Timeline



Jul 7, 2009:

11:46 PM Changeset in webkit [45621] by ggaren@apple.com
  • 2190 edits
    69 copies
    214 adds
    411 deletes in branches/nitro-extreme

Merged r45243:45600 into nitro-extreme branch.

10:38 PM Changeset in webkit [45620] by oliver@apple.com
  • 5 edits in trunk/WebCore

Reduce complexity of lifetime management in DynamicNodeList caches
<https://bugs.webkit.org/show_bug.cgi?id=27068>

Reviewed by Maciej Stachowiak

Switch the Cache object used by DynamicNodeList into a normal
refcounted object rather than having a weird flag controlled
refcounting system, where positive refcount did not automatically
imply the cache object would actually still be live.

10:33 PM Changeset in webkit [45619] by mitz@apple.com
  • 25 edits in trunk/LayoutTests/platform

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

10:26 PM Changeset in webkit [45618] by Simon Fraser
  • 4 edits
    8 adds in trunk

2009-07-07 Simon Fraser <Simon Fraser>

Reviewed by Dan Bernstein.

-webkit-perspective should be a Length
https://bugs.webkit.org/show_bug.cgi?id=27066

-webkit-perspective should not take a magic valueless number, but should
be a normal Length value which responds to zooming. Treat valueless numbers
as pixels for backward compatibility.

Test: transforms/3d/general/perspective-units.html

  • css/CSSParser.cpp: (WebCore::CSSParser::parseValue):
  • css/CSSStyleSelector.cpp: (WebCore::CSSStyleSelector::applyProperty):
7:26 PM Changeset in webkit [45617] by adachan@apple.com
  • 2 edits in trunk/WebKit/win

Reviewed by Dave Hyatt.

Missed a call to FocusController::setFocused(true) when handling WM_SETFOCUS in WebView.
https://bugs.webkit.org/show_bug.cgi?id=27057

  • WebView.cpp: (WebViewWndProc):
6:55 PM Changeset in webkit [45616] by beidson@apple.com
  • 2 edits in trunk/WebKitTools

2009-07-07 Brady Eidson <beidson@apple.com>

Tiger build fix, work around missing NSString API.

  • DumpRenderTree/mac/DumpRenderTree.mm: (dumpHistoryItem):
6:41 PM Changeset in webkit [45615] by beidson@apple.com
  • 5 edits in trunk

WebKitTools:

2009-07-07 Brady Eidson <beidson@apple.com>

Reviewed by Mark Rowe.

https://bugs.webkit.org/show_bug.cgi?id=27049 - In dumpBackForwardList() mode, DRT should normalize file urls.

Make the dump of a history item agnostic to the layout of filesystem on the testing machine.

  • DumpRenderTree/mac/DumpRenderTree.mm: (dumpHistoryItem):
  • DumpRenderTree/win/DumpRenderTree.cpp: (dumpHistoryItem):

LayoutTests:

2009-07-07 Brady Eidson <beidson@apple.com>

Reviewed by Mark Rowe.

https://bugs.webkit.org/show_bug.cgi?id=27049 - In dumpBackForwardList() mode, DRT should normalize file urls.

  • fast/loader/subframe-navigate-during-main-frame-load-expected.txt: Make the results agnostic to the layout of filesystem on the testing machine.
6:19 PM Changeset in webkit [45614] by pdherbemont@apple.com
  • 5 edits
    3 adds in trunk

WebCore:

2009-07-07 Pierre d'Herbemont <pdherbemont@apple.com>

Reviewed by Simon Fraser.

https://bugs.webkit.org/show_bug.cgi?id=27047

We need to make sure that when we reattach, we also reattach
the children in every MediaControlElement. Else we may end up
having no remaining or elapsed time.

We have to handle that, because we are using a special shadow
tree in the DOM, and that we are ourselves handling
attaching/detaching the renderer.

The strategy here is to try to implement ::attach(), and try
to reuse as much code as we can from the super class, including
children attachement.

Test: media/controls-after-reload.html

  • rendering/MediaControlElements.cpp: (WebCore::MediaControlElement::styleForElement): Code factoring. (WebCore::MediaControlElement::rendererIsNeeded): Code factoring. (WebCore::MediaControlElement::attach): Implement attach and call super class so that children are also attached. (WebCore::MediaControlElement::updateStyle): Use attach() (WebCore::MediaControlInputElement::styleForElement): Code factoring. (WebCore::MediaControlInputElement::rendererIsNeeded): Code factoring. (WebCore::MediaControlInputElement::attach): See above. (WebCore::MediaControlInputElement::updateStyle): Use attach()
  • rendering/MediaControlElements.h:
  • rendering/RenderMedia.cpp: (WebCore::RenderMedia::updateControls): Directly run attach() on the m_panel, which is a root node for our shadow tree.

LayoutTests:

2009-07-07 Pierre d'Herbemont <pdherbemont@apple.com>

Reviewed by Simon Fraser.

https://bugs.webkit.org/show_bug.cgi?id=27047

Make sure nothing changing the src and reloading doesn't
change the look of the controller.

  • media/controls-after-reload.html: Added.
  • platform/mac-leopard/media/controls-after-reload-expected.txt: Added.
  • platform/mac/media/controls-after-reload-expected.txt: Added.
5:19 PM Changeset in webkit [45613] by Simon Fraser
  • 5 edits
    4 adds in trunk

2009-07-07 Simon Fraser <Simon Fraser>

Reviewed by Dan Bernstein.

Repaint issue after layer drops out of composited mode.
<https://bugs.webkit.org/show_bug.cgi?id=27022>

RenderLayers cache repaint rects in the form of m_repaintRect and m_outlineBox,
and expect these to stay valid from one style change to the next. These rects
are relative to the repaint container, so if a layer stops being composited,
we need to recompute them.

Test: compositing/repaint/layer-repaint-rects.html

  • rendering/RenderLayer.cpp: (WebCore::RenderLayer::computeRepaintRects):
  • rendering/RenderLayer.h:
  • rendering/RenderLayerCompositor.cpp: (WebCore::RenderLayerCompositor::updateBacking):
4:45 PM Changeset in webkit [45612] by mitz@apple.com
  • 8 edits
    4 adds in trunk

WebCore:

Reviewed by Dave Hyatt and Darin Adler.

Test: fast/css/pseudo-cache-stale.html

  • dom/Element.cpp: (WebCore::Element::pseudoStyleCacheIsInvalid): Added. Given the old style and the new style, goes over cached pseudo-styles in the old style and re-resolves the same style types off the new style. If any of the new pseudo-styles is different from the currently cached corresponding style, returns true. Otherwise, returns false. (WebCore::Element::recalcStyle): Validate the pseudo-style cache before deciding to keep the existing style.
  • dom/Element.h:
  • rendering/RenderObject.cpp: (WebCore::RenderObject::uncachedFirstLineStyle): Added this version that returns an uncached first-line style based off the given style. (WebCore::RenderObject::getUncachedPseudoStyle): Added the 'ownStyle' parameter.
  • rendering/RenderObject.h:
  • rendering/style/RenderStyle.cpp: (WebCore::RenderStyle::getPseudoStyleCache): Added. Returns the cached pseudo-styles in the passed-in vector.
  • rendering/style/RenderStyle.h:

LayoutTests:

Reviewed by Dave Hyatt and Darin Adler.

  • fast/css/pseudo-cache-stale.html: Added.
  • platform/mac/fast/css/pseudo-cache-stale-expected.checksum: Added.
  • platform/mac/fast/css/pseudo-cache-stale-expected.png: Added.
  • platform/mac/fast/css/pseudo-cache-stale-expected.txt: Added.
4:44 PM Changeset in webkit [45611] by darin@chromium.org
  • 1 edit
    1 move
    1 add in trunk/LayoutTests

2009-07-07 Dirk Pranke <dpranke@chromium.org>

Reviewed by Darin Fisher.

revise test expectations for bug 26088
http://bugs.webkit.org/show_bug.cgi?id=26088

generate pixel expectations, move location of generic expected file
for the crash test.

  • fast/text/text-large-negative-letter-spacing-with-opacity-expected.txt: Copied from platform/mac/fast/text/text-large-negative-letter-spacing-with-opacity-expected.txt.
  • platform/mac/fast/text/text-large-negative-letter-spacing-with-opacity-expected.txt: Removed.
  • platform/mac/fast/text/text-letter-spacing-expected.checksum: Added.
  • platform/mac/fast/text/text-letter-spacing-expected.png: Added.
3:04 PM Changeset in webkit [45610] by pdherbemont@apple.com
  • 1 edit
    1 add in trunk/LayoutTests

2009-07-07 Pierre d'Herbemont <pdherbemont@apple.com>

Reviewed by Simon Fraser.

Add a specific leopard result for controls-styling.html, given that it doesn't
use the new controller look.

  • platform/mac-leopard/media/controls-styling-expected.txt: Added.
2:52 PM Changeset in webkit [45609] by barraclough@apple.com
  • 15 edits in trunk

JavaScriptCore:

2009-07-07 Gavin Barraclough <barraclough@apple.com>

Reviewed by Sam Weinig.

Stop loading constants into the register file.

Instead, use high register values (highest bit bar the sign bit set) to indicate
constants in the instruction stream, and when we encounter such a value load it
directly from the CodeBlock.

Since constants are no longer copied into the register file, this patch renders
the 'unexpected constant' mechanism redundant, and removes it.

2% improvement, thanks to Sam Weinig.

