Timeline
05/24/13: Today
- 11:09 Changeset [150648] by
-
Intermittent crash when typing in Address Bar
https://bugs.webkit.org/show_bug.cgi?id=116728
<rdar://problem/13976781>
Reviewed by Sam Weinig.
If a web process connection has been removed from the storage manager because we disconnected it
since there were no more active pages associated with the web process, we'll fail to find it here
so just return early instead of crashing.
- UIProcess/Storage/StorageManager.cpp:
(WebKit::StorageManager::destroyStorageMap):
- 11:01 Changeset [150647] by
-
Versioning.
- 10:28 Changeset [150646] by
-
Path: upstream the missing bits from the BlackBerry port
https://bugs.webkit.org/show_bug.cgi?id=116658
Patch by Alberto Garcia <agarcia@igalia.com> on 2013-05-24
Reviewed by Xan Lopez.
PathBlackBerry was upstreamed a while ago (r144612) but a few bits
were missing from the cross-platform code.
- platform/graphics/Path.cpp:
(WebCore::Path::addPathForRoundedRect):
- platform/graphics/Path.h:
(Graphics):
(Path):
- 10:16 Changeset [150645] by
-
Improve loader/go-back-cached-main-resource.html test
https://bugs.webkit.org/show_bug.cgi?id=116709
Patch by Manuel Rego Casasnovas <rego@igalia.com> on 2013-05-24
Reviewed by Alexey Proskuryakov.
It is only needed to register the listeners on the initial load as they
are already present when restoring from page cache. Modified the test to
use event.persisted property to know if we are handling the first load
in order to register the listeners only once.
- loader/resources/first-page.html: Rename registerListeners() function
to a more accurate name didShowPage() and use event.persisted to
register the listeners only in the first load.
- loader/resources/other-page.html: Ditto.
- 09:50 WebKit Team edited by
- (diff)
- 09:46 Changeset [150644] by
-
[sh4] Optimize LLINT generated code and fix few bugs in baseline JIT.
https://bugs.webkit.org/show_bug.cgi?id=116716
Patch by Julien Brianceau <jbrianceau@nds.com> on 2013-05-24
Reviewed by Geoffrey Garen.
- assembler/MacroAssemblerSH4.h:
(JSC::MacroAssemblerSH4::mul32): Cosmetic changes.
(JSC::MacroAssemblerSH4::convertInt32ToDouble): Absolute address was not dereferenced.
(JSC::MacroAssemblerSH4::branch32): Absolute address was not dereferenced.
(JSC::MacroAssemblerSH4::revertJumpReplacementToBranchPtrWithPatch): Use all 32 bits of pointer for revertJump call.
- assembler/SH4Assembler.h:
(JSC::SH4Assembler::revertJump): Use changePCrelativeAddress to patch the whole pointer.
(JSC::SH4Assembler::linkJump): Cosmetic change.
- offlineasm/sh4.rb: Optimize LLINT generated code.
- 09:34 Changeset [150643] by
-
WebProcess is crashing on http://achicu.github.io/css-presentation when direct pattern compositing is enabled
https://bugs.webkit.org/show_bug.cgi?id=115135
Patch by Noam Rosenthal <noam@webkit.org> on 2013-05-24
Reviewed by Antti Koivisto.
Source/WebCore:
Issue is invoked when there is a visibility:hidden layer with a non-compositable
background image (e.g. a radial gradient).
The layer is treated as a no-contents layer because of its visibility, however
updateDirectlyCompositedBackgroundImage() later mistakenly treats it as a layer
with direct background-image compositing.
Fixed by returning early from updating directly composited contents for layers
with hidden contents.
Test: compositing/patterns/direct-pattern-compositing-hidden-radial-gradient.html
- rendering/RenderLayerBacking.cpp:
(WebCore::RenderLayerBacking::updateDirectlyCompositedContents):
(WebCore::RenderLayerBacking::updateDirectlyCompositedBackgroundImage):
(WebCore::RenderLayerBacking::paintsChildren):
(WebCore::RenderLayerBacking::isDirectlyCompositedImage):
LayoutTests:
Added a ref-test that makes sure that a layer with a hidden radial gradient does not
cause the web process to crash.
- compositing/patterns/direct-pattern-compositing-hidden-radial-gradient-expected.html: Added.
- compositing/patterns/direct-pattern-compositing-hidden-radial-gradient.html: Added.
- 09:18 Changeset [150642] by
-
Refactor shouldAddBorderPaddingMargin()
https://bugs.webkit.org/show_bug.cgi?id=98803
Reviewed by Ryosuke Niwa.
Make this helper function less clever-stupid in its misguided sacrifice of intelligibility
for the sake of concision.
No new tests, refactoring.
- rendering/RenderBlockLineLayout.cpp:
(WebCore::shouldAddBorderPaddingMargin):
(WebCore::inlineLogicalWidth):
- 08:51 Changeset [150641] by
-
[BlackBerry] Fix bad type warning in InputHandler log.
https://bugs.webkit.org/show_bug.cgi?id=116720
Reviewed by Carlos Garcia Campos.
Fix warning in inputLog string.
- WebKitSupport/InputHandler.cpp:
(BlackBerry::WebKit::InputHandler::handleKeyboardInput):
- 07:48 Changeset [150640] by
-
Update WTF::Vector pretty printers
https://bugs.webkit.org/show_bug.cgi?id=116719
Patch by Simon Pena <simon.pena@samsung.com> on 2013-05-24
Reviewed by Anders Carlsson.
After r148891, WTF::Vector pretty printers were left outdated, as
the WTF::Vector was made to extend a VectorBuffer, instead of keeping
one as a member. This patch updates the pretty printers so they work
again.
- gdb/webkit.py:
(WTFVectorPrinter.children):
(WTFVectorPrinter.to_string):
- 07:20 Changeset [150639] by
-
[Qt] Unreviewed gardening. Update qt-5.0-wk2 expected pictures after r150506.
Patch by Gabor Abraham <abrhm@inf.u-szeged.hu> on 2013-05-24
- platform/qt-5.0-wk2/svg/dynamic-updates/SVGFEDisplacementMapElement-dom-in-attr-expected.png:
- platform/qt-5.0-wk2/svg/dynamic-updates/SVGFEDisplacementMapElement-dom-in2-attr-expected.png:
- platform/qt-5.0-wk2/svg/dynamic-updates/SVGFEDisplacementMapElement-dom-scale-attr-expected.png:
- platform/qt-5.0-wk2/svg/dynamic-updates/SVGFEDisplacementMapElement-dom-xChannelSelector-attr-expected.png:
- platform/qt-5.0-wk2/svg/dynamic-updates/SVGFEDisplacementMapElement-dom-yChannelSelector-attr-expected.png:
- platform/qt-5.0-wk2/svg/dynamic-updates/SVGFEDisplacementMapElement-svgdom-in-prop-expected.png:
- platform/qt-5.0-wk2/svg/dynamic-updates/SVGFEDisplacementMapElement-svgdom-in2-prop-expected.png:
- platform/qt-5.0-wk2/svg/dynamic-updates/SVGFEDisplacementMapElement-svgdom-scale-prop-expected.png:
- platform/qt-5.0-wk2/svg/dynamic-updates/SVGFEDisplacementMapElement-svgdom-xChannelSelector-prop-expected.png:
- platform/qt-5.0-wk2/svg/dynamic-updates/SVGFEDisplacementMapElement-svgdom-yChannelSelector-prop-expected.png:
- platform/qt-5.0-wk2/svg/dynamic-updates/SVGFEImageElement-dom-preserveAspectRatio-attr-expected.png:
- platform/qt-5.0-wk2/svg/dynamic-updates/SVGFEImageElement-svgdom-preserveAspectRatio-prop-expected.png:
- platform/qt-5.0-wk2/svg/filters/feImage-animated-transform-on-target-rect-expected.png:
- platform/qt-5.0-wk2/svg/filters/feImage-late-indirect-update-expected.png: Added.
- platform/qt-5.0-wk2/svg/filters/feImage-reference-invalidation-expected.png: Added.
- platform/qt-5.0-wk2/svg/filters/feImage-target-add-to-document-expected.png: Added.
- platform/qt-5.0-wk2/svg/filters/feImage-target-attribute-change-expected.png: Added.
- platform/qt-5.0-wk2/svg/filters/feImage-target-attribute-change-with-use-indirection-2-expected.png: Added.
- platform/qt-5.0-wk2/svg/filters/feImage-target-attribute-change-with-use-indirection-expected.png: Added.
- platform/qt-5.0-wk2/svg/filters/feImage-target-changes-id-expected.png: Added.
- platform/qt-5.0-wk2/svg/filters/feImage-target-id-change-expected.png: Added.
- platform/qt-5.0-wk2/svg/filters/feImage-target-inline-style-change-expected.png: Added.
- platform/qt-5.0-wk2/svg/filters/feImage-target-property-change-expected.png: Added.
- platform/qt-5.0-wk2/svg/filters/feImage-target-reappend-to-document-expected.png: Added.
- platform/qt-5.0-wk2/svg/filters/feImage-target-remove-from-document-expected.png: Added.
- platform/qt-5.0-wk2/svg/filters/feImage-target-style-change-expected.png: Added.
- 06:25 Changeset [150638] by
-
[Qt] Unreviewed gardening. Skip failing tests.
- platform/qt-5.0-wk2/TestExpectations: Move http/tests/security/cross-frame-access-put.html to qt/TestExpectations.
- platform/qt/TestExpectations: Skip fast/frames/detach-frame-during-focus.html after r146523.
- 05:47 Changeset [150637] by
-
[BlackBerry] Respect tabindex when using form controls.
https://bugs.webkit.org/show_bug.cgi?id=116676
Reviewed by Xan Lopez.
PR 337419.
Update form control navigation tracking to take tabindex
order into account when calculating next and previous nodes.
The order to follow is tab index 1 - N, followed by all items
without a tab index or tab index 0 in rendering order to match
with standard desktop behavior.
Internally Reviewed by Genevieve Mak and Nima Ghanavatian.
- WebKitSupport/InputHandler.cpp:
(BlackBerry::WebKit::InputHandler::updateFormState):
- 05:23 WebKitGTK/TrackingMemoryErrors edited by
- (diff)
- 01:41 Changeset [150636] by
-
Update prepareConsistentTestingEnvironment to use a non-deprecated API.
https://bugs.webkit.org/show_bug.cgi?id=116710
Reviewed by Mark Rowe.
- DumpRenderTree/mac/DumpRenderTree.mm:
(prepareConsistentTestingEnvironment): Use beginActivityWithOptions instead of beginSuspensionOfSystemBehaviors.
- 01:05 Changeset [150635] by
-
[BlackBerry] DRT - crashed on GraphicsContext3D::makeContextCurrent
https://bugs.webkit.org/show_bug.cgi?id=116654
Patch by Xiaobo Wang <xiaobwang@blackberry.com> on 2013-05-24
Reviewed by Carlos Garcia Campos.
PR 331552.
Internally reviewed and amended by Arvid Nilsson.
No new tests, this patch is to fix DRT crashes in fast/canvas/webgl.
The instance of GraphicsContext3D has already been destroyed but
WebGLLayerWebKitThread is not aware of that. This is because compositor
layer of the WebGL graphics context was referenced outside the context.
So when destructing GraphicsContext3D the compositing layer was not
destructed.
Fix by setting m_webGLContext to 0 when destructing GrapthicsContext3D
after releasing the compositor layer, so that it won't be de-referenced
again later.
- platform/graphics/blackberry/GraphicsContext3DBlackBerry.cpp:
(WebCore::GraphicsContext3D::~GraphicsContext3D):
- platform/graphics/blackberry/WebGLLayerWebKitThread.cpp:
(WebCore::WebGLLayerWebKitThread::~WebGLLayerWebKitThread):
(WebCore::WebGLLayerWebKitThread::webGLContextDestroyed):
(WebCore):
- platform/graphics/blackberry/WebGLLayerWebKitThread.h:
(WebGLLayerWebKitThread):
- 00:32 Changeset [150634] by
-
[BlackBerry] Remove SKIA leftovers from WebPage.cpp
https://bugs.webkit.org/show_bug.cgi?id=116656
Patch by Jeff Rogers <jrogers@rim.com> on 2013-05-24
Reviewed by Rob Buis.
- Api/WebPage.cpp:
(BlackBerry::WebKit::WebPagePrivate::setLoadState):
- 00:12 Changeset [150633] by
-
Typing in Safari's unified field causes unnecessary web content repaints.
<http://webkit.org/b/116703>
Reviewed by Andy Estes.
Source/WebCore:
Break out an alternate version of Page::markAllMatchesForText() that only counts the number of times
a string occurs in the page, but doesn't try to mark the occurrences.
This allows Safari to count matches for the 'Find "foo" on This Page' functionality in the unified
location bar without causing tile repaints.
- WebCore.exp.in:
- page/Page.h:
- page/Page.cpp:
(WebCore::Page::findMatchesForText):
(WebCore::Page::markAllMatchesForText):
(WebCore::Page::countFindMatches):
Renamed markAllMatchesForText() to findMatchesForText() and gave it a boolean parameter to control whether
matches should be marked or not. countFindMatches() is the new helper that doesn't mark.
Source/WebKit2:
- WebProcess/WebPage/FindController.cpp:
(WebKit::FindController::countStringMatches):
Call WebCore::Page::countMatchesForText() instead of markAllMatchesForText(). If the search string
is found inside the page content, the renderer containing that text will no longer be repainted.
- 00:05 Changeset [150632] by
-
Add back a failing test expectation on fast/canvas/webgl/object-deletion-behaviour.html
after r150558 on Lion Debug WK1.
- platform/mac-wk2/TestExpectations:
- platform/mac/TestExpectations:
- 00:02 Changeset [150631] by
-
https://bugs.webkit.org/show_bug.cgi?id=116702
REGRESSION (r150399): Scrollbar missing for all overflow scroll
-and corresponding-
<rdar://problem/13963712>
Reviewed by Simon Fraser.
This was a silly copy-paste error.
- rendering/RenderLayer.cpp:
(WebCore::RenderLayer::scrollbarAnimationsAreSuppressed):
05/23/13: Yesterday
- 23:56 Changeset [150630] by
-
CLoop llint backend should not use the d8 register as scratch register
https://bugs.webkit.org/show_bug.cgi?id=116019
Patch by Peter Wang <peter.wang@torchmobile.com.cn> on 2013-05-23
Reviewed by Csaba Osztrogonác.
- offlineasm/cloop.rb:
- 23:26 Changeset [150629] by
-
[BlackBerry] Need to suspend/resume RootLayerCommit when the application becomes inactive/active
https://bugs.webkit.org/show_bug.cgi?id=115245
Patch by Xiaobo Wang <xiaobwang@blackberry.com> on 2013-05-23
Reviewed by Rob Buis.
PR 330917.
Internally reviewed by Arvid Nilsson.
- Suspend/resumeRootLayerCommit when notified app activation state
change.
- Schedule root layer commit in resumeRootLayerCommit() to explicitly
start root layer commit timer, so that there's a commit even if
BackingStore got disabled/removed.
- Api/WebPage.cpp:
(BlackBerry::WebKit::WebPagePrivate::notifyAppActivationStateChange):
(BlackBerry::WebKit::WebPagePrivate::resumeRootLayerCommit):
- 23:23 Changeset [150628] by
-
[BlackBerry] Need to forward the opener frame url to client when creating a new window
https://bugs.webkit.org/show_bug.cgi?id=116566
Patch by Andy Chen <andchen@blackberry.com> on 2013-05-23
Reviewed by Rob Buis.
PR 337935
Internally reviewed by Arvid Nilsson.
When creating a new window, forward the opener frame url to client.
- Api/WebPageClient.h:
- WebCoreSupport/ChromeClientBlackBerry.cpp:
(WebCore::ChromeClientBlackBerry::createWindow):
- WebCoreSupport/FrameLoaderClientBlackBerry.cpp:
(WebCore::FrameLoaderClientBlackBerry::dispatchCreatePage):
- 23:20 Changeset [150627] by
-
[WinCairo] Build fails to link.
https://bugs.webkit.org/show_bug.cgi?id=116671
Patch by peavo@outlook.com <peavo@outlook.com> on 2013-05-23
Reviewed by Brent Fulgham.
- page/FrameView.h: Added guard for ACCELERATED_COMPOSITING.
- 23:15 Changeset [150626] by
-
Add missing dependencies to deps script.
[GTK] Missing packages in install-dependencies script.
https://bugs.webkit.org/show_bug.cgi?id=116632
Reviewed by Martin Robinson.
- gtk/install-dependencies:
- 23:07 Changeset [150625] by
-
[BlackBerry] Compositor API change: Don't pass the unused "viewport" parameter.
https://bugs.webkit.org/show_bug.cgi?id=116545
Internal PR 189775
Patch by Jakob Petsovits <jpetsovits@blackberry.com> on 2013-05-23
Reviewed by Rob Buis.
Also change the name of the "documentContents" parameter
to the more descriptive "documentSrcRect".
- Api/WebPageCompositor.cpp:
(BlackBerry::WebKit::WebPageCompositorPrivate::render):
(BlackBerry::WebKit::WebPageCompositor::render):
- Api/WebPageCompositor.h:
- Api/WebPageCompositor_p.h:
(WebPageCompositorPrivate):
- 22:14 Changeset [150624] by
-
[WK2][EFL][QT] Build break after r150610
https://bugs.webkit.org/show_bug.cgi?id=116699
Patch by Ryuan Choi <ryuan.choi@samsung.com> on 2013-05-23
Reviewed by Csaba Osztrogonác.
- UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::findPlugin): Initialized pluginProcessSandboxPolicy.
- 19:59 Changeset [150623] by
-
New Tag.
- 18:35 Changeset [150622] by
-
fourthTier: rationalize DFG::CapabilityLevel and DFGCapabilities.[h|cpp]
https://bugs.webkit.org/show_bug.cgi?id=116696
Reviewed by Sam Weinig.
Make it so that all capability calculation is funneled through one function, which tells
you everything you wanted to know: can it be inlined, and can it be compiled.
This work will help with https://bugs.webkit.org/show_bug.cgi?id=116557, since now the
JIT has a fairly authoritative answer to the "can it be inlined" question.
- bytecode/CodeBlock.cpp:
(JSC::CodeBlock::CodeBlock):
(JSC::ProgramCodeBlock::capabilityLevelInternal):
(JSC::EvalCodeBlock::capabilityLevelInternal):
(JSC::FunctionCodeBlock::capabilityLevelInternal):
- bytecode/CodeBlock.h:
(CodeBlock):
(JSC::CodeBlock::capabilityLevel):
(JSC::CodeBlock::capabilityLevelState):
(ProgramCodeBlock):
(EvalCodeBlock):
(FunctionCodeBlock):
- dfg/DFGCapabilities.cpp:
(JSC::DFG::debugFail):
(DFG):
(JSC::DFG::canInlineResolveOperations):
(JSC::DFG::capabilityLevel):
- dfg/DFGCapabilities.h:
(DFG):
(JSC::DFG::capabilityLevel):
(JSC::DFG::evalCapabilityLevel):
(JSC::DFG::programCapabilityLevel):
(JSC::DFG::functionForCallCapabilityLevel):
(JSC::DFG::functionForConstructCapabilityLevel):
(JSC::DFG::canInlineFunctionForCall):
(JSC::DFG::canInlineFunctionForClosureCall):
(JSC::DFG::canInlineFunctionForConstruct):
- dfg/DFGCommon.h:
(JSC::DFG::canCompile):
(DFG):
(JSC::DFG::canInline):
(JSC::DFG::leastUpperBound):
- dfg/DFGDriver.cpp:
(JSC::DFG::compile):
- jit/JIT.cpp:
(JSC::JIT::privateCompile):
- jit/JITPropertyAccess.cpp:
(JSC::JIT::privateCompilePutByIdTransition):
- jit/JITPropertyAccess32_64.cpp:
(JSC::JIT::privateCompilePutByIdTransition):
- tools/CodeProfile.cpp:
(JSC::CodeProfile::sample):
- 18:33 Changeset [150621] by
-
[Windows] Rolling back r150600 as it breaks the VS2010 builds.
- wtf/StackBounds.cpp: Rollback.
- 17:43 Changeset [150620] by
-
Shim Flash plug-in to allow its updater to run
https://bugs.webkit.org/show_bug.cgi?id=116692
<rdar://problem/13938510>
Reviewed by Sam Weinig.
- PluginProcess/PluginProcess.h:
- PluginProcess/mac/PluginProcessMac.mm:
(WebKit::replacedNSConcreteTask_launch):
(WebKit::initializeCocoaOverrides):
Patch -[NSConcreteTask launch] to call PluginProcess::launchProcess.
(WebKit::PluginProcess::launchProcess):
Send a LaunchProcess message to the UI process.
- UIProcess/Plugins/PluginProcessProxy.cpp:
- UIProcess/Plugins/PluginProcessProxy.h:
Remove PluginProcessProxy::didReceiveMessage.
(PluginProcessProxy):
- UIProcess/Plugins/PluginProcessProxy.messages.in:
Add LaunchProcess message and remove LegacyReceiver attribute.
- UIProcess/Plugins/mac/PluginProcessProxyMac.mm:
(WebKit::isFlashUpdater):
Helper function to determine whether the process is the flash updater.
(WebKit::shouldLaunchProcess):
Return whether we should launch this process.
(WebKit::PluginProcessProxy::launchProcess):
Determine if we should launch this process and then do so.
- 17:41 Changeset [150619] by
-
[Windows] Unreviewed change for NRWT support.
- Scripts/webkitpy/port/win.py:
(WinPort.default_child_processes): Added a default implementation
that limits NRWT to a single instance.
- 17:03 Changeset [150618] by
-
Unreviewed. Modify WebCore for AppleWin 2010 production build.
- WebCore.vcxproj/WebCoreGenerated.make:
(WebCore::AXObjectCache::frameLoadingEventPlatformNotification):
- platform/graphics/avfoundation/cf/MediaPlayerPrivateAVFoundationCF.cpp:
- 16:44 Changeset [150617] by
-
[EFL] Rebaseline after r150593
https://bugs.webkit.org/show_bug.cgi?id=116695
Unreviewed gardening.
Moved some editing tests from execCommand into pasteboard.
Patch by Seokju Kwon <seokju.kwon@gmail.com> on 2013-05-23
- platform/efl/TestExpectations:
- 16:22 Changeset [150616] by
-
Make Math.imul tests run quicker, and fix some other bugs.
Rubber stampted by Oliver Hunt.
Also make tests not call print().
- fast/js/regress/script-tests/cell-argument.js:
- fast/js/regress/script-tests/imul-double-only.js:
(f):
- fast/js/regress/script-tests/imul-int-only.js:
(f):
- fast/js/regress/script-tests/imul-mixed.js:
(f):
- fast/js/regress/script-tests/object-closure-call.js:
- 16:00 Changeset [150615] by
-
Remove meaningless "test_uses_apache" from Python test cases.
https://bugs.webkit.org/show_bug.cgi?id=38756
Patch by Brent Fulgham <bfulgham@apple.com> on 2013-05-23
Reviewed by Ryosuke Niwa.
- Scripts/webkitpy/port/port_testcase.py:
(PortTestCase.test_driver_cmd_line): Remove the meaningless
"test_uses_apache" test case from the Python test infrastructure.
It was hard-coded to true, even though not all ports use
Apache.
- 14:44 Changeset [150614] by
-
Merged r150610. <rdar://problem/13975569>
- 14:17 Changeset [150613] by
-
We need to clear main resource when detaching DocumentLoader from the frame.
https://bugs.webkit.org/show_bug.cgi?id=116680
Normally, when we detach the documentLoader in DocumentLoader::detachFromFrame, main resource is also cleared
in stopLoading(). There is possibility that main resource not being cleared, and this could cause crash later
since docLoader could still receive resource callbacks. To avoid that, we can remove docLoader from main resource's
client set in detachFromFrame.
Patch by Yongjun Zhang <yongjun_zhang@apple.com> on 2013-05-23
Reviewed by Brady Eidson.
No new tests, this doesn't happen on OS X.
- loader/DocumentLoader.cpp:
(WebCore::DocumentLoader::detachFromFrame):
(WebCore::DocumentLoader::clearMainResource):
- loader/cache/CachedResource.h: make hasClient accessible publicly.
(WebCore::CachedResource::hasClient):
(CachedResource):
- 13:59 Changeset [150612] by
-
[Windows] Update new-run-webkit-tests for Windows port.
https://bugs.webkit.org/show_bug.cgi?id=38756
Patch by Brent Fulgham <bfulgham@webkit.org> on 2013-05-23
Reviewed by Ryosuke Niwa.
- Scripts/new-run-webkit-httpd:
(run): Switch from 'chromium' to 'win'
- Scripts/webkitpy/port/win.py:
(WinPort): Add support methods for lighttpd server.
(WinPort._uses_apache):
(WinPort._path_to_lighttpd):
(WinPort._path_to_lighttpd_modules):
(WinPort._path_to_lighttpd_php):
- 13:56 Changeset [150611] by
-
assertion failed: !node node->isElementNode() in WebCore::RenderBlock::clone https://bugs.webkit.org/show_bug.cgi?id=116614
Patch by Mihai Maerean <mmaerean@adobe.com> on 2013-05-23
Reviewed by David Hyatt.
Source/WebCore:
Add a test for flow threads so they don't end up being cloned in region based multi columns with columns:1 where
the only child has column-span:all.
Test: fast/multicol/newmulticol/direct-child-column-span-all.html
- rendering/RenderBlock.cpp:
(WebCore::RenderBlock::addChildIgnoringAnonymousColumnBlocks):
LayoutTests:
- fast/multicol/newmulticol/direct-child-column-span-all-expected.txt: Added.
- fast/multicol/newmulticol/direct-child-column-span-all.html: Added.
- 13:48 Changeset [150610] by
-
Don't run all plug-ins unsandboxed
https://bugs.webkit.org/show_bug.cgi?id=116690
<rdar://problem/13975569>
Reviewed by Jessie Berlin.
Remove debugging code.
- UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::findPlugin):
- 13:08 Changeset [150609] by
-
Crash in convertMainResourceLoadToDownload when downloading file by option-return
https://bugs.webkit.org/show_bug.cgi?id=116451
Reviewed by Andreas Kling.
Source/WebCore:
Handle MainResourceLoader being null here which will happen when loading a cached main resource.
- loader/DocumentLoader.cpp:
(WebCore::DocumentLoader::continueAfterContentPolicy):
Source/WebKit/mac:
If the main resource loader is null, create a new download instead since there's no existing loader to convert.
- WebCoreSupport/WebFrameLoaderClient.mm:
(WebFrameLoaderClient::convertMainResourceLoadToDownload):
Source/WebKit2:
If the main resource loader is null, create a new download instead since there's no existing loader to convert.
- NetworkProcess/NetworkConnectionToWebProcess.cpp:
(WebKit::NetworkConnectionToWebProcess::convertMainResourceLoadToDownload):
- WebProcess/WebPage/WebFrame.cpp:
(WebKit::WebFrame::convertMainResourceLoadToDownload):
- 12:56 Changeset [150608] by
-
Remove a stale test expectation.
- platform/wk2/TestExpectations:
- 12:49 Changeset [150607] by
-
Update Mac test expectations for the bug 85408 as bugs 116398 and 116477 have been merged into it.
- platform/mac/TestExpectations:
- 12:46 Changeset [150606] by
-
KURL::createCFURL() should return a RetainPtr<CFURLRef>
https://bugs.webkit.org/show_bug.cgi?id=116644
Patch by Benjamin Poulain <bpoulain@apple.com> on 2013-05-23
Reviewed by Andreas Kling.
For consistency with the other createCFType() APIs in WebKit, KURL should
return a RetainPtr<> instead of raw +1 CFURLRef.
No leak were discovered while making this change.
- platform/KURL.h:
- platform/cf/KURLCFNet.cpp:
(WebCore):
(WebCore::createCFURLFromBuffer):
(WebCore::KURL::createCFURL):
(WebCore::KURL::fileSystemPath):
- platform/graphics/avfoundation/cf/MediaPlayerPrivateAVFoundationCF.cpp:
(WebCore::AVFWrapper::createAssetForURL):
- platform/graphics/cg/GraphicsContextCG.cpp:
(WebCore::GraphicsContext::setURLForRect):
- platform/mac/KURLMac.mm:
(WebCore):
(WebCore::KURL::operator NSURL *):
(WebCore::KURL::createCFURL):
- platform/network/cf/CookieJarCFNet.cpp:
(WebCore::setCookiesFromDOM):
(WebCore::cookiesForDOM):
(WebCore::cookieRequestHeaderFieldValue):
(WebCore::getRawCookies):
(WebCore::deleteCookie):
- platform/network/cf/DNSCFNet.cpp:
(WebCore::DNSResolveQueue::platformProxyIsEnabledInSystemPreferences):
- platform/network/cf/ProxyServerCFNet.cpp:
(WebCore::addProxyServersForURL):
- platform/network/cf/ResourceRequestCFNet.cpp:
(WebCore::ResourceRequest::doUpdatePlatformRequest):
(WebCore::ResourceRequest::doUpdatePlatformHTTPBody):
- platform/network/cf/ResourceResponseCFNet.cpp:
(WebCore::ResourceResponse::cfURLResponse):
- platform/network/cf/SocketStreamHandleCFNet.cpp:
(WebCore::SocketStreamHandle::SocketStreamHandle):
- 12:42 Changeset [150605] by
-
Add back failing test expectations for the bugs 88672 and 89528 after r150558 and r150559 on Lion.
- platform/mac/TestExpectations:
- 12:37 Changeset [150604] by
-
fast/workers/storage/use-same-database-in-page-and-workers.html times out on the lion bots
https://bugs.webkit.org/show_bug.cgi?id=81618
Add back the flaky test failure on Mac.
- platform/mac/TestExpectations:
- 12:19 Changeset [150603] by
-
Use correct stack size on Solaris and OpenBSD
https://bugs.webkit.org/show_bug.cgi?id=114978
Reviewed by Oliver Hunt.
Original patch by David Hill <david@wmol.com>.
Use stack_t.ss_size for getting the size of the stack.
- wtf/Platform.h:
- wtf/StackBounds.cpp:
(WTF):
(WTF::StackBounds::initialize):
- 11:55 Changeset [150602] by
-
The ellipsis in a text overflow should not avoid floats
https://bugs.webkit.org/show_bug.cgi?id=115746
Reviewed by David Hyatt.
