WebKit-https.git
6 years agoUse a separate backdrop element to allow cues to have highlight and background color
bfulgham@apple.com [Fri, 11 Jul 2014 17:32:17 +0000 (17:32 +0000)]
Use a separate backdrop element to allow cues to have highlight and background color
https://bugs.webkit.org/show_bug.cgi?id=134821
<rdar://problem/15999721>

Reviewed by Eric Carlson.

Source/WebCore:
Add a new <div> element wrapping the existing cue <span>. This allows
us to have a highlight on the cue (in the <span> background), as well
as an overall background color.

* Modules/mediacontrols/mediaControlsApple.css:
(video::-webkit-media-text-track-display-backdrop): New markup for
the backdrop element of the caption.
* html/track/VTTCue.cpp:
(WebCore::VTTCue::cueBackdropShadowPseudoId): Added to
allow user customization of the cue backdrop.
(WebCore::VTTCue::initialize): Rename the old "m_cueBackgroundBox" to
"m_cueHighlightBox" and add a new "m_cueBackdropBox" member.
(WebCore::VTTCue::updateDisplayTree): Update for m_cueHighlightBox.
(WebCore::VTTCue::getDisplayTree): Make m_cueHighlightBox a child
of the new m_cueBackdropBox element, and add m_cueBackdropBox to
the display tree.
* html/track/VTTCue.h:
(WebCore::VTTCue::element):
* page/CaptionUserPreferencesMediaAF.cpp:
(WebCore::CaptionUserPreferencesMediaAF::setInterestedInCaptionPreferenceChanges):
Fix for missing caption style updates. Even if we are already
listening for caption changes, we still want to update the new
instance's style sheet to match.
(WebCore::CaptionUserPreferencesMediaAF::captionsStyleSheetOverride):
* rendering/RenderVTTCue.cpp:
(WebCore::RenderVTTCue::initializeLayoutParameters): Take the new
<div> into consideration when looking for the Cue text element.

LayoutTests:
Updated tests for new formatting logic.

* platform/mac/media/track/track-cue-rendering-horizontal-expected.png:
* platform/mac/media/track/track-cue-rendering-horizontal-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@171004 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoTighten WebContent sandbox
oliver@apple.com [Fri, 11 Jul 2014 17:15:31 +0000 (17:15 +0000)]
Tighten WebContent sandbox
https://bugs.webkit.org/show_bug.cgi?id=134834

Reviewed by Sam Weinig.

Define a much tighter sandbox profile for the WebContent process

* Resources/SandboxProfiles/ios/com.apple.WebKit.WebContent.sb:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@171003 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION (r170163?): Web content shifts revealing space equivalent to the find...
antti@apple.com [Fri, 11 Jul 2014 16:29:33 +0000 (16:29 +0000)]
REGRESSION (r170163?): Web content shifts revealing space equivalent to the find bar when clicking a link while a phrase is targeted via Cmd+F
https://bugs.webkit.org/show_bug.cgi?id=134833
<rdar://problem/17580021>

Reviewed by Zalan Bujtas.

Some versions of OS X Safari can't handle the new unfreeze timing. Revert back to DidFirstLayout on Mac.

* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::dispatchDidLayout):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@171002 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoSubpixel layout: return integral results for offset*, client*, scroll* by default.
zalan@apple.com [Fri, 11 Jul 2014 15:27:42 +0000 (15:27 +0000)]
Subpixel layout: return integral results for offset*, client*, scroll* by default.
https://bugs.webkit.org/show_bug.cgi?id=134651

Reviewed by Simon Fraser.

Revert to returning integral values for Element.offset* client* scroll* by default.
Fractional values break number of sites(tight design) and JS frameworks(fail to handle fractional values).

Since snapped dimension depends on both the original point and the width/height of the box,
we need to call RenderBoxModelObject::pixelSnapped*() helpers, instead of round().

Covered by existing tests

Source/WebCore:

* dom/Element.cpp:
(WebCore::subpixelMetricsEnabled):
(WebCore::convertToNonSubpixelValueIfNeeded):
(WebCore::Element::offsetLeft):
(WebCore::Element::offsetTop):
(WebCore::Element::offsetWidth):
(WebCore::Element::offsetHeight):
(WebCore::Element::clientLeft):
(WebCore::Element::clientTop):
(WebCore::Element::clientWidth):
(WebCore::Element::clientHeight):
* page/Settings.in:

Source/WebKit/mac:

* WebView/WebPreferences.mm:
(+[WebPreferences initialize]):

Source/WebKit2:

* Shared/WebPreferencesDefinitions.h:
* UIProcess/API/C/WKPreferencesRefPrivate.h:

LayoutTests:

* cssom/subpixel-offsetleft-top-width-height-values-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@171001 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoSubpixel rendering: icloud.com password arrow has clipped circle at some window sizes.
zalan@apple.com [Fri, 11 Jul 2014 14:19:12 +0000 (14:19 +0000)]
Subpixel rendering: icloud.com password arrow has clipped circle at some window sizes.
https://bugs.webkit.org/show_bug.cgi?id=134824
<rdar://problem/17612729>

Reviewed by Simon Fraser.

Transparent layer clipping needs device pixel offset adjusting.

Source/WebCore:
Test: fast/layers/wrong-clipping-semi-transparent-compositing-layer-on-subpixel-position.html

* rendering/RenderLayer.cpp:
(WebCore::RenderLayer::beginTransparencyLayers):
(WebCore::RenderLayer::paintLayer):
(WebCore::RenderLayer::paintLayerContents):
(WebCore::RenderLayer::paintBackgroundForFragments):
(WebCore::RenderLayer::paintForegroundForFragments):
* rendering/RenderLayer.h:

LayoutTests:
* fast/layers/wrong-clipping-semi-transparent-compositing-layer-on-subpixel-position-expected.html: Added.
* fast/layers/wrong-clipping-semi-transparent-compositing-layer-on-subpixel-position.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@171000 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed, rolling out r170995.
commit-queue@webkit.org [Fri, 11 Jul 2014 12:26:08 +0000 (12:26 +0000)]
Unreviewed, rolling out r170995.
https://bugs.webkit.org/show_bug.cgi?id=134831

Causing odd crashes in debug builds (Requested by zdobersek on
#webkit).

Reverted changeset:

"[WTF] Add the move constructor, move assignment operator for
HashTable"
https://bugs.webkit.org/show_bug.cgi?id=130772
http://trac.webkit.org/changeset/170995

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170999 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed. Remove unused file.
carlosgc@webkit.org [Fri, 11 Jul 2014 12:15:48 +0000 (12:15 +0000)]
Unreviewed. Remove unused file.

* html/shadow/MediaControlsGtk.h: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170998 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[GTK] Enable VIDEO_TRACK by default
carlosgc@webkit.org [Fri, 11 Jul 2014 12:12:43 +0000 (12:12 +0000)]
[GTK] Enable VIDEO_TRACK by default
https://bugs.webkit.org/show_bug.cgi?id=134801

Reviewed by Philippe Normand.

* Source/cmake/OptionsGTK.cmake:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170997 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[CSS Grid Layout] Update grid-auto-flow to the new syntax
rego@igalia.com [Fri, 11 Jul 2014 10:42:32 +0000 (10:42 +0000)]
[CSS Grid Layout] Update grid-auto-flow to the new syntax
https://bugs.webkit.org/show_bug.cgi?id=134057

Reviewed by Sergio Villar Senin.

Source/WebCore:
In last versions of the spec grid-auto-flow syntax has changed. New
syntax is:
[ row | column ] && dense? | stack && [ row | column ]?

Implemented parsing for new syntax and added/modified test cases in
current layout tests.

For the moment, the implementation keeps working in the same way, but
using "stack" value for grid-auto-flow property instead of "none". This
should be fixed in a follow-up patch once "stack" is properly
implemented.

Also "dense" needs to be reviewed. Right now auto-placement algorithm is
always "dense" and never "sparse".

No new tests, update current tests to add new cases.

* css/CSSComputedStyleDeclaration.cpp:
(WebCore::ComputedStyleExtractor::propertyValue): Adapt to new syntax.
* css/CSSParser.cpp:
(WebCore::isValidKeywordPropertyAndValue): Removed grid-auto-flow as it
is not a keyword anymore.
(WebCore::isKeywordPropertyID): Ditto.
(WebCore::CSSParser::parseValue): Add specific method for parsing
grid-auto-flow.
(WebCore::CSSParser::parseGridShorthand): Adapt parsing for
grid-auto-flow property inside the shorthand.
(WebCore::isValidGridAutoFlowId): Helper method to check if the CSSValue
id is a valid keyword for grid-auto-flow property.
(WebCore::CSSParser::parseGridAutoFlow): Implement new parsing for
grid-auto-flow.
* css/CSSParser.h: Method header for grid-auto-flow parsing.
* css/CSSPrimitiveValueMappings.h:
(WebCore::CSSPrimitiveValue::operator GridAutoFlow): Deleted.
(WebCore::CSSPrimitiveValue::CSSPrimitiveValue): Deleted.
* css/CSSValueKeywords.in: Add new keywords required by grid-auto-flow:
"dense" and "stack".
* css/DeprecatedStyleBuilder.cpp:
(WebCore::DeprecatedStyleBuilder::DeprecatedStyleBuilder): Remove
default handler as grid-auto-flow is not a keyword now.
* css/StyleResolver.cpp:
(WebCore::StyleResolver::applyProperty): Implement particular handler
for new grid-auto-flow syntax.
* rendering/RenderGrid.cpp:
(WebCore::RenderGrid::placeItemsOnGrid): Keep old "none" behavior as
"stack" behavior for the moment.
(WebCore::RenderGrid::autoPlacementMajorAxisDirection): Use the new
helper methods in RenderStyle.
(WebCore::RenderGrid::autoPlacementMinorAxisDirection): Ditto.
* rendering/style/GridResolvedPosition.cpp:
(WebCore::GridResolvedPosition::resolveGridPositionsFromStyle): Again
keep old "none" behavior for "stack".
* rendering/style/RenderStyle.h: Add new helper methods to know the
direction and algorithm of grid-auto-flow property.
* rendering/style/RenderStyleConstants.h: Redefine GridAutoFlow enum
using flags for algorithm and direction.
* rendering/style/StyleGridData.h: Change m_gridAutoFlow type to
unsigned.

LayoutTests:
Adapted current tests to the new syntax. Keeping old "none" behavior
using "stack".

Add new use cases to cover the new syntax to grid-auto-flow-get-set.html
and grid-shorthand-get-set.html.

* fast/css-grid-layout/grid-auto-flow-get-set-expected.txt:
* fast/css-grid-layout/grid-auto-flow-get-set.html:
* fast/css-grid-layout/grid-auto-flow-resolution.html:
* fast/css-grid-layout/grid-auto-flow-update-expected.txt:
* fast/css-grid-layout/grid-auto-flow-update.html:
* fast/css-grid-layout/grid-item-named-grid-area-resolution.html:
* fast/css-grid-layout/grid-item-z-index-change-repaint.html:
* fast/css-grid-layout/grid-item-z-index-stacking-context.html:
* fast/css-grid-layout/grid-shorthand-get-set-expected.txt:
* fast/css-grid-layout/grid-shorthand-get-set.html:
* fast/css-grid-layout/named-grid-lines-with-named-grid-areas-dynamic-get-set.html:
* fast/css-grid-layout/named-grid-lines-with-named-grid-areas-resolution.html:
* fast/css-grid-layout/resources/grid.css:
(.gridAutoFlowStack):
(.gridAutoFlowNone): Deleted.
* fast/css/getComputedStyle/computed-style-expected.txt:
* fast/css/getComputedStyle/computed-style-without-renderer-expected.txt:
* ietestcenter/css3/grid/grid-column-001.htm:
* ietestcenter/css3/grid/grid-column-002.htm:
* ietestcenter/css3/grid/grid-column-003.htm:
* svg/css/getComputedStyle-basic-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170996 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[WTF] Add the move constructor, move assignment operator for HashTable
zandobersek@gmail.com [Fri, 11 Jul 2014 08:16:36 +0000 (08:16 +0000)]
[WTF] Add the move constructor, move assignment operator for HashTable
https://bugs.webkit.org/show_bug.cgi?id=130772

Reviewed by Darin Adler.

HashTable has both copy constructor and copy assignment operator, meaning that the move constructor
and move assignment operator are implicitly deleted. This patch defines both to avoid unnecessary
copies when moves can be performed.

* wtf/HashTable.h:
(WTF::KeyTraits>::HashTable):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170995 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoVersioning.
lforschler@apple.com [Fri, 11 Jul 2014 06:01:26 +0000 (06:01 +0000)]
Versioning.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170993 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoWeb Inspector: Debugger Continue button is incorrectly sized
commit-queue@webkit.org [Fri, 11 Jul 2014 05:06:55 +0000 (05:06 +0000)]
Web Inspector: Debugger Continue button is incorrectly sized
https://bugs.webkit.org/show_bug.cgi?id=134786

Patch by Joseph Pecoraro <pecoraro@apple.com> on 2014-07-10
Reviewed by Timothy Hatcher.

Before this patch everything in the DebuggerDashboardView was flexing.
Normally, if the contents were not too wide, this would just behave
as expected and be centered. If the content was too wide, then each
element would flex uniformly and overlap uniformly. That was not expected.

This patch changes makes everything but the link have a non-flexible
size. If the link (at the end) flexs it will have a nice ellipsis
text truncation, and the rest of the content will be left justified.

This also fixes a few other minor issues seen in the view.

* UserInterface/Views/DebuggerDashboardView.css:
(.dashboard.debugger .navigation-bar):
Reduce the complexity, give the nav bar (the resume button) some simple padding.

(.dashboard.debugger > div):
Make everything non-flex by default.

(.dashboard.debugger > .location):
The location will have an inner element flex, so make it flex.

(.dashboard.debugger > .location > *):
Make everything in .location non-flex by default.

(.dashboard.debugger > .location .go-to-link):
Make the link flex and truncate cleanly.

(.dashboard.debugger > .location .function-name):
Fix vertical alignment when the toolbar has small icons.

* UserInterface/Views/DebuggerDashboardView.js:
(WebInspector.DebuggerDashboardView.prototype._rebuildLocation):
Fix class name spelling.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170991 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed EFL build fix after r170970.
jinwoo7.song@samsung.com [Fri, 11 Jul 2014 04:26:31 +0000 (04:26 +0000)]
Unreviewed EFL build fix after r170970.

* UIProcess/efl/WebContextEfl.cpp:
(WebKit::WebContext::platformDefaultOpenGLCacheDirectory):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170990 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION(r170970) REGRESSION(r170974): [GTK] Build broken.
commit-queue@webkit.org [Fri, 11 Jul 2014 03:50:15 +0000 (03:50 +0000)]
REGRESSION(r170970) REGRESSION(r170974): [GTK] Build broken.
https://bugs.webkit.org/show_bug.cgi?id=134825

Unreviewed GTK build fix.

