WebKit-https.git
6 years agoREGRESSION(r147602): Search text field doesn't render selection when it has some...
rniwa@webkit.org [Tue, 18 Jun 2013 21:25:07 +0000 (21:25 +0000)]
REGRESSION(r147602): Search text field doesn't render selection when it has some :focus rules
https://bugs.webkit.org/show_bug.cgi?id=117747

Reviewed by Kent Tamura.

.:

* ManualTests/search-select-all-with-focus-style.html: Added.

Source/WebCore:

Merge https://chromium.googlesource.com/chromium/blink/+/7ff656c8239ce3b125246abcc8b149a603fcff28

r147612 added setNeedsLayout(true, MarkOnlyThis) for
TextControlInnerTextElement renderer. It's ok for input[type=text]
because RenderTextControlSingleLine also has needsLayout flag, and it
has only the inner text renderer as a child.

As for input[type=search], it doesn't work. Renderer structure for
input[type=search] is:

RenderTextControlSingleLine
└Renderer for TextControlInnerContainer
  └Renderer for TextControlInnerElement
    └Renderer for TextControlInnerTextElement

After r147612, there is a case that only RenderTextControlSingleLine
and TextControlInnerTextElement renderer have needsLayout flag, and
others don't.  It won't trigger layout for
TextControlInnerTextElement renderer.

We should set needsLayout flags of TextControlInnerContainer renderer
and TextControlInnerElement renderer by MarkContainingBlockChain.

Manual Tests: search-select-all-with-focus-style.html
(I tried but couldn't make a test that works in DRT)

* rendering/RenderTextControlSingleLine.cpp:
(WebCore::RenderTextControlSingleLine::styleDidChange):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151695 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoAttempted build fix for non-Mac.
dino@apple.com [Tue, 18 Jun 2013 20:40:09 +0000 (20:40 +0000)]
Attempted build fix for non-Mac.

* platform/graphics/SimpleFontData.cpp:
(WebCore::SimpleFontData::nonSyntheticItalicFontData):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151694 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoSynthesized vertical italics on rotated glyphs are transformed incorrectly
dino@apple.com [Tue, 18 Jun 2013 20:28:58 +0000 (20:28 +0000)]
Synthesized vertical italics on rotated glyphs are transformed incorrectly
https://bugs.webkit.org/show_bug.cgi?id=117726

Reviewed by Enrica Casucci.

Source/WebCore:

If a font does not have italic glyphs, and is in a vertical flow, then
the synthesis of italics was skewing in the wrong direction. While here,
ensure that no CJK text will get italics in vertical layout.

This is a topic under discussion in the CSS Working Group. The resolution
at the June 2013 face-to-face was that:
- italics in Japanese (and Chinese) text are rare in vertical layout
- most people wanted Latin text to italicize in the direction of the line flow

The final behaviour was left undefined, but we consider the change here
to be sufficient for now. In all cases, the best solution is to avoid
synthesized italics by specifying a font that includes such forms.

Tests: fast/text/international/synthesized-italic-vertical-latin.html
       fast/text/international/synthesized-italic-vertical.html

* platform/graphics/FontGlyphs.cpp:
(WebCore::glyphDataAndPageForCJKCharacterWithoutSyntheticItalic): New function
    that makes a new glyph data pair that is marked as NOT using a synthesized italic.
(WebCore::FontGlyphs::glyphDataAndPageForCharacter): If this is a CJK character and
    has a synthesized oblique, then run it through the function above to make a new glyph data pair.
* platform/graphics/SimpleFontData.cpp:
(WebCore::SimpleFontData::nonSyntheticItalicFontData): Creates/returns a new SimpleFontData that will
    not synthesize italics.
* platform/graphics/SimpleFontData.h: New method nonSyntheticItalicFontData.
* platform/graphics/mac/FontMac.mm:
(WebCore::Font::drawGlyphs): This is the actual bug fix. If you are in a vertical flow, the
    skew direction should be downwards, not leftwards.

LayoutTests:

Two new tests. One ensures that CJK characters do not get italics
when vertical. The other ensures that Latin text does get italics in
a font that is not synthesized, but unfortunately has to be a pixel test.
The latter is skipped on non-Mac since they don't support
synthesized obliques in FontPlatformData.

* fast/text/international/synthesized-italic-vertical-expected.html: Added.
* fast/text/international/synthesized-italic-vertical-latin.html: Added.
* fast/text/international/synthesized-italic-vertical.html: Added.
* platform/efl/TestExpectations:
* platform/gtk/TestExpectations:
* platform/mac/fast/text/international/synthesized-italic-vertical-latin-expected.txt: Added.
* platform/qt/TestExpectations:
* platform/win/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151693 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed. Build fix for Mac and Win
roger_fong@apple.com [Tue, 18 Jun 2013 20:28:26 +0000 (20:28 +0000)]
Unreviewed. Build fix for Mac and Win

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151692 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoRe-implement WebFrameNetworkingContext.
roger_fong@apple.com [Tue, 18 Jun 2013 20:14:32 +0000 (20:14 +0000)]
Re-implement WebFrameNetworkingContext.
<rdar://problem/13174821>.

Reviewed by Alexey Proskuryakov.

* WebProcess/WebCoreSupport/mac/WebFrameNetworkingContext.h: Added.
(WebKit::WebFrameNetworkingContext::create):
(WebKit::WebFrameNetworkingContext::WebFrameNetworkingContext):
* WebProcess/WebCoreSupport/mac/WebFrameNetworkingContext.mm: Added.
(WebKit::WebFrameNetworkingContext::setPrivateBrowsingStorageSessionIdentifierBase):
(WebKit::WebFrameNetworkingContext::ensurePrivateBrowsingSession):
(WebKit::WebFrameNetworkingContext::destroyPrivateBrowsingSession):
(WebKit::WebFrameNetworkingContext::setCookieAcceptPolicyForAllContexts):
(WebKit::WebFrameNetworkingContext::needsSiteSpecificQuirks):
(WebKit::WebFrameNetworkingContext::localFileContentSniffingEnabled):
(WebKit::WebFrameNetworkingContext::scheduledRunLoopPairs):
(WebKit::WebFrameNetworkingContext::sourceApplicationAuditData):
(WebKit::WebFrameNetworkingContext::blockedError):
(WebKit::WebFrameNetworkingContext::storageSession):
* WebCoreSupport/WebFrameNetworkingContext.h: Added.
(WebFrameNetworkingContext::create):
(WebFrameNetworkingContext::WebFrameNetworkingContext):
(WebFrameNetworkingContext::userAgent):
* WebCoreSupport/WebFrameNetworkingContext.mm: Added.
(WebFrameNetworkingContext::ensurePrivateBrowsingSession):
(WebFrameNetworkingContext::destroyPrivateBrowsingSession):
(WebFrameNetworkingContext::needsSiteSpecificQuirks):
(WebFrameNetworkingContext::localFileContentSniffingEnabled):
(WebFrameNetworkingContext::scheduledRunLoopPairs):
(WebFrameNetworkingContext::sourceApplicationAuditData):
(WebFrameNetworkingContext::blockedError):
(WebFrameNetworkingContext::storageSession):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151691 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago <rdar://problem/13174821> Remove files with an incorrect license.
ap@apple.com [Tue, 18 Jun 2013 20:05:08 +0000 (20:05 +0000)]
    <rdar://problem/13174821> Remove files with an incorrect license.

        Rubber-stamped by Maciej Stachowiak.

        This will break the build, stay tuned for a fix coming soon.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151690 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoPackedIntVector: check template arguments at compile time
mikhail.pozdnyakov@intel.com [Tue, 18 Jun 2013 19:58:32 +0000 (19:58 +0000)]
PackedIntVector: check template arguments at compile time
https://bugs.webkit.org/show_bug.cgi?id=117737

Use 'COMPILE_ASSERT' instead of 'ASSERT' in PackedIntVector class
constructor.

Reviewed by Sam Weinig.

* wtf/PackedIntVector.h:
(WTF::PackedIntVector::PackedIntVector):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151689 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoIf the tree outline is processing a selection currently, then don't change the selection.
timothy@apple.com [Tue, 18 Jun 2013 19:38:01 +0000 (19:38 +0000)]
If the tree outline is processing a selection currently, then don't change the selection.

This is needed to allow breakpoints tree elements to be selected without jumping back to
selecting the resource tree element.

https://bugs.webkit.org/show_bug.cgi?id=117746

Reviewed by Joseph Pecoraro.

* UserInterface/Main.js:
(WebInspector._revealAndSelectRepresentedObjectInNavigationSidebar):
* UserInterface/TreeOutline.js:
(TreeElement.prototype.select):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151688 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUpdate breakpoints to match Xcode 5.
timothy@apple.com [Tue, 18 Jun 2013 18:49:28 +0000 (18:49 +0000)]
Update breakpoints to match Xcode 5.

This updates the breakpoint images, line colors, and moves breakpoints toggle button
to the Debugger sidebar navigation bar.

https://bugs.webkit.org/show_bug.cgi?id=117723

Reviewed by Joseph Pecoraro.

* Localizations/en.lproj/localizedStrings.js: Updated.
* UserInterface/DebuggerSidebarPanel.css: Removed custom toggle button rules.
* UserInterface/DebuggerSidebarPanel.js:
(WebInspector.DebuggerSidebarPanel):
(WebInspector.DebuggerSidebarPanel.prototype._breakpointsToggleButtonClicked):
* UserInterface/Images/Breakpoint.png: Updated.
* UserInterface/Images/Breakpoints.pdf: Added.
* UserInterface/Images/Breakpoint@2x.png: Updated.
* UserInterface/Images/BreakpointButton.pdf: Updated.
* UserInterface/Images/BreakpointInactive.png: Updated.
* UserInterface/Images/BreakpointInactive@2x.png: Updated.
* UserInterface/Images/BreakpointInactiveButton.pdf: Updated.
* UserInterface/Images/InstructionPointer.png: Updated.
* UserInterface/Images/InstructionPointer@2x.png: Updated.
* UserInterface/SourceCodeTextEditor.css:
(.source-code.text-editor > .CodeMirror .error): Updated background-color and removed box-shadow.
(.source-code.text-editor > .CodeMirror .warning): Ditto.
* UserInterface/TextEditor.css:
(.text-editor > .CodeMirror .has-breakpoint .CodeMirror-linenumber::before): Changed left position to 0.
(.text-editor > .CodeMirror .execution-line .CodeMirror-linenumber::after): Updated position.
(.text-editor > .CodeMirror .execution-line): Updated background-color and removed box-shadow.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151687 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoThe page title of technical-articles.html is incorrect.
commit-queue@webkit.org [Tue, 18 Jun 2013 17:51:42 +0000 (17:51 +0000)]
The page title of technical-articles.html is incorrect.
https://bugs.webkit.org/show_bug.cgi?id=117727

Patch by Denis Nomiyama <d.nomiyama@samsung.com> on 2013-06-18
Reviewed by Ryosuke Niwa.

Changed the page title of technical-articles.html to "WebKit Technical Articles".

* coding/technical-articles.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151686 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed, rolling out r151549.
commit-queue@webkit.org [Tue, 18 Jun 2013 17:38:07 +0000 (17:38 +0000)]
Unreviewed, rolling out r151549.
http://trac.webkit.org/changeset/151549
https://bugs.webkit.org/show_bug.cgi?id=117741

broke selection in the web inspector source after scrolling
(Requested by smfr on #webkit).

Source/WebCore:

* rendering/RenderBox.cpp:
(WebCore::RenderBox::computeRectForRepaint):
* rendering/RenderLayerModelObject.cpp:
* rendering/RenderLayerModelObject.h:

LayoutTests:

* fast/repaint/change-overflow-and-display-of-relative-expected.txt: Removed.
* fast/repaint/change-overflow-and-display-of-relative.html: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151685 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago<video> element delays document load event for ~ 3 seconds
commit-queue@webkit.org [Tue, 18 Jun 2013 15:16:02 +0000 (15:16 +0000)]
<video> element delays document load event for ~ 3 seconds
https://bugs.webkit.org/show_bug.cgi?id=90272

Patch by Santosh Mahto <santosh.ma@samsung.com> on 2013-06-18
Reviewed by Eric Carlson.

Source/WebCore:

Test: media/media-preload-no-delay-loadevent.html

* html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::prepareForLoad):
No need to delay document load event here  when preload="none"

LayoutTests:

* media/media-preload-no-delay-loadevent-expected.txt: Added.
* media/media-preload-no-delay-loadevent.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151684 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[GTK] Remove comment about possibility of handling ObjCPolymorphic (it's used by...
commit-queue@webkit.org [Tue, 18 Jun 2013 13:04:58 +0000 (13:04 +0000)]
[GTK] Remove comment about possibility of handling ObjCPolymorphic (it's used by ObjC bindings only)
https://bugs.webkit.org/show_bug.cgi?id=117546

Patch by Diego Pino Garcia <dpino@igalia.com> on 2013-06-18
Reviewed by Xan Lopez.

According to the WebKit IDL spec the modifier ObjCPolymorphic only
applies to the Objective-C bindings generator

* bindings/scripts/CodeGeneratorGObject.pm:
(IsPolymorphic): Remove comment. Search parameter 'type' in array, instead of
concatening a series of 'or' conditionals.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151682 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoFix after r151673
commit-queue@webkit.org [Tue, 18 Jun 2013 12:24:02 +0000 (12:24 +0000)]
Fix after r151673
https://bugs.webkit.org/show_bug.cgi?id=116042

Patch by Víctor Manuel Jáquez Leal <vjaquez@igalia.com> on 2013-06-18
Reviewed by Philippe Normand.

Removed a spurious semicolon in the video sink caps definition.

No new tests, no behavior change.

* platform/graphics/gstreamer/VideoSinkGStreamer.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151681 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[WebSpeech] Speech Recognition requires convertValue support in JSDictionary
commit-queue@webkit.org [Tue, 18 Jun 2013 11:23:37 +0000 (11:23 +0000)]
[WebSpeech] Speech Recognition requires convertValue support in JSDictionary
https://bugs.webkit.org/show_bug.cgi?id=117731

Patch by Praveen R Jadhav <praveen.j@samsung.com> on 2013-06-18
Reviewed by Christophe Dumez.

Build fails after enabling ENABLE_SCRIPTED_SPEECH feature. This patch
resolves the issue.

No new tests. No change in behaviour.

* bindings/js/JSDictionary.cpp:
(WebCore::JSDictionary::convertValue):
* bindings/js/JSDictionary.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151680 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed. Update NEWS and Versions.m4 for 2.1.2 release.
carlosgc@webkit.org [Tue, 18 Jun 2013 10:30:22 +0000 (10:30 +0000)]
Unreviewed. Update NEWS and Versions.m4 for 2.1.2 release.

.:

* Source/autotools/Versions.m4: Bump version numbers.

Source/WebKit/gtk:

* NEWS: Added release notes for 2.1.2.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151679 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Qt] Unreviewed gardening. Rebaseline and skipping some tests.
zarvai@inf.u-szeged.hu [Tue, 18 Jun 2013 09:35:40 +0000 (09:35 +0000)]
[Qt] Unreviewed gardening. Rebaseline and skipping some tests.

