Timeline



Jun 6, 2014:

6:29 PM Changeset in webkit [169670] by gyuyoung.kim@samsung.com
  • 9 edits in trunk

Scheme of content utils should be compared in an ASCII case-insensitive manner
​https://bugs.webkit.org/show_bug.cgi?id=133502

Reviewed by Darin Adler.

Source/WebCore:
According to spec, scheme must be compared in an ASCII case-insensitive manner by user agents.

Covered by existing tests.

  • Modules/navigatorcontentutils/NavigatorContentUtils.cpp:

(WebCore::isProtocolWhitelisted):

LayoutTests:
Tests are updated to compare protocol in ASCII case-insensitive manner.

  • fast/dom/NavigatorContentUtils/register-protocol-handler-expected.txt:
  • fast/dom/NavigatorContentUtils/register-protocol-handler.html:
  • fast/dom/NavigatorContentUtils/unregister-protocol-handler-expected.txt:
  • fast/dom/NavigatorContentUtils/unregister-protocol-handler.html:
  • platform/efl/fast/dom/NavigatorContentUtils/register-protocol-handler-expected.txt:
  • platform/efl/fast/dom/NavigatorContentUtils/unregister-protocol-handler-expected.txt:
5:30 PM Changeset in webkit [169669] by enrica@apple.com
  • 7 edits in trunk/Source/WebKit2

REGRESSION (iOS WebKit2): backspace key does not repeat beyond a single word.
​https://bugs.webkit.org/show_bug.cgi?id=133593
<rdar://problem/16352228>

Reviewed by Benjamin Poulain.

We need to provide a way for the keyboard layer to select from
the current position to the beginning of the word moving backwards
and implement the method to know when we have reached the beginning of the document.

  • UIProcess/WebPageProxy.h:
  • UIProcess/ios/WKContentViewInteraction.mm:

(-[WKContentView selectWordBackward]):
(-[WKContentView _selectionAtDocumentStart]):

  • UIProcess/ios/WebPageProxyIOS.mm:

(WebKit::WebPageProxy::selectWordBackward):

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

(WebKit::WebPage::selectWordBackward):

4:56 PM Changeset in webkit [169668] by akling@apple.com
  • 11 edits in trunk/Source

Indexed getters should return values directly on the PropertySlot.
<​https://webkit.org/b/133586>

Source/JavaScriptCore:
Remove PropertySlot's custom index mode.

Reviewed by Darin Adler.

  • runtime/JSObject.h:

(JSC::PropertySlot::getValue):

  • runtime/PropertySlot.h:

(JSC::PropertySlot::setCustomIndex): Deleted.

Source/WebCore:
Make indexed getters more efficient by using PropertySlot::setValue()
to pass the value directly back through the slot, instead of giving it
a function pointer that JSC would then immediately call back through
to retrieve the value.

The function pointer approach would make sense if we did inline caching
of indexed getters but since we currently don't, this is pure overhead.

Reviewed by Darin Adler.

  • bindings/js/JSCSSStyleDeclarationCustom.cpp:

(WebCore::JSCSSStyleDeclaration::getOwnPropertySlotDelegate):
(WebCore::cssPropertyGetterPixelOrPosPrefixCallback): Deleted.
(WebCore::cssPropertyGetterCallback): Deleted.

  • bindings/js/JSDOMWindowCustom.cpp:

(WebCore::JSDOMWindow::getOwnPropertySlot):
(WebCore::JSDOMWindow::getOwnPropertySlotByIndex):
(WebCore::indexGetter): Deleted.

  • bindings/scripts/CodeGeneratorJS.pm:

(GenerateGetOwnPropertySlotBody):
(GenerateHeader):
(GetIndexedGetterExpression):
(GenerateImplementation):

  • bridge/runtime_array.cpp:

(JSC::RuntimeArray::getOwnPropertySlot):
(JSC::RuntimeArray::getOwnPropertySlotByIndex):
(JSC::RuntimeArray::indexGetter): Deleted.

  • bridge/runtime_array.h:
4:32 PM Changeset in webkit [169667] by matthew_hanson@apple.com
  • 2 edits in tags/Safari-538.39/Source/WebCore

Merge r169665. <rdar://problem/15840022>

4:18 PM Changeset in webkit [169666] by Brent Fulgham
  • 3 edits in trunk/Source/WebCore

GenericCueData elements prematurely removed
​https://bugs.webkit.org/show_bug.cgi?id=133588

Reviewed by Eric Carlson.

Add a 'removeCompletedCues' method instead of using the 'resetCueValues' method
after processing new cues. We do not want to reset cue state if we have pending
cues that are being extended by future updates (such as paint-on cues).

  • platform/graphics/avfoundation/InbandTextTrackPrivateAVF.cpp:

(WebCore::InbandTextTrackPrivateAVF::processCue): Call new 'removeCompletedCues' method.
(WebCore::InbandTextTrackPrivateAVF::removeCompletedCues): Added.

  • platform/graphics/avfoundation/InbandTextTrackPrivateAVF.h:
4:07 PM Changeset in webkit [169665] by beidson@apple.com
  • 2 edits in trunk/Source/WebCore

Initialize a char* that needs to be initialized.
<rdar://problem/15840022> and ​https://bugs.webkit.org/show_bug.cgi?id=133585

Reviewed by Mark Rowe.

r152134 (erroneously) removed this initialization.

  • platform/sql/SQLiteStatement.cpp:

(WebCore::SQLiteStatement::prepare): tail should start out initialized to 0.

3:06 PM Changeset in webkit [169664] by enrica@apple.com
  • 8 edits in trunk/Source/WebKit2

Adopt the new UIAlertController API to replace UIActionSheet.
​https://bugs.webkit.org/show_bug.cgi?id=133582
<rdar://problem/16845223>

Reviewed by Joseph Pecoraro.

UIActionSheet is deprecated. We need to move out action sheets
implementation to UIAlertView.

  • UIProcess/API/Cocoa/_WKElementAction.h:
  • UIProcess/API/Cocoa/_WKElementAction.mm:
  • UIProcess/ios/WKActionSheet.h:
  • UIProcess/ios/WKActionSheet.mm:

(-[WKActionSheet presentSheetFromRect:]):
(-[WKActionSheet doneWithSheet]):
(-[WKActionSheet willRotate]):
(-[WKActionSheet updateSheetPosition]):
(-[WKActionSheet _didRotateAndLayout]):
(-[WKActionSheet didRotate]):

  • UIProcess/ios/WKActionSheetAssistant.h:
  • UIProcess/ios/WKActionSheetAssistant.mm:

(-[WKActionSheetAssistant _createSheetWithElementActions:showLinkTitle:]):
(-[WKActionSheetAssistant showImageSheet]):
(-[WKActionSheetAssistant showLinkSheet]):
(-[WKActionSheetAssistant showDataDetectorsSheet]):
(-[WKActionSheetAssistant cleanupSheet]):
(-[WKActionSheetAssistant actionSheet:clickedButtonAtIndex:]): Deleted.

  • UIProcess/ios/WKContentViewInteraction.mm:

(-[WKContentView _longPressRecognized:]):

2:43 PM Changeset in webkit [169663] by barraclough@apple.com
  • 2 edits in trunk/Source/WebKit2

Should boost ChildProcesses on iOS
​https://bugs.webkit.org/show_bug.cgi?id=133583

Reviewed by Sam Weinig.

  • UIProcess/Launcher/mac/ProcessLauncherMac.mm:

(WebKit::connectToService):

  • boost all child processes on iOS.
2:08 PM Changeset in webkit [169662] by Simon Fraser
  • 2 edits in trunk/Tools

32-bit build fix.

