WebKit-https.git
5 years ago[iOS] run-webkit-tests should check that simctl can boot and shutdown simulator devic...
dbates@webkit.org [Tue, 17 Feb 2015 22:30:07 +0000 (22:30 +0000)]
[iOS] run-webkit-tests should check that simctl can boot and shutdown simulator device before running tests
https://bugs.webkit.org/show_bug.cgi?id=141718

Reviewed by Alex Christensen.

We should only run layout tests if simctl can successfully boot and shutdown the testing device.

* Scripts/webkitpy/port/ios.py:
(IOSSimulatorPort.check_sys_deps): Modified to call Simulator.check_simulator_device_and_erase_if_needed().
Log a error and return False if we are unable to boot the simulator device so that the caller can take
appropriate action, say exit(3) before running any layout tests.
* Scripts/webkitpy/xcode/simulator.py:
(Simulator._boot_and_shutdown_simulator_device): Added. Boot and then shut down the simulator device
with the specified UDID.
(Simulator.check_simulator_device_and_erase_if_needed): Added. Checks that simulator device
with the specified UDID can successfully boot and shut down. We make at most two attempts to
boot and shut down the device, erasing the device between tries so as to restore the device
to a known good state.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180239 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agorun_webkit_tests.py should not know about iOS Simulator details
dbates@webkit.org [Tue, 17 Feb 2015 22:13:11 +0000 (22:13 +0000)]
run_webkit_tests.py should not know about iOS Simulator details
https://bugs.webkit.org/show_bug.cgi?id=141711

Reviewed by Alex Christensen.

Run_webkit_tests.py should only parse the iOS-specific command
line options --runtime and --device-type. Let the port-specific
logic in ios.py validate the parsed options and instantiate
internal data structures.

Currently run_webkit_tests.py parses the iOS-specific command
line options --runtime and --device-type, validates them, and
instantiates internal data structures from the parsed strings.
Instead the validation logic and instantiation of internal
data structures should be handled by the iOS port object.

* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
(_set_up_derived_options): Remove logic to validate --runtime and --device-type
and instantiate internal data structures for them.
* Scripts/webkitpy/port/driver.py:
(IOSSimulatorDriver.cmd_line): Modified to reference IOSSimulatorPort.simulator_runtime
and IOSSimulatorPort.simulator_device_type for the iOS Simulator runtime and device type,
respectively.
* Scripts/webkitpy/port/ios.py: Sort the list of imports.
(IOSSimulatorPort.simulator_runtime): Added. Instantiates a Runtime
object from --runtime, if specified. Otherwise, instantiates a Runtime
object for the latest installed iphonesimulator SDK version.
(IOSSimulatorPort.simulator_device_type): Added. Instantiates a DeviceType
object from --device-type, if specified. Otherwise, instantiates
a DeviceType object for a iPhone 5 or iPhone 5s when on a 32-bit and 64-bit
machine, respectively.
(IOSSimulatorPort.check_sys_deps): Added. Validate if the chosen iOS simulator
runtime is available to use.
(IOSSimulatorPort.testing_device): Modified to make use of properties simulator_device_type
and simulator_runtime for the iOS Simulator device type and runtime, respectively.
* Scripts/webkitpy/xcode/simulator.py:
(Runtime.from_version_string): Added. Turns around and calls Runtime.from_identifier()
with a runtime identifier for the specified iOS version.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180238 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoStackLayoutPhase should use CodeBlock::usesArguments rather than FunctionExecutable...
fpizlo@apple.com [Tue, 17 Feb 2015 21:41:25 +0000 (21:41 +0000)]
StackLayoutPhase should use CodeBlock::usesArguments rather than FunctionExecutable::usesArguments
https://bugs.webkit.org/show_bug.cgi?id=141721
rdar://problem/17198633

Reviewed by Michael Saboff.

I've seen cases where the two are out of sync.  We know we can trust the CodeBlock::usesArguments because
we use it everywhere else.

No test because I could never reproduce the crash.

* dfg/DFGGraph.h:
(JSC::DFG::Graph::usesArguments):
* dfg/DFGStackLayoutPhase.cpp:
(JSC::DFG::StackLayoutPhase::run):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180237 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoWeb Inspector: Improved Console Support for Bound Functions
joepeck@webkit.org [Tue, 17 Feb 2015 20:17:28 +0000 (20:17 +0000)]
Web Inspector: Improved Console Support for Bound Functions
https://bugs.webkit.org/show_bug.cgi?id=141635

Reviewed by Timothy Hatcher.

Source/JavaScriptCore:

* inspector/JSInjectedScriptHost.cpp:
(Inspector::JSInjectedScriptHost::getInternalProperties):
Expose internal properties of a JSBoundFunction.

LayoutTests:

* inspector/model/remote-object-get-properties-expected.txt:
* inspector/model/remote-object-get-properties.html:
Show that boundFunction has extra properties (targetFunction, boundThis, boundArgs).

* inspector/model/remote-object-expected.txt:
* inspector/model/remote-object.html:
Include a bound function. Much different from a function now though.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180236 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoWeb Inspector: ES6: Improved Console Support for Promise Objects
joepeck@webkit.org [Tue, 17 Feb 2015 20:17:16 +0000 (20:17 +0000)]
Web Inspector: ES6: Improved Console Support for Promise Objects
https://bugs.webkit.org/show_bug.cgi?id=141634

Reviewed by Timothy Hatcher.

Source/JavaScriptCore:

* inspector/InjectedScript.cpp:
(Inspector::InjectedScript::getInternalProperties):
* inspector/InjectedScriptSource.js:
Include internal properties in previews. Share code
with normal internal property handling.

* inspector/JSInjectedScriptHost.cpp:
(Inspector::constructInternalProperty):
(Inspector::JSInjectedScriptHost::getInternalProperties):
Provide internal state of Promises.

* inspector/protocol/Runtime.json:
Provide an optional field to distinguish if a PropertyPreview
is for an Internal property or not.

Source/WebInspectorUI:

* UserInterface/Protocol/RemoteObject.js:
For now just add Internal Properties and regular properties.
We will address the UI later.

LayoutTests:

* inspector/model/remote-object-expected.txt:
* inspector/model/remote-object.html:
Include more tests for Promises.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180235 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoThrowing from an FTL call IC slow path may result in tag registers being clobbered...
fpizlo@apple.com [Tue, 17 Feb 2015 20:10:00 +0000 (20:10 +0000)]
Throwing from an FTL call IC slow path may result in tag registers being clobbered on 64-bit CPUs
https://bugs.webkit.org/show_bug.cgi?id=141717
rdar://problem/19863382

Reviewed by Geoffrey Garen.

The best solution is to ensure that the engine catching an exception restores tag registers.

Each of these new test cases reliably crashed prior to this patch and they don't crash at all now.

* jit/JITOpcodes.cpp:
(JSC::JIT::emit_op_catch):
* llint/LowLevelInterpreter.asm:
* llint/LowLevelInterpreter64.asm:
* tests/stress/throw-from-ftl-call-ic-slow-path-cells.js: Added.
* tests/stress/throw-from-ftl-call-ic-slow-path-undefined.js: Added.
* tests/stress/throw-from-ftl-call-ic-slow-path.js: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180234 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUnreviewed GTK Gardening 17th February.
commit-queue@webkit.org [Tue, 17 Feb 2015 19:47:11 +0000 (19:47 +0000)]
Unreviewed GTK Gardening 17th February.

https://bugs.webkit.org/show_bug.cgi?id=141712

Patch by Marcos Chavarría Teijeiro <chavarria1991@gmail.com> on 2015-02-17

* platform/gtk/TestExpectations:
* platform/gtk/accessibility/press-targets-center-point-expected.txt: Added. Rebaselined after r180107.
* platform/gtk/svg/custom/glyph-selection-non-bmp-expected.txt: Added. Rebaselined after r177774.
* platform/gtk/svg/text/textPathBoundsBug-expected.txt: Added. Rebaselined after r177774.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180233 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[ARM] Add the necessary setupArgumentsWithExecState after bug141332
ossy@webkit.org [Tue, 17 Feb 2015 19:09:53 +0000 (19:09 +0000)]
[ARM] Add the necessary setupArgumentsWithExecState after bug141332
https://bugs.webkit.org/show_bug.cgi?id=141714

Reviewed by Michael Saboff.

* jit/CCallHelpers.h:
(JSC::CCallHelpers::setupArgumentsWithExecState):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180232 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoREGRESSION: Video control fails to hide after playback begins in reflowable EPUB...
dino@apple.com [Tue, 17 Feb 2015 19:07:03 +0000 (19:07 +0000)]
REGRESSION: Video control fails to hide after playback begins in reflowable EPUB files
https://bugs.webkit.org/show_bug.cgi?id=141689
<rdar://problem/19689286>

Reviewed by Eric Carlson.

The iBooks app turns off the requirement for a user gesture
before triggering playback. When we moved to script-based
media controls there were a few regressions in this setup.
This makes them behave a lot more like iOS 7.

* Modules/mediacontrols/mediaControlsApple.js:
(Controller.prototype.setPlaying): We don't want to hide the controls
yet, just start a hiding timer.
* Modules/mediacontrols/mediaControlsiOS.js:
(ControllerIOS.prototype.shouldHaveStartPlaybackButton): If we've ever
played before, we don't want to show the big start button. If we
are in the middle of construction, we do want to show it. We no
longer care about the case where you're not in setup and have
the relaxed restrictions (we still want to show the start button
in that case).
(ControllerIOS.prototype.showControls): When we show the controls,
make sure our timeline is up to date. This is necessary for the
cases where we automatically show the controls, such as when we hit
the end of a video.
(ControllerIOS.prototype.updateTime): Make sure to pass the forceUpdate
parameter on to the super-method.
(ControllerIOS.prototype.setPlaying):
(ControllerIOS.prototype.progressFillStyle): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180231 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[Win] Document some accessibility and animation debug assertions (Unreviewed)
bfulgham@apple.com [Tue, 17 Feb 2015 18:47:55 +0000 (18:47 +0000)]
[Win] Document some accessibility and animation debug assertions (Unreviewed)

* platform/win/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180229 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoOnly create the WebKit.framework symlink in PrivateFrameworks when the platform is...
andersca@apple.com [Tue, 17 Feb 2015 18:42:55 +0000 (18:42 +0000)]
Only create the WebKit.framework symlink in PrivateFrameworks when the platform is iphoneos
https://bugs.webkit.org/show_bug.cgi?id=141710
<rdar://problem/19719748>

Reviewed by Andreas Kling.

Don't create the symlink when building for the simulator.

* WebKit2.xcodeproj/project.pbxproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180228 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoASan does not like JSC::MachineThreads::tryCopyOtherThreadStack
dburkart@apple.com [Tue, 17 Feb 2015 18:06:25 +0000 (18:06 +0000)]
ASan does not like JSC::MachineThreads::tryCopyOtherThreadStack
https://bugs.webkit.org/show_bug.cgi?id=141672

Reviewed by David Kilzer.

* asan/webkit-asan-ignore.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180227 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoAccess MemoryPressureHandler global instance via a singleton() static member function
cdumez@apple.com [Tue, 17 Feb 2015 17:30:34 +0000 (17:30 +0000)]
Access MemoryPressureHandler global instance via a singleton() static member function
https://bugs.webkit.org/show_bug.cgi?id=141691

Reviewed by Andreas Kling.

Access MemoryPressureHandler global instance via a singleton() static
Source/WebCore:

member function as per coding style. Also make all other member
functions non-static as callers can just use singleton() to get the
instance and access methods. This avoid having to call
MemoryPressureHandler::singleton() from member functions.

* bindings/js/ScriptController.cpp:
(WebCore::collectGarbageAfterWindowShellDestruction):
* history/PageCache.cpp:
(WebCore::PageCache::canCache):
* loader/FrameLoader.cpp:
(WebCore::FrameLoader::commitProvisionalLoad):
* page/FrameView.cpp:
(WebCore::FrameView::willPaintContents):
(WebCore::FrameView::didPaintContents):
(WebCore::FrameView::computeCoverageRect):
* platform/MemoryPressureHandler.cpp:
(WebCore::MemoryPressureHandler::singleton):
(WebCore::MemoryPressureHandler::MemoryPressureHandler):
(WebCore::MemoryPressureHandler::releaseCriticalMemory):
(WebCore::memoryPressureHandler): Deleted.
* platform/MemoryPressureHandler.h:
(WebCore::MemoryPressureHandler::setLowMemoryHandler):
* platform/cocoa/MemoryPressureHandlerCocoa.mm:
(WebCore::MemoryPressureHandler::platformReleaseMemory):
(WebCore::MemoryPressureHandler::install):
(WebCore::MemoryPressureHandler::holdOff):
(WebCore::respondToMemoryPressureCallback):
* platform/graphics/FontCache.cpp:
(WebCore::FontCache::purgeInactiveFontDataIfNeeded):
* platform/ios/LegacyTileCache.mm:
(WebCore::LegacyTileCache::createTilesInActiveGrid):
* platform/ios/LegacyTileGrid.mm:
(WebCore::LegacyTileGrid::shouldUseMinimalTileCoverage):
* platform/ios/LegacyTileLayerPool.mm:
(WebCore::LegacyTileLayerPool::addLayer):
* platform/ios/TileControllerMemoryHandlerIOS.cpp:
(WebCore::TileControllerMemoryHandler::tileControllerGainedUnparentedTiles):
* platform/linux/MemoryPressureHandlerLinux.cpp:
(WebCore::MemoryPressureHandler::waitForMemoryPressureEvent):

Source/WebKit/mac:

member function.

