⚠ Archived content — this site is no longer maintained.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

Timeline



Aug 22, 2011:

10:45 PM Changeset in webkit [93580] by commit-queue@webkit.org
  • 5 edits in trunk

Implemented skia support for caching resizes of cropped images.
​https://bugs.webkit.org/show_bug.cgi?id=65587

Patch by John Bates <​jbates@google.com> on 2011-08-22
Reviewed by Darin Fisher.

Previously, resizes of cropped images would not be cached. This causes various websites to have janky CSS animations in software compositing mode.

  • platform/graphics/skia/ImageSkia.cpp:

(WebCore::drawResampledBitmap): Changed to use new APIs for subset caching.
(WebCore::Image::drawPattern): Added allowCaching parameter.

  • platform/graphics/skia/NativeImageSkia.cpp:

(WebCore::NativeImageSkia::NativeImageSkia):
(WebCore::NativeImageSkia::CachedImageInfo::CachedImageInfo):
(WebCore::NativeImageSkia::CachedImageInfo::isEqual):
(WebCore::NativeImageSkia::CachedImageInfo::set):
(WebCore::NativeImageSkia::hasResizedBitmap): Changed this method so that it does not modify caching data. Added a second version used for cropped image resizes.
(WebCore::NativeImageSkia::resizedBitmap): Added parameter to let caller specify whether caching is allowed.
(WebCore::NativeImageSkia::shouldCacheResampling): Added a second version used for cropped image resizes.
(WebCore::NativeImageSkia::shouldCacheResamplingInternal): Both shouldCacheResampling methods call down to this for the shared logic.

  • platform/graphics/skia/NativeImageSkia.h: Added CachedImageInfo to uniquely identify the cached or requested image resize operation.
9:09 PM Changeset in webkit [93579] by tonyg@chromium.org
  • 2 edits in trunk/Source/WebCore

[chromium] Fonts returned by FontCache::getFontDataForCharacters() are never released
​https://bugs.webkit.org/show_bug.cgi?id=66673

Reviewed by James Robinson.

This was fixed for other ports in ​http://trac.webkit.org/changeset/88260.
However, it was missed in FontCacheLinux likely due to similar param in the nested call to getCachedFontPlatformData().

No new tests because no directly observable change in behavior.

  • platform/graphics/chromium/FontCacheLinux.cpp:

(WebCore::FontCache::getFontDataForCharacters):

8:55 PM Changeset in webkit [93578] by commit-queue@webkit.org
  • 4 edits
    2 deletes in trunk

Unreviewed, rolling out r93565.
​http://trac.webkit.org/changeset/93565
​https://bugs.webkit.org/show_bug.cgi?id=66745

Breaks layout tests (Requested by koz2 on #webkit).

Patch by Sheriff Bot <​webkit.review.bot@gmail.com> on 2011-08-22

Source/WebCore:

  • accessibility/AccessibilityRenderObject.cpp:

(WebCore::AccessibilityRenderObject::ariaLabeledByElements):
(WebCore::AccessibilityRenderObject::ariaLabeledByAttribute):
(WebCore::AccessibilityRenderObject::title):
(WebCore::AccessibilityRenderObject::hasTextAlternative):

  • accessibility/AccessibilityRenderObject.h:

LayoutTests:

  • accessibility/aria-labelledby-overrides-aria-labeledby-expected.txt: Removed.
  • accessibility/aria-labelledby-overrides-aria-labeledby.html: Removed.
7:53 PM Changeset in webkit [93577] by commit-queue@webkit.org
  • 4 edits in trunk/Source/WebKit/chromium

Unreviewed, rolling out r93575.
​http://trac.webkit.org/changeset/93575
​https://bugs.webkit.org/show_bug.cgi?id=66744

Does not compile on chromium canaries (Requested by jamesr on
#webkit).

Patch by Sheriff Bot <​webkit.review.bot@gmail.com> on 2011-08-22

  • public/WebSettings.h:
  • src/WebSettingsImpl.cpp:

(WebKit::WebSettingsImpl::setStandardFontFamily):
(WebKit::WebSettingsImpl::setFixedFontFamily):
(WebKit::WebSettingsImpl::setSerifFontFamily):
(WebKit::WebSettingsImpl::setSansSerifFontFamily):
(WebKit::WebSettingsImpl::setCursiveFontFamily):
(WebKit::WebSettingsImpl::setFantasyFontFamily):

  • src/WebSettingsImpl.h:
7:53 PM Changeset in webkit [93576] by jamesr@google.com
  • 2 edits in branches/chromium/835/Source/WebCore

Chromium crash fix, just for this branch.

BUG=93824

  • platform/graphics/chromium/LayerRendererChromium.cpp:

(WebCore::LayerRendererChromium::paintLayerContents):

7:15 PM Changeset in webkit [93575] by commit-queue@webkit.org
  • 4 edits in trunk/Source/WebKit/chromium

[chromium] Update WebSettings to support per-script font settings
​https://bugs.webkit.org/show_bug.cgi?id=66220

Patch by Matt Falkenhagen <​falken@chromium.org> on 2011-08-22
Reviewed by Tony Chang.

  • public/WebSettings.h:
  • src/WebSettingsImpl.cpp:

(WebKit::WebSettingsImpl::setStandardFontFamily):
(WebKit::WebSettingsImpl::setFixedFontFamily):
(WebKit::WebSettingsImpl::setSerifFontFamily):
(WebKit::WebSettingsImpl::setSansSerifFontFamily):
(WebKit::WebSettingsImpl::setCursiveFontFamily):
(WebKit::WebSettingsImpl::setFantasyFontFamily):

  • src/WebSettingsImpl.h:
7:07 PM Changeset in webkit [93574] by commit-queue@webkit.org
  • 2 edits in trunk/Source/WebCore

Range::setEnd(const Position&, ExceptionCode) calls setStart() instead of setEnd()
​https://bugs.webkit.org/show_bug.cgi?id=66638

Patch by Alice Boxhall <​aboxhall@chromium.org> on 2011-08-22
Reviewed by Ryosuke Niwa.

Relevant layout test coming in a future patch (see ​https://bugs.webkit.org/show_bug.cgi?id=65900)

  • dom/Range.cpp:

(WebCore::Range::setEnd):

7:05 PM Changeset in webkit [93573] by abarth@webkit.org
  • 19 edits in trunk/Source/WebCore

Remove PLATFORM(ANDROID)
​https://bugs.webkit.org/show_bug.cgi?id=66741

Reviewed by Darin Adler.

PLATFORM(CHROMIUM) && OS(ANDROID) is the new hotness. These special
cases can be removed now that Android shares more code with Chromium.

  • bindings/js/JSDOMWindowBase.cpp:

(WebCore::JSDOMWindowBase::supportsRichSourceInfo):

  • bindings/v8/ScriptCachedFrameData.cpp:
  • bindings/v8/ScriptCachedFrameData.h:
  • bindings/v8/V8GCController.cpp:
  • bridge/jni/jsc/JavaInstanceJSC.cpp:

(JavaInstance::invokeMethod):

  • history/HistoryItem.cpp:

(WebCore::HistoryItem::~HistoryItem):
(WebCore::HistoryItem::setDocumentState):
(WebCore::HistoryItem::clearDocumentState):
(WebCore::HistoryItem::setIsTargetItem):
(WebCore::HistoryItem::addChildItem):
(WebCore::HistoryItem::setFormInfoFromRequest):

  • history/HistoryItem.h:
  • loader/cache/CachedFont.cpp:
  • platform/DragImage.h:
  • platform/PlatformTouchEvent.h:
  • platform/PlatformTouchPoint.h:

(WebCore::PlatformTouchPoint::PlatformTouchPoint):

  • platform/Widget.h:
  • platform/network/NetworkStateNotifier.cpp:
  • platform/network/NetworkStateNotifier.h:
  • platform/network/NetworkingContext.h:
  • platform/network/ResourceHandle.h:
  • platform/network/ResourceHandleInternal.h:
  • platform/network/ResourceRequestBase.cpp:
7:02 PM Changeset in webkit [93572] by commit-queue@webkit.org
  • 4 edits in trunk/Source/WebKit/efl

[EFL] Implement FrameLoaderClient::dispatchDidClearWindowObjectInWorld
​https://bugs.webkit.org/show_bug.cgi?id=66685

Patch by Raphael Kubo da Costa <​kubo@profusion.mobi> on 2011-08-22
Reviewed by Kent Tamura.

Implement dispatchDidClearWindowObjectInWorld, as it is needed by
DumpRenderTree to, among other things, properly install the event
sender in the window object.

This implementation works by emitting a "window,object,cleared" signal
with a private struct containing the required information (context,
window object and frame).

  • WebCoreSupport/FrameLoaderClientEfl.cpp:

(WebCore::FrameLoaderClientEfl::dispatchDidClearWindowObjectInWorld):

  • WebCoreSupport/FrameLoaderClientEfl.h:
  • ewk/ewk_private.h:
6:35 PM Changeset in webkit [93571] by enne@google.com
  • 4 edits in trunk/Source/WebCore

[chromium] Remove compositing assertions about creating layers when painting
​https://bugs.webkit.org/show_bug.cgi?id=66731

Reviewed by James Robinson.

