WebKit-https.git
6 years ago[iOS WebKit2] Web process should try to shrink its memory footprint when going into...
akling@apple.com [Fri, 23 May 2014 01:36:10 +0000 (01:36 +0000)]
[iOS WebKit2] Web process should try to shrink its memory footprint when going into background.
<https://webkit.org/b/133197>
<rdar://problem/17011561>

Source/WebCore:
Make releaseMemory() public so we can call it from the process-will-suspend callback.

Reviewed by Gavin Barraclough.

* WebCore.exp.in:
* platform/MemoryPressureHandler.h:

Source/WebKit2:
Try to free up as much memory as possible before going into background.

Reviewed by Gavin Barraclough.

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

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169237 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoHook up a setting for showing detailed logging during memory pressure relief.
akling@apple.com [Fri, 23 May 2014 01:34:26 +0000 (01:34 +0000)]
Hook up a setting for showing detailed logging during memory pressure relief.
<https://webkit.org/b/133194>

Source/WebCore:
Make the pressure relief logger opt-in. Also make it flush memory back to the
OS (for both malloc and FastMalloc) to get more accurate numbers at each step.

Reviewed by Gavin Barraclough.

* WebCore.exp.in:
* platform/MemoryPressureHandler.cpp:
* platform/MemoryPressureHandler.h:
(WebCore::MemoryPressureHandler::ReliefLogger::ReliefLogger):
(WebCore::MemoryPressureHandler::ReliefLogger::~ReliefLogger):
(WebCore::MemoryPressureHandler::ReliefLogger::setLoggingEnabled):
* platform/cocoa/MemoryPressureHandlerCocoa.mm:
(WebCore::MemoryPressureHandler::ReliefLogger::platformMemoryUsage):

Source/WebKit2:
Plumb through the same setting that we used for detailed logging in WK1.

Reviewed by Gavin Barraclough.

* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::WebProcessCreationParameters):
(WebKit::WebProcessCreationParameters::encode):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/mac/WebContextMac.mm:
(WebKit::WebContext::platformInitializeWebProcess):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169236 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoDon't scan for phone numbers in editable regions
beidson@apple.com [Fri, 23 May 2014 01:29:30 +0000 (01:29 +0000)]
Don't scan for phone numbers in editable regions
<rdar://problem/16949846> and https://bugs.webkit.org/show_bug.cgi?id=133192

Reviewed by Enrica Casucci.

No new tests (Currently untested WK2-only feature)

* editing/Editor.cpp:
(WebCore::Editor::scanRangeForTelephoneNumbers): Skip the range if the Node is editable.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169235 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoCaptions layout incorrectly in fullscreen.
commit-queue@webkit.org [Fri, 23 May 2014 01:04:43 +0000 (01:04 +0000)]
Captions layout incorrectly in fullscreen.
https://bugs.webkit.org/show_bug.cgi?id=133175

Patch by Jeremy Jones <jeremyj@apple.com> on 2014-05-22
Reviewed by Eric Carlson.

When doing layout for fullscreen, geometry should be absolute so it is not influenced
by the rest of the page. This change adds style for CSSPropertyPosition, CSSPropertyLeft,
and CSSPropertyTop.

* html/shadow/MediaControlElements.cpp:
(WebCore::MediaControlTextTrackContainerElement::updateDisplay):
Refactor style changes into updateStyleForTextTrackRepresentation().
(WebCore::MediaControlTextTrackContainerElement::updateTimerFired):
Refactor style changes into updateStyleForTextTrackRepresentation().
(WebCore::MediaControlTextTrackContainerElement::clearTextTrackRepresentation):
Refactor style changes into updateStyleForTextTrackRepresentation().
(WebCore::MediaControlTextTrackContainerElement::updateStyleForTextTrackRepresentation):
Consolidate style changes here. Include the two existing and three new style changes.
* html/shadow/MediaControlElements.h:
Declare updateStyleForTextTrackRepresentation().

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169234 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[iOS] Send shareable resources to QuickLook if enabled
aestes@apple.com [Fri, 23 May 2014 00:40:23 +0000 (00:40 +0000)]
[iOS] Send shareable resources to QuickLook if enabled
https://bugs.webkit.org/show_bug.cgi?id=133189
<rdar://problem/17003995>

Reviewed by Brady Eidson.

Like we do for didReceiveData(), we need to pass an incoming ShareableResource to QuickLook if a QuickLook
handle exists.

* Shared/ShareableResource.cpp:
(WebKit::ShareableResource::Handle::tryWrapInCFData): Return the shared resource in a CFDataRef.
(WebKit::ShareableResource::Handle::tryWrapInSharedBuffer): Implemented in terms of tryWrapInCFData().
* Shared/ShareableResource.h:
* WebProcess/Network/WebResourceLoader.cpp:
(WebKit::WebResourceLoader::didReceiveResource): If a QuickLook handle exists, send the shareable resource to
it via a CFDataRef.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169233 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoFix the compositing/transitions/transform-on-large-layer.html test
simon.fraser@apple.com [Fri, 23 May 2014 00:39:36 +0000 (00:39 +0000)]
Fix the  compositing/transitions/transform-on-large-layer.html test
https://bugs.webkit.org/show_bug.cgi?id=133195

Reviewed by Dean Jackson.

This test used window resize when it doesn't need to and was flakey.

* compositing/transitions/transform-on-large-layer.html:
* platform/efl/TestExpectations:
* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169232 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION (r155977): Very stuttery 3D css animation on jtechcommunications.com
simon.fraser@apple.com [Thu, 22 May 2014 23:57:46 +0000 (23:57 +0000)]
REGRESSION (r155977): Very stuttery 3D css animation on jtechcommunications.com
https://bugs.webkit.org/show_bug.cgi?id=133179
<rdar://problem/16864666>

Reviewed by Dean Jackson.

Source/WebCore:

Revert code added in r155977 to try to pick a good contentsScale for layers
based on a root-relative transform. This resulted in pages with too much
backing store, and too many cases where layers repainted during animations,
causing stutter.

* platform/graphics/ca/GraphicsLayerCA.cpp:
(WebCore::GraphicsLayerCA::GraphicsLayerCA):
(WebCore::GraphicsLayerCA::flushCompositingState):
(WebCore::GraphicsLayerCA::recursiveCommitChanges):
(WebCore::GraphicsLayerCA::commitLayerChangesBeforeSublayers):
(WebCore::GraphicsLayerCA::updateContentsOpaque):
(WebCore::GraphicsLayerCA::updateAnimations):
(WebCore::GraphicsLayerCA::appendToUncommittedAnimations):
(WebCore::GraphicsLayerCA::setTransformAnimationEndpoints):
(WebCore::GraphicsLayerCA::setTransformAnimationKeyframes):
(WebCore::GraphicsLayerCA::updateContentsScale):
(WebCore::clampedContentsScaleForScale): Deleted.
(WebCore::maxScaleFromTransform): Deleted.
(WebCore::GraphicsLayerCA::updateRootRelativeScale): Deleted.
(WebCore::GraphicsLayerCA::getTransformFromAnimationsWithMaxScaleImpact): Deleted.
* platform/graphics/ca/GraphicsLayerCA.h:

LayoutTests:

Rebaseline.

* compositing/contents-scale/animating-expected.txt:
* compositing/contents-scale/rounded-contents-scale-expected.txt:
* compositing/contents-scale/scaled-ancestor-expected.txt:
* compositing/contents-scale/simple-scale-expected.txt:
* compositing/contents-scale/z-translate-expected.txt:
* platform/mac/compositing/overflow/composited-scrolling-paint-phases-expected.txt:
* platform/mac/compositing/tiling/rotated-tiled-clamped-expected.txt:
* platform/mac/compositing/tiling/rotated-tiled-preserve3d-clamped-expected.txt:
* platform/mac/compositing/visible-rect/nested-transform-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169229 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[iOS] Enable -apple-system- styled elements to respond to system font size changes.
mhock@apple.com [Thu, 22 May 2014 21:46:23 +0000 (21:46 +0000)]
[iOS] Enable -apple-system- styled elements to respond to system font size changes.
https://bugs.webkit.org/show_bug.cgi?id=133186
<rdar://problem/16583782>

Reviewed by Enrica Casucci.

Source/WebCore:
* WebCore.exp.in:
* WebCore.xcodeproj/project.pbxproj:
* rendering/RenderThemeIOS.h:
* rendering/RenderThemeIOS.mm:
(WebCore::_contentSizeCategory):
(WebCore::RenderThemeIOS::contentSizeCategory):
(WebCore::RenderThemeIOS::setContentSizeCategory):

Source/WebKit2:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView initWithFrame:configuration:]):
(-[WKWebView _contentSizeCategoryDidChange:]):
(-[WKWebView _contentSizeCategory]):
* UIProcess/WebPageProxy.h:
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::contentSizeCategoryDidChange):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::contentSizeCategoryDidChange):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169223 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION(r154797): Debugger crashes when stepping over an uncaught exception.
mark.lam@apple.com [Thu, 22 May 2014 21:40:21 +0000 (21:40 +0000)]
REGRESSION(r154797): Debugger crashes when stepping over an uncaught exception.
<https://webkit.org/b/133182>

Reviewed by Oliver Hunt.

Source/JavaScriptCore:
Before r154797, we used to clear the VM exception before calling into the
debugger.  After r154797, we don't.  This patch will restore this clearing
of the exception before calling into the debugger.

Also added assertions after returning from calls into the debugger to
ensure that the debugger did not introduce any exceptions.

* interpreter/Interpreter.cpp:
(JSC::unwindCallFrame):
(JSC::Interpreter::unwind):
(JSC::Interpreter::debug):
- Fixed the assertion here.  Interpreter::debug() should never be called
  with a pending exception.  Debugger callbacks for exceptions should be
  handled by Interpreter::unwind() and Interpreter::unwindCallFrame().

LayoutTests:
* inspector-protocol/debugger/regress-133182-expected.txt: Added.
* inspector-protocol/debugger/regress-133182.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169221 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoAdd option to run-jsc-stress-tests to use installed jsc
msaboff@apple.com [Thu, 22 May 2014 21:11:55 +0000 (21:11 +0000)]
Add option to run-jsc-stress-tests to use installed jsc
https://bugs.webkit.org/show_bug.cgi?id=133102

Reviewed by Geoffrey Garen.

Added --no-copy option to not copy a JavaScriptCore framework, but use the one
provied with the --jsc option.

* Scripts/run-jsc-stress-tests:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169217 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[iOS] Add CFBundleAllowMixedLocalizations to WebContentService
ap@apple.com [Thu, 22 May 2014 20:35:23 +0000 (20:35 +0000)]
[iOS] Add CFBundleAllowMixedLocalizations to WebContentService
https://bugs.webkit.org/show_bug.cgi?id=133187
<rdar://problem/16561084>

Reviewed by Tim Horton.

* WebProcess/EntryPoint/mac/XPCService/WebContentService/Info-iOS.plist:
We already had it in WebContentService.Development, just not here.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169214 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[MSE] Stored samples are not freed when SourceBuffer is removed from MediaSource
jer.noble@apple.com [Thu, 22 May 2014 18:25:46 +0000 (18:25 +0000)]
[MSE] Stored samples are not freed when SourceBuffer is removed from MediaSource
https://bugs.webkit.org/show_bug.cgi?id=133174

Reviewed by Eric Carlson.

Clear out stored MediaSamples from SourceBuffer's TrackBuffer storage when
aborting loading. Also, report the memory cost of those samples, so that the
SourceBuffer will be GCd more readily.

Add a mechanism for reporting the size of a MediaSample:
* Modules/mediasource/SourceBuffer.h:
* platform/MediaSample.h:
* platform/graphics/avfoundation/objc/SourceBufferPrivateAVFObjC.mm:
* platform/mock/mediasource/MockSourceBufferPrivate.cpp:

Track the memory usage of a SampleMap when adding and removing samples, as well
as allowing the SampleMap to be cleared wholesale:
* Modules/mediasource/SampleMap.cpp:
(WebCore::SampleMap::clear): Release all stored MediaSamples.
(WebCore::SampleMap::addSample): Update m_totalSize.
(WebCore::SampleMap::removeSample): Ditto.
* Modules/mediasource/SampleMap.h:
(WebCore::SampleMap::SampleMap): Initialize m_totalSize.
(WebCore::SampleMap::sizeInBytes): Simple accessor.

Clear the stored samples when loading is aborted, and report the extra memory
cost
* Modules/mediasource/SourceBuffer.cpp:
(WebCore::SourceBuffer::SourceBuffer): Initialize m_reportedExtraMemoryCost.
(WebCore::SourceBuffer::removedFromMediaSource): Clear all stored samples.
(WebCore::SourceBuffer::sourceBufferPrivateAppendComplete): Call reportExtraMemoryCost().
(WebCore::SourceBuffer::reportExtraMemoryCost): Inform the vm of the new
    extra memory cost incurred by the object.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169209 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[iOS][WK2] Application cache is broken
ap@apple.com [Thu, 22 May 2014 18:23:21 +0000 (18:23 +0000)]
[iOS][WK2] Application cache is broken
https://bugs.webkit.org/show_bug.cgi?id=133172
<rdar://problem/16994593>

Reviewed by Andy Estes.

* UIProcess/mac/WebContextMac.mm:
(WebKit::WebContext::platformDefaultApplicationCacheDirectory): Use a proper
path for application cache database (same as WebKit1).

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169208 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION (WebKit2): Keyboard should have Search button in duckduckgo.com.
enrica@apple.com [Thu, 22 May 2014 18:03:54 +0000 (18:03 +0000)]
REGRESSION (WebKit2): Keyboard should have Search button in duckduckgo.com.
https://bugs.webkit.org/show_bug.cgi?id=133183
<rdar://problem/17004207>

Reviewed by Geoff Garen.

