Timeline


and

01/13/09:

23:17 Changeset [39884] by ap@webkit.org

Reviewed by Anders Carlsson.

Test: http/tests/appcache/xhr-foreign-resource.html

https://bugs.webkit.org/show_bug.cgi?id=23256
Implement application cache foreign entries

  • loader/DocumentLoader.cpp: (WebCore::DocumentLoader::shouldLoadResourceFromApplicationCache): Removed a misplaced check for foreign resources that prevented them from being loaded altogether.
  • loader/appcache/ApplicationCacheGroup.cpp: (WebCore::ApplicationCacheGroup::selectCache): If a document that was loaded from appcache should be associated with a different cache, mark the resource as foreign in the original cache and start over.
  • loader/appcache/ApplicationCacheResource.cpp: (WebCore::ApplicationCacheResource::addType): This function can now be called after a cache is stored, so it no longer asserts the opposite.
  • loader/appcache/ApplicationCacheStorage.cpp: (WebCore::ApplicationCacheStorage::cacheGroupForURL): Fix database pass to correctly ignore in-memory caches that were already rejected. (WebCore::ApplicationCacheStorage::fallbackCacheGroupForURL): Ditto. (WebCore::ApplicationCacheStorage::storeUpdatedType): Added a function that updates type of an already stored cached resource.
  • loader/appcache/ApplicationCacheStorage.h: Added storeUpdatedType().
23:14 Changeset [39883] by ap@webkit.org

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=23286
Add a test for non-ASCII URIs in application cache

  • http/tests/appcache/cyrillic-uri-expected.txt: Added.
  • http/tests/appcache/cyrillic-uri.html: Added.
  • http/tests/appcache/resources/cyrillic-uri-form.html: Added.
  • http/tests/appcache/resources/cyrillic-uri.manifest: Added.
  • http/tests/appcache/resources/intercept: Added.
  • http/tests/appcache/resources/intercept/.htaccess: Added.
  • http/tests/appcache/resources/print-uri.php: Added.
17:44 Changeset [39882] by bdakin@apple.com

JavaScriptCore:

2009-01-13 Beth Dakin <bdakin@apple.com>

Reviewed by Darin Adler and Oliver Hunt.

<rdar://problem/6489314> REGRESSION: Business widget's front side
fails to render correctly when flipping widget

The problem here is that parseInt was parsing NaN as 0. This patch
corrects that by parsing NaN as NaN. This matches our old behavior
and Firefox.

  • runtime/JSGlobalObjectFunctions.cpp: (JSC::globalFuncParseInt):

LayoutTests:

2009-01-13 Beth Dakin <bdakin@apple.com>

Reviewed by Darin Adler and Oliver Hunt.

Updated test and results for <rdar://problem/6489314> REGRESSION:
Business widget's front side fails to render correctly when
flipping widget

parseInt(NaN) should be NaN.

  • fast/js/numeric-conversion-expected.txt:
  • fast/js/resources/numeric-conversion.js:
17:24 Changeset [39881] by mrowe@apple.com

Fix the Tiger WebCore build after r39880.

  • editing/Editor.cpp:

(WebCore::Editor::markBadGrammar):

  • loader/mac/DocumentLoaderMac.cpp:

(WebCore::DocumentLoader::schedule):
(WebCore::DocumentLoader::unschedule):

  • platform/mac/PasteboardMac.mm:

(WebCore::Pasteboard::writeSelection):

  • platform/mac/WebCoreObjCExtras.mm:

(WebCoreObjCFinalizeOnMainThread):

  • platform/network/mac/ResourceHandleMac.mm:

(WebCore::ResourceHandle::schedule):
(WebCore::ResourceHandle::unschedule):

16:16 Changeset [39880] by darin@apple.com

2009-01-13 Darin Adler <darin@apple.com>

Reviewed by Sam Weinig.

Bug 23102: turn on unused parameter warnings in WebCore
https://bugs.webkit.org/show_bug.cgi?id=23102

Last step: Turn on the warning and fix the last few instances.

  • Configurations/Base.xcconfig: Removed -Wno-unused-parameter from WARNING_CFLAGS_BASE, so the unused parameter warning will fire.
  • bindings/js/JSDOMWindowBase.cpp: (jsDOMWindowBaseMessageChannel): Separate the version for use when CHANNEL_MESSAGING is turned off, so we can omit the argument names. It would be better to not even have this property in that case, rather than having the function return undefined, but I don't know how to do that with the bindings script.
  • bindings/scripts/CodeGeneratorJS.pm: Leave out argument names in functions where they're not used (the SVG context, and the property name in the index getter). Use UNUSED_PARAM where arguments are sometimes not used.
  • bridge/runtime_object.cpp: (JSC::RuntimeObjectImp::put): Use UNUSED_PARAM for non-QT.
  • css/MediaQueryEvaluator.cpp: (WebCore::animationMediaFeatureEval): Omit argument names for style and frame. (WebCore::transitionMediaFeatureEval): Ditto. (WebCore::transform_2dMediaFeatureEval): Ditto. (WebCore::transform_3dMediaFeatureEval): Ditto.
  • dom/make_names.pl: Leave out the createdByParser property name in cases where it's not used.
  • inspector/JavaScriptDebugServer.cpp: (WebCore::JavaScriptDebugServer::setJavaScriptPaused): Separate the version for the Mac platform so we can omit the argument names.
  • loader/EmptyClients.h: (WebCore::EmptyChromeClient::contentsSizeChanged): Removed argument name.
  • loader/FrameLoaderClient.h: (WebCore::FrameLoaderClient::shouldUsePluginDocument): Commented out argument name.
  • loader/icon/IconFetcher.cpp: (WebCore::IconFetcher::didReceiveResponse): Use ASSERT_UNUSED. (WebCore::IconFetcher::didReceiveData): Ditto. (WebCore::IconFetcher::didFinishLoading): Ditto. (WebCore::IconFetcher::didFail): Ditto.
  • platform/KURL.cpp: (WebCore::assertProtocolIsGood): Separate out the NDEBUG version so we can omit the argument names.
  • platform/Timer.cpp: (WebCore::TimerHeapIterator::checkConsistency): Use ASSERT_UNUSED.
  • platform/graphics/GlyphBuffer.h: (WebCore::GlyphBuffer::offsetAt): Use UNUSED_PARAM in the non-WIN case. (WebCore::GlyphBuffer::add): Use UNUSED_PARAM in the non-CAIRO case.
  • platform/graphics/cg/ImageSourceCG.cpp: (WebCore::ImageSource::clear): Removed argument name.
  • rendering/SVGRenderSupport.cpp: (WebCore::prepareToRenderSVGContent): Use UNUSED_PARAM when SVG_FILTERS is not enabled. (WebCore::finishRenderSVGContent): Ditto.
16:15 Changeset [39879] by barraclough@apple.com

2009-01-13 Gavin Barraclough <barraclough@apple.com>

Reviewed by Oliver Hunt.

Fix for: https://bugs.webkit.org/show_bug.cgi?id=23292

Implementation of two argument canDoFastAdditiveOperations does not correlate well with reality.

  • runtime/JSImmediate.h: (JSC::JSFastMath::canDoFastAdditiveOperations):
15:40 ApplicationsGtk edited by greut.lists@dosimple.ch
fixed Empathy's link label (diff)
14:47 Changeset [39878] by cmarrin@apple.com

Implemented https://bugs.webkit.org/show_bug.cgi?id=23298
Simple fix to convertToLength to catch style==null cases

  • css/CSSStyleSelector.cpp: (WebCore::convertToLength):
14:36 Changeset [39877] by andersca@apple.com

WebKitTools:

2009-01-13 Anders Carlsson <andersca@apple.com>

Reviewed by Sam Weinig.

Add NPRuntime test.

  • DumpRenderTree/TestNetscapePlugIn.subproj/PluginObject.cpp: (testNPRuntime):
  • DumpRenderTree/TestNetscapePlugIn.subproj/PluginObject.h:
  • DumpRenderTree/TestNetscapePlugIn.subproj/main.cpp: (NPP_New): (NPP_GetValue):

LayoutTests:

2009-01-13 Anders Carlsson <andersca@apple.com>

Reviewed by Sam Weinig.

Add NPRuntime test.


  • plugins/npruntime-expected.txt: Added.
  • plugins/npruntime.html: Added.
13:15 Changeset [39876] by eric@webkit.org

Reviewed by Mark Rowe.

Speculative fix for an Uninitialized Memory Read (UMR)
seen by purify in chormium's equivalent V8 file.

  • bindings/js/JSSVGPODTypeWrapper.h: (WebCore::PODTypeWrapperCacheInfo::PODTypeWrapperCacheInfo):
12:55 Changeset [39875] by dglazkov@chromium.org

2009-01-13 Dimitri Glazkov <dglazkov@chromium.org>

Reviewed by Eric Seidel.

Land portions of platform/graphics/chromium.
https://bugs.webkit.org/show_bug.cgi?id=23297

  • platform/graphics/chromium/FontCustomPlatformData.cpp: Added. (WebCore::FontCustomPlatformData::~FontCustomPlatformData): (WebCore::FontCustomPlatformData::fontPlatformData): (WebCore::EOTStream::EOTStream): (WebCore::EOTStream::read): (WebCore::readEmbedProc): (WebCore::createUniqueFontName): (WebCore::createFontCustomPlatformData):
  • platform/graphics/chromium/FontCustomPlatformData.h: Added. (WebCore::FontCustomPlatformData::FontCustomPlatformData):
  • platform/graphics/chromium/FontLinux.cpp: Added. (WebCore::Font::drawGlyphs): (WebCore::Font::drawComplexText): (WebCore::Font::floatWidthForComplexText): (WebCore::Font::offsetForPositionForComplexText): (WebCore::Font::selectionRectForComplexText):
  • platform/graphics/chromium/FontPlatformData.h: Added.
12:44 Changeset [39874] by hausmann@webkit.org

2009-01-08 Yongjun Zhang <yongjun.zhang@nokia.com>

Reviewed by Simon Hausmann.

https://bugs.webkit.org/show_bug.cgi?id=23187

Update webview with the intersected rect.

In ChromeClientQt::repaint, view should be updated with the
intersected rect, not the whole windowRect;

This generally is not a problem for normal viewport setup where
viewport size is the same as the qwebview widget size. However, if we
set the viewport size smaller than qwebkit widget, we will see
unwanted painting outside the viewport.

12:43 Changeset [39873] by ariya@webkit.org

2009-01-13 Ariya Hidayat <ariya.hidayat@trolltech.com>

Rubber-stamped by Tor Arne Vestbø.

[Qt] Missing piece from my previous commit:
Disable and enable the cursor navigation actions when contentEditable
is changed.

  • Api/qwebpage.cpp: (QWebPagePrivate::updateEditorActions):
12:24 Changeset [39872] by mrowe@apple.com

Don't copy DOMStringList.idl into the framework bundle.

Rubber-stamped by Oliver Hunt.

  • WebCore.xcodeproj/project.pbxproj:
12:22 WebKit Team edited by vestbo@webkit.org
(diff)
12:07 Changeset [39871] by mrowe@apple.com

https://bugs.webkit.org/show_bug.cgi?id=23290
Fix JSImmediate::isImmediate(src) to !src->isCell()

Reviewed by Darin Adler.

  • interpreter/Interpreter.cpp:

(JSC::Interpreter::privateExecute):

12:07 Changeset [39870] by ariya@webkit.org

2009-01-13 Ariya Hidayat <ariya.hidayat@trolltech.com>

Reviewed by Tor Arne Vestbø.

[Qt] Create actions for cursor navigation for QWebPage.
Also properly disable and enable them when contentEditable is toggled.

  • Api/qwebpage.cpp: (QWebPagePrivate::updateAction): (QWebPage::action):
11:26 Changeset [39869] by andersca@apple.com

2009-01-12 Anders Carlsson <andersca@apple.com>

Reviewed by Darin Adler.

Add a bunch of methods to WebKitPluginClient.defs, and implement them.


  • Plugins/Hosted/NetscapePluginHostProxy.mm: (WebKit::fromUTF8WithLatin1Fallback): If the length isn't specified, get it by calling strlen.


(WKPCEvaluate):
Evaluate doesn't take any arguments.


(WKPCGetIntIdentifier):
Call _NPN_GetIntIdentifier.


(identifierFromServerIdentifier):
New helper function that returns a JSC Identifier from an NPIdentifier.


(WKPCInvoke):
Call identifierFromServerIdentifier.


(WKPCRemoveProperty):
(WKPCHasProperty):
(WKPCHasMethod):
Call NetscapePluginInstanceProxy.

  • Plugins/Hosted/NetscapePluginInstanceProxy.h:
  • Plugins/Hosted/NetscapePluginInstanceProxy.mm: (WebKit::NetscapePluginInstanceProxy::removeProperty): (WebKit::NetscapePluginInstanceProxy::hasProperty): (WebKit::NetscapePluginInstanceProxy::hasMethod):
  • Plugins/Hosted/WebKitPluginClient.defs: Add new definitions.
11:25 Changeset [39868] by christian@webkit.org

2009-01-13 Christian Dywan <christian@twotoasts.de>

Build fix, StorageEvent is only defined if we have DOM_STORAGE

  • dom/Document.cpp: (WebCore::Document::createEvent): Conditionalize StorageEvent
11:17 Changeset [39867] by andersca@apple.com

2009-01-13 Anders Carlsson <andersca@apple.com>

Fix build.

  • WebView/WebView.mm: (-[WebView _initWithArguments:]):
11:13 Changeset [39866] by dglazkov@chromium.org

2009-01-13 Dmitry Titov <dimich@chromium.org>

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=23281
Fix the Chromium Win build.
Need to use PLATFORM(WIN_OS) instead of PLATFORM(WIN).
Moved GTK and WX up in #if sequence because they could come with WIN_OS too,
while they have their own implementation even on Windows.

  • wtf/CurrentTime.cpp: (WTF::currentTime):
10:12 Changeset [39865] by timothy@apple.com

Adds a workaround for the flip4mac installer plugin decoding a WebView from a NIB on a secondary thread.

<rdar://problem/6489788> New WebKit thread checks break installation of flip4mac (thread violation)

Reviewed by Darin Adler.

  • Misc/WebKitVersionChecks.h: Add WEBKIT_FIRST_VERSION_WITHOUT_WEBVIEW_INIT_THREAD_WORKAROUND.
  • Misc/WebNSObjectExtras.h: Add _webkit_performSelectorOnMainThread:withObject:.
  • Misc/WebNSObjectExtras.mm: (-[NSObject _webkit_performSelectorWithArguments:]): Renamed from _webkit_getPropertyWithArguments. Passes the optional object to the selector. (-[NSObject _webkit_performSelectorOnMainThread:withObject:]): Renamed from _webkit_getPropertyOnMainThread:. Put the optional object into the arguments dictionary. (-[NSObject _webkit_getPropertyOnMainThread:]): Call _webkit_performSelectorOnMainThread with a nil object.
  • WebView/WebResource.mm: (-[WebResource _ignoreWhenUnarchiving]): Use _cmd instead of making the selector again. (-[WebResource _initWithData:URL:MIMEType:textEncodingName:frameName:response:copyData:]): Use the new _webkit_performSelectorOnMainThread:withObject: method instead of performSelectorOnMainThread.
  • WebView/WebView.mm: (-[WebView _initWithArguments:]): Added. Pulls arguments out of the dictionary and calls the right init method. (needsWebViewInitThreadWorkaround): Checks if the thead is not the main thread and if we are in the Installer bundle. (-[WebView initWithFrame:frameName:groupName:]): Call needsWebViewInitThreadWorkaround and use _webkit_performSelectorOnMainThread to call _initWithArguments: passing the frame, frameName and groupName. (-[WebView initWithCoder:]): Ditto, except pass the coder to _initWithArguments:.
09:59 Changeset [39864] by darin@apple.com

2009-01-13 Darin Adler <darin@apple.com>

Reviewed by Adele Peterson.

Bug 23277: space bar scrolls the page when typed in any text field
https://bugs.webkit.org/show_bug.cgi?id=23277
rdar://problem/6490249

There's already a regression test for this, since bug 16421 was the same problem, but on
Mac. The regression test is fast/events/space-scroll-event.html, so I won't write a new one.

