Timeline
Apr 28, 2014:
- 9:20 PM Changeset in webkit [167923] by
-
- 3 edits in trunk/Source/WebKit2
<rdar://problem/16750708> REGRESSION (r159358): Crash in InjectedBundleHitTestResult::frame
https://bugs.webkit.org/show_bug.cgi?id=132318
Reviewed by Andy Estes.
- UIProcess/Cocoa/NavigationState.mm:
(WebKit::NavigationState::PolicyClient::decidePolicyForNavigationAction): Fixed a typo: set
the source frame, not the destination frame.
- WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::dispatchDecidePolicyForNavigationAction): Changed to get the
frame for the action not from a hit test result, because some link activation actions are
not mouse events, but rather from the event’s target.
- 8:10 PM Changeset in webkit [167922] by
-
- 2 edits in trunk/Source/JavaScriptCore
Simplify tryCacheGetById
https://bugs.webkit.org/show_bug.cgi?id=132314
Reviewed by Oliver Hunt and Filip Pizlo.
This is neutral across all benchmarks we track, although it looks like a wee 0.5% progression on sunspider.
- jit/Repatch.cpp:
(JSC::tryCacheGetByID): If we fail to cache on self, we just repatch to call tryBuildGetByIDList next time.
- 7:29 PM Changeset in webkit [167921] by
-
- 3 edits2 adds in trunk
SelectorCodeGenerator::generateElementIsNthChild() leaks the parent register :nth-child() is non-filtering
https://bugs.webkit.org/show_bug.cgi?id=132311
Reviewed by Andreas Kling.
Source/WebCore:
In two cases, the parent register was never returned to the register allocator:
-Non filtering selectors (the early return).
-Non marking selectors (at the moment: only querySelector API).
Unfortunately, generateElementIsNthChild() makes function call, which forces us to do this manual allocation
of the parentElement register. Long term, I want RegisterAllocator and FunctionCall to have a special type for that.
Test: fast/selectors/several-nth-child.html
- cssjit/SelectorCompiler.cpp:
(WebCore::SelectorCompiler::SelectorCodeGenerator::generateElementIsNthChild):
LayoutTests:
- fast/selectors/several-nth-child-expected.txt: Added.
- fast/selectors/several-nth-child.html: Added.
- 6:58 PM Changeset in webkit [167920] by
-
- 3 edits4 adds in trunk
CSS JIT: backtracking with current / parent element for child
https://bugs.webkit.org/show_bug.cgi?id=132057
Patch by Yusuke Suzuki <Yusuke Suzuki> on 2014-04-28
Reviewed by Benjamin Poulain.
Calculate appropriate backtracking start height from the closest
descendant. And at first, we use it for a simple optimization.
- When backtracking start height equals to current height, we
can simply jump to a descendant element check phase.
- When backtracking start height equals to current height + 1, we
can simply jump to a descendant element traversing phase.
We can apply this optimization to fragments with adjacent combinators.
But, in the meantime, we start to implement it for a fragment with
child combinator.
Source/WebCore:
- cssjit/SelectorCompiler.cpp:
(WebCore::SelectorCompiler::SelectorFragment::SelectorFragment):
(WebCore::SelectorCompiler::TagNamePattern::TagNamePattern):
(WebCore::SelectorCompiler::solveDescendantBacktrackingActionForChild):
(WebCore::SelectorCompiler::solveBacktrackingAction):
(WebCore::SelectorCompiler::equalTagNames):
(WebCore::SelectorCompiler::equalTagNamePatterns):
(WebCore::SelectorCompiler::computeBacktrackingStartHeightFromDescendant):
(WebCore::SelectorCompiler::computeBacktrackingHeightFromDescendant):
(WebCore::SelectorCompiler::requiresAdjacentTail):
(WebCore::SelectorCompiler::requiresDescendantTail):
(WebCore::SelectorCompiler::SelectorCodeGenerator::computeBacktrackingInformation):
(WebCore::SelectorCompiler::SelectorCodeGenerator::generateSelectorChecker):
(WebCore::SelectorCompiler::SelectorCodeGenerator::generateWalkToParentNode):
(WebCore::SelectorCompiler::SelectorCodeGenerator::generateWalkToParentElement):
(WebCore::SelectorCompiler::SelectorCodeGenerator::generateParentElementTreeWalker):
(WebCore::SelectorCompiler::SelectorCodeGenerator::generateAncestorTreeWalker):
(WebCore::SelectorCompiler::SelectorCodeGenerator::generateDirectAdjacentTreeWalker):
(WebCore::SelectorCompiler::SelectorCodeGenerator::generateIndirectAdjacentTreeWalker):
(WebCore::SelectorCompiler::SelectorCodeGenerator::generateElementMatching):
LayoutTests:
- fast/selectors/backtracking-child-combinator-with-tail-expected.txt: Added.
- fast/selectors/backtracking-child-combinator-with-tail.html: Added.
- fast/selectors/backtracking-child-combinator-without-tail-expected.txt: Added.
- fast/selectors/backtracking-child-combinator-without-tail.html: Added.
- 6:54 PM Changeset in webkit [167919] by
-
- 5 edits in trunk/Source
Versioning.
- 6:52 PM Changeset in webkit [167918] by
-
- 1 copy in tags/Safari-538.32
New Tag.
- 6:36 PM Changeset in webkit [167917] by
-
- 2 edits in trunk/Source/JavaScriptCore
REGRESSION(r153142) ASSERT from CodeBlock::dumpBytecode dumping String Switch Jump Tables
https://bugs.webkit.org/show_bug.cgi?id=132315
Reviewed by Mark Hahnenberg.
Used the StringImpl version of utf8() instead of creating a String first.
- bytecode/CodeBlock.cpp:
(JSC::CodeBlock::dumpBytecode):
- 6:30 PM Changeset in webkit [167916] by
-
- 11 edits in trunk/Source
[iOS][WK2] Restore the scroll position and scale from the HistoryItem (mostly)
https://bugs.webkit.org/show_bug.cgi?id=132307
Source/WebCore:
Patch by Benjamin Poulain <bpoulain@apple.com> on 2014-04-28
Reviewed by Simon Fraser.
- WebCore.exp.in:
Source/WebKit2:
<rdar://problem/16031704>
Patch by Benjamin Poulain <bpoulain@apple.com> on 2014-04-28
Reviewed by Simon Fraser.
Restore the scroll position and scale on back-forward by restoring the values from HistoryItem.
This covers the common cases, see <rdar://problem/16031704>.
- UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _didCommitLayerTree:WebKit::]):
(-[WKWebView _didCommitLoadForMainFrame]): Deleted.
- UIProcess/API/Cocoa/WKWebViewInternal.h:
- UIProcess/ios/WKContentView.mm:
(-[WKContentView _didCommitLoadForMainFrame]):
Get rid of the old code restoring the scroll position. This is now done by the WebProcess.
- UIProcess/mac/RemoteLayerTreeDrawingAreaProxy.mm:
(WebKit::RemoteLayerTreeDrawingAreaProxy::commitLayerTree):
Update the scrolling tree after invoking didCommitLayerTree() on the WebPageProxy.
WKContentView and WKWebView states are updated in response to didCommitLayerTree(), the content size
and scale in particular are set there.
After that, the content size/scale is in the same state as the layerTree update, and we can scroll
the content as needed.
- WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
- WebProcess/WebCoreSupport/ios/WebFrameLoaderClientIOS.mm:
(WebKit::WebFrameLoaderClient::restoreViewState):
iOS WebKit1 uses a custom restoreViewState(), do the same for WebKit2 since the generic code is #ifdefed out.
This is also necessary for the main frame's state update.
- WebProcess/WebPage/WebPage.h:
- WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::restorePageState):
Restore the scale and position of the main frame. For now, we assume m_userHasChangedPageScaleFactor is set otherwise
the initial-scale would override the restored value. We will need to fix that, the value should probably be saved with
the history.
- 6:13 PM Changeset in webkit [167915] by
-
- 4 edits in trunk/LayoutTests
Unreviewed EFL gardening.
Skip editing/pasteboard which EFL port does not support.
- platform/efl-wk1/TestExpectations:
- platform/efl-wk2/TestExpectations:
- platform/efl/TestExpectations:
- 5:46 PM Changeset in webkit [167914] by
-
- 3 edits in trunk/Source/WebKit2
[iOS][WK2] iOS fix after r167867
https://bugs.webkit.org/show_bug.cgi?id=132302
Patch by Benjamin Poulain <bpoulain@apple.com> on 2014-04-28
Reviewed by Simon Fraser.
iOS Touch Points uses "phase" for describing the touch points state for historical reasons.
This patch adds state() as a synonym for phase, and revert r167893.
- Shared/WebEvent.h:
(WebKit::WebPlatformTouchPoint::state):
- UIProcess/WebPageProxy.cpp:
(WebKit::areAllTouchPointsReleased):
- 5:31 PM Changeset in webkit [167913] by
-
- 6 edits in trunk/Tools
Crash in platform/mac/accessibility/table-visible-rows.html
https://bugs.webkit.org/show_bug.cgi?id=132146
Reviewed by Filip Pizlo.
Instead of creating a local JSValueRef array on the stack and passing that to
JSObjectMakeArray(), changed to create an empty JSArray and then populate the
values using the JSObjectSetPropertyAtIndex() API.
- DumpRenderTree/AccessibilityUIElement.cpp:
(elementsForRangeCallback):
(convertElementsToObjectArray):
- DumpRenderTree/mac/TestRunnerMac.mm:
(originsArrayToJS):
- WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
(WTR::EventSendingController::contextClick):
- WebKitTestRunner/InjectedBundle/TestRunner.cpp:
(WTR::stringArrayToJS):
- WebKitTestRunner/InjectedBundle/mac/AccessibilityUIElementMac.mm:
(WTR::convertElementsToObjectArray):
- 4:23 PM Changeset in webkit [167912] by
-
- 2 edits in trunk/Source/WebKit2
Fix a transcription error from r167901.
When addressing Darin's review feedback, I moved the iteration of dataArray from
QuickLookDocumentData::append() to WebQuickLookHandleClient::didReceiveDataArray(),
but I forgot to update the context argument to CFArrayApplyFunction() accordingly.
- WebProcess/WebCoreSupport/ios/WebQuickLookHandleClient.cpp:
(WebKit::WebQuickLookHandleClient::didReceiveDataArray):
- 4:19 PM Changeset in webkit [167911] by
-
- 2 edits in trunk/Source/WebCore
Scrollbars do not update properly when topContentInset changes dynamically
https://bugs.webkit.org/show_bug.cgi?id=132309
-and corresponding-
<rdar://problem/16642232>
Reviewed by Tim Horton.
It is not sufficient to do a layout and call updateScrollbars(). We must also call
RenderLayerCompositor::frameViewDidChangeSize() in order to properly adjust the
size and position of all of the scrolling-related layers.