  • bytecode/CodeBlock.cpp: (JSC::CodeBlock::dump): (JSC::CodeBlock::CodeBlock): (JSC::CodeBlock::mark): (JSC::CodeBlock::shrinkToFit):
  • bytecode/CodeBlock.h: (JSC::CodeBlock::isTemporaryRegisterIndex): (JSC::CodeBlock::constantRegister): (JSC::CodeBlock::isConstantRegisterIndex): (JSC::CodeBlock::getConstant): (JSC::ExecState::r):
  • bytecode/Opcode.h:
  • bytecompiler/BytecodeGenerator.cpp: (JSC::BytecodeGenerator::preserveLastVar): (JSC::BytecodeGenerator::BytecodeGenerator): (JSC::BytecodeGenerator::addConstantValue): (JSC::BytecodeGenerator::emitEqualityOp): (JSC::BytecodeGenerator::emitLoad): (JSC::BytecodeGenerator::emitResolveBase): (JSC::BytecodeGenerator::emitResolveWithBase): (JSC::BytecodeGenerator::emitNewError):
  • bytecompiler/BytecodeGenerator.h: (JSC::BytecodeGenerator::emitNode):
  • interpreter/CallFrame.h: (JSC::ExecState::noCaller): (JSC::ExecState::hasHostCallFrameFlag): (JSC::ExecState::addHostCallFrameFlag): (JSC::ExecState::removeHostCallFrameFlag):
  • interpreter/Interpreter.cpp: (JSC::Interpreter::resolve): (JSC::Interpreter::resolveSkip): (JSC::Interpreter::resolveGlobal): (JSC::Interpreter::resolveBase): (JSC::Interpreter::resolveBaseAndProperty): (JSC::Interpreter::resolveBaseAndFunc): (JSC::Interpreter::dumpRegisters): (JSC::Interpreter::throwException): (JSC::Interpreter::createExceptionScope): (JSC::Interpreter::privateExecute): (JSC::Interpreter::retrieveArguments):
  • jit/JIT.cpp: (JSC::JIT::privateCompileMainPass):
  • jit/JITInlineMethods.h: (JSC::JIT::emitLoadDouble): (JSC::JIT::emitLoadInt32ToDouble):
  • jit/JITOpcodes.cpp: (JSC::JIT::emit_op_new_error): (JSC::JIT::emit_op_enter): (JSC::JIT::emit_op_enter_with_activation):
  • parser/Nodes.cpp: (JSC::DeleteResolveNode::emitBytecode): (JSC::DeleteValueNode::emitBytecode): (JSC::PrefixResolveNode::emitBytecode):
  • runtime/JSActivation.cpp: (JSC::JSActivation::JSActivation):
  • wtf/Platform.h:

LayoutTests:

2009-07-07 Gavin Barraclough <barraclough@apple.com>

Reviewed by Geoff Garen.

fast/js/global-recursion-on-full-stack.html is a little finicky.

The test recurses down the JS stack to find the bottom (catching this with an exception),
then tries to call a host function (document.write), switch writes new '<script>' code,
and expects this code to be run, then expects this code to try to call 'f();' again,
which it expects to fail, and it expects to catch that exception. However it is possible
that one of the earlier stages (the call to document.write, entering the interpreter to
run the new global code) will fail, depending on exactly how much stack space was free at
the point the last call to f() failed.

Tweak the test to make it work.

  • fast/js/global-recursion-on-full-stack.html:
2:36 PM Changeset in webkit [45608] by mitz@apple.com
  • 5 edits
    8 adds in trunk

WebCore:

Reviewed by Simon Fraser.

Tests: fast/repaint/float-in-new-block-with-layout-delta.html

fast/repaint/float-new-in-block.html

  • rendering/RenderBlock.cpp: (WebCore::RenderBlock::layoutBlockChildren): If the child has never been laid out before, paint its overhanging floats in addition to itself.
  • rendering/RenderBlock.h: (WebCore::RenderBlock::FloatWithRect::FloatWithRect): Added and initialized a boolean everHadLayout member.
  • rendering/bidi.cpp: (WebCore::RenderBlock::layoutInlineChildren): Paint floats that have never been laid out before and did not move from (0, 0).

LayoutTests:

Reviewed by Simon Fraser.

  • fast/repaint/float-in-new-block-with-layout-delta.html: Added.
  • fast/repaint/float-new-in-block.html: Added.
  • platform/mac/fast/repaint/float-in-new-block-with-layout-delta-expected.checksum: Added.
  • platform/mac/fast/repaint/float-in-new-block-with-layout-delta-expected.png: Added.
  • platform/mac/fast/repaint/float-in-new-block-with-layout-delta-expected.txt: Added.
  • platform/mac/fast/repaint/float-new-in-block-expected.checksum: Added.
  • platform/mac/fast/repaint/float-new-in-block-expected.png: Added.
  • platform/mac/fast/repaint/float-new-in-block-expected.txt: Added.
2:27 PM Changeset in webkit [45607] by andersca@apple.com
  • 5 edits in trunk/WebKit/mac

2009-07-07 Anders Carlsson <andersca@apple.com>

Reviewed by Sam Weinig.

  • Fix <rdar://problem/6544693>.


For Flash, don't cache which methods or fields in an object are missing, since they can
be added at any time.


  • Plugins/Hosted/NetscapePluginHostManager.mm: (WebKit::NetscapePluginHostManager::hostForPackage):
  • Plugins/Hosted/NetscapePluginHostProxy.h: (WebKit::NetscapePluginHostProxy::shouldCacheMissingPropertiesAndMethods):
  • Plugins/Hosted/NetscapePluginHostProxy.mm: (WebKit::NetscapePluginHostProxy::NetscapePluginHostProxy):
  • Plugins/Hosted/ProxyInstance.mm: (WebKit::ProxyInstance::methodsNamed): (WebKit::ProxyInstance::fieldNamed):
2:19 PM Changeset in webkit [45606] by beidson@apple.com
  • 6 edits
    3 adds in trunk

WebCore:

2009-07-07 Brady Eidson <beidson@apple.com>

Contributions from both Darin Adler and Brady Eidson.
Reviewed by Darin Adler.

<rdar://problem/7024039> REGRESSION (r42158): Back-forward navigation does not work correctly on nytimes.com

In r42158 we lost some exclusions to making back/forward items for subframe navigations when a main frame
navigation was still in progress.

This patch makes things even better than it used to be by:

  • Locking back/forward history at the time the navigation is scheduled, not after the load has committed.
  • Locking back/forward history if *any* ancestor frame is still loading instead of just the main frame.

Test: fast/loader/subframe-navigate-during-main-frame-load.html

  • loader/FrameLoader.cpp: (WebCore::FrameLoader::submitForm): Move out the history locking logic into a standalone function so it can be used from multiple sites. (WebCore::mustLockBackForwardList): Returns true if any ancestor frame is still loading. (WebCore::FrameLoader::scheduleLocationChange): Respect mustLockBackForwardList() in addition to the passed in lockBackForwardList flag. (WebCore::FrameLoader::scheduleFormSubmission): Ditto.

LayoutTests:

2009-07-07 Brady Eidson <beidson@apple.com>

Reviewed by Darin Adler.

<rdar://problem/7024039> REGRESSION (r42158): Back-forward navigation does not work correctly on nytimes.com

  • fast/loader/resources/subframe-navigate-during-main-frame-load2.html: Added.
  • fast/loader/subframe-navigate-during-main-frame-load-expected.txt: Added.
  • fast/loader/subframe-navigate-during-main-frame-load.html: Added.
  • http/tests/navigation/javascriptlink-subframeload-expected.txt: Updated results to reflect the progression!
  • platform/mac/http/tests/navigation/javascriptlink-frames-expected.txt: Ditto!
  • http/tests/misc/redirect-to-about-blank.html: Add missing semi-colons.
2:15 PM Changeset in webkit [45605] by Nate Chapin
  • 1 edit
    1 add in trunk/WebCore

2009-07-07 Nate Chapin <Nate Chapin>

Reviewed by Darin Fisher.

Upstream DOMObjectsInclude.h from src.chromium.org.

https://bugs.webkit.org/show_bug.cgi?id=27035

  • bindings/v8/DOMObjectsInclude.h: Added.
1:16 PM Changeset in webkit [45604] by mrowe@apple.com
  • 1 edit in trunk/WebKitLibraries/win/tools/scripts/PRODUCTVERSION

Versioning.k

1:02 PM Changeset in webkit [45603] by Nate Chapin
  • 17 edits
    1 copy in trunk/WebCore

2009-07-07 Nate Chapin <Nate Chapin>

Reviewed by David Levin.

Relanding r45559, which was rolled back at r45574.

https://bugs.webkit.org/show_bug.cgi?id=26857

12:05 PM Changeset in webkit [45602] by Chris Fleizach
  • 2 edits in trunk/WebCore

Bug 26995 - AX: Some webpages do not send AXLoadComplete
https://bugs.webkit.org/show_bug.cgi?id=26995

11:02 AM Changeset in webkit [45601] by pdherbemont@apple.com
  • 4 edits
    4 adds in trunk

WebCore:

2009-07-06 Pierre d'Herbemont <pdherbemont@apple.com>

Reviewed by Simon Fraser.

https://bugs.webkit.org/show_bug.cgi?id=27015

Make sure that the CSS properties letter-spacing, word-spacing,
line-height, text-transform, text-indent, text-shadow,
text-decoration and color do not affect the media element controls,
that display text.

Controls that display text are only present in when the theme
MediaControllerThemeQT is being used.

Test: media/controls-styling.html

  • css/mediaControlsQT.css:
  • rendering/MediaControlElements.cpp: (WebCore::MediaControlElement::updateStyle): Special case for text-decoration. text-decoration can't be overriden from CSS, because text-decoration is additive for historical reasons.

LayoutTests:

2009-07-06 Pierre d'Herbemont <pdherbemont@apple.com>

Reviewed by Simon Fraser.

Make sure overriding properties won't affect the media controller.

  • media/controls-styling.html: Added.
10:54 AM Changeset in webkit [45600] by mrowe@apple.com
  • 2 edits in trunk/JavaScriptCore

Bug 27025: Crashes and regression test failures related to regexps in 64-bit
Fix <https://bugs.webkit.org/show_bug.cgi?id=27025> / <rdar://problem/7033448>.

Reviewed by Darin Adler.

For x86_64 RegexGenerator uses rbx, a callee-save register, as a scratch register but
neglects to save and restore it. The change in handling of the output vector in r45545
altered code generation so that the RegExp::match was now storing important data in rbx,
which caused crashes and bogus results when it was clobbered.

  • yarr/RegexJIT.cpp:

(JSC::Yarr::RegexGenerator::generateEnter): Save rbx.
(JSC::Yarr::RegexGenerator::generateReturn): Restore rbx.

10:30 AM Changeset in webkit [45599] by dimich@chromium.org
  • 2 edits in trunk/WebCore

2009-07-07 Albert Wong <ajwong@chromium.org>

Not reviewed, Chromium build fix.

Add in missing header + fix syntax issue that crept into last patch.
https://bugs.webkit.org/show_bug.cgi?id=27027

  • rendering/RenderThemeChromiumMac.mm: (WebCore::RenderThemeChromiumMac::paintMediaFullscreenButton):
10:15 AM Changeset in webkit [45598] by Simon Fraser
  • 9 edits in trunk/WebCore

2009-07-07 Simon Fraser <Simon Fraser>

Reviewed by Dave Hyatt.

Make use of geometry information to decide which layers become composited.

<rdar://problem/7011947>
<https://bugs.webkit.org/show_bug.cgi?id=27021>

In addition to looking at painting order, also, optionally, take layer
overlap into account when deciding which RenderLayers need to be composited.

No testcase because DRT doesn't dump which layers are composited.

