Should support value profiling on CPU(X86)
https://bugs.webkit.org/show_bug.cgi?id=68575

Reviewed by Sam Weinig.

Fix verbose profiling in ToT (SlowCaseProfile had been
partially renamed to RareCaseProfile), add in-memory
bucket counter for CPU(X86), move JIT::m_canBeOptimized
out of the DFG_JIT ifdef.

* bytecode/CodeBlock.cpp:
(JSC::CodeBlock::resetRareCaseProfiles):
(JSC::CodeBlock::dumpValueProfiles):
* bytecode/CodeBlock.h:
* dfg/DFGByteCodeParser.cpp:
(JSC::DFG::ByteCodeParser::makeSafe):
* jit/JIT.cpp:
(JSC::JIT::privateCompileSlowCases):
(JSC::JIT::privateCompile):
* jit/JIT.h:
* jit/JITInlineMethods.h:
(JSC::JIT::emitValueProfilingSite):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95676 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remove checks for Leopard-only Objective-C APIs
https://bugs.webkit.org/show_bug.cgi?id=68579

Reviewed by Sam Weinig.

We can always assume that OBJC_API_VERSION is at least 2, it's been 2 since Leopard.

../WebCore:

* bridge/objc/objc_class.mm:
(JSC::Bindings::ObjcClass::methodsNamed):
(JSC::Bindings::ObjcClass::fieldNamed):
* bridge/objc/objc_instance.mm:
(allocateAutoReleasePool):
* bridge/objc/objc_runtime.mm:
(JSC::Bindings::ObjcField::ObjcField):
* page/mac/EventHandlerMac.mm:

../WebKit/mac:

* Misc/WebNSObjectExtras.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95675 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fix resolve-Changelog's strange butchery of the LayoutTests/Changelog.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95674 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remove Widget::beforeMouseDown and Widget::afterMouseDown
https://bugs.webkit.org/show_bug.cgi?id=68570

Reviewed by Sam Weinig.

These two member functions were originally added to avoid crashes due to removing NSViews while they were
being tracked by AppKit. Since they were added, we've moved away from NSViews for form controls, and the bugs in
AppKit that lead to crashes have been fixed.

* page/mac/EventHandlerMac.mm:
(WebCore::EventHandler::passMouseDownEventToWidget):
* platform/Widget.h:
* platform/mac/WidgetMac.mm:
(WebCore::Widget::Widget):
(WebCore::Widget::removeFromSuperview):

* rendering/RenderWidget.cpp:
Change widgetHierarchyUpdateSuspendCount to be an unsigned integer instead of a size_t.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95673 268f45cc-cd09-0410-ab3c-d52691b4dbfc

DFG does not support compiling functions as constructors
https://bugs.webkit.org/show_bug.cgi?id=68500

Reviewed by Oliver Hunt.

This adds support for compiling constructors to the DFG. It's a
1% speed-up on V8, mostly due to a 6% speed-up on early-boyer.
It's also a 13% win on access-binary-trees, but it's neutral in
the SunSpider and Kraken averages.

* dfg/DFGByteCodeParser.cpp:
(JSC::DFG::ByteCodeParser::parseBlock):
* dfg/DFGCapabilities.h:
(JSC::DFG::mightCompileFunctionForConstruct):
(JSC::DFG::canCompileOpcode):
* dfg/DFGNode.h:
* dfg/DFGOperations.cpp:
* dfg/DFGOperations.h:
* dfg/DFGPropagator.cpp:
(JSC::DFG::Propagator::propagateNodePredictions):
(JSC::DFG::Propagator::performNodeCSE):
* dfg/DFGSpeculativeJIT.cpp:
(JSC::DFG::SpeculativeJIT::compile):
* runtime/Executable.cpp:
(JSC::FunctionExecutable::compileOptimizedForConstruct):
(JSC::FunctionExecutable::compileForConstructInternal):
* runtime/Executable.h:
(JSC::FunctionExecutable::compileForConstruct):
(JSC::FunctionExecutable::compileFor):
(JSC::FunctionExecutable::compileOptimizedFor):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95672 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Source/WebCore: Not use anonymousContainer on beforeChild calculation
and let later code handle it. Fix style update on block-level
anonymous objects.
https://bugs.webkit.org/show_bug.cgi?id=67656

Reviewed by David Hyatt.

Test: fast/table/quote-text-around-iframe.html

* rendering/RenderTable.cpp:
(WebCore::RenderTable::addChild):
* rendering/RenderTableCell.cpp:
(WebCore::RenderTableCell::styleDidChange):
* rendering/RenderTableRow.cpp:
(WebCore::RenderTableRow::addChild):
* rendering/RenderObject.cpp:
(WebCore::RenderObject::propagateStyleToAnonymousChildren):

LayoutTests: Not use anonymousContainer on beforeChild calculation
and let later code handle it.
https://bugs.webkit.org/show_bug.cgi?id=67656

Reviewed by David Hyatt.

* fast/table/quote-text-around-iframe.html: Added.
* fast/table/table-row-style-not-updated-with-after-content-expected.txt:
* platform/mac/fast/table/quote-text-around-iframe-expected.png: Added.
* platform/mac/fast/table/quote-text-around-iframe-expected.txt: Added.
and lot of style propagation rebaselines.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95671 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Source/WebCore: https://bugs.webkit.org/show_bug.cgi?id=68562

Pagination struts don't update properly when collapsing margin estimates are wrong.

Refactor the pagination adjustments that happen during block layout to allow for
a 3-pass layout system instead of just a 2-pass layout system. The basic idea is
to come up with an estimate that includes collapsing, clearance and pagination.

If the estimate only with collapsing and clearance was wrong, then we lay out again
using the correct collapsing margin values and clearance before applying pagination breaks.

After this second layout, the normal code that runs when unpaginated compares estimates
and lays out a third time to account for adjustments as a result of pagination movement.

Reviewed by Sam Weinig.

Added fast/regions/text-regions-split-small-pagination.html

* rendering/RenderBlock.cpp:
(WebCore::RenderBlock::estimateLogicalTopPosition):
(WebCore::RenderBlock::layoutBlockChild):
(WebCore::RenderBlock::adjustBlockChildForPagination):
* rendering/RenderBlock.h:

LayoutTests: https://bugs.webkit.org/show_bug.cgi?id=68562

Pagination struts don't update properly when collapsing margin estimates are wrong.

Add a new regions result to test the bug fix.

Reviewed by Sam Weinig.

* fast/regions/text-region-split-small-pagination.html: Added.
* platform/mac/fast/regions/text-region-split-small-pagination-expected.png: Added.
* platform/mac/fast/regions/text-region-split-small-pagination-expected.txt: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95670 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remove old and busted context notification APIs
https://bugs.webkit.org/show_bug.cgi?id=68381

Reviewed by David Levin.

* public/WebFrameClient.h:
* src/FrameLoaderClientImpl.cpp:
(WebKit::FrameLoaderClientImpl::didCreateScriptContext):
(WebKit::FrameLoaderClientImpl::willReleaseScriptContext):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95669 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[chromium] SharedGraphicsContext3D's context should not be destroyed during atexit
https://bugs.webkit.org/show_bug.cgi?id=68558

Reviewed by Stephen White.

The SharedGraphicsContext3D's internal context is lazily created and then leaked intentionally, but since the
last ref is in a function-static RefPtr<> it ends up being destroyed at process exit time.  This is bad, since
by the time we get this far in process shutdown we can't actually cleanly run the destruction logic and we
crash.  Since we are explicitly leaking this for the lifetime of the process there's no point in keeping it in a
RefPtr<>

* platform/graphics/gpu/SharedGraphicsContext3D.cpp:
(WebCore::SharedGraphicsContext3D::get):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95668 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Chromium] Protect the Frame in V8HTMLDocument::openCallback
https://bugs.webkit.org/show_bug.cgi?id=68555

Patch by Sergey Glazunov <serg.glazunov@gmail.com> on 2011-09-21
Reviewed by Nate Chapin.

Source/WebCore:

Test: fast/dom/frame-deleted-in-document-open.html

* bindings/v8/custom/V8HTMLDocumentCustom.cpp:
(WebCore::V8HTMLDocument::openCallback):

LayoutTests:

* fast/dom/frame-deleted-in-document-open-expected.txt: Added.
* fast/dom/frame-deleted-in-document-open.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95667 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Replace jsFunctionVPtr compares with a type check on the Structure.
https://bugs.webkit.org/show_bug.cgi?id=68557

Reviewed by Oliver Hunt.

This will permit calls to still optimize to subclasses of JSFunction
that have the correct type (but a different C++ vptr).

This patch stops passing the globalData into numerous functions.

