Timeline
May 28, 2015:
- 11:38 PM Changeset in webkit [184982] by
-
- 5 edits2 deletes in trunk/Source/WebKit2
Change method of signaling "should open external urls" to WebKit.
https://bugs.webkit.org/show_bug.cgi?id=145454
Reviewed by Dan Bernstein.
Making this a property on the request was never a good idea.
Adding a new extensible load SPI seems much better.
- Shared/API/Cocoa/_WKNSURLRequestExtras.h: Removed.
- Shared/API/Cocoa/_WKNSURLRequestExtras.mm: Removed.
- UIProcess/API/Cocoa/WKBrowsingContextController.mm:
(-[WKBrowsingContextController loadRequest:userData:]): Since this API is deprecated,
it always gets the default restrictive behavior.
- UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView loadRequest:]):
(-[WKWebView loadRequest:withOptions:]): Take an options dictionary that, for now,
only supports one option.
- UIProcess/API/Cocoa/WKWebViewPrivate.h:
- WebKit2.xcodeproj/project.pbxproj:
- 11:17 PM Changeset in webkit [184981] by
-
- 2 edits in trunk/LayoutTests
[EFL] Unreviewed gardening on 29th May
Unskip passing tests for viewport-percentage, text, sub-pixel
and update new failing accessibility, fast/overflow and so on.
- platform/efl/TestExpectations:
- 9:56 PM Changeset in webkit [184980] by
-
- 2 edits in trunk/Tools
Update results of WebKit1.StringTruncator after r184965. I missed one letter.
- TestWebKitAPI/Tests/mac/StringTruncator.mm: (TestWebKitAPI::TEST):
- 9:53 PM Changeset in webkit [184979] by
-
- 9 edits in trunk/Source/WebCore
Replaced 0 with nullptr in WebCore/accessibility.
https://bugs.webkit.org/show_bug.cgi?id=145309
Patch by Hunseop Jeong <Hunseop Jeong> on 2015-05-28
Reviewed by Chris Fleizach.
No new tests, no behavior change.
- accessibility/AccessibilityRenderObject.cpp:
(WebCore::AccessibilityRenderObject::mathPrescripts):
(WebCore::AccessibilityRenderObject::mathPostscripts):
- accessibility/AccessibilityScrollView.cpp:
(WebCore::AccessibilityScrollView::scrollBar):
- accessibility/atk/AXObjectCacheAtk.cpp:
(WebCore::notifyChildrenSelectionChange):
- accessibility/atk/WebKitAccessibleHyperlink.cpp:
(webkitAccessibleHyperlinkInit):
- accessibility/atk/WebKitAccessibleInterfaceDocument.cpp:
(webkitAccessibleDocumentGetAttributes):
- accessibility/atk/WebKitAccessibleInterfaceText.cpp:
(getAttributeSetForAccessibilityObject):
(attributeSetDifference):
- accessibility/atk/WebKitAccessibleUtil.h:
- accessibility/atk/WebKitAccessibleWrapperAtk.cpp:
(webkitAccessibleRefChild):
(webkitAccessibleGetAttributes):
(webkitAccessibleGetObjectLocale):
(getInterfaceMaskFromObject):
(objectFocusedAndCaretOffsetUnignored):
(cacheAndReturnAtkProperty):
- 9:50 PM Changeset in webkit [184978] by
-
- 4 edits in trunk/Source/WebInspectorUI
Web Inspector: Update toolbar styles
https://bugs.webkit.org/show_bug.cgi?id=145419
Also, use hairline borders (1 phisical pixel) on retina (DPI >= 2).
Reviewed by Timothy Hatcher.
- UserInterface/Views/ButtonToolbarItem.css:
(.toolbar .item.button): Deleted.
(.toolbar.icon-and-label-vertical .item.button): Deleted.
(.toolbar.icon-and-label-horizontal .item.button): Deleted.
(.toolbar .item.button > .glyph): Deleted.
(.toolbar.small-size .item.button > .glyph): Deleted.
- UserInterface/Views/DashboardContainerView.css:
(.toolbar .dashboard-container): Deleted.
(.toolbar.collapsed .dashboard-container): Deleted.
(.toolbar.small-size:matches(.icon-only, .icon-and-label-vertical, .icon-and-label-horizontal) .dashboard-container): Deleted.
(.toolbar .dashboard.slide-out-down): Deleted.
(.toolbar .dashboard.slide-in-up): Deleted.
- UserInterface/Views/Toolbar.css:
(.toolbar .search-bar > input[type="search"]):
(.toolbar .search-bar > input[type="search"]::-webkit-textfield-decoration-container):
(body:not(.unknown-mac) .toolbar .dashboard-container):
(body:not(.unknown-mac) .toolbar .search-bar > input[type="search"]):
(body:not(.unknown-mac) .toolbar .search-bar > input[type="search"]:focus):
(body:not(.unknown-mac) .toolbar .item.button:active):
(body.window-inactive:not(.unknown-mac) .toolbar .dashboard-container):
(body.unknown-mac .toolbar .dashboard-container):
(body.unknown-mac .toolbar .search-bar > input[type="search"]):
(body.unknown-mac .toolbar .search-bar > input[type="search"]:focus):
(@media (-webkit-min-device-pixel-ratio: 2)):
(body.unknown-mac .toolbar .item.button:active):
(body.unknown-mac.window-inactive .toolbar .dashboard-container):
- 9:47 PM Changeset in webkit [184977] by
-
- 8 edits in trunk/Source/WebInspectorUI
Web Inspector: Jump from a computed style to the rule it came from
https://bugs.webkit.org/show_bug.cgi?id=120640
Patch by Devin Rousso <Devin Rousso> on 2015-05-28
Reviewed by Timothy Hatcher.
- UserInterface/Views/CSSStyleDeclarationTextEditor.js:
(WebInspector.CSSStyleDeclarationTextEditor.prototype._createTextMarkerForPropertyIfNeeded):
If the delegate of CSSStyleDeclarationTextEditor has cssStyleDeclarationTextEditorShouldAddPropertyGoToArrows set to true, add a goToArrow to all CSS property entries.
(WebInspector.CSSStyleDeclarationTextEditor.prototype.highlightProperty.propertiesMatch):
(WebInspector.CSSStyleDeclarationTextEditor.prototype.highlightProperty.hasMatchingLonghandProperty):
(WebInspector.CSSStyleDeclarationTextEditor.prototype.highlightProperty):
Determines if a given CSS property is in the CSS style section and if so, select that CSS property, focus on the section containing that CSS property, and return true.
- UserInterface/Views/CSSStyleDeclarationSection.js:
(WebInspector.CSSStyleDeclarationSection.prototype.highlightProperty):
Searches through the properties of the section for a given property and scrolls to it if found.
- UserInterface/Views/CSSStyleDetailsSidebarPanel.js:
(WebInspector.CSSStyleDetailsSidebarPanel.prototype.computedStyleDetailsPanelShowProperty):
Switches to the rulesStyleDetailsPanel and scrolls to and hightlights a given property in that panel.
(WebInspector.CSSStyleDetailsSidebarPanel.prototype._navigationItemSelected):
(WebInspector.CSSStyleDetailsSidebarPanel.prototype._switchPanels):
Moved this function out of _navigationItemSelected for better reusablity.
- UserInterface/Views/ComputedStyleDetailsPanel.js:
(WebInspector.ComputedStyleDetailsPanel.prototype.cssStyleDeclarationTextEditorShowProperty):
Function that calls the delegate (which should be CSSStyleDetailsSidebarPanel) function computedStyleDetailsPanelShowProperty.
- UserInterface/Views/RulesStyleDetailsPanel.js:
(WebInspector.RulesStyleDetailsPanel):
(WebInspector.RulesStyleDetailsPanel.prototype.refresh):
(WebInspector.RulesStyleDetailsPanel.prototype.scrollToSectionAndHighlightProperty):
Searches through all the sections of the RulesStyleDetailsPanel for a given CSS property.
(WebInspector.RulesStyleDetailsPanel.prototype.shown):
(WebInspector.RulesStyleDetailsPanel.prototype.hidden):
(WebInspector.RulesStyleDetailsPanel.prototype.nodeStylesRefreshed):
Added a flag to this function that will highlight and scroll to a given property (_propertyToSelectAndHighlight) if set on load.
- UserInterface/Views/StyleDetailsPanel.js:
(WebInspector.StyleDetailsPanel.prototype.markAsNeedsRefresh):
(WebInspector.StyleDetailsPanel.prototype.nodeStylesRefreshed):
Made into protected function to allow it to be overridden.
(WebInspector.StyleDetailsPanel.prototype._nodeStylesRefreshed): Deleted.
- 9:30 PM Changeset in webkit [184976] by
-
- 1 edit in trunk/Tools/ChangeLog
Crash under ICU with ASAN during editing/selection/move-by-word-visually-crash-test-5.html
https://bugs.webkit.org/show_bug.cgi?id=145429
Unreviewed test gardening.
- TestWebKitAPI/Tests/mac/StringTruncator.mm:
(TestWebKitAPI::TEST): Update test due to the progression.
- 9:03 PM Changeset in webkit [184975] by
-
- 2 edits in trunk/Tools
Update results of WebKit1.StringTruncator after r184965.
- TestWebKitAPI/Tests/mac/StringTruncator.mm: (TestWebKitAPI::TEST):
- TestWebKitAPI/Tests/mac/StringTruncator.mm:
(TestWebKitAPI::TEST):
- 7:06 PM Changeset in webkit [184974] by
-
- 4 edits in trunk/Source/WebInspectorUI
Web Inspector: Should have a keyboard shortcut to switch between inspector tabs
https://bugs.webkit.org/show_bug.cgi?id=144207
Patch by Joseph Pecoraro <Joseph Pecoraro> on 2015-05-28
Reviewed by Timothy Hatcher.
These keyboard shortcuts work well in an undocked inspector window.
Be careful that some keyboard shortcuts, when used inside a text
editor / text field are just text actions. However, some keyboard
shortcuts will work no matter where you trigger them.
- UserInterface/Models/KeyboardShortcut.js:
- UserInterface/Views/TabBar.js:
(WebInspector.TabBar.prototype.selectPreviousTab):
(WebInspector.TabBar.prototype.selectNextTab):
- UserInterface/Views/TabBrowser.js:
(WebInspector.TabBrowser):
(WebInspector.TabBrowser.prototype._showPreviousTab):
(WebInspector.TabBrowser.prototype._showNextTab):
- 6:56 PM Changeset in webkit [184973] by
-
- 2 edits in trunk/Tools
Unreviewed, updated my list of email addresses.
- Scripts/webkitpy/common/config/contributors.json:
- 6:08 PM Changeset in webkit [184972] by
-
- 2 edits in trunk/Source/WTF
Purge PassRefPtr in StringConcatenate.h
https://bugs.webkit.org/show_bug.cgi?id=145376
Reviewed by Darin Adler.
Replace PassRefPtr with RefPtr. Additionally use nullptr instead of 0.
- wtf/text/StringConcatenate.h:
(WTF::tryMakeString):
(WTF::makeString):
- 5:20 PM Changeset in webkit [184971] by
-
- 2 edits in trunk/Source/WebInspectorUI
Web Inspector: NewTabContentView should update button disabled state as other tabs are added/removed
https://bugs.webkit.org/show_bug.cgi?id=145448
Patch by Joseph Pecoraro <Joseph Pecoraro> on 2015-05-28
Reviewed by Timothy Hatcher.
- UserInterface/Views/NewTabContentView.js:
(WebInspector.NewTabContentView):
Always add a click listener and add the type as a private property on the element.
(WebInspector.NewTabContentView.prototype.shown):
(WebInspector.NewTabContentView.prototype.hidden):
Add / remove event listeners for tab changes. Also update the view when shown.
(WebInspector.NewTabContentView.prototype._createNewTab):
A disabled button an now be clicked. Do nothing if a new tab of this type is not allowed.
(WebInspector.NewTabContentView.prototype._updateTabItems):
Refresh the disabled state for each of the buttons.
- 5:15 PM Changeset in webkit [184970] by
-
- 5 edits2 adds in trunk
Subpixel rendering: Pixel crack in text selection of simple text in <textarea>.
https://bugs.webkit.org/show_bug.cgi?id=145393
rdar://problem/19918941
Reviewed by Darin Adler.
Float to LayoutUnit conversion is lossy. To ensure that selection
painting always lines up (snaps) properly, the calculated width needs to
be adjusted by ceiling the float to the next LayoutUnit value.
Source/WebCore:
Test: fast/text/hidpi-text-selection-gap-between-words.html
- platform/graphics/FontCascade.cpp:
(WebCore::FontCascade::adjustSelectionRectForSimpleText):
- platform/graphics/cocoa/FontCascadeCocoa.mm:
(WebCore::FontCascade::adjustSelectionRectForComplexText):
LayoutTests:
- fast/text/hidpi-text-selection-gap-between-words-expected.html: Added.
- fast/text/hidpi-text-selection-gap-between-words.html: Added.
- platform/mac/platform/mac/editing/input/caret-primary-bidi-expected.txt:
- 4:51 PM Changeset in webkit [184969] by
-
- 2 edits in trunk/Source/JavaScriptCore
mozilla/js1_5/Array/regress-154338.js test causes ARM 32 bit iOS devices to run out of memory
https://bugs.webkit.org/show_bug.cgi?id=145444
Reviewed by Geoffrey Garen.