Source/WebCore:
When placing an ellipsis for text overflow we shouldn't avoid any floats on the line.
That is to say, if there is a float in the box we overflow we should let our no-wrap text,
including the ellipsis, render inside the float just as we would if overflow was hidden but
a text-overflow rendering had not been specified.
Test: fast/css/text-overflow-ellipsis-behind-floats.html
- rendering/RenderBlockLineLayout.cpp:
(WebCore::RenderBlock::checkLinesForTextOverflow):
LayoutTests:
- fast/css/text-overflow-ellipsis-behind-floats-expected.html: Added.
- fast/css/text-overflow-ellipsis-behind-floats.html: Added.
- fast/css/text-overflow-ellipsis-full-truncate-rtl-expected.html:
- fast/css/text-overflow-ellipsis-full-truncate-rtl.html: Change this test to expect the text to cross into the floating input element and render the ellipsis at the left extreme of the element. The previous expected rendering relied on the incorrect premise that an overflow ellipsis should avoid floated elements.
- 11:51 Changeset [150601] by
-
Merged r150556. <rdar://problem/13918391>
- 11:51 Changeset [150600] by
-
[WIN] Implement correct detection of stack size
https://bugs.webkit.org/show_bug.cgi?id=116661
Reviewed by Oliver Hunt.
- wtf/StackBounds.cpp:
(WTF):
(WTF::StackBounds::initialize):
- 11:49 Changeset [150599] by
-
<rdar://problem/13849260> Sandbox violations reading .CFUserTextEncoding
Reviewed by Dan Bernstein.
- WebProcess/com.apple.WebProcess.sb.in: Allow this.
- 11:48 Changeset [150598] by
-
Merged r150543. <rdar://problem/13904054>
- 11:44 Changeset [150597] by
-
Merged r150537. <rdar://problem/13964515>
- 11:30 Changeset [150596] by
-
[Lion] [WK2] compositing/overflow/overflow-compositing-descendant.html
https://bugs.webkit.org/show_bug.cgi?id=116684
Add a timeout/failure test expectation. Also update the expected result since the first frame
of the video must be shown in this test.
- platform/mac-wk2/TestExpectations:
- platform/mac/compositing/overflow/overflow-compositing-descendant-expected.png:
- 11:03 Changeset [150595] by
-
Move posix_spawn onto a zero delay timer
https://bugs.webkit.org/show_bug.cgi?id=116682
Reviewed by Anders Carlson.
<rdar://problem/13973468>
- Shared/EntryPointUtilities/mac/XPCService/XPCServiceMain.Development.mm:
(ReexecInfo):
- struct to hold info from the XPC message.
(WebKit::reexec):
- moved call to posix_spawn out to this helper.
(WebKit::reexecCallBack):
- calls reexec.
(WebKit::XPCServiceEventHandler):
- removed call to posix_spawn; schedule a call to reexecCallBack.
- 10:52 Changeset [150594] by
-
Fix loader/go-back-cached-main-resource.html test
https://bugs.webkit.org/show_bug.cgi?id=116670
Patch by Manuel Rego Casasnovas <rego@igalia.com> on 2013-05-23
Reviewed by Alexey Proskuryakov.
The page does not get an onload when being restored from page cache. The
test case case is modified to handle pageshow event instead.
- loader/go-back-cached-main-resource.html: Wait for the third
'first-page' message to finish the test.
- loader/resources/first-page.html: Use onpageshow instead of onload.
- loader/resources/other-page.html: Ditto.
- 10:50 Changeset [150593] by
-
Several editing tests that copy & paste slow down Mac WK2 EWS
https://bugs.webkit.org/show_bug.cgi?id=116681
Reviewed by Anders Carlsson.
Move editing tests that trigger copy and paste from execCommand into pasteboard.
Since NWRT uses a single instance of WebKitTestRunner to run tests in single directory,
this should mitigate the issue for now.
In the long term, we should really mock out the clipboard in WebKitTestRunner.
- editing/execCommand/4128080-2-expected.txt: Removed.
- editing/execCommand/4128080-2.html: Removed.
- editing/execCommand/5939887-expected.txt: Removed.
- editing/execCommand/5939887.html: Removed.
- editing/execCommand/copy-without-selection-expected.txt: Removed.
- editing/execCommand/copy-without-selection.html: Removed.
- editing/execCommand/paste-2.html: Removed.
- editing/pasteboard/copy-cut-paste-events-fired-when-disabled-expected.txt: Copied from LayoutTests/editing/execCommand/5939887-expected.txt.
- editing/pasteboard/copy-cut-paste-events-fired-when-disabled.html: Copied from LayoutTests/editing/execCommand/5939887.html.
- editing/pasteboard/copy-inside-h1-preserves-h1-expected.txt: Copied from LayoutTests/editing/execCommand/4128080-2-expected.txt.
- editing/pasteboard/copy-inside-h1-preserves-h1.html: Copied from LayoutTests/editing/execCommand/4128080-2.html.
- editing/pasteboard/copy-without-selection-expected.txt: Copied from LayoutTests/editing/execCommand/copy-without-selection-expected.txt.
- editing/pasteboard/copy-without-selection.html: Copied from LayoutTests/editing/execCommand/copy-without-selection.html.
- editing/pasteboard/paste-2.html: Copied from LayoutTests/editing/execCommand/paste-2.html.
- platform/gtk/editing/execCommand/paste-2-expected.png: Removed.
- platform/gtk/editing/execCommand/paste-2-expected.txt: Removed.
- platform/gtk/editing/pasteboard/paste-2-expected.png: Copied from LayoutTests/platform/gtk/editing/execCommand/paste-2-expected.png.
- platform/gtk/editing/pasteboard/paste-2-expected.txt: Copied from LayoutTests/platform/gtk/editing/execCommand/paste-2-expected.txt.
- platform/mac/editing/execCommand/paste-2-expected.png: Removed.
- platform/mac/editing/execCommand/paste-2-expected.txt: Removed.
- platform/mac/editing/pasteboard/paste-2-expected.png: Copied from LayoutTests/platform/mac/editing/execCommand/paste-2-expected.png.
- platform/mac/editing/pasteboard/paste-2-expected.txt: Copied from LayoutTests/platform/mac/editing/execCommand/paste-2-expected.txt.
- platform/qt/editing/execCommand/paste-2-expected.png: Removed.
- platform/qt/editing/execCommand/paste-2-expected.txt: Removed.
- platform/qt/editing/pasteboard/paste-2-expected.png: Copied from LayoutTests/platform/qt/editing/execCommand/paste-2-expected.png.
- platform/qt/editing/pasteboard/paste-2-expected.txt: Copied from LayoutTests/platform/qt/editing/execCommand/paste-2-expected.txt.
- 10:49 Changeset [150592] by
-
Make default bug links in buildbot clickable
https://bugs.webkit.org/show_bug.cgi?id=116673
Reviewed by Ryosuke Niwa.
Use changecommentlink of html.WebStatus to make the links clickable.
- BuildSlaveSupport/build.webkit.org-config/master.cfg:
- 10:39 WebKit Team edited by
- (diff)
- 10:17 Changeset [150591] by
-
EWS bubbles need more space
https://bugs.webkit.org/show_bug.cgi?id=116675
Reviewed by Andreas Kling.
- template/en/custom/attachment/list.html.tmpl:
- 09:45 Changeset [150590] by
-
Unreviewed, rolling out r150586.
http://trac.webkit.org/changeset/150586
https://bugs.webkit.org/show_bug.cgi?id=116660
Those constructors should be exposed on the WorkerContext, not
the DOMWindow
Source/WebCore:
- fileapi/FileReaderSync.idl:
- workers/WorkerLocation.idl:
LayoutTests:
- fast/js/global-constructors-attributes-expected.txt:
- platform/efl/fast/js/constructor-length-expected.txt:
- platform/efl/fast/js/global-constructors-attributes-expected.txt:
- platform/gtk/fast/js/constructor-length-expected.txt:
- platform/gtk/fast/js/global-constructors-attributes-expected.txt:
- platform/mac-lion/fast/js/global-constructors-attributes-expected.txt:
- platform/mac/fast/js/constructor-length-expected.txt:
- platform/qt/fast/js/constructor-length-expected.txt:
- platform/qt/fast/js/global-constructors-attributes-expected.txt:
- 09:13 Changeset [150589] by
-
[CSS Regions][Mac] [WK2] Layout Test fast/regions/fixed-pos-elem-in-region.html is flaky on Debug builders
https://bugs.webkit.org/show_bug.cgi?id=114571
Tests seems to be passing now, removed from TestExpectations.
Patch by Radu Stavila <stavila@adobe.com> on 2013-05-23
Reviewed by Anders Carlsson.
- platform/mac-wk2/TestExpectations:
- 09:11 Changeset [150588] by
-
[BlackBerry] The web context does not contain any link properties if the context node is an image enclosed by a link node
https://bugs.webkit.org/show_bug.cgi?id=116627
Patch by Ed Baker <edbaker@blackberry.com> on 2013-05-23
Reviewed by Xan Lopez.
PR #341084
If the context node has an image tag and an enclosing link node was detected then add the link properties to the context.
Internally Reviewed by Gen Mak.
- Api/WebPage.cpp:
(BlackBerry::WebKit::WebPagePrivate::webContext):
- 08:55 Changeset [150587] by
-
[BlackBerry] Validate form data before doing direct submission.
https://bugs.webkit.org/show_bug.cgi?id=116674
Reviewed by Xan Lopez.
PR 314202.
Check validation of the input form before triggering direct submission.
Internally Reviewed by Nima Ghanavatian.
- WebKitSupport/InputHandler.cpp:
(BlackBerry::WebKit::InputHandler::submitForm):
- 08:20 Changeset [150586] by
-
Remove [NoInterfaceObject] from FileReaderSync and WorkerLocation
https://bugs.webkit.org/show_bug.cgi?id=116660
Reviewed by Kentaro Hara.
Source/WebCore:
Remove [NoInterfaceObject] from FileReaderSync and WorkerLocation to match their
respective specifications:
- http://dev.w3.org/2006/webapi/FileAPI/#FileReaderSync
- http://dev.w3.org/html5/workers/#workerlocation
No new tests, covered by existing tests.
- fileapi/FileReaderSync.idl:
- workers/WorkerLocation.idl:
LayoutTests:
Rebaseline a couple of test cases now that FileReaderSync and WorkerLocation
attributes are exposed on the global Window object.
- fast/js/global-constructors-attributes-expected.txt:
- platform/efl/fast/js/constructor-length-expected.txt:
- platform/efl/fast/js/global-constructors-attributes-expected.txt:
- platform/gtk/fast/js/constructor-length-expected.txt:
- platform/gtk/fast/js/global-constructors-attributes-expected.txt:
- platform/mac-lion/fast/js/global-constructors-attributes-expected.txt:
- platform/mac/fast/js/constructor-length-expected.txt:
- platform/qt/fast/js/constructor-length-expected.txt:
- platform/qt/fast/js/global-constructors-attributes-expected.txt:
- 07:14 Changeset [150585] by
-
[BlackBerry] Properly fill the ResourceError in FrameLoaderClientBlackBerry::cannotShowURLError
https://bugs.webkit.org/show_bug.cgi?id=116603
Patch by Rob Buis <rbuis@rim.com> on 2013-05-23
Reviewed by Xan Lopez.
PR 119789
Internally reviewed by Yong Li.
Provide a domain value for this ResourceError instance. Note that
this does not change behavior.
- WebCoreSupport/FrameLoaderClientBlackBerry.cpp:
(WebCore::FrameLoaderClientBlackBerry::didRestoreFromPageCache):
- 07:00 Changeset [150584] by
-
[Qt] Unreviewed gardening.
- platform/qt-5.0-wk2/TestExpectations: Skip some failing test.
- platform/qt-5.0-wk2/fast/forms/select-baseline-expected.png: Rebaselined after r148594.
- platform/qt-5.0-wk2/fast/forms/select-baseline-expected.txt:
- platform/qt-5.0-wk2/fast/images/repaint-subrect-grid-expected.txt: Rebaselined after r149292.
- 06:42 TestExpectations edited by
- Link to "Writing Reftests" to explain more on reference tests (diff)
- 06:17 Changeset [150583] by
-
[cmake] Fix generating-forwarding-headers rule after r150460
https://bugs.webkit.org/show_bug.cgi?id=116659
Reviewed by Christophe Dumez.
- CMakeLists.txt: Typo fix.
- 05:45 Changeset [150582] by
-
[GTK] Remote Web Inspector sometimes won't serve any content
https://bugs.webkit.org/show_bug.cgi?id=116225
Patch by Anton Obzhirov <a.obzhirov@samsung.com> on 2013-05-23
Reviewed by Gustavo Noronha Silva.
Added an option to use environment variable to set up path to
inspector server files. Implementation is similar to InspectorClientGtk
implementation.
- UIProcess/InspectorServer/WebInspectorServer.h:
(WebInspectorServer):
- UIProcess/InspectorServer/gtk/WebInspectorServerGtk.cpp:
(WebKit::WebInspectorServer::platformResourceForPath):
(WebKit::WebInspectorServer::inspectorServerFilesPath):
(WebKit):
- 05:30 Changeset [150581] by
-
[EFL][WK2] Use WK2 C API in ewk_view_pagination_mode_set|get
https://bugs.webkit.org/show_bug.cgi?id=116657
Reviewed by Christophe Dumez.
WKPageGet|SetPaginationMode APIs is present in WK2 C APIs. So, we need to use them for
ewk_view_pagination_mode_set|get.
Besides we don't need to use WebCore namespace anymore.
- UIProcess/API/efl/ewk_view.cpp:
(ewk_view_pagination_mode_set):
(ewk_view_pagination_mode_get):
- 05:22 Changeset [150580] by
-
Remove special case for MessagePortArray from bindings generator
https://bugs.webkit.org/show_bug.cgi?id=116653
Reviewed by Kentaro Hara.
Use MessagePort[] instead of MessagePortArray type in the IDL as MessagePortArray
is not defined on IDL side. This allows us to remove the special case for
MessagePortArray from the bindings generator.
No new tests, no behavior change.
- bindings/js/JSMessageEventCustom.cpp:
(WebCore::JSMessageEvent::data):
- bindings/scripts/CodeGeneratorJS.pm:
(NativeToJSValue): Remove special case for MessagePortArray.
- bindings/scripts/test/CPP/WebDOMTestSerializedScriptValueInterface.cpp:
- bindings/scripts/test/CPP/WebDOMTestSerializedScriptValueInterface.h:
(WebDOMTestSerializedScriptValueInterface):
- bindings/scripts/test/GObject/WebKitDOMTestSerializedScriptValueInterface.cpp:
(webkit_dom_test_serialized_script_value_interface_get_property):
(webkit_dom_test_serialized_script_value_interface_class_init):
- bindings/scripts/test/GObject/WebKitDOMTestSerializedScriptValueInterface.h:
- bindings/scripts/test/JS/JSTestSerializedScriptValueInterface.cpp:
(WebCore::jsTestSerializedScriptValueInterfacePorts):
- bindings/scripts/test/ObjC/DOMTestSerializedScriptValueInterface.h:
- bindings/scripts/test/ObjC/DOMTestSerializedScriptValueInterface.mm:
- bindings/scripts/test/TestSerializedScriptValueInterface.idl: Replace MessagePortArray
by MessagePort[].
- dom/MessageEvent.h:
(WebCore::MessageEvent::ports): The method should return a Vector, not a Vector*
as the array on JavaScript side is not nullable.
- dom/MessageEvent.idl: Replace MessagePortArray by MessagePort[].
- 03:32 Changeset [150579] by
-
media/W3C/video/networkState/networkState_during_progress.html is flaky
https://bugs.webkit.org/show_bug.cgi?id=76280
Unreviewed gardening.
Change expectation to the correct output and mark as flaky.
- TestExpectations:
- media/W3C/video/networkState/networkState_during_progress-expected.txt:
- 02:26 Changeset [150578] by
-
Support frame instrumentation for coordinated graphics
https://bugs.webkit.org/show_bug.cgi?id=115260
Patch by Roland Takacs <rtakacs@inf.u-szeged.hu> on 2013-05-23
Reviewed by Andreas Kling.
Enabled supportsFrameInstrumentation on ports that use coordinated graphics
by calling didBeginFrame().
- WebProcess/WebCoreSupport/WebInspectorClient.cpp:
(WebKit::WebInspectorClient::supportsFrameInstrumentation):
(WebKit):
- WebProcess/WebCoreSupport/WebInspectorClient.h:
(WebInspectorClient):
- WebProcess/WebPage/CoordinatedGraphics/CoordinatedLayerTreeHost.cpp:
(WebKit::CoordinatedLayerTreeHost::syncDisplayState):
- 02:20 Changeset [150577] by
-
[Qt] Unreviewed gardening.
- platform/qt-5.0-wk1/TestExpectations: Remove a test failure that skipped in qt.
- platform/qt/TestExpectations: Unskip now passing test.
- 01:39 Changeset [150576] by
-
[Qt] Unreviewed gardening.
- platform/qt-5.0-wk2/TestExpectations: Unskip after rebaseline.
- platform/qt/TestExpectations: Skip flaky test.
- platform/qt/plugins/mouse-events-expected.txt: Rebaselining after r150558.
- 00:06 Changeset [150575] by
-
[EFL] Back, Forward and Stop should be hidden in context menu
https://bugs.webkit.org/show_bug.cgi?id=116615
Patch by Michał Pakuła vel Rutka <m.pakula@samsung.com> on 2013-05-23
Reviewed by Christophe Dumez.
Context menu items: Back, Forward, Stop, Reload, should be shown only when
they are able to perform action.
No new tests - no new functionality.
- page/ContextMenuController.cpp:
(WebCore::ContextMenuController::populate):
05/22/13:
- 23:32 Changeset [150574] by
-
Updated the bug associated with the failure of media/context-menu-actions.html
from the bug 45021 to the bug 116651 since the former is the bug for which this test was added.
Also skip this test it has been timing out intermittently.
- platform/mac/TestExpectations:
- 23:22 Changeset [150573] by
-
Update Mac test expectation for the bug 82980.
The test is consistently timing out instead of being flaky now.
- platform/mac/TestExpectations:
- 23:20 Changeset [150572] by
-
Add back two Lion test expectations after r150558 and r150559.
- platform/mac/TestExpectations:
- 23:03 Changeset [150571] by
-
Make the test expectation for the bug 116649 specific to WK2
as the test is passing on WK1.
- platform/mac-wk2/TestExpectations:
- platform/mac/TestExpectations:
- 23:01 Changeset [150570] by
-
Mark media/audio-repaint.html as always failing on Mac.
Also add back a failing test expectation to compositing/rtl/rtl-fixed-overflow-scrolled.html on Mac WK2.
Also add failing test expectation for the bug 116649.
- platform/mac/TestExpectations:
- 22:18 Changeset [150569] by
-
Use uninitialized register in "JIT::emit_op_neq_null" and "emit_op_eq_null"
https://bugs.webkit.org/show_bug.cgi?id=116593
Patch by Peter Wang <peter.wang@torchmobile.com.cn> on 2013-05-22
Reviewed by Filip Pizlo.
Generated instructions using uninitialized register. It's caused by a mistake of r126494.
Source/JavaScriptCore:
- jit/JITOpcodes32_64.cpp:
(JSC::JIT::emit_op_eq_null):
(JSC::JIT::emit_op_neq_null):
LayoutTests:
- fast/js/neq-null-crash-expected.txt: Added.
- fast/js/neq-null-crash.html: Added.
- fast/js/script-tests/neq-null-crash.js: Added.
- 22:07 Changeset [150568] by
-
Merge http://trac.webkit.org/changeset/149162, since we need it after
we merged some bytecode refactorings.
2013-04-25 Filip Pizlo <fpizlo@apple.com>
DFG doesn't support to_jsnumber
https://bugs.webkit.org/show_bug.cgi?id=115129
Reviewed by Geoffrey Garen.
Based on Oliver's patch. Implements to_jsnumber as Identity(Number:@thingy), and then does
an optimization in Fixup to turn Identity(Number:) into Identity(Int32:) if the predictions
tell us to. Identity is later turned into Phantom.
Also fixed BackPropMask, which appeared to have NodeDoesNotExit included in it. That's
wrong; NodeDoesNotExit is not a backward propagation property.
Also fixed Identity to be marked as CanExit (i.e. not NodeDoesNotExit).
This more than doubles the FPS on ammo.
- dfg/DFGByteCodeParser.cpp: (JSC::DFG::ByteCodeParser::parseBlock):
- dfg/DFGCapabilities.h: (JSC::DFG::canCompileOpcode):
- dfg/DFGFixupPhase.cpp: (JSC::DFG::FixupPhase::fixupNode): (FixupPhase): (JSC::DFG::FixupPhase::observeUseKindOnNode): (JSC::DFG::FixupPhase::observeUseKindOnEdge):
- dfg/DFGNodeFlags.h: (DFG):
- dfg/DFGNodeType.h: (DFG):
- dfg/DFGPredictionPropagationPhase.cpp: (JSC::DFG::PredictionPropagationPhase::propagate):
- 22:00 Changeset [150567] by
-
Fix indentation of CodeBlock.h
Rubber stamped by Mark Hahnenberg.
- bytecode/CodeBlock.h:
- 22:00 Changeset [150566] by
-
Fix indentation of CodeBlock.h
Rubber stampted by Mark Hahnenberg.
- bytecode/CodeBlock.h:
- 21:53 Changeset [150565] by
-
Rename getJITCode and getJITType to jitCode and jitType.
Rubber stampted by Mark Hahnenberg.
- assembler/RepatchBuffer.h:
(JSC::RepatchBuffer::RepatchBuffer):
- bytecode/CodeBlock.cpp:
(JSC::CodeBlock::dump):
(JSC::CodeBlock::visitAggregate):
(JSC::CodeBlock::finalizeUnconditionally):
(JSC::CodeBlock::resetStubInternal):
(JSC::CodeBlock::stronglyVisitWeakReferences):
(JSC::CodeBlock::baselineVersion):
(JSC::CodeBlock::hasOptimizedReplacement):
(JSC::CodeBlock::bytecodeOffset):
(JSC::CodeBlock::codeOriginForReturn):
(JSC::ProgramCodeBlock::compileOptimized):
(JSC::EvalCodeBlock::compileOptimized):
(JSC::FunctionCodeBlock::compileOptimized):
(JSC::ProgramCodeBlock::jettison):
(JSC::EvalCodeBlock::jettison):
(JSC::FunctionCodeBlock::jettison):
(JSC::ProgramCodeBlock::jitCompileImpl):
(JSC::EvalCodeBlock::jitCompileImpl):
(JSC::FunctionCodeBlock::jitCompileImpl):
(JSC::CodeBlock::setOptimizationThresholdBasedOnCompilationResult):
(JSC::CodeBlock::adjustedExitCountThreshold):
(JSC::CodeBlock::tallyFrequentExitSites):
- bytecode/CodeBlock.h:
(JSC::CodeBlock::getCallLinkInfo):
(JSC::CodeBlock::jitCode):
(JSC::CodeBlock::jitCodeWithArityCheck):
(JSC::CodeBlock::jitType):
(JSC::CodeBlock::hasBaselineJITProfiling):
(JSC::CodeBlock::jitCompile):
(JSC::CodeBlock::addFrequentExitSite):
(JSC::CodeBlock::shouldImmediatelyAssumeLivenessDuringScan):
(JSC::ExecState::isInlineCallFrame):
- dfg/DFGAssemblyHelpers.cpp:
(JSC::DFG::AssemblyHelpers::decodedCodeMapFor):
- dfg/DFGAssemblyHelpers.h:
(JSC::DFG::AssemblyHelpers::AssemblyHelpers):
- dfg/DFGDriver.cpp:
(JSC::DFG::compile):
- dfg/DFGOSREntry.cpp:
(JSC::DFG::prepareOSREntry):
- dfg/DFGOSRExit.cpp:
(JSC::DFG::OSRExit::codeLocationForRepatch):
- dfg/DFGOSRExitCompiler.cpp:
- dfg/DFGOSRExitCompilerCommon.cpp:
(JSC::DFG::reifyInlinedCallFrames):
(JSC::DFG::adjustAndJumpToTarget):
- dfg/DFGOperations.cpp:
- dfg/DFGVariableEventStream.cpp:
(JSC::DFG::VariableEventStream::reconstruct):
- ftl/FTLOSRExit.cpp:
(JSC::FTL::OSRExit::codeLocationForRepatch):
- ftl/FTLOSRExitCompiler.cpp:
(JSC::FTL::compileFTLOSRExit):
- heap/DFGCodeBlocks.cpp:
(JSC::DFGCodeBlocks::~DFGCodeBlocks):
(JSC::DFGCodeBlocks::jettison):
(JSC::DFGCodeBlocks::clearMarks):
(JSC::DFGCodeBlocks::deleteUnmarkedJettisonedCodeBlocks):
(JSC::DFGCodeBlocks::traceMarkedCodeBlocks):
- interpreter/Interpreter.cpp:
(JSC::getLineNumberForCallFrame):
(JSC::getCallerInfo):
- jit/JITDriver.h:
(JSC::jitCompileIfAppropriateImpl):
(JSC::jitCompileFunctionIfAppropriateImpl):
- jit/JITStubs.cpp:
(JSC::DEFINE_STUB_FUNCTION):
- llint/LLIntSlowPaths.cpp:
(JSC::LLInt::entryOSR):
(JSC::LLInt::LLINT_SLOW_PATH_DECL):
- runtime/Executable.cpp:
(JSC::jettisonCodeBlock):
(JSC::EvalExecutable::compileOptimized):
(JSC::EvalExecutable::jettisonOptimizedCode):
(JSC::ProgramExecutable::compileOptimized):
(JSC::ProgramExecutable::jettisonOptimizedCode):
(JSC::FunctionExecutable::baselineCodeBlockFor):
(JSC::FunctionExecutable::compileOptimizedForCall):
(JSC::FunctionExecutable::compileOptimizedForConstruct):
(JSC::FunctionExecutable::jettisonOptimizedCodeForCall):
(JSC::FunctionExecutable::jettisonOptimizedCodeForConstruct):
- tools/CodeProfile.cpp:
(JSC::CodeProfile::sample):
- 21:26 Changeset [150564] by
-
REGRESSION(r150491): WebKit2.DOMWindowExtensionNoCache fails
https://bugs.webkit.org/show_bug.cgi?id=116595
Temporarily skip this test.
- TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionNoCache.cpp:
(TestWebKitAPI::TEST):
- 21:15 Changeset [150563] by
-
fourthTier: Race between LLInt->Baseline tier-up and DFG reading Baseline profiling data
https://bugs.webkit.org/show_bug.cgi?id=116633
Reviewed by Mark Hahnenberg.
Previously we would check if we had Baseline JIT profiling data by seeing if the
appropriate vector was non-empty. This is horrible if we're doing LLInt->Baseline
tier-up at the same time. This will happen for code we are inlining, if we're
deciding to inline it before the LLInt->Baseline tier-up happened for that code.
This changes things to take advantage of the fact that the very last thing that
LLInt->Baseline tier-up will do (in JITDriver.h) is setJITCode(). We now precede
the actual work in setJITCode() with a store-store fence to ensure that all
stores to modify the CodeBlock happen before setting the JITCode, and we modify
CodeBlock::getJITType() to use load-load fences to ensure that if you see
JITCode::BaselineJIT then you will also see all of those vectors. Then this
changes all of the code that scrapes Baseline JIT profiles to check if
getJITType() returns JITCode::BaselineJIT instead of checking vector sizes.
The outcome is that for the non-racy cases we behave as we did before (we fall
back on LLInt profiling if the tier-up hasn't happened) and for racy cases we
use LLInt profiling conservatively.
Note that for some (but not all!) of the cases where we scrape Baseline JIT
profiling, we would have anyway been holding the CodeBlock::m_lock so we can also
fix those cases by just having setJITCode graph that lock. This patch does that
also, mainly because although we only call setJITCode() from the main thread, in
general it's dangerous to have a pointer to a ref-counted object being modified
in a racy way. So, this patch just does the most conservative thing possible that
does we can afford to do.
- bytecode/CallLinkStatus.cpp:
(JSC::CallLinkStatus::computeFor):
- bytecode/CodeBlock.h:
(JSC::CodeBlock::setJITCode):
(JSC::CodeBlock::getJITType):
(JSC::CodeBlock::hasBaselineJITProfiling):
(JSC::CodeBlock::likelyToTakeSlowCase):
(JSC::CodeBlock::couldTakeSlowCase):
(JSC::CodeBlock::likelyToTakeSpecialFastCase):
(JSC::CodeBlock::couldTakeSpecialFastCase):
(JSC::CodeBlock::likelyToTakeDeepestSlowCase):
(JSC::CodeBlock::likelyToTakeAnySlowCase):
- bytecode/GetByIdStatus.cpp:
(JSC::GetByIdStatus::computeFor):
- bytecode/PutByIdStatus.cpp:
(JSC::PutByIdStatus::computeFor):
- 20:41 Changeset [150562] by
-
Unreviewed EFL gardening.
Skip 3 more tests which has problem related to ecore library.
- platform/efl/TestExpectations:
- 20:36 Changeset [150561] by
-
Rolling out r150555. It borked about 30 media tests.
Source/WebCore:
- WebCore.xcodeproj/project.pbxproj:
- html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::HTMLMediaElement):
- html/HTMLMediaElement.h:
- platform/audio/AudioSession.cpp:
(WebCore):
(WebCore::AudioSession::AudioSession):
(WebCore::AudioSession::preferredBufferDuration):
(WebCore::AudioSession::setPreferredBufferDuration):
- platform/audio/AudioSession.h:
- platform/audio/AudioSessionListener.h:
(AudioSessionListener):
- platform/audio/AudioSessionManager.cpp: Removed.
- platform/audio/AudioSessionManager.h: Removed.