This bug doesn't affect Mac because the canEdit check in -[WebHTMLView insertText:] prevents
a text input event from being generated at all. That difference may in itself be a bug, but
the only symptom is that you get text input events in more cases on platforms other than Mac.

  • page/EventHandler.cpp: (WebCore::EventHandler::handleTextInputEvent): Return whether the event's default was handled, not the result of the dispatchEvent function, which indicates only whether the default behavior was prevented, not if the default behavior did some insertion. This fix makes it so the keyboard event handler can detect that no space was inserted and know that the space bar keypress event can be treated as a scrolling command. (WebCore::EventHandler::defaultKeyboardEventHandler): Handle the keypress event, not the keydown event. Without the change above, making this change would have restored the old behavior, reintroducing bug 22913 where the space bar doesn't scroll at all on Windows.
09:58 Changeset [39863] by hausmann@webkit.org

2009-01-13 Simon Hausmann <simon.hausmann@nokia.com>

Reviewed by Tor Arne Vestbø.

https://bugs.webkit.org/show_bug.cgi?id=23171

Based on patch by Yael Aharon <yael.aharon@nokia.com>

Make the QWebSettings functions for setting the path for the local
storage database private QtWebkit API, including the ability to
associated a QWebPage with a named page group.

09:24 Changeset [39862] by vestbo@webkit.org

2009-01-13 Tor Arne Vestbø <tavestbo@trolltech.com>

Reviewed by Simon Hausmann.

[Qt] Make sure media elements dispatch the 'loaded' event

We assume that when Phonon goes into paused state that we have the
complete media file. Once we do media loading ourselves we can
distinguish between loading the first frame and the complete media.

08:48 Changeset [39861] by darin@chromium.org

2009-01-13 Darin Fisher <darin@chromium.org>

Reviewed by Eric Seidel.

Landing portions of platform/chromium/, taken from here:
http://src.chromium.org/viewvc/chrome/trunk/deps/third_party/WebKit/WebCore/platform/chromium/

  • platform/chromium/ChromiumBridge.h: Added.
  • platform/chromium/ChromiumDataObject.cpp: Added. (WebCore::ChromiumDataObject::clear): (WebCore::ChromiumDataObject::hasData):
  • platform/chromium/ChromiumDataObject.h: Added. (WebCore::ChromiumDataObject::create): (WebCore::ChromiumDataObject::ChromiumDataObject):
  • platform/chromium/ClipboardChromium.cpp: Added. (WebCore::): (WebCore::clipboardTypeFromMIMEType): (WebCore::ClipboardChromium::ClipboardChromium): (WebCore::ClipboardChromium::create): (WebCore::ClipboardChromium::clearData): (WebCore::ClipboardChromium::clearAllData): (WebCore::ClipboardChromium::getData): (WebCore::ClipboardChromium::setData): (WebCore::ClipboardChromium::types): (WebCore::ClipboardChromium::setDragImage): (WebCore::ClipboardChromium::setDragImageElement): (WebCore::ClipboardChromium::createDragImage): (WebCore::imageToMarkup): (WebCore::getCachedImage): (WebCore::writeImageToDataObject): (WebCore::ClipboardChromium::declareAndWriteDragImage): (WebCore::ClipboardChromium::writeURL): (WebCore::ClipboardChromium::writeRange): (WebCore::ClipboardChromium::hasData):
  • platform/chromium/ClipboardChromium.h: Added. (WebCore::ClipboardChromium::~ClipboardChromium): (WebCore::ClipboardChromium::dataObject):
  • platform/chromium/ClipboardUtilitiesChromium.cpp: Added. (WebCore::replaceNewlinesWithWindowsStyleNewlines): (WebCore::replaceNBSPWithSpace): (WebCore::urlToMarkup):
  • platform/chromium/ClipboardUtilitiesChromium.h: Added.
  • platform/chromium/ContextMenuChromium.cpp: Added. (WebCore::ContextMenu::ContextMenu): (WebCore::ContextMenu::~ContextMenu): (WebCore::ContextMenu::itemCount): (WebCore::ContextMenu::insertItem): (WebCore::ContextMenu::appendItem): (WebCore::ContextMenu::itemWithAction): (WebCore::ContextMenu::itemAtIndex): (WebCore::ContextMenu::setPlatformDescription): (WebCore::ContextMenu::platformDescription): (WebCore::ContextMenu::releasePlatformDescription):
  • platform/chromium/ContextMenuItemChromium.cpp: Added. (WebCore::ContextMenuItem::ContextMenuItem): (WebCore::ContextMenuItem::~ContextMenuItem): (WebCore::ContextMenuItem::releasePlatformDescription): (WebCore::ContextMenuItem::type): (WebCore::ContextMenuItem::action): (WebCore::ContextMenuItem::title): (WebCore::ContextMenuItem::platformSubMenu): (WebCore::ContextMenuItem::setType): (WebCore::ContextMenuItem::setAction): (WebCore::ContextMenuItem::setTitle): (WebCore::ContextMenuItem::setSubMenu): (WebCore::ContextMenuItem::setChecked): (WebCore::ContextMenuItem::setEnabled): (WebCore::ContextMenuItem::enabled):
  • platform/chromium/CursorChromium.cpp: Added. (WebCore::Cursor::Cursor): (WebCore::Cursor::~Cursor): (WebCore::Cursor::operator=): (WebCore::pointerCursor): (WebCore::crossCursor): (WebCore::handCursor): (WebCore::iBeamCursor): (WebCore::waitCursor): (WebCore::helpCursor): (WebCore::eastResizeCursor): (WebCore::northResizeCursor): (WebCore::northEastResizeCursor): (WebCore::northWestResizeCursor): (WebCore::southResizeCursor): (WebCore::southEastResizeCursor): (WebCore::southWestResizeCursor): (WebCore::westResizeCursor): (WebCore::northSouthResizeCursor): (WebCore::eastWestResizeCursor): (WebCore::northEastSouthWestResizeCursor): (WebCore::northWestSouthEastResizeCursor): (WebCore::columnResizeCursor): (WebCore::rowResizeCursor): (WebCore::middlePanningCursor): (WebCore::eastPanningCursor): (WebCore::northPanningCursor): (WebCore::northEastPanningCursor): (WebCore::northWestPanningCursor): (WebCore::southPanningCursor): (WebCore::southEastPanningCursor): (WebCore::southWestPanningCursor): (WebCore::westPanningCursor): (WebCore::moveCursor): (WebCore::verticalTextCursor): (WebCore::cellCursor): (WebCore::contextMenuCursor): (WebCore::aliasCursor): (WebCore::progressCursor): (WebCore::noDropCursor): (WebCore::copyCursor): (WebCore::noneCursor): (WebCore::notAllowedCursor): (WebCore::zoomInCursor): (WebCore::zoomOutCursor): (WebCore::grabCursor): (WebCore::grabbingCursor):
  • platform/chromium/DragDataChromium.cpp: Added. (WebCore::containsHTML): (WebCore::DragData::createClipboard): (WebCore::DragData::containsURL): (WebCore::DragData::asURL): (WebCore::DragData::containsFiles): (WebCore::DragData::asFilenames): (WebCore::DragData::containsPlainText): (WebCore::DragData::asPlainText): (WebCore::DragData::containsColor): (WebCore::DragData::canSmartReplace): (WebCore::DragData::containsCompatibleContent): (WebCore::DragData::asFragment): (WebCore::DragData::asColor):
  • platform/chromium/EditorChromium.cpp: Added. (WebCore::Editor::newGeneralClipboard):
  • platform/chromium/FileChooserChromium.cpp: Added. (WebCore::FileChooser::basenameForWidth):
  • platform/chromium/FileSystemChromium.cpp: Added. (WebCore::deleteFile): (WebCore::deleteEmptyDirectory): (WebCore::getFileSize): (WebCore::getFileModificationTime): (WebCore::directoryName): (WebCore::pathByAppendingComponent): (WebCore::makeAllDirectories): (WebCore::fileExists):
  • platform/chromium/FileSystemChromiumLinux.cpp: Added. (WebCore::pathGetFileName):
  • platform/chromium/FileSystemChromiumMac.mm: Added. (WebCore::pathGetFileName):
  • platform/chromium/FileSystemChromiumWin.cpp: Added. (WebCore::pathGetFileName):
  • platform/chromium/FramelessScrollView.cpp: Added. (WebCore::FramelessScrollView::~FramelessScrollView): (WebCore::FramelessScrollView::invalidateScrollbarRect): (WebCore::FramelessScrollView::isActive): (WebCore::FramelessScrollView::invalidateRect): (WebCore::FramelessScrollView::hostWindow): (WebCore::FramelessScrollView::windowClipRect): (WebCore::FramelessScrollView::paintContents): (WebCore::FramelessScrollView::contentsResized): (WebCore::FramelessScrollView::visibleContentsResized):
  • platform/chromium/FramelessScrollView.h: Added. (WebCore::FramelessScrollView::FramelessScrollView): (WebCore::FramelessScrollView::client): (WebCore::FramelessScrollView::setClient):
  • platform/chromium/FramelessScrollViewClient.h: Added.
  • platform/chromium/ScrollbarThemeChromiumWin.cpp: Added. (WebCore::): (WebCore::ScrollbarThemeChromium::scrollbarThickness): (WebCore::ScrollbarThemeChromium::invalidateOnMouseEnterExit): (WebCore::ScrollbarThemeChromium::paintTrackPiece): (WebCore::ScrollbarThemeChromium::paintButton): (WebCore::ScrollbarThemeChromium::paintThumb): (WebCore::ScrollbarThemeChromium::getThemeState): (WebCore::ScrollbarThemeChromium::getThemeArrowState): (WebCore::ScrollbarThemeChromium::getClassicThemeState):
  • platform/chromium/SoundChromiumPosix.cpp: Added. (WebCore::systemBeep):
  • platform/chromium/SoundChromiumWin.cpp: Added. (WebCore::systemBeep):
  • platform/chromium/SystemTimeChromium.cpp: Added. (WebCore::currentTime): (WebCore::userIdleTime):
  • platform/chromium/WidgetChromium.cpp: Added. (WebCore::Widget::Widget): (WebCore::Widget::~Widget): (WebCore::Widget::show): (WebCore::Widget::hide): (WebCore::Widget::setCursor): (WebCore::Widget::paint): (WebCore::Widget::setFocus): (WebCore::Widget::setIsSelected): (WebCore::Widget::frameRect): (WebCore::Widget::setFrameRect):
  • platform/chromium/WindowsVersion.cpp: Added. (WebCore::isVistaOrNewer):
  • platform/chromium/WindowsVersion.h: Added.
08:38 Changeset [39860] by vestbo@webkit.org

2009-01-13 Tor Arne Vestbø <tavestbo@trolltech.com>

Reviewed by Simon Hausmann.

[Qt] Add more verbose error output when loading NPAPI plugins

08:17 Changeset [39859] by hausmann@webkit.org

2009-01-13 Simon Hausmann <simon.hausmann@nokia.com>

Reviewed by Tor Arne Vestbø.

Fix crash in the Qt port when deleting a popup from within a
JavaScript onchange handler.

06:58 Changeset [39858] by zecke@webkit.org

[GIO] Make sure to untef the GFile and the GInputStream on destruction

Make sure to call cleanupGioOperation from the ResourceHandleInternal
d'tor. Change the signature of cleanupGioOperation so we can do the
cleanup of GIO next to the cleanup of SOUP.

06:58 Changeset [39857] by zecke@webkit.org

[GIO] Attach the ResourceHandle to the GObject and get it from the GObject

https://bugs.webkit.org/show_bug.cgi?id=23116 discussed with Benjamin Otte

Sometimes it is not possible to cancel a pending GIO operation and the
callback might be invoked with a pointer to an already destructed
ResourceHandle. To avoid this issue we will attach the ResourceHandle
to the GObject* and in the callback attempt to get the ResourceHandle
from the GObject*. If no ResourceHandle is attached we silently fail
as the cleanup has been already done.

06:58 Changeset [39856] by zecke@webkit.org

[GIO] Invoke cleanupGioOperation before calling didFail to avoid crashes

https://bugs.webkit.org/show_bug.cgi?id=23116

The ResourceHandle becomes invalid (e.g. got destructed) from within
the invocation of didFail. Avoid this by cleaning up everything before
calling didFail and simply return afterwards.

This is happening with many tests from dom/html/level2/html e.g.
LayoutTests/dom/html/level2/html/HTMLBodyElement07.html.

06:57 Changeset [39855] by zecke@webkit.org

2009-01-13 Alexander V. Butenko <alex@digiqube.com>

Reviewed by Holger Freyther.

https://bugs.webkit.org/show_bug.cgi?id=23279

Fix crash on WebView dispose stage.

Add null checks to webkit_web_view_set_scroll_adjustments and
webkit_web_view_get_accessible as they get called from within
the dispose.

  • webkit/webkitwebview.cpp:
06:51 Changeset [39854] by vestbo@webkit.org

2009-01-13 Tor Arne Vestbø <tavestbo@trolltech.com>

Reviewed by Simon Hausmann.

Change how themes adjust mediaControls.css to match html4/quicks.css

Instead of providing the full style sheet, the themes provide extra
overrides to the default style defined in UserAgentStyleSheetsData.

https://bugs.webkit.org/show_bug.cgi?id=23210

Also, merge WebKitResources.qrc and WebCoreResources.qrc to speed up
build time for the Qt port.

00:11 Changeset [39853] by sfalken@apple.com

Build fix.

  • WebScriptCallFrame.cpp: (WebScriptCallFrame::jsValueToString):

01/12/09:

23:54 Changeset [39852] by mrowe@apple.com

Use the modern spelling of WebKit.

  • Scripts/find-extra-includes:
  • Scripts/report-include-statistics:
  • Scripts/run-webkit-app:
  • Scripts/svn-unapply:
  • Scripts/update-webkit:
20:51 Changeset [39851] by barraclough@apple.com

JavaScriptCore:

2009-01-12 Gavin Barraclough <barraclough@apple.com>

Reviewed by Oliver Hunt.

Make the JSImmediate interface private.

All manipulation of JS values should be through the JSValuePtr class, not by using JSImmediate
directly. The key missing methods on JSValuePtr are:

  • isCell() - check for values that are JSCell*s, and as such where asCell() may be used.
  • isInt32Fast() getInt32Fast() - fast check/access for integer immediates.
  • isUInt32Fast() getUInt32Fast() - ditto for unsigned integer immediates.

The JIT is allowed full access to JSImmediate, since it needs to be able to directly
manipulate JSValuePtrs. The Interpreter is provided access to perform operations directly
on JSValuePtrs through the new JSFastMath interface.

No performance impact.

