⚠ Archived content — this site is no longer maintained.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

Timeline



Nov 5, 2011:

11:10 PM Changeset in webkit [99371] by Martin Robinson
  • 2 edits in trunk/LayoutTests

Skip a newly failing test and reclassify some missing DRT features.

  • platform/gtk/Skipped: Update skipped list.
10:33 PM Changeset in webkit [99370] by Martin Robinson
  • 4 edits
    2 deletes in trunk/Source/WebCore

Remove ContextShadow
​https://bugs.webkit.org/show_bug.cgi?id=71617

Reviewed by Ariya Hidayat.

No new tests. This patch does not change behavior.

Completely remove ContextShadow. It's unused, unmaintained, and
replaced by ShadowBlur.

  • WebCore.gypi: Remove references to deleted files.
  • WebCore.vcproj/WebCore.vcproj: Remove references to deleted files.
  • platform/graphics/ContextShadow.cpp: Removed.
  • platform/graphics/ContextShadow.h: Removed.
  • platform/gtk/WidgetRenderingContext.cpp: Remove now-inaccurate TODO.
9:27 PM Changeset in webkit [99369] by jonlee@apple.com
  • 6 edits in trunk/Source/WebCore

Dragging a file onto <input type="file"> should give distinct visual feedback
​https://bugs.webkit.org/show_bug.cgi?id=13897
<rdar://problem/5232483>

Reviewed by Dan Bernstein.

When hovering over a file input element, we set the button's state to active
to differentiate dragging one file over the input element (which populates that
element) versus over the document (which would load the file into the view).

  • html/HTMLInputElement.cpp:

(WebCore::HTMLInputElement::HTMLInputElement):
(WebCore::HTMLInputElement::canReceiveDroppedFiles):
(WebCore::HTMLInputElement::setCanReceiveDroppedFiles): If set, the element
is updated, which sets the active state on the button control.

  • html/HTMLInputElement.h: Add a boolean member representing whether the file

input can receive dropped files.

  • page/DragController.cpp:

(WebCore::DragController::DragController): Update/set the file input that
can receive dropped files.
(WebCore::DragController::dragExited):
(WebCore::DragController::tryDocumentDrag):
(WebCore::DragController::concludeEditDrag):

  • page/DragController.h:
  • rendering/RenderFileUploadControl.cpp:

(WebCore::RenderFileUploadControl::updateFromElement): Sets the button active
state if the input can receive dropped files.

9:14 PM Changeset in webkit [99368] by dcheng@chromium.org
  • 6 edits in trunk

[chromium] Use the security origin instead of the URL when checking notification permissions
​https://bugs.webkit.org/show_bug.cgi?id=71590

Reviewed by Adam Barth.

Source/WebKit/chromium:

  • public/WebNotificationPresenter.h:
  • src/NotificationPresenterImpl.cpp:

(WebKit::NotificationPresenterImpl::checkPermission):

Tools:

  • DumpRenderTree/chromium/NotificationPresenter.cpp:

(NotificationPresenter::checkPermission):

  • DumpRenderTree/chromium/NotificationPresenter.h:
7:26 PM Changeset in webkit [99367] by Darin Adler
  • 2 edits in trunk/Source/WebCore

Improve pending resource hash table code, including fixing a memory leak
​https://bugs.webkit.org/show_bug.cgi?id=71616

Reviewed by Adam Roben.

Memory leak fix and refactoring covered by existing tests.

  • svg/SVGDocumentExtensions.cpp:

(WebCore::SVGDocumentExtensions::addPendingResource): Replace the combination
of contains, get, and add with just add, removing an extra unneeded hash table
lookup each time this function is called.
(WebCore::SVGDocumentExtensions::isElementInPendingResources): Added a comment
about the performance of this function. Removed unnecessary check for an empty
map; the code already efficiently exits doing nothing without the check.
(WebCore::SVGDocumentExtensions::removeElementFromPendingResources): Removed
unnecessary check for an empty map; the code already efficiently does nothing
without the check. Use removePendingResource rather than calling remove to
avoid leaking the SVGPendingElements set.
(WebCore::SVGDocumentExtensions::removePendingResource): Replace the
combination of get and remove with a call to take, removing an extra
unneeded hash table lookup each time this function is called

6:07 PM Changeset in webkit [99366] by Darin Adler
  • 2 edits in trunk/Tools

Fix a small leak in Mac version of LayoutTestController
​https://bugs.webkit.org/show_bug.cgi?id=71615

Reviewed by Mark Rowe.

  • DumpRenderTree/mac/LayoutTestControllerMac.mm:

(LayoutTestController::applicationCacheDiskUsageForOrigin): Release the origin.
(LayoutTestController::localStorageDiskUsageForOrigin): Ditto.

5:56 PM Changeset in webkit [99365] by Martin Robinson
  • 2 edits in trunk/LayoutTests

Unskip some tests on GTK+ that seem to be passing.

  • platform/gtk/Skipped: Unskip tests.
5:50 PM Changeset in webkit [99364] by commit-queue@webkit.org
  • 2 edits in trunk/Source/WebCore

[GTK] Actually use the user_data arguments to gio async functions
instead of using g_object_set_data() to basically reinvent them.
​https://bugs.webkit.org/show_bug.cgi?id=71614

Patch by Dan Winship <​danw@gnome.org> on 2011-11-05
Reviewed by Martin Robinson.

No new tests; behavior is unchanged

  • platform/network/soup/ResourceHandleSoup.cpp:

(WebCore::ResourceHandleInternal::~ResourceHandleInternal):
(WebCore::cleanupSoupRequestOperation):
(WebCore::sendRequestCallback):
(WebCore::startHTTPRequest):
(WebCore::ResourceHandle::platformSetDefersLoading):
(WebCore::closeCallback):
(WebCore::readCallback):
(WebCore::startNonHTTPRequest):

5:39 PM Changeset in webkit [99363] by commit-queue@webkit.org
  • 2 edits in trunk/Source/WebCore

[GTK] Don't unnecessarily 0-initialize read buffers.
​https://bugs.webkit.org/show_bug.cgi?id=71612

Patch by Dan Winship <​danw@gnome.org> on 2011-11-05
Reviewed by Martin Robinson.

No new tests; behavior is unchanged

  • platform/network/soup/ResourceHandleSoup.cpp:

(WebCore::sendRequestCallback):

5:37 PM Changeset in webkit [99362] by commit-queue@webkit.org
  • 2 edits in trunk/Source/WebCore

[GTK] Remove a bit of dead code
​https://bugs.webkit.org/show_bug.cgi?id=71613

Patch by Dan Winship <​danw@gnome.org> on 2011-11-05
Reviewed by Martin Robinson.

No new tests; behavior is unchanged

  • platform/network/soup/ResourceHandleSoup.cpp:

(WebCore::readCallback): the caller that needed the convertToUTF16
code was removed in r77408, but the code was accidentally left
behind.

5:34 PM Changeset in webkit [99361] by Martin Robinson
  • 20 edits in trunk/LayoutTests

[GTK] Frame flattening tests are super-flaky after switching to backing store-based painting
​https://bugs.webkit.org/show_bug.cgi?id=71576

Reviewed by Antonio Gomes.

Instead of waiting until <body onload> to toggle the frame flattening
setting. Toggle is as soon as possible. This fixes an issue where
layouts that happen before the onload event lead to flaky results.

  • fast/frames/flattening/frameset-flattening-advanced.html:
  • fast/frames/flattening/frameset-flattening-grid.html:
  • fast/frames/flattening/frameset-flattening-simple.html:
  • fast/frames/flattening/frameset-flattening-subframe-resize.html:
  • fast/frames/flattening/frameset-flattening-subframesets.html:
  • fast/frames/flattening/iframe-flattening-crash.html:
  • fast/frames/flattening/iframe-flattening-fixed-height.html:
  • fast/frames/flattening/iframe-flattening-fixed-width-and-height-no-scrolling.html:
  • fast/frames/flattening/iframe-flattening-fixed-width-and-height-zero-size.html:
  • fast/frames/flattening/iframe-flattening-fixed-width-and-height.html:
  • fast/frames/flattening/iframe-flattening-fixed-width.html:
  • fast/frames/flattening/iframe-flattening-nested.html:
  • fast/frames/flattening/iframe-flattening-offscreen.html:
  • fast/frames/flattening/iframe-flattening-out-of-view-and-scroll.html:
  • fast/frames/flattening/iframe-flattening-out-of-view-scroll-and-relayout.html:
  • fast/frames/flattening/iframe-flattening-out-of-view.html:
  • fast/frames/flattening/iframe-flattening-selection-crash.html:
  • fast/frames/flattening/iframe-flattening-simple.html:
  • platform/gtk/Skipped: Unskip tests which are no longer flaky.
4:45 PM Changeset in webkit [99360] by Martin Robinson
  • 2 edits in trunk/LayoutTests

[GTK] Frame flattening tests are super-flaky after switching to backing store-based painting
​https://bugs.webkit.org/show_bug.cgi?id=71576

Patch by Zan Dobersek <​zandobersek@gmail.com> on 2011-11-05
Reviewed by Martin Robinson.

Skip frame flattening tests. These became flaky after backing store for
painting was introduced for the GTK port in r98827.

  • platform/gtk/Skipped:
2:55 PM Changeset in webkit [99359] by abarth@webkit.org
  • 5 edits in trunk/Source/WebCore

Refactor sandbox flag calculation to make implementing CSP sandbox directive easier
​https://bugs.webkit.org/show_bug.cgi?id=71603

Reviewed by Sam Weinig.

This patch changes the way we compute sandbox bits. Instead of caching
the result on Frame, we always compute the sandbox bits for a new
document fresh from the inputs to the process (the sandbox attribute
and the forced sandbox flags).

This patch is just refactoring. It shouldn't have an observable
effects. This patch makes it easier to implement CSP's sandbox
directive in a subsequent patch.

  • dom/Document.cpp:

(WebCore::Document::initSecurityContext):

  • html/HTMLFrameOwnerElement.cpp:

(WebCore::HTMLFrameOwnerElement::setSandboxFlags):

  • loader/FrameLoader.cpp:

(WebCore::FrameLoader::FrameLoader):
(WebCore::FrameLoader::init):
(WebCore::FrameLoader::effectiveSandboxFlags):
(WebCore::createWindow):

  • loader/FrameLoader.h:

(WebCore::FrameLoader::forceSandboxFlags):

1:37 PM Changeset in webkit [99358] by weinig@apple.com
  • 2 edits in trunk/Source/JavaScriptCore

Fix windows build.

1:29 PM Changeset in webkit [99357] by weinig@apple.com
  • 6 edits in trunk/Source/JavaScriptCore

Reduce the number of putWithAttributes
​https://bugs.webkit.org/show_bug.cgi?id=71597

Reviewed by Adam Roben.

Remove exports of removed functions.

  • runtime/JSActivation.cpp:

(JSC::JSActivation::putWithAttributes):
Calling the overload without the extra parameters does the same thing.

  • runtime/JSObject.cpp:

(JSC::JSObject::putWithAttributes):

  • runtime/JSObject.h:

Remove four unused JSObject::putWithAttributes overloads and make one of the remaining
two overloads not virtual, since no one overrides it.

11:16 AM Changeset in webkit [99356] by Martin Robinson
  • 2 edits in trunk/LayoutTests

Unskip a test on GTK+ that doesn't seem to be failing on the bots
or locally.

  • platform/gtk/Skipped: Unskip a test.
9:28 AM Changeset in webkit [99355] by andreas.kling@nokia.com
  • 5 edits in trunk/Source/WebCore

CSSStyleDeclaration: Devirtualize isMutableStyleDeclaration().
​https://bugs.webkit.org/show_bug.cgi?id=71609

Reviewed by Anders Carlsson.

Add a bool member to CSSStyleDeclaration that determines whether
it's a CSSMutableDeclaration.