The trigger for this assertion would cause a real rendering issue, but
it is both intermittent and in the wrong place in the codebase to
catch the source of the problem.

  • platform/graphics/chromium/LayerChromium.cpp:

(WebCore::LayerChromium::LayerChromium):
(WebCore::LayerChromium::~LayerChromium):

  • platform/graphics/chromium/LayerRendererChromium.cpp:

(WebCore::LayerRendererChromium::updateLayers):

  • platform/graphics/chromium/LayerRendererChromium.h:
6:07 PM Changeset in webkit [93570] by pkasting@chromium.org
  • 2 edits in trunk/LayoutTests

[chromium] Mark some tests on Mac as flakily crashing.

Unreviewed, test expectations update.

  • platform/chromium/test_expectations.txt:
6:03 PM Changeset in webkit [93569] by kerz@chromium.org
  • 3 edits
    2 copies in branches/chromium/835

Merge 92298 - Crash in DocumentWriter::endIfNotLoadingMainResource
​https://bugs.webkit.org/show_bug.cgi?id=65581

Reviewed by Nate Chapin.

Source/WebCore:

This function is poorly designed because isLoadingMainResource is a
poor proxy for determing whether to flush/finish the parser. Really,
we should change how loads complete to match the model in HTML5, but
that's pretty tricky. In the meantime, this null check fixes the
crash.

I'm sure there's another bug on file about this crash, but we've never
been able to reproduce it. Thanks to Berend-Jan Wever for the test
case!

Test: fast/loader/reload-zero-byte-plugin.html

  • loader/DocumentWriter.cpp:

(WebCore::DocumentWriter::endIfNotLoadingMainResource):

LayoutTests:

  • fast/loader/reload-zero-byte-plugin-expected.txt: Added.
  • fast/loader/reload-zero-byte-plugin.html: Added.

TBR=​abarth@webkit.org
Review URL: ​http://codereview.chromium.org/7685041

5:49 PM Changeset in webkit [93568] by pkasting@chromium.org
  • 2 edits in trunk/LayoutTests

[chromium] Mark some tests on Windows as flaky.

Unreviewed, test expectations update.

  • platform/chromium/test_expectations.txt:
5:43 PM Changeset in webkit [93567] by ap@apple.com
  • 4 edits in trunk/Source/WebCore

showModalDialog does not correctly return the defined returnValue in case domain relaxing is used
​https://bugs.webkit.org/show_bug.cgi?id=53191
<rdar://problem/8629478>

Reviewed by Geoff Garen.

Cannot test domain relaxing, we only have 127.0.0.1 and localhost.

  • bindings/js/JSDOMWindowCustom.cpp: (WebCore::JSDOMWindow::getOwnPropertySlot): Added a FIXME about a difference with Firefox.

(WebCore::DialogHandler::DialogHandler):
(WebCore::DialogHandler::dialogCreated):
(WebCore::DialogHandler::returnValue):
Changed to fetch returnValue from the global object that's in the frame when the dialog is
dismissed. A dialog can navigate itself, and it also creates a new JSDOMWindow on first load
if the origin doesn't match opener origin (which the case with domain relaxing).
Re-added a security check for returnValue that got lost in r73829, so that we don't send the
result across origins. This matches Firefox.

  • bindings/js/JSDOMWindowShell.cpp: (WebCore::JSDOMWindowShell::setWindow): Added an assertion against a very confusing case that should no longer occur.
  • page/Frame.cpp: (WebCore::Frame::pageDestroyed): Don't clear the window shell, it doesn't seem necessary, but prevents DialogHandler from fetching the return value. Added a keepAlive call to avoid changing life support timer behavior in this patch.
5:37 PM Changeset in webkit [93566] by commit-queue@webkit.org
  • 2 edits in trunk/Source/WebKit/chromium

[chromium] WebViewImpl should not reshape the 3D context
​https://bugs.webkit.org/show_bug.cgi?id=66704

Patch by Iain Merrick <​husky@google.com> on 2011-08-22
Reviewed by James Robinson.

  • src/WebViewImpl.cpp:

(WebKit::WebViewImpl::createLayerTreeHostContext3D):
(WebKit::WebViewImpl::graphicsContext3D):

5:24 PM Changeset in webkit [93565] by commit-queue@webkit.org
  • 4 edits
    2 adds in trunk

Patch by Alice Boxhall <​aboxhall@chromium.org> on 2011-08-22
Reviewed by Chris Fleizach.

aria-labelledby should be used in preference to aria-labeledby
​https://bugs.webkit.org/show_bug.cgi?id=62351

Source/WebCore:

Test: accessibility/aria-labelledby-overrides-aria-labeledby.html

  • accessibility/AccessibilityRenderObject.cpp:

(WebCore::AccessibilityRenderObject::ariaLabelledByElements): Renamed from ariaLabeledByElements.
(WebCore::AccessibilityRenderObject::ariaLabelledByAttribute): Renamed from ariaLabeledByAttribute,
and re-ordered to prefer the standard spelling.
(WebCore::AccessibilityRenderObject::title):
(WebCore::AccessibilityRenderObject::hasTextAlternative):

  • accessibility/AccessibilityRenderObject.h:

LayoutTests:

  • accessibility/aria-labelledby-overrides-aria-labeledby-expected.txt: Added.
  • accessibility/aria-labelledby-overrides-aria-labeledby.html: Added.
5:14 PM Changeset in webkit [93564] by nduca@chromium.org
  • 10 edits in trunk/Source

Make GraphicsContext3D::isResourceSafe a function and, on Chromium, determine its value lazily
​https://bugs.webkit.org/show_bug.cgi?id=66708

Reviewed by Kenneth Russell.

Source/WebCore:

  • platform/graphics/GraphicsContext3D.cpp:

(WebCore::GraphicsContext3D::texImage2DResourceSafe):

  • platform/graphics/GraphicsContext3D.h:
  • platform/graphics/gtk/GraphicsContext3DGtk.cpp:

(WebCore::GraphicsContext3D::GraphicsContext3D):

  • platform/graphics/mac/GraphicsContext3DMac.mm:

(WebCore::GraphicsContext3D::GraphicsContext3D):

  • platform/graphics/opengl/GraphicsContext3DOpenGL.cpp:

(WebCore::GraphicsContext3D::isResourceSafe):

  • platform/graphics/qt/GraphicsContext3DQt.cpp:

(WebCore::GraphicsContext3D::GraphicsContext3D):

Source/WebKit/chromium:

  • src/GraphicsContext3DChromium.cpp:

(WebCore::GraphicsContext3DInternal::GraphicsContext3DInternal):
(WebCore::GraphicsContext3DInternal::isResourceSafe):
(WebCore::GraphicsContext3D::create):
(WebCore::GraphicsContext3D::isResourceSafe):

  • src/GraphicsContext3DInternal.h:
5:05 PM Changeset in webkit [93563] by eric@webkit.org
  • 2 edits in trunk/Tools

Hack start-queue.sh in an attempt to resurrect the cr-jail-1 cr-mac-ews bot.
I won't have ssh access to the bot again for a couple days, so
attempting to fix by changing this script. If this doesn't
work we'll just live with the it hung for now.

Unreviewed.

  • EWSTools/start-queue.sh:
4:53 PM Changeset in webkit [93562] by Dimitri Glazkov
  • 3 edits in trunk/Tools

webkit-patch rebaseline-expectations command should not rearrange the entire test_expectations.txt file.
​https://bugs.webkit.org/show_bug.cgi?id=66727

Reviewed by Adam Barth.

  • Scripts/webkitpy/layout_tests/models/test_expectations.py: Made it possible to run TestExpectationSerializer without a TestConfigurationConverter,

in which case the serializer treats all expectations as unparsed.

  • Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Added unit tests.
4:52 PM Changeset in webkit [93561] by abarth@webkit.org
  • 3 edits
    3 adds in trunk

HTMLSourceTracker crashes when network packets break poorly
​https://bugs.webkit.org/show_bug.cgi?id=66728

Reviewed by Darin Adler.

Source/WebCore:

If there is a network packet boundary in the middle of an attribute
that begins with the letters "on", then the HTMLSourceTracker will get
confused and try to extract too many characters from future input. If
the future input is small enough, that will walk off the end of the
input and crash.

Test: http/tests/security/xssAuditor/crash-while-loading-tag-with-pause.html

  • html/parser/HTMLSourceTracker.cpp:

(WebCore::HTMLSourceTracker::sourceForToken):

LayoutTests:

Test that we don't crash when we get a bad network packet boundary.

  • http/tests/security/xssAuditor/crash-while-loading-tag-with-pause-expected.txt: Added.
  • http/tests/security/xssAuditor/crash-while-loading-tag-with-pause.html: Added.
  • http/tests/security/xssAuditor/resources/tag-with-pause.php: Added.
4:41 PM Changeset in webkit [93560] by fpizlo@apple.com
  • 7 edits
    2 adds in trunk/Source/JavaScriptCore

Sampling counter support is in the bytecode directory
​https://bugs.webkit.org/show_bug.cgi?id=66724

Reviewed by Darin Adler.

Moved SamplingCounter to a separate header in runtime/.