- platform/audio/ios/AudioDestinationIOS.cpp:
(WebCore::AudioDestinationIOS::configure):
- platform/audio/ios/AudioSessionIOS.mm:
(WebCore::AudioSession::preferredBufferDuration):
(WebCore::AudioSession::setPreferredBufferDuration):
- platform/audio/mac/AudioDestinationMac.cpp:
(WebCore::AudioDestination::hardwareSampleRate):
(WebCore::AudioDestinationMac::AudioDestinationMac):
(WebCore::AudioDestinationMac::configure):
- platform/audio/mac/AudioDestinationMac.h:
(AudioDestinationMac):
- platform/audio/mac/AudioSessionMac.cpp: Removed.
- platform/audio/mac/AudioSessionManagerMac.cpp: Removed.
Source/WTF:
- wtf/Platform.h:
- 18:11 Changeset [150560] by
-
Active DOM object resumption should match reason for suspending
https://bugs.webkit.org/show_bug.cgi?id=116476
Add a reason to the arguments for resuming active DOM object and
scheduled tasks and ensure that it matches the suspend reason before
actually resuming the objects.
Reviewed by Joseph Pecoraro.
No new tests, no change in functionality.
- bindings/js/PageScriptDebugServer.cpp:
(WebCore::PageScriptDebugServer::setJavaScriptPaused): Add matching resume reason
- dom/Document.cpp:
(WebCore::Document::resumeActiveDOMObjects): Check for matching resume reason
(WebCore::Document::resumeScheduledTasks): Check for matching resume reason
- dom/Document.h:
(Document):
- dom/ScriptExecutionContext.cpp:
(WebCore::ScriptExecutionContext::resumeActiveDOMObjects): Check for matching resume reason
- dom/ScriptExecutionContext.h:
(ScriptExecutionContext):
(WebCore::ScriptExecutionContext::reasonForSuspendingActiveDOMObjects): Expose suspend reason to subclasses
- history/CachedFrame.cpp:
(WebCore::CachedFrameBase::restore): Add matching resume reason
- page/Frame.cpp:
(WebCore::Frame::resumeActiveDOMObjectsAndAnimations): Add matching resume reason
- page/PageGroupLoadDeferrer.cpp:
(WebCore::PageGroupLoadDeferrer::~PageGroupLoadDeferrer): Add matching resume reason
- 18:07 Changeset [150559] by
-
Remove even more stale Mac test expectations. These tests are passing now!
- platform/mac/TestExpectations:
- 17:46 Changeset [150558] by
-
Even more Mac test expectation updates.
- platform/mac-wk2/TestExpectations:
- platform/mac/TestExpectations:
- platform/wk2/TestExpectations:
- plugins/mouse-events-expected.txt:
- 17:05 Changeset [150557] by
-
Unreviewed build fix; cast OSStatus to an int before passing it to printf.
- platform/audio/mac/AudioSessionMac.cpp:
(WebCore::AudioSession::setPreferredBufferSize):
- 17:02 Changeset [150556] by
-
Don't snapshot plugins that can't be snapshotted
https://bugs.webkit.org/show_bug.cgi?id=116490
Reviewed by Tim Horton.
Some plugins, such as QuickTime, don't support snapshotting. We're
planning on another way to capture snapshots, but in the meantime make
sure that they don't fail to initialise because they were told to
not use accelerated rendering.
Source/WebCore:
- WebCore.exp.in: Export restartSnapshottedPlugin so that it can be used from WK2.
- html/HTMLPlugInImageElement.cpp:
(WebCore::HTMLPlugInImageElement::restartSimilarPlugIns): Don't set the play state before restarting.
(WebCore::HTMLPlugInImageElement::userDidClickSnapshot): Make sure we indicated that this snapshotted plugin was clicked.
- html/HTMLPlugInImageElement.h: Move restartSnapshottedPlugin from protected to public.
- rendering/RenderSnapshottedPlugIn.cpp:
(WebCore::RenderSnapshottedPlugIn::handleEvent): Don't set play state here. Let the restart functions do it.
Source/WebKit2:
Expose a new message on the plugins so that they can tell the PluginView
if they support snapshotting or not.
- PluginProcess/PluginControllerProxy.cpp:
(WebKit::PluginControllerProxy::supportsSnapshotting): New method.
- PluginProcess/PluginControllerProxy.h: Ditto.
- PluginProcess/PluginControllerProxy.messages.in: Message signature.
- WebProcess/Plugins/Netscape/NetscapePlugin.h: Add supportsSnapshotting.
- WebProcess/Plugins/PDF/SimplePDFPlugin.h: Ditto.
- WebProcess/Plugins/Plugin.h: Ditto - pure virtual here.
- WebProcess/Plugins/PluginProxy.cpp:
(WebKit::PluginProxy::supportsSnapshotting): Send the message to the plugin process and return the results.
- WebProcess/Plugins/PluginProxy.h:
- WebProcess/Plugins/PluginView.cpp:
(WebKit::PluginView::didInitializePlugin): If the snapshotting plugin has a layer, make
sure to add it to the tree.
(WebKit::PluginView::isAcceleratedCompositingEnabled): Snapshotting plugins can use
accelerated compositing, just like normal plugins.
(WebKit::PluginView::pluginSnapshotTimerFired): If the plugin doesn't allow snapshotting,
immediately restart it.
- 16:43 Changeset [150555] by
-
Mac: Set the default audio buffer size to a large value for <video> elements.
https://bugs.webkit.org/show_bug.cgi?id=116342
Reviewed by Eric Carlson.
Source/WebCore:
To enable power savings by waking up the audio hardware less often, set the
requested buffer frame size to a large value, such as 4096. Since this results
in approximately 100ms worth of buffer, set the buffer size to a much lower
value when playing WebAudio, which is much more sensitive to latency than video
or audio elements.
Introduce a new class, AudioSessionManager, as well as a helper class,
AudioSessionManagerToken. Audio elements, video elements, and WebAudio destination
nodes will create and retain a token, and release the token in their destructor.
This allows the AudioSessionManager to track how many of what type of audio-
generating objects are in existence.
This requires implementing AudioSession for Mac platforms. Move the implementation
for retrieving the hardware sample rate and setting the buffer duration into
AudioSessionMac from AudioDestinationMac, to be shared with AudioSessionManagerMac.
Change the AudioSession method preferredBufferLength() into preferredBufferSize(),
as the callers really want to specify a buffer size, not a buffer duration. On iOS,
where the available API requires a duration, perform the conversion from duration to
size on behalf of the caller.
- html/HTMLMediaElement.h: Add a AudioSessionManagerToken member.
- html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::HTMLMediaElement): Initialize the token.
- platform/audio/AudioSession.cpp:
(WebCore::AudioSession::preferredBufferSize): Renamed from preferredBufferLength.
(WebCore::AudioSession::setPreferredBufferSize): Renamed from setPreferredBufferLength.
- platform/audio/AudioSession.h:
- platform/audio/AudioSessionManager.cpp:
(AudioSessionManagerToken::create): Simple factory method.
(AudioSessionManagerToken::AudioSessionManagerToken): Call AudioSessionManager::incrementCount().
(AudioSessionManagerToken::~AudioSessionManagerToken): Call AudioSessionManager::decrementCount().
(AudioSessionManager::sharedManager): Simple singleton method.
(AudioSessionManager::AudioSessionManager): Simple constructor.
(AudioSessionManager::has): Return whether the type is present.
(AudioSessionManager::incrementCount): Increment, then call updateSessionState()
(AudioSessionManager::decrementCount): Decrement, then call updateSessionState()
(AudioSessionManager::updateSessionState): Stub, does nothing.
- platform/audio/AudioSessionManager.h:
- platform/audio/ios/AudioDestinationIOS.cpp:
(WebCore::AudioDestinationIOS::configure): Call setPreferredBufferSize() instead of setPreferredBufferLength()
- platform/audio/ios/AudioSessionIOS.mm:
(WebCore::AudioSession::preferredBufferSize): Renamed from preferredBufferLength.
(WebCore::AudioSession::setPreferredBufferSize): Renamed from setPreferredBufferLength
- platform/audio/mac/AudioDestinationMac.cpp:
(WebCore::AudioDestination::hardwareSampleRate): Call AudioSession::sampleRate().
(WebCore::AudioDestinationMac::AudioDestinationMac): Create the AudioSessionManagerToken.
(WebCore::AudioDestinationMac::configure): Do not set the buffer size (this is done in AudioSessionManagerMac).
- platform/audio/mac/AudioDestinationMac.h:
- platform/audio/mac/AudioSessionMac.cpp: Added.
(WebCore::defaultDevice): Added, returns the default audio device.
(WebCore::AudioSession::AudioSession): Simple constructor.
(WebCore::AudioSession::~AudioSession): Simple destructor.
(WebCore::AudioSession::category): Stub, unimplemented.
(WebCore::AudioSession::setCategory): Ditto.
(WebCore::AudioSession::categoryOverride): Ditto.
(WebCore::AudioSession::setCategoryOverride): Ditto.
(WebCore::AudioSession::numberOfOutputChannels): Ditto.
(WebCore::AudioSession::setActive): Ditto.
(WebCore::AudioSession::sampleRate): Use the HAL to return the default audio device sample rate.
(WebCore::AudioSession::preferredBufferSize): Return the current HAL setting.
(WebCore::AudioSession::setPreferredBufferSize): Set the buffer size.
- platform/audio/mac/AudioSessionManagerMac.cpp:
(AudioSessionManager::updateSessionState): Set the buffer size depending on what audio outputs are present.
- WebCore.xcodeproj/project.pbxproj: Add the new files to the project.
Source/WTF:
- wtf/Platform.h: Add a WTF_USE_AUDIO_SESSION setting.
- 16:42 Changeset [150554] by
-
[CSS Shaders] valueForCustomFilterTransformParameter should check that renderer is a RenderBox
https://bugs.webkit.org/show_bug.cgi?id=116464
Reviewed by Alexandru Chiculita.
Source/WebCore:
If the renderer is not a RenderBox, we just use a zero size to compute the transform,
like in CSSComputedStyleDeclaration::computedTransform.
Test: css3/filters/custom/custom-filter-crash-inline-computed-style.html
- css/CSSComputedStyleDeclaration.cpp:
(WebCore::valueForCustomFilterTransformParameter):
LayoutTests:
- css3/filters/custom/custom-filter-crash-inline-computed-style-expected.txt: Added.
- css3/filters/custom/custom-filter-crash-inline-computed-style.html: Added.
- 16:27 Changeset [150553] by
-
Add a missing result after r150551.
- http/tests/inspector/network/network-size-chunked-expected.txt: Added.
- 16:24 Changeset [150552] by
-
Mac test expectation corrections after r150545.
- platform/mac-lion/http/tests/multipart/policy-ignore-crash-expected.txt: Added.
- platform/mac/TestExpectations:
- 15:53 Changeset [150551] by
-
More Mac test expectation cleanups.
- http/tests/inspector/network/network-size-sync-expected.txt:
- platform/mac/TestExpectations:
- platform/mac/webarchive/loading/mainresource-null-mimetype-crash-expected.txt:
- 15:27 Changeset [150550] by
-
Merged r150495. <rdar://problem/13915556>
- 15:24 Changeset [150549] by
-
Merged r150492. <rdar://problem/13921879>
- 15:23 Changeset [150548] by
-
Merged r150490. <rdar://problem/13921879>
- 15:19 Changeset [150547] by
-
Fix assertion in the getComputedStyle-background-shorthand.html test
https://bugs.webkit.org/show_bug.cgi?id=116639
Reviewed by Ryosuke Niwa.
getComputedStyle-background-shorthand.html was asserting under the newly added
code in willBeRemovedFromTree(). The assertion was caused by too many calls to
FrameView::removeSlowRepaintObject(), which happened because we called it for
RenderTexts, which share style with their parent renderer.
RenderText overrides styleWillChange() to do nothing, so we never
call add/removeSlowRepaintObject() for them. Thus we should also skip this
work in willBeRemovedFromTree() for RenderTexts.
- rendering/RenderObject.cpp:
(WebCore::RenderObject::willBeRemovedFromTree):
- 15:18 Changeset [150546] by
-
REGRESSION: fast/files/workers tests crash
https://bugs.webkit.org/show_bug.cgi?id=45575
Unskip the test since they no longer crash. Also rebaseline two tests as needed.
- fast/files/workers/worker-read-blob-async-expected.txt:
- fast/files/workers/worker-read-blob-sync-expected.txt:
- platform/mac/TestExpectations:
- 15:11 Changeset [150545] by
-
Mac test expectation cleanups. Unskip a bunch of tests and rebaseline them as needed.
- accessibility/aria-menubar-menuitems-expected.txt: Added.
- fast/replaced/border-radius-clip-expected.png:
- platform/mac/TestExpectations:
- platform/mac/fast/replaced/border-radius-clip-expected.txt:
- platform/mac/http/tests/multipart/policy-ignore-crash-expected.txt: Copied from LayoutTests/http/tests/multipart/policy-ignore-crash-expected.txt.
- platform/mac/media/video-volume-slider-expected.png: Added.
- platform/mac/media/video-volume-slider-expected.txt: Added.
- 15:08 Changeset [150544] by
-
Heap-use-after-free in WebCore::RenderLayerModelObject::hasSelfPaintingLayer
https://bugs.webkit.org/show_bug.cgi?id=116626
Patch by Bem Jones-Bey <bjonesbe@adobe.com> on 2013-05-22
Reviewed by David Hyatt.
This is a port of a Blink bug fix by Emil Eklund.
Account for saturated estimated position in RenderBlock::layoutBlockChild.
If the estimated top position is saturated the comparison with oldLogicalTop
might yield a false negative as adding and removing margins, borders etc from
a saturated number might yield incorrect results. If this is the case always
mark for layout.
Minimal test case impractical. See bug for raw fuzzer test case.
- platform/LayoutUnit.h:
(WebCore::LayoutUnit::mightBeSaturated): Add helper method for checking whether
a number might be saturated.
- rendering/RenderBlock.cpp:
(WebCore::RenderBlock::layoutBlockChild): Check if logicalTopEstimate is likely
to be saturated and if so mark for layout.
- 14:25 Changeset [150543] by
-
Fix crash when closing a page that's trying to access session storage
https://bugs.webkit.org/show_bug.cgi?id=116634
<rdar://problem/13904054>
Reviewed by Geoffrey Garen.
It is possible for the StorageManager to get messages from a web page that has already been closed by the UI process.
If that happens, just ignore the messages.
- UIProcess/Storage/StorageManager.cpp:
(WebKit::StorageManager::createSessionStorageMap):
(WebKit::StorageManager::getValues):
(WebKit::StorageManager::setItem):
(WebKit::StorageManager::removeItem):
(WebKit::StorageManager::clear):
(WebKit::StorageManager::destroySessionStorageNamespaceInternal):
- 14:22 Changeset [150542] by
-
[CSS Exclusions] property parsing tests should be revised
https://bugs.webkit.org/show_bug.cgi?id=116446
Reviewed by Alexandru Chiculita.
This is just a cleanup of the existing parsing tests. None of the tests have effectively
changed. The cleaned-up version should be simpler, more regular, and shorter:
- Moved all of parsing tests into a fast/exclusions/parsing subdirectory.
- Remove the obsolete "wrap-" prefix from file names.
- Eiminated excessive and needless stringification in many of the test functions.
- Moved common code into parsing-test-utils.js and refactored redundant code.
- fast/exclusions/parsing-wrap-shape-inside-expected.txt: Removed.
- fast/exclusions/parsing-wrap-shape-inside.html: Removed.
- fast/exclusions/parsing-wrap-shape-lengths-expected.txt: Removed.
- fast/exclusions/parsing-wrap-shape-lengths.html: Removed.
- fast/exclusions/parsing-wrap-shape-outside-expected.txt: Removed.
- fast/exclusions/parsing-wrap-shape-outside.html: Removed.
- fast/exclusions/parsing/parsing-shape-inside-expected.txt: Added.
- fast/exclusions/parsing/parsing-shape-inside.html: Added.
- fast/exclusions/parsing/parsing-shape-lengths-expected.txt: Added.
- fast/exclusions/parsing/parsing-shape-lengths.html: Added.
- fast/exclusions/parsing/parsing-shape-margin-expected.txt: Added.
- fast/exclusions/parsing/parsing-shape-margin.html: Added.
- fast/exclusions/parsing/parsing-shape-outside-expected.txt: Added.
- fast/exclusions/parsing/parsing-shape-outside.html: Added.
- fast/exclusions/parsing/parsing-shape-padding-expected.txt: Added.
- fast/exclusions/parsing/parsing-shape-padding.html: Added.
- fast/exclusions/parsing/parsing-wrap-flow-expected.txt: Added.
- fast/exclusions/parsing/parsing-wrap-flow.html: Added.
- fast/exclusions/parsing/parsing-wrap-through-expected.txt: Added.
- fast/exclusions/parsing/parsing-wrap-through.html: Added.
- fast/exclusions/parsing/script-tests/parsing-shape-inside.js: Added.
(string_appeared_here):
- fast/exclusions/parsing/script-tests/parsing-shape-lengths.js: Added.
(getStyleValue):
(testStyleValue):
(testComputedStyleValue):
(testInvalidValue):
- fast/exclusions/parsing/script-tests/parsing-shape-margin.js: Added.
(testShapeSpecifiedProperty):
- fast/exclusions/parsing/script-tests/parsing-shape-outside.js: Added.
(testShapePropertyParsingFailure):
- fast/exclusions/parsing/script-tests/parsing-shape-padding.js: Added.
(testShapeSpecifiedProperty):
- fast/exclusions/parsing/script-tests/parsing-test-utils.js: Added.
(getCSSText):
(getComputedStyleValue):
(getParentAndChildComputedStyles):
(getParentAndChildComputedStylesString):
(getChildComputedStyle):
(testShapeSpecifiedProperty):
(testShapeComputedProperty):
(testNotInheritedShapeChildProperty):
- fast/exclusions/parsing/script-tests/parsing-wrap-flow.js: Added.
- fast/exclusions/parsing/script-tests/parsing-wrap-through.js: Added.
- fast/exclusions/shape-margin-parsing-expected.txt: Removed.
- fast/exclusions/shape-margin-parsing.html: Removed.
- fast/exclusions/shape-padding-parsing-expected.txt: Removed.
- fast/exclusions/shape-padding-parsing.html: Removed.
- fast/exclusions/wrap-flow-parsing-expected.txt: Removed.
- fast/exclusions/wrap-flow-parsing.html: Removed.
- fast/exclusions/wrap-through-parsing-expected.txt: Removed.
- fast/exclusions/wrap-through-parsing.html: Removed.
- 13:58 Changeset [150541] by
-
Merged r150486. <rdar://problem/13822856>
- 13:55 Changeset [150540] by
-
Merged r150484. <rdar://problem/13822856>
- 13:52 Changeset [150539] by
-
Remove CSSPropertySourceData emptyCSSPropertySourceData
https://bugs.webkit.org/show_bug.cgi?id=116517
Reviewed by Darin Adler.
This was added for chromium and does not seem to be used
anywhere anymore.
- css/CSSParser.cpp:
(WebCore::CSSParser::CSSParser):
- css/CSSPropertySourceData.cpp:
- css/CSSPropertySourceData.h:
- 13:52 Changeset [150538] by
-
Versioning.
- 13:50 Changeset [150537] by
-
Crashes in NetworkProcess due to incorrect private browsing session tracking
https://bugs.webkit.org/show_bug.cgi?id=116628
Reviewed by Brady Eidson.
The current API for private browsing makes it extremely difficult to track sessions.
Private browsing is enabled via WKPreferences, but the session is shared, so it
has to be maintained while there is any chance that any page group anywhere still
needs it.
This patch fixes some of the issues, but ultimately, I think that we'll just need
to deprecate and replace the API.
- NetworkProcess/NetworkConnectionToWebProcess.cpp: (WebKit::storageSession): There are valid scenarios where privateBrowsingEnabled is true, but there is no private browsing session. Handle that without crashing, although this unfortunately means that it will be harder to spot logic errors when using a wrong session. (WebKit::NetworkConnectionToWebProcess::registerBlobURL): Removed an obsolete FIXME.
- NetworkProcess/mac/RemoteNetworkingContext.h: Changed privateBrowsingSession() to return a pointer, as no caller could know when it was safe to call it.
- NetworkProcess/mac/RemoteNetworkingContext.mm: (WebKit::RemoteNetworkingContext::storageSession): Handle the case where private browsing session is unexpectedly missing without crashing. (WebKit::RemoteNetworkingContext::privateBrowsingSession): Changed to return a pointer.
- UIProcess/WebContext.cpp: (WebKit::WebContext::ensureNetworkProcess): Actually initialize privateBrowsingEnabled creation parameter. It would be very difficult to figure out 100% reliably whether NetworkProcess needs a private browsing session with the current API, but for existing clients, looking at m_privateBrowsingEnterCount is good enough. Certainly better than not initializing.
- WebProcess/InjectedBundle/InjectedBundle.cpp: (WebKit::InjectedBundle::setPrivateBrowsingEnabled): Added a FIXME.
- 13:42 Changeset [150536] by
-
fourthTier: It should be possible to use more than one compiler thread
https://bugs.webkit.org/show_bug.cgi?id=116630
Reviewed by Mark Hahnenberg.
This gives us the ability to use more compiler threads, but doesn't actually
enable the functionality because it isn't a speed-up on any benchmark. It can
even be a slow-down. This also adds the ability to disable concurrent
compilation if we're on a uniprocessor machine, and adds more logging to the
worklist code to allow us to investigate how many threads are active. It
appears that even on the most compiler-heavy benchmarks, we never have enough
work for more than 4 threads, and even then the 4 threads are all active for
a short time.
Something that having more threads does accomplish is that it shakes out bugs.
This patch fixes a bug with Watchpoint not being thread-safe ref-counted,
which enabling 7 compilation threads did catch.
As it stands, this patch is performance-neutral and just fixes bugs and adds
some options.
- bytecode/Watchpoint.h:
- dfg/DFGCommon.h:
(JSC::DFG::enableConcurrentJIT):
- dfg/DFGWorklist.cpp:
(JSC::DFG::Worklist::Worklist):
(JSC::DFG::Worklist::~Worklist):
(JSC::DFG::Worklist::finishCreation):
(JSC::DFG::Worklist::create):
(JSC::DFG::Worklist::enqueue):
(JSC::DFG::Worklist::waitUntilAllPlansForVMAreReady):
(JSC::DFG::Worklist::dump):
(JSC::DFG::Worklist::runThread):
(JSC::DFG::initializeGlobalWorklistOnce):
- dfg/DFGWorklist.h:
- runtime/Options.cpp:
(JSC::computeNumberOfWorkerThreads):
(JSC):
(JSC::computeNumberOfGCMarkers):
- runtime/Options.h:
(JSC):
- 13:30 Changeset [150535] by
-
Buildfix for !ENABLE(FULLSCREEN_API) platforms.
Rubber-stamped by Ryosuke Niwa.
- rendering/RenderInline.cpp: Missing #if guard added.
(WebCore::RenderInline::splitInlines):
- 13:11 Changeset [150534] by
-
Fix a typo in 150469.
- editing/Editor.cpp:
(WebCore::dispatchEditableContentChangedEvents):
- 13:09 Changeset [150533] by
-
fourthTier: FTL shouldn't use the LLVM global context, and should instead create its own context for each compilation
https://bugs.webkit.org/show_bug.cgi?id=116631
Reviewed by Mark Hahnenberg.
In the future we might want to share contexts for multiple compilations, but for
now using one context per compilation is a progression over just constantly using
the global context.
- dfg/DFGPlan.cpp:
(JSC::DFG::Plan::compileInThread):
(DFG):
(JSC::DFG::Plan::compileInThreadImpl):
- dfg/DFGPlan.h:
- ftl/FTLAbbreviatedTypes.h:
(FTL):
- ftl/FTLAbbreviations.h:
(JSC::FTL::voidType):
(JSC::FTL::int1Type):
(JSC::FTL::int8Type):
(JSC::FTL::int32Type):
(JSC::FTL::int64Type):
(JSC::FTL::intPtrType):
(JSC::FTL::doubleType):
(JSC::FTL::structType):
(JSC::FTL::mdKindID):
(JSC::FTL::mdString):
(JSC::FTL::mdNode):
(JSC::FTL::appendBasicBlock):
(JSC::FTL::insertBasicBlock):
- ftl/FTLAbstractHeap.cpp:
(JSC::FTL::AbstractHeap::tbaaMetadataSlow):
(JSC::FTL::IndexedAbstractHeap::IndexedAbstractHeap):
(JSC::FTL::NumberedAbstractHeap::NumberedAbstractHeap):
(JSC::FTL::AbsoluteAbstractHeap::AbsoluteAbstractHeap):
- ftl/FTLAbstractHeap.h:
(IndexedAbstractHeap):
(NumberedAbstractHeap):
(AbsoluteAbstractHeap):
- ftl/FTLAbstractHeapRepository.cpp:
(JSC::FTL::AbstractHeapRepository::AbstractHeapRepository):
- ftl/FTLAbstractHeapRepository.h:
(AbstractHeapRepository):
- ftl/FTLCommonValues.cpp:
(JSC::FTL::CommonValues::CommonValues):
- ftl/FTLCommonValues.h:
(CommonValues):
- ftl/FTLCompile.cpp:
(JSC::FTL::mmAllocateCodeSection):
- ftl/FTLIntrinsicRepository.cpp:
(JSC::FTL::IntrinsicRepository::IntrinsicRepository):
- ftl/FTLIntrinsicRepository.h:
(FTL):
(IntrinsicRepository):
- ftl/FTLLowerDFGToLLVM.cpp:
(JSC::FTL::LowerDFGToLLVM::LowerDFGToLLVM):
(JSC::FTL::LowerDFGToLLVM::lower):
- ftl/FTLOutput.cpp:
(JSC::FTL::Output::Output):
- ftl/FTLOutput.h:
(Output):
(JSC::FTL::Output::newBlock):
- ftl/FTLState.cpp:
(JSC::FTL::State::State):
(JSC::FTL::State::~State):
(FTL):
- ftl/FTLState.h:
(State):
- runtime/Options.h:
(JSC):
- 12:48 Changeset [150532] by
-
[wk2] Should be able to make tiled WKViews clip to the view's exposed rect without using autolayout
https://bugs.webkit.org/show_bug.cgi?id=116625
<rdar://problem/13962816>
Reviewed by Anders Carlsson.
The two modes are not dependent; it should be possible to enable
clips-to-exposed-rect without using autolayout. This enables apps to
make very large tiled WKViews that don't create tiles for the whole
area of the view (and also disables scrolling, since that is currently
incompatible with clipping to the exposed rect).
- UIProcess/API/mac/WKView.mm:
(-[WKView setFrameSize:]):
(-[WKView _updateWindowAndViewFrames]):
(-[WKView initWithFrame:contextRef:pageGroupRef:relatedToPage:]):
(-[WKView enableFrameSizeUpdates]):
(-[WKView forceAsyncDrawingAreaSizeUpdate:]):
Rename _expandsToFitContentViaAutoLayout to _clipsToVisibleRect for accuracy.
(-[WKView setMinimumWidthForAutoLayout:]):
Factor clips-to-sisible-rect toggle out into setShouldClipToVisibleRect.
(-[WKView shouldClipToVisibleRect]):
(-[WKView setShouldClipToVisibleRect:]):
Allow setting clips-to-visible-rect without autolayout.
- UIProcess/API/mac/WKViewPrivate.h:
Add shouldClipToVisibleRect property.
- 12:21 Changeset [150531] by
-
REGRESSION: ASSERTION FAILED: obj->isRenderInline() obj == this, Bad cast in WebCore::RenderBlock::createLineBoxes https://bugs.webkit.org/show_bug.cgi?id=110622
Reviewed by David Hyatt.
Source/WebCore:
Merge https://chromium.googlesource.com/chromium/blink/+/998ad358eed702b873dd54697b3fa3f952e0feb7
Inserting an element before the fullscreened element could crash if it caused a containing inline to be split,
since the splitting logic doesn't expect the fullscreened element to be wrapped in a RenderFullScreen. This patch changes
inline splitting to be aware of RenderFullScreen.
Test: fullscreen/full-screen-inline-split-crash.html
- rendering/RenderInline.cpp:
(WebCore::RenderInline::splitInlines):
LayoutTests:
Add a regression test.
- fullscreen/full-screen-inline-split-crash-expected.txt: Added.
- fullscreen/full-screen-inline-split-crash.html: Added.
- 12:06 Changeset [150530] by
-
Fix issues with focus rings on search fields
https://bugs.webkit.org/show_bug.cgi?id=116591
Source/WebCore:
Reviewed by Darin Adler.
On a future OS, the rect returned by _focusRingVisibleRect should
be in the coordinate system of the that NSView, not the base
coordinate system, so fix the code that sets and reads the global
focusRingClipRect variable accordingly.
- platform/graphics/mac/WebLayer.mm:
(drawLayerContents):
- platform/mac/ThemeMac.mm:
(-[WebCoreFlippedView _focusRingVisibleRect]):
Source/WebKit/mac:
Reviewed by Darin Adler.
In WebKit1, focus rings didn't render correctly on some OSes until
some other control drew a focus ring, which changes internal AppKit state
on the focused view. Fix by overriding an internal AppKit method to always
disable AppKit's "automatic" focus ring drawing.
- WebView/WebHTMLView.mm:
(-[WebHTMLView _automaticFocusRingDisabled]):
- 12:06 Changeset [150529] by
-
New Flickr doesn't get fast scrolling but should
https://bugs.webkit.org/show_bug.cgi?id=116514
Source/WebCore:
Reviewed by Darin Adler.
RenderObject increments and decrements a counter of slow repaint objects on
FrameView when it sees style changes related to background-attachment:fixed.
However, it omitted to decrement the count when a renderer with a fixed background
was destroyed.
This caused Flickr to never fall into fast scrolling mode, since it toggled
display:none on an element with a fixed background during loading, then removed
the fixed background.
Did some minor cleanup of #ENABLE(FAST_MOBILE_SCROLLING) crap.
Tests: platform/mac-wk2/tiled-drawing/slow-scrolling-background-toggle.html
platform/mac-wk2/tiled-drawing/slow-scrolling-hidden-background-toggle.html
platform/mac-wk2/tiled-drawing/slow-scrolling.html
- rendering/RenderObject.cpp:
(WebCore::shouldRepaintFixedBackgroundsOnScroll):
(WebCore::RenderObject::styleWillChange):
(WebCore::RenderObject::willBeRemovedFromTree):
LayoutTests:
Reviewed by Darin Adler.
slow-scrolling-hidden-background-toggle.html actually tests this patch. The other
two tests were added because there appear to be no tests for basic internals.mainThreadScrollingReasons()
functionality.