Disabled mozilla/js1_5/Array/regress-154338.js when run on iOS ARM 32 bit devices and
the --memory-limited option is passed to run-jsc-stress-tests.
- tests/mozilla/mozilla-tests.yaml:
- 4:48 PM Changeset in webkit [184968] by
-
- 3 edits2 adds in trunk
GraphicsLayerCA should start out with m_intersectsCoverageRect=false
<https://webkit.org/b/145443>
<rdar://problem/19685798>
Reviewed by Simon Fraser.
Source/WebCore:
Since the coverage rect of a newly-constructed GraphicsLayerCA is always empty,
there's no way that the layer's own bounding box can intersect with it.
Hence the m_intersectsCoverageRect flag should be false after construction.
This fixes an issue where compositing layers in subframes would be marked as
non-volatile, despite being far outside the viewport and coverage rect.
Test: compositing/visible-rect/iframe-with-layers-outside-viewport.html
- platform/graphics/ca/GraphicsLayerCA.cpp:
(WebCore::GraphicsLayerCA::GraphicsLayerCA):
LayoutTests:
Add a test (based on iframe-and-layers.html) that makes an iframe with compositing
layers inside it and places it far outside the visible viewport.
The layer dump should show that the "intersects coverage rect" flag is 0 for the
iframe and its sublayers.
- compositing/visible-rect/iframe-with-layers-outside-viewport-expected.txt: Added.
- compositing/visible-rect/iframe-with-layers-outside-viewport.html: Added.
- 4:20 PM Changeset in webkit [184967] by
-
- 2 edits in trunk/Source/WebCore
Unreviewed, rolling out r184796 and r184832.
https://bugs.webkit.org/show_bug.cgi?id=145449
2% performance regression (Requested by litherum on #webkit).
Reverted changesets:
"[Cocoa] Use CTFontDrawGlyphs() instead of
CGContextShowGlyphsWithAdvances()/CGContextShowGlyphsAtPositions()"
https://bugs.webkit.org/show_bug.cgi?id=145234
http://trac.webkit.org/changeset/184796
"Cleanup after r184796"
https://bugs.webkit.org/show_bug.cgi?id=145333
http://trac.webkit.org/changeset/184832
- 4:03 PM Changeset in webkit [184966] by
-
- 8 edits2 moves in trunk/Source/WebCore
Rename HTMLMediaSession to MediaElementSession to reserve "media session" for the Media Session API.
https://bugs.webkit.org/show_bug.cgi?id=145440
Patch by Matt Rajca <mrajca@apple.com> on 2015-05-28
Reviewed by Eric Carlson.
- CMakeLists.txt:
- WebCore.vcxproj/WebCore.vcxproj:
- WebCore.vcxproj/WebCore.vcxproj.filters:
- WebCore.xcodeproj/project.pbxproj:
- html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::HTMLMediaElement):
(WebCore::HTMLMediaElement::loadInternal):
(WebCore::HTMLMediaElement::loadResource):
(WebCore::HTMLMediaElement::suspend):
(WebCore::HTMLMediaElement::resume):
(WebCore::HTMLMediaElement::removeBehaviorsRestrictionsAfterFirstUserGesture):
(WebCore::HTMLMediaElement::mediaState):
- html/HTMLMediaElement.h:
(WebCore::HTMLMediaElement::mediaSession):
- html/MediaElementSession.cpp: Renamed from Source/WebCore/html/HTMLMediaSession.cpp.
(WebCore::restrictionName):
(WebCore::MediaElementSession::MediaElementSession):
(WebCore::MediaElementSession::registerWithDocument):
(WebCore::MediaElementSession::unregisterWithDocument):
(WebCore::MediaElementSession::addBehaviorRestriction):
(WebCore::MediaElementSession::removeBehaviorRestriction):
(WebCore::MediaElementSession::playbackPermitted):
(WebCore::MediaElementSession::dataLoadingPermitted):
(WebCore::MediaElementSession::fullscreenPermitted):
(WebCore::MediaElementSession::pageAllowsDataLoading):
(WebCore::MediaElementSession::pageAllowsPlaybackAfterResuming):
(WebCore::MediaElementSession::showPlaybackTargetPicker):
(WebCore::MediaElementSession::hasWirelessPlaybackTargets):
(WebCore::MediaElementSession::wirelessVideoPlaybackDisabled):
(WebCore::MediaElementSession::setWirelessVideoPlaybackDisabled):
(WebCore::MediaElementSession::setHasPlaybackTargetAvailabilityListeners):
(WebCore::MediaElementSession::setPlaybackTarget):
(WebCore::MediaElementSession::targetAvailabilityChangedTimerFired):
(WebCore::MediaElementSession::externalOutputDeviceAvailableDidChange):
(WebCore::MediaElementSession::canPlayToWirelessPlaybackTarget):
(WebCore::MediaElementSession::isPlayingToWirelessPlaybackTarget):
(WebCore::MediaElementSession::setShouldPlayToPlaybackTarget):
(WebCore::MediaElementSession::mediaStateDidChange):
(WebCore::MediaElementSession::effectivePreloadForElement):
(WebCore::MediaElementSession::requiresFullscreenForVideoPlayback):
(WebCore::MediaElementSession::mediaEngineUpdated):
(WebCore::MediaElementSession::allowsAlternateFullscreen):
(WebCore::MediaElementSession::maximumMediaSourceBufferSize):
- html/MediaElementSession.h: Renamed from Source/WebCore/html/HTMLMediaSession.h.
- testing/Internals.cpp:
(WebCore::Internals::setMediaElementRestrictions):
(WebCore::Internals::setAudioContextRestrictions):
- 3:41 PM Changeset in webkit [184965] by
-
- 2 edits in trunk/Source/WebCore
Crash under ICU with ASAN during editing/selection/move-by-word-visually-crash-test-5.html
https://bugs.webkit.org/show_bug.cgi?id=145429
<rdar://problem/20992218>
Reviewed by Alexey Proskuryakov.
WebKit uses some strings which contain the lower 8-bits of UTF-16 (thereby saving space). However,
ICU doesn't understand this encoding. When we want to use ICU functions with strings in this encoding,
we create a UTextProvider which converts our encoded strings to UTF-16 for ICU, one chunk at a time.
This object contains a vtable which we populate to perform the conversion.
The WebKit function which actually returns the UTF-16 chunks has two relevant arguments: an index into
the encoded string which ICU is requesting, and a direction from that index which ICU is interested
in. This function populates a "chunk" which is characterized by a pointer to a buffer, the length of
the populated data in the buffer, and an offset into the chunk which represents the index that the
requested character was put into.
When ICU requests data going backward, we fill in the chunk accordingly, with the requested character
all the way at the end. We then set the offset equal to the length of the buffer. However, this length
value is stale from the previous time the function ran. Therefore, ICU was reading the wrong index in
the chunk when expecting the requested character.
Covered by editing/selection/move-by-word-visually-crash-test-5.html.
- platform/text/icu/UTextProviderLatin1.cpp:
(WebCore::uTextLatin1Access):
- 3:40 PM Changeset in webkit [184964] by
-
- 2 edits in trunk/Source/WebKit2
Make sure that we create the private WebKit.framework symlinks last
https://bugs.webkit.org/show_bug.cgi?id=145442
<rdar://problem/21142213>
Reviewed by Dan Bernstein.
Move the "Add Symlinks in /System/Library/PrivateFrameworks/WebKit.framework" build phase to
the "All" aggregate target. Also update the script now that INSTALL_PATH is different.
- WebKit2.xcodeproj/project.pbxproj:
- 3:31 PM Changeset in webkit [184963] by
-
- 2 edits in trunk/Source/WebCore
[iOS] Fix controls bug caused by r184794
https://bugs.webkit.org/show_bug.cgi?id=145445
Reviewed by Dean Jackson.
- Modules/mediacontrols/mediaControlsiOS.js:
(ControllerIOS.prototype.showControls): Check video.controls instead of trying to call
showInlinePlaybackPlaceholderOnly, which doesn't exist.
- 2:55 PM Changeset in webkit [184962] by
-
- 18 edits in trunk/Source
Code clean up for extracting information from the mix of WritingMode and TextDirection
https://bugs.webkit.org/show_bug.cgi?id=143399
Patch by Said Abou-Hallawa <sabouhallawa@apple.com> on 2015-05-28
Reviewed by Darin Adler.
Source/WebCore:
There should be eight TextFlows. A TextFlow consists of two directions.
The first one controls the inline growing direction and the second one
controls the block growing direction. We are going to use the cardinal
directions to define these combinations.
We need to extract the following information easily from the TextFlow
-- isTextFlowReversed() RTL
-- isTextFlowFlipped() the opposite direction to normal
-- isTextFlowVertical() The line grows vertically
Mixing WritingMode and TextDirection into TextFlow and choosing their values
carefully to ease implementing the above functions can make things clearer.
- Modules/notifications/Notification.h:
- platform/graphics/TextRun.h:
- platform/ios/SelectionRect.h:
- platform/text/BidiResolver.h:
- platform/text/StringWithDirection.h:
- platform/text/TextFlags.h:
(WebCore::isLeftToRightDirection): Deleted.
enum TextDirection is moved from TextFlags.h to WritingMode.h. So include
WritingMode.h instead of including TextFlags.h.
- css/CSSProperty.cpp:
(WebCore::resolveToPhysicalProperty): Instead of handling the eight possible
cases of the text flows one by one, we can map from a LogicalSide to a
PhysicalSide using the TextFlow value.
- platform/LengthBox.h:
(WebCore::BoxExtent::before):
(WebCore::BoxExtent::after):
(WebCore::BoxExtent::start):
(WebCore::BoxExtent::end):
Use mapLogicalSideToPhysicalSide() to convert from a LogicalSide to a PhysicalSide.
- platform/text/WritingMode.h:
(WebCore::isLeftToRightDirection): Usually the default enum is defined
first. And usually the default value is equal to zero. Since LTR is the
default for the text direction, it should be defined before RTL.
(WebCore::makeTextFlow): Returns a TextFlow given a WritingMode and a
TextDirection.
(WebCore::isReversedTextFlow): Returns true if the text direction is RTL.
(WebCore::isFlippedTextFlow):
(WebCore::isFlippedWritingMode): Returns true if the block growing direction
is in the opposite direction of normal.
(WebCore::isVerticalTextFlow):
(WebCore::isVerticalWritingMode): Returns true if the line grows vertically.
(WebCore::isHorizontalWritingMode):
(WebCore::isFlippedLinesWritingMode): Use isVerticalWritingMode() and
isFlippedWritingMode() to answer these questions.
(WebCore::isHorizontalPhysicalSide):
(WebCore::mirrorPhysicalSide): top <-> bottom and left <-> right conversion.
(WebCore::rotatePhysicalSide): top <-> left and right <-> bottom conversion.
(WebCore::mapLogicalSideToPhysicalSide): Maps a LogicalSide to a PhysicalSide
using either the TextFlow or the WritingMode.
(WebCore::isFlippedBlocksWritingMode): Deleted.
- rendering/RenderMultiColumnSet.cpp:
(WebCore::RenderMultiColumnSet::collectLayerFragments):
(WebCore::RenderMultiColumnSet::columnTranslationForOffset):
- rendering/shapes/Shape.cpp:
(WebCore::physicalRectToLogical):
(WebCore::physicalPointToLogical):
- rendering/style/RenderStyle.h:
Rename isFlippedBlocksWritingMode() to be isFlippedWritingMode() so we can
be consistent with other WritingMode query functions.
Source/WebKit2:
- Shared/WebPopupItem.h:
- UIProcess/API/C/efl/WKAPICastEfl.h:
- UIProcess/WebPopupMenuProxy.h:
enum TextDirection is moved from TextFlags.h to WritingMode.h. So include
WritingMode.h instead of including TextFlags.h.
- 2:05 PM Changeset in webkit [184961] by
-
- 2 edits in trunk/Source/WebKit2
Disable network cache for old clients
https://bugs.webkit.org/show_bug.cgi?id=145418
Reviewed by Andreas Kling.
Follow-up: Don't do linked-on-or-after checks for iOS Safari.
- UIProcess/Cocoa/VersionChecks.mm:
(WebKit::linkTimeVersion):
(WebKit::linkedOnOrAfter):
- 1:57 PM Changeset in webkit [184960] by
-
- 2 edits2 adds in trunk/Source/JavaScriptCore
[iOS8][ARMv7(s)] Optimized Object.create in 'use strict' context sometimes breaks.
https://bugs.webkit.org/show_bug.cgi?id=138038
Reviewed by Michael Saboff.
TL;DR: sometimes the baseline JIT could accidentally nuke the tag before calling
to C++, making put_by_id behave erratically.
The bug was that put_by_id would randomly not work correctly in 32bits. It happened
in the baseline JIT if we were unlucky enough:
-The code get hot enough and the structure is stable so we get a fast path for
put_by_id.
-We repatch the fast-path branch with a stub generated by
emitPutTransitionStubAndGetOldStructure().
-In emitPutTransitionStubAndGetOldStructure(), we only preserve the payload of the base
register, the tag register is ignored.
-emitPutTransitionStubAndGetOldStructure() allocate 2 to 3 registers. Any of those
could be the one used for the base's tag before the fast path and the value is trashed.
-If we hit one of the failure case, we fallback to the slow path, but we destroyed
the tag pointer.
-We now have unrelated bits in the tag, the most likely value type is now "double"
and we fail the put_by_id because we try to set a property on a number.
The most obvious solution would be to change emitPutTransitionStubAndGetOldStructure()
to preserve the tag register in addition to the value register.