  • WebKitTestRunner/mac/EventSenderProxy.mm:

(WTR::EventSenderProxy::mouseMoveTo):
(WTR::EventSenderProxy::mouseScrollByWithWheelAndMomentumPhases):

2:07 PM Changeset in webkit [169661] by matthew_hanson@apple.com
  • 7 edits in tags/Safari-538.39/Source

Rollout r169656. <rdar://problem/17095692>

2:04 PM Changeset in webkit [169660] by ap@apple.com
  • 3 edits in trunk/Source/WebCore

[iOS] Support pasting HTML
​https://bugs.webkit.org/show_bug.cgi?id=133581
<rdar://problem/17070305>

Reviewed by Enrica Casucci.

We can probably share a lot of code between Mac and iOS here, but for now, just
implement the unimplemented functions.

  • editing/ios/EditorIOS.mm: (WebCore::Editor::WebContentReader::readHTML):

Implemented based on WebArchive reading code above, and on Mac version of readHTML.
Unlike on Mac, we don't need to support Microsoft CF_HTML format, as nothing
generates it.

  • platform/ios/PasteboardIOS.mm:

(WebCore::Pasteboard::read):
(WebCore::Pasteboard::supportedPasteboardTypes):
Added HTML support, as the second preferred format after WebArchive.

1:59 PM Changeset in webkit [169659] by Simon Fraser
  • 4 edits in trunk/Tools

Fix builds on 10.8; -firstObject isn't declared on NSArray there.

  • DumpRenderTree/mac/DumpRenderTree.mm:

(createWebViewAndOffscreenWindow):

  • DumpRenderTree/mac/EventSendingController.mm:

(-[EventSendingController mouseScrollByX:andY:continuously:]):
(-[EventSendingController mouseScrollByX:andY:withWheel:andMomentumPhases:]):

  • WebKitTestRunner/mac/EventSenderProxy.mm:

(WTR::EventSenderProxy::mouseScrollBy):
(WTR::EventSenderProxy::mouseScrollByWithWheelAndMomentumPhases):

1:34 PM Changeset in webkit [169658] by ap@apple.com
  • 5 edits in trunk/Source/WebKit2

Build fix after <​http://trac.webkit.org/changeset/169655>.

The SPIs used do not exist on Mac, limit the code to iOS.

  • Shared/Authentication/mac/AuthenticationManager.mac.mm:
  • Shared/WebCoreArgumentCoders.cpp:

(IPC::ArgumentCoder<Credential>::encode):
(IPC::ArgumentCoder<Credential>::decode):

  • Shared/cf/ArgumentCodersCF.cpp:

(IPC::typeFromCFTypeRef):
(IPC::encode):
(IPC::decode):

  • Shared/cf/ArgumentCodersCF.h:
1:07 PM Changeset in webkit [169657] by ap@apple.com
  • 2 edits in trunk/Source/WebCore

Try to fix the build after ​http://trac.webkit.org/changeset/169655

Unfortunately, the fix is very fragile, and will likely not work for some SDKs
or possibly local headers - all OS versions have multiple SecBase.h headers with
different definitions for SecIdentityRef.

We'll see.

  • WebCore.exp.in:
12:57 PM Changeset in webkit [169656] by matthew_hanson@apple.com
  • 7 edits in tags/Safari-538.39/Source

Merge r169655. <rdar://problem/17095692>

12:04 PM Changeset in webkit [169655] by mitz@apple.com
  • 7 edits in trunk/Source

Source/WebCore: WebCore part of <rdar://problem/17095692> [iOS] Client-certificate authentication isn’t working
​https://bugs.webkit.org/show_bug.cgi?id=133527

Reviewed by Darin Adler.

  • WebCore.exp.in: Exported some Credential member functions.

Source/WebKit2: <rdar://problem/17095692> [iOS] Client-certificate authentication isn’t working
​https://bugs.webkit.org/show_bug.cgi?id=133527

Reviewed by Darin Adler.

  • Configurations/Network-iOS.entitlements: Enabled the Network process to access the keys

needed to create identities to authenticate with.

  • Shared/WebCoreArgumentCoders.cpp:

(IPC::ArgumentCoder<Credential>::encode): Encode the credential type, and if it is a client
certificate, encode the identity and the certificates.
(IPC::ArgumentCoder<Credential>::decode): Decode the credential type. If it is a client
certificate, decode the identity and the certificates and use the proper Credential
constructor.

  • Shared/cf/ArgumentCodersCF.cpp:

(IPC::typeFromCFTypeRef): Handle SecIdentityRef.
(IPC::encode): Encode an identity by encoding its certificate and a persistent reference to
its key.
(IPC::decode): Decode a certificate and a persistent reference to a key, find the key, and
create an identity.

  • Shared/cf/ArgumentCodersCF.h:
11:37 AM Changeset in webkit [169654] by dino@apple.com
  • 3 edits in trunk/Source/WebCore

[iOS WebGL] Float texture extension has a slightly different name
​https://bugs.webkit.org/show_bug.cgi?id=133580
<rdar://problem/17113451>

Reviewed by Brady Eidson.

Floating point textures were not available on iOS, because we were
looking for the wrong extension.

Covered by the Khronos test (and in LayoutTests):
conformance/extensions/oes-texture-float.html

  • platform/graphics/opengl/Extensions3DOpenGL.cpp:

(WebCore::Extensions3DOpenGL::supportsExtension): Look for GL_OES_texture_float
as well as GL_ARB_texture_float.

  • platform/graphics/opengl/GraphicsContext3DOpenGL.cpp:

(WebCore::GraphicsContext3D::texImage2D): Increase the set of things skipped
on PLATFORM(IOS) since they are supported natively there.

11:00 AM Changeset in webkit [169653] by matthew_hanson@apple.com
  • 5 edits in trunk/Source

Versioning.

10:59 AM Changeset in webkit [169652] by matthew_hanson@apple.com
  • 2 edits in tags/Safari-538.39/Source/WebCore

Add UNUSED_PARAM when GRID is off.

10:57 AM Changeset in webkit [169651] by hyatt@apple.com
  • 7 edits in trunk

REGRESSION: Hit testing of composited elements is broken in new multicolumn layout.
​https://bugs.webkit.org/show_bug.cgi?id=133522 (<rdar://problem/17060556>)

Reviewed by Simon Fraser.

Source/WebCore:
Hit testing for the new columns doesn't care about compositing but does still
need a pagination layer set to work. The old column code didn't care about this
because it was top down (and wrong). The new code lets each layer paint and hit
test its own fragments, so having a pagination layer set suddenly becomes relevant
for composited layers.

The solution is to cache whether or not the pagination layer crosses a compositing
boundary, and then to force call sites of enclosingPaginationLayer to pass an argument
indicating whether compositing layers should be included. Typically painting wants to
exclude them but hit testing wants to include them.

  • page/FrameView.cpp:

(WebCore::updateLayerPositionFlags):

  • rendering/RenderLayer.cpp:

(WebCore::RenderLayer::RenderLayer):
(WebCore::RenderLayer::updateLayerPositions):
(WebCore::RenderLayer::updatePagination):
(WebCore::transparencyClipBox):
(WebCore::RenderLayer::paintLayer):
(WebCore::RenderLayer::paintLayerContents):
(WebCore::RenderLayer::enclosingPaginationLayerInSubtree):
(WebCore::RenderLayer::collectFragments):
(WebCore::RenderLayer::paintTransformedLayerIntoFragments):
(WebCore::RenderLayer::hitTestLayer):
(WebCore::RenderLayer::hitTestTransformedLayerInFragments):
(WebCore::RenderLayer::calculateClipRects):

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

(WebCore::RenderLayerCompositor::requiresCompositingForPosition):

LayoutTests:

  • compositing/columns/hittest-composited-in-paginated-expected.txt:
10:41 AM Changeset in webkit [169650] by Simon Fraser
  • 18 edits in trunk

Latched scrolling tests are flakey on Mavericks
​https://bugs.webkit.org/show_bug.cgi?id=133578
<rdar://problem/17180591>

Reviewed by Brent Fulgham.