Patch by Gabor Abraham <abrhm@inf.u-szeged.hu> on 2013-06-18

* platform/qt-5.0-wk2/fast/repaint/table-cell-collapsed-border-scroll-expected.png: Added.
* platform/qt/TestExpectations:
* platform/qt/fast/repaint/table-cell-collapsed-border-scroll-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151678 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[BlackBerry] only notify download client when download error
mary.wu@torchmobile.com.cn [Tue, 18 Jun 2013 09:13:13 +0000 (09:13 +0000)]
[BlackBerry] only notify download client when download error
https://bugs.webkit.org/show_bug.cgi?id=117687

Reviewed by Rob Buis.

When met network error in downloading, we notify the error to download
client through download stream, we shouldn't notify page client the error.
RIM JIRA 419985

* platform/network/blackberry/NetworkJob.cpp:
(WebCore::NetworkJob::shouldNotifyClientFailed):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151677 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoShader compiler not properly configured for GLES on cairo based ports
mario@webkit.org [Tue, 18 Jun 2013 08:35:20 +0000 (08:35 +0000)]
Shader compiler not properly configured for GLES on cairo based ports
https://bugs.webkit.org/show_bug.cgi?id=117705

Reviewed by Martin Robinson.

Configure the shader compiler with SH_ESSL_OUTPUT for GLESv2
compliant platforms, and with SH_GLSL_OUTPUT otherwise.

* platform/graphics/cairo/GraphicsContext3DCairo.cpp:
(WebCore::GraphicsContext3D::GraphicsContext3D):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151676 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agotouching any idl rebuilds all derived sources
ch.dumez@sisa.samsung.com [Tue, 18 Jun 2013 08:29:20 +0000 (08:29 +0000)]
touching any idl rebuilds all derived sources
https://bugs.webkit.org/show_bug.cgi?id=117708

Reviewed by Kentaro Hara.

Fix preprocess-idls.pl script to update the following files only
if they have changed:
DerivedSources/WebCore/supplemental_dependency.tmp
DerivedSources/WebCore/DOMWindowConstructors.idl
DerivedSources/WebCore/WorkerContextConstructors.idl

This avoids triggering uselessly bindings generation for all IDL
files whenever an IDL file is touched.

No new tests, no behavior change.

* bindings/scripts/preprocess-idls.pl:
(WriteFileIfChanged):
(GeneratePartialInterface):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151675 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[GTK][GStreamer] Fullscreen option in video element context menu not working
commit-queue@webkit.org [Tue, 18 Jun 2013 08:18:40 +0000 (08:18 +0000)]
[GTK][GStreamer] Fullscreen option in video element context menu not working
https://bugs.webkit.org/show_bug.cgi?id=105191

.:

Fullscreen with native controls is outdated and even broken in
[GTK][WK2], so they are deactivated for now.

Patch by Xabier Rodriguez Calvar <calvaris@igalia.com> on 2013-06-18
Reviewed by Philippe Normand.

* Source/autotools/SetupAutoconfHeader.m4: Removed the use of
fullscreen native media controls.

LayoutTests:

Patch by Xabier Rodriguez Calvar <calvaris@igalia.com> on 2013-06-18
Reviewed by Philippe Normand.

* platform/gtk-wk1/TestExpectations:
fullscreen/video-controls-drag.html does not time out anymore.
* platform/gtk-wk2/TestExpectations:
media/video-display-aspect-ratio.html crashes sometimes.
* platform/gtk/TestExpectations:
media/video-playing-and-pause.html was already flaky.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151674 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[GStreamer] [texmap] upload a video buffer into the video texture
commit-queue@webkit.org [Tue, 18 Jun 2013 08:09:55 +0000 (08:09 +0000)]
[GStreamer] [texmap] upload a video buffer into the video texture
https://bugs.webkit.org/show_bug.cgi?id=116042

Patch by Víctor Manuel Jáquez Leal <vjaquez@igalia.com> on 2013-06-18
Reviewed by Philippe Normand.

This patch prepares more quickly the texture when the video frame is
already in the GPU memory.

It is done using a new buffer's metadata available in GStreamer 1.2,
and its purpose is to upload buffers into a OpenGL texture.

If the decoder provides buffers with this metadata, the buffer will be
uploaded into the texture used for the video display and it will be
rendered, avoiding a expensive mem copies. This is particularly useful
for Full HD videos, where all the processing and display will be done
in the GPU.

No new tests, covered by existing tests.

* platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.cpp:
(WebCore::MediaPlayerPrivateGStreamerBase::MediaPlayerPrivateGStreamerBase):
(WebCore::MediaPlayerPrivateGStreamerBase::updateTexture):
* platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.h:
* platform/graphics/gstreamer/VideoSinkGStreamer.cpp:
(webkitVideoSinkProposeAllocation):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151673 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed GTK gardening.
zandobersek@gmail.com [Tue, 18 Jun 2013 07:20:45 +0000 (07:20 +0000)]
Unreviewed GTK gardening.

* platform/gtk/TestExpectations: Mark two tests that are flakily timing out as slow.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151672 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoA static_assert() in RetainPtr.h sometimes causes the build to fail.
aestes@apple.com [Tue, 18 Jun 2013 07:02:00 +0000 (07:02 +0000)]
A static_assert() in RetainPtr.h sometimes causes the build to fail.

Wrap the first argument to a static_assert() in parentheses since it
contains a comma, which sometimes tricks the preprocessor into thinking
the macro is being passed three arguments rather than two.

Reviewed by Brian Weinstein.

* wtf/RetainPtr.h:
(WTF::RetainPtr::RetainPtr):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151671 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION (r149184): Build errors in RefPtr.h when building with Clang, C++98 standard
zandobersek@gmail.com [Tue, 18 Jun 2013 07:00:19 +0000 (07:00 +0000)]
REGRESSION (r149184): Build errors in RefPtr.h when building with Clang, C++98 standard
https://bugs.webkit.org/show_bug.cgi?id=116352

Reviewed by Anders Carlsson.

Use the __has_feature macro instead of the __has_extension one when detecting what features the Clang compiler
that's being used is capable to provide. With the project moving onto using C++11 throughout its components,
there currently exist cases where C++11-specific functionality (for instance std::move) is used despite being
guarded by compiler-specific feature guards that still get enabled even when compiling in C++98 with Clang.
The problematic features are enabled because they're available as C++98 language extensions by Clang.

Using __has_feature instead enables the feature guard only if the tested feature is actually standardized in
the currently used language standard.

* wtf/Compiler.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151670 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoExtra whitespace in blank line in PlatformPopupMenuData.cpp
commit-queue@webkit.org [Tue, 18 Jun 2013 06:24:07 +0000 (06:24 +0000)]
Extra whitespace in blank line in PlatformPopupMenuData.cpp
https://bugs.webkit.org/show_bug.cgi?id=117627

Patch by Santosh Mahto <santosh.ma@samsung.com> on 2013-06-17
Reviewed by Brent Fulgham.

* Shared/PlatformPopupMenuData.cpp:
(WebKit::PlatformPopupMenuData::decode):
Removing extra space

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151669 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[curl] Set the http response status text
commit-queue@webkit.org [Tue, 18 Jun 2013 06:09:40 +0000 (06:09 +0000)]
[curl] Set the http response status text
https://bugs.webkit.org/show_bug.cgi?id=117307

Patch by Peter Gal <galpeter@inf.u-szeged.hu> on 2013-06-17
Reviewed by Brent Fulgham.

No new tests, covered by existing ones.

* platform/network/curl/ResourceHandleManager.cpp:
(WebCore::headerCallback):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151668 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Windows] Unreviewed AX gardening of windows-specific results.
bfulgham@apple.com [Tue, 18 Jun 2013 05:59:33 +0000 (05:59 +0000)]
[Windows] Unreviewed AX gardening of windows-specific results.

* platform/win/accessibility/aria-combobox-expected.txt: Added.
* platform/win/aria-labelledby-on-input-expected.txt: Added.
* platform/win/aria-menubar-menuitems-expected.txt: Added.
* platform/win/aria-roles-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151667 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[EFL] Remove unneeded WebCore namespace in FrameLoaderClientEfl
commit-queue@webkit.org [Tue, 18 Jun 2013 05:59:03 +0000 (05:59 +0000)]
[EFL] Remove unneeded WebCore namespace in FrameLoaderClientEfl
https://bugs.webkit.org/show_bug.cgi?id=117718

Patch by Seokju Kwon <seokju.kwon@gmail.com> on 2013-06-17
Reviewed by Christophe Dumez.

* WebCoreSupport/FrameLoaderClientEfl.cpp: Remove 'using namespace WebCore'.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151666 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoAX: Correct accessibility role when -webkit-box:display is used.
bfulgham@apple.com [Tue, 18 Jun 2013 01:30:44 +0000 (01:30 +0000)]
AX: Correct accessibility role when -webkit-box:display is used.
https://bugs.webkit.org/show_bug.cgi?id=117706

Reviewed by Chris Fleizach.

Source/WebCore:

accessibility/box-styled-lists.html

* accessibility/AccessibilityRenderObject.cpp:
(WebCore::AccessibilityRenderObject::determineAccessibilityRole): If HTML token is
of the list element type, treat it as a ListItemRole accessibility role, regardless
of what specific renderer is being used.

LayoutTests:

* accessibility/box-styled-lists.html: Added.
* platform/efl/TestExpectations:
* platform/gtk/TestExpectations:
* platform/mac/accessibility/box-styled-lists-expected.txt: Added.
* platform/win/TestExpectations:
* platform/win/accessibility/box-styled-lists-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151665 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoMac rebaseline.
rniwa@webkit.org [Tue, 18 Jun 2013 01:19:13 +0000 (01:19 +0000)]
Mac rebaseline.

* fast/repaint/table-cell-collapsed-border-scroll-expected.png:
* fast/repaint/table-cell-collapsed-border-scroll-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151664 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoInitialize AtomicStringTable in WTFThreadData's constructor
rniwa@webkit.org [Tue, 18 Jun 2013 01:08:33 +0000 (01:08 +0000)]
Initialize AtomicStringTable in WTFThreadData's constructor
https://bugs.webkit.org/show_bug.cgi?id=117671

Reviewed by Geoffrey Garen.

Extracted AtomicStringTable from AtomicString.cpp into AtomicStringTable.h/cpp
and made WTFThreadDada::WTFThreadData create the atomic string table for the thread.

This eliminates a branch from stringTable() in AtomicString.cpp.

* GNUmakefile.list.am:
* WTF.pro:
* WTF.vcxproj/WTF.vcxproj:
* WTF.vcxproj/WTF.vcxproj.filters:
* WTF.xcodeproj/project.pbxproj:
* wtf/CMakeLists.txt:
* wtf/WTFThreadData.cpp:
(WTF::WTFThreadData::WTFThreadData):
* wtf/text/AtomicString.cpp:
(WTF::stringTable):
* wtf/text/AtomicStringTable.cpp: Added.
(WTF::AtomicStringTable::create):
(WTF::AtomicStringTable::destroy):
* wtf/text/AtomicStringTable.h: Added.
(WTF::AtomicStringTable::table):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151663 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoAdd some UNUSED_PARAMs to RenderBlock.cpp so that it builds properly if CSS_EXCLUSION...
roger_fong@apple.com [Tue, 18 Jun 2013 00:54:50 +0000 (00:54 +0000)]
Add some UNUSED_PARAMs to RenderBlock.cpp so that it builds properly if CSS_EXCLUSIONS is disabled.
https://bugs.webkit.org/show_bug.cgi?id=117722.
Rubberstamped by Jon Lee.
* rendering/RenderBlock.cpp:
(WebCore::RenderBlock::updateRegionsAndExclusionsAfterChildLayout):
(WebCore::RenderBlock::logicalLeftOffsetForLine):
(WebCore::RenderBlock::logicalRightOffsetForLine):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151662 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoPageBanners appear over HTML5 video when media element is in full screen mode
bdakin@apple.com [Tue, 18 Jun 2013 00:41:43 +0000 (00:41 +0000)]
PageBanners appear over HTML5 video when media element is in full screen mode
https://bugs.webkit.org/show_bug.cgi?id=117721
-and corresponding-
<rdar://problem/13686998>

Reviewed by Sam Weinig.

Source/WebCore:

Update the scrolling tree when a header/footer has been removed.
* rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::updateLayerForHeader):
(WebCore::RenderLayerCompositor::updateLayerForFooter):

Source/WebKit2:

When an element enters fullscreen, hide the banners. Show them again when the
element exits fullscreen.
* WebProcess/FullScreen/WebFullScreenManager.cpp:
(WebKit::WebFullScreenManager::willEnterFullScreen):
(WebKit::WebFullScreenManager::willExitFullScreen):

Handle hiding and showing by removing or re-creating a parent layer for m_layer.
* WebProcess/WebPage/PageBanner.cpp:
(WebKit::PageBanner::hide):
(WebKit::PageBanner::showIfHidden):
* WebProcess/WebPage/PageBanner.h:
* WebProcess/WebPage/mac/PageBannerMac.mm:
(WebKit::PageBanner::PageBanner):
(WebKit::PageBanner::hide):
(WebKit::PageBanner::showIfHidden):
(WebKit::PageBanner::mouseEvent):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::hidePageBanners):
(WebKit::WebPage::showPageBanners):
* WebProcess/WebPage/WebPage.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151661 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoModify Windows makefiles to copy some bin output into Program Files.
roger_fong@apple.com [Tue, 18 Jun 2013 00:23:38 +0000 (00:23 +0000)]
Modify Windows makefiles to copy some bin output into Program Files.
https://bugs.webkit.org/show_bug.cgi?id=117714.
<rdar://problem/14179054>

Reviewed by Brent Fulgham.

* WTF.vcxproj/WTF.make:
* JavaScriptCore.vcxproj/JavaScriptCore.make:
* WebCore.vcxproj/WebCore.make:
* WebKit.vcxproj/WebKit.make:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151660 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoCrash in loadPendingShaders
mvujovic@adobe.com [Mon, 17 Jun 2013 22:42:36 +0000 (22:42 +0000)]
Crash in loadPendingShaders
https://bugs.webkit.org/show_bug.cgi?id=117665

Reviewed by Dean Jackson.

Speculative fix because the crash reports do not contain a test case.

Rearrange the conditions to check m_state.hasPendingShaders() first. Additionally, check if
m_state.style() is non-null.

No new tests. We don't know how to reproduce this crash yet.

* css/StyleResolver.cpp:
(WebCore::StyleResolver::loadPendingShaders):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151657 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION: Important controls are missing from <video> element UI, when compared...
commit-queue@webkit.org [Mon, 17 Jun 2013 21:50:55 +0000 (21:50 +0000)]
REGRESSION: Important controls are missing from <video> element UI, when compared to QuickTime
https://bugs.webkit.org/show_bug.cgi?id=20599

Add a "Download Video"/"Download Audio" context menu item to download media
elements.

Patch by Ruth Fong <ruth_fong@apple.com> on 2013-06-17
Reviewed by Beth Dakin.

Source/WebCore:

No new tests. media/context-menu-action.html,
which has been disabled by bug 116651, is used to test context menus.

* English.lproj/Localizable.strings: Updated to
include localizable strings for "Download Video" and "Download Audio".

