⚠ Archived content — this site is no longer maintained.   Current WebKit documentation is at docs.webkit.org.

Timeline



Apr 26, 2014:

9:09 PM Changeset in webkit [167851] by Darin Adler
  • 4 edits in trunk/Source/WebCore

Frame and page lifetime fixes in WebCore::createWindow
​https://bugs.webkit.org/show_bug.cgi?id=132089

Reviewed by Sam Weinig.

Speculative fix because I was unable to reproduce the crash that was
reported with the test case attached to this bug.

  • loader/FrameLoader.cpp:

(WebCore::createWindow): Changed code to remove the assumption that calls
out will not destroy the page or frame. Use RefPtr for the frame, and
added early exits if frame->page() becomes null at any point before we
use a page pointer.

7:48 PM Changeset in webkit [167850] by ap@apple.com
  • 4 edits
    2 adds in trunk

Local files should not be allowed to read pasteboard data during drag
​https://bugs.webkit.org/show_bug.cgi?id=131767

Reviewed by Sam Weinig.

Source/WebCore:
Test: fast/files/local-file-drag-security.html

  • page/DragController.cpp:

(WebCore::DragController::dragExited):
(WebCore::DragController::tryDHTMLDrag):
Make an old Dashboard quirk really Dashboard only.

LayoutTests:

  • fast/files/local-file-drag-security-expected.txt: Added.
  • fast/files/local-file-drag-security.html: Added.
  • platform/wk2/TestExpectations:
7:43 PM Changeset in webkit [167849] by timothy_horton@apple.com
  • 2 edits in trunk/Source/WebKit2

REGRESSION (r167775): Safari crashes in ViewSnapshotStore::pruneSnapshots after loading 20 pages
​https://bugs.webkit.org/show_bug.cgi?id=132204
<rdar://problem/16735622>

Reviewed by Dan Bernstein and Sam Weinig.

  • UIProcess/mac/ViewSnapshotStore.mm:

(WebKit::ViewSnapshotStore::pruneSnapshots):
(WebKit::ViewSnapshotStore::recordSnapshot):
Fix a bug where the count of snapshots with live images was too high
because we were failing to decrement it when replacing a snapshot of
an existing item with a fresh one.

5:03 PM Changeset in webkit [167848] by bshafiei@apple.com
  • 5 edits in tags/Safari-538.31.2/Source

Versioning.

5:01 PM Changeset in webkit [167847] by bshafiei@apple.com
  • 1 copy in tags/Safari-538.31.2

New tag.

4:29 PM Changeset in webkit [167846] by mitz@apple.com
  • 3 edits in trunk/Source/WebKit2

[Cocoa] Rename a bundle form delegate method
​https://bugs.webkit.org/show_bug.cgi?id=132221

Reviewed by Adele Peterson.

  • WebProcess/InjectedBundle/API/Cocoa/WKWebProcessPlugInFormDelegatePrivate.h: Renamed.
  • WebProcess/InjectedBundle/API/mac/WKWebProcessPlugInBrowserContextController.mm:

(-[WKWebProcessPlugInBrowserContextController _setFormDelegate:]): Check for the method
under the old (“new”) name and under the new name.

2:28 PM Changeset in webkit [167845] by Darin Adler
  • 3 edits
    4 adds in trunk

REGRESSION (r164133): Selection doesn't paint when scrolling some pages
​https://bugs.webkit.org/show_bug.cgi?id=132172

Source/WebCore:
rdar://problem/16719473

Reviewed by Brent Fulgham.

Tests: fast/dynamic/remove-invisible-node-inside-selection.html

fast/dynamic/remove-node-inside-selection.html

  • editing/FrameSelection.cpp:

(WebCore::clearRenderViewSelection): Changed to take a Node& because having
this take a Position& was unnecessary and strange, when really it just needs
to take a document as an argument.
(WebCore::DragCaretController::nodeWillBeRemoved): Updated for the above.
(WebCore::FrameSelection::respondToNodeModification): Added code to set the
m_pendingSelectionUpdate flag and call RenderView::setNeedsLayout so the
selection will be recomputed after it's temporarily cleared when one of
the selected nodes is removed.

LayoutTests:

Reviewed by Brent Fulgham.

  • fast/dynamic/remove-invisible-node-inside-selection-expected.html: Added.
  • fast/dynamic/remove-invisible-node-inside-selection.html: Added.
  • fast/dynamic/remove-node-inside-selection-expected.html: Added.
  • fast/dynamic/remove-node-inside-selection.html: Added.
12:15 PM Changeset in webkit [167844] by bshafiei@apple.com
  • 5 edits in tags/Safari-538.31.1/Source

Versioning.

12:13 PM Changeset in webkit [167843] by bshafiei@apple.com
  • 1 copy in tags/Safari-538.31.1

New tag.

Apr 25, 2014:

11:00 PM Changeset in webkit [167842] by akling@apple.com
  • 4 edits in trunk/Source/JavaScriptCore

Inline (C++) GetByVal with numeric indices more aggressively.
<​https://webkit.org/b/132218>

We were already inlining the string indexed GetByVal path pretty well,
while the path for numeric indices got neglected. No more!

~9.5% improvement on Dromaeo/dom-traverse.html on my MBP:

Before: 199.50 runs/s

After: 218.58 runs/s

Reviewed by Phil Pizlo.

  • dfg/DFGOperations.cpp:
  • runtime/JSCJSValueInlines.h:

(JSC::JSValue::get):

ALWAYS_INLINE all the things.

  • runtime/JSObject.h:

(JSC::JSObject::getPropertySlot):

Avoid fetching the Structure more than once. We have the same
optimization in the string-indexed code path.

10:53 PM Changeset in webkit [167841] by mitz@apple.com
  • 2 edits in trunk/Source/WebKit2

[Cocoa] Synthesized getter for WKNavigationAction's _userInitiated property has the wrong name
​https://bugs.webkit.org/show_bug.cgi?id=132219

Reviewed by Sam Weinig.

  • UIProcess/API/Cocoa/WKNavigationActionInternal.h:
9:41 PM Changeset in webkit [167840] by rniwa@webkit.org
  • 4 edits
    2 adds in trunk

REGRESSION (r167689): Hovering file name in a file input causes a crash
​https://bugs.webkit.org/show_bug.cgi?id=132214

Reviewed by Andreas Kling.

Source/WebCore:
The bug was caused by callDefaultEventHandlersInTheBubblingOrder unconditionally
accessing path.contextAt(0) even if the event path was empty.

Fixed the bug by exiting early when the event path is empty.

Test: fast/events/shadow-event-path-2.html

  • dom/EventDispatcher.cpp:

(WebCore::callDefaultEventHandlersInTheBubblingOrder):

LayoutTests:
Add a regression test.

  • fast/events/shadow-event-path-2-expected.txt: Added.
  • fast/events/shadow-event-path-2.html: Added.
  • fast/events/shadow-event-path.html:
8:54 PM Changeset in webkit [167839] by dino@apple.com
  • 2 edits in trunk/Source/WebKit2

[iOS] getAssistedNodeInformation crashes getting the bounding box if it doesn't have a renderer
​https://bugs.webkit.org/show_bug.cgi?id=132217
<rdar://problem/16671662>

Patch by Simon Fraser <Simon Fraser> on 2014-04-25
Reviewed by Dean Jackson.

  • WebProcess/WebPage/ios/WebPageIOS.mm:

(WebKit::WebPage::getAssistedNodeInformation): Check that m_assistedNode has
a renderer before asking for its bounding box.

8:30 PM Changeset in webkit [167838] by akling@apple.com
  • 7 edits
    1 add in trunk/PerformanceTests

Bump jQuery to same version that dromaeo.com uses.

Rubber-stamped by Ryosuke Niwa.