Tools:
The latched scrolling tests rely on synthetic mousewheel events whose coordinates can
be wrong on machines with more than one screen. This is because we use
[NSScreen mainScreen] when computing the global coordinates of the events, but
[NSScreen mainScreen] returns the screen with the key window, not the first screen.

So replace calls to [NSScreen mainScreen] with [[NSScreen screens] firstObject]. Also
add comments clarifying why the coordinate math works (which is not obvious given
the events have no target window).

Some scrolling tests also dispatched events outside the 800x600 test window, so
fix those tests, and add logging that will appear in test output when this happens.

  • DumpRenderTree/mac/DumpRenderTree.mm:

(createWebViewAndOffscreenWindow):

  • DumpRenderTree/mac/EventSendingController.mm:

(-[EventSendingController mouseScrollByX:andY:continuously:]):
(-[EventSendingController mouseScrollByX:andY:withWheel:andMomentumPhases:]):

  • WebKitTestRunner/mac/EventSenderProxy.mm:

(WTR::EventSenderProxy::mouseMoveTo):
(WTR::EventSenderProxy::mouseScrollBy):
(WTR::EventSenderProxy::continuousMouseScrollBy):
(WTR::EventSenderProxy::mouseScrollByWithWheelAndMomentumPhases):

LayoutTests:
Fix these tests to not dispatch events outside the window, and correct some of the wording
in the tests.

  • platform/mac-wk2/tiled-drawing/fast-scroll-div-latched-div-expected.txt:
  • platform/mac-wk2/tiled-drawing/fast-scroll-div-latched-div-with-handler-expected.txt:
  • platform/mac-wk2/tiled-drawing/fast-scroll-div-latched-div-with-handler.html:
  • platform/mac-wk2/tiled-drawing/fast-scroll-div-latched-div.html:
  • platform/mac-wk2/tiled-drawing/fast-scroll-div-latched-mainframe-with-handler.html:
  • platform/mac-wk2/tiled-drawing/fast-scroll-div-latched-mainframe.html:
  • platform/mac-wk2/tiled-drawing/fast-scroll-iframe-latched-mainframe.html:
  • platform/mac-wk2/tiled-drawing/fast-scroll-select-latched-mainframe-with-handler.html:
  • platform/mac-wk2/tiled-drawing/fast-scroll-select-latched-mainframe.html:
  • platform/mac-wk2/tiled-drawing/fast-scroll-select-latched-select-expected.txt:
  • platform/mac-wk2/tiled-drawing/fast-scroll-select-latched-select-with-handler-expected.txt:
  • platform/mac-wk2/tiled-drawing/fast-scroll-select-latched-select-with-handler.html:
  • platform/mac-wk2/tiled-drawing/fast-scroll-select-latched-select.html:
9:54 AM Changeset in webkit [169649] by commit-queue@webkit.org
  • 2 edits in trunk/LayoutTests

Run passing WebGL tests on Windows.
​https://bugs.webkit.org/show_bug.cgi?id=133563

Patch by Alex Christensen <​achristensen@webkit.org> on 2014-06-06
Reviewed by Brent Fulgham.

  • platform/win/TestExpectations:

Replaced all WebGL tests with lists of failing and crashing tests to fix.

9:33 AM Changeset in webkit [169648] by commit-queue@webkit.org
  • 7 edits in trunk

[MediaStream] 'get' prefix is missing for capabilities and constraints.
​https://bugs.webkit.org/show_bug.cgi?id=133573

Patch by Kiran <​kiran.guduru@samsung.com> on 2014-06-06
Reviewed by Eric Carlson.

Source/WebCore:
Modified the existing MediaStreamTrack.html to test the changes.

As per the Media Capture and Streams spec method names constreaints and
capabilities should be prefixed by 'get'. So updated the method names.

  • Modules/mediastream/MediaStreamTrack.cpp:

(WebCore::MediaStreamTrack::getConstraints):
(WebCore::MediaStreamTrack::getCapabilities):
(WebCore::MediaStreamTrack::constraints): Deleted.
(WebCore::MediaStreamTrack::capabilities): Deleted.

  • Modules/mediastream/MediaStreamTrack.h:
  • Modules/mediastream/MediaStreamTrack.idl:

LayoutTests:

  • fast/mediastream/MediaStreamTrack-expected.txt:
  • fast/mediastream/MediaStreamTrack.html:
9:20 AM Changeset in webkit [169647] by Brent Fulgham
  • 4 edits in trunk/Source/WebCore

REGRESSION(169440): Repro crash when playing a video with captions
​https://bugs.webkit.org/show_bug.cgi?id=133459
<rdar://problem/17079081>

Reviewed by Eric Carlson.

The change I made in r169440 introduced a crash for paint-on captions. This
resolves the problem as follows:

  1. No longer attempt to modify the start/endTime in the TextTrack implementation.
  2. Check for extensions to existing cues when we are about to add new arriving cues. If they extend a current pending cue, don't set an end-time for the pending cue, and throw away the new duplicate. Otherwise set the end time as previously and continue.
  • html/track/TextTrack.cpp:

(WebCore::TextTrack::hasCue): Remove 'doesExtendCue' case since this is
no longer possible.

  • platform/graphics/InbandTextTrackPrivateClient.h:

(WebCore::GenericCueData::doesExtendCueData): Added.

  • platform/graphics/avfoundation/InbandTextTrackPrivateAVF.cpp:

(WebCore::InbandTextTrackPrivateAVF::processCue): Check for cues extending
other cues and handle appropriately.

9:19 AM Changeset in webkit [169646] by mario.prada@samsung.com
  • 4 edits
    1 copy in trunk/LayoutTests

[GTK] Layout Test accessibility/table-fallback-roles-expose-element-attributes.html is failing
​https://bugs.webkit.org/show_bug.cgi?id=133519

Reviewed by Chris Fleizach.

Adapted test to play nicely with the different types of
accessibility hierarchies exposed for Mac and ATK.

  • accessibility/table-fallback-roles-expose-element-attributes.html:

Updated text to print information of cells from every platform but
only print information of their children in the Mac.

  • accessibility/table-fallback-roles-expose-element-attributes-expected.txt:

Updated expectation to match ATK platforms (GTK & EFL).

  • platform/mac/accessibility/table-fallback-roles-expose-element-attributes-expected.txt:

Copied from LayoutTests/accessibility/table-fallback-roles-expose-element-attributes-expected.txt.

  • platform/gtk/TestExpectations: Removed expected failure.
8:49 AM Changeset in webkit [169645] by Brent Fulgham
  • 4 edits in trunk/Source/WebCore

ROLLOUT: r169630: Caused a retain cycle.
​https://bugs.webkit.org/show_bug.cgi?id=133463
<rdar://problem/17098100>

Rolling out this change as it introduced a retain cycle.

  • Modules/mediacontrols/MediaControlsHost.cpp:

(WebCore::MediaControlsHost::setControllerJSObject): Deleted.

  • Modules/mediacontrols/MediaControlsHost.h:

(WebCore::MediaControlsHost::controllerJSValue):
(WebCore::MediaControlsHost::setControllerJSValue):
(WebCore::MediaControlsHost::controllerJSObject): Deleted.

  • html/HTMLMediaElement.cpp:

(WebCore::HTMLMediaElement::parseAttribute):

12:37 AM Changeset in webkit [169644] by fred.wang@free.fr
  • 1 edit
    7 adds in trunk/LayoutTests

Add test references for horizontal stretchy operators.
​https://bugs.webkit.org/show_bug.cgi?id=72828

Unreviewed Win and EFL gardening.