  • css/CSSMutableStyleDeclaration.cpp:

(WebCore::CSSMutableStyleDeclaration::CSSMutableStyleDeclaration):

  • css/CSSMutableStyleDeclaration.h:
  • css/CSSStyleDeclaration.cpp:

(WebCore::CSSStyleDeclaration::CSSStyleDeclaration):

  • css/CSSStyleDeclaration.h:

(WebCore::CSSStyleDeclaration::isMutableStyleDeclaration):

9:25 AM Changeset in webkit [99354] by andreas.kling@nokia.com
  • 2 edits in trunk/Source/WebCore

Web Inspector: Simplify InspectorCSSAgent::inlineStyleElement().
​https://bugs.webkit.org/show_bug.cgi?id=71608

Reviewed by Pavel Feldman.

Use CSSMutableStyleDeclaration::isInlineStyleDeclaration() instead
of doing the same checks manually.

  • inspector/InspectorCSSAgent.cpp:

(WebCore::InspectorCSSAgent::inlineStyleElement):

7:50 AM Changeset in webkit [99353] by Simon Hausmann
  • 2 edits
    1 delete in trunk/Source/WebCore

Reviewed by Andreas Kling.

Removed duplicated npruntime_internal.h from bindings/v8
in favour of the one from bridge/, which has a few more

​https://bugs.webkit.org/show_bug.cgi?id=45617

  • bindings/v8/npruntime_internal.h: Removed.
  • Target.pri: Removed strange occurence here.
5:15 AM Changeset in webkit [99352] by Simon Hausmann
  • 2 edits
    1 delete in trunk/Source/WebCore

Remove empty and unused dom/PositionCreationFunctions.h header file
​https://bugs.webkit.org/show_bug.cgi?id=71552

Reviewed by Kenneth Rohde Christiansen.

The file is empty (0 bytes) and not used anywhere. It was added in
2009 in r48234 but hasn't been touched since then.

  • WebCore.gypi:
  • dom/PositionCreationFunctions.h: Removed.
12:00 AM Changeset in webkit [99351] by abarth@webkit.org
  • 2 edits in trunk/Source/WebCore

Fix assert after running tests in dumpAsText folders. See
​https://bugs.webkit.org/show_bug.cgi?id=71599 for discussion. This
patch reverts part of a slightly over-agressive refactoring in
​http://trac.webkit.org/changeset/99347.

  • dom/Document.cpp:

(WebCore::Document::setIsViewSource):

Nov 4, 2011:

10:50 PM Changeset in webkit [99350] by tony@chromium.org
  • 2 edits in trunk/LayoutTests

Update a timing out gpu test.

  • platform/chromium/test_expectations.txt:
6:59 PM Changeset in webkit [99349] by commit-queue@webkit.org
  • 5 edits
    3 adds in trunk

Canvas drawImage(canvas) with SourceIn, DestinationIn, SourceOut, DestinationAtop and Copy have errors
​https://bugs.webkit.org/show_bug.cgi?id=71537

Patch by Ben Wells <​benwells@chromium.org> on 2011-11-04
Reviewed by Stephen White.

Source/WebCore:

drawImage(canvas) now uses the same approach as drawImage(image) to fix these errors.

Test: fast/canvas/canvas-composite-canvas.html

  • html/canvas/CanvasRenderingContext2D.cpp:

(WebCore::CanvasRenderingContext2D::drawImage):
(WebCore::drawImageToContext):
(WebCore::CanvasRenderingContext2D::fullCanvasCompositedDrawImage):

  • html/canvas/CanvasRenderingContext2D.h:

LayoutTests:

  • fast/canvas/canvas-composite-canvas-expected.txt: Added.
  • fast/canvas/canvas-composite-canvas.html: Added.
  • fast/canvas/canvas-composite-image.html:
  • fast/canvas/resources/canvas-composite-image-common.js: Added.
6:35 PM Changeset in webkit [99348] by cevans@google.com
  • 1 edit
    2 copies in branches/chromium/874

Merge 98763
BUG=99597
Review URL: ​http://codereview.chromium.org/8478022

6:32 PM Changeset in webkit [99347] by abarth@webkit.org
  • 16 edits in trunk

JavaScript URLs execute in sandboxed iframes
​https://bugs.webkit.org/show_bug.cgi?id=71599

Reviewed by Eric Seidel.

Source/WebCore:

This patch fixes the intentional regression I introduced earlier today
by moving the sandbox bits from SecurityOrigin to Document. In the
process, I renamed SecurityOrigin::createEmpty to
SecurityOrigin::createUnique to better align with HTML5 terminology.

  • WebCore.exp.in:
  • bindings/ScriptControllerBase.cpp:

(WebCore::ScriptController::canExecuteScripts):

  • dom/Document.cpp:

(WebCore::Document::setIsViewSource):
(WebCore::Document::initSecurityContext):

  • dom/ScriptExecutionContext.cpp:

(WebCore::ScriptExecutionContext::ScriptExecutionContext):

  • dom/ScriptExecutionContext.h:

(WebCore::ScriptExecutionContext::sandboxFlags):
(WebCore::ScriptExecutionContext::enforceSandboxFlags):
(WebCore::ScriptExecutionContext::isSandboxed):

  • html/HTMLAppletElement.cpp:

(WebCore::HTMLAppletElement::canEmbedJava):

  • loader/DocumentWriter.cpp:

(WebCore::DocumentWriter::begin):

  • loader/FrameLoader.cpp:

(WebCore::isDocumentSandboxed):
(WebCore::FrameLoader::addHTTPOriginIfNeeded):

  • loader/PolicyChecker.cpp:

(WebCore::PolicyChecker::checkNewWindowPolicy):

  • loader/SubframeLoader.cpp:

(WebCore::SubframeLoader::requestPlugin):

  • page/SecurityOrigin.cpp:

(WebCore::SecurityOrigin::SecurityOrigin):
(WebCore::SecurityOrigin::create):
(WebCore::SecurityOrigin::createUnique):

  • page/SecurityOrigin.h:

LayoutTests:

Update results to show progression.

  • fast/frames/sandboxed-iframe-scripting-expected.txt:
  • fast/frames/sandboxed-iframe-scripting.html:
6:30 PM Changeset in webkit [99346] by cevans@google.com
  • 1 edit
    2 copies in branches/chromium/874

Merge 98561
BUG=100526
Review URL: ​http://codereview.chromium.org/8480023

6:26 PM Changeset in webkit [99345] by cevans@google.com
  • 5 edits
    5 copies in branches/chromium/874

Merge 98033
BUG=99603
Review URL: ​http://codereview.chromium.org/8477022

6:26 PM Changeset in webkit [99344] by commit-queue@webkit.org
  • 2 edits in trunk/Source/WebCore

Potential Unused Param Build Issue
​https://bugs.webkit.org/show_bug.cgi?id=71598

Patch by Joseph Pecoraro <Joseph Pecoraro> on 2011-11-04
Reviewed by Kent Tamura.

  • html/HTMLSelectElement.cpp:

(WebCore::HTMLSelectElement::platformHandleKeydownEvent):

6:20 PM Changeset in webkit [99343] by cevans@google.com
  • 2 edits
    2 copies in branches/chromium/874

Merge 98010
BUG=100177
Review URL: ​http://codereview.chromium.org/8474016

6:00 PM Changeset in webkit [99342] by cevans@google.com
  • 1 edit
    2 copies in branches/chromium/874

Merge 98935
BUG=101018
Review URL: ​http://codereview.chromium.org/8479023

5:58 PM Changeset in webkit [99341] by eric@webkit.org
  • 2 edits in trunk/Tools

Upgrade to the latest Mechanize
​https://bugs.webkit.org/show_bug.cgi?id=71594

Reviewed by Adam Barth.

I noticed our mechanize was somewhat out of date when removing Python 2.5 support.
Mechanize is no longer two separate libraries, so importing it is much simpler.

  • Scripts/webkitpy/thirdparty/init.py:
5:58 PM Changeset in webkit [99340] by cevans@google.com
  • 1 edit in branches/chromium/874/Source/WebCore/editing/ReplaceNodeWithSpanCommand.cpp

Merge 98796
BUG=102242
Review URL: ​http://codereview.chromium.org/8479022

5:51 PM Changeset in webkit [99339] by commit-queue@webkit.org
  • 8 edits
    23 adds in trunk

Crash in ScrollAnimator.cpp
​https://bugs.webkit.org/show_bug.cgi?id=69865

The code in ScrollAnimator assumes that horizontal per-page mouse
wheel events cannot happen, which is not true. This patch adds layout
tests for all paging wheel event situations and fixes the broken
horizontal case.

Source/WebCore:

Patch by Stephen Chenney <​schenney@chromium.org> on 2011-11-04
Reviewed by Anders Carlsson

Tests: fast/events/platform-wheelevent-paging-x-in-non-scrolling-div.html

fast/events/platform-wheelevent-paging-x-in-non-scrolling-page.html
fast/events/platform-wheelevent-paging-x-in-scrolling-div.html
fast/events/platform-wheelevent-paging-x-in-scrolling-page.html
fast/events/platform-wheelevent-paging-xy-in-scrolling-div.html
fast/events/platform-wheelevent-paging-xy-in-scrolling-page.html
fast/events/platform-wheelevent-paging-y-in-non-scrolling-div.html
fast/events/platform-wheelevent-paging-y-in-non-scrolling-page.html
fast/events/platform-wheelevent-paging-y-in-scrolling-div.html
fast/events/platform-wheelevent-paging-y-in-scrolling-page.html

  • platform/PlatformWheelEvent.h: Modify the comment to reflect the new

reality.

  • platform/ScrollAnimator.cpp:

(WebCore::ScrollAnimator::handleWheelEvent): Modified the code to handle the
horizontal per-page wheel event case.

Tools:

Patch by Stephen Chenney <​schenney@chromium.org> on 2011-11-04
Reviewed by Anders Carlsson

  • DumpRenderTree/chromium/EventSender.cpp:

(EventSender::handleMouseWheel): Added the ability to specify that a
mouse wheel event should be paging.

LayoutTests:

Patch by Stephen Chenney <​schenney@chromium.org> on 2011-11-04
Reviewed by Anders Carlsson

  • fast/events/platform-wheelevent-paging-x-in-non-scrolling-div-expected.txt: Added.
  • fast/events/platform-wheelevent-paging-x-in-non-scrolling-div.html: Added.
  • fast/events/platform-wheelevent-paging-x-in-non-scrolling-page-expected.txt: Added.
  • fast/events/platform-wheelevent-paging-x-in-non-scrolling-page.html: Added.
  • fast/events/platform-wheelevent-paging-x-in-scrolling-div-expected.txt: Added.
  • fast/events/platform-wheelevent-paging-x-in-scrolling-div.html: Added.
  • fast/events/platform-wheelevent-paging-x-in-scrolling-page-expected.txt: Added.
  • fast/events/platform-wheelevent-paging-x-in-scrolling-page.html: Added.
  • fast/events/platform-wheelevent-paging-xy-in-scrolling-div-expected.txt: Added.
  • fast/events/platform-wheelevent-paging-xy-in-scrolling-div.html: Added.
  • fast/events/platform-wheelevent-paging-xy-in-scrolling-page-expected.txt: Added.
  • fast/events/platform-wheelevent-paging-xy-in-scrolling-page.html: Added.
  • fast/events/platform-wheelevent-paging-y-in-non-scrolling-div-expected.txt: Added.
  • fast/events/platform-wheelevent-paging-y-in-non-scrolling-div.html: Added.
  • fast/events/platform-wheelevent-paging-y-in-non-scrolling-page-expected.txt: Added.
  • fast/events/platform-wheelevent-paging-y-in-non-scrolling-page.html: Added.
  • fast/events/platform-wheelevent-paging-y-in-scrolling-div-expected.txt: Added.
  • fast/events/platform-wheelevent-paging-y-in-scrolling-div.html: Added.
  • fast/events/platform-wheelevent-paging-y-in-scrolling-page-expected.txt: Added.
  • fast/events/platform-wheelevent-paging-y-in-scrolling-page.html: Added.
  • platform/chromium-mac/fast/events/platform-wheelevent-paging-x-in-scrolling-page-expected.txt: Added.
  • platform/chromium-mac/fast/events/platform-wheelevent-paging-xy-in-scrolling-page-expected.txt: Added.
  • platform/chromium-mac/fast/events/platform-wheelevent-paging-y-in-scrolling-page-expected.txt: Added.
  • platform/mac/Skipped: Added skip for the new tests, as the platform

does not support it.