- page/FrameView.cpp:
(WebCore::FrameView::topContentInsetDidChange):
- 3:35 PM Changeset in webkit [167910] by
-
- 4 edits in trunk/Source/WebCore
[New Multicolumn] event.offsetX/offsetY don't work correctly
https://bugs.webkit.org/show_bug.cgi?id=132284
Reviewed by Simon Fraser.
- rendering/RenderBoxModelObject.cpp:
(WebCore::RenderBoxModelObject::mapAbsoluteToLocalPoint):
Change the code here to only trigger for CSS Regions and not for
multi-column. Add a FIXME pointing out that the code is broken even for
CSS regions.
- rendering/RenderMultiColumnFlowThread.cpp:
(WebCore::RenderMultiColumnFlowThread::mapAbsoluteToLocalPoint):
Override mapAbsoluteToLocalPoint in order to guess a region and attempt
to translate the point from that region to the flow thread local coordinate
space.
(WebCore::RenderMultiColumnFlowThread::physicalTranslationFromRegionToFlow):
- rendering/RenderMultiColumnFlowThread.h:
New helper function that just wraps calling the region's logical translation
function and converting to/from physical coordinates.
- 3:00 PM Changeset in webkit [167909] by
-
- 2 edits in trunk/Source/WebKit2
[WK2 iOS] Need to show title in the action sheet for images not inside an anchor element.
https://bugs.webkit.org/show_bug.cgi?id=132304
<rdar://problem/16535227>
Reviewed by Benjamin Poulain.
If the image in not inside an anchor tag, we must use the title attribute as
title for the action sheet.
- UIProcess/ios/WKActionSheetAssistant.mm:
(-[WKActionSheetAssistant _createSheetWithElementActions:showLinkTitle:]):
- 2:54 PM Changeset in webkit [167908] by
-
- 2 edits in trunk/Source/WebCore
Unreviewed, rolling out r167857.
https://bugs.webkit.org/show_bug.cgi?id=132305
the change was rolled out, roll out the new expectations as
well (Requested by thorton on #webkit).
Reverted changeset:
"Unreviewed. Updating one bindings test baseline after
r167855."
http://trac.webkit.org/changeset/167857
- 2:17 PM Changeset in webkit [167907] by
-
- 2 edits in trunk/Source/WebCore
[Mac] Use the animated version of setHighlighted on NSButtonCell where available
https://bugs.webkit.org/show_bug.cgi?id=132295
<rdar://problem/16747240>
Reviewed by Beth Dakin.
Like setState, there is an animated version of setHighlighted available.
- platform/mac/ThemeMac.mm:
(WebCore::updateStates): Use private version of _setHighlighted where possible.
- 2:14 PM Changeset in webkit [167906] by
-
- 4 edits in trunk
Source/JavaScriptCore: The LLInt is awesome and it should get more of the action.
Rubber stamped by Geoffrey Garen.
5% speed-up on JSBench and no meaningful regressions. Should be a PLT/DYE speed-up also.
- runtime/Options.h:
Tools: Make sure that the coverage of our tests doesn't change much if we change the thresholds now and in the future, but making
"no-cjit" runs force the old-school threshold.
Rubber stamped by Mark Hahnenberg.
- Scripts/run-jsc-stress-tests:
- 2:07 PM Changeset in webkit [167905] by
-
- 7 edits in trunk/Source/WebKit2
WebKit2 View Gestures (Swipe): Discard snapshots made with a different view size/pixel density
https://bugs.webkit.org/show_bug.cgi?id=132115
Reviewed by Simon Fraser.
We shouldn't try to use snapshots if they no longer match the destination view size.
To do this, we need an accurate understanding of the topContentInset, both for the Web view,
and for any custom views, because we don't want to include the inset area in the snapshot,
and we don't want to reject snapshots that differ from the current view only because they have an inset or not.
- UIProcess/API/Cocoa/WKViewPrivate.h:
- UIProcess/API/mac/WKView.mm:
(-[WKView _takeViewSnapshot]):
(-[WKView _setCustomSwipeViewsTopContentInset:]):
Add and plumb through a topContentInset property to apply to the custom swipe views.
For simplicitly's sake in the multi-view case, it is applied in window coordinates,
so clients will have to do conversions if necessary.
Don't include the top content inset in the snapshot that is taken.
- UIProcess/mac/ViewGestureController.h:
(WebKit::ViewGestureController::setCustomSwipeViewsTopContentInset):
- UIProcess/mac/ViewGestureControllerMac.mm:
(WebKit::ViewGestureController::ViewGestureController):
(WebKit::ViewGestureController::windowRelativeBoundsForCustomSwipeViews):
Apply the aforementioned custom swipe view top content inset.
(WebKit::ViewGestureController::determineSnapshotLayerParent):
(WebKit::ViewGestureController::determineLayerAdjacentToSnapshotForParent):
Factor these out of beginSwipeGesture.
(WebKit::ViewGestureController::retrieveSnapshotForItem):
Don't return a snapshot if the current deviceScaleFactor doesn't match that
at which the snapshot was taken, or if the size of the snapshot doesn't match
the unobscured part of the snapshot layer (the layer's size sans top content inset).
(WebKit::layerGeometryFlippedToRoot): Added.
Determine whether geometry flipping occurs between a layer and the root of the tree it is in.
(WebKit::ViewGestureController::applyDebuggingPropertiesToSwipeViews):
Factor out debugging layer properties, and add some borders and backgrounds to our layers.
I switched to invert instead of blur because blur makes the borders hard to use.
(WebKit::ViewGestureController::beginSwipeGesture):
Add an additional layer, the swipeLayer, which is the parent of the swipeSnapshotLayer.
We use this layer to compensate for geometry flipping, since we don't necessarily
know the structure of the layer tree ahead of time when inserting snapshots into the
client's layer tree, in the custom swipe view case.
The swipe layer is always the full size of the view, including topContentInset, and
has a white background in order to fill in that space. The snapshot layer is the size
of the view *excluding* the topContentInset, which should match the size of the snapshot image
itself (and we reject the image and just paint white if this is not the case).
(WebKit::ViewGestureController::handleSwipeGesture):
(WebKit::ViewGestureController::removeSwipeSnapshot):
Move and remove the new swipeLayer.
- UIProcess/mac/ViewSnapshotStore.h:
- UIProcess/mac/ViewSnapshotStore.mm:
(WebKit::ViewSnapshotStore::recordSnapshot):
Store the deviceScaleFactor at the time the snapshot was taken.
Fix a bug where the count of snapshots with live images was too high
because we were failing to decrement it when replacing a snapshot of
an existing item with a fresh one.
- 1:55 PM Changeset in webkit [167904] by
-
- 35 edits in trunk/Source/WebCore
Unreviewed, rolling out r167855.
https://bugs.webkit.org/show_bug.cgi?id=132301
Broke the windows build (Requested by bfulgham on #webkit).
Reverted changeset:
"ScriptExecutionContext::Task should work well with C++11
lambdas"
https://bugs.webkit.org/show_bug.cgi?id=129795
http://trac.webkit.org/changeset/167855
- 1:50 PM Changeset in webkit [167903] by
-
- 2 edits in trunk/Source/WebKit2
WK2 iOS: crash when handling a synthetic keyboard event.
https://bugs.webkit.org/show_bug.cgi?id=132296
<rdar://problem/16469726>
Reviewed by Benjamin Poulain.
Adding a null check, since synthetic keyboard events do not
have a corresponding platform event. We already do this on OS X.
- WebProcess/WebPage/ios/WebPageIOS.mm:
(WebKit::WebPage::handleEditingKeyboardEvent):
- 1:39 PM Changeset in webkit [167902] by
-
- 2 edits in trunk/Tools
[GTK] install-dependencies should also install ruby-json and ruby-highline
https://bugs.webkit.org/show_bug.cgi?id=132292
Patch by Carlos Alberto Lopez Perez <clopez@igalia.com> on 2014-04-28
Reviewed by Philippe Normand.
- gtk/install-dependencies: Install ruby json and highline packages.
- 1:02 PM Changeset in webkit [167901] by
-
- 15 edits5 copies1 add in trunk/Source
[iOS] Implement WebQuickLookHandleClient for WebKit2
https://bugs.webkit.org/show_bug.cgi?id=132157
Reviewed by Darin Adler.
Source/WebCore:
- WebCore.exp.in: Exported QuickLookHandle::previewUTI().
- platform/network/ios/QuickLook.h:
- platform/network/ios/QuickLook.mm:
(WebCore::QuickLookHandle::previewFileName): Changed to return a WTF::String.
(WebCore::QuickLookHandle::previewUTI): Added.
Source/WebKit2:
- Shared/API/Cocoa/_WKNSFileManagerExtras.h: Added.
- Shared/API/Cocoa/_WKNSFileManagerExtras.mm: Added.
(+[NSFileManager _web_createTemporaryFileForQuickLook:]): Called WebCore::createTemporaryFileForQuickLook().
- Shared/ios/QuickLookDocumentData.cpp: Added.
(WebKit::QuickLookDocumentData::QuickLookDocumentData):
(WebKit::QuickLookDocumentData::append): Appened each CFDataRef to m_data and added their sizes to m_size.
(WebKit::QuickLookDocumentData::decodedData): Returned the combined CFDataRef created by decode().
Named this method 'decodedData' to indicate that it should only be called on objects populated by decode().
(WebKit::QuickLookDocumentData::clear): Cleared m_data and reset m_size to 0.
(WebKit::QuickLookDocumentData::encode): Implemented a custom encoding function to avoid having to combine
m_data into contiguous memory at encoding time. Encoded m_size then sequentially encoded each member of m_data.
(WebKit::QuickLookDocumentData::decode): Decoded the byte stream into a single CFDataRef. Eagerly copied the
decoded data rather than using a DataReference since we know this will outlive the decoder.
- Shared/ios/QuickLookDocumentData.h: Added.
- UIProcess/API/APILoaderClient.h:
(API::LoaderClient::didStartLoadForQuickLookDocumentInMainFrame):
(API::LoaderClient::didFinishLoadForQuickLookDocumentInMainFrame):
- UIProcess/API/Cocoa/WKNavigationDelegatePrivate.h:
- UIProcess/Cocoa/NavigationState.h:
- UIProcess/Cocoa/NavigationState.mm:
(WebKit::NavigationState::setNavigationDelegate):
(WebKit::NavigationState::LoaderClient::didStartLoadForQuickLookDocumentInMainFrame):
(WebKit::NavigationState::LoaderClient::didFinishLoadForQuickLookDocumentInMainFrame):
- UIProcess/WebPageProxy.h:
- UIProcess/WebPageProxy.messages.in:
- UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::didStartLoadForQuickLookDocumentInMainFrame):
(WebKit::WebPageProxy::didFinishLoadForQuickLookDocumentInMainFrame):
- WebKit2.xcodeproj/project.pbxproj:
- WebProcess/WebCoreSupport/WebFrameLoaderClient.h:
- WebProcess/WebCoreSupport/ios/WebFrameLoaderClientIOS.mm:
(WebKit::WebFrameLoaderClient::didCreateQuickLookHandle):
- WebProcess/WebCoreSupport/ios/WebQuickLookHandleClient.cpp: Added.