  • platform/efl/mathml/opentype/horizontal-LatinModern-expected.txt: Added.
  • platform/efl/mathml/opentype/horizontal-LatinModern-munderover-expected.txt: Added.
  • platform/efl/mathml/opentype/opentype-stretchy-horizontal-expected.txt: Added.
  • platform/win/mathml/opentype/horizontal-LatinModern-expected.txt: Added.
  • platform/win/mathml/opentype/horizontal-LatinModern-munderover-expected.txt: Added.
  • platform/win/mathml/opentype/opentype-stretchy-horizontal-expected.txt: Added.
12:26 AM Changeset in webkit [169643] by fred.wang@free.fr
  • 1 edit
    3 adds in trunk/LayoutTests

Add test references for horizontal stretchy operators.
​https://bugs.webkit.org/show_bug.cgi?id=72828

Unreviewed Mac gardening.

  • platform/mac/mathml/horizontal-LatinModern-expected.txt: Added.
  • platform/mac/mathml/horizontal-LatinModern-munderover-expected.txt: Added.
  • platform/mac/mathml/opentype-stretchy-horizontal-expected.txt: Added.
12:19 AM Changeset in webkit [169642] by matthew_hanson@apple.com
  • 8 edits in tags/Safari-538.39/Source

Custom patch for <rdar://problem/16650338>.

12:08 AM Changeset in webkit [169641] by Carlos Garcia Campos
  • 6 edits in releases/WebKitGTK/webkit-2.2

Merge r154727 - [Soup] WebTiming information not shown in the inspector
​https://bugs.webkit.org/show_bug.cgi?id=118395

Reviewed by Martin Robinson.

Source/WebCore:

WebTiming information was not correctly provided to WebCore
because the gotHeadersCallback was incorrectly resetting the
original ResourceResponse (which had the ResourceLoadTiming
object) instead of simply updating their contents using the
SoupMessage.

No new test required as this feature is already covered by the
existing webtiming tests. In any case this change includes a fix
for the http/tests/misc/webtiming-ssl.php test which was not
failing even if it should because it was not correct.

  • platform/network/soup/ResourceHandleSoup.cpp:

(WebCore::gotHeadersCallback):
(WebCore::restartedCallback): restartedCallback should be
available only if WEB_TIMING is defined.
(WebCore::createSoupMessageForHandleAndRequest): Ditto.

LayoutTests:

Fixed the webiming-ssl.php test that was incorrectly considering a
FAIL as the expected result. It was probably just a legacy
decision which came from the times where DRT was responsible of
reporting WebTiming information.

Also added a new test that fails due to wkb.ug/103927. It was not
detected before because there were no timing information in the
response.

  • http/tests/misc/resources/webtiming-ssl.html:
  • http/tests/misc/webtiming-ssl-expected.txt:
  • platform/gtk/TestExpectations: added

http/tests/w3c/webperf/submission/Google/resource-timing/html/test_resource_attribute_order.html.

Jun 5, 2014:

11:53 PM Changeset in webkit [169640] by dino@apple.com
  • 2 edits in trunk/Source/WebCore

StyleResolver doesn't compile without CSS_GRID_LAYOUT enabled
​https://bugs.webkit.org/show_bug.cgi?id=133568

Reviewed by Tim Horton.

  • css/StyleResolver.cpp:

(WebCore::isDisplayGridBox): Add UNUSED_PARAM.

11:44 PM Changeset in webkit [169639] by mihnea@adobe.com
  • 3 edits
    2 adds in trunk

[CSSRegions] Region box incorrectly overlaps floating box
​https://bugs.webkit.org/show_bug.cgi?id=133545

Reviewed by David Hyatt.

Source/WebCore:
A region box creates a BFC and it should not overlap a floating box in the same formatting context: ​http://www.w3.org/TR/CSS21/visuren.html#bfc-next-to-float

Test: fast/regions/region-as-bfc-avoids-float.html

  • rendering/RenderBlock.cpp:

(WebCore::RenderBlock::avoidsFloats):

LayoutTests:
Add test to check that a region does not incorrectly overlap a floating
element in the same formatting context.

  • fast/regions/region-as-bfc-avoids-float-expected.html: Added.
  • fast/regions/region-as-bfc-avoids-float.html: Added.
11:43 PM Changeset in webkit [169638] by matthew_hanson@apple.com
  • 1 copy in tags/Safari-538.39

Create the Safari-538.39 tag.

11:06 PM Changeset in webkit [169637] by yoav@yoav.ws
  • 17 edits
    1 move
    2 deletes in trunk

Align srcset parser with recent spec changes
​https://bugs.webkit.org/show_bug.cgi?id=133504

Reviewed by Darin Adler.

Source/WebCore:
New tests were added to fast/hidpi/image-srcset-invalid-descriptor.html

The srcset parser changes align it with the spec:
​http://picture.responsiveimages.org/#parse-srcset-attr.
The spec changes are meant to give the parser better future
compatibility, and enable it to work well with the 'w' descriptor, the
soon to be added 'h' descriptor and function descriptors (e.g.
future-descriptor(a, b, c) ).

  • CMakeLists.txt:
  • WebCore.vcxproj/WebCore.vcxproj:
  • WebCore.vcxproj/WebCore.vcxproj.filters:
  • WebCore.xcodeproj/project.pbxproj:
  • html/HTMLImageElement.cpp:

(WebCore::HTMLImageElement::setBestFitURLAndDPRFromImageCandidate):
(WebCore::HTMLImageElement::parseAttribute):

  • html/HTMLImageElement.h:
  • html/parser/HTMLParserIdioms.h:

(WebCore::isHTMLSpace):
(WebCore::isComma):
(WebCore::isHTMLSpaceOrComma):

  • html/parser/HTMLPreloadScanner.cpp:

(WebCore::TokenPreloadScanner::StartTagScanner::processAttributes):

  • html/parser/HTMLSrcsetParser.cpp:

(WebCore::compareByDensity):
(WebCore::appendDescriptorAndReset):
(WebCore::appendCharacter):
(WebCore::isEOF):
(WebCore::tokenizeDescriptors):
(WebCore::stringViewToFloat):
(WebCore::stringViewToInt):
(WebCore::parseDescriptors):
(WebCore::parseImageCandidatesFromSrcsetAttribute):
(WebCore::pickBestImageCandidate):
(WebCore::bestFitSourceForImageAttributes):
(WebCore::compareByScaleFactor): Deleted.
(WebCore::parseImagesWithScaleFromSrcsetAttribute): Deleted.

  • html/parser/HTMLSrcsetParser.h:

(WebCore::DescriptorParsingResult::DescriptorParsingResult):
(WebCore::DescriptorParsingResult::hasDensity):
(WebCore::DescriptorParsingResult::hasWidth):
(WebCore::DescriptorParsingResult::hasHeight):
(WebCore::DescriptorParsingResult::density):
(WebCore::DescriptorParsingResult::resourceWidth):
(WebCore::DescriptorParsingResult::resourceHeight):
(WebCore::DescriptorParsingResult::setResourceWidth):
(WebCore::DescriptorParsingResult::setResourceHeight):
(WebCore::DescriptorParsingResult::setDensity):
(WebCore::ImageCandidate::ImageCandidate):
(WebCore::ImageCandidate::toString):
(WebCore::ImageCandidate::url):
(WebCore::ImageCandidate::setDensity):
(WebCore::ImageCandidate::density):
(WebCore::ImageCandidate::resourceWidth):
(WebCore::ImageCandidate::srcOrigin):
(WebCore::ImageCandidate::isEmpty):
(WebCore::ImageWithScale::ImageWithScale): Deleted.
(WebCore::ImageWithScale::imageURL): Deleted.
(WebCore::ImageWithScale::scaleFactor): Deleted.

