkCORSPreflightChecker.cpp: Added.
(WebKit::NetworkCORSPreflightChecker::NetworkCORSPreflightChecker):
(WebKit::NetworkCORSPreflightChecker::~NetworkCORSPreflightChecker):
(WebKit::NetworkCORSPreflightChecker::startPreflight):
(WebKit::NetworkCORSPreflightChecker::willPerformHTTPRedirection):
(WebKit::NetworkCORSPreflightChecker::didReceiveChallenge):
(WebKit::NetworkCORSPreflightChecker::didReceiveResponseNetworkSession):
(WebKit::NetworkCORSPreflightChecker::didReceiveData):
(WebKit::NetworkCORSPreflightChecker::didCompleteWithError):
(WebKit::NetworkCORSPreflightChecker::didSendData):
(WebKit::NetworkCORSPreflightChecker::wasBlocked):
(WebKit::NetworkCORSPreflightChecker::cannotShowURL):
* NetworkProcess/NetworkCORSPreflightChecker.h: Added.
* NetworkProcess/NetworkConnectionToWebProcess.cpp:
(WebKit::NetworkConnectionToWebProcess::loadPing):
* NetworkProcess/NetworkConnectionToWebProcess.h:
* NetworkProcess/NetworkResourceLoadParameters.cpp:
(WebKit::NetworkResourceLoadParameters::encode const):
(WebKit::NetworkResourceLoadParameters::decode):
* NetworkProcess/NetworkResourceLoadParameters.h:
* NetworkProcess/PingLoad.cpp: Added.
(WebKit::PingLoad::PingLoad):
(WebKit::PingLoad::~PingLoad):
(WebKit::PingLoad::startNetworkLoad):
(WebKit::PingLoad::willPerformHTTPRedirection):
(WebKit::PingLoad::didReceiveChallenge):
(WebKit::PingLoad::didReceiveResponseNetworkSession):
(WebKit::PingLoad::didReceiveData):
(WebKit::PingLoad::didCompleteWithError):
(WebKit::PingLoad::didSendData):
(WebKit::PingLoad::wasBlocked):
(WebKit::PingLoad::cannotShowURL):
(WebKit::PingLoad::timeoutTimerFired):
(WebKit::PingLoad::needsCORSPreflight const):
(WebKit::PingLoad::doCORSPreflight):
* NetworkProcess/PingLoad.h:
* WebKit.xcodeproj/project.pbxproj:
* WebProcess/Network/WebLoaderStrategy.cpp:
(WebKit::WebLoaderStrategy::createPingHandle):
* WebProcess/Network/WebLoaderStrategy.h:

Source/WebKitLegacy:

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

* WebCoreSupport/WebResourceLoadScheduler.cpp:
(WebResourceLoadScheduler::createPingHandle):
* WebCoreSupport/WebResourceLoadScheduler.h:

LayoutTests:

Add layout test coverage.

* http/wpt/beacon/cors/cors-preflight-arraybufferview-failure-expected.txt: Added.
* http/wpt/beacon/cors/cors-preflight-arraybufferview-failure.html: Added.
* http/wpt/beacon/cors/cors-preflight-arraybufferview-success-expected.txt: Added.
* http/wpt/beacon/cors/cors-preflight-arraybufferview-success.html: Added.
* http/wpt/beacon/cors/cors-preflight-blob-failure-expected.txt: Added.
* http/wpt/beacon/cors/cors-preflight-blob-failure.html: Added.
* http/wpt/beacon/cors/cors-preflight-blob-success-expected.txt: Added.
* http/wpt/beacon/cors/cors-preflight-blob-success.html: Added.
* http/wpt/beacon/cors/cors-preflight-cookie-expected.txt: Added.
* http/wpt/beacon/cors/cors-preflight-cookie.html: Added.
* http/wpt/beacon/resources/beacon-preflight.py: Added.
(respondToCORSPreflight):
(main):
* http/wpt/beacon/resources/set-cookie.py: Added.
(main):
* platform/mac-wk1/TestExpectations:
* platform/mac-wk2/TestExpectations:
* platform/win/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220442 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoICs should do caging
fpizlo@apple.com [Wed, 9 Aug 2017 03:48:44 +0000 (03:48 +0000)]
ICs should do caging
https://bugs.webkit.org/show_bug.cgi?id=175295

Reviewed by Saam Barati.

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

* bytecode/AccessCase.cpp:
(JSC::AccessCase::generateImpl):
* bytecode/InlineAccess.cpp:
(JSC::InlineAccess::dumpCacheSizesAndCrash):
(JSC::InlineAccess::generateSelfPropertyAccess):
(JSC::InlineAccess::generateSelfPropertyReplace):
(JSC::InlineAccess::generateArrayLength):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220441 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoAddress review feedback from https://bugs.webkit.org/show_bug.cgi?id=175246.
weinig@apple.com [Wed, 9 Aug 2017 03:24:31 +0000 (03:24 +0000)]
Address review feedback from https://bugs.webkit.org/show_bug.cgi?id=175246.

* bindings/js/DOMPromiseProxy.h:
* css/FontFaceSet.cpp:
* css/FontFaceSet.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220440 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoWeb Inspector: Canvas: support editing WebGL shaders
drousso@apple.com [Wed, 9 Aug 2017 01:55:53 +0000 (01:55 +0000)]
Web Inspector: Canvas: support editing WebGL shaders
https://bugs.webkit.org/show_bug.cgi?id=124211
<rdar://problem/15448958>

Reviewed by Matt Baker.

Source/JavaScriptCore:

* inspector/protocol/Canvas.json:
Add `updateShader` command that will change the given shader's source to the provided string,
recompile, and relink it to its associated program.
Drive-by: add description to `requestShaderSource` command.
Source/WebCore:

Test: inspector/canvas/updateShader.html

* inspector/InspectorCanvasAgent.h:
* inspector/InspectorCanvasAgent.cpp:
(WebCore::InspectorCanvasAgent::updateShader):

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

Source/WebInspectorUI:

* UserInterface/Models/ShaderProgram.js:
(WI.ShaderProgram.prototype.updateVertexShader):
(WI.ShaderProgram.prototype.updateFragmentShader):
(WI.ShaderProgram.prototype._updateShader):

* UserInterface/Views/ShaderProgramContentView.js:
(WI.ShaderProgramContentView):
(WI.ShaderProgramContentView.prototype._contentDidChange):

LayoutTests:

* inspector/canvas/updateShader-expected.txt: Added.
* inspector/canvas/updateShader.html: Added.

* platform/win/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220436 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoRemove old and unused pointIsInAssistedNode definition
megan_gardner@apple.com [Wed, 9 Aug 2017 01:39:37 +0000 (01:39 +0000)]
Remove old and unused pointIsInAssistedNode definition
https://bugs.webkit.org/show_bug.cgi?id=175335

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

Reviewed by Tim Horton.