Source/WebCore:
Adding new export.

* WebCore.exp.in:

Source/WebKit2:
The keyboard type should be search if the input type is search
or if it is text and it is inside a form with an action and either
name, id or title contain the word search. This matches the
heuristics we have in WK1 for iOS.

* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::getAssistedNodeInformation):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169207 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION(r163712): [GTK] Misspelling and grammar underline marks are no longer...
carlosgc@webkit.org [Thu, 22 May 2014 16:23:05 +0000 (16:23 +0000)]
REGRESSION(r163712): [GTK] Misspelling and grammar underline marks are no longer drawn
https://bugs.webkit.org/show_bug.cgi?id=133047

Reviewed by Darin Adler.

Change two conditions changed by mistake in r163712.

* editing/TextCheckingHelper.cpp:
(WebCore::findMisspellings): Enter the loop also when wordStart is 0.
(WebCore::TextCheckingHelper::findFirstMisspelling): Skip the work
when the text is a single character. Also reworked it to use a for
loop to improve the readability.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169203 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed EFL gardening
m.pakula@samsung.com [Thu, 22 May 2014 15:51:47 +0000 (15:51 +0000)]
Unreviewed EFL gardening

Add test expectations for failing tests.

* platform/efl-wk2/TestExpectations:
* platform/efl/editing/apply-style-iframe-crash-expected.txt: Added after r168641.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169202 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoVideo is resumed with old playback rate.
commit-queue@webkit.org [Thu, 22 May 2014 14:57:59 +0000 (14:57 +0000)]
Video is resumed with old playback rate.
https://bugs.webkit.org/show_bug.cgi?id=132905

Patch by Piotr Grad <p.grad@samsung.com> on 2014-05-22
Reviewed by Philippe Normand.

Source/WebCore:
Setting '0' playback rate is causing pipeline to pause.
GStreamer player impl. exposed this information to upper layers but it should not.
Solution is to hidden such situation behind m_playbackRatePause flag.

Test: media/video-paused-0-rate.html

* platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
(WebCore::MediaPlayerPrivateGStreamer::MediaPlayerPrivateGStreamer):
(WebCore::MediaPlayerPrivateGStreamer::play):
(WebCore::MediaPlayerPrivateGStreamer::pause):
(WebCore::MediaPlayerPrivateGStreamer::doSeek):
(WebCore::MediaPlayerPrivateGStreamer::updatePlaybackRate):
(WebCore::MediaPlayerPrivateGStreamer::paused):
(WebCore::MediaPlayerPrivateGStreamer::setRate):
(WebCore::MediaPlayerPrivateGStreamer::updateStates):
* platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.h:

LayoutTests:
* media/video-paused-0-rate-expected.txt: Added.
* media/video-paused-0-rate.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169200 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoText markers don't paint on simple lines
antti@apple.com [Thu, 22 May 2014 13:05:20 +0000 (13:05 +0000)]
Text markers don't paint on simple lines
https://bugs.webkit.org/show_bug.cgi?id=133177

Reviewed by Anders Carlsson.

Source/WebCore:

Marker painting code does not yet support simple lines.

Tests: fast/text/mark-matches-rendering-simple-lines.html
       fast/text/mark-matches-rendering.html

* dom/DocumentMarkerController.cpp:
(WebCore::DocumentMarkerController::addMarker):

    Force text blocks with markers to use line boxes.

* testing/Internals.cpp:
(WebCore::Internals::setMarkedTextMatchesAreHighlighted):

    Expose this so we can make reftests for marker rendering.

* testing/Internals.h:
* testing/Internals.idl:

LayoutTests:

* fast/text/mark-matches-rendering-expected.html: Added.
* fast/text/mark-matches-rendering-simple-lines-expected.html: Added.
* fast/text/mark-matches-rendering-simple-lines.html: Added.
* fast/text/mark-matches-rendering.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169199 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[CSS Grid Layout] Split the grid-item-margin-auto-columns-rows.html test.
commit-queue@webkit.org [Thu, 22 May 2014 12:01:32 +0000 (12:01 +0000)]
[CSS Grid Layout] Split the grid-item-margin-auto-columns-rows.html test.
https://bugs.webkit.org/show_bug.cgi?id=131064

Patch by Javier Fernandez <jfernandez@igalia.com> on 2014-05-22
Reviewed by Benjamin Poulain.

Splitting the "grid-item-margin-auto-columns-rows" into different writing-modes
related tests, so besides increasing clarity, it makes it fit in the 800x600 viewport.

* fast/css-grid-layout/grid-item-margin-auto-columns-rows-expected.html: Removed.
* fast/css-grid-layout/grid-item-margin-auto-columns-rows.html: Removed.
* fast/css-grid-layout/grid-item-margin-auto-columns-rows-horizontal-ltr-expected.html: Added.
* fast/css-grid-layout/grid-item-margin-auto-columns-rows-horizontal-ltr.html: Added.
* fast/css-grid-layout/grid-item-margin-auto-columns-rows-horizontal-rtl-expected.html: Added.
* fast/css-grid-layout/grid-item-margin-auto-columns-rows-horizontal-rtl.html: Added.
* fast/css-grid-layout/grid-item-margin-auto-columns-rows-vertical-ltr-expected.html: Added.
* fast/css-grid-layout/grid-item-margin-auto-columns-rows-vertical-ltr.html: Added.
* fast/css-grid-layout/grid-item-margin-auto-columns-rows-vertical-rtl-expected.html: Added.
* fast/css-grid-layout/grid-item-margin-auto-columns-rows-vertical-rtl.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169197 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[CSS Grid Layout] Guard RenderObject::isRenderGrid() method
rego@igalia.com [Thu, 22 May 2014 11:54:24 +0000 (11:54 +0000)]
[CSS Grid Layout] Guard RenderObject::isRenderGrid() method
https://bugs.webkit.org/show_bug.cgi?id=132380

Reviewed by Benjamin Poulain.

Guard RenderObject::isRenderGrid() method under ENABLE_CSS_GRID_LAYOUT compilation flag.

* rendering/RenderBox.cpp:
(WebCore::RenderBox::computeLogicalWidthInRegion):
* rendering/RenderObject.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169195 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Win64] Unreviewed build fix.
achristensen@apple.com [Thu, 22 May 2014 06:44:44 +0000 (06:44 +0000)]
[Win64] Unreviewed build fix.

* WebKit.vcxproj/WebKitExportGenerator/WebKitExports.def.in:
Corrected 64-bit linker symbol.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169190 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION(r167870): Crash in simple line layout code with :after
antti@apple.com [Thu, 22 May 2014 06:12:34 +0000 (06:12 +0000)]
REGRESSION(r167870): Crash in simple line layout code with :after
https://bugs.webkit.org/show_bug.cgi?id=133155

Source/WebCore:
<rdar://problem/16977696>

Reviewed by Darin Adler.

Fix https://bugs.webkit.org/show_bug.cgi?id=132241 in a safer way.
The underline behavior is tested by the existing fast/text/simple-lines-hover-underline.html

Test: fast/text/simple-lines-hover-after.html

* rendering/RenderBlock.cpp:
(WebCore::RenderBlock::invalidateLineLayoutPath): Deleted.

    Move to RenderBlockFlow.

* rendering/RenderBlock.h:
(WebCore::RenderBlock::invalidateLineLayoutPath):
* rendering/RenderBlockFlow.cpp:
(WebCore::RenderBlockFlow::styleDidChange):

    Invalidate layout if style changes in a manner that makes us ineligible to use the simple line layout path.

(WebCore::RenderBlockFlow::invalidateLineLayoutPath):

    Drop the simple line layout on path invalidation if it exists. It may not be valid anymore.
    Also invalidate the layout if this happens so we'll reconstruct the lines later.

(WebCore::RenderBlockFlow::simpleLineLayout): Deleted.
(WebCore::RenderBlockFlow::ensureLineBoxes):
(WebCore::RenderBlockFlow::createLineBoxes): Deleted.

    Revert some of the changes made it r167870.

* rendering/RenderBlockFlow.h:
(WebCore::RenderBlockFlow::simpleLineLayout):

    Add strong validity assert.

LayoutTests:

Reviewed by Darin Adler.

* fast/text/simple-lines-hover-after-expected.html: Added.
* fast/text/simple-lines-hover-after.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169189 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoStore barrier elision should run after DCE in both the DFG path and the FTL path
fpizlo@apple.com [Thu, 22 May 2014 01:36:23 +0000 (01:36 +0000)]
Store barrier elision should run after DCE in both the DFG path and the FTL path
https://bugs.webkit.org/show_bug.cgi?id=129718

Rubber stamped by Mark Hahnenberg.

* dfg/DFGPlan.cpp:
(JSC::DFG::Plan::compileInThreadImpl):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169184 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[iOS] two media control button strings are not localized
eric.carlson@apple.com [Thu, 22 May 2014 01:18:27 +0000 (01:18 +0000)]
[iOS] two media control button strings are not localized
https://bugs.webkit.org/show_bug.cgi?id=133160

Reviewed by Geoffrey Garen.

* English.lproj/mediaControlsLocalizedStrings.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169183 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[EFL] Add include path of compact_unwind_encoding.h if FTL JIT is enabled
commit-queue@webkit.org [Thu, 22 May 2014 00:12:47 +0000 (00:12 +0000)]
[EFL] Add include path of compact_unwind_encoding.h if FTL JIT is enabled
https://bugs.webkit.org/show_bug.cgi?id=132907

Patch by Zsolt Borbely <zsborbely.u-szeged@partner.samsung.com> on 2014-05-21
Reviewed by Gyuyoung Kim.

.:
* Source/cmake/FindLIBCXXABI.cmake: Added.
* Source/cmake/OptionsEfl.cmake:

Source/JavaScriptCore:
* CMakeLists.txt:

Tools:
We need the mach-o/compact_unwind_encoding.h header from libc++abi-dev package
in JavaScriptCore/ftl/FTLUnwindInfo.cpp when FTL JIT is enabled.

* efl/install-dependencies:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169181 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoOnly enable sandbox extensions on mac.
oliver@apple.com [Thu, 22 May 2014 00:11:07 +0000 (00:11 +0000)]
Only enable sandbox extensions on mac.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169180 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed Mac gardening after r169161.
zalan@apple.com [Wed, 21 May 2014 21:50:39 +0000 (21:50 +0000)]
Unreviewed Mac gardening after r169161.

* platform/mac-wk2/tiled-drawing/tile-size-slow-zoomed-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169178 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[iOS] Update some of the media controls buttons to be bigger
dino@apple.com [Wed, 21 May 2014 21:35:53 +0000 (21:35 +0000)]
[iOS] Update some of the media controls buttons to be bigger
https://bugs.webkit.org/show_bug.cgi?id=133158
<rdar://problem/16475828>

Reviewed by Simon Fraser.

The buttons in the media controls on iOS are too small
to be comfortably pressed. Make them the size of the controls
panel and the expected iOS button size (44x44px).

The artwork for the buttons now includes the padding. I've
added a file from which you can edit the artwork, then
copy it into the CSS (because editing directly in CSS is
a huge pain).

So far just the play/pause and fullscreen button have been
updated. There will be a follow-up patch to fix the remaining
buttons and the scrubber.

* Modules/mediacontrols/assets-apple-iOS.svg: Added.
* Modules/mediacontrols/mediaControlsiOS.css:
(audio::-webkit-media-controls-panel):
(audio::-webkit-media-controls-fullscreen-button):
(audio::-webkit-media-controls-play-button):
(audio::-webkit-media-controls-play-button:active):
(audio::-webkit-media-controls-play-button.paused):
(audio::-webkit-media-controls-fullscreen-button:active):
(@media only screen and (-webkit-min-device-pixel-ratio: 2)): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169177 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[iOS] Enable sandboxing for the database process
oliver@apple.com [Wed, 21 May 2014 20:52:21 +0000 (20:52 +0000)]
[iOS] Enable sandboxing for the database process
https://bugs.webkit.org/show_bug.cgi?id=132963

Reviewed by Alexey Proskuryakov.

Simply ensure that the database process is running in a tight
sandbox. As part of this i've separated out ChildProcessIOS
and ensures that the database process loads a custom sandbox.

* Configurations/WebKit.xcconfig:
* DatabaseProcess/ios/DatabaseProcessIOS.mm:
(WebKit::DatabaseProcess::initializeSandbox):
* DatabaseProcess/ios/com.apple.WebKit.DatabasesIOS.sb: Added.
* Shared/ios/ChildProcessIOS.mm: Added.
Essentially cloned from the Mac version, only all the OSX
specific logic has been stripped.
(WebKit::ChildProcess::setApplicationIsDaemon):
(WebKit::ChildProcess::platformInitialize):
(WebKit::ChildProcess::initializeSandbox):
(WebKit::ChildProcess::setQOS):
* Shared/mac/ChildProcessMac.mm:
Remove all the IOS ifdefs.
(WebKit::ChildProcess::setApplicationIsDaemon):
(WebKit::ChildProcess::platformInitialize):
(WebKit::ChildProcess::initializeSandbox):
(WebKit::ChildProcess::setQOS):
* WebKit2.xcodeproj/project.pbxproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169176 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[iOS] Create a UIPrintFormatter for WKWebView
aestes@apple.com [Wed, 21 May 2014 20:39:36 +0000 (20:39 +0000)]
[iOS] Create a UIPrintFormatter for WKWebView
https://bugs.webkit.org/show_bug.cgi?id=133005

Reviewed by Tim Horton.