  • html/parser/ParsingUtilities.h: Renamed from Source/WebCore/html/track/ParsingUtilities.h.

(skipExactly):
(skipUntil):
(skipWhile):
(reverseSkipWhile):

Source/WTF:
I've added toInt and toFloat methods to StringView, and exposed
charactersToInt so that they can be used by
HTMLSrcsetParser.cpp.

  • wtf/text/WTFString.h:
  • wtf/text/StringView.h:

(WTF::StringView::toInt):
(WTF::StringView::toFloat):

LayoutTests:
I have changed the "invalid-descriptor" test and removed the "invalid
inputs correct src" test to reflect the fact that invalid descriptors
are now treated identically to null descriptors. That is according to
the spec.
I've also added new tests that make sure that the spec's future
compatible changes are working.

  • fast/hidpi/image-srcset-invalid-descriptor-expected.txt:
  • fast/hidpi/image-srcset-invalid-descriptor.html:
  • fast/hidpi/image-srcset-invalid-inputs-correct-src-expected.txt: Removed.
  • fast/hidpi/image-srcset-invalid-inputs-correct-src.html: Removed.
10:11 PM Changeset in webkit [169636] by jonlee@apple.com
  • 3 edits
    2 adds in trunk

Source/WebCore: Current media controls look different from previous media controls. Adjusted CSS to make media elements more similar by adding padding, adjusting hues/svg, resizing elements, etc.

​https://bugs.webkit.org/show_bug.cgi?id=133521

Patch by Wenson Hsieh <Wenson Hsieh> on 2014-06-05
Reviewed by Eric Carlson.

The main issues:

  • The widths between elements of both audio and video players are incorrect.
  • The hues of audio player elements are incorrect.
  • The hues of video player elements are content dependent (refer to full-screen mode), and this cannot be exactly replicated using just css. This patch only attempts to make them more similar.
  • The height of the audio player is incorrect.

Test: platform/ios-sim/media/audio-width.html

  • Modules/mediacontrols/mediaControlsiOS.css: Updated style elements for audio/video elements.

(::-webkit-media-controls): Enforce a minimum width on audio/video controls.
(audio): Enforce a minimum width on audio elements.
(audio::-webkit-media-controls): see below
(audio::-webkit-media-controls-wireless-playback-picker-button): modified SVG, hue and width/height.
(audio::-webkit-media-controls-panel): adjusted hue, added left/right padding.
(audio::-webkit-media-controls-fullscreen-button): removed. There's no fullscreen for audio anyways.
(audio::-webkit-media-controls-current-time-display): adjusted font size, added left padding.
(audio::-webkit-media-controls-time-remaining-display): adjusted font size, added right padding.

(video::-webkit-media-controls-panel): adjusted hue, added left/right padding.
(video::-webkit-media-controls-current-time-display): added left padding.
(video::-webkit-media-controls-time-remaining-display): added right padding.
(video::-webkit-media-controls-wireless-playback-picker-button): modified SVG, hue, opacity and width/height.
(video::-webkit-media-controls-fullscreen-button): modified SVG, hue, opacity, and width/height. In particular, made the svg seem sharper.
(video::-webkit-media-controls-play-button): modified SVG, hue, opacity, and width/height.

LayoutTests: REGRESSION (Pluginless): Media controls look different from iOS7 media controls.
​https://bugs.webkit.org/show_bug.cgi?id=133521

Patch by Wenson Hsieh <Wenson Hsieh> on 2014-06-05
Reviewed by Eric Carlson.

Tests that audio elements are properly sized. Will add tests for video players in the (near) future.

  • platform/ios-sim/media/audio-width-expected.txt: Checks that audio width tests were OK'd
  • platform/ios-sim/media/audio-width.html: Checks that minimum audio width is properly enforced
7:22 PM Changeset in webkit [169635] by commit-queue@webkit.org
  • 2 edits in trunk/Source/WebCore

Unreviewed, rolling out r169570.
​https://bugs.webkit.org/show_bug.cgi?id=133565

It breaks Baidu (Requested by benjaminp on #webkit).

Reverted changeset:

"[iOS] Start ignoring common bogus viewport parameters"
​https://bugs.webkit.org/show_bug.cgi?id=133466
​http://trac.webkit.org/changeset/169570

7:02 PM Changeset in webkit [169634] by dburkart@apple.com
  • 5 edits in branches/safari-538.34-branch/Source

Bump versioning

6:45 PM Changeset in webkit [169633] by dburkart@apple.com
  • 1 copy in tags/Safari-538.34.49

Tagging Safari-538.34.49

5:23 PM Changeset in webkit [169632] by commit-queue@webkit.org
  • 3 edits
    2 adds in trunk

AX: Add support for ARIA 1.1 'none' role
​https://bugs.webkit.org/show_bug.cgi?id=133367

Patch by Michael Timbrook <​timbrook@apple.com> on 2014-06-05
Reviewed by Chris Fleizach.

Source/WebCore:
Added support for the 'none' role by added a lookup for it to go to PresentationalRole
Test: accessibility/aria-none-role.html

  • accessibility/AccessibilityObject.cpp:

(WebCore::initializeRoleMap):

LayoutTests:
Added the tests to cover the behavior of the 'none' role

  • accessibility/aria-none-role-expected.txt: Added.
  • accessibility/aria-none-role.html: Added.
4:53 PM Changeset in webkit [169631] by jdiggs@igalia.com
  • 6 edits
    3 adds in trunk

AX: [ATK] Children of tables and lists with role='presentation' should each be exposed as if they were div elements
​https://bugs.webkit.org/show_bug.cgi?id=133133

Reviewed by Chris Fleizach.

Source/WebCore:
When a table or list has role='presentation', accessibility is "ignored"
for that element, but not for the required children. The default roles
for the children (CellRole and ListItemRole) do not make sense without
the parent, therefore map them to DivRole in ATK if they inherit the
presentational role.

Test: accessibility/aria-inherits-presentational.html

  • accessibility/AccessibilityObject.h:

(WebCore::AccessibilityObject::inheritsPresentationalRole):

  • accessibility/AccessibilityRenderObject.cpp:

(WebCore::AccessibilityRenderObject::inheritsPresentationalRole):

  • accessibility/AccessibilityRenderObject.h:
  • accessibility/atk/WebKitAccessibleWrapperAtk.cpp:

(atkRole):

LayoutTests:
Added a new test with platform-specific expectations due to differences
in accessible hierarchies.

  • accessibility/aria-inherits-presentational.html: Added.
  • platform/gtk/accessibility/aria-inherits-presentational-expected.txt: Added.
  • platform/mac/accessibility/aria-inherits-presentational-expected.txt: Added.
4:46 PM Changeset in webkit [169630] by Brent Fulgham
  • 4 edits in trunk/Source/WebCore

REGRESSION (r167962): Out of bounds read in JSC::StructureIDTable::get()
​https://bugs.webkit.org/show_bug.cgi?id=133463
<rdar://problem/17098100>

Reviewed by Dean Jackson.

Revise MediaControllerHost implementation so that it holds a Strong
reference to the JavaScript 'Controller' object it mirrors. The
MediaControllerHost is constructed when we create an HTMLMediaElement,
and is the logical place to retain ownership of the JS object so that
it does not get garbage collected as long as HTMLMediaElement exists.

  • Modules/mediacontrols/MediaControlsHost.cpp:

(WebCore::MediaControlsHost::setControllerJSObject): Create a strong
reference to the JavaScript Controller object.

  • Modules/mediacontrols/MediaControlsHost.h: Use JSC::Strong to hold a

JSObject, rather than a bare JSValue.
(WebCore::MediaControlsHost::controllerJSObject): Change name to reflect
that we are working with JSObjects, not JSValues.
(WebCore::MediaControlsHost::controllerJSValue): Deleted.
(WebCore::MediaControlsHost::setControllerJSValue): Deleted.