I decided against that option because of the added complexity. The DFG does not need
that case, so I would have to add branches everywhere to distinguish the cases
were we need to preserve the tag or not.
Instead, I just load the tag back from memory in the slow path. The function in the slow
path is several order of magnitude slower than a load, it is not worth eliminating it,
especially in baseline JIT.
I also discovered 4 useless loads in the fast path, so even with my extra load, this patch
makes the baseline faster :)
- jit/JITPropertyAccess32_64.cpp:
(JSC::JIT::emitSlow_op_put_by_id):
(JSC::JIT::emit_op_put_by_id): Deleted.
- tests/stress/put-by-id-on-new-object-after-prototype-transition-non-strict.js: Added.
(opaqueNewObject):
(putValueOnNewObject):
- tests/stress/put-by-id-on-new-object-after-prototype-transition-strict.js: Added.
(string_appeared_here.opaqueNewObject):
(putValueOnNewObject):
- 1:46 PM Changeset in webkit [184959] by
-
- 4 edits in trunk/Source/JavaScriptCore
[JSC] reduction the iteration count of the DoubleRep stress tests
Once again, I used big numbers for manual testing and I forgot to fix them before landing.
- tests/stress/double-rep-with-non-cell.js:
- tests/stress/double-rep-with-null.js:
- tests/stress/double-rep-with-undefined.js:
- 12:14 PM Changeset in webkit [184958] by
-
- 2 edits in trunk/Source/JavaScriptCore
Add debug mode assertions for accessors casting JSC::DFG::Node.m_opInfo
https://bugs.webkit.org/show_bug.cgi?id=145441
Reviewed by Filip Pizlo.
Most accessor functions casting m_opInfo in JSC::DFG::Node are
performing debug checks that they are only accessed for node types that
should have them. This patch adds similar checks for the accessors that
were missing them.
- dfg/DFGNode.h:
(JSC::DFG::Node::watchpointSet):
(JSC::DFG::Node::storagePointer):
(JSC::DFG::Node::multiGetByOffsetData):
(JSC::DFG::Node::multiPutByOffsetData):
(JSC::DFG::Node::hasTypeLocation):
(JSC::DFG::Node::typeLocation):
(JSC::DFG::Node::hasBasicBlockLocation):
(JSC::DFG::Node::basicBlockLocation):
- 10:45 AM Changeset in webkit [184957] by
-
- 5 edits in branches/safari-601.1.32-branch/Source
Versioning.
- 10:40 AM Changeset in webkit [184956] by
-
- 8 edits in trunk/Source
Add ENABLE_MEDIA_SESSION feature flag (which is off by default).
https://bugs.webkit.org/show_bug.cgi?id=145415
Patch by Matt Rajca <mrajca@apple.com> on 2015-05-28
Reviewed by Eric Carlson.
- Configurations/FeatureDefines.xcconfig:
- 5:16 AM Changeset in webkit [184955] by
-
- 9 edits in trunk
[Streams API] ReadableStreamReader should not be exposed
https://bugs.webkit.org/show_bug.cgi?id=143564
Reviewed by Darin Adler.
Source/WebCore:
Covered by rebased tests.
- Modules/streams/ReadableStreamReader.idl: Adding NoInterfaceObject to ReadableStreamReader.idl.
LayoutTests:
Rebasing tests as ReadableStreamReader is no longer defined globally.
- js/dom/global-constructors-attributes-expected.txt:
- platform/efl/js/dom/global-constructors-attributes-expected.txt:
- platform/gtk/js/dom/global-constructors-attributes-expected.txt:
- platform/mac-mavericks/js/dom/global-constructors-attributes-expected.txt:
- platform/mac/js/dom/global-constructors-attributes-expected.txt:
- platform/win/js/dom/global-constructors-attributes-expected.txt:
- 4:30 AM Changeset in webkit [184954] by
-
- 9 edits1 copy1 delete in trunk
[CMake] Improve detection and usage of GL/GLES/EGL libraries.
https://bugs.webkit.org/show_bug.cgi?id=145408
Reviewed by Carlos Garcia Campos.
.:
- Source/cmake/FindEGL.cmake: Improve detection of EGL libraries.
- Source/cmake/FindGLES.cmake: Removed. It was used by the EGL port.
Remove it and make the EGL port use the improved FindOpenGLES2.cmake
instead.
- Source/cmake/FindOpenGL.cmake: Added. Add module to detect OpenGL
libraries. Detect also GLX libraries.
- Source/cmake/FindOpenGLES2.cmake: Improve detection of OpenGLES-v2
libraries. Use find_path() to get the include path.
- Source/cmake/OptionsEfl.cmake: Use now the improved FindOpenGLES2
module.
- Source/cmake/OptionsGTK.cmake: Set default value for ENABLE_GLES2
depending on the libraries found on the system.
Move the detection of GLX (and the include of CMakePushCheckState)
to FindOpenGL.cmake.
Ensure that we only define USE_GLX when we build with OpenGL
(but not with GLESv2).
Source/WebCore:
No new tests, no behavior changes.
- CMakeLists.txt: Ensure that we include the libraries and includes
for the GL/GLESv2/EGL libraries before including the ANGLE directories.
Define also any CFLAG that the system GL/GLESv2/EGL libraries may need.
- PlatformEfl.cmake: Remove some includes that are now unneeded,
because we are including now the EGL libraries on CMakeLists.txt
- PlatformGTK.cmake: Remove unneeded include (We are including the EGL
libraries now on CMakeLists.txt)
- 3:12 AM Changeset in webkit [184953] by
-
- 2 edits9 adds in trunk/Source/WebCore
Binding generator should support interfaces with CustomConstructor and NoInterfaceObject
https://bugs.webkit.org/show_bug.cgi?id=145016
Reviewed by Darin Adler.
Updated code generator to generate the necessary declarations (constructor property in the prototype, the constructor class).
In the case of CustomConstructor with NoInterfaceObject, the implementation of the constructor property related JS function is changed as follow:
It creates a constructor object and shadows the constructor property JS function with this object.
This ensures that only one constructor object is created per prototype.
The constructor is not exposed as it is not added to the global object constructor map.
NoInterfaceObject is asking to not make the interface visible, which is contradictory with having a visible constructor of the interface.
The case for having NoInterfaceObject and CustomConstructor is the following:
We do not want to expose the interface but we still want to have a constructor property on the prototype of objects implementing the interface.
One such case is https://streams.spec.whatwg.org/#globals
Added TestCustomConstructorWithNoInterfaceObject.idl to check that case.
- bindings/scripts/CodeGeneratorJS.pm:
(GenerateHeader):
(GenerateAttributesHashTable):
(GenerateImplementation):
(HasConstructorProperty):
- bindings/scripts/test/GObject/WebKitDOMTestCustomConstructorWithNoInterfaceObject.cpp: Added.
(WebKit::kit):
(WebKit::core):
(WebKit::wrapTestCustomConstructorWithNoInterfaceObject):
(webkit_dom_test_custom_constructor_with_no_interface_object_finalize):
(webkit_dom_test_custom_constructor_with_no_interface_object_constructor):
(webkit_dom_test_custom_constructor_with_no_interface_object_class_init):
(webkit_dom_test_custom_constructor_with_no_interface_object_init):
- bindings/scripts/test/GObject/WebKitDOMTestCustomConstructorWithNoInterfaceObject.h: Added.
- bindings/scripts/test/GObject/WebKitDOMTestCustomConstructorWithNoInterfaceObjectPrivate.h: Added.
- bindings/scripts/test/JS/JSTestCustomConstructorWithNoInterfaceObject.cpp: Added.
(WebCore::JSTestCustomConstructorWithNoInterfaceObjectPrototype::create):
(WebCore::JSTestCustomConstructorWithNoInterfaceObjectPrototype::createStructure):
(WebCore::JSTestCustomConstructorWithNoInterfaceObjectPrototype::JSTestCustomConstructorWithNoInterfaceObjectPrototype):
(WebCore::JSTestCustomConstructorWithNoInterfaceObjectConstructor::create):
(WebCore::JSTestCustomConstructorWithNoInterfaceObjectConstructor::createStructure):
(WebCore::JSTestCustomConstructorWithNoInterfaceObjectConstructor::JSTestCustomConstructorWithNoInterfaceObjectConstructor):
(WebCore::JSTestCustomConstructorWithNoInterfaceObjectConstructor::finishCreation):
(WebCore::JSTestCustomConstructorWithNoInterfaceObjectConstructor::getConstructData):
(WebCore::JSTestCustomConstructorWithNoInterfaceObjectPrototype::finishCreation):
(WebCore::JSTestCustomConstructorWithNoInterfaceObject::JSTestCustomConstructorWithNoInterfaceObject):
(WebCore::JSTestCustomConstructorWithNoInterfaceObject::createPrototype):
(WebCore::JSTestCustomConstructorWithNoInterfaceObject::getPrototype):
(WebCore::JSTestCustomConstructorWithNoInterfaceObject::destroy):
(WebCore::JSTestCustomConstructorWithNoInterfaceObject::~JSTestCustomConstructorWithNoInterfaceObject):
(WebCore::jsTestCustomConstructorWithNoInterfaceObjectConstructor):
(WebCore::JSTestCustomConstructorWithNoInterfaceObject::getConstructor):
(WebCore::JSTestCustomConstructorWithNoInterfaceObjectOwner::isReachableFromOpaqueRoots):
(WebCore::JSTestCustomConstructorWithNoInterfaceObjectOwner::finalize):
(WebCore::toJS):
(WebCore::JSTestCustomConstructorWithNoInterfaceObject::toWrapped):
- bindings/scripts/test/JS/JSTestCustomConstructorWithNoInterfaceObject.h: Added.
(WebCore::wrapperOwner):
(WebCore::toJS):
- bindings/scripts/test/ObjC/DOMTestCustomConstructorWithNoInterfaceObject.h: Added.
- bindings/scripts/test/ObjC/DOMTestCustomConstructorWithNoInterfaceObject.mm: Added.
(-[DOMTestCustomConstructorWithNoInterfaceObject dealloc]):
(-[DOMTestCustomConstructorWithNoInterfaceObject finalize]):
(core):
(kit):
- bindings/scripts/test/ObjC/DOMTestCustomConstructorWithNoInterfaceObjectInternal.h: Added.
- bindings/scripts/test/TestCustomConstructor.idl: Added.
- 2:40 AM Changeset in webkit [184952] by
-
- 2 edits in trunk/Source/WebKit2
Unreviewed, Fix build break since r184930.
- UIProcess/Storage/StorageManager.cpp:
(WebKit::StorageManager::createTransientLocalStorageMap): Pass a pointer of SecurityOrigin
in isSameSchemeHostPort().
- 2:23 AM Changeset in webkit [184951] by
-
- 2 edits in trunk/Source/WebKit2
Unreviewed. Fixing the build after r184950.
- UIProcess/Storage/StorageManager.cpp:
(WebKit::StorageManager::createSessionStorageMap): getOrCreateStorageArea()
now expects a Ref<WebCore::SecurityOrigin> object.
- 1:50 AM WebKitGTK/2.8.x edited by
- (diff)
- 1:18 AM Changeset in webkit [184950] by
-
- 14 edits in trunk/Source/WebKit2
[WK2] Handle WebCore::SecurityOrigin objects through Ref<> where possible
https://bugs.webkit.org/show_bug.cgi?id=145373
Reviewed by Andreas Kling.
WebCore::SecurityOrigin creation functions in most cases return a
Ref<WebCore::SecurityOrigin> object. In the WebKit2 layer, those objects
were still largely converted to RefPtr<>s.
This patch switches to using Ref<> where possible. The remaining uses of
RefPtr<WebCore::SecurityOrigin> are present due to HashTables, to-be-copied
Vectors, WebCore::DatabaseManager::origins() API and code where non-null
WebCore::SecurityOrigin can't be guaranteed.
- DatabaseProcess/DatabaseProcess.cpp:
(WebKit::DatabaseProcess::deleteIndexedDatabaseEntriesForOrigin):
- NetworkProcess/NetworkProcess.cpp:
(WebKit::cfURLCacheOrigins):
- UIProcess/Storage/LocalStorageDatabase.cpp:
(WebKit::LocalStorageDatabase::create):
(WebKit::LocalStorageDatabase::LocalStorageDatabase):
(WebKit::LocalStorageDatabase::openDatabase):
(WebKit::LocalStorageDatabase::close):
- UIProcess/Storage/LocalStorageDatabase.h:
- UIProcess/Storage/LocalStorageDatabaseTracker.cpp:
(WebKit::LocalStorageDatabaseTracker::origins):
- UIProcess/Storage/LocalStorageDatabaseTracker.h:
- UIProcess/Storage/StorageManager.cpp:
(WebKit::StorageManager::StorageArea::securityOrigin):
(WebKit::StorageManager::TransientLocalStorageNamespace::getOrCreateStorageArea):
(WebKit::StorageManager::TransientLocalStorageNamespace::origins):
(WebKit::StorageManager::TransientLocalStorageNamespace::clearStorageAreasMatchingOrigin):
(WebKit::StorageManager::StorageArea::create):
(WebKit::StorageManager::StorageArea::StorageArea):
(WebKit::StorageManager::StorageArea::openDatabaseAndImportItemsIfNeeded):
(WebKit::StorageManager::LocalStorageNamespace::getOrCreateStorageArea):
(WebKit::StorageManager::LocalStorageNamespace::didDestroyStorageArea):
(WebKit::StorageManager::SessionStorageNamespace::origins):
(WebKit::StorageManager::SessionStorageNamespace::clearStorageAreasMatchingOrigin):
(WebKit::StorageManager::SessionStorageNamespace::getOrCreateStorageArea):
- UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::reachedApplicationCacheOriginQuota):
- WebProcess/ApplicationCache/WebApplicationCacheManager.cpp:
(WebKit::WebApplicationCacheManager::getApplicationCacheOrigins):
- WebProcess/InjectedBundle/InjectedBundle.cpp:
(WebKit::InjectedBundle::setApplicationCacheOriginQuota):
(WebKit::InjectedBundle::resetApplicationCacheOriginQuota):
- WebProcess/ResourceCache/WebResourceCacheManager.cpp:
(WebKit::WebResourceCacheManager::returnCacheOrigins):
(WebKit::WebResourceCacheManager::clearCacheForOrigin):
- WebProcess/WebCoreSupport/WebDatabaseManager.cpp:
(WebKit::WebDatabaseManager::deleteDatabaseWithNameForOrigin):
(WebKit::WebDatabaseManager::deleteDatabasesForOrigin):
(WebKit::WebDatabaseManager::setQuotaForOrigin):
- 1:13 AM Changeset in webkit [184949] by
-
- 2 edits in trunk/Source/WTF
Clean up HashTable constructors
https://bugs.webkit.org/show_bug.cgi?id=145369
Reviewed by Andreas Kling.