Implemented a UIPrintFormatter subclass to support printing WKWebViews using UIPrintInteractionController.
Overrode -_recalcPageCount: to synchronously ask the web process to lay out for printing and return the number
of page rects. The web process returns the page rects immediately and then starts rendering the page to a PDF,
returning the PDF data to the UI process asynchronously. When UIPrintInteractionController later calls our
overridden -drawInRect:forPageAtIndex:, we block and wait for the PDF data to be returned. Since
UIPrintInteractionController draws each page after a delay, it's possible that the web process has already
returned the PDF data to us by the time we print the first page.

* Shared/WebPreferencesStore.h: Correctly initialized shouldPrintBackgrounds for iOS.
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _dataForDisplayedPDF]): Now that WKPDFView vends its CGPDFDocument directly, extract its data here.
(-[WKWebView _printFormatterClass]): Tell UIView which UIPrintFormatter to alloc.
(-[WKWebView _computePageCountAndStartDrawingToPDFWithPrintInfo:WebKit::firstPage:computedTotalScaleFactor:]):
If we're displaying a PDF, return its page count directly. Otherwise, synchronously ask the web process to
compute page rects for printing and then return the number of rects.
(-[WKWebView _endPrinting]): Cleared state and sent the WebPage::EndPrinting message to the web process.
(-[WKWebView _printedDocument]): If we're displaying a PDF, return its data directly. Otherwise, if we haven't
received PDF data from the web process yet, wait for it indefinitely (there's no sensible timeout to use here
since failing to receive the data results in printing some number of blank pages).
(-[WKWebView _setPrintedDocument:]): Set the CGPDFDocument received from the web process.
* UIProcess/API/Cocoa/WKWebViewInternal.h: Declared methods called by WKWebViewPrintFormatter.
* UIProcess/PageClient.h: Declared didFinishDrawingPagesToPDF.
* UIProcess/WKWebViewPrintFormatter.h: Added.
* UIProcess/WKWebViewPrintFormatter.mm: Added.
(-[WKWebViewPrintFormatter dealloc]): Cleared state and called -[WKWebView _endPrinting].
(-[WKWebViewPrintFormatter webView]): Returned the WKWebView for this print formatter.
(-[WKWebViewPrintFormatter _recalcPageCount]): Set up a PrintInfo object given the formatter's content rect
and called -_computePageCountAndStartDrawingToPDFWithPrintInfo:...
(-[WKWebViewPrintFormatter rectForPageAtIndex:]): Called -_recalcIfNecessary and then returned the rect from
-_pageContentRect.
(-[WKWebViewPrintFormatter drawInRect:forPageAtIndex:]): Taught to draw the given page of the printed PDF
document into the given rect and context provided to us.
* UIProcess/WebPageProxy.h: Declared didFinishDrawingPagesToPDF.
* UIProcess/WebPageProxy.messages.in: Added DidFinishDrawingPagesToPDF. Made this a delayed syncrhonous message
so that additoinal work could be performed after sending the response.
* UIProcess/ios/PageClientImplIOS.h:
* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::didFinishDrawingPagesToPDF): Created a CGPDFDocument from the data sent to us by the
web process and called -[WKWebView _setPrintedDocument:].
* UIProcess/ios/WKPDFView.h:
* UIProcess/ios/WKPDFView.mm:
(-[WKPDFView pdfDocument]): Returned the underlying CGPDFDocument.
(-[WKPDFView documentData]): Deleted.
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::didFinishDrawingPagesToPDF):
* WebKit2.xcodeproj/project.pbxproj:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::computePagesForPrinting): Changed to call computePagesForPrintingImpl.
(WebKit::WebPage::computePagesForPrintingImpl): Moved the common logic of computePagesForPrinting to here so it
could be called by computePagesForPrintingAndStartDrawingToPDF.
(WebKit::WebPage::drawPagesToPDF): Changed to call drawPagesToPDFImpl
(WebKit::WebPage::drawPagesToPDFImpl): Moved the common logic of drawPagesToPDF to here so it could be called
by computePagesForPrintingAndStartDrawingToPDF.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::computePagesForPrintingAndStartDrawingToPDF): Computed page rects and returned them as a
synchronous response. Afterwards called drawPagesToPDFImpl() and returned the PDF data by sending
WebPageProxy::DidFinishDrawingPagesToPDF.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169175 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed. Actually enable snapshot tests on mac wk2.
roger_fong@apple.com [Wed, 21 May 2014 19:45:30 +0000 (19:45 +0000)]
Unreviewed. Actually enable snapshot tests on mac wk2.
        https://bugs.webkit.org/show_bug.cgi?id=131871

        * platform/mac-wk1/TestExpectations: Skip them here.
        * platform/mac/TestExpectations: Unskip them here.

        Some slight modifications to some of the tests to reduce flakiness of tests.
        * plugins/snapshotting/autoplay-dominant.html:
        * plugins/snapshotting/autoplay-plugin-blocked-by-image.html:
        * plugins/snapshotting/autoplay-plugin-mostly-blocked-by-image.html:
        * plugins/snapshotting/snapshot-plugin-not-quite-blocked-by-image.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169173 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[EFL] Turn on ENABLE_CSS_FILTERS
m.pakula@samsung.com [Wed, 21 May 2014 18:22:28 +0000 (18:22 +0000)]
[EFL] Turn on ENABLE_CSS_FILTERS
https://bugs.webkit.org/show_bug.cgi?id=133153

Reviewed by Gyuyoung Kim.

.:
* Source/cmake/OptionsEfl.cmake: Enable CSS_FILTERS as default option value.

Source/WebKit2:
Make WebKit2 compile after r166741 when CSS_FILTERS are enabled.

* Shared/CoordinatedGraphics/CoordinatedGraphicsArgumentCoders.cpp:
(IPC::ArgumentCoder<WebCore::FilterOperations>::decode):

LayoutTests:
Remove failure test expectations for tests passing after turning on CSS_FILTERS.

* platform/efl-wk2/TestExpectations:
* platform/efl/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169172 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoSpecial handling of telephone number detection menu.
beidson@apple.com [Wed, 21 May 2014 17:37:48 +0000 (17:37 +0000)]
Special handling of telephone number detection menu.
<rdar://problem/16967971> and https://bugs.webkit.org/show_bug.cgi?id=133143

Reviewed by Geoff Garen.

Add a bool to ContextMenuContextData to represent telephone number menus:
* Shared/ContextMenuContextData.cpp:
(WebKit::ContextMenuContextData::ContextMenuContextData):
(WebKit::ContextMenuContextData::operator=):
(WebKit::ContextMenuContextData::encode):
(WebKit::ContextMenuContextData::decode):
* Shared/ContextMenuContextData.h:
(WebKit::ContextMenuContextData::ContextMenuContextData):
(WebKit::ContextMenuContextData::isTelephoneNumberContext):

* UIProcess/mac/WebContextMenuProxyMac.mm:
(WebKit::WebContextMenuProxyMac::showContextMenu): If the menu is a telephone
  number menu then display it using a specific AppKit API.

* UIProcess/mac/WebPageProxyMac.mm:
(WebKit::WebPageProxy::showTelephoneNumberMenu): Filter the array of menu items
  through a new WKSI call, and then show it using a telephone number context.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169170 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Mac] REGRESSION (Async text input): Asian text input doesn't work in Flash
ap@apple.com [Wed, 21 May 2014 17:04:12 +0000 (17:04 +0000)]
[Mac] REGRESSION (Async text input): Asian text input doesn't work in Flash
https://bugs.webkit.org/show_bug.cgi?id=133128
<rdar://problem/16890920>

Reviewed by Anders Carlsson.

* UIProcess/API/mac/WKView.mm:
(-[WKView _interpretKeyEvent:completionHandler:]): Added a separate code path for
plug-in input. Before async text input, we used to provide a nil input context due
to having _data->_interpretKeyEventsParameters while this function was executing,
and thus took a shortcut. The new behavior is different from pre-async in that
we don't collect commands at all, but I couldn't find any case where that mattered.
(-[WKView inputContext]): The _collectedKeypressCommands check was not a correct
replacement for _interpretKeyEventsParameters one, because this variable is not
always set within -_interpretKeyEvent:completionHandler:.

* UIProcess/mac/WKTextInputWindowController.mm:
(-[WKTextInputPanel _interpretKeyEvent:usingLegacyCocoaTextInput:string:]):
While at it, fixed a separate issue with Cangjie predictive input. We shouldn't
keep prediction pop-up on screen after hiding the bottom input window.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169167 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed Mac gardening after r169161.
zalan@apple.com [Wed, 21 May 2014 16:56:40 +0000 (16:56 +0000)]
Unreviewed Mac gardening after r169161.

* platform/mac-wk2/tiled-drawing/tile-size-slow-zoomed-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169166 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[CMake] Improve handling of LIB_INSTALL_DIR, EXEC_INSTALL_DIR, and LIBEXEC_INSTALL_DIR
mrobinson@webkit.org [Wed, 21 May 2014 16:39:17 +0000 (16:39 +0000)]
[CMake] Improve handling of LIB_INSTALL_DIR, EXEC_INSTALL_DIR, and LIBEXEC_INSTALL_DIR
https://bugs.webkit.org/show_bug.cgi?id=132819

Reviewed by Carlos Garcia Campos.

.:
* CMakeLists.txt: Hard-code CMAKE_ARCHIVE_OUTPUT_DIRECTORY etc, so that we can accept absolute paths for LIB_INSTALL_DIR
and friends.
* Source/cmake/OptionsCommon.cmake: Get rid of LIB_SUFFIX and try to make all path variables absolute. Also set
CMAKE_ARCHIVE_OUTPUT_DIRECTORY, etc using the last path component of LIB_INSTALL_DIR, etc, so that absolute paths
are supported.
* Source/cmake/OptionsGTK.cmake: Ditto. Also get rid of all pkg-config specific variables.

Source/JavaScriptCore:
* javascriptcoregtk.pc.in: Instead of using the special pkg-config variables,
use the common CMake ones directly.

Source/WebKit/efl:
* EWebKitConfig.cmake.in: LIB_INSTALL_DIR is always absolute now, so we don't need
to prefix it with the prefix.
* ewebkit.pc.in: Instead of hard-coding the path to the library installation directory use LIB_INSTALL_DIR.

Source/WebKit2:
* PlatformEfl.cmake: No need to prefix EXEC_INSTALL_DIR with the prefix any longer.
* PlatformGtk.cmake: No longer use CMAKE_INSTALL_FULL_FOO variables, instead using the WebKit-specific ones.
* efl/EWebKit2Config.cmake.in: Ditto.
* efl/ewebkit2.pc.in: Instead of hard-coding the path to the library installation directory use LIB_INSTALL_DIR.
* webkit2gtk-web-extension.pc.in: No longer use the pkg-config only variables. Instead use
the common CMake ones.
* webkit2gtk.pc.in: Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169165 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed, roll out http://trac.webkit.org/changeset/169159.
fpizlo@apple.com [Wed, 21 May 2014 16:33:37 +0000 (16:33 +0000)]
Unreviewed, roll out trac.webkit.org/changeset/169159.

This was a unilateral change and wasn't properly reviewed.

PerformanceTests/SunSpider:
* profiler-test.yaml:

Source/JavaScriptCore:
* tests/mozilla/mozilla-tests.yaml:

LayoutTests:
* js/script-tests/function-apply-many-args.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169164 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION(r168625): [GTK] Assertion failure in LocalStorageDatabaseTracker
carlosgc@webkit.org [Wed, 21 May 2014 16:19:20 +0000 (16:19 +0000)]
REGRESSION(r168625): [GTK] Assertion failure in LocalStorageDatabaseTracker
https://bugs.webkit.org/show_bug.cgi?id=133118

Reviewed by Philippe Normand.

Call WebContext::applyPlatformSpecificConfigurationDefaults() to
initialize the WebContextConfiguration before creating the default WebContext.

* UIProcess/API/gtk/WebKitWebContext.cpp:
(createDefaultWebContext):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169163 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoArray.prototype.find and findIndex should skip holes
graouts@webkit.org [Wed, 21 May 2014 16:14:55 +0000 (16:14 +0000)]
Array.prototype.find and findIndex should skip holes
https://bugs.webkit.org/show_bug.cgi?id=132658

Reviewed by Geoffrey Garen.

Source/JavaScriptCore:
Skip holes in the array when iterating such that callback isn't called.

* builtins/Array.prototype.js:
(find):
(findIndex):

LayoutTests:
* js/array-find-expected.txt:
* js/array-findIndex-expected.txt:
* js/script-tests/array-find.js:
* js/script-tests/array-findIndex.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169162 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoGarbage when rubber-banding at the right edge of a page zoomed to non-integral scale.
zalan@apple.com [Wed, 21 May 2014 16:06:55 +0000 (16:06 +0000)]
Garbage when rubber-banding at the right edge of a page zoomed to non-integral scale.
https://bugs.webkit.org/show_bug.cgi?id=133139
<rdar://problem/16503353>

Reviewed by Simon Fraser.

Do not pixel align the root content layer. The alignment code expands the graphics's layer size
which makes the right and bottom tiles bigger than the content.
Painting the body's background color produces pixel cracks, because the content can not
fill the expanded tiles completely.

Not testable.

* platform/graphics/GraphicsLayerClient.h:
(WebCore::GraphicsLayerClient::needsPixelAligment):
* platform/graphics/ca/GraphicsLayerCA.cpp:
(WebCore::GraphicsLayerCA::updateGeometry):
(WebCore::GraphicsLayerCA::computePixelAlignment):
* platform/graphics/ca/mac/TileGrid.mm:
(WebCore::TileGrid::rectForTileIndex):
* rendering/RenderLayerBacking.cpp:
(WebCore::RenderLayerBacking::needsPixelAligment):
* rendering/RenderLayerBacking.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169161 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION (r168046): Invalid layout in WebCore::RenderBox::containingBlockLogicalWid...
stavila@adobe.com [Wed, 21 May 2014 15:31:20 +0000 (15:31 +0000)]
REGRESSION (r168046): Invalid layout in WebCore::RenderBox::containingBlockLogicalWidthForPositioned
https://bugs.webkit.org/show_bug.cgi?id=132933

Reviewed by Darin Adler.