  • Dromaeo/resources/dromaeo/web/lib/jquery-1.10.2.min.js: Added.
  • Dromaeo/resources/dromaeo/web/tests/cssquery-jquery.html:
  • Dromaeo/resources/dromaeo/web/tests/jslib-attr-jquery.html:
  • Dromaeo/resources/dromaeo/web/tests/jslib-event-jquery.html:
  • Dromaeo/resources/dromaeo/web/tests/jslib-modify-jquery.html:
  • Dromaeo/resources/dromaeo/web/tests/jslib-style-jquery.html:
  • Dromaeo/resources/dromaeo/web/tests/jslib-traverse-jquery.html:
5:31 PM Changeset in webkit [167837] by timothy_horton@apple.com
  • 2 edits in trunk/Source/WebCore

REGRESSION (r167828): http/tests/media/hls/video-controls-live-stream.html fails

  • English.lproj/mediaControlsLocalizedStrings.js:

Add 'Live Broadcast' to the localized strings list.

5:12 PM Changeset in webkit [167836] by benjamin@webkit.org
  • 2 edits in trunk/Source/WebKit2

[iOS][WK2] Add an intermediary view to do the target transform and adjustment transform
​https://bugs.webkit.org/show_bug.cgi?id=132175

Patch by Benjamin Poulain <​bpoulain@apple.com> on 2014-04-25
Reviewed by Enrica Casucci.

Some utility views of WKContentView account for the transforms between the content view
and the utility view.

Since the dynamic resize relies on setting subLayerTransforms, it is getting in the way.

This patch adds a layer in between for the animation. That way the content view remains
unscaled.

  • UIProcess/API/Cocoa/WKWebView.mm:

(-[WKWebView _didCommitLayerTree:WebKit::]):
(-[WKWebView _dynamicViewportUpdateChangedTargetToScale:position:]):
(-[WKWebView _beginAnimatedResizeWithUpdates:]):
(-[WKWebView _endAnimatedResize]):

5:06 PM Changeset in webkit [167835] by enrica@apple.com
  • 4 edits in trunk/Source/WebKit2

REGRESSION (iOS WebKit2): Selection is not being repainted during live resize.
​https://bugs.webkit.org/show_bug.cgi?id=132216
<rdar://problem/16628819>

Reviewed by Benjamin Poulain.

Since the selection is drawn in the UIProcess on iOS,
we need to update the selection rects and force a repaint
during a live resize.

  • WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:

(WebKit::WebFrameLoaderClient::dispatchDidLayout):

  • WebProcess/WebPage/WebPage.h:
  • WebProcess/WebPage/ios/WebPageIOS.mm:

(WebKit::WebPage::updateSelectionAppearance):

4:48 PM Changeset in webkit [167834] by jcraig@apple.com
  • 2 edits in trunk/Tools

Set prepare-Changelog flag so that it no longer runs check-webkit-style by default
​https://bugs.webkit.org/show_bug.cgi?id=132209

Reviewed by Daniel Bates.

Set default to --no-style to avoid additional spew from false positives in check-webkit-style.
Run check-webkit-style before generating changelogs to avoid additional spurious errors.

  • Scripts/prepare-ChangeLog:

(main):

4:43 PM Changeset in webkit [167833] by commit-queue@webkit.org
  • 3 edits in trunk/Source/WebInspectorUI

Web Inspector: DOM Tree debugger popover obscures dashboard
​https://bugs.webkit.org/show_bug.cgi?id=132205

Patch by Jonathan Wells <​jonowells@apple.com> on 2014-04-25
Reviewed by Timothy Hatcher.

  • UserInterface/Views/DashboardContainerView.css:

(.toolbar .dashboard.visible): Change z-index to 1. Force stacking context.

  • UserInterface/Views/Popover.css:

(.popover): Add z-index: 100.

4:14 PM Changeset in webkit [167832] by oliver@apple.com
  • 3 edits
    3 adds in trunk

Need earlier cell test
​https://bugs.webkit.org/show_bug.cgi?id=132211

Reviewed by Mark Lam.

Source/JavaScriptCore:
Move cell test to before the function call repatch
location, as the repatch logic for 32bit assumes that the
caller will already have performed a cell check.

  • jit/JITCall32_64.cpp:

(JSC::JIT::compileOpCall):

LayoutTests:
Tests

  • js/regress/polymorphic-array-call-expected.txt: Added.
  • js/regress/polymorphic-array-call.html: Added.
  • js/regress/script-tests/polymorphic-array-call.js: Added.

(func.C.this.m):
(func.C):
(func):

4:08 PM Changeset in webkit [167831] by akling@apple.com
  • 2 edits in trunk/Source/JavaScriptCore

Un-fast-allocate JSGlobalObjectRareData because Windows doesn't build and I'm not in the mood.

  • runtime/JSGlobalObject.h:

(JSC::JSGlobalObject::JSGlobalObjectRareData::JSGlobalObjectRareData):
(JSC::JSGlobalObject::JSGlobalObjectRareData::~JSGlobalObjectRareData): Deleted.

4:05 PM Applications using WebKit edited by yves@mailhot.com
(diff)
4:04 PM Applications using WebKit edited by yves@mailhot.com
(diff)
4:03 PM Applications using WebKit edited by yves@mailhot.com
(diff)
3:22 PM Changeset in webkit [167830] by beidson@apple.com
  • 11 edits
    5 adds in trunk/Source

Add a selection overlay.
<rdar://problem/16727797> and ​https://bugs.webkit.org/show_bug.cgi?id=132200

Reviewed by David Hyatt.

Source/WebCore:
No new tests (WK2 feature in development).

  • WebCore.xcodeproj/project.pbxproj:

Add a SelectionRectGathterer class.
It creates a Notifier (given a RenderView) then accepts Rects from the RenderView.
When the Notifier is destroyed, the appropriate EditorClient is notified of the gathered rects.

  • editing/SelectionRectGatherer.cpp: Added.

(WebCore::SelectionRectGatherer::SelectionRectGatherer):
(WebCore::SelectionRectGatherer::addRect):
(WebCore::SelectionRectGatherer::addRects):
(WebCore::SelectionRectGatherer::Notifier::Notifier):
(WebCore::SelectionRectGatherer::Notifier::~Notifier):
(WebCore::SelectionRectGatherer::clearAndCreateNotifier):

  • editing/SelectionRectGatherer.h: Added.
  • page/EditorClient.h:

(WebCore::EditorClient::selectionRectsDidChange):

  • rendering/RenderView.cpp:

(WebCore::RenderView::RenderView):
(WebCore::RenderView::setSelection): Clear the gatherer’s rects then create a Notifier.
(WebCore::RenderView::setSubtreeSelection): Give all the selection rects to be painted to

the SelectionRectGatherer.

  • rendering/RenderView.h:

Source/WebKit2:

  • WebKit2.xcodeproj/project.pbxproj:
  • WebProcess/WebCoreSupport/WebEditorClient.cpp:

(WebKit::WebEditorClient::selectionRectsDidChange): Pass the new selection rects on

to the selection overlay controller.

  • WebProcess/WebCoreSupport/WebEditorClient.h:
  • WebProcess/WebPage/SelectionOverlayController.cpp: Added.

(WebKit::SelectionOverlayController::SelectionOverlayController):
(WebKit::SelectionOverlayController::createOverlayIfNeeded):
(WebKit::SelectionOverlayController::destroyOverlay):
(WebKit::SelectionOverlayController::pageOverlayDestroyed):
(WebKit::SelectionOverlayController::willMoveToWebPage):
(WebKit::SelectionOverlayController::didMoveToWebPage):
(WebKit::SelectionOverlayController::selectionRectsDidChange):
(WebKit::SelectionOverlayController::drawRect):
(WebKit::SelectionOverlayController::mouseEvent):

  • WebProcess/WebPage/SelectionOverlayController.h: Added.

(WebKit::SelectionOverlayController::create):

  • WebProcess/WebPage/mac/SelectionOverlayControllerMac.mm: Added.