  • html/HTMLMediaElement.cpp:

(WebCore::HTMLMediaElement::parseAttribute): Revise for changed method
signature.

4:19 PM Changeset in webkit [169629] by g.czajkowski@samsung.com
  • 4 edits
    2 adds
    2 deletes in trunk/LayoutTests

Share platform/mac/editing/spelling/move-cursor-around-misspelled-word.html with other platforms
​https://bugs.webkit.org/show_bug.cgi?id=133208

Reviewed by Darin Adler.

Move the test to common editing/spelling since it just verifies spelling
markers and other platforms might be interested in running it as well.

Due to bug 125688, the test starts using asynchronous spellchecking.
Add additional test case which checks spelling marker after typing
the text, before any caret change.

  • editing/spelling/move-cursor-around-misspelled-word-expected.txt:

Renamed from platform/mac/editing/spelling/move-cursor-around-misspelled-word-expected.txt

  • editing/spelling/move-cursor-around-misspelled-word.html:

Renamed from platform/mac/editing/spelling/move-cursor-around-misspelled-word.html

  • platform/gtk/TestExpectations:
  • platform/mac-wk2/TestExpectations:
  • platform/win/TestExpectations:

Add failing expectation for platforms which didn't switch to
asynchronous spellchecking.

3:54 PM Changeset in webkit [169628] by fpizlo@apple.com
  • 29 edits
    4 adds
    1 delete in branches/ftlopt/Source/JavaScriptCore

Added system for inlining native functions via the FTL.
​https://bugs.webkit.org/show_bug.cgi?id=131515

Patch by Matthew Mirman <​mmirman@apple.com> on 2014-06-04
Reviewed by Filip Pizlo.

Also fixed the build to not compress the bitcode and to
include all of the relevant runtime. With GCC_GENERATE_DEBUGGING_SYMBOLS = NO,
the produced bitcode files are a 100th the size they were before.
Now we can include all of the relevant runtime files with only a 3mb overhead.
This is the same overhead as for two compressed files before,
but done more efficiently (on both ends) and with less code.

Deciding whether to inline native functions is left up to LLVM.
The entire module containing the function is linked into the current
compiled JS so that inlining the native functions shouldn't make them smaller.

Rather than loading Runtime.symtbl at runtime FTLState.cpp now generates a file
InlineRuntimeSymbolTable.h which statically builds the symbol table hash table.

  • JavaScriptCore.xcodeproj/project.pbxproj: Added back runtime files to compile.
  • build-symbol-table-index.py: Changed bitcode suffix.

Added inclusion of only tested symbols.
Added output to InlineRuntimeSymbolTable.h.

  • build-symbol-table-index.sh: Changed bitcode suffix.
  • copy-llvm-ir-to-derived-sources.sh: Removed gzip compression.
  • tested-symbols.symlst: Added.
  • dfg/DFGByteCodeParser.cpp:

(JSC::DFG::ByteCodeParser::handleCall):
Now sets the knownFunction of the call node if such a function exists
and emits a check that during runtime the callee is in fact known.

  • dfg/DFGNode.h:

Added functions to set the known function of a call node.
(JSC::DFG::Node::canBeKnownFunction): Added.
(JSC::DFG::Node::hasKnownFunction): Added.
(JSC::DFG::Node::knownFunction): Added.
(JSC::DFG::Node::giveKnownFunction): Added.

  • ftl/FTLAbbreviatedTypes.h: Added a typedef for LLVMMemoryBufferRef
  • ftl/FTLAbbreviations.h: Added some abbreviations.
  • ftl/FTLLowerDFGToLLVM.cpp:

(JSC::FTL::LowerDFGToLLVM::isInlinableSize): Added. Hardcoded threshold to 275.
(JSC::FTL::LowerDFGToLLVM::getModuleByPathForSymbol): Added.
(JSC::FTL::LowerDFGToLLVM::getFunctionBySymbol): Added.
(JSC::FTL::LowerDFGToLLVM::possiblyCompileInlineableNativeCall): Added.
(JSC::FTL::LowerDFGToLLVM::compileCallOrConstruct):
Added call to possiblyCompileInlineableNativeCall

  • ftl/FTLOutput.h:

(JSC::FTL::Output::allocaName): Added. Useful for debugging.

  • ftl/FTLState.cpp:

(JSC::FTL::State::State): Added an include for InlineRuntimeSymbolTable.h

  • ftl/FTLState.h: Added symbol table hash table.
  • ftl/FTLCompile.cpp:

(JSC::FTL::compile): Added inlining and dead function elimination passes.

  • heap/HandleStack.h: Added JS_EXPORT_PRIVATE to a few functions to get inlining to compile.
  • llvm/InitializeLLVMMac.mm: Deleted.
  • llvm/InitializeLLVMMac.cpp: Added.
  • llvm/LLVMAPIFunctions.h: Added macros to include Bitcode parsing and linking functions.
  • llvm/LLVMHeaders.h: Added includes for Bitcode parsing and linking.
  • runtime/BundlePath.h: Added.
  • runtime/BundlePath.mm: Added.
  • runtime/DateInstance.h: Added JS_EXPORT_PRIVATE to a few functions to get inlining to compile.
  • runtime/DateInstance.h: ditto.
  • runtime/DateConversion.h: ditto.
  • runtime/ExceptionHelpers.h: ditto.
  • runtime/JSCJSValue.h: ditto.
  • runtime/JSArray.h: ditto.
  • runtime/JSDateMath.h: ditto.
  • runtime/JSObject.h: ditto.
  • runtime/JSObject.h: ditto.
  • runtime/RegExp.h: ditto.
  • runtime/Structure.h: ditto.
  • runtime/Options.h: Added maximumLLVMInstructionCountForNativeInlining.
2:40 PM Changeset in webkit [169627] by psolanki@apple.com
  • 3 edits in trunk/Source/WebKit/mac

Remove unsused method _setAcceleratedImageDecoding
​https://bugs.webkit.org/show_bug.cgi?id=133554

Reviewed by Jon Honeycutt.

  • WebView/WebView.mm:

(+[WebView _setAcceleratedImageDecoding:]): Deleted.

  • WebView/WebViewPrivate.h:
2:36 PM Changeset in webkit [169626] by benjamin@webkit.org
  • 2 edits in trunk/Source/WebKit2

[iOS][WK2] WebPage sometimes forces two resize events on animated resize
​https://bugs.webkit.org/show_bug.cgi?id=133534

Patch by Benjamin Poulain <​bpoulain@apple.com> on 2014-06-05
Reviewed by Enrica Casucci.

When I updated the patch for resize events (r168556) to use the unobscured rect in scrollview coordinates,
I forgot to update a snippet of code in dynamicViewportSizeUpdate(). Since that code was still using
the scaled unobscured rect, there was sometimes a difference of 1 pixel between the size set by that
call site and the others.

This path nukes the bad code and use the unobscured rect in scrollview coordinates given by the UIProcess.

  • WebProcess/WebPage/ios/WebPageIOS.mm:

(WebKit::WebPage::dynamicViewportSizeUpdate):

2:14 PM Changeset in webkit [169625] by benjamin@webkit.org
  • 26 edits in trunk/Source

[iOS][WK2] Add device orientation
​https://bugs.webkit.org/show_bug.cgi?id=133530

Source/WebCore:

Patch by Benjamin Poulain <​bpoulain@apple.com> on 2014-06-05
Reviewed by Tim Horton.

Previously, WebCore was getting the orientation directly from the WebKit layer.
This had to be done during the frame creation and was very fragile.

Frame is changed to pull the value from the chrome client. That way we avoid races
and we can maintain a single value for all frames.

Sending the values to all the subframes is aslo done here so that we do not have to repeat
it in both WebKit layers.