* WebView/WebView.mm:
(-[WebView _commonInitializationWithFrameName:groupName:]):
(+[WebView registerForMemoryNotifications]):
(+[WebView _isUnderMemoryPressure]):
(+[WebView _clearMemoryPressure]):
(+[WebView _shouldWaitForMemoryClearMessage]):
(WebInstallMemoryPressureHandler):

Source/WebKit2:

member function.

* NetworkProcess/NetworkProcess.cpp:
(WebKit::NetworkProcess::initializeNetworkProcess):
(WebKit::NetworkProcess::lowMemoryHandler): Deleted.
* NetworkProcess/NetworkProcess.h:
* PluginProcess/PluginProcess.cpp:
(WebKit::PluginProcess::initializeProcess):
(WebKit::PluginProcess::lowMemoryHandler): Deleted.
* PluginProcess/PluginProcess.h:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::updateVisibleContentRects):
* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::initializeWebProcess):
(WebKit::WebProcess::processWillSuspend):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180225 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoRemove WebCore.exp.in and clean up.
achristensen@apple.com [Tue, 17 Feb 2015 17:23:54 +0000 (17:23 +0000)]
Remove WebCore.exp.in and clean up.
https://bugs.webkit.org/show_bug.cgi?id=141491

Reviewed by Andreas Kling.

Source/WebCore:

* Configurations/WebCore.xcconfig:
Don't use exported symbols files.
* DerivedSources.make:
Don't generate export files.
* WebCore.exp.in: Removed.
* WebCore.xcodeproj/project.pbxproj:
Removed WebCoreExportFileGenerator targets.
* make-export-file-generator: Removed.

Tools:

* Scripts/sort-export-file: Removed.
* Scripts/webkitpy/style/checker.py:
(_all_categories):
(FileType):
(CheckerDispatcher._file_type):
(CheckerDispatcher._create_checker):
Don't check if the export files are sorted.
* Scripts/webkitpy/style/checkers/exportfile.py: Removed.
* Scripts/webkitpy/style/checkers/exportfile_unittest.py: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180224 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoDocument some debug assertions in Accessibility tests (Unreviewed)
bfulgham@apple.com [Tue, 17 Feb 2015 17:09:45 +0000 (17:09 +0000)]
Document some debug assertions in Accessibility tests (Unreviewed)

* platform/win/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180223 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUnreviewed. Fix GTK+ make distcheck.
carlosgc@webkit.org [Tue, 17 Feb 2015 16:38:38 +0000 (16:38 +0000)]
Unreviewed. Fix GTK+ make distcheck.

Do not exclude bmalloc directory from the tarball.

* gtk/manifest.txt.in:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180221 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[CMAKE] Remove CMakeLists.txt in WK1 port
gyuyoung.kim@samsung.com [Tue, 17 Feb 2015 16:31:00 +0000 (16:31 +0000)]
[CMAKE] Remove CMakeLists.txt in WK1 port
https://bugs.webkit.org/show_bug.cgi?id=141617

Reviewed by Anders Carlsson.

.:

Nobody uses CMake in WK1 port. Remove it.

* CMakeLists.txt:

Source/WebKit:

* CMakeLists.txt: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180220 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUpdate svg/animations/animate-text-nested-transforms.html after r177774
mmaxfield@apple.com [Tue, 17 Feb 2015 16:03:06 +0000 (16:03 +0000)]
Update svg/animations/animate-text-nested-transforms.html after r177774
https://bugs.webkit.org/show_bug.cgi?id=141686

Reviewed by Andreas Kling.

Text metrics are expected to change.

* platform/mac/TestExpectations:
* svg/animations/animate-text-nested-transforms-expected.txt:
* svg/animations/script-tests/animate-text-nested-transforms.js:
(text.appendChild.rootSVGElement.appendChild.startSample):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180219 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUpdate expected result of fast/forms/cursor-at-editable-content-boundary.html after...
mmaxfield@apple.com [Tue, 17 Feb 2015 16:01:00 +0000 (16:01 +0000)]
Update expected result of fast/forms/cursor-at-editable-content-boundary.html after r177774
https://bugs.webkit.org/show_bug.cgi?id=141684

Reviewed by Andreas Kling.

One check no longer fails, and another fails in the same way it did before (but with 1-different
measurement values)

* platform/mac/TestExpectations:
* platform/mac/fast/forms/cursor-at-editable-content-boundary-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180218 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUse HashMap::add instead of get/contains + set in DOMObjectCache
carlosgc@webkit.org [Tue, 17 Feb 2015 11:57:43 +0000 (11:57 +0000)]
Use HashMap::add instead of get/contains + set in DOMObjectCache
https://bugs.webkit.org/show_bug.cgi?id=141558

Rubber-stamped by Žan Doberšek.

* bindings/gobject/DOMObjectCache.cpp:
(WebKit::getOrCreateDOMObjectCacheFrameObserver):
(WebKit::DOMObjectCache::put):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180216 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUnreviewed. Fix /webkit2/WebKitDOMNode/dom-cache after r180214.
carlosgc@webkit.org [Tue, 17 Feb 2015 10:33:57 +0000 (10:33 +0000)]
Unreviewed. Fix /webkit2/WebKitDOMNode/dom-cache after r180214.

I forgot to add the return of a bool function.

* TestWebKitAPI/Tests/WebKit2Gtk/DOMNodeTest.cpp:
(WebKitDOMNodeTest::testDOMCache):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180215 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[GTK] GObject DOM bindings object are cached forever
carlosgc@webkit.org [Tue, 17 Feb 2015 09:57:42 +0000 (09:57 +0000)]
[GTK] GObject DOM bindings object are cached forever
https://bugs.webkit.org/show_bug.cgi?id=141558

Reviewed by Sergio Villar Senin.

Source/WebCore:

Rework the DOMObjectCache to avoid having to manually clear the
objects when the frame is destroyed, using a FrameDestructionObserver
instead. When a new object associated to a Frame is added to the
cache, a FrameDestructionObserver is created for the frame if
needed, and the DOM object is tracked by the observer too. When
the frame is detached from the page all its objects are cleared,
and if the aren't additional references, the object is finalized
and removed from the cache normally.
This patch also simplifies and modernizes the code to make it
easier to read an maintain.

* bindings/gobject/DOMObjectCache.cpp:
(WebKit::DOMObjectCacheData::DOMObjectCacheData): Add constructor
to initialize its members and simplify the callers.
(WebKit::DOMObjectCacheData::clearObject): Remove the references
added by the cache, ensuring the GObject is not finalized until
the method returns.
(WebKit::DOMObjectCacheData::refObject): Adds a reference owned by
the cache.
(WebKit::domObjectCacheFrameObservers): Map a frame to a FrameDestructionObserver.
(WebKit::getOrCreateDOMObjectCacheFrameObserver): Create a new
FrameDestructionObserver for the given Frame or return the
existing one.
(WebKit::domObjects): Map wrapped object to wrapper object.
(WebKit::DOMObjectCache::forget): Remove the wrapped object from
the cache.
(WebKit::DOMObjectCache::get): Return the wrapped object if it is
in the cache, increasing the cache references.
(WebKit::DOMObjectCache::put): Add the wrapper object to the cache
for the given wrapped object. If it's a Node and has a frame add
the node to the FrameDestructionObserver corresponding to the frame.
(WebKit::getFrameFromHandle): Deleted.
(WebKit::weakRefNotify): Deleted.
(WebKit::DOMObjectCache::clearByFrame): Deleted.
(WebKit::DOMObjectCache::~DOMObjectCache): Deleted.
* bindings/gobject/DOMObjectCache.h:

Tools:

Add checks for all DOM objects to ensure they are not leaked. Also
add a dedicated test for the DOM Object Cache.

* TestWebKitAPI/Tests/WebKit2Gtk/DOMNodeFilterTest.cpp:
(WebKitDOMNodeFilterTest::testTreeWalker):
(WebKitDOMNodeFilterTest::testNodeIterator):
* TestWebKitAPI/Tests/WebKit2Gtk/DOMNodeTest.cpp:
(WebKitDOMNodeTest::testHierarchyNavigation):
(WebKitDOMNodeTest::testInsertion):
(WebKitDOMNodeTest::testTagNames):
(WebKitDOMNodeTest::testDOMCache):
(registerTests):
* TestWebKitAPI/Tests/WebKit2Gtk/DOMXPathNSResolverTest.cpp:
(WebKitDOMXPathNSResolverTest::evaluateFooChildTextAndCheckResult):
(WebKitDOMXPathNSResolverTest::testXPathNSResolverNative):
(WebKitDOMXPathNSResolverTest::testXPathNSResolverCustom):
* TestWebKitAPI/Tests/WebKit2Gtk/TestDOMNode.cpp:
(testWebKitDOMObjectCache):
(beforeAll):
* TestWebKitAPI/Tests/WebKit2Gtk/WebExtensionTest.cpp:
(documentLoadedCallback):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180214 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoREGRESSION(r180050): It broke the !ENABLE(CSS_GRID_LAYOUT) build
changseok.oh@collabora.com [Tue, 17 Feb 2015 09:42:13 +0000 (09:42 +0000)]
REGRESSION(r180050): It broke the !ENABLE(CSS_GRID_LAYOUT) build
https://bugs.webkit.org/show_bug.cgi?id=141647

Reviewed by Csaba Osztrogonác.

No new tests since this just fixes a build break.

* dom/Position.cpp: Add a build guard ENABLE(CSS_GRID_LAYOUT) for RenderGrid.
(WebCore::Position::isCandidate):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180213 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agofast/selectors/nth-last-child-of-register-requirement.html is extremely slow in debug...
ap@apple.com [Tue, 17 Feb 2015 08:58:41 +0000 (08:58 +0000)]
fast/selectors/nth-last-child-of-register-requirement.html is extremely slow in debug builds
https://bugs.webkit.org/show_bug.cgi?id=141695

* TestExpectations: Fixed a copy/paste mistake.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180212 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[GTK] WebKitFrame objects are never released
carlosgc@webkit.org [Tue, 17 Feb 2015 07:34:29 +0000 (07:34 +0000)]
[GTK] WebKitFrame objects are never released
https://bugs.webkit.org/show_bug.cgi?id=141641

Reviewed by Martin Robinson.

Source/WebKit2:

Use a FrameDestructionObserver derived class to wrap our
WebKitFrame objects and delete them when the frame is destroyed,
instead of using willDestroyFrame callback of WKBundlePageLoaderClient
that has never worked.

* WebProcess/InjectedBundle/API/gtk/WebKitFrame.cpp:
(webkitFrameGetWebFrame):
* WebProcess/InjectedBundle/API/gtk/WebKitFramePrivate.h:
* WebProcess/InjectedBundle/API/gtk/WebKitWebPage.cpp:
(webkitFrameGetOrCreate):
(webkitFrameDestroy):
(webkitWebPageCreate):
(willDestroyFrame): Deleted.

Tools:

Add a way to check GObjects leaks for WebProcess tests and check
WebKitFrame objects are not leaked.

* TestWebKitAPI/Tests/WebKit2Gtk/FrameTest.cpp:
(WebKitFrameTest::testMainFrame):
(WebKitFrameTest::testURI):
(WebKitFrameTest::testJavaScriptContext):
* TestWebKitAPI/Tests/WebKit2Gtk/WebProcessTest.cpp:
(WebProcessTest::assertObjectIsDeletedWhenTestFinishes):
(runTest):
* TestWebKitAPI/Tests/WebKit2Gtk/WebProcessTest.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180211 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoFix the !USE(IOSURFACE) build
timothy_horton@apple.com [Tue, 17 Feb 2015 07:25:13 +0000 (07:25 +0000)]
Fix the !USE(IOSURFACE) build

* Shared/mac/RemoteLayerBackingStore.mm:
SOFT_LINK_MAY_FAIL makes a function that will be unused if !USE(IOSURFACE).

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180210 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoREGRESSION(?): inspector/css/selector-dynamic-specificity.html sometimes crashes
ap@apple.com [Tue, 17 Feb 2015 07:17:19 +0000 (07:17 +0000)]
REGRESSION(?): inspector/css/selector-dynamic-specificity.html sometimes crashes
https://bugs.webkit.org/show_bug.cgi?id=141118

This happens on WK1 too.

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

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180209 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agofast/selectors/nth-last-child-of-register-requirement.html is extremely slow in debug...
ap@apple.com [Tue, 17 Feb 2015 06:42:39 +0000 (06:42 +0000)]
fast/selectors/nth-last-child-of-register-requirement.html is extremely slow in debug builds
https://bugs.webkit.org/show_bug.cgi?id=141695

* TestExpectations: Marked as slow.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180208 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoKeep all memory cache resources in ListHashSets
cdumez@apple.com [Tue, 17 Feb 2015 04:54:56 +0000 (04:54 +0000)]
Keep all memory cache resources in ListHashSets
https://bugs.webkit.org/show_bug.cgi?id=141667

Reviewed by Andreas Kling.