- platform/mac-wk2/tiled-drawing/slow-scrolling-background-toggle-expected.txt: Added.
- platform/mac-wk2/tiled-drawing/slow-scrolling-background-toggle.html: Added.
- platform/mac-wk2/tiled-drawing/slow-scrolling-expected.txt: Added.
- platform/mac-wk2/tiled-drawing/slow-scrolling-hidden-background-toggle-expected.txt: Added.
- platform/mac-wk2/tiled-drawing/slow-scrolling-hidden-background-toggle.html: Added.
- platform/mac-wk2/tiled-drawing/slow-scrolling.html: Added.
- 12:01 Changeset [150528] by
-
New Branch.
- 11:59 Changeset [150527] by
-
Need to Remove Anonymous Wrappers When All Children Become Inline
https://bugs.webkit.org/show_bug.cgi?id=115687
Reviewed by David Hyatt.
Source/WebCore:
Remove a simple subset of anonymous wrappers, i.e. anonymous blocks without continuations,
when a block becoming float or positioned removes the need for them.
Test: fast/block/remove-anonymous-wrappers-when-element-changes-to-float-or-positioned.html
- rendering/RenderBlock.h:
(RenderBlock):
- rendering/RenderObject.cpp:
(WebCore):
(WebCore::RenderObject::removeAnonymousWrappersFromLineIfNecessary):
(WebCore::RenderObject::styleWillChange):
(WebCore::RenderObject::styleDidChange):
- rendering/RenderObject.h:
(RenderObject):
LayoutTests:
- fast/block/remove-anonymous-wrappers-when-element-changes-to-float-or-positioned-expected.txt: Added.
- fast/block/remove-anonymous-wrappers-when-element-changes-to-float-or-positioned.html: Added.
- platform/qt/fast/block/float/float-not-removed-from-next-sibling2-expected.txt:
- platform/qt/fast/dynamic/002-expected.txt:
- 11:47 Changeset [150526] by
-
Updated LLVM drops to include MCJIT memory management APIs.
Rubber stamped by Mark Hahnenberg.
- LLVMIncludesMountainLion.tar.bz2:
- LLVMLibrariesMountainLion.tar.bz2:
- 11:38 Changeset [150525] by
-
[css] Update ETextDecorations enum to TextDecorations
https://bugs.webkit.org/show_bug.cgi?id=116581
Reviewed by Andreas Kling.
Currently, ETextDecorations enum uses a C-like coding style for its
enumeration values, as well as a 'E' prefix that surely had some meaning
in the past, but not nowadays.
No behavior changes, covered by existing tests.
- accessibility/AccessibilityRenderObject.cpp:
(WebCore::AccessibilityRenderObject::hasPlainText):
(WebCore::AccessibilityRenderObject::hasUnderline):
- accessibility/atk/WebKitAccessibleInterfaceText.cpp:
(getAttributeSetForAccessibilityObject):
- accessibility/ios/WebAccessibilityObjectWrapperIOS.mm:
(AXAttributeStringSetStyle):
- accessibility/mac/WebAccessibilityObjectWrapperMac.mm:
(AXAttributeStringSetStyle):
- css/CSSComputedStyleDeclaration.cpp:
(WebCore::renderTextDecorationFlagsToCSSValue):
- css/CSSPrimitiveValueMappings.h:
(WebCore::CSSPrimitiveValue::operator TextDecoration):
- css/DeprecatedStyleBuilder.cpp:
(WebCore::ApplyPropertyTextDecoration::applyValue):
(WebCore::ApplyPropertyTextDecoration::createHandler):
- editing/mac/EditorMac.mm:
(WebCore::Editor::fontAttributesForSelectionStart):
- platform/mac/HTMLConverter.mm:
(+[WebHTMLConverter editingAttributedStringFromRange:]):
- rendering/InlineTextBox.cpp:
(WebCore::InlineTextBox::paint):
(WebCore::InlineTextBox::paintDecoration):
- rendering/InlineTextBox.h:
- rendering/RenderObject.cpp:
(WebCore::RenderObject::getTextDecorationColors):
- rendering/style/RenderStyle.h:
- rendering/style/RenderStyleConstants.h:
(WebCore::operator|):
(WebCore::operator|=):
- rendering/style/StyleVisualData.h:
(StyleVisualData):
- rendering/svg/SVGInlineTextBox.cpp:
(WebCore::SVGInlineTextBox::paint):
(WebCore::positionOffsetForDecoration):
(WebCore::thicknessForDecoration):
(WebCore::findRenderObjectDefininingTextDecoration):
(WebCore::SVGInlineTextBox::paintDecoration):
(WebCore::SVGInlineTextBox::paintDecorationWithStyle):
- rendering/svg/SVGInlineTextBox.h:
- 11:30 Changeset [150524] by
-
[GStreamer] Allow multiple log levels in media player
https://bugs.webkit.org/show_bug.cgi?id=116618
Patch by Lori Anderson <lori@rkymtnhi.com> on 2013-05-22
Reviewed by Philippe Normand.
No new tests, just changing logging.
- platform/graphics/gstreamer/GStreamerUtilities.h:
- platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
(WebCore::MediaPlayerPrivateGStreamer::load):
(WebCore::MediaPlayerPrivateGStreamer::pause):
(WebCore::MediaPlayerPrivateGStreamer::seek):
(WebCore::MediaPlayerPrivateGStreamer::setRate):
(WebCore::MediaPlayerPrivateGStreamer::handleMessage):
- 11:23 WebKit Team edited by
- added self (diff)
- 11:13 Changeset [150523] by
-
Unreviewed. Added myself to contributors.json.
- Scripts/webkitpy/common/config/contributors.json:
Added myself.
- 11:06 Changeset [150522] by
-
The LayoutMilestones didLayout callback sometimes fires with a milestone value of 0
https://bugs.webkit.org/show_bug.cgi?id=116623
Reviewed by Geoffrey Garen.
Make sure we have actually achieved some milestones before calling didLayout().
- page/FrameView.cpp:
(WebCore::FrameView::performPostLayoutTasks):
- 10:09 Changeset [150521] by
-
Remove unused private browsing override support
<rdar://problem/11969491>
Rubber-stamped by Sam Weinig.
Private browsing override is unused, and it complicates tracking of private
browsing sessions in UI process (in fact, it doesn't plug into this tracking at all,
so we would just assert if we used the API).
- 10:01 Changeset [150520] by
-
FTL should force LLVM to use our own JIT memory allocator, and we shouldn't have to keep around an LLVMExecutionEngineRef to keep code alive
https://bugs.webkit.org/show_bug.cgi?id=113619
Reviewed by Geoffrey Garen.
This uses new API that I've exposed, which allows for memory manager callbacks
from within LLVM. LLVM may allocate multiple independent chunks of memory for
a module, and we track all of those in a Vector in FTL::JITCode.
- ftl/FTLCompile.cpp:
(JSC::FTL::mmAllocateCodeSection):
(FTL):
(JSC::FTL::mmAllocateDataSection):
(JSC::FTL::mmApplyPermissions):
(JSC::FTL::mmDestroy):
(JSC::FTL::compile):
- ftl/FTLJITCode.cpp:
(JSC::FTL::JITCode::JITCode):
(JSC::FTL::JITCode::~JITCode):
(JSC::FTL::JITCode::addHandle):
(FTL):
(JSC::FTL::JITCode::initializeCode):
- ftl/FTLJITCode.h:
(JITCode):
(JSC::FTL::JITCode::handles):
- ftl/FTLJITFinalizer.cpp:
(JSC::FTL::JITFinalizer::~JITFinalizer):
(JSC::FTL::JITFinalizer::finalizeFunction):
- ftl/FTLJITFinalizer.h:
(JSC::FTL::JITFinalizer::initializeEntrypointLinkBuffer):
(JITFinalizer):
- ftl/FTLLink.cpp:
(JSC::FTL::link):
- ftl/FTLState.cpp:
(JSC::FTL::State::State):
- ftl/FTLState.h:
(State):
- 09:42 Changeset [150519] by
-
[sh4] Remove MacroAssemblerSH4.cpp file.
https://bugs.webkit.org/show_bug.cgi?id=116596.
Patch by Julien Brianceau <jbrianceau@nds.com> on 2013-05-22
Reviewed by Geoffrey Garen.
Move linkCall and repatchCall implementations from MacroAssemblerSH4.cpp
to MacroAssemblerSH4.h and remove MacroAssemblerSH4.cpp, as it is done
for other architectures.
- GNUmakefile.list.am:
- JavaScriptCore.xcodeproj/project.pbxproj:
- Target.pri:
- assembler/MacroAssemblerSH4.cpp: Removed.
- assembler/MacroAssemblerSH4.h:
(JSC::MacroAssemblerSH4::linkCall):
(MacroAssemblerSH4):
(JSC::MacroAssemblerSH4::repatchCall):
- 09:22 Changeset [150518] by
-
[GTK] Reimplement atk_text_get_text_*_offset for CHAR boundary
https://bugs.webkit.org/show_bug.cgi?id=114870
Reviewed by Martin Robinson.
Re-implement this functions without using GailTextUtil nor Pango.
- accessibility/atk/WebKitAccessibleInterfaceText.cpp:
(webkitAccessibleTextGetChar): New function.
(webkitAccessibleTextGetTextForOffset): Call the new function for CHAR.
- 09:18 Changeset [150517] by
-
[Qt][WK2] Unreviewed gardening. Skipping failing printing tests after r150046.
https://bugs.webkit.org/show_bug.cgi?id=116617
- platform/qt-5.0-wk2/TestExpectations:
- 09:11 Changeset [150516] by
-
[Mac] Captions menu isn't internationalized, doesn't use rtl layout for rtl languages
https://bugs.webkit.org/show_bug.cgi?id=116605
Reviewed by Eric Carlson.
Ensure that rtl languages are properly recognized by the captions menu such that text
layout and alignment is natural.
- css/mediaControls.css:
(::-webkit-media-controls):
Remove the forced "direction: ltr" property which was forcing all layout to be
left-to-right.
- css/mediaControlsQuickTime.css:
(video::-webkit-media-controls-closed-captions-track-list):
Remove the forced "text-align: left" property which was forcing all caption
titles to be left-aligned.
(video::-webkit-media-controls-closed-captions-track-list h3):
Use "-webkit-margin-start" instead of "padding-left" to position the heading
such that it adds the margin on the left for ltr languages and on the right
for rtl languages.
(video::-webkit-media-controls-closed-captions-track-list li):
Set "position: relative" such that the ::before pseudo-class can use "position: absolute"
to position itself relative to the caption title rather than the list element.
(video::-webkit-media-controls-closed-captions-track-list li.selected::before):
(video::-webkit-media-controls-closed-captions-track-list li.selected:hover::before):
Make the checkmarks a ::before pseudo-class rather than a background-image such that
we may control its positioning using the internationlization-friendly "-webkit-margin-start"
property.
- html/shadow/MediaControlElements.cpp:
(WebCore::MediaControlClosedCaptionsContainerElement::create):
Set the "dir" attribute to "auto" such that rtl languages are correctly handled. This
means that the first string displayed within the captions container, currently the
"Subtitles" <h3> element, defines the direction for the whole container. It is therefore
crucial that this string is properly localized in rtl languages as falling back to
english would mean the entire caption container is using ltr.
- 07:16 Changeset [150515] by
-
[EFL][WK2] Add sub menus to MiniBrowser
https://bugs.webkit.org/show_bug.cgi?id=114728
Patch by Michał Pakuła vel Rutka <m.pakula@samsung.com> on 2013-05-22
Reviewed by Christophe Dumez.
Elementary Ctxpopup widget API does not provide support for submenu item.
To implement ewk submenus in Minibrowser, Ctxpopup has to be replaced with
Elementary Menu widget.
- MiniBrowser/efl/main.c:
(_Browser_Window):
(on_window_resize):
(window_free):
(context_menu_item_selected_cb):
(context_menu_populate):
(on_context_menu_show):
(on_context_menu_hide):
(window_create):
- 07:15 Changeset [150514] by
-
[Qt] Unreviewed gardening. Skip failing tests after r150065 and r150498.
Patch by Gabor Abraham <abrhm@inf.u-szeged.hu> on 2013-05-22
- platform/qt/TestExpectations:
- 06:45 Changeset [150513] by
-
WebCore fails to compile with -Werror=maybe-uninitialized on GCC 4.8.0
https://bugs.webkit.org/show_bug.cgi?id=116340
Reviewed by Christophe Dumez.
No new tests, just a simple build fix.
- inspector/InspectorDOMStorageAgent.cpp:
(WebCore::InspectorDOMStorageAgent::setDOMStorageItem): Init frame variable.
(WebCore::InspectorDOMStorageAgent::removeDOMStorageItem): Init frame variable.
- 06:36 Changeset [150512] by
-
[CSSRegions] Remove platform-specific differences from offsetLeft/offsetTop tests
https://bugs.webkit.org/show_bug.cgi?id=116607
Patch by Radu Stavila <stavila@adobe.com> on 2013-05-22
Reviewed by Antti Koivisto.
- fast/regions/offsetLeft-offsetTop-in-multiple-regions-expected.txt:
- fast/regions/offsetLeft-offsetTop-in-multiple-regions.html:
- fast/regions/offsetLeft-offsetTop-in-region-float-vert-rl.html:
- fast/regions/offsetLeft-offsetTop-inlines-region-in-element.html:
- 05:49 Changeset [150511] by
-
[EFL] Add command line option to MiniBrowser to set cookies policy.
https://bugs.webkit.org/show_bug.cgi?id=115028
Patch by Andrzej Badowski <a.badowski@samsung.com> on 2013-05-22
Reviewed by Christophe Dumez.
Add command line option to MiniBrowser EFL to set cookies policy:
-p parameter with value from (always, never, no-third-party) respectively (EWK_COOKIE_ACCEPT_POLICY_ALWAYS,
EWK_COOKIE_ACCEPT_POLICY_NEVER, EWK_COOKIE_ACCEPT_POLICY_NO_THIRD_PARTY).
- MiniBrowser/efl/main.c:
(elm_main):
- 05:47 WebKitGTK/2.0.x edited by
- Proposed changes for 2.0.3 (diff)
- 05:20 Changeset [150510] by
-
[Qt] Unreviewed gardening, updating expected files after r150508.
Patch by Gabor Abraham <abrhm@inf.u-szeged.hu> on 2013-05-22
- platform/qt-5.0-wk2/svg/transforms/text-with-pattern-inside-transformed-html-expected.png: Added.
- platform/qt-5.0/svg/transforms/text-with-pattern-inside-transformed-html-expected.txt: Removed.
- 05:04 Changeset [150509] by
-
Have IDL interface names match their global constructor
https://bugs.webkit.org/show_bug.cgi?id=116520
Reviewed by Kentaro Hara.
Have IDL interface names match their global constructor name on DOMWindow.idl
using [InterfaceName] extended attribute. This affects in particular webkit
prefixed constructors. This is mandated by Web IDL specification:
http://dev.w3.org/2006/webapi/WebIDL/#es-interfaces
As a consequence, we can now get rid of [NoInterfaceObject] extended attribute
for those interfaces and have their global constructors automatically generated
by the bindings generator.
No new tests, no behavior change for layout tests.
- Modules/mediasource/MediaSource.idl:
- Modules/mediasource/SourceBuffer.idl:
- Modules/mediasource/SourceBufferList.idl:
- Modules/mediastream/MediaStream.idl:
- Modules/mediastream/RTCPeerConnection.idl:
- Modules/speech/SpeechGrammar.idl:
- Modules/speech/SpeechGrammarList.idl:
- Modules/speech/SpeechRecognition.idl:
- Modules/speech/SpeechRecognitionError.idl:
- Modules/speech/SpeechRecognitionEvent.idl:
- Modules/webaudio/OfflineAudioContext.idl:
- Modules/webaudio/PannerNode.idl:
- dom/ShadowRoot.idl:
- page/DOMWindow.idl:
- 04:42 Changeset [150508] by
-
[Qt] Unreviewed gardening. Unskipping some tests.
https://bugs.webkit.org/show_bug.cgi?id=26830
https://bugs.webkit.org/show_bug.cgi?id=99870
https://bugs.webkit.org/show_bug.cgi?id=108429
https://bugs.webkit.org/show_bug.cgi?id=23166
https://bugs.webkit.org/show_bug.cgi?id=98523
https://bugs.webkit.org/show_bug.cgi?id=110654
Patch by Gabor Abraham <abrhm@inf.u-szeged.hu> on 2013-05-22
- platform/qt-5.0/svg/transforms/text-with-pattern-inside-transformed-html-expected.png: Removed.
- platform/qt/TestExpectations:
- platform/qt/css2.1/20110323/replaced-elements-001-expected.png:
- platform/qt/css2.1/20110323/replaced-elements-001-expected.txt:
- platform/qt/css3/flexbox/button-expected.png: Added.
- platform/qt/css3/flexbox/button-expected.txt: Added.
- platform/qt/fast/css/background-clip-radius-values-expected.png: Added.
- platform/qt/fast/css/background-clip-radius-values-expected.txt: Added.
- platform/qt/fast/images/repaint-subrect-grid-expected.txt: Added.
- platform/qt/fast/multicol/single-line-expected.png: Added.
- platform/qt/fast/multicol/single-line-expected.txt: Added.
- platform/qt/ietestcenter/css3/bordersbackgrounds/border-radius-clip-002-expected.png:
- platform/qt/svg/custom/foreign-object-skew-expected.txt:
- platform/qt/svg/custom/js-late-pattern-creation-expected.png:
- platform/qt/svg/custom/pattern-scaling-expected.png: Added.
- platform/qt/svg/custom/pattern-scaling-expected.txt: Added.
- platform/qt/svg/custom/pattern-skew-transformed-expected.png:
- platform/qt/svg/custom/pattern-skew-transformed-expected.txt: Added.
- platform/qt/svg/custom/pattern-with-transformation-expected.png:
- platform/qt/svg/custom/text-ctm-expected.png:
- platform/qt/svg/custom/text-ctm-expected.txt:
- platform/qt/svg/transforms/text-with-pattern-inside-transformed-html-expected.png:
- platform/qt/svg/transforms/text-with-pattern-inside-transformed-html-expected.txt:
- platform/qt/tables/mozilla/bugs/bug92647-2-expected.png:
- platform/qt/tables/mozilla/bugs/bug92647-2-expected.txt:
- 02:37 Changeset [150507] by
-
[EFL][WK2] Change Ewk_Context_Menu_Item's parent menu to raw pointer
https://bugs.webkit.org/show_bug.cgi?id=116549
Patch by Michał Pakuła vel Rutka <m.pakula@samsung.com> on 2013-05-22
Reviewed by Christophe Dumez.
A context menu item should have only a pointer to parent menu instead
of owning parent menu.
- UIProcess/API/efl/ewk_context_menu_item.cpp:
(EwkContextMenuItem::EwkContextMenuItem):
- UIProcess/API/efl/ewk_context_menu_item_private.h:
(EwkContextMenuItem::create):
(EwkContextMenuItem::parentMenu):
(EwkContextMenuItem):
- 02:25 Changeset [150506] by
-
[Qt] Unreviewed gardening. Unskipping some tests.
https://bugs.webkit.org/show_bug.cgi?id=105437
Patch by Gabor Abraham <abrhm@inf.u-szeged.hu> on 2013-05-22
- platform/qt/svg/W3C-SVG-1.1-SE/filters-image-03-f-expected.png:
- platform/qt/svg/W3C-SVG-1.1-SE/filters-image-05-f-expected.png:
- platform/qt/svg/W3C-SVG-1.1/filters-displace-01-f-expected.png:
- platform/qt/svg/dynamic-updates/SVGFEDisplacementMapElement-dom-in-attr-expected.png: Added.
- platform/qt/svg/dynamic-updates/SVGFEDisplacementMapElement-dom-in-attr-expected.txt: Added.
- platform/qt/svg/dynamic-updates/SVGFEDisplacementMapElement-dom-in2-attr-expected.png: Added.
- platform/qt/svg/dynamic-updates/SVGFEDisplacementMapElement-dom-in2-attr-expected.txt: Added.
- platform/qt/svg/dynamic-updates/SVGFEDisplacementMapElement-dom-scale-attr-expected.png: Added.
- platform/qt/svg/dynamic-updates/SVGFEDisplacementMapElement-dom-scale-attr-expected.txt: Added.
- platform/qt/svg/dynamic-updates/SVGFEDisplacementMapElement-dom-xChannelSelector-attr-expected.png: Added.
- platform/qt/svg/dynamic-updates/SVGFEDisplacementMapElement-dom-xChannelSelector-attr-expected.txt: Added.
- platform/qt/svg/dynamic-updates/SVGFEDisplacementMapElement-dom-yChannelSelector-attr-expected.png: Added.
- platform/qt/svg/dynamic-updates/SVGFEDisplacementMapElement-dom-yChannelSelector-attr-expected.txt: Added.
- platform/qt/svg/dynamic-updates/SVGFEDisplacementMapElement-svgdom-in-prop-expected.png: Added.
- platform/qt/svg/dynamic-updates/SVGFEDisplacementMapElement-svgdom-in-prop-expected.txt: Added.
- platform/qt/svg/dynamic-updates/SVGFEDisplacementMapElement-svgdom-in2-prop-expected.png: Added.
- platform/qt/svg/dynamic-updates/SVGFEDisplacementMapElement-svgdom-in2-prop-expected.txt: Added.
- platform/qt/svg/dynamic-updates/SVGFEDisplacementMapElement-svgdom-scale-prop-expected.png: Added.
- platform/qt/svg/dynamic-updates/SVGFEDisplacementMapElement-svgdom-scale-prop-expected.txt: Added.
- platform/qt/svg/dynamic-updates/SVGFEDisplacementMapElement-svgdom-xChannelSelector-prop-expected.png: Added.
- platform/qt/svg/dynamic-updates/SVGFEDisplacementMapElement-svgdom-xChannelSelector-prop-expected.txt: Added.
- platform/qt/svg/dynamic-updates/SVGFEDisplacementMapElement-svgdom-yChannelSelector-prop-expected.png: Added.
- platform/qt/svg/dynamic-updates/SVGFEDisplacementMapElement-svgdom-yChannelSelector-prop-expected.txt: Added.
- platform/qt/svg/filters/feImage-animated-transform-on-target-rect-expected.png: Copied from LayoutTests/platform/qt/svg/filters/feImage-multiple-targets-id-change-expected.png.
- platform/qt/svg/filters/feImage-animated-transform-on-target-rect-expected.txt: Added.
- platform/qt/svg/filters/feImage-filterUnits-objectBoundingBox-primitiveUnits-objectBoundingBox-expected.png:
- platform/qt/svg/filters/feImage-filterUnits-userSpaceOnUse-primitiveUnits-objectBoundingBox-expected.png:
- platform/qt/svg/filters/feImage-filterUnits-userSpaceOnUse-primitiveUnits-userSpaceOnUse-expected.png:
- platform/qt/svg/filters/feImage-late-indirect-update-expected.png:
- platform/qt/svg/filters/feImage-late-indirect-update-expected.txt: Added.
- platform/qt/svg/filters/feImage-multiple-targets-id-change-expected.png:
- platform/qt/svg/filters/feImage-multiple-targets-id-change-expected.txt: Added.
- platform/qt/svg/filters/feImage-position-expected.png:
- platform/qt/svg/filters/feImage-position-expected.txt: Added.
- platform/qt/svg/filters/feImage-preserveAspectratio-expected.png:
- platform/qt/svg/filters/feImage-preserveAspectratio-expected.txt: Added.
- platform/qt/svg/filters/feImage-reference-invalidation-expected.png:
- platform/qt/svg/filters/feImage-reference-invalidation-expected.txt: Added.
- platform/qt/svg/filters/feImage-subregions-expected.png:
- platform/qt/svg/filters/feImage-subregions-expected.txt: Added.
- platform/qt/svg/filters/feImage-subregions-preseveAspectRatio-none-expected.png:
- platform/qt/svg/filters/feImage-subregions-preseveAspectRatio-none-expected.txt: Added.
- platform/qt/svg/filters/feImage-subregions-preseveAspectRatio-none-with-viewBox-expected.png:
- platform/qt/svg/filters/feImage-subregions-preseveAspectRatio-none-with-viewBox-expected.txt: Added.
- platform/qt/svg/filters/feImage-target-add-to-document-expected.png:
- platform/qt/svg/filters/feImage-target-add-to-document-expected.txt: Added.
- platform/qt/svg/filters/feImage-target-attribute-change-expected.png:
- platform/qt/svg/filters/feImage-target-attribute-change-expected.txt: Added.
- platform/qt/svg/filters/feImage-target-attribute-change-with-use-indirection-2-expected.png:
- platform/qt/svg/filters/feImage-target-attribute-change-with-use-indirection-2-expected.txt: Added.
- platform/qt/svg/filters/feImage-target-attribute-change-with-use-indirection-expected.png:
- platform/qt/svg/filters/feImage-target-attribute-change-with-use-indirection-expected.txt: Added.
- platform/qt/svg/filters/feImage-target-changes-id-expected.png:
- platform/qt/svg/filters/feImage-target-changes-id-expected.txt: Added.
- platform/qt/svg/filters/feImage-target-id-change-expected.png:
- platform/qt/svg/filters/feImage-target-id-change-expected.txt: Added.
- platform/qt/svg/filters/feImage-target-inline-style-change-expected.png:
- platform/qt/svg/filters/feImage-target-inline-style-change-expected.txt: Added.
- platform/qt/svg/filters/feImage-target-property-change-expected.png:
- platform/qt/svg/filters/feImage-target-property-change-expected.txt: Added.
- platform/qt/svg/filters/feImage-target-reappend-to-document-expected.png:
- platform/qt/svg/filters/feImage-target-reappend-to-document-expected.txt: Added.
- platform/qt/svg/filters/feImage-target-remove-from-document-expected.png:
- platform/qt/svg/filters/feImage-target-remove-from-document-expected.txt: Added.
- platform/qt/svg/filters/feImage-target-style-change-expected.png:
- platform/qt/svg/filters/feImage-target-style-change-expected.txt: Added.
- 00:24 Changeset [150505] by
-
[EFL] accessibility/label-for-control-hittest.html is failing
https://bugs.webkit.org/show_bug.cgi?id=112029
Unreviewed EFL gardening.
accessibility/label-for-control-hittest.html passes after r150439.
Patch by Krzysztof Czech <k.czech@samsung.com> on 2013-05-22
- platform/efl-wk2/TestExpectations:
- 00:07 Changeset [150504] by
-
[Qt][WK2] Unreviewed buildfix after r150484.
- UIProcess/Plugins/qt/PluginProcessProxyQt.cpp:
(WebKit::PluginProcessProxy::platformGetLaunchOptions):
05/21/13:
- 23:26 Changeset [150503] by
-
Source/WebCore: Update the background blending implementation to match the changes done
in the spec. This implies that background color blending is removed and
the background layers won't blend with the content behind the element
any more.
https://bugs.webkit.org/show_bug.cgi?id=116212
Patch by Mihai Tica <mitica@adobe.com> on 2013-05-21
Reviewed by Dean Jackson.
- rendering/RenderBox.cpp:
(WebCore::RenderBox::paintFillLayers):
- rendering/RenderBoxModelObject.cpp:
(WebCore::RenderBoxModelObject::paintFillLayerExtended):
LayoutTests: Updating element background blending to conform to the changes in the spec.
This implies that the background layers won't longer blend with the content
behind the element.
Removing the background color tests, as background color blending is no longer
a valid use case, also modified the current background image tests.
https://bugs.webkit.org/show_bug.cgi?id=116212
Patch by Mihai Tica <mitica@adobe.com> on 2013-05-21
Reviewed by Dean Jackson.
- css3/compositing/effect-background-blend-mode-color-expected.txt: Removed.
- css3/compositing/effect-background-blend-mode-color.html: Removed.
- css3/compositing/effect-background-blend-mode-color2-expected.txt: Removed.
- css3/compositing/effect-background-blend-mode-color2.html: Removed.
- css3/compositing/effect-background-blend-mode-stacking-expected.png: Removed.
- css3/compositing/effect-background-blend-mode-stacking.html:
- platform/mac/css3/compositing/effect-background-blend-mode-color-expected.png: Removed.
- platform/mac/css3/compositing/effect-background-blend-mode-color2-expected.png: Removed.
- platform/mac/css3/compositing/effect-background-blend-mode-expected.png:
- platform/mac/css3/compositing/effect-background-blend-mode-stacking-expected.png:
- 23:18 Changeset [150502] by
-
[EFL] Cleanup TestExpectations to pass --lint-test-files
https://bugs.webkit.org/show_bug.cgi?id=116590
Unreviewed EFL gardening.
http/tests/websocket/tests/hybi/close-on-navigate-new-location.html was skipped in r150497.
Patch by Seokju Kwon <seokju.kwon@gmail.com> on 2013-05-21
- platform/efl/TestExpectations: Remove duplicate entry.
- 23:09 Changeset [150501] by
-
Web Inspector: [Mac] inspector/geolocation-error.html is flaky
https://bugs.webkit.org/show_bug.cgi?id=116592
Add a flaky test expectation.
- platform/mac/TestExpectations:
- 22:49 Changeset [150500] by
-
REGRESSION(r150491): WebKit2.CloseThenTerminate asserting below WebContext::disconnectProcess().
<http://webkit.org/b/116588>
Reviewed by Sam Weinig.
Do an early return from WebProcessProxy::requestTermination() if there is no child process
connection to terminate.
- UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::requestTermination):
- 20:07 Changeset [150499] by
-
[Windows, curl] WebDownloadCurl.cpp should not include SystemTime.h
https://bugs.webkit.org/show_bug.cgi?id=116584
Patch by Mark Salisbury <mark.salisbury@hp.com> on 2013-05-21
Reviewed by Brent Fulgham.
SystemTime header and implementation are no longer used and were removed
recently - http://trac.webkit.org/changeset/150216.
- WebDownloadCurl.cpp:
- 19:36 Changeset [150498] by
-
Use-after-free in DOMSelection::containsNode
https://bugs.webkit.org/show_bug.cgi?id=116468
Reviewed by Andreas Kling.