  • API/JSCallbackObjectFunctions.h: (JSC::::toNumber):
  • API/JSValueRef.cpp: (JSValueIsEqual): (JSValueIsStrictEqual):
  • JavaScriptCore.exp:
  • bytecode/CodeBlock.h: (JSC::CodeBlock::isKnownNotImmediate):
  • bytecompiler/BytecodeGenerator.cpp: (JSC::keyForImmediateSwitch):
  • bytecompiler/BytecodeGenerator.h: (JSC::BytecodeGenerator::JSValueHashTraits::constructDeletedValue): (JSC::BytecodeGenerator::JSValueHashTraits::isDeletedValue):
  • interpreter/Interpreter.cpp: (JSC::jsLess): (JSC::jsLessEq): (JSC::jsAdd): (JSC::jsIsObjectType): (JSC::cachePrototypeChain): (JSC::Interpreter::tryCachePutByID): (JSC::Interpreter::tryCacheGetByID): (JSC::Interpreter::privateExecute): (JSC::Interpreter::tryCTICachePutByID): (JSC::Interpreter::tryCTICacheGetByID): (JSC::Interpreter::cti_op_add): (JSC::Interpreter::cti_op_get_by_id_self_fail): (JSC::Interpreter::cti_op_get_by_id_proto_list): (JSC::Interpreter::cti_op_instanceof): (JSC::Interpreter::cti_op_mul): (JSC::Interpreter::cti_op_get_by_val): (JSC::Interpreter::cti_op_get_by_val_byte_array): (JSC::Interpreter::cti_op_sub): (JSC::Interpreter::cti_op_put_by_val): (JSC::Interpreter::cti_op_put_by_val_array): (JSC::Interpreter::cti_op_put_by_val_byte_array): (JSC::Interpreter::cti_op_negate): (JSC::Interpreter::cti_op_div): (JSC::Interpreter::cti_op_eq): (JSC::Interpreter::cti_op_lshift): (JSC::Interpreter::cti_op_bitand): (JSC::Interpreter::cti_op_rshift): (JSC::Interpreter::cti_op_bitnot): (JSC::Interpreter::cti_op_neq): (JSC::Interpreter::cti_op_urshift): (JSC::Interpreter::cti_op_call_eval): (JSC::Interpreter::cti_op_throw): (JSC::Interpreter::cti_op_is_undefined): (JSC::Interpreter::cti_op_stricteq): (JSC::Interpreter::cti_op_nstricteq): (JSC::Interpreter::cti_op_switch_imm): (JSC::Interpreter::cti_vm_throw):
  • interpreter/Interpreter.h: (JSC::Interpreter::isJSArray): (JSC::Interpreter::isJSString): (JSC::Interpreter::isJSByteArray):
  • jit/JIT.cpp: (JSC::JIT::compileOpStrictEq): (JSC::JIT::privateCompileMainPass):
  • jit/JIT.h: (JSC::JIT::isStrictEqCaseHandledInJITCode):
  • jit/JITArithmetic.cpp: (JSC::JIT::compileFastArith_op_rshift): (JSC::JIT::compileFastArith_op_bitand): (JSC::JIT::compileFastArith_op_mod):
  • jit/JITCall.cpp: (JSC::JIT::unlinkCall): (JSC::JIT::compileOpCall):
  • jit/JITInlineMethods.h: (JSC::JIT::getConstantOperandImmediateInt): (JSC::JIT::isOperandConstantImmediateInt):
  • parser/Nodes.cpp: (JSC::processClauseList):
  • runtime/ArrayPrototype.cpp: (JSC::arrayProtoFuncIndexOf): (JSC::arrayProtoFuncLastIndexOf):
  • runtime/BooleanPrototype.cpp: (JSC::booleanProtoFuncValueOf):
  • runtime/Collector.cpp: (JSC::Heap::protect): (JSC::Heap::unprotect): (JSC::Heap::heap):
  • runtime/JSByteArray.cpp: (JSC::JSByteArray::getOwnPropertySlot):
  • runtime/JSByteArray.h: (JSC::JSByteArray::getIndex):
  • runtime/JSCell.cpp:
  • runtime/JSCell.h: (JSC::JSValuePtr::isNumberCell): (JSC::JSValuePtr::asCell): (JSC::JSValuePtr::isNumber):
  • runtime/JSGlobalObjectFunctions.cpp: (JSC::globalFuncParseInt):
  • runtime/JSImmediate.h: (JSC::js0): (JSC::jsImpossibleValue): (JSC::JSValuePtr::toInt32): (JSC::JSValuePtr::toUInt32): (JSC::JSValuePtr::isCell): (JSC::JSValuePtr::isInt32Fast): (JSC::JSValuePtr::getInt32Fast): (JSC::JSValuePtr::isUInt32Fast): (JSC::JSValuePtr::getUInt32Fast): (JSC::JSValuePtr::makeInt32Fast): (JSC::JSValuePtr::areBothInt32Fast): (JSC::JSFastMath::canDoFastBitwiseOperations): (JSC::JSFastMath::equal): (JSC::JSFastMath::notEqual): (JSC::JSFastMath::andImmediateNumbers): (JSC::JSFastMath::xorImmediateNumbers): (JSC::JSFastMath::orImmediateNumbers): (JSC::JSFastMath::canDoFastRshift): (JSC::JSFastMath::canDoFastUrshift): (JSC::JSFastMath::rightShiftImmediateNumbers): (JSC::JSFastMath::canDoFastAdditiveOperations): (JSC::JSFastMath::addImmediateNumbers): (JSC::JSFastMath::subImmediateNumbers): (JSC::JSFastMath::incImmediateNumber): (JSC::JSFastMath::decImmediateNumber):
  • runtime/JSNumberCell.h: (JSC::JSValuePtr::asNumberCell): (JSC::jsNumber): (JSC::JSValuePtr::uncheckedGetNumber): (JSC::JSNumberCell::toInt32): (JSC::JSNumberCell::toUInt32): (JSC::JSValuePtr::toJSNumber): (JSC::JSValuePtr::getNumber): (JSC::JSValuePtr::numberToInt32): (JSC::JSValuePtr::numberToUInt32):
  • runtime/JSObject.h: (JSC::JSValuePtr::isObject): (JSC::JSValuePtr::get): (JSC::JSValuePtr::put):
  • runtime/JSValue.cpp: (JSC::JSValuePtr::toInteger): (JSC::JSValuePtr::toIntegerPreserveNaN):
  • runtime/JSValue.h:
  • runtime/Operations.cpp: (JSC::JSValuePtr::equalSlowCase): (JSC::JSValuePtr::strictEqualSlowCase):
  • runtime/Operations.h: (JSC::JSValuePtr::equal): (JSC::JSValuePtr::equalSlowCaseInline): (JSC::JSValuePtr::strictEqual): (JSC::JSValuePtr::strictEqualSlowCaseInline):
  • runtime/Protect.h: (JSC::gcProtect): (JSC::gcUnprotect):
  • runtime/StringPrototype.cpp: (JSC::stringProtoFuncCharAt): (JSC::stringProtoFuncCharCodeAt):
  • runtime/Structure.cpp: (JSC::Structure::createCachedPrototypeChain):

WebCore:

2009-01-12 Gavin Barraclough <barraclough@apple.com>

Reviewed by Oliver Hunt.

Deprecate JSValuePtr::getNumber() - two ways to get a number should be enough.

  • bindings/js/JSSQLTransactionCustom.cpp: (WebCore::JSSQLTransaction::executeSql):
  • bindings/objc/WebScriptObject.mm: (+[WebScriptObject _convertValueToObjcValue:originRootObject:rootObject:]):

WebKit/mac:

2009-01-12 Gavin Barraclough <barraclough@apple.com>

Reviewed by Oliver Hunt.

Deprecate JSValuePtr::getNumber() - two ways to get a number should be enough.

  • WebView/WebView.mm: (aeDescFromJSValue):
18:33 Changeset [39850] by beidson@apple.com

2009-01-12 Brady Eidson <beidson@apple.com>

Reviewed by NOBODY (OOPS!).


<rdar://problem/6490446> - Crash when going back to a cached page

  • WebCoreSupport/WebFrameLoaderClient.cpp: (WebFrameLoaderClient::savePlatformDataToCachedPage): ENABLE(CFNETWORK) needed to be USE(CFNETWORK)
17:09 Changeset [39849] by cmarrin@apple.com

Fixed https://bugs.webkit.org/show_bug.cgi?id=22919

I added a common getElapsedTime() function to AnimationBase which uses a combination of
the techniques from KeyframeAnimation and AnimationBase to compute elapsedTime in all
states.

  • page/animation/AnimationBase.cpp: (WebCore::AnimationBase::progress): (WebCore::AnimationBase::getElapsedTime):
  • page/animation/AnimationBase.h:
  • page/animation/KeyframeAnimation.cpp: (WebCore::KeyframeAnimation::animate):
16:22 Changeset [39848] by bfulgham@webkit.org

Correct scrolldelay units (s -> ms)

16:20 Changeset [39847] by simon.fraser@apple.com

2009-01-12 Simon Fraser <simon.fraser@apple.com>

Reviewed by Dan Bernstein

https://bugs.webkit.org/show_bug.cgi?id=23273

In RenderBox::repaintLayerRectsForImage(), the repaint rect does not have
to be computed in absolute coordintes. Instead, we compute a repaintRect
relative to the RenderObject, which can repaint itself.

  • rendering/RenderBox.cpp: (WebCore::RenderBox::repaintLayerRectsForImage):
16:15 Changeset [39846] by hyatt@apple.com

2009-01-12 David Hyatt <hyatt@apple.com>

Apply the same idea as the previous fix but for matched declarations.

Reviewed by Dan Bernstein

  • css/CSSStyleSelector.h:
16:15 Changeset [39845] by beidson@apple.com

WebCore:

2009-01-12 Brady Eidson <beidson@apple.com>

Reviewed by Darin Adler

<rdar://problem/6468274> - Track Non-get requests in global history

  • WebCore.base.exp:
  • history/HistoryItem.cpp: (WebCore::HistoryItem::HistoryItem):
  • history/HistoryItem.h: (WebCore::HistoryItem::lastVisitWasHTTPNonGet): (WebCore::HistoryItem::setLastVisitWasHTTPNonGet):

WebKit/mac:

2009-01-12 Brady Eidson <beidson@apple.com>

Reviewed by Darin Adler

<rdar://problem/6468274> - Track Non-get requests in global history

  • History/WebHistory.mm: (-[WebHistory _visitedURL:withTitle:method:wasFailure:]):
  • History/WebHistoryInternal.h:
  • History/WebHistoryItem.mm: (-[WebHistoryItem initFromDictionaryRepresentation:]): (-[WebHistoryItem dictionaryRepresentation]): (-[WebHistoryItem _lastVisitWasHTTPNonGet]):
  • History/WebHistoryItemPrivate.h:
  • WebCoreSupport/WebFrameLoaderClient.mm: (WebFrameLoaderClient::updateGlobalHistory): Only pass the method through if it was an HTTP load

WebKit/win:

2009-01-12 Brady Eidson <beidson@apple.com>

Reviewed by Darin Adler

<rdar://problem/6468274> - Track Non-get requests in global history

  • WebCoreSupport/WebFrameLoaderClient.cpp: (WebFrameLoaderClient::updateGlobalHistory):
  • WebHistory.cpp: (WebHistory::addItem):
  • WebHistory.h:
  • Interfaces/IWebHistoryItemPrivate.idl:
  • WebHistoryItem.cpp: (WebHistoryItem::initFromDictionaryRepresentation): (WebHistoryItem::dictionaryRepresentation): (WebHistoryItem::lastVisitWasHTTPNonGet): (WebHistoryItem::setLastVisitWasHTTPNonGet):
  • WebHistoryItem.h:
15:49 Changeset [39844] by kevino@webkit.org

wx build fix. JSValue* -> JSValuePtr.

15:47 Changeset [39843] by kevino@webkit.org

!ENABLE(SVG_FONTS) build fix. Move defaultUnitsPerEm into a non-SVG header so it can be used by all builds.

15:41 Changeset [39842] by kevino@webkit.org

wx build typo fix.

15:39 Changeset [39841] by kevino@webkit.org

Since platform date time functions have moved to JSC, now the wx port JSC needs to depend on wx.

15:18 Changeset [39840] by andersca@apple.com

2009-01-12 Anders Carlsson <andersca@apple.com>

Reviewed by Sam Weinig.

Move marshalling into NetscapePluginInstanceProxy.


Add support for marshallin strings.


  • Plugins/Hosted/NetscapePluginHostProxy.mm: (WKPCEvaluate): (WKPCInvoke):
  • Plugins/Hosted/NetscapePluginInstanceProxy.h:
  • Plugins/Hosted/NetscapePluginInstanceProxy.mm: (WebKit::NetscapePluginInstanceProxy::evaluate): (WebKit::NetscapePluginInstanceProxy::invoke): (WebKit::NetscapePluginInstanceProxy::marshalValue):
  • Plugins/Hosted/WebKitPluginHostTypes.h:
15:01 Changeset [39839] by hyatt@apple.com

2009-01-12 David Hyatt <hyatt@apple.com>

Make sure the set of matched rules has a reserved capacity of 32. This prevents remalloc thrash as
elements are matched, since the Vector is shrinking and growing over and over. 1% PLT speedup.

Reviewed by Oliver Hunt

  • css/CSSStyleSelector.h:
14:37 Changeset [39838] by hyatt@apple.com

2009-01-12 David Hyatt <hyatt@apple.com>

Optimize Element::attributeChanged to just check if accessibility is enabled up front before looking at the
two accessibility-related attributes. This avoids the rest of the code for all attributes on the PLT.
Tiny speedup.

Reviewed by Oliver Hunt

  • dom/Element.cpp: (WebCore::Element::attributeChanged):
14:34 Changeset [39837] by hyatt@apple.com

2009-01-12 David Hyatt <hyatt@apple.com>

Optimize the retrieval of line height. If the document is not using any first-line rules, just immediately
default to the normal case. Tiny speedup on PLT.

Reviewed by Oliver Hunt

  • rendering/RenderFlow.cpp: (WebCore::RenderFlow::lineHeight):
14:25 Changeset [39836] by andersca@apple.com

2009-01-12 Anders Carlsson <andersca@apple.com>

Reviewed by Sam Weinig.

Implement WKPCInvoke.


  • Plugins/Hosted/NetscapePluginHostProxy.mm: (WKPCEvaluate): (WKPCInvoke):
  • Plugins/Hosted/NetscapePluginInstanceProxy.h:
  • Plugins/Hosted/NetscapePluginInstanceProxy.mm: (WebKit::NetscapePluginInstanceProxy::idForObject): (WebKit::NetscapePluginInstanceProxy::invoke):
  • Plugins/Hosted/WebKitPluginClient.defs:
14:15 QtWebKitTodo edited by hausmann@webkit.org
(diff)
14:15 Changeset [39835] by hyatt@apple.com

2009-01-12 David Hyatt <hyatt@apple.com>

Fix 0.6% regression in PLT. Make sure not to waste time updating NSScrollView's can blit on scroll state
in the ScrollView constructor, since this is always going to be properly updated by FrameView anyway.

Reviewed by Oliver Hunt

  • platform/ScrollView.cpp: (WebCore::ScrollView::ScrollView):
13:42 Changeset [39834] by simon.fraser@apple.com

2009-01-12 Simon Fraser <simon.fraser@apple.com>

Reviewed by Dan Bernstein

https://bugs.webkit.org/show_bug.cgi?id=23220

Merge layoutDelta into LayoutState. This requires that we
push even when layoutState is disabled, so that layoutDelta
can still be stored.

13:38 Changeset [39833] by andersca@apple.com

2009-01-12 Anders Carlsson <andersca@apple.com>

Reviewed by Sam Weinig.

Move marshalling code to NetscapePluginInstanceProxy. Add support for marshalling JS objects.


  • Plugins/Hosted/NetscapePluginHostProxy.mm: (WKPCEvaluate):
  • Plugins/Hosted/NetscapePluginInstanceProxy.h:
  • Plugins/Hosted/NetscapePluginInstanceProxy.mm: (WebKit::NetscapePluginInstanceProxy::marshalValue):
  • Plugins/Hosted/WebKitPluginHostTypes.h:
13:31 Changeset [39832] by jchaffraix@webkit.org

2009-01-12 Julien Chaffraix <jchaffraix@pleyo.com>

Reviewed by Holger Freyther.

[Qt] Build fix after turning the fontCache to a singleton
https://bugs.webkit.org/show_bug.cgi?id=23253

Add a stub implementation for WebCore::fontCache() and FontCache::FontCache().

  • platform/graphics/qt/FontCacheQt.cpp: (WebCore::fontCache): (WebCore::FontCache::FontCache):
13:07 Changeset [39831] by dglazkov@chromium.org

2009-01-12 Eric Roman <eroman@chromium.org>

Reviewed by Darin Adler.


Fix some bugs with Selection::appendTrailingWhitespace().
https://bugs.webkit.org/show_bug.cgi?id=23232


Test: editing/selection/doubleclick-whitespace-crash.html


  • editing/Selection.cpp: (WebCore::makeSearchRange): (WebCore::Selection::appendTrailingWhitespace):

2009-01-12 Eric Roman <eroman@chromium.org>

Reviewed by Darin Adler.
https://bugs.webkit.org/show_bug.cgi?id=23232

  • editing/selection/doubleclick-whitespace-crash-expected.txt: Added.
  • editing/selection/doubleclick-whitespace-crash.html: Added.
  • editing/selection/doubleclick-whitespace-expected.txt:
  • editing/selection/doubleclick-whitespace.html:
12:26 Changeset [39830] by eric@webkit.org

2009-01-12 Dimitri Glazkov <dglazkov@chromium.org>

Reviewed by Eric Seidel.