  • platform/win/Skipped: Added skip for the new tests, as the platform

does not support it.

5:44 PM Changeset in webkit [99338] by commit-queue@webkit.org
  • 14 edits
    1 copy
    1 add in trunk/Source/WebCore

[MutationObservers] Refactor MutationObserverRegistration into its own class that is
referenced by registration points

​https://bugs.webkit.org/show_bug.cgi?id=71577

Patch by Rafael Weinstein <​rafaelw@chromium.org> on 2011-11-04
Reviewed by Ojan Vafai.

MutationObserverRegistration is now owned by the node which is observed. If transient
registrations are created, they hold a reference to this object.

The ownership relationship between Node, MutationObserverRegistration &
WebKitMutationObserver now goes like this: WebKitMutationObserver is RefCounted -
and only script and MutationObserverRegistration take references to it. Exactly
one MutationObserverRegistration can exist for any given tuple of
<Node, WebKitMutationObserver>. This represents an observer observing at a given
Node. Further observation by the same observer has the effect of *resetting*
the existing observation. The Node owns the MutationObserverRegistration.
Transient registrations are only pointers to the registration. The registration
keeps a RefPtr to its registration node and all of its transient registration nodes
as long as *any* transient registrations exist. This ensures the registration
and all registration nodes stay alive at least until the end of the microtask.

No tests required - refactor only.

  • GNUmakefile.list.am:
  • WebCore.gypi:
  • WebCore.pro:
  • WebCore.vcproj/WebCore.vcproj:
  • WebCore.xcodeproj/project.pbxproj:
  • dom/CharacterData.cpp:

(WebCore::enqueueCharacterDataMutationRecord):

  • dom/ChildListMutationScope.cpp:

(WebCore::MutationAccumulationRouter::MutationAccumulationRouter::incrementScopingLevel):

  • dom/Element.cpp:

(WebCore::enqueueAttributesMutationRecord):

  • dom/MutationObserverRegistration.cpp: Added.

(WebCore::MutationObserverRegistration::create):
(WebCore::MutationObserverRegistration::MutationObserverRegistration):
(WebCore::MutationObserverRegistration::~MutationObserverRegistration):
(WebCore::MutationObserverRegistration::resetObservation):
(WebCore::MutationObserverRegistration::observedSubtreeNodeWillDetach):
(WebCore::MutationObserverRegistration::clearTransientRegistrations):
(WebCore::MutationObserverRegistration::unregister):
(WebCore::MutationObserverRegistration::shouldReceiveMutationFrom):

  • dom/MutationObserverRegistration.h: Copied from Source/WebCore/dom/WebKitMutationObserver.h.

(WebCore::MutationObserverRegistration::observer):
(WebCore::MutationObserverRegistration::deliveryOptions):

  • dom/Node.cpp:

(WebCore::Node::clearRareData):
(WebCore::Node::mutationObserverRegistry):
(WebCore::Node::transientMutationObserverRegistry):
(WebCore::addObserverToDeliverySet):
(WebCore::Node::collectMatchingObserversForMutation):
(WebCore::Node::getRegisteredMutationObserversOfType):
(WebCore::Node::registerMutationObserver):
(WebCore::Node::unregisterMutationObserver):
(WebCore::Node::registerTransientMutationObserver):
(WebCore::Node::unregisterTransientMutationObserver):
(WebCore::Node::notifyMutationObserversNodeWillDetach):

  • dom/Node.h:
  • dom/NodeRareData.h:

(WebCore::NodeRareData::mutationObserverRegistry):
(WebCore::NodeRareData::ensureMutationObserverRegistry):
(WebCore::NodeRareData::transientMutationObserverRegistry):
(WebCore::NodeRareData::ensureTransientMutationObserverRegistry):

  • dom/WebKitMutationObserver.cpp:

(WebCore::WebKitMutationObserver::~WebKitMutationObserver):
(WebCore::WebKitMutationObserver::observe):
(WebCore::WebKitMutationObserver::disconnect):
(WebCore::WebKitMutationObserver::observationStarted):
(WebCore::WebKitMutationObserver::observationEnded):
(WebCore::WebKitMutationObserver::deliver):

  • dom/WebKitMutationObserver.h:
5:34 PM Changeset in webkit [99337] by commit-queue@webkit.org
  • 10 edits in trunk/Source/WebCore

Add methods to compute magnitude and phase response for biquads
​https://bugs.webkit.org/show_bug.cgi?id=71055

Reviewed by Kenneth Russell.

Patch by Raymond Toy <Raymond Toy> on 2011-11-04

  • platform/audio/Biquad.cpp:

(WebCore::Biquad::getFrequencyResponse):
Computes the magnitude and phase (radians) response for the given
biquad at the specified set of (normalized) frequencies.

  • platform/audio/Biquad.h:

Declare getFrequencyResponse.

  • webaudio/BiquadDSPKernel.cpp:

(WebCore::BiquadDSPKernel::updateCoefficientsIfNecessary):
Factor out the code that updates filter coefficients. Allow the
caller to specify whether the smoothed values are used or not and
whether we do the update even if the coefficients are not dirty.
(WebCore::BiquadDSPKernel::process):
Use updateCoefficientsIfNecessary to update.
(WebCore::BiquadDSPKernel::getFrequencyResponse):
Implmentation of getFrequencyResponse.

  • webaudio/BiquadDSPKernel.h:

Declare getFrequencyResponse.

  • webaudio/BiquadFilterNode.cpp:

(WebCore::BiquadFilterNode::getFrequencyResponse):
Implementation of getFrequencyResponse

  • webaudio/BiquadFilterNode.h:

Declare getFrequencyResponse.

  • webaudio/BiquadFilterNode.idl:

Define interface to getFrequencyResponse.

  • webaudio/BiquadProcessor.cpp:

(WebCore::BiquadProcessor::checkForDirtyCoefficients):
Factor out code for checking for dirty coefficients.
(WebCore::BiquadProcessor::process):
Use checkForDirtyCoefficients.
(WebCore::BiquadProcessor::getFrequencyResponse):
Implementation of getFrequencyResponse

  • webaudio/BiquadProcessor.h:

Declare getFrequencyResponse.

5:29 PM Changeset in webkit [99336] by commit-queue@webkit.org
  • 2 edits
    2 moves
    5 adds in trunk/LayoutTests

Baselines for new test introduced in r99043
​https://bugs.webkit.org/show_bug.cgi?id=71528

Patch by Ben Wells <​benwells@chromium.org> on 2011-11-04
Reviewed by Kenneth Russell.

  • platform/chromium-cg-mac-snowleopard/fast/canvas/canvas-composite-fill-repaint-expected.png: Added.
  • platform/chromium-gpu-mac/fast/canvas/canvas-composite-fill-repaint-expected.png: Added.
  • platform/chromium-gpu-win/fast/canvas/canvas-composite-fill-repaint-expected.png: Renamed from LayoutTests/platform/chromium-gpu-linux/fast/canvas/canvas-composite-fill-repaint-expected.png.
  • platform/chromium-mac-snowleopard/fast/canvas/canvas-composite-fill-repaint-expected.png: Added.
  • platform/chromium-win/fast/canvas/canvas-composite-fill-repaint-expected.png: Renamed from LayoutTests/platform/chromium-linux/fast/canvas/canvas-composite-fill-repaint-expected.png.
  • platform/chromium/test_expectations.txt:
5:27 PM Changeset in webkit [99335] by eric@webkit.org
  • 2 edits in trunk/Tools

Remove deprecated free functions in port.factory
​https://bugs.webkit.org/show_bug.cgi?id=71494

Unreviewed. Fix exception when running new-run-webkit-httpd.

  • Scripts/new-run-webkit-httpd:
5:02 PM Changeset in webkit [99334] by commit-queue@webkit.org
  • 3 edits in trunk/Source/WebCore

[Mac] ResourceRequest's nsURLRequest() does not differentiate null and empty URLs with CFNetwork
​https://bugs.webkit.org/show_bug.cgi?id=71539

Patch by Benjamin Poulain <​bpoulain@apple.com> on 2011-11-04
Reviewed by David Kilzer.

In order to have CFURL and NSURL to be consistent when both are used on Mac,
KURL::createCFURL() is changed to support empty URL values.

  • platform/cf/KURLCFNet.cpp:

(WebCore::createCFURLFromBuffer):
(WebCore::KURL::createCFURL):

  • platform/mac/KURLMac.mm:

(WebCore::KURL::operator NSURL *):
(WebCore::KURL::createCFURL):

4:51 PM Changeset in webkit [99333] by psolanki@apple.com
  • 2 edits in trunk/Source/JavaScriptCore

sqrtDouble and andnotDouble should be declared noreturn
​https://bugs.webkit.org/show_bug.cgi?id=71592

Reviewed by Sam Weinig.

  • assembler/MacroAssemblerARMv7.h:
4:07 PM Changeset in webkit [99332] by fsamuel@chromium.org
  • 15 edits
    4 adds in trunk

CSS Aspect Ratio Property Parsing Stage
​https://bugs.webkit.org/show_bug.cgi?id=70707

Reviewed by Ojan Vafai.

Source/WebCore:

Added parsing support for -webkit-aspect-ratio CSS property.
The spec can be found here: ​http://www.xanthir.com/blog/b4810

Test: fast/css/aspect-ratio-parsing-tests.html

  • CMakeLists.txt:
  • GNUmakefile.list.am:
  • Target.pri:
  • WebCore.gypi:
  • WebCore.vcproj/WebCore.vcproj:
  • WebCore.xcodeproj/project.pbxproj:
  • css/CSSAspectRatioValue.cpp: Added.

(WebCore::CSSAspectRatioValue::~CSSAspectRatioValue):
(WebCore::CSSAspectRatioValue::cssText):

  • css/CSSAspectRatioValue.h: Added.

(WebCore::CSSAspectRatioValue::create):
(WebCore::CSSAspectRatioValue::numeratorValue):
(WebCore::CSSAspectRatioValue::denominatorValue):
(WebCore::CSSAspectRatioValue::CSSAspectRatioValue):

  • css/CSSComputedStyleDeclaration.cpp:

(WebCore::CSSComputedStyleDeclaration::getPropertyCSSValue):

  • css/CSSMutableStyleDeclaration.cpp:
  • css/CSSParser.cpp:

(WebCore::CSSParser::parseValue):
(WebCore::CSSParser::parseAspectRatio):

  • css/CSSParser.h:
  • css/CSSProperty.cpp:

(WebCore::CSSProperty::isInheritedProperty):

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

(WebCore::CSSStyleSelector::applyProperty):

LayoutTests:

  • fast/css/aspect-ratio-parsing-tests-expected.txt: Added.
  • fast/css/aspect-ratio-parsing-tests.html: Added.
3:52 PM Changeset in webkit [99331] by Adam Roben
  • 2 edits in trunk/Tools

Remove Leopard-specific code from build.webkit.org

We don't have any Leopard bots anymore.

Fixes <​http://webkit.org/b/71583> build.webkit.org configuration contains obsolete Leopard code

Reviewed by Mark Rowe.

  • BuildSlaveSupport/build.webkit.org-config/master.cfg: Removed now-unused StartATSServer

and StopATSServer steps.
(unitTestsSupported): Removed check for mac-leopard.
(TestFactory.init): Removed Leopard-specific steps.

3:50 PM Changeset in webkit [99330] by abarth@webkit.org
  • 9 edits in trunk

Delete FrameLoader::isSandboxed
​https://bugs.webkit.org/show_bug.cgi?id=71591

Reviewed by Eric Seidel.

Source/WebCore:

We should always use document->securityOrigin()->isSandboxed because
that picks up the sandbox bits that are frozen on the document rather
than the ones that could change on the Frame.

This patch starts preparing us to implement the CSP sandbox directive,
which will cause use to have document sandbox bits without any attribute.