  • page/FrameView.h:
  • page/FrameView.cpp: (WebCore::FrameView::updateCompositingLayers): Removed the unused CompositingUpdate parameter, and just do an early return when there is no view.
  • rendering/RenderLayer.h:
  • rendering/RenderLayer.cpp: (WebCore::RenderLayer::rendererContentChanged): (WebCore::RenderLayer::dirtyZOrderLists): (WebCore::RenderLayer::dirtyNormalFlowList): (WebCore::RenderLayer::styleChanged): (WebCore::RenderLayer::mustOverlapCompositedLayers): (WebCore::RenderLayer::setMustOverlapCompositedLayers): Change "overlay" to "overlap" for consistency with the use of "overlap" elsewhere.
  • rendering/RenderLayerBacking.cpp: (WebCore::RenderLayerBacking::updateAfterLayout): compositingLayersNeedUpdate() renamed to compositingLayersNeedRebuild().

(WebCore::RenderLayerBacking::startAnimation):
(WebCore::RenderLayerBacking::startTransition):
Tell the compositor that we started an accelerated animation or transition.

  • rendering/RenderLayerCompositor.h: (WebCore::RenderLayerCompositor::compositingLayersNeedRebuild): Renamed, since it's explicitly about changes in hierarchy now.

(WebCore::RenderLayerCompositor::setCompositingConsultsOverlap):
(WebCore::RenderLayerCompositor::compositingConsultsOverlap):
New getter and setter for whether compositing should take overlap into account.

  • rendering/RenderLayerCompositor.cpp: (WebCore::RenderLayerCompositor::RenderLayerCompositor): Add m_compositingConsultsOverlap, re-order some member variables.

(WebCore::RenderLayerCompositor::cacheAcceleratedCompositingEnabledFlag):
Whitespace cleanup and method rename.

(WebCore::RenderLayerCompositor::setCompositingLayersNeedRebuild):
Method and variable renaming.

(WebCore::RenderLayerCompositor::updateCompositingLayers):
Now we have to run through the layer hierarchy every time if looking
at overlap, because changes in layout can now alter compositing behavior.
We minimize work by tracking whether we actually need to change the layer
hierarchy via needLayerRebuild.

(WebCore::RenderLayerCompositor::updateBacking):
3D transforms turn off overlap mode.

(WebCore::RenderLayerCompositor::layerWasAdded):
(WebCore::RenderLayerCompositor::layerWillBeRemoved):
Method rename.

(WebCore::RenderLayerCompositor::addToOverlapMap):
(WebCore::RenderLayerCompositor::overlapsCompositedLayers):
Utility methods to add a layer to the overlapMap (computing the absolute
bounding box only if we haven't already done so), and testing the map
entries.

(WebCore::RenderLayerCompositor::computeCompositingRequirements):
If a layer would composite only because it comes after other compositing
layers in the painting order, then consult the overlap map to determine whether
it overlaps, and thus actually needs to composite.

Add layers to the map when they must be composited.

(WebCore::RenderLayerCompositor::rebuildCompositingLayerTree):
Only do re-parenting work if the updateHierarchy flag is set.

(WebCore::RenderLayerCompositor::didStartAcceleratedAnimation):
When we start an accelerated transition or animation, we stop looking at
overlap because we can no longer guarantee correct front-to-back ordering while the
accelerated animation is running.

(WebCore::RenderLayerCompositor::needsToBeComposited):
Method renames.

  • rendering/RenderView.cpp: (WebCore::RenderView::setMaximalOutlineSize): Add comment indicating that this could be optimized.
9:52 AM Changeset in webkit [45597] by ddkilzer@apple.com
  • 4 edits
    35 adds in trunk/BugsSite

Updated BugsSite with local changes on server

The new files were created by checksetup.pl during the upgrade
to Bugzilla 3.2.3. The rest are self-explanatory.

  • data: Added "bugzilla-update.xml" to svn:ignore.
  • data/attachments: Added "groups.*" to svn:ignore.
  • data/params: Updated for a few changed parameters.
  • lib/.htaccess: Added.
  • skins/contrib/Dusk/IE-fixes.css: Added.
  • skins/contrib/Dusk/admin.css: Added.
  • skins/contrib/Dusk/create_attachment.css: Added.
  • skins/contrib/Dusk/dependency-tree.css: Added.
  • skins/contrib/Dusk/duplicates.css: Added.
  • skins/contrib/Dusk/editusers.css: Added.
  • skins/contrib/Dusk/help.css: Added.
  • skins/contrib/Dusk/index.css: Added.
  • skins/contrib/Dusk/panel.css: Added.
  • skins/contrib/Dusk/params.css: Added.
  • skins/contrib/Dusk/release-notes.css: Added.
  • skins/contrib/Dusk/show_bug.css: Added.
  • skins/contrib/Dusk/show_multiple.css: Added.
  • skins/contrib/Dusk/summarize-time.css: Added.
  • skins/contrib/Dusk/voting.css: Added.
  • skins/contrib/Dusk/yui: Added.
  • skins/contrib/Dusk/yui/calendar.css: Added.
  • skins/custom/IE-fixes.css: Added.
  • skins/custom/admin.css: Added.
  • skins/custom/buglist.css: Added.
  • skins/custom/create_attachment.css: Added.
  • skins/custom/dependency-tree.css: Added.
  • skins/custom/duplicates.css: Added.
  • skins/custom/editusers.css: Added.
  • skins/custom/help.css: Added.
  • skins/custom/panel.css: Added.
  • skins/custom/params.css: Added.
  • skins/custom/release-notes.css: Added.
  • skins/custom/show_bug.css: Added.
  • skins/custom/show_multiple.css: Added.
  • skins/custom/summarize-time.css: Added.
  • skins/custom/voting.css: Added.
  • skins/custom/yui: Added.
  • skins/custom/yui/calendar.css: Added.
9:16 AM Changeset in webkit [45596] by darin@chromium.org
  • 3 edits in trunk/WebCore

2009-07-07 Anton Muhin <antonm@chromium.org>

Reviewed by Darin Fisher.

Speed up creation of V8 wrappers for DOM nodes.

https://bugs.webkit.org/show_bug.cgi?id=26882

This patch doesn't require new tests as it a set of refactorings
to speed up wrapper creation.

  • bindings/v8/V8Proxy.cpp:
  • bindings/v8/V8Proxy.h:
8:51 AM Changeset in webkit [45595] by jmalonzo@webkit.org
  • 5 edits in trunk

2009-07-07 Jan Michael Alonzo <jmalonzo@webkit.org>

Reviewed by Gustavo Noronha.

[GTK] textarea height property works only if other property are defined
https://bugs.webkit.org/show_bug.cgi?id=18984

Let WebCore handle textarea's metrics instead of readjusting it RenderthemeGtk.

Test: fast/forms/textarea-metrics.html

  • platform/gtk/RenderThemeGtk.cpp:
  • platform/gtk/RenderThemeGtk.h:
  • LayoutTests/platform/gtk/Skipped:
7:29 AM Changeset in webkit [45594] by jmalonzo@webkit.org
  • 6 edits
    3 adds in trunk

2009-07-07 Ben Murdoch <benm@google.com>

Reviewed by Antti Koivisto.

HTML5 Database becomes locked if a transaction is in progress when the page is refreshed.
https://bugs.webkit.org/show_bug.cgi?id=25711

Tests fix for https://bugs.webkit.org/show_bug.cgi?id=25711, where
web storage databases could become locked until the browser is
shut down if the page is refreshed whilst a transaction is in
progress.

  • storage/database-lock-after-reload-expected.txt: Added.
  • storage/database-lock-after-reload.html: Added.
  • storage/resources/database-lock-after-reload-2.html: Added.

Fix for https://bugs.webkit.org/show_bug.cgi?id=25711 where web
storage databases could become locked until the browser is shut
down if the page is refreshed whilst a transaction is in progress.

Test: storage/database-lock-after-reload.html

  • storage/Database.cpp: (WebCore::Database::Database): (WebCore::Database::close): add code to inform the database thread we've closed the database. (WebCore::Database::performOpenAndVerify): add code to inform the database thread we've opened a database.
  • storage/Database.h: (WebCore::Database::opened): return true iff the underlying sqlite database has been opened but not closed.
  • storage/DatabaseThread.cpp: (WebCore::DatabaseThread::databaseThread): Before the database thread terminates, close any databases that ran transactions in this thread. (WebCore::DatabaseThread::recordDatabaseOpen): Records a database that executed a transaction in this thread. (WebCore::DatabaseThread::recordDatabaseClosed): Removes a database from the set of open databases.
  • storage/DatabaseThread.h: (WebCore::DatabaseThread::getThreadID): return the thread id for the database thread.
7:04 AM Changeset in webkit [45593] by Simon Hausmann
  • 6 edits in trunk

WebKit/qt:

2009-07-07 Simon Hausmann <Simon Hausmann>

Reviewed by Holger Freyther.

Add Qt DRT hook for clearing the frame name.

  • Api/qwebframe.cpp: (qt_drt_clearFrameName):

WebKitTools:

2009-07-07 Simon Hausmann <Simon Hausmann>

Reviewed by Holger Freyther.

Clear the main frame's name between loading pages, like in r36652. This
reduces the side-effects between test cases.

  • DumpRenderTree/qt/DumpRenderTree.cpp: (WebCore::DumpRenderTree::open):

LayoutTests:

2009-07-07 Simon Hausmann <Simon Hausmann>

Reviewed by Holger Freyther.

Removed subframe-self-close.html from the Qt skip list.

  • platform/qt/Skipped:
5:52 AM Changeset in webkit [45592] by jmalonzo@webkit.org
  • 2 edits
    1 add in trunk/WebCore

2009-07-07 Jiahua Huang <jhuangjiahua@gmail.com>

Reviewed by Jan Alonzo.

[Gtk] Paste of rich text from firefox results garbled markup
https://bugs.webkit.org/show_bug.cgi?id=26791

Fix problem with UTF-16 clipboard pasted.

  • manual-tests/gtk/paste-richtext-from-firefox.html: Added.
  • platform/gtk/PasteboardGtk.cpp: (WebCore::Pasteboard::documentFragment):
5:35 AM Changeset in webkit [45591] by pfeldman@chromium.org
  • 2 edits in trunk/WebCore

2009-07-07 Yury Semikhatsky <yurys@chromium.org>

Reviewed by Timothy Hatcher.

Change 'Continue debug' shortcut from F5 to F8 for consistency with Firebug.

https://bugs.webkit.org/show_bug.cgi?id=23849

  • inspector/front-end/ScriptsPanel.js: (WebInspector.ScriptsPanel):
3:48 AM Changeset in webkit [45590] by Simon Hausmann
  • 2 edits in trunk/LayoutTests

2009-07-07 Simon Hausmann <Simon Hausmann>

Reviewed by Holger Freyther.

Added all current failing Qt DRT tests to the skip list, for a fresh start.

  • platform/qt/Skipped:
3:27 AM Changeset in webkit [45589] by Simon Hausmann
  • 2 edits in trunk/WebKitTools

2009-07-06 Simon Hausmann <Simon Hausmann>

Reviewed by Holger Freyther.