* dfg/DFGByteCodeParser.cpp:
(JSC::DFG::ByteCodeParser::parseBlock):
* dfg/DFGGraph.h:
(JSC::DFG::Graph::isFunctionConstant):
(JSC::DFG::Graph::valueOfFunctionConstant):
* dfg/DFGJITCompiler.h:
(JSC::DFG::JITCompiler::isFunctionConstant):
(JSC::DFG::JITCompiler::valueOfFunctionConstant):
* dfg/DFGOperations.cpp:
* interpreter/Interpreter.cpp:
(JSC::Interpreter::privateExecute):
* jit/JIT.h:
* jit/JITCall.cpp:
(JSC::JIT::compileOpCallVarargs):
(JSC::JIT::compileOpCallSlowCase):
* jit/JITCall32_64.cpp:
(JSC::JIT::compileOpCallVarargs):
(JSC::JIT::compileOpCallSlowCase):
* jit/JITInlineMethods.h:
(JSC::JIT::emitJumpIfNotType):
* jit/JITStubs.cpp:
(JSC::DEFINE_STUB_FUNCTION):
* runtime/Executable.h:
(JSC::isHostFunction):
* runtime/JSFunction.h:
(JSC::JSFunction::createStructure):
* runtime/JSObject.cpp:
(JSC::JSObject::put):
(JSC::JSObject::putWithAttributes):
* runtime/JSObject.h:
(JSC::getJSFunction):
(JSC::JSObject::putDirect):
(JSC::JSObject::putDirectWithoutTransition):
* runtime/JSType.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95666 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed. Remove unused ENABLE_AS_IMAGE flag after r95234 from
Options{Efl,WinCE}.cmake.

* Source/cmake/OptionsEfl.cmake:
* Source/cmake/OptionsWinCE.cmake:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95665 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Clean up CSS Counter code
https://bugs.webkit.org/show_bug.cgi?id=68492

Reviewed by Tony Chang.

Source/WebCore:

Several cleanups/fixes:
- Make Counter::listStyle() return the actual list style string,
  rather than empty string.
- Handle a specified list style of "none" correctly when generating
  cssText and in Counter::listStyle().
- Use StringBuilder when generating cssText.

Test: fast/css/counters/getCounterValue.html

* css/CSSParser.cpp:
(WebCore::CSSParser::parseCounterContent):
Simplify processing (moving conversion to CSSStyleSelector),
pass a CSS_IDENT as the listStyle to Counter::create.
* css/CSSPrimitiveValue.cpp:
(WebCore::CSSPrimitiveValue::cssText):
Use StringBuilder.
* css/CSSStyleSelector.cpp:
(WebCore::CSSStyleSelector::applyProperty):
Convert from a CSS_IDENT to an EListStyleType.
* css/Counter.h:
(WebCore::Counter::listStyleIdent):
Renamed from listStyleNumber to match its new functionality.

LayoutTests:

Added tests for setting the list style to "none"
as well as tests for getCounterValue.

* fast/css/counters/counter-cssText-expected.txt:
* fast/css/counters/counter-cssText.html:
* fast/css/counters/getCounterValue-expected.txt: Added.
* fast/css/counters/getCounterValue.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95664 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Removed WTFTHREADDATA_MULTITHREADED, making it always true
https://bugs.webkit.org/show_bug.cgi?id=68549

Reviewed by Darin Adler.

Another part of making threads exist in WebKit.

* wtf/WTFThreadData.cpp:
* wtf/WTFThreadData.h:
(WTF::wtfThreadData):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95663 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-09-21  Chris Marrin  <cmarrin@apple.com>

        Check in WebCore.vcproj file unchanged but written like MSDev likes it

        Unreviewed.

        * WebCode.vcproj/WebCore.vcproj:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95662 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Use the prefix "WK" instead of "Web" for Objective-C classes defined in WebKit2.

Reviewed by Anders Carlsson.

Renamed WebInspectorProxyObjCAdapter to WKWebInspectorProxyObjCAdapter.
Renamed WebInspectorWKView to WKWebInspectorWKView.
Renamed WebFindIndicatorWindowAnimation to WKFindIndicatorWindowAnimation.
Renamed WebFindIndicatorView to WKFindIndicatorView.
Renamed WebUserDataWrapper to WKUserDataWrapper.

Updated the build phase checking for inappropriate Objective-C class names to disallow the prefix
"Web" and only allow "WK".

* UIProcess/API/mac/FindIndicatorWindow.h:
* UIProcess/API/mac/FindIndicatorWindow.mm:
(WebKit::FindIndicatorWindow::setFindIndicator):
(WebKit::FindIndicatorWindow::startFadeOutTimerFired):
* UIProcess/WebInspectorProxy.h:
* UIProcess/mac/WebContextMenuProxyMac.mm:
(-[WKMenuTarget forwardContextMenuAction:]):
(WebKit::nsMenuItemVector):
* UIProcess/mac/WebInspectorProxyMac.mm:
(WebKit::WebInspectorProxy::platformCreateInspectorPage):
(WebKit::WebInspectorProxy::platformOpen):
* WebKit2.xcodeproj/project.pbxproj:
* WebProcess/FullScreen/mac/WebFullScreenManagerMac.mm:
(WebKit::WebFullScreenManagerMac::WebFullScreenManagerMac):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95661 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[WK2] [Mac] Implement a more-complete MouseDown/MouseUp/MouseMoveTo functions for WebKit2 EventSender
https://bugs.webkit.org/show_bug.cgi?id=68108

This patch implements the event sender mouse events through WebKit2 UIProcess which is
closer to the real simulation than a WebProcess-only approach. The patch only supports Mac
platform as the first step so the existing code is still kept for other platforms.

Source/WebKit2:

Patch by Chang Shu <cshu@webkit.org> on 2011-09-21
Reviewed by Darin Adler.

* UIProcess/API/C/WKPage.cpp:
(WKPageSetShouldSendEventsSynchronously):
* UIProcess/API/C/WKPagePrivate.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::WebPageProxy):
(WebKit::WebPageProxy::handleMouseEvent):
(WebKit::WebPageProxy::handleKeyboardEvent):
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::setShouldSendEventsSynchronously):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::mouseEventSyncForTesting):
* WebProcess/WebPage/WebPage.h:
* WebProcess/WebPage/WebPage.messages.in:

Tools:

Patch by Chang Shu <cshu@webkit.org> on 2011-09-21
Reviewed by Darin Adler.

* WebKitTestRunner/EventSenderProxy.h:
(WTR::EventSenderProxy::EventSenderProxy):
(WTR::EventSenderProxy::leapForward):
(WTR::EventSenderProxy::currentEventTime):
* WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
(WTR::EventSendingController::EventSendingController):
(WTR::EventSendingController::mouseDown):
(WTR::EventSendingController::mouseUp):
(WTR::EventSendingController::mouseMoveTo):
(WTR::EventSendingController::leapForward):
(WTR::EventSendingController::keyDown):
* WebKitTestRunner/InjectedBundle/EventSendingController.h:
* WebKitTestRunner/TestController.cpp:
(WTR::TestController::didReceiveSynchronousMessageFromInjectedBundle):
* WebKitTestRunner/mac/EventSenderProxy.mm:
(WTR::eventTypeForMouseButtonAndAction):
(WTR::operator==):
(WTR::EventSenderProxy::updateClickCountForButton):
(WTR::EventSenderProxy::mouseDown):
(WTR::EventSenderProxy::mouseUp):
(WTR::EventSenderProxy::mouseMoveTo):
(WTR::EventSenderProxy::keyDown):
* WebKitTestRunner/qt/EventSenderProxyQt.cpp:
(WTR::EventSenderProxy::keyDown):
(WTR::EventSenderProxy::mouseDown):
(WTR::EventSenderProxy::mouseUp):
(WTR::EventSenderProxy::mouseMoveTo):

LayoutTests:

Unskip the passed tests.

Patch by Chang Shu <cshu@webkit.org> on 2011-09-21
Reviewed by Darin Adler.

* platform/mac-wk2/Skipped:
* platform/wk2/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95660 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[EFL] Revert pointer operator coding style in ewk
https://bugs.webkit.org/show_bug.cgi?id=68231

Reviewed by Antonio Gomes.

The WebKit coding style mandates that "(...) in C++ code - Both pointer
types and reference types should be written with no space between the
type name and the * or &.". Until now EFL port was using the
coding-style of the library for platform  code instead of WebKit's.

We are now trying to be more compliant with WebKit coding-style so it's
easier for other reviewers to review EFL code. The first step in this
direction is to move the '*' and '&' of pointers and references to the
right place.

This patch was generated with 'uncrustify' with the configure file
attached to the respective but report. There was one fix that had to be
manually made in order to keep the 'check-webkit-style' happy: a
function definition with a meaningless name. Other functions like this
exist but they will be fixed by a later patch.

Some nice side-effects of using uncrustify:

- Whitespace cleanup on comments
- Function parameters are aligned to the open parenthesis
- Remove tab characters

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95659 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[chromium] Optimize ImageBuffer constructor when accelerated
https://bugs.webkit.org/show_bug.cgi?id=68501

If the accelerated rendering mode is passed in to ImageBuffer, instead
of creating a bitmap PlatformCanvas and subsequently replacing its
device with an SkGpuDevice, create an SkCanvas and SkGpuDevice first,
and fall back to raster creation only failure.  This saves on system
RAM for the backing store (and possibly GDI resources).

Reviewed by James Robinson.