Source/WebCore:
Invalid layout is performed when calling containingBlockLogicalWidthForPositioned for a RenderTableSection object.

Test: fast/multicol/newmulticol/table-section-crash.html

* rendering/RenderBox.cpp:
(WebCore::RenderBox::containingBlockLogicalWidthForPositioned):

LayoutTests:
Added test for crash caused by invalid layout for RenderTableSection.

* fast/multicol/newmulticol/table-section-crash-expected.txt: Added.
* fast/multicol/newmulticol/table-section-crash.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169160 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION(r169092 and r169102): Skip failing JSC tests on ARM64 properly
commit-queue@webkit.org [Wed, 21 May 2014 14:56:52 +0000 (14:56 +0000)]
REGRESSION(r169092 and r169102): Skip failing JSC tests on ARM64 properly
https://bugs.webkit.org/show_bug.cgi?id=133149

Patch by Eva Balazsfalvi <evab.u-szeged@partner.samsung.com> on 2014-05-21
Reviewed by Csaba Osztrogonác.

PerformanceTests/SunSpider:
* profiler-test.yaml:

Source/JavaScriptCore:
* tests/mozilla/mozilla-tests.yaml:

LayoutTests:
* js/script-tests/function-apply-many-args.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169159 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed EFL gardening
m.pakula@samsung.com [Wed, 21 May 2014 14:43:44 +0000 (14:43 +0000)]
Unreviewed EFL gardening

Rebaseline test expectations after r169048.

* platform/efl/css1/box_properties/float_on_text_elements-expected.txt:
* platform/efl/css1/text_properties/text_indent-expected.txt:
* platform/efl/css2.1/t0905-c414-flt-wrap-00-e-expected.txt:
* platform/efl/css2.1/t0905-c5525-fltwidth-00-c-g-expected.txt:
* platform/efl/css2.1/t100801-c544-valgn-00-a-ag-expected.txt:
* platform/efl/css2.1/t100801-c544-valgn-02-d-agi-expected.txt:
* platform/efl/css2.1/t100801-c544-valgn-03-d-agi-expected.txt:
* platform/efl/css2.1/t100801-c544-valgn-04-d-agi-expected.txt:
* platform/efl/css2.1/t1601-c547-indent-00-b-a-expected.txt:
* platform/efl/fast/block/float/026-expected.txt:
* platform/efl/fast/block/float/028-expected.txt:
* platform/efl/fast/css/empty-pseudo-class-expected.txt:
* platform/efl/fast/css/first-child-pseudo-class-expected.txt:
* platform/efl/fast/css/first-of-type-pseudo-class-expected.txt:
* platform/efl/fast/css/last-child-pseudo-class-expected.txt:
* platform/efl/fast/css/last-of-type-pseudo-class-expected.txt:
* platform/efl/fast/css/only-child-pseudo-class-expected.txt:
* platform/efl/fast/css/only-of-type-pseudo-class-expected.txt:
* platform/efl/fast/line-grid/line-align-left-edges-expected.txt:
* platform/efl/fast/multicol/column-count-with-rules-expected.txt:
* platform/efl/fast/multicol/float-multicol-expected.txt:
* platform/efl/fast/multicol/layers-in-multicol-expected.txt:
* platform/efl/fast/multicol/vertical-lr/float-multicol-expected.txt:
* platform/efl/fast/multicol/vertical-rl/float-multicol-expected.txt:
* platform/efl/fast/repaint/line-flow-with-floats-1-expected.txt:
* platform/efl/fast/repaint/line-flow-with-floats-10-expected.txt:
* platform/efl/fast/repaint/line-flow-with-floats-2-expected.txt:
* platform/efl/fast/repaint/line-flow-with-floats-3-expected.txt:
* platform/efl/fast/repaint/line-flow-with-floats-4-expected.txt:
* platform/efl/fast/repaint/line-flow-with-floats-5-expected.txt:
* platform/efl/fast/repaint/line-flow-with-floats-6-expected.txt:
* platform/efl/fast/repaint/line-flow-with-floats-7-expected.txt:
* platform/efl/fast/repaint/line-flow-with-floats-8-expected.txt:
* platform/efl/fast/repaint/line-flow-with-floats-9-expected.txt:
* platform/efl/fast/text/whitespace/024-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169158 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[CMake] Support building with Debug Fission
gns@gnome.org [Wed, 21 May 2014 13:08:35 +0000 (13:08 +0000)]
[CMake] Support building with Debug Fission
https://bugs.webkit.org/show_bug.cgi?id=131177

Reviewed by Philippe Normand.

* Source/cmake/OptionsCommon.cmake: add a DEBUG_FISSION option to enable usage
of http://gcc.gnu.org/wiki/DebugFission.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169157 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION(r169092): It broke run-jsc-stress tests on non Mac platforms
ossy@webkit.org [Wed, 21 May 2014 12:01:42 +0000 (12:01 +0000)]
REGRESSION(r169092): It broke run-jsc-stress tests on non Mac platforms
https://bugs.webkit.org/show_bug.cgi?id=133145

Patch by Eva Balazsfalvi <evab.u-szeged@partner.samsung.com> on 2014-05-21
Reviewed by Csaba Osztrogonác.

* Scripts/run-jsc-stress-tests:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169156 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed EFL gardening.
jinwoo7.song@samsung.com [Wed, 21 May 2014 11:55:30 +0000 (11:55 +0000)]
Unreviewed EFL gardening.

* platform/efl/fast/css/absolute-child-with-percent-height-inside-relative-parent-expected.txt: Added.
Rebaseline after r169048.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169155 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[GTK] Unreviewed GTK gardening
berto@igalia.com [Wed, 21 May 2014 09:28:54 +0000 (09:28 +0000)]
[GTK] Unreviewed GTK gardening

Patch by Lorenzo Tilve <ltilve@igalia.com> on 2014-05-21

* platform/gtk/TestExpectations: Report and mark new tests failing.
* platform/gtk/fast/css/absolute-child-with-percent-height-inside-relative-parent-expected.txt: Added.
Rebaseline after r169048.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169154 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[EFL] Ensure EGLOffScreenContext::initialize calls platformMakeCurrent on the surface
simon.pena@samsung.com [Wed, 21 May 2014 09:11:14 +0000 (09:11 +0000)]
[EFL] Ensure EGLOffScreenContext::initialize calls platformMakeCurrent on the surface
https://bugs.webkit.org/show_bug.cgi?id=124832

Reviewed by Antonio Gomes.

The original code was only calling platformMakeCurrent when the
GL_EXT_robustness extension is supported: the alternate code path
would only create a context, but it wouldn't make that context
current. This patch ensures platformMakeCurrent is also called for
the alternate code path, and, if it fails to make the context
current, will destroy the context. Effectively, this makes both
code paths behave in the same way.

* platform/graphics/surfaces/egl/EGLContext.cpp:
(WebCore::EGLOffScreenContext::initialize):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169153 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[WebKit2] Cleanup the build from unused variable in WebProcess module
commit-queue@webkit.org [Wed, 21 May 2014 08:12:50 +0000 (08:12 +0000)]
[WebKit2] Cleanup the build from unused variable in WebProcess module
https://bugs.webkit.org/show_bug.cgi?id=133062

Patch by Shivakumar JM <shiva.jm@samsung.com> on 2014-05-21
Reviewed by Csaba Osztrogonác.

Fix unused parameter by using UNUSED_PARAM macro

* WebProcess/UserContent/WebUserContentController.cpp:
(WebKit::WebUserContentController::addUserScriptMessageHandlers):
(WebKit::WebUserContentController::removeUserScriptMessageHandler):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169152 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoRolled out <http://trac.webkit.org/changeset/166184>
ggaren@apple.com [Wed, 21 May 2014 06:55:02 +0000 (06:55 +0000)]
Rolled out <trac.webkit.org/changeset/166184>
https://bugs.webkit.org/show_bug.cgi?id=133144

Reviewed by Gavin Barraclough.

It caused a performance regression.

../JavaScriptCore:
* heap/BlockAllocator.cpp:
(JSC::BlockAllocator::blockFreeingThreadStartFunc):

../WTF:
* wtf/FastMalloc.cpp:
(WTF::TCMalloc_PageHeap::runScavengerThread):
* wtf/Threading.h:
* wtf/ThreadingPthreads.cpp:
(WTF::createThreadInternal):
(WTF::setCurrentThreadQOSUtility): Deleted.
* wtf/ThreadingWin.cpp:
(WTF::setCurrentThreadQOSUtility): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169151 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoVersioning.
lforschler@apple.com [Wed, 21 May 2014 06:01:19 +0000 (06:01 +0000)]
Versioning.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169150 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Mac] DataCues do not work properly when rewinding video
bfulgham@apple.com [Wed, 21 May 2014 04:08:33 +0000 (04:08 +0000)]
[Mac] DataCues do not work properly when rewinding video
https://bugs.webkit.org/show_bug.cgi?id=133138
<rdar://problem/16979086>

Reviewed by Eric Carlson.

Make the TrackPrivateBase responsible for knowing if a type of track needs a non-zero
startTimeVariance value.

Also, correct a bug in the equality test for DataCue objects.

* html/track/DataCue.cpp:
(WebCore::DataCue::isEqual): Handle the JSValue data types consistently, so that we don't attempt
to compare a JSNull against a default JSValue object.
* html/track/InbandTextTrack.cpp:
(WebCore::InbandTextTrack::startTimeVariance): Added.
* html/track/InbandTextTrack.h:
* html/track/TextTrack.cpp:
(WebCore::TextTrack::hasCue): Use new startTimeVariance method location.
* html/track/TextTrack.h:
(WebCore::TextTrack::startTimeVariance): Added.
* html/track/TextTrackCue.cpp:
(WebCore::TextTrackCue::hasEquivalentStartTime): Use new startTimeVariance location.
* html/track/TextTrackCue.h:
(WebCore::TextTrackCue::startTimeVariance): Deleted.
* html/track/TextTrackCueGeneric.h:
* platform/graphics/TrackPrivateBase.h:
(WebCore::TrackPrivateBase::startTimeVariance): Added.
* platform/graphics/avfoundation/InbandTextTrackPrivateAVF.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169146 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoDFG prediction propagation should agree with fixup phase over the return type of...
fpizlo@apple.com [Wed, 21 May 2014 03:49:16 +0000 (03:49 +0000)]
DFG prediction propagation should agree with fixup phase over the return type of GetByVal
https://bugs.webkit.org/show_bug.cgi?id=133134

Reviewed by Mark Hahnenberg.

Make prediction propagator use ArrayMode refinement to decide the return type.

Also introduce a heap prediction intrinsic that allows us to test weird corner cases
like this. The only way we'll see a mismatch like this in the real world is probably
through a gnarly race condition.

* dfg/DFGByteCodeParser.cpp:
(JSC::DFG::ByteCodeParser::handleIntrinsic):
* dfg/DFGNode.h:
(JSC::DFG::Node::setHeapPrediction):
* dfg/DFGPredictionPropagationPhase.cpp:
(JSC::DFG::PredictionPropagationPhase::propagate):
* jsc.cpp:
(GlobalObject::finishCreation):
(functionFalse1):
(functionFalse2):
(functionUndefined1):
(functionUndefined2):
(functionFalse): Deleted.
(functionOtherFalse): Deleted.
(functionUndefined): Deleted.
* runtime/Intrinsic.h:
* tests/stress/get-by-val-double-predicted-int.js: Added.
(foo):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169145 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION (WK2): Tapping on the input field in duckduckgo.com shows the keyboard...
enrica@apple.com [Wed, 21 May 2014 00:13:30 +0000 (00:13 +0000)]
REGRESSION (WK2): Tapping on the input field in duckduckgo.com shows the keyboard but not the caret.
https://bugs.webkit.org/show_bug.cgi?id=133140
<rdar://problem/16980898>

Reviewed by Benjamin Poulain.

When a page focuses a field on load, we receive the call to startAssistingNode
but we don't create a text interaction assistant if the focus is not originated
by a user action. Upon receiving the selection change notification we don't
call the text interaction assistant, since we don't have one, but we do call
the web selection assistant. We don't have any valid rects for web selection
and we clear the selection on the page, causing the caret not to appear and
the typing to have no effect.

* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _updateChangedSelection]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169142 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoWatchdog timer should be lazily allocated
mhahnenberg@apple.com [Tue, 20 May 2014 22:55:49 +0000 (22:55 +0000)]
Watchdog timer should be lazily allocated
https://bugs.webkit.org/show_bug.cgi?id=133135

Reviewed by Geoffrey Garen.

Source/JavaScriptCore:
We incur a noticeable amount of overhead on some benchmarks due to checking if the Watchdog ever fired.
There is no reason to do this checking if we never activated the Watchdog, which can only be done through
JSContextGroupSetExecutionTimeLimit or JSContextGroupClearExecutionTimeLimit.

By allocating the Watchdog lazily on the VM we can avoid all of the associated overhead when we don't use
these two API functions (which is true of most clients).