Use nullptr to initialize pointer member variables in the HashTable
default constructor. Copy and move constructors can use constructor
delegation instead of replicating all member initializations. Move
constructor should simply call HashTable::swap() instead of replicating
all the swap operations.
- wtf/HashTable.h:
(WTF::KeyTraits>::HashTable):
- 12:44 AM Changeset in webkit [184948] by
-
- 2 edits in trunk/Source/WebKit/mac
Build fix after r184946.
- WebView/WebPreferences.mm:
(-[WebPreferences allowsAirPlayForMediaPlayback]):
(-[WebPreferences setAllowsAirPlayForMediaPlayback:]):
- 12:34 AM Changeset in webkit [184947] by
-
- 19 edits in trunk/Source
[iOS] When viewing an MJPEG stream as the main resource, only the first
frame paints
<https://bugs.webkit.org/show_bug.cgi?id=145185>
<rdar://problem/20124694>
Source/WebCore:
This bug is caused by state not being restored between successive loads
of multipart/x-mixed-replace content. Each part of the stream is not
treated as a wholly new load, so if state is cleared as part of the
previous page being destroyed, it may not be restored when loading
the next stream part.
On WK1, tile cache updates are disabled in FrameView::clear(), which is
called as the previous page is destroyed, but were not being
reenabled when loading new replacement content from the stream.
On WK2, the exposed content rect was being reset to an empty rect as
the previous page was destroyed, but it was not being reset by loading
replacement content.
To fix these issues, I added two WebFrameLoaderClient callbacks;
willReplaceMultipartContent() and didReplaceMultipartContent(). These
are used to save and restore state between successive loads.
Reviewed by Darin Adler.
- loader/DocumentLoader.cpp:
Call the new client callbacks.
(WebCore::DocumentLoader::commitLoad):
If replacing content in a multipart/x-mixed-replace stream, call the
WebFrameLoaderClient's didReplaceMultipartContent().
(WebCore::DocumentLoader::setupForReplace):
Call the WebFrameLoaderClient's willReplaceMultipartContent().
- loader/EmptyClients.h:
- loader/FrameLoaderClient.h:
- page/FrameView.cpp:
(WebCore::FrameView::didReplaceMultipartContent):
Re-enable tile cache updates that were disabled in FrameView::clear().
This fixes the issue for WK1.
- page/FrameView.h:
Declare an exported function, didReplaceMultipartContent(), that WebKit
can call.
Source/WebKit/mac:
Reviewed by Darin Adler.
- WebCoreSupport/WebFrameLoaderClient.h:
- WebCoreSupport/WebFrameLoaderClient.mm:
(WebFrameLoaderClient::didReplaceMultipartContent):
Call the FrameView's didReplaceMultipartContent().
Source/WebKit/win:
Reviewed by Darin Adler.
- WebCoreSupport/WebFrameLoaderClient.h:
Stubbed new functions that aren't used on Windows.
Source/WebKit2:
Reviewed by Darin Adler.
- WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::willReplaceMultipartContent):
Tell the WebPage that we will replace multipart content.
(WebKit::WebFrameLoaderClient::didReplaceMultipartContent):
Tell the WebPage that we did replace multipart content.
- WebProcess/WebCoreSupport/WebFrameLoaderClient.h:
- WebProcess/WebPage/DrawingArea.h:
- WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::willReplaceMultipartContent):
Save the exposed content rect.
(WebKit::WebPage::didReplaceMultipartContent):
Restore the exposed content rect that was saved before this load. This
fixes the issue for WK2.
- WebProcess/WebPage/WebPage.h:
- WebProcess/WebPage/mac/RemoteLayerTreeDrawingArea.h:
- WebProcess/WebPage/mac/RemoteLayerTreeDrawingArea.mm:
(WebKit::RemoteLayerTreeDrawingArea::exposedContentRect):
- 12:27 AM Changeset in webkit [184946] by
-
- 9 edits in trunk/Source
Fix macros for wireless playback
https://bugs.webkit.org/show_bug.cgi?id=145239
rdar://problem/21049541
Reviewed by Anders Carlsson.
Encompass the logic within ENABLE(WIRELESS_TARGET_PLAYBACK). Update property name.
Source/WebKit/mac:
- WebView/WebPreferenceKeysPrivate.h:
- WebView/WebPreferences.h:
- WebView/WebPreferences.mm:
(+[WebPreferences initialize]): Wrap in ENABLE.
(-[WebPreferences allowsAirPlayForMediaPlayback]): Ditto.
(-[WebPreferences setAllowsAirPlayForMediaPlayback:]): Ditto.
- WebView/WebView.mm:
(-[WebView _preferencesChanged:]): Ditto.
Source/WebKit2:
- UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView initWithFrame:configuration:]):
- UIProcess/API/Cocoa/WKWebViewConfiguration.h: Update names. Add a WKDeprecated
category section.
- UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
(-[WKWebViewConfiguration init]): Update name of variables to reflect new property names.
(-[WKWebViewConfiguration copyWithZone:]): Ditto.
(-[WKWebViewConfiguration mediaPlaybackAllowsAirPlay]): Get and set off of the new
variable.
(-[WKWebViewConfiguration setMediaPlaybackAllowsAirPlay:]): Ditto.
(-[WKWebViewConfiguration mediaPlaybackRequiresUserAction]): Ditto.
(-[WKWebViewConfiguration setMediaPlaybackRequiresUserAction:]): Ditto.
May 27, 2015:
- 11:47 PM WebKitGTK/2.8.x edited by
- (diff)
- 11:14 PM Changeset in webkit [184945] by
-
- 6 edits in trunk/Source/WebCore
Remove unused ResourceRequest "user initiated" flag.
https://bugs.webkit.org/show_bug.cgi?id=145428
Reviewed by Alexey Proskuryakov.
No new tests (No change in behavior).
- platform/ios/WebCoreSystemInterfaceIOS.mm:
- platform/mac/WebCoreSystemInterface.mm:
- platform/network/ResourceRequestBase.h:
(WebCore::ResourceRequestBase::initiatedByUserGesture): Deleted.
(WebCore::ResourceRequestBase::setInitiatedByUserGesture): Deleted.
(WebCore::ResourceRequestBase::encodeWithoutPlatformData): Deleted.
(WebCore::ResourceRequestBase::decodeWithoutPlatformData): Deleted.
- platform/network/cf/ResourceRequestCFNet.cpp:
(WebCore::ResourceRequest::doUpdatePlatformRequest): Deleted.
(WebCore::ResourceRequest::doUpdateResourceRequest): Deleted.
- platform/network/cocoa/ResourceRequestCocoa.mm:
(WebCore::ResourceRequest::doUpdateResourceRequest): Deleted.
(WebCore::ResourceRequest::doUpdatePlatformRequest): Deleted.
- 11:12 PM Changeset in webkit [184944] by
-
- 2 edits in trunk/Source/WTF
[Mac] Fix clang static analyzer build
https://bugs.webkit.org/show_bug.cgi?id=145426
rdar://problem/20947408
Reviewed by Dan Bernstein.
- wtf/spi/darwin/XPCSPI.h: Correct declarations of xpc_retain and xpc_release.
- 10:52 PM Changeset in webkit [184943] by
-
- 1 edit3 deletes in trunk/LayoutTests
Delete media/controls-layers.html. It's a bad test that fails and asserts, and there
are apparently no plans to fix any of that (see https://bugs.webkit.org/show_bug.cgi?id=145420).
- media/controls-layers.html: Removed.
- platform/mac-mavericks/media/controls-layers-expected.txt: Removed.
- platform/mac/media/controls-layers-expected.txt: Removed.
- 10:47 PM Changeset in webkit [184942] by
-
- 6 edits in trunk
Array.of should work with other constructors
https://bugs.webkit.org/show_bug.cgi?id=145365
Source/JavaScriptCore:
Per https://people.mozilla.org/~jorendorff/es6-draft.html#sec-array.of
step 4
Patch by Jordan Harband <ljharb@gmail.com> on 2015-05-27
Reviewed by Yusuke Suzuki.
- builtins/ArrayConstructor.js:
(of):
- runtime/ArrayConstructor.cpp:
(JSC::arrayConstructorOf): Deleted.
LayoutTests:
Patch by Jordan Harband <ljharb@gmail.com> on 2015-05-27
Reviewed by Yusuke Suzuki.
- js/array-of-expected.txt:
- js/script-tests/array-of.js:
(Foo):
- 9:52 PM Changeset in webkit [184941] by
-
- 26 edits in trunk/Source
Add a "should open urls externally" flag to DocumentLoader.
rdar://problem/21025301 and https://bugs.webkit.org/show_bug.cgi?id=145417
Reviewed by Geoff Garen.
Source/WebCore:
No new tests (No change in behavior, testable at API level with a followup patch)
- loader/DocumentLoader.cpp:
(WebCore::DocumentLoader::setTriggeringAction): Apply the current flag value to the
new NavigationAction so it can be used later.
- loader/DocumentLoader.h:
(WebCore::DocumentLoader::setShouldOpenExternalURLsPolicy):
(WebCore::DocumentLoader::shouldOpenExternalURLsPolicy):
(WebCore::DocumentLoader::setTriggeringAction): Deleted.
FrameLoadRequest should hold the new flag to pass into FrameLoader machinery:
- loader/FrameLoadRequest.h:
(WebCore::FrameLoadRequest::setShouldOpenExternalURLsPolicy):
(WebCore::FrameLoadRequest::shouldOpenExternalURLsPolicy):
- loader/FrameLoader.cpp:
(WebCore::FrameLoader::load): Set the flag on the newly created DocumentLoader but
only if it is for the main frame.
- loader/FrameLoaderTypes.h: Add a new enum for the flag.
Navigation action should hold the new flag to pass to the policy delegate:
- loader/NavigationAction.cpp:
(WebCore::NavigationAction::NavigationAction):
- loader/NavigationAction.h:
(WebCore::NavigationAction::shouldOpenExternalURLsPolicy):
(WebCore::NavigationAction::setShouldOpenExternalURLsPolicy):
- loader/PolicyChecker.cpp:
(WebCore::PolicyChecker::checkNavigationPolicy): Include the flag from the DocumentLoader
in the NavigationAction.
Source/WebKit2:
Rename "user initiated" to "should open external URLs":
- Shared/API/Cocoa/_WKNSURLRequestExtras.h:
- Shared/API/Cocoa/_WKNSURLRequestExtras.mm:
(-[NSURLRequest _web_shouldOpenExternalURLs]):
(-[NSMutableURLRequest _web_setShouldOpenExternalURLs:]):
(-[NSURLRequest _web_isUserInitiated]): Deleted.
(-[NSMutableURLRequest _web_setIsUserInitiated:]): Deleted.
Hold the new flag as a bool, and encode/decode it across IPC:
- Shared/NavigationActionData.cpp:
(WebKit::NavigationActionData::encode):
(WebKit::NavigationActionData::decode):
(WebKit::NavigationActionData::NavigationActionData): Deleted.
- Shared/NavigationActionData.h:
- UIProcess/API/APINavigationAction.h: Add accessor for the bool flag.
- UIProcess/API/C/WKPage.cpp:
(WKPageLoadURLWithUserData): Use "ShouldNotAllow" in the legacy C API for now.
(WKPageLoadURLRequestWithUserData): Ditto.
- UIProcess/API/Cocoa/WKBrowsingContextController.mm:
(-[WKBrowsingContextController loadRequest:userData:]): Set the value of the flag
based on whether or not the initiating NSURLRequest was user initiated.
- UIProcess/API/Cocoa/WKNavigationAction.mm:
(-[WKNavigationAction _shouldOpenExternalURLs]): Add accessor for the bool flag.
- UIProcess/API/Cocoa/WKNavigationActionPrivate.h:
- UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView loadRequest:]): Set the value of the flag based on whether or not the
initiating NSURLRequest was user initiated.
- UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::loadRequest): Pass the value down to the WebProcess to load.
(WebKit::WebPageProxy::loadFile): Ditto, but always with "ShouldNotAllow".