* page/ContextMenuController.cpp: Updated to
include and handle a "Download Video/Audio" context menu item.

* platform/ContextMenuItem.h:

* platform/LocalizedStrings.cpp:
* platform/LocalizedStrings.h:
* platform/efl/LocalizedStringsEfl.cpp:
* platform/gtk/LocalizedStringsGtk.cpp:
* platform/qt/LocalizedStringsQt.cpp:
Updated to include localizable strings for "Download Video" and "Download Audio".

Source/WebKit/efl:

* WebCoreSupport/AssertMatchingEnums.cpp:
* ewk/ewk_contextmenu.h:
Updated to include new EWK enum type for "Download Video/Audio" context menu item.

Source/WebKit2:

* Shared/API/c/WKContextMenuItemTypes.h: Updated to include
an enum type for the "Download Video/Audio" context menu item.

* Shared/API/c/WKSharedAPICast.h: Updated to associate
the enums in WebKit and WebCore that handle the "Download Video/Audio"
context menu item.

* UIProcess/API/efl/ewk_context_menu.cpp:
* UIProcess/API/efl/ewk_context_menu_item.cpp:
* UIProcess/API/efl/ewk_context_menu_item.h:
Updated to include new EWK enum type for "Download Video/Audio" context menu item.

* UIProcess/WebPageProxy.cpp: Updated to
handle the "Download Video/Audio" context menu item mouse click
and download the media item in the UI process.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151656 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[CoordinatedGraphics][CSS Shaders] Use forwarding header for TextureMapperPlatformCom...
commit-queue@webkit.org [Mon, 17 Jun 2013 21:18:33 +0000 (21:18 +0000)]
[CoordinatedGraphics][CSS Shaders] Use forwarding header for TextureMapperPlatformCompiledProgram.h
https://bugs.webkit.org/show_bug.cgi?id=117712

Patch by Ralph Thomas <ralpht@gmail.com> on 2013-06-17
Reviewed by Noam Rosenthal.

* Shared/CoordinatedGraphics/WebCustomFilterProgramProxy.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151653 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[CSS Shapes] Consider bottom borders when calculating the position of the overflow
zoltan@webkit.org [Mon, 17 Jun 2013 20:23:58 +0000 (20:23 +0000)]
[CSS Shapes] Consider bottom borders when calculating the position of the overflow
https://bugs.webkit.org/show_bug.cgi?id=117663

Reviewed by Alexandru Chiculita.

When you have a shape and the content overflows from the shape we need to push the overflow below the content box,
not below the border box. We didn't consider the bottom borders of the content box, now I fixed it too and the overflow
just starts after the content box. I added a new helper function called pushShapeContentOverflowBelowTheContentBox. Now
both the shape in flow thread and the shape overflow use the same function. I added a new test and modified the existing
tests to cover all the affected cases.

Source/WebCore:

Test: fast/exclusions/shape-inside/shape-inside-overflow-fixed-dimensions-block-content.html
Existing overflow tests also modified to test the behavior:
      fast/exclusions/shape-inside/shape-inside-empty-expected.html
      fast/exclusions/shape-inside/shape-inside-overflow-fixed-dimensions-block-content-expected.html
      fast/exclusions/shape-inside/shape-inside-overflow-fixed-dimensions-block-content.html
      fast/exclusions/shape-inside/shape-inside-overflow-fixed-dimensions-expected.html
      fast/exclusions/shape-inside/shape-inside-overflow-fixed-dimensions.html
      fast/regions/shape-inside/shape-inside-on-regions-block-content-overflow-multiple-shapes-expected.html
      fast/regions/shape-inside/shape-inside-on-regions-block-content-overflow-multiple-shapes.html

* rendering/RenderBlockLineLayout.cpp:
(WebCore::pushShapeContentOverflowBelowTheContentBox):
(WebCore::RenderBlock::updateShapeAndSegmentsForCurrentLine):
(WebCore::RenderBlock::updateShapeAndSegmentsForCurrentLineInFlowThread):

LayoutTests:

* fast/exclusions/shape-inside/shape-inside-empty-expected.html:
* fast/exclusions/shape-inside/shape-inside-overflow-fixed-dimensions-block-content-expected.html: Copied from LayoutTests/fast/exclusions/shape-inside/shape-inside-overflow-fixed-dimensions-expected.html.
* fast/exclusions/shape-inside/shape-inside-overflow-fixed-dimensions-block-content.html: Copied from LayoutTests/fast/exclusions/shape-inside/shape-inside-overflow-fixed-dimensions.html.
* fast/exclusions/shape-inside/shape-inside-overflow-fixed-dimensions-expected.html:
* fast/exclusions/shape-inside/shape-inside-overflow-fixed-dimensions.html:
* fast/regions/shape-inside/shape-inside-on-regions-block-content-overflow-multiple-shapes-expected.html:
* fast/regions/shape-inside/shape-inside-on-regions-block-content-overflow-multiple-shapes.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151652 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[CSS Exclusions] Implement text alignment for shape-inside
betravis@adobe.com [Mon, 17 Jun 2013 18:53:56 +0000 (18:53 +0000)]
[CSS Exclusions] Implement text alignment for shape-inside
https://bugs.webkit.org/show_bug.cgi?id=101086

Reviewed by Alexandru Chiculita.

Adding tests to make sure that shape-inside layout correctly supports
the different values of text-align.

* fast/exclusions/resources/multi-segment-polygon.js:
(createOrInsert):
* fast/exclusions/shape-inside/shape-inside-text-align-expected.html: Added.
* fast/exclusions/shape-inside/shape-inside-text-align.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151650 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoCache FatFinger Text Result.
commit-queue@webkit.org [Mon, 17 Jun 2013 18:06:27 +0000 (18:06 +0000)]
Cache FatFinger Text Result.
https://bugs.webkit.org/show_bug.cgi?id=107403.

Patch by Tiancheng Jiang <tijiang@rim.com> on 2013-06-17
Reviewed by Rob Buis.

Internally Reviewed by Genevieve Mak.
Cache the FatFinger text result for later use in TouchEventHandler.

* Api/WebPage.cpp:
(BlackBerry::WebKit::WebPagePrivate::contextNode):
* WebKitSupport/SelectionHandler.cpp:
(BlackBerry::WebKit::SelectionHandler::selectAtPoint):
* WebKitSupport/TouchEventHandler.h:
(BlackBerry::WebKit::TouchEventHandler::cacheTextResult):
(TouchEventHandler):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151648 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[CSS Regions] ::before and ::after pseudo-elements are not displayed for regions
igor.o@sisa.samsung.com [Mon, 17 Jun 2013 18:05:19 +0000 (18:05 +0000)]
[CSS Regions] ::before and ::after pseudo-elements are not displayed for regions
https://bugs.webkit.org/show_bug.cgi?id=80163

Source/WebCore:

When the implementation of the before/after was moved to the DOM, before/after generated
content stoppped to work with css regions. The problem happens when:
1. RenderRegion can not have children and when the PseudoElement::attach tries to create a renderer for the
generated content, it fails.
2. RenderRegion::canHaveGeneratedChildren should not be false because regions can have generated content.

Reviewed by David Hyatt.

Tests: fast/regions/region-dynamic-after-before.html
       fast/regions/region-generated-content-before-after.html

* dom/NodeRenderingContext.cpp:
(WebCore::NodeRenderingContext::shouldCreateRenderer):
* rendering/RenderRegion.h:

LayoutTests:

Add two simple tests verifying if region works with before/after generated content.

Reviewed by David Hyatt.

* fast/regions/region-dynamic-after-before.html: Added.
* fast/regions/region-generated-content-before-after.html: Added.
* platform/mac/fast/regions/region-dynamic-after-before-expected.png: Added.
* platform/mac/fast/regions/region-dynamic-after-before-expected.txt: Added.
* platform/mac/fast/regions/region-generated-content-before-after-expected.png: Added.
* platform/mac/fast/regions/region-generated-content-before-after-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151647 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago<rdar://problem/13145014> Allow CoreIPC messages to transmit importance boosts
mrowe@apple.com [Mon, 17 Jun 2013 17:53:58 +0000 (17:53 +0000)]
<rdar://problem/13145014> Allow CoreIPC messages to transmit importance boosts

Transmitting the importance boost from the UI process to the web process is particularly
important when disabling process suppression on a web process, as it will minimize the
delays that can occur before the message instructing process suppression to be disabled
is processed by the web process.

We keep the importance boost alive for the duration of the CoreIPC message handler. We
achieve this by having an ImportanceAssertion object that owns the assertion, and whose
lifetime is tied to that of the MessageDecoder.

Reviewed by Anders Carlsson.

* Platform/CoreIPC/MessageDecoder.cpp:
(CoreIPC::MessageDecoder::setImportanceAssertion): Associate an importance assertion
with this message.
* Platform/CoreIPC/MessageDecoder.h:
* Platform/CoreIPC/mac/ConnectionMac.cpp:
(CoreIPC::Connection::open): Mark our receive port as allowing importance boosts to be received.
(CoreIPC::Connection::receiveSourceEventHandler):
* Platform/CoreIPC/mac/ImportanceAssertion.h:
(CoreIPC::ImportanceAssertion::create):
(CoreIPC::ImportanceAssertion::~ImportanceAssertion): Release the assertion.
(CoreIPC::ImportanceAssertion::ImportanceAssertion): Take an assertion if a boost was present in
the Mach message.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151646 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago<rdar://problem/14051357> Web processes sometimes remain in a suppressed
mrowe@apple.com [Mon, 17 Jun 2013 16:40:22 +0000 (16:40 +0000)]
<rdar://problem/14051357> Web processes sometimes remain in a suppressed
state after their window is unoccluded.

The code in WKView that responds to occlusion notifications was incorrectly
treating the window's occlusion state as being an enum rather than the set of
flags that it is.

Reviewed by Oliver Hunt.

* UIProcess/API/mac/WKView.mm:
(-[WKView _windowDidChangeOcclusionState:]): Test only the visibility state.
(-[WKView setWindowOcclusionDetectionEnabled:]): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151645 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Mac] Remove unmaintained GStreamer specific video code.
michael.bruning@digia.com [Mon, 17 Jun 2013 15:26:24 +0000 (15:26 +0000)]
[Mac] Remove unmaintained GStreamer specific video code.
https://bugs.webkit.org/show_bug.cgi?id=117694

Reviewed by Philippe Normand.

No new tests, no behaviour change.

* platform/mac/WebVideoFullscreenController.mm:
(-[WebVideoFullscreenController setupVideoOverlay:]):
(-[WebVideoFullscreenController windowDidLoad]):
(-[WebVideoFullscreenController setMediaElement:]):
(-[WebVideoFullscreenController windowDidExitFullscreen]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151643 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Qt] Remove Qt specific QTKIT flagged code.
michael.bruning@digia.com [Mon, 17 Jun 2013 14:35:05 +0000 (14:35 +0000)]
[Qt] Remove Qt specific QTKIT flagged code.
https://bugs.webkit.org/show_bug.cgi?id=117635

Reviewed by Simon Hausmann.

Due to disabling QTKIT for Qt in r151546, the
code the flags that are not taken into account
anymore and the code that has been rendered
unreachable by this are removed.

.:

* Source/widgetsapi.pri:

Source/WebCore:

No new tests, no behavioural change.

* Target.pri:
* WebCore.exp.in:
* page/Settings.cpp:
* page/Settings.h:
* platform/KURL.h:
* platform/SharedBuffer.h:
* platform/cf/KURLCFNet.cpp:
(WebCore::KURL::fileSystemPath):
* platform/cf/SharedBufferCF.cpp:
* platform/graphics/FloatSize.h:
* platform/graphics/IntRect.h:
* platform/graphics/IntSize.h:
* platform/graphics/MediaPlayer.cpp:
(WebCore::installedMediaEngines):
* platform/graphics/cg/FloatSizeCG.cpp:
* platform/graphics/cg/IntRectCG.cpp:
* platform/graphics/mac/MediaPlayerPrivateQTKit.h:
* platform/graphics/mac/MediaPlayerPrivateQTKit.mm:
(WebCore::MediaPlayerPrivateQTKit::createQTMovieLayer):
(WebCore::MediaPlayerPrivateQTKit::preferredRenderingMode):
(WebCore::MediaPlayerPrivateQTKit::paint):
(-[WebCoreMovieObserver layerHostChanged:]):
* platform/mac/SharedBufferMac.mm:
(+[WebCoreSharedBufferData initialize]):
(WebCore::SharedBuffer::createWithContentsOfFile):

Source/WebKit:

* WebKit1.pro:

Source/WebKit/mac:

* WebView/WebView.mm:
(-[WebView _preferencesChanged:]):

Source/WebKit/qt:

* WebCoreSupport/ChromeClientQt.cpp:
(WebCore::ChromeClientQt::ChromeClientQt):
(WebCore::ChromeClientQt::~ChromeClientQt):
* WebCoreSupport/ChromeClientQt.h:
* WebCoreSupport/FullScreenVideoQt.cpp:
(WebCore::FullScreenVideoQt::FullScreenVideoQt):
(WebCore::FullScreenVideoQt::~FullScreenVideoQt):
(WebCore::FullScreenVideoQt::enterFullScreenForNode):
(WebCore::FullScreenVideoQt::exitFullScreenForNode):
(WebCore::FullScreenVideoQt::isValid):
* WebCoreSupport/FullScreenVideoQt.h:
* WebCoreSupport/InitWebCoreQt.cpp:
(WebCore::initializeWebCoreQt):
* WebCoreSupport/QTKitFullScreenVideoHandler.h: Removed.
* WebCoreSupport/QTKitFullScreenVideoHandler.mm: Removed.
* WebCoreSupport/WebSystemInterface.h: Removed.
* WebCoreSupport/WebSystemInterface.mm: Removed.

Source/WebKit2:

* Target.pri:
* UIProcess/qt/WebContextQt.cpp:
(WebKit::WebContext::platformInitializeWebProcess):
* WebProcess/WebCoreSupport/qt/WebSystemInterface.h: Removed.
* WebProcess/WebCoreSupport/qt/WebSystemInterface.mm: Removed.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updatePreferences):
* WebProcess/qt/WebProcessMainQt.cpp:
(WebKit::WebProcessMainQt):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151642 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoRemove invalid checks from block-cursor-overtype-mode.html
sergio@webkit.org [Mon, 17 Jun 2013 14:32:02 +0000 (14:32 +0000)]
Remove invalid checks from block-cursor-overtype-mode.html
https://bugs.webkit.org/show_bug.cgi?id=117587

Reviewed by Darin Adler.

The test uses an array of ">" "<" symbols to perform comparisons
between consecutive positions of the block cursor. Since those
checks are done for each pair of consecutive characters the total
amount of checks is lengthOfTheWord - 1. We were adding an extra
symbol that was not used at all in the test.

* editing/selection/block-cursor-overtype-mode.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151641 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSSION(r151632) : Build error on ASSERT(WTF_USE_GRAMMAR_CHECKING)
allan.jensen@digia.com [Mon, 17 Jun 2013 13:57:45 +0000 (13:57 +0000)]
REGRESSSION(r151632) : Build error on ASSERT(WTF_USE_GRAMMAR_CHECKING)
https://bugs.webkit.org/show_bug.cgi?id=117692

Reviewed by Jocelyn Turcotte.