Patch by Carlos Alberto Lopez Perez <clopez@igalia.com> on 2014-07-10

* Shared/SessionState.h: Put ifdefs for ViewSnapshot on Mac port.
* Shared/WebBackForwardListItem.h: Idem.
* UIProcess/gtk/WebContextGtk.cpp:
(WebKit::WebContext::platformDefaultOpenGLCacheDirectory): Implement skeleton.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170989 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[iOS][WK2] It should be safe to call WKContentViewInteraction's cleanupInteraction...
benjamin@webkit.org [Fri, 11 Jul 2014 02:29:33 +0000 (02:29 +0000)]
[iOS][WK2] It should be safe to call WKContentViewInteraction's cleanupInteraction multiple times
https://bugs.webkit.org/show_bug.cgi?id=134820

Patch by Benjamin Poulain <bpoulain@apple.com> on 2014-07-10
Reviewed by Andreas Kling.

If a view is destroyed just after a crash, "cleanupInteraction" is called twice: once on crash,
once on dealloc.

The code handling _interactionViewsContainerView is using KVO to monitor transform changes. It is not safe
to remove the observer if we are not already observing on that view.

To solve the problem, this patch makes the cleanup actually remove the view so that setup and cleanup
are completely symmetrical. If cleanup is called twice, the second time would not enter the branch because
the view is already nil.

* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView setupInteraction]):
(-[WKContentView cleanupInteraction]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170988 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[iOS WK2] Move WKInspectorHighlightView to its own file
simon.fraser@apple.com [Fri, 11 Jul 2014 00:49:02 +0000 (00:49 +0000)]
[iOS WK2] Move WKInspectorHighlightView to its own file
https://bugs.webkit.org/show_bug.cgi?id=134819

Reviewed by Joseph Pecoraro.

WKInspectorHighlightView brought a lot of path/quad-related code into WKContentView.mm,
so move it into its own file.

* UIProcess/WKInspectorHighlightView.h: Added.
* UIProcess/WKInspectorHighlightView.mm: Added.
(-[WKInspectorHighlightView dealloc]):
(-[WKInspectorHighlightView _removeAllLayers]):
(-[WKInspectorHighlightView _createLayers:]):
(findIntersectionOnLineBetweenPoints):
(quadIntersection):
(layerPathWithHole):
(layerPath):
(-[WKInspectorHighlightView _layoutForNodeHighlight:]):
(-[WKInspectorHighlightView _layoutForRectsHighlight:]):
(-[WKInspectorHighlightView update:]):
* UIProcess/ios/WKContentView.mm:
(-[WKInspectorHighlightView initWithFrame:]): Deleted.
(-[WKInspectorHighlightView dealloc]): Deleted.
(-[WKInspectorHighlightView _removeAllLayers]): Deleted.
(-[WKInspectorHighlightView _createLayers:]): Deleted.
(findIntersectionOnLineBetweenPoints): Deleted.
(quadIntersection): Deleted.
(layerPathWithHole): Deleted.
(layerPath): Deleted.
(-[WKInspectorHighlightView _layoutForNodeHighlight:]): Deleted.
(-[WKInspectorHighlightView _layoutForRectsHighlight:]): Deleted.
(-[WKInspectorHighlightView update:]): Deleted.
* WebKit2.xcodeproj/project.pbxproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170987 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[iOS] Pasting rich content does not perform a two-step paste
mmaxfield@apple.com [Fri, 11 Jul 2014 00:12:02 +0000 (00:12 +0000)]
[iOS] Pasting rich content does not perform a two-step paste
https://bugs.webkit.org/show_bug.cgi?id=134795

Reviewed by Benjamin Poulain.

Source/WebCore:
UIKit requires writing something (anything!) to a particular pasteboard
during copy in order to flag the content as rich. When a paste occurs,
UIKit will read this flag and perform a two-step paste.

Test: platform/ios-sim/editing/pasteboard/two-step-paste.html

* platform/ios/PlatformPasteboardIOS.mm:
(WebCore::PlatformPasteboard::write):

LayoutTests:
Copy and paste some text, and make sure that the two step paste occurred.
The mechanism for detecting this is whether or not styles have changed appropriately.

* platform/ios-sim/editing/pasteboard/two-step-paste-expected.txt: Added.
* platform/ios-sim/editing/pasteboard/two-step-paste.html: Added.

Tools:
Two-step paste requires setting InputViewObeysDOMFocus on the UIWebBrowserView.

* DumpRenderTree/mac/DumpRenderTree.mm:
(createWebViewAndOffscreenWindow):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170986 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUpdate the ChangeLog per Pratik's review comments
aestes@apple.com [Fri, 11 Jul 2014 00:04:55 +0000 (00:04 +0000)]
Update the ChangeLog per Pratik's review comments

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170985 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[iOS] Some QuickLook documents are not displayed as previews
aestes@apple.com [Fri, 11 Jul 2014 00:02:51 +0000 (00:02 +0000)]
[iOS] Some QuickLook documents are not displayed as previews
https://bugs.webkit.org/show_bug.cgi?id=134788

Reviewed by Pratik Solanki.

ResourceRequest::isMainResourceRequest() is not reliable (e.g. ResourceRequests reconstituted from an
NSURLRequest have no way to set this bit properly) and should probably be removed. Instead of using
isMainResourceRequest() to check if a QuickLookHandle should be created when the Network Process is in use,
just check if the current ResourceLoader is the same as the DocumentLoader's ResourceLoader.

No new tests. QuickLook is not testable from WebKit.

* platform/network/ios/QuickLook.mm:
(WebCore::QuickLookHandle::create):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170984 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION (r170935): WKWebView is always transparent until the first layer tree...
timothy_horton@apple.com [Thu, 10 Jul 2014 23:46:43 +0000 (23:46 +0000)]
REGRESSION (r170935): WKWebView is always transparent until the first layer tree commit
https://bugs.webkit.org/show_bug.cgi?id=134818
<rdar://problem/17632468>

Reviewed by Anders Carlsson.

* UIProcess/API/Cocoa/WKWebView.mm:
(scrollViewBackgroundColor):
r170935 made it so that we would initialize the scroll view background color to an
invalid color until the first layer tree commit. We should go with white instead.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170982 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoAdd a mechanism to notify the UIProcess when an editing command is done executing.
enrica@apple.com [Thu, 10 Jul 2014 23:26:19 +0000 (23:26 +0000)]
Add a mechanism to notify the UIProcess when an editing command is done executing.
https://bugs.webkit.org/show_bug.cgi?id=134807

Reviewed by Tim Horton.

Some editing commands have an effect on some parts of the system that
run inside the UIProcess. A good example are the cursor movement commands
that require an update of the autocorrection/autosuggestion machinery.
This patch adds a way to reliably know when the command has been executed
in the WebProcess. A previous attempt at solving this problem was added in
r170858 and was partially reverted in r170948.
The change also removes the selectionWillChange notification added in r170858.

* UIProcess/PageClient.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::editorStateChanged):
* UIProcess/WebPageProxy.h:
* UIProcess/ios/PageClientImplIOS.h:
* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::selectionWillChange): Deleted.
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView executeEditCommandWithCallback:]):
(-[WKContentView _moveUp:withHistory:]):
(-[WKContentView _moveDown:withHistory:]):
(-[WKContentView _moveLeft:withHistory:]):
(-[WKContentView _moveRight:withHistory:]):
(-[WKContentView _moveToStartOfWord:withHistory:]):
(-[WKContentView _moveToStartOfParagraph:withHistory:]):
(-[WKContentView _moveToStartOfLine:withHistory:]):
(-[WKContentView _moveToStartOfDocument:withHistory:]):
(-[WKContentView _moveToEndOfWord:withHistory:]):
(-[WKContentView _moveToEndOfParagraph:withHistory:]):
(-[WKContentView _moveToEndOfLine:withHistory:]):
(-[WKContentView _moveToEndOfDocument:withHistory:]):
(-[WKContentView _selectionWillChange]): Deleted.
* UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::executeEditCommand):
(WebKit::WebPageProxy::notifySelectionWillChange): Deleted.
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::executeEditCommandWithCallback):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170981 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Mac] NSWindow warning: adding an unknown subview opening detached Inspector
commit-queue@webkit.org [Thu, 10 Jul 2014 23:20:51 +0000 (23:20 +0000)]
[Mac] NSWindow warning: adding an unknown subview opening detached Inspector
https://bugs.webkit.org/show_bug.cgi?id=134813

Patch by Joseph Pecoraro <pecoraro@apple.com> on 2014-07-10
Reviewed by Timothy Hatcher.

Source/WebKit/mac:
* WebCoreSupport/WebInspectorClient.mm:
(-[WebInspectorWindowController window]):
Use a selector that will avoid the warning message.

Source/WebKit2:
* UIProcess/mac/WebInspectorProxyMac.mm:
(WebKit::WebInspectorProxy::createInspectorWindow):
Use a selector that will avoid the warning message.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170980 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoRemove use of container relative restrictions in the network process sandbox
oliver@apple.com [Thu, 10 Jul 2014 23:17:12 +0000 (23:17 +0000)]
Remove use of container relative restrictions in the network process sandbox
https://bugs.webkit.org/show_bug.cgi?id=134816

Reviewed by Anders Carlsson.

As i'm tidying up the various sandboxes and that's meaning we
need to reduce some file restrictions in the network process.

* Resources/SandboxProfiles/ios/com.apple.WebKit.Networking.sb:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170979 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[iOS WebKit2] Don't uninstall the memory pressure listener after first signal.
akling@apple.com [Thu, 10 Jul 2014 22:26:02 +0000 (22:26 +0000)]
[iOS WebKit2] Don't uninstall the memory pressure listener after first signal.
<https://webkit.org/b/134814>

The call to uninstall() should not be made on iOS since iOS doesn't use the
hold-off mechanism in MemoryPressureHandler.

Reviewed by Benjamin Poulain.

* platform/cocoa/MemoryPressureHandlerCocoa.mm:
(WebCore::MemoryPressureHandler::respondToMemoryPressure):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170978 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed iOS build fix after r170974. Define id if building a non ObjC file.
psolanki@apple.com [Thu, 10 Jul 2014 21:35:52 +0000 (21:35 +0000)]
Unreviewed iOS build fix after r170974. Define id if building a non ObjC file.

* UIProcess/mac/ViewSnapshotStore.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170977 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[MSE] Overlapping appended ranges must cause the decoder to flush and re-enqueue.
jer.noble@apple.com [Thu, 10 Jul 2014 21:28:05 +0000 (21:28 +0000)]
[MSE] Overlapping appended ranges must cause the decoder to flush and re-enqueue.
https://bugs.webkit.org/show_bug.cgi?id=134805

Reviewed by Eric Carlson.

When appending a range of media data which overlaps with an existing range, SourceBuffer
must cause the decoder to flush and re-enqueue samples. Those samples removed by appending
the overlapping range may have already been enqueued for display, and unless the decode
queue is flushed, corruption or decode errors may occur as the new samples are enqueued for
display.

Add a boolean flag onto TrackBuffer to indicate whether the decoder needs to be flushed and
re-enqueued the next time an append operation completes. Set this flag whenever samples are
removed due to overlapping or an explicit call to removeCodedFrames(). Move the contents of
sourceBufferPrivateSeekToTime() (which previously did flushing and re-enqueueing) into a new
function, reenqueueMediaForTime(), which can be called from sourceBufferPrivateAppendComplete().

Drive-by fix: findSyncSampleAfterDecodeIterator() would return the passed in iterator if that
sample is a sync-sample. Fix this to correctly return the next sync sample.

Drive-by fix: Use a SampleMap, rather than a DecodeOrderSampleMap, to track erased samples
so that the erasedSamples are correctly accounted for in both presentation and decode orders.

* Modules/mediasource/SampleMap.cpp:
(WebCore::SampleMap::empty): Add convenience method.
(WebCore::DecodeOrderSampleMap::findSyncSampleAfterDecodeIterator): Drive-by fix.
* Modules/mediasource/SampleMap.h:
(WebCore::SampleMap::decodeOrder): Added const accessor.
(WebCore::SampleMap::presentationOrder): Ditto.
(WebCore::SampleMap::addRange): Added.
* Modules/mediasource/SourceBuffer.cpp:
(WebCore::SourceBuffer::TrackBuffer::TrackBuffer): Add needsReenqueeing flag.
(WebCore::SourceBuffer::sourceBufferPrivateSeekToTime): Move contents into reenqueueMediaForTime().
(WebCore::SourceBuffer::sourceBufferPrivateAppendComplete): Call reenqueMediaForTime() if necessary.
(WebCore::SourceBuffer::removeCodedFrames): Set needsReenqueing.
(WebCore::SourceBuffer::sourceBufferPrivateDidReceiveSample): Ditto.
(WebCore::SourceBuffer::provideMediaData): Drive-by fix.
(WebCore::SourceBuffer::reenqueueMediaForTime): Moved from sourceBufferPrivateSeekToTime().
* Modules/mediasource/SourceBuffer.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170976 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoASSERT in SharedBuffer::maybeAppendDataArray() on MobileSafari launch
psolanki@apple.com [Thu, 10 Jul 2014 20:48:44 +0000 (20:48 +0000)]
ASSERT in SharedBuffer::maybeAppendDataArray() on MobileSafari launch
https://bugs.webkit.org/show_bug.cgi?id=134812
<rdar://problem/17628434>

Reviewed by Joseph Pecoraro.

Fix bug in my fix in r170930. Initialize the badly named m_shouldUsePurgeableMemory field to
false. This field indicates when it is okay to use purgeable memory and is set to true once
the resource is finished loading. By setting it to true in the constructor we were creating
purgeable memory while the resource was still being loaded and this triggered the assert.

No new tests. Should be covered by existing tests.

* platform/cf/SharedBufferCF.cpp:
(WebCore::SharedBuffer::SharedBuffer):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170975 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoStore ViewSnapshots directly on the WebBackForwardListItem
timothy_horton@apple.com [Thu, 10 Jul 2014 20:33:05 +0000 (20:33 +0000)]
Store ViewSnapshots directly on the WebBackForwardListItem
https://bugs.webkit.org/show_bug.cgi?id=134667
<rdar://problem/17082639>

Reviewed by Dan Bernstein.

Make ViewSnapshot a refcounted class. Store it directly on the back-forward item
instead of in a side map referenced by UUID. Switch to a very simple LRU eviction model for now.
This fixes a ton of snapshot management bugs; for example, we would start throwing out snapshots
in the page that was actively being interacted with *first* when evicting snapshots, instead of
preferring older snapshots. Additionally, we would not throw away snapshots when back forward items
became unreachable.

There is definitely room for improvement of the eviction mechanism, but this is closer to a time-tested implementation.

* Shared/SessionState.h:
Keep a ViewSnapshot instead of a UUID on the BackForwardListItemState.

* Shared/WebBackForwardListItem.h:
Fix some indented namespace contents.

