Timeline
Jul 2, 2014:
- 9:25 PM Changeset in webkit [170746] by
-
- 4 edits in trunk/Source/WebCore
Removing an element with CSS -webkit-overflow-scrolling: touch breaks
subsequent touch events
<https://bugs.webkit.org/show_bug.cgi?id=134584>
<rdar://problem/17202021>
This issue is caused by the order of teardown for RenderLayers.
RenderLayer clears its m_backing in its destructor, calling the
RenderLayerBacking destructor. The RenderLayerBacking destructor calls
RenderLayerCompositor::willRemoveScrollingLayer(), which tries to
access the RenderLayer's backing that has already been cleared. This
causes us to skip the call to ChromeClient::removeScrollingLayer(),
resulting in a stale view that can block touch events.
Reviewed by Simon Fraser.
- rendering/RenderLayerBacking.cpp:
(WebCore::RenderLayerBacking::updateScrollingLayers):
When calling willRemoveScrollingLayerWithBacking(), pass both the layer
and the RenderLayerBacking.
- rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::willRemoveScrollingLayerWithBacking):
Renamed. Use the passed RenderLayerBacking, as the RenderLayer's
backing might have already been cleared.
(WebCore::RenderLayerCompositor::willRemoveScrollingLayer): Deleted.
- rendering/RenderLayerCompositor.h:
Renamed willRemoveScrollingLayer() to
willRemoveScrollingLayerWithBacking() and added a parameter.
- 9:03 PM Changeset in webkit [170745] by
-
- 5 edits1 add in trunk
Tools: [GTK] Unreviewed gardening.
Patch by Carlos Alberto Lopez Perez <clopez@igalia.com> on 2014-07-02
- Scripts/run-gtk-tests:
(TestRunner): Report and mark flaky test WebKit2Gtk/TestCookieManager:persistent-storage
LayoutTests: Unreviewed gardening.
Patch by Carlos Alberto Lopez Perez <clopez@igalia.com> on 2014-07-02
- platform/gtk/TestExpectations: Report and mark several flaky tests.
- platform/gtk/fast/text/indic-expected.txt: Added. Add baseline for GTK.
- platform/mac-wk2/TestExpectations: Report and mark flaky test http/tests/cache/iframe-304-crash.html.
- 8:52 PM Changeset in webkit [170744] by
-
- 2 edits in trunk/Source/WebKit2
[iOS][WK2] Make is safe/fast to use the animated resize API without resizing anything
https://bugs.webkit.org/show_bug.cgi?id=134570
Reviewed by Enrica Casucci.
Today, if someone calls _beginAnimatedResizeWithUpdates-_endAnimatedResize, it is always a very heavy process
forcing a relayout and can cause synchronous operations.
We should not force the caller of the API to maintain their own state tracking, we should do that for them.
With this patch, we track everything we need for a resize and only do the heavy lifting if anything has actually changed.
- UIProcess/API/Cocoa/WKWebView.mm:
(activeMinimumLayoutSize):
(activeMinimumLayoutSizeForMinimalUI):
(activeMaximumUnobscuredSize):
(activeOrientation):
This code is used in both _didRelaunchProcess and _beginAnimatedResizeWithUpdates. It is moved to static functions
to avoid having it in two places.
(-[WKWebView _didRelaunchProcess]):
(-[WKWebView _beginAnimatedResizeWithUpdates:]):
The updates now goes like this:
1) Set _isAnimatingResize to prevent the undesired side effects updateBlock() if we really need to resize.
2) Get all the original values before calling the update block.
3) Call the update block.
4a) If nothing useful for resize has changed, reset _isAnimatingResize to false.
We also need to update the visible content rect because the update block may have changed something unrelated
to the view size (scale, scroll position, etc).
4b) If we really need to resize, proceed as usual.
(-[WKWebView _endAnimatedResize]):
With the changes in _beginAnimatedResizeWithUpdates:, _isAnimatingResize is only set for real cases of animated resize,
bail out early if that flag isn't set.
The remaining code is unchanged, it is just not indented due to the removal of the if() branch.
- 8:31 PM Changeset in webkit [170743] by
-
- 3 edits in trunk/Source/WebKit2
Improve handling of primary offscreen plugins.
https://bugs.webkit.org/show_bug.cgi?id=134528.
<rdar://problem/17471864>
Reviewed by Dean Jackson.
- WebProcess/Plugins/PluginView.cpp:
Determine whether or not the plugin starts offscreen when the plugin is initialized.
(WebKit::PluginView::initializePlugin):
(WebKit::PluginView::pluginSnapshotTimerFired):
- WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::createPlugin):
Use the PluginProcessTypeNormal for offscreen plugins that are potentially primary plugins.
- 8:03 PM WebKitGTK/KeepingTheTreeGreen edited by
- (diff)
- 7:47 PM Changeset in webkit [170742] by
-
- 4 edits in tags/Safari-538.44.1/Source/WebCore
Roll out r170578.
- 7:31 PM Changeset in webkit [170741] by
-
- 6 edits in trunk/Tools
Clean up unused EFL buildbot information in BuildSlaveSupport and TestResultServer
https://bugs.webkit.org/show_bug.cgi?id=134581
Reviewed by Tim Horton.
Should be removed unused EFL bot info.
- BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
- BuildSlaveSupport/build.webkit.org-config/public_html/dashboard/Scripts/EWS.js:
(EWS):
- BuildSlaveSupport/build.webkit.org-config/public_html/dashboard/Scripts/WebKitBuildbot.js:
(WebKitBuildbot):
- TestResultServer/static-dashboards/builders.jsonp:
- TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
(test):
- 7:14 PM Changeset in webkit [170740] by
-
- 5 edits in tags/Safari-538.44.1/Source
Versioning.
- 7:03 PM Changeset in webkit [170739] by
-
- 1 copy in tags/Safari-538.44.1
New tag.
- 5:59 PM Changeset in webkit [170738] by
-
- 3 edits in trunk/Source/WebKit2
REGRESSION(WK2): Undo does not work in text fields in Safari.
https://bugs.webkit.org/show_bug.cgi?id=134572
<rdar://problem/17542238>
Reviewed by Benjamin Poulain.
This patch adds for iOS the same implementation we have already for OS X.
- UIProcess/ios/PageClientImplIOS.h:
- UIProcess/ios/PageClientImplIOS.mm:
(-[WKEditCommandObjC initWithWebEditCommandProxy:]):
(-[WKEditCommandObjC command]):
(-[WKEditorUndoTargetObjC undoEditing:]):
(-[WKEditorUndoTargetObjC redoEditing:]):
(WebKit::PageClientImpl::PageClientImpl):
(WebKit::PageClientImpl::registerEditCommand):
(WebKit::PageClientImpl::clearAllEditCommands):
(WebKit::PageClientImpl::canUndoRedo):
(WebKit::PageClientImpl::executeUndoRedo):
- 5:55 PM Changeset in webkit [170737] by
-
- 4 edits in trunk/Source/WebKit2
ProcessAssertion should also prevent UIApp suspension
https://bugs.webkit.org/show_bug.cgi?id=134563
Reviewed by Dan Bernstein.
If the application suspends then the child processes will, too.
Use beginBackgroundTaskWithName:expirationHandler: to request that the application remain runnable
while waiting for background tasks to complete.
- UIProcess/ProcessAssertion.cpp:
(WebKit::ProcessAssertion::~ProcessAssertion):
- Added destructor.
- UIProcess/ProcessAssertion.h:
- Added destructor.
- UIProcess/ios/ProcessAssertionIOS.mm:
(+[WKProcessAssertionBackgroundTaskManager shared]):
- singleton WKProcessAssertionBackgroundTaskManager.
(-[WKProcessAssertionBackgroundTaskManager init]):
(-[WKProcessAssertionBackgroundTaskManager dealloc]):
- register/remove notification handlers.
(-[WKProcessAssertionBackgroundTaskManager _updateBackgroundTask]):
- if we need to be runnable make sure we're holding a BackgroundTask, if not release it.
(-[WKProcessAssertionBackgroundTaskManager _applicationDidEnterBackgroundOrWillEnterForeground:]):
- detect when the app enters/leaves foreground; calls to _updateBackgroundTask.
(-[WKProcessAssertionBackgroundTaskManager incrementNeedsToRunInBackgroundCount]):
(-[WKProcessAssertionBackgroundTaskManager decrementNeedsToRunInBackgroundCount]):
- interface to update the count; calls to _updateBackgroundTask.
(WebKit::ProcessAssertion::ProcessAssertion):
(WebKit::ProcessAssertion::~ProcessAssertion):
- count ProcessAssertions in and out of existance.
(WebKit::ProcessAssertion::setState):
- count when the state changes.
- 4:41 PM Changeset in webkit [170736] by
-
- 12 edits2 adds in branches/ftlopt/Source/JavaScriptCore
Added In for cells to the FTL
- 4:02 PM Changeset in webkit [170735] by
-
- 5 edits in trunk/Source/WebCore
Ensure that the WebKit bundle version in the user agent string continues to match the current format.
<https://webkit.org/b/134524> / <rdar://problem/17447771>
Reviewed by Simon Fraser.
- page/cocoa/UserAgent.h:
- page/cocoa/UserAgent.mm:
(WebCore::userVisibleWebKitBundleVersionFromFullVersion): Updated to take an NSString now that it's internal
to the file.
(WebCore::userAgentBundleVersionFromFullVersionString): Limit the bundle version included in the user agent
string to three components.
- page/ios/UserAgentIOS.mm:
(WebCore::standardUserAgentWithApplicationName): Update to call userAgentBundleVersionFromFullVersionString.
- page/mac/UserAgentMac.mm:
(WebCore::standardUserAgentWithApplicationName): Ditto.
- 4:02 PM Changeset in webkit [170734] by
-
- 10 edits in trunk/Source
<https://webkit.org/b/134522> Remove duplication in code that prepares the user agent string on Mac and iOS
Reviewed by Simon Fraser.
Source/WebCore:
- page/cocoa/UserAgent.h:
- page/cocoa/UserAgent.mm:
(WebCore::userVisibleWebKitBundleVersionFromFullVersion): Moved from WebKit2.
- page/ios/UserAgentIOS.mm:
(WebCore::standardUserAgentWithApplicationName): Pass the WebKit bundle version through userVisibleWebKitBundleVersionFromFullVersion
before including it in the user agent string.
- page/mac/UserAgentMac.mm:
(WebCore::standardUserAgentWithApplicationName): Ditto.
Source/WebKit/mac:
- WebView/WebView.mm:
(webKitBundleVersionString): Return the entire CFBundleVersion now that WebCore handles formatting it.
(+[WebView _standardUserAgentWithApplicationName:]):
Source/WebKit2:
- UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::webKitBundleVersionString): Return the entire CFBundleVersion now that WebCore handles formatting it.
(WebKit::WebPageProxy::standardUserAgent):
- UIProcess/mac/WebPageProxyMac.mm:
(WebKit::webKitBundleVersionString): Ditto.
(WebKit::WebPageProxy::standardUserAgent):
- 3:54 PM Changeset in webkit [170733] by
-
- 10 edits in trunk/Source/WebKit2
Restrict network process sandbox
https://bugs.webkit.org/show_bug.cgi?id=134360
Reviewed by Sam Weinig.
Add more restrictions to the network process sandbox.
- NetworkProcess/cocoa/NetworkProcessCocoa.mm: (WebKit::NetworkProcess::platformInitializeNetworkProcessCocoa):
Always use the cache directory provided in the initialization parameters,
and make sure we consume the cookie directory extension.
- Resources/SandboxProfiles/ios/com.apple.WebKit.Networking.sb:
Make the sandbox profile much more restrictive.
- Shared/Network/NetworkProcessCreationParameters.cpp: (WebKit::NetworkProcessCreationParameters::encode): (WebKit::NetworkProcessCreationParameters::decode):
- Shared/Network/NetworkProcessCreationParameters.h:
The network process now requires an extension to access
its cookie storage.
- Shared/mac/SandboxUtilities.cpp: (WebKit::pathForProcessContainer):
- Shared/mac/SandboxUtilities.h:
We need to be able to get hold of our container so
that we can get the correct cookie storage directory.
- UIProcess/WebContext.cpp: (WebKit::WebContext::ensureNetworkProcess):
We have to pass in the an extension for the cookie storage directory when
initalising the network process
- UIProcess/mac/WebContextMac.mm: (WebKit::WebContext::platformDefaultCookieStorageDirectory):
Make sure we provide the correct location on IOS
- WebProcess/cocoa/WebProcessCocoa.mm: (WebKit::WebProcess::platformInitializeWebProcess):
Consume the cookie storage extension
- 3:52 PM Changeset in webkit [170732] by
-
- 2 edits in trunk/Source/WebKit2
URTBF after r170725.
- CMakeLists.txt: Removed Shared/LegacySessionState.cpp.
- 3:52 PM Changeset in webkit [170731] by
-
- 2 edits in trunk/Source/WebKit2
URTBF after r170719 for !CF platforms.
- UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::sessionStateData): Deleted.
(WebKit::WebPageProxy::restoreFromSessionStateData): Deleted.
- 3:38 PM Changeset in webkit [170730] by
-
- 7 edits1 copy3 moves1 add in trunk/Source
<https://webkit.org/b/134521> iOS should use shared code to determine the system marketing version
Reviewed by Simon Fraser.
Source/WebCore:
- WebCore.xcodeproj/project.pbxproj: Add the new files, and sort the groups they're in.
- page/cocoa/UserAgent.h: Copied from Source/WebCore/page/mac/UserAgent.h.
- page/cocoa/UserAgent.mm: Renamed from Source/WebCore/page/mac/UserAgent.h.
Move systemMarketingVersionForUserAgentString to a location where it can be shared between
Mac and iOS.
- page/ios/UserAgentIOS.mm:
(WebCore::standardUserAgentWithApplicationName): Switch to systemMarketingVersionForUserAgentString.
- page/mac/UserAgentMac.mm:
- platform/cocoa/SystemVersion.h: Renamed from Source/WebCore/platform/mac/SystemVersionMac.h.
- platform/cocoa/SystemVersion.mm: Renamed from Source/WebCore/platform/mac/SystemVersionMac.mm.
Move to a location that makes it clear this is shared between Mac and iOS. Enable the modern Mac
codepath for iOS as well.
Source/WebKit2:
- Shared/ios/ChildProcessIOS.mm: Update #import.
- Shared/mac/ChildProcessMac.mm: Ditto.
- 3:34 PM Changeset in webkit [170729] by
-
- 4 edits in branches/ftlopt/Source/JavaScriptCore
[ftlopt] Allocate a new GetterSetter if we change the value of any of its entries other than when they were previously null, so that if we constant-infer an accessor slot then we immediately get the function constant for free
https://bugs.webkit.org/show_bug.cgi?id=134518
Reviewed by Mark Hahnenberg.
This has no real effect right now, particularly since almost all uses of
setSetter/setGetter were already allocating a branch new GetterSetter. But once we start
doing more aggressive constant property inference, this change will allow us to remove
all runtime checks from getter/setter calls.
- runtime/GetterSetter.cpp:
(JSC::GetterSetter::withGetter):
(JSC::GetterSetter::withSetter):
- runtime/GetterSetter.h:
(JSC::GetterSetter::setGetter):
(JSC::GetterSetter::setSetter):
- runtime/JSObject.cpp:
(JSC::JSObject::defineOwnNonIndexProperty):
- 3:14 PM Changeset in webkit [170728] by
-
- 3 edits in branches/ftlopt/Source/JavaScriptCore
[ftlopt] Rename notifyTransitionFromThisStructure to didTransitionFromThisStructure
Rubber stamped by Mark Hahnenberg.
- runtime/Structure.cpp:
(JSC::Structure::Structure):
(JSC::Structure::nonPropertyTransition):
(JSC::Structure::didTransitionFromThisStructure):
(JSC::Structure::notifyTransitionFromThisStructure): Deleted.
- runtime/Structure.h:
- 3:12 PM Changeset in webkit [170727] by
-
- 3 edits in trunk/Source/WebCore
Remove keyed coding from FormData
https://bugs.webkit.org/show_bug.cgi?id=134565
Reviewed by Tim Horton.
- platform/network/FormData.cpp:
(WebCore::encodeElement): Deleted.
(WebCore::decodeElement): Deleted.
(WebCore::FormData::encode): Deleted.
(WebCore::FormData::decode): Deleted.
- platform/network/FormData.h:
- 3:04 PM Changeset in webkit [170726] by
-
- 4 edits in trunk/Source/WebCore
Get rid of the HistoryItem encoding/decoding in WebCore
https://bugs.webkit.org/show_bug.cgi?id=134564
Reviewed by Beth Dakin.
- WebCore.exp.in:
- history/HistoryItem.cpp:
(WebCore::HistoryItem::encodeBackForwardTree): Deleted.
(WebCore::encodeRect): Deleted.
(WebCore::encodeSize): Deleted.
(WebCore::HistoryItem::encodeBackForwardTreeNode): Deleted.
(WebCore::DecodeRecursionStackElement::DecodeRecursionStackElement): Deleted.
(WebCore::decodeRect): Deleted.
(WebCore::decodeSize): Deleted.
(WebCore::HistoryItem::decodeBackForwardTree): Deleted.
- history/HistoryItem.h:
- 2:49 PM Changeset in webkit [170725] by
-
- 6 edits2 deletes in trunk/Source/WebKit2
Rip out more dead code
https://bugs.webkit.org/show_bug.cgi?id=134562
Reviewed by Andreas Kling.
- Shared/LegacySessionState.cpp: Removed.
- Shared/LegacySessionState.h: Removed.
- Shared/WebBackForwardListItem.cpp:
(WebKit::WebBackForwardListItem::WebBackForwardListItem): Deleted.
(WebKit::WebBackForwardListItem::backForwardData): Deleted.
(WebKit::WebBackForwardListItem::setBackForwardData): Deleted.
(WebKit::WebBackForwardListItem::encode): Deleted.
(WebKit::WebBackForwardListItem::decode): Deleted.
- Shared/WebBackForwardListItem.h:
(WebKit::WebBackForwardListItem::create): Deleted.
(WebKit::WebBackForwardListItem::setOriginalURL): Deleted.
(WebKit::WebBackForwardListItem::setURL): Deleted.
(WebKit::WebBackForwardListItem::setTitle): Deleted.
- UIProcess/WebPageProxy.cpp:
- WebKit2.xcodeproj/project.pbxproj:
- WebProcess/WebPage/WebPage.cpp:
- 2:35 PM Changeset in webkit [170724] by
-
- 5 edits in branches/ftlopt/Source/JavaScriptCore
[ftlopt] Remove the functionality for cloning StructureRareData since we never do that anymore.
Rubber stamped by Mark Hahnenberg.
- runtime/Structure.cpp:
(JSC::Structure::Structure):
(JSC::Structure::cloneRareDataFrom): Deleted.
- runtime/Structure.h:
- runtime/StructureRareData.cpp:
(JSC::StructureRareData::clone): Deleted.
(JSC::StructureRareData::StructureRareData): Deleted.
- runtime/StructureRareData.h:
(JSC::StructureRareData::needsCloning): Deleted.
- 2:31 PM Changeset in webkit [170723] by
-
- 10 edits in trunk/Source/WebKit2
RestoreSession should take a vector of BackForwardListItemStates
https://bugs.webkit.org/show_bug.cgi?id=134558
Reviewed by Tim Horton.
- Scripts/webkit2/messages.py:
(struct_or_class):
(headers_for_type):
- Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode):
(WebKit::WebPageCreationParameters::decode):
- Shared/WebPageCreationParameters.h:
- UIProcess/WebBackForwardList.cpp:
(WebKit::WebBackForwardList::itemStates):
- UIProcess/WebBackForwardList.h:
- UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::restoreFromSessionState):
(WebKit::WebPageProxy::creationParameters):
- WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::WebPage):
(WebKit::WebPage::restoreSession):
(WebKit::WebPage::restoreSessionAndNavigateToCurrentItem): Deleted.
- WebProcess/WebPage/WebPage.h:
- WebProcess/WebPage/WebPage.messages.in:
- 2:13 PM Changeset in webkit [170722] by
-
- 7 edits in trunk/Source/WebKit2
Store the back forward list item id inside BackForwardListItemState
https://bugs.webkit.org/show_bug.cgi?id=134557
Reviewed by Tim Horton.
- Shared/SessionState.cpp:
(WebKit::BackForwardListItemState::encode):
(WebKit::BackForwardListItemState::decode):
- Shared/SessionState.h:
- Shared/WebBackForwardListItem.cpp:
(WebKit::WebBackForwardListItem::create):
(WebKit::WebBackForwardListItem::WebBackForwardListItem):
(WebKit::WebBackForwardListItem::encode):
- Shared/WebBackForwardListItem.h:
(WebKit::WebBackForwardListItem::itemID):
- UIProcess/WebBackForwardList.cpp:
(WebKit::WebBackForwardList::restoreFromState):
- UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::addBackForwardItem):
- 2:05 PM Changeset in webkit [170721] by
-
- 2 edits in trunk/Source/WebKit2
Unreviewed. EFL and GTK build fix after r170716.
- UIProcess/LegacySessionStateCodingNone.cpp:
(WebKit::decodeLegacySessionState): Update method signature.
- 2:03 PM Changeset in webkit [170720] by
-
- 2 edits in tags/Safari-538.44/Source/WTF
Merged r170713.
- 1:43 PM Changeset in webkit [170719] by
-
- 5 edits1 delete in trunk/Source/WebKit2
Begin ripping out the old session state code
https://bugs.webkit.org/show_bug.cgi?id=134556
Reviewed by Andreas Kling.
- UIProcess/WebBackForwardList.h:
- UIProcess/WebPageProxy.h:
- UIProcess/cf/WebBackForwardListCF.cpp: Removed.
- UIProcess/cf/WebPageProxyCF.cpp:
(WebKit::WebPageProxy::sessionStateData): Deleted.
(WebKit::WebPageProxy::restoreFromSessionStateData): Deleted.
- WebKit2.xcodeproj/project.pbxproj:
- 1:29 PM Changeset in webkit [170718] by
-
- 2 edits in trunk/Source/WebCore
[WebGL] Fix crash when glsl expressions are too complex.
https://bugs.webkit.org/show_bug.cgi?id=134554
<rdar://problem/17540462>
Reviewed by Dean Jackson.
This fixes the Khronos test conformance/glsl/bugs/long-expressions-should-not-crash.html.
- platform/graphics/opengl/Extensions3DOpenGLCommon.cpp:
(WebCore::Extensions3DOpenGLCommon::getTranslatedShaderSourceANGLE):
Add SH_LIMIT_EXPRESSION_COMPLEXITY to the ANGLE compile options.
- 1:22 PM Changeset in webkit [170717] by
-
- 1 edit2 adds in trunk/LayoutTests
REGRESSION(160908): vube.com video won't play after going into and out of fullscreen
https://bugs.webkit.org/show_bug.cgi?id=134489
Reviewed by Zalan Bujtas.