(WebKit::SelectionOverlayController::drawRect):
(WebKit::SelectionOverlayController::mouseEvent):
(WebKit::SelectionOverlayController::handleClick):
(WebKit::SelectionOverlayController::clearHighlight):

  • WebProcess/WebPage/WebPage.cpp:

(WebKit::WebPage::selectionOverlayController):

  • WebProcess/WebPage/WebPage.h:
3:18 PM Changeset in webkit [167829] by dino@apple.com
  • 2 edits in trunk/WebKitLibraries

Update WKSI for ​https://webkit.org/b/132195
<rdar://problem/16175750>

Sorry, I forgot to update the header. This should fix
the Mac builds.

  • WebKitSystemInterface.h:
3:18 PM Changeset in webkit [167828] by jer.noble@apple.com
  • 10 edits
    6 adds in trunk

Support "Live" streams in media controls.
​https://bugs.webkit.org/show_bug.cgi?id=131390

Reviewed by Brent Fulgham.

Source/WebCore:
Test: http/tests/media/hls/video-controls-live-stream.html

Support "Live" streams by adding an isLive property to our media controls.

  • Modules/mediacontrols/mediaControlsApple.css:

(audio::-webkit-media-controls-status-display):
(video:-webkit-full-screen::-webkit-media-controls-status-display):

  • Modules/mediacontrols/mediaControlsApple.js:

(Controller): isLive defaults to false.
(Controller.prototype.setIsLive): Set the isLive property and conditionally reconfigure the controls.
(Controller.prototype.configureInlineControls): Don't add the timeline if we are live.
(Controller.prototype.configureFullScreenControls): Ditto.
(Controller.prototype.updateStatusDisplay): Added.
(Controller.prototype.handleLoadStart): Call updateStatusDisplay().
(Controller.prototype.handleError): Ditto.
(Controller.prototype.handleAbort): Ditto.
(Controller.prototype.handleSuspend): Ditto.
(Controller.prototype.handleStalled): Ditto.
(Controller.prototype.handleWaiting): Ditto.
(Controller.prototype.updateDuration): Ditto.
(Controller.prototype.updateReadyState): Ditto.

LayoutTests:

  • http/tests/media/hls/video-controls-live-stream-expected.txt: Added.
  • http/tests/media/hls/video-controls-live-stream.html: Added.
  • http/tests/media/resources/hls/test-live.php: Added.
  • http/tests/media/resources/hls/test-vod.m3u8: Added.
  • http/tests/media/resources/hls/test.ts: Added.
  • platform/efl/TestExpectations:
  • platform/gtk/TestExpectations:
  • platform/mac/fast/hidpi/video-controls-in-hidpi-expected.txt:
  • platform/mac/fast/layers/video-layer-expected.txt:
  • platform/mac/media/media-controls-clone-expected.txt:
  • platform/wincairo/TestExpectations:
3:02 PM Changeset in webkit [167827] by timothy_horton@apple.com
  • 2 edits in trunk/LayoutTests

REGRESSION (r167818): editing/inserting/typing-space-to-trigger-smart-link.html fails on WebKit1 bots
​https://bugs.webkit.org/show_bug.cgi?id=132207

  • platform/mac-wk1/TestExpectations:

Mark as failing in WebKit1.

2:47 PM Changeset in webkit [167826] by enrica@apple.com
  • 2 edits in trunk/Source/WebKit2

Remove duplicate member variable in WKAutocorrectionRects and WKAutocorrectionContext.
​https://bugs.webkit.org/show_bug.cgi?id=132206

Reviewed by Benjamin Poulain.

  • UIProcess/ios/WKContentViewInteraction.mm:
2:42 PM Changeset in webkit [167825] by dino@apple.com
  • 18 edits in trunk/Source

Allow a platform-specific size enumeration to be passed into popup-menu display
​https://bugs.webkit.org/show_bug.cgi?id=132195

Reviewed by Brent Fulgham. With some in-person review comments from Sam Weinig.

Platforms like OS X use a set of predefined sizes for built-in controls
used for <select>: normal, small and mini. Expose that information to
the PopupMenuClient via the PopupMenuStyle, allowing it to be passed
into the platform code in WebKitSystemInterface.

Source/WebCore:

  • platform/PopupMenuStyle.h: Add a menu size enum.

(WebCore::PopupMenuStyle::PopupMenuStyle):
(WebCore::PopupMenuStyle::menuSize):

  • platform/mac/WebCoreSystemInterface.h: Pass in NSControlSize as a parameter

to WKPopupMenu.

  • platform/mac/WebCoreSystemInterface.mm: Ditto.
  • rendering/RenderMenuList.cpp:

(RenderMenuList::menuStyle): Ask the RenderTheme to calculate the size
of the menu button, so that it can be added to the PopupMenuStyle.

  • rendering/RenderTheme.h: New method to retrieve the menu size.

(WebCore::RenderTheme::popupMenuSize): Convert an NSControlSize into a PopupMenuSize.

  • rendering/RenderThemeMac.h: Override the base function, and add a

controlSizeForCell helper.

  • rendering/RenderThemeMac.mm:

(WebCore::RenderThemeMac::controlSizeForCell): Used by this new
code and the old setControlSizeForCell to calculate the NSControlSize
that would be used for the button.
(WebCore::RenderThemeMac::setControlSize): Call the new helper.
(WebCore::RenderThemeMac::popupMenuSize): Return the value from the helper.

Source/WebKit/mac:

  • WebCoreSupport/PopupMenuMac.mm:

(PopupMenuMac::show): Convert the WebCore menu size type into the
WebCoreSystemInterface type.

Source/WebKit2:

  • Shared/PlatformPopupMenuData.cpp: Encode and decode the PopupMenuSize enum.

(WebKit::PlatformPopupMenuData::encode):
(WebKit::PlatformPopupMenuData::decode):

  • Shared/PlatformPopupMenuData.h: Add PopupMenuSize member variable.
  • UIProcess/mac/WebPopupMenuProxyMac.mm:

(WebKit::WebPopupMenuProxyMac::showPopupMenu): Convert the PopupMenuSize
into a WKControlSize, and pass that into WKPopupMenu.

  • WebProcess/WebCoreSupport/mac/WebPopupMenuMac.mm:

(WebKit::WebPopupMenu::setUpPlatformData): Add the PopupMenuSize to
the PlatformPopupMenuData.

2:40 PM Changeset in webkit [167824] by dino@apple.com
  • 4 edits in trunk/WebKitLibraries

Update WKSI for ​https://webkit.org/b/132195
<rdar://problem/16175750>

Reviewed by Jessie Berlin.

  • WebKitSystemInterface.h:
  • libWebKitSystemInterfaceMavericks.a:
  • libWebKitSystemInterfaceMountainLion.a:
2:32 PM Changeset in webkit [167823] by akling@apple.com
  • 2 edits in trunk/Source/JavaScriptCore

Windows build fix attempt.

  • runtime/JSGlobalObject.h:

(JSC::JSGlobalObject::JSGlobalObjectRareData::~JSGlobalObjectRareData):

2:28 PM Changeset in webkit [167822] by timothy_horton@apple.com
  • 3 edits in trunk/Source/WebKit2

REGRESSION (r167775): Safari crashes in ViewSnapshotStore::pruneSnapshots after loading 20 pages
​https://bugs.webkit.org/show_bug.cgi?id=132204
<rdar://problem/16729123>

Reviewed by Anders Carlsson.

  • UIProcess/mac/ViewSnapshotStore.h:
  • UIProcess/mac/ViewSnapshotStore.mm:

(WebKit::ViewSnapshotStore::ViewSnapshotStore):
(WebKit::ViewSnapshotStore::pruneSnapshots):
(WebKit::ViewSnapshotStore::recordSnapshot):
Keep track of the number of snapshots that actually have live images; the
size of the snapshot map no longer represents that.
Also, fix the crash by using the UUID from the current item instead of from
the (potentially null) most distant item; r167775 accidentally typo'd this.