(WebKit::WebBackForwardListItem::snapshot):
(WebKit::WebBackForwardListItem::setSnapshot):
(WebKit::WebBackForwardListItem::setSnapshotUUID): Deleted.
(WebKit::WebBackForwardListItem::snapshotUUID): Deleted.
Switch the snapshot getter/setter to operate on ViewSnapshots instead of UUIDs.

* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _takeViewSnapshot]):
* UIProcess/API/Cocoa/WKWebViewInternal.h:
* UIProcess/API/mac/WKView.mm:
(-[WKView _takeViewSnapshot]):
* UIProcess/API/mac/WKViewInternal.h:
* UIProcess/PageClient.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::takeViewSnapshot):
* UIProcess/WebPageProxy.h:
* UIProcess/ios/PageClientImplIOS.h:
* UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::takeViewSnapshot):
* UIProcess/mac/PageClientImpl.h:
* UIProcess/mac/PageClientImpl.mm:
(WebKit::PageClientImpl::takeViewSnapshot):
Adopt ViewSnapshot::create, return a PassRefPtr, and class-ify ViewSnapshot.

* UIProcess/ios/ViewGestureControllerIOS.mm:
(WebKit::ViewGestureController::beginSwipeGesture):
(WebKit::ViewGestureController::endSwipeGesture):
* UIProcess/mac/ViewGestureController.h:
* UIProcess/mac/ViewGestureControllerMac.mm:
(WebKit::ViewGestureController::shouldUseSnapshotForSize):
(WebKit::ViewGestureController::beginSwipeGesture):
(WebKit::ViewGestureController::endSwipeGesture):
Grab the ViewSnapshot directly from the WebBackForwardListItem, and adopt the new functions.

* UIProcess/ios/WebMemoryPressureHandlerIOS.mm:
(WebKit::WebMemoryPressureHandler::WebMemoryPressureHandler):
Rename discardSnapshots to discardSnapshotImages, because we're really only discarding
the images; the render tree size/background color "snapshot" remains and is useful.

* UIProcess/mac/ViewSnapshotStore.h:
(WebKit::ViewSnapshot::setRenderTreeSize):
(WebKit::ViewSnapshot::renderTreeSize):
(WebKit::ViewSnapshot::setBackgroundColor):
(WebKit::ViewSnapshot::backgroundColor):
(WebKit::ViewSnapshot::setDeviceScaleFactor):
(WebKit::ViewSnapshot::deviceScaleFactor):
(WebKit::ViewSnapshot::imageSizeInBytes):
(WebKit::ViewSnapshot::surface):
(WebKit::ViewSnapshot::size):
(WebKit::ViewSnapshot::creationTime):
Make ViewSnapshot a refcounted class.
Add create functions which take an image (or slot ID), and relevant sizes.
It is expected that a ViewSnapshot is created with an image, and it is only possible
to remove that image, never to replace it. A new ViewSnapshot is required in that case.
Add setters for things that ViewSnapshotStore sets on the snapshot after the PageClient
retrieves it from the view. Add getters for things that the ViewGestureControllers need.

Remove removeSnapshotImage, getSnapshot, and the snapshot map.

* UIProcess/mac/ViewSnapshotStore.mm:
(WebKit::ViewSnapshotStore::~ViewSnapshotStore):
(WebKit::ViewSnapshotStore::didAddImageToSnapshot):
(WebKit::ViewSnapshotStore::willRemoveImageFromSnapshot):
Manage m_snapshotCacheSize and m_snapshotsWithImages via didAddImageToSnapshot and willRemoveImageFromSnapshot.
willRemoveImageFromSnapshot will -always- be called before the ViewSnapshot is destroyed.

(WebKit::ViewSnapshotStore::pruneSnapshots):
Switch to a simple LRU eviction model. As previously mentioned, it's possible to do better, but
this is much less broken than the previous implementation.

(WebKit::ViewSnapshotStore::recordSnapshot):
(WebKit::ViewSnapshotStore::discardSnapshotImages):
(WebKit::ViewSnapshot::create):
(WebKit::ViewSnapshot::ViewSnapshot):
(WebKit::ViewSnapshot::~ViewSnapshot):
(WebKit::ViewSnapshot::hasImage):
(WebKit::ViewSnapshot::clearImage):
(WebKit::ViewSnapshot::asLayerContents):
If a surface is Empty when it comes back from being volatile, throw away the surface
and notify the Store to remove it from m_snapshotCacheSize (via clearImage()).

(WebKit::ViewSnapshotStore::removeSnapshotImage): Deleted.
(WebKit::ViewSnapshotStore::getSnapshot): Deleted.
(WebKit::ViewSnapshotStore::discardSnapshots): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170974 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[iOS WebKit2] Some memory pressure relief tweaks.
akling@apple.com [Thu, 10 Jul 2014 20:24:12 +0000 (20:24 +0000)]
[iOS WebKit2] Some memory pressure relief tweaks.
<https://webkit.org/b/134811>

Split memory pressure relief into critical and non-critical sections.
Non-critical relief is for clearing out things that are really not
essential, e.g unused font data, text measurement caches, etc.

On iOS, only flip the "WebKit is under memory pressure" flag when we
are under *critical* memroy pressure, rather than doing it early on
and gimping ourselves because other processes are too big.

Also added logging for when we transition in/out of system pressure.

Reviewed by Geoffrey Garen.

* platform/MemoryPressureHandler.cpp:
(WebCore::MemoryPressureHandler::releaseNoncriticalMemory):
(WebCore::MemoryPressureHandler::releaseCriticalMemory):
(WebCore::MemoryPressureHandler::releaseMemory):
* platform/MemoryPressureHandler.h:
(WebCore::MemoryPressureHandler::ReliefLogger::loggingEnabled):
* platform/cocoa/MemoryPressureHandlerCocoa.mm:
(WebCore::MemoryPressureHandler::install):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170973 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoNeed Setting/WKPreference that allows clients to prevent scrollbars from drawing
bdakin@apple.com [Thu, 10 Jul 2014 20:04:42 +0000 (20:04 +0000)]
Need Setting/WKPreference that allows clients to prevent scrollbars from drawing
on a secondary thread
https://bugs.webkit.org/show_bug.cgi?id=134778
-and corresponding-
<rdar://problem/17595333>

Reviewed by Tim Horton.

Source/WebCore:
This is a requirement for some types of performance tests. The patch adds a new
virtual function to ScrollableArea that forces subclasses to indicate the value of
the Setting. This is required because Scrollbar and ScrollableArea can’t get to
Settings on their own.

* page/FrameView.cpp:
(WebCore::FrameView::forceUpdateScrollbarsOnMainThreadForPerformanceTesting):
* page/FrameView.h:
* page/Settings.in:
* platform/ScrollableArea.h:
* platform/Scrollbar.cpp:
(WebCore::Scrollbar::supportsUpdateOnSecondaryThread):
* platform/win/PopupMenuWin.h:
* rendering/RenderLayer.cpp:
(WebCore::RenderLayer::forceUpdateScrollbarsOnMainThreadForPerformanceTesting):
* rendering/RenderLayer.h:
* rendering/RenderListBox.cpp:
(WebCore::RenderListBox::forceUpdateScrollbarsOnMainThreadForPerformanceTesting):
* rendering/RenderListBox.h:

Source/WebKit2:
This is a requirement for some types of performance tests.

New pref.
* Shared/WebPreferencesDefinitions.h:

PDFPlugin has to implement this new ScrollableArea virtual function to indicate
the Setting’s value.
* WebProcess/Plugins/PDF/PDFPlugin.h:
* WebProcess/Plugins/PDF/PDFPlugin.mm:
(WebKit::PDFPlugin::forceUpdateScrollbarsOnMainThreadForPerformanceTesting):

New pref.
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::updatePreferences):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170972 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[iOS] Frequent assertion failures when swiping back
timothy_horton@apple.com [Thu, 10 Jul 2014 19:31:54 +0000 (19:31 +0000)]
[iOS] Frequent assertion failures when swiping back

Reviewed by Dan Bernstein.

* UIProcess/mac/RemoteLayerTreeDrawingAreaProxy.mm:
(WebKit::RemoteLayerTreeDrawingAreaProxy::dispatchAfterEnsuringDrawing):
Don't create an unused VoidCallback. It will assert when destroyed without being called.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170971 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoPass sandbox extension for GL cache over to webprocess
oliver@apple.com [Thu, 10 Jul 2014 18:33:39 +0000 (18:33 +0000)]
Pass sandbox extension for GL cache over to webprocess
https://bugs.webkit.org/show_bug.cgi?id=134806

Reviewed by Anders Carlsson.

Add additional WebProcess parameters to pass an extension
that allows access to the opengl cache directory in the
host application's container.

* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
* UIProcess/WebContext.cpp:
(WebKit::WebContext::createNewWebProcess):
(WebKit::WebContext::openGLCacheDirectory):
* UIProcess/WebContext.h:
* UIProcess/mac/WebContextMac.mm:
(WebKit::WebContext::platformDefaultOpenGLCacheDirectory):
* WebProcess/cocoa/WebProcessCocoa.mm:
(WebKit::WebProcess::platformInitializeWebProcess):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170970 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoFix my mis-merged ChangeLog from r170966
beidson@apple.com [Thu, 10 Jul 2014 18:24:23 +0000 (18:24 +0000)]
Fix my mis-merged ChangeLog from r170966

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170969 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoEnable same features on all Windows ports.
commit-queue@webkit.org [Thu, 10 Jul 2014 17:52:19 +0000 (17:52 +0000)]
Enable same features on all Windows ports.
https://bugs.webkit.org/show_bug.cgi?id=134715

Patch by Alex Christensen <achristensen@webkit.org> on 2014-07-10
Reviewed by Brent Fulgham.

Source/WTF:
* wtf/Platform.h:
Use REQUEST_ANIMATION_FRAME_TIMER on WinCairo.

WebKitLibraries:
* win/tools/vsprops/FeatureDefinesCairo.props:
Made identical to FeatureDefines.props.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170968 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoiOS build fix.
mitz@apple.com [Thu, 10 Jul 2014 17:07:41 +0000 (17:07 +0000)]
iOS build fix.

* WebProcess/WebPage/ServicesOverlayController.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170967 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoPhone number highlights should always be visible if the mouse hovers over.
beidson@apple.com [Thu, 10 Jul 2014 16:19:56 +0000 (16:19 +0000)]
Phone number highlights should always be visible if the mouse hovers over.
<rdar://problem/17527476> and https://bugs.webkit.org/show_bug.cgi?id=134784

Reviewed by Tim Horton.

Source/WebCore:
* WebCore.exp.in:

Source/WebKit2:
This is a fairly extensive rewrite of ServicesOverlayController.
It allows one selection highlight for the entire selection, and as many telephone number highlights as there are numbers.
If a telephone number highlight is hovered over, it wins and is painted.
If no telephone number highlight is hovered but the selection highlight is, then it is painted.

The purposes of each method are self evident by their name, and the concepts are mostly the same as they used to be.

The exception is establishHoveredTelephoneHighlight which gets a more detailed explanation below.

* Platform/Logging.h: Add a Services logging channel.

* WebProcess/WebPage/ServicesOverlayController.h:
(WebKit::TelephoneNumberData::TelephoneNumberData):

* WebProcess/WebPage/mac/ServicesOverlayController.mm:
(WebKit::ServicesOverlayController::ServicesOverlayController):
(WebKit::ServicesOverlayController::selectionRectsDidChange):
(WebKit::ServicesOverlayController::selectedTelephoneNumberRangesChanged):
(WebKit::ServicesOverlayController::clearHighlightState):
(WebKit::ServicesOverlayController::drawRect):
(WebKit::ServicesOverlayController::drawSelectionHighlight):
(WebKit::ServicesOverlayController::maybeDrawTelephoneNumberHighlight):
(WebKit::ServicesOverlayController::drawHighlight):
(WebKit::ServicesOverlayController::clearSelectionHighlight):
(WebKit::ServicesOverlayController::clearHoveredTelephoneNumberHighlight):
(WebKit::ServicesOverlayController::establishHoveredTelephoneHighlight): Starts walking the telephone number ranges and
    creating a highlight for each one that doesn’t already have a highlight. If that highlight is also being hovered by
    the mouse, then it is set as the hovered telephone number highlight and the method stops creating new highlights.
(WebKit::ServicesOverlayController::maybeCreateSelectionHighlight):
(WebKit::ServicesOverlayController::mouseEvent):
(WebKit::ServicesOverlayController::handleClick):
(WebKit::ServicesOverlayController::drawTelephoneNumberHighlight): Deleted.
(WebKit::ServicesOverlayController::drawCurrentHighlight): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170966 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[GTK][CMake] Add a 'distcheck' target
carlosgc@webkit.org [Thu, 10 Jul 2014 16:13:54 +0000 (16:13 +0000)]
[GTK][CMake] Add a 'distcheck' target
https://bugs.webkit.org/show_bug.cgi?id=130675

.:
Reviewed by Gustavo Noronha Silva.

* Source/PlatformGTK.cmake: Add distcheck target.

Tools:
Patch by Carlos Garcia Campos  <cgarcia@igalia.com> and Martin Robinson <mrobinson@igalia.com> on 2014-07-10
Reviewed by Gustavo Noronha Silva.

Add --check option to make-dist script that builds and installs the tarball.

* gtk/make-dist.py:
(Distcheck.__init__):
(Distcheck.extract_tarball): Extratc the tarball.
(Distcheck.configure): Run cmake.
(Distcheck.configure.create_dir): Helper function to create a directory.
(Distcheck.build): Run make.
(Distcheck.install): Run make install.
(Distcheck.clean): Clean up the extracted tarball.
(Distcheck.check): Run all previous methods in order.
* gtk/manifest.txt: Add missing file.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170965 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[GTK] Use the same default options for production builds that previous stable releases
carlosgc@webkit.org [Thu, 10 Jul 2014 16:11:11 +0000 (16:11 +0000)]
[GTK] Use the same default options for production builds that previous stable releases
https://bugs.webkit.org/show_bug.cgi?id=134589

Reviewed by Martin Robinson.

.:
Change the default value of some features to match our stable releases.
Add FindCairoGL to find cairo-gl libraries and make accelearetd 2D
canvas depend on whether cairo-gl is found.

* Source/cmake/FindCairoGL.cmake: Added.
* Source/cmake/OptionsGTK.cmake:

Tools:
Make sure all features enabled in production builds are also
enabled in developer mode.

* Scripts/webkitperl/FeatureList.pm: Enable ACCELERATED_2D_CANVAS
and CSS_FILTERS for the GTK port.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170964 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoAssertions or crashes under _takeViewSnapshot when restoring windows
timothy_horton@apple.com [Thu, 10 Jul 2014 15:39:49 +0000 (15:39 +0000)]
Assertions or crashes under _takeViewSnapshot when restoring windows
https://bugs.webkit.org/show_bug.cgi?id=134792

Reviewed by Simon Fraser.