* Platform/spi/ios/UIKitSPI.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView pointIsInAssistedNode:]): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220435 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoCreated Individual EWS Pages
commit-queue@webkit.org [Wed, 9 Aug 2017 01:24:25 +0000 (01:24 +0000)]
Created Individual EWS Pages
https://bugs.webkit.org/show_bug.cgi?id=174477

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

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

* QueueStatusServer/handlers/patch.py:
(Patch.get): Handles the case for individual ews queue as well as for all EWS queues.
* QueueStatusServer/handlers/statusbubble.py:
(StatusBubble._build_bubble): Updated url for specific EWS.
* QueueStatusServer/main.py: Added a link to the individual ews in the routes dictionary.
* QueueStatusServer/templates/patch.html: Added a link to go directly to the page with
all ews queues.
* QueueStatusServer/templates/statusbubble.html: If you click on a statusbubble it
takes you directly to the individual ews page.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220434 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[WebIDL] Add support for Promise<> attributes
weinig@apple.com [Wed, 9 Aug 2017 01:01:44 +0000 (01:01 +0000)]
[WebIDL] Add support for Promise<> attributes
https://bugs.webkit.org/show_bug.cgi?id=175246

Reviewed by Yusuke Suzuki.

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

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

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

* Modules/encryptedmedia/MediaKeySession.cpp:
(WebCore::MediaKeySession::sessionClosed):
(WebCore::MediaKeySession::registerClosedPromise): Deleted.
* Modules/encryptedmedia/MediaKeySession.h:
* Modules/encryptedmedia/MediaKeySession.idl:

    Replace custom promise code using the registerPromise idiom with DOMPromiseProxy.

* bindings/IDLTypes.h:

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

(WebCore::IDLWrapper::convertToParameterType):

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

* bindings/js/DOMPromiseProxy.h: Added.
(WebCore::DOMPromiseProxy<IDLType>):

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

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

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

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

(WebCore::DOMPromiseProxy<IDLVoid>):

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

(WebCore::DOMPromiseProxyWithResolveCallback<IDLType>):

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

* bindings/js/JSDOMConvertPromise.h:
(WebCore::JSConverter<IDLPromise<T>>::convert):
Add converter that calls through to the DOMPromiseProxy function promise.

* bindings/js/JSDOMPromiseDeferred.cpp:
(WebCore::DeferredPromise::callFunction):
(WebCore::DeferredPromise::reject):
* bindings/js/JSDOMPromiseDeferred.h:
(WebCore::DeferredPromise::create):
(WebCore::DeferredPromise::DeferredPromise):

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

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

* bindings/scripts/CodeGeneratorJS.pm:
(AddToIncludesForIDLType):
(GenerateParametersCheck):
(GenerateImplementationFunctionCall):
(NativeToJSValueDOMConvertNeedsState):
(NativeToJSValueDOMConvertNeedsGlobalObject):

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

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

* bindings/scripts/IDLAttributes.json:

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

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

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

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

    Replace custom promise code using the registerPromise idiom with DOMPromiseProxy.

* page/NavigatorBase.cpp:
(WebCore::NavigatorBase::NavigatorBase):
* page/NavigatorBase.h:

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

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

    Replace custom promise code using the registerPromise idiom with DOMPromiseProxy.

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

    Add / update tests.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220433 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoMake JSC_validateExceptionChecks=1 succeed on JSTests/slowMicrobenchmarks/spread...
commit-queue@webkit.org [Wed, 9 Aug 2017 00:26:04 +0000 (00:26 +0000)]
Make JSC_validateExceptionChecks=1 succeed on JSTests/slowMicrobenchmarks/spread-small-array.js.
https://bugs.webkit.org/show_bug.cgi?id=175347

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

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

* bytecode/CodeBlock.cpp:
(JSC::CodeBlock::finishCreation):
(JSC::CodeBlock::setConstantIdentifierSetRegisters):
(JSC::CodeBlock::setConstantRegisters):
* bytecode/CodeBlock.h:
* runtime/ScriptExecutable.cpp:
(JSC::ScriptExecutable::newCodeBlockFor):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220432 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoSeparate jsc stress test script writer from run-jsc-stress-tests
commit-queue@webkit.org [Wed, 9 Aug 2017 00:18:08 +0000 (00:18 +0000)]
Separate jsc stress test script writer from run-jsc-stress-tests
https://bugs.webkit.org/show_bug.cgi?id=175216

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

* Scripts/run-jsc-stress-tests:
* Scripts/webkitruby/jsc-stress-test-writer-default.rb: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220431 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoResource Load Statistics: Merge of dataRecordsRemoved should use std::max()
wilander@apple.com [Wed, 9 Aug 2017 00:03:55 +0000 (00:03 +0000)]
Resource Load Statistics: Merge of dataRecordsRemoved should use std::max()
https://bugs.webkit.org/show_bug.cgi?id=175348
<rdar://problem/33785695>

Reviewed by Chris Dumez.

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

* loader/ResourceLoadStatistics.cpp:
(WebCore::ResourceLoadStatistics::merge):
    Now uses std::max() instead of addition.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220430 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoDon't enable default icon loading in WK1 for apps linked against old SDKs.
beidson@apple.com [Tue, 8 Aug 2017 23:53:04 +0000 (23:53 +0000)]
Don't enable default icon loading in WK1 for apps linked against old SDKs.
<rdar://problem/33724060> and https://bugs.webkit.org/show_bug.cgi?id=175342

Reviewed by Andy Estes.

* Misc/WebKitVersionChecks.h:
* WebCoreSupport/WebFrameLoaderClient.mm:
(WebFrameLoaderClient::getLoadDecisionForIcons): Don't do any icon loading unless
  the app is linked against new WebKit with the new behavior.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220429 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoMark media/modern-media-controls/css/webkit-cursor-visibility-auto-hide.html as flaky.
ryanhaddad@apple.com [Tue, 8 Aug 2017 23:36:49 +0000 (23:36 +0000)]
Mark media/modern-media-controls/css/webkit-cursor-visibility-auto-hide.html as flaky.
https://bugs.webkit.org/show_bug.cgi?id=175195

Unreviewed test gardening.

* platform/mac-wk1/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220428 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoSandbox flags do not support document.domain control
bfulgham@apple.com [Tue, 8 Aug 2017 23:27:08 +0000 (23:27 +0000)]
Sandbox flags do not support document.domain control
https://bugs.webkit.org/show_bug.cgi?id=175281
<rdar://problem/33778936>

Reviewed by Chris Dumez.

Source/WebCore:

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

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

* dom/Document.cpp:
(WebCore::Document::setDomain): Add check for sandbox flag (with appropriate error message)
* dom/SecurityContext.h:

LayoutTests:

* fast/frames/resources/sandboxed-iframe-set-domain.html: Added.
* fast/frames/sandboxed-iframe-domain.html: Added.
* fast/frames/sandboxed-iframe-domain-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220427 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoSkipping imported/w3c/IndexedDB-private-browsing/idbfactory_open12.html
jlewis3@apple.com [Tue, 8 Aug 2017 23:24:08 +0000 (23:24 +0000)]
Skipping imported/w3c/IndexedDB-private-browsing/idbfactory_open12.html
<rdar://problem/33555052>

Unreviewed test gardening.

* platform/ios-wk2/TestExpectations:
* platform/mac-highsierra/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220426 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUnreviewed, fix Ubuntu LTS build
mcatanzaro@igalia.com [Tue, 8 Aug 2017 23:03:21 +0000 (23:03 +0000)]
Unreviewed, fix Ubuntu LTS build
https://bugs.webkit.org/show_bug.cgi?id=174490

Source/JavaScriptCore:

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

Source/WebDriver:

* glib/SessionHostGlib.cpp:

Source/WebKit:

* UIProcess/glib/RemoteInspectorClient.cpp:
(WebKit::RemoteInspectorClient::RemoteInspectorClient):

Tools:

* TestWebKitAPI/Tests/WebKitGLib/TestAutomationSession.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220425 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUnreviewed, fix the Mac debug build after r220418.
wenson_hsieh@apple.com [Tue, 8 Aug 2017 21:06:18 +0000 (21:06 +0000)]
Unreviewed, fix the Mac debug build after r220418.

* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::prepareForMoveToWindow):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220420 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoSkip plugins/crash-restoring-plugin-page-from-page-cache.html on macOS debug.
ryanhaddad@apple.com [Tue, 8 Aug 2017 20:55:09 +0000 (20:55 +0000)]
Skip plugins/crash-restoring-plugin-page-from-page-cache.html on macOS debug.
https://bugs.webkit.org/show_bug.cgi?id=175270

Unreviewed test gardening.

* platform/mac-wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220419 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[Mac WK2] WebViewImpl should hold a weak pointer to its WKWebView or WKView
wenson_hsieh@apple.com [Tue, 8 Aug 2017 20:45:21 +0000 (20:45 +0000)]
[Mac WK2] WebViewImpl should hold a weak pointer to its WKWebView or WKView
https://bugs.webkit.org/show_bug.cgi?id=175302
<rdar://problem/33762116>

Reviewed by Tim Horton.

Source/WebKit:

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

Test: WKWebViewMacEditingTests.DoNotCrashWhenInterpretingKeyEventWhileDeallocatingView.

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

Tools:

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

* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/Tests/TestWebKitAPI/mac/AppKitSPI.h: Renamed from Tools/TestWebKitAPI/Tests/TestWebKitAPI/mac/NSTextInputClientSPI.h.
* TestWebKitAPI/Tests/mac/WKWebViewMacEditingTests.mm: Renamed from Tools/TestWebKitAPI/Tests/mac/WKWebViewSelectionTests.mm.
(-[SlowTextInputContext handleEventByInputMethod:completionHandler:]):
(-[SlowTextInputContext handleEvent:completionHandler:]):
(-[SlowInputWebView inputContext]):
(-[SlowInputWebView _web_superInputContext]):
(TEST):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220418 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoMark fast/images/async-image-body-background-image.html as flaky on macOS.
ryanhaddad@apple.com [Tue, 8 Aug 2017 19:58:26 +0000 (19:58 +0000)]
Mark fast/images/async-image-body-background-image.html as flaky on macOS.
https://bugs.webkit.org/show_bug.cgi?id=175193

Unreviewed test gardening.

* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220417 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoBaseline JIT should do caging
fpizlo@apple.com [Tue, 8 Aug 2017 19:44:20 +0000 (19:44 +0000)]
Baseline JIT should do caging
https://bugs.webkit.org/show_bug.cgi?id=175037

Reviewed by Mark Lam.
Source/bmalloc:

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

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

* bmalloc/Gigacage.cpp:
(Gigacage::primitiveGigacageDisabled):
(Gigacage::disableDisablingPrimitiveGigacageIfShouldBeEnabled):
(Gigacage::isDisablingPrimitiveGigacageDisabled):
* bmalloc/Gigacage.h:
(Gigacage::isPrimitiveGigacagePermanentlyEnabled):
(Gigacage::canPrimitiveGigacageBeDisabled):

Source/JavaScriptCore:

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

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

Relanded with fixed AssemblyHelpers::cageConditionally().

* bytecode/AccessCase.cpp:
(JSC::AccessCase::generateImpl):
* bytecode/InlineAccess.cpp:
(JSC::InlineAccess::dumpCacheSizesAndCrash):
(JSC::InlineAccess::generateSelfPropertyAccess):
(JSC::InlineAccess::generateSelfPropertyReplace):
(JSC::InlineAccess::generateArrayLength):
* ftl/FTLLowerDFGToB3.cpp:
(JSC::FTL::DFG::LowerDFGToB3::caged):
* jit/AssemblyHelpers.h:
(JSC::AssemblyHelpers::cage):
(JSC::AssemblyHelpers::cageConditionally):
* jit/JITPropertyAccess.cpp:
(JSC::JIT::emitDoubleLoad):
(JSC::JIT::emitContiguousLoad):
(JSC::JIT::emitArrayStorageLoad):
(JSC::JIT::emitGenericContiguousPutByVal):
(JSC::JIT::emitArrayStoragePutByVal):
(JSC::JIT::emit_op_get_from_scope):
(JSC::JIT::emit_op_put_to_scope):
(JSC::JIT::emitIntTypedArrayGetByVal):
(JSC::JIT::emitFloatTypedArrayGetByVal):
(JSC::JIT::emitIntTypedArrayPutByVal):
(JSC::JIT::emitFloatTypedArrayPutByVal):
* jsc.cpp:
(jscmain):
(primitiveGigacageDisabled): Deleted.

Source/WebKit:

Use a better API to disable disabling the primitive gigacage.

* WebProcess/WebProcess.cpp:
(WebKit::m_webSQLiteDatabaseTracker):
(WebKit::primitiveGigacageDisabled): Deleted.

Source/WTF:

* wtf/Gigacage.h:
(Gigacage::disableDisablingPrimitiveGigacageIfShouldBeEnabled):
(Gigacage::isDisablingPrimitiveGigacageDisabled):
(Gigacage::isPrimitiveGigacagePermanentlyEnabled):
(Gigacage::canPrimitiveGigacageBeDisabled):
(Gigacage::basePtr):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220416 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUnreviewed, gardening test262 for Promise resolve / reject function length
utatane.tea@gmail.com [Tue, 8 Aug 2017 19:16:10 +0000 (19:16 +0000)]
Unreviewed, gardening test262 for Promise resolve / reject function length
https://bugs.webkit.org/show_bug.cgi?id=175333

* test262.yaml:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220415 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoChange fast seek logic to prevent ping-ponging.
commit-queue@webkit.org [Tue, 8 Aug 2017 19:03:57 +0000 (19:03 +0000)]
Change fast seek logic to prevent ping-ponging.
https://bugs.webkit.org/show_bug.cgi?id=175167
rdar://problem/32936116

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

Source/WebCore:

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

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

* html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::fastSeek):

LayoutTests:

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

* media/video-fast-seek.html:
* media/media-source/media-source-fastseek.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220414 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoRemove obsolete failure expectations for event_loadedmetadata.html.
commit-queue@webkit.org [Tue, 8 Aug 2017 17:22:07 +0000 (17:22 +0000)]
Remove obsolete failure expectations for event_loadedmetadata.html.
https://bugs.webkit.org/show_bug.cgi?id=175027

Unreviewed test gardening.

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

* TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220413 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoResource Load Statistics: Clear web processes' state on history removal
wilander@apple.com [Tue, 8 Aug 2017 17:18:24 +0000 (17:18 +0000)]
Resource Load Statistics: Clear web processes' state on history removal
https://bugs.webkit.org/show_bug.cgi?id=175271
<rdar://problem/33756560>

Reviewed by Chris Dumez.