Keep all memory cache resources in ListHashSets instead of manual linked
lists. This simplifies the code a lot and is also more efficient for
retrieving / removing particular CachedResources.

* loader/cache/CachedResource.cpp:
(WebCore::CachedResource::CachedResource):
* loader/cache/CachedResource.h:
* loader/cache/MemoryCache.cpp:
(WebCore::MemoryCache::pruneDeadResourcesToSize):
(WebCore::MemoryCache::removeFromLRUList):
(WebCore::MemoryCache::insertInLRUList):
(WebCore::MemoryCache::dumpLRULists):
(WebCore::MemoryCache::lruListFor): Deleted.
* loader/cache/MemoryCache.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180207 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoCSS JIT: finish :nth-last-child()
benjamin@webkit.org [Tue, 17 Feb 2015 04:26:58 +0000 (04:26 +0000)]
CSS JIT: finish :nth-last-child()
https://bugs.webkit.org/show_bug.cgi?id=141629

Reviewed by Andreas Kling.

Source/WebCore:

This patch adds the matcher for :nth-child(An+B of selector list) and
fix a small bug I discovered while working on it.

The matcher is straightforward: count the next siblings matching the selector,
nothing fancy.

While working on it I noticed I forgot the test for IsParsingChildrenFinished
on the simple version of :nth-last-child(). I add it in both matcher, write
a couple of tests, the first part of the tests now succeed, but the second part
fails...

What happened is:
1) We interupt the parsing to execute the JavaScript.
   From there, we force the style resolution to get the computed style.
2) When resolving the style, the early check for isFinishedParsingChildren()
   quits the function early. This is done *before* we marked the parent
   for :nth-last-child() style resolution.
3) After the script, parsing resume and the following elements are added.
4) When resolving the style, only the new elements are marked dirty,
   the elements pending their :nth-last-child() style never get udpated.

To fix the problem, I moved the test for FinishedParsingChildren after
the parent marking.

Honestly, those early return for FinishedParsingChildren need to be refined
and they should be tested properly. We should not do this kind of things
for Query for example.

Tests: fast/selectors/nth-last-child-cannot-match-during-parsing-1.html
       fast/selectors/nth-last-child-cannot-match-during-parsing-2.html
       fast/selectors/nth-last-child-of-cannot-match-during-parsing-1.html
       fast/selectors/nth-last-child-of-cannot-match-during-parsing-2.html
       fast/selectors/nth-last-child-of-register-requirement.html

* css/SelectorChecker.cpp:
(WebCore::SelectorChecker::checkOne):
Fix the tree marking.

* cssjit/SelectorCompiler.cpp:
(WebCore::SelectorCompiler::addPseudoClassType):
The fragment creation was already done, all I had to do was
accept the compile.

(WebCore::SelectorCompiler::SelectorCodeGenerator::generateElementMatching):
:nth-child() and :nth-last-child() with a selector list are heavier than :not()
and :matches(), move them accordingly.

(WebCore::SelectorCompiler::SelectorCodeGenerator::generateElementIsNthLastChild):
More the early return after the tree marking.

(WebCore::SelectorCompiler::setParentAffectedByLastChildOf):
(WebCore::SelectorCompiler::SelectorCodeGenerator::generateElementIsNthLastChildOf):
New matcher, nothing optimized yet.

LayoutTests:

* fast/selectors/nth-last-child-cannot-match-during-parsing-1-expected.txt: Added.
* fast/selectors/nth-last-child-cannot-match-during-parsing-1.html: Added.
* fast/selectors/nth-last-child-cannot-match-during-parsing-2-expected.txt: Added.
* fast/selectors/nth-last-child-cannot-match-during-parsing-2.html: Added.
* fast/selectors/nth-last-child-of-cannot-match-during-parsing-1-expected.txt: Added.
* fast/selectors/nth-last-child-of-cannot-match-during-parsing-1.html: Added.
* fast/selectors/nth-last-child-of-cannot-match-during-parsing-2-expected.txt: Added.
* fast/selectors/nth-last-child-of-cannot-match-during-parsing-2.html: Added.
Those test cover the bugs related to IsParsingChildrenFinished and style update.

I had to duplicate the tests to have a simple selector because :nth-last-child(n)
and :nth-last-child(2n) are optimized differently.

* fast/selectors/nth-last-child-of-register-requirement-expected.txt: Added.
* fast/selectors/nth-last-child-of-register-requirement.html: Added.
The ususal register pressure tests. This one is taken directly from
the :nth-child() tests, I just changed the expected values.

* fast/selectors/several-nth-last-child.html:
Due to the changes with IsParsingChildrenFinished, this test can no longer
use body:nth-last-child() during parsing. I changed it to an async test.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180206 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoWeb Inspector: Tweak the styles on the Console
commit-queue@webkit.org [Tue, 17 Feb 2015 04:20:21 +0000 (04:20 +0000)]
Web Inspector: Tweak the styles on the Console
https://bugs.webkit.org/show_bug.cgi?id=140580

    - Use non-monospace (default sans-serif) font for console messages, but keep using monospace font
      for objects, functions, and anything else that presumably code.
    - Display all expandable objects on a separate row.

Patch by Nikita Vasilyev <nvasilyev@apple.com> on 2015-02-16
Reviewed by Timothy Hatcher.

* UserInterface/Views/ConsoleCommand.js:
(WebInspector.ConsoleCommand.prototype.decorateMessageElement):
(WebInspector.ConsoleCommand.prototype.toMessageElement):
* UserInterface/Views/ConsoleGroup.js:
(WebInspector.ConsoleGroup.prototype.addMessage):
* UserInterface/Views/ConsoleMessageImpl.js:
(WebInspector.ConsoleMessageImpl.prototype._format):
(WebInspector.ConsoleMessageImpl.prototype._isExpandable):
(WebInspector.ConsoleMessageImpl.prototype._formatParameter):
(WebInspector.ConsoleMessageImpl.prototype.append):
(WebInspector.ConsoleMessageImpl.prototype._formatWithSubstitutionString):
(WebInspector.ConsoleMessageImpl.prototype.decorateMessageElement):
(WebInspector.ConsoleMessageImpl.prototype.toMessageElement):
(WebInspector.ConsoleMessageImpl.prototype._formatParameterAsString): Deleted.
* UserInterface/Views/LogContentView.css:
(.console-messages):
(.console-formatted-object):
(.expandable.source-code):
(.console-message):
(.console-item.selected::after):
(.console-messages:focus .console-item.selected::after):
(.console-user-command-result.console-log-level::before):
(.console-item::before):
(.console-message .bubble):
(.console-error-level):
(.console-error-level .section .header .title):
(.console-error-level::before):
(.console-warning-level):
(.console-warning-level .console-message-text):
(.console-warning-level::before):
(.console-user-command::before):
(.console-group-messages .section):
(.console-group-messages .section .header::before):
(.console-group-messages .section .header .title):
(.console-formatted-object, .console-formatted-node, .console-formatted-error, .console-formatted-map, .console-formatted-set, .console-formatted-weakmap):
(.outline-disclosure li):
(.outline-disclosure li.parent::before):
(.console-item): Deleted.
(.console-item.selected): Deleted.
(.console-messages:focus .console-item.selected): Deleted.
(.console-item.selected .console-message::after): Deleted.
(.console-messages:focus .console-item.selected .data-grid tr.selected): Deleted.
(.console-messages:focus .console-item.selected .console-message::after): Deleted.
(.console-message, .console-user-command): Deleted.
(.console-message::before, .console-user-command::before, .javascript-prompt::before, .console-group-title::before): Deleted.
(.console-group): Deleted.
(.console-debug-level .console-message-text): Deleted.
(.console-debug-level::before): Deleted.
(.console-group-messages .section .header): Deleted.
(.console-group-messages .outline-disclosure): Deleted.
(.console-group-messages .outline-disclosure > ol): Deleted.
(.outline-disclosure li.elements-drag-over .selection): Deleted.
(.outline-disclosure ol:focus li.selected .selection): Deleted.
(.outline-disclosure > ol): Deleted.
* UserInterface/Views/LogContentView.js:
(WebInspector.LogContentView.prototype._mousemove):
(WebInspector.LogContentView.prototype._updateMessagesSelection):
(WebInspector.LogContentView.prototype._isMessageVisible):
(WebInspector.LogContentView.prototype._isMessageSelected):
(WebInspector.LogContentView.prototype._selectAllMessages):
(WebInspector.LogContentView.prototype._unfilteredMessages):
(WebInspector.LogContentView.prototype._visibleMessages):
(WebInspector.LogContentView.prototype._filterMessages):
(WebInspector.LogContentView.prototype._leftArrowWasPressed):
(WebInspector.LogContentView.prototype._propertiesSectionDidUpdateContent):
(WebInspector.LogContentView.prototype._performSearch):
* UserInterface/Views/Section.css:
(.section .header .title::before):
(.section.expanded .header .title::before):
(.section .header .title, .event-bar .header .title):
(.properties-tree li.parent::before):
(.section .header::before): Deleted.
(.section.expanded .header::before): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180205 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoSometimes RemoteLayerBackingStore ends up non-volatile while the process is suspended
timothy_horton@apple.com [Tue, 17 Feb 2015 02:30:11 +0000 (02:30 +0000)]
Sometimes RemoteLayerBackingStore ends up non-volatile while the process is suspended
<rdar://problem/19842957>
https://bugs.webkit.org/show_bug.cgi?id=141675

Reviewed by Simon Fraser.

Previously, it was possible to get a layer tree flush in between the
process suspension cleanup code making surfaces volatile and the process
actually being suspended (it was racy because the suspension requires
a few IPC round trips). Depending on how far through the flush we got,
we could sometimes end up either making new non-volatile backing store,
or switching some recently-made-volatile backing store back to non-volatile.
We don't want to have any non-volatile backing store while suspended.

* UIProcess/ProcessThrottler.cpp:
(WebKit::ProcessThrottler::updateAssertion):
Inform the WebProcess when it's going from suspended to runnable state.

* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::sendProcessDidResume):
* UIProcess/WebProcessProxy.h:
Forward the message along.

* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::processWillSuspend):
(WebKit::WebProcess::cancelProcessWillSuspend):
(WebKit::WebProcess::setAllLayerTreeStatesFrozen):
(WebKit::WebProcess::processDidResume):
Freeze all of this process' pages' layer trees when we start trying to suspend,
and un-freeze them when either suspension is cancelled or we resume.

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

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180204 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoAdopt CAMachPort-as-layer-contents
timothy_horton@apple.com [Tue, 17 Feb 2015 02:28:52 +0000 (02:28 +0000)]
Adopt CAMachPort-as-layer-contents
https://bugs.webkit.org/show_bug.cgi?id=141687
<rdar://problem/19393233>

Reviewed by Simon Fraser.

No new tests, just a performance bump.

* platform/cocoa/MachSendRight.h:
(WebCore::MachSendRight::operator bool):
Add an operator bool() that checks if the underlying port is nonnull.

* platform/spi/cocoa/QuartzCoreSPI.h:
Add some SPI.

* Shared/mac/RemoteLayerBackingStore.h:
* Shared/mac/RemoteLayerBackingStore.mm:
(WebKit::RemoteLayerBackingStore::decode):
If we have CAMachPort, just keep the MachSendRight around.

(WebKit::RemoteLayerBackingStore::applyBackingStoreToLayer):
If we have CAMachPort, make one and leak our send right into it. CAMachPort
will adopt the port and destroy it when needed.

(WebKit::RemoteLayerBackingStore::setBufferVolatility):
Tiny style fix.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180203 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUpdate cross-frame http tests after r177774
mmaxfield@apple.com [Tue, 17 Feb 2015 02:21:09 +0000 (02:21 +0000)]
Update cross-frame http tests after r177774
https://bugs.webkit.org/show_bug.cgi?id=141679

Reviewed by Alexey Proskuryakov.

Simply a spacing change

* http/tests/security/XFrameOptions/x-frame-options-deny-multiple-clients-expected.txt:
* http/tests/security/cross-frame-access-callback-explicit-domain-ALLOW-expected.txt:
* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180202 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUpdate accessibility/table-cells.html after r177774
mmaxfield@apple.com [Tue, 17 Feb 2015 02:17:37 +0000 (02:17 +0000)]
Update accessibility/table-cells.html after r177774
https://bugs.webkit.org/show_bug.cgi?id=141680

Reviewed by Chris Fleizach.

Simply needs a rebaseline.

* platform/mac/TestExpectations:
* platform/mac/accessibility/table-cells-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180201 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[OS X] Reordering TestExpectations
mmaxfield@apple.com [Tue, 17 Feb 2015 02:10:52 +0000 (02:10 +0000)]
[OS X] Reordering TestExpectations

Unreviewed.

* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180200 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agofast/text/glyph-reordering.html no longer fails
mmaxfield@apple.com [Tue, 17 Feb 2015 01:33:59 +0000 (01:33 +0000)]
fast/text/glyph-reordering.html no longer fails
https://bugs.webkit.org/show_bug.cgi?id=141678

Reviewed by Sam Weinig.

* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180197 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUpdate fast/text/decomposed-after-stacked-diacritics.html after r177774
mmaxfield@apple.com [Tue, 17 Feb 2015 01:28:42 +0000 (01:28 +0000)]
Update fast/text/decomposed-after-stacked-diacritics.html after r177774
https://bugs.webkit.org/show_bug.cgi?id=141677

Reviewed by Dan Bernstein.

Apply kerning to both spans in the same way.

* fast/text/decomposed-after-stacked-diacritics-expected.txt:
* fast/text/decomposed-after-stacked-diacritics.html:
* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180195 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUpdate fast/regions/cssom/get-regions-by-content-horiz-*.html after r177774
mmaxfield@apple.com [Tue, 17 Feb 2015 01:16:21 +0000 (01:16 +0000)]
Update fast/regions/cssom/get-regions-by-content-horiz-*.html after r177774
https://bugs.webkit.org/show_bug.cgi?id=141674

Reviewed by Zalan Bujtas.

Text metrics changed, so the text ended up no longer spanning a region boundary.

* fast/regions/cssom/get-regions-by-content-horiz-bt.html:
* fast/regions/cssom/get-regions-by-content-horiz-tb.html:
* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180194 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoAdd experimental <attachment> element support
weinig@apple.com [Tue, 17 Feb 2015 01:09:52 +0000 (01:09 +0000)]
Add experimental <attachment> element support
https://bugs.webkit.org/show_bug.cgi?id=141626

Reviewed by Tim Horton.

Source/JavaScriptCore:

* Configurations/FeatureDefines.xcconfig:

Source/WebCore:

Adds scaffolding for an experimental implementation of an <attachment> element
which can represent a file attachment (e.g. NSTextAttachment in NSAttributedString
parlance).

The implementation is guarded by both an #ifdef (ENABLE(ATTACHMENT_ELEMENT)) and
a setting (Settings::attachmentElementEnabled()).

Tests: fast/attachment/attachment-disabled-dom.html
       fast/attachment/attachment-disabled-rendering.html
       fast/attachment/attachment-dom.html
       fast/attachment/attachment-rendering.html

* Configurations/FeatureDefines.xcconfig:
Add new ENABLE_ATTACHMENT_ELEMENT macro.

* CMakeLists.txt:
* WebCore.vcxproj/WebCore.vcxproj:
* WebCore.xcodeproj/project.pbxproj:
Add new files.

* dom/make_names.pl:
Add support for checking a setting to determine if a particular tag should be exposed.

* html/HTMLAttachmentElement.cpp: Added.
* html/HTMLAttachmentElement.h: Added.
Stub out the basics of the new element.

* html/HTMLElementsAllInOne.cpp:
Add HTMLAttachmentElement.cpp.

* html/HTMLTagNames.in:
Add 'attachment'.

* page/Settings.in:
Add attachmentElementEnabled setting.

* rendering/RenderAttachment.cpp: Added.
* rendering/RenderAttachment.h: Added.
Stub out the basics of the new render. We will probably want to replace this with
a non-replaced render, but this will do as a placeholder for now.

* rendering/RenderObject.h:
(WebCore::RenderObject::isAttachment):
Add predicate for type casting.

* rendering/RenderingAllInOne.cpp:
Add RenderAttachment.cpp.

Source/WebKit/mac:

* Configurations/FeatureDefines.xcconfig:

Source/WebKit2:

* Configurations/FeatureDefines.xcconfig:

LayoutTests:

* fast/attachment: Added.
* fast/attachment/attachment-disabled-dom-expected.txt: Added.
* fast/attachment/attachment-disabled-dom.html: Added.
* fast/attachment/attachment-disabled-rendering-expected.txt: Added.
* fast/attachment/attachment-disabled-rendering.html: Added.
* fast/attachment/attachment-dom-expected.txt: Added.
* fast/attachment/attachment-dom.html: Added.
* fast/attachment/attachment-rendering-expected.txt: Added.
* fast/attachment/attachment-rendering.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180193 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoGC (almost) immediately when navigating under memory pressure.
akling@apple.com [Tue, 17 Feb 2015 01:06:45 +0000 (01:06 +0000)]
GC (almost) immediately when navigating under memory pressure.
<https://webkit.org/b/141663>

Reviewed by Geoffrey Garen.

Since the PageCache is already disabled in memory pressure situations,
we know that detaching the old window shell on navigation is basically
guaranteed to generate a bunch of garbage, we can soften the memory
peak a bit by doing a GC right away instead of scheduling one for soon(tm).

* bindings/js/GCController.cpp:
(WebCore::GCController::GCController):
(WebCore::GCController::garbageCollectSoon):
(WebCore::GCController::garbageCollectOnNextRunLoop):
(WebCore::GCController::gcTimerFired):
* bindings/js/GCController.h:

    Add a GCController::garbageCollectOnNextRunLoop() complement to the
    "soon" and "now" options. There was already a zero timer in here for
    non-CF builds, so I just used that same timer to implement this
    and have the non-CF code path call garbageCollectOnNextRunLoop().

* bindings/js/ScriptController.cpp:
(WebCore::collectGarbageAfterWindowShellDestruction):
(WebCore::ScriptController::~ScriptController):
(WebCore::ScriptController::clearWindowShell):

    Under system memory pressure conditions, schedule a full GC on next
    runloop iteration instead of just asking for it to happen soon.
    We do it on next runloop to ensure that there's no pointer to the
    window object on the stack.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180192 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoEmoji sequences do not render properly.
enrica@apple.com [Tue, 17 Feb 2015 01:05:28 +0000 (01:05 +0000)]
Emoji sequences do not render properly.
https://bugs.webkit.org/show_bug.cgi?id=141661
rdar://problem/19820463

Reviewed by Sam Weinig.

Source/WebCore:

Emoji sequences and emoji with variations should be rendered
using the Complex code path and should be treated as graphemes.
This change modifies advanceByCombiningCharacterSequence to add
this logic.

Test: fast/text/emoji.html

* WebCore.xcodeproj/project.pbxproj:
* platform/graphics/FontCascade.cpp:
(WebCore::FontCascade::characterRangeCodePath):
* platform/graphics/mac/ComplexTextController.cpp:
(WebCore::advanceByCombiningCharacterSequence): Implements a simple
logic to treat emoji sequences and emoji with variations as graphemes.
* platform/text/CharacterProperties.h: Added.
(WebCore::isEmojiGroupCandidate):
(WebCore::isEmojiModifier):
(WebCore::isVariationSelector):
* rendering/RenderText.cpp:
(WebCore::isEmojiGroupCandidate): Deleted.
(WebCore::isEmojiModifier): Deleted.

LayoutTests:

* TestExpectations:
* fast/text/emoji-expected.txt: Added.
* fast/text/emoji.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180191 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoRenderTableRow should check if it has access to its ancestor chain.
zalan@apple.com [Tue, 17 Feb 2015 01:02:17 +0000 (01:02 +0000)]
RenderTableRow should check if it has access to its ancestor chain.
https://bugs.webkit.org/show_bug.cgi?id=141668

Reviewed by Andreas Kling.

Preventive fix.

* rendering/RenderTableRow.cpp:
(WebCore::RenderTableRow::styleDidChange):
(WebCore::RenderTableRow::addChild):
* rendering/RenderTableRow.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180190 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[iOS] Build fix: declare undeclared identifier 'credential'.
jer.noble@apple.com [Tue, 17 Feb 2015 00:26:27 +0000 (00:26 +0000)]
[iOS] Build fix: declare undeclared identifier 'credential'.

* platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.mm:
(WebCore::WebCoreNSURLAuthenticationChallengeClient::receivedCredential):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180187 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoREGRESSION(r180060): C Loop crashes
msaboff@apple.com [Tue, 17 Feb 2015 00:08:17 +0000 (00:08 +0000)]
REGRESSION(r180060): C Loop crashes
https://bugs.webkit.org/show_bug.cgi?id=141671

Reviewed by Geoffrey Garen.

Fixed a typo that only affected the C Loop in the prologue() macro in LowLevelInterpreter.asm.
After the stackHeightOKGetCodeBlock label, codeBlockSetter(t1) should be codeBlockGetter(t1).
Fixed the processing of an out of stack exception in llint_stack_check to not get the caller's
frame.  This isn't needed, since this helper is only called to check the stack on entry.  Any
exception will be handled by a call ancestor.

* llint/LLIntSlowPaths.cpp:
(JSC::LLInt::llint_stack_check): Changed to use the current frame for processing an exception.
* llint/LowLevelInterpreter.asm: Fixed a typo.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180184 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[iOS] Unable to play .mp4 file over http with basic http authentication
jer.noble@apple.com [Tue, 17 Feb 2015 00:02:44 +0000 (00:02 +0000)]
[iOS] Unable to play .mp4 file over http with basic http authentication
https://bugs.webkit.org/show_bug.cgi?id=141503
rdar://problem/15799844

Reviewed by Alexey Proskuryakov.

On iOS, CFNetwork is used for authentication, so the NSURLAuthenticationChallenge
provided by AVAssetResourceLoader needs to be shoehorned into a AuthenticationChallenge
object by way of CFURLAuthChallengeRef.

Create a new class, WebCoreNSURLAuthenticationChallengeClient, whose sole purpose
is to take AuthenticationChallengeClient callbacks and pass them along to a
NSURLAuthenticationChallenge sender.

Create a NSURLAuthenticationChallenge out of the CF version through an SPI, and add
that SPI to a new header. Drive-by fix: take two of our existing SPI calls and move
them into that same header.

* WebCore.xcodeproj/project.pbxproj: Add CFNSURLConnectionSPI.h to project.
* platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.mm:
(WebCore::WebCoreNSURLAuthenticationChallengeClient::create): Factory.
(WebCore::WebCoreNSURLAuthenticationChallengeClient::WebCoreNSURLAuthenticationChallengeClient): Simple constructor.
(WebCore::WebCoreNSURLAuthenticationChallengeClient::receivedCredential): Pass to m_challenge.
(WebCore::WebCoreNSURLAuthenticationChallengeClient::receivedRequestToContinueWithoutCredential): Ditto.
(WebCore::WebCoreNSURLAuthenticationChallengeClient::receivedCancellation): Ditto.
(WebCore::WebCoreNSURLAuthenticationChallengeClient::receivedRequestToPerformDefaultHandling): Ditto.
(WebCore::WebCoreNSURLAuthenticationChallengeClient::receivedChallengeRejection): Ditto.
(WebCore::MediaPlayerPrivateAVFoundationObjC::shouldWaitForResponseToAuthenticationChallenge):
    Create an AuthenticationChallenge out of the nsChallenge and client and pass it up
    to the HTMLMediaElement.
* platform/network/mac/AuthenticationMac.mm:
* platform/spi/cocoa/CFNSURLConnectionSPI.h: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180183 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoREGRESSION (r176459): Process suspension cleanup timer sometimes never stops
timothy_horton@apple.com [Mon, 16 Feb 2015 23:24:09 +0000 (23:24 +0000)]
REGRESSION (r176459): Process suspension cleanup timer sometimes never stops
https://bugs.webkit.org/show_bug.cgi?id=141669

Reviewed by Simon Fraser.

* WebProcess/WebProcess.cpp:
(WebKit::WebProcess::markAllLayersVolatileIfPossible):
r176459 accidentally removed the code to stop the cleanup timer
in the case where we successfully finish marking layers volatile,
causing the timer to continue running once the process comes back
from a suspended state.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180180 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoRevert a change in SVGRenderSupport::mapLocalToContainer committed for fixing <https...
commit-queue@webkit.org [Mon, 16 Feb 2015 23:19:23 +0000 (23:19 +0000)]
Revert a change in SVGRenderSupport::mapLocalToContainer committed for fixing <https://bugs.webkit.org/show_bug.cgi?id=119626>.
https://bugs.webkit.org/show_bug.cgi?id=138439

Patch by Said Abou-Hallawa <sabouhallawa@apple.com> on 2015-02-16
Reviewed by Darin Adler.

The change http://trac.webkit.org/changeset/164861 was ported from Blink
but it was was revert later because it broke Chrome SVG rendering. It also
broke the WebKit SVG text search highlighting.

Tests: svg/transforms/svg-geometry-crash.html: This test was added by the
blamed change. A new but correct solution is required to fix the assertion.

* rendering/svg/SVGRenderSupport.h:
* rendering/svg/SVGRenderSupport.cpp:
Notice that TransformState.applyTransform() does matrix-left-multiplication,
i.e. state = transform * state. But operator*() of AffineTransform and
MatrixTransform  does right multiply, i.e. this = this * transform.

(WebCore::SVGRenderSupport::localToParentTransform): Have the calculation
of the SVG localToParentTransform in one shared function.

(WebCore::SVGRenderSupport::mapLocalToContainer): Revert the Blink change.
If the parent is the SVG root, the transform state should be equal to
transform = svg_to_css_mapping * local_to_parent_mapping * transform.