* UIProcess/API/mac/WKView.mm:
(-[WKView _takeViewSnapshot]):
Taking a window-server snapshot of a non-visible window tends to not succeed.

* platform/graphics/cocoa/IOSurface.mm:
(IOSurface::createFromImage):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170963 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[EFL] Unreviewed EFL gardening.
jinwoo7.song@samsung.com [Thu, 10 Jul 2014 12:23:59 +0000 (12:23 +0000)]
[EFL] Unreviewed EFL gardening.

Rebaseline tests after r170304 and r170619.

* platform/efl/fast/multicol/pagination/RightToLeft-rl-dynamic-expected.png: Added after r170304.
* platform/efl/fast/multicol/pagination/RightToLeft-rl-dynamic-expected.txt: Added after r170304.
* platform/efl/fast/text/indic-expected.png: Added after r170619.
* platform/efl/fast/text/indic-expected.txt: Added after r170619.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170962 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[EFL] Unreviewed EFL gardening.
jinwoo7.song@samsung.com [Thu, 10 Jul 2014 12:14:00 +0000 (12:14 +0000)]
[EFL] Unreviewed EFL gardening.

Rebaseline tests after r170207.

* platform/efl/fast/text/khmer-lao-font-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170961 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[XHR] overrideMimeType() should be able to change encoding in HEADERS RECEIVED state
commit-queue@webkit.org [Thu, 10 Jul 2014 10:09:41 +0000 (10:09 +0000)]
[XHR] overrideMimeType() should be able to change encoding in HEADERS RECEIVED state
https://bugs.webkit.org/show_bug.cgi?id=128968

Patch by Youenn Fablet <youenn.fablet@crf.canon.fr> on 2014-07-10
Reviewed by Alexey Proskuryakov.

Source/WebCore:
Moved response encoding computation from didReceiveResponse to didReceiveData, just before the decoder is instantiated.
This allows overrideMimeType to be changed within readystatechange event callback and have an impact on selected encoding.

Test: http/tests/xmlhttprequest/overridemimetype-headers-received-state-force-shiftjis.html

* xml/XMLHttpRequest.cpp:
(WebCore::XMLHttpRequest::didReceiveResponse): Removed m_responseEncoding computation from response and mimeTypeOverride.
(WebCore::XMLHttpRequest::didReceiveData): Added m_responseEncoding computation from response and mimeTypeOverride.

LayoutTests:
Added test equivalent to http://w3c-test.org/web-platform-tests/master/XMLHttpRequest/overridemimetype-headers-received-state-force-shiftjis.htm
The test sets overrideMimeType when xhr state is changed to HEADERS RECEIVED.

* http/tests/xmlhttprequest/overridemimetype-headers-received-state-force-shiftjis-expected.txt: Added.
* http/tests/xmlhttprequest/overridemimetype-headers-received-state-force-shiftjis.html: Added.
* http/tests/xmlhttprequest/resources/get-content.php: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170960 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoBuffer CSS and JS resources in network process before sending over to web process
psolanki@apple.com [Thu, 10 Jul 2014 06:14:35 +0000 (06:14 +0000)]
Buffer CSS and JS resources in network process before sending over to web process
https://bugs.webkit.org/show_bug.cgi?id=134560
<rdar://problem/16737186>

Reviewed by Antti Koivisto.

For CSS and JS resources, ask the network process to buffer the entire resource instead of
sending it to web process in chunks since the web process can't do anything with a partial
css or js file.

* NetworkProcess/NetworkResourceLoader.cpp:
(WebKit::NetworkResourceLoader::NetworkResourceLoader):
* Shared/Network/NetworkResourceLoadParameters.cpp:
(WebKit::NetworkResourceLoadParameters::NetworkResourceLoadParameters):
(WebKit::NetworkResourceLoadParameters::encode):
(WebKit::NetworkResourceLoadParameters::decode):
* Shared/Network/NetworkResourceLoadParameters.h:
* WebProcess/Network/WebResourceLoadScheduler.cpp:
(WebKit::WebResourceLoadScheduler::scheduleLoad):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170958 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoFire connected/disconnected events for Gamepads.
beidson@apple.com [Thu, 10 Jul 2014 06:08:58 +0000 (06:08 +0000)]
Fire connected/disconnected events for Gamepads.
https://bugs.webkit.org/show_bug.cgi?id=134386

Reviewed by Dean Jackson.

No new tests (No effect in a currently tested config)

* Modules/gamepad/GamepadManager.cpp:
(WebCore::navigatorGamepadFromDOMWindow): Handling converting a possibly-null Navigator into
    a possibly null NavigatorGamepad.
(WebCore::GamepadManager::platformGamepadConnected): Notify blind Navigator/DOMWindows of all
    previously attached Gamepads, then notify everybody of this new gamepad.
(WebCore::GamepadManager::platformGamepadDisconnected): Handle dispatching the disconnected
    event to all registered DOMWindows.
(WebCore::GamepadManager::platformGamepadInputActivity): Notify blind Navigator/DOMWindows of all
    attached Gamepads.
(WebCore::GamepadManager::makeGamepadVisible): Handles notifying setting up a new gamepads
    with all NavigatorGamepads as well as dispatching the connected even to DOMWindows.
(WebCore::GamepadManager::registerDOMWindow):
(WebCore::GamepadManager::unregisterDOMWindow):
(WebCore::GamepadManager::makeGamepadsVisibileToBlindNavigators): Deleted.
* Modules/gamepad/GamepadManager.h:

* Modules/gamepad/NavigatorGamepad.cpp:
(WebCore::NavigatorGamepad::gamepadAtIndex):
* Modules/gamepad/NavigatorGamepad.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170957 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Win] Unreviewed build fix.
bfulgham@apple.com [Thu, 10 Jul 2014 05:32:40 +0000 (05:32 +0000)]
[Win] Unreviewed build fix.

Build bots don't have a three-part default version string,
breaking pattern matching. Handle this case.

* win/tools/scripts/auto-version.pl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170954 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[iOS][WK2] Disable text quantization while actively changing the page's scale factor
benjamin@webkit.org [Thu, 10 Jul 2014 05:06:30 +0000 (05:06 +0000)]
[iOS][WK2] Disable text quantization while actively changing the page's scale factor
https://bugs.webkit.org/show_bug.cgi?id=134781

Patch by Benjamin Poulain <bpoulain@apple.com> on 2014-07-09
Reviewed by Tim Horton and Myles C. Maxfield.

Source/WebCore:
Query the chrome client to setup quantization on each layers.

* page/ChromeClient.h:
(WebCore::ChromeClient::hasStablePageScaleFactor):
* platform/graphics/mac/FontMac.mm:
(WebCore::Font::drawGlyphs):
* rendering/RenderLayer.cpp:
(WebCore::RenderLayer::setupFontSubpixelQuantization):

Source/WebKit2:
While zooming a page, text quantization causes glyphs to "move" in order to get to the closest
boundary for the current scale factor.

We do not want this to happen while dynamically changing the scale factor because the effect
is visible. To avoid this, we disable text quantization if the page's scale factor changes
in response to a non-stable contentRect update.

* WebProcess/WebCoreSupport/WebChromeClient.h:
* WebProcess/WebCoreSupport/ios/WebChromeClientIOS.mm:
(WebKit::WebChromeClient::hasStablePageScaleFactor):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::WebPage):
* WebProcess/WebPage/WebPage.h:
(WebKit::WebPage::hasStablePageScaleFactor):
* WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::updateVisibleContentRects):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170953 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[iOS] Use UIAlertController API in WKFileUploadPanel instead of SPI
commit-queue@webkit.org [Thu, 10 Jul 2014 04:38:13 +0000 (04:38 +0000)]
[iOS] Use UIAlertController API in WKFileUploadPanel instead of SPI
https://bugs.webkit.org/show_bug.cgi?id=134777

Patch by Joseph Pecoraro <pecoraro@apple.com> on 2014-07-09
Reviewed by Sam Weinig.

* UIProcess/ios/forms/WKFileUploadPanel.mm:
(-[WKFileUploadPanel _showMediaSourceSelectionSheet]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170952 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[EFL][WK2] Add new Public API in ewk_download_job.h to get size of the data already...
commit-queue@webkit.org [Thu, 10 Jul 2014 04:13:37 +0000 (04:13 +0000)]
[EFL][WK2] Add new Public API in ewk_download_job.h to get size of the data already downloaded.
https://bugs.webkit.org/show_bug.cgi?id=134759

Patch by Shivakumar JM <shiva.jm@samsung.com> on 2014-07-09
Reviewed by Gyuyoung Kim.

Add new API in ewk_download_job.h to get size of the data already downloaded.

* UIProcess/API/efl/ewk_download_job.cpp:
(ewk_download_job_received_data_length_get):
(EwkDownloadJob::receivedData):
* UIProcess/API/efl/ewk_download_job.h:
* UIProcess/API/efl/ewk_download_job_private.h:
* UIProcess/API/efl/tests/test_ewk2_download_job.cpp:
(EWK2DownloadJobTest::on_download_requested):
(EWK2DownloadJobTest::on_download_finished):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170951 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUse 16bits instructions for push/pop on ARMv7 when possible
benjamin@webkit.org [Thu, 10 Jul 2014 04:11:34 +0000 (04:11 +0000)]
Use 16bits instructions for push/pop on ARMv7 when possible
https://bugs.webkit.org/show_bug.cgi?id=134753

Reviewed by Geoffrey Garen.

The patch r170839 mixed the code for push/pop pair and single push/pop.
That part was reverted in r170909.

This patch puts the code back but specialized for single push/pop.

* assembler/ARMv7Assembler.h:
(JSC::ARMv7Assembler::pop):
(JSC::ARMv7Assembler::push):
* assembler/MacroAssemblerARMv7.h:
(JSC::MacroAssemblerARMv7::pop):
(JSC::MacroAssemblerARMv7::push):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170950 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Curl] Cache entry invalidated too early.
commit-queue@webkit.org [Thu, 10 Jul 2014 03:11:26 +0000 (03:11 +0000)]
[Curl] Cache entry invalidated too early.
https://bugs.webkit.org/show_bug.cgi?id=134681

Patch by peavo@outlook.com <peavo@outlook.com> on 2014-07-09
Reviewed by Alex Christensen.

When a cache entry has expired, it is invalidated when a request for the corresponding url is started.
This is too early, since the resource is possibly not modified (even though it has expired),
and the server might respond with a 304 (not modified) response.
When we then receive a 304 response, the cache entry is deleted, and we have no cached response/data to provide.
This can be solved by not invalidating the entry when a request for that url is started,
but instead invalidate when a 200 OK response is received (which means the resource has been modified).

* platform/network/curl/CurlCacheManager.cpp:
(WebCore::CurlCacheManager::isCached): Avoid invalidating cache entry when it has expired, wait until we know the server response.
* platform/network/curl/CurlCacheManager.h: Added const modifier.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170949 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION(r170858): Safari freezes upon making a search on a website (yelp.com).
enrica@apple.com [Thu, 10 Jul 2014 02:18:30 +0000 (02:18 +0000)]
REGRESSION(r170858): Safari freezes upon making a search on a website (yelp.com).
https://bugs.webkit.org/show_bug.cgi?id=134791
<rdar://problem/17616971>

Reviewed by Benjamin Poulain.

After r170858 we notify the keyboard too often about
the changed selection. This patch removes the notification
until we find a better way to do it that doesn't cause
deadlocks.

* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _selectionWillChange]):
(-[WKContentView _selectionChanged]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170948 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoRevert r170413 and r170390
mmaxfield@apple.com [Thu, 10 Jul 2014 01:42:43 +0000 (01:42 +0000)]
Revert r170413 and r170390
https://bugs.webkit.org/show_bug.cgi?id=134741

Reviewed by Benjamin Poulain.

These patches caused a variety of failures. I'm going to break them up into
smaller chunks and commit them separately at some point in the future.

Requesting a review from Ben Poulain because there were some conflicts with
r170561 in InlineTextBox.cpp and InlineTextBox.h.