  • bindings/ScriptControllerBase.cpp:

(WebCore::ScriptController::canExecuteScripts):

  • This call site was the only functional site left where these two could be different. This patch causes one progression and one regression. The progression is that we now correctly freeze the allow-scripts bit when a document is created, but the regression is we now allow the execution of JavaScript URLs, as noted in fast/frames/sandboxed-iframe-scripting.html. That's even more of an edge case, so I think it's a win overall.
  • loader/DocumentWriter.cpp:

(WebCore::DocumentWriter::begin):

  • I don't think this part of the change is testable. There's no time to execute script between when the bits get copied off the Frame and when they're checked, so there's no time to change them.
  • loader/FrameLoader.cpp:
    • Update FIXME comment that is now fixed.
  • loader/FrameLoader.h:
    • Remove wrong API.

LayoutTests:

This patch updates this test to check a few more cases and notes a bug
in our current implementation.

  • fast/frames/resources/sandboxed-iframe-script-dynamic.html:
  • fast/frames/sandboxed-iframe-scripting-expected.txt:
  • fast/frames/sandboxed-iframe-scripting.html:
3:32 PM Changeset in webkit [99329] by abarth@webkit.org
  • 3 edits
    4 adds in trunk

When CSP blocks mixed content, we shouldn't also warn about mixed content
​https://bugs.webkit.org/show_bug.cgi?id=71588

Reviewed by Eric Seidel.

Source/WebCore:

Getting both warnings confused some early adopters. They thought the
insecure loads were happening even though they blocked them with CSP.
This patch restricts the mixed content warnings to show only when the
load isn't already blocked by CSP or by the embedder.

Test: http/tests/security/contentSecurityPolicy/block-mixed-content-hides-warning.html

  • loader/cache/CachedResourceLoader.cpp:

(WebCore::CachedResourceLoader::canRequest):

LayoutTests:

These tests check what kinds of warnings we generate when blocking insecure scripts with CSP.

  • http/tests/security/contentSecurityPolicy/block-mixed-content-hides-warning-expected.txt: Added.
  • http/tests/security/contentSecurityPolicy/block-mixed-content-hides-warning.html: Added.
  • http/tests/security/contentSecurityPolicy/resources/alert-fail.js: Added.
  • http/tests/security/contentSecurityPolicy/resources/mixed-content-with-csp.html: Added.
3:24 PM Changeset in webkit [99328] by commit-queue@webkit.org
  • 5 edits in trunk/Source/WebCore

Allow ScriptExecutionContext::addMessage to be called from background threads.
​https://bugs.webkit.org/show_bug.cgi?id=71575

Patch by Michael Nordman <​michaeln@google.coom> on 2011-11-04
Reviewed by Nate Chapin.

No new tests.

  • dom/Document.cpp:

(WebCore::Document::addMessage):

  • dom/ScriptExecutionContext.cpp:
  • dom/ScriptExecutionContext.h:

(WebCore::ScriptExecutionContext::AddConsoleMessageTask::create):
(WebCore::ScriptExecutionContext::AddConsoleMessageTask::performTask):
(WebCore::ScriptExecutionContext::AddConsoleMessageTask::AddConsoleMessageTask):

  • workers/WorkerContext.cpp:

(WebCore::WorkerContext::addMessage):

3:16 PM Changeset in webkit [99327] by commit-queue@webkit.org
  • 16 edits in trunk/Source

[chromium] Re-named the original scissorRect to clipRect
​https://bugs.webkit.org/show_bug.cgi?id=71580

Patch by Shawn Singh <​shawnsingh@chromium.org> on 2011-11-04
Reviewed by James Robinson.

Source/WebCore:

Existing tests are updated appropriately; and no new behavior is
introduced by this patch.

  • platform/graphics/chromium/LayerChromium.cpp:

(WebCore::LayerChromium::LayerChromium):

  • platform/graphics/chromium/LayerChromium.h:

(WebCore::LayerChromium::setUsesLayerClipping):
(WebCore::LayerChromium::usesLayerClipping):
(WebCore::LayerChromium::clipRect):
(WebCore::LayerChromium::setClipRect):

  • platform/graphics/chromium/LayerRendererChromium.cpp:

(WebCore::LayerRendererChromium::drawLayersInternal):
(WebCore::LayerRendererChromium::drawLayer):

  • platform/graphics/chromium/RenderSurfaceChromium.h:

(WebCore::RenderSurfaceChromium::clipRect):
(WebCore::RenderSurfaceChromium::setClipRect):

  • platform/graphics/chromium/cc/CCLayerImpl.cpp:

(WebCore::CCLayerImpl::CCLayerImpl):

  • platform/graphics/chromium/cc/CCLayerImpl.h:

(WebCore::CCLayerImpl::setUsesLayerClipping):
(WebCore::CCLayerImpl::usesLayerClipping):
(WebCore::CCLayerImpl::clipRect):
(WebCore::CCLayerImpl::setClipRect):

  • platform/graphics/chromium/cc/CCLayerTreeHost.cpp:

(WebCore::CCLayerTreeHost::updateLayers):

  • platform/graphics/chromium/cc/CCLayerTreeHostCommon.cpp:

(WebCore::layerShouldBeSkipped):
(WebCore::calculateDrawTransformsAndVisibilityInternal):

  • platform/graphics/chromium/cc/CCLayerTreeHostCommon.h:

(WebCore::CCLayerTreeHostCommon::calculateVisibleLayerRect):

  • platform/graphics/chromium/cc/CCRenderSurface.cpp:

(WebCore::CCRenderSurface::draw):

  • platform/graphics/chromium/cc/CCRenderSurface.h:

(WebCore::CCRenderSurface::setClipRect):
(WebCore::CCRenderSurface::clipRect):

Source/WebKit/chromium:

  • tests/CCLayerImplTest.cpp:

(WebCore::TEST):

  • tests/CCLayerTreeHostCommonTest.cpp:

(WebCore::TEST):

  • tests/LayerChromiumTest.cpp:
3:09 PM Changeset in webkit [99326] by eric@webkit.org
  • 7 edits in trunk/Tools

new-run-webkit-tests autoinstalls python-irclib even though it doesn't need to
​https://bugs.webkit.org/show_bug.cgi?id=71549

Reviewed by Adam Barth.

Several members of Host didn't really belong there
and were instead specific to WebKitPatch, so I've moved
them up onto that class.

  • Scripts/webkitpy/common/host.py:
  • Scripts/webkitpy/common/host_mock.py:
  • Scripts/webkitpy/tool/main.py:
  • Scripts/webkitpy/tool/mocktool.py:
  • Scripts/webkitpy/tool/steps/runtests_unittest.py:
3:01 PM Changeset in webkit [99325] by tony@chromium.org
  • 6 edits
    1 move in trunk/Tools

garden-o-matic: bring back party time!
​https://bugs.webkit.org/show_bug.cgi?id=71582

Reviewed by Adam Barth.

  • BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html: Allow images from file:
  • BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/images/partytime.gif: Renamed from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/partytime.gif.
  • BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base.js: Add length() for UpdateTracker.
  • BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js: Plumb through length().
  • BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js: Toggle partytime.gif via css class.
  • BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/onebar.css: Add css styling for partytime.
2:43 PM Changeset in webkit [99324] by dslomov@google.com
  • 8 edits in trunk/Source/WebCore

Add the ability to transfer ArrayBuffer and "neuter" it.
​https://bugs.webkit.org/show_bug.cgi?id=71535

Reviewed by David Levin.

  • html/canvas/ArrayBuffer.cpp:

(WebCore::ArrayBuffer::create):
(WebCore::ArrayBuffer::ArrayBuffer):
(WebCore::ArrayBuffer::data):
(WebCore::ArrayBuffer::byteLength):
(WebCore::ArrayBuffer::transfer):
(WebCore::ArrayBufferContents::~ArrayBufferContents):
(WebCore::ArrayBufferContents::tryAllocate):
(WebCore::ArrayBuffer::addView):
(WebCore::ArrayBuffer::removeView):

  • html/canvas/ArrayBuffer.h:

(WebCore::ArrayBufferContents::ArrayBufferContents):
(WebCore::ArrayBufferContents::data):
(WebCore::ArrayBufferContents::sizeInBytes):
(WebCore::ArrayBufferContents::release):
(WebCore::ArrayBuffer::~ArrayBuffer):

  • html/canvas/ArrayBufferView.cpp:

(WebCore::ArrayBufferView::ArrayBufferView):
(WebCore::ArrayBufferView::~ArrayBufferView):
(WebCore::ArrayBufferView::neuter):

  • html/canvas/ArrayBufferView.h:
  • html/canvas/DataView.cpp:

(WebCore::DataView::neuter):
(WebCore::DataView::neuterBinding):

  • html/canvas/DataView.h:
  • html/canvas/TypedArrayBase.h:

(WebCore::TypedArrayBase::neuter):
(WebCore::TypedArrayBase::neuterBinding):

2:42 PM Changeset in webkit [99323] by commit-queue@webkit.org
  • 2 edits
    5 copies
    6 moves
    2 adds in trunk/LayoutTests

Unreviewed, rolling out r99315.
​http://trac.webkit.org/changeset/99315
​https://bugs.webkit.org/show_bug.cgi?id=71589

baselines still broken after running rebaseline script
(Requested by enne on #webkit).

Patch by Sheriff Bot <​webkit.review.bot@gmail.com> on 2011-11-04

  • platform/chromium-cg-mac/compositing/scaling/tiled-layer-recursion-expected.png: Renamed from LayoutTests/platform/chromium-cg-mac-snowleopard/compositing/scaling/tiled-layer-recursion-expected.png.
  • platform/chromium-gpu-cg-mac/compositing/geometry/fixed-position-expected.png: Copied from LayoutTests/platform/chromium-gpu/compositing/geometry/fixed-position-expected.png.
  • platform/chromium-gpu-cg-mac/compositing/geometry/horizontal-scroll-composited-expected.png: Copied from LayoutTests/platform/chromium-gpu/compositing/geometry/horizontal-scroll-composited-expected.png.
  • platform/chromium-gpu-cg-mac/compositing/geometry/vertical-scroll-composited-expected.png: Copied from LayoutTests/platform/chromium-gpu/compositing/geometry/vertical-scroll-composited-expected.png.
  • platform/chromium-gpu-cg-mac/compositing/overflow/fixed-position-ancestor-clip-expected.png: Copied from LayoutTests/platform/chromium-gpu/compositing/overflow/fixed-position-ancestor-clip-expected.png.
  • platform/chromium-gpu-cg-mac/compositing/scaling/tiled-layer-recursion-expected.png: Copied from LayoutTests/platform/chromium-gpu/compositing/scaling/tiled-layer-recursion-expected.png.
  • platform/chromium-gpu-mac/compositing/geometry/fixed-position-expected.png: Renamed from LayoutTests/platform/chromium-gpu/compositing/geometry/fixed-position-expected.png.
  • platform/chromium-gpu-mac/compositing/geometry/horizontal-scroll-composited-expected.png: Renamed from LayoutTests/platform/chromium-gpu/compositing/geometry/horizontal-scroll-composited-expected.png.
  • platform/chromium-gpu-mac/compositing/geometry/vertical-scroll-composited-expected.png: Renamed from LayoutTests/platform/chromium-gpu/compositing/geometry/vertical-scroll-composited-expected.png.
  • platform/chromium-gpu-mac/compositing/overflow/fixed-position-ancestor-clip-expected.png: Renamed from LayoutTests/platform/chromium-gpu/compositing/overflow/fixed-position-ancestor-clip-expected.png.
  • platform/chromium-gpu-mac/compositing/scaling/tiled-layer-recursion-expected.png: Renamed from LayoutTests/platform/chromium-gpu/compositing/scaling/tiled-layer-recursion-expected.png.
  • platform/chromium-gpu-win/compositing/geometry/horizontal-scroll-composited-expected.txt: Added.
  • platform/chromium-gpu-win/compositing/geometry/vertical-scroll-composited-expected.txt: Added.
  • platform/chromium/test_expectations.txt:
2:39 PM Changeset in webkit [99322] by Philippe Normand
  • 4 edits in trunk/LayoutTests

Unreviewed, GTK rebaseline after r99285.

  • platform/gtk/Skipped:
  • platform/gtk/fast/dom/prototype-inheritance-2-expected.txt:
  • platform/gtk/fast/js/global-constructors-expected.txt:
2:32 PM Changeset in webkit [99321] by abarth@webkit.org
  • 1 edit
    2 adds in trunk/LayoutTests

Add a test for how CSP interacts with about:blank iframes
​https://bugs.webkit.org/show_bug.cgi?id=71578

Reviewed by Dimitri Glazkov.