  • WebCore.exp.in:
  • loader/EmptyClients.h:
  • page/ChromeClient.h:
  • page/Frame.cpp:

(WebCore::Frame::Frame):
(WebCore::Frame::orientationChanged):
(WebCore::Frame::orientation):
(WebCore::Frame::sendOrientationChangeEvent): Deleted.

  • page/Frame.h:

(WebCore::Frame::orientation): Deleted.

Source/WebKit/ios:

Patch by Benjamin Poulain <​bpoulain@apple.com> on 2014-06-05
Reviewed by Tim Horton.

Change WebKit1 to provide the deviceOrientation as pull instead of push.

  • DefaultDelegates/WebDefaultUIKitDelegate.m:

(-[WebDefaultUIKitDelegate deviceOrientation]):

  • WebCoreSupport/WebChromeClientIOS.h:
  • WebCoreSupport/WebChromeClientIOS.mm:

(WebChromeClientIOS::deviceOrientation):

  • WebView/WebUIKitDelegate.h:

Source/WebKit/mac:

Patch by Benjamin Poulain <​bpoulain@apple.com> on 2014-06-05
Reviewed by Tim Horton.

  • WebView/WebFrame.mm:

(-[WebFrame deviceOrientationChanged]):
(-[WebFrame sendOrientationChangeEvent:]):

  • WebView/WebFramePrivate.h:

Source/WebKit2:
<rdar://problem/16680041>

Patch by Benjamin Poulain <​bpoulain@apple.com> on 2014-06-05
Reviewed by Tim Horton.

Add device orientation for WebKit2.

For the public API, WKWebView get the updates through the notification UIWindowDidRotateNotification.
We do not have any control over how the API is used, but we can expect the size will be updated before
the end of rotation in most cases and the events should be sent in the right order.

For Safari, we use an override to support animated resize (and some tabs corner cases).

On the WebProcess side, we just get the new orientation directly or on DynamicViewportSizeUpdate.

  • UIProcess/API/Cocoa/WKWebView.mm:

(deviceOrientationForUIInterfaceOrientation):
(deviceOrientation):
(-[WKWebView initWithFrame:configuration:]):
(-[WKWebView _windowDidRotate:]):
(-[WKWebView _setInterfaceOrientationOverride:]):
(-[WKWebView _interfaceOrientationOverride]):
(-[WKWebView _beginAnimatedResizeWithUpdates:]):

  • UIProcess/API/Cocoa/WKWebViewPrivate.h:
  • UIProcess/WebPageProxy.cpp:

(WebKit::WebPageProxy::WebPageProxy):

  • UIProcess/WebPageProxy.h:

(WebKit::WebPageProxy::deviceOrientation):

  • UIProcess/ios/WebPageProxyIOS.mm:

(WebKit::WebPageProxy::dynamicViewportSizeUpdate):
(WebKit::WebPageProxy::setDeviceOrientation):

  • WebProcess/WebCoreSupport/WebChromeClient.h:
  • WebProcess/WebCoreSupport/ios/WebChromeClientIOS.mm:

(WebKit::WebChromeClient::deviceOrientation):

  • WebProcess/WebPage/WebPage.cpp:

(WebKit::WebPage::WebPage):

  • WebProcess/WebPage/WebPage.h:

(WebKit::WebPage::deviceOrientation):

  • WebProcess/WebPage/WebPage.messages.in:
  • WebProcess/WebPage/ios/WebPageIOS.mm:

(WebKit::WebPage::setDeviceOrientation):
(WebKit::WebPage::dynamicViewportSizeUpdate):

2:10 PM Changeset in webkit [169624] by ap@apple.com
  • 2 edits in trunk/LayoutTests

media/video-rtl.html is flaky
​https://bugs.webkit.org/show_bug.cgi?id=133363

  • platform/mac/TestExpectations: Marked as such.
2:00 PM Changeset in webkit [169623] by timothy_horton@apple.com
  • 2 edits in trunk/Source/WebCore

Fix the !CACHE_SUBIMAGES build

  • platform/graphics/cg/GraphicsContextCG.cpp:

(WebCore::GraphicsContext::drawNativeImage):

12:49 PM Changeset in webkit [169622] by oliver@apple.com
  • 2 edits in trunk/Source/WebKit2

Enable SANDBOX_EXTENSIONS build flag universally on cocoa
​https://bugs.webkit.org/show_bug.cgi?id=133556

Reviewed by Alexey Proskuryakov.

Remove unnecessary iOS exception.

  • WebKit2Prefix.h:
11:40 AM Changeset in webkit [169621] by commit-queue@webkit.org
  • 2 edits in trunk/Source/WebCore

CSS JIT: Clean up StackAllocator
​https://bugs.webkit.org/show_bug.cgi?id=133506

Patch by Yusuke Suzuki <Yusuke Suzuki> on 2014-06-05
Reviewed by Benjamin Poulain.

Cleaning up CSS JIt StackAllocator code. Sharing push / pop code and
using information provided by MacroAssembler more aggressively
to drop ifdefs.

  • cssjit/StackAllocator.h:

(WebCore::StackAllocator::allocateUninitialized):
(WebCore::StackAllocator::push):
(WebCore::StackAllocator::pop):
(WebCore::StackAllocator::popAndDiscard):
(WebCore::StackAllocator::popAndDiscardUpTo):
(WebCore::StackAllocator::alignStackPreFunctionCall):
(WebCore::StackAllocator::unalignStackPostFunctionCall):
(WebCore::StackAllocator::stackUnitInBytes):

11:18 AM Changeset in webkit [169620] by Alan Bujtas
  • 4 edits
    2 adds in trunk

Subpixel rendering: border-radius painting falls back to rectangle when the snapped rounded rect becomes non-renderable.
​https://bugs.webkit.org/show_bug.cgi?id=133491

Reviewed by Simon Fraser.

Pixel snapping can change the rectangle's size when it is on a certain subpixel position. (usually it does not)
This patch ensures that the snapped rect is still renderable by adjusting the radii as well.

Source/WebCore:
Test: fast/borders/hidpi-border-radius-with-subpixel-margin-not-renderable.html

  • platform/graphics/RoundedRect.cpp:

(WebCore::RoundedRect::Radii::scale):
(WebCore::RoundedRect::pixelSnappedRoundedRectForPainting):

  • platform/graphics/RoundedRect.h:

LayoutTests:

  • fast/borders/hidpi-border-radius-with-subpixel-margin-not-renderable-expected.html: Added.
  • fast/borders/hidpi-border-radius-with-subpixel-margin-not-renderable.html: Added.
10:21 AM Changeset in webkit [169619] by commit-queue@webkit.org
  • 2 edits in trunk/Source/WebCore

Unreviewed, rolling out r169587.
​https://bugs.webkit.org/show_bug.cgi?id=133552

Calls a pure virtual method, crashes (Requested by ap on
#webkit).

Reverted changeset:

"WebKit should adopt journal_mode=wal for all SQLite
databases."
​https://bugs.webkit.org/show_bug.cgi?id=133496
​http://trac.webkit.org/changeset/169587

10:21 AM Changeset in webkit [169618] by ap@apple.com
  • 3 edits in trunk/Source/WebKit2

[Mac] Tweak plug-in sandbox profiles
​https://bugs.webkit.org/show_bug.cgi?id=133549

Fix <rdar://problem/17068055> and <rdar://problem/17148837>.

Reviewed by Sam Weinig and Simon Cooper.

  • Resources/PlugInSandboxProfiles/com.apple.ist.ds.appleconnect.webplugin.sb:
  • Resources/PlugInSandboxProfiles/com.microsoft.SilverlightPlugin.sb:
9:23 AM Changeset in webkit [169617] by fred.wang@free.fr
  • 3 edits
    2 adds in trunk

Apply Character-level mirroring to stretchy operators in RTL mode.
​https://bugs.webkit.org/show_bug.cgi?id=130839

Reviewed by Chris Fleizach.