(WebCore::SVGRenderSupport::pushMappingToContainer): Get the localToParent
transform to be pushed in the geometryMap. If the parent is the SVG root,
localToParent = svg_to_css_mapping * local_to_parent_mapping. The original
code was doing the opposite and this is the cause of the assertion which
was fixed wrongly by reversing the correct multiplication order in
SVGRenderSupport::mapLocalToContainer().

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180179 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUpdate fast/dom/{Element,Range}/getClientRects.html after r177774
mmaxfield@apple.com [Mon, 16 Feb 2015 23:11:23 +0000 (23:11 +0000)]
Update fast/dom/{Element,Range}/getClientRects.html after r177774
https://bugs.webkit.org/show_bug.cgi?id=141665

Reviewed by Alexey Proskuryakov.

Line breaking is different due to different text metrics.

* fast/dom/Element/getClientRects-expected.txt:
* fast/dom/Element/getClientRects.html:
* fast/dom/Range/getClientRects-expected.txt:
* fast/dom/Range/getClientRects.html:
* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180178 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoDelete LayoutTests/platform/mac-mountainlion directory
mmaxfield@apple.com [Mon, 16 Feb 2015 23:01:54 +0000 (23:01 +0000)]
Delete LayoutTests/platform/mac-mountainlion directory

Rubber-stamped by Brent Fulgham.

The Win port falls back to mac-mountainlion, so this patch moves all files in
platform/mac-mountainlion that aren't already in platform/win into platform/win. It then deletes
the remaining files in platform/mac-mountainlion.

* platform/mac-mountainlion: Deleted.
* platform/win/{many files}: Moved from platform/mac-mountainlion

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180176 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoCheck for the assume_nonnull feature instead of noescape
andersca@apple.com [Mon, 16 Feb 2015 22:40:51 +0000 (22:40 +0000)]
Check for the assume_nonnull feature instead of noescape
https://bugs.webkit.org/show_bug.cgi?id=141666

Reviewed by Dan Bernstein.

* Shared/API/Cocoa/WKFoundation.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180175 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoRenderTableCell can't access its parent while in detached state.
zalan@apple.com [Mon, 16 Feb 2015 22:38:29 +0000 (22:38 +0000)]
RenderTableCell can't access its parent while in detached state.
https://bugs.webkit.org/show_bug.cgi?id=141639
rdar://problem/19850760

Reviewed by Simon Fraser.

Null check against ancestor chain so that certain methods in RenderTableCell can
be called even if the renderer is not yet attached.

Source/WebCore:

Test: fast/table/table-cell-crash-when-detached-state.html

* rendering/RenderTableCell.cpp:
(WebCore::RenderTableCell::borderLeft):
(WebCore::RenderTableCell::borderRight):
(WebCore::RenderTableCell::borderTop):
(WebCore::RenderTableCell::borderBottom):
(WebCore::RenderTableCell::borderStart):
(WebCore::RenderTableCell::borderEnd):
(WebCore::RenderTableCell::borderBefore):
(WebCore::RenderTableCell::borderAfter):
* rendering/RenderTableCell.h:

LayoutTests:

* fast/table/table-cell-crash-when-detached-state-expected.txt: Added.
* fast/table/table-cell-crash-when-detached-state.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180174 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoWeb Inspector: Scope details sidebar should label objects with constructor names
commit-queue@webkit.org [Mon, 16 Feb 2015 22:10:12 +0000 (22:10 +0000)]
Web Inspector: Scope details sidebar should label objects with constructor names
https://bugs.webkit.org/show_bug.cgi?id=139449

Patch by Joseph Pecoraro <pecoraro@apple.com> on 2015-02-16
Reviewed by Timothy Hatcher.

Source/JavaScriptCore:

* inspector/JSInjectedScriptHost.cpp:
(Inspector::JSInjectedScriptHost::internalConstructorName):
* runtime/Structure.cpp:
(JSC::Structure::toStructureShape):
Share calculatedClassName.

* runtime/JSObject.h:
* runtime/JSObject.cpp:
(JSC::JSObject::calculatedClassName):
Elaborate on a way to get an Object's class name.

LayoutTests:

* inspector/model/remote-object-expected.txt:
* inspector/model/remote-object.html:
Improve the test to include Objects where previously
we would have had poorer class name descriptions.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180173 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUpdate expected result for editing/pasteboard/5761530-1.html after r177774
mmaxfield@apple.com [Mon, 16 Feb 2015 21:58:33 +0000 (21:58 +0000)]
Update expected result for editing/pasteboard/5761530-1.html after r177774
https://bugs.webkit.org/show_bug.cgi?id=141603

Reviewed by Alexey Proskuryakov.

The test used to have incorrect expected output, but the new results are correct.

* LayoutTests/editing/pasteboard/5761530-1-expected.txt:
* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180172 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUpdate fast/css/css3-ch-unit.html after r177774
mmaxfield@apple.com [Mon, 16 Feb 2015 21:51:45 +0000 (21:51 +0000)]
Update fast/css/css3-ch-unit.html after r177774
https://bugs.webkit.org/show_bug.cgi?id=141657

Reviewed by Zalan Bujtas.

The alphabet is not expected to have the same width as 26 times the average character width.

* fast/css/css3-ch-unit-expected.txt:
* fast/css/css3-ch-unit.html:
* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180171 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agofullscreen/full-screen-plugin.html is very flaky on Yosemite WK2
ap@apple.com [Mon, 16 Feb 2015 21:44:10 +0000 (21:44 +0000)]
fullscreen/full-screen-plugin.html is very flaky on Yosemite WK2
https://bugs.webkit.org/show_bug.cgi?id=141364

* platform/mac-wk2/TestExpectations: Marking as such.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180170 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoinspector/css/stylesheet-with-mutations.html is a flaky timeout
ap@apple.com [Mon, 16 Feb 2015 21:26:24 +0000 (21:26 +0000)]
inspector/css/stylesheet-with-mutations.html is a flaky timeout
https://bugs.webkit.org/show_bug.cgi?id=141601

* platform/mac-wk2/TestExpectations: Marking as such.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180168 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUpdate fast/css/content/content-quotes-*.html after r177774
mmaxfield@apple.com [Mon, 16 Feb 2015 20:41:23 +0000 (20:41 +0000)]
Update fast/css/content/content-quotes-*.html after r177774
https://bugs.webkit.org/show_bug.cgi?id=141654

Reviewed by Zalan Bujtas.

Because of printer fonts and LayoutUnits, putting text inside an anonymous renderer
has different metrics than the raw text itself.

* fast/css/content/content-quotes-01.html:
* fast/css/content/content-quotes-02.html:
* fast/css/content/content-quotes-03.html:
* fast/css/content/content-quotes-04.html:
* fast/css/content/content-quotes-05.html:
* fast/css/content/content-quotes-06.html:
* fast/css/content/content-quotes-07.html:
* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180166 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUpdate expectations for css2.1/t1202-counter-04-b.html and css2.1/t1202-counters...
ap@apple.com [Mon, 16 Feb 2015 20:13:54 +0000 (20:13 +0000)]
Update expectations for css2.1/t1202-counter-04-b.html and css2.1/t1202-counters-04-b.html
rdar://problem/19848737

* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180165 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoFix the MiniBrowser build with newer clang
timothy_horton@apple.com [Mon, 16 Feb 2015 19:56:02 +0000 (19:56 +0000)]
Fix the MiniBrowser build with newer clang

* MiniBrowser/mac/SettingsController.m:
(-[SettingsController _toggleBooleanDefault:]):
* MiniBrowser/mac/WK1BrowserWindowController.m:
(-[WK1BrowserWindowController windowWillClose:]):
* MiniBrowser/mac/WK2BrowserWindowController.m:
(-[WK2BrowserWindowController windowWillClose:]):
NSApp is id and clang gets confused about which -delegate we're talking about.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180164 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUpdate expected result for compositing/fixed-positioned-pseudo-content-no-compositing...
mmaxfield@apple.com [Mon, 16 Feb 2015 19:43:21 +0000 (19:43 +0000)]
Update expected result for compositing/fixed-positioned-pseudo-content-no-compositing.html after r177774
https://bugs.webkit.org/show_bug.cgi?id=141602

Reviewed by Simon Fraser.

Not sure if the change to contentsOpaque is right, so CC'ing smfr to see what he thinks.

* compositing/fixed-positioned-pseudo-content-no-compositing-expected.txt:
* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180163 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoAdd nullability qualifiers to all API headers
andersca@apple.com [Mon, 16 Feb 2015 19:38:08 +0000 (19:38 +0000)]
Add nullability qualifiers to all API headers
https://bugs.webkit.org/show_bug.cgi?id=141652
rdar://problem/19793630

Reviewed by Dan Bernstein.

* Shared/API/Cocoa/WKFoundation.h:
Add a WK_NULLABLE_SPECIFIER macro.

* UIProcess/API/Cocoa/WKBackForwardListItem.h:
* UIProcess/API/Cocoa/WKFrameInfo.h:
* UIProcess/API/Cocoa/WKNavigationAction.h:
* UIProcess/API/Cocoa/WKNavigationDelegate.h:
* UIProcess/API/Cocoa/WKNavigationResponse.h:
* UIProcess/API/Cocoa/WKScriptMessage.h:
* UIProcess/API/Cocoa/WKScriptMessageHandler.h:
* UIProcess/API/Cocoa/WKUIDelegate.h:
* UIProcess/API/Cocoa/WKUserContentController.h:
* UIProcess/API/Cocoa/WKUserScript.h:
* UIProcess/API/Cocoa/WKWebView.h:
* UIProcess/API/Cocoa/WKWebViewConfiguration.h:
* UIProcess/API/Cocoa/WKWindowFeatures.h:

* mac/postprocess-framework-headers.sh:
Handle the WK_NULLABLE_SPECIFIER macro. Change the WK_NULLABLE sed command to replace
all occurrences of WK_NULLABLE instead of just the first one.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180162 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[iOS] Gardening; update test expectations
dbates@webkit.org [Mon, 16 Feb 2015 19:32:30 +0000 (19:32 +0000)]
[iOS] Gardening; update test expectations

* platform/ios-simulator-wk1/TestExpectations:
* platform/ios-simulator-wk2/TestExpectations:
* platform/ios-simulator/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180161 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoDFG SSA should use GetLocal for arguments, and the GetArgument node type should be...
fpizlo@apple.com [Mon, 16 Feb 2015 19:27:37 +0000 (19:27 +0000)]
DFG SSA should use GetLocal for arguments, and the GetArgument node type should be removed
https://bugs.webkit.org/show_bug.cgi?id=141623

Reviewed by Oliver Hunt.

During development of https://bugs.webkit.org/show_bug.cgi?id=141332, I realized that I
needed to use GetArgument for loading something that has magically already appeared on the
stack, so currently trunk sort of allows this. But then I realized three things:

- A GetArgument with a non-JSValue flush format means speculating that the value on the
  stack obeys that format, rather than just assuming that that it already has that format.
  In bug 141332, I want it to assume rather than speculate. That also happens to be more
  intuitive; I don't think I was wrong to expect that.

- The node I really want is GetLocal. I'm just getting the value of the local and I don't
  want to do anything else.

- Maybe it would be easier if we just used GetLocal for all of the cases where we currently
  use GetArgument.

This changes the FTL to do argument speculations in the prologue just like the DFG does.
This brings some consistency to our system, and allows us to get rid of the GetArgument
node. The speculations that the FTL must do are now made explicit in the m_argumentFormats
vector in DFG::Graph. This has natural DCE behavior: even if all uses of the argument are
dead we will still speculate. We already have safeguards to ensure we only speculate if
there are uses that benefit from speculation (which is a much more conservative criterion
than DCE).