- fullscreen/full-screen-plugin-expected.txt: Added.
- fullscreen/full-screen-plugin.html: Added.
- 1:19 PM Changeset in webkit [170716] by
-
- 8 edits in trunk/Source/WebKit2
Use legacy state coding directly in WKWebView
https://bugs.webkit.org/show_bug.cgi?id=134553
Reviewed by Dan Bernstein.
- UIProcess/API/C/WKPage.cpp:
(WKPageRestoreFromSessionState):
- UIProcess/API/C/WKSessionStateRef.cpp:
(WKSessionStateCreateFromData):
- UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _sessionStateData]):
(-[WKWebView _sessionState]):
(-[WKWebView _restoreFromSessionStateData:]):
(-[WKWebView _restoreFromSessionState:]):
(releaseNSData): Deleted.
- UIProcess/LegacySessionStateCoding.h:
- UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::restoreFromSessionState):
(WebKit::WebPageProxy::restoreFromState): Deleted.
- UIProcess/WebPageProxy.h:
- UIProcess/mac/LegacySessionStateCoding.cpp:
(WebKit::decodeLegacySessionState):
- 12:55 PM Changeset in webkit [170715] by
-
- 2 edits in trunk/Tools
Unreviewed. Moved self to reviewers list.
- Scripts/webkitpy/common/config/contributors.json:
Moved self to reviewers list, added WebGL to expertise.
- 12:48 PM Changeset in webkit [170714] by
-
- 5 edits in trunk/Source/WebCore
WebCore JIT: rename registerCount to something less generic and add new types for lists of registers and lists of stack references
https://bugs.webkit.org/show_bug.cgi?id=134552
Reviewed by Alex Christensen.
Little cleanup:
-The name registerCount was a little too generic. Rename that to "maximumRegisterCount" to avoid confusion.
-Add a new type RegisterVector for any vector holding registers. This is just to avoid repeating the inline
size everywhere, no functional change.
-Same idea for the stack: welcome StackReferenceVector!
- cssjit/FunctionCall.h:
(WebCore::FunctionCall::saveAllocatedCallerSavedRegisters):
Remove the appendVector here. It was unnecessarily cautious, StackAllocator already protect us
from mistakes.
- cssjit/RegisterAllocator.h:
(WebCore::RegisterAllocator::allocatedRegisters):
- cssjit/SelectorCompiler.cpp:
(WebCore::SelectorCompiler::SelectorCodeGenerator::generateSelectorChecker):
Changing from the count of "calleeSavedRegisterCount" to "maximumRegisterCount" will cause
calleeSavedRegisterStackReferences to always overallocate.
The code generator is never on the heap, so that should not change anything.
- cssjit/StackAllocator.h:
(WebCore::StackAllocator::push):
(WebCore::StackAllocator::pop):
- 12:40 PM Changeset in webkit [170713] by
-
- 2 edits in trunk/Source/WTF
iOS build fix after http://trac.webkit.org/changeset/170697.
Unreviewed.
- wtf/Deque.h:
- 11:30 AM Changeset in webkit [170712] by
-
- 2 edits in trunk/LayoutTests
Skip fast/canvas/webgl/tex-image-and-sub-image-2d-with-potentially-subsampled-image.html
on Mountain Lion. It doesn't support such large textures (or at least our
bots don't).
- platform/mac/TestExpectations:
- 10:40 AM Changeset in webkit [170711] by
-
- 3 edits in trunk/Source/WebKit2
Add -[WKWebView _sessionStateData] and -[WKWebView _restoreFromSessionStateData:]
https://bugs.webkit.org/show_bug.cgi?id=134549
Reviewed by Dan Bernstein.
- UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _sessionState]):
(-[WKWebView _sessionStateData]):
(-[WKWebView _restoreFromSessionState:]):
(-[WKWebView _restoreFromSessionStateData:]):
- UIProcess/API/Cocoa/WKWebViewPrivate.h:
- 10:33 AM Changeset in webkit [170710] by
-
- 16 edits in trunk/Source
Remove some code only needed for versions of Safari that are no longer supported
https://bugs.webkit.org/show_bug.cgi?id=134537
Reviewed by Anders Carlsson.
Source/WebCore:
- bindings/objc/DOMHTML.mm:
(-[DOMHTMLInputElement _rectOnScreen]): Deleted.
(-[DOMHTMLInputElement _replaceCharactersInRange:withString:selectingFromIndex:]): Deleted.
(-[DOMHTMLInputElement _selectedRange]): Deleted.
- bindings/objc/DOMPrivate.h:
- editing/EditorCommand.cpp:
(WebCore::createCommandMap):
Source/WebKit/mac:
- History/WebHistoryItem.mm:
- Misc/WebKitVersionChecks.h:
- WebCoreSupport/WebSecurityOrigin.mm:
(-[WebSecurityOrigin domain]): Deleted.
- WebInspector/WebInspector.mm:
(-showTimeline:): Deleted.
(+webInspector): Deleted.
(-setWebFrame:): Deleted.
(-window): Deleted.
(-showWindow:): Deleted.
- WebView/WebFrame.mm:
(-[WebFrame reload]):
- WebView/WebPDFView.mm:
(-[WebPDFView menuForEvent:]):
(-[WebPDFView _anyPDFTagsFoundInMenu:]): Deleted.
- WebView/WebView.mm:
(-[WebView _initWithFrame:frameName:groupName:]):
(-[WebView _preferencesChanged:]):
(-[WebView initWithFrame:frameName:groupName:]):
(-[WebView _initWithFrame:frameName:groupName:usesDocumentViews:]): Deleted.
(+[WebView _setAlwaysUseATSU:]): Deleted.
(-[WebView _needsUnrestrictedGetMatchedCSSRules]): Deleted.
- WebView/WebViewPrivate.h:
Source/WTF:
- wtf/Assertions.cpp:
- wtf/Assertions.h:
- 9:02 AM Changeset in webkit [170709] by
-
- 2 edits in trunk/Source/WebCore
[GStreamer] MediaPlayerPrivateGStreamer.cpp fails to build when VIDEO_TRACK is enabled and using GSTREAMER_MPEGTS
https://bugs.webkit.org/show_bug.cgi?id=134548
Reviewed by Philippe Normand.
Include AtomicStringHash.h when ENABLE(VIDEO_TRACK) && USE(GSTREAMER_MPEGTS)
since it's used by HashMap<AtomicString, RefPtr<InbandMetadataTextTrackPrivateGStreamer>>
- platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.h:
- 8:24 AM Changeset in webkit [170708] by
-
- 2 edits in trunk/Source/WebCore
[GTK] Unstable headers should include webkitdomdefines-unstable.h
https://bugs.webkit.org/show_bug.cgi?id=134540
Reviewed by Martin Robinson.
Instead of the class header that can only be included from webkitdom.h.
- bindings/scripts/CodeGeneratorGObject.pm:
(WriteData):
- 8:22 AM Changeset in webkit [170707] by
-
- 3 edits in trunk/Source/WebKit2
[GTK] WebKitVersion.h should be shared between UI and Web Process APIs
https://bugs.webkit.org/show_bug.cgi?id=134538
Reviewed by Martin Robinson.
To be able to check current version from web extensions too.
- UIProcess/API/gtk/WebKitVersion.h.in: Allow to include WebKitVersion.h also from webkit-web-extension.h.
- WebProcess/InjectedBundle/API/gtk/webkit-web-extension.h: Include WebKitVersion.h.
- 8:13 AM Changeset in webkit [170706] by
-
- 2 edits in trunk/Tools
[GTK] Make dist only works when run from the source directory
https://bugs.webkit.org/show_bug.cgi?id=134543
Reviewed by Martin Robinson.
Since the paths in the manifest are relative to the source dir,
and the script expects the current working directory to be the
source directory, resolve all passed in relative paths while
parsing the arguments and then change the current working
directory.
- gtk/make-dist.py:
- 8:11 AM Changeset in webkit [170705] by
-
- 5 edits in trunk
[GTK] make dist is broken
https://bugs.webkit.org/show_bug.cgi?id=134542
Reviewed by Martin Robinson.
.:
- Source/PlatformGTK.cmake: Remove ENABLE_WEBKIT check.
Tools:
- gtk/make-dist.py: Rename command line options to match what cmake uses.
- gtk/manifest.txt: Remove WebKit1 references and add WebKitDOM documentation.
- 4:03 AM WebKitGTK/KeepingTheTreeGreen edited by
- (diff)
- 2:52 AM Changeset in webkit [170704] by
-
- 4 edits in trunk
REGRESSION(r170008): [GTK] Layout Tests fast/forms/option-constructor-selected.html and fast/forms/select-live-pseudo-selectors.html crash.
https://bugs.webkit.org/show_bug.cgi?id=134399
Reviewed by Chris Fleizach.
Source/WebCore:
Check that the value of optionIndex passed as a parameter is valid
before calling didUpdateActiveOption() over the item of the menu
list, to prevent crashes in platforms where renderers (and therefore
accessibility objects) might not be created for items at the time
of calling AccessibilityMenuList::didUpdateActiveOption().
- accessibility/AccessibilityMenuList.cpp:
(WebCore::AccessibilityMenuList::didUpdateActiveOption): Added check.
LayoutTests:
Removing Crash expectations for two layout tests.
- platform/gtk/TestExpectations: Updated.
- 1:22 AM Changeset in webkit [170703] by
-
- 1 copy in tags/Safari-538.44
New Tag.
- 12:19 AM Changeset in webkit [170702] by
-
- 22 edits3 adds in trunk
[GTK] WebKitWebView::create should receive information about the navigation action
https://bugs.webkit.org/show_bug.cgi?id=133680
Reviewed by Martin Robinson.
Source/WebKit2:
Add WebKitNavigationAction boxed type representing a navigation
action to be passed as parameter of WebKitWebView::create
signal. In the future it could be also for the
WebKitNavigationPolicyDecision class. The class contains
information about a navigation action like the navigation type,
the request, the mouse button and key modifiers and whether it was
started by a user gesture. This information is needed to be able
to implement a proper popup blocker. Since all that information is
not provided by the C API, we have switched the UI client to use a
custom client derived from API::UIClient. This also avoids a lot
of toAPI -> toImpl unneeded conversions.
- PlatformGTK.cmake: Add new files to compilation.
- UIProcess/API/gtk/WebKitHitTestResult.cpp:
(webkitHitTestResultCreate): Use a WebHitTestResult::Data const
reference instead of a pointer to a WebHitTestResult.
(webkitHitTestResultCompare): Ditto.
- UIProcess/API/gtk/WebKitHitTestResultPrivate.h:
- UIProcess/API/gtk/WebKitNavigationAction.cpp: Added.
(webkitNavigationActionCreate): Create a WebKitNavigationAction
for the given ResourceRequest and NavigationActionData.
(webkit_navigation_action_copy):
(webkit_navigation_action_free):
(webkit_navigation_action_get_navigation_type):
(webkit_navigation_action_get_mouse_button):
(webkit_navigation_action_get_modifiers):
(webkit_navigation_action_get_request):
(webkit_navigation_action_is_user_gesture):
- UIProcess/API/gtk/WebKitNavigationAction.h: Added.
- UIProcess/API/gtk/WebKitNavigationActionPrivate.h: Added.
- UIProcess/API/gtk/WebKitNavigationPolicyDecision.h:
- UIProcess/API/gtk/WebKitPrivate.cpp:
(toGdkModifiers):
(toWebKitNavigationType):
(toWebKitMouseButton):
- UIProcess/API/gtk/WebKitPrivate.h:
- UIProcess/API/gtk/WebKitUIClient.cpp:
(UIClient::UIClient): Implement the UIClient using a custom class.
(attachUIClientToView): Set the UIClient directly to the WebPageProxy.
(createNewPage): Deleted.
(showPage): Deleted.
(closePage): Deleted.
(runJavaScriptAlert): Deleted.
(runJavaScriptConfirm): Deleted.
(runJavaScriptPrompt): Deleted.
(toolbarsAreVisible): Deleted.
(setToolbarsAreVisible): Deleted.
(menuBarIsVisible): Deleted.
(setMenuBarIsVisible): Deleted.
(statusBarIsVisible): Deleted.
(setStatusBarIsVisible): Deleted.
(isResizable): Deleted.
(setIsResizable): Deleted.
(getWindowFrame): Deleted.
(setWindowFrame): Deleted.
(mouseDidMoveOverElement): Deleted.
(printFrame): Deleted.
(runOpenPanel): Deleted.
(decidePolicyForGeolocationPermissionRequest): Deleted.
(runModal): Deleted.
- UIProcess/API/gtk/WebKitWebView.cpp:
(webkitWebViewCreate): Add WebKitNavigationAction parameter.
(webkit_web_view_class_init): Add WebKitNavigationAction paramter
to CREATE signal.
(webkitWebViewCreateNewPage): Add WebKitNavigationAction parameter
and pass it to the signal callbacks.
(webkitWebViewMouseTargetChanged): Use a WebHitTestResult::Data
const reference.
(webkitWebViewPopulateContextMenu): Ditto.
- UIProcess/API/gtk/WebKitWebView.h:
- UIProcess/API/gtk/WebKitWebViewPrivate.h:
- UIProcess/API/gtk/WebKitWindowProperties.cpp:
(webkitWindowPropertiesUpdateFromWebWindowFeatures): Use a
WebCore::WindowFeatures const reference.
- UIProcess/API/gtk/WebKitWindowPropertiesPrivate.h:
- UIProcess/API/gtk/docs/webkit2gtk-docs.sgml: Add new section.
- UIProcess/API/gtk/docs/webkit2gtk-sections.txt: Add new symbols.
- UIProcess/API/gtk/webkit2.h: Include WebKitNavigationAction.h.
Tools:
Add unit test to check the WebKitNavigationAction passed to
WebKitWebView::create callback.
- MiniBrowser/gtk/BrowserWindow.c:
(webViewCreate): Add WebKitNavigationAction parameter.
- TestWebKitAPI/Tests/WebKit2Gtk/TestMultiprocess.cpp:
- TestWebKitAPI/Tests/WebKit2Gtk/TestPrinting.cpp:
- TestWebKitAPI/Tests/WebKit2Gtk/TestUIClient.cpp:
(testWebViewCreateNavigationData):
(beforeAll):
- TestWebKitAPI/Tests/WebKit2Gtk/TestWebKitPolicyClient.cpp:
(createCallback):
- 12:09 AM Changeset in webkit [170701] by
-
- 8 edits in trunk
[GTK] Do not add forward declaration of class struct for interfaces
https://bugs.webkit.org/show_bug.cgi?id=134297
Reviewed by Martin Robinson.
Source/WebCore:
- PlatformGTK.cmake: Make files generated by
gobject-generate-headers.pl script depend on it.
- bindings/gobject/WebKitDOMEventTarget.h: Remove forward declaration.
- bindings/gobject/WebKitDOMNodeFilter.h: Ditto
- bindings/gobject/WebKitDOMXPathNSResolver.h: Ditto.
- bindings/scripts/gobject-generate-headers.pl: Use Iface instead
of Class suffix for interfaces.
Tools:
Add Iface symbol to documentation instead of Class for interfaces.
- gtk/webkitdom.py:
(WebKitDOMDocGeneratorSections.write_section):
- 12:04 AM Changeset in webkit [170700] by
-
- 7 edits7 adds in trunk
[GTK] XPathNSResolver should be exposed as an interface instead of as a class
https://bugs.webkit.org/show_bug.cgi?id=134028
Reviewed by Gustavo Noronha Silva.
Source/WebCore:
Add custom implementation of WebKitDOMXPathNSResolver to expose it
as an interface instead of a class. It also includes an internal
default implementation of the interface to support methods that
create a XPathNSResolver using the native implementation.
- PlatformGTK.cmake: Add new files to compilation.
- bindings/gobject/GObjectXPathNSResolver.cpp: Added.
(WebCore::GObjectXPathNSResolver::~GObjectXPathNSResolver):
(WebCore::GObjectXPathNSResolver::lookupNamespaceURI): Call
webkit_dom_xpath_ns_resolver_lookup_namespace_uri().
- bindings/gobject/GObjectXPathNSResolver.h: Added.
(WebCore::GObjectXPathNSResolver::create):
(WebCore::GObjectXPathNSResolver::GObjectXPathNSResolver): Create
a new GObjectXPathNSResolver for the given WebKitDOMXPathNSResolver.
- bindings/gobject/WebKitDOMXPathNSResolver.cpp: Added.
(webkit_dom_xpath_ns_resolver_default_init):
(webkit_dom_xpath_ns_resolver_lookup_namespace_uri):
(webkitDOMNativeXPathNSResolverFinalize):
(webkitDOMNativeXPathNSResolverConstructor):
(webkit_dom_native_xpath_ns_resolver_init):
(webkit_dom_native_xpath_ns_resolver_class_init):
(webkitDOMNativeXPathNSResolverLookupNamespaceURI):
(webkitDOMXPathNSResolverIfaceInit):
(WebKit::core): Return the native XPathNSResolver when using the
internal implementation or create a custom XPathNSResolver when
using a user provided WebKitDOMXPathNSResolver.
(WebKit::kit):
- bindings/gobject/WebKitDOMXPathNSResolver.h: Added.
- bindings/gobject/WebKitDOMXPathNSResolverPrivate.h: Added.
- bindings/scripts/CodeGeneratorGObject.pm:
(GenerateFunction): Add exceptions for XPathNSResolver since
the core method returns a PassRefPtr.
Tools:
Add unit tests to check WebKitDOMXPathNSResolver API when using the
native one and a user provided one.
- Scripts/webkitpy/style/checker.py: Add WebKitDOMXPathNSResolver
to the list of GTK+ public API exceptions.
- TestWebKitAPI/Tests/WebKit2Gtk/CMakeLists.txt: Add new files to compilation.
- TestWebKitAPI/Tests/WebKit2Gtk/DOMXPathNSResolverTest.cpp: Added.
(webkitXPathNSResolverLookupNamespaceURI):
(webkitXPathNSResolverDOMXPathNSResolverIfaceInit):
(webkit_xpath_ns_resolver_init):
(webkit_xpath_ns_resolver_class_init):
(WebKitDOMXPathNSResolverTest::create):
(WebKitDOMXPathNSResolverTest::webPageFromArgs):
(WebKitDOMXPathNSResolverTest::evaluateFooChildTextAndCheckResult):
(WebKitDOMXPathNSResolverTest::testXPathNSResolverNative):
(WebKitDOMXPathNSResolverTest::testXPathNSResolverCustom):
(WebKitDOMXPathNSResolverTest::runTest):
(registerTests):
- TestWebKitAPI/Tests/WebKit2Gtk/TestDOMXPathNSResolver.cpp: Added.
(runTest):
(testWebKitDOMXPathNSResolverNative):
(testWebKitDOMXPathNSResolverCustom):
(serverCallback):
(beforeAll):
(afterAll):
- gtk/webkitdom.py:
(WebKitDOMDocGeneratorSections.write_section): Add
WebKitDOMXPathNSResolver to the list of interfaces.
Jul 1, 2014:
- 11:44 PM Changeset in webkit [170699] by
-
- 6 edits4 deletes in trunk
The OpenType MATH table is not read on Mac
https://bugs.webkit.org/show_bug.cgi?id=133569
Reviewed by Chris Fleizach.
Source/WebCore:
An incorrect tag format is used to load the OpenType MATH table on Mac.
This commit makes the format consistent with what is used for other tables on Mac.
This will allow to load the MATH table and to use it for the MathML rendering.
No new tests. platform/graphics/opentype already contains tests for the OpenType MATH support.
- platform/graphics/opentype/OpenTypeMathData.cpp:
LayoutTests:
We update the references for some tests in LayoutTests/mathml/opentype/ and remove the failure from TestExpectations.
Other tests require a font with a MATH table to be installed on Mac bots and to be used by default (bug 133603).
- platform/mac-wk2/mathml/opentype/opentype-stretchy-expected.txt: Removed.
- platform/mac/TestExpectations: Enable mathml/opentype/opentype-stretchy-horizontal.html and mathml/opentype/opentype-stretchy.html.
- platform/mac/mathml/horizontal-LatinModern-expected.txt: Removed. This file should not have been in this directory.
- platform/mac/mathml/horizontal-LatinModern-munderover-expected.txt: Removed. This file should not have been in this directory.
- platform/mac/mathml/opentype-stretchy-horizontal-expected.txt: Removed. This file should not have been in this directory.
- platform/mac/mathml/opentype/opentype-stretchy-expected.txt: Add text reference.
- platform/mac/mathml/opentype/opentype-stretchy-horizontal-expected.txt: Add text reference.
- 10:57 PM Changeset in webkit [170698] by
-
- 5 edits in trunk/Source
[Win64] Unreviewed build fix.
Source/WebCore:
- WebCore.vcxproj/WebCore.vcxproj:
Exclude CA files from 32 and 64 bit WinCairo builds.
- WebCore.vcxproj/WebCoreTestSupport.vcxproj:
Don't use a precompiled header in 32 or 63 bit builds of MockCDM.cpp.
Source/WebKit:
- WebKit.vcxproj/WebKitExportGenerator/WebKitExports.def.in:
Corrected 64-bit linker symbols.
- 9:49 PM Changeset in webkit [170697] by
-
- 7 edits in trunk/Source
Reduce dynamic memory allocation in css jit.
https://bugs.webkit.org/show_bug.cgi?id=134416
Reviewed by Benjamin Poulain.
Source/WebCore:
- cssjit/FunctionCall.h:
(WebCore::FunctionCall::saveAllocatedRegisters):
- cssjit/RegisterAllocator.h:
- cssjit/SelectorCompiler.cpp:
(WebCore::SelectorCompiler::addPseudoClassType):
(WebCore::SelectorCompiler::SelectorCodeGenerator::generateSelectorChecker):
(WebCore::SelectorCompiler::SelectorCodeGenerator::generateElementIsNthChild):
- cssjit/StackAllocator.h:
(WebCore::StackAllocator::push):
(WebCore::StackAllocator::pop):
Use inlineCapacity template parameter to reduce malloc calls.
I use 32 as the inline capacity for non-register related vectors because that
is probably big enough for the selector compiler, and if it is not then a malloc
call will not be significant because the selector is very complex.
Source/WTF:
- wtf/Deque.h:
Added inlineCapacity optional template parameter.
- 9:49 PM Changeset in webkit [170696] by
-
- 2 edits in trunk/Source/WebKit2
[EFL] Fix occurrence of two cursors on WK2 EFL.
https://bugs.webkit.org/show_bug.cgi?id=134436
Patch by KwangHyuk Kim <hyuki.kim@samsung.com> on 2014-07-01
Reviewed by Anders Carlsson.
ecore_x_window_cursor is released when ecore_evas_object_cursor should be shown in order to avoid occurrence of two cursors.
- UIProcess/API/efl/EwkView.cpp:
(EwkView::updateCursor):
- 8:44 PM Changeset in webkit [170695] by
-
- 4 edits in trunk
Fix build break on EFL and GTK ports since r170683
https://bugs.webkit.org/show_bug.cgi?id=134536
Unreviewed, build fix.