Fix various layout tests with Qt's DRT.

Remember to set the focus on the web page before
loading.

  • DumpRenderTree/qt/DumpRenderTree.cpp: (WebCore::DumpRenderTree::open):
3:25 AM Changeset in webkit [45588] by Simon Hausmann
  • 3 edits in trunk/WebKitTools

2009-07-07 Simon Hausmann <Simon Hausmann>

Reviewed by Tor Arne Vestbø and Holger Freyther.

Fix timeout timer handling in Qt DRT.

Use QBasicTimer instead of startTimer/killTimer, to fix the
problem that starting the timeout timer did not stop a
previously started timer, causing multiple emissions of timeout.
The timerEvent() implementation now also checks the timer id, to
protect against double timeouts.

  • DumpRenderTree/qt/jsobjects.cpp: (LayoutTestController::LayoutTestController): (LayoutTestController::reset): (LayoutTestController::waitUntilDone): (LayoutTestController::notifyDone): (LayoutTestController::timerEvent):
  • DumpRenderTree/qt/jsobjects.h:

Jul 6, 2009:

11:42 PM Changeset in webkit [45587] by ggaren@apple.com
  • 5 edits in branches/nitro-extreme

JavaScriptCore:

2009-07-06 Geoffrey Garen <ggaren@apple.com>

Reviewed by Sam Weinig (?).


Fixed an assertion seen during the stress test.


Don't assume that, if op1 is constant, op2 is not, and vice versa. Sadly,
not all constants get folded.

  • jit/JITArithmetic.cpp: (JSC::JIT::emit_op_jnless): (JSC::JIT::emitSlow_op_jnless): (JSC::JIT::emit_op_jnlesseq): (JSC::JIT::emitSlow_op_jnlesseq):

LayoutTests:

2009-07-06 Geoffrey Garen <ggaren@apple.com>

Reviewed by Sam Weinig (?).

Test for an assertion seen during the stress test.


  • fast/js/arithmetic-expected.txt:
  • fast/js/resources/arithmetic.js:
10:34 PM Changeset in webkit [45586] by barraclough@apple.com
  • 16 edits in branches/nitro-extreme

JavaScriptCore:

2009-07-06 Gavin Barraclough <barraclough@apple.com>

Reviewed by Sam Weinig.

Stop loading constants into the register file.

Instead, use high register values (highest bit bar the sign bit set) to indicate
constants in the instruction stream, and when we encounter such a value load it
directly from the CodeBlock.

Since constants are no longer copied into the register file, this patch renders
the 'unexpected constant' mechanism redundant, and removes it.

2% improvement, thanks to Sam Weinig.

  • bytecode/CodeBlock.cpp: (JSC::CodeBlock::dump): (JSC::CodeBlock::CodeBlock): (JSC::CodeBlock::mark): (JSC::CodeBlock::shrinkToFit):
  • bytecode/CodeBlock.h: (JSC::CodeBlock::isTemporaryRegisterIndex): (JSC::CodeBlock::constantRegister): (JSC::CodeBlock::isConstantRegisterIndex): (JSC::CodeBlock::getConstant): (JSC::ExecState::r):
  • bytecode/Opcode.h:
  • bytecompiler/BytecodeGenerator.cpp: (JSC::BytecodeGenerator::preserveLastVar): (JSC::BytecodeGenerator::BytecodeGenerator): (JSC::BytecodeGenerator::addConstantValue): (JSC::BytecodeGenerator::emitEqualityOp): (JSC::BytecodeGenerator::emitLoad): (JSC::BytecodeGenerator::emitResolveBase): (JSC::BytecodeGenerator::emitResolveWithBase): (JSC::BytecodeGenerator::emitNewError):
  • bytecompiler/BytecodeGenerator.h: (JSC::BytecodeGenerator::emitNode):
  • interpreter/CallFrame.h: (JSC::ExecState::noCaller): (JSC::ExecState::hasHostCallFrameFlag): (JSC::ExecState::addHostCallFrameFlag): (JSC::ExecState::removeHostCallFrameFlag):
  • interpreter/Interpreter.cpp: (JSC::Interpreter::resolve): (JSC::Interpreter::resolveSkip): (JSC::Interpreter::resolveGlobal): (JSC::Interpreter::resolveBase): (JSC::Interpreter::resolveBaseAndProperty): (JSC::Interpreter::resolveBaseAndFunc): (JSC::Interpreter::dumpRegisters): (JSC::Interpreter::throwException): (JSC::Interpreter::createExceptionScope): (JSC::Interpreter::privateExecute): (JSC::Interpreter::retrieveArguments):
  • jit/JIT.cpp: (JSC::JIT::privateCompileMainPass):
  • jit/JITInlineMethods.h: (JSC::JIT::emitLoadDouble): (JSC::JIT::emitLoadInt32ToDouble):
  • jit/JITOpcodes.cpp: (JSC::JIT::emit_op_new_error): (JSC::JIT::emit_op_enter): (JSC::JIT::emit_op_enter_with_activation):
  • parser/Nodes.cpp: (JSC::DeleteResolveNode::emitBytecode): (JSC::DeleteValueNode::emitBytecode): (JSC::PrefixResolveNode::emitBytecode):
  • runtime/JSActivation.cpp: (JSC::JSActivation::JSActivation):
  • wtf/Platform.h:

LayoutTests:

2009-07-06 Gavin Barraclough <barraclough@apple.com>

Reviewed by Geoff Garen.

fast/js/global-recursion-on-full-stack.html is a little finicky.

The test recurses down the JS stack to find the bottom (catching this with an exception),
then tries to call a host function (document.write), switch writes new '<script>' code,
and expects this code to be run, then expects this code to try to call 'f();' again,
which it expects to fail, and it expects to catch that exception. However it is possible
that one of the earlier stages (the call to document.write, entering the interpreter to
run the new global code) will fail, depending on exactly how much stack space was free at
the point the last call to f() failed.

Tweak the test to make it work.

  • fast/js/global-recursion-on-full-stack.html:
8:40 PM Changeset in webkit [45585] by eric@webkit.org
  • 3 edits in trunk/WebCore

2009-07-06 Albert J. Wong <ajwong@chromium.org>

Reviewed by Maciej Stachowiak.

Update RenderThemeChromiumMac for wkDrawMediaUIPart and wkDrawMediaSlider API
change.

Fix compile RenderThemeChromiumMac.mm due to API changes from r45572
https://bugs.webkit.org/show_bug.cgi?id=27018

This is mainly an API change update to fix the chromium compile. It
also has a small fix of a typo in RenderThemeMac.mm.

Compilation against the head of chromium trunk passed. Since this is
a compile fix, that should be sufficient.

  • rendering/RenderThemeChromiumMac.mm: (WebCore::): (WebCore::RenderThemeChromiumMac::paintMediaFullscreenButton): (WebCore::RenderThemeChromiumMac::paintMediaMuteButton): (WebCore::RenderThemeChromiumMac::paintMediaPlayButton): (WebCore::RenderThemeChromiumMac::paintMediaSeekBackButton): (WebCore::RenderThemeChromiumMac::paintMediaSeekForwardButton): (WebCore::RenderThemeChromiumMac::paintMediaSliderTrack): (WebCore::RenderThemeChromiumMac::paintMediaSliderThumb):
  • rendering/RenderThemeMac.mm: (WebCore::):
7:49 PM Changeset in webkit [45584] by ddkilzer@apple.com
  • 2 edits in trunk/WebKit/mac

Bug 27006: Build fix when MAC_JAVA_BRIDGE is disabled

<https://bugs.webkit.org/show_bug.cgi?id=27006>

Reviewed by Darin Adler.

  • WebCoreSupport/WebFrameLoaderClient.mm: (WebFrameLoaderClient::createJavaAppletWidget): Wrap contents of the method in #if ENABLE(MAC_JAVA_BRIDGE)/#endif and return 0 when this feature is disabled.
7:44 PM Changeset in webkit [45583] by ddkilzer@apple.com
  • 3 edits in trunk/WebKit/mac

Bug 27005: Build fixes when NETSCAPE_PLUGIN_API is disabled

<https://bugs.webkit.org/show_bug.cgi?id=27005>

Reviewed by Geoff Garen.

  • Plugins/WebPluginController.mm: (-[WebPluginController destroyPlugin:]): Wrap call to ScriptController::cleanupScriptObjectsForPlugin() in #if ENABLE(NETSCAPE_PLUGIN_API)/#endif macro. (-[WebPluginController destroyAllPlugins]): Ditto.
  • Plugins/WebPluginDatabase.mm: (-[WebPluginDatabase removePluginInstanceViewsFor:]): Wrap WebBaseNetscapePluginView class checks in #if ENABLE(NETSCAPE_PLUGIN_API)/#endif macro. (-[WebPluginDatabase destroyAllPluginInstanceViews]): Ditto.
7:39 PM Changeset in webkit [45582] by ddkilzer@apple.com
  • 4 edits in trunk/WebKit/mac

BUILD FIX: Use ENABLE(NETSCAPE_PLUGIN_API) instead of USE(PLUGIN_HOST_PROCESS)

In r45579, #if/#endif macros for USE(PLUGIN_HOST_PROCESS) were
used, but ENABLE(NETSCAPE_PLUGIN_API) should have been used
instead.

  • Plugins/WebNetscapeContainerCheckContextInfo.h:
  • Plugins/WebNetscapeContainerCheckContextInfo.mm:
  • Plugins/WebNetscapeContainerCheckPrivate.mm:
7:18 PM Changeset in webkit [45581] by ddkilzer@apple.com
  • 2 edits in trunk/WebKit/mac

Bug 27004: Build fix for ENABLE(PLUGIN_PROXY_FOR_VIDEO) after r42618

<https://bugs.webkit.org/show_bug.cgi?id=27004>

Reviewed by Geoff Garen.

  • Plugins/WebPluginController.mm: (mediaProxyClient): Use core() method to convert from DOMElement to WebCore::Element.
7:18 PM Changeset in webkit [45580] by eric@webkit.org
  • 3 edits in trunk/WebKitTools

2009-07-06 Eric Seidel <eric@webkit.org>

Reviewed by Maciej Stachowiak.

bugzilla-tool fails to close bugs with bugzilla 3.0
https://bugs.webkit.org/show_bug.cgi?id=27008

Update bugzilla.py to match changes in the bug page "changeform"
Update scm.py to add bug_ids to attachments to make error reporting nicer.

  • Scripts/modules/bugzilla.py:
  • Scripts/modules/scm.py:
7:14 PM Changeset in webkit [45579] by ddkilzer@apple.com
  • 4 edits in trunk/WebKit/mac

Bug 27003: Build fix when USE(PLUGIN_HOST_PROCESS) is disabled

<https://bugs.webkit.org/show_bug.cgi?id=27003>

Reviewed by Geoff Garen.