Source/WebCore:
Retain the node pointer. Also bail out early if the node was not in the document
since Range::compareBoundaryPoints sets ec to WRONG_DOCUMENT_ERR otherwise.
Test: editing/selection/contains-node-crash.html
- page/DOMSelection.cpp:
(WebCore::DOMSelection::containsNode):
- page/DOMSelection.h:
(DOMSelection):
LayoutTests:
Add a regression test from https://chromium.googlesource.com/chromium/blink/+/40bb8089352b15dd034641b4c131111cd79b44f1.
- editing/selection/contains-node-crash-expected.txt: Added.
- editing/selection/contains-node-crash.html: Added.
- 19:35 Changeset [150497] by
-
Unreviewed EFL gardening.
Skip a test which has problem related to ecore library.
- platform/efl/TestExpectations:
- 19:13 Changeset [150496] by
-
Web Inspector: InspectorFrontendHost::loadResourceSynchronously() builds ASCII-only results
https://bugs.webkit.org/show_bug.cgi?id=116569
Merge: https://chromium.googlesource.com/chromium/blink/+/7f1cfecfe34f4dbc6ce1ef0250b13f2058128eeb
Patch by Joseph Pecoraro <pecoraro@apple.com> on 2013-05-21
Reviewed by Timothy Hatcher.
Source/WebCore:
Test: inspector/load-resource-synchronously-utf8.html
- inspector/InspectorFrontendHost.cpp:
(WebCore::InspectorFrontendHost::loadResourceSynchronously):
LayoutTests:
- inspector/load-resource-synchronously-utf8-expected.txt: Added.
- inspector/load-resource-synchronously-utf8.html: Added.
- 19:13 Changeset [150495] by
-
<rdar://problem/13915556> Sluggish scrolling in web content
Reviewed by Gavin Barraclough.
- Shared/mac/ChildProcessMac.mm:
(WebKit::ChildProcess::setProcessSuppressionEnabled): When process suppression is disabled, low latency is important.
- 18:47 Changeset [150494] by
-
Fix build break after r150484
Unreviewed, fix build break on EFL, GTK.
- UIProcess/Plugins/PluginInfoStore.cpp:
(WebKit::PluginInfoStore::defaultLoadPolicyForPlugin):
- UIProcess/Plugins/unix/PluginProcessProxyUnix.cpp:
(WebKit::PluginProcessProxy::platformGetLaunchOptions):
- 17:25 Changeset [150493] by
-
REGRESSION(r150369): WebKit2WillLoadTest tests are failing
https://bugs.webkit.org/show_bug.cgi?id=116585
Reviewed by Alexey Proskuryakov.
willLoadURLRequest and willLoadDataRequest require the API version 6.
- TestWebKitAPI/Tests/WebKit2/WillLoad_Bundle.cpp:
- 17:07 Changeset [150492] by
-
Check in the test and result I forgot to 'git add' before committing r150490.
- http/tests/security/cross-origin-session-storage-third-party-blocked-expected.txt: Added.
- http/tests/security/cross-origin-session-storage-third-party-blocked.html: Added.
- 17:07 Changeset [150491] by
-
PPT: Closing tab that is hung or chewing 100% CPU leaves abandoned WebProcess.
<http://webkit.org/b/116464>
<rdar://problem/10103795>
Reviewed by Anders Carlsson.
Simplify the web process sacrifice ritual. Instead of murderizing the process when its last tab closes,
just shut down the CoreIPC connection from the UI process. The web process, upon noticing this,
will start a 10 second watchdog timer, and if it hasn't shut down by the time it fires, well, it'll shut down.
(WebKit::WebProcessProxy::removeWebPage):
When the last page is removed from a WebProcessProxy, disconnect the process so that it'll die off
eventually, even if the WebPage::Close message never gets properly handled.
- Shared/ChildProcessProxy.h:
- Shared/ChildProcessProxy.cpp:
(WebKit::ChildProcessProxy::abortProcessLaunchIfNeeded):
Added a helper to abort any in-progress launch of a new web process. Otherwise we'll have a race
where the exiting web process may get re-used for a new tab.
- UIProcess/WebProcessProxy.h:
- UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::WebProcessProxy):
(WebKit::WebProcessProxy::enableSuddenTermination):
(WebKit::WebProcessProxy::disableSuddenTermination):
Remove WebProcessProxy's sudden termination counter. The web process will do what it needs in
response to the WebPage::Close message. We still forward these calls to NSProcessInfo on Mac though.
- 17:03 Changeset [150490] by
-
Allow session storage for third-party origins even if third-party data access is blocked.
https://bugs.webkit.org/show_bug.cgi?id=116532
Reviewed by Anders Carlsson.
Source/WebCore:
In http://trac.webkit.org/changeset/149326 we relaxed our policy for
localStorage when third-party data blocking is enabled to create a
transient, session-scoped storage area for the third-party origin
instead of throwing an exception. Since this gives essentially the same
behavior as sessionStorage, we might as well also allow sessionStorage
for third-party origins.
Test: http/tests/security/cross-origin-session-storage-third-party-blocked.html
- WebCore.exp.in: Updated the symbol for
SecurityOrigin::canAccessStorage.
- page/DOMWindow.cpp:
(WebCore::DOMWindow::sessionStorage): Called canAccessSessionStorage()
rather than canAccessLocalStorage().
- page/SecurityOrigin.cpp:
(WebCore::SecurityOrigin::canAccessStorage): Passed in an enum
indicating if access should be allowed from third-party origins.
- page/SecurityOrigin.h:
(WebCore::SecurityOrigin::canAccessSessionStorage): Called
canAccessStorage() with AlwaysAllowFromThirdParty.
(SecurityOrigin):
LayoutTests:
- http/tests/security/cross-origin-session-storage-third-party-blocked-expected.txt: Renamed from LayoutTests/http/tests/security/cross-origin-session-storage-expected.txt.
- http/tests/security/cross-origin-session-storage-third-party-blocked.html: Renamed from LayoutTests/http/tests/security/cross-origin-session-storage.html.
- 16:53 Changeset [150489] by
-
[Windows] Unreviewed speculative fix for test-bots.
Add export declaration for WTFInvokeCrashHook to avoid runtime
load error on test bots.
- JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreExports.def:
- JavaScriptCore.vcxproj/JavaScriptCoreExportGenerator/JavaScriptCoreExports.def.in:
- 16:34 Changeset [150488] by
-
Web Inspector: inspector/styles/import-pseudoclass-crash.html intermittently hits an assertion in HTMLLinkElement::startLoadingDynamicSheet
https://bugs.webkit.org/show_bug.cgi?id=116582
Add a flaky crash expectation.
- platform/mac/TestExpectations:
- 16:09 Changeset [150487] by
-
Update Mountain Lion test expectations per bug 116477.
- platform/mac/TestExpectations:
- 16:02 Changeset [150486] by
-
Fix build.
- WebProcess/Plugins/PluginProcessConnectionManager.cpp:
(WebKit::PluginProcessConnectionManager::getPluginProcessConnection):
Stop using lambdas :(
- 16:01 Changeset [150485] by
-
[EFL] Add missing #if guards on drag support methods
https://bugs.webkit.org/show_bug.cgi?id=116574
Reviewed by Christophe Dumez.
- platform/efl/PasteboardEfl.cpp:
(WebCore):
- 15:54 Changeset [150484] by
-
Provide an API to run sandboxed plug-ins outside of their sandbox
https://bugs.webkit.org/show_bug.cgi?id=116578
Reviewed by Andreas Kling.
- Platform/CoreIPC/HandleMessage.h:
(CoreIPC::callMemberFunction):
Add new overload.
- PluginProcess/EntryPoint/mac/LegacyProcess/PluginProcessMain.mm:
(WebKit::PluginProcessMainDelegate::getExtraInitializationData):
Handle "disable-sandbox".
- PluginProcess/EntryPoint/mac/XPCService/PluginServiceEntryPoint.mm:
(WebKit::PluginServiceInitializerDelegate::getExtraInitializationData):
Handle "disable-sandbox".
- PluginProcess/PluginProcess.h:
Move Type to PluginProcessAttributes and remove hash traits for type.
- PluginProcess/mac/PluginProcessMac.mm:
(WebKit::PluginProcess::platformInitializePluginProcess):
Update for new type name.
(WebKit::PluginProcess::initializeSandbox):
Return early if the sandbox is disabled.
- Shared/Plugins/Netscape/PluginInformation.cpp:
(WebKit::getPluginModuleInformation):
PluginInfoStore::policyForPlugin has been renamed to PluginInfoStore::defaultLoadPolicyForPlugin.
- Shared/Plugins/PluginModuleInfo.h:
Add PluginModuleLoadUnsandboxed.
- Shared/Plugins/PluginProcessAttributes.h: Added.
Add new header with various plug-in process attributes.
- Shared/Plugins/PluginProcessCreationParameters.h:
Change type enum.
- UIProcess/API/C/WKAPICast.h:
(WebKit::toWKPluginLoadPolicy):
(WebKit::toPluginModuleLoadPolicy):
Handle the unsandboxed enum.
- UIProcess/API/C/WKPage.h:
Add kWKPluginLoadPolicyLoadUnsandboxed.
- UIProcess/Plugins/PluginInfoStore.h:
Rename policyForPlugin to defaultLoadPolicyForPlugin.
- UIProcess/Plugins/PluginProcessManager.cpp:
(WebKit::PluginProcessManager::pluginProcessToken):
Add new helper function for getting a token given plug-in attributes.
(WebKit::PluginProcessManager::getPluginProcessConnection):
This now takes a pluginProcessToken.
(WebKit::PluginProcessManager::getSitesWithData):
Make sure to call pluginProcessToken.
(WebKit::PluginProcessManager::clearSiteData):
Ditto.
(WebKit::PluginProcessManager::getOrCreatePluginProcess):
This now takes a token.
- UIProcess/Plugins/PluginProcessProxy.cpp:
(WebKit::PluginProcessProxy::create):
(WebKit::PluginProcessProxy::PluginProcessProxy):
Pass the PluginProcessAttributes as well as the plug-in process token.
(WebKit::PluginProcessProxy::getLaunchOptions):
Pass the plug-in process attributes to platformGetLaunchOptions.
(WebKit::PluginProcessProxy::didClose):
Pass the token to PluginProcessCrashed.
(WebKit::PluginProcessProxy::didFinishLaunching):
Update for type change.
- UIProcess/Plugins/mac/PluginInfoStoreMac.mm:
(WebKit::shouldBlockPlugin):
(WebKit::PluginInfoStore::defaultLoadPolicyForPlugin):
Rename policyForPlugin to defaultLoadPolicyForPlugin.
- UIProcess/Plugins/mac/PluginProcessProxyMac.mm:
(WebKit::PluginProcessProxy::platformGetLaunchOptions):
Set the "disable-sandbox" data if needed.
(WebKit::PluginProcessProxy::platformInitializePluginProcess):
(WebKit::PluginProcessProxy::openPluginPreferencePane):
Get the module info from m_pluginProcessAttributes.
- UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::findPlugin):
This now returns a token.
- UIProcess/WebPageProxy.messages.in:
FindPlugin now returns a token instead of the path.
- UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::getPluginProcessConnection):
Pass the token to the plug-in process manager.
- UIProcess/WebProcessProxy.messages.in:
This now takes a token.
- WebKit2.xcodeproj/project.pbxproj:
Add new files.
- WebProcess/Plugins/PluginProcessConnection.cpp:
(WebKit::PluginProcessConnection::PluginProcessConnection):
This now takes a token.
- WebProcess/Plugins/PluginProcessConnection.h:
(WebKit::PluginProcessConnection::create):
(WebKit::PluginProcessConnection::pluginProcessToken):
Create this with a token instead of a path and process type.
- WebProcess/Plugins/PluginProcessConnectionManager.cpp:
(WebKit::PluginProcessConnectionManager::getPluginProcessConnection):
This now takes a token.
(WebKit::PluginProcessConnectionManager::pluginProcessCrashed):
Ditto.
- WebProcess/Plugins/PluginProcessConnectionManager.messages.in:
PluginProcessCrashed now takes a token.
- WebProcess/Plugins/PluginProxy.cpp:
(WebKit::PluginProxy::create):
(WebKit::PluginProxy::PluginProxy):
These now take tokens.
(WebKit::PluginProxy::initialize):
Pass the token when getting the connection.
- WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::createPlugin):
Get the token from the UI process and pass it to the PluginProxy.
(WebKit::WebPage::canPluginHandleResponse):
Update now that FindPlugin returns a token.
- 15:48 Changeset [150483] by
-
Remove stub HTMLContentElement
https://bugs.webkit.org/show_bug.cgi?id=116580
Reviewed by Andreas Kling.
Inherit DetailsContentElement and DetailsSummaryElement directly from InsertionPoint instead.
- html/HTMLDetailsElement.cpp:
(WebCore::DetailsContentElement::DetailsContentElement):
(WebCore::DetailsSummaryElement::DetailsSummaryElement):
- html/HTMLSummaryElement.cpp:
(WebCore::SummaryContentElement::SummaryContentElement):
- html/shadow/HTMLContentElement.cpp:
(WebCore::HTMLContentElement::contentTagName):
- html/shadow/HTMLContentElement.h:
(WebCore::isHTMLContentElement):
- html/shadow/InsertionPoint.cpp:
(WebCore::InsertionPoint::setResetStyleInheritance):
- html/shadow/InsertionPoint.h:
Also remove some leftover selector code.
(WebCore::InsertionPoint::insertionPointType):
- 15:20 Changeset [150482] by
-
Robustify repaint of previous caret node when moving FrameSelection.
<http://webkit.org/b/116558>
Reviewed by Ryosuke Niwa.
As a follow-up to r150396, also repaint the caret rects if the previous node is contenteditable.
- editing/FrameSelection.cpp:
(WebCore::FrameSelection::recomputeCaretRect):
- 15:06 Changeset [150481] by
-
Unreviewed build fix after r150469. We need to bail out here when DOM is mutated by an event listener such that
the split text nodes no longer exist right next to each other.
- editing/InsertParagraphSeparatorCommand.cpp:
(WebCore::InsertParagraphSeparatorCommand::doApply):
- 14:38 Changeset [150480] by
-
Remove ScopeContentDistribution
https://bugs.webkit.org/show_bug.cgi?id=116576
Reviewed by Andreas Kling.
With maximum one ShadowRoot per Element this can be smashed into ContentDistributor.
- dom/ShadowRoot.cpp:
(WebCore):
- dom/ShadowRoot.h:
(WebCore):
- html/shadow/ContentDistributor.cpp:
(WebCore::ContentDistributor::ContentDistributor):
(WebCore):
(WebCore::ContentDistributor::~ContentDistributor):
(WebCore::ContentDistributor::invalidateInsertionPointList):
(WebCore::ContentDistributor::ensureInsertionPointList):
(WebCore::ContentDistributor::distribute):
(WebCore::ContentDistributor::invalidate):
- html/shadow/ContentDistributor.h:
(WebCore):
(ContentDistributor):
- html/shadow/InsertionPoint.cpp:
(WebCore::InsertionPoint::InsertionPoint):
(WebCore::InsertionPoint::insertedInto):
(WebCore::InsertionPoint::removedFrom):
Simplify insertion point list invalidation.
- html/shadow/InsertionPoint.h:
- 14:28 Changeset [150479] by
-
REGRESSION(r150393): editing/inserting/typing-at-end-of-line.html fails
https://bugs.webkit.org/show_bug.cgi?id=116516
Reviewed by Alexey Proskuryakov.
Reset states in DumpRenderTree. The fix for WebKitTestRunner should be implemented in the bug 116577.
- DumpRenderTree/mac/DumpRenderTree.mm:
(createWebViewAndOffscreenWindow): Match the ordering.
(resetWebViewToConsistentStateBeforeTesting): Reset text replacement states.
- 14:04 Changeset [150478] by
-
[CSS Regions][CSS Exclusions] Multiple regions with shape-insides should respect positioned shapes and overflow
https://bugs.webkit.org/show_bug.cgi?id=115001
Reviewed by David Hyatt.
In r150375 I implemented the new overflow behavior for shape-inside on regions, but it covered only the simple cases
when you had only one region. This patch adds support for the new overflow behavior for multiple regions with multiple
positioned shape-insides and overflow.
Source/WebCore:
Tests: fast/regions/shape-inside/shape-inside-on-regions-block-content-overflow-multiple-shapes.html
fast/regions/shape-inside/shape-inside-on-regions-inline-content-overflow-multiple-shapes.html
- rendering/RenderBlockLineLayout.cpp:
(WebCore::RenderBlock::updateLineBoundariesForExclusions): Cover cases when content flows into the next region, update the
line segments for the overflow.
(WebCore::RenderBlock::adjustLogicalLineTopAndLogicalHeightIfNeeded): In flow thread case we don't want to adjust the height
in two places.
LayoutTests:
- fast/regions/shape-inside/shape-inside-on-regions-block-content-overflow-multiple-shapes-expected.html: Added.
- fast/regions/shape-inside/shape-inside-on-regions-block-content-overflow-multiple-shapes.html: Added.
- fast/regions/shape-inside/shape-inside-on-regions-inline-content-overflow-multiple-shapes-expected.html: Added.
- fast/regions/shape-inside/shape-inside-on-regions-inline-content-overflow-multiple-shapes.html: Added.
- 14:00 Changeset [150477] by
-
[Windows] Reenable storage tests
https://bugs.webkit.org/show_bug.cgi?id=116559
Unreviewed update to TestExpectations:
- Storage features are present on Windows, so test them.
- Disable (for now) the four websql tests that fail.
- platform/win/TestExpectations:
- 13:52 Changeset [150476] by
-
Unreviewed gardening; rebaseline and re-enable media/track/track-cue-rendering-snap-to-lines-not-set.html
- media/track/track-cue-rendering-snap-to-lines-not-set-expected.txt:
- platform/mac/TestExpectations:
- platform/mac/media/track/track-cue-rendering-snap-to-lines-not-set-expected.txt: Removed.
- 13:42 Changeset [150475] by
-
Unreviewed, rolling out r150425.
http://trac.webkit.org/changeset/150425
https://bugs.webkit.org/show_bug.cgi?id=116573
Not the right fix for WebKit2 (Requested by rniwa on #webkit).
Patch by Commit Queue <commit-queue@webkit.org> on 2013-05-21
- testing/Internals.cpp:
(WebCore::Internals::resetToConsistentState):
- 13:38 Changeset [150474] by
-
<rdar://problem/13929704> [WK2] WebKitTestRunner always crashes on exit
https://bugs.webkit.org/show_bug.cgi?id=116568
Reviewed by Anders Carlsson.
- WebKitTestRunner/WebNotificationProvider.cpp: (WTR::WebNotificationProvider::~WebNotificationProvider):
- WebKitTestRunner/WebNotificationProvider.h: When the provider is destroyed, it should no longer be registered with the manager.
- WebKitTestRunner/GeolocationProviderMock.cpp: (WTR::GeolocationProviderMock::~GeolocationProviderMock):
- WebKitTestRunner/GeolocationProviderMock.h: I wasn't seeing a crash here, but this code has the same issue as notification provider, so applying the same fix.
- WebKitTestRunner/TestController.cpp: (WTR::TestController::~TestController): Close icon database. I started seeing ASSERT(!isOpen()) in WebCore IconDatabase destructor having fixed the above.
- 13:32 Changeset [150473] by
-
Update test expectations after r150472 for the bug 116125.
- platform/wk2/TestExpectations:
- 13:25 WebKitGTK/2.0.x edited by
- (diff)
- 13:21 Changeset [150472] by
-
Unreviewed, rolling out r150386 and r150397.
http://trac.webkit.org/changeset/150386
http://trac.webkit.org/changeset/150397
https://bugs.webkit.org/show_bug.cgi?id=116572
Broke a test; Alex is going to try again later. (Requested by
thorton on #webkit).
Tools:
- WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::pathSuitableForTestResult):
(WTR::dumpRequestDescriptionSuitableForTestResult):
(WTR::dumpResponseDescriptionSuitableForTestResult):
(WTR::InjectedBundlePage::willPerformClientRedirectForFrame):
(WTR::InjectedBundlePage::didInitiateLoadForResource):
(WTR::InjectedBundlePage::willSendRequestForFrame):
(WTR::InjectedBundlePage::didReceiveResponseForResource):
- WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
(InjectedBundlePage):
LayoutTests:
- platform/wk2/TestExpectations:
- 13:02 Changeset [150471] by
-
Unreviewed gardening; rebaseline and re-enable media/track/track-cue-rendering-snap-to-lines-not-set.html
https://bugs.webkit.org/show_bug.cgi?id=116567
- media/track/track-cue-rendering-snap-to-lines-not-set-expected.txt:
- platform/mac/TestExpectations:
- platform/mac/media/track/track-cue-rendering-snap-to-lines-not-set-expected.txt: Removed.
- 12:43 Changeset [150470] by
-
fourthTier: FTL shouldn't use FastISel and Small code model should be turned off for now
https://bugs.webkit.org/show_bug.cgi?id=115998
Reviewed by Oliver Hunt.
This switches off FastISel and makes it possible to turn off Small code model.
- ftl/FTLCompile.cpp:
(JSC::FTL::compile):
- runtime/Options.h:
(JSC):
- 12:34 Changeset [150469] by
-
dispatchEvent call can execute javascript and blow away endRoot from underneath
https://bugs.webkit.org/show_bug.cgi?id=116483
Source/WebCore:
Reviewed by Andreas Kling.
Merge https://chromium.googlesource.com/chromium/blink/+/798cba0af9b2aff21e475e2e08ea3ca5e97dfc2c.
Test: editing/undo/undo-after-event-edited.html
- editing/Editor.cpp:
(WebCore::dispatchEditableContentChangedEvents):
LayoutTests:
Reviewed by Andreas Kling.
Add a regression test.
- editing/undo/undo-after-event-edited-expected.txt: Added.
- editing/undo/undo-after-event-edited.html: Added.
- 12:32 Changeset [150468] by
-
media/track/track-cue-rendering-snap-to-lines-not-set.html is failing
https://bugs.webkit.org/show_bug.cgi?id=116567
Add a failing test expectation to the test added in r150452.
Patch by Ryosuke Niwa <rniwa@webkit.org> on 2013-05-21
- platform/mac/TestExpectations:
- platform/mac/media/track/track-cue-rendering-snap-to-lines-not-set-expected.txt: Added.
- 12:28 Changeset [150467] by
-
media/track/track-cue-rendering-snap-to-lines-not-set.html is failing
https://bugs.webkit.org/show_bug.cgi?id=116567
Add a failing test expectation to the test added in r150452.
- platform/mac/TestExpectations:
- platform/mac/media/track/track-cue-rendering-snap-to-lines-not-set-expected.txt: Added.
- 12:26 Changeset [150466] by
-
fourthTier: should use ConcurrentJITLock[er] directly and not through typedef
https://bugs.webkit.org/show_bug.cgi?id=116561
Rubber stamped by Geoffrey Garen.
- JavaScriptCore.xcodeproj/project.pbxproj:
- bytecode/ArrayProfile.cpp:
(JSC::ArrayProfile::computeUpdatedPrediction):
(JSC::ArrayProfile::briefDescription):
- bytecode/ArrayProfile.h:
(ArrayProfile):
(JSC::ArrayProfile::expectedStructure):
(JSC::ArrayProfile::structureIsPolymorphic):
(JSC::ArrayProfile::hasDefiniteStructure):
(JSC::ArrayProfile::observedArrayModes):
(JSC::ArrayProfile::mayInterceptIndexedAccesses):
(JSC::ArrayProfile::mayStoreToHole):
(JSC::ArrayProfile::outOfBounds):
(JSC::ArrayProfile::usesOriginalArrayStructures):
- bytecode/CallLinkStatus.cpp:
(JSC::CallLinkStatus::computeFor):
- bytecode/CodeBlock.cpp:
(JSC::CodeBlock::dumpValueProfiling):
(JSC::CodeBlock::dumpArrayProfiling):
(JSC::CodeBlock::updateAllPredictionsAndCountLiveness):
(JSC::CodeBlock::updateAllArrayPredictions):
(JSC::CodeBlock::nameForRegister):
- bytecode/CodeBlock.h:
(JSC::CodeBlock::valueProfilePredictionForBytecodeOffset):
(CodeBlock):
- bytecode/CodeBlockLock.h: Removed.
- bytecode/GetByIdStatus.cpp:
(JSC::GetByIdStatus::computeFor):
- bytecode/LazyOperandValueProfile.cpp:
(JSC::CompressedLazyOperandValueProfileHolder::computeUpdatedPredictions):
(JSC::CompressedLazyOperandValueProfileHolder::add):
(JSC::LazyOperandValueProfileParser::initialize):
(JSC::LazyOperandValueProfileParser::prediction):
- bytecode/LazyOperandValueProfile.h:
(CompressedLazyOperandValueProfileHolder):
(LazyOperandValueProfileParser):
- bytecode/MethodOfGettingAValueProfile.cpp:
(JSC::MethodOfGettingAValueProfile::getSpecFailBucket):
- bytecode/PutByIdStatus.cpp:
(JSC::PutByIdStatus::computeFor):
- bytecode/ResolveGlobalStatus.cpp:
(JSC::ResolveGlobalStatus::computeFor):
- bytecode/ValueProfile.h:
(JSC::ValueProfileBase::briefDescription):
(JSC::ValueProfileBase::computeUpdatedPrediction):
- bytecompiler/BytecodeGenerator.cpp:
(JSC::BytecodeGenerator::addVar):
- dfg/DFGArrayMode.cpp:
(JSC::DFG::ArrayMode::fromObserved):
- dfg/DFGArrayMode.h:
(ArrayMode):
(JSC::DFG::ArrayMode::withProfile):
- dfg/DFGByteCodeParser.cpp:
(JSC::DFG::ByteCodeParser::injectLazyOperandSpeculation):
(JSC::DFG::ByteCodeParser::getPredictionWithoutOSRExit):
(JSC::DFG::ByteCodeParser::getArrayMode):
(JSC::DFG::ByteCodeParser::getArrayModeAndEmitChecks):
(JSC::DFG::ByteCodeParser::parseResolveOperations):
(JSC::DFG::ByteCodeParser::parseBlock):
(JSC::DFG::ByteCodeParser::InlineStackEntry::InlineStackEntry):
- dfg/DFGFixupPhase.cpp:
(JSC::DFG::FixupPhase::fixupNode):
- dfg/DFGPredictionInjectionPhase.cpp:
(JSC::DFG::PredictionInjectionPhase::run):
- jit/JITInlines.h:
(JSC::JIT::chooseArrayMode):
- jit/JITStubs.cpp:
(JSC::tryCachePutByID):
(JSC::tryCacheGetByID):
(JSC::DEFINE_STUB_FUNCTION):
(JSC::lazyLinkFor):
- llint/LLIntSlowPaths.cpp:
(JSC::LLInt::LLINT_SLOW_PATH_DECL):
(JSC::LLInt::setUpCall):
- profiler/ProfilerBytecodeSequence.cpp:
(JSC::Profiler::BytecodeSequence::BytecodeSequence):
- runtime/Executable.cpp:
(JSC::ProgramExecutable::addGlobalVar):
- runtime/JSActivation.cpp:
(JSC::JSActivation::getOwnNonIndexPropertyNames):
(JSC::JSActivation::symbolTablePutWithAttributes):
- runtime/JSScope.cpp:
(JSC::JSScope::resolveContainingScopeInternal):
(JSC::JSScope::resolvePut):
- runtime/JSSegmentedVariableObject.cpp:
(JSC::JSSegmentedVariableObject::findRegisterIndex):
(JSC::JSSegmentedVariableObject::addRegisters):
- runtime/JSSegmentedVariableObject.h:
(JSSegmentedVariableObject):
- runtime/JSSymbolTableObject.cpp:
(JSC::JSSymbolTableObject::getOwnNonIndexPropertyNames):
- runtime/JSSymbolTableObject.h:
(JSC::symbolTableGet):
(JSC::symbolTablePut):
(JSC::symbolTablePutWithAttributes):
- runtime/Structure.cpp:
(JSC::Structure::materializePropertyMap):
(JSC::Structure::addPropertyTransitionToExistingStructureConcurrently):
(JSC::Structure::addPropertyTransition):
(JSC::Structure::takePropertyTableOrCloneIfPinned):
(JSC::Structure::nonPropertyTransition):
(JSC::Structure::putSpecificValue):
(JSC::Structure::remove):
(JSC::Structure::createPropertyMap):
- runtime/Structure.h:
(Structure):
- runtime/SymbolTable.h:
(SymbolTable):
(JSC::SymbolTable::find):
(JSC::SymbolTable::get):
(JSC::SymbolTable::inlineGet):
(JSC::SymbolTable::begin):
(JSC::SymbolTable::end):
(JSC::SymbolTable::size):
(JSC::SymbolTable::add):
(JSC::SymbolTable::set):
(JSC::SymbolTable::contains):
- 12:10 Changeset [150465] by
-
fourthTier: DFG should be able to run on a separate thread
https://bugs.webkit.org/show_bug.cgi?id=112839
Source/JavaScriptCore:
Reviewed by Geoffrey Garen.
This is the final bit of concurrent JITing. The idea is that there is a
single global worklist, and a single global thread, that does all
optimizing compilation. This is the DFG::Worklist. It contains a queue of
DFG::Plans, and a map from CodeBlock* (the baseline code block we're
trying to optimize) to DFG::Plan. If the DFGDriver tries to concurrently
compile something, it puts the Plan on the Worklist. The Worklist's
thread will compile that Plan eventually, and when it's done, it will
signal its completion by (1) notifying anyone waiting for the Worklist to
be done, and (2) forcing the CodeBlock::m_jitExecuteCounter to take slow
path. The next Baseline JIT cti_optimize call will then install all ready
(i.e. compiled) Plans for that VM. Note that (1) is only for the GC and
VM shutdown, which will want to ensure that there aren't any outstanding
async compilations before proceeding. They do so by simply waiting for
all of the plans for the current VM to complete. (2) is the actual way
that code typically gets installed.