Ifdef the feature-depending code and fix the assertions to check runtime conditions.

* editing/Editor.cpp:
(WebCore::Editor::advanceToNextMisspelling):
(WebCore::Editor::markMisspellingsOrBadGrammar):
(WebCore::Editor::markBadGrammar):
* editing/TextCheckingHelper.cpp:
(WebCore::findBadGrammars):
(WebCore::TextCheckingHelper::findFirstGrammarDetail):
(WebCore::TextCheckingHelper::findFirstBadGrammar):
(WebCore::TextCheckingHelper::isUngrammatical):
(WebCore::TextCheckingHelper::markAllBadGrammar):
(WebCore::checkTextOfParagraph):
* editing/TextCheckingHelper.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151640 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed. Add my email to a DOM watchlist.
kangil.han@samsung.com [Mon, 17 Jun 2013 13:43:48 +0000 (13:43 +0000)]
Unreviewed. Add my email to a DOM watchlist.

* Scripts/webkitpy/common/config/watchlist:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151639 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed, rolling out r151632.
g.czajkowski@samsung.com [Mon, 17 Jun 2013 13:32:09 +0000 (13:32 +0000)]
Unreviewed, rolling out r151632.
http://trac.webkit.org/changeset/151632
https://bugs.webkit.org/show_bug.cgi?id=117585

Debug build error ASSERT(WTF_USE_GRAMMAR_CHECKING) for non MAC
platforms

Source/WebCore:

* page/ContextMenuController.cpp:
(WebCore::ContextMenuController::contextMenuItemSelected):
(WebCore::ContextMenuController::createAndAppendSpellingAndGrammarSubMenu):
(WebCore::ContextMenuController::populate):
(WebCore::ContextMenuController::checkOrEnableIfNeeded):
* platform/LocalizedStrings.cpp:
(WebCore::contextMenuItemTagSpellingMenu):
(WebCore::contextMenuItemTagShowSpellingPanel):
(WebCore::contextMenuItemTagCheckGrammarWithSpelling):
* platform/LocalizedStrings.h:
* platform/efl/LocalizedStringsEfl.cpp:
(WebCore::contextMenuItemTagSpellingMenu):
(WebCore::contextMenuItemTagShowSpellingPanel):
(WebCore::contextMenuItemTagCheckGrammarWithSpelling):
* platform/gtk/LocalizedStringsGtk.cpp:
(WebCore::contextMenuItemTagSpellingMenu):
(WebCore::contextMenuItemTagShowSpellingPanel):
(WebCore::contextMenuItemTagCheckGrammarWithSpelling):
* platform/qt/LocalizedStringsQt.cpp:
(WebCore::contextMenuItemTagSpellingMenu):
(WebCore::contextMenuItemTagShowSpellingPanel):
(WebCore::contextMenuItemTagCheckGrammarWithSpelling):

Source/WebKit/gtk:

* webkit/webkitglobals.cpp:
(webkit_context_menu_item_get_action):

Source/WebKit/mac:

* WebCoreSupport/WebContextMenuClient.mm:
(fixMenusReceivedFromOldClients):

Source/WebKit2:

* Shared/API/c/WKSharedAPICast.h:
(WebKit::toAPI):
(WebKit::toImpl):
* UIProcess/API/efl/ewk_context_menu_item.cpp:
(getEwkActionFromWKTag):
* UIProcess/API/efl/tests/test_ewk2_text_checker.cpp:
(onContextMenuShow):
(TEST_F):
* UIProcess/API/gtk/WebKitContextMenuActions.cpp:
(webkitContextMenuActionGetActionTag):
(webkitContextMenuActionGetForContextMenuItem):
(webkitContextMenuActionGetLabel):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::contextMenuItemSelected):

Source/WTF:

* wtf/Platform.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151638 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed. Fix make distcheck.
carlosgc@webkit.org [Mon, 17 Jun 2013 12:24:28 +0000 (12:24 +0000)]
Unreviewed. Fix make distcheck.

* GNUmakefile.list.am: Remove non existent files from compilation
and add a missing header file.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151637 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Qt] Unreviewed gardening. Correcting rebaselining after r151634.
zarvai@inf.u-szeged.hu [Mon, 17 Jun 2013 09:44:13 +0000 (09:44 +0000)]
[Qt] Unreviewed gardening. Correcting rebaselining after r151634.

* platform/qt-5.0-wk2/compositing/repaint/fixed-background-scroll-expected.txt: Renamed from LayoutTests/platform/qt-5.0-wk2/fast/repaint/fixed-background-scroll-expected.txt.
* platform/qt-5.0-wk2/compositing/repaint/positioned-movement-expected.txt: Renamed from LayoutTests/platform/qt-5.0-wk2/fast/repaint/positioned-movement-expected.txt.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151636 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed. Rename gobject_introspection_required variable.
carlosgc@webkit.org [Mon, 17 Jun 2013 08:39:58 +0000 (08:39 +0000)]
Unreviewed. Rename gobject_introspection_required variable.

As gobject_introspection_required_version for consistency with all
other required_version variables.

* Source/autotools/FindDependencies.m4:
* Source/autotools/Versions.m4:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151635 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Qt] Unreviewed gardening. Rebaselining repaint tests, see bug 115372.
zarvai@inf.u-szeged.hu [Mon, 17 Jun 2013 08:17:51 +0000 (08:17 +0000)]
[Qt] Unreviewed gardening. Rebaselining repaint tests, see bug 115372.

* platform/qt-5.0-wk2/fast/repaint/change-overflow-and-display-of-relative-expected.txt: Rebaselining after r151549.
* platform/qt-5.0-wk2/fast/repaint/fixed-background-scroll-expected.txt: Rebaselining after r151624.
* platform/qt-5.0-wk2/fast/repaint/positioned-movement-expected.txt: Rebaselining after r151622.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151634 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed GTK gardening.
zandobersek@gmail.com [Mon, 17 Jun 2013 07:43:58 +0000 (07:43 +0000)]
Unreviewed GTK gardening.

* Scripts/run-gtk-tests:
(TestRunner): Skip three WebKit2APITests suites that are currently timing out.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151633 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoContext menu grammar checking items are available when GRAMMAR_CHECKING macro is off
g.czajkowski@samsung.com [Mon, 17 Jun 2013 07:28:52 +0000 (07:28 +0000)]
Context menu grammar checking items are available when GRAMMAR_CHECKING macro is off
https://bugs.webkit.org/show_bug.cgi?id=117585

Reviewed by Anders Carlsson.

Source/WebCore:

Do not add "Check Grammar With Spelling" and "Ignore Grammar" to the context menu
when GRAMMAR_CHECKING is off.
Replace "Spelling and Grammar" with "Spelling" as the sub menu title and "Show/Hide Spelling
and Grammar" with "Show/Hide Spelling" when GRAMMAR_CHECKING is off.

Additionally, guard grammar checking in context menu code to not necessarily
compile it.

No new layout tests because every port has a different way of showing
spelling/grammar context menu items.

* page/ContextMenuController.cpp:
(WebCore::ContextMenuController::contextMenuItemSelected):
(WebCore::ContextMenuController::createAndAppendSpellingAndGrammarSubMenu):
(WebCore::ContextMenuController::populate):
(WebCore::ContextMenuController::checkOrEnableIfNeeded):
* platform/LocalizedStrings.cpp:
(WebCore::contextMenuItemTagCheckGrammarWithSpelling):
(WebCore::contextMenuItemTagSpellingMenu):
(WebCore::contextMenuItemTagShowSpellingPanel):
* platform/LocalizedStrings.h:
* platform/efl/LocalizedStringsEfl.cpp:
(WebCore::contextMenuItemTagCheckGrammarWithSpelling):
(WebCore::contextMenuItemTagSpellingMenu):
(WebCore::contextMenuItemTagShowSpellingPanel):
* platform/gtk/LocalizedStringsGtk.cpp:
(WebCore::contextMenuItemTagCheckGrammarWithSpelling):
(WebCore::contextMenuItemTagSpellingMenu):
(WebCore::contextMenuItemTagShowSpellingPanel):
* platform/qt/LocalizedStringsQt.cpp:
(WebCore::contextMenuItemTagCheckGrammarWithSpelling):
(WebCore::contextMenuItemTagSpellingMenu):
(WebCore::contextMenuItemTagShowSpellingPanel):

Source/WebKit/gtk:

Add GRAMMAR_CHECKING guard to the context menu grammar items.

* webkit/webkitglobals.cpp:
(webkit_context_menu_item_get_action):

Source/WebKit/mac:

Add GRAMMAR_CHECKING guard to the context menu grammar items.

* WebCoreSupport/WebContextMenuClient.mm:
(fixMenusReceivedFromOldClients):

Source/WebKit2:

Add GRAMMAR_CHECKING guard to the context menu grammar items.

* Shared/API/c/WKSharedAPICast.h:
(WebKit::toAPI):
(WebKit::toImpl):
* UIProcess/API/efl/ewk_context_menu_item.cpp:
(getEwkActionFromWKTag):
* UIProcess/API/efl/tests/test_ewk2_text_checker.cpp:
(checkSpellingItemsAvailability):
Change callback name, to be consistent with a newly added.

(checkSpellingAndGrammarSubmenuAvailability):
(checkGrammarWithSpellingItemAvailability):
Add two unit tests for EFL to check whether context menu grammar items are
available only when GRAMMAR_CHECKING macro is enabled.

* UIProcess/API/gtk/WebKitContextMenuActions.cpp:
(webkitContextMenuActionGetActionTag):
(webkitContextMenuActionGetForContextMenuItem):
(webkitContextMenuActionGetLabel):
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::contextMenuItemSelected):

Source/WTF:

Enable GRAMMAR_CHECKING for WebKit ports that use/implement it.

* wtf/Platform.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151632 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed GTK gardening.
zandobersek@gmail.com [Mon, 17 Jun 2013 07:23:47 +0000 (07:23 +0000)]
Unreviewed GTK gardening.

* platform/gtk-wk1/TestExpectations: Add a crashing expectation for a test that most likely regressed with r151021.
* platform/gtk/TestExpectations: Add failure expectations for a couple of tests failing due to stack size misbehavior.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151631 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[CSS Blending] Update the background-blend-mode-image-color-dynamic test to use the...
commit-queue@webkit.org [Mon, 17 Jun 2013 06:53:02 +0000 (06:53 +0000)]
[CSS Blending] Update the background-blend-mode-image-color-dynamic test to use the repaint test logic

Updating testcase verifying if background blend modes are updated dynamically from script. This now
properly uses the repaint test mechanism. Also, the background image resource is addded.

https://bugs.webkit.org/show_bug.cgi?id=117686

Patch by Mihai Tica <mitica@adobe.com> on 2013-06-16
Reviewed by Dean Jackson.

* fast/repaint/background-blend-mode-image-color-dynamic-expected.html:
* fast/repaint/background-blend-mode-image-color-dynamic.html:
* fast/repaint/resources/ducky.png: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151630 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoFix test assertion after r151624
simon.fraser@apple.com [Mon, 17 Jun 2013 04:23:15 +0000 (04:23 +0000)]
Fix test assertion after r151624

An assertion was hit in RenderObject::willBeDestroyed()
for fast/css/getComputedStyle/getComputedStyle-background-shorthand.html
because the code asserted that the RenderObject had been removed from
the FrameView's slowRepaintObject set before remove() was called,
so move the assertion to after that call.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::willBeDestroyed):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151629 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoRemove a redundant virtual call to hostWindow() in FrameView::invalidateRect()
gyuyoung.kim@samsung.com [Mon, 17 Jun 2013 03:56:41 +0000 (03:56 +0000)]
Remove a redundant virtual call to hostWindow() in FrameView::invalidateRect()
https://bugs.webkit.org/show_bug.cgi?id=117685

Reviewed by Andreas Kling.

From Blink r152490 by <vivek.vg@samsung.com>

* page/FrameView.cpp:
(WebCore::FrameView::invalidateRect): Avoid hostWindow() twice call.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151628 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoFix two assertion failures in Range::insertNode
tkent@chromium.org [Mon, 17 Jun 2013 02:25:37 +0000 (02:25 +0000)]
Fix two assertion failures in Range::insertNode
https://bugs.webkit.org/show_bug.cgi?id=116511

Reviewed by Ryosuke Niwa.

Source/WebCore:

> ASSERTION FAILED: childBefore == (offset ? container->childNode(offset - 1) : 0)
> third_party/WebKit/Source/WebCore/dom/RangeBoundaryPoint.h(115) : void WebCore::RangeBoundaryPoint::set(PassRefPtr<WebCore::Node>, int, WebCore::Node *)

> ASSERTION FAILED: child->parentNode()
> ../../third_party/WebKit/Source/core/dom/RangeBoundaryPoint.h(133) : void WebCore::RangeBoundaryPoint::setToBeforeChild(WebCore::Node *)
>  1   0x87594b2 WebCore::RangeBoundaryPoint::setToBeforeChild(WebCore::Node*)
>  2   0x87534a9 WebCore::Range::insertNode(WTF::PassRefPtr<WebCore::Node>, int&)

Range::insertNode calls Node::insertBefore, in which an event handler
can update the DOM structure so that RangeBoundaryPoint don't like. We
postpone event dispatching by EventQueueScope.

Also, remove old comments about Acid3. The behavior is standardized.
(Step 9 of http://dom.spec.whatwg.org/#dom-range-insertnode)

This patch imports http://src.chromium.org/viewvc/blink?view=revision&revision=150470 .

Test: fast/dom/Range/range-insertNode-assertion.html

* dom/Range.cpp:
(WebCore::Range::insertNode):
 - Add EventQueueScope
 - Remove obsolete comments

LayoutTests:

We need to update fast/dom/insertBefore-refChild-crash.html because the
test caused recursive calls to the event handler. container.innerHTML=''
did nothing before this CL. Now container has the newChild because
DOMNodeRemoved event dispatching is delayed until Range::insertNode
completion.

We need to update fast/text/split-text-crash.xhtml so that it doesn't
stop when the event handler is called twice. I'm not sure why the test
worked before this CL.

* fast/dom/Range/range-insertNode-assertion-expected.txt: Added.
* fast/dom/Range/range-insertNode-assertion.html: Added.
* fast/dom/insertBefore-refChild-crash.html:
* fast/text/split-text-crash.xhtml:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151627 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed. Added new baselines for some of editing/style tests, which works on EFL...
gyuyoung.kim@samsung.com [Mon, 17 Jun 2013 01:58:38 +0000 (01:58 +0000)]
Unreviewed. Added new baselines for some of editing/style tests, which works on EFL port.

* platform/efl/TestExpectations:
* platform/efl/editing/style/5017613-1-expected.png: Added.
* platform/efl/editing/style/5017613-1-expected.txt: Added.
* platform/efl/editing/style/5046875-1-expected.png: Added.
* platform/efl/editing/style/5046875-1-expected.txt: Added.
* platform/efl/editing/style/5065910-expected.png: Added.
* platform/efl/editing/style/5065910-expected.txt: Added.
* platform/efl/editing/style/5228141-expected.png: Added.
* platform/efl/editing/style/5228141-expected.txt: Added.
* platform/efl/editing/style/5279521-expected.png: Added.
* platform/efl/editing/style/5279521-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151626 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed. Added new baselines for some of editing/selection tests as r151533 on...
gyuyoung.kim@samsung.com [Mon, 17 Jun 2013 01:13:06 +0000 (01:13 +0000)]
Unreviewed. Added new baselines for some of editing/selection tests as r151533 on EFL port.