* dfg/DFGAbstractInterpreterInlines.h:
(JSC::DFG::AbstractInterpreter<AbstractStateType>::executeEffects):
* dfg/DFGClobberize.h:
(JSC::DFG::clobberize):
* dfg/DFGDCEPhase.cpp:
(JSC::DFG::DCEPhase::run):
* dfg/DFGDoesGC.cpp:
(JSC::DFG::doesGC):
* dfg/DFGFixupPhase.cpp:
(JSC::DFG::FixupPhase::fixupNode):
* dfg/DFGFlushFormat.h:
(JSC::DFG::typeFilterFor):
* dfg/DFGGraph.cpp:
(JSC::DFG::Graph::dump):
* dfg/DFGGraph.h:
(JSC::DFG::Graph::valueProfileFor):
(JSC::DFG::Graph::methodOfGettingAValueProfileFor):
* dfg/DFGInPlaceAbstractState.cpp:
(JSC::DFG::InPlaceAbstractState::initialize):
* dfg/DFGNode.cpp:
(JSC::DFG::Node::hasVariableAccessData):
* dfg/DFGNodeType.h:
* dfg/DFGOSRAvailabilityAnalysisPhase.cpp:
(JSC::DFG::OSRAvailabilityAnalysisPhase::run):
(JSC::DFG::LocalOSRAvailabilityCalculator::executeNode):
* dfg/DFGPredictionPropagationPhase.cpp:
(JSC::DFG::PredictionPropagationPhase::propagate):
* dfg/DFGPutLocalSinkingPhase.cpp:
* dfg/DFGSSAConversionPhase.cpp:
(JSC::DFG::SSAConversionPhase::run):
* dfg/DFGSafeToExecute.h:
(JSC::DFG::safeToExecute):
* dfg/DFGSpeculativeJIT32_64.cpp:
(JSC::DFG::SpeculativeJIT::compile):
* dfg/DFGSpeculativeJIT64.cpp:
(JSC::DFG::SpeculativeJIT::compile):
* ftl/FTLCapabilities.cpp:
(JSC::FTL::canCompile):
* ftl/FTLLowerDFGToLLVM.cpp:
(JSC::FTL::LowerDFGToLLVM::lower):
(JSC::FTL::LowerDFGToLLVM::compileNode):
(JSC::FTL::LowerDFGToLLVM::compileGetLocal):
(JSC::FTL::LowerDFGToLLVM::compileGetArgument): Deleted.
* tests/stress/dead-speculating-argument-use.js: Added.
(foo):
(o.valueOf):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180160 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUpdate editing/selection/move-by-word-visually-mac.html after r177774
mmaxfield@apple.com [Mon, 16 Feb 2015 19:05:06 +0000 (19:05 +0000)]
Update editing/selection/move-by-word-visually-mac.html after r177774
https://bugs.webkit.org/show_bug.cgi?id=141609

Reviewed by Alexey Proskuryakov.

Line break occurs in a different place, so option-left and option-right
have slightly different behavior.

* editing/selection/move-by-word-visually-mac.html:
* editing/selection/move-by-word-visually-mac-expected.txt:
* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180159 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUpdate SPI availability annotations
mitz@apple.com [Mon, 16 Feb 2015 18:56:06 +0000 (18:56 +0000)]
Update SPI availability annotations
https://bugs.webkit.org/show_bug.cgi?id=141588

Reviewed by Anders Carlsson.

* Shared/API/Cocoa/WKFoundation.h: Added a definition of __NSi_8_3 for when building with
an older SDK.
* UIProcess/API/Cocoa/WKErrorPrivate.h:
* UIProcess/API/Cocoa/WKUIDelegatePrivate.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180158 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoRemove unused InjectedBundle::setCacheModel.
antti@apple.com [Mon, 16 Feb 2015 18:37:09 +0000 (18:37 +0000)]
Remove unused InjectedBundle::setCacheModel.

Rubber-stamped by Anders Carlsson.

* WebProcess/InjectedBundle/InjectedBundle.cpp:
(WebKit::InjectedBundle::setCacheModel): Deleted.
* WebProcess/InjectedBundle/InjectedBundle.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180156 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoAdd ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
hyatt@apple.com [Mon, 16 Feb 2015 18:04:13 +0000 (18:04 +0000)]
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):
Add ifdefs to avoid adjusting Ruby selection rects to avoid overlap on iOS.
https://bugs.webkit.org/show_bug.cgi?id=141651
<rdar://problem/19182526>

Reviewed by Simon Fraser.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180155 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoSoftLinking.h: Update copyright and license; clean up whitespace
ddkilzer@apple.com [Mon, 16 Feb 2015 18:02:19 +0000 (18:02 +0000)]
SoftLinking.h: Update copyright and license; clean up whitespace

* platform/mac/SoftLinking.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180154 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoFix typo.
andersca@apple.com [Mon, 16 Feb 2015 17:55:43 +0000 (17:55 +0000)]
Fix typo.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180153 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoBuild fix for versions of Xcode that support nullability but don't have the AppKit...
andersca@apple.com [Mon, 16 Feb 2015 17:50:26 +0000 (17:50 +0000)]
Build fix for versions of Xcode that support nullability but don't have the AppKit #defines.

* Shared/API/Cocoa/WKFoundation.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180152 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agotext-underline-position:under has multiple correctness issues
hyatt@apple.com [Mon, 16 Feb 2015 17:47:27 +0000 (17:47 +0000)]
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.
text-underline-position:under has multiple correctness issues
https://bugs.webkit.org/show_bug.cgi?id=141528

Reviewed by Dean Jackson.

Source/WebCore:

Added a bunch of new tests in fast/text

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::maxLogicalBottomForTextDecorationLine):
(WebCore::InlineFlowBox::minLogicalTopForTextDecorationLine):
(WebCore::InlineFlowBox::computeMaxLogicalBottom): Deleted.
* rendering/InlineFlowBox.h:
These functions have been re-written to take an enclosing renderer that specified the
decoration. This way they can properly limit the bottom/top computation to only line boxes
that are contained inside the renderer.

* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paintDecoration):
Tweak the call to get the decoration colors now that quirks mode has been removed.

* rendering/RenderElement.cpp:
(WebCore::RenderElement::enclosingRendererWithTextDecoration):
* rendering/RenderElement.h:
New function that finds the enclosing renderer that specified a text decoration. For now
this is only used in the "under" position computation, but soon we'll be using it
everywhere.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
* rendering/RenderObject.h:
Remove the quirks mode argument, since we were always passing true anyway (making the argument dead).

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::maxLogicalBottom): Deleted.
* rendering/RootInlineBox.h:
Get rid of these functions and just have InlineTextBoxStyle's computeUnderLineOffset function call
the InlineFlowBox functions directly.

* style/InlineTextBoxStyle.cpp:
(WebCore::computeUnderlineOffset):
Re-written to fetch the enclosingRendererWithTextDecoration so that it can be passed to the logical top/bottom
computation to limit which line boxes get included.

LayoutTests:

* fast/text/text-underline-first-line-decoration-expected.html: Added.
* fast/text/text-underline-first-line-decoration.html: Added.
* fast/text/text-underline-position-mixed-fonts-expected.html: Added.
* fast/text/text-underline-position-mixed-fonts.html: Added.
* fast/text/text-underline-position-subscript-expected.html: Added.
* fast/text/text-underline-position-subscript.html: Added.
* fast/text/text-underline-vertical-first-line-decoration-expected.html: Added.
* fast/text/text-underline-vertical-first-line-decoration.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180150 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoDevelop/Disable Caches does not properly disable disk cache
antti@apple.com [Mon, 16 Feb 2015 17:17:39 +0000 (17:17 +0000)]
Develop/Disable Caches does not properly disable disk cache
https://bugs.webkit.org/show_bug.cgi?id=141636

Reviewed by Anders Carlsson.

Source/WebKit2:

The menu item sets the cache model to "document viewer" to disable the caches (cache size 0).
However the cache model was never passed to the network process after the initialization.

* NetworkProcess/NetworkProcess.messages.in:
* NetworkProcess/cache/NetworkCacheStorageCocoa.mm:
(WebKit::NetworkCacheStorage::retrieve):
(WebKit::NetworkCacheStorage::store):
(WebKit::NetworkCacheStorage::update):

    Don't try to store or retrieve when the maximum cache size is set to zero.

* UIProcess/WebProcessPool.cpp:
(WebKit::WebProcessPool::setCacheModel):

   Message the network process too.

* WebProcess/InjectedBundle/API/c/WKBundle.cpp:
(WKBundleSetCacheModel): Deleted.

    Remove the now unused SPI.

* WebProcess/InjectedBundle/API/c/WKBundlePrivate.h:

Tools:

Make testRunner.setCacheModel() use the WKContext interface so the model gets passed to
the network process too.

This makes setCacheModel asynchronous but there don't seem to be any tests that expect otherwise.

* WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
(WTR::InjectedBundle::beginTesting):
(WTR::InjectedBundle::setCacheModel):
* WebKitTestRunner/InjectedBundle/InjectedBundle.h:
* WebKitTestRunner/InjectedBundle/TestRunner.cpp:

    Forward to TestController.

(WTR::TestRunner::setCacheModel):
* WebKitTestRunner/TestInvocation.cpp:
(WTR::TestInvocation::didReceiveMessageFromInjectedBundle):

LayoutTests:

* http/tests/cache/disk-cache-disable-expected.txt: Added.
* http/tests/cache/disk-cache-disable.html: Added.
* http/tests/cache/resources/cache-test.js:
(runTests): Add optional completion handler.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180149 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoAssertion in disk cache code with redirect to a non-http resource
antti@apple.com [Mon, 16 Feb 2015 17:16:16 +0000 (17:16 +0000)]
Assertion in disk cache code with redirect to a non-http resource
https://bugs.webkit.org/show_bug.cgi?id=141644

Reviewed by Anders Carlsson.

Source/WebKit2:

* NetworkProcess/NetworkResourceLoader.cpp:
(WebKit::NetworkResourceLoader::didFinishLoading):

    Check that response is HTTP before calling computeFreshnessLifetimeForHTTPFamily.

* NetworkProcess/cache/NetworkCache.cpp:
(WebKit::canStore):

    Deny storing non-HTTP responses to cache (though the above already covers this).

LayoutTests:

* http/tests/cache/disk-cache-redirect-to-data-expected.txt: Added.
* http/tests/cache/disk-cache-redirect-to-data.html: Added.
* http/tests/cache/resources/redirect-to-data.php: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180148 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoFEGaussianBlur::calculateUnscaledKernelSize does unspeakable things with signed and...
bfulgham@apple.com [Mon, 16 Feb 2015 17:07:08 +0000 (17:07 +0000)]
FEGaussianBlur::calculateUnscaledKernelSize does unspeakable things with signed and unsigned values
https://bugs.webkit.org/show_bug.cgi?id=141596
<rdar://problem/19837103>

Reviewed by Zalan Bujtas.

No new tests. Covered by css3/filters/huge-blur-value.html

Avoid overflowing the signed integer values by not converting from unsigned
until the maximum size has been clamped to the expected max.

* platform/graphics/filters/FEGaussianBlur.cpp:
(WebCore::FEGaussianBlur::calculateUnscaledKernelSize):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180147 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoAdd support for nullability to the header postprocessing
andersca@apple.com [Mon, 16 Feb 2015 16:36:39 +0000 (16:36 +0000)]
Add support for nullability to the header postprocessing
https://bugs.webkit.org/show_bug.cgi?id=141573

Reviewed by Darin Adler.

* Shared/API/Cocoa/WKFoundation.h:
* UIProcess/API/Cocoa/WKBackForwardList.h:
* mac/postprocess-framework-headers.sh:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180146 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[EFL] Rename WebContextEfl to WebProcessPoolEfl
g.czajkowski@samsung.com [Mon, 16 Feb 2015 16:10:27 +0000 (16:10 +0000)]
[EFL] Rename WebContextEfl to WebProcessPoolEfl
https://bugs.webkit.org/show_bug.cgi?id=141642

Reviewed by Csaba Osztrogonác.

Although we've renamed almost everything in r177716, this is still
required after r177692.

* UIProcess/efl/WebProcessPoolEfl.cpp: Renamed from Source/WebKit2/UIProcess/efl/WebContextEfl.cpp.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180145 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUnreviewed. Add myself as a committer.
a.badowski@samsung.com [Mon, 16 Feb 2015 15:15:22 +0000 (15:15 +0000)]
Unreviewed. Add myself as a committer.

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

Change-Id: I5d745aac5c20ab589446a69ef943202a6db4947a

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180144 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUnreviewed. Uploading a new test that should have been uploaded as
svillar@igalia.com [Mon, 16 Feb 2015 12:36:47 +0000 (12:36 +0000)]
Unreviewed. Uploading a new test that should have been uploaded as
part of r180142.

* fast/css-grid-layout/grid-initialize-span-one-items-expected.txt: Added.
* fast/css-grid-layout/grid-initialize-span-one-items.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180143 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[CSS Grid Layout] Invalid initialization of track sizes with non spanning grid items
svillar@igalia.com [Mon, 16 Feb 2015 12:00:17 +0000 (12:00 +0000)]
[CSS Grid Layout] Invalid initialization of track sizes with non spanning grid items
https://bugs.webkit.org/show_bug.cgi?id=140763

Reviewed by Antti Koivisto.

Source/WebCore:

Content sized tracks with non-spanning grid items were not
properly sized because the growth limit was sometimes infinity
(-1) after calling resolveContentBasedTrackSizingFunctions() when
it should not. This patch adds an special initialization phase for
non-spanning grid items as the new track sizing algorithm
describes.

Granted, that was handled in the old algorithm in
distributeSpaceToTracks() as a special case. The problem is that
it regressed after the optimization added in r173868 because that
method is no longer called when the space to distribute is 0.

That's why we could fix this by allowing calls to
distributeSpaceToTracks() with spaceToDistribute>=0 but by fixing
it with an explicit initialization our implementation becomes
closer to the new algorithm and the initialization is now explicit
in the code instead of a side effect of calling
distributeSpaceToTracks() with no space to be distributed. It also
brings a slight performance improvement as we save sorts and hash
lookups.

I also took the change to add caching to several GridTrackSize
methods that were hot on the profiler (each one accounted for ~1%
of the total time, now they account for ~0.3% each).