Covered by existing tests.

* platform/graphics/gpu/SharedGraphicsContext3D.cpp:
(WebCore::SharedGraphicsContext3D::get):
* platform/graphics/gpu/SharedGraphicsContext3D.h:
Rename "SharedGraphicsContext3D::create()" to "get()" to more
accurately reflect its function.
* platform/graphics/skia/ImageBufferSkia.cpp:
(WebCore::createAcceleratedCanvas):
Refactor creation of accelerated ImageBuffer.
(WebCore::ImageBuffer::ImageBuffer):
Attempt creation of accelerated ImageBuffer first, and only fall back
to bitmap-backed canvas creation if it fails.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95658 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[EFL] Don't try to free user-provided list
https://bugs.webkit.org/show_bug.cgi?id=68356

Reviewed by Antonio Gomes.

A coding-style cleanup revealed that we were using EINA_LIST_FREE in a
wrong manner. If we intended to free the data, we should call free()
(or whatever function is required to destroy it). However not only did
it use the wrong EFL api but it also introduced a change in behavior,
freeing user-provided data.

* ewk/ewk_frame.cpp:
(ewk_frame_feed_touch_event): do not free user-provided data.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95656 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Source/JavaScriptCore: JavaScriptCore Part of: Prevent the WebKit frameworks from defining inappropriately-named Objective-C classes
https://bugs.webkit.org/show_bug.cgi?id=68451

Reviewed by Darin Adler.

* JavaScriptCore.xcodeproj/project.pbxproj: Added a script build phase that invokes
check-for-inappropriate-objc-class-names, allowing only class names prefixed with "JS".

Source/WebCore: WebCore part of: Prevent the WebKit frameworks from defining inappropriately-named Objective-C classes
https://bugs.webkit.org/show_bug.cgi?id=68451

Reviewed by Darin Adler.

* WebCore.xcodeproj/project.pbxproj: Added a script build phase that invokes
check-for-inappropriate-objc-class-names, allowing only class names prefixed with "DOM" or "Web".

Source/WebKit: WebKit part of: Prevent the WebKit frameworks from defining inappropriately-named Objective-C classes
https://bugs.webkit.org/show_bug.cgi?id=68451

Reviewed by Darin Adler.

* WebKit.xcodeproj/project.pbxproj: Added a script build phase that invokes
check-for-inappropriate-objc-class-names, allowing only class names prefixed with "Web" or "_Web".

Source/WebKit2: WebKit2 part of: Prevent the WebKit frameworks from defining inappropriately-named Objective-C classes
https://bugs.webkit.org/show_bug.cgi?id=68451

Reviewed by Darin Adler.

* WebKit2.xcodeproj/project.pbxproj: Added a script build phase that invokes
check-for-inappropriate-objc-class-names, allowing only class names prefixed with "WK" or "Web".

Tools: Tools part of: Prevent the WebKit frameworks from defining inappropriately-named Objective-C classes
https://bugs.webkit.org/show_bug.cgi?id=68451

Reviewed by Darin Adler.

* Scripts/check-for-inappropriate-objc-class-names: Added. Checks for
Objective-C classes with names not having one of a list of prefixes
passed on the command line.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95655 268f45cc-cd09-0410-ab3c-d52691b4dbfc

<rdar://problem/9768483> REGRESSION: Crash in RenderBlock::removeFloatingObjectsBelow()
https://bugs.webkit.org/show_bug.cgi?id=68550

Reviewed by Darin Adler.

Source/WebCore:

Test: fast/multicol/null-lastFloat-in-removeFloatingObjectsBelow.html

* rendering/RenderBlock.cpp:
(WebCore::RenderBlock::removeFloatingObjectsBelow): Break out of the while() loop when the
floating object set is empty, which can happen if there were no floats on previous lines
(in which case, lastFloat is 0).

LayoutTests:

* fast/multicol/null-lastFloat-in-removeFloatingObjectsBelow-expected.txt: Added.
* fast/multicol/null-lastFloat-in-removeFloatingObjectsBelow.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95654 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fixed bug number in test expectations.

* platform/chromium/test_expectations.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95653 268f45cc-cd09-0410-ab3c-d52691b4dbfc

MacroAssembler fixes.
https://bugs.webkit.org/show_bug.cgi?id=68494

Reviewed by Sam Weinig.

Add X86-64's 3 operand or32 to other MacroAssembler, fix load32's [const] void* mismatch

* assembler/MacroAssembler.h:
(JSC::MacroAssembler::orPtr):
(JSC::MacroAssembler::loadPtr):
* assembler/MacroAssemblerARM.h:
(JSC::MacroAssemblerARM::or32):
* assembler/MacroAssemblerARMv7.h:
(JSC::MacroAssemblerARMv7::or32):
* assembler/MacroAssemblerMIPS.h:
(JSC::MacroAssemblerMIPS::or32):
* assembler/MacroAssemblerSH4.h:
(JSC::MacroAssemblerSH4::or32):
(JSC::MacroAssemblerSH4::load32):
* assembler/MacroAssemblerX86.h:
(JSC::MacroAssemblerX86::load32):
* assembler/MacroAssemblerX86_64.h:
(JSC::MacroAssemblerX86_64::load32):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95652 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[GTK] Fix build break when enabling webgl on r95593
https://bugs.webkit.org/show_bug.cgi?id=68502

Patch by ChangSeok Oh <shivamidow@gmail.com> on 2011-09-21
Reviewed by Martin Robinson.

Add PassOwnPtr.h to GC3DPrivate.h and remove useless one from GC3DGtk.cpp & GC3DPrivate.cpp

No tests required.

* platform/graphics/gtk/GraphicsContext3DGtk.cpp:
* platform/graphics/gtk/GraphicsContext3DPrivate.cpp:
* platform/graphics/gtk/GraphicsContext3DPrivate.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95651 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Branch builds don't launch on Windows when run against trunk WebKit
https://bugs.webkit.org/show_bug.cgi?id=68546

Reviewed by Darin Adler.

* Interfaces/IWebPreferences.idl: Swap avFoundation and pictographicFontFamily
in IDL for binary compatibility between branch and trunk.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95650 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: clear resource agent resource cache upon resource agent's disable.
https://bugs.webkit.org/show_bug.cgi?id=67995

Reviewed by Pavel Feldman.

Source/WebCore:

Test: http/tests/inspector/network/network-disabling-check-no-memory-leak.html

* inspector/InspectorResourceAgent.cpp:
(WebCore::InspectorResourceAgent::disable):

LayoutTests:

* http/tests/inspector/network/network-disabling-check-no-memory-leak-expected.txt: Added.
* http/tests/inspector/network/network-disabling-check-no-memory-leak.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95649 268f45cc-cd09-0410-ab3c-d52691b4dbfc

https://bugs.webkit.org/show_bug.cgi?id=68543

Make sure that when regions change dimensions we recompute pagination slices.

Reviewed by Simon Fraser.

Source/WebCore:

Added new tests in fast/regions.

* rendering/LayoutState.cpp:
(WebCore::LayoutState::LayoutState):
* rendering/LayoutState.h:
* rendering/RenderFlowThread.cpp:
(WebCore::RenderFlowThread::layout):
(WebCore::RenderFlowThread::renderRegionForLine):
(WebCore::RenderFlowThread::regionLogicalWidthForLine):
* rendering/RenderRegion.cpp:
(WebCore::RenderRegion::layout):
* rendering/RenderView.cpp:
(WebCore::RenderView::pushLayoutState):
* rendering/RenderView.h:
(WebCore::LayoutStateMaintainer::LayoutStateMaintainer):
(WebCore::LayoutStateMaintainer::push):

LayoutTests:

* fast/regions/text-region-split-after-resize-expected.txt: Added.
* fast/regions/text-region-split-after-resize.html: Added.
* fast/regions/webkit-flow-floats-inside-regions-bounds-expected.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95648 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Ensure TestWebKitAPI works on mac, win, chromium-mac and chromium-linux
https://bugs.webkit.org/show_bug.cgi?id=68206

Patch by Xianzhu Wang <wangxianzhu@chromium.org> on 2011-09-21
Reviewed by Tony Chang.

Source/WebCore:

Added two new wtf headers into ForwardingHeaders/wtf to make
#include <wtf/...h> cross-platform.
No new tests needed.

* ForwardingHeaders/wtf/MetaAllocator.h: Added.
* ForwardingHeaders/wtf/RedBlackTree.h: Added.

Source/WebKit/chromium:

* WebKit.gyp:

Tools:

* Scripts/webkitdirs.pm:
(determineBaseProductDir): add support for chromium-mac
* TestWebKitAPI/Tests/WTF/MetaAllocator.cpp:
* TestWebKitAPI/Tests/WTF/RedBlackTree.cpp:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95647 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[WinCairo] IconDatabase::defaultIcon always fails for non-CAN_THEME_URL_ICON builds.
https://bugs.webkit.org/show_bug.cgi?id=67469

Patch by David Delaune <david.delaune@googlemail.com> on 2011-09-21
Reviewed by Brent Fulgham.