(WebKit::WebQuickLookHandleClient::WebQuickLookHandleClient): Sent DidStartLoadForQuickLookDocumentInMainFrame
to WebPageProxy.
(WebKit::WebQuickLookHandleClient::didReceiveDataArray): Appended dataArray to m_data.
(WebKit::WebQuickLookHandleClient::didFinishLoading): Sent DidFinishLoadForQuickLookDocumentInMainFrame to
WebPageProxy and then cleared m_data.
(WebKit::WebQuickLookHandleClient::didFail): Cleared m_data.
- WebProcess/WebCoreSupport/ios/WebQuickLookHandleClient.h: Added.
(WebKit::WebQuickLookHandleClient::create):
- 1:01 PM Changeset in webkit [167900] by
-
- 2 edits1 add in trunk/LayoutTests
Layout Test fast/events/shadow-event-path[-2].html is failing
https://bugs.webkit.org/show_bug.cgi?id=132238
Add the file that was supposed to be committed in r167840.
- TestExpectations:
- fast/events/resources/shadow-event-path-shared.js: Added.
- 12:57 PM Changeset in webkit [167899] by
-
- 2 edits in trunk/Source/WebCore
REGRESSION: Intermittent crash in SpeechSynthesis::didFinishSpeaking
https://bugs.webkit.org/show_bug.cgi?id=111613
Reviewed by Mark Lam.
I think it's possible that didFinishSpeaking ends up calling directly back into start speaking, and the utterance reference
we were holding can get cleared, so protecting this should avoid a few asserts.
- platform/mock/PlatformSpeechSynthesizerMock.cpp:
(WebCore::PlatformSpeechSynthesizerMock::speakingFinished):
- 12:56 PM Changeset in webkit [167898] by
-
- 2 edits in trunk/LayoutTests
REGRESSION (r167845): ASSERT(!m_renderView.needsLayout()) in svg/custom/bug79798.html
https://bugs.webkit.org/show_bug.cgi?id=132297
- platform/wk2/TestExpectations:
Skip the test.
- 12:01 PM Changeset in webkit [167897] by
-
- 22 edits in trunk/Source
GC should be able to remove things from the DFG worklist and cancel on-going compilations if it knows that the compilation would already be invalidated
https://bugs.webkit.org/show_bug.cgi?id=132166
Reviewed by Oliver Hunt and Mark Hahnenberg.
The GC can aid type inference by removing structures that are dead and jettisoning
code that relies on those structures. This can dramatically accelerate type inference
for some tricky programs.
Unfortunately, we previously pinned any structures that enqueued compilations depended
on. This means that if you're on a machine that only runs a single compilation thread
and where compilations are relatively slow, you have a high chance of large numbers of
structures being pinned during any GC since the compilation queue is likely to be full
of random stuff.
This comprehensively fixes this issue by allowing the GC to remove compilation plans
if the things they depend on are dead, and to even cancel safepointed compilations.
- bytecode/CodeBlock.cpp:
(JSC::CodeBlock::shouldImmediatelyAssumeLivenessDuringScan):
(JSC::CodeBlock::isKnownToBeLiveDuringGC):
(JSC::CodeBlock::finalizeUnconditionally):
- bytecode/CodeBlock.h:
(JSC::CodeBlock::shouldImmediatelyAssumeLivenessDuringScan): Deleted.
- dfg/DFGDesiredIdentifiers.cpp:
(JSC::DFG::DesiredIdentifiers::DesiredIdentifiers):
- dfg/DFGDesiredIdentifiers.h:
- dfg/DFGDesiredWatchpoints.h:
- dfg/DFGDesiredWeakReferences.cpp:
(JSC::DFG::DesiredWeakReferences::DesiredWeakReferences):
- dfg/DFGDesiredWeakReferences.h:
- dfg/DFGGraphSafepoint.cpp:
(JSC::DFG::GraphSafepoint::GraphSafepoint):
- dfg/DFGGraphSafepoint.h:
- dfg/DFGPlan.cpp:
(JSC::DFG::Plan::Plan):
(JSC::DFG::Plan::compileInThread):
(JSC::DFG::Plan::compileInThreadImpl):
(JSC::DFG::Plan::notifyCompiling):
(JSC::DFG::Plan::notifyCompiled):
(JSC::DFG::Plan::notifyReady):
(JSC::DFG::Plan::checkLivenessAndVisitChildren):
(JSC::DFG::Plan::isKnownToBeLiveDuringGC):
(JSC::DFG::Plan::cancel):
(JSC::DFG::Plan::visitChildren): Deleted.
- dfg/DFGPlan.h:
- dfg/DFGSafepoint.cpp:
(JSC::DFG::Safepoint::Result::~Result):
(JSC::DFG::Safepoint::Result::didGetCancelled):
(JSC::DFG::Safepoint::Safepoint):
(JSC::DFG::Safepoint::~Safepoint):
(JSC::DFG::Safepoint::checkLivenessAndVisitChildren):
(JSC::DFG::Safepoint::isKnownToBeLiveDuringGC):
(JSC::DFG::Safepoint::cancel):
(JSC::DFG::Safepoint::visitChildren): Deleted.
- dfg/DFGSafepoint.h:
(JSC::DFG::Safepoint::Result::Result):
- dfg/DFGWorklist.cpp:
(JSC::DFG::Worklist::compilationState):
(JSC::DFG::Worklist::waitUntilAllPlansForVMAreReady):
(JSC::DFG::Worklist::removeAllReadyPlansForVM):
(JSC::DFG::Worklist::completeAllReadyPlansForVM):
(JSC::DFG::Worklist::visitWeakReferences):
(JSC::DFG::Worklist::removeDeadPlans):
(JSC::DFG::Worklist::runThread):
(JSC::DFG::Worklist::visitChildren): Deleted.
- dfg/DFGWorklist.h:
- ftl/FTLCompile.cpp:
(JSC::FTL::compile):
- ftl/FTLCompile.h:
- heap/CodeBlockSet.cpp:
(JSC::CodeBlockSet::rememberCurrentlyExecutingCodeBlocks):
- heap/Heap.cpp:
(JSC::Heap::markRoots):
(JSC::Heap::visitCompilerWorklistWeakReferences):
(JSC::Heap::removeDeadCompilerWorklistEntries):
(JSC::Heap::visitWeakHandles):
(JSC::Heap::collect):
(JSC::Heap::visitCompilerWorklists): Deleted.
- heap/Heap.h:
- 11:45 AM Changeset in webkit [167896] by
-
- 5 edits in trunk/Source/WebKit2
Hide service controls for selection behind a runtime switch
<rdar://problem/16736616> and https://bugs.webkit.org/show_bug.cgi?id=132291
Reviewed by Tim Horton.
- Shared/WebPreferencesStore.h:
- WebProcess/WebCoreSupport/WebEditorClient.cpp:
(WebKit::WebEditorClient::selectionRectsDidChange): Do nothing if the UI is disabled.
- WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::WebPage):
(WebKit::WebPage::updatePreferences):
- WebProcess/WebPage/WebPage.h:
(WebKit::WebPage::serviceControlsEnabled):
- 11:05 AM Changeset in webkit [167895] by
-
- 7 edits2 deletes in trunk
Unreviewed, rolling out r167871.
https://bugs.webkit.org/show_bug.cgi?id=132290
broke a newmulticol test (spanner-nested-dynamic) (Requested
by thorton on #webkit).
Reverted changeset:
"Store the containing region map inside the flow thread"
https://bugs.webkit.org/show_bug.cgi?id=131647
http://trac.webkit.org/changeset/167871
- 10:57 AM Changeset in webkit [167894] by
-
- 9 edits in trunk/Source
Unreviewed, rolling out r167853.
https://bugs.webkit.org/show_bug.cgi?id=132288
caused crashes+timeouts+layout test failures described in the
bug (Requested by thorton on #webkit).
Reverted changeset:
"Coalesce responses on network process side"
https://bugs.webkit.org/show_bug.cgi?id=132229
http://trac.webkit.org/changeset/167853
- 10:38 AM Changeset in webkit [167893] by
-
- 2 edits in trunk/Source/WebKit2
Fix the iOS buld.
- UIProcess/WebPageProxy.cpp:
(WebKit::areAllTouchPointsReleased):
- 10:28 AM Changeset in webkit [167892] by
-
- 2 edits in trunk/Source/WebKit2
[iOS WK2] Make subview rebuilding more incremental
https://bugs.webkit.org/show_bug.cgi?id=132162
<rdar://problem/16718916>
Reviewed by Sam Weinig.
Previously we did bulk replacement of subviews, but that triggers
UIKit to send spurious will/DidMoveToWindow messages, which can
interfere with UIScrollView scrolling.
Make _web_setSubviews more incremental, only removing
deleted subviews, adding new ones, and moving existing ones
into the correct order.
- Shared/mac/RemoteLayerTreePropertyApplier.mm:
(-[UIView _web_setSubviews:]):
- 10:28 AM Changeset in webkit [167891] by
-
- 4 edits in trunk/Source
[iOS WK2] flickery scrolling with overflow-scrolling:touch
https://bugs.webkit.org/show_bug.cgi?id=132150
Source/WebCore:
Reviewed by Tim Horton.
Fix typo in a comment.
- page/scrolling/AsyncScrollingCoordinator.cpp:
(WebCore::AsyncScrollingCoordinator::scheduleUpdateScrollPositionAfterAsyncScroll):
Source/WebKit2:
<rdar://problem/16683006>
Reviewed by Tim Horton.
Maintain a flag that tells us if the user is interacting
with the UIScrollView, and while set, don't clobber the contentOffset
via scrolling tree updates.
- UIProcess/Scrolling/ios/ScrollingTreeOverflowScrollingNodeIOS.mm:
(-[WKOverflowScrollViewDelegate initWithScrollingTreeNode:WebKit::]):
(-[WKOverflowScrollViewDelegate scrollViewWillBeginDragging:]):
(-[WKOverflowScrollViewDelegate scrollViewDidEndDragging:willDecelerate:]):
(-[WKOverflowScrollViewDelegate scrollViewDidEndDecelerating:]): We need this one,
not scrollViewDidEndScrollingAnimation:, to detect the end of decleration.
(WebKit::ScrollingTreeOverflowScrollingNodeIOS::updateAfterChildren):
(-[WKOverflowScrollViewDelegate scrollViewDidEndScrollingAnimation:]): Deleted.
- 10:27 AM Changeset in webkit [167890] by
-
- 2 edits in trunk/Source/WebKit2
[X11] Add missing case for mime type application/x-webkit-test-netscape
https://bugs.webkit.org/show_bug.cgi?id=132278
Patch by Krzysztof Wolanski <k.wolanski@samsung.com> on 2014-04-28
Reviewed by Darin Adler.
Case for x-webkit-test-netscape has been omitted along with r167759.
- WebProcess/Plugins/Netscape/NetscapePlugin.cpp:
(WebKit::NetscapePlugin::initialize):
- 10:26 AM Changeset in webkit [167889] by
-
- 7 edits4 adds in trunk
Deleting properties poisons objects
https://bugs.webkit.org/show_bug.cgi?id=131551
Reviewed by Oliver Hunt.