Source/WebKit2:
- CMakeLists.txt:
Tools:
- TestWebKitAPI/Tests/WebKit2/RestoreSessionStateContainingFormData.cpp:
- 8:35 PM Changeset in webkit [170694] by
-
- 24 edits in trunk
WebVTT percentage value can be a float
https://bugs.webkit.org/show_bug.cgi?id=134511
Reviewed by Brent Fulgham.
Source/WebCore:
No new tests, existing tests modified.
- html/track/TextTrackCueGeneric.cpp:
(WebCore::TextTrackCueGeneric::setLine): int -> double.
(WebCore::TextTrackCueGeneric::setPosition): Ditto.
- html/track/TextTrackCueGeneric.h:
- html/track/VTTCue.cpp:
(WebCore::VTTCue::setLine): Ditto.
(WebCore::VTTCue::setPosition): Ditto.
(WebCore::VTTCue::setCueSettings): Support floating point values for line, position, and size.
(WebCore::scanPercentage): Deleted.
- html/track/VTTCue.h:
- html/track/VTTScanner.cpp:
(WebCore::VTTScanner::scanFloat): Support negative values.
- html/track/VTTScanner.h:
- html/track/WebVTTParser.cpp: Include parseFloatPercentageValue in non-WEBVTT_REGIONS builds.
- html/track/WebVTTParser.h:
- platform/graphics/avfoundation/InbandTextTrackPrivateAVF.cpp:
(WebCore::InbandTextTrackPrivateAVF::processCueAttributes): Drive by change to take a cue
reference rather than a pointer.
(WebCore::InbandTextTrackPrivateAVF::processAttributedStrings): Ditto.
(WebCore::InbandTextTrackPrivateAVF::removeCompletedCues): Loop over all cues, don't LOG that
we are asking the client to delete the cue because the client will have already deleted
it when status when to GenericCueData::Complete.
- platform/graphics/avfoundation/InbandTextTrackPrivateAVF.h:
LayoutTests:
- media/track/captions-webvtt/tc015-positioning-ltr.vtt:
- media/track/captions-webvtt/tc015-positioning.vtt:
- media/track/captions-webvtt/tc017-line-position.vtt:
- media/track/captions-webvtt/tc018-align-text-line-position.vtt:
- media/track/captions-webvtt/tc019-cue-size.vtt:
- media/track/track-webvtt-tc015-positioning-expected.txt:
- media/track/track-webvtt-tc015-positioning.html:
- media/track/track-webvtt-tc017-line-position-expected.txt:
- media/track/track-webvtt-tc017-line-position.html:
- media/track/track-webvtt-tc018-align-text-line-position-expected.txt:
- media/track/track-webvtt-tc018-align-text-line-position.html:
- media/video-test.js:
(testCues):
- 8:21 PM Changeset in webkit [170693] by
-
- 2 edits in trunk/Source/WebKit2
Make LogsPageMessagesToSystemConsoleEnabled a global debug preference
https://bugs.webkit.org/show_bug.cgi?id=134534
Reviewed by Joseph Pecoraro.
- Shared/WebPreferencesDefinitions.h:
- 8:06 PM Changeset in webkit [170692] by
-
- 2 edits in trunk/Source/WebKit2
Handle invalid data more gracefully.
Reviewed by Anders Carlsson.
- UIProcess/mac/LegacySessionStateCoding.cpp:
(WebKit::decodeFormData):
(WebKit::decodeBackForwardTreeNode):
- 7:50 PM Changeset in webkit [170691] by
-
- 2 edits in trunk/LayoutTests
Fix this test; it was waiting about 32 years for the end of an animation.
What we really care about is the start of the animation.
- compositing/animation/keyframe-order.html:
- 7:43 PM Changeset in webkit [170690] by
-
- 2 edits in trunk/Source/WebCore
Improve behavior of position:sticky on zoomed pages
https://bugs.webkit.org/show_bug.cgi?id=134535
<rdar://problem/12898829>
Reviewed by Dean Jackson.
Despite r138036, position:sticky still had incorrect behavior in zoomed pages.
Fix by removing the scaling of the rect from viewportConstrainedVisibleContentRect(),
since all the other math happens in unscaled coordinates.
- rendering/RenderBoxModelObject.cpp:
(WebCore::RenderBoxModelObject::constrainingRectForStickyPosition):
- 7:24 PM Changeset in webkit [170689] by
-
- 2 edits in trunk/Source/WebKit2
Add a missing return statement in WKPageCopySessionState.
Reviewed by Anders Carlsson.
- UIProcess/API/C/WKPage.cpp:
(WKPageCopySessionState):
- 6:36 PM Changeset in webkit [170688] by
-
- 2 edits in trunk/Source/WebKit/mac
Remove WebChromeClient code that was only needed for Safari versions earlier than 3.1
https://bugs.webkit.org/show_bug.cgi?id=134531
Reviewed by Daniel Bates.
- WebCoreSupport/WebChromeClient.mm:
(WebChromeClient::exceededDatabaseQuota):
- 6:27 PM Changeset in webkit [170687] by
-
- 2 edits in trunk/Source/WebKit2
Don't encode/decode the snapshot UUID
https://bugs.webkit.org/show_bug.cgi?id=134532
Reviewed by Sam Weinig.
- UIProcess/mac/LegacySessionStateCoding.cpp:
(WebKit::encodeSessionHistory):
(WebKit::decodeSessionHistoryEntry):
- 6:24 PM Changeset in webkit [170686] by
-
- 10 edits in trunk/Source/WebKit2
Unreviewed, rolling out r170608.
https://bugs.webkit.org/show_bug.cgi?id=134533
Safari keeps crashing on device due to sandbox violation.
(Requested by enrica on #webkit).
Reverted changeset:
"Restrict network process sandbox"
https://bugs.webkit.org/show_bug.cgi?id=134360
http://trac.webkit.org/changeset/170608
- 5:57 PM Changeset in webkit [170685] by
-
- 6 edits in trunk
Encode/decode CFURLRequestRefs when USE(CFNETWORK) is enabled
https://bugs.webkit.org/show_bug.cgi?id=134454
<rdar://problem/17510980>
Reviewed by Andreas Kling.
Source/WebKit2:
Use new helper methods to serialize/deserialize CFURLRequestRef directly so we can avoid
creating NSURLRequest.
- Shared/mac/WebCoreArgumentCodersMac.mm:
(IPC::ArgumentCoder<ResourceRequest>::encodePlatformData):
(IPC::ArgumentCoder<ResourceRequest>::decodePlatformData):
WebKitLibraries:
- WebKitSystemInterface.h:
- libWebKitSystemInterfaceMavericks.a:
- libWebKitSystemInterfaceMountainLion.a:
- 5:49 PM Changeset in webkit [170684] by
-
- 2 edits in trunk/Source/WebKit2
[iOS][WK2] Fix a race between the short tap and long tap highlight
https://bugs.webkit.org/show_bug.cgi?id=134530
Reviewed by Enrica Casucci.
There was a potential race of event that can theoretically cause WKContentViewInteraction
to call [WKContentView _showTapHighlight] after all interactions have been cancelled.
The race would be like this:
1) On a short tap, _singleTapRecognized: is called, a tap highlight ID is defined and
_potentialTapInProgress is set to YES.
2) For some reason, the gesture is cancelled. The method _singleTapDidReset is called,
setting _potentialTapInProgress but leaving the tap highlight ID as valid.
3) The UIProcess receives the tap highlight information from the WebProcess, _didGetTapHighlightForRequest:
has a valid ID, _potentialTapInProgress is false -> the highlight is shown right away as if a long tap
was in progress.
The missing piece that causes this is _singleTapDidReset: must also invalidate the tap highlight ID. This is done
in the new static function cancelPotentialTapIfNecessary().
Just invalidating the ID would create another race:
1) Short tap gesture recognizer starts.
2) The long press recognizer starts before (1) is commited.
3) The long press recognizers sets up its own tap highlight ID.
4) The short tap gesture recognizer resets, erasing the tap highlight ID defined in (3).
To avoid this, the long press gesture recognizers immediately cancels any potential tap in progress.
If _singleTapDidReset: is called before (3), this does nothing. If the reset is called after (3),
_singleTapDidReset does nothing.
- UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _highlightLongPressRecognized:]):
(cancelPotentialTapIfNecessary):
(-[WKContentView _singleTapDidReset:]):
- 5:47 PM Changeset in webkit [170683] by
-
- 8 edits in trunk
Add ABI hacks to allow WKPageRef to use WKSessionStateRef
https://bugs.webkit.org/show_bug.cgi?id=134529
Reviewed by Sam Weinig.
- UIProcess/API/APISessionState.h:
Add a sessionState() getter.
- UIProcess/API/C/WKPage.cpp:
(WKPageCopySessionState):
Add a hack where we return a WKSessionStateRef if the least significant bit of the context pointer
is 1, and a WKDataRef otherwise.
(WKPageRestoreFromSessionState):
Handle both WKDataRef and WKSessionStateref.
- UIProcess/API/C/WKPage.h:
Use WKTypeRefs for state saving and restoration.
- UIProcess/API/C/WKSessionStateRef.cpp:
(WKSessionStateCopyData):
Add helper function.
- 5:43 PM Changeset in webkit [170682] by
-
- 2 edits in trunk/Source/WebCore
[WebGL] Fix PVRTC extension.
https://bugs.webkit.org/show_bug.cgi?id=134525
Patch by Alex Christensen <achristensen@webkit.org> on 2014-07-01
Reviewed by Dean Jackson.
- html/canvas/WebGLRenderingContext.cpp:
(WebCore::WebGLRenderingContext::validateCompressedTexDimensions):
Remove the return false which I mistakenly committed in r170128
and make the gl error match the S3TC error and the khronos test.
- 5:13 PM Changeset in webkit [170681] by
-
- 3 edits in trunk/Source/WebInspectorUI
Web Inspector: Hexadecimal color values in all CSS editors are purple when they should be blue
https://bugs.webkit.org/show_bug.cgi?id=133997
Reviewed by Joseph Pecoraro.
CodeMirror 4 removed the style for hex colors from the CSS tokenizer.
Feature has been added to CodeMirrorAdditions.js.
- UserInterface/Views/CodeMirrorAdditions.js: Added hex-color to style for a valid hexadecimal color.
- UserInterface/Views/SyntaxHighlightingDefaultTheme.css: Added selector for hexadecimal color.
- 5:12 PM Changeset in webkit [170680] by
-
- 17 edits1 delete in branches/ftlopt/Source
[ftlopt] DebuggerCallFrame::scope() should return a DebuggerScope.
<https://webkit.org/b/134420>
Reviewed by Geoffrey Garen.
Source/JavaScriptCore:
Previously, DebuggerCallFrame::scope() returns a JSActivation (and relevant
peers) which the WebInspector will use to introspect CallFrame variables.
Instead, we should be returning a DebuggerScope as an abstraction layer that
provides the introspection functionality that the WebInspector needs. This
is the first step towards not forcing every frame to have a JSActivation
object just because the debugger is enabled.
- Instantiate the debuggerScopeStructure as a member of the JSGlobalObject instead of the VM. This allows JSObject::globalObject() to be able to return the global object for the DebuggerScope.
- On the DebuggerScope's life-cycle management:
The DebuggerCallFrame is designed to be "valid" only during a debugging session
(while the debugger is broken) through the use of a DebuggerCallFrameScope in
Debugger::pauseIfNeeded(). Once the debugger resumes from the break, the
DebuggerCallFrameScope destructs, and the DebuggerCallFrame will be invalidated.
We can't guarantee (from this code alone) that the Inspector code isn't still
holding a ref to the DebuggerCallFrame (though they shouldn't), but by contract,
the frame will be invalidated, and any attempt to query it will return null values.
This is pre-existing behavior.
Now, we're adding the DebuggerScope into the picture. While a single debugger
pause session is in progress, the Inspector may request the scope from the
DebuggerCallFrame. While the DebuggerCallFrame is still valid, we want
DebuggerCallFrame::scope() to always return the same DebuggerScope object.
This is why we hold on to the DebuggerScope with a strong ref.
If we use a weak ref instead, the following cooky behavior can manifest:
- The Inspector calls Debugger::scope() to get the top scope.
- The Inspector iterates down the scope chain and is now only holding a reference to a parent scope. It is no longer referencing the top scope.
- A GC occurs, and the DebuggerCallFrame's weak m_scope ref to the top scope gets cleared.
- The Inspector calls DebuggerCallFrame::scope() to get the top scope again but gets a different DebuggerScope instance.
- The Inspector iterates down the scope chain but never sees the parent scope instance that retained a ref to in step 2 above. This is because when iterating this new DebuggerScope instance (which has no knowledge of the previous parent DebuggerScope instance), a new DebuggerScope instance will get created for the same parent scope.
Since the DebuggerScope is a JSObject, it's liveness is determined by its reachability.
However, it's "validity" is determined by the life-cycle of its owner DebuggerCallFrame.
When the owner DebuggerCallFrame gets invalidated, its debugger scope chain (if
instantiated) will also get invalidated. This is why we need the
DebuggerScope::invalidateChain() method. The Inspector should not be using the
DebuggerScope instance after its owner DebuggerCallFrame is invalidated. If it does,
those methods will do nothing or returned a failed status.
- debugger/Debugger.h:
- debugger/DebuggerCallFrame.cpp:
(JSC::DebuggerCallFrame::scope):
(JSC::DebuggerCallFrame::evaluate):
(JSC::DebuggerCallFrame::invalidate):
(JSC::DebuggerCallFrame::vm):
(JSC::DebuggerCallFrame::lexicalGlobalObject):
- debugger/DebuggerCallFrame.h:
- debugger/DebuggerScope.cpp:
(JSC::DebuggerScope::DebuggerScope):
(JSC::DebuggerScope::finishCreation):
(JSC::DebuggerScope::visitChildren):
(JSC::DebuggerScope::className):
(JSC::DebuggerScope::getOwnPropertySlot):
(JSC::DebuggerScope::put):
(JSC::DebuggerScope::deleteProperty):
(JSC::DebuggerScope::getOwnPropertyNames):
(JSC::DebuggerScope::defineOwnProperty):
(JSC::DebuggerScope::next):
(JSC::DebuggerScope::invalidateChain):
(JSC::DebuggerScope::isWithScope):
(JSC::DebuggerScope::isGlobalScope):
(JSC::DebuggerScope::isFunctionScope):
- debugger/DebuggerScope.h:
(JSC::DebuggerScope::create):
(JSC::DebuggerScope::Iterator::Iterator):
(JSC::DebuggerScope::Iterator::get):
(JSC::DebuggerScope::Iterator::operator++):
(JSC::DebuggerScope::Iterator::operator==):
(JSC::DebuggerScope::Iterator::operator!=):
(JSC::DebuggerScope::isValid):
(JSC::DebuggerScope::jsScope):
(JSC::DebuggerScope::begin):
(JSC::DebuggerScope::end):
- inspector/JSJavaScriptCallFrame.cpp:
(Inspector::JSJavaScriptCallFrame::scopeType):
(Inspector::JSJavaScriptCallFrame::scopeChain):
- inspector/JavaScriptCallFrame.h:
(Inspector::JavaScriptCallFrame::scopeChain):
- inspector/ScriptDebugServer.cpp:
- runtime/JSGlobalObject.cpp:
(JSC::JSGlobalObject::reset):
(JSC::JSGlobalObject::visitChildren):
- runtime/JSGlobalObject.h:
(JSC::JSGlobalObject::debuggerScopeStructure):
- runtime/JSObject.h:
(JSC::JSObject::isWithScope):
- runtime/JSScope.h:
- runtime/VM.cpp:
(JSC::VM::VM):
- runtime/VM.h:
Source/WebCore:
No new tests.
- ForwardingHeaders/debugger/DebuggerCallFrame.h: Removed.
- This is not in use. Hence, we can remove it.
- bindings/js/ScriptController.cpp:
(WebCore::ScriptController::attachDebugger):
- We should acquire the JSLock before modifying a JS global object.
- 5:08 PM Changeset in webkit [170679] by
-
- 4 edits in trunk/Source/WebKit2
WKPageRestoreFromSessionState should use the new session state restore code path
https://bugs.webkit.org/show_bug.cgi?id=134526
Reviewed by Tim Horton.
- UIProcess/API/C/WKPage.cpp:
(WKPageRestoreFromSessionState):
Use the new code path.
- UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::restoreFromState):
Only restore the back-forward state if we have a back-forward list.
- UIProcess/mac/LegacySessionStateCoding.cpp:
(WebKit::encodeSessionHistory):
Save the original URL string as well.
(WebKit::decodeV1SessionHistory):
If we don't have a current index, set it to Nullopt instead of 0.
- 4:57 PM Changeset in webkit [170678] by
-
- 3 edits2 adds in trunk
AX: iOS does not expose ARIA toggle buttons at all (not even as a regular button)
https://bugs.webkit.org/show_bug.cgi?id=133875
Reviewed by Enrica Casucci.
Source/WebCore:
Support the toggle button type in iOS.
Test: platform/ios-sim/accessibility/toggle-button.html
- accessibility/ios/WebAccessibilityObjectWrapperIOS.mm:
(-[WebAccessibilityObjectWrapper accessibilityCanFuzzyHitTest]):
(-[WebAccessibilityObjectWrapper accessibilityTraits]):
(-[WebAccessibilityObjectWrapper determineIsAccessibilityElement]):
LayoutTests:
- platform/ios-sim/accessibility/toggle-button-expected.txt: Added.
- platform/ios-sim/accessibility/toggle-button.html: Added.
- 4:40 PM Changeset in webkit [170677] by
-
- 4 edits in trunk/Source/JavaScriptCore
Debugger's breakpoint list should not be a Vector.
<https://webkit.org/b/134514>
Reviewed by Geoffrey Garen.
The debugger currently stores breakpoint data as entries in a Vector (see
BreakpointsInLine). It also keeps a fast map look up of breakpoint IDs to
the breakpoint data (see m_breakpointIDToBreakpoint). Because a Vector can
compact or reallocate its backing store, this can causes all sorts of havoc.
The m_breakpointIDToBreakpoint map assumes that the breakpoint data doesn't
move in memory.
The fix is to replace the BreakpointsInLine Vector with a BreakpointsList
doubly linked list.
- debugger/Breakpoint.h:
(JSC::Breakpoint::Breakpoint):
(JSC::BreakpointsList::~BreakpointsList):
- debugger/Debugger.cpp:
(JSC::Debugger::setBreakpoint):
(JSC::Debugger::removeBreakpoint):
(JSC::Debugger::hasBreakpoint):
- debugger/Debugger.h:
- 4:40 PM Changeset in webkit [170676] by
-
- 2 edits in trunk/Source/WebKit2
Unreviewed. EFL and GTK build fix since r170654.
- UIProcess/LegacySessionStateCodingNone.cpp: Add a dummy encodeLegacySessionState(const SessionState&).
(WebKit::encodeLegacySessionState):
- 4:26 PM Changeset in webkit [170675] by
-
- 6 edits6 adds in trunk
[iOS] Subsampled JPEG images do not draw correctly via the canvas APIs
https://bugs.webkit.org/show_bug.cgi?id=134513
<rdar://problem/12078860>
<rdar://problem/16745393>
Reviewed by Tim Horton.
Source/WebCore:
Subsampled images (e.g. JPEG) were not consistently using
their original dimensions and subsampled dimensions. This caused
things like texImage2D to pack the pixels incorrectly, or drawImage
to squish the rendering.
Renamed m_scale to m_subsamplingScale on FrameData.
Tests: fast/canvas/image-potential-subsample.html
fast/canvas/webgl/tex-image-and-sub-image-2d-with-potentially-subsampled-image.html
- platform/graphics/BitmapImage.cpp:
(WebCore::BitmapImage::cacheFrame): Rename to m_subsamplingScale.
(WebCore::BitmapImage::frameAtIndex): Ditto.
- platform/graphics/BitmapImage.h:
(WebCore::FrameData::FrameData): Ditto.
- platform/graphics/cg/BitmapImageCG.cpp:
(WebCore::FrameData::clear): Ditto.
(WebCore::BitmapImage::BitmapImage): Ditto.
(WebCore::BitmapImage::draw): Use a scaledSrcRect that reflects the subsampled size,
rather than assuming the srcRect accurately reflects how many pixels we have
in the Bitmap.
(WebCore::BitmapImage::copyUnscaledFrameAtIndex):
- platform/graphics/cg/GraphicsContext3DCG.cpp:
(WebCore::GraphicsContext3D::ImageExtractor::extractImage): Similar fix, although this
time we just ask the image decoder to take into account the subsampled size
when it is "generating" a frame, causing it to use the bitmap data it has already
decoded.
LayoutTests:
Add Canvas2D and WebGL tests that exercise a very large JPEG image.
The WebGL test is mostly copied from the WebGL test suite, so please
excuse the coding style.
- fast/canvas/image-potential-subsample-expected.txt: Added.
- fast/canvas/image-potential-subsample.html: Added.
- fast/canvas/resources/image-8000x8000.jpg: Added.
- fast/canvas/webgl/resources/tex-image-and-sub-image-2d-with-potentially-subsampled-image.js: Added.
(.init):
(.runOneIteration):
(.runTestOnImage):
(.runTest):
(generateTest):
- fast/canvas/webgl/tex-image-and-sub-image-2d-with-potentially-subsampled-image-expected.txt: Added.
- fast/canvas/webgl/tex-image-and-sub-image-2d-with-potentially-subsampled-image.html: Added.
- 4:22 PM Changeset in webkit [170674] by
-
- 2 edits in trunk/Source/WebKit2
Crash in WebDocumentLoader::setNavigationID
https://bugs.webkit.org/show_bug.cgi?id=134520
Reviewed by Anders Carlsson.
- WebProcess/WebPage/WebFrame.cpp:
(WebKit::WebFrame::didReceivePolicyDecision): Null-check the policy document loader.
- 4:01 PM Changeset in webkit [170673] by
-
- 2 edits in trunk/Source/WebKit2
Build fix.
- UIProcess/API/APILoaderClient.h:
(API::LoaderClient::didDestroyNavigation):
- 3:40 PM Changeset in webkit [170672] by
-
- 15 edits9 adds in branches/ftlopt/Source/JavaScriptCore
[ftlopt] DFG bytecode parser should turn PutById with nothing but a Setter stub as stuff+handleCall, and handleCall should be allowed to inline if it wants to
https://bugs.webkit.org/show_bug.cgi?id=130756
Reviewed by Oliver Hunt.
The enables exposing the call to setters in the DFG, and then inlining it. Previously we
already supproted inlined-cached calls to setters from within put_by_id inline caches,
and the DFG could certainly emit such IC's. Now, if an IC had a setter call, then the DFG
will either emit the GetGetterSetterByOffset/GetSetter/Call combo, or it will do one
better and inline the call.