2:17 PM Changeset in webkit [167821] by commit-queue@webkit.org
  • 6 edits in trunk

REGRESSION(r167799): ASSERTION in parseGridTemplateShorthand in fast/css-grid-layout/grid-template-shorthand-get-set.html
​https://bugs.webkit.org/show_bug.cgi?id=132194

Patch by Javier Fernandez <​jfernandez@igalia.com> on 2014-04-25
Reviewed by Martin Robinson.

Source/WebCore:
Properly resolving the grid-template shorthand for the corresponding longhand
properties.

No new tests, grid-template-shorthand-get-set.html already covers this case.

  • css/StyleProperties.cpp:

(WebCore::StyleProperties::getPropertyValue):

  • css/StylePropertyShorthand.cpp:

(WebCore::shorthandForProperty):
(WebCore::matchingShorthandsForLonghand):

  • css/StyleResolver.cpp:

(WebCore::StyleResolver::applyProperty):

LayoutTests:

  • TestExpectations: Unskip fast/css-grid-layout/grid-template-shorthand-get-set.html.
2:05 PM Changeset in webkit [167820] by hyatt@apple.com
  • 6 edits
    3 adds in trunk

Column rules not respecting scroll offsets.
​https://bugs.webkit.org/show_bug.cgi?id=109683

Reviewed by Dean Jackson.

Source/WebCore:
Added fast/multicol/scrolling-column-rules.html

  • rendering/RenderBlock.cpp:

(WebCore::RenderBlock::paintColumnRules):
Make paintColumnRules virtual so that it works with both column implementations.

(WebCore::RenderBlock::paintObject):
Changed to call paintColumnRules with the adjusted scroll offset and to do it after
bailing on the root background only check.

  • rendering/RenderBlock.h:

paintColumnRules is now virtual.

  • rendering/RenderBlockFlow.cpp:

(WebCore::RenderBlockFlow::paintColumnRules):
(WebCore::RenderBlockFlow::paintBoxDecorations): Deleted.

  • rendering/RenderBlockFlow.h:

Got rid of paintBoxDecorations override since it failed when hasBoxDecorations was false
anyway. Override paintColumnRules instead to paint at the right time.

LayoutTests:

  • fast/multicol/scrolling-column-rules.html: Added.
  • platform/mac/fast/multicol/scrolling-column-rules-expected.png: Added.
  • platform/mac/fast/multicol/scrolling-column-rules-expected.txt: Added.
1:34 PM Changeset in webkit [167819] by msaboff@apple.com
  • 4 edits in trunk/Tools

Crash in platform/mac/accessibility/table-visible-rows.html
​https://bugs.webkit.org/show_bug.cgi?id=132146

Reviewed by Mark Lam.

Changed to use a local JSValueRef array temporary instead of a
std::make_unique<JSValueRef[]> when making an array of JSValues so that the temporary
JSValues are visited during garbage collection when the stack is scanned. Otherwise,
the temporary values could be collected.

  • DumpRenderTree/AccessibilityUIElement.cpp:

(convertElementsToObjectArray):

  • WebKitTestRunner/InjectedBundle/EventSendingController.cpp:

(WTR::EventSendingController::contextClick):

  • WebKitTestRunner/InjectedBundle/mac/AccessibilityUIElementMac.mm:

(WTR::convertElementsToObjectArray):

1:30 PM Changeset in webkit [167818] by jhoneycutt@apple.com
  • 3 edits
    2 adds in trunk

Crash applying editing commands from iframe onload event

<​https://bugs.webkit.org/show_bug.cgi?id=132103>
<rdar://problem/15696351>

Source/WebCore:
This patch merges the Chromium bug workaround from
<​http://src.chromium.org/viewvc/blink?revision=162080&view=revision>,
which prevents reentrancy in CompositeEditCommand::apply().

Reviewed by Darin Adler.

Test: editing/apply-style-iframe-crash.html

  • editing/CompositeEditCommand.cpp:

(WebCore::HTMLNames::ReentrancyGuard::isRecursiveCall):
(WebCore::HTMLNames::ReentrancyGuard::Scope::Scope):
(WebCore::HTMLNames::ReentrancyGuard::Scope::~Scope):
(WebCore::CompositeEditCommand::apply):
If this is a recursive call, return early.

LayoutTests:
Reviewed by Darin Adler.

  • editing/apply-style-iframe-crash-expected.txt: Added.
  • editing/apply-style-iframe-crash.html: Added.
1:20 PM Changeset in webkit [167817] by hyatt@apple.com
  • 6 edits
    16 adds in trunk

[New Multicolumn] fast/multicol/hit-test-* layout tests all fail
​https://bugs.webkit.org/show_bug.cgi?id=132081

Reviewed by Dean Jackson.

Source/WebCore:
Added a bunch of tests in fast/multicol/newmulticol/compare-with-old-impl/hit-test-*.html

  • rendering/RenderBlock.h:

Make offsetForContents public, since I need to call it from RenderMultiColumnSet.

  • rendering/RenderMultiColumnFlowThread.cpp:

(WebCore::RenderMultiColumnFlowThread::populate):
Stop an ASSERT in the new columns code on the hit tests by making sure layout state is
disabled when moving children around.

  • rendering/RenderMultiColumnSet.cpp:

(WebCore::RenderMultiColumnSet::positionForPoint):
Refactor this function to call a helper function instead, translateRegionPointToFlowThread.

(WebCore::RenderMultiColumnSet::translateRegionPointToFlowThread):
This function is logical (unlike the mistaken physical function I first implemented in
positionForPoint).

(WebCore::RenderMultiColumnSet::updateHitTestResult):

  • rendering/RenderMultiColumnSet.h:

Overridden to fill in the correct local coordinate when the HTML document is inside a paginated
RenderView. Note that column spans don't actually work, but once we move over to a
non-column based pagination API, that will become irrelevant.

LayoutTests:

  • fast/multicol/newmulticol/compare-with-old-impl/hit-test-above-or-below-expected.txt: Added.
  • fast/multicol/newmulticol/compare-with-old-impl/hit-test-above-or-below.html: Added.
  • fast/multicol/newmulticol/compare-with-old-impl/hit-test-block-axis-flipped-expected.txt: Added.
  • fast/multicol/newmulticol/compare-with-old-impl/hit-test-block-axis-flipped.html: Added.
  • fast/multicol/newmulticol/compare-with-old-impl/hit-test-end-of-column-expected.txt: Added.
  • fast/multicol/newmulticol/compare-with-old-impl/hit-test-end-of-column-with-line-height-expected.txt: Added.
  • fast/multicol/newmulticol/compare-with-old-impl/hit-test-end-of-column-with-line-height.html: Added.
  • fast/multicol/newmulticol/compare-with-old-impl/hit-test-end-of-column.html: Added.
  • fast/multicol/newmulticol/compare-with-old-impl/hit-test-float-expected.txt: Added.
  • fast/multicol/newmulticol/compare-with-old-impl/hit-test-float.html: Added.
  • fast/multicol/newmulticol/compare-with-old-impl/hit-test-gap-between-pages-expected.txt: Added.
  • fast/multicol/newmulticol/compare-with-old-impl/hit-test-gap-between-pages-flipped-expected.txt: Added.
  • fast/multicol/newmulticol/compare-with-old-impl/hit-test-gap-between-pages-flipped.html: Added.
  • fast/multicol/newmulticol/compare-with-old-impl/hit-test-gap-between-pages.html: Added.
  • fast/multicol/newmulticol/compare-with-old-impl/hit-test-gap-block-axis-expected.txt: Added.
  • fast/multicol/newmulticol/compare-with-old-impl/hit-test-gap-block-axis.html: Added.
12:45 PM Changeset in webkit [167816] by mark.lam@apple.com
  • 6 edits in trunk/Source/JavaScriptCore

Refactor debugging code to use BreakpointActions instead of Vector<ScriptBreakpointAction>.
<​https://webkit.org/b/132201>

Reviewed by Joseph Pecoraro.

BreakpointActions is Vector<ScriptBreakpointAction>. Let's just consistently use
BreakpointActions everywhere.