Source/WebKit:

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

* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::clearResourceLoadStatistics):
* UIProcess/WebProcessPool.h:
* UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::removeData):
(WebKit::WebsiteDataStore::clearResourceLoadStatisticsInWebProcesses):
* UIProcess/WebsiteData/WebsiteDataStore.h:
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::clearResourceLoadStatistics):
* WebProcess/WebProcess.h:
* WebProcess/WebProcess.messages.in:

Tools:

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

* WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
* WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
(WTR::InjectedBundle::didReceiveMessageToPage):
* WebKitTestRunner/InjectedBundle/TestRunner.cpp:
(WTR::TestRunner::statisticsClearThroughWebsiteDataRemoval):
(WTR::TestRunner::statisticsCallClearThroughWebsiteDataRemovalCallback):
* WebKitTestRunner/InjectedBundle/TestRunner.h:
* WebKitTestRunner/TestController.cpp:
(WTR::TestController::statisticsClearThroughWebsiteDataRemoval):
* WebKitTestRunner/TestController.h:
* WebKitTestRunner/TestInvocation.cpp:
(WTR::TestInvocation::didReceiveSynchronousMessageFromInjectedBundle):
(WTR::TestInvocation::didClearStatisticsThroughWebsiteDataRemoval):
* WebKitTestRunner/TestInvocation.h:
* WebKitTestRunner/cocoa/TestControllerCocoa.mm:
(WTR::TestController::statisticsClearThroughWebsiteDataRemoval):

LayoutTests:

* http/tests/loading/resourceLoadStatistics/user-interaction-reported-after-website-data-removal-expected.txt: Added.
* http/tests/loading/resourceLoadStatistics/user-interaction-reported-after-website-data-removal.html: Added.
* platform/mac-wk2/TestExpectations:
    user-interaction-reported-after-website-data-removal.html marked as [ Pass ].

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220412 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUnreviewed. Correct the function name in a debug message.
commit-queue@webkit.org [Tue, 8 Aug 2017 17:15:49 +0000 (17:15 +0000)]
Unreviewed. Correct the function name in a debug message.
https://bugs.webkit.org/show_bug.cgi?id=175262

No new tests: debug logs are not tested.

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

* platform/audio/PlatformMediaSessionManager.cpp:
(WebCore::PlatformMediaSessionManager::applicationDidBecomeActive const):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220411 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoFix baseline for http/tests/security/set-domain-remove-subdomain-for-ip-address.html.
ryanhaddad@apple.com [Tue, 8 Aug 2017 17:14:01 +0000 (17:14 +0000)]
Fix baseline for http/tests/security/set-domain-remove-subdomain-for-ip-address.html.

Unreviewed test gardening.

* http/tests/security/set-domain-remove-subdomain-for-ip-address-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220410 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoFixed rebaselined expectations for js/dom/global-constructors-attributes.html.
jlewis3@apple.com [Tue, 8 Aug 2017 17:00:12 +0000 (17:00 +0000)]
Fixed rebaselined expectations for js/dom/global-constructors-attributes.html.
https://bugs.webkit.org/show_bug.cgi?id=175201

Unreviewed test gardening.

* platform/mac-elcapitan-wk2/js/dom/global-constructors-attributes-expected.txt: Renamed from LayoutTests/platform/mac-elcapitan-wk2/js/dom/global-constructors-attributes-dedicated-worker-expected.txt.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220409 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[WPE] Implement WebsiteDataStore::defaultApplicationCacheDirectory() and friends
aperez@igalia.com [Tue, 8 Aug 2017 16:49:15 +0000 (16:49 +0000)]
[WPE] Implement WebsiteDataStore::defaultApplicationCacheDirectory() and friends
https://bugs.webkit.org/show_bug.cgi?id=175322

Reviewed by Carlos Garcia Campos.

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

* PlatformGTK.cmake: Add APIWebsiteDataStoreGLib.cpp to the build sources and remove APIWebsiteDataStoreGtk.cpp.
* PlatformWPE.cmake: Add APIWebsiteDataStoreGLib.cpp to the built sources.
* UIProcess/API/APIWebsiteDataStore.cpp: Remove now dead code.
* UIProcess/API/glib/APIWebsiteDataStoreGLib.cpp: Renamed from Source/WebKit/UIProcess/API/gtk/APIWebsiteDataStoreGtk.cpp.
(API::WebsiteDataStore::defaultApplicationCacheDirectory):
(API::WebsiteDataStore::defaultIndexedDBDatabaseDirectory):
(API::WebsiteDataStore::defaultLocalStorageDirectory):
(API::WebsiteDataStore::defaultMediaKeysStorageDirectory):
(API::WebsiteDataStore::defaultWebSQLDatabaseDirectory):
(API::WebsiteDataStore::defaultResourceLoadStatisticsDirectory):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220408 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoAdd Alicia Boya García as contributor
mcatanzaro@igalia.com [Tue, 8 Aug 2017 16:41:50 +0000 (16:41 +0000)]
Add Alicia Boya García as contributor
https://bugs.webkit.org/show_bug.cgi?id=175326

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

* Scripts/webkitpy/common/config/contributors.json:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220407 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUnreviewed test results fix after r220376.
bfulgham@apple.com [Tue, 8 Aug 2017 16:24:19 +0000 (16:24 +0000)]
Unreviewed test results fix after r220376.

Rebaseline for new SecurityException message text.

* fast/dom/Document/invalid-domain-change-throws-exception-expected.txt:
* http/tests/dom/document-attributes-null-handling-expected.txt:
* http/tests/security/set-domain-remove-subdomain-for-ip-address-expected.txt:
* http/tests/security/xss-DENIED-invalid-domain-change-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220406 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUnreviewed build fix after r220376.
bfulgham@apple.com [Tue, 8 Aug 2017 16:13:48 +0000 (16:13 +0000)]
Unreviewed build fix after r220376.

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

* dom/Document.cpp:
(WebCore::Document::domainIsRegisterable const):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220405 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUnreviewed, rolling out r220368.
ryanhaddad@apple.com [Tue, 8 Aug 2017 16:00:06 +0000 (16:00 +0000)]
Unreviewed, rolling out r220368.

This change caused WK1 tests to exit early with crashes.

Reverted changeset:

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

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220404 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[CMake] Properly test if compiler supports compiler flags
mcatanzaro@igalia.com [Tue, 8 Aug 2017 15:03:48 +0000 (15:03 +0000)]
[CMake] Properly test if compiler supports compiler flags
https://bugs.webkit.org/show_bug.cgi?id=174490

Reviewed by Konstantin Tokarev.

.:

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

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

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

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

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

Source/JavaScriptCore:

* API/tests/PingPongStackOverflowTest.cpp:
(testPingPongStackOverflow):
* API/tests/testapi.c:
* b3/testb3.cpp:
(JSC::B3::testPatchpointLotsOfLateAnys):

Source/ThirdParty:

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

Source/WebCore:

* CMakeLists.txt:
* PlatformGTK.cmake:
* PlatformWPE.cmake:

Source/WebDriver:

* WebDriverService.cpp:
(WebDriver::WebDriverService::run):
* glib/SessionHostGlib.cpp:

Source/WebKit:

* CMakeLists.txt:
* PlatformGTK.cmake:

Source/WTF:

* wtf/Compiler.h:

Tools:

* DumpRenderTree/TestNetscapePlugIn/CMakeLists.txt:
* MiniBrowser/gtk/CMakeLists.txt:
* TestRunnerShared/Bindings/JSWrapper.cpp:
(WTR::JSWrapper::initialize):
* TestWebKitAPI/CMakeLists.txt:
* TestWebKitAPI/PlatformGTK.cmake:
* TestWebKitAPI/Tests/WTF/CheckedArithmeticOperations.cpp:
(TestWebKitAPI::CheckedArithmeticTester::run):
* TestWebKitAPI/Tests/WebKitGLib/TestAutomationSession.cpp:
* TestWebKitAPI/Tests/WebKitGLib/TestWebExtensions.cpp:
* TestWebKitAPI/Tests/WebKitGLib/WebExtensionTest.cpp:
(formControlsAssociatedCallback):
* TestWebKitAPI/glib/CMakeLists.txt:
* TestWebKitAPI/glib/WebKitGLib/TestMain.h:
(Test::getResourcesDir):
* WebKitTestRunner/CMakeLists.txt:
* WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
(WTR::menuItemClickCallback):
(WTR::staticConvertMenuItemToType):
* WebKitTestRunner/InjectedBundle/TestRunner.cpp:
(WTR::TestRunner::setUseDashboardCompatibilityMode):
* WebKitTestRunner/InjectedBundle/atk/AccessibilityNotificationHandlerAtk.cpp:
(WTR::AccessibilityNotificationHandler::disconnectAccessibilityCallbacks):
* WebKitTestRunner/InjectedBundle/atk/AccessibilityUIElementAtk.cpp:
(WTR::AccessibilityUIElement::helpText const):
(WTR::AccessibilityUIElement::attributedStringForRange):
* WebKitTestRunner/gtk/EventSenderProxyGtk.cpp:
(WTR::EventSenderProxy::updateTouchPoint):
(WTR::EventSenderProxy::releaseTouchPoint):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220403 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[GStreamer] Don't use GraphicsContext3D in VideoTextureCoperGStreamer
zandobersek@gmail.com [Tue, 8 Aug 2017 13:29:30 +0000 (13:29 +0000)]
[GStreamer] Don't use GraphicsContext3D in VideoTextureCoperGStreamer
https://bugs.webkit.org/show_bug.cgi?id=174774

Reviewed by Carlos Garcia Campos.

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

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

No new tests -- no change in behavior.

* platform/graphics/gstreamer/VideoTextureCopierGStreamer.cpp:
(WebCore::VideoTextureCopierGStreamer::VideoTextureCopierGStreamer):
(WebCore::VideoTextureCopierGStreamer::~VideoTextureCopierGStreamer):
(WebCore::VideoTextureCopierGStreamer::copyVideoTextureToPlatformTexture):
* platform/graphics/gstreamer/VideoTextureCopierGStreamer.h:
(WebCore::VideoTextureCopierGStreamer::resultTexture):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220402 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[Linux] Clear WasmMemory with madvice instead of memset
utatane.tea@gmail.com [Tue, 8 Aug 2017 13:11:00 +0000 (13:11 +0000)]
[Linux] Clear WasmMemory with madvice instead of memset
https://bugs.webkit.org/show_bug.cgi?id=175150

Reviewed by Filip Pizlo.

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

We also commit grown pages in all OSes.

* wasm/WasmMemory.cpp:
(JSC::Wasm::commitZeroPages):
(JSC::Wasm::Memory::create):
(JSC::Wasm::Memory::grow):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220401 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[TexMap] Drop GC3D* type usage from TextureMapperPlatformLayer{Buffer,Proxy}
zandobersek@gmail.com [Tue, 8 Aug 2017 13:05:22 +0000 (13:05 +0000)]
[TexMap] Drop GC3D* type usage from TextureMapperPlatformLayer{Buffer,Proxy}
https://bugs.webkit.org/show_bug.cgi?id=175314

Reviewed by Carlos Garcia Campos.

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

No new tests -- no change in behavior.

* platform/graphics/texmap/TextureMapperPlatformLayerBuffer.cpp:
(WebCore::TextureMapperPlatformLayerBuffer::TextureMapperPlatformLayerBuffer):
(WebCore::TextureMapperPlatformLayerBuffer::canReuseWithoutReset):
* platform/graphics/texmap/TextureMapperPlatformLayerBuffer.h:
* platform/graphics/texmap/TextureMapperPlatformLayerProxy.cpp:
(WebCore::TextureMapperPlatformLayerProxy::getAvailableBuffer):
* platform/graphics/texmap/TextureMapperPlatformLayerProxy.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220400 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[TexMap] Don't use GraphicsContext3D in ClipStack
zandobersek@gmail.com [Tue, 8 Aug 2017 13:04:40 +0000 (13:04 +0000)]
[TexMap] Don't use GraphicsContext3D in ClipStack
https://bugs.webkit.org/show_bug.cgi?id=174776

Reviewed by Carlos Garcia Campos.

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

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

No new tests -- no change in behavior.

* platform/graphics/texmap/BitmapTextureGL.cpp:
(WebCore::BitmapTextureGL::clearIfNeeded):
(WebCore::BitmapTextureGL::bindAsSurface):
* platform/graphics/texmap/ClipStack.cpp:
(WebCore::ClipStack::apply):
(WebCore::ClipStack::applyIfNeeded):
* platform/graphics/texmap/ClipStack.h:
* platform/graphics/texmap/TextureMapperGL.cpp:
(WebCore::TextureMapperGL::bindDefaultSurface):
(WebCore::TextureMapperGL::beginScissorClip):
(WebCore::TextureMapperGL::beginClip):
(WebCore::TextureMapperGL::endClip):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220399 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoNot possible to remove the 'li' element inside the table cell
jfernandez@igalia.com [Tue, 8 Aug 2017 12:22:33 +0000 (12:22 +0000)]
Not possible to remove the 'li' element inside the table cell
https://bugs.webkit.org/show_bug.cgi?id=173148

Reviewed by Ryosuke Niwa.

Source/WebCore:

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

Tests: editing/deleting/delete-list-items-in-table-cell-1.html
       editing/deleting/delete-list-items-in-table-cell-2.html
       editing/deleting/delete-list-items-in-table-cell-3.html
       editing/deleting/delete-list-items-in-table-cell-4.html
       editing/deleting/delete-list-items-in-table-cell-5.html
       editing/deleting/delete-list-items-in-table-cell-6.html
       editing/deleting/delete-list-items-in-table-cell-7.html
       editing/deleting/delete-list-items-in-table-cell-8.html

* editing/TypingCommand.cpp:
(WebCore::TypingCommand::deleteKeyPressed):

LayoutTests:

Regression tests for different scenarios of list items removal.

* editing/deleting/delete-list-items-in-table-cell-1-expected.txt: Added.
* editing/deleting/delete-list-items-in-table-cell-1.html: Added.
* editing/deleting/delete-list-items-in-table-cell-2-expected.txt: Added.
* editing/deleting/delete-list-items-in-table-cell-2.html: Added.
* editing/deleting/delete-list-items-in-table-cell-3-expected.txt: Added.
* editing/deleting/delete-list-items-in-table-cell-3.html: Added.
* editing/deleting/delete-list-items-in-table-cell-4-expected.txt: Added.
* editing/deleting/delete-list-items-in-table-cell-4.html: Added.
* editing/deleting/delete-list-items-in-table-cell-5-expected.txt: Added.
* editing/deleting/delete-list-items-in-table-cell-5.html: Added.
* editing/deleting/delete-list-items-in-table-cell-6-expected.txt: Added.
* editing/deleting/delete-list-items-in-table-cell-6.html: Added.
* editing/deleting/delete-list-items-in-table-cell-7-expected.txt: Added.
* editing/deleting/delete-list-items-in-table-cell-7.html: Added.
* editing/deleting/delete-list-items-in-table-cell-8-expected.txt: Added.
* editing/deleting/delete-list-items-in-table-cell-8.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220398 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[TexMap] Isolate the TextureMapperPlatformLayerProxyProvider class
zandobersek@gmail.com [Tue, 8 Aug 2017 10:27:11 +0000 (10:27 +0000)]
[TexMap] Isolate the TextureMapperPlatformLayerProxyProvider class
https://bugs.webkit.org/show_bug.cgi?id=175316

Reviewed by Carlos Garcia Campos.