* API/JSContextRef.cpp:
(JSContextGroupSetExecutionTimeLimit):
(JSContextGroupClearExecutionTimeLimit):
* dfg/DFGByteCodeParser.cpp:
(JSC::DFG::ByteCodeParser::parseBlock):
* dfg/DFGSpeculativeJIT32_64.cpp:
(JSC::DFG::SpeculativeJIT::compile):
* dfg/DFGSpeculativeJIT64.cpp:
(JSC::DFG::SpeculativeJIT::compile):
* interpreter/Interpreter.cpp:
(JSC::Interpreter::execute):
(JSC::Interpreter::executeCall):
(JSC::Interpreter::executeConstruct):
* jit/JITOpcodes.cpp:
(JSC::JIT::emit_op_loop_hint):
(JSC::JIT::emitSlow_op_loop_hint):
* jit/JITOperations.cpp:
* llint/LLIntSlowPaths.cpp:
(JSC::LLInt::LLINT_SLOW_PATH_DECL):
* runtime/VM.h:
* runtime/Watchdog.cpp:
(JSC::Watchdog::Scope::Scope): Deleted.
(JSC::Watchdog::Scope::~Scope): Deleted.
* runtime/Watchdog.h:
(JSC::Watchdog::Scope::Scope):
(JSC::Watchdog::Scope::~Scope):

Source/WebCore:
No new tests.

We incur a noticeable amount of overhead on some benchmarks due to checking if the Watchdog ever fired.
There is no reason to do this checking if we never activated the Watchdog, which can only be done through
JSContextGroupSetExecutionTimeLimit or JSContextGroupClearExecutionTimeLimit.

By allocating the Watchdog lazily on the VM we can avoid all of the associated overhead when we don't use
these two API functions (which is true of most clients).

* bindings/js/JSEventListener.cpp:
(WebCore::JSEventListener::handleEvent):
* bindings/js/WorkerScriptController.cpp:
(WebCore::WorkerScriptController::evaluate):
(WebCore::WorkerScriptController::scheduleExecutionTermination):
(WebCore::WorkerScriptController::isExecutionTerminating):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169139 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Mac] Allow popup menus to override default appearance
dino@apple.com [Tue, 20 May 2014 21:03:54 +0000 (21:03 +0000)]
[Mac] Allow popup menus to override default appearance
https://bugs.webkit.org/show_bug.cgi?id=133129

Reviewed by Tim Horton.

Source/WebCore:
WebKitSystemInterface's WKPopupMenu has a flag to hide
the arrows at the end of a popup-menu (e.g. <select>).
Expose that via adding a hasDefaultAppearance flag to
PopupMenuStyle, which then passes it on to WKSI.

* platform/PopupMenuStyle.h:
(WebCore::PopupMenuStyle::PopupMenuStyle): New flag in constructor.
(WebCore::PopupMenuStyle::hasDefaultAppearance): New flag.
* platform/mac/WebCoreSystemInterface.h: wkPopupMenu signature has changed to accept
the new flag.
* platform/mac/WebCoreSystemInterface.mm: Ditto.
* rendering/RenderMenuList.cpp:
(RenderMenuList::itemStyle): Pass in true as default appearance when creating
the PopupMenuStyle. We don't need it.
(RenderMenuList::menuStyle): In this case we pass in the negated value of
style().hasAppearance(), which indicates if we are overriding the built-in drawing.
* rendering/RenderSearchField.cpp:
(WebCore::RenderSearchField::menuStyle): Use true as the default value.

Source/WebKit/mac:
Replace the call to wkPopupMenuWithSize with wkPopupMenu, now
with extra parameters to indicate the control size and whether
or not to draw the arrows at the end.

* WebCoreSupport/PopupMenuMac.mm:
(PopupMenuMac::show):

Source/WebKit2:
Add an extra "hideArrows" flag to PlatformPopupMenuData
and replace the call to WKPopupMenuWithSize with WKPopupMenu.

* Shared/PlatformPopupMenuData.cpp: Encode and decode the new data value.
(WebKit::PlatformPopupMenuData::encode):
(WebKit::PlatformPopupMenuData::decode):
* Shared/PlatformPopupMenuData.h: Add hideArrows boolean.
* UIProcess/mac/WebPopupMenuProxyMac.mm:
(WebKit::WebPopupMenuProxyMac::showPopupMenu):
* WebProcess/WebCoreSupport/mac/WebPopupMenuMac.mm:
(WebKit::WebPopupMenu::setUpPlatformData): The hideArrows flag is
true if we don't have the default appearance.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169134 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Mac] Allow popup menus to override default appearance
dino@apple.com [Tue, 20 May 2014 20:59:58 +0000 (20:59 +0000)]
[Mac] Allow popup menus to override default appearance
https://bugs.webkit.org/show_bug.cgi?id=133129

Reviewed by Tim Horton.

Add a new parameter to WKPopupMenu.

* WebKitSystemInterface.h:
* libWebKitSystemInterfaceMavericks.a:
* libWebKitSystemInterfaceMountainLion.a:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169133 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Mac] AVAssets are never destroyed; lack of an autorelease pool when calling callOnMa...
jer.noble@apple.com [Tue, 20 May 2014 20:57:17 +0000 (20:57 +0000)]
[Mac] AVAssets are never destroyed; lack of an autorelease pool when calling callOnMainThread.
https://bugs.webkit.org/show_bug.cgi?id=133130

Reviewed by Geoff Garen.

Wrap the function to be called in an autorelease pool, so that autoreleased objects are cleaned
up immediately after the end of the function.

* wtf/mac/MainThreadMac.mm:
(WTF::timerFired):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169131 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Mac] do not deactivate an audio session that has running I/O
eric.carlson@apple.com [Tue, 20 May 2014 20:52:17 +0000 (20:52 +0000)]
[Mac] do not deactivate an audio session that has running I/O
https://bugs.webkit.org/show_bug.cgi?id=133127

Reviewed by Jer Noble.

Source/WebCore:
* page/Settings.h:
(WebCore::Settings::setShouldManageAudioSessionCategory): Renamed from setShouldManageAudioSession.
(WebCore::Settings::shouldManageAudioSessionCategory): Renamed from shouldManageAudioSession.
(WebCore::Settings::setShouldManageAudioSession): Deleted.
(WebCore::Settings::shouldManageAudioSession): Deleted.

* platform/audio/mac/MediaSessionManagerMac.cpp:
(MediaSessionManager::updateSessionState): Don't deactivate the session if there are any
    Video or Audio sessions.

* platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.mm:
(WebCore::MediaPlayerPrivateAVFoundationObjC::assetStatus): Drive-by change to log errors
    returned by -statusOfValueForKey:error: in debug builds.

Source/WebKit/mac:
* WebView/WebView.mm:
(-[WebView _commonInitializationWithFrameName:groupName:]): Settings::setShouldManageAudioSession
    renamed to setShouldManageAudioSessionCategory, deal with it.

Source/WebKit2:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::WebPage): Settings::setShouldManageAudioSession renamed to
    setShouldManageAudioSessionCategory, deal with it.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169130 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION: All non-mainframe scrollbars don't paint after r169065
bdakin@apple.com [Tue, 20 May 2014 20:36:27 +0000 (20:36 +0000)]
REGRESSION: All non-mainframe scrollbars don't paint after r169065
https://bugs.webkit.org/show_bug.cgi?id=133132
-and corresponding-
<rdar://problem/16968850>

Reviewed by Geoff Garen.

We should not universally return true here. This feature will only work for
scrollbars with layers.
* platform/ScrollableArea.h:
* platform/Scrollbar.cpp:
(WebCore::Scrollbar::supportsUpdateOnSecondaryThread):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169129 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoElement within flattened frame may update its scroll state during the layout phase...
dbates@webkit.org [Tue, 20 May 2014 20:34:35 +0000 (20:34 +0000)]
Element within flattened frame may update its scroll state during the layout phase of the wrong RenderView
https://bugs.webkit.org/show_bug.cgi?id=133013
<rdar://problem/16760154>

Reviewed by David Hyatt.

Source/WebCore:
Fixes an issue where the scroll state of an element may be updated during the layout of the wrong
RenderView. In particular, the scroll state of an element inside a CSS flex box in a flattened
frame f is updated during the layout of the RenderView associated with the parent frame of f instead
of during the layout of the RenderView associated with f.

The layout machinery assumes that the scroll state of each scrollable element is updated before the
completion of layout for its associated RenderView. Currently we have logic to defer updating the scroll
state of a scrollable element until completion of recursive layout. For a page with a flattened frame, we
defer such updates until completion of layout for all RenderViews along the ancestor frame hierarchy of
each flattened frame regardless of the RenderView associated with the element that originated the
deferred scroll state request. Instead, only the RenderView associated with the element that deferred its
scroll state update should dispatch the scroll state update.

Tests: fast/frames/flattening/scrollable-flexbox-inside-iframe-with-zero-height-and-needs-full-repaint-crash.html
       fast/frames/flattening/scrollable-flexbox-inside-iframe-with-zero-height-assertion-failure.html

* rendering/RenderBlock.cpp: Removed WTF::-prefix in typedef definition for ContinuationOutlineTableMap.
(WebCore::UpdateScrollInfoAfterLayoutTransaction::UpdateScrollInfoAfterLayoutTransaction): Added; a struct
that represents a RenderView v, nested transaction count, and a list of RenderBlocks in v that will need
to have their scroll state updated.
(WebCore::updateScrollInfoAfterLayoutTransactionStack): Added; returns the global stack of transactions.
(WebCore::RenderBlock::willBeDestroyed): Call removeFromUpdateScrollInfoAfterLayoutTransaction(), which
was formerly named as removeFromDelayedUpdateScrollInfoSet.
(WebCore::currentUpdateScrollInfoAfterLayoutTransaction): Added; returns the top-most transaction in the
global stack.
(WebCore::RenderBlock::beginUpdateScrollInfoAfterLayoutTransaction): Modified as appropriate to make use
of the global stack of transactions; formerly named startDelayUpdateScrollInfo.
(WebCore::RenderBlock::endAndCommitUpdateScrollInfoAfterLayoutTransaction): Modified as appropriate to
make use of the global stack of transactions; formerly named finishDelayUpdateScrollInfo.
(WebCore::RenderBlock::removeFromUpdateScrollInfoAfterLayoutTransaction): Modified as appropriate to make
use of the global stack of transactions; formerly named removeFromDelayedUpdateScrollInfoSet.
(WebCore::RenderBlock::updateScrollInfoAfterLayout): Modified as appropriate to make use of the global
stack of transactions.
(WebCore::RenderBlock::layout): Ditto.
(WebCore::RenderBlock::startDelayUpdateScrollInfo): Deleted.
(WebCore::RenderBlock::finishDelayUpdateScrollInfo): Deleted.
(WebCore::RenderBlock::removeFromDelayedUpdateScrollInfoSet): Deleted.
* rendering/RenderBlock.h:
* rendering/RenderBlockFlow.cpp:
(WebCore::RenderBlockFlow::willBeDestroyed): Call removeFromUpdateScrollInfoAfterLayoutTransaction(), which
was formerly named removeFromDelayedUpdateScrollInfoSet.
* rendering/RenderDeprecatedFlexibleBox.cpp:
(WebCore::RenderDeprecatedFlexibleBox::layoutHorizontalBox): Call {begin, end}UpdateScrollInfoAfterLayoutTransaction(),
which was formerly named {start, end}DelayUpdateScrollInfo.
(WebCore::RenderDeprecatedFlexibleBox::layoutVerticalBox): Ditto.
* rendering/RenderFlexibleBox.cpp:
(WebCore::RenderFlexibleBox::layoutBlock): Ditto.

LayoutTests:
Added tests to ensure that that we update the scroll state of an element during the layout phase of its
associated RenderView.

* fast/frames/flattening/scrollable-flexbox-inside-iframe-with-zero-height-and-needs-full-repaint-crash-expected.txt: Added.
* fast/frames/flattening/scrollable-flexbox-inside-iframe-with-zero-height-and-needs-full-repaint-crash.html: Added.
* fast/frames/flattening/scrollable-flexbox-inside-iframe-with-zero-height-assertion-failure-expected.txt: Added.
* fast/frames/flattening/scrollable-flexbox-inside-iframe-with-zero-height-assertion-failure.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169128 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION (r169065): Mountain Lion run-api-tests failures: ASSERTION FAILED:
bdakin@apple.com [Tue, 20 May 2014 19:31:01 +0000 (19:31 +0000)]
REGRESSION (r169065): Mountain Lion run-api-tests failures: ASSERTION FAILED:
Uncaught exception - -[NSRegularLegacyScrollerImp setPresentationValue:]:
unrecognized selector sent to instance 0x7ff51aa38000
https://bugs.webkit.org/show_bug.cgi?id=133121

Reviewed by Tim Horton.

* platform/mac/ScrollbarThemeMac.mm:
(WebCore::ScrollbarThemeMac::setPaintCharacteristicsForScrollbar):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169124 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION (r169063) Fixed and sticky nodes misplaced on scrolling sometimes
simon.fraser@apple.com [Tue, 20 May 2014 19:13:14 +0000 (19:13 +0000)]
REGRESSION (r169063) Fixed and sticky nodes misplaced on scrolling sometimes
https://bugs.webkit.org/show_bug.cgi?id=133106
<rdar://problem/16967648>

Reviewed by Sam Weinig.

Fix regression from r169063. That commit removed scrolledContentsLayers from
frame scrolling nodes, but they do actually use them.

So put them back; not in the base class, because they have a somewhat different
meaning for overflow scrolling and frame scrolling.