  • Plugins/WebNetscapeContainerCheckContextInfo.h: Added #if USE(PLUGIN_HOST_PROCESS)/#endif guards.
  • Plugins/WebNetscapeContainerCheckContextInfo.mm: Ditto.
  • Plugins/WebNetscapeContainerCheckPrivate.mm: Ditto.
6:01 PM Changeset in webkit [45578] by ddkilzer@apple.com
  • 2 edits in trunk/WebCore

Bug 27002: Build fix when DASHBOARD_SUPPORT is disabled with -Wunused-parameter

<https://bugs.webkit.org/show_bug.cgi?id=27002>

Reviewed by Geoff Garen.

  • html/CanvasRenderingContext2D.cpp: (WebCore::CanvasRenderingContext2D::prepareGradientForDashboard): Mark gradient as an unused parameter when DASHBOARD_SUPPORT is disabled.
5:59 PM Changeset in webkit [45577] by ddkilzer@apple.com
  • 2 edits in trunk/WebCore

Bug 27001: Fix improper use of PassRefPtr<Node> to RefPtr<Node>

<https://bugs.webkit.org/show_bug.cgi?id=27001>

Reviewed by Geoff Garen.

PassRefPtr<> should only be used for arguments to functions that
take ownership of the object, or as return values from functions
that relinquish ownership of the object.

  • editing/Editor.cpp: (WebCore::Editor::increaseSelectionListLevelOrdered): Changed stack-allocated PassRefPtr<Node> to RefPtr<Node> and call release() on returned object. (WebCore::Editor::increaseSelectionListLevelUnordered): Ditto.
5:44 PM Changeset in webkit [45576] by eric.carlson@apple.com
  • 4 edits in trunk

2009-07-06 Eric Carlson <eric.carlson@apple.com>

Reviewed by Simon Fraser.

<rdar://problem/7035474> Make new media controller UI default on SnowLeopard

  • rendering/RenderThemeMac.mm: (WebCore::mediaControllerTheme):

New media controller UI is enabled by default on SnowLeopard.

2009-07-06 Eric Carlson <eric.carlson@apple.com>

Reviewed by Simon Fraser.

<rdar://problem/7035474> Make new media controller UI default on SnowLeopard

Disable media tests that depend on controller size/position on SnowLeopard until
<rdar://problem/6502936> Update run-webkit-tests to deal with new media controller look
has been fixed

  • platform/mac-snowleopard/Skipped:
4:54 PM Changeset in webkit [45575] by ggaren@apple.com
  • 4 edits in branches/nitro-extreme/JavaScriptCore

2009-07-06 Geoffrey Garen <ggaren@apple.com>

Reviewed by Sam Weinig.


Include op_convert_this in result caching.


No change on SunSpider or v8.

  • jit/JITOpcodes.cpp: (JSC::JIT::emit_op_convert_this):
  • jit/JITStubs.cpp: (JSC::DEFINE_STUB_FUNCTION):
  • jit/JITStubs.h: (JSC::): Made the op_convert_this JIT stub return an EncodedJSValue, so to maintain the result caching contract that { tag, payload } can be found in { regT1, regT0 }.
4:18 PM Changeset in webkit [45574] by Nate Chapin
  • 17 edits
    1 delete in trunk/WebCore

2009-07-06 Nate Chapin <Nate Chapin>

Unreviewed, build fix.

Roll out r45559 to fix Chromium canary.

https://bugs.webkit.org/show_bug.cgi?id=26857

3:59 PM Changeset in webkit [45573] by ggaren@apple.com
  • 10 edits in branches/nitro-extreme/JavaScriptCore

2009-07-06 Geoffrey Garen <ggaren@apple.com>

Reviewed by Sam Weinig.


Implemented result chaining.


1% faster on SunSpider. 4%-5% faster on v8.

  • assembler/MacroAssemblerX86Common.h: (JSC::MacroAssemblerX86Common::move):
  • assembler/X86Assembler.h: (JSC::X86Assembler::movl_rr): Added an optimization to eliminate no-op mov instructions, to simplify chaining.
  • jit/JIT.cpp: (JSC::JIT::JIT):
  • jit/JIT.h: Added data members and helper functions for recording chained results. We record both a mapping from virtual to machine register and the opcode for which the mapping is valid, to help ensure that the mapping isn't used after the mapped register has been stomped by other instructions.
  • jit/JITCall.cpp: (JSC::JIT::compileOpCallVarargs): (JSC::JIT::compileOpCallVarargsSlowCase): (JSC::JIT::emit_op_ret): (JSC::JIT::emit_op_construct_verify): (JSC::JIT::compileOpCall): (JSC::JIT::compileOpCallSlowCase): Chain function call results.
  • jit/JITInlineMethods.h: (JSC::JIT::emitLoadTag): (JSC::JIT::emitLoadPayload): (JSC::JIT::emitLoad): (JSC::JIT::emitLoad2): (JSC::JIT::isLabeled): (JSC::JIT::map): (JSC::JIT::unmap): (JSC::JIT::isMapped): (JSC::JIT::getMappedPayload): (JSC::JIT::getMappedTag): Use helper functions when loading virtual registers into machine registers, in case the loads can be eliminated by chaining.
  • jit/JITOpcodes.cpp: (JSC::JIT::emit_op_mov): (JSC::JIT::emit_op_end): (JSC::JIT::emit_op_instanceof): (JSC::JIT::emit_op_get_global_var): (JSC::JIT::emit_op_put_global_var): (JSC::JIT::emit_op_get_scoped_var): (JSC::JIT::emit_op_put_scoped_var): (JSC::JIT::emit_op_to_primitive): (JSC::JIT::emit_op_resolve_global): (JSC::JIT::emit_op_jneq_ptr): (JSC::JIT::emit_op_next_pname): (JSC::JIT::emit_op_to_jsnumber): (JSC::JIT::emit_op_catch): Chain results from these opcodes.

(JSC::JIT::emit_op_profile_will_call):
(JSC::JIT::emit_op_profile_did_call): Load the profiler into regT2 to
avoid stomping a chained result.

  • jit/JITPropertyAccess.cpp: (JSC::JIT::emit_op_method_check): (JSC::JIT::emit_op_get_by_val): (JSC::JIT::emit_op_get_by_id): Chain results from these opcodes.
  • jit/JITStubCall.h: (JSC::JITStubCall::addArgument): Always use { regT1, regT0 }, to facilitate chaining.

(JSC::JITStubCall::call): Unmap all mapped registers, since our callee
stub might stomp them.

3:56 PM Changeset in webkit [45572] by eric.carlson@apple.com
  • 13 edits in trunk

2009-07-06 Eric Carlson <eric.carlson@apple.com>

Reviewed by Darin Adler.

<rdar://problem/7008093> Media controller can’t be used to scrub when movie is
narrow — track is too narrow

Do not show media controller time display elements when the a movie is too narrow.

  • platform/mac/WebCoreSystemInterface.h:
  • platform/mac/WebCoreSystemInterface.mm:

The 'state' parameter to wkDrawMediaUIPart is now an unsigned bitfield.

  • rendering/MediaControlElements.cpp: (WebCore::MediaControlTimeDisplayElement::MediaControlTimeDisplayElement):

Give current time and time remaining controls a common base class.

(WebCore::MediaControlTimeDisplayElement::setVisible):

New method, hide and show the element.

  • rendering/MediaControlElements.h:
  • rendering/RenderMedia.cpp: (WebCore::RenderMedia::RenderMedia):

Initialize m_previousVisible.

(WebCore::RenderMedia::layout):

Show/hide the time display elements as the movie width changes.

(WebCore::RenderMedia::createCurrentTimeDisplay):
(WebCore::RenderMedia::createTimeRemainingDisplay):

Base class is now MediaControlTimeDisplayElement, not MediaControlElement.

(WebCore::RenderMedia::shouldShowTimeDisplayControls):

New, decide if time display elements should be visible or not.

  • rendering/RenderMedia.h:
  • rendering/RenderThemeMac.mm: (WebCore::): (WebCore::RenderThemeMac::paintMediaFullscreenButton): (WebCore::RenderThemeMac::paintMediaMuteButton): (WebCore::RenderThemeMac::paintMediaPlayButton): (WebCore::RenderThemeMac::paintMediaSeekBackButton): (WebCore::RenderThemeMac::paintMediaSeekForwardButton): (WebCore::RenderThemeMac::paintMediaSliderTrack): (WebCore::RenderThemeMac::paintMediaSliderThumb): (WebCore::RenderThemeMac::paintMediaRewindButton): (WebCore::RenderThemeMac::paintMediaReturnToRealtimeButton): (WebCore::RenderThemeMac::paintMediaControlsBackground): (WebCore::RenderThemeMac::paintMediaCurrentTime): (WebCore::RenderThemeMac::paintMediaTimeRemaining):

The 'state' parameter to wkDrawMediaUIPart is now an unsigned bitfield.

2009-07-06 Eric Carlson <eric.carlson@apple.com>

Update WebKitSystemInterface for <rdar://problem/7008093>.

  • WebKitSystemInterface.h:
  • libWebKitSystemInterfaceLeopard.a:
  • libWebKitSystemInterfaceSnowLeopard.a:
  • libWebKitSystemInterfaceTiger.a:
3:46 PM Changeset in webkit [45571] by ddkilzer@apple.com
  • 3 edits in trunk/WebCore

Bug 27000: Minor clean up to runtime_root.{cpp|h}

<https://bugs.webkit.org/show_bug.cgi?id=27000>

Reviewed by Geoff Garen.

  • bridge/runtime_root.cpp: Added blank line between license and the first #include statement.
  • bridge/runtime_root.h: Added comment to #endif.
3:42 PM Changeset in webkit [45570] by eric@webkit.org
  • 9 edits in trunk/WebCore

2009-07-06 Roland Steiner <rolandsteiner@google.com>

Reviewed by Maciej Stachowiak.

generalize the special height treatment for SVG (to be re-used for ruby):

renamed InlineBox::m_isSVG to m_hasVirtualHeight
renamed InlineBox::isSVG() to hasVirtualHeight()
renamed InlineBox::setIsSVG() to setHasVirtualHeight()

  • rendering/InlineBox.cpp: (WebCore::InlineBox::height):
  • rendering/InlineBox.h: (WebCore::InlineBox::InlineBox): (WebCore::InlineBox::isText): (WebCore::InlineBox::setIsText): (WebCore::InlineBox::isSVGRootInlineBox): (WebCore::InlineBox::hasVirtualHeight): (WebCore::InlineBox::setHasVirtualHeight): (WebCore::InlineBox::virtualHeight):
  • rendering/RenderSVGInline.cpp: (WebCore::RenderSVGInline::createFlowBox):
  • rendering/RenderSVGInlineText.cpp: (WebCore::RenderSVGInlineText::createTextBox):
  • rendering/RenderSVGText.cpp: (WebCore::RenderSVGText::createRootBox):
  • rendering/SVGInlineFlowBox.h: (WebCore::SVGInlineFlowBox::virtualHeight):
  • rendering/SVGInlineTextBox.h: (WebCore::SVGInlineTextBox::virtualHeight):
  • rendering/SVGRootInlineBox.h: (WebCore::SVGRootInlineBox::virtualHeight):
3:11 PM Changeset in webkit [45569] by alice.liu@apple.com
  • 14 edits in trunk/WebCore

2009-07-06 Alice Liu <alice.liu@apple.com>

Reviewed by Darin Adler.