This is all very racy by design. For example, just as we try to force the
execute counter to take slow path, the main thread may be setting the
execute counter to some other value. The main thread must set it to
another value because (a) JIT code is constantly incrementing the counter
in a racy way, (b) the cti_optimize slow path will set it to some
large-ish negative value to ensure that cti_optimize isn't called
repeatedly, and (c) OSR exits from previously jettisoned code blocks may
still want to reset the counter values. This "race" is made benign, by
ensuring that while there is an asynchronous compilation, we at worse set
the counter to optimizeAfterWarmUp and never to deferIndefinitely. Hence
if the race happens then the worst case is that we wait another ~1000
counts before installing the optimized code. Another defense is that if
any CodeBlock calls into cti_optimize, then it will check for all ready
plans for the VM - so even if a code block has to wait another ~1000
executions before it calls cti_optimize to do the installation, it may
actually end up being installed sooner because a different code block had
called cti_optimize, potentially for an unrelated reason.
Special care is taken to ensure that installing plans informs the GC
about the increased memory usage, but also ensures that we don't recurse
infinitely - since at start of GC we try to install outstanding plans.
This is done by introducing a new GC deferral mechanism (the DeferGC
block-scoped thingy), which will ensure that GCs don't happen in the
scope but are allowed to happen after. This still leaves the strange
corner case that cti_optimize may install outstanding plans, then GC, and
that GC may jettison the code block that was installed. This, and the
fact that the plan that we took slow path to install could have been a
failed or invalid compile, mean that we have to take special precautions
in cti_optimize.
This patch also fixes a number of small concurrency bugs that I found
when things started running. There are probably more of those bugs still
left to fix. This patch just fixes the ones I know about.
Concurrent compilation is right now only enabled on X86_64 Mac. We need
platforms that are sufficiently CAStastic so that we can do the various
memory fence and CAS tricks that make this safe. We also need a platform
that uses JSVALUE64. And we need pthread_once. So, that pretty much means
just X64_64 for now. Enabling Linux-64_64 should be a breeze, but I'll
leave that up to the Qt and GTK+ ports to do at their discretion.
This is a solid speed-up on SunSpider (8-9%) and V8Spider (16%), our two
main compile-time benchmarks. Most peculiarly, this also appears to
reduce measurement noise, rather than increasing it as you would have
expected. I don't understand that result but I like it anyway. On the
other hand, this is a slight (1%) slow-down on V8v7. I will continue to
investigate this but I think that the results are already good enough
that we should land this as-is. So far, it appears that the slow-down is
due to this breaking the don't-compile-inlineables heuristics. See
investigation in https://bugs.webkit.org/show_bug.cgi?id=116556 and the
bug https://bugs.webkit.org/show_bug.cgi?id=116557.
- JavaScriptCore.xcodeproj/project.pbxproj:
- bytecode/CodeBlock.cpp:
(JSC):
(JSC::CodeBlock::finalizeUnconditionally):
(JSC::CodeBlock::resetStubInternal):
(JSC::CodeBlock::baselineVersion):
(JSC::CodeBlock::hasOptimizedReplacement):
(JSC::CodeBlock::optimizationThresholdScalingFactor):
(JSC::CodeBlock::checkIfOptimizationThresholdReached):
(JSC::CodeBlock::optimizeNextInvocation):
(JSC::CodeBlock::dontOptimizeAnytimeSoon):
(JSC::CodeBlock::optimizeAfterWarmUp):
(JSC::CodeBlock::optimizeAfterLongWarmUp):
(JSC::CodeBlock::optimizeSoon):
(JSC::CodeBlock::forceOptimizationSlowPathConcurrently):
(JSC::CodeBlock::setOptimizationThresholdBasedOnCompilationResult):
(JSC::CodeBlock::updateAllPredictionsAndCountLiveness):
(JSC::CodeBlock::updateAllArrayPredictions):
(JSC::CodeBlock::shouldOptimizeNow):
- bytecode/CodeBlock.h:
(CodeBlock):
(JSC::CodeBlock::jitCompile):
- bytecode/CodeBlockLock.h:
(JSC):
- bytecode/ExecutionCounter.cpp:
(JSC::ExecutionCounter::forceSlowPathConcurrently):
(JSC):
(JSC::ExecutionCounter::setThreshold):
- bytecode/ExecutionCounter.h:
(ExecutionCounter):
- debugger/Debugger.cpp:
(JSC::Debugger::recompileAllJSFunctions):
- dfg/DFGByteCodeParser.cpp:
(JSC::DFG::ByteCodeParser::injectLazyOperandSpeculation):
(JSC::DFG::ByteCodeParser::getArrayMode):
(JSC::DFG::ByteCodeParser::getArrayModeAndEmitChecks):
- dfg/DFGCommon.h:
(JSC::DFG::enableConcurrentJIT):
(DFG):
- dfg/DFGDriver.cpp:
(JSC::DFG::compile):
- dfg/DFGGraph.cpp:
(JSC::DFG::Graph::Graph):
- dfg/DFGGraph.h:
(Graph):
- dfg/DFGOSREntry.cpp:
(JSC::DFG::prepareOSREntry):
- dfg/DFGOperations.cpp:
- dfg/DFGPlan.cpp:
(JSC::DFG::Plan::Plan):
(JSC::DFG::Plan::compileInThread):
(JSC::DFG::Plan::key):
(DFG):
- dfg/DFGPlan.h:
(DFG):
(Plan):
- dfg/DFGWorklist.cpp: Added.
(DFG):
(JSC::DFG::Worklist::Worklist):
(JSC::DFG::Worklist::~Worklist):
(JSC::DFG::Worklist::finishCreation):
(JSC::DFG::Worklist::create):
(JSC::DFG::Worklist::enqueue):
(JSC::DFG::Worklist::compilationState):
(JSC::DFG::Worklist::waitUntilAllPlansForVMAreReady):
(JSC::DFG::Worklist::removeAllReadyPlansForVM):
(JSC::DFG::Worklist::completeAllReadyPlansForVM):
(JSC::DFG::Worklist::completeAllPlansForVM):
(JSC::DFG::Worklist::queueLength):
(JSC::DFG::Worklist::dump):
(JSC::DFG::Worklist::runThread):
(JSC::DFG::Worklist::threadFunction):
(JSC::DFG::initializeGlobalWorklistOnce):
(JSC::DFG::globalWorklist):
- dfg/DFGWorklist.h: Added.
(DFG):
(Worklist):
- heap/CopiedSpaceInlines.h:
(JSC::CopiedSpace::allocateBlock):
- heap/DeferGC.h: Added.
(JSC):
(DeferGC):
(JSC::DeferGC::DeferGC):
(JSC::DeferGC::~DeferGC):
- heap/Heap.cpp:
(JSC::Heap::Heap):
(JSC::Heap::reportExtraMemoryCostSlowCase):
(JSC::Heap::collectAllGarbage):
(JSC::Heap::collect):
(JSC::Heap::collectIfNecessaryOrDefer):
(JSC):
(JSC::Heap::incrementDeferralDepth):
(JSC::Heap::decrementDeferralDepthAndGCIfNeeded):
- heap/Heap.h:
(Heap):
(JSC::Heap::isCollecting):
(JSC):
- heap/MarkedAllocator.cpp:
(JSC::MarkedAllocator::allocateSlowCase):
- jit/JIT.cpp:
(JSC::JIT::privateCompile):
- jit/JIT.h:
- jit/JITStubs.cpp:
(JSC::DEFINE_STUB_FUNCTION):
- llint/LLIntSlowPaths.cpp:
(JSC::LLInt::jitCompileAndSetHeuristics):
(JSC::LLInt::entryOSR):
(JSC::LLInt::LLINT_SLOW_PATH_DECL):
- profiler/ProfilerBytecodes.h:
- runtime/ConcurrentJITLock.h: Added.
(JSC):
- runtime/ExecutionHarness.h:
(JSC::replaceWithDeferredOptimizedCode):
- runtime/JSSegmentedVariableObject.cpp:
(JSC::JSSegmentedVariableObject::findRegisterIndex):
(JSC::JSSegmentedVariableObject::addRegisters):
- runtime/JSSegmentedVariableObject.h:
(JSSegmentedVariableObject):
- runtime/Options.h:
(JSC):
- runtime/Structure.h:
(Structure):
- runtime/StructureInlines.h:
(JSC::Structure::propertyTable):
- runtime/SymbolTable.h:
(SymbolTable):
- runtime/VM.cpp:
(JSC::VM::VM):
(JSC::VM::~VM):
(JSC::VM::prepareToDiscardCode):
(JSC):
(JSC::VM::discardAllCode):
(JSC::VM::releaseExecutableMemory):
- runtime/VM.h:
(DFG):
(VM):
Source/WTF:
Reviewed by Geoffrey Garen.
- wtf/ByteSpinLock.h:
Make it non-copyable. We previously had bugs where we used ByteSpinLock as a locker.
Clearly that's bad.
- wtf/MetaAllocatorHandle.h:
Make it thread-safe ref-counted, since we may now be passing them between the
concurrent JIT thread and the main thread.
- wtf/Vector.h:
(WTF::Vector::takeLast):
I've wanted this method for ages, and now I finally added.
- 12:09 Changeset [150464] by
-
Remove ContentDistribution
https://bugs.webkit.org/show_bug.cgi?id=116527
Reviewed by Andreas Kling.
Resolving distribution on traversal is simpler.
- WebCore.exp.in:
- dom/ComposedShadowTreeWalker.cpp:
(WebCore::ComposedShadowTreeWalker::traverseNode):
(WebCore::ComposedShadowTreeWalker::traverseDistributedNodes):
(WebCore::ComposedShadowTreeWalker::traverseSiblingOrBackToInsertionPoint):
- html/HTMLDetailsElement.cpp:
- html/shadow/ContentDistributor.cpp:
(WebCore::ContentDistributor::distributeSelectionsTo):
- html/shadow/ContentDistributor.h:
- html/shadow/HTMLContentElement.idl:
Remove getDistributedNodes().
- html/shadow/InsertionPoint.cpp:
(WebCore::InsertionPoint::InsertionPoint):
(WebCore::InsertionPoint::attach):
(WebCore::InsertionPoint::detach):
(WebCore):
(WebCore::InsertionPoint::firstDistributed):
(WebCore::InsertionPoint::lastDistributed):
(WebCore::InsertionPoint::nextDistributedTo):
(WebCore::InsertionPoint::previousDistributedTo):
Rename for clarity.
- html/shadow/InsertionPoint.h:
(WebCore::InsertionPoint::hasDistribution):
(WebCore::InsertionPoint::setHasDistribution):
(WebCore::InsertionPoint::clearDistribution):
(WebCore::InsertionPoint::matchTypeFor):
(InsertionPoint):
- 12:03 Changeset [150463] by
-
[Windows] Workers are crashing tests.
https://bugs.webkit.org/show_bug.cgi?id=116560.
Unreviewed change to TestExpectations while investigating.
- platform/win/TestExpectations: Disable worker tests under Windows.
- 12:02 Changeset [150462] by
-
Unreviewed. Add myself as committer in contributors.json.
- 11:59 Changeset [150461] by
-
Fix a python unittest after r149419. We need to support Python 2.6 on Windows.
- Scripts/webkitpy/tool/commands/suggestnominations.py:
(AbstractCommitLogCommand._recent_commit_messages):
- 11:49 Changeset [150460] by
-
[GTK] [CMake] Add support for building WebKit2
https://bugs.webkit.org/show_bug.cgi?id=116372
Reviewed by Gustavo Noronha Silva.
.:
- Source/cmake/FindWebP.cmake: Added.
- Source/cmake/OptionsGTK.cmake: Turn on WebKit2 and the plugin process and also look
for WebP. These missing symbols were hidden up until now. ENABLE_TEXTURE_MAPPER was
also incorrect specified.
Source/WebCore:
- PlatformGTK.cmake: Add missing source files, include directories, and WebP properties.
Source/WebKit2:
- CMakeLists.txt: Add common directories to the common include list and remove UIProcess/PageViewportController.cpp
from the source list since it should only be compiled for WebKitEFL. Instead of taking explicit forwarding-header
dependencies accept a list of "tacked-on" dependencies for WebKit2. This allows GTK+ to specify its extra fake
installed header step.
- PlatformEfl.cmake: Remove common include directories and add UIProcess/PageViewportController.cpp to the source
list. Fix up forwarding header dependencies now.
- PlatformGTK.cmake: Added.
- config.h: First check whether this is a CMake build and then fall back to the autotools header.
Tools:
- CMakeLists.txt: Do not build WebKitTestRunner yet for WebKitGTK+.
- MiniBrowser/gtk/CMakeLists.txt: Added.
- 11:41 Changeset [150459] by
-
[Windows] Unreviewed test expectation change - skip failing tests.
https://bugs.webkit.org/show_bug.cgi?id=116564
- Skip a number of failing tests, and investigate the sources of the
problems. Progress will be tracked on the above bug.
- Correct a typo I made in r150382
- Remove skip entries from https://webkit.org/b/113908; these were
replaced with new tests added in https://webkit.org/b/113946.
- platform/win/TestExpectations:
- 11:21 Changeset [150458] by
-
[CSSRegions] Remove percentages from offsetLeft/offsetTop tests to avoid sub-pixel errors
https://bugs.webkit.org/show_bug.cgi?id=116544
Patch by Radu Stavila <stavila@adobe.com> on 2013-05-21
Reviewed by Andreas Kling.
- fast/regions/offsetLeft-offsetTop-in-multiple-regions.html:
- fast/regions/offsetLeft-offsetTop-in-region-absolute-sticky-fixed.html:
- 11:15 Changeset [150457] by
-
Revert r149635, it was too optimistic.
Rubber-stamped by Andreas Kling.
We are not at the point where we can always use more descriptive macros instead of PLATFORM()
- Scripts/webkitpy/style/checkers/cpp.py:
(check_invalid_increment):
(process_line):
(CppChecker):
- Scripts/webkitpy/style/checkers/cpp_unittest.py:
(CppStyleTest.test_plain_integral_bitfields):
- 11:07 WebKitGTK/2.0.x edited by
- (diff)
- 11:04 Changeset [150456] by
-
fourthTier: display-profiler-output should make it even easier to diff the compilation story between two different runs
https://bugs.webkit.org/show_bug.cgi?id=116556
Reviewed by Oliver Hunt.
This adds three new capabilities:
- 'display' now accepts negative compilation indices, and allows specifying wildcard hashes combined with combination indices. Previously you could say 'display <hash>', 'display <hash>-<index>-<engine>', or 'display *'. The latter would display every compilation. This improves this so that you can say 'display *-<index>-<engine>'. It also makes it so that you can give a negative index, which allows you to say things like 'display blah--1-dfg', which displays just the last DFG compilation. Also you can say 'display *--1-dfg', which displays the last DFG compilation for each code block. I realize that this is kind of ugly, but gosh is it practical.
- You can now say 'sort hash', which will henceforth sort all of the output by code hash rather than by the time when it was compiled. This means that if you're doing 'dispay *--1-dfg' and then want to diff the results against something else, you can ensure that this doesn't get confused just by changes in compilation order.
- You can now say 'counts off', which will henceforth disable the display of execution counts from 'bytecode' and 'display'. This is also useful for diffs, since if you're trying to figure out why two compilations are different, they probably have radically different counts. 'counts off' gets this out of the output so that it doesn't confuse your diff.
Note that to use this effectively you should also have a script that scrubs
pointers from text so that the diff doesn't get confused by pointers. I'll
post my elimptr script to the bug. Maybe at some point I'll integrate that
into display-profiler-output.
Put together this is pretty awesome. I was able to do the following:
(echo "counts off" && echo "sort hash" && echo "d *--1-dfg") | Tools/Scripts/display-profiler-output richards-serial.profile | elimptr > richards-serial.asm
(echo "counts off" && echo "sort hash" && echo "d *--1-dfg") | Tools/Scripts/display-profiler-output richards-concurrent.profile | elimptr > richards-concurrent.asm
diff -u richards-serial.asm richards-concurrent.asm
And this immediately told me that the reason why richards is slower in the
concurrent compilation case is just that we end up compiling *tons* more
functions, most of which are trivially inlineable. Basically, concurrent
compilation breaks our previous heuristics for delaying compilation of
inlineables just enough that they never trigger compilation. We should fix
that in a separate bug.
- Scripts/display-profiler-output:
- 10:26 Changeset [150455] by
-
Unreviewed; fix lldb synthetic representation of WTF::Vector after r148891.
- lldb/lldb_webkit.py:
(WTFVectorProvider.num_children):
(WTFVectorProvider.get_child_index):
(WTFVectorProvider.get_child_at_index):
(WTFVectorProvider.update):
- 10:20 Changeset [150454] by
-
REGRESSION(r101810): Media controls status text missing for live-stream videos.
https://bugs.webkit.org/show_bug.cgi?id=116547
Reviewed by Eric Carlson.
Un-reverse the logic in loadedMetadata() so that the status text isn't
hidden (when it should be shown) for live-stream videos.
- html/shadow/MediaControlsApple.cpp:
(WebCore::MediaControlsApple::loadedMetadata):
- 10:16 Changeset [150453] by
-
Implement overlap-avoidance for in-band text track cues.
https://bugs.webkit.org/show_bug.cgi?id=116540
Reviewed by Eric Carlson.
In-band (or Generic) cues need special casing for certain features
present in in-band tracks, like paint-on and roll-up modes. To avoid
the problem of overlap avoidance forcing a caption meant to appear
below a cue to appear above it when a larger font size is selected,
impose an additional sort ordering for "generic cues". Instead of
cues being ordered by the order they appear in the track, "generic
cues" further sorted by their position within the video area, such
that cues at the bottom of the video area appear first, and later
cues are pushed up to avoid them, preserving the desired apparent
ordering.
- html/HTMLMediaElement.cpp:
(WebCore::compareCueInterval): Added; wrapper around
TextTrackCue::isOrderedBefore.
(WebCore::HTMLMediaElement::updateActiveTextTrackCues):
After creating the list of current cues, sort them.
- html/track/TextTrackCue.cpp:
(WebCore::TextTrackCue::isOrderedBefore): Added; implementation moved
from TextTrackCueList::add().
- html/track/TextTrackCue.h:
- html/track/TextTrackCueGeneric.cpp:
(WebCore::TextTrackCueGeneric::isOrderedBefore): Added override;
impose additional oredring on generic cues.
- html/track/TextTrackCueGeneric.h:
- html/track/TextTrackCueList.cpp:
(WebCore::TextTrackCueList::add): Moved ordering test into
isOrderedBefore().
- rendering/RenderTextTrackCue.cpp:
(WebCore::RenderTextTrackCue::repositionGenericCue):
Call repositionCueSnapToLinesNotSet() after positioning the cue.
- 10:14 Changeset [150452] by
-
Implement overlap avoidance for cues with snap-to-lines flag not set
https://bugs.webkit.org/show_bug.cgi?id=84296
Reviewed by Eric Carlson.
Source/WebCore:
Test: media/track/track-cue-overlap-snap-to-lines-not-set.html
Support overlap avoidance for the non-snap-to-lines part of the WebVTT spec.
- rendering/RenderTextTrackCue.cpp:
(WebCore::RenderTextTrackCue::isOutside): Split implementation into rectIsWithinContainer().
(WebCore::RenderTextTrackCue::rectIsWithinContainer): Ditto.
(WebCore::RenderTextTrackCue::isOverlapping): Split into overlappingObject() and overlappingObjectForRect().
(WebCore::RenderTextTrackCue::overlappingObject): Ditto.
(WebCore::RenderTextTrackCue::overlappingObjectForRect): Ditto.
(WebCore::RenderTextTrackCue::moveIfNecessaryToKeepWithinContainer): Added.
(WebCore::RenderTextTrackCue::findNonOverlappingPosition): When an overlapping object is found, move the
cue to just above or below that object and try again.
(WebCore::RenderTextTrackCue::repositionCueSnapToLinesSet): Move implementation into moveIfNecessaryToKeepWithinContainer().
(WebCore::RenderTextTrackCue::repositionCueSnapToLinesNotSet): Add implementanton based on above.
- rendering/RenderTextTrackCue.h:
LayoutTests:
- media/track/track-cue-overlap-snap-to-lines-not-set-expected.txt: Added.
- media/track/track-cue-overlap-snap-to-lines-not-set.html: Added.
- 09:57 Changeset [150451] by
-
Build fix.
- wtf/NeverDestroyed.h:
(NeverDestroyed):
- 09:56 Changeset [150450] by
-
Add WTF::NeverDestroyed and start using it in WTF
https://bugs.webkit.org/show_bug.cgi?id=116472
Reviewed by Benjamin Poulain.
NeverDestroyed is a class template that can be used for singletons and other objects that we never
want to destroy. It's intended as a replacement for WTF_STATIC_LOCAL with the advantage that it doesn't
fragment the heap.
- GNUmakefile.list.am:
- WTF.vcproj/WTF.vcproj:
- WTF.vcxproj/WTF.vcxproj:
- WTF.vcxproj/WTF.vcxproj.filters:
- WTF.xcodeproj/project.pbxproj:
Add NeverDestroyed.h
- wtf/Compiler.h:
Add a helper macro, WTF_DELETED_FUNCTION.
- wtf/CryptographicallyRandomNumber.cpp:
Use NeverDestroyed.
- wtf/NeverDestroyed.h: Added.
- wtf/Noncopyable.h:
Use WTF_DELETED_FUNCTION.
- wtf/text/WTFString.cpp:
(WTF::emptyString):
Use
- 09:44 Changeset [150449] by
-
Added missing assert condition for PositiveOrZero in ARM branch32().
https://bugs.webkit.org/show_bug.cgi?id=116538.
Reviewed by Geoffrey Garen.
- assembler/MacroAssemblerARM.h:
(JSC::MacroAssemblerARM::branchAdd32):
- 09:39 Changeset [150448] by
-
2013-05-21 Lucas Forschler <lforschler@apple.com>
Merge r150441
2013-05-21 Jessie Berlin <jberlin@apple.com>
Expose a way to know when forms are added to a page or when form controls are added to a form
in the injected bundle
https://bugs.webkit.org/show_bug.cgi?id=116334
Reviewed by Alexey Proskuryakov.
Source/WebKit2:
Add shouldNotifyOnFormChanges and didAssociateFormControls to the WKBundlePageFormClient.
- WebProcess/InjectedBundle/API/c/WKBundlePage.h: Add the new callbacks as part of version 2 of the WKBundlePageFormClient.
- WebProcess/InjectedBundle/InjectedBundlePageFormClient.cpp: (WebKit::InjectedBundlePageFormClient::didAssociateFormControls): Pass the message along to the client if the client has a handler. (WebKit::InjectedBundlePageFormClient::shouldNotifyOnFormChanges): Ditto.
- WebProcess/InjectedBundle/InjectedBundlePageFormClient.h:
- WebProcess/WebCoreSupport/WebChromeClient.cpp: (WebKit::WebChromeClient::didAssociateFormControls): Tell the injected bundle form client for the page. (WebKit::WebChromeClient::shouldNotifyOnFormChanges): Ditto.
- WebProcess/WebCoreSupport/WebChromeClient.h:
Tools:
Add tests for the new callbacks.
- TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Add DidAssociateFormControls/_Bundle.cpp and associate-form-controls.html
- TestWebKitAPI/Tests/WebKit2/DidAssociateFormControls.cpp: Added. (TestWebKitAPI::nullJavaScriptCallback): A "null" callback to handle the fact that WKPageRunJavaScriptInMainFrame cannot handle null being passed in for the callback. (TestWebKitAPI::didReceiveMessageFromInjectedBundle): After receiving the message that didAssociateFormControls callback was invoked from adding the form in the onload handler, tell the page to add a password field to the form, which should also invoke didAssociateFormControls callback. (TestWebKitAPI::setInjectedBundleClient): Register to receive messages. (TestWebKitAPI::TEST): Load associate-form-controls.html and wait until the didAssociateFormControls callback has been invoked for both adding the form and for adding a password field to the form.
- TestWebKitAPI/Tests/WebKit2/DidAssociateFormControls_Bundle.cpp: Added. (TestWebKitAPI::shouldNotifyOnFormChanges): Return true so the didAssociateFormControls callback is invoked. (TestWebKitAPI::didAssociateFormControls): Tell the UI process. (TestWebKitAPI::DidAssociateFormControlsTest::DidAssociateFormControlsTest): (TestWebKitAPI::DidAssociateFormControlsTest::didCreatePage): Register for the shouldNotifyOnFormChanges and didAssociateFormControls callbacks.
- TestWebKitAPI/Tests/WebKit2/associate-form-controls.html: Added. Add a form in response to the onload event. Add a button that will add the password field for manual testing.
- 09:21 Changeset [150447] by
-
Versioning.
- 09:09 Changeset [150446] by
-
Remove GraphicsLayerClient::contentsVisible()
https://bugs.webkit.org/show_bug.cgi?id=116523
Patch by Alberto Garcia <agarcia@igalia.com> on 2013-05-21
Reviewed by Darin Adler.
This is dead code from the BlackBerry port.
- platform/graphics/GraphicsLayerClient.h:
- rendering/RenderLayerBacking.cpp:
- 09:05 Changeset [150445] by
-
[GTK] Compile everything in C++11 mode
https://bugs.webkit.org/show_bug.cgi?id=116452
Reviewed by Anders Carlsson.
.:
- Source/autotools/SetupCompilerFlags.m4: Use the C++11 standard by default when compiling C++ source code.
Perform some minor cleanup around the comments and the order of specifying additional CXXFLAGS entries.
Source/WebKit2:
- GNUmakefile.am: Remove the -std=c++11 flags from the various CPPFLAGS lists, the mode is now specified
by default for all the C++ source code in SetupCompilerFlags.m4.
- 09:03 Changeset [150444] by
-
PLATFORM(*) macros used in Source/WebCore/loader/archive/ArchiveFactory.cpp
https://bugs.webkit.org/show_bug.cgi?id=116453
Reviewed by Alexey Proskuryakov.
- loader/archive/ArchiveFactory.cpp: Remove a redundant PLATFORM(QT) macro when checking whether
to include the LegacyWebArchive.h header that's specific to the WebArchive support. The Qt port
does not enable the WebArchive feature at all (which is enabled only on AppleMac, AppleWin and iOS
platforms) so there's no reason to additionally condition the header include with building on
non-Qt platforms.
- 08:58 Changeset [150443] by
-
Add BlackBerry definition of NativeImagePtr
https://bugs.webkit.org/show_bug.cgi?id=116526
Patch by Alberto Garcia <agarcia@igalia.com> on 2013-05-21
Reviewed by Darin Adler.
- platform/graphics/NativeImagePtr.h:
(Graphics):
(WebCore):
- 08:50 Changeset [150442] by
-
[BlackBerry] FontCache::getFontDataForCharacters() returns nullptr
https://bugs.webkit.org/show_bug.cgi?id=116529
Patch by Alberto Garcia <agarcia@igalia.com> on 2013-05-21
Reviewed by Andreas Kling.
This cannot be converted to a PassRefPtr and breaks the build.
- platform/graphics/blackberry/FontCacheBlackBerry.cpp:
(WebCore::FontCache::getFontDataForCharacters):
- 08:12 WebKitGTK/2.0.x edited by
- (diff)
- 06:44 Changeset [150441] by
-
Expose a way to know when forms are added to a page or when form controls are added to a form
in the injected bundle
https://bugs.webkit.org/show_bug.cgi?id=116334
Reviewed by Alexey Proskuryakov.
Source/WebKit2:
Add shouldNotifyOnFormChanges and didAssociateFormControls to the WKBundlePageFormClient.
- WebProcess/InjectedBundle/API/c/WKBundlePage.h:
Add the new callbacks as part of version 2 of the WKBundlePageFormClient.
- WebProcess/InjectedBundle/InjectedBundlePageFormClient.cpp:
(WebKit::InjectedBundlePageFormClient::didAssociateFormControls):
Pass the message along to the client if the client has a handler.
(WebKit::InjectedBundlePageFormClient::shouldNotifyOnFormChanges):
Ditto.
- WebProcess/InjectedBundle/InjectedBundlePageFormClient.h:
- WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::didAssociateFormControls):
Tell the injected bundle form client for the page.
(WebKit::WebChromeClient::shouldNotifyOnFormChanges):
Ditto.
- WebProcess/WebCoreSupport/WebChromeClient.h:
Tools:
Add tests for the new callbacks.
- TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
Add DidAssociateFormControls/_Bundle.cpp and associate-form-controls.html
- TestWebKitAPI/Tests/WebKit2/DidAssociateFormControls.cpp: Added.
(TestWebKitAPI::nullJavaScriptCallback):
A "null" callback to handle the fact that WKPageRunJavaScriptInMainFrame cannot handle null
being passed in for the callback.
(TestWebKitAPI::didReceiveMessageFromInjectedBundle):
After receiving the message that didAssociateFormControls callback was invoked from adding
the form in the onload handler, tell the page to add a password field to the form, which
should also invoke didAssociateFormControls callback.
(TestWebKitAPI::setInjectedBundleClient):
Register to receive messages.
(TestWebKitAPI::TEST):
Load associate-form-controls.html and wait until the didAssociateFormControls callback has
been invoked for both adding the form and for adding a password field to the form.
- TestWebKitAPI/Tests/WebKit2/DidAssociateFormControls_Bundle.cpp: Added.
(TestWebKitAPI::shouldNotifyOnFormChanges):
Return true so the didAssociateFormControls callback is invoked.
(TestWebKitAPI::didAssociateFormControls):
Tell the UI process.
(TestWebKitAPI::DidAssociateFormControlsTest::DidAssociateFormControlsTest):
(TestWebKitAPI::DidAssociateFormControlsTest::didCreatePage):
Register for the shouldNotifyOnFormChanges and didAssociateFormControls callbacks.
- TestWebKitAPI/Tests/WebKit2/associate-form-controls.html: Added.
Add a form in response to the onload event. Add a button that will add the password field
for manual testing.
- 06:37 Changeset [150440] by
-
[GTK] Always use EGL to create the GL context when running on Wayland
https://bugs.webkit.org/show_bug.cgi?id=115721
Patch by Iago Toral Quiroga <itoral@igalia.com> on 2013-05-21
Reviewed by Martin Robinson.
- platform/graphics/cairo/GLContext.cpp:
(WebCore::GLContext::createContextForWindow):
(WebCore::GLContext::createOffscreenContext): Always use EGL to create the GL context
instead of GLX when running on Wayland and make createOffScreenContext reuse the code
in createContextWindow.