- UIProcess/WebPageProxy.h:
- WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::dispatchDecidePolicyForNavigationAction): Translate
the NavigationAction flag to a bool in the NavigationActionData object.
- WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::loadRequest): Set the value of the flag in the FrameLoadRequest that
will be loaded.
- WebProcess/WebPage/WebPage.h:
- WebProcess/WebPage/WebPage.messages.in:
- 9:22 PM Changeset in webkit [184940] by
-
- 66 edits in trunk/Source/WebCore
Purge PassRefPtr in WebCore/Modules - 2
https://bugs.webkit.org/show_bug.cgi?id=145304
Reviewed by Darin Adler.
As a step to remove PassRefPtr, this patch replaces PassRefPtr with Ref or RefPtr.
No new tests, no behavior changes
- Modules/mediasource/SourceBufferList.h:
- Modules/mediastream/AllAudioCapabilities.h:
(WebCore::AllAudioCapabilities::create):
- Modules/mediastream/AllVideoCapabilities.h:
(WebCore::AllVideoCapabilities::create):
- Modules/mediastream/CapabilityRange.cpp:
(WebCore::CapabilityRange::create):
- Modules/mediastream/CapabilityRange.h:
- Modules/mediastream/MediaSourceStates.cpp:
(WebCore::MediaSourceStates::create):
- Modules/mediastream/MediaStream.cpp:
(WebCore::MediaStream::create):
- Modules/mediastream/MediaStreamTrack.cpp:
(WebCore::MediaStreamTrack::create):
- Modules/mediastream/MediaStreamTrackSourcesRequest.cpp:
(WebCore::MediaStreamTrackSourcesRequest::create):
- Modules/mediastream/MediaTrackConstraint.cpp:
(WebCore::MediaTrackConstraint::create):
- Modules/mediastream/MediaTrackConstraint.h:
- Modules/mediastream/RTCConfiguration.h:
(WebCore::RTCConfiguration::create):
- Modules/mediastream/RTCIceServer.h:
(WebCore::RTCIceServer::create):
- Modules/mediastream/RTCSessionDescription.cpp:
(WebCore::RTCSessionDescription::create):
- Modules/mediastream/RTCSessionDescription.h:
- Modules/mediastream/RTCSessionDescriptionRequestImpl.cpp:
(WebCore::RTCSessionDescriptionRequestImpl::create):
- Modules/mediastream/RTCStatsReport.cpp:
(WebCore::RTCStatsReport::create):
- Modules/mediastream/RTCStatsReport.h:
- Modules/mediastream/RTCStatsRequestImpl.cpp:
(WebCore::RTCStatsRequestImpl::create):
- Modules/mediastream/RTCStatsRequestImpl.h:
- Modules/mediastream/RTCStatsResponse.cpp:
(WebCore::RTCStatsResponse::create):
- Modules/mediastream/RTCStatsResponse.h:
- Modules/mediastream/RTCVoidRequestImpl.cpp:
(WebCore::RTCVoidRequestImpl::create):
- Modules/mediastream/RTCVoidRequestImpl.h:
- Modules/mediastream/SourceInfo.cpp:
(WebCore::SourceInfo::create):
- Modules/mediastream/SourceInfo.h:
- Modules/mediastream/UserMediaRequest.cpp:
(WebCore::UserMediaRequest::create):
- Modules/mediastream/UserMediaRequest.h:
- Modules/quota/StorageInfo.h:
(WebCore::StorageInfo::create):
- Modules/quota/StorageQuota.h:
(WebCore::StorageQuota::create):
- Modules/speech/SpeechSynthesis.cpp:
(WebCore::SpeechSynthesis::create):
- Modules/speech/SpeechSynthesis.h:
- Modules/speech/SpeechSynthesisUtterance.cpp:
(WebCore::SpeechSynthesisUtterance::create):
- Modules/speech/SpeechSynthesisUtterance.h:
- Modules/speech/SpeechSynthesisVoice.cpp:
(WebCore::SpeechSynthesisVoice::create):
- Modules/speech/SpeechSynthesisVoice.h:
- Modules/webaudio/ChannelMergerNode.cpp:
(WebCore::ChannelMergerNode::create):
- Modules/webaudio/ChannelMergerNode.h:
- Modules/webaudio/ChannelSplitterNode.cpp:
(WebCore::ChannelSplitterNode::create):
- Modules/webaudio/ChannelSplitterNode.h:
- Modules/webaudio/DefaultAudioDestinationNode.h:
(WebCore::DefaultAudioDestinationNode::create):
- Modules/webaudio/GainNode.h:
(WebCore::GainNode::create):
- Modules/webaudio/MediaElementAudioSourceNode.cpp:
(WebCore::MediaElementAudioSourceNode::create):
- Modules/webaudio/MediaElementAudioSourceNode.h:
- Modules/webaudio/MediaStreamAudioDestinationNode.cpp:
(WebCore::MediaStreamAudioDestinationNode::create):
- Modules/webaudio/MediaStreamAudioDestinationNode.h:
- Modules/webaudio/MediaStreamAudioSource.cpp:
(WebCore::MediaStreamAudioSource::create):
- Modules/webaudio/MediaStreamAudioSource.h:
- Modules/webaudio/MediaStreamAudioSourceNode.cpp:
(WebCore::MediaStreamAudioSourceNode::create):
- Modules/webaudio/MediaStreamAudioSourceNode.h:
- Modules/webaudio/OscillatorNode.cpp:
(WebCore::OscillatorNode::setType):
- Modules/webaudio/PannerNode.h:
(WebCore::PannerNode::create):
- Modules/webaudio/PeriodicWave.cpp:
(WebCore::PeriodicWave::create):
(WebCore::PeriodicWave::createSine):
(WebCore::PeriodicWave::createSquare):
(WebCore::PeriodicWave::createSawtooth):
(WebCore::PeriodicWave::createTriangle):
- Modules/webaudio/PeriodicWave.h:
- Modules/webaudio/ScriptProcessorNode.cpp:
(WebCore::ScriptProcessorNode::create):
- Modules/webaudio/ScriptProcessorNode.h:
- Modules/webaudio/WaveShaperNode.h:
(WebCore::WaveShaperNode::create):
- 7:33 PM Changeset in webkit [184939] by
-
- 2 edits in tags/Safari-601.1.33.1/Source/WebCore
Roll out r184794. rdar://problem/21130157
- 7:30 PM Changeset in webkit [184938] by
-
- 5 edits in tags/Safari-601.1.33.1/Source
Versioning.
- 7:27 PM Changeset in webkit [184937] by
-
- 1 copy in tags/Safari-601.1.33.1
New tag.
- 7:15 PM Changeset in webkit [184936] by
-
- 1 copy in tags/Safari-601.1.32.7
New tag.
- 6:55 PM Changeset in webkit [184935] by
-
- 2 edits in trunk/Source/WebInspectorUI
Web Inspector: Show Page Source does not switch to Resources tab if inspector was closed
https://bugs.webkit.org/show_bug.cgi?id=145423
Patch by Joseph Pecoraro <Joseph Pecoraro> on 2015-05-27
Reviewed by Timothy Hatcher.
- UserInterface/Base/Main.js:
(WebInspector._frameWasAdded.delayedWork):
(WebInspector._frameWasAdded):
In the special case where we are holding to show a frame's source code when
the frame becomes available (_frameIdentifierToShowSourceCodeWhenAvailable)
we should be forcing to show the tab.
- 6:51 PM Changeset in webkit [184934] by
-
- 2 edits in trunk/Source/WebInspectorUI
Web Inspector: Toggling an inherited property sometimes stops working
https://bugs.webkit.org/show_bug.cgi?id=145416
Patch by Joseph Pecoraro <Joseph Pecoraro> on 2015-05-27
Reviewed by Timothy Hatcher.
- UserInterface/Models/DOMNodeStyles.js:
(WebInspector.DOMNodeStyles.prototype._parseStyleDeclarationPayload):
Ensure we update the existing model objects with the new state information
before possibly bailing. This may produce changed events that update
parts of the inspector appropriately.
- 6:30 PM Changeset in webkit [184933] by
-
- 7 edits6 adds in trunk
[JSC] Add undefined->double conversion to DoubleRep
https://bugs.webkit.org/show_bug.cgi?id=145293
Patch by Benjamin Poulain <bpoulain@apple.com> on 2015-05-27
Reviewed by Filip Pizlo.
Source/JavaScriptCore:
This patch adds undefined to double conversion to the DoubleRep
node for the cases were we speculate "undefined" as part of the types
processed.
The use case is doing math with accidental out-of-bounds access. For example,
something like:
for (var i = 0; i <= length; ++i)
ouptput += array[i];
would cause us to OSR exit every time i === length.
When hitting one of those cases, we would already speculate double math,
but the DoubleRep node was unable to convert the undefined and would exit.
With this patch the use kind NotCellUse cover this conversion for DoubleRep.
I have been quite conservative so in general we will not find "undefined"
until a few recompile but being optimistic seems better since this is a corner case.
This patch is a 80% progression on WebXPRT's DNA Sequencing test.
- dfg/DFGAbstractInterpreterInlines.h:
(JSC::DFG::AbstractInterpreter<AbstractStateType>::executeEffects):
- dfg/DFGFixupPhase.cpp:
(JSC::DFG::FixupPhase::injectTypeConversionsForEdge):
- dfg/DFGNode.h:
(JSC::DFG::Node::sawUndefined):
- dfg/DFGPredictionPropagationPhase.cpp:
(JSC::DFG::PredictionPropagationPhase::propagate):
- dfg/DFGSafeToExecute.h:
(JSC::DFG::SafeToExecuteEdge::operator()):
- dfg/DFGSpeculativeJIT.cpp:
(JSC::DFG::SpeculativeJIT::compileDoubleRep):
- dfg/DFGUseKind.cpp:
(WTF::printInternal):
- dfg/DFGUseKind.h:
(JSC::DFG::typeFilterFor):
- ftl/FTLCapabilities.cpp:
(JSC::FTL::canCompile):
- ftl/FTLLowerDFGToLLVM.cpp:
(JSC::FTL::LowerDFGToLLVM::compileDoubleRep):
(JSC::FTL::LowerDFGToLLVM::jsValueToDouble):
- tests/stress/double-rep-with-undefined.js: Added.
(addArgsNumberAndUndefined):
(addArgsInt32AndUndefined):
(testFallbackWithDouble):
(addArgsDoubleAndUndefined):
(testFallbackWithObject.):
(testFallbackWithObject):
(addArgsOnlyUndefined):
(testFallbackWithString):
LayoutTests:
- js/regress/math-with-out-of-bounds-array-values-expected.txt: Added.
- js/regress/math-with-out-of-bounds-array-values.html: Added.
- js/regress/script-tests/math-with-out-of-bounds-array-values.js: Added.
- 5:39 PM Changeset in webkit [184932] by
-
- 4 edits3 adds in trunk
REGRESSION (r183820): webkit.org/blog/ background painting issue on reload, when the page contains videos
https://bugs.webkit.org/show_bug.cgi?id=145420
Reviewed by Dean Jackson.
Source/WebCore:
After r183820, the media controls no longer had a wrapper that created CSS stacking context.
The media controls on Mac use mix-blend-mode, which causes the compositing code to look for
a stacking context ancestor and make it composited. After this change, it would walk up
to a layer outside of the media element (e.g. the document element's layer), and make
that composited. This triggered bugs with root background painting.
Prevent mix-blend-mode affecting content outside the media elements by having the media element's
layer act as a stacking context.
Test: media/controls-layers.html
- rendering/RenderLayer.cpp:
(WebCore::RenderLayer::RenderLayer):
- rendering/RenderLayer.h:
LayoutTests:
Test that dumps compositing layers in a document with media controls.
- media/controls-layers.html: Added.
- platform/mac/media/controls-layers-expected.txt: Added.
- platform/mac-mavericks/media/controls-layers-expected.txt: Added.
- 5:34 PM Changeset in webkit [184931] by
-
- 2 edits in trunk/Source/WebKit2
Can't load local files in WKWebView from containerized app
https://bugs.webkit.org/show_bug.cgi?id=145424
rdar://problem/20831176
Reviewed by Dan Bernstein.
Allow the network process to load local files from its bundle as long as it has
the sandbox extensions to do so.
This was originally added to only allow SSO to read the Info.plist inside its own application
bundle, but being able to read the application bundle from within the network process should be OK.
- Resources/SandboxProfiles/ios/com.apple.WebKit.Networking.sb:
- 5:30 PM Changeset in webkit [184930] by
-
- 13 edits in trunk/Source
[WK2] Local storage areas should get torn down when they have no remaining references.
<https://webkit.org/b/143339>
<rdar://problem/20156436>
Reviewed by Darin Adler.
Source/WebCore:
Add StorageArea::securityOrigin() implementations.
- loader/EmptyClients.cpp:
- storage/StorageArea.h:
Source/WebKit:
Add StorageArea::securityOrigin() implementation.
- Storage/StorageAreaImpl.h:
Source/WebKit2:
Let StorageNamespaceImpl own its StorageAreaMaps weakly instead of through RefPtr.
Ownership is flipped so that StorageAreaMap refs the StorageNamespaceImpl instead.
This allows the StorageAreaMaps to get destroyed once all of its clients are gone.
Practically speaking, this means that the garbage collector now decides when local
storage databases can be closed, instead of us keeping them open for the lifetime
of the web process.
For session storage, it works a bit differently. In the web process, they get torn
down when their last client disappears, but they stay alive in the UI process.