Add the first pieces of platform/graphics/chromium
https://bugs.webkit.org/show_bug.cgi?id=23257

  • platform/graphics/chromium/ColorChromium.cpp: Added. (WebCore::focusRingColor):
  • platform/graphics/chromium/FontCacheChromiumWin.cpp: Added. (WebCore::FontCache::platformInit): (WebCore::isStringASCII): (WebCore::LookupAltName): (WebCore::FontCodepage::): (WebCore::FontCodepage::if): (WebCore::createFontIndirectAndGetWinName): (WebCore::fontContainsCharacter): (WebCore::FontCache::getFontDataForCharacters): (WebCore::FontCache::alternateFamilyName): (WebCore::FontCache::getSimilarFontPlatformData): (WebCore::FontCache::getLastResortFallbackFont): (WebCore::toGDIFontWeight): (WebCore::FontCache::getGenericFontForScript): (WebCore::FillLogFont): (WebCore::FontCache::fontExists): (WebCore::TraitsInFamilyProcData::TraitsInFamilyProcData): (WebCore::traitsInFamilyEnumProc): (WebCore::FontCache::getTraitsInFamily): (WebCore::FontCache::createFontPlatformData):
  • platform/graphics/chromium/FontCacheLinux.cpp: Added. (WebCore::FontCache::platformInit): (WebCore::FontCache::getFontDataForCharacters): (WebCore::FontCache::alternateFamilyName): (WebCore::FontCache::getSimilarFontPlatformData): (WebCore::FontCache::getLastResortFallbackFont): (WebCore::FontCache::getTraitsInFamily): (WebCore::FontCache::createFontPlatformData): (WebCore::FontCache::getGenericFontForScript):
  • platform/graphics/chromium/FontChromiumWin.cpp: Added. (WebCore::windowsCanHandleTextDrawing): (WebCore::skiaDrawText): (WebCore::PaintSkiaText): (WebCore::Font::drawGlyphs): (WebCore::Font::selectionRectForComplexText): (WebCore::Font::drawComplexText): (WebCore::Font::floatWidthForComplexText): (WebCore::Font::offsetForPositionForComplexText):
11:25 Changeset [39829] by darin@chromium.org

2009-01-12 Darin Fisher <darin@chromium.org>

Reviewed by Eric Seidel.

https://bugs.webkit.org/show_bug.cgi?id=23254
Upstream page/chromium/

  • page/chromium/AXObjectCacheChromium.cpp: Added. (WebCore::AXObjectCache::detachWrapper): (WebCore::AXObjectCache::attachWrapper): (WebCore::AXObjectCache::postNotification): (WebCore::AXObjectCache::postNotificationToElement): (WebCore::AXObjectCache::handleFocusedUIElementChanged):
  • page/chromium/AccessibilityObjectChromium.cpp:
  • page/chromium/AccessibilityObjectWrapper.h:
  • page/chromium/ChromeClientChromium.h: Added.
  • page/chromium/DragControllerChromium.cpp: Added. (WebCore::DragController::dragOperation): (WebCore::DragController::isCopyKeyDown): (WebCore::DragController::maxDragImageSize): (WebCore::DragController::cleanupAfterSystemDrag):
  • page/chromium/EventHandlerChromium.cpp: Added. (WebCore::EventHandler::passMousePressEventToSubframe): (WebCore::EventHandler::passMouseMoveEventToSubframe): (WebCore::EventHandler::passMouseReleaseEventToSubframe): (WebCore::EventHandler::passWheelEventToWidget): (WebCore::EventHandler::passWidgetMouseDownEventToWidget): (WebCore::EventHandler::passMouseDownEventToWidget): (WebCore::EventHandler::tabsToAllControls): (WebCore::EventHandler::eventActivatedView): (WebCore::EventHandler::createDraggingClipboard): (WebCore::EventHandler::focusDocumentView): (WebCore::EventHandler::accessKeyModifiers):
  • page/chromium/FrameChromium.cpp: Added. (WebCore::computePageRectsForFrame): (WebCore::Frame::dragImageForSelection):
  • page/chromium/FrameChromium.h: Added.
11:10 Changeset [39828] by dglazkov@chromium.org

2008-01-12 Dimitri Glazkov <dglazkov@chromium.org>

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=22936
Fix uninitialized memory read error, reported by Purify.

  • platform/graphics/SimpleFontData.cpp: (WebCore::SimpleFontData::SimpleFontData): Added initializer for

m_unitsPerEm using cDefaultUnitsPerEm constant.

  • svg/SVGFontFaceElement.cpp: (WebCore::SVGFontFaceElement::unitsPerEm): Replaced literal value with

the constant.

  • svg/SVGFontFaceElement.h: Added cDefaultUnitsPerEm constant.
07:50 Changeset [39827] by treat@webkit.org

Do not set fixedLayoutSize for anything other than the mainframe during
the transition as this is a page level state like the viewportSize.

05:03 Changeset [39826] by vestbo@webkit.org

2009-01-12 Tor Arne Vestbø <tavestbo@trolltech.com>

Reviewed by Simon Hausmann.

[Qt] Ensure that we're always notified of navigation actions for local anchors

FrameLoader::checkNavigationPolicy() does not notify us if the request has
been checked before (which happens for local anchors), but in the case of
our clients not accepting the navigation request we do want to be notified
again later on, so we clear the lastCheckedRequest flag.

04:51 Changeset [39825] by vestbo@webkit.org

2009-01-12 Ariya Hidayat <ariya.hidayat@trolltech.com>

Rubber-stamped by Tor Arne Vestbø.

Build fix: use qobject_cast which does not rely on RTTI. This fixes the
unit test when building it with a version of Qt (e.g. Qt/Embedded)
configured without RTTI.

  • tests/qwebframe/tst_qwebframe.cpp:
03:10 Changeset [39824] by ap@webkit.org

Update test results.

  • fast/dom/Window/window-properties-expected.txt: Remove appcache dynamic entries methods.
02:46 Changeset [39823] by jchaffraix@webkit.org

WebCore:

2009-01-12 Julien Chaffraix <jchaffraix@pleyo.com>

Reviewed by Darin Adler.

Bug 22861: Turn the FontCache into a singleton
https://bugs.webkit.org/show_bug.cgi?id=22861

  • Added fontCache() to get the global FontCache.
  • Made all the methods in FontCache instance method.
  • Changed FontCache:: to fontCache()-> in WebCore.
  • WebCore.base.exp: Now export fontCache().
  • css/CSSFontFaceSource.cpp: (WebCore::CSSFontFaceSource::getFontData):
  • css/CSSFontSelector.cpp: (WebCore::CSSFontSelector::CSSFontSelector): (WebCore::CSSFontSelector::~CSSFontSelector): (WebCore::CSSFontSelector::addFontFaceRule): (WebCore::fontDataForGenericFamily):
  • platform/graphics/Font.cpp: (WebCore::Font::glyphDataForCharacter):
  • platform/graphics/FontCache.cpp: (WebCore::fontCache): (WebCore::FontCache::FontCache):
  • platform/graphics/FontCache.h:
  • platform/graphics/FontFallbackList.cpp: (WebCore::FontFallbackList::FontFallbackList): (WebCore::FontFallbackList::invalidate): (WebCore::FontFallbackList::releaseFontData): (WebCore::FontFallbackList::fontDataAt): (WebCore::FontFallbackList::fontDataForCharacters): (WebCore::FontFallbackList::setPlatformFont):
  • platform/graphics/SimpleFontData.cpp: (WebCore::SimpleFontData::~SimpleFontData):
  • platform/graphics/mac/FontCacheMac.mm: (WebCore::fontCacheATSNotificationCallback):
  • platform/graphics/mac/SimpleFontDataMac.mm: (WebCore::SimpleFontData::smallCapsFontData):
  • platform/graphics/win/FontCacheWin.cpp: (WebCore::getCJKCodePageMasks):
  • platform/graphics/win/SimpleFontDataWin.cpp: (WebCore::SimpleFontData::containsCharacters):

WebKit/mac:

2009-01-12 Julien Chaffraix <jchaffraix@pleyo.com>

Reviewed by Darin Adler.

Bug 22861: Turn the FontCache into a singleton
https://bugs.webkit.org/show_bug.cgi?id=22861

  • Misc/WebCoreStatistics.mm: (+[WebCoreStatistics cachedFontDataCount]): (+[WebCoreStatistics cachedFontDataInactiveCount]): (+[WebCoreStatistics purgeInactiveFontData]): Redirected all the static calls to the global FontCache instance.

WebKit/win:

2009-01-12 Julien Chaffraix <jchaffraix@pleyo.com>

Reviewed by Darin Adler.

Bug 22861: Turn the FontCache into a singleton
https://bugs.webkit.org/show_bug.cgi?id=22861

  • WebCoreStatistics.cpp: (WebCoreStatistics::cachedFontDataCount): (WebCoreStatistics::cachedFontDataInactiveCount): (WebCoreStatistics::purgeInactiveFontData): Redirected all the static calls to the global FontCache instance.
01:58 QtWebKitJournal edited by hausmann@webkit.org
(diff)
01:41 Changeset [39822] by hausmann@webkit.org

2009-01-12 Simon Hausmann <simon.hausmann@nokia.com>

Rubber-stamped by Tor Arne Vestbø.

When the QWebView is disabled context menus implemented in JavaScript should not pop up.

Just like in QWidget::event() we have to check for the disabled state and
ignore context menu events.

01:36 Changeset [39821] by hausmann@webkit.org

2009-01-12 Simon Hausmann <simon.hausmann@nokia.com

Reviewed by Tor Arne Vestbø.

Fix QWebView appearance when showed uninitialized without page.

Only set WA_OpaquePaintEvent if we have a page set that will actually
paint the entire contents, otherwise the optimization does not apply
as we do not have a paintEvent() implementation and Qt has to fill the
view with its default background instead.

01:12 Changeset [39820] by vestbo@webkit.org

2009-01-11 Ariya Hidayat <ariya.hidayat@trolltech.com>

Rubber-stamped by Tor Arne Vestbø.

Do not create a QVariant from QObject* directly, use the
template-specialized function.

  • tests/qwebframe/tst_qwebframe.cpp: (MyQObject::myOverloadedSlot):

01/11/09:

23:48 Changeset [39819] by darin@apple.com

WebCore:

2009-01-11 Darin Adler <darin@apple.com>

Reviewed by Dan Bernstein.

Bug 23247: createSVGPathSegCurvetoCubicSmoothAbs doesn't set y2 correctly
https://bugs.webkit.org/show_bug.cgi?id=23247

Also fix problems where partially parsed paths will still have an effect
rather than failing parsing and being ignored.

Test: dom/path-segments.html

  • platform/graphics/Path.h: (WebCore::Path::swap): Added.
  • svg/SVGParserUtilities.cpp: (WebCore::SVGPathParser::parseSVG): Removed unneeded special case for empty string. (WebCore::PathBuilder::build): Changed to not modify the path if parsing fails. (WebCore::SVGPathSegListBuilder::build): Changed to not modify the segment list if parsing fails. (WebCore::SVGPathSegListBuilder::svgMoveTo): Put the path element into the vector, not directly into the segment list. (WebCore::SVGPathSegListBuilder::svgLineTo): Ditto. (WebCore::SVGPathSegListBuilder::svgLineToHorizontal): Ditto. (WebCore::SVGPathSegListBuilder::svgLineToVertical): Ditto. (WebCore::SVGPathSegListBuilder::svgCurveToCubic): Ditto. (WebCore::SVGPathSegListBuilder::svgCurveToCubicSmooth): Ditto. (WebCore::SVGPathSegListBuilder::svgCurveToQuadratic): Ditto. (WebCore::SVGPathSegListBuilder::svgCurveToQuadraticSmooth): Ditto. (WebCore::SVGPathSegListBuilder::svgArcTo): Ditto. (WebCore::SVGPathSegListBuilder::svgClosePath): Ditto.
  • svg/SVGPathSegCurvetoCubicSmooth.h: (WebCore::SVGPathSegCurvetoCubicSmooth::SVGPathSegCurvetoCubicSmooth): Set y2 to y2, not to y.

WebKitTools:

2009-01-11 Darin Adler <darin@apple.com>

  • Scripts/make-js-test-wrappers: Updated so it won't overwrite tests in the svg/dom directory that aren't using standard wrappers.

LayoutTests:

2009-01-11 Darin Adler <darin@apple.com>

Reviewed by Dan Bernstein.

Bug 23247: createSVGPathSegCurvetoCubicSmoothAbs doesn't set y2 correctly
https://bugs.webkit.org/show_bug.cgi?id=23247

Also fix problems where partially parsed paths will still have an effect
rather than failing parsing and being ignored.

Renamed the path-parser.html test to fuzz-path-parser.html, since it's a fuzz test,
to make room for a non-fuzz test of the path parser.

  • platform/mac/svg/custom/path-bad-data-expected.txt: Updated results, since the old ones were based on incorrect path parsing behavior.
  • svg/dom/fuzz-path-parser-expected.txt: Copied from svg/dom/path-parser-expected.txt. Also updated because the parser no longer tolerates extra characters at the end of the string, which it was doing by accident.
  • svg/dom/fuzz-path-parser.html: Copied from svg/dom/path-parser.html.
  • svg/dom/path-parser-expected.txt: Replaced fuzz test results with new test results.
  • svg/dom/path-parser.html: Replaced fuzz test with a new test.
  • svg/dom/path-segments-expected.txt: Added.
  • svg/dom/path-segments.html: Added.
  • svg/dom/resources/TEMPLATE.html: Copied from fast/js/resources/TEMPLATE.html.
  • svg/dom/resources/path-parser.js: Added. Mostly, this test shows that we are processing the path while parsing it, which is a bug. This test would be much more useful if that was fixed. But it also shows that we were incorrectly tolerating extra characters after the path, which is now fixed.
  • svg/dom/resources/path-segments.js: Added.
23:44 Changeset [39818] by darin@apple.com

2009-01-11 Darin Adler <darin@apple.com>

Reviewed by Oliver Hunt.

Bug 23102: turn on unused parameter warnings in WebCore
https://bugs.webkit.org/show_bug.cgi?id=23102

Fourth step: Remove arguments entirely in the many cases where they are not
needed at all. This step is a little less mechanical than the earlier ones.