This issue came up in the discussion on
​http://code.google.com/p/chromium/issues/detail?id=103038. This patch
just adds a test to make sure we don't regress.

  • http/tests/security/contentSecurityPolicy/frame-src-about-blank-allowed-by-scheme-expected.txt: Added.
  • http/tests/security/contentSecurityPolicy/frame-src-about-blank-allowed-by-scheme.html: Added.
2:27 PM Changeset in webkit [99320] by tony@chromium.org
  • 2 edits in trunk/LayoutTests

stop running perf LayoutTests in debug builds
​https://bugs.webkit.org/show_bug.cgi?id=71513

Reviewed by Ojan Vafai.

This allows us to get rid of SLOW on some perf tests.

  • platform/chromium/test_expectations.txt:
2:14 PM Changeset in webkit [99319] by noel.gordon@gmail.com
  • 10 edits
    4 adds in trunk

[Chromium] Implement canvas.toDataURL("image/webp")
​https://bugs.webkit.org/show_bug.cgi?id=70172

Reviewed by Adam Barth.

Source/WebCore:

Add a webp image encoder for the chromium skia ports, add it to the canvas
element toDataURL() flow with support for an optional quality [0.0-1.0].
Use a webp compression method (3) that provides good speed and compression
performance for this image format, and works well for various input image
types: line-art, screen-shots, and photographic images.

Test: fast/canvas/canvas-toDataURL-webp.html

  • WebCore.gypi:
  • platform/chromium/MIMETypeRegistryChromium.cpp:

(WebCore::MIMETypeRegistry::isSupportedImageMIMETypeForEncoding):

  • platform/graphics/skia/ImageBufferSkia.cpp:

(WebCore::ImageToDataURL):

  • platform/image-encoders/skia/WEBPImageEncoder.cpp: Added.

(WebCore::writeOutput):
(WebCore::importPicture):
(WebCore::importPictureBGRA):
(WebCore::importPictureRGBA):
(WebCore::encodePixels):
(WebCore::WEBPImageEncoder::encode):

  • platform/image-encoders/skia/WEBPImageEncoder.h: Added.

LayoutTests:

  • fast/canvas/canvas-toDataURL-webp-expected.txt: Added.
  • fast/canvas/canvas-toDataURL-webp.html: Added.
  • platform/chromium/test_expectations.txt: skip test on MAC CPU-CG (CoreGraphics)
  • platform/gtk/Skipped:
  • platform/mac/Skipped:
  • platform/qt/Skipped:
  • platform/win/Skipped:
2:04 PM Changeset in webkit [99318] by commit-queue@webkit.org
  • 5 edits in trunk

[EFL] DRT: Fix the order of the items in the dumped history list.
​https://bugs.webkit.org/show_bug.cgi?id=71562

It did not make much sense to hold the history list in a map, as we do not need
the keys. And the printed items must be sorted by target, which is more easily
done with a vector.

Should make tests such as http/tests/navigation/error404-frames.html pass.

Patch by Raphael Kubo da Costa <​kubo@profusion.mobi> on 2011-11-04
Reviewed by Antonio Gomes.

Source/WebKit/efl:

  • WebCoreSupport/DumpRenderTreeSupportEfl.cpp:

(DumpRenderTreeSupportEfl::childHistoryItems):

  • WebCoreSupport/DumpRenderTreeSupportEfl.h:

Tools:

  • DumpRenderTree/efl/DumpHistoryItem.cpp:

(compareHistoryItemsByTarget):
(dumpHistoryItem):

2:00 PM Changeset in webkit [99317] by Patrick Gansterer
  • 5 edits in trunk

[Qt] Remove ENABLE_SQLITE from qmake files
​https://bugs.webkit.org/show_bug.cgi?id=71546

Reviewed by Simon Hausmann.

Source/WebCore:

ENABLE_SQLITE is required to build WebCore, so remove conditions.

  • Target.pri:

Tools:

  • qmake/mkspecs/features/features.prf:
  • qmake/mkspecs/features/webcore.prf:
1:58 PM Changeset in webkit [99316] by arv@chromium.org
  • 3 edits in trunk/LayoutTests

Layout Test storage/indexeddb/exception-in-event-aborts.html is failing
​https://bugs.webkit.org/show_bug.cgi?id=71536

Unreviewed. Updated test and expectations.

  • storage/indexeddb/exception-in-event-aborts-expected.txt:
  • storage/indexeddb/exception-in-event-aborts.html:
1:55 PM Changeset in webkit [99315] by enne@google.com
  • 2 edits
    6 moves
    3 adds
    7 deletes in trunk/LayoutTests

[chromium] Unreviewed expectation rebaseline from r99286.

  • platform/chromium-cg-mac-snowleopard/compositing/scaling/tiled-layer-recursion-expected.png: Renamed from LayoutTests/platform/chromium-cg-mac/compositing/scaling/tiled-layer-recursion-expected.png.
  • platform/chromium-gpu-mac/compositing/geometry/fixed-position-expected.png: Removed.
  • platform/chromium-gpu-mac/compositing/geometry/horizontal-scroll-composited-expected.png: Removed.
  • platform/chromium-gpu-mac/compositing/geometry/vertical-scroll-composited-expected.png: Removed.
  • platform/chromium-gpu-mac/compositing/overflow/fixed-position-ancestor-clip-expected.png: Removed.
  • platform/chromium-gpu-mac/compositing/scaling/tiled-layer-recursion-expected.png: Removed.
  • platform/chromium-gpu-win/compositing/geometry/horizontal-scroll-composited-expected.txt: Removed.
  • platform/chromium-gpu-win/compositing/geometry/vertical-scroll-composited-expected.txt: Removed.
  • platform/chromium-gpu/compositing/geometry/fixed-position-expected.png: Renamed from LayoutTests/platform/chromium-gpu-cg-mac/compositing/geometry/fixed-position-expected.png.
  • platform/chromium-gpu/compositing/geometry/horizontal-scroll-composited-expected.png: Renamed from LayoutTests/platform/chromium-gpu-cg-mac/compositing/geometry/horizontal-scroll-composited-expected.png.
  • platform/chromium-gpu/compositing/geometry/vertical-scroll-composited-expected.png: Renamed from LayoutTests/platform/chromium-gpu-cg-mac/compositing/geometry/vertical-scroll-composited-expected.png.
  • platform/chromium-gpu/compositing/overflow/fixed-position-ancestor-clip-expected.png: Renamed from LayoutTests/platform/chromium-gpu-cg-mac/compositing/overflow/fixed-position-ancestor-clip-expected.png.
  • platform/chromium-gpu/compositing/scaling/tiled-layer-recursion-expected.png: Renamed from LayoutTests/platform/chromium-gpu-cg-mac/compositing/scaling/tiled-layer-recursion-expected.png.
  • platform/chromium/test_expectations.txt:
1:48 PM Changeset in webkit [99314] by abarth@webkit.org
  • 2 edits in trunk/Tools

watchlist for WebIDL shouldn't trigger on Internals.idl
​https://bugs.webkit.org/show_bug.cgi?id=70657

Reviewed by David Levin.

Internals.idl isn't part of the Web-facing IDL in the project.

  • Scripts/webkitpy/common/config/watchlist:
1:43 PM Changeset in webkit [99313] by tony@chromium.org
  • 2 edits in trunk/LayoutTests

[chromium] comment out duplicate suppressions for compositing/zoom-animator-scale-test2.html
​https://bugs.webkit.org/show_bug.cgi?id=71587

  • platform/chromium/test_expectations.txt:
1:37 PM Changeset in webkit [99312] by mhahnenberg@apple.com
  • 15 edits in trunk/Source/JavaScriptCore

De-virtualize JSObject::hasInstance
​https://bugs.webkit.org/show_bug.cgi?id=71430

Reviewed by Darin Adler.

Added hasInstance to the MethodTable, changed all the virtual
implementations of hasInstance to static ones, and replaced
all call sites with corresponding lookups in the MethodTable.

  • API/JSCallbackObject.h:
  • API/JSCallbackObjectFunctions.h:

(JSC::::hasInstance):

  • API/JSValueRef.cpp:

(JSValueIsInstanceOfConstructor):

(JSC::Interpreter::privateExecute):

  • jit/JITStubs.cpp:

(JSC::DEFINE_STUB_FUNCTION):

  • runtime/ClassInfo.h:
  • runtime/JSBoundFunction.cpp:

(JSC::JSBoundFunction::hasInstance):

  • runtime/JSBoundFunction.h:
  • runtime/JSCell.cpp:

(JSC::JSCell::hasInstance):

  • runtime/JSCell.h:
  • runtime/JSObject.cpp:

(JSC::JSObject::hasInstance):

  • runtime/JSObject.h:
1:24 PM Changeset in webkit [99311] by Nate Chapin
  • 3 edits in trunk/Source/WebCore

[V8] Null out V8NPObject::rootObject when the owning
Frame goes away.
​https://bugs.webkit.org/show_bug.cgi?id=71569

Reviewed by Adam Barth.

No new tests, this has only been successfully triggered
using ppapi flash.

  • bindings/v8/V8Helpers.cpp:

(WebCore::toV8Context):

  • bindings/v8/npruntime.cpp:
12:53 PM Changeset in webkit [99310] by commit-queue@webkit.org
  • 4 edits
    21 adds in trunk

Send the submissions character encoding in hidden _charset_ field.
​https://bugs.webkit.org/show_bug.cgi?id=19079

Patch by Vineet Chaudhary <​vineet.chaudhary@motorola.com> on 2011-11-04
Reviewed by Darin Adler.

Source/WebCore:

Spec reference ​http://dev.w3.org/html5/spec/Overview.html#attr-fe-name-charset
If _charset_ used as the name of a hidden control it should send character encoding
as value while submissions.

Tests: http/tests/misc/char-encoding-in-hidden-charset-field-default.html

http/tests/misc/char-encoding-in-hidden-charset-field-with-Big5.html
http/tests/misc/char-encoding-in-hidden-charset-field-with-EUC-JP.html
http/tests/misc/char-encoding-in-hidden-charset-field-with-ISO-2022-JP.html
http/tests/misc/char-encoding-in-hidden-charset-field-with-Shift_JIS.html
http/tests/misc/char-encoding-in-hidden-charset-field-with-accept-charset.html
http/tests/misc/char-encoding-in-hidden-charset-field-with-get-method.html
http/tests/misc/char-encoding-in-hidden-charset-field-with-one-field.html
http/tests/misc/char-encoding-in-text-charset-field-with-value.html
http/tests/misc/char-encoding-without-charset-field.html

  • html/HiddenInputType.cpp:

(WebCore::HiddenInputType::isHiddenType):
(WebCore::HiddenInputType::appendFormData): For _charset_ send value as respective encodingType.

  • html/HiddenInputType.h: Override appendFormData for hidden control types.

LayoutTests:

Spec reference ​http://dev.w3.org/html5/spec/Overview.html#attr-fe-name-charset
If _charset_ used as the name of a hidden control it should send character encoding
as value while submissions.
Added test cases to check this behavior.