* platform/efl/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151625 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoFixed backgrounds in composited layers not repainted on scrolling
simon.fraser@apple.com [Sun, 16 Jun 2013 23:32:22 +0000 (23:32 +0000)]
Fixed backgrounds in composited layers not repainted on scrolling
https://bugs.webkit.org/show_bug.cgi?id=117684

Source/WebCore:

Reviewed by Tim Horton.

FrameView is aware that "slow-repaint objects" (i.e. renderers with
background-attachment:fixed) require a slow-scrolling path. However,
it was ignorant of the fact that such objects could be painting into
compositing layers; it simply dirtied the main tiles, and nothing else.

Fix by having FrameView track the slow-repaints objects explicitly, as we
do for position:fixed, and repaint each of them on scrolling.

Test: compositing/repaint/fixed-background-scroll.html

* page/FrameView.cpp:
(WebCore::FrameView::FrameView): No need to initialize m_slowRepaintObjectCount,
which is now an OwnPtr<RenderObjectSet>.
(WebCore::FrameView::useSlowRepaints): Use hasSlowRepaintObjects() now.
(WebCore::FrameView::addSlowRepaintObject): Now adds the object to a set,
allocating the set if necessary.
(WebCore::FrameView::removeSlowRepaintObject): Remove the object from the set,
and deallocate the set if empty.
(WebCore::FrameView::scrollContentsSlowPath): Call repaintSlowRepaintObjects();
this is the change that fixes the bug for always-composited implementations (e.g.
tile cache). This is a conservative change; we still invalidate the tile cache
as well. This could be optimized later.
(WebCore::FrameView::repaintSlowRepaintObjects): Repaint each object in the set.
Their appropriate compositing ancestor will be repainted.
(WebCore::FrameView::scrollPositionChangedViaPlatformWidget): Call
repaintSlowRepaintObjects() so that fixed backgrounds in composited layers are
correctly repainted; this fixes the bug for WebKit1.
* page/FrameView.h: Replace m_slowRepaintObjectCount with a HashSet of
RenderObjects.
(WebCore::FrameView::hasSlowRepaintObject): Takes a RenderObject* now.
(WebCore::FrameView::hasSlowRepaintObjects): Ditto.
* rendering/RenderObject.cpp:
(WebCore::RenderObject::styleWillChange): Pass the RenderObject.
(WebCore::RenderObject::willBeDestroyed): Assert that this RenderObject
has been removed from FrameView's set of slow-repaint objects already, if
the document is not being destroyed.
(WebCore::RenderObject::willBeRemovedFromTree): Pass the RenderObject.

LayoutTests:

Reviewed by Tim Horton.

Test that verifies that a composited layer with a fixed background is repainted
on scrolling.

* compositing/repaint/fixed-background-scroll-expected.txt: Added.
* compositing/repaint/fixed-background-scroll.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151624 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoSource/WebCore: Painting of fixed background images is wrong in composited layers
simon.fraser@apple.com [Sun, 16 Jun 2013 17:24:40 +0000 (17:24 +0000)]
Source/WebCore: Painting of fixed background images is wrong in composited layers
https://bugs.webkit.org/show_bug.cgi?id=65793

Reviewed by Sam Weinig.

The code that computed background image geometry for background-attachment:fixed
images was unaware of compositing, so often painting the image at the wrong location.

Fix by having RenderBoxModelObject::calculateBackgroundImageGeometry() do the correct
math for fixed backgrounds in composited layer by offsetting the viewport rect by
the paint container's absolute position.

Tests: compositing/backgrounds/fixed-background-on-descendant.html
       compositing/backgrounds/fixed-backgrounds.html

* rendering/RenderBox.cpp:
(WebCore::RenderBox::getBackgroundPaintedExtent): Now returns a bool indicating
whether it is returning a reliable extent rect. It can return false in the case where
a background is fixed, since computing the correct extent would require finding
the appropriate composited ancestor to pass to calculateBackgroundImageGeometry().
This is OK since this function is used for "background opaque" optimizations.
(WebCore::RenderBox::computeBackgroundIsKnownToBeObscured): If getBackgroundPaintedExtent()
returns false, return false.
(WebCore::RenderBox::maskClipRect): We removed mask-attachment, so we never need to
compute the composited ancestor here and can pass null.
(WebCore::RenderBox::repaintLayerRectsForImage): Unwrap a comment.
If the changed image is related to a fixed background, geometry.hasNonLocalGeometry()
will be true. In that cause, just repaint the entire renderer rather than groveling
around for a composited ancestor.
* rendering/RenderBox.h: Changed name and signature of backgroundPaintedExtent.
* rendering/RenderBoxModelObject.cpp:
(WebCore::RenderBoxModelObject::paintFillLayerExtended): calculateBackgroundImageGeometry()
now needs to know the painting container.
(WebCore::RenderBoxModelObject::calculateBackgroundImageGeometry): Now takes a painting
container, that is required to correctly compute the viewport-relative offset for fixed
backgrounds. geometry.setHasNonLocalGeometry() is set for fixed backgrounds to indicate
to callers that, if they didn't pass a paint container, the destRect is not accurate.
The main bug fix is also here: we move the viewportRect by the absolute location of
paint container, which is equivalent to the composited layer offset.
(WebCore::RenderBoxModelObject::getGeometryForBackgroundImage): calculateBackgroundImageGeometry()
takes a paint container.
* rendering/RenderBoxModelObject.h:
(WebCore::RenderBoxModelObject::BackgroundImageGeometry::BackgroundImageGeometry):
(WebCore::RenderBoxModelObject::BackgroundImageGeometry::setHasNonLocalGeometry):
(WebCore::RenderBoxModelObject::BackgroundImageGeometry::hasNonLocalGeometry):
* rendering/RenderImage.cpp:
(WebCore::RenderImage::computeBackgroundIsKnownToBeObscured): If getBackgroundPaintedExtent()
can't cheaply give an accurate answer, return false.
* rendering/RenderLayerBacking.cpp:
(WebCore::RenderLayerBacking::updateDirectlyCompositedBackgroundImage): Pass the paint container,
which is our own renderer.

LayoutTests: Fixed background images behave strangely with webkit transitions
https://bugs.webkit.org/show_bug.cgi?id=65793

Reviewed by Sam Weinig.

Ref tests that compare fixed background rendering after a scroll, with and
without compositing, with a couple of layer configurations.

* compositing/backgrounds/fixed-background-on-descendant-expected.html: Added.
* compositing/backgrounds/fixed-background-on-descendant.html: Added.
* compositing/backgrounds/fixed-backgrounds-expected.html: Added.
* compositing/backgrounds/fixed-backgrounds.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151623 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agowebkit-backface-visibility on a parent element stops background-position from updating
simon.fraser@apple.com [Sun, 16 Jun 2013 16:57:14 +0000 (16:57 +0000)]
webkit-backface-visibility on a parent element stops background-position from updating
https://bugs.webkit.org/show_bug.cgi?id=116319

Source/WebCore:

Reviewed by Darin Adler.

The optimization added in r102952 was incorrect in the case where a style change
resulted in a positioned-movement-only layout but also required a repaint; it assumed
that a composited layer did not need to be repainted for a a positioned-movement-only layout.

Fix by making RenderObject::setNeedsLayoutForPositionedMovement() check whether the
style change requires a repaint, and calling setLayerNeedsFullRepaint() in that situation.

Test: compositing/repaint/positioned-movement.html

* rendering/RenderLayer.h: RepaintStatus values do not need to be bit flags.
* rendering/RenderObject.cpp:
(WebCore::RenderObject::setStyle): Pass the old style to setNeedsPositionedMovementLayout().
(WebCore::RenderObject::styleDidChange): Ditto.
* rendering/RenderObject.h:
(WebCore::RenderObject::setNeedsPositionedMovementLayout): Now takes a const RenderStyle*.
* rendering/style/RenderStyle.cpp:
(WebCore::RenderStyle::diff): Remove a comment now that this bug is fixed.
(WebCore::RenderStyle::diffRequiresRepaint): Public wrapper for changeRequiresRepaint().
That function never uses contextSensitiveProperties so we can safely ignore them.
* rendering/style/RenderStyle.h: Expose a way to call changeRequiresRepaint().

LayoutTests:

Reviewed by Darin Adler.

Test that does a positioned-movement-only layout and dumps a layer tree with
repaint rects.

* compositing/repaint/positioned-movement-expected.txt: Added.
* compositing/repaint/positioned-movement.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151622 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago Try to fix iOS after last Pasteboard change.
darin@apple.com [Sun, 16 Jun 2013 15:06:30 +0000 (15:06 +0000)]
    Try to fix iOS after last Pasteboard change.

        * platform/ios/PasteboardIOS.mm:
        (WebCore::Pasteboard::writeSelection): Put the new code inline here because the
        stringSelectionForPasteboard function is currently Mac-only.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151621 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoMove Pasteboard::getStringSelection to Editor, fixing a layering violation
darin@apple.com [Sun, 16 Jun 2013 02:21:24 +0000 (02:21 +0000)]
Move Pasteboard::getStringSelection to Editor, fixing a layering violation
https://bugs.webkit.org/show_bug.cgi?id=117673

Reviewed by Ryosuke Niwa.

* editing/Editor.h: Added stringSelectionForPasteboardWithImageAltText.

* editing/mac/EditorMac.mm:
(WebCore::Editor::stringSelectionForPasteboard): Moved code here from the Pasteboard
class, since all the Pasteboard class did really was turn around and call back here.
(WebCore::Editor::stringSelectionForPasteboardWithImageAltText): Added. Variant of
the function above that includes image alt text. Separate named functions are often
the best pattern for something like this unless there are many different combinations.

* platform/Pasteboard.h: Remove the Mac-only Pasteboard::getStringSelectioon.

* platform/ios/PasteboardIOS.mm:
(WebCore::Pasteboard::writeSelection):
* platform/mac/PasteboardMac.mm:
(WebCore::Pasteboard::writeSelectionForTypes):
Call the two new functions instead of the old Pasteboard::getStringSelection.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151620 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoSupport using clang instead of gcc for bindings preprocessing
darin@apple.com [Sun, 16 Jun 2013 02:15:41 +0000 (02:15 +0000)]
Support using clang instead of gcc for bindings preprocessing
https://bugs.webkit.org/show_bug.cgi?id=117674

Reviewed by Sam Weinig.

I couldn't build on my computer because I don't have gcc installed.
This is my cut at fixing it.

* bindings/scripts/CodeGeneratorObjC.pm:
(ReadPublicInterfaces): Use clang if present.
* bindings/scripts/preprocessor.pm:
(applyPreprocessor): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151619 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoCrash in supportsSnapshotting
dino@apple.com [Sat, 15 Jun 2013 22:56:00 +0000 (22:56 +0000)]
Crash in supportsSnapshotting
https://bugs.webkit.org/show_bug.cgi?id=117670
<rdar://problem/14059711>

Reviewed by Darin Adler.

Speculative fix for deferencing a null RefPtr in supportsSnapshotting.
It seems like it was possible to get in a state where the plugin machinery
would fire up and attempt to get a screenshot before there was an active
connection to the plugin process (usually under periods of high load, such
as opening a lot of tabs at once). In this case we're now returning false
from supportsSnapshotting, which will trigger the plugin to restart, but
that's better than requesting a snapshot on something that doesn't yet exist.

* WebProcess/Plugins/PluginProxy.cpp:
(WebKit::PluginProxy::supportsSnapshotting): Guard for an empty connection object.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151617 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed. Reorder my email addresses.
commit-queue@webkit.org [Sat, 15 Jun 2013 15:44:18 +0000 (15:44 +0000)]
Unreviewed. Reorder my email addresses.
https://bugs.webkit.org/show_bug.cgi?id=100122

Patch by Bruno de Oliveira Abinader <brunoabinader@gmail.com> on 2013-06-15

* Scripts/webkitpy/common/config/contributors.json:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151614 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION (r148367): Facebook and Twitter icons at macworld.com are stacked vertical...
robert@webkit.org [Sat, 15 Jun 2013 06:59:28 +0000 (06:59 +0000)]
REGRESSION (r148367): Facebook and Twitter icons at macworld.com are stacked vertically, obscuring Twitter one
https://bugs.webkit.org/show_bug.cgi?id=117284

Reviewed by David Hyatt.

Source/WebCore:

Test: fast/text/whitespace/inline-whitespace-wrapping-7.html

We were looking for line breaks in trailing collapsed whitespace even when inside a nowrap inline.
It's wrong to do this as we don't know yet if we will want or need to break - that decision has to
be deferred until we enter a part of the line that is autowrap.

* rendering/RenderBlockLineLayout.cpp:
(WebCore::RenderBlock::LineBreaker::nextSegmentBreak):

LayoutTests:

* fast/text/whitespace/inline-whitespace-wrapping-7-expected.html: Added.
* fast/text/whitespace/inline-whitespace-wrapping-7.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151613 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agobackground-color is not preserved when inserting a new paragraph
commit-queue@webkit.org [Sat, 15 Jun 2013 04:42:23 +0000 (04:42 +0000)]
background-color is not preserved when inserting a new paragraph
https://bugs.webkit.org/show_bug.cgi?id=117138

Patch by Lukasz Gajowy <l.gajowy@samsung.com> on 2013-06-14
Reviewed by Ryosuke Niwa.

Source/WebCore:

The style isn't copied because background-color isn't defined as inheritable property.

Test: editing/inserting/insert-paragraph-not-preserving-background-color.html

* editing/InsertParagraphSeparatorCommand.cpp:
(WebCore::InsertParagraphSeparatorCommand::calculateStyleBeforeInsertion):
Passing EditingPropertiesInEffect instead of implicit parameter OnlyEditingInheritableProperties.

LayoutTests:

Adding layout test and expected result to reproduce the issue.

* editing/inserting/insert-paragraph-with-font-and-background-color-expected.txt: Added.
* editing/inserting/insert-paragraph-with-font-and-background-color.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151612 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoThreaded HTML parser can ASSERT in some situations
abarth@webkit.org [Fri, 14 Jun 2013 23:05:39 +0000 (23:05 +0000)]
Threaded HTML parser can ASSERT in some situations
https://bugs.webkit.org/show_bug.cgi?id=117662

Reviewed by Ryosuke Niwa.

Source/WebCore:

This patch merges https://chromium.googlesource.com/chromium/blink/+/3c0a112b42d72c7623f78463166dd7f04d680b4c

--->8---
The parser is supposed to stop processing data when there's a pending
location change, but before this CL we would actually process the first
token from every chunk when there was a pending location change.

If the first token in a chunk was a "script" start tag, that would put
the tree builder into TextMode, which can only process character
tokens. If the next chunk starts with another start tag token, the tree
builder would be sad and hit an ASSERT.

This CL reorders a couple lines in HTMLDocumentParser so that we check
for pending location changes before processing any tokens. This change
stops us from processing the first token from each chunk while there is
a pending location change.