  • inspector/ScriptBreakpoint.h:

(Inspector::ScriptBreakpoint::ScriptBreakpoint):

  • inspector/ScriptDebugServer.cpp:

(Inspector::ScriptDebugServer::setBreakpoint):
(Inspector::ScriptDebugServer::getActionsForBreakpoint):

  • inspector/ScriptDebugServer.h:
  • inspector/agents/InspectorDebuggerAgent.cpp:

(Inspector::InspectorDebuggerAgent::breakpointActionsFromProtocol):
(Inspector::InspectorDebuggerAgent::setBreakpointByUrl):
(Inspector::InspectorDebuggerAgent::setBreakpoint):
(Inspector::InspectorDebuggerAgent::removeBreakpoint):

  • inspector/agents/InspectorDebuggerAgent.h:
12:41 PM Changeset in webkit [167815] by fpizlo@apple.com
  • 2 edits in trunk/Source/WTF

Figure out how many cores are available for realsies.

Rubber stamped by Andreas Kling.

HW_AVAILCPU is more canonical than HW_NCPU. For example if you use hacks to tell Darwin
to make it seem like there are fewer CPUs, HW_AVAILCPU will change but HW_NCPU won't.
HW_NCPU doesn't reflect how much actual paralellism you'll get; it's merely reporting
facts about the machine you're running on. HW_AVAILCPU is almost always what WebKit
wants since it uses this information to decide how many threads to launch for various
things.

  • wtf/NumberOfCores.cpp:

(WTF::numberOfProcessorCores):

12:39 PM Changeset in webkit [167814] by fpizlo@apple.com
  • 5 edits in trunk/Source/JavaScriptCore

DFG worklist scanning should not treat the key as a separate entity
​https://bugs.webkit.org/show_bug.cgi?id=132167

Reviewed by Mark Hahnenberg.

This simplifies the interface to the GC and will enable more optimizations.

  • dfg/DFGCompilationKey.cpp:

(JSC::DFG::CompilationKey::visitChildren): Deleted.

  • dfg/DFGCompilationKey.h:
  • dfg/DFGPlan.cpp:

(JSC::DFG::Plan::visitChildren):

  • dfg/DFGWorklist.cpp:

(JSC::DFG::Worklist::visitChildren):

11:51 AM Changeset in webkit [167813] by oliver@apple.com
  • 4 edits in trunk/Source/JavaScriptCore

Remove unused parameter from codeblock linking function
​https://bugs.webkit.org/show_bug.cgi?id=132199

Reviewed by Anders Carlsson.

No change in behaviour. This is just a small change to make it
slightly easier to reason about what the offsets in UnlinkedFunctionExecutable
actually mean.

  • bytecode/UnlinkedCodeBlock.cpp:

(JSC::UnlinkedFunctionExecutable::link):

  • bytecode/UnlinkedCodeBlock.h:
  • runtime/Executable.cpp:

(JSC::ProgramExecutable::initializeGlobalProperties):

11:17 AM Changeset in webkit [167812] by timothy_horton@apple.com
  • 2 edits in trunk/LayoutTests

REGRESSION(r167799): ASSERTION in parseGridTemplateShorthand in fast/css-grid-layout/grid-template-shorthand-get-set.html
​https://bugs.webkit.org/show_bug.cgi?id=132194

Skip the test on debug testers.
Also unmark js/slow-stress/new-spread.html because it is fixed (​https://bugs.webkit.org/show_bug.cgi?id=132022)

11:15 AM Changeset in webkit [167811] by akling@apple.com
  • 31 edits in trunk/Source

Mark some things with WTF_MAKE_FAST_ALLOCATED.
<​https://webkit.org/b/132198>

Use FastMalloc for more things.

Reviewed by Anders Carlsson.

Source/JavaScriptCore:

  • builtins/BuiltinExecutables.h:
  • heap/GCThreadSharedData.h:
  • inspector/JSConsoleClient.h:
  • inspector/agents/InspectorAgent.h:
  • runtime/CodeCache.h:
  • runtime/JSGlobalObject.h:
  • runtime/Lookup.cpp:

(JSC::HashTable::createTable):
(JSC::HashTable::deleteTable):

  • runtime/WeakGCMap.h:

Source/WebCore:

  • bindings/js/ScriptController.h:
  • dom/DocumentOrderedMap.h:
  • inspector/InspectorCSSAgent.h:
  • inspector/InspectorDOMAgent.h:
  • inspector/InspectorDOMDebuggerAgent.h:
  • inspector/InspectorDOMStorageAgent.h:
  • inspector/InspectorDatabaseAgent.h:
  • inspector/InspectorLayerTreeAgent.h:
  • inspector/InspectorPageAgent.h:
  • inspector/InspectorResourceAgent.h:
  • inspector/InspectorTimelineAgent.h:
  • inspector/InspectorWorkerAgent.h:
  • inspector/PageRuntimeAgent.h:
  • loader/HistoryController.h:
  • page/DeviceClient.h:
  • page/DeviceController.h:
  • page/EventHandler.h:
  • page/Page.h:
  • page/scrolling/ScrollingStateNode.h:
  • platform/graphics/FontGenericFamilies.h:
  • platform/graphics/FontPlatformData.h:
10:58 AM Changeset in webkit [167810] by stavila@adobe.com
  • 8 edits in trunk/Source/WebCore

[CSS Regions] Rename objectShouldPaintInFlowRegion to something more clear
​https://bugs.webkit.org/show_bug.cgi?id=132050

Reviewed by Andreas Kling.

The objectShouldPaintInFlowRegion is no longer used only by the painting process
but also for hit-testing, so it was renamed to something more generic (objectShouldFragmentInFlowRegion).

No new tests required, it's just a method rename.

  • rendering/RenderBlock.cpp:

(WebCore::RenderBlock::paint):
(WebCore::RenderBlock::selectionGaps):
(WebCore::RenderBlock::nodeAtPoint):

  • rendering/RenderBox.cpp:

(WebCore::RenderBox::positionForPoint):

  • rendering/RenderFlowThread.cpp:

(WebCore::RenderFlowThread::objectShouldFragmentInFlowRegion):
(WebCore::RenderFlowThread::objectShouldPaintInFlowRegion): Deleted.

  • rendering/RenderFlowThread.h:
  • rendering/RenderLayer.cpp:

(WebCore::RenderLayer::paintLayer):
(WebCore::RenderLayer::hitTestLayer):

  • rendering/RenderRegion.cpp:

(WebCore::RenderRegion::ensureOverflowForBox):

  • rendering/RenderReplaced.cpp:

(WebCore::RenderReplaced::shouldPaint):

10:53 AM Changeset in webkit [167809] by enrica@apple.com
  • 3 edits in trunk/Source/WebCore

iOS build fix after ​http://trac.webkit.org/changeset/167803
Unreviewed.

  • page/Frame.cpp:

(WebCore::Frame::checkOverflowScroll):

  • rendering/RenderBlock.cpp:

(WebCore::positionForPointRespectingEditingBoundaries):

9:21 AM Changeset in webkit [167808] by hyatt@apple.com
  • 7 edits
    5 adds in trunk

[New Multicolumn] Add support for offsetLeft and offsetTop.
​https://bugs.webkit.org/show_bug.cgi?id=132080

Reviewed by Andrei Bucur.

Source/WebCore:
Added fast/multicol/client-spanners-complex.html and offset-top-left.html

  • rendering/RenderBoxModelObject.cpp:

(WebCore::RenderBoxModelObject::adjustedPositionRelativeToOffsetParent):
Patch the offsetLeft/Top loop to handle calling into RenderMultiColumnFlowThread
in order to adjust the coordinates for the new multicolumn layout.

  • rendering/RenderMultiColumnFlowThread.cpp:

(WebCore::RenderMultiColumnFlowThread::mapFromFlowToRegion):
Modified to use a helper that can be shared by offsetLeft/Top code. This code
ran for client rects, and I'm moving/refactoring it for more sharing.