  • GNUmakefile.list.am:
  • JavaScriptCore.pro:
  • JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.vcproj:
  • JavaScriptCore.xcodeproj/project.pbxproj:
  • bytecode/SamplingTool.cpp:
  • bytecode/SamplingTool.h:
  • runtime/SamplingCounter.cpp: Added.

(JSC::AbstractSamplingCounter::dump):

  • runtime/SamplingCounter.h: Added.

(JSC::AbstractSamplingCounter::count):
(JSC::AbstractSamplingCounter::addressOfCounter):
(JSC::AbstractSamplingCounter::init):
(JSC::SamplingCounter::SamplingCounter):
(JSC::GlobalSamplingCounter::name):
(JSC::DeletableSamplingCounter::DeletableSamplingCounter):
(JSC::DeletableSamplingCounter::~DeletableSamplingCounter):

4:34 PM Changeset in webkit [93559] by eric@webkit.org
  • 3 edits in trunk/Source/WebCore

Attempt to clean up bidiNext usage
​https://bugs.webkit.org/show_bug.cgi?id=66721

Reviewed by Ryosuke Niwa.

bidiNext and bidiFirst are horribly confusing.
Even worse is that bidiNext takes a bunch of mutually exclusive options.
It appears that there is a "return me every inline, even if its empty"
mode which is only used for simplified inline layout in RenderBlock.cpp.
To support that mode, there is a endOfInline pointer which keeps track
of if we just returned at the end of an inline to so we don't get stuck in
and empty inline (unable to distinguish the start from the finish).

The actual bidi/line-layout code uses bidiNext/bidiFirst in a "skip empty inlines"
mode. (Since empty inlines do not participate in the Unicode Bidi Algorithm.)

This change renames bidiNext to bidiNextShared (still a horrible name) and moves
all callers to explicitly calling bidiNextSkippingEmptyInlines or bidiNextIncludingEmptyInlines.
It becomes obvious which code uses which.

In reviewing this code be aware that the previous bidiNext default was to "skip empty inlines" (skipInlines = true).
Thus any caller who didn't pass true/false should now be calling bidiNextSkippingEmptyInlines instead.

No functional change, thus no tests.

  • rendering/InlineIterator.h:

(WebCore::bidiNextShared):
(WebCore::bidiNextSkippingEmptyInlines):
(WebCore::bidiNextIncludingEmptyInlines):
(WebCore::bidiFirstSkippingEmptyInlines):
(WebCore::bidiFirstIncludingEmptyInlines):
(WebCore::InlineWalker::InlineWalker):
(WebCore::InlineWalker::advance):
(WebCore::InlineIterator::increment):
(WebCore::InlineBidiResolver::appendRun):

  • rendering/RenderBlockLineLayout.cpp:

(WebCore::RenderBlock::determineStartPosition):
(WebCore::shouldSkipWhitespaceAfterStartObject):
(WebCore::RenderBlock::LineBreaker::nextLineBreak):

4:30 PM Changeset in webkit [93558] by levin@chromium.org
  • 6 edits in trunk/Source/WebCore

Fix usage of PassRefPtr in postMessage and structured clone code.
​https://bugs.webkit.org/show_bug.cgi?id=66725

Reviewed by Darin Adler.

No new functionality exposed so no new tests.

  • bindings/js/JSDOMWindowCustom.cpp:

(WebCore::JSDOMWindow::postMessage): Changed PassRefPtr to RefPtr and added release when handed off.

  • bindings/js/JSMessageEventCustom.cpp:

(WebCore::JSMessageEvent::initMessageEvent): Ditto.

  • bindings/js/JSMessagePortCustom.h:

(WebCore::handlePostMessage): Ditto.

  • bindings/js/SerializedScriptValue.cpp:

(WebCore::SerializedScriptValue::create): Ditto.

  • workers/SharedWorkerContext.cpp:

(WebCore::createConnectEvent): Added release() when a RefPtr is returned.

4:24 PM Changeset in webkit [93557] by macpherson@chromium.org
  • 3 edits in trunk/Source/WebCore

Implement cast between CSSPrimitiveValue and EBackfaceVisibility, use in CSSStyleSelector::applyProperty.
​https://bugs.webkit.org/show_bug.cgi?id=66648

Reviewed by Dan Bernstein.

No new tests / no functionality changed.

  • css/CSSPrimitiveValueMappings.h:

(WebCore::CSSPrimitiveValue::CSSPrimitiveValue):
Support cast from EBackfaceVisibility.
(WebCore::CSSPrimitiveValue::operator EBackfaceVisibility):
Support cast from EBackfaceVisibility.

  • css/CSSStyleSelector.cpp:

(WebCore::CSSStyleSelector::applyProperty):
Use HANDLE_INHERIT_AND_INITIAL_AND_PRIMITIVE macro.

4:15 PM Changeset in webkit [93556] by pkasting@chromium.org
  • 11 edits
    6 deletes in trunk

Unreviewed, rolling out r93538.
​http://trac.webkit.org/changeset/93538
​https://bugs.webkit.org/show_bug.cgi?id=66143

Broke Leopard, SnowLeopard, and Chromium Mac bots

Source/WebCore:

  • rendering/LayoutState.cpp:

(WebCore::LayoutState::LayoutState):

  • rendering/RenderBlock.cpp:

(WebCore::RenderBlock::layoutBlock):
(WebCore::RenderBlock::logicalRightOffsetForLine):

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

(WebCore::RenderFlowThread::RenderFlowThread):
(WebCore::RenderFlowThread::layout):
(WebCore::RenderFlowThread::repaintRectangleInRegions):

  • rendering/RenderFlowThread.h:
  • rendering/RenderRegion.cpp:
  • rendering/RenderRegion.h:
  • rendering/RenderView.cpp:

(WebCore::RenderView::RenderView):

  • rendering/RenderView.h:

(WebCore::RenderView::pushLayoutState):

LayoutTests:

  • fast/regions/webkit-flow-floats-inside-regions-bounds-expected.txt: Removed.
  • fast/regions/webkit-flow-floats-inside-regions-bounds.html: Removed.
  • fast/regions/webkit-flow-inlines-inside-regions-bounds-expected.txt: Removed.
  • fast/regions/webkit-flow-inlines-inside-regions-bounds-vertical-lr-expected.txt: Removed.
  • fast/regions/webkit-flow-inlines-inside-regions-bounds-vertical-lr.html: Removed.
  • fast/regions/webkit-flow-inlines-inside-regions-bounds.html: Removed.
4:08 PM Changeset in webkit [93555] by oliver@apple.com
  • 7 edits in trunk/Source/WebKit2

Delay GC triggered NP object destruction to the next runloop cycle
​https://bugs.webkit.org/show_bug.cgi?id=66717

Reviewed by Anders Carlsson.

Delay destruction of plugin objects caused by GC until the next
runloop cycle so that they can execute JS in their finalizers.
We do this using a zero delay timer coupled with a queue of
objects to be finalised.

  • WebProcess/Plugins/Netscape/JSNPObject.cpp:

(WebKit::JSNPObject::releaseObject):

  • WebProcess/Plugins/Netscape/JSNPObject.h:
  • WebProcess/Plugins/Netscape/NPRuntimeObjectMap.cpp:

(WebKit::NPRuntimeObjectMap::NPRuntimeObjectMap):
(WebKit::NPRuntimeObjectMap::invalidate):
(WebKit::NPRuntimeObjectMap::invalidateQueuedObjects):
(WebKit::NPRuntimeObjectMap::addToInvalidationQueue):
(WebKit::NPRuntimeObjectMap::finalize):

  • WebProcess/Plugins/Netscape/NPRuntimeObjectMap.h:
  • WebProcess/Plugins/Netscape/NPRuntimeUtilities.cpp:

(WebKit::trySafeReleaseNPObject):

  • WebProcess/Plugins/Netscape/NPRuntimeUtilities.h:
4:08 PM Changeset in webkit [93554] by tony@chromium.org
  • 1 edit
    1 copy
    1 delete in trunk/LayoutTests

Move chromium-mac/plugins/keyboard-events-expected.txt to
chromium-cg-mac/plugins/keyboard-events-expected.txt since the results
for chromium-mac are currently in chromium-cg-mac. This result was
added in r93519 and put in the wrong place.

  • platform/chromium-cg-mac/plugins/keyboard-events-expected.txt: Copied from LayoutTests/platform/chromium-mac/plugins/keyboard-events-expected.txt.
  • platform/chromium-mac/plugins: Removed.
  • platform/chromium-mac/plugins/keyboard-events-expected.txt: Removed.
3:44 PM Changeset in webkit [93553] by adamk@chromium.org
  • 4 edits in trunk/Source/WebKit/chromium

[chromium] Remove deprecated and unused bits of WebIDB* headers
​https://bugs.webkit.org/show_bug.cgi?id=66711

Reviewed by Tony Chang.

The Chromium side of this patch was submitted as
​http://crrev.com/97687.

  • public/WebIDBCallbacks.h:
  • public/WebIDBCursor.h:

(WebKit::WebIDBCursor::primaryKey):
(WebKit::WebIDBCursor::value):
(WebKit::WebIDBCursor::deleteFunction):

  • public/WebIDBDatabase.h:

(WebKit::WebIDBDatabase::transaction):

3:35 PM Changeset in webkit [93552] by Dimitri Glazkov
  • 3 edits in trunk/Tools

Update unit test expectations after r93530.