  • dom/DOMImplementation.cpp: (WebCore::DOMImplementation::isXMLMIMEType): Pass TextCaseSensitive when creating the regular expression, since we already specify a-zA-Z in our expression.
  • dom/Document.cpp: (WebCore::Document::databaseThread): Don't pass in the document when creating a DatabaseThread.
  • dom/Element.cpp: (WebCore::Element::createRareData): Don't pass the element when creating an ElementRareData. (WebCore::Element::detach): Don't pass the element when calling resetComputedStyle. (WebCore::Element::recalcStyle): Ditto.
  • dom/ElementRareData.h: Remove the unused Element* from the constructor and the resetComputedStyle function.
  • dom/EventTargetNode.cpp: (WebCore::EventTargetNode::dispatchEvent): Don't pass ec to dispatchGenericEvent. (WebCore::EventTargetNode::dispatchGenericEvent): Remove the unused ec argument. (WebCore::EventTargetNode::dispatchWindowEvent): Don't pass ec to dispatchGenericEvent.
  • dom/EventTargetNode.h: Remove the ExceptionCode& argument form disaptchGenericEvent.
  • editing/ApplyStyleCommand.cpp: Remove the empty removeBlockStyle function. (WebCore::ApplyStyleCommand::pushDownTextDecorationStyleAroundNode): Remove the unused start and end arguments. (WebCore::ApplyStyleCommand::pushDownTextDecorationStyleAtBoundaries): Don't pass start and end to pushDownTextDecorationStyleAroundNode.
  • editing/ApplyStyleCommand.h: Remove the empty removeBlockStyle function. Removed the unused start and end arguments from pushDownTextDecorationStyleAroundNode.
  • editing/htmlediting.cpp: Removed include of unused RegularExpression.h header.
  • html/HTMLTokenizer.cpp: (WebCore::Token::addAttribute): Removed unused Document* argument. (WebCore::HTMLTokenizer::parseTag): Don't pass document to addAttribute.
  • html/HTMLTokenizer.h: Removed unused Document* argument to to addAttribute.
  • page/DOMWindow.cpp: (WebCore::DOMWindow::localStorage): Don't pass the frame to storageArea.
  • page/DragController.cpp: (WebCore::DragController::dragIsMove): Removed unused DragData argument. (WebCore::DragController::performDrag): Call concludeEditDrag rather than concludeDrag. (WebCore::DragController::tryDocumentDrag): Don't pass dragData to dragIsMove. (WebCore::DragController::concludeEditDrag): Renamed to make it clear this should only be called for edit drags. Removed the action mask argument. Also changed dragIsMove call sites to not pass dragData.
  • page/DragController.h: Updated for above changes and tweaked comment format.
  • page/Frame.cpp: (WebCore::createRegExpForLabels): Pass TextCaseSensitive and TextCaseInsensitive when creating regular expressions for clarity. Call match instead of search, since both do the same thing except for with negative start offsets, and none used here are negative. (WebCore::Frame::matchLabelsAgainstElement): Ditto.
  • page/Page.h: Moved TextCaseSensitivity from here to StringImpl.h in platform.
  • page/mac/AccessibilityObjectWrapper.mm: (AXAttributeStringSetHeadingLevel): Removed unused object argument. (AXAttributedStringAppendText): Ditto. (-[AccessibilityObjectWrapper doAXAttributedStringForTextMarkerRange:]): Ditto.
  • page/mac/FrameMac.mm: (WebCore::regExpForLabels): Pass TextCaseSensitive and TextCaseInsensitive when creating regular expressions for clarity. Call match instead of search, since both do the same thing except for with negative start offsets, and none used here are negative. (WebCore::Frame::matchLabelsAgainstElement): Ditto.
  • platform/graphics/SegmentedFontData.cpp: (WebCore::SegmentedFontData::containsCharacter): Added. (WebCore::SegmentedFontData::containsCharacters): Updated to check all characters instead of just the first one, using containsCharacter.
  • platform/graphics/SegmentedFontData.h: Made more members private. Added containsCharacter function member.
  • platform/mac/DragImageMac.mm: (WebCore::dragImageSize): Changed to use RetainPtr<NSImage> directly instead of DragImageRef for clarity, since this is platform-specific code. (WebCore::deleteDragImage): Ditto. Also remove argument name to get rid of unused variable warning. (WebCore::scaleDragImage): Ditto. (WebCore::dissolveDragImageToFraction): Ditto. (WebCore::createDragImageFromImage): Ditto. (WebCore::createDragImageIconForCachedImage): Ditto.
  • platform/mac/WebFontCache.mm: (acceptableChoice): Removed the unused weight-related arguments. (betterChoice): Ditto. (+[WebFontCache internalFontWithFamily:traits:weight:size:]): Ditto.
  • platform/posix/FileSystemPOSIX.cpp: Removed the implementation of listDirectory because we'd rather get a link error rather than a runtime error if we use it for anything, and right now it's not used.
  • platform/text/RegularExpression.cpp: Removed unused private data, including pattern, lastMatchString, lastMatchOffsets, lastMatchCount, and lastMatchPos. Made the RegularExpression::Private::regex be private and use an accessor function to get at it. (WebCore::RegularExpression::RegularExpression): Removed the const char* version of the constructor because it's not any more efficient. Changed the boolean constructor argument to use TextCaseSensitivity instead to make it less likely we'll use it wrong. (WebCore::RegularExpression::operator=): Since the only member is a single RefPtr, just use plain old assignment on it. The only reason we do this instead of using the compiler generated assignment operator is that we want the Private structure to be internal to the .cpp file. (WebCore::RegularExpression::match): Added a null check so we won't crash if we fail to compile the regular expression. Removed the code to handle multiple match strings, since that was unused.
  • platform/text/RegularExpression.h: Changed the case sensitivity argument to use TextCaseSensitivity and not have a default value. Removed the default constructor, and the pattern, match, and pos functions.
  • platform/text/StringImpl.h: Moved the TextCaseSensitivity enum here from Page.h, because we should eventually use it in all the places where we have an argument to control case sensitivity, starting with the functions in this class.
  • platform/text/mac/ShapeArabic.c: (shapeUnicode): Removed the unused options and pErrorCode arguments. Even though this is code borrowed from ICU, this seems like a simple safe change to make. (shapeArabic): Don't pass options or pErrorCode in to shapeUnicode.
  • rendering/RenderPath.cpp: (WebCore::RenderPath::RenderPath): Removed the unused style argument. Also removed the type check. The argument already has the correct type, so it's not helpful to upcast it and then check its type again, unless we have some reason to believe the type is not reliable. And we don't.
  • rendering/RenderPath.h: Removed the RenderStyle* argument to the constructor. Also removed the explicit declaration of the destructor. It's not helpful or needed.
  • rendering/RenderSVGInlineText.cpp: (WebCore::RenderSVGInlineText::localCaretRect): Updated the comment here to make it clear that just returning an empty rect is not a good implementation. Removed the argument names to get rid of the warning.
  • rendering/RenderTreeAsText.cpp: (WebCore::printBorderStyle): Removed the unused RenderObject argument. (WebCore::operator<<): Updated for the change to printBorderStyle function.
  • storage/DatabaseThread.cpp: (WebCore::DatabaseThread::DatabaseThread): Removed the unused document argument.
  • storage/DatabaseThread.h: Ditto.
  • storage/LocalStorage.cpp: (WebCore::LocalStorage::storageArea): Removed the unused Frame* argument.
  • storage/LocalStorage.h: Ditto.
  • svg/SVGElement.cpp: (WebCore::SVGElement::sendSVGLoadEventIfPossible): Don't pass an exception code location to dispatchGenericEvent.
  • svg/SVGElementInstance.cpp: (WebCore::SVGElementInstance::dispatchEvent): Ditto.
  • svg/SVGStyledTransformableElement.cpp: Removed include of unused RegularExpression.h. (WebCore::SVGStyledTransformableElement::createRenderer): Don't pass style in to the RenderPath constructor.
  • svg/SVGTransformable.cpp: Removed include of unused RegularExpression.h.
  • xml/XSLTProcessor.cpp: (WebCore::createFragmentFromSource): Removed unused sourceNode argument. (WebCore::XSLTProcessor::transformToFragment): Don't pass sourceNode to createFragmentFromSource.
23:34 Changeset [39817] by ap@webkit.org

2009-01-11 David Levin <levin@chromium.org>

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=23245

Add initializeThreading to key places in JS API to ensure that
UString is properly initialized.

  • API/JSContextRef.cpp: (JSContextGroupCreate): (JSGlobalContextCreate):
  • API/JSObjectRef.cpp: (JSClassCreate):
  • API/JSStringRef.cpp: (JSStringCreateWithCharacters): (JSStringCreateWithUTF8CString):
  • API/JSStringRefCF.cpp: (JSStringCreateWithCFString):
23:21 Changeset [39816] by ap@webkit.org

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=23165
Disable application cache dynamic entries support.

19:19 Changeset [39815] by darin@apple.com

2009-01-11 David Levin <levin@chromium.org>

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=23175

Separate out BaseString information from UString::Rep and make all baseString access go through
a member function, so that it may be used for something else (in the future) in the BaseString
case.

  • runtime/SmallStrings.cpp: (JSC::SmallStringsStorage::rep): (JSC::SmallStringsStorage::SmallStringsStorage): (JSC::SmallStrings::SmallStrings): (JSC::SmallStrings::mark): Adjust to account for the changes in UString and put the UString in place in SmallStringsStorage to aid in locality of reference among the UChar[] and UString::Rep's.
  • runtime/SmallStrings.h:
  • runtime/UString.cpp: (JSC::initializeStaticBaseString): (JSC::initializeUString): (JSC::UString::Rep::create): (JSC::UString::Rep::destroy): (JSC::UString::Rep::checkConsistency): (JSC::expandCapacity): (JSC::UString::expandPreCapacity): (JSC::concatenate): (JSC::UString::append): (JSC::UString::operator=):
  • runtime/UString.h: (JSC::UString::Rep::baseIsSelf): (JSC::UString::Rep::setBaseString): (JSC::UString::Rep::baseString): (JSC::UString::Rep::): (JSC::UString::Rep::null): (JSC::UString::Rep::empty): (JSC::UString::Rep::data): (JSC::UString::cost): Separate out the items out used by base strings from those used in Rep's that only point to base strings. (This potentially saves 24 bytes per Rep.)
19:17 Changeset [39814] by darin@apple.com

2009-01-11 Darin Adler <darin@apple.com>

  • fast/frames/frame-deep-nested-resize-expected.txt: This time for sure.
18:07 Changeset [39813] by darin@apple.com

2009-01-11 Darin Adler <darin@apple.com>

Fix failure of new test seen on buildbots.

  • fast/frames/frame-deep-nested-resize-expected.txt: Updated newlines to match what DumpRenderTree generates on Mac. I believe Adam generated his results under Qt DumpRenderTree. If so, then the Qt version will need to change to match the Mac version, since all the expected results were generated on Mac.
16:54 Changeset [39812] by treat@webkit.org

Correct coordinate transformation for deeply nested frames on resizing. This
fixes https://bugs.webkit.org/show_bug.cgi?id=20766.

15:55 Changeset [39811] by mrowe@apple.com

Tiger build fix. The connection parameter is still used on Tiger in this method.

  • platform/network/mac/ResourceHandleMac.mm:

(-[WebCoreResourceHandleAsDelegate connection:willCacheResponse:]):

14:01 Changeset [39810] by darin@apple.com

2009-01-11 Darin Adler <darin@apple.com>

Reviewed by Dan Bernstein.

Bug 23102: turn on unused parameter warnings in WebCore
https://bugs.webkit.org/show_bug.cgi?id=23102

Third step: Deal with cases of arguments used only in assertions.

  • dom/MessagePort.cpp: (WebCore::MessagePortCloseEventTask::performTask): Use ASSERT_UNUSED.
  • dom/Worker.cpp: (WebCore::Worker::notifyFinished): Ditto.
  • dom/XMLTokenizer.cpp: (WebCore::XMLTokenizer::notifyFinished): Ditto.
  • loader/appcache/ApplicationCacheGroup.cpp: (WebCore::ApplicationCacheGroup::failedLoadingMainResource): Ditto.
  • loader/appcache/ApplicationCacheStorage.cpp: (WebCore::ApplicationCacheStorage::verifySchemaVersion): Ditto.
  • loader/icon/IconLoader.cpp: (WebCore::IconLoader::didReceiveData): Since loader and size are only used when logging, use UNUSED_PARAM when logging is disabled.
  • platform/mac/WidgetMac.mm: (WebCore::Widget::beforeMouseDown): Use ASSERT_UNUSED.
  • platform/network/mac/ResourceHandleMac.mm: (WebCore::ResourceHandle::didCancelAuthenticationChallenge): Use the argument since it's probably a tiny bit more efficient and gets rid of the unused argument warning.
  • rendering/RenderListMarker.cpp: (WebCore::RenderListMarker::createInlineBox): Use ASSERT_UNUSED.
  • rendering/RenderObject.cpp: (WebCore::RenderObject::createInlineBox): Ditto.
  • rendering/RenderSVGInline.cpp: (WebCore::RenderSVGInline::createInlineBox): Use UNUSED_PARAM. I couldn't use ASSERT_UNUSED because it's a single assertion, but it's the only use for two different arguments.
  • rendering/RenderText.cpp: (WebCore::RenderText::createInlineBox): Use ASSERT_UNUSED.
  • rendering/RenderThemeMac.mm: (-[WebCoreRenderThemeNotificationObserver systemColorsDidChange:]): Ditto.
13:58 Changeset [39809] by darin@apple.com

2009-01-11 Darin Adler <darin@apple.com>

Reviewed by Dan Bernstein.

Bug 23239: improve handling of unused arguments in JavaScriptCore
https://bugs.webkit.org/show_bug.cgi?id=23239

  • runtime/DatePrototype.cpp: Moved LocaleDateTimeFormat enum outside #if so we can use this on all platforms. Changed valueOf to share the same function with getTime, since the contents of the two are identical. Removed a FIXME since the idea isn't really specific enough or helpful enough to need to sit here in the source code. (JSC::formatLocaleDate): Changed the Mac version of this function to take the same arguments as the non-Mac version so the caller doesn't have to special-case the two platforms. Also made the formatString array be const; before the characters were, but the array was a modifiable global variable. (JSC::dateProtoFuncToLocaleString): Changed to call the new unified version of formatLocaleDate and remove the ifdef. (JSC::dateProtoFuncToLocaleDateString): Ditto. (JSC::dateProtoFuncToLocaleTimeString): Ditto.
  • runtime/JSNotAnObject.cpp: (JSC::JSNotAnObject::toObject): Use the new ASSERT_UNUSED instead of the old UNUSED_PARAM.
  • runtime/RegExp.cpp: (JSC::RegExp::RegExp): Changed to only use UNUSED_PARAM when the parameter is actually unused.
  • wtf/TCSystemAlloc.cpp: (TCMalloc_SystemRelease): Changed to only use UNUSED_PARAM when the parameter is actually unused. (TCMalloc_SystemCommit): Changed to omit the argument names instead of using UNUSED_PARAM.
13:42 Changeset [39808] by mitz@apple.com

Set the svn:mime-type property of recently-added PNG files to image/png.

13:22 Changeset [39807] by simon.fraser@apple.com

2009-01-11 Simon Fraser <simon.fraser@apple.com>

Reviewed by Oliver Hunt

https://bugs.webkit.org/show_bug.cgi?id=23242

Fix CanvasRenderingContext2D::transform to do a pre-multiply,
rather than a post-multiply into m_transform. This bug did not affect
drawing, but did cause m_transform to be incorrect, which impacted
willDraw(), and isPointInPath.

Test: fast/canvas/canvas-incremental-repaint-2.html

  • html/CanvasRenderingContext2D.cpp: (WebCore::CanvasRenderingContext2D::transform):
12:05 Changeset [39806] by darin@apple.com

2009-01-11 Darin Adler <darin@apple.com>

Reviewed by Dan Bernstein.

Bug 23102: turn on unused parameter warnings in WebCore
https://bugs.webkit.org/show_bug.cgi?id=23102

Second step: Fix simple Objective-C cases where we need to use UNUSED_PARAM.
That's because you can't omit an argument name in Objective-C methods.