If/when the web process asks the UI process to open session storage for an origin,
the UI process checks if one already exists, and if so, just updates the ID of the
old storage with the new one provided by the web process.
- UIProcess/Storage/StorageManager.cpp:
(WebKit::StorageManager::StorageArea::isSessionStorage):
(WebKit::StorageManager::createTransientLocalStorageMap):
(WebKit::StorageManager::createSessionStorageMap):
(WebKit::StorageManager::destroyStorageMap):
- WebProcess/Storage/StorageAreaImpl.cpp:
(WebKit::StorageAreaImpl::securityOrigin):
- WebProcess/Storage/StorageAreaImpl.h:
- WebProcess/Storage/StorageAreaMap.cpp:
(WebKit::StorageAreaMap::StorageAreaMap):
(WebKit::StorageAreaMap::~StorageAreaMap):
- WebProcess/Storage/StorageAreaMap.h:
(WebKit::StorageAreaMap::securityOrigin):
- WebProcess/Storage/StorageNamespaceImpl.cpp:
(WebKit::StorageNamespaceImpl::didDestroyStorageAreaMap):
(WebKit::StorageNamespaceImpl::storageArea):
- WebProcess/Storage/StorageNamespaceImpl.h:
- 5:25 PM Changeset in webkit [184929] by
-
- 3 edits2 adds in trunk/Source/WebKit2
Disable network cache for old clients
https://bugs.webkit.org/show_bug.cgi?id=145418
rdar://problem/21126587
Reviewed by Andy Estes.
Old clients might use NSURLCache API to clear the cache. New cache requires use of new APIs.
- UIProcess/Cocoa/VersionChecks.h: Added.
- UIProcess/Cocoa/VersionChecks.mm: Added.
Add linked-on-or-after check mechanism similar to WebKit1.
(WebKit::linkTimeVersion):
(WebKit::linkedOnOrAfter):
- UIProcess/Cocoa/WebProcessPoolCocoa.mm:
(WebKit::WebProcessPool::platformInitializeNetworkProcess):
Use it when deciding whether to enable the cache.
- WebKit2.xcodeproj/project.pbxproj:
- 5:16 PM Changeset in webkit [184928] by
-
- 25 edits in trunk
img.currentSrc problem in strict mode with old picturefill
https://bugs.webkit.org/show_bug.cgi?id=144095
<rdar://problem/21087013>
Reviewed by Simon Fraser.
.:
Add a PICTURE_SIZES flag.
- Source/cmake/OptionsEfl.cmake:
- Source/cmake/OptionsGTK.cmake:
- Source/cmake/OptionsMac.cmake:
- Source/cmake/OptionsWindows.cmake:
- Source/cmake/WebKitFeatures.cmake:
Source/JavaScriptCore:
Add a PICTURE_SIZES flag.
- Configurations/FeatureDefines.xcconfig:
Source/WebCore:
Add a PICTURE_SIZES flag. This allows us to easily remove the currentSrc
attribute on HTMLImageElement, which is causing some issues with
<picture> polyfills.
- Configurations/FeatureDefines.xcconfig:
- html/HTMLImageElement.cpp:
(WebCore::HTMLImageElement::setBestFitURLAndDPRFromImageCandidate): Add guard around use of m_currentSrc.
- html/HTMLImageElement.h:
- html/HTMLImageElement.idl: Guard the currentSrc attribute.
- html/parser/HTMLPreloadScanner.cpp:
(WebCore::TokenPreloadScanner::StartTagScanner::processAttributes):
(WebCore::TokenPreloadScanner::StartTagScanner::processAttribute):
Source/WebKit/mac:
Add a PICTURE_SIZES flag.
- Configurations/FeatureDefines.xcconfig:
Source/WebKit2:
Add a PICTURE_SIZES flag.
- Configurations/FeatureDefines.xcconfig:
Source/WTF:
Add a PICTURE_SIZES flag.
- wtf/FeatureDefines.h:
Tools:
Add a PICTURE_SIZES flag.
- Scripts/webkitperl/FeatureList.pm:
WebKitLibraries:
Add a PICTURE_SIZES flag.
- win/tools/vsprops/FeatureDefines.props:
- win/tools/vsprops/FeatureDefinesCairo.props:
- 4:47 PM Changeset in webkit [184927] by
-
- 2 edits in trunk/Source/JavaScriptCore
LazyNode comparison can return incorrect results when comparing an empty value
https://bugs.webkit.org/show_bug.cgi?id=145421
Reviewed by Geoffrey Garen.
When comparing a LazyNode to another, we compare the value pointers if
we have one, and otherwise compare the nodes.
We should be comparing value pointers if the other LazyNode has one as
well, otherwise we risk an incoherency when we are a empty LazyNode
being compared to a FrozenValue without node.
Note that this is not a problem in any other case because if we don't
have a FrozenValue and we are not an empty LazyNode, we are a
non-constant node, and comparing the node pointers is correct.
- dfg/DFGLazyNode.h:
(JSC::DFG::LazyNode::operator==):
- 4:43 PM Changeset in webkit [184926] by
-
- 3 edits in trunk/Source/JavaScriptCore
REGRESSION: These sorting idioms used by Peacekeeper and Browsermark are ~20X slower
https://bugs.webkit.org/show_bug.cgi?id=145412
Reviewed by Benjamin Poulain.
Cache strings when doing a string-converting sort.
This is a 21% speedup.
- builtins/Array.prototype.js:
(sort.stringComparator): Use subtraction instead of branching because
it's slightly faster.
(sort.comparatorSort):
(sort.stringSort):
(sort): Add a special case for string sorting to avoid redundant string
conversion.
- parser/Parser.cpp:
(JSC::Parser<LexerType>::createBindingPattern): Names can be empty if
they are private names.
- 4:21 PM Changeset in webkit [184925] by
-
- 2 edits in trunk/Source/WebInspectorUI
Web Inspector: REGRESSION(r179286) Editing Style Metrics Values no longer works
https://bugs.webkit.org/show_bug.cgi?id=143164
Patch by Joseph Pecoraro <Joseph Pecoraro> on 2015-05-27
Reviewed by Brian Burg.
- UserInterface/Views/BoxModelDetailsSectionRow.js:
(WebInspector.BoxModelDetailsSectionRow.prototype._updateMetrics.createElement):
(WebInspector.BoxModelDetailsSectionRow.prototype._applyUserInput.resolvedNode.toggleInlineStyleProperty):
(WebInspector.BoxModelDetailsSectionRow.prototype._applyUserInput.resolvedNode.didToggle):
(WebInspector.BoxModelDetailsSectionRow.prototype._applyUserInput.resolvedNode):
(WebInspector.BoxModelDetailsSectionRow.prototype._applyUserInput):
Make it so editing in the metrics pane just evaluates elem.style.setProperty
on the inspected page. Use "!important" to try and give the maximum priority
possible, which is a change from older behavior. Finally, refresh the sidebar
to update all values, and update the UI if bad input didn't change styles.
- 3:11 PM Changeset in webkit [184924] by
-
- 2 edits in trunk/Source/WebCore
Attempt to fix internal build following <http://trac.webkit.org/changeset/184760>
(https://bugs.webkit.org/show_bug.cgi?id=145289)
Reviewed by Jer Noble.
- platform/spi/mac/AVFoundationSPI.h:
- 2:51 PM Changeset in webkit [184923] by
-
- 2 edits in trunk/Source/WebKit2
Assertion hit in WebPage::didChangeSelection()
https://bugs.webkit.org/show_bug.cgi?id=145413
<rdar://problem/21001129>
Reviewed by Ryosuke Niwa.
We sometimes hit the "ASSERT(layoutCount == view->layoutCount())"
assertion in WebPage::didChangeSelection(). We manage to prevent
synchronous layouts in most cases when calling editorState(). However,
it seems it can still happen in some cases. Crashing in this case seems
overkill and can be annoying to other developers, especially
considering the implications are only on performance.
This patch drops the assertion.
- WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::didChangeSelection): Deleted.
- 2:41 PM Changeset in webkit [184922] by
-
- 5 edits in trunk/Source/WebCore
Refactor WebKit1 specific threading code out of WebVideoFullscreen code that is shared with WebKit2.
https://bugs.webkit.org/show_bug.cgi?id=143954
Patch by Jeremy Jones <jeremyj@apple.com> on 2015-05-27
Reviewed by Darin Adler.
WebVideoFullscreenInterfaceAVKit is used in both WebKit1 and WebKit2. In WebKit1, the model runs in the WebThread, while
the interface is on the main thread. So there is code to dispatch between these two thread when communicating between the
interface and the model. In WebKit2, this is handled automatically by the IPC mechanism. As a result, the threading code
in WebVideoFullscreenbInterfaceAVKit and in WebVideoFullscreenModelVideoElement is redundant in WebKit2 and relies on
WebThreadRun being a no-op in WebKit2.
This change clarifies this and simplifies the WebKit2 path by moving the WebKit1 specific threading code into
WebVideoFullscreenControllerContext, which is the WebKit1 specific controller.
- platform/ios/WebVideoFullscreenControllerAVKit.mm:
(WebVideoFullscreenControllerContext::WebVideoFullscreenControllerContext):
(WebVideoFullscreenControllerContext::setController):
(WebVideoFullscreenControllerContext::didSetupFullscreen):
(WebVideoFullscreenControllerContext::didExitFullscreen):
(WebVideoFullscreenControllerContext::didCleanupFullscreen):
(WebVideoFullscreenControllerContext::fullscreenMayReturnToInline):
(WebVideoFullscreenControllerContext::resetMediaState):
(WebVideoFullscreenControllerContext::setDuration):
(WebVideoFullscreenControllerContext::setCurrentTime):
(WebVideoFullscreenControllerContext::setBufferedTime):
(WebVideoFullscreenControllerContext::setRate):
(WebVideoFullscreenControllerContext::setVideoDimensions):
(WebVideoFullscreenControllerContext::setSeekableRanges):
(WebVideoFullscreenControllerContext::setCanPlayFastReverse):
(WebVideoFullscreenControllerContext::setAudioMediaSelectionOptions):
(WebVideoFullscreenControllerContext::setLegibleMediaSelectionOptions):
(WebVideoFullscreenControllerContext::setExternalPlayback):
(WebVideoFullscreenControllerContext::play):
(WebVideoFullscreenControllerContext::pause):
(WebVideoFullscreenControllerContext::togglePlayState):
(WebVideoFullscreenControllerContext::beginScrubbing):
(WebVideoFullscreenControllerContext::endScrubbing):
(WebVideoFullscreenControllerContext::seekToTime):
(WebVideoFullscreenControllerContext::fastSeek):
(WebVideoFullscreenControllerContext::beginScanningForward):
(WebVideoFullscreenControllerContext::beginScanningBackward):
(WebVideoFullscreenControllerContext::endScanning):
(WebVideoFullscreenControllerContext::requestExitFullscreen):
(WebVideoFullscreenControllerContext::setVideoLayerFrame):
(WebVideoFullscreenControllerContext::setVideoLayerGravity):
(WebVideoFullscreenControllerContext::selectAudioMediaOption):
(WebVideoFullscreenControllerContext::selectLegibleMediaOption):
(WebVideoFullscreenControllerContext::fullscreenModeChanged):
(WebVideoFullscreenControllerContext::setupFullscreen):
(WebVideoFullscreenControllerContext::exitFullscreen):
(WebVideoFullscreenControllerContext::requestHideAndExitFullscreen):
(-[WebVideoFullscreenController init]):
(-[WebVideoFullscreenController dealloc]):
(-[WebVideoFullscreenController enterFullscreen:mode:]):
(-[WebVideoFullscreenController exitFullscreen]):
(-[WebVideoFullscreenController requestHideAndExitFullscreen]):
(-[WebVideoFullscreenController didFinishFullscreen:]):
(WebVideoFullscreenControllerChangeObserver::setTarget): Deleted.
(-[WebVideoFullscreenController didSetupFullscreen]): Deleted.
(-[WebVideoFullscreenController didEnterFullscreen]): Deleted.
(-[WebVideoFullscreenController didExitFullscreen]): Deleted.
(-[WebVideoFullscreenController didCleanupFullscreen]): Deleted.
(-[WebVideoFullscreenController fullscreenMayReturnToInline]): Deleted.