  • Scripts/webkitpy/layout_tests/port/factory_unittest.py: Updated.
  • Scripts/webkitpy/tool/commands/rebaseline_unittest.py: Ditto.
3:20 PM Changeset in webkit [93551] by tony@chromium.org
  • 14 edits in trunk

Add CSS parsing of -webkit-flex-pack
​https://bugs.webkit.org/show_bug.cgi?id=66701

Reviewed by Eric Seidel.

Source/WebCore:

  • css/CSSComputedStyleDeclaration.cpp:

(WebCore::CSSComputedStyleDeclaration::getPropertyCSSValue):

  • css/CSSParser.cpp:

(WebCore::CSSParser::parseValue):

  • css/CSSPrimitiveValueMappings.h:

(WebCore::CSSPrimitiveValue::CSSPrimitiveValue):
(WebCore::CSSPrimitiveValue::operator EFlexPack):

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

(WebCore::CSSStyleApplyProperty::CSSStyleApplyProperty):

  • css/CSSStyleSelector.cpp:

(WebCore::CSSStyleSelector::applyProperty):

  • rendering/style/RenderStyle.h:

(WebCore::InheritedFlags::flexPack):
(WebCore::InheritedFlags::setFlexPack):
(WebCore::InheritedFlags::initialFlexPack):

  • rendering/style/RenderStyleConstants.h:
  • rendering/style/StyleFlexibleBoxData.cpp:

(WebCore::StyleFlexibleBoxData::StyleFlexibleBoxData):
(WebCore::StyleFlexibleBoxData::operator==):

  • rendering/style/StyleFlexibleBoxData.h:

LayoutTests:

  • css3/flexbox/css-properties-expected.txt:
  • css3/flexbox/script-tests/css-properties.js:
3:08 PM Changeset in webkit [93550] by andersca@apple.com
  • 4 edits in trunk/Source/WebKit2

Move code from PageClientImpl::doneWithKeyEvent to WKView
​https://bugs.webkit.org/show_bug.cgi?id=66722

Reviewed by Darin Adler.

  • UIProcess/API/mac/PageClientImpl.mm:

(WebKit::PageClientImpl::doneWithKeyEvent):

  • UIProcess/API/mac/WKView.mm:

(-[WKView _doneWithKeyEvent:WebKit::eventWasHandled:]):

  • UIProcess/API/mac/WKViewInternal.h:
3:01 PM Changeset in webkit [93549] by rniwa@webkit.org
  • 2 edits in trunk/LayoutTests

Add editing/selection/caret-at-bidi-boundary.html back to the Qt's skipped list
because r93529 didn't fix it. The failure is still tracked by the bug 66514.

  • platform/qt/Skipped:
2:56 PM Changeset in webkit [93548] by andersca@apple.com
  • 2 edits in trunk/Source/WebKit2

Update the text input context when the plug-in focus changes and not when the complex text input state changes
​https://bugs.webkit.org/show_bug.cgi?id=66716

Reviewed by Darin Adler.

  • UIProcess/API/mac/WKView.mm:

(-[WKView _pluginFocusOrWindowFocusChanged:pluginComplexTextInputIdentifier:]):
(-[WKView _setPluginComplexTextInputState:WebKit::pluginComplexTextInputIdentifier:]):

2:49 PM Changeset in webkit [93547] by tony@chromium.org
  • 6 edits
    2 adds in trunk

handle child margin values when flexing
​https://bugs.webkit.org/show_bug.cgi?id=65887

Reviewed by David Hyatt.

Source/WebCore:

Test: css3/flexbox/002.html

  • rendering/RenderBox.cpp:

(WebCore::RenderBox::computeLogicalWidth):

  • rendering/RenderFlexibleBox.cpp:

(WebCore::RenderFlexibleBox::layoutHorizontalBlock):
(WebCore::RenderFlexibleBox::computePreferredSize): If the margin is

auto, treat it as flex(1 0 0).

  • rendering/RenderObject.cpp:

(WebCore::RenderObject::isFlexibleBoxIncludingDeprecated): Added.

LayoutTests:

  • css3/flexbox/002-expected.txt: Added.
  • css3/flexbox/002.html: Added.
  • css3/flexbox/resources/flexbox.js:

(checkHorizontalBoxen):

2:46 PM Changeset in webkit [93546] by msaboff@apple.com
  • 3 edits in trunk/Source/WebKit2

REGRESSION (r92231): Apple campus proposal PDF doesn't display in Safari
​https://bugs.webkit.org/show_bug.cgi?id=66464

Changed ArgumentEncoder to use system malloc instead of fastMalloc.
FastMalloc uses madvise(MADV_FREE_REUSABLE) which is incompatible with
mach message Out Of Line (OOL) messages that use MACH_MSG_VIRTUAL_COPY.
The system malloc has no such limitation.
Changed sendOutgoingMessage to use MACH_MSG_VIRTUAL_COPY again as it
doesn't have size limitations that MACH_MSG_PHYSICAL_COPY.

Reviewed by Anders Carlsson.

  • Platform/CoreIPC/ArgumentEncoder.cpp:

(CoreIPC::ArgumentEncoder::~ArgumentEncoder):
(CoreIPC::ArgumentEncoder::grow):

  • Platform/CoreIPC/mac/ConnectionMac.cpp:

(CoreIPC::Connection::sendOutgoingMessage):

2:45 PM Changeset in webkit [93545] by Dimitri Glazkov
  • 3 edits in trunk/Tools

[Chromium] Expand CPU/GPU/Skia/CG into a matrix in graphics_type.
​https://bugs.webkit.org/show_bug.cgi?id=66705

Reviewed by Adam Barth.

  • Scripts/webkitpy/layout_tests/port/chromium.py: Added plumbing for CG graphics types.
  • Scripts/webkitpy/tool/servers/gardeningserver_unittest.py: Adjusted unit test.
2:38 PM Changeset in webkit [93544] by andersca@apple.com
  • 2 edits in trunk/Source/WebKit2

Use -[NSApplication updateWindows] to update the current input context
​https://bugs.webkit.org/show_bug.cgi?id=66715

Reviewed by Alexey Proskuryakov.

  • UIProcess/API/mac/WKView.mm:

(-[WKView _setPluginComplexTextInputState:WebKit::pluginComplexTextInputIdentifier:]):

2:35 PM Changeset in webkit [93543] by andersca@apple.com
  • 19 edits in trunk/Source/WebKit2

The UI process should keep track of whether a plug-in has focus or not
​https://bugs.webkit.org/show_bug.cgi?id=66712

Reviewed by Darin Adler.

This is in preparation for implementing the updated Cocoa text input model.

Pipe through whether a plug-in and its containing window has focus or not, from the
plug-in process and all the way to the UI process.

  • PluginProcess/PluginControllerProxy.h:
  • PluginProcess/mac/PluginControllerProxyMac.mm:

(WebKit::PluginControllerProxy::pluginFocusOrWindowFocusChanged):

  • UIProcess/API/mac/PageClientImpl.h:
  • UIProcess/API/mac/PageClientImpl.mm:

(WebKit::PageClientImpl::pluginFocusOrWindowFocusChanged):

  • UIProcess/API/mac/WKView.mm:

(-[WKView _pluginFocusOrWindowFocusChanged:pluginComplexTextInputIdentifier:]):

  • UIProcess/API/mac/WKViewInternal.h:
  • UIProcess/PageClient.h:
  • UIProcess/WebPageProxy.h:
  • UIProcess/WebPageProxy.messages.in:
  • UIProcess/mac/WebPageProxyMac.mm:

(WebKit::WebPageProxy::pluginFocusOrWindowFocusChanged):

  • WebProcess/Plugins/Netscape/NetscapePlugin.h:
  • WebProcess/Plugins/Netscape/mac/NetscapePluginMac.mm:

(WebKit::NetscapePlugin::platformSetFocus):
(WebKit::NetscapePlugin::windowFocusChanged):
(WebKit::NetscapePlugin::pluginFocusOrWindowFocusChanged):

  • WebProcess/Plugins/Netscape/mac/PluginProxyMac.mm:

(WebKit::PluginProxy::pluginFocusOrWindowFocusChanged):

  • WebProcess/Plugins/PluginController.h:
  • WebProcess/Plugins/PluginProxy.h:
  • WebProcess/Plugins/PluginProxy.messages.in:
  • WebProcess/Plugins/PluginView.cpp:

(WebKit::PluginView::pluginFocusOrWindowFocusChanged):

  • WebProcess/Plugins/PluginView.h:
2:23 PM Changeset in webkit [93542] by nduca@chromium.org
  • 2 edits in trunk/Source/WebCore

[chromium] Make compositor context current before querying for extensions
​https://bugs.webkit.org/show_bug.cgi?id=66696

Reviewed by James Robinson.

  • platform/graphics/chromium/LayerRendererChromium.cpp:

(WebCore::LayerRendererChromium::initialize):

2:13 PM Changeset in webkit [93541] by pkasting@chromium.org
  • 2 edits in trunk/LayoutTests

[chromium] Disable compositing tests for Leopard.