- 06:16 Changeset [150439] by
-
[GTK][WK2] accessibility/label-for-control-hittest.html is failing
https://bugs.webkit.org/show_bug.cgi?id=106341
Patch by Krzysztof Czech <k.czech@samsung.com> on 2013-05-21
Reviewed by Martin Robinson.
Tools:
Adds support for retrieving accessibility element at exact point.
- WebKitTestRunner/InjectedBundle/atk/AccessibilityUIElementAtk.cpp:
(WTR::AccessibilityUIElement::elementAtPoint):
LayoutTests:
Fixing failing test. The reason of failure is missing
possibility of getting accessibility object at exact position.
- platform/gtk-wk2/TestExpectations:
- 06:14 Changeset [150438] by
-
[EFL] accessibility/accessibility-node-reparent.html is failing
https://bugs.webkit.org/show_bug.cgi?id=112026
Unreviewed EFL gardening.
accessibility/accessibility-node-reparent.html passes after r150431.
Patch by Krzysztof Czech <k.czech@samsung.com> on 2013-05-21
- platform/efl-wk2/TestExpectations:
- 05:50 Changeset [150437] by
-
Add FloatRect::normalized() for BlackBerry
https://bugs.webkit.org/show_bug.cgi?id=116531
Patch by Alberto Garcia <agarcia@igalia.com> on 2013-05-21
Reviewed by Carlos Garcia Campos.
This was added to FloatRectBlackBerry.cpp in r111072 when the
graphics platform code was upstreamed, but this change was
missing.
- platform/graphics/FloatRect.h:
(FloatRect):
- 05:21 Changeset [150436] by
-
[EFL] accessibility/accessibility-node-memory-management.html is failing
https://bugs.webkit.org/show_bug.cgi?id=112025
Unreviewed EFL gardening
accessibility/accessibility-node-memory-management.html passes after r150431.
Patch by Krzysztof Czech <k.czech@samsung.com> on 2013-05-21
- platform/efl-wk2/TestExpectations:
- 04:35 Changeset [150435] by
-
[GStreamer] cleanup duration query
https://bugs.webkit.org/show_bug.cgi?id=116228
Reviewed by Philippe Normand.
Covered by existing tests.
- platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
(WebCore::MediaPlayerPrivateGStreamer::duration):
We can also cache the duration here if the query succeeds.
(MediaPlayerPrivateGStreamer::updateStates):
Don't query when we are in GST_STATE_READY state because it never succeeds.
Do it instead when we reached a stable state.
(MediaPlayerPrivateGStreamer::cacheDuration): Return early if it is already
cached. Recaching is not necessary and now we call it a lot of times.
Only look into the state if the query failed.
- platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.h:
(MediaPlayerPrivateGStreamer): Made m_mediaDuration mutable to be
able to cache it in duration(). Internal cached values is one of
the sensible uses of mutable.
- 03:55 Changeset [150434] by
-
[BlackBerry] Make PagePopup implementation independent from WebCore
https://bugs.webkit.org/show_bug.cgi?id=116448
Patch by Carlos Garcia Campos <cgarcia@igalia.com> on 2013-05-21
Reviewed by Anders Carlsson.
.:
- Source/cmake/OptionsBlackBerry.cmake: Do not enable PAGE_POPUP
feature.
Source/WebKit:
- PlatformBlackBerry.cmake: Add
blackberry/WebCoreSupport/PagePopupBlackBerryClient.cpp to
compilation.
Source/WebKit/blackberry:
Add our own implementation of PagePopupClient and make all the
pickers inherit from it. Unused methods have been removed and
common implementation have been moved from the pickers to the
parent class.
- Api/WebPage.cpp:
(BlackBerry::WebKit::WebPagePrivate::~WebPagePrivate): Destroy the
popup if there's one active.
(BlackBerry::WebKit::WebPagePrivate::setVisible): Call
closePagePopup() directly.
(BlackBerry::WebKit::WebPagePrivate::openPagePopup): Create a new
PagePopupBlackBerry for the given client.
(BlackBerry::WebKit::WebPagePrivate::closePagePopup): Close the
active popup if there's one.
(BlackBerry::WebKit::WebPagePrivate::hasOpenedPopup): Return
whether there's an active popup.
- Api/WebPage.h:
- Api/WebPage_p.h:
- WebCoreSupport/ChromeClientBlackBerry.cpp:
(WebCore::ChromeClientBlackBerry::chromeDestroyed): Remove
closePagePopup() call since this is now handled by
WebPagePrivate.
(WebCore::ChromeClientBlackBerry::hasOpenedPopup): Call
WebPagePrivate::hasOpenedPopup().
- WebCoreSupport/ChromeClientBlackBerry.h:
(ChromeClientBlackBerry):
- WebCoreSupport/ColorPickerClient.cpp:
(BlackBerry::WebKit::ColorPickerClient::ColorPickerClient):
(BlackBerry::WebKit::ColorPickerClient::didClosePopup):
- WebCoreSupport/ColorPickerClient.h:
(ColorPickerClient):
- WebCoreSupport/DatePickerClient.cpp:
(BlackBerry::WebKit::DatePickerClient::DatePickerClient):
(BlackBerry::WebKit::DatePickerClient::didClosePopup):
- WebCoreSupport/DatePickerClient.h:
(DatePickerClient):
- WebCoreSupport/PagePopupBlackBerry.cpp:
(BlackBerry::WebKit::PagePopupBlackBerry::PagePopupBlackBerry):
(BlackBerry::WebKit::PagePopupBlackBerry::closePopup):
- WebCoreSupport/PagePopupBlackBerry.h:
(PagePopupBlackBerry):
(BlackBerry::WebKit::PagePopupBlackBerry::SharedClientPointer::SharedClientPointer):
(BlackBerry::WebKit::PagePopupBlackBerry::SharedClientPointer::get):
(SharedClientPointer):
- WebCoreSupport/PagePopupBlackBerryClient.cpp: Added.
(BlackBerry::WebKit::PagePopupBlackBerryClient::PagePopupBlackBerryClient):
(BlackBerry::WebKit::PagePopupBlackBerryClient::closePopup):
(BlackBerry::WebKit::PagePopupBlackBerryClient::didClosePopup):
(BlackBerry::WebKit::PagePopupBlackBerryClient::contentSize):
(BlackBerry::WebKit::PagePopupBlackBerryClient::writeDocument):
- WebCoreSupport/PagePopupBlackBerryClient.h: Added.
(PagePopupBlackBerryClient):
(BlackBerry::WebKit::PagePopupBlackBerryClient::~PagePopupBlackBerryClient):
- WebCoreSupport/SelectPopupClient.cpp:
(BlackBerry::WebKit::SelectPopupClient::SelectPopupClient):
(BlackBerry::WebKit::SelectPopupClient::didClosePopup):
- WebCoreSupport/SelectPopupClient.h:
(SelectPopupClient):
- WebKitSupport/InputHandler.cpp:
(BlackBerry::WebKit::InputHandler::openDatePopup): Do not call
closePagePopup() since this is already done by openPagePopup() and
use openPagePopup() from WebPagePrivate directly.
(BlackBerry::WebKit::InputHandler::openColorPopup): Ditto.
(BlackBerry::WebKit::InputHandler::openSelectPopup): Ditto.
- 02:31 Changeset [150433] by
-
Build fix for !USE(ACCELERATED_COMPOSITING) after r150307.
- rendering/RenderLayerCompositor.h: Added missing #if USE(ACCELERATED_COMPOSITING).
- 02:04 Changeset [150432] by
-
[Qt][WK2] Fix Mountain Lion builds after r150277.
https://bugs.webkit.org/show_bug.cgi?id=116467
Reviewed by Jocelyn Turcotte.
- Shared/API/c/WKBase.h:
- 01:43 Changeset [150431] by
-
[ATK] [WebKit2] AccessibilityUIElement often leaks its AtkObject
https://bugs.webkit.org/show_bug.cgi?id=112913
Patch by Krzysztof Czech <k.czech@samsung.com> on 2013-05-21
Reviewed by Martin Robinson.
AccessibilityUIElement is often created by passing new reference to AtkObject (transfer full).
AccessibilityUIElement::~AccessibilityUIElement does not decrement the reference of the element,
so the leak appears.
- WebKitTestRunner/InjectedBundle/AccessibilityController.cpp:
(WTR):
- WebKitTestRunner/InjectedBundle/AccessibilityUIElement.h:
- WebKitTestRunner/InjectedBundle/atk/AccessibilityControllerAtk.cpp:
(WTR::AccessibilityController::rootElement):
(WTR):
(WTR::AccessibilityController::focusedElement):
- WebKitTestRunner/InjectedBundle/atk/AccessibilityUIElementAtk.cpp:
(WTR::checkElementState):
(WTR::indexRangeInTable):
(WTR::alterCurrentValue):
(WTR::AccessibilityUIElement::getChildren):
(WTR::AccessibilityUIElement::getChildrenWithRange):
(WTR::AccessibilityUIElement::childrenCount):
(WTR::AccessibilityUIElement::indexOfChild):
(WTR::AccessibilityUIElement::titleUIElement):
(WTR::AccessibilityUIElement::parentElement):
(WTR::AccessibilityUIElement::allAttributes):
(WTR::AccessibilityUIElement::role):
(WTR::AccessibilityUIElement::title):
(WTR::AccessibilityUIElement::description):
(WTR::AccessibilityUIElement::orientation):
(WTR::AccessibilityUIElement::stringValue):
(WTR::AccessibilityUIElement::language):
(WTR::AccessibilityUIElement::x):
(WTR::AccessibilityUIElement::y):
(WTR::AccessibilityUIElement::width):
(WTR::AccessibilityUIElement::height):
(WTR::AccessibilityUIElement::intValue):
(WTR::AccessibilityUIElement::minValue):
(WTR::AccessibilityUIElement::maxValue):
(WTR::AccessibilityUIElement::isEnabled):
(WTR::AccessibilityUIElement::isFocused):
(WTR::AccessibilityUIElement::isSelected):
(WTR::AccessibilityUIElement::isExpanded):
(WTR::AccessibilityUIElement::isChecked):
(WTR::AccessibilityUIElement::rowCount):
(WTR::AccessibilityUIElement::columnCount):
(WTR::AccessibilityUIElement::rowIndexRange):
(WTR::AccessibilityUIElement::columnIndexRange):
(WTR::AccessibilityUIElement::cellForColumnAndRow):
(WTR::AccessibilityUIElement::increment):
(WTR::AccessibilityUIElement::decrement):
(WTR::AccessibilityUIElement::press):
(WTR::AccessibilityUIElement::documentEncoding):
(WTR::AccessibilityUIElement::documentURI):
(WTR::AccessibilityUIElement::isFocusable):
(WTR::AccessibilityUIElement::isSelectable):
(WTR::AccessibilityUIElement::isMultiSelectable):
(WTR::AccessibilityUIElement::isVisible):
- 01:40 Changeset [150430] by
-
Simplify Shadow DOM distribution code
https://bugs.webkit.org/show_bug.cgi?id=116454
Reviewed by Andreas Kling.
Remove code supporting nested insertion points and distributing to multiple insertion points.
- html/HTMLDetailsElement.cpp:
(DetailsContentElement):
Make distribution to <details> and <summary> mutually exclusive.
- html/shadow/ContentDistributor.cpp:
(WebCore):
(WebCore::ContentDistributor::distribute):
(WebCore::ContentDistributor::invalidate):
(WebCore::ContentDistributor::distributeSelectionsTo):
- html/shadow/ContentDistributor.h:
- html/shadow/InsertionPoint.cpp:
(WebCore::resolveReprojection):
- 01:07 Changeset [150429] by
-
New Tag.
- 01:06 Changeset [150428] by
- 01:02 Changeset [150427] by
-
[CSSRegions] Constrain auto-height region computation in a different way
https://bugs.webkit.org/show_bug.cgi?id=116310
Reviewed by Alexandru Chiculita.
After https://bugs.webkit.org/show_bug.cgi?id=74132, the region is a render block. We can use RenderBox::constrainContentBoxLogicalHeightByMinMax
instead of RenderBox::computeReplacedLogicalHeightRespectingMinMaxHeight to better reflect the fact that the region is not a replaced element.
Covered by existing regions tests, since when replaced-based the region had an intrinsic height of 0 and now, block-based, it does not have any children.
- rendering/RenderFlowThread.cpp:
(WebCore::RenderFlowThread::addForcedRegionBreak):
- 00:48 Changeset [150426] by
-
[Mac] svg/batik/paints/patternRegions-positioned-objects.svg is flaky
https://bugs.webkit.org/show_bug.cgi?id=116519
Add a flaky test expectation.
- platform/mac/TestExpectations:
- 00:39 Changeset [150425] by
-
REGRESSION(r150393): editing/inserting/typing-at-end-of-line.html fails
https://bugs.webkit.org/show_bug.cgi?id=116516
Reviewed by Antti Koivisto.
The bug was caused by DRT not resetting various auto correction states.
Do that in Internals::resetToConsistentState.
- testing/Internals.cpp:
(WebCore::Internals::resetToConsistentState):
- 00:00 Changeset [150424] by
-
Source/WebKit2: Rollout r150398.
Tools: Rollout 150398.
05/20/13:
- 23:39 Changeset [150423] by
-
New Tag.
- 23:39 Changeset [150422] by
- 23:16 Changeset [150421] by
-
[GTK] Fix concatenations of string literals that are not C++11-compliant
https://bugs.webkit.org/show_bug.cgi?id=116449
Reviewed by Martin Robinson.
Fix the problematic concatenations of string literals so they are compilable under the C++11 standard. This
is simply a matter of separating the string and the string literal that are being concatenated with a space.
Source/WebCore:
- platform/graphics/gtk/FullscreenVideoControllerGtk.cpp:
- platform/gtk/FileSystemGtk.cpp:
(WebCore::sharedResourcesPath):
Source/WebKit2:
- UIProcess/API/gtk/tests/TestPrinting.cpp:
(testPrintOperationPrint):
- UIProcess/API/gtk/tests/WebExtensionTest.cpp:
(methodCallCallback):
Tools:
- ImageDiff/gtk/ImageDiff.cpp:
(printImage):
- 23:13 Changeset [150420] by
-
[GTK] Pass GTEST_HAS_RTTI=0 cppflag to TestWebKitAPI programs at compilation-time
https://bugs.webkit.org/show_bug.cgi?id=116450
Reviewed by Martin Robinson.
The -fno-rtti flag is used when compiling C++ source code throughout the project. The TestWebKitAPI
programs fail to compile because of this when using the C++11 standard as the GTest headers assume by
default that the RTTI support is available. This can be turned off by passing GTEST_HAS_RTTI=0 to the
TestWebKitAPI programs when compiling them.
- TestWebKitAPI/GNUmakefile.am:
- 23:11 Changeset [150419] by
-
Use 'typeof' keyword to avoid problems when compiling in C++11 standard compliance
https://bugs.webkit.org/show_bug.cgi?id=116447
Reviewed by Darin Adler.
The 'typeof' keyword is a GNU C extension which becomes unavailable when compiling in ANSI or specific standard mode
(such as -std=c++11). The 'typeof' keyword should be used instead as it's available in all modes both under GCC and Clang.
- wtf/FastMalloc.cpp:
(WTF): Replace the 'typeof' keyword with 'typeof'.
- 23:08 Changeset [150418] by
-
[GTK] Support application/x-mimearchive mimetype for MHTML archives instead of message/rfc822
https://bugs.webkit.org/show_bug.cgi?id=116442
Reviewed by Martin Robinson.
- loader/archive/ArchiveFactory.cpp:
(WebCore::archiveMIMETypes): Like EFL and Qt, register the application/x-mimearchive mime type as a MHTML archive type
and stop propagating the message/rfc822 mime type as such. This also removes a couple of PLATFORM ifdefs.
- 22:57 Changeset [150417] by
-
Minor String fixes in CSS
https://bugs.webkit.org/show_bug.cgi?id=116291
Patch by Benjamin Poulain <bpoulain@apple.com> on 2013-05-20
Reviewed by Darin Adler.
Some trival changes to make up for a bad day :)
- css/CSSImageSetValue.cpp:
(WebCore::CSSImageSetValue::customCssText):
- css/CSSImageValue.cpp:
(WebCore::CSSImageValue::customCssText):
- css/CSSImportRule.cpp:
(WebCore::CSSImportRule::cssText):
- css/CSSInheritedValue.cpp:
(WebCore::CSSInheritedValue::customCssText):
- css/CSSInitialValue.cpp:
(WebCore::CSSInitialValue::customCssText):
- css/CSSPrimitiveValue.cpp:
(WebCore::CSSPrimitiveValue::customCssText):
- css/CSSPropertySourceData.cpp:
(WebCore::CSSPropertySourceData::toString):
- css/CSSSelector.cpp:
(WebCore::CSSSelector::selectorText):
- css/CSSStyleRule.cpp:
(WebCore::CSSStyleRule::generateSelectorText):
- css/CSSStyleSheet.h:
(WebCore::CSSStyleSheet::type):
- css/CSSTimingFunctionValue.cpp:
(WebCore::CSSLinearTimingFunctionValue::customCssText):
(WebCore::CSSCubicBezierTimingFunctionValue::customCssText):
- css/CSSValueList.cpp:
(WebCore::CSSValueList::customCssText):
(WebCore::CSSValueList::customSerializeResolvingVariables):
- css/MediaQueryExp.cpp:
(WebCore::MediaQueryExp::serialize):
- css/WebKitCSSFilterValue.cpp:
(WebCore::WebKitCSSFilterValue::customCssText):
- css/WebKitCSSKeyframesRule.cpp:
(WebCore::StyleRuleKeyframes::findKeyframeIndex):
(WebCore::WebKitCSSKeyframesRule::cssText):
- css/WebKitCSSMatFunctionValue.cpp:
(WebCore::WebKitCSSMatFunctionValue::customCssText):
- css/WebKitCSSMixFunctionValue.cpp:
(WebCore::WebKitCSSMixFunctionValue::customCssText):
- 22:50 Changeset [150416] by
-
Remove DriverProxy._running_drivers from webkitpy.layout_tests.port
https://bugs.webkit.org/show_bug.cgi?id=116492
Reviewed by Dirk Pranke.
Merge https://chromium.googlesource.com/chromium/blink/+/1af2f0be5a1746784ff114952f29f7b8fa9895f3.
The DriverProxy class had the concept of keeping a pool of started DRTs around (indexed by command line)
so that we didn't necessarily need to restart DRT when the command line for a test changed (e.g., when
starting to run tests in a virtual suite). However, the pool was per-worker (so not shared across the machine)
and we didn't cap the number of processes started, so we would end up with a whole bunch of DRTs (3x-4x
the number of workers started). If anything, this seemed to make the tests slightly slower and probably added
to memory pressure on the machine.
This change removes the pool code; now we'll simply restart DRT when the command line changes. With this change,
we can actually fold the restarting code into Driver itself (and get rid of DriverProxy), but I'll leave that
for a subsequent patch just to make things easier to follow.
- Scripts/webkitpy/port/driver.py:
(DriverProxy.init):
(DriverProxy.run_test):
(DriverProxy.has_crashed):
(DriverProxy.stop):
- 22:33 Changeset [150415] by
-
Change EFL Linux 64-bit Release with EFL Linux 64-bit Release WK1
https://bugs.webkit.org/show_bug.cgi?id=116423
Reviewed by Ryosuke Niwa.
EFL Linux 64-bit Release starts to support WK1's build and layout test from now on.
- BuildSlaveSupport/build.webkit.org-config/config.json:
- 21:55 Changeset [150414] by
-
New Tag.
- 21:52 Changeset [150413] by
-
Remove remaining link-prerender code
https://bugs.webkit.org/show_bug.cgi?id=116429
Patch by Seokju Kwon <seokju.kwon@gmail.com> on 2013-05-20
Reviewed by Darin Adler.
Tools:
Link prerendering code was removed in r150356.
- Scripts/webkitperl/FeatureList.pm:
- qmake/mkspecs/features/features.pri:
LayoutTests:
Remove test cases related to link-prerender.
- fast/dom/HTMLLinkElement/prerender-insert-after-stop-expected.txt: Removed.
- fast/dom/HTMLLinkElement/prerender-insert-after-stop.html: Removed.
- fast/dom/HTMLLinkElement/prerender-remove-after-stop-expected.txt: Removed.
- fast/dom/HTMLLinkElement/prerender-remove-after-stop.html: Removed.
- fast/dom/HTMLLinkElement/resources/empty.html: Removed.
- fast/dom/HTMLLinkElement/resources/empty2.html: Removed.
- fast/dom/HTMLLinkElement/resources/simple-prerender.html: Removed.
- platform/mac/TestExpectations:
- platform/qt/TestExpectations:
- platform/win/TestExpectations:
- platform/wincairo/TestExpectations:
- platform/wk2/TestExpectations:
- 21:51 Changeset [150412] by
- 21:50 Changeset [150411] by
-
New Tag.
- 21:50 Changeset [150410] by
-
New Tag.
- 21:49 Changeset [150409] by
- 21:42 Changeset [150408] by
-
Unreviewed, rolling out r150384.
http://trac.webkit.org/changeset/150384
https://bugs.webkit.org/show_bug.cgi?id=116513
Not ready for primetime. (Requested by kling on #webkit).
Patch by Commit Queue <commit-queue@webkit.org> on 2013-05-20
- UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::WebProcessProxy):
(WebKit::WebProcessProxy::removeWebPage):
- UIProcess/WebProcessProxy.h:
(WebProcessProxy):
- 21:35 Changeset [150407] by
-
Versioning.
- 21:34 Changeset [150406] by
-
Unreviewed EFL gardening. Below tests don't come to crash anymore after r150368.
But, pending-activity.html still comes to crash on wk2.
storage/indexeddb/pending-activity.html
storage/indexeddb/transaction-coordination-across-databases.html
- platform/efl-wk2/TestExpectations:
- platform/efl/TestExpectations:
- 21:30 Changeset [150405] by
-
New Tag.
- 20:58 Changeset [150404] by
-
commit-queue should check for oops in changelog entries
https://bugs.webkit.org/show_bug.cgi?id=116395
Reviewed by Martin Robinson.
Make changelog validation fail when it contains oops!.
- Scripts/webkitpy/tool/bot/commitqueuetask.py:
(CommitQueueTask._validate_changelog):
- Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
(CommitQueueTaskTest._run_through_task):
- Scripts/webkitpy/tool/commands/queues_unittest.py:
- Scripts/webkitpy/tool/steps/validatechangelogs.py:
(ValidateChangeLogs.options):
(ValidateChangeLogs._check_changelog_diff):
(ValidateChangeLogs._changelog_contains_oops):
(ValidateChangeLogs.run):
- Scripts/webkitpy/tool/steps/validatechangelogs_unittest.py:
(ValidateChangeLogsTest._assert_start_line_produces_output):
(ValidateChangeLogsTest.test_check_changelog_diff):
(ValidateChangeLogsTest):
(ValidateChangeLogsTest.test_changelog_contains_oops):
- 20:43 Changeset [150403] by
-
Add tests for named access checks on DOMWindow
https://bugs.webkit.org/show_bug.cgi?id=116512
Reviewed by Geoffrey Garen.
Merge https://chromium.googlesource.com/chromium/blink/+/9207625a100abe9b72333ac5dba02b38ca5f4466
- http/tests/security/xss-DENIED-window-name-alert-expected.txt: Added.
- http/tests/security/xss-DENIED-window-name-alert.html: Added.
- http/tests/security/xss-DENIED-window-name-navigator-expected.txt: Added.
- http/tests/security/xss-DENIED-window-name-navigator.html: Added.
- 20:18 Changeset [150402] by
-
Null pointer deference in WebCore::AppendNodeCommand::create
https://bugs.webkit.org/show_bug.cgi?id=116479
Source/WebCore:
Reviewed by Andreas Kling.
Merge https://chromium.googlesource.com/chromium/blink/+/5cb43002a44f67a60ecf5a7ed76de2d0bcf89eb2
DeleteSelection::makeStylingElementsDirectChildrenOfEditableRootToPreventStyleLoss() make style and link elements
to be the direct children of the editable root. However, these style and link elements are not necessary editable
and WebKit crashes when they are not.
Test: editing/deleting/delete-uneditable-style.html
- editing/DeleteSelectionCommand.cpp:
(WebCore::DeleteSelectionCommand::makeStylingElementsDirectChildrenOfEditableRootToPreventStyleLoss):
LayoutTests:
Reviewed by Andreas Kling.
Add a regression test.
- editing/deleting/delete-uneditable-style-expected.txt: Added.
- editing/deleting/delete-uneditable-style.html: Added.
- 20:13 Changeset [150401] by
-
Disable SuperRegion
https://bugs.webkit.org/show_bug.cgi?id=116362
Rubber stamped by Geoff Garen.
- heap/Region.h:
- 19:39 Changeset [150400] by
-
Remove SharedMemoryGtk.cpp
https://bugs.webkit.org/show_bug.cgi?id=116457
Reviewed by Andreas Kling.
- Platform/gtk/SharedMemoryGtk.cpp: Removed.
- 19:29 Changeset [150399] by
-
Scrollbars still show sometimes even when WKPageSetSuppressScrollbarAnimations()
is set
https://bugs.webkit.org/show_bug.cgi?id=116493
-and corresponding-
<rdar://problem/13912871>
Reviewed by Dean Jackson.
We need to be more aggressive with our approach since AppKit may still ask the
scrollbars to paint here.
scrollbarAnimationsAreSuppressed() will allow us to find out if the setting has
been set.
- page/FrameView.cpp:
(WebCore::FrameView::scrollbarAnimationsAreSuppressed):
- page/FrameView.h:
- platform/ScrollableArea.h:
(WebCore::ScrollableArea::scrollbarAnimationsAreSuppressed):
- rendering/RenderLayer.cpp:
(WebCore::RenderLayer::scrollbarAnimationsAreSuppressed):
- rendering/RenderLayer.h:
(RenderLayer):
- rendering/RenderListBox.cpp:
(WebCore::RenderListBox::scrollbarAnimationsAreSuppressed):
- rendering/RenderListBox.h:
Don’t allow animations if scrollbars are suppressed.
- platform/mac/ScrollAnimatorMac.mm:
(-[WebScrollbarPainterDelegate scrollerImp:animateKnobAlphaTo:duration:]):
Call cancelAnimations() to prevent a just-started animation from continuing.
(WebCore::ScrollAnimatorMac::finishCurrentScrollAnimations):
- 19:07 Changeset [150398] by
-
Expose a way to know when forms are added to a page or when form controls are added to a form
in the injected bundle
https://bugs.webkit.org/show_bug.cgi?id=116334
Reviewed by Alexey Proskuryakov.
Source/WebKit2:
Add shouldNotifyOnFormChanges and didAssociateFormControls to the WKBundlePageFormClient.
- WebProcess/InjectedBundle/API/c/WKBundlePage.h:
Add the new callbacks as part of version 2 of the WKBundlePageFormClient.
- WebProcess/InjectedBundle/InjectedBundlePageFormClient.cpp:
(WebKit::InjectedBundlePageFormClient::didAssociateFormControls):
Pass the message along to the client if the client has a handler.
(WebKit::InjectedBundlePageFormClient::shouldNotifyOnFormChanges):
Ditto.
- WebProcess/InjectedBundle/InjectedBundlePageFormClient.h:
- WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::didAssociateFormControls):
Tell the injected bundle form client for the page.
(WebKit::WebChromeClient::shouldNotifyOnFormChanges):
Ditto.
- WebProcess/WebCoreSupport/WebChromeClient.h:
Tools:
Add tests for the new callbacks.
- TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
Add DidAssociateFormControls/_Bundle.cpp and associate-form-controls.html
- TestWebKitAPI/Tests/WebKit2/DidAssociateFormControls.cpp: Added.
(TestWebKitAPI::nullJavaScriptCallback):
A "null" callback to handle the fact that WKPageRunJavaScriptInMainFrame cannot handle null
being passed in for the callback.
(TestWebKitAPI::didReceiveMessageFromInjectedBundle):
After receiving the message that didAssociateFormControls callback was invoked from adding
the form in the onload handler, tell the page to add a password field to the form, which
should also invoke didAssociateFormControls callback.
(TestWebKitAPI::setInjectedBundleClient):
Register to receive messages.
(TestWebKitAPI::TEST):
Load associate-form-controls.html and wait until the didAssociateFormControls callback has
been invoked for both adding the form and for adding a password field to the form.
- TestWebKitAPI/Tests/WebKit2/DidAssociateFormControls_Bundle.cpp: Added.
(TestWebKitAPI::shouldNotifyOnFormChanges):
Return true so the didAssociateFormControls callback is invoked.
(TestWebKitAPI::didAssociateFormControls):
Tell the UI process.
(TestWebKitAPI::DidAssociateFormControlsTest::DidAssociateFormControlsTest):
(TestWebKitAPI::DidAssociateFormControlsTest::didCreatePage):
Register for the shouldNotifyOnFormChanges and didAssociateFormControls callbacks.
- TestWebKitAPI/Tests/WebKit2/associate-form-controls.html: Added.
Add a form in response to the onload event. Add a button that will add the password field
for manual testing.
- 19:04 Changeset [150397] by
-
REGRESSION(r150386): [WK2] loader/go-back-cached-main-resource.html fails
https://bugs.webkit.org/show_bug.cgi?id=116491
Add a test expectation.
- platform/wk2/TestExpectations:
- 18:53 Changeset [150396] by
-
Avoid caret repaints if we're not showing carets anyway.
<http://webkit.org/b/116489>
Reviewed by Simon Fraser.
When moving the selection, check if we're inside a contenteditable element (or in caret browsing mode)
before repainting the caret's previous location.
- editing/FrameSelection.cpp:
(WebCore::FrameSelection::recomputeCaretRect):
- 18:43 Changeset [150395] by
-
position:sticky should stick for the enclosing overflow ancestor
https://bugs.webkit.org/show_bug.cgi?id=100054
Source/WebCore:
Reviewed by Beth Dakin.
Make position: -webkit-sticky be constrained by an enclosing ancestor
with non-visible overflow if there is one, rather than being constained
always by the viewport.
Test: fast/css/sticky/sticky-top-overflow.html
- rendering/RenderBoxModelObject.cpp:
(WebCore::RenderBoxModelObject::computeStickyPositionConstraints): Rename
viewportRect to constrainingRect, since it isn't just about the viewport any more.