(WebCore::RenderMultiColumnFlowThread::physicalTranslationOffsetFromFlowToRegion):
(WebCore::RenderMultiColumnFlowThread::physicalTranslationFromFlowToRegion):

  • rendering/RenderMultiColumnFlowThread.h:

New functions that beef up what was in mapFromFlowToRegion and fix some bugs
with the translation code.

  • rendering/RenderMultiColumnSet.cpp:

(WebCore::RenderMultiColumnSet::columnTranslationForOffset):
Patched to factor in the flow thread portion rect of the specific multicolumn set.

  • rendering/RenderObject.h:

(WebCore::RenderObject::isRenderMultiColumnFlowThread):
Added for toRenderMultiColumnFlowThread capability.

LayoutTests:

  • fast/multicol/client-rects-spanners-complex.html: Added.
  • fast/multicol/client-rects-spanners.html:
  • fast/multicol/offset-top-left-expected.txt: Added.
  • fast/multicol/offset-top-left.html: Added.
  • platform/mac/fast/multicol/client-rects-spanners-complex-expected.png: Added.
  • platform/mac/fast/multicol/client-rects-spanners-complex-expected.txt: Added.
8:27 AM Changeset in webkit [167807] by Brent Fulgham
  • 3 edits
    2 adds in trunk

ScrollingCoordinator is unaware of topContentInset
​https://bugs.webkit.org/show_bug.cgi?id=132158
<rdar://problem/16706152>

Reviewed by Darin Adler.

Source/WebCore:
Test: platform/mac/fast/scrolling/scroll-select-bottom-test.html

The calculation of non-fast-scrollable regions does not currently take
the topContentOffset into account. Consequently, the logic that decides
whether to stay on the scrolling thread, or drop down to an individual
page element, can make the wrong choice. This is especially true for
small scrollable regions (such as <select> elements), where the
topContentInset may be quite close to the size of the scrollable

element itself.

  • page/scrolling/ScrollingCoordinator.cpp:

(WebCore::ScrollingCoordinator::computeNonFastScrollableRegion): Also
include the topContentInset value in our calculation.

LayoutTests:

  • platform/mac/fast/scrolling/scroll-select-bottom-test-expected.txt: Added.
  • platform/mac/fast/scrolling/scroll-select-bottom-test.html: Added.
8:15 AM Changeset in webkit [167806] by commit-queue@webkit.org
  • 2 edits in trunk/Source/WebCore

REGRESSION(r167799): Breaks debug build
​https://bugs.webkit.org/show_bug.cgi?id=132194

Patch by Javier Fernandez <​jfernandez@igalia.com> on 2014-04-25
Reviewed by Andrei Bucur.

Fix the debug bots after r167799

No new tests, no new functionality.

  • css/CSSParser.cpp:

(WebCore::CSSParser::parseGridTemplateShorthand):

7:28 AM Changeset in webkit [167805] by commit-queue@webkit.org
  • 3 edits
    2 adds in trunk

Web process is crashed during dispatching touchEvent created by JS.
​https://bugs.webkit.org/show_bug.cgi?id=113225

Patch by Miyoung Shin <​myid.shin@samsung.com> on 2014-04-25
Reviewed by Benjamin Poulain.

TouchEvent created by JS should have the necessary attributes
of touches, targetTouches and changedTouches.
It should be verified weather there are touchLists before dispatching touch event.

Source/WebCore:
Test: fast/events/touch/create-touch-event-without-touchList.html

  • dom/EventDispatcher.cpp:

(WebCore::EventDispatcher::dispatchEvent):
(WebCore::EventPath::updateTouchLists):
(WebCore::addRelatedNodeResolversForTouchList): Deleted.

LayoutTests:

  • fast/events/touch/create-touch-event-without-touchList-expected.txt: Added.
  • fast/events/touch/create-touch-event-without-touchList.html: Added.
7:17 AM Changeset in webkit [167804] by commit-queue@webkit.org
  • 4 edits in trunk/Source/WebCore

[GTK] File webkitRelativePath attribute was removed in r163483
​https://bugs.webkit.org/show_bug.cgi?id=132193

Patch by Philippe Normand <​pnormand@igalia.com> on 2014-04-25
Reviewed by Carlos Garcia Campos.

Add the removed getter as deprecated API to keep backwards compatibility.

  • bindings/gobject/WebKitDOMDeprecated.cpp:

(webkit_dom_file_get_webkit_relative_path):

  • bindings/gobject/WebKitDOMDeprecated.h:
  • bindings/gobject/WebKitDOMDeprecated.symbols:
6:56 AM Changeset in webkit [167803] by stavila@adobe.com
  • 34 edits
    6 adds in trunk

[CSS Regions] Overflow selection doesn't work properly
​https://bugs.webkit.org/show_bug.cgi?id=130715

Reviewed by David Hyatt.

Source/WebCore:
When hit-testing, painting block selection gaps and searching for the node at a specific point inside a flow thread,
the region range of the box being checked must be validated in order to not return false positives. Otherwise, hit-testing
at the top of region B could hit elements that overflow the bottom of region A.

Tests: fast/regions/selection-in-overflow-hit-testing.html

fast/regions/selection-in-overflow.html
fast/regions/selection-in-text-after-overflow-hit-testing.html

  • accessibility/AccessibilityRenderObject.cpp:

(WebCore::AccessibilityRenderObject::visiblePositionRangeForLine):
(WebCore::AccessibilityRenderObject::visiblePositionForPoint):

  • dom/Document.cpp:

(WebCore::Document::caretRangeFromPoint):

  • editing/FrameSelection.cpp:

(WebCore::FrameSelection::contains):

  • editing/VisibleUnits.cpp:

(WebCore::previousLinePosition):
(WebCore::nextLinePosition):

  • page/EventHandler.cpp:

(WebCore::EventHandler::selectClosestWordFromHitTestResult):
(WebCore::EventHandler::selectClosestWordOrLinkFromMouseEvent):
(WebCore::EventHandler::handleMousePressEventTripleClick):
(WebCore::EventHandler::handleMousePressEventSingleClick):
(WebCore::selectionExtentRespectingEditingBoundary):
(WebCore::EventHandler::updateSelectionForMouseDrag):
(WebCore::EventHandler::handleMouseReleaseEvent):

  • page/Frame.cpp:

(WebCore::Frame::visiblePositionForPoint):

  • rendering/RenderBlock.cpp:

(WebCore::RenderBlock::selectionGaps):
(WebCore::RenderBlock::nodeAtPoint):
(WebCore::positionForPointRespectingEditingBoundaries):
(WebCore::RenderBlock::positionForPointWithInlineChildren):
(WebCore::isChildHitTestCandidate):
(WebCore::RenderBlock::positionForPoint):

  • rendering/RenderBlock.h:
  • rendering/RenderBlockFlow.cpp:

(WebCore::RenderBlockFlow::positionForPointWithInlineChildren):
(WebCore::RenderBlockFlow::positionForPoint):

  • rendering/RenderBlockFlow.h:
  • rendering/RenderBox.cpp:

(WebCore::RenderBox::positionForPoint):

  • rendering/RenderBox.h:
  • rendering/RenderFileUploadControl.cpp:

(WebCore::RenderFileUploadControl::positionForPoint):

  • rendering/RenderFileUploadControl.h:
  • rendering/RenderInline.cpp:

(WebCore::RenderInline::positionForPoint):

  • rendering/RenderInline.h:
  • rendering/RenderLineBreak.cpp:

(WebCore::RenderLineBreak::positionForPoint):

  • rendering/RenderLineBreak.h:
  • rendering/RenderMultiColumnSet.cpp:

(WebCore::RenderMultiColumnSet::positionForPoint):

  • rendering/RenderMultiColumnSet.h:
  • rendering/RenderObject.cpp:

(WebCore::RenderObject::positionForPoint):

  • rendering/RenderObject.h:
  • rendering/RenderRegion.cpp:

(WebCore::RenderRegion::positionForPoint):

  • rendering/RenderRegion.h:
  • rendering/RenderReplaced.cpp:

(WebCore::RenderReplaced::positionForPoint):

  • rendering/RenderReplaced.h:
  • rendering/RenderText.cpp:

(WebCore::RenderText::positionForPoint):

  • rendering/RenderText.h:
  • rendering/svg/RenderSVGInlineText.cpp:

(WebCore::RenderSVGInlineText::positionForPoint):

  • rendering/svg/RenderSVGInlineText.h:
  • rendering/svg/RenderSVGText.cpp:

(WebCore::RenderSVGText::positionForPoint):

  • rendering/svg/RenderSVGText.h:

LayoutTests:
Added tests for hit-testing and painting selection gaps in the region's overflow area.