Reviewed by James Robinson.

  • platform/chromium/test_expectations.txt:
2:07 PM Changeset in webkit [93540] by caryclark@google.com
  • 2 edits in trunk/Source/WebCore

Webkit Mac10.6 Builder-Tester (Skia)(deps)(dbg) has compile failures
​https://bugs.webkit.org/show_bug.cgi?id=66703

This error was detected by the debug build of the Chromium WebKit
Canary.

Reviewed by James Robinson.

No new tests. This change suppresses a compile-time warning
but does not affect the generated code.

  • platform/graphics/mac/FontCustomPlatformData.h:

Replaced struct with class in a forward reference.

2:05 PM Changeset in webkit [93539] by andersca@apple.com
  • 2 edits in trunk/Source/WebKit2

Update the current NSTextInputContext whenever the plug-in complex text input state changes
​https://bugs.webkit.org/show_bug.cgi?id=66709

Reviewed by Sam Weinig.

Call [NSTextInputContext currentInputContext] which will force AppKit to update the current input
context. Fixes a rare case when the out-of-line editing window wouldn't appear.

  • UIProcess/API/mac/WKView.mm:

(-[WKView _setPluginComplexTextInputState:WebKit::pluginComplexTextInputIdentifier:]):

2:04 PM Changeset in webkit [93538] by commit-queue@webkit.org
  • 11 edits
    6 adds in trunk

[CSSRegions] RenderFlowThread layout should use the attached region sizes
​https://bugs.webkit.org/show_bug.cgi?id=66143

Source/WebCore:

When RenderFlowThread is doing the layout it also sets the current RenderFlowThread on
the RenderView. RenderFlowThread cannot be nested inside a single RenderView, so there's no
need to put that on the LayoutState.

The right offset used to layout the inlines inside a RenderFlowThread
is now adjusted to fit inside the current region.

Patch by Alexandru Chiculita <Alexandru Chiculita> on 2011-08-22
Reviewed by David Hyatt.

Tests: fast/regions/webkit-flow-floats-inside-regions-bounds.html

fast/regions/webkit-flow-inlines-inside-regions-bounds-vertical-lr.html
fast/regions/webkit-flow-inlines-inside-regions-bounds.html

  • rendering/LayoutState.cpp:

(WebCore::LayoutState::LayoutState):

  • rendering/RenderBlock.cpp:

(WebCore::RenderBlock::layoutBlock):
(WebCore::RenderBlock::adjustForRegionFittingIfNeeded):
(WebCore::RenderBlock::logicalRightOffsetForLine):

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

(WebCore::RenderFlowThread::RenderFlowThread):
(WebCore::CurrentRenderFlowThreadMaintainer::CurrentRenderFlowThreadMaintainer):
(WebCore::CurrentRenderFlowThreadMaintainer::~CurrentRenderFlowThreadMaintainer):
(WebCore::RenderFlowThread::layout):
(WebCore::RenderFlowThread::renderRegionForLine):
(WebCore::RenderFlowThread::regionLogicalWidthForLine):

  • rendering/RenderFlowThread.h:

(WebCore::RegionFittingDisabler::RegionFittingDisabler):
(WebCore::RegionFittingDisabler::~RegionFittingDisabler):

  • rendering/RenderRegion.cpp:

(WebCore::RenderRegion::layout):

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

(WebCore::RenderView::RenderView):

  • rendering/RenderView.h:

(WebCore::RenderView::hasRenderFlowThread):
(WebCore::RenderView::currentRenderFlowThread):
(WebCore::RenderView::setCurrentRenderFlowThread):
(WebCore::RenderView::pushLayoutState):

LayoutTests:

Patch by Alexandru Chiculita <Alexandru Chiculita> on 2011-08-22
Reviewed by David Hyatt.

  • fast/regions/webkit-flow-floats-inside-regions-bounds-expected.txt: Added.
  • fast/regions/webkit-flow-floats-inside-regions-bounds.html: Added.
  • fast/regions/webkit-flow-inlines-inside-regions-bounds-expected.txt: Added.
  • fast/regions/webkit-flow-inlines-inside-regions-bounds-vertical-lr-expected.txt: Added.
  • fast/regions/webkit-flow-inlines-inside-regions-bounds-vertical-lr.html: Added.
  • fast/regions/webkit-flow-inlines-inside-regions-bounds.html: Added.
1:53 PM Changeset in webkit [93537] by Dimitri Glazkov
  • 2 edits in trunk/Source/WebKit/chromium

Remove third_party/leveldb, because it's no longer used and doesn't exist downstream.

  • DEPS: Done it.
1:45 PM Changeset in webkit [93536] by levin@chromium.org
  • 1 edit
    6 adds in trunk/LayoutTests

REGRESSION (r89086): All worker xhr requests trigger preflight requests.
​https://bugs.webkit.org/show_bug.cgi?id=66340

Reviewed by Darin Adler.

Added tests which are similar to the non worker test with the same name.

  • http/tests/xmlhttprequest/workers/access-control-basic-get-fail-non-simple-expected.txt: Added.
  • http/tests/xmlhttprequest/workers/access-control-basic-get-fail-non-simple.html: Added.
  • http/tests/xmlhttprequest/workers/resources/access-control-basic-get-fail-non-simple-test.js: Added.
  • http/tests/xmlhttprequest/workers/resources/access-control-basic-get-fail-non-simple.js: Added.
  • http/tests/xmlhttprequest/workers/shared-worker-access-control-basic-get-fail-non-simple-expected.txt: Added.
  • http/tests/xmlhttprequest/workers/shared-worker-access-control-basic-get-fail-non-simple.html: Added.
1:39 PM Changeset in webkit [93535] by Dimitri Glazkov
  • 2 edits in trunk/Source/WebKit/chromium

Roll Chromium DEPS to 97698.

  • DEPS: Rolled.
1:38 PM Changeset in webkit [93534] by jamesr@google.com
  • 1 edit
    1 move in trunk/LayoutTests

[chromium] Rename chromium-gpu-mac to chromium-gpu-cg-mac

  • platform/chromium-gpu-cg-mac: Copied from LayoutTests/platform/chromium-gpu-mac.
  • platform/chromium-gpu-mac: Removed.
1:34 PM Changeset in webkit [93533] by jam@chromium.org
  • 5 edits in branches/chromium/835/Source

Merge 93296 - Prepare frames for history navigation.

Before beginning an explicit history navigation in a newly initialized
frame, it's important to give a "previous" item to avoid crashes, and
give a state to the state machine to avoid extra validating loads.

​https://bugs.webkit.org/show_bug.cgi?id=66322

Reviewed by Darin Fisher.

Source/WebCore:

No new tests, as history navigation is very difficult to test in
DumpRenderTree.

  • loader/FrameLoader.cpp:

(WebCore::FrameLoader::prepareForHistoryNavigation):

  • loader/FrameLoader.h:

Source/WebKit/chromium:

  • src/WebFrameImpl.cpp:

(WebKit::WebFrameImpl::loadHistoryItem):

TBR=​gavinp@chromium.org
Review URL: ​http://codereview.chromium.org/7711011

1:23 PM Changeset in webkit [93532] by pkasting@chromium.org
  • 3 edits
    3 deletes in trunk

Unreviewed, rolling out r93517.
​http://trac.webkit.org/changeset/93517
​https://bugs.webkit.org/show_bug.cgi?id=63553

Breaks GTK Linux and Chromium Win

Source/WebCore:

  • svg/animation/SVGSMILElement.cpp:

(WebCore::SVGSMILElement::calculateAnimationPercentAndRepeat):

LayoutTests:

  • svg/animations/animate-end-attribute-expected.txt: Removed.
  • svg/animations/animate-end-attribute.html: Removed.
  • svg/animations/script-tests/animate-end-attribute.js: Removed.
1:05 PM Changeset in webkit [93531] by commit-queue@webkit.org
  • 4 edits in trunk/Source/WebCore

FrameSelection::setNonDirectionalSelectionIfNeeded should not have a FrameSelection passed to it
​https://bugs.webkit.org/show_bug.cgi?id=66636

Patch by Wyatt Carss <​wcarss@chromium.org> on 2011-08-22
Reviewed by Darin Adler.

setNonDirectionalSelection was moved into FrameSelection in r93134 for
bug 60529, and still took a FrameSelection as an argument, which has been removed.

  • editing/FrameSelection.cpp:

(WebCore::FrameSelection::setNonDirectionalSelectionIfNeeded):

12:35 PM Changeset in webkit [93530] by jamesr@google.com
  • 4 edits in trunk/Tools

[chromium] Update builder names for chromium gpu bots due to CG switchover
​https://bugs.webkit.org/show_bug.cgi?id=66693

Reviewed by Dirk Pranke.

Update the port and builder names for the chromium mac gpu bots to
account for the CG transition.

  • Scripts/webkitpy/layout_tests/port/builders.py:
12:12 PM Changeset in webkit [93529] by rniwa@webkit.org
  • 5 edits in trunk/LayoutTests

editing/selection/caret-at-bidi-boundary.html timing out on Windows since it was added
​https://bugs.webkit.org/show_bug.cgi?id=66551

Reviewed by Adam Roben.