This issue can reproduce on any platform, but it reproduced often on
Android because mobile markup often elides spaces between consecutive
script tags. If there are space characters between the script tags,
then those space characters will be the first ones processes in the
chunk, which doesn't trigger the ASSERT.
---8<---

Since we landed this patch in Blink, we've been able to reproduce this
issue on desktops as well. I spoke with rniwa in #webkit and he said
that WebKit doesn't have any current plans to enable the threaded
parser but that it was probably worth merging this patch anyway.

Test: http/tests/navigation/pending-location-change-assert.html

* html/parser/HTMLDocumentParser.cpp:
(WebCore::HTMLDocumentParser::processParsedChunkFromBackgroundParser):

LayoutTests:

* http/tests/navigation/pending-location-change-assert-expected.txt: Added.
* http/tests/navigation/pending-location-change-assert.html: Added.
* http/tests/resources/slow-notify-done.php: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151611 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoIntroduce USE(WINGDI) for the Windows port
paroga@webkit.org [Fri, 14 Jun 2013 22:49:14 +0000 (22:49 +0000)]
Introduce USE(WINGDI) for the Windows port
https://bugs.webkit.org/show_bug.cgi?id=116138

Reviewed by Ryosuke Niwa.

Using USE(WINGDI) instead of OS(WINCE) will allow us to
compile the GDI based Windows port on WinNT too.

.:

* Source/cmake/OptionsWinCE.cmake:

Source/WebCore:

* config.h:
* platform/graphics/BitmapImage.h:
* platform/graphics/FontPlatformData.h:
* platform/graphics/GlyphBuffer.h:
(WebCore):
(GlyphBuffer):
* platform/graphics/Gradient.h:
(Gradient):
* platform/graphics/GraphicsContext.cpp:
(WebCore):
* platform/graphics/GraphicsContext.h:
(WebCore):
(GraphicsContext):
* platform/graphics/ImageBufferData.h:
* platform/graphics/NativeImagePtr.h:
(WebCore):
* platform/graphics/Path.h:
* platform/graphics/Pattern.h:
* plugins/win/PluginViewWin.cpp:
(WebCore::PluginView::paintWindowedPluginIntoContext):
(WebCore::PluginView::paint):
(WebCore::PluginView::snapshot):

Source/WTF:

* wtf/Platform.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151610 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoReport the memory cost of HTMLMediaElements to GC.
jer.noble@apple.com [Fri, 14 Jun 2013 22:40:10 +0000 (22:40 +0000)]
Report the memory cost of HTMLMediaElements to GC.
https://bugs.webkit.org/show_bug.cgi?id=117608

Reviewed by Darin Adler.

Report the extra memory cost of a HTMLMediaElement to the VM, in order to
encourage GC to occur after a media element is removed from the DOM.

Because we cannot know for sure the memory cost of our underlying media
framework objects, use the worst case scenario for the media memory cost:
the buffered percentage of movie * the total data length of the movie.

* html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::HTMLMediaElement): Initialize m_reportedExtraMemoryCost.
(WebCore::HTMLMediaElement::removedFrom): Report the extra memory cost.
* html/HTMLMediaElement.h:
* html/TimeRanges.cpp:
(TimeRanges::totalDuration): Added convenience function.
* html/TimeRanges.h:
* platform/graphics/MediaPlayer.cpp:
(WebCore::MediaPlayer::extraMemoryCost): Pass to MediaPlayerPrivate.
* platform/graphics/MediaPlayer.h:
* platform/graphics/MediaPlayerPrivate.h:
(WebCore::MediaPlayerPrivateInterface::extraMemoryCost): Default to 0.
* platform/graphics/avfoundation/MediaPlayerPrivateAVFoundation.cpp:
(WebCore::MediaPlayerPrivateAVFoundation::extraMemoryCost): Return the percentage
    loaded * total data length.
* platform/graphics/avfoundation/MediaPlayerPrivateAVFoundation.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151609 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoAdd support for AVFoundation-backed HTMLMediaElements in the WebVideoFullScreenContro...
jer.noble@apple.com [Fri, 14 Jun 2013 22:27:35 +0000 (22:27 +0000)]
Add support for AVFoundation-backed HTMLMediaElements in the WebVideoFullScreenController.
https://bugs.webkit.org/show_bug.cgi?id=117597

Reviewed by Eric Carlson.

Source/WebCore:

Check for media elements with AVFoundation media types and create a AVPlayerLayer
rather than a QTMovieLayer.

* platform/mac/WebVideoFullscreenController.h:
* platform/mac/WebVideoFullscreenController.mm:
(-[WebVideoFullscreenController setupVideoOverlay:]): Take a CALayer rather than a
    QTMovieLayer. Move the code which connects the layer with it's media source into
    setMediaElement.
(-[WebVideoFullscreenController windowDidLoad]): Give the window's contentView a generic
    CALayer instaead of a QTMovieLayer.
(-[WebVideoFullscreenController setMediaElement:]): Test whether the incoming media
    element is backed by a QTMovie or an AVPlayer, and create the appropriate layer.
(-[WebVideoFullscreenController windowDidExitFullscreen]): Remove the observer of
    AVPlayer's rate property.
(-[WebVideoFullscreenController observeValueForKeyPath:ofObject:change:context:]):
    If the keyPath is "rate", call rateChanged.

Source/WebKit/mac:

Allow AVFoundation to be enabled even when the FullScreen API is disabled.

* WebView/WebView.mm:
(-[WebView _preferencesChanged:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151608 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoWKPageFindStringMatches ignores the kWKFindOptionsBackwards option.
enrica@apple.com [Fri, 14 Jun 2013 22:21:13 +0000 (22:21 +0000)]
WKPageFindStringMatches ignores the kWKFindOptionsBackwards option.
https://bugs.webkit.org/show_bug.cgi?id=117647
<rdar://problem/13881024>

Source/WebCore:

Reviewed by Darin Adler.

The API returns the matched ranges in the DOM order regardless of the
find direction, but the index of the first match after the user selection
should take the find direction into account.

Extended existing test in TestWebKitAPI.

* page/Page.cpp:
(WebCore::Page::findStringMatchingRanges): Added handling of the Backwards case.
* page/Page.h: Fixed incorrect name of the enum.

Source/WebKit2:

Reviewed by Darin Adler.

The API returns the matched ranges in the DOM order regardless of the
find direction, but the index of the first match after the user selection
should take the find direction into account.

Extended existing test in TestWebKitAPI.

* Shared/API/c/WKFindOptions.h: Added enum for the case where there are
no matches after the user selection in the direction of the find.

Tools:

Reviewed by Darin Adler.

The test now uses content with a selection and tests both
forwards and backward find as well as the case of a find
that has no matches after the user selection.

* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/Tests/WebKit2/FindMatches.mm:
(TestWebKitAPI::didFindStringMatches):
* TestWebKitAPI/Tests/WebKit2/findRanges.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151607 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION (r150651): Web Audio doesn't work at all (silence or squawks) on Mountain...
jer.noble@apple.com [Fri, 14 Jun 2013 21:55:48 +0000 (21:55 +0000)]
REGRESSION (r150651): Web Audio doesn't work at all (silence or squawks) on Mountain Lion
https://bugs.webkit.org/show_bug.cgi?id=117652

Reviewed by Eric Carlson.

r150651 disabled setting the buffer size to a large value, but in so doing, also
disabled setting the buffer size to a small value for Web Audio. Narrow the Mountain
Lion protection to just the <video> element case.

* platform/audio/mac/AudioSessionManagerMac.cpp:
(AudioSessionManager::updateSessionState):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151606 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoFunction names on Object.prototype should be common identifiers
rniwa@webkit.org [Fri, 14 Jun 2013 20:36:33 +0000 (20:36 +0000)]
Function names on Object.prototype should be common identifiers
https://bugs.webkit.org/show_bug.cgi?id=117614

Reviewed by Darin Adler.

Source/JavaScriptCore:

Patch written by Sam Weinig. Make Object's prototype function names common identififers since they're used frequently.

* runtime/CommonIdentifiers.h:
* runtime/FunctionConstructor.cpp:
(JSC::constructFunction):
* runtime/JSGlobalObject.cpp:
(JSC::JSGlobalObject::reset):
* runtime/JSObject.h:
* runtime/ObjectPrototype.cpp:
(JSC::ObjectPrototype::finishCreation):
* runtime/StringPrototype.cpp:
(JSC::StringPrototype::finishCreation):

Source/WebCore:

Use the added common identifiers.

* bindings/js/JSDOMWindowBase.cpp:
(WebCore::JSDOMWindowBase::finishCreation):
(WebCore::JSDOMWindowBase::updateDocument):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151605 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoEditing: wrong text position when you click enter on the text behind the image
a.bah@samsung.com [Fri, 14 Jun 2013 20:13:57 +0000 (20:13 +0000)]
Editing: wrong text position when you click enter on the text behind the image
https://bugs.webkit.org/show_bug.cgi?id=115023

Reviewed by Ryosuke Niwa.

Source/WebCore:

When trying to break a line after an image followed by some text, contrary
to the expected behavior the text doesn't move to a new line, instead, the
caret position just shifts to the end of line.

As per the existing implementation for inserting a paragraph separator,
the insertion position obtained corresponds to the point after the image
element.
Since the insertion position doesn't resolve to being offset in the
following text node (0 offset), no splitting of text occurs and the block
to be inserted is placed after the containing (start) block (thereby
causing the caret position to shift).

If the computed insertionPosition points to an element that shall be
ignored for editing purposes (i.e. cannot have a VisiblePosition inside
it), and the position lies either at the start or at the end of such
an element, we should move our Position either upstream or
downstream (respectively) so as to obtain a valid position which can
be used further for splitting of the text.

Test: editing/inserting/insert-paragraph-after-non-editable-node-before-text.html

* editing/InsertParagraphSeparatorCommand.cpp:
(WebCore::InsertParagraphSeparatorCommand::doApply):
Moving the insertionPosition either upstream or downstream, if the point
lies either at the start or at the end of the anchor node.
The ensuing position, if a text node, can then be used for splitting of
the text correctly.

LayoutTests:

* editing/inserting/insert-paragraph-after-non-editable-node-before-text-expected.txt: Added.
* editing/inserting/insert-paragraph-after-non-ediatble-node-before-text.html: Added.
Layout test added for verifying the insert paragraph behavior between an inline node
which is ignored for editing purposes and some text.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151604 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[BlackBerry] Remove implementation of ContextMenu classes
commit-queue@webkit.org [Fri, 14 Jun 2013 20:02:46 +0000 (20:02 +0000)]
[BlackBerry] Remove implementation of ContextMenu classes
https://bugs.webkit.org/show_bug.cgi?id=114860

Patch by Alberto Garcia <agarcia@igalia.com> on 2013-06-14
Reviewed by Rob Buis.

This code is not being used so we can safely remove it.

Source/WebCore:

* PlatformBlackBerry.cmake:
* platform/blackberry/ContextMenuBlackBerry.cpp: Removed.
* platform/blackberry/ContextMenuItemBlackBerry.cpp: Removed.
* platform/blackberry/LocalizedStringsBlackBerry.cpp:

Source/WebKit:

* PlatformBlackBerry.cmake:

Source/WebKit/blackberry:

* Api/WebPage.cpp:
(BlackBerry::WebKit::WebPagePrivate::init):
* WebCoreSupport/ContextMenuClientBlackBerry.cpp: Removed.
* WebCoreSupport/ContextMenuClientBlackBerry.h: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151603 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoShrink larger images to fit the width of the blog.
timothy@apple.com [Fri, 14 Jun 2013 19:37:09 +0000 (19:37 +0000)]
Shrink larger images to fit the width of the blog.

* blog-files/new-inspector/hero-labeled.png:
* blog-files/new-inspector/hero.png:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151602 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Windows] Provide simple <meter> drawing logic for testing.
bfulgham@apple.com [Fri, 14 Jun 2013 19:34:07 +0000 (19:34 +0000)]
[Windows] Provide simple <meter> drawing logic for testing.
https://bugs.webkit.org/show_bug.cgi?id=117645.

Reviewed by Anders Carlsson.

accessbility/meter-element.html

* rendering/RenderThemeSafari.cpp: Simple test implementation
(WebCore::RenderThemeSafari::adjustMeterStyle): Added.
(WebCore::RenderThemeSafari::supportsMeter): Added.
(WebCore::RenderThemeSafari::meterSizeForBounds): Added.
(WebCore::RenderThemeSafari::paintMeter): Added.
* rendering/RenderThemeSafari.h:
* rendering/RenderThemeWin.cpp:
(WebCore::RenderThemeWin::progressBarTheme): Added.
(WebCore::RenderThemeWin::getClassicThemeData): Support meters.
(WebCore::RenderThemeWin::getThemeData): Support meter theme.
(WebCore::RenderThemeWin::adjustMeterStyle): Added.
(WebCore::RenderThemeWin::supportsMeter): Added.
(WebCore::RenderThemeWin::meterSizeForBounds): Added.
(WebCore::RenderThemeWin::paintMeter): Added.
* rendering/RenderThemeWin.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151601 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoPotential use-after-free with an event fired at a HTMLMediaElement which is currently...
eric.carlson@apple.com [Fri, 14 Jun 2013 17:18:59 +0000 (17:18 +0000)]
Potential use-after-free with an event fired at a HTMLMediaElement which is currently being deleted
https://bugs.webkit.org/show_bug.cgi?id=117466

Reviewed by Oliver Hunt.

Merge https://chromium.googlesource.com/chromium/blink/+/f4200a0093b3d9376f703961615359ec7fb712b4

If an event is created using as target an HTMLMediaElement which is
currently being deleted it becomes a heap-use-after free situation.

The GenericEventQueue instance is already owned by the HTMLMediaElement,
and there already is an underlying mechanism to set the target of the
event to NULL, if their target is owner of the queue.

In order to avoid creating this reference in the first place, we enqueue
the event with a NULL target to defer the refcount increment until the
timer for dispatching the event happens (which won't happen at all if
garbage collection is already destroying the objects).

Source/WebCore:

Test: media/track/media-element-enqueue-event-crash.html

* dom/GenericEventQueue.cpp:
(WebCore::GenericEventQueue::enqueueEvent): Don't ASSERT if the event has no target.

* html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::~HTMLMediaElement): Close the event queue so it won't try to
    dispatch any pending events.
(WebCore::HTMLMediaElement::scheduleEvent): Don't set the event target, it will happen just
    prior to event dispatch.
(WebCore::HTMLMediaElement::stop): Close the event queue.

LayoutTests:

* media/track/media-element-enqueue-event-crash-expected.txt: Added.
* media/track/media-element-enqueue-event-crash.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151600 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoClicking on snapshotting plug-ins does not restart them
dino@apple.com [Fri, 14 Jun 2013 17:09:49 +0000 (17:09 +0000)]
Clicking on snapshotting plug-ins does not restart them
https://bugs.webkit.org/show_bug.cgi?id=117620
<rdar://problem/13821729>

Follow-up review comments from Darin Adler, with some
discussion on IRC.

Sprinkle OVERRIDE on virtual functions where needed. Also
make sure that the Event is a MouseEvent before casting to it.