No new functionality added in this change.

* loader/icon/IconDatabase.cpp: defining CAN_THEME_URL_ICON for the WIN_CAIRO

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95646 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Span element gets produced using backspace/delete to merge header with paragraph
https://bugs.webkit.org/show_bug.cgi?id=68413

Reviewed by Darin Adler.

Source/WebCore:

Add p element to the list of elements to retain appearance. Also modified removeStyleFromNode
(and renamed to removeStyleFromRules) not to remove inline style declarations because
inline styles need to stay on copy.

Test: editing/deleting/merge-paragraph-from-p-with-style.html

* editing/DeleteSelectionCommand.cpp:
(WebCore::DeleteSelectionCommand::saveTypingStyleState):
* editing/EditingStyle.cpp:
(WebCore::EditingStyle::removeStyleFromRules): Renamed from removeStyleFromNode.
* editing/EditingStyle.h:
* editing/markup.cpp:
(WebCore::isBlockNodeToRetainAppearance):
(WebCore::StyledMarkupAccumulator::serializeNodes):

LayoutTests:

Add a test to remove the line break before a p element. WebKit should not preserve
the font color from the style rule but should preserve the font style from inline style declaration.

Also fix merge-paragraph-from-h6* to actually h6 instead of h1.

* editing/deleting/merge-paragraph-from-h6-expected.txt:
* editing/deleting/merge-paragraph-from-h6-with-style-expected.txt:
* editing/deleting/merge-paragraph-from-h6-with-style.html:
* editing/deleting/merge-paragraph-from-h6.html:
* editing/deleting/merge-paragraph-from-p-with-style-expected.txt: Added.
* editing/deleting/merge-paragraph-from-p-with-style.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95645 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[flexbox] rename classnames and methods in the tests to not assume horizontal
https://bugs.webkit.org/show_bug.cgi?id=68498

Reviewed by Tony Chang.

* css3/flexbox/001.html:
* css3/flexbox/002.html:
* css3/flexbox/003.html:
* css3/flexbox/004.html:
* css3/flexbox/resources/flexbox.js:
(checkFlexBoxen):
* css3/flexbox/writing-modes.html:
* platform/mac/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95644 268f45cc-cd09-0410-ab3c-d52691b4dbfc

need rebaseline for gradient speedup in skia
https://bugs.webkit.org/show_bug.cgi?id=68531

Reviewed by Stephen White.

* platform/chromium/test_expectations.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95643 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, move today's skipped flaky tests to test_expectations
as suggested by mrobinson.

* platform/gtk/Skipped:
* platform/gtk/test_expectations.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95642 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Report AXValueChanged when value changes in element with role=textbox.
https://bugs.webkit.org/show_bug.cgi?id=68201

Patch by Alice Boxhall <aboxhall@chromium.org> on 2011-09-21
Reviewed by Chris Fleizach.

Source/WebCore:

Test: platform/mac/accessibility/textbox-role-reports-notifications.html

* accessibility/AccessibilityObject.cpp:
(WebCore::AccessibilityObject::isARIATextControl):
* accessibility/AccessibilityObject.h:
* accessibility/AccessibilityRenderObject.cpp:
(WebCore::AccessibilityRenderObject::contentChanged):
(WebCore::AccessibilityRenderObject::childrenChanged):

LayoutTests:

* platform/mac/accessibility/textbox-role-reports-notifications-expected.txt: Added.
* platform/mac/accessibility/textbox-role-reports-notifications.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95641 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[chromium]ClearFocusedNode doesn't clear the cursor or selection in editable div element.
https://bugs.webkit.org/show_bug.cgi?id=68351

Patch by chandra shekar vallala <chandra.vallala@gmail.com> on 2011-09-21
Reviewed by Darin Fisher.

* src/WebViewImpl.cpp:
(WebKit::WebViewImpl::clearFocusedNode):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95640 268f45cc-cd09-0410-ab3c-d52691b4dbfc

New layout tests for the JS bindings related to <track>:
    e.g. TextTrack, TextTrackCue, TextTrackCueList, MutableTextTrack.
https://bugs.webkit.org/show_bug.cgi?id=68194

Reviewed by Eric Carlson.

* media/track/track-mutable-expected.txt: Added.
* media/track/track-mutable.html: Added.
* media/track/track-text-track-cue-list-expected.txt: Added.
* media/track/track-text-track-cue-list.html: Added.
* media/track/track-text-track-expected.txt: Added.
* media/track/track-text-track.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95639 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Actually skip the test (forgot in the previous commit). Sorry for the noise.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95638 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt][WK2] plugins/mouse-click-iframe-to-plugin.html fails since enabled in r95626
https://bugs.webkit.org/show_bug.cgi?id=68539

Unreviewed gardening.

* platform/qt-wk2/Skipped: skip the test.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95637 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: introduce Page.enable and Page.disable
https://bugs.webkit.org/show_bug.cgi?id=68530

Otherwise protocol is violating the contract of not sending
notifications by default.

Reviewed by Yury Semikhatsky.

* inspector/Inspector.json:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95636 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed; new baselines (Skia on Mac)

This separates test failures that are specific to CG Mac from failures
for Skia on Mac. Most of these changes are specific to Leopard, but a
handful add new fail flavors for all Mac variations.

* LayoutTests/platform/chromium/test_expectations.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95635 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: add minimal error handling to the DOM breakpoints.
https://bugs.webkit.org/show_bug.cgi?id=68537

Reviewed by Yury Semikhatsky.

* inspector/InspectorDOMAgent.h:
* inspector/InspectorDOMDebuggerAgent.cpp:
(WebCore::domTypeForName):
(WebCore::InspectorDOMDebuggerAgent::setDOMBreakpoint):
(WebCore::InspectorDOMDebuggerAgent::removeDOMBreakpoint):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95634 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, GTK baselines for
cross-origin-xsl-redirect-BLOCKED.html which were removed by mistake.

* platform/gtk/http/tests/security/cross-origin-xsl-redirect-BLOCKED-expected.txt: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95633 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: make WebInspector a bit more timeline friendly.
https://bugs.webkit.org/show_bug.cgi?id=68534

I found that the maximum captured stack frame depth tracked by timeline is not
very convenient because the average depth of call stack considerably more than five.
Another useful trick is console.timeStamp for all backend calls.

Reviewed by Yury Semikhatsky.

* inspector/CodeGeneratorInspector.pm:
* inspector/front-end/TimelinePanel.js:
(WebInspector.TimelinePanel.prototype._toggleTimelineButtonClicked):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95632 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt][WK2] Implement Download support in WebProcess
https://bugs.webkit.org/show_bug.cgi?id=68153

Patch by Jesus Sanchez-Palencia <jesus.palencia@openbossa.org> on 2011-09-21
Reviewed by Andreas Kling.

Source/WebCore:

Refactored QNetworkReplyHandler::finish() in order to add
and use the static function QNetworkReplyHandler::errorForReply().
This will be used by our Download implementation in WebKit2 (WebProcess)
when handling ResourceError.

* platform/network/qt/QNetworkReplyHandler.cpp:
(WebCore::QNetworkReplyHandler::finish):
(WebCore::QNetworkReplyHandler::errorForReply):
* platform/network/qt/QNetworkReplyHandler.h:
* platform/network/qt/ResourceRequest.h:

Source/WebKit2:

We implement the necessary functions of Download.h, and our QtFileDownloader
to handle all network communication and call the necessary functions of Download.

We use the download policy for any MIME type not supported by WebKit. This
behaves like Qt non-WebKit2 except that we don't force download when
we encounter Content-Disposition: attachment. We still use the "filename="
field for file name suggestion though.

Based on original patches by: Kimmo Kinnunen <kimmo.t.kinnunen@nokia.com>,
Jocelyn Turcotte <jocelyn.turcotte@nokia.com>, Simon Hausmann <simon.hausmann@nokia.com>
and Zalan Bujtas <zalan.bujtas@nokia.com>.

* UIProcess/API/qt/qweberror.cpp: Adding DownloadError
(QWebError::type):
* UIProcess/API/qt/qweberror.h: Adding DownloadError
* WebKit2.pro:
* WebProcess/Downloads/Download.cpp: Adding QtFileDownloader, Qt platform specific
(WebKit::Download::Download):
* WebProcess/Downloads/Download.h: Adding QtFileDownloader, Qt platform specific
* WebProcess/Downloads/qt/DownloadQt.cpp:
(WebKit::Download::start):
(WebKit::Download::startWithHandle):
(WebKit::Download::cancel):
(WebKit::Download::platformInvalidate):
(WebKit::Download::didDecideDestination):
* WebProcess/Downloads/qt/QtFileDownloader.cpp: Added.
(WebKit::QtFileDownloader::QtFileDownloader):
(WebKit::QtFileDownloader::~QtFileDownloader):
(WebKit::QtFileDownloader::determineFilename):
(WebKit::QtFileDownloader::decidedDestination):
(WebKit::QtFileDownloader::abortDownloadWritingAndEmitError):
(WebKit::QtFileDownloader::onReadyRead):
(WebKit::QtFileDownloader::onFinished):
(WebKit::QtFileDownloader::onError):
(WebKit::QtFileDownloader::cancel):
* WebProcess/Downloads/qt/QtFileDownloader.h: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95631 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, skip flaky media test on GTK.