Replaced leapForward by dummy clicks to make it work on Windows and other ports.

  • editing/selection/caret-at-bidi-boundary.html:
  • platform/gtk/Skipped:
  • platform/qt/Skipped:
  • platform/win/Skipped:
12:08 PM Changeset in webkit [93528] by tony@chromium.org
  • 13 edits in trunk

Add CSS parsing of -webkit-flex-order
​https://bugs.webkit.org/show_bug.cgi?id=65384

Source/WebCore:

The property is an int described here:
​http://dev.w3.org/csswg/css3-flexbox/#flex-order

Doesn't do anything yet, just hooking it up.

Reviewed by Ojan Vafai.

  • css/CSSComputedStyleDeclaration.cpp:

(WebCore::CSSComputedStyleDeclaration::getPropertyCSSValue):

  • css/CSSParser.cpp:

(WebCore::CSSParser::parseValue):

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

(WebCore::CSSStyleApplyProperty::CSSStyleApplyProperty):

  • css/CSSStyleSelector.cpp:

(WebCore::CSSStyleSelector::applyProperty):

  • rendering/style/RenderStyle.h:

(WebCore::InheritedFlags::flexOrder):
(WebCore::InheritedFlags::setFlexOrder):
(WebCore::InheritedFlags::initialFlexOrder):

  • rendering/style/StyleFlexibleBoxData.cpp:

(WebCore::StyleFlexibleBoxData::StyleFlexibleBoxData):
(WebCore::StyleFlexibleBoxData::operator==):

  • rendering/style/StyleFlexibleBoxData.h:

LayoutTests:

Reviewed by Ojan Vafai.

  • css3/flexbox/css-properties-expected.txt:
  • css3/flexbox/css-properties.html:
  • css3/flexbox/script-tests/css-properties.js:
12:03 PM Changeset in webkit [93527] by dpranke@chromium.org
  • 2 edits in trunk/Tools

fix the GPU bot names on the layout test dashboard
​https://bugs.webkit.org/show_bug.cgi?id=66697

Reviewed by Dimitri Glazkov.

  • TestResultServer/static-dashboards/builders.js:
11:49 AM Changeset in webkit [93526] by enne@google.com
  • 2 edits in trunk/Source/WebCore

[chromium] Fix nullptr deref when switching composited tabs
​https://bugs.webkit.org/show_bug.cgi?id=66691

Reviewed by James Robinson.

  • platform/graphics/chromium/TiledLayerChromium.cpp:

(WebCore::TiledLayerChromium::protectTileTextures):

11:30 AM Changeset in webkit [93525] by kerz@chromium.org
  • 4 edits in branches/chromium/835/Source/WebCore

Merge 93042 - Web Inspector: [V8] crash upon stepIn while not on pause.
​https://bugs.webkit.org/show_bug.cgi?id=66221

Reviewed by Yury Semikhatsky.

  • inspector/InspectorDebuggerAgent.cpp:

(WebCore::InspectorDebuggerAgent::resume):
(WebCore::InspectorDebuggerAgent::stepOver):
(WebCore::InspectorDebuggerAgent::stepInto):
(WebCore::InspectorDebuggerAgent::stepOut):
(WebCore::InspectorDebuggerAgent::assertPaused):

  • inspector/InspectorDebuggerAgent.h:
  • inspector/front-end/ScriptsPanel.js:

(WebInspector.ScriptsPanel.prototype._stepOverClicked):
(WebInspector.ScriptsPanel.prototype._stepIntoClicked):
(WebInspector.ScriptsPanel.prototype._stepOutClicked):

TBR=​pfeldman@chromium.org
Review URL: ​http://codereview.chromium.org/7708013

11:28 AM Changeset in webkit [93524] by kerz@chromium.org
  • 4 edits in branches/chromium/835

Merge 92764 - Web Inspector: [REGRESSION] Editor lost after committing a CSS property value for inline style
​https://bugs.webkit.org/show_bug.cgi?id=65918

Reviewed by Pavel Feldman.

Source/WebCore:

  • inspector/front-end/StylesSidebarPane.js:

(WebInspector.StylesSidebarPane.prototype._innerUpdate.stylesCallback):
(WebInspector.StylesSidebarPane.prototype._innerUpdate.computedStyleCallback):
(WebInspector.StylesSidebarPane.prototype._innerUpdate):
(WebInspector.StylePropertyTreeElement.prototype):
(WebInspector.StylePropertyTreeElement.prototype.event):
(WebInspector.StylePropertyTreeElement.prototype.styleText.updateInterface.majorChange.isRevert.parentPane):
(WebInspector.StylePropertyTreeElement.prototype.styleText.updateInterface.majorChange.isRevert):

LayoutTests:

  • inspector/styles/styles-commit-editing.html:

TBR=​apavlov@chromium.org
Review URL: ​http://codereview.chromium.org/7708012

11:18 AM Changeset in webkit [93523] by commit-queue@webkit.org
  • 2 edits in trunk/Source/WebCore

[chromium] Content and render surface textures not freed when compositor context destroyed
​https://bugs.webkit.org/show_bug.cgi?id=66589

Patch by James Robinson <​jamesr@chromium.org> on 2011-08-22
Reviewed by Kenneth Russell.

Explicitly delete all managed textures when destroying the compositor
context. This regressed due to a combination of using share groups
(so destroying a context no longer implicitly releases all resources
in that context) and switching the TextureManager over to support
deferred texture destruction.

We sadly don't have a way to construct automated tests for resource
leaks currently.

  • platform/graphics/chromium/LayerRendererChromium.cpp:

(WebCore::LayerRendererChromium::cleanupSharedObjects):

11:15 AM Changeset in webkit [93522] by Dimitri Glazkov
  • 6 edits in trunk/Tools

garden-o-matic's relative time display should be updated regularly.
​https://bugs.webkit.org/show_bug.cgi?id=66682

Reviewed by Adam Barth.

  • BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js: Moved Time to ui.Time, added code for regular updates.
  • BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js: Changed to use ui.Time.
  • BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js: Added tests.
11:12 AM Changeset in webkit [93521] by Nate Chapin
  • 3 edits
    4 adds in trunk

Source/WebCore: Unload events can crash us when they blank out
a parent frame.
​https://bugs.webkit.org/show_bug.cgi?id=64741

Reviewed by Adam Barth.

Test: fast/loader/document-destruction-within-unload.html

  • loader/FrameLoader.cpp:

(WebCore::FrameLoader::stopLoading): Prevent unload events

from going into infinite recursion.

(WebCore::FrameLoader::setDocumentLoader): Ensure we don't

let set m_documentLoader to a DocumentLoader with a null Frame*.

(WebCore::FrameLoader::detachChildren): Save off a vector of

children to detach, rather than doing it inline.

LayoutTests: Test for ​https://bugs.webkit.org/show_bug.cgi?id=64741.

Reviewed by Adam Barth.

  • fast/loader/document-destruction-within-unload-expected.txt: Added.
  • fast/loader/document-destruction-within-unload.html: Added.
  • fast/loader/resources/document-destruction-within-unload-iframe.html: Added.
  • fast/loader/resources/document-destruction-within-unload.svg: Added.
11:07 AM Changeset in webkit [93520] by enne@google.com
  • 2 edits in trunk/LayoutTests

[chromium] Skip rtl-composited, as it causes other tests to crash.

Unreviewed expectation change.

  • platform/chromium/test_expectations.txt:
11:06 AM Changeset in webkit [93519] by commit-queue@webkit.org
  • 2 edits
    3 adds in trunk/LayoutTests

[Chromium] plugins/keyboard-events.html should pass on linux/mac
​https://bugs.webkit.org/show_bug.cgi?id=65964

Patch by Noel Gordon <​noel.gordon@gmail.com> on 2011-08-22
Reviewed by Tony Chang.

Update the Chromium Linux test results following r88712. Also rebaseline test results
for Chromium Mac. Note the WebKit Mac and Win ports both report 0 for the line number
for all console.log() events. The Chromium port reports the actual line number, hence
the need for separate test results for Chromium. The test now runs, and passes on all
the Chromium ports, following r87855. The test confirms user-plugin interactivity, so
we don't regress sites like youtube.com. Refer to bug 34936.

  • platform/chromium-linux/plugins/keyboard-events-expected.txt: Added.
  • platform/chromium-mac/plugins/keyboard-events-expected.txt: Added.
  • platform/chromium/test_expectations.txt:
11:03 AM Changeset in webkit [93518] by commit-queue@webkit.org
  • 4 edits
    3 moves
    2 deletes in trunk/LayoutTests

Rebaseline for webkit bug 65295
​https://bugs.webkit.org/show_bug.cgi?id=66637

Patch by Ben Wells <​benwells@chromium.org> on 2011-08-22
Reviewed by Kenneth Russell.