  • bindings/objc/DOMObject.mm: (-[DOMObject copyWithZone:]): Use UNUSED_PARAM. Also rename argument to include the word "unused" so we don't just start using it without removing the macro.
  • bindings/objc/WebScriptObject.mm: (+[WebUndefined allocWithZone:]): Ditto. (-[WebUndefined initWithCoder:]): Ditto. (-[WebUndefined encodeWithCoder:]): Ditto. (-[WebUndefined copyWithZone:]): Ditto.
  • page/mac/WebDashboardRegion.m: (-[WebDashboardRegion copyWithZone:]): Ditto.
  • platform/graphics/mac/MediaPlayerPrivateQTKit.mm: (WebCore::mainThreadSetNeedsDisplay): Omit argument name. (-[WebCoreMovieObserver loadStateChanged:]): Use UNUSED_PARAM as above. (-[WebCoreMovieObserver rateChanged:]): Ditto. (-[WebCoreMovieObserver sizeChanged:]): Ditto. (-[WebCoreMovieObserver timeChanged:]): Ditto. (-[WebCoreMovieObserver didEnd:]): Ditto. (-[WebCoreMovieObserver newImageAvailable:]): Ditto.
  • platform/mac/ScrollbarThemeMac.mm: (+[ScrollbarPrefsObserver appearancePrefsChanged:]): Ditto. (+[ScrollbarPrefsObserver behaviorPrefsChanged:]): Ditto.
  • platform/mac/SharedTimerMac.mm: (-[WebCorePowerNotifier didWake:]): Ditto.
  • platform/mac/WebCoreKeyGenerator.m: (-[WebCoreKeyGenerator signedPublicKeyAndChallengeStringWithStrengthIndex:challenge:pageURL:]): Ditto.
  • platform/network/mac/ResourceHandleMac.mm: (-[WebCoreResourceHandleAsDelegate connection:willSendRequest:redirectResponse:]): Ditto. (-[WebCoreResourceHandleAsDelegate connectionShouldUseCredentialStorage:]): Ditto. (-[WebCoreResourceHandleAsDelegate connection:didReceiveAuthenticationChallenge:]): Ditto. (-[WebCoreResourceHandleAsDelegate connection:didCancelAuthenticationChallenge:]): Ditto. (-[WebCoreResourceHandleAsDelegate connection:didReceiveResponse:]): Ditto. (-[WebCoreResourceHandleAsDelegate connection:didReceiveData:lengthReceived:]): Ditto. (-[WebCoreResourceHandleAsDelegate connection:willStopBufferingData:]): Ditto. (-[WebCoreResourceHandleAsDelegate connection:didSendBodyData:totalBytesWritten:totalBytesExpectedToWrite:]): Ditto. (-[WebCoreResourceHandleAsDelegate connectionDidFinishLoading:]): Ditto. (-[WebCoreResourceHandleAsDelegate connection:didFailWithError:]): Ditto. (-[WebCoreResourceHandleAsDelegate connection:willCacheResponse:]): Ditto. (-[WebCoreSynchronousLoader connection:willSendRequest:redirectResponse:]): Ditto. (-[WebCoreSynchronousLoader connection:didReceiveAuthenticationChallenge:]): Ditto. (-[WebCoreSynchronousLoader connection:didReceiveResponse:]): Ditto. (-[WebCoreSynchronousLoader connection:didReceiveData:]): Ditto. (-[WebCoreSynchronousLoader connectionDidFinishLoading:]): Ditto. (-[WebCoreSynchronousLoader connection:didFailWithError:]): Ditto.
11:59 Changeset [39805] by krit@webkit.org

2009-01-11 Dirk Schulze <krit@webkit.org>

Reviewed by Darin Adler.

Move platform dependent strokeContains back to RenderPath, Path and delete
unneeded Code in svg.

RenderPath clean-up for strokeContains
https://bugs.webkit.org/show_bug.cgi?id=22957

  • GNUmakefile.am:
  • WebCore.pro:
  • WebCore.vcproj/WebCore.vcproj:
  • WebCore.xcodeproj/project.pbxproj:
  • platform/graphics/Path.h:
  • platform/graphics/cairo/PathCairo.cpp: (WebCore::Path::strokeContains):
  • platform/graphics/cg/PathCG.cpp: (WebCore::Path::strokeContains):
  • platform/graphics/qt/PathQt.cpp: (WebCore::Path::strokeContains):
  • rendering/RenderPath.cpp: (WebCore::RenderPath::strokeContains):
  • svg/graphics/cairo/RenderPathCairo.cpp: Removed.
  • svg/graphics/cg/CgSupport.cpp: Removed.
  • svg/graphics/cg/CgSupport.h: Removed.
  • svg/graphics/cg/RenderPathCg.cpp: Removed.
  • svg/graphics/cg/SVGPaintServerPatternCg.cpp:
  • svg/graphics/cg/SVGResourceMaskerCg.mm:
  • svg/graphics/qt/RenderPathQt.cpp: Removed.
11:59 Changeset [39804] by zecke@webkit.org

2009-01-11 Xan Lopez <xan@gnome.org>

Reviewed by Holger Freyther.

Bump version to 1.1.0 since we are breaking ABI and adding new
major features.

  • configure.ac:
11:57 Changeset [39803] by zecke@webkit.org

2009-01-11 Xan Lopez <xan@gnome.org>

Reviewed by Holger Freyther.

Use NULL instead of 0 when dealing with pointers, as agreed for
the coding style of the WebKit GTK port for its GTK+ specific
files.

  • webkit/webkitwebview.cpp:
11:55 Changeset [39802] by zecke@webkit.org

2009-01-11 Xan Lopez <xan@gnome.org>

Reviewed by Holger Freyther.

Add padding to the class structs to avoid breaking ABI each time
we add stuff there.

  • webkit/webkitnetworkrequest.h:
  • webkit/webkitwebbackforwardlist.h:
  • webkit/webkitwebhistoryitem.h:
  • webkit/webkitwebnavigationaction.h:
  • webkit/webkitwebpolicydecision.h:
  • webkit/webkitwebview.h:
11:47 Changeset [39801] by zecke@webkit.org

[GTK] make distcheck complains that distclean is not deleting JSWorkerContextBase.lut.h

make distcheck pass by adding JSWorkerContextBase.lut.h to the
CLEANFILES that will be cleaned on a make distclean.

11:47 Changeset [39800] by oliver@apple.com

build fix

11:12 Changeset [39799] by oliver@apple.com

Bug 23128: get/put_by_val need to respecialise in the face of ByteArray

Reviewed by Darin Adler and Anders Carlsson

Restructure the code slightly, and add comments per Darin's suggestions

10:56 Changeset [39798] by oliver@apple.com

Fix exception check for string indexing

Reviewed by Anders Carlsson.

Whoops, I accidentally removed an exception check from fast the
fast path for string indexing when i originally landed the
byte array logic.

08:54 Changeset [39797] by oliver@apple.com

attempt to fix win build

08:48 Changeset [39796] by oliver@apple.com

Bug 23128: get/put_by_val need to respecialise in the face of ByteArray
<https://bugs.webkit.org/show_bug.cgi?id=23128>

Reviewed by Anders Carlsson.

Fairly simple patch, add specialised versions of cti_op_get/put_by_val
that assume ByteArray, thus avoiding a few branches in the case of bytearray
manipulation.

No effect on SunSpider. 15% win on the original testcase.

03:06 Changeset [39795] by ap@webkit.org

2009-01-11 Zalan Bujtas <zbujtas@gmail.com>

Reviewed by Alexey Proskuryakov.

https://bugs.webkit.org/show_bug.cgi?id=23235
Build fix for non-VIDEO builds.

  • rendering/RenderTheme.cpp:
  • rendering/RenderTheme.h:
03:03 Changeset [39794] by ap@webkit.org

Try to fix Windows build.

  • wtf/CurrentTime.cpp: Added a definition of msPerSecond (previously, this code was in DateMath.cpp, with constant definition in DateTime.h)
02:36 Changeset [39793] by ap@webkit.org

2009-01-11 Alexey Proskuryakov <ap@webkit.org>

Try to fix Windows build.

  • wtf/CurrentTime.cpp: Include <sys/types.h> and <sys/timeb.h>, as MSDN says to.
02:27 Changeset [39792] by ap@webkit.org

2009-01-11 Gabriella Toth <gtoth@inf.u-szeged.hu>

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=19287
return value of malloc() is not checked in npruntime.cpp

Checking whether malloc can allocate memory or not. If it can't, CRASH macro is invoked
(like in fastMalloc).


  • bridge/npruntime.cpp: (_NPN_GetStringIdentifier): (_NPN_GetIntIdentifier): (NPN_InitializeVariantWithStringCopy): (_NPN_CreateObject):
02:24 Changeset [39791] by ap@webkit.org

2009-01-11 Glenn Wilson <gwilson@google.com>

Reviewed by Darin Adler.

Changes radio buttons so they can be 'checked' even if they
don't have the name attribute set.

See https://bugs.webkit.org/show_bug.cgi?id=21534

Test: fast/html/select-unnamed-radio.html

  • html/HTMLInputElement.cpp: (WebCore::HTMLInputElement::preDispatchEventHandler):
01:48 Changeset [39790] by ap@webkit.org

2009-01-11 tony chang <tony@chromium.org>

Reviewed by Darin Adler.

Remove svn:eol-style native from some utf16 files so
windows-svn doesn't alter them on checkout.

  • fast/xsl/xslt-enc16.xml: remove svn:eol-style
  • fast/xsl/xslt-enc16to16.xml: remove svn:eol-style
01:44 Changeset [39789] by ap@webkit.org

Fix ChangeLog date.

01:43 Changeset [39788] by ap@webkit.org

2009-01-11 Robert Blaut <webkit@blaut.biz>

Reviewed by Eric Seidel.


<https://bugs.webkit.org/show_bug.cgi?id=23134>
Update bisect-builds for Safari 3.2 to prevent crashes

  • Scripts/bisect-builds: Added Safari 3.2 and the corresponding minimal revision, r37348.
01:40 Changeset [39787] by ap@webkit.org

2009-01-11 Jungshik Shin <jshin@chromium.org>

Reviewed by Alexey Proskuryakov

  • Make fast/encoding/char-encoding test work with Chromium by replacing 'data:' with 'file:' pointing to resources/dummy.html (an empty file).
  • Put old Mac encodings in a separate char-encoding-mac.html to make it possible to to skip it.
  • Factor out common functions in char-encoding.html and char-encoding-mac.html and put them in char-encoding-utils.js

See: https://bugs.webkit.org/show_bug.cgi?id=22882

  • fast/encoding/char-encoding-mac-expected.txt: Added.
  • fast/encoding/char-encoding-mac.html: Added.
  • fast/encoding/char-encoding.html:
  • fast/encoding/resources/char-encoding-utils.js: Added.
  • fast/encoding/resources/dummy.html: Added.
01:36 Changeset [39786] by ap@webkit.org

2008-12-18 Bernhard Rosenkraenzer <bero@arklinux.ch>

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=22205
Fix compatibility with bison 2.4, partially based on older patch by
Priit Laes <amd@store20.com>

  • WebCore/css/CSSGrammar.y: Made compatible with bison 2.4
01:18 Changeset [39785] by ap@webkit.org

2009-01-11 Jeremy Moskovich <jeremy@chromium.org>

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=16001
Incorrect assumption that selection is always contiguous LTR.

Test: editing/selection/extend-selection-bidi.html

  • rendering/RootInlineBox.cpp: (WebCore::RootInlineBox::fillLineSelectionGap): Make this function not assume a contiguous visual LTR selection range.
00:14 Changeset [39784] by ap@webkit.org

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=23207
Moved currentTime() to from WebCore to WTF.

00:06 Changeset [39783] by ap@webkit.org

2009-01-11 Dmitry Titov <dimich@chromium.org>

Reviewed by Darin Adler.

Make ScheduledAction (Timeout callback) able to execute the callback under WorkerContext.
Part of 'Timers in Workers' work.
https://bugs.webkit.org/show_bug.cgi?id=23223

  • bindings/js/ScheduledAction.cpp: (WebCore::ScheduledAction::execute): new method that takes WorkerContext. (WebCore::ScheduledAction::executeFunctionInContext): Executes a JSFunction in proper context with 'this' object.
  • bindings/js/ScheduledAction.h:

01/10/09:

23:45 Changeset [39782] by eric.carlson@apple.com

2009-01-10 Eric Carlson <eric.carlson@apple.com>

Reviewed by Eric Seidel.

https://bugs.webkit.org/show_bug.cgi?id=23210
Make it easier for ports to define custom UI for media controls

  • css/CSSStyleSelector.cpp: (WebCore::CSSStyleSelector::styleForElement): call theme()->styleSheetForMediaControls to get the media controller style sheet
  • rendering/MediaControlElements.cpp: (WebCore::MediaControlInputElement::hitTest): Added
  • rendering/MediaControlElements.h: Updated
  • rendering/RenderMedia.cpp: (WebCore::RenderMedia::forwardEvent): Call element hitTest() method instead of local function so control elements don't necessarily have to be rectangular
  • rendering/RenderTheme.cpp: (WebCore::RenderTheme::styleSheetForMediaControls): Added (WebCore::RenderTheme::hitTestMediaControlPart): Added
  • rendering/RenderTheme.h: Updated
18:47 Changeset [39781] by simon.fraser@apple.com

2009-01-10 Simon Fraser <simon.fraser@apple.com>

Reviewed by Dan Bernstein

https://bugs.webkit.org/show_bug.cgi?id=23222

We need to disable LayoutState when an object has transforms,
because LayoutState is not transform-aware and therefore
repaint rects can be computed incorrectly.

Test: fast/repaint/transform-disable-layoutstate.html

  • rendering/RenderBlock.cpp: (WebCore::RenderBlock::layoutBlock): (WebCore::RenderBlock::layoutOnlyPositionedObjects):
  • rendering/RenderFlexibleBox.cpp: (WebCore::RenderFlexibleBox::layoutBlock):
16:05 Changeset [39780] by abarth@webkit.org

Revert 39779.

15:16 Changeset [39779] by abarth@webkit.org

WebCore:

2009-01-10 Adam Barth <abarth@webkit.org>

Reviewed by Darin Adler.

Fix https://bugs.webkit.org/show_bug.cgi?id=21456

We shouldn't update the document pointer of inactive windows. This
behavior differs slightly from Firefox, I'd argue that this behavior is
more correct (we both differ from IE). The HTML 5 spec is a bit in flux
on this point.

Test: http/tests/security/xss-inactive-closure.html

  • bindings/js/JSDOMWindowBase.cpp: (WebCore::JSDOMWindowBase::~JSDOMWindowBase):
  • bindings/js/ScriptController.cpp: (WebCore::ScriptController::clearWindowShell): (WebCore::ScriptController::updateDocument):
  • bindings/js/ScriptController.h:

LayoutTests:

2009-01-10 Adam Barth <abarth@webkit.org>

Reviewed by Darin Adler.

Test that we don't do goofy things with the document pointer after
navigation.

  • fast/dom/Window/dom-access-from-closure-iframe-expected.txt:
  • fast/dom/Window/dom-access-from-closure-window-expected.txt:
  • http/tests/security/resources/childWithButton.html: Added.
  • http/tests/security/resources/xss-inactive-closure-child-2.html: Added.
  • http/tests/security/resources/xss-inactive-closure-child.html: Added.
  • http/tests/security/xss-inactive-closure-expected.txt: Added.
  • http/tests/security/xss-inactive-closure.html: Added.
14:03 Changeset [39778] by darin@apple.com

2009-01-10 Darin Adler <darin@apple.com>

Reviewed by Sam Weinig.

<rdar://problem/5845089> REGRESSION (r30044): Mail custom stationery missing images
because of change to -[HTMLObjectElement data]

  • WebView/WebView.mm: (-[WebView _commonInitializationWithFrameName:groupName:usesDocumentViews:]): Added a thread violation check because I saw this being done off the main thread while testing Mail, and it caused problems. Put all the one time initialization under a single guard to make things just a little faster other times, and to make it clearer which things are one-time. Added a call to the new patchMailRemoveAttributesMethod function. (-[WebView initWithFrame:frameName:groupName:]): Added a thread violation check here too, because I assumed it would be slightly better to have a public method name in the violation message. This calls commonInitialization later, so it will hit that one eventually. (objectElementDataAttribute): Added. Just returns the value of the "data" attribute. (recursivelyRemoveMailAttributes): Added. Patch to an internal Mail method that in turn patches a WebKit method and removes the patch again on the way out. (patchMailRemoveAttributesMethod): Added. On Leopard only, checks the Mail version, and then applies the patch that fixes this bug.

01/09/09:

19:46 Changeset [39777] by zecke@webkit.org

[QT] Make SharedTimer use QBasicTimer

Kill the SharedTimerQt.h and the usage of signal and slots
by using QBasicTimer.

19:46 Changeset [39776] by zecke@webkit.org

2009-01-09 Benjamin Otte <otte@gnome.org>

Reviewed by Holger Freyther.

https://bugs.webkit.org/show_bug.cgi?id=23194

make the finalize function the dispose function. Not releasing held
objects in dispose can cause crashers later.