Test: fast/css-grid-layout/grid-initialize-span-one-items.html

* rendering/RenderGrid.cpp:
(WebCore::GridItemWithSpan::span): New helper method for ASSERTs.
(WebCore::RenderGrid::resolveContentBasedTrackSizingFunctions):
Exclude non spanning grid items from the calls to
resolveContentBasedTrackSizingFunctionsForItems().
(WebCore::RenderGrid::resolveContentBasedTrackSizingFunctionsForNonSpanningItems):
New method to resolve track sizes only using non-spanning grid
items.
(WebCore::RenderGrid::resolveContentBasedTrackSizingFunctionsForItems):
Ensure that it isn't called for non-spanning grid items.
* rendering/RenderGrid.h:
* rendering/style/GridTrackSize.h:
(WebCore::GridTrackSize::GridTrackSize): Cache return values.
(WebCore::GridTrackSize::setLength): Ditto.
(WebCore::GridTrackSize::setMinMax): Ditto.
(WebCore::GridTrackSize::cacheMinMaxTrackBreadthTypes): New method
that caches the return values for hasXXXTrackBreadth() methods.
(WebCore::GridTrackSize::hasMinOrMaxContentMinTrackBreadth): Use
the cached return value.
(WebCore::GridTrackSize::hasMaxContentMaxTrackBreadth): Ditto.
(WebCore::GridTrackSize::hasMinContentMaxTrackBreadth): Ditto.
(WebCore::GridTrackSize::hasMinOrMaxContentMaxTrackBreadth): Ditto.
(WebCore::GridTrackSize::hasMaxContentMinTrackBreadth): Ditto.
(WebCore::GridTrackSize::hasMinContentMinTrackBreadth): Ditto.
(WebCore::GridTrackSize::hasMinContentMinTrackBreadthAndMinOrMaxContentMaxTrackBreadth):
Ditto.
(WebCore::GridTrackSize::hasMaxContentMinTrackBreadthAndMaxContentMaxTrackBreadth):
Ditto.

LayoutTests:

* fast/css-grid-layout/grid-content-sized-columns-resolution-expected.txt:
* fast/css-grid-layout/grid-content-sized-columns-resolution.html:
* fast/css-grid-layout/grid-initialize-span-one-items-expected.txt: Added.
* fast/css-grid-layout/grid-initialize-span-one-items.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180142 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[GTK] Loading page into WebView shows g_closure_unref warning
commit-queue@webkit.org [Mon, 16 Feb 2015 10:24:10 +0000 (10:24 +0000)]
[GTK] Loading page into WebView shows g_closure_unref warning
https://bugs.webkit.org/show_bug.cgi?id=127474

Patch by Milan Crha <mcrha@redhat.com> on 2015-02-16
Reviewed by Carlos Garcia Campos.

* bindings/gobject/GObjectEventListener.cpp:
(WebCore::GObjectEventListener::gobjectDestroyed):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180141 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[CSS Grid Layout] Remove the usage of Length(Undefined) in GridLength
svillar@igalia.com [Mon, 16 Feb 2015 09:42:25 +0000 (09:42 +0000)]
[CSS Grid Layout] Remove the usage of Length(Undefined) in GridLength
https://bugs.webkit.org/show_bug.cgi?id=141562

Reviewed by Chris Dumez.

From Blink r164154 by <timloh@chromium.org>

The patch simplifies some of the logic in creating and applying
GridLength and GridTrackSize values to RenderStyles by using the
constructors for initialisation instead of deferring to setters.

The rationale behind this change is that neither of
createGridTrackSize() nor createGridTrackBreadth() should be able
to fail, so these are changed to return objects directly instead
of taking an out reference (note that in general failing in
applying properties is incorrect and we should catch these in the
parser).

The quirk value across to the Lengths was also removed because it
doesn't apply to Grid.

* css/CSSPropertyNames.in:
* css/StyleBuilderConverter.h:
(WebCore::StyleBuilderConverter::createGridTrackBreadth): Return a
GridLength instead of a bool.
(WebCore::StyleBuilderConverter::createGridTrackSize): Return a
GridTrackSize instead of a bool.
(WebCore::StyleBuilderConverter::createGridTrackList): Do not
create temporary GridTrackSizes with undefined Lengths.
(WebCore::StyleBuilderConverter::convertGridTrackSize): Ditto.
* rendering/style/GridLength.h:
(WebCore::GridLength::GridLength): It now requires a mandatory
Length argument.
(WebCore::GridLength::length): Deleted.
(WebCore::GridLength::setFlex): Deleted.
* rendering/style/GridTrackSize.h:
(WebCore::GridTrackSize::GridTrackSize): It now requires a
mandatory GridLength argument.
(WebCore::GridTrackSize::length): Removed no longer valid ASSERTs.
(WebCore::GridTrackSize::minTrackBreadth): Ditto.
(WebCore::GridTrackSize::maxTrackBreadth): Ditto.
(WebCore::GridTrackSize::setLength): Deleted.
(WebCore::GridTrackSize::setMinMax): Deleted.
* rendering/style/RenderStyle.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180140 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoRemoving text node does not remove its associated markers
g.czajkowski@samsung.com [Mon, 16 Feb 2015 08:25:52 +0000 (08:25 +0000)]
Removing text node does not remove its associated markers
https://bugs.webkit.org/show_bug.cgi?id=140999

Reviewed by Ryosuke Niwa.

Removing text node via script does not remove its markers.
For example, running the following script:

<script>
    var div = document.getElementById("testElement");
    div.focus();
    document.execCommand("InsertText", false, "welllcome ");

    // Give a time async spellchecker to show its markers and
    // remove the node.
    setTimeout(function() { div.removeChild(div.firstChild); }, 100);
</script>

will cause that DocumentMarkerController still stores the markers for
detached node.

The same issue occurs when the text gets cleared, for example,
<script>
    input.value = ""; // for HTMLTextAreaElement, HTMLInputElement
    // or
    div.innerHTML = ""; // for content editable
</script>

No new tests. Internals marker APIs operate on text node attached
to the element. To test it we could expose document().markers().hasMarkers()
but there is no more useful cases where it could be used.
Another obstacle is that we are in isInShadowTree() so that there is
no possible to register on "DOMSubtreeModified" event.
Test cases attached to the bug:
1. input.value = ""  https://bug-140999-attachments.webkit.org/attachment.cgi?id=245704
2. elem.removeChild(textNode) https://bug-140999-attachments.webkit.org/attachment.cgi?id=246515

Remove markers when text nodes are about to remove.
* dom/Document.cpp:
(WebCore::Document::nodeChildrenWillBeRemoved):
Fixes input.value = ""
(WebCore::Document::nodeWillBeRemoved):
Fixes elem.removeChild(textNode)

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180139 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agojs/regress/script-tests/elidable-new-object-roflcopter.js should tier up sooner so...
fpizlo@apple.com [Mon, 16 Feb 2015 06:20:51 +0000 (06:20 +0000)]
js/regress/script-tests/elidable-new-object-roflcopter.js should tier up sooner so that it's more likely to test things
https://bugs.webkit.org/show_bug.cgi?id=141633

Reviewed by Michael Saboff.

More function invocations, fewer loop iterations, but otherwise keep the number of
allocations the same.

* js/regress/script-tests/elidable-new-object-roflcopter.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180138 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoRare case profiling should actually work
fpizlo@apple.com [Mon, 16 Feb 2015 06:19:00 +0000 (06:19 +0000)]
Rare case profiling should actually work
https://bugs.webkit.org/show_bug.cgi?id=141632

Reviewed by Michael Saboff.

This simple adjustment appears to be a 2% speed-up on Octane. Over time, the slow case
heuristic has essentially stopped working because the typical execution count threshold for a
bytecode instruction is around 66 while the slow case threshold is 100: virtually
guaranteeing that the DFG will never think that a bytecode instruction has taken the slow
case even if it took it every single time. So, this changes the slow case threshold to 20.

I checked if we could lower this down further, like to 10. That is worse than 20, and about
as bad as 100.

* runtime/Options.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180137 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUnreviewed, EFL gardening. Mark 5 form tests and an accessibility test to failure.
gyuyoung.kim@samsung.com [Mon, 16 Feb 2015 04:52:33 +0000 (04:52 +0000)]
Unreviewed, EFL gardening. Mark 5 form tests and an accessibility test to failure.

* platform/efl/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180136 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoREGRESSION (r180082): WebCore Debug builds fail on Mavericks due to weak export symbols
ddkilzer@apple.com [Mon, 16 Feb 2015 04:13:21 +0000 (04:13 +0000)]
REGRESSION (r180082): WebCore Debug builds fail on Mavericks due to weak export symbols
<http://webkit.org/b/141607>

Source/WebCore:

This should fix the build, but leaves a FIXME list in
WebCore.unexp.

* Configurations/WebCore.unexp: Add weak externals with a FIXME
statement so they can be removed later.  It's more important to
fix the build first.

* Configurations/WebCore.xcconfig: Change XCODE_VERSION_MINOR to
XCODE_VERSION_MAJOR since the internal bots use version 5.1.1
which still has the error.

Tools:

* Scripts/check-for-weak-vtables-and-externals: Ignore weak
externals in Source/WebCore/Configurations/WebCore.unexp when
building with Xcode 5.x.
(readXcode5SymbolsToIgnore): Add.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180135 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years ago[EFL] fix Evas GL engine error.
hyuki.kim@samsung.com [Mon, 16 Feb 2015 03:13:01 +0000 (03:13 +0000)]
[EFL] fix Evas GL engine error.
https://bugs.webkit.org/show_bug.cgi?id=141621

Reviewed by Gyuyoung Kim.

As new elm engine api was added from efl 1.10, replace elm_config_preferred_engine_set with elm_config_accel_preference_set.

* MiniBrowser/efl/main.c:
(elm_main):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180134 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoWeb Inspector: remove unused XHR replay code
said@apple.com [Mon, 16 Feb 2015 02:21:43 +0000 (02:21 +0000)]
Web Inspector: remove unused XHR replay code
https://bugs.webkit.org/show_bug.cgi?id=141622

Patch by Brian J. Burg <burg@cs.washington.edu> on 2015-02-15
Reviewed by Timothy Hatcher.

XHR Replay functionality became unused with the last frontend rewrite.
Remove instrumentation and data storage classes only used by this feature.

* inspector/InspectorInstrumentation.cpp:
(WebCore::InspectorInstrumentation::documentThreadableLoaderStartedLoadingForClientImpl): Deleted.
(WebCore::InspectorInstrumentation::willLoadXHRImpl): Deleted.
(WebCore::InspectorInstrumentation::didFailXHRLoadingImpl): Deleted.
* inspector/InspectorInstrumentation.h:
(WebCore::InspectorInstrumentation::documentThreadableLoaderStartedLoadingForClient): Deleted.
(WebCore::InspectorInstrumentation::willLoadXHR): Deleted.
(WebCore::InspectorInstrumentation::didFailXHRLoading): Deleted.
* inspector/InspectorResourceAgent.cpp:
(WebCore::InspectorResourceAgent::didLoadResourceFromMemoryCache):
(WebCore::InspectorResourceAgent::didFinishXHRLoading):
(WebCore::InspectorResourceAgent::documentThreadableLoaderStartedLoadingForClient): Deleted.
(WebCore::InspectorResourceAgent::willLoadXHR): Deleted.
(WebCore::InspectorResourceAgent::didFailXHRLoading): Deleted.
(WebCore::InspectorResourceAgent::replayXHR): Deleted.
* inspector/InspectorResourceAgent.h:
* inspector/NetworkResourcesData.cpp:
(WebCore::NetworkResourcesData::clear):
(WebCore::XHRReplayData::create): Deleted.
(WebCore::XHRReplayData::XHRReplayData): Deleted.
(WebCore::NetworkResourcesData::xhrReplayData): Deleted.
(WebCore::NetworkResourcesData::setXHRReplayData): Deleted.
(WebCore::NetworkResourcesData::reuseXHRReplayData): Deleted.
* inspector/NetworkResourcesData.h:
(WebCore::XHRReplayData::method): Deleted.
(WebCore::XHRReplayData::url): Deleted.
(WebCore::XHRReplayData::async): Deleted.
(WebCore::XHRReplayData::formData): Deleted.
(WebCore::XHRReplayData::headers): Deleted.
(WebCore::XHRReplayData::includeCredentials): Deleted.
(WebCore::NetworkResourcesData::ResourceData::xhrReplayData): Deleted.
(WebCore::NetworkResourcesData::ResourceData::setXHRReplayData): Deleted.
* loader/DocumentThreadableLoader.cpp:
(WebCore::DocumentThreadableLoader::loadRequest):
* xml/XMLHttpRequest.cpp:
(WebCore::XMLHttpRequest::createRequest):
(WebCore::XMLHttpRequest::internalAbort):
(WebCore::XMLHttpRequest::sendForInspectorXHRReplay): Deleted.
* xml/XMLHttpRequest.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180133 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoWeb Inspector: remove unused XHR replay code
burg@cs.washington.edu [Mon, 16 Feb 2015 02:17:38 +0000 (02:17 +0000)]
Web Inspector: remove unused XHR replay code
https://bugs.webkit.org/show_bug.cgi?id=141622

Reviewed by Timothy Hatcher.