- platform/ios/WebVideoFullscreenInterfaceAVKit.h:
- platform/ios/WebVideoFullscreenInterfaceAVKit.mm:
(-[WebAVVideoLayer setBounds:]):
(-[WebAVVideoLayer resolveBounds]):
(WebVideoFullscreenInterfaceAVKit::resetMediaState):
(WebVideoFullscreenInterfaceAVKit::setDuration):
(WebVideoFullscreenInterfaceAVKit::setCurrentTime):
(WebVideoFullscreenInterfaceAVKit::setBufferedTime):
(WebVideoFullscreenInterfaceAVKit::setRate):
(WebVideoFullscreenInterfaceAVKit::setVideoDimensions):
(WebVideoFullscreenInterfaceAVKit::setSeekableRanges):
(WebVideoFullscreenInterfaceAVKit::setCanPlayFastReverse):
(WebVideoFullscreenInterfaceAVKit::setAudioMediaSelectionOptions):
(WebVideoFullscreenInterfaceAVKit::setLegibleMediaSelectionOptions):
(WebVideoFullscreenInterfaceAVKit::setExternalPlayback):
(WebVideoFullscreenInterfaceAVKit::setupFullscreen):
(WebVideoFullscreenInterfaceAVKit::enterFullscreen):
(WebVideoFullscreenInterfaceAVKit::enterFullscreenStandard):
(WebVideoFullscreenInterfaceAVKit::exitFullscreen):
(WebVideoFullscreenInterfaceAVKit::exitFullscreenInternal):
(WebVideoFullscreenInterfaceAVKit::cleanupFullscreen):
(WebVideoFullscreenInterfaceAVKit::cleanupFullscreenInternal):
(WebVideoFullscreenInterfaceAVKit::requestHideAndExitFullscreen):
(WebVideoFullscreenInterfaceAVKit::preparedToReturnToInline):
(WebVideoFullscreenInterfaceAVKit::willStartOptimizedFullscreen):
(WebVideoFullscreenInterfaceAVKit::didStartOptimizedFullscreen):
(WebVideoFullscreenInterfaceAVKit::willStopOptimizedFullscreen):
(WebVideoFullscreenInterfaceAVKit::didStopOptimizedFullscreen):
(WebVideoFullscreenInterfaceAVKit::willCancelOptimizedFullscreen):
(WebVideoFullscreenInterfaceAVKit::didCancelOptimizedFullscreen):
(WebVideoFullscreenInterfaceAVKit::setVideoLayerFrame):
(WebVideoFullscreenInterfaceAVKit::setupFullscreenInternal): Deleted.
- platform/ios/WebVideoFullscreenModel.h:
- platform/ios/WebVideoFullscreenModelVideoElement.h:
- platform/ios/WebVideoFullscreenModelVideoElement.mm:
(WebVideoFullscreenModelVideoElement::setVideoFullscreenLayer):
(WebVideoFullscreenModelVideoElement::play):
(WebVideoFullscreenModelVideoElement::pause):
(WebVideoFullscreenModelVideoElement::togglePlayState):
(WebVideoFullscreenModelVideoElement::beginScrubbing):
(WebVideoFullscreenModelVideoElement::endScrubbing):
(WebVideoFullscreenModelVideoElement::seekToTime):
(WebVideoFullscreenModelVideoElement::fastSeek):
(WebVideoFullscreenModelVideoElement::beginScanningForward):
(WebVideoFullscreenModelVideoElement::beginScanningBackward):
(WebVideoFullscreenModelVideoElement::endScanning):
(WebVideoFullscreenModelVideoElement::requestExitFullscreen):
(WebVideoFullscreenModelVideoElement::fullscreenModeChanged):
(WebVideoFullscreenModelVideoElement::videoLayerFrame): Deleted.
(WebVideoFullscreenModelVideoElement::videoLayerGravity): Deleted.
- 2:32 PM Changeset in webkit [184921] by
-
- 5 edits in branches/safari-601.1.32.2-branch/Source
Versioning.
- 2:29 PM Changeset in webkit [184920] by
-
- 6 edits in trunk/Source/JavaScriptCore
JIT-generated store barrier code should assume the buffer pointer and capacity to be compile-time constants
https://bugs.webkit.org/show_bug.cgi?id=145404
Reviewed by Andreas Kling.
We never change the capacity of a write barrier buffer. We never repoint the buffer
pointer. So, the JIT shouldn't load those from memory; it should take advantage of the
fact that these are compile-time constants.
- dfg/DFGSpeculativeJIT.cpp:
(JSC::DFG::SpeculativeJIT::storeToWriteBarrierBuffer):
- ftl/FTLLowerDFGToLLVM.cpp:
(JSC::FTL::LowerDFGToLLVM::emitStoreBarrier):
- heap/WriteBarrierBuffer.h:
(JSC::WriteBarrierBuffer::currentIndexAddress):
(JSC::WriteBarrierBuffer::capacity):
(JSC::WriteBarrierBuffer::buffer):
(JSC::WriteBarrierBuffer::currentIndexOffset): Deleted.
(JSC::WriteBarrierBuffer::capacityOffset): Deleted.
(JSC::WriteBarrierBuffer::bufferOffset): Deleted.
- jit/Repatch.cpp:
(JSC::emitPutTransitionStubAndGetOldStructure):
- 2:21 PM Changeset in webkit [184919] by
-
- 1 copy in tags/Safari-601.1.32.2.9
New tag.
- 2:13 PM Changeset in webkit [184918] by
-
- 2 edits in trunk/LayoutTests
Update Mac WebKit1 TestExpectations for platform/mac/fast/ruby/ruby-expansion-cjk-2.html
- platform/mac-wk1/TestExpectations: It's an image failure, so counter-intuitively,
Failure expectation is not the right one.
- 1:42 PM Changeset in webkit [184917] by
-
- 2 edits in trunk/Source/JavaScriptCore
REGRESSION: These sorting idioms used by Peacekeeper and Browsermark are ~20X slower
https://bugs.webkit.org/show_bug.cgi?id=145412
Reviewed by Darin Adler.
Use @toString instead of the String constructor because calls to the
String constructor are never optimized. (See
https://bugs.webkit.org/show_bug.cgi?id=144458.)
This is a ~2X speedup.
- builtins/Array.prototype.js:
(sort.stringComparator):
- 1:33 PM Changeset in webkit [184916] by
-
- 2 edits in trunk/Source/WebCore
[Mac] short-circuit MIME type lookup when possible
https://bugs.webkit.org/show_bug.cgi?id=145362
Reviewed by Jer Noble.
- platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.mm:
(WebCore::isUnsupportedMIMEType): Renamed from unsupportedMIMEType. Convert type to lower case
once instead of calling equalIgnoringCase many times.
(WebCore::staticMIMETypeList): Renamed from staticMimeTypeCache. Initialize static set in a lambda,
using an array of C strings in a loop.
(WebCore::avfMIMETypes): Renamed from avfMimeTypeCache. Initialize static set in a lambda.
(WebCore::MediaPlayerPrivateAVFoundationObjC::getSupportedTypes): avfMimeTypeCache -> avfMIMETypes.
(WebCore::keySystemIsSupported): equalIgnoringCase -> equalIgnoringASCIICase.
(WebCore::MediaPlayerPrivateAVFoundationObjC::supportsType): Ditto.
(WebCore::MediaPlayerPrivateAVFoundationObjC::supportsKeySystem): unsupportedMIMEType ->
isUnsupportedMIMEType, equalIgnoringCase -> equalIgnoringASCIICase, staticMimeTypeCache ->
staticMIMETypeList, avfMimeTypeCache -> avfMIMETypes.
(WebCore::unsupportedMIMEType): Deleted.
(WebCore::staticMimeTypeCache): Deleted.
(WebCore::avfMimeTypeCache): Deleted.
- 1:14 PM Changeset in webkit [184915] by
-
- 2 edits in trunk/LayoutTests
Update Mac WebKit1 TestExpectations for platform/mac/fast/ruby/ruby-expansion-cjk-2.html
- platform/mac-wk1/TestExpectations:
- 12:17 PM Changeset in webkit [184914] by
-
- 6 edits in trunk/LayoutTests
Fix Mac Yosemite layout test failure following <http://trac.webkit.org/changeset/184899>
(https://bugs.webkit.org/show_bug.cgi?id=145336)
Mark test platform/mac/fast/ruby/ruby-expansion-cjk-2.html as failing on Yosemite.
Also remove references to fast/ruby/ruby-expansion-cjk*.html from TestExpectations files
of other ports since we consider these tests specific to Mac.
- platform/efl/TestExpectations:
- platform/gtk/TestExpectations:
- platform/ios-simulator/TestExpectations:
- platform/mac-wk1/TestExpectations:
- platform/win/TestExpectations:
- 12:00 PM Changeset in webkit [184913] by
-
- 2 edits in trunk/Source/WebKit2
Address a follow-up review comment from Darin.
- UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::removeMediaKeys):
- 11:53 AM Changeset in webkit [184912] by
-
- 9 edits1 copy in trunk/Source/WebKit2
WKWebsiteDataStore should handle media keys
https://bugs.webkit.org/show_bug.cgi?id=145394
rdar://problem/20617794.
Reviewed by Darin Adler.
- Shared/WebsiteData/WebsiteDataTypes.h:
Add WebsiteDataTypeMediaKeys enum.
- UIProcess/API/Cocoa/APIWebsiteDataStoreCocoa.mm:
(API::WebsiteDataStore::defaultDataStoreConfiguration):
Set up mediaKeysStorageDirectory.
- UIProcess/API/Cocoa/WKWebsiteDataRecord.mm:
(dataTypesToString):
- UIProcess/API/Cocoa/WKWebsiteDataRecordInternal.h:
(WebKit::toWebsiteDataTypes):
(WebKit::toWKWebsiteDataTypes):
- UIProcess/API/Cocoa/WKWebsiteDataRecordPrivate.h: Added.
Add _WKWebsiteDataTypeMediaKeys.
- UIProcess/WebProcessPool.cpp:
(WebKit::legacyWebsiteDataStoreConfiguration):
Set up mediaKeysStorageDirectory.
- UIProcess/WebsiteData/WebsiteDataStore.cpp:
(WebKit::WebsiteDataStore::WebsiteDataStore):
Set m_mediaKeysStorageDirectory.
(WebKit::WebsiteDataStore::fetchData):
Make CallbackAggregator ThreadSafeRefCounted.
Handle WebsiteDataTypeMediaKeys by calling mediaKeyOrigins on our background queue.
(WebKit::WebsiteDataStore::removeData):
Make CallbackAggregator ThreadSafeRefCounted.
Handle WebsiteDataTypeMediaKeys by calling removeMediaKeys on our background queue.
(WebKit::WebsiteDataStore::mediaKeyOrigins):
Compute the origins.
(WebKit::computeMediaKeyFile):
Add helper function that returns the media key given a containing directory.
(WebKit::WebsiteDataStore::removeMediaKeys):
Remove media key files.
- UIProcess/WebsiteData/WebsiteDataStore.h:
Add new members.
- WebKit2.xcodeproj/project.pbxproj:
Add new files.
- 11:39 AM Changeset in webkit [184911] by
-
- 3 edits2 deletes in branches/safari-601.1.32-branch
Roll out r184894. rdar://problem/21097484
- 11:37 AM Changeset in webkit [184910] by
-
- 2 edits in trunk/Source/WebCore
[Mac] short-circuit MIME type lookup when possible
https://bugs.webkit.org/show_bug.cgi?id=145362
Reviewed by Dean Jackson.
- platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.mm:
(WebCore::unsupportedMIMEType): New, reject types known to not be supported.
(WebCore::staticMimeTypeCache): Accept MIME types known to be supported.
(WebCore::avfMimeTypeCache): Renamed from mimeTypeCache.
(WebCore::MediaPlayerPrivateAVFoundationObjC::getSupportedTypes): mimeTypeCache -> staticMimeTypeCache.
(WebCore::MediaPlayerPrivateAVFoundationObjC::supportsType): Return immediately if
unsupportedMIMEType returns true, don't call AVFoundation if staticMimeTypeCache returns true.
(WebCore::MediaPlayerPrivateAVFoundationObjC::supportsKeySystem): Ditto.
(WebCore::mimeTypeCache): Deleted.
- 11:31 AM Changeset in webkit [184909] by
-
- 2 edits in trunk/Source/WebCore
[Mac] occasional crash in MediaPlayerPrivateAVFoundationObjC::didStopLoadingRequest
https://bugs.webkit.org/show_bug.cgi?id=145409
Reviewed by Dean Jackson.
- platform/graphics/avfoundation/objc/WebCoreAVFResourceLoader.mm:
(WebCore::WebCoreAVFResourceLoader::stopLoading): NULL-check m_avRequest.
- 11:15 AM Changeset in webkit [184908] by
-
- 14 edits2 adds in trunk
Backdrop filters don't animate
https://bugs.webkit.org/show_bug.cgi?id=145386
<rdar://problem/21110037>
Reviewed by Simon Fraser.
Source/WebCore:
Add support for animation of backdrop filters.
Note that, at the moment, we can only animate/transition
backdrop-filter if it is already present on the element. See
https://bugs.webkit.org/show_bug.cgi?id=145107
Test: css3/filters/backdrop/animation.html
- page/animation/AnimationBase.h: Add m_backdropFilterFunctionListsMatch and backdropFilterFunctionListsMatch.
(WebCore::AnimationBase::backdropFilterFunctionListsMatch):
- page/animation/CSSPropertyAnimation.cpp:
(WebCore::blendFunc): backdrop-filter is also supported.
(WebCore::PropertyWrapperAcceleratedBackdropFilter::PropertyWrapperAcceleratedBackdropFilter): Added. Works
similarly to the PropertyWrapperAcceleratedFilter.
(WebCore::PropertyWrapperAcceleratedBackdropFilter::animationIsAccelerated):
(WebCore::PropertyWrapperAcceleratedBackdropFilter::blend):
(WebCore::CSSPropertyAnimationWrapperMap::CSSPropertyAnimationWrapperMap): Construct the
PropertyWrapperAcceleratedBackdropFilter.
- page/animation/ImplicitAnimation.cpp:
(WebCore::ImplicitAnimation::reset): Call checkForMatchingBackdropFilterFunctionLists.
(WebCore::ImplicitAnimation::validateTransformFunctionList): Fix typo.
(WebCore::ImplicitAnimation::checkForMatchingFilterFunctionLists): Remove whitespace.
(WebCore::ImplicitAnimation::checkForMatchingBackdropFilterFunctionLists): New method that
checks if the individual filters in a filter list match.
- page/animation/ImplicitAnimation.h: Add checkForMatchingBackdropFilterFunctionLists.