  • http/tests/misc/char-encoding-in-hidden-charset-field-default-expected.txt: Added.
  • http/tests/misc/char-encoding-in-hidden-charset-field-default.html: Added. For Default Encoding.
  • http/tests/misc/char-encoding-in-hidden-charset-field-with-Big5-expected.txt: Added.
  • http/tests/misc/char-encoding-in-hidden-charset-field-with-Big5.html: Added. For Big5 Encoding.
  • http/tests/misc/char-encoding-in-hidden-charset-field-with-EUC-JP-expected.txt: Added.
  • http/tests/misc/char-encoding-in-hidden-charset-field-with-EUC-JP.html: Added. For EUC-JP Encoding.
  • http/tests/misc/char-encoding-in-hidden-charset-field-with-ISO-2022-JP-expected.txt: Added.
  • http/tests/misc/char-encoding-in-hidden-charset-field-with-ISO-2022-JP.html: Added. For ISO-2022-JP Encoding.
  • http/tests/misc/char-encoding-in-hidden-charset-field-with-Shift_JIS-expected.txt: Added.
  • http/tests/misc/char-encoding-in-hidden-charset-field-with-Shift_JIS.html: Added. For Shift_JIS Encoding.
  • http/tests/misc/char-encoding-in-hidden-charset-field-with-accept-charset-expected.txt: Added.
  • http/tests/misc/char-encoding-in-hidden-charset-field-with-accept-charset.html: Added. For two input fields.
  • http/tests/misc/char-encoding-in-hidden-charset-field-with-get-method-expected.txt: Added.
  • http/tests/misc/char-encoding-in-hidden-charset-field-with-get-method.html: Added. For GET method.
  • http/tests/misc/char-encoding-in-hidden-charset-field-with-one-field-expected.txt: Added.
  • http/tests/misc/char-encoding-in-hidden-charset-field-with-one-field.html: Added. For one input fields.
  • http/tests/misc/char-encoding-in-text-charset-field-with-value-expected.txt: Added.
  • http/tests/misc/char-encoding-in-text-charset-field-with-value.html: Added. If input type is text.
  • http/tests/misc/char-encoding-without-charset-field-expected.txt: Added.
  • http/tests/misc/char-encoding-without-charset-field.html: Added. If no hidden input field.
  • http/tests/misc/resources/char-encoding-in-hidden-charset-field.php: Added. Server side script for test.
12:29 PM Changeset in webkit [99309] by pfeldman@chromium.org
  • 3 edits in trunk/LayoutTests

2011-11-04 Pavel Feldman <​pfeldman@chromium.org>

Not reviewed: marked inspector/debugger/script-formatter.html
as skipped for now - sorting it out.

  • platform/mac/Skipped:
  • platform/win/Skipped:
12:12 PM Changeset in webkit [99308] by commit-queue@webkit.org
  • 4 edits in trunk/Source

[chromium] Use Chromium's copy of libWebKitSystemInterfaceLeopard.a
​https://bugs.webkit.org/show_bug.cgi?id=71563

Patch by Robert Sesek <​rsesek@chromium.org> on 2011-11-04
Reviewed by Adam Barth.

Source/WebCore:

Use libWebKitSystemInterfaceLeopard.a that's in the Chromium port,
rather in WebKitLibraries/.

  • WebCore.gyp/WebCore.gyp:

Source/WebKit/chromium:

Roll Chromium DEPS r108107:108476 to pull in
libWebKitSystemInterfaceLeopard.a.

  • DEPS:
11:52 AM Changeset in webkit [99307] by jchaffraix@webkit.org
  • 2 edits in trunk/Source/WebCore

RenderLayer::styleChanged invalidates the GraphicsLayer needlessly
​https://bugs.webkit.org/show_bug.cgi?id=71517

Reviewed by Simon Fraser.

Unused code removal covered by existing tests.

The current code is invalidating the whole RenderLayerBacking as part of
a style change. However this is redundant with the repainting logic which
would invalidate only the necessary bits.

  • rendering/RenderLayer.cpp:

(WebCore::RenderLayer::styleChanged):
Remove the invalidation code as it duplicates and defeats the repaint logic.

11:38 AM Changeset in webkit [99306] by mihnea@adobe.com
  • 3 edits
    2 adds in trunk

Correct parsing of incomplete @-webkit-region rules.
​https://bugs.webkit.org/show_bug.cgi?id=71514

Reviewed by Adam Barth.

Source/WebCore:

Test: fast/regions/parse-incomplete-region-rule.html

Make sure to set the region rule to 0 on the fail branch.

  • css/CSSGrammar.y:

LayoutTests:

  • fast/regions/parse-incomplete-region-rule-expected.txt: Added.
  • fast/regions/parse-incomplete-region-rule.html: Added.
11:34 AM Changeset in webkit [99305] by adamk@chromium.org
  • 4 edits in trunk/Source/WebCore

Reduce refcount churn in ChildListMutationScope
​https://bugs.webkit.org/show_bug.cgi?id=71527

Reviewed by Ryosuke Niwa.

Only the ChildListMutationAccumulator needs to hold a reference to the
target node, and the added/removed children only need be ref'd if a
ChildListMutationAccumulator exists (i.e., if there are interested
observers).

  • dom/ChildListMutationScope.cpp:

(WebCore::MutationAccumulationRouter::ChildListMutationAccumulator::enqueueMutationRecord):
(WebCore::MutationAccumulationRouter::MutationAccumulationRouter::childAdded):
(WebCore::MutationAccumulationRouter::MutationAccumulationRouter::willRemoveChild):
(WebCore::ChildListMutationScope::ChildListMutationScope):
(WebCore::ChildListMutationScope::~ChildListMutationScope):
(WebCore::ChildListMutationScope::childAdded):
(WebCore::ChildListMutationScope::willRemoveChild):

  • dom/ChildListMutationScope.h:
  • dom/ContainerNode.cpp:

(WebCore::dispatchChildInsertionEvents):
(WebCore::dispatchChildRemovalEvents):

11:28 AM Changeset in webkit [99304] by Simon Hausmann
  • 2 edits in trunk/LayoutTests

[Qt] Unreviewed gardening.

  • platform/qt/Skipped: Forgot to skip this test also in r99296.
11:27 AM Changeset in webkit [99303] by leviw@chromium.org
  • 4 edits in trunk/Source/WebCore

Correct usage of LayoutUnits in RenderImage and SVGImage
​https://bugs.webkit.org/show_bug.cgi?id=71525

Reviewed by Darin Adler.

Correcting incorrect usage of LayoutUnits for image sizes that should be integers and fixing overloaded functions that
should be using LayoutUnits.

No new tests -- no change in behavior.

  • rendering/RenderImage.cpp:

(WebCore::RenderImage::imageDimensionsChanged):
(WebCore::RenderImage::paintIntoRect):
(WebCore::RenderImage::nodeAtPoint):
(WebCore::RenderImage::computeReplacedLogicalWidth):

  • rendering/RenderImage.h:
  • svg/graphics/SVGImage.cpp:

(WebCore::SVGImageChromeClient::invalidateContentsAndWindow):

11:20 AM Changeset in webkit [99302] by andersca@apple.com
  • 2 edits in trunk/Tools

Fix WKTR crash when running NPN_ConvertPoint test.

Call notifyDone in a timeout - otherwise we'll end up entering layout from layout.

  • DumpRenderTree/TestNetscapePlugIn/Tests/mac/ConvertPoint.cpp:

(ConvertPoint::NPP_SetWindow):

11:14 AM Changeset in webkit [99301] by abarth@webkit.org
  • 9 edits in trunk

Implement allow-popups for iframe@sandbox
​https://bugs.webkit.org/show_bug.cgi?id=66505

Reviewed by Darin Adler.

Source/WebCore:

Enable allow-popups sandbox bit. My previous patch attempted to
forward the sandbox bits during setOpener, but that interacted poorly
with out (internal) sandboxing of the initial document. This patch
forwards the bits during createWindow, which is more targeted to the
case when a sandbox frame actually creates a new window.

  • loader/FrameLoader.cpp:

(WebCore::createWindow):

  • loader/PolicyChecker.cpp:

(WebCore::PolicyChecker::checkNewWindowPolicy):

LayoutTests:

Update test results to show that we now pass.

  • http/tests/security/popup-allowed-by-sandbox-is-sandboxed-control-expected.txt:
  • http/tests/security/popup-allowed-by-sandbox-is-sandboxed-control.html:
  • http/tests/security/popup-allowed-by-sandbox-is-sandboxed.html:
  • http/tests/security/popup-allowed-by-sandbox-when-allowed-expected.txt:
  • http/tests/security/popup-allowed-by-sandbox-when-allowed.html:
11:12 AM Changeset in webkit [99300] by andersca@apple.com
  • 9 edits in trunk

NPN_ConvertPoint plug-in test should be called from NPP_SetWindow
​https://bugs.webkit.org/show_bug.cgi?id=71570

Reviewed by Simon Fraser.

Tools:

  • DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:

(PluginTest::NPP_SetWindow):

  • DumpRenderTree/TestNetscapePlugIn/PluginTest.h:

Remove the NPP parameter from NPP_SetWindow.

  • DumpRenderTree/TestNetscapePlugIn/Tests/NPPSetWindowCalledDuringDestruction.cpp:

(NPPSetWindowCalledDuringDestruction::NPP_SetWindow):
Remove the NPP parameter from NPP_SetWindow.

  • DumpRenderTree/TestNetscapePlugIn/Tests/PassDifferentNPPStruct.cpp:

(PassDifferentNPPStruct::NPP_SetWindow):
Remove the NPP parameter from NPP_SetWindow and use PluginTest::log for logging.

  • DumpRenderTree/TestNetscapePlugIn/Tests/mac/ConvertPoint.cpp:

(ConvertPoint::NPP_SetWindow):
Run the test from NPP_SetWindow and call notifyDone() when done.

  • DumpRenderTree/TestNetscapePlugIn/main.cpp:

(NPP_SetWindow):
Remove unnecessary parameter.

LayoutTests:

Add a call to layoutTestController.waitUntilDone(); the plug-in will call notifyDone() when it's done.

  • platform/mac/plugins/convert-point.html:
11:11 AM Changeset in webkit [99299] by commit-queue@webkit.org
  • 4 edits in trunk/Source/WebCore

Remove initMediaStreamEvent method
​https://bugs.webkit.org/show_bug.cgi?id=71344

This method has been removed from the spec draft.
​http://www.whatwg.org/specs/web-apps/current-work/multipage/video-conferencing-and-peer-to-peer-communication.html#mediastreamevent

Patch by Kaustubh Atrawalkar <Kaustubh Atrawalkar> on 2011-11-04
Reviewed by Adam Barth.

No new tests. Removed method.

  • mediastream/MediaStreamEvent.cpp:
  • mediastream/MediaStreamEvent.h:
  • mediastream/MediaStreamEvent.idl:
11:09 AM Changeset in webkit [99298] by abarth@webkit.org
  • 3 edits
    2 adds in trunk

Anonymous CORS fetch for WebGL texture fails when there is no appropriate server response even for the same origin requests
​https://bugs.webkit.org/show_bug.cgi?id=71053

Reviewed by Darin Adler.

Source/WebCore:

The crossorigin attribute should behave like XMLHttpRequest:
same-origin images pass without and CORS headers, but CORS checks are
performed for cross-origin loads. This patch better aligns our
behavior with Firefox, as discussed in the bug.

Test: http/tests/security/img-crossorigin-loads-same-origin.html

  • loader/ImageLoader.cpp:

(WebCore::ImageLoader::notifyFinished):

LayoutTests:

Test that we succeed in loading a same-origin image without the help of
CORS even if the image has the crossorigin attribute.

  • http/tests/security/img-crossorigin-loads-same-origin-expected.txt: Added.
  • http/tests/security/img-crossorigin-loads-same-origin.html: Added.
10:55 AM Changeset in webkit [99297] by commit-queue@webkit.org
  • 3 edits
    3 adds in trunk

HTMLKeygenElement: Fix "keytype" handling in appendFormData.
​https://bugs.webkit.org/show_bug.cgi?id=70617

This fixes a bug introduced by ​http://trac.webkit.org/changeset/97658
which causes all HTML Keygen elements with a keytype optional attribute
to be considered as unsupported.

A regression test is included.

Patch by Gaurav Shah <​gauravsh@chromium.org> on 2011-11-04
Reviewed by Darin Adler.

Source/WebCore:

Test: http/tests/misc/submit-post-keygen.html

  • html/HTMLKeygenElement.cpp:

LayoutTests:

  • http/tests/misc/resources/check-keygen-post.php: Added.
  • http/tests/misc/submit-post-keygen-expected.txt: Added.
  • http/tests/misc/submit-post-keygen.html: Added.
10:51 AM Changeset in webkit [99296] by Simon Hausmann
  • 2 edits in trunk/LayoutTests

[Qt] Temporarily skip failing tests after the build system refactoring.

Rubber-stamped by Ossy.

Bug #71571 tracks the failure so we can look into it next week.