A lot of the core functionality was already available from the previous work to inline
getters. So, there are some refactorings in this patch that move preexisting
functionality around. For example, the work to figure out how the DFG should go about
getting to what we call the "loaded value" - i.e. the GetterSetter object reference in
the case of accessors - is now shared in ComplexGetStatus, and both GetByIdStatus and
PutByIdStatus use it. This means that we can keep the safety checks common. This patch
also does additional refactorings in DFG::ByteCodeParser so that we can continue to reuse
handleCall() for all of the various kinds of calls we can now emit.
83% speed-up on getter-richards, 2% speed-up on box2d.
- CMakeLists.txt:
- JavaScriptCore.vcxproj/JavaScriptCore.vcxproj:
- JavaScriptCore.xcodeproj/project.pbxproj:
- bytecode/ComplexGetStatus.cpp: Added.
(JSC::ComplexGetStatus::computeFor):
- bytecode/ComplexGetStatus.h: Added.
(JSC::ComplexGetStatus::ComplexGetStatus):
(JSC::ComplexGetStatus::skip):
(JSC::ComplexGetStatus::takesSlowPath):
(JSC::ComplexGetStatus::kind):
(JSC::ComplexGetStatus::attributes):
(JSC::ComplexGetStatus::specificValue):
(JSC::ComplexGetStatus::offset):
(JSC::ComplexGetStatus::chain):
- bytecode/GetByIdStatus.cpp:
(JSC::GetByIdStatus::computeForStubInfo):
- bytecode/GetByIdVariant.cpp:
(JSC::GetByIdVariant::GetByIdVariant):
- bytecode/PolymorphicPutByIdList.h:
(JSC::PutByIdAccess::PutByIdAccess):
(JSC::PutByIdAccess::setter):
(JSC::PutByIdAccess::structure):
(JSC::PutByIdAccess::chainCount):
- bytecode/PutByIdStatus.cpp:
(JSC::PutByIdStatus::computeFromLLInt):
(JSC::PutByIdStatus::computeFor):
(JSC::PutByIdStatus::computeForStubInfo):
(JSC::PutByIdStatus::makesCalls):
- bytecode/PutByIdStatus.h:
(JSC::PutByIdStatus::makesCalls): Deleted.
- bytecode/PutByIdVariant.cpp:
(JSC::PutByIdVariant::PutByIdVariant):
(JSC::PutByIdVariant::operator=):
(JSC::PutByIdVariant::replace):
(JSC::PutByIdVariant::transition):
(JSC::PutByIdVariant::setter):
(JSC::PutByIdVariant::writesStructures):
(JSC::PutByIdVariant::reallocatesStorage):
(JSC::PutByIdVariant::makesCalls):
(JSC::PutByIdVariant::dumpInContext):
- bytecode/PutByIdVariant.h:
(JSC::PutByIdVariant::PutByIdVariant):
(JSC::PutByIdVariant::structure):
(JSC::PutByIdVariant::oldStructure):
(JSC::PutByIdVariant::alternateBase):
(JSC::PutByIdVariant::specificValue):
(JSC::PutByIdVariant::callLinkStatus):
(JSC::PutByIdVariant::replace): Deleted.
(JSC::PutByIdVariant::transition): Deleted.
- dfg/DFGByteCodeParser.cpp:
(JSC::DFG::ByteCodeParser::addCallWithoutSettingResult):
(JSC::DFG::ByteCodeParser::addCall):
(JSC::DFG::ByteCodeParser::handleCall):
(JSC::DFG::ByteCodeParser::handleInlining):
(JSC::DFG::ByteCodeParser::handleGetById):
(JSC::DFG::ByteCodeParser::handlePutById):
(JSC::DFG::ByteCodeParser::parseBlock):
- jit/Repatch.cpp:
(JSC::tryCachePutByID):
(JSC::tryBuildPutByIdList):
- runtime/IntendedStructureChain.cpp:
(JSC::IntendedStructureChain::takesSlowPathInDFGForImpureProperty):
- runtime/IntendedStructureChain.h:
- tests/stress/exit-from-setter.js: Added.
- tests/stress/poly-chain-setter.js: Added.
(Cons):
(foo):
(test):
- tests/stress/poly-chain-then-setter.js: Added.
(Cons1):
(Cons2):
(foo):
(test):
- tests/stress/poly-setter-combo.js: Added.
(Cons1):
(Cons2):
(foo):
(test):
(.test):
- tests/stress/poly-setter-then-self.js: Added.
(foo):
(test):
(.test):
- tests/stress/weird-setter-counter.js: Added.
(foo):
(test):
- tests/stress/weird-setter-counter-syntactic.js: Added.
(foo):
(test):
- 3:34 PM Changeset in webkit [170671] by
-
- 3 edits in trunk/Source/WebKit2
Use the new legacy session state coding in WKPageCopySessionState
https://bugs.webkit.org/show_bug.cgi?id=134516
Reviewed by Tim Horton.
- UIProcess/API/C/WKPage.cpp:
(WKPageCopySessionState):
Call encodeLegacySessionState.
- UIProcess/mac/LegacySessionStateCoding.cpp:
(WebKit::encodeSessionHistoryEntryData):
Use leakPtr() so we won't attempt to double-free the buffer.
(WebKit::createDictionary):
Fix a typo.
(WebKit::encodeLegacySessionState):
Don't encode the provisional URL if it's empty.
- 3:25 PM Changeset in webkit [170670] by
-
- 2 edits in trunk/Source/WebKit2
AX: [iOS WebKit2] Support for Speak selection
https://bugs.webkit.org/show_bug.cgi?id=134512
Reviewed by Darin Adler.
Implement a method that Speak Selection can use to retrieve the selected content.
- UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView accessibilityRetrieveSpeakSelectionContent]):
- 3:19 PM Changeset in webkit [170669] by
-
- 2 edits in trunk/Source/WebKit2
Update the highest back-forward list item ID in the other WebBackForwardListItem constructor
https://bugs.webkit.org/show_bug.cgi?id=134515
Reviewed by Tim Horton.
- Shared/WebBackForwardListItem.cpp:
(WebKit::WebBackForwardListItem::WebBackForwardListItem):
- 2:21 PM Changeset in webkit [170668] by
-
- 2 edits in trunk/Source/WebKit2
Fix ServicesOverlayController teardown
<rdar://problem/17523998> and https://bugs.webkit.org/show_bug.cgi?id=134510
Reviewed by Mark Rowe.
- WebProcess/WebPage/mac/ServicesOverlayController.mm:
(WebKit::ServicesOverlayController::~ServicesOverlayController): Flipped logic on null-checking the overlay.
- 2:16 PM Changeset in webkit [170667] by
-
- 7 edits in trunk
[Cocoa][Mac] WKWebView should automatically set topContentInset when beneath a toolbar
<rdar://problem/17523159>
https://bugs.webkit.org/show_bug.cgi?id=134496
Reviewed by Beth Dakin.
Source/WebKit2:
- Adds _automaticallyAdjustsContentInsets SPI to both WKView and WKWebView.
- Defaults _automaticallyAdjustsContentInsets to NO for WKView to keep legacy clients working.
- Defaults _automaticallyAdjustsContentInsets to YES for WKWebView.
- UIProcess/API/Cocoa/WKViewPrivate.h:
Add _automaticallyAdjustsContentInsets SPI.
- UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView initWithFrame:configuration:]):
Default _automaticallyAdjustsContentInsets to YES for WKWebView.
(-[WKWebView _setTopContentInset:]):
(-[WKWebView _topContentInset]):
Change to forward to the WKView, since we now do coalescing and caching there.
(-[WKWebView _setAutomaticallyAdjustsContentInsets:]):
(-[WKWebView _automaticallyAdjustsContentInsets]):
Forward to the WKView.
- UIProcess/API/Cocoa/WKWebViewPrivate.h:
Add _automaticallyAdjustsContentInsets SPI.
- UIProcess/API/mac/WKView.mm:
(-[WKView renewGState]):
Ensure the contentInset is updated when the view changes, as its relative location to the toolbar may have changed.
(-[WKView addWindowObserversForWindow:]):
(-[WKView removeWindowObservers]):
(-[WKView observeValueForKeyPath:ofObject:change:context:]):
Add observers for the contentLayoutRect and titlebarAppearsTransparent properties on the window. If either of them
change, we need to recalculate the content inset.
(-[WKView _updateContentInsetsIfAutomatic]):
If _automaticallyAdjustsContentInsets is YES, follow the same rules as AppKit in setting the top content inset. Like
AppKit this is only done when:
- The window's style mask has the NSFullSizeContentViewWindowMask bit.
- The window does not have titlebarAppearsTransparent set.
- And we are not in an enclosing NSScrollView.
(-[WKView _setTopContentInset:]):
(-[WKView _topContentInset]):
Coalesce setting the topContentInset to avoid a visual lag when resizing the window that was caused by us sending
multiple topContentInsets to the WebContentProcess per runloop cycle. The reason for the lag was that at the time
we observe the contentLayoutRect changing, our view may not yet be in its final position yet, so the inset will be
temporarily wrong. When the view is finally positioned correctly, we will get a renewGState, at which point we will
again calculate the inset, which now will be correct. Since these both happen in the same runloop iteration, we can
just defer sending the inset to the WebContentProcess.
(-[WKView _setAutomaticallyAdjustsContentInsets:]):
(-[WKView _automaticallyAdjustsContentInsets]):
Add the new SPI.
Tools:
- MiniBrowser/mac/BrowserWindowController.m:
(-[BrowserWindowController windowDidLoad]):
Use the NSFullSizeContentViewWindowMask bit on Yosemite and later to enable toolbar blurring.
- 2:13 PM Changeset in webkit [170666] by
-
- 1 delete in trunk/Source/WebKit2/UIProcess/WebPageProxy.cpp.rej
Oops, didn't mean to commit this.
- 2:12 PM Changeset in webkit [170665] by
-
- 2 edits in trunk/Source/WebCore
Web Inspector: Selected DOM element highlights invisible near bottom of the viewport (topContentInset?)
https://bugs.webkit.org/show_bug.cgi?id=133818
Patch by Joseph Pecoraro <Joseph Pecoraro> on 2014-07-01
Reviewed by Beth Dakin.
- inspector/InspectorOverlay.cpp:
(WebCore::InspectorOverlay::update):
Take the topContentInset into account when sizing the InspectorOverlay's FrameView.
- 2:07 PM Changeset in webkit [170664] by
-
- 6 edits1 add in trunk/Source/WebKit2
Add a function for restoring page state given a SessionState object
https://bugs.webkit.org/show_bug.cgi?id=134509
Reviewed by Tim Horton.
- Shared/WebBackForwardListItem.h:
(WebKit::WebBackForwardListItem::itemState):
- UIProcess/WebBackForwardList.cpp:
(WebKit::WebBackForwardList::saveState):
- UIProcess/WebBackForwardList.h:
- UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::saveState):
- UIProcess/WebPageProxy.h:
- 1:45 PM Changeset in webkit [170663] by
-
- 7 edits in trunk
[CMake] Add necessary support for building for the Wayland target
https://bugs.webkit.org/show_bug.cgi?id=134160
Reviewed by Darin Adler.
.:
- Source/cmake/FindGTK3.cmake: Check for the gtk+-x11-3.0 package if the X11 target
is enabled, and that its version matches the version of the generic package. Same
for the gtk+-wayland-3.0 package and the Wayland target.
- Source/cmake/OptionsCommon.cmake: Don't add the -DXP_UNIX definition here.
- Source/cmake/OptionsEfl.cmake: Add -DXP_UNIX here for the EFL port, under the same
condition as in OptionsCommon.cmake. ENABLE_X11_TARGET is enabled by default for EFL
so they keep building the TestNetscapePlugin target.
- Source/cmake/OptionsGTK.cmake: Enable the X11 target and disable the Wayland target
by default. Group all the X11-specific macro definitions (WTF_PLATFORM_X11, MOZ_X11,
XP_UNIX) and only add them if the X11 target is enabled. Similar goes for the Wayland
target and WTF_PLATFORM_WAYLAND. The Wayland target requires GTK+ 3.12.0, so that's
the minimum required version as soon as that target is enabled. Only search for GLX
if the X11 target is enabled.
Tools:
- CMakeLists.txt: Only build the TestNetscapePlugIn target in case
the X11 windowing target is enabled.
- 1:38 PM Changeset in webkit [170662] by
-
- 2 edits in trunk/Source/WebKit2
Followup for: Combine the Telephone and Selection overlay controllers, updating UI behavior.
https://bugs.webkit.org/show_bug.cgi?id=134461
Rubberstamped by Tim Horton.
- WebProcess/WebPage/mac/ServicesOverlayController.mm: Include the correct constant.
- 1:29 PM Changeset in webkit [170661] by
-
- 8 edits1 copy in trunk
AX: HTML indeterminate IDL attribute not mapped to checkbox value=2
https://bugs.webkit.org/show_bug.cgi?id=134492
Reviewed by Andreas Kling.
Source/WebCore:
Support the indeterminate attribute in AX code.
Test: Update existing test: accessibility/aria-checked-mixed-value
- accessibility/AccessibilityObject.cpp:
(WebCore::AccessibilityObject::checkboxOrRadioValue):
Tools:
- DumpRenderTree/mac/AccessibilityUIElementMac.mm:
(AccessibilityUIElement::isIndeterminate):
- WebKitTestRunner/InjectedBundle/mac/AccessibilityUIElementMac.mm:
(WTR::AccessibilityUIElement::isIndeterminate):
LayoutTests:
- accessibility/aria-checked-mixed-value-expected.txt:
- accessibility/aria-checked-mixed-value.html:
- platform/mac/accessibility/aria-checked-mixed-value-expected.txt: Added.
- 1:22 PM Changeset in webkit [170660] by
-
- 21 edits in trunk/Source/WebKit2
[Cocoa] Navigation parameter in navigation delegate messages is nil for navigations started by the Web Content process
https://bugs.webkit.org/show_bug.cgi?id=134482
Reviewed by Tim Horton.
- UIProcess/API/APILoaderClient.h:
(API::LoaderClient::didDestroyNavigation): New client function, called when a navigation ID
is not going to be used anymore.
- UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _restoreFromSessionState:]): If restoring triggered a navigation, create a
WKNavigation for its ID.
- UIProcess/Cocoa/NavigationState.h:
- UIProcess/Cocoa/NavigationState.mm:
(WebKit::NavigationState::createLoadRequestNavigation): Removed FIXME, because we now remove
navigations from the map in LoaderClient::didDestroyNavigation.
(WebKit::NavigationState::createBackForwardNavigation): Ditto.
(WebKit::NavigationState::createReloadNavigation): Ditto.
(WebKit::NavigationState::createLoadDataNavigation): Ditto.
(WebKit::NavigationState::PolicyClient::decidePolicyForNavigationAction): If we are only
getting a navigation ID now, create a WKNavigation for it. Also changed to create the
NSURLRequest lazily.
(WebKit::NavigationState::LoaderClient::didStartProvisionalLoadForFrame): Assert that
navigationID is not zero and that it maps to an existing WKNavigation.
(WebKit::NavigationState::LoaderClient::didReceiveServerRedirectForProvisionalLoadForFrame):
Ditto.
(WebKit::NavigationState::LoaderClient::didFailProvisionalLoadWithErrorForFrame): Ditto.
(WebKit::NavigationState::LoaderClient::didCommitLoadForFrame): Ditto.
(WebKit::NavigationState::LoaderClient::didFinishDocumentLoadForFrame): Ditto.
(WebKit::NavigationState::LoaderClient::didFinishLoadForFrame): Ditto.
(WebKit::NavigationState::LoaderClient::didFailLoadWithErrorForFrame): Ditto.
(WebKit::NavigationState::LoaderClient::didDestroyNavigation): Override the new client
function to remove the navigation from the map.
(WebKit::NavigationState::LoaderClient::processDidCrash): Clear the navigations map.
- UIProcess/WebFrameListenerProxy.cpp:
(WebKit::WebFrameListenerProxy::WebFrameListenerProxy): Initialize new m_navigationID
member variable.
(WebKit::WebFrameListenerProxy::receivedPolicyDecision): Pass the navigation ID to
WebFrame::receivedPolicyDecision.
- UIProcess/WebFrameListenerProxy.h:
(WebKit::WebFrameListenerProxy::navigationID): Added this accessor.
(WebKit::WebFrameListenerProxy::setNavigationID): Ditto.
- UIProcess/WebFrameProxy.cpp:
(WebKit::WebFrameProxy::receivedPolicyDecision): Added navigationID parameter, which is
passed along to the WebPageProxy.
- UIProcess/WebFrameProxy.h:
- UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::receivedPolicyDecision): Added navigationID parameter, which is
passed along in the message to the Web Content process.
(WebKit::WebPageProxy::restoreFromSessionStateData): Changed to return a navigation ID of 0.
(WebKit::WebPageProxy::restoreFromState): Changed to return a navigation ID if one was
started.
(WebKit::WebPageProxy::didDestroyNavigation): Added. Calls the new client function.
(WebKit::WebPageProxy::decidePolicyForNavigationAction): Added navigationID and
newNavigationID parameters. If a main-frame navigation doesn’t already have an ID, assign it
a new ID and return it in the newNavigationID parmeter as well as setting it on the listener
for the asynchronous case.
- UIProcess/WebPageProxy.h:
- UIProcess/WebPageProxy.messages.in: Added navigationID and newNavigationID parameters.
- UIProcess/cf/WebPageProxyCF.cpp:
(WebKit::WebPageProxy::restoreFromSessionStateData): Changed to return a navigation ID if
one was started.
- WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::dispatchDecidePolicyForResponse): Updated for additional
parameter.
(WebKit::WebFrameLoaderClient::dispatchDecidePolicyForNavigationAction): Send the navigation
ID to the UI process, get the new navigation ID from the reply, and set it on the document
loader.
- WebProcess/WebPage/WebDocumentLoader.cpp:
(WebKit::WebDocumentLoader::detachFromFrame): Override to let the UI process know that the
navigation ID will not be used anymore.
- WebProcess/WebPage/WebDocumentLoader.h:
- WebProcess/WebPage/WebFrame.cpp:
(WebKit::WebFrame::didReceivePolicyDecision): Added navigationID parameter, set it on the
document loader.
(WebKit::WebFrame::documentLoaderDetached): Pass the message along to the UI process.
- WebProcess/WebPage/WebFrame.h:
- WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::didReceivePolicyDecision): Pass new navigationID parameter along.
- WebProcess/WebPage/WebPage.h:
- WebProcess/WebPage/WebPage.messages.in: Added navigationID parameter.
- 12:43 PM Changeset in webkit [170659] by
-
- 2 edits in trunk/Source/WebKit2
Fix build.
- WebKit2.xcodeproj/project.pbxproj:
- 12:42 PM Changeset in webkit [170658] by
-
- 2 edits in trunk/Source/WebKit2
Temporarily encode and decode the snapshot UUID
https://bugs.webkit.org/show_bug.cgi?id=134507
Reviewed by Tim Horton.
- UIProcess/mac/LegacySessionStateCoding.cpp:
(WebKit::encodeSessionHistory):
(WebKit::decodeSessionHistoryEntry):
(WebKit::decodeSessionHistoryEntries):
- 12:34 PM Changeset in webkit [170657] by
-
- 3 edits in trunk/Source/WebKit2
Implement encodeSessionHistoryEntryData
https://bugs.webkit.org/show_bug.cgi?id=134505
Reviewed by Tim Horton.
- UIProcess/API/C/WKPage.cpp:
(WKPageCopySessionState):
Add back a missing null check for the filter pointer.
- UIProcess/mac/LegacySessionStateCoding.cpp:
(WebKit::encodeSessionHistoryEntryData):
- 12:15 PM Changeset in webkit [170656] by
-
- 7 edits2 adds in trunk
[UI-side compositing] Bad spinner on news.google.com: animations need to be ordered
https://bugs.webkit.org/show_bug.cgi?id=134504
<rdar://problem/17507892>
Reviewed by Tim Horton.
Source/WebKit2:
The layer's addedAnimations property needs to maintain order, since the order
in which transforms are applied is important.
- Shared/mac/RemoteLayerTreeTransaction.h: Use a Vector<pair<>> for addedAnimations.
- Shared/mac/RemoteLayerTreeTransaction.mm:
(WebKit::dumpChangedLayers):
- WebProcess/WebPage/mac/PlatformCAAnimationRemote.h:
- WebProcess/WebPage/mac/PlatformCAAnimationRemote.mm:
(WebKit::PlatformCAAnimationRemote::updateLayerAnimations):
- WebProcess/WebPage/mac/PlatformCALayerRemote.cpp:
(WebKit::PlatformCALayerRemote::addAnimationForKey): If this is a new entry, we
can just append to addedAnimations, otherwise we have to find the existing one
and update its properties.
(WebKit::PlatformCALayerRemote::removeAnimationForKey): Do linear search to
find the animation to remove (this list will normally be short).
LayoutTests:
Test that the transforms from the animation are applied in the correct order.
- compositing/animation/keyframe-order-expected.html: Added.
- compositing/animation/keyframe-order.html: Added.
- 12:10 PM Changeset in webkit [170655] by
-
- 2 edits in trunk/Source/WebCore
Typing an automatic text replacement phrase after a br in contenteditable is not rendered as expected
https://bugs.webkit.org/show_bug.cgi?id=133883
Reviewed by Enrica Casucci.
enclosingDeletableElement() makes sure that the element's container is editable, but not
that it is capable of having content inside it (like how a <br> can't).
No new tests. I don't think this is testable because it requires setting some state in System Preferences.
- editing/DeleteButtonController.cpp:
(WebCore::enclosingDeletableElement):
- 12:04 PM Changeset in webkit [170654] by
-
- 3 edits in trunk/Source/WebKit2
Add a encodeLegacySessionState function
https://bugs.webkit.org/show_bug.cgi?id=134502
Reviewed by Tim Horton.
- UIProcess/LegacySessionStateCoding.h:
- UIProcess/mac/LegacySessionStateCoding.cpp:
(WebKit::encodeSessionHistoryEntryData):
(WebKit::encodeLegacySessionHistoryEntryData):
(WebKit::createDictionary):
(WebKit::encodeSessionHistory):
(WebKit::encodeLegacySessionState):
- 11:41 AM Changeset in webkit [170653] by
-
- 9 edits in trunk/Source/WebKit2
[Cocoa] WebProcess doesn't follow localization of UI process when run as a service
https://bugs.webkit.org/show_bug.cgi?id=133126
<rdar://problem/13396515>
Patch by Alexey Proskuryakov.
Reviewed by Tim Horton.
- Configurations/WebContentService.Development.xcconfig:
- Configurations/WebContentService.xcconfig:
Figure out which of the two keys to use. CFBundleAllowMixedLocalizations was
a workaround before we got CFBundleFollowParentLocalization, and these cannot be
used together.
- UIProcess/Launcher/mac/ProcessLauncherMac.mm: (WebKit::connectToService): Set up
XPC bootstrap for CFBundle to use.
- WebKit2.xcodeproj/project.pbxproj: Added a script that applies the key to Info.plist.