Source/JavaScriptCore:
This is ~3% progression on Dromaeo with a ~6% progression on the jslib portion of Dromaeo in particular.
- runtime/JSPropertyNameIterator.cpp:
(JSC::JSPropertyNameIterator::create):
- runtime/PropertyMapHashTable.h:
(JSC::PropertyTable::hasDeletedOffset):
(JSC::PropertyTable::hadDeletedOffset): If we ever had deleted properties we can no longer cache offsets when
iterating properties because we're required to iterate properties in insertion order.
- runtime/Structure.cpp:
(JSC::Structure::Structure):
(JSC::Structure::materializePropertyMap): We now re-use deleted properties when materializing the property map.
(JSC::Structure::removePropertyTransition): We allow up to 5 deletes for a particular path through the tree of
Structure transitions. After that, we convert to an uncacheable dictionary like we used to. We don't cache
delete transitions, but we allow transitioning from them.
(JSC::Structure::changePrototypeTransition):
(JSC::Structure::despecifyFunctionTransition):
(JSC::Structure::attributeChangeTransition):
(JSC::Structure::toDictionaryTransition):
(JSC::Structure::preventExtensionsTransition):
(JSC::Structure::addPropertyWithoutTransition):
(JSC::Structure::removePropertyWithoutTransition):
(JSC::Structure::pin): Now does only what it says it does--marks the property table as pinned.
(JSC::Structure::pinAndPreventTransitions): More descriptive version of what the old pin() was doing.
- runtime/Structure.h:
- runtime/StructureInlines.h:
(JSC::Structure::setEnumerationCache):
(JSC::Structure::hadDeletedOffsets):
(JSC::Structure::propertyTable):
(JSC::Structure::checkOffsetConsistency): Rearranged variables to be more sensible.
- tests/stress/for-in-after-delete.js: Added.
(foo):
LayoutTests:
New JS regress test. We're ~3.5x faster on this microbenchmark now.
- js/regress/delete-a-few-properties-then-get-by-id-expected.txt: Added.
- js/regress/delete-a-few-properties-then-get-by-id.html: Added.
- js/regress/script-tests/delete-a-few-properties-then-get-by-id.js: Added.
(MyObject):
(foo):
- 10:11 AM WebKitGTK/2.4.x edited by
- (diff)
- 10:09 AM WebKitGTK/2.4.x edited by
- (diff)
- 9:24 AM WebKitGTK/KeepingTheTreeGreen edited by
- (diff)
- 8:59 AM Changeset in webkit [167888] by
-
- 2 edits in trunk/LayoutTests
Unreviewed GTK gardening.
- platform/gtk/TestExpectations: Flagged some tests.
- 8:18 AM Changeset in webkit [167887] by
-
- 2 edits in trunk/Source/WebCore
[GTK] Builtin cursors do not properly handle transparency
https://bugs.webkit.org/show_bug.cgi?id=131866
Reviewed by Gustavo Noronha Silva.
Tested by ManualTests/cursor.html.
- platform/gtk/CursorGtk.cpp:
(WebCore::createNamedCursor): Instead of interpreting the source bitmap as an A1 image, use
it as a 1-bit black and white image. We do this by:
- Painting the result to a full color image with transparency instead of an alpha-only surface.
- Masking a white background using the cursor alpha surface.
- Painting the black parts of the cursor by painting the source surface, where black pixels will be interpreted as full opaque pixels.
- 8:06 AM Changeset in webkit [167886] by
-
- 2 edits in trunk/LayoutTests
Unreviewed GTK gardening.
- platform/gtk/TestExpectations: Flagged some tests.
- 7:27 AM Changeset in webkit [167885] by
-
- 3 edits in trunk/Source/WebKit2
[WK2][X11] NetscapePluginModule::scanPlugin() should write UTF-8 strings to stdout
https://bugs.webkit.org/show_bug.cgi?id=132245
Reviewed by Carlos Garcia Campos.
NetscapePluginModule::scanPlugin(), in the helper writeLine function, takes each character
of the given string and writes it out as a UChar, doing two fputc calls for each byte of the
16-bit type.
This fails badly with characters with integer value less than 256 as the most significant byte
of the UChar is written out as a null character. This effectively chops the output that's gathered
in the UIProcess and is parsed in PluginProcessProxy::scanPlugin().
To avoid all this, the UTF-8 encoding of the string is written out in the PluginProcess, and
String::fromUTF8() is called in the UIProcess to properly decode the received string.
- Shared/Plugins/Netscape/x11/NetscapePluginModuleX11.cpp:
(WebKit::writeCharacter):
(WebKit::writeLine):
(WebKit::writeByte): Deleted.
- UIProcess/Plugins/unix/PluginProcessProxyUnix.cpp:
(WebKit::PluginProcessProxy::scanPlugin):
- 7:24 AM Changeset in webkit [167884] by
-
- 2 edits in trunk/Source/WebKit2
[GTK][WK2] Missing return statement in webkit_plugin_get_description()
https://bugs.webkit.org/show_bug.cgi?id=132263
Reviewed by Carlos Garcia Campos.
- UIProcess/API/gtk/WebKitPlugin.cpp:
(webkit_plugin_get_description): Actually return the data of the cached
plugin description CString.
- 4:10 AM Changeset in webkit [167883] by
-
- 2 edits in trunk/Source/WebKit2
[GTK] Crash in debug build with removing windowed plugin child widgets from the view
https://bugs.webkit.org/show_bug.cgi?id=132252
Reviewed by Philippe Normand.
It crashes due to an assert in HashTable that checks the iterators
validity. The problem is that we are iterating the children map
and the callback called on every iteration might modify the map,
making the iterators invalid. This happens when the WebView is
destroyed, GtkContainer calls gtk_container_foreach() with
gtk_widget_destroy as callback. When a widget inside a container
is destroyed, it's removed from the container, and in our case,
the child widget is removed from the map. This fixes several
crashes when running layout tests in debug bot.
- UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseContainerForall): Use copyKeysToVector() instead
of using a range iterator for the map keys and check in every
iteration that the child widget from the keys vector is still
present in the map before calling the callback.
- 3:57 AM WebKitGTK/2.4.x edited by
- (diff)
- 3:50 AM Changeset in webkit [167882] by
-
- 2 edits in trunk/LayoutTests
Unreviewed GTK gardening. Adding failure expectations for 6
failing and flaky tests. Removed 1.
- platform/gtk/TestExpectations:
- 2:35 AM Changeset in webkit [167881] by
-
- 2 edits in trunk/LayoutTests
[GTK] Review media bugs flags
https://bugs.webkit.org/show_bug.cgi?id=132246
Reviewed by Philippe Normand.
- platform/gtk/TestExpectations: Review flags of the GTK media
bugs.
- 1:29 AM Changeset in webkit [167880] by
-
- 2 edits in trunk/Source/WebCore
Unreviewed. Update GObject DOM bindings symbols file.
- bindings/gobject/webkitdom.symbols: Add missing symbols.
- 1:27 AM Changeset in webkit [167879] by
-
- 6 edits in trunk/Source/WebCore
OrderIterator refactoring to avoid extra loops
https://bugs.webkit.org/show_bug.cgi?id=119061
Reviewed by Darin Adler.
This patch removes order values Vector and use a Vector of pairs instead. The pairs are formed by a child
(RenderBox) and the index of this child. In addition, OrderIterator code is simplified.
It provides a helper class OrderIteratorPopulator, used for manipulating the Vector directly. Which allows to
consolidate the code into a single implementation across flexbox and grid. OrderIteratorPopulator part is based
on a patch from Blink r153971 by <jchaffraix@chromium.org>.
Current implementation is O(number of children * number of order values). Now it will just do a sort operation
and then a regular loop. So if you have different order values in a flexbox or grid the performance will
improve.
Comparing results of perf-tests:
- Layout/auto-grid-lots-of-data: ~0.5% worse.
- Layout/fixed-grid-lots-of-data: ~0.5% worse.
- Layout/fixed-grid-lots-of-data (setting 100 different order values): ~50% better.
- Layout/flexbox-lots-of-data: ~5% better.
No new tests, already covered by current tests.
- rendering/OrderIterator.cpp:
(WebCore::OrderIterator::currentChild): Return current child according to m_childrenIndex.
(WebCore::OrderIterator::first): Initialize m_childrenIndex and return current child.
(WebCore::OrderIterator::next): Increase m_childrenIndex and return current child.
(WebCore::compareByOrderValueAndIndex): Sorts the Vector by order value and index.
(WebCore::OrderIteratorPopulator::~OrderIteratorPopulator): Calls compareByOrderValueAndIndex() if there is any
child with non default order value.
(WebCore::OrderIteratorPopulator::collectChild): Adds the child and index to the Vector. Update
m_allChildrenHaveDefaultOrderValue accordingly.
(WebCore::OrderIterator::OrderIterator): Deleted.
(WebCore::OrderIterator::setOrderValues): Deleted.
(WebCore::OrderIterator::reset): Deleted.
- rendering/OrderIterator.h:
(WebCore::OrderIteratorPopulator::OrderIteratorPopulator): New helper class to manipulate the Vector.
(WebCore::OrderIterator::currentChild): Deleted.
- rendering/RenderFlexibleBox.cpp:
(WebCore::RenderFlexibleBox::RenderFlexibleBox): Remove OrderIterator intialization.
(WebCore::RenderFlexibleBox::layoutBlock): Remove unneeded code related to old OrderValues vector.
(WebCore::RenderFlexibleBox::prepareOrderIteratorAndMargins): Populate OrderIterator using collectChild().
(WebCore::RenderFlexibleBox::computeMainAxisPreferredSizes): Deleted.
- rendering/RenderFlexibleBox.h: Rename computeMainAxisPreferredSizes() to prepareOrderIteratorAndMargins().
- rendering/RenderGrid.cpp:
(WebCore::RenderGrid::RenderGrid): Remove OrderIterator initialization.
(WebCore::RenderGrid::populateExplicitGridAndOrderIterator): Populate OrderIterator using collectChild().
- 1:16 AM Changeset in webkit [167878] by
-
- 2 edits in trunk/Source/WebCore
std::bitset<>::test() does unnecessary bounds checks on CSSPropertyID bitsets
https://bugs.webkit.org/show_bug.cgi?id=131685
Patch by Zan Dobersek <zdobersek@igalia.com> on 2014-04-28
Reviewed by Darin Adler.
Use std::bitset<>::operator[]() instead of std::bitset<>::test() to avoid
bounds checks which are not necessary as long as a CSSPropertyID value is used.
- css/CSSParser.cpp:
(WebCore::filterProperties):
- css/StyleProperties.cpp:
(WebCore::StyleProperties::asText):
- css/StyleResolver.cpp:
(WebCore::StyleResolver::CascadedProperties::hasProperty):
(WebCore::StyleResolver::CascadedProperties::set):
- 1:12 AM Changeset in webkit [167877] by
-
- 4 edits in trunk/Source/WebCore
std::bitset<>::test() does unnecessary bounds checks on CSSPropertyID bitsets
https://bugs.webkit.org/show_bug.cgi?id=131685
Reviewed by Darin Adler.