* platform/gtk/Skipped: Skip media/media-blocked-by-beforeload.html

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95630 268f45cc-cd09-0410-ab3c-d52691b4dbfc

garden-o-matic command should explain that it is starting a local server
https://bugs.webkit.org/show_bug.cgi?id=68486

Reviewed by Dimitri Glazkov.

In the user study, both participants didn't understand why the
garden-o-matic command appeared to hang.  This patch adds an
explaination of what's going on.  It also adds a confirmation prompt
before launching the browser to give the user a chance to read the
text.  If that's too much of a speed bump, we can iterate.

* Scripts/webkitpy/tool/commands/gardenomatic.py:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95629 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Syntax error in media/audio-delete-while-step-button-clicked.html
https://bugs.webkit.org/show_bug.cgi?id=68507

Reviewed by Eric Carlson.

* media/audio-delete-while-step-button-clicked.html: Some curly
braces were missing in a try/catch statement.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95628 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[EFL] Fix alphabetical sorting problem
https://bugs.webkit.org/show_bug.cgi?id=68505

Reviewed by Philippe Normand.

* ewk/ewk_util.cpp: Fix wrong alphabetical sorting problem.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95627 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt] Implement KeyDown function for WebKit2 EventSender.
https://bugs.webkit.org/show_bug.cgi?id=56485

Tools:

Implement the KeyDown function on Qt platform.

Patch by Chang Shu <cshu@webkit.org> on 2011-09-21
Reviewed by Andreas Kling.

* WebKitTestRunner/PlatformWebView.h:
* WebKitTestRunner/TestController.cpp:
(WTR::TestController::TestController):
(WTR::TestController::didReceiveSynchronousMessageFromInjectedBundle):
* WebKitTestRunner/qt/EventSenderProxyQt.cpp: Added.
(WTR::getModifiers):
(WTR::EventSenderProxy::keyDown):
(buildModifierFlags):
(EventSenderProxy::keyDown):
* WebKitTestRunner/qt/PlatformWebViewQt.cpp:
(WTR::PlatformWebView::sendEvent):
* WebKitTestRunner/qt/WebKitTestRunner.pro:

LayoutTests:

Unskip passed tests.

Patch by Chang Shu <cshu@webkit.org> on 2011-09-21
Reviewed by Andreas Kling.

* platform/qt-wk2/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95626 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: UI performance test for Resource panel first open experience.
https://bugs.webkit.org/show_bug.cgi?id=68527

Reviewed by Pavel Feldman.

* inspector/performance/resources/first-open-resources.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95625 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: extend UI perf tests coverage. Elements panel opens first time.
https://bugs.webkit.org/show_bug.cgi?id=68432

Reviewed by Pavel Feldman.

* inspector/performance/resources/first-open-elements.html: Added.
* inspector/performance/resources/performance-test.js:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95624 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, skip some more flaky SVG tests on GTK.

* platform/gtk/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95623 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, fix baselines for cross-origin-xsl-redirect-BLOCKED

* http/tests/security/cross-origin-xsl-redirect-BLOCKED-expected.txt:
* platform/chromium-cg-mac/http/tests/security/cross-origin-xsl-redirect-BLOCKED-expected.txt:
* platform/chromium-mac/http/tests/security/cross-origin-xsl-redirect-BLOCKED-expected.txt:
* platform/chromium-win/http/tests/security/cross-origin-xsl-redirect-BLOCKED-expected.txt:
* platform/gtk/http/tests/security/cross-origin-xsl-redirect-BLOCKED-expected.txt: Removed
* platform/mac-wk2/http/tests/security/cross-origin-xsl-redirect-BLOCKED-expected.txt:
* platform/qt-wk2/Skipped:
* platform/qt-wk2/http/tests/security/cross-origin-xsl-redirect-BLOCKED-expected.txt:
* platform/qt/http/tests/security/cross-origin-xsl-redirect-BLOCKED-expected.png: Removed
* platform/qt/http/tests/security/cross-origin-xsl-redirect-BLOCKED-expected.txt: Removed
* platform/win-wk2/http/tests/security/cross-origin-xsl-redirect-BLOCKED-expected.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95622 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: paint box model colors in Metrics sidebar at all times, do not draw box outlines.
https://bugs.webkit.org/show_bug.cgi?id=68240

Today we paint backgrounds in Metrics box model on hover only - should be painted at all
times for the reference. Outlining boxes is highlight is wrong since outlines are outside
the corresponding box regions. We've seen few reports on that + Firebug does not do borders
for that reason.

Reviewed by Yury Semikhatsky.

* inspector/DOMNodeHighlighter.cpp:
* inspector/DOMNodeHighlighter.h:
* inspector/Inspector.json:
* inspector/InspectorDOMAgent.cpp:
(WebCore::InspectorDOMAgent::setHighlightDataFromConfig):
* inspector/front-end/Color.js:
* inspector/front-end/MetricsSidebarPane.js:
(WebInspector.MetricsSidebarPane.prototype._highlightDOMNode):
(WebInspector.MetricsSidebarPane.prototype._updateMetrics):
* inspector/front-end/inspector.css:
(.metrics .label):
* inspector/front-end/inspector.js:
(WebInspector.buildHighlightConfig):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95621 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[chromium] rebaseline caused by r95567

* platform/chromium-cg-mac-leopard/fast/clip/overflow-border-radius-fixed-position-expected.png: Added.
* platform/chromium-mac-leopard: Added.
* platform/chromium-mac-leopard/fast: Added.
* platform/chromium-mac-leopard/fast/clip: Added.
* platform/chromium-mac-leopard/fast/clip/overflow-border-radius-fixed-position-expected.png: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95620 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Protect against misuse of EventListenerIterator.
https://bugs.webkit.org/show_bug.cgi?id=68364

Patch by Andreas Kling <kling@webkit.org> on 2011-09-21
Reviewed by Darin Adler.

In debug mode, keep track of the number of active EventListenerIterators
on an EventListenerMap, and assert that there are no iterators when the
map is being modified.

* dom/EventListenerMap.cpp:
(WebCore::EventListenerMap::EventListenerMap):
(WebCore::EventListenerMap::clear):
(WebCore::EventListenerMap::add):
(WebCore::EventListenerMap::remove):
(WebCore::EventListenerMap::find):
(WebCore::EventListenerMap::removeFirstEventListenerCreatedFromMarkup):
(WebCore::EventListenerMap::copyEventListenersNotCreatedFromMarkupToTarget):
(WebCore::EventListenerIterator::EventListenerIterator):
(WebCore::EventListenerIterator::~EventListenerIterator):
* dom/EventListenerMap.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95619 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[chromium] rebaseline caused by r95599

* fast/dom/HTMLMeterElement/meter-element-repaint-on-update-value-expected.txt: Added.
* platform/chromium-mac-leopard: Removed.
* platform/chromium-mac/fast/dom/HTMLMeterElement/meter-element-repaint-on-update-value-expected.png: Added.
* platform/chromium-win/fast/dom/HTMLMeterElement/meter-element-repaint-on-update-value-expected.png: Added.
* platform/chromium-win/fast/dom/HTMLMeterElement/meter-element-repaint-on-update-value-expected.txt: Added.
* platform/chromium/test_expectations.txt:
* platform/gtk/fast/dom/HTMLMeterElement/meter-element-repaint-on-update-value-expected.txt: Removed.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95618 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, skip 2 GTK flaky tests.

* platform/gtk/Skipped: Skip
http/tests/websocket/tests/hixie76/error-detect.html and
svg/zoom/page/zoom-svg-through-object-with-auto-size.html

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95617 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt][WK2] Unskip some cookies related layout tests.
https://bugs.webkit.org/show_bug.cgi?id=68484

Reviewed by Csaba Osztrogonác.

Unskip some cookies related layout tests as they seems to work fine with
--iterations 1000.

* platform/qt-wk2/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95616 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt][Mac] Unreviewed gardening after r95601.

* platform/qt-mac/Skipped: Paint the Qt-Mac bot green.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95615 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: minor fixes in the DOM domain of the protocol.
https://bugs.webkit.org/show_bug.cgi?id=68517

Reviewed by Yury Semikhatsky.

* inspector/Inspector.json:
* inspector/InspectorDOMAgent.cpp:
(WebCore::InspectorDOMAgent::setAttributesAsText):
* inspector/InspectorDOMAgent.h:
* inspector/front-end/DOMAgent.js:
(WebInspector.DOMNode.prototype.setAttribute):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95614 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: limit the use of "Resource" term in the protocol.
https://bugs.webkit.org/show_bug.cgi?id=68250

Reviewed by Yury Semikhatsky.