* html/HTMLPlugInElement.h: Move defaultEventHandler back to protected and add OVERRIDE.
* html/HTMLPlugInImageElement.cpp:
(WebCore::HTMLPlugInImageElement::defaultEventHandler): Check for MouseEvent and
use toMouseEvent().
* html/HTMLPlugInImageElement.h: Sprinkle some OVERRIDEs, and move defaultEventHandler
to private.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151599 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoRemove front-end from the WebCore Xcode project.
timothy@apple.com [Fri, 14 Jun 2013 16:52:40 +0000 (16:52 +0000)]
Remove front-end from the WebCore Xcode project.
https://bugs.webkit.org/show_bug.cgi?id=117640

Reviewed by Darin Adler.

* WebCore.xcodeproj/project.pbxproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151598 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoCSSParser::parseImageSet() doesn't need a parameter.
commit-queue@webkit.org [Fri, 14 Jun 2013 16:36:45 +0000 (16:36 +0000)]
CSSParser::parseImageSet() doesn't need a parameter.
https://bugs.webkit.org/show_bug.cgi?id=117617

Patch by Jaehun Lim <ljaehun.lim@samsung.com> on 2013-06-14
Reviewed by Darin Adler.

parseImageSet() can get m_valueList directly.
And minor fixes.

No new tests, no behavior change.

* css/CSSParser.cpp:
(WebCore::CSSParser::parseValue):
(WebCore::CSSParser::parseContent):
(WebCore::CSSParser::parseFillImage):
(WebCore::CSSParser::parseBorderImage):
(WebCore::CSSParser::parseImageSet):
* css/CSSParser.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151597 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[BlackBerry] Dropdown button UX updates.
commit-queue@webkit.org [Fri, 14 Jun 2013 16:25:51 +0000 (16:25 +0000)]
[BlackBerry] Dropdown button UX updates.
https://bugs.webkit.org/show_bug.cgi?id=117642.

Patch by Tiancheng Jiang <tijiang@rim.com> on 2013-06-14
Reviewed by Rob Buis.

JIRA 388652
Internally Reviewed by Jeff Rogers.

Change arrow assert.

* platform/blackberry/RenderThemeBlackBerry.cpp:
(WebCore::RenderThemeBlackBerry::paintMenuList):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151596 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[BlackBerry] LayerTiler fails to tile really big layers
anilsson@rim.com [Fri, 14 Jun 2013 15:00:38 +0000 (15:00 +0000)]
[BlackBerry] LayerTiler fails to tile really big layers
https://bugs.webkit.org/show_bug.cgi?id=117211

Reviewed by Carlos Garcia Campos.

PR 273550

Source/WebCore:

The cause for this bug was that LayerTiler computed which tiles are
visible by exhaustively mapping every tile to normalized device
coordinates and checking if it intersected the viewport. If there was a
lot of tiles, it would get stuck in this loop, iterating over all
tiles, for a very long time. Also, the visibility information was
transferred to the WebKit thread using mutexes, which added unnecessary
mutex contention to the mix.

Fixed by doing the reverse calculation, unprojecting the visible part
of the layer to layer coordinate space to find which tiles are visible.
Instead of doing a traditional unprojection of point to line and see
where the line intersects the layer, this patch uses the w-coordinates
of triangle vertices to perform "perspective correct texturing" of the
intersection points in device space, which is equivalent to
unprojection since perspective correct texturing computes a 2D
coordinate in (normalized) layer coordinate space which is easily
scaled up to the layer bounds and get the visible region expressed in
layer coordinate space.

The visible area is approximated by a rectangle, and a set of tiles
needing render are added to the mix, and  we have a swappable data
structure so visbility can be transferred to WebKit thread without
mutexes, but instead using atomic swap.

No new tests, covered by existing tests.

* platform/graphics/blackberry/CanvasLayerWebKitThread.cpp:
(CanvasLayerCompositingThreadClient):
(WebCore::CanvasLayerCompositingThreadClient::drawTextures):
* platform/graphics/blackberry/EGLImageLayerCompositingThreadClient.cpp:
(WebCore::EGLImageLayerCompositingThreadClient::drawTextures):
* platform/graphics/blackberry/EGLImageLayerCompositingThreadClient.h:
(EGLImageLayerCompositingThreadClient):
* platform/graphics/blackberry/LayerCompositingThread.cpp:
(WebCore::LayerCompositingThread::drawTextures):
(WebCore::LayerCompositingThread::drawSurface):
* platform/graphics/blackberry/LayerCompositingThread.h:
(LayerCompositingThread):
* platform/graphics/blackberry/LayerCompositingThreadClient.h:
(WebCore):
(LayerCompositingThreadClient):
* platform/graphics/blackberry/LayerRenderer.cpp:
(WebCore::LayerRenderer::compositeLayersRecursive):
* platform/graphics/blackberry/LayerTile.cpp:
(WebCore::LayerTile::LayerTile):
(WebCore::LayerTile::setContents):
(WebCore::LayerTile::updateContents):
(WebCore::LayerTile::discardContents):
* platform/graphics/blackberry/LayerTile.h:
(LayerTile):
(WebCore::LayerTile::contentsScale):
(WebCore::LayerTile::setContentsDirty):
* platform/graphics/blackberry/LayerTiler.cpp:
(LayerVisibility):
(WebCore::LayerVisibility::LayerVisibility):
(WebCore::LayerVisibility::visibleRect):
(WebCore::LayerVisibility::setVisibleRect):
(WebCore::LayerVisibility::needsRender):
(WebCore::LayerVisibility::tileNeedsRender):
(WebCore::LayerVisibility::willRenderTile):
(WebCore::LayerVisibility::swapTilesNeedingRender):
(WebCore::LayerVisibility::merge):
(WebCore):
(WebCore::LayerTiler::LayerTiler):
(WebCore::LayerTiler::~LayerTiler):
(WebCore::LayerTiler::updateTextureContentsIfNeeded):
(WebCore::LayerTiler::swapFrontVisibility):
(WebCore::LayerTiler::setFrontVisibility):
(WebCore::LayerTiler::layerVisibilityChanged):
(WebCore::LayerTiler::uploadTexturesIfNeeded):
(WebCore::LayerTiler::processTextureJob):
(WebCore::LayerTiler::addTileJob):
(WebCore::LayerTiler::performTileJob):
(WebCore::LayerTiler::drawTile):
(WebCore::inflateViewport):
(WebCore::LayerTiler::drawTextures):
(WebCore::LayerTiler::pruneTextures):
(WebCore::LayerTiler::rectForTile):
* platform/graphics/blackberry/LayerTiler.h:
(WebCore):
(LayerTiler):
(WebCore::LayerTiler::TextureJob::TextureJob):
(WebCore::LayerTiler::takeFrontVisibility):
* platform/graphics/blackberry/LayerUtilities.h:
(WebCore):
(WebCore::det):
(WebCore::dot):
(LayerClipEdge):
(WebCore::LayerClipEdge::LayerClipEdge):
(WebCore::LayerClipEdge::isPointInside):
(WebCore::LayerClipEdge::computeIntersection):
(WebCore::intersectPolygonWithRect):
(WebCore::computeBarycentricCoordinates):
(WebCore::manhattanDistanceToViewport):
(UnprojectionVertex):
(WebCore::compareManhattanDistanceToViewport):
(WebCore::unproject):

Source/WebKit/blackberry:

Adapt to changes in LayerCompositingThreadClient interface.

* Api/WebOverlay.cpp:
(BlackBerry::WebKit::WebOverlayLayerCompositingThreadClient::drawTextures):
* Api/WebOverlay_p.h:
(WebOverlayLayerCompositingThreadClient):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151595 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[GTK] Compilation problems - hb.h not found
commit-queue@webkit.org [Fri, 14 Jun 2013 14:53:47 +0000 (14:53 +0000)]
[GTK] Compilation problems - hb.h not found
https://bugs.webkit.org/show_bug.cgi?id=117636

Unreviewed GTK+ build fix.

Patch by Andres Gomez <agomez@igalia.com> on 2013-06-14

* GNUmakefile.am: Added freetype cflags for WebKit2 GTK+ Platform
library.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151594 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Qt] Unreviewed gardening. Generate new baseline to some tests and bot greening.
zarvai@inf.u-szeged.hu [Fri, 14 Jun 2013 14:50:06 +0000 (14:50 +0000)]
[Qt] Unreviewed gardening. Generate new baseline to some tests and bot greening.

Patch by Gabor Abraham <abrhm@inf.u-szeged.hu> on 2013-06-14

* platform/qt-5.0-wk2/TestExpectations:
* platform/qt-5.0-wk2/animations/3d/change-transform-in-end-event-expected.png: Added.
* platform/qt-5.0-wk2/fast/borders/border-left-right-same-bottom-different-color-expected.png: Added.
* platform/qt-5.0-wk2/fast/borders/border-radius-percent-expected.png: Added.
* platform/qt-5.0-wk2/fast/borders/border-radius-wide-border-05-expected.png: Added.
* platform/qt-5.0-wk2/fast/borders/border-radius-with-box-shadow-01-expected.png: Added.
* platform/qt-5.0-wk2/fast/borders/border-radius-with-box-shadow-expected.png: Added.
* platform/qt-5.0-wk2/fast/css-generated-content/details-summary-before-after-expected.png: Added.
* platform/qt-5.0-wk2/fast/css-generated-content/details-summary-before-after-expected.txt: Added.
* platform/qt-5.0-wk2/fast/forms/zoomed-controls-expected.png: Added.
* platform/qt-5.0-wk2/fast/gradients/unprefixed-color-stop-units-expected.png: Added.
* platform/qt-5.0-wk2/fast/gradients/unprefixed-generated-gradients-expected.png: Added.
* platform/qt-5.0-wk2/fast/gradients/unprefixed-linear-angle-gradients-expected.png: Added.
* platform/qt-5.0-wk2/fast/gradients/unprefixed-linear-right-angle-gradients-expected.png: Added.
* platform/qt-5.0-wk2/fast/gradients/unprefixed-list-item-gradient-expected.png: Added.
* platform/qt-5.0-wk2/fast/gradients/unprefixed-radial-gradients-expected.png: Added.
* platform/qt-5.0-wk2/fast/gradients/unprefixed-radial-gradients3-expected.png: Added.
* platform/qt-5.0-wk2/fast/gradients/unprefixed-repeating-end-fill-expected.png: Added.
* platform/qt-5.0-wk2/fast/gradients/unprefixed-repeating-linear-gradient-expected.png: Added.
* platform/qt-5.0-wk2/fast/gradients/unprefixed-repeating-radial-gradients-expected.png: Added.
* platform/qt-5.0-wk2/fast/gradients/unprefixed-zero-range-repeating-gradient-hang-expected.png: Added.
* platform/qt-5.0-wk2/svg/custom/marker-orient-auto-expected.png: Added.
* platform/qt-5.0-wk2/svg/custom/marker-orient-auto-expected.txt: Added.
* platform/qt-5.0-wk2/svg/filters/feDropShadow-zero-deviation-expected.png: Added.
* platform/qt-5.0-wk2/svg/filters/feDropShadow-zero-deviation-expected.txt: Added.
* platform/qt-5.0-wk2/svg/filters/feGaussianBlur-zero-deviation-expected.png: Added.
* platform/qt-5.0-wk2/svg/filters/feGaussianBlur-zero-deviation-expected.txt: Added.
* platform/qt/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151593 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Mac] clicking caption track glyph should dismiss menu
graouts@apple.com [Fri, 14 Jun 2013 14:43:01 +0000 (14:43 +0000)]
[Mac] clicking caption track glyph should dismiss menu
https://bugs.webkit.org/show_bug.cgi?id=117621

We now correctly capture all click events while the captions track menu
is showing, and ensure that no other control in the media player can be
activated as the track list gets dismissed by clicking anywhere but on
itself.

Reviewed by Eric Carlson.

* html/shadow/MediaControlsApple.cpp:
(WebCore::MediaControlsApple::showClosedCaptionTrackList):
Make the controls panel non-interactive and track click events in the
capture phase also on the controls container itself such that we can
identify correctly whether a click in the <video> shadow tree is on
the track list or another element (the target could only be resolved
to a shadow tree node if the event is handled within the shadow tree).

(WebCore::MediaControlsApple::hideClosedCaptionTrackList):
Make the controls panel interactive again and remove the new click
event handler on the controls container.

(WebCore::MediaControlsApple::handleClickEvent):
Capture any click event that targets a node anywhere outside of the track
list container and its subtree and hide the track list in this condition.

(WebCore::MediaControlsAppleEventListener::handleEvent):
Call handleClickEvent() on the media controls when a click event is handled.

* html/shadow/MediaControlsApple.h:
New public handleClickEvent() API such that it can be called from
MediaControlsAppleEventListener::handleEvent().

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151592 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[BlackBerry] Accelerated compositing layers that intersect the image plane are incorr...
anilsson@rim.com [Fri, 14 Jun 2013 14:04:20 +0000 (14:04 +0000)]
[BlackBerry] Accelerated compositing layers that intersect the image plane are incorrectly transformed
https://bugs.webkit.org/show_bug.cgi?id=117067

Reviewed by Carlos Garcia Campos.

PR 273550

The BlackBerry port used to mathematically project each corner of each
layer bounds rectangle, which gives the wrong results when the layer
intersects the image plane.

For display lists, we still got the correct appearance of the layer
contents, since the display list results in geometry that's transformed
in the vertex shader, on the GPU.

However, the transformed bounds are used for drawing the debug border,
performing visibility calculations on the CPU and drawing of WebGL,
masks/reflections and filters. Using the mathematical projection of
points that lie behind the image plane gives the wrong results.

The most important consequence of this bug was that the wrong area of
the layer would be considered visible, so a tiled layer would not
populate the right tiles (or any tiles at all).

Fixed by implementing something along the lines of the recipe in
section 6.2 of http://www.w3.org/TR/css3-transforms.

No new tests, manually testable by enabling debug border on
https://developer.mozilla.org/en-US/demos/detail/the-box/launch.

* platform/graphics/blackberry/EGLImageLayerCompositingThreadClient.cpp:
(WebCore::EGLImageLayerCompositingThreadClient::drawTextures):
* platform/graphics/blackberry/LayerCompositingThread.cpp:
(WebCore::LayerCompositingThread::LayerCompositingThread):
(WebCore::LayerCompositingThread::setDrawTransform):
(WebCore):
(WebCore::LayerCompositingThread::textureCoordinates):
(WebCore::LayerCompositingThread::drawTextures):
* platform/graphics/blackberry/LayerCompositingThread.h:
(LayerCompositingThread):
(WebCore::LayerCompositingThread::transformedBounds):
(WebCore::LayerCompositingThread::ws):
(WebCore::LayerCompositingThread::centerW):
* platform/graphics/blackberry/LayerFilterRenderer.cpp:
(WebCore::LayerFilterRenderer::applyActions):
* platform/graphics/blackberry/LayerRenderer.cpp:
(WebCore::compareLayerW):
(WebCore::LayerRenderer::compositeLayers):
(WebCore::LayerRenderer::drawDebugBorder):
(WebCore):
(WebCore::LayerRenderer::updateLayersRecursive):
(WebCore::LayerRenderer::compositeLayersRecursive):
* platform/graphics/blackberry/LayerRenderer.h:
(WebCore):
(LayerRenderer):
* platform/graphics/blackberry/LayerRendererSurface.cpp:
(WebCore::LayerRendererSurface::boundingBox):
(WebCore::LayerRendererSurface::transformedBounds):
* platform/graphics/blackberry/LayerRendererSurface.h:
(WebCore::LayerRendererSurface::setDrawTransform):
(WebCore::LayerRendererSurface::setReplicaDrawTransform):
(LayerRendererSurface):
* platform/graphics/blackberry/LayerUtilities.h: Added.
(WebCore):
(LayerClipPlane):
(WebCore::LayerClipPlane::LayerClipPlane):
(WebCore::LayerClipPlane::isPointInside):
(WebCore::LayerClipPlane::computeIntersection):
(WebCore::intersect):
(WebCore::boundingBox):
(WebCore::multVecMatrix):
(WebCore::toVector):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151591 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Qt] Unreviewed gardening. Some more rebaselining after r151205 and r150798.
zarvai@inf.u-szeged.hu [Fri, 14 Jun 2013 12:53:37 +0000 (12:53 +0000)]
[Qt] Unreviewed gardening. Some more rebaselining after r151205 and r150798.