  • webkit/webkitwebview.cpp:
19:46 Changeset [39775] by zecke@webkit.org

2009-01-09 Benjamin Otte <otte@gnome.org>

Reviewed by Holger Freyther.

https://bugs.webkit.org/show_bug.cgi?id=23194

Remove broken code that uses a fallback. The previous code did 2
policy decisions which caused crashes.
Download is still notImplemented();

  • webkit/webkitwebpolicydecision.cpp: (_WebKitWebPolicyDecisionPrivate::webkit_web_policy_decision_download):
19:46 Changeset [39774] by zecke@webkit.org

2009-01-09 Adam Bergkvist <adam.bergkvist@ericsson.com>

Reviewed by Holger Freyther.

[CURL] POST without body becomes GET
https://bugs.webkit.org/show_bug.cgi?id=23182

Make sure that method is always set to POST even when the body is
empty. Cleaned up httpBody null check.

  • platform/network/curl/ResourceHandleManager.cpp: (WebCore::ResourceHandleManager::setupPOST):
18:00 Changeset [39773] by treat@webkit.org

Build fix.

17:12 Changeset [39772] by adele@apple.com

2009-01-09 Adele Peterson <adele@apple.com>

Reviewed by Darin Adler.

Fix infinite recursion in clientPaddingLeft and clientPaddingRight. This changes the code back to
how it was before RenderTextControlSingleLine was split out from RenderTextControl.


No test added since this code is only exercised when trying to place a search field's recent searches popup menu.

  • rendering/RenderTextControlSingleLine.cpp: (WebCore::RenderTextControlSingleLine::clientPaddingLeft): (WebCore::RenderTextControlSingleLine::clientPaddingRight):
16:41 Changeset [39771] by darin@apple.com

WebCore:

2009-01-09 Darin Adler <darin@apple.com>

Reviewed by Jon Honeycutt.

Bug 22913: REGRESSION: Space bar doesn't scroll on windows
https://bugs.webkit.org/show_bug.cgi?id=22913
rdar://problem/6479834

  • page/EventHandler.cpp: (WebCore::EventHandler::defaultKeyboardEventHandler): Added code to call defaultSpaceEventHandler. (WebCore::EventHandler::defaultSpaceEventHandler): Added. Scrolls down or up based on the shift key.
  • page/EventHandler.h: Added defaultSpaceEventHandler.
  • platform/mac/ScrollViewMac.mm: (WebCore::ScrollView::platformScroll): Return false, because this function does not scroll. In an earlier version of this patch, I used this to prevent the patch from affecting the Mac, but I decided to use #if instead.

WebKit/win:

2009-01-09 Darin Adler <darin@apple.com>

Reviewed by Jon Honeycutt.

Bug 22913: REGRESSION: Space bar doesn't scroll on windows
https://bugs.webkit.org/show_bug.cgi?id=22913
rdar://problem/6479834

  • WebView.cpp: (WebView::keyPress): Removed code to handle space bar here; put it in WebCore instead.
16:38 Changeset [39770] by weinig@apple.com

2009-01-09 Sam Weinig <sam@webkit.org>

Check in correct results for failing test.

  • fast/dom/Window/window-properties-expected.txt:
16:14 Changeset [39769] by barraclough@apple.com

JavaScriptCore:

2009-01-09 Gavin Barraclough <barraclough@apple.com>

Reviewed by Oliver Hunt.

Stage two of converting JSValue from a pointer to a class type.
Remove the class JSValue. The functionallity has been transitioned
into the wrapper class type JSValuePtr.

The last stage will be to rename JSValuePtr to JSValue, remove the
overloaded -> operator, and switch operations on JSValuePtrs from
using '->' to use '.' instead.

  • API/APICast.h:
  • JavaScriptCore.exp:
  • runtime/JSCell.h: (JSC::asCell): (JSC::JSValuePtr::asCell): (JSC::JSValuePtr::isNumber): (JSC::JSValuePtr::isString): (JSC::JSValuePtr::isGetterSetter): (JSC::JSValuePtr::isObject): (JSC::JSValuePtr::getNumber): (JSC::JSValuePtr::getString): (JSC::JSValuePtr::getObject): (JSC::JSValuePtr::getCallData): (JSC::JSValuePtr::getConstructData): (JSC::JSValuePtr::getUInt32): (JSC::JSValuePtr::getTruncatedInt32): (JSC::JSValuePtr::getTruncatedUInt32): (JSC::JSValuePtr::mark): (JSC::JSValuePtr::marked): (JSC::JSValuePtr::toPrimitive): (JSC::JSValuePtr::getPrimitiveNumber): (JSC::JSValuePtr::toBoolean): (JSC::JSValuePtr::toNumber): (JSC::JSValuePtr::toString): (JSC::JSValuePtr::toObject): (JSC::JSValuePtr::toThisObject): (JSC::JSValuePtr::needsThisConversion): (JSC::JSValuePtr::toThisString): (JSC::JSValuePtr::getJSNumber):
  • runtime/JSImmediate.h: (JSC::JSValuePtr::isUndefined): (JSC::JSValuePtr::isNull): (JSC::JSValuePtr::isUndefinedOrNull): (JSC::JSValuePtr::isBoolean): (JSC::JSValuePtr::getBoolean): (JSC::JSValuePtr::toInt32): (JSC::JSValuePtr::toUInt32):
  • runtime/JSNumberCell.h: (JSC::JSValuePtr::uncheckedGetNumber): (JSC::JSValuePtr::toJSNumber):
  • runtime/JSObject.h: (JSC::JSValuePtr::isObject): (JSC::JSValuePtr::get): (JSC::JSValuePtr::put):
  • runtime/JSString.h: (JSC::JSValuePtr::toThisJSString):
  • runtime/JSValue.cpp: (JSC::JSValuePtr::toInteger): (JSC::JSValuePtr::toIntegerPreserveNaN): (JSC::JSValuePtr::toInt32SlowCase): (JSC::JSValuePtr::toUInt32SlowCase):
  • runtime/JSValue.h: (JSC::JSValuePtr::makeImmediate): (JSC::JSValuePtr::immediateValue): (JSC::JSValuePtr::JSValuePtr): (JSC::JSValuePtr::operator->): (JSC::JSValuePtr::operator bool): (JSC::JSValuePtr::operator==): (JSC::JSValuePtr::operator!=): (JSC::JSValuePtr::encode): (JSC::JSValuePtr::decode): (JSC::JSValuePtr::toFloat): (JSC::JSValuePtr::asValue): (JSC::operator==): (JSC::operator!=):

WebCore:

2009-01-09 Gavin Barraclough <barraclough@apple.com>

Reviewed by Oliver Hunt.

Delete references to JSValue, removing this class.

  • bindings/js/JSWorkerContextCustom.cpp: (WebCore::JSWorkerContext::self):
  • bindings/js/ScriptCallStack.h:
15:42 Changeset [39768] by eric@webkit.org

2009-01-09 Dimitri Glazkov <dglazkov@chromium.org>

Reviewed by Eric Seidel.

Add svg/graphics/skia files.
https://bugs.webkit.org/show_bug.cgi?id=23147

  • svg/graphics/skia/RenderPathSkia.cpp: Added. (WebCore::RenderPath::strokeContains):
  • svg/graphics/skia/SVGPaintServerPatternSkia.cpp: Added. (WebCore::SVGPaintServerPattern::setup):
  • svg/graphics/skia/SVGPaintServerSkia.cpp: Added. (WebCore::SVGPaintServer::draw): (WebCore::SVGPaintServer::teardown): (WebCore::SVGPaintServer::renderPath):
  • svg/graphics/skia/SVGResourceFilterSkia.cpp: Added. (WebCore::SVGResourceFilter::createPlatformData): (WebCore::SVGResourceFilter::prepareFilter): (WebCore::SVGResourceFilter::applyFilter):
  • svg/graphics/skia/SVGResourceMaskerSkia.cpp: Added. (WebCore::SVGResourceMasker::applyMask):
15:04 Changeset [39767] by eric@webkit.org

2009-01-09 Dimitri Glazkov <dglazkov@chromium.org>

Reviewed by Eric Seidel.

Add SkiaFontWin to platform/graphics/chromium
https://bugs.webkit.org/show_bug.cgi?id=23215

  • platform/graphics/skia/SkiaFontWin.cpp: Added. (WebCore::CachedOutlineKey::CachedOutlineKey): (WebCore::operator==): (WebCore::CachedOutlineKeyHash::hash): (WebCore::CachedOutlineKeyHash::equal): (WebCore::FIXEDToSkScalar): (WebCore::deleteOutline): (WebCore::addPolyCurveToPath): (WebCore::getPathForGlyph): (WebCore::SkiaWinOutlineCache::lookupOrCreatePathForGlyph): (WebCore::SkiaWinOutlineCache::removePathsForFont):
  • platform/graphics/skia/SkiaFontWin.h: Added.
14:55 Changeset [39766] by eric@webkit.org

2009-01-09 Dimitri Glazkov <dglazkov@chromium.org>

Reviewed by Eric Seidel.

Add PlatformContextSkia and PlatformGraphics.h to platform/graphics/skia
https://bugs.webkit.org/show_bug.cgi?id=23215

  • platform/graphics/skia/PlatformContextSkia.cpp: Added. (PlatformContextSkia::State::State): (PlatformContextSkia::State::~State): (PlatformContextSkia::State::applyAlpha): (PlatformContextSkia::PlatformContextSkia): (PlatformContextSkia::~PlatformContextSkia): (PlatformContextSkia::setCanvas): (PlatformContextSkia::save): (PlatformContextSkia::restore): (PlatformContextSkia::drawRect): (PlatformContextSkia::setupPaintCommon): (PlatformContextSkia::setupPaintForFilling): (PlatformContextSkia::setupPaintForStroking): (PlatformContextSkia::setDrawLooper): (PlatformContextSkia::setMiterLimit): (PlatformContextSkia::setAlpha): (PlatformContextSkia::setLineCap): (PlatformContextSkia::setLineJoin): (PlatformContextSkia::setPorterDuffMode): (PlatformContextSkia::setFillColor): (PlatformContextSkia::getDrawLooper): (PlatformContextSkia::getStrokeStyle): (PlatformContextSkia::setStrokeStyle): (PlatformContextSkia::setStrokeColor): (PlatformContextSkia::getStrokeThickness): (PlatformContextSkia::setStrokeThickness): (PlatformContextSkia::getTextDrawingMode): (PlatformContextSkia::setTextDrawingMode): (PlatformContextSkia::setUseAntialiasing): (PlatformContextSkia::fillColor): (PlatformContextSkia::beginPath): (PlatformContextSkia::addPath): (PlatformContextSkia::setFillRule): (PlatformContextSkia::setGradient): (PlatformContextSkia::setPattern): (PlatformContextSkia::setDashPathEffect): (PlatformContextSkia::paintSkPaint): (PlatformContextSkia::bitmap): (PlatformContextSkia::isPrinting):
  • platform/graphics/skia/PlatformContextSkia.h: Added. (PlatformContextSkia::currentPath): (PlatformContextSkia::canvas): (PlatformContextSkia::gdk_skia):
  • platform/graphics/skia/PlatformGraphics.h: Added.
14:35 Changeset [39765] by andersca@apple.com

2009-01-09 Anders Carlsson <andersca@apple.com>

Fix build.


  • WebCore.LP64.exp:
14:34 Changeset [39764] by eric@webkit.org

2009-01-09 Dimitri Glazkov <dglazkov@chromium.org>

Reviewed by Eric Seidel.

Add TransformationMatrixSkia to platform/graphics/skia
https://bugs.webkit.org/show_bug.cgi?id=23215

  • platform/graphics/skia/TransformationMatrixSkia.cpp: Added. (WebCore::TransformationMatrix::TransformationMatrix): (WebCore::TransformationMatrix::setMatrix): (WebCore::TransformationMatrix::map): (WebCore::TransformationMatrix::mapRect): (WebCore::TransformationMatrix::isIdentity): (WebCore::TransformationMatrix::reset): (WebCore::TransformationMatrix::scale): (WebCore::TransformationMatrix::rotate): (WebCore::TransformationMatrix::translate): (WebCore::TransformationMatrix::shear): (WebCore::TransformationMatrix::det): (WebCore::TransformationMatrix::inverse): (WebCore::TransformationMatrix::operator SkMatrix): (WebCore::TransformationMatrix::operator==): (WebCore::TransformationMatrix::operator*=): (WebCore::TransformationMatrix::operator*): (WebCore::TransformationMatrix::a): (WebCore::TransformationMatrix::setA): (WebCore::TransformationMatrix::b): (WebCore::TransformationMatrix::setB): (WebCore::TransformationMatrix::c): (WebCore::TransformationMatrix::setC): (WebCore::TransformationMatrix::d): (WebCore::TransformationMatrix::setD): (WebCore::TransformationMatrix::e): (WebCore::TransformationMatrix::setE): (WebCore::TransformationMatrix::f): (WebCore::TransformationMatrix::setF):
14:33 Changeset [39763] by ap@webkit.org

2009-01-09 Brent Fulgham <bfulgham@gmail.com>

Reviewed by Adam Roben.

Fixes: http://bugs.webkit.org/show_bug.cgi?id=23092
Separate CFNetwork-specific code in WebKit.dll into separate
implementation files. Create cURL stub files to be populated
with new implementation in a future bug report.

  • WebCookieManager.cpp: Move CFNetwork-specific routines to the new WebCookieManagerCFNet.cpp file. (WebCookieManager::Release):
  • WebCookieManagerCFNet.cpp: Copied from WebCookieManager.cpp. (WebCookieManager::cookieStorage):
  • WebCookieManagerCurl.cpp: Added. (WebCookieManager::cookieStorage): (WebCookieManager::setCookieStorage):
  • WebCoreSupport/WebFrameLoaderClient.cpp: Conditionalize two methods in this file. (WebFrameLoaderClient::dispatchDidReceiveAuthenticationChallenge): (WebFrameLoaderClient::savePlatformDataToCachedPage):
  • WebDownload.cpp: Move CFNetwork-specific logic to the new WebDownloadCFNet.cpp (WebDownload::bundleExtension): (WebDownload::bundleMagicNumber): (WebDownload::bundlePathForTargetPath): (WebDownload::extractResumeDataFromBundle): Made static class method so it could be used in WebDownload.cpp and WebDownloadCFNet.cpp (WebDownload::appendResumeDataToBundle): Made static class method so it could be used in WebDownload.cpp and WebDownloadCFNet.cpp
  • WebDownload.h:
  • WebDownloadCFNet.cpp: Copied from WebDownload.cpp. (WebDownload::initToResumeWithBundle): (WebDownload::setDestination): (didFailCallback):
  • WebDownloadCurl.cpp: Added. (WebDownload::init): (WebDownload::initWithRequest): (WebDownload::initToResumeWithBundle): (WebDownload::start): (WebDownload::cancel): (WebDownload::cancelForResume): (WebDownload::deletesFileUponFailure): (WebDownload::setDeletesFileUponFailure): (WebDownload::setDestination): (WebDownload::cancelAuthenticationChallenge): (WebDownload::continueWithoutCredentialForAuthenticationChallenge): (WebDownload::useCredential):
  • WebKit.vcproj/WebKit.vcproj: Modify Apple targets to exclude the new cURL files. Modify Cairo targets to exclude CFNet files.
  • WebMutableURLRequest.cpp: Conditionalize the single method with a CFNetwork-speicic call. (WebMutableURLRequest::mutableCopy):
  • WebURLAuthenticationChallenge.cpp: Conditionalize one line that causes build problems for cURL. This will be removed later. (WebURLAuthenticationChallenge::initWithProtectionSpace):
  • WebURLAuthenticationChallengeSender.cpp:
  • WebURLAuthenticationChallengeSenderCFNet.cpp: Copied from WebURLAuthenticationChallengeSender.cpp. (WebURLAuthenticationChallengeSender::useCredential):
  • WebURLAuthenticationChallengeSenderCurl.cpp: Added. (WebURLAuthenticationChallengeSender::cancelAuthenticationChallenge): (WebURLAuthenticationChallengeSender::continueWithoutCredentialForAuthenticationChallenge): (WebURLAuthenticationChallengeSender::useCredential):
  • WebURLResponse.cpp: Exclude WebKitSystemInterface when compiling for the cURL target.
14:21 Changeset [39762] by ap@webkit.org

Reviewed by Darin Adler.