* WebCore.exp.in:
* platform/graphics/Font.cpp:
(WebCore::Font::drawText):
(WebCore::Font::drawEmphasisMarks):
(WebCore::Font::adjustSelectionRectForText):
(WebCore::computeUnderlineType):
* platform/graphics/Font.h:
* platform/graphics/FontFastPath.cpp:
(WebCore::Font::getGlyphsAndAdvancesForSimpleText):
(WebCore::Font::drawSimpleText):
(WebCore::Font::drawEmphasisMarksForSimpleText):
(WebCore::Font::drawGlyphBuffer):
(WebCore::Font::drawEmphasisMarks):
(WebCore::Font::adjustSelectionRectForSimpleText):
* platform/graphics/GlyphBuffer.h:
(WebCore::GlyphBuffer::size):
(WebCore::GlyphBuffer::glyphs):
(WebCore::GlyphBuffer::advances):
(WebCore::GlyphBuffer::fontDataAt):
(WebCore::GlyphBuffer::advanceAt):
(WebCore::GlyphBuffer::offsetAt):
(WebCore::GlyphBuffer::reverse):
(WebCore::GlyphBuffer::offsetInString):
(WebCore::GlyphBuffer::swap):
* platform/graphics/GraphicsContext.cpp:
* platform/graphics/Latin1TextIterator.h:
(WebCore::Latin1TextIterator::Latin1TextIterator):
(WebCore::Latin1TextIterator::currentCharacter):
* platform/graphics/SimpleFontData.h:
* platform/graphics/SurrogatePairAwareTextIterator.cpp:
(WebCore::SurrogatePairAwareTextIterator::SurrogatePairAwareTextIterator):
* platform/graphics/SurrogatePairAwareTextIterator.h:
(WebCore::SurrogatePairAwareTextIterator::currentCharacter):
* platform/graphics/TextRun.h:
(WebCore::TextRun::length):
(WebCore::TextRun::charactersLength):
* platform/graphics/WidthIterator.cpp:
(WebCore::WidthIterator::glyphDataForCharacter):
(WebCore::applyFontTransforms):
(WebCore::WidthIterator::advanceInternal):
(WebCore::WidthIterator::advance):
(WebCore::WidthIterator::advanceOneCharacter):
* platform/graphics/WidthIterator.h:
* platform/graphics/cairo/FontCairo.cpp:
(WebCore::Font::drawGlyphs):
* platform/graphics/cairo/FontCairoHarfbuzzNG.cpp:
(WebCore::Font::drawComplexText):
(WebCore::Font::drawEmphasisMarksForComplexText):
(WebCore::Font::adjustSelectionRectForComplexText):
* platform/graphics/harfbuzz/HarfBuzzShaper.cpp:
(WebCore::HarfBuzzShaper::setNormalizedBuffer):
* platform/graphics/mac/FontComplexTextMac.cpp:
(WebCore::Font::adjustSelectionRectForComplexText):
(WebCore::Font::getGlyphsAndAdvancesForComplexText):
(WebCore::Font::drawComplexText):
(WebCore::Font::drawEmphasisMarksForComplexText):
* platform/graphics/mac/FontMac.mm:
(WebCore::Font::drawGlyphs):
* platform/graphics/win/FontCGWin.cpp:
(WebCore::Font::drawGlyphs):
* platform/graphics/win/FontWin.cpp:
(WebCore::Font::adjustSelectionRectForComplexText):
(WebCore::Font::getGlyphsAndAdvancesForComplexText):
(WebCore::Font::drawComplexText):
(WebCore::Font::drawEmphasisMarksForComplexText):
* platform/graphics/wince/FontWinCE.cpp:
(WebCore::Font::drawGlyphs):
(WebCore::Font::drawComplexText):
(WebCore::Font::drawEmphasisMarksForComplexText):
(WebCore::Font::selectionRectForComplexText):
* rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::isSelected):
(WebCore::InlineTextBox::selectionState):
(WebCore::InlineTextBox::localSelectionRect):
(WebCore::InlineTextBox::paint):
(WebCore::InlineTextBox::selectionStartEnd):
(WebCore::InlineTextBox::paintSelection):
(WebCore::InlineTextBox::paintCompositionBackground):
(WebCore::InlineTextBox::paintDocumentMarker):
(WebCore::InlineTextBox::paintTextMatchMarker):
(WebCore::InlineTextBox::computeRectForReplacementMarker):
(WebCore::InlineTextBox::paintCompositionUnderline):
(WebCore::InlineTextBox::positionForOffset):
(WebCore::InlineTextBox::constructTextRun):
* rendering/InlineTextBox.h:
(WebCore::InlineTextBox::truncation):
* rendering/RenderCombineText.cpp:
(WebCore::RenderCombineText::getStringToRender):
* rendering/RenderCombineText.h:
* rendering/RenderTextLineBoxes.cpp:
(WebCore::ellipsisRectForBox):
* rendering/svg/SVGInlineFlowBox.cpp:
(WebCore::SVGInlineFlowBox::computeTextMatchMarkerRectForRenderer):
* rendering/svg/SVGInlineTextBox.cpp:
(WebCore::SVGInlineTextBox::positionForOffset):
(WebCore::SVGInlineTextBox::localSelectionRect):
(WebCore::SVGInlineTextBox::paintSelectionBackground):
(WebCore::SVGInlineTextBox::mapStartEndPositionsIntoFragmentCoordinates):
(WebCore::SVGInlineTextBox::paintText):
* rendering/svg/SVGInlineTextBox.h:
* rendering/svg/SVGTextMetricsBuilder.cpp:
(WebCore::SVGTextMetricsBuilder::currentCharacterStartsSurrogatePair):
(WebCore::SVGTextMetricsBuilder::advance):
* rendering/svg/SVGTextQuery.cpp:
(WebCore::SVGTextQuery::mapStartEndPositionsIntoFragmentCoordinates):
(WebCore::SVGTextQuery::modifyStartEndPositionsRespectingLigatures):
(WebCore::SVGTextQuery::subStringLengthCallback):
(WebCore::SVGTextQuery::startPositionOfCharacterCallback):
(WebCore::SVGTextQuery::endPositionOfCharacterCallback):
(WebCore::SVGTextQuery::rotationOfCharacterCallback):
(WebCore::SVGTextQuery::extentOfCharacterCallback):
(WebCore::SVGTextQuery::characterNumberAtPositionCallback):
* rendering/svg/SVGTextQuery.h:
* rendering/svg/SVGTextRunRenderingContext.cpp:
(WebCore::SVGGlyphToPathTranslator::SVGGlyphToPathTranslator):
(WebCore::SVGTextRunRenderingContext::glyphDataForCharacter):
* rendering/svg/SVGTextRunRenderingContext.h:
* svg/SVGFontData.cpp:
(WebCore::SVGFontData::applySVGGlyphSelection):
* svg/SVGFontData.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170947 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoFix flaky loading tests
jpfau@apple.com [Thu, 10 Jul 2014 01:13:45 +0000 (01:13 +0000)]
Fix flaky loading tests
https://bugs.webkit.org/show_bug.cgi?id=133696

Reviewed by Ryosuke Niwa.

Source/WebCore:
In some cases, the Frame may be destroyed while the loader is
initializing. Make sure we don't delete it too early.

* loader/FrameLoader.cpp:
(WebCore::FrameLoader::init):

LayoutTests:
* platform/wk2/TestExpectations: Unskip no-longer failing tests

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170946 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoSafari showing blank pages
andersca@apple.com [Thu, 10 Jul 2014 01:05:23 +0000 (01:05 +0000)]
Safari showing blank pages
https://bugs.webkit.org/show_bug.cgi?id=134790
<rdar://problem/17617166>

Reviewed by Simon Fraser.

* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView initWithFrame:configuration:]):
Make sure to add the content view to the scroll view.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170945 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoMake IndirectCompositingReason an enum class
simon.fraser@apple.com [Thu, 10 Jul 2014 01:01:52 +0000 (01:01 +0000)]
Make IndirectCompositingReason an enum class
https://bugs.webkit.org/show_bug.cgi?id=134789

Reviewed by Tim Horton.

Make RenderLayer::IndirectCompositingReason an enum class, and change some code
in RenderLayerCompositor::reasonsForCompositing() to use a switch.

* rendering/RenderLayer.cpp:
(WebCore::RenderLayer::RenderLayer):
* rendering/RenderLayer.h:
* rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::computeCompositingRequirements):
(WebCore::RenderLayerCompositor::requiresOwnBackingStore):
(WebCore::RenderLayerCompositor::reasonsForCompositing):
(WebCore::RenderLayerCompositor::requiresCompositingForIndirectReason):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170944 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoClosed web views should never create new web processes
andersca@apple.com [Thu, 10 Jul 2014 00:50:25 +0000 (00:50 +0000)]
Closed web views should never create new web processes
https://bugs.webkit.org/show_bug.cgi?id=134787
<rdar://problem/16892526>

Reviewed by Simon Fraser.

* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView loadHTMLString:baseURL:]):
(-[WKWebView reload]):
(-[WKWebView reloadFromOrigin]):
Return nil if the returned navigation ID is 0.

* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::reattachToWebProcess):
Assert that the page is not closed.

(WebKit::WebPageProxy::reattachToWebProcessWithItem):
(WebKit::WebPageProxy::loadRequest):
(WebKit::WebPageProxy::loadFile):
(WebKit::WebPageProxy::loadData):
(WebKit::WebPageProxy::loadHTMLString):
(WebKit::WebPageProxy::loadAlternateHTMLString):
(WebKit::WebPageProxy::loadPlainTextString):
(WebKit::WebPageProxy::loadWebArchiveData):
Add early returns if the page is closed.

* UIProcess/WebPageProxy.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170943 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[EFL] gardening after r170868
ryuan.choi@samsung.com [Thu, 10 Jul 2014 00:29:10 +0000 (00:29 +0000)]
[EFL] gardening after r170868
https://bugs.webkit.org/show_bug.cgi?id=134717

Unreviewed, EFL gardening.

Four tests have been rebaselined, and disable the gamepad tests

Patch by Hunseop Jeong <hs85.jeong@samsung.com> on 2014-07-09

* platform/efl/TestExpectations: Disable the gamepad tests
* platform/efl/fast/text/atsui-multiple-renderers-expected.png:
* platform/efl/fast/text/atsui-multiple-renderers-expected.txt:
* platform/efl/fast/text/complex-text-opacity-expected.png:
* platform/efl/fast/text/complex-text-opacity-expected.txt:
* platform/efl/fast/text/midword-break-before-surrogate-pair-expected.png:
* platform/efl/fast/text/midword-break-before-surrogate-pair-expected.txt:
* platform/efl/transforms/2d/hindi-rotated-expected.png:
* platform/efl/transforms/2d/hindi-rotated-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170942 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[MSE] http/tests/media/media-source/mediasource-closed.html is failing.
jer.noble@apple.com [Thu, 10 Jul 2014 00:25:52 +0000 (00:25 +0000)]
[MSE] http/tests/media/media-source/mediasource-closed.html is failing.
https://bugs.webkit.org/show_bug.cgi?id=134775

Reviewed by Eric Carlson.

Update the test to use isNaN rather than Number.isNaN.

* http/tests/media/media-source/mediasource-closed.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170941 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Win] Unreviewed gardening.
bfulgham@apple.com [Thu, 10 Jul 2014 00:24:43 +0000 (00:24 +0000)]
[Win] Unreviewed gardening.

* win/tools/scripts/auto-version.pl: Protect against
malformed paths (trailing backslashes).
* win/tools/scripts/feature-defines.pl: Make sure result
has an EOL character.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170940 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Win] Unreviewed build fix after r170930.
bfulgham@apple.com [Thu, 10 Jul 2014 00:22:35 +0000 (00:22 +0000)]
[Win] Unreviewed build fix after r170930.

* WTF.vcxproj/copy-files.cmd: Include cf files in
post-build copy operation.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170939 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoMake SharedBuffer::append(SharedBuffer*) be smarter about CFData and data arrays
psolanki@apple.com [Thu, 10 Jul 2014 00:18:17 +0000 (00:18 +0000)]
Make SharedBuffer::append(SharedBuffer*) be smarter about CFData and data arrays
https://bugs.webkit.org/show_bug.cgi?id=134731

Reviewed by Antti Koivisto.

If the target SharedBuffer has a CFDataRef or a data array then we can simply retain that
CFDataRef or data array elements in the SharedBuffer being appended to. This avoids
unnecessary copying.

No new tests because no functional changes.

* platform/SharedBuffer.cpp:
(WebCore::SharedBuffer::append):
(WebCore::SharedBuffer::maybeAppendPlatformData):
* platform/SharedBuffer.h:
* platform/cf/SharedBufferCF.cpp:
(WebCore::SharedBuffer::maybeAppendPlatformData):
(WebCore::SharedBuffer::maybeAppendDataArray):
* platform/soup/SharedBufferSoup.cpp:
(WebCore::SharedBuffer::maybeAppendPlatformData):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170938 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Win] Remove uses of 'bash' in build system
bfulgham@apple.com [Wed, 9 Jul 2014 23:48:26 +0000 (23:48 +0000)]
[Win] Remove uses of 'bash' in build system
https://bugs.webkit.org/show_bug.cgi?id=134782
<rdar://problem/17615533>

Reviewed by Dean Jackson.

Remove uses of 'bash' by replacing Windows-specific bash scripts
with Perl equivalents.

Source/JavaScriptCore:
* JavaScriptCore.vcxproj/JavaScriptCoreGenerated.make:
* JavaScriptCore.vcxproj/JavaScriptCoreGenerated.vcxproj:
* JavaScriptCore.vcxproj/JavaScriptCoreGenerated.vcxproj.filters:
* JavaScriptCore.vcxproj/JavaScriptCorePreBuild.cmd:
* JavaScriptCore.vcxproj/LLInt/LLIntAssembly/LLIntAssembly.make:
* JavaScriptCore.vcxproj/LLInt/LLIntAssembly/LLIntAssembly.vcxproj:
* JavaScriptCore.vcxproj/LLInt/LLIntAssembly/build-LLIntAssembly.pl: Copied from Source/JavaScriptCore/JavaScriptCore.vcxproj/LLInt/LLIntAssembly/build-LLIntAssembly.sh.
* JavaScriptCore.vcxproj/LLInt/LLIntAssembly/build-LLIntAssembly.sh: Removed.
* JavaScriptCore.vcxproj/LLInt/LLIntDesiredOffsets/LLIntDesiredOffsets.make:
* JavaScriptCore.vcxproj/LLInt/LLIntDesiredOffsets/LLIntDesiredOffsets.vcxproj:
* JavaScriptCore.vcxproj/LLInt/LLIntDesiredOffsets/build-LLIntDesiredOffsets.pl: Copied from Source/JavaScriptCore/JavaScriptCore.vcxproj/LLInt/LLIntDesiredOffsets/build-LLIntDesiredOffsets.sh.
* JavaScriptCore.vcxproj/LLInt/LLIntDesiredOffsets/build-LLIntDesiredOffsets.sh: Removed.
* JavaScriptCore.vcxproj/build-generated-files.pl: Copied from Source/JavaScriptCore/JavaScriptCore.vcxproj/build-generated-files.sh.
* JavaScriptCore.vcxproj/build-generated-files.sh: Removed.
* JavaScriptCore.vcxproj/jsc/jscPreBuild.cmd:
* JavaScriptCore.vcxproj/testRegExp/testRegExpPreBuild.cmd:
* JavaScriptCore.vcxproj/testapi/testapiPreBuild.cmd:

Source/WebCore:
* WebCore.vcxproj/WebCoreGenerated.make:
* WebCore.vcxproj/WebCoreGenerated.vcxproj:
* WebCore.vcxproj/WebCoreGenerated.vcxproj.filters:
* WebCore.vcxproj/WebCorePreBuild.cmd:
* WebCore.vcxproj/build-generated-files.pl: Copied from Source/WebCore/WebCore.vcxproj/build-generated-files.sh.
* WebCore.vcxproj/build-generated-files.sh: Removed.
* WebCore.vcxproj/migrate-scripts.pl: Copied from Source/WebCore/WebCore.vcxproj/migrate-scripts.sh.
* WebCore.vcxproj/migrate-scripts.sh: Removed.

Source/WebInspectorUI:
* WebInspectorUI.vcxproj/WebInspectorUI.make:
* WebInspectorUI.vcxproj/WebInspectorUI.vcxproj:
* WebInspectorUI.vcxproj/WebInspectorUI.vcxproj.filters:
* WebInspectorUI.vcxproj/build-webinspectorui.pl: Copied from Source/WebInspectorUI/WebInspectorUI.vcxproj/build-webinspectorui.sh.
* WebInspectorUI.vcxproj/build-webinspectorui.sh: Removed.

Source/WebKit:
* WebKit.vcxproj/Interfaces/FixMIDLHeaders.pl:
* WebKit.vcxproj/Interfaces/InterfacesPreBuild.cmd:
* WebKit.vcxproj/WebKit/WebKitPreBuild.cmd:
* WebKit.vcxproj/WebKit/WebKitVersion.cmd:
* WebKit.vcxproj/WebKitExportGenerator/WebKitExportGeneratorBuildCmd.cmd:
* WebKit.vcxproj/WebKitExportGenerator/WebKitExportGeneratorPreBuild.cmd:
* WebKit.vcxproj/WebKitGUID/WebKitGUIDPreBuild.cmd:

Source/WTF:
* WTF.vcxproj/WTFGenerated.make:
* WTF.vcxproj/WTFGenerated.vcxproj:
* WTF.vcxproj/WTFGenerated.vcxproj.filters:
* WTF.vcxproj/WTFPreBuild.cmd:
* WTF.vcxproj/build-generated-files.pl: Copied from Source/WTF/WTF.vcxproj/build-generated-files.sh.
* WTF.vcxproj/build-generated-files.sh: Removed.