* platform/qt-5.0-wk2/css2.1/20110323/absolute-replaced-height-004-expected.png: Added.
* platform/qt-5.0-wk2/css2.1/20110323/absolute-replaced-height-022-expected.png: Added.
* platform/qt-5.0-wk2/css2.1/t0805-c5518-brdr-t-01-e-expected.png: Added.
* platform/qt-5.0-wk2/css2.1/t170602-bdr-conflct-w-09-d-expected.png: Added.
* platform/qt-5.0-wk2/css3/selectors3/html/css3-modsel-70-expected.png: Added.
* platform/qt-5.0-wk2/css3/selectors3/xhtml/css3-modsel-131b-expected.png: Added.
* platform/qt-5.0-wk2/css3/unicode-bidi-isolate-basic-expected.png:
* platform/qt-5.0-wk2/editing/deleting/delete-image-002-expected.png:
* platform/qt-5.0-wk2/editing/deleting/smart-delete-004-expected.png:
* platform/qt-5.0-wk2/editing/selection/caret-rtl-2-expected.png:
* platform/qt-5.0-wk2/editing/selection/caret-rtl-2-left-expected.png:
* platform/qt-5.0-wk2/editing/selection/caret-rtl-expected.png:
* platform/qt-5.0-wk2/editing/selection/caret-rtl-right-expected.png: Added.
* platform/qt-5.0-wk2/editing/selection/extend-selection-bidi-expected.png:
* platform/qt-5.0-wk2/editing/selection/vertical-rl-rtl-extend-line-backward-br-expected.png:
* platform/qt-5.0-wk2/editing/selection/vertical-rl-rtl-extend-line-backward-p-expected.png:
* platform/qt-5.0-wk2/editing/selection/vertical-rl-rtl-extend-line-forward-br-expected.png:
* platform/qt-5.0-wk2/editing/selection/vertical-rl-rtl-extend-line-forward-p-expected.png:
* platform/qt-5.0-wk2/fast/block/positioning/auto/vertical-rl/001-expected.png: Added.
* platform/qt-5.0-wk2/fast/borders/rtl-border-05-expected.png:
* platform/qt-5.0-wk2/fast/box-shadow/box-shadow-transformed-expected.png:
* platform/qt-5.0-wk2/fast/css/error-in-last-decl-expected.png: Added.
* platform/qt-5.0-wk2/fast/css/font-size-negative-expected.png: Added.
* platform/qt-5.0-wk2/fast/css/namespaces/004-expected.png: Added.
* platform/qt-5.0-wk2/fast/css/table-text-align-strict-expected.png: Added.
* platform/qt-5.0-wk2/fast/dom/34176-expected.png:
* platform/qt-5.0-wk2/fast/dom/inner-text-expected.png: Added.
* platform/qt-5.0-wk2/fast/forms/HTMLOptionElement_label06-expected.png:
* platform/qt-5.0-wk2/fast/forms/HTMLOptionElement_label07-expected.png:
* platform/qt-5.0-wk2/fast/forms/button-cannot-be-nested-expected.png: Added.
* platform/qt-5.0-wk2/fast/forms/form-element-geometry-expected.png:
* platform/qt-5.0-wk2/fast/forms/input-appearance-default-bkcolor-expected.png: Added.
* platform/qt-5.0-wk2/fast/forms/listbox-bidi-align-expected.png:
* platform/qt-5.0-wk2/fast/forms/search-rtl-expected.png:
* platform/qt-5.0-wk2/fast/forms/select-baseline-expected.png:
* platform/qt-5.0-wk2/fast/inline/inline-box-background-expected.png:
* platform/qt-5.0-wk2/fast/inline/inline-box-background-long-image-expected.png:
* platform/qt-5.0-wk2/fast/inline/inline-box-background-repeat-x-expected.png:
* platform/qt-5.0-wk2/fast/inline/inline-box-background-repeat-y-expected.png:
* platform/qt-5.0-wk2/fast/invalid/014-expected.png: Added.
* platform/qt-5.0-wk2/fast/overflow/childFocusRingClip-expected.png: Added.
* platform/qt-5.0-wk2/fast/overflow/hit-test-overflow-controls-expected.png: Added.
* platform/qt-5.0-wk2/fast/repaint/float-in-new-block-with-layout-delta-expected.png: Added.
* platform/qt-5.0-wk2/fast/replaced/width-and-height-of-positioned-replaced-elements-expected.png: Added.
* platform/qt-5.0-wk2/fast/runin/generated-expected.png: Added.
* platform/qt-5.0-wk2/fast/table/multiple-captions-display-expected.png: Added.
* platform/qt-5.0-wk2/fast/table/remove-td-display-none-expected.png: Added.
* platform/qt-5.0-wk2/fast/text/international/bidi-ignored-for-first-child-inline-expected.png:
* platform/qt-5.0-wk2/fast/text/international/unicode-bidi-plaintext-in-textarea-expected.png:
* platform/qt-5.0-wk2/ietestcenter/css3/bordersbackgrounds/border-radius-applies-to-017-expected.png: Added.
* platform/qt-5.0-wk2/svg/W3C-SVG-1.1/animate-elem-16-t-expected.png: Added.
* platform/qt-5.0-wk2/svg/W3C-SVG-1.1/animate-elem-23-t-expected.png: Added.
* platform/qt-5.0-wk2/svg/W3C-SVG-1.1/animate-elem-31-t-expected.png: Added.
* platform/qt-5.0-wk2/svg/W3C-SVG-1.1/animate-elem-85-t-expected.png: Added.
* platform/qt-5.0-wk2/svg/W3C-SVG-1.1/coords-trans-04-t-expected.png: Added.
* platform/qt-5.0-wk2/svg/W3C-SVG-1.1/coords-units-02-b-expected.png: Added.
* platform/qt-5.0-wk2/svg/W3C-SVG-1.1/fonts-glyph-04-t-expected.png: Added.
* platform/qt-5.0-wk2/svg/W3C-SVG-1.1/painting-marker-01-f-expected.png: Added.
* platform/qt-5.0-wk2/svg/W3C-SVG-1.1/paths-data-01-t-expected.png: Added.
* platform/qt-5.0-wk2/svg/W3C-SVG-1.1/pservers-grad-10-b-expected.png: Added.
* platform/qt-5.0-wk2/svg/W3C-SVG-1.1/pservers-grad-16-b-expected.png: Added.
* platform/qt-5.0-wk2/svg/W3C-SVG-1.1/script-handle-01-b-expected.png: Added.
* platform/qt-5.0-wk2/svg/W3C-SVG-1.1/shapes-rect-01-t-expected.png: Added.
* platform/qt-5.0-wk2/svg/W3C-SVG-1.1/struct-cond-03-t-expected.png: Added.
* platform/qt-5.0-wk2/svg/W3C-SVG-1.1/struct-defs-01-t-expected.png: Added.
* platform/qt-5.0-wk2/svg/W3C-SVG-1.1/text-align-02-b-expected.png: Added.
* platform/qt-5.0-wk2/svg/W3C-SVG-1.1/text-align-08-b-expected.png: Added.
* platform/qt-5.0-wk2/svg/W3C-SVG-1.1/text-text-03-b-expected.png: Added.
* platform/qt-5.0-wk2/svg/custom/text-ctm-expected.png: Added.
* platform/qt-5.0-wk2/svg/custom/use-instanceRoot-event-listeners-expected.png:
* platform/qt-5.0-wk2/tables/mozilla/bugs/bug81934-expected.png: Added.
* platform/qt-5.0-wk2/tables/mozilla/marvin/tables_bgcolor_red-expected.png: Added.
* platform/qt-5.0-wk2/tables/mozilla/marvin/tfoot_align_left-expected.png: Added.
* platform/qt-5.0-wk2/tables/mozilla/marvin/x_td_align_right-expected.png: Added.
* platform/qt-5.0-wk2/tables/mozilla/marvin/x_th_class-expected.png: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151590 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Qt] Unreviewed TestExpectation cleanup
commit-queue@webkit.org [Fri, 14 Jun 2013 11:43:19 +0000 (11:43 +0000)]
[Qt] Unreviewed TestExpectation cleanup
https://bugs.webkit.org/show_bug.cgi?id=117626

Unreviewed gardening.

Patch by Seokju Kwon <seokju.kwon@gmail.com> on 2013-06-14

* platform/qt-5.0-wk1/TestExpectations:
* platform/qt/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151589 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Qt] Map rewind and fast forward keys to match GoogleTV
allan.jensen@digia.com [Fri, 14 Jun 2013 11:34:27 +0000 (11:34 +0000)]
[Qt] Map rewind and fast forward keys to match GoogleTV
https://bugs.webkit.org/show_bug.cgi?id=117634

Reviewed by Jocelyn Turcotte.

Set the same keycode for media keys rewind and fast-forward as
GoogleTV does.

* platform/qt/PlatformKeyboardEventQt.cpp:
(WebCore::windowsKeyCodeForKeyEvent):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151588 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed GTK build fix.
zandobersek@gmail.com [Fri, 14 Jun 2013 11:26:39 +0000 (11:26 +0000)]
Unreviewed GTK build fix.

* GNUmakefile.am: Add the Freetype CFLAGS to the list of libPlatformGtk's CPPFLAGS. These are required as there are
specific build targets that end up including the HarfBuzz headers which are included in the Freetype CFLAGS variable.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151587 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoAvoid unnecessary data copies when loading subresources with DoNotBufferData option
carlosgc@webkit.org [Fri, 14 Jun 2013 10:59:11 +0000 (10:59 +0000)]
Avoid unnecessary data copies when loading subresources with DoNotBufferData option
https://bugs.webkit.org/show_bug.cgi?id=115804

Reviewed by Darin Adler.

When DoNotBufferData option is used to load a resource its data
is always copied before sending it to the CachedResource. Most
of the cached resources ignore the incremental data and wait
until all data has been received to save the ResourceBuffer,
that will be NULL anyway when DoNotBufferData is used.
CachedRawResource notifies its clients about the incremental
data, but it doesn't save the data when DoNotBufferData option
is present. In those cases we are unnecessary copying the data.

CachedResource::data has been split into
CachedResource::addDataBuffer() used for incremental data chunks
when buffering data, CachedResource::addData() used for
incremental data chunks when not buffering and
CachedResource::finishLoading() used to finish the loading. This
way we get rid of the allDataReceived boolean parameter and cached
resources not interested in incremenetal data chunks only have to
implement finishLoading() without having to check if all data have
been received or not.

SubresourceLoader::sendDataToResource was always called after
checking if loading multipart content, and then it was checked
again to decided whether to copy the data or not. It has been
removed in favor of calling directly the resource methods,
finishLoading for multipart content, addDataBuffer for data chunks
when buffering and addData for data chunks when not buffering.

No new functionality, covered by existing tests.

* html/ImageDocument.cpp:
(WebCore::ImageDocumentParser::appendBytes): Update to API changes.
(WebCore::ImageDocumentParser::finish): Ditto.
* loader/SubresourceLoader.cpp:
(WebCore::SubresourceLoader::didReceiveResponse): Call
finishLoading() for multipart content.
(WebCore::SubresourceLoader::didReceiveDataOrBuffer): Add data to
the resource using addDataBuffer or addData depending on whether
we are buffering or not.
(WebCore::SubresourceLoader::didFinishLoading): Call
finishLoading() for the cached resource instead of data.
* loader/SubresourceLoader.h:
* loader/cache/CachedCSSStyleSheet.cpp:
(WebCore::CachedCSSStyleSheet::finishLoading):
* loader/cache/CachedCSSStyleSheet.h:
* loader/cache/CachedFont.cpp:
(WebCore::CachedFont::finishLoading):
* loader/cache/CachedFont.h:
* loader/cache/CachedImage.cpp:
(WebCore::CachedImage::isValidDecodedImageSize): Helper function
to check if the image size is valid.
(WebCore::CachedImage::addIncrementalDataBuffer): Helper function
to add incremental data buffer.
(WebCore::CachedImage::addDataBuffer): Call
addIncrementalDataBuffer().
(WebCore::CachedImage::addData): Create a ResourceBuffer and call
addIncrementalDataBuffer().
(WebCore::CachedImage::finishLoading):
* loader/cache/CachedImage.h:
* loader/cache/CachedRawResource.cpp:
(WebCore::CachedRawResource::calculateIncrementalDataChunk):
Returns a pointer to the data corresponding to the current chunk
and its length.
(WebCore::CachedRawResource::addDataBuffer): Assert to make sure
this is only called when BufferData option is present. Use
calculateIncrementalDataChunk().
(WebCore::CachedRawResource::addData): Assert to make sure this is
only called when DoNotBufferData option is present.
(WebCore::CachedRawResource::finishLoading):
(WebCore::CachedRawResource::notifyClientsDataWasReceived): Helper
private function to notify the clients about data received.
* loader/cache/CachedRawResource.h:
* loader/cache/CachedResource.cpp:
(WebCore::CachedResource::addDataBuffer):
(WebCore::CachedResource::addData):
(WebCore::CachedResource::finishLoading):
* loader/cache/CachedResource.h:
* loader/cache/CachedSVGDocument.cpp:
(WebCore::CachedSVGDocument::finishLoading):
* loader/cache/CachedSVGDocument.h:
* loader/cache/CachedScript.cpp:
(WebCore::CachedScript::finishLoading):
* loader/cache/CachedScript.h:
* loader/cache/CachedShader.cpp:
(WebCore::CachedShader::finishLoading):
* loader/cache/CachedShader.h:
* loader/cache/CachedTextTrack.cpp:
(WebCore::CachedTextTrack::addDataBuffer):
(WebCore::CachedTextTrack::finishLoading):
* loader/cache/CachedTextTrack.h:
* loader/cache/CachedXSLStyleSheet.cpp:
(WebCore::CachedXSLStyleSheet::finishLoading):
* loader/cache/CachedXSLStyleSheet.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151586 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoRemove myself from Qt build system watchlist
vestbo@webkit.org [Fri, 14 Jun 2013 10:03:32 +0000 (10:03 +0000)]
Remove myself from Qt build system watchlist

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@151585 268f45cc-cd09-0410-ab3c-d52691b4dbfc