- page/animation/KeyframeAnimation.cpp:
(WebCore::KeyframeAnimation::KeyframeAnimation): Call checkForMatchingBackdropFilterFunctionLists.
(WebCore::KeyframeAnimation::checkForMatchingBackdropFilterFunctionLists): Copied from
checkForMatchingFilterFunctionLists, but calls backdropFilter() instead of filter().
(WebCore::KeyframeAnimation::checkForMatchingFilterFunctionLists): This was accidentally checking
for backdrop filters as well, but it wouldn't have worked.
- page/animation/KeyframeAnimation.h: Add checkForMatchingBackdropFilterFunctionLists.
- platform/graphics/GraphicsLayer.cpp:
(WebCore::GraphicsLayer::validateFilterOperations): Modify the ASSERT to allow AnimatedPropertyWebkitBackdropFilter.
- platform/graphics/GraphicsLayerClient.h: Add AnimatedPropertyWebkitBackdropFilter to the enum.
- platform/graphics/ca/GraphicsLayerCA.cpp:
(WebCore::propertyIdToString): Support new enum.
(WebCore::GraphicsLayerCA::addAnimation): Support AnimatedPropertyWebkitBackdropFilter.
(WebCore::GraphicsLayerCA::createFilterAnimationsFromKeyframes): Ditto.
(WebCore::GraphicsLayerCA::animatedLayer): Use a switch statement now that we have more than
two options, and handle AnimatedPropertyWebkitBackdropFilter.
(WebCore::GraphicsLayerCA::updateAnimations): Deleted a blank line.
- rendering/RenderLayerBacking.cpp:
(WebCore::RenderLayerBacking::startAnimation): Support animation of backdrop-filter.
(WebCore::RenderLayerBacking::startTransition): Ditto. Copied the code from the filter transition.
(WebCore::RenderLayerBacking::graphicsLayerToCSSProperty):
(WebCore::RenderLayerBacking::cssToGraphicsLayerProperty):
- rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::requiresCompositingForAnimation):
LayoutTests:
Add a test for animation of backdrop-filter, and do some
minor cleanups in related files.
- animations/resources/animation-test-helpers.js:
(parseFilterImage): Fix a typo.
(getPropertyValue): Support webkitBackdropFilter.
(comparePropertyValue): Ditto.
- css3/filters/backdrop/animation-expected.txt: Added.
- css3/filters/backdrop/animation.html: Added.
- 10:58 AM WebInspectorCodingStyleGuide edited by
- Fix class skeleton to use ES6 class syntax; promise nit (diff)
- 10:12 AM Changeset in webkit [184907] by
-
- 4 edits in trunk/Source/WebCore
Handle case where -startOptimizedFullscreen fails.
https://bugs.webkit.org/show_bug.cgi?id=145340
Patch by Jeremy Jones <jeremyj@apple.com> on 2015-05-27
Reviewed by Eric Carlson.
Add new AVPlayerViewController delegate methods and remove old ones.
-startOptimizedFullscreen can either fail silently because it is not allowed or it will call a delegate.
We prevent silent failure by testing preconditions in enterFullscreenOptimized, and add the
delegate -playerViewControllerFailedToStartOptimizedFullscreen to handle explicitly failure.
- platform/ios/WebVideoFullscreenInterfaceAVKit.h: added/removed method declarations.
- platform/ios/WebVideoFullscreenInterfaceAVKit.mm:
(-[WebAVPlayerController playerViewControllerFailedToStartOptimizedFullscreen:withError:]): Added.
(WebVideoFullscreenInterfaceAVKit::enterFullscreenOptimized): Handle silent failure.
(WebVideoFullscreenInterfaceAVKit::didStartOptimizedFullscreen): Hide window on main thread.
(WebVideoFullscreenInterfaceAVKit::failedToStartOptimizedFullscreen): Added
(-[WebAVPlayerController playerViewControllerWillCancelOptimizedFullscreen:]): Deleted.
(-[WebAVPlayerController playerViewControllerDidCancelOptimizedFullscreen:]): Deleted.
(WebVideoFullscreenInterfaceAVKit::willCancelOptimizedFullscreen): Deleted.
(WebVideoFullscreenInterfaceAVKit::didCancelOptimizedFullscreen): Deleted.
- platform/spi/cocoa/AVKitSPI.h: Remove dead SPI. Add isOptimizedFullscreenPossible.
- 9:52 AM Changeset in webkit [184906] by
-
- 8 edits in trunk/Source/JavaScriptCore
Remove JSC_OBJC_API_AVAILABLE_MAC_OS_X_1080
https://bugs.webkit.org/show_bug.cgi?id=145403
Reviewed by Anders Carlsson.
JSC_OBJC_API_AVAILABLE_MAC_OS_X_1080 was used to enable the JavaScriptCore Objective-C API
for WebKit and Safari projects building with JavaScriptCore targeting OS X 10.8. We don’t
need it anymore.
- API/JSBase.h:
- API/JSContext.h:
- API/JSManagedValue.h:
- API/JSValue.h:
- API/JSVirtualMachine.h:
- Configurations/Base.xcconfig:
- postprocess-headers.sh:
- 9:08 AM Changeset in webkit [184905] by
-
- 3 edits1 copy4 adds in trunk/LayoutTests
[iOS] Rebaseline expected results for tests in LayoutTests/compositing
https://bugs.webkit.org/show_bug.cgi?id=145401
Reviewed by Simon Fraser.
Rebaselined iOS tests failing due to expected differences in converage rect, scrollbar and contentsScale.
- platform/ios-simulator/compositing/layer-creation/scale-rotation-animation-overlap-expected.txt: Added.
- platform/ios-simulator/compositing/masks/compositing-clip-path-change-no-repaint-expected.txt: Added.
- platform/ios-simulator/compositing/tiling/rotated-tiled-clamped-expected.txt:
- platform/ios-simulator/compositing/tiling/rotated-tiled-preserve3d-clamped-expected.txt:
- platform/ios-simulator/compositing/tiling/transform-origin-tiled-expected.txt: Copied from LayoutTests/platform/ios-simulator/compositing/tiling/rotated-tiled-clamped-expected.txt.
- platform/ios-simulator/compositing/visible-rect/backing-change-height-from-zero-expected.txt: Added.
- 8:59 AM Changeset in webkit [184904] by
-
- 3 edits2 adds in branches/safari-601.1.32-branch
Merge r184894. <rdar://problem/21097484>
- 7:32 AM WebKitGTK/2.8.x edited by
- (diff)
- 2:55 AM Changeset in webkit [184903] by
-
- 1 copy in releases/WebKitGTK/webkit-2.9.2
WebKitGTK+ 2.9.2
- 2:50 AM Changeset in webkit [184902] by
-
- 4 edits in trunk
Unreviewed. Update OptionsGTK.cmake and NEWS for 2.9.2 release.
.:
- Source/cmake/OptionsGTK.cmake: Bump version numbers.
Source/WebKit2:
- gtk/NEWS: Add release notes for 2.9.2.
- 1:37 AM Changeset in webkit [184901] by
-
- 2 edits in trunk/Source/WebKit2
[SOUP] Network Cache: Handle the case when we fail to create the IO stream
https://bugs.webkit.org/show_bug.cgi?id=145406
Reviewed by Sergio Villar Senin.
We were asserting in that case, but it can happen that we
fail to create the stream. It happened to me after r184690, that
changed the NetworkCache::Key hash. Since this was not expected to
happen, the async operation never finished and the completion
handler never called, leaving resources loading forever. We
should make sure we call the completion handler with an error
code, so that the load finishes and the entry is silently removed
from the cache.
- NetworkProcess/cache/NetworkCacheIOChannelSoup.cpp:
(WebKit::NetworkCache::IOChannel::read):
(WebKit::NetworkCache::IOChannel::readSync):
(WebKit::NetworkCache::IOChannel::write):
- 12:23 AM Changeset in webkit [184900] by
-
- 2 edits in branches/safari-601.1.32.2-branch/Source/WebKit2
Merged r184893. rdar://problem/21111232
- 12:14 AM Changeset in webkit [184899] by
-
- 9 edits6 copies10 moves5 adds1 delete in trunk
[iOS] Arabic ligatures are broken in Google Maps
https://bugs.webkit.org/show_bug.cgi?id=145336
<rdar://problem/20689607>
Reviewed by Darin Adler.
Source/WebCore:
When we fall off the end of the font fallback list, we ask the platform for which particular font we should use
to display a particular codepoint. When we do this, we supply the font which is first in the fallback list so
that the platform can try to match it as best it can. However, web fonts' FontPlatformData have a null m_font
property, which means that we were passing nullptr to this platform fallback function.
The solution is something I've wanted to do for a while, which is migrate webfonts from using
CGDataProviderCreateWithCFData() to using CTFontManagerCreateFontDescriptorFromData(). This call creates a
CoreText font for an in-memory buffer (the old call was creating a CoreGraphics front from the same thing).
This means that we can use the FontPlatformData constructor which accepts a CoreText font, thereby making web
fonts not have a null m_font property. Note that not all platforms can use this call, which means that this
new codepath is gated to only certain OSes (which is okay because only certain OSes use the
CTFontCreatePhysicalFontDescriptorForCharactersWithLanguage() call inside FontCache). Because web fonts are the
only user of the constructor which just takes a CG font, we can make the constructor private to the
FontPlatformData class (as soon as all platforms can use this new codepath).
Test: platform/ios-simulator/fast/text/arabic-with-no-supporting-webfont.html
- platform/graphics/FontPlatformData.h: Make the FontPlatformData constructor which takes a CGFontRef private.
- platform/graphics/mac/FontCustomPlatformData.cpp:
(WebCore::FontCustomPlatformData::fontPlatformData): Use the CoreText type instead of the CoreGraphics type.
(WebCore::createFontCustomPlatformData): Migrate to CTFontManagerCreateFontDescriptorFromData().
- platform/graphics/mac/FontCustomPlatformData.h:
(WebCore::FontCustomPlatformData::FontCustomPlatformData): Store a CTFontDescriptorRef instead of a CGFontRef.
LayoutTests:
This patch adds a DRT test to make sure that text gets drawn with the correct font.
It also updates expected results for 4 tests, which I have verified are still correct. However,
because the behavior of Mavericks differs from Yosemite, I copied the old expected results to
platform/mac-mavericks before updating the results in platform/mac.
This patch also moves fast/ruby/ruby-expansion-cjk*.html to platform/mac because they rely on
platform-specific screen-space quantization. This patch also skips these tests on Mavericks
because I can't add platform-specific expected results for reftests.
- platform/ios-simulator/fast/text/arabic-with-no-supporting-webfont-expected.txt:
- platform/ios-simulator/fast/text/arabic-with-no-supporting-webfont.html:
- platform/mac-mavericks/fast/css/font-face-opentype-expected.txt: Copied from platform/mac.
- platform/mac-mavericks/svg/batik/text/xmlSpace-expected.txt: Ditto.
- platform/mac-mavericks/svg/custom/svg-fonts-fallback-expected.txt: Ditto.
- platform/mac-mavericks/svg/custom/svg-fonts-without-missing-glyph-expected.txt: Ditto.
- platform/mac-mavericks/TestExpectations: Skip fast/ruyb/ruby-expansion-cjk*.html
- platform/mac/fast/css/font-face-opentype-expected.txt: Updated.
- platform/mac/svg/batik/text/xmlSpace-expected.txt: Ditto.
- platform/mac/svg/custom/svg-fonts-fallback-expected.txt: Ditto.
- platform/mac/svg/custom/svg-fonts-without-missing-glyph-expected.txt: Ditto.
- LayoutTests/platform/mac/fast/ruby/resources/green.png: Moved from fast/ruby/resources/green.png
- LayoutTests/platform/mac/fast/ruby/resources/ruby-expansion.svg: Moved from fast/ruby/resources/ruby-expansion.svg
- LayoutTests/platform/mac/fast/ruby/ruby-expansion-cjk-2-expected.html: Moved from fast/ruby/ruby-expansion-cjk-2-expected.html
- LayoutTests/platform/mac/fast/ruby/ruby-expansion-cjk-2.html: Moved from fast/ruby/ruby-expansion-cjk-2.html
- LayoutTests/platform/mac/fast/ruby/ruby-expansion-cjk-3-expected.html: Moved from fast/ruby/ruby-expansion-cjk-3-expected.html
- LayoutTests/platform/mac/fast/ruby/ruby-expansion-cjk-3.html: Moved from fast/ruby/ruby-expansion-cjk-3.html
- LayoutTests/platform/mac/fast/ruby/ruby-expansion-cjk-4-expected.html: Moved from fast/ruby/ruby-expansion-cjk-4-expected.html
- LayoutTests/platform/mac/fast/ruby/ruby-expansion-cjk-4.html: Moved from fast/ruby/ruby-expansion-cjk-4.html
- LayoutTests/platform/mac/fast/ruby/ruby-expansion-cjk-5-expected.html: Moved from fast/ruby/ruby-expansion-cjk-5-expected.html
- LayoutTests/platform/mac/fast/ruby/ruby-expansion-cjk-5.html: Moved from fast/ruby/ruby-expansion-cjk-5.html
- LayoutTests/platform/mac/fast/ruby/ruby-expansion-cjk-expected.html: Moved from fast/ruby/ruby-expansion-cjk-expected.html
- LayoutTests/platform/mac/fast/ruby/ruby-expansion-cjk.html: Moved from fast/ruby/ruby-expansion-cjk.html