Use std::bitset<>::operator[]() instead of std::bitset<>::test() to avoid
bounds checks which are not necessary as long as a CSSPropertyID value is used.
- css/CSSParser.cpp:
(WebCore::filterProperties):
- css/StyleProperties.cpp:
(WebCore::StyleProperties::asText):
- css/StyleResolver.cpp:
(WebCore::StyleResolver::CascadedProperties::hasProperty):
(WebCore::StyleResolver::CascadedProperties::set):
- 1:06 AM Changeset in webkit [167876] by
-
- 4 edits in trunk/Source/WebCore
[GTK] TextTrack kind and mode attributes are enums since r166180
https://bugs.webkit.org/show_bug.cgi?id=132228
Reviewed by Martin Robinson.
We don't support enum values yet in GObject DOM bindings, but they
are internally strings anyway, so we can keep the old
implementations using strings as custom functions until we
properly support enums.
- bindings/gobject/WebKitDOMCustom.cpp:
(webkit_dom_text_track_get_kind):
(webkit_dom_text_track_get_mode):
(webkit_dom_text_track_set_mode):
- bindings/gobject/WebKitDOMCustom.h:
- bindings/gobject/WebKitDOMCustom.symbols:
- 1:03 AM Changeset in webkit [167875] by
-
- 6 edits in trunk/Source/WebCore
[GTK] TextTrack::addCue can raise an exception since r163974
https://bugs.webkit.org/show_bug.cgi?id=132227
Reviewed by Martin Robinson.
webkit_dom_text_track_add_cue() now receives a GError paramater
which is an API break. Add
webkit_dom_text_track_add_cue_with_error and keep
webkit_dom_text_track_add_cue as deprecated to keep API
compatibility.
- bindings/gobject/WebKitDOMDeprecated.cpp:
(webkit_dom_text_track_add_cue):
- bindings/gobject/WebKitDOMDeprecated.h:
- bindings/gobject/WebKitDOMDeprecated.symbols:
- bindings/gobject/webkitdom.symbols:
- bindings/scripts/CodeGeneratorGObject.pm:
(GetEffectiveFunctionName):
- 12:57 AM Changeset in webkit [167874] by
-
- 9 edits in trunk/Source/WebCore
[GTK] TextTrackCue API changed in r163649
https://bugs.webkit.org/show_bug.cgi?id=132226
Reviewed by Martin Robinson.
TextTrackCue is now a base class and part of its API was moved to
the derived class VTTCue. Update the GObject DOM bindings to keep
backwards compatibility.
- PlatformGTK.cmake: Generate bindings for DataCue and VTTCue.
- bindings/gobject/WebKitDOMDeprecated.cpp:
(webkit_dom_text_track_cue_get_cue_as_html): Mark as deprecated in
favor of VTTCue API.
(webkit_dom_text_track_cue_get_vertical): Ditto.
(webkit_dom_text_track_cue_set_vertical): Ditto.
(webkit_dom_text_track_cue_get_snap_to_lines): Ditto.
(webkit_dom_text_track_cue_set_snap_to_lines): Ditto.
(webkit_dom_text_track_cue_get_line): Ditto.
(webkit_dom_text_track_cue_set_line): Ditto.
(webkit_dom_text_track_cue_get_position): Ditto.
(webkit_dom_text_track_cue_set_position): Ditto.
(webkit_dom_text_track_cue_get_size): Ditto.
(webkit_dom_text_track_cue_set_size): Ditto.
(webkit_dom_text_track_cue_get_align): Ditto.
(webkit_dom_text_track_cue_set_align): Ditto.
(webkit_dom_text_track_cue_get_text): Ditto.
(webkit_dom_text_track_cue_set_text): Ditto.
- bindings/gobject/WebKitDOMDeprecated.h:
- bindings/gobject/WebKitDOMDeprecated.symbols: Add new deprecated symbols.
- bindings/gobject/WebKitDOMPrivate.cpp:
(WebKit::wrap): Add generic wrap for TextTrackCue now that it's a
base class to generate DataCue or VTTCue objects.
- bindings/gobject/WebKitDOMPrivate.h:
- bindings/gobject/webkitdom.symbols: Add DataCue symbols.
- bindings/scripts/CodeGeneratorGObject.pm:
(IsPolymorphic): Add TextTrackCue to the list of polymorphic classes.
- 12:51 AM Changeset in webkit [167873] by
-
- 2 edits in trunk/Source/WebKit2
[GTK] GObject introspection links to installed libs when using jhbuild
https://bugs.webkit.org/show_bug.cgi?id=132220
Reviewed by Martin Robinson.
The problem is that gobject-introspection is linking the temporary
binaries adding the library paths present LDFLAGS environment
variable first, taking precedence over the libraries in the build
dir. Since the libraries paths of the dependencies are already
correctly deduced by gobject-introspection using ldd, we can
just unset the LDFLAGS variable before calling g-ir-scanner to
make sure the libraries in the build dir take predence.
- PlatformGTK.cmake:
Apr 27, 2014:
- 11:15 PM Changeset in webkit [167872] by
-
- 4 edits in trunk
[EFL] Remove indexedDB's dependency on leveldb and disable indexedDB
https://bugs.webkit.org/show_bug.cgi?id=132176
Patch by Joonghun Park <jh718.park@samsung.com> on 2014-04-27
Reviewed by Gyuyoung Kim.
.:
EFL port will use sqlite to support indexedDB. So, this patch removes levelDB dependency for now.
- Source/cmake/OptionsEfl.cmake:
Tools:
EFL port will use sqlite to support indexedDB. So, this patch removes levelDB dependency
and turn off indexedDB feature until the sqlite implementation is introduced.
- Scripts/webkitperl/FeatureList.pm:
- 10:52 PM Changeset in webkit [167871] by
-
- 7 edits2 adds in trunk
Store the containing region map inside the flow thread
https://bugs.webkit.org/show_bug.cgi?id=131647
Reviewed by Mihnea Ovidenie.
Source/WebCore:
The patch moves the containing region map inside the flow thread where
it can be better handled in case the region chain changes and the map
needs to be cleared.
As a result of this move we are able to also cleanup the lines region
information of a block flow when it is removed from the tree.
Test: fast/regions/inline-strike-through.html
- rendering/InlineFlowBox.h:
(WebCore::InlineFlowBox::InlineFlowBox):
- rendering/RenderBlockLineLayout.cpp:
(WebCore::RenderBlockFlow::addOverflowFromInlineChildren):
- rendering/RenderFlowThread.cpp:
(WebCore::RenderFlowThread::removeFlowChildInfo):
(WebCore::RenderFlowThread::invalidateRegions):
(WebCore::RenderFlowThread::removeLineRegionInfo):
(WebCore::RenderFlowThread::checkLinesConsistency):
(WebCore::RenderFlowThread::containingRegionMap):
- rendering/RenderFlowThread.h:
- rendering/RootInlineBox.cpp:
(WebCore::containingRegionMap):
(WebCore::RootInlineBox::~RootInlineBox):
(WebCore::RootInlineBox::paint):
(WebCore::RootInlineBox::containingRegion):
(WebCore::RootInlineBox::clearContainingRegion):
(WebCore::RootInlineBox::setContainingRegion):
LayoutTests:
Add a test that verifies the containing region map is properly cleared
when the region chain changes.
- fast/regions/inline-strike-through-expected.txt: Added.
- fast/regions/inline-strike-through.html: Added.
- 9:19 PM Changeset in webkit [167870] by
-
- 5 edits2 adds in trunk
REGRESSION (r159345): The hover state for links in the top navigation of Yahoo.com doesn't work
https://bugs.webkit.org/show_bug.cgi?id=132241
rdar://problem/16501924
Reviewed by Andreas Kling.
Source/WebCore:
Test: fast/text/simple-lines-hover-underline.html
Checked that this does not hurt performance by running the
run-perf-tests PerformanceTests/Layout/line-layout.html command before and after.
- rendering/RenderBlockFlow.cpp:
(WebCore::RenderBlockFlow::hitTestInlineChildren): Use simpleLineLayout function rather than
getting at the data member directly.
(WebCore::RenderBlockFlow::firstLineBaseline): Ditto.
(WebCore::RenderBlockFlow::inlineBlockBaseline): Ditto.
(WebCore::RenderBlockFlow::lineCount): Ditto.
(WebCore::RenderBlockFlow::paintInlineChildren): Ditto.
(WebCore::RenderBlockFlow::hasLines): Ditto.
(WebCore::RenderBlockFlow::simpleLineLayout): Added logic to determine which path to use if
m_lineLayoutPath is undetermined, and call createLineBoxes if it's not simple.
(WebCore::RenderBlockFlow::ensureLineBoxes): Factored out most of the code into a new
createLineBoxes function.
(WebCore::RenderBlockFlow::createLineBoxes): Ditto.
- rendering/RenderBlockFlow.h: Made simpleLineLayout function no longer an inline.
Added a private createLineBoxes function.
LayoutTests:
- fast/text/simple-lines-hover-underline-expected.html: Added.
- fast/text/simple-lines-hover-underline.html: Added.
- fast/text/simple-lines-hover.html: Removed an unneeded style element with a style rule that
does nothing.
- 9:07 PM Changeset in webkit [167869] by
-
- 2 edits in trunk/Source/WebKit2
REGRESSION (r164702): Double tap doesn't stay under the new element once the animation finishes
https://bugs.webkit.org/show_bug.cgi?id=132239
<rdar://problem/16192842>
Reviewed by Darin Adler.
- WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::scalePage):
The change in r167864 broke iOS animated resize, because it was depending on
the dynamic size update code not running if the scale wasn't going to change.
So, as a band-aid we should bail from doing that work if the scales aren't different.
In the long term we should try to untangle this code and make it less platform dependent.
- 7:37 PM Changeset in webkit [167868] by
-
- 10 edits in trunk
[MediaStream] .ended shouldn't be part of MediaStream IDL
https://bugs.webkit.org/show_bug.cgi?id=132104
Patch by Praveen R Jadhav <praveen.j@samsung.com> on 2014-04-27
Reviewed by Eric Carlson.
Source/WebCore:
.ended attribute is spec'ed out of MediaStream IDL. Instead, .active
is introduced to handle more scenarios. This patch replaces all 'ended'
attribute calls with corresponding 'active' attributes.
MediaStream-add-remove-tracks.html is updated.
- Modules/mediastream/MediaStream.cpp:
(WebCore::MediaStream::addTrack): Replaced ended() with active().
(WebCore::MediaStream::removeTrack): setEnded() isn't called. setActive()
is retained which propagates oninactive event.
(WebCore::MediaStream::trackDidEnd): setEnded() isn't called.
(WebCore::MediaStream::removeRemoteSource): Replaced ended() with active().
(WebCore::MediaStream::addRemoteTrack): Replaced ended() with active().