- WebProcess/EntryPoint/mac/XPCService/WebContentService.Development/Info-OSX.plist:
- WebProcess/EntryPoint/mac/XPCService/WebContentService.Development/Info-iOS.plist:
- WebProcess/EntryPoint/mac/XPCService/WebContentService/Info-OSX.plist:
- WebProcess/EntryPoint/mac/XPCService/WebContentService/Info-iOS.plist:
Removed CFBundleAllowMixedLocalizations, we now add it programmatically.
- 11:29 AM Changeset in webkit [170652] by
-
- 6 edits in trunk/Source
[iOS] Unreviewed build fix after r170640.
Source/WebCore:
- page/EditorClient.h:
(WebCore::EditorClient::selectedTelephoneNumberRangesChanged):
Add ifdefs to make everything compile and link.
Source/WebKit2:
- WebProcess/WebCoreSupport/WebEditorClient.cpp:
(WebKit::WebEditorClient::selectedTelephoneNumberRangesChanged):
(WebKit::WebEditorClient::selectionRectsDidChange):
- WebProcess/WebCoreSupport/WebEditorClient.h:
- WebProcess/WebPage/WebPage.h:
Add ifdefs to make everything compile and link.
- 11:00 AM Changeset in webkit [170651] by
-
- 3 edits in trunk/Source/WebKit2
Add WKSessionStateGetTypeID
https://bugs.webkit.org/show_bug.cgi?id=134499
Reviewed by Tim Horton.
- UIProcess/API/C/WKSessionStateRef.cpp:
(WKSessionStateGetTypeID):
- UIProcess/API/C/WKSessionStateRef.h:
- 10:49 AM Changeset in webkit [170650] by
-
- 13 edits in trunk
Remove unnecessary calls to std::move()
https://bugs.webkit.org/show_bug.cgi?id=134493
Reviewed by Anders Carlsson.
Source/WebCore:
- Modules/encryptedmedia/CDM.cpp:
(WebCore::CDM::CDM):
- css/CSSGrammar.y.in:
- css/CSSPrimitiveValue.cpp:
(WebCore::CSSPrimitiveValue::formatNumberValue):
- css/MediaQuery.cpp:
(WebCore::MediaQuery::MediaQuery):
- platform/graphics/mac/FontMac.mm:
(WebCore::Font::dashesForIntersectionsWithRect):
Source/WebKit2:
- Shared/ShareableResource.cpp:
(WebKit::ShareableResource::Handle::tryWrapInCFData):
- WebProcess/WebPage/WebFrame.cpp:
(WebKit::WebFrame::certificateInfo): Return a WebCore::CertificateInfo instead
of const WebCore::CertificateInfo& to avoid returning a dangling reference to
a local variable. Also, remove the unnecessary calls to std::move() as the compiler
will invoke the move constructor for WebCore::CertificateInfo on return from the
function.
- WebProcess/WebPage/WebFrame.h:
Tools:
- TestWebKitAPI/Tests/WTF/HashSet.cpp:
(TestWebKitAPI::TEST):
- TestWebKitAPI/Tests/WTF/ns/RetainPtr.mm:
(TestWebKitAPI::TEST):
- 10:46 AM Changeset in webkit [170649] by
-
- 8 edits in trunk/Source/WebKit2
Add a BackForwardListItemState struct and put the snapshot UUID there
https://bugs.webkit.org/show_bug.cgi?id=134497
Reviewed by Sam Weinig.
- Shared/SessionState.cpp:
(WebKit::PageState::encode):
(WebKit::PageState::decode):
(WebKit::BackForwardListItemState::encode):
(WebKit::BackForwardListItemState::decode):
- Shared/SessionState.h:
- Shared/WebBackForwardListItem.cpp:
(WebKit::WebBackForwardListItem::create):
(WebKit::WebBackForwardListItem::WebBackForwardListItem):
(WebKit::WebBackForwardListItem::backForwardData):
(WebKit::WebBackForwardListItem::setBackForwardData):
(WebKit::WebBackForwardListItem::encode):
- Shared/WebBackForwardListItem.h:
(WebKit::WebBackForwardListItem::setPageState):
(WebKit::WebBackForwardListItem::setOriginalURL):
(WebKit::WebBackForwardListItem::originalURL):
(WebKit::WebBackForwardListItem::setURL):
(WebKit::WebBackForwardListItem::url):
(WebKit::WebBackForwardListItem::setTitle):
(WebKit::WebBackForwardListItem::title):
(WebKit::WebBackForwardListItem::setSnapshotUUID):
(WebKit::WebBackForwardListItem::snapshotUUID):
- UIProcess/WebBackForwardList.cpp:
(WebKit::WebBackForwardList::restoreFromState):
- UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::addBackForwardItem):
- UIProcess/mac/LegacySessionStateCoding.cpp:
(WebKit::decodeSessionHistoryEntries):
- 10:33 AM Changeset in webkit [170648] by
-
- 2 edits in trunk/Source/WebKit/ios
[iOS] Compile fix with pointer lock enabled.
https://bugs.webkit.org/show_bug.cgi?id=134472
Reviewed by Pratik Solanki.
- WebView/WebPDFViewPlaceholder.mm:
(-[WebPDFViewPlaceholder simulateClickOnLinkToURL:]):
Add parameters required for pointer lock.
- 10:29 AM Changeset in webkit [170647] by
-
- 2 edits in trunk/Tools
Unreviewed. Adding myself to the reviewers list to make commit bot happy.
- Scripts/webkitpy/common/config/contributors.json:
- 10:25 AM Changeset in webkit [170646] by
-
- 3 edits2 adds in trunk
Subpixel rendering: Pixel crack in breadcrumbs at devforums.apple.com.
https://bugs.webkit.org/show_bug.cgi?id=134491
Reviewed by Simon Fraser.
Do not early round geometry values, while computing background image position. It changes the final
subpixel values and that could lead to wrong snap positions.
pixelSnapBackgroundImageGeometryForPainting() takes care of geometry snapping.
Source/WebCore:
Test: fast/backgrounds/background-image-size-double-rounding.html
- rendering/RenderBoxModelObject.cpp:
(WebCore::RenderBoxModelObject::calculateBackgroundImageGeometry):
LayoutTests:
- fast/backgrounds/background-image-size-double-rounding-expected.html: Added.
- fast/backgrounds/background-image-size-double-rounding.html: Added.
- 10:14 AM Changeset in webkit [170645] by
-
- 2 edits in trunk/Source/WebKit2
Add encodeInteger and decodeInteger for remote object encoder and decoder.
https://bugs.webkit.org/show_bug.cgi?id=134453
In 64bit system, encodeInteger/decodeInteger is by default mapped into encodeInt64/decodeInt64. For
32bit system, they are mapped to encodeInt32/decodeInt32. However, since we don't have encodeInt32/decodeInt32
implementation in WKRemoteObjectCoder, we could hit crash in 32bit systems.
Patch by Yongjun Zhang <yongjun_zhang@apple.com> on 2014-07-01
Reviewed by Sam Weinig.
- Shared/API/Cocoa/WKRemoteObjectCoder.mm:
(-[WKRemoteObjectEncoder encodeInteger:forKey:]): Instead of creating a new number type UInt32, use encodeInt64 for
encoding NSInteger.
(-[WKRemoteObjectDecoder decodeIntegerForKey:]):
- 10:04 AM Changeset in webkit [170644] by
-
- 8 edits in trunk/Source/WebKit2
Use an std::function for filtering session state data
https://bugs.webkit.org/show_bug.cgi?id=134481
Reviewed by Sam Weinig.
- UIProcess/API/C/WKPage.cpp:
(WKPageCopySessionState):
- UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _sessionState]):
- UIProcess/WebBackForwardList.h:
- UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::sessionStateData):
- UIProcess/WebPageProxy.h:
- UIProcess/cf/WebBackForwardListCF.cpp:
(WebKit::WebBackForwardList::createCFDictionaryRepresentation):
- UIProcess/cf/WebPageProxyCF.cpp:
(WebKit::WebPageProxy::sessionStateData):
- 9:59 AM Changeset in webkit [170643] by
-
- 2 edits in trunk/Source/WebKit2
Clean up the WebKit build from unused parameter warning in Webkit2/UIProcess module
https://bugs.webkit.org/show_bug.cgi?id=134294
Patch by Rohit Kumar <kumar.rohit@samsung.com> on 2014-07-01
Reviewed by Darin Adler.
- UIProcess/API/APIUIClient.h:
(API::UIClient::reachedApplicationCacheOriginQuota):
- 9:45 AM Changeset in webkit [170642] by
-
- 4 edits in trunk/Source/WebCore
Create NSURLRequest lazily when USE(CFNETWORK) is enabled
https://bugs.webkit.org/show_bug.cgi?id=134441
Reviewed by Andreas Kling.
No new tests. Should be covered by existing tests.
- platform/network/cf/ResourceRequest.h:
(WebCore::ResourceRequest::ResourceRequest):
(WebCore::ResourceRequest::encodingRequiresPlatformData):
- platform/network/cf/ResourceRequestCFNet.cpp:
(WebCore::ResourceRequest::doUpdatePlatformRequest):
(WebCore::ResourceRequest::doUpdatePlatformHTTPBody):
(WebCore::ResourceRequest::setStorageSession):
- platform/network/cocoa/ResourceRequestCocoa.mm:
(WebCore::ResourceRequest::nsURLRequest):
- 9:43 AM Changeset in webkit [170641] by
-
- 2 edits in trunk/Source/WebKit2
Build fix.
Forgot to commit locally before pushing.
- WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::servicesOverlayController): s/this/*this/
- 9:22 AM Changeset in webkit [170640] by
-
- 9 edits1 move1 add5 deletes in trunk/Source
Combine the Telephone and Selection overlay controllers, updating UI behavior.
https://bugs.webkit.org/show_bug.cgi?id=134461
Reviewed by Tim Horton.
Source/WebCore:
- WebCore.exp.in:
- dom/Range.cpp:
(WebCore::Range::contains): Returns true if the Range completely contains the passed-in Range.
- dom/Range.h:
Source/WebKit2:
- WebKit2.xcodeproj/project.pbxproj:
- WebProcess/WebCoreSupport/WebEditorClient.cpp:
(WebKit::WebEditorClient::selectedTelephoneNumberRangesChanged):
(WebKit::WebEditorClient::selectionRectsDidChange):
Remove the separate Selection overlay controller:
- WebProcess/WebPage/SelectionOverlayController.cpp: Removed.
- WebProcess/WebPage/ServicesOverlayController.h: Renamed from Source/WebKit2/WebProcess/WebPage/SelectionOverlayController.h.
(WebKit::ServicesOverlayController::create):
- WebProcess/WebPage/mac/SelectionOverlayControllerMac.mm: Removed.
Remove the separate TelephoneNumber overlay controller:
- WebProcess/WebPage/TelephoneNumberOverlayController.cpp: Removed.
- WebProcess/WebPage/TelephoneNumberOverlayController.h: Removed.
- WebProcess/WebPage/mac/TelephoneNumberOverlayControllerMac.mm: Removed.
- WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::servicesOverlayController):
(WebKit::WebPage::telephoneNumberOverlayController): Deleted.
(WebKit::WebPage::selectionOverlayController): Deleted.
(WebKit::WebPage::didChangeScrollOffsetForFrame): Deleted.
- WebProcess/WebPage/WebPage.h:
(WebKit::WebPage::serviceControlsEnabled): Deleted.
ServicesOverlayController is a combination of the old TelephoneNumber and Selection overlay controllers.
A lot of code duplication is resolved, and things are generally easier to follow.
It enforces displaying a telephone number highlight if precisely one telephone number is selected, or
the selection highlight if the selection contains more than a single phone number
- WebProcess/WebPage/mac/ServicesOverlayController.mm: Added.
(WebKit::textQuadsToBoundingRectForRange):
(WebKit::ServicesOverlayController::ServicesOverlayController):
(WebKit::ServicesOverlayController::destroyOverlay):
(WebKit::ServicesOverlayController::pageOverlayDestroyed):
(WebKit::ServicesOverlayController::willMoveToWebPage):
(WebKit::ServicesOverlayController::didMoveToWebPage):
(WebKit::ServicesOverlayController::createOverlayIfNeeded):
(WebKit::ServicesOverlayController::selectionRectsDidChange):
(WebKit::ServicesOverlayController::selectedTelephoneNumberRangesChanged):
(WebKit::ServicesOverlayController::clearHighlightState):
(WebKit::ServicesOverlayController::drawRect):
(WebKit::ServicesOverlayController::drawSelectionHighlight):
(WebKit::ServicesOverlayController::drawTelephoneNumberHighlight):
(WebKit::ServicesOverlayController::drawCurrentHighlight):
(WebKit::ServicesOverlayController::mouseEvent):
(WebKit::ServicesOverlayController::handleClick):
- 8:22 AM Changeset in webkit [170639] by
-
- 2 edits in trunk/Tools
webkit-patch apply-from-bug / apply-attachment should not ask for credentials if none are required
https://bugs.webkit.org/show_bug.cgi?id=40095
Patch by Youenn Fablet <youenn.fablet@crf.canon.fr> on 2014-07-01
Reviewed by Csaba Osztrogonác.
- Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
(Bugzilla.fetch_attachment_contents): make a first fetch on bugzilla without authentication. If failing, use authentication (similar to fetch_bug_dictionary).
(Bugzilla):
(Bugzilla.bug_id_for_attachment_id): ditto.
- 8:16 AM Changeset in webkit [170638] by
-
- 4 edits in trunk/Source/WebCore
REGRESSION(160908): vube.com video won't play after going into and out of fullscreen
https://bugs.webkit.org/show_bug.cgi?id=134489
Reviewed by Zalan Bujtas.
Test: fullscreen/full-screen-plugin.html
It is difficult to restore the render tree correctly in all cases after removing a full screen
renderer from the tree. r160908 avoided dealing with this by simply always reconstructing the subtree.
Unfortunately plugin lifetime is currently tied to its renderer so this would cause the plugin to restart.
With this patch we avoid reconstruction in normal cases and only force it if the render tree is complicated.
- dom/Document.cpp:
(WebCore::unwrapFullScreenRenderer):
Force reconstruction conditionally.
(WebCore::Document::webkitWillEnterFullScreenForElement):
(WebCore::Document::webkitDidExitFullScreenForElement):
- rendering/RenderFullScreen.cpp:
(WebCore::RenderFullScreen::wrapRenderer):
(WebCore::RenderFullScreen::unwrapRenderer):
Deal with the simple case of single child, possibly in anonymous wrapper.
In other cases request reconstruction.
This is covered by the existing fullscreen tests.
- rendering/RenderFullScreen.h:
- 6:10 AM Changeset in webkit [170637] by
-
- 19 edits in trunk/Tools
[webkitpy] Refactor pluralizer in grammar.py
https://bugs.webkit.org/show_bug.cgi?id=133067
Patch by Eva Balazsfalvi <evab.u-szeged@partner.samsung.com> on 2014-07-01
Reviewed by Csaba Osztrogonác.
- Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
(LayoutTestRunner.run_tests):
- Scripts/webkitpy/layout_tests/controllers/manager.py:
(Manager.run):
- Scripts/webkitpy/layout_tests/views/printing.py:
(Printer.print_found):
(Printer.print_workers_and_shards):
(Printer._print_one_line_summary):
- Scripts/webkitpy/test/printer.py:
(Printer.print_result):
- Scripts/webkitpy/tool/bot/feeders.py:
(EWSFeeder.feed):
- Scripts/webkitpy/tool/bot/flakytestreporter.py:
(FlakyTestReporter._lookup_bug_for_flaky_test):
- Scripts/webkitpy/tool/bot/irc_command.py:
(Rollout._check_diff_failure):
- Scripts/webkitpy/tool/bot/irc_command_unittest.py:
(IRCCommandTest.test_rollout):
- Scripts/webkitpy/tool/commands/download.py:
(AbstractPatchProcessingCommand._process_patch.execute):
(ProcessBugsMixin._fetch_list_of_patches_to_process):
(ProcessURLsMixin._fetch_list_of_patches_to_process):
- Scripts/webkitpy/tool/commands/queries.py:
(WhatBroke.execute):
- Scripts/webkitpy/tool/commands/suggestnominations.py:
(SuggestNominations._print_nominations):
(SuggestNominations._print_counts):
- Scripts/webkitpy/tool/commands/suggestnominations_unittest.py:
(_make_options):
(test_basic):
(test_with_verbose):
- Scripts/webkitpy/tool/commands/upload.py:
(PostCommits.execute):
- Scripts/webkitpy/tool/grammar.py:
(pluralize):
- Scripts/webkitpy/tool/grammar_unittest.py:
(GrammarTest.test_join_with_separators):
(GrammarTest):
(GrammarTest.test_plural):
(GrammarTest.test_pluralize):
- Scripts/webkitpy/tool/multicommandtool.py:
(Command.check_arguments_and_execute):
- Scripts/webkitpy/tool/steps/obsoletepatches.py:
(ObsoletePatches.run):
- Scripts/webkitpy/tool/steps/preparechangelogforrevert.py:
(PrepareChangeLogForRevert._message_for_revert):
- 5:07 AM WebKitGTK/2.4.x edited by
- (diff)
- 5:07 AM Changeset in webkit [170636] by
-
- 14 edits2 adds in releases/WebKitGTK/webkit-2.4
Merge r169799 - [GTK] Use a different user agent string depending on the site
https://bugs.webkit.org/show_bug.cgi?id=132681
Reviewed by Anders Carlsson.
Source/WebCore:
We have changed the user agent string several times to try to fix
broken websites that require specific things in the UA string to
properly work. But everytime we change the UA string to fix a
website we break others. We could use different UA string
depending on the website. UserAgentGtk code has also been cleaned
up, using NeverDestroyed instead of DEPRECATED_DEFINE_STATIC_LOCAL
and avoiding unneeded conversions to UTF-8.
- platform/gtk/UserAgentGtk.cpp:
(WebCore::UserAgentQuirks::UserAgentQuirks): New helper private
class to handle user agent quirks.
(WebCore::UserAgentQuirks::add):
(WebCore::UserAgentQuirks::contains):
(WebCore::UserAgentQuirks::isEmpty):
(WebCore::platformForUAString): Bring back this method that was
removed to always pretend to be Macintosh.
(WebCore::platformVersionForUAString): Return a different platform
version depending on the actual platform.
(WebCore::versionForUAString): Return the WebKit version.
(WebCore::buildUserAgentString): Helper function to build the user
agent taking into account the UserAgentQuirks received.
(WebCore::standardUserAgentStatic): Standard user agent string
when no quirks are present.
(WebCore::standardUserAgent):
(WebCore::standardUserAgentForURL): New method that returns the
user agent string for the given URL.
- platform/gtk/UserAgentGtk.h:
Source/WebKit2:
- UIProcess/API/gtk/WebKitSettings.cpp:
(webkit_settings_class_init): Enable site specific quirks setting
by default.
- WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::userAgent): Pass the given URL to WebPage.
- WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::userAgent): Try to get the user agent for the
URL received falling back to the current one otherwise.
- WebProcess/WebPage/WebPage.h:
(WebKit::WebPage::platformUserAgent): Added.
- WebProcess/WebPage/efl/WebPageEfl.cpp:
(WebKit::WebPage::platformUserAgent): Return null String.
- WebProcess/WebPage/gtk/WebPageGtk.cpp:
(WebKit::WebPage::platformUserAgent): Use WebCore::standardUserAgentForURL() when site specific quirks
setting is enabled.
- WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::platformUserAgent): Return null String.
- WebProcess/WebPage/mac/WebPageMac.mm:
(WebKit::WebPage::platformUserAgent): Return null String.
Tools:
Add a unit test to check user agent quirks.
- TestWebKitAPI/PlatformGTK.cmake:
- TestWebKitAPI/Tests/WebCore/gtk/UserAgentQuirks.cpp: Added.
(TestWebKitAPI::TEST):
- TestWebKitAPI/Tests/WebKit2Gtk/TestWebKitSettings.cpp:
(testWebKitSettings): Site specific quirks setting is now enabled
by default.
- 4:32 AM Changeset in webkit [170635] by
-
- 5 edits in trunk/Source/WebKit2
[WK2] IPC messages' arguments() method returns const std::tuple<> copies
https://bugs.webkit.org/show_bug.cgi?id=131975
Reviewed by Darin Adler.
The arguments() method of the different IPC message objects should return a const reference
to the std::tuple<> containing the arguments, instead of a const value. This avoids copying
the std::tuple<> every time the arguments are encoded, and works best with the
ArgumentEncoder::encode<std::tuple<...>>(const std::tuple<...>&) template instantiation.
The unit test baselines are updated accordingly.
- Scripts/webkit2/LegacyMessages-expected.h:
(Messages::WebPage::LoadURL::arguments):
(Messages::WebPage::LoadSomething::arguments):
(Messages::WebPage::TouchEvent::arguments):
(Messages::WebPage::AddEvent::arguments):
(Messages::WebPage::LoadSomethingElse::arguments):
(Messages::WebPage::DidReceivePolicyDecision::arguments):
(Messages::WebPage::Close::arguments):
(Messages::WebPage::PreferencesDidChange::arguments):
(Messages::WebPage::SendDoubleAndFloat::arguments):
(Messages::WebPage::SendInts::arguments):
(Messages::WebPage::CreatePlugin::arguments):
(Messages::WebPage::RunJavaScriptAlert::arguments):
(Messages::WebPage::GetPlugins::arguments):
(Messages::WebPage::GetPluginProcessConnection::arguments):
(Messages::WebPage::TestMultipleAttributes::arguments):
(Messages::WebPage::TestParameterAttributes::arguments):
(Messages::WebPage::TemplateTest::arguments):
(Messages::WebPage::SetVideoLayerID::arguments):
(Messages::WebPage::DidCreateWebProcessConnection::arguments):
(Messages::WebPage::InterpretKeyEvent::arguments):
(Messages::WebPage::DeprecatedOperation::arguments):
(Messages::WebPage::ExperimentalOperation::arguments):
- Scripts/webkit2/Messages-expected.h:
(Messages::WebPage::LoadURL::arguments):
(Messages::WebPage::LoadSomething::arguments):
(Messages::WebPage::TouchEvent::arguments):
(Messages::WebPage::AddEvent::arguments):
(Messages::WebPage::LoadSomethingElse::arguments):
(Messages::WebPage::DidReceivePolicyDecision::arguments):
(Messages::WebPage::Close::arguments):
(Messages::WebPage::PreferencesDidChange::arguments):
(Messages::WebPage::SendDoubleAndFloat::arguments):
(Messages::WebPage::SendInts::arguments):
(Messages::WebPage::CreatePlugin::arguments):
(Messages::WebPage::RunJavaScriptAlert::arguments):
(Messages::WebPage::GetPlugins::arguments):
(Messages::WebPage::GetPluginProcessConnection::arguments):
(Messages::WebPage::TestMultipleAttributes::arguments):
(Messages::WebPage::TestParameterAttributes::arguments):
(Messages::WebPage::TemplateTest::arguments):
(Messages::WebPage::SetVideoLayerID::arguments):
(Messages::WebPage::DidCreateWebProcessConnection::arguments):
(Messages::WebPage::InterpretKeyEvent::arguments):
(Messages::WebPage::DeprecatedOperation::arguments):
(Messages::WebPage::ExperimentalOperation::arguments):
- Scripts/webkit2/MessagesSuperclass-expected.h:
(Messages::WebPage::LoadURL::arguments):
- Scripts/webkit2/messages.py:
(message_to_struct_declaration):
- 4:29 AM Changeset in webkit [170634] by
-
- 5 edits in trunk/Source/WebCore
Remove remaining Vector<> copies in WebCore accessibility code
https://bugs.webkit.org/show_bug.cgi?id=133263
Reviewed by Darin Adler.