Fix an issue where just adding stickyLocation to absContainerFrame.location() was
incorrect when the container is scrolled; we also have to take the scroll offset
into account, and that offset is stored in the layer tree.
(WebCore::RenderBoxModelObject::stickyPositionOffset):
Look for an enclosing layer with overflow clipping, and, if found, use it to
compute the sticky constraint rect.
- rendering/RenderLayer.h: Add a helpful IncludeSelfOrNot, and fix several member functions
to use it place of hard-to-read boolean arguments.
Add enclosingOverflowClipLayer(IncludeSelfOrNot).
(WebCore::RenderLayer::ancestorCompositingLayer):
- rendering/RenderLayer.cpp:
(WebCore::RenderLayer::enclosingOverflowClipLayer): New function, finds the enclosing
layer with overflow clip.
(WebCore::RenderLayer::enclosingCompositingLayer): Use IncludeSelfOrNot.
(WebCore::RenderLayer::enclosingCompositingLayerForRepaint): Ditto.
(WebCore::RenderLayer::enclosingFilterLayer): Ditto.
- rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::repaintInCompositedAncestor): Use IncludeSelfOrNot.
(WebCore::isViewportConstrainedFixedOrStickyLayer): Renamed from isRootmostFixedOrStickyLayer
and moved up so we can use it in requiresCompositingForPosition().
(WebCore::RenderLayerCompositor::requiresCompositingForPosition): Only make sticky composited if
it's viewport-constrained.
(WebCore::RenderLayerCompositor::updateViewportConstraintStatus):
(WebCore::RenderLayerCompositor::computeStickyViewportConstraints): Declare StickyPositionViewportConstraints
one line down to just before it gets used.
LayoutTests:
Reviewed by Beth Dakin.
Reference test for sticky behavior inside overflow:.
- fast/css/sticky/sticky-top-overflow-expected.html: Added.
- fast/css/sticky/sticky-top-overflow.html: Added.
- 18:25 Changeset [150394] by
-
Remove DISABLE_ROUNDED_CORNER_CLIPPING
https://bugs.webkit.org/show_bug.cgi?id=115531
Patch by Jeff Rogers <jrogers@blackberry.com> on 2013-05-20
Reviewed by Andreas Kling.
BlackBerry port is no longer using this.
- rendering/RenderLayer.cpp:
(WebCore::inContainingBlockChain):
(WebCore::RenderLayer::clipToRect):
- 17:57 Changeset [150393] by
-
Add a test that inserting a character inside a URL doesn't move caret per auto correction
https://bugs.webkit.org/show_bug.cgi?id=116458
Reviewed by Darin Adler.
Add a test. This is broken on Safari 6 but fixed on ToT.
- editing/inserting/inserting-slash-inside-url-with-smart-link-expected.txt: Added.
- editing/inserting/inserting-slash-inside-url-with-smart-link.html: Added.
- platform/mac-wk2/TestExpectations:
- 17:49 Changeset [150392] by
-
[GTK][WK2] Unflag tests related to layoutTestController.setSmartInsertDeleteEnabled
https://bugs.webkit.org/show_bug.cgi?id=116433
Unreviewed gardening.
Patch by Manuel Rego Casasnovas <rego@igalia.com> on 2013-05-20
- platform/gtk-wk2/TestExpectations: Unflag tests as they are passing
since r150289.
- 17:24 Changeset [150391] by
-
[EFL][JHBUILD] Make sure eet is built before evas
https://bugs.webkit.org/show_bug.cgi?id=116440
Reviewed by Gyuyoung Kim.
Mark eet as a dependency of evas since evas requires eet to enable some of its
functionality.
- efl/jhbuild.modules:
- 17:10 Changeset [150390] by
-
[Mac] [WK2] storage/websql/open-database-creation-callback-isolated-world.html, statement-error-callback-isolated-world.html,
and transaction-success-callback-isolated-world.html are flaky
https://bugs.webkit.org/show_bug.cgi?id=116478
Add flaky test expectations.
- platform/mac-wk2/TestExpectations:
- 17:07 Changeset [150389] by
-
svg/batik/text/textEffect.svg is flaky on Mountain Lion builders
https://bugs.webkit.org/show_bug.cgi?id=116477
Add a failing test expectation.
- platform/mac/TestExpectations:
- 16:43 Changeset [150388] by
-
Clients should have a way to extend rendering suppression
https://bugs.webkit.org/show_bug.cgi?id=116463
<rdar://problem/13738496>
Reviewed by Andy Estes.
- WebProcess/InjectedBundle/API/c/WKBundlePage.cpp:
(WKBundlePageExtendIncrementalRenderingSuppression): Added.
(WKBundlePageStopExtendingIncrementalRenderingSuppression): Added.
- WebProcess/InjectedBundle/API/c/WKBundlePagePrivate.h:
These only have any effect if SuppressesIncrementalRendering is also enabled.
- WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::transitionToCommittedForNewPage):
Inform the FrameView if there are any outstanding rendering suppression tokens.
- WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::WebPage):
(WebKit::WebPage::extendIncrementalRenderingSuppression):
Return the next available rendering suppression token.
Add it to the list of active tokens.
Inform the FrameView that we want to suppress rendering.
(WebKit::WebPage::stopExtendingIncrementalRenderingSuppression):
Remove the given token from the list of active tokens.
Un-suppress rendering if appropriate.
- WebProcess/WebPage/WebPage.h:
(WebKit::WebPage::shouldExtendIncrementalRenderingSuppression):
Return if there are any active rendering suppression tokens.
- WebCore.exp.in:
Export FrameView::setVisualUpdatesAllowedByClient.
- dom/Document.cpp:
(WebCore::Document::setVisualUpdatesAllowed):
Move rendering-suppression watchdog into setVisualUpdatesAllowed(bool).
setVisualUpdatesAllowed(ReadyState) should not re-enable visual updates
if the client is itself disabling visual updates.
(WebCore::Document::visualUpdatesSuppressionTimerFired):
If the watchdog fires and the client is still disabling visual
updates, we should not re-enable visual updates, but instead
wait for the client.
(WebCore::Document::setVisualUpdatesAllowedByClient):
Actually re-enable visual updates if we deferred the re-enabling above
(when the page finishes loading, but the client had them disabled).
- dom/Document.h:
(Document): Add setVisualUpdatesAllowedByClient.
- page/FrameView.cpp:
(WebCore::FrameView::FrameView): Visual updates are allowed by default.
(WebCore::FrameView::setVisualUpdatesAllowedByClient):
Forward through to the document, but we also need to persist the state
here in FrameView so that it survives through navigation.
- page/FrameView.h:
(WebCore::FrameView::setVisualUpdatesAllowedByClient): Added.
(WebCore::FrameView::visualUpdatesAllowedByClient): Added.
Added storage for m_visualUpdatesAllowedByClient.
- 16:30 Changeset [150387] by
-
[CSS Exclusions] Add CSS parsing support for image URI shape-inside and shape-outside values
https://bugs.webkit.org/show_bug.cgi?id=116349
Reviewed by Alexandru Chiculita.
Source/WebCore:
Add support for parsing image URI values for the shape-inside and shape-outside CSS properties.
Extended the ExclusionShapeValue class to support tracking the property's StyleImage value.
The ExclusionShapeValue::type enums were changed from all upper case to "Intercase" by order
of the style bot.
The existing CSS parsing tests were extended to check URI values.
- css/CSSComputedStyleDeclaration.cpp:
(WebCore::CSSComputedStyleDeclaration::getPropertyCSSValue):
- css/CSSParser.cpp:
(WebCore::CSSParser::parseValue):
- css/DeprecatedStyleBuilder.cpp:
(WebCore::ApplyPropertyExclusionShape::applyValue):
- css/StyleResolver.cpp:
(WebCore::StyleResolver::loadPendingImages):
- rendering/style/ExclusionShapeValue.h:
(WebCore::ExclusionShapeValue::createImageValue):
(ExclusionShapeValue):
(WebCore::ExclusionShapeValue::image):
(WebCore::ExclusionShapeValue::setImage):
(WebCore::ExclusionShapeValue::ExclusionShapeValue):
LayoutTests:
Check that shape values specified with a URI are parsed correctly.
- fast/exclusions/parsing-wrap-shape-inside-expected.txt:
- fast/exclusions/parsing-wrap-shape-outside-expected.txt:
- fast/exclusions/script-tests/parsing-wrap-shape-inside.js:
- fast/exclusions/script-tests/parsing-wrap-shape-outside.js:
- 16:28 Changeset [150386] by
-
WTR::pathSuitableForTestResult should behave the same as _drt_descriptionSuitableForTestResult so we can unskip tests.
https://bugs.webkit.org/show_bug.cgi?id=116125
Patch by Alex Christensen <achristensen@apple.com> on 2013-05-20
Reviewed by Tim Horton.
Tools:
- WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::pathSuitableForTestResult):
Changed string operations to match _drt_descriptionSuitableForTestResult.
(WTR::dumpRequestDescriptionSuitableForTestResult):
(WTR::dumpResponseDescriptionSuitableForTestResult):
(WTR::InjectedBundlePage::willPerformClientRedirectForFrame):
(WTR::InjectedBundlePage::didInitiateLoadForResource):
(WTR::InjectedBundlePage::willSendRequestForFrame):
(WTR::InjectedBundlePage::didReceiveResponseForResource):
Pass the main frame's URL to pathSuitableForTestResult.
- WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
Added WKBundlePageRef to willPerformClientRedirectForFrame.
LayoutTests:
- platform/wk2/TestExpectations:
Unskipped working tests.
- 16:05 Changeset [150385] by
-
Add a failing test expectation to editing/selection/user-drag-element-and-user-select-none.html
after r150354. The failure is tracked by the bug 116473.
- 15:55 Changeset [150384] by
-
PPT: Closing tab that is hung or chewing 100% CPU leaves abandoned WebProcess.
<http://webkit.org/b/116464>
<rdar://problem/10103795>
Reviewed by Anders Carlsson.
Give the web process a 10 second chance to exit nicely after closing the last tab belonging to it.
This code only runs if there was something on the page (e.g an unload/beforeunload event handler)
preventing the UI process from killing it right away.
- UIProcess/WebProcessProxy.cpp:
(WebKit::WebProcessProxy::WebProcessProxy):
(WebKit::WebProcessProxy::removeWebPage):
(WebKit::WebProcessProxy::forcefulTerminationTimerFired):
- UIProcess/WebProcessProxy.h:
(WebProcessProxy):
- 14:38 Changeset [150383] by
-
[CSSRegions] Fix offsetLeft / offsetTop for elements inside named flow
https://bugs.webkit.org/show_bug.cgi?id=115899
Patch by Radu Stavila <stavila@adobe.com> on 2013-05-20
Reviewed by David Hyatt.
Source/WebCore:
Elements in named flows that have the body as their offsetParent, need to compute their
offsetLeft and offsetTop values relative to the body.
Tests: fast/regions/offsetLeft-offsetTop-in-multiple-regions.html
fast/regions/offsetLeft-offsetTop-in-region-absolute-sticky-fixed.html
fast/regions/offsetLeft-offsetTop-in-region-float-vert-rl.html
fast/regions/offsetLeft-offsetTop-in-region-float.html
fast/regions/offsetLeft-offsetTop-inlines-region-in-element.html
- rendering/RenderBoxModelObject.cpp:
(WebCore::RenderBoxModelObject::adjustedPositionRelativeToOffsetParent):
- rendering/RenderFlowThread.cpp:
(WebCore):
(WebCore::RenderFlowThread::adjustedPositionRelativeToOffsetParent):
- rendering/RenderFlowThread.h:
LayoutTests:
Added tests for the implementation of offsetLeft and offsetTop for objects inside named flows
for which the offsetParent is the body.
- fast/regions/offsetLeft-offsetTop-in-multiple-regions-expected.txt: Added.
- fast/regions/offsetLeft-offsetTop-in-multiple-regions.html: Added.
- fast/regions/offsetLeft-offsetTop-in-region-absolute-sticky-fixed-expected.txt: Added.
- fast/regions/offsetLeft-offsetTop-in-region-absolute-sticky-fixed.html: Added.
- fast/regions/offsetLeft-offsetTop-in-region-float-expected.txt: Added.
- fast/regions/offsetLeft-offsetTop-in-region-float-vert-rl-expected.txt: Added.
- fast/regions/offsetLeft-offsetTop-in-region-float-vert-rl.html: Added.
- fast/regions/offsetLeft-offsetTop-in-region-float.html: Added.
- fast/regions/offsetLeft-offsetTop-inlines-region-in-element-expected.txt: Added.
- fast/regions/offsetLeft-offsetTop-inlines-region-in-element.html: Added.
- 14:17 Changeset [150382] by
-
[Windows] Unreviewed TestExpectation cleanup. Reordered lines
to match Mac version to make it easier to understand what features
are uniquely missing in Windows port.
- platform/win/TestExpectations: Reordering of lines, adds some
bug ID's to existing skipped tests.
- 14:10 Changeset [150381] by
-
Make C API more robust against null contexts
https://bugs.webkit.org/show_bug.cgi?id=116462
Reviewed by Anders Carlsson.
Handle null contexts in a non-crashy way. It's a bug to ever call the
API with a null context, and the absence of a context means we can't
produce a meaningful result, so we still assert in debug builds.
Now where possible we detect and early return, returning null for any
pointer type, NaN for doubles, and false for any boolean result.
- API/JSBase.cpp:
(JSEvaluateScript):
(JSCheckScriptSyntax):
(JSReportExtraMemoryCost):
- API/JSContextRef.cpp:
(JSContextGetGlobalObject):
(JSContextGetGroup):
(JSContextGetGlobalContext):
(JSContextCreateBacktrace):
- API/JSObjectRef.cpp:
(JSObjectMake):
(JSObjectMakeFunctionWithCallback):
(JSObjectMakeConstructor):
(JSObjectMakeFunction):
(JSObjectMakeArray):
(JSObjectMakeDate):
(JSObjectMakeError):
(JSObjectMakeRegExp):
(JSObjectGetPrototype):
(JSObjectSetPrototype):
(JSObjectHasProperty):
(JSObjectGetProperty):
(JSObjectSetProperty):
(JSObjectGetPropertyAtIndex):
(JSObjectSetPropertyAtIndex):
(JSObjectDeleteProperty):
(JSObjectCopyPropertyNames):
- API/JSValueRef.cpp:
(JSValueGetType):
(JSValueIsUndefined):
(JSValueIsNull):
(JSValueIsBoolean):
(JSValueIsNumber):
(JSValueIsString):
(JSValueIsObject):
(JSValueIsObjectOfClass):
(JSValueIsEqual):
(JSValueIsStrictEqual):
(JSValueIsInstanceOfConstructor):
(JSValueMakeUndefined):
(JSValueMakeNull):
(JSValueMakeBoolean):
(JSValueMakeNumber):
(JSValueMakeString):
(JSValueMakeFromJSONString):
(JSValueCreateJSONString):
(JSValueToBoolean):
(JSValueToNumber):
(JSValueToStringCopy):
(JSValueToObject):
(JSValueProtect):
- API/JSWeakObjectMapRefPrivate.cpp:
- 13:42 Changeset [150380] by
-
fix w3c test importer unit tests after r150324
https://bugs.webkit.org/show_bug.cgi?id=116459
Reviewed by Ryosuke Niwa.
This patch fixes the unit tests to not actually call out to
'hg' (by using a mock executive) and also changes the real
directory we crawl from Source/WebCore/css to
Tools/Scripts/webkitpy/w3c to run much more quickly. We should
still change this to a mock filesystem, but that's a separate change.
- Scripts/webkitpy/w3c/test_importer.py:
(main):
(TestImporter.init):
(TestImporter.load_changeset):
- Scripts/webkitpy/w3c/test_importer_unittest.py:
(TestImporterTest.test_import_dir_with_no_tests_and_no_hg):
(TestImporterTest.test_import_dir_with_no_tests):
- 13:34 Changeset [150379] by
-
Add helper function for converting a KeyedCodingValue to a CFTypeRef
https://bugs.webkit.org/show_bug.cgi?id=116461
Reviewed by Tim Horton.
- Shared/cf/KeyedCodingValueCF.cpp: Added.
- Shared/cf/KeyedCodingValueCF.h: Added.
- WebKit2.xcodeproj/project.pbxproj:
- 13:00 Changeset [150378] by
-
[Qt] REGRESSION(r149457): some fast/exclusions tests are failing
https://bugs.webkit.org/show_bug.cgi?id=115511
Reviewed by Andreas Kling.
Unskipping the exclusions tests on Qt. They were fixed by https://bugs.webkit.org/show_bug.cgi?id=116115.
- platform/qt/TestExpectations:
- 12:55 Changeset [150377] by
-
Begin stubbing out a new KeyedEncoder class
https://bugs.webkit.org/show_bug.cgi?id=116456
Reviewed by Sam Weinig.
This class (and it's coming counterpart, KeyedDecoder) will be used for serializing
object graphs to a tree-like structure of dictionaries. It is up to individual ports to
serialize them further (for example, on Mac they could be serialized to CFPropertyList objects).
This will be used for serializing session storage state but I hope that it can be used for serializing
back forward state in the future as our current implementation of back forward state serialization
is fragile and depends on our IPC layer's binary encoding(!).
- Shared/KeyedCodingValue.h: Added.
- Shared/KeyedEncoder.cpp: Added.
- Shared/KeyedEncoder.h: Added.
- WebKit2.xcodeproj/project.pbxproj:
- 12:13 Changeset [150376] by
-
Unreviewed EFL build fix.
Fix linking errors due to undefined symbols on Ubuntu 13.04.
- DumpRenderTree/efl/CMakeLists.txt:
- EWebLauncher/CMakeLists.txt:
- MiniBrowser/efl/CMakeLists.txt:
- 12:00 Changeset [150375] by
-
[CSS Regions][CSS Exclusions] shape-inside on regions should respect positioned shapes and overflow
https://bugs.webkit.org/show_bug.cgi?id=116252
Reviewed by David Hyatt.
Regions should respect shape-inside properties which have specified top offset. Since the content which overflows from the
shape should be pushed after the content box (for details check out r148975), I implemented overflowthe behavior for simple
cases when you have only one region. I'm going to implement it for additional regions in a follow up patch.
Tests: fast/regions/shape-inside/shape-inside-on-regions-block-content-basic-overflow-shape-top-offset.html
fast/regions/shape-inside/shape-inside-on-regions-inline-content-basic-overflow-shape-top-offset.html
Source/WebCore:
- rendering/RenderBlockLineLayout.cpp:
(WebCore::RenderBlock::updateLineBoundariesForExclusions): Handle the flow thread case, push the overflowing content after the
content box. Respect existing shape-inside on region tests.
(WebCore::RenderBlock::layoutRunsAndFloatsInRange): Set the correct logical top position for the shape-inside in
the case of regions. Respect the region's margins and paddings.
LayoutTests:
- fast/regions/shape-inside/shape-inside-on-regions-block-content-basic-overflow-shape-top-offset-expected.html: Added.
- fast/regions/shape-inside/shape-inside-on-regions-block-content-basic-overflow-shape-top-offset.html: Added.
- fast/regions/shape-inside/shape-inside-on-regions-inline-content-basic-overflow-shape-top-offset-expected.html: Added.
- fast/regions/shape-inside/shape-inside-on-regions-inline-content-basic-overflow-shape-top-offset.html: Added.
- 11:45 Changeset [150374] by
-
[GTK] Cleanup TestExpectations to pass --lint-test-files
https://bugs.webkit.org/show_bug.cgi?id=116419
Unreviewed GTK gardening.
Remove one of two http/tests/cache/willsendrequest-returns-null-for-memory-cache-load.html
Patch by Seokju Kwon <seokju.kwon@gmail.com> on 2013-05-20
- platform/gtk/TestExpectations:
- 11:10 Changeset [150373] by
-
Remove custom binding for the Clipboard clearData function
https://bugs.webkit.org/show_bug.cgi?id=116421
Reviewed by Andreas Kling.
- bindings/js/JSClipboardCustom.cpp: Removed the clearData binding.
- dom/Clipboard.idl: Removed [Custom].
- dom/Clipboard.cpp:
(WebCore::Clipboard::clearData): Renamed from clearAllData.
- dom/Clipboard.h: Rename clearAllData to clearData so we just use
overloading for the one with and without a MIME type.
- platform/ios/ClipboardIOS.h: Rename clearAllData to clearData.
- platform/ios/ClipboardIOS.mm:
(WebCore::ClipboardIOS::clearData): Ditto.
- platform/qt/ClipboardQt.cpp:
(WebCore::ClipboardQt::clearData). Ditto.
- platform/qt/ClipboardQt.h: Ditto.
- platform/win/ClipboardWin.cpp:
(WebCore::ClipboardWin::clearData): Ditto.
- platform/win/ClipboardWin.h: Ditto.
- 10:21 Changeset [150372] by
-
[iOS] media/event-queue-crash.html ASSERTs in notifyChildInserted
https://bugs.webkit.org/show_bug.cgi?id=116444
Reviewed by Jer Noble.
No new tests, covered by existing tests.
- html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::configureMediaControls): Only create media controls when the
element is in a Document.
- 09:47 Changeset [150371] by
-
[Mac] captions menu should behave more like a menu
https://bugs.webkit.org/show_bug.cgi?id=116436
Reviewed by Eric Carlson.
Let the captions menu behave more like a native Mac menu by hiding it when
clicking anywhere on the page and preventing the page from scrolling when
attempting to scroll the captions menu when it cannot scroll in the requested
direction.
Test: media/video-controls-captions-trackmenu-hide-on-click-outside.html
- dom/EventListener.h:
Add the new MediaControlsAppleEventListenerType.
- html/shadow/MediaControlsApple.cpp:
(WebCore::MediaControlsApple::defaultEventHandler):
(WebCore::MediaControlsApple::hide):
(WebCore::MediaControlsApple::makeTransparent):
(WebCore::MediaControlsApple::changedClosedCaptionsVisibility):
(WebCore::MediaControlsApple::reportedError):
(WebCore::MediaControlsApple::toggleClosedCaptionTrackList):
Use the hideClosedCaptionTrackList() and showClosedCaptionTrackList()
methods to hide and show the captions menu instead of calling hide()
and show() directly on m_closedCaptionsContainer.
(WebCore::MediaControlsApple::showClosedCaptionTrackList):
Show the m_closedCaptionsContainer and register a "mousewheel" event
listener on it as well as a "click" event listener on the entire document.
(WebCore::MediaControlsApple::hideClosedCaptionTrackList):
Hide the m_closedCaptionsContainer and remove the "mousewheel" event
listener on it as well as the "click" event listener on the entire document.
(WebCore::MediaControlsApple::shouldClosedCaptionsContainerPreventPageScrolling):
New private utility to determine whether the m_closedCaptionsContainer can scroll
in the provided scroll direction.
(WebCore::MediaControlsApple::eventListener):
Obtain the event listener used for "mousewheel" and "click" event handlers.
(WebCore::MediaControlsAppleEventListener::handleEvent):
Event handler for the "mousewheel" and "click" events. If we get a "click" event, we
toggle the captions menu visibility and if we get a "mousewheel" event, we call into
shouldClosedCaptionsContainerPreventPageScrolling() to see if we can scroll in the
current scroll direction, and if not prevent the event from resulting in a scroll by
calling preventDefault().
(WebCore::MediaControlsAppleEventListener::operator==):
Required for the successful subclassing of EventListener.
- html/shadow/MediaControlsApple.h:
(MediaControlsAppleEventListener):
(WebCore::MediaControlsAppleEventListener::create):
(WebCore::MediaControlsAppleEventListener::cast):
(WebCore::MediaControlsAppleEventListener::MediaControlsAppleEventListener):
New subclass of EventListener required to provide a custom event listener for the
"mousewheel" and "click" events registered in showClosedCaptionTrackList() and
hideClosedCaptionTrackList().
- 09:39 Changeset [150370] by
-
[Cairo] Canvas-shadow behavior is not being as expected
https://bugs.webkit.org/show_bug.cgi?id=108897
Patch by Rashmi Shyamasundar <rashmi.s2@samsung.com> on 2013-05-18
Reviewed by Martin Robinson.
Source/WebCore:
ShadowBlur::endShadowLayer copies the image from shadowContext to cairoContext.
CairoContext-path should be empty for doing this copy. Otherwise, the
original-image area will also get filled with the shadow.
Test: fast/canvas/canvas-image-shadow.html
- platform/graphics/cairo/GraphicsContextCairo.cpp:
(WebCore::drawPathShadow):
LayoutTests:
Test to verify the shadow of an image drawn on canvas.
This test uses an image whose size is smaller than,
the size of the rectangle which should be filled with the image.
- fast/canvas/canvas-image-shadow-expected.txt: Added.
- fast/canvas/canvas-image-shadow.html: Added.
- fast/canvas/script-tests/canvas-image-shadow.js: Added.
(draw):
- 09:21 Changeset [150369] by
-
Bump WKBundlePageLoaderClient version number so as not to break some clients
https://bugs.webkit.org/show_bug.cgi?id=116426
Reviewed by Anders Carlsson.
- Shared/APIClientTraits.cpp:
(WebKit):
- Shared/APIClientTraits.h:
- WebProcess/InjectedBundle/API/c/WKBundlePage.h:
- 09:00 Changeset [150368] by
-
[EFL] Reenabled INDEXED_DATABASE after r150344
https://bugs.webkit.org/show_bug.cgi?id=116430
Reviewed by Gyuyoung Kim.
.:
Reenable INDEXED_DATABASE flag for EFL port now that WebKit2
build was fixed in r150344.
- Source/cmake/OptionsEfl.cmake:
Tools:
Reenable INDEXED_DATABASE flag for EFL port now that WebKit2
build was fixed in r150344.
- Scripts/webkitperl/FeatureList.pm:
LayoutTests:
Unskip indexeddb layout tests for EFL port now that the functionality
is reenabled at compile time.
- platform/efl/TestExpectations:
- 07:44 Changeset [150367] by
-
Synchronize FeatureDefines.xcconfig
Source/JavaScriptCore:
- Configurations/FeatureDefines.xcconfig: Remove
ENABLE_LINK_PRERENDER. This was missed in r150356.
Source/WebKit/mac:
- Configurations/FeatureDefines.xcconfig: Add a newline to the
end of the file.
- 06:44 Changeset [150366] by
-
-webkit-text-underline-position should not be inherited
https://bugs.webkit.org/show_bug.cgi?id=116363
Patch by Lamarque V. Souza <Lamarque.Souza@basyskom.com> on 2013-05-20
Reviewed by Dean Jackson.
Source/WebCore:
Specification says text-underline-position should not be inherited.
No new tests, this updates existing tests.
- css/CSSProperty.cpp:
(WebCore::CSSProperty::isInheritedProperty): Make
CSSPropertyWebkitTextUnderlinePosition return false.
- rendering/style/RenderStyle.h: Treat TextUnderlinePosition as
non-inherited.
- rendering/style/StyleRareInheritedData.cpp:
(WebCore::StyleRareInheritedData::StyleRareInheritedData):
(WebCore::StyleRareInheritedData::operator==): Remove m_textUnderlinePosition.
- rendering/style/StyleRareInheritedData.h:
(StyleRareInheritedData):
- rendering/style/StyleRareNonInheritedData.cpp:
(WebCore::StyleRareNonInheritedData::StyleRareNonInheritedData):
(WebCore::StyleRareNonInheritedData::operator==): Add m_textUnderlinePosition.
- rendering/style/StyleRareNonInheritedData.h:
(StyleRareNonInheritedData):
LayoutTests:
Update expected results.
- fast/css3-text/css3-text-decoration/getComputedStyle/getComputedStyle-text-underline-position-expected.txt:
- fast/css3-text/css3-text-decoration/getComputedStyle/script-tests/getComputedStyle-text-underline-position.js:
- 06:36 Changeset [150365] by
-
Add compile asserts for the size of RootInlineBox and InlineTextbox
https://bugs.webkit.org/show_bug.cgi?id=116337
Patch by Lamarque V. Souza <Lamarque.Souza@basyskom.com> on 2013-05-20
Reviewed by Darin Adler.
Add compile asserts to ensure RootInlineBox and InlineTextbox stay small.
- rendering/InlineTextBox.cpp:
(SameSizeAsInlineTextBox):
(WebCore):
- rendering/RootInlineBox.cpp:
(SameSizeAsRootInlineBox):
(WebCore):
- 05:03 Changeset [150364] by
-
Unreviewed EFL gardening.
Adding failure expectations for spelling tests.
- platform/efl-wk2/TestExpectations:
- 03:45 WebKit Team edited by
- (diff)
- 03:11 Changeset [150363] by
-
Unreviewed EFL gardening.
spelling-double-clicked-word.html passes after r150289.
- platform/efl-wk2/TestExpectations:
- 02:37 Changeset [150362] by
-
[BlackBerry] LocalizedStringsBlackBerry: add dummy text track strings
https://bugs.webkit.org/show_bug.cgi?id=116300
Patch by Xan Lopez <xlopez@igalia.com> on 2013-05-20
Reviewed by Rob Buis.
These are dummy implementations necessary to make WebKit link.
- platform/blackberry/LocalizedStringsBlackBerry.cpp:
(WebCore::weekFormatInLDML):
(WebCore):
(WebCore::textTrackClosedCaptionsText):
(WebCore::textTrackSubtitlesText):
(WebCore::textTrackOffText):
(WebCore::textTrackNoLabelText):
- 01:30 Changeset [150361] by
-
[BlackBerry] Do not use skia in FrameLoaderClientBlackBerry::dispatchDidReceiveIcon()
https://bugs.webkit.org/show_bug.cgi?id=116302
Patch by Carlos Garcia Campos <cgarcia@igalia.com> on 2013-05-20
Reviewed by Rob Buis.
TiledImage is now the NativeImage of the BlackBerry port.
- WebCoreSupport/FrameLoaderClientBlackBerry.cpp:
(WebCore::FrameLoaderClientBlackBerry::dispatchDidReceiveIcon):
- 01:02 EFLWebKit edited by
- buildbots (diff)
- 01:01 EFLWebKitBuildBots edited by
- (diff)
- 01:00 EFLWebKitBuildBots edited by
- removing cq requestees. (diff)
- 01:00 EFLWebKitBuildBots edited by
- removing intel gardeners (diff)
- 00:24 WebKitGTK/2.0.x edited by
- (diff)