  • fast/box-shadow/box-shadow-radius-expected.txt: Renamed from LayoutTests/platform/gtk/fast/box-shadow/box-shadow-radius-expected.txt.
  • fast/layers/self-painting-outline-expected.txt: Renamed from LayoutTests/platform/gtk/fast/layers/self-painting-outline-expected.txt.
  • platform/chromium-win/fast/box-shadow/box-shadow-radius-expected.png:
  • platform/chromium-win/fast/layers/self-painting-outline-expected.png:
  • platform/chromium/test_expectations.txt:
  • platform/gtk/fast/box-shadow/box-shadow-radius-expected.png: Renamed from LayoutTests/fast/box-shadow/box-shadow-radius-expected.png.
  • platform/mac/fast/layers/self-painting-outline-expected.txt: Removed.
  • platform/qt/fast/layers/self-painting-outline-expected.txt: Removed.
11:00 AM Changeset in webkit [93517] by commit-queue@webkit.org
  • 3 edits
    3 adds in trunk

Patch by Young Han Lee <​joybro@company100.net> on 2011-08-22
Reviewed by Dirk Schulze.

SVG animation fill="freeze" doesn't set baseVal to current animVal if animation stops before reaching the end
​https://bugs.webkit.org/show_bug.cgi?id=63553

calculateAnimationPercentAndRepeat() is returning 1, which means 100%, whenever
elapsed >= m_intervalEnd, but this is wrong because m_intervalEnd can be in the middle
of the animation duration. (e.g. begin="0s" end="2s" dur="3s")

This change makes the function return the animations's true progress instead of 100%
when the animation ends.

Source/WebCore:

Test: svg/animations/animate-end-attribute.html

  • svg/animation/SVGSMILElement.cpp:

(WebCore::SVGSMILElement::calculateAnimationPercentAndRepeat):

LayoutTests:

  • svg/animations/animate-end-attribute-expected.txt: Added.
  • svg/animations/animate-end-attribute.html: Added.
  • svg/animations/script-tests/animate-end-attribute.js: Added.

(sample1):
(sample2):
(sample3):
(executeTest):

10:49 AM Changeset in webkit [93516] by gyuyoung.kim@samsung.com
  • 2 edits in trunk/Source/WebCore

Fix build error when inspector is disabled.
​https://bugs.webkit.org/show_bug.cgi?id=66644

Reviewed by Darin Adler.

  • inspector/InspectorDatabaseResource.h: Add ENABLE(INSPECTOR) macro.
10:49 AM Changeset in webkit [93515] by commit-queue@webkit.org
  • 3 edits in trunk/Tools

Pass --makeargs of build-webkit to chromium linux builder
​https://bugs.webkit.org/show_bug.cgi?id=66361

Patch by Xianzhu Wang <​wangxianzhu@chromium.org> on 2011-08-22
Reviewed by Tony Chang.

  • Scripts/build-webkit:
  • Scripts/webkitdirs.pm:

(buildChromiumMakefile):
(buildChromium):

10:33 AM Changeset in webkit [93514] by inferno@chromium.org
  • 3 edits
    2 adds in trunk

Crash in FocusController::advanceFocusInDocumentOrder
​https://bugs.webkit.org/show_bug.cgi?id=66678

Source/WebCore:

RefPtr the focusable node to prevent getting deleted by mutation
event.

Reviewed by Dave Hyatt.

Test: fast/frames/focus-controller-crash-change-event.html

  • page/FocusController.cpp:

(WebCore::FocusController::advanceFocusInDocumentOrder):

LayoutTests:

Reviewed by Dave Hyatt.

  • fast/frames/focus-controller-crash-change-event-expected.txt: Added.
  • fast/frames/focus-controller-crash-change-event.html: Added.
10:07 AM Changeset in webkit [93513] by Martin Robinson
  • 3 edits in trunk/LayoutTests

[GTK] Some GTK+-specific font-face tests fail on the bots
​https://bugs.webkit.org/show_bug.cgi?id=65820

Fix the timing issues in this test.

Reviewed by Philippe Normand.

  • platform/gtk/fonts/custom-font-missing-glyphs.html: Delay the test completion until

after document.readyState moves to "complete" and a short timeout.

  • platform/gtk/fonts/font-face-with-complex-text.html: Ditto.
9:02 AM Changeset in webkit [93512] by commit-queue@webkit.org
  • 4 edits in trunk/Source

Source/WebCore: [Chromium] Crash when allocation of very large canvas fails
​https://bugs.webkit.org/show_bug.cgi?id=66563
Replacing the crash with a more graceful failure: canvas
will not render.

Patch by Justin Novosad <​junov@chromium.org> on 2011-08-22
Reviewed by Stephen White.

  • platform/graphics/skia/ImageBufferSkia.cpp:

(WebCore::ImageBuffer::ImageBuffer):
Invoking TryCreateBitmapCanvas instead of CreateBitmapCanvas
This is the non-crashing version of the factory function.
It returns NULL if allocation fails. Code for handling
NULL pointer was already in place.

Source/WebKit/chromium: [Chromium] Crash when allocation of very large canvas fails
​https://bugs.webkit.org/show_bug.cgi?id=66563

Patch by Justin Novosad <​junov@chromium.org> on 2011-08-22
Reviewed by Stephen White.

  • DEPS:

Bumping up the chrome DEPS to pick up 97474, which is
required by this patch.

8:52 AM Changeset in webkit [93511] by commit-queue@webkit.org
  • 2 edits in trunk/Tools

Add myself to the contributor list.
​https://bugs.webkit.org/show_bug.cgi?id=66674

Patch by Peter Beverloo <​peter@chromium.org> on 2011-08-22
Reviewed by Tony Gentilcore.

  • Scripts/webkitpy/common/config/committers.py:
8:36 AM Changeset in webkit [93510] by loislo@chromium.org
  • 14 edits in trunk/Source/WebCore

Web Inspector: bind registerDomainDispatcher to domain names.
​https://bugs.webkit.org/show_bug.cgi?id=66670

jscompiler can't be used for validating domain events dispatchers because registerDomainDispatcher is too generic.
I'd like to introduce named versions per each domain.

Reviewed by Pavel Feldman.

  • inspector/CodeGeneratorInspector.pm:

(GenerateInterface):
(generateBackendStubJS):

  • inspector/front-end/ApplicationCacheItemsView.js:
  • inspector/front-end/ConsoleView.js:
  • inspector/front-end/DOMAgent.js:

(WebInspector.DOMAgent):

  • inspector/front-end/DOMStorage.js:
  • inspector/front-end/Database.js:
  • inspector/front-end/DebuggerModel.js:

(WebInspector.DebuggerModel):

  • inspector/front-end/NetworkManager.js:

(WebInspector.NetworkDispatcher):

  • inspector/front-end/ProfilesPanel.js:
  • inspector/front-end/ResourceTreeModel.js:

(WebInspector.ResourceTreeModel):

  • inspector/front-end/TimelineManager.js:

(WebInspector.TimelineDispatcher):

  • inspector/front-end/WorkerManager.js:

(WebInspector.WorkerManager):

  • inspector/front-end/inspector.js:
8:17 AM Changeset in webkit [93509] by kbalazs@webkit.org
  • 2 edits in trunk/Tools

[Qt] ImageDiff does not consider --tolerance
​https://bugs.webkit.org/show_bug.cgi?id=66299

Reviewed by Csaba Osztrogonác.

  • DumpRenderTree/qt/ImageDiff.cpp:

(main):

6:43 AM Changeset in webkit [93508] by Csaba Osztrogonác
  • 2 edits in trunk/LayoutTests

[NRWT] Fix "Only unexpected results" option of results.html
​https://bugs.webkit.org/show_bug.cgi?id=66657

Patch by Kristóf Kosztyó <​kkristof@inf.u-szeged.hu> on 2011-08-22
Reviewed by Csaba Osztrogonác.

  • fast/harness/results.html:
6:26 AM Changeset in webkit [93507] by benjamin.poulain@nokia.com
  • 2 edits in trunk/Source/WebCore

Fix generate-inspector-idl for Python3 after r93396
​https://bugs.webkit.org/show_bug.cgi?id=66663

Reviewed by Andreas Kling.

In Python3, print is a function instead of a language keyword. Using parentheses
around the argument ensures the code works for both Python 2 and 3.

  • inspector/generate-inspector-idl:
5:22 AM Changeset in webkit [93506] by vitalyr@chromium.org
  • 2 edits in trunk/LayoutTests

2011-08-22 Vitaly Repeshko <​vitalyr@chromium.org>

[chromium] Updating test expectations.

Unreviewed.

  • platform/chromium/test_expectations.txt:
5:06 AM Changeset in webkit [93505] by commit-queue@webkit.org
  • 3 edits in trunk/LayoutTests

[Chromium] [WebSocket] Add a test to reproduce chromium bug on close before open.
​https://bugs.webkit.org/show_bug.cgi?id=66649

Patch by Takashi Toyoshima <​toyoshim@chromium.org> on 2011-08-22
Reviewed by Kent Tamura.

  • http/tests/websocket/tests/hybi/close-before-open-expected.txt:
  • http/tests/websocket/tests/hybi/close-before-open.html:

Add new URL to reproduce the problem.

5:03 AM Changeset in webkit [93504] by Csaba Osztrogonác
  • 2 edits in trunk/Tools

Update contributor list.

Reviewed by Zoltan Herczeg.