Source/WebCore:
* WebCore.exp.in:
* page/scrolling/AsyncScrollingCoordinator.cpp:
(WebCore::AsyncScrollingCoordinator::frameViewRootLayerDidChange):
(WebCore::AsyncScrollingCoordinator::updateFrameScrollingNode):
* page/scrolling/AsyncScrollingCoordinator.h:
* page/scrolling/ScrollingCoordinator.h:
(WebCore::ScrollingCoordinator::updateFrameScrollingNode):
* page/scrolling/ScrollingStateFrameScrollingNode.cpp:
(WebCore::ScrollingStateFrameScrollingNode::ScrollingStateFrameScrollingNode):
(WebCore::ScrollingStateFrameScrollingNode::setScrolledContentsLayer):
* page/scrolling/ScrollingStateFrameScrollingNode.h:
* page/scrolling/mac/ScrollingTreeFrameScrollingNodeMac.mm:
(WebCore::ScrollingTreeFrameScrollingNodeMac::updateBeforeChildren):
* rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::updateScrollCoordinatedLayer):

Source/WebKit2:
* Shared/Scrolling/RemoteScrollingCoordinatorTransaction.cpp:
(ArgumentCoder<ScrollingStateFrameScrollingNode>::encode):
(ArgumentCoder<ScrollingStateFrameScrollingNode>::decode):
(WebKit::RemoteScrollingTreeTextStream::dump):
* UIProcess/Scrolling/RemoteScrollingCoordinatorProxy.cpp:
(WebKit::RemoteScrollingCoordinatorProxy::connectStateNodeLayers):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169123 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Mac] WebProcess doesn't follow localization of UI process when run as a service
ap@apple.com [Tue, 20 May 2014 18:44:52 +0000 (18:44 +0000)]
[Mac] WebProcess doesn't follow localization of UI process when run as a service
https://bugs.webkit.org/show_bug.cgi?id=133126
<rdar://problem/13396515>
<rdar://problem/16561084>

Reviewed by Sam Weinig and Tim Horton.

Source/WebKit2:
* Shared/EntryPointUtilities/mac/XPCService/XPCServiceMain.Development.mm:
(WebKit::handleXPCBootstrap):
(main):
* Shared/EntryPointUtilities/mac/XPCService/XPCServiceMain.mm:
(WebKit::handleXPCBootstrap):
(main):
Set localization very early in process lifetime (in development builds, that's
even before WebKit framework is loaded).

* UIProcess/Launcher/mac/ProcessLauncherMac.mm: (WebKit::connectToService):
Pass UI process localization to the service, computing it in the same way we do
for child processes.

Tools:
* WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
(WTR::InjectedBundle::platformInitialize): It was too late to override language
here, and it's not necessary any more. Keeping the code for older OS versions in
case it did have effect in some code paths.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169122 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoJSArray::shiftCountWith* could be more efficient
mhahnenberg@apple.com [Tue, 20 May 2014 18:07:48 +0000 (18:07 +0000)]
JSArray::shiftCountWith* could be more efficient
https://bugs.webkit.org/show_bug.cgi?id=133011

Reviewed by Geoffrey Garen.

Source/JavaScriptCore:
Our current implementations of shiftCountWithAnyIndexingType and shiftCountWithArrayStorage
are scared of the presence of any holes in the array. We can mitigate this somewhat by enabling
them to correctly handle holes, thus avoiding the slowest of slow paths in most cases.

* runtime/ArrayStorage.h:
(JSC::ArrayStorage::indexingHeader):
(JSC::ArrayStorage::length):
(JSC::ArrayStorage::hasHoles):
* runtime/IndexingHeader.h:
(JSC::IndexingHeader::publicLength):
(JSC::IndexingHeader::from):
* runtime/JSArray.cpp:
(JSC::JSArray::shiftCountWithArrayStorage):
(JSC::JSArray::shiftCountWithAnyIndexingType):
(JSC::JSArray::unshiftCountWithArrayStorage):
* runtime/JSArray.h:
(JSC::JSArray::shiftCountForShift):
(JSC::JSArray::shiftCountForSplice):
(JSC::JSArray::shiftCount):
* runtime/Structure.cpp:
(JSC::Structure::holesRequireSpecialBehavior):
* runtime/Structure.h:

LayoutTests:
Added a performance regression test which tests the contiguous shift case. We're ~85% faster
on this microbenchmark with this patch.

Also added a bunch of tests for each permutation of contiguous/array storage with no holes, holes,
and holes that require special behavior (e.g. indexed properties in the prototype chain).

* js/array-storage-splice-holes-expected.txt: Added.
* js/array-storage-splice-holes-require-special-behavior-expected.txt: Added.
* js/array-storage-splice-holes-require-special-behavior.html: Added.
* js/array-storage-splice-holes.html: Added.
* js/array-storage-splice-no-holes-expected.txt: Added.
* js/array-storage-splice-no-holes.html: Added.
* js/contiguous-splice-holes-expected.txt: Added.
* js/contiguous-splice-holes-require-special-behavior-expected.txt: Added.
* js/contiguous-splice-holes-require-special-behavior.html: Added.
* js/contiguous-splice-holes.html: Added.
* js/contiguous-splice-no-holes-expected.txt: Added.
* js/contiguous-splice-no-holes.html: Added.
* js/regress/array-splice-contiguous-expected.txt: Added.
* js/regress/array-splice-contiguous.html: Added.
* js/regress/script-tests/array-splice-contiguous.js: Added.
(foo):
* js/script-tests/array-storage-splice-holes-require-special-behavior.js: Added.
* js/script-tests/array-storage-splice-holes.js: Added.
* js/script-tests/array-storage-splice-no-holes.js: Added.
* js/script-tests/contiguous-splice-holes-require-special-behavior.js: Added.
* js/script-tests/contiguous-splice-holes.js: Added.
* js/script-tests/contiguous-splice-no-holes.js: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169121 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION: [CSS Regions] Content flowed directly into the flow thread that ends...
stavila@adobe.com [Tue, 20 May 2014 17:38:04 +0000 (17:38 +0000)]
REGRESSION: [CSS Regions] Content flowed directly into the flow thread that ends up in the second region is not properly repainted
https://bugs.webkit.org/show_bug.cgi?id=133111

Reviewed by David Hyatt.

Source/WebCore:
When computing the repaint rect of an element flowed into a flow thread, if the element is flowed directly into
the flow thread (meaning its containing block is the flow thread itself), the region's position within the flow
must no longer be taken into consideration, because its already included in the element's |topLeft|.

Test: fast/regions/hover-element-flowed-second-region.html

* rendering/RenderBox.cpp:
(WebCore::RenderBox::computeRectForRepaint):

LayoutTests:
Added test for the correct repaint of an element flowed directly into the second region of a flow thread.

* fast/regions/hover-element-flowed-second-region-expected.html: Added.
* fast/regions/hover-element-flowed-second-region.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169120 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoFix web timing assertion failure.
commit-queue@webkit.org [Tue, 20 May 2014 17:29:35 +0000 (17:29 +0000)]
Fix web timing assertion failure.
https://bugs.webkit.org/show_bug.cgi?id=133094
<rdar://problem/16966032>

Patch by Alex Christensen <achristensen@webkit.org> on 2014-05-20
Reviewed by Alexey Proskuryakov.

* platform/network/mac/ResourceHandleMac.mm:
(WebCore::ResourceHandle::getConnectionTimingData):
Set requestStart and responseStart to 0 instead of -1 to match the
ResourceLoadTiming constructor and prevent the assertion failure in
PerformanceTiming::responseStart.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169119 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoOnly set title on SVG documents
commit-queue@webkit.org [Tue, 20 May 2014 17:13:35 +0000 (17:13 +0000)]
Only set title on SVG documents
https://bugs.webkit.org/show_bug.cgi?id=133068

Patch by Prashant Hiremath <hiremathprashants@gmail.com> on 2014-05-20
Reviewed by Dirk Schulze.

Source/WebCore:
Test: svg/custom/html_document_set_title.html

* svg/SVGTitleElement.cpp:
(WebCore::SVGTitleElement::insertedInto):
(WebCore::SVGTitleElement::removedFrom):
(WebCore::SVGTitleElement::childrenChanged):

LayoutTests:
* svg/custom/html_document_set_title-expected.txt: Added.
* svg/custom/html_document_set_title.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169118 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION(r169105): fast/regions/cssom/region-range-for-box-crash.html is more crash...
ap@apple.com [Tue, 20 May 2014 17:08:59 +0000 (17:08 +0000)]
REGRESSION(r169105): fast/regions/cssom/region-range-for-box-crash.html is more crashy than before
https://bugs.webkit.org/show_bug.cgi?id=133124

* TestExpectations: Skip it.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169117 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoSkip http/tests/inspector/replay/document-last-modified-fallback-value.html since...
andersca@apple.com [Tue, 20 May 2014 17:06:29 +0000 (17:06 +0000)]
Skip http/tests/inspector/replay/document-last-modified-fallback-value.html since it asserts.

* TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169116 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Curl] Invalid content in cache file, causes broken rendering.
commit-queue@webkit.org [Tue, 20 May 2014 17:01:47 +0000 (17:01 +0000)]
[Curl] Invalid content in cache file, causes broken rendering.
https://bugs.webkit.org/show_bug.cgi?id=133069

Patch by peavo@outlook.com <peavo@outlook.com> on 2014-05-20
Reviewed by Brent Fulgham.

When data for a url is received as multiple parts, the cache file for the url is truncated when opened for writing,
and will only contain the last part of data received.
This is fixed by only opening the file once, and close it after all data has been received.

* platform/network/curl/CurlCacheEntry.cpp:
(WebCore::CurlCacheEntry::CurlCacheEntry): Initialize cache file handle member.
(WebCore::CurlCacheEntry::~CurlCacheEntry): Close cache file.
(WebCore::CurlCacheEntry::saveCachedData): Only open cache file once to avoid truncating.
(WebCore::CurlCacheEntry::didFail): Close cache file.
(WebCore::CurlCacheEntry::didFinishLoading): Ditto.
(WebCore::CurlCacheEntry::openContentFile): Added method to open cache file.
(WebCore::CurlCacheEntry::closeContentFile): Added method to close cache file.
* platform/network/curl/CurlCacheEntry.h: Added file handle member, and methods to open and close cache file.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169115 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoFix a header comment
andersca@apple.com [Tue, 20 May 2014 16:59:53 +0000 (16:59 +0000)]
Fix a header comment
https://bugs.webkit.org/show_bug.cgi?id=133123
<rdar://problem/16973783>

Reviewed by Geoffrey Garen.

* UIProcess/API/Cocoa/WKUserContentController.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169114 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoMove WebKitTestRunner to std::unique_ptr
dfarler@apple.com [Tue, 20 May 2014 16:23:16 +0000 (16:23 +0000)]
Move WebKitTestRunner to std::unique_ptr
https://bugs.webkit.org/show_bug.cgi?id=133081

Reviewed by Daniel Bates.

* WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
(WTR::InjectedBundle::didCreatePage):
* WebKitTestRunner/InjectedBundle/InjectedBundle.h:
* WebKitTestRunner/TestController.cpp:
(WTR::TestController::initialize):
(WTR::TestController::createWebViewWithOptions):
(WTR::TestController::resetStateToConsistentValues):
(WTR::TestController::runTest):
* WebKitTestRunner/TestController.h:
* WebKitTestRunner/TestInvocation.cpp:
* WebKitTestRunner/TestInvocation.h:
* WebKitTestRunner/WorkQueueManager.cpp:
(WTR::WorkQueueManager::processWorkQueue):
(WTR::WorkQueueManager::enqueue):
* WebKitTestRunner/WorkQueueManager.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169113 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[GTK] WebKitWebPage::send-request always pass a valid pointer for redirected response
carlosgc@webkit.org [Tue, 20 May 2014 15:20:52 +0000 (15:20 +0000)]
[GTK] WebKitWebPage::send-request always pass a valid pointer for redirected response
https://bugs.webkit.org/show_bug.cgi?id=133119

Reviewed by Sergio Villar Senin.

Source/WebKit2:
The problem is that we were checking whether the passed in
WKURLResponseRef is NULL or not, but it's always a valid object
that can contain a NULL WebCore::ResourceResponse.

* WebProcess/InjectedBundle/API/gtk/WebKitWebPage.cpp:
(willSendRequestForFrame): Pass NULL as redirect-response
parameter of WebKitWebPage::send-request signal when the
WebCore::ResourceResponse of the passed in WKURLResponseRef is NULL.

Tools:
Check that redirect response parameter of
WebKitWebPage::send-request signal is NULL when not redirecting
and a valid WebKitURIResponse object when redirecting.

* TestWebKitAPI/Tests/WebKit2Gtk/TestResources.cpp:
(testWebResourceSendRequest):
(serverCallback):
* TestWebKitAPI/Tests/WebKit2Gtk/WebExtensionTest.cpp:
(sendRequestCallback):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169112 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoMake fast/dom/remove-element-with-{id|name}-that-was-inserted-on-DOMNodeRemoved.html...
ddkilzer@apple.com [Tue, 20 May 2014 14:47:02 +0000 (14:47 +0000)]
Make fast/dom/remove-element-with-{id|name}-that-was-inserted-on-DOMNodeRemoved.html reliable
<http://webkit.org/b/133088>

Reviewed by Alexey Proskuryakov.

* fast/dom/remove-element-with-id-that-was-inserted-on-DOMNodeRemoved.html:
* fast/dom/remove-element-with-name-that-was-inserted-on-DOMNodeRemoved.html:
- Use waitUntilDone()/notifyDone() to make the tests reliable
  since they use setTimeout().

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169111 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[CSS Regions] Block incorrectly sized when containing an unsplittable box
stavila@adobe.com [Tue, 20 May 2014 14:34:56 +0000 (14:34 +0000)]
[CSS Regions] Block incorrectly sized when containing an unsplittable box
https://bugs.webkit.org/show_bug.cgi?id=132601

Reviewed by Antti Koivisto.

Source/WebCore:
When laying out elements in a region, when an inline element is encountered
the size of its parent must not be increased beyond the bottom of the current region,
unless if its the last region. This will ensure that the next sibling of the
inline element is correctly laid out at the top of the next region, instead
of leaving an empty space equal to the height of the overflow, as it did until now.