- accessibility/AXObjectCache.cpp:
(WebCore::AXObjectCache::notificationPostTimerFired): Don't copy the Vector member and
then clear it -- move it into the local variable instead.
- accessibility/AccessibilityNodeObject.cpp:
(WebCore::AccessibilityNodeObject::ariaLabeledByText): Move the axElements Vector into
the AccessibilityText constructor.
(WebCore::AccessibilityNodeObject::stringValue): Create a const reference to the Vector of
HTMLSelectElement's list items instead of copying it.
- accessibility/AccessibilityObject.h:
(WebCore::AccessibilityText::AccessibilityText): Take the Vector parameter by non-const
value and move it into the member variable.
- accessibility/AccessibilityRenderObject.cpp:
(WebCore::AccessibilityRenderObject::stringValue): Create a const reference to the Vector
of HTMLSelectElement's list items instead of copying it.
(WebCore::AccessibilityRenderObject::ariaSelectedRows): Wrap the AccessibilityChildrenVector
iteration in a lambda. Use it to iterate over either the newly-constructed Vector (in case
the object has the tree role) or the reference to the vector returned by
AccessibilityTable::rows() (in case the object is an AccessibilityTable instance).
- 2:10 AM Changeset in webkit [170633] by
-
- 2 edits in trunk/Source/WebKit2
[GTK] Annotations for WebResource.get_data produce wrong signature in Vala binding
https://bugs.webkit.org/show_bug.cgi?id=134476
Patch by Evan Nemerson <evan@nemerson.com> on 2014-07-01
Reviewed by Carlos Garcia Campos.
- UIProcess/API/gtk/WebKitWebResource.cpp: Add array length and
element-type annotations to return value of webkit_web_resource_get_data_finish().
- 1:17 AM Changeset in webkit [170632] by
-
- 2 edits in trunk/Source/WebKit2
[WK2] Can not convert ‘sessionState.WebKit::SessionState::provisionalURL’ from ‘WebCore::URL’ to ‘bool’
https://bugs.webkit.org/show_bug.cgi?id=134487
Rubber-stamped by Carlos Garcia Campos.
- UIProcess/WebPageProxy.cpp: Use isNull().
(WebKit::WebPageProxy::restoreFromState):
- 12:57 AM Changeset in webkit [170631] by
-
- 3 edits1 copy1 move in trunk/Source/WebKit2
Fix build break on EFL and GTK ports since r170611 and r170614
https://bugs.webkit.org/show_bug.cgi?id=134479
Reviewed by Carlos Garcia Campos.
Add LegacySessionStateCodingNone.cpp in order to fix build break on EFL and GTK ports. Additionally
SessionStateConversion.cpp is added to CMakeLists.txt as well.
- CMakeLists.txt:
- UIProcess/LegacySessionStateCoding.h: Copied from Source/WebKit2/UIProcess/mac/LegacySessionStateCoding.h.
- UIProcess/LegacySessionStateCodingNone.cpp: Renamed from Source/WebKit2/UIProcess/mac/LegacySessionStateCoding.h.
(WebKit::encodeLegacySessionHistoryEntryData):
(WebKit::decodeLegacySessionState):
(WebKit::decodeLegacySessionHistoryEntryData):
- UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::loadFile):
(WebKit::WebPageProxy::restoreFromState):
- WebKit2.xcodeproj/project.pbxproj:
Jun 30, 2014:
- 9:16 PM Changeset in webkit [170630] by
-
- 7 edits in trunk/Source
Unreviewed, rolling out r170605.
https://bugs.webkit.org/show_bug.cgi?id=134484
The changes cause crashes on ARM64 (Requested by benjaminp on
#webkit).
Reverted changeset:
"Reduce dynamic memory allocation in css jit."
https://bugs.webkit.org/show_bug.cgi?id=134416
http://trac.webkit.org/changeset/170605
- 8:12 PM Changeset in webkit [170629] by
-
- 2 edits in trunk/Source/WebKit2
[iOS][WK2] The long press gesture recognizers should not block the pinch gesture
https://bugs.webkit.org/show_bug.cgi?id=134477
<rdar://problem/17514936>
Reviewed by Tim Horton.
- UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView gestureRecognizer:canPreventGestureRecognizer:]):
- 7:19 PM Changeset in webkit [170628] by
-
- 8 edits21 adds in branches/ftlopt
Merge r169121 from trunk.
2014-05-19 Mark Hahnenberg <mhahnenberg@apple.com>
JSArray::shiftCountWith* could be more efficient
https://bugs.webkit.org/show_bug.cgi?id=133011
Reviewed by Geoffrey Garen.
Source/JavaScriptCore:
Our current implementations of shiftCountWithAnyIndexingType and shiftCountWithArrayStorage
are scared of the presence of any holes in the array. We can mitigate this somewhat by enabling
them to correctly handle holes, thus avoiding the slowest of slow paths in most cases.
- runtime/ArrayStorage.h: (JSC::ArrayStorage::indexingHeader): (JSC::ArrayStorage::length): (JSC::ArrayStorage::hasHoles):
- runtime/IndexingHeader.h: (JSC::IndexingHeader::publicLength): (JSC::IndexingHeader::from):
- runtime/JSArray.cpp: (JSC::JSArray::shiftCountWithArrayStorage): (JSC::JSArray::shiftCountWithAnyIndexingType): (JSC::JSArray::unshiftCountWithArrayStorage):
- runtime/JSArray.h: (JSC::JSArray::shiftCountForShift): (JSC::JSArray::shiftCountForSplice): (JSC::JSArray::shiftCount):
- runtime/Structure.cpp: (JSC::Structure::holesRequireSpecialBehavior):
- runtime/Structure.h:
LayoutTests:
Added a performance regression test which tests the contiguous shift case. We're ~85% faster
on this microbenchmark with this patch.
Also added a bunch of tests for each permutation of contiguous/array storage with no holes, holes,
and holes that require special behavior (e.g. indexed properties in the prototype chain).
- js/array-storage-splice-holes-expected.txt: Added.
- js/array-storage-splice-holes-require-special-behavior-expected.txt: Added.
- js/array-storage-splice-holes-require-special-behavior.html: Added.
- js/array-storage-splice-holes.html: Added.
- js/array-storage-splice-no-holes-expected.txt: Added.
- js/array-storage-splice-no-holes.html: Added.
- js/contiguous-splice-holes-expected.txt: Added.
- js/contiguous-splice-holes-require-special-behavior-expected.txt: Added.
- js/contiguous-splice-holes-require-special-behavior.html: Added.
- js/contiguous-splice-holes.html: Added.
- js/contiguous-splice-no-holes-expected.txt: Added.
- js/contiguous-splice-no-holes.html: Added.
- js/regress/array-splice-contiguous-expected.txt: Added.
- js/regress/array-splice-contiguous.html: Added.
- js/regress/script-tests/array-splice-contiguous.js: Added. (foo):
- js/script-tests/array-storage-splice-holes-require-special-behavior.js: Added.
- js/script-tests/array-storage-splice-holes.js: Added.
- js/script-tests/array-storage-splice-no-holes.js: Added.
- js/script-tests/contiguous-splice-holes-require-special-behavior.js: Added.
- js/script-tests/contiguous-splice-holes.js: Added.
- js/script-tests/contiguous-splice-no-holes.js: Added.
- 6:40 PM Changeset in webkit [170627] by
-
- 3 edits in trunk/Source/WebKit2
Add WebPageProxy::restoreFromState
https://bugs.webkit.org/show_bug.cgi?id=134480
Reviewed by Tim Horton.
- UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::restoreFromState):
- UIProcess/WebPageProxy.h:
- 6:29 PM Changeset in webkit [170626] by
-
- 4 edits in trunk/Source/WebKit2
Add a function to restore a WebBackForwardList object from a BackForwardListState object
https://bugs.webkit.org/show_bug.cgi?id=134478
Reviewed by Tim Horton.
- UIProcess/WebBackForwardList.cpp:
(WebKit::generateWebBackForwardItemID):
(WebKit::WebBackForwardList::restoreFromState):
- UIProcess/WebBackForwardList.h:
- UIProcess/cf/WebBackForwardListCF.cpp:
(WebKit::generateWebBackForwardItemID): Deleted.
- 6:06 PM Changeset in webkit [170625] by
-
- 10 edits in trunk/Source
Change the AddBackForwardItem message to take a page state object
https://bugs.webkit.org/show_bug.cgi?id=134475
Reviewed by Andreas Kling.
Source/WebCore:
- WebCore.exp.in:
Source/WebKit2:
- Scripts/webkit2/messages.py:
(struct_or_class):
(headers_for_type):
- Shared/WebBackForwardListItem.cpp:
(WebKit::WebBackForwardListItem::create):
(WebKit::WebBackForwardListItem::WebBackForwardListItem):
- Shared/WebBackForwardListItem.h:
(WebKit::WebBackForwardListItem::setPageState):
- UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::addBackForwardItem):
- UIProcess/WebProcessProxy.h:
- UIProcess/WebProcessProxy.messages.in:
- WebProcess/WebPage/WebBackForwardListProxy.cpp:
(WebKit::updateBackForwardItem):
- 5:50 PM Changeset in webkit [170624] by
-
- 3 edits in trunk/Source/WebKit2
[WK2] Add a flatter find-in-page current match indicator style
https://bugs.webkit.org/show_bug.cgi?id=134434
<rdar://problem/16225673>
Reviewed by Simon Fraser.
- UIProcess/FindIndicator.cpp:
(WebKit::findIndicatorsForTextRectsOverlap):
(WebKit::FindIndicator::frameRect):
(WebKit::flatHighlightColor):
(WebKit::flatRimShadowColor):
(WebKit::flatDropShadowColor):
(WebKit::FindIndicator::draw):
Add a flatter find indicator, with a bigger shadow.
- WebProcess/WebPage/FindController.cpp:
(WebKit::FindController::updateFindUIAfterPageScroll):
(WebKit::FindController::getFindIndicatorBitmapAndRect):
(WebKit::FindController::hideFindIndicator):
(WebKit::FindController::drawRect):
Don't paint a shadow behind the secondary matches if we're using the new style.
- 5:45 PM Changeset in webkit [170623] by
-
- 3 edits in trunk/Tools
[iOS] GSFontAddCGFont is deprecated and we use it in DumpRenderTree
https://bugs.webkit.org/show_bug.cgi?id=134133
Reviewed by Simon Fraser.
Replace it with CTFontManagerRegisterGraphicsFont
- DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
- DumpRenderTree/mac/DumpRenderTree.mm:
(activateFontIOS):
- 5:42 PM Changeset in webkit [170622] by
-
- 2 edits in trunk/Source/WebKit2
[iOS][WK2] Improve double-tap-to-scroll on image documents
https://bugs.webkit.org/show_bug.cgi?id=134474
<rdar://problem/17496778>
Reviewed by Enrica Casucci.
On image document, we were always rendering the center of the image as the origin, which broke
double tap to scroll.
This patch improves on this in two ways:
1) If the hit testing already got the image, only change the type to replaced (to get the right
scaling behavior for images since the type is block on iOS's image document).
2) If the hit testing is outside the image, only center the axis that is not in the image.
This way, we "fix" the component that is not valid, and keep the vali component.
This works great for viewing comics on iPad.
- WebProcess/WebPage/ViewGestureGeometryCollector.cpp:
(WebKit::ViewGestureGeometryCollector::collectGeometryForSmartMagnificationGesture):
- 5:38 PM Changeset in webkit [170621] by
-
- 2 edits in trunk/Source/WebCore
[iOS][WK2] Do not put tap highlight on images that are not in links, and on applet/embed/object
https://bugs.webkit.org/show_bug.cgi?id=134471
<rdar://problem/17513542>
Reviewed by Enrica Casucci.
UIKit has some special code to avoid tap highlight on certain type of object. This is not very flexible
since the page has no way of overriding that behavior.
This patch implement the same feature by changing the default stylesheet. The tap highlight is disabled
by default for those object, and enabled for images that are in a link.
- css/html.css:
(applet, embed, object, img):
(:-webkit-any-link img):
- 5:27 PM Changeset in webkit [170620] by
-
- 2 edits in trunk/Source/WebKit2
REGRESSION (WK2): Weird selection behavior on autos.yahoo.com, en.wikipedia.org.
https://bugs.webkit.org/show_bug.cgi?id=134466
<rdar://problem/16817263>
Reviewed by Benjamin Poulain.
Avoid selecting blocks across frame boundaries and skip non-selectable
blocks. If the only block we find is almost the same height as the
visible area, we should not select at all.
This patch also fixes the logic to compute the next block when
shrinking the selection. When calculating the new range after shrinking,
we should not include the block that corresponds to the current handle position.
- WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::rangeForWebSelectionAtPosition):
(WebKit::WebPage::expandedRangeFromHandle):
(WebKit::WebPage::contractedRangeFromHandle):
(WebKit::WebPage::updateSelectionWithTouches):
- 5:17 PM Changeset in webkit [170619] by
-
- 3 edits4 adds in trunk
All Indic text is rendered as boxes on iOS
https://bugs.webkit.org/show_bug.cgi?id=134464
Reviewed by Simon Fraser.
Source/WebCore:
I made a typo in r170207.
Test: fast/text/indic.html
- platform/graphics/ios/FontCacheIOS.mm:
(WebCore::languageSpecificFallbackFont):
LayoutTests:
Simply draw some indic text and test that against currently good metric values.
- fast/text/indic.html: Added
- platform/ios-sim/fast/text/indic-expected.txt: Added.
- platform/mac/fast/text/indic-expected.txt: Added.
- platform/mac/fast/text/indic-expected.png: Added.
- 5:14 PM Changeset in webkit [170618] by
-
- 2 edits in trunk/Source/WebCore
[iOS] languageSpecificFallbackFont() is not consistent
https://bugs.webkit.org/show_bug.cgi?id=134473
Reviewed by Darin Adler.
It can use "<" throughout the function as well as keeping the fallthrough behavior consistent.
No tests because there is no behavior change.
- platform/graphics/ios/FontCacheIOS.mm:
(WebCore::languageSpecificFallbackFont):
- 5:11 PM Changeset in webkit [170617] by
-
- 2 edits in trunk/Source/WebCore
[Win] Unreviewed gardening.
- WebCore.vcxproj/WebCore.vcxproj: Do not compile JSFileException
and JSFileReaderSync multiple times.
- 5:04 PM Changeset in webkit [170616] by
-
- 2 edits in trunk/Source/WebCore
Avoid copying function object in lambda function in ScrollingThread::dispatchBarrier(); actually use move semantics
https://bugs.webkit.org/show_bug.cgi?id=134470
Reviewed by Anders Carlsson.
Currently we always copy construct the captured std::function object when calling callOnMainThread()
in the lambda function created in ScrollingThread::dispatchBarrier() because captured variables are
const in a lambda expression's body by default. That is, the std::function object is captured as const.
Instead, we should mark this lambda expression as mutable so the captured std::function object is
non-const and hence we can use move semantics when passing it to callOnMainThread().
- page/scrolling/ScrollingThread.cpp:
(WebCore::ScrollingThread::dispatchBarrier):
- 5:00 PM Changeset in webkit [170615] by
-
- 2 edits in trunk/Source/WTF
Make TransformIterator::TransformIterator() take rvalue references
https://bugs.webkit.org/show_bug.cgi?id=134468
Reviewed by Darin Adler.
It's unnecessary to support TransformIterator::TransformIterator() taking const lvalue references
as we only call it with rvalues. As a side benefit, we can use move semantics to initialize the
members of TransformIterator.
- wtf/IteratorAdaptors.h:
(WTF::TransformIterator::TransformIterator): Make constructor take arguments as rvalue references.
(WTF::makeTransformIterator): Substitute std::move() for std::forward().
- 4:39 PM Changeset in webkit [170614] by
-
- 5 edits in trunk/Source/WebKit2
WebBackForwardListItem should not store back-forward data
https://bugs.webkit.org/show_bug.cgi?id=134469
Reviewed by Darin Adler.
Change WebBackForwardListItem::backForwardData to encode the main frame state lazily,
and change WebBackForwardListItem::setBackForwardData to decode it into the main frame state.
- Shared/WebBackForwardListItem.cpp:
(WebKit::WebBackForwardListItem::backForwardData):
(WebKit::WebBackForwardListItem::setBackForwardData):
(WebKit::WebBackForwardListItem::encode):
- Shared/WebBackForwardListItem.h:
(WebKit::WebBackForwardListItem::backForwardData): Deleted.
- UIProcess/cf/WebBackForwardListCF.cpp:
(WebKit::WebBackForwardList::createCFDictionaryRepresentation):
- WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::restoreSession):
- 4:26 PM Changeset in webkit [170613] by
-
- 3 edits in trunk/Source/WebKit2
WebBackForwardListItems should hold on to PageState objects
https://bugs.webkit.org/show_bug.cgi?id=134467
Reviewed by Darin Adler.
- Shared/WebBackForwardListItem.cpp:
(WebKit::WebBackForwardListItem::WebBackForwardListItem):
(WebKit::WebBackForwardListItem::encode):
- Shared/WebBackForwardListItem.h:
(WebKit::WebBackForwardListItem::setOriginalURL):
(WebKit::WebBackForwardListItem::originalURL):
(WebKit::WebBackForwardListItem::setURL):
(WebKit::WebBackForwardListItem::url):
(WebKit::WebBackForwardListItem::setTitle):
(WebKit::WebBackForwardListItem::title):
- 4:24 PM Changeset in webkit [170612] by
-
- 2 edits in trunk/Source/WebKit2
Add a missing semicolon in WKContentViewInteraction
- UIProcess/ios/WKContentViewInteraction.mm:
(nsSizeForTapHighlightBorderRadius):
I messed up this when fixing the patch on landing.
- 4:05 PM Changeset in webkit [170611] by
-
- 6 edits in trunk/Source
Adopt the legacy session decoding inside WebPage::restoreSession for now
https://bugs.webkit.org/show_bug.cgi?id=134465
Reviewed by Sam Weinig.
Source/WebCore:
- WebCore.exp.in:
Source/WebKit2:
- UIProcess/mac/LegacySessionStateCoding.cpp:
(WebKit::decodeLegacySessionHistoryEntryData):
- UIProcess/mac/LegacySessionStateCoding.h:
- WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::restoreSession):
- 3:58 PM Changeset in webkit [170610] by
-
- 2 edits in trunk/Source/WebKit2
[iOS WK2] Don't crash on status bar tap with overflow scroll views
https://bugs.webkit.org/show_bug.cgi?id=134459
Reviewed by Tim Horton.
The tiled scrolling indicator parents UIViews under CALayers, but that causes
those views to miss -willMoveToWindow:nil notifications on teardown. This can leave
deleted UIScrollViews in UIKit's "scroll to top" list.
Fix by not making UIScrollViews in the debug layer tree host.
- UIProcess/ios/RemoteLayerTreeHostIOS.mm:
(WebKit::RemoteLayerTreeHost::createLayer):
- 3:52 PM Changeset in webkit [170609] by
-
- 4 edits in trunk/Source/WebCore
Unreviewed, rolling out r170578.
https://bugs.webkit.org/show_bug.cgi?id=134462
causes assertions loading nytimes.com on iOS (Requested by
smfr on #webkit).
Reverted changeset:
"Create NSURLRequest lazily when USE(CFNETWORK) is enabled"
https://bugs.webkit.org/show_bug.cgi?id=134441
http://trac.webkit.org/changeset/170578
- 3:51 PM Changeset in webkit [170608] by
-
- 10 edits in trunk/Source/WebKit2
Restrict network process sandbox
https://bugs.webkit.org/show_bug.cgi?id=134360
Reviewed by Sam Weinig.
Add more restrictions to the network process sandbox.
- NetworkProcess/cocoa/NetworkProcessCocoa.mm: (WebKit::NetworkProcess::platformInitializeNetworkProcessCocoa):
Always use the cache directory provided in the initialization parameters,
and make sure we consume the cookie directory extension.
- Resources/SandboxProfiles/ios/com.apple.WebKit.Networking.sb:
Make the sandbox profile much more restrictive.
- Shared/Network/NetworkProcessCreationParameters.cpp: (WebKit::NetworkProcessCreationParameters::encode): (WebKit::NetworkProcessCreationParameters::decode):
- Shared/Network/NetworkProcessCreationParameters.h:
The network process now requires an extension to access
its cookie storage.
- Shared/mac/SandboxUtilities.cpp: (WebKit::pathForProcessContainer):
- Shared/mac/SandboxUtilities.h:
We need to be able to get hold of our container so
that we can get the correct cookie storage directory.
- UIProcess/WebContext.cpp: (WebKit::WebContext::ensureNetworkProcess):
We have to pass in the an extension for the cookie storage directory when
initalising the network process
- UIProcess/mac/WebContextMac.mm: (WebKit::WebContext::platformDefaultCookieStorageDirectory):
Make sure we provide the correct location on IOS
- WebProcess/cocoa/WebProcessCocoa.mm: (WebKit::WebProcess::platformInitializeWebProcess):
Consume the cookie storage extension
- 3:50 PM Changeset in webkit [170607] by
-
- 2 edits in trunk/Source/WebKit2
[iOS WK2] Turn off scrollsToTop on overflow UIScrollViews
https://bugs.webkit.org/show_bug.cgi?id=134456
Reviewed by Tim Horton.
Set scrollsToTop to NO on our UIScrollViews created for overflow scrolling, since
we don't have a good strategy for when to allow it, and doing so would also require
some smarts in the main UIScrollView.
- UIProcess/Scrolling/ios/ScrollingTreeOverflowScrollingNodeIOS.mm:
(WebKit::ScrollingTreeOverflowScrollingNodeIOS::updateAfterChildren):
- 3:50 PM Changeset in webkit [170606] by
-
- 2 edits in trunk/Source/WebKit2
REGRESSION (Okemo): The contextual menu on tap and hold does not cancel the other gestures.
https://bugs.webkit.org/show_bug.cgi?id=134463
<rdar://problem/17388907>
Reviewed by Benjamin Poulain.
- UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView _longPressRecognized:]):
- 3:48 PM Changeset in webkit [170605] by
-
- 7 edits in trunk/Source
Reduce dynamic memory allocation in css jit.
https://bugs.webkit.org/show_bug.cgi?id=134416
Reviewed by Benjamin Poulain.