  • platform/qt/Skipped:
10:45 AM Changeset in webkit [99295] by enne@google.com
  • 11 edits in trunk/Source

[chromium] Implement checkerboarding for missing layer tiles
​https://bugs.webkit.org/show_bug.cgi?id=69585

Reviewed by James Robinson.

Source/WebCore:

For tiles that have no texture or haven't even been created yet,
draw using the background color of the layer.

The only background color set is currently the non-composited content,
and all other layers use transparent black by default.

  • platform/graphics/chromium/LayerChromium.cpp:

(WebCore::LayerChromium::pushPropertiesTo):

  • platform/graphics/chromium/NonCompositedContentHost.cpp:

(WebCore::NonCompositedContentHost::setBackgroundColor):

  • platform/graphics/chromium/NonCompositedContentHost.h:
  • platform/graphics/chromium/cc/CCLayerImpl.cpp:

(WebCore::CCLayerImpl::setBackgroundColor):

  • platform/graphics/chromium/cc/CCLayerImpl.h:

(WebCore::CCLayerImpl::backgroundColor):

  • platform/graphics/chromium/cc/CCTiledLayerImpl.cpp:

(WebCore::CCTiledLayerImpl::drawTiles):

Source/WebKit/chromium:

To pipe the background color of the document to the non-composited
host, set it during paint. This may be more frequent than necessary,
but it'll catch all style changes.

  • src/WebViewImpl.cpp:

(WebKit::WebViewImpl::nonCompositedContentHost):
(WebKit::WebViewImplContentPainter::paint):

  • src/WebViewImpl.h:
  • tests/CCLayerImplTest.cpp:

(WebCore::TEST):

10:41 AM Changeset in webkit [99294] by Simon Hausmann
  • 2 edits in trunk/Tools

[Qt] Fix run-qtwebkit-tests execution on the build bot(s).

Rubber-stamped by Tor Arne Vestbø.

  • BuildSlaveSupport/build.webkit.org-config/master.cfg: The path to the tests

changed slightly after the refactoring.

10:37 AM Changeset in webkit [99293] by thakis@chromium.org
  • 4 edits in trunk/Source/WebCore

[chromium] Remove most exit time destructors
​https://bugs.webkit.org/show_bug.cgi?id=71524

Do this by using the DEFINE_STATIC_LOCAL macro, which allocates things on the heaps and then leaks them.

Reviewed by Dimitri Glazkov.

No behavior change, so no new tests.

  • platform/chromium/CursorChromium.cpp:

(WebCore::pointerCursor):
(WebCore::crossCursor):
(WebCore::handCursor):
(WebCore::iBeamCursor):
(WebCore::waitCursor):
(WebCore::helpCursor):
(WebCore::eastResizeCursor):
(WebCore::northResizeCursor):
(WebCore::northEastResizeCursor):
(WebCore::northWestResizeCursor):
(WebCore::southResizeCursor):
(WebCore::southEastResizeCursor):
(WebCore::southWestResizeCursor):
(WebCore::westResizeCursor):
(WebCore::northSouthResizeCursor):
(WebCore::eastWestResizeCursor):
(WebCore::northEastSouthWestResizeCursor):
(WebCore::northWestSouthEastResizeCursor):
(WebCore::columnResizeCursor):
(WebCore::rowResizeCursor):
(WebCore::middlePanningCursor):
(WebCore::eastPanningCursor):
(WebCore::northPanningCursor):
(WebCore::northEastPanningCursor):
(WebCore::northWestPanningCursor):
(WebCore::southPanningCursor):
(WebCore::southEastPanningCursor):
(WebCore::southWestPanningCursor):
(WebCore::westPanningCursor):
(WebCore::moveCursor):
(WebCore::verticalTextCursor):
(WebCore::cellCursor):
(WebCore::contextMenuCursor):
(WebCore::aliasCursor):
(WebCore::progressCursor):
(WebCore::noDropCursor):
(WebCore::copyCursor):
(WebCore::noneCursor):
(WebCore::notAllowedCursor):
(WebCore::zoomInCursor):
(WebCore::zoomOutCursor):
(WebCore::grabCursor):
(WebCore::grabbingCursor):

  • platform/chromium/LanguageChromium.cpp:

(WebCore::platformDefaultLanguage):

  • platform/chromium/MIMETypeRegistryChromium.cpp:

(WebCore::dummyHashSet):

10:29 AM Changeset in webkit [99292] by enne@google.com
  • 2 edits in trunk/LayoutTests

[chromium] Unreviewed expectations update.

Temporarily mark a few tests as failing from r99286 due to tiny
differences on scrollbars. Can be rebaselined once all bots cycle.

  • platform/chromium/test_expectations.txt:
10:10 AM Changeset in webkit [99291] by fsamuel@chromium.org
  • 4 edits in trunk/Source/WebKit/chromium

[Chromium] Expose deviceScaleFactor to WebKit API
​https://bugs.webkit.org/show_bug.cgi?id=70614

Reviewed by Darin Fisher.

This exposes the device scale factor to the Chromium-WebKit API.
This scale factor is the ratio between the current device's DPI
and the target device's DPI.

For details on how this can be accessed from CSS and javascript, please
see here:

​http://developer.android.com/guide/webapps/targeting.html#DensityCSS

  • public/WebView.h:
  • src/WebViewImpl.cpp:

(WebKit::WebViewImpl::deviceScaleFactor):
(WebKit::WebViewImpl::setDeviceScaleFactor):

  • src/WebViewImpl.h:
10:06 AM Changeset in webkit [99290] by vestbo@webkit.org
  • 2 edits in trunk/Tools

[Qt] Don't warn about disabling WebKit2 when qmake does recursive includes

Reviewed by Simon Hausmann.

  • qmake/mkspecs/features/default_pre.prf:
10:04 AM Changeset in webkit [99289] by andersca@apple.com
  • 6 edits in trunk/Source

Add NetscapePlugin::convertFromRootView
​https://bugs.webkit.org/show_bug.cgi?id=71526

Reviewed by Sam Weinig.

Source/WebCore:

Export some symbols that WebKit2 needs.

  • WebCore.exp.in:

Source/WebKit2:

  • WebProcess/Plugins/Netscape/NetscapePlugin.cpp:

(WebKit::NetscapePlugin::convertFromRootView):
New function that converts a point from root view coordinates to plug-in coordinates.
Returns false if the conversion can't be done.

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

(WebKit::fillInCocoaEventFromMouseEvent):
(WebKit::initializeMouseEvent):
(WebKit::NetscapePlugin::platformHandleMouseEvent):
(WebKit::NetscapePlugin::platformHandleWheelEvent):
(WebKit::NetscapePlugin::platformHandleMouseEnterEvent):
(WebKit::NetscapePlugin::platformHandleMouseLeaveEvent):
Use convertFromRootView instead of subtracting m_frameRectInWindowCoordinates.location() from the
event position (which is in root view coordinates).

10:00 AM Changeset in webkit [99288] by tony@chromium.org
  • 1 edit in trunk/LayoutTests/ChangeLog

Fix svn properties on layout test result files I checked in.

  • platform/chromium-linux/printing/single-line-must-not-be-split-into-two-pages-expected.png: Added property svn:mime-type.
  • platform/chromium-win/printing/setPrinting-expected.txt: Added property svn:eol-style.
  • platform/chromium-win/printing/single-line-must-not-be-split-into-two-pages-expected.png: Added property svn:mime-type.
  • platform/chromium-win/printing/single-line-must-not-be-split-into-two-pages-expected.txt: Added property svn:eol-style.
  • platform/mac/printing/setPrinting-expected.txt: Added property svn:eol-style.
  • platform/mac/printing/single-line-must-not-be-split-into-two-pages-expected.txt: Added property svn:eol-style.
9:55 AM Changeset in webkit [99287] by vestbo@webkit.org
  • 2 edits in trunk/Tools

[Qt] Don't error out on wrong Qt version when qmake does recursive includes

Reviewed by Simmon Hausmann.

9:41 AM Changeset in webkit [99286] by enne@google.com
  • 15 edits in trunk/Source

[chromium] Fix incorrect visibility/scissor rect for threaded compositing
​https://bugs.webkit.org/show_bug.cgi?id=70962

Reviewed by James Robinson.

Source/WebCore:

In order to properly scroll independent of layout on the compositor
thread, we have to apply the scroll to the right layer. Previously,
the NonCompositedContentHost was the root of the graphics layer tree
with all other layers underneath it. However, applying a scroll to
that layer would also scroll the clip layer for the main frame
underneath it, causing visibility issues.

This patch moves the NonCompositedHost to be a child of the
RenderLayerCompositor's scroll layer. This was exposed on
RenderLayerCompositor, because there's no other way to get to this
layer in a way that doesn't make assumptions about the structure of
the layers on a frame.

This also removes the hacks in calculateVisibleLayerRect and
tilingTransform for scroll position.

Partially tested by existing compositor tests. Manually tested
threaded compositing scrolling.

  • platform/graphics/chromium/NonCompositedContentHost.cpp:

(WebCore::NonCompositedContentHost::setScrollLayer):
(WebCore::NonCompositedContentHost::setViewport):
(WebCore::NonCompositedContentHost::scrollLayer):

  • platform/graphics/chromium/NonCompositedContentHost.h:
  • platform/graphics/chromium/cc/CCLayerTreeHost.cpp:

(WebCore::CCLayerTreeHost::create):
(WebCore::CCLayerTreeHost::CCLayerTreeHost):
(WebCore::CCLayerTreeHost::applyScrollDeltas):

  • platform/graphics/chromium/cc/CCLayerTreeHost.h:

(WebCore::CCLayerTreeHost::setRootLayer):

  • platform/graphics/chromium/cc/CCLayerTreeHostCommon.h:

(WebCore::CCLayerTreeHostCommon::calculateVisibleLayerRect):

  • platform/graphics/chromium/cc/CCLayerTreeHostImpl.cpp:

(WebCore::findScrollLayer):
(WebCore::CCLayerTreeHostImpl::setRootLayer):
(WebCore::CCLayerTreeHostImpl::scrollRootLayer):
(WebCore::CCLayerTreeHostImpl::processScrollDeltas):

  • platform/graphics/chromium/cc/CCLayerTreeHostImpl.h:
  • platform/graphics/chromium/cc/CCTiledLayerImpl.cpp:

(WebCore::CCTiledLayerImpl::tilingTransform):

  • rendering/RenderLayerCompositor.cpp:

(WebCore::RenderLayerCompositor::scrollLayer):

  • rendering/RenderLayerCompositor.h:

Source/WebKit/chromium:

  • src/WebLayerTreeViewImpl.cpp:

(WebKit::WebLayerTreeViewImpl::WebLayerTreeViewImpl):

  • src/WebViewImpl.cpp:

(WebKit::WebViewImpl::setRootGraphicsLayer):
(WebKit::WebViewImpl::setIsAcceleratedCompositingActive):

  • tests/CCLayerTreeHostTest.cpp:

(WTF::MockLayerTreeHost::MockLayerTreeHost):

9:09 AM Changeset in webkit [99285] by Philippe Normand
  • 7 edits in trunk

[GTK] Disable VIDEO_TRACK for now
​https://bugs.webkit.org/show_bug.cgi?id=71547

Reviewed by Gustavo Noronha Silva.

.:

  • configure.ac: VIDEO_TRACK is still a moving target, it's best to

disable it for now.

LayoutTests:

Skip media/track tests and rebaseline involved fast/dom test results.

  • platform/gtk/Skipped:
  • platform/gtk/fast/dom/Window/window-properties-expected.txt:
  • platform/gtk/fast/dom/Window/window-property-descriptors-expected.txt:
  • platform/gtk/fast/dom/prototype-inheritance-2-expected.txt:
8:48 AM Changeset in webkit [99284] by Csaba Osztrogonác
  • 2 edits in trunk/LayoutTests

[Qt] Unreviewed gardening.

  • platform/qt/Skipped: Disable inspector tests temporarily because of a bug near build system refactoring.
8:31 AM Changeset in webkit [99283] by caryclark@google.com
  • 478 edits
    131 adds in trunk/LayoutTests

Unreviewed gardening: 8th chunk of baselines for Chromium-Skia
on Leopard.