(WebCore::MediaStream::removeRemoteTrack): Replaced ended() with active().
(WebCore::MediaStream::ended): Deleted.
(WebCore::MediaStream::setEnded): Deleted.
(WebCore::MediaStream::streamDidEnd): Deleted.
- Modules/mediastream/MediaStream.h:
- Modules/mediastream/MediaStream.idl:
- Modules/mediastream/RTCPeerConnection.cpp:
(WebCore::RTCPeerConnection::didRemoveRemoteStream): Replaced setEnded()
with setActive().
- platform/mediastream/MediaStreamPrivate.cpp:
(WebCore::MediaStreamPrivate::MediaStreamPrivate):
(WebCore::MediaStreamPrivate::setActive): Updated comment from bug
https://bugs.webkit.org/show_bug.cgi?id=131973
(WebCore::MediaStreamPrivate::setEnded): Deleted.
- platform/mediastream/MediaStreamPrivate.h:
(WebCore::MediaStreamPrivate::ended): Deleted.
LayoutTests:
.ended is not part of MediaStream IDL. Updated test case.
- fast/mediastream/MediaStream-add-remove-tracks-expected.txt:
- fast/mediastream/MediaStream-add-remove-tracks.html:
- 7:09 PM Changeset in webkit [167867] by
-
- 2 edits in trunk/Source/WebKit2
TouchEvent is not handled after releasing any point among touched points.
https://bugs.webkit.org/show_bug.cgi?id=132043
Reviewed by Benjamin Poulain.
Handle TouchEvent until all touched points are released by setting
trackingTouchEvents variable to false when all points are released.
- UIProcess/WebPageProxy.cpp:
(WebKit::areAllTouchPointsReleased):
(WebKit::WebPageProxy::handleTouchEventSynchronously):
(WebKit::WebPageProxy::handleTouchEventAsynchronously):
(WebKit::WebPageProxy::handleTouchEvent):
- 6:55 PM Changeset in webkit [167866] by
-
- 4 edits in trunk/Source/WebKit2
Don't use DispatchMessageEvenWhenWaitingForSyncReply for messages from NetworkProcess
https://bugs.webkit.org/show_bug.cgi?id=132144
Reviewed by Darin Adler.
- NetworkProcess/AsynchronousNetworkLoaderClient.cpp:
(WebKit::AsynchronousNetworkLoaderClient::willSendRequest):
(WebKit::AsynchronousNetworkLoaderClient::canAuthenticateAgainstProtectionSpace):
Dont use the IPC::DispatchMessageEvenWhenWaitingForSyncReply flag. It's not needed,
almost never works in NetworkProcess, but may cause trouble if it did.
- NetworkProcess/NetworkConnectionToWebProcess.cpp:
(WebKit::NetworkConnectionToWebProcess::NetworkConnectionToWebProcess): Removed
setOnlySendMessagesAsDispatchWhenWaitingForSyncReplyWhenProcessingSuchAMessage()
function call, because there are no more any messages in NetworkProcess it could affect.
- Platform/IPC/Connection.cpp: (IPC::Connection::dispatchMessage): Added a FIXME
for an unrelated bug Anders and myself noticed while looking into this.
- 4:13 PM Changeset in webkit [167865] by
-
- 2 edits in trunk/Source/WebKit2
[Cocoa] -[WKWebProcessPlugInController parameters] returns nil if no parameters have been set yet
https://bugs.webkit.org/show_bug.cgi?id=132223
Reviewed by Sam Weinig.
- WebProcess/InjectedBundle/mac/InjectedBundleMac.mm:
(WebKit::InjectedBundle::bundleParameters): Create the WKWebProcessBundleParameters if
needed.
- 3:47 PM Changeset in webkit [167864] by
-
- 2 edits in trunk/Source/WebKit2
REGRESSION (r164702): Double tap doesn't stay under the new element once the animation finishes
https://bugs.webkit.org/show_bug.cgi?id=132239
<rdar://problem/16192842>
Reviewed by Sam Weinig.
- WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::scalePage):
The early-return added to WebPage::scalePage breaks callers who depend
on being able to call scalePage() with the same scale but a different
origin and having that change take effect.
Page::setPageScaleFactor already has the requisite logic, so move
the early return down after that call, and guard only notification
of page scale changes.
- 3:28 PM Changeset in webkit [167863] by
-
- 8 edits1 delete in trunk/Source
[iOS] Stop creating a WKWebResourceQuickLookDelegate for every WebResourceLoader
https://bugs.webkit.org/show_bug.cgi?id=132215
Reviewed by Dan Bernstein.
Source/WebCore:
- WebCore.exp.in: Moved QuickLook symbols to the USE(QUICK_LOOK) stanza.
- platform/network/ios/QuickLook.h: Changed one of the create() overloads to no longer take a delegate argument.
- platform/network/ios/QuickLook.mm: Moved WKWebResourceQuickLookDelegate to here and renamed to WebResourceLoaderQuickLookDelegate.
(WebCore::QuickLookHandle::create): Created a WebResourceLoaderQuickLookDelegate only if QuickLook can handle the response.
Source/WebKit2:
- WebKit2.xcodeproj/project.pbxproj:
- WebProcess/Network/WebResourceLoader.cpp:
(WebKit::WebResourceLoader::didReceiveResponseWithCertificateInfo): Called QuickLookHandle::create() directly.
- WebProcess/Network/WebResourceLoader.h:
- WebProcess/ios/WebResourceLoaderIOS.mm: Removed.
- 2:33 PM Changeset in webkit [167862] by
-
- 2 edits in trunk/LayoutTests
REGRESSION (167840): Layout Test fast/events/shadow-event-path[-2].html is failing
Mark as failing until the shared JS file can be recovered.
- 1:43 PM Changeset in webkit [167861] by
-
- 16 edits in trunk/Source
[iOS WebKit2] Add support for text autosizing
<rdar://problem/16545245>
https://bugs.webkit.org/show_bug.cgi?id=132237
Reviewed by Tim Horton.
../WebCore:
Move text autosizing width from Frame to Page, as it is a Page level concept.
- WebCore.exp.in:
- page/Frame.cpp:
(WebCore::Frame::textAutosizingWidth): Deleted.
(WebCore::Frame::setTextAutosizingWidth): Deleted.
- page/Frame.h:
- page/FrameView.cpp:
(WebCore::FrameView::layout):
- page/Page.cpp:
(WebCore::Page::Page):
- page/Page.h:
(WebCore::Page::textAutosizingWidth):
(WebCore::Page::setTextAutosizingWidth):
../WebKit/mac:
- WebView/WebFrame.mm:
(-[WebFrame _setTextAutosizingWidth:]):
Forward setting of the text autosizing width to the Page.
../WebKit2:
- Shared/WebPageCreationParameters.cpp:
(WebKit::WebPageCreationParameters::encode):
(WebKit::WebPageCreationParameters::decode):
- Shared/WebPageCreationParameters.h:
- UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::creationParameters):
- UIProcess/WebPageProxy.h:
- UIProcess/ios/WebPageProxyIOS.mm:
(WebKit::WebPageProxy::textAutosizingWidth):
- WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::WebPage):
Pass the text autosizing width from the UIProcess to WebProcess.
- 1:42 PM Changeset in webkit [167860] by
-
- 3 edits in trunk/Source/WebKit2
WebKit2 View Gestures (Zoom): Pages with 'background-attachment: fixed' don't behave correctly when zoomed
https://bugs.webkit.org/show_bug.cgi?id=132225
<rdar://problem/15729975>
Reviewed by Darin Adler.
- WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.h:
- WebProcess/WebPage/mac/TiledCoreAnimationDrawingArea.mm:
(WebKit::TiledCoreAnimationDrawingArea::layerForTransientZoom):
(WebKit::TiledCoreAnimationDrawingArea::shadowLayerForTransientZoom):
(WebKit::TiledCoreAnimationDrawingArea::adjustTransientZoom):
(WebKit::TiledCoreAnimationDrawingArea::commitTransientZoom):
(WebKit::TiledCoreAnimationDrawingArea::applyTransientZoomToPage):
Factor out code to choose which layer (and shadow layer) to apply the transient zoom to.
If we have a contentsContainmentLayer (because we have composited background-attachment: fixed),
it applies page scale, so we should apply the transient zoom to that layer
instead of the RenderView's main GraphicsLayer.
- 12:47 PM Changeset in webkit [167859] by
-
- 2 edits in trunk/Source/WebKit2
Unreviewed. iOS build fix.
- UIProcess/ios/SmartMagnificationController.h:
- 11:32 AM Changeset in webkit [167858] by
-
- 2 edits in trunk/LayoutTests
Unreviewed GTK gardening. Adding failure expectations for 6 failing reference tests.
- platform/gtk/TestExpectations:
- 10:03 AM Changeset in webkit [167857] by
-
- 2 edits in trunk/Source/WebCore
Unreviewed. Updating one bindings test baseline after r167855.
- bindings/scripts/test/JS/JSTestCallback.cpp:
(WebCore::JSTestCallback::~JSTestCallback):
- 9:06 AM Changeset in webkit [167856] by
-
- 9 edits in trunk/Source/WebCore
Webpages can trigger loads with invalid URLs
https://bugs.webkit.org/show_bug.cgi?id=132224
rdar://problem/16697142
Reviewed by Alexey Proskuryakov.
Invalid URLs can be a way to trick the user about what website they
are looking at. Still trying to figure out a good way to regression-test this.
- dom/Document.cpp:
(WebCore::Document::processHttpEquiv): Pass a URL rather than a String to
the navigation scheduler.
- loader/FrameLoader.cpp:
(WebCore::FrameLoader::receivedFirstData): Ditto.
- loader/NavigationScheduler.cpp:
(WebCore::ScheduledURLNavigation::ScheduledURLNavigation): Take a URL rather
than a string.
(WebCore::ScheduledURLNavigation::url): Ditto.
(WebCore::ScheduledRedirect::ScheduledRedirect): Ditto.
(WebCore::ScheduledLocationChange::ScheduledLocationChange): Ditto.
(WebCore::ScheduledRefresh::ScheduledRefresh): Ditto.
(WebCore::NavigationScheduler::shouldScheduleNavigation): Added a check that
prevents navigation to any URL that is invalid, except for JavaScript URLs,
which need not be valid.
(WebCore::NavigationScheduler::scheduleRedirect): Use URL instead of String.
(WebCore::NavigationScheduler::scheduleLocationChange): Use URL instead of
String. Also got rid of empty string check since empty URLs are also invalid,
and so shouldScheduleNavigation will take care of it.
(WebCore::NavigationScheduler::scheduleRefresh): Use URL instead of String.
- loader/NavigationScheduler.h: Take URL instead of String. Also removed some
unneeded incldues and uses of WTF_MAKE_NONCOPYABLE. NavigationScheduler is
already noncopyable because it has a reference for a data member, and the
disabler doesn't have any real reason to be noncopyable.
- loader/SubframeLoader.cpp:
(WebCore::SubframeLoader::loadOrRedirectSubframe): Pass a URL rather than a
String to the NavigationScheduler.