REGRESSION(r45285): focus rings are black on windows safari
https://bugs.webkit.org/show_bug.cgi?id=26821
<rdar://problem/7018252>

Add a mechanism for setting a custom focus ring color, and relocate the
determination of focus ring color to base class RenderTheme.

No new tests. Already-existing tests run in pixel mode would have caught this regression

  • css/CSSStyleSelector.cpp: (WebCore::CSSStyleSelector::getColorFromPrimitiveValue): Move the responsibility of determining focus color to base class RenderTheme.
  • rendering/RenderTheme.cpp: (WebCore::customFocusRingColor): Added static local. (WebCore::RenderTheme::setCustomFocusRingColor): Added (WebCore::RenderTheme::focusRingColor): Use custom color over platform color.
  • rendering/RenderTheme.h: (WebCore::RenderTheme::platformFocusRingColor): default implementation, black color
  • rendering/RenderThemeSafari.cpp: (WebCore::RenderTheme::themeForPage): Circumstances that lead to returning the RenderThemeWin are the same under which we want to use the focus ring color from SafariTheme. (WebCore::RenderThemeSafari::platformFocusRingColor): Renamed from focusRingColor
  • rendering/RenderThemeSafari.h:

Renaming focusRingColor to platformFocusRingColor in these files:

  • rendering/RenderThemeChromiumMac.h:
  • rendering/RenderThemeChromiumMac.mm: (WebCore::RenderThemeChromiumMac::platformFocusRingColor):
  • rendering/RenderThemeChromiumSkia.cpp: (WebCore::RenderThemeChromiumSkia::platformFocusRingColor):
  • rendering/RenderThemeChromiumSkia.h:
  • rendering/RenderThemeMac.h:
  • rendering/RenderThemeMac.mm: (WebCore::RenderThemeMac::platformFocusRingColor):

Cleanup leftover from removal of WebCore::Color in r45285 in these files:

  • WebCore.order:
  • platform/graphics/Color.h:
2:53 PM Changeset in webkit [45568] by eric@webkit.org
  • 5 edits
    4 adds in trunk

2009-07-06 Shinichiro Hamaji <hamaji@chromium.org>

Reviewed by David Hyatt.

Bug 15135: REGRESSION (r19843-r19850): Changing a flexbox's
contents makes its container scroll to the top
https://bugs.webkit.org/show_bug.cgi?id=15135

Delay updateing scroll bar of descendants of flexbox until their
positions are determined. In this way we can prevent descendants
of flexible boxes from changing positions of their scrollbars
using tentative positions.

  • fast/flexbox/repaint-scrollbar-expected.txt: Added.
  • fast/flexbox/repaint-scrollbar.html: Added.
  • fast/flexbox/resources/repaint-scrollbar.js: Added.

2009-07-06 Shinichiro Hamaji <hamaji@chromium.org>

Reviewed by David Hyatt.

Bug 15135: REGRESSION (r19843-r19850): Changing a flexbox's
contents makes its container scroll to the top
https://bugs.webkit.org/show_bug.cgi?id=15135

Delay updateing scroll bar of descendants of flexbox until their
positions are determined. In this way we can prevent descendants
of flexible boxes from changing positions of their scrollbars
using tentative positions.

Test: fast/flexbox/repaint-scrollbar.html

  • rendering/RenderBlock.cpp: (WebCore::RenderBlock::startDelayUpdateScrollInfo): (WebCore::RenderBlock::finishDelayUpdateScrollInfo): (WebCore::RenderBlock::updateScrollInfoAfterLayout): (WebCore::RenderBlock::layoutBlock): (WebCore::RenderBlock::layoutOnlyPositionedObjects):
  • rendering/RenderBlock.h:
  • rendering/RenderFlexibleBox.cpp: (WebCore::RenderFlexibleBox::layoutHorizontalBox): (WebCore::RenderFlexibleBox::layoutVerticalBox):
2:33 PM Changeset in webkit [45567] by eric@webkit.org
  • 3 edits
    2 adds in trunk

2009-07-06 Hironori Bono <hbono@chromium.org>

Reviewed by Maciej Stachowiak.

Make unconfirmed IME text affect textarea's value.
This matches input and contentEditable elements as well
IE and Firefox.

This fixes https://bugs.webkit.org/show_bug.cgi?id=25061.

Input elements would go down this code path because it
would always get a null compositionNode from frame->editor().
Special casing compositionNodes is wrong because we explicitly
want unconfirmed IME input in the textarea's value (assuming we
want to match IE and Firefox here).

This change is originally created by Ojan Vafai <ojan@chromium.org> and
I just changed its manual tests with an automated test on his behalf.

  • platform/mac/editing/input/text-control-ime-input-expected.txt: Added.
  • platform/mac/editing/input/text-control-ime-input.html: Added.

2009-07-06 Hironori Bono <hbono@chromium.org>

Reviewed by Maciej Stachowiak.

Make unconfirmed IME text affect textarea's value.
This matches input and contentEditable elements as well
IE and Firefox.

This fixes https://bugs.webkit.org/show_bug.cgi?id=25061.

Input elements would go down this code path because it
would always get a null compositionNode from frame->editor().
Special casing compositionNodes is wrong because we explicitly
want unconfirmed IME input in the textarea's value (assuming we
want to match IE and Firefox here).

This change is originally created by Ojan Vafai <ojan@chromium.org> and
I just changed its manual tests with an automated test on his behalf.

Test: platform/mac/editing/input/text-control-ime-input.html

  • rendering/RenderTextControl.cpp: (WebCore::RenderTextControl::text): (WebCore::RenderTextControl::textWithHardLineBreaks):
2:13 PM Changeset in webkit [45566] by adachan@apple.com
  • 2 edits in trunk/JavaScriptCore

Reviewed by Darin Adler and Mark Rowe.

Decommitted spans are added to the list of normal spans rather than
the returned spans in TCMalloc_PageHeap::Delete().
https://bugs.webkit.org/show_bug.cgi?id=26998


In TCMalloc_PageHeap::Delete(), the deleted span can be decommitted in
the process of merging with neighboring spans that are also decommitted.
The merged span needs to be placed in the list of returned spans (spans
whose memory has been returned to the system). Right now it's always added
to the list of the normal spans which can theoretically cause thrashing.

  • wtf/FastMalloc.cpp: (WTF::TCMalloc_PageHeap::Delete):
1:59 PM Changeset in webkit [45565] by andersca@apple.com
  • 9 edits in trunk

WebCore:

2009-07-06 Anders Carlsson <andersca@apple.com>

Reviewed by Adele Peterson.

Add the ability for wheel events to latch to a node.


  • WebCore.base.exp:
  • page/EventHandler.cpp: (WebCore::EventHandler::EventHandler): (WebCore::EventHandler::clear): (WebCore::EventHandler::handleWheelEvent):
  • page/EventHandler.h:
  • page/mac/EventHandlerMac.mm: (WebCore::EventHandler::wheelEvent):
  • platform/mac/WebCoreSystemInterface.h:
  • platform/mac/WebCoreSystemInterface.mm:

WebKit/mac:

2009-07-06 Anders Carlsson <andersca@apple.com>

Reviewed by Adele Peterson.

Initialize wkIsLatchingWheelEvent.


  • WebCoreSupport/WebSystemInterface.m: (InitWebCoreSystemInterface):
1:55 PM Changeset in webkit [45564] by Nate Chapin
  • 2 edits in trunk/WebCore

2009-07-06 Nate Chapin <Nate Chapin>

Reviewed by Dimitri Glazkov.

Fix a bunch of layout test crahses in Chromium caused by a bad usage of DEFINE_STATIC_LOCAL.

https://bugs.webkit.org/show_bug.cgi?id=26997

  • bindings/v8/V8Binding.cpp: (WebCore::v8ValueToWebCoreString): Use a regular static declaration instead of DEFINE_STATIC_LOCAL.
1:48 PM Changeset in webkit [45563] by andersca@apple.com
  • 5 edits in trunk/WebKitLibraries

2009-07-06 Anders Carlsson <andersca@apple.com>

Update WebKitSystemInterface.

  • WebKitSystemInterface.h:
  • libWebKitSystemInterfaceLeopard.a:
  • libWebKitSystemInterfaceSnowLeopard.a:
  • libWebKitSystemInterfaceTiger.a:
1:47 PM Changeset in webkit [45562] by eric@webkit.org
  • 3 edits in trunk/WebCore

2009-07-06 Pavel Feldman <pfeldman@chromium.org>

Reviewed by Timothy Hatcher.

WebInspector: suggest global properties based on async evaluation.

https://bugs.webkit.org/show_bug.cgi?id=26976

Before this change, empty string was evaluated to the global object
(or scope chain object) synchronously. This is now fixed and global
object is evaluated using the same control flow.

  • inspector/front-end/Console.js: (WebInspector.Console.prototype.completions): (WebInspector.Console.prototype._evalInInspectedWindow):
  • inspector/front-end/ScriptsPanel.js: (WebInspector.ScriptsPanel.prototype._variablesInScope):
1:37 PM Changeset in webkit [45561] by ojan@chromium.org
  • 5 edits
    2 adds in trunk

WebCore:

2009-07-06 Ojan Vafai <ojan@chromium.org>

Reviewed by Eric Seidel.

Fix crash when indenting in an empty table cell.
https://bugs.webkit.org/show_bug.cgi?id=26872

The crash is that we would call splitTreeToNode where the node
and the nodeToSplitTo were the same node.

Test: editing/execCommand/indent-empty-table-cell.html

  • editing/CompositeEditCommand.cpp: (WebCore::CompositeEditCommand::splitTreeToNode): Added an assert in that node and nodeToSplitTo are different nodes to make this assumption explicit.
  • editing/IndentOutdentCommand.cpp: (WebCore::IndentOutdentCommand::isAtUnsplittableElement): (WebCore::IndentOutdentCommand::indentRegion):
  • editing/IndentOutdentCommand.h:

LayoutTests:

2009-07-06 Ojan Vafai <ojan@chromium.org>

Reviewed by Eric Seidel.