<rdar://problem/6477502> Repro crash reloading page using HTML5 AppCache on Windows

Test: http/tests/appcache/reload.html

On Windows, reloading resulted in the following sequence of events:
1) A new main resource loader was created and associated with the cache. Its document loader
was not associated with the cache yet.
2) The old document loader was destroyed, and since it was the last one, ApplicationCacheGroup
teardown started, with newestCache being zeroed out.
3) The new document loader was associated with the cache group, and cache update started.
4) Since newestCache was null, this caused many problems.

  • loader/appcache/ApplicationCacheGroup.cpp: (WebCore::ApplicationCacheGroup::documentLoaderDestroyed): Improve comments and assertions, no functional changes. (WebCore::ApplicationCacheGroup::associateDocumentLoaderWithCache): Revive the cache group if its newest cache pointer is already saved for teardown. Note that the cache pointer validity is ensured by MainResourceLoader that holds a reference to it.
14:20 Changeset [39761] by ap@webkit.org

Reviewed by Darin Adler.

Fix many appcache tests on Windows by making synchronous XMLHttpRequest throw on cache miss.

  • WebFrame.cpp: (WebFrame::cannotShowURLError): Implement, so that synchronous XHR can raise exceptions, as needed for AppCache tests. I do not know why this error was not implemented.
14:12 Changeset [39760] by darin@chromium.org

2009-01-09 Darin Fisher <darin@chromium.org>

Reviewed by Eric Seidel.

https://bugs.webkit.org/show_bug.cgi?id=23214
Upstream platform/network/chromium/

  • platform/network/chromium/AuthenticationChallenge.h: Added. (WebCore::AuthenticationChallenge::AuthenticationChallenge): (WebCore::AuthenticationChallenge::sourceHandle):
  • platform/network/chromium/AuthenticationChallengeChromium.cpp: Added. (WebCore::AuthenticationChallenge::platformCompare):
  • platform/network/chromium/CookieJarChromium.cpp: Added. (WebCore::setCookies): (WebCore::cookies): (WebCore::cookiesEnabled):
  • platform/network/chromium/DNSChromium.cpp: Added. (WebCore::prefetchDNS):
  • platform/network/chromium/NetworkStateNotifierChromium.cpp: Added. (WebCore::NetworkStateNotifier::updateState): (WebCore::NetworkStateNotifier::NetworkStateNotifier):
  • platform/network/chromium/NetworkStateNotifierPrivate.h: Added.
  • platform/network/chromium/ResourceError.h: Added. (WebCore::ResourceError::ResourceError):
  • platform/network/chromium/ResourceRequest.h: Added. (WebCore::ResourceRequest::): (WebCore::ResourceRequest::ResourceRequest): (WebCore::ResourceRequest::frame): (WebCore::ResourceRequest::setFrame): (WebCore::ResourceRequest::setTargetType): (WebCore::ResourceRequest::targetType): (WebCore::ResourceRequest::originPid): (WebCore::ResourceRequest::setOriginPid): (WebCore::ResourceRequest::securityInfo): (WebCore::ResourceRequest::setSecurityInfo): (WebCore::ResourceRequest::doUpdatePlatformRequest): (WebCore::ResourceRequest::doUpdateResourceRequest):
  • platform/network/chromium/ResourceResponse.h: (WebCore::ResourceResponse::ResourceResponse): (WebCore::ResourceResponse::getSecurityInfo): (WebCore::ResourceResponse::setSecurityInfo): (WebCore::ResourceResponse::isContentFiltered): (WebCore::ResourceResponse::setIsContentFiltered): (WebCore::ResourceResponse::doUpdateResourceResponse):
13:44 Changeset [39759] by barraclough@apple.com

fix tyop in Changelog

13:42 Changeset [39758] by mitz@apple.com

Reviewed by Darin Adler.

  • fixed <rdar://problem/6234347> Add/change conditional key bindings for changing paragraph- and character-level writing direction (to match NSTextView)
  • WebView/WebHTMLView.mm: (-[WebHTMLView validateUserInterfaceItemWithoutDelegate:]): Added code to validate makeBaseWritingDirectionLeftToRight: and makeBaseWritingDirectionRightToLeft:. (writingDirectionKeyBindingsEnabled): Changed this function to always return YES, except on Tiger and Leopard. (-[WebHTMLView makeBaseWritingDirectionLeftToRight:]): Renamed changeBaseWritingDirectionToLTR: to this. (-[WebHTMLView makeBaseWritingDirectionRightToLeft:]): Renamed changeBaseWritingDirectionToRTL: to this. (-[WebHTMLView changeBaseWritingDirectionToLTR:]): Now calls makeBaseWritingDirectionLeftToRight:. (-[WebHTMLView changeBaseWritingDirectionToRTL:]): Now calls makeBaseWritingDirectionRightToLeft:.
  • WebView/WebView.mm: Added makeBaseWritingDirectionLeftToRight and makeBaseWritingDirectionRightToLeft to FOR_EACH_RESPONDER_SELECTOR.
13:01 Changeset [39757] by adele@apple.com

2009-01-09 Darin Adler <darin@apple.com>

Reviewed and landed by Adele Peterson.

Bug 23160: add setMemoryCacheClientCallsEnabled SPI so Safari can be faster with activity window closed
https://bugs.webkit.org/show_bug.cgi?id=23160

  • Interfaces/IWebViewPrivate.idl: Added setMemoryCacheDelegateCallsEnabled.
  • WebView.cpp: (WebView::setMemoryCacheDelegateCallsEnabled): Ditto.
  • WebView.h: Ditto.
12:53 Changeset [39756] by wsiegrist@apple.com

2009-01-09 William Siegrist <wsiegrist@apple.com>
Reviewed by David Kilzer

Update URLs which go through redirects by changing bugs.webkit.org URLs to use https and removing /projects/webkit from Trac URLs.

Replace tabs in bugpriorities.html with spaces.

12:17 Changeset [39755] by oliver@apple.com

2009-01-09 David Levin <levin@chromium.org>

Reviewed by Oliver Hunt.

https://bugs.webkit.org/show_bug.cgi?id=23175

Adjustment to previous patch. Remove call to initilizeThreading from JSGlobalCreate
and fix jsc.cpp instead.

11:42 Changeset [39754] by abarth@webkit.org

2009-01-09 Joel Weinberger <jww@cs.berkeley.edu>

Reviewed by Darin Adler.
Landed by Adam Barth.

The test intended to test reload, but reference reload in one place and
replace in another. This makes them consistent.

  • http/tests/security/resources/xss-DENIED-assign-location-reload-attacker.html:
11:12 Changeset [39753] by hausmann@webkit.org

2009-01-09 Simon Hausmann <simon.hausmann@nokia.com>

Rubber-stamped by Tor Arne Vestbø.

Documentation fixes, remove duplicated docs, document missing
parameters and return values.

10:47 Changeset [39752] by weinig@apple.com

2009-01-09 Sam Weinig <sam@webkit.org>

Roll r39720 back in with a working interpreted mode.

10:46 Changeset [39751] by pkasting@chromium.org

2009-01-09 Peter Kasting <pkasting@google.com>

Reviewed by David Hyatt.

https://bugs.webkit.org/show_bug.cgi?id=22995
Fix flaky image animation by resetting the metadata for a frame when
we get new data for that frame. This fixes animations that halted
when the next frame of the animation wasn't complete by the time the
current frame's duration expired.

  • platform/graphics/BitmapImage.cpp: (WebCore::BitmapImage::destroyDecodedData): (WebCore::BitmapImage::dataChanged):
  • platform/graphics/BitmapImage.h: (WebCore::FrameData::~FrameData):
  • platform/graphics/cairo/ImageCairo.cpp: (WebCore::FrameData::clear):
  • platform/graphics/cg/ImageCG.cpp: (WebCore::FrameData::clear):
  • platform/graphics/qt/ImageQt.cpp: (WebCore::FrameData::clear):
  • platform/graphics/wx/ImageWx.cpp: (WebCore::FrameData::clear):
10:28 Changeset [39750] by oliver@apple.com

Whoops, include actual file the the forwarding headers point to.

09:58 Changeset [39749] by hausmann@webkit.org

2009-01-09 Simon Hausmann <simon.hausmann@nokia.com>

Rubber-stamped by Tor Arne Vestbø.

Swap the .h and the .cpp output in the Qt build for the
UserAgentStyleSheetData files, to put the definition in the .cpp file
and the declaration in the header file.

09:54 Changeset [39748] by oliver@apple.com

include files missed in last commit.

09:53 Changeset [39747] by oliver@apple.com

2009-01-09 David Levin <levin@chromium.org>

Reviewed by Oliver Hunt.

https://bugs.webkit.org/show_bug.cgi?id=23175

Added a template to make the pointer and flags combination
in UString more readable and less error prone.

09:47 Changeset [39746] by hausmann@webkit.org

2009-01-09 Simon Hausmann <simon.hausmann@nokia.com>

Reviewed by Tor Arne Vestbø.

Fix linking against QtWebKit on Windows, by defining the EXPORT macros
_exactly_ the same way as in qglobal.h, depending on QT_NODLL,
QT_MAKEDLL and QT_SHARED.

08:51 Changeset [39745] by vestbo@webkit.org

2009-01-09 Tor Arne Vestbø <tavestbo@trolltech.com>

Reviewed by Simon Hausmann.

Prevent qmake from generating duplicate rules for embedded stylesheets

The STYLESHEETS_EMBED variable used to contain only one file name, but
now that it is a list of files we need to change it to be a dependency
for the stylesheet generator instead of the input.

05:03 Changeset [39744] by hausmann@webkit.org

WebCore:

2009-01-09 Simon Hausmann <simon.hausmann@nokia.com>

Reviewed by Tor Arne Vestbø.

Fix qmake warning about missing test function, a contains() call was
missing.

WebKit/qt:

2009-01-09 Simon Hausmann <simon.hausmann@nokia.com>

Reviewed by Tor Arne Vestbø.

Remove qwebnetworkinterface.h from WEBKIT_API_HEADERS as it's not
public Qt 4.4/4.5 API.

04:32 Changeset [39743] by ap@webkit.org

https://bugs.webkit.org/show_bug.cgi?id=23165

  • fast/dom/Window/window-properties-expected.txt: Forgot to commit this with r39699, too.
04:16 Changeset [39742] by hausmann@webkit.org

2009-01-09 Simon Hausmann <simon.hausmann@nokia.com>

Unreviewed Qt build fix, remove qcookiejar.h from the API headers and
therefore HEADERS. It's not part of the API and it moc shouldn't be
called on it.

03:10 Changeset [39741] by hausmann@webkit.org

2009-01-09 Simon Hausmann <simon.hausmann@nokia.com>

Rubber-stamped by Tor Arne Vestbø.

Add the Qt API headers to HEADERS, too, for improved completion in IDEs.

01:42 Changeset [39740] by ap@webkit.org

2009-01-09 David Levin <levin@chromium.org>

Reviewed by Alexey Proskuryakov.

https://bugs.webkit.org/show_bug.cgi?id=23199

Fix deref's of string happening on two different threads.

No observable change in behavior, so no test. However, there are asserts
being added for https://bugs.webkit.org/show_bug.cgi?id=23175 which
will detect this issue while running the normal worker tests.

  • dom/WorkerThread.cpp: (WebCore::WorkerThreadStartupData::create): (WebCore::WorkerThreadStartupData::WorkerThreadStartupData): (WebCore::WorkerThread::WorkerThread): (WebCore::WorkerThread::workerThread):
  • dom/WorkerThread.h:
00:12 Changeset [39739] by barraclough@apple.com

2009-01-08 Gavin Barraclough <barraclough@apple.com>

RS by Oliver Hunt.

Looks like this should have been updated with r39699.

  • fast/js/global-constructors-expected.txt:
00:11 Changeset [39738] by barraclough@apple.com

2009-01-08 Gavin Barraclough <barraclough@apple.com>

Reviewed by Oliver Hunt.

Encode immediates in the low word of JSValuePtrs, on x86-64.

On 32-bit platforms a JSValuePtr may represent a 31-bit signed integer.
On 64-bit platforms, if USE(ALTERNATE_JSIMMEDIATE) is defined, a full
32-bit integer may be stored in an immediate.


Presently USE(ALTERNATE_JSIMMEDIATE) uses the same encoding as the default
immediate format - the value is left shifted by one, so a one bit tag can
be added to indicate the value is an immediate. However this means that
values must be commonly be detagged (by right shifting by one) before
arithmetic operations can be performed on immediates. This patch modifies
the formattting so the the high bits of the immediate mark values as being
integer.

  • assembler/MacroAssembler.h: (JSC::MacroAssembler::not32): (JSC::MacroAssembler::orPtr): (JSC::MacroAssembler::zeroExtend32ToPtr): (JSC::MacroAssembler::jaePtr): (JSC::MacroAssembler::jbPtr): (JSC::MacroAssembler::jnzPtr): (JSC::MacroAssembler::jzPtr):
  • assembler/X86Assembler.h: (JSC::X86Assembler::): (JSC::X86Assembler::notl_r): (JSC::X86Assembler::testq_i32r):
  • jit/JIT.cpp: (JSC::JIT::privateCompileMainPass): (JSC::JIT::privateCompileSlowCases): (JSC::JIT::privateCompileCTIMachineTrampolines):
  • jit/JIT.h:
  • jit/JITArithmetic.cpp: (JSC::JIT::compileFastArith_op_lshift): (JSC::JIT::compileFastArith_op_rshift): (JSC::JIT::compileFastArith_op_bitand): (JSC::JIT::compileFastArithSlow_op_bitand): (JSC::JIT::compileFastArith_op_mod): (JSC::JIT::compileFastArithSlow_op_mod): (JSC::JIT::compileFastArith_op_add): (JSC::JIT::compileFastArith_op_mul): (JSC::JIT::compileFastArith_op_post_inc): (JSC::JIT::compileFastArith_op_post_dec): (JSC::JIT::compileFastArith_op_pre_inc): (JSC::JIT::compileFastArith_op_pre_dec): (JSC::JIT::putDoubleResultToJSNumberCellOrJSImmediate): (JSC::JIT::compileBinaryArithOp):
  • jit/JITCall.cpp: (JSC::JIT::compileOpCallSlowCase):
  • jit/JITInlineMethods.h: (JSC::JIT::emitJumpIfJSCell): (JSC::JIT::emitJumpIfNotJSCell): (JSC::JIT::emitJumpIfImmNum): (JSC::JIT::emitJumpSlowCaseIfNotImmNum): (JSC::JIT::emitJumpSlowCaseIfNotImmNums): (JSC::JIT::emitFastArithDeTagImmediate): (JSC::JIT::emitFastArithDeTagImmediateJumpIfZero): (JSC::JIT::emitFastArithReTagImmediate): (JSC::JIT::emitFastArithImmToInt): (JSC::JIT::emitFastArithIntToImmNoCheck): (JSC::JIT::emitTagAsBoolImmediate):
  • jit/JITPropertyAccess.cpp: (JSC::resizePropertyStorage): (JSC::JIT::privateCompilePutByIdTransition): (JSC::JIT::privateCompilePatchGetArrayLength): (JSC::JIT::privateCompileGetByIdSelf): (JSC::JIT::privateCompileGetByIdProto): (JSC::JIT::privateCompileGetByIdChain): (JSC::JIT::privateCompilePutByIdReplace):
  • runtime/JSImmediate.h: (JSC::JSImmediate::isNumber): (JSC::JSImmediate::isPositiveNumber): (JSC::JSImmediate::areBothImmediateNumbers): (JSC::JSImmediate::xorImmediateNumbers): (JSC::JSImmediate::rightShiftImmediateNumbers): (JSC::JSImmediate::canDoFastAdditiveOperations): (JSC::JSImmediate::addImmediateNumbers): (JSC::JSImmediate::subImmediateNumbers): (JSC::JSImmediate::makeInt): (JSC::JSImmediate::toBoolean):
  • wtf/Platform.h:
Note: See TracTimeline for information about the timeline view.