  • fast/regions/selection-in-overflow-expected.html: Added.
  • fast/regions/selection-in-overflow-hit-testing-expected.html: Added.
  • fast/regions/selection-in-overflow-hit-testing.html: Added.
  • fast/regions/selection-in-overflow.html: Added.
  • fast/regions/selection-in-text-after-overflow-hit-testing-expected.html: Added.
  • fast/regions/selection-in-text-after-overflow-hit-testing.html: Added.
6:35 AM Changeset in webkit [167802] by commit-queue@webkit.org
  • 4 edits in trunk/Source/WebCore

[GTK] Iframe seamless support was removed in r163427
​https://bugs.webkit.org/show_bug.cgi?id=132192

Patch by Philippe Normand <​pnormand@igalia.com> on 2014-04-25
Reviewed by Carlos Garcia Campos.

Add the removed get and set methods as deprecated API to keep
backwards compatibility.

  • bindings/gobject/WebKitDOMDeprecated.cpp:

(webkit_dom_processing_instruction_set_data):
(webkit_dom_html_iframe_element_get_seamless):
(webkit_dom_html_iframe_element_set_seamless):

  • bindings/gobject/WebKitDOMDeprecated.h:
  • bindings/gobject/WebKitDOMDeprecated.symbols:
6:24 AM Changeset in webkit [167801] by commit-queue@webkit.org
  • 4 edits in trunk/Source/WebCore

[GTK] HTMLInputElement webkitdirectory property was removed in r163483
​https://bugs.webkit.org/show_bug.cgi?id=132191

Patch by Philippe Normand <​pnormand@igalia.com> on 2014-04-25
Reviewed by Carlos Garcia Campos.

Add the removed get and set methods as deprecated API to keep
backwards compatibility.

  • bindings/gobject/WebKitDOMDeprecated.cpp:

(webkit_dom_html_input_element_get_webkitdirectory):
(webkit_dom_html_input_element_set_webkitdirectory):

  • bindings/gobject/WebKitDOMDeprecated.h:
  • bindings/gobject/WebKitDOMDeprecated.symbols:
5:38 AM Changeset in webkit [167800] by commit-queue@webkit.org
  • 2 edits in trunk/LayoutTests

Unreviewed GTK gardening.

Added test expectations for the following flaky tests:

Patch by Enrique Ocaña González <​eocanha@igalia.com> on 2014-04-25

  • platform/gtk/TestExpectations:

fast/loader/submit-form-while-parsing-2.html
css3/calc/transitions-dependent.html
fast/dom/adopt-node-crash-2.html
fast/multicol/span/span-as-immediate-columns-child-dynamic.html
platform/gtk/editing/pasteboard/middle-button-paste.html

5:06 AM Changeset in webkit [167799] by commit-queue@webkit.org
  • 9 edits
    3 adds in trunk

[CSS Grid Layout] Implementation of the grid-template shorthand.
​https://bugs.webkit.org/show_bug.cgi?id=128980

Patch by Javier Fernandez <​jfernandez@igalia.com> on 2014-04-25
Reviewed by Darin Adler.

Source/WebCore:
This shorthand sets the values for the grid-template-columns,
grid-template-rows and grid-template-areas, so the implementation
tries to reuse as much available parsing functions as possible.

The "parsingGridTrackList" was refactored to return a CSSValue and
let the "parseValue" function to assign the property value. The
"forwardSlash" operator is now valid when the track-list clause is
part of a shorthand. The "parseValue" function checkouts that only
additional clauses are allowed when processing shorthands; the
grid-columns-rows-get-set.html tests was modified to verify this.

The "parseGridTemplateAreas" was refactored too, in order to
process single areas's rows. This is very useful for the
gris-template secondary syntax, which mixes areas and rows values.

Finally, the "parseGirdLineNames" function was modified as well by
defining an new argument to concatenate head/tail custom-ident
elements and ensure the identList is at the heading index, since
it's now possible the parseList was rewound.

The implementation of the grid-template shorthand tries first to
match the <grid-template-columns> / <grid-template-rows> syntax,
failing back to the secondary syntax if needed. This approach
requires to rewind the parseList but it produces a clearer code.

Test: fast/css-grid-layout/grid-template-shorthand-get-set.html

  • css/CSSComputedStyleDeclaration.cpp:

(WebCore::ComputedStyleExtractor::propertyValue):

  • css/CSSParser.cpp:

(WebCore::CSSParser::parseValue):
(WebCore::CSSParser::parseGridTemplateRowsAndAreas):
(WebCore::CSSParser::parseGridTemplateShorthand):
(WebCore::CSSParser::parseGridLineNames):
(WebCore::CSSParser::parseGridTrackList):
(WebCore::CSSParser::parseGridTemplateAreasRow):
(WebCore::CSSParser::parseGridTemplateAreas):

  • css/CSSParser.h:
  • css/CSSParserValues.h:

(WebCore::CSSParserValueList::setCurrentIndex):

  • css/CSSPropertyNames.in:
  • css/StylePropertyShorthand.cpp:

(WebCore::webkitGridTemplateShorthand):

  • css/StylePropertyShorthand.h:

LayoutTests:
Layout Test for the basic functionality of the grid-template shorthand. It was
also added a new javascript file with some utility functions.

This shorthand sets the values for the grid-template-columns,
grid-template-rows and grid-template-areas, so the implementation
tries to reuse as much available parsing functions as possible.

  • fast/css-grid-layout/grid-template-shorthand-get-set-expected.txt: Added.
  • fast/css-grid-layout/grid-template-shorthand-get-set.html: Added.
  • fast/css-grid-layout/resources/grid-template-shorthand-parsing-utils.js: Added.

(testGridDefinitionsValues):
(testGridDefinitionsSetJSValues):
(testNonGridDefinitionsSetJSValues):
(checkGridDefinitionsSetJSValues):
(testGridDefinitionsSetBadJSValues):

4:27 AM WebKitGTK/KeepingTheTreeGreen edited by eocanha@igalia.com
(diff)
2:23 AM Changeset in webkit [167798] by akling@apple.com
  • 3 edits in trunk/Source/WebCore

Remove two unused SVGDocument functions.
<​https://webkit.org/b/132178>

Reviewed by Antti Koivisto.

  • svg/SVGDocument.cpp:

(WebCore::SVGDocument::dispatchZoomEvent): Deleted.
(WebCore::SVGDocument::dispatchScrollEvent): Deleted.

  • svg/SVGDocument.h:
1:25 AM Changeset in webkit [167797] by graouts@webkit.org
  • 6 edits
    6 adds in trunk

Implement Array.prototype.find()
​https://bugs.webkit.org/show_bug.cgi?id=130966

Reviewed by Oliver Hunt.

Source/JavaScriptCore:
Implement Array.prototype.find() and Array.prototype.findIndex() as proposed in the Harmony spec.