* inspector/Inspector.json:
* inspector/InspectorResourceAgent.cpp:
(WebCore::InspectorResourceAgent::markResourceAsCached):
(WebCore::InspectorResourceAgent::didLoadResourceFromMemoryCache):
(WebCore::InspectorResourceAgent::setExtraHTTPHeaders):
(WebCore::InspectorResourceAgent::getResponseBody):
* inspector/InspectorResourceAgent.h:
* inspector/front-end/ExtensionServer.js:
(WebInspector.ExtensionServer.prototype._onAddRequestHeaders):
* inspector/front-end/NetworkManager.js:
(WebInspector.NetworkManager.prototype.requestContent):
(WebInspector.NetworkDispatcher.prototype.requestServedFromCache):
(WebInspector.NetworkDispatcher.prototype.requestServedFromMemoryCache):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95613 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, rolling out r95608.
http://trac.webkit.org/changeset/95608
https://bugs.webkit.org/show_bug.cgi?id=68521

breaks GTK and Windows builds. Please check EWS next time.
(Requested by philn-tp on #webkit).

Patch by Sheriff Bot <webkit.review.bot@gmail.com> on 2011-09-21

Source/WebCore:

* html/HTMLBodyElement.cpp:
(WebCore::adjustForZoom):
(WebCore::HTMLBodyElement::setScrollLeft):
(WebCore::HTMLBodyElement::setScrollTop):
* loader/HistoryController.cpp:
(WebCore::HistoryController::saveScrollPositionAndViewStateToItem):
* page/DOMWindow.cpp:
(WebCore::DOMWindow::scrollX):
(WebCore::DOMWindow::scrollY):
(WebCore::DOMWindow::scrollTo):
* page/Frame.cpp:
(WebCore::Frame::pageScaleFactor):
* page/Frame.h:
* page/FrameView.cpp:
(WebCore::FrameView::applyOverflowToViewport):
(WebCore::FrameView::scrollXForFixedPosition):
(WebCore::FrameView::scrollYForFixedPosition):

LayoutTests:

* fast/events/scale-and-scroll-body-expected.png: Removed.
* fast/events/scale-and-scroll-body-expected.txt: Removed.
* fast/events/scale-and-scroll-body.html: Removed.
* fast/events/scale-and-scroll-iframe-body-expected.png: Removed.
* fast/events/scale-and-scroll-iframe-body-expected.txt: Removed.
* fast/events/scale-and-scroll-iframe-body.html: Removed.
* fast/events/scale-and-scroll-iframe-window-expected.png: Removed.
* fast/events/scale-and-scroll-iframe-window-expected.txt: Removed.
* fast/events/scale-and-scroll-iframe-window.html: Removed.
* fast/events/scale-and-scroll-window-expected.png: Removed.
* fast/events/scale-and-scroll-window-expected.txt: Removed.
* fast/events/scale-and-scroll-window.html: Removed.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95612 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: console panel's scroller is not properly aligned.
https://bugs.webkit.org/show_bug.cgi?id=68448

Reviewed by Yury Semikhatsky.

* inspector/front-end/inspector.css:
(#console-view):
(#console-messages):
(#drawer #console-messages):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95611 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, GTK baseslines for 2 new tests and skip a flaky svg test.

* platform/gtk/Skipped: Skip svg/W3C-SVG-1.1-SE/filters-image-05-f.svg
* platform/gtk/fast/clip/overflow-border-radius-fixed-position-expected.png: Added.
* platform/gtk/fast/clip/overflow-border-radius-fixed-position-expected.txt: Added.
* platform/gtk/fast/dom/HTMLMeterElement/meter-element-repaint-on-update-value-expected.png: Added.
* platform/gtk/fast/dom/HTMLMeterElement/meter-element-repaint-on-update-value-expected.txt: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95610 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt][WK2] http/tests/security/cross-origin-xsl-redirect-BLOCKED.html fails
https://bugs.webkit.org/show_bug.cgi?id=68515

* platform/qt-wk2/Skipped: Skip http/tests/security/cross-origin-xsl-redirect-BLOCKED.html.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95609 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Source/WebCore: https://bugs.webkit.org/show_bug.cgi?id=68081
Fix scroll in page scaling mode.

Reviewed by Simon Fraser.

The following things are done in this patch:
1. Rename Frame::pageScaleFactor() to Frame::frameScaleFactor() to better
   reflect its purpose.
2. Frame::frameScaleFactor() returns the scale factor of this frame with
   respect to the container. So for the main frame it'll return the scale
   factor of the page, inner frames will return 1.0
3. scrollBy(), scrollX() and scrollY() to take into account of the frame
   scale factor, causing incorrect coordinates reported to Javascript.

Tests: fast/events/scale-and-scroll-body.html
       fast/events/scale-and-scroll-iframe-body.html
       fast/events/scale-and-scroll-iframe-window.html
       fast/events/scale-and-scroll-window.html

* html/HTMLBodyElement.cpp:
(WebCore::adjustForZoom):
(WebCore::HTMLBodyElement::setScrollLeft):
(WebCore::HTMLBodyElement::setScrollTop):
* loader/HistoryController.cpp:
(WebCore::HistoryController::saveScrollPositionAndViewStateToItem):
* page/DOMWindow.cpp:
(WebCore::DOMWindow::scrollX):
(WebCore::DOMWindow::scrollY):
(WebCore::DOMWindow::scrollTo):
* page/Frame.cpp:
(WebCore::Frame::frameScaleFactor): Return page's scale factor if this is
                                    main frame, otherwise 1.0
* page/Frame.h:
* page/FrameView.cpp:
(WebCore::FrameView::applyOverflowToViewport):
(WebCore::FrameView::scrollXForFixedPosition):
(WebCore::FrameView::scrollYForFixedPosition):

LayoutTests: https://bugs.webkit.org/show_bug.cgi?id=68081
Add layout test for page scale and scrolling.

Reviewed by Simon Fraser.

Add tests to cover the following cases:
1. Scale and scroll window
2. Scale and scroll document.body
3. Scale and scroll iframe's contentWindow
4. Scale and scroll iframe's contentDocument.body

* fast/events/scale-and-scroll-body-expected.png: Added.
* fast/events/scale-and-scroll-body-expected.txt: Added.
* fast/events/scale-and-scroll-body.html: Added.
* fast/events/scale-and-scroll-iframe-body-expected.png: Added.
* fast/events/scale-and-scroll-iframe-body-expected.txt: Added.
* fast/events/scale-and-scroll-iframe-body.html: Added.
* fast/events/scale-and-scroll-iframe-window-expected.png: Added.
* fast/events/scale-and-scroll-iframe-window-expected.txt: Added.
* fast/events/scale-and-scroll-iframe-window.html: Added.
* fast/events/scale-and-scroll-window-expected.png: Added.
* fast/events/scale-and-scroll-window-expected.txt: Added.
* fast/events/scale-and-scroll-window.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95608 268f45cc-cd09-0410-ab3c-d52691b4dbfc

<rdar://problem/9890932> "Open PDF in Preview" on two PDFs of the same name does nothing the second time

The logic in PDFViewController::pathToPDFOnDisk that deals with constructing a unique
file name if a file of the suggested name already exists was incorrect. It was appending
the template used by mkstemps to the directory name component of the path rather than
using it as a prefix on the filename portion of the path. This resulted in a template path
like WebKitPDFs-ABCDEFXXXXXX-/foo.pdf where WebKitPDFs-ABCDEF/XXXXXX-foo.pdf was expected.

Reviewed by Jon Honeycutt.

* UIProcess/API/mac/PDFViewController.mm:
(WebKit::PDFViewController::pathToPDFOnDisk): Append the template as a new path component,
then append the suggested file name to that component.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95607 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, GTK rebaseline, skip some flaky worker tests and 2
failing http tests.

* platform/gtk/Skipped:
* platform/gtk/editing/deleting/regional-indicators-expected.txt:
* platform/gtk/fast/text/justify-ideograph-complex-expected.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95606 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, GTK rebaseline and skip and a11y failing test.

* platform/gtk/Skipped: Skip accessibility/aria-hidden-updates-alldescendants.html
* platform/gtk/editing/deleting/merge-whitespace-pre-expected.txt:
Rebaseline after r95335.
* platform/gtk/http/tests/security/cross-origin-xsl-redirect-BLOCKED-expected.txt: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95605 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt] Remove Qt specific code from css/SelectorChecker.cpp
https://bugs.webkit.org/show_bug.cgi?id=67702

Reviewed by Csaba Osztrogonác.

Source/WebCore:

No new tests needed.

* css/SelectorChecker.cpp: Move Qt-specific code to the Qt-specific WebPlatformStrategies.
(WebCore::SelectorChecker::determineLinkStateSlowCase):
* platform/VisitedLinkStrategy.h:

Source/WebKit/mac:

* WebCoreSupport/WebPlatformStrategies.h:
* WebCoreSupport/WebPlatformStrategies.mm:
(WebPlatformStrategies::isLinkVisited):

Source/WebKit/qt:

* WebCoreSupport/WebPlatformStrategies.cpp: Check the Qt global history here, instead of a platform independent file.
(WebPlatformStrategies::isLinkVisited):
* WebCoreSupport/WebPlatformStrategies.h:

Source/WebKit/win:

* WebCoreSupport/WebPlatformStrategies.cpp: Modify to match new prototype.
(WebPlatformStrategies::isLinkVisited):
* WebCoreSupport/WebPlatformStrategies.h: Ditto.