Source/WebCore:
- cssjit/FunctionCall.h:
(WebCore::FunctionCall::FunctionCall):
(WebCore::FunctionCall::saveAllocatedCallerSavedRegisters):
- cssjit/RegisterAllocator.h:
- cssjit/SelectorCompiler.cpp:
(WebCore::SelectorCompiler::addPseudoClassType):
(WebCore::SelectorCompiler::SelectorCodeGenerator::generatePrologue):
(WebCore::SelectorCompiler::SelectorCodeGenerator::generateSelectorChecker):
(WebCore::SelectorCompiler::SelectorCodeGenerator::generateElementIsNthChild):
Use inlineCapacity template parameter to reduce malloc calls.
I use 16 and 32 as the inline capacity for non-register related vectors because that
is probably big enough for the selector compiler, and if it is not then a malloc
call will not be significant because the selector is very complex.
- cssjit/StackAllocator.h:
(WebCore::StackAllocator::push):
Pass a vector of stack references to match calls to pop and to avoid a call to appendVector.
Source/WTF:
- wtf/Deque.h:
Added inlineCapacity optional template parameter.
- 3:22 PM Changeset in webkit [170604] by
-
- 2 edits in branches/ftlopt/Source/JavaScriptCore
CodeBlock::stronglyVisitWeakReferences should mark DFG::CommonData::weakStructureReferences
https://bugs.webkit.org/show_bug.cgi?id=134455
Reviewed by Geoffrey Garen.
Otherwise we get hanging pointers which can cause us to die later.
- bytecode/CodeBlock.cpp:
(JSC::CodeBlock::stronglyVisitWeakReferences):
- 3:14 PM Changeset in webkit [170603] by
-
- 2 edits in trunk/Source/WebKit2
Implement the last pieces of encodeFrameStateNode
https://bugs.webkit.org/show_bug.cgi?id=134460
Reviewed by Andreas Kling.
- UIProcess/mac/LegacySessionStateCoding.cpp:
(WebKit::HistoryEntryDataEncoder::operator<<):
(WebKit::isValidEnum):
(WebKit::encodeFormDataElement):
(WebKit::encodeFormData):
(WebKit::encodeFrameStateNode):
- 3:13 PM Changeset in webkit [170602] by
-
- 3 edits in trunk/LayoutTests
js/script-tests/function-apply-many-args.js fails on ARM
https://bugs.webkit.org/show_bug.cgi?id=126588
Reviewed by Mark Lam.
Removed three test cases that fail on some platforms and pass on others due
to stack size. Other test cases cover what will fit on a stack or what will
throw an out of stack exception.
- js/function-apply-many-args-expected.txt:
- js/script-tests/function-apply-many-args.js:
- 3:12 PM Changeset in webkit [170601] by
-
- 6 edits in trunk
Add option to run-jsc-stress-testes to filter out tests that use large heaps
https://bugs.webkit.org/show_bug.cgi?id=134458
Reviewed by Filip Pizlo.
Source/JavaScriptCore:
Added test to skip js1_5/Regress/regress-159334.js when testing on a memory limited device.
- tests/mozilla/mozilla-tests.yaml:
Tools:
Added new --memory-limited to filter out tests that have large heap usage.
Such tests are either tagged with @largeHeap or logic is added to their .yaml file.
- Scripts/run-jsc-stress-tests:
LayoutTests:
- js/regress/script-tests/lots-of-fields.js: Added @largeHeap tag to skip when
running on a memoryLimited device.
- 3:10 PM Changeset in webkit [170600] by
-
- 5 edits in trunk/Source/WebKit2
[iOS][WK2] Move tap highlight to the inverseScaleRootView
https://bugs.webkit.org/show_bug.cgi?id=134424
<rdar://problem/17480880>
Patch by Benjamin Poulain <bpoulain@apple.com> on 2014-06-30
Reviewed by Tim Horton.
Enrica introduced the inverseScaleRootView that is a dynamic version of _highlightRootView.
This patch moves the tap highlight from its own inverse root to inverseScaleRootView. This provides better handling
of scale, better behavior on crash, and remove one extra view from the hierarchy.
The code is pretty much the same, just moved around. Some of the changes:
Now the code needs to handle repainting live when the view scale. To do that, _showTapHighlightXXX was split in two:
-_showTapHighlight to setup the view.
-_updateTapHighlight to update the view based on the current scale.
Since the view updates live on scaling, we need to recompute the coordinates for each update. To do that, the _potentialTapHighlightInformation
was generalized to handle all cases of highlight.
Since we can no longer test for the nullity of _potentialTapHighlightInformation,
a new attribute is introduced for that: _hasTapHighlightForPotentialTap.
The last bit of change concern reentering the tap highlight. This becomes quite simple:
1) _isTapHighlightIDValid is updated before the animation starts
so that it can be set again during an animation.
2) The animation checks the "finished" flags before removing the view from its superview
to avoid breaking the new animation.
- UIProcess/API/Cocoa/WKWebView.mm:
(withinEpsilon): Deleted.
- UIProcess/ios/WKContentView.mm:
(-[WKContentView didUpdateVisibleRect:unobscuredRect:unobscuredRectInScrollViewCoordinates:scale:minimumScale:inStableState:isChangingObscuredInsetsInteractively:]):
Make _updateUnscaledView conditional to scale changes, that thing isn't cheap!
- UIProcess/ios/WKContentViewInteraction.h:
(withinEpsilon):
Move withinEpsilon() from WKContentView to use it from both classes.
- UIProcess/ios/WKContentViewInteraction.mm:
(-[WKContentView cleanupInteraction]):
(-[WKContentView _updateUnscaledView]):
(-[WKContentView _updateTapHighlight]):
(-[WKContentView _showTapHighlight]):
(-[WKContentView _didGetTapHighlightForRequest:color:quads:topLeftRadius:topRightRadius:bottomLeftRadius:bottomRightRadius:]):
(-[WKContentView _cancelInteraction]):
(-[WKContentView _finishInteraction]):
(-[WKContentView _singleTapDidReset:]):
(-[WKContentView _singleTapCommited:]):
(-[WKContentView _showTapHighlightWithColor:quads:topLeftRadius:topRightRadius:bottomLeftRadius:bottomRightRadius:]): Deleted.
- 2:53 PM Changeset in webkit [170599] by
-
- 5 edits in trunk/Source
Add code for encoding legacy session history entries
https://bugs.webkit.org/show_bug.cgi?id=134452
Reviewed by Andreas Kling.
Source/WebKit2:
- UIProcess/mac/LegacySessionStateCoding.cpp:
(WebKit::HistoryEntryDataEncoder::HistoryEntryDataEncoder):
(WebKit::HistoryEntryDataEncoder::operator<<):
(WebKit::HistoryEntryDataEncoder::finishEncoding):
(WebKit::HistoryEntryDataEncoder::encodeArithmeticType):
(WebKit::HistoryEntryDataEncoder::encodeFixedLengthData):
(WebKit::HistoryEntryDataEncoder::grow):
(WebKit::HistoryEntryDataEncoder::growCapacity):
(WebKit::encodeFrameStateNode):
(WebKit::encodeLegacySessionHistoryEntryData):
(WebKit::decodeSessionHistoryEntryData):
(WebKit::decodeLegacySessionHistoryEntryData):
- UIProcess/mac/LegacySessionStateCoding.h:
Source/WTF:
- wtf/MallocPtr.h:
(WTF::MallocPtr::malloc):
(WTF::MallocPtr::realloc):
- 2:25 PM Changeset in webkit [170598] by
-
- 3 edits in trunk/Source/WebCore
Use non-thumb registers in armv7 css jit.
https://bugs.webkit.org/show_bug.cgi?id=134450
Reviewed by Geoff Garen.
- cssjit/RegisterAllocator.h:
(WebCore::RegisterAllocator::isValidRegister):
(WebCore::RegisterAllocator::isCallerSavedRegister):
Use additional general purpose registers.
- cssjit/SelectorCompiler.cpp:
(WebCore::SelectorCompiler::SelectorCodeGenerator::compile):
(WebCore::SelectorCompiler::SelectorCodeGenerator::generateSelectorChecker):
Always succeed when compiling css selectors.
- 2:21 PM Changeset in webkit [170597] by
-
- 5 edits2 adds in trunk
Tiles on bottom of screen are not always allocated when necessary
https://bugs.webkit.org/show_bug.cgi?id=134272
Reviewed by Simon Fraser.
Source/WebCore:
The initial visibleRect is in the coordinate of the root layer, so its origin
is at the top left of the view. The initial rect we were using doesn't
include the contents inset, so it was too short, which was causing tiles near
the bottom of the screen to not always be allocated if the tile threshold was
close to the bottom of the view. Instead, we want to include the contents
inset size so the visible rect includes the entire view.
GraphicsLayerCA::recursiveCommitChanges() takes care of mapping and cropping
the visible rect into the tiled layer's coordinate system, at which point it
is used for visible tile logic.
Test: platform/mac-wk2/tiled-drawing/visible-rect-content-inset.html
- platform/graphics/ca/TileCoverageMap.cpp:
(WebCore::TileCoverageMap::update):
- rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::flushPendingLayerChanges):
LayoutTests:
Apply a content inset, then dump visible rects.
- platform/mac-wk2/tiled-drawing/visible-rect-content-inset-expected.txt: Added.
- platform/mac-wk2/tiled-drawing/visible-rect-content-inset.html: Added.
- compositing/visible-rect/iframe-no-layers.html: For WK2, the size of the visibleRect
includes the size of the border of the iframe. In order to elliminate platform
differences, set this border to 0.
- 1:57 PM Changeset in webkit [170596] by
-
- 2 edits in trunk/Tools
DRT in iOS sim crashes while loading fonts
https://bugs.webkit.org/show_bug.cgi?id=134451
Reviewed by Geoffrey Garen.
The magic linker symbols had the same name, and were therefore being unified
to a single font file. Then, adding the same file twice was failing.
- DumpRenderTree/mac/DumpRenderTree.mm:
(activateFontsIOS):
- 12:01 PM Changeset in webkit [170595] by
-
- 3 edits in trunk/Source/WebCore
[Win] Unreviewed project gardening.
- WebCore.vcxproj/WebCore.vcxproj: Clean up path information so
files appear in appropriate places in project view.
- WebCore.vcxproj/WebCore.vcxproj.filters: Ditto.
- 11:48 AM Changeset in webkit [170594] by
-
- 4 edits in trunk/Source/JavaScriptCore
Avoid copying closed variables vector; actually use move semantics
Rubber-stamped by Oliver Hunt.
Currently we always copy the closed variables vector passed by Parser::closedVariables()
to ProgramNode::setClosedVariables() because these member functions return and take a const
rvalue reference, respectively. Instead, these member functions should take an return a non-
constant rvalue reference so that we actually move the closed variables vector from the Parser
object to the Node object.
- parser/Nodes.cpp:
(JSC::ProgramNode::setClosedVariables): Remove const qualifier for argument.
- parser/Nodes.h:
(JSC::ScopeNode::setClosedVariables): Ditto.
- parser/Parser.h:
(JSC::Parser::closedVariables): Remove const qualifier on return type.
(JSC::parse): Remove extraneous call to std::move(). Calling std::move() is unnecessary here
because Parser::closedVariables() returns an rvalue reference.
- 11:27 AM Changeset in webkit [170593] by
-
- 2 edits in tags/Safari-538.43.40/Source/WebKit2
Merged r170569. <rdar://problem/17500555>
- 11:24 AM Changeset in webkit [170592] by
-
- 2 edits in tags/Safari-538.43.40/Source/WebCore
Merged r170533. <rdar://problem/16801069>
- 11:22 AM Changeset in webkit [170591] by
-
- 2 edits in trunk/LayoutTests
Reduce memory required for js/typedarray-zero-size.js
https://bugs.webkit.org/show_bug.cgi?id=134448
Reviewed by Mark Hahnenberg.
Reduce the loop count from 10000 to 4000 to further reduce memory requirements
on older devices.
- js/script-tests/typedarray-zero-size.js:
- 11:06 AM Changeset in webkit [170590] by
-
- 5 edits in trunk/Source/WebCore
Remove redundant repaintCompositedLayers() parameter and its dependencies.
https://bugs.webkit.org/show_bug.cgi?id=134431
Reviewed by Darin Adler.
No change in functionality.
- rendering/RenderLayer.cpp:
(WebCore::RenderLayer::convertToPixelSnappedLayerCoords): Deleted.
(WebCore::RenderLayer::convertToLayerCoords): Deleted.
- rendering/RenderLayer.h:
- rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::repaintCompositedLayers):
(WebCore::RenderLayerCompositor::recursiveRepaintLayer):
- rendering/RenderLayerCompositor.h:
- 10:57 AM Changeset in webkit [170589] by
-
- 10 edits1 copy in trunk/Source/JavaScriptCore
JSContext Inspection: Provide a way to use a non-Main RunLoop for Inspector JavaScript Evaluations
https://bugs.webkit.org/show_bug.cgi?id=134371
Reviewed by Timothy Hatcher.
- API/JSContextPrivate.h:
- API/JSContext.mm:
(-[JSContext _debuggerRunLoop]):
(-[JSContext _setDebuggerRunLoop:]):
Private API for setting the CFRunLoop for a debugger to evaluate in.
- API/JSContextRefInternal.h: Added.
- API/JSContextRef.cpp:
(JSGlobalContextGetDebuggerRunLoop):
(JSGlobalContextSetDebuggerRunLoop):
Internal API for setting a CFRunLoop on a JSContextRef.
Set this on the debuggable.
- inspector/remote/RemoteInspectorDebuggable.h:
- inspector/remote/RemoteInspectorDebuggableConnection.h:
(Inspector::RemoteInspectorBlock::RemoteInspectorBlock):
(Inspector::RemoteInspectorBlock::~RemoteInspectorBlock):
(Inspector::RemoteInspectorBlock::operator=):
(Inspector::RemoteInspectorBlock::operator()):
Moved into the header.
- runtime/JSGlobalObject.h:
(JSC::JSGlobalObject::inspectorDebuggable):
Lets store the RunLoop on the debuggable instead of this core
platform agnostic class, so expose the debuggable.
- inspector/remote/RemoteInspectorDebuggableConnection.mm:
(Inspector::RemoteInspectorHandleRunSourceGlobal):
(Inspector::RemoteInspectorQueueTaskOnGlobalQueue):
(Inspector::RemoteInspectorInitializeGlobalQueue):
Rename the global functions for clarity.
(Inspector::RemoteInspectorHandleRunSourceWithInfo):
Handler for private run loops.
(Inspector::RemoteInspectorDebuggableConnection::RemoteInspectorDebuggableConnection):
(Inspector::RemoteInspectorDebuggableConnection::~RemoteInspectorDebuggableConnection):
(Inspector::RemoteInspectorDebuggableConnection::dispatchAsyncOnDebuggable):
(Inspector::RemoteInspectorDebuggableConnection::setupRunLoop):
(Inspector::RemoteInspectorDebuggableConnection::teardownRunLoop):
(Inspector::RemoteInspectorDebuggableConnection::queueTaskOnPrivateRunLoop):
Setup and teardown and use private run loop sources if the debuggable needs it.
- 10:42 AM Changeset in webkit [170588] by
-
- 5 edits in tags/Safari-538.43.40/Source
Versioning.
- 10:39 AM Changeset in webkit [170587] by
-
- 1 copy in tags/Safari-538.43.40
New Tag.
- 10:37 AM Changeset in webkit [170586] by
-
- 2 edits in trunk/LayoutTests
Reduce memory required for js/typedarray-zero-size.js
https://bugs.webkit.org/show_bug.cgi?id=134448
Reviewed by Darin Adler.
Making the loop count down reduces memory usage from ~190MB to ~22MB while
still testing the original issue.
- js/script-tests/typedarray-zero-size.js:
- 9:59 AM Changeset in webkit [170585] by
-
- 11 edits in trunk/Source/WebCore
Remove webkit prefix from pointer lock.
https://bugs.webkit.org/show_bug.cgi?id=134312
Patch by Alex Christensen <achristensen@webkit.org> on 2014-06-30
Reviewed by Dean Jackson.
- dom/Document.cpp:
(WebCore::Document::exitPointerLock):
(WebCore::Document::pointerLockElement):
(WebCore::Document::webkitExitPointerLock): Deleted.
(WebCore::Document::webkitPointerLockElement): Deleted.
- dom/Document.h:
- dom/Document.idl:
- dom/Element.cpp:
(WebCore::Element::requestPointerLock):
(WebCore::Element::webkitRequestPointerLock): Deleted.
- dom/Element.h:
- dom/Element.idl:
- dom/EventNames.h:
- dom/MouseEvent.idl:
- dom/MouseRelatedEvent.h:
(WebCore::MouseRelatedEvent::movementX):
(WebCore::MouseRelatedEvent::movementY):
(WebCore::MouseRelatedEvent::webkitMovementX): Deleted.
(WebCore::MouseRelatedEvent::webkitMovementY): Deleted.
- page/PointerLockController.cpp:
(WebCore::PointerLockController::PointerLockController):
Initialize all data members.
(WebCore::PointerLockController::requestPointerLock):
(WebCore::PointerLockController::didAcquirePointerLock):
(WebCore::PointerLockController::didNotAcquirePointerLock):
(WebCore::PointerLockController::didLosePointerLock):
- 8:13 AM Changeset in webkit [170584] by
-
- 2 edits in trunk/Source/WebCore
U16_IS_SURROGATE should not be called with a UChar32
https://bugs.webkit.org/show_bug.cgi?id=134440
Reviewed by Darin Adler.
No new tests because there is no behavior change.
- platform/graphics/mac/ComplexTextController.cpp:
(WebCore::advanceByCombiningCharacterSequence):
- 7:48 AM Changeset in webkit [170583] by
-
- 3 edits in trunk/Source/JavaScriptCore
Add missing ENABLE(DFG_JIT) guards
https://bugs.webkit.org/show_bug.cgi?id=134444
Patch by Tibor Meszaros <tmeszaros.u-szeged@partner.samsung.com> on 2014-06-30
Reviewed by Darin Adler.
- dfg/DFGFunctionWhitelist.cpp:
- dfg/DFGFunctionWhitelist.h:
- 7:45 AM Changeset in webkit [170582] by
-
- 3 edits in trunk/Tools
Remove Qt cruft from webkitpy
https://bugs.webkit.org/show_bug.cgi?id=134445
Patch by Tibor Meszaros <tmeszaros.u-szeged@partner.samsung.com> on 2014-06-30
Reviewed by Csaba Osztrogonác.
- Scripts/webkitpy/common/config/ports.py:
(DeprecatedPort.port):
(QtPort): Deleted.
(QtPort.build_webkit_command): Deleted.
(QtPort.run_webkit_tests_command): Deleted.
(QtWK2Port): Deleted.
(QtWK2Port.build_webkit_command): Deleted.
(QtWK2Port.run_webkit_tests_command): Deleted.
- Scripts/webkitpy/common/config/ports_unittest.py:
(DeprecatedPortTest): Deleted.
(DeprecatedPortTest.test_qt_port): Deleted.
(DeprecatedPortTest.test_qt_wk2_port): Deleted.
- 1:46 AM Changeset in webkit [170581] by
-
- 5 edits in trunk/Source/WebCore
[EFL] Remove m_evasObject from Widget
https://bugs.webkit.org/show_bug.cgi?id=134442
Reviewed by Gyuyoung Kim.
Since WebKit1/EFL was dropped, we don't need to keep the m_evasObject
in the Widget.cpp because it's not used for WebKit2/Efl.
- platform/Widget.h:
(WebCore::Widget::evasObject): Deleted.
- platform/efl/ScrollbarEfl.cpp:
Removed WebKit1/Efl specific code which uses m_evasObject.
It will be a dummy until WebKit2/Efl support proper scrollbar implementation.
(WebCore::ScrollbarEfl::ScrollbarEfl):
(WebCore::ScrollbarEfl::~ScrollbarEfl):
(WebCore::ScrollbarEfl::setParent):
(WebCore::ScrollbarEfl::frameRectsChanged):
(WebCore::ScrollbarEfl::invalidate):
(ScrollbarEfl::ScrollbarEfl): Deleted.
(ScrollbarEfl::~ScrollbarEfl): Deleted.
(scrollbarEflEdjeMessage): Deleted.
(ScrollbarEfl::setParent): Deleted.
(ScrollbarEfl::updateThumbPosition): Deleted.
(ScrollbarEfl::updateThumbProportion): Deleted.
(ScrollbarEfl::updateThumbPositionAndProportion): Deleted.
(ScrollbarEfl::frameRectsChanged): Deleted.
(ScrollbarEfl::invalidate): Deleted.
- platform/efl/ScrollbarEfl.h:
Removed unnecessary methods. In addition, applied final, override keyword.
(WebCore::ScrollbarEfl::handleMouseMoveEvent): Deleted.
(WebCore::ScrollbarEfl::handleMouseOutEvent): Deleted.
(WebCore::ScrollbarEfl::handleMousePressEvent): Deleted.
(WebCore::ScrollbarEfl::handleMouseReleaseEvent): Deleted.
- platform/efl/WidgetEfl.cpp:
(WebCore::Widget::Widget):
(WebCore::Widget::setEvasObject): Deleted.
- 1:35 AM Changeset in webkit [170580] by
-
- 3 edits in trunk/Source/WebCore
Fix deprecated warning with bison 3
https://bugs.webkit.org/show_bug.cgi?id=133791
Patch by Eva Balazsfalvi <evab.u-szeged@partner.samsung.com> on 2014-06-30
Reviewed by Csaba Osztrogonác.
No new tests because no functional changes.
- css/CSSGrammar.y.in:
- xml/XPathGrammar.y:
- 1:32 AM Changeset in webkit [170579] by
-
- 3 edits in trunk/Source/WebCore
Remove HarfBuzzShaper::setDrawRange()
https://bugs.webkit.org/show_bug.cgi?id=134243
Reviewed by Gyuyoung Kim.
setDrawRange() is not used anywhere. Remove it.
And m_fromIndex and m_toIndex are changed by only setDrawRange().
So remove them too and use the default values instead of them.
No new tests, no behavior changes.
- platform/graphics/harfbuzz/HarfBuzzShaper.cpp:
(WebCore::HarfBuzzShaper::HarfBuzzShaper):
(WebCore::HarfBuzzShaper::fillGlyphBufferFromHarfBuzzRun):
currentCharacterIndex is unsigned, always greater than 0. Clean up if statements.
(WebCore::HarfBuzzShaper::setDrawRange): Deleted.
- platform/graphics/harfbuzz/HarfBuzzShaper.h:
Jun 29, 2014:
- 11:36 PM Changeset in webkit [170578] by
-
- 4 edits in trunk/Source/WebCore
Create NSURLRequest lazily when USE(CFNETWORK) is enabled
https://bugs.webkit.org/show_bug.cgi?id=134441
Reviewed by Sam Weinig.
No new tests. Should be covered by exsting tests.