Tests: fast/regions/inline-block-inside-anonymous-overflow.html
       fast/regions/inline-block-overflow.html

* rendering/RenderBlockFlow.cpp:
(WebCore::RenderBlockFlow::adjustLinePositionForPagination):
(WebCore::RenderBlockFlow::hasNextPage):
* rendering/RenderBlockFlow.h:
* rendering/RenderBlockLineLayout.cpp:
(WebCore::RenderBlockFlow::layoutRunsAndFloatsInRange):
(WebCore::RenderBlockFlow::linkToEndLineIfNeeded):
(WebCore::RenderBlockFlow::determineStartPosition):
(WebCore::RenderBlockFlow::checkPaginationAndFloatsAtEndLine):

LayoutTests:
Added tests for the layout of elements following inline-block elements
that overflow their region, with and without anonymous blocks.

* fast/regions/inline-block-inside-anonymous-overflow-expected.html: Added.
* fast/regions/inline-block-inside-anonymous-overflow.html: Added.
* fast/regions/inline-block-overflow-expected.html: Added.
* fast/regions/inline-block-overflow.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169110 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoDouble-tap zoom does not take obscuring insets into account
antti@apple.com [Tue, 20 May 2014 13:56:27 +0000 (13:56 +0000)]
Double-tap zoom does not take obscuring insets into account
https://bugs.webkit.org/show_bug.cgi?id=133116
<rdar://problem/16765604>

Reviewed by Anders Carlsson.

* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _zoomToRect:WebCore::atScale:origin:WebCore::]):

    Compute the zoom target taking insets into account.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169109 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[GTK] Rename translation domain as WebKit2GTK-3.0
carlosgc@webkit.org [Tue, 20 May 2014 13:15:50 +0000 (13:15 +0000)]
[GTK] Rename translation domain as WebKit2GTK-3.0
https://bugs.webkit.org/show_bug.cgi?id=132953

Reviewed by Gustavo Noronha Silva.

.:
* Source/cmake/OptionsGTK.cmake:

Source/WebCore/platform/gtk/po:
* CMakeLists.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169108 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[CSS Regions] Add polygon tests for shapes on regions and shapes on the content flow
zoltan@webkit.org [Tue, 20 May 2014 12:34:01 +0000 (12:34 +0000)]
[CSS Regions] Add polygon tests for shapes on regions and shapes on the content flow
https://bugs.webkit.org/show_bug.cgi?id=133107

Reviewed by Mihnea Ovidenie.

* fast/regions/shape-outside-floats/regions-shape-outside-region-polygon-floats-expected.html: Added.
* fast/regions/shape-outside-floats/regions-shape-outside-region-polygon-floats.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169107 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION(CMAKE): [GTK] Several WebKit2GTK+ unit tests are no longer built
carlosgc@webkit.org [Tue, 20 May 2014 12:17:56 +0000 (12:17 +0000)]
REGRESSION(CMAKE): [GTK] Several WebKit2GTK+ unit tests are no longer built
https://bugs.webkit.org/show_bug.cgi?id=133117

Reviewed by Philippe Normand.

Bring back missing tests to the build.

* TestWebKitAPI/Tests/WebKit2Gtk/CMakeLists.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169106 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[CSS Regions] Crash while painting block selection gaps in regions
mihnea@adobe.com [Tue, 20 May 2014 10:57:58 +0000 (10:57 +0000)]
[CSS Regions] Crash while painting block selection gaps in regions
https://bugs.webkit.org/show_bug.cgi?id=132720

Reviewed by David Hyatt.

Source/WebCore:
The fix for WebKit bug https://bugs.webkit.org/show_bug.cgi?id=131511
allowed selection highlight to match the DOM selection when the start
and end point of the selection were in different flow threads. In order
to enable that, the selection was performed separately on view and
render flow threads, considered selection subtrees.

However, the start and end points for each selection subtree were computed
by means of Range class but it is not always possible to construct a valid
Range from two pairs of RenderObjects and offsets.

This patch keeps the substrees approach but instead of storing the endpoints
for each subtree in a Range and continuously extending the Range, it stores them
using the already available SelectionSubtreeRoot class. After the end points are
computed for each subtree and before processing the subtree selection, the end points
are adjusted in a similar fashion as the one used in FrameSelection::updateAppearance(),
to make sure we are passing the same expected information to the method implementing
the visible selection processing.

Test: fast/regions/selection-gaps-paint-crash.html

* rendering/RenderView.cpp:
(WebCore::RenderView::splitSelectionBetweenSubtrees):
* rendering/SelectionSubtreeRoot.cpp:
(WebCore::SelectionSubtreeRoot::adjustForVisibleSelection):
* rendering/SelectionSubtreeRoot.h:
(WebCore::SelectionSubtreeRoot::selectionClear):

LayoutTests:
* TestExpectations: Unskip a test that was crashing
* fast/regions/selection-gaps-paint-crash-expected.txt: Added.
* fast/regions/selection-gaps-paint-crash.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169105 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[GTK] Unreviewed GTK gardening
commit-queue@webkit.org [Tue, 20 May 2014 09:52:02 +0000 (09:52 +0000)]
[GTK] Unreviewed GTK gardening

Rebaseline affected tests by changes on integral snapping of inlines
introduced at r169048.

Patch by Lorenzo Tilve <ltilve@igalia.com> on 2014-05-20

* platform/gtk/css1/box_properties/float_on_text_elements-expected.txt:
* platform/gtk/css1/text_properties/text_indent-expected.txt:
* platform/gtk/css2.1/t0905-c414-flt-wrap-00-e-expected.txt:
* platform/gtk/css2.1/t0905-c5525-fltwidth-00-c-g-expected.txt:
* platform/gtk/css2.1/t100801-c544-valgn-00-a-ag-expected.txt:
* platform/gtk/css2.1/t100801-c544-valgn-02-d-agi-expected.txt:
* platform/gtk/css2.1/t100801-c544-valgn-03-d-agi-expected.txt:
* platform/gtk/css2.1/t100801-c544-valgn-04-d-agi-expected.txt:
* platform/gtk/css2.1/t1601-c547-indent-00-b-a-expected.txt:
* platform/gtk/fast/block/float/026-expected.txt:
* platform/gtk/fast/block/float/028-expected.txt:
* platform/gtk/fast/css/empty-pseudo-class-expected.txt:
* platform/gtk/fast/css/first-child-pseudo-class-expected.txt:
* platform/gtk/fast/css/first-of-type-pseudo-class-expected.txt:
* platform/gtk/fast/css/last-child-pseudo-class-expected.txt:
* platform/gtk/fast/css/last-of-type-pseudo-class-expected.txt:
* platform/gtk/fast/css/only-child-pseudo-class-expected.txt:
* platform/gtk/fast/css/only-of-type-pseudo-class-expected.txt:
* platform/gtk/fast/dom/navigator-detached-no-crash-expected.txt:
* platform/gtk/fast/line-grid/line-align-left-edges-expected.txt:
* platform/gtk/fast/multicol/column-count-with-rules-expected.txt:
* platform/gtk/fast/multicol/float-multicol-expected.txt:
* platform/gtk/fast/multicol/layers-in-multicol-expected.txt:
* platform/gtk/fast/multicol/vertical-lr/float-multicol-expected.txt:
* platform/gtk/fast/multicol/vertical-rl/float-multicol-expected.txt:
* platform/gtk/fast/regions/multiple-directionality-changes-in-variable-width-regions-expected.txt:
* platform/gtk/fast/regions/repaint/line-flow-with-floats-in-regions-expected.txt:
* platform/gtk/fast/repaint/line-flow-with-floats-1-expected.txt:
* platform/gtk/fast/repaint/line-flow-with-floats-10-expected.txt:
* platform/gtk/fast/repaint/line-flow-with-floats-2-expected.txt:
* platform/gtk/fast/repaint/line-flow-with-floats-3-expected.txt:
* platform/gtk/fast/repaint/line-flow-with-floats-4-expected.txt:
* platform/gtk/fast/repaint/line-flow-with-floats-5-expected.txt:
* platform/gtk/fast/repaint/line-flow-with-floats-6-expected.txt:
* platform/gtk/fast/repaint/line-flow-with-floats-7-expected.txt:
* platform/gtk/fast/repaint/line-flow-with-floats-8-expected.txt:
* platform/gtk/fast/repaint/line-flow-with-floats-9-expected.txt:
* platform/gtk/fast/text/whitespace/024-expected.txt:
* platform/gtk/svg/wicd/rightsizing-grid-expected.txt:
* platform/gtk/svg/zoom/page/zoom-replaced-intrinsic-ratio-001-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169104 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoTest gardening: skip some failing tests on not-X86.
fpizlo@apple.com [Tue, 20 May 2014 04:40:13 +0000 (04:40 +0000)]
Test gardening: skip some failing tests on not-X86.

Source/JavaScriptCore:
* tests/mozilla/mozilla-tests.yaml:

LayoutTests:
* js/script-tests/function-apply-many-args.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169102 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoScrolling pages with large TiledBacking content layers creates way too many tiles
simon.fraser@apple.com [Tue, 20 May 2014 02:31:19 +0000 (02:31 +0000)]
Scrolling pages with large TiledBacking content layers creates way too many tiles
https://bugs.webkit.org/show_bug.cgi?id=133101

Reviewed by Tim Horton.

Fix the GraphicsLayerCA::adjustTiledLayerVisibleRect() logic to avoid making giant rects if the old
and new visible rects don't overlap.

* platform/graphics/TiledBacking.h:
* platform/graphics/ca/GraphicsLayerCA.cpp:
(WebCore::GraphicsLayerCA::adjustTiledLayerVisibleRect):
* platform/graphics/ca/mac/TileController.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169097 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoCorrectly maintain the "isInWindow" state for all TiledBackings
simon.fraser@apple.com [Tue, 20 May 2014 02:31:17 +0000 (02:31 +0000)]
Correctly maintain the "isInWindow" state for all TiledBackings
https://bugs.webkit.org/show_bug.cgi?id=133095

Reviewed by Tim Horton.

We never updated the "isInWindow" state for TiledBacking that wasn't the RenderView's
backing. Fix it by having RenderLayerCompositor::setIsInWindow() recurse through all
layers (if necessary), updating their state. Also set the state on newly created
TiledBackings.

* rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::layerTiledBackingUsageChanged):
(WebCore::RenderLayerCompositor::setIsInWindowForLayerIncludingDescendants):
(WebCore::RenderLayerCompositor::setIsInWindow):
(WebCore::RenderLayerCompositor::pageTiledBacking): Deleted.
* rendering/RenderLayerCompositor.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169096 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoWebKit2/iOS: parented view should be considered hidden if app is backgrounded
barraclough@apple.com [Tue, 20 May 2014 02:22:35 +0000 (02:22 +0000)]
WebKit2/iOS: parented view should be considered hidden if app is backgrounded
https://bugs.webkit.org/show_bug.cgi?id=133099
radar://16942183

Reviewed by Simon Fraser.

* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::isViewWindowActive):
    - implemented in terms of isViewVisible.
    - FIXME: https://bugs.webkit.org/show_bug.cgi?id=133098
(WebKit::PageClientImpl::isViewFocused):
    - implemented in terms of isViewWindowActive.
    - FIXME: https://bugs.webkit.org/show_bug.cgi?id=133098
(WebKit::PageClientImpl::isViewVisible):
    - take UIApplication applicationState into account in determining visibility.
(WebKit::PageClientImpl::isViewVisibleOrOccluded):
    - implemented in terms of isViewVisible.
(WebKit::PageClientImpl::isVisuallyIdle):
    - implemented in terms of isViewVisible.
* UIProcess/ios/WKContentView.mm:
(-[WKContentView initWithFrame:context:WebKit::configuration:WebKit::webView:]):
    - register for UIApplicationDidEnterBackgroundNotification.
(-[WKContentView _applicationDidEnterBackground:]):
(-[WKContentView _applicationWillEnterForeground:]):
    - when the application state changes all flags except IsInWindow may be affected.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169095 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agooperationOptimize() should defer the GC for a while.
mark.lam@apple.com [Tue, 20 May 2014 02:03:47 +0000 (02:03 +0000)]
operationOptimize() should defer the GC for a while.
<https://webkit.org/b/133103>

Reviewed by Filip Pizlo.

Currently, operationOptimize() only defers the GC until its end.  As a result,
a GC may be triggered just before we return from operationOptimize(), and it may
jettison the optimize codeBlock that we're planning to OSR enter into when we
return from this function.  This is because the OSR entry on-ramp code hasn't
been executed yet, and hence, there is not yet a reference to this new codeBlock
from the stack, and there won't be until we've had a chance to return out of
operationOptimize() to run the OSR entry on-ramp code.

This issue is now fixed by using DeferGCForAWhile instead of DeferGC.  This
ensures that the GC will be deferred until after the OSR entry on-ramp can be
executed.

* jit/JITOperations.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169094 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoAX: VoiceOver sees the WebArea out of order when topContentInset is used
cfleizach@apple.com [Tue, 20 May 2014 01:56:41 +0000 (01:56 +0000)]
AX: VoiceOver sees the WebArea out of order when topContentInset is used
https://bugs.webkit.org/show_bug.cgi?id=133091

Unreviewed.
Make the layout test more resilient to platform differences in frame positions.

* platform/mac-wk2/accessibility/content-inset-scrollview-frame-expected.txt:
* platform/mac-wk2/accessibility/content-inset-scrollview-frame.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169093 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoTake care of some ARM64 test failures
fpizlo@apple.com [Tue, 20 May 2014 01:53:56 +0000 (01:53 +0000)]
Take care of some ARM64 test failures
https://bugs.webkit.org/show_bug.cgi?id=133090

Reviewed by Geoffrey Garen.

PerformanceTests/SunSpider:

* profiler-test.yaml: Skip these on not-x86.

Source/JavaScriptCore:

Constant blinding on ARM64 cannot use the scratch register.