Source/WebKit/wince:

* WebCoreSupport/PlatformStrategiesWinCE.cpp:
(PlatformStrategiesWinCE::addVisitedLink):
* WebCoreSupport/PlatformStrategiesWinCE.h:

Source/WebKit2:

* WebProcess/WebCoreSupport/WebPlatformStrategies.cpp: Modify to match new prototype.
(WebKit::WebPlatformStrategies::isLinkVisited):
(WebKit::WebPlatformStrategies::addVisitedLink):
* WebProcess/WebCoreSupport/WebPlatformStrategies.h: Ditto.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95604 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, GTK rebaseline after r95566.

* platform/gtk/mathml/xHeight-expected.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95603 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed Skipped list update for meter-element-repaint-on-update-value.html,
which needs windows

* platform/win/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95602 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt] Unreviewed gardening.

* platform/qt-mac/Skipped:
* platform/qt-wk2/Skipped:
* platform/qt/fast/clip/overflow-border-radius-fixed-position-expected.png: Added.
* platform/qt/fast/clip/overflow-border-radius-fixed-position-expected.txt: Added.
* platform/qt/fast/dom/HTMLMeterElement/meter-element-repaint-on-update-value-expected.png: Added.
* platform/qt/fast/dom/HTMLMeterElement/meter-element-repaint-on-update-value-expected.txt: Added.
* platform/qt/http/tests/security/cross-origin-xsl-redirect-BLOCKED-expected.png: Added.
* platform/qt/http/tests/security/cross-origin-xsl-redirect-BLOCKED-expected.txt: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95601 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Style changes on forwarded shadow children should trigger shadow renderer reconstruction.
https://bugs.webkit.org/show_bug.cgi?id=68228

Reviewed by Dimitri Glazkov.

Currently style changes on forwarded shadow child triggers
Node::reattach() only for the forwarded node. But it should
invalidate whole renderers on the shadow tree because
ShadowContentElement expects its inclusions to be attached and to
create renderers in the inclusion order. But single node
reattach() violates the expectation and possibly creates
inconsistent render tree.

This change makes NodeRenderingContext to go
AttachContentForwarded, where the ShadowContentElement::attach()
happens, only during ShadowRoot::attach(), not for style-change
triggered Node::attach() on the forwarded nodes.

Note that the altering phase AttachContentLight doesn't create
renderers for the forwarded nodes. Instead of that, it possibly
triggers shadow subtree reconstruction on
NodeRenderingContext::hostChildrenChanged().

No new tests. This is a defensive change. There is no visible
difference at this time.

* dom/NodeRenderingContext.cpp:
(WebCore::NodeRenderingContext::NodeRenderingContext):
* dom/ShadowRoot.cpp:
(WebCore::ShadowRoot::isInclusionSelectorActive):
* dom/ShadowRoot.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95600 268f45cc-cd09-0410-ab3c-d52691b4dbfc

<meter> doesn't update rendering when its value is changed.
https://bugs.webkit.org/show_bug.cgi?id=67614

Reviewed by Dimitri Glazkov.

Source/WebCore:

It lacks repaint() when the value is changed.

For non-native appearance it works because the indicator is
implemented in shadow DOM, for native appearance, repainting should be
handled explicitly.

Test: fast/dom/HTMLMeterElement/meter-element-repaint-on-update-value.html

* html/HTMLMeterElement.cpp:
(WebCore::HTMLMeterElement::didElementStateChange):
* rendering/RenderMeter.cpp:
(WebCore::RenderMeter::updateFromElement):
* rendering/RenderMeter.h:
* rendering/RenderProgress.cpp:
(WebCore::RenderProgress::updateAnimationState):

LayoutTests:

* fast/dom/HTMLMeterElement/meter-element-repaint-on-update-value-expected.png: Added.
* platform/chromium/test_expectations.txt:
* platform/mac/dom/HTMLMeterElement/meter-element-repaint-on-update-value-expected.txt: Added.
* platform/mac/dom/HTMLMeterElement/meter-element-repaint-on-update-value.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95599 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Move RenderObject::markContainingBlocksForLayout() to RenderObject.cpp.
https://bugs.webkit.org/show_bug.cgi?id=67989

Reviewed by Ryosuke Niwa.

No new tests because of no behavior change.

* rendering/RenderObject.cpp:
(WebCore::objectIsRelayoutBoundary): Moved from RenderObject.h.
(WebCore::RenderObject::markContainingBlocksForLayout): Moved from RenderObject.h
* rendering/RenderObject.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95598 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[chromium] changed test expectations for zoom-svg-through-object-with-auto-size.html:wq

* platform/chromium/test_expectations.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95597 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Some Heap cleanup.

Reviewed by Beth Dakin.

* heap/MarkedBlock.cpp:
(JSC::MarkedBlock::blessNewBlock): Removed blessNewBlockForSlowPath()
because it was unused; renamed blessNewBlockForFastPath() to blessNewBlock()
since there is only one now.

* heap/MarkedBlock.h: Removed ownerSet-related stuff since it was unused.
Updated mark bit overhead calculation. Deployed atomsPerBlock in one
place where we were recalculating it.

* heap/MarkedSpace.cpp:
(JSC::MarkedSpace::addBlock): Updated for rename.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95596 268f45cc-cd09-0410-ab3c-d52691b4dbfc

garden-o-matic unit test fail because Safari 5.1 doesn't enumerate dictionaries in the same order as V8
https://bugs.webkit.org/show_bug.cgi?id=68496

Reviewed by Dimitri Glazkov.

This patch adds a bunch of sorting to make sure things happen in a
predictable order in both JavaScriptCore and V8.  We've talked before
about whether these two JavaScript engines should enumerate
dictionaries in the same order, but the consensus has been that they
shouldn't.  Personally, I disagree with that consensus, but such is life.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base.js:
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base_unittests.js:
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures_unittests.js:
* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95595 268f45cc-cd09-0410-ab3c-d52691b4dbfc

DFG JIT always speculates integer on modulo
https://bugs.webkit.org/show_bug.cgi?id=68485

Reviewed by Oliver Hunt.

Added support for double modulo, which is a call to fmod().
Also added support for recording the old JIT's statistics
on op_mod and propagating them along the graph. Finally,
fixed a goof in the ArithNodeFlags propagation logic that
was made obvious when I started testing ArithMod.

* dfg/DFGByteCodeParser.cpp:
(JSC::DFG::ByteCodeParser::makeSafe):
(JSC::DFG::ByteCodeParser::parseBlock):
* dfg/DFGNode.h:
(JSC::DFG::Node::hasArithNodeFlags):
* dfg/DFGPropagator.cpp:
(JSC::DFG::Propagator::propagateArithNodeFlags):
(JSC::DFG::Propagator::propagateNodePredictions):
(JSC::DFG::Propagator::fixupNode):
* dfg/DFGSpeculativeJIT.cpp:
(JSC::DFG::SpeculativeJIT::compile):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95594 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Support for multiple <link rel="icon"> favicon elements.
https://bugs.webkit.org/show_bug.cgi?id=65564

Patch by Rachel Blum <groby@chromium.org> on 2011-09-20
Reviewed by Darin Fisher.

No tests - purely an API change. (And API is not exposed to LayoutTests)

* dom/Document.cpp:
(WebCore::Document::iconURLs):
(WebCore::Document::addIconURL):
* dom/Document.h:
* dom/IconURL.cpp:
(WebCore::toIconIndex):
* dom/IconURL.h:
(WebCore::IconURL::IconURL):
* html/HTMLLinkElement.cpp:
(WebCore::HTMLLinkElement::process):
* loader/LinkLoader.cpp:
(WebCore::LinkLoader::loadLink):
* loader/LinkLoader.h:
* loader/icon/IconController.cpp:
(WebCore::IconController::urlsForTypes):
(WebCore::IconController::appendToIconURLs):
(WebCore::IconController::defaultURL):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95593 268f45cc-cd09-0410-ab3c-d52691b4dbfc

garden-o-matic doesn't work in Safari 5.1 because Safari 5.1 lacks Function.prototype.bind
https://bugs.webkit.org/show_bug.cgi?id=68495

Reviewed by Dimitri Glazkov.

This patch adds an implementation of Function.prototype.bind if the browser lacks one.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base.js:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95592 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[css3-flexbox] cleanup padding width calculations
https://bugs.webkit.org/show_bug.cgi?id=68490

Reviewed by Tony Chang.

No new tests. Existing tests cover the refactor.

* rendering/RenderFlexibleBox.cpp:
(WebCore::RenderFlexibleBox::logicalBorderAndPaddingWidthForChild):
(WebCore::RenderFlexibleBox::logicalScrollbarHeightForChild):
(WebCore::RenderFlexibleBox::preferredLogicalContentWidthForFlexItem):
(WebCore::RenderFlexibleBox::computePreferredLogicalWidth):
(WebCore::RenderFlexibleBox::layoutAndPlaceChildrenInlineDirection):
* rendering/RenderFlexibleBox.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95591 268f45cc-cd09-0410-ab3c-d52691b4dbfc

This CL makes it possible to create a blob from Chromium.
https://bugs.webkit.org/show_bug.cgi?id=68286

Reviewed by Darin Fisher.