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

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

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

No new tests -- no change in behavior.

* platform/graphics/cairo/ImageBufferCairo.cpp:
(WebCore::ImageBufferData::proxy const):
* platform/graphics/cairo/ImageBufferDataCairo.h:
* platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.cpp:
(WebCore::MediaPlayerPrivateGStreamerBase::updateTexture):
(WebCore::MediaPlayerPrivateGStreamerBase::proxy const):
(WebCore::MediaPlayerPrivateGStreamerBase::swapBuffersIfNeeded):
* platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.h:
* platform/graphics/texmap/TextureMapperGC3DPlatformLayer.cpp:
* platform/graphics/texmap/TextureMapperGC3DPlatformLayer.h:
* platform/graphics/texmap/TextureMapperPlatformLayerProxy.h:
* platform/graphics/texmap/TextureMapperPlatformLayerProxyProvider.h: Added.
* platform/graphics/texmap/coordinated/CoordinatedGraphicsLayer.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220397 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUnreviewed. Follow-up to r220392 that fixes build on configurations
zandobersek@gmail.com [Tue, 8 Aug 2017 09:03:40 +0000 (09:03 +0000)]
Unreviewed. Follow-up to r220392 that fixes build on configurations
that disable USE(GSTREAMER_GL).

* platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.cpp:
(WebCore::MediaPlayerPrivateGStreamerBase::pushTextureToCompositor):
When creating the BitmapTextureGL object, also provide an initialized
TextureMapperContextAttributes object.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220396 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[TexMap] Group GL header file inclusions in a single header file
zandobersek@gmail.com [Tue, 8 Aug 2017 08:36:38 +0000 (08:36 +0000)]
[TexMap] Group GL header file inclusions in a single header file
https://bugs.webkit.org/show_bug.cgi?id=175313

Reviewed by Carlos Garcia Campos.

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

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

* platform/graphics/texmap/TextureMapperContextAttributes.cpp:
* platform/graphics/texmap/TextureMapperGLHeaders.h: Copied from Source/WebCore/platform/graphics/texmap/TextureMapperContextAttributes.cpp.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220395 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoWeb Automation: setUserInputForCurrentJavaScriptPrompt should fail if current dialog...
carlosgc@webkit.org [Tue, 8 Aug 2017 08:29:19 +0000 (08:29 +0000)]
Web Automation: setUserInputForCurrentJavaScriptPrompt should fail if current dialog is not a prompt
https://bugs.webkit.org/show_bug.cgi?id=175261

Reviewed by Brian Burg.

Source/WebDriver:

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

Source/WebKit:

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

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

This fixes selenium test testSettingTheValueOfAnAlertThrows.

* UIProcess/API/APIAutomationSessionClient.h:
(API::AutomationSessionClient::typeOfCurrentJavaScriptDialogOnPage):
* UIProcess/API/glib/WebKitAutomationSession.cpp:
* UIProcess/API/glib/WebKitWebView.cpp:
(webkitWebViewGetCurrentScriptDialogType):
* UIProcess/API/glib/WebKitWebViewPrivate.h:
* UIProcess/Automation/Automation.json:
* UIProcess/Automation/WebAutomationSession.cpp:
(WebKit::WebAutomationSession::setUserInputForCurrentJavaScriptPrompt):
* UIProcess/Cocoa/AutomationSessionClient.h:
* UIProcess/Cocoa/AutomationSessionClient.mm:
(WebKit::AutomationSessionClient::typeOfCurrentJavaScriptDialogOnPage):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220394 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[iOS WK2] WKWebView schedules nonstop layout after pressing cmb+b,i,u inside a conten...
wenson_hsieh@apple.com [Tue, 8 Aug 2017 08:10:23 +0000 (08:10 +0000)]
[iOS WK2] WKWebView schedules nonstop layout after pressing cmb+b,i,u inside a contenteditable div
https://bugs.webkit.org/show_bug.cgi?id=175116
<rdar://problem/28279301>

Reviewed by Darin Adler and Ryosuke Niwa.

Source/WebCore:

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

Tests: EditorStateTests.TypingAttributesBold
       EditorStateTests.TypingAttributesItalic
       EditorStateTests.TypingAttributesUnderline
       EditorStateTests.TypingAttributesTextAlignmentAbsoluteAlignmentOptions
       EditorStateTests.TypingAttributesTextAlignmentStartEnd
       EditorStateTests.TypingAttributesTextAlignmentDirectionalText
       EditorStateTests.TypingAttributesTextColor
       EditorStateTests.TypingAttributesMixedStyles
       EditorStateTests.TypingAttributesLinkColor

* css/StyleProperties.cpp:
(WebCore::StyleProperties::propertyAsColor const):
(WebCore::StyleProperties::propertyAsValueID const):

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

* css/StyleProperties.h:
* editing/EditingStyle.cpp:
(WebCore::EditingStyle::hasStyle):

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

* editing/EditingStyle.h:
* editing/Editor.cpp:
(WebCore::Editor::selectionStartHasStyle const):

Source/WebKit:

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

* UIProcess/API/Cocoa/WKUIDelegatePrivate.h:
* UIProcess/API/Cocoa/WKWebView.mm:
(nsTextAlignment):
(dictionaryRepresentationForEditorState):
(-[WKWebView _didChangeEditorState]):

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

(-[WKWebView _web_editorStateDidChange]):
(-[WKWebView _executeEditCommand:argument:completion:]):
* UIProcess/API/Cocoa/WKWebViewInternal.h:
* UIProcess/API/Cocoa/WKWebViewPrivate.h:
* UIProcess/API/mac/WKView.mm:
(-[WKView _web_editorStateDidChange]):
* UIProcess/Cocoa/WebViewImpl.h:
* UIProcess/Cocoa/WebViewImpl.mm:
(WebKit::WebViewImpl::selectionDidChange):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::executeEditCommand):

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

* UIProcess/WebPageProxy.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView executeEditCommandWithCallback:]):
(-[WKContentView _selectionChanged]):
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::executeEditCommand): Deleted.

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

* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::editorState const):

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

(WebKit::shouldEnsureEditorStateUpdateAfterExecutingCommand):
(WebKit::WebPage::executeEditCommandWithCallback):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::executeEditCommandWithCallback): Deleted.

Tools:

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

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

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

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

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

(-[EditingTestHarness latestEditorStateContains:]):
(-[EditingTestHarness _webView:editorStateDidChange:]):
* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/Tests/WebKit2Cocoa/EditorStateTests.mm: Added.
(TestWebKitAPI::setUpEditorStateTestHarness):
(TestWebKitAPI::TEST):
* TestWebKitAPI/Tests/WebKit2Cocoa/editor-state-test-harness.html: Added.

LayoutTests:

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

* platform/ios-wk2/editing/inserting/insert-div-024-expected.txt:
* platform/ios-wk2/editing/inserting/insert-div-026-expected.txt:
* platform/ios-wk2/editing/style/5084241-expected.txt:
* platform/ios-wk2/editing/style/unbold-in-bold-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220393 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[TexMap] Add TextureMapperContextAttributes
zandobersek@gmail.com [Tue, 8 Aug 2017 07:46:42 +0000 (07:46 +0000)]
[TexMap] Add TextureMapperContextAttributes
https://bugs.webkit.org/show_bug.cgi?id=175311

Reviewed by Carlos Garcia Campos.