[Omitting 602 of 608 files for brevity.]

  • platform/chromium-mac-leopard/css3/bdi-element-expected.png: Added.
  • platform/chromium-mac-leopard/css3/bdi-element-expected.txt: Added.
  • platform/chromium-mac-leopard/fast/text/international/bidi-mirror-he-ar-expected.png: Added.

...

  • platform/chromium-mac-leopard/transforms/no_transform_hit_testing-expected.png: Added.
  • platform/chromium-mac-leopard/transforms/svg-vs-css-expected.png: Added.
  • platform/chromium-mac-leopard/transitions/move-after-transition-expected.png: Added.
8:29 AM Changeset in webkit [99282] by vestbo@webkit.org
  • 2 edits in trunk/Tools

[Qt] Disable STDIN check before cleaning build dir

It didn't work on the bots.

Reviewed by Ossy.

  • Scripts/webkitdirs.pm:
8:20 AM Changeset in webkit [99281] by vestbo@webkit.org
  • 5 edits in trunk

[Qt] Don't build all of webkit when running build-jsc

In case the whole webkit project was built, but we then subsequently
want to build only JSC, we need to run make in the proper subdirectory.

This also means the incremental target needs to be added to all
makefiles.

Reviewed by Simon Hausmann.

8:20 AM Changeset in webkit [99280] by vestbo@webkit.org
  • 1 edit in trunk/Tools/Scripts/webkitdirs.pm

[Qt] Look for STDIN instead of STDOUT when asking to clean the build dir

Reviewed by Ossy.

7:54 AM Changeset in webkit [99279] by Simon Hausmann
  • 2 edits in trunk/Tools

Fix run-javascriptcore-tests for Qt when it's executed by the bot.

Reviewed by Tor Arne Vestbø.

  • Scripts/webkitdirs.pm:

(jscProductDir): jsc path changed.

7:03 AM Changeset in webkit [99278] by vestbo@webkit.org
  • 2 edits in trunk/Tools

[Qt] Error out early if trying to build against Qt < 4.8

Reviewed by Simon Hausmann.

6:47 AM Changeset in webkit [99277] by Simon Hausmann
  • 2 edits in trunk/Tools

Fix run-launcher for Qt and Gtk.

Rubber-stamped by Tor Arne Vestbø.

  • Scripts/webkitdirs.pm:

(isWK2): Re-add isWK2(), as it's used in run-launcher.

6:39 AM Changeset in webkit [99276] by vestbo@webkit.org
  • 2 edits in trunk/Tools

[Qt] Ignore '-2' if passed to build-webkit. WebKit2 is now built by default.

Reviewed by Simon Hausmann.

6:37 AM Changeset in webkit [99275] by Carlos Garcia Campos
  • 2 edits in trunk/Tools

[GTK] Use web view title as window title in MiniBrowser
​https://bugs.webkit.org/show_bug.cgi?id=71544

Reviewed by Philippe Normand.

  • MiniBrowser/gtk/BrowserWindow.c:

(webViewURIChanged): Removed extra space.
(webViewTitleChanged): Set window title using
webkit_web_view_get_title().
(browser_window_init): Set initial default window title.
(browserWindowConstructed): Connect to WebView notify::title
signal.

6:19 AM Changeset in webkit [99274] by vestbo@webkit.org
  • 50 edits
    4 copies
    9 moves
    31 adds
    16 deletes in trunk

[Qt] Refactor and clean up the qmake build system

The qmake build system has accumulated a bit of cruft and redundancy
over time. There's also a fairly tight coupling between how to build
the various targets, and _what_ to build, making it harder to add new
rules or sources. This patch aims to elevate these issues somewhat.

This is a short-list of the changes:

  • The rules for how to build targets are now mostly contained as prf-files in Tools/qmake/mkspecs/features. Using mkspecs also allows us to do pre- and post-processing of each project file, which helps to clean up the actual project files.
  • Derived sources are no longer generated as a separate make-step but is part of each target's project file as a subdir. Makefile rules are used to ensure that we run make on the derived sources before running qmake on the actual target makefile. This makes it easier to keep a proper dependency between derived sources and the target.
  • We use GNU make and the compiler to generate dependencies on UNIX-based systems running Qt 5. This allows us to lessen the need to run qmake, which should reduce compile time.
  • WebKit2 is now build by default if building with Qt 5. It can be disabled by passing --no-webkit2 to build-webkit.

The result of these changes are hopefully a cleaner and easier
build system to modify, and faster build times due to no longer
running qmake on every single build. It's also a first step
towards possibly generating the list of sources using another
build system.

​https://bugs.webkit.org/show_bug.cgi?id=71222

Reviewed by Simon Hausmann.

5:28 AM Changeset in webkit [99273] by kenneth@webkit.org
  • 11 edits in trunk/Source/WebKit2

[Qt] Support programmatic scrolling with the QTouchWebView

Reviewed by Simon Hausmann.

As more code can be shared between the programmatic scrolling and the
method for ensuring that the contents are within valid bounds, this
patch includes a bit of refactoring as well.

  • UIProcess/API/qt/qdesktopwebview_p.h:

(QDesktopWebViewPrivate::scrollPositionRequested):

  • UIProcess/API/qt/qtouchwebview.cpp:

(QTouchWebViewPrivate::scrollPositionRequested):

  • UIProcess/API/qt/qtouchwebview_p.h:
  • UIProcess/qt/QtTouchViewInterface.cpp:

(WebKit::QtTouchViewInterface::scrollPositionRequested):

  • UIProcess/qt/QtTouchViewInterface.h:
  • UIProcess/qt/QtViewInterface.h:
  • UIProcess/qt/QtViewportInteractionEngine.cpp:

(WebKit::QtViewportInteractionEngine::event):
(WebKit::boundPosition):
(WebKit::QtViewportInteractionEngine::pagePositionRequest):
(WebKit::QtViewportInteractionEngine::computePosRangeForItemAtScale):
(WebKit::QtViewportInteractionEngine::ensureContentWithinViewportBoundary):
(WebKit::QtViewportInteractionEngine::setConstraints):
(WebKit::QtViewportInteractionEngine::pinchGestureEnded):
(WebKit::QtViewportInteractionEngine::contentViewportChanged):

  • UIProcess/qt/QtViewportInteractionEngine.h:
  • UIProcess/qt/QtWebPageProxy.cpp:

(QtWebPageProxy::pageDidRequestScroll):

  • UIProcess/qt/QtWebPageProxy.h:
4:23 AM Changeset in webkit [99272] by commit-queue@webkit.org
  • 2 edits in trunk/Source/WebCore

[chromium] MediaStream API: Fixing wrong path for PeerConnectionHandler.h in WebCore.gyp
​https://bugs.webkit.org/show_bug.cgi?id=71548

Patch by Tommy Widenflycht <​tommyw@google.com> on 2011-11-04
Reviewed by Tony Gentilcore.

  • WebCore.gypi:
2:20 AM Changeset in webkit [99271] by eric@webkit.org
  • 4 edits in trunk/Tools

Remove deprecated free functions in port.factory
​https://bugs.webkit.org/show_bug.cgi?id=71494

Unreviewed. Fixing exception seen when running test-webkitpy.

Sorry, I previously had deleted the relevant rebaseline-chromium-webkit-tests
unittest, since that script is nearly ready to be deleted. But I added
it back at the last second and failed to run the tests. :(
This fixes the exception the bots were seeing.

  • Scripts/webkitpy/style/checkers/test_expectations.py:
  • Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
  • Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests_unittest.py:
2:13 AM Changeset in webkit [99270] by Patrick Gansterer
  • 3 edits in trunk/Source/WebKit/wince

[WINCE] Use default LocalizationStrategy
​https://bugs.webkit.org/show_bug.cgi?id=71495

Reviewed by Adam Roben.

Use LocalizationStrategy from WebCore instead of using a copy of the strings in WebKit.

  • WebCoreSupport/PlatformStrategiesWinCE.cpp:
  • WebCoreSupport/PlatformStrategiesWinCE.h:
2:04 AM Changeset in webkit [99269] by Philippe Normand
  • 2 edits in trunk/LayoutTests

Unreviewed, skip some fast/viewport failing tests on GTK after r99195.

  • platform/gtk/Skipped:
1:53 AM Changeset in webkit [99268] by gyuyoung.kim@samsung.com
  • 21 edits in trunk/Source/WebKit/efl

[EFL] Use standard booleand data type.
​https://bugs.webkit.org/show_bug.cgi?id=71433

This is a fourth step in order to be more compliant with WebKit coding style.
Use standard boolean data type instead of efl boolean data type. It makes efl port
is more close to webkit coding style.

Reviewed by Zoltan Herczeg.

  • WebCoreSupport/ChromeClientEfl.cpp:
  • WebCoreSupport/FrameLoaderClientEfl.cpp:
  • ewk/ewk_contextmenu.cpp:
  • ewk/ewk_cookies.cpp:
  • ewk/ewk_frame.cpp:
  • ewk/ewk_history.cpp:
  • ewk/ewk_js.cpp:
  • ewk/ewk_main.cpp:
  • ewk/ewk_network.cpp:
  • ewk/ewk_private.h:
  • ewk/ewk_protocol_handler.cpp:
  • ewk/ewk_protocol_handler_soup.cpp:
  • ewk/ewk_settings.cpp:
  • ewk/ewk_tiled_backing_store.cpp:
  • ewk/ewk_tiled_matrix.cpp:
  • ewk/ewk_tiled_model.cpp:
  • ewk/ewk_util.cpp:
  • ewk/ewk_view.cpp:
  • ewk/ewk_view_single.cpp:
  • ewk/ewk_view_tiled.cpp:
1:20 AM Changeset in webkit [99267] by Patrick Gansterer
  • 2 edits in trunk/Source/WebCore

Unreviewed CMake build fix for !ENABLE(SQL_DATABASE).

  • CMakeLists.txt: Always compile the SQLite source files, since they are required for build.
1:06 AM Changeset in webkit [99266] by yutak@chromium.org
  • 2 edits in trunk/LayoutTests

[Chromium] Unreviewed gardening.

  • platform/chromium/test_expectations.txt: Add more expectations for recent test flakinesses.
12:45 AM Changeset in webkit [99265] by commit-queue@webkit.org
  • 4 edits in trunk/Source/JavaScriptCore

More code clean-up in DFG 32_64
​https://bugs.webkit.org/show_bug.cgi?id=71540

Remove unnecessary code duplications, and fix compilation warnings.

Patch by Yuqiang Xian <​yuqiang.xian@intel.com> on 2011-11-04
Reviewed by Gavin Barraclough.

  • dfg/DFGJITCompiler.cpp:

(JSC::DFG::JITCompiler::emitCount):
(JSC::DFG::JITCompiler::setSamplingFlag):
(JSC::DFG::JITCompiler::clearSamplingFlag):
(JSC::DFG::JITCompiler::jitAssertIsCell):

  • dfg/DFGJITCompiler32_64.cpp:
  • dfg/DFGSpeculativeJIT32_64.cpp:

(JSC::DFG::SpeculativeJIT::compile):

12:35 AM Changeset in webkit [99264] by Csaba Osztrogonác
  • 15 edits in trunk/Source/JavaScriptCore

De-virtualize JSObject::hasInstance
​https://bugs.webkit.org/show_bug.cgi?id=71430

Unreviewed rolling out r99238, because it made a test crash on all platform.

  • API/JSCallbackObject.h:
  • API/JSCallbackObjectFunctions.h:

(JSC::::hasInstance):

  • API/JSValueRef.cpp:

(JSValueIsInstanceOfConstructor):

(JSC::Interpreter::privateExecute):

  • jit/JITStubs.cpp:

(JSC::DEFINE_STUB_FUNCTION):

  • runtime/ClassInfo.h:
  • runtime/JSBoundFunction.cpp:

(JSC::JSBoundFunction::hasInstance):

  • runtime/JSBoundFunction.h:
  • runtime/JSCell.cpp:
  • runtime/JSCell.h:
  • runtime/JSObject.cpp:

(JSC::JSObject::hasInstance):

  • runtime/JSObject.h:
Note: See TracTimeline for information about the timeline view.