* WebKit.gyp:
* public/WebBlob.h: Added.
* src/WebBlob.cpp: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95590 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[chromium] rebaseline caused by r95567

* platform/chromium-cg-mac/fast/clip/overflow-border-radius-fixed-position-expected.png: Added.
* platform/chromium-linux-x86/fast/clip: Removed.
* platform/chromium-linux/fast/clip/overflow-border-radius-fixed-position-expected.png: Added.
* platform/chromium-mac/fast/clip/overflow-border-radius-fixed-position-expected.png: Added.
* platform/chromium-win/fast/clip/overflow-border-radius-fixed-position-expected.png: Added.
* platform/chromium-win/fast/clip/overflow-border-radius-fixed-position-expected.txt: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95589 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[NRWT] REGRESSION: Local loader tests are failing on machines that lost /tmp/LayoutTests symlink
https://bugs.webkit.org/show_bug.cgi?id=65781

Reviewed by Ryosuke Niwa.

Instead of making NRWT create the symlink, I made DumpRenderTree smart enough
to resolve the passed in url relative to the absolute url for the test.
For http tests, since the test url is an http url, we can't resolve relative
to the test path, and thus use a new LOCAL_RESOURCE_ROOT environment variable
for resolving.

I believe this is a better approach than the on used in the Qt and Chromium DRT's
(which resolves the path relative to the built location of the DRT executable)
and we should move this new code into a shared location in a follow-up patch.

It turns out that there was a second use for pathToLocalResource, used by one
test (http/tests/plugins/post-url-file.html) used for getting a path to /tmp.
To support this test I made the new pathToLocalResource smart enough to map
/tmp to DUMPRENDERTREE_TEMP (which is already defined for all ports).

* DumpRenderTree/mac/LayoutTestControllerMac.mm:
(LayoutTestController::pathToLocalResource):
* Scripts/webkitpy/layout_tests/port/webkit.py: Set LOCAL_RESOURCE_ROOT for use by DRT.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95588 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Need to specify a django version, instead of using default
https://bugs.webkit.org/show_bug.cgi?id=68398

Patch by Tom Zakrajsek <tomz@codeaurora.org> on 2011-09-20
Reviewed by Eric Seidel.

* QueueStatusServer/__init__.py:
* QueueStatusServer/main.py:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95587 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fix clang compile errors in Web Audio
https://bugs.webkit.org/show_bug.cgi?id=68384

Reviewed by David Levin.

No new tests; no net change in functionality.

* platform/graphics/MediaPlayer.h: Forward declare (instead of including directly)
    AudioSourceProvider.
* webaudio/AudioChannelMerger.cpp:
(WebCore::AudioChannelMerger::checkNumberOfChannelsForInput): Remove unused parameter name.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95586 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Get rid of PluginView::setBoundsSize from WebKit2
https://bugs.webkit.org/show_bug.cgi?id=68489

Reviewed by Dan Bernstein.

Remove m_boundsSize and the setter since they aren't used anywhere.

* WebProcess/Plugins/PluginView.cpp:
* WebProcess/Plugins/PluginView.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95585 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Update committers.py with a few more contributor contacts
https://bugs.webkit.org/show_bug.cgi?id=68487

Patch by Tom Zakrajsek <tomz@codeaurora.org> on 2011-09-20
Reviewed by Eric Seidel.

* Scripts/webkitpy/common/config/committers.py:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95584 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fix webkitpy unit tests after 95547 and 95549
https://bugs.webkit.org/show_bug.cgi?id=68483

Patch by Leandro Pereira <leandro@profusion.mobi> on 2011-09-20
Reviewed by Adam Barth.

Fix tests for ChangeLog.touched_files() and ChangeLog.parse_entries_from_file().

* Scripts/webkitpy/common/checkout/changelog_unittest.py: While testing
parse_entries_from_file(), use StringIO and entry.reviewer_text().
While testing parse_latest_entry_from_file(), expect the correct
touched_files() list.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95583 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Directory Upload: parent path can truncate first char of the correct path
https://bugs.webkit.org/show_bug.cgi?id=66695

Source/WebCore:

In some cases, if the parent path we compute ends with a separator
character like '/' or 'C:\', by adding one in an attempt to grab the
subsequent separator we actually grab a character of the real path,
which is incorrect.

Reviewed by Darin Fisher.

* html/FileInputType.cpp:
(WebCore::FileInputType::setFileList):

LayoutTests:

Add a test case for when the parent path is '/'.

Reviewed by Darin Fisher.

* fast/forms/input-file-directory-upload.html:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95582 268f45cc-cd09-0410-ab3c-d52691b4dbfc

garden-o-matic makes it hard to compare images when scroll bars appear
https://bugs.webkit.org/show_bug.cgi?id=68482

Reviewed by Eric Seidel.

Because we autoscale image width, the images rescale if we overflow
vertically and get a scrollbar.  That makes it hard to compare two sets
of images.  This patch reserves space for the scrollbar so we won't
have the layout change.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/common.css:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95581 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Invoke CachedResourceLoader::canRequest for all URLs in a redirect chain
https://bugs.webkit.org/show_bug.cgi?id=68279

Reviewed by Adam Barth.

Source/WebCore:

* loader/cache/CachedResourceLoader.h:
* loader/cache/CachedResourceRequest.cpp:
(WebCore::CachedResourceRequest::willSendRequest):

LayoutTests:

* http/tests/security/resources/cross-origin-xsl-redirect.xml:
* http/tests/security/cross-origin-xsl-redirect-BLOCKED-expected.txt:
* http/tests/security/cross-origin-xsl-redirect-BLOCKED.html:
* platform/chromium/http/tests/permissionclient/image-permissions-expected.txt:
* platform/chromium/http/tests/permissionclient/image-permissions.html:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95580 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Generate WebKit team's page out of committers.py
https://bugs.webkit.org/show_bug.cgi?id=68045

Fix per Daniel's review comment since I had misunderstood the code.

* team.html:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95579 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Generate WebKit team's page out of committers.py
https://bugs.webkit.org/show_bug.cgi?id=68045

Reviewed by Daniel Bates.

Add the initial implementation of team.html. This page automatically pulls data from committers.py
There are no areas of expertise but that'll be added once the bug 68061 is fixed.

* team.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95578 268f45cc-cd09-0410-ab3c-d52691b4dbfc

change RenderFlexibleBox to act on logical coordinates
https://bugs.webkit.org/show_bug.cgi?id=68129

Reviewed by David Hyatt.

Source/WebCore:

This makes RenderFlexibleBox respect direction and writing-mode.
We now properly support the default flex-flow value of "row".

Test: css3/flexbox/writing-modes.html

* rendering/RenderBlock.cpp:
(WebCore::RenderBlock::setLogicalLocationForChild):
* rendering/RenderBlock.h:
* rendering/RenderFlexibleBox.cpp:
(WebCore::RenderFlexibleBox::layoutBlock):
(WebCore::RenderFlexibleBox::logicalBorderWidthForChild):
(WebCore::RenderFlexibleBox::logicalPaddingWidthForChild):
(WebCore::RenderFlexibleBox::logicalScrollbarHeightForChild):
(WebCore::RenderFlexibleBox::marginStartStyleForChild):
(WebCore::RenderFlexibleBox::marginEndStyleForChild):
(WebCore::RenderFlexibleBox::preferredLogicalContentWidthForFlexItem):
(WebCore::RenderFlexibleBox::layoutInlineDirection):
(WebCore::RenderFlexibleBox::logicalPositiveFlexForChild):
(WebCore::RenderFlexibleBox::logicalNegativeFlexForChild):
(WebCore::RenderFlexibleBox::computePreferredLogicalWidth):
(WebCore::RenderFlexibleBox::runFreeSpaceAllocationAlgorithmInlineDirection):
(WebCore::RenderFlexibleBox::setLogicalOverrideSize):
(WebCore::RenderFlexibleBox::layoutAndPlaceChildrenInlineDirection):
* rendering/RenderFlexibleBox.h:

LayoutTests:

* css3/flexbox/resources/flexbox.js:
* css3/flexbox/writing-modes-expected.txt: Added.
* css3/flexbox/writing-modes.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95577 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Use DEFINE_STATIC_LOCAL to intentionally leak the static RefCountedHFONT object returned from
FontPlatformData::hashTableDeletedFontValue() to avoid global destructor ordering problems.
https://bugs.webkit.org/show_bug.cgi?id=67906

Patch by Marshall Greenblatt <marshall@chromium.org> on 2011-09-20
Reviewed by Adam Barth.

* platform/graphics/chromium/FontPlatformDataChromiumWin.cpp:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@95576 268f45cc-cd09-0410-ab3c-d52691b4dbfc