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

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

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

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

No new tests -- no change in behavior.

* platform/TextureMapper.cmake:
* platform/graphics/texmap/BitmapTextureGL.cpp:
(WebCore::BitmapTextureGL::BitmapTextureGL):
(WebCore::BitmapTextureGL::updateContentsNoSwizzle):
(WebCore::BitmapTextureGL::updateContents):
(WebCore::driverSupportsSubImage): Deleted.
* platform/graphics/texmap/BitmapTextureGL.h:
(WebCore::BitmapTextureGL::create):
* platform/graphics/texmap/BitmapTexturePool.cpp:
(WebCore::BitmapTexturePool::BitmapTexturePool):
(WebCore::BitmapTexturePool::createTexture):
* platform/graphics/texmap/BitmapTexturePool.h:
* platform/graphics/texmap/TextureMapperContextAttributes.cpp: Added.
(WebCore::TextureMapperContextAttributes::initialize):
* platform/graphics/texmap/TextureMapperContextAttributes.h: Added.
* platform/graphics/texmap/TextureMapperGL.cpp:
(WebCore::TextureMapperGL::TextureMapperGL):
(WebCore::TextureMapperGL::drawTexture):
(WebCore::TextureMapperGL::drawTexturedQuadWithProgram):
(WebCore::TextureMapperGL::createTexture):
(WebCore::driverSupportsNPOTTextures): Deleted.
* platform/graphics/texmap/TextureMapperGL.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220392 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[TexMap] Don't expose GraphicsContext3D object
zandobersek@gmail.com [Tue, 8 Aug 2017 07:44:44 +0000 (07:44 +0000)]
[TexMap] Don't expose GraphicsContext3D object
https://bugs.webkit.org/show_bug.cgi?id=175310

Reviewed by Carlos Garcia Campos.

Source/WebCore:

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

* platform/graphics/texmap/TextureMapper.h:
* platform/graphics/texmap/TextureMapperGL.cpp:
(WebCore::TextureMapperGL::clearColor):
* platform/graphics/texmap/TextureMapperGL.h:
(WebCore::TextureMapperGL::graphicsContext3D const): Deleted.

Source/WebKit:

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

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220391 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoWebDriver: implement unhandled prompt behavior
carlosgc@webkit.org [Tue, 8 Aug 2017 06:43:25 +0000 (06:43 +0000)]
WebDriver: implement unhandled prompt behavior
https://bugs.webkit.org/show_bug.cgi?id=175184

Reviewed by Brian Burg.

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

* Capabilities.h: Add UnhandledPromptBehavior capability.
* CommandResult.cpp:
(WebDriver::CommandResult::httpStatusCode const): Add UnexpectedAlertOpen error.
(WebDriver::CommandResult::errorString const): Ditto.
* CommandResult.h:
(WebDriver::CommandResult::setAdditonalErrorData): New method to set an additional data object that will be sent
as part of the result error message.
(WebDriver::CommandResult::additionalErrorData const): Return the additional data object.
* Session.cpp:
(WebDriver::Session::handleUserPrompts): Check if there's an active JavaScript dialog and deal with it depeding
on the unhandled prompt behavior.
(WebDriver::Session::reportUnexpectedAlertOpen): Generate an error message with UnexpectedAlertOpen error and
including the alert text as additional error data.
(WebDriver::Session::go): Handle user prompts before running the command.
(WebDriver::Session::getCurrentURL): Ditto.
(WebDriver::Session::back): Ditto.
(WebDriver::Session::forward): Ditto.
(WebDriver::Session::refresh): Ditto.
(WebDriver::Session::getTitle): Ditto.
(WebDriver::Session::closeWindow): Ditto.
(WebDriver::Session::switchToFrame): Ditto.
(WebDriver::Session::switchToParentFrame): Ditto.
(WebDriver::Session::isElementSelected): Ditto.
(WebDriver::Session::getElementText): Ditto.
(WebDriver::Session::getElementTagName): Ditto.
(WebDriver::Session::getElementRect): Ditto.
(WebDriver::Session::isElementEnabled): Ditto.
(WebDriver::Session::isElementDisplayed): Ditto.
(WebDriver::Session::getElementAttribute): Ditto.
(WebDriver::Session::elementSendKeys): Ditto.
(WebDriver::Session::elementSubmit): Ditto.
(WebDriver::Session::executeScript): Ditto.
* Session.h:
* WebDriverService.cpp:
(WebDriver::WebDriverService::sendResponse const): Send data object as part of the result error message if present.
(WebDriver::deserializeUnhandledPromptBehavior):
(WebDriver::WebDriverService::parseCapabilities const):
(WebDriver::WebDriverService::validatedCapabilities const):
(WebDriver::WebDriverService::newSession):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220388 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years ago[GTK] Implement JavaScript dialog methods of API::AutomationSessionClient
carlosgc@webkit.org [Tue, 8 Aug 2017 06:25:23 +0000 (06:25 +0000)]
[GTK] Implement JavaScript dialog methods of API::AutomationSessionClient
https://bugs.webkit.org/show_bug.cgi?id=175259

Reviewed by Michael Catanzaro.

Source/WebCore/platform/gtk/po:

* POTFILES.in: Add WebKitScriptDialogGtk.cpp.

Source/WebKit:

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

* PlatformGTK.cmake: Add new files to compilation.
* PlatformWPE.cmake: Ditto.
* UIProcess/API/glib/WebKitAutomationSession.cpp:
(webkitAutomationSessionCreate): Pass the WebKitWebContext to the constructor and keep a pointer to it in session.
* UIProcess/API/glib/WebKitAutomationSessionPrivate.h:
* UIProcess/API/glib/WebKitScriptDialogPrivate.h:
* UIProcess/API/glib/WebKitWebContext.cpp:
* UIProcess/API/glib/WebKitWebView.cpp:
(webkitWebViewRunJavaScriptAlert): Set the currently script dialog for the scope of the function.
(webkitWebViewRunJavaScriptConfirm): Ditto.
(webkitWebViewRunJavaScriptPrompt): Ditto.
(webkitWebViewRunJavaScriptBeforeUnloadConfirm): Ditto.
(webkitWebViewIsShowingScriptDialog): Ask current dialog if there's one.
(webkitWebViewGetCurrentScriptDialogMessage): Ditto.
(webkitWebViewSetCurrentScriptDialogUserInput): Ditto.
(webkitWebViewAcceptCurrentScriptDialog): Ditto.
(webkitWebViewDismissCurrentScriptDialog): Ditto.
* UIProcess/API/glib/WebKitWebViewPrivate.h:
* UIProcess/API/gtk/WebKitScriptDialogGtk.cpp: Added.
(webkitWebViewCreateJavaScriptDialog): Moved from WebKitWebViewGtk.cpp.
(webkitScriptDialogRun): Run the dialog, this code is moved from WebKitWebViewGtk.cpp.
(webkitScriptDialogIsRunning): Return true if the script dialog has a native dialog running.
(webkitScriptDialogAccept): Send Ok or Close response to the native dialog depending on the dialog type.
(webkitScriptDialogDismiss): Send Close response to the native dialog.
(webkitScriptDialogSetUserInput): Set the given text on the prompt dialog entry.
* UIProcess/API/gtk/WebKitWebViewGtk.cpp:
(webkitWebViewScriptDialog): Simply call webkitScriptDialogRun() now.
* UIProcess/API/wpe/WebKitScriptDialogWPE.cpp: Copied from Source/WebKit/UIProcess/API/glib/WebKitAutomationSessionPrivate.h.
(webkitScriptDialogRun):
(webkitScriptDialogIsRunning):
(webkitScriptDialogAccept):
(webkitScriptDialogDismiss):
(webkitScriptDialogSetUserInput):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220387 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoWebDriver: implement user prompt commands
carlosgc@webkit.org [Tue, 8 Aug 2017 06:22:02 +0000 (06:22 +0000)]
WebDriver: implement user prompt commands
https://bugs.webkit.org/show_bug.cgi?id=174614

Reviewed by Brian Burg.