  • builtins/Array.prototype.js:

(find):
(findIndex):

  • runtime/ArrayPrototype.cpp:

LayoutTests:

  • js/Object-getOwnPropertyNames-expected.txt:
  • js/array-find-expected.txt: Added.
  • js/array-find.html: Added.
  • js/array-findIndex-expected.txt: Added.
  • js/array-findIndex.html: Added.
  • js/script-tests/Object-getOwnPropertyNames.js:
  • js/script-tests/array-find.js: Added.
  • js/script-tests/array-findIndex.js: Added.
1:13 AM WebKitGTK/2.4.x edited by Carlos Garcia Campos
(diff)
1:04 AM Changeset in webkit [167796] by commit-queue@webkit.org
  • 3 edits
    2 adds in trunk

Incomplete body painting when using blend modes
​https://bugs.webkit.org/show_bug.cgi?id=131889

Source/WebCore:
The incomplete painting was caused by the transparency layer created for
the root renderer. We can safely skip creating this transparency layer at
the root level, as there is nothing else being painted behind this layer that
could be used erroneously as a backdrop.

Patch by Ion Rosca <Ion Rosca> on 2014-04-25
Reviewed by Simon Fraser.

Test: css3/compositing/blend-mode-with-body.html

  • rendering/RenderLayer.h:

Changing RenderLayer::paintsWithTransparency so that it will not
return true when the root renderer needs to isolate blending.

LayoutTests:
Patch by Ion Rosca <Ion Rosca> on 2014-04-25
Reviewed by Simon Fraser.

  • css3/compositing/blend-mode-with-body-expected.html: Added.
  • css3/compositing/blend-mode-with-body.html: Added.
12:58 AM Changeset in webkit [167795] by Carlos Garcia Campos
  • 2 edits in trunk/Source/WebKit2

[GTK] Plugin process crashes with GTK2 windowed plugins
​https://bugs.webkit.org/show_bug.cgi?id=132127

Reviewed by Martin Robinson.

It happens sometimes because the socket is used before the plug
has been added. A runtime critical warnings is shown and it
sometimes ends up crashing.

  • WebProcess/Plugins/Netscape/x11/NetscapePluginX11.cpp:

(WebKit::NetscapePlugin::platformPostInitializeWindowed): Do not
show the plug widget until the socket is connected.

12:55 AM Changeset in webkit [167794] by Darin Adler
  • 29 edits in trunk/Source/WebCore
ASSERTION FAILED: "!m_isolatedWorld->isNormal()
m_wrapper !m_jsFunction" in svg/custom/use-instanceRoot-event-listeners.xhtml

​https://bugs.webkit.org/show_bug.cgi?id=132148

Reviewed by Andreas Kling.

Changed how JSCustomMarkFunction generation works. Instead of leaving out
the generated visitChildren function, just generate a call to visitAdditionalChildren.
This eliminates the need to repeat boilerplate.

The fix for the above bug was to correct mistaken logic where JSSVGElementInstance
had a visitChildren that did not properly mark event listeners because it explicitly
did not call through to the base class visitChildren. The new arrangement makes that
mistake impossible.

  • bindings/js/JSAttrCustom.cpp:

(WebCore::JSAttr::visitAdditionalChildren): Use this instead of visitChildren.

  • bindings/js/JSAudioTrackCustom.cpp:

(WebCore::JSAudioTrack::visitAdditionalChildren): Ditto.

  • bindings/js/JSAudioTrackListCustom.cpp:

(WebCore::JSAudioTrackList::visitAdditionalChildren): Ditto.

  • bindings/js/JSCSSRuleCustom.cpp:

(WebCore::JSCSSRule::visitAdditionalChildren): Ditto.

  • bindings/js/JSCSSStyleDeclarationCustom.cpp:

(WebCore::JSCSSStyleDeclaration::visitAdditionalChildren): Ditto.

  • bindings/js/JSCanvasRenderingContextCustom.cpp:

(WebCore::JSCanvasRenderingContext::visitAdditionalChildren): Ditto.

  • bindings/js/JSCryptoKeyPairCustom.cpp:

(WebCore::JSCryptoKeyPair::visitAdditionalChildren): Ditto.

  • bindings/js/JSDOMWindowCustom.cpp:

(WebCore::JSDOMWindow::visitAdditionalChildren): Ditto.

  • bindings/js/JSMessageChannelCustom.cpp:

(WebCore::JSMessageChannel::visitAdditionalChildren): Ditto.

  • bindings/js/JSMessagePortCustom.cpp:

(WebCore::JSMessagePort::visitAdditionalChildren): Ditto.

  • bindings/js/JSNodeCustom.cpp:

(WebCore::JSNode::visitAdditionalChildren): Ditto.

  • bindings/js/JSNodeFilterCustom.cpp:

(WebCore::JSNodeFilter::visitAdditionalChildren): Ditto.

  • bindings/js/JSNodeIteratorCustom.cpp:

(WebCore::JSNodeIterator::visitAdditionalChildren): Ditto.

  • bindings/js/JSSVGElementInstanceCustom.cpp:

(WebCore::JSSVGElementInstance::visitAdditionalChildren): Ditto.

  • bindings/js/JSSharedWorkerCustom.cpp:

(WebCore::JSSharedWorker::visitAdditionalChildren): Ditto.

  • bindings/js/JSStyleSheetCustom.cpp:

(WebCore::JSStyleSheet::visitAdditionalChildren): Ditto.

  • bindings/js/JSTextTrackCueCustom.cpp:

(WebCore::JSTextTrackCue::visitAdditionalChildren): Ditto.

  • bindings/js/JSTextTrackCustom.cpp:

(WebCore::JSTextTrack::visitAdditionalChildren): Ditto.

  • bindings/js/JSTextTrackListCustom.cpp:

(WebCore::JSTextTrackList::visitAdditionalChildren): Ditto.

  • bindings/js/JSTreeWalkerCustom.cpp:

(WebCore::JSTreeWalker::visitAdditionalChildren): Ditto.

  • bindings/js/JSVideoTrackCustom.cpp:

(WebCore::JSVideoTrack::visitAdditionalChildren): Ditto.

  • bindings/js/JSVideoTrackListCustom.cpp:

(WebCore::JSVideoTrackList::visitAdditionalChildren): Ditto.

  • bindings/js/JSWebGLRenderingContextCustom.cpp:

(WebCore::JSWebGLRenderingContext::visitAdditionalChildren): Ditto.

  • bindings/js/JSWorkerGlobalScopeCustom.cpp:

(WebCore::JSWorkerGlobalScope::visitAdditionalChildren): Ditto.

  • bindings/js/JSXMLHttpRequestCustom.cpp:

(WebCore::JSXMLHttpRequest::visitAdditionalChildren): Ditto.

  • bindings/js/JSXPathResultCustom.cpp:

(WebCore::JSXPathResult::visitAdditionalChildren): Ditto.

  • bindings/js/JSDOMGlobalObject.cpp:

(WebCore::JSDOMGlobalObject::visitChildren): Rewrote to use modern for loops.

  • bindings/scripts/CodeGeneratorJS.pm:

(GenerateHeader): Generate declaration of visitAdditionalChildren.
(GenerateImplementation): Generate call to visitAdditionalChildren.

12:19 AM Changeset in webkit [167793] by akling@apple.com
  • 6 edits in trunk/Source

[iOS WebKit2] Enable optimization to mmap downloaded resources once they become file-backed.
<​https://webkit.org/b/132171>
<rdar://problem/16720733>

Source/WebCore:
Add a missing export for the USE(CFNETWORK) + WebKit2 combo.

Reviewed by Antti Koivisto.

  • WebCore.exp.in:

Source/WebKit2:
Implement a CFNetwork-based version of NetworkResourceLoader::willCacheResponseAsync()
and activate the DiskCacheMonitor code path. This means that once resources go into
file system cache, we get notified and mmap them from disk, saving heaps of memory.

Reviewed by Antti Koivisto.

  • NetworkProcess/NetworkResourceLoader.h:
  • NetworkProcess/mac/DiskCacheMonitor.mm:
  • NetworkProcess/mac/NetworkResourceLoaderMac.mm:

(WebKit::NetworkResourceLoader::willCacheResponseAsync):

Note: See TracTimeline for information about the timeline view.