Tools:
* DumpRenderTree/DumpRenderTree.vcxproj/DumpRenderTree/DumpRenderTreePreBuild.cmd:
* DumpRenderTree/DumpRenderTree.vcxproj/ImageDiff/ImageDiffPreBuild.cmd:
* DumpRenderTree/DumpRenderTree.vcxproj/TestNetscapePlugin/TestNetscapePluginPreBuild.cmd:
* TestWebKitAPI/TestWebKitAPI.vcxproj/TestWebKitAPIPreBuild.cmd:
* WinLauncher/WinLauncher.vcxproj/WinLauncherLibPreBuild.cmd:
* WinLauncher/WinLauncher.vcxproj/WinLauncherPreBuild.cmd:

WebKitLibraries:
* win/tools/scripts/feature-defines.pl: Copied from WebKitLibraries/win/tools/scripts/feature-defines.sh.
* win/tools/scripts/feature-defines.sh: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170937 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoSupport transparent WKWebViews
andersca@apple.com [Wed, 9 Jul 2014 22:06:29 +0000 (22:06 +0000)]
Support transparent WKWebViews
https://bugs.webkit.org/show_bug.cgi?id=134779
<rdar://problem/17351058>

Reviewed by Tim Horton.

Source/WebCore:
Schedule rebuilding the compositing layers if a FrameView's transparency changes.

* page/FrameView.cpp:
(WebCore::FrameView::setTransparent):

Source/WebKit2:
* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView initWithFrame:configuration:]):
Call _updateScrollViewBackground instead of setting the background color.

(contentZoomScale):
Use dot notation.

(scrollViewBackgroundColor):
Helper function that returns the scroll view background color.
If the web view isn't opaque, we want the scroll view to be transparent.

(-[WKWebView _updateScrollViewBackground]):
Call scrollViewBackgroundColor.

(-[WKWebView setOpaque:]):
Call WebPageProxy::setDrawsBackground and update the scroll view background.

(-[WKWebView setBackgroundColor:]):
Call setBackgroundColor on the content view.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170935 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[iOS] WebKit can crash under QuickLookDocumentData::encode() when viewing a QuickLook...
aestes@apple.com [Wed, 9 Jul 2014 21:58:13 +0000 (21:58 +0000)]
[iOS] WebKit can crash under QuickLookDocumentData::encode() when viewing a QuickLook preview
https://bugs.webkit.org/show_bug.cgi?id=134780

Reviewed by Tim Horton.

Don't use CFDataCreateWithBytesNoCopy() when we can't guarantee the lifetime of the copied-from DataReference
will match or exceed that of the CFDataRef. Copy the data instead.

* WebProcess/Network/WebResourceLoader.cpp:
(WebKit::WebResourceLoader::didReceiveData):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170934 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoCSS canvas color parsing accepts invalid color identifiers
jfernandez@igalia.com [Wed, 9 Jul 2014 21:56:05 +0000 (21:56 +0000)]
CSS canvas color parsing accepts invalid color identifiers
https://bugs.webkit.org/show_bug.cgi?id=134661

Reviewed by Benjamin Poulain.

Source/WebCore:
Current implementation of the CSSParser::parseSystemColor assumes
that if a valid cssValueKeywordID is got then it has to be a valid
color. Such assumption is wrong and lead to many bugs and layout
test failures.

The parseSystemFunction determines now whether the parsed color is
valid or not.

Addtionally, a new method has been added to share the logic of
determining whether a CSSValueID is a valid primitive values for
colors or not. Generally, we should avoid passing invalid color
identifiers to the theming API.

No new tests, but added additional cases to the
canvas-color-serialization.html, test-setting-canvas-color and
rgb-color-parse test.

* css/CSSParser.cpp:
(WebCore::validPrimitiveValueColor): Added.
(WebCore::parseColorValue):
(WebCore::CSSParser::parseSystemColor):

LayoutTests:
The parseSystemFunction determines now whether the parsed color is
valid or not.

The rgb-color-parser covers css style color parsing cases, which
already provide coverage for invalid color identifiers. I've added
a few more, though.

* fast/canvas/canvas-color-serialization-expected.txt:
* fast/canvas/script-tests/canvas-color-serialization.js:
* fast/css/test-setting-canvas-color-expected.txt:
* fast/css/test-setting-canvas-color.html:
* svg/dom/rgb-color-parser-expected.txt:
* svg/dom/rgb-color-parser.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170933 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoSource/WebCore: [MSE] http/tests/media/media-source/mediasource-remove.html is failing
jer.noble@apple.com [Wed, 9 Jul 2014 21:37:17 +0000 (21:37 +0000)]
Source/WebCore: [MSE] http/tests/media/media-source/mediasource-remove.html is failing
https://bugs.webkit.org/show_bug.cgi?id=134768

Reviewed by Eric Carlson.

Fix multiple bugs causing mediasource-remove.html to fail:

Separate out setDuration() into setDurationInternal() so that steps which require
us to run the "duration change algorithm" don't bail out if the SoureBuffer is
inside updating().

* Modules/mediasource/MediaSource.cpp:
(WebCore::MediaSource::setDuration):
(WebCore::MediaSource::setDurationInternal):
* Modules/mediasource/MediaSource.h:
* Modules/mediasource/SourceBuffer.cpp:
(WebCore::SourceBuffer::sourceBufferPrivateDidReceiveInitializationSegment):
(WebCore::SourceBuffer::sourceBufferPrivateDidReceiveSample):

When removing coded frames, do so starting from the first sample in the range in
decode order, so that frames dependant on removed frames are themselves removed. Add
a convenience method in SampleMap findSampleWithPresentationTime(), and rename
findSampleAfterPresentationTime() to findSampleOnOrAfterPresentationTime() to correctly
reflect what the method does, and simplify its implementation by searching the map's keys
directly.

* Modules/mediasource/SampleMap.cpp:
(WebCore::PresentationOrderSampleMap::findSampleWithPresentationTime):
(WebCore::PresentationOrderSampleMap::findSampleOnOrAfterPresentationTime):
(WebCore::DecodeOrderSampleMap::findSyncSampleAfterPresentationTime):
(WebCore::PresentationOrderSampleMap::findSampleAfterPresentationTime): Deleted.
* Modules/mediasource/SampleMap.h:
* Modules/mediasource/SourceBuffer.cpp:
(WebCore::decodeTimeComparator):
(WebCore::SourceBuffer::removeCodedFrames):

Throw the correct exception (INVALID_STATE_ERR) from SourceBuffer::remove().

* Modules/mediasource/SourceBuffer.cpp:
(WebCore::SourceBuffer::remove):

LayoutTests: [MSE] http/tests/media/media-source/mediasource-remove.html is failing.
https://bugs.webkit.org/show_bug.cgi?id=134768

Reviewed by Eric Carlson.

Update mediasource-remove.html with the correct locations of sync-samples.

* http/tests/media/media-source/mediasource-remove.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170932 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION(r167770): Cannot login in Evernote Safari extension
rniwa@webkit.org [Wed, 9 Jul 2014 21:36:16 +0000 (21:36 +0000)]
REGRESSION(r167770): Cannot login in Evernote Safari extension
https://bugs.webkit.org/show_bug.cgi?id=134776

Reviewed by Enrica Casucci.

Unfortunately, NSWindowDidBecomeKeyNotification does not notify us when a popover becomes key
so we need to keep using the SPI (_windowChangedKeyState) I removed in r167770.

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

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170931 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoAdd SharedBuffer::wrapCFDataArray() and use it
psolanki@apple.com [Wed, 9 Jul 2014 21:19:47 +0000 (21:19 +0000)]
Add SharedBuffer::wrapCFDataArray() and use it
https://bugs.webkit.org/show_bug.cgi?id=134733

Reviewed by Antti Koivisto.

No new tests. Should be covered by existing tests.

* platform/SharedBuffer.h:
* platform/cf/SharedBufferCF.cpp:
(WebCore::SharedBuffer::wrapCFDataArray):
(WebCore::SharedBuffer::SharedBuffer):
* platform/network/ResourceHandle.h:
* platform/network/cf/ResourceHandleCFNet.cpp:
(WebCore::ResourceHandle::handleDataArray): Deleted.
* platform/network/cf/ResourceHandleCFURLConnectionDelegateWithOperationQueue.cpp:
(WebCore::ResourceHandleCFURLConnectionDelegateWithOperationQueue::didReceiveDataArray):
* platform/network/cf/SynchronousResourceHandleCFURLConnectionDelegate.cpp:
(WebCore::SynchronousResourceHandleCFURLConnectionDelegate::didReceiveDataArray):
* platform/network/mac/WebCoreResourceHandleAsDelegate.mm:
(-[WebCoreResourceHandleAsDelegate connection:didReceiveDataArray:]):
* platform/network/mac/WebCoreResourceHandleAsOperationQueueDelegate.mm:
(-[WebCoreResourceHandleAsOperationQueueDelegate connection:didReceiveDataArray:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170930 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoMove resource buffering from SynchronousNetworkLoaderClient to NetworkResourceLoader
psolanki@apple.com [Wed, 9 Jul 2014 20:55:35 +0000 (20:55 +0000)]
Move resource buffering from SynchronousNetworkLoaderClient to NetworkResourceLoader
https://bugs.webkit.org/show_bug.cgi?id=134732

Reviewed by Darin Adler.

Source/WebCore:
No new tests because no functional changes.

* WebCore.exp.in:

Source/WebKit2:
Buffer the resource in NetworkResourceLoader instead of SynchronousNetworkLoaderClient. This
is in preparation for bug 134560 where we will be supporting JS and CSS resource buffering
that uses AsynchronousNetworkLoaderClient.

* NetworkProcess/NetworkResourceLoader.cpp:
(WebKit::NetworkResourceLoader::NetworkResourceLoader):
(WebKit::NetworkResourceLoader::didReceiveBuffer):
(WebKit::NetworkResourceLoader::didFinishLoading):
* NetworkProcess/NetworkResourceLoader.h:
(WebKit::NetworkResourceLoader::bufferedData):
* NetworkProcess/SynchronousNetworkLoaderClient.cpp:
(WebKit::SynchronousNetworkLoaderClient::didReceiveBuffer):
(WebKit::SynchronousNetworkLoaderClient::didFinishLoading):
(WebKit::SynchronousNetworkLoaderClient::didFail):
(WebKit::SynchronousNetworkLoaderClient::sendDelayedReply):
* NetworkProcess/SynchronousNetworkLoaderClient.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170928 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoFix the !USE(IOSURFACE) build.
timothy_horton@apple.com [Wed, 9 Jul 2014 20:52:35 +0000 (20:52 +0000)]
Fix the !USE(IOSURFACE) build.

* WebCore.exp.in:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170927 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[iOS][WK2] subviews of the unscaled view drift out during CA animations
benjamin@webkit.org [Wed, 9 Jul 2014 19:47:44 +0000 (19:47 +0000)]
[iOS][WK2] subviews of the unscaled view drift out during CA animations
https://bugs.webkit.org/show_bug.cgi?id=134751

Patch by Benjamin Poulain <bpoulain@apple.com> on 2014-07-09
Reviewed by Enrica Casucci.

It is not possible to animate the WKContentView and the inverse view in such a way
that the combined matrix remain the identity for every frame of the animation.

This patch solves the issue by moving the unscaled view as a sibling of WKContentView
instead of a child so that we do not need to update two scales simultaneously.

* UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView initWithFrame:configuration:]):
(-[WKWebView _processDidExit]):
(-[WKWebView _dynamicViewportUpdateChangedTargetToScale:position:]):
Set the z scale to 1 or no coordinate transform will work with this view.

(-[WKWebView _beginAnimatedResizeWithUpdates:]):
(-[WKWebView _endAnimatedResize]):
* UIProcess/ios/WKContentView.mm:
(-[WKContentView didUpdateVisibleRect:unobscuredRect:unobscuredRectInScrollViewCoordinates:scale:minimumScale:inStableState:isChangingObscuredInsetsInteractively:]):
* UIProcess/ios/WKContentViewInteraction.h:
* UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView setupInteraction]):
Since we need to observe changes inside the animation block, the code now use KVO to observe changes of scale.

(-[WKContentView cleanupInteraction]):
(-[WKContentView unscaledView]):
(-[WKContentView inverseScale]):
(-[WKContentView observeValueForKeyPath:ofObject:change:context:]):
We update the utility views as usual. The extra bits here are to deal with views coming in during an animation.

If a new utility view comes in during a scaling animation, we do not want to start a new animation with the same curve
to end up at the right place. To avoid any issue, we just hide the view until the animation is finished.

(-[WKContentView hitTest:withEvent:]):
(-[WKContentView _showTapHighlight]):
(-[WKContentView _updateUnscaledView]): Deleted.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170926 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUse IOSurface ViewSnapshots everywhere on Mac, remove JPEG encoding path
timothy_horton@apple.com [Wed, 9 Jul 2014 19:31:33 +0000 (19:31 +0000)]
Use IOSurface ViewSnapshots everywhere on Mac, remove JPEG encoding path
https://bugs.webkit.org/show_bug.cgi?id=134773

Reviewed by Anders Carlsson.

* UIProcess/API/mac/WKView.mm:
(-[WKView _takeViewSnapshot]):
* UIProcess/mac/ViewSnapshotStore.h:
* UIProcess/mac/ViewSnapshotStore.mm:
(WebKit::ViewSnapshotStore::ViewSnapshotStore):
(WebKit::ViewSnapshotStore::~ViewSnapshotStore):
(WebKit::ViewSnapshotStore::recordSnapshot):
(WebKit::ViewSnapshot::clearImage):
(WebKit::ViewSnapshot::asLayerContents):
(WebKit::createIOSurfaceFromImage): Deleted.
(WebKit::compressImageAsJPEG): Deleted.
(WebKit::ViewSnapshotStore::reduceSnapshotMemoryCost): Deleted.
(WebKit::ViewSnapshotStore::didCompressSnapshot): Deleted.
Remove all ViewSnapshot(Store) code related to JPEG-encoded snapshots.
Remove the "image" member on ViewSnapshot; Mac will always start out with an IOSurface instead.
Adopt WebCore::IOSurface::createFromImage to make that happen.
Add a comment noting that if a snapshot comes back empty, we should throw it away completely.

* WebCore.exp.in:
* platform/graphics/cocoa/IOSurface.h:
* platform/graphics/cocoa/IOSurface.mm:
(IOSurface::createFromImage):
Move make-an-IOSurface-from-a-CGImageRef into WebCore::IOSurface.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170925 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoImplement Editor::fontAttributesForSelectionStart() or iOS.
enrica@apple.com [Wed, 9 Jul 2014 19:11:47 +0000 (19:11 +0000)]
Implement Editor::fontAttributesForSelectionStart() or iOS.
https://bugs.webkit.org/show_bug.cgi?id=134771
<rdar://problem/16167838>

Reviewed by Ryosuke Niwa.

Source/WebCore:
Implement for iOS the equivalent function we have for OS X.