* CommandResult.cpp:
(WebDriver::CommandResult::CommandResult): Handle NoJavaScriptDialog protocol error.
(WebDriver::CommandResult::httpStatusCode const): Add NoSuchAlert.
(WebDriver::CommandResult::errorString const): Ditto.
* CommandResult.h:
* Session.cpp:
(WebDriver::Session::dismissAlert):
(WebDriver::Session::acceptAlert):
(WebDriver::Session::getAlertText):
(WebDriver::Session::sendAlertText):
* Session.h:
* WebDriverService.cpp:
(WebDriver::WebDriverService::dismissAlert):
(WebDriver::WebDriverService::acceptAlert):
(WebDriver::WebDriverService::getAlertText):
(WebDriver::WebDriverService::sendAlertText):
* WebDriverService.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220386 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoUse more accurate Russian keywords for RexBench/FlightPlanner Unicode mode
msaboff@apple.com [Tue, 8 Aug 2017 03:20:19 +0000 (03:20 +0000)]
Use more accurate Russian keywords for RexBench/FlightPlanner Unicode mode
https://bugs.webkit.org/show_bug.cgi?id=175298

Reviewed by Saam Barati.

With the help of Alexey Proskuryakov updated the Russian keywords.

* RexBench/FlightPlanner/use_unicode.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220384 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoRenderStyle:diff() was inadvertently doing deep compares of StyleRareNonInheritedData etc
simon.fraser@apple.com [Tue, 8 Aug 2017 02:52:10 +0000 (02:52 +0000)]
RenderStyle:diff() was inadvertently doing deep compares of StyleRareNonInheritedData etc
https://bugs.webkit.org/show_bug.cgi?id=175304

Reviewed by Tim Horton.

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

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

* rendering/style/DataRef.h:
(WebCore::DataRef::ptr const):
* rendering/style/RenderStyle.cpp:
(WebCore::RenderStyle::inheritedDataShared const):
(WebCore::RenderStyle::changeAffectsVisualOverflow const):
(WebCore::RenderStyle::changeRequiresLayout const):
(WebCore::RenderStyle::changeRequiresRecompositeLayer const):
(WebCore::RenderStyle::listStyleImage const): Expand the function onto multiple lines.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220383 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoAdd a fast path for rotate() and rotateZ() transform parsing
simon.fraser@apple.com [Tue, 8 Aug 2017 02:52:08 +0000 (02:52 +0000)]
Add a fast path for rotate() and rotateZ() transform parsing
https://bugs.webkit.org/show_bug.cgi?id=175308

Reviewed by Zalan Bujtas.

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

* css/parser/CSSParserFastPaths.cpp:
(WebCore::parseSimpleAngle):
(WebCore::parseTransformAngleArgument):
(WebCore::parseSimpleTransformValue):
(WebCore::transformCanLikelyUseFastPath):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220382 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoRe-order the tests in RenderLayerCompositor::requiresCompositingLayer() for performance
simon.fraser@apple.com [Tue, 8 Aug 2017 02:52:06 +0000 (02:52 +0000)]
Re-order the tests in RenderLayerCompositor::requiresCompositingLayer() for performance
https://bugs.webkit.org/show_bug.cgi?id=175306

Reviewed by Tim Horton.

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

* rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::requiresCompositingLayer const):
(WebCore::RenderLayerCompositor::requiresOwnBackingStore const):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220381 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoAvoid a hash lookup in FilterInfo::remove()
simon.fraser@apple.com [Tue, 8 Aug 2017 02:52:05 +0000 (02:52 +0000)]
Avoid a hash lookup in FilterInfo::remove()
https://bugs.webkit.org/show_bug.cgi?id=175301

Reviewed by Sam Weinig.

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

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

* rendering/RenderLayerFilterInfo.cpp:
(WebCore::RenderLayer::FilterInfo::getIfExists):
(WebCore::RenderLayer::FilterInfo::remove):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220380 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoMake TransformOperation::type() non-virtual
simon.fraser@apple.com [Tue, 8 Aug 2017 02:46:11 +0000 (02:46 +0000)]
Make TransformOperation::type() non-virtual
https://bugs.webkit.org/show_bug.cgi?id=175297

Reviewed by Sam Weinig.

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

Small perf win on some benchmarks.

* platform/graphics/transforms/IdentityTransformOperation.h:
* platform/graphics/transforms/Matrix3DTransformOperation.h:
* platform/graphics/transforms/MatrixTransformOperation.h:
* platform/graphics/transforms/PerspectiveTransformOperation.h:
* platform/graphics/transforms/RotateTransformOperation.cpp:
(WebCore::RotateTransformOperation::blend):
* platform/graphics/transforms/RotateTransformOperation.h:
* platform/graphics/transforms/ScaleTransformOperation.cpp:
(WebCore::ScaleTransformOperation::blend):
* platform/graphics/transforms/ScaleTransformOperation.h:
* platform/graphics/transforms/SkewTransformOperation.cpp:
(WebCore::SkewTransformOperation::blend):
* platform/graphics/transforms/SkewTransformOperation.h:
* platform/graphics/transforms/TransformOperation.h:
(WebCore::TransformOperation::TransformOperation):
(WebCore::TransformOperation::type const):
(WebCore::TransformOperation::isSameType const):
* platform/graphics/transforms/TranslateTransformOperation.cpp:
(WebCore::TranslateTransformOperation::blend):
* platform/graphics/transforms/TranslateTransformOperation.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220379 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoInline ~Color and Color::isExtended()
simon.fraser@apple.com [Tue, 8 Aug 2017 02:46:06 +0000 (02:46 +0000)]
Inline ~Color and Color::isExtended()
https://bugs.webkit.org/show_bug.cgi?id=175293

Reviewed by Zalan Bujtas.

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

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

* platform/LengthSize.h:
(WebCore::operator==):
* platform/graphics/Color.cpp:
(WebCore::Color::~Color): Deleted.
(WebCore::Color::isExtended const): Deleted.
* platform/graphics/Color.h:
(WebCore::Color::~Color):
(WebCore::Color::isExtended const):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220378 268f45cc-cd09-0410-ab3c-d52691b4dbfc

3 years agoGetOwnProperty of TypedArray indexed fields is wrongly configurable
commit-queue@webkit.org [Tue, 8 Aug 2017 02:29:42 +0000 (02:29 +0000)]
GetOwnProperty of TypedArray indexed fields is wrongly configurable
https://bugs.webkit.org/show_bug.cgi?id=175307

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

JSTests:

* stress/typedarray-getownproperty-not-configurable.js: Added.
(assert):
(foo):

Source/JavaScriptCore:

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

* runtime/JSGenericTypedArrayViewInlines.h:
(JSC::JSGenericTypedArrayView<Adaptor>::getOwnPropertySlotByIndex):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@220377 268f45cc-cd09-0410-ab3c-d52691b4dbfc