Source/WebCore:
We add support for character-level mirroring for stretchy MathML operators in RTL mode.
Glyph-level mirroring (via the rtlm OpenType feature) will be handled in bug 130840.

Test: mathml/presentation/stretchy-rtl.html

  • rendering/mathml/RenderMathMLOperator.cpp: reorganize the Unicode stretchyCharacters table to make mirroring easy.

(WebCore::RenderMathMLOperator::computePreferredLogicalWidths): pass the right boolean to glyphDataForCharacter.
(WebCore::RenderMathMLOperator::getDisplayStyleLargeOperator): ditto
(WebCore::RenderMathMLOperator::findStretchyData): ditto and change the index for the stretchyCharacters fallback.
(WebCore::RenderMathMLOperator::updateStyle): pass the right boolean to glyphDataForCharacter.
(WebCore::RenderMathMLOperator::paint): remove the FIXME comment.
(WebCore::RenderMathMLOperator::paintVerticalGlyphAssembly): switch the leading/trailing space in RTL mode.

LayoutTests:
Add some tests to verify that stretchy operators are correctly mirrored in RTL mode.

  • mathml/presentation/stretchy-rtl-expected.html: Added.
  • mathml/presentation/stretchy-rtl.html: Added.
9:16 AM Changeset in webkit [169616] by zoltan@webkit.org
  • 1 edit
    2 adds in trunk/LayoutTests

[CSS Regions][CSS Shapes] Add Shapes tests which flow through multiple regions
​https://bugs.webkit.org/show_bug.cgi?id=133487

Reviewed by Mihnea Ovidenie.

The test verifies if the content with the shape wraps correctly through a region chain.

  • fast/regions/shape-outside-floats/regions-shape-outside-content-flows-multiple-regions-expected.html: Added.
  • fast/regions/shape-outside-floats/regions-shape-outside-content-flows-multiple-regions.html: Added.
8:59 AM Changeset in webkit [169615] by Alan Bujtas
  • 6 edits
    2 adds in trunk

Subpixel rendering: Buttons in default media controls shift vertically when controls fade in or out.
​https://bugs.webkit.org/show_bug.cgi?id=133488

Reviewed by Simon Fraser.

Filter effects are not subpixel aware yet. They paint to integrally sized image buffers.
In order to ensure that the content is rendered to the the same position on the filter
source context as if there was no filter, the repaint rects need to be adjusted to match with
the internal filter snapping (enclosingIntRect).
This is a temporary change until after filter effects get subpixelized.

However, this changeset regresses RenderNameFlow clipping when the content is on subpixel position
and filter effect is present. Instead of adding the failing test cases to TestExpectations, I removed
the video controls from the <video> to make the tests pass, when possible (they just happen to have <video>)
and filed webkit.org/b/133523 to track the regression.

Source/WebCore:

Test: css3/filters/filter-is-on-subpixel-position.html

  • rendering/RenderLayer.cpp:

(WebCore::RenderLayer::setupFilters):

LayoutTests:

  • css3/filters/filter-is-on-subpixel-position-expected.html: Added.
  • css3/filters/filter-is-on-subpixel-position.html: Added.
  • fast/regions/inline-block-inside-anonymous-overflow.html:
  • fast/regions/inline-block-overflow.html:
6:45 AM Changeset in webkit [169614] by commit-queue@webkit.org
  • 4 edits in trunk/Tools

[GTK] run-perf-tests should run the tests inside the jhbuild shell.
​https://bugs.webkit.org/show_bug.cgi?id=133533

Patch by Carlos Alberto Lopez Perez <​clopez@igalia.com> on 2014-06-05
Reviewed by Carlos Garcia Campos.

  • Scripts/webkitpy/common/multiprocessing_bootstrap.py:

(run): Remove the jhbuild-wrapper wrapper from here. It should be
on Scripts/webkitpy/port/gtk.py.

  • Scripts/webkitpy/port/base.py:

(Port.set_option): Add a set_option(name, value) function to set
the optparse parameter "name" to the value "value", even if the
parameter "name" is already initialized.

  • Scripts/webkitpy/port/gtk.py:

(GtkPort.init): Set the default wrapper to jhbuild-wrapper.
If another wrapper is specified via the option "-l --wrapper=x",
then prepend the jhbuild-wrapper to the one specified.

5:47 AM Changeset in webkit [169613] by commit-queue@webkit.org
  • 2 edits in trunk/LayoutTests

[EFL] Change expectations for three layout tests.
​https://bugs.webkit.org/show_bug.cgi?id=133538

Unreviewed EFL gardening.

Patch by Andrzej Badowski <​a.badowski@samsung.com> on 2014-06-05

  • platform/efl-wk2/TestExpectations:
4:30 AM Changeset in webkit [169612] by commit-queue@webkit.org
  • 2 edits in trunk/Source/WebKit2

Resolve compiler error (call of overloaded 'schedule' function is ambiguous)
​https://bugs.webkit.org/show_bug.cgi?id=133467

Patch by Tanay C <​tanay.c@samsung.com> on 2014-06-05
Reviewed by Carlos Garcia Campos.

  • Shared/Downloads/soup/DownloadSoup.cpp:

(WebKit::DownloadClient::handleResponseLater): Explicitly casting second parameter to std::function<void()>

2:41 AM Changeset in webkit [169611] by commit-queue@webkit.org
  • 3 edits
    2 adds in trunk

[MediaStream] MediaStream.addTrack should not check for tracks ended state.
​https://bugs.webkit.org/show_bug.cgi?id=133510

Patch by Kiran <​kiran.guduru@samsung.com> on 2014-06-05
Reviewed by Eric Carlson.

Source/WebCore:
Adding and removing of tracks to a MediaStream should not have dependency
on tracks state. So the check for MediaStreamTrack's ended state has been
removed while adding a track in MediaStreamPrivate.cpp.

Test: fast/mediastream/MediaStream-add-ended-tracks.html

  • platform/mediastream/MediaStreamPrivate.cpp:

(WebCore::MediaStreamPrivate::addTrack):

LayoutTests:

  • fast/mediastream/MediaStream-add-ended-tracks-expected.txt: Added.
  • fast/mediastream/MediaStream-add-ended-tracks.html: Added.
2:25 AM Changeset in webkit [169610] by yoon@igalia.com
  • 3 edits
    1 delete in trunk/Source/WebCore

[GTK] Remove ScrollViewGtk.cpp
​https://bugs.webkit.org/show_bug.cgi?id=133535

Reviewed by Carlos Garcia Campos.

Because GTK+ WK1 was removed, we don't have to consider special cases
handled in ScrollViewGtk.cpp.

No new tests are necessary because there is no behavior change.

  • PlatformGTK.cmake:
  • platform/ScrollView.cpp:

(WebCore::ScrollView::setScrollbarModes):
(WebCore::ScrollView::visibleContentRectInternal):

  • platform/gtk/ScrollViewGtk.cpp: Removed.
2:23 AM Changeset in webkit [169609] by commit-queue@webkit.org
  • 3 edits in trunk/LayoutTests

[EFL] Add expectations for two accessibility layout tests.
​https://bugs.webkit.org/show_bug.cgi?id=133485

Unreviewed EFL gardening.

Patch by Andrzej Badowski <​a.badowski@samsung.com> on 2014-06-05

  • platform/efl-wk2/TestExpectations:
  • platform/efl-wk2/accessibility/lists-expected.txt:
12:31 AM Changeset in webkit [169608] by bshafiei@apple.com
  • 5 edits in trunk/Source

Versioning.

Note: See TracTimeline for information about the timeline view.