* editing/ios/EditorIOS.mm:
(WebCore::Editor::fontAttributesForSelectionStart):

Source/WebKit/mac:
* WebView/WebViewPrivate.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170924 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoRemoteLayerBackingStore::ensureBackingStore should ensure that the entire backing...
andersca@apple.com [Wed, 9 Jul 2014 19:10:21 +0000 (19:10 +0000)]
RemoteLayerBackingStore::ensureBackingStore should ensure that the entire backing store gets redrawn
https://bugs.webkit.org/show_bug.cgi?id=134772

Reviewed by Tim Horton.

* Shared/mac/RemoteLayerBackingStore.h:
(WebKit::RemoteLayerBackingStore::Buffer::operator bool):
* Shared/mac/RemoteLayerBackingStore.mm:
(WebKit::RemoteLayerBackingStore::ensureBackingStore):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170923 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Win] Remove use of 'grep' in build steps
bfulgham@apple.com [Wed, 9 Jul 2014 18:05:06 +0000 (18:05 +0000)]
[Win] Remove use of 'grep' in build steps
https://bugs.webkit.org/show_bug.cgi?id=134770
<rdar://problem/17608783>

Reviewed by Tim Horton.

Replace uses of the grep command in Windows builds with the equivalent
Perl program.

Source/JavaScriptCore:
* JavaScriptCore.vcxproj/JavaScriptCorePreBuild.cmd:
* JavaScriptCore.vcxproj/jsc/jscPreBuild.cmd:
* JavaScriptCore.vcxproj/testRegExp/testRegExpPreBuild.cmd:
* JavaScriptCore.vcxproj/testapi/testapiPreBuild.cmd:

Source/WebCore:
* WebCore.vcxproj/WebCoreGenerated.make:
* WebCore.vcxproj/WebCorePreBuild.cmd:
* platform/graphics/avfoundation/cf/WebCoreAVCFResourceLoader.cpp:
(WebCore::WebCoreAVCFResourceLoader::notifyFinished):

Source/WebKit:
* WebKit.vcxproj/Interfaces/InterfacesPreBuild.cmd:
* WebKit.vcxproj/WebKit/WebKitPreBuild.cmd:
* WebKit.vcxproj/WebKitExportGenerator/WebKitExportGeneratorPreBuild.cmd:
* WebKit.vcxproj/WebKitGUID/WebKitGUIDPreBuild.cmd:

Source/WTF:
* WTF.vcxproj/WTFPreBuild.cmd:

Tools:
* DumpRenderTree/DumpRenderTree.vcxproj/DumpRenderTree/DumpRenderTreePreBuild.cmd:
* DumpRenderTree/DumpRenderTree.vcxproj/ImageDiff/ImageDiffPreBuild.cmd:
* DumpRenderTree/DumpRenderTree.vcxproj/TestNetscapePlugin/TestNetscapePluginPreBuild.cmd:
* TestWebKitAPI/TestWebKitAPI.vcxproj/TestWebKitAPIPreBuild.cmd:
* WinLauncher/WinLauncher.vcxproj/WinLauncherLibPreBuild.cmd:
* WinLauncher/WinLauncher.vcxproj/WinLauncherPreBuild.cmd:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170922 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoAdded css jit profiler, disabled by default.
achristensen@apple.com [Wed, 9 Jul 2014 17:49:59 +0000 (17:49 +0000)]
Added css jit profiler, disabled by default.
https://bugs.webkit.org/show_bug.cgi?id=134695

Reviewed by Benjamin Poulain.

* css/ElementRuleCollector.cpp:
(WebCore::ElementRuleCollector::ruleMatches):
Tell the RuleData when its compiled selector is used.
* css/RuleSet.cpp:
(WebCore::RuleData::RuleData):
Initialize the compiled selector use count to 0.
* css/RuleSet.h:
(WebCore::RuleData::~RuleData):
(WebCore::RuleData::compiledSelectorUsed):
Count and log the number of times a compiled selector is used.
* cssjit/SelectorCompiler.h:
Added option for CSS_SELECTOR_JIT_PROFILING set to 0 by default.
* dom/SelectorQuery.cpp:
(WebCore::SelectorDataList::executeCompiledSimpleSelectorChecker):
(WebCore::SelectorDataList::execute):
Tell the SelectorData when its compiled selector is used.
* dom/SelectorQuery.h:
(WebCore::SelectorDataList::SelectorData::SelectorData):
(WebCore::SelectorDataList::SelectorData::~SelectorData):
(WebCore::SelectorDataList::SelectorData::compiledSelectorUsed):
Count and log the number of times a compiled selector is used.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170921 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[iOS] caption size is sometimes incorrect in fullscreen
eric.carlson@apple.com [Wed, 9 Jul 2014 17:17:16 +0000 (17:17 +0000)]
[iOS] caption size is sometimes incorrect in fullscreen
https://bugs.webkit.org/show_bug.cgi?id=134740

Reviewed by Jer Noble.

Captions on iOS are displayed in fullscreen with a TextTrackRepresentation object. Because
the fullscreen video presentation is controlled by code in the UI process running on the
UI thread, WebCore is notified of changes to fullscreen state asynchronously. This resulted
in the TextTrackRepresentation object being created and/or destroyed too late some of the
time, which caused us to sometimes display captions incorrectly. Fix this by setting up and
tearing down the TextTrackRepresentation object when WebCore's 'webkitfullscreenchange'
event fires.

* Modules/mediacontrols/MediaControlsHost.cpp:
(WebCore::MediaControlsHost::enteredFullscreen): Notify text track container.
(WebCore::MediaControlsHost::exitedFullscreen): Ditto.
* Modules/mediacontrols/MediaControlsHost.h:
* Modules/mediacontrols/MediaControlsHost.idl:

* Modules/mediacontrols/mediaControlsApple.js:
(Controller.prototype.handleFullscreenChange): Notify host of fullscreen change.

* html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::requiresTextTrackRepresentation): Only return true when in
    fullscreen.
(WebCore::HTMLMediaElement::setVideoFullscreenLayer): Call updateTextTrackDisplay.

* html/shadow/MediaControlElements.cpp:
(WebCore::MediaControlTextTrackContainerElement::MediaControlTextTrackContainerElement): Initialize
    m_updateTextTrackRepresentationStyle to false.
(WebCore::MediaControlTextTrackContainerElement::updateDisplay): Move logic for creating
    TextTrackRepresentation to updateTextTrackRepresentation.
(WebCore::MediaControlTextTrackContainerElement::updateActiveCuesFontSize): New, split out of updateTimerFired.
(WebCore::MediaControlTextTrackContainerElement::updateTimerFired): Move code to force immediate
    font size change to updateActiveCuesFontSize.
(WebCore::MediaControlTextTrackContainerElement::updateTextTrackRepresentation): New, update
    text track representation, creating first if necessary.
(WebCore::MediaControlTextTrackContainerElement::clearTextTrackRepresentation): Add an early
    return if we don't have a text track representation.
(WebCore::MediaControlTextTrackContainerElement::updateStyleForTextTrackRepresentation): Early
    return if there is nothing to be done.
(WebCore::MediaControlTextTrackContainerElement::enteredFullscreen): Force a caption update
    if there are visible captions.
(WebCore::MediaControlTextTrackContainerElement::updateSizes): Set m_updateTextTrackRepresentationStyle
    to true.
(WebCore::MediaControlTextTrackContainerElement::textTrackRepresentationBoundsChanged):  Force a
    caption update if there are visible captions.
* html/shadow/MediaControlElements.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170920 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agofast/dom/HTMLImageElement/sizes/image-sizes-js-change.html is failing in some bots
antti@apple.com [Wed, 9 Jul 2014 17:12:39 +0000 (17:12 +0000)]
fast/dom/HTMLImageElement/sizes/image-sizes-js-change.html is failing in some bots
https://bugs.webkit.org/show_bug.cgi?id=134769

* TestExpectations: Skipping the test.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170919 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[MSE] http/tests/media/media-source/mediasource-endofstream-invaliderror.html is...
jer.noble@apple.com [Wed, 9 Jul 2014 16:29:32 +0000 (16:29 +0000)]
[MSE] http/tests/media/media-source/mediasource-endofstream-invaliderror.html is failing.
https://bugs.webkit.org/show_bug.cgi?id=134413

Reviewed by Eric Carlson.

Source/WebCore:
Update MediaSource.endOfStream() to take an optional enum instead of an optional String.

* Modules/mediasource/MediaSource.cpp:
(WebCore::MediaSource::endOfStream):
* Modules/mediasource/MediaSource.h:
* Modules/mediasource/MediaSource.idl:

LayoutTests:
Fix an inadevntant editing error in mediasource-endofstream-invaliderror.html, and fix missing
whitespace at the end of http/tests/media/media-source/mediasource-endofstream-invaliderror-expected.txt.

* http/tests/media/media-source/mediasource-endofstream-invaliderror-expected.txt:
* http/tests/media/media-source/mediasource-endofstream-invaliderror.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170918 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed, update my list of email addresses.
rgf748@motorola.com [Wed, 9 Jul 2014 15:49:08 +0000 (15:49 +0000)]
Unreviewed, update my list of email addresses.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170917 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[GTK] Unreviewed GTK gardening.
commit-queue@webkit.org [Wed, 9 Jul 2014 13:54:16 +0000 (13:54 +0000)]
[GTK] Unreviewed GTK gardening.

Patch by Carlos Alberto Lopez Perez <clopez@igalia.com> on 2014-07-09

* platform/gtk/TestExpectations: Report and mark new failing and flaky tests. Adjust some tests expectations.
* platform/gtk/css1/formatting_model/inline_elements-expected.txt: Rebaseline after r170875.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170916 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed, rolling out r170889.
commit-queue@webkit.org [Wed, 9 Jul 2014 09:12:53 +0000 (09:12 +0000)]
Unreviewed, rolling out r170889.
https://bugs.webkit.org/show_bug.cgi?id=134754

This patch is wrong (Requested by philn on #webkit).

Reverted changeset:

"[GTK] Gstreamer missing from the install-dependencies script"
https://bugs.webkit.org/show_bug.cgi?id=134727
http://trac.webkit.org/changeset/170889

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170915 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[EFL] Fix crash caused by invalid cursor image.
commit-queue@webkit.org [Wed, 9 Jul 2014 07:29:18 +0000 (07:29 +0000)]
[EFL] Fix crash caused by invalid cursor image.
https://bugs.webkit.org/show_bug.cgi?id=134663

Patch by KwangHyuk Kim <hyuki.kim@samsung.com> on 2014-07-09
Reviewed by Gyuyoung Kim.

Remove calling of updateCursor since the custom cursor image is invalid once a mouse is out of the webview.

* UIProcess/API/efl/EwkView.cpp:
(EwkViewEventHandler<EVAS_CALLBACK_MOUSE_IN>::handleEvent):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170914 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoVersioning.
lforschler@apple.com [Wed, 9 Jul 2014 07:01:48 +0000 (07:01 +0000)]
Versioning.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170913 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoDOMWindows should register themselves with the GamepadManager when they have event...
beidson@apple.com [Wed, 9 Jul 2014 06:50:35 +0000 (06:50 +0000)]
DOMWindows should register themselves with the GamepadManager when they have event listeners
https://bugs.webkit.org/show_bug.cgi?id=134669

Reviewed by Dean Jackson.

No new tests (No effect in a currently tested config)

By having DOMWindows register with the GamepadManager when they have/don’t have gamepad event
listeners, we accomplish two things:
1 - It’s simple for GamepadManager to dispatch gamepad events directly to the DOMWindows interested.
2 - We know when to start monitoring for gamepad connections.

* Modules/gamepad/GamepadManager.cpp:
(WebCore::GamepadManager::GamepadManager):
(WebCore::GamepadManager::registerNavigator):
(WebCore::GamepadManager::unregisterNavigator):
(WebCore::GamepadManager::registerDOMWindow):
(WebCore::GamepadManager::unregisterDOMWindow):
(WebCore::GamepadManager::maybeStartMonitoringGamepads): Start monitoring gamepads if there are any
  registered DOMWindows or NavigatorGamepads, and we aren’t already monitoring them.
(WebCore::GamepadManager::maybeStopMonitoringGamepads): Stop monitoring gamepads if there aren’t any
  registered DOMWindows or NavigatorGameads, and we have been monitoring them before now.
* Modules/gamepad/GamepadManager.h:

* dom/EventNames.h:
(WebCore::EventNames::isGamepadEventType):

* page/DOMWindow.cpp:
(WebCore::DOMWindow::DOMWindow):
(WebCore::DOMWindow::~DOMWindow): If registered for gamepad events, unregister now.
(WebCore::DOMWindow::incrementGamepadEventListenerCount): If this is the first gamepad event
  listener, register with the GamepadManager now.
(WebCore::DOMWindow::decrementGamepadEventListenerCount): If this was the last gamepad event
  listener, unregister from the GamepadManager now.
(WebCore::DOMWindow::addEventListener):
(WebCore::DOMWindow::removeEventListener):
* page/DOMWindow.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170911 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago.: [EFL] Support Encrypted Media Extensions.
commit-queue@webkit.org [Wed, 9 Jul 2014 05:56:45 +0000 (05:56 +0000)]
.: [EFL] Support Encrypted Media Extensions.
https://bugs.webkit.org/show_bug.cgi?id=134750

Patch by Sun-woo Nam <sunny.nam@samsung.com> on 2014-07-08
Reviewed by Gyuyoung Kim.

Webkit needs to play encrypted media contents when media player
is played by Media source extensions and normal video procedure.

* Source/cmake/OptionsEfl.cmake: Turn on ENABLE_ENCRYPTED_MEDIA_V2

Source/WebCore: [EFL] Fix the build break when ENABLE_ENCRYPTED_MEDIA_V2 is turned on.
https://bugs.webkit.org/show_bug.cgi?id=134750

Patch by Sun-woo Nam <sunny.nam@samsung.com> on 2014-07-08
Reviewed by Gyuyoung Kim.

When turning ENABLE_ENCRYPTED_MEDIA_V2 on, build break occurs.
Some source files need to be included in CMakeLists.txt file for build.
And wrap the included header file which is built only on mac and wince.

* CMakeLists.txt:
* Modules/encryptedmedia/CDMPrivateMediaPlayer.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170910 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoRestore the assertion changed with 170839
benjamin@webkit.org [Wed, 9 Jul 2014 05:13:01 +0000 (05:13 +0000)]
Restore the assertion changed with 170839

* assembler/ARMv7Assembler.h:
(JSC::ARMv7Assembler::pop):
(JSC::ARMv7Assembler::push):
Revert the Assembler part of 170839. The assertions do not match both encoding.

I'll add specific version of push and pop instead.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170909 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[iOS] user caption preferences not applied
eric.carlson@apple.com [Wed, 9 Jul 2014 04:53:03 +0000 (04:53 +0000)]
[iOS] user caption preferences not applied
https://bugs.webkit.org/show_bug.cgi?id=134599

Reviewed by Sam Weinig.