Fix crash when indenting in an empty table cell.
https://bugs.webkit.org/show_bug.cgi?id=26872

The crash is that we would call splitTreeToNode where the node
and the nodeToSplitTo were the same node.

  • editing/execCommand/indent-empty-table-cell-expected.txt: Added.
  • editing/execCommand/indent-empty-table-cell.html: Added.
11:10 AM Changeset in webkit [45560] by cmarrin@apple.com
  • 2 edits in trunk/WebKitSite/specs/CSSVisualEffects/css3-2d-transforms

added pseudocode form decomposition method

10:38 AM Changeset in webkit [45559] by Nate Chapin
  • 17 edits
    1 add in trunk/WebCore

2009-07-06 Nate Chapin <Nate Chapin>

Reviewed by David Levin.

Upstream V8Binding and update some function names.

https://bugs.webkit.org/show_bug.cgi?id=26857

7:37 AM Changeset in webkit [45558] by kov@webkit.org
  • 5 edits in trunk

WebCore

2009-07-06 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>

Reviewed by Jan Alonzo.

Use soup's content sniffing
https://bugs.webkit.org/show_bug.cgi?id=26982

Drop our hackish content sniffing code, and use the new libsoup
feature to do that job for us.

Testing this is in a cross-platform way is not obvious or possible
while we are using platform-specific code for sniffing.

  • platform/network/ResourceHandleInternal.h: (WebCore::ResourceHandleInternal::ResourceHandleInternal):
  • platform/network/soup/ResourceHandleSoup.cpp: (WebCore::gotHeadersCallback): (WebCore::contentSniffedCallback): (WebCore::gotChunkCallback): (WebCore::ResourceHandle::startHttp):

WebKit/gtk

2009-07-06 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>

Reviewed by Jan Alonzo.

Use soup's content sniffing
https://bugs.webkit.org/show_bug.cgi?id=26982

Add the new Content Sniffing feature provided by libsoup.

  • webkit/webkitprivate.cpp: (webkit_init):

Jul 5, 2009:

10:22 PM Changeset in webkit [45557] by Darin Adler
  • 2 edits in trunk/WebCore

2009-07-05 Antonio Gomes <antonio.gomes@openbossa.org>

Reviewed by Darin Adler.

REGRESSION (r40499): fast/dom/cssTarget-crash.html fails
https://bugs.webkit.org/show_bug.cgi?id=20342

Re-added code removed by commit r40499.
Without this, both Qt and Mac were crashing while running the test.

Note that this does not entirely fix the bug. It fixes the WebCore
crash, but the test no longer seems to work due to loader changes.
So this patch does not reenable the test. The test probably has to
be rewritten.

  • dom/ContainerNode.cpp: (WebCore::ContainerNode::removedFromDocument): Re-added code to set the CSS target of the document to 0.
6:46 PM Changeset in webkit [45556] by Simon Fraser
  • 6 edits
    3 adds in trunk

2009-07-05 Chris Marrin <cmarrin@apple.com>

Reviewed by Simon Fraser.

https://bugs.webkit.org/show_bug.cgi?id=26943

When one transition finishes slightly before another the longer
one will fire a second time. This is because the second
ImplicitAnmation object is culled too early, before its final
RenderStyle is in place. This is done by cleanupFinishedAnimations()
so I got rid of that method completely and now cleanup each
transition or animation at the point where I am setting the final
style, or when I detect that the transition or animation has been
terminated early (which happens when you remove it from the style).

Test: transitions/extra-transition.html

  • page/animation/AnimationController.cpp: (WebCore::AnimationControllerPrivate::updateStyleIfNeededDispatcherFired):
  • page/animation/CompositeAnimation.cpp: (WebCore::CompositeAnimation::updateTransitions): (WebCore::CompositeAnimation::updateKeyframeAnimations): (WebCore::CompositeAnimation::animate):
  • page/animation/CompositeAnimation.h:
  • page/animation/AnimationBase.cpp: (WebCore::AnimationBase::getTimeToNextEvent): Avoid a divide by zero if m_animation->duration() is zero, which can happen if the duration is changed to zero while the animation is running.
12:39 PM Changeset in webkit [45555] by Simon Fraser
  • 6 edits
    3 deletes in trunk

2009-07-05 Simon Fraser <Simon Fraser>

Revert the previous commit because it broke the
animations/transition-and-animation-1.html
testcase.

  • page/animation/AnimationBase.cpp: (WebCore::AnimationBase::getTimeToNextEvent):
  • page/animation/AnimationController.cpp: (WebCore::AnimationControllerPrivate::updateStyleIfNeededDispatcherFired):
  • page/animation/CompositeAnimation.cpp: (WebCore::CompositeAnimation::updateTransitions): (WebCore::CompositeAnimation::updateKeyframeAnimations): (WebCore::CompositeAnimation::animate): (WebCore::CompositeAnimation::cleanupFinishedAnimations):
  • page/animation/CompositeAnimation.h:
12:18 PM Changeset in webkit [45554] by Simon Fraser
  • 6 edits
    4 adds in trunk

2009-07-05 Chris Marrin <cmarrin@apple.com>

Reviewed by Simon Fraser.

https://bugs.webkit.org/show_bug.cgi?id=26943

When one transition finishes slightly before another the longer
one will fire a second time. This is because the second
ImplicitAnmation object is culled too early, before its final
RenderStyle is in place. This is done by cleanupFinishedAnimations()
so I got rid of that method completely and now cleanup each
transition or animation at the point where I am setting the final
style, or when I detect that the transition or animation has been
terminated early (which happens when you remove it from the style).

Test: transitions/extra-transition.html

  • page/animation/AnimationController.cpp: (WebCore::AnimationControllerPrivate::updateStyleIfNeededDispatcherFired):
  • page/animation/CompositeAnimation.cpp: (WebCore::CompositeAnimation::updateTransitions): (WebCore::CompositeAnimation::updateKeyframeAnimations): (WebCore::CompositeAnimation::animate):
  • page/animation/CompositeAnimation.h:
  • page/animation/AnimationBase.cpp: (WebCore::AnimationBase::getTimeToNextEvent): Avoid a divide by zero if m_animation->duration() is zero, which can happen if the duration is changed to zero while the animation is running.
8:45 AM Changeset in webkit [45553] by Simon Hausmann
  • 6 edits in trunk

JavaScriptCore:

2009-07-05 Lars Knoll <lars.knoll@nokia.com>

Reviewed by Maciej Stachowiak.

https://bugs.webkit.org/show_bug.cgi?id=26843

Fix run-time crashes in JavaScriptCore with the Metrowerks compiler on Symbian.

The Metrowerks compiler on the Symbian platform moves the globally
defined Hashtables into read-only memory, despite one of the members
being mutable. This causes crashes at run-time due to write access to
read-only memory.

Avoid the use of const with this compiler by introducing the
JSC_CONST_HASHTABLE macro.

Based on idea by Norbert Leser.

  • runtime/Lookup.h: Define JSC_CONST_HASHTABLE as const for !WINSCW.
  • create_hash_table: Use JSC_CONST_HASHTABLE for hashtables.
  • runtime/JSGlobalData.cpp: Import various global hashtables via the macro.

WebCore:

2009-07-05 Lars Knoll <lars.knoll@nokia.com>

Reviewed by Maciej Stachowiak.

https://bugs.webkit.org/show_bug.cgi?id=26843

Fix run-time crashes in JavaScriptCore with the Metrowerks compiler on Symbian.

The Metrowerks compiler on the Symbian platform moves the globally
defined Hashtables into read-only memory, despite one of the members
being mutable. This causes crashes at run-time due to write access to
read-only memory.

Avoid the use of const with this compiler by introducing the
JSC_CONST_HASHTABLE macro.

Based on idea by Norbert Leser.

  • bindings/scripts/CodeGeneratorJS.pm: Use JSC_CONST_HASHTABLE for hash tables define in the bindings.
7:53 AM Changeset in webkit [45552] by zecke@webkit.org
  • 7 edits in trunk/WebKit/gtk

2009-07-05 Holger Hans Peter Freyther <zecke@selfish.org>

Reviewed by Xan Lopez.

Always initialize WebCore.

It is possible that WebKitWebSettings and other classes get
constructed before the WebKitWebView. In this case WebCore is
not yet properly initialized. Add webkit_init to every class
that can be constructed by API users.

  • webkit/webkitdownload.cpp: (webkit_download_class_init):
  • webkit/webkitnetworkrequest.cpp: (webkit_network_request_class_init):
  • webkit/webkitwebbackforwardlist.cpp: (webkit_web_back_forward_list_class_init):
  • webkit/webkitwebhistoryitem.cpp: (webkit_web_history_item_class_init):
  • webkit/webkitwebsettings.cpp: (webkit_web_settings_class_init):
  • webkit/webkitwebwindowfeatures.cpp: (webkit_web_window_features_class_init):
7:33 AM Changeset in webkit [45551] by Simon Hausmann
  • 3 edits in trunk/WebKit/qt

2009-07-05 Simon Hausmann <Simon Hausmann>

Reviewed by Holger Freyther.

Fix two qdoc warnings.

Added missing \property for QWebFrame::hasFocus and added \a
tag for pos of QWebPage::frameAt.

  • Api/qwebframe.cpp:
  • Api/qwebpage.cpp:
6:06 AM Changeset in webkit [45550] by rwlbuis@webkit.org
  • 3 edits
    4 adds in trunk

Reviewed by Maciej Stachowiak.

Bug 26897 - Dynamic SVG images do not display correctly
https://bugs.webkit.org/show_bug.cgi?id=26897

Use repaint() to fix a rendering problem with a SVG image embedded in xhtml.

Test: svg/custom/createImageElement2.xhtml

  • rendering/RenderSVGImage.cpp: (WebCore::RenderSVGImage::imageChanged):
3:43 AM Changeset in webkit [45549] by xan@webkit.org
  • 2 edits in trunk/WebCore

2009-07-05 Xan Lopez <xlopez@igalia.com>

Reviewed by Jan Alonzo.

https://bugs.webkit.org/show_bug.cgi?id=26960
[Gtk] caret offset not updated when selecting text

Report the caret offset from the end of the selection so it works
correctly for multi-char selections too (ie, anything that is not
the zero width caret).

  • accessibility/gtk/AccessibilityObjectWrapperAtk.cpp: (webkit_accessible_text_get_caret_offset):
12:28 AM Changeset in webkit [45548] by zecke@webkit.org
  • 2 edits in trunk/WebCore

Fix Qt link error by adding two new localized strings

In r45474 two new strings got added to LocalizedStrings.h add
the definition of them to Qt.

12:27 AM Changeset in webkit [45547] by zecke@webkit.org
  • 2 edits in trunk/WebKit/qt

2009-07-04 Holger Hans Peter Freyther <zecke@selfish.org>

Reviewed by Simon Hausmann.