Source/JavaScriptCore:

* inspector/protocol/Network.json: remove XHR replay methods.

Source/WebCore:

XHR Replay functionality became unused with the last frontend rewrite.
Remove instrumentation and data storage classes only used by this feature.

* inspector/InspectorInstrumentation.cpp:
(WebCore::InspectorInstrumentation::documentThreadableLoaderStartedLoadingForClientImpl): Deleted.
(WebCore::InspectorInstrumentation::willLoadXHRImpl): Deleted.
(WebCore::InspectorInstrumentation::didFailXHRLoadingImpl): Deleted.
* inspector/InspectorInstrumentation.h:
(WebCore::InspectorInstrumentation::documentThreadableLoaderStartedLoadingForClient): Deleted.
(WebCore::InspectorInstrumentation::willLoadXHR): Deleted.
(WebCore::InspectorInstrumentation::didFailXHRLoading): Deleted.
* inspector/InspectorResourceAgent.cpp:
(WebCore::InspectorResourceAgent::didLoadResourceFromMemoryCache):
(WebCore::InspectorResourceAgent::didFinishXHRLoading):
(WebCore::InspectorResourceAgent::documentThreadableLoaderStartedLoadingForClient): Deleted.
(WebCore::InspectorResourceAgent::willLoadXHR): Deleted.
(WebCore::InspectorResourceAgent::didFailXHRLoading): Deleted.
(WebCore::InspectorResourceAgent::replayXHR): Deleted.
* inspector/InspectorResourceAgent.h:
* inspector/NetworkResourcesData.cpp:
(WebCore::NetworkResourcesData::clear):
(WebCore::XHRReplayData::create): Deleted.
(WebCore::XHRReplayData::XHRReplayData): Deleted.
(WebCore::NetworkResourcesData::xhrReplayData): Deleted.
(WebCore::NetworkResourcesData::setXHRReplayData): Deleted.
(WebCore::NetworkResourcesData::reuseXHRReplayData): Deleted.
* inspector/NetworkResourcesData.h:
(WebCore::XHRReplayData::method): Deleted.
(WebCore::XHRReplayData::url): Deleted.
(WebCore::XHRReplayData::async): Deleted.
(WebCore::XHRReplayData::formData): Deleted.
(WebCore::XHRReplayData::headers): Deleted.
(WebCore::XHRReplayData::includeCredentials): Deleted.
(WebCore::NetworkResourcesData::ResourceData::xhrReplayData): Deleted.
(WebCore::NetworkResourcesData::ResourceData::setXHRReplayData): Deleted.
* loader/DocumentThreadableLoader.cpp:
(WebCore::DocumentThreadableLoader::loadRequest):
* xml/XMLHttpRequest.cpp:
(WebCore::XMLHttpRequest::createRequest):
(WebCore::XMLHttpRequest::internalAbort):
(WebCore::XMLHttpRequest::sendForInspectorXHRReplay): Deleted.
* xml/XMLHttpRequest.h:

Source/WebInspectorUI:

* UserInterface/Protocol/Legacy/7.0/InspectorBackendCommands.js:
* UserInterface/Protocol/Legacy/8.0/InspectorBackendCommands.js:
* Versions/Inspector-iOS-7.0.json:
* Versions/Inspector-iOS-8.0.json:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180132 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoUnreviewed, EFL gardening. Mark fixed position tests to failure.
gyuyoung.kim@samsung.com [Mon, 16 Feb 2015 02:16:23 +0000 (02:16 +0000)]
Unreviewed, EFL gardening. Mark fixed position tests to failure.

* platform/efl/TestExpectations: Rearrange tests order as well.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180131 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoREGRESSION (r180082): WebCore Debug builds fail on Mavericks due to weak export symbols
ddkilzer@apple.com [Mon, 16 Feb 2015 02:11:02 +0000 (02:11 +0000)]
REGRESSION (r180082): WebCore Debug builds fail on Mavericks due to weak export symbols
<http://webkit.org/b/141607>

More work towards fixing the Mavericks Debug build.

* inspector/ScriptDebugServer.h:
(Inspector::ScriptDebugServer::Task):
* inspector/agents/InspectorDebuggerAgent.h:
(Inspector::InspectorDebuggerAgent::Listener):
- Remove subclass exports. They did not help.

* runtime/JSCJSValue.h:
(JSC::JSValue::toFloat): Do not mark inline method for export.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180130 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoCrash when accessing an item in SVGTransformList and then removing a previous item...
said@apple.com [Mon, 16 Feb 2015 02:08:39 +0000 (02:08 +0000)]
Crash when accessing an item in SVGTransformList and then removing a previous item from this list.
https://bugs.webkit.org/show_bug.cgi?id=141550.

Reviewed by Darin Adler.

Source/WebCore:

Tests: LayoutTests/svg/dom/SVGTransformList-basics.xhtml: This test is modified to
include a new test case.

* svg/properties/SVGMatrixTearOff.h: m_value of SVGMatrixTearOff will be a null
pointer. There is no point in having SVGMatrixTearOff points to the parent and
the property of the parent at the same time.

(WebCore::SVGMatrixTearOff::create): SVGMatrixTearOff will hold a pointer to
the parent SVGPropertyTearOff<SVGTransform>. But it should overrides setValue()
and propertyReference() so it can set and get the SVGMatrix from the SVGTransform
parent.

(WebCore::SVGMatrixTearOff::SVGMatrixTearOff): Pass a nullptr to the base class.
SVGMatrixTearOff will act as a proxy of the parent. It does not hold any data by
itself but it knows what property to set and get from the parent.

* svg/properties/SVGPropertyTearOff.h:
(WebCore::SVGPropertyTearOff::create): Add a create method which can take a pointer value.

(WebCore::SVGPropertyTearOff::propertyReference):
(WebCore::SVGPropertyTearOff::setValue): Make these functions virtual so concrete classes
like SVGMatrixTearOff can override them.

(WebCore::SVGPropertyTearOff::SVGPropertyTearOff): Add a new constructor.

LayoutTests:

* svg/dom/SVGTransformList-basics-expected.txt:
* svg/dom/SVGTransformList-basics.xhtml: Add a new test case to this test. Have a
reference to an SVGMatrix of the last SVGTransform of an SVGTransformList and then
remove the items of this list from the beginning till the end.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180129 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoCrash when accessing an item in SVGLengthList and then replacing it with a previous...
said@apple.com [Mon, 16 Feb 2015 00:13:16 +0000 (00:13 +0000)]
Crash when accessing an item in SVGLengthList and then replacing it with a previous item in the list.
https://bugs.webkit.org/show_bug.cgi?id=141552.

Reviewed by Darin Adler.

Source/WebCore:

Tests: LayoutTests/svg/dom/SVGLengthList-basics.xhtml: This test is modified to
include a new test case.

* svg/properties/SVGListPropertyTearOff.h: Commit the removal of the replacing item
before trying to detach the wrapper of the item which going to be replaced.

LayoutTests:

* svg/dom/SVGLengthList-basics-expected.txt:
* svg/dom/SVGLengthList-basics.xhtml: Add a new test case to this test. Have a
reference to an SVGLength in an SVGLengthList and then replace this SVGLength
with another one which comes before it in the SVGLengthList.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180128 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoCoreText only needs to be soft-linked on Windows
ddkilzer@apple.com [Sun, 15 Feb 2015 20:44:07 +0000 (20:44 +0000)]
CoreText only needs to be soft-linked on Windows

More work towards the Maverick Debug build fix:

REGRESSION (r180082): WebCore Debug builds fail on Mavericks due to weak export symbols
<http://webkit.org/b/141607>

* page/CaptionUserPreferencesMediaAF.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180127 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoMore build fixing.
ap@apple.com [Sun, 15 Feb 2015 08:09:42 +0000 (08:09 +0000)]
More build fixing.

* platform/cocoa/TelephoneNumberDetectorCocoa.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180126 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoMore internal build fixing.
ap@apple.com [Sun, 15 Feb 2015 07:52:39 +0000 (07:52 +0000)]
More internal build fixing.

DDDFACache.h is not an actual private header, so we were using open source
redeclaration of DDDFACacheRef with an internal function prototype from DDDFAScanner.h.

* platform/cocoa/TelephoneNumberDetectorCocoa.cpp:
(WebCore::TelephoneNumberDetector::phoneNumbersScanner):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180125 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoBuild fix after <http://trac.webkit.org/changeset/180122>
ap@apple.com [Sun, 15 Feb 2015 07:16:28 +0000 (07:16 +0000)]
Build fix after <trac.webkit.org/changeset/180122>

* Platform/mac/StringUtilities.mm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180124 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoAdd the initial matching implementation for attribute selectors with case-insensitive...
benjamin@webkit.org [Sun, 15 Feb 2015 06:11:51 +0000 (06:11 +0000)]
Add the initial matching implementation for attribute selectors with case-insensitive value
https://bugs.webkit.org/show_bug.cgi?id=141615

Reviewed by Andreas Kling.

Source/WebCore:

Tests: fast/css/case-insensitive-attribute-selector-specificity.html
       fast/css/case-insensitive-attribute-selector-styling-html-1.html
       fast/css/case-insensitive-attribute-selector-styling-html-2.html
       fast/css/case-insensitive-attribute-selector-styling-html-3.html
       fast/css/case-insensitive-attribute-selector-styling-xhtml-1.xhtml
       fast/css/case-insensitive-attribute-selector-styling-xhtml-2.xhtml
       fast/css/case-insensitive-attribute-selector-styling-xhtml-3.xhtml
       fast/selectors/case-insensitive-attribute-bascis.html
       fast/selectors/case-insensitive-attribute-matching-style-attribute.html
       fast/selectors/case-insensitive-attribute-style-update.html
       fast/selectors/case-insensitive-attribute-with-case-sensitive-name.html

* css/SelectorChecker.cpp:
(WebCore::SelectorChecker::checkOne):
We already had case-insensitive value matching due to some legacy HTML
behaviors where some attribute values would be matched case-insensitively
depending on the attribute name.

For this patch, I just re-use the same mechanism. I used branches to try
to convey the idea that matching is case sensitive by default unless
the selector has a flag or we are in the weird HTML exception.

* cssjit/SelectorCompiler.cpp:
(WebCore::SelectorCompiler::constructFragmentsInternal):
Disable that case in the CSS JIT for now, I'll implement it later.

LayoutTests:

This covers the basics. I will add some more cases as I do the follow up patches.

I avoided the problem of non-ASCII characters, this will need its own follow up
patch that fixes all attribute matching.

* fast/css/case-insensitive-attribute-selector-specificity-expected.html: Added.
* fast/css/case-insensitive-attribute-selector-specificity.html: Added.
* fast/css/case-insensitive-attribute-selector-styling-html-1-expected.html: Added.
* fast/css/case-insensitive-attribute-selector-styling-html-1.html: Added.
* fast/css/case-insensitive-attribute-selector-styling-html-2-expected.html: Added.
* fast/css/case-insensitive-attribute-selector-styling-html-2.html: Added.
* fast/css/case-insensitive-attribute-selector-styling-html-3-expected.html: Added.
* fast/css/case-insensitive-attribute-selector-styling-html-3.html: Added.
* fast/css/case-insensitive-attribute-selector-styling-xhtml-1-expected.html: Added.
* fast/css/case-insensitive-attribute-selector-styling-xhtml-1.xhtml: Added.
* fast/css/case-insensitive-attribute-selector-styling-xhtml-2-expected.html: Added.
* fast/css/case-insensitive-attribute-selector-styling-xhtml-2.xhtml: Added.
* fast/css/case-insensitive-attribute-selector-styling-xhtml-3-expected.html: Added.
* fast/css/case-insensitive-attribute-selector-styling-xhtml-3.xhtml: Added.
* fast/selectors/case-insensitive-attribute-bascis-expected.txt: Added.
* fast/selectors/case-insensitive-attribute-bascis.html: Added.
* fast/selectors/case-insensitive-attribute-matching-style-attribute-expected.txt: Added.
* fast/selectors/case-insensitive-attribute-matching-style-attribute.html: Added.
* fast/selectors/case-insensitive-attribute-style-update-expected.txt: Added.
* fast/selectors/case-insensitive-attribute-style-update.html: Added.
* fast/selectors/case-insensitive-attribute-with-case-sensitive-name-expected.txt: Added.
* fast/selectors/case-insensitive-attribute-with-case-sensitive-name.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180123 268f45cc-cd09-0410-ab3c-d52691b4dbfc

5 years agoDeclare soft-linked functions with extern "C" linkage
ddkilzer@apple.com [Sun, 15 Feb 2015 05:31:53 +0000 (05:31 +0000)]
Declare soft-linked functions with extern "C" linkage

More work towards the Maverick Debug build fix:

REGRESSION (r180082): WebCore Debug builds fail on Mavericks due to weak export symbols
<http://webkit.org/b/141607>

* platform/mac/SoftLinking.h: Try declaring soft-linked
functions with extern "C" linkage to see if they stop appearing
as weak externals.  This has the added benefit that if the
function signature changes, we'll get a build failure.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@180122 268f45cc-cd09-0410-ab3c-d52691b4dbfc