- platform/network/cf/ResourceRequest.h:
(WebCore::ResourceRequest::ResourceRequest):
- platform/network/cf/ResourceRequestCFNet.cpp:
(WebCore::ResourceRequest::doUpdatePlatformRequest):
(WebCore::ResourceRequest::doUpdatePlatformHTTPBody):
(WebCore::ResourceRequest::setStorageSession):
- platform/network/cocoa/ResourceRequestCocoa.mm:
(WebCore::ResourceRequest::nsURLRequest):
- 9:43 PM Changeset in webkit [170577] by
-
- 2 edits2 deletes in trunk/Source/WebCore
[EFL] Remove netscape plugin implementation from WebCore
https://bugs.webkit.org/show_bug.cgi?id=134438
Reviewed by Gyuyoung Kim.
Because WebKit1/Efl was dropped, PluginPackageEfl.cpp and PluginViewEfl.cpp are not necessary.
- PlatformEfl.cmake:
- plugins/efl/PluginPackageEfl.cpp: Removed.
- plugins/efl/PluginViewEfl.cpp: Removed.
- plugins/x11/PluginViewX11.cpp: Removed.
This file is to share common code between WebKit1/Efl and WebKit1/GTk.
Now, both ports are dropped WebKit1 support and the GTK port already removed the related files at r167016.
- 9:07 PM Changeset in webkit [170576] by
-
- 30 edits19 adds in trunk
Add support for HTMLImageElement's sizes attribute
https://bugs.webkit.org/show_bug.cgi?id=133620
Reviewed by Dean Jackson.
.:
Added an ENABLE_PICTURE_SIZES compile flag.
- Source/cmake/WebKitFeatures.cmake:
- Source/cmakeconfig.h.cmake:
Source/JavaScriptCore:
Added an ENABLE_PICTURE_SIZES compile flag.
- Configurations/FeatureDefines.xcconfig:
Source/WebCore:
Tests: fast/dom/HTMLImageElement/sizes/image-sizes-1x.html
fast/dom/HTMLImageElement/sizes/image-sizes-2x.html
fast/dom/HTMLImageElement/sizes/image-sizes-js-change.html
fast/dom/HTMLImageElement/sizes/image-sizes-js-innerhtml.html
http/tests/loading/sizes/preload-image-sizes-2x.html
http/tests/loading/sizes/preload-image-sizes.html
This patch adds support for HTMLImageElement's sizes attribute and the
related srcset extended syntax as defined in
http://picture.responsiveimages.org/.
This sizes attribute syntax is added to the CSSGrammar and parsed by
the CSSParser.
The SourceSizeList class is generated by the parser, and used to get
the final source size.
HTMLImageElement and HTMLPreloadScanner send this value to
HTMLSrcsetParser.
HTMLSrcsetParser uses this value in order to pick the right resource.
- CMakeLists.txt: Added css/SourceSizeList.cpp.
- Configurations/FeatureDefines.xcconfig: Added the PICTURE_SIZES flag.
- WebCore.vcxproj/WebCore.vcxproj: Added css/SourceSizeList.*.
- WebCore.vcxproj/WebCore.vcxproj.filters: Added css/SourceSizeList.*.
- WebCore.xcodeproj/project.pbxproj: Added css/SourceSizeList.*.
- css/CSSGrammar.y.in: Added 4 grammar rules that define a Media condition,
Source size list, Source size length and source size.
- css/CSSParser.cpp:
(WebCore::CSSParser::parseSizesAttribute): Parse the sizes attribute.
(WebCore::CSSParser::detectAtToken): Set the token to SIZESATTR.
- css/CSSParser.h:
- css/SourceSizeList.cpp: Added. Defined the SourceSize and SourceSizeList classes.
(WebCore::SourceSize::match):
(WebCore::computeLength):
(WebCore::defaultValue):
(WebCore::SourceSize::length):
(WebCore::SourceSizeList::parseSizesAttribute):
(WebCore::SourceSizeList::getEffectiveSize):
- css/SourceSizeList.h: Added.
(WebCore::SourceSize::SourceSize):
(WebCore::SourceSizeList::append):
- html/HTMLImageElement.cpp: Integrated sizes attribute parsing and added currentSrc.
(WebCore::HTMLImageElement::setBestFitURLAndDPRFromImageCandidate):
(WebCore::HTMLImageElement::parseAttribute):
(WebCore::HTMLImageElement::currentSrc):
- html/HTMLImageElement.h:
- html/HTMLImageElement.idl:
- html/parser/HTMLDocumentParser.cpp: Added information required for sizes parsing to
HTMLPreloadScanner calls.
(WebCore::HTMLDocumentParser::pumpTokenizer):
(WebCore::HTMLDocumentParser::insert):
(WebCore::HTMLDocumentParser::append):
(WebCore::HTMLDocumentParser::appendCurrentInputStreamToPreloadScannerAndScan):
- html/parser/HTMLPreloadScanner.cpp: Integrated sizes attribute parsing.
(WebCore::TokenPreloadScanner::StartTagScanner::processAttributes):
(WebCore::TokenPreloadScanner::StartTagScanner::processAttribute):
(WebCore::TokenPreloadScanner::scan):
(WebCore::HTMLPreloadScanner::scan):
- html/parser/HTMLPreloadScanner.h:
- html/parser/HTMLSrcsetParser.cpp: Added 'w' descriptor parsing sizes based normalization.
(WebCore::parseDescriptors):
(WebCore::pickBestImageCandidate):
(WebCore::bestFitSourceForImageAttributes):
- html/parser/HTMLSrcsetParser.h:
Source/WebKit/mac:
Added an ENABLE_PICTURE_SIZES compile flag.
- Configurations/FeatureDefines.xcconfig:
Source/WebKit2:
Added an ENABLE_PICTURE_SIZES compile flag.
- Configurations/FeatureDefines.xcconfig:
Tools:
Added an ENABLE_PICTURE_SIZES compile flag.
- Scripts/webkitperl/FeatureList.pm:
LayoutTests:
These layout tests check that the sizes attribute and the related
srcset syntax are well supported, that the right resource is displayed,
that the currentSrc attribute is well supported and also that the right
resource is preloaded.
- fast/dom/HTMLImageElement/resources/currentSrcHelper.js: Added.
(fileName):
(currentSrcFileName):
- fast/dom/HTMLImageElement/resources/image-set-4x.png: Added.
- fast/dom/HTMLImageElement/sizes/image-sizes-1x-expected.txt: Added.
- fast/dom/HTMLImageElement/sizes/image-sizes-1x.html: Added.
- fast/dom/HTMLImageElement/sizes/image-sizes-2x-expected.txt: Added.
- fast/dom/HTMLImageElement/sizes/image-sizes-2x.html: Added.
- fast/dom/HTMLImageElement/sizes/image-sizes-js-change-expected.txt: Added.
- fast/dom/HTMLImageElement/sizes/image-sizes-js-change.html: Added.
- fast/dom/HTMLImageElement/sizes/image-sizes-js-innerhtml-expected.txt: Added.
- fast/dom/HTMLImageElement/sizes/image-sizes-js-innerhtml.html: Added.
- http/tests/loading/sizes/preload-image-sizes-2x-expected.txt: Added.
- http/tests/loading/sizes/preload-image-sizes-2x.html: Added.
- http/tests/loading/sizes/preload-image-sizes-expected.txt: Added.
- http/tests/loading/sizes/preload-image-sizes.html: Added.
- 2:51 PM Changeset in webkit [170575] by
-
- 4 edits in trunk/Source/WebCore
No need to lazily initialize ResourceResponse internals when accessing timing data.
<https://webkit.org/b/134437>
<rdar://problem/17499876>
When building with ENABLE(WEB_TIMING), the last thing that would happen when
serializing a ResourceResponse in the network process is that we'd retrieve
the ResourceLoadTiming to encode it as part of the response. Doing so would
trigger the lazy instantiation of ResourceResponse's internal data structures.
Since timing data is not actually lazily instantiated, we can just not do that.
This shaves off ~550ms of network process main thread time on PLT, reducing
response latency by not doing unnecessary stuff before letting the web process
look at the downloaded data.
Reviewed by Gavin Barraclough.
- WebCore.exp.in:
- platform/network/ResourceResponseBase.cpp:
(WebCore::ResourceResponseBase::adopt):
(WebCore::ResourceResponseBase::resourceLoadTiming): Deleted.
(WebCore::ResourceResponseBase::setResourceLoadTiming): Deleted.
- platform/network/ResourceResponseBase.h:
(WebCore::ResourceResponseBase::resourceLoadTiming):
- 1:59 PM Changeset in webkit [170574] by
-
- 4 edits1 copy2 adds in trunk/Source/WebCore
Refactor ResourceRequest into Cocoa and iOS specific files
https://bugs.webkit.org/show_bug.cgi?id=134430
Reviewed by Andreas Kling.
No new tests because no functional changes.
- WebCore.xcodeproj/project.pbxproj:
- platform/network/cf/ResourceRequestCFNet.cpp:
(WebCore::ResourceRequest::applyWebArchiveHackForMail): Deleted.
- platform/network/cocoa/ResourceRequestCocoa.mm: Copied from Source/WebCore/platform/network/mac/ResourceRequestMac.mm.
(WebCore::ResourceRequest::nsURLRequest):
(WebCore::ResourceRequest::cfURLRequest):
(WebCore::ResourceRequest::doUpdateResourceRequest):
(WebCore::ResourceRequest::doUpdateResourceHTTPBody):
(WebCore::ResourceRequest::doUpdatePlatformRequest):
(WebCore::ResourceRequest::doUpdatePlatformHTTPBody):
(WebCore::ResourceRequest::updateFromDelegatePreservingOldProperties):
(WebCore::ResourceRequest::applyWebArchiveHackForMail):
(WebCore::ResourceRequest::setStorageSession):
- platform/network/ios/ResourceRequestIOS.mm: Added.
(WebCore::ResourceRequest::useQuickLookResourceCachingQuirks):
(WebCore::ResourceRequest::ResourceRequest):
(WebCore::ResourceRequest::updateNSURLRequest):
- platform/network/mac/ResourceRequestMac.mm:
(WebCore::initQuickLookResourceCachingQuirks):
(WebCore::ResourceRequest::useQuickLookResourceCachingQuirks):
(WebCore::ResourceRequest::ResourceRequest):
(WebCore::ResourceRequest::updateNSURLRequest):
(WebCore::ResourceRequest::applyWebArchiveHackForMail):
(WebCore::ResourceRequest::nsURLRequest): Deleted.
(WebCore::ResourceRequest::cfURLRequest): Deleted.
(WebCore::ResourceRequest::doUpdateResourceRequest): Deleted.
(WebCore::ResourceRequest::doUpdateResourceHTTPBody): Deleted.
(WebCore::ResourceRequest::doUpdatePlatformRequest): Deleted.
(WebCore::ResourceRequest::doUpdatePlatformHTTPBody): Deleted.
(WebCore::ResourceRequest::updateFromDelegatePreservingOldProperties): Deleted.
(WebCore::ResourceRequest::setStorageSession): Deleted.
- 10:10 AM Changeset in webkit [170573] by
-
- 3 edits2 adds in trunk
TextCodecICU::encode turns the whole string as yen signs if there is any backslash in it
https://bugs.webkit.org/show_bug.cgi?id=133678
Patch by Youenn Fablet <youenn.fablet@crf.canon.fr> on 2014-06-29
Reviewed by Alexey Proskuryakov.
Source/WebCore:
Test: fast/encoding/backslash-encoding-jp.html
- platform/text/TextCodecICU.cpp:
(WebCore::TextCodecICU::encode): fixed the copy of characters other than backslash.
LayoutTests:
- fast/encoding/backslash-encoding-jp-expected.txt: Added.
- fast/encoding/backslash-encoding-jp.html: Added.
- 8:27 AM BuildingQtOnWindows edited by
- small "typo" (diff)
- 8:18 AM BuildingQtOnWindows edited by
- Added the SQLite dependency information, added some notes and a bit … (diff)
Jun 28, 2014:
- 7:44 PM Changeset in webkit [170572] by
-
- 2 edits in trunk/Source/WebCore
[iOS] Pinching into a video (fullscreen gesture) leaves page zoomed in
https://bugs.webkit.org/show_bug.cgi?id=134433
<rdar://problem/16630794>
Reviewed by Eric Carlson.
Detect the number of target touches (those directly on the
video element) and disable the page zoom if it looks
like the user is trying to trigger fullscreen.
- Modules/mediacontrols/mediaControlsiOS.js:
(ControllerIOS.prototype.handleBaseGestureStart): If we are seeing
two touches in this video element, don't trigger fullscreen. The
exception is if we haven't yet started playback.
(ControllerIOS.prototype.handleBaseGestureChange): Don't allow fullscreen
if we haven't started playback (iOS 7 behaviour).
(ControllerIOS.prototype.handleWrapperTouchStart): Keep track of the
number of targetted touches.
- 4:49 PM Changeset in webkit [170571] by
-
- 17 edits in branches/ftlopt/Source/JavaScriptCore
[ftlopt] Reduce the GC's influence on optimization decisions
https://bugs.webkit.org/show_bug.cgi?id=134427
Reviewed by Oliver Hunt.
This is a slight speed-up on some platforms, that arises from a bunch of fixes that I made
while trying to make the GC keep more structures alive
(https://bugs.webkit.org/show_bug.cgi?id=128072).
The fixes are, roughly:
- If the GC clears an inline cache, then this no longer causes the IC to be forever polymorphic.
- If we exit in inlined code into a function that tries to OSR enter, then we jettison sooner.
- Some variables being uninitialized led to rage-recompilations.
This is a pretty strong step in the direction of keeping more Structures alive and not
blowing away code just because a Structure died. But, it seems like there is still a slight
speed-up to be had from blowing away code that references dead Structures.
- bytecode/CodeBlock.cpp:
(JSC::CodeBlock::dumpAssumingJITType):
(JSC::shouldMarkTransition):
(JSC::CodeBlock::propagateTransitions):
(JSC::CodeBlock::determineLiveness):
- bytecode/GetByIdStatus.cpp:
(JSC::GetByIdStatus::computeForStubInfo):
- bytecode/PutByIdStatus.cpp:
(JSC::PutByIdStatus::computeForStubInfo):
- dfg/DFGCapabilities.cpp:
(JSC::DFG::isSupportedForInlining):
(JSC::DFG::mightInlineFunctionForCall):
(JSC::DFG::mightInlineFunctionForClosureCall):
(JSC::DFG::mightInlineFunctionForConstruct):
- dfg/DFGCapabilities.h:
- dfg/DFGCommonData.h:
- dfg/DFGDesiredWeakReferences.cpp:
(JSC::DFG::DesiredWeakReferences::reallyAdd):
- dfg/DFGOSREntry.cpp:
(JSC::DFG::prepareOSREntry):
- dfg/DFGOSRExitCompilerCommon.cpp:
(JSC::DFG::handleExitCounts):
- dfg/DFGOperations.cpp:
- dfg/DFGOperations.h:
- ftl/FTLForOSREntryJITCode.cpp:
(JSC::FTL::ForOSREntryJITCode::ForOSREntryJITCode): These variables being uninitialized is benign in terms of correctness but can sometimes cause rage-recompilations. For some reason it took this patch to reveal this.
- ftl/FTLOSREntry.cpp:
(JSC::FTL::prepareOSREntry):
- runtime/Executable.cpp:
(JSC::ExecutableBase::destroy):
(JSC::NativeExecutable::destroy):
(JSC::ScriptExecutable::ScriptExecutable):
(JSC::ScriptExecutable::destroy):
(JSC::ScriptExecutable::installCode):
(JSC::EvalExecutable::EvalExecutable):
(JSC::ProgramExecutable::ProgramExecutable):
- runtime/Executable.h:
(JSC::ScriptExecutable::setDidTryToEnterInLoop):
(JSC::ScriptExecutable::didTryToEnterInLoop):
(JSC::ScriptExecutable::addressOfDidTryToEnterInLoop):
(JSC::ScriptExecutable::ScriptExecutable): Deleted.
- runtime/StructureInlines.h:
(JSC::Structure::storedPrototypeObject):
(JSC::Structure::storedPrototypeStructure):
- 3:39 PM Changeset in webkit [170570] by
-
- 10 edits in trunk/Source/WebKit2
[iOS][WK2] PDFs never load inline again after the Web process crashes
https://bugs.webkit.org/show_bug.cgi?id=134432
<rdar://problem/17484205>
Reviewed by Dan Bernstein.
After a crash, the WebPage never gets mimeTypesWithCustomContentProviders filled back in.
- Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode):
(WebKit::WebPageCreationParameters::decode):
- Shared/WebPageCreationParameters.h:
- UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::creationParameters):
- WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::WebPage):
Add mimeTypesWithCustomContentProviders to WebPageCreationParameters.
- UIProcess/Cocoa/WKWebViewContentProviderRegistry.h:
- UIProcess/Cocoa/WKWebViewContentProviderRegistry.mm:
(-[WKWebViewContentProviderRegistry addPage:]):
We don't need to send the MIME types across when a page is added, because it already got them in its creation parameters.
(-[WKWebViewContentProviderRegistry removePage:]):
(-[WKWebViewContentProviderRegistry _mimeTypesWithCustomContentProviders]):
Return a vector of all registered MIME types.
- UIProcess/ios/PageClientImplIOS.h:
- UIProcess/ios/PageClientImplIOS.mm:
(WebKit::PageClientImpl::mimeTypesWithCustomContentProviders):
- UIProcess/PageClient.h:
Add (iOS only for now) mimeTypesWithCustomContentProviders to PageClient.
- 3:31 PM Changeset in webkit [170569] by
-
- 2 edits in trunk/Source/WebKit2
REGRESSION: WebPageProxy::attributedSubstringForCharacterRangeAsync never calls its callback function
https://bugs.webkit.org/show_bug.cgi?id=134429
Reviewed by Tim Horton.
- UIProcess/mac/WebPageProxyMac.mm:
(WebKit::WebPageProxy::attributedSubstringForCharacterRangeAsync):
- 7:50 AM Changeset in webkit [170568] by
-
- 7 edits in trunk/Source
[iOS WK2] position:fixed inside accelerated overflow:scroll is jumpy
https://bugs.webkit.org/show_bug.cgi?id=134426
<rdar://problem/17474523>
Reviewed by Tim Horton.
After committing a new layer tree (with possibly stale position:fixed layer
positions), we need the scrolling tree to update those positions based on
the current scroll offset.
Source/WebCore:
Give ScrollingTreeScrollingNode an implementation of updateLayersAfterAncestorChange()
which is required to update fixed/sticky child nodes.
- WebCore.exp.in:
- page/scrolling/ScrollingTreeScrollingNode.cpp:
(WebCore::ScrollingTreeScrollingNode::updateLayersAfterAncestorChange):
- page/scrolling/ScrollingTreeScrollingNode.h:
Source/WebKit2:
To achieve that, implement ScrollingTreeOverflowScrollingNodeIOS::updateLayersAfterAncestorChange()
and have it add to the cumulative delta the difference between the last committed scroll
position and the current scroll position.
Also make sure that ScrollingTreeOverflowScrollingNodeIOS doesn't call back to scrollViewDidScroll()
when we're updating its scroll position inside a scrolling tree commit.
- UIProcess/Scrolling/ios/ScrollingTreeOverflowScrollingNodeIOS.h:
- UIProcess/Scrolling/ios/ScrollingTreeOverflowScrollingNodeIOS.mm:
(WebKit::ScrollingTreeOverflowScrollingNodeIOS::ScrollingTreeOverflowScrollingNodeIOS):
(WebKit::ScrollingTreeOverflowScrollingNodeIOS::updateAfterChildren):
(WebKit::ScrollingTreeOverflowScrollingNodeIOS::updateLayersAfterAncestorChange):
(WebKit::ScrollingTreeOverflowScrollingNodeIOS::scrollViewDidScroll):
- 2:06 AM Changeset in webkit [170567] by
-
- 11 edits in trunk/Source/WebCore
[GTK] Use public getter/setter in GObject DOM bindings properties implementation.
https://bugs.webkit.org/show_bug.cgi?id=134390.
Patch by Juan A. Suarez Romero <jasuarez@igalia.com> and Carlos Garcia Campos <cgarcia@igalia.com> on 2014-06-28
Reviewed by Carlos Garcia Campos.
The current GObject DOM bindings generator is duplicating code when
implementing the get/set_property methods and the public
getters/setters, instead of making one invoking the other.
This commit changes this behaviour so the code is not duplicated,
making it easier to maintain.
No new tests needed.
- bindings/scripts/CodeGeneratorGObject.pm:
(GenerateProperty):
(GenerateProperties):
(WriteData):
(GetCoreObject): Deleted.
- bindings/scripts/test/GObject/WebKitDOMTestActiveDOMObject.cpp:
(webkit_dom_test_active_dom_object_get_property):
- bindings/scripts/test/GObject/WebKitDOMTestEventConstructor.cpp:
(webkit_dom_test_event_constructor_get_property):
- bindings/scripts/test/GObject/WebKitDOMTestException.cpp:
(webkit_dom_test_exception_get_property):
- bindings/scripts/test/GObject/WebKitDOMTestInterface.cpp:
(webkit_dom_test_interface_set_property):
(webkit_dom_test_interface_get_property):
- bindings/scripts/test/GObject/WebKitDOMTestNondeterministic.cpp:
(webkit_dom_test_nondeterministic_set_property):
(webkit_dom_test_nondeterministic_get_property):
- bindings/scripts/test/GObject/WebKitDOMTestObj.cpp:
(webkit_dom_test_obj_set_property):
(webkit_dom_test_obj_get_property):
- bindings/scripts/test/GObject/WebKitDOMTestSerializedScriptValueInterface.cpp:
(webkit_dom_test_serialized_script_value_interface_get_property):
- bindings/scripts/test/GObject/WebKitDOMTestTypedefs.cpp:
(webkit_dom_test_typedefs_set_property):
(webkit_dom_test_typedefs_get_property):
- bindings/scripts/test/GObject/WebKitDOMattribute.cpp:
(webkit_dom_test_exception_get_property):
(webkit_dom_attribute_get_property):
- 1:44 AM Changeset in webkit [170566] by
-
- 3 edits2 adds in trunk
[New Multicolumn] Elements with rounded corners and overflow:hidden do not properly clip their content
https://bugs.webkit.org/show_bug.cgi?id=133941
Reviewed by Darin Adler.
Source/WebCore:
When having a multicol element inside an element with overflow:hidden and border-radius,
the fragments representing the columns need to have the border radius set also, to ensure
proper clipping.
Test: fast/multicol/newmulticol/multicol-clip-rounded-corners.html
- rendering/RenderLayer.cpp:
(WebCore::RenderLayer::collectFragments):
LayoutTests:
Added test for multicol element inside an element with overflow:hidden and border radius.
- fast/multicol/newmulticol/multicol-clip-rounded-corners-expected.html: Added.
- fast/multicol/newmulticol/multicol-clip-rounded-corners.html: Added.