Use the recently introduced FocusController::setFocused

Use the recently introduced FocusController::setFocused
in the Qt platform. The SelectionController will be updated
from within the FocusController now.

  • Api/qwebpage.cpp: (QWebPagePrivate::focusInEvent): (QWebPagePrivate::focusOutEvent):

Jul 4, 2009:

10:53 AM Changeset in webkit [45546] by mitz@apple.com
  • 2 edits in trunk/JavaScriptCore
  • debug build fix
  • runtime/RegExpConstructor.cpp: (JSC::RegExpConstructor::getLastParen):
7:21 AM Changeset in webkit [45545] by mjs@apple.com
  • 5 edits in trunk/JavaScriptCore

2009-07-03 Yong Li <yong.li@torchmobile.com>

Reviewed by Maciej Stachowiak (and revised slightly)

RegExp::match to be optimized
https://bugs.webkit.org/show_bug.cgi?id=26957

Allow regexp matching to use Vectors with inline capacity instead of
allocating a new ovector buffer every time.


~5% speedup on SunSpider string-unpack-code test, 0.3% on SunSpider overall.

  • runtime/RegExp.cpp: (JSC::RegExp::match):
  • runtime/RegExp.h:
  • runtime/RegExpConstructor.cpp: (JSC::RegExpConstructorPrivate::RegExpConstructorPrivate): (JSC::RegExpConstructorPrivate::lastOvector): (JSC::RegExpConstructorPrivate::tempOvector): (JSC::RegExpConstructorPrivate::changeLastOvector): (JSC::RegExpConstructor::performMatch): (JSC::RegExpMatchesArray::RegExpMatchesArray): (JSC::RegExpMatchesArray::fillArrayInstance): (JSC::RegExpConstructor::getBackref): (JSC::RegExpConstructor::getLastParen): (JSC::RegExpConstructor::getLeftContext): (JSC::RegExpConstructor::getRightContext):
  • runtime/StringPrototype.cpp: (JSC::stringProtoFuncSplit):
7:20 AM Changeset in webkit [45544] by mjs@apple.com
  • 2 edits in trunk/WebKitTools

2009-07-03 Maciej Stachowiak <mjs@apple.com>

Reviewed by Eric Seidel.

bugzilla-tool forgot how to obsolete patches
https://bugs.webkit.org/show_bug.cgi?id=26951

  • Scripts/modules/bugzilla.py:

Jul 3, 2009:

9:50 PM Changeset in webkit [45543] by Darin Adler
  • 1 edit in trunk/WebCore/ChangeLog

Added Radar link I left out by accident the first time.

9:48 PM Changeset in webkit [45542] by Darin Adler
  • 2 edits in trunk/WebCore

2009-07-03 Darin Adler <Darin Adler>

Reviewed by Adele Peterson.

REGRESSION (r44670-r44680): Typing is suprisingly slow in password field on reddit.com
https://bugs.webkit.org/show_bug.cgi?id=26959

The code to handle iteration boundaries was malfunctioning when the boundary was
at the edge of a shadow tree. This happens all the time with <input> elements.

It's not immediately obvious how to make a regression test for this since the
symptom was a performance problem, not incorrect behavior. I'll add a test if I
figure out a way to make one.

  • editing/TextIterator.cpp: (WebCore::parentCrossingShadowBoundaries): Renamed from parentOrShadowParent. (WebCore::depthCrossingShadowBoundaries): Updated for name change. (WebCore::nextInPreOrderCrossingShadowBoundaries): Added. (WebCore::previousInPostOrderCrossingShadowBoundaries): Added. (WebCore::setUpFullyClippedStack): Updated for name change. (WebCore::TextIterator::TextIterator): Updated for name change. Use nextInPreOrderCrossingShadowBoundaries instead of Range::pastLastNode to fix the bug. (WebCore::TextIterator::advance): Updated for name change. (WebCore::SimplifiedBackwardsTextIterator::SimplifiedBackwardsTextIterator): Tweaked formatting. (WebCore::SimplifiedBackwardsTextIterator::advance): Ditto. Changed code that initializes m_pastStartNode to use previousInPostOrderCrossingShadowBoundaries. (WebCore::SimplifiedBackwardsTextIterator::handleNonTextNode): Tweaked formatting. (WebCore::SimplifiedBackwardsTextIterator::exitNode): Ditto.
8:56 PM Changeset in webkit [45541] by mitz@apple.com
  • 2 edits in trunk/WebKit/mac

Reviewed by Simon Fraser.

  • fix <rdar://problem/6964278> REGRESSION (r42118): Scrolling redraw problem in FileMaker Pro
  • WebView/WebHTMLView.mm: (-[WebHTMLView _recursiveDisplayRectIfNeededIgnoringOpacity:isVisibleRect:rectIsVisibleRectForView:topView:]): Perform layout if needed, even on Mac OS X versions that have -viewWillDraw. This prevents attempts to draw without layout in case -viewWillDraw was not called due to NSView issues or the client did something during the display operation that re-invalidated the layout.
8:27 PM Changeset in webkit [45540] by ddkilzer@apple.com
  • 1 edit
    1 add in trunk/BugsSite

Bug 26958: Change edit link to review link in request messages

<https://bugs.webkit.org/show_bug.cgi?id=26958>

Reviewed by Dan Bernstein.

  • template/en/custom/request/email.txt.tmpl: Copied from BugsSite/template/en/default/request/email.txt.tmpl. Changed link from "action=edit" to "action=review".
6:09 PM Changeset in webkit [45539] by Nikolas Zimmermann
  • 2 edits in trunk/LayoutTests

Not reviewed. Forgot to land ontimer-inline-event.html results within my last WML commit.

5:19 PM Changeset in webkit [45538] by ddkilzer@apple.com
  • 2 edits in trunk/BugsSite

Bug 26950: Make the summary and alias fields support click-to-edit
<https://bugs.webkit.org/show_bug.cgi?id=26950>

Reviewed by Maciej Stachowiak.

Original patch by Maciej Stachowiak.

  • js/field.js: (hideEditableField): Updated to add click event listeners to the alias and short description elements to make them easier to edit. Renamed field_id parameter to field2_id (short description id) and added a field1_id parameter (alias id). (showEditableField): If a third parameter is passed in the ContainerInputArray parameter, use it to find the element to focus. Otherwise, fall back to the original behavior of focusing the first input field. (hideAliasAndSummary): Changed to pass the id for the alias element to hideEditableField().
1:54 PM Changeset in webkit [45537] by Simon Fraser
  • 2 edits in trunk/WebCore

2009-07-03 Simon Fraser <Simon Fraser>

Reviewed by Sam Weinig.

Background audio stops playing when JS GC runs
https://bugs.webkit.org/show_bug.cgi?id=26956

Ensure we keep the Audio object around while it is playing, so that it
doesn't get collected.

Not testable because there's no way to know whether the audio keeps
playing without holding a reference to it.

  • bindings/js/JSDOMBinding.cpp: (WebCore::isObservableThroughDOM):
12:34 PM Changeset in webkit [45536] by abarth@webkit.org
  • 2 edits in trunk/WebKitSite

2009-07-03 Adam Barth <abarth@webkit.org>

Reviewed by Sam Weinig.

Adam and Collin's affiliation is incorrect
https://bugs.webkit.org/show_bug.cgi?id=26954

This change corrects their affiliation.

  • security/security-group-members.html:
7:08 AM Changeset in webkit [45535] by xan@webkit.org
  • 2 edits in trunk/WebCore

2009-07-02 Xan Lopez <xlopez@igalia.com>

Reviewed by Jan Alonzo.

https://bugs.webkit.org/show_bug.cgi?id=26814
[Gtk] Caret-moved events are not issued for the correct offset
when text is selected forward

Report the caret offset from the end of the selection, otherwise
we'll report the same offset when moving the caret while
modififying the selection (eg, with Shift + Right).

  • editing/gtk/SelectionControllerGtk.cpp: (WebCore::SelectionController::notifyAccessibilityForSelectionChange):
7:08 AM Changeset in webkit [45534] by xan@webkit.org
  • 2 edits in trunk/WebCore

2009-07-02 Xan Lopez <xlopez@igalia.com>

Reviewed by Jan Alonzo.

https://bugs.webkit.org/show_bug.cgi?id=26815
[Gtk] text-selection-changed events are not issued for the correct
object when the selection spans multiple objects

Get the focused node from the end of the selection, not the start,
so we can detect when we cross object boundaries.

  • editing/gtk/SelectionControllerGtk.cpp: (WebCore::SelectionController::notifyAccessibilityForSelectionChange):
3:47 AM Changeset in webkit [45533] by dimich@chromium.org
  • 2 edits in trunk/LayoutTests

2009-07-03 Dmitry Titov <dimich@chromium.org>

Reviewed by Darin Fisher.

Increase delay in worker-terminate layout test to make it more reliable.
https://bugs.webkit.org/show_bug.cgi?id=26917

This test used time delay to see of terminated worker does not return
messages (terminated). In some test conditions, the current delay (500ms)
is too short and test fails. Move setting the delay to the end of initialization
and increase it 2x to make test more reliable.

  • fast/workers/worker-terminate.html:
3:24 AM Changeset in webkit [45532] by jmalonzo@webkit.org
  • 2 edits in trunk/LayoutTests

2009-07-03 Jan Michael Alonzo <jmalonzo@webkit.org>

Skip tests that require expected results that Gtk doesn't have.

  • platform/gtk/Skipped:
2:57 AM Changeset in webkit [45531] by jmalonzo@webkit.org
  • 11 edits
    1 add in trunk

2009-07-03 Jan Michael Alonzo <jmalonzo@webkit.org>

Reviewed by Xan Lopez and Gustavo Noronha.

Set user-agent from application
https://bugs.webkit.org/show_bug.cgi?id=17375

Add 'user-agent' setter/getter API to WebSettings.

  • WebCoreSupport/FrameLoaderClientGtk.cpp: (WebKit::FrameLoaderClient::FrameLoaderClient): (WebKit::FrameLoaderClient::userAgent):
  • WebCoreSupport/FrameLoaderClientGtk.h:
  • webkit/webkitprivate.h:
  • webkit/webkitversion.h.in:
  • webkit/webkitwebsettings.cpp: (webkit_get_user_agent): (webkit_web_settings_class_init): (webkit_web_settings_finalize): (webkit_web_settings_set_property): (webkit_web_settings_get_property): (webkit_web_settings_copy): (webkit_web_settings_get_user_agent):
  • webkit/webkitwebsettings.h:
  • webkit/webkitwebview.cpp: (webkit_web_view_dispose):

Define UA version macros to be used by the UA string.
Add new WebSettings unit test for the User-Agent string API.

  • GNUmakefile.am:
  • configure.ac:
Note: See TracTimeline for information about the timeline view.