* page/CaptionUserPreferences.cpp:
(WebCore::CaptionUserPreferences::updateCaptionStyleSheetOveride): All Pages in a PageGroup
do not necessarily use the same PageUserContentController, so apply and remove caption user
the style sheet directly to each page rather than to the page group.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170908 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed, rolling out r170894.
commit-queue@webkit.org [Wed, 9 Jul 2014 00:51:07 +0000 (00:51 +0000)]
Unreviewed, rolling out r170894.
https://bugs.webkit.org/show_bug.cgi?id=134749

"It broke the build on mountainlion release 32bit" (Requested
by mmirman on #webkit).

Reverted changeset:

"Added css jit profiler, disabled by default."
https://bugs.webkit.org/show_bug.cgi?id=134695
http://trac.webkit.org/changeset/170894

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170906 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Win] Implement auto-version as a Perl program
bfulgham@apple.com [Wed, 9 Jul 2014 00:01:02 +0000 (00:01 +0000)]
[Win] Implement auto-version as a Perl program
https://bugs.webkit.org/show_bug.cgi?id=134745
<rdar://problem/17597801>

Reviewed by David Kilzer.

Source/WebKit:
Modify build commands to use new Perl version of the
auto-version script.

* WebKit.vcxproj/Interfaces/InterfacesPreBuild.cmd:
* WebKit.vcxproj/WebKit/WebKitPreBuild.cmd:

WebKitLibraries:
Translate the auto-version.sh bash script into an equivalent
Perl script.

* win/tools/scripts/auto-version.pl: Added.
* win/tools/scripts/auto-version.sh: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170905 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUpdate policy for Inactive Committer and Reviewer status
bjonesbe@adobe.com [Tue, 8 Jul 2014 23:58:05 +0000 (23:58 +0000)]
Update policy for Inactive Committer and Reviewer status
https://bugs.webkit.org/show_bug.cgi?id=134403

Reviewed by Nico Weber.

* coding/commit-review-policy.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170904 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoRemoteInspector::shared() should not call WTF::initializeMainThread()
jhoneycutt@apple.com [Tue, 8 Jul 2014 23:31:52 +0000 (23:31 +0000)]
RemoteInspector::shared() should not call WTF::initializeMainThread()
<https://bugs.webkit.org/show_bug.cgi?id=134747>
<rdar://problem/17161482>

Reviewed by Joseph Pecoraro.

* inspector/remote/RemoteInspector.mm:
(Inspector::RemoteInspector::shared):
Don't call WTF::initializeMainThread(). WTF threading is initialized by
JSC::initializeThreading().

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170903 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoRemove WebBackForwardListItems when their owning page goes away
timothy_horton@apple.com [Tue, 8 Jul 2014 23:17:26 +0000 (23:17 +0000)]
Remove WebBackForwardListItems when their owning page goes away
https://bugs.webkit.org/show_bug.cgi?id=134709
<rdar://problem/17584645>

Reviewed by Dan Bernstein.

* Shared/WebBackForwardListItem.cpp:
(WebKit::WebBackForwardListItem::create):
(WebKit::WebBackForwardListItem::WebBackForwardListItem):
* Shared/WebBackForwardListItem.h:
(WebKit::WebBackForwardListItem::pageID):
Add the associated PageID to the WebBackForwardListItem.

* UIProcess/WebBackForwardList.cpp:
(WebKit::WebBackForwardList::restoreFromState):
Push the current PageID onto the WebBackForwardListItem.

* UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::removeWebPage):
Remove all of the WebBackForwardListItems from m_backForwardListItemMap when
the page they are associated with is removed.

(WebKit::WebProcessProxy::addBackForwardItem):
Push the PageID from the WebProcess onto the WebBackForwardListItem.

* UIProcess/WebProcessProxy.h:
* UIProcess/WebProcessProxy.messages.in:
* WebProcess/WebPage/WebBackForwardListProxy.cpp:
(WebKit::idToHistoryItemMap):
(WebKit::historyItemToIDMap):
(WebKit::updateBackForwardItem):
(WebKit::WebBackForwardListProxy::addItemFromUIProcess):
(WebKit::WK2NotifyHistoryItemChanged):
(WebKit::WebBackForwardListProxy::idForItem):
(WebKit::WebBackForwardListProxy::addItem):
(WebKit::WebBackForwardListProxy::goToItem):
(WebKit::WebBackForwardListProxy::close):
* WebProcess/WebPage/WebBackForwardListProxy.h:
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::restoreSession):
Keep track of the PageID that back-forward items were created by.
Pass the PageID along when registering WebBackForwardListItems.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170902 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoWebVideoFullscreenInterfaceAVKit should lazily create m_playerController.
commit-queue@webkit.org [Tue, 8 Jul 2014 23:11:07 +0000 (23:11 +0000)]
WebVideoFullscreenInterfaceAVKit should lazily create m_playerController.
https://bugs.webkit.org/show_bug.cgi?id=134706

Patch by Jeremy Jones <jeremyj@apple.com> on 2014-07-08
Reviewed by Simon Fraser.

* platform/ios/WebVideoFullscreenInterfaceAVKit.mm:
(WebVideoFullscreenInterfaceAVKit::setWebVideoFullscreenModel):
Don't set the delegate if m_playerController isn't created.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170901 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoresolve-ChangeLogs fails when WebKit repository is a git submodule
burg@cs.washington.edu [Tue, 8 Jul 2014 22:54:56 +0000 (22:54 +0000)]
resolve-ChangeLogs fails when WebKit repository is a git submodule
https://bugs.webkit.org/show_bug.cgi?id=134734

Reviewed by David Kilzer.

When using resolve-ChangeLogs in a WebKit git repository with a
non-default arrangement of GIT_DIR and GIT_WORK_TREE, incorrect
paths are computed and the script fails mysteriously.

We were naively computing GIT_WORK_TREE as being one level above
GIT_DIR, but this is not the case for submodules or a manually
configured GIT_WORK_TREE or GIT_DIR location.

* Scripts/VCSUtils.pm:
(gitTreeDirectory): Added. Find the equivalent of GIT_WORK_TREE.
(determineVCSRoot): Calculate the working tree root directly.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170900 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoWeb Inspector: Debugger popovers should not dismiss when mousing over them
commit-queue@webkit.org [Tue, 8 Jul 2014 22:42:57 +0000 (22:42 +0000)]
Web Inspector: Debugger popovers should not dismiss when mousing over them
https://bugs.webkit.org/show_bug.cgi?id=134687

Patch by Joseph Pecoraro <pecoraro@apple.com> on 2014-07-08
Reviewed by David Kilzer.

* UserInterface/Views/SourceCodeTextEditor.js:
(WebInspector.SourceCodeTextEditor.prototype._trackPopoverEvents):
Actually install the listeners in the set. Regression from code refactoring in r165052.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170899 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoVM::lastCachedString should be a Strong, not a Weak.
akling@apple.com [Tue, 8 Jul 2014 22:36:59 +0000 (22:36 +0000)]
VM::lastCachedString should be a Strong, not a Weak.
<https://webkit.org/b/134746>

Using Weak<JSString> for this regressed some of our bindings perf tests
due to Weak having to allocate a new WeakImpl every time the last cached
string changed. Make it a Strong instead should make that problem go away.

Reviewed by Geoffrey Garen.

* runtime/JSString.cpp:
(JSC::jsStringWithCacheSlowCase):
* runtime/VM.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170898 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoRemove webkit prefix from pointer lock tests.
commit-queue@webkit.org [Tue, 8 Jul 2014 22:36:24 +0000 (22:36 +0000)]
Remove webkit prefix from pointer lock tests.
https://bugs.webkit.org/show_bug.cgi?id=134744

Patch by Alex Christensen <achristensen@webkit.org> on 2014-07-08
Reviewed by Brady Eidson.

* http/tests/pointer-lock/iframe-sandboxed-allow-pointer-lock-expected.txt:
* http/tests/pointer-lock/iframe-sandboxed-allow-pointer-lock.html:
* http/tests/pointer-lock/iframe-sandboxed-expected.txt:
* http/tests/pointer-lock/iframe-sandboxed-nested-allow-pointer-lock-expected.txt:
* http/tests/pointer-lock/iframe-sandboxed-nested-allow-pointer-lock.html:
* http/tests/pointer-lock/iframe-sandboxed-nested-disallow-then-allow-pointer-lock-expected.txt:
* http/tests/pointer-lock/iframe-sandboxed-nested-disallow-then-allow-pointer-lock.html:
* http/tests/pointer-lock/iframe-sandboxed.html:
* http/tests/pointer-lock/pointerlockelement-different-origin-expected.txt:
* http/tests/pointer-lock/pointerlockelement-different-origin.html:
* http/tests/pointer-lock/pointerlockelement-same-origin-expected.txt:
* http/tests/pointer-lock/pointerlockelement-same-origin.html:
* http/tests/pointer-lock/requestPointerLock-can-not-transfer-between-documents-expected.txt:
* http/tests/pointer-lock/requestPointerLock-can-not-transfer-between-documents.html:
* http/tests/resources/pointer-lock/iframe-common.js:
(document.onpointerlockchange):
(document.onpointerlockerror):
(document.onwebkitpointerlockchange): Deleted.
(document.onwebkitpointerlockerror): Deleted.
* http/tests/resources/pointer-lock/pointer-lock-test-harness.js:
(expectOnlyChangeEvent):
(expectOnlyErrorEvent):
(expectNoEvents):
* pointer-lock/bug90391-move-then-window-open-crash-expected.txt:
* pointer-lock/bug90391-move-then-window-open-crash.html:
* pointer-lock/lock-already-locked-expected.txt:
* pointer-lock/lock-already-locked.html:
* pointer-lock/lock-element-not-in-dom-expected.txt:
* pointer-lock/lock-element-not-in-dom.html:
* pointer-lock/locked-element-iframe-removed-from-dom-expected.txt:
* pointer-lock/locked-element-iframe-removed-from-dom.html:
* pointer-lock/locked-element-removed-from-dom-expected.txt:
* pointer-lock/locked-element-removed-from-dom.html:
* pointer-lock/mouse-event-api-expected.txt:
* pointer-lock/mouse-event-api.html:
* pointer-lock/mouse-event-delivery-expected.txt:
* pointer-lock/mouse-event-delivery.html:
* pointer-lock/pointer-lock-api-expected.txt:
* pointer-lock/pointer-lock-api.html:
* pointer-lock/pointerlockchange-event-on-lock-lost-expected.txt:
* pointer-lock/pointerlockchange-event-on-lock-lost.html:
* pointer-lock/pointerlockchange-pointerlockerror-events-expected.txt:
* pointer-lock/pointerlockchange-pointerlockerror-events.html:
* pointer-lock/pointerlockelement-null-when-pending-expected.txt:
* pointer-lock/pointerlockelement-null-when-pending.html:
Remove webkit prefix on pointer lock functions and events.
This should have been done in r170585.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170897 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[WK2] Expose a few drawing/compositing settings on WKPreferences(Private)
timothy_horton@apple.com [Tue, 8 Jul 2014 22:24:41 +0000 (22:24 +0000)]
[WK2] Expose a few drawing/compositing settings on WKPreferences(Private)
https://bugs.webkit.org/show_bug.cgi?id=134645

Reviewed by Dan Bernstein.

* UIProcess/API/Cocoa/WKPreferences.mm:
(-[WKPreferences _compositingBordersVisible]):
(-[WKPreferences _setCompositingBordersVisible:]):
(-[WKPreferences _compositingRepaintCountersVisible]):
(-[WKPreferences _setCompositingRepaintCountersVisible:]):
(-[WKPreferences _tiledScrollingIndicatorVisible]):
(-[WKPreferences _setTiledScrollingIndicatorVisible:]):
* UIProcess/API/Cocoa/WKPreferencesPrivate.h:
Expose layer borders, repaint counters, and the tiled scrolling indicator on WKPreferences, as SPI.

* WebView/WebPreferences.mm:
(-[WebPreferences setSubpixelCSSOMElementMetricsEnabled:]):
Remove an extra unnecessary space.

* MiniBrowser/mac/BrowserWindowController.h:
* MiniBrowser/mac/MainMenu.xib:
Add menu items for layer borders and the tiled scrolling indicator.

* MiniBrowser/mac/WK1BrowserWindowController.m:
(-[WK1BrowserWindowController validateMenuItem:]):
(-[WK1BrowserWindowController layerBordersAreVisible]):
(-[WK1BrowserWindowController toggleLayerBordersVisibility:]):
Add getters/setters/validators for layer borders.

* MiniBrowser/mac/WK2BrowserWindowController.m:
(-[WK2BrowserWindowController awakeFromNib]):
(-[WK2BrowserWindowController validateMenuItem:]):
(-[WK2BrowserWindowController layerBordersAreVisible]):
(-[WK2BrowserWindowController toggleLayerBordersVisibility:]):
(-[WK2BrowserWindowController tiledScrollingIndicatorIsVisible]):
(-[WK2BrowserWindowController toggleTiledScrollingIndicatorVisibility:]):
Add getters/setters/validators for layer borders and the tiled scrolling indicator.

(-[WK2BrowserWindowController isSubpixelCSSOMElementMetricsEnabled]): Deleted.
(-[WK2BrowserWindowController toggleSubpixelCSSOMElementMetricsEnabled:]): Deleted.
Remove the WebKit2 version of the subpixel metrics pref, because it doesn't work, because
nothing in WebKit2 reads this default, and there is no WebKit2 API for adjusting this preference.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170896 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoEnsure we compute the height of replaced elements to 'auto' when appropriate.
bjonesbe@adobe.com [Tue, 8 Jul 2014 22:10:43 +0000 (22:10 +0000)]
Ensure we compute the height of replaced elements to 'auto' when appropriate.
https://bugs.webkit.org/show_bug.cgi?id=134700

Reviewed by David Hyatt.

Source/WebCore:
This fixes two issues:

1) If a replaced element has a percentage height specified then its
height should compute to 'auto' when its containing block does not
have a height 'specified explicitly'. We were taking this to mean when
its containing block's specified height value is 'auto' - in fact it
means when the containing block's computed height is auto.

2) Top and bottom on the containing block should only affect the
height of the block if that block has absolute or fixed position.

This brings us into line with IE, Firefox, and Chrome.

This is a port of Blink patches by Rob Hogan and David Vest.

Tests: css2.1/20110323/height-percentage-005.htm
       fast/css/replaced-element-ignore-top-bottom.html

* rendering/RenderBoxModelObject.cpp:
(WebCore::RenderBoxModelObject::hasAutoHeightOrContainingBlockWithAutoHeight):

LayoutTests:
* css2.1/20110323/height-percentage-005-expected.html: Added.
* css2.1/20110323/height-percentage-005.htm: Added.
* css2.1/20110323/support/black96x96.png: Added.
* fast/css/replaced-element-ignore-top-bottom-expected.txt: Added.
* fast/css/replaced-element-ignore-top-bottom.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@170895 268f45cc-cd09-0410-ab3c-d52691b4dbfc