  • Scripts/webkitpy/common/config/committers.py:
4:53 AM Changeset in webkit [93503] by keishi@webkit.org
  • 3 edits in trunk/Source/WebCore

Chrome::setSelectedColorInColorChooser shouldn't be called when color chooser sets a new color
​https://bugs.webkit.org/show_bug.cgi?id=66658

Reviewed by Kent Tamura.

  • html/ColorInputType.cpp:

(WebCore::ColorInputType::colorSelected): Call setValueFromRenderer instead of setValue.

  • html/ColorInputType.h:
3:05 AM Changeset in webkit [93502] by Csaba Osztrogonác
  • 2 edits in trunk/Tools

[ORWT] Fix --reset-results --add-platform-exceptions combo
​https://bugs.webkit.org/show_bug.cgi?id=65464

Patch by Kristóf Kosztyó <​Kosztyo.Kristof@stud.u-szeged.hu> on 2011-08-22
Reviewed by Csaba Osztrogonác.

  • Scripts/old-run-webkit-tests:
2:17 AM Changeset in webkit [93501] by podivilov@chromium.org
  • 8 edits
    3 adds in trunk

Web Inspector: extract breakpoint management code to a separate class and add tests.
​https://bugs.webkit.org/show_bug.cgi?id=66224

Reviewed by Pavel Feldman.

Source/WebCore:

Test: inspector/debugger/breakpoint-manager.html

  • WebCore.gypi:
  • WebCore.vcproj/WebCore.vcproj:
  • inspector/front-end/BreakpointManager.js: Added.

(WebInspector.BreakpointManager):
(WebInspector.BreakpointManager.prototype.uiSourceCodeAdded):
(WebInspector.BreakpointManager.prototype.breakpointsForUISourceCode):
(WebInspector.BreakpointManager.prototype.setBreakpoint):
(WebInspector.BreakpointManager.prototype.removeBreakpoint):
(WebInspector.BreakpointManager.prototype._materializeBreakpoint):
(WebInspector.BreakpointManager.prototype._breakpointDebuggerLocationChanged):
(WebInspector.BreakpointManager.prototype._addBreakpointToUI):
(WebInspector.BreakpointManager.prototype._deleteBreakpointFromUI):
(WebInspector.BreakpointManager.prototype._moveBreakpointInUI):
(WebInspector.BreakpointManager.prototype._breakpoints):
(WebInspector.BreakpointManager.prototype._breakpoint):
(WebInspector.BreakpointManager.prototype._forEachBreakpoint):
(WebInspector.BreakpointManager.prototype._setBreakpointInDebugger):
(WebInspector.BreakpointManager.prototype._removeBreakpointFromDebugger):
(WebInspector.BreakpointManager.prototype._breakpointResolved):
(WebInspector.BreakpointManager.prototype.serializeBreakpoints):
(WebInspector.BreakpointManager.prototype.reset):
(WebInspector.BreakpointManager.prototype.debuggerReset):
(WebInspector.Breakpoint):
(WebInspector.Breakpoint.prototype.serialize):
(WebInspector.Breakpoint.deserialize):

  • inspector/front-end/DebuggerModel.js:

(WebInspector.DebuggerModel.prototype.setBreakpointByScriptLocation):

  • inspector/front-end/SourceFile.js:

(WebInspector.RawSourceCode.prototype.get rawSourceCode):

  • inspector/front-end/WebKit.qrc:
  • inspector/front-end/inspector.html:

LayoutTests:

  • inspector/debugger/breakpoint-manager-expected.txt: Added.
  • inspector/debugger/breakpoint-manager.html: Added.
2:09 AM Changeset in webkit [93500] by zoltan@webkit.org
  • 2 edits in trunk/LayoutTests

[Qt][ARM] Add bug report to flaky tests

http/tests/cache/subresource-multiple-instances.html
​https://bugs.webkit.org/show_bug.cgi?id=66652

http/tests/inspector/resource-tree/resource-tree-errors-reload.html
​https://bugs.webkit.org/show_bug.cgi?id=66653

  • platform/qt-arm/Skipped:
2:07 AM Changeset in webkit [93499] by cevans@google.com
  • 4 edits
    2 copies in branches/chromium/835

Merge 93347
BUG=90357
Review URL: ​http://codereview.chromium.org/7696023

2:04 AM Changeset in webkit [93498] by cevans@google.com
  • 1 edit
    2 copies in branches/chromium/835

Merge 93032
BUG=92769
Review URL: ​http://codereview.chromium.org/7693023

1:56 AM Changeset in webkit [93497] by cevans@google.com
  • 6 edits
    3 copies in branches/chromium/835

Merge 93227
BUG=92959
Review URL: ​http://codereview.chromium.org/7694026

1:49 AM Changeset in webkit [93496] by cevans@google.com
  • 1 edit
    2 copies in branches/chromium/835

Merge 92132
BUG=89991
Review URL: ​http://codereview.chromium.org/7693022

1:39 AM Changeset in webkit [93495] by cevans@google.com
  • 2 edits in branches/chromium/835

Merge 93397
BUG=76771
Review URL: ​http://codereview.chromium.org/7697013

1:05 AM Changeset in webkit [93494] by zoltan@webkit.org
  • 2 edits in trunk/LayoutTests

[Qt][ARM] Skip flaky tests.

  • platform/qt-arm/Skipped:

Aug 21, 2011:

9:04 PM Changeset in webkit [93493] by jamesr@google.com
  • 6 edits in branches/chromium/835/Source/WebCore

Merge 93441 - REGRESSION(r91628): 3 canvas tests crash on Chromium Linux and one test fail on Chromium Mac
​https://bugs.webkit.org/show_bug.cgi?id=65063

Reviewed by Darin Fisher.

Fixes crashes due to a bad cast from Image to BitmapImage on skia ports
that use BitmapImageSingleFrameSkia. In the skia port
Image::isBitmapImage() returning true does not necessarily mean that
Image is of type BitmapImage.

Covered by fast/canvas/canvas-as-image.html and many other canvas
tests.

Patch originally by Tom Hudson.

  • platform/graphics/BitmapImage.h:

(WebCore::BitmapImage::currentFrameHasAlpha):
(WebCore::BitmapImage::notSolidColor):

  • platform/graphics/Image.cpp:

(WebCore::Image::drawTiled):

  • platform/graphics/Image.h:

(WebCore::Image::currentFrameHasAlpha):
(WebCore::Image::notSolidColor):

  • platform/graphics/skia/BitmapImageSingleFrameSkia.h:

(WebCore::BitmapImageSingleFrameSkia::currentFrameHasAlpha):
(WebCore::BitmapImageSingleFrameSkia::notSolidColor):

  • rendering/RenderImage.cpp:

(WebCore::RenderImage::backgroundIsObscured):

BUG=93558
TBR=​jamesr@google.com
Review URL: ​http://codereview.chromium.org/7696019

6:10 PM Changeset in webkit [93492] by Martin Robinson
  • 5 edits in trunk/Source

Fix 'make dist' for WebKitGTK+.

Source/JavaScriptCore:

  • GNUmakefile.list.am: Add a missing header to the sources list.

Source/WebCore:

  • GNUmakefile.am: Add a missing file to the EXTRA_DIST list.
  • GNUmakefile.list.am: Update the sources list to reflect reality.
1:22 PM Changeset in webkit [93491] by Darin Adler
  • 14 edits in trunk/LayoutTests

Update results for more tests affected by the addition of the
HTMLSourceElement class.

  • fast/dom/Window/window-properties-expected.txt:
  • fast/dom/Window/window-property-descriptors-expected.txt:
  • platform/gtk/fast/dom/Window/window-properties-expected.txt:
  • platform/gtk/fast/dom/Window/window-property-descriptors-expected.txt:
  • platform/gtk/fast/js/global-constructors-expected.txt:
  • platform/mac/fast/js/global-constructors-expected.txt:
  • platform/qt-wk2/fast/dom/Window/window-properties-expected.txt:
  • platform/qt-wk2/fast/dom/Window/window-property-descriptors-expected.txt:
  • platform/qt/fast/dom/Window/window-properties-expected.txt:
  • platform/qt/fast/dom/Window/window-property-descriptors-expected.txt:
  • platform/qt/fast/js/global-constructors-expected.txt:
  • platform/win/fast/dom/Window/window-property-descriptors-expected.txt:
  • platform/win/fast/js/global-constructors-expected.txt:
8:53 AM Changeset in webkit [93490] by commit-queue@webkit.org
  • 3 edits in trunk/Source/WebKit/qt

[Qt] Fixes premature plugin unload bug
​https://bugs.webkit.org/show_bug.cgi?id=66628

Patch by Lindsay Mathieson <​lindsay.mathieson@gmail.com> on 2011-08-21
Reviewed by Benjamin Poulain.

The QtPlatformPlugin was allocated on the stack, so its destructor is unloading the
plugin at the end of the scope. This patch moves the local creation of QtPlatformPlugin
to class member (TextCheckerClientQt::m_platformPlugin), preventing premature unloading
of plugin.

  • WebCoreSupport/TextCheckerClientQt.cpp:

(WebCore::TextCheckerClientQt::loadSpellChecker):

  • WebCoreSupport/TextCheckerClientQt.h:
Note: See TracTimeline for information about the timeline view.