- page/DOMWindow.cpp:
(WebCore::DOMWindow::createWindow): Ditto.
- page/SecurityOrigin.cpp:
(WebCore::SecurityOrigin::urlWithUniqueSecurityOrigin): Return a URL instead
of a String.
- page/SecurityOrigin.h: Updated for above change.
- 7:07 AM Changeset in webkit [167855] by
-
- 35 edits in trunk/Source/WebCore
ScriptExecutionContext::Task should work well with C++11 lambdas
https://bugs.webkit.org/show_bug.cgi?id=129795
Reviewed by Darin Adler.
Instead of having classes that subclass ScriptExecutionContext::Task and override
the performTask(ScriptExecutionContext*) method, have the ScriptExecutionContext::Task
take in a std::function<void (ScriptExecutionContext*)>-like object trough the constructor
which would contain the code currently kept in the performTask() methods.
This enables inlining C++11 lambdas into ScriptExecutionContext::postTask() calls. For
cleanup tasks, the Task object can be implicitly constructed by using the initializer list
constructor with the first argument being the ScriptExecutionContext::Task::CleanupTask tag.
The ScriptExecutionContext class remains non-copyable and now stores the passed-in invokable
object in the std::function wrapper, along with a boolean member that indicates whether the
task is of cleanup nature.
- Modules/quota/StorageErrorCallback.cpp:
(WebCore::StorageErrorCallback::CallbackTask::CallbackTask):
(WebCore::StorageErrorCallback::CallbackTask::performTask): Deleted.
- Modules/quota/StorageErrorCallback.h:
(WebCore::StorageErrorCallback::CallbackTask::create): Deleted.
- Modules/webdatabase/Database.cpp:
(WebCore::Database::~Database):
(WebCore::Database::runTransaction):
(WebCore::Database::scheduleTransactionCallback):
(WebCore::DerefContextTask::create): Deleted.
(WebCore::DerefContextTask::performTask): Deleted.
(WebCore::DerefContextTask::isCleanupTask): Deleted.
(WebCore::DerefContextTask::DerefContextTask): Deleted.
(WebCore::callTransactionErrorCallback): Deleted.
(WebCore::DeliverPendingCallbackTask::create): Deleted.
(WebCore::DeliverPendingCallbackTask::performTask): Deleted.
(WebCore::DeliverPendingCallbackTask::DeliverPendingCallbackTask): Deleted.
- Modules/webdatabase/DatabaseManager.cpp:
(WebCore::DatabaseManager::openDatabase):
(WebCore::DatabaseCreationCallbackTask::create): Deleted.
(WebCore::DatabaseCreationCallbackTask::performTask): Deleted.
(WebCore::DatabaseCreationCallbackTask::DatabaseCreationCallbackTask): Deleted.
- Modules/webdatabase/DatabaseSync.cpp:
(WebCore::CloseSyncDatabaseOnContextThreadTask::create): Deleted.
(WebCore::CloseSyncDatabaseOnContextThreadTask::performTask): Deleted.
(WebCore::CloseSyncDatabaseOnContextThreadTask::CloseSyncDatabaseOnContextThreadTask): Deleted.
- Modules/webdatabase/SQLCallbackWrapper.h:
(WebCore::SQLCallbackWrapper::clear):
(WebCore::SQLCallbackWrapper::SafeReleaseTask::create): Deleted.
(WebCore::SQLCallbackWrapper::SafeReleaseTask::performTask): Deleted.
(WebCore::SQLCallbackWrapper::SafeReleaseTask::isCleanupTask): Deleted.
(WebCore::SQLCallbackWrapper::SafeReleaseTask::SafeReleaseTask): Deleted.
- Modules/websockets/ThreadableWebSocketChannelClientWrapper.cpp:
(WebCore::ThreadableWebSocketChannelClientWrapper::didConnect):
(WebCore::ThreadableWebSocketChannelClientWrapper::didReceiveMessage):
(WebCore::ThreadableWebSocketChannelClientWrapper::didReceiveBinaryData):
(WebCore::ThreadableWebSocketChannelClientWrapper::didUpdateBufferedAmount):
(WebCore::ThreadableWebSocketChannelClientWrapper::didStartClosingHandshake):
(WebCore::ThreadableWebSocketChannelClientWrapper::didClose):
(WebCore::ThreadableWebSocketChannelClientWrapper::didReceiveMessageError):
(WebCore::ThreadableWebSocketChannelClientWrapper::processPendingTasks):
- Modules/websockets/ThreadableWebSocketChannelClientWrapper.h:
- Modules/websockets/WorkerThreadableWebSocketChannel.cpp:
(WebCore::WorkerThreadableWebSocketChannel::Bridge::mainThreadInitialize):
(WebCore::WorkerThreadableWebSocketChannel::mainThreadDestroy):
(WebCore::WorkerThreadableWebSocketChannel::Bridge::disconnect):
(WebCore::WorkerThreadableWebSocketChannel::WorkerGlobalScopeDidInitializeTask::create): Deleted.
(WebCore::WorkerThreadableWebSocketChannel::WorkerGlobalScopeDidInitializeTask::~WorkerGlobalScopeDidInitializeTask): Deleted.
(WebCore::WorkerThreadableWebSocketChannel::WorkerGlobalScopeDidInitializeTask::WorkerGlobalScopeDidInitializeTask): Deleted.
- Modules/websockets/WorkerThreadableWebSocketChannel.h:
- bindings/js/JSCallbackData.h:
(WebCore::DeleteCallbackDataTask::DeleteCallbackDataTask):
(WebCore::DeleteCallbackDataTask::create): Deleted.
(WebCore::DeleteCallbackDataTask::performTask): Deleted.
(WebCore::DeleteCallbackDataTask::isCleanupTask): Deleted.
- bindings/js/JSDOMGlobalObjectTask.cpp:
(WebCore::JSGlobalObjectTask::JSGlobalObjectTask):
(WebCore::JSGlobalObjectTask::~JSGlobalObjectTask): Deleted.
(WebCore::JSGlobalObjectTask::performTask): Deleted.
- bindings/js/JSDOMGlobalObjectTask.h:
- bindings/js/JSDOMWindowBase.cpp:
(WebCore::JSDOMWindowBase::queueTaskToEventLoop):
- bindings/js/JSWorkerGlobalScopeBase.cpp:
(WebCore::JSWorkerGlobalScopeBase::queueTaskToEventLoop):
- bindings/scripts/CodeGeneratorJS.pm:
(GenerateCallbackImplementation):
- dom/CrossThreadTask.h:
(WebCore::CrossThreadTask1::performTask):
(WebCore::CrossThreadTask2::performTask):
(WebCore::CrossThreadTask3::performTask):
(WebCore::CrossThreadTask4::performTask):
(WebCore::CrossThreadTask5::performTask):
(WebCore::CrossThreadTask6::performTask):
(WebCore::CrossThreadTask7::performTask):
(WebCore::CrossThreadTask8::performTask):
(WebCore::createCallbackTask):
(WebCore::CrossThreadTask1::create): Deleted.
(WebCore::CrossThreadTask2::create): Deleted.
(WebCore::CrossThreadTask3::create): Deleted.
(WebCore::CrossThreadTask4::create): Deleted.
(WebCore::CrossThreadTask5::create): Deleted.
(WebCore::CrossThreadTask6::create): Deleted.
(WebCore::CrossThreadTask7::create): Deleted.
(WebCore::CrossThreadTask8::create): Deleted.
- dom/Document.cpp:
(WebCore::Document::addConsoleMessage):
(WebCore::Document::addMessage):
(WebCore::Document::postTask):
(WebCore::Document::pendingTasksTimerFired):
(WebCore::PerformTaskContext::PerformTaskContext): Deleted.
(WebCore::Document::didReceiveTask): Deleted.
- dom/Document.h:
- dom/ScriptExecutionContext.cpp:
(WebCore::ScriptExecutionContext::processMessagePortMessagesSoon):
(WebCore::ScriptExecutionContext::timerAlignmentInterval):
(WebCore::ProcessMessagesSoonTask::create): Deleted.
(WebCore::ScriptExecutionContext::AddConsoleMessageTask::performTask): Deleted.
(WebCore::ScriptExecutionContext::Task::~Task): Deleted.
- dom/ScriptExecutionContext.h:
(WebCore::ScriptExecutionContext::Task::Task):
(WebCore::ScriptExecutionContext::Task::performTask):
(WebCore::ScriptExecutionContext::Task::isCleanupTask):
(WebCore::ScriptExecutionContext::AddConsoleMessageTask::AddConsoleMessageTask):
(WebCore::ScriptExecutionContext::AddConsoleMessageTask::create): Deleted.
- dom/StringCallback.cpp:
(WebCore::StringCallback::scheduleCallback):
- loader/appcache/ApplicationCacheGroup.cpp:
(WebCore::ApplicationCacheGroup::postListenerTask):
(WebCore::CallCacheListenerTask::create): Deleted.
(WebCore::CallCacheListenerTask::CallCacheListenerTask): Deleted.
- workers/DefaultSharedWorkerRepository.cpp:
(WebCore::SharedWorkerProxy::postTaskToLoader):
(WebCore::SharedWorkerProxy::postTaskForModeToWorkerGlobalScope):
(WebCore::SharedWorkerConnectTask::SharedWorkerConnectTask):
(WebCore::DefaultSharedWorkerRepository::workerScriptLoaded):
(WebCore::DefaultSharedWorkerRepository::connectToWorker):
(WebCore::SharedWorkerConnectTask::create): Deleted.
(WebCore::SharedWorkerConnectTask::performTask): Deleted.
- workers/WorkerEventQueue.cpp:
(WebCore::WorkerEventQueue::EventDispatcher::EventDispatcher):
(WebCore::WorkerEventQueue::EventDispatcher::~EventDispatcher):
(WebCore::WorkerEventQueue::EventDispatcher::dispatch):
(WebCore::WorkerEventQueue::enqueueEvent):
(WebCore::WorkerEventQueue::cancelEvent):
(WebCore::WorkerEventQueue::close):
- workers/WorkerEventQueue.h:
- workers/WorkerGlobalScope.cpp:
(WebCore::WorkerGlobalScope::close):
(WebCore::WorkerGlobalScope::postTask):
(WebCore::WorkerGlobalScope::addConsoleMessage):
(WebCore::WorkerGlobalScope::addMessage):
(WebCore::CloseWorkerGlobalScopeTask::create): Deleted.
(WebCore::CloseWorkerGlobalScopeTask::performTask): Deleted.
(WebCore::CloseWorkerGlobalScopeTask::isCleanupTask): Deleted.