* assembler/MacroAssembler.h:
(JSC::MacroAssembler::convertInt32ToDouble):
(JSC::MacroAssembler::branchPtr):
(JSC::MacroAssembler::storePtr):
(JSC::MacroAssembler::store64):
* assembler/MacroAssemblerARM64.h:
(JSC::MacroAssemblerARM64::scratchRegisterForBlinding):

Tools:

* Scripts/run-jsc-stress-tests: Make it easier to conditionally skip things.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169092 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoWK2 iOS: Pressing the form navigation button causes the keyboard to be dismissed.
enrica@apple.com [Tue, 20 May 2014 01:09:37 +0000 (01:09 +0000)]
WK2 iOS: Pressing the form navigation button causes the keyboard to be dismissed.
https://bugs.webkit.org/show_bug.cgi?id=133100
<rdar://problem/16808673>

Reviewed by Benjamin Poulain.

The call to focusNextAssistedNode is the result of a user action, since
it is called when the user pressed the form navigation buttons.
Therefore we must set m_userIsInteracting to true before changing the focus
to make sure the focus change is honored.

* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::focusNextAssistedNode):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169091 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[iOS] Use status display for live streams
dino@apple.com [Tue, 20 May 2014 00:55:42 +0000 (00:55 +0000)]
[iOS] Use status display for live streams
https://bugs.webkit.org/show_bug.cgi?id=133097

Reviewed by Sam Weinig.

Follow-on from https://bugs.webkit.org/show_bug.cgi?id=131390
Allow the iOS controls to use the status display field.

* Modules/mediacontrols/mediaControlsiOS.js:
(ControllerIOS.prototype.configureInlineControls): Live streams should
not have a timeline or rewind button.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169090 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoText decorations do not contribute to visual overflow
mmaxfield@apple.com [Tue, 20 May 2014 00:45:52 +0000 (00:45 +0000)]
Text decorations do not contribute to visual overflow
https://bugs.webkit.org/show_bug.cgi?id=132773

Patch by Myles C. Maxfield <litherum@gmail.com> on 2014-05-19
Reviewed by Antti Koivisto.

Source/WebCore:
This patch creates a function, visualOverflowForDecorations, which computes
how much visual overflow to add around a text box due to text decorations. Most of the time,
text decorations are fully contained within the text box, so the result is usually 0.

This function exists within style/InlineTextBoxStyle.cpp, which is an added file. This is
so that it can be called from setLogicalWidthForTextRun() inside RenderBlockLineLayout.cpp
and from RenderStyle::changeAffectsVisualOverflow(). The former case passes in the full
InlineTextBox and the latter case just passes in a RenderStyle (because the InlineTextBox
is unavailable.)

This patch also modifies RenderTableSection::spannedColumns() to fix an off-by-one error
that was causing table borders to not be drawn when they existed right on the edge of
a repaint rect.

Tests: fast/css3-text/css3-text-decoration/repaint/underline-outside-of-layout-rect.html
Tests: fast/repaint/border-collapse-table-off-by-one-expected.html

* WebCore.vcxproj/WebCore.vcxproj: Adding reference to new InlineTextBoxStyle.cpp file
* WebCore.vcxproj/WebCore.vcxproj.filters: Adding reference to new InlineTextBoxStyle files
* WebCore.xcodeproj/project.pbxproj: Adding reference to new InlineTextBoxStyle files
* rendering/InlineTextBox.cpp:
(WebCore::computeUnderlineOffset): Moved to InlineTextBox.cpp
(WebCore::getWavyStrokeParameters): Moved to InlineTextBox.cpp
(WebCore::InlineTextBox::paintDecoration): Update to use newly refactored functions
* rendering/RenderBlockLineLayout.cpp:
(WebCore::setLogicalWidthForTextRun): Call visualOverflowForDecorations()
* rendering/RenderTableSection.cpp:
* rendering/RenderTableSelection.cpp: Fix off-by-one error when the boundary of a repaint
rect lies exactly on top of a table column position
* rendering/style/RenderStyle.cpp:
(WebCore::RenderStyle::changeAffectsVisualOverflow): Inspects shadows and text decorations
(WebCore::RenderStyle::changeRequiresLayout): Calls changeAffectsVisualOverflow()
(WebCore::RenderStyle::changeRequiresRepaintIfTextOrBorderOrOutline): Moved code from here
to changeAffectsVisualOverflow().
* rendering/style/RenderStyle.h: Function signature
* style/InlineTextBoxStyle.cpp: Added.
(WebCore::computeUnderlineOffset): Moved from InlineTextBox.cpp
(WebCore::getWavyStrokeParameters): Moved from InlineTextBox.cpp
(WebCore::extendIntToFloat): Convenience function for dealing with the fact that
underline bounding boxes use floats and GlyphOverflow uses ints
(WebCore::visualOverflowForDecorations): Given
vertical overflow bounds, possibly extend those to include location of
decorations.
* style/InlineTextBoxStyle.h: Added. Function signatures.
(WebCore::textDecorationStrokeThickness): Refactored from InlineTextBox.cpp
(WebCore::wavyOffsetFromDecoration): Refactored from InlineTextBox.cpp
* platform/graphics/Font.h:
(WebCore::GlyphOverflow::isEmpty): Convenience function
(WebCore::GlyphOverflow::extendTo): Convenience function

LayoutTests:
This first test makes sure that repaint rects are extended to include text decorations that may
lie outside of the text layout rects. It compares text with an underline to text that has
had underline applied to it in a timer.

The second test triggers an off-by-one error in collapsed table borders where a border was not
being drawn if it lay right on top of a repaint rect.

* fast/css3-text/css3-text-decoration/repaint/resources/Litherum.svg: Added. This font has a
descent of 0 (so it will not intersect underlines).
* fast/css3-text/css3-text-decoration/repaint/underline-outside-of-layout-rect-expected.html: Added.
Apply the underline without any timeout.
* fast/css3-text/css3-text-decoration/repaint/underline-outside-of-layout-rect.html: Added.
Apply the underline with a timeout.
* fast/css3-text/css3-text-decoration/repaint/underline-outside-of-layout-rect-removed-expected.html: Added.
Draw text without decorations.
* fast/css3-text/css3-text-decoration/repaint/underline-outside-of-layout-rect-removed.html: Added.
Draw text with decorations which contribute to overflow, then remove those decorations on a timer.
* fast/css3-text/css3-text-decoration/repaint/underline-outside-of-layout-rect-altered-expected.html: Added.
Draw text as if the final state of the decorations had always existed.
* fast/css3-text/css3-text-decoration/repaint/underline-outside-of-layout-rect-altered.html: Added.
Make sure that modifying decorations that contribute to overflow redraw properly.
* fast/repaint/border-collapse-table-off-by-one-expected.html: Added.
* fast/repaint/border-collapse-table-off-by-one.html: Added. Trigger the off-by-one error in
RenderTableSection.cpp

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169089 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[iOS] REGRESSION (r168910): webkitdirs.pm: Be more forgiving when parsing --sdk,...
ddkilzer@apple.com [Tue, 20 May 2014 00:43:59 +0000 (00:43 +0000)]
[iOS] REGRESSION (r168910): webkitdirs.pm: Be more forgiving when parsing --sdk, --device and --sim
<http://webkit.org/b/133089>

Reviewed by Darin Adler.

* Scripts/webkitdirs.pm:
(determineXcodeSDK): If both '--sdk SDKROOT' and '--sim' are
specified on the command line, we would remove the '--sdk'
arguments, but leave '--sim', potentially passing it along to
xcodebuild causing a build failure.  This happens when
'old-run-webkit-tests --release --sim' invokes
build-dumprendertree when DumpRenderTree hasn't been built yet.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169088 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoTelephoneNumberOverlayController should use a modern loop for text quad bounding
bdakin@apple.com [Tue, 20 May 2014 00:24:55 +0000 (00:24 +0000)]
TelephoneNumberOverlayController should use a modern loop for text quad bounding
boxes
https://bugs.webkit.org/show_bug.cgi?id=133096

Reviewed by Anders Carlsson.

* WebProcess/WebPage/mac/TelephoneNumberOverlayControllerMac.mm:
(WebKit::textQuadsToBoundingRectForRange):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169087 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoPlatformCALayerRemoteCustom properties aren't initialized to match the PlatformLayer
commit-queue@webkit.org [Tue, 20 May 2014 00:18:38 +0000 (00:18 +0000)]
PlatformCALayerRemoteCustom properties aren't initialized to match the PlatformLayer
https://bugs.webkit.org/show_bug.cgi?id=133025

Patch by Jeremy Jones <jeremyj@apple.com> on 2014-05-19
Reviewed by Simon Fraser.

* WebProcess/WebPage/mac/PlatformCALayerRemoteCustom.mm:
(WebKit::PlatformCALayerRemoteCustom::PlatformCALayerRemoteCustom):
Initialize PlatformCALayerRemote properties from the PlatformLayer.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169086 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed build fix after r169082
achristensen@apple.com [Tue, 20 May 2014 00:01:12 +0000 (00:01 +0000)]
Unreviewed build fix after r169082

* platform/network/ResourceHandle.h:
Added NSDictionary declaration.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169085 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION: Standalone images need to take topContentInset into account
bdakin@apple.com [Mon, 19 May 2014 23:56:55 +0000 (23:56 +0000)]
REGRESSION: Standalone images need to take topContentInset into account
https://bugs.webkit.org/show_bug.cgi?id=133092
-and corresponding-
<rdar://problem/16945791>

Reviewed by Darin Adler.

The FrameView’s size is accurate in terms of the NSWindow size, but that is not
really what the ImageDocument code is looking for here. The ImageDocument wants to
know the size of the unobscured viewport area. We should use visibleSize() for
that.
* html/ImageDocument.cpp:
(WebCore::ImageDocument::scale):
(WebCore::ImageDocument::imageFitsInWindow):
(WebCore::ImageDocument::imageClicked):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169084 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoCollect web timing data on iOS.
achristensen@apple.com [Mon, 19 May 2014 23:45:54 +0000 (23:45 +0000)]
Collect web timing data on iOS.
https://bugs.webkit.org/show_bug.cgi?id=133075

Reviewed by Sam Weinig.

* platform/network/ResourceHandle.h:
Added setCollectsTimingData wrapper and getConnectionTimingData declarations.
getConnectionTimingData with a dictionary parameter is a private helper function.
* platform/network/cf/ResourceHandleCFNet.cpp:
(WebCore::ResourceHandle::start):
Call setCollectsTimingData wrapper when creating a connection to ensure that timing data is collected.
* platform/network/cf/ResourceHandleCFURLConnectionDelegate.cpp:
(WebCore::ResourceHandleCFURLConnectionDelegate::didReceiveResponseCallback):
* platform/network/cf/ResourceHandleCFURLConnectionDelegate.h:
* platform/network/cf/ResourceHandleCFURLConnectionDelegateWithOperationQueue.cpp:
(WebCore::ResourceHandleCFURLConnectionDelegateWithOperationQueue::didReceiveResponse):
* platform/network/cf/ResourceHandleCFURLConnectionDelegateWithOperationQueue.h:
* platform/network/cf/SynchronousResourceHandleCFURLConnectionDelegate.cpp:
(WebCore::SynchronousResourceHandleCFURLConnectionDelegate::didReceiveResponse):
* platform/network/cf/SynchronousResourceHandleCFURLConnectionDelegate.h:
Pass the CFURLConnectionRef to didReceiveResponse to get timing data from.
* platform/network/mac/ResourceHandleMac.mm:
(WebCore::ResourceHandle::getConnectionTimingData):
Moved from getTimingData to compile regardless of using CFNetwork and added wrapper with CFDictionaryRef.
(WebCore::ResourceHandle::setCollectsTimingData):
Added wrapper to call objc function from c++.
(WebCore::ResourceHandle::getTimingData):
Renamed to getConnectionTimingData.
* platform/network/mac/WebCoreResourceHandleAsDelegate.mm:
(-[WebCoreResourceHandleAsDelegate connection:didReceiveResponse:]):
* platform/network/mac/WebCoreResourceHandleAsOperationQueueDelegate.mm:
(-[WebCoreResourceHandleAsOperationQueueDelegate connection:didReceiveResponse:]):
Updated function name from getTimingData to getConnectionTimingData.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169082 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoAX: VoiceOver sees the WebArea out of order when topContentInset is used
cfleizach@apple.com [Mon, 19 May 2014 23:37:56 +0000 (23:37 +0000)]
AX: VoiceOver sees the WebArea out of order when topContentInset is used
https://bugs.webkit.org/show_bug.cgi?id=133091

Reviewed by Simon Fraser.

Source/WebCore:
Modify the scroll view's frame so that it recognizes the content inset.
This is needed so that when VoiceOver sorts elements to create an ordering,
the web area is not ordered at the top of the screen.

Test: platform/mac-wk2/accessibility/content-inset-scrollview-frame.html

* accessibility/AccessibilityScrollView.cpp:
(WebCore::AccessibilityScrollView::elementRect):

LayoutTests:
* platform/mac-wk2/accessibility/content-inset-scrollview-frame-expected.txt: Added.
* platform/mac-wk2/accessibility/content-inset-scrollview-frame.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169081 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoDashboard: OpenSource revision key changed from WebKitOpenSource to WebKit
dfarler@apple.com [Mon, 19 May 2014 23:08:58 +0000 (23:08 +0000)]
Dashboard: OpenSource revision key changed from WebKitOpenSource to WebKit
https://bugs.webkit.org/show_bug.cgi?id=133086

Reviewed by Daniel Bates.

* BuildSlaveSupport/build.webkit.org-config/public_html/dashboard/Scripts/BuildbotIteration.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@169080 268f45cc-cd09-0410-ab3c-d52691b4dbfc