- workers/WorkerGlobalScope.h:
- workers/WorkerLoaderProxy.h:
- workers/WorkerMessagingProxy.cpp:
(WebCore::WorkerMessagingProxy::postMessageToWorkerObject):
(WebCore::WorkerMessagingProxy::postMessageToWorkerGlobalScope):
(WebCore::WorkerMessagingProxy::postTaskForModeToWorkerGlobalScope):
(WebCore::WorkerMessagingProxy::postTaskToLoader):
(WebCore::WorkerMessagingProxy::postExceptionToWorkerObject):
(WebCore::WorkerMessagingProxy::workerThreadCreated):
(WebCore::WorkerMessagingProxy::notifyNetworkStateChange):
(WebCore::WorkerMessagingProxy::workerGlobalScopeDestroyed):
(WebCore::WorkerMessagingProxy::workerGlobalScopeClosed):
(WebCore::WorkerMessagingProxy::postMessageToPageInspector):
(WebCore::WorkerMessagingProxy::confirmMessageFromWorkerObject):
(WebCore::WorkerMessagingProxy::reportPendingActivity):
(WebCore::MessageWorkerGlobalScopeTask::create): Deleted.
(WebCore::MessageWorkerGlobalScopeTask::MessageWorkerGlobalScopeTask): Deleted.
(WebCore::MessageWorkerGlobalScopeTask::performTask): Deleted.
(WebCore::MessageWorkerTask::create): Deleted.
(WebCore::MessageWorkerTask::MessageWorkerTask): Deleted.
(WebCore::MessageWorkerTask::performTask): Deleted.
(WebCore::WorkerExceptionTask::create): Deleted.
(WebCore::WorkerExceptionTask::WorkerExceptionTask): Deleted.
(WebCore::WorkerExceptionTask::performTask): Deleted.
(WebCore::WorkerGlobalScopeDestroyedTask::create): Deleted.
(WebCore::WorkerGlobalScopeDestroyedTask::WorkerGlobalScopeDestroyedTask): Deleted.
(WebCore::WorkerGlobalScopeDestroyedTask::performTask): Deleted.
(WebCore::WorkerTerminateTask::create): Deleted.
(WebCore::WorkerTerminateTask::WorkerTerminateTask): Deleted.
(WebCore::WorkerTerminateTask::performTask): Deleted.
(WebCore::WorkerThreadActivityReportTask::create): Deleted.
(WebCore::WorkerThreadActivityReportTask::WorkerThreadActivityReportTask): Deleted.
(WebCore::WorkerThreadActivityReportTask::performTask): Deleted.
(WebCore::PostMessageToPageInspectorTask::create): Deleted.
(WebCore::PostMessageToPageInspectorTask::PostMessageToPageInspectorTask): Deleted.
(WebCore::PostMessageToPageInspectorTask::performTask): Deleted.
(WebCore::NotifyNetworkStateChangeTask::create): Deleted.
(WebCore::NotifyNetworkStateChangeTask::NotifyNetworkStateChangeTask): Deleted.
(WebCore::NotifyNetworkStateChangeTask::performTask): Deleted.
- workers/WorkerMessagingProxy.h:
- workers/WorkerRunLoop.cpp:
(WebCore::WorkerRunLoop::postTask):
(WebCore::WorkerRunLoop::postTaskAndTerminate):
(WebCore::WorkerRunLoop::postTaskForMode):
(WebCore::WorkerRunLoop::Task::create):
(WebCore::WorkerRunLoop::Task::performTask):
(WebCore::WorkerRunLoop::Task::Task):
- workers/WorkerRunLoop.h:
- workers/WorkerThread.cpp:
(WebCore::WorkerThread::stop):
(WebCore::WorkerThread::releaseFastMallocFreeMemoryInAllThreads):
(WebCore::WorkerThreadShutdownFinishTask::create): Deleted.
(WebCore::WorkerThreadShutdownFinishTask::performTask): Deleted.
(WebCore::WorkerThreadShutdownFinishTask::isCleanupTask): Deleted.
(WebCore::WorkerThreadShutdownStartTask::create): Deleted.
(WebCore::WorkerThreadShutdownStartTask::performTask): Deleted.
(WebCore::WorkerThreadShutdownStartTask::isCleanupTask): Deleted.
- 6:40 AM Changeset in webkit [167854] by
-
- 44 edits in trunk/Source/WebKit2
Move cross-port WebKit2 code to std::unique_ptr
https://bugs.webkit.org/show_bug.cgi?id=129670
Reviewed by Darin Adler.
Replace uses of OwnPtr, PassOwnPtr in cross-port WebKit2 code with std::unique_ptr.
- DatabaseProcess/IndexedDB/sqlite/UniqueIDBDatabaseBackingStoreSQLite.cpp:
(WebKit::UniqueIDBDatabaseBackingStoreSQLite::createIndex):
- DatabaseProcess/IndexedDB/sqlite/UniqueIDBDatabaseBackingStoreSQLite.h:
- NetworkProcess/NetworkProcess.h:
(WebKit::NetworkProcess::addSupplement):
- NetworkProcess/SynchronousNetworkLoaderClient.cpp:
(WebKit::SynchronousNetworkLoaderClient::didReceiveBuffer):
- NetworkProcess/SynchronousNetworkLoaderClient.h:
- Platform/IPC/Connection.cpp:
(IPC::Connection::dispatchWorkQueueMessageReceiverMessage):
- Platform/IPC/Connection.h:
- Platform/IPC/MessageReceiver.h:
- Platform/Module.h:
- Platform/WorkQueue.h:
- Platform/efl/ModuleEfl.cpp:
(WebKit::Module::load):
- Platform/mac/LayerHostingContext.mm:
- Shared/APIURL.h:
- Shared/Network/CustomProtocols/CustomProtocolManager.h:
- Shared/Plugins/NPRemoteObjectMap.cpp:
- Shared/ShareableBitmap.h:
- UIProcess/DrawingAreaProxyImpl.h:
- UIProcess/InspectorServer/WebSocketServer.cpp:
(WebKit::WebSocketServer::didAcceptConnection):
(WebKit::WebSocketServer::didCloseWebSocketServerConnection):
- UIProcess/InspectorServer/WebSocketServer.h:
- UIProcess/InspectorServer/soup/WebSocketServerSoup.cpp:
(WebKit::connectionCallback):
- UIProcess/Network/NetworkProcessProxy.cpp:
(WebKit::NetworkProcessProxy::createDownloadProxy):
- UIProcess/Network/NetworkProcessProxy.h:
- UIProcess/Notifications/WebNotificationManagerProxy.h:
- UIProcess/Plugins/WebPluginSiteDataManager.cpp:
(WebKit::WebPluginSiteDataManager::getSitesWithData):
(WebKit::WebPluginSiteDataManager::clearSiteData):
(WebKit::WebPluginSiteDataManager::didGetSitesWithDataForAllPlugins):
(WebKit::WebPluginSiteDataManager::didClearSiteDataForAllPlugins):
- UIProcess/Plugins/WebPluginSiteDataManager.h:
- UIProcess/Storage/StorageManager.cpp:
(WebKit::callCallbackFunction):
(WebKit::StorageManager::getOriginsInternal):
- UIProcess/WebPageProxy.cpp:
(WebKit::ExceededDatabaseQuotaRecords::areBeingProcessed):
(WebKit::ExceededDatabaseQuotaRecords::createRecord):
(WebKit::ExceededDatabaseQuotaRecords::add):
(WebKit::ExceededDatabaseQuotaRecords::next):
(WebKit::WebPageProxy::handleMouseEvent):
(WebKit::WebPageProxy::handleWheelEvent):
(WebKit::WebPageProxy::processNextQueuedWheelEvent):
(WebKit::WebPageProxy::internalShowContextMenu):
- UIProcess/WebPageProxy.h:
- UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::createDownloadProxy):
- UIProcess/WebProcessProxy.h:
- WebProcess/Plugins/Netscape/NetscapeBrowserFuncs.cpp:
(WebKit::PluginDestructionProtector::PluginDestructionProtector):
- WebProcess/Plugins/Netscape/NetscapePluginStream.cpp:
(WebKit::NetscapePluginStream::deliverData):
- WebProcess/Plugins/Netscape/NetscapePluginStream.h:
- WebProcess/Plugins/PluginProcessConnectionManager.h:
- WebProcess/Plugins/PluginProxy.cpp:
(WebKit::PluginProxy::initialize):
(WebKit::PluginProxy::didCreatePluginInternal):
(WebKit::PluginProxy::didFailToCreatePluginInternal):
- WebProcess/Plugins/PluginProxy.h:
- WebProcess/Plugins/PluginView.cpp:
(WebKit::PluginView::createWebEvent):
(WebKit::PluginView::handleEvent):
- WebProcess/Plugins/PluginView.h:
- WebProcess/WebCoreSupport/WebPopupMenu.h:
- WebProcess/WebPage/EventDispatcher.cpp:
(WebKit::EventDispatcher::EventDispatcher):
- WebProcess/WebPage/EventDispatcher.h:
- WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::WebPage):
(WebKit::WebPage::beginPrinting):
- WebProcess/WebPage/WebPage.h:
- WebProcess/WebProcess.h:
(WebKit::WebProcess::addSupplement):
- 4:53 AM Changeset in webkit [167853] by
-
- 9 edits in trunk/Source
Coalesce responses on network process side
https://bugs.webkit.org/show_bug.cgi?id=132229
Reviewed by Andreas Kling.
Source/WebCore:
- WebCore.exp.in:
Source/WebKit2:
To reduce IPC we should coalesce response data in the network process and send it over with single IPC call.
- NetworkProcess/AsynchronousNetworkLoaderClient.cpp:
(WebKit::AsynchronousNetworkLoaderClient::AsynchronousNetworkLoaderClient):
(WebKit::AsynchronousNetworkLoaderClient::didReceiveResponse):
(WebKit::AsynchronousNetworkLoaderClient::didReceiveBuffer):
(WebKit::AsynchronousNetworkLoaderClient::didFinishLoading):
(WebKit::AsynchronousNetworkLoaderClient::didFail):
(WebKit::AsynchronousNetworkLoaderClient::dispatchPartialCoalescedResponse):
(WebKit::AsynchronousNetworkLoaderClient::clearCoalescedResponse):
(WebKit::AsynchronousNetworkLoaderClient::responseCoalesceTimerFired):
Coalesce the response. Completed response is sent on didFinishLoading. If the coalesce timer fires
before that the data accumulated so far is dispatched.
- NetworkProcess/AsynchronousNetworkLoaderClient.h:
- NetworkProcess/NetworkResourceLoader.h:
- WebProcess/Network/WebResourceLoader.cpp:
(WebKit::WebResourceLoader::didReceiveCompleteResponse):
- WebProcess/Network/WebResourceLoader.h:
- WebProcess/Network/WebResourceLoader.messages.in:
Add a new message type that covers didReceiveResponse, didReceiveBuffer and didFinishLoading in a single message.
- 1:15 AM Changeset in webkit [167852] by
-
- 3 edits in trunk/Source/WebCore
Roll out changes not part of the patch reviewed for Bug 132089
<http://webkit.org/b/132089>
- loader/SubframeLoader.cpp:
(WebCore::SubframeLoader::loadOrRedirectSubframe):
- page/DOMWindow.cpp:
(WebCore::DOMWindow::setLocation):
(WebCore::DOMWindow::createWindow):
(WebCore::DOMWindow::open):
